nephele
/
haikuwebkit
1
0
Fork 0
Code Issues Releases Activity
haikuwebkit/Source/WebKit/ChangeLog-2021-03-18

73541 lines
3.4 MiB
Raw Permalink Blame History

This file contains invisible Unicode characters

This file contains invisible Unicode characters that are indistinguishable to humans but may be processed differently by a computer. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

2021-03-18 Aditya Keerthi <akeerthi@apple.com>
[iOS] Update strings and icons in the file upload panel
https://bugs.webkit.org/show_bug.cgi?id=223463
<rdar://problem/74999459>
Reviewed by Tim Horton.
* UIProcess/ios/forms/WKFileUploadPanel.mm:
(-[WKFileUploadPanel currentAvailableActionTitles]):
Expose the new string for testing.
(-[WKFileUploadPanel _chooseFilesButtonLabel]):
(-[WKFileUploadPanel contextMenuInteraction:configurationForMenuAtLocation:]):
Use the "folder" SF symbol as the icon for the "Choose File" action.
2021-03-18 BJ Burg <bburg@apple.com>
[Cocoa] Web Inspector: load inspector resources using a custom scheme handler
https://bugs.webkit.org/show_bug.cgi?id=179904
<rdar://problem/10887211>
Reviewed by Geoff Garen.
Load WebInspectorUI resources via the inspector-resource:// scheme.
Moving off of file:// URLs will allow for a tighter sandbox and
will fix the longstanding issue that Web Inspector's settings are
part of the file:// scheme and cleared when Safari clears website data.
This also has the effect of resetting Web Inspector settings since
they are no longer associated with file:///.
Loading via this scheme is exercised by existing inspector tests.
* SourcesCocoa.txt: Added new file.
* WebKit.xcodeproj/project.pbxproj:
* UIProcess/Inspector/mac/WKInspectorResourceURLSchemeHandler.h: Copied from Source/WebKit/WebProcess/Inspector/mac/WebInspectorUIMac.mm.
* UIProcess/Inspector/mac/WKInspectorResourceURLSchemeHandler.mm: Added.
(-[WKInspectorResourceURLSchemeHandler webView:startURLSchemeTask:]):
Use a global concurrent queue and NSOperation / NSOperationQueue to support
reading files off the main queue.
(-[WKInspectorResourceURLSchemeHandler webView:stopURLSchemeTask:]):
Translate the request URL to a bundle resource URL, read the data, and send a response
with the file data. Return an error if the file could not be found or read.
* UIProcess/Inspector/mac/WKInspectorViewController.h:
* UIProcess/Inspector/mac/WKInspectorViewController.mm:
(-[WKInspectorViewController webViewConfiguration]):
(+[WKInspectorViewController URLForInspectorResource:]):
(-[WKInspectorViewController webView:decidePolicyForNavigationAction:decisionHandler:]):
Allow navigations to URIs with the inspector-resource:// scheme.
* UIProcess/Inspector/mac/WebInspectorProxyMac.mm:
(WebKit::WebInspectorProxy::inspectorPageURL):
(WebKit::WebInspectorProxy::inspectorTestPageURL):
(WebKit::WebInspectorProxy::inspectorBaseURL):
Updated.
* WebProcess/Inspector/mac/WebInspectorUIMac.mm:
(WebKit::webInspectorUILocalizedStringsURL):
Updated.
2021-03-18 Sihui Liu <sihui_liu@appe.com>
Handle empty bound size during animated resizing
https://bugs.webkit.org/show_bug.cgi?id=223446
Reviewed by Tim Horton.
We don't allow empty bounds in animated resizing: when that happens, we would cancel it. However, a set of
updates involving empty bounds can be merged to a valid update. For example, two changes of frame size
[1] (0, 0, 100, 100) => (0, 0, 0, 50) [2] (0, 0, 0, 50) => (0, 0, 50, 50) can be viewed as single change
(0, 0, 100, 100) => (0, 0, 50, 50), which is valid. We used to ask SPI clients to ensure each update is valid,
but sometimes the frame change is not under client's control. For example when client uses auto layout, empty
frame seems to be a valid mid product.
To handle this case, we can store the old bounds, and actually perform animated resizing on when new valid
bounds show up. For the example above, we would store (0, 0, 100, 100) on [1] and use it to resize from
(0, 0, 100, 100) to (0, 0, 50, 50) on [2].
* UIProcess/API/Cocoa/WKWebViewInternal.h:
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _processWillSwapOrDidExit]):
(-[WKWebView _updateVisibleContentRects]): If we still have stored old bounds in _updateVisibleContentRects, it
means they are not consumed in this cycle. In this case let's just cancel animated resizing and send out updates.
(-[WKWebView _cancelAnimatedResize]):
(-[WKWebView _didCompleteAnimatedResize]):
(-[WKWebView _beginAnimatedResizeWithUpdates:]):
2021-03-18 Jiewen Tan <jiewen_tan@apple.com>
Update Keychain queries according to internal needs
https://bugs.webkit.org/show_bug.cgi?id=223424
rdar://75434045
Reviewed by Brent Fulgham.
Covered by manual tests.
* UIProcess/WebAuthentication/Cocoa/LocalAuthenticator.mm:
(updateQuery):
(WebKit::LocalAuthenticatorInternal::getExistingCredentials):
(WebKit::LocalAuthenticator::clearAllCredentials):
(WebKit::LocalAuthenticator::continueMakeCredentialAfterUserVerification):
(WebKit::LocalAuthenticator::continueGetAssertionAfterUserVerification):
(WebKit::LocalAuthenticator::receiveException const):
(WebKit::LocalAuthenticator::deleteDuplicateCredential const):
(shouldUseAlternateQuery): Deleted.
(alternateQueryForRPID): Deleted.
2021-03-18 Chris Dumez <cdumez@apple.com>
Crash under WebsiteDataStore::terminateNetworkProcess()
https://bugs.webkit.org/show_bug.cgi?id=223482
<rdar://75326967>
Reviewed by Alex Christensen.
Calling WebProcessPool::terminateServiceWorkers() may destroy the WebProcessPool, which would
remove the WebProcessPool from the WebProcessPool::allProcessPools() vector we're iterating
on. To make this safe, I now copy the WebProcessPool into a separate vector and hold a RefPtr
to them. I now iterate over this new vector instead.
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::terminateNetworkProcess):
2021-03-18 Michael Catanzaro <mcatanzaro@gnome.org>
[WPE][GTK] Remove plugin process leftover arg from ProcessLauncher
https://bugs.webkit.org/show_bug.cgi?id=223460
Reviewed by Adrian Perez de Castro.
Remove extra nullptr arg.
* UIProcess/Launcher/glib/ProcessLauncherGLib.cpp:
(WebKit::ProcessLauncher::launchProcess):
2021-03-18 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] editing/selection/character-granularity-rect.html fails on certain internal configurations after r274610
https://bugs.webkit.org/show_bug.cgi?id=223447
Reviewed by Devin Rousso.
On versions of iOS where `ENABLE(IMAGE_EXTRACTION)` is defined but the system feature is disabled, this test
began failing after r274610. This is because the new image extraction deferral gesture now blocks text
interaction gestures, but the code to actually fail this deferring gesture (and thus, "lift" the deferring
gesture gate) is still behind the system flag. As a result, text interactions when the image extraction flag is
disabled will only begin after the touch ends.
This means that when selecting text via a long press, the text selection doesn't appear until lifting the touch
if image extraction is disabled. To fix this, we only enable this deferring gesture in the case where the system
feature is enabled.
* UIProcess/ios/GestureRecognizerConsistencyEnforcer.mm:
(WebKit::GestureRecognizerConsistencyEnforcer::timerFired):
Additionally tweak the `GestureRecognizerConsistencyEnforcer` to only forcibly reset lingering unlifted gesture
recognizers if they are enabled.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView setUpInteraction]):
2021-03-18 Antti Koivisto <antti@apple.com>
Don't delay conditional requests in resource load scheduler
https://bugs.webkit.org/show_bug.cgi?id=223450
Reviewed by Geoffrey Garen.
* NetworkProcess/NetworkLoadScheduler.cpp:
(WebKit::NetworkLoadScheduler::HostContext::schedule):
Conditional requests usually come back as 304s and don't slow other resource loads down.
2021-03-18 Carlos Garcia Campos <cgarcia@igalia.com>
[WPE] Bump API version when building with libsoup3
https://bugs.webkit.org/show_bug.cgi?id=223437
Reviewed by Adrian Perez de Castro.
* PlatformWPE.cmake: Use WPE_API_DOC_VERSION instead of WPE_API_VERSION for gtkdoc config files.
* wpe/wpe-web-extension-uninstalled.pc.in: Add variable for libsoup version.
* wpe/wpe-web-extension.pc.in: Ditto.
* wpe/wpe-webkit-uninstalled.pc.in: Ditto.
* wpe/wpe-webkit.pc.in: Ditto.
2021-03-18 Youenn Fablet <youenn@apple.com>
Geolocation API does not work in WKWebView with custom URL scheme handler
https://bugs.webkit.org/show_bug.cgi?id=222861
<rdar://problem/75394695>
Reviewed by Eric Carlson.
Update geolocation prompt origin to match getUserMedia and show the app name instead of custom scheme origin.
* UIProcess/Cocoa/MediaPermissionUtilities.mm:
(WebKit::alertMessageText):
2021-03-18 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] Web process sometimes crashes under WebPage::positionInformation() after r274031
https://bugs.webkit.org/show_bug.cgi?id=223417
<rdar://problem/75540178>
Reviewed by Devin Rousso.
Add a null check to guard against the case where the hit-tested node is null, but the element responding to
click events exists.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::elementPositionInformation):
2021-03-17 Peng Liu <peng.liu6@apple.com>
[GPUP] WebGL2RenderingContext::texImage2D() does not work with a video as the image source
https://bugs.webkit.org/show_bug.cgi?id=223294
Reviewed by Eric Carlson.
No new tests. Fix a test failure.
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::updateCachedState):
Add a missing piece to implement `MediaPlayerPrivateRemote::videoPlaybackQualityMetrics()`.
2021-03-17 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] Refactor some gesture deferral logic to account for image extraction
https://bugs.webkit.org/show_bug.cgi?id=223400
<rdar://problem/75502285>
Reviewed by Tim Horton.
Refactor some logic around deferring gestures. See below for more details.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView setUpInteraction]):
Rename a few existing deferring gestures, to better reflect that the names are deferral reasons (i.e. touch
start, touch end, or image extraction), rather than the names of the things being deferred.
(-[WKContentView deferringGestures]):
(-[WKContentView _doneDeferringTouchStart:]):
(-[WKContentView _doneDeferringTouchEnd:]):
(-[WKContentView hasSelectablePositionAtPoint:]):
(-[WKContentView textInteractionGesture:shouldBeginAtPoint:]):
(-[WKContentView deferringGestureRecognizer:shouldDeferOtherGestureRecognizer:]):
(-[WKContentView _endImageExtractionGestureDeferral:]):
* UIProcess/ios/WKDeferringGestureRecognizer.h:
Add a property to indicate that the deferring gesture recognizer should immediately fail upon touchend.
* UIProcess/ios/WKDeferringGestureRecognizer.mm:
(-[WKDeferringGestureRecognizer touchesEnded:withEvent:]):
(-[WKDeferringGestureRecognizer endDeferral:]):
(-[WKDeferringGestureRecognizer setDefaultPrevented:]): Deleted.
Rename `-setDefaultPrevented:` to `-endDeferral:` to decouple this terminology from `Event.preventDefault()` in
bindings (which is important, now that we have a new deferrer that is not associated with touch events).
Additionally, change this take an `enum class` instead of a `BOOL` to help clarify the intent of the given flag.
* UIProcess/ios/WKImageExtractionGestureRecognizer.h:
* UIProcess/ios/WKImageExtractionGestureRecognizer.mm:
(-[WKImageExtractionGestureRecognizer setState:]):
Add support for a new image extraction gesture delegate method which informs the client that the gesture has
transitioned to Failed state.
2021-03-17 Peng Liu <peng.liu6@apple.com>
[GPUP] Some media-source tests fail/timeout when media in GPU Process is enabled
https://bugs.webkit.org/show_bug.cgi?id=221686
Reviewed by Jer Noble.
For a `MediaPlayerPrivateRemote`, if it is owned by a video element with a reference
to a `MediaSource`, it needs to use the duration information from the `MediaSource`.
Covered by existing tests.
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::durationMediaTime const):
* WebProcess/GPU/media/MediaPlayerPrivateRemote.h:
* WebProcess/GPU/media/MediaSourcePrivateRemote.h:
2021-03-17 Aditya Keerthi <akeerthi@apple.com>
[iOS] Refactor WKFileUploadPanel to use UniformTypeIdentifiers
https://bugs.webkit.org/show_bug.cgi?id=223000
<rdar://problem/75237774>
Reviewed by Tim Horton.
Uniform type identifier APIs in MobileCoreServices were deprecated in
iOS 14. Instead, use the new APIs in the UniformTypeIdentifiers
framework.
This patch also cleans up instances of code duplication when comparing
uniform type idenfifiers and increases the test coverage for
WKFileUploadPanel.
Test: fast/forms/ios/file-upload-panel-accept.html
* Configurations/WebKit.xcconfig:
Link UniformTypeIdentifiers.framework on watchOS and tvOS.
* UIProcess/API/ios/WKWebViewPrivateForTestingIOS.h:
Expose _filePickerAcceptedTypeIdentifiers to enable testing of accepted
file types.
* UIProcess/API/ios/WKWebViewTestingIOS.mm:
(-[WKWebView _filePickerAcceptedTypeIdentifiers]):
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView filePickerAcceptedTypeIdentifiers]):
* UIProcess/ios/forms/WKFileUploadPanel.h:
* UIProcess/ios/forms/WKFileUploadPanel.mm:
(setContainsUTIThatConformsTo):
Modified this helper method to take in an NSSet rather than an NSArray
to avoid unnecessary conversion. Replaced MobileCoreServices API usage
with UniformTypeIdentifiers API.
(-[WKFileUploadPanel presentWithParameters:resultListener:]):
Determine the accepted UTIs and image picker configuration once before
presentation and store them in member variables to reduce code
duplication.
(-[WKFileUploadPanel currentAvailableActionTitles]):
(-[WKFileUploadPanel acceptedTypeIdentifiers]):
Expose the accepted type identifers as a sorted array for testing.
(-[WKFileUploadPanel _mediaTypesForPickerSourceType:]):
(-[WKFileUploadPanel _cameraButtonLabel]):
(-[WKFileUploadPanel contextMenuInteraction:configurationForMenuAtLocation:]):
(-[WKFileUploadPanel showFilePickerMenu]):
(-[WKFileUploadPanel showDocumentPickerMenu]):
(-[WKFileUploadPanel _uploadItemFromMediaInfo:successBlock:failureBlock:]):
2021-03-17 Brent Fulgham <bfulgham@apple.com>
[Cocoa] Populate NSURLSession with attributed bundle ID
https://bugs.webkit.org/show_bug.cgi?id=223382
<rdar://problem/75498230>
Reviewed by Alex Christensen.
Update NSURLSession property with attributed bundle ID when supported by the platform.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::NetworkSessionCocoa):
(WebKit::NetworkSessionCocoa::initializeEphemeralStatelessSession):
2021-03-17 Alex Christensen <achristensen@webkit.org>
Allow UniqueRef to be the value of a HashMap
https://bugs.webkit.org/show_bug.cgi?id=223240
Reviewed by Youenn Fablet.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::populateFromMemoryStore):
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:
(WebKit::pruneResources):
(WebKit::ResourceLoadStatisticsMemoryStore::aggregatedThirdPartyData const):
(WebKit::ResourceLoadStatisticsMemoryStore::classifyPrevalentResources):
(WebKit::ResourceLoadStatisticsMemoryStore::ensureResourceStatisticsForRegistrableDomain):
(WebKit::ResourceLoadStatisticsMemoryStore::mergeStatistics):
(WebKit::ResourceLoadStatisticsMemoryStore::updateCookieBlocking):
(WebKit::ResourceLoadStatisticsMemoryStore::processStatistics const):
(WebKit::ResourceLoadStatisticsMemoryStore::registrableDomainsToDeleteOrRestrictWebsiteDataFor):
(WebKit::ResourceLoadStatisticsMemoryStore::pruneStatisticsIfNeeded):
(WebKit::ResourceLoadStatisticsMemoryStore::insertExpiredStatisticForTesting):
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.h:
* Shared/Authentication/AuthenticationManager.cpp:
(WebKit::AuthenticationManager::addChallengeToChallengeMap):
(WebKit::AuthenticationManager::didReceiveAuthenticationChallenge):
* Shared/Authentication/AuthenticationManager.h:
2021-03-17 Chris Dumez <cdumez@apple.com>
Maybe-regression(STP121): window.open flakily returning null
https://bugs.webkit.org/show_bug.cgi?id=222590
<rdar://problem/75211786>
Reviewed by Geoffrey Garen.
This was an IPC ordering bug. WebPageProxy::DidCommitLoadForFrame is async and WebPageProxy::CreateNewPage is
sync. As a result, it was possible for the WebPageProxy::CreateNewPage to get processed *BEFORE* the
WebPageProxy::DidCommitLoadForFrame IPC. This was causing trouble because Safari rejects the popup opening if
the main frame is doing a provisional load.
To address the issue, introduce a new IPC::SendSyncOption::MaintainOrderingWithAsyncMessages flag and
use it on WebPageProxy::CreateNewPage sync IPC so that it gets processed in order with surrounding async
messages.
* Platform/IPC/Connection.cpp:
(IPC::Connection::SyncMessageState::processIncomingMessage):
(IPC::Connection::sendSyncMessage):
* Platform/IPC/Connection.h:
* Platform/IPC/Decoder.cpp:
(IPC::Decoder::shouldMaintainOrderingWithAsyncMessages const):
* Platform/IPC/Decoder.h:
* Platform/IPC/Encoder.cpp:
(IPC::Encoder::setShouldMaintainOrderingWithAsyncMessages):
* Platform/IPC/Encoder.h:
* Platform/IPC/MessageFlags.h:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::createWindow):
2021-03-17 Peng Liu <peng.liu6@apple.com>
[GPUP] Some media-source tests fail/timeout when media in GPU Process is enabled
https://bugs.webkit.org/show_bug.cgi?id=221686
Reviewed by Eric Carlson.
With "GPU Process: Media" enabled, a video element with a reference to a `MediaSource`
object will schedule a "resize" event after the source buffer scheduling an "update" event.
The root cause is that the `MediaPlayerPrivateRemote` does not notify its `HTMLMediaElement`
to change its ready state to HAVE_METADATA after the source buffer receives the initialization
segment (in `SourceBuffer::sourceBufferPrivateDidReceiveInitializationSegment()`).
The requirement is defined in section 3.5.7.6.2 of the specification:
[Initialization Segment Received](https://www.w3.org/TR/media-source/#sourcebuffer-init-segment-received).
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::setReadyState):
* WebProcess/GPU/media/MediaPlayerPrivateRemote.h:
* WebProcess/GPU/media/SourceBufferPrivateRemote.cpp:
(WebKit::SourceBufferPrivateRemote::setReadyState):
2021-03-17 Brent Fulgham <bfulgham@apple.com>
[Cocoa] Move loadSimulatedRequest API from Private to Public WKWebView
https://bugs.webkit.org/show_bug.cgi?id=223279
<rdar://problem/75494316>
Reviewed by Darin Adler.
Just moving from SPI headers to API headers (and implementation).
No new tests, since this just changes the visibility of the functionality. The features are
tested with existing TestWebKitAPI tests.
* UIProcess/API/Cocoa/WKWebView.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView loadSimulatedRequest:withResponse:responseData:]):
(-[WKWebView loadSimulatedRequest:withResponseHTMLString:]):
(-[WKWebView loadFileRequest:allowingReadAccessToURL:]):
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
2021-03-17 Youenn Fablet <youenn@apple.com>
Align device orientation delegate to getUserMedia/geolocation delegates
https://bugs.webkit.org/show_bug.cgi?id=223247
Reviewed by Eric Carlson.
Refactor code to move device orientation prompt implementation with other prompts in MediaPermissionUtilities.
Update code to use top level origin for prompt.
Update delegate to provide both top level origin and initiating iframe as parameters.
Update the decision handler to take grant, deny or prompt parameter.
Covered by API tests.
* UIProcess/API/C/WKFrameInfoRef.cpp:
(WKFrameInfoCopySecurityOrigin):
* UIProcess/API/C/WKFrameInfoRef.h:
* UIProcess/API/C/WKPage.cpp:
(WKPageSetPageUIClient):
* UIProcess/API/C/WKPageUIClient.h:
* UIProcess/API/Cocoa/WKUIDelegatePrivate.h:
* UIProcess/Cocoa/MediaPermissionUtilities.mm:
(WebKit::alertMessageText):
(WebKit::allowButtonText):
(WebKit::doNotAllowButtonText):
(WebKit::alertForPermission):
* UIProcess/Cocoa/UIDelegate.h:
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::setDelegate):
(WebKit::UIDelegate::UIClient::shouldAllowDeviceOrientationAndMotionAccess):
* UIProcess/Cocoa/WKOrientationAccessAlert.h: Removed.
* UIProcess/Cocoa/WKOrientationAccessAlert.mm: Removed.
* UIProcess/MediaPermissionUtilities.h:
* UIProcess/WebsiteData/WebDeviceOrientationAndMotionAccessController.cpp:
(WebKit::WebDeviceOrientationAndMotionAccessController::shouldAllowAccess):
* WebKit.xcodeproj/project.pbxproj:
2021-03-17 Kimmo Kinnunen <kkinnunen@apple.com>
Shutdown ANGLE on GPU process when last WebGL context is destroyed
https://bugs.webkit.org/show_bug.cgi?id=223082
Reviewed by Kenneth Russell.
Schedule a check for releasing the ANGLE default display when global
count of remote graphics contexts reach zero. This should decrease the
memory use of sessions where WebGL is not always on.
Dispatch the check 0.2s after hitting zero, so that the optimization still
affects page navigations but maybe does not redundantly deinitialize /
reinitialize ANGLE in the cases where the context #1 is created and destroyed
frequently.
* GPUProcess/graphics/RemoteGraphicsContextGL.cpp:
(WebKit::dispatchReleaseAllResourcesIfUnused):
(WebKit::RemoteGraphicsContextGL::initialize):
(WebKit::RemoteGraphicsContextGL::stopListeningForIPC):
2021-03-17 Youenn Fablet <youenn@apple.com>
Protect RemoteCaptureSampleManager and RemoteAudioSourceProviderManager when hopping to background threads
https://bugs.webkit.org/show_bug.cgi?id=223245
Reviewed by Alex Christensen.
When hopping to a background thread, take a ref to the manager.
Both managers are thread safe refcounted.
* WebProcess/GPU/media/RemoteAudioSourceProviderManager.cpp:
(WebKit::RemoteAudioSourceProviderManager::addProvider):
(WebKit::RemoteAudioSourceProviderManager::removeProvider):
* WebProcess/cocoa/RemoteCaptureSampleManager.cpp:
(WebKit::RemoteCaptureSampleManager::addSource):
(WebKit::RemoteCaptureSampleManager::removeSource):
2021-03-16 Devin Rousso <drousso@apple.com>
[macOS] change for the language/subtitle tracks button to use an `NSMenu` instead of web content
https://bugs.webkit.org/show_bug.cgi?id=223239
<rdar://problem/75462340>
Reviewed by Eric Carlson.
* Shared/ContextMenuContextData.h:
* Shared/ContextMenuContextData.cpp:
(WebKit::ContextMenuContextData::ContextMenuContextData):
Move `WebKit::ContextMenuContextData::Type` to `WebCore::ContextmenuContext::Type` so that
it can be used in WebCore as well.
* Shared/WebContextMenuItemData.h:
* Shared/WebContextMenuItemData.cpp:
(WebKit::WebContextMenuItemData::WebContextMenuItemData):
(WebKit::WebContextMenuItemData::core const):
(WebKit::WebContextMenuItemData::encode const):
(WebKit::WebContextMenuItemData::decode):
(WebKit::WebContextMenuItemData::indentationLevel const): Added.
Provide a way to eventually set the `-[NSMenuItem setIndentationLevel:]`.
* UIProcess/WebContextMenuProxy.h:
* UIProcess/mac/WebContextMenuProxyMac.h:
* UIProcess/mac/WebContextMenuProxyMac.mm:
(-[WKMenuDelegate initWithMenuProxy:]): Added.
(-[WKMenuDelegate menuWillOpen:]): Added.
(-[WKMenuDelegate menuDidClose:]): Added.
(WebKit::WebContextMenuProxyMac::getContextMenuItem):
(WebKit::WebContextMenuProxyMac::useContextMenuItems):
(WebKit::WebContextMenuProxyMac::platformMenu const): Added.
(WebKit::contentsOfContextMenuItem): Added.
(WebKit::WebContextMenuProxyMac::platformData const): Added.
Add support for `-[NSMenuItem setIndentationLevel:]`.
* UIProcess/API/Cocoa/WKWebViewPrivateForTesting.h:
* UIProcess/API/Cocoa/WKWebViewTesting.mm:
(-[WKWebView _contentsOfUserInterfaceItem:]):
(-[WKWebView _didShowContextMenu]): Added.
(-[WKWebView _didDismissContextMenu]): Added.
* UIProcess/API/ios/WKWebViewPrivateForTestingIOS.h:
* UIProcess/API/ios/WKWebViewTestingIOS.mm:
(-[WKWebView _didShowContextMenu]): Deleted.
(-[WKWebView _didDismissContextMenu]): Deleted.
Move these methods so they can be used on macOS too.
* UIProcess/PageClient.h:
(WebKit::PageClient::didShowContextMenu): Added.
(WebKit::PageClient::didDismissContextMenu): Added.
* UIProcess/mac/PageClientImplMac.h:
* UIProcess/mac/PageClientImplMac.mm:
(WebKit::PageClientImpl::didShowContextMenu): Added.
(WebKit::PageClientImpl::didDismissContextMenu): Added.
Add support for `didShowContextMenu`/`didDismissContextMenu` for tests.
* UIProcess/API/mac/WKWebViewTestingMac.mm:
(-[WKWebView _activeMenu]):
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::contentsOfUserInterfaceItem): Added.
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::platformActiveContextMenu const): Added.
Add support for `contentsOfUserInterfaceItem` for tests.
* UIProcess/WebPageProxy.messages.in:
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.cpp:
* UIProcess/ios/PageClientImplIOS.h:
* UIProcess/ios/PageClientImplIOS.mm:
* UIProcess/ios/WKActionSheetAssistant.h:
* UIProcess/ios/WKActionSheetAssistant.mm:
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.cpp:
Be more explicit with compiler flags.
2021-03-16 Kate Cheney <katherine_cheney@apple.com>
[Swift] Ambiguous use of 'closeAllMediaPresentations'
https://bugs.webkit.org/show_bug.cgi?id=223275
<rdar://problem/75447989>
Reviewed by Tim Horton.
We added back deprecated media playback control APIs in
https://trac.webkit.org/changeset/273186/webkit after a name change
to support early adoptors. Importing these methods to Swift causes
a method redefinition error because the completion handler in both
cases is nullable. We should ignore the deprecated APIs for Swift.
* UIProcess/API/Cocoa/WKWebView.h:
2021-03-16 John Wilander <wilander@apple.com>
PCM: Test infrastructure for sending attribution reports to attribution website too
https://bugs.webkit.org/show_bug.cgi?id=223238
<rdar://problem/75462031>
Reviewed by Brent Fulgham.
setPrivateClickMeasurementAttributionReportURLForTesting() now is named
setPrivateClickMeasurementAttributionReportURLsForTesting() in the layers that
eventually reach WebKit::PrivateClickMeasurementManager. It takes two URLs
attributionReportSourceURL and attributionReportAttributeOnURL and sets them
in a new struct AttributionReportTestConfig.
No new tests. This patch just adds test infrastructure for a coming patch.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::setPrivateClickMeasurementAttributionReportURLsForTesting):
(WebKit::NetworkProcess::setPrivateClickMeasurementAttributionReportURLForTesting): Deleted.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/NetworkSession.cpp:
(WebKit::NetworkSession::setPrivateClickMeasurementAttributionReportURLsForTesting):
(WebKit::NetworkSession::setPrivateClickMeasurementAttributionReportURLForTesting): Deleted.
* NetworkProcess/NetworkSession.h:
* NetworkProcess/PrivateClickMeasurementManager.cpp:
(WebKit::PrivateClickMeasurementManager::fireConversionRequestImpl):
(WebKit::PrivateClickMeasurementManager::setAttributionReportURLsForTesting):
(WebKit::PrivateClickMeasurementManager::setAttributionReportURLForTesting): Deleted.
* NetworkProcess/PrivateClickMeasurementManager.h:
* UIProcess/API/C/WKPage.cpp:
(WKPageSetPrivateClickMeasurementAttributionReportURLsForTesting):
(WKPageSetPrivateClickMeasurementAttributionReportURLForTesting): Deleted.
* UIProcess/API/C/WKPagePrivate.h:
* UIProcess/API/Cocoa/WKWebViewPrivateForTesting.h:
* UIProcess/API/Cocoa/WKWebViewTesting.mm:
(-[WKWebView _setPrivateClickMeasurementAttributionReportURLsForTesting:attributeOnURL:completionHandler:]):
(-[WKWebView _setPrivateClickMeasurementAttributionReportURLForTesting:completionHandler:]): Deleted.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::setPrivateClickMeasurementAttributionReportURLsForTesting):
(WebKit::WebPageProxy::setPrivateClickMeasurementAttributionReportURLForTesting): Deleted.
* UIProcess/WebPageProxy.h:
2021-03-16 Per Arne Vollan <pvollan@apple.com>
The WebContent process crashes when launching Safari
https://bugs.webkit.org/show_bug.cgi?id=223264
<rdar://75482851>
Reviewed by Brent Fulgham.
The WebContent process crashes when launching Safari on older OSes, because of unavailable sandbox features.
* WebProcess/com.apple.WebProcess.sb.in:
2021-03-16 Alex Christensen <achristensen@webkit.org>
Terminate network process when making new WebProcessPool in Lutron App
https://bugs.webkit.org/show_bug.cgi?id=223194
<rdar://74674915>
Reviewed by Geoff Garen.
The Lutron App uses allocation of a new WKProcessPool as a way to stop using old session cookies.
Since r267763 we keep the session cookies for the lifetime of a WKWebsiteDataStore instead of making a new process for each WKProcessPool.
Add a quirk for this app until they update to use a new WKWebsiteDataStore or use WKHTTPCookieStore API to clear the cookies properly.
I manually verified this causes the logging out to succeed.
* UIProcess/WebProcessPool.cpp:
2021-03-16 Kate Cheney <katherine_cheney@apple.com>
Add internal additions for WebAuthn compatibility
https://bugs.webkit.org/show_bug.cgi?id=223168
<rdar://problem/74890060>
Reviewed by Brent Fulgham.
* UIProcess/WebAuthentication/Cocoa/LocalAuthenticator.mm:
(shouldUseAlternateQuery):
(alternateQueryForRPID):
(WebKit::LocalAuthenticatorInternal::getExistingCredentials):
2021-03-16 Peng Liu <peng.liu6@apple.com>
[iOS] Scroll event is fired before webkitfullscreenchange event when an element enters fullscreen
https://bugs.webkit.org/show_bug.cgi?id=223221
Reviewed by Eric Carlson.
Defer the "scroll" event until the "webkitfullscreenchange" event is fired
when an element enters fullscreen. We have to do that because some sites use
"scroll" event handler to remove elements out of visible area, which may
conflict with the entering fullscreen process.
No new tests. We cannot test this patch with a layout test because
`InjectedBundlePageFullScreenClient` in WKTR does not support testing the
fullscreen implementation in the UI process, which is related to the "scroll"
event. We cannot add an API test on iOS for it either because the test binary
is not a real "app" (webkit.org/b/212654).
* UIProcess/ios/fullscreen/WKFullScreenWindowControllerIOS.mm:
(-[WKFullScreenWindowController enterFullScreen]):
(-[WKFullScreenWindowController beganEnterFullScreenWithInitialFrame:finalFrame:]):
2021-03-16 Per Arne <pvollan@apple.com>
Add signpost for injected bundle creation
https://bugs.webkit.org/show_bug.cgi?id=223068
Reviewed by Alex Christensen.
Emit signposts for creation of injected bundle.
* WebProcess/InjectedBundle/InjectedBundle.cpp:
(WebKit::InjectedBundle::create):
2021-03-16 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK][WPE] Stop using g_memdup
https://bugs.webkit.org/show_bug.cgi?id=223189
Reviewed by Philippe Normand.
* UIProcess/API/glib/WebKitWebResource.cpp:
(webkit_web_resource_get_data_finish): Use g_malloc + memcpy instead of g_memdup.
* UIProcess/API/glib/WebKitWebView.cpp:
(webkit_web_view_save_finish): Use fastMemDup instead g_memdup.
2021-03-16 Khem Raj <raj.khem@gmail.com>
[CMake] Build fails on RISC-V with GCC 11
https://bugs.webkit.org/show_bug.cgi?id=222959
Reviewed by Carlos Alberto Lopez Perez.
Use renamed ATOMICS_REQUIRE_LIBATOMIC instead of ATOMIC_INT64_REQUIRES_LIBATOMIC.
* CMakeLists.txt:
2021-03-16 Lauro Moura <lmoura@igalia.com>
[SOUP] Fix SOUP3 debug build
https://bugs.webkit.org/show_bug.cgi?id=223237
Reviewed by Carlos Garcia Campos.
* NetworkProcess/soup/NetworkDataTaskSoup.cpp:
(WebKit::NetworkDataTaskSoup::acceptCertificateCallback):
2021-03-16 Youenn Fablet <youenn@apple.com>
Add a new delegate for geolocation permission
https://bugs.webkit.org/show_bug.cgi?id=223183
Reviewed by Eric Carlson.
Refactoring to reuse media capture prompt for geolocation with updated prompt strings.
Introduce a new permission delegate for geolocation which provides both the top level origin as well as frame info.
The decision handler can get grant, deny or prompt.
In case of prompt, we use the updated prompt to get grant or deny decision from the user.
* UIProcess/API/Cocoa/WKUIDelegatePrivate.h:
* UIProcess/Cocoa/MediaPermissionUtilities.mm:
(WebKit::alertMessageText):
(WebKit::allowButtonText):
(WebKit::doNotAllowButtonText):
(WebKit::alertForPermission):
* UIProcess/Cocoa/UIDelegate.h:
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::setDelegate):
(WebKit::UIDelegate::UIClient::decidePolicyForGeolocationPermissionRequest):
* UIProcess/MediaPermissionUtilities.h:
* UIProcess/SpeechRecognitionPermissionManager.cpp:
(WebKit::SpeechRecognitionPermissionManager::decideByDefaultAction):
* UIProcess/UserMediaPermissionRequestProxy.cpp:
(WebKit::UserMediaPermissionRequestProxy::prompt):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::requestGeolocationPermissionForFrame):
2021-03-15 Chris Dumez <cdumez@apple.com>
Unreviewed, revert r272629 as it caused regressions in some apps
https://bugs.webkit.org/show_bug.cgi?id=223232
<rdar://75440378>
* SourcesCocoa.txt:
* UIProcess/API/Cocoa/WKWebView.mm:
* UIProcess/API/Cocoa/WKWebViewInternal.h:
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/Cocoa/PageClientImplCocoa.h:
* UIProcess/Cocoa/PageClientImplCocoa.mm:
* UIProcess/Cocoa/WKBlankOverlayView.h: Removed.
* UIProcess/PageClient.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didCommitLoadForFrame):
(WebKit::WebPageProxy::runJavaScriptAlert):
(WebKit::WebPageProxy::runJavaScriptConfirm):
(WebKit::WebPageProxy::runJavaScriptPrompt):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::didCommitLayerTree):
* UIProcess/mac/WebPageProxyMac.mm:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::didCommitLoad):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::updateRendering):
* WebProcess/WebPage/mac/WebPageMac.mm:
2021-03-15 Jer Noble <jer.noble@apple.com>
[WK2] Can get stuck in fullscreen mode if node is removed prior to receiving willEnterFullscreen()
https://bugs.webkit.org/show_bug.cgi?id=223218
<rdar://75009548>
Reviewed by Eric Carlson.
There are a number of preflight checks made in each of the listed functions below,
and those preflights can corrupt the state machine of the UIProcess's fullscreen
code. If any of the preflights fail, use the new return value of those methods to
close() the fullscreen presentation.
* WebProcess/FullScreen/WebFullScreenManager.cpp:
(WebKit::WebFullScreenManager::willEnterFullScreen):
(WebKit::WebFullScreenManager::didEnterFullScreen):
(WebKit::WebFullScreenManager::willExitFullScreen):
(WebKit::WebFullScreenManager::requestExitFullScreen):
2021-03-15 Chris Dumez <cdumez@apple.com>
Avoid heap allocation in RemoteAudioDestinationProxy::renderQuantum()
https://bugs.webkit.org/show_bug.cgi?id=223228
Reviewed by Darin Adler.
Avoid heap allocation in RemoteAudioDestinationProxy::renderQuantum() since this runs on the
audio thread.
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::connectToGPUProcess):
(WebKit::RemoteAudioDestinationProxy::renderQuantum):
2021-03-15 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] WebPageProxy's EditorState may be stale until the next remote layer tree commit
https://bugs.webkit.org/show_bug.cgi?id=223207
<rdar://problem/75410700>
Reviewed by Tim Horton.
Fixes a race condition wherein it's possible for the cached `EditorState` in the UI process (on `WebPageProxy`)
to be set to a stale (older) value after being set to a newer value. This can happen in the case where:
1. An editor state update is computed and set on a remote layer tree transaction in the web process, inside
`WebPage::willCommitLayerTree`. A pending `BackingStoreFlusher` is created, and we dispatch a task onto a
background queue to flush the layer tree commit.
2. IPC that updates the selection arrives in the web process (e.g. WebPage::selectTextWithGranularityAtPoint),
and causes us to send an `EditorStateChanged` message to the UI process with this new selection.
3. The pending `BackingStoreFlusher` created in step (1) dispatches on the background queue, and sends its
editor state over to the UI process.
As a result, the UI process receives the `EditorStateChanged` message in step (2) before it receives the remote
layer tree transaction sent in step (3), which contains the editor state computed prior to step (2); we then set
`WebPageProxy::m_editorState` from its value computed in step (2) to the value computed in step (1).
To fix this (and any similar issues involving out-of-band editor state updates), we introduce a monotonically
increasing transaction ID on editor state that's similar to the one used for remote layer tree commits; we then
avoid updating the editor state in `WebPageProxy` if we receive an `EditorState` whose transaction ID is less
than our current ID.
Test: fast/images/image-extraction/ios/reselect-word-in-image-overlay.html
* Shared/EditorState.cpp:
(WebKit::EditorState::encode const):
(WebKit::EditorState::decode):
* Shared/EditorState.h:
Add the TransactionID, and add encoding/decoding support for the new member.
* UIProcess/RemoteLayerTree/RemoteLayerTreeDrawingAreaProxy.mm:
(WebKit::RemoteLayerTreeDrawingAreaProxy::commitLayerTree):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::editorStateChanged):
(WebKit::WebPageProxy::updateEditorState):
Refactor this to return a `bool`, indicating whether the editor state was modified (this only returns `false` in
the case where the incoming editor state was older than the one we currently have). We also make this method
platform-agnostic, and move the platform-dependent logic previously implemented by individual `updateEditorState`
implementations into `WebPageProxy::didUpdateEditorState` instead.
* UIProcess/WebPageProxy.h:
* UIProcess/gtk/WebPageProxyGtk.cpp:
(WebKit::WebPageProxy::didUpdateEditorState):
(WebKit::WebPageProxy::updateEditorState): Deleted.
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::didUpdateEditorState):
Add a new platform method hook that is called when `m_editorState` is updated due to an incoming `EditorState`
from the web process. The old and new cached `EditorState` values are passed into this method; note that even
though only the previous value is necessary as an argument (since `m_editorState` is already equal to the new
value here), I chose to pass in both to make it clear that both old and new values can be consulted here.
(WebKit::WebPageProxy::updateEditorState): Deleted.
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::didUpdateEditorState):
(WebKit::WebPageProxy::updateEditorState): Deleted.
* UIProcess/playstation/WebPageProxyPlayStation.cpp:
(WebKit::WebPageProxy::didUpdateEditorState):
(WebKit::WebPageProxy::updateEditorState): Deleted.
* UIProcess/win/WebPageProxyWin.cpp:
(WebKit::WebPageProxy::didUpdateEditorState):
(WebKit::WebPageProxy::updateEditorState): Deleted.
* UIProcess/wpe/WebPageProxyWPE.cpp:
(WebKit::WebPageProxy::didUpdateEditorState):
(WebKit::WebPageProxy::updateEditorState): Deleted.
Rename these method implementations, and adjust them to use `oldEditorState` and `newEditorState`.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::editorState const):
* WebProcess/WebPage/WebPage.h:
2021-03-15 Alex Christensen <achristensen@webkit.org>
REGRESSION(r271642) Another app was relying on DOMWindow reuse
https://bugs.webkit.org/show_bug.cgi?id=223217
<rdar://75186172>
Reviewed by Geoff Garen.
* UIProcess/API/Cocoa/WKProcessPool.mm:
(+[WKProcessPool _setLinkedOnOrBeforeEverythingForTesting]):
* UIProcess/API/Cocoa/WKProcessPoolPrivate.h:
2021-03-15 John Wilander <wilander@apple.com>
PCM: Split attributionReportURL() into attributionReportSourceURL() and attributionReportAttributeOnURL()
https://bugs.webkit.org/show_bug.cgi?id=223210
<rdar://problem/75442246>
Reviewed by Brent Fulgham.
This patch splits WebCore::PrivateClickMeasurement::attributionReportURL() into
WebCore::PrivateClickMeasurement::attributionReportSourceURL() and
WebCore::PrivateClickMeasurement::attributionReportAttributeOnURL() to
facilitate sending attribution reports to both the click source and destination.
* NetworkProcess/PrivateClickMeasurementManager.cpp:
(WebKit::PrivateClickMeasurementManager::storeUnattributed):
Removed a reference to ads in a log statement.
(WebKit::PrivateClickMeasurementManager::fireConversionRequestImpl):
Now uses WebCore::PrivateClickMeasurement::attributionReportSourceURL().
2021-03-15 Per Arne Vollan <pvollan@apple.com>
Enforce subcommand filtering
https://bugs.webkit.org/show_bug.cgi?id=223192
<rdar://75434409>
Reviewed by Brent Fulgham.
Enforce subcommand filtering in the WebContent process' sandbox.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* WebProcess/com.apple.WebProcess.sb.in:
2021-03-15 Jer Noble <jer.noble@apple.com>
[Cocoa] Adopt AudioComponentApplyServerRegistrations()
https://bugs.webkit.org/show_bug.cgi?id=223188
<rdar://75433850>
Reviewed by Eric Carlson.
The WebKit sandbox will block access to the AudioComponent XPC service when the
GPU process is enabled. Adopt new API to fetch the registered component data from
the priviledged UIProcess and consume that registration data in the WebContent
process, making it unnecessary to communicate with the AudioComponent XPC service
to instantiate components.
* UIProcess/Cocoa/WebProcessProxyCocoa.mm:
(WebKit::WebProcessProxy::sendAudioComponentRegistrations):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::initializeNewWebProcess):
* UIProcess/WebProcessProxy.h:
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::consumeAudioComponentRegistrations):
2021-03-15 Kimmo Kinnunen <kkinnunen@apple.com>
WebGL IPC should use shared memory for synchronous messages
https://bugs.webkit.org/show_bug.cgi?id=220974
<rdar://problem/73876947>
Reviewed by Geoffrey Garen.
Send synchronous messages via the stream shared memory.
Improves MotionMark on iMac1,1 from 5000 to ~5300 pts.
If the message is encodable as a value, try to do so.
If the reply is encodable as a value, try to do so.
A message or a reply is not encodable if an argument in the message
must be passed by reference via the OS specific mechanism
(e.g. a iosurface or a file handle cannot be sent as a value).
Similar to asynchronous messages, if the synchronous message fits to
the stream buffer, the message is sent via the buffer. Same with the
reply.
If the message or reply is not sent via the stream buffer, it is sent via
the normal IPC.
The client protocol is:
1) Write the message to the buffer and release it.
2) Wait until the server releases the whole buffer back to the client.
3) Read the reply from index 0.
4) Continue sending next message from index 0.
The server protocol is upon receiving a message that is synchronous:
1) Read the message from the message position, dispatch it.
2) Write the reply to index 0.
3) Release the whole buffer back to the client.
4) Continue reading next message from index 0.
The client will not know at the send time whether the server will reply
via the stream or via normal IPC message. The client will reserve
a IPC sync request ID for normal IPC reply purposes.
In case a message or a reply does not fit to the stream buffer, the
ProcessOutOfStreamMessage message is written to the buffer instead.
This will make the reader to wait for the normal IPC message.
Changes temporarily so that the creation of the context waits for the
confirmation that the creation was done. This is due to synchrononous
stream send unable to deliver the WasCreated message during
wait for reply, at the moment.
* GPUProcess/graphics/RemoteGraphicsContextGL.messages.in:
* Platform/IPC/ArgumentCoder.h:
* Platform/IPC/ArgumentCoders.h:
* Platform/IPC/Connection.cpp:
(IPC::Connection::createSyncMessageEncoder):
(IPC::Connection::pushPendingSyncRequestID):
(IPC::Connection::popPendingSyncRequestID):
(IPC::Connection::sendSyncMessage):
* Platform/IPC/Connection.h:
(IPC::Connection::makeSyncRequestID):
Move the code from sendSyncMessage to individual functions so
that the client can reserve a sync request ID for reply purposes.
* Platform/IPC/StreamClientConnection.h:
(IPC::StreamClientConnection::send):
(IPC::StreamClientConnection::trySendStream):
(IPC::StreamClientConnection::sendSync):
(IPC::StreamClientConnection::trySendSyncStream):
(IPC::StreamClientConnection::tryAcquireAll):
* Platform/IPC/StreamConnectionBuffer.h:
* Platform/IPC/StreamServerConnection.cpp:
(IPC::StreamServerConnectionBase::acquireAll):
(IPC::StreamServerConnectionBase::release):
(IPC::StreamServerConnectionBase::releaseAll):
* Platform/IPC/StreamServerConnection.h:
(IPC::StreamServerConnectionBase::sendSyncReply):
The implementation.
(IPC::StreamServerConnectionBase::tryAcquire):
(IPC::StreamServerConnection<Receiver>::dispatchStreamMessages):
Fix a typo StreamServerConnectionBase::tryAquire
(IPC::StreamServerConnection<Receiver>::dispatchStreamMessage):
If the message comes in as normal IPC message, it must be replied
via normal IPC.
Add a flag m_isDispatchingStreamMessage to indicate whether the
message being processed comes as a normal IPC message or a
stream message. This information must be stateful in the connection
class, as it is not passable as arguments when the execution goes through:
StreamServerConnection<Receiver>::dispatchStreamMessage()
Receiver::didReceiveStreamMessage()
handleMessageSynchronous()
StreamServerConnection<Receiver>::sendSyncReply()
* Scripts/webkit/messages.py:
Add properties NotStreamEncodable, NotStreamEncodableReply
that indicate if the message parameters are such that they
cannot be encoded into plain data buffer as values.
This is encoded as 'constexpr bool isStreamEncodable' and
'constexpr bool isStreamEncodableReply' for the stream messages.
* Scripts/webkit/tests/MessageArgumentDescriptions.cpp:
(IPC::jsValueForArguments):
(IPC::messageArgumentDescriptions):
* Scripts/webkit/tests/MessageNames.cpp:
(IPC::description):
(IPC::receiverName):
(IPC::isValidMessageName):
* Scripts/webkit/tests/MessageNames.h:
* Scripts/webkit/tests/TestWithIfMessageMessages.h:
* Scripts/webkit/tests/TestWithImageDataMessages.h:
* Scripts/webkit/tests/TestWithLegacyReceiverMessages.h:
* Scripts/webkit/tests/TestWithSemaphoreMessages.h:
* Scripts/webkit/tests/TestWithStream.messages.in:
* Scripts/webkit/tests/TestWithStreamBufferMessages.h:
* Scripts/webkit/tests/TestWithStreamMessageReceiver.cpp:
(WebKit::TestWithStream::didReceiveStreamMessage):
* Scripts/webkit/tests/TestWithStreamMessages.h:
(Messages::TestWithStream::SendMachSendRight::name):
(Messages::TestWithStream::SendMachSendRight::SendMachSendRight):
(Messages::TestWithStream::SendMachSendRight::arguments const):
(Messages::TestWithStream::ReceiveMachSendRight::name):
(Messages::TestWithStream::ReceiveMachSendRight::arguments const):
(Messages::TestWithStream::SendAndReceiveMachSendRight::name):
(Messages::TestWithStream::SendAndReceiveMachSendRight::SendAndReceiveMachSendRight):
(Messages::TestWithStream::SendAndReceiveMachSendRight::arguments const):
* Scripts/webkit/tests/TestWithSuperclassMessages.h:
* Scripts/webkit/tests/TestWithoutAttributesMessages.h:
Test changes related to NotStreamEncodable, NotStreamEncodableReply.
Other changes indicate an auxiliary change to message structs:
changing 'const bool isSync' to 'constexpr bool isSync'.
This is done for consistency, since 'isStreamEncodable*' properties
are also added as constexpr.
* Shared/WebCoreArgumentCoders.cpp:
* Shared/WebCoreArgumentCoders.h:
Make types that are WebGL sync message parameters or reply parameters
now encodable to a stream by making the encoder part a template.
Previously these were sent only by normal IPC, and did not need
to be polymorphic to the encoder.
2021-03-15 Youenn Fablet <youenn@apple.com>
Update getUserMedia delegate to expose frame info
https://bugs.webkit.org/show_bug.cgi?id=223178
Reviewed by Eric Carlson.
Covered by API test.
* UIProcess/API/Cocoa/WKUIDelegatePrivate.h:
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::UIClient::decidePolicyForUserMediaPermissionRequest):
Expose frame information in the permission delegate.
2021-03-15 Philippe Normand <pnormand@igalia.com>
REGRESSION(r233845): EME Keysystem permission when MediaStream is disabled
https://bugs.webkit.org/show_bug.cgi?id=223185
Reviewed by Eric Carlson.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::requestMediaKeySystemPermissionForFrame): This method is required for
EME build configurations, not MediaStream...
2021-03-15 Peng Liu <peng.liu6@apple.com>
[GPUP][MSE] Add an IPC message to implement SourceBufferPrivateRemote::canSwitchToType()
https://bugs.webkit.org/show_bug.cgi?id=223152
Reviewed by Eric Carlson.
Add a synchronous message to implement `SourceBufferPrivateRemote::canSwitchToType()`.
No new tests. This patch fixes a test failure.
* GPUProcess/media/RemoteSourceBufferProxy.cpp:
(WebKit::RemoteSourceBufferProxy::canSwitchToType):
* GPUProcess/media/RemoteSourceBufferProxy.h:
* GPUProcess/media/RemoteSourceBufferProxy.messages.in:
* WebProcess/GPU/media/SourceBufferPrivateRemote.cpp:
(WebKit::SourceBufferPrivateRemote::canSwitchToType):
2021-03-15 Daniel Kolesa <dkolesa@igalia.com>
[GLIB] Be more selective about bublewrapping libdirs
https://bugs.webkit.org/show_bug.cgi?id=221709
Reviewed by Carlos Garcia Campos.
On 32-bit systems, bubblewrap lib32 instead of lib64. This unbreaks
sandbox on some distributions that have lib32 (which glibc defaults
its ld cache lookups to) but configure default libdir to lib rather
than lib32.
* UIProcess/Launcher/glib/BubblewrapLauncher.cpp:
(WebKit::bubblewrapSpawn):
2021-03-15 Kimmo Kinnunen <kkinnunen@apple.com>
Move vertex array functionality from ExtensionsGL to GraphicsContextGL
https://bugs.webkit.org/show_bug.cgi?id=221665
Reviewed by Simon Fraser.
Generate WebGL GPUP implementation after the change.
* GPUProcess/graphics/RemoteGraphicsContextGL.messages.in:
* GPUProcess/graphics/RemoteGraphicsContextGLFunctionsGenerated.h:
(getGraphicsResetStatusARB):
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.h:
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxyFunctionsGenerated.cpp:
2021-03-15 Kimmo Kinnunen <kkinnunen@apple.com>
RemoteRenderingBackend commands are not processed in order if GPU process is waiting for replies
https://bugs.webkit.org/show_bug.cgi?id=222099
Reviewed by Simon Fraser.
Mark all asynchronous message sends with DispatchMessageEvenWhenWaitingForSyncReply.
Otherwise synchronous messages will be dispatched before asynchronous messages.
The messages are temporally dependent.
Consider initial message exchange in canvas-blending-color-over-image.html:
GPUConnectionToWebProcess_CreateRenderingBackend
RemoteRenderingBackend_CreateImageBuffer
RemoteRenderingBackend_DidCreateSharedDisplayListHandle
RemoteRenderingBackend_CreateImageBuffer
RemoteRenderingBackend_WakeUpAndApplyDisplayList
RemoteRenderingBackend_GetDataURLForImageBuffer
RemoteRenderingBackendProxy_DidCreateImageBufferBackend
RemoteRenderingBackendProxy_DidCreateImageBufferBackend
If GPU process is waiting for a sync reply when the sequence up
until GetDataURL.. arrives, then the dispatch order is:
GPUConnectionToWebProcess_CreateRenderingBackend
RemoteRenderingBackend_GetDataURLForImageBuffer
[... rest of the messages]
This is a rare case, affects only the case where:
1) These messages arrive before main thread processes the message queue
2) GPU process has sent a synchronous message and is waiting for a reply
In base case after RenderingBackend has been created and the work queue
receive queue has been installed, all messages are moved to the work
queue thread and dispatched in order of arrival.
No new tests, timing dependent and WebKit code cannot be
unit tested.
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::~RemoteRenderingBackendProxy):
(WebKit::RemoteRenderingBackendProxy::reestablishGPUProcessConnection):
(WebKit::RemoteRenderingBackendProxy::createImageBuffer):
(WebKit::RemoteRenderingBackendProxy::cacheNativeImage):
(WebKit::RemoteRenderingBackendProxy::cacheFont):
(WebKit::RemoteRenderingBackendProxy::deleteAllFonts):
(WebKit::RemoteRenderingBackendProxy::releaseRemoteResource):
(WebKit::RemoteRenderingBackendProxy::sendWakeupMessage):
(WebKit::RemoteRenderingBackendProxy::createItemBuffer):
2021-03-15 Kimmo Kinnunen <kkinnunen@apple.com>
Simulated WebGL context screen change events should work with GPU process
https://bugs.webkit.org/show_bug.cgi?id=223121
Reviewed by Darin Adler.
Add a hunk missing from "WebGL context screen change events should work with GPU process"
that implements the context-specific event for simulating the context change.
Fixes
fast/canvas/webgl/webglcontextchangedevent.html
* GPUProcess/graphics/RemoteGraphicsContextGL.cpp:
(WebKit::RemoteGraphicsContextGL::simulateEventForTesting):
2021-03-13 Peng Liu <peng.liu6@apple.com>
[GPUP][MSE] MediaSource::buffered and MediaSource::activeSourceBuffers do not update in the same run loop as MediaSource::endOfStream()
https://bugs.webkit.org/show_bug.cgi?id=221293
Reviewed by Jer Noble.
When a Web process needs to "pull" the latest value of `m_buffered` from the
GPU process (`updateBufferedFromTrackBuffers`), we need to use a synchronous
IPC message.
When the GPU process needs to "push" the latest value of `m_buffered` to
a Web process, we can use an asynchronous message, e.g, `RemoveCodedFrames`
and `SourceBufferPrivateAppendComplete`.
* GPUProcess/media/RemoteSourceBufferProxy.cpp:
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateAppendComplete):
(WebKit::RemoteSourceBufferProxy::updateBufferedFromTrackBuffers):
(WebKit::RemoteSourceBufferProxy::removeCodedFrames):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateBufferedRangesChanged): Deleted.
* GPUProcess/media/RemoteSourceBufferProxy.h:
* GPUProcess/media/RemoteSourceBufferProxy.messages.in:
* WebProcess/GPU/media/SourceBufferPrivateRemote.cpp:
(WebKit::SourceBufferPrivateRemote::updateBufferedFromTrackBuffers):
(WebKit::SourceBufferPrivateRemote::removeCodedFrames):
(WebKit::SourceBufferPrivateRemote::sourceBufferPrivateAppendComplete):
(WebKit::SourceBufferPrivateRemote::sourceBufferPrivateBufferedRangesChanged): Deleted.
* WebProcess/GPU/media/SourceBufferPrivateRemote.h:
* WebProcess/GPU/media/SourceBufferPrivateRemote.messages.in:
2021-03-13 Philippe Normand <pnormand@igalia.com>
Unreviewed, fix build warnings after r273204 and r274323
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::storageQuotaManager): Use portable uint64_t format specifier.
* Platform/IPC/StreamConnectionWorkQueue.cpp:
(IPC::StreamConnectionWorkQueue::StreamConnectionWorkQueue):
* Platform/IPC/StreamConnectionWorkQueue.h: Make m_name COCOA specific to avoid
unused-member on other platforms.
2021-03-13 Tim Horton <timothy_horton@apple.com>
Adopt DDMacAction instead of DDAction on macOS
https://bugs.webkit.org/show_bug.cgi?id=223145
<rdar://problem/70127512>
Reviewed by Megan Gardner.
* Platform/mac/MenuUtilities.mm:
(WebKit::actionForMenuItem):
(WebKit::menuItemForTelephoneNumber):
(WebKit::menuForTelephoneNumber):
Adopt the new class name, when available.
2021-03-12 Jer Noble <jer.noble@apple.com>
[Cocoa][WebM] Hang when reloading page before WebM content is loaded
https://bugs.webkit.org/show_bug.cgi?id=223139
<rdar://75351029>
Reviewed by Darin Adler.
No new tests; a truly deterministic test would require a .cgi script to block loading after
the WebM init segment, but a bug in the platform format reader causes URLs not ending in
.webm to fail to load the format reader plugin. Once this issue is fixed, we can write a
test to cover this behavior.
The WebM TrackEntry "enabled" bit is optional, and WebKit previously waited until any media data
was appended to say whether or not the track is enabled in the absense of an explicit signal.
Instead, assume any track that is not explicity disabled is enabled, for the purpose of the
format reader. This means that "enabled" queries will no longer block, which breaks the deadlock
when tearing down the AVAsset backing the WebM file.
* Shared/mac/MediaFormatReader/MediaTrackReader.cpp:
(WebKit::MediaTrackReader::copyProperty):
2021-03-12 Chris Fleizach <cfleizach@apple.com>
AX: PDF frame conversion routines need to be updated
https://bugs.webkit.org/show_bug.cgi?id=223138
Reviewed by Darin Adler.
PDF bounding boxes are wrong in WebKit because.
1) There's no way for PDF objects to get the primary screen height. So we need to be able to return the primary
screen height from an object in the PDF hierarchy.
2) The WKPDFPluginAccessibilityObject's position was not being converted correctly.
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(-[WKPDFPluginAccessibilityObject accessibilityAttributeValue:]):
(-[WKPDFPluginAccessibilityObject ALLOW_DEPRECATED_IMPLEMENTATIONS_END]):
(WebKit::PDFPlugin::boundsOnScreen const):
* WebProcess/WebPage/mac/WKAccessibilityWebPageObjectMac.mm:
(-[WKAccessibilityWebPageObject accessibilityAttributeValue:]):
2021-03-12 Michael Catanzaro <mcatanzaro@gnome.org>
REGRESSION(r274270): [WPE][GTK] Broke Epiphany test /embed/ephy-web-view/error-pages-not-stored-in-history
https://bugs.webkit.org/show_bug.cgi?id=223140
Reviewed by Alex Christensen.
Convert empty strings to NULL.
* UIProcess/API/glib/WebKitSecurityOrigin.cpp:
(webkit_security_origin_to_string):
2021-03-12 Myles C. Maxfield <mmaxfield@apple.com>
[macOS] MobileAsset fonts are broken in Reader mode in Safari
https://bugs.webkit.org/show_bug.cgi?id=223062
Reviewed by Simon Fraser.
Here is an ASCII-art description of how various kinds of installed fonts work in WebKit:
+----------------------++-----------------------------------------+-----------------------------------------+
| || Safari | Normal WKWebViews |
+======================++=========================================+=========================================+
| Preinstalled Fonts || Just works | Just works |
+----------------------++-----------------------------------------+-----------------------------------------+
| MobileAsset Fonts || Needs access to mobileassetd | Needs access to mobileassetd but not |
| || (and fontd for any subsequent requests | fontd. |
| || after the first MobileAsset font is | |
| || used) | App calls _grantAccessToAssetServices() |
| || | to vend the sandbox extension |
| || App calls _grantAccessToAssetServices() | |
| || to vend the sandbox extension to | And then the app needs to use |
| || mobileassetd | InjectedBundle to activate the fonts in |
| || | the web process |
| || Web process also needs a call to | |
| || CTFontManagerEnableAllUserFonts() and a | |
| || sandbox extension to access fontd | |
| || | |
| || After this call is made, font requests | |
| || go through fontd (like for normal | |
| || WKWebViews) | |
| || | |
| || And then the app needs to use | |
| || InjectedBundle to activate the fonts in | |
| || the web process | |
+----------------------++-----------------------------------------+-----------------------------------------+
| User-installed Fonts || Intentionally doesn't work | Needs access to fontd. |
| || | |
| || | Just works (we already vend the sandbox |
| || | extension upon WKWebView creation) |
+----------------------++-----------------------------------------+-----------------------------------------+
The part that this patch fixes is the "Web process also needs a call to CTFontManagerEnableAllUserFonts()
and a sandbox extension to access fontd" under MobileAsset Fonts / Safari.
From looking at this chart, it becomes clear that a new message is necessary that does:
1. Adds a sandbox extension so the web process can access fontd
2. Calls CTFontManagerEnableAllUserFonts() to cause platform font routines to use fontd
So that's exactly what this patch does. It adds new WKWebView SPI,
_switchFromStaticFontRegistryToUserFontRegistry, which does these two things.
Even when we start using fontd, that doesn't allow user-installed fonts in WebKit in Safari, because we'll still
continue to use kCTFontUserInstalledAttribute / kCTFontFallbackOptionAttribute, just like we do in Big Sur.
Test: WebKit.MobileAssetSandboxCheck
* UIProcess/API/C/WKPreferencesRefPrivate.h: Correct the comment.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _switchFromStaticFontRegistryToUserFontRegistry]):
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::switchFromStaticFontRegistryToUserFontRegistry):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
(WebKit::customizedReaderConfiguration): Deleted.
(WebKit::disableStaticFontRegistry): Deleted.
* UIProcess/WebPageProxy.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::~WebPage):
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::switchFromStaticFontRegistryToUserFontRegistry):
* WebProcess/glib/WebProcessGLib.cpp:
(WebKit::WebProcess::switchFromStaticFontRegistryToUserFontRegistry):
* WebProcess/playstation/WebProcessPlayStation.cpp:
(WebKit::WebProcess::switchFromStaticFontRegistryToUserFontRegistry):
* WebProcess/win/WebProcessWin.cpp:
(WebKit::WebProcess::switchFromStaticFontRegistryToUserFontRegistry):
2021-03-12 Youenn Fablet <youenn@apple.com>
Enable video capture in GPUProcess by default on iOS
https://bugs.webkit.org/show_bug.cgi?id=223061
Reviewed by Eric Carlson.
Covered by existing tests.
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultCaptureVideoInGPUProcessEnabled):
* UIProcess/UserMediaPermissionRequestManagerProxy.cpp:
(WebKit::UserMediaPermissionRequestManagerProxy::finishGrantingRequest):
We need to synchronously fill the granted requests, otherwise there is a risk that
the granted requests will be cleared (on page close for instance) and then later filled
for the page that was gone.
2021-03-12 Chris Dumez <cdumez@apple.com>
Unreviewed, fix debug assertion on bots after r274323.
Was failing to call the completion handler in a early return case.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::setQuotaLoggingEnabled):
2021-03-12 Chris Dumez <cdumez@apple.com>
Unreviewed, partial revert of r274286 because this introduced an assertion failure.
* NetworkProcess/WebStorage/StorageManagerSet.cpp:
(WebKit::StorageManagerSet::waitUntilSyncingLocalStorageFinished):
2021-03-12 Alex Christensen <achristensen@webkit.org>
Reduce maximum HashTable entry size to 400 bytes
https://bugs.webkit.org/show_bug.cgi?id=223106
Reviewed by Youenn Fablet.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::populateFromMemoryStore):
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:
(WebKit::pruneResources):
(WebKit::ResourceLoadStatisticsMemoryStore::aggregatedThirdPartyData const):
(WebKit::ResourceLoadStatisticsMemoryStore::recursivelyGetAllDomainsThatHaveRedirectedToThisDomain const):
(WebKit::ResourceLoadStatisticsMemoryStore::markAsPrevalentIfHasRedirectedToPrevalent):
(WebKit::ResourceLoadStatisticsMemoryStore::classifyPrevalentResources):
(WebKit::ResourceLoadStatisticsMemoryStore::hasHadUserInteraction):
(WebKit::ResourceLoadStatisticsMemoryStore::setPrevalentResource):
(WebKit::ResourceLoadStatisticsMemoryStore::dumpResourceLoadStatistics):
(WebKit::ResourceLoadStatisticsMemoryStore::isPrevalentResource const):
(WebKit::ResourceLoadStatisticsMemoryStore::isVeryPrevalentResource const):
(WebKit::ResourceLoadStatisticsMemoryStore::isRegisteredAsSubresourceUnder const):
(WebKit::ResourceLoadStatisticsMemoryStore::isRegisteredAsSubFrameUnder const):
(WebKit::ResourceLoadStatisticsMemoryStore::isRegisteredAsRedirectingTo const):
(WebKit::ResourceLoadStatisticsMemoryStore::isGrandfathered const):
(WebKit::ResourceLoadStatisticsMemoryStore::ensureResourceStatisticsForRegistrableDomain):
(WebKit::ResourceLoadStatisticsMemoryStore::createEncoderFromData const):
(WebKit::ResourceLoadStatisticsMemoryStore::mergeStatistics):
(WebKit::ResourceLoadStatisticsMemoryStore::updateCookieBlocking):
(WebKit::ResourceLoadStatisticsMemoryStore::processStatistics const):
(WebKit::ResourceLoadStatisticsMemoryStore::registrableDomainsToDeleteOrRestrictWebsiteDataFor):
(WebKit::ResourceLoadStatisticsMemoryStore::pruneStatisticsIfNeeded):
(WebKit::ResourceLoadStatisticsMemoryStore::removeDataForDomain):
(WebKit::ResourceLoadStatisticsMemoryStore::insertExpiredStatisticForTesting):
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.h:
* Shared/Authentication/AuthenticationManager.cpp:
(WebKit::AuthenticationManager::addChallengeToChallengeMap):
(WebKit::AuthenticationManager::shouldCoalesceChallenge const):
(WebKit::AuthenticationManager::coalesceChallengesMatching const):
(WebKit::AuthenticationManager::didReceiveAuthenticationChallenge):
(WebKit::AuthenticationManager::completeAuthenticationChallenge):
* Shared/Authentication/AuthenticationManager.h:
(WebKit::AuthenticationManager::Challenge::Challenge):
* WebProcess/WebCoreSupport/WebResourceLoadObserver.cpp:
(WebKit::WebResourceLoadObserver::ensureResourceStatisticsForRegistrableDomain):
(WebKit::WebResourceLoadObserver::statisticsForURL):
(WebKit::WebResourceLoadObserver::takeStatistics):
* WebProcess/WebCoreSupport/WebResourceLoadObserver.h:
2021-03-12 Youenn Fablet <youenn@apple.com>
[MacOS] Reenable Audio Capture in GPUProcess by default
https://bugs.webkit.org/show_bug.cgi?id=223060
Reviewed by Eric Carlson.
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultCaptureAudioInGPUProcessEnabled):
2021-03-12 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] GTK4 crashes with XVFB: GLXBadWindow
https://bugs.webkit.org/show_bug.cgi?id=223108
Reviewed by Žan Doberšek.
Handle GLXBadWindow errors in AcceleratedBackingStoreX11.
* UIProcess/gtk/AcceleratedBackingStoreX11.cpp:
(WebKit::AcceleratedBackingStoreX11::checkRequirements):
(WebKit::glxErrorCode):
(WebKit::AcceleratedBackingStoreX11::~AcceleratedBackingStoreX11):
(WebKit::AcceleratedBackingStoreX11::update):
2021-03-12 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] Bump API version when building with libsoup3
https://bugs.webkit.org/show_bug.cgi?id=223067
Reviewed by Adrian Perez de Castro.
* PlatformGTK.cmake: Use WEBKITGTK_API_DOC_VERSION instead of WEBKITGTK_API_VERSION for gtkdoc config files
* gtk/webkit2gtk-web-extension.pc.in: Add variables for gtk pkg-config file and libsoup version.
* gtk/webkit2gtk.pc.in: Ditto.
2021-03-11 Saam Barati <sbarati@apple.com>
Adopt VM_FLAGS_PERMANENT for the config vm mapping
https://bugs.webkit.org/show_bug.cgi?id=222086
<rdar://74402690>
Reviewed by Yusuke Suzuki and Mark Lam.
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceMain.mm:
(WebKit::XPCServiceMain):
2021-03-11 Sihui Liu <sihui_liu@appe.com>
Add some logging to help debug flaky quota tests
https://bugs.webkit.org/show_bug.cgi?id=222995
Reviewed by Youenn Fablet.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::setQuotaLoggingEnabled):
(WebKit::NetworkProcess::storageQuotaManager):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreSetQuotaLoggingEnabled):
* UIProcess/API/C/WKWebsiteDataStoreRef.h:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::setQuotaLoggingEnabled):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::setQuotaLoggingEnabled):
* UIProcess/WebsiteData/WebsiteDataStore.h:
2021-03-11 Chris Dumez <cdumez@apple.com>
Drop some unnecessary RunLoop::isMain() / IsMainRunLoop() checks
https://bugs.webkit.org/show_bug.cgi?id=223094
Reviewed by Darin Adler.
Drop some unnecessary RunLoop::isMain() / IsMainRunLoop() checks. callOnMainThread() / ensureOnMainRunLoop()
take care of calling their lambda synchronously when already on the main runloop.
* GPUProcess/graphics/RemoteGraphicsContextGL.cpp:
(WebKit::RemoteGraphicsContextGL::copyTextureFromMedia):
* Shared/mac/MediaFormatReader/MediaFormatReader.cpp:
(WebKit::MediaFormatReader::startOnMainThread):
* UIProcess/API/Cocoa/WKURLSchemeTask.mm:
(getExceptionTypeFromMainRunLoop):
(-[WKURLSchemeTaskImpl dealloc]):
* WebProcess/GPU/media/RemoteImageDecoderAVF.cpp:
(WebKit::RemoteImageDecoderAVF::createFrameImageAtIndex):
* WebProcess/WebPage/mac/WKAccessibilityWebPageObjectBase.mm:
(-[WKAccessibilityWebPageObjectBase accessibilityPluginObject]):
2021-03-11 Eric Carlson <eric.carlson@apple.com>
[GPU Process] http/tests/media/hls/hls-audio-tracks-locale-selection.html fails
https://bugs.webkit.org/show_bug.cgi?id=223102
<rdar://problem/75338802>
Reviewed by Jer Noble.
No new tests, these changes fix an existing test.
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::GPUConnectionToWebProcess): Set user preferred
languages if parameters.overrideLanguages isn't empty.
(WebKit::GPUConnectionToWebProcess::setUserPreferredLanguages): Set user preferred
languages.
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/GPUConnectionToWebProcess.messages.in:
* Shared/GPUProcessConnectionParameters.h: Add overrideLanguages.
(WebKit::GPUProcessConnectionParameters::encode const):
(WebKit::GPUProcessConnectionParameters::decode):
* WebProcess/GPU/GPUProcessConnection.cpp:
(WebKit::languagesChanged): Send new languages to GPU process.
(WebKit::GPUProcessConnection::GPUProcessConnection): Register for language changes.
(WebKit::GPUProcessConnection::~GPUProcessConnection): Unregister.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeGPUProcessConnectionParameters):Add overrideLanguages.
2021-03-11 Brent Fulgham <bfulgham@apple.com>
[iOS, macOS] Re-allow WebContent sandbox permissions for specific /etc files
https://bugs.webkit.org/show_bug.cgi?id=223088
<rdar://problem/75332107>
Reviewed by Per Arne Vollan.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* WebProcess/com.apple.WebProcess.sb.in:
2021-03-11 Devin Rousso <drousso@apple.com>
[macOS] override the background color of PDF documents to match PDFKit
https://bugs.webkit.org/show_bug.cgi?id=223091
<rdar://problem/74584770>
Reviewed by Tim Horton.
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::PDFPlugin):
2021-03-11 Chris Dumez <cdumez@apple.com>
Introduce WorkQueue::main() to get the main thread's work queue
https://bugs.webkit.org/show_bug.cgi?id=223087
Reviewed by Geoffrey Garen.
Introduce WorkQueue::main() to get the main thread's work queue. This allows us to port some more code from
dispatch_queue to WorkQueue. It also simplifies some code that has to deal that sometimes needs to run on
the main thread and other times on a background queue. Having a single WorkQueue type to represent both the
main thread and a background queue makes writing such code more convenient.
* NetworkProcess/cache/CacheStorageEngine.cpp:
(WebKit::CacheStorage::Engine::writeFile):
(WebKit::CacheStorage::Engine::readFile):
* NetworkProcess/cache/CacheStorageEngineCaches.cpp:
(WebKit::CacheStorage::Caches::retrieveOriginFromDirectory):
* NetworkProcess/cache/NetworkCacheIOChannel.h:
* NetworkProcess/cache/NetworkCacheIOChannelCocoa.mm:
(WebKit::NetworkCache::IOChannel::read):
(WebKit::NetworkCache::IOChannel::write):
* NetworkProcess/cache/NetworkCacheIOChannelCurl.cpp:
(WebKit::NetworkCache::IOChannel::read):
(WebKit::NetworkCache::IOChannel::write):
* NetworkProcess/cache/NetworkCacheIOChannelGLib.cpp:
(WebKit::NetworkCache::inputStreamReadReadyCallback):
(WebKit::NetworkCache::IOChannel::read):
(WebKit::NetworkCache::IOChannel::readSyncInThread):
(WebKit::NetworkCache::outputStreamWriteReadyCallback):
(WebKit::NetworkCache::IOChannel::write):
* NetworkProcess/cache/NetworkCacheStorage.cpp:
(WebKit::NetworkCache::Storage::dispatchReadOperation):
(WebKit::NetworkCache::Storage::dispatchWriteOperation):
(WebKit::NetworkCache::Storage::traverse):
* Shared/Cocoa/WebKit2InitializeCocoa.mm:
(WebKit::InitializeWebKit2):
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceMain.mm:
(WebKit::XPCServiceEventHandler):
* UIProcess/API/Cocoa/WKBrowsingContextController.mm:
(+[WKBrowsingContextController registerSchemeForCustomProtocol:]):
(+[WKBrowsingContextController unregisterSchemeForCustomProtocol:]):
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::scheduleActivityStateUpdate):
* UIProcess/PDF/WKPDFHUDView.mm:
(-[WKPDFHUDView initWithFrame:pluginIdentifier:page:]):
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::fetchData):
(WebKit::WebsiteDataStore::fetchDataAndApply):
* UIProcess/WebsiteData/WebsiteDataStore.h:
* UIProcess/ios/ProcessAssertionIOS.mm:
(-[WKProcessAssertionBackgroundTaskManager _scheduleReleaseTask]):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView requestRectsToEvadeForSelectionCommandsWithCompletionHandler:]):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::willOpenAppLink):
* UIProcess/mac/ServicesController.mm:
(WebKit::ServicesController::refreshExistingServices):
2021-03-11 Chris Dumez <cdumez@apple.com>
Use CallOnMainThreadAndWait() instead of CallAndMainThread() + BinarySemaphore
https://bugs.webkit.org/show_bug.cgi?id=223093
Reviewed by Geoffrey Garen.
* NetworkProcess/WebStorage/StorageManagerSet.cpp:
* UIProcess/API/glib/IconDatabase.cpp:
2021-03-11 Tim Horton <timothy_horton@apple.com>
Adopt WKSeparatedModelView for <model>
https://bugs.webkit.org/show_bug.cgi?id=223085
<rdar://problem/75330603>
Reviewed by Sam Weinig.
* UIProcess/RemoteLayerTree/ios/RemoteLayerTreeHostIOS.mm:
(WebKit::RemoteLayerTreeHost::makeNode):
* UIProcess/RemoteLayerTree/ios/RemoteLayerTreeViews.h:
Make a WKSeparatedModelView for <model>'s UI-side view if available.
2021-03-11 Alex Christensen <achristensen@webkit.org>
Revert r260302
https://bugs.webkit.org/show_bug.cgi?id=223048
Reviewed by Geoffrey Garen.
r260302 was based on the faulty assumption that if we receive bytes from a server after a client certificate challenge,
it must've accepted the client certificate we provided and we should always provide this certificate to this server.
This assumption is faulty because there are servers that request a client certificate but provide a response if you do
not provide a certificate that it accepts. I'm removing this code because as part of my investigation of rdar://73974226
I found that it did not help anything, and since then I committed r269162 which did help solve that problem.
* NetworkProcess/cocoa/NetworkSessionCocoa.h:
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(-[WKNetworkSessionDelegate URLSession:task:willPerformHTTPRedirection:newRequest:completionHandler:]):
(-[WKNetworkSessionDelegate URLSession:task:_schemeUpgraded:completionHandler:]):
(-[WKNetworkSessionDelegate URLSession:task:didReceiveChallenge:completionHandler:]):
(-[WKNetworkSessionDelegate URLSession:task:didCompleteWithError:]):
(-[WKNetworkSessionDelegate URLSession:dataTask:didReceiveResponse:completionHandler:]):
(WebKit::NetworkSessionCocoa::clientCertificateSuggestedForHost): Deleted.
(WebKit::NetworkSessionCocoa::taskServerConnectionSucceeded): Deleted.
(WebKit::NetworkSessionCocoa::taskFailed): Deleted.
(WebKit::NetworkSessionCocoa::successfulClientCertificateForHost const): Deleted.
2021-03-11 Chris Dumez <cdumez@apple.com>
Replace some usage of dispatch_queue with WorkQueue now that it supports sync dispatching
https://bugs.webkit.org/show_bug.cgi?id=223073
Reviewed by Darin Adler.
Replace some usage of dispatch_queue with WorkQueue now that it supports sync dispatching
since r274286.
* Shared/Cocoa/DefaultWebBrowserChecks.mm:
(WebKit::itpQueue):
(WebKit::determineITPState):
(WebKit::doesAppHaveITPEnabled):
2021-03-11 Per Arne Vollan <pvollan@apple.com>
[macOS] Add additional telemetry to WebContent sandbox
https://bugs.webkit.org/show_bug.cgi?id=223080
Reviewed by Brent Fulgham.
Add additional telemetry to WebContent sandbox on macOS.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* WebProcess/com.apple.WebProcess.sb.in:
2021-03-11 Philippe Normand <pnormand@igalia.com>
[WPE] WebKitMediaKeySystemPermissionRequest.h missing in top-level header
https://bugs.webkit.org/show_bug.cgi?id=223076
Reviewed by Adrian Perez de Castro.
* UIProcess/API/wpe/WebKitMediaKeySystemPermissionRequest.h: Fix copy/paste mistake from GTK version of this
header.
* UIProcess/API/wpe/webkit.h: The WebKitMediaKeySystemPermissionRequest header has to be
included here so apps can consume this new API.
2021-03-11 Michael Catanzaro <mcatanzaro@gnome.org>
REGRESSION(r274270): Broke WebKitSecurityOrigin docs
https://bugs.webkit.org/show_bug.cgi?id=223077
Reviewed by Darin Adler.
* UIProcess/API/glib/WebKitSecurityOrigin.cpp:
2021-03-11 Per Arne Vollan <pvollan@apple.com>
[macOS] Add message filtering check in sandbox
https://bugs.webkit.org/show_bug.cgi?id=223072
<rdar://75314821>
Reviewed by Brent Fulgham.
Message filters cannot be applied unconditionally; their presence needs to be checked first.
* WebProcess/com.apple.WebProcess.sb.in:
2021-03-11 Chris Dumez <cdumez@apple.com>
Introduce WorkQueue::dispatchSync()
https://bugs.webkit.org/show_bug.cgi?id=223049
Reviewed by Alex Christensen.
Adopt WorkQueue::dispatchSync().
* NetworkProcess/WebStorage/StorageManagerSet.cpp:
(WebKit::StorageManagerSet::waitUntilTasksFinished):
(WebKit::StorageManagerSet::waitUntilSyncingLocalStorageFinished):
* UIProcess/API/glib/IconDatabase.cpp:
(WebKit::IconDatabase::IconDatabase):
(WebKit::IconDatabase::~IconDatabase):
2021-03-11 Youenn Fablet <youenn@apple.com>
Add camera-related sandbox extensions to GPUProcess
https://bugs.webkit.org/show_bug.cgi?id=223059
Reviewed by Eric Carlson.
Align GPU process sandbox with WebProcess one, including additional camera related sandboxes.
Allow reading preferences for "com.apple.cmio"like done for coremedia.
* GPUProcess/GPUProcess.cpp:
(WebKit::GPUProcess::initializeGPUProcess):
* GPUProcess/GPUProcessCreationParameters.cpp:
(WebKit::GPUProcessCreationParameters::encode const):
(WebKit::GPUProcessCreationParameters::decode):
* GPUProcess/GPUProcessCreationParameters.h:
* GPUProcess/mac/com.apple.WebKit.GPUProcess.sb.in:
* UIProcess/GPU/GPUProcessProxy.cpp:
(WebKit::shouldCreateAppleCameraServiceSandboxExtension):
(WebKit::GPUProcessProxy::GPUProcessProxy):
* UIProcess/UserMediaProcessManager.cpp:
* WebProcess/com.apple.WebProcess.sb.in:
2021-03-11 Michael Catanzaro <mcatanzaro@gnome.org>
REGRESSION(r272469): [WPE][GTK] Epiphany UI process crashes when downloading PDFs, WebKitSecurityOrigin should use SecurityOriginData
https://bugs.webkit.org/show_bug.cgi?id=222943
Reviewed by Alex Christensen.
Since r272469, WebCore::SecurityOrigin no longer accepts custom protocols except those
registered with LegacySchemeRegistry. WebPage registers all custom protocols, but
WebPageProxy does not, so WebCore::SecurityOrigin now only supports custom protocols in the
web process, not the UI process. This causes Epiphany to crash when the protocol of its
WebKitSecurityOrigin is unexpectedly NULL.
Alex wants to reduce usage of WebCore::SecurityOrigin outside the web process, so instead of
registering custom protocols with LegacySchemeRegistry in the UI process -- making it harder
to eventually get rid of LegacySchemeRegistry -- we will transition WebKitSecurityOrigin
from WebCore::SecurityOrigin to WebCore::SecurityOriginData, which is a simple data store
for <protocol, host, port>. This is mostly sufficient to implement WebKitSecurityOrigin,
except for webkit_security_origin_is_opaque(). I considered multiple ways to handle this,
but ultimately decided to just deprecate it. Epiphany is the only client using this function
in order to implement a WebKitSecurityOrigin equality operation, and it does so using
origins that should never be opaque, so there are no compatibility concerns here.
* UIProcess/API/glib/WebKitAuthenticationRequest.cpp:
(webkit_authentication_request_get_security_origin):
* UIProcess/API/glib/WebKitSecurityOrigin.cpp:
(_WebKitSecurityOrigin::_WebKitSecurityOrigin):
(webkitSecurityOriginCreate):
(webkitSecurityOriginGetSecurityOriginData):
(webkit_security_origin_new):
(webkit_security_origin_new_for_uri):
(webkit_security_origin_get_protocol):
(webkit_security_origin_get_host):
(webkit_security_origin_get_port):
(webkit_security_origin_is_opaque):
(webkit_security_origin_to_string):
(webkitSecurityOriginGetSecurityOrigin): Deleted.
* UIProcess/API/glib/WebKitSecurityOriginPrivate.h:
* UIProcess/API/glib/WebKitWebContext.cpp:
(addOriginToMap):
* UIProcess/API/gtk/WebKitSecurityOrigin.h:
* UIProcess/API/wpe/WebKitSecurityOrigin.h:
2021-03-10 Peng Liu <peng.liu6@apple.com>
[GPU Process] Assertion under RenderLayerCompositor::computeCompositingRequirements()
https://bugs.webkit.org/show_bug.cgi?id=220375
Reviewed by Eric Carlson.
When "GPU Process: Media" is enabled, `MediaPlayer` will call `mediaPlayerEngineUpdated()`
three times when a media element tries to load a URL. The three calls happen at the
following time points:
1) `MediaPlayer` creates a `MediaPlayerPrivateRemote`.
2) `MediaPlayerPrivateRemote` receives the response of a "createMediaPlayer" message
from the GPU process and gets the initial configurations of the "remote" media player.
3) The `RemoteMediaPlayerProxy` creates a `MediaPlayerPrivate` in the GPU process and
notify the `MediaPlayerPrivateRemote` in the WebContent process.
The second call of `mediaPlayerEngineUpdated()` is unnecessary because the player's
configuration at that time is similar to `NullMediaPlayerPrivate`. This patch removes it.
For the third call of `mediaPlayerEngineUpdated()`, we have to make sure that
the `MediaPlayerPrivateRemote` has received the configuration update from the GPU
process before the call. Therefore, this patch removes the message
`MediaPlayerPrivateRemote::EngineUpdated`, and let `MediaPlayerPrivateRemote`
decide when to call `mediaPlayerEngineUpdated()`.
* GPUProcess/media/RemoteMediaPlayerManagerProxy.cpp:
(WebKit::RemoteMediaPlayerManagerProxy::createMediaPlayer):
* GPUProcess/media/RemoteMediaPlayerManagerProxy.h:
* GPUProcess/media/RemoteMediaPlayerManagerProxy.messages.in:
The `playerConfiguration` is meaningless in this step so we can remove it.
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
(WebKit::RemoteMediaPlayerProxy::mediaPlayerEngineUpdated): Deleted.
* GPUProcess/media/RemoteMediaPlayerProxy.h:
* GPUProcess/media/cocoa/RemoteMediaPlayerProxyCocoa.mm:
(WebKit::RemoteMediaPlayerProxy::prepareForPlayback):
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::MediaPlayerPrivateRemote):
(WebKit::MediaPlayerPrivateRemote::load):
(WebKit::MediaPlayerPrivateRemote::setConfiguration): Deleted.
(WebKit::MediaPlayerPrivateRemote::engineUpdated): Deleted.
* WebProcess/GPU/media/MediaPlayerPrivateRemote.h:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.messages.in:
* WebProcess/GPU/media/RemoteMediaPlayerManager.cpp:
(WebKit::RemoteMediaPlayerManager::createRemoteMediaPlayer):
2021-03-10 Chris Dumez <cdumez@apple.com>
Use RetainPtr<> / OSObjectPtr<> more in WebKit
https://bugs.webkit.org/show_bug.cgi?id=223030
Reviewed by Darin Adler.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::newTestingSession):
(WebKit::NetworkProcess::ensureSession):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/cocoa/NetworkProcessCocoa.mm:
(WebKit::NetworkProcess::clearDiskCache):
* NetworkProcess/mac/NetworkProcessMac.mm:
(WebKit::NetworkProcess::platformTerminate):
* Platform/IPC/Connection.h:
* Platform/IPC/cocoa/ConnectionCocoa.mm:
(IPC::Connection::platformInvalidate):
(IPC::Connection::cancelReceiveSource):
(IPC::Connection::open):
(IPC::Connection::initializeSendSource):
(IPC::Connection::receiveSourceEventHandler):
* Shared/Cocoa/DefaultWebBrowserChecks.mm:
(WebKit::itpQueue):
(WebKit::determineITPState):
(WebKit::doesAppHaveITPEnabled):
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceMain.mm:
(WebKit::XPCServiceEventHandler):
* Shared/ShareableResource.cpp:
(WebKit::createShareableResourceDeallocator):
(WebKit::ShareableResource::wrapInSharedBuffer):
* Shared/cf/ArgumentCodersCF.cpp:
(IPC::decode):
* UIProcess/mac/LegacySessionStateCoding.cpp:
(WebKit::encodeSessionHistoryEntryData):
* WebProcess/WebPage/Cocoa/WebCookieCacheCocoa.mm:
(WebKit::WebCookieCache::inMemoryStorageSession):
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.mm:
(WebKit::RemoteLayerTreeDrawingArea::RemoteLayerTreeDrawingArea):
(WebKit::RemoteLayerTreeDrawingArea::updateRendering):
2021-03-10 Devin Rousso <drousso@apple.com>
Add plumbing for `defaultPlaybackRate` to AVKit
https://bugs.webkit.org/show_bug.cgi?id=222991
<rdar://problem/75012417>
Reviewed by Eric Carlson.
* WebProcess/cocoa/PlaybackSessionManager.messages.in:
* WebProcess/cocoa/PlaybackSessionManager.h:
* WebProcess/cocoa/PlaybackSessionManager.mm:
(WebKit::PlaybackSessionInterfaceContext::rateChanged):
(WebKit::PlaybackSessionManager::rateChanged):
(WebKit::PlaybackSessionManager::setDefaultPlaybackRate): Added.
* UIProcess/Cocoa/PlaybackSessionManagerProxy.messages.in:
* UIProcess/Cocoa/PlaybackSessionManagerProxy.h:
* UIProcess/Cocoa/PlaybackSessionManagerProxy.mm:
(WebKit::PlaybackSessionModelContext::setDefaultPlaybackRate): Added.
(WebKit::PlaybackSessionModelContext::rateChanged):
(WebKit::PlaybackSessionManagerProxy::rateChanged):
(WebKit::PlaybackSessionManagerProxy::setDefaultPlaybackRate): Added.
* UIProcess/ios/fullscreen/WKFullScreenViewController.mm:
(WKFullScreenViewControllerPlaybackSessionModelClient::rateChanged):
2021-03-10 Per Arne Vollan <pvollan@apple.com>
[iOS] Adjust telemetry rule
https://bugs.webkit.org/show_bug.cgi?id=223045
Reviewed by Geoffrey Garen.
Adjust socket-option related telemetry rule on iOS.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2021-03-10 Per Arne <pvollan@apple.com>
[iOS] Add additional telemetry to WebContent sandbox
https://bugs.webkit.org/show_bug.cgi?id=223035
<rdar://75275161>
Reviewed by Geoffrey Garen.
Add additional telemetry to WebContent sandbox on iOS.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2021-03-10 Brent Fulgham <bfulgham@apple.com>
[Cocoa] Add additional bundle ID property to WKWebViewConfiguration
https://bugs.webkit.org/show_bug.cgi?id=222919
<rdar://problem/75013854>
Reviewed by Alex Christensen.
Add an additional property to _WKWebsiteDataStoreConfiguration to help thread the correct bundle
ID of the driving application to lower levels of WebKit so that we can provide better messaging
when a remote service (e.g., Safari View Controller or ASWebAuthenticationSession) is used. Currently
we often lack context and have to report a generic "web content" message that doesn't help a user
understand which app is actually requesting the load.
We cannot use either of the existing bundle ID's for this purpose since we always indicate 'com.apple.Safari'
as the source application to ensure proper handling of web traffic in lower levels of the system Network stack,
and sourceApplicationSecondaryIdentifier is used for Apple Pay purposes and cannot be repurposed for this task.
This first patch adds the property. A follow-up patch will flesh out the implementation.
* NetworkProcess/NetworkSessionCreationParameters.cpp:
(WebKit::NetworkSessionCreationParameters::encode const):
(WebKit::NetworkSessionCreationParameters::decode):
* NetworkProcess/NetworkSessionCreationParameters.h:
* NetworkProcess/cocoa/NetworkSessionCocoa.h:
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::attributedBundleIdentifier const):
(WebKit::NetworkSessionCocoa::NetworkSessionCocoa):
* UIProcess/API/Cocoa/_WKWebsiteDataStoreConfiguration.h:
* UIProcess/API/Cocoa/_WKWebsiteDataStoreConfiguration.mm:
(-[_WKWebsiteDataStoreConfiguration setAttributedBundleIdentifier:]):
(-[_WKWebsiteDataStoreConfiguration attributedBundleIdentifier]):
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::WebsiteDataStore::platformSetNetworkParameters):
* UIProcess/WebsiteData/WebsiteDataStoreConfiguration.cpp:
(WebKit::WebsiteDataStoreConfiguration::copy):
* UIProcess/WebsiteData/WebsiteDataStoreConfiguration.h:
(WebKit::WebsiteDataStoreConfiguration::attributedBundleIdentifier const):
(WebKit::WebsiteDataStoreConfiguration::setAttributedBundleIdentifier):
2021-03-09 Darin Adler <darin@apple.com>
[Cocoa] Make WebKit API Objective-C objects safe to release on non-main threads
https://bugs.webkit.org/show_bug.cgi?id=223013
Reviewed by Geoffrey Garen.
Used ensureOnMainRunLoop in dealloc methods where possible to make them
thread-safe. Used WebCoreObjCScheduleDeallocateOnMainRunLoop instead in cases
that were too complex to be done with ensureOnMainRunLoop.
* Shared/API/APIFrameHandle.cpp:
(API::FrameHandle::~FrameHandle): Deleted.
* Shared/API/APIFrameHandle.h: Use pragma once. Marked the class final.
Removed explicit destructor because it is not helpful.
* Shared/API/Cocoa/_WKFrameHandle.mm:
(-[_WKFrameHandle dealloc]): Use WebCoreObjCScheduleDeallocateOnMainRunLoop.
* Shared/API/Cocoa/_WKHitTestResult.mm:
(-[_WKHitTestResult dealloc]): Ditto.
* Shared/Cocoa/WKNSArray.mm:
(-[WKNSArray dealloc]): Ditto.
* Shared/Cocoa/WKNSData.mm:
(-[WKNSData dealloc]): Ditto.
* Shared/Cocoa/WKNSDictionary.mm:
(-[WKNSDictionary dealloc]): Ditto.
* Shared/Cocoa/WKObject.mm:
(-[WKObject dealloc]): Ditto.
* UIProcess/API/C/mac/WKPagePrivateMac.mm:
(-[WKObservablePageState dealloc]): Use ensureOnMainRunLoop.
* UIProcess/API/Cocoa/PageLoadStateObserver.h: Added clearObject for use in
the WKObservablePageState deallc method. Removed header guards since this is
an Objective-C header and #import takes care of it.
* UIProcess/API/Cocoa/WKBackForwardList.mm:
(-[WKBackForwardList dealloc]): Use WebCoreObjCScheduleDeallocateOnMainRunLoop.
* UIProcess/API/Cocoa/WKBackForwardListItem.mm:
(-[WKBackForwardListItem dealloc]): Ditto.
* UIProcess/API/Cocoa/WKBrowsingContextController.mm:
(-[WKBrowsingContextController dealloc]): Ditto.
* UIProcess/API/Cocoa/WKBrowsingContextGroup.mm:
(-[WKBrowsingContextGroup dealloc]): Ditto.
* UIProcess/API/Cocoa/WKConnection.mm:
(-[WKConnection dealloc]): Ditto.
* UIProcess/API/Cocoa/WKContentRuleList.mm:
(-[WKContentRuleList dealloc]): Ditto.
* UIProcess/API/Cocoa/WKContentRuleListStore.mm:
(-[WKContentRuleListStore dealloc]): Ditto.
* UIProcess/API/Cocoa/WKContentWorld.mm:
(-[WKContentWorld dealloc]): Ditto.
* UIProcess/API/Cocoa/WKDownload.mm:
(-[WKDownload dealloc]): Ditto.
* UIProcess/API/Cocoa/WKFrameInfo.mm:
(-[WKFrameInfo dealloc]): Ditto.
* UIProcess/API/Cocoa/WKHTTPCookieStore.mm:
(-[WKHTTPCookieStore dealloc]): Ditto.
* UIProcess/API/Cocoa/WKNavigation.mm:
(-[WKNavigation dealloc]): Ditto.
* UIProcess/API/Cocoa/WKNavigationAction.mm:
(-[WKNavigationAction dealloc]): Ditto.
* UIProcess/API/Cocoa/WKNavigationData.mm:
(-[WKNavigationData dealloc]): Ditto.
* UIProcess/API/Cocoa/WKNavigationResponse.mm:
(-[WKNavigationResponse dealloc]): Ditto.
* UIProcess/API/Cocoa/WKPreferences.mm:
(-[WKPreferences dealloc]): Ditto.
* UIProcess/API/Cocoa/WKProcessPool.mm:
(-[WKProcessPool dealloc]): Ditto.
* UIProcess/API/Cocoa/WKSecurityOrigin.mm:
(-[WKSecurityOrigin dealloc]): Ditto.
* UIProcess/API/Cocoa/WKUserContentController.mm:
(-[WKUserContentController dealloc]): Ditto.
* UIProcess/API/Cocoa/WKUserScript.mm:
(-[WKUserScript dealloc]): Ditto.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView dealloc]): Ditto.
* UIProcess/API/Cocoa/WKWebpagePreferences.mm:
(-[WKWebpagePreferences dealloc]): Ditto.
* UIProcess/API/Cocoa/WKWebsiteDataRecord.mm:
(-[WKWebsiteDataRecord dealloc]): Ditto.
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
(-[WKWebsiteDataStore dealloc]): Ditto.
* UIProcess/API/Cocoa/WKWindowFeatures.mm:
(-[WKWindowFeatures dealloc]): Ditto.
* UIProcess/API/Cocoa/_WKApplicationManifest.mm:
(-[_WKApplicationManifest dealloc]): Ditto.
* UIProcess/API/Cocoa/_WKAttachment.mm:
(-[_WKAttachment dealloc]): Ditto.
* UIProcess/API/Cocoa/_WKAutomationSession.mm:
(-[_WKAutomationSession dealloc]): Ditto.
* UIProcess/API/Cocoa/_WKContentRuleListAction.mm:
(-[_WKContentRuleListAction dealloc]): Ditto.
* UIProcess/API/Cocoa/_WKCustomHeaderFields.mm:
(-[_WKCustomHeaderFields dealloc]): Ditto.
* UIProcess/API/Cocoa/_WKExperimentalFeature.mm:
(-[_WKExperimentalFeature dealloc]): Ditto.
* UIProcess/API/Cocoa/_WKFrameTreeNode.mm:
(-[_WKFrameTreeNode dealloc]): Ditto.
* UIProcess/API/Cocoa/_WKGeolocationPosition.mm:
(-[_WKGeolocationPosition dealloc]): Ditto.
* UIProcess/API/Cocoa/_WKInspectorConfiguration.mm:
(-[_WKInspectorConfiguration dealloc]): Ditto.
* UIProcess/API/Cocoa/_WKInspectorDebuggableInfo.mm:
(-[_WKInspectorDebuggableInfo dealloc]): Ditto.
* UIProcess/API/Cocoa/_WKInspectorExtension.mm:
(-[_WKInspectorExtension dealloc]): Ditto.
* UIProcess/API/Cocoa/_WKInternalDebugFeature.mm:
(-[_WKInternalDebugFeature dealloc]): Ditto.
* UIProcess/API/Cocoa/_WKProcessPoolConfiguration.mm:
(-[_WKProcessPoolConfiguration dealloc]): Ditto.
* UIProcess/API/Cocoa/_WKResourceLoadInfo.mm:
(-[_WKResourceLoadInfo dealloc]): Ditto.
* UIProcess/API/Cocoa/_WKResourceLoadStatisticsFirstParty.mm:
(-[_WKResourceLoadStatisticsFirstParty dealloc]): Ditto.
* UIProcess/API/Cocoa/_WKResourceLoadStatisticsThirdParty.mm:
(-[_WKResourceLoadStatisticsThirdParty dealloc]): Ditto.
* UIProcess/API/Cocoa/_WKUserInitiatedAction.mm:
(-[_WKUserInitiatedAction dealloc]): Ditto.
* UIProcess/API/Cocoa/_WKUserStyleSheet.mm:
(-[_WKUserStyleSheet dealloc]): Ditto.
* UIProcess/API/Cocoa/_WKVisitedLinkStore.mm:
(-[_WKVisitedLinkStore dealloc]): Ditto.
* UIProcess/API/Cocoa/_WKWebAuthenticationAssertionResponse.mm:
(-[_WKWebAuthenticationAssertionResponse dealloc]): Ditto.
* UIProcess/API/Cocoa/_WKWebAuthenticationPanel.mm:
(-[_WKWebAuthenticationPanel dealloc]): Ditto.
* UIProcess/API/Cocoa/_WKWebsiteDataStoreConfiguration.mm:
(-[_WKWebsiteDataStoreConfiguration dealloc]): Ditto.
* UIProcess/API/mac/WKView.mm:
(-[WKView dealloc]): Ditto.
* UIProcess/Cocoa/WKSafeBrowsingWarning.mm:
(-[WKSafeBrowsingWarning dealloc]): Use ensureOnMainRunLoop.
* UIProcess/mac/WKPrintingView.mm:
(-[WKPrintingView dealloc]): Ditto.
* WebProcess/InjectedBundle/API/Cocoa/WKWebProcessPlugInFrame.mm:
(-[WKWebProcessPlugInFrame dealloc]): Use WebCoreObjCScheduleDeallocateOnMainRunLoop.
* WebProcess/InjectedBundle/API/Cocoa/WKWebProcessPlugInHitTestResult.mm:
(-[WKWebProcessPlugInHitTestResult dealloc]): Ditto.
* WebProcess/InjectedBundle/API/Cocoa/WKWebProcessPlugInNodeHandle.mm:
(-[WKWebProcessPlugInNodeHandle dealloc]): Ditto.
* WebProcess/InjectedBundle/API/Cocoa/WKWebProcessPlugInPageGroup.mm:
(-[WKWebProcessPlugInPageGroup dealloc]): Ditto.
* WebProcess/InjectedBundle/API/Cocoa/WKWebProcessPlugInRangeHandle.mm:
(-[WKWebProcessPlugInRangeHandle dealloc]): Ditto.
* WebProcess/InjectedBundle/API/Cocoa/WKWebProcessPlugInScriptWorld.mm:
(-[WKWebProcessPlugInScriptWorld dealloc]): Ditto.
* WebProcess/InjectedBundle/API/mac/WKDOMNode.mm:
(-[WKDOMNode dealloc]): Use ensureOnMainRunLoop.
* WebProcess/InjectedBundle/API/mac/WKDOMRange.mm:
(-[WKDOMRange dealloc]): Ditto.
* WebProcess/InjectedBundle/API/mac/WKWebProcessPlugIn.mm:
(-[WKWebProcessPlugInController dealloc]): Use WebCoreObjCScheduleDeallocateOnMainRunLoop.
* WebProcess/InjectedBundle/API/mac/WKWebProcessPlugInBrowserContextController.mm:
(-[WKWebProcessPlugInBrowserContextController dealloc]): Ditto.
2021-03-10 Wenson Hsieh <wenson_hsieh@apple.com>
Logic for updating the text selection when dragging selection handles should account for image overlays
https://bugs.webkit.org/show_bug.cgi?id=223010
Reviewed by Tim Horton.
Apply the same logic introduced in r272503 to iOS, when extending the selection by moving selection handles. In
the case where the hit-tested node is the image overlay container (as opposed to any text inside the container),
it's better avoid updating the selection, rather than extend the selection to the start of the image overlay.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::rangeForPointInRootViewCoordinates):
2021-03-10 Youenn Fablet <youenn@apple.com>
Do not send sandbox extensions to WebProcess if capture happens in GPUProcess
https://bugs.webkit.org/show_bug.cgi?id=222961
Reviewed by Eric Carlson.
In case capture does not happen in WebProcess, we do not need to send the tccd sandbox extension.
Ditto for camera/microphone sandbox extensions if capture happens in GPUProcess.
Manually tested.
* UIProcess/UserMediaPermissionRequestManagerProxy.cpp:
(WebKit::c):
(WebKit::UserMediaPermissionRequestManagerProxy::finishGrantingRequest):
* UIProcess/UserMediaProcessManager.cpp:
(WebKit::UserMediaProcessManager::willCreateMediaStream):
2021-03-10 Youenn Fablet <youenn@apple.com>
Remove getUserMedia denied requests if user grants a new getUserMedia request
https://bugs.webkit.org/show_bug.cgi?id=222962
<rdar://74805451>
Reviewed by Eric Carlson.
A user may deny an audio getUserMedia request.
On user gesture, user may be reprompted, in which case user may grant access.
Before the patch, after these two getUserMedia calls, if the web page was trying to call getUserMedia without user gesture, it would fail.
With this patch, we remove the first denied request based on the second granted request.
This allows getUserMedia to be granted, even without a user gesture.
Tests: fast/mediastream/granted-denied-request-management1.html
fast/mediastream/granted-denied-request-management2.html
* UIProcess/UserMediaPermissionRequestManagerProxy.cpp:
(WebKit::UserMediaPermissionRequestManagerProxy::finishGrantingRequest):
(WebKit::isMatchingDeniedRequest):
(WebKit::UserMediaPermissionRequestManagerProxy::wasRequestDenied):
(WebKit::UserMediaPermissionRequestManagerProxy::updateStoredRequests):
(WebKit::UserMediaPermissionRequestManagerProxy::getRequestAction):
* UIProcess/UserMediaPermissionRequestManagerProxy.h:
2021-03-10 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r274186.
https://bugs.webkit.org/show_bug.cgi?id=223023
triggers an infinite network process launch loop in GTK
Reverted changeset:
"REGRESSION (r272376): [iOS] ASSERTION FAILED:
sessionID.isEphemeral() || !path.isEmpty() in
WebKit::NetworkProcess::swServerForSession"
https://bugs.webkit.org/show_bug.cgi?id=222713
https://trac.webkit.org/changeset/274186
2021-03-09 Don Olmstead <don.olmstead@sony.com>
GLib JSC API headers should only include other GLib JSC API headers
https://bugs.webkit.org/show_bug.cgi?id=222803
Reviewed by Michael Catanzaro.
Update the includes and include directories.
* PlatformGTK.cmake:
* PlatformWPE.cmake:
* WebProcess/InjectedBundle/API/glib/DOM/WebKitDOMNode.cpp:
* WebProcess/InjectedBundle/API/glib/WebKitFrame.cpp:
2021-03-09 Chris Dumez <cdumez@apple.com>
REGRESSION (r274174): ASSERTION FAILED: m_shouldBeCalledOnMainThread == isMainThread() under WebKit::RemoteSampleBufferDisplayLayerManager::createLayer
https://bugs.webkit.org/show_bug.cgi?id=223015
<rdar://problem/75248289>
Unreviewed, r274174 mistakenly replaced a call to dispatchToThread() with a call to callOnMainRunLoop().
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayerManager.cpp:
(WebKit::RemoteSampleBufferDisplayLayerManager::createLayer):
2021-03-09 Sam Weinig <weinig@apple.com>
Preferences do not need to be passed to the WebProcess via WebPageCreationParameters since the store already is
https://bugs.webkit.org/show_bug.cgi?id=222945
Reviewed by Simon Fraser.
needsInAppBrowserPrivacyQuirks and textInteractionEnabled were both being passed to the WebProcess twice,
once via the preferences store, and once via explicit serialization.
Use the preferences infrastructure to do this for us instead, and remove a bunch of unneeded code.
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_lastNavigationWasAppBound):
(WebKit::WebPage::updatePreferences):
2021-03-09 Brent Fulgham <bfulgham@apple.com>
REGRESSION (r274109): Catalina Perf tests failing due to unexpected output (222993)
https://bugs.webkit.org/show_bug.cgi?id=222993
<rdar://75235585>
Reviewed by Alex Christensen.
NSURL generates a single line of console output when we use [NSURL init]. We can avoid hitting this
message by using the preferred default initializer.
Found by existing tests (e.g.: Animation/balls.html)
* Shared/Cocoa/ArgumentCodersCocoa.mm:
(-[WKSecureCodingURLWrapper initWithCoder:]):
(-[WKSecureCodingURLWrapper initWithURL:]):
2021-03-09 Chris Dumez <cdumez@apple.com>
Use UniqueRef<> instead of std::unique_ptr<> when passing IPC::Encoder to sendMessage()
https://bugs.webkit.org/show_bug.cgi?id=222992
Reviewed by Alex Christensen.
Use UniqueRef<> instead of std::unique_ptr<> when passing IPC::Encoder to sendMessage(), to
make it clear it cannot be null.
I had to update didReceiveSyncMessage() to return a boolean since we can no longer rely
on the function nulling out the encoder it is given (since I modified the type from
std::unique_ptr to WTF::UniqueRef).
* Platform/IPC/Connection.cpp:
(IPC::Connection::dispatchMessageReceiverMessage):
(IPC::Connection::createSyncMessageEncoder):
(IPC::Connection::sendMessage):
(IPC::Connection::sendSyncReply):
(IPC::Connection::sendSyncMessage):
(IPC::Connection::sendOutgoingMessages):
(IPC::Connection::dispatchSyncMessage):
* Platform/IPC/Connection.h:
(IPC::Connection::send):
(IPC::Connection::sendWithAsyncReply):
(IPC::Connection::sendSync):
* Platform/IPC/Encoder.cpp:
(IPC::Encoder::wrapForTesting):
* Platform/IPC/Encoder.h:
* Platform/IPC/HandleMessage.h:
(IPC::handleMessageSynchronous):
(IPC::handleMessageSynchronousWantsConnection):
(IPC::handleMessageAsync):
(IPC::handleMessageAsyncWantsConnection):
* Platform/IPC/MessageSender.cpp:
(IPC::MessageSender::sendMessage):
* Platform/IPC/MessageSender.h:
* Platform/IPC/StreamServerConnection.h:
(IPC::StreamServerConnectionBase::sendSyncReply):
* Platform/IPC/cocoa/ConnectionCocoa.mm:
(IPC::Connection::open):
(IPC::Connection::sendOutgoingMessage):
* Platform/IPC/unix/ConnectionUnix.cpp:
(IPC::Connection::sendOutgoingMessage):
* Platform/IPC/win/ConnectionWin.cpp:
(IPC::Connection::sendOutgoingMessage):
* Scripts/webkit/messages.py:
* UIProcess/AuxiliaryProcessProxy.cpp:
(WebKit::AuxiliaryProcessProxy::sendMessage):
* UIProcess/AuxiliaryProcessProxy.h:
(WebKit::AuxiliaryProcessProxy::send):
(WebKit::AuxiliaryProcessProxy::sendWithAsyncReply):
* UIProcess/DrawingAreaProxy.cpp:
(WebKit::DrawingAreaProxy::sendMessage):
* UIProcess/DrawingAreaProxy.h:
* UIProcess/ProvisionalPageProxy.cpp:
(WebKit::ProvisionalPageProxy::sendMessage):
* UIProcess/ProvisionalPageProxy.h:
* UIProcess/SuspendedPageProxy.cpp:
(WebKit::SuspendedPageProxy::sendMessage):
* UIProcess/SuspendedPageProxy.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::sendMessage):
* UIProcess/WebPageProxy.h:
* WebProcess/WebPage/IPCTestingAPI.cpp:
(WebKit::IPCTestingAPI::JSIPC::sendMessage):
(WebKit::IPCTestingAPI::JSIPC::sendSyncMessage):
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.mm:
(WebKit::RemoteLayerTreeDrawingArea::updateRendering):
(WebKit::RemoteLayerTreeDrawingArea::BackingStoreFlusher::create):
(WebKit::RemoteLayerTreeDrawingArea::BackingStoreFlusher::BackingStoreFlusher):
(WebKit::RemoteLayerTreeDrawingArea::BackingStoreFlusher::flush):
2021-03-09 Alex Christensen <achristensen@webkit.org>
REGRESSION (r272376): [iOS] ASSERTION FAILED: sessionID.isEphemeral() || !path.isEmpty() in WebKit::NetworkProcess::swServerForSession
https://bugs.webkit.org/show_bug.cgi?id=222713
Reviewed by Geoff Garen.
Because NetworkProcess::CreateNetworkConnectionToWebProcess is sent with SendOption::DispatchMessageEvenWhenWaitingForSyncReply, it is possible
for two messages of type NetworkProcess::AddWebsiteDataStore and two messages of type NetworkProcess::CreateNetworkConnectionToWebProcess to be in the queue,
but the second NetworkProcess::CreateNetworkConnectionToWebProcess jumps to the front of the line while the UI process is waiting for the reply to the first.
Then, when calling NetworkProcess::swServerForSession we expect the session's parameters to have been initialized in the network process already, but we cut
ahead of the parameter initialization message. This is a realistically rare condition that can only be hit when using SPI, but it is hit in the
ResourceLoadStatistics.StoreSuspension API test. To fix this, we send the WebsiteDataStoreParameters from each WebsiteDataStore in the NetworkProcessCreationParameters.
To avoid doing extra work, we introduce an early return in NetworkProcessProxy::addSession if we have already added parameters from this session to the network process.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::initializeNetworkProcess):
(WebKit::NetworkProcess::addSessionStorageQuotaManager):
* NetworkProcess/NetworkProcessCreationParameters.cpp:
(WebKit::NetworkProcessCreationParameters::encode const):
(WebKit::NetworkProcessCreationParameters::decode):
* NetworkProcess/NetworkProcessCreationParameters.h:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::sendCreationParametersToNewProcess):
(WebKit::NetworkProcessProxy::addSession):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::parametersFromEachWebsiteDataStore):
* UIProcess/WebsiteData/WebsiteDataStore.h:
2021-03-09 Chris Dumez <cdumez@apple.com>
Crash under WebViewImpl::flagsChanged()
https://bugs.webkit.org/show_bug.cgi?id=222989
Reviewed by Tim Horton.
Capture a WeakPtr for |this| in the lambda passed to interpretKeyEvent() and make sure |this|
is still alive before getting m_page. This is the same pattern that is used for other
interpretKeyEvent() calls in this file (but this one was missed).
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::flagsChanged):
2021-03-09 Chris Dumez <cdumez@apple.com>
Stop using callOnMainThread() / isMainThread() in WebKit2
https://bugs.webkit.org/show_bug.cgi?id=222986
Reviewed by Alex Christensen.
Use callOnMainRunLoop() / callOnMainRunLoopAndWait() / isMainRunLoop() in WebKit2
instead of callOnMainThread() / callOnMainThreadAndWait() / isMainThread().
Using the "MainThread" variants doesn't do the right thing in the UIProcess when
an iOS app is using both WK1 (with WebThread) and WK2.
* GPUProcess/graphics/RemoteGraphicsContextGL.cpp:
(WebKit::RemoteGraphicsContextGL::copyTextureFromMedia):
* GPUProcess/media/RemoteAudioDestinationManager.cpp:
(WebKit::RemoteAudioDestination::render):
* GPUProcess/media/RemoteMediaResource.cpp:
(WebKit::RemoteMediaResource::~RemoteMediaResource):
(WebKit::RemoteMediaResource::responseReceived):
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayerManager.cpp:
(WebKit::RemoteSampleBufferDisplayLayerManager::close):
(WebKit::RemoteSampleBufferDisplayLayerManager::createLayer):
(WebKit::RemoteSampleBufferDisplayLayerManager::releaseLayer):
* NetworkProcess/Downloads/Download.cpp:
(WebKit::Download::cancel):
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::storageQuotaManager):
* NetworkProcess/cache/CacheStorageEngine.cpp:
(WebKit::CacheStorage::getDirectorySize):
(WebKit::CacheStorage::Engine::diskUsage):
* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::NetworkDataTaskCocoa::resume):
* NetworkProcess/webrtc/NetworkRTCMonitor.cpp:
(WebKit::NetworkManagerWrapper::onNetworksChanged):
* NetworkProcess/webrtc/NetworkRTCProvider.cpp:
(WebKit::NetworkRTCProvider::createServerTCPSocket):
(WebKit::NetworkRTCProvider::createClientTCPSocket):
(WebKit::NetworkRTCProvider::createResolver):
(WebKit::NetworkRTCProvider::stopResolver):
(WebKit::NetworkRTCProvider::closeListeningSockets):
* Platform/IPC/Connection.cpp:
(IPC::Connection::sendMessage):
(IPC::Connection::dispatchMessage):
* Shared/Authentication/cocoa/AuthenticationManagerCocoa.mm:
(WebKit::AuthenticationManager::initializeConnection):
* Shared/mac/MediaFormatReader/MediaFormatReader.cpp:
(WebKit::MediaFormatReader::startOnMainThread):
(WebKit::MediaFormatReader::parseByteSource):
(WebKit::MediaFormatReader::didParseTracks):
(WebKit::MediaFormatReader::didProvideMediaData):
(WebKit::MediaFormatReader::finishParsing):
* Shared/mac/MediaFormatReader/MediaSampleByteRange.cpp:
(WebKit::MediaSampleByteRange::MediaSampleByteRange):
* Shared/mac/MediaFormatReader/MediaTrackReader.cpp:
(WebKit::MediaTrackReader::MediaTrackReader):
(WebKit::MediaTrackReader::addSample):
(WebKit::MediaTrackReader::finishParsing):
* Shared/mac/MediaFormatReader/MediaTrackReader.h:
* UIProcess/API/glib/IconDatabase.cpp:
(WebKit::IconDatabase::IconDatabase):
(WebKit::IconDatabase::~IconDatabase):
(WebKit::IconDatabase::createTablesIfNeeded):
(WebKit::IconDatabase::populatePageURLToIconURLMap):
(WebKit::IconDatabase::clearStatements):
(WebKit::IconDatabase::pruneTimerFired):
(WebKit::IconDatabase::startPruneTimer):
(WebKit::IconDatabase::clearLoadedIconsTimerFired):
(WebKit::IconDatabase::startClearLoadedIconsTimer):
(WebKit::IconDatabase::iconIDForIconURL):
(WebKit::IconDatabase::setIconIDForPageURL):
(WebKit::IconDatabase::iconData):
(WebKit::IconDatabase::addIcon):
(WebKit::IconDatabase::updateIconTimestamp):
(WebKit::IconDatabase::deleteIcon):
(WebKit::IconDatabase::checkIconURLAndSetPageURLIfNeeded):
(WebKit::IconDatabase::loadIconForPageURL):
(WebKit::IconDatabase::iconURLForPageURL):
(WebKit::IconDatabase::setIconForPageURL):
(WebKit::IconDatabase::clear):
* UIProcess/Cocoa/MediaPermissionUtilities.mm:
(WebKit::requestAVCaptureAccessForType):
(WebKit::requestSpeechRecognitionAccess):
* UIProcess/Cocoa/PreferenceObserver.mm:
(-[WKUserDefaults _notifyObserversOfChangeFromValuesForKeys:toValuesForKeys:]):
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::nonBrowserServices):
(WebKit::agxCompilerClasses):
(WebKit::agxCompilerServices):
(WebKit::diagnosticServices):
* UIProcess/Inspector/socket/RemoteInspectorClient.cpp:
(WebKit::RemoteInspectorClient::sendWebInspectorEvent):
(WebKit::RemoteInspectorClient::didClose):
* UIProcess/WebFramePolicyListenerProxy.cpp:
(WebKit::WebFramePolicyListenerProxy::didReceiveSafeBrowsingResults):
* UIProcess/WebPageProxy.cpp:
(WebKit::gpuIOKitClasses):
(WebKit::gpuMachServices):
(WebKit::mediaRelatedMachServices):
* UIProcess/ios/WKPDFView.mm:
(-[WKPDFView web_setContentProviderData:suggestedFilename:]):
* UIProcess/mac/WKPrintingView.mm:
(-[WKPrintingView dealloc]):
* WebProcess/Cache/WebCacheStorageConnection.cpp:
(WebKit::WebCacheStorageConnection::~WebCacheStorageConnection):
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
(WebKit::RemoteImageBufferProxy::waitForDidFlushOnSecondaryThread):
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::load):
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::renderQuantum):
* WebProcess/GPU/media/RemoteAudioSourceProvider.cpp:
(WebKit::RemoteAudioSourceProvider::RemoteAudioSourceProvider):
(WebKit::RemoteAudioSourceProvider::close):
* WebProcess/GPU/media/RemoteImageDecoderAVF.cpp:
(WebKit::RemoteImageDecoderAVF::createFrameImageAtIndex):
* WebProcess/GPU/webrtc/LibWebRTCCodecs.cpp:
(WebKit::LibWebRTCCodecs::setCallbacks):
(WebKit::LibWebRTCCodecs::failedDecoding):
(WebKit::LibWebRTCCodecs::completedDecoding):
(WebKit::LibWebRTCCodecs::completedEncoding):
* WebProcess/Network/webrtc/LibWebRTCResolver.cpp:
(WebKit::LibWebRTCResolver::sendOnMainThread):
* WebProcess/Network/webrtc/LibWebRTCSocketFactory.cpp:
(WebKit::LibWebRTCSocketFactory::createServerTcpSocket):
(WebKit::LibWebRTCSocketFactory::createUdpSocket):
(WebKit::LibWebRTCSocketFactory::createClientTcpSocket):
* WebProcess/Network/webrtc/WebRTCMonitor.cpp:
(WebKit::WebRTCMonitor::sendOnMainThread):
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::pdfLog):
(WebKit::PDFPlugin::logStreamLoader):
(WebKit::PDFPlugin::verboseLog):
(WebKit::PDFPlugin::receivedNonLinearizedPDFSentinel):
(WebKit::dataProviderGetBytesAtPositionCallback):
(WebKit::dataProviderGetByteRangesCallback):
(WebKit::dataProviderReleaseInfoCallback):
(WebKit::PDFPlugin::threadEntry):
(WebKit::PDFPlugin::getResourceBytesAtPositionMainThread):
(WebKit::PDFPlugin::getResourceBytesAtPosition):
(WebKit::PDFPlugin::adoptBackgroundThreadDocument):
(WebKit::PDFPlugin::installPDFDocument):
(WebKit::PDFPlugin::tryRunScriptsInPDFDocument):
* WebProcess/Storage/WebSWClientConnection.cpp:
(WebKit::WebSWClientConnection::didMatchRegistration):
(WebKit::WebSWClientConnection::didGetRegistrations):
(WebKit::WebSWClientConnection::matchRegistration):
(WebKit::WebSWClientConnection::getRegistrations):
* WebProcess/Storage/WebServiceWorkerFetchTaskClient.cpp:
(WebKit::WebServiceWorkerFetchTaskClient::didReceiveFormDataAndFinish):
(WebKit::WebServiceWorkerFetchTaskClient::cleanup):
* WebProcess/WebPage/FindController.cpp:
(WebKit::FindController::drawRect):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::elementDidBlur):
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::handleSyntheticClick):
(WebKit::WebPage::didFinishContentChangeObserving):
(WebKit::WebPage::updateInputContextAfterBlurringAndRefocusingElementIfNeeded):
* WebProcess/WebPage/mac/DrawingAreaMac.cpp:
(WebKit::DisplayRefreshMonitorMac::displayLinkFired):
* WebProcess/WebPage/mac/WKAccessibilityWebPageObjectBase.mm:
(-[WKAccessibilityWebPageObjectBase axObjectCache]):
(-[WKAccessibilityWebPageObjectBase accessibilityPluginObject]):
(-[WKAccessibilityWebPageObjectBase setWebPage:]):
(-[WKAccessibilityWebPageObjectBase setHasMainFramePlugin:]):
(-[WKAccessibilityWebPageObjectBase setRemoteParent:]):
* WebProcess/cocoa/RemoteRealtimeAudioSource.cpp:
(WebKit::RemoteRealtimeAudioSource::remoteAudioSamplesAvailable):
* WebProcess/cocoa/RemoteRealtimeMediaSourceProxy.cpp:
(WebKit::RemoteRealtimeMediaSourceProxy::connection):
2021-03-09 Chris Dumez <cdumez@apple.com>
[IPC Hardening] Protect against bad input in WebProcessProxy::createSpeechRecognitionServer() and MessageReceiverMap
https://bugs.webkit.org/show_bug.cgi?id=222948
<rdar://problem/75191472>
Reviewed by Alex Christensen.
Update MessageReceiverMap so that:
1. Trying to remove a MessageReceiver that is not in the map does not
do dangerous things.
2. It stores weak pointers to the receivers instead of raw pointers. This
would avoid doing bad things when trying to remove a message receiver
that's already been destroyed.
Add a MESSAGE_CHECK() in WebProcessProxy::createSpeechRecognitionServer() to
make sure the identifier is not already in the map. There used to be a debug
assertion but we should MESSAGE_CHECK() too since the value is coming from
IPC.
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/GPUProcess.h:
* GPUProcess/media/RemoteAudioSessionProxy.h:
* GPUProcess/media/RemoteCDMFactoryProxy.h:
* GPUProcess/media/RemoteCDMProxy.h:
* GPUProcess/media/RemoteLegacyCDMFactoryProxy.h:
* GPUProcess/media/RemoteLegacyCDMProxy.h:
* GPUProcess/media/RemoteLegacyCDMSessionProxy.h:
* GPUProcess/media/RemoteMediaEngineConfigurationFactoryProxy.h:
* GPUProcess/media/RemoteMediaPlayerManagerProxy.h:
* GPUProcess/media/RemoteMediaPlayerProxy.h:
* GPUProcess/media/RemoteMediaResourceManager.h:
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayer.h:
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/ServiceWorker/WebSWServerConnection.h:
* NetworkProcess/ServiceWorker/WebSWServerToContextConnection.h:
* Platform/IPC/MessageReceiver.h:
* Platform/IPC/MessageReceiverMap.cpp:
(IPC::MessageReceiverMap::addMessageReceiver):
(IPC::MessageReceiverMap::removeMessageReceiver):
(IPC::MessageReceiverMap::dispatchMessage):
(IPC::MessageReceiverMap::dispatchSyncMessage):
* Platform/IPC/MessageReceiverMap.h:
* Shared/API/Cocoa/RemoteObjectRegistry.h:
* Shared/ApplePay/WebPaymentCoordinatorProxy.h:
* Shared/Authentication/AuthenticationManager.h:
* Shared/AuxiliaryProcess.h:
* UIProcess/GPU/GPUProcessProxy.h:
* UIProcess/Inspector/WebInspectorProxy.h:
* UIProcess/Inspector/WebInspectorUIExtensionControllerProxy.h:
* UIProcess/Media/AudioSessionRoutingArbitratorProxy.h:
* UIProcess/Network/CustomProtocols/LegacyCustomProtocolManagerProxy.h:
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/ProvisionalPageProxy.h:
* UIProcess/SpeechRecognitionRemoteRealtimeMediaSourceManager.h:
* UIProcess/SpeechRecognitionServer.h:
* UIProcess/SuspendedPageProxy.h:
* UIProcess/WebAuthentication/WebAuthnProcessProxy.h:
* UIProcess/WebPageProxy.h:
* UIProcess/WebProcessPool.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::createSpeechRecognitionServer):
* UIProcess/WebProcessProxy.h:
* WebAuthnProcess/WebAuthnProcess.h:
* WebProcess/GPU/GPUProcessConnection.h:
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.h:
* WebProcess/GPU/media/MediaSourcePrivateRemote.h:
* WebProcess/GPU/media/SourceBufferPrivateRemote.h:
* WebProcess/GPU/webrtc/SampleBufferDisplayLayer.h:
* WebProcess/Inspector/WebInspectorUIExtensionController.h:
* WebProcess/Network/WebSocketChannel.h:
* WebProcess/Speech/SpeechRecognitionRealtimeMediaSourceManager.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.h:
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.h:
* WebProcess/WebStorage/StorageAreaMap.h:
2021-03-09 Fujii Hironori <Hironori.Fujii@sony.com>
[WinCairo] Deadlocks in WTF::Thread::ThreadHolder::~ThreadHolder while WebKitWebProcess.exe is shutting down
https://bugs.webkit.org/show_bug.cgi?id=222682
Reviewed by Don Olmstead.
WinCairo WebKit2 layout tests were observing deadlocks in
WTF::Thread::ThreadHolder::~ThreadHolder while
WebKitWebProcess.exe is shutting down in the IPC thread.
WebProcess calls _exit in IPC threads if a IPC connection is
disconnected.
r260911 (Bug 210955) fixed a similar deadlock by skipping the
destruction in ~ThreadHolder if the calling thread is the main
thread.
Microsoft Docs describe how a process is terminated and recommend
TerminateProcess API to avoid such deadlocks.
<https://docs.microsoft.com/en-us/windows/win32/api/processthreadsapi/nf-processthreadsapi-exitprocess>
* NetworkProcess/NetworkProcess.cpp:
(WebKit::callExitSoon): Use TerminateProcess instead of _exit.
* WebProcess/WebProcess.cpp:
(WebKit::callExit): Ditto.
2021-03-09 Ben Nham <nham@apple.com>
Adopt new NSURLSessionConfiguration SPI for connection cache configuration
https://bugs.webkit.org/show_bug.cgi?id=222934
Reviewed by Geoffrey Garen.
This uses the NSURLSessionConfiguration connection cache limit SPI introduced in Big Sur to
properly set the parameters we want for HTTP/1.1 connections. Previously we tried to do this
using _CFNetworkHTTPConnectionCacheSetLimit in NetworkProcessCocoa, but this didn't work
because that SPI only applies to NSURLConnection rather than NSURLSession.
In particular, this meant that the number of priority levels wasn't set correctly, which we
had to work around by constraining the number of priority levels when mapping WebKit
resource priorities to CFNetwork priorities (https://bugs.webkit.org/show_bug.cgi?id=203423).
This patch adopts the SPI and removes that workaround.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::configurationForSessionID):
2021-03-09 Chris Dumez <cdumez@apple.com>
[IPC Hardening] IPC::decode(Decoder& decoder, RetainPtr<CFDictionaryRef>&) should make sure keys & values aren't null
https://bugs.webkit.org/show_bug.cgi?id=222980
Reviewed by Geoffrey Garen.
[NSMutableDictionary setObject:forKey:] throws an exception when the given object or key is nil. The dictionary decoder
should therefore fail nicely when either of these is nil, instead of crashing.
* Shared/cf/ArgumentCodersCF.cpp:
(IPC::decode):
2021-03-09 Chris Dumez <cdumez@apple.com>
[IPC Hardening] SandboxExtension::HandleArray IPC decoder should not call Vector::resize()
https://bugs.webkit.org/show_bug.cgi?id=222977
<rdar://problem/75218451>
Reviewed by Anders Carlsson.
SandboxExtension::HandleArray IPC decoder should not call Vector::resize() with an untrusted size
coming from IPC. Instead, call Vector::append(), like the Vector IPC decoder does.
* Shared/Cocoa/SandboxExtensionCocoa.mm:
(WebKit::SandboxExtension::HandleArray::append):
(WebKit::SandboxExtension::HandleArray::decode):
* Shared/SandboxExtension.h:
(WebKit::SandboxExtension::append):
2021-03-09 Peng Liu <peng.liu6@apple.com>
[GPUP] Test fast/images/animated-image-mp4.html times out when media in GPU Process is enabled
https://bugs.webkit.org/show_bug.cgi?id=221795
Reviewed by Eric Carlson.
Pass the color space along with a `MachSendRight` created from an `IOSurface` to ensure
that we can recreate the image from the `MachSendRight`/`IOSurface` properly.
* GPUProcess/media/RemoteImageDecoderAVFProxy.cpp:
(WebKit::RemoteImageDecoderAVFProxy::createFrameImageAtIndex):
* GPUProcess/media/RemoteImageDecoderAVFProxy.h:
* GPUProcess/media/RemoteImageDecoderAVFProxy.messages.in:
* WebProcess/GPU/media/RemoteImageDecoderAVF.cpp:
(WebKit::RemoteImageDecoderAVF::createFrameImageAtIndex):
This functions needs to return a valid image when it is called in the main thread,
otherwise, the animation will pause.
2021-03-09 Chris Dumez <cdumez@apple.com>
[IPC Hardening] Protect WebPageProxy::willSubmitForm() against bad Strings
https://bugs.webkit.org/show_bug.cgi?id=222955
<rdar://problem/75195062>
Reviewed by Anders Carlsson.
The Strings passed in textFieldValues are used as keys in a HashMap later on so we need
to validate them.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::willSubmitForm):
2021-03-09 Wenson Hsieh <wenson_hsieh@apple.com>
[macOS] Add a way to trigger webpage translation via the context menu
https://bugs.webkit.org/show_bug.cgi?id=222953
<rdar://problem/73901967>
Reviewed by Devin Rousso.
Add support for the new context menu item in WebKit2. See below for more details.
* Shared/API/c/WKContextMenuItemTypes.h:
* Shared/API/c/WKSharedAPICast.h:
(WebKit::toAPI):
(WebKit::toImpl):
Add a C API value for the new WebCore enum.
* Shared/ContextMenuContextData.cpp:
(WebKit::ContextMenuContextData::ContextMenuContextData):
(WebKit::ContextMenuContextData::encode const):
(WebKit::ContextMenuContextData::decode):
* Shared/ContextMenuContextData.h:
(WebKit::ContextMenuContextData::selectionBounds const):
Add a rect, `selectionBounds`, to `ContextMenuContextData` that represents the rect of the current text
selection in root view coordinates.
* UIProcess/API/Cocoa/WKMenuItemIdentifiers.mm:
* UIProcess/API/Cocoa/WKMenuItemIdentifiersPrivate.h:
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::canHandleContextMenuTranslation const):
* UIProcess/Cocoa/WebViewImpl.h:
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::canHandleContextMenuTranslation const):
Consult the class method to return whether or not the menu item should be present. If not, then we filter this
item out from the list of context menu items in `WebContextMenuProxyMac` below.
(WebKit::WebViewImpl::handleContextMenuTranslation):
Handle the translation action by using `NSPopover` to present an `LTUITranslationViewController`.
* UIProcess/PageClient.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::contextMenuItemSelected):
* UIProcess/WebPageProxy.h:
* UIProcess/mac/PageClientImplMac.h:
* UIProcess/mac/PageClientImplMac.mm:
(WebKit::PageClientImpl::canHandleContextMenuTranslation const):
(WebKit::PageClientImpl::handleContextMenuTranslation):
* UIProcess/mac/WebContextMenuProxyMac.mm:
(WebKit::WebContextMenuProxyMac::getContextMenuFromItems):
Remove the translate menu item if we're already presented in the context of a popover, or if the system does not
support the feature (see `WebViewImpl::canHandleContextMenuTranslation`).
(WebKit::menuItemIdentifier):
2021-03-08 Youenn Fablet <youenn@apple.com>
Allow access to more cmio services from Web/GPU processes
https://bugs.webkit.org/show_bug.cgi?id=222465
Reviewed by Per Arne Vollan.
These new services are used when getting camera access.
Manually tested.
* GPUProcess/mac/com.apple.WebKit.GPUProcess.sb.in:
* WebProcess/com.apple.WebProcess.sb.in:
2021-03-08 Chris Dumez <cdumez@apple.com>
Simplify CallbackAggregator in WebsiteDataStore::fetchDataAndApply()
https://bugs.webkit.org/show_bug.cgi?id=222937
Reviewed by Darin Adler.
Simplify CallbackAggregator in WebsiteDataStore::fetchDataAndApply():
1. Get rid of removePendingCallback() / addPendingCallback() / callIfNeeded()
and instead rely on ref counting to determine when we need to call the
apply function.
2. Replaced removePendingCallback() with addWebsiteData() since it takes
a WebsiteData as an object and make it so that it can be called on
any thread for convenience.
* Shared/WebsiteData/WebsiteData.cpp:
(WebKit::WebsiteData::isolatedCopy const):
(WebKit::WebsiteData::Entry::isolatedCopy const):
* Shared/WebsiteData/WebsiteData.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::fetchDataAndApply):
2021-03-08 Chris Dumez <cdumez@apple.com>
Regression(r273875): Potential over-release in WKRemoteObjectCoder's decodeObjCObject()
https://bugs.webkit.org/show_bug.cgi?id=222954
<rdar://75163359>
Reviewed by Darin Adler.
r273875 added an adoptNS() for the result of `[allocation initWithCoder:decoder]`. This would be
fine in general, except that we call awakeAfterUsingCoder on the result right after. As per the
awakeAfterUsingCoder documentation [1], it may return the receiver or a new object. When it
returns a new object, it takes care of releasing the receiver. This is an issue for us here since
we were holding the receiver in a smart pointer.
[1] https://developer.apple.com/documentation/objectivec/nsobject/1417074-awakeafterusingcoder
* Shared/API/Cocoa/WKRemoteObjectCoder.mm:
(decodeObjCObject):
2021-03-08 Jer Noble <jer.noble@apple.com>
[Cocoa][GPUP] Enable WebM/Vorbis/Opus in the GPU process
https://bugs.webkit.org/show_bug.cgi?id=222926
Reviewed by Eric Carlson.
The first time a GPUProcessProxy is created, and whenever a new WebPageProxy is created,
The GPUProcessProxy will query all existing WebPageGroups and their WebPreferences. If
any group enables "Media in the GPU Process", and also WebM, Vorbis, or Opus, those features
which are enabled will also be enabled inside the GPU process.
* GPUProcess/GPUProcess.cpp:
(WebKit::GPUProcess::setWebMParserEnabled):
(WebKit::GPUProcess::setWebMFormatReaderEnabled):
(WebKit::GPUProcess::setOpusDecoderEnabled):
(WebKit::GPUProcess::setVorbisDecoderEnabled):
* GPUProcess/GPUProcess.h:
* GPUProcess/GPUProcess.messages.in:
* UIProcess/GPU/GPUProcessProxy.cpp:
(WebKit::GPUProcessProxy::getOrCreate):
(WebKit::GPUProcessProxy::updatePreferences):
* UIProcess/GPU/GPUProcessProxy.h:
* UIProcess/WebPageGroup.cpp:
(WebKit::WebPageGroup::forEach):
* UIProcess/WebPageGroup.h:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::createWebPage):
2021-03-08 Brent Fulgham <bfulgham@apple.com>
REGRESSION (r273541): Payloads requiring NSURL in serialization fail
https://bugs.webkit.org/show_bug.cgi?id=222856
<rdar://problem/75036452>
Reviewed by Darin Adler.
In Bug 222145 I converted serialization of CFURLRequest to NSURLRequest. As part of that change,
I used a new encoder (WKSecureCodingURLWrapper) to wrap the NSURL to ensure proper handling of
encoded URLs.
The WKSecureCodingURLWrapper should have been declared as a subclass of NSURL, so that NSSecureCoding
would respect it as a valid way of serializing an NSURL.
This patch corrects that oversight.
* Shared/Cocoa/ArgumentCodersCocoa.mm:
2021-03-08 BJ Burg <bburg@apple.com>
[Cocoa] Changes in WebInspectorUI's appearance don't propagate to .effectiveAppearance while docked
https://bugs.webkit.org/show_bug.cgi?id=222925
<rdar://71201454>
Reviewed by Devin Rousso.
Update the appearance property for the underlying webView in addition to
setting it for the inspector NSWindow (undocked only).
This fix is needed for both local and remote inspector code paths.
* UIProcess/Inspector/mac/RemoteWebInspectorProxyMac.mm:
(WebKit::RemoteWebInspectorProxy::platformSetForcedAppearance):
* UIProcess/Inspector/mac/WebInspectorProxyMac.mm:
(WebKit::WebInspectorProxy::applyForcedAppearance):
2021-03-08 Stephan Szabo <stephan.szabo@sony.com>
[WinCairo] Builds with ENABLE_CONTEXT_MENUS=OFF fail
https://bugs.webkit.org/show_bug.cgi?id=222771
Reviewed by Don Olmstead.
Build fix for turning off context menus adding ifdefs as necessary.
* UIProcess/win/PageClientImpl.h:
* UIProcess/win/WebView.cpp:
2021-03-08 John Wilander <wilander@apple.com>
PCM: Align fraud prevention token naming with issue 73 in the Privacy CG repo
https://bugs.webkit.org/show_bug.cgi?id=222838
<rdar://75115427>
Reviewed by Brent Fulgham.
This patch doesn't change functionality and only aligns naming to this W3C Privacy CG issue:
https://github.com/privacycg/private-click-measurement/issues/73
Naming for PCM fraud prevention is updated like this:
/.well-known/private-click-measurement/sign-secret-token/ changed to
/.well-known/private-click-measurement/sign-unlinkable-token/
/.well-known/private-click-measurement/get-unlinkable-token-public-key/ changed to
/.well-known/private-click-measurement/get-token-public-key/
source_secret_token in requests for token signature changed to source_unlinkable_token
source_unlinkable_token in attribution reports changed to source_secret_token
Variable, member, and function names are being renamed similarly.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::buildPrivateClickMeasurementFromDatabase):
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::setPCMFraudPreventionValuesForTesting):
(WebKit::NetworkProcess::setFraudPreventionValuesForTesting): Deleted.
Renamed setPCMFraudPreventionValuesForTesting() to make it more clear.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/NetworkSession.cpp:
(WebKit::NetworkSession::setPCMFraudPreventionValuesForTesting):
(WebKit::NetworkSession::setFraudPreventionValuesForTesting): Deleted.
Renamed setPCMFraudPreventionValuesForTesting() to make it more clear.
* NetworkProcess/NetworkSession.h:
* NetworkProcess/PrivateClickMeasurementManager.cpp:
(WebKit::PrivateClickMeasurementManager::storeUnattributed):
(WebKit::PrivateClickMeasurementManager::getTokenPublicKey):
(WebKit::PrivateClickMeasurementManager::getSignedUnlinkableToken):
(WebKit::PrivateClickMeasurementManager::setPCMFraudPreventionValuesForTesting):
(WebKit::PrivateClickMeasurementManager::getSignedSecretToken): Deleted.
(WebKit::PrivateClickMeasurementManager::setFraudPreventionValuesForTesting): Deleted.
Renamed setPCMFraudPreventionValuesForTesting() to make it more clear.
* NetworkProcess/PrivateClickMeasurementManager.h:
* UIProcess/API/C/WKPage.cpp:
(WKPageSetPCMFraudPreventionValuesForTesting):
(WKPageSetFraudPreventionValuesForTesting): Deleted.
Renamed WKPageSetPCMFraudPreventionValuesForTesting() to make it more clear.
* UIProcess/API/C/WKPagePrivate.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::setPCMFraudPreventionValuesForTesting):
(WebKit::WebPageProxy::setFraudPreventionValuesForTesting): Deleted.
Renamed setPCMFraudPreventionValuesForTesting() to make it more clear.
* UIProcess/WebPageProxy.h:
2021-03-08 Aditya Keerthi <akeerthi@apple.com>
[iOS] <select multiple> group headers are illegible in dark mode
https://bugs.webkit.org/show_bug.cgi?id=222914
<rdar://problem/75020326>
Reviewed by Tim Horton.
* UIProcess/ios/forms/WKFormSelectPicker.mm:
(-[WKSelectPickerTableViewController tableView:heightForHeaderInSection:]):
Do not specify a textColor for the header label. This ensures the default
semantic label color is used, which automatically adapts when dark mode
is enabled.
2021-03-08 Chris Dumez <cdumez@apple.com>
Simplify CallbackAggregator in WebsiteDataStore::removeData()
https://bugs.webkit.org/show_bug.cgi?id=222837
Reviewed by Geoffrey Garen.
Simplify CallbackAggregator in WebsiteDataStore::removeData(). We don't need an explicit
counter data member or explicit calls to addPendingCallback() / removePendingCallback().
We can simply rely on the ref count of the CallbackAggregator.
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::removeData):
2021-03-08 Kimmo Kinnunen <kkinnunen@apple.com>
WebGL asserts after GPU process times out
https://bugs.webkit.org/show_bug.cgi?id=222546
Reviewed by Wenson Hsieh.
IPC::StreamClientConnection::tryAcquire would assert on
ASSERT(untrustedLimit < (dataSize() - 1));
The untrustedLimit value would be then:
StreamConnectionBuffer::serverOffsetClientIsWaitingTag.
This would be equivalent of assertion added in this commit:
ASSERT(clientLimit != ClientLimit::clientIsWaitingTag);
This would happen in case of timeout, because upon entering
the wait sequence:
1) The client itself would first write the tag value to the
offset variable.
2) Wait and timeout the wait.
3) run the next iteration of the acquire loop.
The next iteration would load the tag value written before,
e.g. serverOffsetClientIsWaitingTag.
This would be then used in clampedLimit() call.
Fix this by taking it into account that the variable might be
ClientLimit::clientIsWaitingTag within the acquire loop.
Possibly fix future similar problems by not treating the value as
size_t, rather treat it as enum class ClientOffset : size_t. This
enables the compiler to enforce that when checking for the various tag
values, the holder must hold the enum class. When intending to interpret
the value as size_t, holder must explicitly convert.
Do the above change for the ServerOffset.
Tests: fast/canvas/webgl/lose-context-on-timeout-async.html
fast/canvas/webgl/lose-context-on-timeout.html
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::releaseGraphicsContextGL):
Do not assert that the release message actually releases the context.
For testing, the induce the timeout by releasing the context prematurely
and treating the subsequent messages as if they were non-legit messages.
That should be just skipped.
(WebKit::GPUConnectionToWebProcess::releaseGraphicsContextGLForTesting):
Add a public function to invoke from "ForTesting" codepaths.
(WebKit::GPUConnectionToWebProcess::dispatchMessage):
(WebKit::GPUConnectionToWebProcess::dispatchSyncMessage):
Skip RemoteGraphicsContextGL messages that do not get routed to any
instance. These messages can be "non-legit", ones that are sent before
the "context was lost" message reached the sender, or ones that are
sent during simulated timeout.
These do not contain asserts as it is expected that these will occur in
above situations.
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/graphics/RemoteGraphicsContextGL.cpp:
(WebKit::RemoteGraphicsContextGL::simulateEventForTesting):
* GPUProcess/graphics/RemoteGraphicsContextGL.h:
* GPUProcess/graphics/RemoteGraphicsContextGL.messages.in:
* GPUProcess/graphics/RemoteGraphicsContextGLFunctionsGenerated.h:
Add the simulateEventForTesting function to simulate the
timeout. Simulate via just removing the context, which causes
left-over messages to skipped.
Currently we do not simulate the timeout with an infinite loop
in GPU process threads as that is not supported.
Currently we do not simulate the timeout with a gpu process
crash since it is not supported to be tested (will cause a
reload).
* Platform/IPC/StreamClientConnection.cpp:
(IPC::StreamClientConnection::StreamClientConnection):
* Platform/IPC/StreamClientConnection.h:
(IPC::StreamClientConnection::sendSync):
(IPC::StreamClientConnection::release):
Change size_t usage to -> ClientOffset, ServerOffset
(IPC::StreamClientConnection::tryAcquire):
Fix the assertion of equivalent of ASSERT(clientLimit != ClientLimit::clientIsWaitingTag);
by reorganizing the acquire loop.
(IPC::StreamClientConnection::toLimit const):
Fix the off-by-one in assertion error
( < dataSize() - 1 vs <= dataSize() - 1)
Rename from clampedLimit() to toLimit(), as the
clamping is confusing as the client trusts the
server.
* Platform/IPC/StreamConnectionBuffer.cpp:
(IPC::StreamConnectionBuffer::StreamConnectionBuffer):
(IPC::StreamConnectionBuffer::operator=):
(IPC::StreamConnectionBuffer::encode const):
* Platform/IPC/StreamConnectionBuffer.h:
(IPC::StreamConnectionBuffer::clientOffset):
(IPC::StreamConnectionBuffer::serverOffset):
(IPC::StreamConnectionBuffer::clientWaitSemaphore):
(IPC::StreamConnectionBuffer::maximumSize):
* Platform/IPC/StreamServerConnection.cpp:
(IPC::StreamServerConnectionBase::tryAquire):
(IPC::StreamServerConnectionBase::release):
Change size_t usage to -> ClientOffset, ServerOffset
(IPC::StreamServerConnectionBase::clampedLimit const):
Fix the off-by-one in assertion error
( < dataSize() - 1 vs <= dataSize() - 1)
* Platform/IPC/StreamServerConnection.h:
(IPC::StreamServerConnectionBase::sharedServerLimit):
(IPC::StreamServerConnectionBase::sharedServerOffset):
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.cpp:
(WebKit::RemoteGraphicsContextGLProxy::synthesizeGLError):
(WebKit::RemoteGraphicsContextGLProxy::getError):
For the cases where the timeout would happen in these two functions,
call the markContextLost() instead of wasLost(). This is to clarify
all the context lost handling to use the same pattern.
(WebKit::RemoteGraphicsContextGLProxy::simulateEventForTesting):
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.h:
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxyFunctionsGenerated.cpp:
2021-03-08 Kimmo Kinnunen <kkinnunen@apple.com>
WebGL context screen change events should work with GPU process
https://bugs.webkit.org/show_bug.cgi?id=219669
<rdar://problem/72373371>
Reviewed by Per Arne Vollan.
Pass display configuration change notification from UI
process to GPU process.
Fixes
fast/canvas/webgl/webglcontextchangedevent.html
fast/canvas/webgl/context-update-on-display-configuration.html
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::displayConfigurationChanged):
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/GPUProcess.cpp:
(WebKit::GPUProcess::displayConfigurationChanged):
* GPUProcess/GPUProcess.h:
* GPUProcess/GPUProcess.messages.in:
* GPUProcess/graphics/RemoteGraphicsContextGL.cpp:
(WebKit::RemoteGraphicsContextGL::displayWasReconfigured):
* GPUProcess/graphics/RemoteGraphicsContextGL.h:
* Platform/IPC/ScopedActiveMessageReceiveQueue.h:
(IPC::ScopedActiveMessageReceiveQueue::operator-> const):
* UIProcess/GPU/GPUProcessProxy.cpp:
(WebKit::GPUProcessProxy::displayConfigurationChanged):
* UIProcess/GPU/GPUProcessProxy.h:
* UIProcess/WebProcessPool.cpp:
(WebKit::displayReconfigurationCallBack):
(WebKit::WebProcessPool::sendDisplayConfigurationChangedMessageForTesting):
2021-03-07 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] maild hits a simulated crash when ENABLE(APP_HIGHLIGHTS) is defined
https://bugs.webkit.org/show_bug.cgi?id=222893
<rdar://problem/74230060>
Reviewed by Tim Horton.
Fix the simulated crash by not initializing the shared callout bar and installing menu items for new app
highlight actions in the case where the web view is unparented. Additionally, avoid this work in the case where
the selection is not a range.
Test: AppHighlights.AvoidForcingCalloutBarInitialization
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView didMoveToWindow]):
Call `-setUpAppHighlightMenusIfNeeded` to ensure that we still add app highlight menu items in the case where
we avoided it previously, due to the web view being unparented. Note that if we're being unparented (i.e.
`self.window` has just become `nil`), we will return early from `-setUpAppHighlightMenusIfNeeded`.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _selectionChanged]):
(-[WKContentView setUpAppHighlightMenusIfNeeded]):
(-[WKContentView setUpAppHighlightMenus]): Deleted.
Rename `setUpAppHighlightMenus` to `setUpAppHighlightMenusIfNeeded`, and declare it in the header so that we
can call into it from `-didMoveToWindow` above.
2021-03-06 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Expand WebAuthn Process's sandbox profile on iOS
https://bugs.webkit.org/show_bug.cgi?id=222858
<rdar://problem/74772012>
Reviewed by Brent Fulgham.
Allows WebAuthn process to use AAA again.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebAuthn.sb:
2021-03-06 Tim Horton <timothy_horton@apple.com>
<model> should create a model-owning compositing layer
https://bugs.webkit.org/show_bug.cgi?id=222798
Reviewed by Simon Fraser.
* Shared/RemoteLayerTree/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::display):
* Shared/RemoteLayerTree/RemoteLayerTreeTransaction.h:
* Shared/RemoteLayerTree/RemoteLayerTreeTransaction.mm:
(WebKit::RemoteLayerTreeTransaction::LayerCreationProperties::encode const):
(WebKit::RemoteLayerTreeTransaction::LayerCreationProperties::decode):
(WebKit::RemoteLayerTreeTransaction::description const):
* SourcesCocoa.txt:
* UIProcess/RemoteLayerTree/RemoteLayerTreeHost.mm:
(WebKit::RemoteLayerTreeHost::makeNode):
* UIProcess/RemoteLayerTree/ios/RemoteLayerTreeHostIOS.mm:
(WebKit::RemoteLayerTreeHost::makeNode):
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebPage/RemoteLayerTree/GraphicsLayerCARemote.cpp:
(WebKit::GraphicsLayerCARemote::createPlatformCALayer):
* WebProcess/WebPage/RemoteLayerTree/GraphicsLayerCARemote.h:
* WebProcess/WebPage/RemoteLayerTree/PlatformCALayerRemote.cpp:
(WebKit::PlatformCALayerRemote::create):
(WebKit::PlatformCALayerRemote::populateCreationProperties):
* WebProcess/WebPage/RemoteLayerTree/PlatformCALayerRemote.h:
* WebProcess/WebPage/RemoteLayerTree/PlatformCALayerRemoteCustom.h:
* WebProcess/WebPage/RemoteLayerTree/PlatformCALayerRemoteCustom.mm:
(WebKit::PlatformCALayerRemoteCustom::populateCreationProperties):
* WebProcess/WebPage/RemoteLayerTree/PlatformCALayerRemoteModelHosting.h: Added.
* WebProcess/WebPage/RemoteLayerTree/PlatformCALayerRemoteModelHosting.mm: Added.
(WebKit::PlatformCALayerRemoteModelHosting::create):
(WebKit::PlatformCALayerRemoteModelHosting::PlatformCALayerRemoteModelHosting):
(WebKit::PlatformCALayerRemoteModelHosting::clone const):
(WebKit::PlatformCALayerRemoteModelHosting::populateCreationProperties):
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeContext.mm:
(WebKit::RemoteLayerTreeContext::layerDidEnterContext):
Create a new PlatformCALayerRemote subclass that knows about Model, and
passes its model along in the layer creation properties.
For now, the special model layer type will just make a plain layer.
Future patches will replace it with model-specific rendering.
2021-03-05 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] Implement additional accessibility support for image overlays
https://bugs.webkit.org/show_bug.cgi?id=222811
<rdar://problem/73203775>
Reviewed by Tim Horton and Devin Rousso.
Adds additional support for accessible image extraction. See below for more details.
* UIProcess/ios/WKContentViewInteraction.h:
(WebKit::SuppressInteractionToken::SuppressInteractionToken): Deleted.
(WebKit::SuppressInteractionToken::~SuppressInteractionToken): Deleted.
Remove `SuppressInteractionToken`, along with its only use as a member on `WKContentView`. This was used to
temporarily suppress the image extraction interaction when editing, but is now no longer necessary, since we're
removing all dependence on this interaction object.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView setIsEditable:]):
Remove `_suppressImageExtractionToken`.
(-[WKContentView hasSelectablePositionAtPoint:]):
(-[WKContentView textInteractionGesture:shouldBeginAtPoint:]):
There's no need to cancel the regular text interaction in favor of the image extraction interaction after this
patch, since the former will be used for both. However, we do need to temporarily suppress regular text
interaction while image extraction is still pending, to avoid prematurely selecting the entire image element.
To do this, we keep track of the element pending image extraction (`_elementPendingImageExtraction`), and avoid
allowing text interactions to begin if we're over this exact element.
(-[WKContentView _singleTapIdentified:]):
We don't need to explicitly cancel the image extraction interaction on single tap anymore; this behavior will
automatically fall out of using default text interactions for image extraction.
(-[WKContentView _didStartProvisionalLoadForMainFrame]):
(-[WKContentView _dragInteraction:prepareForSession:completion:]):
Remove explicit image extraction cancellation here as well, and instead bail from the drag only if we're about
to proceed with the image extraction gesture (which is indicated by the argument to the completion handler).
See comment below.
(-[WKContentView _doAfterPendingImageExtraction:]):
Add an enum flag to the completion handler block here, to determine whether or not we'll want to immediately
proceed with image extraction. The value of this flag will be determined by the hit-test after updating the page
with image extraction results in `WebPage::updateWithImageExtractionResult`.
(-[WKContentView _invokeAllActionsToPerformAfterPendingImageExtraction:]):
(-[WKContentView _contextMenuInteraction:configurationForMenuAtLocation:completion:]):
(-[WKContentView _invokeAllActionsToPerformAfterPendingImageExtraction]): Deleted.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updateWithImageExtractionResult):
Remove a FIXME and implement hit-testing logic, to return whether or not the given location hit-tests to the
image overlay we've just injected.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::setFocusedFrameBeforeSelectingTextAtLocation):
(WebKit::elementPositionInformation):
Avoid extracting image data for a position information request, if the hit-tested node is actually a text node
in an image overlay, instead of an image.
(WebKit::selectionPositionInformation):
(WebKit::WebPage::positionInformation):
Allow text interactions to begin in user-agent shadow roots. This matches behavior on macOS, and was
(presumably) initially disallowed on iOS to avoid activating the editable text interaction
(`UIWKTextInteractionAssistant`), as opposed to block text selection interactions, when long pressing text
inside textareas and text fields. However, since we only use the normal text interaction assistant now for both
non-editable and editable text, this distinction no longer matters.
2021-03-05 Chris Dumez <cdumez@apple.com>
[IPC Hardening] Protect against bad navigationID in WebPageProxy::didDestroyNavigation()
https://bugs.webkit.org/show_bug.cgi?id=222842
<rdar://problem/75119560>
Reviewed by Darin Adler.
* UIProcess/WebNavigationState.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didDestroyNavigation):
2021-03-05 Chris Dumez <cdumez@apple.com>
Potential crash under NotificationPermissionRequestManagerProxy::invalidateRequests()
https://bugs.webkit.org/show_bug.cgi?id=222841
Reviewed by Darin Adler.
Make sure we copy the HashMap before iterating over it and calling deny() on the requests.
* UIProcess/Notifications/NotificationPermissionRequestManagerProxy.cpp:
(WebKit::NotificationPermissionRequestManagerProxy::invalidateRequests):
2021-03-05 Chris Dumez <cdumez@apple.com>
Update ApplicationCacheStorage::originsWithCache() to return a HashSet<SecurityOriginData> instead of a Vector<Ref<SecurityOrigin>>
https://bugs.webkit.org/show_bug.cgi?id=222828
Reviewed by Darin Adler.
Update ApplicationCacheStorage::originsWithCache() to return a HashSet<SecurityOriginData> instead of a Vector<Ref<SecurityOrigin>>.
Using a HashSet makes sure we do not return duplicates (which do occur as per ApplicationCacheStorage implementation) and makes typing
more consistent with other similar storage functions. Using SecurityOriginData is more lightweight and sufficient for our use cases.
It also takes care of a FIXME comment in WebsiteDataStore::fetchDataAndApply about switching to SecurityOriginData.
This patch also updates WebsiteDataStore::fetchDataAndApply() to crossThreadCopy() the securityOriginData before sending them from
the data store queue to the main thread. The previous code as sending SecurityOrigins and WebsiteData to the main thread without
any isolated copy, which was fragile and would not have been thread-safe if ApplicationCacheStorage::originsWithCache() had cached
the origins it returned for example.
Also update WebsiteDataStore::removeData() to make sure we crossThreadCopy() the origins before sending them from the main thread
to the data store queue. This code did not look thread-safe.
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::fetchDataAndApply):
(WebKit::WebsiteDataStore::removeData):
* WebProcess/InjectedBundle/API/c/WKBundlePage.cpp:
(WKBundlePageClearApplicationCacheForOrigin):
(WKBundlePageGetAppCacheUsageForOrigin):
(WKBundlePageCopyOriginsWithApplicationCache):
2021-03-05 Chris Dumez <cdumez@apple.com>
Make HTMLMediaElement::clearMediaCacheForOrigins() take in SecurityOriginData objects
https://bugs.webkit.org/show_bug.cgi?id=222832
Reviewed by Geoffrey Garen.
Make HTMLMediaElement::clearMediaCacheForOrigins() take in SecurityOriginData instead of
SecurityOrigin objects. The call site has SecurityOriginData objects and there was a FIXME
comment about switching clearMediaCacheForOrigins() to SecurityOriginData.
SecurityOriginData also has the benefits of being lighter weight, and it can be sent over
IPC, while being sufficient for our use cases here.
* GPUProcess/media/RemoteMediaPlayerManagerProxy.cpp:
(WebKit::RemoteMediaPlayerManagerProxy::clearMediaCacheForOrigins):
* GPUProcess/media/RemoteMediaPlayerManagerProxy.h:
* GPUProcess/media/RemoteMediaPlayerManagerProxy.messages.in:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::removeData):
* WebProcess/GPU/media/RemoteMediaPlayerManager.cpp:
(WebKit::RemoteMediaPlayerManager::clearMediaCacheForOrigins):
* WebProcess/GPU/media/RemoteMediaPlayerManager.h:
2021-03-05 Chris Dumez <cdumez@apple.com>
Update HTMLMediaElement::originsInMediaCache() to return SecurityOriginData instead of SecurityOrigin objects
https://bugs.webkit.org/show_bug.cgi?id=222820
Reviewed by Darin Adler.
Update HTMLMediaElement::originsInMediaCache() to return SecurityOriginData instead of SecurityOrigin
objects. The caller only needs SecurityOriginData objects and there was a FIXME comment about switching
to SecurityOriginData. SecurityOriginData is beneficial here because it is a smaller object and it
can be easily passed over IPC.
Also update the code at WebsiteDataStore::fetchDataAndApply() to crossThreadCopy() the originDatas
before sending them to the main thread. The previous code was passing both the SecurityOrigins
and the WebsiteDatas from the data store queue to the main thread without isolated copy, which was
fragile. For example, if HTMLMediaElement::originsInMediaCache() was caching the security origins
it returned, the code would not have been thread safe.
* GPUProcess/media/RemoteMediaPlayerManagerProxy.cpp:
(WebKit::RemoteMediaPlayerManagerProxy::originsInMediaCache):
* GPUProcess/media/RemoteMediaPlayerManagerProxy.h:
* GPUProcess/media/RemoteMediaPlayerManagerProxy.messages.in:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::fetchDataAndApply):
* WebProcess/GPU/media/RemoteMediaPlayerManager.cpp:
(WebKit::RemoteMediaPlayerManager::originsInMediaCache):
* WebProcess/GPU/media/RemoteMediaPlayerManager.h:
2021-03-05 Chris Dumez <cdumez@apple.com>
Use WTF::DestructionThread::MainRunLoop for CallbackAggregators in WebsiteDataStore
https://bugs.webkit.org/show_bug.cgi?id=222831
Reviewed by Geoffrey Garen.
Use WTF::DestructionThread::MainRunLoop for CallbackAggregators in WebsiteDataStore to make sure
the CallbackAggregators get destroyed on the main thread. Previously, the code was letting the
CallbackAggregators get destroyed on any thread and then dispatching to the main thread in their
destructor their WebsiteDataStore data member to make sure it gets destroyed on the main thread.
It is simpler and safer to have the CallbackAggregators get destroyed on the main thread.
Also add threading assertions whenever the pendingCallbacks is used since this counter is not
atomic.
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::removeData):
2021-03-05 Chris Dumez <cdumez@apple.com>
Fix potential thread safety issue in WebsiteDataStore::fetchDataAndApply()
https://bugs.webkit.org/show_bug.cgi?id=222830
Reviewed by Geoffrey Garen.
The previous code was passing SecurityOriginData objects from the data store queue to the main
thread without isolated copy. This code would not be thread-safe if the function we got the
origins from would keep a copy of the origins (e.g. caching). To make this more robust from a
thread-safety standpoint, we now crossThreadCopy() to origins before passing them to the main
thread.
I also updated these to construct the WebsiteData on the data store queue instead of the main
thread for consistency with other cases inside WebsiteDataStore::fetchDataAndApply().
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::fetchDataAndApply):
2021-03-05 Michael Catanzaro <mcatanzaro@gnome.org>
[GTK] Clean up GTK-specific text checker stuff
https://bugs.webkit.org/show_bug.cgi?id=222818
Reviewed by Don Olmstead.
This moves the GTK-specific classes WebTextChecker and WebTextCheckerClient into
GTK-specific directories.
This also moves most of WKTextChecker, except for WKTextCheckerSetTestingMode, which is
cross-platform. This is not quite how we normally do it -- normally we try to have one
cross-platform header with platform-specific conditionals and only separate out platform-
specific source files -- but in this case the header is almost entirely GTK-specific, so it
makes sense to do it this way.
This also exposes WKTextCheckerSetContinuousSpellCheckingEnabled and uses it in
UIScriptControllerGtk, which is needed for bug #181916.
This code is all pretty messy, especially the WKTextCheckerClient (which actually has its
own WKTextCheckerSetContinuousSpellCheckingEnabled!), but now it's a *platform-specific*
mess, which is a small improvement.
* SourcesGTK.txt:
* UIProcess/API/C/WKTextChecker.cpp:
(WKTextCheckerSetTestingMode):
(WKTextCheckerSetClient): Deleted.
(WKTextCheckerContinuousSpellCheckingEnabledStateChanged): Deleted.
(WKTextCheckerGrammarCheckingEnabledStateChanged): Deleted.
(WKTextCheckerCheckSpelling): Deleted.
(WKTextCheckerChangeSpellingToWord): Deleted.
* UIProcess/API/C/WKTextChecker.h:
* UIProcess/API/C/gtk/WKTextCheckerGtk.cpp:
(WKTextCheckerSetClient):
(WKTextCheckerContinuousSpellCheckingEnabledStateChanged):
(WKTextCheckerGrammarCheckingEnabledStateChanged):
(WKTextCheckerCheckSpelling):
(WKTextCheckerChangeSpellingToWord):
(WKTextCheckerSetContinuousSpellCheckingEnabled):
* UIProcess/API/C/gtk/WKTextCheckerGtk.h:
* UIProcess/gtk/WebTextChecker.cpp: Renamed from Source/WebKit/UIProcess/WebTextChecker.cpp.
(WebKit::WebTextChecker::singleton):
(WebKit::WebTextChecker::WebTextChecker):
(WebKit::WebTextChecker::setClient):
(WebKit::updateStateForAllContexts):
(WebKit::WebTextChecker::continuousSpellCheckingEnabledStateChanged):
(WebKit::WebTextChecker::grammarCheckingEnabledStateChanged):
(WebKit::WebTextChecker::checkSpelling):
(WebKit::WebTextChecker::changeSpellingToWord):
* UIProcess/gtk/WebTextChecker.h: Renamed from Source/WebKit/UIProcess/WebTextChecker.h.
(WebKit::WebTextChecker::client):
* UIProcess/gtk/WebTextCheckerClient.cpp: Renamed from Source/WebKit/UIProcess/WebTextCheckerClient.cpp.
(WebKit::WebTextCheckerClient::continuousSpellCheckingAllowed):
(WebKit::WebTextCheckerClient::continuousSpellCheckingEnabled):
(WebKit::WebTextCheckerClient::setContinuousSpellCheckingEnabled):
(WebKit::WebTextCheckerClient::grammarCheckingEnabled):
(WebKit::WebTextCheckerClient::setGrammarCheckingEnabled):
(WebKit::WebTextCheckerClient::uniqueSpellDocumentTag):
(WebKit::WebTextCheckerClient::closeSpellDocumentWithTag):
(WebKit::WebTextCheckerClient::checkSpellingOfString):
(WebKit::WebTextCheckerClient::checkGrammarOfString):
(WebKit::WebTextCheckerClient::spellingUIIsShowing):
(WebKit::WebTextCheckerClient::toggleSpellingUIIsShowing):
(WebKit::WebTextCheckerClient::updateSpellingUIWithMisspelledWord):
(WebKit::WebTextCheckerClient::updateSpellingUIWithGrammarString):
(WebKit::WebTextCheckerClient::guessesForWord):
(WebKit::WebTextCheckerClient::learnWord):
(WebKit::WebTextCheckerClient::ignoreWord):
* UIProcess/gtk/WebTextCheckerClient.h: Renamed from Source/WebKit/UIProcess/WebTextCheckerClient.h.
2021-03-05 Chris Dumez <cdumez@apple.com>
Regression(r268097): WKWebView.URL is nil in the processDidTerminate delegate
https://bugs.webkit.org/show_bug.cgi?id=222809
Reviewed by Michael Catanzaro.
There was a PageLoadState::Transaction in resetStateAfterProcessTermination() that
was previously making sure we would not clear the WebView's URL before calling the
processDidTerminate client delegate. Now that we call the client delegate in a
separate function (WebPageProxy::dispatchProcessDidTerminate), we need to make move
the PageLoadState::Transaction to the caller in
WebProcessProxy::processDidTerminateOrFailedToLaunch(), so that its scope covers
both resetStateAfterProcessTermination() & dispatchProcessDidTerminate() calls.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::resetStateAfterProcessTermination):
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::processDidTerminateOrFailedToLaunch):
2021-03-05 Chris Dumez <cdumez@apple.com>
Fix potential thread-safety issue in WebsiteDataStore::fetchDataAndApply()
https://bugs.webkit.org/show_bug.cgi?id=222807
Reviewed by Geoffrey Garen.
Fix an issue when WebsiteDataStore::fetchDataAndApply() failed to create an isolated
copy of the WebsiteDataRecords before passing them to its internal dispatch queue.
Also add more threading assertions to help catch issues since we have crashes in this
area of the code.
* UIProcess/WebsiteData/WebsiteDataRecord.cpp:
(WebKit::WebsiteDataRecord::isolatedCopy const):
* UIProcess/WebsiteData/WebsiteDataRecord.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::fetchDataAndApply):
2021-03-05 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed, fix the build with recent SDKs.
* UIProcess/ios/fullscreen/WKFullScreenViewController.mm:
(-[WKFullScreenViewController loadView]):
2021-03-05 Eric Carlson <eric.carlson@apple.com>
[GPU Process] Remote control command should only go to the current NowPlaying session
https://bugs.webkit.org/show_bug.cgi?id=222759
<rdar://problem/75053393>
Reviewed by Jer Noble.
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::clearNowPlayingInfo): Track being the active
NowPlaying process.
(WebKit::GPUConnectionToWebProcess::setNowPlayingInfo): Ditto.
(WebKit::GPUConnectionToWebProcess::updateSupportedRemoteCommands): Pass supported
commands and seeking to NowPlayingManager so it can pass them to the remote command
listener.
(WebKit::GPUConnectionToWebProcess::createRemoteCommandListener): We only ever need
one listener.
(WebKit::GPUConnectionToWebProcess::releaseRemoteCommandListener): Ditto.
(WebKit::GPUConnectionToWebProcess::dispatchMessage): Dispatch RemoteRemoteCommandListenerProxy
messages.
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/media/RemoteRemoteCommandListenerProxy.cpp:
(WebKit::RemoteRemoteCommandListenerProxy::RemoteRemoteCommandListenerProxy): Don't
create a remote command listener, we only need one for the GPU process.
(WebKit::RemoteRemoteCommandListenerProxy::updateSupportedCommands): New.
(WebKit::RemoteRemoteCommandListenerProxy::didReceiveRemoteControlCommand): Deleted.
* GPUProcess/media/RemoteRemoteCommandListenerProxy.h:
(WebKit::RemoteRemoteCommandListenerProxy::create):
(WebKit::RemoteRemoteCommandListenerProxy::supportsSeeking const):
(WebKit::RemoteRemoteCommandListenerProxy::supportedCommands const):
(WebKit::RemoteRemoteCommandListenerProxy::identifier const):
* GPUProcess/media/RemoteRemoteCommandListenerProxy.messages.in:
* WebProcess/GPU/media/RemoteRemoteCommandListener.cpp:
(WebKit::RemoteRemoteCommandListener::RemoteRemoteCommandListener): Use m_process
instead of calling WebProcess::singleton().
(WebKit::RemoteRemoteCommandListener::~RemoteRemoteCommandListener): Ditto.
(WebKit::RemoteRemoteCommandListener::didReceiveRemoteControlCommand): Use client(),
m_client is now private.
(WebKit::RemoteRemoteCommandListener::updateSupportedCommands):
* WebProcess/GPU/media/RemoteRemoteCommandListener.h:
2021-03-05 Chris Dumez <cdumez@apple.com>
Reduce use of CFRetain() / CFRelease() / CFAutoRelease() in WebKit
https://bugs.webkit.org/show_bug.cgi?id=222760
Reviewed by Darin Adler.
Reduce use of CFRetain() / CFRelease() / CFAutoRelease() in WebKit by using RetainPtr<>.
* Platform/mac/StringUtilities.mm:
(WebKit::nsStringFromWebCoreString):
(WebKit::formattedPhoneNumberString):
* PluginProcess/mac/PluginProcessMac.mm:
(WebKit::initializeCocoaOverrides):
* Shared/mac/PasteboardTypes.mm:
(WebKit::PasteboardTypes::forEditing):
(WebKit::PasteboardTypes::forURL):
(WebKit::PasteboardTypes::forImages):
(WebKit::PasteboardTypes::forImagesWithArchive):
(WebKit::PasteboardTypes::forSelection):
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::acceptsFirstMouse):
(WebKit::WebViewImpl::shouldDelayWindowOrderingForEvent):
(WebKit::WebViewImpl::validAttributesForMarkedText):
(WebKit::WebViewImpl::performKeyEquivalent):
* UIProcess/Inspector/ios/WKInspectorHighlightView.mm:
(layerPathWithHole):
(layerPath):
* UIProcess/_WKTouchEventGenerator.mm:
(-[_WKTouchEventGenerator dealloc]):
(-[_WKTouchEventGenerator _sendHIDEvent:]):
* UIProcess/ios/fullscreen/WKFullScreenWindowControllerIOS.mm:
(-[WKFullScreenWindowController _EVOrganizationName]):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::webProcessLoaderAccessibilityBundlePath):
2021-03-05 Michael Catanzaro <mcatanzaro@gnome.org>
[GTK] Bubblewrap sandbox should not break X11 forwarding
https://bugs.webkit.org/show_bug.cgi?id=221990
Reviewed by Carlos Alberto Lopez Perez.
If $DISPLAY points to a TCP socket, or a Unix socket on a different host, then we cannot
isolate the web process from the network and must grant access to the host network
namespace.
Also, clean up some related code by adding PLATFORM(X11) guards where appropriate and
removing a redundant display type check.
* UIProcess/Launcher/glib/BubblewrapLauncher.cpp:
(WebKit::bindWayland):
(WebKit::shouldUnshareNetwork):
(WebKit::bubblewrapSpawn):
2021-03-05 Guido Günther <agx@sigxcpu.org>
Add support for gstreamer's h264 stateless codecs
https://bugs.webkit.org/show_bug.cgi?id=222787
Reviewed by Philippe Normand.
Manually tested, there is currently no automatic tests for it
* UIProcess/Launcher/glib/BubblewrapLauncher.cpp:
(WebKit::bindV4l):
2021-03-05 Said Abou-Hallawa <said@apple.com>
[GPU Process] RemoteImageBuffer recording GraphicsContext should report its backend RenderingMode
https://bugs.webkit.org/show_bug.cgi?id=222772
Reviewed by Simon Fraser.
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
2021-03-04 Alex Christensen <achristensen@webkit.org>
Add internal preference to disable HTTPS upgrade
https://bugs.webkit.org/show_bug.cgi?id=222778
Reviewed by Simon Fraser.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
2021-03-04 Alex Christensen <achristensen@webkit.org>
Fix clean build after r273946
https://bugs.webkit.org/show_bug.cgi?id=222706
* DerivedSources-output.xcfilelist:
* WebKit.xcodeproj/project.pbxproj:
2021-03-04 Alex Christensen <achristensen@webkit.org>
Remove the HTTPSUpgradeEnabled experimental feature
https://bugs.webkit.org/show_bug.cgi?id=222706
Reviewed by Simon Fraser.
* DerivedSources.make:
* Scripts/generate-https-upgrade-database.sh: Removed.
* Shared/HTTPSUpgrade/HTTPSUpgradeList.txt: Removed.
2021-03-04 Chris Dumez <cdumez@apple.com>
NetworkProcessProxy::deleteWebsiteDataInUIProcessForRegistrableDomains() can take domains as an rvalue reference
https://bugs.webkit.org/show_bug.cgi?id=222768
Reviewed by Geoffrey Garen.
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::deleteWebsiteDataInUIProcessForRegistrableDomains):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::fetchDataForRegistrableDomains):
* UIProcess/WebsiteData/WebsiteDataStore.h:
2021-03-04 Devin Rousso <drousso@apple.com>
[Payment Request] remove some unnecessary WebKitAdditions hooks
https://bugs.webkit.org/show_bug.cgi?id=222765
<rdar://problem/75060894>
Reviewed by Andy Estes.
* Shared/ApplePay/cocoa/WebPaymentCoordinatorProxyCocoa.mm:
2021-03-04 Sihui Liu <sihui_liu@appe.com>
Validate documentState of FrameState when setting and getting it
https://bugs.webkit.org/show_bug.cgi?id=222587
Reviewed by Geoffrey Garen.
In rdar://48634553, strings of documentState can be invalid when they are encoded in encodeFrameStateNode
in UI process. To get a better idea of when the strings become invalid, add checks for documentState when
getting and setting it.
No test as no behavior change.
* Shared/SessionState.cpp:
(WebKit::FrameState::encode const):
(WebKit::FrameState::decode):
(WebKit::FrameState::validateDocumentState const):
(WebKit::FrameState::setDocumentState):
* Shared/SessionState.h:
(WebKit::FrameState::FrameState):
(WebKit::FrameState::~FrameState):
(WebKit::FrameState::documentState const):
* Shared/WebBackForwardListItem.cpp:
(WebKit::WebBackForwardListItem::create):
(WebKit::WebBackForwardListItem::~WebBackForwardListItem):
* UIProcess/API/glib/WebKitWebViewSessionState.cpp:
(encodeFrameState):
(decodeFrameState):
* UIProcess/Cocoa/SessionStateCoding.h:
* UIProcess/LegacySessionStateCoding.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::sessionState const):
* UIProcess/mac/LegacySessionStateCoding.cpp:
(WebKit::encodeFrameStateNode):
(WebKit::decodeBackForwardTreeNode):
* WebProcess/WebCoreSupport/SessionStateConversion.cpp:
(WebKit::toFrameState):
(WebKit::applyFrameState):
2021-03-04 Tim Horton <timothy_horton@apple.com>
Fix the build after r273904
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::setPreferenceValue):
auto's evilness rears its head again. -mutableCopy returns `id`, so
we infer RetainPtr<id>, so depending on what headers are included
we get ambiguous selector errors. De-auto to fix.
2021-03-04 Chris Dumez <cdumez@apple.com>
Clean up API::ResourceLoadStatisticsThirdParty / API::ResourceLoadStatisticsFirstParty
https://bugs.webkit.org/show_bug.cgi?id=222701
Reviewed by Geoffrey Garen.
Clean up API::ResourceLoadStatisticsThirdParty / API::ResourceLoadStatisticsFirstParty:
1. Make constructors private since there is a create() factory function.
2. Mark constructor as explicit since they take a single parameter.
3. Update ResourceLoadStatisticsFirstParty constructor to take parameter as a const
reference instead of passing it by value.
4. Add assertions to make sure those are constructed and destroyed on the main thread.
* UIProcess/API/APIResourceLoadStatisticsFirstParty.h:
* UIProcess/API/APIResourceLoadStatisticsThirdParty.h:
2021-03-04 Alex Christensen <achristensen@webkit.org>
Remove the HTTPSUpgradeEnabled experimental feature
https://bugs.webkit.org/show_bug.cgi?id=222706
Reviewed by Simon Fraser.
* NetworkProcess/NetworkHTTPSUpgradeChecker.cpp: Removed.
* NetworkProcess/NetworkHTTPSUpgradeChecker.h: Removed.
* NetworkProcess/NetworkLoadChecker.cpp:
(WebKit::NetworkLoadChecker::NetworkLoadChecker):
(WebKit::NetworkLoadChecker::checkRequest):
(WebKit::NetworkLoadChecker::applyHTTPSUpgradeIfNeeded const): Deleted.
* NetworkProcess/NetworkLoadChecker.h:
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkResourceLoadParameters.cpp:
(WebKit::NetworkResourceLoadParameters::encode const):
(WebKit::NetworkResourceLoadParameters::decode):
* NetworkProcess/NetworkResourceLoadParameters.h:
* NetworkProcess/NetworkResourceLoader.cpp:
* NetworkProcess/cocoa/NetworkProcessCocoa.mm:
(WebKit::NetworkProcess::networkHTTPSUpgradeChecker): Deleted.
* Sources.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/Network/WebLoaderStrategy.cpp:
(WebKit::addParametersShared):
2021-03-04 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] Add support for a new image-extraction-related responder action
https://bugs.webkit.org/show_bug.cgi?id=222743
<rdar://problem/75001149>
Reviewed by Devin Rousso.
Add support for the new action, such that we return `YES` if and only if we're in an editable context, and also
such that we return a non-null target for the new action.
* UIProcess/ios/WKContentViewInteraction.h:
Add a fallback definition for `FOR_EACH_ADDITIONAL_WKCONTENTVIEW_ACTION`, which can be used by WebKitAdditions
to inject additional responder actions when generating selector forwarding code between `WKContentView` and
`WKWebView`.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView canPerformActionForWebView:withSender:]):
2021-03-04 Chris Dumez <cdumez@apple.com>
Unreviewed build fix after r272629.
* UIProcess/Cocoa/WKBlankOverlayView.mm:
2021-03-04 Chris Dumez <cdumez@apple.com>
[macOS][WK2] Changing the system language does not update navigator.language
https://bugs.webkit.org/show_bug.cgi?id=222619
Reviewed by Per Arne Vollan.
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceMain.mm:
(WebKit::setAppleLanguagesPreference):
Fix a memory leak for newArguments.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
If CFPREFS_DIRECT_MODE is not enabled, we need to listen for AppleLanguagePreferencesChangedNotification
inside the WebProcess. This used to happen implicitly inside WTF::platformUserPreferredLanguages() but
it is now explicit since we don't want/need it when using CFPREFS_DIRECT_MODE.
(WebKit::setPreferenceValue):
- If preference is AppleLanguages, set it for the volatile domain to match what we do in XPCServiceMain.mm.
This is needed because the preference in the volatile domain seems to take precedence.
- Call WTF::languageDidChange() when the AppleLanguages preference gets updated so that
language change listeners get notified of the language change (e.g. we fire a languagechange
event at the Window).
2021-03-04 Chris Dumez <cdumez@apple.com>
Set ownership of IOSurfaces from the GPUProcess instead of the WebProcess
https://bugs.webkit.org/show_bug.cgi?id=222391
<rdar://problem/74748353>
Reviewed by Simon Fraser.
Replace use of IOSurfaceSetOwnership() SPI with IOSurfaceSetOwnershipIdentity().
Both do the same thing but IOSurfaceSetOwnershipIdentity() only requires an identity token
for the new owner (instead of a control port). As a result, IOSurfaceSetOwnershipIdentity()
requires a lot less priviledges and can now be used directly in the GPUProcess instead of
the WebProcess.
* GPUProcess/graphics/RemoteGraphicsContextGLCocoa.cpp:
(WebKit::RemoteGraphicsContextGLCocoa::prepareForDisplay):
* GPUProcess/graphics/RemoteImageBuffer.h:
(WebKit::RemoteImageBuffer::setProcessOwnership):
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::createImageBuffer):
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.cpp:
(WebKit::RemoteGraphicsContextGLProxy::prepareForDisplay):
* WebProcess/GPU/graphics/cocoa/ImageBufferShareableMappedIOSurfaceBackend.cpp:
(WebKit::ImageBufferShareableMappedIOSurfaceBackend::create):
(WebKit::ImageBufferShareableMappedIOSurfaceBackend::setProcessOwnership):
* WebProcess/GPU/graphics/cocoa/ImageBufferShareableMappedIOSurfaceBackend.h:
2021-03-04 Youenn Fablet <youenn@apple.com>
REGRESSION (r273732): ASSERTION FAILED: Completion handler should always be called under WebKit::RemoteRealtimeMediaSourceProxy::~RemoteRealtimeMediaSourceProxy
https://bugs.webkit.org/show_bug.cgi?id=222691
<rdar://problem/75006211>
Reviewed by Eric Carlson.
Make sure to return a ApplyConstraints completed message in all cases.
Call completion handler when source gets destroyed.
Covered by existing tests.
* UIProcess/Cocoa/UserMediaCaptureManagerProxy.cpp:
(WebKit::UserMediaCaptureManagerProxy::applyConstraints):
* WebProcess/cocoa/RemoteRealtimeMediaSourceProxy.cpp:
(WebKit::RemoteRealtimeMediaSourceProxy::~RemoteRealtimeMediaSourceProxy):
* WebProcess/cocoa/RemoteRealtimeMediaSourceProxy.h:
* WebProcess/cocoa/RemoteRealtimeVideoSource.cpp:
(WebKit::RemoteRealtimeVideoSource::gpuProcessConnectionDidClose):
2021-03-04 Youenn Fablet <youenn@apple.com>
Update camera and microphone capture state control WKWebView API
https://bugs.webkit.org/show_bug.cgi?id=222166
Reviewed by Eric Carlson.
Introduce KVO states for microphone and camera capture states.
Add two APIs to set microphone and camera state.
This allows to mute/unmute or stop microphone or camera state.
We needed to introduce muting of either microphone or camera since this was tied together before the patch.
These two APIs take a completion handler.
* UIProcess/API/Cocoa/WKUIDelegatePrivate.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _cameraCaptureState]):
(-[WKWebView _microphoneCaptureState]):
(-[WKWebView setMicrophoneCaptureState:completionHandler:]):
(-[WKWebView setCameraCaptureState:completionHandler:]):
(-[WKWebView _stopMediaCapture]):
(-[WKWebView _mediaCaptureState]):
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/Cocoa/MediaUtilities.h:
* UIProcess/Cocoa/MediaUtilities.mm:
(WebKit::toWKMediaCaptureStateDeprecated):
(WebKit::toWKMediaCaptureState): Deleted.
* UIProcess/Cocoa/PageClientImplCocoa.h:
* UIProcess/Cocoa/PageClientImplCocoa.mm:
(WebKit::PageClientImplCocoa::microphoneCaptureChanged):
(WebKit::PageClientImplCocoa::cameraCaptureChanged):
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::UIClient::mediaCaptureStateDidChange):
* UIProcess/PageClient.h:
(WebKit::PageClient::microphoneCaptureChanged):
(WebKit::PageClient::cameraCaptureChanged):
* UIProcess/UserMediaPermissionRequestManagerProxy.cpp:
(WebKit::UserMediaPermissionRequestManagerProxy::~UserMediaPermissionRequestManagerProxy):
(WebKit::UserMediaPermissionRequestManagerProxy::stopCapture):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::setMuted):
(WebKit::WebPageProxy::stopMediaCapture):
(WebKit::WebPageProxy::updateReportedMediaCaptureState):
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::setMuted):
(WebKit::WebPageProxy::stopMediaCapture):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_lastNavigationWasAppBound):
(WebKit::WebPage::setMuted):
(WebKit::WebPage::stopMediaCapture):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2021-03-04 Rob Buis <rbuis@igalia.com>
Null check platformData when decoding
https://bugs.webkit.org/show_bug.cgi?id=221442
Reviewed by Ryosuke Niwa.
Null check platformData when decoding.
Test: ipc/pasteboard-write-custom-data.html
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<PasteboardCustomData::Entry>::decode):
2021-03-04 Kimmo Kinnunen <kkinnunen@apple.com>
IPC stream feature should use client/server naming instead of sender/receiver
https://bugs.webkit.org/show_bug.cgi?id=222592
Reviewed by Antti Koivisto.
Attempt to simplify already complicated implementation of the stream
implementation. Uses of "sender" and "receiver" were a leftover from
early phase of the implementation.
Rename all remaining instances of "sender" with "client".
Rename all remaining instances of "receiver" with "server".
No new tests, a refactor.
* Platform/IPC/StreamClientConnection.cpp:
(IPC::StreamClientConnection::StreamClientConnection):
(IPC::StreamClientConnection::setWakeUpSemaphore):
(IPC::StreamClientConnection::wakeUpServer):
* Platform/IPC/StreamClientConnection.h:
(IPC::StreamClientConnection::send):
(IPC::StreamClientConnection::trySendDestinationIDIfNeeded):
(IPC::StreamClientConnection::tryAcquire):
(IPC::StreamClientConnection::release):
* Platform/IPC/StreamConnectionBuffer.cpp:
(IPC::StreamConnectionBuffer::StreamConnectionBuffer):
(IPC::StreamConnectionBuffer::operator=):
(IPC::StreamConnectionBuffer::encode const):
* Platform/IPC/StreamConnectionBuffer.h:
(IPC::StreamConnectionBuffer::clientOffset):
(IPC::StreamConnectionBuffer::serverOffset):
(IPC::StreamConnectionBuffer::clientWaitSemaphore):
(IPC::StreamConnectionBuffer::maximumSize):
* Platform/IPC/StreamServerConnection.cpp:
(IPC::StreamServerConnectionBase::tryAquire):
(IPC::StreamServerConnectionBase::release):
* Platform/IPC/StreamServerConnection.h:
(IPC::StreamServerConnectionBase::sharedClientOffset):
(IPC::StreamServerConnectionBase::sharedServerOffset):
2021-03-04 Kimmo Kinnunen <kkinnunen@apple.com>
Adding new test conditions for WebGL should be simpler
https://bugs.webkit.org/show_bug.cgi?id=222593
Reviewed by Kenneth Russell.
Replace simulateContextChanged and
setFailNextGPUStatusCheck with simulateEventForTesting.
* GPUProcess/graphics/RemoteGraphicsContextGL.cpp:
(WebKit::RemoteGraphicsContextGL::simulateEventForTesting):
* GPUProcess/graphics/RemoteGraphicsContextGL.h:
* GPUProcess/graphics/RemoteGraphicsContextGL.messages.in:
* GPUProcess/graphics/RemoteGraphicsContextGLFunctionsGenerated.h:
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.cpp:
(WebKit::RemoteGraphicsContextGLProxy::simulateEventForTesting):
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.h:
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxyFunctionsGenerated.cpp:
2021-03-03 Alex Christensen <achristensen@webkit.org>
WKRemoteObjectCoder should be able to handle NSErrors from TLS failures
https://bugs.webkit.org/show_bug.cgi?id=222401
<rdar://problem/72103865>
Reviewed by Chris Dumez.
NSErrors from TLS failures contain values like a SecTrustRef or a SecCertificateRef,
which are not ObjC objects, and they don't like it when you call encodeWithCoder: on them.
Until r273141 it would crash when we do, but even after that we just decode a nil NSError.
Add a special case like we did in encodeNSError to successfully encode and decode these errors.
* Shared/API/Cocoa/WKRemoteObjectCoder.mm:
(decodeObjCObject):
(transformCertificatesToData):
(transformTrustToData):
(encodeError):
(transformDataToCertificates):
(transformDataToTrust):
(decodeError):
(encodeObject):
(decodeObject):
2021-03-03 Alex Christensen <achristensen@webkit.org>
Limit HashTable entry size to 500 bytes
https://bugs.webkit.org/show_bug.cgi?id=222658
Reviewed by Yusuke Suzuki.
* NetworkProcess/NetworkConnectionToWebProcess.h:
(WebKit::NetworkConnectionToWebProcess::getNetworkLoadInformationResponse):
(WebKit::NetworkConnectionToWebProcess::getNetworkLoadIntermediateInformation):
(WebKit::NetworkConnectionToWebProcess::takeNetworkLoadInformationMetrics):
(WebKit::NetworkConnectionToWebProcess::addNetworkLoadInformation):
(WebKit::NetworkConnectionToWebProcess::addNetworkLoadInformationMetrics):
* NetworkProcess/cache/CacheStorageEngineCaches.cpp:
(WebKit::CacheStorage::Caches::writeRecord):
(WebKit::CacheStorage::Caches::readRecord):
* NetworkProcess/cache/CacheStorageEngineCaches.h:
2021-03-03 Chris Dumez <cdumez@apple.com>
Unreviewed, reverting r273851.
Caused some tests failures on macOS Big Sur
Reverted changeset:
"[macOS][WK2] Changing the system language does not update
navigator.language"
https://bugs.webkit.org/show_bug.cgi?id=222619
https://commits.webkit.org/r273851
2021-03-03 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed, reverting r273832.
26 inspector tests failing a new assert added with this change
Reverted changeset:
"Limit HashTable entry size to 500 bytes"
https://bugs.webkit.org/show_bug.cgi?id=222658
https://commits.webkit.org/r273832
2021-03-03 Chris Dumez <cdumez@apple.com>
[macOS][WK2] Changing the system language does not update navigator.language
https://bugs.webkit.org/show_bug.cgi?id=222619
Reviewed by Per Arne Vollan.
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceMain.mm:
(WebKit::setAppleLanguagesPreference):
- When ENABLE(CFPREFS_DIRECT_MODE), we don't need to set the AppleLanguages
in a volatile domain. This does not change behavior but I made this change
so that I don't need to set AppleLanguages in the volatile domain in
WebProcessCocoa when the preference changes.
- Also fix a memory leak for newArguments.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
If CFPREFS_DIRECT_MODE is not enabled, we need to listen for AppleLanguagePreferencesChangedNotification
inside the WebProcess. This used to happen implicitly inside WTF::platformUserPreferredLanguages() but
it is now explicit since we don't want/need it when using CFPREFS_DIRECT_MODE.
(WebKit::setPreferenceValue):
- Call CFPreferencesSetAppValue() instead of CFPreferencesSetValue(). Without this change
trying to set AppleLanguages here would have no effect and CFLocaleCopyPreferredLanguages()
would keep returning outdated languages.
- Call WTF::languageDidChange() when the AppleLanguages preference gets updated so that
language change listeners get notified of the language change (e.g. we fire a languagechange
event at the Window).
2021-03-03 Chris Dumez <cdumez@apple.com>
Use adoptNS() right away after calling [obj copy] / [obj mutableCopy]
https://bugs.webkit.org/show_bug.cgi?id=222634
Reviewed by Darin Adler.
Use adoptNS() right away after calling [obj copy] / [obj mutableCopy] to minimize the chance of leaks.
* Shared/ApplePay/ApplePayPaymentSetupFeatures.mm:
(WebKit::PaymentSetupFeatures::decode):
* Shared/ApplePay/PaymentSetupConfiguration.mm:
(WebKit::PaymentSetupConfiguration::decode):
* Shared/ApplePay/cocoa/PaymentSetupConfiguration.mm:
(WebKitAdditions::PaymentSetupConfiguration::decode):
* UIProcess/Launcher/mac/ProcessLauncherMac.mm:
(WebKit::systemDirectoryPath):
* UIProcess/mac/ServicesController.mm:
(WebKit::ServicesController::refreshExistingServices):
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::temporaryPDFDirectoryPath):
2021-03-03 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r273727.
https://bugs.webkit.org/show_bug.cgi?id=222689
Landed broken test
Reverted changeset:
"Null check platformData when decoding"
https://bugs.webkit.org/show_bug.cgi?id=221442
https://trac.webkit.org/changeset/273727
2021-03-03 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r273814.
https://bugs.webkit.org/show_bug.cgi?id=222676
Unresolved types in JavaScriptCore-4.0.gir
Reverted changeset:
"[CMake] JavaScriptCore GLib headers should be copies"
https://bugs.webkit.org/show_bug.cgi?id=222625
https://trac.webkit.org/changeset/273814
2021-03-03 Tim Horton <timothy_horton@apple.com>
Modernize WebKit2 PDFKit softlinking
https://bugs.webkit.org/show_bug.cgi?id=222643
Reviewed by Alex Christensen.
WebKit2 has long had an unusual way of softlinking PDFKit, using NSBundle.
It has caused much trouble in the past, because it means that running
WebKit against a custom PDFKit requires either installing it into
the system, or changing the hardcoded path.
Instead of this, just use the normal softlinking macros that the rest
of WebKit uses.
* Shared/Cocoa/PDFKitSoftLink.h: Added.
* Shared/Cocoa/PDFKitSoftLink.mm: Added.
* Shared/mac/PDFKitImports.h: Removed.
* Shared/mac/PDFKitImports.mm: Removed.
* SourcesCocoa.txt:
* UIProcess/mac/WKPrintingView.mm:
(-[WKPrintingView _drawPDFDocument:page:atPoint:]):
(-[WKPrintingView drawRect:]):
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/Plugins/PDF/PDFPlugin.h:
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::PDFPlugin):
(WebKit::PDFPlugin::receivedNonLinearizedPDFSentinel):
(WebKit::PDFPlugin::threadEntry):
(WebKit::PDFPlugin::documentDataDidFinishLoading):
(WebKit::PDFPlugin::setActiveAnnotation):
(WebKit::PDFPlugin::nextMatchForString):
(WebKit:: const):
* WebProcess/Plugins/PDF/PDFPluginAnnotation.mm:
(WebKit::PDFPluginAnnotation::create):
* WebProcess/Plugins/PDF/PDFPluginChoiceAnnotation.mm:
* WebProcess/Plugins/PDF/PDFPluginPasswordField.mm:
* WebProcess/Plugins/PDF/PDFPluginTextAnnotation.mm:
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::shouldUsePDFPlugin const):
(WebKit::drawPDFPage):
2021-03-03 Alex Christensen <achristensen@webkit.org>
Limit HashTable entry size to 500 bytes
https://bugs.webkit.org/show_bug.cgi?id=222658
Reviewed by Yusuke Suzuki.
* NetworkProcess/NetworkConnectionToWebProcess.h:
(WebKit::NetworkConnectionToWebProcess::getNetworkLoadInformationResponse):
(WebKit::NetworkConnectionToWebProcess::getNetworkLoadIntermediateInformation):
(WebKit::NetworkConnectionToWebProcess::takeNetworkLoadInformationMetrics):
(WebKit::NetworkConnectionToWebProcess::addNetworkLoadInformation):
(WebKit::NetworkConnectionToWebProcess::addNetworkLoadInformationMetrics):
* NetworkProcess/cache/CacheStorageEngineCaches.cpp:
(WebKit::CacheStorage::Caches::writeRecord):
(WebKit::CacheStorage::Caches::readRecord):
* NetworkProcess/cache/CacheStorageEngineCaches.h:
2021-03-03 Megan Gardner <megan_gardner@apple.com>
Preserve information about the origin of the app highlight request
https://bugs.webkit.org/show_bug.cgi?id=222223
Reviewed by Wenson Hsieh.
* Scripts/webkit/messages.py:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _storeAppHighlight:]):
(-[WKWebView _addAppHighlight]):
* UIProcess/API/Cocoa/_WKAppHighlightDelegate.h:
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::createAppHighlightInSelectedRange):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::contextMenuItemSelected):
* UIProcess/WebPageProxy.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView createHighlightInCurrentGroupWithRange:]):
(-[WKContentView createHighlightInNewGroupWithRange:]):
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::storeAppHighlight const):
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::createAppHighlightInSelectedRange):
* WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::highlightIsNewGroup const):
(WebKit::WebPage::highlightRequestOriginatedInApp const):
* WebProcess/WebPage/WebPage.messages.in:
2021-03-03 Kate Cheney <katherine_cheney@apple.com>
Report the correct document uri in the case of a ContentSecurityPolicyClient
https://bugs.webkit.org/show_bug.cgi?id=222489
<rdar://problem/73774118>
Reviewed by Brent Fulgham.
Pass the document URL from the Network Process when we schedule a load
in case we need to report a CSP violation in NetworkLoadChecker.
* NetworkProcess/NetworkLoadChecker.cpp:
(WebKit::NetworkLoadChecker::NetworkLoadChecker):
(WebKit::NetworkLoadChecker::contentSecurityPolicy):
The regular toString() method sets file:// URLs to null. We should use
toRawString() so we can report the scheme if the source origin is a
local file, as per the W3C spec.
* NetworkProcess/NetworkLoadChecker.h:
* NetworkProcess/NetworkResourceLoadParameters.cpp:
(WebKit::NetworkResourceLoadParameters::encode const):
(WebKit::NetworkResourceLoadParameters::decode):
* NetworkProcess/NetworkResourceLoadParameters.h:
* NetworkProcess/NetworkResourceLoader.cpp:
* NetworkProcess/PingLoad.cpp:
(WebKit::PingLoad::PingLoad):
* WebProcess/Network/WebLoaderStrategy.cpp:
(WebKit::WebLoaderStrategy::scheduleLoadFromNetworkProcess):
2021-03-03 Don Olmstead <don.olmstead@sony.com>
[CMake] JavaScriptCore GLib headers should be copies
https://bugs.webkit.org/show_bug.cgi?id=222625
Reviewed by Michael Catanzaro.
Update the includes for the GLib ports.
* PlatformGTK.cmake:
* PlatformWPE.cmake:
* WebProcess/InjectedBundle/API/glib/DOM/WebKitDOMNode.cpp:
* WebProcess/InjectedBundle/API/glib/WebKitFrame.cpp:
2021-03-02 Myles C. Maxfield <mmaxfield@apple.com>
REGRESSION(r270691): It's impossible to use MobileAsset fonts on macOS
https://bugs.webkit.org/show_bug.cgi?id=222582
<rdar://problem/74877450>
Reviewed by Per Arne Vollan.
Simply apply the iOS codepath to macOS.
There is also an additional change here to the Mac sandbox.
On Big Sur and earlier, before this patch, com.apple.mobileassetd was allowed,
and com.apple.mobileassetd.v2 was disallowed. After this patch, they both are
allowed, but only with the com.apple.webkit.extension.mach extension.
On other OSes, before this patch, nothing was allowed, whereas after this patch,
only com.apple.mobileassetd.v2 is allowed with the com.apple.webkit.extension.mach
extension.
Test: WebKit.MobileAssetSandboxCheck
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _grantAccessToAssetServices]):
(-[WKWebView _revokeAccessToAssetServices]):
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _grantAccessToAssetServices]): Deleted.
(-[WKWebView _revokeAccessToAssetServices]): Deleted.
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::grantAccessToAssetServices):
(WebKit::WebPageProxy::revokeAccessToAssetServices):
* UIProcess/WebPageProxy.h:
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::grantAccessToAssetServices): Deleted.
(WebKit::WebPageProxy::revokeAccessToAssetServices): Deleted.
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::revokeAccessToAssetServices):
* WebProcess/com.apple.WebProcess.sb.in:
* WebProcess/glib/WebProcessGLib.cpp:
(WebKit::WebProcess::grantAccessToAssetServices):
(WebKit::WebProcess::revokeAccessToAssetServices):
* WebProcess/playstation/WebProcessPlayStation.cpp:
(WebKit::WebProcess::grantAccessToAssetServices):
(WebKit::WebProcess::revokeAccessToAssetServices):
* WebProcess/win/WebProcessWin.cpp:
(WebKit::WebProcess::grantAccessToAssetServices):
(WebKit::WebProcess::revokeAccessToAssetServices):
2021-03-02 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] Image extraction should install image overlay content
https://bugs.webkit.org/show_bug.cgi?id=222621
Reviewed by Tim Horton.
Take some initial steps towards refactoring image extraction on iOS. See WebKitAdditions for more details.
* Platform/cocoa/ImageExtractionUtilities.h: Added.
* Platform/cocoa/ImageExtractionUtilities.mm: Added.
Add a new file to contain image-extraction-related functionality; see WebKitAdditions.
* SourcesCocoa.txt:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::updateWithImageExtractionResult):
* UIProcess/WebPageProxy.h:
* UIProcess/ios/WKContentViewInteraction.mm:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updateWithImageExtractionResult):
Add IPC plumbing between `WebPageProxy` (in the UI process) and `WebPage` to allow clients to push image
extraction results down to a given element on the page.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2021-03-02 Peng Liu <peng.liu6@apple.com>
[ BigSur Debug Wk 2 ] ASSERT NOT REACHED in WebCore::DisplayList::PaintFrameForMedia::apply
https://bugs.webkit.org/show_bug.cgi?id=221989
<rdar://problem/74403030>
Reviewed by Jer Noble.
No new tests, this will fix a crash on debug bots.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::applyMediaItem): When we cannot get the player, the player
with the given identifier must have been destroyed. This function can return true to indicate
that the item has been applied.
2021-03-02 Ricky Mondello <rmondello@apple.com>
Unconditionally return information in _autofillContext SPI when a field is focused
https://bugs.webkit.org/show_bug.cgi?id=221828
rdar://74211237
Reviewed by Darin Adler.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _autofillContext]): Change the contract of _autofillContext to return information for any focused
text field. Add a "version" key. Explicitly indicate whether we're in a login context. This SPI remains stringly
typed for flexibility.
2021-03-02 Carlos Garcia Campos <cgarcia@igalia.com>
REGRESSION(r263094): [GTK][WPE] API test /webkit/WebKitWebContext/languages is failing
https://bugs.webkit.org/show_bug.cgi?id=188111
Reviewed by Michael Catanzaro.
The GLib API allows to change the user preferred languages after the web process is created. Since r263094 we
are no loner sending the new overrides to the web process. Instead of calling overrideUserPreferredLanguages()
we now set the overrides in the WebProcessPool configuration, so that we can remove the language observer.
* UIProcess/API/glib/WebKitWebContext.cpp:
(webkit_web_context_set_preferred_languages): Use WebProcessPool::setOverrideLanguages() instead of
overrideUserPreferredLanguages().
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::WebProcessPool) Remove the language observer registration.
(WebKit::WebProcessPool::~WebProcessPool): Remove the language observer unregistration.
(WebKit::WebProcessPool::setOverrideLanguages): Update the language overrides in the configuration and notify
all processes.
(WebKit::WebProcessPool::languageChanged): Deleted.
* UIProcess/WebProcessPool.h:
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::userPreferredLanguagesChanged const): Call overrideUserPreferredLanguages() again here.
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
2021-03-02 Youenn Fablet <youenn@apple.com>
Use capture settings after recovering capture from GPUProcess
https://bugs.webkit.org/show_bug.cgi?id=221126
<rdar://problem/73744819>
Reviewed by Eric Carlson.
When updating constraints, store the constraints so that, on crash recovery,
we recreate the remote source and reapply the last constraints.
Manually tested and partially covered by API test.
* WebProcess/cocoa/RemoteRealtimeAudioSource.cpp:
(WebKit::RemoteRealtimeAudioSource::applyConstraints):
(WebKit::RemoteRealtimeAudioSource::gpuProcessConnectionDidClose):
* WebProcess/cocoa/RemoteRealtimeAudioSource.h:
* WebProcess/cocoa/RemoteRealtimeVideoSource.cpp:
(WebKit::RemoteRealtimeVideoSource::setFrameRateWithPreset):
(WebKit::RemoteRealtimeVideoSource::gpuProcessConnectionDidClose):
* WebProcess/cocoa/RemoteRealtimeVideoSource.h:
2021-03-02 Rob Buis <rbuis@igalia.com>
Null check platformData when decoding
https://bugs.webkit.org/show_bug.cgi?id=221442
Reviewed by Ryosuke Niwa.
Null check platformData when decoding.
Test: ipc/pasteboard-write-custom-data.html
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<PasteboardCustomData::Entry>::decode):
2021-03-02 Youenn Fablet <youenn@apple.com>
Camera capture in iOS with GPUProcess enabled does not always handle orientation well at start up
https://bugs.webkit.org/show_bug.cgi?id=222167
<rdar://problem/74520496>
Reviewed by Eric Carlson.
Before the patch, GPU process orientation monitors were only getting updates after creation of the web process connection.
The initial value was therefore sometimes wrong.
Update this by storing the orientation value in GPU process and initializing it for every web process connection.
Manually tested.
* GPUProcess/GPUProcess.cpp:
(WebKit::GPUProcess::createGPUConnectionToWebProcess):
(WebKit::GPUProcess::setOrientationForMediaCapture):
* GPUProcess/GPUProcess.h:
2021-03-01 Fujii Hironori <Hironori.Fujii@sony.com>
[GTK] Web inspector related tests very slow or timeout in debug mode
https://bugs.webkit.org/show_bug.cgi?id=208000
Reviewed by Carlos Garcia Campos.
setTimeout was very slow in the inspector page because it was
throttled by 1 second due to that the page is hidden. Mac port
disables HiddenPageDOMTimerThrottlingEnabled for the inspector
page under testing. Do the same for GTK port.
* UIProcess/Inspector/gtk/WebInspectorProxyGtk.cpp:
(WebKit::WebInspectorProxy::platformCreateFrontendPage):
setHiddenPageDOMTimerThrottlingEnabled(false) if under test.
2021-03-01 Chris Dumez <cdumez@apple.com>
Have WebProcessProxy::requestTermination() call processDidTerminateOrFailedToLaunch()
https://bugs.webkit.org/show_bug.cgi?id=222577
Reviewed by Geoffrey Garen.
Have WebProcessProxy::requestTermination() call processDidTerminateOrFailedToLaunch() instead of duplicating
the code. It was error-prone to have 2 separate code paths whether the process exited due to a crash or a
termination request. It led to Bug 222574 for example because we added some logic to remove the process from
the cache in processDidTerminateOrFailedToLaunch() but had failed to do so in requestTermination().
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::didClose):
(WebKit::WebProcessProxy::processDidTerminateOrFailedToLaunch):
(WebKit::WebProcessProxy::didFinishLaunching):
(WebKit::WebProcessProxy::requestTermination):
* UIProcess/WebProcessProxy.h:
2021-03-01 Truitt Savell <tsavell@apple.com>
Unreviewed, reverting r273655.
Broke internal Mac
Reverted changeset:
"Set ownership of IOSurfaces from the GPUProcess instead of
the WebProcess"
https://bugs.webkit.org/show_bug.cgi?id=222391
https://commits.webkit.org/r273655
2021-03-01 Chris Dumez <cdumez@apple.com>
Crash under WebProcessPool::pageBeginUsingWebsiteDataStore()
https://bugs.webkit.org/show_bug.cgi?id=222574
<rdar://68340471>
Reviewed by Geoffrey Garen.
The issue was that when WebProcessProxy::requestTermination() was called (e.g. process is killed by WebKit
for using too much memory), we would fail to remove the process from the WebProcessCache. Because the
terminated would stay in the cache (even though WebProcessProxy::shutDown() was called), we could potentially
try and use it later on for a navigation to the same domain. This would lead to crashes because
WebProcessProxy::shutDown() has already been called.
Note that we were previously correctly removing the process from the cache in case of a proper crash, inside
WebProcessProxy::processDidTerminateOrFailedToLaunch(). I have moved the logic to remove from the cache
from processDidTerminateOrFailedToLaunch() to shutDown() to avoid similar issues in the future.
* UIProcess/API/Cocoa/WKProcessPool.mm:
(-[WKProcessPool _requestWebProcessTermination:]):
* UIProcess/API/Cocoa/WKProcessPoolPrivate.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::shutDown):
(WebKit::WebProcessProxy::processDidTerminateOrFailedToLaunch):
2021-03-01 Stephan Szabo <stephan.szabo@sony.com>
[WinCairo] Building with ENABLE_SERVICE_WORKER=OFF fails
https://bugs.webkit.org/show_bug.cgi?id=222566
Unreviewed build fix.
* UIProcess/WebProcessPool.cpp: Only compile a method when service
worker is enabled.
2021-03-01 Chris Dumez <cdumez@apple.com>
Have WebProcessPool::processForRegistrableDomain() return a Ref<WebProcessProxy>
https://bugs.webkit.org/show_bug.cgi?id=222575
Reviewed by Geoffrey Garen.
Have WebProcessPool::processForRegistrableDomain() return a Ref<WebProcessProxy> instead of a WebProcessProxy&,
for extra safety. Returning a WebProcessProxy& is generally safe because the WebProcessPool is expected to hold
a reference to the WebProcessProxy it is returning (via WebProcessPool::m_processes). However, this is a bit
fragile and can lead to bad bugs as we've seen recently.
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::createNewWebProcess):
(WebKit::WebProcessPool::processForRegistrableDomain):
(WebKit::WebProcessPool::createWebPage):
* UIProcess/WebProcessPool.h:
2021-03-01 Said Abou-Hallawa <said@apple.com>
[GPU Process] Some DisplayList items may not be replayed back before calling getImageData()
https://bugs.webkit.org/show_bug.cgi?id=222343
Reviewed by Tim Horton.
Make sure all the pending DisplayList items are pushed to the GPUP and
submitted to RemoteImageBuffer before calling the asynchronous method
getImageData().
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
2021-03-01 Antti Koivisto <antti@apple.com>
Use PriorityQueue in NetworkCache::Storage
https://bugs.webkit.org/show_bug.cgi?id=222555
Reviewed by Chris Dumez.
Replace Vector<Deque<>> with a more flexible, simpler to use data structure.
* NetworkProcess/cache/NetworkCacheStorage.cpp:
(WebKit::NetworkCache::Storage::ReadOperation::ReadOperation):
(WebKit::NetworkCache::Storage::isHigherPriority):
(WebKit::NetworkCache::Storage::cancelAllReadOperations):
(WebKit::NetworkCache::Storage::dispatchPendingReadOperations):
(WebKit::NetworkCache::Storage::retrieve):
* NetworkProcess/cache/NetworkCacheStorage.h:
2021-03-01 Fujii Hironori <Hironori.Fujii@sony.com>
EventDispatcher::wheelEvent is accessing m_scrollingTrees without locking m_scrollingTreesMutex since r271235
https://bugs.webkit.org/show_bug.cgi?id=222529
Reviewed by Carlos Garcia Campos.
r271235 added a code to access m_scrollingTrees, but it didn't
lock m_scrollingTreesMutex. The code can be removed by moving it
into the above lambda expression.
EventDispatcher::wheelEvent was using a local lambda expression.
However the return value wasn't used. And, MSVC can't compile the
lambda expression because it has a bug that 'this' can't be used
in lambda expressions. Converted the lambda expression to a
do-while statement.
* WebProcess/WebPage/EventDispatcher.cpp:
(WebKit::EventDispatcher::wheelEvent):
2021-02-25 Simon Fraser <simon.fraser@apple.com>
Remove ENABLE_WEBPROCESS_WINDOWSERVER_BLOCKING which is always true for macOS
https://bugs.webkit.org/show_bug.cgi?id=222459
Reviewed by Sam Weinig.
ENABLE_WEBPROCESS_WINDOWSERVER_BLOCKING is always defined for PLATFORM(MAC), so remove it,
replacing with PLATFORM(MAC) in a few places. In order to reduce the number PLATFORM(MAC), define
HAVE(CVDISPLAYLINK) and use it in WK2 code that relates to DisplayLinks.
* Platform/cocoa/LayerHostingContext.mm:
(WebKit::LayerHostingContext::createForExternalHostingProcess):
* Shared/mac/HangDetectionDisablerMac.mm:
(WebKit::setClientsMayIgnoreEvents):
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::registerNotificationObservers):
(WebKit::WebProcessPool::unregisterNotificationObservers):
* UIProcess/WebPageProxy.cpp:
(WebKit::ScrollingObserver::singleton):
(WebKit::WebPageProxy::sendWheelEvent):
(WebKit::WebPageProxy::updateCurrentModifierState):
* UIProcess/WebProcessPool.cpp:
(WebKit::displayReconfigurationCallBack):
(WebKit::WebProcessPool::sendDisplayConfigurationChangedMessageForTesting):
* UIProcess/WebProcessPool.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::~WebProcessProxy):
(WebKit::WebProcessProxy::processWillShutDown):
* UIProcess/WebProcessProxy.h:
* UIProcess/WebProcessProxy.messages.in:
* UIProcess/mac/DisplayLink.cpp:
* UIProcess/mac/DisplayLink.h:
* UIProcess/mac/WebProcessProxyMac.mm:
(WebKit::WebProcessProxy::stopDisplayLink):
* WebProcess/InjectedBundle/mac/InjectedBundleMac.mm:
(WebKit::InjectedBundle::initialize):
* WebProcess/Plugins/PDF/PDFLayerControllerSPI.h:
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::PDFPlugin):
* WebProcess/WebPage/DrawingArea.cpp:
* WebProcess/WebPage/EventDispatcher.cpp:
* WebProcess/WebPage/EventDispatcher.h:
* WebProcess/WebPage/EventDispatcher.messages.in:
* WebProcess/WebPage/mac/DrawingAreaMac.cpp:
(WebKit::DrawingArea::createDisplayRefreshMonitor):
* WebProcess/WebProcess.cpp:
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
(WebKit::WebProcess::platformInitializeProcess):
2021-03-01 Chris Dumez <cdumez@apple.com>
Set ownership of IOSurfaces from the GPUProcess instead of the WebProcess
https://bugs.webkit.org/show_bug.cgi?id=222391
<rdar://problem/74748353>
Reviewed by Simon Fraser.
Replace use of IOSurfaceSetOwnership() SPI with IOSurfaceSetOwnershipIdentity().
Both do the same thing but IOSurfaceSetOwnershipIdentity() only requires an identity token
for the new owner (instead of a control port). As a result, IOSurfaceSetOwnershipIdentity()
requires a lot less priviledges and can now be used directly in the GPUProcess instead of
the WebProcess.
* GPUProcess/graphics/RemoteGraphicsContextGLCocoa.cpp:
(WebKit::RemoteGraphicsContextGLCocoa::prepareForDisplay):
* GPUProcess/graphics/RemoteImageBuffer.h:
(WebKit::RemoteImageBuffer::setProcessOwnership):
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::createImageBuffer):
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.cpp:
(WebKit::RemoteGraphicsContextGLProxy::prepareForDisplay):
* WebProcess/GPU/graphics/cocoa/ImageBufferShareableMappedIOSurfaceBackend.cpp:
(WebKit::ImageBufferShareableMappedIOSurfaceBackend::create):
(WebKit::ImageBufferShareableMappedIOSurfaceBackend::setProcessOwnership):
* WebProcess/GPU/graphics/cocoa/ImageBufferShareableMappedIOSurfaceBackend.h:
2021-03-01 Kimmo Kinnunen <kkinnunen@apple.com>
Timeout calculations are error-prone for compound IPC operations
https://bugs.webkit.org/show_bug.cgi?id=222305
Reviewed by Geoffrey Garen.
Add Timeout class to hold the timeout value in blocking functions
of IPC::.
This makes it possible to forward absolute points of time as the timeout
value, as well as preserving the ability to pass points of time relative
to the invocation.
Call the concept Timeout instead of TimeOut since word 'timeout' is an
English word and preferable when used as a noun according to ell stackoverflow.
No new tests, a refactor.
* Platform/IPC/Connection.cpp:
(IPC::Connection::sendMessage):
(IPC::Connection::waitForMessage):
(IPC::Connection::sendSyncMessage):
(IPC::Connection::waitForSyncReply):
* Platform/IPC/Connection.h:
(IPC::Connection::sendSync):
(IPC::Connection::waitForAndDispatchImmediately):
(IPC::Connection::waitForAsyncCallbackAndDispatchImmediately):
* Platform/IPC/IPCSemaphore.h:
* Platform/IPC/MessageSender.h:
(IPC::MessageSender::sendSync):
* Platform/IPC/StreamClientConnection.h:
(IPC::StreamClientConnection::send):
(IPC::StreamClientConnection::sendSync):
(IPC::StreamClientConnection::trySendDestinationIDIfNeeded):
(IPC::StreamClientConnection::tryAcquire):
Here are few examples of compound use-case.
* Platform/IPC/Timeout.h: Copied from Source/WebKit/Platform/IPC/IPCSemaphore.h.
(IPC::Timeout::Timeout):
(IPC::Timeout::infinity):
(IPC::Timeout::operator Seconds const):
(IPC::Timeout::hasPassed const):
* Platform/IPC/darwin/IPCSemaphoreDarwin.cpp:
(IPC::Semaphore::waitFor):
* UIProcess/AuxiliaryProcessProxy.h:
(WebKit::AuxiliaryProcessProxy::sendSync):
2021-03-01 Alberto Garcia <berto@igalia.com>
[GTK] Fails to build in i386: static assertion failed: divisor must be a power of two
https://bugs.webkit.org/show_bug.cgi?id=222480
Reviewed by Carlos Garcia Campos.
* Platform/IPC/StreamConnectionBuffer.h:
(IPC::StreamConnectionBuffer::headerSize):
2021-02-26 Adrian Perez de Castro <aperez@igalia.com>
[WPE][GTK] Bogus Response.{status,ok} for successful fetch() requests to a custom URI scheme handler
https://bugs.webkit.org/show_bug.cgi?id=222471
Reviewed by Carlos Garcia Campos.
No new tests needed.
* UIProcess/API/glib/WebKitURISchemeRequest.cpp:
(webkitURISchemeRequestReadCallback): Mark response as having status "200 OK" when
webkit_uri_scheme_request_finish() gets passed a GInputStream.
2021-02-26 Peng Liu <peng.liu6@apple.com>
[GPUP] Replace references to GPUProcessConnection and Connection with WeakPtr<GPUProcessConnection>
https://bugs.webkit.org/show_bug.cgi?id=222492
Reviewed by Eric Carlson.
As we did in r273473, this patch replaces some references to `GPUProcessConnection` and `Connection`
with `WeakPtr<GPUProcessConnection>`. r273473 was for the GPU process, this patch is for web processes.
When the GPU process crashes, the `GPUProcessConnection` object (as well as its `Connection` object)
will be destroyed, and the reference will be invalid.
No new tests, fixing test crashes on bots.
* WebProcess/GPU/media/AudioTrackPrivateRemote.cpp:
(WebKit::AudioTrackPrivateRemote::AudioTrackPrivateRemote):
(WebKit::AudioTrackPrivateRemote::setEnabled):
* WebProcess/GPU/media/AudioTrackPrivateRemote.h:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::addRemoteAudioTrack):
(WebKit::MediaPlayerPrivateRemote::addRemoteTextTrack):
(WebKit::MediaPlayerPrivateRemote::addRemoteVideoTrack):
* WebProcess/GPU/media/MediaSourcePrivateRemote.cpp:
(WebKit::MediaSourcePrivateRemote::MediaSourcePrivateRemote):
(WebKit::MediaSourcePrivateRemote::~MediaSourcePrivateRemote):
(WebKit::MediaSourcePrivateRemote::addSourceBuffer):
(WebKit::MediaSourcePrivateRemote::durationChanged):
(WebKit::MediaSourcePrivateRemote::setReadyState):
(WebKit::MediaSourcePrivateRemote::setIsSeeking):
(WebKit::MediaSourcePrivateRemote::waitForSeekCompleted):
(WebKit::MediaSourcePrivateRemote::seekCompleted):
(WebKit::MediaSourcePrivateRemote::setTimeFudgeFactor):
* WebProcess/GPU/media/MediaSourcePrivateRemote.h:
* WebProcess/GPU/media/RemoteImageDecoderAVF.cpp:
(WebKit::RemoteImageDecoderAVF::RemoteImageDecoderAVF):
(WebKit::RemoteImageDecoderAVF::createFrameImageAtIndex):
(WebKit::RemoteImageDecoderAVF::setExpectedContentSize):
(WebKit::RemoteImageDecoderAVF::setData):
(WebKit::RemoteImageDecoderAVF::gpuProcessConnection const): Deleted.
* WebProcess/GPU/media/RemoteImageDecoderAVF.h:
* WebProcess/GPU/media/RemoteImageDecoderAVFManager.cpp:
(WebKit::RemoteImageDecoderAVFManager::gpuProcessConnection const):
* WebProcess/GPU/media/RemoteImageDecoderAVFManager.h:
* WebProcess/GPU/media/SourceBufferPrivateRemote.cpp:
(WebKit::SourceBufferPrivateRemote::SourceBufferPrivateRemote):
(WebKit::SourceBufferPrivateRemote::~SourceBufferPrivateRemote):
(WebKit::SourceBufferPrivateRemote::append):
(WebKit::SourceBufferPrivateRemote::abort):
(WebKit::SourceBufferPrivateRemote::resetParserState):
(WebKit::SourceBufferPrivateRemote::removedFromMediaSource):
(WebKit::SourceBufferPrivateRemote::setReadyState):
(WebKit::SourceBufferPrivateRemote::setActive):
(WebKit::SourceBufferPrivateRemote::setMediaSourceEnded):
(WebKit::SourceBufferPrivateRemote::setMode):
(WebKit::SourceBufferPrivateRemote::updateBufferedFromTrackBuffers):
(WebKit::SourceBufferPrivateRemote::removeCodedFrames):
(WebKit::SourceBufferPrivateRemote::evictCodedFrames):
(WebKit::SourceBufferPrivateRemote::addTrackBuffer):
(WebKit::SourceBufferPrivateRemote::resetTrackBuffers):
(WebKit::SourceBufferPrivateRemote::clearTrackBuffers):
(WebKit::SourceBufferPrivateRemote::setAllTrackBuffersNeedRandomAccess):
(WebKit::SourceBufferPrivateRemote::setGroupStartTimestamp):
(WebKit::SourceBufferPrivateRemote::setGroupStartTimestampToEndTimestamp):
(WebKit::SourceBufferPrivateRemote::setShouldGenerateTimestamps):
(WebKit::SourceBufferPrivateRemote::reenqueueMediaIfNeeded):
(WebKit::SourceBufferPrivateRemote::resetTimestampOffsetInTrackBuffers):
(WebKit::SourceBufferPrivateRemote::startChangingType):
(WebKit::SourceBufferPrivateRemote::setTimestampOffset):
(WebKit::SourceBufferPrivateRemote::setAppendWindowStart):
(WebKit::SourceBufferPrivateRemote::setAppendWindowEnd):
(WebKit::SourceBufferPrivateRemote::seekToTime):
(WebKit::SourceBufferPrivateRemote::updateTrackIds):
(WebKit::SourceBufferPrivateRemote::bufferedSamplesForTrackId):
* WebProcess/GPU/media/SourceBufferPrivateRemote.h:
* WebProcess/GPU/media/TextTrackPrivateRemote.cpp:
(WebKit::TextTrackPrivateRemote::TextTrackPrivateRemote):
(WebKit::TextTrackPrivateRemote::setMode):
* WebProcess/GPU/media/TextTrackPrivateRemote.h:
* WebProcess/GPU/media/VideoTrackPrivateRemote.cpp:
(WebKit::VideoTrackPrivateRemote::VideoTrackPrivateRemote):
(WebKit::VideoTrackPrivateRemote::setSelected):
* WebProcess/GPU/media/VideoTrackPrivateRemote.h:
(WebKit::VideoTrackPrivateRemote::create):
2021-02-26 Chris Dumez <cdumez@apple.com>
Reduce explicit usage of [objC retain] in WebKit
https://bugs.webkit.org/show_bug.cgi?id=222439
Reviewed by Darin Adler.
Reduce explicit usage of [objC retain] in WebKit by using RetainPtr<>.
* NetworkProcess/Downloads/cocoa/DownloadCocoa.mm:
(WebKit::Download::resume):
* Shared/Cocoa/WebPreferencesDefaultValuesCocoa.mm:
(WebKit::isFeatureFlagEnabled):
* UIProcess/API/Cocoa/APIContentRuleListStoreCocoa.mm:
(API::ContentRuleListStore::defaultStorePath):
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
(+[WKWebsiteDataStore _allWebsiteDataTypesIncludingPrivate]):
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::WebsiteDataStore::tempDirectoryFileSystemRepresentation):
(WebKit::WebsiteDataStore::cacheDirectoryFileSystemRepresentation):
(WebKit::WebsiteDataStore::websiteDataDirectoryFileSystemRepresentation):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView keyCommands]):
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(-[WKPDFPluginAccessibilityObject ALLOW_DEPRECATED_IMPLEMENTATIONS_END]):
2021-02-26 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r273560.
https://bugs.webkit.org/show_bug.cgi?id=222499
reverting an unnecessary change after confirming it did not
fix build issues
Reverted changeset:
"Convert WKMediaPlaybackState NS_ENUM back to NSUInteger type"
https://bugs.webkit.org/show_bug.cgi?id=222475
https://trac.webkit.org/changeset/273560
2021-02-26 Kate Cheney <katherine_cheney@apple.com>
Remove unnecessary WKMediaPlaybackState header from Xcode project file
https://bugs.webkit.org/show_bug.cgi?id=222496
<rdar://problem/74808525>
Reviewed by Brent Fulgham.
* WebKit.xcodeproj/project.pbxproj:
2021-02-26 John Wilander <wilander@apple.com>
Non-cookie website data not deleted after 7 days of browser use without user interaction
https://bugs.webkit.org/show_bug.cgi?id=222248
<rdar://problem/74612589>
Reviewed by Brent Fulgham and informally by Kate Cheney.
Tests: http/tests/resourceLoadStatistics/operating-dates-all-but-cookies-not-removed-with-user-interaction-6-days-ago.html
http/tests/resourceLoadStatistics/operating-dates-all-but-cookies-removed-with-user-interaction-7-days-ago.html
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::updateOperatingDatesParameters):
Now sets two new member variables:
- Optional<OperatingDate> m_longWindowOperatingDate
- Optional<OperatingDate> m_shortWindowOperatingDate
These are only set if the minimum number of operating days has passed.
This function no longer sets m_leastRecentOperatingDate.
(WebKit::ResourceLoadStatisticsDatabaseStore::hasStatisticsExpired const):
Now checks against the two new member variables m_longWindowOperatingDate
and m_shortWindowOperatingDate. This fixes the bug.
(WebKit::ResourceLoadStatisticsDatabaseStore::insertExpiredStatisticForTesting):
Test infrastructure. Now takes the new parameter numberOfOperatingDaysPassed
with which layout tests can control the exact number of operating days
passed since user interaction.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.h:
Added m_longWindowOperatingDate and m_shortWindowOperatingDate, and
removed m_leastRecentOperatingDate since it's no longer used.
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:
(WebKit::ResourceLoadStatisticsMemoryStore::insertExpiredStatisticForTesting):
Test infrastructure. The same changes as in the DB store.
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.h:
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::insertExpiredStatisticForTesting):
Test infrastructure.
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.h:
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::insertExpiredStatisticForTesting):
Test infrastructure.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreSetStatisticsExpiredStatistic):
Test infrastructure.
* UIProcess/API/C/WKWebsiteDataStoreRef.h:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::insertExpiredStatisticForTesting):
Test infrastructure.
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::insertExpiredStatisticForTesting):
Test infrastructure.
* UIProcess/WebsiteData/WebsiteDataStore.h:
2021-02-26 BJ Burg <bburg@apple.com>
Web Inspector: give WebInspectorUIExtensionControllerProxy a chance to teardown when the frontend is about to close
https://bugs.webkit.org/show_bug.cgi?id=222486
Reviewed by Devin Rousso.
Sometimes when closing Web Inspector, the message receiver for WebInspectorUIExtensionControllerProxy
is not removed. Move this teardown into a separate method. It's too late to do this in the destructor
if a lambda has kept a strong reference to the controller beyond the point when the frontend was closed.
Covered by existing API tests, which will hopefully become less flaky as a result.
* UIProcess/Inspector/RemoteWebInspectorProxy.cpp:
(WebKit::RemoteWebInspectorProxy::closeFrontendPageAndWindow):
* UIProcess/Inspector/WebInspectorProxy.cpp:
(WebKit::WebInspectorProxy::closeFrontendPageAndWindow):
Notify the extensions controller that it's time to teardown.
* UIProcess/Inspector/WebInspectorUIExtensionControllerProxy.h:
* UIProcess/Inspector/WebInspectorUIExtensionControllerProxy.cpp:
(WebKit::WebInspectorUIExtensionControllerProxy::~WebInspectorUIExtensionControllerProxy):
By the time this destructor runs, we should have already gotten the message to teardown.
(WebKit::WebInspectorUIExtensionControllerProxy::inspectorFrontendWillClose): Added.
Use the soon-to-be-gone m_frontendPage to unregister as a message receiver.
2021-02-26 BJ Burg <bburg@apple.com>
[Cocoa] Web Inspector: add support for receiving Web Extension events via _WKInspectorExtensionDelegate
https://bugs.webkit.org/show_bug.cgi?id=222252
<rdar://problem/71206685>
Unreviewed, revert one unintentionally changed error code.
Fixes failing API test WKInspectorExtensionHost.UnregisterExtension.
* UIProcess/Inspector/WebInspectorUIExtensionControllerProxy.cpp:
(WebKit::WebInspectorUIExtensionControllerProxy::unregisterExtension):
2021-02-26 Per Arne <pvollan@apple.com>
[Cocoa] Send QOS parameters as part of Web process creation parameters
https://bugs.webkit.org/show_bug.cgi?id=222474
Reviewed by Chris Dumez.
In order to save one IPC call when starting a WebContent process, send QOS parameters as part of Web process creation parameters.
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::initializeNewWebProcess):
* WebProcess/WebProcess.messages.in:
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2021-02-26 Per Arne <pvollan@apple.com>
[Cocoa] Start WebContent process pre-warming when process is initialized
https://bugs.webkit.org/show_bug.cgi?id=222476
Reviewed by Chris Dumez.
Instead of pre-warming in response to a message, the WebContent process can start pre-warming automatically
if it is of a pre-warmed process type. This saves one IPC message on startup of the WebContent process.
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::initializeNewWebProcess):
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::initializeWebProcess):
* WebProcess/WebProcess.messages.in:
2021-02-26 Devin Rousso <drousso@apple.com>
Provide the image URL for accessibility image extraction
https://bugs.webkit.org/show_bug.cgi?id=222483
<rdar://problem/74144895>
Reviewed by Wenson Hsieh.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::requestImageExtraction):
* UIProcess/WebPageProxy.messages.in:
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::requestImageExtraction):
* UIProcess/PageClient.h:
(WebKit::PageClient::requestImageExtraction):
* UIProcess/mac/PageClientImplMac.h:
* UIProcess/mac/PageClientImplMac.mm:
(WebKit::PageClientImpl::requestImageExtraction):
* UIProcess/Cocoa/WebViewImpl.h:
* UIProcess/Cocoa/WebViewImpl.mm:
2021-02-26 Jer Noble <jer.noble@apple.com>
[GPUP] Allow painting of GPUP hosted video from the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=222461
Reviewed by Eric Carlson.
Add Cocoa implementations for nativeImageForCurrentTime() and pixelBufferForCurrentTime().
Use those methods to implement paintCurrentFrameInContext().
* GPUProcess/media/RemoteMediaPlayerProxy.h:
* GPUProcess/media/RemoteMediaPlayerProxy.messages.in:
* GPUProcess/media/cocoa/RemoteMediaPlayerProxyCocoa.mm:
(WebKit::RemoteMediaPlayerProxy::nativeImageForCurrentTime):
(WebKit::RemoteMediaPlayerProxy::pixelBufferForCurrentTime):
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::paint):
(WebKit::MediaPlayerPrivateRemote::paintCurrentFrameInContext):
(WebKit::MediaPlayerPrivateRemote::pixelBufferForCurrentTime):
* WebProcess/GPU/media/MediaPlayerPrivateRemote.h:
* WebProcess/GPU/media/cocoa/MediaPlayerPrivateRemoteCocoa.mm:
(WebKit::MediaPlayerPrivateRemote::nativeImageForCurrentTime):
(WebKit::MediaPlayerPrivateRemote::pixelBufferForCurrentTime):
2021-02-26 Per Arne <pvollan@apple.com>
[Cocoa] Register VP9 decoders when PlatformMediaSessionManager is created
https://bugs.webkit.org/show_bug.cgi?id=222473
<rdar://problem/74790242>
Reviewed by Eric Carlson.
Currently, VP9 decoders are registered when a Web page is created in the WebContent process. Instead, VP9 decoders can be registered when
PlatformMediaSessionManager is created, which should be a slight performance improvement, since calls into media frameworks will then not
be made unconditionally when creating a Web page, but delayed until required.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_lastNavigationWasAppBound):
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::enableVP9Decoder): Deleted.
(WebKit::WebProcess::enableVP8SWDecoder): Deleted.
(WebKit::WebProcess::enableVP9SWDecoder): Deleted.
* WebProcess/WebProcess.h:
2021-02-26 Kate Cheney <katherine_cheney@apple.com>
Convert WKMediaPlaybackState NS_ENUM back to NSUInteger type
https://bugs.webkit.org/show_bug.cgi?id=222475
<rdar://problem/74787665>
Reviewed by Alex Christensen.
* UIProcess/API/Cocoa/WKWebView.h:
2021-02-26 Tim Horton <timothy_horton@apple.com>
REGRESSION (r269824): macCatalyst WKWebView shows chunks of other tiles in the middle of content
https://bugs.webkit.org/show_bug.cgi?id=222460
<rdar://problem/74102753>
Reviewed by Simon Fraser.
* Shared/RemoteLayerTree/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::swapToValidFrontBuffer):
Reinstate some code that was accidentally deleted in r269824 that ensures
that we do not use an in-use IOSurface as our front buffer, so that
we don't paint into it while it's being composited in the render server.
Also, add some comments, since this code is all a little confusing
with its fronts and backs.
2021-02-25 Brent Fulgham <bfulgham@apple.com>
Serialize NSURLRequest (rather than CFURLRequest) in IPC
https://bugs.webkit.org/show_bug.cgi?id=222145
<rdar://problem/74500963>
Reviewed by Darin Adler and Alex Christensen.
Tested by existing layout and WebKitAPI tests.
Revise IPC serialization so that NSURLConnection (modern) Cocoa builds serialize the
NSURLRequest object, rather than bridging to CFURLRequest and serializing that. This
ensures that the full set of NSURLRequest features are accurately reconstituted on the
receiving side of the connection.
Because NSURL serialization doesn't handle some encoded parameters properly, we implement
an archiving delegate to archive the URL like we do for CFURL.
* Shared/Cocoa/ArgumentCodersCocoa.mm:
(-[WKSecureCodingArchivingDelegate archiver:willEncodeObject:]): Added.
(-[WKSecureCodingArchivingDelegate unarchiver:didDecodeObject:]): Added.
(+[WKEncodedURL supportsSecureCoding]): Added.
(-[WKEncodedURL encodeWithCoder:]): Added.
(-[WKEncodedURL initWithCoder:]): Added.
(-[WKEncodedURL initWithURL:]): Added.
(IPC::encodeSecureCodingInternal): Update to use new delegate.
(IPC::decodeSecureCodingInternal): Ditto.
(-[WKSecureCodingFontAttributeNormalizer archiver:willEncodeObject:]): Deleted.
* Shared/Cocoa/WebCoreArgumentCodersCocoa.mm:
(IPC::ArgumentCoder<WebCore::ResourceRequest>::encodePlatformData): Add NSURLRequest encoding.
(IPC::ArgumentCoder<WebCore::ResourceRequest>::decodePlatformData): Add NSURLRequest decoding.
* Shared/mac/WebCoreArgumentCodersMac.mm:
(IPC::createSerializableRepresentation): Deleted.
(IPC::extractDictionaryValue): Deleted.
(IPC::createArchiveList): Deleted.
(IPC::createCFURLRequestFromSerializableRepresentation): Deleted.
(IPC::createNSURLRequestFromSerializableRepresentation): Deleted.
(IPC::ArgumentCoder<WebCore::ResourceRequest>::encodePlatformData): Deleted.
(IPC::ArgumentCoder<WebCore::ResourceRequest>::decodePlatformData): Deleted.
2021-02-25 Peng Liu <peng.liu6@apple.com>
[GPUP] The GPU Process crashes in WebKit::RemoteLegacyCDMFactoryProxy::getSession() with an assertion failure
https://bugs.webkit.org/show_bug.cgi?id=222435
Reviewed by Eric Carlson.
Fix an assertion failure on debug bots.
No new tests. Fix crashes on the bots.
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
(WebKit::RemoteMediaPlayerProxy::mediaPlayerCachedKeyForKeyId const):
We should not search the hash table with an empty key (`m_legacySession`).
2021-02-25 BJ Burg <bburg@apple.com>
[Cocoa] Web Inspector: add support for receiving Web Extension events via _WKInspectorExtensionDelegate
https://bugs.webkit.org/show_bug.cgi?id=222252
<rdar://problem/71206685>
Reviewed by Devin Rousso.
This patch adds support for clients to receive extension lifecycle events. For now this includes
extension tab shown/hidden, but will eventually include page navigations and other events exposed
via Web Extensions API.
This is implemented at the WebKit level via new InspectorFrontendClient methods. This gets piped to
WebInspectorUIExtensionControllerProxy via IPC, which then passes on the event to _WKInspectorExtensionDelegate.
_WKInspectorExtensionDelegate is a delegate of _WKInspectorExtension. It's for receiving lifecycle events, and
is implemented similarly to _WKInspectorDelegate. WebInspectorUIExtensionControllerProxy now keeps a map of
extension API objects so that it can call out to the the appropriate extension API object's client/delegate.
To better handle the situation where client completion handlers may be invoked after the inspector page has been torn down,
switch WebInspectorUIExtensionControllerProxy to be RefCounted instead of std::unique_ptr so that the appropriate
API object can be retrieved in the async IPC reply block. (IPC message receiver code knows how to ref the receiver
if it expects to deliver a response).
Lastly, since this patch introduces InspectorExtensionError at the WebCore level, I decided to move the contents
of InspectorExtensionTypes.h to the Inspector namespace. It made more sense than typing out WebCore::InspectorExtensionError
and the like. As part of this change, the string type aliases Inspector::ExtensionID and Inspector:ExtensionTabID are
now defined in InspectorFrontendClient.h.
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
Add generated files for new message receiver.
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
Add new files for inspector extension client/delegate boilerplate classes.
* Scripts/webkit/messages.py:
* Shared/InspectorExtensionTypes.h:
* Shared/InspectorExtensionTypes.cpp:
(Inspector::extensionErrorToString):
(WebKit::inspectorExtensionErrorToString): Deleted.
Move these types and helpers to Inspector namespace.
* UIProcess/API/APIInspectorExtension.cpp:
(API::InspectorExtension::setClient): Add a client member which is used to call out to _WKInspectorExtensionDelegate.
(API::InspectorExtension::createTab): Fix namespaces.
(API::InspectorExtension::evaluateScript): Fix namespaces.
* UIProcess/API/APIInspectorExtension.h:
* UIProcess/API/APIInspectorExtensionClient.h: Copied from Source/WebKit/UIProcess/API/Cocoa/_WKInspectorExtensionInternal.h.
(API::InspectorExtensionClient::didShowExtensionTab):
(API::InspectorExtensionClient::didHideExtensionTab):
Added.
* UIProcess/API/Cocoa/_WKInspector.mm:
(-[_WKInspector registerExtensionWithID:displayName:completionHandler:]):
(-[_WKInspector unregisterExtension:completionHandler:]):
Adjust to new type signature that returns an initialized API::InspectorExtension.
* UIProcess/API/Cocoa/_WKRemoteWebInspectorViewController.mm:
(-[_WKRemoteWebInspectorViewController registerExtensionWithID:displayName:completionHandler:]):
(-[_WKRemoteWebInspectorViewController unregisterExtension:completionHandler:]):
Adjust to new type signature that returns an initialized API::InspectorExtension.
* UIProcess/API/Cocoa/_WKInspectorExtensionDelegate.h: Added.
* UIProcess/API/Cocoa/_WKInspectorExtension.h: Add the delegate property and update documentation to mention its intended usage.
* UIProcess/API/Cocoa/_WKInspectorExtensionInternal.h: Add delegate storage.
* UIProcess/API/Cocoa/_WKInspectorExtension.mm:
(-[_WKInspectorExtension delegate]):
(-[_WKInspectorExtension setDelegate:]):
Added.
(-[_WKInspectorExtension createTabWithName:tabIconURL:sourceURL:completionHandler:]): Fix namespaces.
(-[_WKInspectorExtension evaluateScript:frameURL:contextSecurityOrigin:useContentScriptContext:completionHandler:]): Fix namespaces.
* UIProcess/Inspector/Cocoa/InspectorExtensionDelegate.h: Added.
* UIProcess/Inspector/Cocoa/InspectorExtensionDelegate.mm: Added.
(WebKit::InspectorExtensionDelegate::InspectorExtensionDelegate):
(WebKit::InspectorExtensionDelegate::delegate):
(WebKit::InspectorExtensionDelegate::InspectorExtensionClient::InspectorExtensionClient):
(WebKit::InspectorExtensionDelegate::InspectorExtensionClient::~InspectorExtensionClient):
(WebKit::InspectorExtensionDelegate::InspectorExtensionClient::didShowExtensionTab):
(WebKit::InspectorExtensionDelegate::InspectorExtensionClient::didHideExtensionTab):
Added. This is ObjC-to-C++ bridge code to connect API::InspectorExtensionClient (used by WebInspectorUIExtensionControllerProxy)
to _WKInspectorExtensionDelegate (used by client code).
* UIProcess/Inspector/WebInspectorUIExtensionControllerProxy.h: Fix namespaces.
* UIProcess/Inspector/WebInspectorUIExtensionControllerProxy.messages.in: Added.
Make this class an IPC::MessageReceiver. It receives messages for extension lifecycle events.
* UIProcess/Inspector/WebInspectorUIExtensionControllerProxy.cpp:
(WebKit::WebInspectorUIExtensionControllerProxy::WebInspectorUIExtensionControllerProxy):
(WebKit::WebInspectorUIExtensionControllerProxy::~WebInspectorUIExtensionControllerProxy):
Make this class an IPC::MessageReceiver. It receives messages for extension lifecycle events.
(WebKit::WebInspectorUIExtensionControllerProxy::registerExtension):
Expand this code out so that we can save API::InspectorExtension instances into our own map
before sending it to the completionHandler. This map keeps strong references to the extension
API objects since they need to be used to send events at an arbitrary time in the future.
Previously, they were kept alive (or not) by the API client.
(WebKit::WebInspectorUIExtensionControllerProxy::unregisterExtension):
When the extension is no longer registered, remove the associated extension API object (and
our strong reference to it).
(WebKit::WebInspectorUIExtensionControllerProxy::createTabForExtension): Fix namespaces.
(WebKit::WebInspectorUIExtensionControllerProxy::evaluateScriptForExtension): Fix namespaces.
(WebKit::WebInspectorUIExtensionControllerProxy::didShowExtensionTab):
(WebKit::WebInspectorUIExtensionControllerProxy::didHideExtensionTab):
Added. Try to look up the associated API object for the extension and forward the event to
the client of the API object, which is ultimately _WKInspectorExtensionDelegate.
* WebProcess/Inspector/RemoteWebInspectorUI.h:
* WebProcess/Inspector/RemoteWebInspectorUI.cpp:
(WebKit::RemoteWebInspectorUI::supportsWebExtensions):
(WebKit::RemoteWebInspectorUI::didShowExtensionTab):
(WebKit::RemoteWebInspectorUI::didHideExtensionTab):
Added. Defer to the extension controller to handle further event propagation.
* WebProcess/Inspector/WebInspectorUI.h:
* WebProcess/Inspector/WebInspectorUI.cpp:
(WebKit::WebInspectorUI::supportsWebExtensions):
(WebKit::WebInspectorUI::didShowExtensionTab):
(WebKit::WebInspectorUI::didHideExtensionTab):
Added. Defer to the extension controller to handle further event propagation.
* WebProcess/Inspector/WebInspectorUIExtensionController.h:
* WebProcess/Inspector/WebInspectorUIExtensionController.messages.in:
Fix namespaces.
* WebProcess/Inspector/WebInspectorUIExtensionController.cpp:
(WebKit::WebInspectorUIExtensionController::WebInspectorUIExtensionController):
(WebKit::WebInspectorUIExtensionController::~WebInspectorUIExtensionController):
Be more precise about associating this message receiver with the inspector page's identifier.
Also, save the page identifier so that methods that need to send IPC messages can send
the message to the correct destinationID.
(WebKit::WebInspectorUIExtensionController::parseInspectorExtensionErrorFromEvaluationResult):
(WebKit::WebInspectorUIExtensionController::registerExtension):
(WebKit::WebInspectorUIExtensionController::unregisterExtension):
(WebKit::WebInspectorUIExtensionController::createTabForExtension):
(WebKit::WebInspectorUIExtensionController::evaluateScriptForExtension):
Fix namespaces.
(WebKit::WebInspectorUIExtensionController::didShowExtensionTab):
(WebKit::WebInspectorUIExtensionController::didHideExtensionTab):
Added.
2021-02-25 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r273503.
https://bugs.webkit.org/show_bug.cgi?id=222442
Caused a blob loading failure
Reverted changeset:
"Serialize NSURLRequest (rather than CFURLRequest) in IPC"
https://bugs.webkit.org/show_bug.cgi?id=222145
https://trac.webkit.org/changeset/273503
2021-02-25 Alex Christensen <achristensen@webkit.org>
Add stubs to enable SafariForWebKitDevelopment to launch
https://bugs.webkit.org/show_bug.cgi?id=222388
Reviewed by Myles Maxfield.
I manually verified that before this change, DYLD won't let it launch or it calls a missing selector,
but after this change it launches and loads web pages.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _inspectorDelegate]):
(-[WKWebView _setInspectorDelegate:]):
2021-02-25 Chris Dumez <cdumez@apple.com>
Update WebContent sandbox to allow it to call task_create_identity_token()
https://bugs.webkit.org/show_bug.cgi?id=222423
Reviewed by Simon Fraser.
Update WebContent sandbox to allow it to call task_create_identity_token(). We started using
task_create_identity_token() in r273288 and need it for Bug 222391.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-25 Eric Carlson <eric.carlson@apple.com>
[GPUP] Some media-capabilities tests fail when media in GPU Process is enabled
https://bugs.webkit.org/show_bug.cgi?id=221691
<rdar://problem/74439420>
Reviewed by Jer Noble.
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::setMediaOverridesForTesting):
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/GPUConnectionToWebProcess.messages.in:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/GPUProcessConnection.cpp:
(WebKit::GPUProcessConnection::updateMediaConfiguration):
* WebProcess/GPU/GPUProcessConnection.h:
* WebProcess/GPU/media/MediaOverridesForTesting.h: Added.
(WebKit::MediaOverridesForTesting::encode const):
(WebKit::MediaOverridesForTesting::decode):
* WebProcess/GPU/media/RemoteMediaEngineConfigurationFactory.cpp:
(WebKit::RemoteMediaEngineConfigurationFactory::registerFactory):
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::setUseGPUProcessForMedia):
2021-02-25 Brent Fulgham <bfulgham@apple.com>
Serialize NSURLRequest (rather than CFURLRequest) in IPC
https://bugs.webkit.org/show_bug.cgi?id=222145
<rdar://problem/74500963>
Reviewed by Darin Adler and Alex Christensen.
Tested by existing layout and WebKitAPI tests.
Revise IPC serialization so that NSURLConnection (modern) Cocoa builds serialize the
NSURLRequest object, rather than bridging to CFURLRequest and serializing that. This
ensures that the full set of NSURLRequest features are accurately reconstituted on the
receiving side of the connection.
Because NSURL serialization doesn't handle some encoded parameters properly, we implement
an archiving delegate to archive the URL like we do for CFURL.
* Shared/Cocoa/ArgumentCodersCocoa.mm:
(-[WKSecureCodingArchivingDelegate archiver:willEncodeObject:]): Added.
(-[WKSecureCodingArchivingDelegate unarchiver:didDecodeObject:]): Added.
(+[WKEncodedURL supportsSecureCoding]): Added.
(-[WKEncodedURL encodeWithCoder:]): Added.
(-[WKEncodedURL initWithCoder:]): Added.
(-[WKEncodedURL initWithURL:]): Added.
(IPC::encodeSecureCodingInternal): Update to use new delegate.
(IPC::decodeSecureCodingInternal): Ditto.
(-[WKSecureCodingFontAttributeNormalizer archiver:willEncodeObject:]): Deleted.
* Shared/Cocoa/WebCoreArgumentCodersCocoa.mm:
(IPC::ArgumentCoder<WebCore::ResourceRequest>::encodePlatformData): Add NSURLRequest encoding.
(IPC::ArgumentCoder<WebCore::ResourceRequest>::decodePlatformData): Add NSURLRequest decoding.
* Shared/mac/WebCoreArgumentCodersMac.mm:
(IPC::createSerializableRepresentation): Deleted.
(IPC::extractDictionaryValue): Deleted.
(IPC::createArchiveList): Deleted.
(IPC::createCFURLRequestFromSerializableRepresentation): Deleted.
(IPC::createNSURLRequestFromSerializableRepresentation): Deleted.
(IPC::ArgumentCoder<WebCore::ResourceRequest>::encodePlatformData): Deleted.
(IPC::ArgumentCoder<WebCore::ResourceRequest>::decodePlatformData): Deleted.
2021-02-25 Per Arne <pvollan@apple.com>
Temporarily stop enforcing notification filtering
https://bugs.webkit.org/show_bug.cgi?id=222386
<rdar://problem/73987767>
Reviewed by Brent Fulgham.
Temporarily stop enforcing notification filtering on macOS and iOS.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-25 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r273469.
https://bugs.webkit.org/show_bug.cgi?id=222417
Broke 30+ debug layout tests
Reverted changeset:
"Add stubs to enable SafariForWebKitDevelopment to launch"
https://bugs.webkit.org/show_bug.cgi?id=222388
https://trac.webkit.org/changeset/273469
2021-02-24 Peng Liu <peng.liu6@apple.com>
[GPUP] Use a WeakPtr of GPUConnectionToWebProcess instead of a reference in some media related objects
https://bugs.webkit.org/show_bug.cgi?id=222293
Reviewed by Eric Carlson.
A GPUConnectionToWebProcess object may be destroyed when we close a page, then crash will happen
in callbacks accessing the GPUConnectionToWebProcess object. For example, in
`RemoteSourceBufferProxy::sourceBufferPrivateBufferedRangesChanged()`. This patch
replaces the references to a GPUConnectionToWebProcess with WeakPtr(s) in places where the
issue may happen.
No new tests. Fix some test crashes.
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::isAlwaysOnLoggingAllowed const):
(WebKit::GPUConnectionToWebProcess::mediaEngineConfigurationFactoryProxy):
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/media/RemoteAudioHardwareListenerProxy.cpp:
(WebKit::RemoteAudioHardwareListenerProxy::RemoteAudioHardwareListenerProxy):
(WebKit::RemoteAudioHardwareListenerProxy::audioHardwareDidBecomeActive):
(WebKit::RemoteAudioHardwareListenerProxy::audioHardwareDidBecomeInactive):
(WebKit::RemoteAudioHardwareListenerProxy::audioOutputDeviceChanged):
* GPUProcess/media/RemoteAudioHardwareListenerProxy.h:
* GPUProcess/media/RemoteAudioTrackProxy.cpp:
(WebKit::RemoteAudioTrackProxy::RemoteAudioTrackProxy):
(WebKit::RemoteAudioTrackProxy::configurationChanged):
* GPUProcess/media/RemoteAudioTrackProxy.h:
* GPUProcess/media/RemoteCDMFactoryProxy.cpp:
(WebKit::RemoteCDMFactoryProxy::RemoteCDMFactoryProxy):
* GPUProcess/media/RemoteCDMFactoryProxy.h:
* GPUProcess/media/RemoteCDMInstanceProxy.cpp:
(WebKit::RemoteCDMInstanceProxy::unrequestedInitializationDataReceived):
* GPUProcess/media/RemoteCDMInstanceSessionProxy.cpp:
(WebKit::RemoteCDMInstanceSessionProxy::updateKeyStatuses):
(WebKit::RemoteCDMInstanceSessionProxy::sendMessage):
(WebKit::RemoteCDMInstanceSessionProxy::sessionIdChanged):
* GPUProcess/media/RemoteImageDecoderAVFProxy.cpp:
(WebKit::RemoteImageDecoderAVFProxy::RemoteImageDecoderAVFProxy):
(WebKit::RemoteImageDecoderAVFProxy::encodedDataStatusChanged):
* GPUProcess/media/RemoteImageDecoderAVFProxy.h:
* GPUProcess/media/RemoteLegacyCDMFactoryProxy.cpp:
(WebKit::RemoteLegacyCDMFactoryProxy::RemoteLegacyCDMFactoryProxy):
(WebKit::RemoteLegacyCDMFactoryProxy::~RemoteLegacyCDMFactoryProxy):
(WebKit::RemoteLegacyCDMFactoryProxy::addProxy):
(WebKit::RemoteLegacyCDMFactoryProxy::removeProxy):
(WebKit::RemoteLegacyCDMFactoryProxy::addSession):
(WebKit::RemoteLegacyCDMFactoryProxy::removeSession):
* GPUProcess/media/RemoteLegacyCDMFactoryProxy.h:
* GPUProcess/media/RemoteLegacyCDMProxy.cpp:
(WebKit::RemoteLegacyCDMProxy::cdmMediaPlayer const):
* GPUProcess/media/RemoteLegacyCDMSessionProxy.cpp:
(WebKit::RemoteLegacyCDMSessionProxy::sendMessage):
(WebKit::RemoteLegacyCDMSessionProxy::sendError):
(WebKit::RemoteLegacyCDMSessionProxy::mediaKeysStorageDirectory const):
* GPUProcess/media/RemoteMediaEngineConfigurationFactoryProxy.cpp:
(WebKit::RemoteMediaEngineConfigurationFactoryProxy::RemoteMediaEngineConfigurationFactoryProxy):
* GPUProcess/media/RemoteMediaEngineConfigurationFactoryProxy.h:
* GPUProcess/media/RemoteMediaPlayerManagerProxy.cpp:
(WebKit::RemoteMediaPlayerManagerProxy::RemoteMediaPlayerManagerProxy):
(WebKit::RemoteMediaPlayerManagerProxy::createMediaPlayer):
(WebKit::RemoteMediaPlayerManagerProxy::logger):
(WebKit::RemoteMediaPlayerManagerProxy::logger const): Deleted.
(WebKit::RemoteMediaPlayerManagerProxy::logChannel const): Deleted.
* GPUProcess/media/RemoteMediaPlayerManagerProxy.h:
(WebKit::RemoteMediaPlayerManagerProxy::gpuConnectionToWebProcess):
(WebKit::RemoteMediaPlayerManagerProxy::gpuConnectionToWebProcess const): Deleted.
(): Deleted.
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
(WebKit::RemoteMediaPlayerProxy::RemoteMediaPlayerProxy):
(WebKit::RemoteMediaPlayerProxy::loadMediaSource):
(WebKit::RemoteMediaPlayerProxy::requestResource):
(WebKit::RemoteMediaPlayerProxy::mediaPlayerMediaKeysStorageDirectory const):
(WebKit::RemoteMediaPlayerProxy::mediaPlayerMediaCacheDirectory const):
(WebKit::RemoteMediaPlayerProxy::addRemoteAudioTrackProxy):
(WebKit::RemoteMediaPlayerProxy::addRemoteVideoTrackProxy):
(WebKit::RemoteMediaPlayerProxy::addRemoteTextTrackProxy):
(WebKit::RemoteMediaPlayerProxy::mediaPlayerCachedKeyForKeyId const):
(WebKit::RemoteMediaPlayerProxy::setLegacyCDMSession):
(WebKit::RemoteMediaPlayerProxy::cdmInstanceAttached):
(WebKit::RemoteMediaPlayerProxy::cdmInstanceDetached):
(WebKit::RemoteMediaPlayerProxy::attemptToDecryptWithInstance):
* GPUProcess/media/RemoteMediaPlayerProxy.h:
* GPUProcess/media/RemoteMediaSourceProxy.cpp:
(WebKit::RemoteMediaSourceProxy::RemoteMediaSourceProxy):
(WebKit::RemoteMediaSourceProxy::~RemoteMediaSourceProxy):
(WebKit::RemoteMediaSourceProxy::seekToTime):
(WebKit::RemoteMediaSourceProxy::addSourceBuffer):
* GPUProcess/media/RemoteMediaSourceProxy.h:
* GPUProcess/media/RemoteRemoteCommandListenerProxy.cpp:
(WebKit::RemoteRemoteCommandListenerProxy::RemoteRemoteCommandListenerProxy):
(WebKit::RemoteRemoteCommandListenerProxy::didReceiveRemoteControlCommand):
* GPUProcess/media/RemoteRemoteCommandListenerProxy.h:
* GPUProcess/media/RemoteSourceBufferProxy.cpp:
(WebKit::RemoteSourceBufferProxy::RemoteSourceBufferProxy):
(WebKit::RemoteSourceBufferProxy::~RemoteSourceBufferProxy):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateDidReceiveInitializationSegment):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateStreamEndedWithDecodeError):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateAppendError):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateHighestPresentationTimestampChanged):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateDurationChanged):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateDidParseSample):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateDidDropSample):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateAppendComplete):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateDidReceiveRenderingError):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateReportExtraMemoryCost):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateBufferedDirtyChanged):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateBufferedRangesChanged):
* GPUProcess/media/RemoteSourceBufferProxy.h:
* GPUProcess/media/RemoteTextTrackProxy.cpp:
(WebKit::RemoteTextTrackProxy::RemoteTextTrackProxy):
(WebKit::RemoteTextTrackProxy::configurationChanged):
(WebKit::RemoteTextTrackProxy::addDataCue):
(WebKit::RemoteTextTrackProxy::updateDataCue):
(WebKit::RemoteTextTrackProxy::removeDataCue):
(WebKit::RemoteTextTrackProxy::addGenericCue):
(WebKit::RemoteTextTrackProxy::updateGenericCue):
(WebKit::RemoteTextTrackProxy::removeGenericCue):
(WebKit::RemoteTextTrackProxy::parseWebVTTFileHeader):
(WebKit::RemoteTextTrackProxy::parseWebVTTCueData):
* GPUProcess/media/RemoteTextTrackProxy.h:
* GPUProcess/media/RemoteVideoTrackProxy.cpp:
(WebKit::RemoteVideoTrackProxy::RemoteVideoTrackProxy):
(WebKit::RemoteVideoTrackProxy::configurationChanged):
* GPUProcess/media/RemoteVideoTrackProxy.h:
2021-02-24 BJ Burg <bburg@apple.com>
[Cocoa] Web Inspector: expose the extension host that is used to load _WKInspectorExtension tabs
https://bugs.webkit.org/show_bug.cgi?id=222344
<rdar://problem/74678778>
Reviewed by Devin Rousso.
Clients need access to the WKWebView that hosts extension content in order to implement message passing
and event delivery to the extension panel iframes. (Currently, that WebView is
the same underlying WKWebView for the whole interface, but that is open to change.)
* UIProcess/API/Cocoa/_WKInspectorExtensionHost.h: New property.
* UIProcess/API/Cocoa/_WKInspectorPrivate.h:
* UIProcess/API/Cocoa/_WKInspector.mm:
(-[_WKInspector extensionHostWebView]): Added.
(-[_WKInspector handle]): Deleted.
* UIProcess/API/Cocoa/_WKRemoteWebInspectorViewController.mm:
(-[_WKRemoteWebInspectorViewController extensionHostWebView]): Added.
(-[_WKRemoteWebInspectorViewController handle]): Deleted.
* Shared/API/Cocoa/WKBrowsingContextHandle.mm:
(-[WKBrowsingContextHandle description]):
Drive-by, add the standard -description for debugging purposes.
2021-02-24 Alex Christensen <achristensen@webkit.org>
Add stubs to enable SafariForWebKitDevelopment to launch
https://bugs.webkit.org/show_bug.cgi?id=222388
Reviewed by Myles Maxfield.
I manually verified that before this change, DYLD won't let it launch or it calls a missing selector,
but after this change it launches and loads web pages.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _inspectorDelegate]):
(-[WKWebView _setInspectorDelegate:]):
2021-02-24 Kate Cheney <katherine_cheney@apple.com>
App-bound request parameter should be passed to main resource requests not the main frame
https://bugs.webkit.org/show_bug.cgi?id=222241
<rdar://problem/74560966>
Reviewed by Chris Dumez.
Bug fix to set the isAppBound parameter on the resource request if it
is not on the main frame. Previously this only checked if the
resource was not a main resource before setting the parameter.
Otherwise mostly changes to add better testing.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::appBoundNavigationData):
(WebKit::NetworkProcess::clearAppBoundNavigationData):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/NetworkSession.h:
(WebKit::NetworkSession::appBoundNavigationTestingData):
* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::NetworkDataTaskCocoa::NetworkDataTaskCocoa):
* Scripts/webkit/messages.py:
* Shared/NavigatingToAppBoundDomain.h:
(WebKit::AppBoundNavigationTestingData::clearAppBoundNavigationDataTesting):
(WebKit::AppBoundNavigationTestingData::updateAppBoundNavigationTestingData):
(WebKit::AppBoundNavigationTestingData::encode const):
(WebKit::AppBoundNavigationTestingData::decode):
Struct to return data for tests.
* UIProcess/API/Cocoa/WKWebViewPrivateForTesting.h:
* UIProcess/API/Cocoa/WKWebViewTesting.mm:
(-[WKWebView _appBoundNavigationData:]):
(-[WKWebView _clearAppBoundNavigationData:]):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::appBoundNavigationData):
(WebKit::WebPageProxy::clearAppBoundNavigationData):
* UIProcess/WebPageProxy.h:
* WebProcess/Network/WebLoaderStrategy.cpp:
(WebKit::WebLoaderStrategy::loadResource):
The fix.
2021-02-24 Peng Liu <peng.liu6@apple.com>
[GPUP] Refactor RemoteImageDecoderAVF::createFrameImageAtIndex()
https://bugs.webkit.org/show_bug.cgi?id=222300
Reviewed by Jer Noble.
Use callOnMainThreadAndWait() instead of callOnMainThread() and remove the semaphore.
No new tests, this has no behavior change.
* GPUProcess/media/RemoteImageDecoderAVFProxy.messages.in:
* WebProcess/GPU/media/RemoteImageDecoderAVF.cpp:
(WebKit::RemoteImageDecoderAVF::createFrameImageAtIndex):
2021-02-24 Chris Dumez <cdumez@apple.com>
Unreviewed, fix build with the latest iOS SDK.
* Shared/FocusedElementInformation.h:
2021-02-24 Per Arne Vollan <pvollan@apple.com>
[iOS] Get default value in the UI process for whether synchronous XMLHttpRequest are allowed during unload
https://bugs.webkit.org/show_bug.cgi?id=222377
Reviewed by Brent Fulgham.
Getting this default value in the UI process instead of in every WebContent process should be a small speedup,
since getting this default value is a bit costly on iOS.
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
* Shared/ios/WebPreferencesDefaultValuesIOS.h: Added.
* Shared/ios/WebPreferencesDefaultValuesIOS.mm:
(WebKit::cachedAllowsRequest):
(WebKit::allowsDeprecatedSynchronousXMLHttpRequestDuringUnload):
(WebKit::setAllowsDeprecatedSynchronousXMLHttpRequestDuringUnload):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_lastNavigationWasAppBound):
2021-02-24 Russell Epstein <repstein@apple.com>
Revert r272735. rdar://problem/74674837
2021-02-24 Russell Epstein <repstein@apple.com>
Revert r272810. rdar://problem/74674837
2021-02-24 Said Abou-Hallawa <said@apple.com>
[GPU Process] Implement ImageBufferShareableBitmapBackend::bytesPerRow()
https://bugs.webkit.org/show_bug.cgi?id=222348
Reviewed by Simon Fraser.
The default implementation ImageBufferBackend::bytesPerRow() is not correct
for ImageBufferShareableBitmapBackend because ShareableBitmap aligns the
rows using IOSurfaceAlignProperty(kIOSurfaceBytesPerRow, ...).
* Shared/ShareableBitmap.h:
(WebKit::ShareableBitmap::bytesPerRow const):
* WebProcess/GPU/graphics/ImageBufferShareableBitmapBackend.cpp:
(WebKit::ImageBufferShareableBitmapBackend::bytesPerRow const):
* WebProcess/GPU/graphics/ImageBufferShareableBitmapBackend.h:
2021-02-24 Carlos Garcia Campos <cgarcia@igalia.com>
REGRESSION(r267763): [GTK][WPE] user preferred languages are sent to the wrong network process
https://bugs.webkit.org/show_bug.cgi?id=222347
Reviewed by Adrian Perez de Castro.
WebProcessPool::languageChanged() is creating the default website data store, so languages are sent to the newly
created network process. We should send them to the existing network processes instead.
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::languageChanged):
2021-02-24 Carlos Garcia Campos <cgarcia@igalia.com>
Unreviewed. Fix GTK build with neewer wpe and wpe-fdo
We need to explicitly include wpe.h now.
* UIProcess/API/glib/WebKitProtocolHandler.cpp:
* UIProcess/gtk/AcceleratedBackingStoreWayland.cpp:
2021-02-23 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] Add null checks for Node before accessing treeScope() in rangeForPointInRootViewCoordinates
https://bugs.webkit.org/show_bug.cgi?id=222338
<rdar://problem/72292289>
Reviewed by Tim Horton.
Add a null check for the hit test result's `targetNode` inside `rangeForPointInRootViewCoordinates()` to avoid
crashing in the case where the target node is null and either `result <= selectionStart` is `false` or
`selectionEnd <= result` is `false`.
Additionally, avoid a potential null deref on `selectionStart.deepEquivalent().anchorNode()` and
`selectionEnd.deepEquivalent().anchorNode()` by using `Position::treeScope()` (which null checks the anchor
node) instead of directly asking the anchor node.
Note that the call sites of `rangeForPointInRootViewCoordinates()` already handle a `WTF::nullopt` return value
gracefully, by skipping the selection update.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::rangeForPointInRootViewCoordinates):
2021-02-23 Patrick Angle <pangle@apple.com>
[Cocoa] Web Inspector: Add support for reloading the inspected page via _WKInspectorExtension
https://bugs.webkit.org/show_bug.cgi?id=222296
Reviewed by BJ Burg.
Added a new method to _WKInspectorExtension which is used to implement `browser.devtools.inspectedWindow.reload`
in the Web Extensions API.
* UIProcess/API/APIInspectorExtension.cpp:
(API::InspectorExtension::reloadIgnoringCache):
* UIProcess/API/APIInspectorExtension.h:
* UIProcess/API/Cocoa/_WKInspectorExtension.h:
* UIProcess/API/Cocoa/_WKInspectorExtension.mm:
(-[_WKInspectorExtension reloadIgnoringCache:userAgent:injectedScript:completionHandler:]):
* UIProcess/Inspector/WebInspectorUIExtensionControllerProxy.cpp:
(WebKit::WebInspectorUIExtensionControllerProxy::reloadForExtension):
* UIProcess/Inspector/WebInspectorUIExtensionControllerProxy.h:
* WebProcess/Inspector/WebInspectorUIExtensionController.cpp:
(WebKit::WebInspectorUIExtensionController::reloadForExtension):
* WebProcess/Inspector/WebInspectorUIExtensionController.h:
* WebProcess/Inspector/WebInspectorUIExtensionController.messages.in:
2021-02-23 Brent Fulgham <bfulgham@apple.com>
[macOS] Allow additional syscall support in the sandbox
https://bugs.webkit.org/show_bug.cgi?id=222335
<rdar://74473824>
Reviewed by Per Arne Vollan.
Expand the WebContent process sandbox to allow SYS_open_dprotected_np
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-23 Fujii Hironori <Hironori.Fujii@sony.com>
[CoordinatedGraphics] The whole content is unnecessarily repainted by animations in non-AC mode pages
https://bugs.webkit.org/show_bug.cgi?id=221391
Reviewed by Carlos Garcia Campos.
When triggerRenderingUpdate was called back, it added the whole
view area into the dirty region to repaint in non-AC mode pages.
This caused a problem that the whole content was unnecessarily
repainted by animations in non-AC mode pages.
Call scheduleDisplay instead of setNeedsDisplay in
triggerRenderingUpdate.
However, if the dirty region is empty, display() exited early. If
triggerRenderingUpdate is called back,
DrawingAreaCoordinatedGraphics::display should be called to ensure
WebPage::updateRendering() called even if m_dirtyRegion is empty.
Added a new flag m_scheduledWhileWaitingForDidUpdate.
* WebProcess/WebPage/CoordinatedGraphics/DrawingAreaCoordinatedGraphics.cpp:
(WebKit::DrawingAreaCoordinatedGraphics::triggerRenderingUpdate):
Call scheduleDisplay() instead of setNeedsDisplay().
(WebKit::DrawingAreaCoordinatedGraphics::didUpdate):
(WebKit::DrawingAreaCoordinatedGraphics::scheduleDisplay): Don't
return early even if m_dirtyRegion is empty.
(WebKit::DrawingAreaCoordinatedGraphics::display()): Ditto.
(WebKit::DrawingAreaCoordinatedGraphics::display(UpdateInfo&)):
Return early if m_dirtyRegion is empty.
* WebProcess/WebPage/CoordinatedGraphics/DrawingAreaCoordinatedGraphics.h:
Added m_scheduledWhileWaitingForDidUpdate.
2021-02-23 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r273304.
https://bugs.webkit.org/show_bug.cgi?id=222324
Introduced test failure
Reverted changeset:
"[macOS] Crash under AuxiliaryProcess::initializeSandbox"
https://bugs.webkit.org/show_bug.cgi?id=222233
https://trac.webkit.org/changeset/273304
2021-02-23 Kimmo Kinnunen <kkinnunen@apple.com>
REGRESSION(r273204): webkitpy tests fail
https://bugs.webkit.org/show_bug.cgi?id=222275
<rdar://problem/74614397>
Reviewed by Jonathan Bedard.
Regenerate the IPC generation test expectations that were
not part of r273204.
Tested by
./Tools/Scripts/test-webkitpy-python2 -j1 -p Source/WebKit/Scripts/webkit/messages_unittest.py
* Scripts/webkit/tests/MessageArgumentDescriptions.cpp:
(IPC::jsValueForArguments):
(IPC::messageArgumentDescriptions):
* Scripts/webkit/tests/MessageNames.cpp:
(IPC::description):
(IPC::receiverName):
(IPC::isValidMessageName):
* Scripts/webkit/tests/MessageNames.h:
* Scripts/webkit/tests/TestWithSemaphoreMessages.h:
* Scripts/webkit/tests/TestWithStreamMessages.h:
2021-02-23 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] Add a watchdog timer to forcibly reset deferring gesture recognizers
https://bugs.webkit.org/show_bug.cgi?id=222216
Reviewed by Simon Fraser.
Add a mechanism to detect and resolve situations where at least one deferring gesture recognizer
(`WKDeferringGestureRecognizer`) is stuck in Possible state after ending touches, but the touch event gesture
recognizer is unable to be reset (i.e. neither possible or recognizing). This state is indicative of a general
class of bugs regarding native gesture deferral, wherein our "gesture gates" never end up being lifted after
ending all touches in the view. One such example of this type of bug was fixed in r272584, where all interaction
(touch events, scrolling, tapping) due to the touch end deferrers never exiting Possible state after long
pressing to present a context menu.
To make these kinds of bugs both easier to diagnose and less fatal in the future, we introduce a watchdog timer
that dispatches after ending all touches on the content view, and checks the consistency of deferring gesture
recognizers in the view; if we detect stuck deferring gestures, we handle this by forcibly resetting the stuck
deferring gestures and logging a fault message.
I manually verified this change by reverting my fix for the aforementioned bug (r272584), and observing both (1)
a system log fault message, and (2) that interactions continued to work even without the fix in r272584, after
the watchdog timer fired. See below for more details.
* SourcesCocoa.txt:
* UIProcess/ios/GestureRecognizerConsistencyEnforcer.h: Copied from Source/WebKit/UIProcess/ios/WKDeferringGestureRecognizer.h.
* UIProcess/ios/GestureRecognizerConsistencyEnforcer.mm: Added.
(WebKit::GestureRecognizerConsistencyEnforcer::GestureRecognizerConsistencyEnforcer):
(WebKit::GestureRecognizerConsistencyEnforcer::beginTracking):
(WebKit::GestureRecognizerConsistencyEnforcer::endTracking):
(WebKit::GestureRecognizerConsistencyEnforcer::reset):
(WebKit::GestureRecognizerConsistencyEnforcer::timerFired):
Add a helper class that wraps a `RunLoop::Timer` for the main runloop, and schedules a 1 second timer after all
deferring gestures have ended; assuming no new touches begin after this timer is scheduled, we'll perform a
consistency check over the content view's gestures and perform a log fault and force deferring gestures to end
if needed.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView setUpInteraction]):
(-[WKContentView cleanUpInteraction]):
Make sure that we reset any extant `GestureRecognizerConsistencyEnforcer` if the web process crashes.
(-[WKContentView _removeDefaultGestureRecognizers]):
(-[WKContentView _addDefaultGestureRecognizers]):
(-[WKContentView _webTouchEventsRecognized:]):
Adjust a few compile-time guards so that deferring gesture recognizers exist on the view, even if
`ENABLE_IOS_TOUCH_EVENTS` is off. This makes some of the gesture deferral code throughout this file easier to
reason about in non-internal iOS builds, since we don't need to sprinkle as many `ENABLE(IOS_TOUCH_EVENTS)`
checks throughout this class.
(-[WKContentView touchEventGestureRecognizer]):
(-[WKContentView gestureRecognizerConsistencyEnforcer]):
(-[WKContentView deferringGestures]):
(-[WKContentView gestureRecognizer:shouldRecognizeSimultaneouslyWithGestureRecognizer:]):
(-[WKContentView deferringGestureRecognizer:willBeginTouchesWithEvent:]):
Refactor `-deferringGestureRecognizer:shouldDeferGesturesAfterBeginningTouchesWithEvent:` so that it returns an
enum flag indicating whether the given gesture should be allowed to begin deferring native gestures; by making
this `-deferringGestureRecognizer:willBeginTouchesWithEvent:`, we can now put state management logic here that's
responsible for updating the `GestureRecognizerConsistencyEnforcer`.
(-[WKContentView deferringGestureRecognizer:didTransitionToState:]):
(-[WKContentView deferringGestureRecognizer:didEndTouchesWithEvent:]):
(-[WKContentView _deferringGestures]): Deleted.
Rename this to `-deferringGestures` and expose it as a readonly property, for use by the
`GestureRecognizerConsistencyEnforcer`.
(-[WKContentView deferringGestureRecognizer:shouldDeferGesturesAfterBeginningTouchesWithEvent:]): Deleted.
* UIProcess/ios/WKDeferringGestureRecognizer.h:
* UIProcess/ios/WKDeferringGestureRecognizer.mm:
(-[WKDeferringGestureRecognizer touchesBegan:withEvent:]):
(-[WKDeferringGestureRecognizer setState:]):
Add a delegate hook to inform `WKContentView` when the gesture recognizer state changes.
(-[WKDeferringGestureRecognizer deferringGestureDelegate]): Deleted.
* UIProcess/mac/WKPrintingView.h:
Fix a missing include that results in a build failure due to changed unified sources.
* WebKit.xcodeproj/project.pbxproj:
2021-02-23 Per Arne Vollan <pvollan@apple.com>
[macOS] Crash under AuxiliaryProcess::initializeSandbox
https://bugs.webkit.org/show_bug.cgi?id=222233
Unreviewed, address additional review feedback.
Add comment and logging when the WebKit client is setting the user directory suffix.
* Shared/mac/AuxiliaryProcessMac.mm:
(WebKit::initializeSandboxParameters):
2021-02-23 Kimmo Kinnunen <kkinnunen@apple.com>
HTMLCanvasElement::copiedImage() contains old image with GPU Process on
https://bugs.webkit.org/show_bug.cgi?id=222101
Reviewed by Simon Fraser.
Implement RemoteImageBufferProxy::copyImage and
RemoteImageBufferProxy::copyNativeImage.
Implement by asking the data from the GPU process.
Previously, with AcceleratedImageBufferShareableMappedBackend
the HTMLCanvasElement::copiedImage() would construct the copy
by copying the CGContext backed by the IOSurface that is accessible
by both processes. This would succeed. After a draw to the image,
the second copy would not reflect the draw but contain the same
image as the first copy.
Fix by not using the fact that the IOSurface is accessible by both
processes. In the end, IOSurfaces will not be available in
Web process.
Converts NativeImage to ShareableBitmap with WebCore::CompositeOperator::Copy
since the ImageBuffers might have unpremultiplied contents. Also, ShareableBitmap
does not have contents, so blending is overhead.
No new tests, enables ~150 tests.
* GPUProcess/graphics/RemoteGraphicsContextGL.cpp:
(WebKit::RemoteGraphicsContextGL::copyTextureFromMedia):
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::getShareableBitmapForImageBuffer):
* GPUProcess/graphics/RemoteRenderingBackend.h:
* GPUProcess/graphics/RemoteRenderingBackend.messages.in:
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::getShareableBitmap):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
2021-02-22 Don Olmstead <don.olmstead@sony.com>
Non-unified build fixes late February 2021 edition
https://bugs.webkit.org/show_bug.cgi?id=222303
Unreviewed non-unified build fixes.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
2021-02-22 Devin Rousso <drousso@apple.com>
[Payment Request] REGRESSION(r273143): crash shortly after showing the Apple Pay sheet
https://bugs.webkit.org/show_bug.cgi?id=222298
<rdar://problem/74557943>
Reviewed by Wenson Hsieh.
* Platform/cocoa/PaymentAuthorizationPresenter.mm:
(WebKit::toNSErrors):
Don't use `adoptNS` with `+[NSMutableArray array]` since that returns an autoreleased object.
Use the helper `createNSArray` instead (as was previously done before r273143).
2021-02-22 Wenson Hsieh <wenson_hsieh@apple.com>
WebMouseEvent.h should avoid pulling in WebCore headers that know about DOM nodes
https://bugs.webkit.org/show_bug.cgi?id=222291
Reviewed by Darin Adler.
Import `PointerEventTypeNames.h` instead of `PointerEvent.h` in several UI process and shared headers.
* Shared/WebMouseEvent.h:
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(primaryPointerForType):
(webkitWebViewBaseSynthesizeMouseEvent):
* UIProcess/Automation/SimulatedInputDispatcher.cpp:
(WebKit::SimulatedInputDispatcher::transitionInputSourceToState):
* UIProcess/Automation/WebAutomationSession.cpp:
(WebKit::WebAutomationSession::performMouseInteraction):
* UIProcess/gtk/PointerLockManager.cpp:
(WebKit::PointerLockManager::handleMotion):
2021-02-22 Chris Dumez <cdumez@apple.com>
Add assertions to help identify to source of <rdar://68340471>
https://bugs.webkit.org/show_bug.cgi?id=222290
Reviewed by Alex Christensen.
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::pageBeginUsingWebsiteDataStore):
(WebKit::WebProcessPool::pageEndUsingWebsiteDataStore):
(WebKit::WebProcessPool::hasPagesUsingWebsiteDataStore const):
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::addExistingWebPage):
2021-02-22 Said Abou-Hallawa <said@apple.com>
[GPU Process] Implement the ClipToDrawingCommands item by using begin and end markers
https://bugs.webkit.org/show_bug.cgi?id=222230
Reviewed by Simon Fraser.
Replace the DisplayList item ClipToDrawingCommands by the pair items
BeginClipToDrawingCommands and EndClipToDrawingCommands.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::decodeItem):
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
2021-02-22 Eric Carlson <eric.carlson@apple.com>
UserMediaPermissionRequestManagerProxy may be released while computing capture device list
https://bugs.webkit.org/show_bug.cgi?id=222236
<rdar://74480265>
Unreviewed, address post-review comment.
* UIProcess/UserMediaPermissionRequestManagerProxy.cpp:
(WebKit::UserMediaPermissionRequestManagerProxy::computeFilteredDeviceList): Use
captured `this`.
2021-02-22 Chris Dumez <cdumez@apple.com>
Prepare for memory ownership transfer in the GPUProcess
https://bugs.webkit.org/show_bug.cgi?id=222122
Reviewed by Simon Fraser.
Prepare for memory ownership transfer in the GPUProcess:
1. Have the WebProcess create a task identity token and pass it to the GPUProcess
when connecting to it. The WebProcess' task identity token is then stored on
the GPUConnectionToWebProcess object.
2. Add the entitlement to the GPUProcess that is needed to transfer memory ownership
to another process.
This is a first step towards having the GPUProcess transfer memory ownership of its
IOSurfaces and Metal textures to the WebProcesses it is serving.
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::create):
(WebKit::GPUConnectionToWebProcess::GPUConnectionToWebProcess):
* GPUProcess/GPUConnectionToWebProcess.h:
(WebKit::GPUConnectionToWebProcess::webProcessIdentityToken const):
* GPUProcess/GPUProcess.cpp:
(WebKit::GPUProcess::createGPUConnectionToWebProcess):
* GPUProcess/GPUProcess.h:
* GPUProcess/GPUProcess.messages.in:
* Scripts/process-entitlements.sh:
* Shared/GPUProcessConnectionParameters.h: Copied from Source/WTF/wtf/MachSendRight.h.
(WebKit::GPUProcessConnectionParameters::encode const):
(WebKit::GPUProcessConnectionParameters::decode):
* UIProcess/GPU/GPUProcessProxy.cpp:
(WebKit::GPUProcessProxy::getGPUProcessConnection):
* UIProcess/GPU/GPUProcessProxy.h:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::getGPUProcessConnection):
* UIProcess/WebProcessPool.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::getGPUProcessConnection):
* UIProcess/WebProcessProxy.h:
* UIProcess/WebProcessProxy.messages.in:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::platformInitializeGPUProcessConnectionParameters):
(WebKit::WebProcess::getGPUProcessConnection):
* WebProcess/WebProcess.h:
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeGPUProcessConnectionParameters):
2021-02-22 Per Arne Vollan <pvollan@apple.com>
[Cocoa] Send sandbox extensions for Network Extension services in load parameters
https://bugs.webkit.org/show_bug.cgi?id=222284
<rdar://problem/74402532>
Reviewed by Brent Fulgham.
Currently, sandbox extensions for Network Extension services are sent from the UI process to the WebContent process when the
policy decision is made, but that is not soon enough in all cases. They should also be sent as part of the load parameters.
* Shared/Cocoa/LoadParametersCocoa.mm:
(WebKit::LoadParameters::platformEncode const):
(WebKit::LoadParameters::platformDecode):
* Shared/LoadParameters.h:
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::addPlatformLoadParameters):
* WebProcess/WebPage/Cocoa/WebPageCocoa.mm:
(WebKit::WebPage::platformDidReceiveLoadParameters):
2021-02-22 Chris Dumez <cdumez@apple.com>
[Cocoa] Add WKWebView session restoration API
https://bugs.webkit.org/show_bug.cgi?id=220958
<rdar://70956146>
Reviewed by Geoffrey Garen.
Add WKWebView session restoration API so that applications can easily retrieve
view state (such as back-forward list, scroll position, form data...) and
restore it on a WKWebView later on.
* UIProcess/API/Cocoa/WKWebView.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView interactionState]):
(-[WKWebView setInteractionState:]):
2021-02-22 Chris Dumez <cdumez@apple.com>
Use adoptNS() as soon as we [[ObjcClass alloc] init] to avoid leaks
https://bugs.webkit.org/show_bug.cgi?id=222237
Reviewed by Darin Adler.
* UIProcess/API/Cocoa/NSAttributedString.mm:
(globalConfiguration):
(+[_WKAttributedStringWebViewCache configuration]):
(+[_WKAttributedStringWebViewCache clearConfiguration]):
* UIProcess/API/Cocoa/WKNSURLAuthenticationChallenge.mm:
(-[WKNSURLAuthenticationChallenge _web_createTarget]):
* UIProcess/API/Cocoa/WKProcessPool.mm:
(sharedProcessPool):
(-[WKProcessPool encodeWithCoder:]):
(+[WKProcessPool _sharedProcessPool]):
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView loadSimulatedRequest:withResponseHTMLString:]):
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
(+[WKWebsiteDataStore allWebsiteDataTypes]):
* UIProcess/Cocoa/PreferenceObserver.mm:
(+[WKPreferenceObserver sharedInstance]):
* UIProcess/Cocoa/VideoFullscreenManagerProxy.h:
* UIProcess/Cocoa/VideoFullscreenManagerProxy.mm:
(WebKit::VideoFullscreenModelContext::createVideoFullscreenViewController):
* UIProcess/Cocoa/WebInspectorPreferenceObserver.mm:
(+[WKWebInspectorPreferenceObserver sharedInstance]):
* UIProcess/WebAuthentication/Cocoa/AuthenticatorPresenterCoordinator.mm:
(WebKit::AuthenticatorPresenterCoordinator::AuthenticatorPresenterCoordinator):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView supportedPasteboardTypesForCurrentSelection]):
(-[WKContentView assignLegacyDataForContextMenuInteraction]):
* UIProcess/ios/WKMouseDeviceObserver.mm:
(+[WKMouseDeviceObserver sharedInstance]):
* UIProcess/ios/WKStylusDeviceObserver.mm:
(+[WKStylusDeviceObserver sharedInstance]):
* UIProcess/mac/WKTextInputWindowController.mm:
(+[WKTextInputWindowController sharedTextInputWindowController]):
* UIProcess/mac/WebColorPickerMac.mm:
(+[WKPopoverColorWell _colorPopoverCreateIfNecessary:]):
* UIProcess/mac/WebDataListSuggestionsDropdownMac.mm:
(-[WKDataListSuggestionWindow initWithContentRect:styleMask:backing:defer:]):
* WebProcess/InjectedBundle/API/mac/WKDOMInternals.mm:
(WebKit::initWithImpl):
(WebKit::toWKDOMType):
* WebProcess/InjectedBundle/mac/InjectedBundleMac.mm:
(WebKit::InjectedBundle::initialize):
2021-02-22 Per Arne Vollan <pvollan@apple.com>
[macOS] Crash under AuxiliaryProcess::initializeSandbox
https://bugs.webkit.org/show_bug.cgi?id=222233
<rdar://problem/74261611>
Reviewed by Brent Fulgham.
When a WebKit client provides a user directory suffix in the process initialization parameters, confstr with the new
user suffix applied will fail to create the full directory path if it does not exist, and return an empty result.
This will lead to empty paths in the sandbox parameters, which will cause the sandbox to fail to compile, which will
eventually crash the WebKit process. This patch addresses this by making sure the new user directory suffix does not
represent a path, since confstr is not able to handle user directory suffixes containing paths that do not exist.
Additionally, this patch reverts r271417, which was the first attempt at fixing this crash, but was unsuccessful in
doing so.
* Shared/mac/AuxiliaryProcessMac.mm:
(WebKit::initializeSandboxParameters):
(WebKit::AuxiliaryProcess::initializeSandbox):
2021-02-22 Per Arne <pvollan@apple.com>
[macOS] Disabling relaunch on login for the WebContent process is racy
https://bugs.webkit.org/show_bug.cgi?id=222074
<rdar://problem/74230216>
Reviewed by Geoffrey Garen.
When NSApplication is being intialized, the method -[NSApplication disableRelaunchOnLogin] is dispatched on a non-main thread, which is in a race
with the revocation of the Launch Services sandbox extension. This patch addresses this by setting this information synchronously with Launch
Services while the sandbox extension is being held. Additionally, accessibility related code under WebPage::platformInitialize may reach out to
the Launch Services daemon once during the lifetime of the WebContent process. Make sure that call succeeds by making a synchronous call while
a sandbox extension to the Launch Services daemon is being held.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2021-02-22 Sam Sneddon <gsnedders@apple.com>
REGRESSION(r273204): webkitpy tests don't run on Py3
https://bugs.webkit.org/show_bug.cgi?id=222274
Reviewed by Jonathan Bedard.
This doesn't make them pass, but it does get them back to running to completion.
* Scripts/webkit/messages.py: avoid using a relative import
2021-02-22 Eric Carlson <eric.carlson@apple.com>
UserMediaPermissionRequestManagerProxy may be released while computing capture device list
https://bugs.webkit.org/show_bug.cgi?id=222236
<rdar://74480265>
Reviewed by Youenn Fablet.
* UIProcess/UserMediaPermissionRequestManagerProxy.cpp:
(WebKit::UserMediaPermissionRequestManagerProxy::computeFilteredDeviceList): Return early
if `weakThis` is NULL.
2021-02-22 Carlos Garcia Campos <cgarcia@igalia.com>
[SOUP] Add support for libsoup3
https://bugs.webkit.org/show_bug.cgi?id=222093
Reviewed by Adrian Perez de Castro.
* NetworkProcess/soup/NetworkDataTaskSoup.cpp:
(WebKit::NetworkDataTaskSoup::createRequest):
(WebKit::NetworkDataTaskSoup::clearRequest):
(WebKit::NetworkDataTaskSoup::resume):
(WebKit::NetworkDataTaskSoup::cancel):
(WebKit::NetworkDataTaskSoup::tlsConnectionAcceptCertificateCallback):
(WebKit::NetworkDataTaskSoup::acceptCertificateCallback):
(WebKit::NetworkDataTaskSoup::acceptCertificate):
(WebKit::NetworkDataTaskSoup::authenticateCallback):
(WebKit::NetworkDataTaskSoup::authenticate):
(WebKit::NetworkDataTaskSoup::continueAuthenticate):
(WebKit::shouldRedirectAsGET):
(WebKit::NetworkDataTaskSoup::continueHTTPRedirection):
(WebKit::NetworkDataTaskSoup::didGetHeaders):
(WebKit::NetworkDataTaskSoup::wroteBodyDataCallback):
(WebKit::NetworkDataTaskSoup::didWriteBodyData):
(WebKit::NetworkDataTaskSoup::networkEvent):
(WebKit::NetworkDataTaskSoup::hstsEnforced):
(WebKit::NetworkDataTaskSoup::didRestart):
* NetworkProcess/soup/NetworkDataTaskSoup.h:
* NetworkProcess/soup/NetworkSessionSoup.cpp:
(WebKit::NetworkSessionSoup::createWebSocketTask):
* NetworkProcess/soup/WebSocketTaskSoup.cpp:
(WebKit::isConnectionError):
(WebKit::WebSocketTask::WebSocketTask):
2021-02-20 Brent Fulgham <bfulgham@apple.com>
[Cocoa] Update WebCoreArgumentCodersMac with some new additions
https://bugs.webkit.org/show_bug.cgi?id=222242
<rdar://problem/74562457>
Reviewed by Tim Horton.
Will be tested by new API Tests in a subsequent patch.
* Shared/mac/WebCoreArgumentCodersMac.mm:
(IPC::ArgumentCoder<WebCore::ResourceRequest>::encodePlatformData):
(IPC::ArgumentCoder<WebCore::ResourceRequest>::decodePlatformData):
2021-02-20 BJ Burg <bburg@apple.com>
[Cocoa] Web Inspector: expose the WKBrowsingContextHandle associated with Web Inspector's page
https://bugs.webkit.org/show_bug.cgi?id=222226
Reviewed by Timothy Hatcher.
In order to implement browser.devtools.inspectedWindow.tabId for Web Extensions,
we need a way to do a reverse lookup of the tabId using the current browsing context
controller (WKWebProcessPlugInBrowserContextController) in the injected bundle.
This patch exposes the browsing context handle for the underlying WKWebView that
hosts the inspector page. Clients can use this to precompute a tabId for each
inspector browsing context that may be encountered by injected bundle code.
* UIProcess/API/Cocoa/_WKInspectorPrivate.h:
* UIProcess/API/Cocoa/_WKInspector.mm:
(-[_WKInspector handle]): Added.
* WebProcess/InjectedBundle/mac/InjectedBundleMac.mm:
(WebKit::InjectedBundle::classesForCoder):
Drive-by, make it possible to use WKBrowsingContextHandle in injected bundle parameters.
This is easy because it conforms to NSSecureCoding and is simply a pageId + frameId.
2021-02-20 Chris Fleizach <cfleizach@apple.com>
AX: Image should report the embedded accessibility description if available
https://bugs.webkit.org/show_bug.cgi?id=221875
<rdar://problem/74320584>
Reviewed by Jer Noble.
* WebProcess/GPU/media/RemoteImageDecoderAVF.h:
2021-02-20 Peng Liu <peng.liu6@apple.com>
Video elements do not work as sources for TexImage2D calls in GPU Process
https://bugs.webkit.org/show_bug.cgi?id=218184
Reviewed by Jer Noble.
Add an IPC message `RemoteGraphicsContextGL::CopyTextureFromMedia` to implement
`RemoteGraphicsContextGLProxy::copyTextureFromMedia()` in the Web process.
* GPUProcess/graphics/RemoteGraphicsContextGL.cpp:
(WebKit::RemoteGraphicsContextGL::copyTextureFromMedia):
* GPUProcess/graphics/RemoteGraphicsContextGL.h:
We should use `m_gpuConnectionToWebProcess` in the main thread only.
* GPUProcess/graphics/RemoteGraphicsContextGL.messages.in:
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.cpp:
(WebKit::RemoteGraphicsContextGLProxy::copyTextureFromMedia):
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.h:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.h:
2021-02-20 Youenn Fablet <youenn@apple.com>
Add support for adding/removing/clearing mock devices in GPUProcess
https://bugs.webkit.org/show_bug.cgi?id=222174
Reviewed by Eric Carlson.
Make sure to forward mock device orders to GPU process so that it is kept in sync with UIProcess.
Covered by updated test.
* GPUProcess/GPUProcess.cpp:
(WebKit::GPUProcess::addMockMediaDevice):
(WebKit::GPUProcess::clearMockMediaDevices):
(WebKit::GPUProcess::removeMockMediaDevice):
(WebKit::GPUProcess::resetMockMediaDevices):
* GPUProcess/GPUProcess.h:
* GPUProcess/GPUProcess.messages.in:
* UIProcess/GPU/GPUProcessProxy.cpp:
(WebKit::GPUProcessProxy::addMockMediaDevice):
(WebKit::GPUProcessProxy::clearMockMediaDevices):
(WebKit::GPUProcessProxy::removeMockMediaDevice):
(WebKit::GPUProcessProxy::resetMockMediaDevices):
* UIProcess/GPU/GPUProcessProxy.h:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::addMockMediaDevice):
(WebKit::WebProcessPool::clearMockMediaDevices):
(WebKit::WebProcessPool::removeMockMediaDevice):
(WebKit::WebProcessPool::resetMockMediaDevices):
2021-02-20 Jiewen Tan <jiewen_tan@apple.com>
PCM: Store and report source unlinkable tokens
https://bugs.webkit.org/show_bug.cgi?id=222208
<rdar://problem/73582032>
Reviewed by John Wilander.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::setFraudPreventionValuesForTesting):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/NetworkSession.cpp:
(WebKit::NetworkSession::setFraudPreventionValuesForTesting):
* NetworkProcess/NetworkSession.h:
* NetworkProcess/PrivateClickMeasurementManager.cpp:
(WebKit::PrivateClickMeasurementManager::storeUnattributed):
(WebKit::PrivateClickMeasurementManager::getSignedSecretToken):
(WebKit::PrivateClickMeasurementManager::setFraudPreventionValuesForTesting):
* NetworkProcess/PrivateClickMeasurementManager.h:
* UIProcess/API/C/WKPage.cpp:
(WKPageSetFraudPreventionValuesForTesting):
* UIProcess/API/C/WKPagePrivate.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::markPrivateClickMeasurementsAsExpiredForTesting):
(WebKit::WebPageProxy::setFraudPreventionValuesForTesting):
* UIProcess/WebPageProxy.h:
Adds infrastructure to mock the crypto operations such that layout tests can be
conducted without the server side crypto support.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::buildPrivateClickMeasurementFromDatabase):
(WebKit::ResourceLoadStatisticsDatabaseStore::insertPrivateClickMeasurement):
* NetworkProcess/PrivateClickMeasurementManager.cpp:
(WebKit::PrivateClickMeasurementManager::storeUnattributed):
(WebKit::PrivateClickMeasurementManager::getTokenPublicKey):
(WebKit::PrivateClickMeasurementManager::getSignedSecretToken):
(WebKit::PrivateClickMeasurementManager::fireConversionRequest):
(WebKit::PrivateClickMeasurementManager::fireConversionRequestImpl):
* NetworkProcess/PrivateClickMeasurementManager.h:
Connects the fraud prevention feature in PCM to generate, store, and report the unlinkable token.
2021-02-20 Kimmo Kinnunen <kkinnunen@apple.com>
WebGL GPU process IPC should use shared memory for asynchronous messages
https://bugs.webkit.org/show_bug.cgi?id=219641
<rdar://problem/72340651>
Reviewed by Geoff Garen.
Implement a concept of a message buffer. This is a similar concept as
"command buffer" or "display list" in some graphics APIs. Messages
can be added to the buffer by a client and then played back by a
server.
Current implementation adds a specific pseudo-infinite,
semi-duplex message buffer type IPC::StreamConnectionBuffer. This
is a buffer that can be attached to stream connection, an object
that can deliver IPC messages either through the message buffer
or through normal IPC::Connection. The stream connection is
implemented as IPC::StreamClientConnection and IPC::StreamServerConnection.
The connection has server and client distinction, similar to
IPC::Connection. Unlike IPC::Connection, a stream connection is asymmetric
in features: the client can send messages and receive replies, where
as the server can receive messages and send replies.
Messages received through IPC::StreamServerConnection can be
processed by StreamConnectionWorkQueue.
Currently asynchronous messages are added to the stream. Synchronous messages
and their replies are dispatched through normal IPC.
WebGL is implemented by having one StreamConnectionWorkQueue for all WebGL
in the browser. Currently each web process WebGL context in has one
StreamClientConnection attached to one 2mb StreamConnectionBuffer.
Later on, there will be only one WebGL stream connection and connection
buffer shared between all contexts in particular thread in Web process.
The implementation is "optimized" for the later on case. Due to single-
process nature of each page, it does not make much sense to have buffer
per context. However, it is anticipated that many of the successive calls
are to same context. The SHM message format implements an optimization
where the destination id is maintained as part of the stream connection
state.
In GPU process side, all WebGL is processed in one task queue,
"RemoteGraphicsContextGL task queue". It will process up to 1000 commands
per connection per iteration.
Implemented as follows:
Sender shares the shared memory area with the destination when creating
the destination.
If the message fits to the stream shared memory area, write it there.
Otherwise send it via normal IPC as out of line message.
Messages to the destination arrive as before, in order. For each out of line
message, an entry is pushed to the stream. Receiver will stop processing the
stream and processing will continue once the out-of-line
message has been processed.
Sender can write messages to the stream simultaneous to the receiver reading
the messages. This is implemented as a ring buffer with atomic sender and receiver
pointers.
Stream connection processing is implemented as a shared semaphore. This semaphore
is notified if the receiver is waiting on it.
Performance impact (iMac Pro):
- MotionMark triangles: 300 points -> 4000 points
(~7000 points no-GPUP)
- WebGL samples Aquarium: 1000 fish ~60fps -> 5000 fish ~60fps
(~5000 fish with no-GPUP)
No new tests, tested by existing tests.
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::createGraphicsContextGL):
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/GPUConnectionToWebProcess.messages.in:
* GPUProcess/graphics/RemoteGraphicsContextGL.cpp:
(WebKit::remoteGraphicsContextGLStreamWorkQueue):
(WebKit::RemoteGraphicsContextGL::create):
(WebKit::RemoteGraphicsContextGL::RemoteGraphicsContextGL):
(WebKit::RemoteGraphicsContextGL::~RemoteGraphicsContextGL):
(WebKit::RemoteGraphicsContextGL::initialize):
(WebKit::RemoteGraphicsContextGL::connectWebProcessConnection):
(WebKit::RemoteGraphicsContextGL::disconnectWebProcessConnection):
(WebKit::RemoteGraphicsContextGL::workQueueInitialize):
(WebKit::RemoteGraphicsContextGL::workQueueUninitialize):
Initialize the OpenGL context in the work queue, since OpenGL
has various limitations wrt thread mobility.
(WebKit::RemoteGraphicsContextGL::forceContextLost):
(WebKit::RemoteGraphicsContextGL::dispatchContextChangedNotification):
* GPUProcess/graphics/RemoteGraphicsContextGL.h:
(WebKit::RemoteGraphicsContextGL::platformWorkQueueInitialize):
(WebKit::RemoteGraphicsContextGL::send const):
* GPUProcess/graphics/RemoteGraphicsContextGL.messages.in:
* GPUProcess/graphics/RemoteGraphicsContextGLCocoa.cpp:
(WebKit::RemoteGraphicsContextGL::create):
(WebKit::RemoteGraphicsContextGLCocoa::RemoteGraphicsContextGLCocoa):
(WebKit::RemoteGraphicsContextGLCocoa::platformWorkQueueInitialize):
* NetworkProcess/webrtc/NetworkRTCProvider.h:
* Platform/IPC/ArgumentCoder.h:
* Platform/IPC/ArgumentCoders.cpp:
* Platform/IPC/ArgumentCoders.h:
Make it possible to use multiple different Encoder classes to encode
the IPC data.
* Platform/IPC/Connection.cpp:
(IPC::Connection::processIncomingMessage):
(IPC::Connection::dispatchDidReceiveInvalidMessage):
* Platform/IPC/Connection.h:
Add possibility for the stream decoding logic to send the didReceiveInvalidMessage
notification to the connection client.
* Platform/IPC/Decoder.cpp:
(IPC::Decoder::Decoder):
(IPC::m_bufferDeallocator):
(IPC::m_destinationID):
* Platform/IPC/Decoder.h:
Add a constructor for Decoder for constructing a decoder in place in
the stream buffer.
* Platform/IPC/HandleMessage.h:
(IPC::handleMessageSynchronous):
Adds a stream-specific handleMessageSynchronous.
The old ones are unneccessarily complex, as they need redundant calls through redundant
generated send functions.
The old ones need the reply encoder passed in, which is redundant and problematic
for the upcoming case where the stream messaging will mainly reply through the stream itself,
constructing the IPC::StreamConnectionEncoder instead of normal IPC::Encoder.
* Platform/IPC/StreamClientConnection.cpp: Copied from Source/WebKit/Platform/IPC/cocoa/MachPort.h.
(IPC::StreamClientConnection::StreamClientConnection):
(IPC::StreamClientConnection::setWakeUpSemaphore):
Add function to set the connection receive side wakeup semaphore.
Currently the semaphore is the semaphore that StreamConnectionWorkQueue
waits on. All streams processed by the same work queue will receive the same
semaphore and will notify it when they have written new data.
(IPC::StreamClientConnection::wakeUpReceiver):
* Platform/IPC/StreamClientConnection.h: Added.
(IPC::StreamClientConnection::send):
(IPC::StreamClientConnection::sendSync):
(IPC::StreamClientConnection::trySendDestinationIDIfNeeded):
(IPC::StreamClientConnection::sendProcessOutOfStreamMessage):
(IPC::StreamClientConnection::tryAcquire):
(IPC::StreamClientConnection::release):
(IPC::StreamClientConnection::alignedSpan):
(IPC::StreamClientConnection::size):
(IPC::StreamClientConnection::clampedLimit const):
Add the implementation for sending data through the shared memory.
* Platform/IPC/StreamConnectionBuffer.cpp: Added.
(IPC::createMemory):
(IPC::StreamConnectionBuffer::StreamConnectionBuffer):
(IPC::StreamConnectionBuffer::operator=):
(IPC::StreamConnectionBuffer::encode const):
(IPC::StreamConnectionBuffer::decode):
* Platform/IPC/StreamConnectionBuffer.h: Added.
(IPC::StreamConnectionBuffer::wrapOffset const):
(IPC::StreamConnectionBuffer::alignOffset const):
(IPC::StreamConnectionBuffer::senderOffset):
(IPC::StreamConnectionBuffer::receiverOffset):
(IPC::StreamConnectionBuffer::data const):
(IPC::StreamConnectionBuffer::dataSize const):
(IPC::StreamConnectionBuffer::senderWaitSemaphore):
(IPC::StreamConnectionBuffer::header const):
(IPC::StreamConnectionBuffer::headerSize):
Add the common implementation for the shared memory ring buffer.
This is the common implementation used by both the sender side
as well as the receiver side.
Due to the selection of how the size is marked up, the
common implementation does not contain all the non-trivial code.
The "algorithm" uses a "hole indicator" to differentiate between
cases of "buffer is empty" and "buffer is full". The other
alternative could be better and maybe explored later.
* Platform/IPC/StreamConnectionEncoder.h: Added.
* Platform/IPC/StreamConnectionWorkQueue.cpp: Added.
(IPC::StreamConnectionWorkQueue::StreamConnectionWorkQueue):
(IPC::StreamConnectionWorkQueue::dispatch):
(IPC::StreamConnectionWorkQueue::addStreamConnection):
(IPC::StreamConnectionWorkQueue::removeStreamConnection):
(IPC::StreamConnectionWorkQueue::stop):
(IPC::StreamConnectionWorkQueue::wakeUp):
(IPC::StreamConnectionWorkQueue::wakeUpSemaphore):
(IPC::StreamConnectionWorkQueue::wakeUpProcessingThread):
(IPC::StreamConnectionWorkQueue::processStreams):
Work queue contains multiple StreamServerConnection instances.
For each iteration of the work queue thread, it will process
1000 items from each connection. If none of the connections contain
work, the queue will sleep on the work queue semaphore.
The senders will send through StreamClientConnection which
will wake up the work queue by signaling the semaphore.
* Platform/IPC/StreamConnectionWorkQueue.h: Copied from Source/WebKit/Platform/IPC/cocoa/MachPort.h.
* Platform/IPC/StreamServerConnection.cpp: Added.
(IPC::StreamServerConnectionBase::StreamServerConnectionBase):
(IPC::StreamServerConnectionBase::startReceivingMessagesImpl):
(IPC::StreamServerConnectionBase::stopReceivingMessagesImpl):
(IPC::StreamServerConnectionBase::enqueueMessage):
(IPC::StreamServerConnectionBase::tryAquire):
(IPC::StreamServerConnectionBase::release):
(IPC::StreamServerConnectionBase::alignedSpan):
(IPC::StreamServerConnectionBase::size):
(IPC::StreamServerConnectionBase::clampedLimit const):
* Platform/IPC/StreamServerConnection.h: Added.
(IPC::StreamServerConnectionBase::connection):
(IPC::StreamServerConnectionBase::wrapOffset const):
(IPC::StreamServerConnectionBase::alignOffset const):
(IPC::StreamServerConnectionBase::sharedSenderOffset):
(IPC::StreamServerConnectionBase::sharedReceiverOffset):
(IPC::StreamServerConnectionBase::data const):
(IPC::StreamServerConnectionBase::dataSize const):
(IPC::StreamServerConnectionBase::sendSyncReply):
(IPC::StreamServerConnection<Receiver>::startReceivingMessages):
(IPC::StreamServerConnection<Receiver>::stopReceivingMessages):
(IPC::StreamServerConnection<Receiver>::dispatchStreamMessages):
(IPC::StreamServerConnection<Receiver>::processSetStreamDestinationID):
(IPC::StreamServerConnection<Receiver>::dispatchStreamMessage):
(IPC::StreamServerConnection<Receiver>::dispatchOutOfStreamMessage):
StreamServerConnection will process the message buffer and dispatch
messages based on the buffer contents. The class is a template so
that the message data decode switch (generated by the normal IPC
generator) is perhaps inlined to the message unpack loop.
* Platform/IPC/cocoa/MachPort.h:
* Scripts/webkit/messages.py:
* Scripts/webkit/messages_unittest.py:
Implements a 'Stream' attribute for the message receiver. When marked as
'Stream', the receiver message handler is didReceiveStreamMessage and it
will be called for all messages, stream or normal IPC, sync or async.
* Scripts/webkit/model.py:
Adds IPC control message ProcessOutOfStreamMessage that is acted on if
it is part of the stream messages. It means that the stream processing
should stop until one normal IPC message, sync or non-synch, should be
processed.
Adds IPC control message SetStreamDestinationID that is acted on if it
is part of the stream messages. It will set the destination ID for all
the subsequent stream messages, until another SetStreamDestinationID
message is seen.
* Scripts/webkit/tests/Makefile:
* Scripts/webkit/tests/MessageNames.cpp:
(IPC::description):
(IPC::receiverName):
(IPC::isValidMessageName):
* Scripts/webkit/tests/MessageNames.h:
* Scripts/webkit/tests/TestWithStream.messages.in: Copied from Source/WebKit/WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.messages.in.
* Scripts/webkit/tests/TestWithStreamBuffer.messages.in: Copied from Source/WebKit/WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.messages.in.
* Scripts/webkit/tests/TestWithStreamBufferMessageReceiver.cpp: Added.
(WebKit::TestWithStreamBuffer::didReceiveMessage):
* Scripts/webkit/tests/TestWithStreamBufferMessages.h: Added.
(Messages::TestWithStreamBuffer::messageReceiverName):
(Messages::TestWithStreamBuffer::SendStreamBuffer::name):
(Messages::TestWithStreamBuffer::SendStreamBuffer::SendStreamBuffer):
(Messages::TestWithStreamBuffer::SendStreamBuffer::arguments const):
* Scripts/webkit/tests/TestWithStreamBufferMessagesReplies.h: Added.
* Scripts/webkit/tests/TestWithStreamMessageReceiver.cpp: Added.
(WebKit::TestWithStream::didReceiveStreamMessage):
* Scripts/webkit/tests/TestWithStreamMessages.h: Added.
(Messages::TestWithStream::messageReceiverName):
(Messages::TestWithStream::SendString::name):
(Messages::TestWithStream::SendString::SendString):
(Messages::TestWithStream::SendString::arguments const):
(Messages::TestWithStream::SendStringSynchronized::name):
(Messages::TestWithStream::SendStringSynchronized::SendStringSynchronized):
(Messages::TestWithStream::SendStringSynchronized::arguments const):
* Scripts/webkit/tests/TestWithStreamMessagesReplies.h: Added.
* Sources.txt:
* UIProcess/Downloads/DownloadProxy.h:
* UIProcess/ProvisionalPageProxy.cpp:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.cpp:
(WebKit::RemoteGraphicsContextGLProxy::RemoteGraphicsContextGLProxy):
(WebKit::RemoteGraphicsContextGLProxy::wasCreated):
Send the work queue wakeup semaphore as a result of the context creation.
* Platform/IPC/ScopedActiveMessageReceiveQueue.h: Added.
(IPC::ScopedActiveMessageReceiveQueue::ScopedActiveMessageReceiveQueue):
(IPC::ScopedActiveMessageReceiveQueue::operator=):
(IPC::ScopedActiveMessageReceiveQueue::~ScopedActiveMessageReceiveQueue):
(IPC::ScopedActiveMessageReceiveQueue::reset):
(IPC::ScopedActiveMessageReceiveQueue::get const):
(IPC::ScopedActiveMessageReceiveQueue::stopListeningForIPCAndRelease):
Add a holder to replace previous RemoteRenderingBackendWrapper.
Now other classes can be held similarly.
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.h:
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.messages.in:
* WebProcess/Network/webrtc/LibWebRTCNetwork.h:
2021-02-20 Wenson Hsieh <wenson_hsieh@apple.com>
Unreviewed, reverting r273115.
Breaks autocorrect without the accompanying change in
rdar://problem/74211293
Reverted changeset:
"Unconditionally return information in _autofillContext SPI
when a field is focused"
https://bugs.webkit.org/show_bug.cgi?id=221828
https://commits.webkit.org/r273115
2021-02-20 Chris Dumez <cdumez@apple.com>
Unreviewed, fix regression from r273194 that was spotted by Darin Adler after landing.
* Platform/cocoa/WKPaymentAuthorizationDelegate.mm:
(-[WKPaymentAuthorizationDelegate completePaymentMethodSelection:]):
(-[WKPaymentAuthorizationDelegate completeShippingContactSelection:]):
(-[WKPaymentAuthorizationDelegate completeShippingMethodSelection:]):
2021-02-19 Chris Dumez <cdumez@apple.com>
Crash under Decoder::Decoder()
https://bugs.webkit.org/show_bug.cgi?id=222192
<rdar://31392681>
Reviewed by Geoffrey Garen.
We are sometimes crashing under Decoder's copyBuffer(), inside the memcpy()
call, with a null address. I have no idea how this is happening and this
code has not changed in a long time so I have made the following hardening:
1. Update copyBuffer() to use tryFastMalloc() instead of fastMalloc(). Log
and return null if tryFastMalloc() failed instead of calling memcpy().
2. Update Decoder::create() to log and return early if the input buffer
is null.
3. Update Connection's createMessageDecoder() to use CheckedSize when
computing the bodySize that is being passed to Decoder::create(). If
we overflow, log and return null.
No new tests, no idea how this can happen in practice.
* Platform/IPC/Decoder.cpp:
(IPC::copyBuffer):
(IPC::Decoder::create):
(IPC::Decoder::Decoder):
* Platform/IPC/cocoa/ConnectionCocoa.mm:
(IPC::createMessageDecoder):
2021-02-19 BJ Burg <bburg@apple.com>
[Cocoa] WKBrowsingContextHandle should conform to NSCopying
https://bugs.webkit.org/show_bug.cgi?id=222214
Reviewed by Devin Rousso.
* Shared/API/Cocoa/WKBrowsingContextHandle.h:
* Shared/API/Cocoa/WKBrowsingContextHandle.mm:
(-[WKBrowsingContextHandle copyWithZone:]):
2021-02-19 Chris Dumez <cdumez@apple.com>
Review remaining usage of autorelease to make sure it is necessary
https://bugs.webkit.org/show_bug.cgi?id=222177
Reviewed by Geoffrey Garen.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::NetworkSessionCocoa):
* Platform/cocoa/WKPaymentAuthorizationDelegate.mm:
(-[WKPaymentAuthorizationDelegate completePaymentMethodSelection:]):
(-[WKPaymentAuthorizationDelegate completeShippingContactSelection:]):
(-[WKPaymentAuthorizationDelegate completeShippingMethodSelection:]):
* UIProcess/API/Cocoa/NSAttributedString.mm:
(+[_WKAttributedStringWebViewCache configuration]):
(+[NSAttributedString _loadFromHTMLWithOptions:contentLoader:completionHandler:]):
* UIProcess/API/Cocoa/WKBrowsingContextController.mm:
(setUpPagePolicyClient):
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _takePDFSnapshotWithConfiguration:completionHandler:]):
* UIProcess/API/ios/WKWebViewTestingIOS.mm:
(-[WKWebView _requestTextInputContextsInRect:completionHandler:]):
* WebProcess/Plugins/Netscape/mac/NetscapePluginMac.mm:
(WebKit::convertToNPNNString):
* WebProcess/WebPage/mac/WKAccessibilityWebPageObjectMac.mm:
(-[WKAccessibilityWebPageObject ALLOW_DEPRECATED_IMPLEMENTATIONS_END]):
2021-02-19 Devin Rousso <drousso@apple.com>
[Payment Request] add an `object data` to `PaymentShippingOption` so that data specific to Apple Pay can be provided
https://bugs.webkit.org/show_bug.cgi?id=222211
<rdar://problem/72916508>
Reviewed by Tim Horton.
* Shared/ApplePay/cocoa/WebPaymentCoordinatorProxyCocoa.mm:
(WebKit::toPKShippingMethod):
* Platform/cocoa/WKPaymentAuthorizationDelegate.mm:
(toShippingMethod):
2021-02-19 Wenson Hsieh <wenson_hsieh@apple.com>
Unreviewed, fix some internal builds after r273184
Add a forward declaration for this (recently added) SPI, which we can safely revert once internal automation is
using an SDK with the fix for <rdar://problem/73852335>.
* Platform/spi/ios/UIKitSPI.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _dataOwnerForPasteboard:]):
Additionally make this binary compatible with older versions of the system by adding a `-respondsToSelector:`
check before attempting to use the declared SPI.
2021-02-19 Kate Cheney <katherine_cheney@apple.com>
Add back old media playback API but deprecated
https://bugs.webkit.org/show_bug.cgi?id=222206
<rdar://problem/74539125>
We should add these back in case clients adopted them before the name
update in https://trac.webkit.org/changeset/272923/webkit.
Reviewed by Tim Horton.
* UIProcess/API/Cocoa/WKWebView.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView pauseAllMediaPlayback:]):
(-[WKWebView suspendAllMediaPlayback:]):
(-[WKWebView resumeAllMediaPlayback:]):
(-[WKWebView closeAllMediaPresentations:]):
(-[WKWebView requestMediaPlaybackState:]):
2021-02-19 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] Specify a _UIDataOwner when reading or writing from the system pasteboard
https://bugs.webkit.org/show_bug.cgi?id=222072
<rdar://problem/74208576>
Reviewed by Devin Rousso.
Adds WebKit2 support for tagging `PlatformPasteboard` with the content view's pasteboard data owner type. See
below for more details.
* Platform/spi/ios/UIKitSPI.h:
Import `UIPasteboard_Private.h` so that we know about `+_performAsDataOwner:block:` in the API tests.
* UIProcess/Cocoa/WebPasteboardProxyCocoa.mm:
(WebKit::WebPasteboardProxy::getPasteboardTypes):
(WebKit::WebPasteboardProxy::getPasteboardPathnamesForType):
(WebKit::WebPasteboardProxy::getPasteboardStringForType):
(WebKit::WebPasteboardProxy::getPasteboardStringsForType):
(WebKit::WebPasteboardProxy::getPasteboardBufferForType):
(WebKit::WebPasteboardProxy::getPasteboardChangeCount):
(WebKit::WebPasteboardProxy::getPasteboardColor):
(WebKit::WebPasteboardProxy::getPasteboardURL):
(WebKit::WebPasteboardProxy::addPasteboardTypes):
(WebKit::WebPasteboardProxy::setPasteboardTypes):
(WebKit::WebPasteboardProxy::setPasteboardURL):
(WebKit::WebPasteboardProxy::setPasteboardColor):
(WebKit::WebPasteboardProxy::setPasteboardStringForType):
(WebKit::WebPasteboardProxy::containsURLStringSuitableForLoading):
Make all of these methods use `dataOwner()` with either a `Read` or `Write` pasteboard access intent to grab the
data owner type from the view corresponding to the page identifier, and then wrap all pasteboard access within
`PlatformPasteboard`'s new `performAsDataOwner` static helper.
(WebKit::WebPasteboardProxy::urlStringSuitableForLoading):
(WebKit::WebPasteboardProxy::setPasteboardBufferForType):
(WebKit::WebPasteboardProxy::getNumberOfFiles):
(WebKit::WebPasteboardProxy::typesSafeForDOMToReadAndWrite):
(WebKit::WebPasteboardProxy::writeCustomData):
(WebKit::WebPasteboardProxy::allPasteboardItemInfo):
(WebKit::WebPasteboardProxy::informationForItemAtIndex):
(WebKit::WebPasteboardProxy::getPasteboardItemsCount):
(WebKit::WebPasteboardProxy::readStringFromPasteboard):
(WebKit::WebPasteboardProxy::readURLFromPasteboard):
(WebKit::WebPasteboardProxy::readBufferFromPasteboard):
(WebKit::WebPasteboardProxy::containsStringSafeForDOMToReadForType):
(WebKit::WebPasteboardProxy::writeURLToPasteboard):
(WebKit::WebPasteboardProxy::writeWebContentToPasteboard):
(WebKit::WebPasteboardProxy::writeImageToPasteboard):
(WebKit::WebPasteboardProxy::writeStringToPasteboard):
(WebKit::WebPasteboardProxy::updateSupportedTypeIdentifiers):
(WebKit::WebPasteboardProxy::determineDataOwner const):
Use the given page identifier in various pasteboard IPC endpoints in `WebPasteboardProxy` to ask `WebPageProxy`
for its pasteboard data owner type when reading or writing pasteboard data.
* UIProcess/PageClient.h:
Add a client hook to return a `DataOwnerType` enum, given a `PasteboardAccessIntent` -- that is, copy (write) or
paste (read).
(WebKit::PageClient::dataOwnerForPasteboard const):
* UIProcess/PasteboardAccessIntent.h: Added.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::dataOwnerForPasteboard const):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPasteboardProxy.h:
* UIProcess/WebPasteboardProxy.messages.in:
* UIProcess/ios/PageClientImplIOS.h:
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::dataOwnerForPasteboard const):
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(coreDataOwnerType):
(-[WKContentView _dataOwnerForPasteboard:]):
Return `-_dataOwnerForPaste` or `-_dataOwnerForCopy` (depending on the `PasteboardAccessIntent`).
* WebKit.xcodeproj/project.pbxproj:
2021-02-19 Chris Dumez <cdumez@apple.com>
Make format of release logging more consistent to facilitate grepping
https://bugs.webkit.org/show_bug.cgi?id=222200
Reviewed by Simon Fraser.
* NetworkProcess/NetworkCORSPreflightChecker.cpp:
(WebKit::NetworkCORSPreflightChecker::didCompleteWithError):
* NetworkProcess/NetworkLoadChecker.cpp:
(WebKit::NetworkLoadChecker::checkCORSRequestWithPreflight):
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::didReceiveMessage):
(WebKit::NetworkProcess::didReceiveSyncMessage):
(WebKit::NetworkProcess::initializeNetworkProcess):
(WebKit::NetworkProcess::prepareToSuspend):
* NetworkProcess/PingLoad.cpp:
(WebKit::PingLoad::didReceiveResponse):
(WebKit::PingLoad::didCompleteWithError):
* NetworkProcess/PreconnectTask.cpp:
(WebKit::PreconnectTask::didFailLoading):
* NetworkProcess/ServiceWorker/WebSWServerConnection.cpp:
(WebKit::WebSWServerConnection::startFetch):
* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::NetworkDataTaskCocoa::NetworkDataTaskCocoa):
(WebKit::NetworkDataTaskCocoa::willPerformHTTPRedirection):
* Platform/IPC/Connection.cpp:
(IPC::Connection::waitForSyncReply):
* UIProcess/ApplicationStateTracker.mm:
(WebKit::ApplicationStateTracker::ApplicationStateTracker):
* UIProcess/AuxiliaryProcessProxy.cpp:
(WebKit::AuxiliaryProcessProxy::terminate):
* UIProcess/ProcessThrottler.cpp:
(WebKit::ProcessThrottler::setAssertionType):
(WebKit::ProcessThrottler::updateAssertionIfNeeded):
(WebKit::ProcessThrottler::sendPrepareToSuspendIPC):
* UIProcess/ProvisionalPageProxy.cpp:
(WebKit::ProvisionalPageProxy::ProvisionalPageProxy):
(WebKit::ProvisionalPageProxy::processDidTerminate):
(WebKit::ProvisionalPageProxy::cancel):
(WebKit::ProvisionalPageProxy::loadData):
(WebKit::ProvisionalPageProxy::loadRequest):
(WebKit::ProvisionalPageProxy::goToBackForwardItem):
(WebKit::ProvisionalPageProxy::didCreateMainFrame):
(WebKit::ProvisionalPageProxy::didStartProvisionalLoadForFrame):
(WebKit::ProvisionalPageProxy::didFailProvisionalLoadForFrame):
(WebKit::ProvisionalPageProxy::didCommitLoadForFrame):
(WebKit::ProvisionalPageProxy::didCreateContextForVisibilityPropagation):
* UIProcess/WebBackForwardCache.cpp:
(WebKit::WebBackForwardCache::addEntry):
(WebKit::WebBackForwardCache::removeEntry):
(WebKit::WebBackForwardCache::takeSuspendedPage):
* UIProcess/WebBackForwardCacheEntry.cpp:
(WebKit::WebBackForwardCacheEntry::expirationTimerFired):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::receivedNavigationPolicyDecision):
(WebKit::WebPageProxy::commitProvisionalPage):
(WebKit::WebPageProxy::continueNavigationInNewProcess):
(WebKit::WebPageProxy::didStartProvisionalLoadForFrameShared):
(WebKit::WebPageProxy::didReceiveServerRedirectForProvisionalLoadForFrameShared):
(WebKit::WebPageProxy::willPerformClientRedirectForFrame):
(WebKit::WebPageProxy::didCancelClientRedirectForFrame):
(WebKit::WebPageProxy::didFailProvisionalLoadForFrameShared):
(WebKit::WebPageProxy::didCommitLoadForFrame):
(WebKit::WebPageProxy::didFinishDocumentLoadForFrame):
(WebKit::WebPageProxy::didFinishLoadForFrame):
(WebKit::WebPageProxy::didFailLoadForFrame):
(WebKit::WebPageProxy::didSameDocumentNavigationForFrame):
(WebKit::WebPageProxy::decidePolicyForNavigationAction):
(WebKit::WebPageProxy::didPerformClientRedirectShared):
(WebKit::WebPageProxy::dispatchProcessDidTerminate):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::gpuProcessCrashed):
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::updateServiceWorkerProcessAssertion):
* UIProcess/ios/ProcessAssertionIOS.mm:
(-[WKProcessAssertionBackgroundTaskManager _updateBackgroundTask]):
(-[WKProcessAssertionBackgroundTaskManager _handleBackgroundTaskExpiration]):
(-[WKProcessAssertionBackgroundTaskManager _handleBackgroundTaskExpirationOnMainThread]):
(WebKit::ProcessAssertion::ProcessAssertion):
(WebKit::ProcessAssertion::~ProcessAssertion):
(WebKit::ProcessAssertion::processAssertionWasInvalidated):
* UIProcess/ios/WKApplicationStateTrackingView.mm:
(-[WKApplicationStateTrackingView willMoveToWindow:]):
(-[WKApplicationStateTrackingView didMoveToWindow]):
* UIProcess/ios/WKContentView.mm:
(-[WKContentView _setupVisibilityPropagationView]):
(-[WKContentView _setupVisibilityPropagationViewForGPUProcess]):
* WebAuthnProcess/WebAuthnProcess.cpp:
(WebKit::WebAuthnProcess::prepareToSuspend):
* WebProcess/WebCoreSupport/WebResourceLoadObserver.cpp:
(WebKit::WebResourceLoadObserver::logUserInteractionWithReducedTimeResolution):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::didReceivePolicyDecision):
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::initializeWebProcess):
(WebKit::WebProcess::prepareToSuspend):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::updateProcessName):
(WebKit::WebProcess::updateFreezerStatus):
2021-02-19 Alex Christensen <achristensen@webkit.org>
NSError returned by WebKit API should actually conform to NSSecureCoding
https://bugs.webkit.org/show_bug.cgi?id=222204
<rdar://problem/63893583>
Reviewed by Chris Dumez.
When an NSError is encoded, its userInfo is encoded, which means everything we put in userInfo should be encodable.
WKReloadFrameErrorRecoveryAttempter was not, so let's make it! When encoded, it will just encode nothing.
When decoded, it will produce a WKReloadFrameErrorRecoveryAttempter that does nothing, which is valid behavior,
and more desirable than crashing.
* UIProcess/Cocoa/WKReloadFrameErrorRecoveryAttempter.h:
* UIProcess/Cocoa/WKReloadFrameErrorRecoveryAttempter.mm:
(-[WKReloadFrameErrorRecoveryAttempter encodeWithCoder:]):
(-[WKReloadFrameErrorRecoveryAttempter initWithCoder:]):
(+[WKReloadFrameErrorRecoveryAttempter supportsSecureCoding]):
2021-02-19 Alex Christensen <achristensen@webkit.org>
CrashTracer: com.apple.WebKit.Networking at WebKit: WTF::Detail::CallableWrapper<WebKit::PrivateClickMeasurementManager::firePendingAttributionRequests()
https://bugs.webkit.org/show_bug.cgi?id=222201
Reviewed by Chris Dumez.
* NetworkProcess/PrivateClickMeasurementManager.cpp:
(WebKit::PrivateClickMeasurementManager::attribute):
(WebKit::PrivateClickMeasurementManager::firePendingAttributionRequests):
2021-02-19 Per Arne Vollan <pvollan@apple.com>
[iOS] Remove access to IOSurfaceAcceleratorClient
https://bugs.webkit.org/show_bug.cgi?id=222199
<rdar://problem/73518610>
Reviewed by Brent Fulgham.
Remove access to IOSurfaceAcceleratorClient in the WebContent process on iOS when the GPU process is enabled.
This patch also adds the IOKit filter from IOSurfaceRootUserClient when the GPU process is not enabled, to
the rule for IOSurfaceRootUserClient when the GPU process is enabled.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2021-02-19 Jean-Yves Avenard <jya@apple.com>
Legacy CDM doesn't work when GPU media process is enabled
https://bugs.webkit.org/show_bug.cgi?id=222163
rdar://74254335
Reviewed by Eric Carlson.
We don't have an infrastructure to generate CDM keys
* WebProcess/GPU/media/RemoteLegacyCDMSession.cpp:
(WebKit::RemoteLegacyCDMSession::update):
Don't unditionally report that an error occurred, but the result of the actual proxied operation in the content process.
2021-02-19 Per Arne Vollan <pvollan@apple.com>
[iOS] Partially revert r272973
https://bugs.webkit.org/show_bug.cgi?id=222168
Unreviewed, partially reverting r272973.
Partially revert r272973, since it appears to have caused a Safari launch regression.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2021-02-19 Devin Rousso <drousso@apple.com>
[Payment Request] add support for Apple Pay payment method mode
https://bugs.webkit.org/show_bug.cgi?id=222128
<rdar://problem/72320278>
Reviewed by Wenson Hsieh.
Handle the delegate call from `PaymentAuthorization(View)Controller` by passing it along to
the WebProcess to be handled by WebCore. Handle the response sent by the WebProcess by
invoking the callback that was provided by the delegate call (which is saved to an ivar).
* WebProcess/ApplePay/WebPaymentCoordinator.messages.in:
* WebProcess/ApplePay/WebPaymentCoordinator.h:
* WebProcess/ApplePay/WebPaymentCoordinator.cpp:
(WebKit::WebPaymentCoordinator::completePaymentMethodModeChange): Added.
(WebKit::WebPaymentCoordinator::didChangePaymentMethodMode): Added.
* Shared/ApplePay/WebPaymentCoordinatorProxy.messages.in:
* Shared/ApplePay/WebPaymentCoordinatorProxy.h:
* Shared/ApplePay/WebPaymentCoordinatorProxy.cpp:
(WebKit::WebPaymentCoordinatorProxy::completePaymentMethodModeChange): Added.
(WebKit::WebPaymentCoordinatorProxy::presenterDidChangePaymentMethodMode): Added.
(WebKit::WebPaymentCoordinatorProxy::canBegin const):
(WebKit::WebPaymentCoordinatorProxy::canCancel const):
(WebKit::WebPaymentCoordinatorProxy::canCompletePayment const):
(WebKit::WebPaymentCoordinatorProxy::canAbort const):
* Shared/ApplePay/cocoa/WebPaymentCoordinatorProxyCocoa.mm:
(WebKit::WebPaymentCoordinatorProxy::platformCompletePaymentMethodModeChange): Added.
* Platform/cocoa/PaymentAuthorizationPresenter.h:
* Platform/cocoa/PaymentAuthorizationPresenter.mm:
(WebKit::toPKPaymentErrorCode):
(WebKit::PaymentAuthorizationPresenter::completePaymentMethodModeChange): Added.
* Platform/cocoa/PaymentAuthorizationViewController.mm:
* Platform/cocoa/WKPaymentAuthorizationDelegate.h:
* Platform/cocoa/WKPaymentAuthorizationDelegate.mm:
* Platform/ios/PaymentAuthorizationController.mm:
2021-02-19 Per Arne Vollan <pvollan@apple.com>
[macOS] Add IOKit filtering also when the GPU process is enabled
https://bugs.webkit.org/show_bug.cgi?id=222153
<rdar://problem/74505024>
Reviewed by Brent Fulgham.
IOKit filtering is enabled when the GPU process is not enabled. It should also be enabled when the GPU process is enabled.
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-19 Per Arne <pvollan@apple.com>
Add notifyd message filter
https://bugs.webkit.org/show_bug.cgi?id=222129
<rdar://problem/73987767>
Reviewed by Brent Fulgham.
Add notifyd message filter in the WebContent process' sandbox.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-19 Kate Cheney <katherine_cheney@apple.com>
Add better error handling to ResourceLoadStatisticsDatabaseStore::needsUpdatedPrivateClickMeasurementSchema()
https://bugs.webkit.org/show_bug.cgi?id=222169
<rdar://problem/74522283>
Reviewed by Darin Adler.
Add better error handling for the case where the query to find the
UnattributedPrivateClickMeasurement table fails. This is unlikely to
happen in practice, however, if it does, we should add logging and
a debug ASSERT.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::needsUpdatedPrivateClickMeasurementSchema):
2021-02-19 Devin Rousso <drousso@apple.com>
[Payment Request] add an `object data` to `PaymentDetailsBase` so that data specific to Apple Pay can be provided
https://bugs.webkit.org/show_bug.cgi?id=222002
<rdar://problem/72319946>
Reviewed by Tim Horton.
* Platform/cocoa/PaymentAuthorizationPresenter.h:
* Platform/cocoa/PaymentAuthorizationPresenter.mm:
(WebKit::toPKPaymentErrorCode):
(WebKit::toNSError):
(WebKit::toNSErrors):
(WebKit::toPKShippingMethods):
(WebKit::PaymentAuthorizationPresenter::completePaymentMethodSelection):
(WebKit::PaymentAuthorizationPresenter::completeShippingContactSelection):
(WebKit::PaymentAuthorizationPresenter::completeShippingMethodSelection):
* Platform/cocoa/WKPaymentAuthorizationDelegate.mm:
(toShippingMethod):
* Scripts/webkit/messages.py:
* Shared/ApplePay/WebPaymentCoordinatorProxy.messages.in:
* Shared/ApplePay/WebPaymentCoordinatorProxy.h:
* Shared/ApplePay/WebPaymentCoordinatorProxy.cpp:
(WebKit::WebPaymentCoordinatorProxy::completeShippingMethodSelection):
(WebKit::WebPaymentCoordinatorProxy::completeShippingContactSelection):
(WebKit::WebPaymentCoordinatorProxy::completePaymentMethodSelection):
(WebKit::WebPaymentCoordinatorProxy::presenterDidSelectShippingMethod):
* Shared/ApplePay/cocoa/WebPaymentCoordinatorProxyCocoa.h:
* Shared/ApplePay/cocoa/WebPaymentCoordinatorProxyCocoa.mm:
(WebKit::toPKShippingMethod):
(WebKit::WebPaymentCoordinatorProxy::platformPaymentRequest):
(WebKit::WebPaymentCoordinatorProxy::platformCompleteShippingMethodSelection):
(WebKit::WebPaymentCoordinatorProxy::platformCompleteShippingContactSelection):
(WebKit::WebPaymentCoordinatorProxy::platformCompletePaymentMethodSelection):
(WebKit::finishCreating): Deleted.
* Shared/WebCoreArgumentCoders.h:
* Shared/Cocoa/WebCoreArgumentCodersCocoa.mm:
(IPC::ArgumentCoder<WebCore::PaymentAuthorizationResult>::decode):
(IPC::ArgumentCoder<ApplePaySessionPaymentRequest>::decode):
(IPC::ArgumentCoder<Vector<RefPtr<ApplePayError>>>::encode): Added.
(IPC::ArgumentCoder<Vector<RefPtr<ApplePayError>>>::decode): Added.
(IPC::ArgumentCoder<WebCore::PaymentError>::encode): Deleted.
(IPC::ArgumentCoder<WebCore::PaymentError>::decode): Deleted.
(IPC::ArgumentCoder<WebCore::PaymentMethodUpdate>::encode): Deleted.
(IPC::ArgumentCoder<WebCore::PaymentMethodUpdate>::decode): Deleted.
(IPC::ArgumentCoder<ApplePaySessionPaymentRequest::ShippingMethod>::encode): Deleted.
(IPC::ArgumentCoder<ApplePaySessionPaymentRequest::ShippingMethod>::decode): Deleted.
(IPC::ArgumentCoder<ApplePaySessionPaymentRequest::TotalAndLineItems>::encode): Deleted.
(IPC::ArgumentCoder<ApplePaySessionPaymentRequest::TotalAndLineItems>::decode): Deleted.
(IPC::ArgumentCoder<WebCore::ShippingContactUpdate>::encode): Deleted.
(IPC::ArgumentCoder<WebCore::ShippingContactUpdate>::decode): Deleted.
(IPC::ArgumentCoder<WebCore::ShippingMethodUpdate>::encode): Deleted.
(IPC::ArgumentCoder<WebCore::ShippingMethodUpdate>::decode): Deleted.
* WebProcess/ApplePay/WebPaymentCoordinator.messages.in:
* WebProcess/ApplePay/WebPaymentCoordinator.h:
* WebProcess/ApplePay/WebPaymentCoordinator.cpp:
(WebKit::WebPaymentCoordinator::completeShippingMethodSelection):
(WebKit::WebPaymentCoordinator::completeShippingContactSelection):
(WebKit::WebPaymentCoordinator::completePaymentMethodSelection):
(WebKit::WebPaymentCoordinator::didSelectShippingMethod):
Use `ApplePay*Update` instead of duplicating its structure in another object (`ApplePaySessionPaymentRequest::*Update`).
Use `ApplePayShippingMethod` instead of duplicating its structure in another object (`ApplePaySessionPaymentRequest::ShippingMethod`).
Use `ApplePayError*` instead of duplicating its structure in another object (`ApplePaySessionPaymentRequest::PaymentError::*`).
2021-02-19 Chris Dumez <cdumez@apple.com>
Norton Safe Web extension is causing crashes / hangs under [WKRemoteObjectEncoder encodeObject:forKey:]
https://bugs.webkit.org/show_bug.cgi?id=222172
Reviewed by Alex Christensen.
The extension appears to be trying to send a JSValue that is a DOM Node. WebKit makes the following
call to convert it into a NSDictionary:
`[[JSValue valueWithJSValueRef:value inContext:[JSContext contextWithJSGlobalContextRef:JSContextGetGlobalContext(context)]] toObject]`
JSC very aggressively iterates over all of the properties of the DOM Node and recursively ends up
converting the whole DOM tree with all their properties. This leads to a lot of cycles to as JSC
maintains the JSObject <-> NSObject identity during the conversion (Each time the JSDocument is
serialized, the same NSDictionary* pointer is used to represent it).
The logic introduced in r270559 to detect cycles was flawed because it relied on a NSSet of
NSObject* and [NSSet containsObject:] to detect the cycles. The issue is that [NSSet containsObject:]
doesn't do a simple pointer comparison but instead calls [NSObject isEqual:] which is very
expensive for types like NSDictionary and leads to trouble when the dictionary contains a cycle.
To address this I replaced the NSSet with a WTF::HashSet<NSObject *> so that key lookup ends up
doing a simple pointer comparison.
Even after the previous fix, the extension would still cause massive hangs because it would take
a very long time to try and encode the whole DOM tree with all the properties of each Node (even
without cycles). To address this, we now abort encoding when detecting a cycle instead of encoding
an empty object to break the cycle.
After this change, Safari becomes usable with this extension again. However, there are still much
shorter hangs that occur due to the converting of the JSNode into a JSDictionary via
[JSValue toObject]. We should probably improve this in a follow-up.
Easy way to reproduce the crash / hang:
1. Install Norton Safe Web & Norton Password Manager extension (may require a subscription)
2. Make sure the extensions are activated and turned on by clicking on their icons next to the
URL bar
3. Go to https://bugs.webkit.org/attachment.cgi?id=420530&action=edit
4. Click on the combo box next to "Review" -> Hang / Crash
No new tests, covered by WebKit.RemoteObjectRegistry API test.
* Shared/API/Cocoa/WKRemoteObjectCoder.mm:
(-[WKRemoteObjectEncoder init]):
(encodeInvocationArguments):
(encodeObject):
2021-02-19 Kimmo Kinnunen <kkinnunen@apple.com>
RemoteRenderingBackend is accessed in non-thread-safe manner
https://bugs.webkit.org/show_bug.cgi?id=222056
Reviewed by Chris Dumez.
The class calls Connection::addWorkQueueMessageReceiver() during its
construction, passing 'this' as the receiver.
The virtual function pointer is not fully constructed during the call time.
Connection may start dispatching calls to the work queue. The work queue may
dispatch the task function through that pointer immediately after the call. If a
message comes during the constructor call, it may be dispatched during the
constructor still running, before the virtual function pointer being correctly
set up. Accessing virtual function pointer is not thread safe until it is fully
constructed.
No new tests, unit tests for WebKit not possible ATM.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::create):
(WebKit::RemoteRenderingBackend::RemoteRenderingBackend):
(WebKit::RemoteRenderingBackend::startListeningForIPC):
* GPUProcess/graphics/RemoteRenderingBackend.h:
2021-02-19 Kate Cheney <katherine_cheney@apple.com>
Create SQLite skeleton implementation for PCM fraud prevention
https://bugs.webkit.org/show_bug.cgi?id=222149
<rdar://problem/74502655>
Reviewed by Jiewen Tan.
This patch updates the PCM sqlite tables to hold 3 extra values:
token content, a signature, and a keyID. Since we are updating the
schema of an existing table, we need to migrate existing PCM data to
the new tables. These will have "null" values in the new fields.
Extensive local testing.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::unattributedPrivateClickMeasurementSchemaV1):
(WebKit::unattributedPrivateClickMeasurementSchemaV1Alternate):
(WebKit::outdatedUnattributedColumns):
(WebKit::outdatedAttributedColumns):
Strings that will help us determine if the schema needs to be updated
and hold the existing columns so we can populate the new tables with
this data.
(WebKit::ResourceLoadStatisticsDatabaseStore::needsUpdatedPrivateClickMeasurementSchema):
Compare the existing schema with the new one. We only need to do this
for one PCM table because they are being changed in the same patch.
(WebKit::ResourceLoadStatisticsDatabaseStore::missingReferenceToObservedDomains):
Name update now that there are various reasons to migrate data.
(WebKit::columnsToCopy):
I chose to migrate all tables even if only PCM tables need updating
because it was simpler with existing code, and it gives us more
flexibility if we need to change more schemas in the future. Since
the migration happens only once each time we update the schema, this
is probably low cost. The columnsToCopy() function gives us the option
to specify data we need from existing tables when we migrate.
(WebKit::ResourceLoadStatisticsDatabaseStore::migrateDataToNewTablesIfNecessary):
(WebKit::ResourceLoadStatisticsDatabaseStore::buildPrivateClickMeasurementFromDatabase):
(WebKit::ResourceLoadStatisticsDatabaseStore::insertPrivateClickMeasurement):
For now, these are empty values that we will implement when PCM fraud
prevention pieces are in place.
(WebKit::ResourceLoadStatisticsDatabaseStore::isMigrationNecessary): Deleted.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.h:
2021-02-19 Jiewen Tan <jiewen_tan@apple.com>
PCM: Request server public key to generate secret token
https://bugs.webkit.org/show_bug.cgi?id=222141
<rdar://problem/74462955>
Reviewed by John Wilander.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::setPrivateClickMeasurementTokenPublicKeyURLForTesting):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/NetworkSession.cpp:
(WebKit::NetworkSession::setPrivateClickMeasurementTokenPublicKeyURLForTesting):
* NetworkProcess/NetworkSession.h:
Paperwork to add a way to set test URL.
* NetworkProcess/PrivateClickMeasurementManager.cpp:
(WebKit::PrivateClickMeasurementManager::storeUnattributed):
(WebKit::generateNetworkResourceLoadParameters):
(WebKit::generateNetworkResourceLoadParametersForPost):
(WebKit::generateNetworkResourceLoadParametersForGet):
(WebKit::PrivateClickMeasurementManager::getTokenPublicKey):
(WebKit::PrivateClickMeasurementManager::getSignedUnlinkableToken):
(WebKit::PrivateClickMeasurementManager::fireConversionRequest):
(WebKit::PrivateClickMeasurementManager::setTokenPublicKeyURLForTesting):
* NetworkProcess/PrivateClickMeasurementManager.h:
Teaches the PCMM to send the token public key request.
* UIProcess/API/C/WKPage.cpp:
(WKPageSetPrivateClickMeasurementTokenPublicKeyURLForTesting):
* UIProcess/API/C/WKPagePrivate.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::simulateResourceLoadStatisticsSessionRestart):
(WebKit::WebPageProxy::setPrivateClickMeasurementTokenPublicKeyURLForTesting):
* UIProcess/WebPageProxy.h:
Paperwork to add a way to set test URL.
2021-02-19 Michael Catanzaro <mcatanzaro@gnome.org>
[WPE] User-Agent quirks not implemented
https://bugs.webkit.org/show_bug.cgi?id=222109
Reviewed by Carlos Garcia Campos.
Have WPE benefit from GTK's existing user agent code by moving the bits in WebPageGtk.cpp
and WebPageProxyGtk.cpp to WebPageGLib.cpp and a new WebPageProxyGLib.cpp. This also moves
over a few stub functions that are unimplemented in both WPE and GTK. Finally, it moves a
WebsiteDataStore function that had gotten very lost.
No new tests. Unfortunately, our existing user agent quirks tests do not exercise these
higher-level functions, and so were already running and passing for WPE.
* SourcesGTK.txt:
* SourcesWPE.txt:
* UIProcess/glib/WebPageProxyGLib.cpp: Copied from Source/WebKit/UIProcess/wpe/WebPageProxyWPE.cpp.
(WebKit::WebPageProxy::userAgentForURL):
(WebKit::WebPageProxy::standardUserAgent):
(WebKit::WebPageProxy::saveRecentSearches):
(WebKit::WebPageProxy::loadRecentSearches):
* UIProcess/glib/WebsiteDataStoreGLib.cpp:
(WebKit::WebsiteDataStore::platformRemoveRecentSearches):
* UIProcess/gtk/WebPageProxyGtk.cpp:
(WebKit::WebPageProxy::userAgentForURL): Deleted.
(WebKit::WebPageProxy::standardUserAgent): Deleted.
(WebKit::WebPageProxy::saveRecentSearches): Deleted.
(WebKit::WebPageProxy::loadRecentSearches): Deleted.
(WebKit::WebsiteDataStore::platformRemoveRecentSearches): Deleted.
* UIProcess/wpe/WebPageProxyWPE.cpp:
(WebKit::WebPageProxy::platformInitialize):
(WebKit::WebPageProxy::userAgentForURL): Deleted.
(WebKit::WebPageProxy::standardUserAgent): Deleted.
(WebKit::WebPageProxy::saveRecentSearches): Deleted.
(WebKit::WebPageProxy::loadRecentSearches): Deleted.
(WebKit::WebsiteDataStore::platformRemoveRecentSearches): Deleted.
* WebProcess/WebPage/glib/WebPageGLib.cpp:
(WebKit::WebPage::platformUserAgent const):
* WebProcess/WebPage/gtk/WebPageGtk.cpp:
(WebKit::WebPage::platformUserAgent const): Deleted.
* WebProcess/WebPage/wpe/WebPageWPE.cpp:
(WebKit::WebPage::platformUserAgent const): Deleted.
2021-02-19 Youenn Fablet <youenn@apple.com>
Update WKWebView getUserMedia delegate to latest proposal
https://bugs.webkit.org/show_bug.cgi?id=222103
Reviewed by Eric Carlson.
Covered by updated API tests.
* UIProcess/API/Cocoa/WKUIDelegatePrivate.h:
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::UIClient::decidePolicyForUserMediaPermissionRequest):
* UIProcess/UserMediaPermissionRequestProxy.cpp:
(WebKit::UserMediaPermissionRequestProxy::doDefaultAction):
2021-02-18 Chris Dumez <cdumez@apple.com>
Reduce explicit usage of [objC autorelease] in WebKit
https://bugs.webkit.org/show_bug.cgi?id=221932
<rdar://problem/74410900>
Reviewed by Darin Adler.
Follow-up to r272936 in order to address post-landing review comments.
* UIProcess/API/Cocoa/NSAttributedString.mm:
(+[NSAttributedString _loadFromHTMLWithOptions:contentLoader:completionHandler:]):
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView findString:withConfiguration:completionHandler:]):
(-[WKWebView _frames:]):
(-[WKWebView _getContentsAsAttributedStringWithCompletionHandler:]):
* UIProcess/API/Cocoa/_WKAutomationSession.mm:
(-[_WKAutomationSession init]):
* UIProcess/API/Cocoa/_WKInspector.mm:
(-[_WKInspector registerExtensionWithID:displayName:completionHandler:]):
* UIProcess/API/Cocoa/_WKRemoteWebInspectorViewController.mm:
(-[_WKRemoteWebInspectorViewController registerExtensionWithID:displayName:completionHandler:]):
* UIProcess/API/Cocoa/_WKUserContentExtensionStore.mm:
(-[_WKUserContentExtensionStore compileContentExtensionForIdentifier:encodedContentExtension:completionHandler:]):
(-[_WKUserContentExtensionStore lookupContentExtensionForIdentifier:completionHandler:]):
* UIProcess/Automation/ios/WebAutomationSessionIOS.mm:
(WebKit::WebAutomationSession::platformSimulateKeyboardInteraction):
* UIProcess/Cocoa/WKContactPicker.mm:
(-[WKContactPicker _contactsFromJSContacts:]):
2021-02-18 Peng Liu <peng.liu6@apple.com>
[ MacOS Wk2 Debug ] http/tests/misc/copy-resolves-urls.html is a flakey crash in RemoteMediaPlayerProxy
https://bugs.webkit.org/show_bug.cgi?id=222080
Reviewed by Eric Carlson.
No new tests, this will fix some crashing macOS tests.
* GPUProcess/media/RemoteMediaResource.cpp:
(WebKit::RemoteMediaResource::RemoteMediaResource):
(WebKit::RemoteMediaResource::~RemoteMediaResource):
* GPUProcess/media/RemoteMediaResource.h:
* GPUProcess/media/RemoteMediaResourceManager.h:
2021-02-18 Per Arne Vollan <pvollan@apple.com>
[iOS] Remove access to AppleJPEGDriverUserClient
https://bugs.webkit.org/show_bug.cgi?id=221458
<rdar://problem/73518481>
Reviewed by Brent Fulgham.
Remove access to AppleJPEGDriverUserClient in the WebContent process' sandbox on iOS.
* Resources/SandboxProfiles/ios/com.apple.WebKit.GPU.sb:
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebAuthn.sb:
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* UIProcess/WebPageProxy.cpp:
(WebKit::gpuIOKitClasses):
2021-02-18 Megan Gardner <megan_gardner@apple.com>
Correctly retain properties in _WKAppHighlight
https://bugs.webkit.org/show_bug.cgi?id=222151
Reviewed by Wenson Hsieh.
* UIProcess/API/Cocoa/_WKAppHighlight.h:
* UIProcess/API/Cocoa/_WKAppHighlight.mm:
(-[_WKAppHighlight initWithHighlight:text:image:]):
(-[_WKAppHighlight highlight]):
(-[_WKAppHighlight text]):
(-[_WKAppHighlight image]):
2021-02-18 Julian Gonzalez <julian_a_gonzalez@apple.com>
Crash in IPC::decode(Decoder& decoder, RetainPtr<SecKeychainItemRef>& result)
https://bugs.webkit.org/show_bug.cgi?id=222084
Reviewed by Ryosuke Niwa.
Add a check for a CFDataRef of 0 length so that we do not pass those
to SecKeychainItemCopyFromPersistentReference(), which crashes with them.
Test: ipc/argument-coders-crash.html
* Shared/cf/ArgumentCodersCF.cpp:
(IPC::decode):
2021-02-18 Ricky Mondello <rmondello@apple.com>
Unconditionally return information in _autofillContext SPI when a field is focused
https://bugs.webkit.org/show_bug.cgi?id=221828
rdar://74211237
Reviewed by Darin Adler.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _autofillContext]): Change the contract of _autofillContext to return information for any focused
text field. Add a "version" key. Explicitly indicate whether we're in a login context. This SPI remains stringly
typed for flexibility.
2021-02-18 Brent Fulgham <bfulgham@apple.com>
[macOS] Re-add fsctl support for marking files as potentially 'hot'
https://bugs.webkit.org/show_bug.cgi?id=222120
<rdar://problem/74387453>
Reviewed by Per Arne Vollan.
Rolled out r271959.
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-18 Devin Rousso <drousso@apple.com>
[Payment Request] add an `object data` to `PaymentItem` so that data specific to Apple Pay can be provided
https://bugs.webkit.org/show_bug.cgi?id=221970
<rdar://problem/69806999>
Reviewed by Alex Christensen.
* Shared/WebCoreArgumentCoders.h:
* Shared/Cocoa/WebCoreArgumentCodersCocoa.mm:
(IPC::ArgumentCoder<ApplePaySessionPaymentRequest>::decode):
(IPC::ArgumentCoder<ApplePaySessionPaymentRequest::TotalAndLineItems>::decode):
(IPC::ArgumentCoder<ApplePaySessionPaymentRequest::LineItem>::encode): Deleted.
(IPC::ArgumentCoder<ApplePaySessionPaymentRequest::LineItem>::decode): Deleted.
Just use `ApplePayLineItem` instead of duplicating its structure in another object (`ApplePaySessionPaymentRequest::LineItem`).
2021-02-18 Alex Christensen <achristensen@webkit.org>
Add null check in _getMainResourceDataWithCompletionHandler
https://bugs.webkit.org/show_bug.cgi?id=222139
<rdar://problem/74449438>
Reviewed by Tim Horton.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _getMainResourceDataWithCompletionHandler:]):
2021-02-18 BJ Burg <bburg@apple.com>
[Cocoa] Web Inspector: add support for using custom process pool and page group identifier
https://bugs.webkit.org/show_bug.cgi?id=222046
<rdar://74180162>
Reviewed by Timothy Hatcher.
These additions are necessary for clients to inject Web Extensions API into the tabs created
using _WKInspectorExtension.
Currently, Web Inspector pages use a private process pool that is separate from page content.
Since there is no way to set the pool's injected bundle path prior to opening Web Inspector,
we need to allow clients to explicitly configure and set a process pool to be used for
WebInspectorUI and any iframe tabs created using _WKInspectorExtension.
* UIProcess/API/APIInspectorConfiguration.cpp:
(API::InspectorConfiguration::processPool):
(API::InspectorConfiguration::setProcessPool):
* UIProcess/API/APIInspectorConfiguration.h:
* UIProcess/API/Cocoa/_WKInspectorConfiguration.h:
* UIProcess/API/Cocoa/_WKInspectorConfiguration.mm:
(-[_WKInspectorConfiguration setProcessPool:]):
(-[_WKInspectorConfiguration processPool]):
(-[_WKInspectorConfiguration applyToWebViewConfiguration:]):
(-[_WKInspectorConfiguration copyWithZone:]):
Add some plumbing for the new properties.
* UIProcess/Inspector/WebInspectorUtilities.h:
* UIProcess/Inspector/WebInspectorUtilities.cpp:
(WebKit::defaultInspectorPageGroupIdentifierForPage): Renamed.
(WebKit::defaultInspectorProcessPool): Renamed.
(WebKit::allInspectorProcessPools): Move to using a WeakHashSet since there could
be a number of pools specified by the client for different inspector instances.
(WebKit::prepareProcessPoolForInspector): Added.
(WebKit::isInspectorProcessPool): Turn this into a set lookup.
* UIProcess/Inspector/mac/WKInspectorViewController.mm:
(-[WKInspectorViewController webViewConfiguration]):
Add logic to use a custom process pool or a default process pool (existing behavior).
Add logic to use a custom page group identifier or a default value (existing behavior).
* UIProcess/Inspector/gtk/RemoteWebInspectorProxyGtk.cpp:
(WebKit::RemoteWebInspectorProxy::platformCreateFrontendPageAndWindow):
* UIProcess/Inspector/gtk/WebInspectorProxyGtk.cpp:
(WebKit::WebInspectorProxy::platformCreateFrontendPage):
* UIProcess/Inspector/win/RemoteWebInspectorProxyWin.cpp:
(WebKit::RemoteWebInspectorProxy::platformCreateFrontendPageAndWindow):
* UIProcess/Inspector/win/WebInspectorProxyWin.cpp:
(WebKit::WebInspectorProxy::platformCreateFrontendPage):
Adapt to renames.
2021-02-18 Wenson Hsieh <wenson_hsieh@apple.com>
Plumb an Optional<PageIdentifier> through WebPasteboardProxy
https://bugs.webkit.org/show_bug.cgi?id=222071
Reviewed by Tim Horton.
Add `const PasteboardContext*` arguments to the `PasteboardStrategy` methods on `WebPlatformStrategies`, convert
them to `PageIdentifier`s, and send them over to the UI process via `WebPasteboardProxy` methods.
While currently unused inside `WebPasteboardProxy`, the next patch will attempt to use these page identifiers to
retrieve a data owner type from the corresponding `WebPageProxy`, and then forward this data owner type to the
`PlatformPasteboard`.
* UIProcess/Cocoa/WebPasteboardProxyCocoa.mm:
(WebKit::WebPasteboardProxy::getPasteboardTypes):
(WebKit::WebPasteboardProxy::getPasteboardPathnamesForType):
(WebKit::WebPasteboardProxy::getPasteboardStringForType):
(WebKit::WebPasteboardProxy::getPasteboardStringsForType):
(WebKit::WebPasteboardProxy::getPasteboardBufferForType):
(WebKit::WebPasteboardProxy::getPasteboardChangeCount):
(WebKit::WebPasteboardProxy::getPasteboardColor):
(WebKit::WebPasteboardProxy::getPasteboardURL):
(WebKit::WebPasteboardProxy::addPasteboardTypes):
(WebKit::WebPasteboardProxy::setPasteboardTypes):
(WebKit::WebPasteboardProxy::setPasteboardURL):
(WebKit::WebPasteboardProxy::setPasteboardColor):
(WebKit::WebPasteboardProxy::setPasteboardStringForType):
(WebKit::WebPasteboardProxy::containsURLStringSuitableForLoading):
(WebKit::WebPasteboardProxy::urlStringSuitableForLoading):
(WebKit::WebPasteboardProxy::setPasteboardBufferForType):
(WebKit::WebPasteboardProxy::getNumberOfFiles):
(WebKit::WebPasteboardProxy::typesSafeForDOMToReadAndWrite):
(WebKit::WebPasteboardProxy::writeCustomData):
(WebKit::WebPasteboardProxy::allPasteboardItemInfo):
(WebKit::WebPasteboardProxy::informationForItemAtIndex):
(WebKit::WebPasteboardProxy::getPasteboardItemsCount):
(WebKit::WebPasteboardProxy::readStringFromPasteboard):
(WebKit::WebPasteboardProxy::readURLFromPasteboard):
(WebKit::WebPasteboardProxy::readBufferFromPasteboard):
(WebKit::WebPasteboardProxy::containsStringSafeForDOMToReadForType):
(WebKit::WebPasteboardProxy::writeURLToPasteboard):
(WebKit::WebPasteboardProxy::writeWebContentToPasteboard):
(WebKit::WebPasteboardProxy::writeImageToPasteboard):
(WebKit::WebPasteboardProxy::writeStringToPasteboard):
(WebKit::WebPasteboardProxy::updateSupportedTypeIdentifiers):
* UIProcess/WebPasteboardProxy.cpp:
(WebKit::WebPasteboardProxy::typesSafeForDOMToReadAndWrite):
(WebKit::WebPasteboardProxy::writeCustomData):
(WebKit::WebPasteboardProxy::allPasteboardItemInfo):
(WebKit::WebPasteboardProxy::informationForItemAtIndex):
(WebKit::WebPasteboardProxy::getPasteboardItemsCount):
(WebKit::WebPasteboardProxy::readURLFromPasteboard):
(WebKit::WebPasteboardProxy::readBufferFromPasteboard):
(WebKit::WebPasteboardProxy::readStringFromPasteboard):
(WebKit::WebPasteboardProxy::containsStringSafeForDOMToReadForType):
(WebKit::WebPasteboardProxy::containsURLStringSuitableForLoading):
(WebKit::WebPasteboardProxy::urlStringSuitableForLoading):
* UIProcess/WebPasteboardProxy.h:
* UIProcess/WebPasteboardProxy.messages.in:
* UIProcess/gtk/WebPasteboardProxyGtk.cpp:
(WebKit::WebPasteboardProxy::typesSafeForDOMToReadAndWrite):
(WebKit::WebPasteboardProxy::writeCustomData):
* UIProcess/libwpe/WebPasteboardProxyLibWPE.cpp:
(WebKit::WebPasteboardProxy::readStringFromPasteboard):
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::writeItemsToPasteboard):
* WebProcess/WebCoreSupport/WebPlatformStrategies.cpp:
(WebKit::pageIdentifier):
(WebKit::WebPlatformStrategies::getTypes):
(WebKit::WebPlatformStrategies::bufferForType):
(WebKit::WebPlatformStrategies::getPathnamesForType):
(WebKit::WebPlatformStrategies::stringForType):
(WebKit::WebPlatformStrategies::allStringsForType):
(WebKit::WebPlatformStrategies::changeCount):
(WebKit::WebPlatformStrategies::color):
(WebKit::WebPlatformStrategies::url):
(WebKit::WebPlatformStrategies::addTypes):
(WebKit::WebPlatformStrategies::setTypes):
(WebKit::WebPlatformStrategies::setBufferForType):
(WebKit::WebPlatformStrategies::setURL):
(WebKit::WebPlatformStrategies::setColor):
(WebKit::WebPlatformStrategies::setStringForType):
(WebKit::WebPlatformStrategies::getNumberOfFiles):
(WebKit::WebPlatformStrategies::containsURLStringSuitableForLoading):
(WebKit::WebPlatformStrategies::urlStringSuitableForLoading):
(WebKit::WebPlatformStrategies::writeToPasteboard):
(WebKit::WebPlatformStrategies::updateSupportedTypeIdentifiers):
(WebKit::WebPlatformStrategies::typesSafeForDOMToReadAndWrite):
(WebKit::WebPlatformStrategies::writeCustomData):
(WebKit::WebPlatformStrategies::containsStringSafeForDOMToReadForType):
(WebKit::WebPlatformStrategies::getPasteboardItemsCount):
(WebKit::WebPlatformStrategies::allPasteboardItemInfo):
(WebKit::WebPlatformStrategies::informationForItemAtIndex):
(WebKit::WebPlatformStrategies::readBufferFromPasteboard):
(WebKit::WebPlatformStrategies::readURLFromPasteboard):
(WebKit::WebPlatformStrategies::readStringFromPasteboard):
* WebProcess/WebCoreSupport/WebPlatformStrategies.h:
2021-02-18 Kate Cheney <katherine_cheney@apple.com>
Add support for non app-bound requests
https://bugs.webkit.org/show_bug.cgi?id=221909
<rdar://problem/73512988>
Reviewed by Brent Fulgham.
Add a new parameter to the ResourceRequest class specifying app
bound requests and populate subresource loads, redirects, and
preconnect tasks with this value.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::preconnectTo):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/NetworkResourceLoader.cpp:
(WebKit::NetworkResourceLoader::continueWillSendRedirectedRequest):
* Scripts/webkit/messages.py:
* Shared/NavigatingToAppBoundDomain.h:
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
* Shared/mac/WebCoreArgumentCodersMac.mm:
(IPC::ArgumentCoder<WebCore::ResourceRequest>::encodePlatformData):
(IPC::ArgumentCoder<WebCore::ResourceRequest>::decodePlatformData):
* UIProcess/API/Cocoa/WKWebViewPrivateForTesting.h:
* UIProcess/API/Cocoa/WKWebViewTesting.mm:
(-[WKWebView _lastNavigationWasAppBound:]):
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::setLastNavigationWasAppBound):
(WebKit::WebPageProxy::lastNavigationWasAppBound):
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::preconnectTo):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::loadRequest):
(WebKit::WebPageProxy::preconnectTo):
(WebKit::WebPageProxy::creationParameters):
* UIProcess/WebPageProxy.h:
* WebProcess/Network/WebLoaderStrategy.cpp:
(WebKit::WebLoaderStrategy::loadResource):
For the case of a main resource load without a navigation, like from about:blank,
the main resource load uses the original document loader which was created before
WebPage::loadRequest sets the value of m_lastNavigationWasAppBound. For this reason,
we should not check the document loader value for the main resource and
know that the resource request has the correct value.
(WebKit::WebLoaderStrategy::preconnectTo):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_lastNavigationWasAppBound):
(WebKit::WebPage::loadRequest):
(WebKit::WebPage::createDocumentLoader):
We should store lastNavigationWasAppBound in the document loader
instead of the page so we don't affect previous loads while initiating
the navigation to the next site.
(WebKit::WebPage::lastNavigationWasAppBound):
* WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::setLastNavigationWasAppBound):
* WebProcess/WebPage/WebPage.messages.in:
2021-02-18 Brent Fulgham <bfulgham@apple.com>
[macOS] Remove unused IPC shared memory reads in WebContent process
https://bugs.webkit.org/show_bug.cgi?id=222083
<rdar://problem/73509623>
Reviewed by Per Arne Vollan.
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-18 Fujii Hironori <Hironori.Fujii@sony.com>
Unreviewed, reverting r273063.
GTK port layout tests are timint out.
Reverted changeset:
"[CoordinatedGraphics] The whole content is unnecessarily
repainted by animations in non-AC mode pages"
https://bugs.webkit.org/show_bug.cgi?id=221391
https://commits.webkit.org/r273063
2021-02-18 Per Arne <pvollan@apple.com>
[macOS] Remove access to audio IOKit
https://bugs.webkit.org/show_bug.cgi?id=222062
<rdar://problem/70496648>
Reviewed by Brent Fulgham.
Remove access to audio IOKit in the WebContent process on macOS when GPUP is enabled.
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-18 John Wilander <wilander@apple.com>
PCM: Request signature for unlinkable token using attributionSourceNonce
https://bugs.webkit.org/show_bug.cgi?id=222076
<rdar://73581651>
Reviewed by Chris Dumez.
This patch makes PrivateClickMeasurementManager::storeUnattributed() look
for incoming source nonces and if found, call the new function
PrivateClickMeasurementManager::getSignedUnlinkableToken().
A new static function generateNetworkResourceLoadParameters() was added
to reuse code between PrivateClickMeasurementManager::getSignedUnlinkableToken()
and the existing PrivateClickMeasurementManager::fireConversionRequest().
To enable testing, there is a new function called
NetworkSession::setPrivateClickMeasurementTokenSignatureURLForTesting() and
function setPrivateClickMeasurementConversionURLForTesting() was renamed
setPrivateClickMeasurementAttributionReportURLForTesting() in NetworkProcess,
NetworkSession, WKPagePrivate, WKWebViewPrivateForTesting, and
PrivateClickMeasurementManager to align with the naming used in other places.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::setPrivateClickMeasurementTokenSignatureURLForTesting):
(WebKit::NetworkProcess::setPrivateClickMeasurementAttributionReportURLForTesting):
(WebKit::NetworkProcess::setPrivateClickMeasurementConversionURLForTesting): Deleted.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/NetworkSession.cpp:
(WebKit::NetworkSession::setPrivateClickMeasurementTokenSignatureURLForTesting):
(WebKit::NetworkSession::setPrivateClickMeasurementAttributionReportURLForTesting):
(WebKit::NetworkSession::setPrivateClickMeasurementConversionURLForTesting): Deleted.
* NetworkProcess/NetworkSession.h:
* NetworkProcess/PrivateClickMeasurementManager.cpp:
(WebKit::PrivateClickMeasurementManager::storeUnattributed):
(WebKit::generateNetworkResourceLoadParameters):
(WebKit::PrivateClickMeasurementManager::getSignedUnlinkableToken):
(WebKit::PrivateClickMeasurementManager::fireConversionRequest):
(WebKit::PrivateClickMeasurementManager::setTokenSignatureURLForTesting):
(WebKit::PrivateClickMeasurementManager::setAttributionReportURLForTesting):
(WebKit::PrivateClickMeasurementManager::setConversionURLForTesting): Deleted.
* NetworkProcess/PrivateClickMeasurementManager.h:
* UIProcess/API/C/WKPage.cpp:
(WKPageSetPrivateClickMeasurementTokenSignatureURLForTesting):
(WKPageSetPrivateClickMeasurementAttributionReportURLForTesting):
(WKPageSetPrivateClickMeasurementConversionURLForTesting): Deleted.
* UIProcess/API/C/WKPagePrivate.h:
* UIProcess/API/Cocoa/WKWebViewPrivateForTesting.h:
* UIProcess/API/Cocoa/WKWebViewTesting.mm:
(-[WKWebView _setPrivateClickMeasurementAttributionReportURLForTesting:completionHandler:]):
(-[WKWebView _setPrivateClickMeasurementConversionURLForTesting:completionHandler:]): Deleted.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::setPrivateClickMeasurementTokenSignatureURLForTesting):
(WebKit::WebPageProxy::setPrivateClickMeasurementAttributionReportURLForTesting):
(WebKit::WebPageProxy::setPrivateClickMeasurementConversionURLForTesting): Deleted.
* UIProcess/WebPageProxy.h:
2021-02-18 Michael Catanzaro <mcatanzaro@gnome.org>
[WPE][GTK] Avoid another child setup function in process launcher code
https://bugs.webkit.org/show_bug.cgi?id=222049
Reviewed by Carlos Garcia Campos.
Avoiding child setup functions is desirable because it could allow GSubprocess to use
posix_spawn() instead of fork() in the future. That's not possible to do if we have code
that needs to run between fork() and exec().
In this case, the child setup is used only to unset CLOEXEC. We could simply not set it in
the first place. This only fails if a secondary thread decides to launch a subprocess before
XDGDBusProxyLauncher::launch returns. That window already exists in many other places (e.g.
anywhere else setCloseOnExec is called, such as for IPC::Connection objects). Threads should
not do that.
This also fixes a bug where unsetting CLOEXEC would fail if we get unlucky and receive
EINTR. A loop is required here. WTF::setCloseOnExec handles that for us.
* UIProcess/Launcher/glib/BubblewrapLauncher.cpp:
(WebKit::XDGDBusProxyLauncher::launch):
(WebKit::XDGDBusProxyLauncher::childSetupFunc): Deleted.
2021-02-18 Youenn Fablet <youenn@apple.com>
Set ENABLE_VP9 to 1 on IOS
https://bugs.webkit.org/show_bug.cgi?id=222042
<rdar://problem/74433523>
Reviewed by Eric Carlson.
No change of behavior.
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetVP9DecoderEnabled):
(WKPreferencesGetVP9DecoderEnabled):
* UIProcess/API/C/WKPreferencesRef.h:
2021-02-18 Devin Rousso <drousso@apple.com>
Experiment with using the theme-color as the scroll area background if provided
https://bugs.webkit.org/show_bug.cgi?id=222078
<rdar://problem/74158818>
Reviewed by Tim Horton.
* UIProcess/API/ios/WKWebViewIOS.mm:
(baseScrollViewBackgroundColor):
* UIProcess/ViewSnapshotStore.cpp:
(WebKit::ViewSnapshotStore::recordSnapshot):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.cpp:
(WebPageProxy::scrollAreaBackgroundColor): Added.
2021-02-18 Per Arne <pvollan@apple.com>
[macOS] Observe system sleep events in the UI process
https://bugs.webkit.org/show_bug.cgi?id=221996
<rdar://problem/74406570>
Reviewed by Chris Dumez.
System sleep events should be observed in the UI process, which should notify the WebContent process.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::registerNotificationObservers):
(WebKit::WebProcessPool::unregisterNotificationObservers):
(WebKit::WebProcessPool::systemWillSleep):
(WebKit::WebProcessPool::systemDidWake):
* UIProcess/WebProcessPool.h:
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::systemWillSleep):
(WebKit::WebProcess::systemDidWake):
2021-02-18 Kimmo Kinnunen <kkinnunen@apple.com>
WebGL contexts do not work as source for Context2D drawImage calls in GPU process
https://bugs.webkit.org/show_bug.cgi?id=221811
Reviewed by Darin Adler.
Implements rudimentary support for reading the WebGL context and transfering
it to a ImageBuffer.
Fixes many tests that are removed from gpu-process/TestExpectations.
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::createGraphicsContextGL):
* GPUProcess/GPUConnectionToWebProcess.h:
(WebKit::GPUConnectionToWebProcess::RemoteRenderingBackendWrapper::get const):
* GPUProcess/GPUConnectionToWebProcess.messages.in:
* GPUProcess/graphics/RemoteGraphicsContextGL.cpp:
(WebKit::RemoteGraphicsContextGL::create):
(WebKit::RemoteGraphicsContextGL::RemoteGraphicsContextGL):
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::createGraphicsContextGL const):
* GPUProcess/graphics/RemoteGraphicsContextGLCocoa.cpp:
(WebKit::RemoteGraphicsContextGL::create):
(WebKit::RemoteGraphicsContextGLCocoa::RemoteGraphicsContextGLCocoa):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
(WebKit::RemoteRenderingBackendProxy::renderingBackendIdentifier const):
RemoteGraphicsContextGL takes a ref of the respective RemoteRenderingBackend.
RemoteRenderingBackend is the one that can handle ImageBuffers.
(WebKit::RemoteGraphicsContextGL::paintRenderingResultsToCanvas):
(WebKit::RemoteGraphicsContextGL::paintCompositedResultsToCanvas):
(WebKit::RemoteGraphicsContextGL::paintImageDataToImageBuffer):
* GPUProcess/graphics/RemoteGraphicsContextGL.h:
* GPUProcess/graphics/RemoteGraphicsContextGL.messages.in:
Implement the functions by:
1) Wait for the pending changes to the ImageBuffer
2) Send a synchronous message from Web process to GPU process WebGL sequence
3) Reading the buffer, post the task to RenderingBackend sequence
4) Paint the results to the ImageBuffer
5) Return the synchronous message result
(WebKit::RemoteGraphicsContextGLCocoa::prepareForDisplay):
RemoteGraphicsContextGL now retains the reference to display buffer.
Before, just sent the display buffer to Web process and forgot
about it.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::dispatch):
* GPUProcess/graphics/RemoteRenderingBackend.h:
Add function to dispatch work to rendering backend sequence.
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.cpp:
(WebKit::RemoteGraphicsContextGLProxy::create):
(WebKit::RemoteGraphicsContextGLProxy::RemoteGraphicsContextGLProxy):
(WebKit::RemoteGraphicsContextGLProxy::paintRenderingResultsToCanvas):
(WebKit::RemoteGraphicsContextGLProxy::paintCompositedResultsToCanvas):
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.h:
Send the messages.
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::connectToGPUProcess):
Use IPC::SendOption::DispatchMessageEvenWhenWaitingForSyncReply to
send CreateRenderingBackend. It must be dispatched in order compared to other
messages such as CreateRemoteGraphicsContextGL.
2021-02-17 Wenson Hsieh <wenson_hsieh@apple.com>
Plumb an Optional<PageIdentifier> through Pasteboard-related codepaths
https://bugs.webkit.org/show_bug.cgi?id=221987
Reviewed by Tim Horton.
Update more `DragData` and `Pasteboard` initialization sites by passing in a `PageIdentifier` or
`PasteboardContext`, respectively. See WebCore ChangeLog for more details.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<DragData>::encode):
(IPC::ArgumentCoder<DragData>::decode):
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::draggingEntered):
(WebKit::WebViewImpl::draggingUpdated):
(WebKit::WebViewImpl::draggingExited):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView dragDataForDropSession:dragDestinationAction:]):
* WebProcess/WebCoreSupport/gtk/WebEditorClientGtk.cpp:
(WebKit::WebEditorClient::updateGlobalSelection):
* WebProcess/WebCoreSupport/mac/WebDragClientMac.mm:
(WebKit::WebDragClient::declareAndWriteDragImage):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::performDragControllerAction):
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::performActionOnElement):
2021-02-18 Alex Christensen <achristensen@webkit.org>
Add SPI for getting size of PDF document during printing
https://bugs.webkit.org/show_bug.cgi?id=222064
<rdar://problem/74370891>
Reviewed by Geoff Garen and Tim Horton.
When printing a PDF, we want to know its page size. If the print is initiated by the web process (such as when window.print is called
or if the PDF contains JavaScript that tells the browser to print it) then we would like that information with the print request so we
don't have to ask the web process, which is hung waiting for a synchronous message reply. If the print is initiated by the UI process
(such as if the user prints from a menu or keyboard shortcut) then we need a way to ask what the size of a PDF is.
I reused the SPI in WKUIDelegatePrivate because it was only briefly adopted before we knew we needed this and then reverted.
I'm also adding an SPI way to get from a WKPageRef to a WKWebView so this can be adopted without redoing all of Safari.
Covered by API tests.
* UIProcess/API/APIUIClient.h:
(API::UIClient::printFrame):
* UIProcess/API/C/WKPage.cpp:
(WKPageSetPageUIClient):
* UIProcess/API/Cocoa/WKUIDelegatePrivate.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _getPDFFirstPageSizeInFrame:completionHandler:]):
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/Cocoa/UIDelegate.h:
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::setDelegate):
(WebKit::UIDelegate::UIClient::printFrame):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::printFrame):
(WebKit::WebPageProxy::getPDFFirstPageSize):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebProcess/Plugins/PDF/PDFPlugin.h:
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::pdfDocumentSizeForPrinting const):
* WebProcess/Plugins/Plugin.cpp:
(WebKit::Plugin::pdfDocumentSizeForPrinting const):
* WebProcess/Plugins/Plugin.h:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::print):
* WebProcess/WebPage/Cocoa/WebPageCocoa.mm:
(WebKit::WebPage::getPDFFirstPageSize):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::getPDFFirstPageSize):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2021-02-18 Kimmo Kinnunen <kkinnunen@apple.com>
ThreadMessageReceiverRefCounted subclasses are accessed in non-thread-safe way
https://bugs.webkit.org/show_bug.cgi?id=221891
Reviewed by Chris Dumez.
For each ThreadMessageReceiverRefCounted subclass, remove calls
to Connection::addThreadMessageReceiver() from the constructor.
The calls pass 'this' as the ThreadMessageReceiver*. If the
connection tries to dispatch a message during the constructor, this
results in some or all of below:
- non-thread-safe read of mutated virtual function pointer
- call to incorrect member function, mostly to MessageReceiver::didReceiveMessage
- an assert from above
- skipped message
For each ThreadMessageReceiverRefCounted subclass, remove calls
to Connection::removeThreadMessageReceiver() from the destructor.
The message receive destination is still active during destructor,
so Connection might dispatch new message tasks during that time.
This causes ref of the instance already being destroyed, which
then causes the tasks to use deleted instance and then later
deleting it again.
Does not try to fix any other potential thread-related issues in
the modified areas.
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRendererManager.cpp:
(WebKit::RemoteAudioMediaStreamTrackRendererManager::RemoteAudioMediaStreamTrackRendererManager):
(WebKit::RemoteAudioMediaStreamTrackRendererManager::startListeningForIPC):
(WebKit::RemoteAudioMediaStreamTrackRendererManager::close):
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRendererManager.h:
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayerManager.cpp:
(WebKit::RemoteSampleBufferDisplayLayerManager::RemoteSampleBufferDisplayLayerManager):
(WebKit::RemoteSampleBufferDisplayLayerManager::startListeningForIPC):
(WebKit::RemoteSampleBufferDisplayLayerManager::close):
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayerManager.h:
* NetworkProcess/webrtc/NetworkRTCProvider.cpp:
(WebKit::NetworkRTCProvider::NetworkRTCProvider):
(WebKit::NetworkRTCProvider::startListeningForIPC):
* NetworkProcess/webrtc/NetworkRTCProvider.h:
(WebKit::NetworkRTCProvider::create):
* WebProcess/GPU/webrtc/LibWebRTCCodecs.cpp:
(WebKit::LibWebRTCCodecs::LibWebRTCCodecs):
(WebKit::LibWebRTCCodecs::startListeningForIPC):
(WebKit::LibWebRTCCodecs::~LibWebRTCCodecs):
(WebKit::LibWebRTCCodecs::gpuProcessConnectionDidClose):
* WebProcess/GPU/webrtc/LibWebRTCCodecs.h:
(WebKit::LibWebRTCCodecs::create):
* WebProcess/Network/webrtc/LibWebRTCNetwork.cpp:
(WebKit::LibWebRTCNetwork::~LibWebRTCNetwork):
* WebProcess/Network/webrtc/LibWebRTCNetwork.h:
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::libWebRTCCodecs):
* WebProcess/cocoa/RemoteCaptureSampleManager.cpp:
(WebKit::RemoteCaptureSampleManager::~RemoteCaptureSampleManager):
(WebKit::RemoteCaptureSampleManager::stopListeningForIPC):
* WebProcess/cocoa/RemoteCaptureSampleManager.h:
* WebProcess/cocoa/UserMediaCaptureManager.cpp:
(WebKit::UserMediaCaptureManager::~UserMediaCaptureManager):
* WebProcess/glib/UserMediaCaptureManager.cpp:
(WebKit::UserMediaCaptureManager::~UserMediaCaptureManager):
2021-02-17 Megan Gardner <megan_gardner@apple.com>
Return correct target for App Highlight menus.
https://bugs.webkit.org/show_bug.cgi?id=222089
Reviewed by Tim Horton.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView targetForAction:withSender:]):
2021-02-17 Peng Liu <peng.liu6@apple.com>
[GPUP] Remove RemoteLegacyCDMSessionProxy and RemoteLegacyCDMSessionProxy from the message receiver map in the destructor of RemoteLegacyCDMFactoryProxy
https://bugs.webkit.org/show_bug.cgi?id=222079
Reviewed by Jer Noble.
Fix a GPU process crash when running http/tests/navigation/parsed-iframe-dynamic-form-back-entry.html.
* GPUProcess/media/RemoteLegacyCDMFactoryProxy.cpp:
(WebKit::RemoteLegacyCDMFactoryProxy::~RemoteLegacyCDMFactoryProxy):
2021-02-17 Fujii Hironori <Hironori.Fujii@sony.com>
[CoordinatedGraphics] The whole content is unnecessarily repainted by animations in non-AC mode pages
https://bugs.webkit.org/show_bug.cgi?id=221391
Reviewed by Carlos Garcia Campos.
When triggerRenderingUpdate was called back, it added the whole
view area into the dirty region to repaint in non-AC mode pages.
This caused a problem that the whole content was unnecessarily
repainted by animations in non-AC mode pages.
Call scheduleDisplay instead of setNeedsDisplay in
triggerRenderingUpdate.
However, if the dirty region is empty, display() exited early. If
triggerRenderingUpdate is called back,
DrawingAreaCoordinatedGraphics::display should be called to ensure
WebPage::updateRendering() called even if m_dirtyRegion is empty.
Added a new flag m_scheduledWhileWaitingForDidUpdate.
* WebProcess/WebPage/CoordinatedGraphics/DrawingAreaCoordinatedGraphics.cpp:
(WebKit::DrawingAreaCoordinatedGraphics::triggerRenderingUpdate):
Call scheduleDisplay() instead of setNeedsDisplay().
(WebKit::DrawingAreaCoordinatedGraphics::didUpdate):
(WebKit::DrawingAreaCoordinatedGraphics::scheduleDisplay): Don't
return early even if m_dirtyRegion is empty.
(WebKit::DrawingAreaCoordinatedGraphics::display()): Ditto.
(WebKit::DrawingAreaCoordinatedGraphics::display(UpdateInfo&)):
Return early if m_dirtyRegion is empty.
* WebProcess/WebPage/CoordinatedGraphics/DrawingAreaCoordinatedGraphics.h:
Added m_scheduledWhileWaitingForDidUpdate.
2021-02-17 Chris Dumez <cdumez@apple.com>
Reduce explicit usage of [objC autorelease] in WebKit even further
https://bugs.webkit.org/show_bug.cgi?id=222060
Reviewed by Darin Adler.
Reduce explicit usage of [objC autorelease] in WebKit further by using RetainPtr.
* UIProcess/Cocoa/SafeBrowsingWarningCocoa.mm:
(WebKit::safeBrowsingDetailsText):
* UIProcess/WebAuthentication/Mock/MockNfcService.mm:
(WebKit::MockNfcService::transceive):
* UIProcess/ios/DragDropInteractionState.mm:
(WebKit::DragDropInteractionState::updatePreviewsForActiveDragSources):
* UIProcess/ios/ViewGestureControllerIOS.mm:
(-[WKSwipeTransitionController gestureRecognizerForInteractiveTransition:WithTarget:action:]):
* UIProcess/ios/WKActionSheetAssistant.mm:
(appLinkForURL):
* UIProcess/ios/WKContentViewInteraction.mm:
(ensureLocalDragSessionContext):
(-[WKContentView insertTextPlaceholderWithSize:completionHandler:]):
(-[WKContentView continueContextMenuInteraction:]):
(-[WKContentView _dataForPreviewItemController:atPosition:type:]):
(-[WKContentView _presentedViewControllerForPreviewItemController:]):
(-[WKContentView _presentationSnapshotForPreviewItemController:]):
(+[WKTextRange textRangeWithState:isRange:isEditable:startRect:endRect:selectionRects:selectedTextLength:]):
(+[WKTextPosition textPositionWithRect:]):
* UIProcess/ios/WKTextPlaceholder.mm:
(-[WKTextPlaceholder rects]):
* UIProcess/ios/WKWebEvent.mm:
(-[WKWebEvent initWithEvent:]):
* UIProcess/ios/WebDataListSuggestionsDropdownIOS.mm:
(-[WKDataListSuggestionsPopover showSuggestionsDropdown:activationType:]):
(-[WKDataListSuggestionsViewController tableView:cellForRowAtIndexPath:]):
* UIProcess/ios/forms/WKDatePickerViewController.mm:
(-[WKDatePickerViewController _dateComponentForDay:month:year:era:]):
* UIProcess/ios/forms/WKFormSelectPicker.mm:
(-[WKOptionPickerCell initWithOptionItem:]):
(-[WKOptionGroupPickerCell initWithOptionItem:]):
(-[WKMultipleSelectPicker pickerView:viewForRow:forComponent:reusingView:]):
(-[WKSelectSinglePicker pickerView:attributedTitleForRow:forComponent:]):
(-[WKSelectPickerTableViewController tableView:viewForHeaderInSection:]):
(-[WKSelectPickerTableViewController tableView:cellForRowAtIndexPath:]):
* UIProcess/ios/forms/WKFormSelectPopover.mm:
(-[WKSelectTableViewController tableView:cellForRowAtIndexPath:]):
(-[WKSelectPopover initWithView:hasGroups:]):
* UIProcess/ios/forms/WKNumberPadView.mm:
(-[WKNumberPadView _buttonForPosition:]):
* UIProcess/ios/forms/WKNumberPadViewController.mm:
(-[WKNumberPadViewController _handleKeyPress:]):
* UIProcess/ios/forms/WKQuickboardListViewController.mm:
(-[WKQuickboardListViewController actionController]):
* UIProcess/ios/forms/WKTextInputListViewController.mm:
(-[WKTextInputListViewController enterText:]):
* UIProcess/ios/fullscreen/WKFullScreenWindowControllerIOS.mm:
(-[WKFullScreenWindowController enterFullScreen]):
* UIProcess/mac/WKImmediateActionController.mm:
(-[WKImmediateActionController _defaultAnimationController]):
* UIProcess/mac/WKTextInputWindowController.mm:
(-[WKTextInputPanel _interpretKeyEvent:usingLegacyCocoaTextInput:string:]):
* UIProcess/mac/WKViewLayoutStrategy.mm:
(+[WKViewLayoutStrategy layoutStrategyWithPage:view:viewImpl:mode:]):
* UIProcess/mac/WebColorPickerMac.mm:
(-[WKColorPopoverMac setAndShowPicker:withColor:suggestions:]):
* UIProcess/mac/WebDataListSuggestionsDropdownMac.mm:
(-[WKDataListSuggestionsController tableView:rowViewForRow:]):
(-[WKDataListSuggestionsController tableView:viewForTableColumn:row:]):
* WebProcess/InjectedBundle/API/Cocoa/WKWebProcessPlugInFrame.mm:
(-[WKWebProcessPlugInFrame _certificateChain]):
* WebProcess/InjectedBundle/API/Cocoa/WKWebProcessPlugInNodeHandle.mm:
(-[WKWebProcessPlugInNodeHandle renderedImageWithOptions:width:]):
* WebProcess/InjectedBundle/API/mac/WKDOMDocument.mm:
(-[WKDOMDocument parserYieldToken]):
* WebProcess/InjectedBundle/API/mac/WKDOMInternals.mm:
(WebKit::toWKDOMType):
* WebProcess/InjectedBundle/API/mac/WKDOMRange.mm:
(-[WKDOMRange rangeByExpandingToWordBoundaryByCharacters:inDirection:]):
* WebProcess/InjectedBundle/API/mac/WKWebProcessPlugInBrowserContextController.mm:
(-[WKWebProcessPlugInBrowserContextController handle]):
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::platformInitialize):
2021-02-17 Mark Lam <mark.lam@apple.com>
Build fix for Debug builds using --force-opt=O3.
Not reviewed.
#include a needed Inlines.h.
* Platform/IPC/JSIPCBinding.cpp:
2021-02-17 Kate Cheney <katherine_cheney@apple.com>
Add additions to LocalConnection.mm
https://bugs.webkit.org/show_bug.cgi?id=222029
<rdar://problem/69031303>
Reviewed by Jiewen Tan.
* UIProcess/WebAuthentication/Cocoa/AuthenticationServicesCoreSoftLink.h:
* UIProcess/WebAuthentication/Cocoa/AuthenticationServicesCoreSoftLink.mm:
* UIProcess/WebAuthentication/Cocoa/LocalConnection.mm:
(WebKit::LocalConnection::createCredentialPrivateKey const):
2021-02-17 Fujii Hironori <Hironori.Fujii@sony.com>
Unreviewed, reverting r273019.
GTK port layout tests are timint out.
Reverted changeset:
"[CoordinatedGraphics] The whole content is unnecessarily
repainted by animations in non-AC mode pages"
https://bugs.webkit.org/show_bug.cgi?id=221391
https://commits.webkit.org/r273019
2021-02-17 Aditya Keerthi <akeerthi@apple.com>
[iOS][FCR] Add new datalist dropdown for text inputs
https://bugs.webkit.org/show_bug.cgi?id=222008
<rdar://problem/74410513>
Reviewed by Wenson Hsieh.
The new datalist dropdown appears as a popover on both iPhones and iPads.
Previously, iPhones displayed a UIPickerView when tapping on the datalist
button, while iPads displayed a popover. Furthermore, the datalist button
was only visible after focusing the field. Now, the button can be used at
any time.
If the dropdown is presented while the software keyboard is visible, the
keyboard will be dismissed to create space for the dropdown. On devices
with a hardware keyboard, the dropdown is updated as characters are typed.
Tests: fast/forms/ios/form-control-refresh/datalist/datalist-show-hide.html
fast/forms/ios/form-control-refresh/datalist/datalist-textinput-suggestions-order.html
* Platform/spi/ios/UIKitSPI.h:
Added SPI declarations to configure the presentation of the presented
dropdown and to check whether the keyboard is visible.
* Shared/FocusedElementInformation.cpp:
(WebKit::FocusedElementInformation::encode const):
(WebKit::FocusedElementInformation::decode):
* Shared/FocusedElementInformation.h:
Added the isFocusingWithDataListDropdown member to control the presentation
of the keyboard when presenting a datalist dropdown.
* UIProcess/API/ios/WKWebViewPrivateForTestingIOS.h:
* UIProcess/API/ios/WKWebViewTestingIOS.mm:
(-[WKWebView _selectDataListOption:]):
Added a testing hook to select an option from the presented dropdown.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView shouldShowAutomaticKeyboardUI]):
Do not show the keyboard when a datalist dropdown is visible, unless a
hardware keyboard is attached.
(-[WKContentView updateFocusedElementFocusedWithDataListDropdown:]):
Toggle keyboard visibility depending on the visibility of the dropdown.
(-[WKContentView _selectDataListOption:]):
(-[WKContentView _setDataListSuggestionsControl:]):
* UIProcess/ios/WebDataListSuggestionsDropdownIOS.h:
* UIProcess/ios/WebDataListSuggestionsDropdownIOS.mm:
(WebKit::WebDataListSuggestionsDropdownIOS::show):
(-[WKDataListSuggestionsControl initWithInformation:inView:]):
(-[WKDataListSuggestionsViewController reloadData]):
(-[WKDataListSuggestionsDropdown initWithInformation:inView:]):
(-[WKDataListSuggestionsDropdown updateWithInformation:]):
(-[WKDataListSuggestionsDropdown showSuggestionsDropdown:activationType:]):
(-[WKDataListSuggestionsDropdown invalidate]):
(-[WKDataListSuggestionsDropdown didSelectOptionAtIndex:]):
(-[WKDataListSuggestionsDropdown dealloc]):
(-[WKDataListSuggestionsDropdown _displayWithActivationType:]):
(-[WKDataListSuggestionsDropdown _showSuggestions]):
(-[WKDataListSuggestionsDropdown _presentSuggestionsViewController]):
(-[WKDataListSuggestionsDropdown _updateTextSuggestions]):
(-[WKDataListSuggestionsDropdown _removeKeyboardDismissalObserver]):
(-[WKDataListSuggestionsDropdown presentationControllerDidDismiss:]):
(-[WKDataListSuggestionsDropdown adaptivePresentationStyleForPresentationController:traitCollection:]):
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::getFocusedElementInformation):
2021-02-17 Brent Fulgham <bfulgham@apple.com>
Follow-up fix after r273023.
https://bugs.webkit.org/show_bug.cgi?id=221641
<rdar://problem/74166568>
Reviewed by Chris Dumez.
Tested by TestWebKitAPI.
Resource loading expects the passed ResourceResponse to have a certificateInfo object.
This patch ensures that one exists.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::loadSimulatedRequest): Ensure certificate info is included.
2021-02-17 Megan Gardner <megan_gardner@apple.com>
Add SPI to allow apps to initiate highlight creation.
https://bugs.webkit.org/show_bug.cgi?id=222063
Reviewed by Tim Horton.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _addAppHighlight]):
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
2021-02-17 Per Arne Vollan <pvollan@apple.com>
[iOS] Remove access to the IOKIt class AGXDeviceUserClient in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=219733
<rdar://problem/70628137>
Reviewed by Brent Fulgham.
Remove access to the IOKIt class AGXDeviceUserClient in the WebContent process on iOS when the GPU process is enabled, since this is
not needed when allowing access to IOGPU. The IOGPU class is the superclass of AGXDeviceUserClient, and is the preferred IOKit
reference to use according to the Sandbox team.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2021-02-17 Yusuke Suzuki <ysuzuki@apple.com>
[JSC] Rename entitlements for JITCage
https://bugs.webkit.org/show_bug.cgi?id=222077
<rdar://problem/74451093>
Reviewed by Mark Lam.
* Scripts/process-entitlements.sh:
2021-02-17 Truitt Savell <tsavell@apple.com>
Unreviewed, reverting r272905.
caused
TestWebKitAPI.WebKit2.CrashGPUProcessWhileCapturingAndCalling
to timeout
Reverted changeset:
"ThreadMessageReceiverRefCounted subclasses are accessed in
non-thread-safe way"
https://bugs.webkit.org/show_bug.cgi?id=221891
https://commits.webkit.org/r272905
2021-02-17 Chris Dumez <cdumez@apple.com>
Unreviewed attempt to fix GTK build after r273023.
* UIProcess/API/APINavigation.cpp:
2021-02-17 David Kilzer <ddkilzer@apple.com>
Remove missing files from WebKit Xcode project
<https://webkit.org/b/222054>
Reviewed by Youenn Fablet.
Found using tidy-Xcode-project-file --missing (see Bug
188754).
* WebKit.xcodeproj/project.pbxproj:
(ConcreteShareableImageBuffer.h):
- Remove references to this missing file. This regressed in
r270470 when the file was removed.
(ImageDataReference.h):
- Remove references to this missing file. This regressed in
r272786 when the file was removed.
(NetworkCacheFileSystemCocoa.mm):
- Remove references to this missing file. This regressed in
r244921 when the file was removed.
(Recovered References):
- Remove "Recovered References" folder and missing
com.apple.WebKit.GPU.sb file that regressed in
r253098. The file was added back separately in r253131.
(WebRTCSocket.h):
(WebRTCSocket.cpp):
(WebRTCNetwork.messages.in):
- Remove references to these missing files. This apparently
regressed in r261163 when the files were renamed to have a
"Lib" prefix, but the old references were not removed.
2021-02-17 Brent Fulgham <bfulgham@apple.com>
Expand stubs to thread actual NSURLRequest/NSURLResponse for the new simulated-request API
https://bugs.webkit.org/show_bug.cgi?id=221641
<rdar://problem/74166568>
Reviewed by Chris Dumez.
Tested by new API Test LoadSimulatedRequestUpdatesBackForwardList.
This bug builds on the stubs created in Bug 221430 to use actual Request and
Response objects for simulated loads:
1. Thread the NSURLRequest and NSURLResponse objects to the loading layer when handling loads
for HTML strings and data.
2. Loads via the simulated request API should cause entries to be added to the back/forward list.
3. Create a new API Test to confirm the back/forward list works as expected.
* UIProcess/API/APINavigation.cpp:
(API::SubstituteData::SubstituteData): New convenience constructor accepting a ResourceResponse.
(API::Navigation::Navigation): Ditto.
* UIProcess/API/APINavigation.h:
(API::Navigation::create): Added.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView loadSimulatedRequest:withResponse:responseData:]): Switch from using the plain 'loadData'
method to the new 'loadSimulatedRequest' implementation.
* UIProcess/WebNavigationState.cpp:
(WebKit::WebNavigationState::createSimulatedLoadWithDataNavigation): Added.
* UIProcess/WebNavigationState.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::loadSimulatedRequest): New method to support the loadSimulatedRequest WKWebView API.
* UIProcess/WebPageProxy.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::loadSimulatedRequestAndResponse): Added message handler to perform a simulated load using
ResourceRequest and ResourceResponse objects.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in: Added LoadSimulatedRequestAndResponse IPC message.
2021-02-17 Fujii Hironori <Hironori.Fujii@sony.com>
[CoordinatedGraphics] The whole content is unnecessarily repainted by animations in non-AC mode pages
https://bugs.webkit.org/show_bug.cgi?id=221391
Reviewed by Carlos Garcia Campos.
When triggerRenderingUpdate was called back, it added the whole
view area into the dirty region to repaint in non-AC mode pages.
This caused a problem that the whole content was unnecessarily
repainted by animations in non-AC mode pages
If triggerRenderingUpdate is called back,
DrawingAreaCoordinatedGraphics::display should be called to ensure
WebPage::updateRendering() called even if m_dirtyRegion is empty.
* WebProcess/WebPage/CoordinatedGraphics/DrawingAreaCoordinatedGraphics.cpp:
(WebKit::DrawingAreaCoordinatedGraphics::triggerRenderingUpdate):
Call scheduleDisplay() instead of setNeedsDisplay().
(WebKit::DrawingAreaCoordinatedGraphics::didUpdate):
(WebKit::DrawingAreaCoordinatedGraphics::scheduleDisplay): Don't
return early even if m_dirtyRegion is empty.
(WebKit::DrawingAreaCoordinatedGraphics::display()): Ditto.
(WebKit::DrawingAreaCoordinatedGraphics::display(UpdateInfo&)):
Return early if m_dirtyRegion is empty.
2021-02-17 David Kilzer <ddkilzer@apple.com>
REGRESSION (r272625): Fix missing headers in Xcode project
<https://webkit.org/b/222052>
Reviewed by Eric Carlson.
* WebKit.xcodeproj/project.pbxproj:
- Manually reconstruct missing PBXBuildFile entries for Xcode
project.
2021-02-17 Megan Gardner <megan_gardner@apple.com>
Change App Highlight delegate to pass more information about the highlight for display.
https://bugs.webkit.org/show_bug.cgi?id=222014
Reviewed by Tim Horton.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _storeAppHighlight:]):
* UIProcess/API/Cocoa/_WKAppHighlightDelegate.h:
* WebKit.xcodeproj/project.pbxproj:
2021-02-17 Devin Rousso <drousso@apple.com>
[iOS] Add support for additional accessibility image extraction contextmenu actions
https://bugs.webkit.org/show_bug.cgi?id=222022
<rdar://problem/74157750>
Reviewed by Tim Horton.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView continueContextMenuInteraction:]):
(-[WKContentView contextMenuInteraction:willEndForConfiguration:animator:]):
2021-02-17 Devin Rousso <drousso@apple.com>
[iOS] don't "promote" media controls contextmenu items if they are the only item being shown
https://bugs.webkit.org/show_bug.cgi?id=221939
Reviewed by Eric Carlson.
* UIProcess/ios/WKActionSheetAssistant.mm:
(-[WKActionSheetAssistant showMediaControlsContextMenu:items:completionHandler:]):
2021-02-17 Per Arne <pvollan@apple.com>
[macOS] Deny mach-lookup to the fonts service
https://bugs.webkit.org/show_bug.cgi?id=221610
<rdar://problem/69168609>
Reviewed by Brent Fulgham.
This change will by default block access to the font daemon when the platform has support for a static font registry. When the client prefers
to allow user installed fonts, or Safari Reader configuration has been customized, access to the font daemon will still be allowed by issuing
an extension to the font daemon.
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::customizedReaderConfiguration):
(WebKit::WebPageProxy::creationParameters):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
(WebKit::WebPage::~WebPage):
* WebProcess/WebPage/WebPage.h:
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-17 Wenson Hsieh <wenson_hsieh@apple.com>
[macOS] Introduce a new context menu item to preview images
https://bugs.webkit.org/show_bug.cgi?id=221917
<rdar://problem/74363578>
Reviewed by Darin Adler and Tim Horton.
Add plumbing to support the `ContextMenuItemTagRevealImage` action, as well as some WebKitAdditions extension
points. See below for more details. This patch was mostly written by Devin Rousso, with some adjustments from
me.
* Configurations/WebKit.xcconfig:
Link against UniformTypeIdentifiers.
* Shared/API/c/WKContextMenuItemTypes.h:
* Shared/API/c/WKSharedAPICast.h:
(WebKit::toAPI):
(WebKit::toImpl):
Add `kWKContextMenuItemTagRevealImage`, and add helpers to convert from `WebCore::ContextMenuItemTagRevealImage`
to `kWKContextMenuItemTagRevealImage` and vice versa.
* Shared/ContextMenuContextData.h:
(WebKit::ContextMenuContextData::webHitTestResultData):
Add a non-const version of `webHitTestResultData()`, so that we can generate a `CGImage` using
`WebHitTestResultData`'s bitmap data.
* UIProcess/API/Cocoa/WKMenuItemIdentifiers.mm:
* UIProcess/API/Cocoa/WKMenuItemIdentifiersPrivate.h:
* UIProcess/API/Cocoa/_WKElementAction.mm:
(+[_WKElementAction _elementActionWithType:customTitle:assistant:]):
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::handleContextMenuRevealImage):
Add a helper method, `handleContextMenuRevealImage`, to handle the new context menu action; additionally,
introduce another helper method, `revealExtractedImageInPreviewPanel`, to handle the Mac-specific
`QLPreviewPanel` logic when invoking the new item. See WebKitAdditions changes for more detail.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::contextMenuItemSelected):
* UIProcess/WebPageProxy.h:
* UIProcess/mac/WebContextMenuProxyMac.mm:
(WebKit::menuItemIdentifier):
2021-02-17 Per Arne <pvollan@apple.com>
[macOS] Deny access to unused IOKit classes
https://bugs.webkit.org/show_bug.cgi?id=221969
<rdar://problem/74395400>
Reviewed by Brent Fulgham.
Deny access to the IOKit classes AppleIntelMEUserClient and AppleSNBFBUserClient when the GPU process is enabled.
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-17 Youenn Fablet <youenn@apple.com>
RemoteMediaPlayerManager does not need to support media stream code path
https://bugs.webkit.org/show_bug.cgi?id=221956
Reviewed by Eric Carlson.
RemoteMediaPlayerManager does not support isMediaStream code path so it is simpler to return IsNotSupported instead of MaybeSupported.
* WebProcess/GPU/media/RemoteMediaPlayerManager.cpp:
(WebKit::RemoteMediaPlayerManager::supportsTypeAndCodecs):
2021-02-17 Carlos Garcia Campos <cgarcia@igalia.com>
REGRESSION(r272636) [GTK] Crash and failures in API test /WebKit2Gtk/TestResources
https://bugs.webkit.org/show_bug.cgi?id=221672
Reviewed by Michael Catanzaro.
Since r272636 it's no longer possible to differentiate between async reply being cancelled or data just being
empty for WebPageProxy data completion handlers. Use Optional to indicate that the reply was canlled when its
data is nullopt, and ensure we always provide a valid API::Data object when data is not nullopt.
* Scripts/webkit/messages.py: Handle the special case of SharedBufferDataReference as Optional.
* UIProcess/API/glib/WebKitWebResource.cpp:
(resourceDataCallback): Bring back the cancellation error handling.
* UIProcess/WebPageProxy.cpp:
(WebKit::CompletionHandler<void):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::getSelectionAsWebArchiveData):
(WebKit::WebPage::getMainResourceDataOfFrame):
(WebKit::WebPage::getResourceDataFromFrame):
(WebKit::WebPage::getWebArchiveOfFrame):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2021-02-17 Chris Dumez <cdumez@apple.com>
Regression(r268097): Calling processDidTerminate delegate asynchronously is risky compatibility-wise
https://bugs.webkit.org/show_bug.cgi?id=222011
Reviewed by Carlos Garcia Campos.
Calling processDidTerminate delegate asynchronously like we did in r268097 is risky compatibility-wise.
This caused breakage in at least 2 client applications. While this can be dealt with on the client side,
it would be better to fix what r268097 was trying to fix without making the delegate call asynchronous.
The reason calling the delegate asynchronously is risky is because some view state may have time to get
reset by the time the client gets notified on the crash, potentially confusing the crash handling logic
in the client.
No new tests, covered by WKNavigation.ReloadRelatedViewsInProcessDidTerminate API test that is
still passing even though the delegate call is no longer asynchronous.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::commitProvisionalPage):
(WebKit::WebPageProxy::resetStateAfterProcessTermination):
(WebKit::WebPageProxy::dispatchProcessDidTerminate):
* UIProcess/WebPageProxy.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::processDidTerminateOrFailedToLaunch):
(WebKit::WebProcessProxy::requestTermination):
2021-02-17 Youenn Fablet <youenn@apple.com>
Build fix for non MEDIA_STREAM enabled ports after r272858
https://bugs.webkit.org/show_bug.cgi?id=222030
Unreviewed.
* WebProcess/GPU/webrtc/MediaRecorderProvider.h:
(WebKit::MediaRecorderProvider::MediaRecorderProvider):
Do not declare m_webPage in configs that do not use it.
2021-02-16 Peng Liu <peng.liu6@apple.com>
[GPUP] Video as image source triggers IOSurface readback (unnecessarily slow)
https://bugs.webkit.org/show_bug.cgi?id=221993
Reviewed by Tim Horton.
Use IOSurface::sinkIntoImage() instead of IOSurface::createImage() to avoid expensive GPU readback.
Manually tested.
* WebProcess/GPU/media/RemoteImageDecoderAVF.cpp:
(WebKit::RemoteImageDecoderAVF::createFrameImageAtIndex):
2021-02-16 Fujii Hironori <Hironori.Fujii@sony.com>
Unreviewed, reverting r272476.
GTK port is crashing.
Reverted changeset:
"[CoordinatedGraphics] The whole content is unnecessarily
repainted by animations in non-AC mode pages"
https://bugs.webkit.org/show_bug.cgi?id=221391
https://commits.webkit.org/r272476
2021-02-16 Devin Rousso <drousso@apple.com>
Unreviewed, build fix after r272925
* UIProcess/ios/WKContentViewInteraction.mm:
2021-02-16 Per Arne <pvollan@apple.com>
[iOS] Remove access to the Mobile asset service
https://bugs.webkit.org/show_bug.cgi?id=221959
<rdar://problem/73362858>
Reviewed by Brent Fulgham.
Remove access to the Mobile asset service in the WebContent process, since there are no longer reports of this being in use.
* GPUProcess/mac/com.apple.WebKit.GPUProcess.sb.in:
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::nonBrowserServices):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::grantAccessToAssetServices):
* WebAuthnProcess/mac/com.apple.WebKit.WebAuthnProcess.sb.in:
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::grantAccessToAssetServices):
2021-02-16 Per Arne <pvollan@apple.com>
[macOS] Remove access to IOAccelerationUserClient
https://bugs.webkit.org/show_bug.cgi?id=221967
<rdar://problem/70495933>
Reviewed by Brent Fulgham.
Remove access to IOAccelerationUserClient in the WebContent process on macOS when the GPU process is enabled.
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-16 Per Arne <pvollan@apple.com>
Update syntax in the sandbox bootstrap filter
https://bugs.webkit.org/show_bug.cgi?id=221913
<rdar://problem/73463985>
Reviewed by Brent Fulgham.
Update syntax in the sandbox bootstrap filter on macOS and iOS.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-16 Brady Eidson <beidson@apple.com>
REGRESSION (r271493 ) Crash in +[NSAttributedString(WKPrivate) _loadFromHTMLWithOptions:contentLoader:completionHandler:] on background thread
<rdar://problem/74038372> and https://bugs.webkit.org/show_bug.cgi?id=222005
Reviewed by Geoffrey Garen.
Test not possible as TestWebKitAPI seems to always initialize WebKit on the main thread.
* UIProcess/API/Cocoa/NSAttributedString.mm:
(+[NSAttributedString _loadFromHTMLWithOptions:contentLoader:completionHandler:]): Switch this RunLoop::main() back to dispatch_async to main.
2021-02-16 Chris Dumez <cdumez@apple.com>
Reduce explicit usage of [objC autorelease] in WebKit
https://bugs.webkit.org/show_bug.cgi?id=221932
Reviewed by Geoff Garen.
Reduce explicit usage of [objC autorelease] in WebKit by using RetainPtr<>.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(downgradeRequest):
(updateIgnoreStrictTransportSecuritySettingIfNecessary):
(WebKit::proxyDictionary):
(WebKit::NetworkSessionCocoa::NetworkSessionCocoa):
(WebKit::NetworkSessionCocoa::createWebSocketTask):
* Platform/cocoa/WKPaymentAuthorizationDelegate.mm:
(-[WKPaymentAuthorizationDelegate completePaymentMethodSelection:]):
(-[WKPaymentAuthorizationDelegate completeShippingContactSelection:]):
(-[WKPaymentAuthorizationDelegate completeShippingMethodSelection:]):
* Shared/API/Cocoa/WKRemoteObjectCoder.mm:
(encodeObject):
* Shared/API/Cocoa/_WKRemoteObjectInterface.mm:
(+[_WKRemoteObjectInterface remoteObjectInterfaceWithProtocol:]):
* Shared/Cocoa/WKObject.h:
(WebKit::wrapper):
* UIProcess/API/Cocoa/NSAttributedString.mm:
(+[_WKAttributedStringWebViewCache configuration]):
(+[NSAttributedString _loadFromHTMLWithOptions:contentLoader:completionHandler:]):
* UIProcess/API/Cocoa/WKBrowsingContextController.mm:
(makePolicyDecisionBlock):
(setUpPagePolicyClient):
(-[WKBrowsingContextController handle]):
* UIProcess/API/Cocoa/WKContentWorld.mm:
(-[WKContentWorld _userContentWorld]):
* UIProcess/API/Cocoa/WKFrameInfo.mm:
(-[WKFrameInfo securityOrigin]):
(-[WKFrameInfo webView]):
(-[WKFrameInfo _handle]):
(-[WKFrameInfo _parentFrameHandle]):
* UIProcess/API/Cocoa/WKUserScript.mm:
(-[WKUserScript _userContentWorld]):
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView configuration]):
(-[WKWebView findString:withConfiguration:completionHandler:]):
(-[WKWebView _frames:]):
(-[WKWebView _handle]):
(-[WKWebView _takePDFSnapshotWithConfiguration:completionHandler:]):
(-[WKWebView _sessionState]):
(-[WKWebView _sessionStateWithFilter:]):
(-[WKWebView _getContentsAsAttributedStringWithCompletionHandler:]):
* UIProcess/API/Cocoa/WKWebpagePreferences.mm:
(+[WKWebpagePreferences defaultPreferences]):
* UIProcess/API/Cocoa/WKWebsiteDataRecord.mm:
(-[WKWebsiteDataRecord _dataSize]):
* UIProcess/API/Cocoa/_WKActivatedElementInfo.mm:
(+[_WKActivatedElementInfo activatedElementInfoWithInteractionInformationAtPosition:userInfo:]):
(-[_WKActivatedElementInfo image]):
* UIProcess/API/Cocoa/_WKAttachment.mm:
(-[_WKAttachment info]):
* UIProcess/API/Cocoa/_WKAutomationSession.mm:
(-[_WKAutomationSession init]):
(-[_WKAutomationSession configuration]):
* UIProcess/API/Cocoa/_WKDownload.mm:
(+[_WKDownload downloadWithDownload:]):
(-[_WKDownload originatingWebView]):
* UIProcess/API/Cocoa/_WKElementAction.mm:
(+[_WKElementAction elementActionWithTitle:actionHandler:]):
(+[_WKElementAction _elementActionWithType:title:actionHandler:]):
(+[_WKElementAction _elementActionWithType:customTitle:assistant:]):
* UIProcess/API/Cocoa/_WKFrameTreeNode.mm:
(-[_WKFrameTreeNode securityOrigin]):
(-[_WKFrameTreeNode webView]):
(-[_WKFrameTreeNode _handle]):
(-[_WKFrameTreeNode _parentFrameHandle]):
* UIProcess/API/Cocoa/_WKInspector.mm:
(-[_WKInspector registerExtensionWithID:displayName:completionHandler:]):
* UIProcess/API/Cocoa/_WKProcessPoolConfiguration.mm:
(-[_WKProcessPoolConfiguration customClassesForParameterCoder]):
* UIProcess/API/Cocoa/_WKRemoteWebInspectorViewController.mm:
(-[_WKRemoteWebInspectorViewController registerExtensionWithID:displayName:completionHandler:]):
* UIProcess/API/Cocoa/_WKTextManipulationToken.mm:
(-[_WKTextManipulationToken _descriptionPreservingPrivacy:]):
* UIProcess/API/Cocoa/_WKUserContentExtensionStore.mm:
(+[_WKUserContentExtensionStore defaultStore]):
(+[_WKUserContentExtensionStore storeWithURL:]):
(-[_WKUserContentExtensionStore compileContentExtensionForIdentifier:encodedContentExtension:completionHandler:]):
(-[_WKUserContentExtensionStore lookupContentExtensionForIdentifier:completionHandler:]):
* UIProcess/API/Cocoa/_WKUserContentWorld.mm:
(+[_WKUserContentWorld worldWithName:]):
(+[_WKUserContentWorld normalWorld]):
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _uiEventAttribution]):
(-[WKWebView _retainActiveFocusedState]):
* UIProcess/Automation/ios/WebAutomationSessionIOS.mm:
(WebKit::WebAutomationSession::platformSimulateKeyboardInteraction):
* UIProcess/Cocoa/AutomationClient.mm:
(WebKit::AutomationClient::requestAutomationSession):
* UIProcess/Cocoa/SafeBrowsingWarningCocoa.mm:
(WebKit::addLinkAndReplace):
(WebKit::safeBrowsingDetailsText):
* UIProcess/Cocoa/SystemPreviewControllerCocoa.mm:
* UIProcess/Cocoa/VideoFullscreenManagerProxy.mm:
(-[WKLayerHostView makeBackingLayer]):
* UIProcess/Cocoa/WKContactPicker.mm:
(-[WKContactPicker _contactsFromJSContacts:]):
* UIProcess/Cocoa/WKSafeBrowsingWarning.mm:
(makeButton):
(makeLabel):
(-[WKSafeBrowsingWarning addContent]):
(-[WKSafeBrowsingWarning showDetailsClicked]):
(-[WKSafeBrowsingTextView initWithAttributedString:forWarning:]):
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::fullScreenWindow):
2021-02-16 BJ Burg <bburg@apple.com>
REGRESSION(r266890): [Cocoa] Fix InspectorDelegate / API::InspectorClient leak
https://bugs.webkit.org/show_bug.cgi?id=221988
<rdar://73594555>
Reviewed by Devin Rousso.
There was a logic error introduced into the new delegate situation after
moving _WKInspectorDelegate from WKWebView to _WKInspector. When setting
.delegate to nil, we shouldn't allocate dummy API::InspectorClient/
InspectorDelegate instances. As written, these instances form their own retain
cycle and cause a leak if Web Inspector has been opened or if
WKWebView._inspector is accessed (which lazily creates the delegates).
* UIProcess/API/Cocoa/_WKInspector.mm:
(-[_WKInspector setDelegate:]):
Adopt new constructor and pass in the ObjC delegate.
* UIProcess/Inspector/Cocoa/InspectorDelegate.h:
* UIProcess/Inspector/Cocoa/InspectorDelegate.mm:
(WebKit::InspectorDelegate::InspectorDelegate):
(WebKit::InspectorDelegate::createInspectorClient): Deleted.
(WebKit::InspectorDelegate::setDelegate): Deleted.
Clean up this class so that we always receive the ObjC delegate
via the constructor. If a nil delegate is passed---for example, when
closing the WKWebView---then don't create an API::InspectorClient
and set the WebInspectorProxy's client to nullptr.
2021-02-16 Brent Fulgham <bfulgham@apple.com>
Remove unneeded sandbox access to some file paths
https://bugs.webkit.org/show_bug.cgi?id=221946
<rdar://problem/62865856>
Reviewed by Per Arne Vollan.
Block access to /etc/passwd now that none of WebContent's dependencies
expect access.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-16 Megan Gardner <megan_gardner@apple.com>
Change App Highlights API to operate in terms of a single serialized highlight at a time
https://bugs.webkit.org/show_bug.cgi?id=221802
Reviewed by Tim Horton.
Pipe through additional data needed for display, and update to allow for more information to be added more easily.
* Shared/CreateNewGroupForHighlight.h: Removed.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _storeAppHighlight:]):
(-[WKWebView _restoreAppHighlights:]):
(-[WKWebView _updateAppHighlightsStorage:]): Deleted.
* UIProcess/API/Cocoa/WKWebViewInternal.h:
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/Cocoa/PageClientImplCocoa.h:
* UIProcess/Cocoa/PageClientImplCocoa.mm:
(WebKit::PageClientImplCocoa::storeAppHighlight):
(WebKit::PageClientImplCocoa::updateAppHighlightsStorage): Deleted.
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::createAppHighlightInSelectedRange):
(WebKit::WebPageProxy::restoreAppHighlights):
* UIProcess/PageClient.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::contextMenuItemSelected):
(WebKit::WebPageProxy::storeAppHighlight):
(WebKit::WebPageProxy::updateAppHighlightsStorage): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView createHighlightInCurrentGroupWithRange:]):
(-[WKContentView createHighlightInNewGroupWithRange:]):
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::storeAppHighlight const):
(WebKit::WebChromeClient::updateAppHighlightsStorage const): Deleted.
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::createAppHighlightInSelectedRange):
(WebKit::WebPage::restoreAppHighlights):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2021-02-16 Kate Cheney <katherine_cheney@apple.com>
Clean up WKWebView media API
https://bugs.webkit.org/show_bug.cgi?id=221929
<rdar://problem/73460121>
Reviewed by Youenn Fablet.
A few fixes that came from API review:
1) Change closeAllMediaPresentations to be
closeAllMediaPresentationsWithCompletionHandler. Ditto for
pauseAllMediaPlayback and requestMediaPlaybackState. This leads to
better Swift imports.
2) Combine suspendAllMediaPlayback and resumeAllMediaPlayback into a
single setAllMediaPlaybackSuspended:completionHandler function which
takes a boolean. This reinforces the relationship between
suspend/resume and differentiates them from pause.
And a few other drive-by fixes, like moving WKMediaPlaybackState to
the WKWebView class because it is unlikely to be used outside of that
context, changing NSUInteger -> NSInteger, and changing the ordering
of the WKMediaPlaybackState enum to be more natural.
No new tests. No behavior change. Covered by existing API tests.
* UIProcess/API/Cocoa/WKMediaPlaybackState.h: Removed.
* UIProcess/API/Cocoa/WKWebView.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView closeAllMediaPresentationsWithCompletionHandler:]):
(-[WKWebView pauseAllMediaPlaybackWithCompletionHandler:]):
(-[WKWebView setAllMediaPlaybackSuspended:completionHandler:]):
(toWKMediaPlaybackState):
(-[WKWebView requestMediaPlaybackStateWithCompletionHandler:]):
(-[WKWebView _closeAllMediaPresentations]):
(-[WKWebView _stopAllMediaPlayback]):
(-[WKWebView _suspendAllMediaPlayback]):
(-[WKWebView _resumeAllMediaPlayback]):
(-[WKWebView closeAllMediaPresentations:]): Deleted.
(-[WKWebView pauseAllMediaPlayback:]): Deleted.
(-[WKWebView suspendAllMediaPlayback:]): Deleted.
(-[WKWebView resumeAllMediaPlayback:]): Deleted.
(-[WKWebView requestMediaPlaybackState:]): Deleted.
* WebKit.xcodeproj/project.pbxproj:
2021-02-16 Yusuke Suzuki <ysuzuki@apple.com>
[JSC] Enable JITCage on macOS
https://bugs.webkit.org/show_bug.cgi?id=221805
<rdar://problem/74153806>
Reviewed by Mark Lam.
We need to add this entitlement only when building it on macOS 120000 or higher version.
Otherwise, we cannot launch the process. And we attach this entitlement only when building processes
with Apple Internal SDKs.
* Scripts/process-entitlements.sh:
2021-02-16 Alex Christensen <achristensen@webkit.org>
Add API to disable HTTPS upgrade
https://bugs.webkit.org/show_bug.cgi?id=221591
<rdar://problem/56254945>
Reviewed by Geoff Garen.
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
* UIProcess/API/APIPageConfiguration.cpp:
(API::PageConfiguration::copy const):
* UIProcess/API/APIPageConfiguration.h:
(API::PageConfiguration::setHTTPSUpgradeEnabled):
(API::PageConfiguration::httpsUpgradeEnabled const):
* UIProcess/API/Cocoa/WKWebViewConfiguration.h:
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration upgradeCommonKnownRequestsToHTTPS]):
(-[WKWebViewConfiguration setUpgradeCommonKnownRequestsToHTTPS:]):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
2021-02-16 Kimmo Kinnunen <kkinnunen@apple.com>
Connection clients should be able to obtain the messages as data instead of embedded in function references
https://bugs.webkit.org/show_bug.cgi?id=221560
Reviewed by Chris Dumez.
Add IPC::MessageReceiveQueue, an interface which obtains messages from
the message receive thread and forwards them to caller as data.
This is important in order to be able to skip messages that are stale
or originating from a sender which is incorrect. Before, the client
could only get the messages as Function references, and those cannot
be skipped as it would be undefined what not executing a function means.
IPC::MessageReceiveQueue is a generalization and simplification of current
IPC::Connection::ThreadMessageReceiverRefCounted and
IPC::Connection::WorkQueueMessageReceiver. It has different ref-count
semantics, in the way that client controls the life-time with add and
remove.
IPC::MessageReceiveQueue fixes message reordering bug with
WorkQueueMessageReceiver and ThreadMessageReceiver where first few messages
that are routed to main thread message queue can come later than some
other message that gets handled from message receive thread.
* Platform/IPC/Connection.cpp:
(IPC::Connection::addMessageReceiveQueue):
Add plain raw MessageReceiveQueue, to be used in WebGL IPC implementation.
Implement ThreadMessageReceiver and WorkQueueMessageReceiver with the
above.
(IPC::Connection::removeMessageReceiveQueue):
(IPC::Connection::dispatchMessageReceiverMessage):
(IPC::Connection::processIncomingMessage):
(IPC::Connection::enqueueIncomingMessage):
(IPC::Connection::dispatchMessage):
* Platform/IPC/Connection.h:
* Platform/IPC/MessageReceiveQueue.h: Copied from Source/WTF/wtf/FunctionDispatcher.h.
* Platform/IPC/MessageReceiveQueues.h: Added.
* Platform/IPC/MessageReceiveQueueMap.cpp: Added.
(IPC::MessageReceiveQueueMap::addImpl):
(IPC::MessageReceiveQueueMap::remove):
(IPC::MessageReceiveQueueMap::get const):
* Platform/IPC/MessageReceiveQueueMap.h: Copied from Source/WTF/wtf/FunctionDispatcher.h.
(IPC::MessageReceiveQueueMap::isValidMessage):
(IPC::MessageReceiveQueueMap::add):
* Sources.txt:
* WebKit.xcodeproj/project.pbxproj:
2021-02-16 Kimmo Kinnunen <kkinnunen@apple.com>
RemoteAudioSourceProviderManager is accessed in non-thread-safe manner
https://bugs.webkit.org/show_bug.cgi?id=221894
Reviewed by Chris Dumez.
Remove call to Connection::removeWorkQueueMessageReceiver() from the destructor.
The message receive destination is still active during destructor,
so Connection might dispatch new message tasks during that time.
This causes ref of the instance already being destroyed, which
then causes the tasks to use deleted instance and then later
deleting it again.
Other subclasses call add...() correctly.
Other subclasses never call remove..(). Assert that the instances are
never destroyed, as to never leave a dangling reference in the connection.
* UIProcess/mac/SecItemShimProxy.cpp:
(WebKit::SecItemShimProxy::~SecItemShimProxy):
* UIProcess/mac/SecItemShimProxy.h:
* WebProcess/GPU/GPUProcessConnection.cpp:
(WebKit::GPUProcessConnection::~GPUProcessConnection):
* WebProcess/GPU/media/RemoteAudioSourceProviderManager.cpp:
(WebKit::RemoteAudioSourceProviderManager::~RemoteAudioSourceProviderManager):
(WebKit::RemoteAudioSourceProviderManager::stopListeningForIPC):
* WebProcess/GPU/media/RemoteAudioSourceProviderManager.h:
* WebProcess/Inspector/WebInspectorInterruptDispatcher.cpp:
(WebKit::WebInspectorInterruptDispatcher::~WebInspectorInterruptDispatcher):
* WebProcess/Plugins/PluginProcessConnectionManager.cpp:
(WebKit::PluginProcessConnectionManager::~PluginProcessConnectionManager):
* WebProcess/WebPage/EventDispatcher.cpp:
(WebKit::EventDispatcher::~EventDispatcher):
* WebProcess/WebPage/ViewUpdateDispatcher.cpp:
(WebKit::ViewUpdateDispatcher::~ViewUpdateDispatcher):
2021-02-16 Youenn Fablet <youenn@apple.com>
Set a default path for device ID hash salts
https://bugs.webkit.org/show_bug.cgi?id=221955
Reviewed by Eric Carlson.
The default path was the empty string, which leads to not persist device IDs.
Update this to provide a valid default path like for other persisting data types.
Covered by API test.
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::WebsiteDataStore::defaultDeviceIdHashSaltsStorageDirectory):
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
2021-02-16 Alex Christensen <achristensen@webkit.org>
Synthesize range responses if needed in WebCoreNSURLSession
https://bugs.webkit.org/show_bug.cgi?id=221072
Reviewed by Geoff Garen.
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::requestResource):
2021-02-16 Kimmo Kinnunen <kkinnunen@apple.com>
ThreadMessageReceiverRefCounted subclasses are accessed in non-thread-safe way
https://bugs.webkit.org/show_bug.cgi?id=221891
Reviewed by Chris Dumez.
For each ThreadMessageReceiverRefCounted subclass, remove calls
to Connection::addThreadMessageReceiver() from the constructor.
The calls pass 'this' as the ThreadMessageReceiver*. If the
connection tries to dispatch a message during the constructor, this
results in some or all of below:
- non-thread-safe read of mutated virtual function pointer
- call to incorrect member function, mostly to MessageReceiver::didReceiveMessage
- an assert from above
- skipped message
For each ThreadMessageReceiverRefCounted subclass, remove calls
to Connection::removeThreadMessageReceiver() from the destructor.
The message receive destination is still active during destructor,
so Connection might dispatch new message tasks during that time.
This causes ref of the instance already being destroyed, which
then causes the tasks to use deleted instance and then later
deleting it again.
Does not try to fix any other potential thread-related issues in
the modified areas.
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRendererManager.cpp:
(WebKit::RemoteAudioMediaStreamTrackRendererManager::RemoteAudioMediaStreamTrackRendererManager):
(WebKit::RemoteAudioMediaStreamTrackRendererManager::startListeningForIPC):
(WebKit::RemoteAudioMediaStreamTrackRendererManager::close):
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRendererManager.h:
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayerManager.cpp:
(WebKit::RemoteSampleBufferDisplayLayerManager::RemoteSampleBufferDisplayLayerManager):
(WebKit::RemoteSampleBufferDisplayLayerManager::startListeningForIPC):
(WebKit::RemoteSampleBufferDisplayLayerManager::close):
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayerManager.h:
* NetworkProcess/webrtc/NetworkRTCProvider.cpp:
(WebKit::NetworkRTCProvider::NetworkRTCProvider):
(WebKit::NetworkRTCProvider::startListeningForIPC):
* NetworkProcess/webrtc/NetworkRTCProvider.h:
(WebKit::NetworkRTCProvider::create):
* WebProcess/GPU/webrtc/LibWebRTCCodecs.cpp:
(WebKit::LibWebRTCCodecs::LibWebRTCCodecs):
(WebKit::LibWebRTCCodecs::startListeningForIPC):
(WebKit::LibWebRTCCodecs::~LibWebRTCCodecs):
(WebKit::LibWebRTCCodecs::gpuProcessConnectionDidClose):
* WebProcess/GPU/webrtc/LibWebRTCCodecs.h:
(WebKit::LibWebRTCCodecs::create):
* WebProcess/Network/webrtc/LibWebRTCNetwork.cpp:
(WebKit::LibWebRTCNetwork::~LibWebRTCNetwork):
* WebProcess/Network/webrtc/LibWebRTCNetwork.h:
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::libWebRTCCodecs):
* WebProcess/cocoa/RemoteCaptureSampleManager.cpp:
(WebKit::RemoteCaptureSampleManager::~RemoteCaptureSampleManager):
(WebKit::RemoteCaptureSampleManager::stopListeningForIPC):
* WebProcess/cocoa/RemoteCaptureSampleManager.h:
* WebProcess/cocoa/UserMediaCaptureManager.cpp:
(WebKit::UserMediaCaptureManager::~UserMediaCaptureManager):
* WebProcess/glib/UserMediaCaptureManager.cpp:
(WebKit::UserMediaCaptureManager::~UserMediaCaptureManager):
2021-02-16 Youenn Fablet <youenn@apple.com>
Enable GPU WebRTC codecs in GPUProcess by default on iOS
https://bugs.webkit.org/show_bug.cgi?id=221562
Reviewed by Eric Carlson.
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultWebRTCCodecsInGPUProcess):
2021-02-15 Per Arne <pvollan@apple.com>
[macOS] Close XPC connections before entering sandbox
https://bugs.webkit.org/show_bug.cgi?id=221905
<rdar://problem/70359582>
Reviewed by Brent Fulgham.
XPC connections that are now allowed in the sandbox should be closed before entering the sandbox.
* Shared/mac/AuxiliaryProcessMac.mm:
(WebKit::initializeSandboxParameters):
2021-02-15 Per Arne <pvollan@apple.com>
Remove 'com.apple.powerlog.plxpclogger.xpc' from WebKit sandboxes
https://bugs.webkit.org/show_bug.cgi?id=221840
<rdar://problem/73579040>
Reviewed by Brent Fulgham.
Remove this service from WebKit sandboxes, since it is not required for correctness.
* GPUProcess/mac/com.apple.WebKit.GPUProcess.sb.in:
* Resources/SandboxProfiles/ios/com.apple.WebKit.GPU.sb:
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebAuthn.sb:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::diagnosticServices):
* WebAuthnProcess/mac/com.apple.WebKit.WebAuthnProcess.sb.in:
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-15 Truitt Savell <tsavell@apple.com>
Unreviewed, reverting r272867.
Broke 50+ fast/layoutformattingcontext/ tests on Mac Debug WK2
Reverted changeset:
"Page scrolling by wheel events doesn't work in non-AC mode
pages if Hardware Acceleration Policy is changed dynamically
from "always" to "never""
https://bugs.webkit.org/show_bug.cgi?id=221067
https://commits.webkit.org/r272867
2021-02-15 Eric Carlson <eric.carlson@apple.com>
[macOS] Connect Touch Bar to MediaSession
https://bugs.webkit.org/show_bug.cgi?id=221921
<rdar://problem/74363311>
Reviewed by Jer Noble.
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::didReceiveRemoteControlCommand): PlatformMediaSession::RemoteCommandArgument
is a struct.
* GPUProcess/GPUConnectionToWebProcess.h:
* UIProcess/Cocoa/PlaybackSessionManagerProxy.h:
* UIProcess/Cocoa/PlaybackSessionManagerProxy.mm:
(WebKit::PlaybackSessionModelContext::sendRemoteCommand):
(WebKit::PlaybackSessionManagerProxy::sendRemoteCommand):
* WebProcess/GPU/GPUProcessConnection.cpp:
(WebKit::GPUProcessConnection::didReceiveRemoteCommand):
* WebProcess/GPU/GPUProcessConnection.h:
* WebProcess/GPU/GPUProcessConnection.messages.in:
* WebProcess/GPU/media/RemoteRemoteCommandListener.messages.in:
* WebProcess/cocoa/PlaybackSessionManager.h:
* WebProcess/cocoa/PlaybackSessionManager.messages.in:
* WebProcess/cocoa/PlaybackSessionManager.mm:
(WebKit::PlaybackSessionManager::sendRemoteCommand):
2021-02-15 Timothy Hatcher <timothy@apple.com>
Web Inspector: Add a way to wake up debuggables to the remote inspector protocol
https://bugs.webkit.org/show_bug.cgi?id=221871
rdar://70351644
Reviewed by Devin Rousso.
* UIProcess/API/Cocoa/_WKAutomationDelegate.h:
* UIProcess/Cocoa/AutomationClient.h:
* UIProcess/Cocoa/AutomationClient.mm:
(WebKit::AutomationClient::AutomationClient): Check if the delegate responds to new method.
(WebKit::AutomationClient::requestedDebuggablesToWakeUp): Added. Call new process pool delegate method.
2021-02-15 Said Abou-Hallawa <said@apple.com>
[GPU Process] Ensure that no image decoders run in the GPU Process
https://bugs.webkit.org/show_bug.cgi?id=221885
<rdar://problem/74341122>
Reviewed by Simon Fraser.
Make GPU Process force ImageIO to disable all image decoding.
* GPUProcess/GPUProcess.cpp:
(WebKit::GPUProcess::initializeGPUProcess):
2021-02-15 Peng Liu <peng.liu6@apple.com>
[GPUP] Ignore an IPC message if the message receiver has been destroyed
https://bugs.webkit.org/show_bug.cgi?id=221920
Reviewed by Eric Carlson.
No new tests, fixing test crashes due to assertion failures.
* WebProcess/GPU/GPUProcessConnection.cpp:
(WebKit::GPUProcessConnection::dispatchMessage):
2021-02-15 Wenson Hsieh <wenson_hsieh@apple.com>
[macOS] Add a ShareableBitmap with decoded image data to WebHitTestResultData
https://bugs.webkit.org/show_bug.cgi?id=221869
Reviewed by Tim Horton.
Add `WebHitTestResultData::imageBitmap`; we'll use this in future patches to vend decoded image data to system
frameworks for image extraction.
* Shared/WebHitTestResultData.cpp:
(WebKit::WebHitTestResultData::WebHitTestResultData):
(WebKit::WebHitTestResultData::encode const):
(WebKit::WebHitTestResultData::decode):
Add `imageBitmap` and decoding/encoding support using `ShareableBitmap::Handle`.
* Shared/WebHitTestResultData.h:
* Sources.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebCoreSupport/ShareableBitmapUtilities.cpp: Added.
(WebKit::createShareableBitmap):
* WebProcess/WebCoreSupport/ShareableBitmapUtilities.h: Added.
Pull logic to decode rendered image data into a bitmap into a separate helper function; use this helper function
in `WebHitTestResultData`, as well as a couple of existing call sites in `WebPage`.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::requestImageExtraction):
(WebKit::WebPage::shareableBitmap const): Deleted.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::imagePositionInformation):
2021-02-15 Michael Catanzaro <mcatanzaro@gnome.org>
[WPE][GTK] SleepDisabler does not inhibit sleep with bubblewrap sandbox enabled: need to run xdg-dbus-proxy under bwrap or xdg-desktop-portal does not read our app ID
https://bugs.webkit.org/show_bug.cgi?id=219010
Reviewed by Carlos Garcia Campos.
We need to run xdg-dbus-proxy under bubblewrap so that we can mount /.flatpak-info in its
mount namespace. Why? Because xdg-desktop-portal checks the pid of its D-Bus peer using
Unix credentials, then looks in that process's mount namespace to read its app ID from
/.flatpak-info. If /.flatpak-info doesn't exist, then any portals that depend on the app ID
will break, including the inhibit portal that we need to prevent the system from suspending
during video playback. This normally works because flatpak runs xdg-dbus-proxy under
bubblewrap and mounts /.flatpak-info, but it breaks for WebKit because we don't. Start doing
so.
This method of getting the app ID might seem like a strange contortion, but it's really
required beacuse the sandboxed process cannot be trusted to provide its own app ID, and will
not have write access to its /.flatpak-info so we can rely on that.
* UIProcess/Launcher/ProcessLauncher.h:
* UIProcess/Launcher/glib/BubblewrapLauncher.cpp:
(WebKit::createFlatpakInfo):
(WebKit::XDGDBusProxyLauncher::launch):
(WebKit::bubblewrapSpawn):
2021-02-15 Michael Catanzaro <mcatanzaro@gnome.org>
Fix -Wunused-variable warning
https://bugs.webkit.org/show_bug.cgi?id=221653
<rdar://problem/74280146>
Reviewed by Alex Christensen.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::resetState):
2021-02-15 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r272863.
https://bugs.webkit.org/show_bug.cgi?id=221918
broke process launching
Reverted changeset:
"[WPE][GTK] SleepDisabler does not inhibit sleep with
bubblewrap sandbox enabled: need to run xdg-dbus-proxy under
bwrap or xdg-desktop-portal does not read our app ID"
https://bugs.webkit.org/show_bug.cgi?id=219010
https://trac.webkit.org/changeset/272863
2021-02-15 Alex Christensen <achristensen@webkit.org>
Unexpected ASSERT when touch events are dispatched on the main thread
https://bugs.webkit.org/show_bug.cgi?id=221832
Reviewed by Dean Jackson.
The TouchEvent message is special. It's received on a non-main thread and completed on the main thread, which asserts.
Rather than lose our asserts making sure our message CompletionHandlers are called on the correct thread,
mark this one with a new attribute, MainThreadCallback, which indicates that it is ok for the callback to be called on the main thread
like it was before r272558 in a different form.
* Platform/IPC/HandleMessage.h:
(IPC::handleMessageAsync):
* Scripts/webkit/messages.py:
* Scripts/webkit/model.py:
* Scripts/webkit/tests/TestWithImageDataMessages.h:
* Scripts/webkit/tests/TestWithLegacyReceiverMessages.h:
* Scripts/webkit/tests/TestWithSemaphoreMessages.h:
* Scripts/webkit/tests/TestWithSuperclass.messages.in:
* Scripts/webkit/tests/TestWithSuperclassMessages.h:
* Scripts/webkit/tests/TestWithoutAttributesMessages.h:
* WebProcess/WebPage/EventDispatcher.messages.in:
2021-02-15 Fujii Hironori <Hironori.Fujii@sony.com>
Page scrolling by wheel events doesn't work in non-AC mode pages if Hardware Acceleration Policy is changed dynamically from "always" to "never"
https://bugs.webkit.org/show_bug.cgi?id=221067
Reviewed by Carlos Garcia Campos.
GTK port is still supporting non-AC mode in WebKit2, and it can
enable/disable accelerated compositing dynamically.
r271235 fixed the bug that wheel event scrolling didn't work in
non-AC mode pages by checking if EventDispatcher's
m_scrollingTrees has a scrolling tree for a WebKit::WebPage.
However, if accelerated compositing is disabled dynamically,
m_scrollingTrees keeps to have a scrolling tree for the WebPage.
m_scrollingTrees should remove the item if accelerated compositing
is disabled.
r271235 added a code to check m_scrollingTrees, but it didn't lock
m_scrollingTreesMutex. The code can be removed. Removed it by
moving to above code.
EventDispatcher::wheelEvent was using a local lambda expression.
However the return value wasn't used. It can be a do-while
statement.
* WebProcess/WebPage/EventDispatcher.cpp:
(WebKit::EventDispatcher::wheelEvent): Replaced the lambda
expression with a do-while statement. Unified code of the checking
scrolling tree existence.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
(WebKit::WebPage::enterAcceleratedCompositingMode):
(WebKit::WebPage::exitAcceleratedCompositingMode):
(WebKit::WebPage::close):
2021-02-15 Michael Catanzaro <mcatanzaro@gnome.org>
[WPE][GTK] SleepDisabler does not inhibit sleep with bubblewrap sandbox enabled: need to run xdg-dbus-proxy under bwrap or xdg-desktop-portal does not read our app ID
https://bugs.webkit.org/show_bug.cgi?id=219010
Reviewed by Carlos Garcia Campos.
We need to run xdg-dbus-proxy under bubblewrap so that we can mount /.flatpak-info in its
mount namespace. Why? Because xdg-desktop-portal checks the pid of its D-Bus peer using
Unix credentials, then looks in that process's mount namespace to read its app ID from
/.flatpak-info. If /.flatpak-info doesn't exist, then any portals that depend on the app ID
will break, including the inhibit portal that we need to prevent the system from suspending
during video playback. This normally works because flatpak runs xdg-dbus-proxy under
bubblewrap and mounts /.flatpak-info, but it breaks for WebKit because we don't. Start doing
so.
This method of getting the app ID might seem like a strange contortion, but it's really
required beacuse the sandboxed process cannot be trusted to provide its own app ID, and will
not have write access to its /.flatpak-info so we can rely on that.
* UIProcess/Launcher/ProcessLauncher.h:
* UIProcess/Launcher/glib/BubblewrapLauncher.cpp:
(WebKit::createFlatpakInfo):
(WebKit::XDGDBusProxyLauncher::launch):
(WebKit::bubblewrapSpawn):
2021-02-15 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r272831.
https://bugs.webkit.org/show_bug.cgi?id=221907
breaking internal build
Reverted changeset:
"[JSC] Enable JITCage on macOS"
https://bugs.webkit.org/show_bug.cgi?id=221805
https://trac.webkit.org/changeset/272831
2021-02-15 Youenn Fablet <youenn@apple.com>
Move WebRTCPlatformCodecsInGPUProcessEnabled to Page Settings
https://bugs.webkit.org/show_bug.cgi?id=221893
Reviewed by Eric Carlson.
Make LibWebRTCProvider and MediaRecorder provider keep a WebPage reference.
Use it to get WebRTCPlatformCodecsInGPUProcessEnabled setting when needed.
For LibWebRTCProvider, set callback only once before creating the first peer connection.
This makes sure we do not change callbacks while using them in another thread.
* WebProcess/GPU/webrtc/MediaRecorderProvider.cpp:
(WebKit::MediaRecorderProvider::createMediaRecorderPrivate):
* WebProcess/GPU/webrtc/MediaRecorderProvider.h:
(WebKit::MediaRecorderProvider::MediaRecorderProvider):
* WebProcess/Network/webrtc/LibWebRTCProvider.cpp:
(WebKit::LibWebRTCProvider::createPeerConnection):
* WebProcess/Network/webrtc/LibWebRTCProvider.h:
(WebKit::LibWebRTCProvider::LibWebRTCProvider):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
2021-02-15 Youenn Fablet <youenn@apple.com>
Share more code between RemoteRealtimeAudioSource and RemoteRealtimeVideoSource
https://bugs.webkit.org/show_bug.cgi?id=221898
Reviewed by Eric Carlson.
Introduce RemoteRealtimeMediaSourceProxy to share more code between the two classes.
This class implements most of the messaging with capture process.
No change of behavior.
* WebProcess/cocoa/RemoteRealtimeAudioSource.cpp:
(WebKit::RemoteRealtimeAudioSource::createRemoteMediaSource):
* WebProcess/cocoa/RemoteRealtimeAudioSource.h:
(WebKit::RemoteRealtimeAudioSource::identifier const): Deleted.
(WebKit::RemoteRealtimeAudioSource::connection): Deleted.
(WebKit::RemoteRealtimeAudioSource::applyConstraintsFailed): Deleted.
* WebProcess/cocoa/RemoteRealtimeMediaSourceProxy.h:
(WebKit::RemoteRealtimeMediaSourceProxy::isReady const):
* WebProcess/cocoa/RemoteRealtimeVideoSource.cpp:
(WebKit::RemoteRealtimeVideoSource::RemoteRealtimeVideoSource):
(WebKit::RemoteRealtimeVideoSource::createRemoteMediaSource):
(WebKit::RemoteRealtimeVideoSource::~RemoteRealtimeVideoSource):
(WebKit::RemoteRealtimeVideoSource::setShouldApplyRotation):
(WebKit::RemoteRealtimeVideoSource::hasEnded):
(WebKit::RemoteRealtimeVideoSource::generatePresets):
(WebKit::RemoteRealtimeVideoSource::setFrameRateWithPreset):
(WebKit::RemoteRealtimeVideoSource::gpuProcessConnectionDidClose):
(WebKit::RemoteRealtimeVideoSource::whenReady): Deleted.
(WebKit::RemoteRealtimeVideoSource::didFail): Deleted.
(WebKit::RemoteRealtimeVideoSource::setAsReady): Deleted.
(WebKit::RemoteRealtimeVideoSource::connection): Deleted.
(WebKit::RemoteRealtimeVideoSource::startProducingData): Deleted.
(WebKit::RemoteRealtimeVideoSource::stopProducingData): Deleted.
* WebProcess/cocoa/RemoteRealtimeVideoSource.h:
(WebKit::RemoteRealtimeVideoSource::identifier const): Deleted.
2021-02-15 Michael Catanzaro <mcatanzaro@gnome.org>
[WPE][GTK] BubblewrapLauncher should create flatpak-info keyfile only once
https://bugs.webkit.org/show_bug.cgi?id=221224
Reviewed by Adrian Perez de Castro.
BubblewrapLauncher should create its flatpak-info keyfile only once, because its contents
will never change. Makes more sense to cache this tiny string in memory than to recompute it
every time a subprocess is launched.
When working on this, I noticed a quirk in the documentation of g_key_file_to_data: although
the function has a GError parameter, it is legacy and the function is actually guaranteed to
succeed. So icky though it seems, I've removed the error checking for that, as recommended
by its documentation.
* UIProcess/Launcher/glib/BubblewrapLauncher.cpp:
(WebKit::createFlatpakInfo):
2021-02-13 Alexey Proskuryakov <ap@apple.com>
Unreviewed build fix.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView pointerInteraction:styleForRegion:]):
2021-02-12 Yusuke Suzuki <ysuzuki@apple.com>
[JSC] Enable JITCage on macOS
https://bugs.webkit.org/show_bug.cgi?id=221805
<rdar://problem/74153806>
Reviewed by Mark Lam.
We need to add this entitlement only when building it on macOS 120000 or higher version.
Otherwise, we cannot launch the process. And we attach this entitlement only when building processes
with Apple Internal SDKs.
* Scripts/process-entitlements.sh:
2021-02-12 Chris Dumez <cdumez@apple.com>
Reduce explicit usage of [objC release] in WebKit
https://bugs.webkit.org/show_bug.cgi?id=221780
<rdar://problem/74282389>
Reviewed by Darin Adler.
Apply review feedback from Darin Adler and Sam Weinig after r272789.
* UIProcess/Inspector/ios/WKInspectorHighlightView.h:
* UIProcess/ios/WKScrollView.mm:
* UIProcess/mac/WebColorPickerMac.mm:
(+[WKPopoverColorWell _colorPopoverCreateIfNecessary:]):
* WebProcess/WebPage/mac/WKAccessibilityWebPageObjectBase.mm:
(-[WKAccessibilityWebPageObjectBase setRemoteParent:]):
2021-02-12 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Provide SPI to query local credentials
https://bugs.webkit.org/show_bug.cgi?id=221743
<rdar://problem/65849832>
Reviewed by Brent Fulgham.
Covered by API tests.
* UIProcess/API/Cocoa/_WKWebAuthenticationPanel.h:
* UIProcess/API/Cocoa/_WKWebAuthenticationPanel.mm:
(getAllLocalAuthenticatorCredentialsImpl):
(+[_WKWebAuthenticationPanel getAllLocalAuthenticatorCredentials]):
(+[_WKWebAuthenticationPanel getAllLocalAuthenticatorCredentialsWithAccessGroup:]):
(+[_WKWebAuthenticationPanel deleteLocalAuthenticatorCredentialWithID:]):
* UIProcess/API/Cocoa/_WKWebAuthenticationPanelForTesting.h:
2021-02-12 Jer Noble <jer.noble@apple.com>
[Mac] Unable to play WebM/Opus generated from Chrome MediaRecorder
https://bugs.webkit.org/show_bug.cgi?id=221808
Reviewed by Eric Carlson.
In the absense of an explicit duration parsed from the InitializationSegment,
rely instead on the maximum presentation timestamp available in any MediaTrackReader
to generate an explicit duration.
* Shared/mac/MediaFormatReader/MediaFormatReader.cpp:
(WebKit::MediaFormatReader::finishParsing):
(WebKit::MediaFormatReader::copyProperty):
* Shared/mac/MediaFormatReader/MediaTrackReader.cpp:
(WebKit::MediaTrackReader::greatestPresentationTime const):
* Shared/mac/MediaFormatReader/MediaTrackReader.h:
2021-02-12 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed, reverting r272799.
Caused debug WK2 tests to exit early with an assertion failure
Reverted changeset:
"Unexpected ASSERT when touch events are dispatched on the
main thread"
https://bugs.webkit.org/show_bug.cgi?id=221832
https://commits.webkit.org/r272799
2021-02-12 Youenn Fablet <youenn@apple.com>
Enable video capture in GPUProcess by default on MacOS
https://bugs.webkit.org/show_bug.cgi?id=221750
Reviewed by Eric Carlson.
Manually tested.
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultCaptureVideoInGPUProcessEnabled):
2021-02-12 Jer Noble <jer.noble@apple.com>
[Mac] Update Opus setting after a preference change.
https://bugs.webkit.org/show_bug.cgi?id=221807
Reviewed by Eric Carlson.
Make the opusDecoderEnabled() setting dynamic by applying changes
to that setting after a preference update.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences):
2021-02-12 Youenn Fablet <youenn@apple.com>
Remote video capture samples should be processed in a background thread
https://bugs.webkit.org/show_bug.cgi?id=221201
<rdar://problem/73826816>
Reviewed by Eric Carlson.
Make RemoteRealtimeVideoSource register to RemoteCaptureSampleManager to get video samples from a background thread.
Make RemoteCaptureSampleManager able to handle incoming video samples in background thread like for audio.
Make UserMediaCaptureManagerProxy use RemoteCaptureSampleManager instead of UserMediaCaptureManager message.
Covered by existing tests.
* UIProcess/Cocoa/UserMediaCaptureManagerProxy.cpp:
* WebProcess/cocoa/RemoteCaptureSampleManager.cpp:
(WebKit::RemoteCaptureSampleManager::addSource):
(WebKit::RemoteCaptureSampleManager::removeSource):
(WebKit::RemoteCaptureSampleManager::audioStorageChanged):
(WebKit::RemoteCaptureSampleManager::audioSamplesAvailable):
(WebKit::RemoteCaptureSampleManager::videoSampleAvailable):
(WebKit::RemoteCaptureSampleManager::RemoteVideo::RemoteVideo):
(WebKit::RemoteCaptureSampleManager::RemoteVideo::videoSampleAvailable):
* WebProcess/cocoa/RemoteCaptureSampleManager.h:
* WebProcess/cocoa/RemoteCaptureSampleManager.messages.in:
* WebProcess/cocoa/RemoteRealtimeVideoSource.cpp:
(WebKit::RemoteRealtimeVideoSource::create):
(WebKit::RemoteRealtimeVideoSource::videoSampleAvailable):
(WebKit::RemoteRealtimeVideoSource::remoteVideoSampleAvailable): Deleted.
* WebProcess/cocoa/RemoteRealtimeVideoSource.h:
* WebProcess/cocoa/UserMediaCaptureManager.cpp:
(WebKit::UserMediaCaptureManager::remoteVideoSampleAvailable): Deleted.
* WebProcess/cocoa/UserMediaCaptureManager.h:
* WebProcess/cocoa/UserMediaCaptureManager.messages.in:
2021-02-12 Aditya Keerthi <akeerthi@apple.com>
[tvOS] Fix build failure in WKFormSelectPicker
https://bugs.webkit.org/show_bug.cgi?id=221839
<rdar://problem/74287348>
Reviewed by Wenson Hsieh.
UITableViewStyleInsetGrouped is not available on tvOS. Rather than
have two separate codepaths, we can just use UITableViewStyleGrouped
with sectionContentInsetFollowsLayoutMargins set to YES.
* Platform/spi/ios/UIKitSPI.h:
* UIProcess/ios/forms/WKFormSelectPicker.mm:
(-[WKSelectPickerTableViewController initWithView:]):
2021-02-12 Youenn Fablet <youenn@apple.com>
Test landed flaky: [iOS] TestWebKitAPI.WebKit2.CrashGPUProcessWhileCapturing
https://bugs.webkit.org/show_bug.cgi?id=221331
<rdar://problem/73935129>
Reviewed by Eric Carlson.
Now that RemoteRealtimeVideoSource is a RealtimeVideoCaptureSource, we can register RemoteRealtimeVideoSource
to its factory so that it will be muted if another RemoteRealtimeVideoSource is created.
This is now possible as clones of the track will get the same underlying RemoteRealtimeVideoSource.
We can thus only recreate the active source when GPU process crashes.
Covered by test no longer flaky.
* WebProcess/cocoa/RemoteRealtimeAudioSource.cpp:
(WebKit::RemoteRealtimeAudioSource::gpuProcessConnectionDidClose):
* WebProcess/cocoa/RemoteRealtimeVideoSource.cpp:
(WebKit::RemoteRealtimeVideoSource::gpuProcessConnectionDidClose):
* WebProcess/cocoa/UserMediaCaptureManager.cpp:
(WebKit::UserMediaCaptureManager::VideoFactory::setActiveSource): Deleted.
* WebProcess/cocoa/UserMediaCaptureManager.h:
2021-02-12 Dean Jackson <dino@apple.com>
Unexpected ASSERT when touch events are dispatched on the main thread
https://bugs.webkit.org/show_bug.cgi?id=221832
<rdar://problem/74283856>
Reviewed by Sam Weinig.
The change in r272558 causes an ASSERT when running on iOS debug
builds. The touch event CompletionHandlers are created on the EventDispatcher
thread, but serviced on the main thread. We now need to state that this
is ok as we create the CompletionHandler.
This is covered by tests, but the EWS bots don't run iOS tests in
debug mode, and the internal bots hadn't caught up yet!
* Platform/IPC/HandleMessage.h:
(IPC::handleMessageAsync): Pass a CompletionHandlerCallThread into the constructor.
2021-02-12 Brent Fulgham <bfulgham@apple.com>
[macOS] Limit mech-register to required services
https://bugs.webkit.org/show_bug.cgi?id=221768
<rdar://problem/70350150>
Reviewed by Per Arne Vollan.
Limit mach-register to the two cases identified by telemetry.
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-12 Chris Dumez <cdumez@apple.com>
Reduce explicit usage of [objC release] in WebKit
https://bugs.webkit.org/show_bug.cgi?id=221780
Reviewed by Alex Christensen.
Reduce explicit usage of [objC release] in WebKit by using smart pointers.
* Shared/Cocoa/WKObject.mm:
(-[WKObject dealloc]):
(initializeTargetIfNeeded):
(-[WKObject forwardingTargetForSelector:]):
(-[WKObject forwardInvocation:]):
* Shared/RemoteLayerTree/RemoteLayerTreePropertyApplier.mm:
(-[UIView _web_setSubviews:]):
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _snapshotRect:intoImageOfWidth:completionHandler:]):
* UIProcess/Cocoa/LegacyCustomProtocolManagerClient.mm:
(-[WKCustomProtocolLoader initWithLegacyCustomProtocolManagerProxy:customProtocolID:request:]):
(-[WKCustomProtocolLoader dealloc]):
(WebKit::LegacyCustomProtocolManagerClient::startLoading):
* UIProcess/Inspector/ios/WKInspectorHighlightView.h:
* UIProcess/Inspector/ios/WKInspectorHighlightView.mm:
(-[WKInspectorHighlightView initWithFrame:]):
(-[WKInspectorHighlightView dealloc]):
(-[WKInspectorHighlightView _removeAllLayers]):
(-[WKInspectorHighlightView _createLayers:]):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _selectPositionAtPoint:stayingWithinFocusedElement:completionHandler:]):
(-[WKContentView selectPositionAtBoundary:inDirection:fromPoint:completionHandler:]):
(-[WKContentView moveSelectionAtBoundary:inDirection:completionHandler:]):
(-[WKContentView updateSelectionWithExtentPoint:withBoundary:completionHandler:]):
* UIProcess/ios/WKScrollView.mm:
(-[WKScrollView _updateDelegate]):
(-[WKScrollView dealloc]):
* UIProcess/ios/forms/WKFormSelectPopover.mm:
(-[WKSelectPopover initWithView:hasGroups:]):
* UIProcess/mac/WKTextInputWindowController.mm:
(-[WKTextInputPanel dealloc]):
(-[WKTextInputPanel init]):
* UIProcess/mac/WebColorPickerMac.mm:
(+[WKPopoverColorWell _colorPopoverCreateIfNecessary:]):
* WebProcess/WebPage/mac/WKAccessibilityWebPageObjectBase.h:
* WebProcess/WebPage/mac/WKAccessibilityWebPageObjectBase.mm:
(-[WKAccessibilityWebPageObjectBase setRemoteParent:]):
* WebProcess/WebPage/mac/WKAccessibilityWebPageObjectMac.mm:
(-[WKAccessibilityWebPageObject dealloc]):
(-[WKAccessibilityWebPageObject accessibilityAttributeValue:]):
2021-02-12 Kimmo Kinnunen <kkinnunen@apple.com>
GPU process WebGL context toDataURL does not work for non-premultiplied contexts
https://bugs.webkit.org/show_bug.cgi?id=221748
Reviewed by Myles C. Maxfield.
Implement GraphicsContextGL::paintRenderingResultsToImageData(). This
is used for Image.toDataURL() call, and returns something for
WebGL canvases that have non-premultiplied alpha.
No new tests, tested by currently failing WebGL conformance and ad hoc tests like
premultiplyalpha-test.html.
This does not make any failing test pass, as the tests test also premultiplied
case, which needs other currently unimplemented code.
* GPUProcess/graphics/RemoteGraphicsContextGL.messages.in:
* GPUProcess/graphics/RemoteGraphicsContextGLFunctionsGenerated.h:
(paintRenderingResultsToImageData):
Generated GPU process side implementation.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::getImageData):
* GPUProcess/graphics/RemoteRenderingBackend.h:
* GPUProcess/graphics/RemoteRenderingBackend.messages.in:
Change ImageDataReference to RefPtr<ImageData>.
* Platform/IPC/ImageDataReference.h: Removed.
This implementation is redundant. The messages could accept just
RefPtr<WebCore::ImageData>, which is exactly "reference to
Image Data", i.e. ImageDataReference.
RefPtr<WebCore::ImageData> encoder and decoder already exist.
* Scripts/webkit/messages.py:
* Scripts/webkit/messages_unittest.py:
* Scripts/webkit/tests/Makefile:
* Scripts/webkit/tests/MessageArgumentDescriptions.cpp:
(IPC::jsValueForArguments):
(IPC::messageArgumentDescriptions):
* Scripts/webkit/tests/MessageNames.cpp:
(IPC::description):
(IPC::receiverName):
(IPC::isValidMessageName):
* Scripts/webkit/tests/MessageNames.h:
* Scripts/webkit/tests/TestWithImageData.messages.in: Added.
* Scripts/webkit/tests/TestWithImageDataMessageReceiver.cpp: Added.
(WebKit::TestWithImageData::didReceiveMessage):
(WebKit::TestWithImageData::didReceiveSyncMessage):
* Scripts/webkit/tests/TestWithImageDataMessages.h: Added.
(Messages::TestWithImageData::messageReceiverName):
(Messages::TestWithImageData::SendImageData::name):
(Messages::TestWithImageData::SendImageData::SendImageData):
(Messages::TestWithImageData::SendImageData::arguments const):
(Messages::TestWithImageData::ReceiveImageData::name):
(Messages::TestWithImageData::ReceiveImageData::arguments const):
* Scripts/webkit/tests/TestWithImageDataMessagesReplies.h: Added.
Test using RefPtr<WebCore::ImageData> in messages.in.
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.cpp:
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.h:
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxyFunctionsGenerated.cpp:
(WebKit::RemoteGraphicsContextGLProxy::paintRenderingResultsToImageData):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::getImageData):
Generated Web process side implementation.
2021-02-12 Alex Christensen <achristensen@webkit.org>
Stop using GenericCallback from WebPageProxy
https://bugs.webkit.org/show_bug.cgi?id=221653
Reviewed by Chris Dumez.
There are still two uses in the DrawingAreaProxy implementations, but those are quite tangled and deserve their own patch.
* UIProcess/API/APIAttachment.cpp:
(API::Attachment::updateAttributes):
* UIProcess/API/APIAttachment.h:
* UIProcess/API/C/WKPage.cpp:
(CompletionHandler<void):
(WKPageRenderTreeExternalRepresentation):
(WKPageGetSourceForFrame):
(WKPageGetContentsAsString):
(WKPageGetBytecodeProfile):
(WKPageGetSamplingProfilerOutput):
(WTF::Function<void): Deleted.
* UIProcess/API/Cocoa/APIAttachmentCocoa.mm:
(API::Attachment::updateFromSerializedRepresentation):
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _insertAttachmentWithFileWrapper:contentType:completion:]):
(-[WKWebView _getContentsAsStringWithCompletionHandler:]):
(-[WKWebView _getContentsOfAllFramesAsStringWithCompletionHandler:]):
* UIProcess/API/Cocoa/_WKAttachment.mm:
(-[_WKAttachment setFileWrapper:contentType:completion:]):
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::startSpeaking):
* UIProcess/GenericCallback.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::dispatchActivityStateChange):
(WebKit::WebPageProxy::getRenderTreeExternalRepresentation):
(WebKit::WebPageProxy::getSourceForFrame):
(WebKit::WebPageProxy::getContentsAsString):
(WebKit::WebPageProxy::getBytecodeProfile):
(WebKit::WebPageProxy::getSamplingProfilerOutput):
(WebKit::WebPageProxy::getSelectionOrContentsAsString):
(WebKit::WebPageProxy::didCommitLoadForFrame):
(WebKit::WebPageProxy::didFailLoadForFrame):
(WebKit::WebPageProxy::resetState):
(WebKit::WebPageProxy::installActivityStateChangeCompletionHandler):
(WebKit::WebPageProxy::insertAttachment):
(WebKit::WebPageProxy::updateAttachmentAttributes):
(WebKit::WebPageProxy::didInsertAttachmentWithIdentifier):
(WebKit::WebPageProxy::clearLoadDependentCallbacks): Deleted.
(WebKit::WebPageProxy::voidCallback): Deleted.
(WebKit::WebPageProxy::stringCallback): Deleted.
(WebKit::WebPageProxy::invalidateStringCallback): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _shareForWebView:]):
(-[WKContentView _defineForWebView:]):
(-[WKContentView accessibilityRetrieveSpeakSelectionContent]):
(-[WKContentView applyAutocorrection:toString:withCompletionHandler:]):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::applyAutocorrection):
* UIProcess/mac/WKTextFinderClient.mm:
(-[WKTextFinderClient getSelectedText:]):
* WebProcess/Notifications/NotificationPermissionRequestManager.h:
* WebProcess/WebCoreSupport/WebNotificationClient.h:
* WebProcess/WebPage/DrawingArea.h:
(WebKit::DrawingArea::activityStateDidChange):
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.mm:
(WebKit::RemoteLayerTreeDrawingArea::activityStateDidChange):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::reinitializeWebPage):
(WebKit::WebPage::setActivityState):
(WebKit::WebPage::getContentsAsString):
(WebKit::WebPage::getRenderTreeExternalRepresentation):
(WebKit::WebPage::getSelectionOrContentsAsString):
(WebKit::WebPage::getSourceForFrame):
(WebKit::WebPage::getBytecodeProfile):
(WebKit::WebPage::getSamplingProfilerOutput):
(WebKit::WebPage::insertAttachment):
(WebKit::WebPage::updateAttachmentAttributes):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::applyAutocorrection):
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.h:
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::handleActivityStateChangeCallbacks):
(WebKit::TiledCoreAnimationDrawingArea::activityStateDidChange):
2021-02-12 Youenn Fablet <youenn@apple.com>
Make RemoteRealtimeVideoSource a RealtimeVideoCaptureSource
https://bugs.webkit.org/show_bug.cgi?id=221747
Reviewed by Eric Carlson.
All capture video sources are now deriving from RealtimeVideoCaptureSource.
This allows to expose a RealtimeVideoSource that can handle changes of resolution and frame rate in software as needed.
This also allows to better align audio and video sources on iOS: there will only be one running source at any given time.
Covered by existing tests.
* Scripts/webkit/messages.py:
* UIProcess/Cocoa/UserMediaCaptureManagerProxy.cpp:
(WebKit::UserMediaCaptureManagerProxy::capabilities):
(WebKit::UserMediaCaptureManagerProxy::generatePresets):
* UIProcess/Cocoa/UserMediaCaptureManagerProxy.h:
* UIProcess/Cocoa/UserMediaCaptureManagerProxy.messages.in:
* WebProcess/cocoa/RemoteRealtimeAudioSource.cpp:
(WebKit::RemoteRealtimeAudioSource::createRemoteMediaSource):
(WebKit::RemoteRealtimeAudioSource::setAsReady):
* WebProcess/cocoa/RemoteRealtimeVideoSource.cpp:
(WebKit::RemoteRealtimeVideoSource::create):
(WebKit::RemoteRealtimeVideoSource::RemoteRealtimeVideoSource):
(WebKit::RemoteRealtimeVideoSource::createRemoteMediaSource):
(WebKit::RemoteRealtimeVideoSource::setAsReady):
(WebKit::RemoteRealtimeVideoSource::remoteVideoSampleAvailable):
(WebKit::RemoteRealtimeVideoSource::setFrameRateWithPreset):
(WebKit::RemoteRealtimeVideoSource::prefersPreset):
(WebKit::RemoteRealtimeVideoSource::gpuProcessConnectionDidClose):
* WebProcess/cocoa/RemoteRealtimeVideoSource.h:
* WebProcess/cocoa/UserMediaCaptureManager.cpp:
(WebKit::UserMediaCaptureManager::VideoFactory::createVideoCaptureSource):
(WebKit::UserMediaCaptureManager::DisplayFactory::createDisplayCaptureSource):
2021-02-12 Said Abou-Hallawa <said@apple.com>
Make Pattern hold a NativeImage instead of an Image
https://bugs.webkit.org/show_bug.cgi?id=221637
Reviewed by Simon Fraser.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<ImageHandle>::encode): Deleted.
(IPC::ArgumentCoder<ImageHandle>::decode): Deleted.
* Shared/WebCoreArgumentCoders.h:
Delete the encoding and decoding of Image since it is not used anymore
2021-02-12 Antti Koivisto <antti@apple.com>
Increase network layout resource load priority for prioritized loads
https://bugs.webkit.org/show_bug.cgi?id=221751
Reviewed by Geoffrey Garen.
Currently visibility based prioritization only affects resource load scheduler.
We should increase the priority passed to the network layer and so the server too.
* NetworkProcess/NetworkDataTask.h:
(WebKit::NetworkDataTask::setPriority):
* NetworkProcess/NetworkLoad.cpp:
(WebKit::NetworkLoad::reprioritizeRequest):
* NetworkProcess/NetworkLoad.h:
* NetworkProcess/NetworkLoadScheduler.cpp:
(WebKit::NetworkLoadScheduler::HostContext::shouldDelayLowPriority const):
(WebKit::NetworkLoadScheduler::HostContext::schedule):
(WebKit::NetworkLoadScheduler::HostContext::unschedule):
(WebKit::NetworkLoadScheduler::HostContext::prioritize):
* NetworkProcess/cocoa/NetworkDataTaskCocoa.h:
* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::NetworkDataTaskCocoa::setPriority):
2021-02-11 Manuel Rego Casasnovas <rego@igalia.com>
[WPE] event.ctrlKey and other are false in keydown event
https://bugs.webkit.org/show_bug.cgi?id=221760
Reviewed by Adrian Perez de Castro.
This is the same issue that was fixed for WebKitGTK in r272489.
Test: fast/events/keyboardevent-modifier.html
* Shared/libwpe/WebEventFactory.cpp:
(WebKit::modifiersForEventModifiers):
(WebKit::modifiersForKeyboardEvent):
(WebKit::WebEventFactory::createWebKeyboardEvent):
2021-02-11 Alex Christensen <achristensen@webkit.org>
Unreviewed, reverting r272603.
Broke test, has room for improvement
Reverted changeset:
"Synthesize range responses if needed in WebCoreNSURLSession"
https://bugs.webkit.org/show_bug.cgi?id=221072
https://trac.webkit.org/changeset/272603
2021-02-11 Chris Dumez <cdumez@apple.com>
Review usage of adoptNS()
https://bugs.webkit.org/show_bug.cgi?id=221779
Reviewed by Alex Christensen.
Review usage of adoptNS():
1. Make sure we call adoptNS() as soon as we allocate the object instead of doing it later on, as this is less error-prone
2. Fix cases where we called adoptNS() but really should have retained instead of adopting
* UIProcess/RemoteLayerTree/RemoteLayerTreeHost.mm:
(WebKit::RemoteLayerTreeHost::makeNode):
* UIProcess/RemoteLayerTree/ios/RemoteLayerTreeHostIOS.mm:
(WebKit::RemoteLayerTreeHost::makeNode):
2021-02-11 Jer Noble <jer.noble@apple.com>
[Mac] Add Experimental Opus Codec support
https://bugs.webkit.org/show_bug.cgi?id=221745
Reviewed by Eric Carlson.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2021-02-11 David Kilzer <ddkilzer@apple.com>
Remove unused WebKit::LocalStorageDatabaseTracker::deleteAllDatabases()
<https://webkit.org/b/221734>
<rdar://problem/74215959>
Reviewed by Alexey Proskuryakov.
* NetworkProcess/WebStorage/LocalStorageDatabaseTracker.cpp:
(WebKit::LocalStorageDatabaseTracker::deleteAllDatabases): Delete.
* NetworkProcess/WebStorage/LocalStorageDatabaseTracker.h:
(WebKit::LocalStorageDatabaseTracker::deleteAllDatabases): Delete.
- Remove this method since it's not used anywhere. This fixes a
missing symbol linker error for
WebCore::SQLiteFileSystem::deleteEmptyDatabaseDirectory() when
building WebKit.framework with dead code stripping turned off
since that method was not exported from WebCore.framework.
2021-02-11 Aditya Keerthi <akeerthi@apple.com>
[iOS][FCR] Add new picker for <select multiple> elements
https://bugs.webkit.org/show_bug.cgi?id=221636
<rdar://problem/72399337>
Reviewed by Wenson Hsieh.
Test: fast/forms/ios/form-control-refresh/select/select-multiple-picker.html
* UIProcess/ios/forms/WKFormSelectControl.mm:
(-[WKFormSelectControl initWithView:]):
* UIProcess/ios/forms/WKFormSelectPicker.h:
* UIProcess/ios/forms/WKFormSelectPicker.mm:
(-[WKSelectPickerTableViewController initWithView:]):
The new picker is a table view with style UITableStyleInsetGrouped.
(-[WKSelectPickerTableViewController viewWillAppear:]):
Enable/disable the next/previous accessory buttons accordingly.
(-[WKSelectPickerTableViewController numberOfRowsInGroup:]):
(-[WKSelectPickerTableViewController numberOfSectionsInTableView:]):
There will always be at least one section, to display the empty message.
Adding groups will add more sections to the table view.
(-[WKSelectPickerTableViewController tableView:numberOfRowsInSection:]):
(-[WKSelectPickerTableViewController tableView:heightForHeaderInSection:]):
The header for the first section is an inset matching table view inset.
(-[WKSelectPickerTableViewController tableView:heightForFooterInSection:]):
An empty first section should not have a footer if there are more sections
after it. This can occur when the first option belongs to a group.
(-[WKSelectPickerTableViewController tableView:titleForHeaderInSection:]):
(-[WKSelectPickerTableViewController tableView:viewForFooterInSection:]):
Return nil so that the footer height is obeyed.
(-[WKSelectPickerTableViewController tableView:viewForHeaderInSection:]):
The header view displays the group name along with a button to collapse
the section.
(-[WKSelectPickerTableViewController collapseSection:]):
Collapsing sections is implemented by removing the corresponding rows
from the data source. The button is animated by changing its transform.
(-[WKSelectPickerTableViewController findItemIndexAt:]):
(-[WKSelectPickerTableViewController optionItemAtIndexPath:]):
(-[WKSelectPickerTableViewController tableView:cellForRowAtIndexPath:]):
(-[WKSelectPickerTableViewController tableView:didSelectRowAtIndexPath:]):
(-[WKSelectPickerTableViewController next:]):
(-[WKSelectPickerTableViewController previous:]):
(-[WKSelectPickerTableViewController close:]):
(-[WKSelectMultiplePicker initWithView:]):
(-[WKSelectMultiplePicker configurePresentation]):
Display the picker as a half-sheet (that can be dragged up into a
fullscreen view) on iPhones, and as a popover on iPads.
(-[WKSelectMultiplePicker controlView]):
(-[WKSelectMultiplePicker controlBeginEditing]):
(-[WKSelectMultiplePicker controlEndEditing]):
(-[WKSelectMultiplePicker presentationControllerDidDismiss:]):
(-[WKSelectMultiplePicker _indexPathForRow:]):
(-[WKSelectMultiplePicker selectRow:inComponent:extendingSelection:]):
Added testing hook to support layout testing.
2021-02-11 Jer Noble <jer.noble@apple.com>
[Cocoa][GPUP] Move RemoteCommandListener into the GPU Process
https://bugs.webkit.org/show_bug.cgi?id=221732
Reviewed by Eric Carlson.
Add a new Remote/Proxy class pair to facilitate RemoteCommandListener being
created in the GPU Process.
* DerivedSources-input.xcfilelist:
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::createRemoteCommandListener):
(WebKit::GPUConnectionToWebProcess::releaseRemoteCommandListener):
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/GPUConnectionToWebProcess.messages.in:
* GPUProcess/media/RemoteRemoteCommandListenerProxy.cpp:
(WebKit::RemoteRemoteCommandListenerProxy::RemoteRemoteCommandListenerProxy):
(WebKit::RemoteRemoteCommandListenerProxy::didReceiveRemoteControlCommand):
(WebKit::RemoteRemoteCommandListenerProxy::updateSupportedCommands):
* GPUProcess/media/RemoteRemoteCommandListenerProxy.h:
* GPUProcess/media/RemoteRemoteCommandListenerProxy.messages.in:
* Scripts/webkit/messages.py:
* Sources.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/GPUProcessConnection.cpp:
(WebKit::GPUProcessConnection::didReceiveRemoteCommand):
* WebProcess/GPU/media/RemoteRemoteCommandListener.cpp: Added.
(WebKit::RemoteRemoteCommandListener::create):
(WebKit::RemoteRemoteCommandListener::RemoteRemoteCommandListener):
(WebKit::RemoteRemoteCommandListener::~RemoteRemoteCommandListener):
(WebKit::RemoteRemoteCommandListener::gpuProcessConnectionDidClose):
(WebKit::RemoteRemoteCommandListener::didReceiveRemoteControlCommand):
* WebProcess/GPU/media/RemoteRemoteCommandListener.h:
* WebProcess/GPU/media/RemoteRemoteCommandListener.messages.in:
* WebProcess/GPU/media/RemoteRemoteCommandListenerIdentifier.h:
* WebProcess/GPU/media/ios/RemoteMediaSessionHelper.cpp:
(WebKit::RemoteMediaSessionHelper::activeVideoRouteDidChange):
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::setUseGPUProcessForMedia):
2021-02-11 Darin Adler <darin@apple.com>
[Cocoa] IPC decoder is using decoded size to allocate memory for an array
https://bugs.webkit.org/show_bug.cgi?id=221773
Reviewed by Geoffrey Garen.
* Shared/Cocoa/ArgumentCodersCocoa.mm:
(IPC::decodeArrayInternal): As with other similar structures, such a Vector and
CFArray, don't use the size to preallocate space when decoding an NSArray. The
decoded size is potentially incorrect, which we will discover indirectly when
decoding the array elements; we can't safely use the size to make a choice about
allocating memory beforehand.
2021-02-11 Brent Fulgham <bfulgham@apple.com>
[macOS] Add telemetry for a likely-unused call to ipc-posix-shm-read-data
https://bugs.webkit.org/show_bug.cgi?id=221733
<rdar://problem/74214692>
Reviewed by Per Arne Vollan.
Add telemetry to confirm this is unsed.
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-11 Youenn Fablet <youenn@apple.com>
[MacOS] Enable Audio Capture in GPUProcess by default
https://bugs.webkit.org/show_bug.cgi?id=221400
Reviewed by Eric Carlson.
Covered by existing API tests, in particular TestWebKitAPI.WebKit.OnDeviceChangeCrash.
* WebKit/UIProcess/UserMediaPermissionRequestManagerProxy.cpp:
We need to synchronously fill the granted requests, otherwise there is a risk that
the granted requests will be cleared (on page close for instance) and then later filled
for the page that was gone.
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultCaptureAudioInGPUProcessEnabled):
2021-02-11 Sam Weinig <weinig@apple.com>
Rework RGB color types to be more declarative to reduce code duplication and make progress toward better automatic conversion
https://bugs.webkit.org/show_bug.cgi?id=221677
Reviewed by Darin Adler.
* UIProcess/API/wpe/WebKitColor.cpp:
(webkitColorToWebCoreColor):
Update to specify SRGBA<float>, which is now required due to lack of deduction guides through
type aliases.
2021-02-11 Simon Fraser <simon.fraser@apple.com>
Crash when UI-side compositing is enabled on macOS
https://bugs.webkit.org/show_bug.cgi?id=221740
Reviewed by Tim Horton.
When UI-side compositing is enabled on macOS, WebProcess::updatePageScreenProperties()
crashes because allOf(m_pageMap.values(), ...) return a null WebPage. We're inside
WebProcess::createWebPage() here, so haven't set the HashMap value yet.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::updatePageScreenProperties):
2021-02-11 Per Arne <pvollan@apple.com>
[macOS] Add syscall to allow list in WebContent sandbox
https://bugs.webkit.org/show_bug.cgi?id=221705
<rdar://problem/74162777>
Reviewed by Brent Fulgham.
Add required syscall to allow list in WebContent sandbox on macOS.
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-11 Sergio Villar Senin <svillar@igalia.com>
Non unified build fixes for mid February 2021
https://bugs.webkit.org/show_bug.cgi?id=221749
Reviewed by Youenn Fablet.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
* GPUProcess/media/RemoteCDMInstanceProxy.cpp:
2021-02-11 Youenn Fablet <youenn@apple.com>
Split RemoteRealtimeMediaSource in two audio-specific and video-specific classes
https://bugs.webkit.org/show_bug.cgi?id=221608
Reviewed by Eric Carlson.
Split RemoteRealtimeMediaSource in RemoteRealtimeAudioSource and RemoteRealtimeVideoSource.
This will allow RemoteRealtimeVideoSource to become a RealtimeVideoCaptureSource in a follow-up
to properly handle fan-out to multiple clones with various frame rates and resolutions.
This also allows removing differences of behaviors between audio and video classes.
No change of behavior.
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/cocoa/RemoteCaptureSampleManager.cpp:
(WebKit::RemoteCaptureSampleManager::addSource):
(WebKit::RemoteCaptureSampleManager::didUpdateSourceConnection):
(WebKit::RemoteCaptureSampleManager::RemoteAudio::RemoteAudio):
* WebProcess/cocoa/RemoteCaptureSampleManager.h:
* WebProcess/cocoa/RemoteRealtimeAudioSource.cpp: Added.
(WebKit::RemoteRealtimeAudioSource::create):
(WebKit::RemoteRealtimeAudioSource::RemoteRealtimeAudioSource):
(WebKit::RemoteRealtimeAudioSource::createRemoteMediaSource):
(WebKit::RemoteRealtimeAudioSource::~RemoteRealtimeAudioSource):
(WebKit::RemoteRealtimeAudioSource::whenReady):
(WebKit::RemoteRealtimeAudioSource::didFail):
(WebKit::RemoteRealtimeAudioSource::setAsReady):
(WebKit::RemoteRealtimeAudioSource::setCapabilities):
(WebKit::RemoteRealtimeAudioSource::setSettings):
(WebKit::RemoteRealtimeAudioSource::remoteAudioSamplesAvailable):
(WebKit::RemoteRealtimeAudioSource::connection):
(WebKit::RemoteRealtimeAudioSource::startProducingData):
(WebKit::RemoteRealtimeAudioSource::stopProducingData):
(WebKit::RemoteRealtimeAudioSource::capabilities):
(WebKit::RemoteRealtimeAudioSource::applyConstraints):
(WebKit::RemoteRealtimeAudioSource::applyConstraintsSucceeded):
(WebKit::RemoteRealtimeAudioSource::applyConstraintsFailed):
(WebKit::RemoteRealtimeAudioSource::hasEnded):
(WebKit::RemoteRealtimeAudioSource::captureStopped):
(WebKit::RemoteRealtimeAudioSource::captureFailed):
(WebKit::RemoteRealtimeAudioSource::gpuProcessConnectionDidClose):
* WebProcess/cocoa/RemoteRealtimeAudioSource.h: Copied from Source/WebKit/WebProcess/cocoa/RemoteRealtimeMediaSource.h.
(WebKit::RemoteRealtimeAudioSource::identifier const):
* WebProcess/cocoa/RemoteRealtimeVideoSource.cpp: Renamed from Source/WebKit/WebProcess/cocoa/RemoteRealtimeMediaSource.cpp.
(WebKit::RemoteRealtimeVideoSource::create):
(WebKit::RemoteRealtimeVideoSource::RemoteRealtimeVideoSource):
(WebKit::RemoteRealtimeVideoSource::createRemoteMediaSource):
(WebKit::RemoteRealtimeVideoSource::~RemoteRealtimeVideoSource):
(WebKit::RemoteRealtimeVideoSource::whenReady):
(WebKit::RemoteRealtimeVideoSource::didFail):
(WebKit::RemoteRealtimeVideoSource::setAsReady):
(WebKit::RemoteRealtimeVideoSource::clone):
(WebKit::RemoteRealtimeVideoSource::setCapabilities):
(WebKit::RemoteRealtimeVideoSource::setSettings):
(WebKit::RemoteRealtimeVideoSource::remoteVideoSampleAvailable):
(WebKit::RemoteRealtimeVideoSource::connection):
(WebKit::RemoteRealtimeVideoSource::startProducingData):
(WebKit::RemoteRealtimeVideoSource::stopProducingData):
(WebKit::RemoteRealtimeVideoSource::setShouldApplyRotation):
(WebKit::RemoteRealtimeVideoSource::capabilities):
(WebKit::RemoteRealtimeVideoSource::applyConstraints):
(WebKit::RemoteRealtimeVideoSource::applyConstraintsSucceeded):
(WebKit::RemoteRealtimeVideoSource::applyConstraintsFailed):
(WebKit::RemoteRealtimeVideoSource::hasEnded):
(WebKit::RemoteRealtimeVideoSource::captureStopped):
(WebKit::RemoteRealtimeVideoSource::captureFailed):
(WebKit::RemoteRealtimeVideoSource::stopBeingObserved):
(WebKit::RemoteRealtimeVideoSource::requestToEnd):
(WebKit::RemoteRealtimeVideoSource::gpuProcessConnectionDidClose):
* WebProcess/cocoa/RemoteRealtimeVideoSource.h: Renamed from Source/WebKit/WebProcess/cocoa/RemoteRealtimeMediaSource.h.
* WebProcess/cocoa/UserMediaCaptureManager.cpp:
(WebKit::UserMediaCaptureManager::addAudioSource):
(WebKit::UserMediaCaptureManager::removeAudioSource):
(WebKit::UserMediaCaptureManager::addVideoSource):
(WebKit::UserMediaCaptureManager::removeVideoSource):
(WebKit::UserMediaCaptureManager::sourceStopped):
(WebKit::UserMediaCaptureManager::sourceEnded):
(WebKit::UserMediaCaptureManager::captureFailed):
(WebKit::UserMediaCaptureManager::sourceMutedChanged):
(WebKit::UserMediaCaptureManager::sourceSettingsChanged):
(WebKit::UserMediaCaptureManager::remoteVideoSampleAvailable):
(WebKit::UserMediaCaptureManager::applyConstraintsSucceeded):
(WebKit::UserMediaCaptureManager::applyConstraintsFailed):
(WebKit::UserMediaCaptureManager::AudioFactory::createAudioCaptureSource):
(WebKit::UserMediaCaptureManager::VideoFactory::createVideoCaptureSource):
(WebKit::UserMediaCaptureManager::DisplayFactory::createDisplayCaptureSource):
* WebProcess/cocoa/UserMediaCaptureManager.h:
(WebKit::UserMediaCaptureManager::remoteCaptureSampleManager):
2021-02-10 Wenson Hsieh <wenson_hsieh@apple.com>
[Webpage translation] Add support for the -_translate: action on WKContentView
https://bugs.webkit.org/show_bug.cgi?id=221713
Reviewed by Tim Horton.
Add another way to trigger web page translation on iOS, via the `-_translate:` selector.
* Platform/spi/ios/UIKitSPI.h:
Add a new staged SPI declaration on `UIWKTextInteractionAssistant`. Additionally remove an old staging
declaration for a method in `UIWebGeolocationPolicyDecider`.
* UIProcess/API/ios/WKWebViewIOS.h:
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _translateForWebView:]):
Grab the currently selected text and invoke the new method on `UIWKTextInteractionAssistant`.
(-[WKContentView _addShortcutForWebView:]):
Drive-by fix: remove an extraneous `nil` check.
(-[WKContentView canPerformActionForWebView:withSender:]):
Return `YES` if the selection is ranged, and we're not inside a password input.
2021-02-10 Jer Noble <jer.noble@apple.com>
Move AudioHardwareListener into the GPU Process
https://bugs.webkit.org/show_bug.cgi?id=221455
Reviewed by Eric Carlson.
Add a new Remote pair of classes, RemoteAudioHardwareListener/Proxy.
* CMakeLists.txt:
* DerivedSources-input.xcfilelist:
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::createAudioHardwareListener):
(WebKit::GPUConnectionToWebProcess::releaseAudioHardwareListener):
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/GPUConnectionToWebProcess.messages.in:
* GPUProcess/media/RemoteAudioHardwareListenerProxy.cpp:
(WebKit::RemoteAudioHardwareListenerProxy::RemoteAudioHardwareListenerProxy):
(WebKit::RemoteAudioHardwareListenerProxy::audioHardwareDidBecomeActive):
(WebKit::RemoteAudioHardwareListenerProxy::audioHardwareDidBecomeInactive):
(WebKit::RemoteAudioHardwareListenerProxy::audioOutputDeviceChanged):
* GPUProcess/media/RemoteAudioHardwareListenerProxy.h:
* GPUProcess/media/RemoteLegacyCDMFactoryProxy.cpp:
* Scripts/webkit/messages.py:
* Sources.txt:
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/media/RemoteAudioHardwareListener.cpp: Added.
(WebKit::RemoteAudioHardwareListener::create):
(WebKit::RemoteAudioHardwareListener::RemoteAudioHardwareListener):
(WebKit::RemoteAudioHardwareListener::~RemoteAudioHardwareListener):
(WebKit::RemoteAudioHardwareListener::gpuProcessConnectionDidClose):
(WebKit::RemoteAudioHardwareListener::audioHardwareDidBecomeActive):
(WebKit::RemoteAudioHardwareListener::audioHardwareDidBecomeInactive):
(WebKit::RemoteAudioHardwareListener::audioOutputDeviceChanged):
* WebProcess/GPU/media/RemoteAudioHardwareListener.h:
* WebProcess/GPU/media/RemoteAudioHardwareListener.messages.in:
* WebProcess/GPU/media/RemoteAudioHardwareListenerIdentifier.h:
* WebProcess/GPU/webrtc/MediaRecorderPrivate.h:
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::setUseGPUProcessForMedia):
2021-02-10 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Update - [ASCAppleIDCredential initWithUser:...] to match the latest SPI
https://bugs.webkit.org/show_bug.cgi?id=221723
<rdar://problem/74038383>
Reviewed by Brent Fulgham.
No tests.
* Platform/spi/Cocoa/AuthenticationServicesCoreSPI.h:
* UIProcess/WebAuthentication/Cocoa/AuthenticatorPresenterCoordinator.mm:
(WebKit::AuthenticatorPresenterCoordinator::dimissPresenter):
2021-02-10 Rob Buis <rbuis@igalia.com>
Use event loop to set title
https://bugs.webkit.org/show_bug.cgi?id=218496
Reviewed by Darin Adler.
Add title parameter to PrintFrame message.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::printFrame):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::print):
* WebProcess/WebCoreSupport/WebChromeClient.h:
2021-02-10 Myles C. Maxfield <mmaxfield@apple.com>
Move pal/spi/cocoa/CoreTextSPI.h to pal/spi/cf/CoreTextSPI.h
https://bugs.webkit.org/show_bug.cgi?id=221683
Reviewed by Simon Fraser.
Core Text has a C API, and doesn't depend on UIKit/AppKit. It also exists on Windows,
where there is no Cocoa. And pal/spi/cf already includes things like CoreAudioSPI.h
and CoreVideoSPI.h
* Shared/Cocoa/CoreTextHelpers.mm:
* Shared/Cocoa/WebCoreArgumentCodersCocoa.mm:
* UIProcess/ios/forms/WKNumberPadView.mm:
* WebProcess/WebPage/WebPage.cpp:
2021-02-10 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed, reverting r272702.
Caused assertion failure on macOS debug WK2 bots
Reverted changeset:
"Stop using GenericCallback from WebPageProxy"
https://bugs.webkit.org/show_bug.cgi?id=221653
https://trac.webkit.org/changeset/272702
2021-02-10 Alex Christensen <achristensen@webkit.org>
Stop using GenericCallback from WebPageProxy
https://bugs.webkit.org/show_bug.cgi?id=221653
Reviewed by Chris Dumez.
There are still two uses in the DrawingAreaProxy implementations, but those are quite tangled and deserve their own patch.
* UIProcess/API/APIAttachment.cpp:
(API::Attachment::updateAttributes):
* UIProcess/API/APIAttachment.h:
* UIProcess/API/C/WKPage.cpp:
(CompletionHandler<void):
(WKPageRenderTreeExternalRepresentation):
(WKPageGetSourceForFrame):
(WKPageGetContentsAsString):
(WKPageGetBytecodeProfile):
(WKPageGetSamplingProfilerOutput):
(WTF::Function<void): Deleted.
* UIProcess/API/Cocoa/APIAttachmentCocoa.mm:
(API::Attachment::updateFromSerializedRepresentation):
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _insertAttachmentWithFileWrapper:contentType:completion:]):
(-[WKWebView _getContentsAsStringWithCompletionHandler:]):
(-[WKWebView _getContentsOfAllFramesAsStringWithCompletionHandler:]):
* UIProcess/API/Cocoa/_WKAttachment.mm:
(-[_WKAttachment setFileWrapper:contentType:completion:]):
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::startSpeaking):
* UIProcess/GenericCallback.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::dispatchActivityStateChange):
(WebKit::WebPageProxy::getRenderTreeExternalRepresentation):
(WebKit::WebPageProxy::getSourceForFrame):
(WebKit::WebPageProxy::getContentsAsString):
(WebKit::WebPageProxy::getBytecodeProfile):
(WebKit::WebPageProxy::getSamplingProfilerOutput):
(WebKit::WebPageProxy::getSelectionOrContentsAsString):
(WebKit::WebPageProxy::didCommitLoadForFrame):
(WebKit::WebPageProxy::didFailLoadForFrame):
(WebKit::WebPageProxy::resetState):
(WebKit::WebPageProxy::installActivityStateChangeCompletionHandler):
(WebKit::WebPageProxy::insertAttachment):
(WebKit::WebPageProxy::updateAttachmentAttributes):
(WebKit::WebPageProxy::didInsertAttachmentWithIdentifier):
(WebKit::WebPageProxy::clearLoadDependentCallbacks): Deleted.
(WebKit::WebPageProxy::voidCallback): Deleted.
(WebKit::WebPageProxy::stringCallback): Deleted.
(WebKit::WebPageProxy::invalidateStringCallback): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _shareForWebView:]):
(-[WKContentView _defineForWebView:]):
(-[WKContentView accessibilityRetrieveSpeakSelectionContent]):
(-[WKContentView applyAutocorrection:toString:withCompletionHandler:]):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::applyAutocorrection):
* UIProcess/mac/WKTextFinderClient.mm:
(-[WKTextFinderClient getSelectedText:]):
* WebProcess/Notifications/NotificationPermissionRequestManager.h:
* WebProcess/WebCoreSupport/WebNotificationClient.h:
* WebProcess/WebPage/DrawingArea.h:
(WebKit::DrawingArea::activityStateDidChange):
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.mm:
(WebKit::RemoteLayerTreeDrawingArea::activityStateDidChange):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::reinitializeWebPage):
(WebKit::WebPage::setActivityState):
(WebKit::WebPage::getContentsAsString):
(WebKit::WebPage::getRenderTreeExternalRepresentation):
(WebKit::WebPage::getSelectionOrContentsAsString):
(WebKit::WebPage::getSourceForFrame):
(WebKit::WebPage::getBytecodeProfile):
(WebKit::WebPage::getSamplingProfilerOutput):
(WebKit::WebPage::insertAttachment):
(WebKit::WebPage::updateAttachmentAttributes):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::applyAutocorrection):
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.h:
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::handleActivityStateChangeCallbacks):
(WebKit::TiledCoreAnimationDrawingArea::activityStateDidChange):
2021-02-10 Simon Fraser <simon.fraser@apple.com>
REGRESSION (r269824) IOSurface allocation failure causes crash in RemoteLayerBackingStore::display()
https://bugs.webkit.org/show_bug.cgi?id=221729
rdar://72651289
Reviewed by Tim Horton.
The refactoring in r269824 dropped a null check on the front buffer surface (IOSurface allocation
can fail when the process reaches a hardcoded limit). Restore the null check.
* Shared/RemoteLayerTree/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::display):
2021-02-10 Peng Liu <peng.liu6@apple.com>
Remove MediaSourcePrivateClient::monitorSourceBuffers()
https://bugs.webkit.org/show_bug.cgi?id=220945
Reviewed by Youenn Fablet.
* GPUProcess/media/RemoteMediaSourceProxy.cpp:
(WebKit::RemoteMediaSourceProxy::monitorSourceBuffers): Deleted.
* GPUProcess/media/RemoteMediaSourceProxy.h:
2021-02-10 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Produce ClientDataJSON in the SPI
https://bugs.webkit.org/show_bug.cgi?id=221623
<rdar://problem/71509394>
Reviewed by Brent Fulgham.
Instead of asking the clients of the SPI to pre-calculate the ClientDataJSON hash for us, we do
that in the SPI space ourselves.
Covered by API tests.
* UIProcess/API/Cocoa/_WKAuthenticatorAssertionResponse.mm:
(-[_WKAuthenticatorAssertionResponse initWithClientDataJSON:rawId:extensions:authenticatorData:signature:userHandle:]):
(-[_WKAuthenticatorAssertionResponse initWithRawId:extensions:authenticatorData:signature:userHandle:]): Deleted.
* UIProcess/API/Cocoa/_WKAuthenticatorAssertionResponseInternal.h:
* UIProcess/API/Cocoa/_WKAuthenticatorAttestationResponse.mm:
(-[_WKAuthenticatorAttestationResponse initWithClientDataJSON:rawId:extensions:attestationObject:]):
(-[_WKAuthenticatorAttestationResponse initWithRawId:extensions:attestationObject:]): Deleted.
* UIProcess/API/Cocoa/_WKAuthenticatorAttestationResponseInternal.h:
* UIProcess/API/Cocoa/_WKAuthenticatorResponse.h:
* UIProcess/API/Cocoa/_WKAuthenticatorResponse.mm:
(-[_WKAuthenticatorResponse initWithClientDataJSON:rawId:extensions:]):
(-[_WKAuthenticatorResponse initWithRawId:extensions:]): Deleted.
* UIProcess/API/Cocoa/_WKAuthenticatorResponseInternal.h:
Adds a field to return the JSON serialized bytes for the ClientDataJSON.
* UIProcess/API/Cocoa/_WKWebAuthenticationPanel.h:
* UIProcess/API/Cocoa/_WKWebAuthenticationPanel.mm:
(produceClientDataJson):
(produceClientDataJsonHash):
(wkAuthenticatorAttestationResponse):
(-[_WKWebAuthenticationPanel makeCredentialWithChallenge:origin:options:completionHandler:]):
(wkAuthenticatorAssertionResponse):
(-[_WKWebAuthenticationPanel getAssertionWithChallenge:origin:options:completionHandler:]):
(-[_WKWebAuthenticationPanel makeCredentialWithHash:options:completionHandler:]): Deleted.
(-[_WKWebAuthenticationPanel getAssertionWithHash:options:completionHandler:]): Deleted.
Modifies the SPI to accept a challenge and an origin to calculate the ClientDataJSON.
2021-02-10 Don Olmstead <don.olmstead@sony.com>
Non-unified build fixes, early February 2021 edition
https://bugs.webkit.org/show_bug.cgi?id=221701
Unreviewed non-unified build fixes.
* NetworkProcess/PrivateClickMeasurementManager.cpp:
2021-02-10 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed, reverting r272507.
Caused TestWebKitAPI.WebKit.OnDeviceChangeCrash to become a
flaky timeout
Reverted changeset:
"[MacOS] Enable Audio Capture in GPUProcess by default"
https://bugs.webkit.org/show_bug.cgi?id=221400
https://trac.webkit.org/changeset/272507
2021-02-10 BJ Burg <bburg@apple.com>
[Cocoa] Web Inspector: add support for evaluating script on the inspected page via _WKInspectorExtension
https://bugs.webkit.org/show_bug.cgi?id=221567
<rdar://71208534>
Reviewed by Devin Rousso.
This patch adds a new method to _WKInspectorExtension which is used to implement
`browser.devtools.inspectedWindow.eval` in the Web Extensions API.
* Shared/InspectorExtensionTypes.h:
* Shared/InspectorExtensionTypes.cpp:
(WebKit::inspectorExtensionErrorToString):
Add enum value NotImplemented.
* UIProcess/API/APIInspectorExtension.h:
* UIProcess/API/APIInspectorExtension.cpp:
(API::InspectorExtension::evaluateScript):
Plumbing.
* UIProcess/API/Cocoa/WKWebViewInternal.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(nsErrorFromExceptionDetails):
Move this helper up near the top as it is now exposed via <WebKit/WKWebViewInternal.h>.
* UIProcess/API/Cocoa/_WKInspectorExtension.h:
* UIProcess/API/Cocoa/_WKInspectorExtension.mm:
(-[_WKInspectorExtension evaluateScript:frameURL:contextSecurityOrigin:useContentScriptContext:completionHandler:]):
Add new method to evaluate script in the inspected page. The semantics of the parameters
are intended to match those of `browser.devtools.inspectedWindow.eval`.
* UIProcess/Inspector/WebInspectorUIExtensionControllerProxy.h:
* UIProcess/Inspector/WebInspectorUIExtensionControllerProxy.cpp:
(WebKit::WebInspectorUIExtensionControllerProxy::evaluateScriptForExtension):
Plumbing.
* WebProcess/Inspector/WebInspectorUIExtensionController.h:
* WebProcess/Inspector/WebInspectorUIExtensionController.messages.in:
Add new message for evaluateScript.
* WebProcess/Inspector/WebInspectorUIExtensionController.cpp:
(WebKit::WebInspectorUIExtensionController::parseInspectorExtensionErrorFromEvaluationResult):
Support the new error enum value.
(WebKit::WebInspectorUIExtensionController::unregisterExtension):
(WebKit::WebInspectorUIExtensionController::createTabForExtension):
Drive-by, simplify this by passing the EvaluationResult value without unwrapping.
(WebKit::WebInspectorUIExtensionController::evaluateScriptForExtension):
This is the meat of the patch. Call out to InspectorFrontendAPI.evaluateScriptForExtension.
Inspect the return value and invoke the completion handler with the result or an error.
2021-02-10 Kimmo Kinnunen <kkinnunen@apple.com>
WebGL IPC messages are delivered out of order
https://bugs.webkit.org/show_bug.cgi?id=221488
Reviewed by Chris Dumez.
Sync messages are delivered before earlier async messages, if some other
WebKit part waits on sync replies.
No new tests, makes the existing WebGL --use-gpu-process tests a bit less
flakey.
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.h:
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.cpp:
Mark all async messages with IPC::SendOption::DispatchMessageEvenWhenWaitingForSyncReply.
There are no messages in the receiver message list that would make sense
to be delivered out of order.
2021-02-10 Wenson Hsieh <wenson_hsieh@apple.com>
Use HAVE(PEPPER_UI_CORE) instead of PLATFORM(WATCHOS) to guard code that uses PepperUICore
https://bugs.webkit.org/show_bug.cgi?id=221679
Reviewed by Tim Horton.
Use `HAVE(PEPPER_UI_CORE)` instead of `PLATFORM(WATCHOS)` in code that depends on PepperUICore, either directly
or indirectly. While technically equivalent, the former is more semantically precise. Common examples of this
include dependencies on Quickboard for text input, or PepperUICore category extensions on common UIKit classes
(e.g. digital crown support in `WKScrollView.mm`). No change in behavior.
Inspired by <https://bugs.webkit.org/show_bug.cgi?id=221649#c2>.
* Platform/spi/watchos/PepperUICoreSPI.h:
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _willStartScrollingOrZooming]):
(-[WKContentView _didEndScrollingOrZooming]):
(-[WKContentView _updateTextInputTraits:]):
(-[WKContentView _elementDidFocus:userIsInteracting:blurPreviousNode:activityStateChanges:userObject:]):
(-[WKContentView _elementDidBlur]):
(-[WKContentView reloadContextViewForPresentedListViewController]):
(-[WKContentView _wheelChangedWithEvent:]):
(-[WKContentView _simulateTextEntered:]):
(-[WKContentView selectFormAccessoryPickerRow:]):
(-[WKContentView selectFormAccessoryHasCheckedItemAtRow:]):
(-[WKContentView textContentTypeForTesting]):
(-[WKContentView formInputLabel]):
(-[WKContentView setTimePickerValueToHour:minute:]):
* UIProcess/ios/WKScrollView.mm:
(-[WKScrollView initWithFrame:]):
* UIProcess/ios/forms/WKDatePickerViewController.h:
* UIProcess/ios/forms/WKDatePickerViewController.mm:
* UIProcess/ios/forms/WKFocusedFormControlView.h:
* UIProcess/ios/forms/WKFocusedFormControlView.mm:
* UIProcess/ios/forms/WKNumberPadView.h:
* UIProcess/ios/forms/WKNumberPadView.mm:
* UIProcess/ios/forms/WKNumberPadViewController.h:
* UIProcess/ios/forms/WKNumberPadViewController.mm:
* UIProcess/ios/forms/WKQuickboardListViewController.h:
* UIProcess/ios/forms/WKQuickboardListViewController.mm:
* UIProcess/ios/forms/WKSelectMenuListViewController.h:
* UIProcess/ios/forms/WKSelectMenuListViewController.mm:
* UIProcess/ios/forms/WKTextInputListViewController.h:
* UIProcess/ios/forms/WKTextInputListViewController.mm:
* UIProcess/ios/forms/WKTimePickerViewController.h:
* UIProcess/ios/forms/WKTimePickerViewController.mm:
2021-02-10 Alex Christensen <achristensen@webkit.org>
REGRESSION (r272628): [macOS] TestWebKitAPI.WKWebView.SnapshotImageError is consistently failing
https://bugs.webkit.org/show_bug.cgi?id=221702
I got a little carried away with my error reduction. This one needs an error.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView takeSnapshotWithConfiguration:completionHandler:]):
2021-02-10 Kate Cheney <katherine_cheney@apple.com>
PCM: Expired reports get sent at the same time after a session restart
https://bugs.webkit.org/show_bug.cgi?id=221555
<rdar://problem/73724816>
Reviewed by John Wilander.
Since PCM data is now persisted, we need to address the case of a
session-restart after 24-48+ hours. We should not send all overdue
attributions in the same burst in case multiple have the same destination
and could identify a user cross-site.
This patch kicks off the timer to fire pending attributions on session-start
and sends one report at a time. If more than one overdue report exists
at any time, we schedule the timer for a random interval between 15 and
30 minutes.
In theory this could result in some attributions never being sent if a
user keeps quitting and restarting a session. In practice this is
probably unlikely. Protecting the user's privacy is a hard requirement,
so we think possible starvation of some reports is the right tradeoff.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::clearSentAttribution):
(WebKit::ResourceLoadStatisticsDatabaseStore::clearSentAttributions): Deleted.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.h:
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::clearSentAttribution):
(WebKit::WebResourceLoadStatisticsStore::clearSentAttributions): Deleted.
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.h:
Remove unused SQLite query. Update the query to sort attributed PCM
by earliestTimeToSend, which seems important now that we send only
one overdue report at a time. Change the clearSentAttributions
function to take only a single attribution.
* NetworkProcess/NetworkSession.cpp:
(WebKit::NetworkSession::NetworkSession):
Convert m_privateClickMeasurement to a unique_ptr so we can wait to
create it after WebResourceLoadStatisticsStore is created. This ensures
that the call to create the SQLite database will run on a
background thread before we try to collect overdue PCM attributions.
(WebKit::NetworkSession::storePrivateClickMeasurement):
(WebKit::NetworkSession::handlePrivateClickMeasurementConversion):
(WebKit::NetworkSession::dumpPrivateClickMeasurement):
(WebKit::NetworkSession::clearPrivateClickMeasurement):
(WebKit::NetworkSession::clearPrivateClickMeasurementForRegistrableDomain):
(WebKit::NetworkSession::setPrivateClickMeasurementOverrideTimerForTesting):
(WebKit::NetworkSession::markAttributedPrivateClickMeasurementsAsExpiredForTesting):
(WebKit::NetworkSession::setPrivateClickMeasurementConversionURLForTesting):
(WebKit::NetworkSession::markPrivateClickMeasurementsAsExpiredForTesting):
(WebKit::NetworkSession::firePrivateClickMeasurementTimerImmediately):
* NetworkProcess/NetworkSession.h:
* NetworkProcess/PrivateClickMeasurementManager.cpp:
(WebKit::PrivateClickMeasurementManager::PrivateClickMeasurementManager):
(WebKit::PrivateClickMeasurementManager::firePendingAttributionRequest):
(WebKit::PrivateClickMeasurementManager::attribute):
Drive-by fixes to add protectedThis and check if PrivateClickMeasurementManager
is still alive when this lambda is called.
2021-02-10 Wenson Hsieh <wenson_hsieh@apple.com>
[watchOS] Adopt PUICQuickboardController for text input
https://bugs.webkit.org/show_bug.cgi?id=221649
Reviewed by Tim Horton.
Refactor text input on watchOS to use `PUICQuickboardController`, instead of custom WebKit subclasses of
`PUICQuickboardListViewController`. This new API serves the same purpose as the current list view controller
subclass, by providing a view that offers options to dictate, scribble, and choose text suggestions. See below
for more details.
For the time being, this new `PUICQuickboardController` doesn't provide a way to render a custom header view
for us to show the domain name; to avoid introducing a security or privacy bug in the interim, guard this behind
a WebKit-exposed runtime-enabled setting that's off by default. We can remove this setting and instead just use
the `HAVE(QUICKBOARD_CONTROLLER)` build-time flag once support for the custom header view is complete.
* Platform/spi/watchos/PepperUICoreSPI.h:
Include a couple of new headers if `HAVE(QUICKBOARD_CONTROLLER)` is set.
* UIProcess/ios/WKContentViewInteraction.h:
Add a `_presentedQuickboardController` instance variable if `HAVE(QUICKBOARD_CONTROLLER)` is set.
* UIProcess/ios/WKContentViewInteraction.mm:
Add protocol conformance to `PUICQuickboardControllerDelegate`.
(-[WKContentView _updateInteractionTintColor:]):
(-[WKContentView tintColorDidChange]):
Refactor this to take `UITextInputTraits`, instead of using `_traits`.
(-[WKContentView textInputTraits]):
(-[WKContentView _updateTextInputTraits:]):
Refactor this to modify the given set of text input traits given current focused element information state,
rather than operating on the `_traits` instance variable. This refactoring allows us to share logic when setting
up text input traits on both iOS and watchOS; in the former case, we're setting up a set of cached `_traits` on
the content view, but in the latter case, we're modifying a newly created `PUICTextInputContext` that we'll then
set on the `PUICQuickboardController`.
(-[WKContentView _createQuickboardController:]):
Add a helper method to create and return a new `PUICQuickboardController` that's suitable for the current
focused element's state.
(-[WKContentView presentViewControllerForCurrentFocusedElement]):
Present the view controller for the focused element, by either using a custom WebKit-owned
`PUICQuickboardListViewController` subclass, or by using a new `PUICQuickboardController` with a custom text
input context. Note that in the latter case, we may not be able to invoke the UI delegate hook
`-_webView:didPresentFocusedElementViewController:`, since the presenting view controller's transition
coordinator may not have been created yet in the scenario where we're spinning up the Quickboard view service
for the first time, since the call to establish the XPC connection is asynchronous in PepperUICore, and we don't
attempt to present the remote view controller until the connection is established. This makes it so that we
can't rely on the `-didPresentFocusedElementViewController:` hook in WebKitTestRunner to know when a
`PUICQuickboardController` has finished presenting.
To work around this, we teach the test harness to override `-presentViewController:animated:completion:` and
call into the test runner's `WKWebView` subclass when a remote quickboard controller is done presenting. In the
longer term, we will require SPI from PepperUICore to present a `PUICQuickboardController` with a completion
block.
(-[WKContentView _isPresentingFullScreenInputView]):
A "full screen input view" for the focused element is now present if either `_presentedQuickboardController` is
set, or `_presentedFullScreenInputViewController` is set.
(-[WKContentView dismissAllInputViewControllers:]):
Dismiss either the currently presented `_presentedQuickboardController` or
`_presentedFullScreenInputViewController`, and invoke the UI delegate method
`-_webView:didDismissFocusedElementViewController:` when finished.
(-[WKContentView quickboardController:textInputValueDidChange:]):
(-[WKContentView quickboardControllerTextInputValueCancelled:]):
(-[WKContentView dismissQuickboardViewControllerAndRevealFocusedFormOverlayIfNecessary:]):
(-[WKContentView textContentTypeForQuickboard]):
Refactor `-_updateTextInputTraits:` to call a helper method (`-textContentTypeForQuickboard`) on watchOS to
convert the current focused element information to a `UITextContentType`.
(-[WKContentView textContentTypeForListViewController:]):
(-[WKContentView _simulateTextEntered:]):
(-[WKContentView textContentTypeForTesting]):
(-[WKContentView formInputLabel]):
Update a few internal testing hooks to handle the case where we have a `_presentedQuickboardController` instead
of a `_presentedFullScreenInputViewController`.
(-[WKContentView _updateInteractionTintColor]): Deleted.
* UIProcess/ios/forms/WKTextInputListViewController.mm:
(-[WKTextInputListViewController additionalTrayButtons]): Suppress a deprecation warning.
2021-02-10 Myles C. Maxfield <mmaxfield@apple.com>
Replace ChangeLog entry I accidentally deleted in r272657
Unreviewed.
2021-02-10 Myles C. Maxfield <mmaxfield@apple.com>
Revert r272586 because we're not ready yet
We're still trying to figure out the right way to trigger this behavior.
Unreviewed.
* WebProcess/EntryPoint/Cocoa/XPCService/WebContentService/Info-OSX.plist:
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-10 Brent Fulgham <bfulgham@apple.com>
Create stub methods to support finer-grained control over loading
https://bugs.webkit.org/show_bug.cgi?id=221430
<rdar://problem/73999547>
Reviewed by Geoffrey Garen.
Create three new WebKit Cocoa methods that accept NSURLRequest, rather than URL. This allows
us to pass additional hints to the networking subsystem.
Tested with new API Tests.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView loadSimulatedRequest:withResponse:responseData:]):
(-[WKWebView loadSimulatedRequest:withResponseHTMLString:]):
(-[WKWebView loadFileRequest:allowingReadAccessToURL:]):
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
2021-02-10 Sam Weinig <weinig@apple.com>
Workaround some order dependent issues by parenting remote layers before applying other properties
https://bugs.webkit.org/show_bug.cgi?id=221585
Reviewed by Simon Fraser.
This runs the updateChildren (renamed to applyHierarchyUpdates for consistency) part of
the RemoteLayerTreePropertyApplier for all the transaction changes prior to running the
rest of applyProperties to work around some downstream bugs that currently require parented
layers to work. This likely won't be required forever, but it should be harmless to do.
* Shared/RemoteLayerTree/RemoteLayerTreePropertyApplier.h:
* Shared/RemoteLayerTree/RemoteLayerTreePropertyApplier.mm:
(WebKit::RemoteLayerTreePropertyApplier::applyProperties):
(WebKit::RemoteLayerTreePropertyApplier::applyHierarchyUpdates):
(WebKit::RemoteLayerTreePropertyApplier::updateChildren): Deleted.
* UIProcess/RemoteLayerTree/RemoteLayerTreeHost.mm:
(WebKit::RemoteLayerTreeHost::updateLayerTree):
2021-02-10 Kimmo Kinnunen <kkinnunen@apple.com>
RemoteGraphicsContextGLProxy should support losing the context and should lose the context on timeouts
https://bugs.webkit.org/show_bug.cgi?id=221396
Reviewed by Simon Fraser.
Implement support for RemoteGraphicsContextGLProxy acting on
losing the context in RemoteGraphicsContextGL.
Implement support for RemoteGraphicsContextGLProxy timing out on message
sends. Force the context to be lost.
Implement support for GPUProcessConnection disconnect. Force the context
be lost.
* GPUProcess/graphics/RemoteGraphicsContextGL.cpp:
Send a bool via WasCreated method, so that the
receiver knows if creation was successful or not.
Currently the creation always succeeds.
Previously it was designed to use wasLost instead,
but this has the problem that the client will wait
for WasCreated in particular cases to ensure initialization,
and thus we cannot wait for "WasCreated OR WasLost".
(WebKit::RemoteGraphicsContextGL::RemoteGraphicsContextGL):
(WebKit::RemoteGraphicsContextGL::synthesizeGLError):
(WebKit::RemoteGraphicsContextGL::getError):
Manually implement few methods that contain custom client-side
logic wrt. lost context.
* GPUProcess/graphics/RemoteGraphicsContextGL.h:
* GPUProcess/graphics/RemoteGraphicsContextGL.messages.in:
* GPUProcess/graphics/RemoteGraphicsContextGLFunctionsGenerated.h:
(setFailNextGPUStatusCheck):
(getBooleanv):
* WebProcess/GPU/GPUProcessConnection.cpp:
(WebKit::GPUProcessConnection::dispatchMessage):
Skip the RemoteGraphicsContextGLProxy messages that do not have
a receiver. This is expected, as we cannot assume that only expected
messages arrive to the message hander.
The client, i.e. the web process, might
- just delete the proxy
- force the context to be lost
- notice a timeout and force the context to be lost
while the GPU process cannot know this. RemoteGraphicsContextGL might
already have sent messages to the Proxy.
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.cpp:
(WebKit::RemoteGraphicsContextGLProxy::create):
(WebKit::RemoteGraphicsContextGLProxy::RemoteGraphicsContextGLProxy):
(WebKit::RemoteGraphicsContextGLProxy::~RemoteGraphicsContextGLProxy):
Listen to GPUProcessConnection::Client notifications about
connection closure. When the connection closes, mark the context object
lost.
Store the reference to GPUProcessConnection so that when the connection
is shut down and new one is instantiated, RemoteGraphicsContextGLProxy
destruction logic will correctly refer to the one it was using.
(WebKit::RemoteGraphicsContextGLProxy::reshape):
(WebKit::RemoteGraphicsContextGLProxy::prepareForDisplay):
(WebKit::RemoteGraphicsContextGLProxy::ensureExtensionEnabled):
(WebKit::RemoteGraphicsContextGLProxy::notifyMarkContextChanged):
(WebKit::RemoteGraphicsContextGLProxy::synthesizeGLError):
(WebKit::RemoteGraphicsContextGLProxy::getError):
(WebKit::RemoteGraphicsContextGLProxy::wasCreated):
(WebKit::RemoteGraphicsContextGLProxy::wasLost):
When the context lost event happens, disconnect with
the GPUProcessConnection.
(WebKit::RemoteGraphicsContextGLProxy::wasChanged):
(WebKit::RemoteGraphicsContextGLProxy::markContextLost):
(WebKit::RemoteGraphicsContextGLProxy::waitUntilInitialized):
(WebKit::RemoteGraphicsContextGLProxy::gpuProcessConnectionDidClose):
When the connection closes, mark the context object
lost.
(WebKit::RemoteGraphicsContextGLProxy::disconnect):
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.h:
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.messages.in:
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxyFunctionsGenerated.cpp:
Changes generated by the code-generator wrt handling timeouts from the sends.
2021-02-10 Philippe Normand <pnormand@igalia.com>
[WPE] Optionally build Cog as external project and replacement for MiniBrowser
https://bugs.webkit.org/show_bug.cgi?id=221518
Reviewed by Adrian Perez de Castro.
* PlatformWPE.cmake: Produce uninstalled pkg-config files for wpe-webkit and web-extension
libraries. Also populate the JSC ForwardedHeaders path during the build.
* wpe/wpe-web-extension-uninstalled.pc.in: Added.
* wpe/wpe-webkit-uninstalled.pc.in: Added.
2021-02-10 Carlos Garcia Campos <cgarcia@igalia.com>
[SOUP] Simplify ResourceRequest and ResourceResponse
https://bugs.webkit.org/show_bug.cgi?id=221543
Reviewed by Adrian Perez de Castro.
Use new ResourceRequest and ResourceResponse API.
* NetworkProcess/soup/NetworkDataTaskSoup.cpp:
(WebKit::NetworkDataTaskSoup::createRequest): Use ResourceRequest::createSoupMessage().
(WebKit::NetworkDataTaskSoup::didSendRequest): Use ResourceResponse passing the SoupMessage and sniffed content type.
(WebKit::NetworkDataTaskSoup::didRequestNextPart): Create a ResourceResponse passing the url, content type,
expected length and text encoding. Then call ResourceResponse::updateFromSoupMessageHeaders() to set the headers.
* NetworkProcess/soup/NetworkSessionSoup.cpp:
(WebKit::NetworkSessionSoup::createWebSocketTask): Use ResourceRequest::createSoupMessage() and pass the
ResourceRequest to WebSocketTask constructor.
* NetworkProcess/soup/WebSocketTaskSoup.cpp:
(WebKit::WebSocketTask::WebSocketTask): Save the ResourceRequest and then just call
ResourceRequest::updateFromSoupMessageHeaders() to update the headers after the response has been sent by libsoup.
(WebKit::WebSocketTask::didConnect): Use ResourceResponse constructor with the SoupMessage directly.
(WebKit::WebSocketTask::didFail): Ditto.
* NetworkProcess/soup/WebSocketTaskSoup.h:
2021-02-09 Alex Christensen <achristensen@webkit.org>
Use CompletionHandler instead of ApplicationManifestCallback
https://bugs.webkit.org/show_bug.cgi?id=221627
Reviewed by Chris Dumez.
* UIProcess/API/C/WKPage.cpp:
(WKPageGetApplicationManifest_b):
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _getApplicationManifestWithCompletionHandler:]):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::getApplicationManifest):
(WebKit::WebPageProxy::applicationManifestCallback): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::getApplicationManifest):
(WebKit::WebPage::didFinishLoadingApplicationManifest):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2021-02-09 Alex Christensen <achristensen@webkit.org>
Use CompletionHandler instead of DataCallback
https://bugs.webkit.org/show_bug.cgi?id=221639
Reviewed by Chris Dumez.
* UIProcess/API/C/WKFrame.cpp:
(WKFrameGetMainResourceData):
(WKFrameGetResourceData):
(WKFrameGetWebArchive):
* UIProcess/API/C/WKPage.cpp:
(WKPageGetSelectionAsWebArchiveData):
(WKPageDrawPagesToPDF):
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView createWebArchiveDataWithCompletionHandler:]):
(-[WKWebView _getMainResourceDataWithCompletionHandler:]):
* UIProcess/WebFrameProxy.cpp:
(WebKit::WebFrameProxy::getWebArchive):
(WebKit::WebFrameProxy::getMainResourceData):
(WebKit::WebFrameProxy::getResourceData):
* UIProcess/WebFrameProxy.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::CompletionHandler<void):
(WebKit::WebPageProxy::getContentsAsMHTMLData):
(WebKit::WebPageProxy::getSelectionAsWebArchiveData):
(WebKit::WebPageProxy::getMainResourceDataOfFrame):
(WebKit::WebPageProxy::getResourceDataFromFrame):
(WebKit::WebPageProxy::getWebArchiveOfFrame):
(WebKit::WebPageProxy::drawPagesToPDF):
(WebKit::WebPageProxy::dataCallback): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/mac/WKPrintingView.mm:
(-[WKPrintingView _preparePDFDataForPrintingOnSecondaryThread]):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::getContentsAsMHTMLData):
(WebKit::WebPage::getSelectionAsWebArchiveData):
(WebKit::WebPage::getMainResourceDataOfFrame):
(WebKit::WebPage::getResourceDataFromFrame):
(WebKit::WebPage::getWebArchiveOfFrame):
(WebKit::WebPage::drawPagesToPDF):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2021-02-09 Alex Christensen <achristensen@webkit.org>
Use CompletionHandler instead of PrintFinishedCallback
https://bugs.webkit.org/show_bug.cgi?id=221643
Reviewed by Chris Dumez.
* UIProcess/API/gtk/WebKitPrintOperation.cpp:
(webkitPrintOperationPrintPagesForFrame):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::drawPagesForPrinting):
(WebKit::WebPageProxy::printFinishedCallback): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::drawPagesForPrinting):
(WebKit::WebPage::didFinishPrintOperation): Deleted.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/gtk/WebPrintOperationGtk.cpp:
(WebKit::WebPrintOperationGtk::printDone):
* WebProcess/WebPage/gtk/WebPrintOperationGtk.h:
2021-02-09 Devin Rousso <drousso@apple.com>
Unreviewed, build fix after r272630
* UIProcess/ios/WKActionSheetAssistant.h:
* UIProcess/ios/WKActionSheetAssistant.mm:
2021-02-09 Devin Rousso <drousso@apple.com>
[iOS] Add support for the language/subtitle tracks button using `UIMenu`
https://bugs.webkit.org/show_bug.cgi?id=221594
<rdar://problem/74129129>
Reviewed by Eric Carlson and Wenson Hsieh.
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::showMediaControlsContextMenu): Added.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::showMediaControlsContextMenu): Added.
* UIProcess/WebPageProxy.messages.in:
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::showMediaControlsContextMenu): Added.
Send the given `MediaControlsContextMenuItem`s to the UIProcess for the `UIMenu`.
* UIProcess/PageClient.h:
(WebKit::PageClient::showMediaControlsContextMenu): Added.
* UIProcess/ios/PageClientImplIOS.h:
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::showMediaControlsContextMenu): Added.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView actionSheetAssistantDidShowContextMenu:]): Added.
(-[WKContentView actionSheetAssistantDidDismissContextMenu:]): Added.
(-[WKContentView _showMediaControlsContextMenu:items:completionHandler:]): Added.
(-[WKContentView _contentsOfUserInterfaceItem:]):
Pass the given `MediaControlsContextMenuItem`s to the `WKActionSheetAssistant`.
* UIProcess/ios/WKActionSheetAssistant.h:
* UIProcess/ios/WKActionSheetAssistant.mm:
(-[WKActionSheetAssistant dealloc]):
(-[WKActionSheetAssistant _removeContextMenuInteractions]): Added.
(-[WKActionSheetAssistant _ensureDataDetectorContextMenuInteraction]): Added.
(-[WKActionSheetAssistant _ensureMediaControlsContextMenuInteraction]): Added.
(-[WKActionSheetAssistant hasContextMenuInteraction]):
(-[WKActionSheetAssistant showDataDetectorsUIForPositionInformation:]):
(-[WKActionSheetAssistant _uiMenuElementsForMediaControlContextMenuItems:]): Added.
(-[WKActionSheetAssistant showMediaControlsContextMenu:items:completionHandler:]): Added.
(-[WKActionSheetAssistant contextMenuInteraction:configurationForMenuAtLocation:]):
(-[WKActionSheetAssistant contextMenuInteraction:previewForHighlightingMenuWithConfiguration:]):
(-[WKActionSheetAssistant contextMenuInteraction:willDisplayMenuForConfiguration:animator:]): Added.
(-[WKActionSheetAssistant contextMenuInteraction:willEndForConfiguration:animator:]):
(-[WKActionSheetAssistant _contextMenuInteraction:overrideSuggestedActionsForConfiguration:]):
(-[WKActionSheetAssistant _contentsOfContextMenuItem:]): Added.
(-[WKActionSheetAssistant currentlyAvailableActionTitles]): Added.
(-[WKActionSheetAssistant currentlyAvailableMediaControlsContextMenuItems]): Added.
(-[WKActionSheetAssistant currentAvailableActionTitles]): Deleted.
(-[WKActionSheetAssistant removeContextMenuInteraction]): Deleted.
(-[WKActionSheetAssistant ensureContextMenuInteraction]): Deleted.
Instead of assuming that the only `UIContextMenuInteraction` is for data detectors, check
the given `interaction` to see if it's the data detector interaction before doing any work.
This allows the media controls `UIContextMenuInteraction` to exist side-by-side and use the
`WKActionSheetAssistant` as its delegate too (especially since it's not possible for both
contextmenus to be active at the same time). Present the media controls contextmenu without
a preview (it uses an empty `UIView` for proper sizing) at the location provided. Selecting
a `UIAction` will send the `MediaControlsContextMenuItem::ID` back to the WebProcess.
Also add support for
- `UIScriptController.prototype.contentsOfUserInterfaceItem`
- `UIScriptController.prototype.didShowContextMenuCallback`
- `UIScriptController.prototype.didDismissContextMenuCallback`
for usage in the added `LayoutTests/media/modern-media-controls/tracks-support/ios/` tests.
2021-02-09 Chris Dumez <cdumez@apple.com>
Make sure we are no longer show the previous page when running a JS prompt
https://bugs.webkit.org/show_bug.cgi?id=215782
<rdar://problem/67698601>
Reviewed by Simon Fraser.
Make sure we are no longer show the previous page when running a JS prompt.
If we have not yet done a layer tree commit since the last load commit, then
we are likely still showing the previous page. If we are asked to run a JS
prompt / alert / confirm at this point, it would be confusing to still show
the previous page. In order to address the issue, we now make the view blank
in such scenario (ideally, we'd have painted the new page but this is
currently not a trivial thing to do).
To make the view blank, the approach chosen was to add a blank overlay view
on top of the content. This overlay view gets taken down as soon as we
paint the view again.
* SourcesCocoa.txt:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _hasBlankOverlay]):
(-[WKWebView _setHasBlankOverlay:]):
* UIProcess/API/Cocoa/WKWebViewInternal.h:
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/Cocoa/PageClientImplCocoa.h:
* UIProcess/Cocoa/PageClientImplCocoa.mm:
(WebKit::PageClientImplCocoa::setHasBlankOverlay):
* UIProcess/Cocoa/WKBlankOverlayView.h: Added.
* UIProcess/Cocoa/WKBlankOverlayView.mm: Added.
(-[WKBlankOverlayView initWithFrame:]):
* UIProcess/PageClient.h:
(WebKit::PageClient::setHasBlankOverlay):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didFirstLayerTreeCommitAfterCommittingLoad):
(WebKit::WebPageProxy::makeViewBlankIfUnpaintedSinceLastLoadCommit):
(WebKit::WebPageProxy::didCommitLoadForFrame):
(WebKit::WebPageProxy::runJavaScriptAlert):
(WebKit::WebPageProxy::runJavaScriptConfirm):
(WebKit::WebPageProxy::runJavaScriptPrompt):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::didCommitLayerTree):
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::didUpdateRenderingAfterCommittingLoad):
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::didCommitLoad):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::updateRendering):
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::didUpdateRendering):
2021-02-09 Alex Christensen <achristensen@webkit.org>
Use CompletionHandler instead of ImageCallback
https://bugs.webkit.org/show_bug.cgi?id=221626
Reviewed by Chris Dumez.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView takeSnapshotWithConfiguration:completionHandler:]):
* UIProcess/API/Cocoa/_WKThumbnailView.mm:
(-[_WKThumbnailView requestSnapshot]):
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _snapshotRect:intoImageOfWidth:completionHandler:]):
* UIProcess/GenericCallback.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::drawRectToImage):
(WebKit::WebPageProxy::takeSnapshot):
(WebKit::WebPageProxy::imageCallback): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/mac/WKPrintingView.mm:
(-[WKPrintingView _drawPreview:]):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::takeSnapshot):
(WebKit::WebPage::drawRectToImage):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2021-02-09 Alex Christensen <achristensen@webkit.org>
Use CompletionHandler instead of UnsignedCallback
https://bugs.webkit.org/show_bug.cgi?id=221631
Reviewed by Chris Dumez.
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::characterIndexForPoint):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::replaceMatches):
(WebKit::WebPageProxy::characterIndexForPointAsync):
(WebKit::WebPageProxy::unsignedCallback): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView selectTextWithGranularity:atPoint:completionHandler:]):
(-[WKContentView beginSelectionInDirection:completionHandler:]):
(-[WKContentView updateSelectionWithExtentPoint:completionHandler:]):
(-[WKContentView updateSelectionWithExtentPoint:withBoundary:completionHandler:]):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::beginSelectionInDirection):
(WebKit::WebPageProxy::updateSelectionWithExtentPoint):
(WebKit::WebPageProxy::updateSelectionWithExtentPointAndBoundary):
* UIProcess/mac/WKTextFinderClient.mm:
(-[WKTextFinderClient replaceMatches:withString:inSelectionOnly:resultCollector:]):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::replaceMatches):
(WebKit::WebPage::characterIndexForPointAsync):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::beginSelectionInDirection):
(WebKit::WebPage::updateSelectionWithExtentPointAndBoundary):
(WebKit::WebPage::updateSelectionWithExtentPoint):
2021-02-09 Peng Liu <peng.liu6@apple.com>
[GPUP] Run ImageDecoderAVFObjC in the GPU Process
https://bugs.webkit.org/show_bug.cgi?id=221317
Reviewed by Jer Noble.
This patch adds the support to run `ImageDecoderAVFObjC` in the GPU process when the
"GPU Process: Media" preference is on. The `RemoteImageDecoderAVF` running in the Web
process forwards video data to the `RemoteImageDecoderAVFProxy` running in the GPU process,
which parses the data and returns frame information ("hasAlpha" and "duration") to the
Web process. `RemoteImageDecoderAVFProxy` will generate (decode) frames (images) after
receiving requests from `RemoteImageDecoderAVF`. The decoded frames are transferred to
the Web process through IOSurfaces, and the frames will be cached in the Web process.
* DerivedSources-input.xcfilelist:
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::imageDecoderAVFProxy):
(WebKit::GPUConnectionToWebProcess::dispatchMessage):
(WebKit::GPUConnectionToWebProcess::dispatchSyncMessage):
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/media/RemoteImageDecoderAVFProxy.cpp: Added.
(WebKit::RemoteImageDecoderAVFProxy::RemoteImageDecoderAVFProxy):
(WebKit::RemoteImageDecoderAVFProxy::createDecoder):
(WebKit::RemoteImageDecoderAVFProxy::deleteDecoder):
(WebKit::RemoteImageDecoderAVFProxy::encodedDataStatusChanged):
(WebKit::RemoteImageDecoderAVFProxy::setExpectedContentSize):
(WebKit::RemoteImageDecoderAVFProxy::setData):
(WebKit::RemoteImageDecoderAVFProxy::createFrameImageAtIndex):
* GPUProcess/media/RemoteImageDecoderAVFProxy.h: Added.
* GPUProcess/media/RemoteImageDecoderAVFProxy.messages.in: Added.
* Scripts/webkit/messages.py:
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/media/RemoteImageDecoderAVF.cpp: Added.
(WebKit::RemoteImageDecoderAVF::RemoteImageDecoderAVF):
(WebKit::RemoteImageDecoderAVF::~RemoteImageDecoderAVF):
(WebKit::RemoteImageDecoderAVF::gpuProcessConnection const):
(WebKit::RemoteImageDecoderAVF::canDecodeType):
(WebKit::RemoteImageDecoderAVF::supportsMediaType):
(WebKit::RemoteImageDecoderAVF::encodedDataStatus const):
(WebKit::RemoteImageDecoderAVF::setEncodedDataStatusChangeCallback):
(WebKit::RemoteImageDecoderAVF::size const):
(WebKit::RemoteImageDecoderAVF::frameCount const):
(WebKit::RemoteImageDecoderAVF::repetitionCount const):
(WebKit::RemoteImageDecoderAVF::uti const):
(WebKit::RemoteImageDecoderAVF::filenameExtension const):
(WebKit::RemoteImageDecoderAVF::frameSizeAtIndex const):
(WebKit::RemoteImageDecoderAVF::frameIsCompleteAtIndex const):
(WebKit::RemoteImageDecoderAVF::frameMetadataAtIndex const):
(WebKit::RemoteImageDecoderAVF::frameDurationAtIndex const):
(WebKit::RemoteImageDecoderAVF::frameHasAlphaAtIndex const):
(WebKit::RemoteImageDecoderAVF::frameAllowSubsamplingAtIndex const):
(WebKit::RemoteImageDecoderAVF::frameBytesAtIndex const):
(WebKit::RemoteImageDecoderAVF::createFrameImageAtIndex):
(WebKit::RemoteImageDecoderAVF::setExpectedContentSize):
(WebKit::RemoteImageDecoderAVF::setData):
(WebKit::RemoteImageDecoderAVF::clearFrameBufferCache):
(WebKit::RemoteImageDecoderAVF::encodedDataStatusChanged):
* WebProcess/GPU/media/RemoteImageDecoderAVF.h: Added.
* WebProcess/GPU/media/RemoteImageDecoderAVFManager.cpp: Added.
(WebKit::RemoteImageDecoderAVFManager::createImageDecoder):
(WebKit::RemoteImageDecoderAVFManager::deleteRemoteImageDecoder):
(WebKit::RemoteImageDecoderAVFManager::RemoteImageDecoderAVFManager):
(WebKit::RemoteImageDecoderAVFManager::~RemoteImageDecoderAVFManager):
(WebKit::RemoteImageDecoderAVFManager::supplementName):
(WebKit::RemoteImageDecoderAVFManager::gpuProcessConnection const):
(WebKit::RemoteImageDecoderAVFManager::setUseGPUProcess):
(WebKit::RemoteImageDecoderAVFManager::encodedDataStatusChanged):
* WebProcess/GPU/media/RemoteImageDecoderAVFManager.h: Added.
* WebProcess/GPU/media/RemoteImageDecoderAVFManager.messages.in: Added.
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp: Fix unified build failures.
* WebProcess/GPU/webrtc/MediaRecorderPrivate.h: Ditto.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences):
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::setUseGPUProcessForMedia):
2021-02-09 Eric Carlson <eric.carlson@apple.com>
[macOS] Add internal preference to control how AVOutputContext is allocated
https://bugs.webkit.org/show_bug.cgi?id=221583
<rdar://73830632>
Reviewed by Jer Noble.
* UIProcess/WebPageProxy.h:
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::useiTunesAVOutputContext const):
2021-02-09 Martin Robinson <mrobinson@webkit.org>
Implement scroll-snap-stop for scroll snapping
https://bugs.webkit.org/show_bug.cgi?id=197744
<rdar://problem/50708356>
Reviewed by Simon Fraser.
* Shared/RemoteLayerTree/RemoteScrollingCoordinatorTransaction.cpp: Add encoding and decoding support
for the SnapOffset struct.
(ArgumentCoder<SnapOffset<float>>::encode):
(ArgumentCoder<SnapOffset<float>>::decode):
* UIProcess/RemoteLayerTree/ios/RemoteScrollingCoordinatorProxyIOS.mm: Update to use SnapOffset struct.
(WebKit::RemoteScrollingCoordinatorProxy::shouldSnapForMainFrameScrolling const):
(WebKit::RemoteScrollingCoordinatorProxy::hasActiveSnapPoint const):
(WebKit::RemoteScrollingCoordinatorProxy::nearestActiveContentInsetAdjustedSnapOffset const):
2021-02-09 Per Arne <pvollan@apple.com>
[macOS] Deny mach-lookup to the distributed notifications service
https://bugs.webkit.org/show_bug.cgi?id=221604
<rdar://problem/69169123>
Reviewed by Brent Fulgham.
Deny mach-lookup to the distributed notifications service in the WebContent process on macOS.
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-09 Alex Christensen <achristensen@webkit.org>
Use CompletionHandler instead of ComputedPagesCallback
https://bugs.webkit.org/show_bug.cgi?id=221619
Reviewed by Chris Dumez.
* UIProcess/API/C/WKPage.cpp:
(WKPageComputePagesForPrinting):
* UIProcess/GenericCallback.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::computePagesForPrinting):
(WebKit::WebPageProxy::computedPagesCallback): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/mac/WKPrintingView.mm:
(-[WKPrintingView _askPageToComputePageRects]):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::computePagesForPrinting):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2021-02-09 Alex Christensen <achristensen@webkit.org>
Use CompletionHandler instead of NowPlayingInfoCallback
https://bugs.webkit.org/show_bug.cgi?id=221617
Reviewed by Chris Dumez.
* UIProcess/API/Cocoa/WKWebViewTesting.mm:
(-[WKWebView _requestActiveNowPlayingSessionInfo:]):
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::requestActiveNowPlayingSessionInfo):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::requestActiveNowPlayingSessionInfo): Deleted.
(WebKit::WebPageProxy::nowPlayingInfoCallback): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebPage/Cocoa/WebPageCocoa.mm:
(WebKit::WebPage::requestActiveNowPlayingSessionInfo):
* WebProcess/WebPage/WebPage.messages.in:
2021-02-09 Alex Christensen <achristensen@webkit.org>
Synthesize range responses if needed in WebCoreNSURLSession
https://bugs.webkit.org/show_bug.cgi?id=221072
Reviewed by Geoff Garen.
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::requestResource):
2021-02-09 Peng Liu <peng.liu6@apple.com>
[GPUP] Test media/track/audio-track-add-remove.html crashes on debug bots
https://bugs.webkit.org/show_bug.cgi?id=221595
Reviewed by Eric Carlson.
Under stress tests, instances of RemoteAudioTrackProxy, RemoteTextTrackProxy, and
RemoteVideoTrackProxy might be destroyed in the GPU process when an IPC message
comes and leads to an assertion failure.
To fix the assertion failure, this patches removes three IPC message receivers in
the GPU process:
- RemoteAudioTrackProxy
- RemoteTextTrackProxy
- RemoteVideoTrackProxy
Instead of using these three message receivers, this patch adds three new IPC messages
to RemoteMediaPlayerProxy:
- AudioTrackSetEnabled
- TextTrackSetMode
- VideoTrackSetSelected
No new tests, fixing following tests:
- media/track/audio-track-add-remove.html
- media/track/audio-track.html
* CMakeLists.txt:
* DerivedSources.make:
* GPUProcess/media/RemoteAudioTrackProxy.cpp:
(WebKit::RemoteAudioTrackProxy::RemoteAudioTrackProxy):
(WebKit::RemoteAudioTrackProxy::~RemoteAudioTrackProxy):
* GPUProcess/media/RemoteAudioTrackProxy.h:
* GPUProcess/media/RemoteAudioTrackProxy.messages.in: Removed.
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
(WebKit::RemoteMediaPlayerProxy::audioTrackSetEnabled):
(WebKit::RemoteMediaPlayerProxy::videoTrackSetSelected):
(WebKit::RemoteMediaPlayerProxy::textTrackSetMode):
* GPUProcess/media/RemoteMediaPlayerProxy.h:
* GPUProcess/media/RemoteMediaPlayerProxy.messages.in:
* GPUProcess/media/RemoteTextTrackProxy.cpp:
(WebKit::RemoteTextTrackProxy::RemoteTextTrackProxy):
(WebKit::RemoteTextTrackProxy::~RemoteTextTrackProxy):
* GPUProcess/media/RemoteTextTrackProxy.h:
* GPUProcess/media/RemoteTextTrackProxy.messages.in: Removed.
* GPUProcess/media/RemoteVideoTrackProxy.cpp:
(WebKit::RemoteVideoTrackProxy::RemoteVideoTrackProxy):
(WebKit::RemoteVideoTrackProxy::~RemoteVideoTrackProxy):
* GPUProcess/media/RemoteVideoTrackProxy.h:
* GPUProcess/media/RemoteVideoTrackProxy.messages.in: Removed.
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/media/AudioTrackPrivateRemote.cpp:
(WebKit::AudioTrackPrivateRemote::AudioTrackPrivateRemote):
(WebKit::AudioTrackPrivateRemote::setEnabled):
* WebProcess/GPU/media/AudioTrackPrivateRemote.h:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::addRemoteAudioTrack):
(WebKit::MediaPlayerPrivateRemote::addRemoteTextTrack):
(WebKit::MediaPlayerPrivateRemote::addRemoteVideoTrack):
* WebProcess/GPU/media/TextTrackPrivateRemote.cpp:
(WebKit::TextTrackPrivateRemote::TextTrackPrivateRemote):
(WebKit::TextTrackPrivateRemote::setMode):
* WebProcess/GPU/media/TextTrackPrivateRemote.h:
* WebProcess/GPU/media/VideoTrackPrivateRemote.cpp:
(WebKit::VideoTrackPrivateRemote::VideoTrackPrivateRemote):
(WebKit::VideoTrackPrivateRemote::setSelected):
* WebProcess/GPU/media/VideoTrackPrivateRemote.h:
(WebKit::VideoTrackPrivateRemote::create):
2021-02-09 Chris Dumez <cdumez@apple.com>
Unreviewed, fix build with the latest SDK.
* UIProcess/ios/forms/WKFormColorControl.mm:
(-[WKColorPicker selectColor:]):
2021-02-09 Aditya Keerthi <akeerthi@apple.com>
[iOS][FCR] Use UIColorPickerViewController for color inputs
https://bugs.webkit.org/show_bug.cgi?id=221572
<rdar://problem/72183130>
Reviewed by Sam Weinig.
UIColorPickerViewController was added in iOS 14, while WebKit still
uses a custom color picker. To stay consistent with the rest of the
platform, this patch drops the custom color picker and adopts
UIColorPickerViewController for <input type=color> on iOS.
Test: fast/forms/ios/choose-color-from-color-picker.html
* Platform/spi/ios/UIKitSPI.h:
Added SPI declarations for _UISheetPresentationController and
UIColorPickerViewController.
* Shared/FocusedElementInformation.cpp:
(WebKit::FocusedElementInformation::encode const):
(WebKit::FocusedElementInformation::decode):
* Shared/FocusedElementInformation.h:
Added colorValue member to avoid parsing the input's value
in the UIProcess.
* SourcesCocoa.txt:
* UIProcess/API/ios/WKWebViewPrivateForTestingIOS.h:
* UIProcess/API/ios/WKWebViewTestingIOS.mm:
(-[WKWebView setSelectedColorForColorPicker:]):
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _shouldShowAutomaticKeyboardUIIgnoringInputMode]):
(-[WKContentView _elementTypeRequiresAccessoryView:]):
The new color picker does not appear as a keyboard input view.
(-[WKContentView updateFocusedElementValueAsColor:]):
(-[WKContentView setSelectedColorForColorPicker:]):
* UIProcess/ios/forms/WKFormColorControl.h:
* UIProcess/ios/forms/WKFormColorControl.mm:
(-[WKColorPicker initWithView:]):
(-[WKColorPicker selectColor:]):
Note that the delegate method is not called when programmatically
setting the selected color.
(-[WKColorPicker focusedElementSuggestedColors]):
Suggested colors from a <datalist> element are displayed in the
favorites view.
(-[WKColorPicker updateColorPickerState]):
(-[WKColorPicker configurePresentation]):
On iPad, the color picker is displayed a popover. On iPhone, the picker
is displayed as a half-sheet, and can be dragged up into a fullscreen
view.
(-[WKColorPicker controlBeginEditing]):
(-[WKColorPicker controlEndEditing]):
(-[WKColorPicker presentationControllerDidDismiss:]):
(-[WKColorPicker colorPickerViewControllerDidSelectColor:]):
(-[WKColorPicker colorPickerViewControllerDidFinish:]):
(-[WKFormColorControl initWithView:]):
(-[WKFormColorControl selectColor:]):
* UIProcess/ios/forms/WKFormColorPicker.h: Removed.
* UIProcess/ios/forms/WKFormColorPicker.mm: Removed.
* UIProcess/ios/forms/WKFormSelectPicker.h:
Build fix.
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::getFocusedElementInformation):
2021-02-09 Alex Christensen <achristensen@webkit.org>
Use CompletionHandler instead of DrawToPDFCallback
https://bugs.webkit.org/show_bug.cgi?id=220120
Reviewed by Chris Dumez.
Along the way I made all asynchronous message replies use 8 fewer bytes by using the destination ID for the listener ID
instead of encoding 0 as the destination ID and then encoding the listener ID. This was needed in order to make implementing
waitForAsyncCallbackAndDispatchImmediately possible, which looks at the destination IDs of incoming messages.
I also made it so that if we are just snapshotting the first page, we don't send a synchronous message because we already know the page count:
1!
* Platform/IPC/Connection.cpp:
(IPC::Connection::dispatchMessage):
* Platform/IPC/Connection.h:
(IPC::Connection::sendWithAsyncReply):
(IPC::Connection::waitForAndDispatchImmediately):
(IPC::Connection::waitForAsyncCallbackAndDispatchImmediately):
* Platform/IPC/HandleMessage.h:
(IPC::handleMessageAsync):
(IPC::handleMessageAsyncWantsConnection):
* Platform/IPC/JSIPCBinding.cpp:
(IPC::jsValueForDecodedArgumentValue):
* Platform/IPC/JSIPCBinding.h:
* Platform/IPC/MessageSender.h:
(IPC::MessageSender::sendWithAsyncReply):
* Scripts/webkit/messages.py:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView createPDFWithConfiguration:completionHandler:]):
* UIProcess/AuxiliaryProcessProxy.h:
(WebKit::AuxiliaryProcessProxy::sendWithAsyncReply):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::drawToPDF):
(WebKit::WebPageProxy::getLoadDecisionForIcon):
(WebKit::WebPageProxy::drawToPDFCallback): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/ios/WKContentView.mm:
(-[WKContentView _processDidExit]):
(-[WKContentView _wk_pageCountForPrintFormatter:]):
(-[WKContentView _waitForDrawToPDFCallback]):
(-[WKContentView _wk_printedDocument]):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::computePagesForPrintingAndDrawToPDF):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::drawToPDF):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::computePagesForPrintingiOS):
(WebKit::WebPage::drawToPDFiOS):
(WebKit::WebPage::computePagesForPrintingAndDrawToPDF): Deleted.
2021-02-09 Per Arne <pvollan@apple.com>
[macOS] Deny mach-lookup to the fonts service
https://bugs.webkit.org/show_bug.cgi?id=221610
<rdar://problem/69168609>
Reviewed by Brent Fulgham.
By setting the Info.plist key CTIgnoreUserFonts to true, mach-lookup to the fonts service in the WebContent process can
be denied on macOS.
No new tests, covered by existing tests.
* WebProcess/EntryPoint/Cocoa/XPCService/WebContentService/Info-OSX.plist:
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-09 Per Arne <pvollan@apple.com>
[macOS] Deny mach-lookup to the service 'com.apple.trustd.agent'
https://bugs.webkit.org/show_bug.cgi?id=221611
<rdar://problem/68935818>
Reviewed by Brent Fulgham.
Deny mach-lookup to the service 'com.apple.trustd.agent' in the WebContent process on macOS.
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-09 Wenson Hsieh <wenson_hsieh@apple.com>
REGRESSION (r271660): Unable to interact with page after long-pressing image on images.google.com
https://bugs.webkit.org/show_bug.cgi?id=221584
<rdar://problem/74073581>
Reviewed by Andy Estes.
After long pressing on an image with an active touchend event listener, it's possible to sometimes get stuck in
a state where further interaction with the page is impossible, due to all gesture recognizers (except for the
touchend deferral gestures) remaining in either Failed or Ended state.
When presenting the context menu with a long press, the touch event gesture transitions to Failed state due to
the active touch being cancelled. Normally, this invokes `-_webTouchEventsRecognized:` with all touch points
being released, which allows us to "lift" the deferred gesture gate by failing any deferring gesture recognizers
that are still in Possible state (i.e. deferring native gestures).
However, it's possible for touch deferring gestures (in particular, the touchend deferrer) introduced in r271660
to end (i.e. call `-touchesEnded:withEvent:`) before the touch event gesture gets a chance to call
`-_webTouchEventsRecognized:`. In this scenario, the touch end deferral gesture remains in Possible state, and
prevents the touch event gesture from subsequently firing its action (`-_webTouchEventsRecognized:`), presumably
because UIKit is waiting for all other gestures in the same subgraph as the touch event gesture recognizer to
Fail.
This effectively results in a gesture "deadlock", since the web touch event gesture recognizer won't call
`-_webTouchEventsRecognized:` until the touch end deferring gesture has failed, and the touch end deferring
gesture won't fail until `-_webTouchEventsRecognized:` is called. To fix this, we restore a bit of logic that
was removed with r271193, such that we allow our deferring gesture recognizers to transition to Failed state
underneath `-touchesEnded:withEvent:`, as long it's [1] not actively deferring any native gestures, and [2] the
web touch event gesture has already failed, so we aren't expecting a subsequent call to
`-_webTouchEventsRecognized:` until the deferring gesture has failed.
Note that this check for the touch event gesture recognizer's state (condition [2] above) is necessary to
prevent the touchend deferring gesture from failing prematurely (i.e. right before we're about to dispatch
preventable `touchend` events).
Test: fast/events/touch/ios/tap-after-long-press-on-image.html
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _isTouchStartDeferringGesture:]):
(-[WKContentView _isTouchEndDeferringGesture:]):
(-[WKContentView deferringGestureRecognizer:didEndTouchesWithEvent:]):
Add a delegate hook when a deferring gesture's ends touches; use this hook in `WKContentView` to "lift" the
gesture gate if needed, in the case where the touch event gesture recognizer has already failed and we can't
expect `-_webTouchEventsRecognized:` to be called with all touch points released.
(-[WKContentView contextMenuInteraction:willEndForConfiguration:animator:]):
Add a missing call to `-[WKWebView _didDismissContextMenu]` here to make `UIHelper.waitForContextMenuToHide()`
actually work with image and link context menus in WebKitTestRunner.
* UIProcess/ios/WKDeferringGestureRecognizer.h:
* UIProcess/ios/WKDeferringGestureRecognizer.mm:
(-[WKDeferringGestureRecognizer touchesEnded:withEvent:]):
2021-02-09 Philippe Normand <pnormand@igalia.com>
Unreviewed, tvOS build fix after r272573
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::requestUserMediaPermissionForSpeechRecognition): That ifdef was
updated by mistake in r272573.
2021-02-09 Philippe Normand <pnormand@igalia.com>
[GLib] Permission request API for MediaKeySystem access support
https://bugs.webkit.org/show_bug.cgi?id=221199
Reviewed by Carlos Garcia Campos.
Expose a new Permission Request object to the public WPE and GTK APIs, allowing the
application to defer the decision of using a given EME MediaKeySystem to the user's
approval.
Covered by new API test.
* PlatformGTK.cmake:
* PlatformWPE.cmake:
* SourcesGTK.txt:
* SourcesWPE.txt:
* UIProcess/API/glib/WebKitMediaKeySystemPermissionRequest.cpp: Added.
(webkitMediaKeySystemPermissionRequestAllow):
(webkitMediaKeySystemPermissionRequestDeny):
(webkit_permission_request_interface_init):
(webkitMediaKeySystemPermissionRequestDispose):
(webkit_media_key_system_permission_request_class_init):
(webkitMediaKeySystemPermissionRequestCreate):
(webkit_media_key_system_permission_get_name):
* UIProcess/API/glib/WebKitMediaKeySystemPermissionRequestPrivate.h: Added.
* UIProcess/API/glib/WebKitUIClient.cpp:
* UIProcess/API/gtk/WebKitAutocleanups.h:
* UIProcess/API/gtk/WebKitMediaKeySystemPermissionRequest.h: Added.
* UIProcess/API/gtk/docs/webkit2gtk-4.0-sections.txt:
* UIProcess/API/gtk/docs/webkit2gtk-docs.sgml:
* UIProcess/API/gtk/webkit2.h:
* UIProcess/API/wpe/WebKitMediaKeySystemPermissionRequest.h: Added.
* UIProcess/API/wpe/docs/wpe-1.0-sections.txt:
* UIProcess/API/wpe/docs/wpe-docs.sgml:
2021-02-09 Philippe Normand <pnormand@igalia.com>
Permission request API for MediaKeySystem access support
https://bugs.webkit.org/show_bug.cgi?id=221187
Reviewed by Jer Noble.
Add new IPC and C API exposing MediaKeySystem access permissions to the UIProcess. At least
GTK/WPE would like to have API for asking permission to the user when EME's MediaKeyAccess
is being requested by a page. On Apple platforms the permission is granted by default until
a decision is made about supporting this in their APIUIClient.
* CMakeLists.txt:
* Headers.cmake:
* Platform/Logging.h:
* Shared/API/APIObject.h:
* Shared/API/c/WKBase.h:
* Sources.txt:
* UIProcess/API/APIUIClient.h:
(API::UIClient::decidePolicyForMediaKeySystemPermissionRequest):
* UIProcess/API/C/WKAPICast.h:
* UIProcess/API/C/WKMediaKeySystemPermissionCallback.cpp: Added.
(WKMediaKeySystemPermissionCallbackGetTypeID):
(WKMediaKeySystemPermissionCallbackComplete):
* UIProcess/API/C/WKMediaKeySystemPermissionCallback.h: Added.
* UIProcess/API/C/WKPage.cpp:
(WKPageSetPageUIClient):
* UIProcess/API/C/WKPageUIClient.h:
* UIProcess/MediaKeySystemPermissionRequest.h: Added.
(WebKit::MediaKeySystemPermissionRequest::create):
(WebKit::MediaKeySystemPermissionRequest::complete):
(WebKit::MediaKeySystemPermissionRequest::keySystem const):
(WebKit::MediaKeySystemPermissionRequest::MediaKeySystemPermissionRequest):
(WebKit::MediaKeySystemPermissionCallback::create):
(WebKit::MediaKeySystemPermissionCallback::complete):
(WebKit::MediaKeySystemPermissionCallback::MediaKeySystemPermissionCallback):
* UIProcess/MediaKeySystemPermissionRequestManagerProxy.cpp: Added.
(WebKit::MediaKeySystemPermissionRequestManagerProxy::MediaKeySystemPermissionRequestManagerProxy):
(WebKit::MediaKeySystemPermissionRequestManagerProxy::~MediaKeySystemPermissionRequestManagerProxy):
(WebKit::MediaKeySystemPermissionRequestManagerProxy::invalidatePendingRequests):
(WebKit::MediaKeySystemPermissionRequestManagerProxy::denyRequest):
(WebKit::MediaKeySystemPermissionRequestManagerProxy::grantRequest):
(WebKit::MediaKeySystemPermissionRequestManagerProxy::createRequestForFrame):
(WebKit::MediaKeySystemPermissionRequestManagerProxy::logChannel const):
(WebKit::MediaKeySystemPermissionRequestManagerProxy::logger const):
* UIProcess/MediaKeySystemPermissionRequestManagerProxy.h: Added.
(WebKit::MediaKeySystemPermissionRequestManagerProxy::page const):
(WebKit::MediaKeySystemPermissionRequestManagerProxy::denyRequest):
* UIProcess/MediaKeySystemPermissionRequestProxy.cpp: Added.
(WebKit::MediaKeySystemPermissionRequestProxy::MediaKeySystemPermissionRequestProxy):
(WebKit::MediaKeySystemPermissionRequestProxy::allow):
(WebKit::MediaKeySystemPermissionRequestProxy::deny):
(WebKit::MediaKeySystemPermissionRequestProxy::invalidate):
(WebKit::MediaKeySystemPermissionRequestProxy::doDefaultAction):
* UIProcess/MediaKeySystemPermissionRequestProxy.h: Added.
(WebKit::MediaKeySystemPermissionRequestProxy::create):
(WebKit::MediaKeySystemPermissionRequestProxy::isPending const):
(WebKit::MediaKeySystemPermissionRequestProxy::mediaKeySystemID const):
(WebKit::MediaKeySystemPermissionRequestProxy::mainFrameID const):
(WebKit::MediaKeySystemPermissionRequestProxy::frameID const):
(WebKit::MediaKeySystemPermissionRequestProxy::topLevelDocumentSecurityOrigin):
(WebKit::MediaKeySystemPermissionRequestProxy::topLevelDocumentSecurityOrigin const):
(WebKit::MediaKeySystemPermissionRequestProxy::keySystem const):
* UIProcess/SpeechRecognitionRemoteRealtimeMediaSource.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::requestMediaKeySystemPermissionForFrame):
(WebKit::WebPageProxy::mediaKeySystemPermissionRequestManager):
(WebKit::WebPageProxy::requestMediaKeySystemPermissionByDefaultAction):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebProcess/EncryptedMedia/MediaKeySystemPermissionRequestManager.cpp: Added.
(WebKit::MediaKeySystemPermissionRequestManager::MediaKeySystemPermissionRequestManager):
(WebKit::MediaKeySystemPermissionRequestManager::startMediaKeySystemRequest):
(WebKit::MediaKeySystemPermissionRequestManager::sendMediaKeySystemRequest):
(WebKit::MediaKeySystemPermissionRequestManager::cancelMediaKeySystemRequest):
(WebKit::MediaKeySystemPermissionRequestManager::mediaKeySystemWasGranted):
(WebKit::MediaKeySystemPermissionRequestManager::mediaKeySystemWasDenied):
* WebProcess/EncryptedMedia/MediaKeySystemPermissionRequestManager.h: Added.
* WebProcess/WebCoreSupport/WebMediaKeySystemClient.cpp: Added.
(WebKit::WebMediaKeySystemClient::WebMediaKeySystemClient):
(WebKit::WebMediaKeySystemClient::pageDestroyed):
(WebKit::WebMediaKeySystemClient::requestMediaKeySystem):
(WebKit::WebMediaKeySystemClient::cancelMediaKeySystemRequest):
* WebProcess/WebCoreSupport/WebMediaKeySystemClient.h: Added.
(WebKit::WebMediaKeySystemClient::~WebMediaKeySystemClient):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
(WebKit::WebPage::mediaKeySystemWasGranted):
(WebKit::WebPage::mediaKeySystemWasDenied):
* WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::mediaKeySystemPermissionRequestManager):
* WebProcess/WebPage/WebPage.messages.in:
2021-02-08 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r272485.
https://bugs.webkit.org/show_bug.cgi?id=221588
Required revisions were reversed
Reverted changeset:
"[GLib] Permission request API for MediaKeySystem access
support"
https://bugs.webkit.org/show_bug.cgi?id=221199
https://trac.webkit.org/changeset/272485
2021-02-08 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r272480, r272481, and r272500.
https://bugs.webkit.org/show_bug.cgi?id=221586
Caused assertion failure seen with EME tests
Reverted changesets:
"Permission request API for MediaKeySystem access support"
https://bugs.webkit.org/show_bug.cgi?id=221187
https://trac.webkit.org/changeset/272480
"Unreviewed, build fix after r272480"
https://trac.webkit.org/changeset/272481
"Permission request API for MediaKeySystem access support"
https://bugs.webkit.org/show_bug.cgi?id=221187
https://trac.webkit.org/changeset/272500
2021-02-08 Simon Fraser <simon.fraser@apple.com>
Crash with remote layer tree logging
https://bugs.webkit.org/show_bug.cgi?id=221580
Reviewed by Tim Horton.
Null-check animation.timingFunction when dumping remote layer tree animations.
* WebProcess/WebPage/RemoteLayerTree/PlatformCAAnimationRemote.mm:
(WebKit::operator<<):
2021-02-08 Alex Christensen <achristensen@webkit.org>
Use CompletionHandler instead of BoolCallback
https://bugs.webkit.org/show_bug.cgi?id=221575
<rdar://problem/74058604>
Reviewed by Chris Dumez.
This makes the code more modern, and it makes our fuzzers better at reasoning what the boolean value is going to be used for.
It should also fix a rare crash involving BoolCallback by completely removing it.
* UIProcess/GenericCallback.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::handlePreventableTouchEvent):
(WebKit::WebPageProxy::handleUnpreventableTouchEvent):
(WebKit::WebPageProxy::boolCallback): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebPage/EventDispatcher.cpp:
(WebKit::EventDispatcher::touchEventWithoutCallback):
(WebKit::EventDispatcher::touchEvent):
(WebKit::EventDispatcher::dispatchTouchEvents):
* WebProcess/WebPage/EventDispatcher.h:
* WebProcess/WebPage/EventDispatcher.messages.in:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::touchEventSync):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::dispatchAsynchronousTouchEvents):
(WebKit::WebPage::cancelAsynchronousTouchEvents):
2021-02-08 Brady Eidson <beidson@apple.com>
Make the UserContentController for ServiceWorker pages be non-optional.
<rdar://problem/71434565> and https://bugs.webkit.org/show_bug.cgi?id=221503
Reviewed by Alex Christensen.
Normal Pages always have a UserContentProvider, even if they just create an empty default one.
Same should be true for ServiceWorkers.
* Shared/ServiceWorkerInitializationData.cpp:
(WebKit::ServiceWorkerInitializationData::decode):
* Shared/ServiceWorkerInitializationData.h:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::userContentControllerIdentifierForServiceWorkers):
(WebKit::WebProcessPool::createWebPage):
* UIProcess/WebProcessPool.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::enableServiceWorkers):
* UIProcess/WebProcessProxy.h:
* WebProcess/Storage/WebSWContextManagerConnection.cpp:
(WebKit::m_userContentController):
(WebKit::m_userAgent): Deleted.
* WebProcess/Storage/WebSWContextManagerConnection.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
2021-02-08 Chris Dumez <cdumez@apple.com>
Prevent WebProcess to re-enter on sync IPC to the GPUProcess
https://bugs.webkit.org/show_bug.cgi?id=221566
<rdar://problem/74108466>
Reviewed by Ryosuke Niwa.
Prevent WebProcess to re-enter on sync IPC to the GPUProcess, similarly to what we already
do for sync IPC to other processes (Network, UI).
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::GPUConnectionToWebProcess):
2021-02-08 Youenn Fablet <youenn@apple.com>
[MacOS] Enable Audio Capture in GPUProcess by default
https://bugs.webkit.org/show_bug.cgi?id=221400
Reviewed by Eric Carlson.
Manually tested.
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultCaptureAudioInGPUProcessEnabled):
2021-02-08 David Kilzer <ddkilzer@apple.com>
Check that CGPDFDocumentRef is consistently unlocked and printable
<https://webkit.org/b/221512>
<rdar://problem/65277765>
Reviewed by Daniel Bates.
This patch does the following when printing PDFs:
- Checks return value of CGPDFDocumentUnlockWithPassword().
- Calls CGPDFDocumentAllowsPrinting() consistently.
- Handles -_ensureDocumentForPrinting and -_wk_printedDocument
returning nullptr.
* UIProcess/_WKWebViewPrintFormatter.mm:
(-[_WKWebViewPrintFormatter drawInRect:forPageAtIndex:]):
- Check return value of -_wk_printedDocument since it may return
nullptr now.
* UIProcess/ios/WKPDFView.mm:
(-[WKPDFView _ensureDocumentForPrinting]):
- Check return value of CGPDFDocumentUnlockWithPassword(), and
return nullptr if unsuccessful.
- Add CGPDFDocumentAllowsPrinting() check and return nullptr if
it fails.
(-[WKPDFView _wk_pageCountForPrintFormatter:]):
- Move the CGPDFDocumentAllowsPrinting() check into
-[WKPDFView _ensureDocumentForPrinting].
- Check return value of -_ensureDocumentForPrinting since it may
return nullptr now.
2021-02-08 Youenn Fablet <youenn@apple.com>
NetworkRTCSocketCocoa extractDataMessages should not read too much data
https://bugs.webkit.org/show_bug.cgi?id=221544
Reviewed by Eric Carlson.
* NetworkProcess/webrtc/NetworkRTCSocketCocoa.mm:
Make use of WebCore method.
2021-02-08 Youenn Fablet <youenn@apple.com>
Enable GPU WebRTC codecs in GPUProcess by default on MacOS
https://bugs.webkit.org/show_bug.cgi?id=221397
Reviewed by Eric Carlson.
Manually tested.
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultWebRTCCodecsInGPUProcess):
2021-02-08 Manuel Rego Casasnovas <rego@igalia.com>
[GTK] event.ctrlKey and other are false in keydown event
https://bugs.webkit.org/show_bug.cgi?id=221456
Reviewed by Adrian Perez de Castro.
Modifier keys were not being set on "keydown" events.
For example if you check event.ctrlKey (and other) you'll get always false when the user only typed Ctrl key
(it'll be true when it's in combination with some other key, like Ctrl + P).
This is due to some differences on how X system manages the modifier keys (see http://crbug.com/127142#c8),
but this patch will align WebKitGTK with other platforms and browsers on this case.
Test: fast/events/keyboardevent-modifier.html
* Shared/gtk/WebEventFactory.cpp:
(WebKit::modifiersForEvent): In the case of a keyboard press event check if some modifier key has been pressed.
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseSynthesizeKeyEvent): Check if some modifier key has been pressed.
2021-02-08 Antoine Quint <graouts@webkit.org>
Unreviewed unified build fix.
* UIProcess/MediaPermissionUtilities.h:
2021-02-08 Philippe Normand <pnormand@igalia.com>
[GLib] Permission request API for MediaKeySystem access support
https://bugs.webkit.org/show_bug.cgi?id=221199
Reviewed by Carlos Garcia Campos.
Expose a new Permission Request object to the public WPE and GTK APIs, allowing the
application to defer the decision of using a given EME MediaKeySystem to the user's
approval.
Covered by new API test.
* PlatformGTK.cmake:
* PlatformWPE.cmake:
* SourcesGTK.txt:
* SourcesWPE.txt:
* UIProcess/API/glib/WebKitMediaKeySystemPermissionRequest.cpp: Added.
(webkitMediaKeySystemPermissionRequestAllow):
(webkitMediaKeySystemPermissionRequestDeny):
(webkit_permission_request_interface_init):
(webkitMediaKeySystemPermissionRequestDispose):
(webkit_media_key_system_permission_request_class_init):
(webkitMediaKeySystemPermissionRequestCreate):
(webkit_media_key_system_permission_get_name):
* UIProcess/API/glib/WebKitMediaKeySystemPermissionRequestPrivate.h: Added.
* UIProcess/API/glib/WebKitUIClient.cpp:
* UIProcess/API/gtk/WebKitAutocleanups.h:
* UIProcess/API/gtk/WebKitMediaKeySystemPermissionRequest.h: Added.
* UIProcess/API/gtk/docs/webkit2gtk-4.0-sections.txt:
* UIProcess/API/gtk/docs/webkit2gtk-docs.sgml:
* UIProcess/API/gtk/webkit2.h:
* UIProcess/API/wpe/WebKitMediaKeySystemPermissionRequest.h: Added.
* UIProcess/API/wpe/docs/wpe-1.0-sections.txt:
* UIProcess/API/wpe/docs/wpe-docs.sgml:
2021-02-08 Philippe Normand <pnormand@igalia.com>
Unreviewed, unified build fix after r272480
* WebProcess/Speech/SpeechRecognitionRealtimeMediaSourceManager.cpp: Logging.h is needed for
log channel macros usage.
2021-02-08 Philippe Normand <pnormand@igalia.com>
Unreviewed, build fix after r272480
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::requestMediaKeySystemPermissionByDefaultAction):
* UIProcess/WebPageProxy.h:
2021-02-08 Philippe Normand <pnormand@igalia.com>
Permission request API for MediaKeySystem access support
https://bugs.webkit.org/show_bug.cgi?id=221187
Reviewed by Jer Noble.
Add new IPC and C API exposing MediaKeySystem access permissions to the UIProcess. At least
GTK/WPE would like to have API for asking permission to the user when EME's MediaKeyAccess
is being requested by a page. On Apple platforms the permission is granted by default until
a decision is made about supporting this in their APIUIClient.
* CMakeLists.txt:
* Headers.cmake:
* Platform/Logging.h:
* Shared/API/APIObject.h:
* Shared/API/c/WKBase.h:
* Sources.txt:
* UIProcess/API/APIUIClient.h:
(API::UIClient::decidePolicyForMediaKeySystemPermissionRequest):
* UIProcess/API/C/WKAPICast.h:
* UIProcess/API/C/WKMediaKeySystemPermissionCallback.cpp: Added.
(WKMediaKeySystemPermissionCallbackGetTypeID):
(WKMediaKeySystemPermissionCallbackComplete):
* UIProcess/API/C/WKMediaKeySystemPermissionCallback.h: Added.
* UIProcess/API/C/WKPage.cpp:
(WKPageSetPageUIClient):
* UIProcess/API/C/WKPageUIClient.h:
* UIProcess/MediaKeySystemPermissionRequest.h: Added.
(WebKit::MediaKeySystemPermissionRequest::create):
(WebKit::MediaKeySystemPermissionRequest::complete):
(WebKit::MediaKeySystemPermissionRequest::keySystem const):
(WebKit::MediaKeySystemPermissionRequest::MediaKeySystemPermissionRequest):
(WebKit::MediaKeySystemPermissionCallback::create):
(WebKit::MediaKeySystemPermissionCallback::complete):
(WebKit::MediaKeySystemPermissionCallback::MediaKeySystemPermissionCallback):
* UIProcess/MediaKeySystemPermissionRequestManagerProxy.cpp: Added.
(WebKit::MediaKeySystemPermissionRequestManagerProxy::MediaKeySystemPermissionRequestManagerProxy):
(WebKit::MediaKeySystemPermissionRequestManagerProxy::~MediaKeySystemPermissionRequestManagerProxy):
(WebKit::MediaKeySystemPermissionRequestManagerProxy::invalidatePendingRequests):
(WebKit::MediaKeySystemPermissionRequestManagerProxy::denyRequest):
(WebKit::MediaKeySystemPermissionRequestManagerProxy::grantRequest):
(WebKit::MediaKeySystemPermissionRequestManagerProxy::createRequestForFrame):
(WebKit::MediaKeySystemPermissionRequestManagerProxy::logChannel const):
(WebKit::MediaKeySystemPermissionRequestManagerProxy::logger const):
* UIProcess/MediaKeySystemPermissionRequestManagerProxy.h: Added.
(WebKit::MediaKeySystemPermissionRequestManagerProxy::page const):
(WebKit::MediaKeySystemPermissionRequestManagerProxy::denyRequest):
* UIProcess/MediaKeySystemPermissionRequestProxy.cpp: Added.
(WebKit::MediaKeySystemPermissionRequestProxy::MediaKeySystemPermissionRequestProxy):
(WebKit::MediaKeySystemPermissionRequestProxy::allow):
(WebKit::MediaKeySystemPermissionRequestProxy::deny):
(WebKit::MediaKeySystemPermissionRequestProxy::invalidate):
(WebKit::MediaKeySystemPermissionRequestProxy::doDefaultAction):
* UIProcess/MediaKeySystemPermissionRequestProxy.h: Added.
(WebKit::MediaKeySystemPermissionRequestProxy::create):
(WebKit::MediaKeySystemPermissionRequestProxy::isPending const):
(WebKit::MediaKeySystemPermissionRequestProxy::mediaKeySystemID const):
(WebKit::MediaKeySystemPermissionRequestProxy::mainFrameID const):
(WebKit::MediaKeySystemPermissionRequestProxy::frameID const):
(WebKit::MediaKeySystemPermissionRequestProxy::topLevelDocumentSecurityOrigin):
(WebKit::MediaKeySystemPermissionRequestProxy::topLevelDocumentSecurityOrigin const):
(WebKit::MediaKeySystemPermissionRequestProxy::keySystem const):
* UIProcess/SpeechRecognitionRemoteRealtimeMediaSource.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::requestMediaKeySystemPermissionForFrame):
(WebKit::WebPageProxy::mediaKeySystemPermissionRequestManager):
(WebKit::WebPageProxy::requestMediaKeySystemPermissionByDefaultAction):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebProcess/EncryptedMedia/MediaKeySystemPermissionRequestManager.cpp: Added.
(WebKit::MediaKeySystemPermissionRequestManager::MediaKeySystemPermissionRequestManager):
(WebKit::MediaKeySystemPermissionRequestManager::startMediaKeySystemRequest):
(WebKit::MediaKeySystemPermissionRequestManager::sendMediaKeySystemRequest):
(WebKit::MediaKeySystemPermissionRequestManager::cancelMediaKeySystemRequest):
(WebKit::MediaKeySystemPermissionRequestManager::mediaKeySystemWasGranted):
(WebKit::MediaKeySystemPermissionRequestManager::mediaKeySystemWasDenied):
* WebProcess/EncryptedMedia/MediaKeySystemPermissionRequestManager.h: Added.
* WebProcess/WebCoreSupport/WebMediaKeySystemClient.cpp: Added.
(WebKit::WebMediaKeySystemClient::WebMediaKeySystemClient):
(WebKit::WebMediaKeySystemClient::pageDestroyed):
(WebKit::WebMediaKeySystemClient::requestMediaKeySystem):
(WebKit::WebMediaKeySystemClient::cancelMediaKeySystemRequest):
* WebProcess/WebCoreSupport/WebMediaKeySystemClient.h: Added.
(WebKit::WebMediaKeySystemClient::~WebMediaKeySystemClient):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
(WebKit::WebPage::mediaKeySystemWasGranted):
(WebKit::WebPage::mediaKeySystemWasDenied):
* WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::mediaKeySystemPermissionRequestManager):
* WebProcess/WebPage/WebPage.messages.in:
2021-02-07 Fujii Hironori <Hironori.Fujii@sony.com>
[CoordinatedGraphics] The whole content is unnecessarily repainted by animations in non-AC mode pages
https://bugs.webkit.org/show_bug.cgi?id=221391
Reviewed by Carlos Garcia Campos.
When triggerRenderingUpdate was called back, it added the whole
view area into the dirty region to repaint in non-AC mode pages.
This caused a problem that the whole content was unnecessarily
repainted by animations in non-AC mode pages
If triggerRenderingUpdate is called back,
DrawingAreaCoordinatedGraphics::display should be called to ensure
WebPage::updateRendering() called even if m_dirtyRegion is empty.
* WebProcess/WebPage/CoordinatedGraphics/DrawingAreaCoordinatedGraphics.cpp:
(WebKit::DrawingAreaCoordinatedGraphics::triggerRenderingUpdate):
Call scheduleDisplay() instead of setNeedsDisplay().
(WebKit::DrawingAreaCoordinatedGraphics::scheduleDisplay): Don't
return early even if m_dirtyRegion is empty.
(WebKit::DrawingAreaCoordinatedGraphics::display()): Ditto.
(WebKit::DrawingAreaCoordinatedGraphics::display(UpdateInfo&)):
Return early if m_dirtyRegion is empty.
2021-02-07 Sam Weinig <weinig@apple.com>
Differentiate between color spaces used to tag colors and color spaces used to tag destinations
https://bugs.webkit.org/show_bug.cgi?id=221520
Reviewed by Antti Koivisto.
* Scripts/webkit/messages.py:
* Shared/WebCoreArgumentCoders.h:
Coder implementation of EnumTraits for color space types is now
in WebCore.
* GPUProcess/graphics/RemoteImageBuffer.h:
(WebKit::RemoteImageBuffer::create):
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::createImageBuffer):
* GPUProcess/graphics/RemoteRenderingBackend.h:
* GPUProcess/graphics/RemoteRenderingBackend.messages.in:
* Shared/RemoteLayerTree/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::swapToValidFrontBuffer):
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
(WebKit::RemoteImageBufferProxy::create):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::createImageBuffer):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::createImageBuffer const):
* WebProcess/WebCoreSupport/WebChromeClient.h:
Update to use DestinationColorSpace rather than ColorSpace.
2021-02-07 Sam Weinig <weinig@apple.com>
Remove more special cases from color conversion
https://bugs.webkit.org/show_bug.cgi?id=221519
Reviewed by Antti Koivisto.
* UIProcess/API/wpe/WebKitColor.cpp:
(webkitColorToWebCoreColor):
Replace convertTo<> component type conversion to general convertColor<>.
2021-02-06 Wenson Hsieh <wenson_hsieh@apple.com>
[macOS] Add support for accessibility image extraction
https://bugs.webkit.org/show_bug.cgi?id=221262
<rdar://problem/70744978>
Reviewed by Devin Rousso.
Add WebKitAdditions support for image extraction.
* UIProcess/Cocoa/WebViewImpl.h:
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::requestImageExtraction): Deleted.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::requestImageExtraction):
2021-02-06 Alex Christensen <achristensen@webkit.org>
Non-special URLs should have an opaque origin
https://bugs.webkit.org/show_bug.cgi?id=162254
Reviewed by Geoff Garen.
I needed to change API::SecurityOrigin from wrapping a WebCore::SecurityOrigin to wrapping a WebCore::SecurityOriginData,
otherwise it would often be null and we want it to be not null in the UI process.
* Shared/API/APISecurityOrigin.h:
(API::SecurityOrigin::createFromString):
(API::SecurityOrigin::create):
(API::SecurityOrigin::securityOrigin const):
(API::SecurityOrigin::SecurityOrigin):
* Shared/API/c/WKSecurityOriginRef.cpp:
(WKSecurityOriginCopyDatabaseIdentifier):
(WKSecurityOriginCopyProtocol):
(WKSecurityOriginCopyHost):
(WKSecurityOriginGetPort):
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreRemoveFetchCacheForOrigin):
* UIProcess/API/Cocoa/WKSecurityOrigin.mm:
(-[WKSecurityOrigin protocol]):
(-[WKSecurityOrigin host]):
(-[WKSecurityOrigin port]):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::registerURLSchemeHandler):
2021-02-05 Sihui Liu <sihui_liu@appe.com>
REGRESSION(r272337): crash under WebCore::SpeechRecognizer::setInactive()
https://bugs.webkit.org/show_bug.cgi?id=221451
Reviewed by Youenn Fablet.
* UIProcess/SpeechRecognitionServer.cpp:
(WebKit::SpeechRecognitionServer::handleRequest):
2021-02-05 Darin Adler <darin@apple.com>
Add missing null checks to decoding functions involving RetainPtr
https://bugs.webkit.org/show_bug.cgi?id=221441
Reviewed by Anders Carlsson.
* Shared/mac/WebCoreArgumentCodersMac.mm:
(IPC::ArgumentCoder<WebCore::ResourceRequest>::decodePlatformData): Fail if the dictionary
is null: the encoder never encodes a null dictionary.
(IPC::ArgumentCoder<WebCore::CertificateInfo>::decode): Fail if the SecTrustRef is null:
the encoder never encodes a null SecTrustRef. Fail if the certificate chain array is null:
the encoder never encodes a null CFArrayRef.
(IPC::decodeNSError): Return a RetainPtr instead of using a bool return value and a
RetainPtr out argument. Fail if the user info dictionary is null: the encoder never encodes
a null user info dictionary.
(IPC::ArgumentCoder<WebCore::ResourceError>::decodePlatformData): Updated for the change
to the interface of decodeNSError.
(IPC::ArgumentCoder<WebCore::ContentFilterUnblockHandler>::decode): Fail if the CFDataRef
is null: the encoder never encodes a null CFDataRef.
2021-02-05 Brent Fulgham <bfulgham@apple.com>
REGRESSION (r271815): [macOS] /dev/dtracehelper is blocked on customer builds, even with SIP disabled
https://bugs.webkit.org/show_bug.cgi?id=221501
<rdar://73893768>
Reviewed by Per Arne Vollan.
Instead of requiring apple-internal to use /dev/dtracehelper, allow on systems where
SIP has been deactivated.
* WebProcess/com.apple.WebProcess.sb.in: Use CSR filter to gate access to /dev/dtracehelper
2021-02-05 Eric Carlson <eric.carlson@apple.com>
[Mac] Connect MediaSession with MediaRemote and NowPlaying
https://bugs.webkit.org/show_bug.cgi?id=221431
<rdar://problem/74000363>
Reviewed by Jer Noble.
Add a private preference so the new MediaSession API test can enable the feature.
* UIProcess/API/Cocoa/WKPreferences.mm:
(-[WKPreferences _mediaSessionEnabled]):
(-[WKPreferences _setMediaSessionEnabled:]):
* UIProcess/API/Cocoa/WKPreferencesPrivate.h:
2021-02-05 Youenn Fablet <youenn@apple.com>
Enable audio capture for speech recognition in GPUProcess
https://bugs.webkit.org/show_bug.cgi?id=221457
Reviewed by Eric Carlson.
Allow to create remote sources without any constraint.
To do so, we serialize through IPC a MediaConstraints with isValid = false and treat it as no constraint in capture process.
Make sure to send sandbox extensions and authorizations for GPUProcess to capture in case of speech recognition audio capture request.
In case of GPUProcess audio capture, send the request to capture to WebProcess like done for iOS.
WebProcess is then responsible to get audio samples from GPUProcess and forward them to UIProcess.
A future refactoring should move speech recognition to GPUProcess.
* UIProcess/Cocoa/UserMediaCaptureManagerProxy.cpp:
(WebKit::UserMediaCaptureManagerProxy::createMediaSourceForCaptureDeviceWithConstraints):
* UIProcess/UserMediaPermissionRequestManagerProxy.cpp:
(WebKit::UserMediaPermissionRequestManagerProxy::grantRequest):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::createRealtimeMediaSourceForSpeechRecognition):
* WebProcess/Speech/SpeechRecognitionRealtimeMediaSourceManager.cpp:
(WebKit::SpeechRecognitionRealtimeMediaSourceManager::grantSandboxExtensions):
(WebKit::SpeechRecognitionRealtimeMediaSourceManager::createSource):
* WebProcess/cocoa/RemoteRealtimeMediaSource.cpp:
(WebKit::RemoteRealtimeMediaSource::create):
(WebKit::RemoteRealtimeMediaSource::RemoteRealtimeMediaSource):
(WebKit::RemoteRealtimeMediaSource::createRemoteMediaSource):
(WebKit::RemoteRealtimeMediaSource::~RemoteRealtimeMediaSource):
(WebKit::RemoteRealtimeMediaSource::cloneVideoSource):
(WebKit::RemoteRealtimeMediaSource::gpuProcessConnectionDidClose):
* WebProcess/cocoa/RemoteRealtimeMediaSource.h:
* WebProcess/cocoa/UserMediaCaptureManager.cpp:
(WebKit::UserMediaCaptureManager::AudioFactory::createAudioCaptureSource):
(WebKit::UserMediaCaptureManager::VideoFactory::createVideoCaptureSource):
(WebKit::UserMediaCaptureManager::DisplayFactory::createDisplayCaptureSource):
2021-02-05 Kate Cheney <katherine_cheney@apple.com>
ASSERTION FAILED: Completion handler should always be called under WebKit::VideoFullscreenManagerProxy::forEachSession
https://bugs.webkit.org/show_bug.cgi?id=221343
<rdar://problem/73939450>
Reviewed by Youenn Fablet.
If a client tries to exit fullscreen mode when there is no fullscreen
video present, we should call the completion handler immediately and
return.
Rename requestFullscreenModeWithCallback to requestCloseAllMediaPresentations
because the completion handler is only called when all media presentations
are closed. requestFullscreenModeWithCallback makes it seem like we
can use this function with other modes besides VideoFullscreenModeNone.
No new tests, this will fix a crashing iOS test.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView closeAllMediaPresentations:]):
* UIProcess/Cocoa/VideoFullscreenManagerProxy.h:
* UIProcess/Cocoa/VideoFullscreenManagerProxy.mm:
(WebKit::VideoFullscreenModelContext::requestCloseAllMediaPresentations):
(WebKit::VideoFullscreenManagerProxy::requestCloseAllMediaPresentations):
(WebKit::VideoFullscreenModelContext::requestFullscreenModeWithCallback): Deleted.
(WebKit::VideoFullscreenManagerProxy::requestFullscreenModeWithCallback): Deleted.
2021-02-05 Youenn Fablet <youenn@apple.com>
Make sure click attribution is processed in case of redirected kept alive loads
https://bugs.webkit.org/show_bug.cgi?id=221453
<rdar://problem/70896640>
Reviewed by John Wilander.
* NetworkProcess/NetworkResourceLoader.cpp:
(WebKit::NetworkResourceLoader::continueWillSendRedirectedRequest):
Move the click attribution processing before kept alive check.
2021-02-05 Per Arne Vollan <pvollan@apple.com>
Remove access to IOMobileFramebufferUserClient
https://bugs.webkit.org/show_bug.cgi?id=221408
<rdar://problem/70551457>
Reviewed by Brent Fulgham.
Remove access to the IOKit class IOMobileFramebufferUserClient in the WebContent process on macOS and iOS
when the GPU process is enabled.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-05 Kate Cheney <katherine_cheney@apple.com>
CrashTracer: com.apple.WebKit.Networking at WebKit: WebKit::ResourceLoadStatisticsDatabaseStore::setPrevalentResource
https://bugs.webkit.org/show_bug.cgi?id=221432
<rdar://problem/67069819>
Reviewed by John Wilander.
We are seeing crashes in ResourceLoadStatisticsDatabaseStore::setPrevalentResource
as a result of trying to use a nullopt domainID value. In theory this should
never be WTF::nullopt but is because of a failing SQLite query in
ResourceLoadStatisticsDatabaseStore::domainID which reports the error "not an error".
To fix this we should check the domain ID and return early with a
debug assert in setPrevalentResource() if it is WTF::nullopt to avoid
a crash. Additionally, we should add more information to the logging
statement, specifically the SQLite statement string, to try and debug further.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::domainID const):
(WebKit::ResourceLoadStatisticsDatabaseStore::setPrevalentResource):
2021-02-05 Sihui Liu <sihui_liu@appe.com>
SpeechRecognitionPermissionManager should not handle requests that are already cancelled
https://bugs.webkit.org/show_bug.cgi?id=221296
Reviewed by Youenn Fablet.
It is possible that client asks to stop/abort pending requests, which are waiting for permission check to be
done. In our current implementation, SpeechRecognitionPermissionManager would perform permission checks for
these cancelled requests, which is unnecessary and may cause extra prompts. We should let
SpeechRecognitionPermissionManager check if request is still valid before checking permissions.
No test is added as this behavior change is not observable without resetting the TCC permission between requests.
* UIProcess/SpeechRecognitionPermissionManager.cpp:
(WebKit::SpeechRecognitionPermissionManager::request):
(WebKit::SpeechRecognitionPermissionManager::startNextRequest):
(WebKit::SpeechRecognitionPermissionManager::startProcessingRequest):
(WebKit::SpeechRecognitionPermissionManager::continueProcessingRequest):
(WebKit::SpeechRecognitionPermissionManager::requestUserPermission):
* UIProcess/SpeechRecognitionPermissionManager.h:
* UIProcess/SpeechRecognitionPermissionRequest.h:
(WebKit::SpeechRecognitionPermissionRequest::create):
(WebKit::SpeechRecognitionPermissionRequest::~SpeechRecognitionPermissionRequest):
(WebKit::SpeechRecognitionPermissionRequest::request):
(WebKit::SpeechRecognitionPermissionRequest::SpeechRecognitionPermissionRequest):
(WebKit::SpeechRecognitionPermissionRequest::origin const): Deleted.
(WebKit::SpeechRecognitionPermissionRequest::lang const): Deleted.
(WebKit::SpeechRecognitionPermissionRequest::frameIdentifier const): Deleted.
* UIProcess/SpeechRecognitionServer.cpp:
(WebKit::SpeechRecognitionServer::requestPermissionForRequest):
(WebKit::SpeechRecognitionServer::sendUpdate):
* UIProcess/SpeechRecognitionServer.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::requestSpeechRecognitionPermission):
* UIProcess/WebPageProxy.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::createSpeechRecognitionServer):
2021-02-05 Chris Dumez <cdumez@apple.com>
[GPUProcess][iOS] Audio is lost after media playback recovers from the GPUProcess crash
https://bugs.webkit.org/show_bug.cgi?id=221305
<rdar://problem/73904864>
Reviewed by Eric Carlson.
Fix issue where we would sometimes lose the video after recovering from a GPUProcess
crash. The issue is that RemoteMediaPlayerProxy::setVideoInlineSizeFenced() may get
called *before* RemoteMediaPlayerProxy::mediaPlayerFirstVideoFrameAvailable(). As a
result, we would not have a root layer yet and we would fail to set the video
dimensions and they would remain at 0x0.
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
* GPUProcess/media/RemoteMediaPlayerProxy.h:
* GPUProcess/media/cocoa/RemoteMediaPlayerProxyCocoa.mm:
(WebKit::setVideoInlineSizeIfPossible):
(WebKit::RemoteMediaPlayerProxy::mediaPlayerFirstVideoFrameAvailable):
(WebKit::RemoteMediaPlayerProxy::setVideoInlineSizeFenced):
2021-02-05 Carlos Garcia Campos <cgarcia@igalia.com>
REGRESSION(r271879) [SOUP] webrtc/datachannel/gather-candidates-networkprocess-crash.html is crashing in debug mode
https://bugs.webkit.org/show_bug.cgi?id=221379
Reviewed by Youenn Fablet.
* NetworkProcess/soup/WebKitDirectoryInputStream.cpp:
(webkitDirectoryInputStreamRead): Fix bytes read calculation.
2021-02-05 Philippe Normand <pnormand@igalia.com>
[GTK] clang warnings
https://bugs.webkit.org/show_bug.cgi?id=221410
Reviewed by Adrian Perez de Castro.
* NetworkProcess/NetworkResourceLoader.cpp:
(WebKit::NetworkResourceLoader::resourceLoadInfo): Remove ifdef added in r272313, the
corresponding enum value is not ifdeffed either.
* UIProcess/SpeechRecognitionServer.cpp:
(WebKit::SpeechRecognitionServer::sendUpdate): Use portable format specifier for uint64_t.
* WebProcess/Speech/SpeechRecognitionRealtimeMediaSourceManager.cpp: Ifdef member variable
currently used only on Cocoa platforms.
2021-02-04 Adrian Perez de Castro <aperez@igalia.com>
Unreviewed. Update OptionsWPE.cmake and NEWS for the 2.31.1 release
* wpe/NEWS: Add release notes for 2.31.1.
2021-02-04 Ross Kirsling <ross.kirsling@sony.com>
REGRESSION (r272382): 2 webkit.messages_unittest.GeneratedFileContentsTest tests failing after Semaphore.h rename
https://bugs.webkit.org/show_bug.cgi?id=221433
Reviewed by Don Olmstead.
Looks like some changes were accidentally dropped prior to landing.
* Scripts/webkit/tests/MessageArgumentDescriptions.cpp:
* Scripts/webkit/tests/TestWithSemaphoreMessageReceiver.cpp:
2021-02-04 Julian Gonzalez <julian_a_gonzalez@apple.com>
Null dereference in DocumentLoader::commitData()
https://bugs.webkit.org/show_bug.cgi?id=221097
Reviewed by Alex Christensen.
As we add another return-early case to commitData(),
WebFrameLoaderClient::committedLoad() has to allow for
the absence of a Document.
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::committedLoad):
2021-02-04 Kate Cheney <katherine_cheney@apple.com>
Link libnetworkextension at compile time
https://bugs.webkit.org/show_bug.cgi?id=221372
<rdar://problem/73855844>
Reviewed by Brent Fulgham.
Network Extension is not available on tvOS.
* Configurations/WebKit.xcconfig:
2021-02-04 Megan Gardner <megan_gardner@apple.com>
Allow highlight delegate to be set on WKWebView
https://bugs.webkit.org/show_bug.cgi?id=221362
Reviewed by Tim Horton.
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
2021-02-04 Alex Christensen <achristensen@webkit.org>
Don't launch a network process when registering URL schemes
https://bugs.webkit.org/show_bug.cgi?id=221427
Reviewed by Chris Dumez.
When Safari on macOS is launching, it calls registerURLSchemeAsSecure and registerURLSchemeAsBypassingContentSecurityPolicy.
This causes the network process to launch before the app has finished launching. Since values are stored in the UI process
and sent in NetworkProcessCreationParameters whenever a network process starts, so it's not necessary to start the network
process when calling these two functions. This should decrease Safari's launch time.
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::registerURLSchemeAsSecure):
(WebKit::WebProcessPool::registerURLSchemeAsBypassingContentSecurityPolicy):
2021-02-04 Ross Kirsling <ross.kirsling@sony.com>
Don't use Semaphore.h as a filename
https://bugs.webkit.org/show_bug.cgi?id=221418
Reviewed by Don Olmstead.
IPC::Semaphore was recently moved to IPC/Semaphore.h; let's ensure that it's IPC/IPCSemaphore.h instead.
Semaphore.h is a filename that may exist in platform SDKs and thus can cause a WK build break based on ordering.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
* GPUProcess/graphics/RemoteRenderingBackend.h:
* GPUProcess/media/RemoteAudioDestinationManager.cpp:
* GPUProcess/media/RemoteAudioDestinationManager.h:
* Platform/IPC/IPCSemaphore.cpp: Renamed from Source/WebKit/Platform/IPC/Semaphore.cpp.
* Platform/IPC/IPCSemaphore.h: Renamed from Source/WebKit/Platform/IPC/Semaphore.h.
* Platform/IPC/darwin/IPCSemaphoreDarwin.cpp: Renamed from Source/WebKit/Platform/IPC/darwin/SemaphoreDarwin.cpp.
* Scripts/webkit/messages.py:
* Scripts/webkit/tests/MessageArgumentDescriptions.cpp:
* Scripts/webkit/tests/TestWithSemaphoreMessageReceiver.cpp:
* Sources.txt:
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
* WebProcess/GPU/media/RemoteAudioDestinationProxy.h:
2021-02-04 Youenn Fablet <youenn@apple.com>
Test landed flaky: [iOS] TestWebKitAPI.WebKit2.CrashGPUProcessWhileCapturing
https://bugs.webkit.org/show_bug.cgi?id=221331
<rdar://problem/73935129>
Reviewed by Eric Carlson.
For video sources tied to a remote source in GPUProcess, we are calling requestToEnd instead of end
since the same remote source may be shared with other cloned video sources.
In that case, the source in WebProcess should be considered as ended as soon as requestToEnd is called.
This is important as in case of GPUProcess, we are recreating sources that are not ended.
We were thus recreating a source which was actually ended in WebProcess.
This appears in particular on iOS where only one source can be active at a time.
A further refactoring should probably change the way we clone video sources given we are now supporting size and frame rate adaptations in RealtimeVideoSource.
Covered by API test no longer falkily failing.
* WebProcess/cocoa/RemoteRealtimeMediaSource.cpp:
(WebKit::RemoteRealtimeMediaSource::captureFailed):
(WebKit::RemoteRealtimeMediaSource::requestToEnd):
* WebProcess/cocoa/RemoteRealtimeMediaSource.h:
2021-02-04 Alex Christensen <achristensen@apple.com>
REGRESSION(r267763) Network process launches earlier than before
https://bugs.webkit.org/show_bug.cgi?id=221384
<rdar://problem/73507706>
Reviewed by Chris Dumez.
Before r267763 we could have an app launch, allocate a WKWebView, and set _resourceLoadStatisticsEnabled
all without having a network process launch.
When setting _resourceLoadStatisticsEnabled, we don't need to launch a network process if we haven't already
because when we do, WebsiteDataStore::parameters will send the correct value. I added a test that verifies this doesn't
launch the network process if it hasn't already been launched.
We also don't need to update the process assertion or the process suspension state if there is no network process.
sendNetworkProcessXPCEndpointToWebProcess is also called when opening Safari on iOS and would be the next thing to
unnecessarily launch the network process during app initialization. I manually verified that if we add an early return
if there is no network process yet, we do send the XPC endpoint to the web process when the network process does launch
through the call site in sendNetworkProcessXPCEndpointToAllWebProcesses.
I verified on my phone that this is the minimal change needed for Safari on iOS to not launch the network process before
application:didFinishLaunchingWithOptions: is called as it did before r267763.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::updateProcessSuppressionState):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::updateProcessAssertions):
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::WebsiteDataStore::sendNetworkProcessXPCEndpointToWebProcess):
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::setResourceLoadStatisticsEnabled):
* UIProcess/WebsiteData/WebsiteDataStore.h:
(WebKit::WebsiteDataStore::networkProcessIfExists):
2021-02-04 Aditya Keerthi <akeerthi@apple.com>
[macOS] Selecting a date on datetime-local inputs unexpectedly adds second and millisecond fields
https://bugs.webkit.org/show_bug.cgi?id=221350
<rdar://problem/73943517>
Reviewed by Devin Rousso.
* UIProcess/mac/WebDateTimePickerMac.mm:
(-[WKDateTimePicker updatePicker:]):
(-[WKDateTimePicker dateFormatStringForType:]):
Do not use the length of the value to determine whether or seconds and
milliseconds should be present, since the value can be empty.
Instead, use the new information in DateTimeChooserParameters, matching
the visual appearance of the input.
2021-02-04 Lauro Moura <lmoura@igalia.com>
REGRESSION(r267763): [GTK][WPE] Broken main thread assertion in MemoryPressureMonitor
https://bugs.webkit.org/show_bug.cgi?id=217363
Reviewed by Carlos Garcia Campos.
r267763 changed a number of NetworkProcessProxy methods to be
accessible only from the main thread, including allNetworkProcesses().
This function was called from the memory pressure monitor thread
directly to notify high memory usage.
This was causing the debug bots to abort prematurely with many crashes
when in reality just a memory pressure warning would be enough.
Covered by existing tests.
* UIProcess/linux/MemoryPressureMonitor.cpp:
(WebKit::MemoryPressureMonitor::start): Dispatch the memory pressure
notification in the main thread.
2021-02-03 Timothy Horton <timothy_horton@apple.com>
ASSERTION FAILED: !size.isEmpty() under platformInitializeWebProcess
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
Follow-up to r272335; don't call IOSurface::setMaximumSize if the UI process
did not send us a size. Caused all Web Content processes to assert in debug on launch.
2021-02-03 Jer Noble <jer.noble@apple.com>
[Cocoa] WebM audio goes out-of-sync or stops playing after a seek
https://bugs.webkit.org/show_bug.cgi?id=221354
<rdar://73471690>
Reviewed by Eric Carlson.
The MTPluginFormatReader will attempt to determine whether an audio sample needs priming frames
by querying for, among other things, its "syncInfo". Previously, we had just presumed all samples
without a sample attachment dictionary were sync samples, and reported them as such from
MediaSampleCursor, but this causes the format reader plugin to walk backwards all the way to the
beginning of the audio track looking for priming samples. Instead, since the sample in question
has no sync info (it was synthesized by MediaSample), return kCMBaseObjectError_ValueNotAvailable
in this case, which causes the plugin format reader to assume all samples are both sync and do
not require explicit priming frames.
* Shared/mac/MediaFormatReader/MediaSampleCursor.cpp:
(WebKit::MediaSampleCursor::getSampleTiming const):
(WebKit::MediaSampleCursor::getSyncInfo const):
2021-02-03 Andy Estes <aestes@apple.com>
[macOS] Shrink MediaSampleByteRange by 8 bytes
https://bugs.webkit.org/show_bug.cgi?id=221363
Reviewed by Eric Carlson.
Reordered member variables to shrink MediaSampleByteRange from 160 bytes to 152 bytes.
* Shared/mac/MediaFormatReader/MediaSampleByteRange.cpp:
(WebKit::MediaSampleByteRange::MediaSampleByteRange):
* Shared/mac/MediaFormatReader/MediaSampleByteRange.h:
2021-02-03 Kate Cheney <katherine_cheney@apple.com>
PCM: earliestTimeToSend should be treated as an independent time value, not relative to timeOfAdClick
https://bugs.webkit.org/show_bug.cgi?id=221303
<rdar://problem/73902668>
Reviewed by John Wilander.
r270136 assumed that PCM's earliestTimeToSend attribute was the
difference between timeOfAdClick and when the attribution should be
sent. As a result, it began each new browser session by adjusting
this value based on the amount of time the session was closed.
In reality, earliestTimeToSend is a raw secondsSinceEpoch() value, so we don't
need any adjustment after a new session.
No new tests, this is covered by
http/tests/privateClickMeasurement/expired-attribution-report-gets-sent-on-session-start.html
now that the corresponding test SPI has been updated.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::createTableQueries):
(WebKit::ResourceLoadStatisticsDatabaseStore::ResourceLoadStatisticsDatabaseStore):
(WebKit::ResourceLoadStatisticsDatabaseStore::createSchema):
(WebKit::ResourceLoadStatisticsDatabaseStore::destroyStatements):
(WebKit::ResourceLoadStatisticsDatabaseStore::markAttributedPrivateClickMeasurementsAsExpiredForTesting):
Update the testing SPI to insert an expired time of now minus one
hour as the earliestTimeToSend.
(WebKit::ResourceLoadStatisticsDatabaseStore::updateTimerLastFired): Deleted.
(WebKit::ResourceLoadStatisticsDatabaseStore::timerLastFired): Deleted.
(WebKit::ResourceLoadStatisticsDatabaseStore::updatePrivateClickMeasurementAttributionTimes): Deleted.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.h:
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::updateTimerLastFired): Deleted.
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.h:
* NetworkProcess/PrivateClickMeasurementManager.cpp:
(WebKit::PrivateClickMeasurementManager::firePendingAttributionRequests):
(WebKit::PrivateClickMeasurementManager::updateTimerLastFired): Deleted.
* NetworkProcess/PrivateClickMeasurementManager.h:
2021-02-03 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Allow one user gesture free prompt for each navigation
https://bugs.webkit.org/show_bug.cgi?id=220897
<rdar://problem/73541303>
Reviewed by Brent Fulgham.
This is a quirk that only allows certain websites.
Covered by existing tests.
* UIProcess/WebAuthentication/AuthenticatorManager.cpp:
(WebKit::AuthenticatorManager::filterTransports const):
(WebKit::AuthenticatorManager::runPanel):
(WebKit::AuthenticatorManager::runPresenter):
Cancel the whole WebAuthn operation if no user gesture is indicated.
* WebProcess/WebAuthentication/WebAuthenticatorCoordinator.cpp:
(WebKit::WebAuthenticatorCoordinator::makeCredential):
(WebKit::WebAuthenticatorCoordinator::getAssertion):
* WebProcess/WebAuthentication/WebAuthenticatorCoordinator.h:
Set a boolean to indicate a free coupon for each navigation.
2021-02-03 Michael Catanzaro <mcatanzaro@gnome.org>
[GTK] webkit_settings_set_enable_plugins deprecation warning breaks unit tests
https://bugs.webkit.org/show_bug.cgi?id=220620
Reviewed by Carlos Garcia Campos.
Don't print a warning when calling webkit_settings_set_enable_plugins() to disable plugins.
We'll still print a warning if attempting to enable plugins, or if checking whether plugins
are enabled.
* UIProcess/API/glib/WebKitSettings.cpp:
(webkit_settings_get_enable_plugins):
(webkit_settings_set_enable_plugins):
2021-02-03 Sihui Liu <sihui_liu@appe.com>
Create a SpeechRecognizer for each SpeechRecognitionRequest
https://bugs.webkit.org/show_bug.cgi?id=219699
<rdar://problem/72392097>
Reviewed by Youenn Fablet.
We used the same SpeechRecognizer for multiple requests, so we had to reset SpeechRecognizer state between
requests, which makes things complicated and hard to debug (like nested reset calls). Now we have one dedicated
SpeechRecognizer for each request.
* UIProcess/SpeechRecognitionServer.cpp:
(WebKit::SpeechRecognitionServer::requestPermissionForRequest):
(WebKit::SpeechRecognitionServer::handleRequest):
(WebKit::SpeechRecognitionServer::stop):
(WebKit::SpeechRecognitionServer::abort):
(WebKit::SpeechRecognitionServer::invalidate):
* UIProcess/SpeechRecognitionServer.h:
2021-02-03 Tim Horton <timothy_horton@apple.com>
REGRESSION (r270392): Launch time regression under IOSurface::maximumSize()
https://bugs.webkit.org/show_bug.cgi?id=221346
<rdar://problem/73507706>
Reviewed by Sam Weinig.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
Scope r270392 to only the case where it is needed, when DOM rendering in the GPU process
is enabled and the Web Content process does not have access to IOSurface,
because it causes a small launch time regression.
2021-02-03 Aditya Keerthi <akeerthi@apple.com>
[iOS][FCR] Add new picker for select elements
https://bugs.webkit.org/show_bug.cgi?id=221153
<rdar://problem/73770389>
Reviewed by Tim Horton.
Tapping on a select element should display an context menu that allows
users to choose one of the specified options. Rather than presenting a
UIPickerView, tapping on select elements now create UIContextMenuInteractions,
similar to date and file inputs.
Test: fast/forms/ios/form-control-refresh/select/choose-select-option.html
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _shouldShowAutomaticKeyboardUIIgnoringInputMode]):
The new picker does not bring up the keyboard view on all devices.
(-[WKContentView _elementTypeRequiresAccessoryView:]):
Changed from a static method to an instance method, as the returned
value depends on a flag which is only accessible through the instance.
(-[WKContentView requiresAccessoryView]):
(-[WKContentView _formControlRefreshEnabled]):
(-[WKContentView _shouldShowKeyboardForElement:]):
(-[WKContentView _elementDidFocus:userIsInteracting:blurPreviousNode:activityStateChanges:userObject:]):
(-[WKContentView _removeContextMenuViewIfPossible]):
Do not remove the context menu if an select element is actively being
interacted with.
(-[WKContentView selectControl]):
* UIProcess/ios/forms/WKFormSelectControl.mm:
(-[WKFormSelectControl initWithView:]):
* UIProcess/ios/forms/WKFormSelectPicker.h:
* UIProcess/ios/forms/WKFormSelectPicker.mm:
(-[WKSelectPicker initWithView:]):
(-[WKSelectPicker controlView]):
(-[WKSelectPicker controlBeginEditing]):
Ensure the position information is up-to-date prior to presenting the
context menu.
(-[WKSelectPicker controlEndEditing]):
(-[WKSelectPicker dealloc]):
(-[WKSelectPicker didSelectOptionIndex:]):
(-[WKSelectPicker createMenu]):
Build the menu using UIActions and UIMenus. Since optgroup elements
cannot be nested, only the root UIMenu can contain UIMenus. Submenus
can only contain UIActions.
(-[WKSelectPicker actionForOptionItem:withIndex:]):
(-[WKSelectPicker contextMenuInteraction:previewForHighlightingMenuWithConfiguration:]):
(-[WKSelectPicker _contextMenuInteraction:styleForMenuWithConfiguration:]):
(-[WKSelectPicker contextMenuInteraction:configurationForMenuAtLocation:]):
(-[WKSelectPicker contextMenuInteraction:willDisplayMenuForConfiguration:animator:]):
(-[WKSelectPicker contextMenuInteraction:willEndForConfiguration:animator:]):
(-[WKSelectPicker removeContextMenuInteraction]):
(-[WKSelectPicker ensureContextMenuInteraction]):
(-[WKSelectPicker showSelectPicker]):
(-[WKSelectPicker selectRow:inComponent:extendingSelection:]):
Implement method for testing select pickers.
2021-02-03 Per Arne Vollan <pvollan@apple.com>
[macOS] Remove access to graphics related user clients
https://bugs.webkit.org/show_bug.cgi?id=221327
<rdar://problem/70496539>
Reviewed by Brent Fulgham.
Remove access to graphics related IOKit classes in the WebContent process on macOS when the GPU process is enabled.
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-03 Per Arne Vollan <pvollan@apple.com>
[macOS] Remove access to com.apple.cvmsServ
https://bugs.webkit.org/show_bug.cgi?id=221336
<rdar://problem/70495789>
Reviewed by Brent Fulgham.
Remove access to com.apple.cvmsServ in the WebContent process when the GPU Process is enabled.
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-03 Per Arne Vollan <pvollan@apple.com>
[macOS] Remove access to RootDomainUserClient
https://bugs.webkit.org/show_bug.cgi?id=221326
<rdar://problem/73928098>
Reviewed by Brent Fulgham.
After r272241, access to the IOKit class RootDomainUserClient can be removed from the WebContent process.
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-03 Kate Cheney <katherine_cheney@apple.com>
Add internal additions to NetworkSessionCocoa
https://bugs.webkit.org/show_bug.cgi?id=221060
<rdar://problem/68431471>
Reviewed by Brent Fulgham.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(-[WKNetworkSessionDelegate URLSession:task:willPerformHTTPRedirection:newRequest:completionHandler:]):
2021-02-03 Devin Rousso <drousso@apple.com>
[iOS] delay `+[BKSMousePointerService sharedInstance]` so that it doesn't impact launch perf
https://bugs.webkit.org/show_bug.cgi?id=221298
Reviewed by Geoffrey Garen.
* UIProcess/ios/WKMouseDeviceObserver.h:
* UIProcess/ios/WKMouseDeviceObserver.mm:
(-[WKMouseDeviceObserver init]): Added.
(-[WKMouseDeviceObserver startWithCompletionHandler:]): Added.
(-[WKMouseDeviceObserver start]):
(-[WKMouseDeviceObserver stop]):
(-[WKMouseDeviceObserver stopWithCompletionHandler:]): Added.
2021-02-03 Youenn Fablet <youenn@apple.com>
Make sure GPUProcess MediaRecorder handles correctly muted tracks
https://bugs.webkit.org/show_bug.cgi?id=221325
Reviewed by Eric Carlson.
In case of muted audio, zero the audio buffer before writing it in the ring buffer.
In case of muted video, send a black frame instead of the real frame.
For that purpose, create a black frame and convert it to IOSurface.
Apply this to any non IOSurface based video sample.
Fix unified build.
* GPUProcess/webrtc/RemoteMediaRecorder.h:
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
(WebKit::MediaRecorderPrivate::videoSampleAvailable):
(WebKit::MediaRecorderPrivate::audioSamplesAvailable):
* WebProcess/GPU/webrtc/MediaRecorderPrivate.h:
* WebProcess/GPU/webrtc/MediaRecorderPrivate.mm: Added.
* WebProcess/Plugins/PDF/PDFPlugin.h:
* WebProcess/Plugins/PDF/PDFPluginTextAnnotation.mm:
2021-02-03 Alex Christensen <achristensen@webkit.org>
Fix watch and tv build after r272228
https://bugs.webkit.org/show_bug.cgi?id=220683
UIEventAttribution is special in that the symbol is declared as unavailable on watchOS and tvOS.
To fix the build, we need to not use this symbol on those operating systems.
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/API/ios/WKWebViewIOS.mm:
2021-02-03 Antoine Quint <graouts@webkit.org>
Add a loader for <model> resources
https://bugs.webkit.org/show_bug.cgi?id=221255
Reviewed by Youenn Fablet.
* NetworkProcess/NetworkLoadChecker.cpp:
(WebKit::NetworkLoadChecker::isAllowedByContentSecurityPolicy):
* NetworkProcess/NetworkResourceLoader.cpp:
(WebKit::NetworkResourceLoader::resourceLoadInfo):
* WebProcess/Network/WebLoaderStrategy.cpp:
(WebKit::maximumBufferingTime):
2021-02-03 Kimmo Kinnunen <kkinnunen@apple.com>
MachSemaphore does not work well with IPC messages
https://bugs.webkit.org/show_bug.cgi?id=220919
<rdar://problem/73826848>
Reviewed by Sam Weinig.
Move WTF::MachSemaphore to WebKit IPC::Semaphore. Currently the
MachSemaphore is useful only in cross-process IPC, implemented by the
IPC::.
Adds a stub IPC::Semaphore to all non-COCOA platforms, so that
constructors do not need to ifdef'ed to be able to compile platforms
that do not yet support GPU process fully.
Implement move constructor and move assignment operator for Semaphore.
Makes it possible to create types that hold Semaphore but still are
decode()able from a IPC message. The decoded objects are held with
Optional, and this forces the subobjects to have move assignment
operator.
Add asserts that the moved-away object is not used. Add the asserts as
asserts on the call return values, so the values are also checked on
normal objects on normal operation.
Implement an encoder and decoder for IPC::Semaphore. Makes it possible
to send it via IPC as its own type, not as MachSendRight.
As an example, changes existing uses of MachSendRight for semaphores in
messages as Semaphores, described below.
No new tests, a refactor.
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::createRenderingBackend):
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/GPUConnectionToWebProcess.messages.in:
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::create):
(WebKit::RemoteRenderingBackend::RemoteRenderingBackend):
(WebKit::RemoteRenderingBackend::nextDestinationImageBufferAfterApplyingDisplayLists):
* GPUProcess/graphics/RemoteRenderingBackend.h:
* Shared/RemoteRenderingBackendCreationParameters.h: Removed.
Changes RemoteRenderingBackend to receive its parameters from the normal
message parameters instead of redundant
RemoteRenderingBackendCreationParameters. The IPC generator already
generates 1:1 identical structure, it is just the message
Messages::GPUConnectionToWebProcess::CreateRenderingBackend. There is no
need to double-wrap the simple list of parameters. The double-wrapping
is a problem: The message system parameters are of ownership "const
reference", which is what is expected.
RemoteRenderingBackendCreationParameters contents were of ownership
"value", which works as expected only for simple primitive types.
Essentially previously calling the IPC function meant transfering
objects to RemoteRenderingBackendCreationParameters. This is not how the
IPC system is meant to work. The intention of a IPC call is to, by
default, reference its arguments for the purpose of encoding.
* GPUProcess/media/RemoteAudioDestinationManager.cpp:
(WebKit::RemoteAudioDestination::create):
(WebKit::RemoteAudioDestination::RemoteAudioDestination):
(WebKit::RemoteAudioDestinationManager::createAudioDestination):
* GPUProcess/media/RemoteAudioDestinationManager.h:
* GPUProcess/media/RemoteAudioDestinationManager.messages.in:
Receive the semaphore from RemoteAudioDestinationProxy. Previously
it was created during RemoteAudioDestination creation and
be sent back by RemoteAudioDestinationManager.
* Platform/IPC/Semaphore.cpp: Renamed from Source/WTF/wtf/cocoa/MachSemaphore.cpp.
(IPC::Semaphore::encode const):
(IPC::Semaphore::decode):
* Platform/IPC/Semaphore.h: Renamed from Source/WTF/wtf/cocoa/MachSemaphore.h.
(IPC::Semaphore::operator bool const):
* Platform/IPC/cocoa/SemaphoreCocoa.cpp: Added.
(IPC::Semaphore::Semaphore):
(IPC::Semaphore::~Semaphore):
(IPC::Semaphore::operator=):
(IPC::Semaphore::signal):
(IPC::Semaphore::wait):
(IPC::Semaphore::waitFor):
(IPC::Semaphore::createSendRight const):
(IPC::Semaphore::encode const):
(IPC::Semaphore::decode):
(IPC::Semaphore::destroy):
* Scripts/webkit/messages.py:
* Scripts/webkit/messages_unittest.py:
* Scripts/webkit/tests/Makefile:
* Scripts/webkit/tests/MessageArgumentDescriptions.cpp:
(IPC::jsValueForArguments):
(IPC::messageArgumentDescriptions):
* Scripts/webkit/tests/MessageNames.cpp:
(IPC::description):
(IPC::receiverName):
(IPC::isValidMessageName):
* Scripts/webkit/tests/MessageNames.h:
* Scripts/webkit/tests/TestWithMachSemaphore.messages.in: Copied from Source/WebKit/GPUProcess/GPUConnectionToWebProcess.messages.in.
* Scripts/webkit/tests/TestWithMachSemaphoreMessageReceiver.cpp: Added.
* Scripts/webkit/tests/TestWithMachSemaphoreMessages.h: Added.
* Scripts/webkit/tests/TestWithMachSemaphoreMessagesReplies.h: Added.
* Sources.txt:
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::connectToGPUProcess):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::startRenderingThread):
(WebKit::RemoteAudioDestinationProxy::stopRenderingThread):
Change the semaphore usage from unique_ptr to the normal
object access. The semaphore is now created by the proxy,
so there is no unique_ptr.
(WebKit::RemoteAudioDestinationProxy::connectToGPUProcess):
(WebKit::RemoteAudioDestinationProxy::gpuProcessConnectionDidClose):
* WebProcess/GPU/media/RemoteAudioDestinationProxy.h:
Changes RemoteAudioDestinationProxy to send its own semaphore from
WebProcess to GPUProcess. Previously the semaphore would be sent back by
RemoteAudioDestinationManager. Previously the code structure is more
complicated than neccessary, as the semaphore holder was being
null-checked, reset and re-seated. The WebProcess gets already to choose
the semaphores GPUProcess uses, so this should not increase security
surface. The change chosen here side-steps the shortcoming of the IPC
system where unpacking objects from the Reply messages mandates that a
constructed version of that type exists. For previous MachSendRight
empty-constructed object is simple, but for MachSemaphore
empty-constructed object creates unnecessary complication that is not
required otherwise at the moment. The alternative to what was chosen
here would be to hold with Optional<MachSemaphore> and add a
empty-initializing constructor to hold the object where to unpack the
semaphore object from the Reply.
2021-02-03 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK][WPE] Reduce the use of SoupURI in preparation for libsoup3
https://bugs.webkit.org/show_bug.cgi?id=221251
Reviewed by Adrian Perez de Castro.
* UIProcess/API/glib/WebKitURISchemeRequest.cpp:
(webkit_uri_scheme_request_get_scheme):
(webkit_uri_scheme_request_get_path):
* UIProcess/API/glib/WebKitWebView.cpp:
(webkit_web_view_load_uri):
2021-02-02 Per Arne Vollan <pvollan@apple.com>
Register for power notifications in the UI process
https://bugs.webkit.org/show_bug.cgi?id=221258
Reviewed by Brent Fulgham.
Create a PowerObserver object when the Web process pool is registering notification observers. Notify all WebContent processes
When a power on notification is received.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::registerNotificationObservers):
(WebKit::WebProcessPool::unregisterNotificationObservers):
* UIProcess/WebProcessPool.h:
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeProcess):
(WebKit::WebProcess::systemWillPowerOn):
2021-02-02 Kate Cheney <katherine_cheney@apple.com>
PCM: Indicate if network load data is user identifiable
https://bugs.webkit.org/show_bug.cgi?id=221113
<rdar://problem/71348187>
Reviewed by John Wilander.
* NetworkProcess/NetworkDataTask.cpp:
(WebKit::NetworkDataTask::create):
* NetworkProcess/NetworkLoadParameters.h:
* NetworkProcess/NetworkResourceLoadParameters.cpp:
(WebKit::NetworkResourceLoadParameters::encode const):
(WebKit::NetworkResourceLoadParameters::decode):
* NetworkProcess/PrivateClickMeasurementManager.cpp:
(WebKit::PrivateClickMeasurementManager::fireConversionRequest):
* NetworkProcess/cocoa/NetworkDataTaskCocoa.h:
* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::NetworkDataTaskCocoa::NetworkDataTaskCocoa):
2021-02-02 Alex Christensen <achristensen@webkit.org>
Remove all use of _UIEventAttribution
https://bugs.webkit.org/show_bug.cgi?id=221289
<rdar://problem/73629041>
Reviewed by John Wilander.
It has been replaced by UIEventAttribution API.
* Platform/ios/UIKitSoftLink.h: Removed.
* Platform/ios/UIKitSoftLink.mm: Removed.
* Platform/spi/ios/UIKitSPI.h:
* SourcesCocoa.txt:
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _setEventAttribution:]): Deleted.
(-[WKWebView _eventAttribution]): Deleted.
* WebKit.xcodeproj/project.pbxproj:
2021-02-02 Wenson Hsieh <wenson_hsieh@apple.com>
Add scaffolding to request image extraction after a delay when hovering over images
https://bugs.webkit.org/show_bug.cgi?id=221226
Reviewed by Devin Rousso.
Work towards allowing clients to extract image data when hovering over rendered image elements, for new
accessibility-related features. See changes below for more details.
* UIProcess/Cocoa/WebViewImpl.h:
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::requestImageExtraction):
Leave this method stub unimplemented for the time being.
* UIProcess/PageClient.h:
(WebKit::PageClient::requestImageExtraction):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::requestImageExtraction):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/mac/PageClientImplMac.h:
* UIProcess/mac/PageClientImplMac.mm:
(WebKit::PageClientImpl::requestImageExtraction):
Add plumbing to allow the web process to asynchronously request image extraction data from the UI process.
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::requestImageExtraction):
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::didCommitLoad):
(WebKit::WebPage::requestImageExtraction):
* WebProcess/WebPage/WebPage.h:
Add a weak set of elements that we've attempted to extract, so that we won't end up extracting elements more
than once. Although this set this currently only cleared upon navigation, this should eventually be updated when
(for instance) an element's image source changes, such that we can re-extract the new image data if needed.
2021-02-02 Kate Cheney <katherine_cheney@apple.com>
REGRESSION(r271744): Broke Microsoft live login in internal builds
https://bugs.webkit.org/show_bug.cgi?id=221271
<rdar://problem/73848416>
Reviewed by Alex Christensen.
If a user goes straight to login.live.com without first visiting
microsoft.com, we crash when checking for storage access because
we assume the user has interacted with the page. Since this is a
quirked flow, we should not assume that, and should insert the
domain into the database.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::hasUserGrantedStorageAccessThroughPrompt):
(WebKit::ResourceLoadStatisticsDatabaseStore::hasUserGrantedStorageAccessThroughPrompt const): Deleted.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.h:
2021-02-02 Alex Christensen <achristensen@webkit.org>
Adopt UIEventAttribution instead of _UIEventAttribution
https://bugs.webkit.org/show_bug.cgi?id=220683
Reviewed by John Wilander.
UIKit introduced UIEventAttribution, so let's adopt it so they can remove their SPI.
See rdar://73629041
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _setUIEventAttribution:]):
(-[WKWebView _uiEventAttribution]):
2021-02-02 Rini Patel <rini_patel@apple.com>
Remove unused functions from GraphicsContextGL and ExtensionsGL
https://bugs.webkit.org/show_bug.cgi?id=218896
Reviewed by Dean Jackson.
* GPUProcess/graphics/RemoteGraphicsContextGL.messages.in:
* GPUProcess/graphics/RemoteGraphicsContextGLFunctionsGenerated.h:
(getGraphicsResetStatusARB):
(getTranslatedShaderSourceANGLE):
(blitFramebufferANGLE): Deleted.
(renderbufferStorageMultisampleANGLE): Deleted.
(insertEventMarkerEXT): Deleted.
(pushGroupMarkerEXT): Deleted.
(popGroupMarkerEXT): Deleted.
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.h:
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxyFunctionsGenerated.cpp:
(WebKit::RemoteGraphicsContextGLProxy::blitFramebufferANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::renderbufferStorageMultisampleANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::insertEventMarkerEXT): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::pushGroupMarkerEXT): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::popGroupMarkerEXT): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::readPixelsRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::texParameterfvRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::texParameterivRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getQueryivRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getQueryObjectuivRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getBufferPointervRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getVertexAttribIivRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getVertexAttribIuivRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getUniformuivRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getBufferParameteri64vRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::samplerParameterivRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::samplerParameterfvRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getSamplerParameterivRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getSamplerParameterfvRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getFramebufferParameterivRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getProgramInterfaceivRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getBooleani_vRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getMultisamplefvRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getTexLevelParameterivRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getTexLevelParameterfvRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getPointervRobustANGLERobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getnUniformfvRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getnUniformivRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getnUniformuivRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::texParameterIivRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::texParameterIuivRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getTexParameterIivRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getTexParameterIuivRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::samplerParameterIivRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::samplerParameterIuivRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getSamplerParameterIivRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getSamplerParameterIuivRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getQueryObjectivRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getQueryObjecti64vRobustANGLE): Deleted.
(WebKit::RemoteGraphicsContextGLProxy::getQueryObjectui64vRobustANGLE): Deleted.
2021-02-02 Youenn Fablet <youenn@apple.com>
Recover MediaStreamTrack audio rendering from GPUProcess crash
https://bugs.webkit.org/show_bug.cgi?id=221128
Reviewed by Eric Carlson.
Make AudioMediaStreamTrackRenderer observe GPUProcess crash.
Notify its owner in case of crash.
* WebProcess/GPU/webrtc/AudioMediaStreamTrackRenderer.cpp:
(WebKit::AudioMediaStreamTrackRenderer::AudioMediaStreamTrackRenderer):
(WebKit::AudioMediaStreamTrackRenderer::~AudioMediaStreamTrackRenderer):
(WebKit::AudioMediaStreamTrackRenderer::initialize):
(WebKit::AudioMediaStreamTrackRenderer::gpuProcessConnectionDidClose):
* WebProcess/GPU/webrtc/AudioMediaStreamTrackRenderer.h:
2021-02-02 Youenn Fablet <youenn@apple.com>
Make sure NetworkRTCProvider stays valid for all NetworkRTCSocketCocoa callbacks
https://bugs.webkit.org/show_bug.cgi?id=221249
<rdar://problem/72671547>
Reviewed by Eric Carlson.
Make the connection callback take a ref of the NetworkRTCProvider.
The ref cycle will break at NetworkConnectionToWebProcess destruction since it will call NetworkRTCProvider::close
which deletes all sockets owning the connection.
* NetworkProcess/webrtc/NetworkRTCSocketCocoa.mm:
(WebKit::NetworkRTCSocketCocoa::NetworkRTCSocketCocoa):
2021-02-02 Per Arne Vollan <pvollan@apple.com>
Unreviewed build fix after 272154.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
2021-02-02 Chris Dumez <cdumez@apple.com>
[macOS] Regression(r271897) Prewarmed WebProcesses no longer get a user-friendly process name in Activity Monitor
https://bugs.webkit.org/show_bug.cgi?id=221094
<rdar://73658122>
Reviewed by Per Arne Vollan.
Prewarmed WebProcesses currently do not have a network process connection because they do not have a sessionID yet
and initiating a connection to the network process currently requires a sessionID. r271897 made it so that
WebProcess::updateProcessName() early returns when there is no network process connection, which means that our
prewarmed WebProcesses no longer get the "Safari Web Content (prewarmed)" user-friendly name in activity monitor.
This patch relies on the fact that we are allowed to talk to launchservices when WebProcess::updateProcessName()
gets called during process initialization (The sandbox extension gets revoked during process initialization right
after calling updateProcessName). As a result, if updateProcessName() gets called during process initialization,
we now set the process name directly within the process instead of messaging the network process. This fixes the
issue with prewarmed WebProcesses.
This patch also gets rid of the early returns when the network connection is not present and calls
ensureNetworkProcessConnection() instead. We want to make sure to set the process name even if the process does
not have a network process connection yet (either because none has been initiated yet or because the network
process crashed).
Finally, this patch replaces some WTFLogAlways() for error cases with RELEASE_LOG_ERROR() so that the errors
stand out in the logs.
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::tryTakePrewarmedProcess):
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::setIsInProcessCache):
(WebKit::WebProcess::markIsNoLongerPrewarmed):
* WebProcess/WebProcess.h:
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
(WebKit::WebProcess::updateProcessName):
(WebKit::WebProcess::updateActivePages):
2021-02-02 Frédéric Wang <fwang@igalia.com>
[GTK] Initialize PrintInfo::PrintMode
https://bugs.webkit.org/show_bug.cgi?id=221256
Reviewed by Carlos Garcia Campos.
* Shared/PrintInfo.h:
2021-02-02 Youenn Fablet <youenn@apple.com>
Use WebRTC GPU process flag for MediaRecorder
https://bugs.webkit.org/show_bug.cgi?id=220967
Reviewed by Eric Carlson.
* WebProcess/GPU/webrtc/MediaRecorderProvider.cpp:
(WebKit::MediaRecorderProvider::createMediaRecorderPrivate):
2021-02-02 Youenn Fablet <youenn@apple.com>
Only update IPC connection of RemoteCaptureSampleManager for audio sources
https://bugs.webkit.org/show_bug.cgi?id=221200
Reviewed by Eric Carlson.
There is no need to update the connection for sources that are not added to the RemoteCaptureSampleManager.
* WebProcess/cocoa/UserMediaCaptureManager.cpp:
(WebKit::UserMediaCaptureManager::addSource):
(WebKit::UserMediaCaptureManager::didUpdateSourceConnection):
* WebProcess/cocoa/UserMediaCaptureManager.h:
2021-02-02 Antoine Quint <graouts@webkit.org>
WKASCAuthorizationPresenterDelegate needs to import RunLoop.h and BlockPtr.h
https://bugs.webkit.org/show_bug.cgi?id=221250
Reviewed by Antti Koivisto.
* UIProcess/WebAuthentication/Cocoa/WKASCAuthorizationPresenterDelegate.mm:
2021-02-02 Youenn Fablet <youenn@apple.com>
Recover WebRTC codecs in case of GPUProcess crash
https://bugs.webkit.org/show_bug.cgi?id=221138
Reviewed by Eric Carlson.
Make LibWebRTCCodecs listen to GPUProcess crash on creation of LibWebRTCCodec.
Set the connection also at the same time.
This allows to remove the hop to main thread when creating/releasing encoders and decoders.
Everything now happens in the dedicated queue, in particular updating of m_connection, encoder and decoder connections,
as well as encoders and decoders maps.
When using encoder/decoder connection outside of the dedicated queue, a lock is needed.
Update LibWebRTCCodecs use of gpu process when creating a peer connection so that this happens in main thread.
In case of process crash, we do some setup in main thread (listening to IPC messages for instance).
But most of the work is done in a task run in the dedicated queue.
In this task, we set the new IPC connection of all running decoders and encoders.
We also send IPC message to recreate encoders and decoders.
For that purpose, we store some states in encoders and decoders.
Covered by API test.
* WebProcess/GPU/webrtc/LibWebRTCCodecs.cpp:
(WebKit::formatNameFromWebRTCCodecType):
(WebKit::createRemoteDecoder):
(WebKit::LibWebRTCCodecs::LibWebRTCCodecs):
(WebKit::LibWebRTCCodecs::createDecoder):
(WebKit::LibWebRTCCodecs::releaseDecoder):
(WebKit::LibWebRTCCodecs::decodeFrame):
(WebKit::LibWebRTCCodecs::createEncoder):
(WebKit::LibWebRTCCodecs::releaseEncoder):
(WebKit::LibWebRTCCodecs::initializeEncoder):
(WebKit::LibWebRTCCodecs::encodeFrame):
(WebKit::LibWebRTCCodecs::setEncodeRates):
(WebKit::LibWebRTCCodecs::gpuProcessConnectionDidClose):
(WebKit::LibWebRTCCodecs::setConnection): Deleted.
* WebProcess/GPU/webrtc/LibWebRTCCodecs.h:
* WebProcess/Network/webrtc/LibWebRTCProvider.cpp:
(WebKit::LibWebRTCProvider::createPeerConnection):
(WebKit::LibWebRTCProvider::createDecoderFactory):
(WebKit::LibWebRTCProvider::createEncoderFactory):
2021-02-02 Youenn Fablet <youenn@apple.com>
imported/w3c/web-platform-tests/mediacapture-record/MediaRecorder-stop.html is flaky when GPUProcess is enabled
https://bugs.webkit.org/show_bug.cgi?id=221197
Reviewed by Eric Carlson.
* GPUProcess/webrtc/RemoteMediaRecorder.cpp:
(WebKit::RemoteMediaRecorder::stopRecording):
* GPUProcess/webrtc/RemoteMediaRecorder.h:
* GPUProcess/webrtc/RemoteMediaRecorder.messages.in:
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
(WebKit::MediaRecorderPrivate::stopRecording):
* WebProcess/GPU/webrtc/MediaRecorderPrivate.h:
2021-02-02 Carlos Garcia Campos <cgarcia@igalia.com>
[SOUP] Stop using SoupBuffer in preparation for libsoup3
https://bugs.webkit.org/show_bug.cgi?id=221206
Reviewed by Sergio Villar Senin.
Use GBytes instead of SoupBuffer.
* NetworkProcess/cache/NetworkCacheData.h:
(WebKit::NetworkCache::Data::bytes const):
* NetworkProcess/cache/NetworkCacheDataGLib.cpp: Renamed from Source/WebKit/NetworkProcess/cache/NetworkCacheDataSoup.cpp.
(WebKit::NetworkCache::Data::Data):
(WebKit::NetworkCache::Data::empty):
(WebKit::NetworkCache::Data::data const):
(WebKit::NetworkCache::Data::apply const):
(WebKit::NetworkCache::Data::subrange const):
(WebKit::NetworkCache::concatenate):
(WebKit::NetworkCache::Data::adoptMap):
(WebKit::NetworkCache::Data::tryCreateSharedMemory const):
* NetworkProcess/cache/NetworkCacheIOChannel.h:
* NetworkProcess/cache/NetworkCacheIOChannelGLib.cpp: Renamed from Source/WebKit/NetworkProcess/cache/NetworkCacheIOChannelSoup.cpp.
(WebKit::NetworkCache::runTaskInQueue):
(WebKit::NetworkCache::fillDataFromReadBuffer):
(WebKit::NetworkCache::inputStreamReadReadyCallback):
(WebKit::NetworkCache::IOChannel::read):
(WebKit::NetworkCache::IOChannel::readSyncInThread):
(WebKit::NetworkCache::outputStreamWriteReadyCallback):
(WebKit::NetworkCache::IOChannel::write):
* Shared/ShareableResource.cpp:
(WebKit::ShareableResource::wrapInSharedBuffer):
* SourcesGTK.txt:
* SourcesWPE.txt:
2021-02-02 Adrian Perez de Castro <aperez@igalia.com>
Non-unified build fixes, late January 2021 edition
https://bugs.webkit.org/show_bug.cgi?id=221044
Unreviewed non-unified build fixes.
* NetworkProcess/NetworkLoadScheduler.cpp: Add missing wtf/ListHashSet.h header inclusion.
(WebKit::NetworkLoadScheduler::HostContext::schedule): Add missing WebCore:: namespace in
usage of WebCore::ResourceLoadPriority enum.
* Platform/unix/SharedMemoryUnix.cpp: Add missing ArgumentCoders.h header inclusion.
* Shared/API/APIFrameHandle.cpp: Ditto.
* Shared/API/APIGeometry.cpp: Ditto.
* Shared/API/APIPageGroupHandle.cpp: Ditto.
* Shared/API/APIPageHandle.cpp: Ditto.
* Shared/LayerTreeContext.cpp: Ditto.
2021-02-01 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Optimize _WKWebAuthenticationPanelDelegate for native API
https://bugs.webkit.org/show_bug.cgi?id=221125
<rdar://problem/71509141>
Reviewed by Brent Fulgham.
This patch mainly does the following few things:
1. It connects the new SPI with the existing mock testing infrastructure such that we can mock test it.
2. It adds a new mode in AuthenticatorManager to connect it to the new _WKWebAuthenticationPanelDelegate object.
3. It adss a new SPI to _WKWebAuthenticationPanelDelegate: -[panel:requestLAContextForUserVerificationWithCompletionHandler:],
and adds a new method within _WKWebAuthenticationAssertionResponse to accept an LAContext from the caller to unlock
the keychain.
4. It enables -[_WKWebAuthenticationPanel cancel] to use the new cancel method.
Covered by API tests.
* UIProcess/API/APIWebAuthenticationAssertionResponse.h:
Accepts a LAContext object.
* UIProcess/API/APIWebAuthenticationPanel.cpp:
(API::WebAuthenticationPanel::WebAuthenticationPanel):
(API::WebAuthenticationPanel::handleRequest):
(API::WebAuthenticationPanel::cancel const):
(API::WebAuthenticationPanel::setMockConfiguration):
* UIProcess/API/APIWebAuthenticationPanel.h:
* UIProcess/API/APIWebAuthenticationPanelClient.h:
(API::WebAuthenticationPanelClient::requestLAContextForUserVerification const):
Adds new SPI and mocking.
* UIProcess/API/Cocoa/_WKWebAuthenticationAssertionResponse.h:
* UIProcess/API/Cocoa/_WKWebAuthenticationAssertionResponse.mm:
(-[_WKWebAuthenticationAssertionResponse setLAContext:]):
Accepts a LAContext object.
* UIProcess/API/Cocoa/_WKWebAuthenticationPanel.h:
* UIProcess/API/Cocoa/_WKWebAuthenticationPanel.mm:
(-[_WKWebAuthenticationPanel setMockConfiguration:]):
* UIProcess/API/Cocoa/_WKWebAuthenticationPanelForTesting.h:
Adds new SPI and mocking.
* UIProcess/WebAuthentication/AuthenticatorManager.cpp:
(WebKit::AuthenticatorManager::handleRequest):
(WebKit::AuthenticatorManager::enableModernWebAuthentication):
(WebKit::AuthenticatorManager::enableNativeSupport):
(WebKit::AuthenticatorManager::authenticatorAdded):
(WebKit::AuthenticatorManager::requestLAContextForUserVerification):
(WebKit::AuthenticatorManager::runPresenter):
(WebKit::AuthenticatorManager::dispatchPanelClientCall const):
* UIProcess/WebAuthentication/AuthenticatorManager.h:
Adds a new mode to power the native API.
* UIProcess/WebAuthentication/Cocoa/WebAuthenticationPanelClient.h:
* UIProcess/WebAuthentication/Cocoa/WebAuthenticationPanelClient.mm:
(WebKit::WebAuthenticationPanelClient::WebAuthenticationPanelClient):
(WebKit::WebAuthenticationPanelClient::requestLAContextForUserVerification const):
Accepts a LAContext object.
* UIProcess/WebAuthentication/Mock/MockLocalConnection.h:
* UIProcess/WebAuthentication/Mock/MockLocalConnection.mm:
(WebKit::MockLocalConnection::verifyUser):
Adds new mocking functions.
* UIProcess/WebAuthentication/WebAuthenticationRequestData.h:
Adds the new weak panel.
2021-02-01 Michael Catanzaro <mcatanzaro@gnome.org>
[WPE][GTK] flatpak-spawn subsandbox clears environment variables, breaks EphyPermissionsManager
https://bugs.webkit.org/show_bug.cgi?id=219325
Reviewed by Adrian Perez de Castro.
Manually forward all environment variables from the web process to the UI process.
EphyPermissionsManager is breaking because XDG directories are not set in the web process,
because flatpak-spawn launches subprocesses with the environment of the host system rather
than the environment of the parent process. But also, we really need to forward the entire
environment, because otherwise basic expected features like GST_DEBUG, G_MESSAGES_DEBUG, and
such will be broken.
* UIProcess/Launcher/glib/FlatpakLauncher.cpp:
(WebKit::flatpakSpawn):
2021-02-01 Antoine Quint <graouts@webkit.org>
Allow support for CAAnimationGroup
https://bugs.webkit.org/show_bug.cgi?id=221209
Reviewed by Dean Jackson, Sam Weinig and Simon Fraser.
In order to fix bug 219894, we need to use Core Animation's CAAnimationGroup and its
setAnimations() method. Add support for the new "Group" PlatformCAAnimation type for
the remote animations.
* WebProcess/WebPage/RemoteLayerTree/PlatformCAAnimationRemote.h:
* WebProcess/WebPage/RemoteLayerTree/PlatformCAAnimationRemote.mm:
(WebKit::PlatformCAAnimationRemote::Properties::encode const):
(WebKit::PlatformCAAnimationRemote::Properties::decode):
(WebKit::PlatformCAAnimationRemote::copy const):
(WebKit::PlatformCAAnimationRemote::setAnimations):
(WebKit::PlatformCAAnimationRemote::copyAnimationsFrom):
(WebKit::createAnimation):
(WebKit::addAnimationToLayer):
(WebKit::operator<<):
2021-02-01 Wenson Hsieh <wenson_hsieh@apple.com>
Add a cross-platform helper method to generate shareable bitmaps from rendered images
https://bugs.webkit.org/show_bug.cgi?id=221217
Reviewed by Tim Horton.
Pull what is currently iOS-specific logic in `imagePositionInformation` out into a private, cross-platform
helper method on `WebPage`. No change in behavior.
* WebProcess/WebPage/WebPage.cpp:
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::imagePositionInformation):
2021-02-01 Stephan Szabo <stephan.szabo@sony.com>
[PlayStation] Build fix after r272058
https://bugs.webkit.org/show_bug.cgi?id=221141
Unreviewed build fix.
* Platform/unix/SharedMemoryUnix.cpp: Include ArgumentCoders.h on PlayStation
2021-02-01 Sihui Liu <sihui_liu@appe.com>
Use user media permission prompt for speech recognition
https://bugs.webkit.org/show_bug.cgi?id=221082
rdar://problem/73372499
Reviewed by Youenn Fablet.
Make SpeechRecognitionPermissionManager ask UserMediaPermissionRequestManagerProxy for user permission on
microphone.
* UIProcess/SpeechRecognitionPermissionManager.cpp:
(WebKit::SpeechRecognitionPermissionManager::request):
(WebKit::SpeechRecognitionPermissionManager::startProcessingRequest):
(WebKit::SpeechRecognitionPermissionManager::requestUserPermission):
* UIProcess/SpeechRecognitionPermissionManager.h:
* UIProcess/SpeechRecognitionPermissionRequest.h:
(WebKit::SpeechRecognitionPermissionRequest::create):
(WebKit::SpeechRecognitionPermissionRequest::frameIdentifier const):
(WebKit::SpeechRecognitionPermissionRequest::SpeechRecognitionPermissionRequest):
* UIProcess/SpeechRecognitionServer.cpp:
(WebKit::SpeechRecognitionServer::start):
(WebKit::SpeechRecognitionServer::requestPermissionForRequest):
* UIProcess/SpeechRecognitionServer.h:
* UIProcess/SpeechRecognitionServer.messages.in:
* UIProcess/UserMediaPermissionRequestManagerProxy.cpp:
(WebKit::UserMediaPermissionRequestManagerProxy::denyRequest):
(WebKit::UserMediaPermissionRequestManagerProxy::grantRequest):
(WebKit::UserMediaPermissionRequestManagerProxy::checkUserMediaPermissionForSpeechRecognition):
* UIProcess/UserMediaPermissionRequestManagerProxy.h:
* UIProcess/UserMediaPermissionRequestProxy.cpp:
(WebKit::UserMediaPermissionRequestProxy::UserMediaPermissionRequestProxy):
* UIProcess/UserMediaPermissionRequestProxy.h:
(WebKit::UserMediaPermissionRequestProxy::create):
(WebKit::UserMediaPermissionRequestProxy::decisionCompletionHandler):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::requestSpeechRecognitionPermission):
(WebKit::WebPageProxy::requestUserMediaPermissionForSpeechRecognition):
* UIProcess/WebPageProxy.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::createSpeechRecognitionServer):
* WebProcess/WebCoreSupport/WebSpeechRecognitionConnection.cpp:
(WebKit::WebSpeechRecognitionConnection::start):
* WebProcess/WebCoreSupport/WebSpeechRecognitionConnection.h:
2021-02-01 Wenson Hsieh <wenson_hsieh@apple.com>
Add some new data structures to represent image extraction results
https://bugs.webkit.org/show_bug.cgi?id=221214
Reviewed by Sam Weinig.
See WebCore ChangeLog for more details.
* Shared/WebCoreArgumentCoders.cpp:
* Shared/WebCoreArgumentCoders.h:
2021-02-01 Per Arne <pvollan@apple.com>
[macOS] Observe color preference changes in the UI process
https://bugs.webkit.org/show_bug.cgi?id=221096
<rdar://problem/73721275>
Reviewed by Brent Fulgham.
As a step towards blocking the distributed notifications daemon in the WebContent process, color preference changes should be observed in the UI process.
The UI process should notify the WebContent process about color preference changes. In order to be able to notify the AppKit observer in the WebContent
process about changes with a local notification, a specific key/value needs to added to the Info.plist of the WebContent process.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::colorPreferencesDidChangeCallback):
(WebKit::WebProcessPool::registerNotificationObservers):
(WebKit::WebProcessPool::unregisterNotificationObservers):
* UIProcess/WebProcessPool.h:
* WebProcess/EntryPoint/Cocoa/XPCService/WebContentService/Info-OSX.plist:
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::colorPreferencesDidChange):
2021-02-01 Per Arne <pvollan@apple.com>
[macOS] Remove write access to some IOKit properties
https://bugs.webkit.org/show_bug.cgi?id=221137
<rdar://problem/73473658>
Reviewed by Brent Fulgham.
Remove write access to some IOKit properties in the WebContent process' sandbox.
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-01 Per Arne <pvollan@apple.com>
[macOS] Add reporting to mach-register sandbox rule
https://bugs.webkit.org/show_bug.cgi?id=221139
<rdar://problem/70350150>
Reviewed by Brent Fulgham.
Add reporting to mach-register rule in the WebContent process' sandbox.
* WebProcess/com.apple.WebProcess.sb.in:
2021-02-01 Carlos Garcia Campos <cgarcia@igalia.com>
[SOUP] Add support for data URLs in the network process
https://bugs.webkit.org/show_bug.cgi?id=221127
Reviewed by Adrian Perez de Castro.
In r271879 I removed the support for data URLs in the network process for soup, assuming data URLs were always
loaded from the web process, but that's not the case for downloads. I think we should not go to the network
process for downloads either, but for now it's easier to fix this regression by supporting data URLs in the
network process again. Fixes http/tests/security/anchor-download-allow-data.html.
* NetworkProcess/soup/NetworkDataTaskSoup.cpp:
(WebKit::NetworkDataTaskSoup::createRequest): Return early if it's a data URL.
(WebKit::NetworkDataTaskSoup::clearRequest): Clear the pending data URL result.
(WebKit::NetworkDataTaskSoup::resume): Start a data URL decode task if it's a new data URL request, or complete
the pending one.
(WebKit::NetworkDataTaskSoup::didReadDataURL): Complete the data URL request.
* NetworkProcess/soup/NetworkDataTaskSoup.h:
2021-01-29 Darin Adler <darin@apple.com>
Check activeDocumentLoader() for null consistently everywhere
https://bugs.webkit.org/show_bug.cgi?id=221167
Reviewed by Anders Carlsson.
* WebProcess/Plugins/PluginView.cpp:
(WebKit::PluginView::cancelManualStreamLoad): Added null check before calling
isLoadingMainResource.
2021-01-30 Wenson Hsieh <wenson_hsieh@apple.com>
[macOS] Remove support for image controls
https://bugs.webkit.org/show_bug.cgi?id=221156
Reviewed by Tim Horton.
Leave the existing SPI implemented for binary compatability, but make them empty stubs and additionally mark
the property on WKWebViewConfiguration as deprecated.
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetImageControlsEnabled):
(WKPreferencesGetImageControlsEnabled):
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _setupPageConfiguration:]):
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration init]):
(-[WKWebViewConfiguration copyWithZone:]):
(-[WKWebViewConfiguration _imageControlsEnabled]):
(-[WKWebViewConfiguration _setImageControlsEnabled:]):
* UIProcess/API/Cocoa/WKWebViewConfigurationPrivate.h:
2021-01-29 Megan Gardner <megan_gardner@apple.com>
Add plumbing to allow AppHighlights to be restored.
https://bugs.webkit.org/show_bug.cgi?id=221123
Reviewed by Wenson Hsieh.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _restoreAppHighlights:]):
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::restoreAppHighlights):
* UIProcess/WebPageProxy.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::restoreAppHighlights):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2021-01-29 Yusuke Suzuki <ysuzuki@apple.com>
[JSC] Add com.apple.private.securejit entitlement for JITCage
https://bugs.webkit.org/show_bug.cgi?id=221147
Reviewed by Keith Miller.
Add com.apple.private.securejit only for iOS.
* Scripts/process-entitlements.sh:
2021-01-29 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r272065.
https://bugs.webkit.org/show_bug.cgi?id=221155
Introduced crash
Reverted changeset:
"[macOS] Add reporting for some syscalls"
https://bugs.webkit.org/show_bug.cgi?id=221135
https://trac.webkit.org/changeset/272065
2021-01-29 Per Arne <pvollan@apple.com>
[iOS] Stop observing changes in one preference domain
https://bugs.webkit.org/show_bug.cgi?id=221130
<rdar://problem/70729182>
Reviewed by Brent Fulgham.
It has been determined that there is no need to observe preference changes in the domain 'com.apple.Preferences'.
* UIProcess/Cocoa/PreferenceObserver.mm:
(-[WKPreferenceObserver init]):
2021-01-29 Per Arne <pvollan@apple.com>
[macOS] Add audio related shm write reporting
https://bugs.webkit.org/show_bug.cgi?id=221136
<rdar://problem/73510094>
Reviewed by Brent Fulgham.
Add audio related shm write reporting in the WebContent process' sandbox.
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-29 Per Arne <pvollan@apple.com>
[macOS] Remove access to IOSurfaceSendRight
https://bugs.webkit.org/show_bug.cgi?id=221131
<rdar://problem/70496356>
Reviewed by Brent Fulgham.
Remove access to IOSurfaceSendRight in the WebContent process on macOS, since there are no reports of this being used.
* UIProcess/WebPageProxy.cpp:
(WebKit::gpuIOKitClasses):
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-29 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r272020 and r272034.
https://bugs.webkit.org/show_bug.cgi?id=221144
Revert potentially related change while investigating crash
Reverted changesets:
"[macOS] WebProcess::updateProcessName should initiate network
process connection if none exists"
https://bugs.webkit.org/show_bug.cgi?id=221094
https://trac.webkit.org/changeset/272020
"Unreviewed macOS build fix after r272020."
https://trac.webkit.org/changeset/272034
2021-01-29 Per Arne <pvollan@apple.com>
[macOS] Add reporting for some syscalls
https://bugs.webkit.org/show_bug.cgi?id=221135
<rdar://problem/73513138>
Reviewed by Brent Fulgham.
Add reporting for some syscalls in the WebContent process on macOS in order to determine if they are being used.
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-29 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r272052.
https://bugs.webkit.org/show_bug.cgi?id=221132
WPT mediacapture-record/MediaRecorder-stop.html is broken
Reverted changeset:
"Use WebRTC GPU process flag for MediaRecorder"
https://bugs.webkit.org/show_bug.cgi?id=220967
https://trac.webkit.org/changeset/272052
2021-01-29 Kimmo Kinnunen <kkinnunen@apple.com>
Some type serialization encoders are implemented in Encoder class
https://bugs.webkit.org/show_bug.cgi?id=220519
Reviewed by Chris Dumez.
Move arithmetic type and enum type serialization from Encoder to
ArgumentCoder. This makes it a bit simpler to implement multiple
Encoder types, as the serialization code is consistently separate
from the Encoder. WebGL SHM IPC stream extensions would eventually
need a similar but slightly different Encoder.
Since Encoder is intentionally coupled with the Decoder, do the
corresponding modification to Decoder class for consistency.
Reimplement Has{Legacy,Modern}Decoder for ease of forward declaration
and (perhaps?) in a bit simpler terms.
Encoder::operator<<() and Encoder::encode() were the same function.
Remove Encoder::encode().
No new tests, a refactor.
* NetworkProcess/NetworkSocketChannel.h:
* Platform/IPC/ArgumentCoder.h:
(IPC::Detail::TestLegacyDecoder):
(IPC::Detail::TestModernDecoder):
(IPC::ArgumentCoder::encode):
(IPC::ArgumentCoder::decode):
* Platform/IPC/ArgumentCoders.h:
* Platform/IPC/Connection.h:
(IPC::Connection::send):
(IPC::Connection::sendWithAsyncReply):
(IPC::Connection::sendSync):
* Platform/IPC/Decoder.h:
(IPC::Decoder::decode):
(IPC::Decoder::operator>>):
* Platform/IPC/Encoder.h:
* Platform/IPC/MessageSender.h:
* Platform/IPC/cocoa/ConnectionCocoa.mm:
(IPC::Connection::open):
* Platform/cocoa/SharedMemoryCocoa.cpp:
* Shared/RemoteLayerTree/RemoteLayerTreeTransaction.mm:
(WebKit::RemoteLayerTreeTransaction::LayerProperties::encode const):
* Shared/RemoteLayerTree/RemoteScrollingUIState.cpp:
(WebKit::RemoteScrollingUIState::encode const):
* UIProcess/AuxiliaryProcessProxy.h:
(WebKit::AuxiliaryProcessProxy::send):
(WebKit::AuxiliaryProcessProxy::sendWithAsyncReply):
* WebProcess/WebPage/IPCTestingAPI.cpp:
(WebKit::IPCTestingAPI::JSIPC::sendMessage):
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.mm:
(WebKit::RemoteLayerTreeDrawingArea::updateRendering):
2021-01-29 Philippe Normand <pnormand@igalia.com>
[GTK] return-type-c-linkage warnings
https://bugs.webkit.org/show_bug.cgi?id=221047
Reviewed by Carlos Garcia Campos.
Remove extern "C" braces, un-needed for private C++ objects.
* NetworkProcess/soup/WebKitDirectoryInputStream.h:
2021-01-29 Youenn Fablet <youenn@apple.com>
Use WebRTC GPU process flag for MediaRecorder
https://bugs.webkit.org/show_bug.cgi?id=220967
Reviewed by Eric Carlson.
* WebProcess/GPU/webrtc/MediaRecorderProvider.cpp:
(WebKit::MediaRecorderProvider::createMediaRecorderPrivate):
2021-01-29 Youenn Fablet <youenn@apple.com>
Recover audio and video capture from GPUProcess crash
https://bugs.webkit.org/show_bug.cgi?id=221086
Reviewed by Eric Carlson.
Make RemoteRealtimeMediaSource observe GPU process crash.
If RemoteRealtimeMediaSource is live and capturing in GPUProcess, restart capture from WebProcess.
If GPU process crashes, make sure to send back the necessary entitlements and information to GPUProcess
of which processes are allowed to capture.
Covered by API test.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::gpuProcessCrashed):
* WebProcess/cocoa/RemoteCaptureSampleManager.cpp:
(WebKit::RemoteCaptureSampleManager::didUpdateSourceConnection):
* WebProcess/cocoa/RemoteCaptureSampleManager.h:
* WebProcess/cocoa/RemoteRealtimeMediaSource.cpp:
(WebKit::RemoteRealtimeMediaSource::create):
(WebKit::RemoteRealtimeMediaSource::createRemoteMediaSource):
(WebKit::RemoteRealtimeMediaSource::~RemoteRealtimeMediaSource):
(WebKit::RemoteRealtimeMediaSource::gpuProcessConnectionDidClose):
* WebProcess/cocoa/RemoteRealtimeMediaSource.h:
* WebProcess/cocoa/UserMediaCaptureManager.cpp:
(WebKit::UserMediaCaptureManager::AudioFactory::createAudioCaptureSource):
(WebKit::UserMediaCaptureManager::VideoFactory::createVideoCaptureSource):
(WebKit::UserMediaCaptureManager::DisplayFactory::createDisplayCaptureSource):
(WebKit::UserMediaCaptureManager::didUpdateSourceConnection):
* WebProcess/cocoa/UserMediaCaptureManager.h:
2021-01-28 Megan Gardner <megan_gardner@apple.com>
Add plumbing to allow AppHighlights to be stored.
https://bugs.webkit.org/show_bug.cgi?id=221020
Reviewed by Devin Rousso.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _didCreateHighlightsStorageData:]):
(-[WKWebView _appHighlightsStorageDelegate]):
(-[WKWebView _setAppHighlightsStorageDelegate:]):
* UIProcess/API/Cocoa/WKWebViewInternal.h:
* UIProcess/Cocoa/PageClientImplCocoa.mm:
(WebKit::PageClientImplCocoa::didCreateHighlightsStorageData):
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
* UIProcess/PageClient.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didCreateHighlightsStorageData):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::didCreateHighlightsStorageData):
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::createAppHighlightInSelectedRange):
2021-01-28 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Adopt new SPI to show no credentials error for the platform authenticator
https://bugs.webkit.org/show_bug.cgi?id=220894
<rdar://problem/73538568>
Reviewed by Brent Fulgham.
Covered by manual tests.
* Platform/spi/Cocoa/AuthenticationServicesCoreSPI.h:
* UIProcess/WebAuthentication/Cocoa/AuthenticatorPresenterCoordinator.mm:
(WebKit::AuthenticatorPresenterCoordinator::updatePresenter):
2021-01-28 Peng Liu <peng.liu6@apple.com>
[GPUP][MSE] A video element does not fire “canplaythrough” event if SourceBuffer.abort() is called
https://bugs.webkit.org/show_bug.cgi?id=220964
Reviewed by Jer Noble.
* Shared/mac/MediaFormatReader/MediaFormatReader.cpp:
(WebKit::MediaFormatReader::parseByteSource):
2021-01-28 Chris Dumez <cdumez@apple.com>
Unreviewed macOS build fix after r272020.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::updateProcessName):
(WebKit::WebProcess::updateActivePages):
2021-01-28 Chris Dumez <cdumez@apple.com>
[macOS] WebProcess::updateProcessName should initiate network process connection if none exists
https://bugs.webkit.org/show_bug.cgi?id=221094
<rdar://73658122>
Reviewed by Per Arne Vollan.
WebProcess::updateProcessName should initiate network process connection if none exists, instead of giving up
and returning early in this case. Same policy applies to WebProcess::updateActivePages.
This is a follow-up to:
https://trac.webkit.org/changeset/271897/webkit
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::updateProcessName):
(WebKit::WebProcess::updateActivePages):
2021-01-28 Martin Robinson <mrobinson@webkit.org>
Make closestSnapOffset a method on ScrollSnapOffsetsInfo
https://bugs.webkit.org/show_bug.cgi?id=221030
Reviewed by Simon Fraser.
Make closestSnapOffset a method and change all call sites to use the new method.
* UIProcess/RemoteLayerTree/ios/RemoteScrollingCoordinatorProxyIOS.mm:
(WebKit::RemoteScrollingCoordinatorProxy::closestSnapOffsetForMainFrameScrolling const):
* UIProcess/RemoteLayerTree/ios/ScrollingTreeScrollingNodeDelegateIOS.mm:
(-[WKScrollingNodeScrollViewDelegate scrollViewWillEndDragging:withVelocity:targetContentOffset:]):
2021-01-28 Per Arne <pvollan@apple.com>
[macOS] Crash when updating color preferences
https://bugs.webkit.org/show_bug.cgi?id=221088
<rdar://problem/73709142>
Reviewed by Brent Fulgham.
After r271965, the WebContent process is no longer allowed to connect to the Launch Services daemon. This introduced a crash
in the WebContent process when color preferences were changed in System Preferences, since AppKit will then attempt to set
application information with Launch Services, which causes a crash when that fails. This patch addresses this issue by
overriding the AppKit method that updates the specific application information with Launch Services, since updating this
information should be required in the WebContent process.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::updateCanQuitQuietlyAndSafely):
(WebKit::WebProcess::platformInitializeWebProcess):
2021-01-28 Brent Fulgham <bfulgham@apple.com>
[macOS] Remove reference to deprecated SubmitDiagInfo.domains path
https://bugs.webkit.org/show_bug.cgi?id=221056
<rdar://problem/73474697>
Reviewed by Per Arne Vollan.
macOS does not use SubmitDiagInfo.domains anymore, and even when it did it failed safely.
There's no reason to have this access in any of our sandboxes, and we should remove it.
* GPUProcess/mac/com.apple.WebKit.GPUProcess.sb.in:
* NetworkProcess/mac/com.apple.WebKit.NetworkProcess.sb.in:
* PluginProcess/mac/com.apple.WebKit.plugin-common.sb.in:
* WebAuthnProcess/mac/com.apple.WebKit.WebAuthnProcess.sb.in:
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-28 Per Arne <pvollan@apple.com>
[macOS] Remove access to com.apple.BluetoothServices
https://bugs.webkit.org/show_bug.cgi?id=221059
<rdar://problem/70509471>
Reviewed by Alex Christensen.
Remove access to com.apple.BluetoothServices in the WebContent sandbox on macOS, since there are no reports of this being used.
* UIProcess/WebPageProxy.cpp:
(WebKit::mediaRelatedMachServices):
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-28 Carlos Garcia Campos <cgarcia@igalia.com>
[SOUP] Stop using SoupRequest API in preparation for libsoup3
https://bugs.webkit.org/show_bug.cgi?id=220764
Reviewed by Adrian Perez de Castro.
It's gone in libsoup3, we can just use soup_session_send_async() instead, which is what SoupRequest does
internally.
* NetworkProcess/soup/NetworkDataTaskSoup.cpp:
(WebKit::NetworkDataTaskSoup::createRequest): Manually create the SoupMessage. Also connect to content-sniffed
signal that it was handled by SoupRequest internally.
(WebKit::NetworkDataTaskSoup::clearRequest): Remove the SoupRequest reset and clear m_sniffedContentType.
(WebKit::NetworkDataTaskSoup::resume): Use soup_session_send_async(). We need to know the message for passed to
the function from the async ready callback, so we use a struct to pass both the task and the message. In case of
being suspended the struct was attached to the pending result.
(WebKit::NetworkDataTaskSoup::sendRequestCallback): Finish the soup_session_send_async() operation.
(WebKit::NetworkDataTaskSoup::didSendRequest): Set the sniffed content type from m_sniffedContentType.
(WebKit::NetworkDataTaskSoup::tlsConnectionAcceptCertificate): Use SoupMessage instead of SoupRequest to get the URI.
(WebKit::NetworkDataTaskSoup::didSniffContentCallback): Build the sniffed content type and call didSniffContent().
(WebKit::NetworkDataTaskSoup::didSniffContent): Save the passed content type.
(WebKit::NetworkDataTaskSoup::continueAuthenticate): Use m_currentRequest instead of the SoupRequest to build
the authentication error.
(WebKit::NetworkDataTaskSoup::skipInputStreamForRedirectionCallback): Pass the failing URL to the ResourceError constructor.
(WebKit::NetworkDataTaskSoup::continueHTTPRedirection): Ditto.
(WebKit::NetworkDataTaskSoup::readCallback): Ditto.
(WebKit::NetworkDataTaskSoup::requestNextPartCallback): Ditto.
* NetworkProcess/soup/NetworkDataTaskSoup.h:
* WebProcess/InjectedBundle/API/glib/WebKitWebPage.cpp: Stop setting the initiating page ID in the request because
it's unused.
2021-01-28 Carlos Garcia Campos <cgarcia@igalia.com>
Load data URLs in the web process also for synchronous loads
https://bugs.webkit.org/show_bug.cgi?id=220981
Reviewed by Adrian Perez de Castro.
In r271879 I removed the support for data URLs in the network process for soup, assuming data URLs were always
loaded from the web process, but that's not the case for synchronous loads.
* NetworkProcess/soup/NetworkDataTaskSoup.cpp:
(WebKit::NetworkDataTaskSoup::createRequest): Schedule an error if URL is not file or HTTP family.
* WebProcess/Network/WebLoaderStrategy.cpp:
(WebKit::WebLoaderStrategy::loadDataURLSynchronously): Helper to load a Data URL synchronously.
(WebKit::WebLoaderStrategy::loadResourceSynchronously): Call loadDataURLSynchronously for data URLs.
* WebProcess/Network/WebLoaderStrategy.h:
2021-01-27 Devin Rousso <drousso@apple.com>
[iOS] add accessibility image extraction contextmenu item for revealing images
https://bugs.webkit.org/show_bug.cgi?id=221058
<rdar://problem/73382220>
Reviewed by Wenson Hsieh.
* UIProcess/API/Cocoa/_WKElementAction.h:
* UIProcess/API/Cocoa/_WKElementAction.mm:
(+[_WKElementAction _elementActionWithType:customTitle:assistant:]):
(+[_WKElementAction imageForElementActionType:]):
(elementActionTypeToUIActionIdentifier):
(uiActionIdentifierToElementActionType):
* UIProcess/ios/WKActionSheetAssistant.h:
* UIProcess/ios/WKActionSheetAssistant.mm:
(-[WKActionSheetAssistant defaultActionsForLinkSheet:]):
(-[WKActionSheetAssistant defaultActionsForImageSheet:]):
(-[WKActionSheetAssistant handleElementActionWithType:element:needsInteraction:]):
* Configurations/WebKit.xcconfig:
Link UniformTypeIdentifiers.framework for use in WebKitAdditions.
2021-01-27 Chris Dumez <cdumez@apple.com>
[GPUProcess] Move ownership of RingBuffer backing used by RemoteAudioSourceProviderManager::RemoteAudio to WebProcess
https://bugs.webkit.org/show_bug.cgi?id=221048
Reviewed by Geoffrey Garen.
Move ownership of RingBuffer backing used by RemoteAudioSourceProviderManager::RemoteAudio
from the GPUProcess to WebProcess, for jetsam purposes. This moves half a megabyte of memory
from the GPUProcess to the WebProcess on https://foobar404.github.io/Wave.js.
* Platform/SharedMemory.cpp:
(WebKit::SharedMemory::Handle::takeOwnershipOfMemory const):
* Platform/SharedMemory.h:
Add takeOwnershipOfMemory() API to SharedMemory::Handle to take ownership of memory and
assign it to a particular memory ledger, via mach_memory_entry_ownership().
* Platform/cocoa/SharedMemoryCocoa.cpp:
(WebKit::toVMMemoryLedger):
(WebKit::SharedMemory::Handle::takeOwnershipOfMemory const):
Add implementation to take ownership of memory when given a mach port.
(WebKit::SharedMemory::allocate):
Add VM_FLAGS_PURGABLE flag when calling mach_vm_allocate() so that memory ownership can
be transferred later on.
(WebKit::makeMemoryEntry):
First try to call mach_make_memory_entry_64() without the MAP_MEM_VM_SHARE flag since this
flag prevents transfer of the memory ownership via mach_memory_entry_ownership(). If this
fails, we try again with the MAP_MEM_VM_SHARE flag, to be safe. MAP_MEM_VM_SHARE allows
the memory entry to cover any virtual range, possibly backed by multiple VM objects.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* WebProcess/com.apple.WebProcess.sb.in:
Tweak the sandbox to allow the WebProcess to call mach_memory_entry_ownership().
* WebProcess/GPU/media/RemoteAudioSourceProviderManager.cpp:
(WebKit::RemoteAudioSourceProviderManager::RemoteAudio::setStorage):
Take ownership of RingBuffer backing memory.
2021-01-27 Brent Fulgham <bfulgham@apple.com>
[iOS] Conditionalize debugging support on internal builds
https://bugs.webkit.org/show_bug.cgi?id=221053
<rdar://problem/73520300>
Reviewed by Per Arne Vollan.
We should restrict use of the (debugging-support) rules in our iOS sandboxes to internal OS, since those
features are not accessible to production users.
Since these restrictions don't impact end user debugging use cases, we'll also rename the rules to
"internal-debugging-support".
* Resources/SandboxProfiles/ios/com.apple.WebKit.GPU.sb:
* Resources/SandboxProfiles/ios/com.apple.WebKit.Networking.sb:
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebAuthn.sb:
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2021-01-27 Kate Cheney <katherine_cheney@apple.com>
WKWebView closeAllMediaPresentations API does not have a completion handler
https://bugs.webkit.org/show_bug.cgi?id=220741
<rdar://problem/73045904>
Reviewed by Youenn Fablet.
Adds a completion handler to WKWebView closeAllMediaPresentations.
This required handling the case for video fullscreen, picture-in-picture,
and element fullscreen. Closing fullscreen requires multiple async
operations, some of which go out of WebKit scope like PIPViewControllerDelegate
functions, so for simplicity, the callbacks are stored in
the respective *FullscreenManagerProxy object and called when
didExitFullscreen() is executed. Any additional calls to closeAllMediaPresentations
made before the previous has finished will have their callbacks
appended to a vector and called when all presentations have closed.
* UIProcess/API/Cocoa/WKWebView.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView closeAllMediaPresentations:]):
(-[WKWebView _closeAllMediaPresentations]):
(-[WKWebView _allMediaPresentationsClosed]):
New SPI for testing. Previously these tests used JavaScript to check
for inline video but updating the JS value is async and may not be completed
by the time didExitFullscreen() is called. So, this SPI will help avoid
flakiness.
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/Cocoa/VideoFullscreenManagerProxy.h:
* UIProcess/Cocoa/VideoFullscreenManagerProxy.mm:
(WebKit::VideoFullscreenModelContext::requestFullscreenModeWithCallback):
(WebKit::VideoFullscreenManagerProxy::~VideoFullscreenManagerProxy):
(WebKit::VideoFullscreenManagerProxy::forEachSession):
(WebKit::VideoFullscreenManagerProxy::callCloseCompletionHandlers):
(WebKit::VideoFullscreenManagerProxy::requestFullscreenModeWithCallback):
(WebKit::VideoFullscreenManagerProxy::didExitFullscreen):
Add a new requestFullscreenModeWithCallback function
for VideoFullscreenModelContext which is used when closing fullscreen
video. This way we don't need to adjust an overridden function.
* UIProcess/WebFullScreenManagerProxy.cpp:
(WebKit::WebFullScreenManagerProxy::~WebFullScreenManagerProxy):
(WebKit::WebFullScreenManagerProxy::callCloseCompletionHandlers):
(WebKit::WebFullScreenManagerProxy::closeWithCallback):
(WebKit::WebFullScreenManagerProxy::didExitFullScreen):
* UIProcess/WebFullScreenManagerProxy.h:
Ditto, create a new function here to be used only for closing
media presentations so we don't have to pass empty completion handlers
unnecessarily.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::pauseAllMediaPlayback):
(WebKit::WebPageProxy::suspendAllMediaPlayback):
(WebKit::WebPageProxy::resumeAllMediaPlayback):
* UIProcess/WebPageProxy.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::pauseAllMediaPlayback):
(WebKit::WebPage::suspendAllMediaPlayback):
(WebKit::WebPage::resumeAllMediaPlayback):
* WebProcess/WebPage/WebPage.h:
Remove unnecessary voids.
2021-01-27 Devin Rousso <drousso@apple.com>
[iOS] expose existing macOS-only `-[WKWebView _pageExtendedBackgroundColor]` SPI
https://bugs.webkit.org/show_bug.cgi?id=221015
<rdar://problem/73636453>
Reviewed by Tim Horton.
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::didCommitLayerTree):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::pageExtendedBackgroundColorDidChange):
* UIProcess/PageClient.h:
(WebKit::PageClient::pageExtendedBackgroundColorWillChange): Added.
(WebKit::PageClient::pageExtendedBackgroundColorDidChange): Added.
* UIProcess/Cocoa/PageClientImplCocoa.h:
* UIProcess/Cocoa/PageClientImplCocoa.mm:
(WebKit::PageClientImplCocoa::pageExtendedBackgroundColorWillChange): Added.
(WebKit::PageClientImplCocoa::pageExtendedBackgroundColorDidChange): Added.
Add support for ObjC KVO of `-[WKWebView _themeColor]`.
* WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::pageExtendedBackgroundColorDidChange): Added.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::willCommitLayerTree):
(WebKit::WebPage::flushPendingPageExtendedBackgroundColorChange): Added.
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::pageExtendedBackgroundColorDidChange const):
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::updateRendering):
On macOS, instead of messaging the UIProcess as soon as changes happen, keep a flag of
whether the `pageExtendedBackgroundColor` has changed and use it when updating rendering so
that change messages are kept in sync (e.g. modifying `<meta name="theme-color" content="...">`).
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _pageExtendedBackgroundColor]): Added.
* UIProcess/API/mac/WKWebViewMac.mm:
(-[WKWebView _pageExtendedBackgroundColor]): Deleted.
2021-01-27 Per Arne <pvollan@apple.com>
[macOS] Remove CFPBS access in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=221000
<rdar://problem/70501475>
Reviewed by Brent Fulgham.
Remove CFPBS access in the WebContent process on macOS, since there are no reports of this being used.
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-27 Per Arne <pvollan@apple.com>
[macOS] Re-enable ENABLE_SET_WEBCONTENT_PROCESS_INFORMATION_IN_NETWORK_PROCESS
https://bugs.webkit.org/show_bug.cgi?id=221039
<rdar://problem/73665061>
Reviewed by Brent Fulgham.
Work around crash introduced in r271907, by disabling persistent UI. Having persistent UI enabled will exercise code
which tries to connect with the Launch Services daemon, and crash if that fails.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2021-01-27 Per Arne <pvollan@apple.com>
[iOS] Fix sandbox violation file-read-data /dev/dtracehelper in the Networking process
https://bugs.webkit.org/show_bug.cgi?id=221046
<rdar://problem/73625718>
Reviewed by Brent Fulgham.
Reading this file should be allowed in the Networking process on internal installs on iOS.
* Resources/SandboxProfiles/ios/com.apple.WebKit.Networking.sb:
2021-01-27 Per Arne <pvollan@apple.com>
[macOS] WebContent sandbox regex incorrect
https://bugs.webkit.org/show_bug.cgi?id=221045
<rdar://problem/73473985>
Reviewed by Brent Fulgham.
When the regex is intended to match the character '.', the pattern '\.' should be used.
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-27 Per Arne <pvollan@apple.com>
[macOS] Tighten IOUSB access when camera is enabled
https://bugs.webkit.org/show_bug.cgi?id=221042
<rdar://problem/70504956>
Reviewed by Brent Fulgham.
Tighten IOUSB access when camera is enabled in the WebContent process on macOS.
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-27 Per Arne <pvollan@apple.com>
[macOS] Remove shm-write access to com.apple.AppleDatabaseChanged
https://bugs.webkit.org/show_bug.cgi?id=221041
<rdar://problem/70501252>
Reviewed by Brent Fulgham.
Remove shm-write access to com.apple.AppleDatabaseChanged in the WebContent process on maOS, since there are no reports
of this being used.
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-27 Per Arne <pvollan@apple.com>
[macOS] Enforce IOKit message filtering
https://bugs.webkit.org/show_bug.cgi?id=221038
<rdar://problem/71333134>
Reviewed by Brent Fulgham.
Update IOKit message filters based on telemetry, and deny messages not observed being used.
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-27 Per Arne <pvollan@apple.com>
[macOS] Remove fsctl-command in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=221002
<rdar://problem/70504451>
Reviewed by Brent Fulgham.
Remove fsctl-command in the WebContent process on macOS, since there are no reports of this being used.
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-27 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r271940.
https://bugs.webkit.org/show_bug.cgi?id=221040
Some tests are failing on bots
Reverted changeset:
"[MacOS] Enable NSURLSession WebSocket code path in
WebKitTestRunner"
https://bugs.webkit.org/show_bug.cgi?id=220973
https://trac.webkit.org/changeset/271940
2021-01-27 Per Arne <pvollan@apple.com>
Restrict access to syslog in sandbox
https://bugs.webkit.org/show_bug.cgi?id=220983
<rdar://problem/73512626>
Reviewed by Brent Fulgham.
Only allow syslog access on internal builds on macOS and iOS.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-27 Antti Koivisto <antti@apple.com>
Implement visibility based resource load scheduling for low priority resources
https://bugs.webkit.org/show_bug.cgi?id=220728
Reviewed by Geoff Garen.
Add a simple network process side resource load scheduler. When active it limits the number of low priority resource loads
(mostly images) that are passed to network layer. The current limit is 6 per host, reduced by any ongoing higher priority
loads. This reduces impact of low priority loads to higher priority ones, espcially on HTTP/2.
The scheduler also supports reprioritizing loads. This is used to let images currently in the viewport to skip the queue
and start their loads faster.
The feature is not enabled in this patch.
* NetworkProcess/Downloads/Download.cpp:
(WebKit::Download::cancel):
(WebKit::Download::didReceiveData):
(WebKit::Download::didFinish):
(WebKit::Download::didFail):
* NetworkProcess/NetworkConnectionToWebProcess.cpp:
(WebKit::NetworkConnectionToWebProcess::clearPageSpecificData):
(WebKit::NetworkConnectionToWebProcess::removeStorageAccessForFrame):
(WebKit::NetworkConnectionToWebProcess::setResourceLoadSchedulingMode):
(WebKit::NetworkConnectionToWebProcess::prioritizeResourceLoads):
(WebKit::NetworkConnectionToWebProcess::clearPageSpecificDataForResourceLoadStatistics): Deleted.
Rename to clearPageSpecificData and move out of #if.
* NetworkProcess/NetworkConnectionToWebProcess.h:
* NetworkProcess/NetworkConnectionToWebProcess.messages.in:
* NetworkProcess/NetworkLoad.cpp:
(WebKit::NetworkLoad::start):
(WebKit::NetworkLoad::startWithScheduling):
(WebKit::NetworkLoad::~NetworkLoad):
(WebKit::NetworkLoad::didCompleteWithError):
* NetworkProcess/NetworkLoad.h:
* NetworkProcess/NetworkLoadScheduler.cpp: Added.
(WebKit::NetworkLoadScheduler::HostContext::schedule):
(WebKit::NetworkLoadScheduler::HostContext::unschedule):
(WebKit::NetworkLoadScheduler::HostContext::prioritize):
(WebKit::NetworkLoadScheduler::HostContext::start):
(WebKit::NetworkLoadScheduler::HostContext::~HostContext):
(WebKit::NetworkLoadScheduler::schedule):
(WebKit::NetworkLoadScheduler::unschedule):
(WebKit::NetworkLoadScheduler::setResourceLoadSchedulingMode):
(WebKit::NetworkLoadScheduler::prioritizeLoads):
(WebKit::NetworkLoadScheduler::clearPageData):
(WebKit::NetworkLoadScheduler::contextForLoad):
* NetworkProcess/NetworkLoadScheduler.h: Copied from Source/WebCore/loader/LoaderStrategy.cpp.
* NetworkProcess/NetworkResourceLoader.cpp:
(WebKit::NetworkResourceLoader::startNetworkLoad):
* NetworkProcess/NetworkSession.cpp:
(WebKit::NetworkSession::networkLoadScheduler):
* NetworkProcess/NetworkSession.h:
* NetworkProcess/cache/NetworkCacheSpeculativeLoad.cpp:
(WebKit::NetworkCache::SpeculativeLoad::SpeculativeLoad):
* Sources.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/Network/WebLoaderStrategy.cpp:
(WebKit::WebLoaderStrategy::setResourceLoadSchedulingMode):
(WebKit::WebLoaderStrategy::prioritizeResourceLoads):
* WebProcess/Network/WebLoaderStrategy.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::close):
2021-01-27 Youenn Fablet <youenn@apple.com>
[MacOS] Enable NSURLSession WebSocket code path in WebKitTestRunner
https://bugs.webkit.org/show_bug.cgi?id=220973
Reviewed by Darin Adler.
Disable a debug assert unti it get fixed below.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(-[WKNetworkSessionDelegate URLSession:webSocketTask:didCloseWithCode:reason:]):
2021-01-27 Andy Estes <aestes@apple.com>
[Mac] MediaSampleCursor::stepInOrderedMap can hang when stepping to the end of a track that hasn't been fully parsed
https://bugs.webkit.org/show_bug.cgi?id=220555
<rdar://problem/72910000>
Reviewed by Eric Carlson.
When a container has multiple tracks, CoreMedia determines the overall duration by creating
a cursor at a presentation time of +infinity then asking for its timing info, the expectation
being that the cursor will clamp itself to the track duration and report that in
MediaSampleCursor::getSampleTiming.
Previously, this could result in a hang for tracks that had not yet finished parsing, as
MediaSampleCursor was designed to wait for samples outside the playable range. With this
change, MediaSampleCursor no longer waits for samples outside the playable range but
rather reports clamped timing info without waiting for a sample. If CoreMedia asks for
information that requires a sample (e.g., sample location), MediaSampleCursor will respond
with `kMTPluginSampleCursorError_LocationNotAvailable`. So that CoreMedia can determine
the range of playable samples, this patch implements MediaSampleCursor::getPlayableHorizon.
To simplify this implementation, MediaSampleCursor now assumes that decode time equals
presentation time (for video tracks) or is invalid (for audio tracks). While here, also
added media logging that helped debug this issue.
Test: http/tests/media/video-canplaythrough-webm.html
* Shared/mac/MediaFormatReader/MediaSampleByteRange.cpp:
(WebKit::MediaSampleByteRange::MediaSampleByteRange): Asserted our assmpution about the
relationship between presentation time and decode time.
* Shared/mac/MediaFormatReader/MediaSampleCursor.cpp:
(WebKit::assumedDecodeTime): Added to compute decode time from presentation time and a
nominal sample.
(WebKit::upperBound): Changed to compute the decode key based on assumed decode time.
(WebKit::stepIterator): Removed the MediaTime specialization.
(WebKit::stepTime): Added to step a MediaTime rather than an iterator.
(WebKit::MediaSampleCursor::MediaSampleCursor): Initialized logging.
(WebKit::MediaSampleCursor::locateIterator const): Changed to find an iterator based on
presentation time and assumed decode time.
(WebKit::MediaSampleCursor::locateMediaSample const): Ditto.
(WebKit::MediaSampleCursor::locateTiming const): Added to locate timing info even if no
sample is available.
(WebKit::MediaSampleCursor::stepInOrderedMap): Changed to call getSampleMap, and removed
the MediaTime specialization.
(WebKit::MediaSampleCursor::stepInPresentationTime): Added to replace the MediaTime
specialization of stepInOrderedMap.
(WebKit::MediaSampleCursor::getSampleMap const): Added to call
MediaTrackReader::waitForSample and return as soon as at least one sample is available.
Returns kMTPluginSampleCursorError_NoSamples if the track finishes parsing with no
samples.
(WebKit::MediaSampleCursor::getMediaSample const): Changed to call getSampleMap.
(WebKit::MediaSampleCursor::getTiming const): Added to get timing info.
(WebKit::MediaSampleCursor::copyProperty): Added logging.
(WebKit::MediaSampleCursor::stepInDecodeOrderAndReportStepsTaken): Changed to call the new
version of stepInOrderedMap.
(WebKit::MediaSampleCursor::stepInPresentationOrderAndReportStepsTaken): Ditto.
(WebKit::MediaSampleCursor::stepByDecodeTime): Changed to call stepInPresentationTime.
(WebKit::MediaSampleCursor::stepByPresentationTime): Ditto.
(WebKit::MediaSampleCursor::compareInDecodeOrder const): Changed to call getTiming.
(WebKit::MediaSampleCursor::getSampleTiming const): Ditto.
(WebKit::MediaSampleCursor::getPlayableHorizon const): Implemented by taking the
difference between the last sample's end time and the cursor's presentation time.
(WebKit::MediaSampleCursor::logChannel const): Added.
(WebKit::makeIterator): Deleted.
(WebKit::makeTime): Deleted.
(WebKit::MediaSampleCursor::createAtDecodedSample): Deleted.
* Shared/mac/MediaFormatReader/MediaSampleCursor.h:
(WebKit::MediaSampleCursor::logger const): Configured logging.
(WebKit::MediaSampleCursor::logClassName const): Ditto.
(WebKit::MediaSampleCursor::logIdentifier const): Ditto.
* Shared/mac/MediaFormatReader/MediaTrackReader.cpp:
(WebKit::MediaTrackReader::mediaTypeString const): Ditto.
(WebKit::MediaTrackReader::nextSampleCursorLogIdentifier const): Ditto.
* Shared/mac/MediaFormatReader/MediaTrackReader.h:
2021-01-27 Martin Robinson <mrobinson@igalia.com>
Use ScrollSnapOffsetsInfo in the scrolling tree
https://bugs.webkit.org/show_bug.cgi?id=220915
Reviewed by Simon Fraser.
Use ScrollSnapOffsets more often in the scrolling tree to continue to
abstract away the details of how snap offsets are implemented.
* Shared/RemoteLayerTree/RemoteScrollingCoordinatorTransaction.cpp:
(ArgumentCoder<ScrollingStateScrollingNode>::encode):
(ArgumentCoder<ScrollingStateScrollingNode>::decode):
(ArgumentCoder<ScrollSnapOffsetsInfo<float>>::encode):
(ArgumentCoder<ScrollSnapOffsetsInfo<float>>::decode):
(WebKit::dump):
* UIProcess/RemoteLayerTree/ios/RemoteScrollingCoordinatorProxyIOS.mm:
(WebKit::RemoteScrollingCoordinatorProxy::shouldSnapForMainFrameScrolling const):
(WebKit::RemoteScrollingCoordinatorProxy::closestSnapOffsetForMainFrameScrolling const):
(WebKit::RemoteScrollingCoordinatorProxy::hasActiveSnapPoint const):
(WebKit::RemoteScrollingCoordinatorProxy::nearestActiveContentInsetAdjustedSnapOffset const):
* UIProcess/RemoteLayerTree/ios/ScrollingTreeScrollingNodeDelegateIOS.mm:
(-[WKScrollingNodeScrollViewDelegate scrollViewWillEndDragging:withVelocity:targetContentOffset:]):
(WebKit::ScrollingTreeScrollingNodeDelegateIOS::commitStateAfterChildren):
2021-01-26 Sihui Liu <sihui_liu@appe.com>
Mute audio capture for speech recognition based on shouldInterruptAudioOnPageVisibilityChange when page is invisible
https://bugs.webkit.org/show_bug.cgi?id=220960
Reviewed by Youenn Fablet.
We used to abort ongoing recognition when page becomes invisible. To match media capture's behavior, now we only
abort recognition when shouldInterruptAudioOnPageVisibilityChange is true.
Updated API test: WebKit2.SpeechRecognitionPageBecomesInvisible
* UIProcess/SpeechRecognitionServer.cpp:
(WebKit::SpeechRecognitionServer::abortForPageIsBecomingInvisible): Deleted.
* UIProcess/SpeechRecognitionServer.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::pageIsBecomingInvisible):
2021-01-26 Devin Rousso <drousso@apple.com>
Expose the value of `<meta name="theme-color" content="...">` as SPI
https://bugs.webkit.org/show_bug.cgi?id=220944
<rdar://problem/72198083>
Reviewed by Tim Horton.
* WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::themeColorChanged): Added.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::willCommitLayerTree):
(WebKit::WebPage::didCommitLoad):
(WebKit::WebPage::flushPendingThemeColorChange): Added.
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::themeColorChanged const):
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::updateRendering):
* UIProcess/WebPageProxy.messages.in:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::themeColorChanged): Added.
On macOS, keep a flag indicating whether the `themeColor` has changed. Use this flag when
updating rendering to send the `Color` to the UIProcess so that it's kept in sync with other
changes (e.g. modifying the CSS `background-color`).
* Shared/RemoteLayerTree/RemoteLayerTreeTransaction.h:
(WebKit::RemoteLayerTreeTransaction::themeColor const): Added.
(WebKit::RemoteLayerTreeTransaction::setThemeColor): Added.
* Shared/RemoteLayerTree/RemoteLayerTreeTransaction.mm:
(WebKit::RemoteLayerTreeTransaction::encode const):
(WebKit::RemoteLayerTreeTransaction::decode):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::didCommitLayerTree):
On iOS, include the `themeColor` in every `RemoteLayerTreeTransaction` so that it's kept in
sync with other changes (e.g. modifying the CSS `background-color`).
* UIProcess/PageClient.h:
(WebKit::PageClient::themeColorWillChange): Added.
(WebKit::PageClient::themeColorDidChange): Added.
* UIProcess/Cocoa/PageClientImplCocoa.h:
* UIProcess/Cocoa/PageClientImplCocoa.mm:
(WebKit::PageClientImplCocoa::themeColorWillChange): Added.
(WebKit::PageClientImplCocoa::themeColorDidChange): Added.
Add support for ObjC KVO of `-[WKWebView _themeColor]`.
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::themeColor const): Added.
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _themeColor]): Added.
2021-01-26 Brent Fulgham <bfulgham@apple.com>
IPC::Decoder constructor should mark the Decoder as invalid if header decoding fails
https://bugs.webkit.org/show_bug.cgi?id=220990
<rdar://problem/64984186>
Reviewed by Darin Adler.
Mark the decoder as invalid if any of the decode operations fail, not just the first one.
* Platform/IPC/Decoder.cpp:
2021-01-26 Brent Fulgham <bfulgham@apple.com>
Harden NetworkProcess against integer overflow in CacheStorageEngine size calculation
https://bugs.webkit.org/show_bug.cgi?id=220997
<rdar://problem/66116827>
Reviewed by Youenn Fablet.
Since the CacheStorage allocation is based on data provided by the WebContent process, we should
check for integer overflow before making an allocation.
* NetworkProcess/cache/CacheStorageEngineCache.cpp:
(WebKit::CacheStorage::Cache::put): Use checked arithmetic and return a failure if we overflow.
2021-01-26 Per Arne <pvollan@apple.com>
[macOS] Deny mach-lookup to the service 'com.apple.tccd.system' in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=220976
<rdar://problem/68935772>
Reviewed by Brent Fulgham.
Deny mach-lookup to the service 'com.apple.tccd.system' in the WebContent process, since there are no longer reports of this being used.
No new tests, since the result will depend on OS version.
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-26 Chris Dumez <cdumez@apple.com>
Disable TBA availability macro post-processing on iOS 14 / macOS 11
https://bugs.webkit.org/show_bug.cgi?id=221004
Reviewed by Tim Horton.
Disable TBA availability macro post-processing on iOS 14 / macOS 11, now that TBA headers have been updated in r271813.
* Configurations/WebKit.xcconfig:
* Configurations/WebKitTargetConditionals.xcconfig:
2021-01-26 Per Arne <pvollan@apple.com>
[macOS] Close all XPC connections to Launch Services after checking in
https://bugs.webkit.org/show_bug.cgi?id=220952
<rdar://13785139>
Reviewed by Brent Fulgham.
After checking in with Launch Services on macOS, close all open XPC connections to the service.
This patch also adds checks for a valid Network process connection when updating the WebContent
process' information.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
(WebKit::WebProcess::updateProcessName):
(WebKit::WebProcess::updateActivePages):
2021-01-26 Per Arne <pvollan@apple.com>
[macOS] Add local sandbox reporting for access to Metal services in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=220986
Reviewed by Brent Fulgham.
To determine if these services are still used in the WebContent process when the GPU process is enabled, add local sandbox reporting for access to Metal services.
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-26 Brent Fulgham <bfulgham@apple.com>
REGRESSION(r261238): WKWebView crashes on launch inside a quicklook preview
https://bugs.webkit.org/show_bug.cgi?id=219632
<rdar://problem/72154830>
Reviewed by Alexey Proskuryakov.
Partially revert r261238 to allow applications without access to 'nsurlsessiond' to still
be able to use WebKit for local file access. This workaround should be removed once we
complete the work needed to support true offline WebKit use.
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceEntryPoint.mm:
(WebKit::XPCServiceInitializerDelegate::checkEntitlements): Add an early return if the application
possesses the 'com.apple.security.network.client' entitlement.
2021-01-26 Per Arne <pvollan@apple.com>
[macOS] Restrict process-info in sandbox
https://bugs.webkit.org/show_bug.cgi?id=220985
<rdar://problem/70353532>
Reviewed by Brent Fulgham.
Access to process-info should be restricted to self.
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-26 Per Arne <pvollan@apple.com>
[macOS] Deny mach-lookup to the service 'com.apple.audio.SandboxHelper' in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=220978
<rdar://problem/69320352>
Reviewed by Brent Fulgham.
Deny mach-lookup to the service 'com.apple.audio.SandboxHelper' in the WebContent process. Access to this service is only allowed
by extension if Media is running in the WebContent process.
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-26 Per Arne <pvollan@apple.com>
[macOS] Restrict access to OpenGL profiler
https://bugs.webkit.org/show_bug.cgi?id=220980
<rdar://problem/70354481>
Reviewed by Brent Fulgham.
Only allow the OpenGL profiler to attach on internal installs.
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-26 Brent Fulgham <bfulgham@apple.com>
[CoreIPC] Add size check in decodeSharedBuffer to allow it to be validated against the SharedMemory received
https://bugs.webkit.org/show_bug.cgi?id=220962
<rdar://problem/68204552>
Reviewed by Darin Adler.
Add a new check during decoding so that we fail if the expected buffer size is larger than the block of
shared memory supplied in the IPC call.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::decodeSharedBuffer):
2021-01-26 Alex Christensen <achristensen@webkit.org>
Revert part of r271493
https://bugs.webkit.org/show_bug.cgi?id=220066
<rdar://problem/73615999>
We have an internal app that calls registerSchemeForCustomProtocol on a non-main thread before WebKit initializes.
While this is a problem, we have to not crash on launch until they fix this.
* UIProcess/API/Cocoa/WKBrowsingContextController.mm:
(+[WKBrowsingContextController registerSchemeForCustomProtocol:]):
(+[WKBrowsingContextController unregisterSchemeForCustomProtocol:]):
2021-01-26 Per Arne <pvollan@apple.com>
[GPU process][macOS] Only allow access to diagnostic services for internal builds
https://bugs.webkit.org/show_bug.cgi?id=220957
Reviewed by Brent Fulgham.
This patch also removes access to an unused logging service.
* GPUProcess/mac/com.apple.WebKit.GPUProcess.sb.in:
2021-01-26 Chris Dumez <cdumez@apple.com>
Unreviewed, silence deprecation warning for our own API to address build issues.
* UIProcess/API/Cocoa/WKUserContentControllerPrivate.h:
2021-01-26 Chris Dumez <cdumez@apple.com>
Unreviewed, reverting r271882.
Broke the build
Reverted changeset:
"Unreviewed, silence deprecation warning for our own API to
address build issues."
https://trac.webkit.org/changeset/271882
2021-01-26 Chris Dumez <cdumez@apple.com>
Unreviewed, silence deprecation warning for our own API to address build issues.
* UIProcess/API/Cocoa/WKUserContentControllerPrivate.h:
2021-01-26 Kimmo Kinnunen <kkinnunen@apple.com>
WebGL power preference and discrete/internal gpu selection implemented incorrectly with ANGLE
https://bugs.webkit.org/show_bug.cgi?id=220843
Reviewed by Dean Jackson.
Move the high-performance GPU shutdown timer from individual web processes to the main class
in the ui process. This simplifies the implementation and reduces the number of timers.
* UIProcess/mac/HighPerformanceGPUManager.h:
* UIProcess/mac/HighPerformanceGPUManager.mm:
(WebKit::HighPerformanceGPUManager::HighPerformanceGPUManager):
(WebKit::HighPerformanceGPUManager::removeProcessRequiringHighPerformance):
(WebKit::HighPerformanceGPUManager::updateState):
* WebProcess/WebCoreSupport/mac/WebSwitchingGPUClient.cpp:
(WebKit::WebSwitchingGPUClient::requestHighPerformanceGPU):
(WebKit::WebSwitchingGPUClient::releaseHighPerformanceGPU):
* WebProcess/WebCoreSupport/mac/WebSwitchingGPUClient.h:
2021-01-26 Carlos Garcia Campos <cgarcia@igalia.com>
[SOUP] Stop using SoupRequest API to load files in preparation for libsoup3
https://bugs.webkit.org/show_bug.cgi?id=220509
Reviewed by Adrian Perez de Castro.
SoupRequest API is gone in libsoup3 and there's no replacement for file requests. GResource and data URI loads
already happen in the web process so we only need to care about file and directory loads.
* NetworkProcess/soup/NetworkDataTaskSoup.cpp:
(WebKit::NetworkDataTaskSoup::createRequest): Create a GFile for local files.
(WebKit::NetworkDataTaskSoup::clearRequest): Clear the m_file.
(WebKit::NetworkDataTaskSoup::resume): Check also if there's a file operation to resume.
(WebKit::NetworkDataTaskSoup::didSendRequest): We always have a soup message now here because SoupRequest is
only used for HTTP.
(WebKit::NetworkDataTaskSoup::readCallback): In case of failure check if this is an HTTP or file request to
create the error.
(WebKit::NetworkDataTaskSoup::fileQueryInfoCallback): Call didGetFileInfo() and continue reading the file or directory.
(WebKit::NetworkDataTaskSoup::didGetFileInfo): Set URL, content type and length of the response for the given GFileInfo.
(WebKit::NetworkDataTaskSoup::readFileCallback): Complete the g_file_read_async operation.
(WebKit::NetworkDataTaskSoup::enumerateFileChildrenCallback): Create a WebKitDirectoryInputStream.
(WebKit::NetworkDataTaskSoup::didReadFile): Set the input stream and notify the response is ready.
* NetworkProcess/soup/NetworkDataTaskSoup.h:
* NetworkProcess/soup/Resources/directory.css: Added.
* NetworkProcess/soup/Resources/directory.js: Added.
* NetworkProcess/soup/WebKitDirectoryInputStream.cpp: Added.
(webkitDirectoryInputStreamCreateHeader):
(webkitDirectoryInputStreamCreateFooter):
(webkitDirectoryInputStreamCreateRow):
(webkitDirectoryInputStreamReadNextFile):
(webkitDirectoryInputStreamRead):
(webkitDirectoryInputStreamClose):
(webkit_directory_input_stream_class_init):
(webkitDirectoryInputStreamNew):
* NetworkProcess/soup/WebKitDirectoryInputStream.h: Added.
* PlatformGTK.cmake:
* PlatformWPE.cmake:
* SourcesGTK.txt:
* SourcesWPE.txt:
2021-01-25 Alexey Shvayka <shvaikalesh@gmail.com>
REGRESSION (r270874): Some React Native apps are reported broken on iOS
https://bugs.webkit.org/show_bug.cgi?id=220809
Reviewed by Saam Barati.
* WebProcess/Plugins/Netscape/JSNPObject.h:
2021-01-25 Chris Dumez <cdumez@apple.com>
Unreviewed, add missing header includes to address build issues.
* UIProcess/API/Cocoa/WKUserScript.mm:
* UIProcess/API/Cocoa/_WKInspectorExtension.mm:
* UIProcess/Cocoa/AutomationClient.mm:
* UIProcess/WebAuthentication/Cocoa/AuthenticatorPresenterCoordinator.h:
2021-01-25 Devin Rousso <drousso@apple.com>
[iOS] improve support for sequential scribble interactions that involving different inputs
https://bugs.webkit.org/show_bug.cgi?id=220933
<rdar:/problem/71911346>
Reviewed by Tim Horton.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView cleanUpInteraction]):
(-[WKContentView becomeFirstResponderForWebView]):
(-[WKContentView _zoomToRevealFocusedElement]):
(-[WKContentView _didCommitLoadForMainFrame]):
(-[WKContentView _willBeginTextInteractionInTextInputContext:]):
(-[WKContentView _didFinishTextInteractionInTextInputContext:]):
Rather than have a simple "yes/no" `_textInteractionIsHappening` we should keep a count of
scribble interactions (increment in `willBegin*` and decrement in `didFinish*`) so that we
don't send unnecessary/incorrect IPC to the WebProcess after the first scribble commits if
there's a second scribble in progress for another element.
2021-01-25 Aditya Keerthi <akeerthi@apple.com>
[iOS] Obscured elements should not be focusable using the accessory bar
https://bugs.webkit.org/show_bug.cgi?id=220929
<rdar://problem/73239409>
Reviewed by Tim Horton.
Currently, assistable elements hidden behind unassistable elements can
be focused using the next/previous buttons on the accessory bar. This
behavior is undesirable since users cannot see the focused element, but
are still presented UI to enter information into the element.
While the current behavior has been longstanding, its undesirability was
exposed after r269059, which made it possible to use the accessory bar
to focus elements across focusable scopes. As a result, it became possible
to focus a hidden input field on appleid.apple.com using the accessory bar.
Note that the hidden field is inside a subframe, which was previously
excluded from accessory bar next/previous operations. However, the same
bug would still exist even if the field was not inside a subframe, which
means that r269059 is not the root cause of the issue, but simply exposed
a larger underlying problem.
To prevent obscured elements from being focusable using the accessory bar,
we now use hit testing to determine whether or not an element is obscured,
before considering the element "assistable".
Test: fast/forms/ios/accessory-bar-navigation-obscured-input.html
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::isObscuredElement):
An element is considered obscured if a hit test at its center position
relative to the main document targets a different element.
(WebKit::nextAssistableElement):
Obscured elements should not be assistable.
2021-01-25 Per Arne <pvollan@apple.com>
[macOS] Create extension to Metal compiler service based on GPUP flags
https://bugs.webkit.org/show_bug.cgi?id=220931
<rdar://problem/73574476>
Reviewed by Brent Fulgham.
Depending on which GPUP settings are enabled, an extension to the Metal compiler service should be created.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::gpuMachServices):
(WebKit::WebPageProxy::creationParameters):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-25 Per Arne Vollan <pvollan@apple.com>
[macOS] Restrict access to dtracehelper
https://bugs.webkit.org/show_bug.cgi?id=220927
<rdar://70350947>
Reviewed by Brent Fulgham.
Only allow access to dtracehelper for internal installs.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-25 Chris Dumez <cdumez@apple.com>
Update availability annotations to match the macOS 11.0 and iOS 14.0 GM SDKs
https://bugs.webkit.org/show_bug.cgi?id=220874
<rdar://73474368>
Reviewed by Darin Adler.
* UIProcess/API/Cocoa/WKContentWorld.h:
* UIProcess/API/Cocoa/WKError.h:
* UIProcess/API/Cocoa/WKFindConfiguration.h:
* UIProcess/API/Cocoa/WKFindResult.h:
* UIProcess/API/Cocoa/WKFrameInfoPrivate.h:
* UIProcess/API/Cocoa/WKHTTPCookieStorePrivate.h:
* UIProcess/API/Cocoa/WKNavigationDelegate.h:
* UIProcess/API/Cocoa/WKNavigationDelegatePrivate.h:
* UIProcess/API/Cocoa/WKOpenPanelParametersPrivate.h:
* UIProcess/API/Cocoa/WKPDFConfiguration.h:
* UIProcess/API/Cocoa/WKPreferences.h:
* UIProcess/API/Cocoa/WKPreferencesPrivate.h:
* UIProcess/API/Cocoa/WKProcessPoolPrivate.h:
* UIProcess/API/Cocoa/WKScriptMessage.h:
* UIProcess/API/Cocoa/WKScriptMessageHandlerWithReply.h:
* UIProcess/API/Cocoa/WKUIDelegatePrivate.h:
* UIProcess/API/Cocoa/WKURLSchemeTaskPrivate.h:
* UIProcess/API/Cocoa/WKUserContentController.h:
* UIProcess/API/Cocoa/WKUserContentControllerPrivate.h:
* UIProcess/API/Cocoa/WKUserScript.h:
* UIProcess/API/Cocoa/WKUserScriptPrivate.h:
* UIProcess/API/Cocoa/WKWebView.h:
* UIProcess/API/Cocoa/WKWebViewConfiguration.h:
* UIProcess/API/Cocoa/WKWebViewConfigurationPrivate.h:
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/API/Cocoa/WKWebpagePreferences.h:
* UIProcess/API/Cocoa/WKWebpagePreferencesPrivate.h:
* UIProcess/API/Cocoa/WKWebsiteDataRecordPrivate.h:
* UIProcess/API/Cocoa/WKWebsiteDataStorePrivate.h:
* UIProcess/API/Cocoa/_WKDownload.h:
* UIProcess/API/Cocoa/_WKFrameTreeNode.h:
* UIProcess/API/Cocoa/_WKInputDelegate.h:
* UIProcess/API/Cocoa/_WKInspectorDebuggableInfo.h:
* UIProcess/API/Cocoa/_WKProcessPoolConfiguration.h:
* UIProcess/API/Cocoa/_WKResourceLoadDelegate.h:
* UIProcess/API/Cocoa/_WKResourceLoadInfo.h:
* UIProcess/API/Cocoa/_WKResourceLoadStatisticsFirstParty.h:
* UIProcess/API/Cocoa/_WKResourceLoadStatisticsThirdParty.h:
* UIProcess/API/Cocoa/_WKTextManipulationConfiguration.h:
* UIProcess/API/Cocoa/_WKTextManipulationExclusionRule.h:
* UIProcess/API/Cocoa/_WKTextManipulationItem.h:
* UIProcess/API/Cocoa/_WKTextManipulationToken.h:
* UIProcess/API/Cocoa/_WKUserContentWorld.h:
* UIProcess/API/Cocoa/_WKUserStyleSheet.h:
* UIProcess/API/Cocoa/_WKWebAuthenticationAssertionResponse.h:
* UIProcess/API/Cocoa/_WKWebAuthenticationPanel.h:
* UIProcess/API/Cocoa/_WKWebsiteDataStoreConfiguration.h:
* UIProcess/API/Cocoa/_WKWebsitePolicies.h:
2021-01-25 Chris Dumez <cdumez@apple.com>
[iOS] Stop leaking an XPC boost message to XPC services
https://bugs.webkit.org/show_bug.cgi?id=220883
<rdar://47402853>
Reviewed by Geoffrey Garen.
Turn on RB_full_manage_WK_jetsam feature flag so that we stop leaking an XPC boost
message to XPC services on IOS_FAMILY. Instead, RunningBoard will now be setting
the priority of our XPC services based on the RunningBoard assertions we're holding
for them. The feature flag is checked both by RunningBoard and WebKit.
I have verified that our bots have upgraded to a build that supports this.
* FeatureFlags/WebKit-appletvos.plist:
* FeatureFlags/WebKit-ios.plist:
* FeatureFlags/WebKit-watchos.plist:
2021-01-25 Per Arne <pvollan@apple.com>
[macOS] Restrict access to mds data
https://bugs.webkit.org/show_bug.cgi?id=220920
<rdar://70355552>
Reviewed by Brent Fulgham.
Only root show have write access.
* GPUProcess/mac/com.apple.WebKit.GPUProcess.sb.in:
* NetworkProcess/mac/com.apple.WebKit.NetworkProcess.sb.in:
* WebAuthnProcess/mac/com.apple.WebKit.WebAuthnProcess.sb.in:
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-25 Per Arne <pvollan@apple.com>
[Cocoa] Adjust logic for creating sandbox extensions based on GPUP flags
https://bugs.webkit.org/show_bug.cgi?id=220917
Reviewed by Brent Fulgham.
Update which GPUP flag should be used in this logic. This patch also includes many IOKit classes on macOS in the set of
classes that should be extended when the appropriate GPU flags are set. In the case they are not being extended, logging
rules have been added in the sandbox, to determine if they are being used in that case.
* UIProcess/WebPageProxy.cpp:
(WebKit::gpuIOKitClasses):
(WebKit::WebPageProxy::creationParameters):
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-25 Chris Dumez <cdumez@apple.com>
Support AbortSignal in addEventListenerOptions to unsubscribe from events
https://bugs.webkit.org/show_bug.cgi?id=218753
<rdar://problem/71258012>
Reviewed by Darin Adler.
Minor build fixes.
* WebProcess/Plugins/PDF/PDFPluginAnnotation.mm:
2021-01-25 Wenson Hsieh <wenson_hsieh@apple.com>
[GPU Process] Web process should be terminated if DisplayListReaderHandle advances past 0
https://bugs.webkit.org/show_bug.cgi?id=220926
Reviewed by Simon Fraser.
Address another FIXME in the GPU Process by replacing a release assertion with a MESSAGE_CHECK. See below for
more details.
* GPUProcess/graphics/DisplayListReaderHandle.cpp:
(WebKit::DisplayListReaderHandle::advance):
Return `WTF::nullopt` instead of just crashing in the case where the number of bytes to advance exceeds the
number of unread bytes. This can (and should) only happen in the case where a compromised web content process
attempts to overwrite the number of unread bytes in the shared display list header.
* GPUProcess/graphics/DisplayListReaderHandle.h:
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::nextDestinationImageBufferAfterApplyingDisplayLists):
* Shared/SharedDisplayListHandle.h:
Remove this virtual method from the base class, since the reader handle now returns an `Optional<size_t>`
instead of just a `size_t`.
* WebProcess/GPU/graphics/DisplayListWriterHandle.h:
2021-01-24 Simon Fraser <simon.fraser@apple.com>
[iOS WK2] theverge.com - rubber band scrolling at the top of the page causes an abrupt jump
https://bugs.webkit.org/show_bug.cgi?id=220886
<rdar://71177566>
Reviewed by Sam Weinig.
theverge.com on iOS is a page that has long main thread stalls with forced layouts on a timer
that alter the page height; this caused the post-layout updateScrollbars() called from
FrameView::adjustViewSize() to call scrollToPosition() after adjusting the scroll position
to the allowed range.
If the page laid out while rubberbanding was happening, the current scroll position would
be negative, then clamped to 0, then sent to the UI process as a requested scroll to 0,
triggering the jump to top in the UI process.
There's existing code to prevent this from happening if we know that rubberbanding is
happening; this patch makes isRubberBandInProgress() work for iOS WK2. It does so
by having updateVisibleContentRects() push information about rubberbanding nodes onto
RemoteScrollingCoordinator.
We remove an unnecessary shouldUpdateScrollLayerPositionSynchronously() check in
FrameView::isRubberBandInProgress() - if it's true, then the scrolling coordinator
won't see any rubberbanding nodes anyway.
* UIProcess/RemoteLayerTree/ios/RemoteScrollingCoordinatorProxyIOS.mm:
* WebProcess/WebPage/RemoteLayerTree/RemoteScrollingCoordinator.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteScrollingCoordinator.mm:
(WebKit::RemoteScrollingCoordinator::addNodeWithActiveRubberBanding):
(WebKit::RemoteScrollingCoordinator::removeNodeWithActiveRubberBanding):
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::updateVisibleContentRects):
2021-01-23 Per Arne Vollan <pvollan@apple.com>
[macOS] Deny mach lookup access to "com.apple.iconservices" in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=220013
<rdar://problem/60649036>
Reviewed by Brent Fulgham.
Deny mach lookup access to "com.apple.iconservices" in the WebContent process on macOS,
since there are no longer any reports of this being used.
Test: fast/sandbox/mac/sandbox-mach-lookup.html
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-23 Per Arne Vollan <pvollan@apple.com>
[macOS] Deny mach lookup access to "com.apple.iconservices.store" in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=220014
<rdar://problem/68923110>
Reviewed by Brent Fulgham.
Deny mach lookup access to "com.apple.iconservices.store" in the WebContent process on macOS,
since there are no longer any reports of this being used.
Test: fast/sandbox/mac/sandbox-mach-lookup.html
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-23 Per Arne Vollan <pvollan@apple.com>
[macOS] Deny mach-lookup to the service 'com.apple.hiservices-xpcservice' in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=220321
<rdar://problem/68878470>
Reviewed by Darin Adler.
Deny mach-lookup to the service 'com.apple.hiservices-xpcservice' in the WebContent process on macOS,
since there are no longer any reports of this service being used.
Test: fast/sandbox/mac/sandbox-mach-lookup.html
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-22 Wenson Hsieh <wenson_hsieh@apple.com>
The web process should be killed after failing to decode display list items in the GPU process
https://bugs.webkit.org/show_bug.cgi?id=219097
<rdar://problem/71546526>
Reviewed by Chris Dumez.
Handle `StopReplayReason::InvalidItem` by terminating the web process via MESSAGE_CHECK. See below for more
details.
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::didReceiveInvalidMessage):
(WebKit::GPUConnectionToWebProcess::terminateWebProcess):
* GPUProcess/GPUConnectionToWebProcess.h:
Pull logic for terminating the web process out into a separate helper method on `GPUConnectionToWebProcess`, and
use this helper in `GPUConnectionToWebProcess::didReceiveInvalidMessage`, as well as in `RemoteRenderingBackend`
below.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
Add macro definitions for `MESSAGE_CHECK` and `MESSAGE_CHECK_WITH_RETURN_VALUE`. Since the methods on
`RemoteRenderingBackend` that trigger message checks all run on a background queue, the normal (main-thread) way
of defining these macros (`MESSAGE_CHECK_WITH_RETURN_VALUE_BASE` and `MESSAGE_CHECK_BASE`) don't work. We
instead call into the GPU to web process connection object directly to send a `TerminateWebProcess` message to
the parent (UI) process, and additionally log a given failure message.
(WebKit::RemoteRenderingBackend::nextDestinationImageBufferAfterApplyingDisplayLists):
Replace a number FIXMEs throughout these IPC message handlers with `MESSAGE_CHECK`s. Additionally, add a new
`MESSAGE_CHECK` for the case where we stopped replay early due to a corrupted or invalid display list item.
(WebKit::RemoteRenderingBackend::wakeUpAndApplyDisplayList):
(WebKit::RemoteRenderingBackend::setNextItemBufferToRead):
Simply remove the FIXME and early return here; this can only happen in the case where a compromised web content
process attempts to append redundant item buffer change items. However, doing so will either be (1) harmless,
since the pending item buffer information will just be overwritten, or (2) result in hitting a MESSAGE_CHECK
when decoding items if we try to execute the contents of another item buffer that has not been written to.
(WebKit::RemoteRenderingBackend::didCreateSharedDisplayListHandle):
* GPUProcess/graphics/RemoteRenderingBackend.h:
2021-01-22 Simon Fraser <simon.fraser@apple.com>
[iOS WK2] Make the "in stable state" bit in visible content rect updates more fine-grained
https://bugs.webkit.org/show_bug.cgi?id=220875
Reviewed by Tim Horton.
VisibleContentRectUpdateInfo contains a single "in stable state" bit, but in a future
patch the web process will need to know if the UI process is rubberbanding. In addition,
we can fold "isChangingObscuredInsetsInteractively" into the same state bits.
So give VisibleContentRectUpdateInfo a OptionSet<ViewStabilityFlag>.
To give the correct state for non-main UIScrollViews, have _scrollViewIsRubberBanding
take the UIScrollView in question, and only include isChangingObscuredInsetsInteractively
if the question is being asked of the main UIScrollView.
* Shared/VisibleContentRectUpdateInfo.cpp:
(WebKit::VisibleContentRectUpdateInfo::encode const):
(WebKit::VisibleContentRectUpdateInfo::decode):
(WebKit::operator<<):
* Shared/VisibleContentRectUpdateInfo.h:
(WebKit::VisibleContentRectUpdateInfo::VisibleContentRectUpdateInfo):
(WebKit::VisibleContentRectUpdateInfo::inStableState const):
(WebKit::VisibleContentRectUpdateInfo::viewStability const):
(WebKit::VisibleContentRectUpdateInfo::isFirstUpdateForNewViewSize const):
(WebKit::operator==):
(WebKit::VisibleContentRectUpdateInfo::isChangingObscuredInsetsInteractively const): Deleted.
* UIProcess/API/Cocoa/WKWebViewInternal.h:
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _didRelaunchProcess]):
(-[WKWebView _scrollViewIsRubberBanding:]):
(-[WKWebView _viewStabilityState:]):
(-[WKWebView _scheduleVisibleContentRectUpdateAfterScrollInView:]):
(-[WKWebView _updateVisibleContentRects]):
(-[WKWebView _scrollViewIsRubberBanding]): Deleted.
(-[WKWebView _scrollViewIsInStableState:]): Deleted.
* UIProcess/ios/WKContentView.h:
* UIProcess/ios/WKContentView.mm:
(-[WKContentView didUpdateVisibleRect:unobscuredRect:contentInsets:unobscuredRectInScrollViewCoordinates:obscuredInsets:unobscuredSafeAreaInsets:inputViewBounds:scale:minimumScale:viewStability:enclosedInScrollableAncestorView:sendEvenIfUnchanged:]):
(-[WKContentView didUpdateVisibleRect:unobscuredRect:contentInsets:unobscuredRectInScrollViewCoordinates:obscuredInsets:unobscuredSafeAreaInsets:inputViewBounds:scale:minimumScale:inStableState:isChangingObscuredInsetsInteractively:enclosedInScrollableAncestorView:sendEvenIfUnchanged:]): Deleted.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::updateVisibleContentRects):
2021-01-22 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] No error is visible after user enters the incorrect PIN
https://bugs.webkit.org/show_bug.cgi?id=220839
<rdar://problem/73378319>
Reviewed by Brent Fulgham.
To fix that, let's have a boolean to indicate the PIN entry UI has already been presented within this session.
And then rely on the error UI to let users retry the PIN.
Covered by manual tests.
* UIProcess/WebAuthentication/Cocoa/AuthenticatorPresenterCoordinator.h:
* UIProcess/WebAuthentication/Cocoa/AuthenticatorPresenterCoordinator.mm:
(WebKit::AuthenticatorPresenterCoordinator::requestPin):
(WebKit::AuthenticatorPresenterCoordinator::selectAssertionResponse):
Besides above, this patch also fixes some crashes.
2021-01-22 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Should only check default browsers for the Modern WebAuthn
https://bugs.webkit.org/show_bug.cgi?id=220840
<rdar://problem/73375937>
Reviewed by Brent Fulgham.
* WebProcess/WebAuthentication/WebAuthenticatorCoordinator.cpp:
(WebKit::WebAuthenticatorCoordinator::makeCredential):
(WebKit::WebAuthenticatorCoordinator::getAssertion):
(WebKit::WebAuthenticatorCoordinator::isUserVerifyingPlatformAuthenticatorAvailable):
2021-01-21 Peng Liu <peng.liu6@apple.com>
PiP video subtitles stop updating when Safari is backgrounded
https://bugs.webkit.org/show_bug.cgi?id=220660
Reviewed by Darin Adler.
Subtitles in the picture-in-picture window will stop updating when the browser is
in the background because we freeze the layer tree when a browser is in the background.
This patch fixes this issue by avoiding freezing the layer tree if a video is playing
in picture-in-picture when the browser is in the background.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updateDrawingAreaLayerTreeFreezeState):
* WebProcess/cocoa/VideoFullscreenManager.h:
* WebProcess/cocoa/VideoFullscreenManager.mm:
(WebKit::VideoFullscreenManager::videoInPictureInPicture const):
(WebKit::VideoFullscreenManager::enterVideoFullscreenForVideoElement):
(WebKit::VideoFullscreenManager::exitVideoFullscreenForVideoElement):
(WebKit::VideoFullscreenManager::exitVideoFullscreenToModeWithoutAnimation):
2021-01-21 Alex Christensen <achristensen@webkit.org>
Add experimental feature to use network loader
https://bugs.webkit.org/show_bug.cgi?id=220521
<rdar://problem/69394713>
Reviewed by Geoff Garen.
Manually verified that the setting is correctly appearing in Safari and being given to CFNetwork.
* NetworkProcess/NetworkSessionCreationParameters.cpp:
(WebKit::NetworkSessionCreationParameters::encode const):
(WebKit::NetworkSessionCreationParameters::decode):
* NetworkProcess/NetworkSessionCreationParameters.h:
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::NetworkSessionCocoa):
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::experimentalFeatureEnabled):
(WebKit::WebsiteDataStore::http3Enabled):
(WebKit::WebsiteDataStore::useNetworkLoader):
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::parameters):
* UIProcess/WebsiteData/WebsiteDataStore.h:
2021-01-21 Aditya Keerthi <akeerthi@apple.com>
Unreviewed, reverting r271691.
https://bugs.webkit.org/show_bug.cgi?id=220812
Introduced crash, and failing API tests.
Reverted changeset:
"[macOS] Titlebar separator doesn't show when WKWebView is scrolled"
https://bugs.webkit.org/show_bug.cgi?id=220633
https://trac.webkit.org/changeset/271691
2021-01-21 Chris Dumez <cdumez@apple.com>
[GPUProcess][iOS] https://foobar404.github.io/Wave.js/#/ does not output any audio
https://bugs.webkit.org/show_bug.cgi?id=220802
<rdar://problem/73455820>
Reviewed by Per Arne Vollan.
Allow in the sandbox a few things that were needed to use AudioSourceProvider in the GPUProcess.
* Resources/SandboxProfiles/ios/com.apple.WebKit.GPU.sb:
2021-01-21 Per Arne Vollan <pvollan@apple.com>
[macOS] The development WebContent XPC service crashes under WebKit::AuxiliaryProcess::initializeSandbox
https://bugs.webkit.org/show_bug.cgi?id=220777
<rdar://problem/73382584>
Reviewed by Brent Fulgham.
This happens because the development XPC service does not have the required private entitlement to enable message filtering in the sandbox,
and therefore fails to apply the sandbox. The existing runtime switching to enable message filtering based on the presence of the entitlement
is not working correctly. The 'if' clause checking the sandbox parameter in the sandbox cannot be nested inside an 'allow' clause, it appears.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::initializeSandbox):
* WebProcess/com.apple.WebProcess.sb.in:
2021-01-21 Sam Weinig <weinig@apple.com>
Add experimental support for separated layers
https://bugs.webkit.org/show_bug.cgi?id=220734
Reviewed by Simon Fraser.
Plumb a "separated" bit through the graphics layer infrastructure. Currently unused
but will be used for optimization experimentation.
* Shared/RemoteLayerTree/RemoteLayerTreePropertyApplier.mm:
(WebKit::RemoteLayerTreePropertyApplier::applyPropertiesToLayer):
* Shared/RemoteLayerTree/RemoteLayerTreeTransaction.h:
* Shared/RemoteLayerTree/RemoteLayerTreeTransaction.mm:
(WebKit::RemoteLayerTreeTransaction::LayerProperties::LayerProperties):
(WebKit::RemoteLayerTreeTransaction::LayerProperties::encode const):
(WebKit::RemoteLayerTreeTransaction::LayerProperties::decode):
(WebKit::dumpChangedLayers):
* WebProcess/WebPage/RemoteLayerTree/PlatformCALayerRemote.cpp:
(WebKit::PlatformCALayerRemote::isSeparated const):
(WebKit::PlatformCALayerRemote::setSeparated):
* WebProcess/WebPage/RemoteLayerTree/PlatformCALayerRemote.h:
2021-01-21 Sam Weinig <weinig@apple.com>
Rework color clamping logic to be more consistent and clear
https://bugs.webkit.org/show_bug.cgi?id=220716
Reviewed by Simon Fraser.
* UIProcess/API/wpe/WebKitColor.cpp:
(webkitColorToWebCoreColor):
Update to use new convertTo<> functions.
2021-01-21 Aditya Keerthi <akeerthi@apple.com>
[macOS] Titlebar separator doesn't show when WKWebView is scrolled
https://bugs.webkit.org/show_bug.cgi?id=220633
<rdar://problem/71094055>
Reviewed by Tim Horton.
Starting in Big Sur, NSWindows with a titlebar display a separator if
there is a scrolled NSScrollView adjacent to the titlebar. Since
WKWebViews are scrollable views, but not backed by NSScrollView, we
need to adopt SPI to support this functionality.
This patch updates WKWebView to conform to the NSScrollViewSeparatorTrackingAdapter
protocol, ensuring the titlebar separator is displayed when
necessary. Note that since WKWebViews are not actually NSScrollView's we
don't already have the scroll position of the view in the UIProcess. To
determine whether or not the view is scrolled, this patch adds plumbing
so that the WebProcess can tell the UIProcess the new scroll position
when a page is scrolled.
Tests: WKWebViewTitlebarSeparatorTests.BackForwardCache
WKWebViewTitlebarSeparatorTests.ChangeTitlebarAdjacency
WKWebViewTitlebarSeparatorTests.ChangeViewVisibility
WKWebViewTitlebarSeparatorTests.NavigationResetsTitlebarAppearance
WKWebViewTitlebarSeparatorTests.ScrollWithTitlebarAdjacency
WKWebViewTitlebarSeparatorTests.ScrollWithoutTitlebarAdjacency
* Platform/spi/mac/AppKitSPI.h:
* UIProcess/API/mac/WKView.mm:
(-[WKView scrollViewFrame]):
(-[WKView hasScrolledContentsUnderTitlebar]):
(-[WKView _web_registerScrollViewSeparatorTrackingAdapter]):
(-[WKView _web_unregisterScrollViewSeparatorTrackingAdapter]):
* UIProcess/API/mac/WKWebViewMac.mm:
(-[WKWebView scrollViewFrame]):
(-[WKWebView hasScrolledContentsUnderTitlebar]):
(-[WKWebView _web_registerScrollViewSeparatorTrackingAdapter]):
(-[WKWebView _web_unregisterScrollViewSeparatorTrackingAdapter]):
* UIProcess/Cocoa/WebViewImpl.h:
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::updateWindowAndViewFrames):
If the WKWebView's frame changes, update the titlebar adjacency
state and notify observers of the change.
(WebKit::WebViewImpl::viewWillMoveToWindowImpl):
Unregister the WKWebView as an NSScrollViewSeparatorTrackingAdapter if
it is removed from the window.
(WebKit::WebViewImpl::viewDidHide):
Hidden views are not adjacent to the titlebar.
(WebKit::WebViewImpl::viewDidUnhide):
An unhidden view may be adjacent to the titlebar.
(WebKit::WebViewImpl::pageDidScroll):
Use the scroll position of the page to determine whether or not the
WKWebView is scrolled.
(WebKit::WebViewImpl::scrollViewFrame):
Needed to conform to NSScrollViewSeparatorTrackingAdapter.
(WebKit::WebViewImpl::hasScrolledContentsUnderTitlebar):
Needed to conform to NSScrollViewSeparatorTrackingAdapter. Returns true
if the view is registered as an NSScrollViewSeparatorTrackingAdapter
and is scrolled.
(WebKit::WebViewImpl::updateTitlebarAdjacencyState):
The WKWebView needs to be registered as an NSScrollViewSeparatorTrackingAdapter
if it's adjacent to the titlebar and unregistered otherwise.
* UIProcess/PageClient.h:
(WebKit::PageClient::pageDidScroll):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::pageDidScroll):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/mac/PageClientImplMac.h:
* UIProcess/mac/PageClientImplMac.mm:
(WebKit::PageClientImpl::didCommitLoadForMainFrame):
Reset the scroll position upon navigation, as pageDidScroll does not get
called when navigating.
(WebKit::PageClientImpl::pageDidScroll):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::pageDidScroll):
Pass the current scroll position when the page is scrolled, so that the
UIProcess knows whether or not the page has a non-zero scroll position.
2021-01-21 Dave Blanchard <dave@grow.game>
Missing header gtk/gtk.h in PointerLockManagerX11.cpp
https://bugs.webkit.org/show_bug.cgi?id=220152
Reviewed by Carlos Garcia Campos.
* UIProcess/gtk/PointerLockManagerX11.cpp:
2021-01-20 Wenson Hsieh <wenson_hsieh@apple.com>
[macOS] Include an origin identifier when writing promised image data to the drag pasteboard
https://bugs.webkit.org/show_bug.cgi?id=220782
Reviewed by Megan Gardner.
Add plumbing in WebKit to serialize and write the pasteboard origin identifier via custom data when dragging an
image with promised data on macOS. In the case where we're dragging an image into a contenteditable area in a
document with the *same* pasteboard origin, this allows us to avoid sanitizing the web archive data upon
handling the drop.
Test: DragAndDropTests.ProvideImageDataForMultiplePasteboards
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::setPromisedDataForImage):
* UIProcess/Cocoa/WebViewImpl.h:
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::setPromisedDataForImage):
If specified, write the origin to the drag pasteboard by creating a new `PasteboardCustomData` object, setting
its origin, and then serializing the custom data object into an `NSData`.
* UIProcess/PageClient.h:
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/mac/PageClientImplMac.h:
* UIProcess/mac/PageClientImplMac.mm:
(WebKit::PageClientImpl::setPromisedDataForImage):
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::setPromisedDataForImage):
* WebProcess/WebCoreSupport/mac/WebDragClientMac.mm:
(WebKit::WebDragClient::declareAndWriteDragImage):
2021-01-20 Kenneth Russell <kbr@chromium.org>
Support WEBGL_multi_draw extension
https://bugs.webkit.org/show_bug.cgi?id=219139
Reviewed by Dean Jackson.
Add automatically-generated stubs for supporting the multi-draw
extension in the GPU process.
* GPUProcess/graphics/RemoteGraphicsContextGL.messages.in:
* GPUProcess/graphics/RemoteGraphicsContextGLFunctionsGenerated.h:
(multiDrawArraysANGLE):
(multiDrawArraysInstancedANGLE):
(multiDrawElementsANGLE):
(multiDrawElementsInstancedANGLE):
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.h:
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxyFunctionsGenerated.cpp:
(WebKit::RemoteGraphicsContextGLProxy::multiDrawArraysANGLE):
(WebKit::RemoteGraphicsContextGLProxy::multiDrawArraysInstancedANGLE):
(WebKit::RemoteGraphicsContextGLProxy::multiDrawElementsANGLE):
(WebKit::RemoteGraphicsContextGLProxy::multiDrawElementsInstancedANGLE):
2021-01-20 Chris Dumez <cdumez@apple.com>
[GPUProcess] Mark IOSurface backing for RemoteGraphicsContextGL's displayBuffer as owned by the WebProcess
https://bugs.webkit.org/show_bug.cgi?id=220770
Reviewed by Geoffrey Garen.
Mark IOSurface backing for RemoteGraphicsContextGL's displayBuffer as owned by the WebProcess. This is so
that Jetsam knows which process to terminate to reclaim memory.
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.cpp:
(WebKit::RemoteGraphicsContextGLProxy::prepareForDisplay):
* WebProcess/GPU/graphics/cocoa/ImageBufferShareableMappedIOSurfaceBackend.cpp:
(WebKit::ImageBufferShareableMappedIOSurfaceBackend::create):
2021-01-20 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] "touchend" events should be dispatched and handled asynchronously
https://bugs.webkit.org/show_bug.cgi?id=220256
<rdar://problem/64912551>
Reviewed by Tim Horton.
Allow active "touchend" event listeners to send asynchronous (non-blocking) IPC messages to the web process by
augmenting the content view's deferring gesture recognizers. See below for more details.
Test: fast/events/touch/ios/click-event-after-touchend.html
* UIProcess/PageClient.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::handlePreventableTouchEvent):
Send both "touchstart" and "touchend" events asynchronously.
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::isHandlingPreventableTouchEnd const):
In a similar way to how we defer "touchstart", add a counter to keep track of when we're handling an active
(preventable) touchend, such that we can defer all native gestures that can be prevented by touchend until after
the page is done handling the event.
* UIProcess/ios/PageClientImplIOS.h:
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::doneDeferringTouchEnd):
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView setUpInteraction]):
Add a new set of deferring gesture recognizers that can be used to defer all tap gestures on the content view,
due to active "touchend" event listeners. This works because currently, the only default gestures in WebKit
that can be prevented by calling `preventDefault()` on "touchend" events are tap gestures.
(-[WKContentView cleanUpInteraction]):
(-[WKContentView _removeDefaultGestureRecognizers]):
(-[WKContentView _addDefaultGestureRecognizers]):
(-[WKContentView _webTouchEventsRecognized:]):
(-[WKContentView _deferringGestures]):
(-[WKContentView _touchEndDeferringGestures]):
(-[WKContentView _doneDeferringTouchEnd:]):
Add a hook to "lift" the gesture gate upon touchend, which allows all deferred tap gestures to either fail (in
the case where the page prevented default), or recognize (in the case where the page did not prevent default).
(-[WKContentView gestureRecognizer:shouldRecognizeSimultaneouslyWithGestureRecognizer:]):
Replace `_touchStartDeferringGestures` with `_deferringGestures` in several call sites that need to iterate over
all deferring gestures on the content view.
(-[WKContentView _singleTapDidReset:]):
When resetting the single tap gesture, ensure that the tap highlight view is removed by calling
`-_finishInteraction`; this is because an active touchend event handler may now defer `-_singleTapRecognized:`
until after the ending the touch, which causes `-_showTapHighlight` to be called without a matching call to
remove the highlight. Since we know that the synthetic tap gesture must eventually be reset after recognition,
this is a more suitable catch-all for removing the tap highlight.
(-[WKContentView deferringGestureRecognizer:shouldDeferOtherGestureRecognizer:]):
For the "touchend" deferring gesture recognizers, partition all tap gestures on the content view into three
subgraphs that mirror the extant gesture subgraphs currently used for the "touchstart" deferring gestures. This
ensures that synthetic clicks always remain fast when the blocking double tap gesture is disabled, and that
other tap gestures (such as the text interaction multi-tap gesture) are still immediately reset.
2021-01-20 Jer Noble <jer.noble@apple.com>
[Mac] Netflix controls do not fade out after entering fullscreen
https://bugs.webkit.org/show_bug.cgi?id=220472
<rdar://70602577>
Reviewed by Eric Carlson.
Test: TestWebKitAPI/Tests/mac/FullscreenPointerLeave.mm
Netflix uses a "pointerleave" event over their playback controls to trigger setting their "fade out controls"
timer, but a "pointerleave" event isn't fired when the element moves out from under the pointer (either due to
style changes, or the window screen location moving due to entering fullscreen).
Work around this behavior by sending a synthetic mouse event after entering fullscreen, which causes the
"pointerleave" event to be fired if, indeed, the pointer is no longer over the element in question.
* UIProcess/mac/WKFullScreenWindowController.mm:
(-[WKFullScreenWindowController finishedEnterFullScreenAnimation:]):
2021-01-20 Kate Cheney <katherine_cheney@apple.com>
Safari says "Blocked Plug-in" instead showing a PDF
https://bugs.webkit.org/show_bug.cgi?id=220665
<rdar://problem/64372944>
Reviewed by Darin Adler.
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::shouldUsePDFPlugin const):
* WebProcess/WebCoreSupport/WebFrameLoaderClient.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::createPlugin):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::shouldUsePDFPlugin const):
2021-01-20 Michael Catanzaro <mcatanzaro@gnome.org>
[WPE][GTK] Ensure URI scheme handler is not registered multiple times
https://bugs.webkit.org/show_bug.cgi?id=220363
Reviewed by Carlos Garcia Campos.
* UIProcess/API/glib/WebKitWebContext.cpp:
(webkit_web_context_register_uri_scheme):
2021-01-20 Youenn Fablet <youenn@apple.com>
On page close, make sure to notify of capture state immediately
https://bugs.webkit.org/show_bug.cgi?id=220588
Reviewed by Geoffrey Garen.
If page is getting closed or reset (for instance in case of process swap), we immediately notify the client of the capture state,
instead of waiting for 3 seconds if capture just started.
Covered by API test.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::isPlayingMediaDidChange):
(WebKit::WebPageProxy::updatePlayingMediaDidChange):
(WebKit::WebPageProxy::updateReportedMediaCaptureState):
* UIProcess/WebPageProxy.h:
2021-01-19 Sihui Liu <sihui_liu@appe.com>
Update media state for active speech recognition as it uses audio capture
https://bugs.webkit.org/show_bug.cgi?id=220667
Reviewed by Youenn Fablet.
* WebProcess/WebCoreSupport/WebSpeechRecognitionConnection.cpp:
(WebKit::WebSpeechRecognitionConnection::unregisterClient):
* WebProcess/WebCoreSupport/WebSpeechRecognitionConnection.h:
2021-01-19 Chris Dumez <cdumez@apple.com>
[Hardening] Protect against overflows in ArgumentCoder<ArrayReference<T, arrayReferenceDynamicExtent>>::decode()
https://bugs.webkit.org/show_bug.cgi?id=220748
<rdar://73323644>
Reviewed by Geoff Garen.
Make sure `size` and `size * sizeof(T)` do not overflow.
* Platform/IPC/ArgumentCoders.h:
2021-01-19 Chris Dumez <cdumez@apple.com>
[GPUProcess] Claim ownership of Canvas IOSurface backing in the WebProcess
https://bugs.webkit.org/show_bug.cgi?id=220663
Reviewed by Geoffrey Garen.
Claim ownership of Canvas IOSurface backing in the WebProcess. If we don't do this, all
IOSurface memory gets attributed to the GPUProcess, making it a prime candidate for jetsam.
We need to attribute memory usage to the WebProcesses being served by the GPUProcess as much
as possible and this is a step in this direction.
Note that this patch currently only impacts Canvas IOSurface backing when "DOM Rendering in
GPUProcess" is disabled. When "DOM Rendering in GPUProcess" is enabled, we are not supposed
to use the IOSurface API in the WebProcess at all.
* WebProcess/GPU/graphics/cocoa/ImageBufferShareableMappedIOSurfaceBackend.cpp:
(WebKit::ImageBufferShareableMappedIOSurfaceBackend::create):
2021-01-19 Michael Catanzaro <mcatanzaro@gnome.org>
Syscall param sendmsg(msg.msg_iov[1]) points to uninitialised byte(s) in IPC::Connection::sendOutgoingMessage
https://bugs.webkit.org/show_bug.cgi?id=220668
Reviewed by Darin Adler.
We previously discovered that we need to either (a) manually zero the MessageInfo struct
with memcpy (accepted solution), or b) use __attribute__((packed)) to omit struct packing.
This seemed to fix our valgrind warnings when bmalloc was in use, but not when bmalloc is
disabled, as is good practice when valgrinding.
Turns out we need to zero (or pack) the AttachmentInfo struct as well. Let's go with zeroing
it, since all its members are trivially copyable, so that's safe to do.
For good measure, we'll implement a copy constructor and do so there as well. Do this for
MessageInfo too, to round out the original fix to MessageInfo.
Since AttachmentInfo and MessageInfo are no longer trivially-copyable, we now need to cast
them to void* when using memcpy in order to avoid -Wclass-memaccess warnings from GCC.
* Platform/IPC/unix/ConnectionUnix.cpp:
(IPC::AttachmentInfo::AttachmentInfo):
(): Deleted.
* Platform/IPC/unix/UnixMessage.h:
(IPC::MessageInfo::MessageInfo):
(IPC::UnixMessage::UnixMessage):
2021-01-19 Alex Christensen <achristensen@webkit.org>
REGRESSION(r269865) Mail's context menu when right clicking on a link does not contain correct entries
https://bugs.webkit.org/show_bug.cgi?id=220745
Reviewed by Tim Horton.
* Shared/API/c/WKContextMenuItem.cpp:
Add a bunch of static_asserts to verify that our binary values correspond to each other and don't change.
* Shared/API/c/WKContextMenuItemTypes.h:
2021-01-19 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Polish the new WebAuthn UI
https://bugs.webkit.org/show_bug.cgi?id=220617
<rdar://problem/73185470>
Reviewed by Brent Fulgham.
This patch does the following few things:
1. It updates the way how the PIN error for security keys is handled.
2. It uses the credential name to identify a credential that passed to the UI instead of the login choice object
as it turns out that the UI won't return the same object at all.
3. It delays to show the UI if the platform authenticator is involved given the platform authenticator might not contain
the requested credentials. If not, we should either show an error or just requesting the security key ones.
Covered by manual tests.
* Platform/spi/Cocoa/AuthenticationServicesCoreSPI.h:
(NS_ERROR_ENUM):
* UIProcess/WebAuthentication/Cocoa/AuthenticatorPresenterCoordinator.h:
* UIProcess/WebAuthentication/Cocoa/AuthenticatorPresenterCoordinator.mm:
(WebKit::AuthenticatorPresenterCoordinator::AuthenticatorPresenterCoordinator):
(WebKit::AuthenticatorPresenterCoordinator::updatePresenter):
(WebKit::AuthenticatorPresenterCoordinator::selectAssertionResponse):
(WebKit::AuthenticatorPresenterCoordinator::didSelectAssertionResponse):
* UIProcess/WebAuthentication/Cocoa/WKASCAuthorizationPresenterDelegate.mm:
(-[WKASCAuthorizationPresenterDelegate authorizationPresenter:credentialRequestedForLoginChoice:authenticatedContext:completionHandler:]):
2021-01-19 Michael Catanzaro <mcatanzaro@gnome.org>
[WPE][GTK] Remove unnecessary child setup function from process launcher
https://bugs.webkit.org/show_bug.cgi?id=220090
Reviewed by Carlos Garcia Campos.
The process launcher currently uses a child setup function to close the server end of its
IPC socket. But this is totally unnecessary because this socket always uses CLOEXEC. So the
child setup function is redundant.
This is step one towards making Eclipse not crash when its UI process JVM is using a huge
amount of memory. In such conditions, fork() will fail due to OOM, but posix_spawn() will
still succeed. A child setup function forces GLib to use fork() instead of posix_spawn().
However, this commit is not sufficient to fix Eclipse, because GSubprocess itself sets its
own child setup function.
* UIProcess/Launcher/glib/ProcessLauncherGLib.cpp:
(WebKit::ProcessLauncher::launchProcess):
(WebKit::childSetupFunction): Deleted.
2021-01-19 Lauro Moura <lmoura@igalia.com>
REGRESSION(r271580) [GTK] LTS/Debian build failure due to unsupported g-ir-scanner parameter
https://bugs.webkit.org/show_bug.cgi?id=220730
Reviewed by Philippe Normand.
No new behavior. No new tests.
* PlatformGTK.cmake: Expose --sources-top-dirs only if available.
2021-01-19 Keith Rollin <krollin@apple.com>
Modularize WebKitLegacy.framework
https://bugs.webkit.org/show_bug.cgi?id=220026
<rdar://57173237>
Reviewed by Darin Adler.
Add .modulemap files to WebKitLegacy to help speed up client builds.
For MigrateHeadersFromWebKitLegacy.make, remove the newly-created
umbrella header files from WEBKIT_LEGACY_PRIVATE_HEADERS in order to
avoid some processing that the non-umbrella headers receive.
* mac/MigrateHeadersFromWebKitLegacy.make:
2021-01-19 Chris Dumez <cdumez@apple.com>
Unreviewed, drop remaining main-thread assertion in ShareableBitmap that I missed in r271533.
This assertion is hitting on the debug GPUProcess bot. SharedBitmap subclasses ThreadSafeRefCounted
as of r271533 and it is expected that it is used from non-main threads.
* Shared/cg/ShareableBitmapCG.cpp:
(WebKit::ShareableBitmap::createGraphicsContext):
2021-01-19 Rob Buis <rbuis@igalia.com>
Null check ResourceRequest IPC decode
https://bugs.webkit.org/show_bug.cgi?id=220400
Reviewed by Youenn Fablet.
Null check ResourceRequest IPC decode since
the IPC::decode call can succeed yet but set
the dictionary to null.
Test: ipc/web-authenticator-get-assertion.html
* Shared/mac/WebCoreArgumentCodersMac.mm:
(IPC::ArgumentCoder<WebCore::ResourceRequest>::decodePlatformData):
2021-01-18 Michael Catanzaro <mcatanzaro@gnome.org>
[GTK] Multilib conflicts in gir files
https://bugs.webkit.org/show_bug.cgi?id=220636
Reviewed by Carlos Garcia Campos.
By passing --sources-top-dirs to g-ir-scanner, we ensure all paths in the generated gir
files are relative to the toplevel builddir. This avoids multilib conflicts when the
builddir contains the target architecture.
* PlatformGTK.cmake:
2021-01-17 Yusuke Suzuki <ysuzuki@apple.com>
JITOperation does not exist in WebKit
https://bugs.webkit.org/show_bug.cgi?id=220686
Reviewed by Mark Lam.
Now, we do not register host functions as JIT operations. As a result, WebKit.framework does not have any JITOperations.
We should remove population of JITOperations code in WebKit.framework.
* Shared/API/c/WKString.cpp:
(WKStringCopyJSString):
* Shared/Cocoa/WebKit2InitializeCocoa.mm:
(WebKit::runInitializationCode):
* Shared/WebKit2Initialize.cpp:
(WebKit::InitializeWebKit2):
* Shared/WebKitJITOperations.cpp: Removed.
* Sources.txt:
* WebKit.xcodeproj/project.pbxproj:
2021-01-17 Fujii Hironori <Hironori.Fujii@sony.com>
[non-Cocoa] Add a new AuxiliaryProcessMainBase class for auxiliary processes without singleton()
https://bugs.webkit.org/show_bug.cgi?id=220615
Reviewed by Don Olmstead.
non-Cocoa ports auxiliary processes are using AuxiliaryProcessMain
as the entry points. AuxiliaryProcessMain supports both kinds of
auxiliary processes with and without singleton() method by using
initializeAuxiliaryProcess template function. However, all
initializeAuxiliaryProcess look similar code. They can share more
code.
Added a AuxiliaryProcessMainBaseNoSingleton template class for
auxiliary processes without singleton().
Moved the code that was in AuxiliaryProcessMain to
AuxiliaryProcessMainBase::run() to remove
takeInitializationParameters().
* GPUProcess/gstreamer/GPUProcessMainGStreamer.cpp:
(WebKit::GPUProcessMain):
(WebKit::initializeAuxiliaryProcess<GPUProcess>): Deleted.
* GPUProcess/playstation/GPUProcessMainPlayStation.cpp:
(WebKit::GPUProcessMain):
(WebKit::initializeAuxiliaryProcess<GPUProcess>): Deleted.
* GPUProcess/win/GPUProcessMainWin.cpp:
(WebKit::GPUProcessMain):
(WebKit::initializeAuxiliaryProcess<GPUProcess>): Deleted.
* NetworkProcess/curl/NetworkProcessMainCurl.cpp:
(WebKit::NetworkProcessMain):
(WebKit::initializeAuxiliaryProcess<NetworkProcess>): Deleted.
* NetworkProcess/soup/NetworkProcessMainSoup.cpp:
(WebKit::NetworkProcessMain):
(WebKit::initializeAuxiliaryProcess<NetworkProcess>): Deleted.
* Shared/AuxiliaryProcessMain.h:
(WebKit::AuxiliaryProcessMainBase::platformInitialize):
(WebKit::AuxiliaryProcessMainBase::platformFinalize):
(WebKit::AuxiliaryProcessMainBase::initializeAuxiliaryProcess):
(WebKit::AuxiliaryProcessMainBase::run):
(WebKit::AuxiliaryProcessMainBaseNoSingleton::process):
(WebKit::AuxiliaryProcessMain):
(WebKit::AuxiliaryProcessMainBase::initializationParameters): Deleted.
(WebKit::AuxiliaryProcessMainBase::takeInitializationParameters): Deleted.
(WebKit::initializeAuxiliaryProcess): Deleted.
* Shared/unix/AuxiliaryProcessMain.cpp:
(WebKit::AuxiliaryProcessMainCommon::parseCommandLine):
(WebKit::AuxiliaryProcessMainBase::parseCommandLine): Deleted.
* Shared/win/AuxiliaryProcessMainWin.cpp:
(WebKit::AuxiliaryProcessMainCommon::parseCommandLine):
(WebKit::AuxiliaryProcessMainBase::parseCommandLine): Deleted.
* WebProcess/gtk/WebProcessMainGtk.cpp:
(WebKit::WebProcessMain):
* WebProcess/playstation/WebProcessMainPlayStation.cpp:
(WebKit::WebProcessMain):
* WebProcess/win/WebProcessMainWin.cpp:
(WebKit::WebProcessMain):
* WebProcess/wpe/WebProcessMainWPE.cpp:
(WebKit::WebProcessMain):
2021-01-17 Youenn Fablet <youenn@apple.com>
Add CoreMedia aavds entitlement to WebContent and GPU processes
https://bugs.webkit.org/show_bug.cgi?id=220238
Reviewed by Brent Fulgham.
Manually tested.
* Scripts/process-entitlements.sh:
2021-01-15 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] Emoji keyboard covers text field on twitter.com/messages
https://bugs.webkit.org/show_bug.cgi?id=220664
<rdar://problem/68400471>
Reviewed by Devin Rousso.
After iOS 14, the emoji software keyboard layout now includes a search field that can be used to filter for
specific emojis. This slightly increases the overall height of the software keyboard when the emoji keyplane is
active; in turn, this means that if the selection or caret is positioned right above the top of the software
keyboard when the normal (alphabetic) keyplane is active, switching to the emoji keyplane will cause the
keyboard to overlap the selection, making it difficult to see inserted text.
To address this, add a mechanism to detect when a change in the bounds of the software keyboard causes a visible
selection or caret rect to become overlapped, and react by scrolling to keep the selection visible. This has the
effect of fixing this bug by scrolling to reveal the text field after switching to the emoji keyboard, but it
also has the effect of scrolling to keep the selection visible after detaching a connected hardware keyboard,
in the case where it would've otherwise been overlapped by the (much taller) software keyboard that appears.
Test: editing/selection/ios/scroll-to-reveal-selection-when-showing-software-keyboard.html
* UIProcess/API/ios/WKWebViewIOS.h:
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _selectionRectIsFullyVisibleAndNonEmpty]):
Add an internal helper to check whether the selection bounds are fully visible.
(-[WKWebView _scrollToRevealSelectionIfNeeded]):
(-[WKWebView _zoomToFocusRect:selectionRect:fontSize:minimumScale:maximumScale:allowScaling:forceScroll:]):
(-[WKWebView _keyboardChangedWithInfo:adjustScrollView:]):
In the case where changing input view bounds causes a previously visible selection to become overlapped, call
`-_scrollToRevealSelectionIfNeeded` to make the selection visible again.
(-[WKWebView _zoomToFocusRect:selectionRect:insideFixed:fontSize:minimumScale:maximumScale:allowScaling:forceScroll:]): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/ios/WKContentView.h:
* UIProcess/ios/WKContentView.mm:
(-[WKContentView _zoomToFocusRect:selectionRect:fontSize:minimumScale:maximumScale:allowScaling:forceScroll:]):
(-[WKContentView _zoomToFocusRect:selectionRect:insideFixed:fontSize:minimumScale:maximumScale:allowScaling:forceScroll:]): Deleted.
Drive-by fix: remove the unused `insideFixed:` parameter from this adjacent method.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView rectToRevealWhenZoomingToFocusedElement]):
(-[WKContentView _zoomToRevealFocusedElement]):
(rectToRevealWhenZoomingToFocusedElement): Deleted.
Pull this into the `-rectToRevealWhenZoomingToFocusedElement` internal helper method instead, and use the new
`selectionBoundingRectInRootViewCoordinates` method below.
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::selectionBoundingRectInRootViewCoordinates const):
Pull out code to compute the selection bounding rect (for both ranged and caret selections) into a method on
`WebPageProxy`, so that it can be used in `WKContentView` and `WKWebView`.
2021-01-15 Alex Christensen <achristensen@webkit.org>
REGRESSION(r267763) Uploading zipped directories does not work on iOS
https://bugs.webkit.org/show_bug.cgi?id=220666
<rdar://problem/72940505>
Reviewed by Geoffrey Garen.
r267763 removed indexedDatabaseTempBlobDirectoryExtensionHandle with no replacement, which used to give the network process
read/write access to /tmp inside the parent process's container. This seems to have been unused for IndexedDB, but it was used
by createTemporaryZipArchive when uploading directories, such as Pages, Numbers, and Keynote documents.
Unfortunately the unit test added by r248139 is macOS-only because WKOpenPanelParameters is only available on macOS and it would
require a large and risky amount of refactoring to add SPI on iOS to test this because iOS uses WKFileUploadPanel instead.
I did manually verify that the bug is fixed using my phone, though.
* NetworkProcess/NetworkProcessCreationParameters.cpp:
(WebKit::NetworkProcessCreationParameters::encode const):
(WebKit::NetworkProcessCreationParameters::decode):
* NetworkProcess/NetworkProcessCreationParameters.h:
* NetworkProcess/cocoa/NetworkProcessCocoa.mm:
(WebKit::NetworkProcess::platformInitializeNetworkProcessCocoa):
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::sendCreationParametersToNewProcess):
2021-01-15 Chris Dumez <cdumez@apple.com>
[GPUProcess] Move DOM / Canvas rendering off the main thread in the GPUProcess
https://bugs.webkit.org/show_bug.cgi?id=220476
Reviewed by Simon Fraser.
Move DOM / Canvas rendering off the main thread (and on a high-priority serial WorkQueue) in the
GPUProcess by making RemoteRenderingBackend a WorkQueueMessageReceiver. There is a serial WorkQueue
per RemoteRenderingBackend, which means each WebPage gets its own WorkQueue.
A/B testing shows this mostly perf-neutral (could be a slight progression on some hardware). I believe
this new architecture should also give us more optimization opportunities in the future. For example,
we may be able to more aggressively wait on the cross-process semaphore when waiting for new DisplayList
entries without worrying about blocking the main thread.
I have run layout tests in Debug on both macOS & iOS with the GPUProcess enabled and I believe I
have fixed all the crashes that were discovered.
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::GPUConnectionToWebProcess):
(WebKit::GPUConnectionToWebProcess::didClose):
(WebKit::GPUConnectionToWebProcess::createRenderingBackend):
(WebKit::GPUConnectionToWebProcess::RemoteRenderingBackendWrapper::RemoteRenderingBackendWrapper):
(WebKit::GPUConnectionToWebProcess::RemoteRenderingBackendWrapper::~RemoteRenderingBackendWrapper):
* GPUProcess/GPUConnectionToWebProcess.h:
(WebKit::GPUConnectionToWebProcess::remoteMediaPlayerManagerProxy):
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::create):
(WebKit::RemoteRenderingBackend::RemoteRenderingBackend):
(WebKit::RemoteRenderingBackend::~RemoteRenderingBackend):
(WebKit::RemoteRenderingBackend::disconnect):
(WebKit::RemoteRenderingBackend::messageSenderConnection const):
(WebKit::RemoteRenderingBackend::applyMediaItem):
(WebKit::RemoteRenderingBackend::createImageBuffer):
(WebKit::RemoteRenderingBackend::wakeUpAndApplyDisplayList):
(WebKit::RemoteRenderingBackend::getImageData):
(WebKit::RemoteRenderingBackend::getDataURLForImageBuffer):
(WebKit::RemoteRenderingBackend::getDataForImageBuffer):
(WebKit::RemoteRenderingBackend::getBGRADataForImageBuffer):
(WebKit::RemoteRenderingBackend::cacheNativeImage):
(WebKit::RemoteRenderingBackend::cacheFont):
(WebKit::RemoteRenderingBackend::deleteAllFonts):
(WebKit::RemoteRenderingBackend::releaseRemoteResource):
(WebKit::RemoteRenderingBackend::didCreateSharedDisplayListHandle):
* GPUProcess/graphics/RemoteRenderingBackend.h:
* GPUProcess/graphics/RemoteRenderingBackend.messages.in:
* GPUProcess/media/RemoteLegacyCDMProxy.cpp:
(WebKit::RemoteLegacyCDMProxy::cdmMediaPlayer const):
* GPUProcess/media/RemoteMediaPlayerManagerProxy.cpp:
(WebKit::RemoteMediaPlayerManagerProxy::createMediaPlayer):
(WebKit::RemoteMediaPlayerManagerProxy::deleteMediaPlayer):
(WebKit::RemoteMediaPlayerManagerProxy::didReceivePlayerMessage):
(WebKit::RemoteMediaPlayerManagerProxy::didReceiveSyncPlayerMessage):
(WebKit::RemoteMediaPlayerManagerProxy::mediaPlayer):
* GPUProcess/media/RemoteMediaPlayerManagerProxy.h:
* Platform/SharedMemory.h:
* Shared/ShareableBitmap.cpp:
(WebKit::ShareableBitmap::ShareableBitmap):
(WebKit::ShareableBitmap::~ShareableBitmap):
* Shared/ShareableBitmap.h:
* Shared/cg/ShareableBitmapCG.cpp:
(WebKit::ShareableBitmap::createGraphicsContext):
(WebKit::ShareableBitmap::releaseBitmapContextData):
(WebKit::ShareableBitmap::releaseDataProviderData):
2021-01-15 Alex Christensen <achristensen@webkit.org>
Remove non-inclusively named WKBrowsingContextGroup SPI
https://bugs.webkit.org/show_bug.cgi?id=220604
Reviewed by Anders Carlsson.
Once rdar://problem/73167624 has been integrated, we can remove the old SPI, which has been replaced in r263099
* UIProcess/API/Cocoa/WKBrowsingContextGroup.h:
* UIProcess/API/Cocoa/WKBrowsingContextGroup.mm:
(-[WKBrowsingContextGroup addUserStyleSheet:baseURL:whitelistedURLPatterns:blacklistedURLPatterns:mainFrameOnly:]): Deleted.
(-[WKBrowsingContextGroup addUserScript:baseURL:whitelistedURLPatterns:blacklistedURLPatterns:injectionTime:mainFrameOnly:]): Deleted.
2021-01-15 Chris Dumez <cdumez@apple.com>
[GPUProcess] Improve the GPUProcess' memory pressure handler
https://bugs.webkit.org/show_bug.cgi?id=220659
Reviewed by Simon Fraser.
Improve the GPUProcess' memory pressure handler to clear things like the IOSurfacePool
and the SubimageCacheWithTimer.
* GPUProcess/GPUProcess.cpp:
(WebKit::GPUProcess::lowMemoryHandler):
(WebKit::GPUProcess::initializeGPUProcess):
(WebKit::GPUProcess::prepareToSuspend):
* GPUProcess/GPUProcess.h:
2021-01-15 Philippe Normand <pnormand@igalia.com>
[GStreamer] UI<->WebPocess IPC needed for enumerateDevices support
https://bugs.webkit.org/show_bug.cgi?id=220542
Reviewed by Eric Carlson.
For GLib ports the UIProcess will now send a message to the WebProcess to retrieve the
MediaStream devices. This is required because we want to avoid initializing GStreamer in the
UIProcess as much as possible.
* UIProcess/UserMediaPermissionRequestManagerProxy.cpp:
(WebKit::UserMediaPermissionRequestManagerProxy::platformGetMediaStreamDevices):
(WebKit::UserMediaPermissionRequestManagerProxy::computeFilteredDeviceList):
* UIProcess/UserMediaPermissionRequestManagerProxy.h:
* UIProcess/glib/UserMediaPermissionRequestManagerProxyGLib.cpp:
(WebKit::UserMediaPermissionRequestManagerProxy::platformGetMediaStreamDevices):
* WebProcess/glib/UserMediaCaptureManager.cpp:
(WebKit::UserMediaCaptureManager::validateUserMediaRequestConstraints):
(WebKit::UserMediaCaptureManager::getMediaStreamDevices):
* WebProcess/glib/UserMediaCaptureManager.h:
* WebProcess/glib/UserMediaCaptureManager.messages.in:
2021-01-14 Peng Liu <peng.liu6@apple.com>
Add a quirk to disable "return to element fullscreen from picture-in-picture" for some sites
https://bugs.webkit.org/show_bug.cgi?id=220606
Reviewed by Eric Carlson.
* UIProcess/Cocoa/VideoFullscreenManagerProxy.h:
* UIProcess/Cocoa/VideoFullscreenManagerProxy.messages.in:
* UIProcess/Cocoa/VideoFullscreenManagerProxy.mm:
(WebKit::VideoFullscreenManagerProxy::setupFullscreenWithID):
* UIProcess/WebFullScreenManagerProxy.cpp:
(WebKit::WebFullScreenManagerProxy::blocksReturnToFullscreenFromPictureInPicture const):
(WebKit::WebFullScreenManagerProxy::enterFullScreen):
* UIProcess/WebFullScreenManagerProxy.h:
* UIProcess/WebFullScreenManagerProxy.messages.in:
* UIProcess/ios/fullscreen/WKFullScreenWindowControllerIOS.mm:
(-[WKFullScreenWindowController enterFullScreen]):
(-[WKFullScreenWindowController didEnterPictureInPicture]):
* WebProcess/FullScreen/WebFullScreenManager.cpp:
(WebKit::WebFullScreenManager::enterFullScreenForElement):
* WebProcess/InjectedBundle/InjectedBundlePageFullScreenClient.cpp:
(WebKit::InjectedBundlePageFullScreenClient::enterFullScreenForElement):
* WebProcess/InjectedBundle/InjectedBundlePageFullScreenClient.h:
* WebProcess/cocoa/VideoFullscreenManager.mm:
(WebKit::VideoFullscreenManager::enterVideoFullscreenForVideoElement):
2021-01-14 Chris Dumez <cdumez@apple.com>
Unreviewed, apply post-landing review comments from Darin for r271495.
* Shared/cg/ShareableBitmapCG.cpp:
(WebKit::ShareableBitmap::calculateBytesPerRow):
(WebKit::ShareableBitmap::createGraphicsContext):
(WebKit::ShareableBitmap::makeCGImageCopy):
2021-01-14 Peng Liu <peng.liu6@apple.com>
[Cocoa] Add a script to generate the "Feature Flags" plist file
https://bugs.webkit.org/show_bug.cgi?id=220080
Reviewed by Daniel Bates.
Add a script to combine a plist file from WebKitAdditions with the corresponding plist
file in this repo to generate a `WebKit.plist` required by the Feature Flags interface.
* FeatureFlags/WebKit-appletvos.plist: Copied from Source/WebKit/FeatureFlags/WebKit.plist.
* FeatureFlags/WebKit-ios.plist: Copied from Source/WebKit/FeatureFlags/WebKit.plist.
* FeatureFlags/WebKit-macos.plist: Copied from Source/WebKit/FeatureFlags/WebKit.plist.
* FeatureFlags/WebKit-watchos.plist: Renamed from Source/WebKit/FeatureFlags/WebKit.plist.
* Scripts/combine-feature-flags-plist.py: Added.
(main):
* Scripts/generate-feature-flags-plist.sh: Added.
* WebKit.xcodeproj/project.pbxproj:
2021-01-14 Tim Horton <timothy_horton@apple.com>
Exceptions thrown when invoking a <select> on an iPhone-idiom app running on macOS
https://bugs.webkit.org/show_bug.cgi?id=220628
<rdar://problem/70546547>
Reviewed by Wenson Hsieh.
iPhone-idiom apps running on macOS need to behave exactly like iPhone apps,
even in ways that might not totally make sense on the platform (e.g.
using input accessory views for <select> instead of popover) for two reasons:
- we want to maximize compatibility, so they should behave exactly as they do
on the only platform they're tested on
- other frameworks (mostly UIKit) enforce the use of iPhone UI in iPhone-idiom
apps, even on macOS
* Shared/UserInterfaceIdiom.mm:
(WebKit::currentUserInterfaceIdiomIsPadOrMac):
(WebKit::setCurrentUserInterfaceIdiomIsPadOrMac):
(WebKit::updateCurrentUserInterfaceIdiom):
(WebKit::userInterfaceIdiomIsPad): Deleted.
Simplify our UIUserInterfaceIdiom-caching code.
currentUserInterfaceIdiomIsPadOrMac now does what it says on the tin, globally;
previously on macOS it said YES for iPhone-idiom apps, but now it says NO,
like you would expect.
Use UIDevice everywhere instead of the soft-deprecated UICurrentUserInterfaceIdiomIsPad,
both because it is soft-deprecated, and because it lets us elide the static helper.
2021-01-14 Chris Dumez <cdumez@apple.com>
[Hardening] Protect against overflows in calculateBytesPerRow()
https://bugs.webkit.org/show_bug.cgi?id=220625
<rdar://73040899>
Reviewed by Alex Christensen.
Protect against overflows in calculateBytesPerRow().
* Shared/ShareableBitmap.h:
* Shared/cairo/ShareableBitmapCairo.cpp:
(WebKit::ShareableBitmap::calculateBytesPerPixel):
* Shared/cg/ShareableBitmapCG.cpp:
(WebKit::ShareableBitmap::calculateBytesPerRow):
(WebKit::ShareableBitmap::calculateBytesPerPixel):
(WebKit::ShareableBitmap::createGraphicsContext):
(WebKit::ShareableBitmap::createCGImage const):
* Shared/win/ShareableBitmapDirect2D.cpp:
(WebKit::ShareableBitmap::calculateBytesPerPixel):
2021-01-14 Geoffrey Garen <ggaren@apple.com>
Removed most uses of dispatch_async(dispatch_get_main_queue(), ...)
https://bugs.webkit.org/show_bug.cgi?id=220066
Reviewed by Antti Koivisto.
dispatch_async has two downsides:
(1) Its order is undefined (and in practice highly variable) relative to
other WebKit operations. This sometimes causes flakiness.
(2) It doesn't honor the RunLoop first paint optimization.
We can use RunLoop::dispatch() instead.
* Platform/cocoa/WKPaymentAuthorizationDelegate.mm:
(-[WKPaymentAuthorizationDelegate _didRequestMerchantSession:]):
* Shared/Cocoa/DefaultWebBrowserChecks.mm:
(WebKit::determineITPState):
* UIProcess/API/C/mac/WKContextPrivateMac.mm:
(WKContextGetInfoForInstalledPlugIns):
* UIProcess/API/Cocoa/NSAttributedString.mm:
(+[NSAttributedString _loadFromHTMLWithOptions:contentLoader:completionHandler:]):
* UIProcess/API/Cocoa/WKBrowsingContextController.mm:
(+[WKBrowsingContextController registerSchemeForCustomProtocol:]):
(+[WKBrowsingContextController unregisterSchemeForCustomProtocol:]):
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _evaluateJavaScript:asAsyncFunction:withSourceURL:withArguments:forceUserGesture:inFrame:inWorld:completionHandler:]):
(-[WKWebView _internalDoAfterNextPresentationUpdate:withoutWaitingForPainting:withoutWaitingForAnimatedResize:]):
* UIProcess/API/Cocoa/WKWebViewTesting.mm:
(-[WKWebView _doAfterNextVisibleContentRectUpdate:]):
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _scheduleVisibleContentRectUpdateAfterScrollInView:]):
(-[WKWebView _updateVisibleContentRects]):
(-[WKWebView _firePresentationUpdateForPendingStableStatePresentationCallbacks]):
(-[WKWebView _doAfterNextStablePresentationUpdate:]):
* UIProcess/Cocoa/AutomationClient.mm:
(WebKit::AutomationClient::requestAutomationSession):
* UIProcess/Cocoa/NavigationState.mm:
(WebKit::tryInterceptNavigation):
* UIProcess/Cocoa/PreferenceObserver.mm:
(-[WKPreferenceObserver preferenceDidChange:key:encodedValue:]):
* UIProcess/Cocoa/WebInspectorPreferenceObserver.mm:
(-[WKWebInspectorPreferenceObserver observeValueForKeyPath:ofObject:change:context:]):
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::requestThumbnailWithOperation):
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::webProcessPoolHighDynamicRangeDidChangeCallback):
* UIProcess/Cocoa/WebViewImpl.mm:
(-[WKDOMPasteMenuDelegate menuDidClose:]):
(WebKit::WebViewImpl::updateWindowAndViewFrames):
(WebKit::WebViewImpl::setTopContentInset):
(WebKit::WebViewImpl::didBecomeEditable):
(WebKit::WebViewImpl::requestCandidatesForSelectionIfNeeded):
(WebKit::WebViewImpl::performDragOperation):
* UIProcess/EndowmentStateTracker.mm:
(WebKit::EndowmentStateTracker::registerMonitorIfNecessary):
* UIProcess/_WKTouchEventGenerator.mm:
(-[_WKTouchEventGenerator _sendHIDEvent:]):
(-[_WKTouchEventGenerator _sendMarkerHIDEventWithCompletionBlock:]):
* UIProcess/ios/ProcessAssertionIOS.mm:
(-[WKProcessAssertionBackgroundTaskManager _updateBackgroundTask]):
(-[WKProcessAssertionBackgroundTaskManager _handleBackgroundTaskExpirationOnMainThread]):
(-[WKRBSAssertionDelegate assertion:didInvalidateWithError:]):
(WebKit::ProcessAssertion::ProcessAssertion):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView resignFirstResponderForWebView]):
* UIProcess/ios/WKPDFView.mm:
(-[WKPDFView pdfHostViewControllerExtensionProcessDidCrash:]):
* UIProcess/ios/forms/WKFileUploadPanel.mm:
(-[WKFileUploadPanel imagePickerController:didFinishPickingMediaWithInfo:]):
(-[WKFileUploadPanel imagePickerController:didFinishPickingMultipleMediaWithInfo:]):
* UIProcess/ios/fullscreen/WKFullScreenWindowControllerIOS.mm:
(-[WKFullScreenWindowController placeholderWillMoveToSuperview:]):
* WebProcess/WebPage/RemoteLayerTree/PlatformCALayerRemoteCustom.mm:
(WebKit::PlatformCALayerRemoteCustom::clone const):
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.mm:
(WebKit::RemoteLayerTreeDrawingArea::updateRendering):
* WebProcess/cocoa/VideoFullscreenManager.mm:
(WebKit::VideoFullscreenManager::requestVideoContentLayer):
(WebKit::VideoFullscreenManager::returnVideoContentLayer):
(WebKit::VideoFullscreenManager::didSetupFullscreen):
(WebKit::VideoFullscreenManager::willExitFullscreen):
(WebKit::VideoFullscreenManager::didEnterFullscreen):
(WebKit::VideoFullscreenManager::didExitFullscreen):
(WebKit::VideoFullscreenManager::didCleanupFullscreen):
2021-01-14 Youenn Fablet <youenn@apple.com>
Use callOnMainRunLoop in MediaPermissionUtilities
https://bugs.webkit.org/show_bug.cgi?id=220616
<rdar://problem/73002655>
Reviewed by Chris Dumez.
This ensures that we are in main thread and not in web thread.
* UIProcess/Cocoa/MediaPermissionUtilities.mm:
(WebKit::requestAVCaptureAccessForType):
(WebKit::requestSpeechRecognitionAccess):
2021-01-14 Sihui Liu <sihui_liu@appe.com>
Update buttons of media permission prompt on macOS
https://bugs.webkit.org/show_bug.cgi?id=220613
Reviewed by Eric Carlson.
Make allow button on the right and not selected to match TCC prompts.
* UIProcess/Cocoa/MediaPermissionUtilities.mm:
(WebKit::alertForPermission):
2021-01-14 Per Arne Vollan <pvollan@apple.com>
[GPUP][iOS] Create sandbox extensions for cache and temp directory
https://bugs.webkit.org/show_bug.cgi?id=220595
<rdar://problem/72450307>
Reviewed by Chris Dumez.
Create sandbox extensions for GPU process access to cache and temp directory.
* GPUProcess/GPUProcess.cpp:
(WebKit::GPUProcess::initializeGPUProcess):
* GPUProcess/GPUProcessCreationParameters.cpp:
(WebKit::GPUProcessCreationParameters::encode const):
(WebKit::GPUProcessCreationParameters::decode):
* GPUProcess/GPUProcessCreationParameters.h:
* UIProcess/GPU/GPUProcessProxy.cpp:
(WebKit::GPUProcessProxy::GPUProcessProxy):
2021-01-13 Tim Horton <timothy_horton@apple.com>
REGRESSION (r266634): Messages crashes sometimes while scrolling around and playing YouTube videos
https://bugs.webkit.org/show_bug.cgi?id=220602
<rdar://problem/70402593>
Reviewed by Wenson Hsieh.
No new tests; we are unable to API test video full-screen because of the lack of UIApp;
I have written a stand-alone test app that can reliably reproduce before this patch
and not afterwards.
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::scheduleActivityStateUpdate):
We can't call dispatchActivityStateChange directly underneath a post-commit callback,
because it has side-effects (like un-parenting the full-screen window) that may result
in other frameworks (e.g. UIKit) trying to install commit handlers for the same phase,
which is not allowed.
To fix this, add a dispatch_async; we _only_ care that the activity state change
doesn't apply until after the active commit is complete.
2021-01-13 Chris Dumez <cdumez@apple.com>
REGRESSION(r270961): Flipgrid can no longer capture video
https://bugs.webkit.org/show_bug.cgi?id=220609
<rdar://73168009>
Reviewed by Darin Adler.
Add back line in RemoteAudio::setStorage() that was initializing m_buffer and that was
dropped by mistake in the refactoring in r270961. This was causing
RemoteCaptureSampleManager::RemoteAudio::audioSamplesAvailable() to return early because
m_buffer was not initialized.
No new tests, unsure how to test this.
* WebProcess/cocoa/RemoteCaptureSampleManager.cpp:
(WebKit::RemoteCaptureSampleManager::RemoteAudio::setStorage):
2021-01-13 John Wilander <wilander@apple.com>
PCM: Output logs by default, including to Web Inspector
https://bugs.webkit.org/show_bug.cgi?id=220596
<rdar://problem/73159180>
Reviewed by Brent Fulgham.
This change turns on PCM output to Web Inspector and logs
by default. In some of the cases, the syslog output was
deleted since it doesn't make sense to log those messages
now that we have output in Web Inspector.
* NetworkProcess/PrivateClickMeasurementManager.cpp:
(WebKit::PrivateClickMeasurementManager::storeUnattributed):
(WebKit::PrivateClickMeasurementManager::handleAttribution):
(WebKit::PrivateClickMeasurementManager::attribute):
(WebKit::PrivateClickMeasurementManager::fireConversionRequest):
2021-01-13 Jer Noble <jer.noble@apple.com>
[HANG] 496ms to 1360ms in WebCore::AVAudioSessionCaptureDeviceManager::refreshAudioCaptureDevices()
https://bugs.webkit.org/show_bug.cgi?id=220471
Reviewed by Youenn Fablet.
Use the completion-handler version of RealtimeMediaSourceCenter::getMediaStreamDevices().
* UIProcess/UserMediaPermissionRequestManagerProxy.cpp:
(WebKit::UserMediaPermissionRequestManagerProxy::computeFilteredDeviceList):
(WebKit::UserMediaPermissionRequestManagerProxy::enumerateMediaDevicesForFrame):
* UIProcess/UserMediaPermissionRequestManagerProxy.h:
* UIProcess/UserMediaProcessManager.cpp:
(WebKit::UserMediaProcessManager::updateCaptureDevices):
(WebKit::UserMediaProcessManager::beginMonitoringCaptureDevices):
* UIProcess/UserMediaProcessManager.h:
2021-01-13 Per Arne Vollan <pvollan@apple.com>
[Cocoa] Network extension sandbox extensions are sometimes issued too late
https://bugs.webkit.org/show_bug.cgi?id=220525
<rdar://problem/68443565>
Reviewed by Brent Fulgham.
Currently, Network extension sandbox extensions are sent to the WebContent process as part of the load parameters, but this is too late in some cases.
In these cases, the extensions can be sent along with the DidReceivePolicyDecision message.
* Shared/Cocoa/LoadParametersCocoa.mm:
(WebKit::LoadParameters::platformEncode const):
(WebKit::LoadParameters::platformDecode):
* Shared/LoadParameters.h:
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::addPlatformLoadParameters):
* UIProcess/WebPageProxy.cpp:
(WebKit::createNetworkExtensionsSandboxExtensions):
(WebKit::WebPageProxy::decidePolicyForNavigationActionAsyncShared):
(WebKit::WebPageProxy::decidePolicyForNewWindowAction):
(WebKit::WebPageProxy::decidePolicyForResponseShared):
* WebProcess/WebPage/Cocoa/WebPageCocoa.mm:
(WebKit::WebPage::platformDidReceiveLoadParameters):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::didReceivePolicyDecision):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2021-01-13 Jiewen Tan <jiewen_tan@apple.com>
Crash at SOAuthorizationSession::dismissViewController
https://bugs.webkit.org/show_bug.cgi?id=220482
<rdar://problem/72375494>
Reviewed by Darin Adler.
A crash report suggests that SOAuthorizationSession::dismissViewController could crash at evaluating m_page.
This could only happen if the SOAuthorizationSession object is freed. The stack trace starts with callbacks
from NSNotificationCenter, which capture a RefPtr of the SOAuthorizationSession object and should guarantee
the lifetime of the object. So it contradicts the crash report.
One of the possible explanations is that the RefPtr is somehow over-released within NSNotificationCenter since
it's not thread-safe. To fix that, the RefPtr can be made thread-safe.
No tests.
* UIProcess/Cocoa/SOAuthorization/SOAuthorizationSession.mm:
(WebKit::SOAuthorizationSession::dismissViewController):
2021-01-13 Wenson Hsieh <wenson_hsieh@apple.com>
[macOS] "Correct Spelling Automatically" menu items are inconsistent when autocorrect="off"
https://bugs.webkit.org/show_bug.cgi?id=220598
<rdar://problem/71713611>
Reviewed by Tim Horton.
Plumb the new `canEnableAutomaticSpellingCorrection` flag through `EditorState` to the UI process, where we can
consult it when validating user interface items (namely, the "Correct Spelling Automatically" menu item, which
corresponds to the `-toggleAutomaticSpellingCorrection:` selector).
* Shared/EditorState.cpp:
(WebKit::EditorState::PostLayoutData::encode const):
(WebKit::EditorState::PostLayoutData::decode):
(WebKit::operator<<):
* Shared/EditorState.h:
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::validateUserInterfaceItem):
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::getPlatformEditorState const):
2021-01-13 Alex Christensen <achristensen@webkit.org>
Remove deprecated script and stylesheet SPI, replaced with more inclusively-named SPI
https://bugs.webkit.org/show_bug.cgi?id=214937
Reviewed by Youenn Fablet.
Once rdar://problem/66227777 is done we will be able to remove this old SPI.
* UIProcess/API/Cocoa/WKUserScript.mm:
(-[WKUserScript _initWithSource:injectionTime:forMainFrameOnly:legacyWhitelist:legacyBlacklist:userContentWorld:]): Deleted.
(-[WKUserScript _initWithSource:injectionTime:forMainFrameOnly:legacyWhitelist:legacyBlacklist:associatedURL:userContentWorld:]): Deleted.
(-[WKUserScript _initWithSource:injectionTime:forMainFrameOnly:legacyWhitelist:legacyBlacklist:contentWorld:]): Deleted.
(-[WKUserScript _initWithSource:injectionTime:forMainFrameOnly:legacyWhitelist:legacyBlacklist:associatedURL:contentWorld:]): Deleted.
(-[WKUserScript _initWithSource:injectionTime:forMainFrameOnly:legacyWhitelist:legacyBlacklist:associatedURL:contentWorld:deferRunningUntilNotification:]): Deleted.
* UIProcess/API/Cocoa/WKUserScriptPrivate.h:
* UIProcess/API/Cocoa/_WKUserStyleSheet.h:
* UIProcess/API/Cocoa/_WKUserStyleSheet.mm:
(-[_WKUserStyleSheet initWithSource:forWKWebView:forMainFrameOnly:level:userContentWorld:]): Deleted.
(-[_WKUserStyleSheet initWithSource:forWKWebView:forMainFrameOnly:baseURL:level:userContentWorld:]): Deleted.
(-[_WKUserStyleSheet initWithSource:forMainFrameOnly:legacyWhitelist:legacyBlacklist:userContentWorld:]): Deleted.
(-[_WKUserStyleSheet initWithSource:forMainFrameOnly:legacyWhitelist:legacyBlacklist:baseURL:userContentWorld:]): Deleted.
(-[_WKUserStyleSheet initWithSource:forMainFrameOnly:legacyWhitelist:legacyBlacklist:baseURL:level:userContentWorld:]): Deleted.
2021-01-13 Fujii Hironori <Hironori.Fujii@sony.com>
Reimplement WebCore::isInWebProcess() family as cross-platform by using AuxiliaryProcessInitializationParameters.processType
https://bugs.webkit.org/show_bug.cgi?id=220529
Reviewed by Alex Christensen.
Replaced AuxiliaryProcess::ProcessType with WebCore::AuxiliaryProcessType.
* GPUProcess/GPUProcess.h:
* NetworkProcess/NetworkProcess.h:
* PluginProcess/PluginProcess.h:
* Shared/AuxiliaryProcess.cpp:
(WebKit::AuxiliaryProcess::initialize):
* Shared/AuxiliaryProcess.h: Removed ProcessType.
* Shared/AuxiliaryProcessMain.h:
(WebKit::AuxiliaryProcessMainBase::AuxiliaryProcessMainBase):
(WebKit::AuxiliaryProcessMain):
(WebKit::AuxiliaryProcessMainBase::takeInitializationParameters): Deleted.
* Shared/mac/AuxiliaryProcessMac.mm:
(WebKit::SandboxInfo::SandboxInfo):
(WebKit::processStorageClass):
(WebKit::sandboxDirectory):
(WebKit::applySandbox):
* WebAuthnProcess/WebAuthnProcess.h:
* WebProcess/WebProcess.h:
2021-01-13 John Wilander <wilander@apple.com>
PCM: Experimental debug mode needs to be read from defaults differently on iOS
https://bugs.webkit.org/show_bug.cgi?id=220573
<rdar://problem/73092137>
Reviewed by Brent Fulgham.
Experimental features on iOS use the prefix "WebKitExperimental" so we
need to read it that way on that platform. PCM is not supported on other
platforms than macOS and iOS at this point.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeNetworkProcess):
2021-01-13 Megan Gardner <megan_gardner@apple.com>
Defer setting up app highlight menus until needed to not delay launch time.
https://bugs.webkit.org/show_bug.cgi?id=220577
rdar://72940219
Delay adding the app highlight menus to the callout bar until a selection happens,
as the creation of these menus is expensive, and they are only needed if a selection is
made on the page, which is not guaranteed to happen.
Reviewed by Tim Horton.
* UIProcess/ios/WKContentView.mm:
(-[WKContentView _processDidExit]):
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView setUpInteraction]):
(-[WKContentView _selectionChanged]):
(-[WKContentView setupAppHighlightMenus]):
2021-01-13 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r271401.
https://bugs.webkit.org/show_bug.cgi?id=220591
It is breaking iOS audio rendering
Reverted changeset:
"Unmuting a track in case of end of interruption by another
web process tab is not working well with capture muting icons"
https://bugs.webkit.org/show_bug.cgi?id=220058
https://trac.webkit.org/changeset/271401
2021-01-12 BJ Burg <bburg@apple.com>
[Cocoa] Web Inspector: move browser domain activation methods back to WKWebView and UIDelegate
https://bugs.webkit.org/show_bug.cgi?id=220480
Reviewed by Devin Rousso.
Having browser domain activation methods on WKInspectorDelegate means that
the browser domain will not work with remote inspection of WKWebViews on
macOS. Move these methods back to UIDelegate/WKWebView so that it is possible
to remote inspect Safari Technology Preview with another Safari while still
benefiting from the browser domain being turned on.
Covered by existing API tests.
* UIProcess/API/APIInspectorClient.h:
(API::InspectorClient::browserDomainEnabled): Deleted.
(API::InspectorClient::browserDomainDisabled): Deleted.
* UIProcess/API/APIUIClient.h:
(API::UIClient::didEnableInspectorBrowserDomain): Moved.
(API::UIClient::didDisableInspectorBrowserDomain): Moved.
* UIProcess/API/Cocoa/WKUIDelegatePrivate.h:
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _didEnableBrowserExtensions:]): Moved.
(-[WKWebView _didDisableBrowserExtensions:]): Moved.
Moved from WKInspector.mm. Rather than accessing the browser agent
directly from ObjC code, use the page's WebPageInspectorController.
* UIProcess/API/Cocoa/_WKInspectorDelegate.h:
* UIProcess/API/Cocoa/_WKInspectorPrivate.h:
* UIProcess/API/Cocoa/_WKInspector.mm:
(-[_WKInspector _browserExtensionsEnabled:]): Deleted.
(-[_WKInspector _browserExtensionsDisabled:]): Deleted.
* UIProcess/Cocoa/UIDelegate.h:
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::setDelegate):
(WebKit::UIDelegate::UIClient::didEnableInspectorBrowserDomain):
(WebKit::UIDelegate::UIClient::didDisableInspectorBrowserDomain):
* UIProcess/Inspector/Agents/InspectorBrowserAgent.cpp:
(WebKit::InspectorBrowserAgent::enable):
(WebKit::InspectorBrowserAgent::disable):
* UIProcess/Inspector/Cocoa/InspectorDelegate.h:
* UIProcess/Inspector/Cocoa/InspectorDelegate.mm:
(WebKit::InspectorDelegate::setDelegate):
(WebKit::InspectorDelegate::InspectorClient::~InspectorClient):
(WebKit::InspectorDelegate::InspectorClient::browserDomainEnabled): Deleted.
(WebKit::InspectorDelegate::InspectorClient::browserDomainDisabled): Deleted.
* UIProcess/Inspector/WebInspectorProxy.cpp:
(WebKit::WebInspectorProxy::browserExtensionsEnabled): Deleted.
(WebKit::WebInspectorProxy::browserExtensionsDisabled): Deleted.
* UIProcess/Inspector/WebInspectorProxy.h:
* UIProcess/Inspector/WebPageInspectorController.h:
* UIProcess/Inspector/WebPageInspectorController.cpp:
(WebKit::WebPageInspectorController::WebPageInspectorController):
(WebKit::WebPageInspectorController::init):
(WebKit::WebPageInspectorController::connectFrontend):
(WebKit::WebPageInspectorController::disconnectFrontend):
(WebKit::WebPageInspectorController::disconnectAllFrontends):
(WebKit::WebPageInspectorController::setIndicating):
(WebKit::WebPageInspectorController::createInspectorTarget):
(WebKit::WebPageInspectorController::webPageAgentContext):
(WebKit::WebPageInspectorController::setEnabledBrowserAgent): Added.
(WebKit::WebPageInspectorController::browserExtensionsEnabled): Added.
(WebKit::WebPageInspectorController::browserExtensionsDisabled): Added.
Drive-by: rename m_page to m_inspectedPage to emphasize that this
class exists as part of Web Inspector's backend, not its frontend.
2021-01-12 Brent Fulgham <bfulgham@apple.com>
[macOS] GPU Process unable to play MSE content due to missing sandbox rules
https://bugs.webkit.org/show_bug.cgi?id=220549
<rdar://problem/73047290>
Reviewed by Per Arne Vollan.
Add missing sandbox rules required to support MSE playback in the GPU Process. These
rules were previously added to the WebContent process, but should have also been
included in the GPU Process sandbox.
* GPUProcess/mac/com.apple.WebKit.GPUProcess.sb.in:
* Scripts/process-entitlements.sh:
2021-01-12 Per Arne Vollan <pvollan@apple.com>
[macOS] Reset user directory suffix before getting sandbox directory
https://bugs.webkit.org/show_bug.cgi?id=220358
<rdar://problem/57616019>
Reviewed by Alexey Proskuryakov.
Reset the user directory suffix before getting the sandbox data vault directory with confstr. We do not want to include the user
directory suffix, since the sandbox data vault will then end up inside the host process' cache folder, which is undesirable.
Also, creating the data vault directory with confstr can fail under some circumstances if the user directory suffix is not empty.
This patch also changes how we next set the user directory suffix for the process. Instead of setting the environment variable
DIRHELPER_USER_DIR_SUFFIX, we can use the SPI _set_user_dir_suffix. The behavior should be identical, but I believe it is better
to use SPI for this.
* Shared/mac/AuxiliaryProcessMac.mm:
(WebKit::initializeSandboxParameters):
(WebKit::AuxiliaryProcess::initializeSandbox):
2021-01-12 Don Olmstead <don.olmstead@sony.com>
Non-unified build fixes mid January 2021 edition
https://bugs.webkit.org/show_bug.cgi?id=220560
Unreviewed non-unified build fixes.
* UIProcess/SpeechRecognitionPermissionManager.cpp:
(WebKit::SpeechRecognitionPermissionManager::~SpeechRecognitionPermissionManager):
(WebKit::SpeechRecognitionPermissionManager::request):
(WebKit::SpeechRecognitionPermissionManager::startProcessingRequest):
(WebKit::SpeechRecognitionPermissionManager::continueProcessingRequest):
(WebKit::SpeechRecognitionPermissionManager::completeCurrentRequest):
(WebKit::SpeechRecognitionPermissionManager::requestSpeechRecognitionServiceAccess):
(WebKit::SpeechRecognitionPermissionManager::requestMicrophoneAccess):
(WebKit::SpeechRecognitionPermissionManager::requestUserPermission):
2021-01-12 Wenson Hsieh <wenson_hsieh@apple.com>
REGRESSION (r265044): [macOS] Safari autocorrects text when typing in login field on amazon.com
https://bugs.webkit.org/show_bug.cgi?id=220556
<rdar://problem/71602937>
Reviewed by Darin Adler.
Restores the implementation of `WKBundleNodeHandleSetHTMLInputElementSpellcheckEnabled`, which was removed in
r265044, with the reasoning that the code was unused. However, Safari still uses this SPI on macOS to prevent
automatic spelling correction from triggering when editing form fields that are AutoFillable.
Test: WebKit.DisableSpellcheck
* WebProcess/InjectedBundle/API/c/WKBundleNodeHandle.cpp:
(WKBundleNodeHandleSetHTMLInputElementSpellcheckEnabled):
2021-01-12 Wenson Hsieh <wenson_hsieh@apple.com>
ASSERTION FAILED: m_connection under WebKit::WebPageProxy::acceptsFirstMouse()
https://bugs.webkit.org/show_bug.cgi?id=220545
<rdar://problem/73043854>
Reviewed by Tim Horton.
Avoid the assertion by making it safe to call `WebPageProxy::messageSenderConnection` in debug builds, in the
case where the web process is still launching.
* UIProcess/AuxiliaryProcessProxy.h:
(WebKit::AuxiliaryProcessProxy::hasConnection const):
Add a helper method that returns whether or not a process proxy has a non-null connection.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::messageSenderConnection const):
Use the above helper to avoid calling into `AuxiliaryProcessProxy::connection()` when the connection is null.
2021-01-12 Jer Noble <jer.noble@apple.com>
[Cocoa] Support key rotation with HLS-backed encrypted media streams
https://bugs.webkit.org/show_bug.cgi?id=220493
<rdar://68227709>
Reviewed by Eric Carlson.
Support handling the "unexpectedInitializationDataReceived()" client callback from within the GPUP.
* DerivedSources-input.xcfilelist:
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* GPUProcess/media/RemoteCDMInstanceProxy.cpp:
(WebKit::RemoteCDMInstanceProxy::create):
(WebKit::RemoteCDMInstanceProxy::RemoteCDMInstanceProxy):
(WebKit::RemoteCDMInstanceProxy::~RemoteCDMInstanceProxy):
(WebKit::RemoteCDMInstanceProxy::unrequestedInitializationDataReceived):
* GPUProcess/media/RemoteCDMInstanceProxy.h:
* GPUProcess/media/RemoteCDMProxy.cpp:
(WebKit::RemoteCDMProxy::createInstance):
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/media/RemoteCDMInstance.cpp:
(WebKit::RemoteCDMInstance::RemoteCDMInstance):
(WebKit::RemoteCDMInstance::~RemoteCDMInstance):
(WebKit::RemoteCDMInstance::unrequestedInitializationDataReceived):
* WebProcess/GPU/media/RemoteCDMInstance.h:
* WebProcess/GPU/media/RemoteCDMInstance.messages.in: Added.
2021-01-12 Youenn Fablet <youenn@apple.com>
Unmuting a track in case of end of interruption by another web process tab is not working well with capture muting icons
https://bugs.webkit.org/show_bug.cgi?id=220058
Reviewed by Eric Carlson.
Add an observer of AudioSession interruptions that forwards the signal to all tabs of the process.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
(WebKit::PageAudioSessionInterruptionObserver::PageAudioSessionInterruptionObserver):
(WebKit::PageAudioSessionInterruptionObserver::~PageAudioSessionInterruptionObserver):
(WebKit::PageAudioSessionInterruptionObserver::beginAudioSessionInterruption):
(WebKit::PageAudioSessionInterruptionObserver::endAudioSessionInterruption):
(WebKit::listenToAudioSessionInterruption):
2021-01-12 Philippe Normand <pnormand@igalia.com>
[GStreamer] Lazy initialization support
https://bugs.webkit.org/show_bug.cgi?id=209332
Reviewed by Carlos Garcia Campos.
Introduce a GTK/WPE UserMediaCaptureManager that relays permission requests to the
RealtimeMediaSourceCenter running in the WebProcess. We might move this to the GPUProcess at
some point but for the time being we only want to avoid initializing GStreamer from the
UIProcess.
* PlatformGTK.cmake:
* PlatformWPE.cmake:
* UIProcess/UserMediaPermissionRequestManagerProxy.cpp:
* UIProcess/glib/UserMediaPermissionRequestManagerProxyGLib.cpp: Added.
(WebKit::UserMediaPermissionRequestManagerProxy::processUserMediaPermissionRequest):
* WebProcess/glib/UserMediaCaptureManager.cpp: Added.
(WebKit::UserMediaCaptureManager::UserMediaCaptureManager):
(WebKit::UserMediaCaptureManager::~UserMediaCaptureManager):
(WebKit::UserMediaCaptureManager::validateUserMediaRequestConstraints):
* WebProcess/glib/UserMediaCaptureManager.h: Added.
(WebKit::UserMediaCaptureManager::supplementName):
* WebProcess/glib/UserMediaCaptureManager.messages.in: Added.
* WebProcess/glib/WebProcessGLib.cpp:
(WebKit::WebProcess::platformInitializeWebProcess):
2021-01-12 Carlos Garcia Campos <cgarcia@igalia.com>
Unreviewed. Update OptionsGTK.cmake and NEWS for 2.31.1 release
* gtk/NEWS: Add release notes for 2.31.1.
* webkitglib-symbols.map: Remove WebKit::PluginProcessMain.
2021-01-11 Peng Liu <peng.liu6@apple.com>
Receiving two enter/exit fullscreen delegate callbacks on iPad when presenting fullscreen elements
https://bugs.webkit.org/show_bug.cgi?id=218688
Reviewed by Darin Adler.
* UIProcess/Cocoa/VideoFullscreenManagerProxy.mm:
(WebKit::VideoFullscreenManagerProxy::didExitFullscreen):
(WebKit::VideoFullscreenManagerProxy::didEnterFullscreen):
2021-01-11 Megan Gardner <megan_gardner@apple.com>
Double tap to select does not work if the page clears selections on tap, like grammarly.com does
https://bugs.webkit.org/show_bug.cgi?id=220454
rdar://67757411
Reviewed by Wenson Hsieh.
Double tap to select content did not work on grammerly.com on iPad because this is a desktop website run on a touch-based device,
and there was a script running that would clear the selection on a mouseDown event. When we would send synthetic click events to the page
we would first update the selection and then on completion of the tap, we would dispatch a mouseDown and mouseUp event on iOS. On Mac, we
update the selection between the mouseDown and the mouseUp, so in order to bring us more in line with mac/mouse behaviors, we are now saving
the information needed to update the selection when we believe we are in the middle of a potential tap, and then setting the selection between
these two events. This makes for a more expected change of events, and does not let grammerly.com clear a valid selection based off of synthetic
clicks.
Test: fast/events/touch/ios/double-tap-on-editable-content-for-selection-with-clear-on-touch.html
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::completeSyntheticClick):
(WebKit::WebPage::setSelectionRange):
(WebKit::WebPage::selectTextWithGranularityAtPoint):
2021-01-11 Kate Cheney <katherine_cheney@apple.com>
Crash in pageDidComputePageRects()
https://bugs.webkit.org/show_bug.cgi?id=220518
<rdar://problem/71065870>
Reviewed by Ryosuke Niwa.
Make _printOperation a weak pointer and check it before use.
Since WKPrintingView is ref counted, a reference could still exist
even if _printOperation is nil. Remove the comment as it is not a
justification for a raw pointer.
* UIProcess/mac/WKPrintingView.h:
* UIProcess/mac/WKPrintingView.mm:
(-[WKPrintingView _adjustPrintingMarginsForHeaderAndFooter]):
(-[WKPrintingView _firstPrintedPageNumber]):
(-[WKPrintingView _lastPrintedPageNumber]):
(-[WKPrintingView _preparePDFDataForPrintingOnSecondaryThread]):
(pageDidComputePageRects):
(-[WKPrintingView _askPageToComputePageRects]):
(-[WKPrintingView knowsPageRange:]):
(-[WKPrintingView _drawPreview:]):
(-[WKPrintingView drawRect:]):
(-[WKPrintingView _drawPageBorderWithSizeOnMainThread:]):
(-[WKPrintingView drawPageBorderWithSize:]):
(-[WKPrintingView rectForPage:]):
(-[WKPrintingView locationOfPrintRect:]):
(-[WKPrintingView beginDocument]):
(-[WKPrintingView endDocument]):
2021-01-11 Sihui Liu <sihui_liu@appe.com>
Make SpeechRecognition permission error more informative
https://bugs.webkit.org/show_bug.cgi?id=220436
Reviewed by Youenn Fablet.
Make SpeechRecognitionPermissionManager complete requests with an optional SpeechRecognitionError instead of a
boolean value.
* UIProcess/SpeechRecognitionPermissionManager.cpp:
(WebKit::SpeechRecognitionPermissionManager::~SpeechRecognitionPermissionManager):
(WebKit::SpeechRecognitionPermissionManager::request):
(WebKit::SpeechRecognitionPermissionManager::startProcessingRequest):
(WebKit::SpeechRecognitionPermissionManager::continueProcessingRequest):
(WebKit::SpeechRecognitionPermissionManager::completeCurrentRequest):
(WebKit::SpeechRecognitionPermissionManager::requestSpeechRecognitionServiceAccess):
(WebKit::SpeechRecognitionPermissionManager::requestMicrophoneAccess):
(WebKit::SpeechRecognitionPermissionManager::requestUserPermission):
* UIProcess/SpeechRecognitionPermissionManager.h:
* UIProcess/SpeechRecognitionPermissionRequest.h:
(WebKit::SpeechRecognitionPermissionRequest::create):
(WebKit::SpeechRecognitionPermissionRequest::complete):
(WebKit::SpeechRecognitionPermissionRequest::SpeechRecognitionPermissionRequest):
* UIProcess/SpeechRecognitionServer.cpp:
(WebKit::SpeechRecognitionServer::requestPermissionForRequest):
* UIProcess/SpeechRecognitionServer.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::requestSpeechRecognitionPermission):
* UIProcess/WebPageProxy.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::createSpeechRecognitionServer):
2021-01-11 Alex Christensen <achristensen@webkit.org>
Use sendWithAsyncReply instead of dataCallback for icon loading
https://bugs.webkit.org/show_bug.cgi?id=220381
Reviewed by Youenn Fablet.
* Scripts/webkit/messages.py:
* Shared/API/APIData.cpp:
(API::Data::create):
* Shared/API/APIData.h:
(API::Data::create):
* UIProcess/API/APIIconLoadingClient.h:
(API::IconLoadingClient::getLoadDecisionForIcon):
* UIProcess/API/mac/WKView.mm:
(-[WKView maybeInstallIconLoadingClient]):
* UIProcess/Cocoa/IconLoadingDelegate.h:
* UIProcess/Cocoa/IconLoadingDelegate.mm:
(WebKit::IconLoadingDelegate::IconLoadingClient::getLoadDecisionForIcon):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::getLoadDecisionForIcon):
(WebKit::WebPageProxy::finishedLoadingIcon): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::finishedLoadingIcon): Deleted.
* WebProcess/WebCoreSupport/WebFrameLoaderClient.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::didGetLoadDecisionForIcon):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2021-01-11 BJ Burg <bburg@apple.com>
Web Inspector: add nullptr check for WebInspectorProxy::m_extensionsController
https://bugs.webkit.org/show_bug.cgi?id=220485
<rdar://72496401>
Reviewed by Devin Rousso.
* UIProcess/Inspector/WebInspectorProxy.cpp:
(WebKit::WebInspectorProxy::frontendLoaded):
2021-01-11 Youenn Fablet <youenn@apple.com>
WebProcessPool::establishWorkerContextConnectionToNetworkProcess should make sure to remove the selected process from the cache
https://bugs.webkit.org/show_bug.cgi?id=220503
Reviewed by Darin Adler.
In case we reuse an existing web process, make sure to remove it from web process cache.
Covered by API test.
* UIProcess/WebProcessCache.cpp:
(WebKit::WebProcessCache::addProcessIfPossible):
(WebKit::WebProcessCache::addProcess):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::establishWorkerContextConnectionToNetworkProcess):
2021-01-11 Brent Fulgham <bfulgham@apple.com>
[macOS] Enable the client decoder entitlement for the GPU Process
https://bugs.webkit.org/show_bug.cgi?id=220441
<rdar://problem/72899054>
Reviewed by Darin Adler.
The GPU process needs the client decoder entitlement we use in the WebContent process
to properly handle MSE content.
* Scripts/process-entitlements.sh:
2021-01-11 Yoshiaki Jitsukawa <yoshiaki.jitsukawa@sony.com>
[PlayStation] Add MiniBrowser
https://bugs.webkit.org/show_bug.cgi?id=220359
Reviewed by Don Olmstead.
Add WKRunloop API to allow an external program to run WebKit's generic runloop.
* UIProcess/API/C/playstation/WKRunLoop.cpp: Added.
(WKRunLoopRunMain):
(WKRunLoopStopMain):
(WKRunLoopCallOnMainThread):
(WKRunLoopInitializeMain):
* UIProcess/API/C/playstation/WKRunLoop.h: Added.
Add private WKPagePaint() to draw contents to bitmap.
Remove unused WKPageSetSize() implementation.
* UIProcess/API/C/playstation/WKPagePrivatePlayStation.cpp:
(drawPageBackground):
(WKPageHandleMouseEvent): Fix button click event handling.
(WKPagePaint):
* UIProcess/API/C/playstation/WKPagePrivatePlayStation.h:
Add WKView API to set view client along with fullscreen API and visibility state API.
* UIProcess/API/C/playstation/WKView.cpp:
(WKViewGetPage):
(WKViewSetSize):
(WKViewSetFocus):
(WKViewSetActive):
(WKViewSetVisible):
(WKViewWillEnterFullScreen):
(WKViewDidEnterFullScreen):
(WKViewWillExitFullScreen):
(WKViewDidExitFullScreen):
(WKViewRequestExitFullScreen):
(WKViewIsFullScreen):
(WKViewSetViewClient):
* UIProcess/API/C/playstation/WKView.h:
* UIProcess/API/C/playstation/WKViewClient.h: Added.
* UIProcess/API/playstation/APIViewClient.h: Added.
* UIProcess/playstation/PageClientImpl.cpp:
* UIProcess/playstation/PageClientImpl.h:
* UIProcess/playstation/PlayStationWebView.cpp:
* UIProcess/playstation/PlayStationWebView.h:
Allow rendering without accelerated compositing.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences):
* PlatformPlayStation.cmake: Add relevant sources and include directories.
2021-01-11 Kimmo Kinnunen <kkinnunen@apple.com>
WebKit::IPC::Encoder needs definitions of all custom enum values at the Encoder definition time
https://bugs.webkit.org/show_bug.cgi?id=220410
Reviewed by Darin Adler.
Change WTF::EnumTraits to have either EnumTraits::values or EnumTraits::isValidEnum().
Fixes the upcoming problem where Encoder::encode(E) function is moved from Encoder
to other class, such as EncoderBase, and where we do not want to add redundant includes.
See the WTF/ ChangeLog for discussiong wrt why they were needed before.
* Scripts/webkit/messages.py:
* Scripts/webkit/tests/MessageNames.h:
(WTF::EnumTraits<IPC::MessageName>::isValidEnum):
2021-01-11 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] System font scaling not applied to 'font-size: XXXpt'
https://bugs.webkit.org/show_bug.cgi?id=218450
Reviewed by Žan Doberšek.
Apply the text scale factor as WebPageProxy text zoom factor, instead of changing the default font size in
settings that only applies to default fonts and it was incorrectly calculated.
* UIProcess/API/glib/WebKitSettings.cpp:
(webKitSettingsDispose):
(webKitSettingsConstructed):
* UIProcess/API/glib/WebKitWebView.cpp:
(webkitWebViewConstructed):
(webkitWebViewDispose):
(webkit_web_view_set_zoom_level):
(webkit_web_view_get_zoom_level):
2021-01-08 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r271331.
https://bugs.webkit.org/show_bug.cgi?id=220487
Wasn't ready for review.
Reverted changeset:
"[macOS] Reset user directory suffix before getting sandbox
directory"
https://bugs.webkit.org/show_bug.cgi?id=220358
https://trac.webkit.org/changeset/271331
2021-01-08 Per Arne Vollan <pvollan@apple.com>
[macOS] Reset user directory suffix before getting sandbox directory
https://bugs.webkit.org/show_bug.cgi?id=220358
<rdar://problem/57616019>
Reviewed by Brent Fulgham.
Reset user directory suffix before getting sandbox data vault directory. We do not want to include the user directory suffix,
since the compiled sandbox should be shared by all WebKit processes of the same type. Also, creating the data vault directory
can fail under some circumstances if the user directory suffix is not empty.
* Shared/mac/AuxiliaryProcessMac.mm:
(WebKit::AuxiliaryProcess::initializeSandbox):
2021-01-08 Peng Liu <peng.liu6@apple.com>
PlaybackSessionManager::m_clientCounts is not updated correctly when a video enters picture-in-picture from fullscreen
https://bugs.webkit.org/show_bug.cgi?id=220435
Reviewed by Jer Noble.
A follow-up patch to fix an assertion failure on Mac.
`VideoFullscreenManager::exitVideoFullscreenToModeWithoutAnimation()` should not call
`removeClientForContext()` on Mac because `didCleanupFullscreen()` will do that.
Fix layout test: media/element-containing-pip-video-going-into-fullscreen.html
* WebProcess/cocoa/VideoFullscreenManager.mm:
(WebKit::VideoFullscreenManager::exitVideoFullscreenToModeWithoutAnimation):
2021-01-08 Wenson Hsieh <wenson_hsieh@apple.com>
[macOS] -[WKWebView acceptsFirstMouse:] sometimes crashes in IPC::Connection::createSyncMessageEncoder
https://bugs.webkit.org/show_bug.cgi?id=220469
<rdar://problem/72319199>
Reviewed by Chris Dumez.
When sending IPC to the web process from the UI process, it's possible for the `WebProcessProxy`'s IPC
connection to be null in the case where the web process is still in the act of launching. In the case of
asynchronous IPC, we handle this by queueing the IPC message on the `WebProcessProxy` in the case where the
connection hasn't been created yet (refer to `AuxiliaryProcessProxy::sendMessage`). However, in the case where
we're sending a sync message, we assume that the IPC connection exists in `MessageSender::sendSync`, which
causes us to crash with a null dereference.
To fix this, remove the debug assertion in `MessageSender::sendSync` and return a null `SendSyncResult` in the
case where the connection has not been created.
Test: WebKit2.AcceptsFirstMouseDuringWebProcessLaunch
* Platform/IPC/MessageSender.h:
2021-01-08 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Adopt new UI for the update flow
https://bugs.webkit.org/show_bug.cgi?id=219713
<rdar://problem/72154988>
Reviewed by Brent Fulgham.
Covered by manual tests.
* Platform/spi/Cocoa/AuthenticationServicesCoreSPI.h:
(NS_ERROR_ENUM):
* UIProcess/WebAuthentication/Cocoa/AuthenticatorPresenterCoordinator.mm:
(WebKit::AuthenticatorPresenterCoordinator::updatePresenter):
2021-01-08 Youenn Fablet <youenn@apple.com>
Service Worker is no longer inspectable
https://bugs.webkit.org/show_bug.cgi?id=220406
<rdar://problem/72883757>
Reviewed by Per Arne Vollan.
We were sending the sandbox extension later when enabling remote inspector.
But this is not soon enough for inspecting service workers.
For that reason, send the sandbox extension at process init time based on whether develop menu is on or not.
Manually tested by validating that Develop Menu can list running service workers.
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
* UIProcess/Cocoa/WebProcessProxyCocoa.mm:
(WebKit::WebProcessProxy::shouldEnableRemoteInspector):
(WebKit::WebProcessProxy::enableRemoteInspectorIfNeeded):
* UIProcess/WebProcessProxy.h:
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
(WebKit::WebProcess::enableRemoteWebInspector):
2021-01-08 Chris Lord <clord@igalia.com>
[WPE] Enable smooth-motion and kinetic scrolling on touchpads
https://bugs.webkit.org/show_bug.cgi?id=219942
Reviewed by Žan Doberšek.
Interpret axis motion events with a zero value as axis stop events and
send the appropriate wheel event phase. This enables kinetic scrolling
when using touchpads and other smooth-scrolling devices.
* UIProcess/API/wpe/PageClientImpl.cpp:
(WebKit::PageClientImpl::doneWithTouchEvent):
* UIProcess/API/wpe/WPEView.cpp:
(WKWPE::m_backend):
* UIProcess/API/wpe/WPEView.h:
2021-01-08 Youenn Fablet <youenn@apple.com>
Make sure that if NetworkProcess clears DOMCache, it also clears service worker registrations
https://bugs.webkit.org/show_bug.cgi?id=220408
<rdar://problem/72360003>
Reviewed by Alex Christensen.
It is expected that clearing DOM cache mandates clearing service worker registrations
as service workers expect cache entries added at install time to stay.
It is hard for clients to enforce this so it is best to handle it in Network Process.
When DOMCache is cleared, service worker registrations will also be cleared.
The reverse is not guaranteed.
Covered by API test.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::deleteWebsiteData):
(WebKit::NetworkProcess::deleteWebsiteDataForOrigins):
(WebKit::NetworkProcess::deleteAndRestrictWebsiteDataForRegistrableDomains):
2021-01-07 Alex Christensen <achristensen@webkit.org>
Use WeakHashSet instead of HashSet<T*>
https://bugs.webkit.org/show_bug.cgi?id=220455
Reviewed by Geoffrey Garen.
* NetworkProcess/NetworkSession.cpp:
(WebKit::NetworkSession::invalidateAndCancel):
(WebKit::NetworkSession::registerNetworkDataTask):
(WebKit::NetworkSession::unregisterNetworkDataTask):
* NetworkProcess/NetworkSession.h:
(WebKit::NetworkSession::registerNetworkDataTask): Deleted.
(WebKit::NetworkSession::unregisterNetworkDataTask): Deleted.
* UIProcess/WebPageGroup.cpp:
(WebKit::WebPageGroup::addPage):
(WebKit::WebPageGroup::removePage):
(WebKit::WebPageGroup::setPreferences):
* UIProcess/WebPageGroup.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
(WebKit::WebPageProxy::~WebPageProxy):
* UIProcess/WebPreferences.cpp:
(WebKit::WebPreferences::~WebPreferences):
(WebKit::WebPreferences::addPage):
(WebKit::WebPreferences::removePage):
(WebKit::WebPreferences::update):
(WebKit::WebPreferences::updateBoolValueForInternalDebugFeatureKey):
* UIProcess/WebPreferences.h:
* WebProcess/Geolocation/WebGeolocationManager.cpp:
(WebKit::WebGeolocationManager::registerWebPage):
(WebKit::WebGeolocationManager::unregisterWebPage):
(WebKit::WebGeolocationManager::setEnableHighAccuracyForPage):
(WebKit::WebGeolocationManager::isUpdating const):
(WebKit::WebGeolocationManager::isHighAccuracyEnabled const):
* WebProcess/Geolocation/WebGeolocationManager.h:
(WebKit::WebGeolocationManager::isUpdating const): Deleted.
(WebKit::WebGeolocationManager::isHighAccuracyEnabled const): Deleted.
2021-01-07 Per Arne Vollan <pvollan@apple.com>
[iOS] Silence diagnostics sandbox violations
https://bugs.webkit.org/show_bug.cgi?id=220414
Reviewed by Darin Adler.
For performance reasons, silence benign diagnostics sandbox violations in the WebContent process on iOS.
Additionally, remove logging for a sysctl in the Networking process' sandbox, since it is known to be used.
* Resources/SandboxProfiles/ios/com.apple.WebKit.Networking.sb:
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2021-01-07 Kate Cheney <katherine_cheney@apple.com>
Error in layout tests: "Passed ITP enabled state (0) does not match TCC setting (1)"
https://bugs.webkit.org/show_bug.cgi?id=220385
<rdar://problem/70730482>
Reviewed by Darin Adler.
No new tests, this has no behavior change. It will fix excessive
logging when running tests.
We do not need to log a discrepancy between these states
for WebKitTestRunner or TestWebKitAPI because they are sometimes treated
like web browsers but need to turn ITP on/off for specific tests, so
these states will not always match.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::activateSessionCleanup):
* Shared/Cocoa/DefaultWebBrowserChecks.h:
* Shared/Cocoa/DefaultWebBrowserChecks.mm:
(WebKit::isRunningTest):
(WebKit::isParentProcessAFullWebBrowser):
(WebKit::isFullWebBrowser):
(WebKit::shouldBeTreatedAsFullBrowser): Deleted.
2021-01-07 Alex Christensen <achristensen@webkit.org>
Remove assertions in MESSAGE_CHECK definition
https://bugs.webkit.org/show_bug.cgi?id=220401
* Platform/IPC/Connection.h:
The unit test associated with r271263 hits these assertions in debug builds.
Since we plan to increase this type of unit testing, I remove the assertions.
2021-01-07 Mark Lam <mark.lam@apple.com>
Work around Clang bug in __builtin_return_address().
https://bugs.webkit.org/show_bug.cgi?id=220432
rdar://71648468
Reviewed by Yusuke Suzuki.
* PluginProcess/mac/PluginProcessShim.mm:
(WebKit::shimCFStringCompare):
- We go direct to ptrauth.h instead of using the WTF PtrTag abstraction because
this file appears to be going out of its way to avoid importing config.h.
Because of this, importing PtrTag.h results in a lot of build error complications.
Rather than jump thru many hoops to make importing PtrTag.h work and because all
we really want is only to use ptrauth_strip(), importing ptrauth.h is simpler.
2021-01-07 Peng Liu <peng.liu6@apple.com>
PlaybackSessionManager::m_clientCounts is not updated correctly when a video enters picture-in-picture from fullscreen
https://bugs.webkit.org/show_bug.cgi?id=220435
Reviewed by Eric Carlson.
* WebProcess/cocoa/VideoFullscreenManager.mm:
(WebKit::VideoFullscreenManager::enterVideoFullscreenForVideoElement):
When a fullscreen video enters picture-in-picture, we should not call `addClientForContext()`
because `m_clientCounts` was updated when the video entered fullscreen.
(WebKit::VideoFullscreenManager::exitVideoFullscreenToModeWithoutAnimation):
We need to call `removeClientForContext()` in this function like `didCleanupFullscreen()` does.
2021-01-07 Eric Carlson <eric.carlson@apple.com>
[Mac] Add runtime logging to format reader and WebM parser
https://bugs.webkit.org/show_bug.cgi?id=220423
<rdar://problem/72896655>
Reviewed by Andy Estes.
Use the Document::sharedLogger to log format and track reader state changes
and errors.
* Shared/mac/MediaFormatReader/MediaFormatReader.cpp:
(WebKit::nextLogIdentifier):
(WebKit::logChannel):
(WebKit::logClassName):
(WebKit::MediaFormatReader::parseByteSource):
(WebKit::MediaFormatReader::didParseTracks):
(WebKit::MediaFormatReader::copyProperty):
(WebKit::MediaFormatReader::nextTrackReaderLogIdentifier const):
* Shared/mac/MediaFormatReader/MediaFormatReader.h:
* Shared/mac/MediaFormatReader/WebKit::MediaTrackReader.cpp:
(WebKit::MediaTrackReader::MediaTrackReader):
(WebKit::MediaTrackReader::finishParsing):
(WebKit::MediaTrackReader::mediaTypeString const):
(WebKit::MediaTrackReader::copyProperty):
(WebKit::MediaTrackReader::finalize):
(WebKit::MediaTrackReader::logChannel const):
* Shared/mac/MediaFormatReader/WebKit::MediaTrackReader.h:
2021-01-07 Alexey Shvayka <shvaikalesh@gmail.com>
[JSC] Simplify get*PropertyNames() methods and EnumerationMode
https://bugs.webkit.org/show_bug.cgi?id=212954
Reviewed by Yusuke Suzuki.
Adjust for changes in JSC's MethodTable, TypeInfo, and EnumerationMode.
No new tests, no behavior change.
* WebProcess/Plugins/Netscape/JSNPObject.cpp:
(WebKit::JSNPObject::getOwnPropertyNames):
* WebProcess/Plugins/Netscape/JSNPObject.h:
* WebProcess/Plugins/Netscape/NPJSObject.cpp:
(WebKit::NPJSObject::enumerate):
2021-01-07 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r271192.
https://bugs.webkit.org/show_bug.cgi?id=220440
Reverted changeset:
"NSCrossWebsiteTrackingUsageDescription is not working on Mac,
ITP is always enabled"
https://bugs.webkit.org/show_bug.cgi?id=220190
https://trac.webkit.org/changeset/271192
2021-01-07 Alex Christensen <achristensen@webkit.org>
Use smart pointers in UIDelegate and NavigationState
https://bugs.webkit.org/show_bug.cgi?id=220394
<rdar://problem/72496374>
Reviewed by Geoff Garen.
UIDelegate and NavigationState are owned by WKWebView, but the clients are owned by WebPageProxy,
which is independently reference counted and can have a different lifetime.
There should be no raw C++ references between the two, so use WeakPtr instead.
Also, use WeakObjCPtr<WKWebView> instead of a raw C++ WKWebView*
* UIProcess/Cocoa/NavigationState.h:
* UIProcess/Cocoa/NavigationState.mm:
(WebKit::NavigationState::NavigationState):
(WebKit::NavigationState::~NavigationState):
(WebKit::NavigationState::navigationGestureDidBegin):
(WebKit::NavigationState::navigationGestureWillEnd):
(WebKit::NavigationState::navigationGestureDidEnd):
(WebKit::NavigationState::willRecordNavigationSnapshot):
(WebKit::NavigationState::navigationGestureSnapshotWasRemoved):
(WebKit::NavigationState::didRequestPasswordForQuickLookDocument):
(WebKit::NavigationState::didFirstPaint):
(WebKit::NavigationState::NavigationClient::NavigationClient):
(WebKit::NavigationState::NavigationClient::webGLLoadPolicy const):
(WebKit::NavigationState::NavigationClient::resolveWebGLLoadPolicy const):
(WebKit::NavigationState::NavigationClient::didChangeBackForwardList):
(WebKit::NavigationState::NavigationClient::willGoToBackForwardListItem):
(WebKit::NavigationState::NavigationClient::decidePolicyForNavigationAction):
(WebKit::NavigationState::NavigationClient::contentRuleListNotification):
(WebKit::NavigationState::NavigationClient::decidePolicyForNavigationResponse):
(WebKit::NavigationState::NavigationClient::didStartProvisionalNavigation):
(WebKit::NavigationState::NavigationClient::didStartProvisionalLoadForFrame):
(WebKit::NavigationState::NavigationClient::didReceiveServerRedirectForProvisionalNavigation):
(WebKit::NavigationState::NavigationClient::willPerformClientRedirect):
(WebKit::NavigationState::NavigationClient::didPerformClientRedirect):
(WebKit::NavigationState::NavigationClient::didCancelClientRedirect):
(WebKit::NavigationState::NavigationClient::didFailProvisionalNavigationWithError):
(WebKit::NavigationState::NavigationClient::didFailProvisionalLoadWithErrorForFrame):
(WebKit::NavigationState::NavigationClient::didCommitNavigation):
(WebKit::NavigationState::NavigationClient::didCommitLoadForFrame):
(WebKit::NavigationState::NavigationClient::didFinishDocumentLoad):
(WebKit::NavigationState::NavigationClient::didFinishNavigation):
(WebKit::NavigationState::NavigationClient::didFinishLoadForFrame):
(WebKit::NavigationState::NavigationClient::didFailNavigationWithError):
(WebKit::NavigationState::NavigationClient::didFailLoadWithErrorForFrame):
(WebKit::NavigationState::NavigationClient::didSameDocumentNavigation):
(WebKit::NavigationState::NavigationClient::renderingProgressDidChange):
(WebKit::NavigationState::NavigationClient::shouldBypassContentModeSafeguards const):
(WebKit::NavigationState::NavigationClient::didReceiveAuthenticationChallenge):
(WebKit::NavigationState::NavigationClient::shouldAllowLegacyTLS):
(WebKit::NavigationState::NavigationClient::didNegotiateModernTLS):
(WebKit::NavigationState::NavigationClient::processDidTerminate):
(WebKit::NavigationState::NavigationClient::processDidBecomeResponsive):
(WebKit::NavigationState::NavigationClient::processDidBecomeUnresponsive):
(WebKit::NavigationState::NavigationClient::webCryptoMasterKey):
(WebKit::NavigationState::NavigationClient::navigationActionDidBecomeDownload):
(WebKit::NavigationState::NavigationClient::navigationResponseDidBecomeDownload):
(WebKit::NavigationState::NavigationClient::contextMenuDidCreateDownload):
(WebKit::NavigationState::NavigationClient::didStartLoadForQuickLookDocumentInMainFrame):
(WebKit::NavigationState::NavigationClient::didFinishLoadForQuickLookDocumentInMainFrame):
(WebKit::NavigationState::NavigationClient::decidePolicyForSOAuthorizationLoad):
(WebKit::NavigationState::HistoryClient::HistoryClient):
(WebKit::NavigationState::HistoryClient::didNavigateWithNavigationData):
(WebKit::NavigationState::HistoryClient::didPerformClientRedirect):
(WebKit::NavigationState::HistoryClient::didPerformServerRedirect):
(WebKit::NavigationState::HistoryClient::didUpdateHistoryTitle):
* UIProcess/Cocoa/UIDelegate.h:
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::ContextMenuClient::ContextMenuClient):
(WebKit::UIDelegate::ContextMenuClient::menuFromProposedMenu):
(WebKit::UIDelegate::UIClient::UIClient):
(WebKit::UIDelegate::UIClient::createNewPage):
(WebKit::UIDelegate::UIClient::runJavaScriptAlert):
(WebKit::UIDelegate::UIClient::runJavaScriptConfirm):
(WebKit::UIDelegate::UIClient::runJavaScriptPrompt):
(WebKit::UIDelegate::UIClient::requestStorageAccessConfirm):
(WebKit::UIDelegate::UIClient::decidePolicyForGeolocationPermissionRequest):
(WebKit::UIDelegate::UIClient::didResignInputElementStrongPasswordAppearance):
(WebKit::UIDelegate::UIClient::canRunBeforeUnloadConfirmPanel const):
(WebKit::UIDelegate::UIClient::runBeforeUnloadConfirmPanel):
(WebKit::UIDelegate::UIClient::exceededDatabaseQuota):
(WebKit::UIDelegate::UIClient::takeFocus):
(WebKit::UIDelegate::UIClient::handleAutoplayEvent):
(WebKit::UIDelegate::UIClient::canRunModal const):
(WebKit::UIDelegate::UIClient::runModal):
(WebKit::UIDelegate::UIClient::headerHeight):
(WebKit::UIDelegate::UIClient::footerHeight):
(WebKit::UIDelegate::UIClient::drawHeader):
(WebKit::UIDelegate::UIClient::drawFooter):
(WebKit::UIDelegate::UIClient::pageDidScroll):
(WebKit::UIDelegate::UIClient::focus):
(WebKit::UIDelegate::UIClient::unfocus):
(WebKit::UIDelegate::UIClient::unavailablePluginButtonClicked):
(WebKit::UIDelegate::UIClient::didExceedBackgroundResourceLimitWhileInForeground):
(WebKit::UIDelegate::UIClient::didNotHandleWheelEvent):
(WebKit::UIDelegate::UIClient::setIsResizable):
(WebKit::UIDelegate::UIClient::setWindowFrame):
(WebKit::UIDelegate::UIClient::windowFrame):
(WebKit::UIDelegate::UIClient::mouseDidMoveOverElement):
(WebKit::UIDelegate::UIClient::toolbarsAreVisible):
(WebKit::UIDelegate::UIClient::didClickAutoFillButton):
(WebKit::UIDelegate::UIClient::showPage):
(WebKit::UIDelegate::UIClient::saveDataToFileInDownloadsFolder):
(WebKit::UIDelegate::UIClient::decidePolicyForNotificationPermissionRequest):
(WebKit::UIDelegate::UIClient::configurationForLocalInspector):
(WebKit::UIDelegate::UIClient::didAttachLocalInspector):
(WebKit::UIDelegate::UIClient::willCloseLocalInspector):
(WebKit::UIDelegate::UIClient::runOpenPanel):
(WebKit::UIDelegate::UIClient::shouldAllowDeviceOrientationAndMotionAccess):
(WebKit::UIDelegate::UIClient::didChangeFontAttributes):
(WebKit::UIDelegate::UIClient::decidePolicyForUserMediaPermissionRequest):
(WebKit::UIDelegate::UIClient::checkUserMediaPermissionForOrigin):
(WebKit::UIDelegate::UIClient::mediaCaptureStateDidChange):
(WebKit::UIDelegate::UIClient::reachedApplicationCacheOriginQuota):
(WebKit::UIDelegate::UIClient::printFrame):
(WebKit::UIDelegate::UIClient::close):
(WebKit::UIDelegate::UIClient::fullscreenMayReturnToInline):
(WebKit::UIDelegate::UIClient::didEnterFullscreen):
(WebKit::UIDelegate::UIClient::didExitFullscreen):
(WebKit::UIDelegate::UIClient::shouldIncludeAppLinkActionsForElement):
(WebKit::UIDelegate::UIClient::actionsForElement):
(WebKit::UIDelegate::UIClient::didNotHandleTapAsClick):
(WebKit::UIDelegate::UIClient::presentingViewController):
(WebKit::UIDelegate::UIClient::dataDetectionContext):
(WebKit::UIDelegate::UIClient::requestPointerLock):
(WebKit::UIDelegate::UIClient::didLosePointerLock):
(WebKit::UIDelegate::UIClient::didShowSafeBrowsingWarning):
(WebKit::UIDelegate::UIClient::confirmPDFOpening):
(WebKit::UIDelegate::UIClient::runWebAuthenticationPanel):
(WebKit::UIDelegate::UIClient::hasVideoInPictureInPictureDidChange):
(WebKit::UIDelegate::UIClient::imageOrMediaDocumentSizeChanged):
(WebKit::UIDelegate::UIClient::decidePolicyForSpeechRecognitionPermissionRequest):
2021-01-07 Alex Christensen <achristensen@webkit.org>
Assertion failure in WebPageProxy::stopURLSchemeTask,
https://bugs.webkit.org/show_bug.cgi?id=220401
Reviewed by Ryosuke Niwa.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::startURLSchemeTaskShared):
(WebKit::WebPageProxy::stopURLSchemeTask):
(WebKit::WebPageProxy::loadSynchronousURLSchemeTask):
* UIProcess/WebURLSchemeHandler.cpp:
(WebKit::WebURLSchemeHandler::processForTaskIdentifier const):
(WebKit::WebURLSchemeHandler::stopTask):
(WebKit::WebURLSchemeHandler::removeTaskFromPageMap):
2021-01-07 Andy Estes <aestes@apple.com>
[Mac] Prepend "Media" to FormatReader, SampleCursor, and TrackReader class names
https://bugs.webkit.org/show_bug.cgi?id=220424
Reviewed by Eric Carlson.
Also moved MediaSampleByteRange into its own file from MediaTrackReader.cpp.
* MediaFormatReaderPlugIn/MediaFormatReaderPlugIn.cpp:
* Shared/API/c/mac/WKMediaFormatReader.cpp: Renamed from Source/WebKit/Shared/API/c/mac/WKFormatReader.cpp.
* Shared/API/c/mac/WKMediaFormatReader.h: Renamed from Source/WebKit/Shared/API/c/mac/WKFormatReader.h.
* Shared/mac/MediaFormatReader/CoreMediaWrapped.h: Updated DECLARE_CORE_MEDIA_TRAITS to account for renaming.
* Shared/mac/MediaFormatReader/MediaFormatReader.cpp: Renamed from Source/WebKit/Shared/mac/MediaFormatReader/FormatReader.cpp.
* Shared/mac/MediaFormatReader/MediaFormatReader.h: Renamed from Source/WebKit/Shared/mac/MediaFormatReader/FormatReader.h.
* Shared/mac/MediaFormatReader/MediaSampleByteRange.cpp: Added.
* Shared/mac/MediaFormatReader/MediaSampleByteRange.h: Added.
* Shared/mac/MediaFormatReader/MediaSampleCursor.cpp: Renamed from Source/WebKit/Shared/mac/MediaFormatReader/SampleCursor.cpp.
* Shared/mac/MediaFormatReader/MediaSampleCursor.h: Renamed from Source/WebKit/Shared/mac/MediaFormatReader/SampleCursor.h.
* Shared/mac/MediaFormatReader/MediaTrackReader.cpp: Renamed from Source/WebKit/Shared/mac/MediaFormatReader/TrackReader.cpp.
* Shared/mac/MediaFormatReader/MediaTrackReader.h: Renamed from Source/WebKit/Shared/mac/MediaFormatReader/TrackReader.h.
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
2021-01-07 Chris Dumez <cdumez@apple.com>
[GPUProcess] Implement GPUProcess crash recovery for MediaElementAudioSourceNode
https://bugs.webkit.org/show_bug.cgi?id=220391
Reviewed by Geoffrey Garen.
When audio is playing using WebAudio and a MediaElementAudioSourceNode, make sure that
audio resumes playing seemlessly after a GPUProcess crash.
* WebProcess/GPU/GPUProcessConnection.h:
Subclass CanMakeWeakPtr<> so that clients can hold a weak pointer to a GPUProcessConnection.
* WebProcess/GPU/media/RemoteAudioSourceProvider.cpp:
(WebKit::RemoteAudioSourceProvider::create):
(WebKit::RemoteAudioSourceProvider::RemoteAudioSourceProvider):
(WebKit::RemoteAudioSourceProvider::close):
(WebKit::RemoteAudioSourceProvider::hasNewClient):
* WebProcess/GPU/media/RemoteAudioSourceProvider.h:
Associate the RemoteAudioSourceProvider with a specific GPUProcessConnection instead of always
using the latest connection via WebProcess::ensureGPUProcessConnection(). In the event of a
GPUProcess crash, MediaPlayer will reconstruct its MediaPlayerPrivate which will reconstruct
a new RemoteAudioSourceProvider (since RemoteAudioSourceProvider is owned by
MediaPlayerPrivateRemote). As a result, RemoteAudioSourceProvider does not need to do anything
to deal with a GPUProcess crash. However, we need to make sure it only interacts with the
GPUProcessConnection that existed when it was constructed.
2021-01-07 Kimmo Kinnunen <kkinnunen@apple.com>
WebKit IPC message names should indicate if the message is synchronous
https://bugs.webkit.org/show_bug.cgi?id=220319
Reviewed by Chris Dumez.
Order WebKit IPC MessageName enumerators by listing asynchronous
messages first. This enables relatively efficient implementation of
isMessageSync().
Remove MessageFlags::SyncMessage. This information is already stored
in the MessageName -- a message is either synchronous or asynchronous,
but not both.
* Platform/IPC/Connection.cpp:
(IPC::Connection::createSyncMessageEncoder):
* Platform/IPC/Decoder.cpp:
(IPC::Decoder::alignBufferPosition):
* Platform/IPC/Decoder.h:
(IPC::Decoder::isSyncMessage const):
* Platform/IPC/Encoder.cpp:
(IPC::Encoder::grow):
* Platform/IPC/Encoder.h:
* Platform/IPC/MessageArgumentDescriptions.h:
* Platform/IPC/MessageFlags.h:
* Scripts/webkit/messages.py:
* Scripts/webkit/messages_unittest.py:
(GeneratedFileContentsTest):
* Scripts/webkit/model.py:
(check_global_model_inputs):
* Scripts/webkit/model_unittest.py:
(test_mismatch_message_attribute_sync):
* Scripts/webkit/tests/MessageArgumentDescriptions.cpp:
* Scripts/webkit/tests/MessageNames.cpp:
(IPC::description):
(IPC::receiverName):
(IPC::isValidMessageName):
* Scripts/webkit/tests/MessageNames.h:
(IPC::messageIsSync):
2021-01-07 Sihui Liu <sihui_liu@appe.com>
ASSERTION FAILED: !m_messageReceiverMapCount under WebKit::SpeechRecognitionServer::~SpeechRecognitionServer()
https://bugs.webkit.org/show_bug.cgi?id=220404
Reviewed by Youenn Fablet.
API test: WebKit2.SpeechRecognitionPageIsDestroyed
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::destroySpeechRecognitionServer):
2021-01-07 Kimmo Kinnunen <kkinnunen@apple.com>
WebKit IPC builtin messages should be part of the model
https://bugs.webkit.org/show_bug.cgi?id=220318
Reviewed by Chris Dumez.
Add IPC and AsyncReply message receivers to the IPC model
so that their messages can be part of the generator dynamic logic instead of
being part of the static textual templates.
IPC receiver is a built-in message receiver that has few async messages.
AsyncReply receiver is a built-in message receiver that has one async message
for each message declared as 'Async' in any receiver.
This helps to refine features such as "is message asynchronous" more easily.
Adds new test case TestWithIfMessage to check a case that went wrong when
making the change.
Moves parsing tests to parser_unittest.py so that TestWithIfMessage does not
need to be added to the parsing tests. The said test is not intended to exercises
anything particular in the parsing.
Adds test logic for asserting MessageNames.{h,cpp} and MessageArgumentDescriptions.cpp
are as expected.
* Scripts/generate-message-receiver.py:
(main):
* Scripts/webkit/messages.py:
* Scripts/webkit/model.py:
(check_global_model_inputs):
(generate_global_model):
* Scripts/webkit/model_unittest.py:
* Scripts/webkit/tests/MessageNames.cpp:
(IPC::description):
(IPC::receiverName):
(IPC::isValidMessageName):
* Scripts/webkit/tests/MessageNames.h:
2021-01-07 Lauro Moura <lmoura@igalia.com>
REGRESSION(r270425) [GTK] wheel scrolling stopped working
https://bugs.webkit.org/show_bug.cgi?id=219547
Reviewed by Carlos Garcia Campos.
In GTK we may have the case where the async scrolling is disabled at
runtime despite being compiled with ASYNC_SCROLLING/SCROLLING_THREAD.
In these cases, the wheel events still need to be dispatched through
the main thread.
* WebProcess/WebPage/EventDispatcher.cpp:
(WebKit::EventDispatcher::wheelEvent):
2021-01-07 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] Build failures with GTK4 3.99.5.1
https://bugs.webkit.org/show_bug.cgi?id=219844
Reviewed by Philippe Normand.
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseSizeAllocate):
2021-01-07 Youenn Fablet <youenn@apple.com>
WKWebView should expose navigator.mediaDevices when content is loaded from app bundle
https://bugs.webkit.org/show_bug.cgi?id=220184
<rdar://problem/72792032>
Reviewed by Eric Carlson.
In case of non HTTP or HTTPS origins, getUserMedia was always denied.
This patch updates the code by allowing getUserMedia prompts for other protocols.
In that case, the user is presented the prompt with the application name since the origin domain does not mean anything.
Manually tested.
* UIProcess/Cocoa/MediaPermissionUtilities.mm:
(WebKit::alertMessageText):
2021-01-07 Jiewen Tan <jiewen_tan@apple.com>
Unreviewed, a build fix after r271221
* Platform/spi/Cocoa/NearFieldSPI.h:
* UIProcess/WebAuthentication/Cocoa/NfcService.mm:
(WebKit::NfcService::isAvailable):
(WebKit::NfcService::platformStartDiscovery):
Changed back to sharedHardwareManager.
2021-01-06 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Adopt new UI for the Client PIN and dismiss flow
https://bugs.webkit.org/show_bug.cgi?id=219712
<rdar://problem/72154935>
Reviewed by Brent Fulgham.
Covered by manual tests.
This patch does the following few things:
1. It tweaks the WebAuthn process as a UI view service to be able to run NearField in the background.
2. It adopts new SPI for security keys' client pin support.
3. It adopts new SPI to dismiss the UI when necessary.
* Platform/spi/Cocoa/AuthenticationServicesCoreSPI.h:
(NS_ERROR_ENUM):
Adds new SPI.
* Scripts/process-entitlements.sh:
Tweaks WebAuthn process for NFC.
* UIProcess/WebAuthentication/Cocoa/AuthenticationServicesCoreSoftLink.h:
* UIProcess/WebAuthentication/Cocoa/AuthenticationServicesCoreSoftLink.mm:
Paperwork.
* UIProcess/WebAuthentication/Cocoa/AuthenticatorPresenterCoordinator.h:
* UIProcess/WebAuthentication/Cocoa/AuthenticatorPresenterCoordinator.mm:
(WebKit::AuthenticatorPresenterCoordinator::~AuthenticatorPresenterCoordinator):
(WebKit::AuthenticatorPresenterCoordinator::updatePresenter):
(WebKit::AuthenticatorPresenterCoordinator::requestPin):
(WebKit::AuthenticatorPresenterCoordinator::dimissPresenter):
(WebKit::AuthenticatorPresenterCoordinator::setPin):
Adopts new SPIs.
* UIProcess/WebAuthentication/Cocoa/NfcService.mm:
(WebKit::NfcService::platformStartDiscovery):
Tweaks NFC.
* UIProcess/WebAuthentication/Cocoa/WKASCAuthorizationPresenterDelegate.mm:
(-[WKASCAuthorizationPresenterDelegate authorizationPresenter:credentialRequestedForLoginChoice:authenticatedContext:completionHandler:]):
(-[WKASCAuthorizationPresenterDelegate authorizationPresenter:validateUserEnteredPIN:completionHandler:]):
Adopts new SPIs.
2021-01-06 Jer Noble <jer.noble@apple.com>
[Cocoa] Revert audioTimePitchAlgorithm to "TimeDomain" from "Spectral"
https://bugs.webkit.org/show_bug.cgi?id=220341
Reviewed by Youenn Fablet.
Add a private preference to control what pitch correction algorithm will be used by MediaPlayer.
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
(WebKit::RemoteMediaPlayerProxy::setPitchCorrectionAlgorithm):
* GPUProcess/media/RemoteMediaPlayerProxy.h:
* GPUProcess/media/RemoteMediaPlayerProxy.messages.in:
* UIProcess/API/Cocoa/WKPreferences.mm:
(-[WKPreferences _pitchCorrectionAlgorithm]):
(-[WKPreferences _setPitchCorrectionAlgorithm:]):
* UIProcess/API/Cocoa/WKPreferencesPrivate.h:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::setPitchCorrectionAlgorithm):
* WebProcess/GPU/media/MediaPlayerPrivateRemote.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences):
2021-01-06 Yusuke Suzuki <ysuzuki@apple.com>
[JSC] Replace JSBigInt::toUint64 with JSBigInt::toBigUInt64
https://bugs.webkit.org/show_bug.cgi?id=220378
Reviewed by Darin Adler.
* WebProcess/WebPage/IPCTestingAPI.cpp:
(WebKit::IPCTestingAPI::convertToUint64):
(WebKit::IPCTestingAPI::encodeNumericType):
2021-01-06 Andy Estes <aestes@apple.com>
[Mac] Replace most uses of HAVE(MT_PLUGIN_FORMAT_READER) with ENABLE(WEBM_FORMAT_READER)
https://bugs.webkit.org/show_bug.cgi?id=220374
<rdar://problem/72600426>
Reviewed by Eric Carlson.
* Shared/API/c/mac/WKFormatReader.cpp:
(WKFormatReaderCreate):
(WKFormatReaderStartOnMainThread):
* Shared/WebPreferencesDefaultValues.cpp:
* Shared/WebPreferencesDefaultValues.h:
* Shared/mac/MediaFormatReader/CoreMediaWrapped.cpp:
* Shared/mac/MediaFormatReader/CoreMediaWrapped.h:
* Shared/mac/MediaFormatReader/FormatReader.cpp:
* Shared/mac/MediaFormatReader/FormatReader.h:
* Shared/mac/MediaFormatReader/SampleCursor.cpp:
* Shared/mac/MediaFormatReader/SampleCursor.h:
* Shared/mac/MediaFormatReader/TrackReader.cpp:
* Shared/mac/MediaFormatReader/TrackReader.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2021-01-06 Chris Dumez <cdumez@apple.com>
[GPUProcess] RemoteAudioSourceProviderManager should be a WorkQueueMessageReceiver
https://bugs.webkit.org/show_bug.cgi?id=220371
Reviewed by Youenn Fablet.
RemoteAudioSourceProviderManager should be a WorkQueueMessageReceiver (not a ThreadMessageReceiver),
since it uses a WorkQueue (not a WTF::Thread). This simplifies the code a bit.
* WebProcess/GPU/media/RemoteAudioSourceProviderManager.cpp:
(WebKit::RemoteAudioSourceProviderManager::setConnection):
(WebKit::RemoteAudioSourceProviderManager::addProvider):
(WebKit::RemoteAudioSourceProviderManager::removeProvider):
* WebProcess/GPU/media/RemoteAudioSourceProviderManager.h:
2021-01-06 Wenson Hsieh <wenson_hsieh@apple.com>
[Concurrent Display Lists] GPU process should not immediately sleep after reading all available display list items
https://bugs.webkit.org/show_bug.cgi?id=219586
<rdar://problem/72275412>
Reviewed by Chris Dumez.
This patch adds a mechanism for the GPU process to wait for a short duration (~30 microseconds) after it has
finished reading all available data in its shared item buffer; if the web process writes additional data to the
item buffer (thereby bumping the unread bytes counter) during this time, we immediately resume processing the
new display list items in the GPU process, rather than wait for a new wakeup message.
This allows us to avoid the cost of going to sleep just to immediately wake up in the GPU process, in the case
where the web process is writing display list items at a very fast rate and the GPU process just happens to
catch up (i.e. advance `unreadBytes()` to 0).
See below for more details.
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::createRenderingBackend):
Refactor the rendering backend creation message from the web process to the GPU process, so that the rendering
backend creation arguments are encapsulated in a single struct, RemoteRenderingBackendCreationParameters. This
struct contains the rendering backend identifier and, on Cocoa platforms, a mach send right that can be used to
construct the corresponding display list wakeup semaphore in the GPU process.
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/GPUConnectionToWebProcess.messages.in:
* GPUProcess/graphics/DisplayListReaderHandle.h:
(WebKit::DisplayListReaderHandle::startWaiting):
(WebKit::DisplayListReaderHandle::stopWaiting):
Add helper methods for the GPU process to start and stop waiting for new items. See the call site in
RemoteRenderingBackend for more detail, as well as the comments in SharedDisplayListHandle below.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::create):
(WebKit::RemoteRenderingBackend::RemoteRenderingBackend):
Refactor this codepath to take RemoteRenderingBackendCreationParameters instead of just an identifier.
(WebKit::RemoteRenderingBackend::nextDestinationImageBufferAfterApplyingDisplayLists):
In the case where we received the wakeup message due to exceeding the display list item count hysteresis, wait
for a short duration using the semaphore after we hit an unread count of 0. This allows the web process to write
more items and signal the semaphore, so that we can resume reading in the GPU process.
(WebKit::RemoteRenderingBackend::wakeUpAndApplyDisplayList):
(WebKit::RemoteRenderingBackend::setNextItemBufferToRead):
* GPUProcess/graphics/RemoteRenderingBackend.h:
* Shared/GPUProcessWakeupMessageArguments.h:
(WebKit::GPUProcessWakeupMessageArguments::encode const):
(WebKit::GPUProcessWakeupMessageArguments::decode):
Plumb a GPUProcessWakeupReason enum flag over to the GPU process, via wake-up arguments. This flag is used by
the GPU process to determine whether we should expect additional items to eventually enter the stream, and
whether we should eagerly go to sleep after processing all known items. In other words, if the wakeup message is
being sent as a result of flushing the image buffer, we don't want to unnecessarily wait for more items;
however, if the wakeup message is being sent as a result of exceeding the (currently) 512-item hysteresis, then
we can probably expect more items to stream in, so it's more optimal to wait after finishing all known display
list items.
* Shared/RemoteRenderingBackendCreationParameters.h: Copied from Source/WebKit/Shared/GPUProcessWakeupMessageArguments.h.
(WebKit::RemoteRenderingBackendCreationParameters::encode const):
(WebKit::RemoteRenderingBackendCreationParameters::decode):
See above for more details.
* Shared/SharedDisplayListHandle.h:
(WebKit::SharedDisplayListHandle::header const):
Adds a new WaitingStatus enum type internal to SharedDisplayListHandle and its subclasses, which is used to
coordinate the act of waiting for new item data in the GPU process. A shared display list handle now contains
an atomic `waitingStatus` flag indicating whether the GPU process is in the process of waiting for more items,
and also whether the web process has acknowledged the fact that the GPU process is waiting (thereby putting the
GPU process in a state where it is waiting to resume processing). Along with this enum, we also add two new
8-byte values to the header section: an offset to begin reading item data after resuming, and 8 bytes for an
identifier indicating the new destination (for the purposes of display list rendering, this is a
RenderingResourceIdentifier, though this will be different for WebGL).
+--> NotWaiting <--+
| | |
| | | [3a]
| [3b] | [1] |
| | |
Resuming +----> Waiting
^ |
| [2] |
+------------------+
There are three main ways in which this state machine may transition, numbered [1]-[3] in the above diagram.
[1] When the GPU process finishes processing available items, it enters Waiting state, indicating that it is now
waiting for additional item data. This corresponds to the call to `DisplayListReaderHandle::startWaiting`.
[2] When the web process bumps the unread count of an item buffer, if it has a pending wakeup message or would
otherwise need to schedule a pending wakeup message, see if we can instead simply tell the GPU process to
resume processing, instead of sending a wakeup message. This corresponds to a call to the helper method
`DisplayListWriterHandle::tryToResume`.
[3] If the maximum wait duration (~30 microseconds) has passed (i.e. scenario 3a) or if the web process has
transitioned us from Waiting to Resuming state (i.e. scenario 3b), then transition back to NotWaiting. In
both scenarios, this corresponds to a call to `DisplayListReaderHandle::stopWaiting`. In the case where we
transitioned from Resuming state, we can immediately continue processing display list items.
In the case where we successfully resume, we're essentially "re-waking" the GPU process without the overhead of
an additional IPC message, using the same shared display list handle. Since this would've otherwise been a
separate wakeup message, we need to be careful that we continue processing display list items from the correct
offset into the shared display list handle, and using the correct destination image buffer. This is because the
writable offset of the item buffer may have been reset in the middle of waiting by the web process, or the
destination image buffer may have changed while waiting. To handle these scenarios, we introduce the
`ResumeReadingInformation` struct, which contains both of these pieces of information. An instance of this
struct exists in the shared memory header section; this is written by the web process immediately prior to
transitioning to Resuming state, and read by the GPU process when transitioning from Resuming to NotWaiting.
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/graphics/DisplayListWriterHandle.h:
(WebKit::DisplayListWriterHandle::tryToResume):
Add a helper method for the web process to try and notify the GPU process that it should try and resume
display list processing. See the call site in RemoteRenderingBackendProxy for more detail, as well as the
comments in SharedDisplayListHandle above.
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::connectToGPUProcess):
(WebKit::RemoteRenderingBackendProxy::didAppendData):
Rather than always schedule (or send) a wakeup message here in the case where the unread count is 0, see if we
can instead tell the GPU process to stop waiting and resume display list processing. To do this, we use the new
DisplayListWriterHandle::tryToResume method, giving it the new offset to begin reading items from as well as the
destination image buffer to which we should apply display list items. If we successfully tell the GPU process to
resume reading, then we can clear out (or avoid storing) wakeup message arguments and avoid sending an IPC
wakeup message.
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
2021-01-06 Alex Christensen <achristensen@webkit.org>
Add SPI to determine whether a regex is supported in WKContentRuleList
https://bugs.webkit.org/show_bug.cgi?id=220095
<rdar://problem/72058626>
Reviewed by Tim Horton.
This is a more efficient way than making a little rule list and seeing if it compiles.
Here, we don't need access to the disk, and we don't need to hop to another thread and back.
Covered by API tests.
* UIProcess/API/APIContentRuleList.cpp:
(API::ContentRuleList::supportsRegularExpression):
* UIProcess/API/APIContentRuleList.h:
* UIProcess/API/Cocoa/WKContentRuleList.mm:
(+[WKContentRuleList _supportsRegularExpression:]):
* UIProcess/API/Cocoa/WKContentRuleListInternal.h:
* UIProcess/API/Cocoa/WKContentRuleListPrivate.h: Copied from Source/WebKit/UIProcess/API/Cocoa/WKContentRuleListInternal.h.
* WebKit.xcodeproj/project.pbxproj:
2021-01-06 Alex Christensen <achristensen@webkit.org>
Modernize WebContextMenu
https://bugs.webkit.org/show_bug.cgi?id=219969
Reviewed by Tim Horton.
This is old code, complete with a raw pointer.
Use references instead of pointers where possible,
and WeakPtr instead of a raw pointer.
* WebProcess/InjectedBundle/API/c/WKBundlePage.cpp:
(WKBundlePageClickMenuItem):
(WKBundlePageCopyContextMenuItems):
* WebProcess/WebCoreSupport/WebContextMenuClient.cpp:
(WebKit::WebContextMenuClient::showContextMenu):
* WebProcess/WebPage/WebContextMenu.cpp:
(WebKit::WebContextMenu::WebContextMenu):
* WebProcess/WebPage/WebContextMenu.h:
(WebKit::WebContextMenu::create):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::contextMenu):
(WebKit::WebPage::contextMenuAtPointInWindow):
(WebKit::handleContextMenuEvent):
(WebKit::WebPage::contextMenuForKeyEvent):
2021-01-06 Sihui Liu <sihui_liu@appe.com>
Stop speech recognition if page becomes invisible
https://bugs.webkit.org/show_bug.cgi?id=220073
<rdar://problem/72710704>
Reviewed by Youenn Fablet.
We should not allow speech recognition on page invisible to user, because page may get content of captured audio
without user's notice. To do this, we should:
1. deny speech recognition request on invisibile page
2. abort ongoing speech recognition if page becomes invisible
API test: WebKit2.SpeechRecognitionPageBecomesInvisible
* UIProcess/SpeechRecognitionPermissionManager.cpp:
(WebKit::SpeechRecognitionPermissionManager::continueProcessingRequest):
* UIProcess/SpeechRecognitionServer.cpp:
(WebKit::SpeechRecognitionServer::abortForPageIsBecomingInvisible):
* UIProcess/SpeechRecognitionServer.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::dispatchActivityStateChange):
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::pageIsBecomingInvisible):
* UIProcess/WebProcessProxy.h:
2021-01-06 Alex Christensen <achristensen@webkit.org>
Use sendWithAsyncReply instead of iOS-specific GenericCallbacks
https://bugs.webkit.org/show_bug.cgi?id=220346
Reviewed by Darin Adler.
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _accessibilityRetrieveRectsEnclosingSelectionOffset:withGranularity:]):
(-[WKContentView _accessibilityRetrieveRectsAtSelectionOffset:withText:completionHandler:]):
(selectionChangedWithGesture):
(selectionChangedWithTouch):
(-[WKContentView changeSelectionWithGestureAt:withGesture:withState:withFlags:]):
(-[WKContentView changeSelectionWithTouchAt:withSelectionTouch:baseIsStart:withFlags:]):
(-[WKContentView changeSelectionWithTouchesFrom:to:withGesture:withState:]):
(-[WKContentView updateCurrentFocusedElementInformation:]):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::requestFocusedElementInformation):
(WebKit::WebPageProxy::selectWithGesture):
(WebKit::WebPageProxy::updateSelectionWithTouches):
(WebKit::WebPageProxy::selectWithTwoTouches):
(WebKit::WebPageProxy::requestRectsForGranularityWithSelectionOffset):
(WebKit::WebPageProxy::requestRectsAtSelectionOffsetWithText):
(WebKit::WebPageProxy::gestureCallback): Deleted.
(WebKit::WebPageProxy::touchesCallback): Deleted.
(WebKit::WebPageProxy::selectionRectsCallback): Deleted.
(WebKit::WebPageProxy::focusedElementInformationCallback): Deleted.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::requestFocusedElementInformation):
(WebKit::WebPage::selectWithGesture):
(WebKit::WebPage::updateSelectionWithTouches):
(WebKit::WebPage::selectWithTwoTouches):
(WebKit::WebPage::getRectsForGranularityWithSelectionOffset):
(WebKit::WebPage::getRectsAtSelectionOffsetWithText):
2021-01-06 Per Arne Vollan <pvollan@apple.com>
[iOS] Silence selected sandbox violation reports
https://bugs.webkit.org/show_bug.cgi?id=220320
Reviewed by Darin Adler.
For performance reasons, benign sandbox violations in the WebContent process can be silenced on iOS.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2021-01-06 Per Arne Vollan <pvollan@apple.com>
[iOS] Add missing endpoint in sandbox message filter
https://bugs.webkit.org/show_bug.cgi?id=219731
<rdar://problem/71566760>
Reviewed by Darin Adler.
Add missing endpoint in the WebContent process' sandbox message filter on iOS.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2021-01-05 Eric Carlson <eric.carlson@apple.com>
[Cocoa] WebM format reader doesn't work with a url in a <source> element
https://bugs.webkit.org/show_bug.cgi?id=219961
<rdar://problem/72399014>
Reviewed by Andy Estes.
Only enable a WebM track when the it has FlagEnabled element, or when we see that
it has media samples. This is necessary because there are WebM files with empty tracks,
and CoreMedia won't play a file if any enabled track doesn't have samples.
* Shared/mac/MediaFormatReader/FormatReader.cpp:
(WebKit::FormatReader::didParseTracks):
* Shared/mac/MediaFormatReader/TrackReader.cpp:
(WebKit::TrackReader::create):
(WebKit::TrackReader::TrackReader):
(WebKit::TrackReader::finishParsing):
(WebKit::TrackReader::isEnabled const):
(WebKit::TrackReader::copyProperty):
* Shared/mac/MediaFormatReader/TrackReader.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2021-01-05 Wenson Hsieh <wenson_hsieh@apple.com>
Refactor some logic around touch event deferring gesture recognizers
https://bugs.webkit.org/show_bug.cgi?id=220345
Reviewed by Tim Horton.
Rename `_deferringGestureRecognizerForFoo` to `_touchStartDeferringGestureRecognizerForFoo`, to make it clear
that these deferring gestures are only intended to defer native gestures until the "touchstart" event has been
handled.
This distinction becomes important once I introduce deferring gesture recognizers that defer the "touchend"
event, and prevent gestures that activate upon touch end (i.e. various types of tap gesture recognizers).
See below for more details.
* UIProcess/PageClient.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::handlePreventableTouchEvent):
* UIProcess/ios/PageClientImplIOS.h:
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::doneDeferringTouchStart):
(WebKit::PageClientImpl::doneDeferringNativeGestures): Deleted.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView setUpInteraction]):
(-[WKContentView cleanUpInteraction]):
(-[WKContentView _removeDefaultGestureRecognizers]):
(-[WKContentView _addDefaultGestureRecognizers]):
(-[WKContentView _webTouchEventsRecognized:]):
(-[WKContentView _touchStartDeferringGestures]):
(-[WKContentView _doneDeferringTouchStart:]):
(-[WKContentView gestureRecognizer:shouldRecognizeSimultaneouslyWithGestureRecognizer:]):
(-[WKContentView deferringGestureRecognizer:shouldDeferOtherGestureRecognizer:]):
(-[WKContentView _deferringGestureRecognizers]): Deleted.
(-[WKContentView _doneDeferringNativeGestures:]): Deleted.
Rename this to `-_doneDeferringTouchStart:`.
(-[WKContentView deferringGestureRecognizer:shouldDeferGesturesAfterEndingTouchesWithEvent:]):
Deleted. This logic has been moved to WKContentView (see below).
* UIProcess/ios/WKDeferringGestureRecognizer.h:
* UIProcess/ios/WKDeferringGestureRecognizer.mm:
(-[WKDeferringGestureRecognizer touchesEnded:withEvent:]): Deleted.
Additionally refactor this code so that this fallback logic for lifting the gesture gate exists inside
`WKContentView` (after ending all touches), rather than inside the deferring gesture recognizer. While this
doesn't affect the extant touch start deferring gestures, it will ensure that touch end deferring gestures won't
get reset prematurely due to the web touch events gesture recognizer ending before the touch end deferring
gesture.
2021-01-05 Kate Cheney <katherine_cheney@apple.com>
NSCrossWebsiteTrackingUsageDescription is not working on Mac, ITP is always enabled
https://bugs.webkit.org/show_bug.cgi?id=220190
<rdar://problem/72744909>
Reviewed by Brent Fulgham.
Consider the presence of the NSCrossWebsiteTrackingUsageDescription
key as an indication to disable ITP for WKWebView until we implement
a mechanism for disabling it in WKWebView like we do for iOS to avoid
compatibility bugs.
* Shared/Cocoa/DefaultWebBrowserChecks.mm:
(WebKit::determineITPStateInternal):
2021-01-05 Aditya Keerthi <akeerthi@apple.com>
REGRESSION (r261157): Crash in WKSelectPopover when running as iPhone app on iPad
https://bugs.webkit.org/show_bug.cgi?id=220065
<rdar://problem/71932792>
Reviewed by Darin Adler.
r261157 changed how WebKit determined the user interface idiom for
WKWebViews that were created in daemons. Since daemons do not create
UIApplications, WebKit cannot use UIDevice in these instances. Instead,
the user interface idiom determination logic was updated to include
a check for device hardware, using MobileGestalt, in cases where no
UIApplication was created.
Since WebKit only determines the user interface idiom once (then
storing the obtained value), the added determination logic breaks
down for iPhone apps on iPad. Consider the following sequence of
events, eventually leading to a crash when interacting with a
<select> element:
1. A WKWebView is created prior to UIApplication initialization. This
can be achieved by creating one in another class's "load" method.
2. Since the app is physically running on an iPad, WebKit determines
the user interface idiom to be "iPad" and saves this information.
3. Once the app actually launches, UIApplication is initialized.
Since this is an iPhone app on iPad, the actual user interface
idiom is "iPhone". However, WebKit does not know this, since
it uses the saved user interface idiom from (2).
4. When tapping a <select> element, WebKit checks its saved idiom,
an attempts to present a UIPopoverController (the standard behavior
under the iPad idiom).
5. However, since the actual idiom is iPhone, UIKit throws an
NSInvalidArgumentException, since UIPopoverController
should not be used when running under the iPhone idiom.
A simple fix for the crash would be to call into UIKit to check the
idiom each time it is required, rather than relying on the saved bit.
However, this does not address the more general issue, which is that
WebKit's saved idiom and the actual idiom can be out of sync.
Consequently, the approach taken is to update the saved idiom when
it changes. This should only occur when a UIApplication is initialized.
Hence, we listen for UIApplicationDidFinishLaunchingNotification,
and if the actual idiom is different from our saved idiom, we update the
saved idiom and also notify the WebProcess.
* Shared/UserInterfaceIdiom.h:
* Shared/UserInterfaceIdiom.mm:
(WebKit::updateCurrentUserInterfaceIdiom):
This method checks the actual idiom and updates the saved idiom if
they differ. Returns true only if the idiom was updated.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::registerNotificationObservers):
Listen for UIApplicationDidFinishLaunchingNotification if the idiom
can change (UIApplication is uninitialized). Then, if the actual
and saved idiom differ, notify all WebProcesses of the change.
(WebKit::WebProcessPool::unregisterNotificationObservers):
* UIProcess/WebProcessPool.h:
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::userInterfaceIdiomDidChange):
Update the saved idiom using the information from the UIProcess.
2021-01-05 Chris Dumez <cdumez@apple.com>
[iOS] Add a feature flag to stop leaking an XPC boost message to XPC services
https://bugs.webkit.org/show_bug.cgi?id=219453
<rdar://72834999>
Reviewed by Geoff Garen.
Add a feature flag to stop leaking an XPC boost message to XPC services on builds where RunningBoard is
able to give the right priorities to our XPC services based on the type of RunningBoard assertion is being
held. There is no behavior change for the Network/WebContent processes since those were blocklisted in
RunningBoard (Turning on the RB_full_manage_WK_jetsam feature flag gives us the default RunningBoard
behavior without blocklisting). However, this allows the new GPUProcess to move to IDLE jetsam band when
homing out of MobileSafari, which makes the GPUProcess eligible for the freezer. Previously, the leaked
boost message would keep the GPUProcess in a higher jetsam band.
For now, the feature flag is off by default. I will turn it on later on.
* FeatureFlags/WebKit.plist:
Add RB_full_manage_WK_jetsam feature flag but keep it off for now. On recent enough builds of RunningBoard, this
flag disables the blocklisting of the WebContent/Network processes in RunningBoard so that RunningBoard can now
set the priority of those processes (instead of us boosting our priority via boost messages).
* GPUProcess/EntryPoint/Cocoa/XPCService/GPUService/Info-iOS.plist:
* NetworkProcess/EntryPoint/Cocoa/XPCService/NetworkService/Info-iOS.plist:
* WebAuthnProcess/EntryPoint/Cocoa/XPCService/WebAuthnService/Info-iOS.plist:
* WebProcess/EntryPoint/Cocoa/XPCService/WebContentService/Info-iOS.plist:
Set _ProcessType=App so that RunningBoard is able to set the priority of our XPC services on iOS. Without
this change, we get a permission denied when RunningBoard attempts to set the priority of our XPC services.
* UIProcess/Launcher/mac/ProcessLauncherMac.mm:
(WebKit::shouldLeakBoost):
Stop leaking a boost message on iOS when RUNNINGBOARD_WEBKIT_PRIORITY_SUPPORT compile-time flag is enabled
and when the RB_full_manage_WK_jetsam feature flag is enabled.
2021-01-05 Alex Christensen <achristensen@webkit.org>
Use sendWithAsyncReply instead of AttributedStringForCharacterRangeCallback and FontAtSelectionCallback
https://bugs.webkit.org/show_bug.cgi?id=220344
Reviewed by Geoffrey Garen.
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::updateFontManagerIfNeeded):
(WebKit::WebViewImpl::attributedSubstringForProposedRange):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::attributedSubstringForCharacterRangeAsync):
(WebKit::WebPageProxy::fontAtSelection):
(WebKit::WebPageProxy::attributedStringForCharacterRangeCallback): Deleted.
(WebKit::WebPageProxy::fontAtSelectionCallback): Deleted.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::attributedSubstringForCharacterRangeAsync):
(WebKit::WebPage::fontAtSelection):
2021-01-05 John Wilander <wilander@apple.com>
PCM: Experimental debug mode stops working after initial use
https://bugs.webkit.org/show_bug.cgi?id=220336
<rdar://problem/72398086>
Reviewed by Brent Fulgham.
The existing experimental PCM debug mode uses
RuntimeEnabledFeatures::sharedFeatures().privateClickMeasurementDebugModeEnabled()
which is not correct in the network process. This makes the flag to lose its
state after navigations in new tabs.
This patch moves the flag to be alongside the PCM feature flag in
WebKit::NetworkProcess.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::setPrivateClickMeasurementDebugMode):
(WebKit::NetworkProcess::privateClickMeasurementDebugModeEnabled const):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/PrivateClickMeasurementManager.cpp:
(WebKit::PrivateClickMeasurementManager::debugModeEnabled const):
2021-01-05 Kimmo Kinnunen <kkinnunen@apple.com>
WebKit IPC generator tests have duplicate receiver names, limiting the implementation
https://bugs.webkit.org/show_bug.cgi?id=220236
Reviewed by Chris Dumez.
IPC test message receivers defined multiple receivers with same name.
Rename the test message receivers based on what feature they test. No
change in the non-test code.
The IPC works with message identifier defined by the MessageName type.
Having multiple clashing receiver names is problematic because a MessageName
is defined by pair <receiver name, message name>. Since different receivers
might have different kinds of messages with same message name, these might end
up mapping to two conflicting message kinds.
The rename of test classes makes the tests conform to the already used contract
that the MessageName is globally unique to each individual conceptual message
because the receiver names are unique.
This enables further implementations that make decisions based on the
message name. Example is to decide whether a message is synchronous or
not. This cannot be implemented generally if there is multiple different kinds of
messages with same name.
Rename the test message receiver files based on the receiver name. This is
more consistent with the existing convention used in the actual code.
Move the script test results from the Scripts/ directory to a tests/ subdirectory
of the tool. This is to not fill the Scripts/ directory with Scripts/webkit/ related
test result data.
Adds previously missing test result file MessageArgumentDescriptions.cpp, which was not added
when the feature was implemented.
Tested by existing tests and new model_unittest.py.
* Scripts/Makefile: Removed.
* Scripts/MessageNames.cpp: Removed.
* Scripts/MessageNames.h: Removed.
* Scripts/generate-message-receiver.py:
(main):
* Scripts/test-legacyMessageReceiver.cpp: Removed.
* Scripts/testMessageReceiver.cpp: Removed.
* Scripts/webkit/messages.py:
* Scripts/webkit/messages_unittest.py:
(ParsingTest.test_receiver):
(GeneratedFileContentsTest.assertGeneratedFileContentsEqual):
* Scripts/webkit/model.py:
(Parameter.has_attribute):
(check_global_model):
* Scripts/webkit/model_unittest.py: Added.
(ModelCheckTest):
(ModelCheckTest.test_duplicate_receivers):
(test_mismatch_message_attribute_sync):
* Scripts/webkit/tests/Makefile: Added.
* Scripts/webkit/tests/MessageArgumentDescriptions.cpp: Added.
(IPC::jsValueForArguments):
(IPC::jsValueForReplyArguments):
(IPC::messageArgumentDescriptions):
(IPC::messageReplyArgumentDescriptions):
(IPC::messageIsSync):
* Scripts/webkit/tests/MessageNames.cpp: Added.
(IPC::description):
(IPC::receiverName):
(IPC::isValidMessageName):
* Scripts/webkit/tests/MessageNames.h: Added.
(WTF::isValidEnum):
* Scripts/webkit/tests/TestWithLegacyReceiver.messages.in: Renamed from Source/WebKit/Scripts/webkit/test-legacy.messages.in.
* Scripts/webkit/tests/TestWithLegacyReceiverMessageReceiver.cpp: Added.
(Messages::TestWithLegacyReceiver::GetPluginProcessConnection::send):
(Messages::TestWithLegacyReceiver::TestMultipleAttributes::send):
(WebKit::TestWithLegacyReceiver::didReceiveTestWithLegacyReceiverMessage):
(WebKit::TestWithLegacyReceiver::didReceiveSyncTestWithLegacyReceiverMessage):
* Scripts/webkit/tests/TestWithLegacyReceiverMessages.h: Renamed from Source/WebKit/Scripts/test-legacyMessages.h.
(Messages::TestWithLegacyReceiver::messageReceiverName):
(Messages::TestWithLegacyReceiver::LoadURL::name):
(Messages::TestWithLegacyReceiver::LoadSomething::name):
(Messages::TestWithLegacyReceiver::TouchEvent::name):
(Messages::TestWithLegacyReceiver::AddEvent::name):
(Messages::TestWithLegacyReceiver::LoadSomethingElse::name):
(Messages::TestWithLegacyReceiver::DidReceivePolicyDecision::name):
(Messages::TestWithLegacyReceiver::Close::name):
(Messages::TestWithLegacyReceiver::PreferencesDidChange::name):
(Messages::TestWithLegacyReceiver::SendDoubleAndFloat::name):
(Messages::TestWithLegacyReceiver::SendInts::name):
(Messages::TestWithLegacyReceiver::CreatePlugin::name):
(Messages::TestWithLegacyReceiver::RunJavaScriptAlert::name):
(Messages::TestWithLegacyReceiver::GetPlugins::name):
(Messages::TestWithLegacyReceiver::GetPluginProcessConnection::name):
(Messages::TestWithLegacyReceiver::TestMultipleAttributes::name):
(Messages::TestWithLegacyReceiver::TestParameterAttributes::name):
(Messages::TestWithLegacyReceiver::TemplateTest::name):
(Messages::TestWithLegacyReceiver::SetVideoLayerID::name):
(Messages::TestWithLegacyReceiver::DidCreateWebProcessConnection::name):
(Messages::TestWithLegacyReceiver::InterpretKeyEvent::name):
(Messages::TestWithLegacyReceiver::DeprecatedOperation::name):
(Messages::TestWithLegacyReceiver::ExperimentalOperation::name):
* Scripts/webkit/tests/TestWithLegacyReceiverMessagesReplies.h: Renamed from Source/WebKit/Scripts/testMessagesReplies.h.
* Scripts/webkit/tests/TestWithSuperclass.messages.in: Renamed from Source/WebKit/Scripts/webkit/test-superclass.messages.in.
* Scripts/webkit/tests/TestWithSuperclassMessageReceiver.cpp: Renamed from Source/WebKit/Scripts/test-superclassMessageReceiver.cpp.
(WebKit::TestWithSuperclass::didReceiveMessage):
(WebKit::TestWithSuperclass::didReceiveSyncMessage):
* Scripts/webkit/tests/TestWithSuperclassMessages.h: Renamed from Source/WebKit/Scripts/test-superclassMessages.h.
(Messages::TestWithSuperclass::messageReceiverName):
(Messages::TestWithSuperclass::LoadURL::name):
(Messages::TestWithSuperclass::TestAsyncMessage::name):
(Messages::TestWithSuperclass::TestAsyncMessage::asyncMessageReplyName):
(Messages::TestWithSuperclass::TestAsyncMessageWithNoArguments::name):
(Messages::TestWithSuperclass::TestAsyncMessageWithNoArguments::asyncMessageReplyName):
(Messages::TestWithSuperclass::TestAsyncMessageWithMultipleArguments::name):
(Messages::TestWithSuperclass::TestAsyncMessageWithMultipleArguments::asyncMessageReplyName):
(Messages::TestWithSuperclass::TestAsyncMessageWithConnection::name):
(Messages::TestWithSuperclass::TestAsyncMessageWithConnection::asyncMessageReplyName):
(Messages::TestWithSuperclass::TestSyncMessage::name):
(Messages::TestWithSuperclass::TestSynchronousMessage::name):
* Scripts/webkit/tests/TestWithSuperclassMessagesReplies.h: Renamed from Source/WebKit/Scripts/test-superclassMessagesReplies.h.
* Scripts/webkit/tests/TestWithoutAttributes.messages.in: Renamed from Source/WebKit/Scripts/webkit/test.messages.in.
* Scripts/webkit/tests/TestWithoutAttributesMessageReceiver.cpp: Added.
(Messages::TestWithoutAttributes::GetPluginProcessConnection::send):
(Messages::TestWithoutAttributes::TestMultipleAttributes::send):
(WebKit::TestWithoutAttributes::didReceiveMessage):
(WebKit::TestWithoutAttributes::didReceiveSyncMessage):
* Scripts/webkit/tests/TestWithoutAttributesMessages.h: Renamed from Source/WebKit/Scripts/testMessages.h.
(Messages::TestWithoutAttributes::messageReceiverName):
(Messages::TestWithoutAttributes::LoadURL::name):
(Messages::TestWithoutAttributes::LoadSomething::name):
(Messages::TestWithoutAttributes::TouchEvent::name):
(Messages::TestWithoutAttributes::AddEvent::name):
(Messages::TestWithoutAttributes::LoadSomethingElse::name):
(Messages::TestWithoutAttributes::DidReceivePolicyDecision::name):
(Messages::TestWithoutAttributes::Close::name):
(Messages::TestWithoutAttributes::PreferencesDidChange::name):
(Messages::TestWithoutAttributes::SendDoubleAndFloat::name):
(Messages::TestWithoutAttributes::SendInts::name):
(Messages::TestWithoutAttributes::CreatePlugin::name):
(Messages::TestWithoutAttributes::RunJavaScriptAlert::name):
(Messages::TestWithoutAttributes::GetPlugins::name):
(Messages::TestWithoutAttributes::GetPluginProcessConnection::name):
(Messages::TestWithoutAttributes::TestMultipleAttributes::name):
(Messages::TestWithoutAttributes::TestParameterAttributes::name):
(Messages::TestWithoutAttributes::TemplateTest::name):
(Messages::TestWithoutAttributes::SetVideoLayerID::name):
(Messages::TestWithoutAttributes::DidCreateWebProcessConnection::name):
(Messages::TestWithoutAttributes::InterpretKeyEvent::name):
(Messages::TestWithoutAttributes::DeprecatedOperation::name):
(Messages::TestWithoutAttributes::ExperimentalOperation::name):
* Scripts/webkit/tests/TestWithoutAttributesMessagesReplies.h: Renamed from Source/WebKit/Scripts/test-legacyMessagesReplies.h.
2021-01-05 Tim Horton <timothy_horton@apple.com>
SourcesCocoa.txt should not be installed in the Web Content service bundle
https://bugs.webkit.org/show_bug.cgi?id=220338
Reviewed by Wenson Hsieh.
* WebKit.xcodeproj/project.pbxproj:
Remove SourcesCocoa.txt from the target.
2021-01-05 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process][MSE] SourceBuffer fires update and updateend events before the coded frames are removed
https://bugs.webkit.org/show_bug.cgi?id=220334
Reviewed by Eric Carlson.
Update the IPC message RemoteSourceBufferProxy::RemoveCodedFrames to implement
SourceBufferPrivateRemote::removeCodedFrames().
* GPUProcess/media/RemoteSourceBufferProxy.cpp:
(WebKit::RemoteSourceBufferProxy::removeCodedFrames):
* GPUProcess/media/RemoteSourceBufferProxy.h:
* GPUProcess/media/RemoteSourceBufferProxy.messages.in:
* WebProcess/GPU/media/SourceBufferPrivateRemote.cpp:
(WebKit::SourceBufferPrivateRemote::removeCodedFrames):
* WebProcess/GPU/media/SourceBufferPrivateRemote.h:
2021-01-05 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed, reverting r271169.
Caused
TestWebKitAPI.WebKit2.SpeechRecognitionErrorWhenStartingAudioCaptureOnDifferentPage
to time out on macOS.
Reverted changeset:
"Stop speech recognition if page becomes invisible"
https://bugs.webkit.org/show_bug.cgi?id=220073
https://trac.webkit.org/changeset/271169
2021-01-05 Alex Christensen <achristensen@webkit.org>
Make WebPage::ForceRepaint use CompletionHandler instead of VoidCallback
https://bugs.webkit.org/show_bug.cgi?id=212269
Reviewed by Chris Dumez.
Fix a few unsafe pointer uses along the way.
No change in behavior.
* UIProcess/API/C/WKPage.cpp:
(WKPageForceRepaint):
* UIProcess/ViewGestureController.cpp:
(WebKit::ViewGestureController::forceRepaintIfNeeded):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::forceRepaint):
* UIProcess/WebPageProxy.h:
* UIProcess/ios/fullscreen/WKFullScreenWindowControllerIOS.mm:
(-[WKFullScreenWindowController enterFullScreen]):
(-[WKFullScreenWindowController _completedExitFullScreen]):
* UIProcess/mac/WKFullScreenWindowController.h:
* UIProcess/mac/WKFullScreenWindowController.mm:
(-[WKFullScreenWindowController dealloc]):
(-[WKFullScreenWindowController finishedExitFullScreenAnimation:]):
(-[WKFullScreenWindowController completeFinishExitFullScreenAnimationAfterRepaint]):
* WebProcess/WebPage/CoordinatedGraphics/DrawingAreaCoordinatedGraphics.cpp:
(WebKit::DrawingAreaCoordinatedGraphics::forceRepaintAsync):
* WebProcess/WebPage/CoordinatedGraphics/DrawingAreaCoordinatedGraphics.h:
* WebProcess/WebPage/CoordinatedGraphics/LayerTreeHost.cpp:
(WebKit::LayerTreeHost::layerFlushTimerFired):
(WebKit::LayerTreeHost::forceRepaintAsync):
(WebKit::LayerTreeHost::renderNextFrame):
* WebProcess/WebPage/CoordinatedGraphics/LayerTreeHost.h:
(WebKit::LayerTreeHost::forceRepaintAsync):
* WebProcess/WebPage/DrawingArea.h:
(WebKit::DrawingArea::forceRepaintAsync):
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::forceRepaint):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.h:
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::forceRepaintAsync):
2021-01-05 Sihui Liu <sihui_liu@appe.com>
Stop speech recognition if page becomes invisible
https://bugs.webkit.org/show_bug.cgi?id=220073
<rdar://problem/72710704>
Reviewed by Youenn Fablet.
We should not allow speech recognition on page invisible to user, because page may get content of captured audio
without user's notice. To do this, we should:
1. deny speech recognition request on invisibile page
2. abort ongoing speech recognition if page becomes invisible
API test: WebKit2.SpeechRecognitionPageBecomesInvisible
* UIProcess/SpeechRecognitionPermissionManager.cpp:
(WebKit::SpeechRecognitionPermissionManager::continueProcessingRequest):
* UIProcess/SpeechRecognitionServer.cpp:
(WebKit::SpeechRecognitionServer::abortForPageIsBecomingInvisible):
* UIProcess/SpeechRecognitionServer.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::dispatchActivityStateChange):
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::pageIsBecomingInvisible):
* UIProcess/WebProcessProxy.h:
2021-01-05 Alex Christensen <achristensen@webkit.org>
Use sendWithAsyncReply instead of ValidateCommandCallback
https://bugs.webkit.org/show_bug.cgi?id=220124
Reviewed by Chris Dumez.
* UIProcess/API/C/WKPage.cpp:
(WKPageValidateCommand):
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::validateUserInterfaceItem):
* UIProcess/WebCookieManagerProxy.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::validateCommand):
(WebKit::WebPageProxy::validateCommandCallback): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::~WebProcessPool):
* UIProcess/WebProcessPool.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::validateCommand):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2021-01-05 Alex Christensen <achristensen@webkit.org>
Use sendWithAsyncReply instead of FontAttributesCallback
https://bugs.webkit.org/show_bug.cgi?id=220123
Reviewed by Chris Dumez.
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::typingAttributesWithCompletionHandler):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::requestFontAttributesAtSelectionStart):
(WebKit::WebPageProxy::fontAttributesCallback): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::requestFontAttributesAtSelectionStart):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2021-01-05 Alex Christensen <achristensen@webkit.org>
Use sendWithAsyncReply instead of RectForCharacterRangeCallback
https://bugs.webkit.org/show_bug.cgi?id=220125
Reviewed by Chris Dumez.
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::firstRectForCharacterRange):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::firstRectForCharacterRangeAsync):
(WebKit::WebPageProxy::rectForCharacterRangeCallback): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::firstRectForCharacterRangeAsync):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2021-01-05 Sihui Liu <sihui_liu@appe.com>
Fail speech recognition when page is muted for audio capture
https://bugs.webkit.org/show_bug.cgi?id=220133
<rdar://problem/72745232>
Reviewed by Youenn Fablet.
We currently only allow one page to capture media at a time and we did this by muting (stop capture in) other
pages. To make speech recognition work with this behavior, two changes are made:
1. when page is muted, mute audio capture source used for speech recognition on the page. This will
ultimately fail recognition.
2. when speech recognition is about to start, make sure other pages are muted for capture.
* UIProcess/SpeechRecognitionPermissionManager.h:
(WebKit::SpeechRecognitionPermissionManager::page):
* UIProcess/SpeechRecognitionServer.cpp:
(WebKit::SpeechRecognitionServer::handleRequest):
(WebKit::SpeechRecognitionServer::mute):
* UIProcess/SpeechRecognitionServer.h:
* UIProcess/UserMediaProcessManager.cpp:
(WebKit::UserMediaProcessManager::muteCaptureMediaStreamsExceptIn): Deleted.
* UIProcess/UserMediaProcessManager.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::activateMediaStreamCaptureInPage):
(WebKit::WebPageProxy::setMuted):
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::muteCaptureInPagesExcept):
(WebKit::WebProcessProxy::pageMutedStateChanged):
* UIProcess/WebProcessProxy.h:
2021-01-05 Alex Christensen <achristensen@webkit.org>
Use WeakHashSet<Page> instead of HashSet<Page*>
https://bugs.webkit.org/show_bug.cgi?id=220102
Reviewed by Darin Adler.
* WebProcess/InjectedBundle/InjectedBundle.cpp:
(WebKit::InjectedBundle::setAsynchronousSpellCheckingEnabled):
(WebKit::InjectedBundle::setUserStyleSheetLocation):
(WebKit::InjectedBundle::liveDocumentURLs):
* WebProcess/WebStorage/StorageAreaMap.cpp:
(WebKit::StorageAreaMap::dispatchLocalStorageEvent):
2021-01-05 Alex Christensen <achristensen@webkit.org>
Use sendWithAsyncReply instead of SelectionContextCallback
https://bugs.webkit.org/show_bug.cgi?id=220121
Reviewed by Chris Dumez.
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _lookupForWebView:]):
(-[WKContentView requestDictationContext:]):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::requestDictationContext):
(WebKit::WebPageProxy::getSelectionContext):
(WebKit::WebPageProxy::selectionContextCallback): Deleted.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::getSelectionContext):
(WebKit::WebPage::requestDictationContext):
2021-01-05 Youenn Fablet <youenn@apple.com>
Enable WebRTC codecs in GPU process only for --use-gpu-process option
https://bugs.webkit.org/show_bug.cgi?id=220316
Reviewed by Alexey Proskuryakov.
ADD C SPI used by TestController.
Manually tested.
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetWebRTCPlatformCodecsInGPUProcessEnabled):
(WKPreferencesGetWebRTCPlatformCodecsInGPUProcessEnabled):
* UIProcess/API/C/WKPreferencesRefPrivate.h:
2021-01-04 Jeff Miller <jeffm@apple.com>
Update user-visible copyright strings to include 2021
https://bugs.webkit.org/show_bug.cgi?id=219901
Reviewed by Anders Carlsson.
* GPUProcess/EntryPoint/Cocoa/XPCService/GPUService/Info-OSX.plist:
* GPUProcess/EntryPoint/Cocoa/XPCService/GPUService/Info-iOS.plist:
* Info.plist:
* NetworkProcess/EntryPoint/Cocoa/XPCService/NetworkService/Info-OSX.plist:
* NetworkProcess/EntryPoint/Cocoa/XPCService/NetworkService/Info-iOS.plist:
* PluginProcess/EntryPoint/Cocoa/XPCService/PluginService.Info.plist:
* WebAuthnProcess/EntryPoint/Cocoa/XPCService/WebAuthnService/Info-OSX.plist:
* WebAuthnProcess/EntryPoint/Cocoa/XPCService/WebAuthnService/Info-iOS.plist:
* WebProcess/EntryPoint/Cocoa/XPCService/WebContentService/Info-OSX.plist:
* WebProcess/EntryPoint/Cocoa/XPCService/WebContentService/Info-iOS.plist:
2021-01-02 Rob Buis <rbuis@igalia.com>
Remove some FrameLoader::changeLocation parameters
https://bugs.webkit.org/show_bug.cgi?id=220186
Reviewed by Darin Adler.
Adapt to API change.
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::clickedLink):
* WebProcess/WebCoreSupport/WebContextMenuClient.cpp:
(WebKit::WebContextMenuClient::searchWithGoogle):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::navigateToPDFLinkWithSimulatedClick):
2020-12-23 Said Abou-Hallawa <said@apple.com>
[GPU Process]: Don't call flushDrawingContext() in the middle of recording drawing commands
https://bugs.webkit.org/show_bug.cgi?id=220126
Reviewed by Simon Fraser.
There is no need to replay back the DrawImageBuffer immediately after it
was recorded. The worry was the source ImageBuffer may change before
replaying back the DrawImageBuffer into the destination ImageBuffer. But
this should not happen with the shared memory buffer because the order
of drawing among multiple ImageBuffer is preserved.
* Shared/RemoteLayerTree/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::display):
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
2020-12-23 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process][MSE] Clean up SourceBufferPrivateRemote and implement functions with new IPC messages
https://bugs.webkit.org/show_bug.cgi?id=220122
Reviewed by Eric Carlson.
Add IPC messages and handlers in `RemoteSourceBufferProxy` to implement the added functions of `SourceBufferPrivateRemote`.
Also add an IPC message to forward the `HighestPresentationTimestampChanged` callback from `RemoteSourceBufferProxy`
to `SourceBufferPrivateRemote`.
* GPUProcess/media/RemoteSourceBufferProxy.cpp:
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateHighestPresentationTimestampChanged):
(WebKit::RemoteSourceBufferProxy::setMediaSourceEnded):
(WebKit::RemoteSourceBufferProxy::setMode):
(WebKit::RemoteSourceBufferProxy::startChangingType):
(WebKit::RemoteSourceBufferProxy::removeCodedFrames):
(WebKit::RemoteSourceBufferProxy::setAllTrackBuffersNeedRandomAccess):
(WebKit::RemoteSourceBufferProxy::setGroupStartTimestamp):
(WebKit::RemoteSourceBufferProxy::setGroupStartTimestampToEndTimestamp):
(WebKit::RemoteSourceBufferProxy::setShouldGenerateTimestamps):
* GPUProcess/media/RemoteSourceBufferProxy.h:
* GPUProcess/media/RemoteSourceBufferProxy.messages.in:
* Scripts/webkit/messages.py:
* WebProcess/GPU/media/SourceBufferPrivateRemote.cpp:
(WebKit::SourceBufferPrivateRemote::setMediaSourceEnded):
(WebKit::SourceBufferPrivateRemote::setMode):
(WebKit::SourceBufferPrivateRemote::removeCodedFrames):
(WebKit::SourceBufferPrivateRemote::setAllTrackBuffersNeedRandomAccess):
(WebKit::SourceBufferPrivateRemote::setGroupStartTimestamp):
(WebKit::SourceBufferPrivateRemote::setGroupStartTimestampToEndTimestamp):
(WebKit::SourceBufferPrivateRemote::setShouldGenerateTimestamps):
(WebKit::SourceBufferPrivateRemote::startChangingType):
(WebKit::SourceBufferPrivateRemote::sourceBufferPrivateHighestPresentationTimestampChanged):
* WebProcess/GPU/media/SourceBufferPrivateRemote.h:
* WebProcess/GPU/media/SourceBufferPrivateRemote.messages.in:
2020-12-23 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process][MSE] Media element does not fire "resize" event
https://bugs.webkit.org/show_bug.cgi?id=220119
Reviewed by Eric Carlson.
Implement `RemoteMediaPlayerProxy::mediaPlayerSizeChanged()` to forward the video
resize callback from MSE media player to web process.
No new tests, covered by existing failing layout test.
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
(WebKit::RemoteMediaPlayerProxy::mediaPlayerSizeChanged):
* GPUProcess/media/RemoteMediaPlayerProxy.h:
2020-12-23 Simon Fraser <simon.fraser@apple.com>
Standardize on "allows scrolling" terminology instead of asking about scrollbars
https://bugs.webkit.org/show_bug.cgi?id=220111
Reviewed by Wenson Hsieh.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<ScrollableAreaParameters>::encode):
(IPC::ArgumentCoder<ScrollableAreaParameters>::decode):
2020-12-22 Dean Jackson <dino@apple.com>
Use an enum class to identify WebGL versions
https://bugs.webkit.org/show_bug.cgi?id=220099
<rdar://problem/72596828>
Reviewed by Ken Russell.
Encode the WebGL version.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<WebCore::GraphicsContextGLAttributes>::encode):
(IPC::ArgumentCoder<WebCore::GraphicsContextGLAttributes>::decode):
* Shared/WebCoreArgumentCoders.h:
2020-12-22 Kate Cheney <katherine_cheney@apple.com>
Cannot see 'Add to my stations' button in BBC World service Radio
https://bugs.webkit.org/show_bug.cgi?id=219949
<rdar://problem/60319532>
Reviewed by Alex Christensen.
This is a temporary quirk to assist a high-traffic website while they
complete the large task of migrating away from site compatibility using
third party cookies. This quirk will be removed when the site is
updated.
Create a quirk that will call the Storage Access API on behalf of
BBC for radioplayer.co.uk under bbc.co.uk. When the user clicks to open
the radio player pop-up, storage access is requested. If the user grants
permission, then we open the pop-up to static.radioplayer.co.uk to get
first party status, then immediately redirect to the radio player page.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::grantStorageAccess):
(WebKit::ResourceLoadStatisticsDatabaseStore::grantStorageAccessInternal):
(WebKit::ResourceLoadStatisticsDatabaseStore::cookieAccess):
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::requestStorageAccessEphemeral):
Change wording now that quirks are not only being used for login
flows.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::createNewPage):
Mark the quirk-initiated pop-up as user-initiated so it doesn't get
blocked.
2020-12-22 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed, fix the build with recent SDKs.
* UIProcess/PDF/WKPDFHUDView.mm:
(-[WKPDFHUDView _getImageForControlName:]):
* UIProcess/RemoteLayerTree/mac/ScrollerMac.mm:
(-[WKScrollerImpDelegate effectiveAppearanceForScrollerImp:]):
2020-12-22 Youenn Fablet <youenn@apple.com>
[iOS] Do extensive search for microphone devices when trying to capture
https://bugs.webkit.org/show_bug.cgi?id=220010
<rdar://problem/72036034>
Reviewed by Eric Carlson.
Enable exhaustive audio device query in case getUserMedia is called for audio capture.
In case the manager is destroyed or reset after some time without capturing, make sure to disable
exhaustive audio device query if no other page need it.
* UIProcess/UserMediaPermissionRequestManagerProxy.cpp:
(WebKit::UserMediaPermissionRequestManagerProxy::invalidatePendingRequests):
(WebKit::UserMediaPermissionRequestManagerProxy::denyRequest):
(WebKit::UserMediaPermissionRequestManagerProxy::resetAccess):
(WebKit::UserMediaPermissionRequestManagerProxy::startProcessingUserMediaPermissionRequest):
* UIProcess/UserMediaPermissionRequestManagerProxy.h:
2020-12-21 Wenson Hsieh <wenson_hsieh@apple.com>
REGRESSION (iOS 14): Bad access crash in ShareableBitmap::makeCGImageCopy() under assignLegacyDataForContextMenuInteraction()
https://bugs.webkit.org/show_bug.cgi?id=219843
<rdar://problem/72537059>
Reviewed by Geoffrey Garen.
Apply a speculative fix for the crash by making the UI process robust in the case where the requested position
information has the `isImage` flag set, but no native image. Additionally, keep the existing debug assertions by
moving them outside of their current if statements.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView assignLegacyDataForContextMenuInteraction]):
(-[WKContentView continueContextMenuInteraction:]):
2020-12-21 Chris Dumez <cdumez@apple.com>
Simplify CARingBuffer usage with SharedMemory storage
https://bugs.webkit.org/show_bug.cgi?id=220047
Reviewed by Geoff Garen.
Simplify CARingBuffer usage with SharedMemory storage:
1. There are now 2 storage types:
- ReadOnlySharedRingBufferStorage: Used by the reader
- SharedRingBufferStorage: Used by the writer
One type inherits the other. This allows the API to be a bit
clearer and we can drop the m_readyOnly flag.
2. Drop SharedRingBufferStorage::updateReadOnlyStorage() function. Instead, the reader
can now construct a new ReadOnlySharedRingBufferStorage directly of the SharedMemory
handle and pass it to the CARingBuffer constructor.
* GPUProcess/media/RemoteAudioDestinationManager.cpp:
(WebKit::RemoteAudioDestination::audioSamplesStorageChanged):
(WebKit::RemoteAudioDestination::RemoteAudioDestination):
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRenderer.cpp:
(WebKit::RemoteAudioMediaStreamTrackRenderer::RemoteAudioMediaStreamTrackRenderer):
(WebKit::RemoteAudioMediaStreamTrackRenderer::audioSamplesStorageChanged):
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRenderer.h:
* GPUProcess/webrtc/RemoteMediaRecorder.cpp:
(WebKit::RemoteMediaRecorder::RemoteMediaRecorder):
(WebKit::RemoteMediaRecorder::audioSamplesStorageChanged):
* GPUProcess/webrtc/RemoteMediaRecorder.h:
* Shared/Cocoa/SharedRingBufferStorage.cpp:
(WebKit::ReadOnlySharedRingBufferStorage::ReadOnlySharedRingBufferStorage):
(WebKit::ReadOnlySharedRingBufferStorage::data):
(WebKit::ReadOnlySharedRingBufferStorage::sharedFrameBounds const):
(WebKit::ReadOnlySharedRingBufferStorage::getCurrentFrameBounds):
(WebKit::ReadOnlySharedRingBufferStorage::flush):
(WebKit::ReadOnlySharedRingBufferStorage::updateFrameBounds):
(WebKit::SharedRingBufferStorage::setStorage):
(WebKit::SharedRingBufferStorage::allocate):
(WebKit::SharedRingBufferStorage::deallocate):
(WebKit::SharedRingBufferStorage::setCurrentFrameBounds):
* Shared/Cocoa/SharedRingBufferStorage.h:
(WebKit::ReadOnlySharedRingBufferStorage::storage const):
* UIProcess/SpeechRecognitionRemoteRealtimeMediaSource.cpp:
(WebKit::m_ringBuffer):
(WebKit::SpeechRecognitionRemoteRealtimeMediaSource::setStorage):
* WebProcess/GPU/media/RemoteAudioSourceProviderManager.cpp:
(WebKit::RemoteAudioSourceProviderManager::RemoteAudio::RemoteAudio):
(WebKit::RemoteAudioSourceProviderManager::RemoteAudio::setStorage):
* WebProcess/cocoa/RemoteCaptureSampleManager.cpp:
(WebKit::RemoteCaptureSampleManager::RemoteAudio::RemoteAudio):
(WebKit::RemoteCaptureSampleManager::RemoteAudio::setStorage):
2020-12-21 Alex Christensen <achristensen@webkit.org>
Retain IPC::Connection when using it when making a NetworkDataTask
https://bugs.webkit.org/show_bug.cgi?id=220067
Reviewed by Geoffrey Garen.
r265230 helped, but it's still crashing.
Maybe we're getting a non-null pointer to an object that is about to be destroyed on another thread.
There's no reason not to keep it alive here.
Also, don't release assert in isParentProcessAFullWebBrowser if there's no audit token so we don't
just kick this can down the road.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::sessionWrapperForTask):
2020-12-21 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed, reverting r270939 and r270952.
Caused layout test timeouts on internal bots
Reverted changesets:
"[Cocoa] WebM format reader doesn't work with a url in a
<source> element"
https://bugs.webkit.org/show_bug.cgi?id=219961
https://trac.webkit.org/changeset/270939
"[Cocoa] WebM format reader doesn't work with a url in a
<source> element"
https://bugs.webkit.org/show_bug.cgi?id=219961
https://trac.webkit.org/changeset/270952
2020-12-21 Wenson Hsieh <wenson_hsieh@apple.com>
Many websites crash under toDataURL() when enabling GPU process for DOM
https://bugs.webkit.org/show_bug.cgi?id=220051
<rdar://problem/72412509>
Reviewed by Simon Fraser.
Three methods of canvas readback (`toDataURL`, `toData`, and `toBGRAData`) currently cause the web content
process to crash when enabling GPU process for DOM rendering. This is because we attempt to call into the
respective methods on the image buffer backend following r270342, which unconditionally fire release assertions.
Instead of using the proxy image buffer backend object in the web process, we need to communicate with the
platform `RemoteImageBuffer` in the GPU process. To do this, we add some plumbing through
`RemoteRenderingBackendProxy` and `RemoteRenderingBackend`, and override the `toDataURL`, `toData`, and
`toBGRAData` methods on `RemoteImageBufferProxy` in the web process to request data from the remote rendering
backend instead.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::getDataURLForImageBuffer):
(WebKit::RemoteRenderingBackend::getDataForImageBuffer):
(WebKit::RemoteRenderingBackend::getBGRADataForImageBuffer):
* GPUProcess/graphics/RemoteRenderingBackend.h:
* GPUProcess/graphics/RemoteRenderingBackend.messages.in:
* Scripts/webkit/messages.py:
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::getImageData):
(WebKit::RemoteRenderingBackendProxy::getDataURLForImageBuffer):
(WebKit::RemoteRenderingBackendProxy::getDataForImageBuffer):
(WebKit::RemoteRenderingBackendProxy::getBGRADataForImageBuffer):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
2020-12-21 Kate Cheney <katherine_cheney@apple.com>
Missing textInteractionEnabled key in WKPreferences initWithCoder looks like a false value.
https://bugs.webkit.org/show_bug.cgi?id=220061
<rdar://problem/72487342>
Reviewed by Wenson Hsieh.
We should check for the existence of the textInteractionEnabled key
in the NSCoder before updating the default value in WKPreferences,
otherwise a missing key will register as a false value when the default
of textInteractionEnabled should be true.
* UIProcess/API/Cocoa/WKPreferences.mm:
(-[WKPreferences initWithCoder:]):
2020-12-21 Sihui Liu <sihui_liu@apple.com>
Check availability of speech recognition service before requesting permissions
https://bugs.webkit.org/show_bug.cgi?id=220033
Reviewed by Youenn Fablet.
If speech recognition service is not available for target language (maybe the language is not supported, user
turns off dictation in system settings, etc), let's just fail the speech recognition request without prompting.
Manually tested with MiniBrowser.
* UIProcess/Cocoa/MediaPermissionUtilities.mm:
(WebKit::requestSpeechRecognitionAccess):
(WebKit::checkSpeechRecognitionServiceAccess):
(WebKit::checkSpeechRecognitionServiceAvailability):
* UIProcess/MediaPermissionUtilities.h:
* UIProcess/SpeechRecognitionPermissionManager.cpp:
(WebKit::SpeechRecognitionPermissionManager::request):
(WebKit::SpeechRecognitionPermissionManager::startProcessingRequest):
* UIProcess/SpeechRecognitionPermissionManager.h:
* UIProcess/SpeechRecognitionPermissionRequest.h:
(WebKit::SpeechRecognitionPermissionRequest::create):
(WebKit::SpeechRecognitionPermissionRequest::lang const):
(WebKit::SpeechRecognitionPermissionRequest::SpeechRecognitionPermissionRequest):
* UIProcess/SpeechRecognitionServer.cpp:
(WebKit::SpeechRecognitionServer::requestPermissionForRequest):
* UIProcess/SpeechRecognitionServer.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::requestSpeechRecognitionPermission):
* UIProcess/WebPageProxy.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::createSpeechRecognitionServer):
2020-12-21 Alicia Boya García <aboya@igalia.com>
[MSE] Fix allSamplesInTrackEnqueued() handling
https://bugs.webkit.org/show_bug.cgi?id=220055
Reviewed by Philippe Normand.
trySignalAllSamplesInTrackEnqueued() does not need to be exposed
outside of SourceBufferPrivate.
* GPUProcess/media/RemoteSourceBufferProxy.cpp:
(WebKit::RemoteSourceBufferProxy::trySignalAllSamplesInTrackEnqueued): Deleted.
* GPUProcess/media/RemoteSourceBufferProxy.h:
* GPUProcess/media/RemoteSourceBufferProxy.messages.in:
* WebProcess/GPU/media/SourceBufferPrivateRemote.cpp:
(WebKit::SourceBufferPrivateRemote::trySignalAllSamplesInTrackEnqueued): Deleted.
* WebProcess/GPU/media/SourceBufferPrivateRemote.h:
2020-12-20 Chris Dumez <cdumez@apple.com>
Unreviewed, reverting r270969.
Regressed MotionMark with GPUProcess enabled
Reverted changeset:
"[iOS] Stop leaking an XPC boost message to XPC services"
https://bugs.webkit.org/show_bug.cgi?id=219453
https://trac.webkit.org/changeset/270969
2020-12-20 Rob Buis <rbuis@igalia.com>
Convert remaining FrameLoaderType enums to enum classes
https://bugs.webkit.org/show_bug.cgi?id=219898
Reviewed by Sam Weinig.
Adapt to enum class changes.
* WebProcess/InjectedBundle/API/c/WKBundleFrame.cpp:
(WKBundleFrameGetFrameLoadState):
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::dispatchDidReceiveServerRedirectForProvisionalLoad):
2020-12-19 Rob Buis <rbuis@igalia.com>
Move FrameLoader::arePluginsEnabled to Frame
https://bugs.webkit.org/show_bug.cgi?id=220022
Reviewed by Sam Weinig.
Adapt to API change.
* WebProcess/Plugins/WebPluginInfoProvider.cpp:
(WebKit::WebPluginInfoProvider::pluginInfo):
(WebKit::WebPluginInfoProvider::populatePluginCache):
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::objectContentType):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::createPlugin):
(WebKit::WebPage::canPluginHandleResponse):
(WebKit::WebPage::canShowMIMEType const):
2020-12-19 Brent Fulgham <bfulgham@apple.com>
REGRESSION(r270575): [iOS] Correct GPU Process sandbox syntax error
https://bugs.webkit.org/show_bug.cgi?id=220041
<rdar://problem/72508138>
Unreviewed build fix after r270575.
A closing parentheses was omitted when the 'play-audio' rule was changed to
remove its last rule.
* Resources/SandboxProfiles/ios/com.apple.WebKit.GPU.sb:
2020-12-18 Tim Horton <timothy_horton@apple.com>
iOS wheel events don't work on many websites
https://bugs.webkit.org/show_bug.cgi?id=220034
Reviewed by Simon Fraser.
* Shared/ios/WebIOSEventFactory.h:
* Shared/ios/WebIOSEventFactory.mm:
(WebIOSEventFactory::createWebWheelEvent):
It turns out that many sites look at the legacy mousewheel event properties,
so compute wheelTicks in the same way that we do on macOS.
* UIProcess/API/Cocoa/WKWebViewInternal.h:
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _scrollView:asynchronouslyHandleScrollEvent:completion:]):
If the UI process sends a wheel event with delta=0,0 to the Web Content
process, it will not be sent to the page, and thus not handled.
If this is the first event, it will prevent future events from
being cancelable.
Instead, synthetically defer the "Begin" phase to the first event in a
given gesture that has non-zero deltas, and just short-circuit any
zero-delta events entirely.
This way, WebCore can continue with its existing only-the-first-event-is-cancelable
logic, and we are assured that the first event that WebCore sees is also
the first event that the page sees, and is a Begin-phase event.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::dispatchWheelEventWithoutScrolling):
wheelScrollGestureState isn't reset to nullopt until we send the
Begin-phase event to EventHandler, so we must make an exception for
the Begin-phase event, and explicitly make it cancelable. Otherwise,
the state of the previous gesture leaks into the new gesture. Existing
API tests caught this, I just apparently failed to run them after
adopting the shared wheelScrollGestureState().
2020-12-18 Brady Eidson <beidson@apple.com>
Fix some issues with PDFs as <object>.
https://bugs.webkit.org/show_bug.cgi?id=220024
Reviewed by Tim Horton.
No new tests (Unable to write automated tests)
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::dataProviderGetBytesAtPositionCallback):
(WebKit::PDFPlugin::getResourceBytesAtPosition): Set the correct URL.
(WebKit::PDFPlugin::ByteRangeRequest::completeWithAccumulatedData): Try harder to catch mismatches in delivered bytes vs expected bytes.
2020-12-18 Jeff Miller <jeffm@apple.com>
WKProcessPool enhancements for URL schemes
https://bugs.webkit.org/show_bug.cgi?id=220030
Reviewed by Brian Weinstein.
* UIProcess/API/Cocoa/WKProcessPool.mm:
(-[WKProcessPool _registerURLSchemeAsSecure:]):
Added, equivalent to WKContextRegisterURLSchemeAsSecure().
(-[WKProcessPool _registerURLSchemeAsBypassingContentSecurityPolicy:]):
Added, equivalent to WKContextRegisterURLSchemeAsBypassingContentSecurityPolicy().
(-[WKProcessPool _setDomainRelaxationForbiddenForURLScheme:]):
Added, equivalent to WKContextSetDomainRelaxationForbiddenForURLScheme().
* UIProcess/API/Cocoa/WKProcessPoolPrivate.h:
2020-12-18 BJ Burg <bburg@apple.com>
[Cocoa] Web Inspector: clean up _WKRemoteWebInspectorViewController
https://bugs.webkit.org/show_bug.cgi?id=219963
Reviewed by Devin Rousso.
Remove deprecated and unused debuggableType-based SPI to load the frontend.
* UIProcess/API/Cocoa/_WKRemoteWebInspectorViewController.h:
* UIProcess/API/Cocoa/_WKRemoteWebInspectorViewControllerPrivate.h:
Promote the _WKInspectorDebuggableInfo-based SPI to the main header.
Fix availability annotations. This class has never worked on iOS.
Add TBA annotations for things added since the Big Sur SDK was finalized.
* UIProcess/API/Cocoa/_WKRemoteWebInspectorViewController.mm:
(legacyDebuggableTypeToModernDebuggableType): Deleted.
(-[_WKRemoteWebInspectorViewController loadForDebuggableType:backendCommandsURL:]): Deleted.
2020-12-18 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process][MSE] SourceBufferPrivateRemote needs to override some track buffer related functions of SourceBufferPrivate
https://bugs.webkit.org/show_bug.cgi?id=219998
Reviewed by Jer Noble.
Implement some track buffer related functions of `SourceBufferPrivateRemote` with new IPC messages.
* GPUProcess/media/RemoteSourceBufferProxy.cpp:
(WebKit::RemoteSourceBufferProxy::resetTrackBuffers):
(WebKit::RemoteSourceBufferProxy::clearTrackBuffers):
(WebKit::RemoteSourceBufferProxy::resetTimestampOffsetInTrackBuffers):
(WebKit::RemoteSourceBufferProxy::setTimestampOffset):
(WebKit::RemoteSourceBufferProxy::setAppendWindowStart):
(WebKit::RemoteSourceBufferProxy::setAppendWindowEnd):
(WebKit::RemoteSourceBufferProxy::updateTrackIds):
* GPUProcess/media/RemoteSourceBufferProxy.h:
* GPUProcess/media/RemoteSourceBufferProxy.messages.in:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::addRemoteAudioTrack):
(WebKit::MediaPlayerPrivateRemote::addRemoteTextTrack):
(WebKit::MediaPlayerPrivateRemote::addRemoteVideoTrack):
`MediaSource` manages the track information for the MSE media player, so we should not call
`MediaPlayer::addAudioTrack()`, `MediaPlayer::addTextTrack()` and `MediaPlayer::addVideoTrack()`.
* WebProcess/GPU/media/SourceBufferPrivateRemote.cpp:
(WebKit::SourceBufferPrivateRemote::resetTrackBuffers):
(WebKit::SourceBufferPrivateRemote::clearTrackBuffers):
(WebKit::SourceBufferPrivateRemote::resetTimestampOffsetInTrackBuffers):
(WebKit::SourceBufferPrivateRemote::setTimestampOffset):
(WebKit::SourceBufferPrivateRemote::setAppendWindowStart):
(WebKit::SourceBufferPrivateRemote::setAppendWindowEnd):
(WebKit::SourceBufferPrivateRemote::updateTrackIds):
* WebProcess/GPU/media/SourceBufferPrivateRemote.h:
2020-12-18 Sihui Liu <sihui_liu@apple.com>
SpeechRecognitionRemoteRealtimeMediaSourceManager should not issue sandbox extension for mocked devices
https://bugs.webkit.org/show_bug.cgi?id=219972
Reviewed by Eric Carlson.
* UIProcess/SpeechRecognitionPermissionManager.cpp:
(WebKit::SpeechRecognitionPermissionManager::startProcessingRequest): We need to sync the preference not only
on when mockCaptureDevicesEnabled is set, but also when it is unset.
* UIProcess/SpeechRecognitionRemoteRealtimeMediaSourceManager.cpp:
(WebKit::SpeechRecognitionRemoteRealtimeMediaSourceManager::addSource): Grant sandbox extensions to web process
when the first source that uses a real device is created.
(WebKit::SpeechRecognitionRemoteRealtimeMediaSourceManager::removeSource): Revoke sandbox extensions from web
process when the last source that uses a real device is delete.
* UIProcess/SpeechRecognitionRemoteRealtimeMediaSourceManager.h:
* UIProcess/SpeechRecognitionServer.cpp: Correct a typo.
(WebKit::SpeechRecognitionServer::SpeechRecognitionServer):
* UIProcess/SpeechRecognitionServer.h: Correct a typo
2020-12-18 Chris Dumez <cdumez@apple.com>
REGRESSION (r270938): ASSERTION FAILED: !m_renderThread in WebKit::RemoteAudioDestinationProxy::startRenderingThread
https://bugs.webkit.org/show_bug.cgi?id=220017
<rdar://problem/72470176>
Reviewed by Darin Adler.
In case of a GPU process crash, we call stopRenderingThread() and then connectToGPUProcess() which
calls startRenderingThread(). We were hitting an assertion in startRenderingThread() making sure
that m_renderThread is null because stopRenderingThread() was failing to set it to null after
stopping the thread.
No new tests, covered by existing failing API test.
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::stopRenderingThread):
2020-12-18 Megan Gardner <megan_gardner@apple.com>
Gate all of the code associated with app highlights
https://bugs.webkit.org/show_bug.cgi?id=220003
<rdar://problem/72451893>
Reviewed by Tim Horton.
* Shared/API/c/WKSharedAPICast.h:
(WebKit::toAPI):
(WebKit::toImpl):
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultAppHighlightsEnabled):
* Shared/WebPreferencesDefaultValues.h:
* UIProcess/mac/WebContextMenuProxyMac.mm:
(WebKit::menuItemIdentifier):
2020-12-18 Chris Dumez <cdumez@apple.com>
[iOS] Stop leaking an XPC boost message to XPC services
https://bugs.webkit.org/show_bug.cgi?id=219453
Reviewed by Geoff Garen.
Stop leaking an XPC boost message to XPC services on builds where RunningBoard is able to give the right
priorities to our XPC services based on the type of RunningBoard assertion is being held. There is no
behavior change for the Network/WebContent processes since those were blocklisted in RunningBoard (Turning
on the RB_full_manage_WK_jetsam feature flag gives us the default RunningBoard behavior without
blocklisting). However, this allows the new GPUProcess to move to IDLE jetsam band when homing out of
MobileSafari, which makes the GPUProcess eligible for the freezer. Previously, the leaked boost message
would keep the GPUProcess in a higher jetsam band.
* FeatureFlags/WebKit.plist:
Add RB_full_manage_WK_jetsam feature flag and turn it on. On recent enough builds of RunningBoard, this
disables the blocklisting of the WebContent/Network processes in RunningBoard so that RunningBoard can now
set the priority of those processes (instead of us boosting our priority via boost messages).
* GPUProcess/EntryPoint/Cocoa/XPCService/GPUService/Info-iOS.plist:
* NetworkProcess/EntryPoint/Cocoa/XPCService/NetworkService/Info-iOS.plist:
* WebAuthnProcess/EntryPoint/Cocoa/XPCService/WebAuthnService/Info-iOS.plist:
* WebProcess/EntryPoint/Cocoa/XPCService/WebContentService/Info-iOS.plist:
Set _ProcessType=App so that RunningBoard is able to set the priority of our XPC services on iOS. Without
this change, we get a permission denied when RunningBoard attempts to set the priority of our XPC services.
* UIProcess/Launcher/mac/ProcessLauncherMac.mm:
(WebKit::shouldLeakBoost):
Stop leaking a boost message on iOS when RUNNINGBOARD_WEBKIT_PRIORITY_SUPPORT compile-time flag is enabled
and when the RB_full_manage_WK_jetsam feature flag is enabled.
2020-12-17 Kimmo Kinnunen <kkinnunen@apple.com>
REGRESSION (r268386): Snapshots of WebGL content in the tab picker don't work (black map on Google Maps)
https://bugs.webkit.org/show_bug.cgi?id=219946
<rdar://problem/72013784>
Reviewed by Dean Jackson.
Add dummy implementations of GraphicsContextGL::paintCompositedResultsToCanvas.
Painting to canvas is not implemented yet for the remote GraphicsContextGL implementation.
* GPUProcess/graphics/RemoteGraphicsContextGLCocoa.cpp:
(WebKit::RemoteGraphicsContextGLCocoa::displayBuffer const):
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.cpp:
(WebKit::RemoteGraphicsContextGLProxy::paintCompositedResultsToCanvas):
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.h:
2020-12-17 Alex Christensen <achristensen@webkit.org>
REGRESSION(r270638) decidePolicyForNewWindowAction should have non-null request
https://bugs.webkit.org/show_bug.cgi?id=220001
<rdar://72436824>
Reviewed by Tim Horton.
A ResourceRequest was used after being moved in the super old C SPI that we need to remove,
but to keep our client functioning correctly until they do we copy the ResourceRequest instead of moving it.
Covered by an API test.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::decidePolicyForNewWindowAction):
2020-12-17 Chris Dumez <cdumez@apple.com>
[GPUProcess] https://www.waveplayer.info/createmediaelementsource-test/ demo is flaky
https://bugs.webkit.org/show_bug.cgi?id=219951
Reviewed by Geoff Garen.
The issue was with the following line in AudioSourceProviderAVFObjC::prepare:
`m_ringBuffer = m_ringBufferCallback(description, capacity).moveToUniquePtr();`
In the case where m_ringBuffer was non-null before the assignment, we would have
2 RingBuffers that would coexist for a very small period of time. When the new
one was created, we would send an IPC to the remote process with the shared
memory handle of the new RingBuffer. However, very shortly after, the old
ring buffer would get destroyed, causing us to send another IPC to the remote
process with a null handle (since the shared memory associated with the old
ring buffer is getting destroyed). As a result, of this ordering issue, the
remote process would end up with a RingBuffer with a null shared memory handle
and no audio would be rendered.
We could have addressed the issue like so:
```
m_ringBuffer = nullptr;
m_ringBuffer = m_ringBufferCallback(description, capacity).moveToUniquePtr();
```
However, this would be super fragile. Instead, I have made the following changes:
1. If there is already a ringBuffer, reuse it instead of reconstructing it.
Calling allocate() with the new parameters on the existing ring buffer is
sufficient in this case.
2. Because of 1, the ring buffer creation callback no longer needs to call
CARingBuffer::allocate().
I also made the following changes to make the code simpler and to reduce code
duplication:
- The storage change handler passed to SharedRingBufferStorage is now given
as parameter the CAAudioStreamDescription & frameCount. What the handler
always does is send an IPC to the remote process to tell it that the
storage changed and in all cases, it needs to provide these 2 parameters
as well. This is because the remote process will need to call
CARingBuffer::allocate(), which requires those 2 parameters. This
simplifies our code in some cases since we no longer need a mechanism
to retrieve those 2 parameters from inside the storage change handler.
- The logic of the StorageChange IPC recipient to update its ringbuffer
with the new shared memory handle is complicated and was duplicated
in a LOT of places. To address this, I introduced a new
SharedRingBufferStorage::updateReadOnlyStorage() function which does
exactly what we need.
* GPUProcess/media/RemoteAudioDestinationManager.cpp:
(WebKit::RemoteAudioDestination::audioSamplesStorageChanged):
* GPUProcess/media/RemoteAudioSourceProviderProxy.cpp:
(WebKit::RemoteAudioSourceProviderProxy::create):
(WebKit::RemoteAudioSourceProviderProxy::createRingBuffer):
(WebKit::RemoteAudioSourceProviderProxy::storageChanged):
* GPUProcess/media/RemoteAudioSourceProviderProxy.h:
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRenderer.cpp:
(WebKit::RemoteAudioMediaStreamTrackRenderer::audioSamplesStorageChanged):
* GPUProcess/webrtc/RemoteMediaRecorder.cpp:
(WebKit::RemoteMediaRecorder::audioSamplesStorageChanged):
* Shared/Cocoa/SharedRingBufferStorage.cpp:
(WebKit::SharedRingBufferStorage::setStorage):
(WebKit::SharedRingBufferStorage::updateReadOnlyStorage):
(WebKit::SharedRingBufferStorage::allocate):
(WebKit::SharedRingBufferStorage::deallocate):
* Shared/Cocoa/SharedRingBufferStorage.h:
(WebKit::SharedRingBufferStorage::SharedRingBufferStorage):
(WebKit::SharedRingBufferStorage::storage const):
* UIProcess/Cocoa/UserMediaCaptureManagerProxy.cpp:
(WebKit::UserMediaCaptureManagerProxy::SourceProxy::SourceProxy):
(WebKit::UserMediaCaptureManagerProxy::SourceProxy::storageChanged):
* UIProcess/SpeechRecognitionRemoteRealtimeMediaSource.cpp:
(WebKit::SpeechRecognitionRemoteRealtimeMediaSource::setStorage):
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::RemoteAudioDestinationProxy):
(WebKit::RemoteAudioDestinationProxy::storageChanged):
* WebProcess/GPU/media/RemoteAudioDestinationProxy.h:
* WebProcess/GPU/media/RemoteAudioSourceProviderManager.cpp:
(WebKit::RemoteAudioSourceProviderManager::RemoteAudio::setStorage):
* WebProcess/GPU/webrtc/AudioMediaStreamTrackRenderer.cpp:
(WebKit::AudioMediaStreamTrackRenderer::AudioMediaStreamTrackRenderer):
(WebKit::AudioMediaStreamTrackRenderer::storageChanged):
* WebProcess/GPU/webrtc/AudioMediaStreamTrackRenderer.h:
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
(WebKit::MediaRecorderPrivate::startRecording):
(WebKit::MediaRecorderPrivate::storageChanged):
* WebProcess/GPU/webrtc/MediaRecorderPrivate.h:
* WebProcess/cocoa/RemoteCaptureSampleManager.cpp:
(WebKit::RemoteCaptureSampleManager::RemoteAudio::setStorage):
2020-12-17 Chris Dumez <cdumez@apple.com>
Unreviewed build fix after r270938.
Add missing include.
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
2020-12-17 Chris Dumez <cdumez@apple.com>
[WebAudio] Simplify code related to dispatchToRenderThread
https://bugs.webkit.org/show_bug.cgi?id=219990
Reviewed by Geoffrey Garen.
Simplify code related to dispatchToRenderThread in WebAudio:
1. AudioDestination::start() now always gets called with a non-null dispatchToRenderThread lambda.
In the case where there is no AudioWorkletThread to dispatch to, the lambda simply calls its
task synchronously.
2. For Cocoa ports, make it so that only AudioDestinationCocoa needs to worry about the
dispatchToRenderThread lambda. The dispatchToRenderThread lambda is no longer exposed to
subclasses such as MockAudioDestinationCocoa & RemoteAudioDestinationProxy.
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::startRenderingThread):
(WebKit::RemoteAudioDestinationProxy::startRendering):
(WebKit::RemoteAudioDestinationProxy::stopRendering):
(WebKit::RemoteAudioDestinationProxy::gpuProcessConnectionDidClose):
* WebProcess/GPU/media/RemoteAudioDestinationProxy.h:
2020-12-17 Alex Christensen <achristensen@webkit.org>
Fix "Open with Preview" menu item in PDF context menus on Big Sur
https://bugs.webkit.org/show_bug.cgi?id=219986
<rdar://problem/72406073>
Reviewed by Geoffrey Garen.
In r266654 I removed the ability for the web process to open a PDF in Preview on Big Sur.
I overlooked the fact that context menus also allow you to open a PDF in Preview,
which this fixes by having the UI process initiate the Preview opening if the user clicks on
the context menu item with the correct index.
* Shared/mac/PDFContextMenu.h:
(WebKit::PDFContextMenu::encode const):
(WebKit::PDFContextMenu::decode):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::showPDFContextMenu):
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::handleContextMenuEvent):
2020-12-17 Kate Cheney <katherine_cheney@apple.com>
Followup patch: Still can't login to my.playstation.com
https://bugs.webkit.org/show_bug.cgi?id=219900
<rdar://problem/72062985>
Reviewed by Darin Adler.
Using curly quotes when constructing the Storage Access prompt message
may create issues with localizable strings. We should make a separate
localizable string for this quirk case and remove it when Sony
migrates away from third party cookies in their login flow.
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::UIClient::requestStorageAccessConfirm):
(WebKit::buildListForStorageAccessPrompt): Deleted.
* UIProcess/Cocoa/WKStorageAccessAlert.h:
* UIProcess/Cocoa/WKStorageAccessAlert.mm:
(WebKit::presentStorageAccessAlert):
(WebKit::presentStorageAccessAlertQuirk):
(WebKit::displayStorageAccessAlert):
2020-12-17 Eric Carlson <eric.carlson@apple.com>
[Cocoa] WebM format reader doesn't work with a url in a <source> element
https://bugs.webkit.org/show_bug.cgi?id=219961
<rdar://problem/72399014>
Reviewed by Andy Estes.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-12-17 Chris Dumez <cdumez@apple.com>
[GPUProcess] Replace WebAudio rendering timer with a cross-process semaphore
https://bugs.webkit.org/show_bug.cgi?id=219964
Reviewed by Geoff Garen.
r270907 replaced the per-rendering quantum IPC from the GPUProcess to the
WebProcess with a very high-frequency timer in the WebProcess to trigger
the rendering. While this works, this seemed a bit fragile and it also
adds some latency to avoid missed frames.
In this patch, I am replacing the high-frequency timer introduced in r270907
with a cross-process semaphore. When the WebProcess asks the GPU process
to create an audio device, the GPUProcess sends it back a MachSendRight
to the "rendering" semaphore. The WebProcess then starts a high-priority
AudioThread which waits this semaphore to be signaled to render a quantum.
Whenever render() gets called on the GPUProcess side, the GPUProcess
simply signals the "rendering" semaphore, causing the WebProcess to do
the rendering of a quantum. As was the case before, the WebProcess stores
rendered buffers into the RingBuffer it shares with the GPUProcess and
the GPUProcess fetches from the RingBuffer as needed.
* GPUProcess/media/RemoteAudioDestinationManager.cpp:
(WebKit::RemoteAudioDestination::createRenderSemaphoreSendRight):
(WebKit::RemoteAudioDestination::audioSamplesStorageChanged):
(WebKit::RemoteAudioDestination::render):
(WebKit::RemoteAudioDestinationManager::createAudioDestination):
* GPUProcess/media/RemoteAudioDestinationManager.h:
* GPUProcess/media/RemoteAudioDestinationManager.messages.in:
* Shared/Cocoa/SharedRingBufferStorage.cpp:
* Shared/Cocoa/SharedRingBufferStorage.h:
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::RemoteAudioDestinationProxy):
(WebKit::RemoteAudioDestinationProxy::startRenderingThread):
(WebKit::RemoteAudioDestinationProxy::stopRenderingThread):
(WebKit::RemoteAudioDestinationProxy::connectToGPUProcess):
(WebKit::RemoteAudioDestinationProxy::~RemoteAudioDestinationProxy):
(WebKit::RemoteAudioDestinationProxy::start):
(WebKit::RemoteAudioDestinationProxy::stop):
(WebKit::RemoteAudioDestinationProxy::renderQuantum):
(WebKit::RemoteAudioDestinationProxy::storageChanged):
(WebKit::RemoteAudioDestinationProxy::gpuProcessConnectionDidClose):
* WebProcess/GPU/media/RemoteAudioDestinationProxy.h:
2020-12-17 Patrick Angle <pangle@apple.com>
Web Inspector: REGRESSION(r266467): viewing a WebSocket created before Web Inspector was opened doesn't show any message frames
https://bugs.webkit.org/show_bug.cgi?id=219510
Reviewed by Youenn Fablet.
Restored functionality prior to r266467 so that the `progressIdentifier` (previously `channelIdentifier`) is
available from `WebSocketChannel`.
* WebProcess/Network/WebSocketChannel.h:
2020-12-17 Rob Buis <rbuis@igalia.com>
Make FontCascade::CodePath an enum class
https://bugs.webkit.org/show_bug.cgi?id=219947
Reviewed by Alex Christensen.
Adapt to enum class change.
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::setAlwaysUsesComplexTextCodePath):
2020-12-16 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process][MSE] Add an IPC message BufferedSamplesForTrackId for testing purposes
https://bugs.webkit.org/show_bug.cgi?id=219956
Reviewed by Eric Carlson.
Implement `SourceBufferPrivateRemote::bufferedSamplesForTrackId()` with a new IPC message.
* GPUProcess/media/RemoteSourceBufferProxy.cpp:
(WebKit::RemoteSourceBufferProxy::bufferedSamplesForTrackId):
* GPUProcess/media/RemoteSourceBufferProxy.h:
* GPUProcess/media/RemoteSourceBufferProxy.messages.in:
* WebProcess/GPU/media/SourceBufferPrivateRemote.cpp:
(WebKit::SourceBufferPrivateRemote::bufferedSamplesForTrackId):
* WebProcess/GPU/media/SourceBufferPrivateRemote.h:
2020-12-16 Jiewen Tan <jiewen_tan@apple.com>
Unreviewed, a build fix for the latest Internal SDK
* Platform/spi/Cocoa/AuthenticationServicesCoreSPI.h:
2020-12-16 Kate Cheney <katherine_cheney@apple.com>
Still can't login to my.playstation.com
https://bugs.webkit.org/show_bug.cgi?id=219900
<rdar://problem/72062985>
Reviewed by John Wilander.
Adapted from an original patch by John Wilander.
This is a temporary quirk to assist a high-traffic website while they
complete the large task of migrating away from login flows that
require third party cookies. This quirk will be removed when the site
is updated.
Create a quirk to call the Storage Access API on behalf of Sony.
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::buildList):
(WebKit::UIDelegate::UIClient::requestStorageAccessConfirm):
* UIProcess/Cocoa/WKStorageAccessAlert.mm:
(WebKit::presentStorageAccessAlert):
Prompt for multiple domains if site-specific quirks are enabled and
either Sony login domain is being requested.
* WebProcess/WebCoreSupport/WebResourceLoadObserver.cpp:
(WebKit::WebResourceLoadObserver::hasCrossPageStorageAccess const):
(WebKit::WebResourceLoadObserver::setDomainsWithCrossPageStorageAccess):
* WebProcess/WebCoreSupport/WebResourceLoadObserver.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::addDomainWithPageLevelStorageAccess):
Store both Sony login domains as having page level storage access if
quirks are enabled and the domain passed is one of the Sony login
domains. In this case we know that both login domains were prompted
for.
2020-12-16 Chris Dumez <cdumez@apple.com>
[GPUProcess] Avoid doing an IPC per rendering quantum when using WebAudio
https://bugs.webkit.org/show_bug.cgi?id=219818
Reviewed by Geoff Garen.
Previously, the GPUProcess was sending an IPC to the WebProcess to request to rendering
of each quantum, which was very inefficient and a big regression compared to when we
do not use a GPUProcess.
To address the issue, we now start a rendering timer on the WebProcess side when rendering
starts, so that the WebProcess renders each quantum as needed, without needing an IPC from
the GPUProcess. Whenever a quantum is rendered, the WebProcess adds it to the RingBuffer
it shares with the GPUProcess.
To make sure that the WebProcess' writer keeps up with the GPUProcess' reader and that it
does not stray too far ahead, the writer keeps track of the last frame that the reader
attempted to read. It uses this information to try and stay a few rendering quantums ahead
of the GPUProcess, to avoid missed frames.
* CMakeLists.txt:
* DerivedSources-input.xcfilelist:
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* GPUProcess/media/RemoteAudioDestinationManager.cpp:
(WebKit::RemoteAudioDestination::audioSamplesStorageChanged):
(WebKit::RemoteAudioDestination::RemoteAudioDestination):
(WebKit::RemoteAudioDestination::render):
* Platform/SharedMemory.h:
(WebKit::SharedMemory::protection const):
* Shared/Cocoa/SharedRingBufferStorage.cpp:
(WebKit::SharedRingBufferStorage::sharedFrameBounds const):
(WebKit::SharedRingBufferStorage::setCurrentFrameBounds):
(WebKit::SharedRingBufferStorage::setLastReadFrame):
(WebKit::SharedRingBufferStorage::lastReadFrame const):
* Shared/Cocoa/SharedRingBufferStorage.h:
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::RemoteAudioDestinationProxy):
(WebKit::RemoteAudioDestinationProxy::startRenderingThread):
(WebKit::RemoteAudioDestinationProxy::stopRenderingThreadIfNecessary):
(WebKit::RemoteAudioDestinationProxy::connectToGPUProcess):
(WebKit::RemoteAudioDestinationProxy::~RemoteAudioDestinationProxy):
(WebKit::RemoteAudioDestinationProxy::start):
(WebKit::RemoteAudioDestinationProxy::startRendering):
(WebKit::RemoteAudioDestinationProxy::stop):
(WebKit::RemoteAudioDestinationProxy::stopRenderingIfNecessary):
(WebKit::RemoteAudioDestinationProxy::renderQuantum):
(WebKit::RemoteAudioDestinationProxy::storageChanged):
* WebProcess/GPU/media/RemoteAudioDestinationProxy.h:
(WebKit::RemoteAudioDestinationProxy::stopRenderingIfNecessary):
* WebProcess/GPU/media/RemoteAudioDestinationProxy.messages.in: Removed.
2020-12-16 Youenn Fablet <youenn@apple.com>
REGRESSION [iOS] Flaky GPU Process crash under -[RTCVideoEncoderH264 destroyCompressionSession]
https://bugs.webkit.org/show_bug.cgi?id=219899
<rdar://problem/72347652>
Reviewed by Alex Christensen.
We can no longer get the connection when needed as we are now operating in a background thread.
For that reason, capture the IPC connection in the lambda to make sure we have a valid IPC connection.
Remove LibWebRTCCodecsProxy when being closed instead of when being destroyed as well.
Covered by existing tests.
* GPUProcess/webrtc/LibWebRTCCodecsProxy.mm:
(WebKit::LibWebRTCCodecsProxy::~LibWebRTCCodecsProxy):
(WebKit::LibWebRTCCodecsProxy::close):
(WebKit::LibWebRTCCodecsProxy::createH264Decoder):
(WebKit::LibWebRTCCodecsProxy::createH265Decoder):
(WebKit::LibWebRTCCodecsProxy::createVP9Decoder):
(WebKit::LibWebRTCCodecsProxy::createEncoder):
2020-12-16 Wenson Hsieh <wenson_hsieh@apple.com>
Suppress the image extraction interaction while editing text
https://bugs.webkit.org/show_bug.cgi?id=219952
<rdar://problem/72390053>
Reviewed by Devin Rousso.
See radar and WebKitAdditions for more details.
* UIProcess/ios/WKContentViewInteraction.h:
(WebKit::SuppressInteractionToken::SuppressInteractionToken):
(WebKit::SuppressInteractionToken::~SuppressInteractionToken):
Add a helper class to help facilitate the temporary removal of UIInteraction objects from a WKContentView.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView setIsEditable:]):
Use the above helper class to temporarily remove the image extraction interaction while the user is editing
text (and is therefore depending on editable text interaction gestures). Note that we avoid unnecessarily
clearing out and resetting this member variable when changing focus between editable text fields.
(-[WKContentView _didStartProvisionalLoadForMainFrame]):
2020-12-16 Brent Fulgham <bfulgham@apple.com>
REGRESSION (r270657) [AS Only] GGE: Netflix, YouTube, Amazon Prime, and Hulu fail to playback
https://bugs.webkit.org/show_bug.cgi?id=219954
<rdar://problem/72325887>
Reviewed by Per Arne Vollan.
Correct a syntax error in a sandbox rule added in r270657.
* WebProcess/com.apple.WebProcess.sb.in:
2020-12-15 Simon Fraser <simon.fraser@apple.com>
Calls to setFixedLayoutSize() should not be inside ENABLE(TEXT_AUTOSIZING)
https://bugs.webkit.org/show_bug.cgi?id=219932
Reviewed by Wenson Hsieh.
We always want to call setFixedLayoutSize(), whether or not ENABLE(TEXT_AUTOSIZING) is defined.
For code simplicity, make resetTextAutosizing() an empty function when ENABLE(TEXT_AUTOSIZING)
is not defined.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::dynamicViewportSizeUpdate):
(WebKit::WebPage::resetTextAutosizing):
(WebKit::WebPage::viewportConfigurationChanged):
2020-12-16 Per Arne Vollan <pvollan@apple.com>
[macOS] Add sandbox parameter for message filter entitlement status
https://bugs.webkit.org/show_bug.cgi?id=219852
<rdar://problem/72145560>
Reviewed by Brent Fulgham.
Add sandbox parameter to determine whether the WebContent process has message filter entitlement.
This parameter will be used in the sandbox to enable message filtering.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::initializeSandbox):
* WebProcess/com.apple.WebProcess.sb.in:
2020-12-16 Youenn Fablet <youenn@apple.com>
Make sure WebKit WebSocketChannel handle suspend/resume correctly
https://bugs.webkit.org/show_bug.cgi?id=219910
Reviewed by Alex Christensen.
* WebProcess/Network/WebSocketChannel.cpp:
(WebKit::WebSocketChannel::fail):
Make sure fail triggers close event if page resumes.
2020-12-15 Pablo Saavedra <psaavedra@igalia.com>
[CMake][WPE] Add LibWPE before WPEBackend-FDO for qtwpe_INCLUDE_DIRECTORIES
https://bugs.webkit.org/show_bug.cgi?id=219920
Reviewed by Michael Catanzaro.
* PlatformWPE.cmake:
2020-12-15 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r270829.
https://bugs.webkit.org/show_bug.cgi?id=219935
Introduced crash
Reverted changeset:
"[macOS] Add sandbox parameter for message filter entitlement
status"
https://bugs.webkit.org/show_bug.cgi?id=219852
https://trac.webkit.org/changeset/270829
2020-12-15 Andy Estes <aestes@apple.com>
[Mac] Numerous webgl tests continue to time out with the WebM format reader enabled
https://bugs.webkit.org/show_bug.cgi?id=219928
Reviewed by Eric Carlson.
Covered by existing tests.
* Shared/mac/MediaFormatReader/TrackReader.cpp:
(WebKit::MediaSampleByteRange::MediaSampleByteRange):
(WebKit::MediaSampleByteRange::trackID const):
(WebKit::TrackReader::addSample): Store MediaSampleByteRange's track ID as a uint64_t
instead of an AtomString that might've been created on another thread.
(WebKit::TrackReader::finalize): Ensure that m_sampleStorage really is destroyed on the
storage queue by explicitly settings its unique_ptr to nullptr in the lambda. Without
doing this, it's possible that the lambda will execute on the storage queue *before*
WorkQueue::dispatch returns, and since WorkQueue's BlockPtr is still holding a reference
to the block containing the lambda, it would not have been destroyed on the storage queue.
2020-12-15 Per Arne Vollan <pvollan@apple.com>
[macOS, iOS] Add required mach syscall
https://bugs.webkit.org/show_bug.cgi?id=219913
<rdar://problem/72347689>
Reviewed by Brent Fulgham.
Add required mach syscall on macOS and iOS.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* WebProcess/com.apple.WebProcess.sb.in:
2020-12-15 Wenson Hsieh <wenson_hsieh@apple.com>
Image extraction interaction should remain active while scrolling
https://bugs.webkit.org/show_bug.cgi?id=219895
<rdar://problem/72344814>
Reviewed by Devin Rousso.
Make our image extraction interaction behave more like the rest of the system by making it persist until
explicitly dismissed via tap gesture (or any other competing interaction, such as text interactions, drag
interactions and context menu interactions, as long as the interaction location is outside of the active image
extraction interaction).
We currently just dismiss any active image extraction upon starting a new touch, which prevents scrolling from
working as expected with image extraction. See radar for more details.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _webTouchEventsRecognized:]):
(-[WKContentView hasSelectablePositionAtPoint:]):
(-[WKContentView textInteractionGesture:shouldBeginAtPoint:]):
(-[WKContentView _singleTapIdentified:]):
(-[WKContentView _dragInteraction:prepareForSession:completion:]):
(-[WKContentView _contextMenuInteraction:configurationForMenuAtLocation:completion:]):
2020-12-15 Kate Cheney <katherine_cheney@apple.com>
CrashTracer: com.apple.WebKit.Networking at WebKit: WTF::Detail::CallableWrapper<WebKit::PrivateClickMeasurementManager::firePendingAttributionRequests()
https://bugs.webkit.org/show_bug.cgi?id=219860
<rdar://problem/72254207>
Reviewed by Alex Christensen.
Followup patch to add missing completion handlers. This also moves the
check for a null m_networkSession variable to the existing
PrivateClickMeasurementManager::featureEnabled() function for
simplicity.
* NetworkProcess/PrivateClickMeasurementManager.cpp:
(WebKit::PrivateClickMeasurementManager::storeUnattributed):
(WebKit::PrivateClickMeasurementManager::attribute):
(WebKit::PrivateClickMeasurementManager::clearSentAttributions):
(WebKit::PrivateClickMeasurementManager::updateTimerLastFired):
(WebKit::PrivateClickMeasurementManager::firePendingAttributionRequests):
(WebKit::PrivateClickMeasurementManager::clear):
(WebKit::PrivateClickMeasurementManager::clearForRegistrableDomain):
(WebKit::PrivateClickMeasurementManager::clearExpired):
(WebKit::PrivateClickMeasurementManager::toString const):
(WebKit::PrivateClickMeasurementManager::markAllUnattributedAsExpiredForTesting):
(WebKit::PrivateClickMeasurementManager::featureEnabled const):
(WebKit::PrivateClickMeasurementManager::markAttributedPrivateClickMeasurementsAsExpiredForTesting):
2020-12-15 Simon Fraser <simon.fraser@apple.com>
Add batching to WebPreferences updates
https://bugs.webkit.org/show_bug.cgi?id=219878
Reviewed by Sam Weinig.
WebPreferences::resetAllInternalDebugFeatures() and WebPreferences::enableAllExperimentalFeatures()
would trigger tens of calls to WebPageProxy::preferencesDidChange(). Fix by adding batching, and only
firing update() at the end of any prefs changed.
* Scripts/PreferencesTemplates/WebPreferencesExperimentalFeatures.cpp.erb:
* Scripts/PreferencesTemplates/WebPreferencesInternalDebugFeatures.cpp.erb:
* UIProcess/WebPreferences.cpp:
(WebKit::WebPreferences::update):
(WebKit::WebPreferences::startBatchingUpdates):
(WebKit::WebPreferences::endBatchingUpdates):
* UIProcess/WebPreferences.h:
(WebKit::WebPreferences::UpdateBatch::UpdateBatch):
(WebKit::WebPreferences::UpdateBatch::~UpdateBatch):
2020-12-15 Youenn Fablet <youenn@apple.com>
Fix alertMessageText regression from https://trac.webkit.org/changeset/269918
https://bugs.webkit.org/show_bug.cgi?id=219891
Reviewed by Eric Carlson.
alertMessageText was checking twice for audio while it should check once for audio and once for video.
* UIProcess/Cocoa/MediaPermissionUtilities.mm:
(WebKit::alertMessageText):
2020-12-15 Ryan Hostetler <rhost@apple.com>
"make clean" fails on EmptyDSTROOT target
https://bugs.webkit.org/show_bug.cgi?id=219870
rdar://72309233
Reviewed by Alexey Proskuryakov.
make clean can't find PROJECT_NAME for EmptyDSTROOT.
* WebKit.xcodeproj/project.pbxproj:
2020-12-15 Andy Estes <aestes@apple.com>
[Mac] Numerous webgl tests time out with WebM format reader enabled
https://bugs.webkit.org/show_bug.cgi?id=219887
Reviewed by Eric Carlson.
When rendering a video frame to canvas, SampleCursor is invoked on the main thread in
order to find a sync frame to render. This triggered an ASSERT(!isMainThread()) in debug
builds, and possibly a deadlock if the cursor pointed to a sample that hadn't yet been
parsed (because SourceBufferParser calls the 'didProvideMediaData' callback on the main
thread).
Resolve this by having SourceBufferParser provide parsed samples on a background queue. To
avoid creating AtomStrings on this non-main queue, teach TrackReader to access the trackID
as a uint64_t instead.
* Shared/mac/MediaFormatReader/FormatReader.cpp:
(WebKit::FormatReader::startOnMainThread):
(WebKit::readerQueue):
(WebKit::FormatReader::parseByteSource):
(WebKit::FormatReader::didParseTracks):
(WebKit::FormatReader::didProvideMediaData):
(WebKit::FormatReader::finishParsing):
(WebKit::FormatReader::copyTrackArray):
* Shared/mac/MediaFormatReader/TrackReader.cpp:
(WebKit::MediaSampleByteRange::MediaSampleByteRange):
(WebKit::TrackReader::create):
(WebKit::TrackReader::storageQueue):
(WebKit::TrackReader::TrackReader):
(WebKit::TrackReader::addSample):
(WebKit::TrackReader::waitForSample const):
(WebKit::TrackReader::finishParsing):
(WebKit::TrackReader::finalize):
* Shared/mac/MediaFormatReader/TrackReader.h:
2020-12-15 Chris Dumez <cdumez@apple.com>
[GPUProcess] Crash under AudioDestinationCocoa::setIsPlaying(bool)
https://bugs.webkit.org/show_bug.cgi?id=219809
<rdar://problem/72298420>
Reviewed by Youenn Fablet.
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::RemoteAudioDestinationProxy):
2020-12-15 Per Arne Vollan <pvollan@apple.com>
[macOS] Add sandbox parameter for message filter entitlement status
https://bugs.webkit.org/show_bug.cgi?id=219852
<rdar://problem/72145560>
Reviewed by Brent Fulgham.
Add sandbox parameter to determine whether the WebContent process has message filter entitlement.
This parameter will be used in the sandbox to enable message filtering.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::initializeSandbox):
* WebProcess/com.apple.WebProcess.sb.in:
2020-12-15 Carlos Garcia Campos <cgarcia@igalia.com>
[SOUP] WebSockets: wrong data sent in some cases with libsoup < 2.68
https://bugs.webkit.org/show_bug.cgi?id=219850
Reviewed by Michael Catanzaro.
We are using soup_websocket_connection_send_text() when libsoup < 2.68, which expects a null terminated string,
but we are receiving a IPC::DataReference that needs to be converted to a null terminated string by taking into
account its size.
* NetworkProcess/soup/WebSocketTaskSoup.cpp:
(WebKit::WebSocketTask::sendString):
2020-12-14 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process][MSE] VP9 stream playback is not smooth
https://bugs.webkit.org/show_bug.cgi?id=219861
Reviewed by Eric Carlson.
Add two new IPC messages in `RemoteMediaSourceProxy` to set `m_isSeeking` and
`m_timeFudgeFactor` of the `MediaSourcePrivate` running in the GPU process.
* GPUProcess/media/RemoteMediaSourceProxy.cpp:
(WebKit::RemoteMediaSourceProxy::setIsSeeking):
(WebKit::RemoteMediaSourceProxy::setTimeFudgeFactor):
* GPUProcess/media/RemoteMediaSourceProxy.h:
* GPUProcess/media/RemoteMediaSourceProxy.messages.in:
* WebProcess/GPU/media/MediaSourcePrivateRemote.cpp:
(WebKit::MediaSourcePrivateRemote::setIsSeeking):
(WebKit::MediaSourcePrivateRemote::setTimeFudgeFactor):
* WebProcess/GPU/media/MediaSourcePrivateRemote.h:
2020-12-14 Per Arne Vollan <pvollan@apple.com>
REGRESSION (r270677-r270693): Catalina Perf tests failing
https://bugs.webkit.org/show_bug.cgi?id=219802
<rdar://problem/72236108>
Reviewed by Brent Fulgham.
Create extension for RootDomainUserClient also for Intel Macs, and move logging rules outside of
HAVE(SANDBOX_MESSAGE_FILTERING) section in sandbox.
* UIProcess/WebPageProxy.cpp:
(WebKit::mediaRelatedIOKitClasses):
* WebProcess/com.apple.WebProcess.sb.in:
2020-12-14 Alex Christensen <achristensen@webkit.org>
Fix null dereference in WKURLCopyCFURL
https://bugs.webkit.org/show_bug.cgi?id=219877
<rdar://problem/48921304>
Reviewed by Geoffrey Garen.
We had an optimistic assertion and crash reports indicating that it is not always true.
When it's not, return null instead of crashing.
* Shared/API/c/cf/WKURLCF.mm:
(WKURLCopyCFURL):
2020-12-14 Chris Dumez <cdumez@apple.com>
Unreviewed, reverting r270768.
Still crashes in debug in a different location
Reverted changeset:
"[GPUProcess] Crash under
AudioDestinationCocoa::setIsPlaying(bool)"
https://bugs.webkit.org/show_bug.cgi?id=219809
https://trac.webkit.org/changeset/270768
2020-12-14 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process][MSE] Fix crashes in a debug build
https://bugs.webkit.org/show_bug.cgi?id=219874
Reviewed by Eric Carlson.
We need to call the completion handler in error cases.
* GPUProcess/media/RemoteSourceBufferProxy.cpp:
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateDidReceiveInitializationSegment):
* WebProcess/GPU/media/SourceBufferPrivateRemote.cpp:
(WebKit::SourceBufferPrivateRemote::sourceBufferPrivateDidReceiveInitializationSegment):
2020-12-14 Chris Dumez <cdumez@apple.com>
[GPUProcess] Bump QoS to UserInteractive for real-time audio rendering thread
https://bugs.webkit.org/show_bug.cgi?id=219873
Reviewed by Eric Carlson.
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::RemoteAudioDestinationProxy):
2020-12-14 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process][MSE] WebM source buffer parser is not enabled
https://bugs.webkit.org/show_bug.cgi?id=219857
Reviewed by Eric Carlson.
Share the value of `RuntimeEnabledFeatures::sharedFeatures().webMParserEnabled()`
with the GPU process (`RemoteMediaSourceProxy`) when `MediaPlayerPrivateRemote`
tries to load a media source.
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
(WebKit::RemoteMediaPlayerProxy::loadMediaSource):
* GPUProcess/media/RemoteMediaPlayerProxy.h:
* GPUProcess/media/RemoteMediaPlayerProxy.messages.in:
* GPUProcess/media/RemoteMediaSourceProxy.cpp:
(WebKit::RemoteMediaSourceProxy::RemoteMediaSourceProxy):
(WebKit::RemoteMediaSourceProxy::addSourceBuffer):
* GPUProcess/media/RemoteMediaSourceProxy.h:
* Shared/mac/MediaFormatReader/FormatReader.cpp:
(WebKit::FormatReader::parseByteSource):
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::load):
* WebProcess/GPU/media/MediaSourcePrivateRemote.cpp:
(WebKit::MediaSourcePrivateRemote::addSourceBuffer):
* WebProcess/GPU/media/MediaSourcePrivateRemote.h:
2020-12-14 Chris Dumez <cdumez@apple.com>
[GPUProcess] Crash in SharedRingBufferStorage::setStorage() under GuardMalloc
https://bugs.webkit.org/show_bug.cgi?id=219867
Reviewed by Darin Adler.
SharedRingBufferStorage was holding a raw pointer to its client and we would sometimes
crash when dereferencing this client in SharedRingBufferStorage::setStorage() because
the client is not guaranteed to be alive.
To address the issue, I replace the SharedRingBufferStorage::Client with a lambda so
that the client can protect itself in the lambda if needed. In particular,
RemoteAudioSourceProviderProxy now protects itself in the lambda since the lifetime
of the CARingBuffer is not tied to the lifetime of RemoteAudioSourceProviderProxy.
* GPUProcess/media/RemoteAudioDestinationManager.cpp:
(WebKit::RemoteAudioDestination::RemoteAudioDestination):
* GPUProcess/media/RemoteAudioSourceProviderProxy.cpp:
(WebKit::RemoteAudioSourceProviderProxy::createRingBuffer):
* GPUProcess/media/RemoteAudioSourceProviderProxy.h:
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRenderer.cpp:
(WebKit::RemoteAudioMediaStreamTrackRenderer::RemoteAudioMediaStreamTrackRenderer):
* GPUProcess/webrtc/RemoteMediaRecorder.cpp:
(WebKit::RemoteMediaRecorder::RemoteMediaRecorder):
* Shared/Cocoa/SharedRingBufferStorage.cpp:
(WebKit::SharedRingBufferStorage::setStorage):
* Shared/Cocoa/SharedRingBufferStorage.h:
(WebKit::SharedRingBufferStorage::SharedRingBufferStorage):
(WebKit::SharedRingBufferStorage::invalidate):
* UIProcess/Cocoa/UserMediaCaptureManagerProxy.cpp:
(WebKit::UserMediaCaptureManagerProxy::SourceProxy::SourceProxy):
(WebKit::UserMediaCaptureManagerProxy::SourceProxy::storageChanged):
* UIProcess/SpeechRecognitionRemoteRealtimeMediaSource.cpp:
(WebKit::m_ringBuffer):
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::RemoteAudioDestinationProxy):
* WebProcess/GPU/media/RemoteAudioDestinationProxy.h:
* WebProcess/GPU/media/RemoteAudioSourceProviderManager.cpp:
(WebKit::RemoteAudioSourceProviderManager::RemoteAudio::RemoteAudio):
* WebProcess/GPU/webrtc/AudioMediaStreamTrackRenderer.cpp:
(WebKit::AudioMediaStreamTrackRenderer::AudioMediaStreamTrackRenderer):
* WebProcess/GPU/webrtc/AudioMediaStreamTrackRenderer.h:
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
(WebKit::MediaRecorderPrivate::startRecording):
* WebProcess/GPU/webrtc/MediaRecorderPrivate.h:
* WebProcess/Speech/SpeechRecognitionRealtimeMediaSourceManager.cpp:
(WebKit::SpeechRecognitionRealtimeMediaSourceManager::Source::Source):
(WebKit::SpeechRecognitionRealtimeMediaSourceManager::Source::storageChanged):
* WebProcess/cocoa/RemoteCaptureSampleManager.cpp:
(WebKit::RemoteCaptureSampleManager::RemoteAudio::RemoteAudio):
2020-12-14 Kate Cheney <katherine_cheney@apple.com>
CrashTracer: com.apple.WebKit.Networking at WebKit: WTF::Detail::CallableWrapper<WebKit::PrivateClickMeasurementManager::firePendingAttributionRequests()
https://bugs.webkit.org/show_bug.cgi?id=219860
<rdar://problem/72254207>
Reviewed by Alex Christensen.
No new tests, this should fix a crash.
m_networkSession is a weakPtr. We should check that the object it
references exists before trying to use it.
* NetworkProcess/PrivateClickMeasurementManager.cpp:
(WebKit::PrivateClickMeasurementManager::storeUnattributed):
(WebKit::PrivateClickMeasurementManager::attribute):
(WebKit::PrivateClickMeasurementManager::clearSentAttributions):
(WebKit::PrivateClickMeasurementManager::updateTimerLastFired):
(WebKit::PrivateClickMeasurementManager::firePendingAttributionRequests):
(WebKit::PrivateClickMeasurementManager::clear):
(WebKit::PrivateClickMeasurementManager::clearForRegistrableDomain):
(WebKit::PrivateClickMeasurementManager::clearExpired):
(WebKit::PrivateClickMeasurementManager::toString const):
(WebKit::PrivateClickMeasurementManager::markAllUnattributedAsExpiredForTesting):
(WebKit::PrivateClickMeasurementManager::markAttributedPrivateClickMeasurementsAsExpiredForTesting):
2020-12-14 Tim Horton <timothy_horton@apple.com>
Adopt symbol image API instead of SPI in WKPDFHUDView
https://bugs.webkit.org/show_bug.cgi?id=219866
Reviewed by Wenson Hsieh.
* UIProcess/PDF/WKPDFHUDView.mm:
(-[WKPDFHUDView _getImageForControlName:]):
2020-12-14 Pablo Saavedra <psaavedra@igalia.com>
[GLIB] Missing header for std::once_flag in WebKitInitialize.cpp
https://bugs.webkit.org/show_bug.cgi?id=219865
Reviewed by Alex Christensen.
* UIProcess/API/glib/WebKitInitialize.cpp:
2020-12-14 Wenson Hsieh <wenson_hsieh@apple.com>
GPU process sometimes hangs during MotionMark Focus when enabling GPU Process for DOM
https://bugs.webkit.org/show_bug.cgi?id=219839
<rdar://problem/71562635>
Reviewed by Tim Horton.
Fixes an indefinite hang in the GPU process that may occur as a result of this very specific corner case in
concurrent display list processing:
[1] Let p and q represent two item buffers, and let A and B represent two remote image buffers whose drawing
contexts will be destinations for display list playback in the GPU process. Suppose p has k remaining bytes
where k >= 32, and A's writable buffer is mapped to a region closest to the end of p. In practice, when this
hang reproduces, k is a relatively small number (typically 32-40 bytes).
[2] Additionally, suppose that the GPU process is not reading from q (i.e. q's `unreadBytes()` count is 0), and
that the GPU process has created a corresponding RemoteImageBuffer for A, but *not* a corresponding
RemoteImageBuffer for B.
[3] We now append a MetaCommandChangeDestinationImageBuffer item to change destinations from A to B, such that
p now has (k - 16) remaining bytes. In preparation for appending new items for B's display list, the logic
in `RemoteRenderingBackendProxy::willAppendItem` clears B's writable buffer, with the intention that B will
turn around and call out to the client layer for a new writable buffer again once it notices that it has no
writable buffer.
[4] If we then immediately try to append an item to B that is larger than (k - 32) bytes,
`RemoteRenderingBackendProxy::findReusableDisplayListHandle` will see that the most recently used item
buffer p has insufficient space, and will instead return item buffer q. We end up appending the new item to
item buffer q.
[5] In the GPU process, we finish processing all available item data in p. When we encounter the destination
buffer change item, we set `m_pendingWakeupInfo` before going to sleep, due to the fact that the GPU process
has not learned about image buffer B yet (see the last part of step [2]).
[6] Later, in the GPU process, when we create a RemoteImageBuffer for B, we see that we have a
`m_pendingWakeupInfo` with destination image buffer B, so we resume display list processing. However, we're
unable to resume where we stopped, since B's display list (in the web process) immediately appended to q
instead of p, skipping MetaCommandChangeItemBuffer in the process. This manifests as an indefinite hang
inside `RemoteRenderingBackend::nextDestinationImageBufferAfterApplyingDisplayLists`, since we expect p to
have a nonzero `unreadBytes()` count.
[7] Now that the GPU process is in a broken state, the web process follows shortly -- in the Focus subtest of
MotionMark, this manifests as a web process crash soon after the GPU process begins to hang, because
`RemoteImageBufferProxy::ensureBackendCreated()` times out and returns null, and we subsequently dereference
null in `RemoteLayerBackingStore::encode()` after failing to ensure the image buffer backend.
To fix this, we target step [3]. Rather than clear B's writable buffer and expect it to append its new items to
p, we simply hand it p's item handle -- in this case, a handle with (k - 16) available bytes -- and set its
writable buffer before we try to append. In this scenario, doing this ensures that if we are about to append an
item that is larger than (k - 32) bytes, we will first leave a 16-byte MetaCommandChangeItemBuffer item inside p
before moving on to q. This eventually ensures that when the GPU process resumes display list processing after
creating a RemoteImageBuffer for B, it will see 16 unread bytes in p and then immediately continue reading data
from item buffer q.
Finally, note that in this scenario, p is always guaranteed to contain at least 16 available bytes after step
[3] (in other words, k >= 32). This is because all calls to append an item of size n bytes to an item buffer
check that there is a minimum (n + 16) bytes remaining (via logic in `ItemBuffer::swapWritableBufferIfNeeded`),
such that it is, by design, always possible to append a `MetaCommandChangeItemBuffer` item if necessary. As
such, it is safe to return the most recently used item buffer's handle in step [4].
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
(WebKit::RemoteImageBufferProxy::clearDisplayList):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::willAppendItem):
Call into `prepareToAppendDisplayListItems` instead of `clearDisplayList` when switching destination image
buffers. See problem diagnosis above for more detail.
(WebKit::RemoteRenderingBackendProxy::mostRecentlyUsedDisplayListHandle):
Pull logic to retrieve the most recently used display list handle out into a separate helper method.
(WebKit::RemoteRenderingBackendProxy::findReusableDisplayListHandle):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
2020-12-14 Sihui Liu <sihui_liu@apple.com>
WebProcessProxy::destroySpeechRecognitionServer should check if identifier exists before removing MessageReceiver
https://bugs.webkit.org/show_bug.cgi?id=219815
<rdar://problem/71369164>
Reviewed by Youenn Fablet.
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::destroySpeechRecognitionServer):
2020-12-14 Chris Dumez <cdumez@apple.com>
Drop duplicate call to makeUniqueRef() in RemoteAudioSourceProviderProxy::createRingBuffer()
https://bugs.webkit.org/show_bug.cgi?id=219859
Reviewed by Youenn Fablet.
* GPUProcess/media/RemoteAudioSourceProviderProxy.cpp:
(WebKit::RemoteAudioSourceProviderProxy::createRingBuffer):
2020-12-14 Youenn Fablet <youenn@apple.com>
Make sure WebProcessPool::setUseSeparateServiceWorkerProcess handles correctly destroyed process pools
https://bugs.webkit.org/show_bug.cgi?id=219773
<rdar://problem/71938320>
Reviewed by Alex Christensen.
Terminating a service worker process may destroy the web process pool.
Copy the vector of process pools before iterating over it to terminate service worker processes.
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::setUseSeparateServiceWorkerProcess):
2020-12-14 Alex Christensen <achristensen@webkit.org>
Remove WKDownloadDelegate.didWriteData
https://bugs.webkit.org/show_bug.cgi?id=219820
Reviewed by Sam Weinig.
API review revealed it was redundant with WKDownload.progress, which has KVO-compliant totalUnitCount and completedUnitCount.
* UIProcess/API/Cocoa/WKDownload.mm:
(-[WKDownload progress]):
* UIProcess/API/Cocoa/WKDownloadDelegate.h:
2020-12-14 Andy Estes <aestes@apple.com>
[Mac] Enable `webm_format_reader` feature flag by default
https://bugs.webkit.org/show_bug.cgi?id=219858
Reviewed by Eric Carlson.
* FeatureFlags/WebKit.plist:
2020-12-14 Wenson Hsieh <wenson_hsieh@apple.com>
Web process crashes during MotionMark Images when GPU Process for DOM is enabled
https://bugs.webkit.org/show_bug.cgi?id=219838
Reviewed by Simon Fraser.
It's possible to return prematurely with a null backend when waiting for image buffer backend creation under
`RemoteImageBufferProxy::ensureBackendCreated()`. Consider the following scenario (where WEB and GPU denote
events that occur in the web and GPU processes, respectively):
1. (WEB) RemoteImageBufferProxy A is created in the web process.
2. (GPU) RemoteImageBuffer A is created in the GPU process, along with an image buffer backend for A.
3. (WEB) RemoteImageBufferProxy B is created in the web process.
4. (GPU) RemoteImageBuffer B is created in the GPU process, along with an image buffer backend for B.
5. (WEB) Something calls `RemoteImageBufferProxy::ensureBackendCreated()` on B.
6. (WEB) We receive the `RemoteRenderingBackendProxy::DidCreateImageBufferBackend` message for A.
B subsequently finishes waiting for backend creation because it incorrectly believes that its backend has been
initialized, and we end up crashing downstream in `RemoteLayerBackingStore::encode()` because
`ensureBackendCreated()` for B returned null.
To fix this, we adopt a similar strategy to what we use in `waitForDidFlushWithTimeout()`, and repeatedly call
`waitForDidCreateImageBufferBackend()` until the backend has been created (allowing for an arbitrary number of
IPC timeouts or failures). Since this counter is only incremented upon timeout (or any other kind of IPC
communication failure), we'll simply receive `RemoteRenderingBackendProxy::DidCreateImageBufferBackend` messages
until we confirm (in the web process) that the current image buffer's backend has been initialized.
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::waitForDidCreateImageBufferBackend):
Drive-by fix: clarify this code by returning an enum to indicate whether or not we successfully received a
backend creation IPC response from the GPU process, instead of just returning a bool.
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
2020-12-14 Sihui Liu <sihui_liu@apple.com>
Implement recognizer for SpeechRecognition
https://bugs.webkit.org/show_bug.cgi?id=219459
<rdar://problem/71914465>
Reviewed by Youenn Fablet.
* UIProcess/SpeechRecognitionServer.cpp:
(WebKit::SpeechRecognitionServer::SpeechRecognitionServer):
(WebKit::SpeechRecognitionServer::requestPermissionForRequest):
(WebKit::SpeechRecognitionServer::handleRequest):
(WebKit::SpeechRecognitionServer::abort):
(WebKit::SpeechRecognitionServer::invalidate):
* UIProcess/SpeechRecognitionServer.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::createSpeechRecognitionServer):
2020-12-14 Chris Dumez <cdumez@apple.com>
[GPUProcess] Crash under AudioDestinationCocoa::setIsPlaying(bool)
https://bugs.webkit.org/show_bug.cgi?id=219809
Reviewed by Eric Carlson.
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::create):
* WebProcess/GPU/media/RemoteAudioDestinationProxy.h:
* WebProcess/GPU/media/WebMediaStrategy.cpp:
(WebKit::WebMediaStrategy::createAudioDestination):
* WebProcess/GPU/media/WebMediaStrategy.h:
2020-12-14 Carlos Garcia Campos <cgarcia@igalia.com>
[WPE][GTK] Should enable WebProcessCache
https://bugs.webkit.org/show_bug.cgi?id=219689
Reviewed by Adrian Perez de Castro.
* UIProcess/API/glib/WebKitWebContext.cpp:
(webkitWebContextConstructed):
2020-12-13 Yusuke Suzuki <ysuzuki@apple.com>
[JSC] Introduce vmEntryCustomAccessor and vmEntryHostFunction for JITCage
https://bugs.webkit.org/show_bug.cgi?id=219847
Reviewed by Mark Lam.
* Shared/WebKitJITOperations.cpp:
(WebKit::populateJITOperations):
2020-12-13 Andy Estes <aestes@apple.com>
[Mac] Create a MediaToolbox format reader plug-in for WebM
https://bugs.webkit.org/show_bug.cgi?id=218908
<rdar://problem/71373264>
Reviewed by Eric Carlson.
Implemented a MTPluginFormatReader for WebM containers. The format reader is responsible
for parsing a WebM byte stream (using SourceBufferParserWebM) into audio, video, and text
tracks. MediaToolbox can then create cursors to step through the tracks' samples in decode
or presentation order and access each sample's metadata (e.g., format description,
duration, key-frame status) and location in the byte stream.
* FeatureFlags/WebKit.plist:
* MediaFormatReaderPlugIn/Info.plist:
* MediaFormatReaderPlugIn/MediaFormatReaderPlugIn.cpp:
* Shared/API/c/mac/WKFormatReader.cpp: Added.
* Shared/API/c/mac/WKFormatReader.h: Added.
* Shared/WebPreferencesDefaultValues.cpp:
* Shared/WebPreferencesDefaultValues.h:
* Shared/mac/MediaFormatReader/CoreMediaWrapped.cpp: Added.
* Shared/mac/MediaFormatReader/CoreMediaWrapped.h: Added.
* Shared/mac/MediaFormatReader/FormatReader.cpp: Added.
* Shared/mac/MediaFormatReader/FormatReader.h: Added.
* Shared/mac/MediaFormatReader/MediaFormatReader.cpp: Added.
* Shared/mac/MediaFormatReader/MediaFormatReader.h: Added.
* Shared/mac/MediaFormatReader/SampleCursor.cpp: Added.
* Shared/mac/MediaFormatReader/SampleCursor.h: Added.
* Shared/mac/MediaFormatReader/TrackReader.cpp: Added.
* Shared/mac/MediaFormatReader/TrackReader.h: Added.
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
2020-12-13 Tim Horton <timothy_horton@apple.com>
Fix the watchOS build after r270712
<rdar://problem/72265227>
* Shared/ios/WebIOSEventFactory.h:
* Shared/ios/WebIOSEventFactory.mm:
(WebIOSEventFactory::createWebWheelEvent):
2020-12-12 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn][iOS] Turn on modern WebAuthn for default browsers
https://bugs.webkit.org/show_bug.cgi?id=219823
<rdar://problem/72250436>
Reviewed by Brent Fulgham.
Covered by manual tests.
This patch turns on modern WebAuthn on iOS by default. In addition to that,
it turns on WebAuthn for all default browsers.
* WebProcess/WebAuthentication/WebAuthenticatorCoordinator.cpp:
(WebKit::WebCore::isWebBrowser):
(WebKit::WebAuthenticatorCoordinator::makeCredential):
(WebKit::WebAuthenticatorCoordinator::getAssertion):
(WebKit::WebAuthenticatorCoordinator::isUserVerifyingPlatformAuthenticatorAvailable):
For apps that turn on WebAuthn by overriding the experimental feature settings, return silently.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences)
Does the meat.
2020-12-12 Alexander Mikhaylenko <alexm@gnome.org>
[GTK4] Web view is incorrectly scaled on hidpi
https://bugs.webkit.org/show_bug.cgi?id=218353
Reviewed by Adrian Perez de Castro.
* UIProcess/gtk/AcceleratedBackingStoreWayland.cpp:
(WebKit::AcceleratedBackingStoreWayland::snapshot): Account for the scale factor.
2020-12-12 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r270692.
https://bugs.webkit.org/show_bug.cgi?id=219828
Introduced crash
Reverted changeset:
"[macOS] Remove access to com.apple.MTLCompilerService in the
WebContent process"
https://bugs.webkit.org/show_bug.cgi?id=219774
https://trac.webkit.org/changeset/270692
2020-12-12 Wenson Hsieh <wenson_hsieh@apple.com>
Unreviewed, reverting r270674.
Broke web content rendering on iOS
Reverted changeset:
"[iOS] Remove access to the IOKIt class AGXDeviceUserClient in
the WebContent process"
https://bugs.webkit.org/show_bug.cgi?id=219733
https://trac.webkit.org/changeset/270674
2020-12-11 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process][MSE] Implement SourceBuffer::reportExtraMemoryAllocated()
https://bugs.webkit.org/show_bug.cgi?id=219812
Reviewed by Eric Carlson.
Add a new IPC message `SourceBufferPrivateReportExtraMemoryCost` to forward the callback
to the web process.
* GPUProcess/media/RemoteSourceBufferProxy.cpp:
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateReportExtraMemoryCost):
* GPUProcess/media/RemoteSourceBufferProxy.h:
* WebProcess/GPU/media/SourceBufferPrivateRemote.cpp:
(WebKit::SourceBufferPrivateRemote::sourceBufferPrivateReportExtraMemoryCost):
* WebProcess/GPU/media/SourceBufferPrivateRemote.h:
* WebProcess/GPU/media/SourceBufferPrivateRemote.messages.in:
2020-12-11 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Adopt new UI for the Platform Authenticator getAssertion flow
https://bugs.webkit.org/show_bug.cgi?id=219710
<rdar://problem/72154800>
Reviewed by Brent Fulgham.
This patch adopts the new UI for the platform authenticator getAssertion flow. It's more or less
the same as the security key one except an LAContext will be passed in from the UI. Now LocalAuthenticator
will just utilize the LAContext from the UI.
* UIProcess/WebAuthentication/Cocoa/AuthenticatorPresenterCoordinator.h:
* UIProcess/WebAuthentication/Cocoa/AuthenticatorPresenterCoordinator.mm:
(WebKit::AuthenticatorPresenterCoordinator::AuthenticatorPresenterCoordinator):
(WebKit::AuthenticatorPresenterCoordinator::~AuthenticatorPresenterCoordinator):
(WebKit::AuthenticatorPresenterCoordinator::selectAssertionResponse):
(WebKit::AuthenticatorPresenterCoordinator::didSelectAssertionResponse):
* UIProcess/WebAuthentication/Cocoa/LocalAuthenticator.mm:
(WebKit::LocalAuthenticator::continueGetAssertionAfterResponseSelected):
* UIProcess/WebAuthentication/Cocoa/WKASCAuthorizationPresenterDelegate.mm:
(-[WKASCAuthorizationPresenterDelegate authorizationPresenter:credentialRequestedForLoginChoice:authenticatedContext:completionHandler:]):
2020-12-11 John Wilander <wilander@apple.com>
PCM: Change debugModeSecondsUntilSend from 60 seconds to 10 seconds
https://bugs.webkit.org/show_bug.cgi?id=219816
<rdar://problem/70779862>
Unreviewed change to a constant, only used for debugging.
* NetworkProcess/PrivateClickMeasurementManager.cpp:
debugModeSecondsUntilSend changed to 10 seconds.
(WebKit::PrivateClickMeasurementManager::attribute):
Fixed bad indentation.
2020-12-11 Ryosuke Niwa <rniwa@webkit.org>
[GPU Process] Cache Font objects
https://bugs.webkit.org/show_bug.cgi?id=219672
Reviewed by Wenson Hsieh.
Partially based on the patch made by Said Abou-Hallawa.
This patch introduces a mechanism to cache Font objects in the GPU process much the same way
NativeImage objects are cached in the GPU process, and re-enables Web Fonts.
Fonts are removed from the GPU process when either they hadn't been used for 4 rendering updates
or a memory warning has been issued.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::submit): Added the font cache.
(WebKit::RemoteRenderingBackend::cacheFont): Added.
(WebKit::RemoteRenderingBackend::deleteAllFonts): Added.
* GPUProcess/graphics/RemoteRenderingBackend.h:
* GPUProcess/graphics/RemoteRenderingBackend.messages.in: Added CacheFont and DeleteAllFonts
as new IPC messages.
* GPUProcess/graphics/RemoteResourceCache.cpp:
(WebKit::RemoteResourceCache::cacheFont): Added.
(WebKit::RemoteResourceCache::deleteAllFonts): Added. Used when a memory warning is issued.
* GPUProcess/graphics/RemoteResourceCache.h:
(WebKit::RemoteResourceCache::fonts const): Added.
* Platform/IPC/FontReference.h: Added.
(IPC::FontReference): Added.
* Scripts/webkit/messages.py:
(types_that_cannot_be_forward_declared): Added FontReference.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<Ref<Font>>::encode): Encode the rendering resource identifier of Font.
(IPC::ArgumentCoder<Ref<Font>>::decode): Ditto for decoding.
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
(WebKit::RemoteImageBufferProxy::cacheFont): Added.
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::cacheFont): Added.
(WebKit::RemoteRenderingBackendProxy::deleteAllFonts): Added.
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
* WebProcess/GPU/graphics/RemoteResourceCacheProxy.cpp:
(WebKit::RemoteResourceCacheProxy::cacheFont): Added. It updates the total number of fonts
being used since the last rendering update and records the self assigned identifier for
the next rendering update.
(WebKit::RemoteResourceCacheProxy::didFinalizeRenderingUpdate): Added. Called at the end of
each rendering update. It removes all fonts unused in the lsat 4 rendering updates if at least
1/4 of the fonts in the GPU process were not used in this rendering update.
(WebKit::RemoteResourceCacheProxy::releaseMemory): Added. Removes all the fonts to free up
memory when a memory warning is issued.
* WebProcess/GPU/graphics/RemoteResourceCacheProxy.h:
(WebKit::RemoteResourceCacheProxy): Added m_fontLastRenderingUpdateMap which keeps track of
the rendering resource identifiers of the fonts cached in the GPU process.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences): Re-enables the Web Fonts.
(WebKit::WebPage::finalizeRenderingUpdate): Added the call to RemoteRenderingBackendProxy's
didFinalizeRenderingUpdate.
(WebKit::WebPage::releaseMemory): Added. Calls RemoteRenderingBackendProxy's function.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::initializeWebProcess): Added the call to WebPage::releaseMemory in
the memory pressure handler.
(WebKit::WebProcess::prepareToSuspend): Ditto for the process suspension.
2020-12-11 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Adopt new UI for the Security Key getAssertion flow
https://bugs.webkit.org/show_bug.cgi?id=219711
<rdar://problem/72154840>
Reviewed by Brent Fulgham.
This patch adopts the new UI for the security key getAssertion flow which contains two part:
1. showing a informative UI to ask the user to connect their security keys,
2. showing an account picker for users to select a credential to use.
Covered by manual tests.
* Platform/spi/Cocoa/AuthenticationServicesCoreSPI.h:
Paperwork.
* UIProcess/WebAuthentication/Cocoa/AuthenticatorPresenterCoordinator.h:
* UIProcess/WebAuthentication/Cocoa/AuthenticatorPresenterCoordinator.mm:
(WebKit::AuthenticatorPresenterCoordinator::AuthenticatorPresenterCoordinator):
(WebKit::AuthenticatorPresenterCoordinator::selectAssertionResponse):
(WebKit::AuthenticatorPresenterCoordinator::didSelectAssertionResponse):
* UIProcess/WebAuthentication/Cocoa/WKASCAuthorizationPresenterDelegate.mm:
(-[WKASCAuthorizationPresenterDelegate authorizationPresenter:credentialRequestedForLoginChoice:authenticatedContext:completionHandler:]):
Implements the two flows.
2020-12-11 John Wilander <wilander@apple.com>
PCM: Add feature flag support in the network process
https://bugs.webkit.org/show_bug.cgi?id=219811
<rdar://problem/72242633>
Reviewed by Alex Christensen.
No new tests. This just forwards a preference to the network process.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::initializeNetworkProcess):
Stores the incoming enablePrivateClickMeasurement from
WebKit::NetworkProcessCreationParameters.
(WebKit::NetworkProcess::setPrivateClickMeasurementEnabled):
(WebKit::NetworkProcess::privateClickMeasurementEnabled const):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcessCreationParameters.cpp:
New field enablePrivateClickMeasurement.
(WebKit::NetworkProcessCreationParameters::encode const):
(WebKit::NetworkProcessCreationParameters::decode):
* NetworkProcess/NetworkProcessCreationParameters.h:
* NetworkProcess/PrivateClickMeasurementManager.cpp:
(WebKit::PrivateClickMeasurementManager::storeUnattributed):
Early return if the feature is not enabled.
(WebKit::PrivateClickMeasurementManager::handleAttribution):
Early return if the feature is not enabled.
(WebKit::PrivateClickMeasurementManager::fireConversionRequest):
Early return if the feature is not enabled.
(WebKit::PrivateClickMeasurementManager::featureEnabled const):
New convenience function.
* NetworkProcess/PrivateClickMeasurementManager.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeNetworkProcess):
Sets the new enablePrivateClickMeasurement in
WebKit::NetworkProcessCreationParameters.
2020-12-11 Jiewen Tan <jiewen_tan@apple.com>
Unreviewed, build fix after r270694
* UIProcess/WebAuthentication/Cocoa/AuthenticatorPresenterCoordinator.mm:
(WebKit::AuthenticatorPresenterCoordinator::AuthenticatorPresenterCoordinator):
Adds a instance method guard for new SPI.
2020-12-11 Jer Noble <jer.noble@apple.com>
[Cocoa] Add Experimental VP8 support
https://bugs.webkit.org/show_bug.cgi?id=219732
Reviewed by Eric Carlson.
Add new WebPageCreationParameter flags to enable VP8 decoders in the WebContent and GPU
processes when the experimental feature is enabled.
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::enableVP9Decoders):
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/GPUConnectionToWebProcess.messages.in:
* GPUProcess/GPUProcess.cpp:
(WebKit::GPUProcess::enableVP9Decoders):
* GPUProcess/GPUProcess.h:
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultVP8DecoderEnabled):
* Shared/WebPreferencesDefaultValues.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
* WebProcess/GPU/GPUProcessConnection.cpp:
(WebKit::GPUProcessConnection::updateParameters):
* WebProcess/GPU/GPUProcessConnection.h:
(WebKit::GPUProcessConnection::isVP8DecoderEnabled const):
* WebProcess/WebPage/WebPage.cpp:
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::enableVP8SWDecoder):
* WebProcess/WebProcess.h:
2020-12-11 Chris Dumez <cdumez@apple.com>
Potential null dereference of m_frame under DocumentLoader::stopLoading()
https://bugs.webkit.org/show_bug.cgi?id=219786
<rdar://71945402>
Reviewed by Geoffrey Garen.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::close):
Add some release logging.
2020-12-11 Tim Horton <timothy_horton@apple.com>
Fix the iOS !UISCROLLVIEW_ASYNCHRONOUS_SCROLL_EVENT_HANDLING build.
* Shared/ios/WebIOSEventFactory.mm:
(WebIOSEventFactory::createWebWheelEvent):
2020-12-11 Tim Horton <timothy_horton@apple.com>
Trackpad and Mouse scroll events on iPad only fire "pointermove" -- not "wheel"
https://bugs.webkit.org/show_bug.cgi?id=210071
<rdar://problem/54616853>
Reviewed by Simon Fraser.
* Platform/spi/ios/UIKitSPI.h:
Add some SPI.
* Shared/ios/WebIOSEventFactory.h:
* Shared/ios/WebIOSEventFactory.mm:
(toWebPhase):
(WebIOSEventFactory::createWebWheelEvent):
Add a UIScrollEvent->WebWheelEvent conversion helper.
* UIProcess/API/Cocoa/WKWebViewInternal.h:
* UIProcess/API/ios/WKWebViewIOS.h:
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _setupScrollAndContentViews]):
Enable async UIScrollEvent handling for WKScrollView.
(-[WKWebView _scrollView:asynchronouslyHandleScrollEvent:completion:]):
Adopt new UIKit SPI to asynchronously defer UIScrollEvents.
We pass them to the Web Content process, where they are processed
*only* for event handling, not for scrolling.
If the event is not cancelable, we will synchronously reply that it was
not handled; if it is cancelable, or we don't yet know if it will be,
we'll wait to hear back from the Web Content process before replying.
UIKit will wait until our reply to apply the UIScrollEvent to the UIScrollView.
* UIProcess/PageClient.h:
* UIProcess/RemoteLayerTree/ios/RemoteLayerTreeViews.h:
* UIProcess/RemoteLayerTree/ios/RemoteLayerTreeViews.mm:
(WebKit::eventListenerTypesAtPoint):
Expose a mechanism for retrieving the event listener types at a given point,
similar to the existing mechanism for touch event listeners.
(-[WKChildScrollView initWithFrame:]):
Enable async UIScrollEvent handling for WKChildScrollView.
* UIProcess/RemoteLayerTree/ios/ScrollingTreeScrollingNodeDelegateIOS.h:
* UIProcess/RemoteLayerTree/ios/ScrollingTreeScrollingNodeDelegateIOS.mm:
(-[WKScrollingNodeScrollViewDelegate _scrollView:asynchronouslyHandleScrollEvent:completion:]):
(WebKit::ScrollingTreeScrollingNodeDelegateIOS::handleAsynchronousCancelableScrollEvent):
Plumb async scroll events for sub-scrollable regions through PageClient
to WKWebView; we don't actually care which UIScrollView they're handed to,
since we re-hit-test ourselves.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::dispatchWheelEventWithoutScrolling):
* UIProcess/WebPageProxy.h:
* UIProcess/ios/PageClientImplIOS.h:
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::handleAsynchronousCancelableScrollEvent):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::wheelEvent):
(WebKit::WebPage::dispatchWheelEventWithoutScrolling):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-12-11 Brent Fulgham <bfulgham@apple.com>
Expose API for enabling/disabling Private Click Measurement
https://bugs.webkit.org/show_bug.cgi?id=219791
<rdar://problem/70502198>
Reviewed by Alex Christensen.
Tested by existing PCM test cases.
This patch moves the flag for enabling and disabling PCM support from the Experimental
features settings to standard settings in preparation for exposing the option in more
permanent UI. This patch does not make any changes in behavior.
* UIProcess/API/Cocoa/WKPreferences.mm:
(-[WKPreferences _privateClickMeasurementEnabled]):
(-[WKPreferences _setPrivateClickMeasurementEnabled:]):
* UIProcess/API/Cocoa/WKPreferencesPrivate.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences):
2020-12-11 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process][MSE] Implement some required functions with new IPC messages and remove some unneeded functions
https://bugs.webkit.org/show_bug.cgi?id=219788
Reviewed by Eric Carlson.
Implement unimplemented functions and remove unneeded ones.
This patch also cleans up `RemoteSourceBufferProxy`.
* GPUProcess/media/RemoteSourceBufferProxy.cpp:
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateStreamEndedWithDecodeError):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateAppendError):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateDidDropSample):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateDidReceiveRenderingError):
* GPUProcess/media/RemoteSourceBufferProxy.h:
* WebProcess/GPU/media/SourceBufferPrivateRemote.cpp:
(WebKit::SourceBufferPrivateRemote::sourceBufferPrivateStreamEndedWithDecodeError):
(WebKit::SourceBufferPrivateRemote::sourceBufferPrivateAppendError):
(WebKit::SourceBufferPrivateRemote::sourceBufferPrivateDidDropSample):
(WebKit::SourceBufferPrivateRemote::sourceBufferPrivateDidReceiveRenderingError):
(WebKit::SourceBufferPrivateRemote::flush): Deleted.
(WebKit::SourceBufferPrivateRemote::isReadyForMoreSamples): Deleted.
(WebKit::SourceBufferPrivateRemote::notifyClientWhenReadyForMoreSamples): Deleted.
(WebKit::SourceBufferPrivateRemote::canSetMinimumUpcomingPresentationTime const): Deleted.
(WebKit::SourceBufferPrivateRemote::setMinimumUpcomingPresentationTime): Deleted.
(WebKit::SourceBufferPrivateRemote::clearMinimumUpcomingPresentationTime): Deleted.
* WebProcess/GPU/media/SourceBufferPrivateRemote.h:
* WebProcess/GPU/media/SourceBufferPrivateRemote.messages.in:
2020-12-11 Chris Dumez <cdumez@apple.com>
[GPUProcess] WebAudio rendering quantum is 15 instead of 128 when the GPU Process is enabled
https://bugs.webkit.org/show_bug.cgi?id=219797
Reviewed by Eric Carlson.
WebAudio rendering quantum is 15 instead of 128 when the GPU Process is enabled. This generates a lot more
IPC unnecessarily, not to mention how WebAudio code expects a rendering quantum of 128 and could potentially
do bad things with a lower value.
The issue was that RemoteAudioSessionProxyManager::setPreferredBufferSizeForProcess() was iterating over
all RemoteAudioSessionProxy objects to find out the lowest rendering quantum. If there is a proxy with
a preferred buffer size that is not set, the value ends up being 0. When then ask core audio for a rendering
quantum of 0, and they end up using 15 (likely the minimum).
* GPUProcess/media/RemoteAudioSessionProxyManager.cpp:
(WebKit::RemoteAudioSessionProxyManager::setPreferredBufferSizeForProcess):
2020-12-11 Don Olmstead <don.olmstead@sony.com>
[PlayStation] Link additional libraries into processes after r270690
https://bugs.webkit.org/show_bug.cgi?id=219795
Unreviewed build fix after r270690.
Some libraries were being propagated from WebKit before the change so just add
them to the Process_PRIVATE_LIBRARIES.
* PlatformPlayStation.cmake:
2020-12-11 Fujii Hironori <Hironori.Fujii@sony.com>
[WinCairo] Enable accelerated WebGL canvas
https://bugs.webkit.org/show_bug.cgi?id=219768
Reviewed by Don Olmstead.
prepareForRendering should be called before calling
flushPendingLayerChanges to make the GL context current because it
uses OpenGL API.
* WebProcess/WebPage/CoordinatedGraphics/LayerTreeHostTextureMapper.cpp:
(WebKit::LayerTreeHost::compositeLayersToContext):
(WebKit::LayerTreeHost::flushAndRenderLayers):
2020-12-10 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Adopt new UI for the Platform Authenticator makeCredential flow
https://bugs.webkit.org/show_bug.cgi?id=219709
<rdar://problem/72154774>
Reviewed by Brent Fulgham.
This patch adopts the new UI for the platform authenticator makeCredential flow. In this new design,
the UI will be responsible for creating the LAContext and get it verified by the user. Therefore, WebKit
just needs to reverify the LAContext in non-UI mode against the SecAccessControlRef without handling any
LA UI anymore.
Covered by manual tests.
* Platform/spi/Cocoa/AuthenticationServicesCoreSPI.h:
Adds a new class which represents a successful operation.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebAuthn.sb:
Adds rule to allow the WebAuthn process to access Keychain and SEP.
* UIProcess/WebAuthentication/Authenticator.h:
(WebKit::Authenticator::setWebAuthenticationModernEnabled):
(WebKit::Authenticator::webAuthenticationModernEnabled const):
Adds a flag to know which UI to call.
* UIProcess/WebAuthentication/AuthenticatorManager.cpp:
(WebKit::AuthenticatorManager::handleRequest):
(WebKit::AuthenticatorManager::enableModernWebAuthentication):
(WebKit::AuthenticatorManager::authenticatorAdded):
(WebKit::AuthenticatorManager::requestLAContextForUserVerification):
(WebKit::AuthenticatorManager::invokePendingCompletionHandler):
* UIProcess/WebAuthentication/AuthenticatorManager.h:
Adds plumbing to the new UI method.
* UIProcess/WebAuthentication/Cocoa/AuthenticationServicesCoreSoftLink.h:
* UIProcess/WebAuthentication/Cocoa/AuthenticationServicesCoreSoftLink.mm:
Paperwork.
* UIProcess/WebAuthentication/Cocoa/AuthenticatorPresenterCoordinator.h:
(WebKit::AuthenticatorPresenterCoordinator::setCredentialRequestHandler):
This handler is from the UI delegate which is supposed to call if the operation succeeds.
* UIProcess/WebAuthentication/Cocoa/AuthenticatorPresenterCoordinator.mm:
(WebKit::AuthenticatorPresenterCoordinator::AuthenticatorPresenterCoordinator):
(WebKit::AuthenticatorPresenterCoordinator::requestLAContextForUserVerification):
(WebKit::AuthenticatorPresenterCoordinator::dimissPresenter):
(WebKit::AuthenticatorPresenterCoordinator::setLAContext):
* UIProcess/WebAuthentication/Cocoa/LocalAuthenticator.h:
* UIProcess/WebAuthentication/Cocoa/LocalAuthenticator.mm:
(WebKit::LocalAuthenticator::makeCredential):
(WebKit::LocalAuthenticator::continueMakeCredentialAfterReceivingLAContext):
* UIProcess/WebAuthentication/Cocoa/LocalConnection.h:
* UIProcess/WebAuthentication/Cocoa/LocalConnection.mm:
(WebKit::LocalConnection::verifyUser):
* UIProcess/WebAuthentication/Cocoa/WKASCAuthorizationPresenterDelegate.mm:
(-[WKASCAuthorizationPresenterDelegate authorizationPresenter:credentialRequestedForLoginChoice:authenticatedContext:completionHandler:]):
(-[WKASCAuthorizationPresenterDelegate dispatchCoordinatorCallback:]):
Implements the new UI logic.
2020-12-11 Per Arne Vollan <pvollan@apple.com>
[macOS] Remove access to com.apple.MTLCompilerService in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=219774
<rdar://problem/70497100>
Reviewed by Brent Fulgham.
Remove access to com.apple.MTLCompilerService in the WebContent process on macOS, since there are no reports of this being used.
No new tests, since the result will depend on macOS version.
* WebProcess/com.apple.WebProcess.sb.in:
2020-12-11 Per Arne Vollan <pvollan@apple.com>
[macOS] Remove Mobile Asset access from the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=219738
<rdar://problem/60650074>
Reviewed by Brent Fulgham.
Remove Mobile Asset access from the WebContent process on macOS, since there are no reports of this being used.
No new tests, since the result will depend on macOS version.
* WebProcess/com.apple.WebProcess.sb.in:
2020-12-11 Don Olmstead <don.olmstead@sony.com>
[CMake] Determine correct visibility for linked frameworks
https://bugs.webkit.org/show_bug.cgi?id=210366
Reviewed by Michael Catanzaro.
Set WebKit_FRAMEWORKS to determine correct linkage for the library. Remove
explicit setting of STATICALLY_LINKED_WITH_${framework} and $<TARGET_OBJECTS:${framework}>
by ports.
* CMakeLists.txt:
* PlatformFTW.cmake:
* PlatformPlayStation.cmake:
* PlatformWPE.cmake:
* PlatformWin.cmake:
2020-12-11 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r270654.
https://bugs.webkit.org/show_bug.cgi?id=219793
Introduced crash
Reverted changeset:
"[macOS] The development WebContent XPC service is missing a
required entitlement"
https://bugs.webkit.org/show_bug.cgi?id=219730
https://trac.webkit.org/changeset/270654
2020-12-11 Carlos Garcia Campos <cgarcia@igalia.com>
Protect the WebProcessPool earlier in WebProcessCache::addProcessIfPossible
https://bugs.webkit.org/show_bug.cgi?id=219777
Reviewed by Adrian Perez de Castro.
We are taking a reference of the WebProcessPool in the isResponsive() completion handler lambda, but the
WebProcessPool can be destroyed earlier when CachedProcess is created if the WebProcessProxy contains the last
strong reference.
* UIProcess/WebProcessCache.cpp:
(WebKit::WebProcessCache::addProcessIfPossible):
2020-12-11 Brent Fulgham <bfulgham@apple.com>
[macOS] Only extend access to the RootDomainUserClient IOKit class if the GPU Process is not used
https://bugs.webkit.org/show_bug.cgi?id=219758
<rdar://problem/70464431>
Reviewed by Per Arne Vollan.
The RootDomainUserClient IOKit class appears to only be used by media on Apple Silicon. Dynamically extend access to this
class for Apple Silicon builds if the GPU process is not being used.
This patch also includes telemetry to confirm that it is not needed on Intel, or for any other purposes. It does not
block access to the IOKit class yet.
* UIProcess/WebPageProxy.cpp:
(WebKit::mediaRelatedIOKitClasses):
* WebProcess/com.apple.WebProcess.sb.in:
2020-12-11 Don Olmstead <don.olmstead@sony.com>
[GLib] Move GCrypt initialization to platformInitialize
https://bugs.webkit.org/show_bug.cgi?id=219749
Reviewed by Adrian Perez de Castro.
It makes more sense for initialization of GCrypt to occur in platformInitialize
than in the main method of the process. This means that the process does not have
to link gcrypt libraries and can just link WebKit only.
* NetworkProcess/EntryPoint/unix/NetworkProcessMain.cpp:
(main):
* NetworkProcess/soup/NetworkProcessMainSoup.cpp:
* WebProcess/EntryPoint/unix/WebProcessMain.cpp:
(main):
* WebProcess/gtk/WebProcessMainGtk.cpp:
(WebKit::WebProcessMain):
* WebProcess/wpe/WebProcessMainWPE.cpp:
2020-12-11 Per Arne Vollan <pvollan@apple.com>
[iOS] Remove access to the IOKIt class AGXDeviceUserClient in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=219733
<rdar://problem/70628137>
Reviewed by Brent Fulgham.
Remove access to the IOKIt class AGXDeviceUserClient in the WebContent process on iOS, since this is not needed when allowing access to IOGPU.
The IOGPU class is the superclass of AGXDeviceUserClient, and is the preferred IOKit reference to use according to the Sandbox team.
* Resources/SandboxProfiles/ios/com.apple.WebKit.GPU.sb:
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* UIProcess/WebPageProxy.cpp:
(WebKit::gpuIOKitClasses):
2020-12-11 Per Arne Vollan <pvollan@apple.com>
[macOS] Add telemetry for notification message filtering in sandbox
https://bugs.webkit.org/show_bug.cgi?id=219684
<rdar://problem/69819260>
Reviewed by Brent Fulgham.
Add telemetry for notification message filtering in sandbox in the WebContent process on macOS.
No new tests, since this just adds telemetry.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-12-11 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r270655.
https://bugs.webkit.org/show_bug.cgi?id=219772
Introduced sandbox compile error
Reverted changeset:
"[iOS] Add missing endpoint in sandbox message filter"
https://bugs.webkit.org/show_bug.cgi?id=219731
https://trac.webkit.org/changeset/270655
2020-12-10 Alex Christensen <achristensen@webkit.org>
Accept click measurement data from hosting application
https://bugs.webkit.org/show_bug.cgi?id=219763
<rdar://problem/72121094>
Reviewed by John Wilander.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::buildPrivateClickMeasurementFromDatabase):
* Platform/ios/UIKitSoftLink.h: Added.
* Platform/ios/UIKitSoftLink.mm: Added.
* Platform/spi/ios/UIKitSPI.h:
* SourcesCocoa.txt:
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/API/Cocoa/WKWebViewPrivateForTesting.h:
* UIProcess/API/Cocoa/WKWebViewTesting.mm:
(-[WKWebView _setPrivateClickMeasurementOverrideTimerForTesting:completionHandler:]):
(-[WKWebView _setPrivateClickMeasurementConversionURLForTesting:completionHandler:]):
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _setEventAttribution:]):
(-[WKWebView _eventAttribution]):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didCommitLoadForFrame):
(WebKit::WebPageProxy::createNewPage):
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::setPrivateClickMeasurement):
(WebKit::WebPageProxy::privateClickMeasurement const):
* WebKit.xcodeproj/project.pbxproj:
2020-12-10 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process][MSE] Implement several methods in SourceBufferPrivateRemote to make seeking more reliable
https://bugs.webkit.org/show_bug.cgi?id=219757
Reviewed by Eric Carlson.
Add the required IPC messages and implement three functions of `SourceBufferPrivateRemote`.
* GPUProcess/media/RemoteSourceBufferProxy.cpp:
(WebKit::RemoteSourceBufferProxy::abort):
(WebKit::RemoteSourceBufferProxy::resetParserState):
(WebKit::RemoteSourceBufferProxy::removedFromMediaSource):
* GPUProcess/media/RemoteSourceBufferProxy.h:
* GPUProcess/media/RemoteSourceBufferProxy.messages.in:
* WebProcess/GPU/media/SourceBufferPrivateRemote.cpp:
(WebKit::SourceBufferPrivateRemote::abort):
(WebKit::SourceBufferPrivateRemote::resetParserState):
(WebKit::SourceBufferPrivateRemote::removedFromMediaSource):
2020-12-10 Andy Estes <aestes@apple.com>
[Mac] Create a format reader plug-in for WebM
https://bugs.webkit.org/show_bug.cgi?id=218908
<rdar://problem/71373264>
Reviewed by Eric Carlson.
Added a new bundle target that builds a MediaToolbox format reader
plug-in and installs it in the WebContent and GPU XPC service bundles.
The plug-in does nothing for now, but in follow-up patches will learn
how to read WebM containers.
* Configurations/MediaFormatReader.xcconfig: Added.
* MediaFormatReaderPlugIn/Info.plist: Added. Defines a CMClassImplementations array
specifying the plug-in factory function and the supported media types / extensions.
* MediaFormatReaderPlugIn/MediaFormatReaderPlugIn.cpp: Added.
* WebKit.xcodeproj/project.pbxproj:
2020-12-10 Chris Dumez <cdumez@apple.com>
[GPUProcess] Cut in half amount of IPC needed to do WebAudio
https://bugs.webkit.org/show_bug.cgi?id=219715
Reviewed by Geoff Garen.
Cut in half amount of IPC needed to do WebAudio. Previously, we would send 2
IPCs per rendering quantum:
1. RemoteAudioDestinationProxy::RequestBuffer from GPUProcess to WebProcess
2. The response to RemoteAudioDestinationProxy::RequestBuffer which contained
the frame bounds for the RingBuffer.
This patch gets rid of IPC 2 by having the CARingBuffer class store its bounds
in its memory backing. As a result, when the CARingBuffer is using a shared
memory backing, the bounds get shared between processes, without any need for
IPC to sync them.
This is still too much IPC but this is a step in the right direction. We
can hopefully get rid of (or greatly reduce) IPC 1 in a follow-up patch.
* GPUProcess/media/RemoteAudioDestinationManager.cpp:
(WebKit::RemoteAudioDestination::audioSamplesStorageChanged):
(WebKit::RemoteAudioDestination::render):
* GPUProcess/media/RemoteAudioSourceProviderProxy.cpp:
(WebKit::RemoteAudioSourceProviderProxy::storageChanged):
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRenderer.cpp:
(WebKit::RemoteAudioMediaStreamTrackRenderer::audioSamplesStorageChanged):
(WebKit::RemoteAudioMediaStreamTrackRenderer::audioSamplesAvailable):
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRenderer.h:
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRenderer.messages.in:
* GPUProcess/webrtc/RemoteMediaRecorder.cpp:
(WebKit::RemoteMediaRecorder::audioSamplesStorageChanged):
(WebKit::RemoteMediaRecorder::audioSamplesAvailable):
* GPUProcess/webrtc/RemoteMediaRecorder.h:
* GPUProcess/webrtc/RemoteMediaRecorder.messages.in:
* Shared/Cocoa/SharedRingBufferStorage.cpp:
(WebKit::SharedRingBufferStorage::allocate):
(WebKit::SharedRingBufferStorage::sharedFrameBounds):
(WebKit::SharedRingBufferStorage::data):
(WebKit::SharedRingBufferStorage::getCurrentFrameBounds):
(WebKit::SharedRingBufferStorage::setCurrentFrameBounds):
(WebKit::SharedRingBufferStorage::updateFrameBounds):
(WebKit::SharedRingBufferStorage::flush):
* Shared/Cocoa/SharedRingBufferStorage.h:
* UIProcess/Cocoa/UserMediaCaptureManagerProxy.cpp:
* UIProcess/SpeechRecognitionRemoteRealtimeMediaSource.cpp:
(WebKit::SpeechRecognitionRemoteRealtimeMediaSource::remoteAudioSamplesAvailable):
* UIProcess/SpeechRecognitionRemoteRealtimeMediaSource.h:
* UIProcess/SpeechRecognitionRemoteRealtimeMediaSourceManager.cpp:
(WebKit::SpeechRecognitionRemoteRealtimeMediaSourceManager::remoteAudioSamplesAvailable):
* UIProcess/SpeechRecognitionRemoteRealtimeMediaSourceManager.h:
* UIProcess/SpeechRecognitionRemoteRealtimeMediaSourceManager.messages.in:
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::requestBuffer):
(WebKit::RemoteAudioDestinationProxy::storageChanged):
* WebProcess/GPU/media/RemoteAudioDestinationProxy.h:
* WebProcess/GPU/media/RemoteAudioDestinationProxy.messages.in:
* WebProcess/GPU/media/RemoteAudioSourceProviderManager.cpp:
(WebKit::RemoteAudioSourceProviderManager::RemoteAudio::setStorage):
(WebKit::RemoteAudioSourceProviderManager::RemoteAudio::audioSamplesAvailable):
* WebProcess/GPU/webrtc/AudioMediaStreamTrackRenderer.cpp:
(WebKit::AudioMediaStreamTrackRenderer::pushSamples):
(WebKit::AudioMediaStreamTrackRenderer::storageChanged):
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
(WebKit::MediaRecorderPrivate::audioSamplesAvailable):
(WebKit::MediaRecorderPrivate::storageChanged):
* WebProcess/Speech/SpeechRecognitionRealtimeMediaSourceManager.cpp:
* WebProcess/cocoa/RemoteCaptureSampleManager.cpp:
(WebKit::RemoteCaptureSampleManager::audioSamplesAvailable):
(WebKit::RemoteCaptureSampleManager::RemoteAudio::audioSamplesAvailable):
* WebProcess/cocoa/RemoteCaptureSampleManager.h:
* WebProcess/cocoa/RemoteCaptureSampleManager.messages.in:
2020-12-10 Per Arne Vollan <pvollan@apple.com>
[macOS] Deny mach-lookup to the service 'com.apple.SystemConfiguration.configd' in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=219680
<rdar://problem/69019138>
Reviewed by Brent Fulgham.
Deny mach-lookup to the service 'com.apple.SystemConfiguration.configd' in the WebContent process on macOS,
since there are no reports of this being used.
No new tests, since the result will depend on macOS version.
* WebProcess/com.apple.WebProcess.sb.in:
2020-12-10 Per Arne Vollan <pvollan@apple.com>
[macOS] Create sandbox extension for AppleAVDUserClient when Media in GPU process is not enabled
https://bugs.webkit.org/show_bug.cgi?id=219740
<rdar://problem/70496905>
Reviewed by Brent Fulgham.
Create sandbox extension for AppleAVDUserClient for WebContent process when Media in GPU process is not enabled on macOS,
since this IOKit class is only used for Media purposes.
* GPUProcess/mac/com.apple.WebKit.GPUProcess.sb.in:
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::mediaRelatedIOKitClasses):
(WebKit::WebPageProxy::creationParameters):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
* WebProcess/com.apple.WebProcess.sb.in:
2020-12-10 Per Arne Vollan <pvollan@apple.com>
[macOS] Add telemetry for notification message filtering in sandbox
https://bugs.webkit.org/show_bug.cgi?id=219684
<rdar://problem/69819260>
Reviewed by Brent Fulgham.
Add telemetry for notification message filtering in sandbox in the WebContent process on macOS.
No new tests, since this just adds telemetry.
* WebProcess/com.apple.WebProcess.sb.in:
2020-12-10 Per Arne Vollan <pvollan@apple.com>
[iOS] Add missing endpoint in sandbox message filter
https://bugs.webkit.org/show_bug.cgi?id=219731
<rdar://problem/71566760>
Reviewed by Brent Fulgham.
Add missing endpoint in the WebContent process' sandbox message filter on iOS.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-12-10 Per Arne Vollan <pvollan@apple.com>
[macOS] The development WebContent XPC service is missing a required entitlement
https://bugs.webkit.org/show_bug.cgi?id=219730
<rdar://problem/72145560>
Reviewed by Brent Fulgham.
The message filtering entitlement is missing in com.apple.WebKit.WebContent.Development.xpc on macOS.
* Scripts/process-entitlements.sh:
2020-12-10 Andres Gonzalez <andresg_22@apple.com>
Rework of AXObjectCache::isIsolatedTreeEnabled.
https://bugs.webkit.org/show_bug.cgi?id=219722
Reviewed by Chris Fleizach.
- Renamed AXObjecCache::canUseSecondaryAXThread as usedOnAXThread.
- No need for the additional thread check since isIsolatedTreeEnabled
will return true for any call off of the main thread.
* WebProcess/InjectedBundle/API/c/WKBundlePage.cpp:
(WKAccessibilityCanUseSecondaryAXThread):
* WebProcess/WebPage/mac/WKAccessibilityWebPageObjectMac.mm:
(-[WKAccessibilityWebPageObject accessibilityHitTest:]):
2020-12-10 Sihui Liu <sihui_liu@apple.com>
REGRESSION (r270574): [iOS] ASSERTION FAILED: !m_messageReceiverMapCount under WebKit::SpeechRecognitionRemoteRealtimeMediaSourceManager::~SpeechRecognitionRemoteRealtimeMediaSourceManager
https://bugs.webkit.org/show_bug.cgi?id=219745
<rdar://problem/72184764>
Reviewed by Youenn Fablet.
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::~WebProcessProxy): Remove SpeechRecognitionRemoteRealtimeMediaSourceManager from
MessageReceiverMap before destroying it.
2020-12-10 Timothy Hatcher <timothy@apple.com>
Add willInjectUserScriptForFrame to WKWebProcessPlugInLoadDelegate.
https://bugs.webkit.org/show_bug.cgi?id=219727
rdar://71987073
Reviewed by Brady Eidson.
* WebProcess/InjectedBundle/API/Cocoa/WKWebProcessPlugInLoadDelegate.h:
* WebProcess/InjectedBundle/API/mac/WKWebProcessPlugInBrowserContextController.mm:
(willInjectUserScriptForFrame): Added.
(setUpPageLoaderClient): Set client.willInjectUserScriptForFrame.
2020-12-10 Don Olmstead <don.olmstead@sony.com>
[GTK] Add even more exports to support hidden visibility
https://bugs.webkit.org/show_bug.cgi?id=219747
Reviewed by Michael Catanzaro.
Add WK_EXPORT to additional declarations.
* UIProcess/API/glib/WebKitSettingsPrivate.h:
* UIProcess/API/glib/WebKitWebViewInternal.h:
* UIProcess/API/gtk/WebKitWebViewBaseInternal.h:
2020-12-10 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process][MSE] Implement seeking support
https://bugs.webkit.org/show_bug.cgi?id=219658
Reviewed by Jer Noble.
Implement the required functions with IPC to support seeking.
* DerivedSources-input.xcfilelist:
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* GPUProcess/media/RemoteMediaSourceProxy.cpp:
(WebKit::RemoteMediaSourceProxy::seekToTime):
(WebKit::RemoteMediaSourceProxy::waitForSeekCompleted):
(WebKit::RemoteMediaSourceProxy::seekCompleted):
* GPUProcess/media/RemoteMediaSourceProxy.h:
* GPUProcess/media/RemoteMediaSourceProxy.messages.in:
* GPUProcess/media/RemoteSourceBufferProxy.cpp:
(WebKit::RemoteSourceBufferProxy::seekToTime):
* GPUProcess/media/RemoteSourceBufferProxy.h:
* GPUProcess/media/RemoteSourceBufferProxy.messages.in:
* Sources.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/media/MediaSourcePrivateRemote.cpp:
(WebKit::MediaSourcePrivateRemote::MediaSourcePrivateRemote):
(WebKit::MediaSourcePrivateRemote::~MediaSourcePrivateRemote):
(WebKit::MediaSourcePrivateRemote::waitForSeekCompleted):
(WebKit::MediaSourcePrivateRemote::seekCompleted):
(WebKit::MediaSourcePrivateRemote::seekToTime):
* WebProcess/GPU/media/MediaSourcePrivateRemote.h:
* WebProcess/GPU/media/MediaSourcePrivateRemote.messages.in: Added.
* WebProcess/GPU/media/SourceBufferPrivateRemote.cpp:
(WebKit::SourceBufferPrivateRemote::seekToTime):
* WebProcess/GPU/media/SourceBufferPrivateRemote.h:
2020-12-10 Alex Christensen <achristensen@webkit.org>
Introduce new download API
https://bugs.webkit.org/show_bug.cgi?id=217747
Reviewed by Brady Eidson.
Safari currently uses _WKDownload, which has evolved strangely over the last decade.
In order to make a nicer interface, we need to learn from those lessons and take a step back.
I did that, and here's what I came up with: WKDownload and WKDownloadDelegate!
Notable changes include:
1. The delegate now lives on the download object instead of the process pool.
2. WKDownload does not conform to NSCopying because we have NSMapTable instead of NSDictionary.
3. publishProgressAtURL is gone. That will be reimplemented in the UI process of the client that used it.
4. cancel has a completion handler to get the resume data instead of waiting for didCancel then getting it from the download object.
5. didFailWithError also gives you the resume data, so there is no way to get it from the download object. That's more about a fail/cancel
event than a property of the download.
6. wasUserInitiated is gone. Instead, WKNavigationDelegate has a callback that links a navigation action (where _isUserInitiated is exposed) to the download.
7. redirectChain is also gone. That can also be gotten from the link to the navigation and the download delegate's redirect callback.
8. _downloadDidStart is gone. Instead, we have completion handlers that expose a WKDownload once it's started. There's no need to
reference the download object before then anyways. Note: that's when the download "starts" which is before it receives the response,
so there will be no server delay in the difference between download object availabilities.
9. didReceiveServerRedirectToURL now gives you the option of continuing or cancelling. This reflects the amount of control the
WKNavigationDelegate has.
10. decideDestinationWithSuggestedFilename and didReceiveResponse have been merged, since they actually happen at the same time and
NSURLResponse has suggestedFilename API. allowOverwrite is also gone. It wasn't used, and not really needed.
11. didCreateDestination is gone. That callback time wasn't really necessary. If you need to know when it's created, wait until the first
didWriteData callback. The destination has definitely been created by then.
12. shouldDecodeSourceDataOfMIMEType is gone. It's actually not called since we adopted NSURLSession, and we should remove the related dead code.
13. originatingFrame is gone. On _WKDownload it was actually a non-null WKFrameInfo that often contained no information. When information was there,
the information access has been replaced by didBecomeDownload which links the WKNavigationResponse which has frame info on it.
I wrote API tests for everything I could think of.
* NetworkProcess/Downloads/DownloadManager.cpp:
(WebKit::DownloadManager::resumeDownload):
* NetworkProcess/Downloads/DownloadManager.h:
* NetworkProcess/Downloads/cocoa/DownloadCocoa.mm:
(WebKit::Download::resume):
(WebKit::Download::platformCancelNetworkLoad):
* NetworkProcess/NetworkDataTaskBlob.cpp:
(WebKit::NetworkDataTaskBlob::suggestedFilename const):
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::resumeDownload):
(WebKit::NetworkProcess::findPendingDownloadLocation):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(-[WKNetworkSessionDelegate URLSession:task:didCompleteWithError:]):
* Scripts/webkit/messages.py:
* Shared/API/Cocoa/WebKit.h:
* Shared/API/c/WKSharedAPICast.h:
(WebKit::toAPI):
* Shared/Cocoa/APIObject.mm:
(API::Object::newObject):
* Sources.txt:
* SourcesCocoa.txt:
* UIProcess/API/APIDownloadClient.h:
(API::DownloadClient::legacyDidStart):
(API::DownloadClient::didFail):
(API::DownloadClient::didStart): Deleted.
* UIProcess/API/APINavigationAction.h:
* UIProcess/API/APINavigationClient.h:
(API::NavigationClient::navigationResponseDidBecomeDownload):
(API::NavigationClient::navigationActionDidBecomeDownload):
(API::NavigationClient::contextMenuDidCreateDownload):
* UIProcess/API/APIPolicyClient.h:
(API::PolicyClient::decidePolicyForNavigationAction):
(API::PolicyClient::decidePolicyForNewWindowAction):
* UIProcess/API/C/WKContext.cpp:
(WKContextSetDownloadClient): Deleted.
* UIProcess/API/C/WKContext.h:
* UIProcess/API/C/WKContextDownloadClient.h: Removed.
* UIProcess/API/C/WKDownload.cpp: Removed.
* UIProcess/API/C/WKDownloadClient.h: Added.
* UIProcess/API/C/WKDownloadRef.cpp: Added.
(WKDownloadGetTypeID):
(WKDownloadCopyRequest):
(WKDownloadCancel):
(WKDownloadGetOriginatingPage):
(WKDownloadGetWasUserInitiated):
(WKDownloadSetClient):
* UIProcess/API/C/WKDownloadRef.h: Renamed from Source/WebKit/UIProcess/API/C/WKDownload.h.
* UIProcess/API/C/WKPage.cpp:
(WKPageSetPagePolicyClient):
(WKPageSetPageNavigationClient):
* UIProcess/API/C/WKPageNavigationClient.h:
* UIProcess/API/Cocoa/WKDownload.h: Copied from Source/WebKit/UIProcess/API/Cocoa/_WKDownload.h.
* UIProcess/API/Cocoa/WKDownload.mm: Added.
(-[WKDownload cancel:]):
(-[WKDownload originalRequest]):
(-[WKDownload delegate]):
(-[WKDownload setDelegate:]):
(-[WKDownload dealloc]):
(-[WKDownload _apiObject]):
* UIProcess/API/Cocoa/WKDownloadDelegate.h: Added.
* UIProcess/API/Cocoa/WKDownloadInternal.h: Copied from Source/WebKit/UIProcess/API/Cocoa/_WKDownloadInternal.h.
* UIProcess/API/Cocoa/WKNavigationAction.h:
* UIProcess/API/Cocoa/WKNavigationAction.mm:
(-[WKNavigationAction downloadAttribute]):
(-[WKNavigationAction _shouldPerformDownload]):
* UIProcess/API/Cocoa/WKNavigationActionPrivate.h:
* UIProcess/API/Cocoa/WKNavigationDelegate.h:
* UIProcess/API/Cocoa/WKNavigationDelegatePrivate.h:
* UIProcess/API/Cocoa/WKProcessPool.mm:
(-[WKProcessPool _downloadURLRequest:websiteDataStore:originatingWebView:]):
(-[WKProcessPool _resumeDownloadFromData:websiteDataStore:path:originatingWebView:]):
* UIProcess/API/Cocoa/WKProcessPoolPrivate.h:
* UIProcess/API/Cocoa/WKWebView.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView downloadRequest:completionHandler:]):
(-[WKWebView resumeDownloadWithData:completionHandler:]):
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/API/Cocoa/_WKDownload.h:
* UIProcess/API/Cocoa/_WKDownload.mm:
(-[_WKDownload initWithDownload2:]):
(+[_WKDownload downloadWithDownload:]):
(-[_WKDownload cancel]):
(-[_WKDownload publishProgressAtURL:]):
(-[_WKDownload request]):
(-[_WKDownload originatingWebView]):
(-[_WKDownload redirectChain]):
(-[_WKDownload wasUserInitiated]):
(-[_WKDownload resumeData]):
(-[_WKDownload originatingFrame]):
(-[_WKDownload _apiObject]):
(-[_WKDownload dealloc]): Deleted.
* UIProcess/API/Cocoa/_WKDownloadDelegate.h:
* UIProcess/API/Cocoa/_WKDownloadInternal.h:
* UIProcess/API/glib/WebKitDownloadClient.cpp:
* UIProcess/Cocoa/LegacyDownloadClient.h:
* UIProcess/Cocoa/LegacyDownloadClient.mm:
(WebKit::LegacyDownloadClient::legacyDidStart):
(WebKit::LegacyDownloadClient::didReceiveResponse):
(WebKit::LegacyDownloadClient::didReceiveData):
(WebKit::LegacyDownloadClient::didReceiveAuthenticationChallenge):
(WebKit::LegacyDownloadClient::didCreateDestination):
(WebKit::LegacyDownloadClient::processDidCrash):
(WebKit::LegacyDownloadClient::decideDestinationWithSuggestedFilename):
(WebKit::LegacyDownloadClient::didFinish):
(WebKit::LegacyDownloadClient::didFail):
(WebKit::LegacyDownloadClient::legacyDidCancel):
(WebKit::LegacyDownloadClient::willSendRequest):
(WebKit::LegacyDownloadClient::didStart): Deleted.
* UIProcess/Cocoa/NavigationState.h:
* UIProcess/Cocoa/NavigationState.mm:
(WebKit::NavigationState::setNavigationDelegate):
(WebKit::NavigationState::NavigationClient::decidePolicyForNavigationAction):
(WebKit::NavigationState::NavigationClient::decidePolicyForNavigationResponse):
(WebKit::NavigationState::NavigationClient::navigationActionDidBecomeDownload):
(WebKit::NavigationState::NavigationClient::navigationResponseDidBecomeDownload):
(WebKit::NavigationState::NavigationClient::contextMenuDidCreateDownload):
* UIProcess/Cocoa/WebProcessProxyCocoa.mm:
* UIProcess/Downloads/DownloadProxy.cpp:
(WebKit::DownloadProxy::~DownloadProxy):
(WebKit::DownloadProxy::didStart):
(WebKit::DownloadProxy::decideDestinationWithSuggestedFilename):
(WebKit::DownloadProxy::didFail):
(WebKit::DownloadProxy::setClient):
* UIProcess/Downloads/DownloadProxy.h:
(WebKit::DownloadProxy::setDidStartCallback):
(WebKit::DownloadProxy::setSuggestedFilename):
* UIProcess/Network/NetworkProcessProxy.cpp:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::receivedNavigationPolicyDecision):
(WebKit::WebPageProxy::receivedPolicyDecision):
(WebKit::WebPageProxy::decidePolicyForNavigationAction):
(WebKit::WebPageProxy::decidePolicyForNewWindowAction):
(WebKit::WebPageProxy::decidePolicyForResponseShared):
(WebKit::WebPageProxy::resumeDownload):
(WebKit::WebPageProxy::downloadRequest):
(WebKit::WebPageProxy::contextMenuItemSelected):
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::receivedPolicyDecision):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::resumeDownload):
* UIProcess/WebProcessPool.h:
* WebKit.xcodeproj/project.pbxproj:
2020-12-10 Don Olmstead <don.olmstead@sony.com>
[CMake] Use TARGET_PROPERTY to set includes for executables
https://bugs.webkit.org/show_bug.cgi?id=219743
Reviewed by Michael Catanzaro.
Use $<TARGET_PROPERTY:WebKit,INCLUDE_DIRECTORIES> for all executables being
built alongside WebKit. This simplifies the includes for those targets.
* CMakeLists.txt:
2020-12-10 Chris Dumez <cdumez@apple.com>
Unreviewed, reverting r270542.
Seems to have regressed perf benchmarks
Reverted changeset:
"[iOS] Stop leaking an XPC boost message to XPC services"
https://bugs.webkit.org/show_bug.cgi?id=219453
https://trac.webkit.org/changeset/270542
2020-12-10 Youenn Fablet <youenn@apple.com>
ICE does not resolve for `turns` relay candidates rooted in LetsEncrypt CA
https://bugs.webkit.org/show_bug.cgi?id=219274
Reviewed by Eric Carlson.
Use NWConnection for client TCP sockets.
This allows piggy-backing on this API for things such as proxies or certificate validation.
Implementation does opening a TCP socket with TLS using NWConnection:
- If the socket is created for STUN/TURN, it is sending packets as is. For receiving data, the socket
will make sure to split received data in meaningful messages that WebProcess will process.
- If the socket is TCP for data, sent data is prefixed with the size of the data. For received data,
the length of the message is read and we wait to receive the whole message data to send it to WebProcess.
We enable that new code path based on an experimental feature.
Tested manually with TURNS servers and TCP-only connections.
* NetworkProcess/webrtc/NetworkRTCProvider.cpp:
(WebKit::NetworkRTCProvider::createClientTCPSocket):
* NetworkProcess/webrtc/NetworkRTCProvider.h:
(WebKit::NetworkRTCProvider::setPlatformSocketsEnabled):
* NetworkProcess/webrtc/NetworkRTCProvider.messages.in:
* NetworkProcess/webrtc/NetworkRTCSocketSocketCocoa.h: Added.
* NetworkProcess/webrtc/NetworkRTCSocketSocketCocoa.mm: Added.
(WebKit::socketQueue):
(WebKit::NetworkRTCSocketSocketCocoa::createClientTCPSocket):
(WebKit::isStunMessage):
(WebKit::getSTUNOrTURNMessageLength):
(WebKit::extractSTUNOrTURNMessages):
(WebKit::extractMessages):
(WebKit::processMessage):
(WebKit::NetworkRTCSocketSocketCocoa::NetworkRTCSocketSocketCocoa):
(WebKit::NetworkRTCSocketSocketCocoa::close):
(WebKit::NetworkRTCSocketSocketCocoa::setOption):
(WebKit::dataFromVector):
(WebKit::NetworkRTCSocketSocketCocoa::sendTo):
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/Network/webrtc/LibWebRTCSocketFactory.cpp:
(WebKit::LibWebRTCSocketFactory::setConnection):
2020-12-10 Antoine Quint <graouts@webkit.org>
Unified WebKit build fails
https://bugs.webkit.org/show_bug.cgi?id=219734
Reviewed by Tim Horton.
* UIProcess/RemoteLayerTree/mac/ScrollingTreeFrameScrollingNodeRemoteMac.h:
* UIProcess/SpeechRecognitionRemoteRealtimeMediaSource.h:
* UIProcess/SpeechRecognitionRemoteRealtimeMediaSourceManager.cpp:
2020-12-10 Jiewen Tan <jiewen_tan@apple.com>
Unreviewed, build fix after r270616
Adds a platform have flag.
* Platform/spi/Cocoa/AuthenticationServicesCoreSPI.h:
* UIProcess/WebAuthentication/Cocoa/AuthenticationServicesCoreSoftLink.h:
* UIProcess/WebAuthentication/Cocoa/AuthenticationServicesCoreSoftLink.mm:
* UIProcess/WebAuthentication/Cocoa/AuthenticatorPresenterCoordinator.mm:
(WebKit::AuthenticatorPresenterCoordinator::AuthenticatorPresenterCoordinator):
* UIProcess/WebAuthentication/Cocoa/WKASCAuthorizationPresenterDelegate.h:
* UIProcess/WebAuthentication/Cocoa/WKASCAuthorizationPresenterDelegate.mm:
2020-12-10 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Adopt new UI for the Security Key makeCredential flow
https://bugs.webkit.org/show_bug.cgi?id=219708
<rdar://problem/72154735>
Reviewed by Brent Fulgham.
This patch adopts the new AuthenticationServices' UI. The idea is to use AuthenticatorPresenterCoordinator
as the middle layer to connect to the new UI while keeping the existing interfaces the same in the AuthenticatorManager
to handle the interactions between UI and authenticators.
Covered by manual tests.
* Platform/spi/Cocoa/AuthenticationServicesCoreSPI.h: Added.
SPI for the new UI.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebAuthn.sb:
* Scripts/process-entitlements.sh:
Added new sandbox rules for Remote SpringBoard services which are used by the new UI.
* SourcesCocoa.txt:
Paperwork.
* UIProcess/WebAuthentication/AuthenticatorManager.cpp:
(WebKit::AuthenticatorManager::handleRequest):
(WebKit::AuthenticatorManager::cancel):
(WebKit::AuthenticatorManager::enableModernWebAuthentication):
(WebKit::AuthenticatorManager::clearState):
(WebKit::AuthenticatorManager::serviceStatusUpdated):
(WebKit::AuthenticatorManager::authenticatorStatusUpdated):
(WebKit::AuthenticatorManager::requestPin):
(WebKit::AuthenticatorManager::selectAssertionResponse):
(WebKit::AuthenticatorManager::startDiscovery):
(WebKit::AuthenticatorManager::runPanel):
(WebKit::AuthenticatorManager::runPresenter):
(WebKit::AuthenticatorManager::invokePendingCompletionHandler):
* UIProcess/WebAuthentication/AuthenticatorManager.h:
Teached AuthenticatorManager to use the new UI's presenter when it is available. Otherwise, it falls back to the old UI.
* UIProcess/WebAuthentication/Cocoa/AuthenticationServicesCoreSoftLink.h: Added.
* UIProcess/WebAuthentication/Cocoa/AuthenticationServicesCoreSoftLink.mm: Added.
* UIProcess/WebAuthentication/Cocoa/AuthenticatorPresenterCoordinator.h: Added.
* UIProcess/WebAuthentication/Cocoa/AuthenticatorPresenterCoordinator.mm: Added.
(WebKit::AuthenticatorPresenterCoordinator::AuthenticatorPresenterCoordinator):
(WebKit::AuthenticatorPresenterCoordinator::updatePresenter):
(WebKit::AuthenticatorPresenterCoordinator::requestPin):
(WebKit::AuthenticatorPresenterCoordinator::selectAssertionResponse):
(WebKit::AuthenticatorPresenterCoordinator::dimissPresenter):
* UIProcess/WebAuthentication/Cocoa/WKASCAuthorizationPresenterDelegate.h: Added.
* UIProcess/WebAuthentication/Cocoa/WKASCAuthorizationPresenterDelegate.mm: Added.
(-[WKASCAuthorizationPresenterDelegate initWithCoordinator:]):
(-[WKASCAuthorizationPresenterDelegate authorizationPresenter:credentialRequestedForLoginChoice:authenticatedContext:completionHandler:]):
(-[WKASCAuthorizationPresenterDelegate authorizationPresenter:validateUserEnteredPIN:completionHandler:]):
Introduced the AuthenticatorPresenterCoordinator to handle the new UI.
* WebAuthnProcess/WebAuthnProcess.cpp:
(WebKit::WebAuthnProcess::WebAuthnProcess):
(WebKit::WebAuthnProcess::setMockWebAuthenticationConfiguration):
Teached WebAuthnProcess to inform AuthenticatorManager about its existence.
* WebKit.xcodeproj/project.pbxproj:
Paperwork.
2020-12-09 Per Arne Vollan <pvollan@apple.com>
[macOS] Remove com.apple.PowerManagement.control from WebContent sandbox
https://bugs.webkit.org/show_bug.cgi?id=219685
<rdar://problem/70497033>
Reviewed by Darin Adler.
Remove com.apple.PowerManagement.control from WebContent sandbox on macOS, since there are no reports of this being in use.
No new tests, since the result will depend on macOS version.
* WebProcess/com.apple.WebProcess.sb.in:
2020-12-09 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process][MSE] Implement basic video playback
https://bugs.webkit.org/show_bug.cgi?id=219402
Reviewed by Jer Noble.
Implement the necessary functions with IPC to support media playback.
* GPUProcess/media/RemoteAudioTrackProxy.cpp:
(WebKit::RemoteAudioTrackProxy::configuration): Rename `identifier` to `trackId`
because it is easy to confuse with `TrackPrivateRemoteIdentifier`.
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
(WebKit::RemoteMediaPlayerProxy::loadMediaSource):
(WebKit::RemoteMediaPlayerProxy::updateCachedState): For MSE media player, `SourceBuffer`
provides the buffered ranges, so `RemoteMediaPlayerProxy` does not need to report this
information to the `MediaPlayer` object in the Web process.
* GPUProcess/media/RemoteMediaPlayerProxy.h:
* GPUProcess/media/RemoteMediaPlayerProxy.messages.in:
* GPUProcess/media/RemoteMediaSourceProxy.cpp:
(WebKit::RemoteMediaSourceProxy::duration const):
(WebKit::RemoteMediaSourceProxy::buffered const):
(WebKit::RemoteMediaSourceProxy::durationChanged):
(WebKit::RemoteMediaSourceProxy::setReadyState):
* GPUProcess/media/RemoteMediaSourceProxy.h:
* GPUProcess/media/RemoteMediaSourceProxy.messages.in:
* GPUProcess/media/RemoteSourceBufferProxy.cpp:
(WebKit::RemoteSourceBufferProxy::RemoteSourceBufferProxy):
(WebKit::RemoteSourceBufferProxy::~RemoteSourceBufferProxy):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateDidReceiveInitializationSegment):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateDurationChanged):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateDidParseSample):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateAppendComplete):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateDidReceiveRenderingError):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateBufferedDirtyChanged):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateBufferedRangesChanged):
(WebKit::RemoteSourceBufferProxy::setActive):
(WebKit::RemoteSourceBufferProxy::setReadyState):
(WebKit::RemoteSourceBufferProxy::updateBufferedFromTrackBuffers):
(WebKit::RemoteSourceBufferProxy::evictCodedFrames):
(WebKit::RemoteSourceBufferProxy::addTrackBuffer):
(WebKit::RemoteSourceBufferProxy::reenqueueMediaIfNeeded):
(WebKit::RemoteSourceBufferProxy::trySignalAllSamplesInTrackEnqueued):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateHasAudio const): Deleted.
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateHasVideo const): Deleted.
* GPUProcess/media/RemoteSourceBufferProxy.h:
* GPUProcess/media/RemoteSourceBufferProxy.messages.in:
* GPUProcess/media/RemoteTextTrackProxy.cpp:
(WebKit::RemoteTextTrackProxy::configuration):
* GPUProcess/media/RemoteVideoTrackProxy.cpp:
(WebKit::RemoteVideoTrackProxy::configuration):
* GPUProcess/media/TextTrackPrivateRemoteConfiguration.h:
(WebKit::TextTrackPrivateRemoteConfiguration::encode const):
(WebKit::TextTrackPrivateRemoteConfiguration::decode):
* GPUProcess/media/TrackPrivateRemoteConfiguration.h:
(WebKit::TrackPrivateRemoteConfiguration::encode const):
(WebKit::TrackPrivateRemoteConfiguration::decode):
* WebProcess/GPU/media/AudioTrackPrivateRemote.cpp:
(WebKit::AudioTrackPrivateRemote::updateConfiguration):
* WebProcess/GPU/media/AudioTrackPrivateRemote.h:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::load):
* WebProcess/GPU/media/MediaSourcePrivateRemote.cpp:
(WebKit::MediaSourcePrivateRemote::durationChanged):
(WebKit::MediaSourcePrivateRemote::isEnded const):
(WebKit::MediaSourcePrivateRemote::setReadyState):
* WebProcess/GPU/media/MediaSourcePrivateRemote.h:
* WebProcess/GPU/media/SourceBufferPrivateRemote.cpp:
(WebKit::SourceBufferPrivateRemote::setReadyState):
(WebKit::SourceBufferPrivateRemote::flush):
(WebKit::SourceBufferPrivateRemote::setActive):
(WebKit::SourceBufferPrivateRemote::updateBufferedFromTrackBuffers):
(WebKit::SourceBufferPrivateRemote::evictCodedFrames):
(WebKit::SourceBufferPrivateRemote::addTrackBuffer):
(WebKit::SourceBufferPrivateRemote::reenqueueMediaIfNeeded):
(WebKit::SourceBufferPrivateRemote::trySignalAllSamplesInTrackEnqueued):
(WebKit::SourceBufferPrivateRemote::sourceBufferPrivateDidReceiveInitializationSegment):
(WebKit::SourceBufferPrivateRemote::sourceBufferPrivateDurationChanged):
(WebKit::SourceBufferPrivateRemote::sourceBufferPrivateDidParseSample):
(WebKit::SourceBufferPrivateRemote::sourceBufferPrivateBufferedDirtyChanged):
(WebKit::SourceBufferPrivateRemote::sourceBufferPrivateBufferedRangesChanged):
(WebKit::SourceBufferPrivateRemote::enqueueSample): Deleted.
* WebProcess/GPU/media/SourceBufferPrivateRemote.h:
* WebProcess/GPU/media/SourceBufferPrivateRemote.messages.in:
* WebProcess/GPU/media/TextTrackPrivateRemote.cpp:
(WebKit::TextTrackPrivateRemote::updateConfiguration):
* WebProcess/GPU/media/VideoTrackPrivateRemote.cpp:
(WebKit::VideoTrackPrivateRemote::updateConfiguration):
2020-11-30 Brian Burg <bburg@apple.com>
[Cocoa] Web Inspector: add support for creating extension tabs in WebInspectorUI via _WKInspectorExtension
https://bugs.webkit.org/show_bug.cgi?id=219380
Reviewed by Devin Rousso and Timothy Hatcher.
Add a new method to _WKInspectorExtension for creating an extension tab in WebInpectorUI.
This can be used to implement browser.devtools.panels.create() as provided by the Web Extensions API.
* Platform/Logging.h: Add Inspector channel for error logging.
* Shared/InspectorExtensionTypes.h: Add a missing EnumTraits case.
* WebKit.xcodeproj/project.pbxproj:
* Sources.txt: Add files.
* UIProcess/API/APIInspectorExtension.h:
* UIProcess/API/APIInspectorExtension.cpp:
(API::InspectorExtension::InspectorExtension):
(API::InspectorExtension::create):
(API::InspectorExtension::createTab):
Inspector extensions need to be able to invoke commands in the WebProcess-side
extension proxy controller, so keep a WeakRef to the extension controller that
created the extension object. Also, implement the new API by forwarding it on.
* UIProcess/API/Cocoa/_WKInspector.mm:
(-[_WKInspector registerExtensionWithID:displayName:completionHandler:]):
Adapt to the new constructor as explained above.
* UIProcess/API/Cocoa/_WKInspectorExtensionInternal.h:
Expose member m_remoteInspectorPorxy for use in the API::InspectorExtension constructor.
* UIProcess/API/Cocoa/_WKInspectorExtension.h:
* UIProcess/API/Cocoa/_WKInspectorExtension.mm:
(-[_WKInspectorExtension initWithIdentifier:]): Deleted.
(-[_WKInspectorExtension createTabWithName:tabIconURL:sourceURL:completionHandler:]):
Implement new API.
* UIProcess/API/Cocoa/_WKRemoteWebInspectorViewControllerInternal.h: Added.
Make the RefPtr<RemoteWebInspectorProxy> member variable accessible to the API object constructor.
* UIProcess/API/Cocoa/_WKRemoteWebInspectorViewController.mm:
(-[_WKRemoteWebInspectorViewController registerExtensionWithID:displayName:completionHandler:]):
Adapt to new constructor as explained above.
* UIProcess/Inspector/WebInspectorUIExtensionControllerProxy.h:
* UIProcess/Inspector/WebInspectorUIExtensionControllerProxy.cpp:
(WebKit::WebInspectorUIExtensionControllerProxy::createTabForExtension):
Forward the request to the inspector frontend WebProcess.
* WebProcess/Inspector/WebInspectorUIExtensionController.messages.in:
Add a new async IPC command for creating a new inspector extension tab.
* WebProcess/Inspector/WebInspectorUIExtensionController.h:
* WebProcess/Inspector/WebInspectorUIExtensionController.cpp:
(WebKit::WebInspectorUIExtensionController::unwrapEvaluationResultAsObject):
Added. Pull out the code needed for error-handling this common operation.
(WebKit::WebInspectorUIExtensionController::createTabForExtension):
Added. Call into InspectorFrontendAPI to request a new tab. Turn the result
into an InspectorExtensionError or InspectorExtensionTabID and send the IPC reply.
* UIProcess/RemoteLayerTree/mac/ScrollingTreeOverflowScrollingNodeRemoteMac.h:
Fix unified sources fallout.
2020-12-09 Said Abou-Hallawa <said@apple.com>
[GPU Process]: Recording an in-process ImageBuffer drawing has to convert it to a NativeImage first
https://bugs.webkit.org/show_bug.cgi?id=219705
Reviewed by Wenson Hsieh.
Override the virtual method isCachedImageBuffer() which in this case
answers the question: is this a remote ImageBuffer or not?
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
2020-12-09 Per Arne Vollan <pvollan@apple.com>
[macOS] Allow mach-lookup of com.apple.relatived.tempest in WebKit GPU process
https://bugs.webkit.org/show_bug.cgi?id=219701
<rdar://problem/71980348>
Reviewed by Darin Adler.
* GPUProcess/mac/com.apple.WebKit.GPUProcess.sb.in:
2020-12-09 Wenson Hsieh <wenson_hsieh@apple.com>
Some WebGL tests crash in WebKit::RemoteRenderingBackendProxy::willAppendItem(WebCore::RenderingResourceIdentifier)
https://bugs.webkit.org/show_bug.cgi?id=219700
Reviewed by Tim Horton.
Fixes these assertions by resetting the identifier of the "most recently appended" image buffer
(m_currentDestinationImageBufferIdentifier) when the image buffer is released in the rendering
backend. We keep track of this so that we can append MetaCommandChangeDestinationImageBuffer
items when painting into multiple image buffers; the code to do this in willAppendItem is
actually robust against the case where m_currentDestinationImageBufferIdentifier is unknown,
but still debug asserts.
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::releaseRemoteResource):
2020-12-09 Alex Christensen <achristensen@webkit.org>
Fix clicking on links in PDFs on Big Sur after r266654
https://bugs.webkit.org/show_bug.cgi?id=219665
<rdar://problem/72080683>
Reviewed by Tim Horton.
mouseUp: was being called by AppKit on the WKPDFHUDView instead of the WKWebView, causing the event to not make it to the PDFPlugin,
which meant PDFKit wasn't being told it was clicked on, which means it wasn't telling WebKit to change the frame's location.
Overriding hitTest to forward all mouse events to the WKWebView solves this problem.
Covered by the WebKit.PDFLinkReferrer API test, which was failing on Big Sur, where ENABLE(UI_PROCESS_PDF_HUD) is true.
* UIProcess/PDF/WKPDFHUDView.mm:
(-[WKPDFHUDView hitTest:]):
2020-12-09 Wenson Hsieh <wenson_hsieh@apple.com>
[Concurrent display lists] Support playback of display list items with cached resources
https://bugs.webkit.org/show_bug.cgi?id=218614
<rdar://problem/71326662>
Reviewed by Tim Horton.
Teaches the GPU process to resume playback of display list items that were previously unable to be played back
due to missing cached resources. See below for more details.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::createImageBuffer):
See if the newly created image buffer's ID matches that of the pending wakeup message; if so, immediately kick
off the wakeup loop.
(WebKit::RemoteRenderingBackend::nextDestinationImageBufferAfterApplyingDisplayLists):
(WebKit::RemoteRenderingBackend::wakeUpAndApplyDisplayList):
In the case where the next destination image buffer is unknown, bail early and resume when we eventually learn
about the image buffer in the GPU process. This can happen if a MetaCommandChangeDestinationImageBuffer item is
added that references a newly created image buffer.
(WebKit::RemoteRenderingBackend::setNextItemBufferToRead):
(WebKit::RemoteRenderingBackend::cacheNativeImage):
See if the newly cached image's ID matches that of the pending wakeup message; if so, immediately kick off the
wakeup loop.
(WebKit::RemoteRenderingBackend::didCreateSharedDisplayListHandle):
* GPUProcess/graphics/RemoteRenderingBackend.h:
Add a private `PendingWakeupInformation` struct that encapsulates all the information needed to remember that
we stopped display list processing, and resume processing in the future. Currently, this contains a set of
wakeup message arguments, and (optionally) the identifier of the missing resource that we need to receive in
order to continue.
We also use this in place of storing GPUProcessWakeupMessageArguments on RemoteRenderingBackend.
(WebKit::RemoteRenderingBackend::PendingWakeupInformation::shouldPerformWakeup const):
2020-12-09 Kimmo Kinnunen <kkinnunen@apple.com>
Implement RemoteWebGLBackend
https://bugs.webkit.org/show_bug.cgi?id=217216
<rdar://problem/69876258>
Reviewed by Simon Fraser.
Implements a beginning of a remote version of GraphicsContextGL for GPU
process. Current implementation is using WebKit IPC.
Implemented with a generator script implementing the IPC to pass the
GraphicsContextGL member function calls from web process to the GPU
process. The script is run during development time and the results are
committed to the repo for ease of review.
Implement the RemoteGraphicsContextGL in WebKit side. This class will
forward calls received via IPC to the GraphicsContextGLOpenGL. At the
moment RemoteGraphicsContextGL instance is a direct child of
GPUConnectionToWebProcess.
Compositing is implemented only for PLATFORM(COCOA).
Missing:
- Simulated context loss
- Simulated context change
- Paint WebGL contents to 2D canvas
- Upload video to WebGL
- Upload canvas to WebGL
* CMakeLists.txt:
* DerivedSources-input.xcfilelist:
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::createGraphicsContextGL):
(WebKit::GPUConnectionToWebProcess::releaseGraphicsContextGL):
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/GPUConnectionToWebProcess.messages.in:
* GPUProcess/graphics/RemoteGraphicsContextGL.cpp: Added.
(WebKit::RemoteGraphicsContextGL::create):
(WebKit::RemoteGraphicsContextGL::RemoteGraphicsContextGL):
(WebKit::RemoteGraphicsContextGL::~RemoteGraphicsContextGL):
(WebKit::RemoteGraphicsContextGL::gpuConnectionToWebProcess const):
(WebKit::RemoteGraphicsContextGL::messageSenderConnection const):
(WebKit::RemoteGraphicsContextGL::messageSenderDestinationID const):
(WebKit::RemoteGraphicsContextGL::didComposite):
(WebKit::RemoteGraphicsContextGL::forceContextLost):
(WebKit::RemoteGraphicsContextGL::recycleContext):
(WebKit::RemoteGraphicsContextGL::dispatchContextChangedNotification):
(WebKit::RemoteGraphicsContextGL::recycleBuffer):
(WebKit::RemoteGraphicsContextGL::present):
(WebKit::RemoteGraphicsContextGL::detachClient):
(WebKit::RemoteGraphicsContextGL::reshape):
(WebKit::RemoteGraphicsContextGL::prepareForDisplay):
(WebKit::RemoteGraphicsContextGL::ensureExtensionEnabled):
(WebKit::RemoteGraphicsContextGL::notifyMarkContextChanged):
* GPUProcess/graphics/RemoteGraphicsContextGL.h: Added.
* GPUProcess/graphics/RemoteGraphicsContextGL.messages.in: Added.
* GPUProcess/graphics/RemoteGraphicsContextGLFunctionsGenerated.h: Added.
(setFailNextGPUStatusCheck):
(synthesizeGLError):
(moveErrorsToSyntheticErrorList):
(activeTexture):
(attachShader):
(bindAttribLocation):
(bindBuffer):
(bindFramebuffer):
(bindRenderbuffer):
(bindTexture):
(blendColor):
(blendEquation):
(blendEquationSeparate):
(blendFunc):
(blendFuncSeparate):
(checkFramebufferStatus):
(clear):
(clearColor):
(clearDepth):
(clearStencil):
(colorMask):
(compileShader):
(copyTexImage2D):
(copyTexSubImage2D):
(createBuffer):
(createFramebuffer):
(createProgram):
(createRenderbuffer):
(createShader):
(createTexture):
(cullFace):
(deleteBuffer):
(deleteFramebuffer):
(deleteProgram):
(deleteRenderbuffer):
(deleteShader):
(deleteTexture):
(depthFunc):
(depthMask):
(depthRange):
(detachShader):
(disable):
(disableVertexAttribArray):
(drawArrays):
(drawElements):
(enable):
(enableVertexAttribArray):
(finish):
(flush):
(framebufferRenderbuffer):
(framebufferTexture2D):
(frontFace):
(generateMipmap):
(getActiveAttrib):
(getActiveUniform):
(getAttribLocation):
(getBufferParameteri):
(getString):
(getFloatv):
(getIntegerv):
(getInteger64):
(getInteger64i):
(getProgrami):
(getBooleanv):
(getError):
(getFramebufferAttachmentParameteri):
(getProgramInfoLog):
(getRenderbufferParameteri):
(getShaderi):
(getShaderInfoLog):
(getShaderPrecisionFormat):
(getShaderSource):
(getTexParameterf):
(getTexParameteri):
(getUniformfv):
(getUniformiv):
(getUniformuiv):
(getUniformLocation):
(getVertexAttribOffset):
(hint):
(isBuffer):
(isEnabled):
(isFramebuffer):
(isProgram):
(isRenderbuffer):
(isShader):
(isTexture):
(lineWidth):
(linkProgram):
(pixelStorei):
(polygonOffset):
(renderbufferStorage):
(sampleCoverage):
(scissor):
(shaderSource):
(stencilFunc):
(stencilFuncSeparate):
(stencilMask):
(stencilMaskSeparate):
(stencilOp):
(stencilOpSeparate):
(texParameterf):
(texParameteri):
(uniform1f):
(uniform1fv):
(uniform1i):
(uniform1iv):
(uniform2f):
(uniform2fv):
(uniform2i):
(uniform2iv):
(uniform3f):
(uniform3fv):
(uniform3i):
(uniform3iv):
(uniform4f):
(uniform4fv):
(uniform4i):
(uniform4iv):
(uniformMatrix2fv):
(uniformMatrix3fv):
(uniformMatrix4fv):
(useProgram):
(validateProgram):
(vertexAttrib1f):
(vertexAttrib1fv):
(vertexAttrib2f):
(vertexAttrib2fv):
(vertexAttrib3f):
(vertexAttrib3fv):
(vertexAttrib4f):
(vertexAttrib4fv):
(vertexAttribPointer):
(viewport):
(bufferData0):
(bufferData1):
(bufferSubData):
(readnPixels0):
(readnPixels1):
(texImage2D0):
(texImage2D1):
(texSubImage2D0):
(texSubImage2D1):
(compressedTexImage2D0):
(compressedTexImage2D1):
(compressedTexSubImage2D0):
(compressedTexSubImage2D1):
(drawArraysInstanced):
(drawElementsInstanced):
(vertexAttribDivisor):
(createVertexArray):
(deleteVertexArray):
(isVertexArray):
(bindVertexArray):
(copyBufferSubData):
(getBufferSubData):
(blitFramebuffer):
(framebufferTextureLayer):
(invalidateFramebuffer):
(invalidateSubFramebuffer):
(readBuffer):
(renderbufferStorageMultisample):
(texStorage2D):
(texStorage3D):
(texImage3D0):
(texImage3D1):
(texSubImage3D0):
(texSubImage3D1):
(copyTexSubImage3D):
(compressedTexImage3D0):
(compressedTexImage3D1):
(compressedTexSubImage3D0):
(compressedTexSubImage3D1):
(getFragDataLocation):
(uniform1ui):
(uniform2ui):
(uniform3ui):
(uniform4ui):
(uniform1uiv):
(uniform2uiv):
(uniform3uiv):
(uniform4uiv):
(uniformMatrix2x3fv):
(uniformMatrix3x2fv):
(uniformMatrix2x4fv):
(uniformMatrix4x2fv):
(uniformMatrix3x4fv):
(uniformMatrix4x3fv):
(vertexAttribI4i):
(vertexAttribI4iv):
(vertexAttribI4ui):
(vertexAttribI4uiv):
(vertexAttribIPointer):
(drawRangeElements):
(drawBuffers):
(clearBufferiv):
(clearBufferuiv):
(clearBufferfv):
(clearBufferfi):
(createQuery):
(deleteQuery):
(isQuery):
(beginQuery):
(endQuery):
(getQuery):
(getQueryObjectui):
(createSampler):
(deleteSampler):
(isSampler):
(bindSampler):
(samplerParameteri):
(samplerParameterf):
(getSamplerParameterf):
(getSamplerParameteri):
(fenceSync):
(isSync):
(deleteSync):
(clientWaitSync):
(waitSync):
(getSynci):
(createTransformFeedback):
(deleteTransformFeedback):
(isTransformFeedback):
(bindTransformFeedback):
(beginTransformFeedback):
(endTransformFeedback):
(transformFeedbackVaryings):
(getTransformFeedbackVarying):
(pauseTransformFeedback):
(resumeTransformFeedback):
(bindBufferBase):
(bindBufferRange):
(getUniformIndices):
(getActiveUniforms):
(getUniformBlockIndex):
(getActiveUniformBlockName):
(uniformBlockBinding):
(getActiveUniformBlockiv):
(getGraphicsResetStatusARB):
(blitFramebufferANGLE):
(renderbufferStorageMultisampleANGLE):
(createVertexArrayOES):
(deleteVertexArrayOES):
(isVertexArrayOES):
(bindVertexArrayOES):
(getTranslatedShaderSourceANGLE):
(insertEventMarkerEXT):
(pushGroupMarkerEXT):
(popGroupMarkerEXT):
(drawBuffersEXT):
(drawArraysInstancedANGLE):
(drawElementsInstancedANGLE):
(vertexAttribDivisorANGLE):
* GPUProcess/mac/GPUProcessMac.mm:
(WebKit::GPUProcess::initializeProcess):
* GPUProcess/media/RemoteLegacyCDMProxy.h:
* GPUProcess/media/RemoteLegacyCDMSessionProxy.cpp:
* GPUProcess/media/RemoteMediaResourceManager.cpp:
* GPUProcess/media/RemoteMediaResourceManager.h:
* GPUProcess/media/RemoteMediaSourceProxy.h:
* GPUProcess/media/RemoteSourceBufferProxy.cpp:
* GPUProcess/media/RemoteSourceBufferProxy.h:
* GPUProcess/webrtc/LibWebRTCCodecsProxy.h:
* GPUProcess/webrtc/LibWebRTCCodecsProxy.mm:
* GPUProcess/webrtc/RemoteMediaRecorder.h:
* NetworkProcess/CustomProtocols/Cocoa/LegacyCustomProtocolManagerCocoa.mm:
* NetworkProcess/CustomProtocols/LegacyCustomProtocolManager.h:
* NetworkProcess/Downloads/Download.cpp:
* NetworkProcess/Downloads/Download.h:
* NetworkProcess/Downloads/DownloadManager.cpp:
* NetworkProcess/Downloads/DownloadManager.h:
* NetworkProcess/NetworkSocketChannel.cpp:
* NetworkProcess/NetworkSocketChannel.h:
* NetworkProcess/NetworkSocketStream.cpp:
* NetworkProcess/NetworkSocketStream.h:
* NetworkProcess/ServiceWorker/ServiceWorkerFetchTask.h:
* NetworkProcess/WebSocketTask.h:
* NetworkProcess/cocoa/WebSocketTaskCocoa.h:
* NetworkProcess/cocoa/WebSocketTaskCocoa.mm:
* NetworkProcess/soup/WebSocketTaskSoup.cpp:
* NetworkProcess/soup/WebSocketTaskSoup.h:
* Platform/IPC/ArgumentCoders.h:
* Platform/IPC/ArrayReference.h: Copied from Source/WebKit/Platform/IPC/DataReference.h.
(IPC::ArrayReference::ArrayReference):
(IPC::ArrayReference::isEmpty const):
(IPC::ArrayReference::size const):
(IPC::ArrayReference::data const):
(IPC::ArrayReference::vector const):
* Platform/IPC/Connection.h:
(IPC::Connection::sendSync):
* Platform/IPC/DataReference.h:
* Platform/IPC/Decoder.cpp:
(IPC::Decoder::decodeFixedLengthReference):
* Platform/IPC/Decoder.h:
* Platform/IPC/Encoder.cpp:
(IPC::Encoder::wrapForTesting):
* Platform/IPC/Encoder.h:
* Platform/IPC/HandleMessage.h:
* Platform/IPC/MessageSender.h:
(IPC::MessageSender::sendSync):
* Platform/IPC/SharedBufferCopy.cpp:
(IPC::SharedBufferCopy::decode):
* Platform/IPC/SharedBufferDataReference.cpp:
* Platform/IPC/StringReference.cpp:
* PluginProcess/PluginControllerProxy.cpp:
* PluginProcess/PluginControllerProxy.h:
* Scripts/webkit/messages.py:
* Shared/API/APIData.cpp:
* Shared/AuxiliaryProcess.h:
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<WebCore::GraphicsContextGLAttributes>::encode):
(IPC::ArgumentCoder<WebCore::GraphicsContextGLAttributes>::decode):
(IPC::ArgumentCoder<WebCore::GraphicsContextGL::ActiveInfo>::encode):
(IPC::ArgumentCoder<WebCore::GraphicsContextGL::ActiveInfo>::decode):
* Shared/WebCoreArgumentCoders.h:
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
* Shared/cf/ArgumentCodersCF.cpp:
* Shared/soup/WebCoreArgumentCodersSoup.cpp:
(IPC::ArgumentCoder<CertificateInfo>::encode):
(IPC::ArgumentCoder<CertificateInfo>::decode):
* Sources.txt:
* SourcesCocoa.txt:
* UIProcess/API/APIIconLoadingClient.h:
* UIProcess/AuxiliaryProcessProxy.h:
(WebKit::AuxiliaryProcessProxy::sendSync):
* UIProcess/Cocoa/SOAuthorization/SubFrameSOAuthorizationSession.h:
* UIProcess/Network/CustomProtocols/LegacyCustomProtocolManagerProxy.h:
* UIProcess/PageClient.h:
* UIProcess/UserContent/WebUserContentControllerProxy.cpp:
* UIProcess/UserContent/WebUserContentControllerProxy.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
* UIProcess/WebPageProxy.h:
* UIProcess/WebURLSchemeHandler.h:
* UIProcess/WebURLSchemeTask.h:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/ApplePay/WebPaymentCoordinator.h:
* WebProcess/GPU/GPUProcessConnection.h:
* WebProcess/GPU/graphics/GraphicsContextGLIdentifier.h: Renamed from Source/WebKit/Platform/IPC/DataReference.cpp.
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.cpp: Added.
(WebKit::RemoteGraphicsContextGLProxy::create):
(WebKit::RemoteGraphicsContextGLProxy::RemoteGraphicsContextGLProxy):
(WebKit::RemoteGraphicsContextGLProxy::~RemoteGraphicsContextGLProxy):
(WebKit::RemoteGraphicsContextGLProxy::messageSenderConnection const):
(WebKit::RemoteGraphicsContextGLProxy::messageSenderDestinationID const):
(WebKit::RemoteGraphicsContextGLProxy::reshape):
(WebKit::RemoteGraphicsContextGLProxy::prepareForDisplay):
(WebKit::RemoteGraphicsContextGLProxy::ensureExtensionEnabled):
(WebKit::RemoteGraphicsContextGLProxy::notifyMarkContextChanged):
(WebKit::RemoteGraphicsContextGLProxy::simulateContextChanged):
(WebKit::RemoteGraphicsContextGLProxy::paintRenderingResultsToCanvas):
(WebKit::RemoteGraphicsContextGLProxy::paintRenderingResultsToImageData):
(WebKit::RemoteGraphicsContextGLProxy::wasCreated):
(WebKit::RemoteGraphicsContextGLProxy::wasLost):
(WebKit::RemoteGraphicsContextGLProxy::wasChanged):
(WebKit::RemoteGraphicsContextGLProxy::waitUntilInitialized):
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.h: Added.
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxy.messages.in: Copied from Source/WebKit/GPUProcess/GPUConnectionToWebProcess.messages.in.
* WebProcess/GPU/graphics/RemoteGraphicsContextGLProxyFunctionsGenerated.cpp: Added.
(WebKit::RemoteGraphicsContextGLProxy::setFailNextGPUStatusCheck):
(WebKit::RemoteGraphicsContextGLProxy::synthesizeGLError):
(WebKit::RemoteGraphicsContextGLProxy::moveErrorsToSyntheticErrorList):
(WebKit::RemoteGraphicsContextGLProxy::activeTexture):
(WebKit::RemoteGraphicsContextGLProxy::attachShader):
(WebKit::RemoteGraphicsContextGLProxy::bindAttribLocation):
(WebKit::RemoteGraphicsContextGLProxy::bindBuffer):
(WebKit::RemoteGraphicsContextGLProxy::bindFramebuffer):
(WebKit::RemoteGraphicsContextGLProxy::bindRenderbuffer):
(WebKit::RemoteGraphicsContextGLProxy::bindTexture):
(WebKit::RemoteGraphicsContextGLProxy::blendColor):
(WebKit::RemoteGraphicsContextGLProxy::blendEquation):
(WebKit::RemoteGraphicsContextGLProxy::blendEquationSeparate):
(WebKit::RemoteGraphicsContextGLProxy::blendFunc):
(WebKit::RemoteGraphicsContextGLProxy::blendFuncSeparate):
(WebKit::RemoteGraphicsContextGLProxy::checkFramebufferStatus):
(WebKit::RemoteGraphicsContextGLProxy::clear):
(WebKit::RemoteGraphicsContextGLProxy::clearColor):
(WebKit::RemoteGraphicsContextGLProxy::clearDepth):
(WebKit::RemoteGraphicsContextGLProxy::clearStencil):
(WebKit::RemoteGraphicsContextGLProxy::colorMask):
(WebKit::RemoteGraphicsContextGLProxy::compileShader):
(WebKit::RemoteGraphicsContextGLProxy::copyTexImage2D):
(WebKit::RemoteGraphicsContextGLProxy::copyTexSubImage2D):
(WebKit::RemoteGraphicsContextGLProxy::createBuffer):
(WebKit::RemoteGraphicsContextGLProxy::createFramebuffer):
(WebKit::RemoteGraphicsContextGLProxy::createProgram):
(WebKit::RemoteGraphicsContextGLProxy::createRenderbuffer):
(WebKit::RemoteGraphicsContextGLProxy::createShader):
(WebKit::RemoteGraphicsContextGLProxy::createTexture):
(WebKit::RemoteGraphicsContextGLProxy::cullFace):
(WebKit::RemoteGraphicsContextGLProxy::deleteBuffer):
(WebKit::RemoteGraphicsContextGLProxy::deleteFramebuffer):
(WebKit::RemoteGraphicsContextGLProxy::deleteProgram):
(WebKit::RemoteGraphicsContextGLProxy::deleteRenderbuffer):
(WebKit::RemoteGraphicsContextGLProxy::deleteShader):
(WebKit::RemoteGraphicsContextGLProxy::deleteTexture):
(WebKit::RemoteGraphicsContextGLProxy::depthFunc):
(WebKit::RemoteGraphicsContextGLProxy::depthMask):
(WebKit::RemoteGraphicsContextGLProxy::depthRange):
(WebKit::RemoteGraphicsContextGLProxy::detachShader):
(WebKit::RemoteGraphicsContextGLProxy::disable):
(WebKit::RemoteGraphicsContextGLProxy::disableVertexAttribArray):
(WebKit::RemoteGraphicsContextGLProxy::drawArrays):
(WebKit::RemoteGraphicsContextGLProxy::drawElements):
(WebKit::RemoteGraphicsContextGLProxy::enable):
(WebKit::RemoteGraphicsContextGLProxy::enableVertexAttribArray):
(WebKit::RemoteGraphicsContextGLProxy::finish):
(WebKit::RemoteGraphicsContextGLProxy::flush):
(WebKit::RemoteGraphicsContextGLProxy::framebufferRenderbuffer):
(WebKit::RemoteGraphicsContextGLProxy::framebufferTexture2D):
(WebKit::RemoteGraphicsContextGLProxy::frontFace):
(WebKit::RemoteGraphicsContextGLProxy::generateMipmap):
(WebKit::RemoteGraphicsContextGLProxy::getActiveAttrib):
(WebKit::RemoteGraphicsContextGLProxy::getActiveUniform):
(WebKit::RemoteGraphicsContextGLProxy::getAttribLocation):
(WebKit::RemoteGraphicsContextGLProxy::getBufferParameteri):
(WebKit::RemoteGraphicsContextGLProxy::getString):
(WebKit::RemoteGraphicsContextGLProxy::getFloatv):
(WebKit::RemoteGraphicsContextGLProxy::getIntegerv):
(WebKit::RemoteGraphicsContextGLProxy::getInteger64):
(WebKit::RemoteGraphicsContextGLProxy::getInteger64i):
(WebKit::RemoteGraphicsContextGLProxy::getProgrami):
(WebKit::RemoteGraphicsContextGLProxy::getBooleanv):
(WebKit::RemoteGraphicsContextGLProxy::getError):
(WebKit::RemoteGraphicsContextGLProxy::getFramebufferAttachmentParameteri):
(WebKit::RemoteGraphicsContextGLProxy::getProgramInfoLog):
(WebKit::RemoteGraphicsContextGLProxy::getRenderbufferParameteri):
(WebKit::RemoteGraphicsContextGLProxy::getShaderi):
(WebKit::RemoteGraphicsContextGLProxy::getShaderInfoLog):
(WebKit::RemoteGraphicsContextGLProxy::getShaderPrecisionFormat):
(WebKit::RemoteGraphicsContextGLProxy::getShaderSource):
(WebKit::RemoteGraphicsContextGLProxy::getTexParameterf):
(WebKit::RemoteGraphicsContextGLProxy::getTexParameteri):
(WebKit::RemoteGraphicsContextGLProxy::getUniformfv):
(WebKit::RemoteGraphicsContextGLProxy::getUniformiv):
(WebKit::RemoteGraphicsContextGLProxy::getUniformuiv):
(WebKit::RemoteGraphicsContextGLProxy::getUniformLocation):
(WebKit::RemoteGraphicsContextGLProxy::getVertexAttribOffset):
(WebKit::RemoteGraphicsContextGLProxy::hint):
(WebKit::RemoteGraphicsContextGLProxy::isBuffer):
(WebKit::RemoteGraphicsContextGLProxy::isEnabled):
(WebKit::RemoteGraphicsContextGLProxy::isFramebuffer):
(WebKit::RemoteGraphicsContextGLProxy::isProgram):
(WebKit::RemoteGraphicsContextGLProxy::isRenderbuffer):
(WebKit::RemoteGraphicsContextGLProxy::isShader):
(WebKit::RemoteGraphicsContextGLProxy::isTexture):
(WebKit::RemoteGraphicsContextGLProxy::lineWidth):
(WebKit::RemoteGraphicsContextGLProxy::linkProgram):
(WebKit::RemoteGraphicsContextGLProxy::pixelStorei):
(WebKit::RemoteGraphicsContextGLProxy::polygonOffset):
(WebKit::RemoteGraphicsContextGLProxy::renderbufferStorage):
(WebKit::RemoteGraphicsContextGLProxy::sampleCoverage):
(WebKit::RemoteGraphicsContextGLProxy::scissor):
(WebKit::RemoteGraphicsContextGLProxy::shaderSource):
(WebKit::RemoteGraphicsContextGLProxy::stencilFunc):
(WebKit::RemoteGraphicsContextGLProxy::stencilFuncSeparate):
(WebKit::RemoteGraphicsContextGLProxy::stencilMask):
(WebKit::RemoteGraphicsContextGLProxy::stencilMaskSeparate):
(WebKit::RemoteGraphicsContextGLProxy::stencilOp):
(WebKit::RemoteGraphicsContextGLProxy::stencilOpSeparate):
(WebKit::RemoteGraphicsContextGLProxy::texParameterf):
(WebKit::RemoteGraphicsContextGLProxy::texParameteri):
(WebKit::RemoteGraphicsContextGLProxy::uniform1f):
(WebKit::RemoteGraphicsContextGLProxy::uniform1fv):
(WebKit::RemoteGraphicsContextGLProxy::uniform1i):
(WebKit::RemoteGraphicsContextGLProxy::uniform1iv):
(WebKit::RemoteGraphicsContextGLProxy::uniform2f):
(WebKit::RemoteGraphicsContextGLProxy::uniform2fv):
(WebKit::RemoteGraphicsContextGLProxy::uniform2i):
(WebKit::RemoteGraphicsContextGLProxy::uniform2iv):
(WebKit::RemoteGraphicsContextGLProxy::uniform3f):
(WebKit::RemoteGraphicsContextGLProxy::uniform3fv):
(WebKit::RemoteGraphicsContextGLProxy::uniform3i):
(WebKit::RemoteGraphicsContextGLProxy::uniform3iv):
(WebKit::RemoteGraphicsContextGLProxy::uniform4f):
(WebKit::RemoteGraphicsContextGLProxy::uniform4fv):
(WebKit::RemoteGraphicsContextGLProxy::uniform4i):
(WebKit::RemoteGraphicsContextGLProxy::uniform4iv):
(WebKit::RemoteGraphicsContextGLProxy::uniformMatrix2fv):
(WebKit::RemoteGraphicsContextGLProxy::uniformMatrix3fv):
(WebKit::RemoteGraphicsContextGLProxy::uniformMatrix4fv):
(WebKit::RemoteGraphicsContextGLProxy::useProgram):
(WebKit::RemoteGraphicsContextGLProxy::validateProgram):
(WebKit::RemoteGraphicsContextGLProxy::vertexAttrib1f):
(WebKit::RemoteGraphicsContextGLProxy::vertexAttrib1fv):
(WebKit::RemoteGraphicsContextGLProxy::vertexAttrib2f):
(WebKit::RemoteGraphicsContextGLProxy::vertexAttrib2fv):
(WebKit::RemoteGraphicsContextGLProxy::vertexAttrib3f):
(WebKit::RemoteGraphicsContextGLProxy::vertexAttrib3fv):
(WebKit::RemoteGraphicsContextGLProxy::vertexAttrib4f):
(WebKit::RemoteGraphicsContextGLProxy::vertexAttrib4fv):
(WebKit::RemoteGraphicsContextGLProxy::vertexAttribPointer):
(WebKit::RemoteGraphicsContextGLProxy::viewport):
(WebKit::RemoteGraphicsContextGLProxy::bufferData):
(WebKit::RemoteGraphicsContextGLProxy::bufferSubData):
(WebKit::RemoteGraphicsContextGLProxy::readnPixels):
(WebKit::RemoteGraphicsContextGLProxy::texImage2D):
(WebKit::RemoteGraphicsContextGLProxy::texSubImage2D):
(WebKit::RemoteGraphicsContextGLProxy::compressedTexImage2D):
(WebKit::RemoteGraphicsContextGLProxy::compressedTexSubImage2D):
(WebKit::RemoteGraphicsContextGLProxy::drawArraysInstanced):
(WebKit::RemoteGraphicsContextGLProxy::drawElementsInstanced):
(WebKit::RemoteGraphicsContextGLProxy::vertexAttribDivisor):
(WebKit::RemoteGraphicsContextGLProxy::createVertexArray):
(WebKit::RemoteGraphicsContextGLProxy::deleteVertexArray):
(WebKit::RemoteGraphicsContextGLProxy::isVertexArray):
(WebKit::RemoteGraphicsContextGLProxy::bindVertexArray):
(WebKit::RemoteGraphicsContextGLProxy::copyBufferSubData):
(WebKit::RemoteGraphicsContextGLProxy::getBufferSubData):
(WebKit::RemoteGraphicsContextGLProxy::blitFramebuffer):
(WebKit::RemoteGraphicsContextGLProxy::framebufferTextureLayer):
(WebKit::RemoteGraphicsContextGLProxy::invalidateFramebuffer):
(WebKit::RemoteGraphicsContextGLProxy::invalidateSubFramebuffer):
(WebKit::RemoteGraphicsContextGLProxy::readBuffer):
(WebKit::RemoteGraphicsContextGLProxy::renderbufferStorageMultisample):
(WebKit::RemoteGraphicsContextGLProxy::texStorage2D):
(WebKit::RemoteGraphicsContextGLProxy::texStorage3D):
(WebKit::RemoteGraphicsContextGLProxy::texImage3D):
(WebKit::RemoteGraphicsContextGLProxy::texSubImage3D):
(WebKit::RemoteGraphicsContextGLProxy::copyTexSubImage3D):
(WebKit::RemoteGraphicsContextGLProxy::compressedTexImage3D):
(WebKit::RemoteGraphicsContextGLProxy::compressedTexSubImage3D):
(WebKit::RemoteGraphicsContextGLProxy::getFragDataLocation):
(WebKit::RemoteGraphicsContextGLProxy::uniform1ui):
(WebKit::RemoteGraphicsContextGLProxy::uniform2ui):
(WebKit::RemoteGraphicsContextGLProxy::uniform3ui):
(WebKit::RemoteGraphicsContextGLProxy::uniform4ui):
(WebKit::RemoteGraphicsContextGLProxy::uniform1uiv):
(WebKit::RemoteGraphicsContextGLProxy::uniform2uiv):
(WebKit::RemoteGraphicsContextGLProxy::uniform3uiv):
(WebKit::RemoteGraphicsContextGLProxy::uniform4uiv):
(WebKit::RemoteGraphicsContextGLProxy::uniformMatrix2x3fv):
(WebKit::RemoteGraphicsContextGLProxy::uniformMatrix3x2fv):
(WebKit::RemoteGraphicsContextGLProxy::uniformMatrix2x4fv):
(WebKit::RemoteGraphicsContextGLProxy::uniformMatrix4x2fv):
(WebKit::RemoteGraphicsContextGLProxy::uniformMatrix3x4fv):
(WebKit::RemoteGraphicsContextGLProxy::uniformMatrix4x3fv):
(WebKit::RemoteGraphicsContextGLProxy::vertexAttribI4i):
(WebKit::RemoteGraphicsContextGLProxy::vertexAttribI4iv):
(WebKit::RemoteGraphicsContextGLProxy::vertexAttribI4ui):
(WebKit::RemoteGraphicsContextGLProxy::vertexAttribI4uiv):
(WebKit::RemoteGraphicsContextGLProxy::vertexAttribIPointer):
(WebKit::RemoteGraphicsContextGLProxy::drawRangeElements):
(WebKit::RemoteGraphicsContextGLProxy::drawBuffers):
(WebKit::RemoteGraphicsContextGLProxy::clearBufferiv):
(WebKit::RemoteGraphicsContextGLProxy::clearBufferuiv):
(WebKit::RemoteGraphicsContextGLProxy::clearBufferfv):
(WebKit::RemoteGraphicsContextGLProxy::clearBufferfi):
(WebKit::RemoteGraphicsContextGLProxy::createQuery):
(WebKit::RemoteGraphicsContextGLProxy::deleteQuery):
(WebKit::RemoteGraphicsContextGLProxy::isQuery):
(WebKit::RemoteGraphicsContextGLProxy::beginQuery):
(WebKit::RemoteGraphicsContextGLProxy::endQuery):
(WebKit::RemoteGraphicsContextGLProxy::getQuery):
(WebKit::RemoteGraphicsContextGLProxy::getQueryObjectui):
(WebKit::RemoteGraphicsContextGLProxy::createSampler):
(WebKit::RemoteGraphicsContextGLProxy::deleteSampler):
(WebKit::RemoteGraphicsContextGLProxy::isSampler):
(WebKit::RemoteGraphicsContextGLProxy::bindSampler):
(WebKit::RemoteGraphicsContextGLProxy::samplerParameteri):
(WebKit::RemoteGraphicsContextGLProxy::samplerParameterf):
(WebKit::RemoteGraphicsContextGLProxy::getSamplerParameterf):
(WebKit::RemoteGraphicsContextGLProxy::getSamplerParameteri):
(WebKit::RemoteGraphicsContextGLProxy::fenceSync):
(WebKit::RemoteGraphicsContextGLProxy::isSync):
(WebKit::RemoteGraphicsContextGLProxy::deleteSync):
(WebKit::RemoteGraphicsContextGLProxy::clientWaitSync):
(WebKit::RemoteGraphicsContextGLProxy::waitSync):
(WebKit::RemoteGraphicsContextGLProxy::getSynci):
(WebKit::RemoteGraphicsContextGLProxy::createTransformFeedback):
(WebKit::RemoteGraphicsContextGLProxy::deleteTransformFeedback):
(WebKit::RemoteGraphicsContextGLProxy::isTransformFeedback):
(WebKit::RemoteGraphicsContextGLProxy::bindTransformFeedback):
(WebKit::RemoteGraphicsContextGLProxy::beginTransformFeedback):
(WebKit::RemoteGraphicsContextGLProxy::endTransformFeedback):
(WebKit::RemoteGraphicsContextGLProxy::transformFeedbackVaryings):
(WebKit::RemoteGraphicsContextGLProxy::getTransformFeedbackVarying):
(WebKit::RemoteGraphicsContextGLProxy::pauseTransformFeedback):
(WebKit::RemoteGraphicsContextGLProxy::resumeTransformFeedback):
(WebKit::RemoteGraphicsContextGLProxy::bindBufferBase):
(WebKit::RemoteGraphicsContextGLProxy::bindBufferRange):
(WebKit::RemoteGraphicsContextGLProxy::getUniformIndices):
(WebKit::RemoteGraphicsContextGLProxy::getActiveUniforms):
(WebKit::RemoteGraphicsContextGLProxy::getUniformBlockIndex):
(WebKit::RemoteGraphicsContextGLProxy::getActiveUniformBlockName):
(WebKit::RemoteGraphicsContextGLProxy::uniformBlockBinding):
(WebKit::RemoteGraphicsContextGLProxy::getActiveUniformBlockiv):
(WebKit::RemoteGraphicsContextGLProxy::getGraphicsResetStatusARB):
(WebKit::RemoteGraphicsContextGLProxy::blitFramebufferANGLE):
(WebKit::RemoteGraphicsContextGLProxy::renderbufferStorageMultisampleANGLE):
(WebKit::RemoteGraphicsContextGLProxy::createVertexArrayOES):
(WebKit::RemoteGraphicsContextGLProxy::deleteVertexArrayOES):
(WebKit::RemoteGraphicsContextGLProxy::isVertexArrayOES):
(WebKit::RemoteGraphicsContextGLProxy::bindVertexArrayOES):
(WebKit::RemoteGraphicsContextGLProxy::getTranslatedShaderSourceANGLE):
(WebKit::RemoteGraphicsContextGLProxy::insertEventMarkerEXT):
(WebKit::RemoteGraphicsContextGLProxy::pushGroupMarkerEXT):
(WebKit::RemoteGraphicsContextGLProxy::popGroupMarkerEXT):
(WebKit::RemoteGraphicsContextGLProxy::drawBuffersEXT):
(WebKit::RemoteGraphicsContextGLProxy::drawArraysInstancedANGLE):
(WebKit::RemoteGraphicsContextGLProxy::drawElementsInstancedANGLE):
(WebKit::RemoteGraphicsContextGLProxy::vertexAttribDivisorANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getInternalformativ):
(WebKit::RemoteGraphicsContextGLProxy::readPixelsRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::texParameterfvRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::texParameterivRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getQueryivRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getQueryObjectuivRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getBufferPointervRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getInternalformativRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getVertexAttribIivRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getVertexAttribIuivRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getUniformuivRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getBufferParameteri64vRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::samplerParameterivRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::samplerParameterfvRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getSamplerParameterivRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getSamplerParameterfvRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getFramebufferParameterivRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getProgramInterfaceivRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getBooleani_vRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getMultisamplefvRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getTexLevelParameterivRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getTexLevelParameterfvRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getPointervRobustANGLERobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getnUniformfvRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getnUniformivRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getnUniformuivRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::texParameterIivRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::texParameterIuivRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getTexParameterIivRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getTexParameterIuivRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::samplerParameterIivRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::samplerParameterIuivRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getSamplerParameterIivRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getSamplerParameterIuivRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getQueryObjectivRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getQueryObjecti64vRobustANGLE):
(WebKit::RemoteGraphicsContextGLProxy::getQueryObjectui64vRobustANGLE):
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::copyVideoTextureToPlatformTexture):
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::connectToGPUProcess):
* WebProcess/GPU/media/RemoteAudioSession.cpp:
* WebProcess/GPU/media/RemoteAudioSourceProviderManager.cpp:
* WebProcess/GPU/media/RemoteAudioSourceProviderManager.h:
* WebProcess/GPU/media/RemoteLegacyCDMSession.cpp:
* WebProcess/GPU/media/SourceBufferPrivateRemote.h:
* WebProcess/GPU/media/TextTrackPrivateRemote.h:
* WebProcess/GPU/webrtc/LibWebRTCCodecs.cpp:
* WebProcess/GPU/webrtc/LibWebRTCCodecs.h:
* WebProcess/InjectedBundle/InjectedBundle.h:
* WebProcess/Network/NetworkProcessConnection.h:
* WebProcess/Network/WebResourceLoader.cpp:
* WebProcess/Network/WebResourceLoader.h:
* WebProcess/Network/WebSocketChannel.cpp:
* WebProcess/Network/WebSocketChannel.h:
* WebProcess/Network/WebSocketStream.cpp:
* WebProcess/Network/WebSocketStream.h:
* WebProcess/Network/webrtc/LibWebRTCSocket.h:
* WebProcess/Plugins/PDF/PDFPlugin.h:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::createGraphicsContextGL const):
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
(WebKit::WebPage::updatePreferences):
* WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::sendSyncWithDelayedReply):
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::handleEditingKeyboardEvent):
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::setUseGPUProcessForWebGL):
(WebKit::WebProcess::shouldUseRemoteRenderingForWebGL const):
* WebProcess/WebProcess.h:
2020-12-09 Brent Fulgham <bfulgham@apple.com>
REGRESSION(r269647) [Cocoa] Consume GPU sandbox extensions early in Web page creation flow
https://bugs.webkit.org/show_bug.cgi?id=219659
<rdar://problem/72058396>
Reviewed by Jer Noble.
We need to consume GPU sandbox extensions earlier in the Web page creation flow so that API calls to things like
muting video have the relevant mach connections available.
This change corrects a regression introduced in Bug 218708 (r269647).
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::WebPage): Move extension consumption earlier in the constructor.
2020-12-09 Carlos Garcia Campos <cgarcia@igalia.com>
WebDriver: add support for pen pointer events
https://bugs.webkit.org/show_bug.cgi?id=219024
Reviewed by Brian Burg.
Make it possible to synthesize pointer events for any pointer type. Add pen input source to automation and
handle it when dispatching actions.
* Shared/NativeWebMouseEvent.h:
* Shared/WebEventConversion.cpp:
(WebKit::WebKit2PlatformMouseEvent::WebKit2PlatformMouseEvent):
* Shared/WebMouseEvent.cpp:
(WebKit::WebMouseEvent::WebMouseEvent):
(WebKit::WebMouseEvent::encode const):
(WebKit::WebMouseEvent::decode):
* Shared/WebMouseEvent.h:
(WebKit::WebMouseEvent::pointerId const):
(WebKit::WebMouseEvent::pointerType const):
* Shared/gtk/NativeWebMouseEventGtk.cpp:
(WebKit::NativeWebMouseEvent::NativeWebMouseEvent):
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(primaryPointerForType):
(webkitWebViewBaseSynthesizeMouseEvent):
* UIProcess/API/gtk/WebKitWebViewBaseInternal.h:
* UIProcess/Automation/Automation.json:
* UIProcess/Automation/SimulatedInputDispatcher.cpp:
(WebKit::SimulatedInputSourceState::emptyStateForSourceType):
(WebKit::SimulatedInputDispatcher::transitionInputSourceToState):
* UIProcess/Automation/SimulatedInputDispatcher.h:
* UIProcess/Automation/WebAutomationSession.cpp:
(WebKit::WebAutomationSession::simulateMouseInteraction):
(WebKit::WebAutomationSession::performMouseInteraction):
(WebKit::simulatedInputSourceTypeFromProtocolSourceType):
(WebKit::WebAutomationSession::performInteractionSequence):
* UIProcess/Automation/WebAutomationSession.h:
* UIProcess/Automation/gtk/WebAutomationSessionGtk.cpp:
(WebKit::WebAutomationSession::platformSimulateMouseInteraction):
* UIProcess/Automation/mac/WebAutomationSessionMac.mm:
(WebKit::WebAutomationSession::platformSimulateMouseInteraction):
* UIProcess/Automation/wpe/WebAutomationSessionWPE.cpp:
(WebKit::WebAutomationSession::platformSimulateMouseInteraction):
* UIProcess/gtk/PointerLockManager.cpp:
(WebKit::PointerLockManager::handleMotion):
2020-12-09 Stephan Szabo <stephan.szabo@sony.com>
[PlayStation] Fix PlayStation GPU process code after r270563
https://bugs.webkit.org/show_bug.cgi?id=219682
Unreviewed build fix.
Remove mediaPlayerRenderingModeChanged from playstation specific
RemoteMediaPlayerProxyPlayStation file similar to the removals from
other versions in r270563.
* GPUProcess/media/playstation/RemoteMediaPlayerProxyPlayStation.cpp:
2020-12-09 Per Arne Vollan <pvollan@apple.com>
[macOS] Change Universal Access zoom in the UI process
https://bugs.webkit.org/show_bug.cgi?id=219635
Reviewed by Darin Adler.
Implement the method changeUniversalAccessZoomFocus in the WebProcess' chrome client, which will send a message
to the UI process, where the Universal Access zoom will be changed.
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::changeUniversalAccessZoomFocus):
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::changeUniversalAccessZoomFocus):
* WebProcess/WebCoreSupport/WebChromeClient.h:
2020-12-09 Ryosuke Niwa <rniwa@webkit.org>
iOS build fix after r270573.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::handlePreventableTouchEvent):
2020-12-09 Per Arne Vollan <pvollan@apple.com>
Remove access to com.apple.audio.toolbox.reporting.service
https://bugs.webkit.org/show_bug.cgi?id=219638
<rdar://problem/71638615>
Reviewed by Darin Adler.
Remove access to com.apple.audio.toolbox.reporting.service for WebKit processes, since this is a logging service which is not needed for correctness.
* Resources/SandboxProfiles/ios/com.apple.WebKit.GPU.sb:
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* UIProcess/WebPageProxy.cpp:
(WebKit::mediaRelatedMachServices):
* WebProcess/com.apple.WebProcess.sb.in:
2020-12-09 Sihui Liu <sihui_liu@apple.com>
Implement audio capture for SpeechRecognition on iOS
https://bugs.webkit.org/show_bug.cgi?id=219371
<rdar://problem/71824049>
Reviewed by Youenn Fablet.
On iOS, audio capture is done in web process. For speech recognition, recognizer is in UI process. Therefore,
UI process needs to ask web process to capture audio and send back data. To implement this, this patch added:
- SpeechRecognitionRemoteRealtimeMediaSource: subclass WebCore::RealtimeMediaSource and connect to a remote
WebCore::RealtimeMediaSource (owned by SpeechRecognitionRealtimeMediaSourceManager::Source) in web process.
- SpeechRecognitionRemoteRealtimeMediaSourceManager: manage all RemoteRealtimeMediaSources.
- SpeechRecognitionRealtimeMediaSourceManager: manage all SpeechRecognitionRemoteRealtimeMediaSource.
- SpeechRecognitionRealtimeMediaSourceManager::Source: own and monitor a local WebCore::RealtimeMediaSource.
This structure is very similar to how we implement media capture on macOS (RemoteRealtimeMediaSource,
UserMediaCaptureManager and UserMediaCaptureManagerProxy), where we send data from UI process to web process
instead. We should unify the architecture if possible.
* CMakeLists.txt:
* DerivedSources-input.xcfilelist:
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* Sources.txt:
* UIProcess/SpeechRecognitionRemoteRealtimeMediaSource.cpp: Added.
(WebKit::SpeechRecognitionRemoteRealtimeMediaSource::create):
(WebKit::SpeechRecognitionRemoteRealtimeMediaSource::SpeechRecognitionRemoteRealtimeMediaSource):
(WebKit::m_ringBuffer):
(WebKit::SpeechRecognitionRemoteRealtimeMediaSource::~SpeechRecognitionRemoteRealtimeMediaSource):
(WebKit::SpeechRecognitionRemoteRealtimeMediaSource::startProducingData):
(WebKit::SpeechRecognitionRemoteRealtimeMediaSource::stopProducingData):
(WebKit::SpeechRecognitionRemoteRealtimeMediaSource::setStorage):
(WebKit::SpeechRecognitionRemoteRealtimeMediaSource::remoteAudioSamplesAvailable):
(WebKit::SpeechRecognitionRemoteRealtimeMediaSource::remoteCaptureFailed):
(WebKit::SpeechRecognitionRemoteRealtimeMediaSource::remoteSourceStopped):
* UIProcess/SpeechRecognitionRemoteRealtimeMediaSource.h: Added.
(WebKit::SpeechRecognitionRemoteRealtimeMediaSource::identifier const):
* UIProcess/SpeechRecognitionRemoteRealtimeMediaSourceManager.cpp: Added.
(WebKit::SpeechRecognitionRemoteRealtimeMediaSourceManager::SpeechRecognitionRemoteRealtimeMediaSourceManager):
(WebKit::SpeechRecognitionRemoteRealtimeMediaSourceManager::addSource):
(WebKit::SpeechRecognitionRemoteRealtimeMediaSourceManager::removeSource):
(WebKit::SpeechRecognitionRemoteRealtimeMediaSourceManager::remoteAudioSamplesAvailable):
(WebKit::SpeechRecognitionRemoteRealtimeMediaSourceManager::remoteCaptureFailed):
(WebKit::SpeechRecognitionRemoteRealtimeMediaSourceManager::remoteSourceStopped):
(WebKit::SpeechRecognitionRemoteRealtimeMediaSourceManager::messageSenderConnection const):
(WebKit::SpeechRecognitionRemoteRealtimeMediaSourceManager::messageSenderDestinationID const):
(WebKit::SpeechRecognitionRemoteRealtimeMediaSourceManager::setStorage):
* UIProcess/SpeechRecognitionRemoteRealtimeMediaSourceManager.h: Added.
* UIProcess/SpeechRecognitionRemoteRealtimeMediaSourceManager.messages.in: Added.
* UIProcess/SpeechRecognitionServer.cpp:
(WebKit::SpeechRecognitionServer::SpeechRecognitionServer):
(WebKit::SpeechRecognitionServer::handleRequest):
* UIProcess/SpeechRecognitionServer.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::createRealtimeMediaSourceForSpeechRecognition):
* UIProcess/WebPageProxy.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::createSpeechRecognitionServer):
(WebKit::WebProcessProxy::ensureSpeechRecognitionRemoteRealtimeMediaSourceManager):
* UIProcess/WebProcessProxy.h:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/Speech/SpeechRecognitionRealtimeMediaSourceManager.cpp: Added.
(WebKit::SpeechRecognitionRealtimeMediaSourceManager::Source::Source):
(WebKit::SpeechRecognitionRealtimeMediaSourceManager::Source::~Source):
(WebKit::SpeechRecognitionRealtimeMediaSourceManager::Source::start):
(WebKit::SpeechRecognitionRealtimeMediaSourceManager::Source::stop):
(WebKit::SpeechRecognitionRealtimeMediaSourceManager::Source::storage):
(WebKit::SpeechRecognitionRealtimeMediaSourceManager::SpeechRecognitionRealtimeMediaSourceManager):
(WebKit::SpeechRecognitionRealtimeMediaSourceManager::~SpeechRecognitionRealtimeMediaSourceManager):
(WebKit::SpeechRecognitionRealtimeMediaSourceManager::grantSandboxExtensions):
(WebKit::SpeechRecognitionRealtimeMediaSourceManager::revokeSandboxExtensions):
(WebKit::SpeechRecognitionRealtimeMediaSourceManager::createSource):
(WebKit::SpeechRecognitionRealtimeMediaSourceManager::deleteSource):
(WebKit::SpeechRecognitionRealtimeMediaSourceManager::start):
(WebKit::SpeechRecognitionRealtimeMediaSourceManager::stop):
(WebKit::SpeechRecognitionRealtimeMediaSourceManager::messageSenderConnection const):
(WebKit::SpeechRecognitionRealtimeMediaSourceManager::messageSenderDestinationID const):
* WebProcess/Speech/SpeechRecognitionRealtimeMediaSourceManager.h: Added.
* WebProcess/Speech/SpeechRecognitionRealtimeMediaSourceManager.messages.in: Added.
* WebProcess/WebCoreSupport/WebSpeechRecognitionConnection.cpp:
(WebKit::WebSpeechRecognitionConnection::WebSpeechRecognitionConnection):
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::ensureSpeechRecognitionRealtimeMediaSourceManager):
* WebProcess/WebProcess.h:
2020-12-08 Kimmo Kinnunen <kkinnunen@apple.com>
IPC should support non-byte aligned data references
https://bugs.webkit.org/show_bug.cgi?id=219487
Reviewed by Chris Dumez.
Add a templated version of IPC::DataReference: ArrayReference<type, number of elements>.
The number of elements can be omitted, in which case the ArrayReference
size is determined at run-time.
This way pointers of typed arrays such as float arrays can be transferred directly
from the IPC message data store to the target. This is useful when implementing
IPC for WebGL in GPU process.
Changes IPC to support passing data references, i.e.
DataReference or ArrayReference, as a synchronous IPC call return value.
The caller must hold the Connection::syncSend() result (Decoder object)
for the duration of using the data values.
Makes DataReference an alias of uint8_t dynamic array reference.
Unfortunately this makes the forward declaring DataReference hard.
Due to this, changes use declaration sites to #include instead of
using forward declarations. Removes forward declarations for headers
where the type is not used.
Adds various #includes to fix errors hidden by the unified build.
Adds some WebCore:: namespace qualifications to fix errors hidden
by the unified build.
* GPUProcess/media/RemoteAudioSessionProxy.cpp:
* GPUProcess/media/RemoteLegacyCDMProxy.cpp:
* GPUProcess/media/RemoteLegacyCDMProxy.h:
* GPUProcess/media/RemoteLegacyCDMSessionProxy.cpp:
* GPUProcess/media/RemoteMediaResourceManager.cpp:
* GPUProcess/media/RemoteMediaResourceManager.h:
* GPUProcess/media/RemoteMediaSourceProxy.h:
* GPUProcess/media/RemoteSourceBufferProxy.cpp:
* GPUProcess/media/RemoteSourceBufferProxy.h:
* GPUProcess/webrtc/LibWebRTCCodecsProxy.h:
* GPUProcess/webrtc/LibWebRTCCodecsProxy.mm:
* GPUProcess/webrtc/RemoteMediaRecorder.h:
* NetworkProcess/CustomProtocols/Cocoa/LegacyCustomProtocolManagerCocoa.mm:
* NetworkProcess/CustomProtocols/LegacyCustomProtocolManager.h:
* NetworkProcess/Downloads/Download.cpp:
* NetworkProcess/Downloads/Download.h:
* NetworkProcess/Downloads/DownloadManager.cpp:
* NetworkProcess/Downloads/DownloadManager.h:
* NetworkProcess/NetworkSocketChannel.cpp:
* NetworkProcess/NetworkSocketChannel.h:
* NetworkProcess/NetworkSocketStream.cpp:
* NetworkProcess/NetworkSocketStream.h:
* NetworkProcess/ServiceWorker/ServiceWorkerFetchTask.h:
* NetworkProcess/WebSocketTask.h:
* NetworkProcess/cocoa/WebSocketTaskCocoa.h:
* NetworkProcess/cocoa/WebSocketTaskCocoa.mm:
* NetworkProcess/soup/WebSocketTaskSoup.cpp:
* NetworkProcess/soup/WebSocketTaskSoup.h:
* Platform/IPC/ArgumentCoders.h:
* Platform/IPC/ArrayReference.h: Copied from Source/WebKit/Platform/IPC/DataReference.h.
(IPC::ArrayReference::ArrayReference):
(IPC::ArrayReference::isEmpty const):
(IPC::ArrayReference::size const):
(IPC::ArrayReference::data const):
(IPC::ArrayReference::vector const):
* Platform/IPC/Connection.h:
(IPC::Connection::sendSync):
* Platform/IPC/DataReference.cpp: Removed.
* Platform/IPC/DataReference.h:
* Platform/IPC/Decoder.cpp:
(IPC::Decoder::decodeFixedLengthReference):
* Platform/IPC/Decoder.h:
* Platform/IPC/Encoder.cpp:
(IPC::Encoder::wrapForTesting):
* Platform/IPC/Encoder.h:
* Platform/IPC/HandleMessage.h:
* Platform/IPC/MessageSender.h:
(IPC::MessageSender::sendSync):
* Platform/IPC/SharedBufferCopy.cpp:
(IPC::SharedBufferCopy::decode):
* Platform/IPC/SharedBufferDataReference.cpp:
* Platform/IPC/StringReference.cpp:
* PluginProcess/PluginControllerProxy.cpp:
* PluginProcess/PluginControllerProxy.h:
* Scripts/webkit/messages.py:
* Shared/API/APIData.cpp:
* Shared/cf/ArgumentCodersCF.cpp:
* Shared/soup/WebCoreArgumentCodersSoup.cpp:
(IPC::ArgumentCoder<CertificateInfo>::encode):
(IPC::ArgumentCoder<CertificateInfo>::decode):
* Sources.txt:
* UIProcess/API/APIIconLoadingClient.h:
* UIProcess/AuxiliaryProcessProxy.h:
(WebKit::AuxiliaryProcessProxy::sendSync):
* UIProcess/Cocoa/SOAuthorization/SubFrameSOAuthorizationSession.h:
* UIProcess/Network/CustomProtocols/LegacyCustomProtocolManagerProxy.h:
* UIProcess/PageClient.h:
* UIProcess/UserContent/WebUserContentControllerProxy.cpp:
* UIProcess/UserContent/WebUserContentControllerProxy.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
* UIProcess/WebPageProxy.h:
* UIProcess/WebURLSchemeHandler.h:
* UIProcess/WebURLSchemeTask.h:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/ApplePay/WebPaymentCoordinator.h:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::copyVideoTextureToPlatformTexture):
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::connectToGPUProcess):
* WebProcess/GPU/media/RemoteAudioSourceProvider.cpp:
* WebProcess/GPU/media/RemoteAudioSourceProviderManager.cpp:
* WebProcess/GPU/media/RemoteAudioSourceProviderManager.h:
* WebProcess/GPU/media/RemoteLegacyCDMSession.cpp:
* WebProcess/GPU/media/TextTrackPrivateRemote.h:
* WebProcess/GPU/webrtc/LibWebRTCCodecs.cpp:
* WebProcess/GPU/webrtc/LibWebRTCCodecs.h:
* WebProcess/InjectedBundle/InjectedBundle.h:
* WebProcess/Inspector/WebInspectorUIExtensionController.cpp:
(WebKit::WebInspectorUIExtensionController::WebInspectorUIExtensionController):
(WebKit::WebInspectorUIExtensionController::registerExtension):
(WebKit::WebInspectorUIExtensionController::unregisterExtension):
* WebProcess/Network/NetworkProcessConnection.h:
* WebProcess/Network/WebResourceLoader.cpp:
* WebProcess/Network/WebResourceLoader.h:
* WebProcess/Network/WebSocketChannel.cpp:
* WebProcess/Network/WebSocketChannel.h:
* WebProcess/Network/WebSocketStream.cpp:
* WebProcess/Network/WebSocketStream.h:
* WebProcess/Network/webrtc/LibWebRTCSocket.h:
* WebProcess/Plugins/PDF/PDFPlugin.h:
* WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::sendSyncWithDelayedReply):
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::handleEditingKeyboardEvent):
2020-12-08 Wenson Hsieh <wenson_hsieh@apple.com>
[Concurrent Display Lists] Add a new type to represent a set of GPU process wakeup arguments
https://bugs.webkit.org/show_bug.cgi?id=219657
Reviewed by Tim Horton.
Pull the initial item buffer identifier, offset, and destination image buffer identifier used to wake the GPU
process out into a separate struct, in `GPUProcessWakeupMessageArguments.h`, and refactor both the web and GPU
process-side rendering backends to use this new struct.
This will greatly simplify future logic that will be added to address <webkit.org/b/218614>, which requires us
to defer calling into the GPU process wakeup loop in more scenarios. No change in behavior.
* GPUProcess/graphics/RemoteImageBuffer.h:
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::nextDestinationImageBufferAfterApplyingDisplayLists):
(WebKit::RemoteRenderingBackend::wakeUpAndApplyDisplayList):
(WebKit::RemoteRenderingBackend::setNextItemBufferToRead):
(WebKit::RemoteRenderingBackend::didCreateSharedDisplayListHandle):
* GPUProcess/graphics/RemoteRenderingBackend.h:
* GPUProcess/graphics/RemoteRenderingBackend.messages.in:
* Shared/GPUProcessWakeupMessageArguments.h: Added.
(WebKit::GPUProcessWakeupMessageArguments::encode const):
(WebKit::GPUProcessWakeupMessageArguments::decode):
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::sendWakeupMessage):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
2020-12-08 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process][MSE] The GPU Process may crash in the destructor of MediaPlayerPrivateMediaSourceAVFObjC
https://bugs.webkit.org/show_bug.cgi?id=219622
Reviewed by Eric Carlson.
The purpose of `mediaPlayerRenderingModeChanged()` callback is to notify the media element
that the player's rendering mode is changed, including the case that the layer is destroyed.
So we should not operate on the layer in this function.
With this patch, `RemoteMediaPlayerProxy` forwards the callback from the GPU process to
the Web process. This patch also cleans up `RemoteMediaPlayerProxy.h`.
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
(WebKit::RemoteMediaPlayerProxy::mediaPlayerRenderingModeChanged):
* GPUProcess/media/RemoteMediaPlayerProxy.h:
* GPUProcess/media/cocoa/RemoteMediaPlayerProxyCocoa.mm:
(WebKit::RemoteMediaPlayerProxy::mediaPlayerRenderingModeChanged): Deleted.
* GPUProcess/media/gstreamer/RemoteMediaPlayerProxyGStreamer.cpp:
(WebKit::RemoteMediaPlayerProxy::mediaPlayerRenderingModeChanged): Deleted.
* GPUProcess/media/win/RemoteMediaPlayerProxyWin.cpp:
(WebKit::RemoteMediaPlayerProxy::mediaPlayerRenderingModeChanged): Deleted.
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::renderingModeChanged):
* WebProcess/GPU/media/MediaPlayerPrivateRemote.h:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.messages.in:
2020-12-08 Chris Dumez <cdumez@apple.com>
Potential crash under [WKRemoteObjectEncoder encodeObject:forKey:] when the object graph contains a cycle
https://bugs.webkit.org/show_bug.cgi?id=219620
<rdar://71551776>
Reviewed by Geoffrey Garen.
Update WKRemoteObjectEncoder to detect cycles when encoding objects. When a cycle is detected, we
first attempt to encode a default-initialized object of the same type instead. If that fails, we
raise a NSInvalidArgumentException.
Based on crashes in the wild, we have evidence that such cycles are occuring and I suspect this is
caused by Norton Safe Web extension somehow.
* Shared/API/Cocoa/WKRemoteObjectCoder.mm:
(-[WKRemoteObjectEncoder init]):
(encodeObject):
2020-12-08 Simon Fraser <simon.fraser@apple.com>
Make ScrollingPerformanceLoggingEnabled an internal debug preference
https://bugs.webkit.org/show_bug.cgi?id=219647
Reviewed by Sam Weinig.
ScrollingPerformanceLoggingEnabled is used by a Safari performance test, so it needs to be
easily toggleable from a script. That's easier if it's an Internal Debug preference.
Testing also sets ForceUpdateScrollbarsOnMainThreadForPerformanceTesting but we can just use
the ScrollingPerformanceLoggingEnabled flag, but it makes more sense to rename
ScrollingPerformanceLoggingEnabled to ScrollingPerformanceTestingEnabled when we do that.
Also fix how we propagate ScrollingPerformanceTestingEnabled to the scrolling tree;
do so via ScrollingStateFrameScrollingNode as we do for the other tree-wide settings,
removing code in TiledCoreAnimationDrawingArea::updatePreferences() that did this.
* Shared/RemoteLayerTree/RemoteScrollingCoordinatorTransaction.cpp:
(ArgumentCoder<ScrollingStateFrameScrollingNode>::encode):
(ArgumentCoder<ScrollingStateFrameScrollingNode>::decode):
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetScrollingPerformanceLoggingEnabled):
(WKPreferencesGetScrollingPerformanceLoggingEnabled):
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::forceUpdateScrollbarsOnMainThreadForPerformanceTesting const):
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::updatePreferences):
2020-12-08 Per Arne Vollan <pvollan@apple.com>
[macOS] Deny mach-lookup to the service 'com.apple.assertiond.processassertionconnection' in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=219637
<rdar://problem/69019269>
Reviewed by Darin Adler.
Deny mach-lookup to the service 'com.apple.assertiond.processassertionconnection' in the WebContent process on macOS,
since there are no reports of this service being used.
No new tests, since the result will depend on macOS version.
* WebProcess/com.apple.WebProcess.sb.in:
2020-12-08 Alex Christensen <achristensen@webkit.org>
Remove NSString overrelease in _WKUserContentExtensionStore
https://bugs.webkit.org/show_bug.cgi?id=219621
Reviewed by Darin Adler.
Gasp!
* UIProcess/API/Cocoa/WKContentRuleListStore.mm:
(-[WKContentRuleListStore compileContentRuleListForIdentifier:encodedContentRuleList:completionHandler:]):
(-[WKContentRuleListStore _compileContentRuleListForIdentifier:encodedContentRuleList:completionHandler:]):
2020-12-08 Per Arne Vollan <pvollan@apple.com>
[macOS] Create sandbox extension for audio service when Media in GPU process is not enabled
https://bugs.webkit.org/show_bug.cgi?id=219540
<rdar://problem/71978050>
Reviewed by Brent Fulgham.
Create sandbox extension for com.apple.audio.SandboxHelper when Media in GPU process is not enabled.
* UIProcess/WebPageProxy.cpp:
(WebKit::mediaRelatedMachServices):
* WebProcess/com.apple.WebProcess.sb.in:
2020-12-08 Chris Dumez <cdumez@apple.com>
[iOS] Stop leaking an XPC boost message to XPC services
https://bugs.webkit.org/show_bug.cgi?id=219453
Reviewed by Geoffrey Garen.
Stop leaking an XPC boost message to XPC services on builds where RunningBoard is able to give the right
priorities to our XPC services based on the type of RunningBoard assertion is being held. There is no
behavior change for the Network/WebContent processes since those were blocklisted in RunningBoard (Turning
on the RB_full_manage_WK_jetsam feature flag gives us the default RunningBoard behavior without
blocklisting). However, this allows the new GPUProcess to move to IDLE jetsam band when homing out of
MobileSafari, which makes the GPUProcess eligible for the freezer. Previously, the leaked boost message
would keep the GPUProcess in a higher jetsam band.
* FeatureFlags/WebKit.plist:
Add RB_full_manage_WK_jetsam feature flag and turn it on. On recent enough builds of RunningBoard, this
disables the blocklisting of the WebContent/Network processes in RunningBoard so that RunningBoard can now
set the priority of those processes (instead of us boosting our priority via boost messages).
* GPUProcess/EntryPoint/Cocoa/XPCService/GPUService/Info-iOS.plist:
* NetworkProcess/EntryPoint/Cocoa/XPCService/NetworkService/Info-iOS.plist:
* WebAuthnProcess/EntryPoint/Cocoa/XPCService/WebAuthnService/Info-iOS.plist:
* WebProcess/EntryPoint/Cocoa/XPCService/WebContentService/Info-iOS.plist:
Set _ProcessType=App so that RunningBoard is able to set the priority of our XPC services on iOS. Without
this change, we get a permission denied when RunningBoard attempts to set the priority of our XPC services.
* UIProcess/Launcher/mac/ProcessLauncherMac.mm:
(WebKit::shouldLeakBoost):
Stop leaking a boost message on iOS when RUNNINGBOARD_WEBKIT_PRIORITY_SUPPORT compile-time flag is enabled
and when the RB_full_manage_WK_jetsam feature flag is enabled.
2020-12-08 Ryan Hostetler <rhost@apple.com>
Add target to create /usr/local to work around XBS Bug <rdar://problem/20388650>
https://bugs.webkit.org/show_bug.cgi?id=219618
rdar://72064576
Reviewed by Jonathan Bedard.
* WebKit.xcodeproj/project.pbxproj:
Add a build target EmptyDSTROOT with script to create an empty folder "/usr/local".
This allows WebKit to retain a unified XBS project name across different train configurations.
2020-12-07 Said Abou-Hallawa <said@apple.com>
[GPU Process] Ensure the backend of ImageBuffer is created before getting its handle
https://bugs.webkit.org/show_bug.cgi?id=219623
Reviewed by Tim Horton.
In r270458 and in RemoteLayerBackingStore::encode(), casting the ImageBuffer
and calling the overriding createImageBufferBackendHandle() of the superclass
was replaced by casting the backend and calling its createImageBufferBackendHandle()
directly.
This may lead to crash if the backend was not created when calling its
createImageBufferBackendHandle(). The createImageBufferBackendHandle()
of the ImageBuffer superclass calls ensureBackendCreated() before calling
createImageBufferBackendHandle() of the backend.
So we need to replace to call backend() to ensureBackendCreated().
* Shared/RemoteLayerTree/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::encode const):
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
2020-12-07 Brent Fulgham <bfulgham@apple.com>
Add sandbox telemetry for likely unused features
https://bugs.webkit.org/show_bug.cgi?id=219544
<rdar://71985678>
Reviewed by Per Arne Vollan.
Add telemetry to confirm the results of our manual review of the sandbox. Add 'hw.memsize' to
the allowed sysctl's for the Network process on macOS, since it is needed by CFNetwork, and is
already permitted on iOS.
* NetworkProcess/mac/com.apple.WebKit.NetworkProcess.sb.in: Allow the 'hw.memsize' sysctl,
which is needed by CFNetwork, and is already allowed on iOS.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb: Add logging to confirm we
do not need file-read-metadata for syslog.
* WebProcess/com.apple.WebProcess.sb.in: Add logging to confirm BluetoothServices is not
needed.
2020-12-07 Wenson Hsieh <wenson_hsieh@apple.com>
Ensure that DisplayListSharedMemoryHeader is never explicitly created or destroyed at compile-time
https://bugs.webkit.org/show_bug.cgi?id=219609
Reviewed by Tim Horton.
Make sure that a `DisplayListSharedMemoryHeader` is never created or destroyed by marking the constructor and
destructor as deleted. A `DisplayListSharedMemoryHeader` should only ever be `reinterpret_cast`-ed from the
first few bytes of a shared display list handle.
* Shared/SharedDisplayListHandle.h:
2020-12-07 Per Arne Vollan <pvollan@apple.com>
[macOS] Deny mach-lookup to the service 'com.apple.audio.SystemSoundServer-OSX' in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=219590
<rdar://problem/69019399>
Reviewed by Brent Fulgham.
After r269593, mach-lookup to the service 'com.apple.audio.SystemSoundServer-OSX' in the WebContent process can be denied.
Test: fast/sandbox/mac/sandbox-mach-lookup.html
* WebProcess/com.apple.WebProcess.sb.in:
2020-12-07 Don Olmstead <don.olmstead@sony.com>
[CMake] Remove WEBKIT_WRAP_SOURCELIST
https://bugs.webkit.org/show_bug.cgi?id=196916
Reviewed by Michael Catanzaro.
* PlatformFTW.cmake:
* PlatformWin.cmake:
2020-12-07 Tim Horton <timothy_horton@apple.com>
Fix incorrect capitalization of ClearWheelEventTestMonitor IPC message name
https://bugs.webkit.org/show_bug.cgi?id=219606
Reviewed by Wenson Hsieh.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::clearWheelEventTestMonitor):
* WebProcess/WebPage/WebPage.messages.in:
2020-12-07 Chris Dumez <cdumez@apple.com>
[iOS] Add basic GPUProcess crash recovery logic to RemoteMediaSessionHelper
https://bugs.webkit.org/show_bug.cgi?id=219494
Reviewed by Eric Carlson.
* WebProcess/GPU/media/ios/RemoteMediaSessionHelper.cpp:
(WebKit::RemoteMediaSessionHelper::RemoteMediaSessionHelper):
(WebKit::RemoteMediaSessionHelper::connectToGPUProcess):
(WebKit::RemoteMediaSessionHelper::gpuProcessConnectionDidClose):
* WebProcess/GPU/media/ios/RemoteMediaSessionHelper.h:
2020-12-07 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r270498.
https://bugs.webkit.org/show_bug.cgi?id=219596
Broke the build as it landed simultaneously with a more
complete fix
Reverted changeset:
"Unreviewed. Add the switch case handling missed out from
r270415."
https://trac.webkit.org/changeset/270498
2020-12-07 Lauro Moura <lmoura@igalia.com>
Fix warnings when GPUProcess is disabled
https://bugs.webkit.org/show_bug.cgi?id=219587
Reviewed by Žan Doberšek.
* UIProcess/API/glib/WebKitNavigationClient.cpp: Handle new enum value
in switch after r270415.
* WebProcess/WebPage/DrawingArea.cpp:
(WebKit::DrawingArea::supportsGPUProcessRendering): Default returning
false when GPUProcess is disabled.
2020-12-07 Joonghun Park <jh718.park@samsung.com>
Unreviewed. Add the switch case handling missed out from r270415.
This change removes the build warning below.
warning: enumeration value ‘RequestedByGPUProcess’ not handled in switch [-Wswitch]
No new tests, no new behaviors.
* UIProcess/API/glib/WebKitNavigationClient.cpp:
2020-12-06 Samuel Thibault <samuel.thibault@ens-lyon.org>
[WebKit] Fix use of std::max_align_t
https://bugs.webkit.org/show_bug.cgi?id=219579
Reviewed by Fujii Hironori.
Standard says that it's alignof(std::max_align_t) that should be used,
not its size, e.g.
https://en.cppreference.com/w/cpp/types/max_align_t
and indeed at least with gcc 10's stddef.h, max_align_t is a structure
containing a long long, a long double, and a float18, thus not a power
of two.
* Shared/SharedDisplayListHandle.h:
(WebKit::SharedDisplayListHandle::headerSize): Use
alignof(std::max_align_t) instead of sizeof(std::max_align_t).
2020-12-06 Lauro Moura <lmoura@igalia.com>
Unreviewed. Add missing GPU_PROCESS guards around RemoteMediaEngineConfigurationFactory after r270462
https://bugs.webkit.org/show_bug.cgi?id=219578
Was breaking the Ubuntu LTS/Debian builds.
* WebProcess/GPU/media/RemoteMediaEngineConfigurationFactory.cpp:
* WebProcess/WebProcess.cpp:
2020-12-04 Wenson Hsieh <wenson_hsieh@apple.com>
[Concurrent display lists] Synchronize display list rendering across remote image buffers
https://bugs.webkit.org/show_bug.cgi?id=219091
<rdar://problem/71747695>
Reviewed by Geoff Garen.
Refactors the existing concurrent display list processing mechanism, such that we no longer update the unread
bytes count every 512 items (or when flushing the drawing context), and instead update it every time we append
display list item data. In order to achieve this without making it impossible for multiple display-list backed
image buffers to render simultaneously, we also add support for synchronizing display list items across
different image buffers by using the new meta command items added in bug #219262. As detailed in the ChangeLog
for that bug, a stream of display list data written by the web process and consumed by the GPU process is now
structured like so:
Wakeup message initiates display list processing with Image Buffer A and Item Buffer 1
|
| MetaCommandChangeDestinationImageBuffer(B)
| |
| | MetaCommandChangeItemBuffer(2)
| | |
| | |
v v v
+-----------------------+--------------------------------++
| +--------------------+ +------------------------------+ |
| | Destination | | Destination | | Item Buffer 1
| | Image Buffer A | | Image Buffer B | |
| +--------------------+ +------------------------------+ |
+---------------------------------------------------------+
+---------------------------------------------------------+
| +--------------+ +-----------------------+ |
| | Dst. Image | | Destination Image | Unused | Item Buffer 2
| | Buffer B | | Buffer A | Capacity |
| +--------------+ +-----------------------+ |
+-----------------+-------------------------+-------------+
^
|
|
MetaCommandChangeDestinationImageBuffer(A)
(Note that the `MetaCommandEnd` item has been omitted, since we no longer need to explicitly tell the GPU
process to go to sleep with the latest iteration of this patch).
In order to write display list data conforming to this format, we add a mechanism to the web process that keeps
track of the "current destination image buffer" (i.e. the image buffer that has most recently
appended a display list item). Right before this destination is about to change, we append a new
`MetaCommandChangeDestinationImageBuffer` item before proceeding to append display list items for the new
current destination.
Since we now bump the unread count every time, we need to do some additional work to avoid an excessive number
of wakeups and sleeps in the GPU process, which would otherwise cause a massive regression. To do this, we
introduce an item count hysteresis in the web process, such that we will wait for 512 items to be written before
attempting to send the wakeup message. This allows the web process to get a small head start over the GPU
process, and thus allows the GPU process to be more consistently busy over the course of a single frame.
On the reader (GPU process) side, we make some minor adjustments so that the main display list processing loop
of the GPU process is capable of changing destination image buffers while processing a stream of display list
items.
See below for more details.
* GPUProcess/graphics/DisplayListReaderHandle.cpp:
(WebKit::DisplayListReaderHandle::advance):
Update `advance` so that it just performs a single 8-byte atomic `exchangeSub` rather than grabbing a spinlock.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::submit):
(WebKit::RemoteRenderingBackend::nextDestinationImageBufferAfterApplyingDisplayLists):
Rename `applyDisplayListsFromHandle` to `nextDestinationImageBufferAfterApplyingDisplayLists`, and make it
return the current image buffer to use as the destination for replaying display list commands. This helper is
still responsible for replaying a display list item buffer until it has no more data (that is, until we
encounter a `MetaCommandChangeItemBuffer` item informing us of the next item buffer to consume).
(WebKit::RemoteRenderingBackend::wakeUpAndApplyDisplayList):
Adjust this to call `nextDestinationImageBufferAfterApplyingDisplayLists`.
(WebKit::RemoteRenderingBackend::decodeItem):
(WebKit::RemoteRenderingBackend::applyDisplayListsFromHandle): Deleted.
* GPUProcess/graphics/RemoteRenderingBackend.h:
* Shared/SharedDisplayListHandle.h:
Remove `SharedDisplayListHandle::Lock` altogether, now that we just have a single atomic `uint64_t` counter to
represent the number of unread bytes.
(WebKit::SharedDisplayListHandle::unreadBytes):
(WebKit::SharedDisplayListHandle::Lock::Lock): Deleted.
(WebKit::SharedDisplayListHandle::Lock::~Lock): Deleted.
* WebProcess/GPU/graphics/DisplayListWriterHandle.cpp:
(WebKit::DisplayListWriterHandle::advance):
Make this use the single atomic counter, instead of grabbing a spinlock.
(WebKit::DisplayListWriterHandle::moveWritableOffsetToStartIfPossible):
(WebKit::DisplayListWriterHandle::resetWritableOffsetIfPossible): Deleted.
Rename `resetWritableOffsetIfPossible` to the more accurate `moveWritableOffsetToStartIfPossible`.
* WebProcess/GPU/graphics/DisplayListWriterHandle.h:
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
(WebKit::RemoteImageBufferProxy::changeDestinationImageBuffer):
Add a helper method to append a `MetaCommandChangeDestinationImageBuffer` item to the display list, given the
identifier of this next item buffer.
(WebKit::RemoteImageBufferProxy::clearDisplayList):
Add a helper method to clear the display list, such that the `RemoteImageBufferProxy` will be asked for a new
buffer handle the next time we append an item.
(WebKit::RemoteImageBufferProxy::waitForDidFlushWithTimeout):
Make this hit the maximum `waitForAndDispatchImmediately` attempt count and "fail" only if the call to
`waitForAndDispatchImmediately` actually times out. This change is necessary to ensure that the scenario of
multiple image buffers simultaneously appending display list items works when all image buffers perform (non-
blocking) flushes simultaneously, followed by a sync-wait (blocking) flush. In this situation, the web process
may receive more than three `DidFlush` messages in rapid succession, causing it to hit this maximum limit of 3
instantly and avoid waiting for the expected `DidFlush` message to arrive.
(WebKit::RemoteImageBufferProxy::submitDisplayList): Deleted.
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::gpuProcessConnectionDidClose):
Reset state when the GPU process is terminated.
(WebKit::RemoteRenderingBackendProxy::willAppendItem):
Update the current destination image buffer (`m_currentDestinationImageBufferIdentifier`). If this identifier
changed, then we append a `MetaCommandChangeDestinationImageBuffer` item to the previous destination before
setting the new destination image buffer. We also need to clear the new destination image buffer's display list
before writing new items, to prevent it from overwriting shared display list item data.
(WebKit::RemoteRenderingBackendProxy::sendWakeupMessage):
(WebKit::RemoteRenderingBackendProxy::sendDeferredWakeupMessageIfNeeded):
(WebKit::RemoteRenderingBackendProxy::didAppendData):
Implement a new `ItemBufferWritingClient` hook by advancing the corresponding item buffer's unread count.
Additionally, schedule a wakeup message in the case where the unread count was zero when we advanced, and the
item buffer we've appended to is not "connected" to a previous item buffer by way of an item buffer change item.
Additionally, in the case where we don't schedule a new wakeup message, check whether or not there is a deferred
wakeup message; if there is, then decrement the remaining item count before we send the wakeup message, and send
the message if the count reaches 0.
(WebKit::RemoteRenderingBackendProxy::findReusableDisplayListHandle):
Pull logic to find a shared display list handle suitable for reuse into a separate helper method, and adjust it
so that if the most recently used display list handle has run out of available capacity, we move it to the end
of the reuse queue and only reuse the new first shared handle in the queue if its writable offset can be reset.
This ensures that the GPU process can always continue reading display list item data from the start of new item
buffers -- i.e., we don't end up with the web process writing out a stream of display list data like:
(Start) Bytes (16, 65520) in ItemBuffer[1]
Bytes (16, 65536) in ItemBuffer[2]
Bytes (65520, 65536) in ItemBuffer[1]
(WebKit::RemoteRenderingBackendProxy::createItemBuffer):
(WebKit::RemoteRenderingBackendProxy::submitDisplayList): Deleted.
(WebKit::RemoteRenderingBackendProxy::updateReusableHandles): Deleted.
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
2020-12-05 Fujii Hironori <Hironori.Fujii@sony.com>
Remove ENABLE_GRAPHICS_CONTEXT_GL by replacing it with ENABLE(WEBGL)
https://bugs.webkit.org/show_bug.cgi?id=219551
Reviewed by Kenneth Russell.
* CMakeLists.txt:
* UIProcess/API/glib/WebKitProtocolHandler.cpp:
(WebKit::WebKitProtocolHandler::handleGPU):
* UIProcess/gtk/HardwareAccelerationManager.cpp:
(WebKit::HardwareAccelerationManager::HardwareAccelerationManager):
2020-12-04 Said Abou-Hallawa <said@apple.com>
[GPU Process] Remove ConcreteShareableImageBuffer
https://bugs.webkit.org/show_bug.cgi?id=219563
Reviewed by Simon Fraser.
Delete this class since the only function it implements:
'createImageBufferBackendHandle()' is not used.
* Shared/ConcreteShareableImageBuffer.h: Removed.
* Shared/RemoteLayerTree/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::swapToValidFrontBuffer):
2020-12-04 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] Cancel image extraction when navigating
https://bugs.webkit.org/show_bug.cgi?id=219559
<rdar://problem/71997004>
Reviewed by Tim Horton.
Call `-_cancelImageExtraction` during mainframe navigation. See WebKitAdditions change for more information.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _webTouchEventsRecognized:]):
(-[WKContentView _didStartProvisionalLoadForMainFrame]):
2020-12-04 Jer Noble <jer.noble@apple.com>
[Cocoa] Adopt AVOutputDevice.allowsHeadTrackedSpatialAudio
https://bugs.webkit.org/show_bug.cgi?id=219503
Reviewed by Eric Carlson.
Adopt the "com.apple.avfoundation.allow-system-wide-context" entitlement on macOS for both the WebContent and GPU processes.
* Scripts/process-entitlements.sh:
2020-12-04 Jer Noble <jer.noble@apple.com>
[GPUP] Move MediaCapabilities into the GPU process
https://bugs.webkit.org/show_bug.cgi?id=219521
Reviewed by Eric Carlson.
Add a new pair of process supplements, RemoteMediaEngineConfigurationFactory and ..Proxy.
* DerivedSources-input.xcfilelist:
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::mediaEngineConfigurationFactoryProxy):
(WebKit::GPUConnectionToWebProcess::dispatchMessage):
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/media/RemoteMediaEngineConfigurationFactoryProxy.cpp: Added.
(WebKit::RemoteMediaEngineConfigurationFactoryProxy::RemoteMediaEngineConfigurationFactoryProxy):
(WebKit::createDecodingConfiguration):
(WebKit::RemoteMediaEngineConfigurationFactoryProxy::createEncodingConfiguration):
* GPUProcess/media/RemoteMediaEngineConfigurationFactoryProxy.h: Added.
* GPUProcess/media/RemoteMediaEngineConfigurationFactoryProxy.messages.in: Added.
* Sources.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/GPUProcessConnection.cpp:
(WebKit::GPUProcessConnection::mediaEngineConfigurationFactory):
* WebProcess/GPU/GPUProcessConnection.h:
* WebProcess/GPU/media/RemoteAudioSourceProviderManager.cpp:
* WebProcess/GPU/media/RemoteAudioSourceProviderManager.h:
* WebProcess/GPU/media/RemoteMediaEngineConfigurationFactory.cpp: Added.
(WebKit::RemoteMediaEngineConfigurationFactory::RemoteMediaEngineConfigurationFactory):
(WebKit::RemoteMediaEngineConfigurationFactory::registerFactory):
(WebKit::RemoteMediaEngineConfigurationFactory::supplementName):
(WebKit::RemoteMediaEngineConfigurationFactory::gpuProcessConnection):
(WebKit::RemoteMediaEngineConfigurationFactory::createDecodingConfiguration):
(WebKit::RemoteMediaEngineConfigurationFactory::createEncodingConfiguration):
* WebProcess/GPU/media/RemoteMediaEngineConfigurationFactory.h: Copied from Source/WTF/wtf/Algorithms.h.
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::setUseGPUProcessForMedia):
2020-12-04 Said Abou-Hallawa <said@apple.com>
[GPU Process] Clean up recreating the ImageBufferBackend because of GPU crashing
https://bugs.webkit.org/show_bug.cgi?id=219007
Reviewed by Simon Fraser.
There is no need to store the backend parameters in RemoteImageBufferProxy.
These parameters are now stored in the base class ImageBuffer before the
backend is created.
Remove the traits definition for RemoteImageBufferProxy types because
they are wrong. All ImageBuffers have valid renderingResourceIdentifier()
regardless whether they are remote or not. ImageBuffer::setBackend and
ImageBuffer::submitDisplayList() are now implemented in WebCore. So no
casting is needed.
* GPUProcess/graphics/RemoteImageBuffer.h:
(WebKit::RemoteImageBuffer::RemoteImageBuffer):
(WebKit::RemoteImageBuffer::submitDisplayList): Deleted.
submitDisplayList() is now a virtual method of ImageBuffer. So make it
private and ignore the return of DisplayList::Replayer::replay() since
the caller does not use it.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::didCreateImageBufferBackend):
(WebKit::RemoteRenderingBackend::applyDisplayListsFromHandle):
(WebKit::RemoteRenderingBackend::imageBufferBackendWasCreated): Deleted.
* GPUProcess/graphics/RemoteRenderingBackend.h:
Rename 'imageBufferBackendWasCreated()' to 'didCreateImageBufferBackend()'
to be consistent with 'didFlush()'.
* Shared/ConcreteShareableImageBuffer.h:
(WebKit::ConcreteShareableImageBuffer::create):
(WebKit::ConcreteShareableImageBuffer::ConcreteShareableImageBuffer): Deleted.
* Shared/RemoteLayerTree/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::encode const):
(WebKit::RemoteLayerBackingStore::swapToValidFrontBuffer):
Call the function with the new name and delete the FIXME comment.
* WebProcess/GPU/graphics/ImageBufferShareableBitmapBackend.cpp:
(WebKit::ImageBufferShareableBitmapBackend::create):
(WebKit::ImageBufferShareableBitmapBackend::ImageBufferShareableBitmapBackend):
(WebKit::ImageBufferShareableBitmapBackend::backendSize const):
* WebProcess/GPU/graphics/ImageBufferShareableBitmapBackend.h:
Use the new structure 'ImageBufferBackend::Parameters'. Implement the
virtual method 'backendSize()'.
* WebProcess/GPU/graphics/PlatformRemoteImageBufferProxy.h:
(isType): Deleted.
Delete the wrong traits.
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
(WebKit::RemoteImageBufferProxy::create):
(WebKit::RemoteImageBufferProxy::RemoteImageBufferProxy):
Use the new structure 'ImageBufferBackend::Parameters'.
(WebKit::RemoteImageBufferProxy::clearBackend): Deleted.
(WebKit::RemoteImageBufferProxy::createBackend): Deleted.
(WebKit::RemoteImageBufferProxy::didFlush): Deleted.
(WebKit::RemoteImageBufferProxy::size const): Deleted.
(WebKit::RemoteImageBufferProxy::renderingMode const): Deleted.
(WebKit::RemoteImageBufferProxy::colorSpace const): Deleted.
(WebKit::RemoteImageBufferProxy::pixelFormat const): Deleted.
These functions are now implemented in ConcreteImageBuffer.
(WebKit::RemoteImageBufferProxy::submitDisplayList): Deleted.
submitDisplayList() is now virtual.
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::reestablishGPUProcessConnection):
(WebKit::RemoteRenderingBackendProxy::waitForDidCreateImageBufferBackend):
(WebKit::RemoteRenderingBackendProxy::createImageBuffer):
(WebKit::RemoteRenderingBackendProxy::didCreateImageBufferBackend):
(WebKit::RemoteRenderingBackendProxy::didFlush):
(WebKit::recreateImageBuffer): Deleted.
(WebKit::RemoteRenderingBackendProxy::waitForImageBufferBackendWasCreated): Deleted.
(WebKit::RemoteRenderingBackendProxy::imageBufferBackendWasCreated): Deleted.
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
No need to cast the ImageBuffer to RemoteImageBufferProxy since all what
RemoteRenderingBackendProxy needs can be accessed through the base class.
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.messages.in:
Rename 'ImageBufferBackendWasCreated' to 'DidCreateImageBufferBackend'
to be consistent with 'DidFlush'. No need to pass the creation parameters
since they set in the ConcreteImageBuffer. No need to pass the backendSize
since it can be retrieved through the backend itself.
* WebProcess/GPU/graphics/cocoa/ImageBufferShareableIOSurfaceBackend.cpp:
(WebKit::ImageBufferShareableIOSurfaceBackend::create):
* WebProcess/GPU/graphics/cocoa/ImageBufferShareableIOSurfaceBackend.h:
* WebProcess/GPU/graphics/cocoa/ImageBufferShareableMappedIOSurfaceBackend.cpp:
(WebKit::ImageBufferShareableMappedIOSurfaceBackend::create):
* WebProcess/GPU/graphics/cocoa/ImageBufferShareableMappedIOSurfaceBackend.h:
Use the new structure 'ImageBufferBackend::Parameters'.
2020-12-04 John Wilander <wilander@apple.com>
PCM: Switch to 8+4 bits for source IDs and trigger values respectively
https://bugs.webkit.org/show_bug.cgi?id=219519
<rdar://problem/70470036>
Reviewed by Brent Fulgham.
We've received a lot of feedback saying increased entropy on the click side is more
important than the current 6 bits on the conversion side. Some of that conversation
is captured in https://github.com/privacycg/private-click-measurement/issues/28.
This patch switches from 6+6 bits to 8+4 bits. It also makes sure PCM Debug Mode
consistently logs on the LOG level except for real errors.
Existing layout tests and API tests were updated.
* NetworkProcess/PrivateClickMeasurementManager.cpp:
(WebKit::PrivateClickMeasurementManager::storeUnattributed):
(WebKit::PrivateClickMeasurementManager::attribute):
(WebKit::PrivateClickMeasurementManager::fireConversionRequest):
2020-12-04 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r270422.
https://bugs.webkit.org/show_bug.cgi?id=219554
Deleted an SPI that is still in use.
Reverted changeset:
"Introduce new download API"
https://bugs.webkit.org/show_bug.cgi?id=217747
https://trac.webkit.org/changeset/270422
2020-12-04 Brian Burg <bburg@apple.com>
Web Inspector: InspectorFrontendAPIDispatcher should not ignore all exceptions
https://bugs.webkit.org/show_bug.cgi?id=219378
Reviewed by Devin Rousso.
The underlying method used for frontend expression evaluations is
ScriptController::evaluateIgnoringExceptions. This method calls
evaluateInWorld and returns nullopt if an exception happens.
Switch to using evaluateInWorld directly and using the existing ValueOrException
type from in WebCore. Change our EvaluationResult type to use ValueOrException
in place of JSC::JSValue. ValueOrException is Expected<JSC::JSValue, ExceptionDetails>
so this is exposing more error information in addition to the JSC::JSValue.
* Platform/Logging.h: Add 'Inspector' log channel for WebKit.framework.
* WebProcess/Inspector/WebInspectorUIExtensionController.cpp:
(WebKit::WebInspectorUIExtensionController::WebInspectorUIExtensionController):
(WebKit::WebInspectorUIExtensionController::~WebInspectorUIExtensionController):
Remove unnecessary debugging code that was accidentally left in/commented out.
* WebProcess/Inspector/WebInspectorUIExtensionController.h:
(WebKit::WebInspectorUIExtensionController::parseInspectorExtensionErrorFromResult): Deleted.
(WebKit::WebInspectorUIExtensionController::parseInspectorExtensionErrorFromEvaluationResult): Added.
(WebKit::WebInspectorUIExtensionController::registerExtension):
(WebKit::WebInspectorUIExtensionController::unregisterExtension):
Adapt to using the new result type. Use the InspectorExtensionID type where possible.
2020-12-04 Kate Cheney <katherine_cheney@apple.com>
Create API to enable/disable text interaction gestures in WKWebView
https://bugs.webkit.org/show_bug.cgi?id=217784
<rdar://problem/63406241>
Reviewed by Wenson Hsieh.
Create a new API to disable text interaction gestures in a WKWebView.
We have SPI for this for iOS, but it could be useful to other WebKit
clients and on macOS. Add the API to WKPreferences, instead of
WKWebViewConfiguration, so it can be changed on the fly.
Deprecate the SPI.
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
* UIProcess/API/Cocoa/WKPreferences.h:
* UIProcess/API/Cocoa/WKPreferences.mm:
(-[WKPreferences encodeWithCoder:]):
(-[WKPreferences initWithCoder:]):
(-[WKPreferences textInteractionEnabled]):
(-[WKPreferences setTextInteractionEnabled:]):
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration init]):
(-[WKWebViewConfiguration encodeWithCoder:]):
(-[WKWebViewConfiguration initWithCoder:]):
(-[WKWebViewConfiguration copyWithZone:]):
(-[WKWebViewConfiguration _textInteractionEnabled]):
(-[WKWebViewConfiguration _setTextInteractionEnabled:]):
* UIProcess/API/Cocoa/WKWebViewConfigurationPrivate.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView hasSelectablePositionAtPoint:]):
(-[WKContentView pointIsNearMarkedText:]):
(-[WKContentView textInteractionGesture:shouldBeginAtPoint:]):
(-[WKContentView _cascadeInteractionTintColor]):
(-[WKContentView selectForWebView:]):
(-[WKContentView selectAllForWebView:]):
Add check for textInteractionEnabled in these select*
functions because disabling text interaction should disable selection
in WKWebView.
* WebProcess/WebPage/WebPage.cpp:
Update the textInteractionEnabled value in the page when
preferences are updated.
(WebKit::WebPage::updatePreferences):
2020-12-04 Adam Roben <aroben@apple.com>
More FALLBACK_PLATFORM adoption
https://bugs.webkit.org/show_bug.cgi?id=219545
Reviewed by Tim Horton.
* Configurations/SDKVariant.xcconfig:
WK_EMPTY_$(THIS_IS_NOT_EMPTY) evaluates to the empty string, not to
NO.
2020-12-04 Said Abou-Hallawa <said@apple.com>
[GPU Process] Share the NativeImage with GPU Process through a ShareableBitmap
https://bugs.webkit.org/show_bug.cgi?id=218934
Reviewed by Tim Horton.
Instead of relying on the IPC coders to transfer the NativeImage to the
GPUP, we are going to move this logic to RemoteResourceCacheProxy and
RemoteRenderingBackend. The NativeImage will be transfered as a handle of
ShareableBitamp and a renderingResourceIdentifier.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::cacheNativeImage):
* GPUProcess/graphics/RemoteRenderingBackend.h:
* GPUProcess/graphics/RemoteRenderingBackend.messages.in:
* Shared/ShareableBitmap.h:
(WebKit::ShareableBitmap::createPlatformImage):
Add this helper function and avoid adding directive code in many places.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<Ref<NativeImage>>::encode): Deleted.
(IPC::ArgumentCoder<Ref<NativeImage>>::decode): Deleted.
* Shared/WebCoreArgumentCoders.h:
* WebProcess/GPU/graphics/ImageBufferShareableBitmapBackend.cpp:
(WebKit::ImageBufferShareableBitmapBackend::copyNativeImage const):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::cacheNativeImage):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
* WebProcess/GPU/graphics/RemoteResourceCacheProxy.cpp:
(WebKit::createShareableBitmapFromNativeImage):
(WebKit::RemoteResourceCacheProxy::cacheNativeImage):
2020-12-04 Per Arne Vollan <pvollan@apple.com>
[macOS] Deny mach-lookup to the service 'com.apple.print.normalizerd' in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=219541
<rdar://problem/68878604>
Reviewed by Darin Adler.
On macOS, deny mach-lookup to the service 'com.apple.print.normalizerd' in the WebContent process, since there are no
reports of this being used.
No new tests, since the result will depend on the macOS version.
* WebProcess/com.apple.WebProcess.sb.in:
2020-12-04 Devin Rousso <drousso@apple.com>
[iOS] Provide a context menu action to perform accessibility image extraction
https://bugs.webkit.org/show_bug.cgi?id=219524
<rdar://problem/69969613>
Reviewed by Wenson Hsieh.
* UIProcess/API/Cocoa/_WKElementAction.h:
* UIProcess/API/Cocoa/_WKElementAction.mm:
(+[_WKElementAction _elementActionWithType:customTitle:assistant:]):
(+[_WKElementAction imageForElementActionType:]):
(elementActionTypeToUIActionIdentifier):
(uiActionIdentifierToElementActionType):
Add a new `_WKElementActionTypeImageExtraction` enum type with a default icon and title.
* UIProcess/ios/WKActionSheetAssistant.h:
* UIProcess/ios/WKActionSheetAssistant.mm:
(-[WKActionSheetAssistant defaultActionsForLinkSheet:]):
(-[WKActionSheetAssistant defaultActionsForImageSheet:]):
(-[WKActionSheetAssistant handleElementActionWithType:element:needsInteraction:]):
If the `WKActionSheetAssistantDelegate` allows it, add a `_WKElementActionTypeImageExtraction`
action for images (including those inside links). Call out to the `WKActionSheetAssistantDelegate`
when handling the action.
2020-12-04 Peng Liu <peng.liu6@apple.com>
[MSE] Move track buffer management from SourceBuffer to SourceBufferPrivate
https://bugs.webkit.org/show_bug.cgi?id=219332
Reviewed by Daniel Bates.
Update `SourceBufferPrivateRemote` and `RemoteSourceBufferProxy` because of
the changes on `SourceBufferPrivate` and `SourceBufferPrivateClient`.
* GPUProcess/media/RemoteSourceBufferProxy.cpp:
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateAppendError):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateDurationChanged):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateDidParseSample):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateDidDropSample):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateStreamEndedWithDecodeError):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateHasAudio const):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateHasVideo const):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateDidReceiveSample): Deleted.
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateReenqueSamples): Deleted.
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateDidBecomeReadyForMoreSamples): Deleted.
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateFastSeekTimeForMediaTime): Deleted.
* GPUProcess/media/RemoteSourceBufferProxy.h:
* WebProcess/GPU/media/SourceBufferPrivateRemote.cpp:
(WebKit::SourceBufferPrivateRemote::setClient): Deleted.
* WebProcess/GPU/media/SourceBufferPrivateRemote.h:
2020-12-04 Kimmo Kinnunen <kkinnunen@apple.com>
GPU Process WindowServer blocking prevents CGL
https://bugs.webkit.org/show_bug.cgi?id=219431
<rdar://problem/71893390>
Reviewed by Per Arne Vollan.
Mark the GPU process as NSSupportsAutomaticGraphicsSwitching.
This allows CGL to work in cases when WindowServer connection
is blocked.
No new tests, makes the future WebGL in GPU process work.
* GPUProcess/EntryPoint/Cocoa/XPCService/GPUService/Info-OSX.plist:
2020-12-03 Simon Fraser <simon.fraser@apple.com>
Only the first wheel event in a gesture should be cancelable
https://bugs.webkit.org/show_bug.cgi?id=218764
<rdar://problem/71248946>
Reviewed by Tim Horton.
In EventDispatcher::wheelEvent(), all wheel events now bounce through the scrolling
thread, even those destined for main thread scrolling. This allows the scrolling thread
to wait on a condition for the event to come back to the scrolling thread via
handleWheelEventAfterMainThread(), since we have to know whether content called
preventDefault() on the first event before sending subsequent events.
* WebProcess/WebPage/EventDispatcher.cpp:
(WebKit::EventDispatcher::wheelEvent):
2020-12-03 Alex Christensen <achristensen@webkit.org>
Introduce new download API
https://bugs.webkit.org/show_bug.cgi?id=217747
Reviewed by Brady Eidson.
Safari currently uses _WKDownload, which has evolved strangely over the last decade.
In order to make a nicer interface, we need to learn from those lessons and take a step back.
I did that, and here's what I came up with: WKDownload and WKDownloadDelegate!
Notable changes include:
1. The delegate now lives on the download object instead of the process pool.
2. WKDownload does not conform to NSCopying because we have NSMapTable instead of NSDictionary.
3. publishProgressAtURL is gone. That will be reimplemented in the UI process of the client that used it.
4. cancel has a completion handler to get the resume data instead of waiting for didCancel then getting it from the download object.
5. didFailWithError also gives you the resume data, so there is no way to get it from the download object. That's more about a fail/cancel
event than a property of the download.
6. wasUserInitiated is gone. Instead, WKNavigationDelegate has a callback that links a navigation action (where _isUserInitiated is exposed) to the download.
7. redirectChain is also gone. That can also be gotten from the link to the navigation and the download delegate's redirect callback.
8. _downloadDidStart is gone. Instead, we have completion handlers that expose a WKDownload once it's started. There's no need to
reference the download object before then anyways. Note: that's when the download "starts" which is before it receives the response,
so there will be no server delay in the difference between download object availabilities.
9. didReceiveServerRedirectToURL now gives you the option of continuing or cancelling. This reflects the amount of control the
WKNavigationDelegate has.
10. decideDestinationWithSuggestedFilename and didReceiveResponse have been merged, since they actually happen at the same time and
NSURLResponse has suggestedFilename API. allowOverwrite is also gone. It wasn't used, and not really needed.
11. didCreateDestination is gone. That callback time wasn't really necessary. If you need to know when it's created, wait until the first
didWriteData callback. The destination has definitely been created by then.
12. shouldDecodeSourceDataOfMIMEType is gone. It's actually not called since we adopted NSURLSession, and we should remove the related dead code.
13. originatingFrame is gone. On _WKDownload it was actually a non-null WKFrameInfo that often contained no information. When information was there,
the information access has been replaced by didBecomeDownload which links the WKNavigationResponse which has frame info on it.
I wrote API tests for everything I could think of.
* NetworkProcess/Downloads/DownloadManager.cpp:
(WebKit::DownloadManager::resumeDownload):
* NetworkProcess/Downloads/DownloadManager.h:
* NetworkProcess/Downloads/cocoa/DownloadCocoa.mm:
(WebKit::Download::resume):
(WebKit::Download::platformCancelNetworkLoad):
* NetworkProcess/NetworkDataTaskBlob.cpp:
(WebKit::NetworkDataTaskBlob::suggestedFilename const):
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::resumeDownload):
(WebKit::NetworkProcess::findPendingDownloadLocation):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(-[WKNetworkSessionDelegate URLSession:task:didCompleteWithError:]):
* Scripts/webkit/messages.py:
* Shared/API/Cocoa/WebKit.h:
* Shared/API/c/WKSharedAPICast.h:
(WebKit::toAPI):
* Shared/Cocoa/APIObject.mm:
(API::Object::newObject):
* Sources.txt:
* SourcesCocoa.txt:
* UIProcess/API/APIDownloadClient.h:
(API::DownloadClient::legacyDidStart):
(API::DownloadClient::didFail):
(API::DownloadClient::didStart): Deleted.
* UIProcess/API/APINavigationAction.h:
* UIProcess/API/APINavigationClient.h:
(API::NavigationClient::navigationResponseDidBecomeDownload):
(API::NavigationClient::navigationActionDidBecomeDownload):
(API::NavigationClient::contextMenuDidCreateDownload):
* UIProcess/API/APIPolicyClient.h:
(API::PolicyClient::decidePolicyForNavigationAction):
(API::PolicyClient::decidePolicyForNewWindowAction):
* UIProcess/API/C/WKContext.cpp:
(WKContextSetDownloadClient): Deleted.
* UIProcess/API/C/WKContext.h:
* UIProcess/API/C/WKContextDownloadClient.h: Removed.
* UIProcess/API/C/WKDownload.cpp: Removed.
* UIProcess/API/C/WKDownloadClient.h: Added.
* UIProcess/API/C/WKDownloadRef.cpp: Added.
(WKDownloadGetTypeID):
(WKDownloadCopyRequest):
(WKDownloadCancel):
(WKDownloadGetOriginatingPage):
(WKDownloadGetWasUserInitiated):
(WKDownloadSetClient):
* UIProcess/API/C/WKDownloadRef.h: Renamed from Source/WebKit/UIProcess/API/C/WKDownload.h.
* UIProcess/API/C/WKNavigationActionRef.cpp:
(WKNavigationActionGetDownloadAttribute):
(WKNavigationActionShouldPerformDownload): Deleted.
* UIProcess/API/C/WKNavigationActionRef.h:
* UIProcess/API/C/WKPage.cpp:
(WKPageSetPagePolicyClient):
(WKPageSetPageNavigationClient):
* UIProcess/API/C/WKPageNavigationClient.h:
* UIProcess/API/Cocoa/WKDownload.h: Copied from Source/WebKit/UIProcess/API/Cocoa/_WKDownload.h.
* UIProcess/API/Cocoa/WKDownload.mm: Added.
(-[WKDownload cancel:]):
(-[WKDownload originalRequest]):
(-[WKDownload delegate]):
(-[WKDownload setDelegate:]):
(-[WKDownload dealloc]):
(-[WKDownload _apiObject]):
* UIProcess/API/Cocoa/WKDownloadDelegate.h: Added.
* UIProcess/API/Cocoa/WKDownloadInternal.h: Copied from Source/WebKit/UIProcess/API/Cocoa/_WKDownloadInternal.h.
* UIProcess/API/Cocoa/WKNavigationAction.h:
* UIProcess/API/Cocoa/WKNavigationAction.mm:
(-[WKNavigationAction downloadAttribute]):
(-[WKNavigationAction _shouldPerformDownload]):
* UIProcess/API/Cocoa/WKNavigationActionPrivate.h:
* UIProcess/API/Cocoa/WKNavigationDelegate.h:
* UIProcess/API/Cocoa/WKNavigationDelegatePrivate.h:
* UIProcess/API/Cocoa/WKProcessPool.mm:
(-[WKProcessPool _downloadURLRequest:websiteDataStore:originatingWebView:]):
(-[WKProcessPool _resumeDownloadFromData:websiteDataStore:path:originatingWebView:]):
* UIProcess/API/Cocoa/WKProcessPoolPrivate.h:
* UIProcess/API/Cocoa/WKWebView.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView downloadRequest:completionHandler:]):
(-[WKWebView resumeDownloadWithData:completionHandler:]):
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/API/Cocoa/_WKDownload.h:
* UIProcess/API/Cocoa/_WKDownload.mm:
(-[_WKDownload initWithDownload2:]):
(+[_WKDownload downloadWithDownload:]):
(-[_WKDownload cancel]):
(-[_WKDownload publishProgressAtURL:]):
(-[_WKDownload request]):
(-[_WKDownload originatingWebView]):
(-[_WKDownload redirectChain]):
(-[_WKDownload wasUserInitiated]):
(-[_WKDownload resumeData]):
(-[_WKDownload originatingFrame]):
(-[_WKDownload _apiObject]):
(-[_WKDownload dealloc]): Deleted.
* UIProcess/API/Cocoa/_WKDownloadDelegate.h:
* UIProcess/API/Cocoa/_WKDownloadInternal.h:
* UIProcess/API/glib/WebKitDownloadClient.cpp:
* UIProcess/Cocoa/LegacyDownloadClient.h:
* UIProcess/Cocoa/LegacyDownloadClient.mm:
(WebKit::LegacyDownloadClient::legacyDidStart):
(WebKit::LegacyDownloadClient::didReceiveResponse):
(WebKit::LegacyDownloadClient::didReceiveData):
(WebKit::LegacyDownloadClient::didReceiveAuthenticationChallenge):
(WebKit::LegacyDownloadClient::didCreateDestination):
(WebKit::LegacyDownloadClient::processDidCrash):
(WebKit::LegacyDownloadClient::decideDestinationWithSuggestedFilename):
(WebKit::LegacyDownloadClient::didFinish):
(WebKit::LegacyDownloadClient::didFail):
(WebKit::LegacyDownloadClient::legacyDidCancel):
(WebKit::LegacyDownloadClient::willSendRequest):
(WebKit::LegacyDownloadClient::didStart): Deleted.
* UIProcess/Cocoa/NavigationState.h:
* UIProcess/Cocoa/NavigationState.mm:
(WebKit::NavigationState::setNavigationDelegate):
(WebKit::NavigationState::NavigationClient::decidePolicyForNavigationAction):
(WebKit::NavigationState::NavigationClient::decidePolicyForNavigationResponse):
(WebKit::NavigationState::NavigationClient::navigationActionDidBecomeDownload):
(WebKit::NavigationState::NavigationClient::navigationResponseDidBecomeDownload):
(WebKit::NavigationState::NavigationClient::contextMenuDidCreateDownload):
* UIProcess/Cocoa/WebProcessProxyCocoa.mm:
* UIProcess/Downloads/DownloadProxy.cpp:
(WebKit::DownloadProxy::~DownloadProxy):
(WebKit::DownloadProxy::didStart):
(WebKit::DownloadProxy::decideDestinationWithSuggestedFilename):
(WebKit::DownloadProxy::didFail):
(WebKit::DownloadProxy::setClient):
* UIProcess/Downloads/DownloadProxy.h:
(WebKit::DownloadProxy::setDidStartCallback):
(WebKit::DownloadProxy::setSuggestedFilename):
* UIProcess/Network/NetworkProcessProxy.cpp:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::receivedNavigationPolicyDecision):
(WebKit::WebPageProxy::receivedPolicyDecision):
(WebKit::WebPageProxy::decidePolicyForNavigationAction):
(WebKit::WebPageProxy::decidePolicyForNewWindowAction):
(WebKit::WebPageProxy::decidePolicyForResponseShared):
(WebKit::WebPageProxy::resumeDownload):
(WebKit::WebPageProxy::downloadRequest):
(WebKit::WebPageProxy::contextMenuItemSelected):
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::receivedPolicyDecision):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::resumeDownload):
* UIProcess/WebProcessPool.h:
* WebKit.xcodeproj/project.pbxproj:
2020-12-03 Adam Roben <aroben@apple.com>
Adopt FALLBACK_PLATFORM
https://bugs.webkit.org/show_bug.cgi?id=219504
Reviewed by Tim Horton.
* Configurations/SDKVariant.xcconfig:
* Scripts/check-xcfilelists.sh:
Use FALLBACK_PLATFORM it if it's defined, otherwise use PLATFORM_NAME
as before.
2020-12-03 Chris Dumez <cdumez@apple.com>
Drop unimplemented suspension functions on GPUConnectionToWebProcess and WebAuthnConnectionToWebProcess
https://bugs.webkit.org/show_bug.cgi?id=219513
Reviewed by Tim Horton.
These functions are never called and have no implementation. This is just bad copy/paste from
NetworkConnectionToWebProcess.
* GPUProcess/GPUConnectionToWebProcess.h:
* WebAuthnProcess/WebAuthnConnectionToWebProcess.h:
2020-12-03 Brent Fulgham <bfulgham@apple.com>
[macOS] Only extend access to the AppleSNBFBUserClient IOKit class if the GPU Process is not used
https://bugs.webkit.org/show_bug.cgi?id=219014
<rdar://problem/70463873>
Reviewed by Per Arne Vollan.
Instead of globally extending access to the AppleSNBFBUserClient IOKit class,
only extend it when the GPU process is not in use.
* UIProcess/WebPageProxy.cpp:
(WebKit::gpuIOKitClasses): Add 'AppleSNBFBUserClient' as a dynamically-extended
IOKit class.
* WebProcess/com.apple.WebProcess.sb.in: Only allow 'AppleSNBFBUserClient' if it
was dynamically extended.
2020-12-03 Said Abou-Hallawa <said@apple.com>
[GPU Process] Disconnect NativeImages from RemoteResourceCacheProxy when RemoteRenderingBackendProxy is destroyed
https://bugs.webkit.org/show_bug.cgi?id=219417
Reviewed by Tim Horton.
If the WebPage is destroyed before destroying the CachedImages, a crash
may happen. The NativeImage will try to release itself from its observer,
which is RemoteResourceCacheProxy, after it has been freed.
* WebProcess/GPU/graphics/RemoteResourceCacheProxy.cpp:
(WebKit::RemoteResourceCacheProxy::~RemoteResourceCacheProxy):
(WebKit::RemoteResourceCacheProxy::cacheNativeImage):
* WebProcess/GPU/graphics/RemoteResourceCacheProxy.h:
2020-12-03 Chris Dumez <cdumez@apple.com>
Bad IPC from the WebProcess should not terminate the GPUProcess
https://bugs.webkit.org/show_bug.cgi?id=219511
Reviewed by Simon Fraser.
Bad IPC from the WebProcess should not terminate the GPUProcess. The GPUProcess is shared by all
WebProcesses and it is not acceptable for a single bad WebProcess to negatively impact other
WebProcesses. Instead, we should terminate the bad WebProcess, like the NetworkProcess already
does on bad IPC.
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::didReceiveInvalidMessage):
* Shared/ProcessTerminationReason.h:
* UIProcess/API/C/WKAPICast.h:
(WebKit::toAPI):
* UIProcess/Cocoa/NavigationState.mm:
(WebKit::wkProcessTerminationReason):
* UIProcess/GPU/GPUProcessProxy.cpp:
(WebKit::GPUProcessProxy::terminateWebProcess):
* UIProcess/GPU/GPUProcessProxy.h:
* UIProcess/GPU/GPUProcessProxy.messages.in:
* UIProcess/WebPageProxy.cpp:
(WebKit::shouldReloadAfterProcessTermination):
2020-12-03 Alex Christensen <achristensen@webkit.org>
Serialize NFA to disk before converting it to a DFA when compiling a WKContentRuleList
https://bugs.webkit.org/show_bug.cgi?id=219452
Reviewed by Geoffrey Garen.
* UIProcess/API/Cocoa/WKContentRuleListStore.mm:
(-[WKContentRuleListStore _compileContentRuleListForIdentifier:encodedContentRuleList:completionHandler:]):
* UIProcess/API/Cocoa/WKContentRuleListStorePrivate.h:
Remove NS_RELEASES_ARGUMENT because it was incorrect and unnecessary because the WTF::String is copied to a background thread.
2020-12-03 Chris Dumez <cdumez@apple.com>
Make sure the GPUConnectionToWebProcess gets destroyed when the connection to the WebProcess gets severed
https://bugs.webkit.org/show_bug.cgi?id=219508
Reviewed by Geoffrey Garen.
Make sure the GPUConnectionToWebProcess gets destroyed when the connection to the WebProcess gets severed
(WebProcess exited normally or crashed). This is similar to what we do in the NetworkProcess for
NetworkConnectionToWebProcess.
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::didClose):
2020-12-03 Tim Horton <timothy_horton@apple.com>
GPU Process: Sandbox violations under IOSurface::maximumSize in the Web Content process
https://bugs.webkit.org/show_bug.cgi?id=219484
<rdar://problem/71603808>
Reviewed by Ryosuke Niwa.
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
Fetch the maximum allowed size of an IOSurface on the current hardware
in the UI process, and push it to the Web Content process on creation.
2020-12-03 Martin Robinson <mrobinson@igalia.com>
Move code from AxisScrollSnapOffsets to ScrollSnapOffsetsInfo
https://bugs.webkit.org/show_bug.cgi?id=219345
Reviewed by Daniel Bates.
* UIProcess/RemoteLayerTree/ios/RemoteScrollingCoordinatorProxyIOS.mm: Remove AxisScrollSnapOffsets.h include.
* UIProcess/RemoteLayerTree/ios/ScrollingTreeScrollingNodeDelegateIOS.mm: Ditto.
2020-12-02 Tim Horton <timothy_horton@apple.com>
Many different assertion failures on the GPU process bot after r270366
https://bugs.webkit.org/show_bug.cgi?id=219467
Reviewed by Simon Fraser.
* WebProcess/WebPage/DrawingArea.cpp:
(WebKit::DrawingArea::supportsGPUProcessRendering):
* WebProcess/WebPage/DrawingArea.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::WebPage):
(WebKit::WebPage::updatePreferences):
* WebProcess/WebPage/WebPage.h:
Disable DOM rendering in the GPU process if the DrawingArea doesn't support it.
Currently only RemoteLayerTreeDrawingArea does.
2020-12-02 Tim Horton <timothy_horton@apple.com>
GPU Process: Temporarily disable Web Fonts when DOM rendering in the GPU process is enabled
https://bugs.webkit.org/show_bug.cgi?id=219479
Reviewed by Simon Fraser.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences):
Web fonts don't work in the GPU process yet, so disable them.
Unstyled text is better than no text.
2020-12-02 Brent Fulgham <bfulgham@apple.com>
[macOS] WebContent sandbox; remove AppleIntelMEUserClient
https://bugs.webkit.org/show_bug.cgi?id=219012
<rdar://problem/70462796>
Reviewed by Eric Carlson.
Instead of globally extending access to the AppleIntelMEUserClient IOKit class,
only extend it when the GPU process is not in use.
* UIProcess/WebPageProxy.cpp:
(WebKit::gpuIOKitClasses): Add 'AppleIntelMEUserClient' as a dynamically-extended
IOKit class.
* WebProcess/com.apple.WebProcess.sb.in: Only allow 'AppleIntelMEUserClient' if it
was dynamically extended.
2020-12-02 Wenson Hsieh <wenson_hsieh@apple.com>
Unreviewed, fix the iOS build after r270362
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView createHighlightInCurrentGroupWithRange:]):
(-[WKContentView createHighlightInNewGroupWithRange:]):
2020-12-02 Megan Gardner <megan_gardner@apple.com>
Create and draw app highlights
https://bugs.webkit.org/show_bug.cgi?id=219365
Reviewed by Wenson Hsieh.
Handle the creation of App Highlights from selections when the context menu item is selected.
This patch will add the highlights to the Document's app Highlight register, and handle drawing
them in InlineTextBox. Later patches will handle the persistent storage and repopulation of app
highlights on launch or reload of a page.
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::createAppHighlightInSelectedRange):
* UIProcess/PageClient.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::contextMenuItemSelected):
* UIProcess/WebPageProxy.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView createHighlightInCurrentGroupWithRange:]):
(-[WKContentView createHighlightInNewGroupWithRange:]):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::createAppHighlightInSelectedRange):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-12-02 Brent Fulgham <bfulgham@apple.com>
[iOS] Silence sandbox warning for unneeded sysctl-read of "hw.tbfrequency_compat"
https://bugs.webkit.org/show_bug.cgi?id=219414
<rdar://problem/71740719>
Reviewed by Per Arne Vollan.
Silence a spurious log generated when our sandbox denies access to the unused sysctl
"hw.tbfrequency_compat". I have confirmed with the framework that attempts to read this
value, and they confirm they don't need the value, and actually are not using the result
anywhere in their framework.
Let's silence this report.
* Resources/SandboxProfiles/ios/com.apple.WebKit.GPU.sb:
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-12-02 Wenson Hsieh <wenson_hsieh@apple.com>
Entire image elements are sometimes selected after ending a image extraction interaction
https://bugs.webkit.org/show_bug.cgi?id=219435
<rdar://problem/71897557>
Reviewed by Megan Gardner.
If the image extraction interaction is currently active, regular text interactions may need to defer to these
image extraction interactions. See WebKitAdditions changes for more details.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView hasSelectablePositionAtPoint:]):
(-[WKContentView textInteractionGesture:shouldBeginAtPoint:]):
2020-12-02 Wenson Hsieh <wenson_hsieh@apple.com>
Context menu should be shown after a long timeout following image extraction
https://bugs.webkit.org/show_bug.cgi?id=219415
<rdar://problem/71872600>
Reviewed by Andy Estes.
Ensures that the context menu can still be shown when triggering an image extraction gesture. To do this, we
add a new long press gesture recognizer with a much longer delay. See WebKitAdditions patch for more details.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView gestureRecognizer:shouldRecognizeSimultaneouslyWithGestureRecognizer:]):
(-[WKContentView _doAfterPendingImageExtraction:]):
(-[WKContentView _invokeAllActionsToPerformAfterPendingImageExtraction]):
Drive-by fix: move a couple of methods out of WebKitAdditions and into non-internal source.
2020-12-01 Tim Horton <timothy_horton@apple.com>
GPU Process: IOSurfaces should not be mapped into the Web Content Process
https://bugs.webkit.org/show_bug.cgi?id=219368
Reviewed by Said Abou-Hallawa.
Since the Web Content process sandbox will eventually not have access to IOKit,
it will not be able to map/lock/use IOSurfaces. Thankfully, all it really needs
to be able to do is carry a reference to the surface from the GPU process to the
UI process (where it is mapped and applied as layer contents).
* GPUProcess/graphics/PlatformRemoteImageBuffer.h:
* Shared/RemoteLayerTree/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::encode const):
(WebKit::RemoteLayerBackingStore::swapToValidFrontBuffer):
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/graphics/PlatformImageBufferShareableBackend.h:
* WebProcess/GPU/graphics/PlatformRemoteImageBufferProxy.h:
(isType):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::reestablishGPUProcessConnection):
(WebKit::RemoteRenderingBackendProxy::createImageBuffer):
(WebKit::RemoteRenderingBackendProxy::imageBufferBackendWasCreated):
(WebKit::RemoteRenderingBackendProxy::didFlush):
* WebProcess/GPU/graphics/cocoa/ImageBufferShareableIOSurfaceBackend.cpp:
(WebKit::ImageBufferShareableIOSurfaceBackend::create):
(WebKit::ImageBufferShareableIOSurfaceBackend::createImageBufferBackendHandle const):
(WebKit::ImageBufferShareableIOSurfaceBackend::context const):
(WebKit::ImageBufferShareableIOSurfaceBackend::copyNativeImage const):
(WebKit::ImageBufferShareableIOSurfaceBackend::copyImage const):
(WebKit::ImageBufferShareableIOSurfaceBackend::draw):
(WebKit::ImageBufferShareableIOSurfaceBackend::drawPattern):
(WebKit::ImageBufferShareableIOSurfaceBackend::toDataURL const):
(WebKit::ImageBufferShareableIOSurfaceBackend::toData const):
(WebKit::ImageBufferShareableIOSurfaceBackend::toBGRAData const):
(WebKit::ImageBufferShareableIOSurfaceBackend::getImageData const):
(WebKit::ImageBufferShareableIOSurfaceBackend::putImageData):
* WebProcess/GPU/graphics/cocoa/ImageBufferShareableIOSurfaceBackend.h:
* WebProcess/GPU/graphics/cocoa/ImageBufferShareableMappedIOSurfaceBackend.cpp: Copied from Source/WebKit/WebProcess/GPU/graphics/cocoa/ImageBufferShareableIOSurfaceBackend.cpp.
(WebKit::ImageBufferShareableMappedIOSurfaceBackend::create):
(WebKit::ImageBufferShareableMappedIOSurfaceBackend::createImageBufferBackendHandle const):
* WebProcess/GPU/graphics/cocoa/ImageBufferShareableMappedIOSurfaceBackend.h: Copied from Source/WebKit/WebProcess/GPU/graphics/cocoa/ImageBufferShareableIOSurfaceBackend.h.
Rename ImageBufferShareableIOSurfaceBackend to ImageBufferShareableMappedIOSurfaceBackend.
Readd ImageBufferShareableIOSurfaceBackend, the unmapped variant. The mapped variant is only used:
- in the GPU process
- in the Web Content process when DOM rendering in the GPU process is disabled
The unmapped variant cannot perform most ImageBufferBackend duties, except creating an ImageBufferBackendHandle,
which it does by cloning the existing handle. It explicitly does *not* map the IOSurface in the
process, and is meant for use in Web Content processes that do not have access to IOSurface.
2020-12-01 Peng Liu <peng.liu6@apple.com>
[Media In GPU Process][MSE] Add the support to forward initialization segment from the GPU Process to Web processes
https://bugs.webkit.org/show_bug.cgi?id=219227
Reviewed by Eric Carlson.
This patch adds the support to forward `sourceBufferPrivateDidReceiveInitializationSegment()`
callback from the source buffer parser running in the GPU process to the `SourceBufferPrivateRemote`
running in the Web process, which in turn provides the initialization segment to `SourceBuffer`.
No change in behavior.
* CMakeLists.txt:
* DerivedSources-input.xcfilelist:
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* GPUProcess/media/InitializationSegmentInfo.h: Added.
(WebKit::InitializationSegmentInfo::TrackInformation::encode const):
(WebKit::InitializationSegmentInfo::TrackInformation::decode):
(WebKit::InitializationSegmentInfo::encode const):
(WebKit::InitializationSegmentInfo::decode):
* GPUProcess/media/RemoteAudioTrackProxy.cpp:
(WebKit::RemoteAudioTrackProxy::RemoteAudioTrackProxy):
(WebKit::RemoteAudioTrackProxy::~RemoteAudioTrackProxy):
(WebKit::RemoteAudioTrackProxy::configurationChanged):
(WebKit::RemoteAudioTrackProxy::willRemove):
* GPUProcess/media/RemoteAudioTrackProxy.h:
* GPUProcess/media/RemoteAudioTrackProxy.messages.in: Copied from Source/WebKit/WebProcess/GPU/media/SourceBufferPrivateRemote.messages.in.
* GPUProcess/media/RemoteMediaDescription.h: Added.
(WebKit::MediaDescriptionInfo::encode const):
(WebKit::MediaDescriptionInfo::decode):
(WebKit::RemoteMediaDescription::RemoteMediaDescription):
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
(WebKit::RemoteMediaPlayerProxy::loadMediaSource):
(WebKit::RemoteMediaPlayerProxy::addRemoteAudioTrackProxy):
(WebKit::RemoteMediaPlayerProxy::addRemoteVideoTrackProxy):
(WebKit::RemoteMediaPlayerProxy::addRemoteTextTrackProxy):
(WebKit::RemoteMediaPlayerProxy::mediaPlayerDidAddAudioTrack):
(WebKit::RemoteMediaPlayerProxy::mediaPlayerDidRemoveAudioTrack):
(WebKit::RemoteMediaPlayerProxy::mediaPlayerDidAddVideoTrack):
(WebKit::RemoteMediaPlayerProxy::mediaPlayerDidRemoveVideoTrack):
(WebKit::RemoteMediaPlayerProxy::mediaPlayerDidAddTextTrack):
(WebKit::RemoteMediaPlayerProxy::mediaPlayerDidRemoveTextTrack):
(WebKit::RemoteMediaPlayerProxy::audioTrackSetEnabled): Deleted.
(WebKit::RemoteMediaPlayerProxy::videoTrackSetSelected): Deleted.
(WebKit::RemoteMediaPlayerProxy::textTrackSetMode): Deleted.
* GPUProcess/media/RemoteMediaPlayerProxy.h:
* GPUProcess/media/RemoteMediaPlayerProxy.messages.in:
* GPUProcess/media/RemoteMediaSourceProxy.cpp:
(WebKit::RemoteMediaSourceProxy::RemoteMediaSourceProxy):
(WebKit::RemoteMediaSourceProxy::addSourceBuffer):
* GPUProcess/media/RemoteMediaSourceProxy.h:
* GPUProcess/media/RemoteSourceBufferProxy.cpp:
(WebKit::RemoteSourceBufferProxy::create):
(WebKit::RemoteSourceBufferProxy::RemoteSourceBufferProxy):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateDidReceiveInitializationSegment):
* GPUProcess/media/RemoteSourceBufferProxy.h:
* GPUProcess/media/RemoteTextTrackProxy.cpp:
(WebKit::RemoteTextTrackProxy::RemoteTextTrackProxy):
(WebKit::RemoteTextTrackProxy::~RemoteTextTrackProxy):
(WebKit::RemoteTextTrackProxy::configurationChanged):
(WebKit::RemoteTextTrackProxy::willRemove):
(WebKit::RemoteTextTrackProxy::addDataCue):
(WebKit::RemoteTextTrackProxy::updateDataCue):
(WebKit::RemoteTextTrackProxy::removeDataCue):
(WebKit::RemoteTextTrackProxy::addGenericCue):
(WebKit::RemoteTextTrackProxy::updateGenericCue):
(WebKit::RemoteTextTrackProxy::removeGenericCue):
(WebKit::RemoteTextTrackProxy::parseWebVTTFileHeader):
(WebKit::RemoteTextTrackProxy::parseWebVTTCueData):
* GPUProcess/media/RemoteTextTrackProxy.h:
* GPUProcess/media/RemoteTextTrackProxy.messages.in: Copied from Source/WebKit/WebProcess/GPU/media/SourceBufferPrivateRemote.messages.in.
* GPUProcess/media/RemoteVideoTrackProxy.cpp:
(WebKit::RemoteVideoTrackProxy::RemoteVideoTrackProxy):
(WebKit::RemoteVideoTrackProxy::~RemoteVideoTrackProxy):
(WebKit::RemoteVideoTrackProxy::configurationChanged):
(WebKit::RemoteVideoTrackProxy::willRemove):
* GPUProcess/media/RemoteVideoTrackProxy.h:
* GPUProcess/media/RemoteVideoTrackProxy.messages.in: Copied from Source/WebKit/WebProcess/GPU/media/SourceBufferPrivateRemote.messages.in.
* Scripts/webkit/messages.py:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/media/AudioTrackPrivateRemote.cpp:
(WebKit::AudioTrackPrivateRemote::AudioTrackPrivateRemote):
(WebKit::AudioTrackPrivateRemote::setEnabled):
* WebProcess/GPU/media/AudioTrackPrivateRemote.h:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::addRemoteAudioTrack):
(WebKit::MediaPlayerPrivateRemote::addRemoteTextTrack):
(WebKit::MediaPlayerPrivateRemote::addRemoteVideoTrack):
* WebProcess/GPU/media/MediaPlayerPrivateRemote.h:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.messages.in:
* WebProcess/GPU/media/MediaSourcePrivateRemote.cpp:
(WebKit::MediaSourcePrivateRemote::create):
(WebKit::MediaSourcePrivateRemote::MediaSourcePrivateRemote):
(WebKit::MediaSourcePrivateRemote::addSourceBuffer):
(WebKit::MediaSourcePrivateRemote::readyState const):
* WebProcess/GPU/media/MediaSourcePrivateRemote.h:
* WebProcess/GPU/media/SourceBufferPrivateRemote.cpp:
(WebKit::SourceBufferPrivateRemote::create):
(WebKit::SourceBufferPrivateRemote::SourceBufferPrivateRemote):
(WebKit::SourceBufferPrivateRemote::readyState const):
(WebKit::SourceBufferPrivateRemote::sourceBufferPrivateDidReceiveInitializationSegment):
* WebProcess/GPU/media/SourceBufferPrivateRemote.h:
* WebProcess/GPU/media/SourceBufferPrivateRemote.messages.in:
* WebProcess/GPU/media/TextTrackPrivateRemote.cpp:
(WebKit::TextTrackPrivateRemote::TextTrackPrivateRemote):
(WebKit::TextTrackPrivateRemote::setMode):
* WebProcess/GPU/media/TextTrackPrivateRemote.h:
* WebProcess/GPU/media/VideoTrackPrivateRemote.cpp:
(WebKit::VideoTrackPrivateRemote::VideoTrackPrivateRemote):
(WebKit::VideoTrackPrivateRemote::setSelected):
* WebProcess/GPU/media/VideoTrackPrivateRemote.h:
(WebKit::VideoTrackPrivateRemote::create):
2020-12-01 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] Support image extraction interactions for accessibility
https://bugs.webkit.org/show_bug.cgi?id=219398
<rdar://problem/70744914>
Reviewed by Devin Rousso.
Adds support for a new accessibility feature related to images. Refer to the bug for more details.
* Shared/ios/InteractionInformationRequest.cpp:
(WebKit::InteractionInformationRequest::encode const):
(WebKit::InteractionInformationRequest::decode):
(WebKit::InteractionInformationRequest::isValidForRequest const):
* Shared/ios/InteractionInformationRequest.h:
Add a new flag to request image data via position information. This flag works similarly to the existing
`includeSnapshot` flag, except that it only grabs data when the hit-tested node is an image. As such, a position
information request with the `includeSnapshot` flag set is valid for an identical request with this new
`includeImageData` flag set, but not vice versa.
* SourcesCocoa.txt:
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
Add various extension points for new WebKitAdditions snippets.
(-[WKContentView setUpInteraction]):
(-[WKContentView cleanUpInteraction]):
Call new methods to set up and tear down the image extraction interaction (defined in WebKitAdditions).
(-[WKContentView _webTouchEventsRecognized:]):
Upon starting a new gesture, reset the active (or pending) image extraction interaction, if appropriate.
(-[WKContentView gestureRecognizer:shouldRecognizeSimultaneouslyWithGestureRecognizer:]):
Allow the image extraction gesture to fire alongside any other gesture.
(-[WKContentView _dragInteraction:prepareForSession:completion:]):
(-[WKContentView _contextMenuInteraction:configurationForMenuAtLocation:completion:]):
These existing interactions may conflict with pending image extraction interactions. Luckily, since these
interactions are both asynchronous, we're able to defer them until any pending image extraction request has
finished, to determine whether or not they should commence. In both cases, image extraction should take
precedence over dragging and context menus.
* UIProcess/ios/WKImageExtractionGestureRecognizer.h:
* UIProcess/ios/WKImageExtractionGestureRecognizer.mm:
Add a new long press gesture recognizer that fires on a very short (100ms) delay. This gesture is used to
preflight image extraction requests ahead of the other long press gestures (in particular, context menus), such
that these other gestures should be minimally impacted by the time required to perform image extraction.
This is still a long press gesture instead of a regular gesture recognizer, since preflighting image extraction
requests could be expensive if performed during every single gesture (i.e. scrolling, taps, etc).
(-[WKImageExtractionGestureRecognizer initWithDelegate:]):
(-[WKImageExtractionGestureRecognizer setState:]):
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::imagePositionInformation):
2020-12-01 Per Arne Vollan <pvollan@apple.com>
[macCatalyst] Embedded YouTube videos fail to play
https://bugs.webkit.org/show_bug.cgi?id=219412
<rdar://problem/71869657>
Reviewed by Tim Horton.
Add a missing comma in list of Media services, for which sandbox extensions are created. The missing comma
is causing a sandbox violation, since strings are concatenated when there's no comma.
* UIProcess/WebPageProxy.cpp:
(WebKit::mediaRelatedMachServices):
2020-12-01 Per Arne Vollan <pvollan@apple.com>
[macOS] Deny mach lookup of 'com.apple.webinspector' in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=219386
<rdar://problem/60932734>
Reviewed by Darin Adler.
After r270035, mach-lookup of 'com.apple.webinspector' can be denied in the WebContent process on macOS.
* WebProcess/com.apple.WebProcess.sb.in:
2020-12-01 Brent Fulgham <bfulgham@apple.com>
[macOS] Add telemetry for sandbox rules we believe are unused
https://bugs.webkit.org/show_bug.cgi?id=219141
<rdar://problem/71569084>
Reviewed by Per Arne Vollan.
This patch adds telemetry to a set of sandbox features we believe are unused. We should
capture telemetry to verify, and confirm before removing.
* WebProcess/com.apple.WebProcess.sb.in:
2020-12-01 Keith Rollin <krollin@apple.com>
Consolidate header postprocessing scripts
https://bugs.webkit.org/show_bug.cgi?id=219388
<rdar://problem/71840357>
Reviewed by David Kilzer.
Our build system contains the following scripts to perform some
postprocessing of headers that we export to the SDK:
JavaScriptCore/postprocess-headers.sh
WebKit/mac/postprocess-framework-headers.sh
WebKitLegacy/mac/postprocess-headers.sh
The preceding scripts are used when using the non-XCBuild -- or
"legacy" -- Xcode build system. They are invoked in a custom Run
Script build phase after the headers have been exported with the
standard Xcode facility for creating frameworks.
Alternatively, we also have the following postprocessing scripts:
WebKit/Scripts/postprocess-header-rule
JavaScriptCore/Scripts/postprocess-header-rule
WebKitLegacy/scripts/postprocess-header-rule
These scripts are used when using the XCBuild build system. They are
invoked *during* the header export process to copy and postprocess the
headers in one blow. They are part of a Custom Build Rule for
exporting files ending in ".h".
The reason why we have two sets of scripts is because of the different
capabilities of the two Xcode build systems. The legacy system does
not support a custom "export header" step that would allow us to copy
and postprocess each header in a single step. Therefore, when using
the legacy build system, we export in one build step and postprocess
in a subsequent build step. And XCBuild doesn't like the approach
taken by the old build system where files are exported first and then
munged in a separate step, since that confuses its notion of the state
of the build ("Hey! That file I exported in the previous build? I see
now that it's been changed, so I'm going to export it again. And
change its modification date. And then rebuild everything downstream
that uses it."). Therefore, XCBuild added a facility for copying and
postprocessing in one step.
The scripts supporting each of these approaches are very similar to
each other, such that there is a lot of code duplication between them.
At the same time, by having two sets of scripts that are very similar
to each other, we run the risk of "drift", where files in one set may
get updated while their counterparts in the other set are not.
Address this duplication by making the scripts in the "legacy" set be
mere stubs that invoke the scripts in the new "XCBuild" set. In doing
this, we also fix a case of drift: the legacy-based scripts made use
of a timestamp file to determine if headers needed to be reprocessed
and exported, while the XCBuild-based scripts used a "process the
files and export them if any actual changes now exist between this new
version and any previously-exported version" approach.
Along the way, fix a bug in WebKitLegacy's postprocess-header-rule
that resulted in WebKitAvailability.h not being processed. The
practical effect of this bug is that the file ended up with both macOS
and iOS code, along with the #if that controlled which chunk of code
was compiled, instead of just the chunk of code specific to the
targeted SDK. Normally, the unused chunk of code would get removed
through the invocation of `unifdef`. But, because of the bug, the
results of running `unifdef` were being discarded.
* mac/postprocess-framework-headers.sh:
2020-12-01 Kate Cheney <katherine_cheney@apple.com>
ITP logging mixes up the UI process ITP state with the Settings ITP state
https://bugs.webkit.org/show_bug.cgi?id=219359
<rdar://problem/71808489>
Reviewed by Youenn Fablet.
Update an ITP logging message in the network process which mixes up
the variables for the ITP state passed from the UI process and the
ITP value from Settings.
No new tests, this only updates an incorrect logging message.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::activateSessionCleanup):
2020-12-01 Alejandro G. Castro <alex@igalia.com>
[GTK][WPE] Async scrolling udpates are blocked in the main thread
https://bugs.webkit.org/show_bug.cgi?id=219308
Reviewed by Simon Fraser.
We need to use the EventDispatcher from the frameDone callback in
the ThreadedCompositor, for GTK and WPE we receive the frame
information in that thread. We have to notify the scrolling trees
in the scrolling thread even if the main thread is busy or we will
freeze when the main thread is busy. For that we use the API in
the EventDispatcher.
* Shared/CoordinatedGraphics/threadedcompositor/ThreadedCompositor.cpp:
(WebKit::ThreadedCompositor::sceneUpdateFinished): Call the
EventDispatcher to notify the scrolling trees a frame was rendered.
* WebProcess/WebPage/EventDispatcher.cpp: Enable the compilation
of the notify function for all the ports.
* WebProcess/WebPage/EventDispatcher.h: Ditto.
2020-12-01 Tim Horton <timothy_horton@apple.com>
GPU Process: Invalid static_cast from ConcreteImageBuffer to RemoteImageBufferProxy
https://bugs.webkit.org/show_bug.cgi?id=219387
Reviewed by Ryosuke Niwa.
* Shared/ConcreteShareableImageBuffer.h: Copied from Source/WebKit/WebProcess/GPU/graphics/cocoa/ImageBufferShareableIOSurfaceBackend.h.
(WebKit::ConcreteShareableImageBuffer::create):
(WebKit::ConcreteShareableImageBuffer::ConcreteShareableImageBuffer):
(WebKit::ConcreteShareableImageBuffer::createImageBufferBackendHandle):
* Shared/RemoteLayerTree/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::encode const):
(WebKit::RemoteLayerBackingStore::swapToValidFrontBuffer):
* WebKit.xcodeproj/project.pbxproj:
Add and adopt ConcreteShareableImageBuffer. This is a ConcreteImageBuffer that knows how
to create an ImageBufferBackendHandle, with no other behavior change.
Previously, we were (completely incorrectly) casting ConcreteImageBuffer to RemoteImageBufferProxy
to get this behavior, and somehow just got very lucky.
* WebProcess/GPU/media/RemoteAudioSession.cpp:
Fix some unified source issues.
2020-11-30 Per Arne Vollan <pvollan@apple.com>
[macOS] Adopt SPI to prevent establishing XPC connections to Launch Services
https://bugs.webkit.org/show_bug.cgi?id=219173
<rdar://problem/71595536>
Reviewed by Alex Christensen.
Adopt SPI to prevent establishing XPC connections to Launch Services in the WebContent process on macOS.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-11-30 Tim Horton <timothy_horton@apple.com>
Unreviewed, reverting r270275.
Broke canvas painting
Reverted changeset:
"GPU Process: IOSurfaces should not be mapped into the Web
Content Process"
https://bugs.webkit.org/show_bug.cgi?id=219368
https://trac.webkit.org/changeset/270275
2020-11-30 Fujii Hironori <Hironori.Fujii@sony.com>
[WinCairo?] GPU process remains alive even after UI process exited
https://bugs.webkit.org/show_bug.cgi?id=219321
Reviewed by Don Olmstead.
GPUConnectionToWebProcess::didClose is called when web process
exits, and GPUProcess::didClose is called when UI process exits in
GPU process. GPUProcess::didClose should terminate the GPU
process.
AuxiliaryProcess is the base class of GPUProcess.
AuxiliaryProcess::didClose is calling _exit. However, GPUProcess
overrode AuxiliaryProcess::didClose but it did nothing. GPUProcess
doesn't need to override it.
* GPUProcess/GPUProcess.cpp:
(WebKit::GPUProcess::didClose): Deleted.
* GPUProcess/GPUProcess.h:
2020-11-30 Simon Fraser <simon.fraser@apple.com>
Pass OptionSet<EventHandling> to various wheel event handler functions
https://bugs.webkit.org/show_bug.cgi?id=219373
Reviewed by Tim Horton.
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::handleWheelEvent):
2020-11-30 Tim Horton <timothy_horton@apple.com>
GPU Process: IOSurfaces should not be mapped into the Web Content Process
https://bugs.webkit.org/show_bug.cgi?id=219368
Reviewed by Ryosuke Niwa.
Since the Web Content process sandbox will eventually not have access to IOKit,
it will not be able to map/lock/use IOSurfaces. Thankfully, all it really needs
to be able to do is carry a reference to the surface from the GPU process to the
UI process (where it is mapped and applied as layer contents).
* Shared/ConcreteShareableImageBuffer.h: Copied from Source/WebKit/WebProcess/GPU/graphics/cocoa/ImageBufferShareableIOSurfaceBackend.h.
(WebKit::ConcreteShareableImageBuffer::create):
(WebKit::ConcreteShareableImageBuffer::ConcreteShareableImageBuffer):
(WebKit::ConcreteShareableImageBuffer::createImageBufferBackendHandle):
* Shared/RemoteLayerTree/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::encode const):
(WebKit::RemoteLayerBackingStore::swapToValidFrontBuffer):
Add and adopt ConcreteShareableImageBuffer. This is a ConcreteImageBuffer that knows how
to create an ImageBufferBackendHandle, with no other behavior change.
Previously, we were (completely incorrectly) casting ConcreteImageBuffer to RemoteImageBufferProxy
to get this behavior, and somehow just got very lucky.
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/graphics/PlatformImageBufferShareableBackend.h:
* WebProcess/GPU/graphics/PlatformRemoteImageBufferProxy.h:
* WebProcess/GPU/graphics/cocoa/ImageBufferShareableIOSurfaceBackend.cpp:
* WebProcess/GPU/graphics/cocoa/ImageBufferShareableIOSurfaceBackend.h:
Remove the ImageBufferShareableIOSurfaceBackend create() method that takes
an ImageBufferBackendHandle, since we always use the Unmapped variant.
ImageBufferShareableIOSurfaceBackend is now only used:
- in the GPU process
- in the Web Content process when DOM rendering in the GPU process is disabled
In both cases, the process that uses ImageBufferShareableIOSurfaceBackend also
originates the surface, so does not create it from a handle.
* WebProcess/GPU/graphics/cocoa/ImageBufferShareableUnmappedIOSurfaceBackend.cpp: Added.
(WebKit::ImageBufferShareableUnmappedIOSurfaceBackend::create):
(WebKit::ImageBufferShareableUnmappedIOSurfaceBackend::createImageBufferBackendHandle const):
(WebKit::ImageBufferShareableUnmappedIOSurfaceBackend::context const):
(WebKit::ImageBufferShareableUnmappedIOSurfaceBackend::copyNativeImage const):
(WebKit::ImageBufferShareableUnmappedIOSurfaceBackend::copyImage const):
(WebKit::ImageBufferShareableUnmappedIOSurfaceBackend::draw):
(WebKit::ImageBufferShareableUnmappedIOSurfaceBackend::drawPattern):
(WebKit::ImageBufferShareableUnmappedIOSurfaceBackend::toDataURL const):
(WebKit::ImageBufferShareableUnmappedIOSurfaceBackend::toData const):
(WebKit::ImageBufferShareableUnmappedIOSurfaceBackend::toBGRAData const):
(WebKit::ImageBufferShareableUnmappedIOSurfaceBackend::getImageData const):
(WebKit::ImageBufferShareableUnmappedIOSurfaceBackend::putImageData):
* WebProcess/GPU/graphics/cocoa/ImageBufferShareableUnmappedIOSurfaceBackend.h: Added.
Add and adopt ImageBufferShareableUnmappedIOSurfaceBackend, which is a ImageBufferBackend
that cannot perform most ImageBufferBackend duties, except creating an ImageBufferBackendHandle,
which it does by cloning the existing handle. It explicitly does *not* map the IOSurface in the
process, and is meant for use in Web Content processes that do not have access to IOSurface.
* WebProcess/GPU/media/RemoteAudioSession.cpp:
Fix some unified source issues.
2020-11-30 Chris Dumez <cdumez@apple.com>
sessionStorage should not be cloned when a window is opened with rel=noopener
https://bugs.webkit.org/show_bug.cgi?id=218804
<rdar://problem/71286606>
Reviewed by Alex Christensen.
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::dispatchCreatePage):
* WebProcess/WebCoreSupport/WebFrameLoaderClient.h:
2020-11-30 Youenn Fablet <youenn@apple.com>
Introduce an experimental flag specific to VP9 profile 2
https://bugs.webkit.org/show_bug.cgi?id=219350
Reviewed by Eric Carlson.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences):
2020-11-30 Antti Koivisto <antti@apple.com>
Remove throttling code from NetworkLoad
https://bugs.webkit.org/show_bug.cgi?id=219347
Reviewed by Anders Carlsson.
It is not a good latency simulation and no one is using it anyway.
* NetworkProcess/NetworkLoad.cpp:
(WebKit::NetworkLoad::NetworkLoad):
(WebKit::NetworkLoad::didReceiveResponse):
(WebKit::NetworkLoad::didReceiveData):
(WebKit::NetworkLoad::didCompleteWithError):
(WebKit::NetworkLoad::Throttle::Throttle): Deleted.
(WebKit::NetworkLoad::throttleDelayCompleted): Deleted.
* NetworkProcess/NetworkLoad.h:
* NetworkProcess/NetworkSession.h:
(WebKit::NetworkSession::shouldLogCookieInformation const):
(WebKit::NetworkSession::loadThrottleLatency const): Deleted.
* NetworkProcess/NetworkSessionCreationParameters.cpp:
(WebKit::NetworkSessionCreationParameters::encode const):
(WebKit::NetworkSessionCreationParameters::decode):
* NetworkProcess/NetworkSessionCreationParameters.h:
* NetworkProcess/cocoa/NetworkSessionCocoa.h:
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::NetworkSessionCocoa):
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::WebsiteDataStore::platformSetNetworkParameters):
2020-11-29 Adam Mazander <mazander@apple.com>
Deny logs for benign fontservicesd daemon
https://bugs.webkit.org/show_bug.cgi?id=219180
Reviewed by Per Arne Vollan.
After investigating denials of fontservicesd logs over several submissions
and without any fallout we are removing logging.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-11-28 Per Arne Vollan <pvollan@apple.com>
[macOS] Set application information in the Networking process on behalf of the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=218052
<rdar://problem/70586405>
Reviewed by Brent Fulgham.
In preparation for blocking Launch Services in the WebContent process' sandbox, application information for the process
should be set in the Networking process.
* NetworkProcess/NetworkConnectionToWebProcess.h:
* NetworkProcess/NetworkConnectionToWebProcess.messages.in:
* NetworkProcess/mac/NetworkConnectionToWebProcessMac.mm: Added.
(WebKit::NetworkConnectionToWebProcess::updateActivePages):
* Scripts/process-entitlements.sh:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
(WebKit::activePagesOrigins):
(WebKit::WebProcess::updateActivePages):
(WebKit::WebProcess::getActivePagesOriginsForTesting):
* WebProcess/com.apple.WebProcess.sb.in:
2020-11-28 Per Arne Vollan <pvollan@apple.com>
[macOS] Issue sandbox extension to icon services when starting a drag operation
https://bugs.webkit.org/show_bug.cgi?id=219159
<rdar://problem/71585876>
Reviewed by Brent Fulgham.
Issue sandbox extension to icon services when starting a drag operation on macOS, since the WebContent process needs
access to these services when a drag is started. This patch detects when a drag is about to start, and sends sandbox
extensions to the icon services along with the mouse event. The sandbox extensions are then temporarily consumed
while handling the mouse event in the WebContent process.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::processNextQueuedMouseEvent):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::mouseEvent):
(WebKit::WebPage::simulateMouseDown):
(WebKit::WebPage::simulateMouseUp):
(WebKit::WebPage::simulateMouseMotion):
(WebKit::WebPage::consumeSandboxExtensions):
(WebKit::WebPage::revokeSandboxExtensions):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-11-27 Fujii Hironori <Hironori.Fujii@sony.com>
Unreviewed, reverting r270210.
It broke TestWebKitAPI.GPUProcess.CrashWhilePlayingVideo on
iOS
Reverted changeset:
"[WinCairo?] GPU process remains alive even after web process
exited"
https://bugs.webkit.org/show_bug.cgi?id=219321
https://trac.webkit.org/changeset/270210
2020-11-27 Fujii Hironori <Hironori.Fujii@sony.com>
[WinCairo?] GPU process remains alive even after web process exited
https://bugs.webkit.org/show_bug.cgi?id=219321
Reviewed by Don Olmstead.
AuxiliaryProcess::didClose is calling _exit. However, GPUProcess
overrode AuxiliaryProcess::didClose but it did nothing. GPUProcess
doesn't need to override it.
* GPUProcess/GPUProcess.cpp:
(WebKit::GPUProcess::didClose): Deleted.
* GPUProcess/GPUProcess.h:
2020-11-27 Don Olmstead <don.olmstead@sony.com>
Non-unified build fixes, late November 2020 edition, take two
https://bugs.webkit.org/show_bug.cgi?id=219317
Unreviewed non-unified build fixes.
* Shared/WebPreferencesDefaultValues.cpp:
2020-11-27 Don Olmstead <don.olmstead@sony.com>
Fix build when ENABLE_RESOURCE_LOAD_STATISTICS is disabled
https://bugs.webkit.org/show_bug.cgi?id=219316
Unreviewed build fix.
Add ENABLE(RESOURCE_LOAD_STATISTICS) guards around NetworkSession requests for
resource load statistics. Move firePrivateClickMeasurementTimerImmediately out of
a ENABLE(RESOURCE_LOAD_STATISTICS) guard and place it where its ordered in the header.
This originally broke the PlayStation build before ENABLE_RESOURCE_LOAD_STATISTICS
was turned on.
No new tests. No change in behavior.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::simulateResourceLoadStatisticsSessionRestart):
* NetworkProcess/NetworkSession.cpp:
(WebKit::NetworkSession::firePrivateClickMeasurementTimerImmediately):
* NetworkProcess/PrivateClickMeasurementManager.cpp:
(WebKit::PrivateClickMeasurementManager::storeUnattributed):
(WebKit::PrivateClickMeasurementManager::attribute):
(WebKit::PrivateClickMeasurementManager::clearSentAttributions):
(WebKit::PrivateClickMeasurementManager::updateTimerLastFired):
(WebKit::PrivateClickMeasurementManager::firePendingAttributionRequests):
(WebKit::PrivateClickMeasurementManager::clear):
(WebKit::PrivateClickMeasurementManager::clearForRegistrableDomain):
(WebKit::PrivateClickMeasurementManager::clearExpired):
(WebKit::PrivateClickMeasurementManager::toString const):
(WebKit::PrivateClickMeasurementManager::markAllUnattributedAsExpiredForTesting):
(WebKit::PrivateClickMeasurementManager::markAttributedPrivateClickMeasurementsAsExpiredForTesting):
2020-11-27 Wenson Hsieh <wenson_hsieh@apple.com>
Introduce new display list meta commands in preparation for webkit.org/b/219091
https://bugs.webkit.org/show_bug.cgi?id=219262
Reviewed by Ryosuke Niwa.
See WebCore ChangeLog for more information. Additionally, rename `MetaCommandSwitchToItemBuffer` to
`MetaCommandChangeItemBuffer`, for consistency with the new `MetaCommandChangeDestinationImageBuffer` item.
* GPUProcess/graphics/RemoteImageBuffer.h:
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::decodeItem):
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
2020-11-27 Wenson Hsieh <wenson_hsieh@apple.com>
Remove some extraneous PLATFORM(IOS_FAMILY) guards in WKContentViewInteraction
https://bugs.webkit.org/show_bug.cgi?id=219289
Reviewed by Sam Weinig.
Remove several redundant uses of `PLATFORM(IOS_FAMILY)`. The entire class is already conditional on
`IOS_FAMILY`, so adding additional compile-time guards for iOS family is not necessary.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView setTimePickerValueToHour:minute:]):
(-[WKContentView timePickerValueHour]):
(-[WKContentView timePickerValueMinute]):
2020-11-27 Adrian Perez de Castro <aperez@igalia.com>
Non-unified build fixes, late November 2020 edition
https://bugs.webkit.org/show_bug.cgi?id=219306
Unreviewed non-unified build fixes.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp: Add missing
PrivateClickMeasurementManager.h header.
* NetworkProcess/PrivateClickMeasurementManager.cpp: Add missing NetworkSession.h header.
* UIProcess/SpeechRecognitionServer.cpp:
(WebKit::SpeechRecognitionServer::handleRequest): Add missing WebCore:: namespace to uses
of the WebCore::SpeechRecognizer and WebCore::SpeechRecognitionUpdateType types.
2020-11-27 Yoshiaki Jitsukawa <yoshiaki.jitsukawa@sony.com>
[PlayStation] Define platform argument coders for Font
https://bugs.webkit.org/show_bug.cgi?id=219300
Reviewed by Fujii Hironori.
* PlatformPlayStation.cmake:
* Shared/playstation/WebCoreArgumentCodersPlayStation.cpp: Added.
Define required member functions of ArgumentCoder<Ref<WebCore::Font>> class.
2020-11-26 Wenson Hsieh <wenson_hsieh@apple.com>
Calling waitForAndDispatchImmediately<M> on a loop fails when multiple M messages arrive simultaneously
https://bugs.webkit.org/show_bug.cgi?id=219240
Reviewed by Chris Dumez.
Fixes a race that may occur when calling `waitForAndDispatchImmediately<N>` in a loop, when multiple messages
`N` arrive on the IPC thread simultaneously. This may result from the following sequence of events (note that
(Main) and (IPC) in the timeline below refer to the main thread and IPC background thread, respectively):
(Main) Call `waitForAndDispatchImmediately`, and begin waiting.
(IPC) A message `N` arrives, and is handled in `processIncomingMessage` by setting the decoder of
`m_waitingForMessage` and notifying the condition variable.
(Main) The main thread wakes up and starts to process `N`, clearing out `m_waitingForMessage` in the process.
(IPC) A second message `N` arrives. We see that `m_waitingForMessage` is null, so we don't set the decoder
and bail. Instead, we prepare to call `enqueueIncomingMessage` and push the message onto the main
thread, *but importantly*, we haven't done so yet.
(Main) Call `waitForAndDispatchImmediately` again, set `m_waitingForMessage`, and begin waiting. Since the
incoming message that was just received above has not been enqueued yet, we are unable to avoid waiting
due to the incoming message.
(IPC) We finally call `enqueueIncomingMessage`, which pushes the message `N` into `m_incomingMessages` and
dispatches onto the main thread. However, this is too late, since the main thread is already stuck
waiting for the incoming IPC message that we've now just enqueued.
Two minor adjustments are required to fix this, described in the below comments. The combination of these two
changes ensures that the scenario described above is impossible, since we'll either set `m_waitingForMessage`'s
decoder and wake up the main thread in the case where `waitForMessage` is called before `processIncomingMessage`,
or we'll bail early in `waitForMessage` with the enqueued IPC message in the case where `processIncomingMessage`
runs before `waitForMessage`.
* Platform/IPC/Connection.cpp:
(IPC::Connection::waitForMessage):
Move logic that checks the incoming messages queue when calling `Connection::waitForMessage` into the
`m_waitForMessageMutex` critical section.
(IPC::Connection::processIncomingMessage):
Extend the critical section of `m_waitForMessageMutex` when processing an incoming message, such that it
encompasses the part that enqueues the incoming message.
2020-11-26 Lauro Moura <lmoura@igalia.com>
[GTK][GTK4] Building with GObject-Introspection support does not work
https://bugs.webkit.org/show_bug.cgi?id=219221
Reviewed by Carlos Garcia Campos.
* PlatformGTK.cmake: Forward GTK version to gir scanner and use correct pkg-config package.
2020-11-26 Lauro Moura <lmoura@igalia.com>
[GTK4] Declare lambda return type to avoid build error deducing it
https://bugs.webkit.org/show_bug.cgi?id=219268
Reviewed by Adrian Perez de Castro.
* UIProcess/gtk/WebPopupMenuProxyGtk.cpp:
(WebKit::WebPopupMenuProxyGtk::createPopupMenu):
2020-11-26 Fujii Hironori <Hironori.Fujii@sony.com>
[WinCairo] Enable GPU process
https://bugs.webkit.org/show_bug.cgi?id=219294
Reviewed by Don Olmstead.
Added stub functions to build GPU process.
Invoke the following command to enable GPU process:
> reg add HKEY_CURRENT_USER\Software\WebKit /v gpu_process_canvas_rendering /t REG_DWORD /d 1 /f
Invoke the following command to disable GPU process:
> reg delete HKEY_CURRENT_USER\Software\WebKit /v gpu_process_canvas_rendering /f
* GPUProcess/media/win/RemoteMediaPlayerProxyWin.cpp: Added.
(WebKit::RemoteMediaPlayerProxy::prepareForPlayback):
(WebKit::RemoteMediaPlayerProxy::mediaPlayerFirstVideoFrameAvailable):
(WebKit::RemoteMediaPlayerProxy::mediaPlayerRenderingModeChanged):
(WebKit::RemoteMediaPlayerProxy::enterFullscreen):
(WebKit::RemoteMediaPlayerProxy::exitFullscreen):
* GPUProcess/win/GPUProcessMainWin.cpp: Added.
(WebKit::initializeAuxiliaryProcess<GPUProcess>):
(WebKit::GPUProcessMain):
* GPUProcess/win/GPUProcessWin.cpp: Added.
(WebKit::GPUProcess::initializeProcess):
(WebKit::GPUProcess::initializeProcessName):
(WebKit::GPUProcess::initializeSandbox):
* PlatformWin.cmake:
* Shared/WebPreferencesDefaultValues.h:
* Shared/curl/WebCoreArgumentCodersCurl.cpp:
(IPC::ArgumentCoder<FontAttributes>::encodePlatformData): Deleted.
(IPC::ArgumentCoder<FontAttributes>::decodePlatformData): Deleted.
(IPC::ArgumentCoder<Ref<Font>>::encodePlatformData): Deleted.
(IPC::ArgumentCoder<Ref<Font>>::decodePlatformData): Deleted.
* Shared/win/WebCoreArgumentCodersWin.cpp: Added.
(IPC::ArgumentCoder<FontAttributes>::encodePlatformData):
(IPC::ArgumentCoder<FontAttributes>::decodePlatformData):
(IPC::ArgumentCoder<Ref<Font>>::encodePlatformData):
(IPC::ArgumentCoder<Ref<Font>>::decodePlatformData):
* Shared/win/WebPreferencesDefaultValuesWin.cpp: Added.
(WebKit::isFeatureFlagEnabled):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::submitDisplayList):
(WebKit::RemoteRenderingBackendProxy::createItemBuffer):
* WebProcess/GPU/media/win/VideoLayerRemoteWin.cpp: Added.
(WebKit::createVideoLayerRemote):
2020-11-26 Youenn Fablet <youenn@apple.com>
https://collab-project.github.io/videojs-record/demo/video-only.html is not working
https://bugs.webkit.org/show_bug.cgi?id=219258
<rdar://problem/69759808>
Reviewed by Eric Carlson.
Fixed a potential null pointer crash in case load is aborted by WebCore at creation time.
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::requestResource):
2020-11-26 Yoshiaki Jitsukawa <yoshiaki.jitsukawa@sony.com>
[PlayStation] Enable ResourceLoadStatistics
https://bugs.webkit.org/show_bug.cgi?id=219292
Reviewed by Fujii Hironori.
* PlatformPlayStation.cmake:
2020-11-21 Sihui Liu <sihui_liu@apple.com>
Implement audio capture for SpeechRecognition on macOS
https://bugs.webkit.org/show_bug.cgi?id=218855
<rdar://problem/71331001>
Reviewed by Youenn Fablet.
* UIProcess/SpeechRecognitionPermissionManager.cpp:
(WebKit::SpeechRecognitionPermissionManager::startProcessingRequest): Check and enable mock devices based on
preference as SpeechRecognition needs it for testing.
* UIProcess/SpeechRecognitionServer.cpp:
(WebKit::SpeechRecognitionServer::start):
(WebKit::SpeechRecognitionServer::requestPermissionForRequest):
(WebKit::SpeechRecognitionServer::handleRequest):
(WebKit::SpeechRecognitionServer::stop):
(WebKit::SpeechRecognitionServer::abort):
(WebKit::SpeechRecognitionServer::invalidate):
(WebKit::SpeechRecognitionServer::sendUpdate):
(WebKit::SpeechRecognitionServer::stopRequest): Deleted.
(WebKit::SpeechRecognitionServer::abortRequest): Deleted.
* UIProcess/SpeechRecognitionServer.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::syncIfMockDevicesEnabledChanged):
* UIProcess/WebPageProxy.h:
2020-11-21 Simon Fraser <simon.fraser@apple.com>
Propagate the 'wheelEventGesturesBecomeNonBlocking' setting to the ScrollingTree
https://bugs.webkit.org/show_bug.cgi?id=219241
Reviewed by Tim Horton.
Propagate the 'wheelEventGesturesBecomeNonBlocking' setting to the ScrollingTree
via the ScrollingStateFrameScrollingNode, as we do for other settings.
* Shared/RemoteLayerTree/RemoteScrollingCoordinatorTransaction.cpp:
(ArgumentCoder<ScrollingStateFrameScrollingNode>::encode):
2020-11-21 Simon Fraser <simon.fraser@apple.com>
Add an Experimental Features for wheel event gestures becoming non-blocking
https://bugs.webkit.org/show_bug.cgi?id=219236
Reviewed by Sam Weinig.
Add a feature flag for the behavior that is being added via webkit.org/b/218764,
which is that only the first wheel event in a gesture is cancelable.
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultWheelEventGesturesBecomeNonBlocking):
* Shared/WebPreferencesDefaultValues.h:
2020-11-21 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Implement SPI for AuthenticationServices.Framework
https://bugs.webkit.org/show_bug.cgi?id=218893
<rdar://problem/71364731>
Reviewed by Alex Christensen.
This patch implements the SPI for AuthenticationServices.Framework, which is more or less a direct
translation of the WebAuthn Web IDLs and a thin wrapper on top of functionalities of the WebAuthn
process.
It deprecates the existing _WKWebAuthenticationPanel SPI which is for Safari to show UI only. This
patch does not just repurpose the SPI but also modifies some of the implementations. The most noticeable one
is the change of ownerships. Prior to this change, AuthenticatorManager owns the APIWebAuthenticationPanel.
Now, it's the opposite way.
Since there will be a period of time that the deprecated implementation will coexist with the new implementation,
this path also marks things that are meant to deprecate in the future to reduce confusions.
* UIProcess/API/APIWebAuthenticationPanel.cpp:
(API::WebAuthenticationPanel::WebAuthenticationPanel):
(API::WebAuthenticationPanel::handleRequest):
(API::WebAuthenticationPanel::cancel const):
* UIProcess/API/APIWebAuthenticationPanel.h:
Implements the plumbing for the new SPI and owns the AuthenticatorManager now.
* UIProcess/API/Cocoa/_WKWebAuthenticationPanel.h:
* UIProcess/API/Cocoa/_WKWebAuthenticationPanel.mm:
(-[_WKAuthenticationExtensionsClientOutputs initWithAppid:]):
(-[_WKAuthenticatorResponse initWithRawId:extensions:]):
(-[_WKAuthenticatorAssertionResponse initWithRawId:extensions:authenticatorData:signature:userHandle:]):
(-[_WKAuthenticatorAttestationResponse initWithRawId:extensions:attestationObject:]):
(-[_WKAuthenticatorSelectionCriteria init]):
(-[_WKPublicKeyCredentialCreationOptions initWithRp:user:pubKeyCredParams:]):
(-[_WKPublicKeyCredentialDescriptor initWithType:identifier:]):
(-[_WKPublicKeyCredentialEntity initWithName:]):
(-[_WKPublicKeyCredentialParameters initWithType:alg:]):
(-[_WKPublicKeyCredentialRequestOptions init]):
(-[_WKPublicKeyCredentialRpEntity initWithName:]):
(-[_WKPublicKeyCredentialUserEntity initWithName:identifier:displayName:]):
(-[_WKWebAuthenticationPanel init]):
(vectorFromNSData):
(encodeEntity):
(publicKeyCredentialRpEntity):
(publicKeyCredentialUserEntity):
(publicKeyCredentialParameters):
(authenticatorTransport):
(authenticatorTransports):
(publicKeyCredentialDescriptors):
(authenticatorAttachment):
(userVerification):
(authenticatorSelectionCriteria):
(attestationConveyancePreference):
(authenticationExtensionsClientInputs):
(+[_WKWebAuthenticationPanel convertToCoreCreationOptionsWithOptions:]):
(wkAuthenticatorAttestationResponse):
(-[_WKWebAuthenticationPanel makeCredentialWithHash:options:completionHandler:]):
(+[_WKWebAuthenticationPanel convertToCoreRequestOptionsWithOptions:]):
(wkAuthenticatorAssertionResponse):
(-[_WKWebAuthenticationPanel getAssertionWithHash:options:completionHandler:]):
(+[_WKWebAuthenticationPanel isUserVerifyingPlatformAuthenticatorAvailable]):
Implements the SPI.
* UIProcess/WebAuthentication/WebAuthenticationRequestData.h:
Marks things as to deprecate.
2020-11-20 Kate Cheney <katherine_cheney@apple.com>
PCM: Persist pending ad clicks and attributions so they can survive browser restart
https://bugs.webkit.org/show_bug.cgi?id=219134
<rdar://problem/70470129>
Reviewed by John Wilander.
This patch migrates Private Click Measurement to use SQLite,
which is beneficial because it requires less in-memory storage and
persists PCM data across browser sessions. It also updates naming
to match naming agreed upon in standards bodies:
- source -> sourceSite
- campaign/campaignID -> sourceID
- destination -> attributeOnSite
- conversion/conversionValue -> attributionTriggerData
- unconverted -> unattributed
- convert(ed) -> attribute(d)
This adds 3 SQLite tables: one for clicks that haven't been
attributed, one for attributions that haven't been sent, and one to
store the last time the reports were sent to make sure reports get
sent as soon as possible if needed after a browser restart.
Behavior is identical to existing PCM implementation with the addition
of persistence. Existing PCM tests confirm no regressions.
Reviewed by John Wilander.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::createTableQueries):
(WebKit::ResourceLoadStatisticsDatabaseStore::ResourceLoadStatisticsDatabaseStore):
(WebKit::ResourceLoadStatisticsDatabaseStore::createUniqueIndices):
(WebKit::ResourceLoadStatisticsDatabaseStore::createSchema):
(WebKit::ResourceLoadStatisticsDatabaseStore::destroyStatements):
New queries to interact with PCM data.
(WebKit::ResourceLoadStatisticsDatabaseStore::updateTimerLastFired):
(WebKit::ResourceLoadStatisticsDatabaseStore::timerLastFired):
(WebKit::ResourceLoadStatisticsDatabaseStore::updatePrivateClickMeasurementAttributionTimes):
Set earliestTimeToSend to be the original value minus the time passed since the last timer fire
for each entry. If the result is less than 0, set to 0 so the report gets sent immediately.
(WebKit::ResourceLoadStatisticsDatabaseStore::buildPrivateClickMeasurementFromDatabase):
Creates a PCM object from data in the database.
(WebKit::ResourceLoadStatisticsDatabaseStore::findPrivateClickMeasurement):
(WebKit::ResourceLoadStatisticsDatabaseStore::insertPrivateClickMeasurement):
(WebKit::ResourceLoadStatisticsDatabaseStore::markAllUnattributedPrivateClickMeasurementAsExpiredForTesting):
(WebKit::ResourceLoadStatisticsDatabaseStore::removeUnattributed):
(WebKit::ResourceLoadStatisticsDatabaseStore::attributePrivateClickMeasurement):
(WebKit::ResourceLoadStatisticsDatabaseStore::allAttributedPrivateClickMeasurement):
(WebKit::ResourceLoadStatisticsDatabaseStore::clearPrivateClickMeasurement):
(WebKit::ResourceLoadStatisticsDatabaseStore::clearExpiredPrivateClickMeasurement):
(WebKit::ResourceLoadStatisticsDatabaseStore::attributionToString):
(WebKit::ResourceLoadStatisticsDatabaseStore::privateClickMeasurementToString):
(WebKit::ResourceLoadStatisticsDatabaseStore::clearSentAttributions):
These functions use database queries to implement PCM functionality with exactly the same
behavior as the in-memory PCM implementation.
(WebKit::ResourceLoadStatisticsDatabaseStore::markAttributedPrivateClickMeasurementsAsExpiredForTesting):
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.h:
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::updateTimerLastFired):
(WebKit::WebResourceLoadStatisticsStore::insertPrivateClickMeasurement):
(WebKit::WebResourceLoadStatisticsStore::markAllUnattributedPrivateClickMeasurementAsExpiredForTesting):
(WebKit::WebResourceLoadStatisticsStore::attributePrivateClickMeasurement):
(WebKit::WebResourceLoadStatisticsStore::allAttributedPrivateClickMeasurement):
(WebKit::WebResourceLoadStatisticsStore::clearPrivateClickMeasurement):
(WebKit::WebResourceLoadStatisticsStore::clearPrivateClickMeasurementForRegistrableDomain):
(WebKit::WebResourceLoadStatisticsStore::clearExpiredPrivateClickMeasurement):
(WebKit::WebResourceLoadStatisticsStore::privateClickMeasurementToString):
(WebKit::WebResourceLoadStatisticsStore::clearSentAttributions):
(WebKit::WebResourceLoadStatisticsStore::markAttributedPrivateClickMeasurementsAsExpiredForTesting):
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.h:
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::firePrivateClickMeasurementTimerImmediately):
(WebKit::NetworkProcess::simulateResourceLoadStatisticsSessionRestart):
(WebKit::NetworkProcess::markAttributedPrivateClickMeasurementsAsExpiredForTesting):
Test functions to help simulate a browser restart after PCM data has expired during
a session close. This is the only behavior change from the existing PCM implementation.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/NetworkResourceLoader.cpp:
(WebKit::NetworkResourceLoader::willSendRedirectedRequest):
(WebKit::NetworkResourceLoader::continueWillSendRedirectedRequest):
* NetworkProcess/NetworkResourceLoader.h:
* NetworkProcess/NetworkSession.cpp:
(WebKit::NetworkSession::NetworkSession):
(WebKit::NetworkSession::firePrivateClickMeasurementTimerImmediately):
(WebKit::NetworkSession::storePrivateClickMeasurement):
(WebKit::NetworkSession::handlePrivateClickMeasurementConversion):
(WebKit::NetworkSession::markAttributedPrivateClickMeasurementsAsExpiredForTesting):
(WebKit::NetworkSession::markPrivateClickMeasurementsAsExpiredForTesting):
* NetworkProcess/NetworkSession.h:
* NetworkProcess/PrivateClickMeasurementManager.cpp:
(WebKit::PrivateClickMeasurementManager::PrivateClickMeasurementManager):
Move constructor to cpp file to call startTimer(5_s) which will kick
off sending any reports that have expired in the database. We should
wait 5 seconds so we are sure ITP is up and running.
(WebKit::PrivateClickMeasurementManager::storeUnattributed):
(WebKit::PrivateClickMeasurementManager::handleAttribution):
(WebKit::PrivateClickMeasurementManager::startTimer):
(WebKit::PrivateClickMeasurementManager::attribute):
(WebKit::PrivateClickMeasurementManager::fireConversionRequest):
(WebKit::PrivateClickMeasurementManager::clearSentAttributions):
(WebKit::PrivateClickMeasurementManager::updateTimerLastFired):
(WebKit::PrivateClickMeasurementManager::firePendingAttributionRequests):
(WebKit::PrivateClickMeasurementManager::clear):
(WebKit::PrivateClickMeasurementManager::clearForRegistrableDomain):
(WebKit::PrivateClickMeasurementManager::clearExpired):
(WebKit::PrivateClickMeasurementManager::toString const):
(WebKit::PrivateClickMeasurementManager::setConversionURLForTesting):
(WebKit::PrivateClickMeasurementManager::markAllUnattributedAsExpiredForTesting):
(WebKit::PrivateClickMeasurementManager::markAttributedPrivateClickMeasurementsAsExpiredForTesting):
(WebKit::PrivateClickMeasurementManager::storeUnconverted): Deleted.
(WebKit::PrivateClickMeasurementManager::handleConversion): Deleted.
(WebKit::PrivateClickMeasurementManager::convert): Deleted.
(WebKit::PrivateClickMeasurementManager::firePendingConversionRequests): Deleted.
(WebKit::PrivateClickMeasurementManager::markAllUnconvertedAsExpiredForTesting): Deleted.
Implementation moved to ResourceLoadStatisticsDatabaseStore.
* NetworkProcess/PrivateClickMeasurementManager.h:
(WebKit::PrivateClickMeasurementManager::PrivateClickMeasurementManager): Deleted.
Moved to cpp file.
(WebKit::PrivateClickMeasurementManager::m_sessionID): Deleted.
* UIProcess/API/C/WKPage.cpp:
(WKPageMarkAttributedPrivateClickMeasurementsAsExpiredForTesting):
(WKPageSimulateResourceLoadStatisticsSessionRestart):
* UIProcess/API/C/WKPagePrivate.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didCommitLoadForFrame):
(WebKit::WebPageProxy::markAttributedPrivateClickMeasurementsAsExpiredForTesting):
(WebKit::WebPageProxy::simulateResourceLoadStatisticsSessionRestart):
* UIProcess/WebPageProxy.h:
Testing support.
2020-11-20 Simon Fraser <simon.fraser@apple.com>
Dispatch main-thread overflow scrolls to the scrolling thread as we do for page scrolls
https://bugs.webkit.org/show_bug.cgi?id=219213
Reviewed by Tim Horton.
Unify the behavior of overflow scrolling and page scrolling for synchronous scrolls.
Somewhat surprisingly, synchronous page scrolls move layers via a dispatch to the scrolling thread in
FrameView::handleWheelEventForScrolling(), but overflow scrolls just did main thread compositing updates
to set the new layer positions.
A future patch will require that the "began" event for a synchronous scroll gets back to the scrolling thread,
so unify these code paths to have overflow scrolls also leverage handleWheelEventAfterMainThread(), via
RenderLayer::handleWheelEventForScrolling().
There's some fallout from this. ThreadedScrollingTree::handleWheelEventAfterMainThread() calls into handleWheelEventWithNode(),
but in this special case that codepath needs to know that we're in a special "post main thread" mode that should
1. Behave as if the node is latched, i.e. don't propagate to parent nodes,
and target the node if if it's scrolled to a the min or max to allow rubberbanding
2. Scroll even if the node has synchronous scrolling reasons
This mode is represented by the EventTargeting value.
Finally, EventHandler::handleWheelEventInternal() should only clear latching state if the content has called preventDefault()
on the event.
* UIProcess/RemoteLayerTree/mac/ScrollingTreeFrameScrollingNodeRemoteMac.cpp:
(WebKit::ScrollingTreeFrameScrollingNodeRemoteMac::handleWheelEvent):
* UIProcess/RemoteLayerTree/mac/ScrollingTreeFrameScrollingNodeRemoteMac.h:
* UIProcess/RemoteLayerTree/mac/ScrollingTreeOverflowScrollingNodeRemoteMac.cpp:
(WebKit::ScrollingTreeOverflowScrollingNodeRemoteMac::handleWheelEvent):
* UIProcess/RemoteLayerTree/mac/ScrollingTreeOverflowScrollingNodeRemoteMac.h:
2020-11-20 Simon Fraser <simon.fraser@apple.com>
Rename some wheel-event handling functions for clarity
https://bugs.webkit.org/show_bug.cgi?id=219211
Reviewed by Tim Horton.
There are too many functions called wheelEvent() or handleWheelEvent(), making it hard
to know which phase of handling they apply to. So rename some to handleWheelEventForScrolling(),
which applies to the "default handling" phase after DOM event dispatch.
In addition, make ScrollableArea's handleWheelEventForScrolling() virtual and have FrameView
override it (a future patch will also add an override in RenderLayer).
Rename ScrollingCoordinator::performDefaultWheelEventHandling() to use handleWheelEventForScrolling()
for clarity.
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::handleWheelEvent):
2020-11-20 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed fix for my build fix.
* UIProcess/ios/WKActionSheetAssistant.mm:
(-[WKActionSheetAssistant _appendAppLinkOpenActionsForURL:actions:elementInfo:]):
You have to spell DEPRECATED correctly for this to work.
2020-11-20 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed, fix the build with recent SDKs.
* UIProcess/ios/WKActionSheetAssistant.mm:
(-[WKActionSheetAssistant _appendAppLinkOpenActionsForURL:actions:elementInfo:]):
2020-11-20 Philippe Normand <pnormand@igalia.com>
Unreviewed, GTK build warning fix
* UIProcess/Inspector/glib/RemoteInspectorClient.cpp: Mark configurationForRemoteInspector
as override of corresponding parent class method declaration.
2020-11-19 Wenson Hsieh <wenson_hsieh@apple.com>
REGRESSION (r259151): The "Convert Text to Traditional Chinese" services menu action is broken
https://bugs.webkit.org/show_bug.cgi?id=219190
<rdar://problem/71438389>
Reviewed by Tim Horton.
Some system services, such as "Convert Text to (Traditional|Simplified) Chinese", use `-[NSServicesMenuRequestor
readSelectionFromPasteboard:]` to insert the contents of the given pasteboard into the current selection. After
the changes in r259151, this requires the UI process to explicitly grant access to the contents of the given
pasteboard, on behalf of the web content process.
Fix the bug by adding the missing call to `grantAccessToCurrentPasteboardData`. This is only called from API in
the UI process, so it cannot be abused by a compromised web content process to grab the contents of
`NSPasteboard`.
Test: PasteHTML.ReadSelectionFromPasteboard
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::readSelectionFromPasteboard):
2020-11-19 Fujii Hironori <Hironori.Fujii@sony.com>
[TextureMapper] Remove m_textureMapper from TextureMapperLayer
https://bugs.webkit.org/show_bug.cgi?id=219191
Reviewed by Carlos Garcia Campos.
* Shared/CoordinatedGraphics/CoordinatedGraphicsScene.cpp:
(WebKit::CoordinatedGraphicsScene::paintToCurrentGLContext):
(WebKit::CoordinatedGraphicsScene::ensureRootLayer):
* WebProcess/WebPage/CoordinatedGraphics/LayerTreeHostTextureMapper.cpp:
(WebKit::LayerTreeHost::compositeLayersToContext):
(WebKit::LayerTreeHost::flushPendingLayerChanges):
(WebKit::LayerTreeHost::LayerTreeHost):
2020-11-19 Hoa Dinh <dvh@apple.com>
Code formatting: change instances of "@synthesize a=b" to "@synthesize a = b".
https://bugs.webkit.org/show_bug.cgi?id=219094
Reviewed by Wenson Hsieh.
There was a mix of "@synthesize a=b" and "@synthesize a = b" in the codebase.
Most of the instances are "@synthesize a = b", with spaces around the equal sign.
With https://bugs.webkit.org/show_bug.cgi?id=219092, we're changing the behavior of
the code style checker to require spaces around the equal sign.
The change replace all the instances of "@synthesize a=b" with "@synthesize a = b".
* NetworkProcess/CustomProtocols/Cocoa/LegacyCustomProtocolManagerCocoa.mm:
* UIProcess/API/Cocoa/WKPreviewActionItem.mm:
* UIProcess/API/Cocoa/_WKThumbnailView.mm:
* UIProcess/Cocoa/WebViewImpl.mm:
* UIProcess/QuickLookThumbnailLoader.mm:
* UIProcess/WebAuthentication/Mock/MockNfcService.mm:
* UIProcess/ios/fullscreen/WKFullScreenViewController.mm:
* UIProcess/mac/ViewGestureControllerMac.mm:
* UIProcess/mac/WKFullScreenWindowController.mm:
* WebProcess/Plugins/PDF/PDFPlugin.mm:
2020-11-19 Brent Fulgham <bfulgham@apple.com>
Unreviewed rollout of r269890 and r269911 due to EME issues on Mac Mini.
* GPUProcess/mac/com.apple.WebKit.GPUProcess.sb.in:
* WebProcess/com.apple.WebProcess.sb.in:
2020-11-19 Chris Dumez <cdumez@apple.com>
Web Inspector: Unable to Show JavaScript Console in empty tab in Safari Technology Preview
https://bugs.webkit.org/show_bug.cgi?id=211590
<rdar://problem/63000159>
Reviewed by Simon Fraser.
We used to launch the initial process for the inspected page from inside
WebInspectorProxy::connect(). I thought this was a good bottleneck to
make sure the inspected page has a process to inspect.
However, while WebInspectorProxy::connect() gets called by functions by
WebInspectorProxy::show(), it does not get called by
WebInspectorProxy::showConsole() for example.
To address this, I have moved the code that launches the inspected page's
initial process from WebInspectorProxy::connect() to
WebInspectorProxy::createFrontendPage(). This seems like a much better
bottleneck.
* UIProcess/Inspector/WebInspectorProxy.cpp:
(WebKit::WebInspectorProxy::connect):
(WebKit::WebInspectorProxy::createFrontendPage):
2020-11-19 Per Arne Vollan <pvollan@apple.com>
[macOS] Issue sandbox extension to Web Inspector service
https://bugs.webkit.org/show_bug.cgi?id=219041
<rdar://problem/71495287>
Reviewed by Brent Fulgham.
In preparation of blocking this service in the WebContent process, a sandbox extension should be issued if Safari's Develop menu is enabled.
This extension will also be dynamically issued to all WebContent processes, if the Develop menu preference changes, by observing this
preference.
* UIProcess/Cocoa/WebInspectorPreferenceObserver.h: Added.
* UIProcess/Cocoa/WebInspectorPreferenceObserver.mm: Added.
(+[WKWebInspectorPreferenceObserver sharedInstance]):
(-[WKWebInspectorPreferenceObserver init]):
(-[WKWebInspectorPreferenceObserver observeValueForKeyPath:ofObject:change:context:]):
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitialize):
* UIProcess/Cocoa/WebProcessProxyCocoa.mm:
(WebKit::WebProcessProxy::enableRemoteInspectorIfNeeded):
* UIProcess/WebProcessPool.h:
* WebKit.xcodeproj/project.pbxproj:
2020-11-19 Per Arne Vollan <pvollan@apple.com>
[macOS] Issue sandbox extension to audio service if Media in the GPU process is not enabled
https://bugs.webkit.org/show_bug.cgi?id=219051
<rdar://problem/71500898>
Reviewed by Youenn Fablet.
Issue sandbox extension to com.apple.audio.audiohald for the WebContent process on macOS if Media in the GPU process is not enabled.
* UIProcess/WebPageProxy.cpp:
(WebKit::mediaRelatedMachServices):
* WebProcess/com.apple.WebProcess.sb.in:
2020-11-19 Joshua Watt <JPEW.hacker@gmail.com>
[WPE] Prevent elements getting stuck in a pressed state when scrolling on a touchscreen
https://bugs.webkit.org/show_bug.cgi?id=218903
Reviewed by Adrian Perez de Castro.
Stops the scroll gesture controller from eating touch up events and
preventing them from propagating to the page when the controller is
handling a scroll event. If the scroll logic prevents the events from
propagating to the page, elements on the page (e.g. buttons) can get
stuck in a pressed state when the user lifts their finger.
* UIProcess/API/wpe/ScrollGestureController.cpp:
(WebKit::ScrollGestureController::handleEvent):
2020-11-19 Carlos Garcia Campos <cgarcia@igalia.com>
Protect WebSocketChannel before calling client methods
https://bugs.webkit.org/show_bug.cgi?id=216791
Reviewed by Youenn Fablet.
Ensure we keep a reference to the WebSocketChannel before calling client methods that might close the channel.
* WebProcess/Network/WebSocketChannel.cpp:
(WebKit::WebSocketChannel::close):
(WebKit::WebSocketChannel::fail):
(WebKit::WebSocketChannel::didClose):
(WebKit::WebSocketChannel::resume):
2020-11-18 Megan Gardner <megan_gardner@apple.com>
Menu Bar support for app highlights in book.
https://bugs.webkit.org/show_bug.cgi?id=219117
Reviewed by Wenson Hsieh.
Add menu stubs in iOS to support highlights for books in modern webkit.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView setUpInteraction]):
(-[WKContentView canPerformAction:withSender:]):
(-[WKContentView setupAppHighlightMenus]):
(-[WKContentView createHighlightInCurrentGroupWithRange:]):
(-[WKContentView createHighlightInNewGroupWithRange:]):
2020-11-18 Wenson Hsieh <wenson_hsieh@apple.com>
Rename MetaCommandSwitchTo to MetaCommandSwitchToItemBuffer
https://bugs.webkit.org/show_bug.cgi?id=219130
Reviewed by Tim Horton.
* GPUProcess/graphics/RemoteImageBuffer.h:
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::decodeItem):
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
2020-11-18 Wenson Hsieh <wenson_hsieh@apple.com>
[Concurrent display lists] Add a way for display lists to partially replay
https://bugs.webkit.org/show_bug.cgi?id=219067
Reviewed by Simon Fraser.
See WebCore ChangeLog for more details.
* GPUProcess/graphics/RemoteImageBuffer.h:
Make `submitDisplayList` on `RemoteImageBuffer` return a `ReplayResult`.
(WebKit::RemoteImageBuffer::submitDisplayList):
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::applyDisplayListsFromHandle):
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
(WebKit::RemoteImageBufferProxy::submitDisplayList):
2020-11-18 Chris Dumez <cdumez@apple.com>
Drop redundant code that tries to bump the QoS of the WebContent main thread of UserInteractive when handling a sync IPC
https://bugs.webkit.org/show_bug.cgi?id=219110
Reviewed by Alex Christensen.
Drop redundant code that tries to bump the QoS of the WebContent main thread of UserInteractive when
handling a sync IPC. Nowadays, we always run the main thread of the WebContent at UserInteractive QoS
(see WTF::Thread::setCurrentThreadIsUserInteractive() call in WebProcess::initializeWebProcess()) so
there is no need to boost the QoS during a sync IPC.
* Platform/IPC/Connection.cpp:
(IPC::Connection::processIncomingMessage):
* Platform/IPC/Connection.h:
(IPC::Connection::setShouldBoostMainThreadOnSyncMessage): Deleted.
* Platform/IPC/Decoder.h:
(IPC::Decoder::setQOSClassOverride): Deleted.
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::initializeConnection):
2020-11-18 Darin Adler <darin@apple.com>
Remove advanced plug-in feature: small plug-in blocking
https://bugs.webkit.org/show_bug.cgi?id=219101
Reviewed by Anders Carlsson.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::pluginIsSmall): Deleted.
(WebKit::WebPage::createPlugin): Removed code that blocks small plug-ins.
2020-11-18 Simon Fraser <simon.fraser@apple.com>
Propagate wheel event handling back to the scrolling thread
https://bugs.webkit.org/show_bug.cgi?id=219050
Reviewed by Chris Dumez.
For now, use MainThreadForBlockingDOMEventDispatch for the default steps (used by non-macOS platforms).
* UIProcess/RemoteLayerTree/RemoteScrollingCoordinatorProxy.cpp:
(WebKit::RemoteScrollingCoordinatorProxy::handleWheelEvent):
* WebProcess/WebPage/EventDispatcher.cpp:
(WebKit::EventDispatcher::wheelEvent):
2020-11-18 Wenson Hsieh <wenson_hsieh@apple.com>
Clean up some code in SharedDisplayListHandle
https://bugs.webkit.org/show_bug.cgi?id=219089
Reviewed by Geoff Garen.
Currently, `reservedCapacityAtStart` is defined as a constant 16 bytes, which is enough to encompass the
contents of the header structure in a shared display list handle (i.e. an 8-byte atomic for the lock, and
another 8 bytes for the unread count).
Instead of hard-coding this, we could simply make this a constexpr function that returns the size of
`DisplayListSharedMemoryHeader` (rounded up to ensure alignment of all display list item data).
No change in behavior.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::wakeUpAndApplyDisplayList):
(WebKit::RemoteRenderingBackend::didCreateSharedDisplayListHandle):
* Shared/SharedDisplayListHandle.h:
(WebKit::SharedDisplayListHandle::headerSize):
* WebProcess/GPU/graphics/DisplayListWriterHandle.cpp:
(WebKit::DisplayListWriterHandle::resetWritableOffsetIfPossible):
* WebProcess/GPU/graphics/DisplayListWriterHandle.h:
(WebKit::DisplayListWriterHandle::DisplayListWriterHandle):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::createItemBuffer):
Also add a static assert that the size of a newly allocated buffer is larger than the reserved header capacity.
2020-11-18 Per Arne Vollan <pvollan@apple.com>
[macOS] Fix message filter sandbox violation
https://bugs.webkit.org/show_bug.cgi?id=219090
<rdar://problem/71522043>
Reviewed by Geoffrey Garen.
The message filter in the WebContent process on macOS needs to allow some additional messages.
* WebProcess/com.apple.WebProcess.sb.in:
2020-11-18 Chris Dumez <cdumez@apple.com>
[GPUProcess] Main thread of the GPUProcess should have same priority as main thread of the WebContent
https://bugs.webkit.org/show_bug.cgi?id=219057
Reviewed by Simon Fraser.
Main thread of the GPUProcess should have same priority as main thread of the WebContent since it
is doing rendering on its behalf.
For now, this means that on macOS, the main threads of the GPU and WebContent processes with both
get UserInteractive QoS. On iOS, they will both still get UserInitiated QoS until our RunningBoard
foreground process assertion get fixed to get the right priority (<rdar://problem/71499731>).
* GPUProcess/EntryPoint/Cocoa/XPCService/GPUService/Info-OSX.plist:
Use _ProcessType=App, similarly as what we do for the WebContent process. If we don't do
this, the process is an adaptive daemon and its QoS can never be higher than UserInitiated.
We need its priority to go to UserInteractive to match the WebContent process.
* GPUProcess/GPUProcess.cpp:
(WebKit::GPUProcess::initializeGPUProcess):
Request UserInteractive QoS for the main thread of the GPUProcess, similarly to what we do
for the main thread of the WebContent process already in WebProcess::InitializeWebProcess().
2020-11-18 Sam Weinig <weinig@apple.com>
Address additional feedback from https://bugs.webkit.org/show_bug.cgi?id=218960
https://bugs.webkit.org/show_bug.cgi?id=219044
Reviewed by Alex Christensen.
* Shared/API/Cocoa/WKDataDetectorTypesInternal.h:
(fromWKDataDetectorTypes):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::detectDataInAllFrames):
* UIProcess/WebPageProxy.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences):
(WebKit::WebPage::detectDataInAllFrames):
Update for rename of WebCore::DataDetectionTypes to WebCore::DataDetectionType.
2020-11-17 Tim Horton <timothy_horton@apple.com>
GPU-process-hosted RemoteLayerBackingStore should flush off the main thread
https://bugs.webkit.org/show_bug.cgi?id=219063
Reviewed by Simon Fraser.
* GPUProcess/graphics/RemoteRenderingBackend.h:
* Shared/RemoteLayerTree/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::display):
Construct the ImageBuffer flusher /after/ submitting the final DisplayList,
so that it takes the correct flush identifier.
* GPUProcess/graphics/RemoteImageBuffer.h:
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::didFlush):
(WebKit::RemoteRenderingBackend::flushDisplayListWasCommitted): Deleted.
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
(WebKit::RemoteImageBufferProxy::didFlush):
(WebKit::RemoteImageBufferProxy::lastSentFlushIdentifier const):
(WebKit::RemoteImageBufferProxy::waitForDidFlushOnSecondaryThread):
(WebKit::RemoteImageBufferProxy::hasPendingFlush const):
(WebKit::RemoteImageBufferProxy::waitForDidFlushWithTimeout):
(WebKit::ThreadSafeRemoteImageBufferFlusher::ThreadSafeRemoteImageBufferFlusher):
(WebKit::RemoteImageBufferProxy::commitFlushDisplayList): Deleted.
(WebKit::RemoteImageBufferProxy::isPendingFlush const): Deleted.
(WebKit::RemoteImageBufferProxy::timeoutWaitForFlushDisplayListWasCommitted): Deleted.
Some renames:
Rename flushDisplayListWasCommitted to didFlush; this is the back-message from
GPU process to Web Content process that a FlushContext display list item was completed successfully.
Rename isPendingFlush to hasPendingFlush, for grammar.
Rename timeoutWaitForFlushDisplayListWasCommitted to waitForDidFlushWithTimeout, for similar reasons.
Rename waitForFlushDisplayListWasCommitted to waitForDidFlush.
Rename commitFlushDisplayList to didFlush.
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::waitForDidFlush):
(WebKit::RemoteRenderingBackendProxy::didFlush):
(WebKit::RemoteRenderingBackendProxy::waitForFlushDisplayListWasCommitted): Deleted.
(WebKit::RemoteRenderingBackendProxy::flushDisplayListWasCommitted): Deleted.
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.messages.in:
Reimplement ThreadSafeRemoteImageBufferFlusher such that it waits
(on the RemoteLayerTreeDrawingArea CommitQueue, in the only current use of it)
for the correct flush reply to come in before continuing. This is analogous
to what we do for in-process buffers, where we call CGContextFlush()
in the flusher.
2020-11-17 Chris Dumez <cdumez@apple.com>
[iOS] ASSERTION FAILED: Completion handler should always be called under WebKit::GPUProcess::didReceiveMessage
https://bugs.webkit.org/show_bug.cgi?id=219055
<rdar://problem/71508223>
Reviewed by Alex Christensen.
Make sure GPUProcess::prepareToSuspend() calls its completion handler to avoid delaying suspension
and avoid assertion hits in debug.
* GPUProcess/GPUProcess.cpp:
(WebKit::GPUProcess::prepareToSuspend):
2020-11-17 Chris Dumez <cdumez@apple.com>
Make sure ProcessThrottler is always holding a ProcessAssertion for its child process
https://bugs.webkit.org/show_bug.cgi?id=219053
Reviewed by Geoff Garen.
Make sure ProcessThrottler is always holding a ProcessAssertion for its child process. Previously, when changing the assertion
type (e.g. when going from background to foreground) we would release the previous assertion and then take the new one.
RunningBoard developers recommend that we wait until the new ProcessAssertion is taken before we release the previous one to
avoid churn. It is also important we always hold a ProcessAssertion at all times for the child process because the child process
may exit otherwise.
* UIProcess/ProcessThrottler.cpp:
(WebKit::ProcessThrottler::setAssertionType):
2020-11-17 Chris Dumez <cdumez@apple.com>
[macOS] Stop using RunLoopType=_WebKit starting in Big Sur
https://bugs.webkit.org/show_bug.cgi?id=219052
<rdar://problem/61742969>
Reviewed by Geoffrey Garen.
Stop using RunLoopType=_WebKit starting in Big Sur. This was a temporary hack so that our WebProcesses would get the right
scheduling priority. Starting with Big Sur, the right way to do this is to use _ProcessType=App and restore the RunLoopType
to be NSRunLoop.
After this change, I have verified that the WebContent's main thread still runs at UserInteractive QoS. App Nap is also
still working as expected.
* Configurations/WebContentService.xcconfig:
* WebProcess/EntryPoint/Cocoa/XPCService/WebContentService/Info-OSX.plist:
2020-11-17 Per Arne Vollan <pvollan@apple.com>
[macOS] Issue sandbox extension to trust service if Media in the GPU process is not enabled
https://bugs.webkit.org/show_bug.cgi?id=218986
<rdar://problem/71439885>
Reviewed by Brent Fulgham.
Since Media is using the trust service, issue a sandbox extension to the WebContent process if Media in the GPU process is not enabled.
* UIProcess/WebPageProxy.cpp:
(WebKit::mediaRelatedMachServices):
* WebProcess/com.apple.WebProcess.sb.in:
2020-11-17 Sihui Liu <sihui_liu@apple.com>
Add a default action for SpeechRecognition permission request
https://bugs.webkit.org/show_bug.cgi?id=219021
Reviewed by Youenn Fablet.
When there is no delegate or selector for deciding user permission on speech recognition, force an alert and let
user decide. This is currently what we do for UserMedia permission request, and it makes testing easier.
* SourcesCocoa.txt:
* UIProcess/API/APIUIClient.h:
(API::UIClient::decidePolicyForSpeechRecognitionPermissionRequest):
* UIProcess/Cocoa/MediaPermissionUtilities.mm:
(WebKit::visibleDomain):
(WebKit::alertMessageText):
(WebKit::allowButtonText):
(WebKit::doNotAllowButtonText):
(WebKit::alertForPermission):
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::UIClient::decidePolicyForSpeechRecognitionPermissionRequest):
* UIProcess/Cocoa/UserMediaPermissionRequestProxy.mm: Removed. Moved alert code to MediaPermissionUtilities so
it can be shared between UserMedia and SpeechRecognition.
* UIProcess/MediaPermissionUtilities.h:
* UIProcess/SpeechRecognitionPermissionManager.cpp:
(WebKit::SpeechRecognitionPermissionManager::decideByDefaultAction):
* UIProcess/SpeechRecognitionPermissionManager.h:
* UIProcess/UserMediaPermissionRequestProxy.cpp:
(WebKit::UserMediaPermissionRequestProxy::doDefaultAction):
* UIProcess/UserMediaPermissionRequestProxy.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::requestSpeechRecognitionPermissionByDefaultAction):
* UIProcess/WebPageProxy.h:
* WebKit.xcodeproj/project.pbxproj:
2020-11-17 Saagar Jha <saagar@saagarjha.com>
Check com.apple.private.webkit.use-xpc-endpoint only on internal builds
https://bugs.webkit.org/show_bug.cgi?id=215423
Reviewed by Per Arne Vollan.
Signing with com.apple.private entitlements can only be done by Apple internally. Thus,
it should only checked for internal builds.
* Shared/Cocoa/XPCEndpoint.mm:
(WebKit::XPCEndpoint::XPCEndpoint): Wrap check in USE(APPLE_INTERNAL_SDK).
* Shared/Cocoa/XPCEndpointClient.mm:
(WebKit::XPCEndpointClient::setEndpoint): Wrap check in USE(APPLE_INTERNAL_SDK).
2020-11-17 Brent Fulgham <bfulgham@apple.com>
[macOS] Remove access to the unused AppleSNBFBUserClient IOKit class
https://bugs.webkit.org/show_bug.cgi?id=219014
<rdar://problem/70463873>
Reviewed by Eric Carlson.
Although we originally thought this was needed for H.264 decoding, testing shows
that this class is never used in WebKit flows, and is not needed. We should remove it.
* GPUProcess/mac/com.apple.WebKit.GPUProcess.sb.in:
* WebProcess/com.apple.WebProcess.sb.in:
2020-11-17 Chris Dumez <cdumez@apple.com>
[iOS] Stop leaking an XPC transaction in our XPC services
https://bugs.webkit.org/show_bug.cgi?id=219036
Reviewed by Geoffrey Garen.
Stop leaking an XPC transaction in our XPC services on iOS. We were doing this to control the lifetime of our child services
ourselves. However, this is not needed on iOS because the UIProcess takes RunningBoard process assertions on behalf of its
child processes.
I have verified that our child processes do not exit/jetsam early. I have verified that our child processes are in the
correct jetsam band (IDLE/0 when background and FG/10 when foreground). I have also verified that the main thread of these
processes runs at UserInitiated QoS before and after this change.
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceEntryPoint.h:
(WebKit::XPCServiceInitializer):
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceEntryPoint.mm:
(WebKit::XPCServiceExit):
2020-11-17 Peng Liu <peng.liu6@apple.com>
[Media In GPU Process][MSE] Add infrastructure needed to run MediaPlayerPrivateMediaSourceAVFObjC in the GPU process
https://bugs.webkit.org/show_bug.cgi?id=218912
Reviewed by Eric Carlson.
Add MediaSourcePrivateRemote/RemoteMediaSourceProxy and SourceBufferPrivateRemote/RemoteSourceBufferProxy,
and setup IPC connections for them.
* CMakeLists.txt:
* DerivedSources-input.xcfilelist:
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* GPUProcess/GPUProcessSessionParameters.h:
* GPUProcess/media/RemoteLegacyCDMFactoryProxy.cpp:
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
(WebKit::RemoteMediaPlayerProxy::loadMediaSource):
* GPUProcess/media/RemoteMediaPlayerProxy.h:
* GPUProcess/media/RemoteMediaPlayerProxy.messages.in:
* GPUProcess/media/RemoteMediaSourceProxy.cpp: Added.
(WebKit::RemoteMediaSourceProxy::RemoteMediaSourceProxy):
(WebKit::RemoteMediaSourceProxy::~RemoteMediaSourceProxy):
(WebKit::RemoteMediaSourceProxy::setPrivateAndOpen):
(WebKit::RemoteMediaSourceProxy::duration const):
(WebKit::RemoteMediaSourceProxy::buffered const):
(WebKit::RemoteMediaSourceProxy::seekToTime):
(WebKit::RemoteMediaSourceProxy::monitorSourceBuffers):
(WebKit::RemoteMediaSourceProxy::setLogIdentifier):
(WebKit::RemoteMediaSourceProxy::failedToCreateRenderer):
(WebKit::RemoteMediaSourceProxy::addSourceBuffer):
* GPUProcess/media/RemoteMediaSourceProxy.h: Added.
* GPUProcess/media/RemoteMediaSourceProxy.messages.in: Added.
* GPUProcess/media/RemoteSourceBufferIdentifier.h: Added.
* GPUProcess/media/RemoteSourceBufferProxy.cpp: Added.
(WebKit::RemoteSourceBufferProxy::create):
(WebKit::RemoteSourceBufferProxy::RemoteSourceBufferProxy):
(WebKit::RemoteSourceBufferProxy::~RemoteSourceBufferProxy):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateDidReceiveInitializationSegment):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateDidReceiveSample):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateHasAudio const):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateHasVideo const):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateReenqueSamples):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateDidBecomeReadyForMoreSamples):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateFastSeekTimeForMediaTime):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateAppendComplete):
(WebKit::RemoteSourceBufferProxy::sourceBufferPrivateDidReceiveRenderingError):
(WebKit::RemoteSourceBufferProxy::append):
* GPUProcess/media/RemoteSourceBufferProxy.h: Added.
* GPUProcess/media/RemoteSourceBufferProxy.messages.in: Added.
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayerManager.cpp:
* Scripts/webkit/messages.py:
* Sources.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::load):
(WebKit::MediaPlayerPrivateRemote::MediaPlayerPrivateRemote::load): Deleted.
* WebProcess/GPU/media/MediaPlayerPrivateRemote.h:
* WebProcess/GPU/media/MediaSourcePrivateRemote.cpp: Added.
(WebKit::MediaSourcePrivateRemote::create):
(WebKit::MediaSourcePrivateRemote::MediaSourcePrivateRemote):
(WebKit::MediaSourcePrivateRemote::~MediaSourcePrivateRemote):
(WebKit::MediaSourcePrivateRemote::addSourceBuffer):
(WebKit::MediaSourcePrivateRemote::durationChanged):
(WebKit::MediaSourcePrivateRemote::markEndOfStream):
(WebKit::MediaSourcePrivateRemote::unmarkEndOfStream):
(WebKit::MediaSourcePrivateRemote::readyState const):
(WebKit::MediaSourcePrivateRemote::setReadyState):
(WebKit::MediaSourcePrivateRemote::waitForSeekCompleted):
(WebKit::MediaSourcePrivateRemote::seekCompleted):
(WebKit::MediaSourcePrivateRemote::logChannel const):
* WebProcess/GPU/media/MediaSourcePrivateRemote.h: Added.
* WebProcess/GPU/media/RemoteCDMConfiguration.h:
* WebProcess/GPU/media/RemoteCDMInstanceConfiguration.h:
* WebProcess/GPU/media/RemoteMediaPlayerManager.h:
* WebProcess/GPU/media/RemoteMediaSourceIdentifier.h: Added.
* WebProcess/GPU/media/SourceBufferPrivateRemote.cpp: Added.
(WebKit::SourceBufferPrivateRemote::create):
(WebKit::SourceBufferPrivateRemote::SourceBufferPrivateRemote):
(WebKit::SourceBufferPrivateRemote::~SourceBufferPrivateRemote):
(WebKit::SourceBufferPrivateRemote::setClient):
(WebKit::SourceBufferPrivateRemote::append):
(WebKit::SourceBufferPrivateRemote::abort):
(WebKit::SourceBufferPrivateRemote::resetParserState):
(WebKit::SourceBufferPrivateRemote::removedFromMediaSource):
(WebKit::SourceBufferPrivateRemote::readyState const):
(WebKit::SourceBufferPrivateRemote::setReadyState):
(WebKit::SourceBufferPrivateRemote::flush):
(WebKit::SourceBufferPrivateRemote::enqueueSample):
(WebKit::SourceBufferPrivateRemote::isReadyForMoreSamples):
(WebKit::SourceBufferPrivateRemote::setActive):
(WebKit::SourceBufferPrivateRemote::notifyClientWhenReadyForMoreSamples):
(WebKit::SourceBufferPrivateRemote::canSetMinimumUpcomingPresentationTime const):
(WebKit::SourceBufferPrivateRemote::setMinimumUpcomingPresentationTime):
(WebKit::SourceBufferPrivateRemote::clearMinimumUpcomingPresentationTime):
(WebKit::SourceBufferPrivateRemote::canSwitchToType):
(WebKit::SourceBufferPrivateRemote::sourceBufferPrivateAppendComplete):
(WebKit::SourceBufferPrivateRemote::logChannel const):
* WebProcess/GPU/media/SourceBufferPrivateRemote.h: Added.
* WebProcess/GPU/media/SourceBufferPrivateRemote.messages.in: Added.
2020-11-17 Per Arne Vollan <pvollan@apple.com>
[macOS] Perform AX TCC check in the UI process
https://bugs.webkit.org/show_bug.cgi?id=218870
<rdar://problem/71339830>
Reviewed by Brent Fulgham.
On behalf of the WebContent process, perform AX TCC check in the UI process on macOS.
This is in preparation of blocking tccd in the WebContent process.
No new tests. It has been manually tested that the WebContent process is allowing VoiceOver
to perform AX requests with this patch.
* UIProcess/Cocoa/WebProcessProxyCocoa.mm:
(WebKit::WebProcessProxy::isAXAuthenticated):
* UIProcess/WebProcessProxy.h:
* UIProcess/WebProcessProxy.messages.in:
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::isAXAuthenticatedCallback):
(WebKit::WebProcess::platformInitializeWebProcess):
2020-11-17 Brian Burg <bburg@apple.com>
[Cocoa] _WKInspectorExtensionHost should conform to NSObject protocol
https://bugs.webkit.org/show_bug.cgi?id=219035
Reviewed by Alex Christensen.
This was overlooked in the initial patch. Without it, we can't call -isEqual: and
similar basic methods on a type of id<_WKInspectorExtensionHost>.
* UIProcess/API/Cocoa/_WKInspectorExtensionHost.h:
2020-11-17 Tim Horton <timothy_horton@apple.com>
Preemptive build fix for https://bugs.webkit.org/show_bug.cgi?id=219024
* UIProcess/ios/forms/WKAirPlayRoutePicker.mm:
(SOFT_LINK_CLASS):
(-[WKAirPlayRoutePicker show:fromRect:]):
2020-11-17 Tim Horton <timothy_horton@apple.com>
REGRESSION (r269824): YouTube media control bar sometimes flickers
https://bugs.webkit.org/show_bug.cgi?id=219017
<rdar://problem/71454799>
Reviewed by Simon Fraser.
Test: compositing/repaint/transparent-layer-repaint.html
* Shared/RemoteLayerTree/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::display):
I accidentally lost the Copy composite operator in r269824, causing
the repaint copy-forward to blend instead of just copying.
This wreaks havoc on layers with contents with colors that have 0 < alpha < 1,
because every repaint means the colors intensify.
2020-11-16 Brent Fulgham <bfulgham@apple.com>
[macOS] Remove access to the unused AppleIntelMEUserClient IOKit class
https://bugs.webkit.org/show_bug.cgi?id=219012
<rdar://problem/70462796>
Reviewed by Eric Carlson.
Although we originally thought this was needed for H.264 decoding, testing shows that
this class is never used in WebKit flows, and is not needed. We should remove it.
* GPUProcess/mac/com.apple.WebKit.GPUProcess.sb.in:
* WebProcess/com.apple.WebProcess.sb.in:
2020-11-16 Per Arne Vollan <pvollan@apple.com>
[macOS] Create sandbox extension for the power log service on internal installs
https://bugs.webkit.org/show_bug.cgi?id=218984
<rdar://problem/58488936>
Reviewed by Brent Fulgham.
Create a sandbox extension for the WebContent process for the power log service on macOS internal installs.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::diagnosticServices):
* WebProcess/com.apple.WebProcess.sb.in:
2020-11-16 Sam Weinig <weinig@apple.com>
Standardize enums that are used by Settings in preperation for autogeneration
https://bugs.webkit.org/show_bug.cgi?id=218960
Reviewed by Tim Horton.
Update enum uses for renames and scoped syntax use.
* UIProcess/API/C/WKAPICast.h:
(WebKit::toAPI):
(WebKit::toEditableLinkBehavior):
(WebKit::toStorageBlockingPolicy):
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetEditableLinkBehavior):
(WKPreferencesSetStorageBlockingPolicy):
(WKPreferencesGetStorageBlockingPolicy):
* UIProcess/API/Cocoa/WKPreferences.mm:
(toStorageBlockingPolicy):
(toAPI):
(-[WKPreferences _storageBlockingPolicy]):
(-[WKPreferences _setStorageBlockingPolicy:]):
(toEditableLinkBehavior):
(-[WKPreferences _setEditableLinkBehavior:]):
* UIProcess/WebPreferences.cpp:
(WebKit::WebPreferences::createWithLegacyDefaults):
* WebProcess/Storage/WebSWContextManagerConnection.cpp:
(WebKit::WebSWContextManagerConnection::updatePreferencesStore):
* WebProcess/Storage/WebSWContextManagerConnection.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences):
(WebKit::WebPage::detectDataInAllFrames):
2020-11-16 John Wilander <wilander@apple.com>
NetworkDataTaskCocoa::applyCookiePolicyForThirdPartyCNAMECloaking() redacts the cookie name in log output on customer builds
https://bugs.webkit.org/show_bug.cgi?id=218997
<rdar://problem/71460604>
Unreviewed minor change to logging.
NetworkDataTaskCocoa::applyCookiePolicyForThirdPartyCNAMECloaking() redacts the
cookie name in log output on customer builds. This patch changes the format
specifier from %s to %{public}s.
Note that this logging is just for ITP Debug Mode which is off by default, logs
in-memory only, and is disabled again on browser quit.
No new tests. This just changes log output on customer builds.
* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::NetworkDataTaskCocoa::applyCookiePolicyForThirdPartyCNAMECloaking):
2020-11-16 Chris Dumez <cdumez@apple.com>
[GPUProcess] Update GPUProcess process assertion based on active WebProcesses
https://bugs.webkit.org/show_bug.cgi?id=219002
Reviewed by Simon Fraser.
Update GPUProcess process assertion based on active WebProcesses, similarly to what
we do for the network process. Previously, we would never release the GPUProcess
process assertion and it would time out after 30 seconds in the background.
* UIProcess/GPU/GPUProcessProxy.cpp:
(WebKit::GPUProcessProxy::updateProcessAssertion):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::updateProcessAssertions):
2020-11-16 Brent Fulgham <bfulgham@apple.com>
[macOS] Remove remote tcp capability from WebContent Sandbox
https://bugs.webkit.org/show_bug.cgi?id=218999
<rdar://problem/70355789>
Reviewed by Per Arne Vollan.
Since we have moved all network activity (aside from some syslog use) out of the WebContent
process, we do not need the ability to open remote tcp connections. We should deny this
capability from our non-Network sandboxes.
* GPUProcess/mac/com.apple.WebKit.GPUProcess.sb.in:
* WebAuthnProcess/mac/com.apple.WebKit.WebAuthnProcess.sb.in:
* WebProcess/com.apple.WebProcess.sb.in:
2020-11-16 Devin Rousso <drousso@apple.com>
Ignore deprecation warnings for the uniform types C API
https://bugs.webkit.org/show_bug.cgi?id=218989
Reviewed by Tim Horton.
These should eventually be replaced with the ObjC API `UniformTypeIdentifiers.framework`.
See <https://developer.apple.com/documentation/uniformtypeidentifiers>.
* Shared/ios/WebIconUtilities.mm:
(WebKit::iconForFile):
* Shared/mac/PasteboardTypes.mm:
(WebKit::PasteboardTypes::forEditing):
(WebKit::PasteboardTypes::forSelection):
* UIProcess/API/Cocoa/APIAttachmentCocoa.mm:
(API::isDeclaredOrDynamicTypeIdentifier):
(API::Attachment::mimeType const):
(API::Attachment::utiType const):
(API::Attachment::setFileWrapperAndUpdateContentType):
* UIProcess/Automation/cocoa/WebAutomationSessionCocoa.mm:
(WebKit::WebAutomationSession::platformGetBase64EncodedPNGData):
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::setPromisedDataForImage):
* UIProcess/ios/WKContentView.mm:
(-[WKContentView _commonInitializationWithProcessPool:configuration:]):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView supportedPasteboardTypesForCurrentSelection]):
* UIProcess/ios/WKPDFView.mm:
(-[WKPDFView actionSheetAssistant:performAction:]):
* UIProcess/ios/forms/WKFileUploadPanel.mm:
(-[WKFileUploadPanel currentAvailableActionTitles]):
(-[WKFileUploadPanel contextMenuInteraction:configurationForMenuAtLocation:]):
(-[WKFileUploadPanel showFilePickerMenu]):
(-[WKFileUploadPanel showDocumentPickerMenu]):
(-[WKFileUploadPanel _uploadItemFromMediaInfo:successBlock:failureBlock:]):
* UIProcess/mac/WebContextMenuProxyMac.mm:
(WebKit::WebContextMenuProxyMac::setupServicesMenu):
2020-11-16 Don Olmstead <don.olmstead@sony.com>
Non-unified build fixes, mid November 2020 edition
https://bugs.webkit.org/show_bug.cgi?id=218911
Unreviewed build fix.
* UIProcess/SpeechRecognitionServer.cpp:
(WebKit::SpeechRecognitionServer::start):
(WebKit::SpeechRecognitionServer::requestPermissionForRequest):
(WebKit::SpeechRecognitionServer::sendUpdate):
2020-11-16 Per Arne Vollan <pvollan@apple.com>
[macOS] The WebContent sandbox does not apply for open source builds
https://bugs.webkit.org/show_bug.cgi?id=218982
Reviewed by Alexey Proskuryakov.
The WebContent sandbox does not apply for open source builds on macOS, since it has enabled message filtering,
which requires a private entitlement.
* WebProcess/com.apple.WebProcess.sb.in:
2020-11-16 Megan Gardner <megan_gardner@apple.com>
Add menu support for app highlights for books
https://bugs.webkit.org/show_bug.cgi?id=218879
<rdar://problem/71352113>
Reviewed by Alex Christensen.
Add menu items and associated plumbing for support for books highlights in modern WebKit.
* FeatureFlags/WebKit.plist:
* Shared/API/c/WKContextMenuItemTypes.h:
* Shared/API/c/WKSharedAPICast.h:
(WebKit::toAPI):
(WebKit::toImpl):
* Shared/Cocoa/WebPreferencesDefaultValuesCocoa.mm:
(WebKit::isFeatureFlagEnabled):
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultAppHighlightsEnabled):
* Shared/WebPreferencesDefaultValues.h:
* UIProcess/API/Cocoa/WKMenuItemIdentifiers.mm:
* UIProcess/API/Cocoa/WKMenuItemIdentifiersPrivate.h:
* UIProcess/mac/WebContextMenuProxyMac.mm:
(WebKit::menuItemIdentifier):
2020-11-16 Trevor Stevenson <trevor_s@apple.com>
Add didReceiveResponseForResource to WKWebProcessPlugInLoadDelegate
https://bugs.webkit.org/show_bug.cgi?id=218925
Reviewed by Alex Christensen.
Expose the callback for didReceiveResponseForResource in WKWebProcessPlugInLoadDelegate
for use in the injected bundle.
* WebProcess/InjectedBundle/API/Cocoa/WKWebProcessPlugInLoadDelegate.h:
* WebProcess/InjectedBundle/API/mac/WKWebProcessPlugInBrowserContextController.mm:
(didReceiveResponseForResource):
(setUpResourceLoadClient):
2020-11-16 Carlos Garcia Campos <cgarcia@igalia.com>
WebDriver: do not focus the main frame when switching to a window
https://bugs.webkit.org/show_bug.cgi?id=218383
Reviewed by Brian Burg.
The spec doesn't say we should focus the main frame. This is confusing WPT test
infrastructure/testdriver/actions/crossOrigin.sub.html that first focuses an input in an iframe and then it send
keyboard events to the window (after switching to the window again even when it's already the current one, that
causes the iframe focus to be lost).
* UIProcess/Automation/WebAutomationSession.cpp:
(WebKit::WebAutomationSession::switchToBrowsingContext):
* WebProcess/Automation/WebAutomationSessionProxy.cpp:
* WebProcess/Automation/WebAutomationSessionProxy.h:
* WebProcess/Automation/WebAutomationSessionProxy.messages.in:
2020-11-16 Kimmo Kinnunen <kkinnunen@apple.com>
Final refactor for WebGL implementation to use only GraphicsContextGL
https://bugs.webkit.org/show_bug.cgi?id=218333
Reviewed by Simon Fraser.
Change MediaPlayerPrivate to use GraphicsContextGL abstract class
instead of GraphicsContextGLOpenGL concrete class, since the
calling code is doing the same change.
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::copyVideoTextureToPlatformTexture):
* WebProcess/GPU/media/MediaPlayerPrivateRemote.h:
2020-11-16 Devin Rousso <drousso@apple.com>
[iPad] nhl.com cannot select new video to play while video is playing with desktop UA
https://bugs.webkit.org/show_bug.cgi?id=218868
<rdar://problem/67823088>
Reviewed by Antoine Quint.
The desktop "version" of this site uses custom media controls that don't work well on iOS
because they rely on mouse events (which is why it was quirked in r255592). The mobile
"version" uses the default/native media controls, so there is no issue.
Currently, when tapping on another video with a desktop UA, the site uses its touch event
handlers for mouse events. These handlers call `Event.prototype.preventDefault` if the event
is not a touch event, meaning that `"click"` is not fired. This prevents new videos from
being played.
Rather than limit the quirk based on some DOM state (which is fragile if the site changes),
just have the entire site default to a mobile UA.
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::desktopClassBrowsingRecommendedForRequest):
2020-11-16 Zan Dobersek <zdobersek@igalia.com>
[CoordinatedGraphics] Adjust client resizing, render-scope notifying in ThreadedCompositor::renderLayerTree()
https://bugs.webkit.org/show_bug.cgi?id=218701
Reviewed by Carlos Garcia Campos.
Rework the client operations in ThreadedCompositor::renderLayerTree() so
that client resizing is done before the will-render client notification
is dispatched. This ensures everything in the will-render-to-did-render
scope is done for a well-specified size from the client's point of view.
Similar to this change, the GL viewport update is split from the client
resize operation and moved into the aforementioned rendering scope. This
change is mostly cosmetic, but it's done to neatly package all the GL
operations into that scope.
* Shared/CoordinatedGraphics/threadedcompositor/ThreadedCompositor.cpp:
(WebKit::ThreadedCompositor::renderLayerTree):
2020-11-15 Tim Horton <timothy_horton@apple.com>
Initial implementation of DOM rendering via the GPU process
https://bugs.webkit.org/show_bug.cgi?id=218928
Reviewed by Simon Fraser.
* Shared/RemoteLayerTree/RemoteLayerBackingStore.h:
(WebKit::RemoteLayerBackingStore::hasFrontBuffer const):
(WebKit::RemoteLayerBackingStore::Buffer::operator bool const):
* Shared/RemoteLayerTree/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::encode const):
(WebKit::RemoteLayerBackingStore::decode):
(WebKit::RemoteLayerBackingStore::pixelFormat const):
(WebKit::RemoteLayerBackingStore::bytesPerPixel const):
(WebKit::RemoteLayerBackingStore::swapToValidFrontBuffer):
(WebKit::RemoteLayerBackingStore::display):
(WebKit::RemoteLayerBackingStore::applyBackingStoreToLayer):
(WebKit::RemoteLayerBackingStore::takePendingFlusher):
(WebKit::RemoteLayerBackingStore::setBufferVolatility):
(WebKit::RemoteLayerBackingStore::Buffer::discard):
(WebKit::RemoteLayerBackingStore::drawInContext): Deleted.
(WebKit::RemoteLayerBackingStore::takeFrontContextPendingFlush): Deleted.
(WebKit::RemoteLayerBackingStore::surfaceBufferFormat const): Deleted.
Reimplement RemoteLayerBackingStore in terms of ImageBuffer instead of
using ShareableBitmap + IOSurface explicitly.
We use ConcreteImageBuffer versions of the shareable backends in the WebContent process
in the non-GPU-process case, and normal DisplayList-backed RemoteImageBufferProxies
in the GPU process case.
* UIProcess/ViewSnapshotStore.h:
* UIProcess/mac/ViewSnapshotStoreMac.mm:
(WebKit::ViewSnapshot::setVolatile):
(WebKit::ViewSnapshot::asLayerContents):
* WebProcess/GPU/graphics/ImageBufferBackendHandle.h:
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
(WebKit::ThreadSafeRemoteImageBufferFlusher::ThreadSafeRemoteImageBufferFlusher):
Add a hack to emulate ThreadSafeRemoteImageBufferFlusher; right now,
we'll synchronously flush on the main thread when creating the flusher;
a future patch will move this into the flush() method so that RemoteLayerBackingStore
can do it on a secondary queue.
(WebKit::RemoteImageBufferProxy::createImageBufferBackendHandle):
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::createImageBuffer const):
(WebKit::WebChromeClient::ensureRemoteRenderingBackendProxy const): Deleted.
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeContext.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeContext.mm:
(WebKit::RemoteLayerTreeContext::ensureRemoteRenderingBackendProxy):
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.mm:
(WebKit::RemoteLayerTreeDrawingArea::updateRendering):
(WebKit::RemoteLayerTreeDrawingArea::BackingStoreFlusher::create):
(WebKit::RemoteLayerTreeDrawingArea::BackingStoreFlusher::BackingStoreFlusher):
(WebKit::RemoteLayerTreeDrawingArea::BackingStoreFlusher::flush):
Reimplement BackingStoreFlusher in terms of the abstract ThreadSafeImageBufferFlusher.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::ensureRemoteRenderingBackendProxy):
* WebProcess/WebPage/WebPage.h:
Move RemoteRenderingBackendProxy from WebChromeClient to WebPage, so that
other non-WebCore clients can use it without having to think about WebChromeClient.
2020-11-14 Zan Dobersek <zdobersek@igalia.com>
Unreviewed, suppressing GCC compilation warnings.
* UIProcess/API/wpe/ScrollGestureController.cpp:
(WebKit::ScrollGestureController::handleEvent):
Explicitly store the std::abs() results as uint32_t values. This avoids
comparison between the otherwise-signed return values and the unsigned
constants.
2020-11-13 Sihui Liu <sihui_liu@apple.com>
Implement basic permission check for SpeechRecognition
https://bugs.webkit.org/show_bug.cgi?id=218476
<rdar://problem/71222638>
Reviewed by Youenn Fablet.
Introduce SpeechRecognitionPermissionManager, which checks and requests speech recognition permissions before
we actually start capturing audio and perform recognition. SpeechRecognitionPermissionManager is per-page, like
SpeechRecognitionServer. The checks include:
1. Sandbox requirement for microphone
2. TCC check for microphone
3. TCC check for SFSpeechRecognizer
4. User permission on speech recognition for origin
Add a delegate function for requesting user permission. By default, user permission is not granted.
API test: WebKit2.SpeechRecognitionUserPermissionPersistence
* Headers.cmake:
* Shared/API/APIObject.h:
* Shared/API/c/WKBase.h:
* Sources.txt:
* SourcesCocoa.txt:
* UIProcess/API/APIUIClient.h:
(API::UIClient::decidePolicyForSpeechRecognitionPermissionRequest):
* UIProcess/API/C/WKAPICast.h:
* UIProcess/API/C/WKPage.cpp:
(WKPageSetPageUIClient):
* UIProcess/API/C/WKPageUIClient.h:
* UIProcess/API/C/WKSpeechRecognitionPermissionCallback.cpp: Added.
(WKSpeechRecognitionPermissionCallbackGetTypeID):
(WKSpeechRecognitionPermissionCallbackComplete):
* UIProcess/API/C/WKSpeechRecognitionPermissionCallback.h: Added.
* UIProcess/API/Cocoa/WKPreferences.mm:
(-[WKPreferences _speechRecognitionEnabled]):
(-[WKPreferences _setSpeechRecognitionEnabled:]):
* UIProcess/API/Cocoa/WKPreferencesPrivate.h:
* UIProcess/API/Cocoa/WKUIDelegatePrivate.h:
* UIProcess/Cocoa/MediaPermissionUtilities.mm: Added.
(WebKit::checkSandboxRequirementForType):
(WebKit::checkUsageDescriptionStringForType):
(WebKit::checkUsageDescriptionStringForSpeechRecognition):
(WebKit::requestAVCaptureAccessForType):
(WebKit::checkAVCaptureAccessForType):
(WebKit::requestSpeechRecognitionAccess):
(WebKit::checkSpeechRecognitionServiceAccess):
* UIProcess/Cocoa/UIDelegate.h:
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::UIClient::decidePolicyForSpeechRecognitionPermissionRequest):
* UIProcess/Cocoa/UserMediaPermissionRequestManagerProxy.mm:
(WebKit::UserMediaPermissionRequestManagerProxy::permittedToCaptureAudio):
(WebKit::UserMediaPermissionRequestManagerProxy::permittedToCaptureVideo):
(WebKit::UserMediaPermissionRequestManagerProxy::requestSystemValidation):
(WebKit::requestAVCaptureAccessForMediaType): Deleted.
* UIProcess/Cocoa/WebProcessProxyCocoa.mm:
* UIProcess/MediaPermissionUtilities.h: Copied from Added.
* UIProcess/SpeechRecognitionPermissionManager.cpp: Added.
(WebKit::computeMicrophoneAccess):
(WebKit::computeSpeechRecognitionServiceAccess):
(WebKit::SpeechRecognitionPermissionManager::SpeechRecognitionPermissionManager):
(WebKit::SpeechRecognitionPermissionManager::~SpeechRecognitionPermissionManager):
(WebKit::SpeechRecognitionPermissionManager::request):
(WebKit::SpeechRecognitionPermissionManager::startNextRequest):
(WebKit::SpeechRecognitionPermissionManager::startProcessingRequest):
(WebKit::SpeechRecognitionPermissionManager::continueProcessingRequest):
(WebKit::SpeechRecognitionPermissionManager::completeCurrentRequest):
(WebKit::SpeechRecognitionPermissionManager::requestSpeechRecognitionServiceAccess):
(WebKit::SpeechRecognitionPermissionManager::requestMicrophoneAccess):
(WebKit::SpeechRecognitionPermissionManager::requestUserPermission):
* UIProcess/SpeechRecognitionPermissionManager.h: Added.
* UIProcess/SpeechRecognitionPermissionRequest.h: Added.
(WebKit::SpeechRecognitionPermissionRequest::create):
(WebKit::SpeechRecognitionPermissionRequest::complete):
(WebKit::SpeechRecognitionPermissionRequest::origin const):
(WebKit::SpeechRecognitionPermissionRequest::SpeechRecognitionPermissionRequest):
(WebKit::SpeechRecognitionPermissionCallback::create):
(WebKit::SpeechRecognitionPermissionCallback::complete):
(WebKit::SpeechRecognitionPermissionCallback::SpeechRecognitionPermissionCallback):
* UIProcess/SpeechRecognitionServer.cpp:
(WebKit::SpeechRecognitionServer::SpeechRecognitionServer):
(WebKit::SpeechRecognitionServer::start):
(WebKit::SpeechRecognitionServer::requestPermissionForRequest):
(WebKit::SpeechRecognitionServer::stop):
(WebKit::SpeechRecognitionServer::abort):
(WebKit::SpeechRecognitionServer::invalidate):
(WebKit::SpeechRecognitionServer::handleRequest):
(WebKit::SpeechRecognitionServer::stopRequest):
(WebKit::SpeechRecognitionServer::abortRequest):
(WebKit::SpeechRecognitionServer::sendUpdate):
(WebKit::SpeechRecognitionServer::processNextPendingRequestIfNeeded): Deleted.
(WebKit::SpeechRecognitionServer::removePendingRequest): Deleted.
(WebKit::SpeechRecognitionServer::startPocessingRequest): Deleted.
(WebKit::SpeechRecognitionServer::stopProcessingRequest): Deleted.
* UIProcess/SpeechRecognitionServer.h:
* UIProcess/SpeechRecognitionServer.messages.in:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didChangeMainDocument):
(WebKit::WebPageProxy::resetState):
(WebKit::WebPageProxy::requestSpeechRecognitionPermission):
* UIProcess/WebPageProxy.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::createSpeechRecognitionServer):
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebCoreSupport/WebSpeechRecognitionConnection.cpp:
(WebKit::WebSpeechRecognitionConnection::start):
(WebKit::WebSpeechRecognitionConnection::didReceiveUpdate):
* WebProcess/WebCoreSupport/WebSpeechRecognitionConnection.h:
2020-11-13 Chris Dumez <cdumez@apple.com>
[GPUProcess] Add basic GPUProcess crash handling for canvas
https://bugs.webkit.org/show_bug.cgi?id=218924
Reviewed by Simon Fraser.
Update RemoteRenderingBackendProxy to be a GPUProcessConnection::Client so that it
gets notified when the IPC connection to the GPUProcess gets severed. When this
happens, RemoteRenderingBackendProxy clears all its SharedMemory handles and
sends IPC messages to the GPU process to recreate all the ImageBuffers we had.
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
(WebKit::RemoteImageBufferProxy::create):
(WebKit::RemoteImageBufferProxy::clearBackend):
(WebKit::RemoteImageBufferProxy::size const):
(WebKit::RemoteImageBufferProxy::renderingMode const):
(WebKit::RemoteImageBufferProxy::colorSpace const):
(WebKit::RemoteImageBufferProxy::pixelFormat const):
(WebKit::RemoteImageBufferProxy::RemoteImageBufferProxy):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::RemoteRenderingBackendProxy):
(WebKit::RemoteRenderingBackendProxy::connectToGPUProcess):
(WebKit::recreateImage):
(WebKit::RemoteRenderingBackendProxy::gpuProcessConnectionDidClose):
(WebKit::RemoteRenderingBackendProxy::createImageBuffer):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
* WebProcess/GPU/graphics/RemoteResourceCacheProxy.h:
(WebKit::RemoteResourceCacheProxy::imageBuffers const):
2020-11-13 Chris Dumez <cdumez@apple.com>
[GPUProcess] Add basic low memory handling in the GPUProcess
https://bugs.webkit.org/show_bug.cgi?id=218926
<rdar://problem/70300888>
Reviewed by Geoffrey Garen.
Add basic low memory handling in the GPUProcess. For now, the GPUProcess only calls
WTF::releaseFastMallocFreeMemory() on low memory warning.
* GPUProcess/GPUProcess.cpp:
(WebKit::GPUProcess::initializeGPUProcess):
2020-11-13 Kate Cheney <katherine_cheney@apple.com>
Can't login to Microsoft Teams
https://bugs.webkit.org/show_bug.cgi?id=218778
<rdar://problem/36331568>
Reviewed by John Wilander.
This is a temporary quirk to assist a high-traffic website while they
complete the large task of migrating away from login flows that
require third party cookies. This quirk will be removed when the site
is updated.
Create a quirk to call the Storage Access API on behalf of Microsoft.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::grantStorageAccess):
(WebKit::ResourceLoadStatisticsDatabaseStore::grantStorageAccessInternal):
Relax the requirement for user interaction for a specific third party
needing cookies for authentication purposes.
(WebKit::ResourceLoadStatisticsDatabaseStore::clear):
(WebKit::ResourceLoadStatisticsDatabaseStore::cookieAccess):
Relax the requirement for user interaction for a specific third party
needing cookies for authentication purposes.
(WebKit::ResourceLoadStatisticsDatabaseStore::domainsWithStorageAccess const):
(WebKit::ResourceLoadStatisticsDatabaseStore::updateCookieBlocking):
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:
(WebKit::ResourceLoadStatisticsMemoryStore::clear):
(WebKit::ResourceLoadStatisticsMemoryStore::updateCookieBlocking):
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::requestStorageAccessEphemeral):
Relax user interaction for reasons above.
(WebKit::WebResourceLoadStatisticsStore::callUpdatePrevalentDomainsToBlockCookiesForHandler):
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.h:
(WebKit::RegistrableDomainsToBlockCookiesFor::isolatedCopy const):
* Shared/WebProcessDataStoreParameters.h:
(WebKit::WebProcessDataStoreParameters::encode const):
(WebKit::WebProcessDataStoreParameters::decode):
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::setDomainsWithCrossPageStorageAccess):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/Network/NetworkProcessProxy.messages.in:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::webProcessDataStoreParameters):
(WebKit::WebProcessPool::setDomainsWithCrossPageStorageAccess):
* UIProcess/WebProcessPool.h:
* WebProcess/WebCoreSupport/WebResourceLoadObserver.cpp:
(WebKit::WebResourceLoadObserver::hasCrossPageStorageAccess const):
(WebKit::WebResourceLoadObserver::setDomainsWithCrossPageStorageAccess):
* WebProcess/WebCoreSupport/WebResourceLoadObserver.h:
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::setWebsiteDataStoreParameters):
(WebKit::WebProcess::setDomainsWithCrossPageStorageAccess):
Add quirk'd domains to have page level storage access to be able to
use document.cookie.
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
Forward domains with the storage access quirk to the web process to
avoid an async call to the network process.
2020-11-13 Sam Weinig <weinig@apple.com>
Move some more WebKit and WebKitLegacy preferences bound to Settings to WebPreferences.yaml
https://bugs.webkit.org/show_bug.cgi?id=218914
Reviewed by Tim Horton.
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
* UIProcess/WebPageProxy.h:
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::appleMailPaginationQuirkEnabled): Deleted.
(WebKit::WebPageProxy::appleMailLinesClampEnabled): Deleted.
Removes appleMailPaginationQuirkEnabled and appleMailLinesClampEnabled.
appleMailPaginationQuirkEnabled is now handled via the preferences infrastructure
and appleMailLinesClampEnabled was unused.
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultAppleMailPaginationQuirkEnabled):
* Shared/WebPreferencesDefaultValues.h:
Adds default value for AppleMailPaginationQuirkEnabled, which is now in WebPreferences.yaml.
Moved from WebPageProxyMac.mm
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences):
Removes explicit setting of setAppleMailPaginationQuirkEnabled, setContentDispositionAttachmentSandboxEnabled,
setScrollingPerformanceLoggingEnabled and setUseImageDocumentForSubframePDF which are all generated now.
LayoutViewportHeightExpansionFactor was already generated so this call was just redundant.
* WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::setScrollingPerformanceLoggingEnabled): Deleted.
It's also no longer necessary to store m_scrollingPerformanceLoggingEnabled as the value
can be extracted directly from the store in the one place it was read. setScrollingPerformanceLoggingEnabled
was never called, so has been removed.
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::updatePreferences):
Extract scrollingPerformanceLoggingEnabled directly from the store rather than having the page
set it just for this one place to read it.
2020-11-13 Geoffrey Garen <ggaren@apple.com>
Removed DeferrableTaskTimer
https://bugs.webkit.org/show_bug.cgi?id=218874
Reviewed by Chris Dumez.
It was (mostly) redundant.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::updateReportedMediaCaptureState):
* UIProcess/WebPageProxy.h: Use WTF::RunLoop::Timer instead of
WebCore::Timer because WebCore::Timer for WebKit code in the UI process
is a no-no, which can crash if the UI process also uses WebThread.
2020-11-13 Per Arne Vollan <pvollan@apple.com>
[macOS] Issue sandbox extension to the WebContent process for com.apple.lskdd
https://bugs.webkit.org/show_bug.cgi?id=218920
<rdar://problem/69168801>
Reviewed by Geoffrey Garen.
This is a Media related service, and a sandbox extension should be issued to the WebContent process based on GPU runtime settings.
No new tests, covered by existing tests.
* UIProcess/WebPageProxy.cpp:
(WebKit::mediaRelatedMachServices):
* WebProcess/com.apple.WebProcess.sb.in:
2020-11-13 Claudio Saavedra <csaavedra@igalia.com>
Non-unified build fixes.
https://bugs.webkit.org/show_bug.cgi?id=218905
Unreviewed.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
* GPUProcess/graphics/RemoteResourceCache.h:
2020-11-13 Alex Christensen <achristensen@webkit.org>
Remove more SecItemShim leftovers
https://bugs.webkit.org/show_bug.cgi?id=218919
Reviewed by Geoffrey Garen.
* Configurations/SecItemShim.xcconfig: Removed.
* NetworkProcess/EntryPoint/Cocoa/XPCService/NetworkServiceEntryPoint.mm:
(NETWORK_SERVICE_INITIALIZER):
* PlatformMac.cmake:
* WebKit.xcodeproj/project.pbxproj:
2020-11-12 Darin Adler <darin@apple.com>
Remove unused advanced plug-in features: snapshotting and plug-in load policy
https://bugs.webkit.org/show_bug.cgi?id=218835
Reviewed by Tim Horton.
* PluginProcess/mac/PluginProcessMac.mm:
(WebKit::muteAudio): Deleted.
(WebKit::PluginProcess::platformInitializePluginProcess):
* Scripts/webkit/messages.py: Remove PluginProcessType.
* Shared/Plugins/PluginProcessAttributes.h: Ditto.
* Shared/Plugins/PluginProcessCreationParameters.cpp:
(WebKit::PluginProcessCreationParameters::encode const): Ditto.
(WebKit::PluginProcessCreationParameters::decode): Ditto.
* Shared/Plugins/PluginProcessCreationParameters.h: Ditto.
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const): Remove plugInAutoStartOrigins
and pluginLoadClientPolicies.
(WebKit::WebProcessCreationParameters::decode): Ditto.
* Shared/WebProcessCreationParameters.h: Ditto.
* Shared/WebProcessDataStoreParameters.h:
(WebKit::WebProcessDataStoreParameters::encode const): Remove plugInAutoStartOriginHashes.
(WebKit::WebProcessDataStoreParameters::decode): Ditto.
* Sources.txt: Remove PlugInAutoStartProvider.cpp and WebPlugInClient.cpp.
* UIProcess/API/C/WKContext.cpp:
(WKContextCopyPlugInAutoStartOriginHashes): Return nullptr.
(WKContextSetPlugInAutoStartOriginHashes): Do nothing.
(WKContextSetPlugInAutoStartOriginsFilteringOutEntriesAddedAfterTime): Ditto.
(WKContextSetPlugInAutoStartOrigins): Ditto.
* UIProcess/API/C/WKPage.cpp:
(WKPageSetPageNavigationClient): Remove decidePolicyForPluginLoad.
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetPlugInSnapshottingEnabled): Do nothing.
(WKPreferencesGetPlugInSnapshottingEnabled): Return false.
(WKPreferencesSetSnapshotAllPlugIns): Do nothing.
(WKPreferencesGetSnapshotAllPlugIns): Return false.
(WKPreferencesSetAutostartOriginPlugInSnapshottingEnabled): Do nothing.
(WKPreferencesGetAutostartOriginPlugInSnapshottingEnabled): Return false.
(WKPreferencesSetPrimaryPlugInSnapshotDetectionEnabled): Do nothing.
(WKPreferencesGetPrimaryPlugInSnapshotDetectionEnabled): Return false.
* UIProcess/API/C/mac/WKContextPrivateMac.mm:
(WKContextIsPlugInUpdateAvailable): Return false.
(WKContextSetPluginLoadClientPolicy): Do nothing.
(WKContextClearPluginClientPolicies): Do nothing.
* UIProcess/API/Cocoa/WKPreferences.mm:
(-[WKPreferences _setPlugInSnapshottingEnabled:]): Do nothing.
(-[WKPreferences _plugInSnapshottingEnabled]): Return NO.
* UIProcess/API/Cocoa/WKProcessPool.mm:
(toPluginLoadClientPoliciesHashMap): Deleted.
(policiesHashMapToDictionary): Deleted.
(-[WKProcessPool _resetPluginLoadClientPolicies:]): Do nothing.
(-[WKProcessPool _pluginLoadClientPolicies]): Return an empty dictionary.
* UIProcess/Cocoa/NavigationState.h: Remove didFailToInitializePlugIn,
didBlockInsecurePluginVersion, decidePolicyForPluginLoad,
webViewDidFailToInitializePlugInWithInfo,
webViewDidBlockInsecurePluginVersionWithInfo, and
webViewDecidePolicyForPluginLoadWithCurrentPolicyPluginInfoCompletionHandler.
* UIProcess/Cocoa/NavigationState.mm:
(WebKit::NavigationState::setNavigationDelegate): Updated for abvoe removals.
(WebKit::NavigationState::NavigationClient::didFailToInitializePlugIn): Deleted.
(WebKit::NavigationState::NavigationClient::didBlockInsecurePluginVersion): Deleted.
(WebKit::pluginModuleLoadPolicy): Deleted.
(WebKit::wkPluginModuleLoadPolicy): Deleted.
(WebKit::NavigationState::NavigationClient::decidePolicyForPluginLoad): Deleted.
* UIProcess/Plugins/PlugInAutoStartProvider.cpp: Removed.
* UIProcess/Plugins/PlugInAutoStartProvider.h: Removed.
* UIProcess/Plugins/PluginProcessManager.cpp:
(WebKit::PluginProcessManager::pluginProcessToken): Removed the
pluginProcessType argument.
(WebKit::PluginProcessManager::fetchWebsiteData): Ditto.
(WebKit::PluginProcessManager::deleteWebsiteData): Ditto.
(WebKit::PluginProcessManager::deleteWebsiteDataForHostNames): Ditto.
* UIProcess/Plugins/PluginProcessManager.h: Ditto.
* UIProcess/Plugins/PluginProcessProxy.cpp:
(WebKit::PluginProcessProxy::didFinishLaunching): Removed the code to
handle PluginProcessType::Snapshot.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::findPlugin): Removed the processType argument.
* UIProcess/WebPageProxy.h: Ditto.
* UIProcess/WebPageProxy.messages.in: Ditto.
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::webProcessDataStoreParameters): Removed
plugInAutoStartOriginHashes.
(WebKit::WebProcessPool::initializeNewWebProcess): Removed
plugInAutoStartOrigins and pluginLoadClientPolicies.
(WebKit::WebProcessPool::plugInAutoStartOriginHashes const): Deleted.
(WebKit::WebProcessPool::setPlugInAutoStartOriginHashes): Deleted.
(WebKit::WebProcessPool::setPlugInAutoStartOrigins): Deleted.
(WebKit::WebProcessPool::setPlugInAutoStartOriginsFilteringOutEntriesAddedAfterTime): Deleted.
(WebKit::WebProcessPool::setPluginLoadClientPolicy): Deleted.
(WebKit::WebProcessPool::resetPluginLoadClientPolicies): Deleted.
(WebKit::WebProcessPool::clearPluginClientPolicies): Deleted.
* UIProcess/WebProcessPool.h: Removed setPluginLoadClientPolicy,
resetPluginLoadClientPolicies, clearPluginClientPolicies,
pluginLoadClientPolicies, plugInAutoStartOriginHashes,
setPlugInAutoStartOriginHashes, setPlugInAutoStartOrigins,
setPlugInAutoStartOriginsFilteringOutEntriesAddedAfterTime,
plugInAutoStartProvider, m_plugInAutoStartProvider, and
m_pluginLoadClientPolicies.
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::addPlugInAutoStartOriginHash): Deleted.
(WebKit::WebProcessProxy::plugInDidReceiveUserInteraction): Deleted.
* UIProcess/WebProcessProxy.h: Ditto.
* UIProcess/WebProcessProxy.messages.in: Ditto.
* WebKit.xcodeproj/project.pbxproj: Removed PlugInAutoStartProvider.cpp/h
and WebPlugInClient.cpp/h.
* WebProcess/InjectedBundle/InjectedBundlePageUIClient.h:
Removed include of RenderSnapshottedPlugIn.h.
* WebProcess/Plugins/Netscape/mac/PluginProxyMac.mm:
(WebKit::PluginProxy::pluginLayer): Removed code to handle
m_isRestartedProcess.
* WebProcess/Plugins/PDF/PDFPlugin.h: Removed shouldAlwaysAutoStart.
* WebProcess/Plugins/Plugin.h: Ditto.
* WebProcess/Plugins/PluginProxy.cpp:
(WebKit::PluginProxy::create): Removed isRestartedProcess argument.
(WebKit::PluginProxy::PluginProxy): Ditto. Also moved initialization
to the class definition.
* WebProcess/Plugins/PluginProxy.h: Updated for the above changes.
* WebProcess/Plugins/PluginView.cpp:
(WebKit::PluginView::PluginView): Removed m_pluginSnapshotTimer.
(WebKit::PluginView::recreateAndInitialize): Removed code to manage
m_pluginSnapshotTimer.
(WebKit::PluginView::initializePlugin): Removed code to set
m_didPlugInStartOffScreen.
(WebKit::PluginView::didInitializePlugin): Removed code to handle
restarting plug-ins.
(WebKit::PluginView::paint): Ditto.
(WebKit::PluginView::handleEvent): Removed calls to
pluginDidReceiveUserInteraction.
(WebKit::PluginView::shouldNotAddLayer const): Return false.
(WebKit::PluginView::invalidateRect): Removed code to handle
restarting plug-ins.
(WebKit::PluginView::isAcceleratedCompositingEnabled): Removed
some Flash-specific code.
(WebKit::isAlmostSolidColor): Deleted.
(WebKit::PluginView::pluginSnapshotTimerFired): Deleted.
(WebKit::PluginView::beginSnapshottingRunningPlugin): Deleted.
(WebKit::PluginView::shouldAlwaysAutoStart const): Deleted.
(WebKit::PluginView::pluginDidReceiveUserInteraction): Deleted.
* WebProcess/Plugins/PluginView.h: Removed pluginSnapshotTimerFired,
pluginDidReceiveUserInteraction, shouldAlwaysAutoStart,
beginSnapshottingRunningPlugin, m_didPlugInStartOffScreen,
m_pluginSnapshotTimer, m_countSnapshotRetries, m_didReceiveUserInteraction.
* WebProcess/Plugins/WebPluginInfoProvider.cpp:
(WebKit::WebPluginInfoProvider::setPluginLoadClientPolicy): Deleted.
(WebKit::WebPluginInfoProvider::clearPluginClientPolicies): Deleted.
(WebKit::WebPluginInfoProvider::populatePluginCache): Do not call
pluginLoadClientPolicyForHost.
(WebKit::WebPluginInfoProvider::pluginLoadClientPolicyForHost const): Deleted.
(WebKit::WebPluginInfoProvider::longestMatchedWildcardHostForHost const): Deleted.
(WebKit::WebPluginInfoProvider::replaceHostWithMatchedWildcardHost const): Deleted.
* WebProcess/Plugins/WebPluginInfoProvider.h: Removed setPluginLoadClientPolicy,
clearPluginClientPolicies, pluginLoadClientPolicyForHost,
longestMatchedWildcardHostForHost, replaceHostWithMatchedWildcardHost,
PluginLoadClientPoliciesByBundleVersion, PluginPolicyMapsByIdentifier, and
m_hostsToPluginIdentifierData.
* WebProcess/WebCoreSupport/WebPlugInClient.cpp: Removed.
* WebProcess/WebCoreSupport/WebPlugInClient.h: Removed.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::WebPage): Removed code to initialize
m_determinePrimarySnapshottedPlugInTimer, plugInClient, and processType.
(WebKit::WebPage::createPlugin): Removed isRestartedProcess.
(WebKit::WebPage::close): Removed m_determinePrimarySnapshottedPlugInTimer code.
(WebKit::WebPage::addPluginView): Ditto.
(WebKit::WebPage::removePluginView): Ditto.
(WebKit::WebPage::canPluginHandleResponse): Removed plug-in process type.
(WebKit::WebPage::didCommitLoad): Removed call to resetPrimarySnapshottedPlugIn.
(WebKit::WebPage::didFinishLoad): Removed m_determinePrimarySnapshottedPlugInTimer
code.
(WebKit::WebPage::determinePrimarySnapshottedPlugInTimerFired): Deleted.
(WebKit::WebPage::determinePrimarySnapshottedPlugIn): Deleted.
(WebKit::WebPage::resetPrimarySnapshottedPlugIn): Deleted.
(WebKit::WebPage::matchesPrimaryPlugIn const): Deleted.
(WebKit::WebPage::plugInIntersectsSearchRect): Deleted.
(WebKit::WebPage::plugInIsPrimarySize): Deleted.
* WebProcess/WebPage/WebPage.h: Removed determinePrimarySnapshottedPlugIn,
determinePrimarySnapshottedPlugInTimerFired, resetPrimarySnapshottedPlugIn,
matchesPrimaryPlugIn, plugInIntersectsSearchRect, plugInIsPrimarySize,
m_readyToFindPrimarySnapshottedPlugin, m_didFindPrimarySnapshottedPlugin,
m_numberOfPrimarySnapshotDetectionAttempts, m_primaryPlugInPageOrigin,
m_primaryPlugInOrigin, m_primaryPlugInMimeType, and
m_determinePrimarySnapshottedPlugInTimer.
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::initializeWebProcess): Removed code to set
m_plugInAutoStartOrigins and call to resetPluginLoadClientPolicies.
(WebKit::WebProcess::setWebsiteDataStoreParameters): Removed call
to resetPlugInAutoStartOriginHashes.
(WebKit::addCaseFoldedCharacters): Deleted.
(WebKit::hashForPlugInOrigin): Deleted.
(WebKit::WebProcess::isPlugInAutoStartOriginHash): Deleted.
(WebKit::WebProcess::shouldPlugInAutoStartFromOrigin): Deleted.
(WebKit::WebProcess::plugInDidStartFromOrigin): Deleted.
(WebKit::WebProcess::didAddPlugInAutoStartOriginHash): Deleted.
(WebKit::WebProcess::resetPlugInAutoStartOriginHashes): Deleted.
(WebKit::WebProcess::plugInDidReceiveUserInteraction): Deleted.
(WebKit::WebProcess::setPluginLoadClientPolicy): Deleted.
(WebKit::WebProcess::resetPluginLoadClientPolicies): Deleted.
(WebKit::WebProcess::clearPluginClientPolicies): Deleted.
* WebProcess/WebProcess.h: Removed shouldPlugInAutoStartFromOrigin,
plugInDidStartFromOrigin, plugInDidReceiveUserInteraction,
setPluginLoadClientPolicy, resetPluginLoadClientPolicies,
clearPluginClientPolicies, isPlugInAutoStartOriginHash,
didAddPlugInAutoStartOriginHash, resetPlugInAutoStartOriginHashes,
m_plugInAutoStartOriginHashes, and m_plugInAutoStartOrigins.
* WebProcess/WebProcess.messages.in: Removed
DidAddPlugInAutoStartOriginHash, ResetPlugInAutoStartOriginHashes,
SetPluginLoadClientPolicy, ResetPluginLoadClientPolicies, and
ClearPluginClientPolicies.
2020-11-12 Simon Fraser <simon.fraser@apple.com>
Force wheel event listeners on the root to be passive
https://bugs.webkit.org/show_bug.cgi?id=218842
<rdar://problem/71312668>
Reviewed by Chris Dumez.
Following Blink (https://www.chromestatus.com/feature/6662647093133312) force 'wheel' and
'mousewheel' event listeners on root objects (window, document and body) to be passive if
they were not explicitly registered as non-passive.
This behavior is controlled by an experimental feature flag, and a linked-on-or-after check
to avoid changing behavior for apps that embed WebKit until they link against new SDKs.
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultCSSOMViewScrollingAPIEnabled):
(WebKit::defaultPassiveWheelListenersAsDefaultOnDocument):
* Shared/WebPreferencesDefaultValues.h:
2020-11-13 Eric Carlson <eric.carlson@apple.com>
Add _STAttributionDisplayName to macOS WebContent Info.plist
https://bugs.webkit.org/show_bug.cgi?id=218900
Reviewed by Youenn Fablet.
* WebProcess/EntryPoint/Cocoa/XPCService/WebContentService/Info-OSX.plist:
2020-11-13 Miguel Gomez <magomez@igalia.com>
[GTK][WPE] CSS backdrop overlay corners are not rounded on results.webkit.org
https://bugs.webkit.org/show_bug.cgi?id=215445
Reviewed by Carlos Garcia Campos.
Pass the backdropFiltersRect from the state to TextureMapperLayer.
* Shared/CoordinatedGraphics/CoordinatedGraphicsScene.cpp:
(WebKit::CoordinatedGraphicsScene::updateSceneState):
2020-11-13 Kimmo Kinnunen <kkinnunen@apple.com>
Remove SecItemShim.dylib from network service link
https://bugs.webkit.org/show_bug.cgi?id=218892
Reviewed by Antti Koivisto.
Fixes a build failure after
https://bugs.webkit.org/show_bug.cgi?id=218862
* Configurations/NetworkService.xcconfig:
2020-11-12 Youenn Fablet <youenn@apple.com>
Add infrastructure for WebRTC transforms
https://bugs.webkit.org/show_bug.cgi?id=218750
Reviewed by Eric Carlson.
Remove no longer needed const casting.
* Configurations/WebKit.xcconfig:
* WebProcess/GPU/webrtc/LibWebRTCCodecs.cpp:
(WebKit::LibWebRTCCodecs::completedEncoding):
2020-11-12 Alex Christensen <achristensen@webkit.org>
Remove SecItemShim.dylib
https://bugs.webkit.org/show_bug.cgi?id=218862
Reviewed by Alexey Proskuryakov.
This is dead code since r269697.
* Shared/mac/SecItemShim.cpp:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/mac/SecItemShimLibrary.h: Removed.
* WebProcess/mac/SecItemShimLibrary.mm: Removed.
2020-11-12 Tim Horton <timothy_horton@apple.com>
canvas-createPattern-video-modify-crash-log.txt asserts on debug GPUProcess bots
https://bugs.webkit.org/show_bug.cgi?id=218872
Reviewed by Simon Fraser.
* WebProcess/GPU/graphics/ImageBufferShareableBitmapBackend.cpp:
(WebKit::ImageBufferShareableBitmapBackend::create):
This assertion is wrong; make it the same as the one in ImageBufferCGBitmapBackend.
2020-11-12 Said Abou-Hallawa <said@apple.com>
[GPU Process] GraphicsContext::drawPattern() should take a NativeImage argument
https://bugs.webkit.org/show_bug.cgi?id=218865
Reviewed by Simon Fraser.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::decodeItem):
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
2020-11-12 Per Arne Vollan <pvollan@apple.com>
[iOS] IOKit sandbox violation when enabling all GPU runtime flags
https://bugs.webkit.org/show_bug.cgi?id=218820
<rdar://problem/71296116>
Reviewed by Brent Fulgham.
When enabling all GPU runtime flags on iOS, there is an iokit-open sandbox violation for the IOKit class AGXDeviceUserClient
in the WebContent process. When all GPU runtime flags are enabled, IOKit extensions are not provided to the WebContent process,
which should then fall back to a set of rules allowing use of these IOKit classes, with logging. There seems to be a problem
with the fallback for this specific IOKit class, which this patch addresses by adding an extra requirement to the rule. The
extra requirement is that IOKit extensions have not been provided for these IOKit classes.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-11-12 Chris Dumez <cdumez@apple.com>
[GPUProcess] Add basic GPUProcess crash handling for media playback
https://bugs.webkit.org/show_bug.cgi?id=218825
Reviewed by Eric Carlson.
When the GPU process crashes and there are pending media players, we
now relaunch the GPU process and ask those media players to reconstuct
their private media players, reload the file and resume playback if
necessary. This ensures that in case of a GPU process crash while the
user is playing a video, the video just seamlessly keeps playing from
where it was before the crash.
Before this patch, we would end up busy looping and the WebContent process
would use 100% CPU after a GPU process crash.
* WebProcess/GPU/GPUProcessConnection.h:
(WebKit::GPUProcessConnection::addClient):
(WebKit::GPUProcessConnection::removeClient):
* WebProcess/GPU/media/MediaPlayerPrivateRemote.h:
* WebProcess/GPU/media/RemoteMediaPlayerManager.cpp:
(WebKit::proxyConfigurationForPlayer):
(WebKit::RemoteMediaPlayerManager::gpuProcessConnection const):
(WebKit::RemoteMediaPlayerManager::gpuProcessConnectionDidClose):
* WebProcess/GPU/media/RemoteMediaPlayerManager.h:
2020-11-12 Chris Dumez <cdumez@apple.com>
ASSERTION FAILED: isValidIdentifier(m_identifier) seen with TestWebKitAPI.GPUProcess.WebProcessTerminationAfterTooManyGPUProcessCrashes
https://bugs.webkit.org/show_bug.cgi?id=218856
<rdar://problem/71331809>
Reviewed by Tim Horton.
The API test is repeatedly killing the GPU process. As a result, it is possible for the
GPUProcess to crash while RemoteAudioDestinationProxy::connectToGPUProcess() is in the
middle of its RemoteAudioDestinationManager::CreateAudioDestination() synchronous IPC.
The function would fail to check if the IPC was successful and proceed with an invalid
destinationID in such cases, causing the crash.
We now check if the sendSync() was successful. If it wasn't we now log an error and
return early. RemoteAudioDestinationManager::gpuProcessConnectionDidClose() will get
called later on to notify us that the GPU Process crashed and it will call
connectToGPUProcess() again.
No new tests, covered by existing API test that is flakily crashing.
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::connectToGPUProcess):
2020-11-12 Youenn Fablet <youenn@apple.com>
WebProcess should process WebRTC codecs IPC messages from the GPU Process in a background thread
https://bugs.webkit.org/show_bug.cgi?id=218790
Reviewed by Eric Carlson.
Use a WorkQueue to process GPU WebRTC codec IPC messages.
For that reason, handle encoder and decoder map within that queue.
Covered by existing tests.
* GPUProcess/webrtc/LibWebRTCCodecsProxy.mm:
(WebKit::LibWebRTCCodecsProxy::setEncodeRates):
* WebProcess/GPU/GPUProcessConnection.cpp:
(WebKit::GPUProcessConnection::dispatchMessage):
* WebProcess/GPU/webrtc/LibWebRTCCodecs.cpp:
(WebKit::LibWebRTCCodecs::LibWebRTCCodecs):
(WebKit::LibWebRTCCodecs::~LibWebRTCCodecs):
(WebKit::LibWebRTCCodecs::setConnection):
(WebKit::LibWebRTCCodecs::createDecoder):
(WebKit::LibWebRTCCodecs::releaseDecoder):
(WebKit::LibWebRTCCodecs::failedDecoding):
(WebKit::LibWebRTCCodecs::completedDecoding):
(WebKit::LibWebRTCCodecs::createEncoder):
(WebKit::LibWebRTCCodecs::releaseEncoder):
(WebKit::LibWebRTCCodecs::initializeEncoder):
(WebKit::LibWebRTCCodecs::setEncodeRates):
(WebKit::LibWebRTCCodecs::completedEncoding):
(WebKit::LibWebRTCCodecs::dispatchToThread):
* WebProcess/GPU/webrtc/LibWebRTCCodecs.h:
2020-11-12 Diego Pino Garcia <dpino@igalia.com>
[GTK] Unreviewed, non-unified build fix.
* WebProcess/Gamepad/WebGamepadProvider.cpp: Add missing Logging.h header.
2020-11-12 Zalan Bujtas <zalan@apple.com>
Show legacy line layout visual coverage instead of "simple line" layout.
https://bugs.webkit.org/show_bug.cgi?id=218695
Reviewed by Antti Koivisto.
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetLegacyLineLayoutVisualCoverageEnabled):
(WKPreferencesGetLegacyLineLayoutVisualCoverageEnabled):
(WKPreferencesSetSimpleLineLayoutDebugBordersEnabled): Deleted.
(WKPreferencesGetSimpleLineLayoutDebugBordersEnabled): Deleted.
* UIProcess/API/C/WKPreferencesRefPrivate.h:
* UIProcess/API/Cocoa/WKPreferences.mm:
(-[WKPreferences _legacyLineLayoutVisualCoverageEnabled]):
(-[WKPreferences _setLegacyLineLayoutVisualCoverageEnabled:]):
(-[WKPreferences _simpleLineLayoutDebugBordersEnabled]): Deleted.
(-[WKPreferences _setSimpleLineLayoutDebugBordersEnabled:]): Deleted.
* UIProcess/API/Cocoa/WKPreferencesPrivate.h:
2020-11-12 Said Abou-Hallawa <said@apple.com>
[GPU Process] Implement DisplayList::ClipToImageBuffer item
https://bugs.webkit.org/show_bug.cgi?id=218843
Reviewed by Tim Horton.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::decodeItem):
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
2020-11-12 Zan Dobersek <zdobersek@igalia.com>
Unreviewed follow-up to r268965.
* UIProcess/API/glib/WebKitSettings.cpp:
(webkit_settings_get_enable_accelerated_2d_canvas):
Use the FALSE identifier, following GLib styling guidelines.
2020-11-12 Alexander Mikhaylenko <exalm7659@gmail.com>
[GTK] Swipe navigation shadow is blurry on hidpi
https://bugs.webkit.org/show_bug.cgi?id=218800
Reviewed by Carlos Garcia Campos.
* UIProcess/gtk/ViewGestureControllerGtk.cpp:
(WebKit::createElementPattern): Scale surface size and set its scale factor.
(WebKit::ViewGestureController::beginSwipeGesture): Pass scale factor to createElementPattern().
2020-11-11 Brian Burg <bburg@apple.com>
REGRESSION(r267411): Unable to "Start Element Selection" with Web Inspector closed
https://bugs.webkit.org/show_bug.cgi?id=218838
<rdar://problem/71158215>
Reviewed by Devin Rousso.
The inspector side of the connection between WebInspector (inspected WebProcess) and
WebInspectorUI (inspector WebProcess) was mistakenly removed in r267411. Revert those changes.
Without a valid frontend connection, WebInspector cannot forward some IPC messages
to WebInspectorUI including StartElementSelection.
* WebProcess/Inspector/WebInspectorUI.cpp:
(WebKit::WebInspectorUI::updateConnection):
(WebKit::WebInspectorUI::closeWindow):
* WebProcess/Inspector/WebInspectorUI.h:
2020-11-11 Said Abou-Hallawa <said@apple.com>
[GPU Process] Delete the DisplayList items: DrawImage and DrawTiledImage
https://bugs.webkit.org/show_bug.cgi?id=218839
Reviewed by Tim Horton.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::decodeItem):
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
2020-11-11 Chris Dumez <cdumez@apple.com>
Add RunLoop::Timer constructor taking in a WTF::Function
https://bugs.webkit.org/show_bug.cgi?id=218828
Reviewed by Darin Adler.
Use the new RunLoop::Timer constructor to simplify the code a bit.
* UIProcess/WebProcessPool.cpp:
(WebKit::m_resetGPUProcessCrashCountTimer):
* UIProcess/WebProcessPool.h:
2020-11-11 John Wilander <wilander@apple.com>
PCM: Change from ad-click-attribution to private-click-measurement (in all forms, including .well-known URL)
https://bugs.webkit.org/show_bug.cgi?id=218730
<rdar://problem/71094296>
Reviewed by Alex Christensen.
Change to the official name of the proposed standard Private Click Measurement
https://github.com/privacycg/private-click-measurement.
This includes a change of the reporting URL from
"/.well-known/ad-click-attribution/" to
"/.well-known/private-click-measurement/".
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::initializeNetworkProcess):
(WebKit::NetworkProcess::setPrivateClickMeasurementDebugMode):
(WebKit::NetworkProcess::deleteWebsiteData):
(WebKit::NetworkProcess::deleteWebsiteDataForOrigins):
(WebKit::NetworkProcess::storePrivateClickMeasurement):
(WebKit::NetworkProcess::dumpPrivateClickMeasurement):
(WebKit::NetworkProcess::clearPrivateClickMeasurement):
(WebKit::NetworkProcess::setPrivateClickMeasurementOverrideTimerForTesting):
(WebKit::NetworkProcess::setPrivateClickMeasurementConversionURLForTesting):
(WebKit::NetworkProcess::markPrivateClickMeasurementsAsExpiredForTesting):
(WebKit::NetworkProcess::setAdClickAttributionDebugMode): Deleted.
(WebKit::NetworkProcess::storeAdClickAttribution): Deleted.
(WebKit::NetworkProcess::dumpAdClickAttribution): Deleted.
(WebKit::NetworkProcess::clearAdClickAttribution): Deleted.
(WebKit::NetworkProcess::setAdClickAttributionOverrideTimerForTesting): Deleted.
(WebKit::NetworkProcess::setAdClickAttributionConversionURLForTesting): Deleted.
(WebKit::NetworkProcess::markAdClickAttributionsAsExpiredForTesting): Deleted.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/NetworkProcessCreationParameters.cpp:
(WebKit::NetworkProcessCreationParameters::encode const):
(WebKit::NetworkProcessCreationParameters::decode):
* NetworkProcess/NetworkProcessCreationParameters.h:
* NetworkProcess/NetworkResourceLoader.cpp:
(WebKit::NetworkResourceLoader::willSendRedirectedRequest):
(WebKit::NetworkResourceLoader::continueWillSendRedirectedRequest):
* NetworkProcess/NetworkResourceLoader.h:
* NetworkProcess/NetworkSession.cpp:
(WebKit::NetworkSession::NetworkSession):
(WebKit::NetworkSession::storePrivateClickMeasurement):
(WebKit::NetworkSession::handlePrivateClickMeasurementConversion):
(WebKit::NetworkSession::dumpPrivateClickMeasurement):
(WebKit::NetworkSession::clearPrivateClickMeasurement):
(WebKit::NetworkSession::clearPrivateClickMeasurementForRegistrableDomain):
(WebKit::NetworkSession::setPrivateClickMeasurementOverrideTimerForTesting):
(WebKit::NetworkSession::setPrivateClickMeasurementConversionURLForTesting):
(WebKit::NetworkSession::markPrivateClickMeasurementsAsExpiredForTesting):
(WebKit::NetworkSession::storeAdClickAttribution): Deleted.
(WebKit::NetworkSession::handleAdClickAttributionConversion): Deleted.
(WebKit::NetworkSession::dumpAdClickAttribution): Deleted.
(WebKit::NetworkSession::clearAdClickAttribution): Deleted.
(WebKit::NetworkSession::clearAdClickAttributionForRegistrableDomain): Deleted.
(WebKit::NetworkSession::setAdClickAttributionOverrideTimerForTesting): Deleted.
(WebKit::NetworkSession::setAdClickAttributionConversionURLForTesting): Deleted.
(WebKit::NetworkSession::markAdClickAttributionsAsExpiredForTesting): Deleted.
* NetworkProcess/NetworkSession.h:
* NetworkProcess/PrivateClickMeasurementManager.cpp: Renamed from Source/WebKit/NetworkProcess/AdClickAttributionManager.cpp.
(WebKit::PrivateClickMeasurementManager::storeUnconverted):
(WebKit::PrivateClickMeasurementManager::handleConversion):
(WebKit::PrivateClickMeasurementManager::startTimer):
(WebKit::PrivateClickMeasurementManager::convert):
(WebKit::PrivateClickMeasurementManager::fireConversionRequest):
(WebKit::PrivateClickMeasurementManager::firePendingConversionRequests):
(WebKit::PrivateClickMeasurementManager::clear):
(WebKit::PrivateClickMeasurementManager::clearForRegistrableDomain):
(WebKit::PrivateClickMeasurementManager::clearExpired):
(WebKit::PrivateClickMeasurementManager::toString const):
(WebKit::PrivateClickMeasurementManager::setConversionURLForTesting):
(WebKit::PrivateClickMeasurementManager::markAllUnconvertedAsExpiredForTesting):
(WebKit::PrivateClickMeasurementManager::debugModeEnabled const):
* NetworkProcess/PrivateClickMeasurementManager.h: Renamed from Source/WebKit/NetworkProcess/AdClickAttributionManager.h.
(WebKit::PrivateClickMeasurementManager::PrivateClickMeasurementManager):
(WebKit::PrivateClickMeasurementManager::m_sessionID):
(WebKit::PrivateClickMeasurementManager::setPingLoadFunction):
(WebKit::PrivateClickMeasurementManager::setOverrideTimerForTesting):
* Platform/Logging.h:
* Shared/NavigationActionData.cpp:
(WebKit::NavigationActionData::encode const):
(WebKit::NavigationActionData::decode):
* Shared/NavigationActionData.h:
* Shared/WebsiteData/WebsiteData.cpp:
(WebKit::WebsiteData::ownerProcess):
* Shared/WebsiteData/WebsiteDataType.h:
* Sources.txt:
* UIProcess/API/APINavigation.h:
(API::Navigation::privateClickMeasurement const):
(API::Navigation::adClickAttribution const): Deleted.
* UIProcess/API/C/WKPage.cpp:
(WKPageDumpPrivateClickMeasurement):
(WKPageClearPrivateClickMeasurement):
(WKPageSetPrivateClickMeasurementOverrideTimerForTesting):
(WKPageSetPrivateClickMeasurementConversionURLForTesting):
(WKPageMarkPrivateClickMeasurementsAsExpiredForTesting):
(WKPageDumpAdClickAttribution): Deleted.
(WKPageClearAdClickAttribution): Deleted.
(WKPageSetAdClickAttributionOverrideTimerForTesting): Deleted.
(WKPageSetAdClickAttributionConversionURLForTesting): Deleted.
(WKPageMarkAdClickAttributionsAsExpiredForTesting): Deleted.
* UIProcess/API/C/WKPagePrivate.h:
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreClearPrivateClickMeasurementsThroughWebsiteDataRemoval):
(WKWebsiteDataStoreClearAdClickAttributionsThroughWebsiteDataRemoval): Deleted.
* UIProcess/API/C/WKWebsiteDataStoreRef.h:
* UIProcess/API/Cocoa/WKWebsiteDataRecord.mm:
(dataTypesToString):
_WKWebsiteDataTypeAdClickAttributions still supported.
_WKWebsiteDataTypePrivateClickMeasurements added.
* UIProcess/API/Cocoa/WKWebsiteDataRecordInternal.h:
(WebKit::toWebsiteDataType):
_WKWebsiteDataTypeAdClickAttributions still supported and mapped to
WebsiteDataType::PrivateClickMeasurements.
(WebKit::toWKWebsiteDataTypes):
* UIProcess/API/Cocoa/WKWebsiteDataRecordPrivate.h:
_WKWebsiteDataTypeAdClickAttributions still supported.
_WKWebsiteDataTypePrivateClickMeasurements added.
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
_WKWebsiteDataTypeAdClickAttributions still supported.
_WKWebsiteDataTypePrivateClickMeasurements added.
(+[WKWebsiteDataStore _allWebsiteDataTypesIncludingPrivate]):
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeNetworkProcess):
* UIProcess/Inspector/WebInspectorProxy.cpp:
(WebKit::WebInspectorProxy::setDeveloperPreferenceOverride):
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::setPrivateClickMeasurementDebugMode):
(WebKit::NetworkProcessProxy::setAdClickAttributionDebugMode): Deleted.
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didCommitLoadForFrame):
(WebKit::WebPageProxy::createNewPage):
(WebKit::WebPageProxy::dumpPrivateClickMeasurement):
(WebKit::WebPageProxy::clearPrivateClickMeasurement):
(WebKit::WebPageProxy::setPrivateClickMeasurementOverrideTimerForTesting):
(WebKit::WebPageProxy::setPrivateClickMeasurementConversionURLForTesting):
(WebKit::WebPageProxy::markPrivateClickMeasurementsAsExpiredForTesting):
(WebKit::WebPageProxy::dumpAdClickAttribution): Deleted.
(WebKit::WebPageProxy::clearAdClickAttribution): Deleted.
(WebKit::WebPageProxy::setAdClickAttributionOverrideTimerForTesting): Deleted.
(WebKit::WebPageProxy::setAdClickAttributionConversionURLForTesting): Deleted.
(WebKit::WebPageProxy::markAdClickAttributionsAsExpiredForTesting): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::setPrivateClickMeasurementDebugMode):
(WebKit::WebsiteDataStore::setAdClickAttributionDebugMode): Deleted.
* UIProcess/WebsiteData/WebsiteDataStore.h:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::createWindow):
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::dispatchDecidePolicyForNewWindowAction):
(WebKit::WebFrameLoaderClient::dispatchDecidePolicyForNavigationAction):
2020-11-11 Peng Liu <peng.liu6@apple.com>
Update RemoteMediaPlayerMIMETypeCache to add the support for AVFoundationMSE media engine
https://bugs.webkit.org/show_bug.cgi?id=218788
Reviewed by Eric Carlson.
Originally, `RemoteMediaPlayerMIMETypeCache` inherits from `MIMETypeCache`.
It needs to implement three functions: `isAvailable()`, `canDecodeExtendedType()`,
`supportedTypes()` through IPC messages, and the IPC message receivers in
`RemoteMediaPlayerManagerProxy` needs to repeat the work done by `AVAssetMIMETypeCache`
and `AVStreamDataParserMIMETypeCache`.
Actually, `MediaPlayerRemoteFactory` only needs two functions from
`RemoteMediaPlayerMIMETypeCache`: `supportedTypes()` and `supportsTypeAndCodecs()`.
We can implement these two functions with IPC messages to `RemoteMediaPlayerManagerProxy`
and cache the results to avoid too many IPC messages. In addition, the implementation of
the IPC message receivers in `RemoteMediaPlayerManagerProxy` can reuse the implementation
of `AVAssetMIMETypeCache` and `AVStreamDataParserMIMETypeCache`.
* GPUProcess/media/RemoteMediaPlayerManagerProxy.cpp:
(WebKit::RemoteMediaPlayerManagerProxy::supportsTypeAndCodecs):
(WebKit::RemoteMediaPlayerManagerProxy::canDecodeExtendedType): Deleted.
* GPUProcess/media/RemoteMediaPlayerManagerProxy.h:
* GPUProcess/media/RemoteMediaPlayerManagerProxy.messages.in:
* WebProcess/GPU/media/RemoteAudioDestinationProxy.h:
* WebProcess/GPU/media/RemoteCDMInstanceSession.cpp:
* WebProcess/GPU/media/RemoteMediaPlayerMIMETypeCache.cpp:
(WebKit::RemoteMediaPlayerMIMETypeCache::addSupportedTypes):
(WebKit::RemoteMediaPlayerMIMETypeCache::supportedTypes):
(WebKit::RemoteMediaPlayerMIMETypeCache::supportsTypeAndCodecs):
(WebKit::RemoteMediaPlayerMIMETypeCache::mimeCache const): Deleted.
(WebKit::RemoteMediaPlayerMIMETypeCache::staticContainerTypeList): Deleted.
(WebKit::RemoteMediaPlayerMIMETypeCache::isUnsupportedContainerType): Deleted.
(WebKit::RemoteMediaPlayerMIMETypeCache::canDecodeExtendedType): Deleted.
(WebKit::RemoteMediaPlayerMIMETypeCache::initializeCache): Deleted.
* WebProcess/GPU/media/RemoteMediaPlayerMIMETypeCache.h:
* WebProcess/GPU/media/RemoteMediaPlayerManager.cpp:
(WebKit::RemoteMediaPlayerManager::getSupportedTypes):
(RemoteMediaPlayerManager::initialize):
2020-11-11 Tim Horton <timothy_horton@apple.com>
Add minimal support for deep color IOSurface backing store to ImageBuffer
https://bugs.webkit.org/show_bug.cgi?id=218816
Reviewed by Simon Fraser.
* GPUProcess/graphics/RemoteImageBuffer.h:
(WebKit::RemoteImageBuffer::create):
(WebKit::RemoteImageBuffer::RemoteImageBuffer):
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::imageBufferBackendWasCreated):
(WebKit::RemoteRenderingBackend::createImageBuffer):
* GPUProcess/graphics/RemoteRenderingBackend.h:
* GPUProcess/graphics/RemoteRenderingBackend.messages.in:
* Scripts/webkit/messages.py:
* Shared/RemoteLayerTree/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::bytesPerPixel const):
(WebKit::RemoteLayerBackingStore::surfaceBufferFormat const):
* Shared/ShareableBitmap.cpp:
(WebKit::ShareableBitmap::create):
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _takeViewSnapshot]):
* WebProcess/GPU/graphics/ImageBufferShareableBitmapBackend.cpp:
(WebKit::ImageBufferShareableBitmapBackend::create):
(WebKit::ImageBufferShareableBitmapBackend::ImageBufferShareableBitmapBackend):
* WebProcess/GPU/graphics/ImageBufferShareableBitmapBackend.h:
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
(WebKit::RemoteImageBufferProxy::createBackend):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::createImageBuffer):
(WebKit::RemoteRenderingBackendProxy::imageBufferBackendWasCreated):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.messages.in:
* WebProcess/GPU/graphics/cocoa/ImageBufferShareableIOSurfaceBackend.cpp:
(WebKit::ImageBufferShareableIOSurfaceBackend::create):
* WebProcess/GPU/graphics/cocoa/ImageBufferShareableIOSurfaceBackend.h:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::createImageBuffer const):
* WebProcess/WebCoreSupport/WebChromeClient.h:
2020-11-11 Wenson Hsieh <wenson_hsieh@apple.com>
Add system trace points for the GPU process display list processing loop
https://bugs.webkit.org/show_bug.cgi?id=218824
Reviewed by Simon Fraser.
Capture the time spent applying display list items in the GPU process with a pair of new trace points, to aid
performance debugging and analysis.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::wakeUpAndApplyDisplayList):
2020-11-11 Said Abou-Hallawa <said@apple.com>
[GPU Process] Cache the NativeImage in GPU Process and allow referencing it with its RemoteResourceIdentifier
https://bugs.webkit.org/show_bug.cgi?id=217596
Reviewed by Simon Fraser.
Recorder::drawNativeImage() will ensure the NativeImage is cached before
recording the DrawNativeImage item with the renderingResourceIdentifier.
Because RemoteImageBufferProxy is a superclass of DisplayList::Recorder::
Delegate, it will ensure the NativeImage is cached in its RemoteResourceCacheProxy.
If the NativeImage is registered in RemoteResourceCacheProxy for the first
time, it is going to be sent as an IPC message to GPU. Upon receiving this
message, RemoteRenderingBackend will cache this NativeImage in its
RemoteResourceCache. This message has to be received for any NativeImage
only once for its entire life time.
When flushing the DisplayList of any ImageBuffer, RemoteImageBuffer will
ensure the NativeImages are cached in the RemoteResourceCache are passed
to DisplayList::Replayer so it can use them to get NativeImage for every
DrawNativeImage with the same renderingResourceIdentifier it was recorded
with.
When NativeImage is bing destroyed, its destructor will call its observer
virtual method releaseNativeImage(). RemoteResourceCacheProxy will be a
superclass NativeImage::Observer so its overriding method will be called.
When RemoteResourceCacheProxy::releaseNativeImage() is called it is going
to delete the NativeImage from RemoteResourceCacheProxy and is going to
send a message to GPU to delete the corresponding NativeImage from
RemoteResourceCache via RemoteRenderingBackend::releaseRemoteResource().
* GPUProcess/graphics/RemoteImageBuffer.h:
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::cacheNativeImage):
(WebKit::RemoteRenderingBackend::decodeItem):
* GPUProcess/graphics/RemoteRenderingBackend.h:
* GPUProcess/graphics/RemoteRenderingBackend.messages.in:
* GPUProcess/graphics/RemoteResourceCache.cpp:
(WebKit::RemoteResourceCache::cacheNativeImage):
(WebKit::RemoteResourceCache::releaseRemoteResource):
* GPUProcess/graphics/RemoteResourceCache.h:
(WebKit::RemoteResourceCache::nativeImages const):
* Scripts/webkit/messages.py:
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<Ref<NativeImage>>::encode):
(IPC::ArgumentCoder<Ref<NativeImage>>::decode):
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
(WebKit::RemoteImageBufferProxy::~RemoteImageBufferProxy):
(WebKit::RemoteImageBufferProxy::RemoteImageBufferProxy):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::createImageBuffer):
(WebKit::RemoteRenderingBackendProxy::cacheNativeImage):
(WebKit::RemoteRenderingBackendProxy::releaseRemoteResource):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
* WebProcess/GPU/graphics/RemoteResourceCacheProxy.cpp:
(WebKit::RemoteResourceCacheProxy::RemoteResourceCacheProxy):
(WebKit::RemoteResourceCacheProxy::cacheImageBuffer):
(WebKit::RemoteResourceCacheProxy::cachedImageBuffer):
(WebKit::RemoteResourceCacheProxy::cacheNativeImage):
(WebKit::RemoteResourceCacheProxy::releaseNativeImage):
* WebProcess/GPU/graphics/RemoteResourceCacheProxy.h:
2020-11-11 Chris Dumez <cdumez@apple.com>
Terminate WebProcesses if GPUProcess crashes more than twice in 30 seconds
https://bugs.webkit.org/show_bug.cgi?id=218811
<rdar://problem/71292424>
Reviewed by Tim Horton.
If the GPUProcess crashes more than twice in 30 seconds, we now terminate all WebProcesses
to try and get back into a good state. This is to avoid getting into a situation where
the GPUProcess is crash looping.
* UIProcess/AuxiliaryProcessProxy.cpp:
(WebKit::AuxiliaryProcessProxy::terminate):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::gpuProcessCrashed):
(WebKit::WebProcessPool::resetGPUProcessCrashCount):
(WebKit::WebProcessPool::terminateAllWebContentProcesses):
* UIProcess/WebProcessPool.h:
2020-11-11 Brian Burg <bburg@apple.com>
[Cocoa] Inspector Extensions: Add _WKInspectorExtension and related plumbing
https://bugs.webkit.org/show_bug.cgi?id=217783
<rdar://problem/69968787>
Reviewed by Devin Rousso.
_WKInspectorExtension is new SPI that represents a Web Extension that uses the 'devtools'
API to interact with Web Inspector. An extension is associated with its _WKInspectorExtensionHost.
A WebKit client that supports Web Extensions can use this class to implement 'devtools' extension APIs.
Add _WKInspectorExtensionHost methods to register and unregister an extension.
Add plumbing so that the extension in registered in the frontend with WI.InspectorExtensionController.
Later patches will flesh out the needed functionality to implement rest of the 'devtools' API.
* Sources.txt:
* SourcesCocoa.txt:
* DerivedSources-input.xcfilelist:
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* WebKit.xcodeproj/project.pbxproj:
Add new files.
* Scripts/webkit/messages.py: If a message includes InspectorExtensionError,
then generate an include for InspectorExtensionTypes.h where it is declared.
* Shared/API/APIObject.h:
* Shared/Cocoa/APIObject.mm:
(API::Object::newObject):
New API object type.
* Shared/InspectorExtensionTypes.h: Added error code enum and useful type aliases.
* Shared/InspectorExtensionTypes.cpp:
(WebKit::inspectorExtensionErrorToString):
* UIProcess/API/APIInspectorExtension.h:
* UIProcess/API/APIInspectorExtension.cpp:
(API::InspectorExtension::create):
(API::InspectorExtension::InspectorExtension):
(API::InspectorExtension::~InspectorExtension):
Added. This is a placeholder object that's used for the wrapper system,
since we don't have single C++ objects for each extension to back the API object.
* UIProcess/API/Cocoa/_WKInspector.h:
* UIProcess/API/Cocoa/_WKInspector.mm:
(-[_WKInspector registerExtensionWithID:displayName:completionHandler:]):
(-[_WKInspector unregisterExtension:completionHandler:]):
Implement new _WKInspectorExtensionHost methods. Send IPC to the Inspector web process.
* UIProcess/API/Cocoa/_WKInspectorExtension.h:
* UIProcess/API/Cocoa/_WKInspectorExtension.mm:
(-[_WKInspectorExtension initWithIdentifier:]):
(-[_WKInspectorExtension dealloc]):
(-[_WKInspectorExtension _apiObject]):
(-[_WKInspectorExtension extensionID]):
Added. This is a data object that's also used as a token to unregister the extension.
* UIProcess/API/Cocoa/_WKInspectorExtensionHost.h: Add new SPI.
* UIProcess/API/Cocoa/_WKInspectorExtensionInternal.h:
* UIProcess/API/Cocoa/_WKRemoteWebInspectorViewController.h:
* UIProcess/API/Cocoa/_WKRemoteWebInspectorViewController.mm:
(-[_WKRemoteWebInspectorViewController registerExtensionWithID:displayName:completionHandler:]):
(-[_WKRemoteWebInspectorViewController unregisterExtension:completionHandler:]):
Implement new _WKInspectorExtensionHost methods. Send IPC to the Inspector web process.
* UIProcess/Inspector/RemoteWebInspectorProxy.h:
* UIProcess/Inspector/RemoteWebInspectorProxy.messages.in: Add FrontendLoaded.
(WebKit::RemoteWebInspectorProxy::extensionController const): Added.
* UIProcess/Inspector/RemoteWebInspectorProxy.cpp:
(WebKit::RemoteWebInspectorProxy::frontendLoaded):
(WebKit::RemoteWebInspectorProxy::createFrontendPageAndWindow):
Set up the extension controller object after the frontend has been created.
Hook up the FrontendLoaded event sent from Inspector web process so that the
extension controller can be notified of when it's safe to message the frontend.
* UIProcess/Inspector/WebInspectorProxy.cpp:
(WebKit::WebInspectorProxy::createFrontendPage):
(WebKit::WebInspectorProxy::closeFrontendPageAndWindow):
(WebKit::WebInspectorProxy::frontendLoaded):
Set up the extension controller object after the frontend has been created.
* UIProcess/Inspector/WebInspectorProxy.h:
(WebKit::WebInspectorProxy::extensionController const): Added.
* UIProcess/Inspector/WebInspectorUIExtensionControllerProxy.h:
* UIProcess/Inspector/WebInspectorUIExtensionControllerProxy.cpp: Added.
(WebKit::WebInspectorUIExtensionControllerProxy::WebInspectorUIExtensionControllerProxy):
(WebKit::WebInspectorUIExtensionControllerProxy::~WebInspectorUIExtensionControllerProxy):
(WebKit::WebInspectorUIExtensionControllerProxy::whenFrontendHasLoaded):
(WebKit::WebInspectorUIExtensionControllerProxy::inspectorFrontendLoaded):
(WebKit::WebInspectorUIExtensionControllerProxy::registerExtension):
(WebKit::WebInspectorUIExtensionControllerProxy::unregisterExtension):
Added. Forward API requests to the Web Inspector's web process to be evaluated in the frontend.
* WebProcess/Inspector/WebInspectorFrontendAPIDispatcher.h:
* WebProcess/Inspector/WebInspectorFrontendAPIDispatcher.cpp:
(WebKit::WebInspectorFrontendAPIDispatcher::dispatchCommand):
Add a more generic way to encode values to be passed as arguments to InspectorFrontendAPI.dispatch.
* WebProcess/Inspector/WebInspectorUI.cpp:
(WebKit::WebInspectorUI::establishConnection):
(WebKit::WebInspectorUI::closeWindow):
* WebProcess/Inspector/WebInspectorUI.h:
(WebKit::WebInspectorUI::frontendPage const): Added.
(WebKit::WebInspectorUI::frontendAPIDispatcher): Added.
* WebProcess/Inspector/WebInspectorUIExtensionController.h:
* WebProcess/Inspector/WebInspectorUIExtensionController.messages.in: Added.
* WebProcess/Inspector/WebInspectorUIExtensionController.cpp: Added.
(WebKit::WebInspectorUIExtensionController::WebInspectorUIExtensionController):
(WebKit::WebInspectorUIExtensionController::~WebInspectorUIExtensionController):
(WebKit::WebInspectorUIExtensionController::registerExtension):
(WebKit::WebInspectorUIExtensionController::unregisterExtension):
(WebKit::WebInspectorUIExtensionController::parseInspectorExtensionErrorFromResult):
Handle incoming messages by forwarding the associated command to WI.WebExtensionController
via InspectorFrontendAPI evaluations.
* UIProcess/API/Cocoa/_WKUserStyleSheet.h:
* UIProcess/API/Cocoa/_WKUserStyleSheet.mm:
Fallout from unified build chunking shifts.
* UIProcess/WebAuthentication/Cocoa/LocalAuthenticator.mm:
(WebKit::LocalAuthenticatorInternal::getExistingCredentials):
(WebKit::LocalAuthenticator::continueMakeCredentialAfterUserVerification):
Fallout from unified build chunking shifts, but more.
2020-11-11 Alex Christensen <achristensen@webkit.org>
Define USE(LEGACY_CFNETWORK_DOWNLOADS)
https://bugs.webkit.org/show_bug.cgi?id=218814
Reviewed by Geoffrey Garen.
Mojave CFNetwork has a few download bugs I'll need to work around.
With this change I can use USE(LEGACY_CFNETWORK_DOWNLOADS) instead of __MAC_OS_X_VERSION_MIN_REQUIRED,
which is more in line with WebKit style. This will be used more in bug 217747.
* NetworkProcess/Downloads/cocoa/DownloadCocoa.mm:
(WebKit::Download::resume):
2020-11-11 Chris Dumez <cdumez@apple.com>
[GPUProcess] Seamlessly resume WebAudio playback if the GPU Process crashes
https://bugs.webkit.org/show_bug.cgi?id=218806
Reviewed by Geoffrey Garen.
In case of a GPU process crash, we were previously terminating all WebProcesses.
This patch makes it so that WebProcesses are no longer terminated in such cases.
Additionally, this patch makes sure that if WebAudio is rendering audio while
the GPU Process crashes, the GPU process will get relaunched right away and
audio rendering will seamlessly continue.
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::gpuProcessCrashed):
* WebProcess/GPU/GPUProcessConnection.cpp:
(WebKit::GPUProcessConnection::didClose):
* WebProcess/GPU/GPUProcessConnection.h:
(WebKit::GPUProcessConnection::Client::gpuProcessConnectionDidClose):
(WebKit::GPUProcessConnection::addClient):
(WebKit::GPUProcessConnection::removeClient):
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::RemoteAudioDestinationProxy):
(WebKit::RemoteAudioDestinationProxy::connectToGPUProcess):
(WebKit::RemoteAudioDestinationProxy::gpuProcessConnectionDidClose):
* WebProcess/GPU/media/RemoteAudioDestinationProxy.h:
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::gpuProcessConnectionClosed):
* WebProcess/WebProcess.h:
2020-11-11 Alex Christensen <achristensen@webkit.org>
Use _CFURLConnectionSetFrameworkStubs instead of DYLD_INTERPOSE on Intel Macs
https://bugs.webkit.org/show_bug.cgi?id=218810
Reviewed by Geoffrey Garen.
A long, long time ago, we realized that CFNetwork was looking in the keychain for persistent credentials,
and in order to make the network process get credentials as the UI process, we introduced SecItemShim which
used DYLD_INTERPOSE to intercept the calls to the security framework and synchronously message the UI process
for those calls. Over the last 6 years we have realized that doesn't work on iOS or Apple silicon Macs, so
we use _CFURLConnectionSetFrameworkStubs instead. That works on Intel Macs, too, and I'm reasonably certain
that using it everywhere won't break anything, so as a step towards rdar://problem/15588174 which should allow
us to use asynchronous messages and no more globalNetworkProcess pointer, let's stop using DYLD_INTERPOSE.
* Shared/mac/SecItemShim.cpp:
(WebKit::initializeSecItemShim):
2020-11-11 Alex Christensen <achristensen@webkit.org>
Don't look in keychain for SecKeyRef after deserializing SecIdentityRef
https://bugs.webkit.org/show_bug.cgi?id=218809
<rdar://problem/69394018>
Reviewed by Geoffrey Garen.
In ye olden days of WebKit2, we used to use special code to serialize an NSURLCredential with an identity.
We needed to serialize the certificate chain and a way to look in the keychain to find the private key.
That didn't work with third party apps not entitled to look in the keychain, and it also didn't work with hardware
keys that retain the private key themselves, so we introduced SecKeyProxy. Now, all those things work and we don't
serialize SecIdentityRefs any more, so let's remove the problematic dead code that would have tried.
Luckily, we now have API tests that verify that client certificate authentication still works.
r169938 added a restriction to looking in the keychain only if the parent app has entitlements needed to do so.
That restriction isn't needed any more because nobody looks in the keychain from the network process.
r269162 entirely removed looking in the keychain for private keys for client certificate authentication.
This is moving in that same direction.
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceEntryPoint.mm:
(WebKit::XPCServiceInitializerDelegate::checkEntitlements):
* Shared/cf/ArgumentCodersCF.cpp:
(IPC::typeFromCFTypeRef):
(IPC::encode):
(IPC::decode):
(IPC::setAllowsDecodingSecKeyRef): Deleted.
(IPC::copyPersistentRef): Deleted.
* Shared/cf/ArgumentCodersCF.h:
* Shared/mac/WebCoreArgumentCodersMac.mm:
(IPC::ArgumentCoder<WebCore::Credential>::encodePlatformData):
(IPC::ArgumentCoder<WebCore::Credential>::decodePlatformData):
2020-11-11 Sihui Liu <sihui_liu@apple.com>
REGRESSION (r269348): ASSERTION FAILED: !m_messageReceiverMapCount under WebKit::SpeechRecognitionServer::~SpeechRecognitionServer
https://bugs.webkit.org/show_bug.cgi?id=218770
<rdar://problem/71252234>
Reviewed by Youenn Fablet.
Remove SpeechRecognitionServer from MessageReceiverMap before it is destroyed.
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::~WebProcessProxy):
2020-11-11 Chris Dumez <cdumez@apple.com>
REGRESSION (r269657): ASSERTION FAILED: Completion handler should always be called under WebKit::GPUProcessProxy::~GPUProcessProxy
https://bugs.webkit.org/show_bug.cgi?id=218807
<rdar://problem/71285627>
Reviewed by Geoffrey Garen.
GPUProcessProxy::sendPrepareToSuspend() was failing to send the IPC to the GPUProcess and was also
failing to call its completion handler, thus asserting in Debug. Update GPUProcessProxy to actually
send the PrepareToSuspend / ProcessDidResume IPCs to the GPUProcess. The GPUProcess actually already
had an implementation to deal with those IPCs.
No new tests, covered by existing tests that are crashing in debug.
* GPUProcess/GPUProcess.messages.in:
* UIProcess/GPU/GPUProcessProxy.cpp:
(WebKit::GPUProcessProxy::sendPrepareToSuspend):
(WebKit::GPUProcessProxy::sendProcessDidResume):
* UIProcess/GPU/GPUProcessProxy.h:
2020-11-11 Wenson Hsieh <wenson_hsieh@apple.com>
[Concurrent display lists] Add an initial implementation of concurrent display list rendering
https://bugs.webkit.org/show_bug.cgi?id=218426
<rdar://problem/71167220>
Reviewed by Ryosuke Niwa.
This patch adds a first-cut implementation of concurrent display list rendering, which allows the GPU and web
processes to read and write display list data at the same time. To achieve this, we remove the placeholder
shared memory display list encoding mechanism previously added in <webkit.org/b/218406>, and replace it with a
new model in which a writer (i.e. the web process) and a reader (i.e. the GPU process) share a set of reusable
shared memory buffers which contain display list item data. As the writer appends data to shared memory, it
increments a counter in each shared memory buffer that represents the number of bytes that the reader has yet to
read; as the reader reads display list items from shared memory, it decrements this counter.
This patch also implements a simple strategy for reusing these shared item buffers. The web process maintains a
queue of reusable buffers; when display list items are appended and the current writable buffer is out of
capacity, the display list item buffer calls out to WebKit2 via a client method to request more shared memory.
Here, the web process traverses the queue of reusable item buffers in search of a buffer with sufficient
capacity; if found, we hand this back to WebCore as an opaque `ItemBufferHandle`; otherwise, we allocate a new
reusable shared memory buffer. Item buffers that have been used become reusable only after they are (1) no
longer being read from -- i.e., the counter value is 0, and (2) they are not actively being written to.
Buffers that have been used are lazily reintroduced into the reuse queue when a display list requests a handle
to writable item buffer data.
See below for more details.
* GPUProcess/graphics/DisplayListReaderHandle.cpp: Copied from Source/WebKit/Shared/SharedDisplayListHandle.cpp.
(WebKit::DisplayListReaderHandle::advance):
(WebKit::DisplayListReaderHandle::displayListForReading const):
* GPUProcess/graphics/DisplayListReaderHandle.h: Copied from Source/WebKit/Shared/SharedDisplayListHandle.cpp.
(WebKit::DisplayListReaderHandle::create):
(WebKit::DisplayListReaderHandle::DisplayListReaderHandle):
Add a new helper class to represent a shared display list item buffer that the GPU process can use to consume
display list items. Advancing the cursor in `DisplayListReaderHandle` decrements the "remaining bytes" counter;
the class also has a helper method for creating a new `DisplayList` with a given offset into the shared buffer,
as well the number of bytes to read.
* GPUProcess/graphics/RemoteImageBuffer.h:
Handle the `MetaCommandSwitchTo` and `FlushContext` items by (respectively) informing the rendering backend
about the next buffer to read from, and sending a "flush committed" IPC message back to the web process.
(WebKit::RemoteImageBuffer::RemoteImageBuffer):
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::applyDisplayListsFromHandle):
(WebKit::RemoteRenderingBackend::wakeUpAndApplyDisplayList):
This is the main loop in the GPU process that is responsible for applying incoming display list item data. This
IPC endpoint receives information about where to start (i.e. an item buffer ID and an offset into shared memory
in that buffer) and begins reading ranges of display list item data inside that buffer. After it is done
processing all of its unread bytes, it then checks the unread byte count once again, and continues the loop if
the web process has advanced this count.
(WebKit::RemoteRenderingBackend::setNextItemBufferToRead):
(WebKit::RemoteRenderingBackend::didCreateSharedDisplayListHandle):
Add logic to handle the case where we reached the end of an item buffer, but the next buffer identifier is still
unknown to us (since it is still "in transit" via the `DidCreateSharedDisplayListHandle` IPC message). In this
case, we remember the next buffer identifier in `m_nextItemBufferToRead`, and exit the processing loop inside
`wakeUpAndApplyDisplayList`. Upon receiving the shared memory handle in `didCreateSharedDisplayListHandle`, we
then automatically resume display list processing, starting from the beginning of this new buffer.
(WebKit::RemoteRenderingBackend::applyDisplayList): Deleted.
(WebKit::RemoteRenderingBackend::submitDisplayList): Deleted.
(WebKit::RemoteRenderingBackend::flushDisplayListAndCommit): Deleted.
(WebKit::RemoteRenderingBackend::didCreateSharedItemData): Deleted.
We don't need any of these methods anymore, now that there is only one codepath for processing display list
items. See `RemoteRenderingBackend::wakeUpAndApplyDisplayList` above. Also, rename `didCreateSharedItemData`
to `didCreateSharedDisplayListHandle` (also above).
* GPUProcess/graphics/RemoteRenderingBackend.h:
* GPUProcess/graphics/RemoteRenderingBackend.messages.in:
* Shared/SharedDisplayListHandle.h:
(WebKit::SharedDisplayListHandle::sharedMemory):
(WebKit::SharedDisplayListHandle::sharedMemory const):
(WebKit::SharedDisplayListHandle::identifier const):
(WebKit::SharedDisplayListHandle::data const):
(WebKit::SharedDisplayListHandle::unreadBytes):
(WebKit::SharedDisplayListHandle::Lock::Lock):
(WebKit::SharedDisplayListHandle::Lock::~Lock):
(WebKit::SharedDisplayListHandle::SharedDisplayListHandle):
(WebKit::SharedDisplayListHandle::header):
(WebKit::SharedDisplayListHandle::header const):
(WebKit::SharedDisplayListHandle::encode const): Deleted.
(WebKit::SharedDisplayListHandle::decode): Deleted.
Repurpose `SharedDisplayListHandle` to serve as a base class for both `DisplayListReaderHandle` and
`DisplayListWriterHandle`. This base class is aware of the memory layout of the display list item buffer in
shared memory (in particular, the fact that the first 8 bytes are the atomic lock and the next 8 bytes contain
the unread byte counter).
Internally, this uses a protected RAII class, `SharedDisplayListHandle::Lock`, to grab the lock before either
reading or writing the unread bytes count.
* Sources.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/graphics/DisplayListWriterHandle.cpp: Added.
(WebKit::DisplayListWriterHandle::advance):
(WebKit::DisplayListWriterHandle::availableCapacity const):
(WebKit::DisplayListWriterHandle::createHandle const):
(WebKit::DisplayListWriterHandle::resetWritableOffsetIfPossible):
This helper method moves the writable offset (see below) back to the start of the buffer if it is not being read
from by the GPU process.
* WebProcess/GPU/graphics/DisplayListWriterHandle.h: Renamed from Source/WebKit/Shared/SharedDisplayListHandle.cpp.
(WebKit::DisplayListWriterHandle::create):
(WebKit::DisplayListWriterHandle::writableOffset const):
Similar to `DisplayListReaderHandle`, except that this contains logic specific to the web process for writing
display list items into shared memory. Advancing the cursor in `DisplayListWriterHandle` increments the counter.
It also increments the `writableOffset` of the handle, which represents the minimum offset into the shared
buffer that may contain newly appended display list items.
(WebKit::DisplayListWriterHandle::DisplayListWriterHandle):
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::submitDisplayList):
(WebKit::RemoteRenderingBackendProxy::updateReusableHandles):
This private helper iterates the set of buffers that are not currently being written to, and adds each one to
the reusable buffer queue only if its writable offset is at the start.
(WebKit::RemoteRenderingBackendProxy::createItemBuffer):
Implements logic that hands shared item buffer data handles to WebCore for writing display list items. This will
try to reuse an item buffer in the reuse queue, if possible, and fall back to allocating a new chunk of shared
memory if no reusable buffer is found (or the existing buffers have insufficient capacity). The fact that we
iterate through each reusable buffer in order, combined with the above logic in `updateReusableHandles`, ensures
that once we begin writing to a item buffer, we will continue using that item buffer until it exhausts available
capacity.
This property is important because it allows the GPU process to simply start reading item data from a given
offset and buffer ID (specified in the `WakeUpAndApplyDisplayList` IPC message), and seamlessly continue
reading data even after it exhausts all available item data in the current buffer by continuing from the start
of the item buffer that contains the next display list items (which we learn in the GPU process by encountering
a `MetaCommandSwitchTo` item at the end of the previous buffer).
In "steady state" (wherein both processes are concurrently writing and reading items as quickly as possible),
this means that as the web process is busy filling up each item buffer with item data, the GPU process is also
busy applying all the item data in one buffer and automatically moving on to the next buffer with no additional
IPC messages required.
(WebKit::RemoteRenderingBackendProxy::flushDisplayListAndCommit): Deleted.
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
2020-11-11 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r269244.
https://bugs.webkit.org/show_bug.cgi?id=218483
Patch is breaking tests
Reverted changeset:
"Stop sending origins from WebProcess to UIProcess for
getUserMedia requests"
https://bugs.webkit.org/show_bug.cgi?id=218192
https://trac.webkit.org/changeset/269244
2020-11-11 Corentin Noël <corentin.noel@collabora.com>
[GTK] Annotate WebKitUserMessage getters as nullable
https://bugs.webkit.org/show_bug.cgi?id=218761
Reviewed by Carlos Garcia Campos.
These properties can be set to NULL at creation time.
* Shared/API/glib/WebKitUserMessage.cpp:
2020-11-10 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r269660.
https://bugs.webkit.org/show_bug.cgi?id=218786
Crashing in EWS iOS simulator bots
Reverted changeset:
"PCM: Change from ad-click-attribution to private-click-
measurement (in all forms, including .well-known URL)"
https://bugs.webkit.org/show_bug.cgi?id=218730
https://trac.webkit.org/changeset/269660
2020-11-10 Per Arne Vollan <pvollan@apple.com>
[macOS] Enforce message filtering in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=218756
<rdar://problem/71240766>
Reviewed by David Kilzer.
Based on telemetry, we should deny access to messages not being used in the WebContent process on macOS.
No new tests, covered by existing tests.
* WebProcess/com.apple.WebProcess.sb.in:
2020-11-10 Chris Dumez <cdumez@apple.com>
Webkit incorrectly setting visibilityState to "prerender" when opening link in new tab
https://bugs.webkit.org/show_bug.cgi?id=215851
<rdar://problem/67817100>
Reviewed by Geoff Garen.
* Shared/API/c/WKSharedAPICast.h:
(WebKit::toVisibilityState):
* WebProcess/InjectedBundle/API/gtk/DOM/WebKitDOMDocumentGtk.cpp:
(webkit_dom_document_get_visibility_state):
2020-11-10 Tim Horton <timothy_horton@apple.com>
Use GraphicsContext instead of CGContext in RemoteLayerBackingStore
https://bugs.webkit.org/show_bug.cgi?id=218738
Reviewed by Wenson Hsieh.
* Shared/RemoteLayerTree/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::drawInContext):
Make use of the GraphicsContext abstraction instead of dropping down
to CGContext, in preparation for DisplayList rendering.
2020-11-10 Darin Adler <darin@apple.com>
Remove another function that implicitly uses the composed tree (intersects with range)
https://bugs.webkit.org/show_bug.cgi?id=218726
Reviewed by Ryosuke Niwa.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::requestDocumentEditingContext): Refactor to use VisiblePositionRange and
to call intersects on two of them, to improve the code readability and also sidestep the
issue of calling the intersects template function from outside WebCore.
2020-11-10 John Wilander <wilander@apple.com>
PCM: Change from ad-click-attribution to private-click-measurement (in all forms, including .well-known URL)
https://bugs.webkit.org/show_bug.cgi?id=218730
<rdar://problem/71094296>
Reviewed by Devin Rousso.
Change to the official name of the proposed standard Private Click Measurement
https://github.com/privacycg/private-click-measurement.
This includes a change of the reporting URL from
"/.well-known/ad-click-attribution/" to
"/.well-known/private-click-measurement/".
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::initializeNetworkProcess):
(WebKit::NetworkProcess::setPrivateClickMeasurementDebugMode):
(WebKit::NetworkProcess::deleteWebsiteData):
(WebKit::NetworkProcess::deleteWebsiteDataForOrigins):
(WebKit::NetworkProcess::storePrivateClickMeasurement):
(WebKit::NetworkProcess::dumpPrivateClickMeasurement):
(WebKit::NetworkProcess::clearPrivateClickMeasurement):
(WebKit::NetworkProcess::setPrivateClickMeasurementOverrideTimerForTesting):
(WebKit::NetworkProcess::setPrivateClickMeasurementConversionURLForTesting):
(WebKit::NetworkProcess::markPrivateClickMeasurementsAsExpiredForTesting):
(WebKit::NetworkProcess::setAdClickAttributionDebugMode): Deleted.
(WebKit::NetworkProcess::storeAdClickAttribution): Deleted.
(WebKit::NetworkProcess::dumpAdClickAttribution): Deleted.
(WebKit::NetworkProcess::clearAdClickAttribution): Deleted.
(WebKit::NetworkProcess::setAdClickAttributionOverrideTimerForTesting): Deleted.
(WebKit::NetworkProcess::setAdClickAttributionConversionURLForTesting): Deleted.
(WebKit::NetworkProcess::markAdClickAttributionsAsExpiredForTesting): Deleted.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/NetworkProcessCreationParameters.cpp:
(WebKit::NetworkProcessCreationParameters::encode const):
(WebKit::NetworkProcessCreationParameters::decode):
* NetworkProcess/NetworkProcessCreationParameters.h:
* NetworkProcess/NetworkResourceLoader.cpp:
(WebKit::NetworkResourceLoader::willSendRedirectedRequest):
(WebKit::NetworkResourceLoader::continueWillSendRedirectedRequest):
* NetworkProcess/NetworkResourceLoader.h:
* NetworkProcess/NetworkSession.cpp:
(WebKit::NetworkSession::NetworkSession):
(WebKit::NetworkSession::storePrivateClickMeasurement):
(WebKit::NetworkSession::handlePrivateClickMeasurementConversion):
(WebKit::NetworkSession::dumpPrivateClickMeasurement):
(WebKit::NetworkSession::clearPrivateClickMeasurement):
(WebKit::NetworkSession::clearPrivateClickMeasurementForRegistrableDomain):
(WebKit::NetworkSession::setPrivateClickMeasurementOverrideTimerForTesting):
(WebKit::NetworkSession::setPrivateClickMeasurementConversionURLForTesting):
(WebKit::NetworkSession::markPrivateClickMeasurementsAsExpiredForTesting):
(WebKit::NetworkSession::storeAdClickAttribution): Deleted.
(WebKit::NetworkSession::handleAdClickAttributionConversion): Deleted.
(WebKit::NetworkSession::dumpAdClickAttribution): Deleted.
(WebKit::NetworkSession::clearAdClickAttribution): Deleted.
(WebKit::NetworkSession::clearAdClickAttributionForRegistrableDomain): Deleted.
(WebKit::NetworkSession::setAdClickAttributionOverrideTimerForTesting): Deleted.
(WebKit::NetworkSession::setAdClickAttributionConversionURLForTesting): Deleted.
(WebKit::NetworkSession::markAdClickAttributionsAsExpiredForTesting): Deleted.
* NetworkProcess/NetworkSession.h:
* NetworkProcess/PrivateClickMeasurementManager.cpp: Renamed from Source/WebKit/NetworkProcess/AdClickAttributionManager.cpp.
(WebKit::PrivateClickMeasurementManager::storeUnconverted):
(WebKit::PrivateClickMeasurementManager::handleConversion):
(WebKit::PrivateClickMeasurementManager::startTimer):
(WebKit::PrivateClickMeasurementManager::convert):
(WebKit::PrivateClickMeasurementManager::fireConversionRequest):
(WebKit::PrivateClickMeasurementManager::firePendingConversionRequests):
(WebKit::PrivateClickMeasurementManager::clear):
(WebKit::PrivateClickMeasurementManager::clearForRegistrableDomain):
(WebKit::PrivateClickMeasurementManager::clearExpired):
(WebKit::PrivateClickMeasurementManager::toString const):
(WebKit::PrivateClickMeasurementManager::setConversionURLForTesting):
(WebKit::PrivateClickMeasurementManager::markAllUnconvertedAsExpiredForTesting):
(WebKit::PrivateClickMeasurementManager::debugModeEnabled const):
* NetworkProcess/PrivateClickMeasurementManager.h: Renamed from Source/WebKit/NetworkProcess/AdClickAttributionManager.h.
(WebKit::PrivateClickMeasurementManager::PrivateClickMeasurementManager):
(WebKit::PrivateClickMeasurementManager::m_sessionID):
(WebKit::PrivateClickMeasurementManager::setPingLoadFunction):
(WebKit::PrivateClickMeasurementManager::setOverrideTimerForTesting):
* Platform/Logging.h:
* Shared/NavigationActionData.cpp:
(WebKit::NavigationActionData::encode const):
(WebKit::NavigationActionData::decode):
* Shared/NavigationActionData.h:
* Shared/WebsiteData/WebsiteData.cpp:
(WebKit::WebsiteData::ownerProcess):
* Shared/WebsiteData/WebsiteDataType.h:
* Sources.txt:
* UIProcess/API/APINavigation.h:
(API::Navigation::privateClickMeasurement const):
(API::Navigation::adClickAttribution const): Deleted.
* UIProcess/API/C/WKPage.cpp:
(WKPageDumpPrivateClickMeasurement):
(WKPageClearPrivateClickMeasurement):
(WKPageSetPrivateClickMeasurementOverrideTimerForTesting):
(WKPageSetPrivateClickMeasurementConversionURLForTesting):
(WKPageMarkPrivateClickMeasurementsAsExpiredForTesting):
(WKPageDumpAdClickAttribution): Deleted.
(WKPageClearAdClickAttribution): Deleted.
(WKPageSetAdClickAttributionOverrideTimerForTesting): Deleted.
(WKPageSetAdClickAttributionConversionURLForTesting): Deleted.
(WKPageMarkAdClickAttributionsAsExpiredForTesting): Deleted.
* UIProcess/API/C/WKPagePrivate.h:
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreClearPrivateClickMeasurementsThroughWebsiteDataRemoval):
(WKWebsiteDataStoreClearAdClickAttributionsThroughWebsiteDataRemoval): Deleted.
* UIProcess/API/C/WKWebsiteDataStoreRef.h:
* UIProcess/API/Cocoa/WKWebsiteDataRecord.mm:
(dataTypesToString):
* UIProcess/API/Cocoa/WKWebsiteDataRecordInternal.h:
(WebKit::toWebsiteDataType):
(WebKit::toWKWebsiteDataTypes):
* UIProcess/API/Cocoa/WKWebsiteDataRecordPrivate.h:
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
(+[WKWebsiteDataStore _allWebsiteDataTypesIncludingPrivate]):
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeNetworkProcess):
* UIProcess/Inspector/WebInspectorProxy.cpp:
(WebKit::WebInspectorProxy::setDeveloperPreferenceOverride):
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::setPrivateClickMeasurementDebugMode):
(WebKit::NetworkProcessProxy::setAdClickAttributionDebugMode): Deleted.
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didCommitLoadForFrame):
(WebKit::WebPageProxy::createNewPage):
(WebKit::WebPageProxy::dumpPrivateClickMeasurement):
(WebKit::WebPageProxy::clearPrivateClickMeasurement):
(WebKit::WebPageProxy::setPrivateClickMeasurementOverrideTimerForTesting):
(WebKit::WebPageProxy::setPrivateClickMeasurementConversionURLForTesting):
(WebKit::WebPageProxy::markPrivateClickMeasurementsAsExpiredForTesting):
(WebKit::WebPageProxy::dumpAdClickAttribution): Deleted.
(WebKit::WebPageProxy::clearAdClickAttribution): Deleted.
(WebKit::WebPageProxy::setAdClickAttributionOverrideTimerForTesting): Deleted.
(WebKit::WebPageProxy::setAdClickAttributionConversionURLForTesting): Deleted.
(WebKit::WebPageProxy::markAdClickAttributionsAsExpiredForTesting): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::setPrivateClickMeasurementDebugMode):
(WebKit::WebsiteDataStore::setAdClickAttributionDebugMode): Deleted.
* UIProcess/WebsiteData/WebsiteDataStore.h:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::createWindow):
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::dispatchDecidePolicyForNewWindowAction):
(WebKit::WebFrameLoaderClient::dispatchDecidePolicyForNavigationAction):
2020-11-10 Simon Fraser <simon.fraser@apple.com>
Make PlatformWheelEventPhase an enum class
https://bugs.webkit.org/show_bug.cgi?id=218772
Reviewed by Tim Horton.
Change PlatformWheelEventPhase to be an enum class.
Changed code that maps between NSEventPhase and PlatformWheelEventPhase to not
treat NSEventPhase as a set of bits, since it only ever contains one of the bits.
* UIProcess/RemoteLayerTree/mac/ScrollerPairMac.mm:
(WebKit::ScrollerPairMac::handleWheelEvent):
* WebProcess/WebPage/EventDispatcher.cpp:
(WebKit::EventDispatcher::wheelEvent):
2020-11-10 Chris Dumez <cdumez@apple.com>
GPUProcess does not properly get relaunched after crashing
https://bugs.webkit.org/show_bug.cgi?id=218769
Reviewed by Tim Horton.
After the GPUProcess crashes, we need to make sure we construct a new GPUProcessProxy
so that a new process gets relaunched. Previously, this wasn't possible because
the GPUProcessProxy was a "never destroyed" singleton.
The GPUProcessProxy is now kept alive by the WebProcessPool objects refing it.
Upon crashing, all WebProcessPools clear their GPUProcessProxy RefPtr so that the
old GPUProcessProxy gets destroyed. The next time a GPUProcessProxy is needed, a new
one gets constructed. We still make sure that all WebProcessPool objects share the
same GPUProcessProxy, as before this patch.
* UIProcess/API/Cocoa/WKProcessPool.mm:
(-[WKProcessPool _gpuProcessIdentifier]):
* UIProcess/API/Cocoa/WKProcessPoolPrivate.h:
* UIProcess/GPU/GPUProcessProxy.cpp:
(WebKit::GPUProcessProxy::getOrCreate):
(WebKit::GPUProcessProxy::GPUProcessProxy):
(WebKit::GPUProcessProxy::~GPUProcessProxy):
(WebKit::GPUProcessProxy::processWillShutDown):
(WebKit::GPUProcessProxy::gpuProcessCrashed):
* UIProcess/GPU/GPUProcessProxy.h:
* UIProcess/UserMediaPermissionRequestManagerProxy.cpp:
(WebKit::UserMediaPermissionRequestManagerProxy::syncWithWebCorePrefs const):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::willStartCapture):
(WebKit::WebPageProxy::setOrientationForMediaCapture):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::ensureGPUProcess):
(WebKit::WebProcessPool::gpuProcessCrashed):
(WebKit::WebProcessPool::getGPUProcessConnection):
(WebKit::WebProcessPool::updateAudibleMediaAssertions):
* UIProcess/WebProcessPool.h:
* UIProcess/ios/WKContentView.mm:
(-[WKContentView _setupVisibilityPropagationViewForGPUProcess]):
2020-11-10 Brent Fulgham <bfulgham@apple.com>
[macOS] Allow IOGLESBundleName for Apple Silicon Macs
https://bugs.webkit.org/show_bug.cgi?id=218774
<rdar://problem/67473780>
Unreviewed follow-up fix to r269649.
Two additional keys are needed: MetalPluginClassName and MetalPluginName
* WebProcess/com.apple.WebProcess.sb.in:
2020-11-10 Brent Fulgham <bfulgham@apple.com>
[macOS] Allow IOGLESBundleName for Apple Silicon Macs
https://bugs.webkit.org/show_bug.cgi?id=218774
<rdar://problem/67473780>
Reviewed by Tim Horton.
The WebContent should be able to read the IOGLESBundleName IOKit property on
Apple Silicon Macs. This will help the OpenGL stack make better choices about
which bundles to load when processing WebGL content.
* WebProcess/com.apple.WebProcess.sb.in:
2020-11-10 Said Abou-Hallawa <said@apple.com>
[macCatalyst] Fix various build breaks
https://bugs.webkit.org/show_bug.cgi?id=218771
Reviewed by Tim Horton.
Use NO_RETURN_DUE_TO_ASSERT for unreachable function.
* UIProcess/Cocoa/UserMediaPermissionRequestProxy.mm:
(WebKit::UserMediaPermissionRequestProxy::doDefaultAction):
* UIProcess/UserMediaPermissionRequestProxy.h:
2020-11-10 Per Arne Vollan <pvollan@apple.com>
[Cocoa] Send GPU sandbox extensions in Web page creation parameters
https://bugs.webkit.org/show_bug.cgi?id=218708
Reviewed by Brent Fulgham.
To pick up correct runtime settings, the GPU sandbox extensions should be sent to the WebContent process as part of the Web page creation parameters.
No new tests, covered by existing tests.
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
(WebKit::mediaRelatedMachServices): Deleted.
(WebKit::gpuIOKitClasses): Deleted.
* UIProcess/WebPageProxy.cpp:
(WebKit::gpuIOKitClasses):
(WebKit::mediaRelatedMachServices):
(WebKit::WebPageProxy::creationParameters):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-11-10 Per Arne Vollan <pvollan@apple.com>
[macCatalyst] Add entitlement for message filtering
https://bugs.webkit.org/show_bug.cgi?id=218768
<rdar://problem/71247464>
Reviewed by Tim Horton.
Add required message filtering entitlement for Catalyst.
No new tests, covered by existing tests.
* Scripts/process-entitlements.sh:
2020-11-10 Youenn Fablet <youenn@apple.com>
Update WebRTC liwebrtc to M87
https://bugs.webkit.org/show_bug.cgi?id=218436
Reviewed by Eric Carlson.
Update code now that fragmentation headers are computed at packetization time.
* Configurations/WebKit.xcconfig:
* GPUProcess/webrtc/LibWebRTCCodecsProxy.mm:
(WebKit::LibWebRTCCodecsProxy::createEncoder):
* Scripts/webkit/messages.py:
* Shared/RTCNetwork.h:
* WebProcess/GPU/webrtc/LibWebRTCCodecs.cpp:
(WebKit::LibWebRTCCodecs::completedEncoding):
* WebProcess/GPU/webrtc/LibWebRTCCodecs.h:
* WebProcess/GPU/webrtc/LibWebRTCCodecs.messages.in:
2020-11-10 Wenson Hsieh <wenson_hsieh@apple.com>
REGRESSION (r269525): Many layout tests crash when run under ASan
https://bugs.webkit.org/show_bug.cgi?id=218733
<rdar://problem/71206273>
Reviewed by Tim Horton.
Construct the item 8 bytes after the handle offset, instead of 1 byte (also, add a FIXME indicating that we some
way of achieving this without requiring WebKit2 to be aware of an item handle's memory layout).
* GPUProcess/graphics/RemoteRenderingBackend.h:
(WebKit::RemoteRenderingBackend::decodeAndCreate):
2020-11-10 Chris Dumez <cdumez@apple.com>
[GPUProcess] Regression(r268632) Garbage is rendered on speakers when using WebAudio
https://bugs.webkit.org/show_bug.cgi?id=218729
Reviewed by Eric Carlson.
RemoteAudioDestinationProxy::requestBuffer() was calling AudioDestinationCocoa::render()
and expecting RemoteAudioDestinationProxy::renderOnRenderingThead() to get called as
a result. It would take care of writing to the CARingBuffer and sending the IPC back
to the GPU process in renderOnRenderingThead(). The issue was that AudioDestinationCocoa
uses a PushPullFIFO internally for buffering. It first fetches available frames from
the FIFO and then only calls renderOnRenderingThead() with the number of frames that
remain to processed (usually 0). As a result, RemoteAudioDestinationProxy::renderOnRenderingThead()
would often store 0 frames instead of 128 (or sometimes a number of frames less than
128), even though the full 128 frames were actually rendered.
To address the issue, stop overriding renderOnRenderingThead() in
RemoteAudioDestinationProxy. Instead, do the writing to the CARingBuffer and the IPC
response in RemoteAudioDestinationProxy::requestBuffer(), directly after calling
AudioDestinationCocoa::render(). After calling AudioDestinationCocoa::render()
we know that |framesToRender| frames have been rendered / added to the buffer.
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::requestBuffer):
* WebProcess/GPU/media/RemoteAudioDestinationProxy.h:
2020-11-10 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] WebView is flipped
https://bugs.webkit.org/show_bug.cgi?id=218354
Reviewed by Adrian Perez de Castro.
Stop painting y-flipped textures when buildig with GTK4.
* UIProcess/gtk/AcceleratedBackingStoreWayland.cpp:
(WebKit::AcceleratedBackingStoreWayland::snapshot):
* WebProcess/WebPage/libwpe/AcceleratedSurfaceLibWPE.h:
2020-11-10 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] Crash in WebKit::DropTarget::drop
https://bugs.webkit.org/show_bug.cgi?id=217482
Reviewed by Michael Catanzaro.
If we don't have selection data when drop is called, just return early to let leave continue. Also change
accept() to receive the drop context and position to be set after leaving any previous operation.
* UIProcess/API/gtk/DropTarget.h:
* UIProcess/API/gtk/DropTargetGtk3.cpp:
(WebKit::DropTarget::DropTarget):
(WebKit::DropTarget::accept):
(WebKit::DropTarget::drop):
* UIProcess/API/gtk/DropTargetGtk4.cpp:
(WebKit::DropTarget::DropTarget):
(WebKit::DropTarget::accept):
2020-11-10 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK][WPE] WEBKIT_PLUGIN_ERROR_WILL_HANDLE_LOAD returned when plugins are disabled
https://bugs.webkit.org/show_bug.cgi?id=216123
Reviewed by Adrian Perez de Castro.
Undeprecate webkit_plugin_error_quark and WEBKIT_PLUGIN_ERROR_WILL_HANDLE_LOAD, and update documentation.
* UIProcess/API/glib/WebKitWebView.cpp:
(webkitWebViewLoadFail):
* UIProcess/API/gtk/WebKitError.h:
* UIProcess/API/wpe/WebKitError.h:
2020-11-10 Wenson Hsieh <wenson_hsieh@apple.com>
Fix crashes when mapping shared display list item data into display lists in the GPU process
https://bugs.webkit.org/show_bug.cgi?id=218739
Reviewed by Tim Horton.
I added a "stop-gap" mechanism in r269525 to allow the GPU process to handle newly refactored display list
items by encoding display list items directly into shared memory buffers in the web process and reading them
back in the GPU process, using standard IPC messages only, and also without reusing any shared memory buffers.
However, I introduced a bug in this "stop-gap" mechanism when I tried to clean up some GPU-process-side logic
for unregistering these shared memory buffers in the GPU process, by taking shared memory buffers out of the
map in the GPU process without ensuring that they're kept alive over the course of applying the display list.
This code is going away soon anyways (see webkit.org/b/218426), but in the meantime, we should at least ensure
that this temporary shared memory mechanism doesn't cause huge waves of layout test failures.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::applyDisplayList):
2020-11-09 Said Abou-Hallawa <said@apple.com>
[GPU Process] Control the life cycle of the platform image by a new class named NativeImage
https://bugs.webkit.org/show_bug.cgi?id=218427
Reviewed by Simon Fraser.
* Scripts/webkit/messages.py:
* Shared/API/c/cg/WKImageCG.cpp:
(WKImageCreateFromCGImage):
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<Ref<NativeImage>>::encode):
(IPC::ArgumentCoder<Ref<NativeImage>>::decode):
(IPC::encodeNativeImage): Deleted.
(IPC::decodeNativeImage): Deleted.
(IPC::encodeOptionalNativeImage): Deleted.
(IPC::decodeOptionalNativeImage): Deleted.
(IPC::ArgumentCoder<NativeImagePtr>::encode): Deleted.
(IPC::ArgumentCoder<NativeImagePtr>::decode): Deleted.
* Shared/WebCoreArgumentCoders.h:
* Shared/cg/ShareableBitmapCG.cpp:
(WebKit::ShareableBitmap::createImage):
* Shared/win/ShareableBitmapDirect2D.cpp:
(WebKit::ShareableBitmap::paint):
* UIProcess/API/glib/IconDatabase.cpp:
(WebKit::IconDatabase::loadIconForPageURL):
(WebKit::IconDatabase::setIconForPageURL):
* UIProcess/API/glib/IconDatabase.h:
* UIProcess/ios/DragDropInteractionState.mm:
(WebKit::uiImageForImage):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _deliverDelayedDropPreviewIfPossible:]):
(uiImageForImage):
(-[WKContentView _handleDropByInsertingImagePlaceholders:session:]):
(-[WKContentView _presentationSnapshotForPreviewItemController:]):
* WebProcess/GPU/graphics/ImageBufferShareableBitmapBackend.cpp:
(WebKit::ImageBufferShareableBitmapBackend::copyNativeImage const):
* WebProcess/GPU/graphics/ImageBufferShareableBitmapBackend.h:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::nativeImageForCurrentTime):
* WebProcess/GPU/media/MediaPlayerPrivateRemote.h:
* WebProcess/Plugins/PluginView.cpp:
(WebKit::isAlmostSolidColor):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::didChooseFilesForOpenPanelWithDisplayStringAndIcon):
2020-11-09 Chris Dumez <cdumez@apple.com>
Unexpose obsolete HTMLAppletElement interface
https://bugs.webkit.org/show_bug.cgi?id=218677
Reviewed by Darin Adler.
* WebProcess/InjectedBundle/API/gtk/DOM/WebKitDOMHTMLAppletElement.cpp:
(webkit_dom_html_applet_element_dispatch_event):
(webkit_dom_html_applet_element_add_event_listener):
(webkit_dom_html_applet_element_remove_event_listener):
(webkit_dom_html_applet_element_get_align):
(webkit_dom_html_applet_element_set_align):
(webkit_dom_html_applet_element_get_alt):
(webkit_dom_html_applet_element_set_alt):
(webkit_dom_html_applet_element_get_archive):
(webkit_dom_html_applet_element_set_archive):
(webkit_dom_html_applet_element_get_code):
(webkit_dom_html_applet_element_set_code):
(webkit_dom_html_applet_element_get_code_base):
(webkit_dom_html_applet_element_set_code_base):
(webkit_dom_html_applet_element_get_height):
(webkit_dom_html_applet_element_set_height):
(webkit_dom_html_applet_element_get_hspace):
(webkit_dom_html_applet_element_set_hspace):
(webkit_dom_html_applet_element_get_name):
(webkit_dom_html_applet_element_set_name):
(webkit_dom_html_applet_element_get_object):
(webkit_dom_html_applet_element_set_object):
(webkit_dom_html_applet_element_get_vspace):
(webkit_dom_html_applet_element_set_vspace):
(webkit_dom_html_applet_element_get_width):
(webkit_dom_html_applet_element_set_width):
* WebProcess/InjectedBundle/API/gtk/DOM/WebKitDOMHTMLAppletElementPrivate.h:
* WebProcess/InjectedBundle/API/gtk/DOM/WebKitDOMHTMLPrivate.cpp:
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
* WebProcess/WebCoreSupport/WebFrameLoaderClient.h:
2020-11-09 Peng Liu <peng.liu6@apple.com>
REGRESSION (r269557): ASSERTION FAILED: Completion handler should always be called under WebCore::HTMLMediaElement::exitFullscreen
https://bugs.webkit.org/show_bug.cgi?id=218722
Reviewed by Eric Carlson.
No new tests, covered by an existing test that is crashing on the bots.
* WebProcess/cocoa/VideoFullscreenManager.mm:
(WebKit::VideoFullscreenManager::exitVideoFullscreenForVideoElement):
2020-11-09 Wenson Hsieh <wenson_hsieh@apple.com>
Rename ImageBuffer::flushDisplayList to ImageBuffer::submitDisplayList
https://bugs.webkit.org/show_bug.cgi?id=218720
Reviewed by Tim Horton.
See Source/WebCore/ChangeLog for more details.
* GPUProcess/graphics/RemoteImageBuffer.h:
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::applyDisplayList):
(WebKit::RemoteRenderingBackend::submitDisplayList):
(WebKit::RemoteRenderingBackend::flushDisplayList): Deleted.
Additionally rename `flushDisplayList` on `RemoteRenderingBackend`, since this IPC message doesn't actually
perform a context flush.
* GPUProcess/graphics/RemoteRenderingBackend.h:
* GPUProcess/graphics/RemoteRenderingBackend.messages.in:
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::submitDisplayList):
(WebKit::RemoteRenderingBackendProxy::flushDisplayList): Deleted.
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
2020-11-09 Alan Bujtas <zalan@apple.com>
Unreviewed, reverting r269603.
Needs Safari patch
Reverted changeset:
"Show legacy line layout visual coverage instead of "simple
line" layout."
https://bugs.webkit.org/show_bug.cgi?id=218695
https://trac.webkit.org/changeset/269603
2020-11-09 Zalan Bujtas <zalan@apple.com>
Show legacy line layout visual coverage instead of "simple line" layout.
https://bugs.webkit.org/show_bug.cgi?id=218695
Reviewed by Antti Koivisto.
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetLegacyLineLayoutVisualCoverageEnabled):
(WKPreferencesGetLegacyLineLayoutVisualCoverageEnabled):
(WKPreferencesSetSimpleLineLayoutDebugBordersEnabled): Deleted.
(WKPreferencesGetSimpleLineLayoutDebugBordersEnabled): Deleted.
* UIProcess/API/C/WKPreferencesRefPrivate.h:
* UIProcess/API/Cocoa/WKPreferences.mm:
(-[WKPreferences _legacyLineLayoutVisualCoverageEnabled]):
(-[WKPreferences _setLegacyLineLayoutVisualCoverageEnabled:]):
(-[WKPreferences _simpleLineLayoutDebugBordersEnabled]): Deleted.
(-[WKPreferences _setSimpleLineLayoutDebugBordersEnabled:]): Deleted.
* UIProcess/API/Cocoa/WKPreferencesPrivate.h:
2020-11-09 Per Arne Vollan <pvollan@apple.com>
[macOS] System sounds should be played in the UI process
https://bugs.webkit.org/show_bug.cgi?id=218405
<rdar://problem/70898846>
Reviewed by Darin Adler.
Subclass SystemSoundDelegate, which will forward system sound requests to the UI process.
* Sources.txt:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::systemBeep):
* UIProcess/WebProcessProxy.h:
* UIProcess/WebProcessProxy.messages.in:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebSystemSoundDelegate.cpp: Added.
(WebKit::WebSystemSoundDelegate::systemBeep):
* WebProcess/WebSystemSoundDelegate.h: Added.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-11-09 Per Arne Vollan <pvollan@apple.com>
[Cocoa] Diagnostic service sandbox violation loading any webpage
https://bugs.webkit.org/show_bug.cgi?id=218710
<rdar://problem/71124931>
Reviewed by Alexey Proskuryakov.
Loading any Web page will cause a diagnostic service mach-lookup sandbox violation. Consume the extension to
the service as early as possible in WebProcess::platformInitializeWebProcess, since there are code paths in
this method which will try to access the diagnostic service.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-11-09 Devin Rousso <drousso@apple.com>
autofocus of text input should not select text
https://bugs.webkit.org/show_bug.cgi?id=218585
<rdar://problem/60130704>
Reviewed by Wenson Hsieh.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::restoreSelectionInFocusedEditableElement):
Slightly rework `SelectionRestorationMode` to replace `SetDefault` with two new values:
- `PlaceCaretAtStart` puts the caret at the start, regardless of any cached selection
- `SelectAll` selects all text, regardless of any cached selection (existing behavior)
In order to preserve existing behavior, the default `Restore` will have the same effect as
`SelectAll` if there is no cached selection (and is renamed to `RestoreOrSelectAll` as such).
2020-11-09 Per Arne Vollan <pvollan@apple.com>
[macOS] Set preference for overridden languages in the WebContent process after entering the sandbox.
https://bugs.webkit.org/show_bug.cgi?id=218097
<rdar://problem/70586545>
Reviewed by Alexey Proskuryakov.
Currently, the preference for overridden languages in the WebContent process is set before entering the sandbox,
which leaves behind an open connection to opendirectoryd. This preference should be set after entering the
sandbox to avoid this. This patch sets the preference after the initialization function has been executed.
API test: WebKit.OverrideAppleLanguagesPreference
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceMain.mm:
(WebKit::setAppleLanguagesPreference):
(WebKit::XPCServiceEventHandler):
(WebKit::XPCServiceMain):
* WebProcess/com.apple.WebProcess.sb.in:
2020-11-09 Joonghun Park <jh718.park@samsung.com>
Unreviewed. Fix the build warning below since r269525.
warning: control reaches end of non-void function [-Wreturn-type]
No new tests, no new behaviors.
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
2020-11-09 Zan Dobersek <zdobersek@igalia.com>
[WPE] ScrollGestureController should only handle complete touch interactions
https://bugs.webkit.org/show_bug.cgi?id=218329
Reviewed by Carlos Garcia Campos.
ScrollGestureController's purpose is to handle otherwise-unhandled
touch events and translate those into a scrolling gesture.
Until now, it was possible for a touch motion event to trigger that
gesture even when the corresponding touch down event was not handled
by the ScrollGestureController, e.g. it was not produced due to the
touch interaction being started somewhere else, or the Web content
actually handled that event.
A flag is added to ScrollGestureController to track that an unhandled
touch down event was confronted, before allowing subsequent unhandled
touch motion events to trigger the scrolling gesture when conditions
are met. The flag is reset upon the touch up event.
* UIProcess/API/wpe/ScrollGestureController.cpp:
(WebKit::ScrollGestureController::handleEvent):
* UIProcess/API/wpe/ScrollGestureController.h:
2020-11-08 Lauro Moura <lmoura@igalia.com>
[GTK] REGRESSION(r267250) API test /webkit/WebKitWebView/usermedia-enumeratedevices-permission-check is crashing
https://bugs.webkit.org/show_bug.cgi?id=216727
Reviewed by Carlos Garcia Campos.
Coverered by existing tests.
In some cases, like in TestWebKitAPI, a message can arrive to the
GeolocationProvider and it would try to access a WebGeolocationManager
that could already had been deleted.
* UIProcess/API/glib/WebKitGeolocationManager.cpp:
(webkitGeolocationManagerDispose): Added. Removes the provider.
(webkit_geolocation_manager_class_init): Register dispose callback.
2020-11-07 Wenson Hsieh <wenson_hsieh@apple.com>
Clean up some logic in RemoteImageBuffer and RemoteRenderingBackend
https://bugs.webkit.org/show_bug.cgi?id=218689
Reviewed by Tim Horton.
Move the implementation of `decodeItem` out of `RemoteImageBuffer`, and into `RemoteRenderingBackend`. This
allows us to avoid `static_cast`-ing image buffers to their unaccelerated or accelerated versions when setting
the display list's item buffer client in the GPU process.
No change in behavior.
* GPUProcess/graphics/RemoteImageBuffer.h:
(WebKit::RemoteImageBuffer::decodeAndCreate): Deleted.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::createImageBuffer):
(WebKit::RemoteRenderingBackend::applyDisplayList):
Also, remove a couple of unnecessary `WebCore::` namespace prefixes while we're here.
(WebKit::RemoteRenderingBackend::decodeItem):
Move `decodeItem` into the rendering backend. This also has the added bonus of letting us put the method
definition in the implementation file instead of a header (i.e. `RemoteRenderingBackend.cpp`).
* GPUProcess/graphics/RemoteRenderingBackend.h:
(WebKit::RemoteRenderingBackend::decodeAndCreate):
2020-11-06 Peng Liu <peng.liu6@apple.com>
PiP return to element full screen is not smooth on some websites
https://bugs.webkit.org/show_bug.cgi?id=218419
Reviewed by Jer Noble.
In r265562, VideoFullscreenInterfaceAVKit uses `fullscreenMayReturnToInline()`
to request the observer (WKFullScreenWindowController) to enter element fullscreen
before exiting picture-in-picture, and calls `fullscreenWillReturnToInline()` to
proceed returning to inline or element fullscreen after WKFullScreenWindowController
is ready.
In this patch, VideoFullscreenInterfaceAVKit will directly enter a temporary
video fullscreen from picture-in-picture, so the preparation step (entering element
fullscreen) is not needed.
Some important points regarding edge cases:
1) A video enters picture-in-picture and exits picture-in-picture immediately
before WKFullScreenWindowController is notified to exit element fullscreen.
In this case, WKFullScreenWindowController will do nothing when it is notified
that the video has exited picture-in-picture. It will directly tell
VideoFullscreenInterfaceAVKit to enter standby.
2) A video exits picture-in-picture when WKFullScreenWindowController is exiting
element fullscreen but not complete yet. In this case, WKFullScreenWindowController
needs to hold the enter element fullscreen request and exits element fullscreen
first before entering element fullscreen.
3) The ancestor element of the video element exits fullscreen and asks the video
element to exit standby (in web process) while the video is entering picture-in-picture
(in UI process). In this case, VideoFullscreenInterfaceAVKit (in UI process) should
ignore the request to exit standby and tell the web process the request is rejected.
4) A user may try to enter picture-in-picture when the video is exiting picture-in-picture
and returning to element fullscreen. In this case, VideoFullscreenInterfaceAVKit
(in UI process) should temporarily disable the picture-in-picture support.
* UIProcess/Cocoa/VideoFullscreenManagerProxy.h:
* UIProcess/Cocoa/VideoFullscreenManagerProxy.messages.in:
Remove the message `EnterFullscreen` for the iOS port because the UI process can enter
fullscreen/picture-in-picture directly.
* UIProcess/Cocoa/VideoFullscreenManagerProxy.mm:
(WebKit::VideoFullscreenManagerProxy::exitFullscreen):
(WebKit::VideoFullscreenManagerProxy::preparedToReturnToInline):
We need to call `m_page->fullscreenMayReturnToInline()` in this function after
removing `fullscreenWillReturnToInline()`.
(WebKit::VideoFullscreenManagerProxy::didSetupFullscreen):
For the iOS port, we should enter fullscreen directly.
(WebKit::VideoFullscreenManagerProxy::didCleanupFullscreen):
(WebKit::VideoFullscreenManagerProxy::fullscreenMayReturnToInline):
(WebKit::VideoFullscreenModelContext::fullscreenWillReturnToInline): Deleted.
(WebKit::VideoFullscreenManagerProxy::fullscreenWillReturnToInline): Deleted.
* UIProcess/ios/fullscreen/WKFullScreenWindowControllerIOS.mm:
(-[WKFullScreenWindowController beganEnterFullScreenWithInitialFrame:finalFrame:]):
(-[WKFullScreenWindowController requestEnterFullScreen]):
(-[WKFullScreenWindowController exitFullScreen]):
(-[WKFullScreenWindowController _completedExitFullScreen]):
(-[WKFullScreenWindowController didEnterPictureInPicture]):
(-[WKFullScreenWindowController didExitPictureInPicture]):
(-[WKFullScreenWindowController _exitFullscreenImmediately]):
We should force to exit fullscreen because the current exiting fullscreen process
might be stuck.
(-[WKFullScreenWindowController prepareToExitPictureInPicture]): Deleted.
* UIProcess/mac/WKFullScreenWindowController.mm:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::exitVideoFullscreenForVideoElement):
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/cocoa/VideoFullscreenManager.h:
(WebKit::VideoFullscreenManager::exitVideoFullscreenForVideoElement):
* WebProcess/cocoa/VideoFullscreenManager.messages.in:
* WebProcess/cocoa/VideoFullscreenManager.mm:
(WebKit::VideoFullscreenManager::enterVideoFullscreenForVideoElement):
(WebKit::VideoFullscreenManager::exitVideoFullscreenForVideoElement):
The request to exit fullscreen (standby) may be rejected.
(WebKit::VideoFullscreenManager::didSetupFullscreen):
Only the macOS port needs to implement this function.
(WebKit::VideoFullscreenManager::didEnterFullscreen):
(WebKit::VideoFullscreenManager::didExitFullscreen):
(WebKit::VideoFullscreenManager::fullscreenMayReturnToInline):
(WebKit::VideoFullscreenManager::fullscreenWillReturnToInline): Deleted.
2020-11-06 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] [iOS] WebAuthn process doesn't start on iOS devices
https://bugs.webkit.org/show_bug.cgi?id=218479
<rdar://problem/70560399>
Reviewed by Brent Fulgham.
One of the reasons is missing the iOS sandbox profile. Therefore, this patch adds one.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebAuthn.sb: Added.
Copied from the GPU process. Will strip it down once the process is fully functional.
* Scripts/process-entitlements.sh:
Adds some missing entitlements.
* UIProcess/WebAuthentication/WebAuthnProcessProxy.h:
Deals with CompletionHandler assertion.
* WebKit.xcodeproj/project.pbxproj:
Adds some plumbings for the sandbox profile.
2020-11-06 Wenson Hsieh <wenson_hsieh@apple.com>
Add new display list item types in preparation for webkit.org/b/218426
https://bugs.webkit.org/show_bug.cgi?id=218588
Reviewed by Simon Fraser.
Simply treat these new items as no-ops for the time being. See Source/WebCore/ChangeLog for more details.
* GPUProcess/graphics/RemoteImageBuffer.h:
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
2020-11-06 Per Arne Vollan <pvollan@apple.com>
[macOS] Add IOKIt message filtering
https://bugs.webkit.org/show_bug.cgi?id=218657
<rdar://problem/71118545>
Reviewed by Brent Fulgham.
Like on iOS, the WebContent process on macOS should have IOKit message filtering. This patch also allows reading
of some IOKIt properties, which was reported when running layout tests.
No new tests, covered by existing tests.
* WebProcess/com.apple.WebProcess.sb.in:
2020-11-06 Wenson Hsieh <wenson_hsieh@apple.com>
Add a display list item to represent stroking a single line
https://bugs.webkit.org/show_bug.cgi?id=218589
Reviewed by Tim Horton.
* GPUProcess/graphics/RemoteImageBuffer.h:
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
2020-11-06 Sam Weinig <weinig@apple.com>
Further progress towards merging Settings.yaml into WebPreferences.yaml
https://bugs.webkit.org/show_bug.cgi?id=218663
Reviewed by Tim Horton.
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration init]):
Replace call to WebCore::Settings::defaultMediaContentTypesRequiringHardwareSupport() which always returned
the empty string with a literal empty string. This was the only caller so the function is really no needed.
2020-11-06 Aditya Keerthi <akeerthi@apple.com>
[macOS] Align left edge of presented calendar with left edge of date inputs
https://bugs.webkit.org/show_bug.cgi?id=218668
Reviewed by Tim Horton.
When a date input is activated, a calendar is presented by adding an
NSWindow containing an NSDatePicker to the screen. The frame of the
window is set to ensure proper alignment with the date input. However,
since the window's hasShadow is set to true, AppKit draws a thin border
and shadow around the window's contentView. This results in left border
of the window being drawn further out than the left border of the input
itself. If the hasShadow property is set to false, the issue is no longer
observed, and the calendar aligns with the input. However, this is not an
acceptable solution, since we lose the border and the shadow.
To ensure alignment between the border of the calendar view and the border
of the date input, we can inset the edges of the view. Note that the inset
is not applied along the y-axis, to avoid a gap between the bottom of the
date input and the top of the calendar view.
* UIProcess/mac/WebDateTimePickerMac.mm:
(-[WKDateTimePickerBackdropView drawRect:]):
2020-11-06 Alex Christensen <achristensen@webkit.org>
Fix build with different unified source arrangement.
* UIProcess/WebAuthentication/Cocoa/LocalAuthenticator.mm:
2020-11-06 Wenson Hsieh <wenson_hsieh@apple.com>
Move DisplayListFlushIdentifier into WebCore as DisplayList::FlushIdentifier
https://bugs.webkit.org/show_bug.cgi?id=218586
Reviewed by Simon Fraser.
In preparation for adding a display list item to represent flushing the graphics context, move WebKit's
`DisplayListFlushIdentifier` to the `DisplayList` namespace in WebCore, and rename it to
`DisplayList::FlushIdentifier`.
No change in behavior.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::flushDisplayListWasCommitted):
(WebKit::RemoteRenderingBackend::flushDisplayListAndCommit):
* GPUProcess/graphics/RemoteRenderingBackend.h:
* GPUProcess/graphics/RemoteRenderingBackend.messages.in:
* Scripts/webkit/messages.py:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/graphics/DisplayListFlushIdentifier.h: Removed.
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
(WebKit::RemoteImageBufferProxy::commitFlushDisplayList):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::flushDisplayListAndCommit):
(WebKit::RemoteRenderingBackendProxy::flushDisplayListWasCommitted):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.messages.in:
2020-11-06 Alex Christensen <achristensen@webkit.org>
Fix build after adding files to UIProcess/API/Cocoa
https://bugs.webkit.org/show_bug.cgi?id=218664
There is some code that assumes that unified sources always combine sources into the same translation units.
inspectorWebView is used from _WKInspectorTesting.mm, but only defined in _WKInspector.mm so it should be added to a header.
_WKUserContentWorld also doesn't include WKObject.h or RetainPtr.h but uses both.
* UIProcess/API/Cocoa/_WKInspector.mm:
(-[_WKInspector inspectorWebView]): Deleted.
* UIProcess/API/Cocoa/_WKInspectorPrivateForTesting.h:
* UIProcess/API/Cocoa/_WKInspectorTesting.mm:
(-[_WKInspector inspectorWebView]):
* UIProcess/API/Cocoa/_WKUserContentWorldInternal.h:
2020-11-06 Wenson Hsieh <wenson_hsieh@apple.com>
Encoding PutImageData should not serialize separate IPC attachments for the image data
https://bugs.webkit.org/show_bug.cgi?id=218649
Reviewed by Simon Fraser.
The argument coder for `WebCore::ImageData` currently sends its image data by allocating a separate shared
memory buffer and sending it over to the GPU process via a separate `IPC::Attachment`. This isn't compatible
with the new "concurrent display list" model for processing display list items, wherein all out-of-line items
are encoded as raw bytes directly into reusable shared memory buffers and decoded from the corresponding buffer
in the GPU process.
Instead, encode and decode image data directly into and out of the main IPC data buffer. This also addresses a
couple of existing FIXMEs in the ImageData coder methods by avoiding the need for redundant copies to and from
temporary `WebCore::SharedBuffer`s.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<Ref<WebCore::ImageData>>::encode):
(IPC::ArgumentCoder<Ref<WebCore::ImageData>>::decode):
2020-11-06 Wenson Hsieh <wenson_hsieh@apple.com>
[Concurrent display lists] Encode display list items directly into shared memory
https://bugs.webkit.org/show_bug.cgi?id=218406
Reviewed by Tim Horton.
Adjust for changes to display lists and display list items in WebCore (see Source/WebCore/ChangeLog for more
information). In particular, we implement the reading and writing client hooks consulted by ItemBuffer, and we
also add a temporary mechanism that allows RemoteRenderingBackendProxy (in the web process) to send display list
item data through shared memory to the RemoteRenderingBackend (in the GPU process). This temporary mechanism
does not attempt to make any reading or writing in shared memory concurrent between the GPU and web processes,
and exists only to make sure that rendering with the GPU process still works using these new display list items.
In the next patch, I will add a simple concurrent reader/writer model for display list processing between the
web and GPU processes, and (in doing so) revert most of the changes in `RemoteRenderingBackend` and
`RemoteRenderingBackendProxy` below.
See comments below for more detail.
* GPUProcess/graphics/RemoteImageBuffer.h:
(WebKit::RemoteImageBuffer::decodeAndCreate):
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::applyResourceItem):
(WebKit::RemoteRenderingBackend::applyMediaItem):
Refactor these to take `WebCore::DisplayList::ItemHandle`.
(WebKit::RemoteRenderingBackend::applyDisplayList):
(WebKit::RemoteRenderingBackend::flushDisplayList):
(WebKit::RemoteRenderingBackend::flushDisplayListAndCommit):
(WebKit::RemoteRenderingBackend::didCreateSharedItemData):
* GPUProcess/graphics/RemoteRenderingBackend.h:
* GPUProcess/graphics/RemoteRenderingBackend.messages.in:
* Scripts/webkit/messages.py:
* Shared/SharedDisplayListHandle.cpp: Added.
(WebKit::SharedDisplayListHandle::SharedDisplayListHandle):
(WebKit::SharedDisplayListHandle::createDisplayList const):
* Shared/SharedDisplayListHandle.h: Added.
Add a WebKit2 helper class that represents display list data in shared memory that is propagated from the web
process to the GPU process. In the next patch, this class will be rewritten to support concurrent display list
reading and writing, with specialized subclasses in service of both the reader (i.e. the GPU process) and the
writer (i.e. the web process).
(WebKit::SharedDisplayListHandle::SharedDisplayListHandle):
(WebKit::SharedDisplayListHandle::encode const):
(WebKit::SharedDisplayListHandle::decode):
* Sources.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
(WebKit::RemoteImageBufferProxy::RemoteImageBufferProxy):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::flushDisplayList):
(WebKit::RemoteRenderingBackendProxy::flushDisplayListAndCommit):
(WebKit::RemoteRenderingBackendProxy::createItemBuffer):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.messages.in:
* WebProcess/GPU/graphics/RemoteResourceCacheProxy.cpp:
2020-11-06 Truitt Savell <tsavell@apple.com>
Unreviewed, reverting r269486.
Caused 50+ timeouts on Mac Debug WK2
Reverted changeset:
"[Cocoa] Inspector Extensions: Add _WKInspectorExtension and
related plumbing"
https://bugs.webkit.org/show_bug.cgi?id=217783
https://trac.webkit.org/changeset/269486
2020-11-06 Wenson Hsieh <wenson_hsieh@apple.com>
Add some missing header includes in WebAuthenticatorCoordinator.cpp
https://bugs.webkit.org/show_bug.cgi?id=218661
Reviewed by Simon Fraser.
Add the missing headers.
* WebProcess/WebAuthentication/WebAuthenticatorCoordinator.cpp:
2020-11-06 Brent Fulgham <bfulgham@apple.com>
[macOS] Allow IOAVDHEVCDecodeCapabilities on Intel
https://bugs.webkit.org/show_bug.cgi?id=218639
<rdar://problem/71100188>
Reviewed by Eric Carlson.
The current rule for IOAVDHEVCDecodeCapabilities is incorrect for macOS 11. We should
allow this access on Intel, too.
* GPUProcess/mac/com.apple.WebKit.GPUProcess.sb.in:
* WebProcess/com.apple.WebProcess.sb.in:
2020-11-06 Milan Crha <mcrha@redhat.com>
[GTK] Application cannot override drag&drop callbacks
https://bugs.webkit.org/show_bug.cgi?id=218562
Reviewed by Michael Catanzaro.
* UIProcess/API/gtk/DropTargetGtk3.cpp:
(WebKit::DropTarget::DropTarget): Use g_signal_connect_after(), thus
any descendants can override the callbacks.
(WebKit::DropTarget::didPerformAction): Always call gdk_drag_status(),
to have gtk+ notified about drag progress.
2020-11-06 Michael Catanzaro <mcatanzaro@gnome.org>
[GTK] dlopen libWPEBackend-fdo-1.0.so.1 instead of libWPEBackend-fdo-1.0.so
https://bugs.webkit.org/show_bug.cgi?id=218584
Reviewed by Carlos Garcia Campos.
wpe_loader_init() tries to dlopen a development symlink, but these will not usually be
installed. It only works in the GNOME flatpak runtime because we don't use devel split
rules in the GNOME runtime, but we really ought to, and freedesktop-sdk already does, so
it's going to break upstream too once that gets implemented. It's easier for everyone if
we tackle this now. Since all major distros except Arch Linux use devel split, this patch
is really required for almost every distro to use WPE renderer.
The downside is that this makes it harder to update the wpebackend-fdo library version,
since minor changes there will now unnecessarily break WebKit. This is not good, but it's
better than forcing distros to apply this same patch downstream. I considered trying a bunch
of different library versions in a loop in order to maximize the chance of success in case
the wpebackend-fdo library version increases in the future, but it's probably nicer to just
be careful about not changing its library version until we find a better solution for this.
* UIProcess/glib/WebProcessPoolGLib.cpp:
(WebKit::WebProcessPool::platformInitializeWebProcess):
2020-11-05 Said Abou-Hallawa <said@apple.com>
[GPU Process] Use the Ref counting of ImageBuffer to control its life cycle in Web Process and GPU Process
https://bugs.webkit.org/show_bug.cgi?id=218529
Reviewed by Simon Fraser.
When a DrawImageBuffer item is appended to DisplayList, a Ref<ImageBuffer>
will be cached till the DisplayList items list is cleared. So no need to
lock the remote resource in RemoteResourceCacheProxy.
All we need to is to pass the ImageBufferHashMap which is maintained by
the DisplayList or by RemoteResourceCache to the DisplayList::Replayer.
It is going to be used to resolve the RenderingResourceIdentifer to an
ImageBuffer so the DrawImageBuffer can be applied.
No need to special-case DrawImageBuffer::apply() in the RemoteImageBuffer.
* GPUProcess/graphics/RemoteImageBuffer.h:
(WebKit::RemoteImageBuffer::RemoteImageBuffer):
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::createImageBuffer):
(WebKit::RemoteRenderingBackend::applyResourceItem): Deleted.
* GPUProcess/graphics/RemoteRenderingBackend.h:
(WebKit::RemoteRenderingBackend::remoteResourceCache):
* GPUProcess/graphics/RemoteResourceCache.cpp:
(WebKit::RemoteResourceCache::cacheImageBuffer):
* GPUProcess/graphics/RemoteResourceCache.h:
(WebKit::RemoteResourceCache::imageBuffers const):
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
(WebKit::RemoteImageBufferProxy::~RemoteImageBufferProxy):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::releaseRemoteResource):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
* WebProcess/GPU/graphics/RemoteResourceCacheProxy.cpp:
(WebKit::RemoteResourceCacheProxy::releaseImageBuffer):
(WebKit::RemoteResourceCacheProxy::RemoteResourceCacheProxy): Deleted.
(WebKit::RemoteResourceCacheProxy::lockRemoteImageBufferForRemoteClient): Deleted.
(WebKit::RemoteResourceCacheProxy::lockRemoteResourceForRemoteClient): Deleted.
(WebKit::RemoteResourceCacheProxy::releaseRemoteResource): Deleted.
(WebKit::RemoteResourceCacheProxy::unlockRemoteResourcesForRemoteClient): Deleted.
* WebProcess/GPU/graphics/RemoteResourceCacheProxy.h:
2020-11-05 Brent Fulgham <bfulgham@apple.com>
[macOS] Remove mdnsresponder access from WebKit processes
https://bugs.webkit.org/show_bug.cgi?id=218633
<rdar://problem/69133548>
Unreviewed follow-up: Make the sandbox rule conditional on recent macOS to
avoid breaking STP use on older macOS releases.
* GPUProcess/mac/com.apple.WebKit.GPUProcess.sb.in:
* WebProcess/com.apple.WebProcess.sb.in:
2020-11-05 Alex Christensen <achristensen@webkit.org>
Add null checks effectively in UserInputBridge
https://bugs.webkit.org/show_bug.cgi?id=218622
<rdar://problem/70724960>
Reviewed by Wenson Hsieh.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::stopLoadingFrame):
(WebKit::WebPage::stopLoading):
(WebKit::WebPage::reload):
2020-11-05 Brent Fulgham <bfulgham@apple.com>
[macOS] Remove mdnsresponder access from WebKit processes
https://bugs.webkit.org/show_bug.cgi?id=218633
<rdar://problem/69133548>
Reviewed by Per Arne Vollan.
Now that all network loading (including media) happens in the Network process,
we can remove an old permission to open outbound connections to mdnsresponder.
This is also unneeded in the GPU process, for the same reason.
I noticed that the new WebAuthn process sandbox mistakenly included this,
and removed it there, too.
* GPUProcess/mac/com.apple.WebKit.GPUProcess.sb.in:
* WebAuthnProcess/mac/com.apple.WebKit.WebAuthnProcess.sb.in:
* WebProcess/com.apple.WebProcess.sb.in:
2020-11-05 John Wilander <wilander@apple.com>
PCM: Switch to JSON report format
https://bugs.webkit.org/show_bug.cgi?id=218634
<rdar://problem/70469923>
Reviewed by Brent Fulgham.
The standards conversation has landed in attribution reports in a JSON
format as opposed to the currently implemented URL format. Discussion
here: https://github.com/privacycg/private-click-measurement/issues/30
* NetworkProcess/AdClickAttributionManager.cpp:
(WebKit::AdClickAttributionManager::fireConversionRequest):
2020-11-05 Brian Burg <bburg@apple.com>
[Cocoa] Inspector Extensions: Add _WKInspectorExtension and related plumbing
https://bugs.webkit.org/show_bug.cgi?id=217783
<rdar://problem/69968787>
Reviewed by Devin Rousso.
_WKInspectorExtension is new SPI that represents a Web Extension that uses the 'devtools'
API to interact with Web Inspector. An extension is associated with its _WKInspectorExtensionHost.
A WebKit client that supports Web Extensions can use this class to implement 'devtools' extension APIs.
Add _WKInspectorExtensionHost methods to register and unregister an extension.
Add plumbing so that the extension in registered in the frontend with WI.InspectorExtensionController.
Later patches will flesh out the needed functionality to implement rest of the 'devtools' API.
* Sources.txt:
* SourcesCocoa.txt:
* DerivedSources-input.xcfilelist:
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* WebKit.xcodeproj/project.pbxproj:
Add new files.
* Scripts/webkit/messages.py: If a message includes InspectorExtensionError,
then generate an include for InspectorExtensionTypes.h where it is declared.
* Shared/API/APIObject.h:
* Shared/Cocoa/APIObject.mm:
(API::Object::newObject):
New API object type.
* Shared/InspectorExtensionTypes.h: Added error code enum and useful type aliases.
* Shared/InspectorExtensionTypes.cpp:
(WebKit::inspectorExtensionErrorToString):
* UIProcess/API/APIInspectorExtension.h:
* UIProcess/API/APIInspectorExtension.cpp:
(API::InspectorExtension::create):
(API::InspectorExtension::InspectorExtension):
(API::InspectorExtension::~InspectorExtension):
Added. This is a placeholder object that's used for the wrapper system,
since we don't have single C++ objects for each extension to back the API object.
* UIProcess/API/Cocoa/_WKInspector.h:
* UIProcess/API/Cocoa/_WKInspector.mm:
(-[_WKInspector registerExtensionWithID:displayName:completionHandler:]):
(-[_WKInspector unregisterExtension:completionHandler:]):
Implement new _WKInspectorExtensionHost methods. Send IPC to the Inspector web process.
* UIProcess/API/Cocoa/_WKInspectorExtension.h:
* UIProcess/API/Cocoa/_WKInspectorExtension.mm:
(-[_WKInspectorExtension initWithIdentifier:]):
(-[_WKInspectorExtension dealloc]):
(-[_WKInspectorExtension _apiObject]):
(-[_WKInspectorExtension extensionID]):
Added. This is a data object that's also used as a token to unregister the extension.
* UIProcess/API/Cocoa/_WKInspectorExtensionHost.h: Add new SPI.
* UIProcess/API/Cocoa/_WKInspectorExtensionInternal.h:
* UIProcess/API/Cocoa/_WKRemoteWebInspectorViewController.h:
* UIProcess/API/Cocoa/_WKRemoteWebInspectorViewController.mm:
(-[_WKRemoteWebInspectorViewController registerExtensionWithID:displayName:completionHandler:]):
(-[_WKRemoteWebInspectorViewController unregisterExtension:completionHandler:]):
Implement new _WKInspectorExtensionHost methods. Send IPC to the Inspector web process.
* UIProcess/Inspector/RemoteWebInspectorProxy.h:
* UIProcess/Inspector/RemoteWebInspectorProxy.messages.in: Add FrontendLoaded.
(WebKit::RemoteWebInspectorProxy::extensionController const): Added.
* UIProcess/Inspector/RemoteWebInspectorProxy.cpp:
(WebKit::RemoteWebInspectorProxy::frontendLoaded):
(WebKit::RemoteWebInspectorProxy::createFrontendPageAndWindow):
Set up the extension controller object after the frontend has been created.
Hook up the FrontendLoaded event sent from Inspector web process so that the
extension controller can be notified of when it's safe to message the frontend.
* UIProcess/Inspector/WebInspectorProxy.cpp:
(WebKit::WebInspectorProxy::createFrontendPage):
(WebKit::WebInspectorProxy::closeFrontendPageAndWindow):
(WebKit::WebInspectorProxy::frontendLoaded):
Set up the extension controller object after the frontend has been created.
* UIProcess/Inspector/WebInspectorProxy.h:
(WebKit::WebInspectorProxy::extensionController const): Added.
* UIProcess/Inspector/WebInspectorUIExtensionControllerProxy.h:
* UIProcess/Inspector/WebInspectorUIExtensionControllerProxy.cpp: Added.
(WebKit::WebInspectorUIExtensionControllerProxy::WebInspectorUIExtensionControllerProxy):
(WebKit::WebInspectorUIExtensionControllerProxy::~WebInspectorUIExtensionControllerProxy):
(WebKit::WebInspectorUIExtensionControllerProxy::whenFrontendHasLoaded):
(WebKit::WebInspectorUIExtensionControllerProxy::inspectorFrontendLoaded):
(WebKit::WebInspectorUIExtensionControllerProxy::registerExtension):
(WebKit::WebInspectorUIExtensionControllerProxy::unregisterExtension):
Added. Forward API requests to the Web Inspector's web process to be evaluated in the frontend.
* WebProcess/Inspector/WebInspectorFrontendAPIDispatcher.h:
* WebProcess/Inspector/WebInspectorFrontendAPIDispatcher.cpp:
(WebKit::WebInspectorFrontendAPIDispatcher::dispatchCommand):
Add a more generic way to encode values to be passed as arguments to InspectorFrontendAPI.dispatch.
* WebProcess/Inspector/WebInspectorUI.cpp:
(WebKit::WebInspectorUI::establishConnection):
(WebKit::WebInspectorUI::closeWindow):
* WebProcess/Inspector/WebInspectorUI.h:
(WebKit::WebInspectorUI::frontendPage const): Added.
(WebKit::WebInspectorUI::frontendAPIDispatcher): Added.
* WebProcess/Inspector/WebInspectorUIExtensionController.h:
* WebProcess/Inspector/WebInspectorUIExtensionController.messages.in: Added.
* WebProcess/Inspector/WebInspectorUIExtensionController.cpp: Added.
(WebKit::WebInspectorUIExtensionController::WebInspectorUIExtensionController):
(WebKit::WebInspectorUIExtensionController::~WebInspectorUIExtensionController):
(WebKit::WebInspectorUIExtensionController::registerExtension):
(WebKit::WebInspectorUIExtensionController::unregisterExtension):
(WebKit::WebInspectorUIExtensionController::parseInspectorExtensionErrorFromResult):
Handle incoming messages by forwarding the associated command to WI.WebExtensionController
via InspectorFrontendAPI evaluations.
* UIProcess/API/Cocoa/_WKUserStyleSheet.h:
* UIProcess/API/Cocoa/_WKUserStyleSheet.mm:
Fallout from unified build chunking shifts.
* UIProcess/WebAuthentication/Cocoa/LocalAuthenticator.mm:
(WebKit::LocalAuthenticatorInternal::getExistingCredentials):
(WebKit::LocalAuthenticator::continueMakeCredentialAfterUserVerification):
Fallout from unified build chunking shifts, but more.
2020-11-05 Per Arne Vollan <pvollan@apple.com>
[macOS] Update message filters in the WebContent sandbox
https://bugs.webkit.org/show_bug.cgi?id=218520
Reviewed by Brent Fulgham.
Based on telemetry, update the message filters with allowed messages in the WebContent sandbox.
No new tests, covered by existing tests.
* WebProcess/com.apple.WebProcess.sb.in:
2020-11-05 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed, fix the build with recent SDKs.
* UIProcess/Launcher/mac/ProcessLauncherMac.mm:
(WebKit::ProcessLauncher::platformInvalidate):
2020-11-05 Don Olmstead <don.olmstead@sony.com>
Non-unified build fixes, early November 2020 edition
https://bugs.webkit.org/show_bug.cgi?id=218628
Unreviewed non-unified build fix.
* UIProcess/SpeechRecognitionServer.cpp:
(WebKit::SpeechRecognitionServer::start):
(WebKit::SpeechRecognitionServer::removePendingRequest):
* WebProcess/WebCoreSupport/WebSpeechRecognitionConnection.cpp:
2020-11-05 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed, fix the build with recent SDKs.
* Platform/IPC/cocoa/ConnectionCocoa.mm:
(IPC::Connection::kill):
2020-11-05 Brent Fulgham <bfulgham@apple.com>
[iOS] Remove redundant rule using never-granted extension
https://bugs.webkit.org/show_bug.cgi?id=218596
<rdar://problem/66581246>
Reviewed by Per Arne Vollan.
WebKit has long had a sandbox rule granting the ability to issue iokit-open
operations if a particular security extension was granted to the process. The
WebKit processes are known to never grant this extension, so this rule is not
needed.
We should remove the rule to reduce sandbox complexity and increase the speed
of sandbox compilation.
* Resources/SandboxProfiles/ios/com.apple.WebKit.GPU.sb:
* Resources/SandboxProfiles/ios/com.apple.WebKit.Networking.sb:
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-11-05 Brent Fulgham <bfulgham@apple.com>
[iOS] Remove redundant rule for shared-preference.read-write
https://bugs.webkit.org/show_bug.cgi?id=218594
<rdar://problem/66581796>
Reviewed by Per Arne Vollan.
We have long had a rule to allow shared preferences to be read and written if a
particular sandbox extension was granted to the process. However, we do not issue
this extension to any WebKit process, so the rule is unnecessary.
We should remove it to reduce confusion, and make the sandbox smaller (and quicker
to compile).
* Resources/SandboxProfiles/ios/com.apple.WebKit.GPU.sb:
* Resources/SandboxProfiles/ios/com.apple.WebKit.Networking.sb:
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-11-03 Darin Adler <darin@apple.com>
Remove more functions that implicitly use composed tree
https://bugs.webkit.org/show_bug.cgi?id=218544
Reviewed by Ryosuke Niwa.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::constrainRangeToSelection): Factored this function out of the
requestDocumentEditingContext and also rewrote it to use higher level operations.
(WebKit::WebPage::requestDocumentEditingContext): Refactored to use the new
constrainRangeToSelection.
2020-11-05 Alex Christensen <achristensen@webkit.org>
Store WeakPtr<Frame> instead of Frame*
https://bugs.webkit.org/show_bug.cgi?id=218599
Reviewed by Youenn Fablet.
* WebProcess/WebPage/WebFrame.cpp:
(WebKit::WebFrame::initWithCoreMainFrame):
(WebKit::WebFrame::createSubframe):
(WebKit::WebFrame::coreFrame const):
(WebKit::WebFrame::info const):
(WebKit::WebFrame::handlesPageScaleGesture const):
(WebKit::WebFrame::requiresUnifiedScaleFactor const):
* WebProcess/WebPage/WebFrame.h:
(WebKit::WebFrame::coreFrame const): Deleted.
2020-11-05 Alex Christensen <achristensen@webkit.org>
Use fewer raw pointers and more const correctness in Frame.h
https://bugs.webkit.org/show_bug.cgi?id=218598
Reviewed by Youenn Fablet.
* WebProcess/WebCoreSupport/WebFrameLoaderClient.h:
(WebKit::toWebFrameLoaderClient):
2020-11-05 Carlos Garcia Campos <cgarcia@igalia.com>
WebDriver: handle surrogate pairs in keyboard actions
https://bugs.webkit.org/show_bug.cgi?id=218279
Reviewed by Brian Burg.
And fail with invalid argument if the string can be represented as a single code point.
Fixes: imported/w3c/webdriver/tests/perform_actions/key_special_keys.py::test_codepoint_keys_behave_correctly[\U0001f604]
imported/w3c/webdriver/tests/perform_actions/key_special_keys.py::test_codepoint_keys_behave_correctly[\U0001f60d]
imported/w3c/webdriver/tests/perform_actions/key_special_keys.py::test_invalid_multiple_codepoint_keys_fail[fa]
imported/w3c/webdriver/tests/perform_actions/key_special_keys.py::test_invalid_multiple_codepoint_keys_fail[\u0ba8\u0bbfb]
imported/w3c/webdriver/tests/perform_actions/key_special_keys.py::test_invalid_multiple_codepoint_keys_fail[\u0ba8\u0bbf\u0ba8]
imported/w3c/webdriver/tests/perform_actions/key_special_keys.py::test_invalid_multiple_codepoint_keys_fail[\u1100\u1161\u11a8c]
* UIProcess/Automation/WebAutomationSession.cpp:
(WebKit::pressedCharKey):
(WebKit::WebAutomationSession::performInteractionSequence):
2020-11-05 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Determine an AAGUID for the platform authenticators
https://bugs.webkit.org/show_bug.cgi?id=217945
<rdar://problem/70811618>
Reviewed by Brent Fulgham.
Relying parties use the AAGUID to recognize supported authenticators. Using a NULL AAGUID blocks them from recognizing Apple products as valid WebAuthentication targets.
We need to assign ourselves a GUID representing Apple authenticators, then publish with our attestation certificate and with the FIDO Alliance.
Covered by existing tests.
* UIProcess/WebAuthentication/Cocoa/LocalAuthenticator.mm:
(WebKit::LocalAuthenticatorInternal::aaguidVector):
(WebKit::LocalAuthenticator::continueMakeCredentialAfterUserVerification):
2020-11-05 Carlos Garcia Campos <cgarcia@igalia.com>
WebDriver: session can have more than one active input source of the same type
https://bugs.webkit.org/show_bug.cgi?id=218382
Reviewed by Brian Burg.
We are failing in that case, and the dispatcher assumes it's not possible, but the spec doesn't say there's such
limitation and there's a WPT infrastructure test checking that (infrastructure/testdriver/actions/eventOrder.html).
* UIProcess/Automation/SimulatedInputDispatcher.cpp:
(WebKit::SimulatedInputKeyFrame::keyFrameFromStateOfInputSources): Input sources are now stored in a HashMap.
(WebKit::SimulatedInputKeyFrame::keyFrameToResetInputSources): Ditto.
(WebKit::SimulatedInputDispatcher::run): We don't need to copy the given input sources, since they are owned by
the session.
(WebKit::SimulatedInputDispatcher::finishDispatching): Ditto.
* UIProcess/Automation/SimulatedInputDispatcher.h:
* UIProcess/Automation/WebAutomationSession.cpp:
(WebKit::WebAutomationSession::WebAutomationSession): Do not create initial input sources for each type.
(WebKit::WebAutomationSession::performInteractionSequence): Check there aren't multiple input sources with the
same ID, but don't fail if there are multiple input sources with the same type.
(WebKit::WebAutomationSession::cancelInteractionSequence): Clear the active input sources.
(WebKit::WebAutomationSession::inputSourceForType): Removed.
* UIProcess/Automation/WebAutomationSession.h:
2020-11-04 Said Abou-Hallawa <said@apple.com>
[GPU Process] Move the internal GPU rendering flags from WebPage to WebProcess
https://bugs.webkit.org/show_bug.cgi?id=218549
Reviewed by Tim Horton.
Move UseGPUProcessForCanvasRendering from WebPage to WebProcess. Add a
new setting for UseGPUProcessForDOMRendering to WebProcess.
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
Encode/decode the flags from the UI Process to the Web Process.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
Read the flags in from the preferences.
* WebProcess/GPU/media/RemoteMediaPlayerManager.cpp:
(WebKit::RemoteMediaPlayerManager::setUseGPUProcess):
(WebKit::RemoteMediaPlayerManager::updatePreferences): Deleted.
* WebProcess/GPU/media/RemoteMediaPlayerManager.h:
These flags belong to WebKit so they should not be accessed through the
Settings.
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::createImageBuffer const):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences):
(WebKit::WebPage::shouldUseRemoteRenderingFor): Deleted.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::setUseGPUProcessForCanvasRendering):
(WebKit::WebProcess::setUseGPUProcessForDOMRendering):
(WebKit::WebProcess::shouldUseRemoteRenderingFor):
* WebProcess/WebProcess.h:
The rendering and media flags can now be accessed from the WebPorcess.
2020-11-04 Aditya Keerthi <akeerthi@apple.com>
[Contact Picker API] Add support for picker UI on iOS
https://bugs.webkit.org/show_bug.cgi?id=218189
<rdar://problem/69862277>
Reviewed by Devin Rousso.
ContactsManager.select() should present a contact picker on platforms
which support one. This patch enables the end-to-end functionality on
iOS, presenting a contact picker upon a call to the API and returning
the selected contacts upon dismissal.
Tests: contact-picker/contacts-select-after-dismissing-picker.html
contact-picker/contacts-select-while-presenting-picker.html
contact-picker/contacts-select.html
* Platform/spi/Cocoa/ContactsUISPI.h: Added.
* SourcesCocoa.txt:
* UIProcess/API/Cocoa/WKWebViewPrivateForTesting.h:
* UIProcess/API/Cocoa/WKWebViewTesting.mm:
(-[WKWebView _didPresentContactPicker]):
(-[WKWebView _didDismissContactPicker]):
(-[WKWebView _dismissContactPickerWithContacts:]):
* UIProcess/Cocoa/WKContactPicker.h: Added.
* UIProcess/Cocoa/WKContactPicker.mm: Added.
WKContactPicker is a WebKit wrapper around CNContactPickerViewController.
(-[WKCNContactPickerDelegate initWithContactPickerDelegate:]):
WKCNContactPickerDelegate is wrapper around CNContactPickerDelegate.
This is necessary as single/multiple selection in a
CNContactPickerViewController is determined by which delegate methods
are implemented. The two specializations of this class include the
methods necessary to present a single-select and multi-select picker
respectively.
(-[WKCNContactPickerDelegate contactPickerDidCancel:]):
(-[WKCNContactPickerSingleSelectDelegate contactPicker:didSelectContact:]):
(-[WKCNContactPickerMultiSelectDelegate contactPicker:didSelectContacts:]):
(-[WKContactPicker delegate]):
(-[WKContactPicker setDelegate:]):
(-[WKContactPicker initWithView:]):
(-[WKContactPicker presentWithRequestData:completionHandler:]):
(-[WKContactPicker contactPickerDidCancel:]):
This delegate method is called when the picker is dismissed by
tapping the done button or when the picker is dismissed by swiping
down.
(-[WKContactPicker contactPicker:didSelectContact:]):
(-[WKContactPicker contactPicker:didSelectContacts:]):
(-[WKContactPicker _contactPickerDidDismissWithContactInfo:]):
(-[WKContactPicker _contactInfoFromCNContact:]):
(-[WKContactPicker dismissWithContacts:]):
(-[WKContactPicker _contactsFromJSContacts:]):
* UIProcess/ios/PageClientImplIOS.h:
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::showContactPicker):
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _showContactPicker:completionHandler:]):
(-[WKContentView contactPickerDidPresent:]):
(-[WKContentView contactPickerDidDismiss:]):
(-[WKContentView _dismissContactPickerWithContacts:]):
* WebKit.xcodeproj/project.pbxproj:
2020-11-04 Chris Dumez <cdumez@apple.com>
Unreviewed, fix build with very recent SDK by silencing a deprecation warning.
* Platform/IPC/cocoa/ConnectionCocoa.mm:
(IPC::ConnectionTerminationWatchdog::watchdogTimerFired):
2020-11-04 Chris Dumez <cdumez@apple.com>
REGRESSION(r268161?): [ macOS ] imported/w3c/web-platform-tests/webaudio/the-audio-api/the-audiocontext-interface/processing-after-resume.https.html is a flaky failure
https://bugs.webkit.org/show_bug.cgi?id=217542
<rdar://problem/70159546>
Reviewed by Youenn Fablet.
When the script calls suspend(), we call AudioDestinationCocoa::stop() to stop rendering and then resolve the suspend promise.
At this point, the script saves the AudioContext's current sample frame and expects this value to not change until the
AudioContext is resumed. The issue was that we previously were potentially resolving the promise on the main thread while
the last rendering quantum was still being processed on the AudioWorklet thread. This meant that the sample frame could
still increase a bit, a short while after the suspend promise has been resolved. To address the issue, we now to a round
trip to the AudioWorklet thread after stopping rendering, to make sure we only resolve the promise after the last rendering
quantum has been processed on the AudioWorklet thread.
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::stop):
2020-11-04 Said Abou-Hallawa <said@apple.com>
[GPU Process] Enable Document based ImageBitmap and OffscreenCanvas drawing to use GPU Process rendering
https://bugs.webkit.org/show_bug.cgi?id=217735
Reviewed by Simon Fraser.
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
putImageData() might be called immediately after creating the ImageBuffer
and before receiving the RemoteRenderingBackend.CreateImageBufferBackend
message. We need to handle for this case and avoid asserting.
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::createImageBuffer):
If the size of the ImageBuffer is very big, no remote ImageBuffer can be
created. So we have to be bail out quietly in this case.
2020-11-04 David Kilzer <ddkilzer@apple.com>
WebKit should remove unused debug variant support
<https://webkit.org/b/218315>
<rdar://problem/70785369>
Reviewed by Darin Adler.
Remove support for building the debug variant since it is
currently unused. We now set default values for the
DEAD_CODE_STRIPPING, DEBUG_DEFINES, GCC_OPTIMIZATION_LEVEL and
STRIP_INSTALLED_PRODUCT variables.
Also move these values out of the Xcode project into
Base.xcconfig files using the [config=Debug] specifier so that
these overrides are next to the definitions.
Additional changes in behavior are noted below.
* Configurations/Base.xcconfig:
* WebKit.xcodeproj/project.pbxproj:
- STRIP_INSTALLED_PRODUCT is now set to NO for Debug builds.
Previously symbols were always stripped in Debug builds.
2020-11-04 Wenson Hsieh <wenson_hsieh@apple.com>
Add some missing header includes in media sources
https://bugs.webkit.org/show_bug.cgi?id=218568
Reviewed by Eric Carlson.
* GPUProcess/media/ios/RemoteMediaSessionHelperProxy.cpp: Include `WebCoreArgumentCoders.h`, since we need to
know how to encode a `MediaPlaybackTargetContext`.
* WebProcess/GPU/webrtc/MediaRecorderPrivate.h: Add a missing `WebCore::` namespace.
2020-11-04 Luming Yin <luming_yin@apple.com>
[watchOS] Don't allow managed preference to enable legacy sync XHRs on watchOS
https://bugs.webkit.org/show_bug.cgi?id=218548
<rdar://problem/68417252>
Reviewed by Geoffrey Garen.
The managed preference to allow legacy sync XHRs during page dismissal is not applicable to watchOS, but we
still attempt to dlopen ManagedConfiguration to get the managed configuration value. This redundant check
blocks the main thread and leads to hangs.
We now compile out this code for the watchOS platforms.
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultDisallowSyncXHRDuringPageDismissalEnabled):
* Shared/WebPreferencesDefaultValues.h:
* Shared/ios/WebPreferencesDefaultValuesIOS.mm:
2020-11-04 Dean Jackson <dino@apple.com>
WKContentView might get KVO notifications it doesn't expect
https://bugs.webkit.org/show_bug.cgi?id=218558
<rdar://problem/71031491>
Reviewed by Tim Horton.
It's possible that code elsewhere in the system adds KVO
observation to the CALayer of a WKContentView, which is
a problem because our implementation of observeValue only
expects calls that it registered for. Fix this by passing
a context to the registration and passing the observation
call onto the superclass if it isn't the expected context.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView setUpInteraction]):
(-[WKContentView observeValueForKeyPath:ofObject:change:context:]):
2020-11-04 Youenn Fablet <youenn@apple.com>
Make functional WebRTC encoders in GPUProcess
https://bugs.webkit.org/show_bug.cgi?id=218498
Reviewed by Eric Carlson.
Update code according libwebrtc changes.
Add missing support for timestamp.
Also, in case of samples that are non IO Surface and not BGRA, do a conversion to BGRA and then create an IO surface from it.
This allows supporting sending black frames generated in case of muted tracks.
Create a remote factory in LibWebRTCProvider based on whether using gpu process or not.
* GPUProcess/webrtc/LibWebRTCCodecsProxy.h:
* GPUProcess/webrtc/LibWebRTCCodecsProxy.messages.in:
* GPUProcess/webrtc/LibWebRTCCodecsProxy.mm:
(WebKit::LibWebRTCCodecsProxy::encodeFrame):
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/webrtc/LibWebRTCCodecs.cpp:
(WebKit::createVideoDecoder):
(WebKit::encodeVideoFrame):
(WebKit::toWebRTCCodecType):
(WebKit::LibWebRTCCodecs::createEncoder):
(WebKit::LibWebRTCCodecs::encodeFrame):
(WebKit::LibWebRTCCodecs::completedEncoding):
* WebProcess/GPU/webrtc/LibWebRTCCodecs.h:
* WebProcess/GPU/webrtc/LibWebRTCCodecs.mm: Added.
(WebKit::LibWebRTCCodecs::convertToBGRA):
* WebProcess/Network/webrtc/LibWebRTCProvider.cpp:
(WebKit::LibWebRTCProvider::createEncoderFactory):
* WebProcess/Network/webrtc/LibWebRTCProvider.h:
2020-11-03 Sihui Liu <sihui_liu@apple.com>
Set up basic infrastructure for SpeechRecognition
https://bugs.webkit.org/show_bug.cgi?id=218216
<rdar://problem/70703788>
Reviewed by Youenn Fablet.
* CMakeLists.txt:
* DerivedSources-input.xcfilelist:
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* Scripts/webkit/messages.py:
* Sources.txt:
* UIProcess/SpeechRecognitionServer.cpp: Added.
(WebKit::SpeechRecognitionServer::SpeechRecognitionServer):
(WebKit::SpeechRecognitionServer::start):
(WebKit::SpeechRecognitionServer::processNextPendingRequestIfNeeded):
(WebKit::SpeechRecognitionServer::stop):
(WebKit::SpeechRecognitionServer::abort):
(WebKit::SpeechRecognitionServer::removePendingRequest):
(WebKit::SpeechRecognitionServer::invalidate):
(WebKit::SpeechRecognitionServer::startPocessingRequest):
(WebKit::SpeechRecognitionServer::stopProcessingRequest):
(WebKit::SpeechRecognitionServer::messageSenderConnection const):
(WebKit::SpeechRecognitionServer::messageSenderDestinationID const):
* UIProcess/SpeechRecognitionServer.h: Added.
* UIProcess/SpeechRecognitionServer.messages.in: Added.
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::createSpeechRecognitionServer):
(WebKit::WebProcessProxy::destroySpeechRecognitionServer):
* UIProcess/WebProcessProxy.h:
* UIProcess/WebProcessProxy.messages.in:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebCoreSupport/WebSpeechRecognitionConnection.cpp: Added.
(WebKit::WebSpeechRecognitionConnection::create):
(WebKit::WebSpeechRecognitionConnection::WebSpeechRecognitionConnection):
(WebKit::WebSpeechRecognitionConnection::~WebSpeechRecognitionConnection):
(WebKit::WebSpeechRecognitionConnection::registerClient):
(WebKit::WebSpeechRecognitionConnection::start):
(WebKit::WebSpeechRecognitionConnection::stop):
(WebKit::WebSpeechRecognitionConnection::abort):
(WebKit::WebSpeechRecognitionConnection::invalidate):
(WebKit::WebSpeechRecognitionConnection::didReceiveUpdate):
(WebKit::WebSpeechRecognitionConnection::messageSenderConnection const):
(WebKit::WebSpeechRecognitionConnection::messageSenderDestinationID const):
* WebProcess/WebCoreSupport/WebSpeechRecognitionConnection.h: Added.
* WebProcess/WebCoreSupport/WebSpeechRecognitionConnection.messages.in: Added.
* WebProcess/WebCoreSupport/WebSpeechRecognitionProvider.h: Added.
* WebProcess/WebPage/IPCTestingAPI.cpp:
(WebKit::IPCTestingAPI::encodeArgument):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
2020-11-03 Brent Fulgham <bfulgham@apple.com>
[macOS] remove sysctl.oidfmt. access
https://bugs.webkit.org/show_bug.cgi?id=218524
<rdar://problem/70353722>
Reviewed by Per Arne Vollan.
Access to "sysctl.oidfmt." was needed due to a bug <rdar://problem/64208532>.
Since that issue has been fixed, we can remove the access.
* NetworkProcess/mac/com.apple.WebKit.NetworkProcess.sb.in:
* WebProcess/com.apple.WebProcess.sb.in:
2020-11-03 Rini Patel <rini_patel@apple.com>
[GPU Process] Flush canvas displayList from doAfterUpdateRendering
https://bugs.webkit.org/show_bug.cgi?id=218401
Reviewed by Simon Fraser.
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
(WebKit::RemoteImageBufferProxy::flushDrawingContextAndCommit): Deleted.
2020-11-03 Wenson Hsieh <wenson_hsieh@apple.com>
Replace DisplayList::Recorder::Delegate::(will|did)AppendItem with (will|did)AppendItemOfType
https://bugs.webkit.org/show_bug.cgi?id=218518
Reviewed by Simon Fraser.
See WebCore ChangeLog for more information.
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
2020-11-03 Wenson Hsieh <wenson_hsieh@apple.com>
Add helper methods to encode and decode IPC arguments as raw data
https://bugs.webkit.org/show_bug.cgi?id=218516
Reviewed by Geoff Garen.
Add new helper methods to `IPC::Encoder` and `IPC::Decoder` that can be used to convert anything that can be
sent as an IPC argument into raw bytes (i.e. using `WebCore::SharedBuffer`), as long as there are no IPC
attachments in the encoded data.
For more details, see <webkit.org/b/218406>.
* Platform/IPC/Decoder.cpp:
(IPC::Decoder::Decoder):
Add private versions of the Encoder and Decoder constructors that avoid IPC header data. These constructors are
called only from within `encodeSingleObject` and `decodeSingleObject`, respectively.
(IPC::m_bufferDeallocator):
* Platform/IPC/Decoder.h:
(IPC::Decoder::decodeSingleObject):
* Platform/IPC/Encoder.cpp:
(IPC::Encoder::Encoder):
(IPC::Encoder::releaseAttachments):
(IPC::Encoder::hasAttachments const):
* Platform/IPC/Encoder.h:
2020-11-03 Stephan Szabo <stephan.szabo@sony.com>
[WinCairo/PlayStation] ICU 68.1 no longer exposes FALSE and TRUE macros by default
https://bugs.webkit.org/show_bug.cgi?id=218522
Reviewed by Don Olmstead.
* UIProcess/CoordinatedGraphics/DrawingAreaProxyCoordinatedGraphics.cpp: Replace FALSE with false
2020-11-03 Said Abou-Hallawa <said@apple.com>
[GPU Process] Make ImageBuffer RefCounted
https://bugs.webkit.org/show_bug.cgi?id=218472
Reviewed by Simon Fraser.
Currently after a remote ImageBuffer is deleted, its identifier is locked
in the RemoteResourceCacheProxy till no remote client is referencing it.
Then an IPC message is sent to GPU Process to release it.
The plan is to decode the DisplayList::DrawImageBuffer by replacing the
ImageBuffer identifier with a RefPtr to the ImageBuffer itself. Once the
ImageBuffer is removed from the RemoteResourceCache, it is going to be
deleted after the last DrawImageBuffer is replayed back. So no locking
is needed to guarantee the existence of the ImageBuffer when it is drawn.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::createImageBuffer):
(WebKit::RemoteRenderingBackend::flushDisplayList):
(WebKit::RemoteRenderingBackend::flushDisplayListAndCommit):
(WebKit::RemoteRenderingBackend::getImageData):
* GPUProcess/graphics/RemoteResourceCache.cpp:
(WebKit::RemoteResourceCache::cacheImageBuffer):
* GPUProcess/graphics/RemoteResourceCache.h:
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
(WebKit::RemoteImageBufferProxy::create):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::createImageBuffer):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::createImageBuffer const):
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebPage/WebFrame.cpp:
(WebKit::WebFrame::createSelectionSnapshot const):
2020-11-03 Wenson Hsieh <wenson_hsieh@apple.com>
Replace DisplayList::itemCount with DisplayList::isEmpty
https://bugs.webkit.org/show_bug.cgi?id=218517
Reviewed by Geoffrey Garen.
Separately keep track of the item count in the context's display list. See WebCore/ChangeLog for more details.
* GPUProcess/graphics/RemoteImageBuffer.h: Use `isEmpty()` instead of checking `itemCount()`.
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
(WebKit::RemoteImageBufferProxy::flushDrawingContextAndCommit):
2020-11-02 Simon Fraser <simon.fraser@apple.com>
Scroll position can get reset after programmatic scroll
https://bugs.webkit.org/show_bug.cgi?id=218477
Reviewed by Antti Koivisto.
* WebProcess/WebPage/RemoteLayerTree/RemoteScrollingCoordinator.mm:
(WebKit::RemoteScrollingCoordinator::scrollPositionChangedForNode):
2020-11-03 Brent Fulgham <bfulgham@apple.com>
[macOS] Adopt additional QuartzCore entitlement to reduce accessible endpoints
https://bugs.webkit.org/show_bug.cgi?id=218466
<rdar://problem/55790037>
Reviewed by Per Arne Vollan.
In iOS 14 we adopted a new entitlement that limited the set of XPC endpoints available
to QuartzCore clients. We should do the same on macOS to provide a similar level of
protection.
* Scripts/process-entitlements.sh:
2020-11-03 Youenn Fablet <youenn@apple.com>
Add support for WebRTC VP9 decoder in GPU process
https://bugs.webkit.org/show_bug.cgi?id=218445
Reviewed by Eric Carlson.
Add support for VP9 decoder, send key frame size as IPC message specifically for VP9 hardware decoder.
If GPU process does not enable either SW or HW VP9 decoder, we fall back to in process VP9 decoder.
Manually tested by enabling WebRTC codec in GPU process.
Will be covered by existing VP9 tests in GPU process mode once canvas in GPUProcess is fully supported.
* GPUProcess/webrtc/LibWebRTCCodecsProxy.h:
* GPUProcess/webrtc/LibWebRTCCodecsProxy.messages.in:
* GPUProcess/webrtc/LibWebRTCCodecsProxy.mm:
(WebKit::LibWebRTCCodecsProxy::createVP9Decoder):
(WebKit::LibWebRTCCodecsProxy::setFrameSize):
* WebProcess/GPU/GPUProcessConnection.h:
(WebKit::GPUProcessConnection::isVP9DecoderEnabled const):
(WebKit::GPUProcessConnection::isVPSWDecoderEnabled const):
* WebProcess/GPU/webrtc/LibWebRTCCodecs.cpp:
(WebKit::createVideoDecoder):
(WebKit::decodeVideoFrame):
(WebKit::LibWebRTCCodecs::setCallbacks):
(WebKit::LibWebRTCCodecs::createDecoder):
(WebKit::LibWebRTCCodecs::decodeFrame):
(WebKit::formatNameFromCodecType):
* WebProcess/GPU/webrtc/LibWebRTCCodecs.h:
(WebKit::LibWebRTCCodecs::setVP9VTBSupport):
(WebKit::LibWebRTCCodecs::supportVP9VTB const):
2020-11-03 Youenn Fablet <youenn@apple.com>
Allow low latency H264 encoder in GPUProcess
https://bugs.webkit.org/show_bug.cgi?id=218442
Reviewed by Eric Carlson.
Pass useLowLatency boolean from WebProcess to GPUProcess.
* GPUProcess/webrtc/LibWebRTCCodecsProxy.h:
* GPUProcess/webrtc/LibWebRTCCodecsProxy.messages.in:
* GPUProcess/webrtc/LibWebRTCCodecsProxy.mm:
(WebKit::LibWebRTCCodecsProxy::createEncoder):
* WebProcess/GPU/webrtc/LibWebRTCCodecs.cpp:
(WebKit::LibWebRTCCodecs::completedDecoding):
(WebKit::LibWebRTCCodecs::createEncoder):
2020-11-03 Chris Lord <clord@igalia.com>
[GTK] Zooming causes page to scroll to top
https://bugs.webkit.org/show_bug.cgi?id=217955
Reviewed by Carlos Garcia Campos.
Don't reset page scale when setting zoom level on GLIB backend.
* UIProcess/API/glib/WebKitWebView.cpp:
(webkit_web_view_set_zoom_level):
2020-11-03 Adrian Perez de Castro <aperez@igalia.com>
[GTK][WPE] Cannot remove individual scripts/stylesheets using WebKitUserContentManager
https://bugs.webkit.org/show_bug.cgi?id=218403
Reviewed by Carlos Garcia Campos.
Add the missing WebKitUserContentManager functions which allow
removing a single script and style sheet from the manager. Given
that it is possible to remove a single content filter and message
handlers already, this make the API orthogonal for scripts and style
sheets as well.
* UIProcess/API/glib/WebKitUserContentManager.cpp:
(webkit_user_content_manager_remove_style_sheet): Added.
(webkit_user_content_manager_remove_script): Added.
* UIProcess/API/gtk/WebKitUserContentManager.h: Add declarations for the new functions.
* UIProcess/API/gtk/docs/webkit2gtk-4.0-sections.txt: Ditto.
* UIProcess/API/wpe/WebKitUserContentManager.h: Ditto.
* UIProcess/API/wpe/docs/wpe-1.0-sections.txt: Ditto.
2020-11-02 Ryosuke Niwa <rniwa@webkit.org>
Remove some useless comments from .messages.in files
https://bugs.webkit.org/show_bug.cgi?id=218473
Reviewed by Tim Horton.
Remove comments that simply repeats what the code says.
* GPUProcess/GPUProcess.messages.in:
* NetworkProcess/NetworkProcess.messages.in:
* WebProcess/WebProcess.messages.in:
2020-11-02 Devin Rousso <drousso@apple.com>
guard UIScribbleInteraction class property observing behind a LOA check
https://bugs.webkit.org/show_bug.cgi?id=218463
<rdar://problem/70747966>
Reviewed by Tim Horton.
Some apps appear to swizzle `-[NSObject addObserver:forKeyPath:options:context:]` without
support for the fact that the object can be a class rather than an instance and therefore
crash. Use a LOA check to guard observing `+[UIScribbleInteraction isPencilInputExpected]`
so that this doesn't happen, but only until the apps update, at which point they can fix it.
* UIProcess/ios/WKStylusDeviceObserver.mm:
(-[WKStylusDeviceObserver start]):
(-[WKStylusDeviceObserver stop]):
2020-11-02 James Savage <james.savage@apple.com>
Changes to SWIFT_MODULE_ONLY_ARCHS
https://bugs.webkit.org/show_bug.cgi?id=218404
Reviewed by Tim Horton.
* SwiftOverlay/Configurations/WebKitSwiftOverlay.xcconfig: Simplify how
we specify these options.
2020-11-02 Aditya Keerthi <akeerthi@apple.com>
[iOS] Link Contacts.framework during build
https://bugs.webkit.org/show_bug.cgi?id=218289
<rdar://problem/70770410>
Reviewed by Andy Estes.
Contacts.framework is currrently soft linked in PaymentAuthorizationPresenter.
Soft linking is generally discouraged, since it requires running the linker
at runtime, leading to an unwanted performance cost. By normally linking
the framework, the linkage is declared at build time, which enables running
the linker at build/update/install time and avoiding the runtime cost.
Unfortunately, we still have to soft link the framework on macOS, to avoid
introducing a dependency cycle. Upward linking is not an option, since the
Contacts framework does not implement InstallAPI.
The SOFT_LINK_FRAMEWORK macro has been kept in PaymentAuthorizationPresenter
to avoid introducing additional #ifs. This should be fine to keep on iOS,
since the framework will already loaded in the address space, and the call
to dlopen() should be much less expensive.
* Configurations/WebKit.xcconfig:
* Platform/cocoa/PaymentAuthorizationPresenter.mm:
2020-11-02 Chris Dumez <cdumez@apple.com>
Crash under ProcessThrottler::setAssertionType()
https://bugs.webkit.org/show_bug.cgi?id=218448
<rdar://problem/67419221>
Reviewed by Geoffrey Garen.
A ProcessThrottler object is owned by its associated AuxiliaryProcessProxy. The crash was happening
in ProcessThrottler::setAssertionType(), we would replace m_assertion with a new "Suspended"
assertion and then crash on the next line when using m_assertion. The reason we crash is that
when we replaced m_assertion with a new assertion, the destruction of the previous assertion
caused the UIProcess's background task to get released (because this was the last non-suspended
process assertion). When we release the UIProcess' background task, we call
WebProcessPool::notifyProcessPoolsApplicationIsAboutToSuspend(), which destroys non-critical
WebProcesses (e.g. WebProcesses in the back/forward cache), which in turns destroy their
ProcessThrottler. As a result, when replacing m_assertion, the ProcessAssertion may get
destroyed, which is why we would crash on the next line when trying to use m_assertion.
To address the issue, we now release the UIProcess's background task asynchronously when
releasing the last non-suspended ProcessAssertion, making sure we still need to release
the assertion beforehand. This has 2 benefits:
- The ProcessThrottler can no longer get destroyed synchronously when releasing its
ProcessAssertion.
- In the case where the ProcessThrottler replaces a foreground assertion with a background
assertion (or vice-versa) and this is the last non-suspended assertion, this avoids
unnecessarily releasing and retaking the UIProcess's background task. This also avoids
killing non-critical processes unnecessarily.
* UIProcess/ios/ProcessAssertionIOS.mm:
(-[WKProcessAssertionBackgroundTaskManager _updateBackgroundTask]):
2020-11-02 Brent Fulgham <bfulgham@apple.com>
[macOS] Remove unneeded shmem access to ColorSync
https://bugs.webkit.org/show_bug.cgi?id=218395
<rdar://problem/70354916>
Reviewed by Per Arne Vollan.
ColorSync on macOS has not used shared memory since Mac OS X 10.7, and our sandbox
should not have this access. Removing this unused shared memory capability.
* GPUProcess/mac/com.apple.WebKit.GPUProcess.sb.in:
* PluginProcess/mac/com.apple.WebKit.plugin-common.sb.in:
* WebProcess/com.apple.WebProcess.sb.in:
2020-11-01 Darin Adler <darin@apple.com>
Start removing functions that implicitly use composed tree
https://bugs.webkit.org/show_bug.cgi?id=218424
Reviewed by Ryosuke Niwa.
* WebProcess/WebPage/glib/WebPageGLib.cpp:
(WebKit::WebPage::getPlatformEditorState const): Use contains<ComposedTree>.
2020-11-02 David Kilzer <ddkilzer@apple.com>
Fix link error with WebKit.framework
Undefined symbols for architecture x86_64:
"JSC::GenericTypedArrayView<JSC::Uint8Adaptor>::create(WTF::RefPtr<JSC::ArrayBuffer, WTF::RawPtrTraits<JSC::ArrayBuffer>, WTF::DefaultRefDerefTraits<JSC::ArrayBuffer> >&&, unsigned int, unsigned int)", referenced from:
WebKit::convertToUint8Array(IPC::SharedBufferCopy&&) in UnifiedSource7.o
ld: symbol(s) not found for architecture x86_64
* GPUProcess/media/RemoteLegacyCDMSessionProxy.cpp:
- Include <JavaScriptCore/GenericTypedArrayViewInlines.h> to fix
linker error with some builds of WebKit.
- See also WebProcess/GPU/media/RemoteLegacyCDMSession.cpp.
2020-11-02 Youenn Fablet <youenn@apple.com>
Stop sending origins from WebProcess to UIProcess for getUserMedia requests
https://bugs.webkit.org/show_bug.cgi?id=218192
Reviewed by Eric Carlson.
Pass directly a UserMediaRequestIdentifier through IPC.
Stop passing origins from IPC and instead compute them from UIProcess side.
* Scripts/webkit/messages.py:
* UIProcess/UserMediaPermissionRequestManagerProxy.cpp:
(WebKit::UserMediaPermissionRequestManagerProxy::denyRequest):
(WebKit::UserMediaPermissionRequestManagerProxy::grantRequest):
(WebKit::UserMediaPermissionRequestManagerProxy::finishGrantingRequest):
(WebKit::UserMediaPermissionRequestManagerProxy::requestUserMediaPermissionForFrame):
(WebKit::UserMediaPermissionRequestManagerProxy::processUserMediaPermissionRequest):
(WebKit::UserMediaPermissionRequestManagerProxy::processUserMediaPermissionInvalidRequest):
(WebKit::UserMediaPermissionRequestManagerProxy::processUserMediaPermissionValidRequest):
(WebKit::UserMediaPermissionRequestManagerProxy::getUserMediaPermissionInfo):
(WebKit::UserMediaPermissionRequestManagerProxy::enumerateMediaDevicesForFrame):
* UIProcess/UserMediaPermissionRequestManagerProxy.h:
* UIProcess/UserMediaPermissionRequestProxy.cpp:
(WebKit::UserMediaPermissionRequestProxy::UserMediaPermissionRequestProxy):
* UIProcess/UserMediaPermissionRequestProxy.h:
(WebKit::UserMediaPermissionRequestProxy::create):
(WebKit::UserMediaPermissionRequestProxy::userMediaID const):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::requestUserMediaPermissionForFrame):
(WebKit::WebPageProxy::enumerateMediaDevicesForFrame):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebProcess/MediaStream/UserMediaPermissionRequestManager.cpp:
(WebKit::UserMediaPermissionRequestManager::sendUserMediaRequest):
(WebKit::UserMediaPermissionRequestManager::userMediaAccessWasGranted):
(WebKit::UserMediaPermissionRequestManager::userMediaAccessWasDenied):
(WebKit::UserMediaPermissionRequestManager::enumerateMediaDevices):
* WebProcess/MediaStream/UserMediaPermissionRequestManager.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::userMediaAccessWasGranted):
(WebKit::WebPage::userMediaAccessWasDenied):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-11-01 Sam Weinig <weinig@apple.com>
Remove remaining alternative preference setting mechanisms from LayoutTests
https://bugs.webkit.org/show_bug.cgi?id=218417
Reviewed by Simon Fraser.
Remove now unused bundle SPI that was only used by the WebKitTestRuner.
* WebProcess/InjectedBundle/API/c/WKBundle.cpp:
(WKBundleOverrideBoolPreferenceForTestRunner): Deleted.
(WKBundleSetAllowUniversalAccessFromFileURLs): Deleted.
(WKBundleSetAllowFileAccessFromFileURLs): Deleted.
(WKBundleSetAllowStorageAccessFromFileURLS): Deleted.
(WKBundleSetMinimumLogicalFontSize): Deleted.
(WKBundleSetFrameFlatteningEnabled): Deleted.
(WKBundleSetJavaScriptCanAccessClipboard): Deleted.
(WKBundleSetPopupBlockingEnabled): Deleted.
(WKBundleSetAuthorAndUserStylesEnabled): Deleted.
* WebProcess/InjectedBundle/API/c/WKBundlePrivate.h:
* WebProcess/InjectedBundle/InjectedBundle.cpp:
(WebKit::InjectedBundle::addOriginAccessAllowListEntry):
(WebKit::InjectedBundle::overrideBoolPreferenceForTestRunner): Deleted.
(WebKit::InjectedBundle::setAllowUniversalAccessFromFileURLs): Deleted.
(WebKit::InjectedBundle::setAllowFileAccessFromFileURLs): Deleted.
(WebKit::InjectedBundle::setNeedsStorageAccessFromFileURLsQuirk): Deleted.
(WebKit::InjectedBundle::setMinimumLogicalFontSize): Deleted.
(WebKit::InjectedBundle::setFrameFlatteningEnabled): Deleted.
(WebKit::InjectedBundle::setAsyncFrameScrollingEnabled): Deleted.
(WebKit::InjectedBundle::setJavaScriptCanAccessClipboard): Deleted.
(WebKit::InjectedBundle::setPopupBlockingEnabled): Deleted.
(WebKit::InjectedBundle::setAuthorAndUserStylesEnabled): Deleted.
* WebProcess/InjectedBundle/InjectedBundle.h:
2020-10-31 Simon Fraser <simon.fraser@apple.com>
Clean up BoxSide and BorderEdge code
https://bugs.webkit.org/show_bug.cgi?id=218197
Reviewed by Sam Weinig.
* UIProcess/ios/WKKeyboardScrollingAnimator.mm:
(boxSide):
2020-10-30 Brian Burg <bburg@apple.com>
Web Inspector: move InspectorFrontendAPIDispatcher to WebCore, clean up uses
https://bugs.webkit.org/show_bug.cgi?id=217835
<rdar://problem/70384407>
Reviewed by Devin Rousso.
Adopt the JSONValue-based InspectorFrontendAPIDispatcher::dispatch() method
in WebInspectorUI and RemoteInspectorUI methods that dispatch to the frontend.
* Sources.txt:
* WebKit.xcodeproj/project.pbxproj: Move files.
* WebProcess/Inspector/RemoteWebInspectorUI.h:
* WebProcess/Inspector/RemoteWebInspectorUI.cpp:
(WebKit::RemoteWebInspectorUI::RemoteWebInspectorUI):
(WebKit::RemoteWebInspectorUI::initialize):
(WebKit::RemoteWebInspectorUI::updateFindString):
(WebKit::RemoteWebInspectorUI::didSave):
(WebKit::RemoteWebInspectorUI::didAppend):
(WebKit::RemoteWebInspectorUI::frontendLoaded):
(WebKit::RemoteWebInspectorUI::sendMessageToFrontend):
(WebKit::RemoteWebInspectorUI::pagePaused):
(WebKit::RemoteWebInspectorUI::pageUnpaused):
(WebKit::RemoteWebInspectorUI::setDiagnosticLoggingAvailable):
* WebProcess/Inspector/WebInspectorUI.h:
* WebProcess/Inspector/WebInspectorUI.cpp:
(WebKit::WebInspectorUI::WebInspectorUI):
(WebKit::WebInspectorUI::setDockSide):
(WebKit::WebInspectorUI::setDockingUnavailable):
(WebKit::WebInspectorUI::setIsVisible):
(WebKit::WebInspectorUI::updateFindString):
(WebKit::WebInspectorUI::setDiagnosticLoggingAvailable):
(WebKit::WebInspectorUI::showConsole):
(WebKit::WebInspectorUI::showResources):
(WebKit::WebInspectorUI::showMainResourceForFrame):
(WebKit::WebInspectorUI::startPageProfiling):
(WebKit::WebInspectorUI::stopPageProfiling):
(WebKit::WebInspectorUI::startElementSelection):
(WebKit::WebInspectorUI::stopElementSelection):
(WebKit::WebInspectorUI::didSave):
(WebKit::WebInspectorUI::didAppend):
(WebKit::WebInspectorUI::sendMessageToFrontend):
(WebKit::WebInspectorUI::evaluateInFrontendForTesting):
(WebKit::WebInspectorUI::pagePaused):
(WebKit::WebInspectorUI::pageUnpaused):
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
Add missing header includes (related to unified sources changes from r269168).
2020-10-30 Brent Fulgham <bfulgham@apple.com>
Correct sandbox violations in GPU Process
https://bugs.webkit.org/show_bug.cgi?id=218356
<rdar://problem/70574999>
Reviewed by Eric Carlson.
The new GPU Process sandbox is missing 'sysctl.name2oid' from the allow list, even though
we allow it everywhere else. We need this for proper function.
* GPUProcess/mac/com.apple.WebKit.GPUProcess.sb.in:
* Resources/SandboxProfiles/ios/com.apple.WebKit.GPU.sb:
2020-10-30 Simon Fraser <simon.fraser@apple.com>
Convert ScrollingTreeNode change flags to an OptionSet<>
https://bugs.webkit.org/show_bug.cgi?id=218374
Reviewed by Antti Koivisto.
Address the FIXME in ScrollingStateNode.h and use an OptionSet<> for the
ScrollingState tree change flags. This required moving them all into
the same enum class in ScrollingStateNode.
* Shared/RemoteLayerTree/RemoteScrollingCoordinatorTransaction.cpp:
(ArgumentCoder<ScrollingStateNode>::encode):
(ArgumentCoder<ScrollingStateNode>::decode):
(ArgumentCoder<ScrollingStateScrollingNode>::encode):
(ArgumentCoder<ScrollingStateFrameScrollingNode>::encode):
(ArgumentCoder<ScrollingStateOverflowScrollProxyNode>::encode):
(ArgumentCoder<ScrollingStateScrollingNode>::decode):
(ArgumentCoder<ScrollingStateFrameScrollingNode>::decode):
(ArgumentCoder<ScrollingStateOverflowScrollProxyNode>::decode):
(ArgumentCoder<ScrollingStateFixedNode>::encode):
(ArgumentCoder<ScrollingStateFixedNode>::decode):
(ArgumentCoder<ScrollingStateStickyNode>::encode):
(ArgumentCoder<ScrollingStateStickyNode>::decode):
(ArgumentCoder<ScrollingStatePositionedNode>::encode):
(ArgumentCoder<ScrollingStatePositionedNode>::decode):
(WebKit::dump):
* UIProcess/RemoteLayerTree/RemoteScrollingCoordinatorProxy.cpp:
(WebKit::RemoteScrollingCoordinatorProxy::connectStateNodeLayers):
* UIProcess/RemoteLayerTree/ios/RemoteScrollingCoordinatorProxyIOS.mm:
(WebKit::RemoteScrollingCoordinatorProxy::connectStateNodeLayers):
* UIProcess/RemoteLayerTree/ios/ScrollingTreeFrameScrollingNodeRemoteIOS.mm:
(WebKit::ScrollingTreeFrameScrollingNodeRemoteIOS::commitStateBeforeChildren):
(WebKit::ScrollingTreeFrameScrollingNodeRemoteIOS::commitStateAfterChildren):
* UIProcess/RemoteLayerTree/ios/ScrollingTreeOverflowScrollingNodeIOS.mm:
(WebKit::ScrollingTreeOverflowScrollingNodeIOS::commitStateBeforeChildren):
* UIProcess/RemoteLayerTree/ios/ScrollingTreeScrollingNodeDelegateIOS.mm:
(WebKit::ScrollingTreeScrollingNodeDelegateIOS::commitStateBeforeChildren):
(WebKit::ScrollingTreeScrollingNodeDelegateIOS::commitStateAfterChildren):
* UIProcess/RemoteLayerTree/mac/ScrollingTreeFrameScrollingNodeRemoteMac.cpp:
(WebKit::ScrollingTreeFrameScrollingNodeRemoteMac::commitStateBeforeChildren):
* UIProcess/RemoteLayerTree/mac/ScrollingTreeOverflowScrollingNodeRemoteMac.cpp:
(WebKit::ScrollingTreeOverflowScrollingNodeRemoteMac::commitStateBeforeChildren):
2020-10-29 Aditya Keerthi <akeerthi@apple.com>
[Cocoa] Remove soft linking of NetworkExtension.framework
https://bugs.webkit.org/show_bug.cgi?id=218314
<rdar://problem/70785239>
Reviewed by Andy Estes.
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::addPlatformLoadParameters):
Replaced the soft linked method call with a new method exported from
WebCore.
* WebProcess/WebPage/Cocoa/WebPageCocoa.mm:
(WebKit::WebPage::platformDidReceiveLoadParameters):
2020-10-29 Chris Dumez <cdumez@apple.com>
Regression(PSON): Back/forward navigation may hang
https://bugs.webkit.org/show_bug.cgi?id=216611
<rdar://problem/68992714>
Reviewed by Geoffrey Garen.
The scenario was as follows:
1. Load A in Process P1
2. Load A#foo in Process P1 (fragment navigation)
3. Load B in Process P2
4. history.go(-2) to load A again in P1
5. history.go(2) to load B again in P2
6. history.back() to go back to A#foo in P1
-> Hang
We doing the process swap at step 3, we would suspend the page containing A#foo
in P1 when doing the new load in P2. When process swapping again at step 4,
we would go back to process P1 (because it is same origin) but we did not have
any suspended WebPage for A so we would create a new WebPage (with a new
WebPageIdentifier). The reason we do this is to allow the suspended WebPage
for A#foo to remain in process, as we may want it for another back/forward
navigation later.
The issue was that when constructing the new WebPage in P1 for A at step 4,
we would restore all history items from the WebPageProxy's BackForwardList
and overwrite any existing HistoryItems with the same IDs in the WebProcess.
This meant that WebKit layer would construct new HistoryItems for
HistoryItems that already existing and are used by the Suspended WebPage
containing A#foo.
Later on, at step 6, when trying to go back A#foo in P1, we would ask
the Suspended WebPage in P1 to load a given HistoryItem. This HistoryItem
is supposed to be present in the Page's back/forward list and is supposed
to have an associated back/forward cache entry. However, it was not the
case here because we ended up with a new HistoryItem instance that was
restored from the UIProcess earlier.
To address the issue, The WebPage constructor no longer overwrites
existing HistoryItems. Note that before prior to r231048, the WebPage
constructor was NOT overwriting existing HistoryItems so this restores
previous behavior. Presumably, we started overwriting to stop hitting
the assertion in WebBackForwardListProxy::addItemFromUIProcess() that
checked that we either wanted to overwrite or that there was no
existing HistoryItem with the given ID. This is no longer an issue
since I replaced this assertion with an if-check and an early return.
* WebProcess/WebPage/WebBackForwardListProxy.cpp:
(WebKit::WebBackForwardListProxy::addItemFromUIProcess):
* WebProcess/WebPage/WebPage.cpp:
2020-10-29 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Make WebContent process talk to the WebAuthn process for WebAuthn requests
https://bugs.webkit.org/show_bug.cgi?id=218070
<rdar://problem/70384404>
Reviewed by Brent Fulgham.
Once modern web authentication is enabled, let WebContent process talk to the WebAuthn process
for WebAuthn requests.
* Scripts/process-entitlements.sh:
Adds necessary entitlements for the WebAuthn process.
* UIProcess/WebAuthentication/AuthenticatorManager.cpp:
(WebKit::AuthenticatorManager::handleRequest):
Adds a way to distinguish if itself is in the UI Process or in the WebAuthn process.
* WebAuthnProcess/WebAuthnConnectionToWebProcess.cpp:
(WebKit::WebAuthnConnectionToWebProcess::makeCredential):
(WebKit::WebAuthnConnectionToWebProcess::getAssertion):
(WebKit::WebAuthnConnectionToWebProcess::handleRequest):
(WebKit::WebAuthnConnectionToWebProcess::isUserVerifyingPlatformAuthenticatorAvailable):
(WebKit::WebAuthnConnectionToWebProcess::setMockWebAuthenticationConfiguration):
* WebAuthnProcess/WebAuthnConnectionToWebProcess.h:
* WebAuthnProcess/WebAuthnConnectionToWebProcess.messages.in:
Adds receivers for messages coming from the web content processes.
* WebAuthnProcess/WebAuthnProcess.cpp:
(WebKit::WebAuthnProcess::WebAuthnProcess):
(WebKit::WebAuthnProcess::setMockWebAuthenticationConfiguration):
* WebAuthnProcess/WebAuthnProcess.h:
(WebKit::WebAuthnProcess::authenticatorManager):
Handles the authenticator managers.
* WebAuthnProcess/mac/com.apple.WebKit.WebAuthnProcess.sb.in:
Relaxes sandbox for necessary hardware accesses.
* WebProcess/WebAuthentication/WebAuthenticatorCoordinator.cpp:
(WebKit::WebAuthenticatorCoordinator::makeCredential):
(WebKit::WebAuthenticatorCoordinator::getAssertion):
(WebKit::WebAuthenticatorCoordinator::isUserVerifyingPlatformAuthenticatorAvailable):
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::setMockWebAuthenticationConfiguration):
Sends messages to the WebAuthn process accordingly.
2020-10-29 Per Arne Vollan <pvollan@apple.com>
[macOS] Add required entitlement for message filtering
https://bugs.webkit.org/show_bug.cgi?id=218342
<rdar://problem/70810567>
Reviewed by Brent Fulgham.
The entitlement 'com.apple.private.security.message-filter' needs to be set to YES. This patch also
adds telemetry for mach-bootstrap filtering.
* Scripts/process-entitlements.sh:
* WebProcess/com.apple.WebProcess.sb.in:
2020-10-29 Alex Christensen <achristensen@webkit.org>
Don't look in the keychain for a preferred client certificate in the network process
https://bugs.webkit.org/show_bug.cgi?id=218322
<rdar://problem/64931374>
Reviewed by Geoffrey Garen.
Often, based on a race condition of preconnecting to a server we have just been told to load and
the use of Safari's BackgroundLoad class when navigating to a page we have not received any data from yet,
and with an installed client certificate and an identity preference in the keychain, the user will be asked
if com.apple.WebKit.Networking can access a private key in the keychain instead of if Safari can access a
private key in the keychain. If the user types in the password and clicks "Always Allow" this does not make
it always allowed, but it would have if Safari had asked instead of com.apple.WebKit.Networking.
This is because Safari is responding to WKNavigationDelegate's didReceiveAuthenticationChallenge with
NSURLSessionAuthChallengeUseCredential and nil, which would cause CFNetwork to search in the keychain as the network
process for the preferred client certificate. What we want Safari's network process to do is not search in the keychain
for this preconnect request's challenge but wait until the actual request, at which time Safari will use its proper logic
to find the correct client certificate and AuthenticationManager::initializeConnection will create a SecKeyProxy
to do the signing in the UI process. Third party applications will not be affected because the SecKeyProxy path is the
only one that works for applications lacking Safari's entitlements. I used the steps in the radar to verify that this
is fixed. Unfortunately, it is not practical to make a unit test that installs a system client certificate and an
identity preference because doing so would require the entry of the login keychain password while running the unit test.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::configurationForSessionID):
(WebKit::NetworkSessionCocoa::NetworkSessionCocoa):
2020-10-29 Said Abou-Hallawa <said@apple.com>
REGRESSION(269065): [GPU Process]: Order of drawing has to be preserved when drawing a canvas to another canvas
https://bugs.webkit.org/show_bug.cgi?id=218324
Reviewed by Simon Fraser.
When drawing an ImageBuffer to another ImageBuffer, the DrawingContext of
the source and the destination ImageBuffers have to be flushed immediately.
Otherwise an older version or a newer version of the source ImageBuffer
might be drawn to the destination ImageBuffer.
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
2020-10-29 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed, reverting r269109.
Introduced dependency cycle
Reverted changeset:
"[Cocoa] Remove soft linking of Contacts.framework"
https://bugs.webkit.org/show_bug.cgi?id=218289
https://trac.webkit.org/changeset/269109
2020-10-29 Per Arne Vollan <pvollan@apple.com>
[macOS] Add telemetry for IOKit use
https://bugs.webkit.org/show_bug.cgi?id=218343
<rdar://problem/70811875>
Reviewed by Geoffrey Garen.
Add telemetry with backtrace for IOKIt use in the WebContent process.
* WebProcess/com.apple.WebProcess.sb.in:
2020-10-29 Chris Lord <clord@igalia.com>
[WPE] Add axis-locking to kinetic scrolling
https://bugs.webkit.org/show_bug.cgi?id=209729
Reviewed by Adrian Perez de Castro.
Add axis-locking to scroll gestures on WPE, and new WPE-specific
settings to control the process.
* UIProcess/API/wpe/ScrollGestureController.cpp:
(WebKit::ScrollGestureController::handleEvent):
* UIProcess/API/wpe/ScrollGestureController.h:
2020-10-29 Youenn Fablet <youenn@apple.com>
Improve LibWebRTCSocketClient logging
https://bugs.webkit.org/show_bug.cgi?id=218336
Reviewed by Eric Carlson.
Output webrtc socket identifier in case of error.
Fix the case of LibWebRTCSocketClient::sendTo that can log too much.
No observable change of behavior.
* NetworkProcess/webrtc/LibWebRTCSocketClient.cpp:
(WebKit::LibWebRTCSocketClient::sendTo):
(WebKit::LibWebRTCSocketClient::close):
(WebKit::LibWebRTCSocketClient::setOption):
2020-10-29 Chris Lord <clord@igalia.com>
[GTK] Smooth scrolling should not apply to continuous scrolling with sync scrolling
https://bugs.webkit.org/show_bug.cgi?id=218133
Reviewed by Adrian Perez de Castro.
Set hasPreciseScrollDeltas appropriately on scroll events created from GdkEvent.
* Shared/gtk/WebEventFactory.cpp:
(WebKit::WebEventFactory::createWebWheelEvent):
2020-10-28 John Wilander <wilander@apple.com>
PCM: Accept ad click data when the link opens a new window
https://bugs.webkit.org/show_bug.cgi?id=214176
<rdar://problem/65358005>
Reviewed by Brent Fulgham.
A link with the attribute target="_blank" takes another code path for
navigation which involves the creation of a new window and webpage. That
code path needs to transfer ad click attribution data to the new webpage
where it can be picked up in WebPageProxy::didCommitLoadForFrame().
The ad click attribution data sits in the NavigationAction which is not
available in the completion handler WebPageProxy::createNewPage().
The client, which differs between TestRunner and e.g. Safari, consumes
the NavigationAction. I don't want to risk a regression in e.g. Safari
while still passing the test because TestRunner hasn't regressed.
Test: http/tests/adClickAttribution/attribution-conversion-through-image-redirect-in-new-window.html
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didCommitLoadForFrame):
Now also checks for pending ad click attribution data in its new
member variable m_newPageNavigationAdClickAttribution.
(WebKit::WebPageProxy::createNewPage):
Now forwards optional ad click attribution data to the completion
handler where it can be stored on the newly created webpage in the
new member variable m_newPageNavigationAdClickAttribution.
* UIProcess/WebPageProxy.h:
Added m_newPageNavigationAdClickAttribution.
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::createWindow):
Added missing navigationAction.adClickAttribution() copy.
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::dispatchDecidePolicyForNewWindowAction):
Added missing navigationAction.adClickAttribution() copy.
2020-10-28 Said Abou-Hallawa <said@apple.com>
[GPU Process] Eagerly flush the PutImageData item to the GPU Process
https://bugs.webkit.org/show_bug.cgi?id=218116
Reviewed by Simon Fraser.
Flush the DrawingContext of the RemoteImageBufferProxy once the PutImageData
item is recorded. So no expensive operation is going to block the painting.
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
2020-10-28 Tim Horton <timothy_horton@apple.com>
macCatalyst WebGL on Apple Silicon devices is using a software renderer
https://bugs.webkit.org/show_bug.cgi?id=218303
<rdar://problem/70587571>
Reviewed by Geoffrey Garen.
* UIProcess/mac/HighPerformanceGPUManager.mm:
(WebKit::HighPerformanceGPUManager::addProcessRequiringHighPerformance):
(WebKit::HighPerformanceGPUManager::removeProcessRequiringHighPerformance):
(WebKit::HighPerformanceGPUManager::updateState):
(WebKit::isiOSAppOnMac): Deleted.
HighPerformanceGPUManager is PLATFORM(MAC)-only, which is not true
for MACCATALYST, so delete this dead code.
2020-10-28 Sam Weinig <weinig@apple.com>
Reduce Preference Override Methods: TabsToLinks/SpatialNavigation
https://bugs.webkit.org/show_bug.cgi?id=218288
Reviewed by Tim Horton.
Removes WKPreferencesGet/SetTabToLinksEnabled preference as it duplicates functionality of
WKPreferencesGet/SetTabsToLinks. The former was added for https://bugs.webkit.org/show_bug.cgi?id=95329
but was redundant even at the time. We can now use test header commands for all of its
use cases.
Also removes testing only bundle SPI for enabling spatial navigation which also can be
set via test headers instead.
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetTabToLinksEnabled): Deleted.
(WKPreferencesGetTabToLinksEnabled): Deleted.
* UIProcess/API/C/WKPreferencesRefPrivate.h:
* WebProcess/InjectedBundle/API/c/WKBundle.cpp:
(WKBundleSetSpatialNavigationEnabled): Deleted.
* WebProcess/InjectedBundle/API/c/WKBundlePrivate.h:
* WebProcess/InjectedBundle/InjectedBundle.cpp:
(WebKit::InjectedBundle::overrideBoolPreferenceForTestRunner):
(WebKit::InjectedBundle::setSpatialNavigationEnabled): Deleted.
* WebProcess/InjectedBundle/InjectedBundle.h:
* WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::setTabToLinksEnabled): Deleted.
(WebKit::WebPage::tabToLinksEnabled const): Deleted.
2020-10-28 Aditya Keerthi <akeerthi@apple.com>
[Cocoa] Remove soft linking of Contacts.framework
https://bugs.webkit.org/show_bug.cgi?id=218289
Reviewed by Geoff Garen.
Since Contacts.framework does not depend on WebKit and soft linking is
discouraged, WebKit should link Contacts.framework normally.
Soft linking requires running the linker at runtime, leading to an
unwanted performance cost. By normally linking (on iOS) and weak linking
(on macOS), the linkage is declared at build time, which enables running
the linker at build/update/install time and avoiding the runtime cost.
* Configurations/WebKit.xcconfig:
On macOS, weak link the framework since Contacts.framework is not
available on the Base System.
* Platform/cocoa/PaymentAuthorizationPresenter.mm:
(WebKit::toNSError):
2020-10-28 Adrian Perez de Castro <aperez@igalia.com>
[GTK4] Build broken with GTK 3.99.3
https://bugs.webkit.org/show_bug.cgi?id=218270
Reviewed by Carlos Garcia Campos.
No new tests needed.
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(toplevelWindowStateChanged): Replace GDK_SURFACE_STATE_MINIMIZED with GDK_TOPLEVEL_STATE_MINIMIZED.
* UIProcess/API/gtk/WebKitWebViewGtk.cpp:
(surfaceStateChangedCallback): Ditto, and GDK_SURFACE_STATE_MAXIMIZED with GDK_TOPLEVEL_STATE_MAXIMIZED.
* UIProcess/Gamepad/gtk/UIGamepadProviderGtk.cpp:
(WebKit::getWebPageProxy): Write an implementation which works for GTK4.
(WebKit::UIGamepadProvider::platformWebPageProxyForGamepadInput): Accomodate for windows
being a bit less special in GTK4 and use gtk_widget_has_focus() directly.
2020-10-27 Alex Christensen <achristensen@webkit.org>
Use _CFURLConnectionSetFrameworkStubs for SecItemShim instead of DYLD_INTERPOSE on Apple Silicon Macs
https://bugs.webkit.org/show_bug.cgi?id=218269
<rdar://problem/70491533>
Reviewed by Darin Adler.
r171066 introduced the use of _CFURLConnectionSetFrameworkStubs on iOS for CFNetwork to be able
to get and set credentials as the UI process. This is also needed on Apple Silicon Macs.
We should eventually replace it with an even cleaner per-NSURLSession solution, but this is a step
in the right direction, and I verified manually that it fixes the radar.
Covered by an API test that used to fail on Apple Silicon Macs.
* Shared/mac/SecItemShim.cpp:
(WebKit::initializeSecItemShim):
2020-10-27 Said Abou-Hallawa <said@apple.com>
Make RenderingMode a bool enum and remove ShouldAccelerate
https://bugs.webkit.org/show_bug.cgi?id=218264
Reviewed by Tim Horton.
When creating a remote ImageBuffer in WebKit, all we need to know is whether
the backend is accelerated or not. RemoteImageBufferProxy has to be backed
by a DisplayList. RemoteImageBuffer has to own the real backend. Two enum
values for RenderingMode is sufficient for creating the remote ImageBuffer.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::createImageBuffer):
* GPUProcess/graphics/RemoteRenderingBackend.h:
* Shared/WebCoreArgumentCoders.h:
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::createImageBuffer):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::createImageBuffer const):
* WebProcess/WebCoreSupport/WebChromeClient.h:
2020-10-27 Tim Horton <timothy_horton@apple.com>
Adopt the UIPointerInteraction API
https://bugs.webkit.org/show_bug.cgi?id=218266
<rdar://problem/70732850>
Reviewed by Wenson Hsieh.
No new tests, just moving from deprecated SPI to API.
* Platform/spi/ios/UIKitSPI.h:
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView setUpInteraction]):
(-[WKContentView cleanUpInteraction]):
(-[WKContentView setUpPointerInteraction]):
(-[WKContentView _pointerInteraction:regionForRequest:defaultRegion:completion:]):
(-[WKContentView pointerRegionForPositionInformation:point:]):
(-[WKContentView pointerInteraction:styleForRegion:]):
(-[WKContentView setUpCursorInteraction]): Deleted.
(-[WKContentView _cursorInteraction:regionForLocation:defaultRegion:completion:]): Deleted.
(-[WKContentView cursorRegionForPositionInformation:point:]): Deleted.
(-[WKContentView cursorInteraction:styleForRegion:modifiers:]): Deleted.
2020-10-27 Chris Dumez <cdumez@apple.com>
[GPUProcess] Use async IPC for RemoteAudioDestinationManager's StartAudioDestination / StopAudioDestination
https://bugs.webkit.org/show_bug.cgi?id=218251
Reviewed by Geoffrey Garen.
Use async IPC for RemoteAudioDestinationManager's StartAudioDestination / StopAudioDestination.
* GPUProcess/media/RemoteAudioDestinationManager.messages.in:
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::start):
(WebKit::RemoteAudioDestinationProxy::stop):
* WebProcess/GPU/media/RemoteAudioDestinationProxy.h:
2020-10-27 Brian Burg <bburg@apple.com>
[Cocoa] Introduce _WKInspectorConfiguration for customizing local and remote Web Inspectors
https://bugs.webkit.org/show_bug.cgi?id=217896
<rdar://problem/70355910>
Reviewed by Devin Rousso.
Introduce _WKInspectorConfiguration for customizing the behavior of Web Inspector instances.
The initial customization is to allow for custom WKURLSchemeHandlers to be used by Web Inspector's
WebView to load resources from client-controlled locations. This can be used to implement loading
of extension resources using a custom scheme such as web-extension://.
Scheme handlers need to be registered at WebView creation time via WKWebViewConfiguration. In
order to configure a inspector page summoned from within WebKit (i.e., Inspect Element context menu item),
we need to add a method to the UI delegate to get a configuration when the page is being created.
This configuration object is used in two different SPI (local and remote cases):
- As part of WKUIDelegatePrivate, to retrieve a _WKInspectorConfiguration given a _WKInspector.
- As an argument to the _WKRemoteWebInspectorViewController initializer. It's used later as needed.
New API test: WKInspectorDelegate.InspectorConfiguration.
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
Add new files.
* UIProcess/API/APIInspectorConfiguration.h: Added.
* UIProcess/API/APIInspectorConfiguration.cpp: Added.
* UIProcess/API/Cocoa/_WKInspectorConfiguration.h: Added.
* UIProcess/API/Cocoa/_WKInspectorConfigurationInternal.h: Added.
* UIProcess/API/Cocoa/_WKInspectorConfiguration.mm: Added.
(-[_WKInspectorConfiguration init]):
(-[_WKInspectorConfiguration dealloc]):
(-[_WKInspectorConfiguration _apiObject]):
(-[_WKInspectorConfiguration setURLSchemeHandler:forURLScheme:]):
(-[_WKInspectorConfiguration applyToWebViewConfiguration:]):
(-[_WKInspectorConfiguration copyWithZone:]):
Create _WKInspectorConfiguration and add a method to register WKURLSchemeHandlers.
* Shared/API/APIObject.h:
* Shared/Cocoa/APIObject.mm:
(API::Object::newObject):
Add new API object types. Add missing InspectorExtension.
* UIProcess/API/Cocoa/WKUIDelegatePrivate.h:
* UIProcess/API/APIUIClient.h:
(API::UIClient::configurationForLocalInspector):
* UIProcess/Cocoa/UIDelegate.h:
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::setDelegate):
(WebKit::UIDelegate::UIClient::configurationForLocalInspector):
Add new client/delegate method to fetch an inspector configuration as needed.
* UIProcess/API/Cocoa/_WKRemoteWebInspectorViewController.h:
* UIProcess/API/Cocoa/_WKRemoteWebInspectorViewControllerPrivate.h:
* UIProcess/API/Cocoa/_WKRemoteWebInspectorViewController.mm:
(-[_WKRemoteWebInspectorViewController initWithConfiguration:]): Renamed from -init.
(-[_WKRemoteWebInspectorViewController init]): Deleted.
(-[_WKRemoteWebInspectorViewController loadForDebuggableType:backendCommandsURL:]):
(-[_WKRemoteWebInspectorViewController configurationForDebuggable:]):
(-[_WKRemoteWebInspectorViewController _setDiagnosticLoggingDelegate:]):
Store a _WKInspectorConfiguration and provide it when asked by RemoteWebInspectorProxy.
* UIProcess/Inspector/RemoteWebInspectorProxy.h:
* UIProcess/Inspector/RemoteWebInspectorProxy.cpp:
(WebKit::RemoteWebInspectorProxy::load):
Store m_debuggableInfo before creating the page and window. It's used from inside
platformCreateFrontendPageAndWindow to pass as an argument to the delegate method.
* UIProcess/Inspector/mac/RemoteWebInspectorProxyMac.mm:
(WebKit::RemoteWebInspectorProxy::platformCreateFrontendPageAndWindow):
* UIProcess/Inspector/mac/WebInspectorProxyMac.mm:
(WebKit::WebInspectorProxy::platformCreateFrontendPage):
Obtain a configuration and use it to initialize the WKInspectorViewController.
* UIProcess/Inspector/mac/WKInspectorViewController.h:
* UIProcess/Inspector/mac/WKInspectorViewController.mm:
(-[WKInspectorViewController initWithConfiguration:inspectedPage:]):
(-[WKInspectorViewController webView]):
(-[WKInspectorViewController webViewConfiguration]): Renamed from -configuration.
(-[WKInspectorViewController initWithInspectedPage:]): Deleted.
(-[WKInspectorViewController configuration]): Deleted.
Apply the URL scheme handlers registered in the _WKInspectorConfiguration to the
WKWebViewConfiguration.
* UIProcess/Inspector/glib/RemoteInspectorClient.cpp:
* UIProcess/Inspector/socket/RemoteInspectorClient.cpp:
Stub out RemoteWebInspectorProxyClient::configurationForRemoteInspector().
* UIProcess/API/Cocoa/_WKUserStyleSheet.h:
* UIProcess/API/Cocoa/_WKUserStyleSheet.mm:
Fix build problems caused by repartitioning of unified sources.
2020-10-27 Said Abou-Hallawa <sabouhallawa@apple.com>
[GPU Process]: Implement DisplayList::DrawImageBuffer item
https://bugs.webkit.org/show_bug.cgi?id=217566
Reviewed by Simon Fraser.
The sequence in the Web Process is the following:
-- GraphicsContext::drawImageBuffer() uses DisplayList::Recorder to check
whether it is appropriate to create a DrawImageBuffer item or not.
-- DisplayList::Recorder::drawImageBuffer() uses the delegate to see if
the ImageBuffer is remote or not and whether it can be locked till
replaying back the DisplayList in the GPU side.
-- RemoteImageBufferProxy inherits DisplayList::Recorder::Delegate. So
its lockRemoteImageBuffer() is called.
-- If the ImageBuffer can be locked, a DrawImageBufferItem is created with
the ImageBuffer::renderingResourceIdentifier().
The sequence in the GPU Process is the following:
-- When replaying back the DisplayList, DisplayList::Replayer will call
its delegate to check if it wants to apply the DrawImageBuffer.
-- Because RemoteImageBuffer inherits DisplayList::Replayer::Delegate,
RemoteImageBuffer::apply() will be called.
-- RemoteImageBuffer will call RemoteRenderingBackend::applyResourceItem()
which will check whether the item's renderingResourceIdentifier is one
of the ImageBuffers in its RemoteResourceCache or not.
-- If there is a cached ImageBuffer, RemoteRenderingBackend::applyResourceItem()
will draw it in the GraphicsContext.
* GPUProcess/graphics/RemoteImageBuffer.h:
(WebKit::RemoteImageBuffer::apply): Deleted.
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::applyResourceItem):
(WebKit::RemoteRenderingBackend::applyMediaItem):
* GPUProcess/graphics/RemoteRenderingBackend.h:
* GPUProcess/graphics/RemoteResourceCache.cpp:
(WebKit::RemoteResourceCache::cachedImageBuffer):
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
* WebProcess/GPU/graphics/RemoteResourceCacheProxy.cpp:
(WebKit::RemoteResourceCacheProxy::releaseImageBuffer):
(WebKit::RemoteResourceCacheProxy::lockRemoteImageBufferForRemoteClient):
(WebKit::RemoteResourceCacheProxy::lockRemoteResourceForRemoteClient):
(WebKit::RemoteResourceCacheProxy::releaseRemoteResource):
(WebKit::RemoteResourceCacheProxy::unlockRemoteResourcesForRemoteClient):
* WebProcess/GPU/graphics/RemoteResourceCacheProxy.h:
2020-10-27 Alex Christensen <achristensen@webkit.org>
Use NSURLSessionAuthChallengeRejectProtectionSpace if WKNavigationDelegate didReceiveAuthenticationChallenge is not implemented
https://bugs.webkit.org/show_bug.cgi?id=218008
Reviewed by Darin Adler.
This is documented in WKNavigationDelegate.h, and the behavior is basically identical.
The difference is if there is a challenge with multiple protection spaces and a delegate that implements didReceiveAuthenticationChallenge
is attached between the delegate callbacks, it will receive the second callback. This is unlikely and next to impossible to test, but
out of principle we should behave like we have documented that we do.
* UIProcess/Cocoa/NavigationState.mm:
(WebKit::NavigationState::NavigationClient::didReceiveAuthenticationChallenge):
2020-10-27 Sam Weinig <weinig@apple.com>
Rename WKPreferencesSet*ValueForKey SPI to WKPreferencesSet*ValueForKeyForTesting
https://bugs.webkit.org/show_bug.cgi?id=218240
Reviewed by Darin Adler.
Addresses feedback from initial naming to make it clear what they are used
for and for consistency with WebKitLegacy.
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetBoolValueForKeyForTesting):
(WKPreferencesSetDoubleValueForKeyForTesting):
(WKPreferencesSetUInt32ValueForKeyForTesting):
(WKPreferencesSetStringValueForKeyForTesting):
(WKPreferencesSetBoolValueForKey): Deleted.
(WKPreferencesSetDoubleValueForKey): Deleted.
(WKPreferencesSetUInt32ValueForKey): Deleted.
(WKPreferencesSetStringValueForKey): Deleted.
* UIProcess/API/C/WKPreferencesRefPrivate.h:
2020-10-27 Carlos Garcia Campos <cgarcia@igalia.com>
WebDriver: sequence of char key press is not supported
https://bugs.webkit.org/show_bug.cgi?id=217951
Reviewed by Brian Burg.
We are assuming there can be only one char key pressed at a time. Use a HashSet to store the currently pressed
char keys and the handle them the same way we do with virtual keys.
Fixes: imported/w3c/webdriver/tests/perform_actions/key_events.py::test_sequence_of_keydown_printable_keys_sends_events
* UIProcess/Automation/SimulatedInputDispatcher.cpp:
(WebKit::SimulatedInputDispatcher::transitionInputSourceToState):
* UIProcess/Automation/SimulatedInputDispatcher.h:
* UIProcess/Automation/WebAutomationSession.cpp:
(WebKit::WebAutomationSession::performInteractionSequence):
2020-10-27 Tetsuharu Ohzeki <tetsuharu.ohzeki@gmail.com>
Make WebCore::FocusDirection to enum class
https://bugs.webkit.org/show_bug.cgi?id=218162
Reviewed by Darin Adler.
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::takeFocus):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::takeFocus):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::takeFocus):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::handleKeyEventByRelinquishingFocusToChrome):
(WebKit::WebPage::setInitialFocus):
2020-10-26 Said Abou-Hallawa <sabouhallawa@apple.com>
[GPU Process]: Introduce RemoteResourceCacheProxy to manage the remote resources in Web Process
https://bugs.webkit.org/show_bug.cgi?id=217558
Reviewed by Simon Fraser.
Remove RemoteImageBufferMessageHandlerProxy and move sending the messages
to RemoteRenderingBackendProxy.
To allow casting ImageBuffer to RemoteImageBufferProxy, a new method named
remoteResourceIdentifier() is added to IamgeBuffer and it is only overridden
by RemoteImageBufferProxy. This method is used in SPECIALIZE_TYPE_TRAITS
macros along with IamgeBuffer::isAccelerated().
Introduce RemoteResourceCacheProxy which will manage caching and releasing
the RemoteImageBufferProxy. Caching NativeImage will be added to it in
future patches. More about controlling the life cycles of remote resources
will be added also when supporting drawing remote resources to remote client
in future patches.
* GPUProcess/graphics/RemoteImageBuffer.h:
(WebKit::RemoteImageBuffer::create):
(WebKit::RemoteImageBuffer::RemoteImageBuffer):
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::imageBufferBackendWasCreated):
(WebKit::RemoteRenderingBackend::flushDisplayListWasCommitted):
(WebKit::RemoteRenderingBackend::createImageBuffer):
(WebKit::RemoteRenderingBackend::flushDisplayList):
(WebKit::RemoteRenderingBackend::flushDisplayListAndCommit):
(WebKit::RemoteRenderingBackend::getImageData):
(WebKit::RemoteRenderingBackend::releaseRemoteResource):
* GPUProcess/graphics/RemoteRenderingBackend.h:
* GPUProcess/graphics/RemoteRenderingBackend.messages.in:
* GPUProcess/graphics/RemoteResourceCache.cpp:
(WebKit::RemoteResourceCache::cacheImageBuffer):
(WebKit::RemoteResourceCache::cachedImageBuffer):
(WebKit::RemoteResourceCache::releaseRemoteResource):
* GPUProcess/graphics/RemoteResourceCache.h:
* Scripts/webkit/messages.py:
* Sources.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/graphics/PlatformRemoteImageBufferProxy.h:
(isType):
* WebProcess/GPU/graphics/RemoteImageBufferMessageHandlerProxy.cpp: Removed.
* WebProcess/GPU/graphics/RemoteImageBufferMessageHandlerProxy.h: Removed.
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
(WebKit::RemoteImageBufferProxy::create):
(WebKit::RemoteImageBufferProxy::~RemoteImageBufferProxy):
(WebKit::RemoteImageBufferProxy::createBackend):
(WebKit::RemoteImageBufferProxy::commitFlushDisplayList):
(WebKit::RemoteImageBufferProxy::RemoteImageBufferProxy):
(WebKit::RemoteImageBufferProxy::isPendingFlush const):
(WebKit::RemoteImageBufferProxy::timeoutWaitForFlushDisplayListWasCommitted):
(WebKit::RemoteImageBufferProxy::flushDrawingContextAndCommit):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::createImageBuffer):
(WebKit::RemoteRenderingBackendProxy::getImageData):
(WebKit::RemoteRenderingBackendProxy::flushDisplayList):
(WebKit::RemoteRenderingBackendProxy::flushDisplayListAndCommit):
(WebKit::RemoteRenderingBackendProxy::releaseRemoteResource):
(WebKit::RemoteRenderingBackendProxy::imageBufferBackendWasCreated):
(WebKit::RemoteRenderingBackendProxy::flushDisplayListWasCommitted):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
(WebKit::RemoteRenderingBackendProxy::remoteResourceCacheProxy):
(WebKit::RemoteRenderingBackendProxy::renderingBackendIdentifier const): Deleted.
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.messages.in:
* WebProcess/GPU/graphics/RemoteResourceCacheProxy.cpp: Copied from Source/WebKit/GPUProcess/graphics/RemoteResourceCache.cpp.
(WebKit::RemoteResourceCacheProxy::RemoteResourceCacheProxy):
(WebKit::RemoteResourceCacheProxy::cacheImageBuffer):
(WebKit::RemoteResourceCacheProxy::cachedImageBuffer):
(WebKit::RemoteResourceCacheProxy::releaseImageBuffer):
* WebProcess/GPU/graphics/RemoteResourceCacheProxy.h: Copied from Source/WebKit/GPUProcess/graphics/RemoteResourceCache.h.
2020-10-26 Alex Christensen <achristensen@webkit.org>
Add null checks and smart pointers in PDF and Plugin code
https://bugs.webkit.org/show_bug.cgi?id=218144
<rdar://problem/69931543>
Reviewed by Youenn Fablet.
It's nice when things don't crash.
isBeingAsynchronouslyInitialized wasn't called anywhere, so I removed it.
* WebProcess/Plugins/PDF/PDFPlugin.h:
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::PDFPlugin):
(WebKit::PDFPlugin::~PDFPlugin):
(WebKit::PDFPlugin::getResourceBytesAtPosition):
(WebKit::PDFPlugin::updateScrollbars):
(WebKit::PDFPlugin::createScrollbar):
(WebKit::PDFPlugin::isActive const):
(WebKit::PDFPlugin::forceUpdateScrollbarsOnMainThreadForPerformanceTesting const):
(WebKit::PDFPlugin::jsPDFDocPrint):
(WebKit::PDFPlugin::installPDFDocument):
(WebKit::PDFPlugin::updatePageAndDeviceScaleFactors):
(WebKit::PDFPlugin::calculateSizes):
(WebKit::PDFPlugin::willDetachRenderer):
(WebKit::PDFPlugin::destroy):
(WebKit::PDFPlugin::paintControlForLayerInContext):
(WebKit::PDFPlugin::convertFromPDFViewToScreen const):
(WebKit::PDFPlugin::boundsOnScreen const):
(WebKit::PDFPlugin::visibilityDidChange):
(WebKit::PDFPlugin::showContextMenuAtPoint):
(WebKit::PDFPlugin::handleContextMenuEvent):
(WebKit::PDFPlugin::isFullFramePlugin const):
(WebKit::PDFPlugin::handlesPageScaleFactor const):
(WebKit::PDFPlugin::clickedLink):
(WebKit::PDFPlugin::save):
(WebKit::PDFPlugin::openWithPreview):
(WebKit::PDFPlugin::saveToPDF):
(WebKit::PDFPlugin::openWithNativeApplication):
(WebKit::PDFPlugin::showDefinitionForAttributedString):
(WebKit::PDFPlugin::notifySelectionChanged):
(WebKit::PDFPlugin::notifyCursorChanged):
(WebKit::PDFPlugin::axObjectCache const):
(WebKit::PDFPlugin::performWebSearch):
(WebKit::PDFPlugin::performSpotlightSearch):
* WebProcess/Plugins/Plugin.cpp:
(WebKit::Plugin::Plugin):
(WebKit::Plugin::initialize):
(WebKit::Plugin::controller):
(WebKit::Plugin::controller const):
(WebKit::Plugin::~Plugin): Deleted.
* WebProcess/Plugins/Plugin.h:
(WebKit::Plugin::controller): Deleted.
(WebKit::Plugin::controller const): Deleted.
* WebProcess/Plugins/PluginController.h:
* WebProcess/WebPage/WebFrame.h:
(WebKit::WebFrame::setLoadListener):
(WebKit::WebFrame::loadListener const):
2020-10-26 Youenn Fablet <youenn@apple.com>
Add support for VP9 Profile 2 (10-bit color) in WebRTC
https://bugs.webkit.org/show_bug.cgi?id=217673
<rdar://problem/70283885>
Reviewed by Eric Carlson.
* WebProcess/GPU/webrtc/LibWebRTCCodecs.cpp:
(WebKit::LibWebRTCCodecs::pixelBufferPool):
2020-10-26 Zan Dobersek <zdobersek@igalia.com>
Remove Accelerated2dCanvasEnabled WebPreferences entry
https://bugs.webkit.org/show_bug.cgi?id=218114
Reviewed by Adrian Perez de Castro.
With the removal of the Accelerated2dCanvasEnabled preference key,
relevant API functions in the WK2 C API are made no-op. Similar
thing is done for the relevant GLib API, with the addition of marking
relevant API funcions as deprecated.
* Shared/API/c/WKDeprecatedFunctions.cpp:
(WKPreferencesSetAccelerated2DCanvasEnabled):
(WKPreferencesGetAccelerated2DCanvasEnabled):
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetAccelerated2DCanvasEnabled): Deleted.
(WKPreferencesGetAccelerated2DCanvasEnabled): Deleted.
* UIProcess/API/glib/WebKitSettings.cpp:
(webKitSettingsSetProperty):
(webKitSettingsGetProperty):
(webkit_settings_class_init):
(webkit_settings_get_enable_accelerated_2d_canvas):
(webkit_settings_set_enable_accelerated_2d_canvas):
* UIProcess/API/gtk/WebKitSettings.h:
* UIProcess/API/wpe/WPEView.cpp:
(WKWPE::m_backend):
* UIProcess/API/wpe/WebKitSettings.h:
2020-10-22 Ryosuke Niwa <rniwa@webkit.org>
IPC testing API should expose ArrayBuffer for an unknown decoded argument
https://bugs.webkit.org/show_bug.cgi?id=218113
Reviewed by Darin Adler.
Expose the decoded argument as an ArrayBuffer when it can't be encoded as a JS value.
Also make the treatment of a JS array as a tuple or a Vector explicit in argument encoding.
Test: TestWebKitAPI.IPCTestingAPI.CanInterceptAlert
* Platform/IPC/Decoder.h:
(IPC::Decoder::currentBufferPosition const): Added.
* Platform/IPC/JSIPCBinding.cpp:
(IPC::putJSValueForDecodedArgumentAtIndexOrArrayBufferIfUndefined): Added.
* Platform/IPC/JSIPCBinding.h:
(IPC::DecodedArgumentJSValueConverter): Deleted.
(IPC::putJSValueForDecodeArgumentInArray): Replaced DecodedArgumentJSValueConverter.
Take Decoder and decode each tuple argument and create a JS value at a time.
(IPC::jsValueForArgumentTuple): Deleted. Merged into jsValueForDecodedArguments.
(IPC::jsValueForDecodedArguments):
* WebProcess/WebPage/IPCTestingAPI.cpp:
(WebKit::IPCTestingAPI::VectorEncodeHelper::encode const):
(WebKit::IPCTestingAPI::encodeArrayArgument): Extracted out of encodeArgument.
(WebKit::IPCTestingAPI::encodeArgument): Treat an array as a tuple unless it's explicitly
specified that the argument's type is a Vector.
2020-10-23 Eric Carlson <eric.carlson@apple.com>
[Media in GPU Process] in-band metadata cues sometimes crash
https://bugs.webkit.org/show_bug.cgi?id=218106
<rdar://problem/70226170>
Reviewed by Jer Noble.
* Shared/mac/WebCoreArgumentCodersMac.mm:
(IPC::ArgumentCoder<WebCore::SerializedPlatformDataCueValue>::encodePlatformData):
SerializedPlatformDataCueValue::nativeValue() is a RetainPtr<>.
2020-10-23 Miguel Gomez <magomez@igalia.com>
[GTK][WPE] Implement antialiased rounded rectangle clipping in TextureMapper
https://bugs.webkit.org/show_bug.cgi?id=174457
Reviewed by Carlos Garcia Campos.
Use the new TextureMapper::beginClip() method that receives a FloatRoundedRect.
* Shared/CoordinatedGraphics/CoordinatedGraphicsScene.cpp:
(WebKit::CoordinatedGraphicsScene::paintToCurrentGLContext):
2020-10-23 Michael Catanzaro <mcatanzaro@gnome.org>
[SOUP] Fix crash in WebSocketTask
https://bugs.webkit.org/show_bug.cgi?id=217892
Reviewed by Carlos Garcia Campos.
The WebSocketTask connects to the "starting" signal of its SoupMessage and never disconnects
this signal, which is only safe if it is guaranteed to outlive its SoupMessage. However, it
is not. We crash when the signal is emitted after the WebSocketTask is destroyed. To solve
this, we just need to disconnect the signal when required. Normally that would be done in
the destructor, but the WebSocketTask drops its ownership of the SoupMessage prior to that
point, so we need to disconnect on each possible paths.
* NetworkProcess/soup/WebSocketTaskSoup.cpp:
(WebKit::WebSocketTask::~WebSocketTask):
(WebKit::WebSocketTask::didConnect):
(WebKit::WebSocketTask::didFail):
2020-10-22 Aditya Keerthi <akeerthi@apple.com>
[Contact Picker API] Add skeleton implementation of ContactsManager.select()
https://bugs.webkit.org/show_bug.cgi?id=218050
<rdar://problem/69862186>
Reviewed by Devin Rousso.
Added the necessary plumbing in order for the UIProcess to display a
contact picker after a call to ContactsManager.select() is made.
* UIProcess/PageClient.h:
(WebKit::PageClient::showContactPicker):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::showContactPicker):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::showContactPicker):
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::showContactPicker):
* WebProcess/WebPage/WebPage.h:
2020-10-22 Ryosuke Niwa <rniwa@webkit.org>
IPC testing API should have the capability to observe messages being sent and received
https://bugs.webkit.org/show_bug.cgi?id=217870
Unreviewed ASAN release build fix.
* Platform/IPC/JSIPCBinding.cpp:
* WebProcess/WebPage/IPCTestingAPI.cpp:
(WebKit::IPCTestingAPI::JSMessageListener::JSMessageListener):
2020-10-22 Eric Carlson <eric.carlson@apple.com>
UIClient isn't notified when page muted state changes
https://bugs.webkit.org/show_bug.cgi?id=218085
<rdar://problem/70462420>
Reviewed by Youenn Fablet.
API test: WKWebView.MediaMuted
* UIProcess/API/C/WKPage.cpp:
(WKPageGetMediaState): reportedMediaCaptureState -> reportedMediaState.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _mediaCaptureState]): Ditto.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::updateReportedMediaCaptureState): Only record current capture
state in m_reportedMediaCaptureState. m_delayStopCapturingReporting isn't necessary,
remove it.
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::reportedMediaState const): Renamed from reportedMediaCaptureState
to reflect what it returns.
(WebKit::WebPageProxy::mediaStateFlags const): Deleted.
(WebKit::WebPageProxy::reportedMediaCaptureState const): Deleted.
2020-10-22 Youenn Fablet <youenn@apple.com>
Introduce worklet destinations and allow to fetch modules with CORS based on destination
https://bugs.webkit.org/show_bug.cgi?id=218019
<rdar://problem/70526201>
Reviewed by Chris Dumez.
Add support to new fetch destinations in enumerations.
* NetworkProcess/NetworkLoadChecker.cpp:
(WebKit::NetworkLoadChecker::isAllowedByContentSecurityPolicy):
* NetworkProcess/NetworkResourceLoader.cpp:
(WebKit::NetworkResourceLoader::resourceLoadInfo):
2020-10-22 Nitzan Uziely <linkgoron@gmail.com>
Elements in Shadow DOM are wrongly marked as stale by the WebDriver
https://bugs.webkit.org/show_bug.cgi?id=217635
Reviewed by Brian Burg.
Fixed the focus script to work for elements in Shadow DOM
instead of declaring them as not interactable.
* WebProcess/Automation/WebAutomationSessionProxy.js:
(let.AutomationSessionProxy.prototype._clearStaleNodes):
(let.AutomationSessionProxy):
2020-10-22 Aditya Keerthi <akeerthi@apple.com>
[iOS] Prevent presentation of input peripherals when focusing form controls with a validation message
https://bugs.webkit.org/show_bug.cgi?id=218004
<rdar://problem/70507678>
Reviewed by Wenson Hsieh.
Interactive form validation can result in the presentation of a
validation message bubble near the first form control that has invalid
data. Prior to displaying the message, the invalid control is focused.
On iOS, this also has the effect of also presenting a virtual keyboard
or another custom input peripheral, such as a context menu for date
inputs.
Attempting to present both the validation message and custom input
peripheral can leave the view in an inconsistent state. For example,
<select> popovers have a strange flashing behavior when presented
alongside a validation message, and context menus can fail to present
entirely.
In order to address these issues, we should never attempt to present
both a validation message and an input peripheral. Instead, we can
prevent the presentation of input peripherals when the focused control
is presenting a validation message. This behavior matches macOS. Note
that we still present the keyboard for controls that have a keyboard
view, since the keyboard area does overlap the area where a validation
message is presented.
* Shared/FocusedElementInformation.cpp:
(WebKit::FocusedElementInformation::encode const):
(WebKit::FocusedElementInformation::decode):
* Shared/FocusedElementInformation.h:
Added isFocusingWithValidationMessage to the struct, so that the UIProcess
knows that the element gained focus due to the presentation of a
validation message.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _elementDidFocus:userIsInteracting:blurPreviousNode:activityStateChanges:userObject:]):
Prevent an input view from being shown if the control does not present
a keyboard and was focused with a validation message.
(-[WKContentView _elementDidBlur]):
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::getFocusedElementInformation):
2020-10-22 Carlos Garcia Campos <cgarcia@igalia.com>
WebDriver: handle key events with non-ASCII unicode code point
https://bugs.webkit.org/show_bug.cgi?id=217950
Reviewed by Carlos Alberto Lopez Perez.
We are assuming that key events are limited to ASCII, but they contain a unicode code point.
Fixes: imported/w3c/webdriver/tests/perform_actions/key_events.py::test_printable_key_sends_correct_events[\xe0-]
imported/w3c/webdriver/tests/perform_actions/key_events.py::test_printable_key_sends_correct_events[\u0416-]
imported/w3c/webdriver/tests/perform_actions/key_events.py::test_printable_key_sends_correct_events[\u2603-]
imported/w3c/webdriver/tests/perform_actions/key_events.py::test_printable_key_sends_correct_events[\uf6c2-]
* UIProcess/Automation/SimulatedInputDispatcher.h: Use UChar32 instead of char for CharKey.
* UIProcess/Automation/gtk/WebAutomationSessionGtk.cpp:
(WebKit::WebAutomationSession::platformSimulateKeyboardInteraction): Handle Charkey as a unicode code point.
* UIProcess/Automation/wpe/WebAutomationSessionWPE.cpp:
(WebKit::WebAutomationSession::platformSimulateKeyboardInteraction): Ditto.
2020-10-21 Ryosuke Niwa <rniwa@webkit.org>
Move function definitions in JSIPCBinding.h into cpp file
https://bugs.webkit.org/show_bug.cgi?id=218065
Reviewed by Wenson Hsieh.
Added JSIPCBinding.cpp and moved the code from JSIPCBinding.h to JSIPCBinding.cpp.
No new tests since there is no behavior change.
* Platform/IPC/JSIPCBinding.cpp: Added.
(IPC::jsValueForDecodedStringArgumentValue): Moved from .h.
(IPC::jsValueForDecodedArgumentValue<String>): Ditto.
(IPC::jsValueForDecodedArgumentValue<URL>): Ditto.
(IPC::jsValueForDecodedArgumentValue<RegistrableDomain>): Ditto.
(IPC::jsValueForDecodedNumericArgumentValue): Ditto.
(IPC::jsValueForDecodedArgumentValue<double>): Ditto.
(IPC::jsValueForDecodedArgumentValue<float>): Ditto.
(IPC::jsValueForDecodedArgumentValue<int8_t>): Ditto.
(IPC::jsValueForDecodedArgumentValue<int16_t>): Ditto.
(IPC::jsValueForDecodedArgumentValue<int32_t>): Ditto.
(IPC::jsValueForDecodedArgumentValue<int64_t>): Ditto.
(IPC::jsValueForDecodedArgumentValue<uint8_t>): Ditto.
(IPC::jsValueForDecodedArgumentValue<uint16_t>): Ditto.
(IPC::jsValueForDecodedArgumentValue<uint32_t>): Ditto.
(IPC::jsValueForDecodedArgumentValue<uint64_t>): Ditto.
(IPC::jsValueForDecodedArgumentRect): Ditto.
(IPC::jsValueForDecodedArgumentValue<IntRect>): Ditto.
(IPC::jsValueForDecodedArgumentValue<FloatRect>): Ditto.
* Platform/IPC/JSIPCBinding.h:
(IPC::jsValueForDecodedArgumentValue):
* Sources.txt:
* WebKit.xcodeproj/project.pbxproj:
2020-10-21 Ryosuke Niwa <rniwa@webkit.org>
IPC testing API should have the capability to observe messages being sent and received
https://bugs.webkit.org/show_bug.cgi?id=217870
Reviewed by Darin Adler.
Added IPC.addIncomingMessageListener and IPC.addOutgoingMessageListener which allows JavaScript
to observe IPC messages being sent or received by WebContent process. We use the generated code
added in r268503 to decode the IPC arguments.
Tests: TestWebKitAPI.IPCTestingAPI.CanInterceptAlert
TestWebKitAPI.IPCTestingAPI.CanInterceptHasStorageAccess
TestWebKitAPI.IPCTestingAPI.CanInterceptFindString
* Platform/IPC/Connection.cpp:
(IPC::Connection::sendMessage): Added the code to invoke MessageObserver::willSendMessage.
Also remove any stale MessageObserver as neded.
(IPC::Connection::addMessageObserver): Added.
(IPC::Connection::dispatchMessage): Added the code to invoke MessageObserver::didReceiveMessage.
Also remove any stale MessageObserver as neded.
* Platform/IPC/Connection.h:
(IPC::Connection::MessageObserver): Added. A pure virtual interface for observing IPC messages.
* Platform/IPC/JSIPCBinding.h:
(IPC::jsValueForDecodedStringArgumentValue): Extracted from jsValueForDecodedArgumentValue<String>.
Now takes the type name as an argument.
(IPC::jsValueForDecodedArgumentValue<URL>): Use "URL" as the type name.
(IPC::jsValueForDecodedArgumentValue<RegistrableDomain>): Use "RegistrableDomain" as the type name.
(IPC::jsValueForDecodedArgumentValue<OptionSet<U>>): Added. Specializations for OptionSet<U>
* WebProcess/WebPage/IPCTestingAPI.cpp:
(WebKit::IPCTestingAPI::JSMessageListener): Added. Implements IPC::MessageObserver.
(WebKit::IPCTestingAPI::JSIPC::staticFunctions):
(WebKit::IPCTestingAPI::createTypeError): Moved.
(WebKit::IPCTestingAPI::JSIPC::addMessageListener): Added.
(WebKit::IPCTestingAPI::JSIPC::addIncomingMessageListener): Added.
(WebKit::IPCTestingAPI::JSIPC::addOutgoingMessageListener): Added.
(WebKit::IPCTestingAPI::JSMessageListener::JSMessageListener): Added.
(WebKit::IPCTestingAPI::JSMessageListener::didReceiveMessage): Added.
(WebKit::IPCTestingAPI::JSMessageListener::willSendMessage): Added.
(WebKit::IPCTestingAPI::JSMessageListener::jsDescriptionFromDecoder): Added.
2020-10-21 Aditya Keerthi <akeerthi@apple.com>
Remove unused UIDocumentMenuViewController SPI declarations
https://bugs.webkit.org/show_bug.cgi?id=218053
Reviewed by Tim Horton.
UIDocumentMenuViewController is deprecated and its last
use was removed in r258092.
* Platform/spi/ios/UIKitSPI.h:
2020-10-21 Alex Christensen <achristensen@webkit.org>
Don't crash when deallocating WKWebView during TLS handshake
https://bugs.webkit.org/show_bug.cgi?id=218025
<rdar://problem/70225969>
Reviewed by Tim Horton.
NetworkProcessProxy::didReceiveAuthenticationChallenge would sometimes dereference an unchecked
Optional<SecurityOriginData> which would result in a null dereference crash. Also, sometimes
Connection::initializeSendSource would assert because it was trying to set up a cancel handler for
a send port that had not been successfully set up yet. I added a test that reproduces both of these
issues most of the time.
* Platform/IPC/cocoa/ConnectionCocoa.mm:
(IPC::Connection::initializeSendSource):
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::didReceiveAuthenticationChallenge):
2020-10-21 Carlos Garcia Campos <cgarcia@igalia.com>
WebDriver: add support for wheel actions
https://bugs.webkit.org/show_bug.cgi?id=217174
Reviewed by Brian Burg.
* UIProcess/Automation/Automation.json: Add scroll delta to action state.
* UIProcess/Automation/SimulatedInputDispatcher.cpp:
(WebKit::SimulatedInputSourceState::emptyStateForSourceType): Initialize scrollDelta for wheel actions.
(WebKit::SimulatedInputDispatcher::transitionInputSourceToState): Handle SimulatedInputSourceType::Wheel.
* UIProcess/Automation/SimulatedInputDispatcher.h:
* UIProcess/Automation/WebAutomationSession.cpp:
(WebKit::WebAutomationSession::WebAutomationSession): Add SimulatedInputSourceType::Wheel.
(WebKit::WebAutomationSession::terminate): Handle pending wheel events.
(WebKit::WebAutomationSession::willShowJavaScriptDialog): Ditto.
(WebKit::WebAutomationSession::wheelEventsFlushedForPage): Ditto.
(WebKit::WebAutomationSession::willClosePage): Ditto.
(WebKit::WebAutomationSession::isSimulatingUserInteraction const): Return true if there are pending wheel events too.
(WebKit::WebAutomationSession::simulateWheelInteraction): Handle the wheel action.
(WebKit::simulatedInputSourceTypeFromProtocolSourceType): Handle Inspector::Protocol::Automation::InputSourceType::Wheel.
(WebKit::WebAutomationSession::performInteractionSequence): Initialize the scroll delta for wheel action.
* UIProcess/Automation/WebAutomationSession.h:
* UIProcess/Automation/gtk/WebAutomationSessionGtk.cpp:
(WebKit::WebAutomationSession::platformSimulateWheelInteraction): Synthesize a wheel event.
* UIProcess/Automation/wpe/WebAutomationSessionWPE.cpp:
(WebKit::WebAutomationSession::platformSimulateWheelInteraction): Ditto.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::isProcessingWheelEvents const): Return whether page has pending wheel events.
(WebKit::WebPageProxy::didReceiveEvent): Notify automation that pending wheel events have been processed.
* UIProcess/WebPageProxy.h:
* config.h:
2020-10-20 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process] Some WebAudio layout tests generate strange noises
https://bugs.webkit.org/show_bug.cgi?id=217921
Reviewed by Eric Carlson.
RemoteAudioDestination::render() should not return `noErr` unless we can provide the requested
samples to the provided AudioBufferList. Otherwise, the audio output unit of CoreAudio will output
the samples in the AudioBufferList, which might be invalid data at the beginning of a rendering.
We have observed that happens in some layout tests and some WebAudio example pages.
Currently, RemoteAudioDestination::render() always returns `noErr` in the render thread (immediately),
but the AudioBufferList (ioData) is updated in the main thread (later). This patch fixes that by only
setting the bounds of CARingBuffer in the completion handler of sendWithAsyncReply() in the main thread,
and fetching AudioBuffer(s) from the CARingBuffer in the render thread. Also, RemoteAudioDestination
tracks the progress of fetching, so RemoteAudioDestinationProxy does not need to send `startFrame`
and `numberOfFramesToRender` to RemoteAudioDestination in response to a buffer request.
* GPUProcess/media/RemoteAudioDestinationManager.cpp:
(WebKit::RemoteAudioDestination::audioSamplesStorageChanged):
(WebKit::RemoteAudioDestination::render): Only return `noErr` if the function renders the requested
sample successfully.
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::requestBuffer): Remove unused parameters.
(WebKit::RemoteAudioDestinationProxy::renderOnRenderingThead): Ditto.
* WebProcess/GPU/media/RemoteAudioDestinationProxy.h: Ditto.
* WebProcess/GPU/media/RemoteAudioDestinationProxy.messages.in: Ditto.
2020-10-20 Chris Dumez <cdumez@apple.com>
Drop legacy code using AssertionServices
https://bugs.webkit.org/show_bug.cgi?id=217787
<rdar://problem/70160942>
Reviewed by Darin Adler.
Drop legacy code using AssertionServices as this framework is deprecated in favor of
RunningBoard. All our code has been ported to RunningBoard already.
* Configurations/WebKit.xcconfig:
* Platform/spi/ios/AssertionServicesSPI.h: Removed.
* Shared/Cocoa/ProcessTaskStateObserver.h: Removed.
* Shared/Cocoa/ProcessTaskStateObserver.mm: Removed.
* UIProcess/ApplicationStateTracker.mm:
* UIProcess/ProcessAssertion.h:
* UIProcess/ios/PageClientImplIOS.mm:
* UIProcess/ios/ProcessAssertionIOS.mm:
(-[WKProcessAssertionBackgroundTaskManager init]):
(-[WKProcessAssertionBackgroundTaskManager _hasBackgroundTask]):
(-[WKProcessAssertionBackgroundTaskManager _updateBackgroundTask]):
(-[WKProcessAssertionBackgroundTaskManager assertion:didInvalidateWithError:]):
(-[WKProcessAssertionBackgroundTaskManager _releaseBackgroundTask]):
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/cocoa/WebProcessCocoa.mm:
2020-10-20 Sam Weinig <weinig@apple.com>
Tweak WebPreferences*.yaml "exposed" key to only indicate that the key should not be changeable by the frontend
https://bugs.webkit.org/show_bug.cgi?id=217918
Reviewed by Darin Adler.
Update to use new "exposed" prefix bindings and use default value explicitly
when updating settings for non-exposed preferences.
* Scripts/PreferencesTemplates/WebPageUpdatePreferences.cpp.erb:
* Scripts/PreferencesTemplates/WebPreferencesDefinitions.h.erb:
* Scripts/PreferencesTemplates/WebPreferencesExperimentalFeatures.cpp.erb:
* Scripts/PreferencesTemplates/WebPreferencesGetterSetters.cpp.erb:
* Scripts/PreferencesTemplates/WebPreferencesInternalDebugFeatures.cpp.erb:
* Scripts/PreferencesTemplates/WebPreferencesKeys.cpp.erb:
* Scripts/PreferencesTemplates/WebPreferencesKeys.h.erb:
* Scripts/PreferencesTemplates/WebPreferencesStoreDefaultsMap.cpp.erb:
2020-10-20 Carlos Garcia Campos <cgarcia@igalia.com>
WebDriver: add support for right variations of virtual keys
https://bugs.webkit.org/show_bug.cgi?id=217819
Reviewed by Brian Burg.
Instead of using a HashSet the pressed virtual keys are now stored in a HashMap where the key is the normalized
virtual key and the value is the raw virtual key.
* UIProcess/Automation/Automation.json:
* UIProcess/Automation/SimulatedInputDispatcher.cpp:
(WebKit::SimulatedInputDispatcher::transitionInputSourceToState):
* UIProcess/Automation/SimulatedInputDispatcher.h:
* UIProcess/Automation/WebAutomationSession.cpp:
(WebKit::normalizedVirtualKey):
(WebKit::WebAutomationSession::performInteractionSequence):
* UIProcess/Automation/gtk/WebAutomationSessionGtk.cpp:
(WebKit::keyCodeForVirtualKey):
(WebKit::modifiersForKeyCode):
* UIProcess/Automation/mac/WebAutomationSessionMac.mm:
(WebKit::keyCodeForVirtualKey):
* UIProcess/Automation/wpe/WebAutomationSessionWPE.cpp:
(WebKit::keyCodeForVirtualKey):
(WebKit::modifiersForKeyCode):
2020-10-19 Peng Liu <peng.liu6@apple.com>
Two feature flag names are incorrect
https://bugs.webkit.org/show_bug.cgi?id=217913
Reviewed by Eric Carlson.
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultCaptureAudioInGPUProcessEnabled): The correct feature flag
is gpu_process_media.
(WebKit::defaultVP9SWDecoderEnabledOnBattery): The correct feature flag
is sw_vp9_decoder_on_battery.
2020-10-19 Chris Dumez <cdumez@apple.com>
[GPU Process] RemoteAudioDestination::render() should not need to dispatch to the main thread to do IPC
https://bugs.webkit.org/show_bug.cgi?id=217920
Reviewed by Alex Christensen.
RemoteAudioDestination::render() should not need to dispatch to the main thread to do IPC as this is bad
for performance and IPC::Connection::sendWithAsyncReply() already does locking internally for thread-safety.
The only thing that was preventing us from calling sendWithAsyncReply() from the audio rendering thread was
a threading assertion in CompletionHandler that made sure that the completion handler was always called on
the thread the completion handler was constructed on. To avoid this issue, I added a paramater to the
CompletionHandler constructor allowing the caller to indicate it expects the completion handler to get called
on the main thread. I am now using this new flag in sendWithAsyncReply() and at its call site in
RemoteAudioDestination::render().
No new tests, no Web-facing behavior change.
* GPUProcess/media/RemoteAudioDestinationManager.cpp:
(WebKit::RemoteAudioDestination::render):
* Platform/IPC/Connection.h:
(IPC::Connection::sendWithAsyncReply):
2020-10-19 Said Abou-Hallawa <sabouhallawa@apple.com>
[GPU Process] Clean the DisplayList recording of the NativeImage
https://bugs.webkit.org/show_bug.cgi?id=217573
Reviewed by Simon Fraser.
Support encoding and decoding NativeImagePtr directly.
* Scripts/webkit/messages.py:
* Shared/WebCoreArgumentCoders.cpp:
(IPC::encodeNativeImage):
(IPC::encodeOptionalNativeImage):
(IPC::ArgumentCoder<NativeImagePtr>::encode):
(IPC::ArgumentCoder<NativeImagePtr>::decode):
(IPC::ArgumentCoder<NativeImageHandle>::encode): Deleted.
(IPC::ArgumentCoder<NativeImageHandle>::decode): Deleted.
* Shared/WebCoreArgumentCoders.h:
* Shared/cg/ShareableBitmapCG.cpp:
(WebKit::ShareableBitmap::paint):
2020-10-18 Fujii Hironori <Hironori.Fujii@sony.com>
[WinCairo] sluggish page scrolling for wheel events since r268499
https://bugs.webkit.org/show_bug.cgi?id=217850
Reviewed by Simon Fraser.
* WebProcess/WebPage/EventDispatcher.cpp:
(WebKit::EventDispatcher::wheelEvent): Include
WheelEventProcessingSteps::MainThreadForScrolling to
processingSteps for !ENABLE(SCROLLING_THREAD).
2020-10-17 David Kilzer <ddkilzer@apple.com>
REGRESSION (r191637): Fix -Wmisleading-indentation warnings in open source clang
<https://webkit.org/b/217882>
Reviewed by Tim Horton.
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::accessibilityAttributeValue):
- Fix indentation of source to match what it is doing.
2020-10-16 Said Abou-Hallawa <sabouhallawa@apple.com>
[GPU Process] Introduce RemoteResourceCacheProxy to manage the resources in the GPU Process
https://bugs.webkit.org/show_bug.cgi?id=217554
Reviewed by Simon Fraser.
RemoteResourceCache will be responsible for caching the remote resources.
RemoteRenderingBackend will delegate all the resource caching to its
RemoteResourceCache.
We can get rid of RemoteImageBufferMessageHandler because all it does is
is sending messages to WebProcess. Removing it will simplify the interface
of RemoteImageBuffer. Sending the messages will be moved to RemoteRenderingBackend.
Rename flushDrawingContext() to flushDisplayList() because the name is
confusing. This function flushes a DisplayList to the context. No
DrawingContext is involved here and there is another function with the
same name which does not take any argument.
Rename ImageBufferFlushIdentifier to DisplayListFlushIdentifier to make
more general name and be ready for GPU DOM rendering.
Rename the RemoteRenderingBackendProxy messages: CreateImageBufferBackend
and CommitImageBufferFlushContext to be ImageBufferBackendWasCreated and
FlushDisplayListWasCommitted. We need to include passive verbs in them to
reflect that the actions were taken place in the GPU Process. They are
sent only to inform the Web Process with the current state.
* GPUProcess/graphics/RemoteImageBuffer.h:
(WebKit::RemoteImageBuffer::RemoteImageBuffer):
(WebKit::RemoteImageBuffer::apply):
* GPUProcess/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::gpuConnectionToWebProcess const):
(WebKit::RemoteRenderingBackend::imageBufferBackendWasCreated):
(WebKit::RemoteRenderingBackend::flushDisplayListWasCommitted):
(WebKit::RemoteRenderingBackend::createImageBuffer):
(WebKit::RemoteRenderingBackend::flushDisplayList):
(WebKit::RemoteRenderingBackend::flushDisplayListAndCommit):
(WebKit::RemoteRenderingBackend::getImageData):
We should call the completion handler even if the ImageBuffer was not found.
(WebKit::RemoteRenderingBackend::releaseRemoteResource):
(WebKit::RemoteRenderingBackend::flushImageBufferDrawingContext): Deleted.
(WebKit::RemoteRenderingBackend::flushImageBufferDrawingContextAndCommit): Deleted.
* GPUProcess/graphics/RemoteRenderingBackend.h:
(WebKit::RemoteRenderingBackend::renderingBackendIdentifier const): Deleted.
* GPUProcess/graphics/RemoteRenderingBackend.messages.in:
* GPUProcess/graphics/RemoteResourceCache.cpp: Renamed from Source/WebKit/GPUProcess/graphics/RemoteImageBufferMessageHandler.cpp.
(WebKit::RemoteResourceCache::cacheImageBuffer):
(WebKit::RemoteResourceCache::cachedImageBuffer):
(WebKit::RemoteResourceCache::releaseRemoteResource):
* GPUProcess/graphics/RemoteResourceCache.h: Renamed from Source/WebKit/GPUProcess/graphics/RemoteImageBufferMessageHandler.h.
* Scripts/webkit/messages.py:
* Sources.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/graphics/DisplayListFlushIdentifier.h: Renamed from Source/WebKit/WebProcess/GPU/graphics/ImageBufferFlushIdentifier.h.
* WebProcess/GPU/graphics/RemoteImageBufferMessageHandlerProxy.cpp:
(WebKit::RemoteImageBufferMessageHandlerProxy::waitForImageBufferBackendWasCreated):
(WebKit::RemoteImageBufferMessageHandlerProxy::waitForFlushDisplayListWasCommitted):
(WebKit::RemoteImageBufferMessageHandlerProxy::flushDisplayList):
(WebKit::RemoteImageBufferMessageHandlerProxy::flushDisplayListAndWaitCommit):
(WebKit::RemoteImageBufferMessageHandlerProxy::commitFlushContext):
(WebKit::RemoteImageBufferMessageHandlerProxy::waitForCreateImageBufferBackend): Deleted.
(WebKit::RemoteImageBufferMessageHandlerProxy::waitForCommitImageBufferFlushContext): Deleted.
(WebKit::RemoteImageBufferMessageHandlerProxy::flushDrawingContext): Deleted.
(WebKit::RemoteImageBufferMessageHandlerProxy::flushDrawingContextAndWaitCommit): Deleted.
* WebProcess/GPU/graphics/RemoteImageBufferMessageHandlerProxy.h:
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h:
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::waitForImageBufferBackendWasCreated):
(WebKit::RemoteRenderingBackendProxy::waitForFlushDisplayListWasCommitted):
(WebKit::RemoteRenderingBackendProxy::imageBufferBackendWasCreated):
(WebKit::RemoteRenderingBackendProxy::flushDisplayListWasCommitted):
(WebKit::RemoteRenderingBackendProxy::waitForCreateImageBufferBackend): Deleted.
(WebKit::RemoteRenderingBackendProxy::waitForCommitImageBufferFlushContext): Deleted.
(WebKit::RemoteRenderingBackendProxy::createImageBufferBackend): Deleted.
(WebKit::RemoteRenderingBackendProxy::commitImageBufferFlushContext): Deleted.
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h:
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.messages.in:
2020-10-16 Wenson Hsieh <wenson_hsieh@apple.com>
Add system trace points for flushing remote image buffers
https://bugs.webkit.org/show_bug.cgi?id=217853
Reviewed by Simon Fraser.
Add trace points to capture drawing context flushes. This encapsulates time spent encoding display lists in
preparation for IPC.
* WebProcess/GPU/graphics/RemoteImageBufferMessageHandlerProxy.cpp:
(WebKit::RemoteImageBufferMessageHandlerProxy::flushDrawingContext):
(WebKit::RemoteImageBufferMessageHandlerProxy::flushDrawingContextAndWaitCommit):
2020-10-16 Devin Rousso <drousso@apple.com>
Web Inspector: rename `Highlight` so it doesn't conflict with CSS `Highlight`
https://bugs.webkit.org/show_bug.cgi?id=217858
Reviewed by Megan Gardner.
No new tests as there is no behavior change since this is purely a rename.
* Scripts/webkit/messages.py:
* Shared/WebCoreArgumentCoders.h:
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<InspectorOverlay::Highlight>::encode): Added.
(IPC::ArgumentCoder<InspectorOverlay::Highlight>::decode): Added.
(IPC::ArgumentCoder<Highlight>::encode): Deleted.
(IPC::ArgumentCoder<Highlight>::decode): Deleted.
* WebProcess/Inspector/WebInspectorClient.cpp:
(WebKit::WebInspectorClient::highlight):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::showInspectorHighlight):
* UIProcess/ios/WKContentView.h:
* UIProcess/ios/WKContentView.mm:
(-[WKContentView _showInspectorHighlight:]):
* UIProcess/WebPageProxy.messages.in:
* UIProcess/WebPageProxy.h:
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::showInspectorHighlight):
* UIProcess/PageClient.h:
* UIProcess/ios/PageClientImplIOS.h:
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::showInspectorHighlight):
* UIProcess/Inspector/ios/WKInspectorHighlightView.h:
* UIProcess/Inspector/ios/WKInspectorHighlightView.mm:
(-[WKInspectorHighlightView _layoutForNodeHighlight:offset:]):
(-[WKInspectorHighlightView _layoutForNodeListHighlight:]):
(-[WKInspectorHighlightView _layoutForRectsHighlight:]):
(-[WKInspectorHighlightView update:]):
2020-10-16 Ryosuke Niwa <rniwa@webkit.org>
IPC testing API should expose whether a given IPC message has sync reply or not
https://bugs.webkit.org/show_bug.cgi?id=217861
Reviewed by Darin Adler.
This patch adds IPC.messages.*.isSync to indicate whether a given IPC message has a sync reply or not.
Test: TestWebKitAPI.IPCTestingAPI.AlertIsSyncMessage
* Platform/IPC/MessageArgumentDescriptions.h:
* Scripts/webkit/messages.py:
(generate_js_value_conversion_function): Fixed a typo.
(generate_js_argument_descriptions): Ditto.
(generate_message_argument_description_implementation): Generate messageIsSync function.
* WebProcess/WebPage/IPCTestingAPI.cpp:
(WebKit::JSIPC::wrapperClass): Updated the class name to match the implementation.
(WebKit::IPCTestingAPI::JSIPC::messages): Added isSync as a boolean property on message description.
2020-10-16 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process] Use CARingBuffer to transfer audio data of an audio destination from a web process to the GPU Process
https://bugs.webkit.org/show_bug.cgi?id=217715
Reviewed by Jer Noble.
The current implementation of RemoteAudioDestination essentially runs the
AudioDestinationCocoa in the GPU process, which is not good because we are adding
AudioWorklet stuff into AudioDestinationCocoa, but the AudioWorklet stuff is
irrelevant to the GPU process.
What really needs to run in the GPU process (for Cocoa platforms at least) is the
audio output unit. This patch adds WebCore::AudioOutputUnitAdaptor and
WebCore::AudioUnitRenderer to implement that. With this patch, RemoteAudioDestination
in the GPU process is an AudioUnitRenderer and it embeds an AudioOutputUnitAdaptor
like AudioDestinationCocoa does in the Web process. Essentially, we run AudioOutputUnitAdaptor
remotely from the AudioDestinationCocoa's perspective.
Also, in the current implementation, the RemoteAudioDestinationProxy in a Web
process transfers audio data to the RemoteAudioDestination in the GPU process
with IPC messages (RemoteAudioBusData). This is not efficient because of the
large number of IPC messages for audio samples.
With this patch, RemoteAudioDestinationProxy (Web process) transfers audio samples
to RemoteAudioDestination(GPU process) with a CARingBuffer providing shared
buffers for the Web process and the GPU process.
* GPUProcess/media/RemoteAudioDestinationManager.cpp:
(WebKit::RemoteAudioDestination::create):
(WebKit::RemoteAudioDestination::~RemoteAudioDestination):
(WebKit::RemoteAudioDestination::audioSamplesStorageChanged):
(WebKit::RemoteAudioDestination::start):
(WebKit::RemoteAudioDestination::stop):
(WebKit::RemoteAudioDestination::isPlaying const):
(WebKit::RemoteAudioDestination::RemoteAudioDestination):
(WebKit::RemoteAudioDestination::storage):
(WebKit::RemoteAudioDestination::render):
(WebKit::RemoteAudioDestinationManager::createAudioDestination):
(WebKit::RemoteAudioDestinationManager::audioSamplesStorageChanged):
(WebKit::RemoteAudioDestination::isPlaying): Deleted.
(WebKit::RemoteAudioDestination::framesPerBuffer const): Deleted.
* GPUProcess/media/RemoteAudioDestinationManager.h:
(WebKit::RemoteAudioDestinationManager::didReceiveSyncMessageFromWebProcess):
* GPUProcess/media/RemoteAudioDestinationManager.messages.in:
Add a message AudioSamplesStorageChanged to support using CARingBuffer to
transmit audio samples to the GPU process.
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/media/RemoteAudioBusData.h: Removed.
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::RemoteAudioDestinationProxy):
(WebKit::RemoteAudioDestinationProxy::start):
(WebKit::RemoteAudioDestinationProxy::stop):
(WebKit::RemoteAudioDestinationProxy::requestBuffer):
(WebKit::RemoteAudioDestinationProxy::renderOnRenderingThead):
(WebKit::RemoteAudioDestinationProxy::storageChanged):
(WebKit::RemoteAudioDestinationProxy::renderBuffer): Deleted.
(WebKit::RemoteAudioDestinationProxy::didChangeIsPlaying): Deleted.
* WebProcess/GPU/media/RemoteAudioDestinationProxy.h:
RemoteAudioDestinationProxy is a subclass of AudioDestinationCocoa. It pulls audio
samples from the WebAudio side and use a CARingBuffer to push audio samples to
the audio output unit (managed by RemoteAudioDestination) in the GPU process.
* WebProcess/GPU/media/RemoteAudioDestinationProxy.messages.in:
Add a message RequestBuffer, which can be used by RemoteAudioDestination to ask
for (pull) audio samples.
* WebProcess/GPU/webrtc/AudioMediaStreamTrackRenderer.cpp:
(WebKit::AudioMediaStreamTrackRenderer::pushSamples): A minor fix to get rid of
a type conversion.
2020-10-16 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Add an experimental feature flag: WebAuthenticationModernEnabled
https://bugs.webkit.org/show_bug.cgi?id=217843
<rdar://problem/70384187>
Reviewed by Brent Fulgham.
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetWebAuthenticationModernEnabled):
(WKPreferencesGetWebAuthenticationModernEnabled):
* UIProcess/API/C/WKPreferencesRefPrivate.h:
2020-10-16 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Remove experimental feature flag: WebAuthenticationLocalAuthenticatorEnabled
https://bugs.webkit.org/show_bug.cgi?id=217796
<rdar://problem/70358912>
Reviewed by Brent Fulgham.
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetWebAuthenticationLocalAuthenticatorEnabled): Deleted.
(WKPreferencesGetWebAuthenticationLocalAuthenticatorEnabled): Deleted.
* UIProcess/API/C/WKPreferencesRefPrivate.h:
* UIProcess/WebAuthentication/AuthenticatorManager.cpp:
(WebKit::AuthenticatorManager::filterTransports const):
(WebKit::WebCore::isFeatureEnabled): Deleted.
2020-10-16 Said Abou-Hallawa <sabouhallawa@apple.com>
[GPU Process] Exchange RemoteImageBuffer and RemoteImageBufferProxy
https://bugs.webkit.org/show_bug.cgi?id=217809
Reviewed by Simon Fraser.
The names of the GPU rendering classes have been confusing. RemoteImageBuffer
should be the real object in the GPU Process because it owns the backend.
And RemoteImageBufferProxy should be a proxy to this object in the Web
Process. So we are going to exchange the following classes:
-- RemoteImageBuffer <-> RemoteImageBufferProxy
-- RemoteRenderingBackend <-> RemoteRenderingBackendProxy
-- RemoteImageBufferMessageHandler <-> RemoteImageBufferMessageHandlerProxy
* CMakeLists.txt:
* DerivedSources-input.xcfilelist:
* DerivedSources.make:
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::createRenderingBackend):
(WebKit::GPUConnectionToWebProcess::releaseRenderingBackend):
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/graphics/PlatformRemoteImageBuffer.h: Renamed from Source/WebKit/WebProcess/GPU/graphics/PlatformRemoteImageBuffer.h.
* GPUProcess/graphics/RemoteImageBuffer.h: Renamed from Source/WebKit/GPUProcess/graphics/RemoteImageBufferProxy.h.
(WebKit::RemoteImageBuffer::create):
(WebKit::RemoteImageBuffer::RemoteImageBuffer):
(WebKit::RemoteImageBuffer::~RemoteImageBuffer):
(WebKit::RemoteImageBuffer::apply):
* GPUProcess/graphics/RemoteImageBufferMessageHandler.cpp: Renamed from Source/WebKit/GPUProcess/graphics/RemoteImageBufferMessageHandlerProxy.cpp.
(WebKit::RemoteImageBufferMessageHandler::RemoteImageBufferMessageHandler):
(WebKit::RemoteImageBufferMessageHandler::createBackend):
(WebKit::RemoteImageBufferMessageHandler::commitFlushContext):
* GPUProcess/graphics/RemoteImageBufferMessageHandler.h: Renamed from Source/WebKit/GPUProcess/graphics/RemoteImageBufferMessageHandlerProxy.h.
(WebKit::RemoteImageBufferMessageHandler::backend):
* GPUProcess/graphics/RemoteRenderingBackend.cpp: Renamed from Source/WebKit/GPUProcess/graphics/RemoteRenderingBackendProxy.cpp.
(WebKit::RemoteRenderingBackend::create):
(WebKit::RemoteRenderingBackend::RemoteRenderingBackend):
(WebKit::RemoteRenderingBackend::~RemoteRenderingBackend):
(WebKit::RemoteRenderingBackend::messageSenderConnection const):
(WebKit::RemoteRenderingBackend::messageSenderDestinationID const):
(WebKit::RemoteRenderingBackend::createImageBuffer):
(WebKit::RemoteRenderingBackend::releaseRemoteResource):
(WebKit::RemoteRenderingBackend::getImageData):
(WebKit::RemoteRenderingBackend::flushImageBufferDrawingContext):
(WebKit::RemoteRenderingBackend::flushImageBufferDrawingContextAndCommit):
(WebKit::RemoteRenderingBackend::gpuConnectionToWebProcess const):
* GPUProcess/graphics/RemoteRenderingBackend.h: Renamed from Source/WebKit/GPUProcess/graphics/RemoteRenderingBackendProxy.h.
(WebKit::RemoteRenderingBackend::renderingBackendIdentifier const):
* GPUProcess/graphics/RemoteRenderingBackend.messages.in: Renamed from Source/WebKit/GPUProcess/graphics/RemoteRenderingBackendProxy.messages.in.
* GPUProcess/media/RemoteCDMInstanceSessionProxy.cpp:
* Sources.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/graphics/PlatformRemoteImageBufferProxy.h: Renamed from Source/WebKit/GPUProcess/graphics/PlatformRemoteImageBufferProxy.h.
* WebProcess/GPU/graphics/RemoteImageBufferMessageHandler.cpp: Removed.
* WebProcess/GPU/graphics/RemoteImageBufferMessageHandlerProxy.cpp: Added.
(WebKit::RemoteImageBufferMessageHandlerProxy::RemoteImageBufferMessageHandlerProxy):
(WebKit::RemoteImageBufferMessageHandlerProxy::~RemoteImageBufferMessageHandlerProxy):
(WebKit::RemoteImageBufferMessageHandlerProxy::getImageData const):
(WebKit::RemoteImageBufferMessageHandlerProxy::waitForCreateImageBufferBackend):
(WebKit::RemoteImageBufferMessageHandlerProxy::waitForCommitImageBufferFlushContext):
(WebKit::RemoteImageBufferMessageHandlerProxy::flushDrawingContext):
(WebKit::RemoteImageBufferMessageHandlerProxy::flushDrawingContextAndWaitCommit):
(WebKit::RemoteImageBufferMessageHandlerProxy::commitFlushContext):
* WebProcess/GPU/graphics/RemoteImageBufferMessageHandlerProxy.h: Renamed from Source/WebKit/WebProcess/GPU/graphics/RemoteImageBufferMessageHandler.h.
(WebKit::RemoteImageBufferMessageHandlerProxy::remoteResourceIdentifier const):
(WebKit::RemoteImageBufferMessageHandlerProxy::isPendingFlush const):
* WebProcess/GPU/graphics/RemoteImageBufferProxy.h: Renamed from Source/WebKit/WebProcess/GPU/graphics/RemoteImageBuffer.h.
(WebKit::RemoteImageBufferProxy::create):
(WebKit::RemoteImageBufferProxy::~RemoteImageBufferProxy):
(WebKit::RemoteImageBufferProxy::RemoteImageBufferProxy):
(WebKit::RemoteImageBufferProxy::putImageData):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.cpp: Renamed from Source/WebKit/WebProcess/GPU/graphics/RemoteRenderingBackend.cpp.
(WebKit::RemoteRenderingBackendProxy::create):
(WebKit::RemoteRenderingBackendProxy::RemoteRenderingBackendProxy):
(WebKit::RemoteRenderingBackendProxy::~RemoteRenderingBackendProxy):
(WebKit::RemoteRenderingBackendProxy::messageSenderConnection const):
(WebKit::RemoteRenderingBackendProxy::messageSenderDestinationID const):
(WebKit::RemoteRenderingBackendProxy::waitForCreateImageBufferBackend):
(WebKit::RemoteRenderingBackendProxy::waitForCommitImageBufferFlushContext):
(WebKit::RemoteRenderingBackendProxy::createImageBuffer):
(WebKit::RemoteRenderingBackendProxy::releaseRemoteResource):
(WebKit::RemoteRenderingBackendProxy::createImageBufferBackend):
(WebKit::RemoteRenderingBackendProxy::commitImageBufferFlushContext):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.h: Renamed from Source/WebKit/WebProcess/GPU/graphics/RemoteRenderingBackend.h.
(WebKit::RemoteRenderingBackendProxy::renderingBackendIdentifier const):
* WebProcess/GPU/graphics/RemoteRenderingBackendProxy.messages.in: Renamed from Source/WebKit/WebProcess/GPU/graphics/RemoteRenderingBackend.messages.in.
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::ensureRemoteRenderingBackendProxy const):
(WebKit::WebChromeClient::createImageBuffer const):
(WebKit::WebChromeClient::ensureRemoteRenderingBackend const): Deleted.
* WebProcess/WebCoreSupport/WebChromeClient.h:
2020-10-16 Eric Carlson <eric.carlson@apple.com>
[GPU Process] Implement VideoPlaybackMetrics
https://bugs.webkit.org/show_bug.cgi?id=217439
<rdar://problem/60179370>
Reviewed by Jer Noble.
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
(WebKit::RemoteMediaPlayerProxy::outOfBandTrackSources):
(WebKit::RemoteMediaPlayerProxy::updateCachedState):
(WebKit::RemoteMediaPlayerProxy::setShouldUpdatePlaybackMetrics):
* GPUProcess/media/RemoteMediaPlayerProxy.h:
* GPUProcess/media/RemoteMediaPlayerProxy.messages.in:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::videoPlaybackQualityMetrics):
* WebProcess/GPU/media/MediaPlayerPrivateRemote.h:
* WebProcess/GPU/media/RemoteMediaPlayerState.h:
(WebKit::RemoteMediaPlayerState::encode const):
(WebKit::RemoteMediaPlayerState::decode):
2020-10-16 Eric Carlson <eric.carlson@apple.com>
[GPU Process] Implement mediaPlayerGetRawCookies
https://bugs.webkit.org/show_bug.cgi?id=217739
<rdar://problem/70313370>
Reviewed by Youenn Fablet.
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
(WebKit::RemoteMediaPlayerProxy::mediaPlayerGetRawCookies const):
* GPUProcess/media/RemoteMediaPlayerProxy.h:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::getRawCookies const):
* WebProcess/GPU/media/MediaPlayerPrivateRemote.h:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.messages.in:
2020-10-16 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Implement a dummy WebAuthnProcess
https://bugs.webkit.org/show_bug.cgi?id=217559
<rdar://problem/70168749>
Reviewed by Brent Fulgham.
Previous patch r268248 introduces a new daemon to handle all WebAuthn requests. However, after more deeper investigations, the daemon
story is not very compelling for WebKit. Here are the two major road blockers:
1) The major one is on macOS. New daemons cannot be installed without a system update. It means relocatable Safari (e.g. STP)
and staged Safari on downlevel OSes will not get this feature. And relocatable STP can only utilize the system daemon. This
seems a bummer especially for the STP case given it is heavily used to test latest web features.
2) Additional plumbing is needed for testing and engineering builds. This problem is specific to WebKit and is on both macOS and iOS.
WebKit doesn't build roots to test engineering builds, which is required to update the plist on the above location. Hence, a separate
configuration is needed to generate a different plist that points to the right engineering binary and additional scripts will be needed
to ask launchd to load the plist before running any tests. Some extra exercises are probably needed to not confuse launchd about the
engineering binary and system binary as well.
Given the above constraints, a new architecture is proposed to make the daemon a WebKit XPC service/process instead and leave the launchd
event handler to a new Safari daemon. The launchd related feature is only on iOS and therefore it really doesn't make sense to overcome
all the above macOS constraints. Having different architectures on iOS (a daemon) and on macOS (a XPC service) accordingly will introduce
tons of overhead for maintainance as well. That's why relying on a Safari daemon to do the job is the most reasonable option.
This patch therefore focuses on making the WebAuthn process happen and removes the WebAuthn daemon.
* CMakeLists.txt:
* Configurations/WebAuthnService.xcconfig: Renamed from Source/WebKit/Configurations/WebAuthenticationAgent.xcconfig.
* Configurations/WebKit.xcconfig:
* Daemons/WebAuthenticationAgent/WebAuthenticationAgent.entitlements: Removed.
* Daemons/WebAuthenticationAgent/com.apple.webkit.WebAuthenticationAgent.plist: Removed.
* DerivedSources-input.xcfilelist:
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
Paperwork for building the new WebAuthn process.
* Platform/ExtraPrivateSymbolsForTAPI.h:
* Shared/AuxiliaryProcess.h:
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceEntryPoint.h:
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceMain.mm:
(WebKit::XPCServiceEventHandler):
* Shared/mac/AuxiliaryProcessMac.mm:
(WebKit::processStorageClass):
(WebKit::sandboxDirectory):
Paperwork for introducing the new WebAuthn process. Mostly copied from GPU process.
* Sources.txt:
* SourcesCocoa.txt:
Paperwork for building the new WebAuthn process.
* UIProcess/AuxiliaryProcessProxy.cpp:
(WebKit::AuxiliaryProcessProxy::getLaunchOptions):
* UIProcess/Launcher/ProcessLauncher.h:
* UIProcess/Launcher/mac/ProcessLauncherMac.mm:
(WebKit::serviceName):
(WebKit::shouldLeakBoost):
* UIProcess/WebAuthentication/WebAuthnProcessProxy.cpp: Added.
(WebKit::WebAuthnProcessProxy::singleton):
(WebKit::WebAuthnProcessProxy::WebAuthnProcessProxy):
(WebKit::WebAuthnProcessProxy::getLaunchOptions):
(WebKit::WebAuthnProcessProxy::connectionWillOpen):
(WebKit::WebAuthnProcessProxy::processWillShutDown):
(WebKit::WebAuthnProcessProxy::getWebAuthnProcessConnection):
(WebKit::WebAuthnProcessProxy::webAuthnProcessCrashed):
(WebKit::WebAuthnProcessProxy::didClose):
(WebKit::WebAuthnProcessProxy::didReceiveInvalidMessage):
(WebKit::WebAuthnProcessProxy::didFinishLaunching):
(WebKit::WebAuthnProcessProxy::updateProcessAssertion):
* UIProcess/WebAuthentication/WebAuthnProcessProxy.h: Added.
* UIProcess/WebAuthentication/WebAuthnProcessProxy.messages.in: Added.
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::getWebAuthnProcessConnection):
* UIProcess/WebProcessPool.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::getWebAuthnProcessConnection):
* UIProcess/WebProcessProxy.h:
* UIProcess/WebProcessProxy.messages.in:
* WebAuthnProcess/EntryPoint/Cocoa/XPCService/WebAuthnService/Info-OSX.plist: Copied from Source/WebKit/Daemons/WebAuthenticationAgent/Info.plist.
* WebAuthnProcess/EntryPoint/Cocoa/XPCService/WebAuthnService/Info-iOS.plist: Renamed from Source/WebKit/Daemons/WebAuthenticationAgent/Info.plist.
* WebAuthnProcess/EntryPoint/Cocoa/XPCService/WebAuthnServiceEntryPoint.mm: Copied from Source/WebKit/Daemons/WebAuthenticationAgent/main.mm.
(WebKit::WebAuthnServiceInitializerDelegate::WebAuthnServiceInitializerDelegate):
(WebKit::initializeAuxiliaryProcess<WebAuthnProcess>):
(WEBAUTHEN_SERVICE_INITIALIZER):
* WebAuthnProcess/WebAuthnConnectionToWebProcess.cpp: Added.
(WebKit::WebAuthnConnectionToWebProcess::create):
(WebKit::WebAuthnConnectionToWebProcess::WebAuthnConnectionToWebProcess):
(WebKit::WebAuthnConnectionToWebProcess::~WebAuthnConnectionToWebProcess):
(WebKit::WebAuthnConnectionToWebProcess::didClose):
(WebKit::WebAuthnConnectionToWebProcess::didReceiveInvalidMessage):
* WebAuthnProcess/WebAuthnConnectionToWebProcess.h: Copied from Source/WebKit/Daemons/WebAuthenticationAgent/main.mm.
(WebKit::WebAuthnConnectionToWebProcess::connection):
(WebKit::WebAuthnConnectionToWebProcess::WebAuthnProcessProcess):
(WebKit::WebAuthnConnectionToWebProcess::webProcessIdentifier const):
* WebAuthnProcess/WebAuthnConnectionToWebProcess.messages.in: Added.
* WebAuthnProcess/WebAuthnProcess.cpp: Added.
(WebKit::WebAuthnProcess::WebAuthnProcess):
(WebKit::WebAuthnProcess::~WebAuthnProcess):
(WebKit::WebAuthnProcess::createWebAuthnConnectionToWebProcess):
(WebKit::WebAuthnProcess::removeWebAuthnConnectionToWebProcess):
(WebKit::WebAuthnProcess::connectionToWebProcessClosed):
(WebKit::WebAuthnProcess::shouldTerminate):
(WebKit::WebAuthnProcess::didClose):
(WebKit::WebAuthnProcess::lowMemoryHandler):
(WebKit::WebAuthnProcess::initializeWebAuthnProcess):
(WebKit::WebAuthnProcess::prepareToSuspend):
(WebKit::WebAuthnProcess::processDidResume):
(WebKit::WebAuthnProcess::resume):
(WebKit::WebAuthnProcess::processDidTransitionToForeground):
(WebKit::WebAuthnProcess::processDidTransitionToBackground):
(WebKit::WebAuthnProcess::webProcessConnection const):
* WebAuthnProcess/WebAuthnProcess.h: Added.
* WebAuthnProcess/WebAuthnProcess.messages.in: Added.
* WebAuthnProcess/WebAuthnProcessCreationParameters.cpp: Copied from Source/WebKit/Daemons/WebAuthenticationAgent/main.mm.
(WebKit::WebAuthnProcessCreationParameters::encode const):
(WebKit::WebAuthnProcessCreationParameters::decode):
* WebAuthnProcess/WebAuthnProcessCreationParameters.h: Copied from Source/WebKit/Daemons/WebAuthenticationAgent/main.mm.
* WebAuthnProcess/ios/WebAuthnProcessIOS.mm: Copied from Source/WebKit/Daemons/WebAuthenticationAgent/main.mm.
(WebKit::WebAuthnProcess::initializeProcess):
(WebKit::WebAuthnProcess::initializeProcessName):
(WebKit::WebAuthnProcess::initializeSandbox):
* WebAuthnProcess/mac/WebAuthnProcessMac.mm: Added.
(WebKit::WebAuthnProcess::initializeProcess):
(WebKit::WebAuthnProcess::initializeProcessName):
(WebKit::WebAuthnProcess::initializeSandbox):
Paperwork for introducing the new WebAuthn process. Mostly copied from GPU process.
* WebAuthnProcess/mac/com.apple.WebKit.WebAuthnProcess.sb.in: Added.
The sandbox profile is originally from the GPU Process with IOKit related rules removed. Will tighten it again after the process is fully functional.
* WebKit.xcodeproj/project.pbxproj:
Paperwork for building the new WebAuthn process.
* WebProcess/WebAuthentication/WebAuthnProcessConnection.cpp: Copied from Source/WebKit/Daemons/WebAuthenticationAgent/main.mm.
(WebKit::WebAuthnProcessConnection::WebAuthnProcessConnection):
(WebKit::WebAuthnProcessConnection::~WebAuthnProcessConnection):
(WebKit::WebAuthnProcessConnection::didClose):
(WebKit::WebAuthnProcessConnection::didReceiveInvalidMessage):
* WebProcess/WebAuthentication/WebAuthnProcessConnection.h: Renamed from Source/WebKit/Daemons/WebAuthenticationAgent/main.mm.
(WebKit::WebAuthnProcessConnection::create):
(WebKit::WebAuthnProcessConnection::connection):
* WebProcess/WebAuthentication/WebAuthnProcessConnection.messages.in: Added.
* WebProcess/WebAuthentication/WebAuthnProcessConnectionInfo.h: Added.
(WebKit::WebAuthnProcessConnectionInfo::identifier const):
(WebKit::WebAuthnProcessConnectionInfo::releaseIdentifier):
(WebKit::WebAuthnProcessConnectionInfo::encode const):
(WebKit::WebAuthnProcessConnectionInfo::decode):
* WebProcess/WebProcess.cpp:
(WebKit::getWebAuthnProcessConnection):
(WebKit::WebProcess::ensureWebAuthnProcessConnection):
(WebKit::WebProcess::webAuthnProcessConnectionClosed):
* WebProcess/WebProcess.h:
(WebKit::WebProcess::existingWebAuthnProcessConnection):
Paperwork for introducing the new WebAuthn process. Mostly copied from GPU process.
2020-10-16 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Remove the "alg" field in the attestation statement
https://bugs.webkit.org/show_bug.cgi?id=217720
<rdar://problem/70349734>
Reviewed by Brent Fulgham.
This old field was used in an unreleased old attestation statement format and is not used by
the final attestation statement format that will be added to the spec:
https://github.com/w3c/webauthn/pull/1491.
Therefore, remove it to resolve confusion.
Covered by existing tests.
* UIProcess/WebAuthentication/Cocoa/LocalAuthenticator.mm:
(WebKit::LocalAuthenticator::continueMakeCredentialAfterAttested):
2020-10-16 Youenn Fablet <youenn@apple.com>
Add support for GPUProcess WebAudio media element providers
https://bugs.webkit.org/show_bug.cgi?id=217704
Reviewed by Eric Carlson.
Introduce RemoteAudioSourceProviderManager which receives IPC messages from GPU process for web audio data.
RemoteAudioSourceProviderManager sends it to the identified RemoteAudioSourceProvider.
Introduce RemoteAudioSourceProviderProxy which is a client to the AV AudioSourceProviderAVFObjC.
RemoteAudioSourceProviderProxy will send IPC messages to RemoteAudioSourceProviderManager to send audio data and description.
* CMakeLists.txt:
* DerivedSources-input.xcfilelist:
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* GPUProcess/media/RemoteAudioSourceProviderProxy.cpp: Added.
(WebKit::RemoteAudioSourceProviderProxy::create):
(WebKit::RemoteAudioSourceProviderProxy::RemoteAudioSourceProviderProxy):
(WebKit::RemoteAudioSourceProviderProxy::createRingBuffer):
(WebKit::RemoteAudioSourceProviderProxy::newAudioSamples):
(WebKit::RemoteAudioSourceProviderProxy::storageChanged):
* GPUProcess/media/RemoteAudioSourceProviderProxy.h: Added.
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
(WebKit::RemoteMediaPlayerProxy::~RemoteMediaPlayerProxy):
(WebKit::RemoteMediaPlayerProxy::createAudioSourceProvider):
(WebKit::RemoteMediaPlayerProxy::setShouldEnableAudioSourceProvider):
* GPUProcess/media/RemoteMediaPlayerProxy.h:
* GPUProcess/media/RemoteMediaPlayerProxy.messages.in:
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/GPUProcessConnection.cpp:
(WebKit::GPUProcessConnection::audioSourceProviderManager):
* WebProcess/GPU/GPUProcessConnection.h:
* WebProcess/GPU/media/RemoteAudioSourceProvider.cpp:
(WebKit::RemoteAudioSourceProvider::create):
(WebKit::RemoteAudioSourceProvider::RemoteAudioSourceProvider):
(WebKit::RemoteAudioSourceProvider::close):
(WebKit::RemoteAudioSourceProvider::hasNewClient):
* WebProcess/GPU/media/RemoteAudioSourceProviderManager.cpp: Added.
(WebKit::RemoteAudioSourceProviderManager::RemoteAudioSourceProviderManager):
(WebKit::RemoteAudioSourceProviderManager::~RemoteAudioSourceProviderManager):
(WebKit::RemoteAudioSourceProviderManager::setConnection):
(WebKit::RemoteAudioSourceProviderManager::addProvider):
(WebKit::RemoteAudioSourceProviderManager::removeProvider):
(WebKit::RemoteAudioSourceProviderManager::dispatchToThread):
(WebKit::RemoteAudioSourceProviderManager::audioStorageChanged):
(WebKit::RemoteAudioSourceProviderManager::audioSamplesAvailable):
(WebKit::RemoteAudioSourceProviderManager::RemoteAudio::RemoteAudio):
(WebKit::RemoteAudioSourceProviderManager::RemoteAudio::setStorage):
(WebKit::RemoteAudioSourceProviderManager::RemoteAudio::audioSamplesAvailable):
* WebProcess/GPU/media/RemoteAudioSourceProviderManager.h: Added.
(WebKit::RemoteAudioSourceProviderManager::create):
* WebProcess/GPU/media/RemoteAudioSourceProviderManager.messages.in: Added.
* WebProcess/cocoa/RemoteCaptureSampleManager.cpp:
2020-10-15 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Move AppAttestInternal related code from WKA to OpenSource
https://bugs.webkit.org/show_bug.cgi?id=217790
<rdar://problem/59613406>
Reviewed by Brent Fulgham.
Covered by manual testing.
* Platform/spi/Cocoa/AppAttestSPI.h: Copied from Source/WebKit/UIProcess/WebAuthentication/Cocoa/AppAttestInternalSoftLink.mm.
* SourcesCocoa.txt:
* UIProcess/WebAuthentication/Cocoa/AppAttestInternalSoftLink.h: Added.
* UIProcess/WebAuthentication/Cocoa/AppAttestInternalSoftLink.mm: Added.
* UIProcess/WebAuthentication/Cocoa/LocalConnection.mm:
(WebKit::LocalConnection::getAttestation const):
* UIProcess/WebAuthentication/Cocoa/LocalService.mm:
(WebKit::LocalService::isAvailable):
* WebKit.xcodeproj/project.pbxproj:
2020-10-15 David Kilzer <ddkilzer@apple.com>
v2: WebKit::XPCServiceEventHandler block should call exit() on the main thread
<https://webkit.org/b/217773>
<rdar://problem/70275659>
Reviewed by Geoff Garen.
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceMain.mm:
(WebKit::XPCServiceEventHandler):
- Use NSRunLoop instead of libdispatch to run code on the main
thread since libdispatch can randomly pick a new main thread when
exit() is called on the original main thread.
2020-10-15 Tim Horton <timothy_horton@apple.com>
iPhone apps and iPad apps linked before iOS 13 running on macOS get desktop UA unexpectedly
https://bugs.webkit.org/show_bug.cgi?id=217798
<rdar://problem/70254509>
Reviewed by Wenson Hsieh.
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::desktopClassBrowsingRecommended):
Share more of the code with iOS, for compatibility and consistency's sake:
iOS apps running on macOS should use the mobile UA if they are iPhone-only,
or if they were linked on an SDK earlier than iOS 13. This matches their
behavior on iPad.
2020-10-15 Per Arne Vollan <pvollan@apple.com>
[macOS] Issue temporary sandbox extension to the Launch Services daemon
https://bugs.webkit.org/show_bug.cgi?id=217781
<rdar://problem/70350924>
Reviewed by Brent Fulgham.
To support blocking of the Launch Services daemon in the WebContent process, issue temporary sandbox extension to this service on macOS.
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
* WebProcess/com.apple.WebProcess.sb.in:
2020-10-15 Per Arne Vollan <pvollan@apple.com>
[macOS] Add telemetry for distributed notification service
https://bugs.webkit.org/show_bug.cgi?id=217762
<rdar://problem/70341045>
Reviewed by Brent Fulgham.
Add telemetry with backtrace for the distributed notification service on macOS.
* WebProcess/com.apple.WebProcess.sb.in:
2020-10-15 David Kilzer <ddkilzer@apple.com>
PDFPlugin.mm:2303:139: error: type 'SEL _Nullable' cannot be narrowed to 'bool' in initializer list [-Wc++11-narrowing]
<https://webkit.org/b/217791>
<rdar://problem/70349804>
Reviewed by Tim Horton.
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::handleContextMenuEvent):
- Use `!!` to convert result to `bool`.
2020-10-15 Aditya Keerthi <akeerthi@apple.com>
Crash in -[WKDateTimePicker setDateTimePickerToInitialValue]
https://bugs.webkit.org/show_bug.cgi?id=217763
<rdar://problem/68635008>
Reviewed by Wenson Hsieh.
Crash occurs due to the use of an autoreleased NSString in
setDateTimePickerToInitialValue. To fix, make the variable
a RetainPtr.
* UIProcess/ios/forms/WKDateTimeInputControl.mm:
(-[WKDateTimePicker setDateTimePickerToInitialValue]):
2020-10-15 Chris Lord <clord@igalia.com>
[GTK][WPE] Add support for smooth scrolling animation with async scrolling
https://bugs.webkit.org/show_bug.cgi?id=210382
Reviewed by Carlos Garcia Campos.
Add hasPreciseScrollingDeltas to GTK/WPE wheel events.
No new tests needed.
* Shared/WebEventConversion.cpp:
(WebKit::WebKit2PlatformWheelEvent::WebKit2PlatformWheelEvent):
* Shared/WebWheelEvent.cpp:
(WebKit::WebWheelEvent::WebWheelEvent):
(WebKit::WebWheelEvent::encode const):
(WebKit::WebWheelEvent::decode):
* Shared/WebWheelEvent.h:
* Shared/WebWheelEventCoalescer.cpp:
(WebKit::WebWheelEventCoalescer::canCoalesce):
(WebKit::WebWheelEventCoalescer::coalesce):
* Shared/gtk/NativeWebWheelEventGtk.cpp:
(WebKit::NativeWebWheelEvent::NativeWebWheelEvent):
* Shared/gtk/WebEventFactory.cpp:
(WebKit::WebEventFactory::createWebWheelEvent):
* Shared/libwpe/WebEventFactory.cpp:
(WebKit::WebEventFactory::createWebWheelEvent):
2020-10-15 Youenn Fablet <youenn@apple.com>
Introduce a RemoteAudioSourceProvider to receive audio samples for web audio from GPUProcess
https://bugs.webkit.org/show_bug.cgi?id=217710
Reviewed by Eric Carlson.
Make use of WebAudioSourceProviderAVFObjC to implement RemoteAudioSourceProvider.
RemoteAudioSourceProvider currently implements an AudioSourceProvider but it is not yet connected to GPUProcess media players.
This will be implemented in a follow-up.
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::~MediaPlayerPrivateRemote):
(WebKit::MediaPlayerPrivateRemote::audioSourceProvider):
* WebProcess/GPU/media/MediaPlayerPrivateRemote.h:
* WebProcess/GPU/media/RemoteAudioSourceProvider.cpp: Added.
(WebKit::RemoteAudioSourceProvider::create):
(WebKit::RemoteAudioSourceProvider::RemoteAudioSourceProvider):
(WebKit::RemoteAudioSourceProvider::~RemoteAudioSourceProvider):
(WebKit::RemoteAudioSourceProvider::close):
(WebKit::RemoteAudioSourceProvider::hasNewClient):
(WebKit::RemoteAudioSourceProvider::audioSamplesAvailable):
(WebKit::RemoteAudioSourceProvider::logChannel const):
* WebProcess/GPU/media/RemoteAudioSourceProvider.h: Added.
* WebProcess/GPU/media/cocoa/MediaPlayerPrivateRemoteCocoa.mm:
2020-10-15 Adrian Perez de Castro <aperez@igalia.com>
Non-unified build fixes, mid October 2020, take two
https://bugs.webkit.org/show_bug.cgi?id=217734
Unreviewed build fix.
* UIProcess/WebProcessProxy.cpp: Add missing WebPreferencesKeys.h header.
2020-10-14 Brent Fulgham <bfulgham@apple.com>
[iOS] Allow additional sysctl reads needed by image decoding
https://bugs.webkit.org/show_bug.cgi?id=217733
<rdar://problem/68649171>
Reviewed by Per Arne Vollan.
Telemetry on iOS 14 shows that we are hitting some sandbox violations during image decoding. We should
grant access to hw.byteorder, hw.cachelinesize_compat, and hw.vectorunit. We should silence warnings
about hw.cpufrequency_compat since it is not needed in web-facing use cases.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-10-14 Ryosuke Niwa <rniwa@webkit.org>
Unreviewed build fix for non-Cocoa platforms & release builds.
* Platform/IPC/Connection.cpp:
(IPC::Connection::dispatchSyncMessage):
2020-10-14 Ryosuke Niwa <rniwa@webkit.org>
IPC testing JS API should expose a reply and describe the list of arguments for each message
https://bugs.webkit.org/show_bug.cgi?id=217565
Reviewed by Geoffrey Garen.
This patch makes IPC.sendMessage and IPC.sendSyncMessage decode the reply. IPC.sendSyncMessage now returns
a dictionary with two keys: "buffer" and "arguments", the first of which returns ArrayBuffer of the decoded
message reply and the second of which is an array of decoded arguments where there was an appropriate JS
binding code existed for the argument type. IPC.sendMessage now returns a Promise which can either resolve
with the same dictionary as IPC.sendSyncMessage or reject when the decoding fails.
In addition, this patch exposes a dictionary describing each IPC message argument's type and parameter name.
In order to add these two functionalities, this patch adds a new step in generate-message-receiver.py to
generate MessageArgumentDescriptions.cpp, which contains functions which know how to decode arguments of
any IPC message and create a JS dictionary describing it as well as functions that return descriptions of
arguments or reply arguments.
Finally, this patch adds encoders for a few types found to be very common after r268239 had been landed.
Tests: TestWebKitAPI.IPCTestingAPI.DecodesReplyArgumentsForPrompt
TestWebKitAPI.IPCTestingAPI.DecodesReplyArgumentsForAsyncMessage
TestWebKitAPI.IPCTestingAPI.DescribesArguments
* CMakeLists.txt:
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* Platform/IPC/Decoder.h:
(IPC::Decoder::buffer const):
* Platform/IPC/JSIPCBinding.h: Added.
(jsValueForDecodedArgumentValue): Added. A template function to construct a JS value for a given C++ value.
(jsValueForDecodedNumericArgumentValue): Added. A helper for constructing a JS value for numeric values.
(jsValueForDecodedArgumentRect): Added. Ditto for IntRect and FloatRect.
(DecodedArgumentJSValueConverter): Added. A helper class to construct JS values for a tuple of values using
partial template specializations.
(DecodedArgumentJSValueConverter::convert): Added.
(jsValueForArgumentTuple): Added. A helper to construct a JS array for the decoded IPC arguments.
(jsValueForDecodedArguments): Added.
* Platform/IPC/MessageArgumentDescriptions.h: Added.
(IPC::ArgumentDescription): Added.
* Scripts/generate-message-receiver.py:
(main): Generate MessageArgumentDescriptions.cpp.
* Scripts/webkit/messages.py:
(headers_for_type): Removed the special case for PaymentMethodUpdate now that it's in its own header. Also
added made webrtc::WebKitEncodedFrameInfo include LibWebRTCEnumTraits.h as it uses webrtc::VideoFrameType.
(collect_header_conditions_for_receiver): Extracted from generate_message_handler.
(generate_header_includes_from_conditions): Ditto.
(generate_message_handler):
(generate_js_value_conversion_function): Added.
(generate_js_argument_descriptions): Added.
(generate_message_argument_description_implementation): Added.
* Shared/ApplePay/ApplePayPaymentSetupFeaturesWebKit.h: Fixed a bug that we were not forward declaring NSArray.
* SourcesCocoa.txt: Added MessageArgumentDescriptions.cpp as a non-unified cpp file as its size is around 1MB.
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebPage/IPCTestingAPI.cpp:
(WebKit::IPCTestingAPI::JSIPC::staticValues): Added the session ID and page ID as static variables.
(WebKit::IPCTestingAPI::encodePointType): Added.
(WebKit::IPCTestingAPI::encodeRectType): Fixed the bug was that this code wasn't checking for any exceptions.
(WebKit::IPCTestingAPI::encodeNumericType): Renamed from encodeIntegralType since this function is now used
to encode double and float, not just integral types.
(WebKit::IPCTestingAPI::encodeArgument): Added the support for IntPoint, FloatPoint, URL, RegistrableDomain,
double, and float all of which turned out to be in the top 20 most common types.
(WebKit::IPCTestingAPI::jsResultFromReplyDecoder): Added.
(WebKit::IPCTestingAPI::JSIPC::sendMessage): Added the code to return Promise when there is a reply and resolve
it with the JS object describing the decoded argument. We use messageReplyArgumentDescriptions to figure out
whether there is a reply or not.
(WebKit::IPCTestingAPI::JSIPC::sendSyncMessage): Decode the reply and return a JS object which describes it.
(WebKit::IPCTestingAPI::JSIPC::frameID): Renamed from frameIdentifier to be consistent.
(WebKit::IPCTestingAPI::JSIPC::pageID): Added.
(WebKit::IPCTestingAPI::JSIPC::sessionID): Added.
(WebKit::IPCTestingAPI::createJSArrayForArgumentDescriptions): Added.
(WebKit::IPCTestingAPI::JSIPC::messages): Added the code to generate descriptions for arguments.
2020-10-14 Chris Dumez <cdumez@apple.com>
REGRESSION(r268423) SHOULD NEVER BE REACHED: GPUConnectionToWebProcess::didReceiveMessage(IPC::Connection &, IPC::Decoder &)
https://bugs.webkit.org/show_bug.cgi?id=217730
<rdar://problem/70308112>
Reviewed by Geoffrey Garen.
r268423 added support for registering IPC WorkQueueMessageReceivers & ThreadMessageReceivers with
a given identifier. When receiving an IPC message, we would look up the key [ReceiverName, ID] in
the receiver map and pass the message to this receiver.
This introduced a crash because of an existing ThreadMessageReceiver (RemoteAudioMediaStreamTrackRenderer)
was registering itself as a message receiver without an identifier (ID=0). However, the IPC senders
on GPU process side would pass a non-0 identifier when sending the IPC. As a result, we would fail to
look up the receiver in the map when receiving those IPC messages.
While existing code should probably be fixed so that senders and receivers agree on on identifier,
I have decided to fix the issue in this patch at IPC::Connection level. When receiving an IPC
message, we first check if there is a "global" receiver in the map by looking up key
[ReceiverName, 0]. If there is, we pass the IPC to this one. If there isn't, we look up the key
[ReceiverName, ID] in the map and use the more specific receiver.
The reason I chose to fix it this way is that the code in MessageReceiverMap (which is used for
receivers that are not using a WorkQueue or a Thread) behaves in this fashion. I believe it is
good to have consistency between all kinds of receivers.
No new tests, covered by existing test that is crashing on the bots.
* Platform/IPC/Connection.cpp:
(IPC::Connection::threadMessageReceiver):
(IPC::Connection::workQueueMessageReceiver):
(IPC::Connection::dispatchMessageToWorkQueueReceiver):
(IPC::Connection::dispatchMessageToThreadReceiver):
* Platform/IPC/Connection.h:
2020-10-14 Simon Fraser <simon.fraser@apple.com>
Scrolls in the passive handler region should be reported as handled promptly
https://bugs.webkit.org/show_bug.cgi?id=217718
Reviewed by Tim Horton.
The UI process will enqueue wheel events to send later if the WebContent process doesn't
promptly respond that it has handled the previous event. When scrolling in the passive
event region, we want the WebContent process to appear responsive, so respond to the UI
process when we've scrolled (which is before the event has been dispatched for DOM events).
Hard to test because CPU-constrained bots don't send wheel events in a predictable way.
* WebProcess/WebPage/EventDispatcher.cpp:
(WebKit::EventDispatcher::wheelEvent):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::wheelEvent):
2020-10-14 Ryosuke Niwa <rniwa@webkit.org>
Enabling IPC testing API should prevent the termination of WebContent process which sends an invalid IPC
https://bugs.webkit.org/show_bug.cgi?id=217698
Reviewed by Geoffrey Garen.
A part of this was landed in r268431 but this patch formally disables UI process' default behavior
to terminate a Web process upon receiving an invalid message from it.
Tests: IPCTestingAPI.CanSendInvalidAsyncMessageWithoutTermination
IPCTestingAPI.CanSendInvalidMessageWithoutTermination
* Platform/IPC/Connection.cpp:
(IPC::Connection::dispatchSyncMessage): Disable the debug assertion if the IPC testing API is enabled.
* Platform/IPC/Connection.h:
(IPC::Connection::setIgnoreInvalidMessageForTesting): Added.
(IPC::Connection::ignoreInvalidMessageForTesting const): Added.
(IPC::Connection::m_ignoreInvalidMessageForTesting): Added.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::WebPageProxy): Set the flag on WebProcessProxy to trigger the behavior.
(WebKit::WebPageProxy::launchProcess): Ditto.
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::didReceiveInvalidMessage): Use the flag on Connection instead of reaching
out to the default page group.
(WebKit::WebProcessProxy::setIgnoreInvalidMessageForTesting): Added. Remember the fact we've enabled IPC
testing API in a member variable and propagate the flag to Connection if a Web process is already running.
(WebKit::WebProcessProxy::didFinishLaunching): Propagte the flag to the newly launched Web process.
* UIProcess/WebProcessProxy.h:
(WebKit::WebProcessProxy::m_ignoreInvalidMessageForTesting): Added.
2020-10-14 Per Arne Vollan <pvollan@apple.com>
[macOS] Issue sandbox extension to additional icon service when attachment element is enabled.
https://bugs.webkit.org/show_bug.cgi?id=217706
<rdar://problem/70291100>
Reviewed by Brent Fulgham.
Issue sandbox extension to 'com.apple.iconservices.store' when attachment element is enabled on macOS.
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::attachmentElementServices):
(WebKit::WebPageProxy::creationParameters):
* WebProcess/WebPage/WebPage.cpp:
* WebProcess/com.apple.WebProcess.sb.in:
2020-10-14 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r268431 and r268443.
https://bugs.webkit.org/show_bug.cgi?id=217728
Broke debug builds of GTK, WinCairo and Playstation ports
Reverted changesets:
"IPC testing JS API should expose a reply and describe the
list of arguments for each message"
https://bugs.webkit.org/show_bug.cgi?id=217565
https://trac.webkit.org/changeset/268431
"IPC testing JS API should expose a reply and describe the
list of arguments for each message"
https://bugs.webkit.org/show_bug.cgi?id=217565
https://trac.webkit.org/changeset/268443
2020-10-14 Don Olmstead <don.olmstead@sony.com>
Non-unified build fixes, mid October 2020
https://bugs.webkit.org/show_bug.cgi?id=217721
Reviewed by Yusuke Suzuki.
* NetworkProcess/Downloads/DownloadManager.cpp:
* UIProcess/Downloads/DownloadProxyMap.cpp:
2020-10-14 Devin Rousso <drousso@apple.com>
REGRESSION (r268384): ASSERTION FAILED: _startCount > 1 in -[WKMouseDeviceObserver stop]
https://bugs.webkit.org/show_bug.cgi?id=217684
<rdar://problem/70272646>
Reviewed by Tim Horton.
The `ASSERT` was incorrect in that it's possible for `_startCount` to be equal to `1`. Also
add a check to make sure `_startCount` doesn't overflow.
* UIProcess/ios/WKMouseDeviceObserver.mm:
(-[WKMouseDeviceObserver stop]):
* UIProcess/ios/WKStylusDeviceObserver.mm:
(-[WKStylusDeviceObserver stop]):
2020-10-14 Sergio Villar Senin <svillar@igalia.com>
[GLIB] Fix linking error for systems not providing <sys/memfd.h>
https://bugs.webkit.org/show_bug.cgi?id=217514
Reviewed by Carlos Garcia Campos.
* UIProcess/Launcher/glib/BubblewrapLauncher.cpp: Unconditionally check for <sys/mman.h> which should always exist instead
of <sys/memfd.h>. Also define the fallback with syscall() only for linux systems (as it requires <linux/memfd.h>).
2020-10-14 Kate Cheney <katherine_cheney@apple.com>
Remove ResourceLoadStatistics telemetry
https://bugs.webkit.org/show_bug.cgi?id=217665
<rdar://problem/69748988>
Reviewed by Sam Weinig.
Telemetry calculation in ITP is unused, very complicated, a potential
privacy risk, and it causes flaky tests. We should remove it.
No new tests. Existing ITP tests will check for no behavior change.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::ResourceLoadStatisticsDatabaseStore):
(WebKit::makeMedianWithUIQuery): Deleted.
(WebKit::buildQueryStartAndEnd): Deleted.
(WebKit::makeMedianWithoutUIQuery): Deleted.
(WebKit::getMedianOfPrevalentResourcesWithUserInteraction): Deleted.
(WebKit::ResourceLoadStatisticsDatabaseStore::getNumberOfPrevalentResources const): Deleted.
(WebKit::ResourceLoadStatisticsDatabaseStore::getNumberOfPrevalentResourcesWithUI const): Deleted.
(WebKit::ResourceLoadStatisticsDatabaseStore::getTopPrevelentResourceDaysSinceUI const): Deleted.
(WebKit::getMedianOfPrevalentResourceWithoutUserInteraction): Deleted.
(WebKit::getNumberOfPrevalentResourcesInTopResources): Deleted.
(WebKit::makeStatisticQuery): Deleted.
(WebKit::ResourceLoadStatisticsDatabaseStore::getNumberOfPrevalentResourcesWithoutUI const): Deleted.
(WebKit::ResourceLoadStatisticsDatabaseStore::calculateTelemetryData const): Deleted.
(WebKit::ResourceLoadStatisticsDatabaseStore::calculateAndSubmitTelemetry const): Deleted.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:
(WebKit::ResourceLoadStatisticsMemoryStore::ResourceLoadStatisticsMemoryStore):
(WebKit::ResourceLoadStatisticsMemoryStore::calculateAndSubmitTelemetry const): Deleted.
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.cpp:
(WebKit::ResourceLoadStatisticsStore::setShouldSubmitTelemetry): Deleted.
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.h:
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::performDailyTasks):
(WebKit::WebResourceLoadStatisticsStore::setShouldSubmitTelemetry): Deleted.
(WebKit::WebResourceLoadStatisticsStore::submitTelemetry): Deleted.
(WebKit::WebResourceLoadStatisticsStore::notifyPageStatisticsTelemetryFinished const): Deleted.
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.h:
* NetworkProcess/Classifier/WebResourceLoadStatisticsTelemetry.cpp: Removed.
* NetworkProcess/Classifier/WebResourceLoadStatisticsTelemetry.h: Removed.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::submitTelemetry): Deleted.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/NetworkSession.cpp:
(WebKit::NetworkSession::notifyPageStatisticsTelemetryFinished): Deleted.
* NetworkProcess/NetworkSession.h:
* SourcesCocoa.txt:
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreStatisticsSubmitTelemetry): Deleted.
* UIProcess/API/C/WKWebsiteDataStoreRef.h:
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::submitTelemetry): Deleted.
(WebKit::NetworkProcessProxy::notifyResourceLoadStatisticsTelemetryFinished): Deleted.
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/Network/NetworkProcessProxy.messages.in:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::notifyPageStatisticsTelemetryFinished): Deleted.
* UIProcess/WebProcessProxy.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::submitTelemetry): Deleted.
* UIProcess/WebsiteData/WebsiteDataStore.h:
* WebKit.xcodeproj/project.pbxproj:
* PlatformFTW.cmake:
* SourcesGTK.txt:
* PlatformPlayStation.cmake:
* PlatformWin.cmake:
* SourcesWPE.txt:
2020-10-14 Kate Cheney <katherine_cheney@apple.com>
Remove _WKWebViewCategory from WKWebViewConfiguration
https://bugs.webkit.org/show_bug.cgi?id=217679
<rdar://problem/70260337>
Reviewed by Sam Weinig.
This property is not used in any meaningful way. Its original purpose
was to categorize WKWebViews for app-bound domains, but this is
now done using the isNavigatingToAppBoundDomain property on
WebPageProxy. We should remove it.
No new tests, behavior confirmed by existing tests.
* Shared/WebViewCategory.h: Removed.
* UIProcess/API/APIPageConfiguration.cpp:
(API::PageConfiguration::copy const):
* UIProcess/API/APIPageConfiguration.h:
(API::PageConfiguration::webViewCategory const): Deleted.
(API::PageConfiguration::setWebViewCategory): Deleted.
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(toWebKitWebViewCategory): Deleted.
(toWKWebViewCategory): Deleted.
(-[WKWebViewConfiguration _webViewCategory]): Deleted.
(-[WKWebViewConfiguration _setWebViewCategory:]): Deleted.
* UIProcess/API/Cocoa/WKWebViewConfigurationPrivate.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::setIsNavigatingToAppBoundDomainAndCheckIfPermitted):
* WebKit.xcodeproj/project.pbxproj:
2020-10-14 Joonghun Park <jh718.park@samsung.com>
Unreviewed. Fix the build warning below since r268312.
warning: control reaches end of non-void function [-Wreturn-type]
No new tests, no new behaviors.
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::numberOfConnectedGamepadsForTesting):
2020-10-14 Joonghun Park <jh718.park@samsung.com>
Unreviewed. Fix the build warning below since r222233.
warning: redundant move in return statement [-Wredundant-move]
No new tests, no new behaviors.
* Shared/Gamepad/GamepadData.cpp:
(WebKit::GamepadData::decode):
2020-10-13 Julian Gonzalez <julian_a_gonzalez@apple.com>
Null dereference in PDFPlugin::snapshot()
https://bugs.webkit.org/show_bug.cgi?id=217668
Reviewed by Ryosuke Niwa.
Add a null pointer check in PDFPlugin's snapshot method
to prevent a crash if the backing store's size
isn't sufficient for the ShareableBitmap.
Test: plugins/pdf-plugin-null-onloaddeddata.html
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::snapshot):
2020-10-12 Ryosuke Niwa <rniwa@webkit.org>
IPC testing JS API should expose a reply and describe the list of arguments for each message
https://bugs.webkit.org/show_bug.cgi?id=217565
Reviewed by Geoffrey Garen.
This patch makes IPC.sendMessage and IPC.sendSyncMessage decode the reply. IPC.sendSyncMessage now returns
a dictionary with two keys: "buffer" and "arguments", the first of which returns ArrayBuffer of the decoded
message reply and the second of which is an array of decoded arguments where there was an appropriate JS
binding code existed for the argument type. IPC.sendMessage now returns a Promise which can either resolve
with the same dictionary as IPC.sendSyncMessage or reject when the decoding fails.
In addition, this patch exposes a dictionary describing each IPC message argument's type and parameter name.
In order to add these two functionalities, this patch adds a new step in generate-message-receiver.py to
generate MessageArgumentDescriptions.cpp, which contains functions which know how to decode arguments of
any IPC message and create a JS dictionary describing it as well as functions that return descriptions of
arguments or reply arguments.
Finally, this patch adds encoders for a few types found to be very common after r268239 had been landed.
Tests: TestWebKitAPI.IPCTestingAPI.DecodesReplyArgumentsForPrompt
TestWebKitAPI.IPCTestingAPI.DecodesReplyArgumentsForAsyncMessage
TestWebKitAPI.IPCTestingAPI.DescribesArguments
* CMakeLists.txt:
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* Platform/IPC/Decoder.h:
(IPC::Decoder::buffer const):
* Platform/IPC/JSIPCBinding.h: Added.
(jsValueForDecodedArgumentValue): Added. A template function to construct a JS value for a given C++ value.
(jsValueForDecodedNumericArgumentValue): Added. A helper for constructing a JS value for numeric values.
(jsValueForDecodedArgumentRect): Added. Ditto for IntRect and FloatRect.
(DecodedArgumentJSValueConverter): Added. A helper class to construct JS values for a tuple of values using
partial template specializations.
(DecodedArgumentJSValueConverter::convert): Added.
(jsValueForArgumentTuple): Added. A helper to construct a JS array for the decoded IPC arguments.
(jsValueForDecodedArguments): Added.
* Platform/IPC/MessageArgumentDescriptions.h: Added.
(IPC::ArgumentDescription): Added.
* Scripts/generate-message-receiver.py:
(main): Generate MessageArgumentDescriptions.cpp.
* Scripts/webkit/messages.py:
(headers_for_type): Removed the special case for PaymentMethodUpdate now that it's in its own header. Also
added made webrtc::WebKitEncodedFrameInfo include LibWebRTCEnumTraits.h as it uses webrtc::VideoFrameType.
(collect_header_conditions_for_receiver): Extracted from generate_message_handler.
(generate_header_includes_from_conditions): Ditto.
(generate_message_handler):
(generate_js_value_conversion_function): Added.
(generate_js_argument_descriptions): Added.
(generate_message_argument_description_implementation): Added.
* Shared/ApplePay/ApplePayPaymentSetupFeaturesWebKit.h: Fixed a bug that we were not forward declaring NSArray.
* SourcesCocoa.txt: Added MessageArgumentDescriptions.cpp as a non-unified cpp file as its size is around 1MB.
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebPage/IPCTestingAPI.cpp:
(WebKit::IPCTestingAPI::JSIPC::staticValues): Added the session ID and page ID as static variables.
(WebKit::IPCTestingAPI::encodePointType): Added.
(WebKit::IPCTestingAPI::encodeRectType): Fixed the bug was that this code wasn't checking for any exceptions.
(WebKit::IPCTestingAPI::encodeNumericType): Renamed from encodeIntegralType since this function is now used
to encode double and float, not just integral types.
(WebKit::IPCTestingAPI::encodeArgument): Added the support for IntPoint, FloatPoint, URL, RegistrableDomain,
double, and float all of which turned out to be in the top 20 most common types.
(WebKit::IPCTestingAPI::jsResultFromReplyDecoder): Added.
(WebKit::IPCTestingAPI::JSIPC::sendMessage): Added the code to return Promise when there is a reply and resolve
it with the JS object describing the decoded argument. We use messageReplyArgumentDescriptions to figure out
whether there is a reply or not.
(WebKit::IPCTestingAPI::JSIPC::sendSyncMessage): Decode the reply and return a JS object which describes it.
(WebKit::IPCTestingAPI::JSIPC::frameID): Renamed from frameIdentifier to be consistent.
(WebKit::IPCTestingAPI::JSIPC::pageID): Added.
(WebKit::IPCTestingAPI::JSIPC::sessionID): Added.
(WebKit::IPCTestingAPI::createJSArrayForArgumentDescriptions): Added.
(WebKit::IPCTestingAPI::JSIPC::messages): Added the code to generate descriptions for arguments.
2020-10-13 Ellie Epskamp-Hunt <eepskamphunt@apple.com>
Expose a property on WKWebView to know if it's being inspected.
https://bugs.webkit.org/show_bug.cgi?id=217629
rdar://70073369
Reviewed by Brian Burg.
Move an existing property from WKWebViewPrivateForTesting to WKWebViewPrivate so Safari can
know when a web view is being inspected.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _isBeingInspected]):
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/API/Cocoa/WKWebViewPrivateForTesting.h:
* UIProcess/API/Cocoa/WKWebViewTesting.mm:
(-[WKWebView _hasInspectorFrontend]): Deleted.
2020-10-13 Chris Dumez <cdumez@apple.com>
WebAudio tests are crashing in debug when enabling the GPU process
https://bugs.webkit.org/show_bug.cgi?id=217663
Reviewed by Geoff Garen.
WebAudio tests were crashing in debug when enabling the GPU process because it did audio
processing on the WebContent process's main thread. To address the issue, I made
RemoteAudioDestinationProxy a ThreadMessageReceiver so that it receives IPC on an audio
thread instead of the main thread. IPC messages are processed directly on the AudioWorklet
thread when active or on an audio thread constructed by RemoteAudioDestinationProxy
otherwise.
* GPUProcess/media/RemoteAudioDestinationManager.cpp:
(WebKit::RemoteAudioDestination::RemoteAudioDestination):
Use a PushPullFIFO structure in render() to avoid hanging the audio rendering thread
on a semaphore. Hanging the rendering thread was terrible for performance and was also
a source of deadlock since the underlying framework is holding a lock while render() is
called. We could process a RemoteAudioDestination::CreateAudioDestination sync IPC on the
main thread and deadlock on that lock.
* GPUProcess/webrtc/LibWebRTCCodecsProxy.h:
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRendererManager.h:
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayerManager.h:
* NetworkProcess/IndexedDB/WebIDBServer.h:
* NetworkProcess/webrtc/NetworkRTCProvider.h:
* WebProcess/Network/webrtc/LibWebRTCNetwork.h:
* WebProcess/cocoa/RemoteCaptureSampleManager.h:
Use ThreadMessageReceiverRefCounted instead of ThreadMessageReceiver since those classes
do not provide their own RefCounting.
* Platform/IPC/Connection.cpp:
(IPC::Connection::addWorkQueueMessageReceiver):
(IPC::Connection::removeWorkQueueMessageReceiver):
(IPC::Connection::addThreadMessageReceiver):
(IPC::Connection::removeThreadMessageReceiver):
(IPC::Connection::processIncomingMessage):
(IPC::Connection::dispatchMessageToWorkQueueReceiver):
(IPC::Connection::dispatchMessageToThreadReceiver):
* Platform/IPC/Connection.h:
(IPC::Connection::ThreadMessageReceiver::ref):
(IPC::Connection::ThreadMessageReceiver::deref):
- Add support for passing a destinationID when registering a WorkQueueMessageReceiver or a
ThreadMessageReceiver, similarly to regular MessageReceivers. This was needed here since
The GPUProcess sends IPC messages to the RemoteAudioDestinationProxy with a given
destinationID and since RemoteAudioDestinationProxy is now a ThreadMessageReceiver.
- Stop having ThreadMessageReceiver subclass ThreadSafeRefCounted since
RemoteAudioDestinationProxy already subclasses ThreadSafeRefCounted indirectly. A new
ThreadMessageReceiverRefCounted class was added for convenience for existing code that
relied on its refcounting.
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::RemoteAudioDestinationProxy):
(WebKit::RemoteAudioDestinationProxy::~RemoteAudioDestinationProxy):
(WebKit::RemoteAudioDestinationProxy::start):
(WebKit::RemoteAudioDestinationProxy::stop):
(WebKit::RemoteAudioDestinationProxy::renderBuffer):
(WebKit::RemoteAudioDestinationProxy::didChangeIsPlaying):
(WebKit::RemoteAudioDestinationProxy::dispatchToThread):
* WebProcess/GPU/media/RemoteAudioDestinationProxy.h:
Use a PushPullFIFO container in render() to avoid handing the audio rendering thread on
a semaphore while the Render IPC is getting processed by the WebProcess.
2020-10-13 Per Arne Vollan <pvollan@apple.com>
[iOS 14] Hang in -[WKUserDefaults _notifyObserversOfChangeFromValuesForKeys:toValuesForKeys:]
https://bugs.webkit.org/show_bug.cgi?id=217183
<rdar://problem/69916673>
Reviewed by Geoffrey Garen.
Dispatch this work on a non-main thread, in order to avoid blocking the main thread.
* UIProcess/Cocoa/PreferenceObserver.mm:
(-[WKUserDefaults _notifyObserversOfChangeFromValuesForKeys:toValuesForKeys:]):
2020-10-13 Keith Rollin <krollin@apple.com>
Remove leftover MACOSX_DEPLOYMENT_TARGET_macosx support
https://bugs.webkit.org/show_bug.cgi?id=217649
<rdar://problem/70236877>
Reviewed by Darin Adler.
Bug 42796 introduced MACOSX_DEPLOYMENT_TARGET_<PLATFORM> as "support
for compiling WebKit against iOS SDKs". Support for the iOS part of
this feature was later removed in several changes, including Bug
139212, Bug 139463 and Bug 144762. However, vestiges have remained for
five or six years in the form of MACOSX_DEPLOYMENT_TARGET_macosx. The
inclusion of the platform in MACOSX_DEPLOYMENT_TARGET is no longer
needed and can be removed.
This changes brings most projects in conformance with other projects
that don't support including the platform in MACOSX_DEPLOYMENT_TARGET,
including WebEditingTester, gtest, WebKitTestRunner, MiniBrowser, and
TestWebKitAPI.
Along the way, remove a couple of left-over references to macOS 10.16,
and a couple of places where [sdk=macosx*] was still being used.
With this change, initialization of MACOSX_DEPLOYMENT_TARGET should be
consistent across all projects, with two exceptions: WebKitLauncher
(which hardcodes it to 10.12) and libwebrtc's copy of googletest
(which hardcodes it to 10.4). The reasons for these hard-coded values
is not apparent, so leave them be.
* Configurations/DebugRelease.xcconfig:
2020-10-13 Simon Fraser <simon.fraser@apple.com>
Plumb OptionSet<WheelEventProcessingSteps> through the wheel event handling code
https://bugs.webkit.org/show_bug.cgi?id=217660
Reviewed by Tim Horton.
Async scrolling in passive wheel event regions will require that the main thread event
handling code knows that the scrolling thread did the scrolling. To propagate this
information, plumb OptionSet<WheelEventProcessingSteps> through.
No behavior change.
* WebProcess/WebPage/EventDispatcher.cpp:
(WebKit::EventDispatcher::wheelEvent):
(WebKit::EventDispatcher::dispatchWheelEventViaMainThread):
(WebKit::EventDispatcher::dispatchWheelEvent):
* WebProcess/WebPage/EventDispatcher.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::handleWheelEvent):
(WebKit::WebPage::wheelEvent):
* WebProcess/WebPage/WebPage.h:
2020-10-13 Alex Christensen <achristensen@webkit.org>
Allow legacy custom scheme registration on non-main threads
https://bugs.webkit.org/show_bug.cgi?id=217656
<rdar://problem/70228036>
Reviewed by Brady Eidson.
r267802 was quite necessary, but it exposed an internal client calling our API on a background thread,
which crashes now. Make it not crash. There is a race condition in their registration, but there already was.
* UIProcess/API/Cocoa/WKBrowsingContextController.mm:
(+[WKBrowsingContextController registerSchemeForCustomProtocol:]):
(+[WKBrowsingContextController unregisterSchemeForCustomProtocol:]):
2020-10-13 Kimmo Kinnunen <kkinnunen@apple.com>
Cocoa: Make WebGLLayer not dependent on GraphicsContextGLOpenGL
https://bugs.webkit.org/show_bug.cgi?id=217212
<rdar://problem/69876022>
Reviewed by Dean Jackson.
Removed unneeded inclusions of GraphicsContextGLOpenGL.h. The
file is now using non-public headers.
* WebProcess/WebPage/WebPage.cpp:
* WebProcess/WebPage/mac/WebPageMac.mm:
2020-10-12 Devin Rousso <drousso@apple.com>
CSS "any-pointer:fine" media query false on iPad/Pencil
https://bugs.webkit.org/show_bug.cgi?id=212580
<rdar://problem/63813283>
Reviewed by Wenson Hsieh.
Use `+[UIScribbleInteraction isPencilInputExpected]` as a proxy for the presence of stylus
devices as there's no way of directly querying for connected stylus devices (especially if
there is no active connection). When the value is changed to `YES`, notify all WebProcess
immediately. When the value is changed to `NO`, use an `NSTimer` to delay notifying all
WebProcess for 10min in case the user starts using their stylus again.
For live-on/tweaking purposes this 10min timeout can be adjusted before the UIProcess is
created with the `WKStylusDeviceObserverChangeTimeInterval` key in `NSUserDefaults`.
* UIProcess/ios/WKStylusDeviceObserver.h: Added.
* UIProcess/ios/WKStylusDeviceObserver.mm: Added.
(+[WKStylusDeviceObserver sharedInstance]):
(-[WKStylusDeviceObserver init]):
(-[WKStylusDeviceObserver setHasStylusDevice:]):
(-[WKStylusDeviceObserver start]):
(-[WKStylusDeviceObserver stop]):
(-[WKStylusDeviceObserver observeValueForKeyPath:ofObject:change:context:]):
(-[WKStylusDeviceObserver startChangeTimer:]):
(-[WKStylusDeviceObserver changeTimerFired:]):
Use ObjC KVO to listen for changes to `+[UIScribbleInteraction isPencilInputExpected]`.
* UIProcess/WebProcessProxy.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::notifyHasStylusDeviceChanged): Added.
* UIProcess/ios/WebProcessProxyIOS.mm:
(WebKit::WebProcessProxy::platformInitialize):
(WebKit::WebProcessProxy::platformDestroy):
Tell the shared `WKStylusDeviceObserver` to start listening when creating the WebProcess.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
* Shared/WebProcessCreationParameters.h:
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
Use the current state of the shared `WKStylusDeviceObserver` when creating WebProcess
instead of sending a `SetHasStylusDevice` message immediately after the WebProcess is
created.
* WebProcess/WebProcess.messages.in:
* WebProcess/WebProcess.h:
(WebKit::WebProcess::hasStylusDevice const): Added.
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::setHasStylusDevice): Added.
(WebKit::WebProcess::initializeWebProcess):
Update style (including `any-pointer`) whenever stylus devices change.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::hasStylusDevice): Added.
(WebKit::WebPage::pointerCharacteristicsOfAllAvailablePointingDevices const):
Include `fine` when there is a stylus device.
* UIProcess/ios/WKMouseDeviceObserver.h:
* UIProcess/ios/WKMouseDeviceObserver.mm:
(-[WKMouseDeviceObserver start]):
(-[WKMouseDeviceObserver stop]):
(-[WKMouseDeviceObserver dealloc]): Removed.
Drive-by: use a count of `start`/`stop` instead of `WebProcessProxy::allProcesses` size.
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
2020-10-12 Darin Adler <darin@apple.com>
Eliminate some "using namespace WebCore" cases in global scope
https://bugs.webkit.org/show_bug.cgi?id=217644
Reviewed by Tim Horton.
* NetworkProcess/ServiceWorker/ServiceWorkerFetchTask.cpp:
* NetworkProcess/cocoa/WebSocketTaskCocoa.mm:
* WebProcess/Automation/WebAutomationSessionProxy.cpp:
Move "using namespace WebCore" inside WebKit namespace.
2020-10-12 Tim Horton <timothy_horton@apple.com>
Fix the macCatalyst build ("reference to 'Rect' is ambiguous")
https://bugs.webkit.org/show_bug.cgi?id=217642
<rdar://problem/70224629>
Reviewed by Wenson Hsieh.
* UIProcess/ios/forms/WKDateTimeInputControl.mm:
(-[WKDateTimePicker initWithView:datePickerMode:]):
(-[WKDateTimePicker shouldPresentGregorianCalendar:]):
(-[WKDateTimePicker controlBeginEditing]):
(-[WKDateTimeInputControl initWithView:]):
An earlier unified source's `using namespace WebCore` inside `namespace WebKit`
and then WKDateTimeInputControl's `using namespace WebKit` in the global namespace
resulted in once-again conflicting definitions of `Rect` between CSS
code and MacTypes.h.
2020-10-12 Andy Estes <aestes@apple.com>
Unreviewed build fix after r268367.
* Configurations/WebKit.xcconfig: Corrected the definition of
WK_PDFKIT_LDFLAGS_maccatalyst.
2020-10-12 Tim Horton <timothy_horton@apple.com>
Cursor window is transparent - but not hidden - when `cursor: none` is specified
https://bugs.webkit.org/show_bug.cgi?id=217639
<rdar://problem/70006186>
Reviewed by Simon Fraser.
* Platform/spi/mac/AppKitSPI.h:
* UIProcess/mac/PageClientImplMac.mm:
(WebKit::PageClientImpl::setCursor):
Adopt new AppKit SPI to actually hide the cursor when we set it to the "none" shape.
This hiding lives until the next time the app-global cursor shape is changed by any client.]
We'll re-set it the next time we set the "none" shape.
This is important, because on some platforms, hiding the cursor is a performance improvement
over simply making it transparent.
2020-10-12 Said Abou-Hallawa <sabouhallawa@apple.com>
[GPU Process]: Replace "ImageBufferIdentifier" by a generic identifier named "RemoteResourceIdentifier"
https://bugs.webkit.org/show_bug.cgi?id=217550
Reviewed by Simon Fraser.
GPU Process needs a universal identifier for drawing cached resources and
drawing buffers and layers. The meaning of RemoteResourceIdentifier will
depend on the context it is used in.
* GPUProcess/graphics/RemoteImageBufferMessageHandlerProxy.cpp:
(WebKit::RemoteImageBufferMessageHandlerProxy::RemoteImageBufferMessageHandlerProxy):
(WebKit::RemoteImageBufferMessageHandlerProxy::createBackend):
(WebKit::RemoteImageBufferMessageHandlerProxy::commitFlushContext):
* GPUProcess/graphics/RemoteImageBufferMessageHandlerProxy.h:
* GPUProcess/graphics/RemoteImageBufferProxy.h:
(WebKit::RemoteImageBufferProxy::create):
(WebKit::RemoteImageBufferProxy::RemoteImageBufferProxy):
* GPUProcess/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::createImageBuffer):
(WebKit::RemoteRenderingBackendProxy::releaseRemoteResource):
(WebKit::RemoteRenderingBackendProxy::getImageData):
(WebKit::RemoteRenderingBackendProxy::flushImageBufferDrawingContext):
(WebKit::RemoteRenderingBackendProxy::flushImageBufferDrawingContextAndCommit):
(WebKit::RemoteRenderingBackendProxy::releaseImageBuffer): Deleted.
* GPUProcess/graphics/RemoteRenderingBackendProxy.h:
* GPUProcess/graphics/RemoteRenderingBackendProxy.messages.in:
* Scripts/webkit/messages.py:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/graphics/RemoteImageBufferMessageHandler.cpp:
(WebKit::RemoteImageBufferMessageHandler::RemoteImageBufferMessageHandler):
(WebKit::RemoteImageBufferMessageHandler::~RemoteImageBufferMessageHandler):
(WebKit::RemoteImageBufferMessageHandler::getImageData const):
(WebKit::RemoteImageBufferMessageHandler::flushDrawingContext):
(WebKit::RemoteImageBufferMessageHandler::flushDrawingContextAndWaitCommit):
* WebProcess/GPU/graphics/RemoteImageBufferMessageHandler.h:
(WebKit::RemoteImageBufferMessageHandler::remoteResourceIdentifier const):
(WebKit::RemoteImageBufferMessageHandler::imageBufferIdentifier const): Deleted.
* WebProcess/GPU/graphics/RemoteRenderingBackend.cpp:
(WebKit::RemoteRenderingBackend::createImageBuffer):
(WebKit::RemoteRenderingBackend::releaseRemoteResource):
(WebKit::RemoteRenderingBackend::createImageBufferBackend):
(WebKit::RemoteRenderingBackend::commitImageBufferFlushContext):
(WebKit::RemoteRenderingBackend::releaseImageBuffer): Deleted.
* WebProcess/GPU/graphics/RemoteRenderingBackend.h:
* WebProcess/GPU/graphics/RemoteRenderingBackend.messages.in:
2020-10-12 Luming Yin <luming_yin@apple.com>
[macOS] Workaround for MAC_OS_X_VERSION_MAJOR incorrectly including minor version when building
with Xcode 12 on macOS Big Sur SUs
https://bugs.webkit.org/show_bug.cgi?id=217602
rdar://70194453
Reviewed by Darin Adler.
The previous workaround turns out to be ineffective because we can't set the value of
TARGET_MAC_OS_X_VERSION_MAJOR based on a previous value of itself. Introduce a new
variable TARGET_MAC_OS_X_VERSION_MAJOR to determine whether we need to explicitly
adjust MAC_OS_X_VERSION_MAJOR to 110000.
* Configurations/DebugRelease.xcconfig:
2020-10-12 Andy Estes <aestes@apple.com>
[macCatalyst] Enable WKPDFView
https://bugs.webkit.org/show_bug.cgi?id=217403
<rdar://problem/48217791>
Reviewed by Tim Horton.
* Configurations/WebKit.xcconfig: Linked PDFKit.framework when building for the Catalyst
variant on Big Sur or later.
2020-10-12 Youenn Fablet <youenn@apple.com>
Add support for MediaRecorder bitrate getters
https://bugs.webkit.org/show_bug.cgi?id=217383
Reviewed by Eric Carlson.
Pipe up bit rates through IPC
* GPUProcess/webrtc/RemoteMediaRecorder.h:
(WebKit::RemoteMediaRecorder::audioBitRate const):
(WebKit::RemoteMediaRecorder::videoBitRate const):
* GPUProcess/webrtc/RemoteMediaRecorderManager.cpp:
(WebKit::RemoteMediaRecorderManager::createRecorder):
* GPUProcess/webrtc/RemoteMediaRecorderManager.h:
* GPUProcess/webrtc/RemoteMediaRecorderManager.messages.in:
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
(WebKit::MediaRecorderPrivate::startRecording):
2020-10-12 Luming Yin <luming_yin@apple.com>
[macOS] Workaround for MAC_OS_X_VERSION_MAJOR incorrectly including minor version when building
with Xcode 12 on macOS Big Sur SUs
https://bugs.webkit.org/show_bug.cgi?id=217602
rdar://70194453
Reviewed by Darin Adler.
Due to a bug in Xcode (rdar://70185899), Xcode 12.0 and Xcode 12.1 Beta incorrectly includes the
minor release number in MAC_OS_X_VERSION_MAJOR, which causes Debug and Release builds of WebKit
to be misconfigured when building on macOS Big Sur SUs, leading to webpages failing to load.
To work around the Xcode bug, when the MAC_OS_X_VERSION_MAJOR includes the minor version number,
drop the minor version number by explicitly setting TARGET_MAC_OS_X_VERSION_MAJOR to 110000.
Note: This change should be reverted after <rdar://70185899> is resolved.
* Configurations/DebugRelease.xcconfig:
2020-10-12 Alex Christensen <achristensen@webkit.org>
DownloadProxy::cancel should take a CompletionHandler
https://bugs.webkit.org/show_bug.cgi?id=217547
Reviewed by Youenn Fablet.
This paves the way for a new download cancel API that looks like this:
- (void)cancelByProducingResumeData:(void (^)(NSData *resumeData))completionHandler;
In order to do this, we push the legacy didCancel client call into the legacy API implementations.
The resume data will be accessible through this completion handler and the didFail callback
instead of querying the DownloadProxy wrapper for it, so I renamed it to legacyResumeData.
No change in behavior, as verified by existing tests.
* UIProcess/API/APIDownloadClient.h:
(API::DownloadClient::legacyDidCancel):
(API::DownloadClient::didCancel): Deleted.
* UIProcess/API/C/WKContext.cpp:
(WKContextSetDownloadClient):
* UIProcess/API/C/WKDownload.cpp:
(WKDownloadGetResumeData):
(WKDownloadCancel):
* UIProcess/API/Cocoa/_WKDownload.mm:
(-[_WKDownload cancel]):
(-[_WKDownload resumeData]):
* UIProcess/API/glib/WebKitDownload.cpp:
(webkit_download_cancel):
* UIProcess/API/glib/WebKitDownloadClient.cpp:
* UIProcess/Cocoa/LegacyDownloadClient.h:
* UIProcess/Cocoa/LegacyDownloadClient.mm:
(WebKit::LegacyDownloadClient::legacyDidCancel):
(WebKit::LegacyDownloadClient::didCancel): Deleted.
* UIProcess/Downloads/DownloadProxy.cpp:
(WebKit::DownloadProxy::cancel):
(WebKit::DownloadProxy::didFail):
* UIProcess/Downloads/DownloadProxy.h:
(WebKit::DownloadProxy::legacyResumeData const):
(WebKit::DownloadProxy::client):
(WebKit::DownloadProxy::resumeData const): Deleted.
2020-10-11 Sam Weinig <weinig@apple.com>
[Preferences] Introduce string based SPI for WKPreferences to allow tests to change internal behavior without always adding additional SPI
https://bugs.webkit.org/show_bug.cgi?id=217582
Reviewed by Darin Adler.
Add SPI for setting any WebKit preference using the key as defined in the WebPreferences*.yaml
files. This will allow adding testing of non-default behavior that we don't necessarily want to
expose via its own API or SPI.
* CMakeLists.txt:
* DerivedSources-input.xcfilelist:
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* Scripts/PreferencesTemplates/WebPreferencesGetterSetters.cpp.erb: Added.
* WebKit.xcodeproj/project.pbxproj:
Add new generated file, WebPreferencesGetterSetters.cpp.
* Scripts/PreferencesTemplates/WebPreferencesExperimentalFeatures.cpp.erb:
* Scripts/PreferencesTemplates/WebPreferencesInternalDebugFeatures.cpp.erb:
Moved getter/setter definitions to WebPreferencesGetterSetters.cpp.erb
* UIProcess/WebPreferences.cpp:
* UIProcess/WebPreferences.h:
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetBoolValueForKey):
(WKPreferencesSetDoubleValueForKey):
(WKPreferencesSetUInt32ValueForKey):
(WKPreferencesSetStringValueForKey):
* UIProcess/API/C/WKPreferencesRefPrivate.h:
Expose new SPI.
2020-10-11 Luming Yin <luming_yin@apple.com>
Strip patch version from TARGET_MAC_OS_X_VERSION_MAJOR when building for macOS Big Sur
or later
https://bugs.webkit.org/show_bug.cgi?id=217594
rdar://70188497
Reviewed by Darin Adler.
To ensure successful Mac Catalyst WebKit builds, strip the patch version from
TARGET_MAC_OS_X_VERSION_MAJOR by using two `base:`s on MACOSX_DEPLOYMENT_TARGET.
* Configurations/Base.xcconfig:
2020-10-11 Michael Catanzaro <mcatanzaro@gnome.org>
-Wreorder in MediaPlayerPrivateRemote.cpp
https://bugs.webkit.org/show_bug.cgi?id=217590
Reviewed by Eric Carlson.
Reorder initializer list to avoid misleading initialization order.
It might be nicer to actually reorder the member variables in the class itself, which would
probably be safe, but that would be a behavior change and I'd rather avoid changing
behavior.
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::MediaPlayerPrivateRemote):
2020-10-11 Luming Yin <luming_yin@apple.com>
Ignore deployment suffix and identifier when computing major OS version for macOS
Big Sur and newer
https://bugs.webkit.org/show_bug.cgi?id=217584
rdar://70168426
Reviewed by Darin Adler.
Stop using MACOSX_DEPLOYMENT_TARGET:suffix:identifier to compute major OS versions.
Only use the deployment target base for macOS Big Sur and newer. Keep the manual
definitions for legacy versions of macOS.
* Configurations/Base.xcconfig:
2020-10-10 Sam Weinig <weinig@apple.com>
Use WebPreference definitions from shared configuration files in WebCore (Part 1)
https://bugs.webkit.org/show_bug.cgi?id=217551
Reviewed by Darin Adler.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::sendViewportAttributesChanged):
Add explicit type to support layoutFallbackWidth now being an unsigned int
in Settings (it was already an unsigned int at the WebKit level).
2020-10-10 Adrian Perez de Castro <aperez@igalia.com>
[GTK] Build broken with ENABLE_GAMEPAD enabled
https://bugs.webkit.org/show_bug.cgi?id=217226
Reviewed by Darin Adler.
* UIProcess/Gamepad/UIGamepadProvider.h: Mark class as final, because the destructor is
final itself.
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::numberOfConnectedGamepadsForTesting): Reorganize PLATFORM(MAC)
guards to use GameControllerGamepadProvider only on macOS, as it is unavailable on other
ports.
* WebProcess/Gamepad/WebGamepadProvider.h: Mark class as final, because the destructor
is final itself.
2020-10-09 Wenson Hsieh <wenson_hsieh@apple.com>
[GPU Process] Add additional support for painting video elements to 2D contexts
https://bugs.webkit.org/show_bug.cgi?id=217532
Reviewed by Tim Horton.
See WebCore/ChangeLog for more details.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::shouldUseRemoteRenderingFor):
2020-10-09 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process] Enable audio routing arbitration on Mac
https://bugs.webkit.org/show_bug.cgi?id=217528
Reviewed by Eric Carlson.
Use the audio routing arbitration implementation in the Web process and UI process
before moving AudioSessionRoutingArbitratorProxy to the GPU process (which will
require some sandbox changes).
* WebProcess/GPU/media/RemoteAudioSession.cpp:
(WebKit::RemoteAudioSession::setCategory):
(WebKit::RemoteAudioSession::category const):
* WebProcess/GPU/media/RemoteAudioSession.h:
2020-10-09 Ryan Haddad <ryanhaddad@apple.com>
Promote WKWebView media playback SPI to API
https://bugs.webkit.org/show_bug.cgi?id=217335
Unreviewed build fix for macCatalyst.
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::isUserFacingChanged):
2020-10-09 Simon Fraser <simon.fraser@apple.com>
Refactor wheel event coalescing into its own class
https://bugs.webkit.org/show_bug.cgi?id=217523
Reviewed by Sam Weinig.
WebPageProxy had some confusing code for coalescing wheel events, first added in r116466.
To prepare this code for partial re-use in another place, factor it into its own class, WebWheelEventCoalescer.
WebWheelEventCoalescer holds a queue of pending NativeWebWheelEvents, and a queue of sequences of
NativeWebWheelEvents, where each sequence has been coalesced into a single WebWeelEvent which was
sent to the Web Process.
When the UI process receives a message back from the Web Process, it calls takeOldestEventBeingProcessed()
which removes the first sequence from m_eventsBeingProcessed. This code has to track NativeWebWheelEvents, because
we need to dig the NSEvent out of the NativeWebWheelEvent at this stage.
* Platform/Logging.h: Add a WheelEvents log channel.
* Shared/NativeWebWheelEvent.h:
* Shared/WebWheelEventCoalescer.cpp: Added.
(WebKit::operator<<):
(WebKit::WebWheelEventCoalescer::canCoalesce):
(WebKit::WebWheelEventCoalescer::coalesce):
(WebKit::WebWheelEventCoalescer::shouldDispatchEventNow const):
(WebKit::WebWheelEventCoalescer::nextEventToDispatch):
(WebKit::WebWheelEventCoalescer::shouldDispatchEvent):
(WebKit::WebWheelEventCoalescer::takeOldestEventBeingProcessed):
(WebKit::WebWheelEventCoalescer::clear):
* Shared/WebWheelEventCoalescer.h: Added.
(WebKit::WebWheelEventCoalescer::hasEventsBeingProcessed const):
* Sources.txt:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::handleWheelEvent):
(WebKit::WebPageProxy::wheelEventCoalescer):
(WebKit::WebPageProxy::didReceiveEvent):
(WebKit::WebPageProxy::resetStateAfterProcessExited):
(WebKit::canCoalesce): Deleted.
(WebKit::coalesce): Deleted.
(WebKit::coalescedWheelEvent): Deleted.
(WebKit::WebPageProxy::processNextQueuedWheelEvent): Deleted.
(WebKit::WebPageProxy::shouldProcessWheelEventNow const): Deleted.
* UIProcess/WebPageProxy.h:
* WebKit.xcodeproj/project.pbxproj:
2020-10-09 Per Arne Vollan <pvollan@apple.com>
[GPU Process] Use correct GPU runtime settings when creating the dynamic IOKit sandbox extensions for the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=217491
<rdar://problem/70066087>
Reviewed by Youenn Fablet.
Currently, the creation of the dynamic IOKit sandbox extensions for the WebContent process is incorrectly using the same GPU
runtime settings as when creating Media related dynamic mach-lookup extensions. There are several other GPU runtime flags
that needs to be checked before creating the extensions.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
2020-10-09 Chris Fleizach <cfleizach@apple.com>
AX: Mark malware safe browser as a header for VoiceOver
https://bugs.webkit.org/show_bug.cgi?id=217505
Reviewed by Sam Weinig.
Mark this UI element as a header for VoiceOver on iOS.
* UIProcess/Cocoa/WKSafeBrowsingWarning.mm:
(makeLabel):
2020-10-09 Alex Christensen <achristensen@webkit.org>
Change RELEASE_ASSERT to ASSERT in networkProcessesSet
https://bugs.webkit.org/show_bug.cgi?id=217524
<rdar://problem/70115841>
Reviewed by Tim Horton.
* UIProcess/Network/NetworkProcessProxy.cpp:
2020-10-09 Per Arne Vollan <pvollan@apple.com>
[macOS] Issue sandbox extension for icon service when attachment element is enabled
https://bugs.webkit.org/show_bug.cgi?id=217483
<rdar://problem/70099680>
Reviewed by Brent Fulgham.
This has previously been done on iOS, and should be done on macOS as well.
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
* UIProcess/WebProcessProxy.h:
* WebProcess/WebPage/WebPage.cpp:
* WebProcess/com.apple.WebProcess.sb.in:
2020-10-09 Alex Christensen <achristensen@webkit.org>
Merge DownloadProxy::DidReceiveResponse and DownloadProxy::DecideDestinationWithSuggestedFilename into one message
https://bugs.webkit.org/show_bug.cgi?id=217519
Reviewed by Youenn Fablet.
No change in behavior. The current API has them as separate callbacks so we can have them as separate messages,
but future API probably won't, so let's merge them.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::findPendingDownloadLocation):
* UIProcess/API/APIDownloadClient.h:
(API::DownloadClient::didReceiveAuthenticationChallenge):
(API::DownloadClient::decideDestinationWithSuggestedFilename):
(API::DownloadClient::didReceiveResponse): Deleted.
* UIProcess/API/C/WKContext.cpp:
(WKContextSetDownloadClient):
* UIProcess/Cocoa/LegacyDownloadClient.h:
* UIProcess/Cocoa/LegacyDownloadClient.mm:
(WebKit::LegacyDownloadClient::decideDestinationWithSuggestedFilename):
* UIProcess/Downloads/DownloadProxy.cpp:
(WebKit::DownloadProxy::decideDestinationWithSuggestedFilename):
(WebKit::DownloadProxy::didReceiveResponse): Deleted.
* UIProcess/Downloads/DownloadProxy.h:
* UIProcess/Downloads/DownloadProxy.messages.in:
2020-10-09 Kate Cheney <katherine_cheney@apple.com>
Promote WKWebView media playback SPI to API
https://bugs.webkit.org/show_bug.cgi?id=217335
<rdar://problem/63406100>
Reviewed by Jer Noble.
Currently the only way to interact with video in WKWebView is by injecting
JavaScript into a page. It would be useful to promote various media
playback SPI to API to make this easier for clients, including adding
completion handlers to the new API calls. Along with this,
we should add new API to check for media playback state.
This patch keeps the SPI around for compatibility.
* Shared/API/Cocoa/WebKit.h:
* Shared/MediaPlaybackState.h: Added.
* UIProcess/API/Cocoa/WKMediaPlaybackState.h: Added.
* UIProcess/API/Cocoa/WKWebView.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView closeAllMediaPresentations]):
(-[WKWebView pauseAllMediaPlayback:]):
(-[WKWebView suspendAllMediaPlayback:]):
(-[WKWebView resumeAllMediaPlayback:]):
(toWKMediaPlaybackState):
(-[WKWebView requestMediaPlaybackState:]):
(-[WKWebView _closeAllMediaPresentations]):
(-[WKWebView _stopAllMediaPlayback]):
(-[WKWebView _suspendAllMediaPlayback]):
(-[WKWebView _resumeAllMediaPlayback]):
* UIProcess/ios/fullscreen/WKFullScreenViewController.mm:
(-[WKFullScreenViewController _showPhishingAlert]):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::requestMediaPlaybackState):
(WebKit::WebPageProxy::pauseAllMediaPlayback):
(WebKit::WebPageProxy::suspendAllMediaPlayback):
(WebKit::WebPageProxy::resumeAllMediaPlayback):
(WebKit::WebPageProxy::stopAllMediaPlayback): Deleted.
* UIProcess/WebPageProxy.h:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::requestMediaPlaybackState):
(WebKit::WebPage::pauseAllMediaPlayback):
(WebKit::WebPage::suspendAllMediaPlayback):
(WebKit::WebPage::resumeAllMediaPlayback):
(WebKit::WebPage::stopAllMediaPlayback): Deleted.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-10-09 Alex Christensen <achristensen@webkit.org>
Use sendWithAsyncReply for NetworkProcess::CancelDownload
https://bugs.webkit.org/show_bug.cgi?id=217420
Reviewed by Youenn Fablet.
This paves the way for a good download API that has a completion handler when you call cancel that gives you the resume data if it can.
Currently you have to call cancel on the download, wait for didCancel on the delegate, then get the resume data from the download again.
This has the side effect of DownloadMonitor cancellations calling didFailWithError instead of didCancel, which makes sense and shouldn't
cause any problems in practice. Safari is the only API client that uses this, and it behaves similarly in didCancel and didFailWithError.
* NetworkProcess/Downloads/Download.cpp:
(WebKit::Download::cancel):
(WebKit::Download::didCancel):
* NetworkProcess/Downloads/Download.h:
* NetworkProcess/Downloads/DownloadManager.cpp:
(WebKit::DownloadManager::cancelDownload):
* NetworkProcess/Downloads/DownloadManager.h:
* NetworkProcess/Downloads/DownloadMonitor.cpp:
(WebKit::DownloadMonitor::timerFired):
* NetworkProcess/Downloads/PendingDownload.cpp:
(WebKit::PendingDownload::cancel):
* NetworkProcess/Downloads/PendingDownload.h:
* NetworkProcess/Downloads/cocoa/WKDownloadProgress.mm:
(-[WKDownloadProgress performCancel]):
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::cancelDownload):
(WebKit::NetworkProcess::findPendingDownloadLocation):
(WebKit::NetworkProcess::pendingDownloadCanceled): Deleted.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* UIProcess/Downloads/DownloadProxy.cpp:
(WebKit::createData):
(WebKit::DownloadProxy::cancel):
(WebKit::DownloadProxy::didCancel): Deleted.
* UIProcess/Downloads/DownloadProxy.h:
* UIProcess/Downloads/DownloadProxy.messages.in:
2020-10-09 Youenn Fablet <youenn@apple.com>
Switch Allow/Don't Allow buttons in getUserMedia prompt
https://bugs.webkit.org/show_bug.cgi?id=217513
<rdar://problem/70134429>
Reviewed by Eric Carlson.
Return NotAllowedError instead of SecurityError in case user denies with default prompt.
Make sure to have the Allow button in iOS prompt on the right instead of the left.
Tested in MiniBrowser and MobileMiniBrowser.
* UIProcess/Cocoa/UserMediaPermissionRequestProxy.mm:
(WebKit::UserMediaPermissionRequestProxy::doDefaultAction):
2020-10-09 Youenn Fablet <youenn@apple.com>
Implement TCC checks in UserMediaPermissionRequestManagerProxy
https://bugs.webkit.org/show_bug.cgi?id=217508
<rdar://problem/70129993>
Reviewed by Eric Carlson.
TCC checks must be done before asking the page client to decide whether to grant or not camera and/or microphone.
This ensures that, should the default action be done (default prompt), for instance in case the WKWebView application does not provide any delegate,
the TCC checks will happen before the default prompt.
For that reason, we move the TCC checks from UIDelegate::UIClient::decidePolicyForUserMediaPermissionRequest to UserMediaPermissionRequestManagerProxy::requestSystemValidation.
Manually tested in Safari and MiniBrowser.
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::UIClient::decidePolicyForUserMediaPermissionRequest):
(WebKit::requestUserMediaAuthorizationForFrame): Deleted.
(WebKit::requestAVCaptureAccessForMediaType): Deleted.
* UIProcess/Cocoa/UserMediaPermissionRequestManagerProxy.mm:
(WebKit::requestAVCaptureAccessForMediaType):
(WebKit::UserMediaPermissionRequestManagerProxy::requestSystemValidation):
* UIProcess/UserMediaPermissionRequestManagerProxy.cpp:
(WebKit::UserMediaPermissionRequestManagerProxy::processUserMediaPermissionValidRequest):
(WebKit::UserMediaPermissionRequestManagerProxy::decidePolicyForUserMediaPermissionRequest):
(WebKit::UserMediaPermissionRequestManagerProxy::requestSystemValidation):
* UIProcess/UserMediaPermissionRequestManagerProxy.h:
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::preferences const):
2020-10-08 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Implement a dummy WebAuthenticationAgent
https://bugs.webkit.org/show_bug.cgi?id=217401
<rdar://problem/70012011>
Reviewed by Darin Adler.
Part 1.
This patch introduces a new dummy WebAuthenticationAgent where in the future it will take over all WebAuthn operations
from the UI Process such that we can isolate high privileged entitlements to this standalone daemon and therefore are able
offer WebAuthn to third party WKWebView clients. One of the future feautres will require this new process to listen to
LaunchEvents, which only daemons can do. That's why it is implemented as a user agent instead of a XPC service.
This is the first part of establishing such a dummy daemon. What it does is to set up a new build target for the daemon and
hook it up with WebKit's build system. One can manually load the launchd plist and load the daemon to verify this change.
Commands are:
launchctl load /to/your/path/com.apple.webkit.WebAuthenticationAgent.plist
launchctl start com.apple.webkit.WebAuthenticationAgent.Development
Covered by manual tests.
* Configurations/WebAuthenticationAgent.xcconfig: Added.
Somehow the executable has a suffix .Development for debug/release builds. Will fix that in the later part.
* Daemons/WebAuthenticationAgent/Info.plist: Added.
* Daemons/WebAuthenticationAgent/WebAuthenticationAgent.entitlements: Added.
* Daemons/WebAuthenticationAgent/com.apple.webkit.WebAuthenticationAgent.plist: Added.
The executable location is hardcoded for now. It will be fixed in Part 2 to dynamically point to the right binary.
* Daemons/WebAuthenticationAgent/main.mm: Added.
(main):
* WebKit.xcodeproj/project.pbxproj:
Not sure why libWTF.a is needed. Will fix that in the later part.
2020-10-08 Chris Fleizach <cfleizach@apple.com>
AX: Website Warning page is inaccessible
https://bugs.webkit.org/show_bug.cgi?id=217489
Reviewed by Zalan Bujtas.
If we have a safe browsing window, we need to use that instead of the content view.
* UIProcess/Cocoa/WKSafeBrowsingWarning.h:
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::accessibilityAttributeValue):
2020-10-08 Alex Christensen <achristensen@webkit.org>
REGRESSION (r267763): SHOULD NEVER BE REACHED in WebKit::contentRuleListsFromIdentifier
https://bugs.webkit.org/show_bug.cgi?id=217499
There was an invalid assertion. NBD.
* UIProcess/WebProcessProxy.cpp:
(WebKit::contentRuleListsFromIdentifier):
2020-10-08 Ryosuke Niwa <rniwa@webkit.org>
Make it possible to send an arbitrary IPC message from JavaScript
https://bugs.webkit.org/show_bug.cgi?id=217423
<rdar://problem/69969351>
Reviewed by Geoffrey Garen.
This patch introduces the JavaScript API (window.IPC) to send IPC out of WebContent process.
The feature is compiled in under ASAN and Debug builds and can be enabled at runtime.
window.IPC has two methods: sendMessage and sendSyncMessage which sends an async and sync IPC respectively.
It takes the destination process name (UI, GPU, or Networking), the destination ID (e.g. WebPageProxy ID),
message ID, timeout for sendSyncMessage, and optionally IPC message arguments. The message arguments can be
passed in as a TypedArray or ArrayBuffer, or a JavaScript array that recursively describes encoded objects.
Each object can be either a TypedArray or ArrayBuffer, which will be treated as encoded message, an array
which will be encoded as a Vector with each item within the array encoded recursively, or a dictionary which
describes a specific type.
When a specific type is described via a dictionary, "value" is encoed based on "type" as follows:
- When "type" is "String", "value" is encoded as a WTF::String, treating null or undefined as a null string.
- When "type" is "bool", "int8_t", "int16_t", "int32_t", "int64_t", "uint8_t", "uint16_t", "uint32_t",
or "uint64_t", "value" (which can be BigInt or a number) is encoded as the respective C++ type.
- When "type" is "RGBA", "value" is used as PackedColor::RGBA to construct WebCore::Color to be encoded.
- When "type" is "IntRect" or "FloatRect", "x", "y", "width", and "height" are treated as respective values
of IntRect or FloatRect C++ objects, and the constructed *Rect is encoded.
- When "type" is "FrameInfoData", the context object's WebFrame's FrameInfoData is encoded.
The list of IPC messages are exposed on window.IPC.messages, and VisitedLinkStore ID, WebPageProxy ID,
and frame identifiers are also exposed as static variables on window.IPC.
* Sources.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::dispatchDidClearWindowObjectInWorld): Inject the API if enabled.
* WebProcess/WebPage/IPCTestingAPI.cpp: Added.
(WebKit::IPCTestingAPI::JSIPC::create): Added.
(WebKit::IPCTestingAPI::JSIPC::webFrame): Added.
(WebKit::IPCTestingAPI::JSIPC::JSIPC): Added.
(WebKit::IPCTestingAPI::JSIPC::wrapperClass): Added.
(WebKit::IPCTestingAPI::JSIPC::unwrap): Added.
(WebKit::IPCTestingAPI::JSIPC::toWrapped): Added.
(WebKit::IPCTestingAPI::JSIPC::initialize): Added.
(WebKit::IPCTestingAPI::JSIPC::finalize): Added.
(WebKit::IPCTestingAPI::JSIPC::staticFunctions): Added.
(WebKit::IPCTestingAPI::JSIPC::staticValues): Added.
(WebKit::IPCTestingAPI::convertToUint64): Added.
(WebKit::IPCTestingAPI::processTargetFromArgument): Added.
(WebKit::IPCTestingAPI::destinationIDFromArgument): Added.
(WebKit::IPCTestingAPI::messageIDFromArgument): Added.
(WebKit::IPCTestingAPI::encodeTypedArray): Added.
(WebKit::IPCTestingAPI::createTypeError): Added.
(WebKit::IPCTestingAPI::encodeRectType): Added.
(WebKit::IPCTestingAPI::encodeIntegralType): Added.
(WebKit::IPCTestingAPI::VectorEncodeHelper::encode const): Added.
(WebKit::IPCTestingAPI::encodeArgument): Added.
(WebKit::IPCTestingAPI::JSIPC::sendMessage): Added.
(WebKit::IPCTestingAPI::JSIPC::sendSyncMessage): Added.
(WebKit::IPCTestingAPI::JSIPC::visitedLinkStoreID): Added.
(WebKit::IPCTestingAPI::JSIPC::webPageProxyID): Added.
(WebKit::IPCTestingAPI::JSIPC::frameIdentifier): Added.
(WebKit::IPCTestingAPI::JSIPC::retrieveID): Added.
(WebKit::IPCTestingAPI::JSIPC::messages): Added.
(WebKit::IPCTestingAPI::inject):
* WebProcess/WebPage/IPCTestingAPI.h: Added.
* WebProcess/WebPage/WebFrame.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
(WebKit::WebPage::updatePreferences):
* WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::ipcTestingAPIEnabled const):
(WebKit::WebPage::webPageProxyID const):
(WebKit::WebPage::visitedLinkTableID const):
2020-10-08 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process] Cannot activate or deactivate an audio session
https://bugs.webkit.org/show_bug.cgi?id=217492
Reviewed by Eric Carlson.
No new tests, covered by existing tests.
* WebProcess/GPU/media/RemoteAudioSession.cpp:
(WebKit::RemoteAudioSession::tryToSetActiveInternal):
2020-10-08 Hector Lopez <hector_i_lopez@apple.com>
Unreviewed, reverting r268155.
This reverts r268155 becasue it broke internal builds
Reverted changeset:
"[WebAuthn] Implement a dummy WebAuthenticationAgent"
https://bugs.webkit.org/show_bug.cgi?id=217401
https://trac.webkit.org/changeset/268155
2020-10-08 Alex Christensen <achristensen@webkit.org>
REGRESSION (r267763): [ iOS wk2 ] http/tests/in-app-browser-privacy/non-app-bound-domain-does-not-get-app-bound-session.html is a constant failure
https://bugs.webkit.org/show_bug.cgi?id=217386
Reviewed by Brent Fulgham.
Use SPI instead of NSUserDefaults to turn on test mode for app browser privacy.
This is more reliable with the timing change of when a WebsiteDataStore's parameters are sent to the network process.
* UIProcess/API/Cocoa/_WKWebsiteDataStoreConfiguration.h:
* UIProcess/API/Cocoa/_WKWebsiteDataStoreConfiguration.mm:
(-[_WKWebsiteDataStoreConfiguration enableInAppBrowserPrivacyForTesting]):
(-[_WKWebsiteDataStoreConfiguration setEnableInAppBrowserPrivacyForTesting:]):
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::WebsiteDataStore::platformSetNetworkParameters):
(WebKit::WebsiteDataStore::ensureAppBoundDomains const):
* UIProcess/WebsiteData/WebsiteDataStore.h:
* UIProcess/WebsiteData/WebsiteDataStoreConfiguration.cpp:
(WebKit::WebsiteDataStoreConfiguration::copy const):
* UIProcess/WebsiteData/WebsiteDataStoreConfiguration.h:
(WebKit::WebsiteDataStoreConfiguration::enableInAppBrowserPrivacyForTesting const):
(WebKit::WebsiteDataStoreConfiguration::setEnableInAppBrowserPrivacyForTesting):
2020-10-08 Chris Dumez <cdumez@apple.com>
Crash under WebKit::WebProcessCache::clear()
https://bugs.webkit.org/show_bug.cgi?id=217480
Reviewed by Geoffrey Garen.
Protect |process| in the responsiveness check lambda in WebProcessCache::addProcessIfPossible().
If we fail to do so and WebProcessCache::clear() gets called while the responsiveness check is
pending, the WebProcessProxy destructor may get called while clear() clears m_pendingAddRequests,
which would resolve the responsiveness check with responsive=false, and cause the lambda to
try and remove the entry from m_pendingAddRequests (while clear() is clearing it).
* UIProcess/WebProcessCache.cpp:
(WebKit::WebProcessCache::addProcessIfPossible):
2020-10-08 David Quesada <david_quesada@apple.com>
WKWebViewConfiguration._shouldRelaxThirdPartyCookieBlocking should be available on iOS
https://bugs.webkit.org/show_bug.cgi?id=217443
rdar://problem/69900089
Reviewed by Youenn Fablet.
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration _setShouldRelaxThirdPartyCookieBlocking:]):
Allow this method to be used by Safari and SafariViewService on iOS-family platforms.
2020-10-08 Peng Liu <peng.liu6@apple.com>
RemoteAudioSession::category() returns an incorrect value after category changing
https://bugs.webkit.org/show_bug.cgi?id=217459
Reviewed by Youenn Fablet.
No new tests, covered by existing tests.
* WebProcess/GPU/media/RemoteAudioSession.cpp:
(WebKit::RemoteAudioSession::setCategory): Update `m_configuration.category`
and `m_configuration.routeSharingPolicy` so that RemoteAudioSession::category()
will return the correct value.
* WebProcess/GPU/media/RemoteAudioSession.h:
2020-10-08 Aditya Keerthi <akeerthi@apple.com>
[iOS] Blurring a text input and focusing a date input changes the scrollable area
https://bugs.webkit.org/show_bug.cgi?id=217325
<rdar://problem/69979171>
Reviewed by Wenson Hsieh.
iOS 14 changed the presentation style of date/time pickers on iPhones.
Previously, date/time pickers would show up in the keyboard view. Now,
pickers are shown in UIContextMenus.
When the change was made, shouldShowKeyboardForElement in WKContentViewInteraction
was not updated to reflect the new behavior, and continued to return
true for date/time inputs.
This inaccuracy led to WKContentView ignoring the UIKeyboardWillHideNotification
when blurring a text input and focusing a date input, since it assumed
the keyboard would stay presented. Consequently, the UIScrollView
adjustments made when presenting the keyboard for the text input are not
reverted when presenting the date picker, leading to a permanent change
in the scrollable area.
To fix, update shouldShowKeyboardForElement to return false for date/time
inputs. Note that the original patch which changed the behavior
correctly updated [WKContentView requiresAccessoryView] but did not
update shouldShowKeyboardForElement. To prevent a similar mistake from
happening again, the logic which determines which input types have a
keyboard view was moved into elementTypeRequiresAccessoryView.
Test: fast/forms/ios/constant-scroll-area-when-moving-focus-between-fields.html
* UIProcess/ios/WKContentViewInteraction.mm:
(elementTypeRequiresAccessoryView):
(-[WKContentView requiresAccessoryView]):
(shouldShowKeyboardForElement):
2020-10-08 Per Arne Vollan <pvollan@apple.com>
[macOS] Deny access to performance analysis service in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=217452
<rdar://problem/68878318>
Reviewed by Youenn Fablet.
It has been confirmed that mach-lookup to 'com.apple.PerformanceAnalysis.animationperfd' can be denied in the WebContent process on macOS.
Test: fast/sandbox/sandbox-mach-lookup.html
* WebProcess/com.apple.WebProcess.sb.in:
2020-10-08 Saagar Jha <saagar@saagarjha.com>
Replace unqualified use of PlatformDisplayID in WebViewImpl
https://bugs.webkit.org/show_bug.cgi?id=217422
Reviewed by Youenn Fablet.
WebViewImpl utilizes things from WebCore but doesn't use the namespace, instead choosing to
qualify each reference. However, there is one use of WebCore::PlatformDisplayID that
doesn't use the the namespace qualifier. Usually this "works" because the code is in the
WebKit namespace, and other files that get included above this one often contain code of
the form
namespace WebKit {
using WebCore;
// code
}
which hides the issue by making PlatformDisplayID visible to this file as well without the
prefix. If we use auto none of this matters because it'll deduce the prefix even if
WebViewImpl.mm ends up near the top of a unified source file, before another file with the
using declaration shows up.
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::windowDidChangeScreen): Replace PlatformDisplayID with auto.
2020-10-07 Joonghun Park <jh718.park@samsung.com>
Unreviewed. Remove the build warning below since r238330.
warning: unused variable ‘connectionIdentifier’ [-Wunused-variable]
No new tests, no new behaviors.
* WebProcess/Inspector/WebInspectorUI.cpp:
(WebKit::WebInspectorUI::updateConnection):
2020-10-07 Keith Rollin <krollin@apple.com>
Update post-processing rules for headers to not unnecessarily change timestamps
https://bugs.webkit.org/show_bug.cgi?id=217371
<rdar://problem/69992230>
Reviewed by Darin Adler.
Under XCBuild, the scripts employed in custom build rules can be
invoked in innocuous situations. A common example is when the user is
building from the command-line and they change the `make` output from
stdout to a file, or vice-versa. Changing the output changes the
setting of the COLOR_DIAGNOSTICS environment variable, which is enough
to cause XCBuild to think something is different and that the custom
build rule needs to be invoked. For the script's part, nothing
significant has changed, yet it post-processes the header files,
causing their modification dates to change, causing downstream
rebuilds to occur.
Fix this problem by adopting an approach that doesn't modify the
post-processed header files unless their contents actually change.
No new tests -- this is a build change only.
* Scripts/postprocess-header-rule:
2020-10-07 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Implement a dummy WebAuthenticationAgent
https://bugs.webkit.org/show_bug.cgi?id=217401
<rdar://problem/70012011>
Reviewed by Darin Adler.
Part 1.
This patch introduces a new dummy WebAuthenticationAgent where in the future it will take over all WebAuthn operations
from the UI Process such that we can isolate high privileged entitlements to this standalone daemon and therefore are able
offer WebAuthn to third party WKWebView clients. One of the future feautres will require this new process to listen to
LaunchEvents, which only daemons can do. That's why it is implemented as a user agent instead of a XPC service.
This is the first part of establishing such a dummy daemon. What it does is to set up a new build target for the daemon and
hook it up with WebKit's build system. One can manually load the launchd plist and load the daemon to verify this change.
Commands are:
launchctl load /to/your/path/com.apple.webkit.WebAuthenticationAgent.plist
launchctl start com.apple.webkit.WebAuthenticationAgent.Development
Covered by manual tests.
* Configurations/WebAuthenticationAgent.xcconfig: Added.
Somehow the executable has a suffix .Development for debug/release builds. Will fix that in the later part.
* Daemons/WebAuthenticationAgent/Info.plist: Added.
* Daemons/WebAuthenticationAgent/WebAuthenticationAgent.entitlements: Added.
* Daemons/WebAuthenticationAgent/com.apple.webkit.WebAuthenticationAgent.plist: Added.
The executable location is hardcoded for now. It will be fixed in Part 2 to dynamically point to the right binary.
* Daemons/WebAuthenticationAgent/main.mm: Added.
(main):
* WebKit.xcodeproj/project.pbxproj:
Not sure why libWTF.a is needed. Will fix that in the later part.
2020-10-07 Devin Rousso <drousso@apple.com>
Add missing `#define` for `PENCILKIT_TEXT_INPUT` flag
https://bugs.webkit.org/show_bug.cgi?id=217436
<rdar://problem/69720219>
Reviewed by Wenson Hsieh.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView setUpInteraction]):
(-[WKContentView cleanUpInteraction]):
(-[WKContentView _shouldSimulateKeyboardInputOnTextInsertion]):
Change from `ENABLE` to `HAVE`.
2020-10-07 Wenson Hsieh <wenson_hsieh@apple.com>
[GPU Process] Support CanvasRenderingContext2D.drawImage() with HTMLVideoElement
https://bugs.webkit.org/show_bug.cgi?id=217339
<rdar://problem/69409029>
Reviewed by Darin Adler.
Implements support for painting the current video frame into a canvas 2D graphics context. See below (and
Source/WebCore/ChangeLog) for more details.
* GPUProcess/graphics/RemoteImageBufferMessageHandlerProxy.h:
(WebKit::RemoteImageBufferMessageHandlerProxy::backend):
* GPUProcess/graphics/RemoteImageBufferProxy.h:
(WebKit::RemoteImageBufferProxy::apply):
Refactor `apply` so that it treats `PutImageData` and `PaintFrameForMedia` as special cases, and otherwise
returns `false` by default. Implement `apply` in the case where the display list item is `PaintFrameForMedia` by
using the `MediaPlayerIdentifier` to look up the platform `MediaPlayer` instance, and then calling
`GraphicsContext::paintFrameForMedia` with this `MediaPlayer`.
* GPUProcess/graphics/RemoteRenderingBackendProxy.cpp:
(WebKit::RemoteRenderingBackendProxy::gpuConnectionToWebProcess const):
* GPUProcess/graphics/RemoteRenderingBackendProxy.h:
2020-10-07 Youenn Fablet <youenn@apple.com>
Add support for BlobEvent.timecode
https://bugs.webkit.org/show_bug.cgi?id=217379
Reviewed by Eric Carlson.
Exchange timecode through IPC.
* GPUProcess/webrtc/RemoteMediaRecorder.cpp:
(WebKit::RemoteMediaRecorder::fetchData):
* GPUProcess/webrtc/RemoteMediaRecorder.h:
* GPUProcess/webrtc/RemoteMediaRecorder.messages.in:
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
(WebKit::MediaRecorderPrivate::fetchData):
* WebProcess/GPU/webrtc/MediaRecorderPrivate.h:
2020-10-07 Youenn Fablet <youenn@apple.com>
Add support for MediaRecorder pause/resume
https://bugs.webkit.org/show_bug.cgi?id=217375
Reviewed by Eric Carlson.
Add IPC support for sending pause/resume orders.
* GPUProcess/webrtc/RemoteMediaRecorder.cpp:
(WebKit::RemoteMediaRecorder::pause):
(WebKit::RemoteMediaRecorder::resume):
* GPUProcess/webrtc/RemoteMediaRecorder.h:
* GPUProcess/webrtc/RemoteMediaRecorder.messages.in:
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
(WebKit::MediaRecorderPrivate::pauseRecording):
(WebKit::MediaRecorderPrivate::resumeRecording):
* WebProcess/GPU/webrtc/MediaRecorderPrivate.h:
2020-10-07 Per Arne Vollan <pvollan@apple.com>
[macOS] Remove 'com.apple.cookied' from the WebContent process sandbox.
https://bugs.webkit.org/show_bug.cgi?id=206088
<rdar://problem/58489040>
Reviewed by Darin Adler.
It has been confirmed that access to the cookied service can be removed from the WebContent process.
Test: fast/sandbox/sandbox-mach-lookup.html
* WebProcess/com.apple.WebProcess.sb.in:
2020-10-07 Aditya Keerthi <akeerthi@apple.com>
zappos.com: Dropdown - 'Quantity' option size different
https://bugs.webkit.org/show_bug.cgi?id=217395
<rdar://problem/60261396>
Reviewed by Tim Horton.
The select element used in this instance uses a custom font (Open Sans)
with a 16px font size. When presenting the select dropdown a call to
WebPopupMenuProxyMac::showPopupMenu is made with those attributes.
However, since Open Sans is not a system font, the call to
[NSFont fontWithDescriptor:size:] returns nil. This results in the
presented view falling back to the default font size, which is smaller
than the expected 16px size, leading to smaller options.
While other browsers do not honor custom fonts in their select dropdowns,
they do honor the font size. Consequently, this dropdown appears
different in Safari. To fix, ensure that the font size is always honored
if the corresponding attribute is present.
* UIProcess/mac/WebPopupMenuProxyMac.mm:
(WebKit::WebPopupMenuProxyMac::showPopupMenu):
2020-10-07 Youenn Fablet <youenn@apple.com>
Make sure to fire the correct set of events in case MediaRecorder stream has track changes
https://bugs.webkit.org/show_bug.cgi?id=217316
Reviewed by Eric Carlson.
Make sure to not set audio/video tracks if the recorder was stopped.
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
(WebKit::MediaRecorderPrivate::startRecording):
(WebKit::MediaRecorderPrivate::~MediaRecorderPrivate):
(WebKit::MediaRecorderPrivate::stopRecording):
* WebProcess/GPU/webrtc/MediaRecorderPrivate.h:
2020-10-06 Peng Liu <peng.liu6@apple.com>
[Media in GPU process] Don't use 'id' as a parameter or variable name
https://bugs.webkit.org/show_bug.cgi?id=206189
Reviewed by Tim Horton.
Mass rename `id` to `identifier`.
No new tests, no functional change.
* GPUProcess/media/RemoteAudioDestinationManager.cpp:
(WebKit::RemoteAudioDestination::create):
(WebKit::RemoteAudioDestination::RemoteAudioDestination):
(WebKit::RemoteAudioDestinationManager::deleteAudioDestination):
(WebKit::RemoteAudioDestinationManager::startAudioDestination):
(WebKit::RemoteAudioDestinationManager::stopAudioDestination):
* GPUProcess/media/RemoteAudioDestinationManager.messages.in:
* GPUProcess/media/RemoteAudioTrackProxy.cpp:
(WebKit::RemoteAudioTrackProxy::RemoteAudioTrackProxy):
(WebKit::RemoteAudioTrackProxy::configuration):
* GPUProcess/media/RemoteAudioTrackProxy.h:
* GPUProcess/media/RemoteCDMFactoryProxy.cpp:
(WebKit::RemoteCDMFactoryProxy::addProxy):
(WebKit::RemoteCDMFactoryProxy::removeProxy):
(WebKit::RemoteCDMFactoryProxy::addInstance):
(WebKit::RemoteCDMFactoryProxy::removeInstance):
(WebKit::RemoteCDMFactoryProxy::getInstance):
(WebKit::RemoteCDMFactoryProxy::addSession):
(WebKit::RemoteCDMFactoryProxy::removeSession):
* GPUProcess/media/RemoteCDMFactoryProxy.messages.in:
* GPUProcess/media/RemoteCDMInstanceProxy.messages.in:
* GPUProcess/media/RemoteCDMProxy.messages.in:
* GPUProcess/media/RemoteLegacyCDMFactoryProxy.cpp:
(WebKit::RemoteLegacyCDMFactoryProxy::addProxy):
(WebKit::RemoteLegacyCDMFactoryProxy::removeProxy):
(WebKit::RemoteLegacyCDMFactoryProxy::addSession):
(WebKit::RemoteLegacyCDMFactoryProxy::removeSession):
(WebKit::RemoteLegacyCDMFactoryProxy::getSession const):
* GPUProcess/media/RemoteLegacyCDMFactoryProxy.messages.in:
* GPUProcess/media/RemoteLegacyCDMProxy.messages.in:
* GPUProcess/media/RemoteMediaPlayerManagerProxy.cpp:
(WebKit::RemoteMediaPlayerManagerProxy::createMediaPlayer):
(WebKit::RemoteMediaPlayerManagerProxy::deleteMediaPlayer):
(WebKit::RemoteMediaPlayerManagerProxy::getProxy):
* GPUProcess/media/RemoteMediaPlayerManagerProxy.messages.in:
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
(WebKit::RemoteMediaPlayerProxy::RemoteMediaPlayerProxy):
* GPUProcess/media/RemoteMediaResource.cpp:
(WebKit::RemoteMediaResource::create):
(WebKit::RemoteMediaResource::RemoteMediaResource):
* GPUProcess/media/RemoteMediaResourceManager.cpp:
(WebKit::RemoteMediaResourceManager::responseReceived):
(WebKit::RemoteMediaResourceManager::redirectReceived):
(WebKit::RemoteMediaResourceManager::dataSent):
(WebKit::RemoteMediaResourceManager::dataReceived):
(WebKit::RemoteMediaResourceManager::accessControlCheckFailed):
(WebKit::RemoteMediaResourceManager::loadFailed):
(WebKit::RemoteMediaResourceManager::loadFinished):
* GPUProcess/media/RemoteMediaResourceManager.messages.in:
* GPUProcess/media/RemoteTextTrackProxy.cpp:
(WebKit::RemoteTextTrackProxy::RemoteTextTrackProxy):
(WebKit::RemoteTextTrackProxy::configuration):
* GPUProcess/media/RemoteTextTrackProxy.h:
* GPUProcess/media/RemoteVideoTrackProxy.cpp:
(WebKit::RemoteVideoTrackProxy::RemoteVideoTrackProxy):
(WebKit::RemoteVideoTrackProxy::configuration):
* GPUProcess/media/RemoteVideoTrackProxy.h:
* GPUProcess/media/TextTrackPrivateRemoteConfiguration.h:
(WebKit::TextTrackPrivateRemoteConfiguration::encode const):
(WebKit::TextTrackPrivateRemoteConfiguration::decode):
* GPUProcess/media/TrackPrivateRemoteConfiguration.h:
(WebKit::TrackPrivateRemoteConfiguration::encode const):
(WebKit::TrackPrivateRemoteConfiguration::decode):
* WebProcess/GPU/media/AudioTrackPrivateRemote.cpp:
(WebKit::AudioTrackPrivateRemote::updateConfiguration):
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::remoteTextTrackConfigurationChanged):
(WebKit::MediaPlayerPrivateRemote::removeRemoteVideoTrack):
(WebKit::MediaPlayerPrivateRemote::remoteVideoTrackConfigurationChanged):
* WebProcess/GPU/media/RemoteCDM.cpp:
(WebKit::RemoteCDM::create):
(WebKit::RemoteCDM::RemoteCDM):
(WebKit::RemoteCDM::createInstance):
* WebProcess/GPU/media/RemoteCDMFactory.cpp:
(WebKit::RemoteCDMFactory::createCDM):
(WebKit::RemoteCDMFactory::removeSession):
* WebProcess/GPU/media/RemoteCDMInstance.cpp:
(WebKit::RemoteCDMInstance::create):
(WebKit::RemoteCDMInstance::RemoteCDMInstance):
(WebKit::RemoteCDMInstance::createSession):
* WebProcess/GPU/media/RemoteCDMInstanceSession.cpp:
(WebKit::RemoteCDMInstanceSession::create):
(WebKit::RemoteCDMInstanceSession::RemoteCDMInstanceSession):
* WebProcess/GPU/media/RemoteLegacyCDM.cpp:
(WebKit::RemoteLegacyCDM::create):
(WebKit::RemoteLegacyCDM::RemoteLegacyCDM):
(WebKit::RemoteLegacyCDM::createSession):
(WebKit::RemoteLegacyCDM::setPlayerId):
* WebProcess/GPU/media/RemoteLegacyCDMFactory.cpp:
(WebKit::RemoteLegacyCDMFactory::createCDM):
(WebKit::RemoteLegacyCDMFactory::addSession):
(WebKit::RemoteLegacyCDMFactory::removeSession):
* WebProcess/GPU/media/RemoteLegacyCDMSession.cpp:
(WebKit::RemoteLegacyCDMSession::create):
(WebKit::RemoteLegacyCDMSession::RemoteLegacyCDMSession):
* WebProcess/GPU/media/RemoteMediaPlayerManager.cpp:
(WebKit::RemoteMediaPlayerManager::deleteRemoteMediaPlayer):
* WebProcess/GPU/media/RemoteMediaResourceProxy.cpp:
(WebKit::RemoteMediaResourceProxy::RemoteMediaResourceProxy):
* WebProcess/GPU/media/TextTrackPrivateRemote.cpp:
(WebKit::TextTrackPrivateRemote::updateConfiguration):
* WebProcess/GPU/media/VideoTrackPrivateRemote.cpp:
(WebKit::VideoTrackPrivateRemote::updateConfiguration):
2020-10-06 Devin Rousso <drousso@apple.com>
Move platform logic for CSS `hover` and `pointer` media queries outside of WebCore
https://bugs.webkit.org/show_bug.cgi?id=217413
Reviewed by Tim Horton.
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::hoverSupportedByPrimaryPointingDevice const): Added.
(WebKit::WebChromeClient::pointerCharacteristicsOfPrimaryPointingDevices const): Added.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/gtk/WebPageGtk.cpp:
(WebKit::WebPage::hoverSupportedByPrimaryPointingDevice const): Added.
(WebKit::WebPage::pointerCharacteristicsOfPrimaryPointingDevices const): Added.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::hoverSupportedByPrimaryPointingDevice const): Added.
(WebKit::WebPage::pointerCharacteristicsOfPrimaryPointingDevices const): Added.
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::hoverSupportedByPrimaryPointingDevice const): Added.
(WebKit::WebPage::pointerCharacteristicsOfPrimaryPointingDevices const): Added.
* WebProcess/WebPage/playstation/WebPagePlayStation.cpp:
(WebKit::WebPage::hoverSupportedByPrimaryPointingDevice const): Added.
(WebKit::WebPage::pointerCharacteristicsOfPrimaryPointingDevices const): Added.
* WebProcess/WebPage/win/WebPageWin.cpp:
(WebKit::WebPage::hoverSupportedByPrimaryPointingDevice const): Added.
(WebKit::WebPage::pointerCharacteristicsOfPrimaryPointingDevices const): Added.
* WebProcess/WebPage/wpe/WebPageWPE.cpp:
(WebKit::WebPage::hoverSupportedByPrimaryPointingDevice const): Added.
(WebKit::WebPage::pointerCharacteristicsOfPrimaryPointingDevices const): Added.
Provide a way for WebCore to get information about the primary pointing device.
Refactor logic in `WebCore::anyHoverEvaluate` and `WebCore::anyPointerEvaluate` to move
platform code into the respective `WebPage` file.
2020-10-06 Chris Dumez <cdumez@apple.com>
Reloading a view in its processTerminationHandler does not work reliably when using related views
https://bugs.webkit.org/show_bug.cgi?id=217407
Reviewed by Geoff Garen.
Related web views share the same WebContent process. When this process crashes, we iterate over
the list of WebPageProxy objects sharing this process and let them know that their process has
crashed. This causes the WebPageProxy to reset its state (so that it is aware it no longer has
a running process) and to notify the client application. Because we were notifying the client
application synchronously, the client could trigger a load in the view synchronously while we
are still iterating over the WebPageProxy objects. When triggering a load in a web view that
has no running process, we normally relaunch one. However, in the case of related web views,
the view uses its related webview's process, relaunching it if necessary. The 'relaunching
if necessary' part was not reliably happening here because the related web view may not have
been notified yet that its WebProcess has crashed (since we are still iterating over the
pages to notify them).
To address the issue, we now notify the client asynchronously of the process termination.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::dispatchProcessDidTerminate):
2020-10-06 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process] The seekable attribute of HTMLMediaElement has an incorrect value
https://bugs.webkit.org/show_bug.cgi?id=217410
Reviewed by Eric Carlson.
Covered by existing tests.
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
(WebKit::RemoteMediaPlayerProxy::mediaPlayerReadyStateChanged):
Update the state variables related to the `seekable` attribute when the media player is ready.
(WebKit::RemoteMediaPlayerProxy::updateCachedState):
Add a missing state variable setting.
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::seekable const): Deleted.
* WebProcess/GPU/media/MediaPlayerPrivateRemote.h:
Instead of overriding seekable() here, we should use MediaPlayerPrivateInterface::seekable(),
which has the correct behavior.
2020-10-06 Devin Rousso <drousso@apple.com>
CSS hover and "pointer: fine" media queries do not evaluate to true with iOS 13.4 mouse support
https://bugs.webkit.org/show_bug.cgi?id=209292
<rdar://problem/60663124>
Reviewed by Tim Horton.
On iOS, the primary pointer will always be touch input, so the `hover`/`pointer` media query
should never change. If a mouse is connected, however, `any-hover`/`any-pointer` should
change as now there is at least one device that supports `hover`/`fine`. Note that in the
case of `any-pointer` this means that both `coarse` and `fine` will apply. Catalyst assumes
that there is always a mouse connected.
Spec: https://drafts.csswg.org/mediaqueries-4/#any-input
* UIProcess/ios/WKMouseDeviceObserver.h: Added.
* UIProcess/ios/WKMouseDeviceObserver.mm: Added.
(+[WKMouseDeviceObserver sharedInstance]):
(-[WKMouseDeviceObserver dealloc]):
(-[WKMouseDeviceObserver start]):
(-[WKMouseDeviceObserver stop]):
(-[WKMouseDeviceObserver mousePointerDevicesDidChange:]):
(-[WKMouseDeviceObserver _setHasMouseDeviceForTesting:]):
Use `BackBoardServices` to listen for mouse devices changes and notify all WebProcess.
* UIProcess/WebProcessProxy.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::WebProcessProxy):
(WebKit::WebProcessProxy::platformInitialize): Added.
(WebKit::WebProcessProxy::~WebProcessProxy):
(WebKit::WebProcessProxy::platformDestroy): Added.
* UIProcess/ios/WebProcessProxyIOS.mm:
(WebKit::WebProcessProxy::platformInitialize): Added.
(WebKit::WebProcessProxy::platformDestroy): Added.
(WebKit::WebProcessProxy::notifyHasMouseDeviceChanged): Added.
Tell the shared `WKMouseDeviceObserver` to start listening for mouse device changes when
creating the WebProcess so that the shared `WKMouseDeviceObserver` knows to notify all
instances when mouse devices change. Make `allProcesses` into a `private static` member
function so that it can also be used in `UIProcess/ios/WebProcessProxyIOS.mm`.
* Shared/WebProcessCreationParameters.h:
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
Use the current state of the shared `WKMouseDeviceObserver` when creating WebProcess instead
of sending a `SetHasMouseDevice` message immediately after the WebProcess is created.
* WebProcess/WebProcess.messages.in:
* WebProcess/WebProcess.h:
(WebKit::WebProcess::hasMouseDevice const): Added.
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::setHasMouseDevice): Added.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
Update style (including `any-hover` and `any-pointer`) whenever mouse devices change.
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::hoverSupportedByAnyAvailablePointingDevice const): Added.
(WebKit::WebChromeClient::pointerCharacteristicsOfAllAvailablePointingDevices const): Added.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/gtk/WebPageGtk.cpp:
(WebKit::WebPage::hoverSupportedByAnyAvailablePointingDevice const): Added.
(WebKit::WebPage::pointerCharacteristicsOfAllAvailablePointingDevices const): Added.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::hasMouseDevice): Added.
(WebKit::WebPage::hoverSupportedByAnyAvailablePointingDevice const): Added.
(WebKit::WebPage::pointerCharacteristicsOfAllAvailablePointingDevices const): Added.
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::hoverSupportedByAnyAvailablePointingDevice const): Added.
(WebKit::WebPage::pointerCharacteristicsOfAllAvailablePointingDevices const): Added.
* WebProcess/WebPage/playstation/WebPagePlayStation.cpp:
(WebKit::WebPage::hoverSupportedByAnyAvailablePointingDevice const): Added.
(WebKit::WebPage::pointerCharacteristicsOfAllAvailablePointingDevices const): Added.
* WebProcess/WebPage/win/WebPageWin.cpp:
(WebKit::WebPage::hoverSupportedByAnyAvailablePointingDevice const): Added.
(WebKit::WebPage::pointerCharacteristicsOfAllAvailablePointingDevices const): Added.
* WebProcess/WebPage/wpe/WebPageWPE.cpp:
(WebKit::WebPage::hoverSupportedByAnyAvailablePointingDevice const): Added.
(WebKit::WebPage::pointerCharacteristicsOfAllAvailablePointingDevices const): Added.
Provide a way for WebCore to get the current state of whether there are any mouse devices.
Refactor logic in `WebCore::anyHoverEvaluate` and `WebCore::anyPointerEvaluate` to move
platform code into the respective `WebPage` file.
* Configurations/WebKit.xcconfig:
* Platform/spi/ios/BackBoardServicesSPI.h:
* Platform/spi/ios/BaseBoardSPI.h: Added.
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
2020-10-06 Simon Fraser <simon.fraser@apple.com>
Redundant rendering updates can be scheduled from inside Page::updateRendering()
https://bugs.webkit.org/show_bug.cgi?id=216726
Reviewed by Tim Horton.
Schedule for layer flush.
* WebProcess/WebPage/CoordinatedGraphics/LayerTreeHostTextureMapper.cpp:
(WebKit::LayerTreeHost::layerFlushTimerFired):
2020-10-06 Alex Christensen <achristensen@webkit.org>
Use sendWithAsyncReply to evaluate JavaScript in a WebPage
https://bugs.webkit.org/show_bug.cgi?id=217351
Reviewed by Youenn Fablet.
No change in meaningful behavior, but complexity is removed.
* UIProcess/API/C/WKPage.cpp:
(WKPageRunJavaScriptInMainFrame):
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _evaluateJavaScript:asAsyncFunction:withSourceURL:withArguments:forceUserGesture:inFrame:inWorld:completionHandler:]):
(-[WKWebView takeSnapshotWithConfiguration:completionHandler:]):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::runJavaScriptInMainFrame):
(WebKit::WebPageProxy::runJavaScriptInFrameInScriptWorld):
(WebKit::WebPageProxy::scriptValueCallback): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::runJavaScript):
(WebKit::WebPage::runJavaScriptInFrameInScriptWorld):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-10-06 Per Arne Vollan <pvollan@apple.com>
[Cocoa] The WebContent development XPC service is not allowed to connect to the XPC endpoint in the Networking process
https://bugs.webkit.org/show_bug.cgi?id=217389
<rdar://problem/68585497>
Reviewed by Brent Fulgham.
The WebContent development XPC service is not allowed to connect to the XPC endpoint in the Networking process,
because this requires a private entitlement. The WebContent development XPC service is not allowed to have private
entitlements. The WebContent process needs to connect to the XPC endpoint in the Networking process in order to
receive the Launch Services database XPC object. This patch addresses this issue by allowing platform binaries
to connect to the XPC endpoint.
* Shared/Cocoa/XPCEndpoint.mm:
(WebKit::XPCEndpoint::XPCEndpoint):
2020-10-06 Wenson Hsieh <wenson_hsieh@apple.com>
Rename MediaPlayerPrivateRemoteIdentifier to MediaPlayerIdentifier
https://bugs.webkit.org/show_bug.cgi?id=217397
Reviewed by Tim Horton.
Mass rename `MediaPlayerPrivateRemoteIdentifier` to `WebCore::MediaPlayerIdentifier`. See WebCore/ChangeLog for
more details.
* GPUProcess/media/RemoteLegacyCDMFactoryProxy.cpp:
(WebKit::RemoteLegacyCDMFactoryProxy::createCDM):
* GPUProcess/media/RemoteLegacyCDMFactoryProxy.h:
* GPUProcess/media/RemoteLegacyCDMFactoryProxy.messages.in:
* GPUProcess/media/RemoteLegacyCDMProxy.cpp:
(WebKit::RemoteLegacyCDMProxy::create):
(WebKit::RemoteLegacyCDMProxy::RemoteLegacyCDMProxy):
(WebKit::RemoteLegacyCDMProxy::setPlayerId):
* GPUProcess/media/RemoteLegacyCDMProxy.h:
* GPUProcess/media/RemoteLegacyCDMProxy.messages.in:
* GPUProcess/media/RemoteMediaPlayerManagerProxy.cpp:
(WebKit::RemoteMediaPlayerManagerProxy::createMediaPlayer):
(WebKit::RemoteMediaPlayerManagerProxy::deleteMediaPlayer):
(WebKit::RemoteMediaPlayerManagerProxy::didReceivePlayerMessage):
(WebKit::RemoteMediaPlayerManagerProxy::didReceiveSyncPlayerMessage):
(WebKit::RemoteMediaPlayerManagerProxy::getProxy):
* GPUProcess/media/RemoteMediaPlayerManagerProxy.h:
* GPUProcess/media/RemoteMediaPlayerManagerProxy.messages.in:
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
(WebKit::RemoteMediaPlayerProxy::RemoteMediaPlayerProxy):
* GPUProcess/media/RemoteMediaPlayerProxy.h:
* Scripts/webkit/messages.py:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::MediaPlayerPrivateRemote):
(WebKit::MediaPlayerPrivateRemote::identifier const):
* WebProcess/GPU/media/MediaPlayerPrivateRemote.h:
* WebProcess/GPU/media/RemoteLegacyCDM.cpp:
(WebKit::RemoteLegacyCDM::setPlayerId):
* WebProcess/GPU/media/RemoteLegacyCDM.h:
* WebProcess/GPU/media/RemoteLegacyCDMFactory.cpp:
(WebKit::RemoteLegacyCDMFactory::createCDM):
* WebProcess/GPU/media/RemoteMediaPlayerManager.cpp:
(WebKit::RemoteMediaPlayerManager::createRemoteMediaPlayer):
(WebKit::RemoteMediaPlayerManager::deleteRemoteMediaPlayer):
(WebKit::RemoteMediaPlayerManager::findRemotePlayerId):
(WebKit::RemoteMediaPlayerManager::didReceivePlayerMessage):
* WebProcess/GPU/media/RemoteMediaPlayerManager.h:
* WebProcess/GPU/media/cocoa/MediaPlayerPrivateRemoteCocoa.mm:
(WebKit::MediaPlayerPrivateRemote::MediaPlayerPrivateRemote):
2020-10-06 Alex Christensen <achristensen@webkit.org>
Move ownership of DownloadClient from WebProcessPool to DownloadProxy
https://bugs.webkit.org/show_bug.cgi?id=217399
Reviewed by Tim Horton.
To keep binary compatibility, the WebProcessPool has a nullable DownloadClient that is used if it has been set.
This will allow us to get a DownloadClient from elsewhere, such as having the API Download object have its own delegate
or having the WKWebsiteDataStore have a delegate, which are two of the solutions I've considered. We definitely don't
want it on the WebProcessPool, though.
* SourcesCocoa.txt:
* UIProcess/API/APIDownloadClient.h:
* UIProcess/API/C/WKContext.cpp:
(WKContextSetDownloadClient):
* UIProcess/API/Cocoa/WKProcessPool.mm:
(-[WKProcessPool _setDownloadDelegate:]):
* UIProcess/API/glib/WebKitDownloadClient.cpp:
(attachDownloadClientToContext):
* UIProcess/API/glib/WebKitWebContext.cpp:
(webkitWebContextDispose):
* UIProcess/Cocoa/LegacyDownloadClient.h: Renamed from Source/WebKit/UIProcess/Cocoa/DownloadClient.h.
* UIProcess/Cocoa/LegacyDownloadClient.mm: Renamed from Source/WebKit/UIProcess/Cocoa/DownloadClient.mm.
(WebKit::LegacyDownloadClient::LegacyDownloadClient):
(WebKit::LegacyDownloadClient::didStart):
(WebKit::LegacyDownloadClient::didReceiveResponse):
(WebKit::LegacyDownloadClient::didReceiveData):
(WebKit::LegacyDownloadClient::didReceiveAuthenticationChallenge):
(WebKit::LegacyDownloadClient::didCreateDestination):
(WebKit::LegacyDownloadClient::processDidCrash):
(WebKit::LegacyDownloadClient::decideDestinationWithSuggestedFilename):
(WebKit::LegacyDownloadClient::didFinish):
(WebKit::LegacyDownloadClient::didFail):
(WebKit::LegacyDownloadClient::didCancel):
(WebKit::LegacyDownloadClient::willSendRequest):
(WebKit::LegacyDownloadClient::takeActivityToken):
(WebKit::LegacyDownloadClient::releaseActivityTokenIfNecessary):
* UIProcess/Downloads/DownloadProxy.cpp:
(WebKit::DownloadProxy::DownloadProxy):
(WebKit::DownloadProxy::cancel):
(WebKit::DownloadProxy::invalidate):
(WebKit::DownloadProxy::processDidClose):
(WebKit::DownloadProxy::didStart):
(WebKit::DownloadProxy::didReceiveAuthenticationChallenge):
(WebKit::DownloadProxy::willSendRequest):
(WebKit::DownloadProxy::didReceiveResponse):
(WebKit::DownloadProxy::didReceiveData):
(WebKit::DownloadProxy::decideDestinationWithSuggestedFilename):
(WebKit::DownloadProxy::didCreateDestination):
(WebKit::DownloadProxy::didFinish):
(WebKit::DownloadProxy::didFail):
(WebKit::DownloadProxy::didCancel):
(WebKit::DownloadProxy::~DownloadProxy): Deleted.
* UIProcess/Downloads/DownloadProxy.h:
* UIProcess/Downloads/DownloadProxyMap.cpp:
(WebKit::DownloadProxyMap::createDownloadProxy):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::WebProcessPool):
(WebKit::WebProcessPool::setLegacyDownloadClient):
(WebKit::WebProcessPool::setDownloadClient): Deleted.
* UIProcess/WebProcessPool.h:
* WebKit.xcodeproj/project.pbxproj:
2020-10-06 Youenn Fablet <youenn@apple.com>
Enable video capture in WebProcess by default on MacOS
https://bugs.webkit.org/show_bug.cgi?id=217385
<rdar://problem/69999542>
Reviewed by Sam Weinig.
Manually tested in MiniBrowser.
* Shared/WebPreferencesDefaultValues.cpp:
* Shared/WebPreferencesDefaultValues.h:
2020-10-06 Simon Fraser <simon.fraser@apple.com>
Rename scheduleTimedRenderingUpdate() to scheduleRenderingUpdate() everywhere
https://bugs.webkit.org/show_bug.cgi?id=217367
Reviewed by Alex Christensen.
The previously named scheduleImmediateRenderingUpdate() is no more, so all rendering
updates that WebCore knows about are of the timed variety. So just call them
scheduleRenderingUpdate() everywhere.
* WebProcess/WebPage/CoordinatedGraphics/LayerTreeHostTextureMapper.cpp:
(WebKit::LayerTreeHost::layerFlushTimerFired):
2020-10-06 Adrian Perez de Castro <aperez@igalia.com>
REGRESSION(r267399): [WPE] Build broken with ENABLE_ACCESSIBILITY=OFF
https://bugs.webkit.org/show_bug.cgi?id=217380
Reviewed by Philippe Normand.
Before the ATK::Bridge target got introduced in r267399 a couple of CMake variables were
used, which were empty when the accessility support was disabled and nothing was being
added to the WebKit and WPEToolingBackends targets. Now that an imported target is used
it is needed to add it as a dependency conditionally to avoid using an undefined target
name.
No new tests needed.
* PlatformWPE.cmake: Use the ATK::Bridge target only when ENABLE_ACCESSIBILITY is set.
2020-10-06 Adrian Perez de Castro <aperez@igalia.com>
Non-unified build fixes, early October 2020 edition, take two
https://bugs.webkit.org/show_bug.cgi?id=217344
Unreviewed build fix.
* UIProcess/WebProcessProxy.cpp: Add missing WebAutomationSession.h header.
2020-10-05 Sihui Liu <sihui_liu@apple.com>
Break reference cycle with NSURLSessionDataTask in NetworkDataTaskCocoa
https://bugs.webkit.org/show_bug.cgi?id=217347
<rdar://problem/69483575>
Reviewed by Geoffrey Garen.
* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::NetworkDataTaskCocoa::applyCookiePolicyForThirdPartyCNAMECloaking):
2020-10-05 Simon Fraser <simon.fraser@apple.com>
Use the "triggerRenderingUpdate" terminology in ChromeClient
https://bugs.webkit.org/show_bug.cgi?id=217355
Reviewed by Tim Horton.
ChromeClient confusingly had both scheduleRenderingUpdate() and scheduleTimedRenderingUpdate().
The former is for triggering an update soon, and the latter is a normal schedule, used by
SVGImages because they have their own Page.
Rename the one that the WebKits have to implement to triggerRenderingUpdate().
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::triggerRenderingUpdate):
(WebKit::WebChromeClient::scheduleRenderingUpdate): Deleted.
* WebProcess/WebCoreSupport/WebChromeClient.h:
2020-10-05 Wenson Hsieh <wenson_hsieh@apple.com>
REGRESSION(r267915): Many layout tests are asserting under ~CompletionHandler()
https://bugs.webkit.org/show_bug.cgi?id=217345
<rdar://problem/69973775>
Reviewed by Tim Horton.
I did *not* intend to schedule an async task after every single image load in r267915. Apart from triggering
assertions on various tests, this also causes us to send excessive IPC messages to the UI process when loading
images, which can't be good for performance. Fix this by only calling into `computeAndSendEditDragSnapshot()`
when we've finished loading all dropped image elements.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::didFinishLoadingImageForElement):
2020-10-05 Brady Eidson <beidson@apple.com>
Change a PDF RELEASE_ASSERT to an early return.
rdar://problem/69805240 and https://bugs.webkit.org/show_bug.cgi?id=217346
Reviewed by Tim Horton.
No new tests (Wish they were possible)
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::installPDFDocument): This release assert is getting hit regularly.
Change to an early return while we continue tracking this down.
2020-10-05 Alex Christensen <achristensen@webkit.org>
Make DownloadID an ObjectIdentifier
https://bugs.webkit.org/show_bug.cgi?id=203962
Reviewed by Youenn Fablet and Geoffrey Garen.
Also deprecate the unused WKDownloadGetID because this identifier is internal.
* NetworkProcess/Downloads/Download.cpp:
(WebKit::Download::Download):
(WebKit::Download::didReceiveData):
(WebKit::Download::didFinish):
(WebKit::Download::didFail):
(WebKit::Download::didCancel):
(WebKit::Download::messageSenderDestinationID const):
* NetworkProcess/Downloads/DownloadID.h:
(WebKit::DownloadID::DownloadID): Deleted.
(WebKit::DownloadID::operator== const): Deleted.
(WebKit::DownloadID::operator!= const): Deleted.
(WebKit::DownloadID::downloadID const): Deleted.
(IPC::ArgumentCoder<WebKit::DownloadID>::encode): Deleted.
(IPC::ArgumentCoder<WebKit::DownloadID>::decode): Deleted.
(WTF::DownloadIDHash::hash): Deleted.
(WTF::DownloadIDHash::equal): Deleted.
(WTF::HashTraits<WebKit::DownloadID>::emptyValue): Deleted.
(WTF::HashTraits<WebKit::DownloadID>::constructDeletedValue): Deleted.
(WTF::HashTraits<WebKit::DownloadID>::isDeletedValue): Deleted.
* NetworkProcess/Downloads/DownloadMonitor.cpp:
(WebKit::DownloadMonitor::applicationWillEnterForeground):
(WebKit::DownloadMonitor::applicationDidEnterBackground):
(WebKit::DownloadMonitor::timerFired):
* NetworkProcess/Downloads/PendingDownload.cpp:
(WebKit::PendingDownload::messageSenderDestinationID const):
* NetworkProcess/Downloads/PendingDownload.h:
* NetworkProcess/NetworkDataTask.h:
(WebKit::NetworkDataTask::setPendingDownloadID):
(WebKit::NetworkDataTask::isDownload const):
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::pendingDownloadCanceled):
(WebKit::NetworkProcess::findPendingDownloadLocation):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkResourceLoader.cpp:
(WebKit::NetworkResourceLoader::convertToDownload):
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(-[WKNetworkSessionDelegate URLSession:task:didCompleteWithError:]):
(-[WKNetworkSessionDelegate URLSession:downloadTask:didFinishDownloadingToURL:]):
(-[WKNetworkSessionDelegate URLSession:downloadTask:didWriteData:totalBytesWritten:totalBytesExpectedToWrite:]):
(WebKit::NetworkSessionCocoa::continueDidReceiveChallenge):
* Scripts/webkit/messages.py:
* Shared/Authentication/AuthenticationManager.h:
* UIProcess/API/C/WKDownload.cpp:
(WKDownloadGetID):
* UIProcess/API/C/WKDownload.h:
* UIProcess/Downloads/DownloadProxy.cpp:
(WebKit::DownloadProxy::DownloadProxy):
(WebKit::generateDownloadID): Deleted.
* UIProcess/Downloads/DownloadProxy.h:
* UIProcess/Downloads/DownloadProxyMap.cpp:
(WebKit::DownloadProxyMap::createDownloadProxy):
(WebKit::DownloadProxyMap::downloadFinished):
(WebKit::DownloadProxyMap::invalidate):
* WebProcess/WebPage/WebFrame.cpp:
(WebKit::WebFrame::startDownload):
(WebKit::WebFrame::convertMainResourceLoadToDownload):
* WebProcess/WebPage/WebFrame.h:
* WebProcess/WebPage/WebPage.h:
2020-10-05 Alex Christensen <achristensen@webkit.org>
Fix tvOS and watchOS build
* NetworkProcess/NetworkProcess.cpp:
I noticed it was broken in an EWS build, and it is missing this include.
2020-10-05 Simon Fraser <simon.fraser@apple.com>
Rename DrawingArea::scheduleRenderingUpdate()
https://bugs.webkit.org/show_bug.cgi?id=217337
Reviewed by Tim Horton.
DrawingArea::scheduleRenderingUpdate() always schedules an "immediate" update (either
on a zero-delay timer, or in the current runloop). To make this more clear, change
the terminology to "triggerRenderingUpdate()."
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::scheduleRenderingUpdate):
* WebProcess/WebPage/CoordinatedGraphics/DrawingAreaCoordinatedGraphics.cpp:
(WebKit::DrawingAreaCoordinatedGraphics::triggerRenderingUpdate):
(WebKit::DrawingAreaCoordinatedGraphics::scheduleRenderingUpdate): Deleted.
* WebProcess/WebPage/CoordinatedGraphics/DrawingAreaCoordinatedGraphics.h:
* WebProcess/WebPage/DrawingArea.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDisplayRefreshMonitor.mm:
(WebKit::RemoteLayerTreeDisplayRefreshMonitor::requestRefreshCallback):
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.mm:
(WebKit::RemoteLayerTreeDrawingArea::setRootCompositingLayer):
(WebKit::RemoteLayerTreeDrawingArea::updateGeometry):
(WebKit::RemoteLayerTreeDrawingArea::setExposedContentRect):
(WebKit::RemoteLayerTreeDrawingArea::triggerRenderingUpdate):
(WebKit::RemoteLayerTreeDrawingArea::didUpdate):
(WebKit::RemoteLayerTreeDrawingArea::addTransactionCallbackID):
(WebKit::RemoteLayerTreeDrawingArea::scheduleRenderingUpdate): Deleted.
* WebProcess/WebPage/RemoteLayerTree/RemoteScrollingCoordinator.mm:
(WebKit::RemoteScrollingCoordinator::scheduleTreeStateCommit):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::scheduleFullEditorStateUpdate):
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::dynamicViewportSizeUpdate):
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.h:
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::triggerRenderingUpdate):
(WebKit::TiledCoreAnimationDrawingArea::attachViewOverlayGraphicsLayer):
(WebKit::TiledCoreAnimationDrawingArea::setShouldScaleViewToFitDocument):
(WebKit::TiledCoreAnimationDrawingArea::addTransactionCallbackID):
(WebKit::TiledCoreAnimationDrawingArea::activityStateDidChange):
(WebKit::TiledCoreAnimationDrawingArea::scheduleRenderingUpdate): Deleted.
2020-10-05 Eric Carlson <eric.carlson@apple.com>
Disable media capture if an app's entitlements won't allow access to capture devices
https://bugs.webkit.org/show_bug.cgi?id=217319
<rdar://problem/69956112>
Reviewed by Youenn Fablet.
No new tests, tested manually with MiniBrowser and Safari.
* UIProcess/Cocoa/UserMediaPermissionRequestManagerProxy.mm:
(WebKit::UserMediaPermissionRequestManagerProxy::permittedToCaptureAudio): If the
app is sandboxed, check for the entitlement required for audio capture. Don't
bother checking for usage string if TCC has already authorized capture.
(WebKit::UserMediaPermissionRequestManagerProxy::permittedToCaptureVideo): Ditto
for video capture.
* UIProcess/Cocoa/WebPreferencesCocoa.mm:
(WebKit::WebPreferences::platformInitializeStore): Initialize `mediaDevicesEnabled`
manually because its default value is always `false` in the WebProcess and may be
`true` in the UIProcess. WebPreferences assumes the default will always be the same
in both processes, and so has an optimization that only sends values to the
WebProcess if they differ from the default value. This means if the default in the
UIProcess is true it won't be pushed to the WebProcess, so capture won't work.
* Platform/spi/ios/TCCSPI.h: Update TCCAccessPreflightResult.
2020-10-05 Alex Christensen <achristensen@webkit.org>
Use sendWithAsyncReply for DecideDestinationWithSuggestedFilename
https://bugs.webkit.org/show_bug.cgi?id=209195
Reviewed by Tim Horton.
This makes the download flow more clear.
No change in behavior.
* NetworkProcess/Downloads/DownloadID.h:
* NetworkProcess/Downloads/DownloadManager.cpp:
(WebKit::DownloadManager::downloadDestinationDecided):
(WebKit::DownloadManager::cancelDownload):
(WebKit::DownloadManager::willDecidePendingDownloadDestination): Deleted.
(WebKit::DownloadManager::continueDecidePendingDownloadDestination): Deleted.
* NetworkProcess/Downloads/DownloadManager.h:
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::findPendingDownloadLocation):
(WebKit::NetworkProcess::continueDecidePendingDownloadDestination): Deleted.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* Scripts/webkit/messages.py:
* Shared/Authentication/AuthenticationManager.h:
* UIProcess/API/APIDownloadClient.h:
* UIProcess/Downloads/DownloadProxy.cpp:
(WebKit::DownloadProxy::decideDestinationWithSuggestedFilename):
(WebKit::DownloadProxy::decideDestinationWithSuggestedFilenameAsync): Deleted.
* UIProcess/Downloads/DownloadProxy.h:
* UIProcess/Downloads/DownloadProxy.messages.in:
2020-10-05 David Kilzer <ddkilzer@apple.com>
Make NetscapePlugInStreamLoaderClient a weak pointer on NetscapePlugInStreamLoader
<https://webkit.org/b/217237>
<rdar://problem/69776714>
Reviewed by Alex Christensen.
Making NetscapePlugInStreamLoaderClient inherit from
CanMakeWeakPtr<> caused PDFPlugin to fail to compile because one
of its parent classes already inherited from CanMakeWeakPtr<>.
To fix this, extract a PDFPluginStreamLoaderClient class from
PDFPlugin and link the lifetime of PDFPluginStreamLoaderClient
to PDFPlugin.
After moving the NetscapePlugInStreamLoaderClient virtual
methods from PDFPlugin to PDFPluginStreamLoaderClient, add some
simple methods to PDFPlugin so that PDFPluginStreamLoaderClient
doesn't access the instance variables directly. Also make use
of these new methods in other parts of PDFPlugin.
* WebProcess/Plugins/PDF/PDFPlugin.h:
(WebKit::PDFPlugin::documentFinishedLoading): Add.
(WebKit::PDFPlugin::identifierForLoader): Add.
(WebKit::PDFPlugin::removeOutstandingByteRangeRequest): Add.
(WebKit::PDFPlugin::PDFPluginStreamLoaderClient): Add class.
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::PDFPlugin):
(WebKit::PDFPlugin::receivedNonLinearizedPDFSentinel):
(WebKit::PDFPlugin::getResourceBytesAtPosition):
(WebKit::PDFPlugin::PDFPluginStreamLoaderClient::willSendRequest): Move from PDFPlugin.
(WebKit::PDFPlugin::PDFPluginStreamLoaderClient::didReceiveResponse): Ditto.
(WebKit::PDFPlugin::PDFPluginStreamLoaderClient::didReceiveData): Ditto.
(WebKit::PDFPlugin::PDFPluginStreamLoaderClient::didFail): Ditto.
(WebKit::PDFPlugin::PDFPluginStreamLoaderClient::didFinishLoading): Ditto.
(WebKit::PDFPlugin::byteRangeRequestForLoader):
(WebKit::PDFPlugin::forgetLoader):
(WebKit::PDFPlugin::willSendRequest): Move to PDFPluginStreamLoaderClient.
(WebKit::PDFPlugin::didReceiveResponse): Ditto.
(WebKit::PDFPlugin::didReceiveData): Ditto.
(WebKit::PDFPlugin::didFail): Ditto.
(WebKit::PDFPlugin::didFinishLoading): Ditto.
2020-10-05 Carlos Garcia Campos <cgarcia@igalia.com>
REGRESSION(r267763) [SOUP] Some tests flaky crashing in NetworkSessionSoup::soupSession()
https://bugs.webkit.org/show_bug.cgi?id=217163
Reviewed by Alex Christensen.
Remove code to set the SoupSession on DNSResolveQueue.
* NetworkProcess/NetworkProcess.cpp:
2020-10-05 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r267986.
https://bugs.webkit.org/show_bug.cgi?id=217329
Introduced layout test failures
Reverted changeset:
"[Cocoa] Reduce time waiting for Launch Services database"
https://bugs.webkit.org/show_bug.cgi?id=217253
https://trac.webkit.org/changeset/267986
2020-10-05 Aditya Keerthi <akeerthi@apple.com>
[Contact Picker API] Introduce bindings for the Contact Picker API
https://bugs.webkit.org/show_bug.cgi?id=216793
<rdar://problem/69317957>
Reviewed by Wenson Hsieh.
Added a new experimental feature flag. See WebCore Changelog for more details.
* Shared/WebPreferencesExperimental.yaml:
2020-10-05 Per Arne Vollan <pvollan@apple.com>
[Cocoa] Reduce time waiting for Launch Services database
https://bugs.webkit.org/show_bug.cgi?id=217253
<rdar://problem/68585497>
Reviewed by Brent Fulgham.
Currently, the WebContent process is waiting up to 5 seconds for the Launch Services database to be updated.
This is too long, since the process may be considered unresponsive by the system. This patch also addresses
an issue where the required entitlement 'com.apple.private.webkit.use-xpc-endpoint' is missing from the
WebContent development XPC service. This entitlement is required in order to receive the Launch Services
database XPC object in the WebContent process.
* Scripts/process-entitlements.sh:
* WebProcess/WebPage/Cocoa/WebPageCocoa.mm:
(WebKit::WebPage::platformDidReceiveLoadParameters):
2020-10-05 Wenson Hsieh <wenson_hsieh@apple.com>
REGRESSION (r267602): editing/selection/ios/select-all-in-readonly-input-does-not-overflow.html fails
https://bugs.webkit.org/show_bug.cgi?id=217321
Reviewed by Dean Jackson.
Fixes an existing test by restoring the commented-out code.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _selectionClipRect]):
2020-10-05 Nitzan Uziely <nitzan@testim.io>
WebDriver Input clear/value commands fails when target is inside shadow dom
https://bugs.webkit.org/show_bug.cgi?id=209233
Reviewed by Brian Burg.
Fixed nodeIsElement which would return true for non-element nodes,
which caused an exception to be thrown when getComputedStyle was called
on non shadow elements.
Fixed computeElementLayout to consider shadow root descendants as descendants
when executing elementsAtPoint.
Test for WPT (https://github.com/web-platform-tests/wpt/pull/25962)
* UIProcess/Automation/atoms/ElementDisplayed.js:
(isShown.nodeIsElement):
(isShown.enclosingNodeOrSelfMatchingPredicate):
(isShown.enclosingElementOrSelfMatchingPredicate):
* WebProcess/Automation/WebAutomationSessionProxy.cpp:
(WebKit::WebAutomationSessionProxy::computeElementLayout):
2020-10-05 John Wilander <wilander@apple.com>
Storage Access API: Enable per-page storage access scope and align test cases
https://bugs.webkit.org/show_bug.cgi?id=217077
<rdar://problem/69017878>
Reviewed by Brent Fulgham.
It was decided in https://github.com/privacycg/storage-access/issues/3 that
browsers should grant storage access for all same-site resources on the whole
page, not just the requesting iframe.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::hasStorageAccess):
Added a call to WebPage::hasPageLevelStorageAccess()
and an early return if it returns true.
(WebKit::WebPage::clearPageLevelStorageAccess):
New function to clear the webpage's m_domainsWithPageLevelStorageAccess
map.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::clearResourceLoadStatistics):
Now clears all webpages' page level storage access maps.
2020-10-05 Sam Weinig <weinig@apple.com>
Remove support for enabling subpixel CSSOM values, it's off by default everywhere and known to be not-compatible with the web
https://bugs.webkit.org/show_bug.cgi?id=217300
Reviewed by Simon Fraser.
Deprecate SPI for enabling subpixel CSSOM values. It is not web compatible.
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetSubpixelCSSOMElementMetricsEnabled):
(WKPreferencesGetSubpixelCSSOMElementMetricsEnabled):
* UIProcess/API/C/WKPreferencesRefPrivate.h:
* UIProcess/API/Cocoa/WKPreferences.mm:
(-[WKPreferences _setSubpixelCSSOMElementMetricsEnabled:]):
(-[WKPreferences _subpixelCSSOMElementMetricsEnabled]):
* UIProcess/API/Cocoa/WKPreferencesPrivate.h:
2020-10-05 Alex Christensen <achristensen@webkit.org>
Unify linkedOnOrAfter duplicate code and use SPI/WebCoreTestSupport instead of NSUserDefaults
https://bugs.webkit.org/show_bug.cgi?id=217239
Reviewed by Darin Adler.
* NetworkProcess/WebStorage/ios/LocalStorageDatabaseTrackerIOS.mm:
* NetworkProcess/cocoa/NetworkProcessCocoa.mm:
* Shared/Cocoa/CompletionHandlerCallChecker.mm:
(WebKit::shouldThrowExceptionForDuplicateCompletionHandlerCall):
* Shared/Cocoa/DefaultWebBrowserChecks.mm:
(WebKit::determineITPState):
* Shared/Cocoa/WebKit2InitializeCocoa.mm:
(WebKit::InitializeWebKit2):
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultPassiveTouchListenersAsDefaultOnDocument):
* SourcesCocoa.txt:
* UIProcess/API/C/WKPage.cpp:
(WKPageReload):
* UIProcess/API/Cocoa/WKProcessPool.mm:
(+[WKProcessPool _setLinkedOnOrAfterEverythingForTesting]):
* UIProcess/API/Cocoa/WKProcessPoolPrivate.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(shouldAllowSettingAnyXHRHeaderFromFileURLs):
(shouldRestrictBaseURLSchemes):
(-[WKWebView _initializeWithConfiguration:]):
(-[WKWebView _setupPageConfiguration:]):
(-[WKWebView reload]):
(-[WKWebView takeSnapshotWithConfiguration:completionHandler:]):
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(defaultShouldDecidePolicyBeforeLoadingQuickLookPreview):
(-[WKWebViewConfiguration init]):
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
(-[WKWebsiteDataStore init]):
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _computedObscuredInset]):
* UIProcess/API/mac/WKWebViewMac.mm:
(-[WKWebView _web_dragDestinationActionForDraggingInfo:]):
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
(WebKit::WebProcessPool::platformInitializeNetworkProcess):
* UIProcess/Inspector/mac/WKInspectorViewController.mm:
(-[WKInspectorViewController inspectorWKWebViewReload:]):
* UIProcess/RemoteLayerTree/RemoteScrollingCoordinatorProxy.cpp:
(WebKit::RemoteScrollingCoordinatorProxy::hasScrollableOrZoomedMainFrame const):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::maybeInitializeSandboxExtensionHandle):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::createWebPage):
* UIProcess/WebProcessProxy.cpp:
(WebKit::isMainThreadOrCheckDisabled):
* UIProcess/ios/WKContentView.mm:
(-[WKContentView _commonInitializationWithProcessPool:configuration:]):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _shouldUseContextMenus]):
(applicationIsKnownToIgnoreMouseEvents):
* UIProcess/ios/WKScrollView.mm:
(-[WKScrollView _systemContentInset]):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::desktopClassBrowsingRecommended):
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/cocoa/WebProcessCocoa.mm:
2020-10-05 Simon Fraser <simon.fraser@apple.com>
Clean up "rendering update" terminology in DrawingArea
https://bugs.webkit.org/show_bug.cgi?id=217307
Reviewed by Tim Horton.
DrawingArea doesn't know about "immediate" vs. "normal" rendering updates; for it,
all requests are for immediate rendering updates; the only difference is whether
layer tree freezing is respected.
This is the first step to correcting the naming; scheduleImmediateRenderingUpdate()
doesn't have to be exposed at all. A future patch will rename DrawingArea::scheduleRenderingUpdate()
to make it clear that it's not about timed updates.
* WebProcess/WebPage/CoordinatedGraphics/DrawingAreaCoordinatedGraphics.h:
* WebProcess/WebPage/DrawingArea.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.mm:
(WebKit::RemoteLayerTreeDrawingArea::setLayerTreeStateIsFrozen):
(WebKit::RemoteLayerTreeDrawingArea::startRenderingUpdateTimer):
(WebKit::RemoteLayerTreeDrawingArea::scheduleRenderingUpdate):
(WebKit::RemoteLayerTreeDrawingArea::activityStateDidChange):
(WebKit::RemoteLayerTreeDrawingArea::scheduleImmediateRenderingUpdate): Deleted.
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.h:
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::scheduleRenderingUpdate):
(WebKit::TiledCoreAnimationDrawingArea::scheduleImmediateRenderingUpdate): Deleted.
2020-10-05 Per Arne Vollan <pvollan@apple.com>
[GPU Process] Make it possible to dynamically block IOKit in the WebContent sandbox
https://bugs.webkit.org/show_bug.cgi?id=217129
<rdar://problem/69466039>
Reviewed by Geoffrey Garen.
We need to be able to enable IOKit blocking via the WebContent sandbox dynamically.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::gpuIOKitClasses):
(WebKit::WebProcessPool::platformInitializeWebProcess):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-10-04 Lauro Moura <lmoura@igalia.com>
webkitpy: Update messages reference files after r267916
https://bugs.webkit.org/show_bug.cgi?id=217304
Reviewed by Simon Fraser.
Covered by existing tests.
* Scripts/test-legacyMessageReceiver.cpp:
* Scripts/testMessageReceiver.cpp:
2020-10-04 Brian Burg <bburg@apple.com>
[Cocoa] Add WKUIDelegate SPI to inform clients when a _WKInspector is about to close
https://bugs.webkit.org/show_bug.cgi?id=217233
Reviewed by Timothy Hatcher and Devin Rousso.
This is complementary to _webView:didAttachLocalInspector: and works similarly.
Clients need to be careful in depending on this during teardown, because it won't
come if the UIDelegate has already been cleared out while the inspected page is being closed.
API test: WKInspectorDelegate.WillCloseLocalInspector
* UIProcess/API/APIUIClient.h:
(API::UIClient::willCloseLocalInspector):
* UIProcess/API/Cocoa/WKUIDelegatePrivate.h:
* UIProcess/Cocoa/UIDelegate.h:
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::setDelegate):
(WebKit::UIDelegate::UIClient::willCloseLocalInspector):
* UIProcess/Inspector/WebInspectorProxy.h:
* UIProcess/Inspector/WebInspectorProxy.cpp:
(WebKit::WebInspectorProxy::closeFrontendPageAndWindow):
Add a guard for reentrant calls to close().
2020-10-03 Yusuke Suzuki <ysuzuki@apple.com>
[JSC] Introduce JITOperationList to validate JIT-caged pointers
https://bugs.webkit.org/show_bug.cgi?id=217261
Reviewed by Saam Barati.
* Shared/API/c/WKString.cpp:
(WKStringCopyJSString):
* Shared/Cocoa/WebKit2InitializeCocoa.mm:
(WebKit::runInitializationCode):
* Shared/WebKit2Initialize.cpp:
(WebKit::InitializeWebKit2):
* Shared/WebKitJITOperations.cpp: Copied from Source/WebKit/Shared/WebKit2Initialize.cpp.
(WebKit::populateJITOperations):
* Shared/WebKitJITOperations.h: Copied from Source/WebKit/Shared/WebKit2Initialize.cpp.
* Sources.txt:
* WebKit.xcodeproj/project.pbxproj:
2020-10-03 Myles C. Maxfield <mmaxfield@apple.com>
[GPU Process] Support drawing text in 2D canvas with font features
https://bugs.webkit.org/show_bug.cgi?id=206118
Reviewed by Wenson Hsieh.
Simply serialize and deserialize all the non-derived data in Font and PlatformFontData.
* Shared/Cocoa/WebCoreArgumentCodersCocoa.mm:
(IPC::ArgumentCoder<Ref<Font>>::encodePlatformData):
(IPC::ArgumentCoder<Ref<Font>>::decodePlatformData):
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<Ref<Font>>::encode):
(IPC::ArgumentCoder<Ref<Font>>::decode):
* Shared/WebCoreArgumentCoders.h:
2020-10-03 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process] Web process is not allowed to set the AudioSession category
https://bugs.webkit.org/show_bug.cgi?id=217236
Reviewed by Eric Carlson.
We disabled setting the audio session category in the Web process when media is playing
in the GPU Process (r257367). But since RemoteAudioSession and RemoteAudioSessionProxy
are working now in the Web process and GPU process, we need to allow setting the audio
session category in the Web process.
This patch also includes an optimization to reduce the IPC traffic of the audio session
category changes.
* GPUProcess/media/RemoteAudioSessionProxy.cpp:
(WebKit::RemoteAudioSessionProxy::setCategory):
* GPUProcess/media/RemoteAudioSessionProxyManager.cpp:
(WebKit::RemoteAudioSessionProxyManager::setCategoryForProcess):
We need to check the category and policy in RemoteAudioSessionProxy instead of
RemoteAudioSessionProxyManager. Otherwise, the change will be ignored.
* WebProcess/GPU/media/RemoteAudioSession.cpp:
(WebKit::RemoteAudioSession::setCategory):
Only send RemoteAudioSessionProxy::SetCategory message if the category or the policy
is changed.
* WebProcess/GPU/media/RemoteAudioSession.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
Revert the change in r257367.
2020-10-02 Carlos Garcia Campos <cgarcia@igalia.com>
WebDriver: add support for same site cookies
https://bugs.webkit.org/show_bug.cgi?id=217222
Reviewed by Brian Burg.
* UIProcess/Automation/Automation.json: Add CookieSameSitePolicy enum and sameSite member to Cookie object.
* UIProcess/Automation/WebAutomationSession.cpp:
(WebKit::toProtocolSameSitePolicy): Convert WebCore same site policy to inspector protocol value.
(WebKit::toWebCoreSameSitePolicy): Convert inspector protocol same site policy to WebCore value.
(WebKit::buildObjectForCookie): Add sameSite.
(WebKit::WebAutomationSession::addSingleCookie): Set sameSite policy to cookie.
2020-10-02 Carlos Garcia Campos <cgarcia@igalia.com>
WebDriver: several issues when switching to new browser context
https://bugs.webkit.org/show_bug.cgi?id=217217
Reviewed by Brian Burg.
* UIProcess/Automation/WebAutomationSession.cpp:
(WebKit::WebAutomationSession::didDestroyFrame): Remove the frame references from maps.
(WebKit::WebAutomationSession::navigationOccurredForFrame): Only clear the frame references from the maps for
the frames in the given frame's page.
* UIProcess/Automation/WebAutomationSession.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::didDestroyFrame): Notify automation session about frame being destroyed.
2020-10-02 Simon Fraser <simon.fraser@apple.com>
Move WebEvent subclass declarations to their own files
https://bugs.webkit.org/show_bug.cgi?id=217259
Reviewed by Tim Horton.
Make new header files for WebMouseEvent, WebKeyboardEvent, WebWheelEvent, WebTouchEvent and
fix the fallout.
* Scripts/webkit/messages.py:
* Shared/API/c/WKSharedAPICast.h:
* Shared/NativeWebGestureEvent.h:
* Shared/NativeWebKeyboardEvent.h:
* Shared/NativeWebMouseEvent.h:
* Shared/NativeWebTouchEvent.h:
* Shared/NativeWebWheelEvent.h:
* Shared/NavigationActionData.h:
* Shared/WebEvent.cpp:
(WebKit::WebTouchEvent::allTouchPointsAreReleased const): Deleted.
* Shared/WebEvent.h:
(WebKit::WebMouseEvent::button const): Deleted.
(WebKit::WebMouseEvent::buttons const): Deleted.
(WebKit::WebMouseEvent::position const): Deleted.
(WebKit::WebMouseEvent::globalPosition const): Deleted.
(WebKit::WebMouseEvent::deltaX const): Deleted.
(WebKit::WebMouseEvent::deltaY const): Deleted.
(WebKit::WebMouseEvent::deltaZ const): Deleted.
(WebKit::WebMouseEvent::clickCount const): Deleted.
(WebKit::WebMouseEvent::eventNumber const): Deleted.
(WebKit::WebMouseEvent::menuTypeForEvent const): Deleted.
(WebKit::WebMouseEvent::force const): Deleted.
(WebKit::WebMouseEvent::syntheticClickType const): Deleted.
(WebKit::WebWheelEvent::position const): Deleted.
(WebKit::WebWheelEvent::globalPosition const): Deleted.
(WebKit::WebWheelEvent::delta const): Deleted.
(WebKit::WebWheelEvent::wheelTicks const): Deleted.
(WebKit::WebWheelEvent::granularity const): Deleted.
(WebKit::WebWheelEvent::directionInvertedFromDevice const): Deleted.
(WebKit::WebWheelEvent::phase const): Deleted.
(WebKit::WebWheelEvent::momentumPhase const): Deleted.
(WebKit::WebWheelEvent::hasPreciseScrollingDeltas const): Deleted.
(WebKit::WebWheelEvent::scrollCount const): Deleted.
(WebKit::WebWheelEvent::unacceleratedScrollingDelta const): Deleted.
(WebKit::WebKeyboardEvent::text const): Deleted.
(WebKit::WebKeyboardEvent::unmodifiedText const): Deleted.
(WebKit::WebKeyboardEvent::key const): Deleted.
(WebKit::WebKeyboardEvent::code const): Deleted.
(WebKit::WebKeyboardEvent::keyIdentifier const): Deleted.
(WebKit::WebKeyboardEvent::windowsVirtualKeyCode const): Deleted.
(WebKit::WebKeyboardEvent::nativeVirtualKeyCode const): Deleted.
(WebKit::WebKeyboardEvent::macCharCode const): Deleted.
(WebKit::WebKeyboardEvent::handledByInputMethod const): Deleted.
(WebKit::WebKeyboardEvent::preeditUnderlines const): Deleted.
(WebKit::WebKeyboardEvent::preeditSelectionRange const): Deleted.
(WebKit::WebKeyboardEvent::commands const): Deleted.
(WebKit::WebKeyboardEvent::isAutoRepeat const): Deleted.
(WebKit::WebKeyboardEvent::isKeypad const): Deleted.
(WebKit::WebKeyboardEvent::isSystemKey const): Deleted.
(WebKit::WebPlatformTouchPoint::WebPlatformTouchPoint): Deleted.
(WebKit::WebPlatformTouchPoint::identifier const): Deleted.
(WebKit::WebPlatformTouchPoint::location const): Deleted.
(WebKit::WebPlatformTouchPoint::phase const): Deleted.
(WebKit::WebPlatformTouchPoint::state const): Deleted.
(WebKit::WebPlatformTouchPoint::setRadiusX): Deleted.
(WebKit::WebPlatformTouchPoint::radiusX const): Deleted.
(WebKit::WebPlatformTouchPoint::setRadiusY): Deleted.
(WebKit::WebPlatformTouchPoint::radiusY const): Deleted.
(WebKit::WebPlatformTouchPoint::setRotationAngle): Deleted.
(WebKit::WebPlatformTouchPoint::rotationAngle const): Deleted.
(WebKit::WebPlatformTouchPoint::setForce): Deleted.
(WebKit::WebPlatformTouchPoint::force const): Deleted.
(WebKit::WebPlatformTouchPoint::setAltitudeAngle): Deleted.
(WebKit::WebPlatformTouchPoint::altitudeAngle const): Deleted.
(WebKit::WebPlatformTouchPoint::setAzimuthAngle): Deleted.
(WebKit::WebPlatformTouchPoint::azimuthAngle const): Deleted.
(WebKit::WebPlatformTouchPoint::setTouchType): Deleted.
(WebKit::WebPlatformTouchPoint::touchType const): Deleted.
(WebKit::WebTouchEvent::WebTouchEvent): Deleted.
(WebKit::WebTouchEvent::touchPoints const): Deleted.
(WebKit::WebTouchEvent::position const): Deleted.
(WebKit::WebTouchEvent::isPotentialTap const): Deleted.
(WebKit::WebTouchEvent::isGesture const): Deleted.
(WebKit::WebTouchEvent::gestureScale const): Deleted.
(WebKit::WebTouchEvent::gestureRotation const): Deleted.
(WebKit::WebTouchEvent::canPreventNativeGestures const): Deleted.
(WebKit::WebTouchEvent::setCanPreventNativeGestures): Deleted.
(WebKit::WebPlatformTouchPoint::id const): Deleted.
(WebKit::WebPlatformTouchPoint::screenPosition const): Deleted.
(WebKit::WebPlatformTouchPoint::position const): Deleted.
(WebKit::WebPlatformTouchPoint::radius const): Deleted.
(WebKit::WebPlatformTouchPoint::setState): Deleted.
* Shared/WebEventConversion.cpp:
* Shared/WebKeyboardEvent.cpp:
* Shared/WebKeyboardEvent.h: Added.
(WebKit::WebKeyboardEvent::text const):
(WebKit::WebKeyboardEvent::unmodifiedText const):
(WebKit::WebKeyboardEvent::key const):
(WebKit::WebKeyboardEvent::code const):
(WebKit::WebKeyboardEvent::keyIdentifier const):
(WebKit::WebKeyboardEvent::windowsVirtualKeyCode const):
(WebKit::WebKeyboardEvent::nativeVirtualKeyCode const):
(WebKit::WebKeyboardEvent::macCharCode const):
(WebKit::WebKeyboardEvent::handledByInputMethod const):
(WebKit::WebKeyboardEvent::preeditUnderlines const):
(WebKit::WebKeyboardEvent::preeditSelectionRange const):
(WebKit::WebKeyboardEvent::commands const):
(WebKit::WebKeyboardEvent::isAutoRepeat const):
(WebKit::WebKeyboardEvent::isKeypad const):
(WebKit::WebKeyboardEvent::isSystemKey const):
* Shared/WebMouseEvent.cpp:
* Shared/WebMouseEvent.h: Added.
(WebKit::WebMouseEvent::button const):
(WebKit::WebMouseEvent::buttons const):
(WebKit::WebMouseEvent::position const):
(WebKit::WebMouseEvent::globalPosition const):
(WebKit::WebMouseEvent::deltaX const):
(WebKit::WebMouseEvent::deltaY const):
(WebKit::WebMouseEvent::deltaZ const):
(WebKit::WebMouseEvent::clickCount const):
(WebKit::WebMouseEvent::eventNumber const):
(WebKit::WebMouseEvent::menuTypeForEvent const):
(WebKit::WebMouseEvent::force const):
(WebKit::WebMouseEvent::syntheticClickType const):
* Shared/WebTouchEvent.cpp:
(WebKit::WebTouchEvent::allTouchPointsAreReleased const):
* Shared/WebTouchEvent.h: Added.
(WebKit::WebPlatformTouchPoint::WebPlatformTouchPoint):
(WebKit::WebPlatformTouchPoint::identifier const):
(WebKit::WebPlatformTouchPoint::location const):
(WebKit::WebPlatformTouchPoint::phase const):
(WebKit::WebPlatformTouchPoint::state const):
(WebKit::WebPlatformTouchPoint::setRadiusX):
(WebKit::WebPlatformTouchPoint::radiusX const):
(WebKit::WebPlatformTouchPoint::setRadiusY):
(WebKit::WebPlatformTouchPoint::radiusY const):
(WebKit::WebPlatformTouchPoint::setRotationAngle):
(WebKit::WebPlatformTouchPoint::rotationAngle const):
(WebKit::WebPlatformTouchPoint::setForce):
(WebKit::WebPlatformTouchPoint::force const):
(WebKit::WebPlatformTouchPoint::setAltitudeAngle):
(WebKit::WebPlatformTouchPoint::altitudeAngle const):
(WebKit::WebPlatformTouchPoint::setAzimuthAngle):
(WebKit::WebPlatformTouchPoint::azimuthAngle const):
(WebKit::WebPlatformTouchPoint::setTouchType):
(WebKit::WebPlatformTouchPoint::touchType const):
(WebKit::WebTouchEvent::WebTouchEvent):
(WebKit::WebTouchEvent::touchPoints const):
(WebKit::WebTouchEvent::position const):
(WebKit::WebTouchEvent::isPotentialTap const):
(WebKit::WebTouchEvent::isGesture const):
(WebKit::WebTouchEvent::gestureScale const):
(WebKit::WebTouchEvent::gestureRotation const):
(WebKit::WebTouchEvent::canPreventNativeGestures const):
(WebKit::WebTouchEvent::setCanPreventNativeGestures):
(WebKit::WebPlatformTouchPoint::id const):
(WebKit::WebPlatformTouchPoint::screenPosition const):
(WebKit::WebPlatformTouchPoint::position const):
(WebKit::WebPlatformTouchPoint::radius const):
(WebKit::WebPlatformTouchPoint::setState):
* Shared/WebWheelEvent.cpp:
(WebKit::WebWheelEvent::WebWheelEvent):
* Shared/WebWheelEvent.h: Added.
(WebKit::WebWheelEvent::position const):
(WebKit::WebWheelEvent::globalPosition const):
(WebKit::WebWheelEvent::delta const):
(WebKit::WebWheelEvent::wheelTicks const):
(WebKit::WebWheelEvent::granularity const):
(WebKit::WebWheelEvent::directionInvertedFromDevice const):
(WebKit::WebWheelEvent::phase const):
(WebKit::WebWheelEvent::momentumPhase const):
(WebKit::WebWheelEvent::hasPreciseScrollingDeltas const):
(WebKit::WebWheelEvent::scrollCount const):
(WebKit::WebWheelEvent::unacceleratedScrollingDelta const):
* Shared/ios/WebIOSEventFactory.h:
* Shared/ios/WebPlatformTouchPointIOS.cpp:
* Shared/ios/WebTouchEventIOS.cpp:
* Shared/mac/NativeWebGestureEventMac.mm:
* Shared/mac/WebEventFactory.h:
* Shared/mac/WebGestureEvent.h:
* UIProcess/API/APINavigationClient.h:
* UIProcess/API/APIPolicyClient.h:
* UIProcess/Automation/SimulatedInputDispatcher.h:
* UIProcess/WebPageProxy.cpp:
* UIProcess/ios/WKMouseGestureRecognizer.mm:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/InjectedBundle/InjectedBundleNavigationAction.h:
* WebProcess/Plugins/Netscape/mac/NetscapePluginMac.mm:
* WebProcess/Plugins/PDF/PDFPlugin.mm:
* WebProcess/Plugins/PluginProxy.cpp:
* WebProcess/Plugins/PluginView.cpp:
* WebProcess/WebPage/EventDispatcher.cpp:
* WebProcess/WebPage/EventDispatcher.h:
* WebProcess/WebPage/PageBanner.h:
* WebProcess/WebPage/WebPage.cpp:
* WebProcess/WebPage/ios/WebPageIOS.mm:
* WebProcess/WebPage/mac/WebPageMac.mm:
2020-10-02 Wenson Hsieh <wenson_hsieh@apple.com>
[ iOS ] ASSERTION FAILED: ScriptDisallowedScope::InMainThread::isEventDispatchAllowedInSubtree
https://bugs.webkit.org/show_bug.cgi?id=217203
<rdar://problem/69864747>
Reviewed by Tim Horton.
Refactor logic that removes drop placeholder attributes such that it runs as an asynchronously task rather than
underneath the scope of `didFinishLoadingImageForElement`. This is because the image load client call may occur
in the middle of a rendering update, during which it is not safe to change element attribute values, since that
may trigger script execution.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::didFinishLoadingImageForElement):
2020-10-02 Kate Cheney <katherine_cheney@apple.com>
Multiple calls to suspend media playback for the same page may result in resuming media playback too soon
https://bugs.webkit.org/show_bug.cgi?id=217228
rdar://problem/69709346
Reviewed by Eric Carlson.
Calls to suspend/resume media playback should always be called in
pairs. The current implementation resumes media playback on the first
call to resumeAllMediaPlayback even if multiple calls to suspendAllMediaPlayback
have been made. This patch adds a counter to track how many calls to
suspendAllMediaPlayback have occurred, and only resumes once the
counter reaches 0.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _suspendMediaPlaybackCounter]):
SPI for test purposes.
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::suspendAllMediaPlayback):
(WebKit::WebPageProxy::resumeAllMediaPlayback):
Never let the counter get below 0.
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::suspendMediaPlaybackCounter):
2020-10-02 Truitt Savell <tsavell@apple.com>
Unreviewed, reverting r267817.
Broke two tests on Mac
Reverted changeset:
"Storage Access API: Enable per-page storage access scope and
align test cases"
https://bugs.webkit.org/show_bug.cgi?id=217077
https://trac.webkit.org/changeset/267817
2020-10-02 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process] Remove requiresTextTrackRepresentation in RemoteMediaPlayerState
https://bugs.webkit.org/show_bug.cgi?id=217230
Reviewed by Jer Noble.
A Web process can figure out whether the fullscreen/picture-in-picture subtitle
layer (for text track representation) is required by itself, so it does not need
to get the information from the GPU process.
* WebProcess/GPU/media/RemoteMediaPlayerState.h:
(WebKit::RemoteMediaPlayerState::encode const):
(WebKit::RemoteMediaPlayerState::decode):
2020-10-02 Brian Burg <bburg@apple.com>
[Cocoa] Web Inspector: add an ObjC protocol for extension support
https://bugs.webkit.org/show_bug.cgi?id=217191
<rdar://problem/69797960>
Reviewed by Timothy Hatcher.
* UIProcess/API/Cocoa/_WKInspector.h:
2020-10-02 Wenson Hsieh <wenson_hsieh@apple.com>
[GPU Process] fast/canvas/canvas-createPattern-video-loading.html times out
https://bugs.webkit.org/show_bug.cgi?id=217227
Reviewed by Eric Carlson.
Even after supporting `createPattern` with video elements using the GPU process for both media and canvas
rendering, we still fail this test due to a security error that is thrown when consulting
`CanvasRenderingContext::wouldTaintOrigin`. This, in turn, is due to the fact that we never attempt to update
cached `wouldTaintDocumentSecurityOrigin` state on `MediaPlayerPrivateRemote`, so it remains `true`.
This change will be covered by fast/canvas/canvas-createPattern-video-loading.html, which should now fail with a
text diff instead of timing out due to lack of support for creating a `CanvasPattern` out of a video element.
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::updateCachedState):
2020-10-02 Sam Weinig <weinig@apple.com>
Remove code behind ENABLE(MEDIA_SESSION), no ports enable it
https://bugs.webkit.org/show_bug.cgi?id=216831
Reviewed by Alex Christensen.
Removes all code in ENABLE(MEDIA_SESSION) and all media session
related SPI. It was never enabled and nothing ever adopted it.
* Sources.txt:
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
Remove media session related files.
* Shared/API/APIObject.h:
* Shared/API/c/WKBase.h:
* Shared/WebCoreArgumentCoders.cpp:
* Shared/WebCoreArgumentCoders.h:
* Shared/WebMediaSessionMetadata.cpp: Removed.
* Shared/WebMediaSessionMetadata.h: Removed.
* UIProcess/API/APIUIClient.h:
* UIProcess/API/C/WKAPICast.h:
* UIProcess/API/C/WKContext.cpp:
* UIProcess/API/C/WKContext.h:
* UIProcess/API/C/WKMediaSessionFocusManager.cpp: Removed.
* UIProcess/API/C/WKMediaSessionFocusManager.h: Removed.
* UIProcess/API/C/WKMediaSessionMetadata.cpp: Removed.
* UIProcess/API/C/WKMediaSessionMetadata.h: Removed.
* UIProcess/API/C/WKPage.cpp:
* UIProcess/API/C/WKPagePrivate.h:
* UIProcess/API/C/WKPageUIClient.h:
* UIProcess/WebMediaSessionFocusManager.cpp: Removed.
* UIProcess/WebMediaSessionFocusManager.h: Removed.
* UIProcess/WebMediaSessionFocusManagerClient.cpp: Removed.
* UIProcess/WebMediaSessionFocusManagerClient.h: Removed.
* UIProcess/WebPageProxy.cpp:
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/WebProcessPool.cpp:
* UIProcess/WebProcessPool.h:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebPage/WebPage.cpp:
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
Remove all code in ENABLE(MEDIA_SESSION).
2020-10-02 Karl Rackler <rackler@apple.com>
Unreviewed, reverting r267796.
This change is breaking the build on AzulE debug bots.
Reverted changeset:
"Remove code behind ENABLE(MEDIA_SESSION), no ports enable it"
https://bugs.webkit.org/show_bug.cgi?id=216831
https://trac.webkit.org/changeset/267796
2020-10-02 Alexander Mikhaylenko <alexm@gnome.org>
[GTK] WPE and backend versions are mixed up in webkit://gpu
https://bugs.webkit.org/show_bug.cgi?id=217210
Reviewed by Adrian Perez de Castro.
* UIProcess/API/glib/WebKitProtocolHandler.cpp:
(WebKit::WebKitProtocolHandler::handleGPU):
2020-10-02 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r267864.
https://bugs.webkit.org/show_bug.cgi?id=217220
Added broken tests
Reverted changeset:
"[GPU Process] Support drawing text in 2D canvas with font
features"
https://bugs.webkit.org/show_bug.cgi?id=206118
https://trac.webkit.org/changeset/267864
2020-10-01 Myles C. Maxfield <mmaxfield@apple.com>
[GPU Process] Support drawing text in 2D canvas with font features
https://bugs.webkit.org/show_bug.cgi?id=206118
Reviewed by Wenson Hsieh.
Simply serialize and deserialize all the non-derived data in Font and PlatformFontData.
* Shared/Cocoa/WebCoreArgumentCodersCocoa.mm:
(IPC::ArgumentCoder<Ref<Font>>::encodePlatformData):
(IPC::ArgumentCoder<Ref<Font>>::decodePlatformData):
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<Ref<Font>>::encode):
(IPC::ArgumentCoder<Ref<Font>>::decode):
* Shared/WebCoreArgumentCoders.h:
2020-10-01 Chris Dumez <cdumez@apple.com>
Add basic infrastructure for AudioWorklet
https://bugs.webkit.org/show_bug.cgi?id=217153
Reviewed by Sam Weinig.
Update RemoteAudioDestinationProxy to match what was done in AudioDestinationCocoa.
RemoteAudioDestinationProxy is used when the GPU process is enabled.
* GPUProcess/media/RemoteAudioDestinationManager.cpp:
(WebKit::RemoteAudioDestination::start):
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::create):
(WebKit::RemoteAudioDestinationProxy::start):
(WebKit::RemoteAudioDestinationProxy::stop):
(WebKit::RemoteAudioDestinationProxy::renderBuffer):
* WebProcess/GPU/media/RemoteAudioDestinationProxy.h:
* WebProcess/GPU/media/WebMediaStrategy.cpp:
(WebKit::WebMediaStrategy::createAudioDestination):
* WebProcess/GPU/media/WebMediaStrategy.h:
2020-10-01 Adrian Perez de Castro <aperez@igalia.com> and Don Olmstead <don.olmstead@sony.com>
Non-unified build fixes, early October 2020 edition
https://bugs.webkit.org/show_bug.cgi?id=217165
Reviewed by Yusuke Suzuki.
* WebProcess/GPU/graphics/ImageBufferShareableBitmapBackend.h:
* WebProcess/WebPage/atk/WebKitWebPageAccessibilityObject.cpp:
2020-10-01 Eric Carlson <eric.carlson@apple.com>
Prevent media capture unless the application has correct data for system permission prompt
https://bugs.webkit.org/show_bug.cgi?id=217104
<rdar://problem/69761929>
Reviewed by Darin Adler.
Tested manually with MiniBrowser.
* SourcesCocoa.txt:
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::requestAVCaptureAccessForMediaType): Renamed from requestAccessForMediaType.
(WebKit::UIDelegate::UIClient::decidePolicyForUserMediaPermissionRequest): Deny
requests of the application is not permitted to capture. Restructure logic to
make it easier to understand.
(WebKit::requestAccessForMediaType): Deleted.
* UIProcess/Cocoa/UserMediaPermissionRequestManagerProxy.mm: Added.
(WebKit::UserMediaPermissionRequestManagerProxy::permittedToCaptureAudio): Check
for required Info.plist string.
(WebKit::UserMediaPermissionRequestManagerProxy::permittedToCaptureVideo): Ditto.
* UIProcess/Cocoa/WebPageProxyCocoa.mm: Update for unified build.
* UIProcess/UserMediaPermissionRequestManagerProxy.cpp:
(WebKit::UserMediaPermissionRequestManagerProxy::permittedToCaptureAudio):
(WebKit::UserMediaPermissionRequestManagerProxy::permittedToCaptureVideo):
* UIProcess/UserMediaPermissionRequestManagerProxy.h:
* WebKit.xcodeproj/project.pbxproj:
2020-10-01 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process] Enable IPC message handlers of RemoteAudioSession
https://bugs.webkit.org/show_bug.cgi?id=217144
Reviewed by Jer Noble.
This patch enables the RemoteAudioSession in the Web process to handle
IPC messages from the GPU process. It also fixes a bug preventing the
audio playback to start.
No new tests, no functional change.
* GPUProcess/media/RemoteAudioSessionProxyManager.cpp:
(WebKit::RemoteAudioSessionProxyManager::tryToSetActiveForProcess):
Don't interrupt the audio session trying to start.
* GPUProcess/media/RemoteAudioSessionProxyManager.messages.in: Removed.
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/GPUProcessConnection.cpp:
(WebKit::GPUProcessConnection::dispatchMessage): Clean-up.
* WebProcess/GPU/media/RemoteAudioSession.cpp:
(WebKit::RemoteAudioSession::RemoteAudioSession):
(WebKit::RemoteAudioSession::~RemoteAudioSession):
We need to remove the "destinationID" parameter, otherwise, RemoteAudioSession
won't receive the IPC messages.
* WebProcess/GPU/media/RemoteAudioSession.h: Clean-up.
* WebProcess/GPU/media/WebMediaStrategy.h: Clean-up.
2020-10-01 Youenn Fablet <youenn@apple.com>
MediaRecorder should support MediaRecorderOptions.mimeType
https://bugs.webkit.org/show_bug.cgi?id=215018
<rdar://problem/66681508>
Reviewed by Eric Carlson.
Update code to pass mimeType as part of startRecording callback.
* GPUProcess/webrtc/RemoteMediaRecorder.h:
(WebKit::RemoteMediaRecorder::mimeType const):
* GPUProcess/webrtc/RemoteMediaRecorderManager.cpp:
(WebKit::RemoteMediaRecorderManager::createRecorder):
* GPUProcess/webrtc/RemoteMediaRecorderManager.h:
* GPUProcess/webrtc/RemoteMediaRecorderManager.messages.in:
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
(WebKit::MediaRecorderPrivate::startRecording):
* WebProcess/GPU/webrtc/MediaRecorderPrivate.h:
2020-10-01 Youenn Fablet <youenn@apple.com>
MediaRecorder should support isTypeSupported
https://bugs.webkit.org/show_bug.cgi?id=216856
<rdar://problem/69767695>
Reviewed by Darin Adler.
Stop passing mimeType for each data blob.
Instead, compute it on WebProcess side.
* GPUProcess/webrtc/RemoteMediaRecorder.cpp:
(WebKit::RemoteMediaRecorder::fetchData):
* GPUProcess/webrtc/RemoteMediaRecorder.h:
* GPUProcess/webrtc/RemoteMediaRecorder.messages.in:
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
(WebKit::MediaRecorderPrivate::MediaRecorderPrivate):
(WebKit::MediaRecorderPrivate::fetchData):
(WebKit::MediaRecorderPrivate::mimeType const):
* WebProcess/GPU/webrtc/MediaRecorderPrivate.h:
2020-09-30 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process] Enable IPC message handlers of RemoteMediaSessionHelper
https://bugs.webkit.org/show_bug.cgi?id=217152
Reviewed by Eric Carlson.
* WebProcess/GPU/GPUProcessConnection.cpp:
(WebKit::GPUProcessConnection::dispatchMessage):
* WebProcess/GPU/media/ios/RemoteMediaSessionHelper.cpp:
(WebKit::RemoteMediaSessionHelper::RemoteMediaSessionHelper):
2020-09-30 John Wilander <wilander@apple.com>
Storage Access API: Enable per-page storage access scope and align test cases
https://bugs.webkit.org/show_bug.cgi?id=217077
<rdar://problem/69017878>
Reviewed by Brent Fulgham.
It was decided in https://github.com/privacycg/storage-access/issues/3 that
browsers should grant storage access for all same-site resources on the whole
page, not just the requesting iframe.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::hasStorageAccess):
Added a call to WebPage::hasPageLevelStorageAccess()
and an early return if it returns true.
2020-09-30 Wenson Hsieh <wenson_hsieh@apple.com>
[GPU Process] Canvas layout tests occasionally crash the web process under PutImageData::apply
https://bugs.webkit.org/show_bug.cgi?id=217140
Reviewed by Tim Horton.
When destroying a `RemoteImageBuffer`, logic exists in the superclass (`DisplayList::ImageBuffer`) that attempts
to flush any remaining display list items. However, this currently doesn't work as intended, since we've already
finished destroying the subclass, so the superclass' call to `flushDrawingContext()` calls `DisplayList::
ImageBuffer::flushDrawingContext()` rather than `RemoteImageBuffer::flushDrawingContext()`. Rather than flushing
display list items to the GPU process, this has the effect of attempting to replay these display list items to
the drawing context (which is backed by a `DisplayList` anyways), so this ends up having no tangible effect.
However, if one of these display list items happens to be `PutImageData`, we'll end up hitting an assertion due
to being unable to apply it without a `Delegate`. To fix this, hoist the call to `flushDrawingContext()` out to
`~RemoteImageBuffer()`, rather than `~DisplayList::ImageBuffer()`.
* WebProcess/GPU/graphics/RemoteImageBuffer.h:
(WebKit::RemoteImageBuffer::~RemoteImageBuffer):
2020-09-30 Per Arne Vollan <pvollan@apple.com>
[macOS] Deny mach-lookup to the analytics service in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=217137
<rdar://problem/69019193>
Reviewed by Brent Fulgham.
Only internal OS variants should have access to the 'com.apple.analyticsd' service.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::diagnosticServices):
* WebProcess/com.apple.WebProcess.sb.in:
2020-09-30 Alex Christensen <achristensen@webkit.org>
REGRESSION[267763] RELEASE_ASSERT in allNetworkProcesses hit when registering legacy protocol handlers before instantiating any WebKit objects
https://bugs.webkit.org/show_bug.cgi?id=217142
<rdar://problem/69778879>
Reviewed by Brady Eidson.
RELEASE_ASSERT(RunLoop::isMain()) was crashing because WTF::initializeMainThread hadn't been called.
This wasn't a problem before r267763 because the assertion did not exist. Rather than remove the assertion (which I think is important to have)
let's just initialize WebKit threading at these SPI entry points.
This unfortunately can't be tested with our current infrastructure because TestsController's constructor calls initializeMainThread.
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::registerGlobalURLSchemeAsHavingCustomProtocolHandlers):
(WebKit::WebProcessPool::unregisterGlobalURLSchemeAsHavingCustomProtocolHandlers):
2020-09-30 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed, reverting r267788.
Broke internal builds.
Reverted changeset:
"MediaRecorder should support isTypeSupported"
https://bugs.webkit.org/show_bug.cgi?id=216856
https://trac.webkit.org/changeset/267788
2020-09-30 Sam Weinig <weinig@apple.com>
Remove code behind ENABLE(MEDIA_SESSION), no ports enable it
https://bugs.webkit.org/show_bug.cgi?id=216831
Reviewed by Alex Christensen.
Removes all code in ENABLE(MEDIA_SESSION) and all media session
related SPI. It was never enabled and nothing ever adopted it.
* Sources.txt:
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
Remove media session related files.
* Shared/API/APIObject.h:
* Shared/API/c/WKBase.h:
* Shared/WebCoreArgumentCoders.cpp:
* Shared/WebCoreArgumentCoders.h:
* Shared/WebMediaSessionMetadata.cpp: Removed.
* Shared/WebMediaSessionMetadata.h: Removed.
* UIProcess/API/APIUIClient.h:
* UIProcess/API/C/WKAPICast.h:
* UIProcess/API/C/WKContext.cpp:
* UIProcess/API/C/WKContext.h:
* UIProcess/API/C/WKMediaSessionFocusManager.cpp: Removed.
* UIProcess/API/C/WKMediaSessionFocusManager.h: Removed.
* UIProcess/API/C/WKMediaSessionMetadata.cpp: Removed.
* UIProcess/API/C/WKMediaSessionMetadata.h: Removed.
* UIProcess/API/C/WKPage.cpp:
* UIProcess/API/C/WKPagePrivate.h:
* UIProcess/API/C/WKPageUIClient.h:
* UIProcess/WebMediaSessionFocusManager.cpp: Removed.
* UIProcess/WebMediaSessionFocusManager.h: Removed.
* UIProcess/WebMediaSessionFocusManagerClient.cpp: Removed.
* UIProcess/WebMediaSessionFocusManagerClient.h: Removed.
* UIProcess/WebPageProxy.cpp:
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/WebProcessPool.cpp:
* UIProcess/WebProcessPool.h:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebPage/WebPage.cpp:
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
Remove all code in ENABLE(MEDIA_SESSION).
2020-09-30 Brian Burg <bburg@apple.com>
Web Inspector: exempt API::SharedJSContext from remote inspection and automatic inspection
https://bugs.webkit.org/show_bug.cgi?id=217109
<rdar://problem/69386559>
Reviewed by Mark Lam.
We have seen occasional unexpected spins underneath pauseWaitingForRemoteInspection() waiting
for this JSContext. Since it does not evaluate JavaScript and is WebKit-internal, simply disable
remote inspection for the JSContext.
* UIProcess/API/Cocoa/APISerializedScriptValueCocoa.mm:
(API::SharedJSContext::ensureContext):
2020-09-30 Tim Horton <timothy_horton@apple.com>
REGRESSION (r265009): Web Share API can no longer be invoked if a previous invocation was dismissed using the close button
https://bugs.webkit.org/show_bug.cgi?id=216913
<rdar://problem/69659878>
Reviewed by Darin Adler.
No new tests; not able to test without UI tests that actually use UIActivityViewController.
* UIProcess/Cocoa/WKShareSheet.mm:
(-[WKShareSheet presentWithShareDataArray:inRect:]):
-presentationControllerDidDismiss is not called if the UIActivityViewController is
programmatically dismissed, so we need to revert to observing completion via
UIActivityViewController's -completionWithItemsHandler.
In order to avoid regressing bug 214894 by doing so, we also need to
make sure that we're actually not presented anymore, as -completionWithItemsHandler
can be called multiple times before the share sheet is actually dismissed.
2020-09-30 Youenn Fablet <youenn@apple.com>
MediaRecorder should support isTypeSupported
https://bugs.webkit.org/show_bug.cgi?id=216856
Reviewed by Darin Adler.
Stop passing mimeType for each data blob.
Instead, compute it on WebProcess side.
* GPUProcess/webrtc/RemoteMediaRecorder.cpp:
(WebKit::RemoteMediaRecorder::fetchData):
* GPUProcess/webrtc/RemoteMediaRecorder.h:
* GPUProcess/webrtc/RemoteMediaRecorder.messages.in:
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
(WebKit::MediaRecorderPrivate::MediaRecorderPrivate):
(WebKit::MediaRecorderPrivate::fetchData):
(WebKit::MediaRecorderPrivate::mimeType const):
* WebProcess/GPU/webrtc/MediaRecorderPrivate.h:
2020-09-30 Myles C. Maxfield <mmaxfield@apple.com>
[GPU Process] Remove FontHandle
https://bugs.webkit.org/show_bug.cgi?id=217076
Reviewed by Wenson Hsieh.
* Shared/Cocoa/WebCoreArgumentCodersCocoa.mm:
(IPC::ArgumentCoder<Ref<Font>>::encodePlatformData):
(IPC::ArgumentCoder<Ref<Font>>::decodePlatformData):
(IPC::ArgumentCoder<FontHandle>::encodePlatformData): Deleted.
(IPC::ArgumentCoder<FontHandle>::decodePlatformData): Deleted.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<Ref<Font>>::encode):
(IPC::createFont):
(IPC::ArgumentCoder<Ref<Font>>::decode):
(IPC::ArgumentCoder<FontHandle>::encode): Deleted.
(IPC::ArgumentCoder<FontHandle>::decode): Deleted.
* Shared/WebCoreArgumentCoders.h:
2020-09-29 Sam Weinig <weinig@apple.com>
[Preferences] Adopt shared preferences configuration and script in WebKit
https://bugs.webkit.org/show_bug.cgi?id=217075
Reviewed by Darin Adler.
* Shared/WebPreferences.yaml: Removed.
* Shared/WebPreferencesDebug.yaml: Removed.
* Shared/WebPreferencesExperimental.yaml: Removed.
* Shared/WebPreferencesInternal.yaml: Removed.
* Scripts/GeneratePreferences.rb: Removed.
* WebKit.xcodeproj/project.pbxproj:
Remove yaml preferences and preference generator in favor of shared one
in WTF.
* CMakeLists.txt:
* DerivedSources.make:
Update preference generation to use the shared script and preference files.
* Configurations/BaseTarget.xcconfig:
Add new variable, WTF_BUILD_SCRIPTS_DIR, that points to directory where the
shared generator and preference files live for use as input to the generation.
* DerivedSources-input.xcfilelist:
Update for new location of script and preference files.
* Scripts/PreferencesTemplates/WebPreferencesDefinitions.h.erb:
* Scripts/PreferencesTemplates/WebPreferencesExperimentalFeatures.cpp.erb:
* Scripts/PreferencesTemplates/WebPreferencesInternalDebugFeatures.cpp.erb:
* Scripts/PreferencesTemplates/WebPreferencesStoreDefaultsMap.cpp.erb:
Port support for multiple conditionally-defined defaultValues from WebKitLegacy,
keeping the same DEFAULT_VALUE_FOR_* macro naming and using the macro definitions
in place of existing <%= @pref.defaultValue %> idiom. Also updates for rename
from @internalDebugFeatures to @internalFeatures.
* Shared/WebPreferencesDefaultValues.cpp:
* Shared/WebPreferencesDefaultValues.h:
Remove defaults that are now fully specified in the preference files.
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration init]):
Replace removed DEFAULT_APPLE_PAY_ENABLED, with the now generated DEFAULT_VALUE_FOR_ApplePayEnabled.
2020-09-29 Wenson Hsieh <wenson_hsieh@apple.com>
[GPU Process] RemoteImageBufferProxy should ensure the state stack of the context is cleared before destruction
https://bugs.webkit.org/show_bug.cgi?id=208876
Reviewed by Tim Horton.
When running canvas layout tests in debug WebKit, the assertion in `~GraphicsContext()` which verifies that
`m_stack.isEmpty()` fires. This occurs because the matching call to restore may still be buffered in the web
process when the remote image buffer proxy in the GPU process is destroyed; however, this may also occur if the
web process terminates without having sent the matching calls to restore the graphics context.
We can fix both of these scenarios by manually unwinding the context's state stack when destroying the
`RemoteImageBufferProxy`. Original patch by Said Abou-Hallawa.
* GPUProcess/graphics/RemoteImageBufferProxy.h:
(WebKit::RemoteImageBufferProxy::~RemoteImageBufferProxy):
2020-09-29 Stephan Szabo <stephan.szabo@sony.com>
[PlayStation] Build fix for GPU Process after r267725
https://bugs.webkit.org/show_bug.cgi?id=217098
Unreviewed build fix.
Update playstation specific code for signature changes in the above.
Fix MediaPlayerPrivateRemote constructor when RELEASE_LOG_DISABLED was specified.
* GPUProcess/media/playstation/RemoteMediaPlayerProxyPlayStation.cpp:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
2020-09-29 Alex Christensen <achristensen@webkit.org>
Move NetworkProcessProxy ownership from WebProcessPool to WebsiteDataStore
https://bugs.webkit.org/show_bug.cgi?id=216041
Reviewed by Brady Eidson.
This patch moves the things on WebProcessPool associated with NetworkProcessProxy ownership and moves them
to WebsiteDataStore. It's pretty mechanical, but it reaches its fingers into a lot of things. Some comments
on interesting parts are below.
Why am I doing this? I'm glad you asked.
WKWebView had a fundamental cookie problem before this change. It had a WKWebsiteDataStore that could point to a cookie store on disk
that could be used in multiple network processes at the same time. This led to problems with cookies being received over the network
in those different processes being unable to be shared until they were written to disk, something that does not happen immediately, especially
on iOS. This problem is even worse with session cookies, which should be kept in memory and never be written to disk. Those processes
did not use shared memory to share the cookies. They just didn't share them, which led to cookie incorrectness. This was unfortunately
the default configuration if you call [[WKWebView alloc] init] more than once, which many developers do. There was also occasional cookie
corruption and loss from multiple processes writing cookies to the same file. All these problems can be worked around by doing what Safari does:
manually setting the WKProcessPool of all WKWebViewConfigurations to the same process pool to only use one network process. This is not intuitive
or well documented, and even though I have spent the majority of WWDC labs for the last 3 years telling developers to do it, most developers do not.
The default behavior should not lead to cookie incorrectness, corruption, or loss. This accomplishes that.
A bigger problem that can't really be worked around well is what we were doing in our WKHTTPCookieStore API. Many developers call setCookie:completionHandler:
to "log in" the user by setting a cookie, then in the completion handler, they open a WKWebView to their page and expect the cookie to be sent. Before this
change, we would not know which process pool to set the cookie in so we would guess. If there were no process pools, we would start writing to disk and hope
it writes fast enough. We can't wait for it to complete because it causes hangs like rdar://problem/66961066. This solution allows to start the network
process of that WebsiteDataStore and set the cookie in that process. We now know that is the only process this WebsiteDataStore will ever use. Now, users
will be able to actually be logged in all the time in such apps.
* NetworkProcess/Cookies/WebCookieManager.cpp:
(WebKit::WebCookieManager::setHTTPCookieAcceptPolicy):
(WebKit::WebCookieManager::getHTTPCookieAcceptPolicy):
* NetworkProcess/Cookies/WebCookieManager.h:
* NetworkProcess/Cookies/WebCookieManager.messages.in:
* NetworkProcess/Cookies/curl/WebCookieManagerCurl.cpp:
(WebKit::WebCookieManager::platformSetHTTPCookieAcceptPolicy):
* NetworkProcess/Cookies/mac/WebCookieManagerMac.mm:
(WebKit::WebCookieManager::platformSetHTTPCookieAcceptPolicy):
* NetworkProcess/Cookies/soup/WebCookieManagerSoup.cpp:
(WebKit::WebCookieManager::platformSetHTTPCookieAcceptPolicy):
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::initializeNetworkProcess):
(WebKit::NetworkProcess::ensureSession):
(WebKit::NetworkProcess::storageSession const):
(WebKit::NetworkProcess::forEachNetworkStorageSession):
(WebKit::NetworkProcess::defaultStorageSession const): Deleted.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcessCreationParameters.cpp:
(WebKit::NetworkProcessCreationParameters::encode const):
(WebKit::NetworkProcessCreationParameters::decode):
* NetworkProcess/NetworkProcessCreationParameters.h:
* NetworkProcess/WebStorage/LocalStorageDatabase.cpp:
(WebKit::LocalStorageDatabase::close):
* NetworkProcess/cocoa/NetworkProcessCocoa.mm:
(WebKit::NetworkProcess::platformInitializeNetworkProcessCocoa):
(WebKit::saveCookies):
(WebKit::NetworkProcess::platformCreateDefaultStorageSession const): Deleted.
* NetworkProcess/curl/NetworkProcessCurl.cpp:
(WebKit::NetworkProcess::platformCreateDefaultStorageSession const): Deleted.
* NetworkProcess/mac/RemoteNetworkingContext.mm:
(WebKit::RemoteNetworkingContext::ensureWebsiteDataStoreSession):
* NetworkProcess/soup/NetworkProcessSoup.cpp:
(WebKit::NetworkProcess::platformCreateDefaultStorageSession const): Deleted.
* Platform/IPC/Attachment.h:
* Shared/WebsiteDataStoreParameters.cpp:
(WebKit::WebsiteDataStoreParameters::encode const):
(WebKit::WebsiteDataStoreParameters::decode):
* Shared/WebsiteDataStoreParameters.h:
* UIProcess/API/APIHTTPCookieStore.cpp:
(API::HTTPCookieStore::HTTPCookieStore):
(API::HTTPCookieStore::~HTTPCookieStore):
(API::HTTPCookieStore::cookies):
(API::HTTPCookieStore::cookiesForURL):
(API::HTTPCookieStore::setCookies):
(API::HTTPCookieStore::deleteCookie):
(API::HTTPCookieStore::deleteAllCookies):
(API::HTTPCookieStore::setHTTPCookieAcceptPolicy):
(API::HTTPCookieStore::registerObserver):
(API::HTTPCookieStore::cookieManagerDestroyed):
(API::HTTPCookieStore::registerForNewProcessPoolNotifications): Deleted.
(API::HTTPCookieStore::unregisterForNewProcessPoolNotifications): Deleted.
* UIProcess/API/APIHTTPCookieStore.h:
* UIProcess/API/C/WKContext.cpp:
(WKContextSetUseSeparateServiceWorkerProcess):
(WKContextSetPrimaryWebsiteDataStore):
(WKContextGetCookieManager): Deleted.
(WKContextSetAllowsAnySSLCertificateForServiceWorkerTesting): Deleted.
(WKContextTerminateNetworkProcess): Deleted.
(WKContextGetNetworkProcessIdentifier): Deleted.
* UIProcess/API/C/WKContext.h:
* UIProcess/API/C/WKContextPrivate.h:
* UIProcess/API/C/WKPage.cpp:
(WKPageGetWebsiteDataStore):
* UIProcess/API/C/WKPage.h:
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreTerminateNetworkProcess):
(WKWebsiteDataStoreGetNetworkProcessIdentifier):
* UIProcess/API/C/WKWebsiteDataStoreRef.h:
* UIProcess/API/Cocoa/WKProcessPool.mm:
(-[WKProcessPool _setUseSeparateServiceWorkerProcess:]):
(-[WKProcessPool _terminateNetworkProcess]): Deleted.
(-[WKProcessPool _sendNetworkProcessWillSuspendImminently]): Deleted.
(-[WKProcessPool _sendNetworkProcessPrepareToSuspend:]): Deleted.
(-[WKProcessPool _sendNetworkProcessDidResume]): Deleted.
(-[WKProcessPool _networkProcessIdentifier]): Deleted.
(-[WKProcessPool _makeNextNetworkProcessLaunchFailForTesting]): Deleted.
(-[WKProcessPool _synthesizeAppIsBackground:]): Deleted.
(-[WKProcessPool _setAllowsAnySSLCertificateForServiceWorker:]): Deleted.
(-[WKProcessPool _networkProcessHasEntitlementForTesting:]): Deleted.
* UIProcess/API/Cocoa/WKProcessPoolPrivate.h:
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
(-[WKWebsiteDataStore _networkProcessHasEntitlementForTesting:]):
(-[WKWebsiteDataStore _terminateNetworkProcess]):
(-[WKWebsiteDataStore _sendNetworkProcessPrepareToSuspend:]):
(-[WKWebsiteDataStore _sendNetworkProcessWillSuspendImminently]):
(-[WKWebsiteDataStore _sendNetworkProcessDidResume]):
(-[WKWebsiteDataStore _synthesizeAppIsBackground:]):
(-[WKWebsiteDataStore _networkProcessIdentifier]):
(+[WKWebsiteDataStore _makeNextNetworkProcessLaunchFailForTesting]):
* UIProcess/API/Cocoa/WKWebsiteDataStorePrivate.h:
* UIProcess/API/glib/WebKitCookieManager.cpp:
(_WebKitCookieManagerPrivate::cookieManager const):
(_WebKitCookieManagerPrivate::~_WebKitCookieManagerPrivate):
(webkitCookieManagerCreate):
(webkit_cookie_manager_set_persistent_storage):
(webkit_cookie_manager_set_accept_policy):
(webkit_cookie_manager_get_accept_policy):
(webkit_cookie_manager_add_cookie):
(webkit_cookie_manager_get_cookies):
(webkit_cookie_manager_delete_cookie):
* UIProcess/API/glib/WebKitWebContext.cpp:
(webkitWebContextConstructed):
(webkit_web_context_prefetch_dns):
* UIProcess/Automation/WebAutomationSession.cpp:
(WebKit::WebAutomationSession::addSingleCookie):
(WebKit::WebAutomationSession::deleteAllCookies):
* UIProcess/AuxiliaryProcessProxy.h:
* UIProcess/Cocoa/LegacyCustomProtocolManagerClient.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::updateProcessSuppressionState):
(WebKit::WebProcessPool::platformInitialize):
(WebKit::WebProcessPool::platformInitializeNetworkProcess):
(WebKit::WebProcessPool::networkProcessHasEntitlementForTesting): Deleted.
(WebKit::WebProcessPool::xpcEndpointMessage const): Deleted.
(WebKit::WebProcessPool::sendNetworkProcessXPCEndpointToWebProcess): Deleted.
* UIProcess/Downloads/DownloadProxy.cpp:
(WebKit::DownloadProxy::cancel):
(WebKit::DownloadProxy::publishProgress):
(WebKit::DownloadProxy::didReceiveAuthenticationChallenge):
(WebKit::DownloadProxy::willSendRequest):
(WebKit::DownloadProxy::decideDestinationWithSuggestedFilenameAsync):
* UIProcess/Downloads/DownloadProxyMap.cpp:
(WebKit::DownloadProxyMap::DownloadProxyMap):
(WebKit::DownloadProxyMap::applicationDidEnterBackground):
(WebKit::DownloadProxyMap::applicationWillEnterForeground):
(WebKit::DownloadProxyMap::createDownloadProxy):
(WebKit::DownloadProxyMap::downloadFinished):
(WebKit::DownloadProxyMap::invalidate):
* UIProcess/Downloads/DownloadProxyMap.h:
* UIProcess/Network/CustomProtocols/LegacyCustomProtocolManagerProxy.cpp:
(WebKit::LegacyCustomProtocolManagerProxy::startLoading):
(WebKit::LegacyCustomProtocolManagerProxy::stopLoading):
(WebKit::LegacyCustomProtocolManagerProxy::invalidate):
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::allNetworkProcesses):
(WebKit::defaultProcess):
(WebKit::NetworkProcessProxy::defaultNetworkProcess):
(WebKit::NetworkProcessProxy::terminate):
(WebKit::NetworkProcessProxy::didTerminate):
(WebKit::NetworkProcessProxy::sendCreationParametersToNewProcess):
(WebKit::anyProcessPoolAlwaysRunsAtBackgroundPriority):
(WebKit::anyProcessPoolShouldTakeUIBackgroundAssertion):
(WebKit::NetworkProcessProxy::NetworkProcessProxy):
(WebKit::NetworkProcessProxy::~NetworkProcessProxy):
(WebKit::NetworkProcessProxy::getLaunchOptions):
(WebKit::NetworkProcessProxy::createDownloadProxy):
(WebKit::NetworkProcessProxy::networkProcessCrashed):
(WebKit::NetworkProcessProxy::didReceiveMessage):
(WebKit::NetworkProcessProxy::didClose):
(WebKit::NetworkProcessProxy::didReceiveAuthenticationChallenge):
(WebKit::NetworkProcessProxy::didFinishLaunching):
(WebKit::NetworkProcessProxy::setDomainsWithUserInteraction):
(WebKit::NetworkProcessProxy::addSession):
(WebKit::NetworkProcessProxy::removeSession):
(WebKit::NetworkProcessProxy::websiteDataStoreFromSessionID):
(WebKit::NetworkProcessProxy::establishWorkerContextConnectionToNetworkProcess):
(WebKit::anyProcessPoolHasForegroundWebProcesses):
(WebKit::anyProcessPoolHasBackgroundWebProcesses):
(WebKit::NetworkProcessProxy::updateProcessAssertion):
(WebKit::NetworkProcessProxy::hasSession const): Deleted.
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/Network/NetworkProcessProxyCocoa.mm:
(WebKit::NetworkProcessProxy::XPCEventHandler::handleXPCEvent const):
* UIProcess/WebCookieManagerProxy.cpp:
(WebKit::WebCookieManagerProxy::WebCookieManagerProxy):
(WebKit::WebCookieManagerProxy::~WebCookieManagerProxy):
(WebKit::WebCookieManagerProxy::getHostnamesWithCookies):
(WebKit::WebCookieManagerProxy::deleteCookiesForHostnames):
(WebKit::WebCookieManagerProxy::deleteAllCookies):
(WebKit::WebCookieManagerProxy::deleteCookie):
(WebKit::WebCookieManagerProxy::deleteAllCookiesModifiedSince):
(WebKit::WebCookieManagerProxy::setCookies):
(WebKit::WebCookieManagerProxy::getAllCookies):
(WebKit::WebCookieManagerProxy::getCookies):
(WebKit::WebCookieManagerProxy::startObservingCookieChanges):
(WebKit::WebCookieManagerProxy::stopObservingCookieChanges):
(WebKit::WebCookieManagerProxy::setHTTPCookieAcceptPolicy):
(WebKit::WebCookieManagerProxy::getHTTPCookieAcceptPolicy):
(WebKit::WebCookieManagerProxy::supplementName): Deleted.
(WebKit::WebCookieManagerProxy::create): Deleted.
(WebKit::WebCookieManagerProxy::processPoolDestroyed): Deleted.
(WebKit::WebCookieManagerProxy::processDidClose): Deleted.
(WebKit::WebCookieManagerProxy::refWebContextSupplement): Deleted.
(WebKit::WebCookieManagerProxy::derefWebContextSupplement): Deleted.
* UIProcess/WebCookieManagerProxy.h:
(WebKit::WebCookieManagerProxy::create):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::setControlledByAutomation):
(WebKit::WebPageProxy::disableServiceWorkerEntitlementInNetworkProcess):
(WebKit::WebPageProxy::clearServiceWorkerEntitlementOverride):
(WebKit::WebPageProxy::preconnectTo):
(WebKit::WebPageProxy::didCommitLoadForFrame):
(WebKit::WebPageProxy::logFrameNavigation):
(WebKit::WebPageProxy::dumpAdClickAttribution):
(WebKit::WebPageProxy::clearAdClickAttribution):
(WebKit::WebPageProxy::setAdClickAttributionOverrideTimerForTesting):
(WebKit::WebPageProxy::setAdClickAttributionConversionURLForTesting):
(WebKit::WebPageProxy::markAdClickAttributionsAsExpiredForTesting):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::urlSchemesWithCustomProtocolHandlers):
(WebKit::WebProcessPool::WebProcessPool):
(WebKit::WebProcessPool::~WebProcessPool):
(WebKit::WebProcessPool::languageChanged):
(WebKit::WebProcessPool::sendMemoryPressureEvent):
(WebKit::WebProcessPool::networkProcessCrashed):
(WebKit::WebProcessPool::establishWorkerContextConnectionToNetworkProcess):
(WebKit::WebProcessPool::removeFromServiceWorkerProcesses):
(WebKit::WebProcessPool::tryTakePrewarmedProcess):
(WebKit::WebProcessPool::initializeNewWebProcess):
(WebKit::WebProcessPool::processDidFinishLaunching):
(WebKit::WebProcessPool::createWebPage):
(WebKit::WebProcessPool::updateServiceWorkerUserAgent):
(WebKit::WebProcessPool::pageBeginUsingWebsiteDataStore):
(WebKit::WebProcessPool::download):
(WebKit::WebProcessPool::resumeDownload):
(WebKit::WebProcessPool::registerURLSchemeAsSecure):
(WebKit::WebProcessPool::registerURLSchemeAsBypassingContentSecurityPolicy):
(WebKit::WebProcessPool::registerURLSchemeAsLocal):
(WebKit::WebProcessPool::registerURLSchemeAsNoAccess):
(WebKit::WebProcessPool::registerGlobalURLSchemeAsHavingCustomProtocolHandlers):
(WebKit::WebProcessPool::unregisterGlobalURLSchemeAsHavingCustomProtocolHandlers):
(WebKit::WebProcessPool::setCacheModel):
(WebKit::WebProcessPool::setCacheModelSynchronouslyForTesting):
(WebKit::WebProcessPool::createDownloadProxy):
(WebKit::WebProcessPool::terminateServiceWorkers):
(WebKit::WebProcessPool::serviceWorkerProcesses):
(WebKit::WebProcessPool::updateProcessAssertions):
(WebKit::WebProcessPool::isServiceWorkerPageID const):
(WebKit::WebProcessPool::setUseSeparateServiceWorkerProcess):
(WebKit::WebProcessPool::hasServiceWorkerForegroundActivityForTesting const):
(WebKit::WebProcessPool::hasServiceWorkerBackgroundActivityForTesting const):
(WebKit::WebProcessPool::setLegacyCustomProtocolManagerClient): Deleted.
(WebKit::WebProcessPool::networkingProcessConnection): Deleted.
(WebKit::WebProcessPool::ensureNetworkProcess): Deleted.
There seems to be a lot of code removal here, but most of it was duplicate code with
WebsiteDataStore::parameters which is used instead. That which was not duplicate was
moved to NetworkProcessProxy::sendCreationParametersToNewProcess. This was the hardest part of this patch.
(WebKit::WebProcessPool::getNetworkProcessConnection): Deleted.
(WebKit::WebProcessPool::networkProcessIdentifier): Deleted.
(WebKit::WebProcessPool::synthesizeAppIsBackground): Deleted.
(WebKit::WebProcessPool::sendSyncToNetworkingProcess): Deleted.
(WebKit::WebProcessPool::clearCachedCredentials): Deleted.
(WebKit::WebProcessPool::terminateNetworkProcess): Deleted.
(WebKit::WebProcessPool::sendNetworkProcessPrepareToSuspendForTesting): Deleted.
(WebKit::WebProcessPool::sendNetworkProcessWillSuspendImminentlyForTesting): Deleted.
(WebKit::WebProcessPool::sendNetworkProcessDidResume): Deleted.
(WebKit::WebProcessPool::flushCookies): Deleted.
(WebKit::WebProcessPool::didCommitCrossSiteLoadWithDataTransfer): Deleted.
* UIProcess/WebProcessPool.h:
(WebKit::WebProcessPool::sendToNetworkingProcess): Deleted.
(WebKit::WebProcessPool::sendToNetworkingProcessRelaunchingIfNecessary): Deleted.
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::setWebsiteDataStore):
(WebKit::WebProcessProxy::getNetworkProcessConnection):
(WebKit::WebProcessProxy::didFinishLaunching):
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::WebsiteDataStore::networkProcessHasEntitlementForTesting):
(WebKit::WebsiteDataStore::sendNetworkProcessXPCEndpointToWebProcess):
(WebKit::WebsiteDataStore::sendNetworkProcessXPCEndpointToAllWebProcesses):
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::forEachWebsiteDataStore):
(WebKit::WebsiteDataStore::WebsiteDataStore):
(WebKit::WebsiteDataStore::~WebsiteDataStore):
(WebKit::WebsiteDataStore::registerWithSessionIDMap):
(WebKit::WebsiteDataStore::existingDataStoreForSessionID):
(WebKit::WebsiteDataStore::networkProcess):
(WebKit::WebsiteDataStore::networkProcess const):
(WebKit::WebsiteDataStore::fetchDataAndApply):
(WebKit::WebsiteDataStore::removeData):
(WebKit::WebsiteDataStore::setServiceWorkerTimeoutForTesting):
(WebKit::WebsiteDataStore::resetServiceWorkerTimeoutForTesting):
(WebKit::WebsiteDataStore::setMaxStatisticsEntries):
(WebKit::WebsiteDataStore::setPruneEntriesDownTo):
(WebKit::WebsiteDataStore::setGrandfatheringTime):
(WebKit::WebsiteDataStore::setMinimumTimeBetweenDataRecordsRemoval):
(WebKit::WebsiteDataStore::dumpResourceLoadStatistics):
(WebKit::WebsiteDataStore::isPrevalentResource):
(WebKit::WebsiteDataStore::isGrandfathered):
(WebKit::WebsiteDataStore::setPrevalentResource):
(WebKit::WebsiteDataStore::setPrevalentResourceForDebugMode):
(WebKit::WebsiteDataStore::isVeryPrevalentResource):
(WebKit::WebsiteDataStore::setVeryPrevalentResource):
(WebKit::WebsiteDataStore::setShouldClassifyResourcesBeforeDataRecordsRemoval):
(WebKit::WebsiteDataStore::setSubframeUnderTopFrameDomain):
(WebKit::WebsiteDataStore::isRegisteredAsSubFrameUnder):
(WebKit::WebsiteDataStore::setSubresourceUnderTopFrameDomain):
(WebKit::WebsiteDataStore::isRegisteredAsSubresourceUnder):
(WebKit::WebsiteDataStore::setSubresourceUniqueRedirectTo):
(WebKit::WebsiteDataStore::setSubresourceUniqueRedirectFrom):
(WebKit::WebsiteDataStore::setTopFrameUniqueRedirectTo):
(WebKit::WebsiteDataStore::setTopFrameUniqueRedirectFrom):
(WebKit::WebsiteDataStore::isRegisteredAsRedirectingTo):
(WebKit::WebsiteDataStore::clearPrevalentResource):
(WebKit::WebsiteDataStore::resetParametersToDefaultValues):
(WebKit::WebsiteDataStore::submitTelemetry):
(WebKit::WebsiteDataStore::scheduleClearInMemoryAndPersistent):
(WebKit::WebsiteDataStore::getResourceLoadStatisticsDataSummary):
(WebKit::WebsiteDataStore::scheduleCookieBlockingUpdate):
(WebKit::WebsiteDataStore::scheduleStatisticsAndDataRecordsProcessing):
(WebKit::WebsiteDataStore::statisticsDatabaseHasAllTables):
(WebKit::WebsiteDataStore::setLastSeen):
(WebKit::WebsiteDataStore::domainIDExistsInDatabase):
(WebKit::WebsiteDataStore::mergeStatisticForTesting):
(WebKit::WebsiteDataStore::insertExpiredStatisticForTesting):
(WebKit::WebsiteDataStore::setNotifyPagesWhenDataRecordsWereScanned):
(WebKit::WebsiteDataStore::setIsRunningResourceLoadStatisticsTest):
(WebKit::WebsiteDataStore::getAllStorageAccessEntries):
(WebKit::WebsiteDataStore::setTimeToLiveUserInteraction):
(WebKit::WebsiteDataStore::logUserInteraction):
(WebKit::WebsiteDataStore::hasHadUserInteraction):
(WebKit::WebsiteDataStore::isRelationshipOnlyInDatabaseOnce):
(WebKit::WebsiteDataStore::clearUserInteraction):
(WebKit::WebsiteDataStore::setGrandfathered):
(WebKit::WebsiteDataStore::setUseITPDatabase):
(WebKit::WebsiteDataStore::setCrossSiteLoadWithLinkDecorationForTesting):
(WebKit::WebsiteDataStore::resetCrossSiteLoadsWithLinkDecorationForTesting):
(WebKit::WebsiteDataStore::deleteCookiesForTesting):
(WebKit::WebsiteDataStore::hasLocalStorageForTesting const):
(WebKit::WebsiteDataStore::hasIsolatedSessionForTesting const):
(WebKit::WebsiteDataStore::setResourceLoadStatisticsShouldDowngradeReferrerForTesting):
(WebKit::WebsiteDataStore::setThirdPartyCookieBlockingMode):
(WebKit::WebsiteDataStore::setResourceLoadStatisticsShouldEnbleSameSiteStrictEnforcementForTesting):
(WebKit::WebsiteDataStore::setResourceLoadStatisticsFirstPartyWebsiteDataRemovalModeForTesting):
(WebKit::WebsiteDataStore::setResourceLoadStatisticsToSameSiteStrictCookiesForTesting):
(WebKit::WebsiteDataStore::setResourceLoadStatisticsFirstPartyHostCNAMEDomainForTesting):
(WebKit::WebsiteDataStore::setResourceLoadStatisticsThirdPartyCNAMEDomainForTesting):
(WebKit::WebsiteDataStore::syncLocalStorage):
(WebKit::WebsiteDataStore::setCacheMaxAgeCapForPrevalentResources):
(WebKit::WebsiteDataStore::resetCacheMaxAgeCapForPrevalentResources):
(WebKit::WebsiteDataStore::processPools const):
(WebKit::WebsiteDataStore::allowSpecificHTTPSCertificateForHost):
(WebKit::WebsiteDataStore::getNetworkProcessConnection):
(WebKit::WebsiteDataStore::networkProcessCrashed):
(WebKit::WebsiteDataStore::terminateNetworkProcess):
(WebKit::WebsiteDataStore::sendNetworkProcessPrepareToSuspendForTesting):
(WebKit::WebsiteDataStore::sendNetworkProcessWillSuspendImminentlyForTesting):
(WebKit::WebsiteDataStore::sendNetworkProcessDidResume):
(WebKit::WebsiteDataStore::setResourceLoadStatisticsEnabled):
(WebKit::WebsiteDataStore::setStatisticsTestingCallback):
(WebKit::WebsiteDataStore::setResourceLoadStatisticsDebugMode):
(WebKit::WebsiteDataStore::isResourceLoadStatisticsEphemeral const):
(WebKit::WebsiteDataStore::setAdClickAttributionDebugMode):
(WebKit::WebsiteDataStore::flushCookies):
(WebKit::WebsiteDataStore::setAllowsAnySSLCertificateForWebSocket):
(WebKit::WebsiteDataStore::clearCachedCredentials):
(WebKit::WebsiteDataStore::parameters):
(WebKit::WebsiteDataStore::getLocalStorageDetails):
(WebKit::WebsiteDataStore::resetQuota):
(WebKit::WebsiteDataStore::networkProcessHasEntitlementForTesting):
(WebKit::WebsiteDataStore::renameOriginInWebsiteData):
(WebKit::WebsiteDataStore::hasAppBoundSession const):
(WebKit::WebsiteDataStore::clearAppBoundSession):
(WebKit::WebsiteDataStore::forwardAppBoundDomainsToITPIfInitialized):
(WebKit::WebsiteDataStore::setAppBoundDomainsForITP):
(WebKit::WebsiteDataStore::updateBundleIdentifierInNetworkProcess):
(WebKit::WebsiteDataStore::clearBundleIdentifierInNetworkProcess):
(WebKit::WebsiteDataStore::makeNextNetworkProcessLaunchFailForTesting):
(WebKit::WebsiteDataStore::shouldMakeNextNetworkProcessLaunchFailForTesting):
(WebKit::WebsiteDataStore::maybeRegisterWithSessionIDMap): Deleted.
(WebKit::WebsiteDataStore::existingNonDefaultDataStoreForSessionID): Deleted.
(WebKit::WebsiteDataStore::processPoolForCookieStorageOperations): Deleted.
(WebKit::WebsiteDataStore::isAssociatedProcessPool const): Deleted.
(WebKit::WebsiteDataStore::pendingCookies const): Deleted.
(WebKit::WebsiteDataStore::addPendingCookie): Deleted.
(WebKit::WebsiteDataStore::removePendingCookie): Deleted.
(WebKit::WebsiteDataStore::clearPendingCookies): Deleted.
(WebKit::WebsiteDataStore::didCreateNetworkProcess): Deleted.
* UIProcess/WebsiteData/WebsiteDataStore.h:
* UIProcess/WebsiteData/curl/WebsiteDataStoreCurl.cpp:
(WebKit::WebsiteDataStore::setNetworkProxySettings):
* UIProcess/WebsiteData/soup/WebsiteDataStoreSoup.cpp:
(WebKit::WebsiteDataStore::platformSetNetworkParameters):
(WebKit::WebsiteDataStore::setPersistentCredentialStorageEnabled):
* UIProcess/glib/WebProcessProxyGLib.cpp:
(WebKit::WebProcessProxy::platformGetLaunchOptions):
* UIProcess/soup/WebCookieManagerProxySoup.cpp:
(WebKit::WebCookieManagerProxy::setCookiePersistentStorage):
* UIProcess/soup/WebProcessPoolSoup.cpp:
(WebKit::WebProcessPool::platformInitializeNetworkProcess):
(WebKit::WebProcessPool::setIgnoreTLSErrors):
(WebKit::WebProcessPool::setNetworkProxySettings):
* WebProcess/Network/NetworkProcessConnectionInfo.h:
2020-09-29 Tim Horton <timothy_horton@apple.com>
Enable the Web Share Level 2 feature wherever Web Share is enabled
https://bugs.webkit.org/show_bug.cgi?id=217105
Reviewed by Beth Dakin.
No new tests; there are already tests that enable this.
* Shared/WebPreferencesExperimental.yaml:
2020-09-29 Aditya Keerthi <akeerthi@apple.com>
[macCatalyst] Focus rings are not painted
https://bugs.webkit.org/show_bug.cgi?id=217073
<rdar://problem/46794111>
Reviewed by Darin Adler.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
Ensure that a default focus ring color is specified on macCatalyst.
2020-09-29 Alex Christensen <achristensen@webkit.org>
Remove plist-based ResourceLoadStatistics storage, which has been replaced by database-based storage
https://bugs.webkit.org/show_bug.cgi?id=217063
Reviewed by John Wilander.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:
(WebKit::ResourceLoadStatisticsMemoryStore::setPersistentStorage): Deleted.
(WebKit::ResourceLoadStatisticsMemoryStore::syncStorageIfNeeded): Deleted.
(WebKit::ResourceLoadStatisticsMemoryStore::syncStorageImmediately): Deleted.
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsPersistentStorage.cpp: Removed.
* NetworkProcess/Classifier/ResourceLoadStatisticsPersistentStorage.h: Removed.
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.cpp:
(WebKit::ResourceLoadStatisticsStore::processStatisticsAndDataRecords):
(WebKit::ResourceLoadStatisticsStore::grandfatherExistingWebsiteData):
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.h:
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::WebResourceLoadStatisticsStore):
(WebKit::WebResourceLoadStatisticsStore::~WebResourceLoadStatisticsStore):
(WebKit::WebResourceLoadStatisticsStore::didDestroyNetworkSession):
(WebKit::WebResourceLoadStatisticsStore::populateMemoryStoreFromDisk):
(WebKit::WebResourceLoadStatisticsStore::scheduleClearInMemoryAndPersistent):
(WebKit::WebResourceLoadStatisticsStore::flushAndDestroyPersistentStore): Deleted.
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.h:
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::didClose):
(WebKit::NetworkProcess::initializeNetworkProcess):
(WebKit::NetworkProcess::setUseITPDatabase): Deleted.
* NetworkProcess/NetworkProcess.h:
(WebKit::NetworkProcess::isITPDatabaseEnabled const): Deleted.
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/NetworkProcessCreationParameters.cpp:
(WebKit::NetworkProcessCreationParameters::encode const):
(WebKit::NetworkProcessCreationParameters::decode):
* NetworkProcess/NetworkProcessCreationParameters.h:
* NetworkProcess/NetworkSession.cpp:
(WebKit::NetworkSession::destroyResourceLoadStatistics):
(WebKit::NetworkSession::flushAndDestroyPersistentStore): Deleted.
* NetworkProcess/NetworkSession.h:
* Shared/WebPreferencesInternal.yaml:
* Sources.txt:
* SourcesCocoa.txt:
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetIsITPDatabaseEnabled): Deleted.
(WKPreferencesGetIsITPDatabaseEnabled): Deleted.
* UIProcess/API/C/WKPreferencesRef.h:
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreStatisticsResetToConsistentState):
(WKWebsiteDataStoreSetUseITPDatabase): Deleted.
* UIProcess/API/C/WKWebsiteDataStoreRef.h:
* UIProcess/API/Cocoa/WKPreferences.mm:
(-[WKPreferences _isITPDatabaseEnabled]): Deleted.
* UIProcess/API/Cocoa/WKPreferencesPrivate.h:
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
(-[WKWebsiteDataStore _setUseITPDatabase:completionHandler:]): Deleted.
* UIProcess/API/Cocoa/WKWebsiteDataStorePrivate.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeNetworkProcess):
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::setUseITPDatabase): Deleted.
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::setUseITPDatabase): Deleted.
* UIProcess/WebsiteData/WebsiteDataStore.h:
* UIProcess/ios/ResourceLoadStatisticsPersistentStorageIOS.mm: Removed.
* WebKit.xcodeproj/project.pbxproj:
2020-09-29 Chris Dumez <cdumez@apple.com>
Add stubs for AudioWorklet
https://bugs.webkit.org/show_bug.cgi?id=217059
Reviewed by Eric Carlson.
Add experimental feature flag for AudioWorklet API.
* Shared/WebPreferencesExperimental.yaml:
2020-09-29 Per Arne Vollan <pvollan@apple.com>
[macOS] Deny access to 'com.apple.awdd' in the WebContent sandbox
https://bugs.webkit.org/show_bug.cgi?id=216975
Reviewed by Darin Adler.
This has already been done on iOS in <https://trac.webkit.org/changeset/254376>, and should be done on macOS as well.
* WebProcess/com.apple.WebProcess.sb.in:
2020-09-29 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] Stop using the default website data store for the inspector
https://bugs.webkit.org/show_bug.cgi?id=217034
Reviewed by Michael Catanzaro.
The GTK should never use the default website data store. Use a dedicated website data store for the inspector
with its own configuration to store persistent data.
* UIProcess/Inspector/gtk/WebInspectorProxyGtk.cpp:
(WebKit::inspectorWebsiteDataStore):
(WebKit::WebInspectorProxy::platformCreateFrontendPage):
2020-09-28 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process] Use VideoLayerManager to manage layers of MediaPlayerPrivateRemote
https://bugs.webkit.org/show_bug.cgi?id=216995
Reviewed by Eric Carlson.
This patch removes the remote hosting layer in the GPU process for video fullscreen and
picture-in-picture. We don't need to keep a layer in the GPU process for video fullscreen
and picture-in-picture because no rendering will be done for that layer. We don't need
EnterFullscreen, ExitFullscreen and SetVideoFullscreenFrameFenced IPC messages after removing
the layer because the RemoteMediaPlayerPrivateRemote in the web process can deal with
the video presentation mode change.
* GPUProcess/media/RemoteMediaPlayerProxy.h:
* GPUProcess/media/RemoteMediaPlayerProxy.messages.in:
* GPUProcess/media/cocoa/RemoteMediaPlayerProxyCocoa.mm:
(WebKit::RemoteMediaPlayerProxy::prepareForPlayback):
(WebKit::RemoteMediaPlayerProxy::enterFullscreen): Deleted.
(WebKit::RemoteMediaPlayerProxy::exitFullscreen): Deleted.
(WebKit::RemoteMediaPlayerProxy::setVideoFullscreenFrameFenced): Deleted.
Remove unused IPC messages after removing the remote hosting layer for video fullscreen
and picture-in-picture.
* GPUProcess/media/gstreamer/RemoteMediaPlayerProxyGStreamer.cpp:
(WebKit::RemoteMediaPlayerProxy::prepareForPlayback):
(WebKit::RemoteMediaPlayerProxy::enterFullscreen): Deleted.
(WebKit::RemoteMediaPlayerProxy::exitFullscreen): Deleted.
Ditto.
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::MediaPlayerPrivateRemote):
(WebKit::MediaPlayerPrivateRemote::~MediaPlayerPrivateRemote):
(WebKit::MediaPlayerPrivateRemote::prepareForPlayback):
Remove the code related to the remote hosting layer for video fullscreen and picture-in-picture.
(WebKit::MediaPlayerPrivateRemote::platformLayer const):
(WebKit::MediaPlayerPrivateRemote::setVideoFullscreenLayer):
Use VideoLayerManager to manage the layer.
(WebKit::MediaPlayerPrivateRemote::setVideoFullscreenFrame): Ditto.
(WebKit::MediaPlayerPrivateRemote::requiresTextTrackRepresentation const): Ditto.
(WebKit::MediaPlayerPrivateRemote::setTextTrackRepresentation): Ditto.
(WebKit::MediaPlayerPrivateRemote::syncTextTrackBounds): Ditto.
(WebKit::MediaPlayerPrivateRemote::setVideoFullscreenFrameFenced): Deleted.
* WebProcess/GPU/media/MediaPlayerPrivateRemote.h:
Need to move the Logger up in the data members so that other members can use it
in the constructor, e.g., VideoLayerManager.
* WebProcess/GPU/media/cocoa/MediaPlayerPrivateRemoteCocoa.mm:
(WebKit::MediaPlayerPrivateRemote::MediaPlayerPrivateRemote):
(WebKit::MediaPlayerPrivateRemote::createVideoFullscreenLayer):
(WebKit::MediaPlayerPrivateRemote::setVideoFullscreenFrame): Deleted.
(WebKit::MediaPlayerPrivateRemote::setTextTrackRepresentation): Deleted.
(WebKit::MediaPlayerPrivateRemote::syncTextTrackBounds): Deleted.
Simplify the implementation by managing layers with VideoLayerManager.
2020-09-28 Eric Carlson <eric.carlson@apple.com>
[GPUP] Out-of-band TextTracks
https://bugs.webkit.org/show_bug.cgi?id=217062
<rdar://problem/68739969>
Reviewed by Jer Noble.
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
(WebKit::RemoteMediaPlayerProxy::outOfBandTrackSources):
* GPUProcess/media/RemoteMediaPlayerProxyConfiguration.h:
(WebKit::RemoteMediaPlayerProxyConfiguration::encode const):
(WebKit::RemoteMediaPlayerProxyConfiguration::decode):
* Scripts/webkit/messages.py:
* WebProcess/GPU/media/RemoteMediaPlayerManager.cpp:
(WebKit::RemoteMediaPlayerManager::createRemoteMediaPlayer):
* WebProcess/GPU/media/RemoteMediaPlayerManager.h:
2020-09-28 Devin Rousso <drousso@apple.com>
[iOS] unable to airplay directly loaded fullscreen video
https://bugs.webkit.org/show_bug.cgi?id=216858
<rdar://problem/68746321>
Reviewed by Eric Carlson.
It's possible that `MediaPlayerPrivate::wirelessVideoPlaybackDisabled` is queried before a
"real" `MediaPlayerPrivate` (i.e. not `NullMediaPlayerPrivate`) is initialized through one
of the registered `MediaPlayerFactory` "engine"s. It's further possible that the value
returned by `MediaPlayerPrivate::wirelessVideoPlaybackDisabled` will change as the new
"real" `MediaPlayerPrivate` may have different logic. In this case, the `MediaPlayer`
doesn't update its client (and therefore nothing "up" the tree, including the client's
clients) with the new `MediaPlayerPrivate::wirelessVideoPlaybackDisabled` value, meaning
that the old value will still be used.
Add piping such that when the `MediaPlayerFactory` "engine" changes, the `MediaPlayer`
notifies its client (which is further piped "up" the tree to the client's clients) which
eventually causes `MediaPlayerPrivate::wirelessVideoPlaybackDisabled` to be re-evaluated.
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::playbackControlsMediaEngineChanged): Added.
* WebProcess/cocoa/PlaybackSessionManager.h:
* WebProcess/cocoa/PlaybackSessionManager.mm:
(WebKit::PlaybackSessionManager::mediaEngineChanged): Added.
* UIProcess/API/Cocoa/WKWebViewPrivateForTesting.h:
* UIProcess/API/Cocoa/WKWebViewTesting.mm:
(-[WKWebView _wirelessVideoPlaybackDisabled]): Added.
* UIProcess/Cocoa/PlaybackSessionManagerProxy.h:
* UIProcess/Cocoa/PlaybackSessionManagerProxy.mm:
(WebKit::PlaybackSessionManagerProxy::wirelessVideoPlaybackDisabled): Added.
Create additional functions for piping data to tests.
2020-09-28 Youenn Fablet <youenn@apple.com>
Make sure our calls to AVCaptureDevice requestAccessForMediaType do processing on the main thread
https://bugs.webkit.org/show_bug.cgi?id=216974
Reviewed by Darin Adler.
The completion handler to [AVCaptureDeviceClass requestAccessForMediaType:] may sometimes be called in a background thread on iOS.
Make sure to hop to the main thread if that is the case.
Also make sure to ref/weakref lambda captured variables.
* UIProcess/Cocoa/UIDelegate.h:
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::requestAccessForMediaType):
(WebKit::UIDelegate::UIClient::decidePolicyForUserMediaPermissionRequest):
2020-09-27 Lauro Moura <lmoura@igalia.com>
REGRESSION(r267688) [GTK] Many compositing timeouts
https://bugs.webkit.org/show_bug.cgi?id=217044
Reviewed by Darin Adler.
Especulative fix by adding missing finalize step to
updateRendering (or calling isolatedUpdateRendering).
Covered by existing tests.
* WebProcess/WebPage/CoordinatedGraphics/DrawingAreaCoordinatedGraphics.cpp:
(WebKit::DrawingAreaCoordinatedGraphics::updateBackingStoreState):
Call finalizeUpdateRendering.
(WebKit::DrawingAreaCoordinatedGraphics::display): Ditto.
* WebProcess/WebPage/CoordinatedGraphics/LayerTreeHostTextureMapper.cpp:
(WebKit::LayerTreeHost::flushAndRenderLayers): Call
isolatedUpdateRendering instead of updateRendering on the corePage().
2020-09-25 Simon Fraser <simon.fraser@apple.com>
WebKitLegacy should call Page::finalizeRenderingUpdate()
https://bugs.webkit.org/show_bug.cgi?id=216958
Reviewed by Tim Horton.
dynamicViewportSizeUpdate() needs to call isolatedUpdateRendering() because it isn't followed
by a finalizeRenderingUpdate().
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::dynamicViewportSizeUpdate):
2020-09-27 Carlos Garcia Campos <cgarcia@igalia.com>
[SOUP] WebSocket: cookies set in request don't appear in the inspector
https://bugs.webkit.org/show_bug.cgi?id=217012
Reviewed by Michael Catanzaro.
Cookies are set by libsoup and we are notifying about the handshake being sent before the cookie header is added
to the request.
* NetworkProcess/soup/WebSocketTaskSoup.cpp:
(WebKit::WebSocketTask::WebSocketTask): Connect to SoupMessage::starting to notify the web process the handshake
request has been sent.
2020-09-26 Carlos Garcia Campos <cgarcia@igalia.com>
[SOUP] Do not set site for cookies twice in case of redirection
https://bugs.webkit.org/show_bug.cgi?id=217010
Reviewed by Michael Catanzaro.
ResourceRequest::updateSoupMessage() already fills the same site information, but we are duplicating the code in
NetworkDataTaskSoup::continueHTTPRedirection().
* NetworkProcess/soup/NetworkDataTaskSoup.cpp:
(WebKit::NetworkDataTaskSoup::continueHTTPRedirection): Do not set same site information here, it will be done
again in NetworkDataTaskSoup::createRequest().
2020-09-25 Hoa Dinh <dvh@apple.com>
Add -[WKWebView _loadRequest:shouldOpenExternalURLsPolicy:] to provide a way to allow app link
https://bugs.webkit.org/show_bug.cgi?id=216977
Reviewed by Alex Christensen.
The method -[WKWebView _loadRequest:shouldOpenExternalURLs:] would only allow opening external or not.
It wasn't providing the ability to allow external URL but not app links.
This patch adds -[WKWebView _loadRequest:shouldOpenExternalURLsPolicy:]. The parameters value for
shouldOpenExternalURLsPolicy has the same meaning than WebCore::ShouldOpenExternalURLsPolicy.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _loadRequest:shouldOpenExternalURLsPolicy:]):
* UIProcess/API/Cocoa/WKWebViewPrivate.h: Added enum _WKShouldOpenExternalURLsPolicy,
2020-09-25 Alex Christensen <achristensen@webkit.org>
Unreviewed, reverting r267608.
Caused API test failures
Reverted changeset:
"Move NetworkProcessProxy ownership from WebProcessPool to
WebsiteDataStore"
https://bugs.webkit.org/show_bug.cgi?id=216041
https://trac.webkit.org/changeset/267608
2020-09-25 Tim Horton <timothy_horton@apple.com>
Crunchyroll playback controls do not work on iPad with trackpad
https://bugs.webkit.org/show_bug.cgi?id=217000
<rdar://problem/66362029>
Reviewed by Wenson Hsieh.
* UIProcess/ios/WKContentViewInteraction.mm:
(applicationIsKnownToIgnoreMouseEvents):
Add it to the list.
2020-09-25 Alex Christensen <achristensen@webkit.org>
Move NetworkProcessProxy ownership from WebProcessPool to WebsiteDataStore
https://bugs.webkit.org/show_bug.cgi?id=216041
Reviewed by Brady Eidson.
This patch moves the things on WebProcessPool associated with NetworkProcessProxy ownership and moves them
to WebsiteDataStore. It's pretty mechanical, but it reaches its fingers into a lot of things. Some comments
on interesting parts are below.
Why am I doing this? I'm glad you asked.
WKWebView had a fundamental cookie problem before this change. It had a WKWebsiteDataStore that could point to a cookie store on disk
that could be used in multiple network processes at the same time. This led to problems with cookies being received over the network
in those different processes being unable to be shared until they were written to disk, something that does not happen immediately, especially
on iOS. This problem is even worse with session cookies, which should be kept in memory and never be written to disk. Those processes
did not use shared memory to share the cookies. They just didn't share them, which led to cookie incorrectness. This was unfortunately
the default configuration if you call [[WKWebView alloc] init] more than once, which many developers do. There was also occasional cookie
corruption and loss from multiple processes writing cookies to the same file. All these problems can be worked around by doing what Safari does:
manually setting the WKProcessPool of all WKWebViewConfigurations to the same process pool to only use one network process. This is not intuitive
or well documented, and even though I have spent the majority of WWDC labs for the last 3 years telling developers to do it, most developers do not.
The default behavior should not lead to cookie incorrectness, corruption, or loss. This accomplishes that.
A bigger problem that can't really be worked around well is what we were doing in our WKHTTPCookieStore API. Many developers call setCookie:completionHandler:
to "log in" the user by setting a cookie, then in the completion handler, they open a WKWebView to their page and expect the cookie to be sent. Before this
change, we would not know which process pool to set the cookie in so we would guess. If there were no process pools, we would start writing to disk and hope
it writes fast enough. We can't wait for it to complete because it causes hangs like rdar://problem/66961066. This solution allows to start the network
process of that WebsiteDataStore and set the cookie in that process. We now know that is the only process this WebsiteDataStore will ever use. Now, users
will be able to actually be logged in all the time in such apps.
* NetworkProcess/Cookies/WebCookieManager.cpp:
(WebKit::WebCookieManager::setHTTPCookieAcceptPolicy):
(WebKit::WebCookieManager::getHTTPCookieAcceptPolicy):
* NetworkProcess/Cookies/WebCookieManager.h:
* NetworkProcess/Cookies/WebCookieManager.messages.in:
* NetworkProcess/Cookies/curl/WebCookieManagerCurl.cpp:
(WebKit::WebCookieManager::platformSetHTTPCookieAcceptPolicy):
* NetworkProcess/Cookies/mac/WebCookieManagerMac.mm:
(WebKit::WebCookieManager::platformSetHTTPCookieAcceptPolicy):
* NetworkProcess/Cookies/soup/WebCookieManagerSoup.cpp:
(WebKit::WebCookieManager::platformSetHTTPCookieAcceptPolicy):
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::initializeNetworkProcess):
(WebKit::NetworkProcess::ensureSession):
(WebKit::NetworkProcess::storageSession const):
(WebKit::NetworkProcess::forEachNetworkStorageSession):
(WebKit::NetworkProcess::defaultStorageSession const): Deleted.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcessCreationParameters.cpp:
(WebKit::NetworkProcessCreationParameters::encode const):
(WebKit::NetworkProcessCreationParameters::decode):
* NetworkProcess/NetworkProcessCreationParameters.h:
* NetworkProcess/WebStorage/LocalStorageDatabase.cpp:
(WebKit::LocalStorageDatabase::close):
* NetworkProcess/cocoa/NetworkProcessCocoa.mm:
(WebKit::NetworkProcess::platformInitializeNetworkProcessCocoa):
(WebKit::saveCookies):
(WebKit::NetworkProcess::platformCreateDefaultStorageSession const): Deleted.
* NetworkProcess/curl/NetworkProcessCurl.cpp:
(WebKit::NetworkProcess::platformCreateDefaultStorageSession const): Deleted.
* NetworkProcess/mac/RemoteNetworkingContext.mm:
(WebKit::RemoteNetworkingContext::ensureWebsiteDataStoreSession):
* NetworkProcess/soup/NetworkProcessSoup.cpp:
(WebKit::NetworkProcess::platformCreateDefaultStorageSession const): Deleted.
* Platform/IPC/Attachment.h:
* Shared/WebsiteDataStoreParameters.cpp:
(WebKit::WebsiteDataStoreParameters::encode const):
(WebKit::WebsiteDataStoreParameters::decode):
* Shared/WebsiteDataStoreParameters.h:
* UIProcess/API/APIHTTPCookieStore.cpp:
(API::HTTPCookieStore::HTTPCookieStore):
(API::HTTPCookieStore::~HTTPCookieStore):
(API::HTTPCookieStore::cookies):
(API::HTTPCookieStore::cookiesForURL):
(API::HTTPCookieStore::setCookies):
(API::HTTPCookieStore::deleteCookie):
(API::HTTPCookieStore::deleteAllCookies):
(API::HTTPCookieStore::setHTTPCookieAcceptPolicy):
(API::HTTPCookieStore::registerObserver):
(API::HTTPCookieStore::cookieManagerDestroyed):
(API::HTTPCookieStore::registerForNewProcessPoolNotifications): Deleted.
(API::HTTPCookieStore::unregisterForNewProcessPoolNotifications): Deleted.
* UIProcess/API/APIHTTPCookieStore.h:
* UIProcess/API/C/WKContext.cpp:
(WKContextSetUseSeparateServiceWorkerProcess):
(WKContextSetPrimaryWebsiteDataStore):
(WKContextGetCookieManager): Deleted.
(WKContextSetAllowsAnySSLCertificateForServiceWorkerTesting): Deleted.
(WKContextTerminateNetworkProcess): Deleted.
(WKContextGetNetworkProcessIdentifier): Deleted.
* UIProcess/API/C/WKContext.h:
* UIProcess/API/C/WKContextPrivate.h:
* UIProcess/API/C/WKPage.cpp:
(WKPageGetWebsiteDataStore):
* UIProcess/API/C/WKPage.h:
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreTerminateNetworkProcess):
(WKWebsiteDataStoreGetNetworkProcessIdentifier):
* UIProcess/API/C/WKWebsiteDataStoreRef.h:
* UIProcess/API/Cocoa/WKProcessPool.mm:
(-[WKProcessPool _setUseSeparateServiceWorkerProcess:]):
(-[WKProcessPool _terminateNetworkProcess]): Deleted.
(-[WKProcessPool _sendNetworkProcessWillSuspendImminently]): Deleted.
(-[WKProcessPool _sendNetworkProcessPrepareToSuspend:]): Deleted.
(-[WKProcessPool _sendNetworkProcessDidResume]): Deleted.
(-[WKProcessPool _networkProcessIdentifier]): Deleted.
(-[WKProcessPool _makeNextNetworkProcessLaunchFailForTesting]): Deleted.
(-[WKProcessPool _synthesizeAppIsBackground:]): Deleted.
(-[WKProcessPool _setAllowsAnySSLCertificateForServiceWorker:]): Deleted.
(-[WKProcessPool _networkProcessHasEntitlementForTesting:]): Deleted.
* UIProcess/API/Cocoa/WKProcessPoolPrivate.h:
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
(-[WKWebsiteDataStore _networkProcessHasEntitlementForTesting:]):
(-[WKWebsiteDataStore _terminateNetworkProcess]):
(-[WKWebsiteDataStore _sendNetworkProcessPrepareToSuspend:]):
(-[WKWebsiteDataStore _sendNetworkProcessWillSuspendImminently]):
(-[WKWebsiteDataStore _sendNetworkProcessDidResume]):
(-[WKWebsiteDataStore _synthesizeAppIsBackground:]):
(-[WKWebsiteDataStore _networkProcessIdentifier]):
(+[WKWebsiteDataStore _makeNextNetworkProcessLaunchFailForTesting]):
* UIProcess/API/Cocoa/WKWebsiteDataStorePrivate.h:
* UIProcess/API/glib/WebKitCookieManager.cpp:
(_WebKitCookieManagerPrivate::cookieManager const):
(_WebKitCookieManagerPrivate::~_WebKitCookieManagerPrivate):
(webkitCookieManagerCreate):
(webkit_cookie_manager_set_persistent_storage):
(webkit_cookie_manager_set_accept_policy):
(webkit_cookie_manager_get_accept_policy):
(webkit_cookie_manager_add_cookie):
(webkit_cookie_manager_get_cookies):
(webkit_cookie_manager_delete_cookie):
* UIProcess/API/glib/WebKitWebContext.cpp:
(webkitWebContextConstructed):
(webkit_web_context_prefetch_dns):
* UIProcess/Automation/WebAutomationSession.cpp:
(WebKit::WebAutomationSession::addSingleCookie):
(WebKit::WebAutomationSession::deleteAllCookies):
* UIProcess/AuxiliaryProcessProxy.h:
* UIProcess/Cocoa/LegacyCustomProtocolManagerClient.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::updateProcessSuppressionState):
(WebKit::WebProcessPool::platformInitialize):
(WebKit::WebProcessPool::platformInitializeNetworkProcess):
(WebKit::WebProcessPool::networkProcessHasEntitlementForTesting): Deleted.
(WebKit::WebProcessPool::xpcEndpointMessage const): Deleted.
(WebKit::WebProcessPool::sendNetworkProcessXPCEndpointToWebProcess): Deleted.
* UIProcess/Downloads/DownloadProxy.cpp:
(WebKit::DownloadProxy::cancel):
(WebKit::DownloadProxy::publishProgress):
(WebKit::DownloadProxy::didReceiveAuthenticationChallenge):
(WebKit::DownloadProxy::willSendRequest):
(WebKit::DownloadProxy::decideDestinationWithSuggestedFilenameAsync):
* UIProcess/Downloads/DownloadProxyMap.cpp:
(WebKit::DownloadProxyMap::DownloadProxyMap):
(WebKit::DownloadProxyMap::applicationDidEnterBackground):
(WebKit::DownloadProxyMap::applicationWillEnterForeground):
(WebKit::DownloadProxyMap::createDownloadProxy):
(WebKit::DownloadProxyMap::downloadFinished):
(WebKit::DownloadProxyMap::invalidate):
* UIProcess/Downloads/DownloadProxyMap.h:
* UIProcess/Network/CustomProtocols/LegacyCustomProtocolManagerProxy.cpp:
(WebKit::LegacyCustomProtocolManagerProxy::startLoading):
(WebKit::LegacyCustomProtocolManagerProxy::stopLoading):
(WebKit::LegacyCustomProtocolManagerProxy::invalidate):
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::allNetworkProcesses):
(WebKit::defaultProcess):
(WebKit::NetworkProcessProxy::defaultNetworkProcess):
(WebKit::NetworkProcessProxy::terminate):
(WebKit::NetworkProcessProxy::didTerminate):
(WebKit::NetworkProcessProxy::sendCreationParametersToNewProcess):
(WebKit::anyProcessPoolAlwaysRunsAtBackgroundPriority):
(WebKit::anyProcessPoolShouldTakeUIBackgroundAssertion):
(WebKit::NetworkProcessProxy::NetworkProcessProxy):
(WebKit::NetworkProcessProxy::~NetworkProcessProxy):
(WebKit::NetworkProcessProxy::getLaunchOptions):
(WebKit::NetworkProcessProxy::createDownloadProxy):
(WebKit::NetworkProcessProxy::networkProcessCrashed):
(WebKit::NetworkProcessProxy::didReceiveMessage):
(WebKit::NetworkProcessProxy::didClose):
(WebKit::NetworkProcessProxy::didReceiveAuthenticationChallenge):
(WebKit::NetworkProcessProxy::didFinishLaunching):
(WebKit::NetworkProcessProxy::setDomainsWithUserInteraction):
(WebKit::NetworkProcessProxy::addSession):
(WebKit::NetworkProcessProxy::removeSession):
(WebKit::NetworkProcessProxy::websiteDataStoreFromSessionID):
(WebKit::NetworkProcessProxy::establishWorkerContextConnectionToNetworkProcess):
(WebKit::anyProcessPoolHasForegroundWebProcesses):
(WebKit::anyProcessPoolHasBackgroundWebProcesses):
(WebKit::NetworkProcessProxy::updateProcessAssertion):
(WebKit::NetworkProcessProxy::hasSession const): Deleted.
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/Network/NetworkProcessProxyCocoa.mm:
(WebKit::NetworkProcessProxy::XPCEventHandler::handleXPCEvent const):
* UIProcess/WebCookieManagerProxy.cpp:
(WebKit::WebCookieManagerProxy::WebCookieManagerProxy):
(WebKit::WebCookieManagerProxy::~WebCookieManagerProxy):
(WebKit::WebCookieManagerProxy::getHostnamesWithCookies):
(WebKit::WebCookieManagerProxy::deleteCookiesForHostnames):
(WebKit::WebCookieManagerProxy::deleteAllCookies):
(WebKit::WebCookieManagerProxy::deleteCookie):
(WebKit::WebCookieManagerProxy::deleteAllCookiesModifiedSince):
(WebKit::WebCookieManagerProxy::setCookies):
(WebKit::WebCookieManagerProxy::getAllCookies):
(WebKit::WebCookieManagerProxy::getCookies):
(WebKit::WebCookieManagerProxy::startObservingCookieChanges):
(WebKit::WebCookieManagerProxy::stopObservingCookieChanges):
(WebKit::WebCookieManagerProxy::setHTTPCookieAcceptPolicy):
(WebKit::WebCookieManagerProxy::getHTTPCookieAcceptPolicy):
(WebKit::WebCookieManagerProxy::supplementName): Deleted.
(WebKit::WebCookieManagerProxy::create): Deleted.
(WebKit::WebCookieManagerProxy::processPoolDestroyed): Deleted.
(WebKit::WebCookieManagerProxy::processDidClose): Deleted.
(WebKit::WebCookieManagerProxy::refWebContextSupplement): Deleted.
(WebKit::WebCookieManagerProxy::derefWebContextSupplement): Deleted.
* UIProcess/WebCookieManagerProxy.h:
(WebKit::WebCookieManagerProxy::create):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::setControlledByAutomation):
(WebKit::WebPageProxy::disableServiceWorkerEntitlementInNetworkProcess):
(WebKit::WebPageProxy::clearServiceWorkerEntitlementOverride):
(WebKit::WebPageProxy::preconnectTo):
(WebKit::WebPageProxy::didCommitLoadForFrame):
(WebKit::WebPageProxy::logFrameNavigation):
(WebKit::WebPageProxy::dumpAdClickAttribution):
(WebKit::WebPageProxy::clearAdClickAttribution):
(WebKit::WebPageProxy::setAdClickAttributionOverrideTimerForTesting):
(WebKit::WebPageProxy::setAdClickAttributionConversionURLForTesting):
(WebKit::WebPageProxy::markAdClickAttributionsAsExpiredForTesting):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::urlSchemesWithCustomProtocolHandlers):
(WebKit::WebProcessPool::WebProcessPool):
(WebKit::WebProcessPool::~WebProcessPool):
(WebKit::WebProcessPool::languageChanged):
(WebKit::WebProcessPool::sendMemoryPressureEvent):
(WebKit::WebProcessPool::networkProcessCrashed):
(WebKit::WebProcessPool::establishWorkerContextConnectionToNetworkProcess):
(WebKit::WebProcessPool::removeFromServiceWorkerProcesses):
(WebKit::WebProcessPool::tryTakePrewarmedProcess):
(WebKit::WebProcessPool::initializeNewWebProcess):
(WebKit::WebProcessPool::processDidFinishLaunching):
(WebKit::WebProcessPool::createWebPage):
(WebKit::WebProcessPool::updateServiceWorkerUserAgent):
(WebKit::WebProcessPool::pageBeginUsingWebsiteDataStore):
(WebKit::WebProcessPool::download):
(WebKit::WebProcessPool::resumeDownload):
(WebKit::WebProcessPool::registerURLSchemeAsSecure):
(WebKit::WebProcessPool::registerURLSchemeAsBypassingContentSecurityPolicy):
(WebKit::WebProcessPool::registerURLSchemeAsLocal):
(WebKit::WebProcessPool::registerURLSchemeAsNoAccess):
(WebKit::WebProcessPool::registerGlobalURLSchemeAsHavingCustomProtocolHandlers):
(WebKit::WebProcessPool::unregisterGlobalURLSchemeAsHavingCustomProtocolHandlers):
(WebKit::WebProcessPool::setCacheModel):
(WebKit::WebProcessPool::setCacheModelSynchronouslyForTesting):
(WebKit::WebProcessPool::createDownloadProxy):
(WebKit::WebProcessPool::terminateServiceWorkers):
(WebKit::WebProcessPool::serviceWorkerProcesses):
(WebKit::WebProcessPool::updateProcessAssertions):
(WebKit::WebProcessPool::isServiceWorkerPageID const):
(WebKit::WebProcessPool::setUseSeparateServiceWorkerProcess):
(WebKit::WebProcessPool::hasServiceWorkerForegroundActivityForTesting const):
(WebKit::WebProcessPool::hasServiceWorkerBackgroundActivityForTesting const):
(WebKit::WebProcessPool::setLegacyCustomProtocolManagerClient): Deleted.
(WebKit::WebProcessPool::networkingProcessConnection): Deleted.
(WebKit::WebProcessPool::ensureNetworkProcess): Deleted.
There seems to be a lot of code removal here, but most of it was duplicate code with
WebsiteDataStore::parameters which is used instead. That which was not duplicate was
moved to NetworkProcessProxy::sendCreationParametersToNewProcess. This was the hardest part of this patch.
(WebKit::WebProcessPool::getNetworkProcessConnection): Deleted.
(WebKit::WebProcessPool::networkProcessIdentifier): Deleted.
(WebKit::WebProcessPool::synthesizeAppIsBackground): Deleted.
(WebKit::WebProcessPool::sendSyncToNetworkingProcess): Deleted.
(WebKit::WebProcessPool::clearCachedCredentials): Deleted.
(WebKit::WebProcessPool::terminateNetworkProcess): Deleted.
(WebKit::WebProcessPool::sendNetworkProcessPrepareToSuspendForTesting): Deleted.
(WebKit::WebProcessPool::sendNetworkProcessWillSuspendImminentlyForTesting): Deleted.
(WebKit::WebProcessPool::sendNetworkProcessDidResume): Deleted.
(WebKit::WebProcessPool::flushCookies): Deleted.
(WebKit::WebProcessPool::didCommitCrossSiteLoadWithDataTransfer): Deleted.
* UIProcess/WebProcessPool.h:
(WebKit::WebProcessPool::sendToNetworkingProcess): Deleted.
(WebKit::WebProcessPool::sendToNetworkingProcessRelaunchingIfNecessary): Deleted.
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::setWebsiteDataStore):
(WebKit::WebProcessProxy::getNetworkProcessConnection):
(WebKit::WebProcessProxy::didFinishLaunching):
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::WebsiteDataStore::networkProcessHasEntitlementForTesting):
(WebKit::WebsiteDataStore::sendNetworkProcessXPCEndpointToWebProcess):
(WebKit::WebsiteDataStore::sendNetworkProcessXPCEndpointToAllWebProcesses):
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::forEachWebsiteDataStore):
(WebKit::WebsiteDataStore::WebsiteDataStore):
(WebKit::WebsiteDataStore::~WebsiteDataStore):
(WebKit::WebsiteDataStore::registerWithSessionIDMap):
(WebKit::WebsiteDataStore::existingDataStoreForSessionID):
(WebKit::WebsiteDataStore::networkProcess):
(WebKit::WebsiteDataStore::networkProcess const):
(WebKit::WebsiteDataStore::fetchDataAndApply):
(WebKit::WebsiteDataStore::removeData):
(WebKit::WebsiteDataStore::setServiceWorkerTimeoutForTesting):
(WebKit::WebsiteDataStore::resetServiceWorkerTimeoutForTesting):
(WebKit::WebsiteDataStore::setMaxStatisticsEntries):
(WebKit::WebsiteDataStore::setPruneEntriesDownTo):
(WebKit::WebsiteDataStore::setGrandfatheringTime):
(WebKit::WebsiteDataStore::setMinimumTimeBetweenDataRecordsRemoval):
(WebKit::WebsiteDataStore::dumpResourceLoadStatistics):
(WebKit::WebsiteDataStore::isPrevalentResource):
(WebKit::WebsiteDataStore::isGrandfathered):
(WebKit::WebsiteDataStore::setPrevalentResource):
(WebKit::WebsiteDataStore::setPrevalentResourceForDebugMode):
(WebKit::WebsiteDataStore::isVeryPrevalentResource):
(WebKit::WebsiteDataStore::setVeryPrevalentResource):
(WebKit::WebsiteDataStore::setShouldClassifyResourcesBeforeDataRecordsRemoval):
(WebKit::WebsiteDataStore::setSubframeUnderTopFrameDomain):
(WebKit::WebsiteDataStore::isRegisteredAsSubFrameUnder):
(WebKit::WebsiteDataStore::setSubresourceUnderTopFrameDomain):
(WebKit::WebsiteDataStore::isRegisteredAsSubresourceUnder):
(WebKit::WebsiteDataStore::setSubresourceUniqueRedirectTo):
(WebKit::WebsiteDataStore::setSubresourceUniqueRedirectFrom):
(WebKit::WebsiteDataStore::setTopFrameUniqueRedirectTo):
(WebKit::WebsiteDataStore::setTopFrameUniqueRedirectFrom):
(WebKit::WebsiteDataStore::isRegisteredAsRedirectingTo):
(WebKit::WebsiteDataStore::clearPrevalentResource):
(WebKit::WebsiteDataStore::resetParametersToDefaultValues):
(WebKit::WebsiteDataStore::submitTelemetry):
(WebKit::WebsiteDataStore::scheduleClearInMemoryAndPersistent):
(WebKit::WebsiteDataStore::getResourceLoadStatisticsDataSummary):
(WebKit::WebsiteDataStore::scheduleCookieBlockingUpdate):
(WebKit::WebsiteDataStore::scheduleStatisticsAndDataRecordsProcessing):
(WebKit::WebsiteDataStore::statisticsDatabaseHasAllTables):
(WebKit::WebsiteDataStore::setLastSeen):
(WebKit::WebsiteDataStore::domainIDExistsInDatabase):
(WebKit::WebsiteDataStore::mergeStatisticForTesting):
(WebKit::WebsiteDataStore::insertExpiredStatisticForTesting):
(WebKit::WebsiteDataStore::setNotifyPagesWhenDataRecordsWereScanned):
(WebKit::WebsiteDataStore::setIsRunningResourceLoadStatisticsTest):
(WebKit::WebsiteDataStore::getAllStorageAccessEntries):
(WebKit::WebsiteDataStore::setTimeToLiveUserInteraction):
(WebKit::WebsiteDataStore::logUserInteraction):
(WebKit::WebsiteDataStore::hasHadUserInteraction):
(WebKit::WebsiteDataStore::isRelationshipOnlyInDatabaseOnce):
(WebKit::WebsiteDataStore::clearUserInteraction):
(WebKit::WebsiteDataStore::setGrandfathered):
(WebKit::WebsiteDataStore::setUseITPDatabase):
(WebKit::WebsiteDataStore::setCrossSiteLoadWithLinkDecorationForTesting):
(WebKit::WebsiteDataStore::resetCrossSiteLoadsWithLinkDecorationForTesting):
(WebKit::WebsiteDataStore::deleteCookiesForTesting):
(WebKit::WebsiteDataStore::hasLocalStorageForTesting const):
(WebKit::WebsiteDataStore::hasIsolatedSessionForTesting const):
(WebKit::WebsiteDataStore::setResourceLoadStatisticsShouldDowngradeReferrerForTesting):
(WebKit::WebsiteDataStore::setThirdPartyCookieBlockingMode):
(WebKit::WebsiteDataStore::setResourceLoadStatisticsShouldEnbleSameSiteStrictEnforcementForTesting):
(WebKit::WebsiteDataStore::setResourceLoadStatisticsFirstPartyWebsiteDataRemovalModeForTesting):
(WebKit::WebsiteDataStore::setResourceLoadStatisticsToSameSiteStrictCookiesForTesting):
(WebKit::WebsiteDataStore::setResourceLoadStatisticsFirstPartyHostCNAMEDomainForTesting):
(WebKit::WebsiteDataStore::setResourceLoadStatisticsThirdPartyCNAMEDomainForTesting):
(WebKit::WebsiteDataStore::syncLocalStorage):
(WebKit::WebsiteDataStore::setCacheMaxAgeCapForPrevalentResources):
(WebKit::WebsiteDataStore::resetCacheMaxAgeCapForPrevalentResources):
(WebKit::WebsiteDataStore::processPools const):
(WebKit::WebsiteDataStore::allowSpecificHTTPSCertificateForHost):
(WebKit::WebsiteDataStore::getNetworkProcessConnection):
(WebKit::WebsiteDataStore::networkProcessCrashed):
(WebKit::WebsiteDataStore::terminateNetworkProcess):
(WebKit::WebsiteDataStore::sendNetworkProcessPrepareToSuspendForTesting):
(WebKit::WebsiteDataStore::sendNetworkProcessWillSuspendImminentlyForTesting):
(WebKit::WebsiteDataStore::sendNetworkProcessDidResume):
(WebKit::WebsiteDataStore::setResourceLoadStatisticsEnabled):
(WebKit::WebsiteDataStore::setStatisticsTestingCallback):
(WebKit::WebsiteDataStore::setResourceLoadStatisticsDebugMode):
(WebKit::WebsiteDataStore::isResourceLoadStatisticsEphemeral const):
(WebKit::WebsiteDataStore::setAdClickAttributionDebugMode):
(WebKit::WebsiteDataStore::flushCookies):
(WebKit::WebsiteDataStore::setAllowsAnySSLCertificateForWebSocket):
(WebKit::WebsiteDataStore::clearCachedCredentials):
(WebKit::WebsiteDataStore::parameters):
(WebKit::WebsiteDataStore::getLocalStorageDetails):
(WebKit::WebsiteDataStore::resetQuota):
(WebKit::WebsiteDataStore::networkProcessHasEntitlementForTesting):
(WebKit::WebsiteDataStore::renameOriginInWebsiteData):
(WebKit::WebsiteDataStore::hasAppBoundSession const):
(WebKit::WebsiteDataStore::clearAppBoundSession):
(WebKit::WebsiteDataStore::forwardAppBoundDomainsToITPIfInitialized):
(WebKit::WebsiteDataStore::setAppBoundDomainsForITP):
(WebKit::WebsiteDataStore::updateBundleIdentifierInNetworkProcess):
(WebKit::WebsiteDataStore::clearBundleIdentifierInNetworkProcess):
(WebKit::WebsiteDataStore::makeNextNetworkProcessLaunchFailForTesting):
(WebKit::WebsiteDataStore::shouldMakeNextNetworkProcessLaunchFailForTesting):
(WebKit::WebsiteDataStore::maybeRegisterWithSessionIDMap): Deleted.
(WebKit::WebsiteDataStore::existingNonDefaultDataStoreForSessionID): Deleted.
(WebKit::WebsiteDataStore::processPoolForCookieStorageOperations): Deleted.
(WebKit::WebsiteDataStore::isAssociatedProcessPool const): Deleted.
(WebKit::WebsiteDataStore::pendingCookies const): Deleted.
(WebKit::WebsiteDataStore::addPendingCookie): Deleted.
(WebKit::WebsiteDataStore::removePendingCookie): Deleted.
(WebKit::WebsiteDataStore::clearPendingCookies): Deleted.
(WebKit::WebsiteDataStore::didCreateNetworkProcess): Deleted.
* UIProcess/WebsiteData/WebsiteDataStore.h:
* UIProcess/WebsiteData/curl/WebsiteDataStoreCurl.cpp:
(WebKit::WebsiteDataStore::setNetworkProxySettings):
* UIProcess/WebsiteData/soup/WebsiteDataStoreSoup.cpp:
(WebKit::WebsiteDataStore::platformSetNetworkParameters):
(WebKit::WebsiteDataStore::setPersistentCredentialStorageEnabled):
* UIProcess/glib/WebProcessProxyGLib.cpp:
(WebKit::WebProcessProxy::platformGetLaunchOptions):
* UIProcess/soup/WebCookieManagerProxySoup.cpp:
(WebKit::WebCookieManagerProxy::setCookiePersistentStorage):
* UIProcess/soup/WebProcessPoolSoup.cpp:
(WebKit::WebProcessPool::platformInitializeNetworkProcess):
(WebKit::WebProcessPool::setIgnoreTLSErrors):
(WebKit::WebProcessPool::setNetworkProxySettings):
* WebProcess/Network/NetworkProcessConnectionInfo.h:
2020-09-24 Dean Jackson <dino@apple.com>
WebGL should use GLES in iOS apps running on Apple Silicon
https://bugs.webkit.org/show_bug.cgi?id=216722
<rdar://problem/68976337>
Reviewed by Tim Horton.
* Configurations/WebKit.xcconfig: Remove OpenGL linking.
* UIProcess/mac/HighPerformanceGPUManager.mm: Renamed to .mm. Use the PAL soft-linked OpenGL API.
2020-09-25 Peng Liu <peng.liu6@apple.com>
Replace the usages of (IPC::Attachment fencePort) with IPC::MachPort
https://bugs.webkit.org/show_bug.cgi?id=207683
Reviewed by Daniel Bates.
No new tests, no functional change.
* UIProcess/Cocoa/VideoFullscreenManagerProxy.mm:
(WebKit::VideoFullscreenManagerProxy::setVideoLayerFrame): Replace the
IPC::Attachment argument with a MachSendRight.
* WebProcess/cocoa/VideoFullscreenManager.h:
* WebProcess/cocoa/VideoFullscreenManager.messages.in:
* WebProcess/cocoa/VideoFullscreenManager.mm:
(WebKit::VideoFullscreenManager::setVideoLayerFrameFenced):
r232451 adds a check for mach port disposition, which is not necessary now.
Because ArgumentCoder<MachSendRight>::decode() provides the check.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::setTopContentInset): Replace the IPC::Attachment
argument with a MachSendRight.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::setTopContentInsetFenced): Ditto.
r232451 adds a check for mach port disposition, which is not necessary now.
Because ArgumentCoder<MachSendRight>::decode() provides the check.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-09-25 Wenson Hsieh <wenson_hsieh@apple.com>
Tweak the names of the GPU process audio/video capture internal feature flags
https://bugs.webkit.org/show_bug.cgi?id=216992
Reviewed by Tim Horton.
Keep all 6 of the GPU process feature flags together by adjusting the human readable names of
`CaptureAudioInGPUProcessEnabled` and `CaptureVideoInGPUProcessEnabled` to start with "GPU Process:", after
<https://trac.webkit.org/r267516>.
* Shared/WebPreferencesInternal.yaml:
2020-09-25 Yusuke Suzuki <ysuzuki@apple.com>
[JSC] Introduce JSC_DECLARE_HOST_FUNCTION / JSC_DEFINE_HOST_FUNCTION to make host function definition easy-to-scanned for JIT-caging
https://bugs.webkit.org/show_bug.cgi?id=216966
Reviewed by Saam Barati.
* WebProcess/Plugins/Netscape/JSNPMethod.cpp:
(WebKit::JSC_DEFINE_HOST_FUNCTION):
(WebKit::callMethod): Deleted.
* WebProcess/Plugins/Netscape/JSNPObject.cpp:
(WebKit::JSC_DEFINE_HOST_FUNCTION):
(WebKit::callNPJSObject): Deleted.
(WebKit::constructWithConstructor): Deleted.
2020-09-25 Antoine Quint <graouts@webkit.org>
Add an experimental feature flag for CSS individual transform properties
https://bugs.webkit.org/show_bug.cgi?id=216271
<rdar://problem/68914539>
Reviewed by Dean Jackson.
* Shared/WebPreferencesExperimental.yaml:
2020-09-25 Per Arne Vollan <pvollan@apple.com>
[macOS] Deny access to diagnosticd in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=216929
<rdar://problem/58496834>
Reviewed by Darin Adler.
After the change-set <https://trac.webkit.org/changeset/267498> landed, an extension to diagnostics
is issued when required, so the unconditional rule can be removed from the WebContent sandbox.
* WebProcess/com.apple.WebProcess.sb.in:
2020-09-25 Youenn Fablet <youenn@apple.com>
Implement default behavior for getUserMedia requests in case where there is no delegate set
https://bugs.webkit.org/show_bug.cgi?id=216968
Reviewed by Eric Carlson.
In case page UIDelegate is not set, make the getUserMedia request do its default action.
In case there is a UIDelegate but no getUserMedia specific delegates, make the getUserMedia request do its default action.
Default action is deny for getDisplayMedia and for non cocoa ports.
For cocoa ports, a prompt is shown.
Manually tested with MiniBrowser and MobileMiniBrowser.
* SourcesCocoa.txt:
* UIProcess/API/APIUIClient.h:
(API::UIClient::decidePolicyForUserMediaPermissionRequest):
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::requestUserMediaAuthorizationForFrame):
* UIProcess/Cocoa/UserMediaPermissionRequestProxy.mm: Renamed from Source/WebKit/UIProcess/Cocoa/WKUserMediaCaptureAccessAlert.mm.
(WebKit::alertMessageText):
(WebKit::UserMediaPermissionRequestProxy::doDefaultAction):
* UIProcess/Cocoa/WKUserMediaCaptureAccessAlert.h: Removed.
* UIProcess/UserMediaPermissionRequestProxy.cpp:
(WebKit::UserMediaPermissionRequestProxy::doDefaultAction):
* UIProcess/UserMediaPermissionRequestProxy.h:
* WebKit.xcodeproj/project.pbxproj:
2020-09-25 Youenn Fablet <youenn@apple.com>
Introduce WKWebViewConfiguration mediaCaptureEnabled
https://bugs.webkit.org/show_bug.cgi?id=216872
Reviewed by Alex Christensen.
Add a new boolean configuration to enable/disable mediaDevices exposure.
Value is off by default.
A future patch will remove the mediaDevices WebPreferences based API, given there is no need
to enable/disable mediaDevices during the lifetime of a page.
Covered by existing tests.
* UIProcess/API/APIPageConfiguration.cpp:
(API::PageConfiguration::copy const):
* UIProcess/API/APIPageConfiguration.h:
(API::PageConfiguration::setMediaDevicesEnabled):
(API::PageConfiguration::mediaDevicesEnabled const):
* UIProcess/API/Cocoa/WKPreferencesPrivate.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _setupPageConfiguration:]):
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration _mediaDevicesEnabled]):
(-[WKWebViewConfiguration _setMediaDevicesEnabled:]):
* UIProcess/API/Cocoa/WKWebViewConfigurationPrivate.h:
2020-09-25 Adrian Perez de Castro <aperez@igalia.com>
Non-unified build fixes, late September 2020 edition
https://bugs.webkit.org/show_bug.cgi?id=216950
Unreviewed build fix.
* NetworkProcess/NetworkSessionCreationParameters.cpp: Arrange guards to include
WebCoreArgumentCoders.h with USE(SOUP).
* UIProcess/WebsiteData/soup/WebsiteDataStoreSoup.cpp: Add missing WebCoreArgumentCoders.h
include.
2020-09-24 Brian Burg <bburg@apple.com>
REGRESSION(iOS WebDriver): action chains with 'Touch' pointerType don't work on Mac
https://bugs.webkit.org/show_bug.cgi?id=216937
<rdar://problem/69149171>
Reviewed by Darin Adler.
Before iOS WebDriver shipped, 'Touch' was unconditionally aliased to 'Mouse.'
Make sure the aliasing happens for macOS so the command doesn't fail frivolously.
* UIProcess/Automation/WebAutomationSession.cpp:
(WebKit::WebAutomationSession::performInteractionSequence):
Alias 'Touch' to 'Mouse' when touch is not available and mouse is.
This is the case for macOS.
2020-09-24 Kate Cheney <katherine_cheney@apple.com>
CrashTracer: com.apple.WebKit.Networking in NetworkSession::firstPartyHostCNAMEDomain() code
https://bugs.webkit.org/show_bug.cgi?id=216934
<rdar://problem/69216768>
Reviewed by Alex Christensen.
Add a check for a valid key.
* NetworkProcess/NetworkSession.cpp:
(WebKit::NetworkSession::firstPartyHostCNAMEDomain):
2020-09-24 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed, fix the build for recent SDKs.
rdar://69394018
* Shared/cf/ArgumentCodersCF.cpp:
(IPC::decode):
2020-09-24 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK][WPE] Move proxy settings from WebKitContext to WebKitWebsiteDataManager
https://bugs.webkit.org/show_bug.cgi?id=216927
Reviewed by Alex Christensen.
Make proxy settings per session instead of global to the network process. The web context API now simply
forwards the request to the website data manager.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/NetworkProcessCreationParameters.cpp:
(WebKit::NetworkProcessCreationParameters::encode const):
(WebKit::NetworkProcessCreationParameters::decode):
* NetworkProcess/NetworkProcessCreationParameters.h:
* NetworkProcess/NetworkSessionCreationParameters.cpp:
(WebKit::NetworkSessionCreationParameters::encode const):
(WebKit::NetworkSessionCreationParameters::decode):
* NetworkProcess/NetworkSessionCreationParameters.h:
* NetworkProcess/soup/NetworkProcessSoup.cpp:
(WebKit::NetworkProcess::platformInitializeNetworkProcess):
(WebKit::NetworkProcess::setNetworkProxySettings):
* NetworkProcess/soup/NetworkSessionSoup.cpp:
(WebKit::NetworkSessionSoup::NetworkSessionSoup):
(WebKit::NetworkSessionSoup::setProxySettings):
* NetworkProcess/soup/NetworkSessionSoup.h:
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/API/glib/WebKitAutomationSession.cpp:
(webkitAutomationSessionCreate):
* UIProcess/API/glib/WebKitWebContext.cpp:
(webkit_web_context_set_network_proxy_settings):
* UIProcess/API/glib/WebKitWebView.cpp:
(webkitWebViewConstructed):
* UIProcess/API/glib/WebKitWebsiteDataManager.cpp:
(webkit_website_data_manager_set_network_proxy_settings):
* UIProcess/API/gtk/WebKitNetworkProxySettings.h:
* UIProcess/API/gtk/WebKitWebContext.h:
* UIProcess/API/gtk/WebKitWebsiteDataManager.h:
* UIProcess/API/gtk/docs/webkit2gtk-4.0-sections.txt:
* UIProcess/API/wpe/WebKitNetworkProxySettings.h:
* UIProcess/API/wpe/WebKitWebContext.h:
* UIProcess/API/wpe/WebKitWebsiteDataManager.h:
* UIProcess/API/wpe/docs/wpe-1.0-sections.txt:
* UIProcess/WebProcessPool.h:
* UIProcess/WebsiteData/WebsiteDataStore.h:
(WebKit::WebsiteDataStore::networkProxySettings const):
* UIProcess/WebsiteData/soup/WebsiteDataStoreSoup.cpp:
(WebKit::WebsiteDataStore::platformSetNetworkParameters):
(WebKit::WebsiteDataStore::setNetworkProxySettings):
* UIProcess/glib/WebProcessPoolGLib.cpp:
(WebKit::WebProcessPool::platformInitializeWebProcess):
* UIProcess/soup/WebProcessPoolSoup.cpp:
(WebKit::WebProcessPool::platformInitializeNetworkProcess):
(WebKit::WebProcessPool::setNetworkProxySettings): Deleted.
2020-09-24 Carlos Garcia Campos <cgarcia@igalia.com>
REGRESSION(r267502): [GTK][WPE] ephemeral WebKitWebsiteDataManager ignores TLS errors by default
https://bugs.webkit.org/show_bug.cgi?id=216919
Reviewed by Adrian Perez de Castro.
Initialize ignoreTLSErrors on website data store for ephemeral sessions.
* UIProcess/API/glib/WebKitWebsiteDataManager.cpp:
(webkitWebsiteDataManagerConstructed):
2020-09-24 Carlos Garcia Campos <cgarcia@igalia.com>
REGRESSION(r267502) [GLIB] Make WTR ignoreTLSErrors
https://bugs.webkit.org/show_bug.cgi?id=216911
Reviewed by Adrian Perez de Castro.
I forgot to initialize ignoreTLSErrors in network session parameters.
* UIProcess/WebsiteData/soup/WebsiteDataStoreSoup.cpp:
(WebKit::WebsiteDataStore::platformSetNetworkParameters):
2020-09-23 Peng Liu <peng.liu6@apple.com>
Add feature flags for GPU Process features
https://bugs.webkit.org/show_bug.cgi?id=216901
Reviewed by Tim Horton.
Add feature flags for the following GPU Process features:
- 2D canvas in GPU process
- WebGL in GPU process
- DOM rendering in GPU Process
Also, clean up some related feature/function names.
* FeatureFlags/WebKit.plist:
Add keys for the GPU process features and sort all keys.
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultUseGPUProcessForCanvasRenderingEnabled):
(WebKit::defaultUseGPUProcessForDOMRenderingEnabled):
(WebKit::defaultUseGPUProcessForMediaEnabled):
(WebKit::defaultUseGPUProcessForWebGLEnabled):
(WebKit::defaultCaptureAudioInGPUProcessEnabled):
(WebKit::defaultCaptureVideoInGPUProcessEnabled):
(WebKit::defaultWebRTCCodecsInGPUProcess):
(WebKit::defaultUseGPUProcessForMedia): Deleted.
(WebKit::defaultRenderCanvasInGPUProcessEnabled): Deleted.
* Shared/WebPreferencesDefaultValues.h:
* Shared/WebPreferencesInternal.yaml:
Add preferences and corresponding default values.
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
Rename useGPUProcessForMedia to useGPUProcessForMediaEnabled.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
Use useGPUProcessForCanvasRenderingEnabled instead of renderCanvasInGPUProcessEnabled.
* WebProcess/GPU/media/RemoteMediaPlayerManager.cpp:
(WebKit::RemoteMediaPlayerManager::updatePreferences):
Use useGPUProcessForMediaEnabled instead of useGPUProcessForMedia.
* WebProcess/InjectedBundle/InjectedBundle.cpp:
(WebKit::InjectedBundle::overrideBoolPreferenceForTestRunner): Ditto.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains): Ditto.
(WebKit::WebPage::updatePreferences): Ditto.
2020-09-23 Chris Dumez <cdumez@apple.com>
Turn off the legacy prefixed WebAudio API
https://bugs.webkit.org/show_bug.cgi?id=216886
Reviewed by Sam Weinig.
Turn off the legacy prefixed WebAudio API, now that the modern unprefixed API has been enabled
in r267488.
* Shared/WebPreferencesInternal.yaml:
2020-09-23 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK][WPE] Move TLS errors API from WebKitContext to WebKitWebsiteDataManager
https://bugs.webkit.org/show_bug.cgi?id=216822
Reviewed by Alex Christensen.
Make TLS errors policy per session instead of global to the network process. The web context API now simply
forwards the request to the website data manager. The website data manager created for an ephemeral web view
will inherit the network settings from the web context website data manager. This ensures backwards
compatibility, but it's also convenient, since it's the desired behavior in most of the cases. Apps can still
change the network settings of the web view by getting the website data manager.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/NetworkProcessCreationParameters.cpp:
(WebKit::NetworkProcessCreationParameters::encode const):
(WebKit::NetworkProcessCreationParameters::decode):
* NetworkProcess/NetworkProcessCreationParameters.h:
* NetworkProcess/NetworkSessionCreationParameters.cpp:
(WebKit::NetworkSessionCreationParameters::encode const):
(WebKit::NetworkSessionCreationParameters::decode):
* NetworkProcess/NetworkSessionCreationParameters.h:
* NetworkProcess/soup/NetworkDataTaskSoup.cpp:
(WebKit::NetworkDataTaskSoup::tlsConnectionAcceptCertificate):
* NetworkProcess/soup/NetworkProcessSoup.cpp:
(WebKit::NetworkProcess::platformInitializeNetworkProcess):
(WebKit::NetworkProcess::setIgnoreTLSErrors):
* NetworkProcess/soup/NetworkSessionSoup.cpp:
(WebKit::NetworkSessionSoup::NetworkSessionSoup):
(WebKit::webSocketAcceptCertificateCallback):
(WebKit::webSocketMessageNetworkEventCallback):
(WebKit::NetworkSessionSoup::createWebSocketTask):
(WebKit::NetworkSessionSoup::setIgnoreTLSErrors):
* NetworkProcess/soup/NetworkSessionSoup.h:
* UIProcess/API/glib/WebKitAutomationSession.cpp:
(webkitAutomationSessionCreate):
* UIProcess/API/glib/WebKitWebContext.cpp:
(webkitWebContextConstructed):
(webkit_web_context_set_tls_errors_policy):
(webkit_web_context_get_tls_errors_policy):
* UIProcess/API/glib/WebKitWebView.cpp:
(webkitWebViewConstructed):
(webkit_web_view_class_init):
(webkitWebViewLoadFailedWithTLSErrors):
* UIProcess/API/glib/WebKitWebsiteDataManager.cpp:
(webkitWebsiteDataManagerConstructed):
(webkitWebsiteDataManagerGetDataStore):
(webkit_website_data_manager_set_tls_errors_policy):
(webkit_website_data_manager_get_tls_errors_policy):
* UIProcess/API/gtk/WebKitWebContext.h:
* UIProcess/API/gtk/WebKitWebsiteDataManager.h:
* UIProcess/API/gtk/docs/webkit2gtk-4.0-sections.txt:
* UIProcess/API/wpe/WebKitWebContext.h:
* UIProcess/API/wpe/WebKitWebsiteDataManager.h:
* UIProcess/API/wpe/docs/wpe-1.0-sections.txt:
* UIProcess/WebProcessPool.h:
* UIProcess/WebsiteData/WebsiteDataStore.h:
(WebKit::WebsiteDataStore::ignoreTLSErrors const):
* UIProcess/WebsiteData/soup/WebsiteDataStoreSoup.cpp:
(WebKit::WebsiteDataStore::setIgnoreTLSErrors):
* UIProcess/soup/WebProcessPoolSoup.cpp:
(WebKit::WebProcessPool::platformInitializeNetworkProcess):
(WebKit::WebProcessPool::setIgnoreTLSErrors): Deleted.
2020-09-23 Per Arne Vollan <pvollan@apple.com>
[macOS] Issue sandbox extension for the diagnostic daemon
https://bugs.webkit.org/show_bug.cgi?id=216834
<rdar://problem/58496834>
Reviewed by Brent Fulgham.
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::isInternalInstall):
(WebKit::agxCompilerClasses):
(WebKit::diagnosticServices):
(WebKit::WebProcessPool::platformInitializeWebProcess):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
* WebProcess/com.apple.WebProcess.sb.in:
2020-09-23 Fujii Hironori <Hironori.Fujii@sony.com>
Remove ENABLE_SVG_FONTS macro
https://bugs.webkit.org/show_bug.cgi?id=216850
Reviewed by Don Olmstead.
* WebProcess/Network/WebLoaderStrategy.cpp:
(WebKit::maximumBufferingTime):
2020-09-23 Per Arne Vollan <pvollan@apple.com>
[macOS] Change name of client decoder entitlement
https://bugs.webkit.org/show_bug.cgi?id=216670
<rdar://problem/69101997>
Reviewed by Darin Adler.
The client decoder entitlement landed in <https://trac.webkit.org/changeset/266613/webkit> has changed name, and should be updated.
* Scripts/process-entitlements.sh:
2020-09-23 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process] Implement caption support in video fullscreen and PiP
https://bugs.webkit.org/show_bug.cgi?id=208752
Reviewed by Eric Carlson.
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
(WebKit::RemoteMediaPlayerProxy::syncTextTrackBounds): Deleted.
* GPUProcess/media/RemoteMediaPlayerProxy.h:
* GPUProcess/media/RemoteMediaPlayerProxy.messages.in:
Remove syncTextTrackBounds() because the GPU process does not need to deal with
the text track bounds.
* WebProcess/GPU/GPUProcessConnection.cpp:
(WebKit::GPUProcessConnection::dispatchMessage):
Add a missing IPC message receiver.
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::setVideoFullscreenLayer): Set the flag
m_requiresTextTrackRepresentation when changing video presentation mode.
(WebKit::MediaPlayerPrivateRemote::requiresTextTrackRepresentation const): Deleted.
(WebKit::MediaPlayerPrivateRemote::setTextTrackRepresentation): Deleted.
(WebKit::MediaPlayerPrivateRemote::syncTextTrackBounds): Deleted.
Move the implementation of those functions to MediaPlayerPrivateRemoteCocoa.
* WebProcess/GPU/media/MediaPlayerPrivateRemote.h:
Add the flag m_requiresTextTrackRepresentation and add a reference
to the layer used to render text track (in fullscreen or picture-in-picture).
* WebProcess/GPU/media/cocoa/MediaPlayerPrivateRemoteCocoa.mm:
(WebKit::MediaPlayerPrivateRemote::setVideoFullscreenFrame):
(WebKit::MediaPlayerPrivateRemote::setTextTrackRepresentation):
(WebKit::MediaPlayerPrivateRemote::syncTextTrackBounds):
Those functions share a similar implementation as VideoLayerManagerObjC.
2020-09-23 Per Arne Vollan <pvollan@apple.com>
[macOS] Start collecting telemetry from a set of XPC services
<https://bugs.webkit.org/show_bug.cgi?id=216797>
Reviewed by Brent Fulgham.
* WebProcess/com.apple.WebProcess.sb.in:
2020-09-23 Chris Dumez <cdumez@apple.com>
Turn on the modern unprefixed WebAudio API
https://bugs.webkit.org/show_bug.cgi?id=216885
Reviewed by Eric Carlson.
Turn on the modern unprefixed WebAudio API now that we support most of the API
besides AudioWorklet and now that we have a good pass rate on WPT tests.
* Shared/WebPreferencesExperimental.yaml:
2020-09-23 Truitt Savell <tsavell@apple.com>
Unreviewed, reverting r267403.
Caused 50+ Crashes on iOS Debug
Reverted changeset:
"Move Media in GPU Process flag from internal to experimental"
https://bugs.webkit.org/show_bug.cgi?id=216473
https://trac.webkit.org/changeset/267403
2020-09-23 Youenn Fablet <youenn@apple.com>
Add support for HTMLMediaElement.setSinkId
https://bugs.webkit.org/show_bug.cgi?id=216696
Reviewed by Eric Carlson.
Add internal flag to enable/disable user gesture requirement for setting audio output device,
and for per media element setting of audio output device.
* Shared/WebPreferencesExperimental.yaml:
* Shared/WebPreferencesInternals.yaml:
2020-09-22 Wenson Hsieh <wenson_hsieh@apple.com>
[GPU Process] fast/canvas/canvas-blend-image.html and fast/canvas/canvas-blend-solid.html fail on macOS
https://bugs.webkit.org/show_bug.cgi?id=216860
Reviewed by Said Abou-Hallawa.
These two tests disable accelerated drawing for canvas elements; when using the GPU process, this causes us to
use `ImageBufferShareableBitmapBackend` as the back-end for the image buffer backing the canvas element.
On both macOS and iOS, the image buffer is backed by BGRA image data. However, we override
`backendColorFormat()` and return `ColorFormat::BGRA` only for `PLATFORM(IOS_FAMILY)`, causing `putImageData`
and `getImageData` to result in flipped red and blue channels on macOS, which subsequently causes these two
layout tests to fail.
Fix these tests by removing the `PLATFORM(IOS_FAMILY)` guard.
* WebProcess/GPU/graphics/ImageBufferShareableBitmapBackend.h:
2020-09-22 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r267421.
https://bugs.webkit.org/show_bug.cgi?id=216864
Broke a test on macOS
Reverted changeset:
"[macOS] Start collecting telemetry from a set of XPC
services"
https://bugs.webkit.org/show_bug.cgi?id=216797
https://trac.webkit.org/changeset/267421
2020-09-22 Lauro Moura <lmoura@igalia.com>
[GTK] REGRESSION(r267398): several hidpi tests are failing
https://bugs.webkit.org/show_bug.cgi?id=216830
Reviewed by Adrian Perez de Castro.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::setCustomDeviceScaleFactor): Remove check for removed define.
2020-09-22 Peng Liu <peng.liu6@apple.com>
[Media in GPU Process] Add missing IPC message receivers
https://bugs.webkit.org/show_bug.cgi?id=216829
Reviewed by Jer Noble.
Add missing IPC message receivers and clean up some macro usages.
Also remove the empty RemoteMediaPlayerManager.messages.in.
No new tests, no functional change.
* CMakeLists.txt:
* DerivedSources-input.xcfilelist:
* DerivedSources.make:
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::dispatchMessage):
(WebKit::GPUConnectionToWebProcess::dispatchSyncMessage):
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/media/ios/RemoteMediaSessionHelperProxy.h:
(WebKit::RemoteMediaSessionHelperProxy::didReceiveMessageFromWebProcess):
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/GPU/GPUProcessConnection.cpp:
(WebKit::GPUProcessConnection::dispatchMessage):
* WebProcess/GPU/media/RemoteMediaPlayerManager.messages.in: Removed.
2020-09-22 Per Arne Vollan <pvollan@apple.com>
[macOS] Start collecting telemetry from a set of XPC services
<https://bugs.webkit.org/show_bug.cgi?id=216797>
Reviewed by Brent Fulgham.
* WebProcess/com.apple.WebProcess.sb.in:
2020-09-22 Youenn Fablet <youenn@apple.com>
Implement a default prompt for getUserMedia
https://bugs.webkit.org/show_bug.cgi?id=216821
Reviewed by Eric Carlson.
Move preferences used by WebRTC develop menu to internals.
This allows having them in MiniBrowser.
Enable video capture in UIProcess by default for non Safari applications.
Add support for a getUserMedia prompt, very similar to iOS current prompt.
Use the prompt in case the application does not implement any of the two delegates.
* Shared/WebPreferences.yaml:
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultCaptureVideoInUIProcessEnabled):
* Shared/WebPreferencesDefaultValues.h:
* Shared/WebPreferencesExperimental.yaml:
* Shared/WebPreferencesInternal.yaml:
* UIProcess/Cocoa/UIDelegate.h:
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::setDelegate):
(WebKit::requestUserMediaAuthorizationForFrame):
(WebKit::UIDelegate::UIClient::decidePolicyForUserMediaPermissionRequest):
* UIProcess/Cocoa/WKUserMediaCaptureAccessAlert.h: Added.
* UIProcess/Cocoa/WKUserMediaCaptureAccessAlert.mm: Added.
(WebKit::visibleDomain):
(WebKit::alertMessageText):
(WebKit::presentUserMediaCaptureAccessAlert):
* WebKit.xcodeproj/project.pbxproj:
2020-09-22 Brady Eidson <beidson@apple.com>
Crashtracer inside PDFPlugin::createScrollbar.
<rdar://problem/69256031> and https://bugs.webkit.org/show_bug.cgi?id=216810
Reviewed by Tim Horton.
To quote Tim from r264945:
No new tests; timing is such that I can't reproduce without inserting
intentional delays into the main thread hops, which is further than
I'm willing to go for a test.
This is a speculative fix due to the aforementioned reproducibility issue.
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::installPDFDocument): With all the past fixes in place, its apparent
the plug-in HAS been torn down, and it's somewhat common to bypass the other "hasBeenDestroyed"
checks. So put an explicit check here followed by an explicit release assert.
2020-09-22 Brian Burg <bburg@apple.com>
[Cocoa] _WKInspectorDelegate should handle showing external resources
https://bugs.webkit.org/show_bug.cgi?id=216334
Reviewed by Devin Rousso.
WebKit clients need a way to control the loading and presentation of
external resources that are linked in Web Inspector's user interface.
Rename InspectorFrontendHost.openInNewTab to openURLExternally. Change
the implementation to forward the request to UIProcess rather than the
inspected WebProcess.
When a navigation is triggered in WKInspectorViewController's WKWebView,
allow the delegate to open the requested resource. Otherwise, redirect
the navigation to the inspected WebView or open it using NSWorkspace
(for the remote case).
New API test: WKInspectorDelegate.OpenURLExternally.
* UIProcess/API/Cocoa/_WKInspectorPrivateForTesting.h: Added.
* UIProcess/API/Cocoa/_WKInspectorTesting.mm: Added.
(snippetToOpenURLExternally):
(-[_WKInspector _openURLExternallyForTesting:useFrontendAPI:]):
Add some helpers for writing API tests.
* UIProcess/API/APIInspectorClient.h:
(API::InspectorClient::openURLExternally):
* UIProcess/API/Cocoa/_WKInspectorDelegate.h:
* UIProcess/Cocoa/PageClientImplCocoa.mm:
* UIProcess/Inspector/Cocoa/InspectorDelegate.h:
* UIProcess/Inspector/Cocoa/InspectorDelegate.mm:
(WebKit::InspectorDelegate::setDelegate):
(WebKit::InspectorDelegate::InspectorClient::openURLExternally):
Add new delegate method to _WKInspectorDelegate.
* UIProcess/Inspector/RemoteWebInspectorProxy.messages.in:
* UIProcess/Inspector/RemoteWebInspectorProxy.h:
* UIProcess/Inspector/RemoteWebInspectorProxy.cpp:
(WebKit::RemoteWebInspectorProxy::openURLExternally):
(WebKit::RemoteWebInspectorProxy::platformOpenURLExternally):
(WebKit::RemoteWebInspectorProxy::openInNewTab): Deleted.
(WebKit::RemoteWebInspectorProxy::platformOpenInNewTab): Deleted.
* UIProcess/Inspector/gtk/RemoteWebInspectorProxyGtk.cpp:
(WebKit::RemoteWebInspectorProxy::platformURLExternally):
(WebKit::RemoteWebInspectorProxy::platformOpenInNewTab): Deleted.
* UIProcess/Inspector/mac/RemoteWebInspectorProxyMac.mm:
(WebKit::RemoteWebInspectorProxy::platformOpenURLExternally):
(WebKit::RemoteWebInspectorProxy::platformOpenInNewTab): Deleted.
* UIProcess/Inspector/win/RemoteWebInspectorProxyWin.cpp:
(WebKit::RemoteWebInspectorProxy::platformOpenURLExternally):
(WebKit::RemoteWebInspectorProxy::platformOpenInNewTab): Deleted.
Rename openInNewTab to openURLExternally. In some cases the client
may not choose to present a new tab, so the name is no longer accurate.
* UIProcess/Inspector/WebInspectorProxy.messages.in:
* UIProcess/Inspector/WebInspectorProxy.h:
* UIProcess/Inspector/WebInspectorProxy.cpp:
(WebKit::WebInspectorProxy::inspectorWindow const):
(WebKit::WebInspectorProxy::openURLExternally):
(WebKit::WebInspectorProxy::evaluateInFrontendForTesting):
* UIProcess/Inspector/mac/WebInspectorProxyMac.mm:
(-[WKWebInspectorProxyObjCAdapter inspectorViewController:openURLExternally:]):
Add some helpers for writing API tests. Add an IPC message
for evaluating a JavaScript expression in the frontend page.
Also, add a message receiver for the OpenURLExternally message
which passes the request to the API layer.
* UIProcess/Inspector/mac/WKInspectorViewController.h:
* UIProcess/Inspector/mac/WKInspectorViewController.mm:
(-[WKInspectorViewController webView:decidePolicyForNavigationAction:decisionHandler:]):
Call the delegate method if a navigation inside the inspector WKWebView is an external URL.
Specifically the request is to load something other than the Web Inspector's main HTML page.
* WebProcess/Inspector/WebInspector.cpp:
(WebKit::WebInspector::openInNewTab): Deleted.
* WebProcess/Inspector/WebInspector.h:
* WebProcess/Inspector/WebInspector.messages.in:
No need to handle this message anymore in the inspected WebProcess.
* WebProcess/Inspector/RemoteWebInspectorUI.h:
* WebProcess/Inspector/RemoteWebInspectorUI.cpp:
(WebKit::RemoteWebInspectorUI::openURLExternally):
(WebKit::RemoteWebInspectorUI::openInNewTab): Deleted.
* WebProcess/Inspector/WebInspectorUI.messages.in:
* WebProcess/Inspector/WebInspectorUI.h:
* WebProcess/Inspector/WebInspectorUI.cpp:
(WebKit::WebInspectorUI::updateConnection):
(WebKit::WebInspectorUI::closeWindow):
(WebKit::WebInspectorUI::openURLExternally):
(WebKit::WebInspectorUI::evaluateInFrontendForTesting):
(WebKit::WebInspectorUI::openInNewTab): Deleted.
When we need to open the URL externally, send an IPC message to UIProcess.
The old implementation triggered the navigation from the inspected WebProcess.
That approach is no longer desirable. Remove the unneeded IPC connection.
* WebProcess/Inspector/WebInspectorFrontendAPIDispatcher.h:
* WebProcess/Inspector/WebInspectorFrontendAPIDispatcher.cpp:
(WebKit::WebInspectorFrontendAPIDispatcher::evaluateExpressionForTesting):
Fulfill the request to evaluate the expression in the frontend page.
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
Some files were added.
2020-09-22 Per Arne Vollan <pvollan@apple.com>
[macOS] Start collecting telemetry from message filtering
<https://bugs.webkit.org/show_bug.cgi?id=216803>
Reviewed by Darin Adler.
Start collecting telemetry for message filtering in the WebContent sandbox on macOS.
* WebProcess/com.apple.WebProcess.sb.in:
2020-09-22 Youenn Fablet <youenn@apple.com>
Move Media in GPU Process flag from internal to experimental
https://bugs.webkit.org/show_bug.cgi?id=216473
<rdar://problem/69318786>
Reviewed by Eric Carlson.
No change of behavior.
* Shared/WebPreferencesExperimental.yaml:
* Shared/WebPreferencesInternal.yaml:
2020-09-21 Youenn Fablet <youenn@apple.com>
Enable MediaRecorder by default on iOS
https://bugs.webkit.org/show_bug.cgi?id=216664
Reviewed by Darin Adler.
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultMediaRecorderEnabled):
2020-09-21 Adrian Perez de Castro <aperez@igalia.com>
[CMake] Use imported targets in find module for the ATK SPI2 bridge dependency
https://bugs.webkit.org/show_bug.cgi?id=216773
Reviewed by Don Olmstead.
No new tests needed.
* PlatformWPE.cmake: Use the ATK::Bridge imported target.
2020-09-21 Adrian Perez de Castro <aperez@igalia.com>
[GTK] Bump cairo version to support HiDPI
https://bugs.webkit.org/show_bug.cgi?id=133378
Reviewed by Carlos Garcia Campos.
Remove conditional compilation around cairo_{g,s}et_device_scale() as there is no need to
support older versions of Cairo which lack the functions. The minimum version of Cairo
being required by CMake is 1.14.0, which already includes them.
* Shared/cairo/ShareableBitmapCairo.cpp:
(WebKit::ShareableBitmap::paint): use cairo_surface_set_device_scale() directly.
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseTakeViewSnapshot): Ditto.
* UIProcess/gtk/AcceleratedBackingStoreWayland.cpp:
(WebKit::AcceleratedBackingStoreWayland::displayBuffer): Ditto.
(WebKit::AcceleratedBackingStoreWayland::downloadTexture): Ditto.
* UIProcess/gtk/AcceleratedBackingStoreX11.cpp:
(WebKit::AcceleratedBackingStoreX11::update): Ditto.
2020-09-21 Tim Horton <timothy_horton@apple.com>
paper.io ad close buttons cannot be iteracted with via trackpad on iPad
https://bugs.webkit.org/show_bug.cgi?id=216812
<rdar://problem/68738585>
Reviewed by Wenson Hsieh.
* UIProcess/ios/WKContentViewInteraction.mm:
(applicationIsKnownToIgnoreMouseEvents):
2020-09-21 Timothy Hatcher <timothy@apple.com>
Add some WKWebProcessPlugInFrame SPI needed by Safari.
https://bugs.webkit.org/show_bug.cgi?id=216809
rdar://problem/69336928
Reviewed by Brian Weinstein.
* WebProcess/InjectedBundle/API/Cocoa/WKWebProcessPlugInFrame.mm:
(+[WKWebProcessPlugInFrame lookUpFrameFromJSContext:]): Added.
(-[WKWebProcessPlugInFrame _securityOrigin]): Added.
* WebProcess/InjectedBundle/API/Cocoa/WKWebProcessPlugInFramePrivate.h:
2020-09-21 Andres Gonzalez <andresg_22@apple.com>
Make accessibility isolated tree a non-experimental feature.
https://bugs.webkit.org/show_bug.cgi?id=216792
Reviewed by Chris Fleizach.
Promote accessibility isolated tree from an experimental feature to a
non-experimental one. This avoids the problem where a user can turn off
the feature through the Safari > Develop > Experimental Features submenu,
while the accessibility request are still coming on a secondary thread,
which leads to random crashes.
* Shared/WebPreferences.yaml:
* Shared/WebPreferencesExperimental.yaml:
2020-09-21 Basuke Suzuki <basuke.suzuki@sony.com>
[WinCairo][PlayStation] Support different instances of listener client.
https://bugs.webkit.org/show_bug.cgi?id=216733
Reviewed by Don Olmstead.
Follows the change of RemoteInspectorSocketEndpoint::Client interface change.
No new tests because there's no behaivior change.
* UIProcess/Inspector/socket/RemoteInspectorClient.cpp:
(WebKit::RemoteInspectorClient::didClose):
* UIProcess/Inspector/socket/RemoteInspectorClient.h:
2020-09-21 Youenn Fablet <youenn@apple.com>
[iOS] MediaRecorder incorrect screen orientation handling
https://bugs.webkit.org/show_bug.cgi?id=198912
<rdar://problem/51802521>
Reviewed by Eric Carlson.
* GPUProcess/webrtc/RemoteMediaRecorder.cpp:
(WebKit::RemoteMediaRecorder::videoSampleAvailable):
2020-09-21 Keith Miller <keith_miller@apple.com>
Functions should consistently enumerate length before name
https://bugs.webkit.org/show_bug.cgi?id=216789
Reviewed by Yusuke Suzuki.
JSNPMethod inherits from InternalFunction which now sets length by default.
I chose 0 because it looks like this function doesn't directly look at any
arguments and I doubt anyone is looking at the length anyway.
* WebProcess/Plugins/Netscape/JSNPMethod.cpp:
(WebKit::JSNPMethod::finishCreation):
2020-09-20 Darin Adler <darin@apple.com>
Next step towards eliminating deprecated legacy positions
https://bugs.webkit.org/show_bug.cgi?id=216751
Reviewed by Sam Weinig.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::selectWithGesture):
(WebKit::WebPage::extendSelection):
(WebKit::WebPage::updateSelectionWithExtentPointAndBoundary):
(WebKit::WebPage::requestAutocorrectionData):
(WebKit::WebPage::applyAutocorrectionInternal):
(WebKit::focusedElementPositionInformation):
(WebKit::dataDetectorLinkPositionInformation):
(WebKit::WebPage::requestDocumentEditingContext):
Replaced calls to createLegacyEditingPosition with calls to
makeDeprecatedLegacyPosition or makeContainerOffsetPosition.
2020-09-21 Wenson Hsieh <wenson_hsieh@apple.com>
Adjust the human readable names some GPU process internal feature flags
https://bugs.webkit.org/show_bug.cgi?id=216794
Reviewed by Darin Adler.
Change the human readable names so that all the GPU-process-specific internal features start are of the form
"Use GPU Process…".
* Shared/WebPreferencesInternal.yaml:
2020-09-21 James Savage <james.savage@apple.com>
WKWebView Swift overlay has mis-annotated nullability for evaluateJavaScript
<http://webkit.org/b/216198>
<rdar://problem/68035950>
Reviewed by Darin Adler.
Due to a mistranslation of evaluateJavaScript, we are vending an API which does not expect
to receive nil as a valid result value. This change fixes the crash, but does not yet fix
the API to have the correct signature. That will come in a later patch.
To fix the crash, we need to produce a valid Result<Any, Error> to pass back to clients.
Fortunately, instead of inventing something clever, we can just use nil. It's valid to box
optional values into Any, and clients can technically retrieve them with the right dynamic cast
as well. Since client code must be using dynamic casting to convert the result Any to a usable
type, and because in the case where we now return a new value at runtime we would have previously
crashed, this shouldn't have any binary compatibility impact either.
To better validate these changes, I also add new unit tests for the conversion of JavaScript
results into Swift values, including a test for the deprecated API.
* SwiftOverlay/SwiftOverlay/ObjectiveCBlockConversions.swift: Added. For clarity I'm factoring
helper methods into a single namespace, as it also makes a nice place to document their expectations.
(ObjectiveCBlockConversion.exclusive.exclusive(_:)): This is renamed from the free function,
makeResultHandler(_:). It still has the same fatalError (now precondition) as before, but
hopefully a better name to clarify that it expects exactly-one value.
(ObjectiveCBlockConversions.boxingNilAsAnyForCompatibility(_:)): This is a variant of exclusive(_:)
that makes the tradeoff of boxing any nil values as Any to avoid crashing. This is still safe,
since as mentioned our clients will need to cast the value they recieve to do anything with it,
and since the deprecated API expects `Any`, no one could have been successfully comparing it
to `nil` today anyways.
* SwiftOverlay/Tests/JavaScriptToSwiftTypeConversions.swift: Added.
(JavaScriptToSwiftConversions.setUp): Construct a new web view, and add it to a window so that it is
in an expected state. I'm using about:blank as the URL, since page content doesn't matter for
these tests and I want the web content to be ready immediately.
(JavaScriptToSwiftConversions.tearDown): Just perform some window cleanup.
(JavaScriptToSwiftConversions.evaluateJavaScript(_:andExpect:)): Helper method to evaluate script and
check its result. I'm using String.debugDescription because it escapes quotes and special characters
which makes the readout easier to parse.
(JavaScriptToSwiftConversions.testNull): JavaScript's null is actually mapped to NSNull, not nil.
(JavaScriptToSwiftConversions.testInteger): Some standard type coercion tests. The underlying value for
all number types should be NSNumber, so this is actually check against a float or integer type
without issue.
(JavaScriptToSwiftConversions.testDecimal): Ditto.
(JavaScriptToSwiftConversions.testBoolean): Ditto.
(JavaScriptToSwiftConversions.testString): Ditto.
(JavaScriptToSwiftConversions.testArray): Ditto.
(JavaScriptToSwiftConversions.testDictionary): Ditto, only you can't evaluate an object literal directly
so I need to store it in a temporary location first.
(JavaScriptToSwiftConversions.testUndefined): Test our boxing of nil. The exact value matters less than
not crashing at all.
* SwiftOverlay/WebKitSwiftOverlay.xcodeproj/project.pbxproj: Added new files to project. I kept the
new test file and helper files within the SwiftOverlay group, because they do not contribute any API.
* UIProcess/API/Cocoa/WebKitSwiftOverlay.swift:
(WKWebView.callAsyncJavaScript(_:arguments:in:in:completionHandler:)): Switch to a conversion which does
not trap on nil.
(WKWebView.createPDF(_:completionHandler:)): Updated to use new helper method name.
(WKWebView.createWebArchiveData(_:)): Ditto.
(WKWebView.evaluateJavaScript(_:in:in:completionHandler:Error:)): See above.
(makeResultHandler(_:): This has been subsumed by ObjCBlockConversions.
2020-09-21 Kate Cheney <katherine_cheney@apple.com>
Allow about:blank subframe loads for non app-bound top frames
https://bugs.webkit.org/show_bug.cgi?id=216715
<rdar://problem/69070720>
Reviewed by Darin Adler.
In the original implementation of App Bound Domains, any loads using
certain protocols like about: were automatically considered app-bound.
When a subframe loads about:blank for example, this forces web views
into app-bound mode even if the top frame is not app-bound, which
should not be the case.
These special protocols inherit security origins from their top-frame
loads, so we can refer to those to determine whether subframe loads
should be considered app-bound. For about: loads in a main frame there
is no behavior change and WKWebViews will still automatically be considered
app-bound in this case.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::setIsNavigatingToAppBoundDomainAndCheckIfPermitted):
(WebKit::WebPageProxy::decidePolicyForNavigationAction):
Refer to the security origin if the subframe load is one of the
special protocols caught by the shouldTreatURLProtocolAsAppBound() function.
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::schemeOrDomainIsAppBound):
(WebKit::WebsiteDataStore::beginAppBoundDomainCheck):
* UIProcess/WebsiteData/WebsiteDataStore.h:
2020-09-20 Simon Fraser <simon.fraser@apple.com>
Sort the project file.
* WebKit.xcodeproj/project.pbxproj:
2020-09-19 Sam Weinig <weinig@apple.com>
Remove runtime setting for enabling/disabling the CrossOriginResourcePolicy API
https://bugs.webkit.org/show_bug.cgi?id=216561
Reviewed by Darin Adler.
Remove runtime setting for enabling/disabling CrossOriginResourcePolicy. It has been
enabled by default for a while and there are no known reasons one would need to
disable it.
* Shared/WebPreferences.yaml:
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetCrossOriginResourcePolicyEnabled):
(WKPreferencesGetCrossOriginResourcePolicyEnabled):
* UIProcess/API/C/WKPreferencesRef.h:
* WebProcess/Network/WebLoaderStrategy.cpp:
(WebKit::WebLoaderStrategy::scheduleLoadFromNetworkProcess):
2020-09-18 Wenson Hsieh <wenson_hsieh@apple.com>
[macOS] REGRESSION (r265702): System Services receive 0 bytes when extracting selected content as rich text data
https://bugs.webkit.org/show_bug.cgi?id=216718
<rdar://problem/69150358>
Reviewed by Tim Horton.
Test: CopyHTML.WriteRichTextSelectionToPasteboard
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::dataSelectionForPasteboard):
After r265702, the `IPCHandle` received in the UI process was always being converted into a buffer of size 0,
due to using the `size` local variable (which is no longer set as an outparam of the sync IPC message). Instead,
use `ipcHandle.dataSize`.
2020-09-18 Megan Gardner <megan_gardner@apple.com>
Remove unneeded respondsToSelector staging for keyboard.
https://bugs.webkit.org/show_bug.cgi?id=216421
Staging needed for dual work in UIKit in 2019, no longer needed.
Reviewed by Tim Horton.
No behavior change, covered by existing tests.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView insertText:]):
(-[WKContentView handleKeyWebEvent:withCompletionHandler:]):
(-[WKContentView _interpretKeyEvent:isCharEvent:]):
2020-09-18 Sihui Liu <sihui_liu@apple.com>
REGRESSION (r266634): fast/selectors/text-field-selection-stroke-color.html and fast/selectors/text-field-selection-window-inactive-stroke-color.html are flaky failures with pixel noise
https://bugs.webkit.org/show_bug.cgi?id=216394
<rdar://problem/68679551>
Reviewed by Tim Horton.
In WebKitTestRunner, we reset the states and load about:blank page between tests. The resetting may cause some
activity state changes of web page, like WindowIsActive. Before r266634, these changes will be dispatched
quickly enough (before runloop waits or when runloop exits) to web process, so web process can do rendering with
updated activity state. After r266634, dispatch of the changes may be delayed (until CATransaction commits).
Page activity state can affect scrollability (in FrameView::computeScrollability()) and scrollability can affect
tiling size (in TileController::computeTileSize()). Different tilings may cause the tiny pixel difference we see
in the test results.
To fix this, create an SPI for sending out activity state updates in next runloop cycle and adopt it in WTR.
* UIProcess/API/C/WKPage.cpp:
(WKPageDispatchActivityStateUpdateForTesting):
* UIProcess/API/C/WKPagePrivate.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::dispatchActivityStateUpdateForTesting):
* UIProcess/WebPageProxy.h:
2020-09-18 Alex Christensen <achristensen@webkit.org>
WKWebView._tryClose should return whether it succeeded or not
https://bugs.webkit.org/show_bug.cgi?id=216660
Reviewed by Tim Horton.
The return value of WKPageTryClose is used by Safari.
In order to migrate to WKWebView _tryClose, we need to expose this return value.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _tryClose]):
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
2020-09-18 Sihui Liu <sihui_liu@apple.com>
Webpages flash when getting closed
https://bugs.webkit.org/show_bug.cgi?id=216131
<rdar://problem/62264106>
Reviewed by Chris Dumez.
Closing a page in web process will detach root CA layer and clears content in view immediately. If this happens
in web process before transaction, which contains view hierachy change or layer change, is committed in UI
process, we would see a white flash. To try fixing this issue, we explicitly delay sending the close message
to next runloop cycle.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::close):
(WebKit::WebPageProxy::receivedNavigationPolicyDecision):
* UIProcess/WebProcessProxy.cpp:
(WebKit::m_shutdownPreventingScopeCounter):
(WebKit::WebProcessProxy::canTerminateAuxiliaryProcess):
* UIProcess/WebProcessProxy.h: remove class ScopePreventingShutdown and use RefCounter to prevent process
shutdown.
(WebKit::WebProcessProxy::shutdownPreventingScope):
(WebKit::WebProcessProxy::ScopePreventingShutdown::ScopePreventingShutdown): Deleted.
(WebKit::WebProcessProxy::ScopePreventingShutdown::~ScopePreventingShutdown): Deleted.
(WebKit::WebProcessProxy::makeScopePreventingShutdown): Deleted.
2020-09-18 Youenn Fablet <youenn@apple.com>
Add internal flag to enable/disable H264 hardware encoder
https://bugs.webkit.org/show_bug.cgi?id=216534
Reviewed by Eric Carlson.
* Shared/WebPreferencesInternal.yaml:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences):
2020-09-18 Youenn Fablet <youenn@apple.com>
Hash deviceIds in WebProcess instead of UIProcess to allow audio output device IDs in WebProcess
https://bugs.webkit.org/show_bug.cgi?id=216693
Reviewed by Eric Carlson.
* UIProcess/UserMediaPermissionRequestManagerProxy.cpp:
(WebKit::UserMediaPermissionRequestManagerProxy::computeFilteredDeviceList):
(WebKit::UserMediaPermissionRequestManagerProxy::enumerateMediaDevicesForFrame):
* UIProcess/UserMediaPermissionRequestManagerProxy.h:
2020-09-18 Youenn Fablet <youenn@apple.com>
Unified build fixes in media code
https://bugs.webkit.org/show_bug.cgi?id=216691
Reviewed by Sam Weinig.
Add missing includes and/or missing undef.
No change of behavior.
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRenderer.cpp:
* GPUProcess/webrtc/RemoteMediaRecorder.cpp:
* WebProcess/WebPage/WebPageOverlay.h:
* WebProcess/cocoa/RemoteRealtimeMediaSource.cpp:
* WebProcess/cocoa/RemoteRealtimeMediaSource.h:
2020-09-18 Noam Rosenthal <noam@webkit.org>
Enable paint timing by default
https://bugs.webkit.org/show_bug.cgi?id=211736
Reviewed by Simon Fraser.
* Shared/WebPreferencesExperimental.yaml:
PaintTimingEnabled default to true.
2020-09-18 Youenn Fablet <youenn@apple.com>
XHR.timeout is affected by long tasks
https://bugs.webkit.org/show_bug.cgi?id=216266
<rdar://problem/68908150>
Reviewed by Alex Christensen.
Go to network process to know whether a load is finished or not.
* NetworkProcess/NetworkConnectionToWebProcess.cpp:
(WebKit::NetworkConnectionToWebProcess::isResourceLoadFinished):
* NetworkProcess/NetworkConnectionToWebProcess.h:
* NetworkProcess/NetworkConnectionToWebProcess.messages.in:
* WebProcess/Network/WebLoaderStrategy.cpp:
(WebKit::WebLoaderStrategy::isResourceLoadFinished):
* WebProcess/Network/WebLoaderStrategy.h:
2020-09-18 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] Epiphany does not make connections in Incognito Mode after clearing .cache
https://bugs.webkit.org/show_bug.cgi?id=216430
Reviewed by Adrian Perez de Castro.
The problem is that when using an ephemeral context, the download is started before the ephemeral session has
been registered in the network process, so it's just ignored.
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::createDownloadProxy): Pass the given WebsiteDataStore to ensureNetworkProcess().
2020-09-18 Youenn Fablet <youenn@apple.com>
Enable MediaRecorder by default on MacOS
https://bugs.webkit.org/show_bug.cgi?id=216663
Reviewed by Eric Carlson.
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultMediaRecorderEnabled):
* Shared/WebPreferencesDefaultValues.h:
* Shared/WebPreferencesExperimental.yaml:
2020-09-17 Sam Weinig <weinig@apple.com>
Remove runtime setting for enabling/disabling the UserTiming and ResourceTiming APIs
https://bugs.webkit.org/show_bug.cgi?id=216556
Reviewed by Darin Adler.
Remove runtime setting for enabling/disabling UserTiming and ResourceTiming
APIs. They have both been on for a while in WebKit2 and there aren't any
known uses cases for disabling them. In addition to removing the setting,
they are now enabled in LegacyWebKit as well.
* Shared/WebPreferences.yaml:
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetUserTimingEnabled):
(WKPreferencesGetUserTimingEnabled):
(WKPreferencesSetResourceTimingEnabled):
(WKPreferencesGetResourceTimingEnabled):
* UIProcess/API/C/WKPreferencesRefPrivate.h:
2020-09-16 Darin Adler <darin@apple.com>
Selection API: Introduce LiveRangeSelectionEnabled, off by default
https://bugs.webkit.org/show_bug.cgi?id=216656
Reviewed by Sam Weinig.
* Shared/WebPreferencesInternal.yaml: Added LiveRangeSelectionEnabled.
2020-09-17 Tim Horton <timothy_horton@apple.com>
Per-app accent color overrides are broken
https://bugs.webkit.org/show_bug.cgi?id=216674
<rdar://problem/67745464>
Reviewed by Wenson Hsieh.
No new tests; I've no idea how to write a test that would have caught this;
any test I can imagine would have also written to _accentColor and worked fine.
* UIProcess/mac/PageClientImplMac.mm:
(WebKit::PageClientImpl::accentColor):
AppKit changed the meaning of the IPI. Use the correct version.
We still write to _accentColor; it is now purely an override.
2020-09-17 Kate Cheney <katherine_cheney@apple.com>
Enable ITP in WKWebViews for apps with the full browser entitlement who are not linked to iOS 14.0
https://bugs.webkit.org/show_bug.cgi?id=216655
<rdar://problem/69081227>
Reviewed by Brent Fulgham.
Apps with the full browser entitlement who are not linked to iOS 14.0
should have ITP enabled in WKWebViews by default.
No new tests, no way currently to test apps with different iOS
versions. Tested manually.
* Shared/Cocoa/DefaultWebBrowserChecks.mm:
(WebKit::determineITPStateInternal):
Ignore the early return for the linked-on-or-after check if the
app has the default browser entitlement so we still enable ITP based
on the user's preference.
2020-09-17 David Kilzer <ddkilzer@apple.com>
WebKit::XPCServiceEventHandler block should call exit() on the main thread
<https://webkit.org/b/216594>
<rdar://problem/68053217>
Reviewed by Chris Dumez.
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceMain.mm:
(WebKit::XPCServiceEventHandler):
- Call exit() on the main thread. I chose dispatch_sync() here
so that if other crashes occur on the main thread, we will
know if this background thread is waiting to call exit() at
the same time. Also, if a different background thread calls
exit() in the future with this block running on the main
thread, we'll know where the block came from since the XPC
handler thread will be waiting for the dispatch_sync() to
return (instead of having a mysterious block calling exit() on
the main thread and not knowing where it came from).
- Add logging (including simulated crash logs) when either of
these exit() code paths is taken.
2020-09-17 Hoa Dinh <dvh@apple.com>
Rename ShouldAllowExternalSchemes to ShouldAllowExternalSchemesButNotAppLinks to clarify what this enum value is for.
https://bugs.webkit.org/show_bug.cgi?id=216651
Reviewed by Alex Christensen.
Rename ShouldAllowExternalSchemes to ShouldAllowExternalSchemesButNotAppLinks and reflects the change where it's used.
Source/WebKit:
* Shared/SessionState.cpp:
(WebKit::isValidEnum):
* UIProcess/API/APINavigationAction.h:
* UIProcess/API/glib/WebKitWebViewSessionState.cpp:
(toExternalURLsPolicy):
(toWebCoreExternalURLsPolicy):
* UIProcess/WebPageProxy.h:
* UIProcess/mac/LegacySessionStateCoding.cpp:
(WebKit::decodeSessionHistoryEntry):
* WebProcess/InjectedBundle/InjectedBundleNavigationAction.cpp:
(WebKit::InjectedBundleNavigationAction::InjectedBundleNavigationAction):
2020-09-17 Sam Weinig <weinig@apple.com>
We shouldn't need to specify a category for preferences now that they are split into separate files
https://bugs.webkit.org/show_bug.cgi?id=216648
Reviewed by Tim Horton.
Remove need / ability to specify a category in the yaml, rather, infer the category
from which file it is in.
* Scripts/GeneratePreferences.rb:
* Shared/WebPreferencesDebug.yaml:
* Shared/WebPreferencesExperimental.yaml:
* Shared/WebPreferencesInternal.yaml:
2020-09-17 Antoine Quint <graouts@webkit.org>
Remove the legacy animation code
https://bugs.webkit.org/show_bug.cgi?id=216608
Reviewed by Dean Jackson.
Since we now have a single animation engine, the runtime flags to optionally enable the Web Animations engine are
not needed anymore and are now removed.
* Shared/WebPreferencesExperimental.yaml:
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetWebAnimationsEnabled): Deleted.
(WKPreferencesGetWebAnimationsEnabled): Deleted.
(WKPreferencesSetWebAnimationsCSSIntegrationEnabled): Deleted.
(WKPreferencesGetWebAnimationsCSSIntegrationEnabled): Deleted.
* UIProcess/API/C/WKPreferencesRefPrivate.h:
* UIProcess/API/Cocoa/WKPreferences.mm:
(-[WKPreferences _setWebAnimationsCSSIntegrationEnabled:]): Deleted.
(-[WKPreferences _webAnimationsCSSIntegrationEnabled]): Deleted.
* UIProcess/API/Cocoa/WKPreferencesPrivate.h:
* WebProcess/InjectedBundle/InjectedBundle.cpp:
(WebKit::InjectedBundle::overrideBoolPreferenceForTestRunner):
(WebKit::InjectedBundle::setWebAnimationsEnabled): Deleted.
(WebKit::InjectedBundle::setWebAnimationsCSSIntegrationEnabled): Deleted.
* WebProcess/InjectedBundle/InjectedBundle.h:
2020-09-17 Rob Buis <rbuis@igalia.com>
Clean up lazy frame loading
https://bugs.webkit.org/show_bug.cgi?id=216560
Reviewed by Youenn Fablet.
Allow web preference to be consistent with lazy image loading.
* Shared/WebPreferencesExperimental.yaml:
2020-09-17 Jer Noble <jer.noble@apple.com>
Unified build fix for "redefinition of 'kCMVideoCodecType_VP9"
https://bugs.webkit.org/show_bug.cgi?id=216205
Reviewed by Youenn Fablet.
Wrap calls to WebCore VP9 functions in ENABLE(VP9) checks.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
* WebProcess/WebProcess.cpp:
* WebProcess/WebProcess.h:
2020-09-16 Sam Weinig <weinig@apple.com>
WebPreferences.yaml is overwhelming, what if we split it up a bit
https://bugs.webkit.org/show_bug.cgi?id=216627
Reviewed by Tim Horton.
Split out experimental, internal and debug preferences into their own
yaml files to add a bit of order.
* CMakeLists.txt:
* DerivedSources.make:
* Shared/WebPreferences.yaml:
* Shared/WebPreferencesDebug.yaml: Copied from Source/WebKit/Shared/WebPreferences.yaml.
* Shared/WebPreferencesExperimental.yaml: Copied from Source/WebKit/Shared/WebPreferences.yaml.
* Shared/WebPreferencesInternal.yaml: Copied from Source/WebKit/Shared/WebPreferences.yaml.
* WebKit.xcodeproj/project.pbxproj:
2020-09-16 Kate Cheney <katherine_cheney@apple.com>
Revert _setIgnoresAppBoundDomains SPI to compile for macOS
https://bugs.webkit.org/show_bug.cgi?id=216628
<rdar://problem/69025655>
Reviewed by Tim Horton.
We need to give users of the _setIgnoresAppBoundDomains SPI time to
convert uses to iOS only, so in the meantime we should still compile
the SPI for macOS so we don't break anything.
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration _ignoresAppBoundDomains]):
(-[WKWebViewConfiguration _setIgnoresAppBoundDomains:]):
* UIProcess/API/Cocoa/WKWebViewConfigurationPrivate.h:
2020-09-16 Sam Weinig <weinig@apple.com>
Remove runtime setting for enabling/disabling CSS shadow parts
https://bugs.webkit.org/show_bug.cgi?id=216567
Reviewed by Tim Horton.
Remove runtime setting for enabling/disabling CSS shadow parts. It is
on by default and there is no known reason one would need to disable it.
* Shared/WebPreferences.yaml:
2020-09-16 Sam Weinig <weinig@apple.com>
Remove runtime setting for enabling/disabling the isSecureContext API?
https://bugs.webkit.org/show_bug.cgi?id=216562
Reviewed by Darin Adler.
Remove runtime setting for enabling/disabling the the isSecureContext attribute.
on window and the worker global scope. There is no known reason one would need
to disable it.
* Shared/WebPreferences.yaml:
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetIsSecureContextAttributeEnabled):
(WKPreferencesGetIsSecureContextAttributeEnabled):
* UIProcess/API/C/WKPreferencesRefPrivate.h:
* WebProcess/InjectedBundle/InjectedBundle.cpp:
(WebKit::InjectedBundle::overrideBoolPreferenceForTestRunner):
2020-09-16 Youenn Fablet <youenn@apple.com>
RemoteSampleBufferDisplayLayer should process its IPC messages from a background thread
https://bugs.webkit.org/show_bug.cgi?id=216475
Reviewed by Eric Carlson.
Use a process wide work queue to process all track video renderer messages in a background thread.
This queue is shared by all RemoteSampleBufferDisplayLayerManager and RemoteSampleBufferDisplayLayer objects, which are per web process.
Since LocalSampleBufferDisplayLayer must currently be created in main thread, hop to main thread to create/delete renderers.
We update messages.py to support WantsAsyncDispatchMessage, which removes the sync handling with dispatchMessage.
This allows RemoteSampleBufferDisplayLayerManager to also handle messages from RemoteSampleBufferDisplayLayer.
Update RemoteSampleBufferDisplayLayerManager accordingly as well.
Covered by webrtc/video.html with GPU process enabled.
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::GPUConnectionToWebProcess):
(WebKit::GPUConnectionToWebProcess::dispatchMessage):
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/GPUProcess.cpp:
(WebKit::GPUProcess::videoMediaStreamTrackRendererQueue):
* GPUProcess/GPUProcess.h:
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRendererManager.cpp:
(WebKit::RemoteAudioMediaStreamTrackRendererManager::dispatchMessage):
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRendererManager.h:
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRendererManager.messages.in:
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayerManager.cpp:
(WebKit::RemoteSampleBufferDisplayLayerManager::RemoteSampleBufferDisplayLayerManager):
(WebKit::RemoteSampleBufferDisplayLayerManager::~RemoteSampleBufferDisplayLayerManager):
(WebKit::RemoteSampleBufferDisplayLayerManager::close):
(WebKit::RemoteSampleBufferDisplayLayerManager::dispatchToThread):
(WebKit::RemoteSampleBufferDisplayLayerManager::dispatchMessage):
(WebKit::RemoteSampleBufferDisplayLayerManager::createLayer):
(WebKit::RemoteSampleBufferDisplayLayerManager::releaseLayer):
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayerManager.h:
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayerManager.messages.in:
* Scripts/webkit/messages.py:
* SourcesCocoa.txt:
2020-09-16 Alex Christensen <achristensen@webkit.org>
Move TLS certificate bypass SPI from WebProcessPool to WebsiteDataStore
https://bugs.webkit.org/show_bug.cgi?id=216585
Reviewed by Geoffrey Garen.
It will eventually be replaced by WKNavigationDelegate's didReceiveAuthenticationChallenge in rdar://30655740
but the SPI needs to continue to exist somewhere until that is complete.
In the meantime, it needs to be moved from the WebProcessPool for https://bugs.webkit.org/show_bug.cgi?id=216041
and rdar://problem/68942815 tracks Safari's adoption of the SPI introduced here, which I will fix within an hour of landing this.
* UIProcess/API/C/WKContext.cpp:
(WKContextAllowSpecificHTTPSCertificateForHost):
* UIProcess/API/C/WKContextPrivate.h:
* UIProcess/API/Cocoa/WKProcessGroup.mm:
(-[WKProcessGroup _setAllowsSpecificHTTPSCertificate:forHost:]): Deleted.
* UIProcess/API/Cocoa/WKProcessGroupPrivate.h:
* UIProcess/API/Cocoa/WKProcessPool.mm:
(-[WKProcessPool _setAllowsSpecificHTTPSCertificate:forHost:]):
* UIProcess/API/Cocoa/WKProcessPoolPrivate.h:
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
(-[WKWebsiteDataStore _allowTLSCertificate:forHost:]):
* UIProcess/API/Cocoa/WKWebsiteDataStorePrivate.h:
* UIProcess/API/glib/WebKitWebContext.cpp:
(webkit_web_context_allow_tls_certificate_for_host):
Use the default data store until this API can be replaced by API in a different place.
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::allowSpecificHTTPSCertificateForHost): Deleted.
* UIProcess/WebProcessPool.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::allowSpecificHTTPSCertificateForHost):
* UIProcess/WebsiteData/WebsiteDataStore.h:
2020-09-16 Alex Christensen <achristensen@webkit.org>
Remove unneeded HAVE_FOUNDATION_WITH_SAVE_COOKIES_WITH_COMPLETION_HANDLER
https://bugs.webkit.org/show_bug.cgi?id=216572
Reviewed by Youenn Fablet.
It's available everywhere we support.
* NetworkProcess/cocoa/NetworkProcessCocoa.mm:
(WebKit::saveCookies):
(WebKit::NetworkProcess::platformFlushCookies):
2020-09-16 Youenn Fablet <youenn@apple.com>
Move LibWebRTCCodecsProxy message handling out of the main thread
https://bugs.webkit.org/show_bug.cgi?id=216531
Reviewed by Eric Carlson.
Use a process wide work queue to process all LibWebRTCCodecsProxy messages in a background thread.
This queue is shared by all LibWebRTCCodecsProxy objects, which are per web process.
Covered by webrtc/video.html with GPU process enabled.
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::GPUConnectionToWebProcess):
(WebKit::GPUConnectionToWebProcess::~GPUConnectionToWebProcess):
(WebKit::GPUConnectionToWebProcess::dispatchMessage):
(WebKit::GPUConnectionToWebProcess::libWebRTCCodecsProxy): Deleted.
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/GPUProcess.cpp:
(WebKit::GPUProcess::libWebRTCCodecsQueue):
* GPUProcess/GPUProcess.h:
* GPUProcess/webrtc/LibWebRTCCodecsProxy.h:
(WebKit::LibWebRTCCodecsProxy::create):
(WebKit::LibWebRTCCodecsProxy::didReceiveMessageFromWebProcess): Deleted.
* GPUProcess/webrtc/LibWebRTCCodecsProxy.mm:
(WebKit::LibWebRTCCodecsProxy::LibWebRTCCodecsProxy):
(WebKit::LibWebRTCCodecsProxy::~LibWebRTCCodecsProxy):
(WebKit::LibWebRTCCodecsProxy::dispatchToThread):
(WebKit::LibWebRTCCodecsProxy::close):
2020-09-16 Adrian Perez de Castro <aperez@igalia.com>
Non-unified build fixes, early September 2020 edition
https://bugs.webkit.org/show_bug.cgi?id=216599
Unreviewed build fix.
Largely based on a patch by Lauro Moura <lmoura@igalia.com>
* GPUProcess/media/RemoteCDMProxy.cpp: Add missing WebCore/SharedBuffer.h include.
* GPUProcess/media/RemoteMediaResourceLoader.cpp: Add missing WebCore/ResourceError.h
include.
* NetworkProcess/webrtc/NetworkRTCProvider.cpp: Add missing DataReference.h include.
* UIProcess/gtk/WebDataListSuggestionsDropdownGtk.cpp:
(WebKit::WebDataListSuggestionsDropdownGtk::show): Add missing namespace to
WebCore::monitorWorkArea() call.
* WebProcess/Network/WebSocketChannelManager.cpp: Add missing
WebCore/WebSocketIdentifier.h include.
(WebKit::WebSocketChannelManager::didReceiveMessage): Add missing namespace to
WebCore::WebSocketIdentifierType usage.
2020-09-15 Said Abou-Hallawa <sabouhallawa@apple.com>
REGRESSION(r266677): ASSERTION FAILED in PaintFrequencyTracker::end()
https://bugs.webkit.org/show_bug.cgi?id=216591
<rdar://problem/68635418>
Reviewed by Simon Fraser.
timestampForPaintFrequencyTracking() do not have to be in ChromeClient.
RenderLayer can access it through its page(). No need to special case
isSVGImageChromeClient() because we are going to track the painting that
goes through Page::updateRendering(). This means all the painting of the
SVG image RenderLayers will be tracked by the same timestamp. Currently
we do not support compositing inside the SVG image which means there is
only one RenderLayer per SVG image.
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::timestampForPaintFrequencyTracking const): Deleted.
* WebProcess/WebCoreSupport/WebChromeClient.h:
2020-09-15 Aditya Keerthi <akeerthi@apple.com>
[macOS] Date pickers should respect the document's color scheme
https://bugs.webkit.org/show_bug.cgi?id=216514
<rdar://problem/68889548>
Reviewed by Tim Horton.
Use the appearance property of NSWindow to ensure the appearance of the
contained NSDatePicker matches the value encoded in DateTimeChooserParameters.
* UIProcess/mac/WebDateTimePickerMac.mm:
(-[WKDateTimePicker initWithParams:inView:]):
2020-09-15 Alex Christensen <achristensen@webkit.org>
Move local storage sync SPI from WebProcessPool to WebsiteDataStore
https://bugs.webkit.org/show_bug.cgi?id=216581
Reviewed by Geoffrey Garen.
This move is needed for https://bugs.webkit.org/show_bug.cgi?id=216041
We no longer need to sync the legacy private browsing storage because it is no longer used by WebKitTestRunner. Thank goodness.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::clearLegacyPrivateBrowsingLocalStorage): Deleted.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* UIProcess/API/C/WKContext.cpp:
(WKContextSyncLocalStorage): Deleted.
(WKContextClearLegacyPrivateBrowsingLocalStorage): Deleted.
* UIProcess/API/C/WKContextPrivate.h:
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreSyncLocalStorage):
* UIProcess/API/C/WKWebsiteDataStoreRef.h:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::syncLocalStorage): Deleted.
(WebKit::WebProcessPool::clearLegacyPrivateBrowsingLocalStorage): Deleted.
* UIProcess/WebProcessPool.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::syncLocalStorage):
* UIProcess/WebsiteData/WebsiteDataStore.h:
2020-09-15 Alex Christensen <achristensen@webkit.org>
Move service worker testing timeout setting from WebProcessPool to WebsiteDataStore
https://bugs.webkit.org/show_bug.cgi?id=216578
Reviewed by Geoffrey Garen.
This is needed for https://bugs.webkit.org/show_bug.cgi?id=216041
* UIProcess/API/C/WKContext.cpp:
(WKContextSetServiceWorkerFetchTimeoutForTesting): Deleted.
(WKContextResetServiceWorkerFetchTimeoutForTesting): Deleted.
* UIProcess/API/C/WKContext.h:
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreSetServiceWorkerFetchTimeoutForTesting):
(WKWebsiteDataStoreResetServiceWorkerFetchTimeoutForTesting):
* UIProcess/API/C/WKWebsiteDataStoreRef.h:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::setServiceWorkerTimeoutForTesting): Deleted.
(WebKit::WebProcessPool::resetServiceWorkerTimeoutForTesting): Deleted.
* UIProcess/WebProcessPool.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::setServiceWorkerTimeoutForTesting):
(WebKit::WebsiteDataStore::resetServiceWorkerTimeoutForTesting):
* UIProcess/WebsiteData/WebsiteDataStore.h:
2020-09-15 Peng Liu <peng.liu6@apple.com>
REGRESSION (r267053): ASSERTION FAILED: videoFullscreenManager->client() == &_videoFullscreenManagerProxyClient in [WKFullScreenWindowController _completedExitFullScreen]
https://bugs.webkit.org/show_bug.cgi?id=216571
Reviewed by Jer Noble.
It is possible that an element exits fullscreen before it completely enters
fullscreen. In that case, the client of VideoFullscreenManagerProxy is nullptr.
No new tests, fixing an assertion failure when running
TestWebKitAPI.WKWebViewCloseAllMediaPresentations.ElementFullscreen.
* UIProcess/ios/fullscreen/WKFullScreenWindowControllerIOS.mm:
(-[WKFullScreenWindowController _completedExitFullScreen]):
2020-09-15 Sam Weinig <weinig@apple.com>
Remove runtime setting for enabling/disabling the fetch API
https://bugs.webkit.org/show_bug.cgi?id=216519
Reviewed by Alex Christensen.
Remove runtime setting for enabling/disabling the fetch API. There is no
reason one needs to disable it now and it has been on by default for
a long enough time.
Also cleans up WKPreferencesRefPrivate.h/WKPreferences.cpp to move all the obsolete
preferences to the bottom with a comment about removing them when possible.
* Shared/WebPreferences.yaml:
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesResetTestRunnerOverrides):
(WKPreferencesSetPrivateBrowsingEnabled):
(WKPreferencesGetPrivateBrowsingEnabled):
(WKPreferencesSetIgnoreViewportScalingConstraints):
(WKPreferencesGetIgnoreViewportScalingConstraints):
(WKPreferencesSetWebRTCLegacyAPIEnabled):
(WKPreferencesGetWebRTCLegacyAPIEnabled):
(WKPreferencesSetNewBlockInsideInlineModelEnabled):
(WKPreferencesGetNewBlockInsideInlineModelEnabled):
(WKPreferencesSetLongMousePressEnabled):
(WKPreferencesGetLongMousePressEnabled):
(WKPreferencesSetAntialiasedFontDilationEnabled):
(WKPreferencesGetAntialiasedFontDilationEnabled):
(WKPreferencesSetHixie76WebSocketProtocolEnabled):
(WKPreferencesGetHixie76WebSocketProtocolEnabled):
(WKPreferencesSetApplicationChromeModeEnabled):
(WKPreferencesGetApplicationChromeModeEnabled):
(WKPreferencesSetInspectorUsesWebKitUserInterface):
(WKPreferencesGetInspectorUsesWebKitUserInterface):
(WKPreferencesSetFetchAPIEnabled):
(WKPreferencesGetFetchAPIEnabled):
* UIProcess/API/C/WKPreferencesRefPrivate.h:
2020-09-15 Sam Weinig <weinig@apple.com>
Remove runtime setting for enabling/disabling custom elements
https://bugs.webkit.org/show_bug.cgi?id=216518
Reviewed by Tim Horton.
Remove runtime setting for enabling/disabling custom elements. There is no
reason one need to disable it now, and it has been on by default for
a long enough time.
* Shared/WebPreferences.yaml:
2020-09-15 Simon Fraser <simon.fraser@apple.com>
REGRESSION (r267002): ASSERTION FAILED:[ iOS wk2 Debug ] !HashTranslator::equal(KeyTraits::emptyValue(), key) on compositing/ios/overflow-scroll-update-overlap.html
https://bugs.webkit.org/show_bug.cgi?id=216520
<rdar://problem/68894931>
Reviewed by Tim Horton.
Avoid hash lookups with zero nodeIDs.
* WebProcess/WebPage/RemoteLayerTree/RemoteScrollingCoordinator.mm:
(WebKit::RemoteScrollingCoordinator::isRubberBandInProgress const):
(WebKit::RemoteScrollingCoordinator::isUserScrollInProgress const):
(WebKit::RemoteScrollingCoordinator::isScrollSnapInProgress const):
2020-09-15 Alex Christensen <achristensen@webkit.org>
Move WKContextClearCachedCredentials SPI to WKWebsiteDataStoreClearCachedCredentials
https://bugs.webkit.org/show_bug.cgi?id=216513
Reviewed by Chris Dumez.
The credentials are owned by session anyways.
This is needed for https://bugs.webkit.org/show_bug.cgi?id=216041
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::clearCachedCredentials):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* UIProcess/API/C/WKContext.cpp:
(WKContextClearCachedCredentials): Deleted.
* UIProcess/API/C/WKContextPrivate.h:
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreClearCachedCredentials):
* UIProcess/API/C/WKWebsiteDataStoreRef.h:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::clearCachedCredentials):
* UIProcess/WebProcessPool.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::clearCachedCredentials):
* UIProcess/WebsiteData/WebsiteDataStore.h:
2020-09-15 Alex Christensen <achristensen@webkit.org>
Remove unused plumbing for setting whether the storage access API is enabled
https://bugs.webkit.org/show_bug.cgi?id=216516
Reviewed by Youenn Fablet.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/NetworkProcessCreationParameters.cpp:
(WebKit::NetworkProcessCreationParameters::encode const):
(WebKit::NetworkProcessCreationParameters::decode):
* NetworkProcess/NetworkProcessCreationParameters.h:
* NetworkProcess/cocoa/NetworkProcessCocoa.mm:
(WebKit::NetworkProcess::platformInitializeNetworkProcessCocoa):
(WebKit::NetworkProcess::setStorageAccessAPIEnabled): Deleted.
* UIProcess/API/C/WKContext.cpp:
(WKContextSetStorageAccessAPIEnabled): Deleted.
* UIProcess/API/C/WKContext.h:
* UIProcess/API/C/WKCookieManager.cpp:
(WKCookieManagerSetStorageAccessAPIEnabled): Deleted.
* UIProcess/API/C/WKCookieManager.h:
* UIProcess/API/Cocoa/WKProcessPool.mm:
(-[WKProcessPool _isStorageAccessAPIEnabled]): Deleted.
(-[WKProcessPool _setStorageAccessAPIEnabled:]): Deleted.
* UIProcess/API/Cocoa/WKProcessPoolPrivate.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeNetworkProcess):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::setStorageAccessAPIEnabled): Deleted.
* UIProcess/WebProcessPool.h:
2020-09-15 Alex Christensen <achristensen@webkit.org>
Move WebSocket TLS bypass SPI from WebProcessPool to WebsiteDataStore
https://bugs.webkit.org/show_bug.cgi?id=216509
Reviewed by Youenn Fablet.
After https://bugs.webkit.org/show_bug.cgi?id=216041 we will be unable to determine which WebsiteDataStore's
NetworkProcess to add the TLS bypass to, so we need to move the SPI to the WebsiteDataStore object. This breaks
off a piece of that large change into a small, easy-to-review change.
* UIProcess/API/C/WKContext.cpp:
(WKContextSetAllowsAnySSLCertificateForWebSocketTesting): Deleted.
* UIProcess/API/C/WKContextPrivate.h:
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreSetAllowsAnySSLCertificateForWebSocketTesting):
* UIProcess/API/C/WKWebsiteDataStoreRef.h:
* UIProcess/API/Cocoa/WKProcessPool.mm:
(-[WKProcessPool _allowAnyTLSCertificateForWebSocketTesting]): Deleted.
* UIProcess/API/Cocoa/WKProcessPoolPrivate.h:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::setAllowsAnySSLCertificateForWebSocket): Deleted.
* UIProcess/WebProcessPool.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::setAllowsAnySSLCertificateForWebSocket):
* UIProcess/WebsiteData/WebsiteDataStore.h:
2020-09-15 Aditya Keerthi <akeerthi@apple.com>
[macOS] Update date picker appearance to match system date pickers
https://bugs.webkit.org/show_bug.cgi?id=216517
<rdar://problem/68890678>
Reviewed by Wenson Hsieh.
System date pickers in macOS do not use NSVisualEffectView. Instead, use
a solid background view with three rounded corners.
* UIProcess/mac/WebDateTimePickerMac.mm:
(-[WKDateTimePickerWindow initWithContentRect:styleMask:backing:defer:]):
(-[WKDateTimePickerBackdropView drawRect:]):
(-[WKDateTimePicker initWithParams:inView:]):
2020-09-15 Rob Buis <rbuis@igalia.com>
Convert runtime flag to setting for lazy image loading
https://bugs.webkit.org/show_bug.cgi?id=215610
Reviewed by Youenn Fablet.
Remove lazy image loading runtime flag.
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetLazyImageLoadingEnabled): Deleted.
(WKPreferencesGetLazyImageLoadingEnabled): Deleted.
* UIProcess/API/C/WKPreferencesRefPrivate.h:
2020-09-15 Youenn Fablet <youenn@apple.com>
End of media capture should not be reported before 3 seconds of the start of capture
https://bugs.webkit.org/show_bug.cgi?id=216415
<rdar://problem/68512358>
Reviewed by Eric Carlson.
Add support for delaying of end of capture notification to the application.
This allows to ensure that a capture indicator stays for long enough for the user to notice it.
A capture indicator should be visible to the user for at least three seconds.
A timer is scheduled when starting a capture and there is no ongoing capture.
As long as the timer is active, the capture state cannot be transitioned to no capture.
Other transitions are allowed.
Once the timer kicks in, any capture state update is done synchronously.
Ensure to update the capture state when the timer kicks in.
Note that even navigations will not allow transitioning the capture state sooner.
This is done to ensure a page does not try to capture one frame before navigating to another page.
In practice, very few pages should navigate quickly after starting capture.
* UIProcess/API/C/WKPage.cpp:
(WKPageGetMediaState):
(WKPageSetMediaCaptureReportingDelayForTesting):
* UIProcess/API/C/WKPagePrivate.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _mediaCaptureState]):
* UIProcess/API/Cocoa/WKWebViewPrivateForTesting.h:
* UIProcess/API/Cocoa/WKWebViewTesting.mm:
(-[WKWebView _mediaCaptureReportingDelayForTesting]):
(-[WKWebView _setMediaCaptureReportingDelayForTesting:]):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::updatePlayingMediaDidChange):
(WebKit::WebPageProxy::updateReportedMediaCaptureState):
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::reportedMediaCaptureState const):
(WebKit::WebPageProxy::mediaCaptureReportingDelay const):
(WebKit::WebPageProxy::setMediaCaptureReportingDelay):
2020-09-15 Youenn Fablet <youenn@apple.com>
Enable VP9 in GPUProcess if page needs it
https://bugs.webkit.org/show_bug.cgi?id=216472
Reviewed by Eric Carlson.
In case a page is using GPUProcess for media, send the VP9 parameters to GPUProcess.
GPUProcess will enable VP9 decoders based on that input.
Covered by LayoutTests/media/vp9.html with GPU process enabled.
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::enableVP9Decoders):
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/GPUConnectionToWebProcess.messages.in:
* GPUProcess/GPUProcess.cpp:
(WebKit::GPUProcess::enableVP9Decoders):
* GPUProcess/GPUProcess.h:
* WebProcess/GPU/GPUProcessConnection.cpp:
(WebKit::GPUProcessConnection::updateParameters):
* WebProcess/GPU/GPUProcessConnection.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
2020-09-14 Sam Weinig <weinig@apple.com>
Remove runtime setting for enabling/disabling ShadowDOM
https://bugs.webkit.org/show_bug.cgi?id=216511
Reviewed by Ryosuke Niwa.
Remove runtime setting for enabling/disabling ShadowDOM. There is no
reason one need to disable it now, and it has been on by default for
a long enough time.
* Shared/WebPreferences.yaml:
2020-09-14 Aditya Keerthi <akeerthi@apple.com>
[macOS] Do not dismiss date picker when a value is selected
https://bugs.webkit.org/show_bug.cgi?id=216515
<rdar://problem/68889944>
Reviewed by Darin Adler.
The calendar presented when activating a date input should not be
dismissed after the user selects a date. This matches the behavior
of date pickers in Calendar.app.
* UIProcess/mac/WebDateTimePickerMac.mm:
(WebKit::WebDateTimePickerMac::didChooseDate):
2020-09-14 Peng Liu <peng.liu6@apple.com>
Returning to element fullscreen from PiP is not stable under stress tests
https://bugs.webkit.org/show_bug.cgi?id=216287
Reviewed by Jer Noble.
* UIProcess/Cocoa/VideoFullscreenManagerProxy.h:
(WebKit::VideoFullscreenManagerProxyClient::~VideoFullscreenManagerProxyClient):
(WebKit::VideoFullscreenManagerProxy::setClient):
(WebKit::VideoFullscreenManagerProxy::client const):
* UIProcess/Cocoa/VideoFullscreenManagerProxy.mm:
(WebKit::VideoFullscreenModelContext::didEnterPictureInPicture):
(WebKit::VideoFullscreenModelContext::didExitPictureInPicture):
(WebKit::VideoFullscreenModelContext::willEnterPictureInPicture):
(WebKit::VideoFullscreenManagerProxy::hasVideoInPictureInPictureDidChange):
(WebKit::VideoFullscreenManagerProxy::fullscreenMayReturnToInline):
(WebKit::VideoFullscreenModelContext::prepareToExitFullscreen): Deleted.
Add the interface VideoFullscreenManagerProxyClient, which is used by WKFullScreenWindowController
to implement the support of "returning to element fullscreen from PiP". Using VideoFullscreenModelClient
for that purpose is not a good idea because the instance observed by VideoFullscreenModelClient may be
destroyed when a video element is exiting picture-in-picture if we don't call
VideoFullscreenManagerProxy::addClientForContext() and VideoFullscreenManagerProxy::removeClientForContext()
properly.
* UIProcess/ios/fullscreen/WKFullScreenWindowControllerIOS.mm:
(-[WKFullScreenWindowController initWithWebView:]):
(-[WKFullScreenWindowController dealloc]):
(-[WKFullScreenWindowController beganEnterFullScreenWithInitialFrame:finalFrame:]):
(-[WKFullScreenWindowController _completedExitFullScreen]):
(-[WKFullScreenWindowController videoControlsManagerDidChange]):
(-[WKFullScreenWindowController prepareToExitPictureInPicture]):
(-[WKFullScreenWindowController didExitPictureInPicture]):
(-[WKFullScreenWindowController _videoFullscreenManager]):
(WKFullScreenWindowControllerVideoFullscreenModelClient::setParent): Deleted.
(WKFullScreenWindowControllerVideoFullscreenModelClient::setInterface): Deleted.
(WKFullScreenWindowControllerVideoFullscreenModelClient::interface const): Deleted.
(-[WKFullScreenWindowController willEnterPictureInPicture]): Deleted.
(-[WKFullScreenWindowController failedToEnterPictureInPicture]): Deleted.
Use VideoFullscreenManagerProxyClient instead of VideoFullscreenModelClient to implement
the support of "returning to element fullscreen from PiP" on iOS.
A user can "return to element fullscreen from PiP" by clicking the "return to fullscreen"
button on the PiP window, or by clicking the "fullscreen" button on the page. This patch
adds support for the latter case.
* UIProcess/mac/WKFullScreenWindowController.h:
* UIProcess/mac/WKFullScreenWindowController.mm:
(-[WKFullScreenWindowController initWithWindow:webView:page:]):
(-[WKFullScreenWindowController dealloc]):
(-[WKFullScreenWindowController videoControlsManagerDidChange]):
(-[WKFullScreenWindowController didExitPictureInPicture]):
(-[WKFullScreenWindowController windowDidEnterFullScreen:]):
(-[WKFullScreenWindowController _videoFullscreenManager]):
(WebKit::WKFullScreenWindowControllerVideoFullscreenModelClient::setParent): Deleted.
(WebKit::WKFullScreenWindowControllerVideoFullscreenModelClient::setInterface): Deleted.
(WebKit::WKFullScreenWindowControllerVideoFullscreenModelClient::interface const): Deleted.
Use VideoFullscreenManagerProxyClient instead of VideoFullscreenModelClient to implement the
support of "exiting fullscreen after entering PiP" on Mac. We may implement the support of
"returning to element fullscreen from PiP" on Mac in the future.
2020-09-14 Alex Christensen <achristensen@webkit.org>
Move cookie flushing SPI from WKProcessPool to WKHTTPCookieStore
https://bugs.webkit.org/show_bug.cgi?id=216493
Reviewed by Chris Dumez.
In order for https://bugs.webkit.org/show_bug.cgi?id=216041 to be possible, our interface for flushing cookies to disk can't be
on the WKProcessPool. It should be with the other cookie operations, in WKHTTPCookieStore.
AuxiliaryProcessProxy::sendWithAsyncReply takes care of the background assertion, so we remove that code from NetworkProcessProxy::syncAllCookies.
rdar://problem/68872711 tracks adoption in Safari, which I will land within an hour of landing this to avoid cookie regressions.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::syncAllCookies):
(WebKit::NetworkProcess::didSyncAllCookies): Deleted.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/cocoa/NetworkProcessCocoa.mm:
(WebKit::NetworkProcess::syncAllCookies):
* UIProcess/API/APIHTTPCookieStore.cpp:
(API::HTTPCookieStore::syncCookies):
* UIProcess/API/APIHTTPCookieStore.h:
* UIProcess/API/Cocoa/WKHTTPCookieStore.mm:
(-[WKHTTPCookieStore _flushCookiesToDiskWithCompletionHandler:]):
* UIProcess/API/Cocoa/WKHTTPCookieStorePrivate.h:
* UIProcess/API/Cocoa/WKProcessPool.mm:
(-[WKProcessPool _syncNetworkProcessCookies]):
* UIProcess/API/Cocoa/WKProcessPoolPrivate.h:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::didClose):
(WebKit::NetworkProcessProxy::syncAllCookies):
(WebKit::NetworkProcessProxy::didSyncAllCookies): Deleted.
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/Network/NetworkProcessProxy.messages.in:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::syncCookies):
(WebKit::WebProcessPool::syncNetworkProcessCookies): Deleted.
* UIProcess/WebProcessPool.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::syncCookies):
* UIProcess/WebsiteData/WebsiteDataStore.h:
2020-09-14 Fujii Hironori <Hironori.Fujii@sony.com>
[WinCairo][MediaFoundation] r264188 introduced linker warning: LNK4199: /DELAYLOAD:mf.dll ignored; no imports found from mf.dll
https://bugs.webkit.org/show_bug.cgi?id=214170
Reviewed by Alex Christensen.
* PlatformWin.cmake:
2020-09-14 Devin Rousso <drousso@apple.com>
Web Inspector: docking buttons don't work when already docked if window is too small
https://bugs.webkit.org/show_bug.cgi?id=216488
<rdar://problem/68242717>
Reviewed by Timothy Hatcher.
* UIProcess/Inspector/WebInspectorProxy.cpp:
(WebKit::WebInspectorProxy::attach):
(WebKit::WebInspectorProxy::attachAvailabilityChanged):
If Web Inspector is already attached, it can re-attach with a different configuration.
2020-09-14 Dean Jackson <dino@apple.com>
Enable WebGL2 by default
https://bugs.webkit.org/show_bug.cgi?id=216464
<rdar://problem/68811378>
Reviewed by Sam Weinig.
Our WebGL2 implementation is in good enough shape that we should
enable it by default for broader testing.
Remove "defaultWebGL2Enabled" since it wasn't being used anywhere.
* Shared/WebPreferences.yaml:
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultWebGL2Enabled): Deleted.
* Shared/WebPreferencesDefaultValues.h:
2020-09-14 Youenn Fablet <youenn@apple.com>
RemoteAudioMediaStreamTrackRenderer should process its IPC messages from a background thread
https://bugs.webkit.org/show_bug.cgi?id=216474
Reviewed by Eric Carlson.
Register RemoteAudioMediaStreamTrackRendererManager as a thread message receiver
for both RemoteAudioMediaStreamTrackRenderer and RemoteAudioMediaStreamTrackRendererManager messages.
All processing is sent to a high priority work queue.
Given AudioMediaStreamTrackRendererUnit is not fully thread safe, we use a single WorkQueue for the whole GPUProcess.
All operations will happen in this thread, while in WebProcess, create/start/stop/release are main thread.
We no longer compile the generated RemoteAudioMediaStreamTrackRendererManager::didReceiveMessage
and instead implement our own version to handle manager and renderer messages.
Covered by WebRTC audio tests run with GPU process enabled.
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::GPUConnectionToWebProcess):
(WebKit::GPUConnectionToWebProcess::~GPUConnectionToWebProcess):
(WebKit::GPUConnectionToWebProcess::dispatchMessage):
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/GPUProcess.cpp:
(WebKit::GPUProcess::audioMediaStreamTrackRendererQueue):
* GPUProcess/GPUProcess.h:
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRenderer.h:
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRendererManager.cpp:
(WebKit::RemoteAudioMediaStreamTrackRendererManager::RemoteAudioMediaStreamTrackRendererManager):
(WebKit::RemoteAudioMediaStreamTrackRendererManager::~RemoteAudioMediaStreamTrackRendererManager):
(WebKit::RemoteAudioMediaStreamTrackRendererManager::close):
(WebKit::RemoteAudioMediaStreamTrackRendererManager::dispatchToThread):
(WebKit::RemoteAudioMediaStreamTrackRendererManager::didReceiveMessage):
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRendererManager.h:
* SourcesCocoa.txt:
2020-09-13 Simon Fraser <simon.fraser@apple.com>
Overflow:scroll rubberbanding is interrupted by post-layout scrolling
https://bugs.webkit.org/show_bug.cgi?id=216463
<rdar://problem/67095741>
Reviewed by Darin Adler.
When rubberbanding overflow:scroll RenderLayer has an overscrolled scroll offset.
If RenderLayer::updateScrollInfoAfterLayout() happens when in this state, it can
clamp the scroll offset, causing the rubberband to collapse which interferes with
the user interaction. This happend on Gmail when composing a reply.
Fix by tracking the rubberbanding state in the scrolling tree, and having RenderLayer
query this state via the ScrollingCoordinator. RenderLayer::updateScrollInfoAfterLayout()
already tested isRubberBandInProgress(). This is similar to how isUserScrollInProgress()
and isScrollSnapInProgress() work.
This patch also fixes the tracking of rubberbanding state. Previously setMainFrameIsRubberBanding()
was just based on when the timer was started and stopped, which did not match the
implementation of ScrollController::isRubberBandInProgress(). Now ScrollController
correctly notifies its clients when the rubberbanding state changes by updating that
state whenever any of the conditions consulted in isRubberBandInProgressInternal() change.
* WebProcess/WebPage/RemoteLayerTree/RemoteScrollingCoordinator.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteScrollingCoordinator.mm:
(WebKit::RemoteScrollingCoordinator::isRubberBandInProgress const):
(WebKit::RemoteScrollingCoordinator::scrollingStateInUIProcessChanged):
2020-09-13 Wenson Hsieh <wenson_hsieh@apple.com>
Add a key to the text manipulation userInfo dictionary indicating whether the translated item is on-screen
https://bugs.webkit.org/show_bug.cgi?id=216452
<rdar://problem/68785397>
Reviewed by Darin Adler.
Add `_WKTextManipulationTokenUserInfoVisibilityKey` and set its value to the value of the `isVisible` member in
`ManipulationTokenInfo`. See WebCore ChangeLog for more details.
* UIProcess/API/Cocoa/WKWebView.mm:
(createUserInfo):
* UIProcess/API/Cocoa/_WKTextManipulationToken.h:
* UIProcess/API/Cocoa/_WKTextManipulationToken.mm:
2020-09-13 Pablo Saavedra <psaavedra@igalia.com>
[GTK][WPE] Build fails when -DENABLE_SERVICE_WORKER=OFF
https://bugs.webkit.org/show_bug.cgi?id=216455
Unreviewed build fix.
* UIProcess/API/glib/WebKitWebsiteData.cpp:
(recordContainsSupportedDataTypes):
(toWebKitWebsiteDataTypes):
* UIProcess/API/glib/WebKitWebsiteDataManager.cpp:
(toWebsiteDataTypes):
* WebProcess/Network/WebLoaderStrategy.cpp:
(WebKit::WebLoaderStrategy::scheduleLoad):
2020-09-11 Alex Christensen <achristensen@webkit.org>
Move _setCookieAcceptPolicy to WKHTTPCookieStore
https://bugs.webkit.org/show_bug.cgi?id=216375
Reviewed by Geoffrey Garen.
It is only used by Safari, which I am changing along with this in rdar://problem/68659937
I need it to not be on the process pool for https://bugs.webkit.org/show_bug.cgi?id=216041
and because the cookie store's job is to manage cookie storage. The fact that it was on the process pool
is just a historical artifact of us trying to make it global.
WKContextUseTestingNetworkSession has had its important parts replaced by WKWebsiteDataStoreConfigurationSetTestingSessionEnabled
but it was still being used to determinte whether the UI process should persist the cookie accept policy. We don't need this
any more because we send a message to the network process and let it manage the cookie accept policy.
* UIProcess/API/C/WKContext.cpp:
(WKContextUseTestingNetworkSession): Deleted.
* UIProcess/API/C/WKContextPrivate.h:
* UIProcess/API/Cocoa/WKHTTPCookieStore.mm:
(-[WKHTTPCookieStore _setCookieAcceptPolicy:completionHandler:]):
* UIProcess/API/Cocoa/WKHTTPCookieStorePrivate.h:
* UIProcess/API/Cocoa/WKProcessPool.mm:
(-[WKProcessPool _setCookieAcceptPolicy:]):
* UIProcess/API/Cocoa/WKProcessPoolPrivate.h:
* UIProcess/WebCookieManagerProxy.cpp:
(WebKit::WebCookieManagerProxy::setHTTPCookieAcceptPolicy):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::useTestingNetworkSession): Deleted.
* UIProcess/WebProcessPool.h:
2020-09-11 Ben Nham <nham@apple.com>
Fix undeclared identifier issue due to clashes in soft link headers
https://bugs.webkit.org/show_bug.cgi?id=216412
Reviewed by Geoffrey Garen.
Use the workaround constant kPALFigPhotoContainerFormat_JFIF instead of the enum value
kFigPhotoContainerFormat_JFIF.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-09-11 Jer Noble <jer.noble@apple.com>
[iOS] Add Experimental setting to disable calling PID Inheritance API.
https://bugs.webkit.org/show_bug.cgi?id=216387
<rdar://problem/66286777>
Reviewed by Eric Carlson.
* Shared/WebPreferences.yaml:
2020-09-11 Kate Cheney <katherine_cheney@apple.com>
Followup to Bug 215027: address comments to improve APP_BOUND_DOMAINS macro use
https://bugs.webkit.org/show_bug.cgi?id=216373
<rdar://problem/68645704>
Reviewed by Darin Adler.
Use a local variable to avoid awkward #if ENABLE() macro placement.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::swServerForSession):
2020-09-10 Joonghun Park <jh718.park@samsung.com>
Unreviewed. Remove the build warning below since r266885.
warning: redundant move in return statement [-Wredundant-move]
Because return statement already returns rvalue reference,
we don't need WTFMove at return.
* UIProcess/Automation/WebAutomationSession.cpp:
(WebKit::WebAutomationSession::getSessionPermissions):
2020-09-10 Ben Nham <nham@apple.com>
Disable hardware JPEG decoding on x86 Mac
https://bugs.webkit.org/show_bug.cgi?id=216377
Reviewed by Geoff Garen.
In Big Sur, ImageIO uses hardware JPEG decoding automatically for certain JPEGs. This caused
a small regression in our battery life benchmark since the one-time setup cost for the
decoder (~20-30 ms per process) was higher than the per-decode win we got from the HW decoder.
For now, we're reverting to the Catalina behavior of not using hardware JPEG on x86 Macs by
calling a MediaToolbox SPI. (The tradeoffs for Apple Silicon are different so we use
ImageIO's default heuristics on that architecture.) We plan to revisit this decision once we
move to a GPUProcess world.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-09-10 Chris Dumez <cdumez@apple.com>
Unreviewed, reverting r266842 and r266883.
Causes some assertions to be hit in debug builds
Reverted changesets:
"Some WebAudio tests give different output on different
machines"
https://bugs.webkit.org/show_bug.cgi?id=216371
https://trac.webkit.org/changeset/266842
"Unreviewed Windows build fix after r266842."
https://trac.webkit.org/changeset/266883
2020-09-09 Brian Burg <bburg@apple.com>
Web Inspector: _WKInspectorDelegate should be attached to _WKInspector not WKWebView
https://bugs.webkit.org/show_bug.cgi?id=215961
Reviewed by Devin Rousso.
Move _WKInspectorDelegate to be a property of _WKInspector. If there
is no _WKInspector available (i.e. for iOS), the delegate cannot be set.
As part of this change, move -didAttachLocalInspector: back to UIDelegatePrivate.
This delegate method is called just as the inspector is shown, so there would
be no way to act upon this message if it were part of _WKInspectorDelegate.
Now, -didAttachLocalInspector: is the designated time to attach _WKInspectorDelegate.
Covered by existing API tests.
* UIProcess/API/APIInspectorClient.h:
(API::InspectorClient::browserDomainEnabled):
(API::InspectorClient::browserDomainDisabled):
(API::InspectorClient::didAttachLocalInspector): Deleted.
* UIProcess/API/APIUIClient.h:
(API::UIClient::didAttachLocalInspector):
* UIProcess/API/Cocoa/WKUIDelegatePrivate.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _initializeWithConfiguration:]):
(-[WKWebView _inspectorDelegate]): Deleted.
(-[WKWebView _setInspectorDelegate:]): Deleted.
* UIProcess/API/Cocoa/WKWebViewInternal.h:
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/API/Cocoa/_WKInspector.h:
* UIProcess/API/Cocoa/_WKInspector.mm:
(-[_WKInspector dealloc]): Added.
(-[_WKInspector delegate]): Added.
(-[_WKInspector setDelegate:]): Added. Since _WKInspector/WebInspectorProxy is always
created internally instead of via ObjC initializer, we need to initialize ivars before
using them, as there is no other place for them to be initialized.
* UIProcess/API/Cocoa/_WKInspectorDelegate.h:
* UIProcess/API/Cocoa/_WKInspectorInternal.h:
* UIProcess/Cocoa/UIDelegate.h:
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::setDelegate):
(WebKit::UIDelegate::UIClient::didAttachLocalInspector):
* UIProcess/Inspector/Agents/InspectorBrowserAgent.cpp:
(WebKit::InspectorBrowserAgent::enable):
(WebKit::InspectorBrowserAgent::disable):
* UIProcess/Inspector/Cocoa/InspectorDelegate.h:
* UIProcess/Inspector/Cocoa/InspectorDelegate.mm:
(WebKit::InspectorDelegate::InspectorDelegate):
(WebKit::InspectorDelegate::setDelegate):
(WebKit::InspectorDelegate::InspectorClient::browserDomainEnabled):
(WebKit::InspectorDelegate::InspectorClient::browserDomainDisabled):
(WebKit::InspectorDelegate::InspectorClient::didAttachLocalInspector): Deleted.
* UIProcess/Inspector/WebInspectorProxy.h:
(WebKit::WebInspectorProxy::inspectorClient):
* UIProcess/Inspector/WebInspectorProxy.cpp:
(WebKit::WebInspectorProxy::WebInspectorProxy):
(WebKit::WebInspectorProxy::setInspectorClient):
(WebKit::WebInspectorProxy::openLocalInspectorFrontend):
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::inspectorClient): Deleted.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::WebPageProxy):
(WebKit::WebPageProxy::setInspectorClient): Deleted.
Similar to the Cocoa SPI, move InspectorClient from WebPageProxy
to WebInspectorProxy.
2020-09-10 Devin Rousso <drousso@apple.com>
Web Inspector: modernize generated backend protocol code
https://bugs.webkit.org/show_bug.cgi?id=216302
<rdar://problem/68547649>
Reviewed by Brian Burg.
Previously, the inspector protocol was expressed in code in a somewhat confusing way:
- the error string was the first argument
- required parameters were `T` or `const T&`
- optional parameters were `const T*`
- enum parameters were the underlying type requiring the backend dispatcher handler to
process it instead of it being preprocessed
- required returns were `T&`
- optional returns were `T*`
This doesn't really make for easy/obvious reading of code since the order of arguments is
not weird (e.g. error string first), and that there are references/pointers to primitive
types.
This patch cleans up the generated inspector protocol code to be:
- required parameters are `T` or `Ref<T>&&`
- optional parameters are `Optional<T>&&` or `RefPtr<T>&&`
- enum parameters are preprocessed and passed to the backend dispatcher handler if valid
- synchronous commands return `Expected<X, ErrorString>` using the same types/rules above
where `X` is either a single return or a `std::tuple` of multiple returns
The one exception to the above is `String`, which is already a tri-state of `nullString()`,
`emptyString()`, and something set, so there's no need to use `Optional<String>`.
Also use `Protocol` objects/`typedefs` wherever possible to further relate the protocol
JSON and the actual backend dispatcher handler implementation.
* UIProcess/Automation/Automation.json:
`CoordinateSystem` has `Page` and `Viewport` enum values, but `WebAutomationSession` checks
for `"Page"` and `"LayoutViewport"`. Add a `LayoutViewport` enum value now that enums are
processed before being passed to backend dispacher handlers to preserve functionality.
* UIProcess/Inspector/Agents/InspectorBrowserAgent.h:
* UIProcess/Inspector/Agents/InspectorBrowserAgent.cpp:
* UIProcess/Automation/WebAutomationSessionMacros.h:
* UIProcess/Automation/WebAutomationSession.h:
* UIProcess/Automation/WebAutomationSession.cpp:
* UIProcess/Automation/mac/WebAutomationSessionMac.mm:
Elided backend dispatcher handler changes describe above.
* UIProcess/Inspector/socket/RemoteInspectorClient.cpp:
(WebKit::RemoteInspectorClient::setTargetList):
`JSON` classes now use `Ref&&` wherever possible and `Optional` instead of an out parameter
for `get*`/`as*` so that values can be more easily manipulated and can be confidently known
to exist.
2020-09-10 Chris Dumez <cdumez@apple.com>
Unreviewed Windows build fix after r266842.
* WebProcess/InjectedBundle/InjectedBundle.cpp:
(WebKit::InjectedBundle::setHardwareSampleRateOverride):
2020-09-10 Sihui Liu <sihui_liu@apple.com>
REGRESSION(r266634): [macOS release] 4 layout tests became flaky failures
https://bugs.webkit.org/show_bug.cgi?id=216275
<rdar://problem/68515242>
Reviewed by Tim Horton.
After r266634, activity state changes caused by adding/removing view will not be dispatched to web process until
transaction is committed in UI process. To make sure web process picks up the changes, we need to invoke the
callbacks after activity state changes are dispatched.
* UIProcess/API/Cocoa/WKWebViewPrivateForTesting.h:
* UIProcess/API/Cocoa/WKWebViewTesting.mm:
(-[WKWebView _doAfterActivityStateUpdate:]):
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::addActivityStateUpdateCompletionHandler):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::dispatchActivityStateChange):
* UIProcess/WebPageProxy.h:
2020-09-10 Chris Dumez <cdumez@apple.com>
Some WebAudio tests give different output on different machines
https://bugs.webkit.org/show_bug.cgi?id=216371
Reviewed by Alex Christensen.
Add InjectedBundle SPI to set the hardware sample rate.
* WebProcess/InjectedBundle/API/c/WKBundle.cpp:
(WKBundleSetHardwareSampleRateOverride):
* WebProcess/InjectedBundle/API/c/WKBundlePrivate.h:
* WebProcess/InjectedBundle/InjectedBundle.cpp:
(WebKit::InjectedBundle::setHardwareSampleRateOverride):
* WebProcess/InjectedBundle/InjectedBundle.h:
2020-09-10 Sihui Liu <sihui_liu@apple.com>
REGRESSION(r266634): fast/animation/request-animation-frame-throttling-lowPowerMode.html is flaky
https://bugs.webkit.org/show_bug.cgi?id=216261
<rdar://problem/68513429>
Reviewed by Tim Horton.
r266634 makes a behavior change that when there is an activity state change, TiledCoreAnimationDrawingArea
schedules a rendering update. This breaks timed rendering update from ScriptedAnimationController in the
test. If rendering update happens before timer fires, the callback of ScriptedAnimationController will be
invoked earlier than expected, thus more callbacks executed in the same time frame.
To fix this, let's only schedule rendering update when we need to handle activity state change callbacks,
as before r266634, where we don't schedule activity state timer if there is no callback.
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::activityStateDidChange):
2020-09-10 Tim Horton <timothy_horton@apple.com>
Upstream additional linked-on-or-after version checks
https://bugs.webkit.org/show_bug.cgi?id=216365
Reviewed by Sam Weinig.
* UIProcess/Cocoa/VersionChecks.h:
2020-09-10 Adrian Perez de Castro <aperez@igalia.com>
[CMake] Use imported targets in find module for libseccomp
https://bugs.webkit.org/show_bug.cgi?id=216329
Reviewed by Don Olmstead.
* PlatformGTK.cmake: Remove unneeded usage of LIBSECCOMP_INCLUDE_DIRS.
* PlatformWPE.cmake: Take the Libseccomp::Libseccomp imported target into use.
2020-09-10 Aditya Keerthi <akeerthi@apple.com>
[macOS] Add editability to input type=datetime-local
https://bugs.webkit.org/show_bug.cgi?id=216311
Reviewed by Devin Rousso.
* UIProcess/mac/WebDateTimePickerMac.mm:
(-[WKDateTimePicker initWithParams:inView:]):
Move initialization that relies on DateTimeChooserParameters to
updatePicker: to reduce code duplication.
(-[WKDateTimePicker updatePicker:]):
(-[WKDateTimePicker dateFormatStringForType:value:]):
datetime-local inputs can have three different formats. The default
contains hour and minute fields, but second and millisecond fields
may also be present. Update this method to return the appropriate
format string based on the current value of the input.
(-[WKDateTimePicker initialDateForEmptyValue]):
The initial value displayed on the picker when the input's value is
empty should match the user's system time. Since the date picker
maintains UTC time, return [NSDate date] with an appropriate offset.
2020-09-10 Kate Cheney <katherine_cheney@apple.com>
Clean up App-Bound Domains code to only compile for iOS with its own macro
https://bugs.webkit.org/show_bug.cgi?id=215027
<rdar://problem/63688232>
Reviewed by Darin Adler.
App-bound domains code should only be compiled on iOS, so this patch
adds additional '#if ENABLE()' checks to ensure that. Additionally,
for clarity this patch creates a new APP_BOUND_DOMAINS macro and
replaces existing IOS_FAMILY checks for App-Bound Domains code with
this new macro.
LoadParameters, PolicyDecision and WKPreferences will be in another
patch, this one is huge already.
No new tests, behavior should not change so testing is covered by
all existing App-Bound Domains layout and API tests.
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.cpp:
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.h:
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::resetParametersToDefaultValues):
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.h:
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::swServerForSession):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/NetworkSession.h:
* NetworkProcess/cocoa/NetworkSessionCocoa.h:
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::NetworkSessionCocoa):
(WebKit::NetworkSessionCocoa::sessionWrapperForTask):
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
* UIProcess/API/APIHTTPCookieStore.cpp:
(API::HTTPCookieStore::filterAppBoundCookies):
* UIProcess/API/APIPageConfiguration.cpp:
(API::PageConfiguration::copy const):
* UIProcess/API/APIPageConfiguration.h:
(API::PageConfiguration::ignoresAppBoundDomains const):
(API::PageConfiguration::setIgnoresAppBoundDomains):
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreHasAppBoundSession):
(WKWebsiteDataStoreSetAppBoundDomainsForTesting):
(WKWebsiteDataStoreClearAppBoundSession):
(WKWebsiteDataStoreReinitializeAppBoundDomains):
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _isNavigatingToAppBoundDomain:]): Deleted.
(-[WKWebView _isForcedIntoAppBoundMode:]): Deleted.
Moved to WKWebViewIOS.
* UIProcess/API/Cocoa/WKWebViewConfiguration.h:
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration limitsNavigationsToAppBoundDomains]):
(-[WKWebViewConfiguration setLimitsNavigationsToAppBoundDomains:]):
(-[WKWebViewConfiguration _ignoresAppBoundDomains]):
(-[WKWebViewConfiguration _setIgnoresAppBoundDomains:]):
* UIProcess/API/Cocoa/WKWebViewConfigurationPrivate.h:
Moved App-Bound Domain functions to the part of the file that only compiles
on iOS and removed macos tags for WK_API_AVAILABLE, because both the
API and SPI are iOS only.
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
Moved App-Bound Domain functions to the part of the file that only compiles on iOS.
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
(-[WKWebsiteDataStore _appBoundDomains:]):
(-[WKWebsiteDataStore _appBoundSchemes:]):
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _isNavigatingToAppBoundDomain:]):
(-[WKWebView _isForcedIntoAppBoundMode:]):
Moved these from WKWebView.
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::getAppBoundDomains):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/Network/NetworkProcessProxy.messages.in:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::loadData):
(WebKit::shouldTreatURLProtocolAsAppBound):
(WebKit::WebPageProxy::setIsNavigatingToAppBoundDomainAndCheckIfPermitted):
(WebKit::WebPageProxy::disableServiceWorkerEntitlementInNetworkProcess):
(WebKit::WebPageProxy::clearServiceWorkerEntitlementOverride):
(WebKit::WebPageProxy::preconnectTo):
Now that m_isNavigatingToAppBoundDomain is iOS only, I changed WebPageProxy::preconnectTo
to call the isNavigatingToAppBoundDomain() function, which is available
on all platforms but returns WTF::nullopt unless the platform is iOS.
(WebKit::WebPageProxy::decidePolicyForNavigationAction):
(WebKit::WebPageProxy::creationParameters):
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::isNavigatingToAppBoundDomain const):
isNavigatingToAppBoundDomain() is used very frequently, it is cleaner
to have two functions -- one for iOS which returns the actual value
and one for non-iOS platforms which returns WTF::nullopt -- rather
than have #if ENABLE(APP_BOUND_DOMAIN) checks everywhere.
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::download):
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::appBoundDomainQueue):
(WebKit::WebsiteDataStore::platformInitialize):
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::parameters):
(WebKit::WebsiteDataStore::renameOriginInWebsiteData):
* UIProcess/WebsiteData/WebsiteDataStore.h:
* WebProcess/Network/WebLoaderStrategy.cpp:
(WebKit::WebLoaderStrategy::scheduleLoadFromNetworkProcess):
(WebKit::WebLoaderStrategy::loadResourceSynchronously):
(WebKit::WebLoaderStrategy::startPingLoad):
(WebKit::WebLoaderStrategy::preconnectTo):
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::sendH2Ping):
* WebProcess/WebCoreSupport/WebFrameLoaderClient.h:
* WebProcess/WebPage/WebFrame.cpp:
* WebProcess/WebPage/WebFrame.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
(WebKit::WebPage::loadRequest):
(WebKit::WebPage::loadDataImpl):
(WebKit::WebPage::didReceivePolicyDecision):
(WebKit::WebPage::runJavaScript):
(WebKit::WebPage::updatePreferences):
Only check for the m_limitsNavigationsToAppBoundDomains flag if the
platform is iOS. Otherwise it should always be false, because it will
never be enabled on a non-iOS platform.
* WebProcess/WebPage/WebPage.h:
2020-09-09 Tim Horton <timothy_horton@apple.com>
macCatalyst: Form controls behave strangely (like iPhone) in macOS-idiom apps
https://bugs.webkit.org/show_bug.cgi?id=216326
Reviewed by Wenson Hsieh.
* Shared/UserInterfaceIdiom.h:
* Shared/UserInterfaceIdiom.mm:
(WebKit::userInterfaceIdiomIsPad):
(WebKit::currentUserInterfaceIdiomIsPadOrMac):
(WebKit::setCurrentUserInterfaceIdiomIsPadOrMac):
(WebKit::currentUserInterfaceIdiomIsPad): Deleted.
(WebKit::setCurrentUserInterfaceIdiomIsPad): Deleted.
* Shared/ios/WebPreferencesDefaultValuesIOS.mm:
(WebKit::defaultTextAutosizingUsesIdempotentMode):
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration init]):
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
* UIProcess/ios/SmartMagnificationController.mm:
(WebKit::SmartMagnificationController::didCollectGeometryForSmartMagnificationGesture):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKFormInputSession setAccessoryViewCustomButtonTitle:]):
(-[WKContentView endEditingAndUpdateFocusAppearanceWithReason:]):
(-[WKContentView _shouldShowAutomaticKeyboardUIIgnoringInputMode]):
(-[WKContentView _zoomToRevealFocusedElement]):
(-[WKContentView requiresAccessoryView]):
(-[WKContentView _updateAccessory]):
(shouldShowKeyboardForElement):
(-[WKContentView _shouldUseLegacySelectPopoverDismissalBehavior]):
* UIProcess/ios/WebDataListSuggestionsDropdownIOS.mm:
(WebKit::WebDataListSuggestionsDropdownIOS::show):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::desktopClassBrowsingSupported):
* UIProcess/ios/forms/WKAirPlayRoutePicker.mm:
(-[WKAirPlayRoutePicker show:fromRect:]):
* UIProcess/ios/forms/WKFileUploadPanel.mm:
(-[WKFileUploadPanel _showPhotoPickerWithSourceType:]):
(-[WKFileUploadPanel _presentMenuOptionForCurrentInterfaceIdiom:]):
* UIProcess/ios/forms/WKFormColorControl.mm:
(-[WKFormColorControl initWithView:]):
* UIProcess/ios/forms/WKFormColorPicker.mm:
(-[WKColorPicker initWithView:inPopover:]):
(-[WKColorPicker drawSelectionIndicatorForColorButton:]):
* UIProcess/ios/forms/WKFormSelectControl.mm:
(-[WKFormSelectControl initWithView:]):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
Rename currentUserInterfaceIdiomIsPad to currentUserInterfaceIdiomIsPadOrMac,
and force it to YES on macCatalyst since internally we use it to distinguish
between iPhone and iPad behavior, and (for our behaviors, at least) macCatalyst
should always follow iPad.
We should clean this up, and make all callers make their
iPhone vs. iPad vs. macOS decisions more explicit.
2020-09-09 Wenson Hsieh <wenson_hsieh@apple.com>
Text copied and pasted from Mac Catalyst apps appears larger than expected
https://bugs.webkit.org/show_bug.cgi?id=215971
<rdar://problem/65768907>
Reviewed by Tim Horton.
Various pieces of platform logic in Mac Catalyst depend on the user interface idiom (i.e. Mac or iPad) of the
app, as well as the application-wide "scale factor". In the context of this bug, NSAttributedString to RTF data
conversion methods in `UIFoundation` consult the scale factor of `UIiOSMacIdiomManager` to determine whether to
emit 0 (standard) or 1 (iOS) for the `\cocoatextscaling` attribute. The fact that the web process' scale factor
may be out of sync with the UI process' scale factor leads to copied RTF data in the web process appearing
either larger or smaller than expected, due to an incorrect `NSTextScalingType` value.
To mitigate this (as well as any other issues), we add a mechanism for the UI process to forward the global
scale factor and user interface idiom to the web process inside of the web process creation parameters, and then
use new UIKit SPI to override the scale factor and idiom within the web process.
* Scripts/process-entitlements.sh:
Add a new entitlement needed to use `_UIApplicationCatalystRequestViewServiceIdiomAndScaleFactor`.
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
Use `_UIApplicationCatalystUserInterfaceIdiom` and `_UIApplicationCatalystScaleFactor` to grab the global user
interface idiom and scale factor, respectively.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
Override the idiom and scale factor in the web process with the idiom and scale factor from the UI process,
using the new SPI `_UIApplicationCatalystRequestViewServiceIdiomAndScaleFactor`. As this function hasn't landed
yet, we soft link the function for now to avoid causing the web process to instantly crash on any builds without
the fix for <rdar://problem/68524148>.
2020-09-09 Brady Eidson <beidson@apple.com>
Make sure WKWebsiteDataStore operations reuse existing process pools even when all WKWebViews have closed.
<rdar://problem/62978295> and https://bugs.webkit.org/show_bug.cgi?id=216317
Reviewed by Geoffrey Garen.
Covered by new API test.
When WebsiteDataStores are gathering all the NetworkProcesses they might need to message, they miss some
obvious candidates if there are no longer any related WKWebViews.
Fix that by tracking which sessions a NetworkProcess knows about.
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::addSession):
(WebKit::NetworkProcessProxy::hasSession const):
(WebKit::NetworkProcessProxy::removeSession):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::isAssociatedProcessPool const):
2020-09-09 Chris Dumez <cdumez@apple.com>
Move lazy DisplayLink tear down logic from the WebProcess to the UIProcess
https://bugs.webkit.org/show_bug.cgi?id=216195
Reviewed by Simon Fraser.
Move lazy DisplayLink tear down logic from the WebProcess to the UIProcess, now that the
DisplayLink has been moved to the UIProcess due to sandboxing.
After a DisplayLink no longer has any clients, we keep it firing up to 20 times without
any clients in case a new client gets added shortly after. The idea was to avoid killing
and respawning too many threads when adding and removing clients in quick succession.
However, now that the DisplayLink lives in the UIProcess side and sends IPC to the
WebProcesses every time it fires, it makes a lot more sense to implement this logic in
the UIProcess side, to avoid sending unnecessary IPC to processes that do not care about
it.
* UIProcess/mac/DisplayLink.cpp:
(WebKit::DisplayLink::addObserver):
(WebKit::DisplayLink::removeObserver):
(WebKit::DisplayLink::removeObservers):
(WebKit::DisplayLink::displayLinkCallback):
(WebKit::DisplayLink::hasObservers const): Deleted.
* UIProcess/mac/DisplayLink.h:
2020-09-08 Chris Dumez <cdumez@apple.com>
Regression(r260614) Power usage has increased due to extra thread hopping
https://bugs.webkit.org/show_bug.cgi?id=216296
<rdar://problem/67719299>
Reviewed by Simon Fraser.
Power usage has increased after r260614 due to extra thread hopping. To recover,
we now process the DisplayWasRefreshed IPC to a background queue if there is
scrolling going on and responsiveness is thus critical. In the common case, where
the user is not scrolling, we keep processing the IPC on the main thread, like
before r260614. This avoids extra thread hopping and saves power.
* UIProcess/WebPageProxy.cpp:
(WebKit::ScrollingObserver::willSendWheelEvent):
(WebKit::ScrollingObserver::ScrollingObserver):
(WebKit::ScrollingObserver::singleton):
(WebKit::WebPageProxy::sendWheelEvent):
* UIProcess/mac/DisplayLink.cpp:
(WebKit::DisplayLink::displayLinkCallback):
* UIProcess/mac/DisplayLink.h:
(WebKit::DisplayLink::setShouldSendIPCOnBackgroundQueue):
* WebProcess/WebPage/EventDispatcher.cpp:
(WebKit::EventDispatcher::notifyScrollingTreesDisplayWasRefreshed):
(WebKit::EventDispatcher::displayWasRefreshed):
* WebProcess/WebPage/EventDispatcher.h:
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::displayWasRefreshed):
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
2020-09-08 Megan Gardner <megan_gardner@apple.com>
Removed unneeded respondsToSelector checks on UIWKTextInteractionAssistant in WKContentViewInteraction.
https://bugs.webkit.org/show_bug.cgi?id=216204
Reviewed by Wenson Hsieh.
Over the years, we have had many selection and text interaction assistants, that may not have all
responded to the selectors we expect them to. This has left many respondsToSelector checks around that
should be removed, as they are unnecessary and an unneeded performance hit.
No behavior change, covered by existing tests.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _willStartScrollingOrZooming]):
(-[WKContentView _didEndScrollingOrZooming]):
2020-09-08 Lauro Moura <lmoura@igalia.com>
NetworkProcess: Check for invalid handle before writing salt file
https://bugs.webkit.org/show_bug.cgi?id=216285
Reviewed by Sihui Liu.
Covered by existing tests.
Fixes crashes in GTK/WPE after r266742.
* NetworkProcess/cache/NetworkCacheData.cpp:
(WebKit::NetworkCache::readOrMakeSalt):
2020-09-08 Brady Eidson <beidson@apple.com>
Crashtracer inside PDFPlugin::createScrollbar.
<rdar://problem/68474062> and https://bugs.webkit.org/show_bug.cgi?id=216286
Reviewed by Tim Horton.
To quote Tim from r264945:
No new tests; timing is such that I can't reproduce without inserting
intentional delays into the main thread hops, which is further than
I'm willing to go for a test.
This is a speculative fix due to the aforementioned reproducibility issue.
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::createScrollbar): Plugin hasn't been destroyed, and there's still a PluginView,
but the PluginView isn't in a frame. So null check that.
2020-09-08 Alex Christensen <achristensen@webkit.org>
Null check m_layerHostingContext in TiledCoreAnimationDrawingArea::sendDidFirstLayerFlushIfNeeded
https://bugs.webkit.org/show_bug.cgi?id=216282
<rdar://problem/64933878>
Reviewed by Tim Horton.
There's no reason to dereference null here. Let's not crash.
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::sendDidFirstLayerFlushIfNeeded):
2020-09-08 Youenn Fablet <youenn@apple.com>
Tighten checks when creating an audio buffer list
https://bugs.webkit.org/show_bug.cgi?id=216237
<rdar://problem/68271376>
Reviewed by Geoffrey Garen.
Add message checks to verify that no message integer overflows happen when processing audio buffer list messages.
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::audioTrackRendererManager):
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRenderer.cpp:
(WebKit::RemoteAudioMediaStreamTrackRenderer::RemoteAudioMediaStreamTrackRenderer):
(WebKit::RemoteAudioMediaStreamTrackRenderer::audioSamplesStorageChanged):
(WebKit::RemoteAudioMediaStreamTrackRenderer::audioSamplesAvailable):
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRenderer.h:
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRendererManager.cpp:
(WebKit::RemoteAudioMediaStreamTrackRendererManager::RemoteAudioMediaStreamTrackRendererManager):
(WebKit::RemoteAudioMediaStreamTrackRendererManager::createRenderer):
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRendererManager.h:
* GPUProcess/webrtc/RemoteMediaRecorder.cpp:
(WebKit::RemoteMediaRecorder::audioSamplesStorageChanged):
(WebKit::RemoteMediaRecorder::audioSamplesAvailable):
* GPUProcess/webrtc/RemoteMediaRecorder.h:
* WebProcess/cocoa/RemoteCaptureSampleManager.cpp:
(WebKit::RemoteCaptureSampleManager::RemoteAudio::audioSamplesAvailable):
2020-09-08 Tim Horton <timothy_horton@apple.com>
iOS: <attachment>'s QuickLook thumbnails can appear squished
https://bugs.webkit.org/show_bug.cgi?id=216209
<rdar://problem/67817706>
Reviewed by Wenson Hsieh.
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::convertPlatformImageToBitmap):
Propagate an image of the same aspect ratio that QuickLook provided,
instead of squishing it to square.
* UIProcess/QuickLookThumbnailLoader.mm:
(-[WKQLThumbnailLoadOperation start]):
Only request full thumbnails; we do not want the icon form, since <attachment>
already has one without QuickLook's help; if we can't get a full thumbnail,
we'll just leave it alone.
2020-09-08 Sihui Liu <sihui_liu@apple.com>
Remove storage WorkQueue in NetworkProcess
https://bugs.webkit.org/show_bug.cgi?id=216199
Reviewed by Youenn Fablet.
The only meaningful usage of this WorkQueue is iterating IndexedDB directories and collecting origins, while
there can be some IDB thread accessing the same directories. To remove the race, Let's just drop this WorkQueue
and fetch origins via WebIDBServer.
* NetworkProcess/IndexedDB/WebIDBServer.cpp:
(WebKit::WebIDBServer::getOrigins):
* NetworkProcess/IndexedDB/WebIDBServer.h:
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::addSessionStorageQuotaManager):
(WebKit::NetworkProcess::fetchWebsiteData):
(WebKit::NetworkProcess::deleteAndRestrictWebsiteDataForRegistrableDomains):
(WebKit::NetworkProcess::registrableDomainsWithWebsiteData):
(WebKit::NetworkProcess::addIndexedDatabaseSession):
(WebKit::NetworkProcess::addServiceWorkerSession):
(WebKit::NetworkProcess::ensurePathExists): Deleted.
(WebKit::NetworkProcess::postStorageTask): Deleted.
(WebKit::NetworkProcess::performNextStorageTask): Deleted.
(WebKit::NetworkProcess::collectIndexedDatabaseOriginsForVersion): Deleted.
(WebKit::NetworkProcess::indexedDatabaseOrigins): Deleted.
* NetworkProcess/NetworkProcess.h:
2020-09-08 Peng Liu <peng.liu6@apple.com>
Clean up functions and state variables related to the picture-in-picture implementation
https://bugs.webkit.org/show_bug.cgi?id=215972
Reviewed by Jer Noble.
This patch adds an optional FloatSize parameter to the VideoFullscreenManager::DidEnterFullscreen message.
Also, it enables the picture-in-picture and fullscreen events when a video element enters picture-in-picture
from the fullscreen mode.
* UIProcess/Cocoa/VideoFullscreenManagerProxy.h:
* UIProcess/Cocoa/VideoFullscreenManagerProxy.mm:
(WebKit::VideoFullscreenModelContext::didEnterFullscreen):
(WebKit::VideoFullscreenModelContext::prepareToExitFullscreen):
(WebKit::VideoFullscreenModelContext::willEnterPictureInPicture):
(WebKit::VideoFullscreenModelContext::didEnterPictureInPicture):
(WebKit::VideoFullscreenModelContext::failedToEnterPictureInPicture):
(WebKit::VideoFullscreenModelContext::willExitPictureInPicture):
(WebKit::VideoFullscreenModelContext::didExitPictureInPicture):
(WebKit::VideoFullscreenManagerProxy::ensureClientForContext):
(WebKit::VideoFullscreenManagerProxy::setupFullscreenWithID):
(WebKit::VideoFullscreenManagerProxy::enterFullscreen):
(WebKit::VideoFullscreenManagerProxy::didEnterFullscreen):
Use ensureClientForContext() instead of addClientForContext() in setupFullscreenWithID()
to avoid leaking VideoFullscreenInterfaceAVKit instances.
Iterate a copy of the HashSet of VideoFullscreenModelClient pointers because a callback may
remove the client itself during the iteration.
* UIProcess/ios/fullscreen/WKFullScreenWindowControllerIOS.mm:
(-[WKFullScreenWindowController prepareToExitPictureInPicture]):
* UIProcess/mac/WKFullScreenWindowController.h:
* UIProcess/mac/WKFullScreenWindowController.mm:
(-[WKFullScreenWindowController didExitPictureInPicture]):
Fix leaking of VideoFullscreenInterfaceMac.
* WebProcess/cocoa/VideoFullscreenManager.h:
* WebProcess/cocoa/VideoFullscreenManager.messages.in:
* WebProcess/cocoa/VideoFullscreenManager.mm:
(WebKit::VideoFullscreenManager::fullscreenModeChanged):
We need to change the fullscreen mode in the VideoFullscreenInterfaceContext instance as well.
(WebKit::VideoFullscreenManager::didEnterFullscreen):
(WebKit::VideoFullscreenManager::didCleanupFullscreen):
Call didExitFullscreenOrPictureInPicture() instead of didStopBeingFullscreenElement(), which
is for the FullscreenManager.
2020-09-08 Mike Gorse <mgorse@suse.com>
[GTK] AcceleratedBackingStoreWayland might erroneously try to use SHM with older wpebackend-fdo
https://bugs.webkit.org/show_bug.cgi?id=216254
Reviewed by Adrian Perez de Castro.
* UIProcess/gtk/AcceleratedBackingStoreWayland.cpp:
(WebKit::tryInitializeSHM): Return false for wpebackend-fdo < 1.7.0.
2020-09-08 Pablo Saavedra <psaavedra@igalia.com>
[GLIB] RemoteInspectorServer is not started if WebKitWebContext is not created already created
https://bugs.webkit.org/show_bug.cgi?id=216120
Reviewed by Carlos Garcia Campos.
Added a WebKit/UIProcess/API/glib/WebKitInitialize what implements
its own webkitInitialize() to ensure the inspector server is
initialized as early as possible, before other api calls that would
depend on the inspector server being running.
The RemoteInspectorServer initialization is not longer
asociated to the initialization of the WebProcessPoolGLib.
The webkitInitialize() extends the shared InitializeWebKit2();
and also includes the initialization of the RemoteInspectorServer
so the initialization of the remote inspector server can be now
triggered by the initialization of many other API objects too:
WebKitInputMethodContext, WebKitSettings, WebKitUserContentManager,
WebKitWebContext, WebKitWebsiteDataManager,
WebKitUserContentFilterStore, WebKitWebViewBase.
* SourcesGTK.txt:
* SourcesWPE.txt:
* UIProcess/API/glib/WebKitInitialize.cpp: Added.
(WebKit::initializeRemoteInspectorServer):
(WebKit::webkitInitialize):
* UIProcess/API/glib/WebKitInitialize.h: Added.
* UIProcess/API/glib/WebKitInputMethodContext.cpp:
(webkit_input_method_context_class_init):
* UIProcess/API/glib/WebKitSettings.cpp:
(webkit_settings_class_init):
* UIProcess/API/glib/WebKitUserContentFilterStore.cpp:
(webkit_user_content_filter_store_class_init):
* UIProcess/API/glib/WebKitUserContentManager.cpp:
* UIProcess/API/glib/WebKitWebContext.cpp:
(webkit_web_context_class_init):
* UIProcess/API/glib/WebKitWebsiteDataManager.cpp:
(webkit_website_data_manager_class_init):
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkit_web_view_base_class_init):
* UIProcess/glib/WebProcessPoolGLib.cpp:
(WebKit::WebProcessPool::platformInitialize):
(WebKit::initializeRemoteInspectorServer): Deleted.
2020-09-07 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r266645.
https://bugs.webkit.org/show_bug.cgi?id=216251
Caused MotionMark regression
Reverted changeset:
"Move lazy DisplayLink tear down logic from the WebProcess to
the UIProcess"
https://bugs.webkit.org/show_bug.cgi?id=216195
https://trac.webkit.org/changeset/266645
2020-09-07 Mike Gorse <mgorse@suse.com>
Build failure; cannot find seccomp.h
https://bugs.webkit.org/show_bug.cgi?id=216226
Reviewed by Darin Adler.
No new tests (build fix only).
* PlatformGTK.cmake: Add LIBSECCOMP_INCLUDE_DIRS to
WebKit_SYSTEM_INCLUDE_DIRECTORIES.
2020-09-05 Wenson Hsieh <wenson_hsieh@apple.com>
[MotionMark] RenderLayer::paintLayerContents spends ~5% of the time in MonotonicTime::now() in Multiply
https://bugs.webkit.org/show_bug.cgi?id=216190
Reviewed by Darin Adler.
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::timestampForPaintFrequencyTracking const):
In WebKit2, we can assume (with the exception of SVG pages) that we must've performed a rendering update prior
to tracking painting frequencies. As such, we can use the page's rendering update timestamp instead of the real
current time (`MonotonicTime::now()`).
Note that in WebKit1, it is possible for any client to force a synchronous paint of the page before the page has
performed a rendering update, which triggers assertions in `SinglePaintFrequencyTracking::end()`. As such, we
stick with `MonotonicTime::now()` in WebKit1.
* WebProcess/WebCoreSupport/WebChromeClient.h:
2020-09-05 Sam Weinig <weinig@apple.com>
[WebIDL] Realign our IDL extended attribute names with those specified in WebIDL
https://bugs.webkit.org/show_bug.cgi?id=216194
Reviewed by Darin Adler.
* WebProcess/InjectedBundle/InjectedBundleScriptWorld.cpp:
(WebKit::InjectedBundleScriptWorld::disableOverrideBuiltinsBehavior):
Update for new name WebCore::DOMWrapperWorld.
2020-09-04 Alex Christensen <achristensen@webkit.org>
Move PDF heads-up display to UI process on macOS
https://bugs.webkit.org/show_bug.cgi?id=215780
<rdar://problem/58715847>
Reviewed by Tim Horton.
This has many benefits including the one in the radar, testability, and moving WebKit-specific code from PDFKit to WebKit.
Now that the HUD is displayed in the UI process instead of in the web content, zooming in keeps the HUD the same size.
A 3d-transformed iframe with a PDF in it now shows the HUD in the place you can click on it, instead of in the transformed PDF.
Also, moving the mouse to the HUD within the first 3 seconds prevents it from hiding, which it does when the mouse leaves the HUD.
The rest of the behavior I think should be the same. Most users should not notice any difference.
Functionality covered by API tests, the first tests for PDF content interaction!
* Scripts/webkit/messages.py:
* UIProcess/API/Cocoa/WKWebViewInternal.h:
* UIProcess/API/mac/WKView.mm:
(-[WKView _createPDFHUD:rect:]):
(-[WKView _pdfHUD:changedLocation:]):
(-[WKView _removePDFHUD:]):
(-[WKView _removeAllPDFHUDs]):
* UIProcess/API/mac/WKWebViewMac.mm:
(-[WKWebView mouseMoved:]):
(-[WKWebView mouseDown:]):
(-[WKWebView mouseUp:]):
(-[WKWebView _createPDFHUD:rect:]):
(-[WKWebView _pdfHUD:changedLocation:]):
(-[WKWebView _removePDFHUD:]):
(-[WKWebView _removeAllPDFHUDs]):
* UIProcess/API/mac/WKWebViewPrivateForTestingMac.h:
* UIProcess/API/mac/WKWebViewTestingMac.mm:
(-[WKWebView pdfHUDs]):
* UIProcess/Cocoa/WebViewImpl.h:
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::createPDFHUD):
(WebKit::WebViewImpl::updatePDFHUDLocation):
(WebKit::WebViewImpl::removePDFHUD):
(WebKit::WebViewImpl::removeAllPDFHUDs):
* UIProcess/PageClient.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didCommitLoadForFrame):
(WebKit::WebPageProxy::processDidTerminate):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::savePDFToTemporaryFolderAndOpenWithNativeApplication): Deleted.
(WebKit::WebPageProxy::openPDFFromTemporaryFolderWithNativeApplication): Deleted.
* UIProcess/mac/PageClientImplMac.h:
* UIProcess/mac/PageClientImplMac.mm:
(WebKit::PageClientImpl::createPDFHUD):
(WebKit::PageClientImpl::updatePDFHUDLocation):
(WebKit::PageClientImpl::removePDFHUD):
(WebKit::PageClientImpl::removeAllPDFHUDs):
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::savePDFToTemporaryFolderAndOpenWithNativeApplication):
(WebKit::WebPageProxy::createPDFHUD):
(WebKit::WebPageProxy::removePDFHUD):
(WebKit::WebPageProxy::updatePDFHUDLocation):
(WebKit::WebPageProxy::pdfZoomIn):
(WebKit::WebPageProxy::pdfZoomOut):
(WebKit::WebPageProxy::pdfSaveToPDF):
(WebKit::WebPageProxy::pdfOpenWithPreview):
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/Plugins/PDF/PDFLayerControllerSPI.h:
* WebProcess/Plugins/PDF/PDFPlugin.h:
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(-[WKPDFLayerControllerDelegate openWithNativeApplication]):
(-[WKPDFLayerControllerDelegate saveToPDF]):
(WebKit::PDFPlugin::PDFPlugin):
(WebKit::PDFPlugin::~PDFPlugin):
(WebKit::PDFPlugin::frameForHUD const):
(WebKit::PDFPlugin::calculateSizes):
(WebKit::PDFPlugin::convertFromPDFViewToRootView const):
(WebKit::PDFPlugin::visibilityDidChange):
(WebKit::PDFPlugin::zoomIn):
(WebKit::PDFPlugin::zoomOut):
(WebKit::PDFPlugin::save):
(WebKit::PDFPlugin::openWithPreview):
* WebProcess/WebPage/WebPage.cpp:
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::zoomPDFIn):
(WebKit::WebPage::zoomPDFOut):
(WebKit::WebPage::savePDF):
(WebKit::WebPage::openPDFWithPreview):
(WebKit::WebPage::createPDFHUD):
(WebKit::WebPage::updatePDFHUDLocation):
(WebKit::WebPage::removePDFHUD):
2020-09-04 Chris Dumez <cdumez@apple.com>
handleXPCEndpointMessages does not check XPC object is an xpc_endpoint_t
https://bugs.webkit.org/show_bug.cgi?id=216196
<rdar://problem/68198293>
Reviewed by Per Arne Vollan.
Add a check to have sure the XPC object is indeed an xpc_endpoint_t before
using it.
* WebProcess/cocoa/HandleXPCEndpointMessages.mm:
(WebKit::handleXPCEndpointMessages):
2020-09-04 Chris Dumez <cdumez@apple.com>
Move lazy DisplayLink tear down logic from the WebProcess to the UIProcess
https://bugs.webkit.org/show_bug.cgi?id=216195
Reviewed by Simon Fraser.
Move lazy DisplayLink tear down logic from the WebProcess to the UIProcess, now that the
DisplayLink has been moved to the UIProcess due to sandboxing.
After a DisplayLink no longer has any clients, we keep it firing up to 20 times without
any clients in case a new client gets added shortly after. The idea was to avoid killing
and respawning too many threads when adding and removing clients in quick succession.
However, now that the DisplayLink lives in the UIProcess side and sends IPC to the
WebProcesses every time it fires, it makes a lot more sense to implement this logic in
the UIProcess side, to avoid sending unnecessary IPC to processes that do not care about
it.
* UIProcess/mac/DisplayLink.cpp:
(WebKit::DisplayLink::addObserver):
(WebKit::DisplayLink::removeObserver):
(WebKit::DisplayLink::removeObservers):
(WebKit::DisplayLink::displayLinkCallback):
(WebKit::DisplayLink::hasObservers const): Deleted.
* UIProcess/mac/DisplayLink.h:
2020-09-04 Sihui Liu <sihui_liu@apple.com>
Webpages flash when switching between windows
https://bugs.webkit.org/show_bug.cgi?id=216012
Reviewed by Tim Horton.
Based on patches from Tim Horton. To avoid flash, we need to make sure view to be selected updates its content
before UI process commits its transaction. We did this by making UI process block until receiving a reply
DidUpdateActivityState from web process. However, web process did not make sure the reply would be sent after
normal rendering update and corresponding transasction commit. Instead, it flushed transactions in progress and
replied. To fix this, now we make web process reply in transaction commit handler after rendering update, if
there is an active transaction.
In the switching case, view to be unselected will detach from root layer in its web process, which makes its
content empty. This change is independent from the UI process commit, so we want this to happen after UI process
commits (which submits the view hierachy change). Otherwise, empty content(white flash) on the unselected view
will be displayed. To fix this, we let UI process send the activity state change message in transaction commit
handler, if there is an active transaction.
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::scheduleActivityStateUpdate):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::setSuppressVisibilityUpdates):
(WebKit::WebPageProxy::activityStateDidChange):
(WebKit::WebPageProxy::dispatchActivityStateChange):
* UIProcess/WebPageProxy.h:
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.h:
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::TiledCoreAnimationDrawingArea):
(WebKit::TiledCoreAnimationDrawingArea::updateRendering):
(WebKit::TiledCoreAnimationDrawingArea::handleActivityStateChangeCallbacks):
(WebKit::TiledCoreAnimationDrawingArea::handleActivityStateChangeCallbacksIfNeeded):
(WebKit::TiledCoreAnimationDrawingArea::activityStateDidChange):
(WebKit::TiledCoreAnimationDrawingArea::didUpdateActivityStateTimerFired): Deleted.
2020-09-04 Per Arne Vollan <pvollan@apple.com>
[macOS] Add client decoder entitlement
https://bugs.webkit.org/show_bug.cgi?id=216175
<rdar://problem/67532128>
Reviewed by Brent Fulgham.
Add client decoder entitlement in the WebContent process on macOS.
* Scripts/process-entitlements.sh:
2020-09-04 Youenn Fablet <youenn@apple.com>
MediaRecorder timeslice parameter causing internal error on longer videos
https://bugs.webkit.org/show_bug.cgi?id=216076
<rdar://problem/68209422>
Reviewed by Eric Carlson.
Test: http/wpt/fetch/blob-range.html
* NetworkProcess/NetworkDataTaskBlob.cpp:
(WebKit::NetworkDataTaskBlob::readData):
readData can be re-entrant so we need to reset m_currentItemReadSize before consuming data.
2020-09-04 Aditya Keerthi <akeerthi@apple.com>
REGRESSION (r241734): [iOS] Changing the value of a select element and tapping on another results in two changes
https://bugs.webkit.org/show_bug.cgi?id=216077
Reviewed by Darin Adler.
r241734 added a call to [self _endEditing] in
[WKContentView _elementDidBlur] to ensure that any open form controls
are dismissed when using the keyboard to change focus.
[self _endEditing] calls [_inputPeripheral endEditing], which
eventually calls [WKFormControl controlEndEditing].
However, if the focus is changed by tapping on another form control, a
a gesture recognizer calls [_inputPeripheral endEditing]. Consequently,
changing the focused element by tapping results in two calls to
[WKFormControl controlEndEditing]. First from the gesture recognizer,
and then from [WKContentView _elementDidBlur].
For certain form controls on iPhone, such as the select element,
controlEndEditing is responsible for updating the element's value in
the WebProcess. Note that the same is not true on iPad, as
controlEndEditing simply dismisses a popover. Now, the first call to
[WKSelectSinglePicker controlEndEditing] correctly updates the expected
select element. However, when tapping on another select element, the
second call to [WKSelectSinglePicker controlEndEditing] also attempts
to update the value of the focused element. By this time, the focused
element in the WebProcess is the second select element. Consequently,
changing the value in one select element and then tapping another
results in updates to both elements.
To fix this issue, while preserving the intent of r241734, we should
prevent callers from being able to end editing in a control twice. Note
that WKFormPeripheralBase already keeps track of the editing state of a
control in a boolean variable _editing. If the control has finished
being edited, _editing is set to false. This patch adds an early return
to [WKFormPeripheralBase endEditing] if _editing is false, ensuring
controlEndEditing is never called twice. For parity, also ensure that
controlBeginEditing can only be called if the control is not already
being edited.
Test: fast/forms/ios/select-picker-change-and-focus-another-select.html
* UIProcess/ios/forms/WKFormPeripheralBase.mm:
(-[WKFormPeripheralBase beginEditing]): Early return if control is already being edited.
(-[WKFormPeripheralBase endEditing]): Early return if control is not being edited.
2020-09-04 Wenson Hsieh <wenson_hsieh@apple.com>
[macOS] UI process sometimes crashes under -[WKSharingServicePickerDelegate sharingService:didShareItems:]
https://bugs.webkit.org/show_bug.cgi?id=216167
<rdar://problem/68287449>
Reviewed by Anders Carlsson.
It's not clear what causes the array of items in this delegate method to be empty, but we should at least be
robust against this scenario and avoid crashing the application process.
* UIProcess/mac/WKSharingServicePickerDelegate.mm:
(-[WKSharingServicePickerDelegate sharingService:didShareItems:]):
2020-09-04 Kate Cheney <katherine_cheney@apple.com>
WebProcessProxy::checkURLReceivedFromWebProcess() does not always need to check the back/forward list
https://bugs.webkit.org/show_bug.cgi?id=216109
<rdar://problem/59867349>
Reviewed by Brady Eidson.
We should not need to check the back/forward list when setting a
pasteboard URL because any valid pasteboard URL will be handled
by one of the other if-statements in
WebProcessProxy::checkURLReceivedFromWebProcess(), therefore checking
again is a waste of time.
This also hardens the process against potential abuse by preventing
the possibility of an invalid URL in the back/forward list being
saved in the pasteboard property list.
* UIProcess/Cocoa/WebPasteboardProxyCocoa.mm:
(WebKit::WebPasteboardProxy::setPasteboardURL):
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::checkURLReceivedFromWebProcess):
* UIProcess/WebProcessProxy.h:
2020-09-03 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] Unexpected User-Agent on redirect
https://bugs.webkit.org/show_bug.cgi?id=191858
Reviewed by Youenn Fablet.
Clear the user agent on new request after a redirect to ensure a new one is computed taking into account the
quirks if needed. Also ensure quirks are applied for downloads started in the UI process.
* NetworkProcess/soup/NetworkDataTaskSoup.cpp:
(WebKit::NetworkDataTaskSoup::continueHTTPRedirection): Clear the user agent on new request.
* UIProcess/API/glib/WebKitDownloadClient.cpp:
* UIProcess/WebPageProxy.h: Add userAgentForURL().
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::download): Use userAgentForURL() instead of userAgent().
* UIProcess/gtk/WebPageProxyGtk.cpp:
(WebKit::WebPageProxy::userAgentForURL): Implement it to apply quirks if needed.
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::userAgentForURL): Just return the web page user agent.
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::userAgentForURL): Ditto.
* UIProcess/playstation/WebPageProxyPlayStation.cpp:
(WebKit::WebPageProxy::userAgentForURL): Ditto.
* UIProcess/win/WebPageProxyWin.cpp:
(WebKit::WebPageProxy::userAgentForURL): Ditto.
* UIProcess/wpe/WebPageProxyWPE.cpp:
(WebKit::WebPageProxy::userAgentForURL): Ditto.
2020-09-02 Darin Adler <darin@apple.com>
Remove EAffinity, UPSTREAM, DOWNSTREAM, SEL_DEFAULT_AFFINITY, and VisibleSelection::selectionType
https://bugs.webkit.org/show_bug.cgi?id=216102
Reviewed by Sam Weinig.
* WebProcess/InjectedBundle/API/APIInjectedBundleEditorClient.h:
(API::InjectedBundle::EditorClient::shouldChangeSelectedRange): Use Affinity.
* WebProcess/InjectedBundle/API/c/WKBundleAPICast.h:
(WebKit::toAPI): Dito.
* WebProcess/InjectedBundle/API/mac/WKWebProcessPlugInBrowserContextController.mm:
(-[WKWebProcessPlugInBrowserContextController _setEditingDelegate:]): Ditto.
* WebProcess/InjectedBundle/InjectedBundlePageEditorClient.cpp:
(WebKit::InjectedBundlePageEditorClient::shouldChangeSelectedRange): Ditto.
* WebProcess/InjectedBundle/InjectedBundlePageEditorClient.h: Ditto.
* WebProcess/WebCoreSupport/WebEditorClient.cpp:
(WebKit::WebEditorClient::shouldChangeSelectedRange): Ditto.
* WebProcess/WebCoreSupport/WebEditorClient.h: Ditto.
* WebProcess/WebPage/Cocoa/WebPageCocoa.mm:
(WebKit::WebPage::insertDictatedTextAsync): Rely on default affinity.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::insertTextAsync): Ditto.
(WebKit::WebPage::setCompositionAsync): Ditto.
(WebKit::WebPage::deleteSurrounding): Ditto.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::didConcludeEditDrag): Use Affinity.
(WebKit::WebPage::selectPositionAtBoundaryWithDirection): Ditto.
(WebKit::WebPage::moveSelectionAtBoundaryWithDirection): Ditto.
(WebKit::WebPage::selectTextWithGranularityAtPoint): Ditto.
(WebKit::WebPage::updateSelectionWithExtentPointAndBoundary): Ditto.
(WebKit::WebPage::updateSelectionWithExtentPoint): Ditto.
(WebKit::WebPage::replaceSelectedText): Ditto.
(WebKit::WebPage::replaceDictatedText): Ditto.
(WebKit::WebPage::applyAutocorrectionInternal): Ditto.
(WebKit::WebPage::updateSelectionWithDelta): Ditto.
2020-09-02 John Wilander <wilander@apple.com>
Remove custom domain from SameSite=strict jail protection
https://bugs.webkit.org/show_bug.cgi?id=216083
<rdar://problem/68220223>
Reviewed by Alex Christensen.
No new tests. This was a site-specific opt-in rule. The feature
itself has tests.
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.cpp:
(WebKit::ResourceLoadStatisticsStore::shouldEnforceSameSiteStrictForSpecificDomain const):
2020-09-01 Darin Adler <darin@apple.com>
Remove comparePositions and make VisiblePosition improvements
https://bugs.webkit.org/show_bug.cgi?id=215982
Reviewed by Sam Weinig.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::rangeForPointInRootViewCoordinates): Use <= instead of
comparePositions.
2020-09-02 Timothy Horton <timothy_horton@apple.com>
iPad + Trackpad: JW Library highlighting and edit menus don't appear
https://bugs.webkit.org/show_bug.cgi?id=216086
Reviewed by Wenson Hsieh.
* UIProcess/ios/WKContentViewInteraction.mm:
(applicationIsKnownToIgnoreMouseEvents):
Add JW Library to the list.
2020-09-02 Youenn Fablet <youenn@apple.com>
Introduce a ThreadableWebSocketChannel identifier
https://bugs.webkit.org/show_bug.cgi?id=216034
Reviewed by Alex Christensen.
* NetworkProcess/NetworkConnectionToWebProcess.cpp:
* NetworkProcess/NetworkConnectionToWebProcess.h:
* NetworkProcess/NetworkConnectionToWebProcess.messages.in:
* NetworkProcess/NetworkSocketChannel.h:
* NetworkProcess/NetworkSocketStream.h:
* Scripts/webkit/messages.py:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/Network/WebSocketChannel.cpp:
(WebKit::WebSocketChannel::WebSocketChannel):
* WebProcess/Network/WebSocketChannel.h:
(WebKit::WebSocketChannel::identifier const): Deleted.
* WebProcess/Network/WebSocketChannelManager.h:
* WebProcess/Network/WebSocketIdentifier.h: Removed.
* WebProcess/Network/WebSocketProvider.cpp:
(WebKit::WebSocketProvider::createSocketStreamHandle):
* WebProcess/Network/WebSocketProvider.h:
* WebProcess/Network/WebSocketStream.cpp:
(WebKit::WebSocketStream::create):
(WebKit::WebSocketStream::WebSocketStream):
* WebProcess/Network/WebSocketStream.h:
2020-09-02 Aditya Keerthi <akeerthi@apple.com>
[macOS] Update date picker when the inner control is edited
https://bugs.webkit.org/show_bug.cgi?id=216004
Reviewed by Wenson Hsieh.
* UIProcess/mac/WebDateTimePickerMac.mm:
(WebKit::WebDateTimePickerMac::showDateTimePicker):
If showDateTimePicker is called while a picker is already being
displayed, call updatePicker: rather than showPicker:.
(-[WKDateTimePicker initWithParams:inView:]):
The NSDatePicker and NSDateFormatter should use a UTC timezone.
This is necessary as all double values passed into WKDateTimePicker
are UTC timestamps. This has no effect on the value returned to
the WebProcess on user selection, as a timezone-agnostic format
string is used.
(-[WKDateTimePicker updatePicker:]):
Set the date value of the owned NSDatePicker to the value in the
DateTimeChooserParameters.
2020-09-01 Hector Lopez <hector_i_lopez@apple.com>
Unreviewed, reverting r266408.
Revision caused constant crashes on iOS and macOS
Reverted changeset:
"Convert runtime flag to setting for lazy image loading"
https://bugs.webkit.org/show_bug.cgi?id=215610
https://trac.webkit.org/changeset/266408
2020-09-01 Kate Cheney <katherine_cheney@apple.com>
Pass the correct data size for the SharedMemory::Handle in WebCompiledContentRuleListData::encode
https://bugs.webkit.org/show_bug.cgi?id=216049
<rdar://problem/60758278>
Reviewed by Alex Christensen.
No new tests, validated by existing API tests.
* Shared/WebCompiledContentRuleListData.cpp:
(WebKit::WebCompiledContentRuleListData::encode const):
2020-09-01 Kate Cheney <katherine_cheney@apple.com>
WebPasteboardProxy::getPasteboardStringsForType() and WebPasteboardProxy::readURLFromPasteboard() should check return value of SharedMemory::createHandle()
https://bugs.webkit.org/show_bug.cgi?id=211002
<rdar://problem/60103950>
Reviewed by Wenson Hsieh.
* UIProcess/Cocoa/WebPasteboardProxyCocoa.mm:
(WebKit::WebPasteboardProxy::getPasteboardBufferForType):
(WebKit::WebPasteboardProxy::readBufferFromPasteboard):
Check result of SharedMemory::createHandle() and return early
on failure.
2020-09-01 Hector Lopez <hector_i_lopez@apple.com>
Unreviewed, reverting r266384.
Revision is causing 5 API faliures/timeouts and build failures
on macOS
Reverted changeset:
"Webpages flash when switching between windows"
https://bugs.webkit.org/show_bug.cgi?id=216012
https://trac.webkit.org/changeset/266384
2020-09-01 Per Arne Vollan <pvollan@apple.com>
[iOS] AGX compiler service sandbox violation
https://bugs.webkit.org/show_bug.cgi?id=216042
<rdar://problem/68111667>
Reviewed by Brent Fulgham.
For a set of devices, mach-lookup sandbox violations have been observed for an AGX compiler service. For these devices,
we currently issue an extension for one AGX compiler service, but this is not sufficient since there is a similar
service name that needs to be added as well.
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::agxCompilerServices):
(WebKit::WebProcessPool::platformInitializeWebProcess):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-09-01 Rob Buis <rbuis@igalia.com>
Convert runtime flag to setting for lazy image loading
https://bugs.webkit.org/show_bug.cgi?id=215610
Reviewed by Youenn Fablet.
Remove lazy image loading runtime flag.
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetLazyImageLoadingEnabled): Deleted.
(WKPreferencesGetLazyImageLoadingEnabled): Deleted.
* UIProcess/API/C/WKPreferencesRefPrivate.h:
2020-09-01 Tim Horton <timothy_horton@apple.com>
Some apps linked beyond iOS 13.4 don't respect mouse events, only touch events
https://bugs.webkit.org/show_bug.cgi?id=216021
<rdar://problem/64830335>
Reviewed by Wenson Hsieh.
* UIProcess/Cocoa/VersionChecks.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(applicationIsKnownToIgnoreMouseEvents):
(-[WKContentView shouldUseMouseGestureRecognizer]):
Remove Feedly from the list, they are no longer affected.
Add a once-per-app-launch error-level log message to warn developers
of the impending behavior change.
Add three new apps with a later linked-on target.
2020-09-01 Megan Gardner <megan_gardner@apple.com>
Switch to class method for smart insert/delete
https://bugs.webkit.org/show_bug.cgi?id=215871
<rdar://problem/67281776>
Reviewed by Darin Adler.
No process changes. Covered by existing tests.
* Platform/spi/ios/UIKitSPI.h:
* UIProcess/ios/TextCheckerIOS.mm:
(WebKit::TextChecker::isSmartInsertDeleteEnabled):
2020-09-01 Tim Horton <timothy_horton@apple.com>
Upstream PencilKit text input support
https://bugs.webkit.org/show_bug.cgi?id=216030
Reviewed by Anders Carlsson.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView setUpInteraction]):
(-[WKContentView cleanUpInteraction]):
(-[WKContentView _shouldSimulateKeyboardInputOnTextInsertion]):
(-[WKContentView setUpScribbleInteraction]):
(-[WKContentView cleanUpScribbleInteraction]):
(-[WKContentView _textInputContextByScribbleIdentifier:]):
(-[WKContentView _elementForTextInputContextIsFocused:]):
(-[WKContentView indirectScribbleInteraction:requestElementsInRect:completion:]):
(-[WKContentView indirectScribbleInteraction:isElementFocused:]):
(-[WKContentView indirectScribbleInteraction:focusElementIfNeeded:referencePoint:completion:]):
(-[WKContentView indirectScribbleInteraction:frameForElement:]):
(-[WKContentView indirectScribbleInteraction:willBeginWritingInElement:]):
(-[WKContentView indirectScribbleInteraction:didFinishWritingInElement:]):
2020-09-01 Alex Christensen <achristensen@webkit.org>
Fix crash during WebPage constructor
https://bugs.webkit.org/show_bug.cgi?id=216028
<rdar://problem/67781279>
Reviewed by Timothy Hatcher.
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::ensureNetworkProcessConnection):
2020-09-01 Youenn Fablet <youenn@apple.com>
Enable inspection of WebSocket when opening Web Inspector in the middle of the connection like done for the legacy WebSocket implementation
https://bugs.webkit.org/show_bug.cgi?id=215928
Reviewed by Devin Rousso.
Implement inspector related API and add additional messages for WebSocket frames.
* WebProcess/Network/WebSocketChannel.cpp:
(WebKit::WebSocketChannel::close):
(WebKit::WebSocketChannel::didClose):
(WebKit::WebSocketChannel::didSendHandshakeRequest):
(WebKit::WebSocketChannel::didReceiveHandshakeResponse):
(WebKit::WebSocketChannel::clientHandshakeRequest const):
(WebKit::WebSocketChannel::serverHandshakeResponse const):
(WebKit::WebSocketChannel::handshakeMode const):
* WebProcess/Network/WebSocketChannel.h:
2020-09-01 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK][WPE] Do not use the default WebsiteDataStore for downloads
https://bugs.webkit.org/show_bug.cgi?id=216001
Reviewed by Adrian Perez de Castro.
Use the web context one instead.
* UIProcess/API/glib/WebKitWebContext.cpp:
(webkitWebContextStartDownload):
2020-08-31 Sihui Liu <sihui_liu@apple.com>
Webpages flash when switching between windows
https://bugs.webkit.org/show_bug.cgi?id=216012
Reviewed by Darin Adler.
Based on patches from Tim Horton. To avoid flash, we need to make sure view to be selected updates its content
before UI process commits its transaction. We did this by making UI process block until receiving a reply
DidUpdateActivityState from web process. However, web process did not make sure the reply would be sent after
normal rendering update and corresponding transasction commit. Instead, it flushed transactions in progress and
replied. To fix this, now we make web process reply in transaction commit handler after rendering update.
In the switching case, view to be unselected will detach from root layer in its web process, which makes its
content empty. This change is independent from the UI process commit, so we want this to happen after UI process
commits (which submits the view hierachy change). Otherwise, empty content(white flash) on the unselected view
will be displayed. To fix this, we let UI process send the activity state update message in transaction commit
handler.
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::scheduleActivityStateUpdate):
* UIProcess/WebPageProxy.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
(WebKit::WebPageProxy::setSuppressVisibilityUpdates):
(WebKit::WebPageProxy::activityStateDidChange):
(WebKit::WebPageProxy::dispatchActivityStateChange):
* UIProcess/WebPageProxy.h:
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.h:
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::TiledCoreAnimationDrawingArea):
(WebKit::TiledCoreAnimationDrawingArea::handleActivityStateChangeCallbacksIfNeeded):
(WebKit::TiledCoreAnimationDrawingArea::updateRendering):
(WebKit::TiledCoreAnimationDrawingArea::activityStateDidChange):
(WebKit::TiledCoreAnimationDrawingArea::didUpdateActivityStateTimerFired): Deleted.
2020-08-31 Tim Horton <timothy_horton@apple.com>
Upstream support for image picker confirmation and downscaling
https://bugs.webkit.org/show_bug.cgi?id=215997
Reviewed by Anders Carlsson.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _showRunOpenPanel:frameInfo:resultListener:]):
* UIProcess/ios/forms/WKFileUploadPanel.mm:
(-[WKFileUploadPanel _showPhotoPickerWithSourceType:]):
(-[WKFileUploadPanel _configureImagePicker:]): Deleted.
2020-08-31 Sam Weinig <weinig@apple.com>
Many settings in RuntimeEnabledSettings should autogenerated Settings: Part 1
https://bugs.webkit.org/show_bug.cgi?id=215981
Reviewed by Darin Adler.
This moves an initial chunk of RuntimeEnabledSettings settings to Settings.yaml, focusing
on settings that don't require substantial refactoring or new features to work in Settings.yaml.
In WebKit, this means mostly removing now unnecessary use of the webcoreBinding key in
WebPreferences.yaml and removing now superfluous cases in InjectedBundle::overrideBoolPreferenceForTestRunner.
* Shared/WebPreferences.yaml:
Removing now unnecessary use of the webcoreBinding key and fix indentation of a few cases.
* WebProcess/InjectedBundle/InjectedBundle.cpp:
(WebKit::InjectedBundle::overrideBoolPreferenceForTestRunner):
Remove now superfluous cases, as each boolean Settings based setting gets handled automatically.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences):
Remove explicit setting of setInteractiveFormValidationEnabled now that we don't have the
RuntimeEnabledSettings variant being set by the generator.
2020-08-31 Hector Lopez <hector_i_lopez@apple.com>
Unreviewed, reverting r266350.
Revision is causing a constant crash on both macOS and iOS
Reverted changeset:
"Convert runtime flag to setting for lazy image loading"
https://bugs.webkit.org/show_bug.cgi?id=215610
https://trac.webkit.org/changeset/266350
2020-08-31 Rob Buis <rbuis@igalia.com>
Convert runtime flag to setting for lazy image loading
https://bugs.webkit.org/show_bug.cgi?id=215610
Reviewed by Youenn Fablet.
Remove lazy image loading runtime flag.
* Shared/WebPreferences.yaml:
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetLazyImageLoadingEnabled): Deleted.
(WKPreferencesGetLazyImageLoadingEnabled): Deleted.
* UIProcess/API/C/WKPreferencesRefPrivate.h:
2020-08-31 Youenn Fablet <youenn@apple.com>
[Cocoa] Disable sniffing for WebSocket handshakes
https://bugs.webkit.org/show_bug.cgi?id=215932
Reviewed by Darin Adler.
Covered by http/tests/websocket/tests/hybi/handshake-error.html no longer timing out with NSURLSession WebSocket code path.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::createWebSocketTask):
2020-08-30 Tim Horton <timothy_horton@apple.com>
Remove unused editable image code
https://bugs.webkit.org/show_bug.cgi?id=215991
Reviewed by Simon Fraser.
* DerivedSources.make:
* Platform/spi/ios/PencilKitSPI.h: Removed.
* Shared/FocusedElementInformation.cpp:
(WebKit::FocusedElementInformation::encode const):
(WebKit::FocusedElementInformation::decode):
* Shared/FocusedElementInformation.h:
* Shared/RemoteLayerTree/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::drawInContext):
* Shared/RemoteLayerTree/RemoteLayerTreeTransaction.h:
* Shared/RemoteLayerTree/RemoteLayerTreeTransaction.mm:
(WebKit::RemoteLayerTreeTransaction::LayerCreationProperties::LayerCreationProperties):
(WebKit::RemoteLayerTreeTransaction::LayerCreationProperties::encode const):
(WebKit::RemoteLayerTreeTransaction::LayerCreationProperties::decode):
* Shared/WebPreferences.yaml:
* SourcesCocoa.txt:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _setupPageConfiguration:]):
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration init]):
(-[WKWebViewConfiguration copyWithZone:]):
(-[WKWebViewConfiguration _setEditableImagesEnabled:]): Deleted.
(-[WKWebViewConfiguration _editableImagesEnabled]): Deleted.
* UIProcess/API/Cocoa/WKWebViewConfigurationPrivate.h:
* UIProcess/API/ios/WKWebViewIOS.h:
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _setupScrollAndContentViews]):
(-[WKWebView _stylusTapGestureShouldCreateEditableImage]): Deleted.
* UIProcess/PageClient.h:
(WebKit::PageClient::createDrawingView): Deleted.
* UIProcess/RemoteLayerTree/RemoteLayerTreeHost.h:
* UIProcess/RemoteLayerTree/RemoteLayerTreeHost.mm:
(WebKit::RemoteLayerTreeHost::layerWillBeRemoved):
(WebKit::RemoteLayerTreeHost::clearLayers):
(WebKit::RemoteLayerTreeHost::makeNode):
* UIProcess/RemoteLayerTree/ios/RemoteLayerTreeHostIOS.mm:
(WebKit::RemoteLayerTreeHost::makeNode):
(WebKit::RemoteLayerTreeHost::createEmbeddedView): Deleted.
* UIProcess/RemoteLayerTree/ios/RemoteLayerTreeViews.h:
* UIProcess/RemoteLayerTree/ios/RemoteLayerTreeViews.mm:
(-[WKEmbeddedView initWithEmbeddedViewID:]): Deleted.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didAttachToRunningProcess):
(WebKit::WebPageProxy::resetState):
(WebKit::WebPageProxy::willUpdateAttachmentAttributes):
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::editableImageController): Deleted.
* UIProcess/ios/EditableImageController.h: Removed.
* UIProcess/ios/EditableImageController.messages.in: Removed.
* UIProcess/ios/EditableImageController.mm: Removed.
* UIProcess/ios/PageClientImplIOS.h:
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::createDrawingView): Deleted.
* UIProcess/ios/PencilKitSoftLink.h: Removed.
* UIProcess/ios/PencilKitSoftLink.mm: Removed.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView setUpInteraction]):
(-[WKContentView cleanUpInteraction]):
(-[WKContentView _removeDefaultGestureRecognizers]):
(-[WKContentView _addDefaultGestureRecognizers]):
(-[WKContentView gestureRecognizerShouldBegin:]):
(-[WKContentView _elementDidFocus:userIsInteracting:blurPreviousNode:activityStateChanges:userObject:]):
(-[WKContentView _elementDidBlur]):
(-[WKContentView _stylusSingleTapRecognized:]): Deleted.
(-[WKContentView _drawingCoordinator]): Deleted.
* UIProcess/ios/WKDrawingCoordinator.h: Removed.
* UIProcess/ios/WKDrawingCoordinator.mm: Removed.
* UIProcess/ios/WKDrawingView.h: Removed.
* UIProcess/ios/WKDrawingView.mm: Removed.
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::handleStylusSingleTapAtPoint): Deleted.
* UIProcess/ios/fullscreen/WKFullScreenWindowControllerIOS.mm:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebCoreSupport/ios/WebChromeClientIOS.mm:
(WebKit::WebChromeClient::associateEditableImageWithAttachment): Deleted.
(WebKit::WebChromeClient::didCreateEditableImage): Deleted.
(WebKit::WebChromeClient::didDestroyEditableImage): Deleted.
* WebProcess/WebPage/RemoteLayerTree/GraphicsLayerCARemote.cpp:
(WebKit::GraphicsLayerCARemote::createPlatformCALayerForEmbeddedView): Deleted.
* WebProcess/WebPage/RemoteLayerTree/GraphicsLayerCARemote.h:
* WebProcess/WebPage/RemoteLayerTree/PlatformCALayerRemote.cpp:
(WebKit::PlatformCALayerRemote::createForEmbeddedView): Deleted.
(WebKit::PlatformCALayerRemote::embeddedViewID const): Deleted.
* WebProcess/WebPage/RemoteLayerTree/PlatformCALayerRemote.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeContext.mm:
(WebKit::RemoteLayerTreeContext::layerDidEnterContext):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::isAssistableElement):
(WebKit::WebPage::getFocusedElementInformation):
(WebKit::WebPage::handleStylusSingleTapAtPoint): Deleted.
2020-08-28 Tim Horton <timothy_horton@apple.com>
Sites built with flambe.js don't work with trackpad on iPad
https://bugs.webkit.org/show_bug.cgi?id=215954
Reviewed by Wenson Hsieh.
New API Test: iOSMouseSupport.MouseTimestampTimebase
* UIProcess/ios/WKMouseGestureRecognizer.mm:
(-[WKMouseGestureRecognizer createMouseEventWithType:]):
Use the timestamp property on UITouch instead of GSCurrentEventTimestamp().
This property is in the same units as mach_absolute_time/MonotonicTime.
Previously, treating a monotonic time as wall time resulted in it getting clamped to 0,
breaking sites that depend on it.
2020-08-28 Per Arne Vollan <pvollan@apple.com>
[macOS] Fix iokit-get-properties sandbox violations
https://bugs.webkit.org/show_bug.cgi?id=215925
Reviewed by Brent Fulgham.
Add iokit-property to WebContent sandbox that should have been a part of r266282.
* WebProcess/com.apple.WebProcess.sb.in:
2020-08-28 Darin Adler <darin@apple.com>
Remove almost all the remaining uses of live ranges
https://bugs.webkit.org/show_bug.cgi?id=215755
Reviewed by Sam Weinig.
* WebProcess/InjectedBundle/API/gtk/DOM/WebKitDOMRange.cpp:
(webkit_dom_range_compare_boundary_points): Update for rename of
compareBoundaryPointsForBindings.
(webkit_dom_range_intersects_node): Removed exception handling.
(webkit_dom_range_get_text): Use plainText.
* WebProcess/InjectedBundle/API/mac/WKDOMRange.mm:
(-[WKDOMRange initWithDocument:]): Tightened up into a one-liner.
(-[WKDOMRange text]): Use plainText.
(-[WKDOMRange textRects]): Stopped using Range::ownerDocument.
* WebProcess/InjectedBundle/DOM/InjectedBundleRangeHandle.cpp:
(WebKit::InjectedBundleRangeHandle::document): Stopped using
Range::ownerDocument.
(WebKit::InjectedBundleRangeHandle::boundingRectInWindowCoordinates const):
Ditto.
(WebKit::InjectedBundleRangeHandle::renderedImage): Ditto.
(WebKit::InjectedBundleRangeHandle::text const): Use plainText.
* WebProcess/WebPage/glib/WebPageGLib.cpp:
(WebKit::WebPage::getPlatformEditorState const): Use contains.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::requestDocumentEditingContext): Use intersects.
2020-08-28 Alex Christensen <achristensen@webkit.org>
REGRESSION(r266187): ARGUMENT BAD: WebIconUtilities.mm(138) : RetainPtr<UIImage> WebKit::iconForFile(NSURL *) file, [file isFileURL]
https://bugs.webkit.org/show_bug.cgi?id=215931
Reviewed by Wenson Hsieh.
It's possible for JavaScript to make a filename that doesn't convert to a file URL, such as the empty string.
When this happens, we would've returned null for the icon later anyways, but adding a null check prevents an assertion.
Covered by 7 existing tests which would have hit that assertion.
* WebProcess/WebCoreSupport/ios/WebChromeClientIOS.mm:
(WebKit::WebChromeClient::createIconForFiles):
2020-08-28 Youenn Fablet <youenn@apple.com>
Enable TransformStream by default
https://bugs.webkit.org/show_bug.cgi?id=215891
Reviewed by Darin Adler.
* Shared/WebPreferences.yaml:
2020-08-28 Per Arne Vollan <pvollan@apple.com>
[macOS] Fix iokit-get-properties sandbox violations
https://bugs.webkit.org/show_bug.cgi?id=215925
Reviewed by Darin Adler.
A few iokit properties needs to be added to the WebContent process' sandbox after observing these being in use.
* WebProcess/com.apple.WebProcess.sb.in:
2020-08-27 John Wilander <wilander@apple.com>
Remove the feature flag for capped cookies set in 3rd-party CNAME cloaked HTTP responses and add ITP debug logging
https://bugs.webkit.org/show_bug.cgi?id=215902
<rdar://problem/66699731>
Reviewed by Brent Fulgham.
This change removes the feature flag, effectively turning the feature on.
It also adds log output of cookie names that have been capped when ITP
Debug Mode is enabled.
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::setResourceLoadStatisticsDebugMode):
Now sets the flag in WebCore::NetworkStorageSession through the new function
WebCore::NetworkStorageSession::setResourceLoadStatisticsDebugLoggingEnabled().
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::resetParametersToDefaultValues):
(WebKit::NetworkProcess::setIsRunningResourceLoadStatisticsTest):
* NetworkProcess/NetworkSession.cpp:
(WebKit::NetworkSession::NetworkSession):
(WebKit::NetworkSession::setFirstPartyHostCNAMEDomain):
(WebKit::NetworkSession::firstPartyHostCNAMEDomain):
* NetworkProcess/NetworkSession.h:
(WebKit::NetworkSession::setCNAMECloakingMitigationEnabled): Deleted.
(WebKit::NetworkSession::cnameCloakingMitigationEnabled const): Deleted.
* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::NetworkDataTaskCocoa::updateFirstPartyInfoForSession):
(WebKit::NetworkDataTaskCocoa::applyCookiePolicyForThirdPartyCNAMECloaking):
* Shared/ResourceLoadStatisticsParameters.h:
(WebKit::ResourceLoadStatisticsParameters::encode const):
(WebKit::ResourceLoadStatisticsParameters::decode):
* Shared/WebPreferences.yaml:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::ensureNetworkProcess):
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::WebsiteDataStore::platformSetNetworkParameters):
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::parameters):
2020-08-27 Devin Rousso <drousso@apple.com>
[iOS] provide a way to get previously inserted alternatives for the selected text
https://bugs.webkit.org/show_bug.cgi?id=215816
<rdar://problem/66646042>
Reviewed by Darin Adler.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView alternativesForSelectedText]):
* UIProcess/WebPageProxy.h:
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::platformDictationAlternatives): Added.
* UIProcess/PageClient.h:
* UIProcess/Cocoa/PageClientImplCocoa.h:
* UIProcess/Cocoa/PageClientImplCocoa.mm:
(WebKit::PageClientImplCocoa::dictationAlternatives):
(WebKit::PageClientImplCocoa::platformDictationAlternatives): Added.
Provide a way to get the raw `NSTextAlternatives *` for a given `WebCore::DictationContext`.
* Shared/EditorState.h:
* Shared/EditorState.cpp:
(WebKit::EditorState::PostLayoutData::encode const):
(WebKit::EditorState::PostLayoutData::decode):
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::getPlatformEditorState const):
Include a `Vector<WebCore::DictationContext>` as part of the `EditorState` that contains all
of the `WebCore::DictationContext` that exist in the currently selected range (or the range
of the word containing the cursor if nothing is selected).
* Platform/spi/ios/UIKitSPI.h:
2020-08-27 Jer Noble <jer.noble@apple.com>
REGRESSION(r264710): Initializing the AVPlayer Obj-C class at process start up causes a regression in power-use tests
https://bugs.webkit.org/show_bug.cgi?id=215861
Reviewed by Eric Carlson.
Calling +instancesRespondToSelector: will cause the underyling Obj-C class to be initialized, which in the case of
AVPlayer does a bunch of work which shows up on power-use tests. Because the API being checked for will always exist
in macOS 11, and that code is already protected by a HAVE(AVPLAYER_VIDEORANGEOVERRIDE) pragma, just return early
here without running the +instancesRespondToSelector: check.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::updatePageScreenProperties):
2020-08-27 Chris Dumez <cdumez@apple.com>
Unreviewed, add missing double quote in r266216.
* WebProcess/com.apple.WebProcess.sb.in:
2020-08-27 Kate Cheney <katherine_cheney@apple.com>
syntax error in "DELETE FROM StorageAccessUnderTopFrameDomains"
https://bugs.webkit.org/show_bug.cgi?id=215813
<rdar://problem/67743521>
Reviewed by Darin Adler.
Delete unnecessary code. ResourceLoadStatisticsDatabaseStore::clearExpiredUserInteractions()
is not needed. User interaction and storage access will be cleared for
expired statistics in
ResourceLoadStatisticsDatabaseStore::hasHadUnexpiredRecentUserInteraction(),
which exactly matches behavior in the memory store. If we clear the
expired statistics before setting the local var oldestUserInteraction,
then that var will likely always be equal to zero, and the check for
if (!parameters().isRunningTest && now - oldestUserInteraction < parameters().minimumTimeBetweenDataRecordsRemoval)
will always be false.
The only reason we did not notice this extra code is because the
SQLite query in clearExpiredUserInteractions() was incorrect, causing
an early return. So the function never worked properly. Fixing the
query revealed incorrect behavior.
No new tests, behavior should not change because this code was not
being run due to the SQLite error.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::registrableDomainsToDeleteOrRestrictWebsiteDataFor):
(WebKit::ResourceLoadStatisticsDatabaseStore::findExpiredUserInteractions const): Deleted.
(WebKit::ResourceLoadStatisticsDatabaseStore::clearExpiredUserInteractions): Deleted.
(WebKit::ResourceLoadStatisticsDatabaseStore::statisticsExpirationTime const): Deleted.
2020-08-27 Youenn Fablet <youenn@apple.com>
Add support for TransformStream
https://bugs.webkit.org/show_bug.cgi?id=215607
Reviewed by Alex Christensen.
Add a runtime flag for TransformStream, off by default.
* Shared/WebPreferences.yaml:
* WebProcess/InjectedBundle/InjectedBundle.cpp:
(WebKit::InjectedBundle::overrideBoolPreferenceForTestRunner):
2020-08-26 Per Arne Vollan <pvollan@apple.com>
[macOS] Stop logging sandbox violations of com.apple.CoreDisplay.Notification
https://bugs.webkit.org/show_bug.cgi?id=215869
Reviewed by Darin Adler.
On macOS, stop logging mach-lookup sandbox violations of com.apple.CoreDisplay.Notification for performance reasons.
No new tests, since there is no change in behavior. The service was already being denied in the sandbox.
* WebProcess/com.apple.WebProcess.sb.in:
2020-08-26 Kate Cheney <katherine_cheney@apple.com>
Resource Load Statistics data summary does not report data which is held up in the web content process.
https://bugs.webkit.org/show_bug.cgi?id=215822
<rdar://problem/66682044>
Reviewed by Chris Dumez.
No new tests, this fixes a timing bug that is flaky to reproduce, so I
was unable to write a test case. Non-regressed behavior is confirmed
with existing API testing.
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::resourceLoadStatisticsUpdated):
Most of the time, the completion handler will be an empty function,
but we should handle it in the network process so we fix the
case where we wait to send the full data summary until the update has
finished.
(WebKit::WebResourceLoadStatisticsStore::aggregatedThirdPartyData):
Delete extra space.
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.h:
* NetworkProcess/NetworkConnectionToWebProcess.cpp:
(WebKit::NetworkConnectionToWebProcess::resourceLoadStatisticsUpdated):
* NetworkProcess/NetworkConnectionToWebProcess.h:
* NetworkProcess/NetworkConnectionToWebProcess.messages.in:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::sendResourceLoadStatisticsDataImmediately):
* UIProcess/WebProcessPool.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::getResourceLoadStatisticsDataSummary):
Don't ask the network process for data until any lingering data
in the web content process has been sent first.
* WebProcess/InjectedBundle/API/c/WKBundle.cpp:
(WKBundleResourceLoadStatisticsNotifyObserver):
* WebProcess/WebCoreSupport/WebResourceLoadObserver.cpp:
(WebKit::WebResourceLoadObserver::WebResourceLoadObserver):
(WebKit::WebResourceLoadObserver::~WebResourceLoadObserver):
(WebKit::WebResourceLoadObserver::updateCentralStatisticsStore):
* WebProcess/WebCoreSupport/WebResourceLoadObserver.h:
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::flushResourceLoadStatistics):
(WebKit::WebProcess::sendResourceLoadStatisticsDataImmediately):
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
2020-08-26 Andres Gonzalez <andresg_22@apple.com>
Buttons with aria-haspopup attribute are not exposed to accessibility clients as form controls.
https://bugs.webkit.org/show_bug.cgi?id=215866
Reviewed by Chris Fleizach.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::handleSyntheticClick): isARIAControl false implies
isARIAInput false, thus removed the || isARIAInput call.
2020-08-26 Aditya Keerthi <akeerthi@apple.com>
[iOS] Disabled options in the multi-select picker should not be selectable
https://bugs.webkit.org/show_bug.cgi?id=201458
<rdar://problem/55018179>
Reviewed by Wenson Hsieh.
WKMultipleSelectPicker is displayed when a <select multiple> or a
<select> with an <optgroup> is activated. Due to <rdar://problem/18745253>,
group rows and disabled rows were selectable in the picker. r175266 added a
workaround for this issue, preventing group rows from being selectable.
However, the workaround did not account for disabled rows, which means that
it is possible to select disabled options in the multi-select picker.
To fix this behavior, the same fix that was applied to group rows is now
applied to disabled rows. `pickerView:row:column:checked:` resets the style
for the associated view if `item.disabled` is true.
Test: fast/forms/ios/disabled-options-in-multi-select-picker.html
* UIProcess/ios/forms/WKFormSelectPicker.mm:
(-[WKMultipleSelectPicker pickerView:row:column:checked:]):
2020-08-26 Youenn Fablet <youenn@apple.com>
Enable WritableStream by default
https://bugs.webkit.org/show_bug.cgi?id=149842
Reviewed by Alex Christensen.
Implementation is now ready and is up to date with the latest specification.
* Shared/WebPreferences.yaml:
2020-08-26 Chris Dumez <cdumez@apple.com>
REGRESSION (r265908): Crash under Blob::arrayBuffer() / Blob::text() in stress GC
https://bugs.webkit.org/show_bug.cgi?id=215832
<rdar://problem/67741677>
Reviewed by Ryosuke Niwa.
* WebProcess/Automation/WebAutomationSessionProxy.cpp:
(WebKit::WebAutomationSessionProxy::setFilesForInputFileUpload):
2020-08-26 Youenn Fablet <youenn@apple.com>
enumerateDevices should expose audiooutput devices that are tied to an audio input device
https://bugs.webkit.org/show_bug.cgi?id=215806
Reviewed by Eric Carlson.
Introduce an experimental flag (off by default) to cover speaker selection, including exposing speaker devices to web pages.
Update UserMediaPermissionRequestManagerProxy::computeFilteredDeviceList to only expose speakers that are tied to a microphone that is exposed.
This aligns with https://w3c.github.io/mediacapture-output/#privacy-obtaining-consent.
* GPUProcess/GPUConnectionToWebProcess.cpp:
* Shared/WebPreferences.yaml:
* UIProcess/API/C/WKMockMediaDevice.cpp:
(WKAddMockMediaDevice):
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetExposeSpeakersEnabled):
(WKPreferencesGetExposeSpeakersEnabled):
* UIProcess/API/C/WKPreferencesRefPrivate.h:
* UIProcess/Cocoa/UserMediaCaptureManagerProxy.cpp:
(WebKit::UserMediaCaptureManagerProxy::createMediaSourceForCaptureDeviceWithConstraints):
* UIProcess/UserMediaPermissionRequestManagerProxy.cpp:
(WebKit::isMicrophoneDevice):
(WebKit::UserMediaPermissionRequestManagerProxy::computeFilteredDeviceList):
* WebProcess/cocoa/RemoteRealtimeMediaSource.cpp:
(WebKit::sourceTypeFromDeviceType):
(WebKit::RemoteRealtimeMediaSource::RemoteRealtimeMediaSource):
(WebKit::RemoteRealtimeMediaSource::~RemoteRealtimeMediaSource):
* WebProcess/cocoa/UserMediaCaptureManager.h:
2020-08-25 Tim Horton <timothy_horton@apple.com>
Web Share API Level 2 functions even when its experimental feature flag is disabled
https://bugs.webkit.org/show_bug.cgi?id=215831
<rdar://problem/67760687>
Reviewed by Darin Adler.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::showShareSheet):
If the Web Content process sends us files, but the Level 2 feature flag is disabled,
something fishy is happening, so fire a MESSAGE_CHECK.
2020-08-25 Ryosuke Niwa <rniwa@webkit.org>
HashMap<Ref<T>>::take should return RefPtr<T>
https://bugs.webkit.org/show_bug.cgi?id=215830
Reviewed by Darin Adler.
* GPUProcess/media/RemoteAudioDestinationManager.cpp:
(WebKit::RemoteAudioDestinationManager::deleteAudioDestination):
* NetworkProcess/NetworkResourceLoadMap.cpp:
(WebKit::NetworkResourceLoadMap::take):
* UIProcess/Automation/WebAutomationSession.cpp:
(WebKit::WebAutomationSession::willClosePage):
* UIProcess/WebURLSchemeHandler.cpp:
(WebKit::WebURLSchemeHandler::taskCompleted):
* WebProcess/MediaStream/UserMediaPermissionRequestManager.cpp:
(WebKit::UserMediaPermissionRequestManager::userMediaAccessWasGranted):
(WebKit::UserMediaPermissionRequestManager::userMediaAccessWasDenied):
2020-08-25 Tim Horton <timothy_horton@apple.com>
Web Share API can share non-HTTP(S) URLs
https://bugs.webkit.org/show_bug.cgi?id=215823
<rdar://problem/62083130>
Reviewed by Wenson Hsieh.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::showShareSheet):
Ensure that only HTTP family or data: URLs are shared.
2020-08-25 Megan Gardner <megan_gardner@apple.com>
Trying to lookup when WebView is in a popover causes process to hang. Fix for Legacy WebView.
https://bugs.webkit.org/show_bug.cgi?id=215792
Reviewed by Tim Horton.
Update code based on Darin's suggestion for mirror fix in legacy webkit.
* UIProcess/mac/WebContextMenuProxyMac.mm:
(WebKit::WebContextMenuProxyMac::getContextMenuFromItems):
2020-08-25 Carlos Alberto Lopez Perez <clopez@igalia.com>
[WPE][Qt] Fix qt-wpe-minibrowser on Debian10
https://bugs.webkit.org/show_bug.cgi?id=215730
Reviewed by Philippe Normand.
QWheelEvent position is not available on Qt < 5.14. Use posF instead in that case.
Tested with Qt 5.11
No new tests, is a build fix.
* UIProcess/API/wpe/qt/WPEQtViewBackend.cpp:
(WPEQtViewBackend::dispatchWheelEvent):
2020-08-25 Jer Noble <jer.noble@apple.com>
[Mac] REGRESSION(r262322): Focusable elements are focused when exiting from video fullscreen mode.
https://bugs.webkit.org/show_bug.cgi?id=215660
<rdar://problem/65263150>
Reviewed by Darin Adler.
API Test: Fullscreen.Focus.
The default behavior of NSWindow is to change its -keyViewSelectionDirection to NSSelectingNext when activating
and changing the firstResponder. This causes WKWebView to change it's focus in response to becoming first responder.
The cheap fix is to ensure WKWebView is already the firstResponder when the window is activated. Then NSWindow won't
try to change the firstResponder in response to activation.
* UIProcess/mac/WKFullScreenWindowController.mm:
(-[WKFullScreenWindowController beganEnterFullScreenWithInitialFrame:finalFrame:]):
2020-08-25 Alex Christensen <achristensen@webkit.org>
Clean up deprecated and unused SPI to set HSTS storage on ProcessPoolConfiguration
https://bugs.webkit.org/show_bug.cgi?id=215788
Reviewed by Chris Dumez.
After rdar://problem/65047971 this is no longer needed because the replacement on WebsiteDataStoreConfiguration has been adopted.
The selector on _WKProcessPoolConfiguration had already been deprecated, and is left for now to allow easier debugging with OSes that use it still.
* NetworkProcess/NetworkProcessCreationParameters.cpp:
(WebKit::NetworkProcessCreationParameters::encode const):
(WebKit::NetworkProcessCreationParameters::decode):
* NetworkProcess/NetworkProcessCreationParameters.h:
* NetworkProcess/cocoa/NetworkProcessCocoa.mm:
(WebKit::NetworkProcess::platformInitializeNetworkProcessCocoa):
* NetworkProcess/soup/NetworkProcessSoup.cpp:
(WebKit::NetworkProcess::platformInitializeNetworkProcess):
* UIProcess/API/APIProcessPoolConfiguration.cpp:
(API::ProcessPoolConfiguration::copy):
* UIProcess/API/APIProcessPoolConfiguration.h:
* UIProcess/API/Cocoa/_WKProcessPoolConfiguration.mm:
(-[_WKProcessPoolConfiguration setHSTSStorageDirectory:]):
(-[_WKProcessPoolConfiguration hstsStorageDirectory]):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::ensureNetworkProcess):
2020-08-25 Eric Carlson <eric.carlson@apple.com>
[macOS] Update audio arbitration manager when audio transport changes
https://bugs.webkit.org/show_bug.cgi?id=215781
<rdar://problem/65920613>
Reviewed by Jer Noble.
* UIProcess/API/Cocoa/WKWebViewPrivateForTesting.h:
* UIProcess/API/Cocoa/WKWebViewTesting.mm:
(-[WKWebView _audioRoutingArbitrationUpdateTime]):
* UIProcess/Media/AudioSessionRoutingArbitratorProxy.h:
(WebKit::AudioSessionRoutingArbitratorProxy::arbitrationUpdateTime const):
* UIProcess/Media/cocoa/AudioSessionRoutingArbitratorProxyCocoa.mm:
(WebKit::AudioSessionRoutingArbitratorProxy::beginRoutingArbitrationWithCategory):
2020-08-25 Youenn Fablet <youenn@apple.com>
Add support for MediaRecorder bitrate options
https://bugs.webkit.org/show_bug.cgi?id=214973
Reviewed by Eric Carlson.
Serialize options when creating remote media recorder.
* GPUProcess/webrtc/RemoteMediaRecorder.cpp:
(WebKit::RemoteMediaRecorder::create):
* GPUProcess/webrtc/RemoteMediaRecorder.h:
* GPUProcess/webrtc/RemoteMediaRecorderManager.cpp:
(WebKit::RemoteMediaRecorderManager::createRecorder):
* GPUProcess/webrtc/RemoteMediaRecorderManager.h:
* GPUProcess/webrtc/RemoteMediaRecorderManager.messages.in:
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
(WebKit::MediaRecorderPrivate::MediaRecorderPrivate):
(WebKit::MediaRecorderPrivate::startRecording):
* WebProcess/GPU/webrtc/MediaRecorderPrivate.h:
* WebProcess/GPU/webrtc/MediaRecorderProvider.cpp:
(WebKit::MediaRecorderProvider::createMediaRecorderPrivate):
* WebProcess/GPU/webrtc/MediaRecorderProvider.h:
2020-08-25 Carlos Garcia Campos <cgarcia@igalia.com>
[SOUP] Initialize HSTS storage directory from network session initialization
https://bugs.webkit.org/show_bug.cgi?id=215802
Reviewed by Alex Christensen.
Instead of network process initialization, since it's not actually a global setting.
* NetworkProcess/soup/NetworkProcessSoup.cpp:
(WebKit::NetworkProcess::platformInitializeNetworkProcess): Remove code to initialize HSTS storage.
* NetworkProcess/soup/NetworkSessionSoup.cpp:
(WebKit::NetworkSessionSoup::NetworkSessionSoup): Call SoupNetworkSession::setHSTSPersistentStorage() if the
path is not mpety.
* UIProcess/API/glib/WebKitWebContext.cpp:
(webkitWebContextConstructed): Remove the ode to set the HSTS storage directory as process pool configuration.
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::ensureNetworkProcess): Initialize hstsStorageDirectory in default network session parameters.
2020-08-24 Brady Eidson <beidson@apple.com>
CrashTracer: com.apple.WebKit.WebContent at com.apple.WebKit: WebKit::PDFPlugin::createScrollbar
<rdar://problem/67473335> and https://bugs.webkit.org/show_bug.cgi?id=215787
Reviewed by Tim Horton.
To quote Tim from r264945:
No new tests; timing is such that I can't reproduce without inserting
intentional delays into the main thread hops, which is further than
I'm willing to go for a test.
This is a speculative fix due to the aforementioned reproducibility issue.
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::receivedNonLinearizedPDFSentinel): Check on the main thread whenever this is called,
instead of when deciding to dispatch to the main thread.
(WebKit::PDFPlugin::threadEntry): We can't do this check on the background thread when considering the dispatch
to the main thread, as the flag might've changed by then. Let's *just* check it on the main thread.
(WebKit::PDFPlugin::adoptBackgroundThreadDocument): We can't do the check on the background thread when
(WebKit::PDFPlugin::updateScrollbars): This is where the crash itself is. All of the Obj-C code in here is
safe to do after destroy(), up until the very end when we get into pluginView() derefencing.
So it seems prudent to add another check here.
(WebKit::PDFPlugin::documentDataDidFinishLoading): In addition to receivedNonLinearizedPDFSentinel and
adoptBackgroundThreadDocument, this is the final of the (3) calls that end up calling installPDFDocument,
so for added coverage it seems like a prudent place to add the check.
2020-08-24 Wenson Hsieh <wenson_hsieh@apple.com>
Unreviewed, fix the internal iOS 13.4 build
The -_selectionClipRect SPI method isn't declared in any private headers on iOS 13.4 and prior. This isn't a
problem when building with a non-internal SDK since the SPI is declared in the `!USE(APPLE_INTERNAL_SDK)`
section in `UIKitSPI.h`, but on internal builds of iOS 13.4, this declaration is absent from both the internal
SDK and `UIKitSPI.h`.
To fix the build, simply declare it in the IPI section — once we stop supporting iOS 13.4 as a build target,
this can be moved back into the non-internal section only.
* Platform/spi/ios/UIKitSPI.h:
2020-08-24 Alex Christensen <achristensen@webkit.org>
Implement Request/Response consuming as FormData
https://bugs.webkit.org/show_bug.cgi?id=215671
Reviewed by Darin Adler.
* WebProcess/Storage/WebServiceWorkerFetchTaskClient.cpp:
(WebKit::WebServiceWorkerFetchTaskClient::didReceiveFormDataAndFinish):
Add a fast path that allows non-blob FormData responses from service workers to not hang.
This part is covered by this layout test:
imported/w3c/web-platform-tests/service-workers/service-worker/fetch-event-respond-with-custom-response.https.html
2020-08-24 Alex Christensen <achristensen@webkit.org>
Make _WKWebsiteDataStoreConfiguration SPI for HSTS storage to replace _WKProcessPoolConfiguration.hstsStorageDirectory
https://bugs.webkit.org/show_bug.cgi?id=213048
Reviewed by Youenn Fablet.
This uses CFNetwork SPI introduced in rdar://problem/50109631 to allow HSTS storage per NSURLSession.
To be complete, I also deprecated our UI process HSTS state removal attempt SPIs, WKContextResetHSTSHosts and
WKContextResetHSTSHostsAddedAfterDate, which had their last use removed in rdar://problem/64220838.
I manually verified that this new SPI puts HSTS data in the specified location, and I also verified that HSTS
state querying and removal works with the new CFNetwork SPI as it did with the old one.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::fetchWebsiteData):
(WebKit::NetworkProcess::deleteWebsiteData):
(WebKit::NetworkProcess::deleteWebsiteDataForOrigins):
(WebKit::NetworkProcess::deleteAndRestrictWebsiteDataForRegistrableDomains):
(WebKit::NetworkProcess::registrableDomainsWithWebsiteData):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkSessionCreationParameters.cpp:
(WebKit::NetworkSessionCreationParameters::encode const):
(WebKit::NetworkSessionCreationParameters::decode):
* NetworkProcess/NetworkSessionCreationParameters.h:
* NetworkProcess/cocoa/NetworkProcessCocoa.mm:
(WebKit::NetworkProcess::hostNamesWithHSTSCache const):
(WebKit::NetworkProcess::deleteHSTSCacheForHostNames):
(WebKit::NetworkProcess::clearHSTSCache):
(WebKit::NetworkProcess::getHostNamesWithHSTSCache): Deleted.
* NetworkProcess/cocoa/NetworkSessionCocoa.h:
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::hstsStorage const):
(WebKit::NetworkSessionCocoa::NetworkSessionCocoa):
* NetworkProcess/soup/NetworkProcessSoup.cpp:
(WebKit::NetworkProcess::hostNamesWithHSTSCache const):
(WebKit::NetworkProcess::deleteHSTSCacheForHostNames):
(WebKit::NetworkProcess::clearHSTSCache):
(WebKit::NetworkProcess::getHostNamesWithHSTSCache): Deleted.
* UIProcess/API/C/mac/WKContextPrivateMac.h:
* UIProcess/API/C/mac/WKContextPrivateMac.mm:
(WKContextResetHSTSHosts):
(WKContextResetHSTSHostsAddedAfterDate):
* UIProcess/API/Cocoa/_WKProcessPoolConfiguration.h:
* UIProcess/API/Cocoa/_WKWebsiteDataStoreConfiguration.h:
* UIProcess/API/Cocoa/_WKWebsiteDataStoreConfiguration.mm:
(-[_WKWebsiteDataStoreConfiguration hstsStorageDirectory]):
(-[_WKWebsiteDataStoreConfiguration setHSTSStorageDirectory:]):
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::privateBrowsingSession): Deleted.
(WebKit::WebProcessPool::resetHSTSHosts): Deleted.
(WebKit::WebProcessPool::resetHSTSHostsAddedAfterDate): Deleted.
* UIProcess/WebProcessPool.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::resolveDirectoriesIfNecessary):
(WebKit::WebsiteDataStore::parameters):
* UIProcess/WebsiteData/WebsiteDataStore.h:
(WebKit::WebsiteDataStore::resolvedHSTSStorageDirectory const):
* UIProcess/WebsiteData/WebsiteDataStoreConfiguration.cpp:
(WebKit::WebsiteDataStoreConfiguration::copy const):
* UIProcess/WebsiteData/WebsiteDataStoreConfiguration.h:
2020-08-24 Aditya Keerthi <akeerthi@apple.com>
[macOS] Show picker for date and datetime-local input types
https://bugs.webkit.org/show_bug.cgi?id=214946
Reviewed by Darin Adler and Wenson Hsieh.
Created WKDateTimePicker as a wrapper around NSDatePicker. The picker is
displayed in its own NSWindow, ensuring the view is always above the page.
WebPageProxy and WKDateTimePicker communicate through WebDateTimePickerMac,
in order for the picker to be initialized with the correct initial, minimum,
and maximum date, and so that the chosen date can be sent back to the
WebProcess.
Added IPC messages to enable communication between the UIProcess and the
WebProcess necessary for showing and hiding the picker.
* Sources.txt:
* SourcesCocoa.txt:
* UIProcess/PageClient.h:
* UIProcess/WebDateTimePicker.cpp: Added.
(WebKit::WebDateTimePicker::WebDateTimePicker):
(WebKit::WebDateTimePicker::~WebDateTimePicker):
(WebKit::WebDateTimePicker::endPicker):
* UIProcess/WebDateTimePicker.h: Added.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::showDateTimePicker):
(WebKit::WebPageProxy::endDateTimePicker):
(WebKit::WebPageProxy::didChooseDate):
(WebKit::WebPageProxy::didEndDateTimePicker):
(WebKit::WebPageProxy::closeOverlayedViews):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/ios/PageClientImplIOS.h:
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::createDateTimePicker):
* UIProcess/mac/PageClientImplMac.h:
* UIProcess/mac/PageClientImplMac.mm:
(WebKit::PageClientImpl::createDateTimePicker):
* UIProcess/mac/WebDateTimePickerMac.h: Added.
* UIProcess/mac/WebDateTimePickerMac.mm: Added.
(WebKit::WebDateTimePickerMac::create):
(WebKit::WebDateTimePickerMac::~WebDateTimePickerMac):
(WebKit::WebDateTimePickerMac::WebDateTimePickerMac):
(WebKit::WebDateTimePickerMac::endPicker):
(WebKit::WebDateTimePickerMac::showDateTimePicker):
(WebKit::WebDateTimePickerMac::didChooseDate):
(-[WKDateTimePickerWindow initWithContentRect:styleMask:backing:defer:]):
(-[WKDateTimePickerWindow canBecomeKeyWindow]):
(-[WKDateTimePickerWindow hasKeyAppearance]):
(-[WKDateTimePickerWindow shadowOptions]):
(-[WKDateTimePicker initWithParams:inView:]):
(-[WKDateTimePicker showPicker:]):
(-[WKDateTimePicker invalidate]):
(-[WKDateTimePicker didChooseDate:]):
(-[WKDateTimePicker dateFormatStringForType:]):
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::createDateTimeChooser):
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebCoreSupport/WebDateTimeChooser.cpp: Added.
(WebKit::WebDateTimeChooser::WebDateTimeChooser):
(WebKit::WebDateTimeChooser::didChooseDate):
(WebKit::WebDateTimeChooser::didEndChooser):
(WebKit::WebDateTimeChooser::endChooser):
(WebKit::WebDateTimeChooser::showChooser):
* WebProcess/WebCoreSupport/WebDateTimeChooser.h: Added.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::setActiveDateTimeChooser):
(WebKit::WebPage::didChooseDate):
(WebKit::WebPage::didEndDateTimePicker):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-08-24 Adrian Perez de Castro <aperez@igalia.com>
Non-unified build fixes, late August 2020 edition
https://bugs.webkit.org/show_bug.cgi?id=215768
Unreviewed build fix.
* UIProcess/WebURLSchemeTask.cpp: Add missing SharedBufferDataReference.h header.
2020-08-24 Youenn Fablet <youenn@apple.com>
Cocoa WebSocketTask should expose WebSocket server extensions
https://bugs.webkit.org/show_bug.cgi?id=215696
Reviewed by Darin Adler.
Send back to WebProcess the value of server WebSocket extensions.
* NetworkProcess/cocoa/WebSocketTaskCocoa.mm:
(WebKit::WebSocketTask::didConnect):
2020-08-24 Carlos Garcia Campos <cgarcia@igalia.com>
Unreviewed. Fix GTK4 build
* PlatformGTK.cmake:
* UIProcess/API/gtk/WebKitAuthenticationDialog.cpp:
(webkitAuthenticationDialogInitialize):
* UIProcess/API/gtk/WebKitScriptDialogImpl.cpp:
(webkitScriptDialogImplConstructed):
(webkit_script_dialog_impl_class_init):
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseEvent):
(webkit_web_view_base_class_init):
* UIProcess/API/gtk/WebKitWebViewDialog.cpp:
(webkitWebViewDialogConstructed):
(webkitWebViewDialogSetChild):
* UIProcess/Inspector/gtk/WebKitInspectorWindow.cpp:
(webkit_inspector_window_init):
* UIProcess/gtk/WebDataListSuggestionsDropdownGtk.cpp:
(WebKit::WebDataListSuggestionsDropdownGtk::show):
* UIProcess/gtk/WebPageProxyGtk.cpp:
(WebKit::WebPageProxy::bindAccessibilityTree):
* UIProcess/gtk/WebPopupMenuProxyGtk.cpp:
(WebKit::WebPopupMenuProxyGtk::showPopupMenu):
2020-08-23 Wenson Hsieh <wenson_hsieh@apple.com>
Programmatic selection of text in a text field causes the highlight overlay to spill out
https://bugs.webkit.org/show_bug.cgi?id=215647
<rdar://problem/67404979>
Reviewed by Darin Adler.
UIKit consults the SPI method `-_selectionClipRect` on WKContentView to determine the maximum bounds in which it
is allowed to show text selection UI (i.e. the text caret view, selection highlight views, and selection
grabbers). Normally, when contentEditable elements and editable text form controls are focused, we plumb the
bounds of the focused element to the UI process via `focusedElementRect` in editor state's post layout data.
However, in this case, the selection is inside a readonly text field that is *not* focused; this causes us to
return `CGRectNull` from `-_selectionClipRect`, which means that UIKit selection UI is not clipped at all and
instead overflows the bounds of the input element.
To fix this, rearrange some logic in `WebPage::getPlatformEditorState` in `WebPageIOS.mm` such that we compute
and send the selection clipping rect if the selection is inside a text form control, even if it is not the
focused element. In doing this, we also rename some confusingly-named members in `EditorState::PostLayoutData`
(see below for more detail).
Test: editing/selection/ios/select-all-in-readonly-input-does-not-overflow.html
* Platform/spi/ios/UIKitSPI.h:
Add a declaration for `-_selectionClipRect`.
* Shared/EditorState.cpp:
(WebKit::EditorState::PostLayoutData::encode const):
(WebKit::EditorState::PostLayoutData::decode):
(WebKit::operator<<):
* Shared/EditorState.h:
Split the existing rect member `focusedElementRect` into two: `selectionClipRect` on iOS, and
`selectionBoundingRect` on macOS. Previously, `focusedElementRect` was set to the focused element's bounding
rect on iOS (as expected), but on macOS, we set to the bounds of the ranged or caret selection; we proceed to
use this rect to mean the bounds of the selection anyways in macOS-specific code, so it makes more sense to just
move this into the macOS-specific section.
Additionally, after the below change in WebPageIOS, `focusedElementRect` is no longer specific to the focused
element, so rename it instead to `selectionClipRect` instead. In `WKContentViewInteraction.mm`, this rect is
effectively only used in two ways: as the selection clip rect for UIKit, and to determine if the selection would
be completely clipped anyways (in which case we suppress UIKit text interactions), so `selectionClipRect` is a
name that is suitable for both purposes.
* UIProcess/API/mac/WKWebViewTestingMac.mm:
(-[WKWebView _candidateRect]):
* UIProcess/Cocoa/WebViewImpl.mm:
Rename `focusedElementRect` to `selectionBoundingRect` on macOS.
(WebKit::WebViewImpl::handleRequestedCandidates):
* UIProcess/ios/WKContentViewInteraction.mm:
Rename `focusedElementRect` to `selectionClipRect` on iOS.
(WebKit::WKSelectionDrawingInfo::WKSelectionDrawingInfo):
(-[WKContentView _selectionClipRect]):
Change this to not require a focused element when returning the selection clip rect, and instead return the
editor state's selection clip rect as long as it is not empty.
(-[WKContentView _updateSelectionAssistantSuppressionState]):
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::getPlatformEditorState const):
Refactor this code to compute and send selectionClipRect as long as the selection is inside a form control or
editing host (not just when there is a focused element). Also leave a drive-by FIXME about using the focused
element (instead of the selection container node) when computing the caret color.
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::getPlatformEditorState const):
2020-08-19 Darin Adler <darin@apple.com>
Move node geometry functions from Range to RenderObject
https://bugs.webkit.org/show_bug.cgi?id=215677
Reviewed by Zalan Bujtas.
* WebProcess/InjectedBundle/DOM/InjectedBundleRangeHandle.cpp:
(WebKit::InjectedBundleRangeHandle::boundingRectInWindowCoordinates const):
Call unionRectIgnoringZeroRects(RenderObject::absoluteBorderAndTextRects).
(WebKit::InjectedBundleRangeHandle::renderedImage): Call
unionRectIgnoringZeroRects(RenderObject::absoluteBorderAndTextRects).
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::requestEvasionRectsAboveSelection): Call
unionRect(RenderObject::absoluteTextRects).
(WebKit::WebPage::requestDocumentEditingContext): Call
unionRect(RenderObject::absoluteTextRects).
2020-08-17 Darin Adler <darin@apple.com>
Create documentOrder function, start refactoring to use it instead of Range::compare functions
https://bugs.webkit.org/show_bug.cgi?id=215551
Reviewed by Chris Dumez.
* WebProcess/InjectedBundle/API/mac/WKDOMRange.mm:
(-[WKDOMRange rangeByExpandingToWordBoundaryByCharacters:inDirection:]):
Use makeDeprecatedLegacyPosition instead of Range::start/endPosition.
This was the only place in the WebKit framework still using
Range::start/endPosition and over time we would like to move from
makeDeprecatedLegacyPosition to makeContainerOffsetPosition, so it's good
to have it be explicit here rather than hidden inside the Range class.
2020-08-22 Philippe Normand <pnormand@igalia.com>
[GTK][WPE] Compilation warnings with clang/gcc 10
https://bugs.webkit.org/show_bug.cgi?id=215699
Reviewed by Darin Adler.
* UIProcess/API/wpe/PageClientImpl.cpp:
(WebKit::PageClientImpl::doneWithTouchEvent): Use null input touch event enum value for touch event handling.
2020-08-21 Wenson Hsieh <wenson_hsieh@apple.com>
[macOS] Unable to copy text from the function browser panel in Numbers
https://bugs.webkit.org/show_bug.cgi?id=215740
<rdar://problem/65189303>
Reviewed by Darin Adler.
On some versions of Xcode, using Interface Builder to set up WKWebViews causes the WKWebView to contain two
WKFlippedViews as subviews; one of these is created and added in the normal initialization path of the web view
underneath WebViewImpl's constructor, and the other is created and added underneath the superclass (NSView's)
`-initWithCoder:` implementation.
This causes issues when hit-testing WKWebView using the `-hitTest:` method, which is expected to return the
WKWebView itself instead of the inner WKFlippedView. The logic that tries to implement this behavior is in
`WebViewImpl::hitTest`, which calls into WKWebView's `-hitTest:` method and returns the WKWebView if the hit-
tested view ended up being equal to the `m_layerHostingView`. Since we end up with two layer hosting views (only
one of which is the real `m_layerHostingView`), we fail the check and end up returning the other WKFlippedView.
In the context of this bug, this erroneous hit-test causes AppKit to ask if the WKFlippedView can become the
first responder upon mousedown (which it cannot), and so the window's first responder remains the same (instead
of changing to WKWebView).
To fix this, adjust the constructor of WebViewImpl to avoid making an extra layer hosting view in the case where
decoding the WKWebView already initialized and placed the layer hosting view, and just use the existing one
instead.
Test: WKWebView.HitTestAfterInitializingFromCoder
* UIProcess/Cocoa/WebViewImpl.mm:
(-[WKFlippedView initWithFrame:]):
(-[WKFlippedView initWithCoder:]):
(-[WKFlippedView _commonInitialize]):
Override both designated initialization codepaths so that we call the `-_commonInitialize` helper, which sets
the autoresizing mask. This was previously called immediately after creating the WKFlippedView in the
constructor of WebViewImpl.
(WebKit::WebViewImpl::WebViewImpl):
2020-08-21 Andy Estes <aestes@apple.com>
IPC::encodeSharedBuffer combines SharedBuffer data segments when copying to SharedMemory
https://bugs.webkit.org/show_bug.cgi?id=215737
<rdar://problem/67568099>
Reviewed by Darin Adler.
IPC::encodeSharedBuffer called SharedBuffer::data() when copying data into SharedMemory,
requiring the SharedBuffer to combine its data segments. This can be avoided by using
SharedMemory::copyBuffer() instead.
* Platform/SharedMemory.cpp:
(WebKit::SharedMemory::copyBuffer): Modernized to use auto and range-based for loops.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::encodeSharedBuffer): Used SharedMemory::copyBuffer() to create a copy of the buffer in
a new SharedMemory. Calling this function does not combine the buffer's data segments.
* Platform/SharedMemory.h: Added a FIXME for how SharedMemory's factory functions could be
improved.
2020-08-21 Jiten Mehta <jmehta@apple.com>
Use closeReason provided in didCloseWithCode delegate for the NSURLSessionWebSocket code path
https://bugs.webkit.org/show_bug.cgi?id=215519
Reviewed by Youenn Fablet.
Covered by existing protocol tests with NSURLSession code path enabled.
http/tests/websocket/tests/hybi/workers/close-code-and-reason.html
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(-[WKNetworkSessionDelegate URLSession:webSocketTask:didCloseWithCode:reason:]):
2020-08-21 Andy Estes <aestes@apple.com>
REGRESSION (r257667): 1.9x more CPU time in IPC::SharedBufferDataReference decoding during Netflix playback
https://bugs.webkit.org/show_bug.cgi?id=215232
<rdar://problem/66637920>
Reviewed by Geoff Garen.
Prior to r257667, an IPC message defined a IPC::SharedBufferDataReference argument to allow
its sender to easily encode SharedBuffer data as a DataReference. SharedBufferDataReference
cost the same to encode as a DataReference -- both copied data into the Encoder -- but did
so by iterating over the SharedBuffer's data segments rather than combining them into a
contiguous allocation. The message argument was received as a decoded IPC::DataReference,
which stores a pointer into the IPC::Decoder's buffer rather than making a copy.
In r257667, IPC::SharedBufferDataReference switched its encoder to use IPC::ArgumentCoder's
SharedBuffer specialization, which works by allocating a SharedMemory buffer, copying
SharedBuffer data into it, creating a SharedMemory::IPCHandle from it, and encoding the
handle. It makes only one copy like the old encoder did, but pays the added cost of
combining SharedBuffer data segments prior to copying by needlessly calling
SharedBuffer::data() (this should be fixed independently).
Since this serialization could no longer be decoded by DataReference's decoder,
SharedBufferDataReference implemented a decoder, also in terms of ArgumentCoder's SharedBuffer
specialization. It works by mapping a SharedMemory buffer from a decoded
SharedMemory::IPCHandle and copying that buffer into a new SharedBuffer. This copy is a cost
that the old decoder did not incur, but in some cases the message receiver would have made
such a copy anyway and can instead just ref the decoded SharedBuffer. In other cases though,
the receiver either made no copies or couldn't use the decoded SharedBuffer to avoid making
its own copy so creating the SharedBuffer was pure added cost.
There are additional costs from allocating, mapping, and deleting shared memory and Mach
send rights, but they are offset by savings due to smaller Encoder and Decoder buffers. The
additional SharedBuffer combining and copying is not offset elsewhere, and on net makes
sending and receiving IPC::SharedBufferDataReferences use up to 2x more CPU time when
playing a video on netflix.com in Safari.
Resolve this slowdown by reverting the argument encoding changes made to
IPC::SharedBufferDataReference in r257667. For compatibility with code added after r257667,
the current IPC::SharedBufferDataReference was copied to IPC::SharedBufferCopy and retains
the behavior of copying into a SharedBuffer when decoded. IPC::SharedBufferDataReference now
has the pre-r257667 behavior of decoding to an IPC::DataReference. Messages added after
r257667 whose receivers were designed to handle a SharedBuffer were changed to use
IPC::SharedBufferCopy, and all other receivers of messages with IPC::SharedBufferDataReference
arguments were changed to handle an IPC::DataReference.
* GPUProcess/media/RemoteCDMInstanceProxy.cpp:
* GPUProcess/media/RemoteCDMInstanceProxy.h:
* GPUProcess/media/RemoteCDMInstanceProxy.messages.in:
* GPUProcess/media/RemoteCDMInstanceSessionProxy.cpp:
* GPUProcess/media/RemoteCDMInstanceSessionProxy.h:
* GPUProcess/media/RemoteCDMInstanceSessionProxy.messages.in:
* GPUProcess/media/RemoteCDMProxy.cpp:
* GPUProcess/media/RemoteCDMProxy.h:
* GPUProcess/media/RemoteLegacyCDMSessionProxy.cpp:
* GPUProcess/media/RemoteLegacyCDMSessionProxy.h:
* GPUProcess/media/RemoteLegacyCDMSessionProxy.messages.in:
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::platformRegisterAttachment):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::registerAttachmentIdentifierFromData):
(WebKit::WebPageProxy::platformRegisterAttachment):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/WebPasteboardProxy.h:
* UIProcess/WebPasteboardProxy.messages.in:
* UIProcess/gtk/WebPasteboardProxyGtk.cpp:
* WebProcess/GPU/media/RemoteCDMInstance.cpp:
* WebProcess/GPU/media/RemoteCDMInstanceSession.cpp:
* WebProcess/GPU/media/RemoteCDMInstanceSession.h:
* WebProcess/GPU/media/RemoteCDMInstanceSession.messages.in:
* WebProcess/GPU/media/RemoteLegacyCDMSession.cpp:
* WebProcess/GPU/media/RemoteLegacyCDMSession.h:
* WebProcess/GPU/media/RemoteLegacyCDMSession.messages.in:
* WebProcess/WebCoreSupport/WebEditorClient.cpp:
* WebProcess/WebCoreSupport/WebPlatformStrategies.cpp: Replaced
IPC::SharedBufferDataReference with IPC::SharedBufferCopy.
* NetworkProcess/NetworkResourceLoader.cpp:
* NetworkProcess/ServiceWorker/ServiceWorkerFetchTask.cpp:
* NetworkProcess/ServiceWorker/ServiceWorkerFetchTask.h:
* NetworkProcess/ServiceWorker/ServiceWorkerFetchTask.messages.in:
* WebProcess/Network/WebResourceLoader.cpp:
* WebProcess/Network/WebResourceLoader.h:
* WebProcess/Network/WebResourceLoader.messages.in:
* WebProcess/Storage/WebServiceWorkerFetchTaskClient.cpp: Removed the DidReceiveSharedBuffer
message and reverted the DidReceiveData message back to taking an
IPC::SharedBufferDataReference.
* Platform/IPC/Encoder.h: Made Encoder::reserve(size_t) public so
SharedBufferCopy::encode(Encoder&) can call it.
* Platform/IPC/HandleMessage.h: Restored the CodingType specialization that decodes
SharedBufferDataReference arguments as DataReferences.
* Platform/IPC/SharedBufferCopy.cpp:
* Platform/IPC/SharedBufferCopy.h: Copied from SharedBufferDataReference.h. Implemented the
interface and decoding behavior that SharedBufferDataReference had between r257667 and this
change.
* Platform/IPC/SharedBufferDataReference.cpp:
* Platform/IPC/SharedBufferDataReference.h: Reverted to the pre-r257667 interface and
decoding behvior.
* Scripts/webkit/messages.py: Removed an unneeded special case.
* Sources.txt:
* WebKit.xcodeproj/project.pbxproj: Added SharedBuffer{Copy,DataReference}.cpp.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::dataCallback):
(WebKit::WebPageProxy::finishedLoadingIcon):
* UIProcess/WebPageProxy.h:
* WebProcess/WebPage/WebPage.cpp:
* WebProcess/WebPage/WebPage.h: Used `const IPC::DataReference&` in message receiver
signatures for SharedBufferDataReference arguments.
* UIProcess/WebURLSchemeTask.cpp:
* WebProcess/GPU/media/RemoteAudioBusData.h:
* WebProcess/GPU/media/RemoteCDM.cpp:
* WebProcess/Network/WebLoaderStrategy.cpp: Removed unused includes.
2020-08-20 Peng Liu <peng.liu6@apple.com>
Mac Catalyst Apps require "playsinline" attribute to support playing videos in the inline mode
https://bugs.webkit.org/show_bug.cgi?id=215724
Reviewed by Tim Horton.
Currently WebKit::currentUserInterfaceIdiomIsPad() returns NO on Mac Catalyst apps.
Therefore, the behavior of Mac Catalyst apps regarding the inline media playback
support is not consistent with iPad apps.
This patch removes the usage of WebKit::currentUserInterfaceIdiomIsPad() to decide
the settings of the inline media playback support on Mac Catalyst apps.
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration init]):
2020-08-20 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK][WPE] Debug crashes in backdrop filter tests
https://bugs.webkit.org/show_bug.cgi?id=215209
Reviewed by Adrian Perez de Castro.
* WebProcess/WebPage/CoordinatedGraphics/CompositingCoordinator.cpp:
(WebKit::CompositingCoordinator::~CompositingCoordinator): Use invalidateCoordinator() instead of setCoordinator()
that has been removed.
(WebKit::CompositingCoordinator::createGraphicsLayer): Use setCoordinatorIncludingSubLayersIfNeeded() instead of
attaching the layer directly.
(WebKit::CompositingCoordinator::attachLayer): Do not call setCoordinator(), the caller is expected to set the
coordinator.
2020-08-20 Per Arne Vollan <pvollan@apple.com>
[macOS] Web pages are not responding correctly to changes in "Reduce motion" setting
https://bugs.webkit.org/show_bug.cgi?id=215664
Reviewed by Darin Adler.
This happens because there is a race between the preference change and notification being received in the WebContent
process. This race is usually won by the notification, which then reads the wrong preference value when being handled.
This bug was introduced when access to the preference daemon was closed in the WebContent process, and preferences
were updated from the UI process. This patch addresses this issue by re-posting the notification when the preference
value is updated in the WebContent process.
API test: WebKit.AccessibilityReduceMotion
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::reduceMotionPreferenceKey):
(WebKit::dispatchSimulatedNotificationsForPreferenceChange):
2020-08-20 James Savage <james.savage@apple.com>
Could not find module 'WebKit' for target 'armv7-apple-ios'
<https://bugs.webkit.org/show_bug.cgi?id=215190>
<rdar://problem/65642049>
Reviewed by Brady Eidson.
The config changes here got added to the wrong config file. They need to
be for the overlay, not its tests.
* SwiftOverlay/Configurations/WebKitSwiftOverlay.xcconfig:
* SwiftOverlay/Configurations/WebKitSwiftOverlayTests.xcconfig:
2020-08-20 Aditya Keerthi <akeerthi@apple.com>
Add runtime setting for editable components in date/time inputs
https://bugs.webkit.org/show_bug.cgi?id=215705
Reviewed by Wenson Hsieh.
Added DateTimeInputsEditableComponentsEnabled as a runtime setting. This
setting is enabled by default on macOS.
* Shared/WebPreferences.yaml:
* Shared/WebPreferencesDefaultValues.h:
2020-08-20 Chris Dumez <cdumez@apple.com>
Sandbox violation when calling [NSHTTPCookieStorage sharedHTTPCookieStorage] from the WebProcess
https://bugs.webkit.org/show_bug.cgi?id=215701
<rdar://problem/67069826>
Reviewed by Geoffrey Garen.
The sandbox does not allow calling [NSHTTPCookieStorage sharedHTTPCookieStorage] from the WebProcess.
createPrivateStorageSession() was calling [NSHTTPCookieStorage sharedHTTPCookieStorage] to retrieve
the cookieAcceptPolicy and createPrivateStorageSession() is used by the cookie cache in the WebProcess.
To address the issue, we now allow the caller of createPrivateStorageSession() to provide the
cookieAcceptPolicy and we have the cookie cache retrieve the cookieAcceptPolicy from the
NetworkProcessConnection.
* WebProcess/Network/NetworkProcessConnection.h:
(WebKit::NetworkProcessConnection::cookieAcceptPolicy const):
* WebProcess/WebPage/Cocoa/WebCookieCacheCocoa.mm:
(WebKit::WebCookieCache::inMemoryStorageSession):
2020-08-20 Youenn Fablet <youenn@apple.com>
Align console log message of new WebSocket code path with the legacy code path
https://bugs.webkit.org/show_bug.cgi?id=215695
Reviewed by Alex Christensen.
No observable change of behavior from the page.
Console log messages appear in expected.txt files when NSURLSession code path is enabled.
* WebProcess/Network/WebSocketChannel.cpp:
(WebKit::WebSocketChannel::connect):
(WebKit::WebSocketChannel::didReceiveMessageError):
* WebProcess/Network/WebSocketChannel.h:
2020-08-20 Lauro Moura <lmoura@igalia.com>
REGRESSION(r265856) [GTK][WPE] hybi websockets tests failing
https://bugs.webkit.org/show_bug.cgi?id=215679
Reviewed by Carlos Garcia Campos.
r265856 changed WebSocketChannel to expect AbnormalClosure when the
closing handshake is not received.
Covered by existing tests.
* NetworkProcess/soup/WebSocketTaskSoup.cpp:
(WebKit::WebSocketTask::didFail):
2020-08-20 Youenn Fablet <youenn@apple.com>
WebSocketTask::close should correctly compute reason byte size in UTF-8
https://bugs.webkit.org/show_bug.cgi?id=215645
Reviewed by Darin Adler.
Covered by http/tests/websocket/tests/hybi/close.html with NSURLSession code path enabled.
* NetworkProcess/cocoa/WebSocketTaskCocoa.mm:
(WebKit::WebSocketTask::close):
Instead of calling sizeInBytes, we should just get the utf8 CString and gets its size directly.
2020-08-19 Fujii Hironori <Hironori.Fujii@sony.com>
IPC message can't be decoded due to uninitialized NavigationActionData member variables with the latest MSVC
https://bugs.webkit.org/show_bug.cgi?id=215662
Reviewed by Darin Adler.
WinCairo WebKit2 is crashing in AuxiliaryProcess::didReceiveInvalidMessage
if it is compiled by the latest MSVC (Visual Studio 2019 16.7) with optimizer (/O2).
LockHistory is a bool based enum type. NavigationActionData
doesn't initialize the member variable of it. IPC::Encoder
converts a LockHistory value to the underlying type bool. The
uninitialized value is converted 0 or 1 in GCC, Clang and old
MSVC. However, the latest MSVC doesn't convert and IPC::Decoder
detects an invalid enum value.
* Shared/NavigationActionData.h: Added initializes for userGestureTokenIdentifier, lockHistory and lockBackForwardList.
2020-08-19 Brady Eidson <beidson@apple.com>
Improperly linearized PDFs show up as blank gray on Big Sur.
rdar://problem/67117968 and https://bugs.webkit.org/show_bug.cgi?id=215683
Reviewed by Darin Adler.
Will add a new test once an in-house expert helps us construct an example PDF that
we will have the legal right to commit.
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::threadEntry): When PDFKit indicates a PDFDocument is *improperly* linearized
(as opposed to not linearized at all), we failed to appopriately notify the main thread that
it needs to start handling the PDF.
Failure to create a PDFDocument is equivalent to receiving the "nonlinearized sentinel", so
adding that call fixes this issue.
2020-08-19 Simon Fraser <simon.fraser@apple.com>
Fix possible crash when webAnimationsCSSIntegrationEnabled is false
https://bugs.webkit.org/show_bug.cgi?id=215668
<rdar://problem/67402003>
Reviewed by Dean Jackson.
When webAnimationsCSSIntegrationEnabled is false, GraphicsLayerCA::setupAnimation() doesn't
set a timing function on the animation, causing a null de-ref when the animation is being encoded.
Fix the encoder to handle null timing functions.
Test: animations/legacy-encoding-timing-function.html
* WebProcess/WebPage/RemoteLayerTree/PlatformCAAnimationRemote.mm:
(WebKit::PlatformCAAnimationRemote::Properties::encode const):
(WebKit::PlatformCAAnimationRemote::Properties::decode):
2020-08-19 Alex Christensen <achristensen@webkit.org>
Fix null pointer crashes in network process after r265835
​https://bugs.webkit.org/show_bug.cgi?id=215626
<rdar://problem/67268892>
Call the correct CompletionHandler. completionHandler has already been moved from at this point.
This crash was being hit in the test, but the test recovered from it.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::continueDidReceiveChallenge):
2020-08-19 Kate Cheney <katherine_cheney@apple.com>
Third party domains are not stored in the case of back/forward navigations
https://bugs.webkit.org/show_bug.cgi?id=215595
<rdar://problem/66642893>
Reviewed by Chris Dumez.
This patch adds the ability to retrieve loaded third party domains from the page
synchronously to be stored in the CachedPage in the case of a
back/forward navigation.
It also changes the functions loadedThirdPartyDomains and
getLoadedThirdPartyDomains to loadedSubresourceDomains and
getLoadedSubresourceDomains respectively based on the conversation from
https://bugs.webkit.org/show_bug.cgi?id=215595.
* UIProcess/API/C/WKPage.cpp:
(WKPageLoadedSubresourceDomains):
(WKPageClearLoadedSubresourceDomains):
(WKPageLoadedThirdPartyDomains): Deleted.
(WKPageClearLoadedThirdPartyDomains): Deleted.
* UIProcess/API/C/WKPagePrivate.h:
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
(-[WKWebsiteDataStore _loadedSubresourceDomainsFor:completionHandler:]):
(-[WKWebsiteDataStore _clearLoadedSubresourceDomainsFor:]):
(-[WKWebsiteDataStore _loadedThirdPartyDomainsFor:completionHandler:]): Deleted.
(-[WKWebsiteDataStore _clearLoadedThirdPartyDomainsFor:]): Deleted.
* UIProcess/API/Cocoa/WKWebsiteDataStorePrivate.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::getLoadedSubresourceDomains):
(WebKit::WebPageProxy::clearLoadedSubresourceDomains):
(WebKit::WebPageProxy::loadedThirdPartyDomains): Deleted.
(WebKit::WebPageProxy::clearLoadedThirdPartyDomains): Deleted.
* UIProcess/WebPageProxy.h:
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::loadedSubresourceDomains const):
* WebProcess/WebCoreSupport/WebFrameLoaderClient.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::didCommitLoad):
(WebKit::WebPage::didLoadFromRegistrableDomain):
(WebKit::WebPage::getLoadedSubresourceDomains):
(WebKit::WebPage::clearLoadedSubresourceDomains):
(WebKit::WebPage::loadedThirdPartyDomains): Deleted.
(WebKit::WebPage::clearLoadedThirdPartyDomains): Deleted.
* WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::loadedSubresourceDomains const):
* WebProcess/WebPage/WebPage.messages.in:
2020-08-19 Chris Dumez <cdumez@apple.com>
REGRESSION (r261407): Sharing rich links from Twitter/Netflix via Messages is unreliable
https://bugs.webkit.org/show_bug.cgi?id=215656
<rdar://problem/67087352>
Reviewed by Tim Horton.
The WKWebView._clientNavigationsRunAtForegroundPriority SPI was not working for the
MessagesViewService because the view is unparented and PageClient::isApplicationVisible()
would not work return accurate results when [view window] returns nil. It is very
unfortunate to have to rely on the view or the view's window to determine application
visibility so this patch fixes this. We now rely on UIKit's visibility endowment to
determine if the UIProcess is visible, using RunningBoard API. I have verified that the
visibility endowment properly gets propagated from the host application to the view
service when the application is a view service and I have verified that this detects
the MessagesViewService as visible for the use-case at <rdar://problem/67087352>, which
causes the WKWebView._clientNavigationsRunAtForegroundPriority SPI to work as intended.
Note that when the application does not have the visibility endowment, we still allow
it to take a foreground assertion for client navigations if the process is allowed to
run in the background for an extended period of time, according to RunningBoard. This
allows the SPI to work for daemons (like maild) which rely on it.
* UIProcess/PageClient.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::runJavaScriptInFrameInScriptWorld):
* UIProcess/ios/PageClientImplIOS.h:
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::canTakeForegroundAssertions):
(WebKit::PageClientImpl::isApplicationVisible): Deleted.
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::shouldForceForegroundPriorityForClientNavigation const):
2020-08-19 Youenn Fablet <youenn@apple.com>
WebSocketChannel::didClose should correctly compute whether a closing handshake was received
https://bugs.webkit.org/show_bug.cgi?id=215644
Reviewed by Darin Adler.
Covered by LayoutTests/http/tests/websocket/tests/hybi/close-code-and-reason.html with NSURLSession code path enabled.
* WebProcess/Network/WebSocketChannel.cpp:
(WebKit::WebSocketChannel::didClose):
As noted by Jiten Mehta, absence of closing handshake from the server is notified by WebSocketChannel::CloseEventCodeAbnormalClosure.
Update check accordingly.
2020-08-18 Alex Christensen <achristensen@webkit.org>
WKWebViews using fastServerTrustEvaluationEnabled should only allow legacy TLS for main resource loads
https://bugs.webkit.org/show_bug.cgi?id=215626
<rdar://problem/67268892>
Reviewed by Darin Adler.
We have introduced public API webView:authenticationChallenge:shouldAllowDeprecatedTLS: in WKNavigationDelegate to allow
applications to choose whether to allow TLS 1.0 or 1.1 connections. We don't want to break this API or break existing third party
apps that load pages that load third party subresources that use TLS 1.0 or 1.1.
However, we do want Safari, which uses fastServerTrustEvaluationEnabled SPI, to silently fail subresource loads that use TLS 1.0 or 1.1.
This matches the current behavior of Chrome and Firefox, which was not implemented in those other browsers when we decided to ask about subresources.
Covered by an API test.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::continueDidReceiveChallenge):
2020-08-18 Jer Noble <jer.noble@apple.com>
[Mac] Unhandled keyDown: events cause a system beep in full screen mode
https://bugs.webkit.org/show_bug.cgi?id=215596
<rdar://problem/65823962>
Reviewed by Darin Adler.
* UIProcess/mac/WKFullScreenWindowController.mm:
(-[WKFullScreenWindowController noResponderFor:]):
2020-08-18 Youenn Fablet <youenn@apple.com>
NetworkSocketChannel should delay sending back error message as done by NetworkSocketStream
https://bugs.webkit.org/show_bug.cgi?id=215563
Reviewed by Alex Christensen.
Apply same policy as NetworkSocketStream in case of WebSocket connection error.
Covered by http/tests/websocket/tests/hybi/closed-port-delay.html with NSURLSession WebSocket code path.
* NetworkProcess/NetworkSocketChannel.cpp:
(WebKit::NetworkSocketChannel::NetworkSocketChannel):
(WebKit::NetworkSocketChannel::didClose):
(WebKit::NetworkSocketChannel::didReceiveMessageError):
(WebKit::NetworkSocketChannel::sendDelayedError):
* NetworkProcess/NetworkSocketChannel.h:
2020-08-17 Chris Dumez <cdumez@apple.com>
AudioContext.getOutputTimestamp() is missing
https://bugs.webkit.org/show_bug.cgi?id=215591
Reviewed by Darin Adler.
* GPUProcess/media/RemoteAudioDestinationManager.cpp:
* WebProcess/GPU/media/RemoteAudioBusData.h:
(WebKit::RemoteAudioBusData::encode const):
(WebKit::RemoteAudioBusData::decode):
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::renderBuffer):
2020-08-17 Per Arne Vollan <pvollan@apple.com>
Assert failure after r265715
https://bugs.webkit.org/show_bug.cgi?id=215592
Reviewed by Darin Adler.
The change set r265715 introduced an assert failure in AuthenticationManager::initializeConnection. This
happens because after r265715 the new XPC event handler set in AuthenticationManager::initializeConnection,
will no longer be called on the main thread. This patch addresses this issue by dispatching this work on
the main thread.
No new tests, covered by existing tests.
* Shared/Authentication/cocoa/AuthenticationManagerCocoa.mm:
(WebKit::AuthenticationManager::initializeConnection):
2020-08-17 Per Arne Vollan <pvollan@apple.com>
[Cocoa] Avoid waiting for Launch Services on every load
https://bugs.webkit.org/show_bug.cgi?id=215569
Reviewed by Darin Adler.
Currently, we are waiting for the Launch Services database to be present on every load in the WebContent process.
It should be sufficient to wait only on the first load. Since we now are waiting only once per WebContent process,
increase the maximum wait time from 1s to 5s.
No new tests, covered by existing tests.
* WebProcess/WebPage/Cocoa/WebPageCocoa.mm:
(WebKit::WebPage::platformDidReceiveLoadParameters):
2020-08-17 Jer Noble <jer.noble@apple.com>
[Mac] Add Experimental Feature preference for SW VP9
https://bugs.webkit.org/show_bug.cgi?id=215043
<rdar://problem/66400034>
Reviewed by Darin Adler.
Follow up fix: during a rebase, a line deletion was dropped which causes the SW decoder to
always be registered regardless of the new setting.
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::enableVP9Decoder):
(WebKit::WebProcess::enableVP9SWDecoder):
2020-08-17 Youenn Fablet <youenn@apple.com>
Remove message check in WebSocketTask::readNextMessage
https://bugs.webkit.org/show_bug.cgi?id=215561
Reviewed by Alex Christensen.
This if check was added as a workaround to cases where both error and message were null.
This is no longer the case as per running existing tests with NSURLSession WebSocket code path on BigSur.
* NetworkProcess/cocoa/WebSocketTaskCocoa.mm:
(WebKit::WebSocketTask::readNextMessage):
2020-08-17 Youenn Fablet <youenn@apple.com>
Remove Cocoa WebSocketTask::send workaround
https://bugs.webkit.org/show_bug.cgi?id=215564
Reviewed by Alex Christensen.
Covered by existing tests run in Debug.
* NetworkProcess/cocoa/WebSocketTaskCocoa.mm:
(WebKit::WebSocketTask::sendString):
(WebKit::WebSocketTask::sendData):
2020-08-17 David Kilzer <ddkilzer@apple.com>
Clean up DragApplicationFlags after switch to OptionSet<>
<https://webkit.org/b/215349>
Reviewed by Darin Adler.
Change WebCore::DragApplicationFlags to an enum class.
* UIProcess/API/gtk/DropTargetGtk3.cpp:
(WebKit::DropTarget::drop):
- Switch to use OptionSet<WebCore::DragApplicationFlags>.
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::applicationFlagsForDrag):
* WebProcess/WebPage/WebPage.h:
- Update forward declaration.
2020-08-17 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK][WPE] Remove support for NPAPI plugins
https://bugs.webkit.org/show_bug.cgi?id=215503
Reviewed by Darin Adler.
* PlatformGTK.cmake:
* PluginProcess/unix/PluginControllerProxyUnix.cpp: Removed.
* PluginProcess/unix/PluginProcessMainUnix.cpp: Removed.
* PluginProcess/unix/PluginProcessUnix.cpp: Removed.
* Shared/Plugins/Netscape/unix/NetscapePluginModuleUnix.cpp: Removed.
* Shared/Plugins/unix/PluginSearchPath.cpp: Removed.
* Shared/Plugins/unix/PluginSearchPath.h: Removed.
* SourcesGTK.txt:
* SourcesWPE.txt:
* UIProcess/API/glib/WebKitMimeInfo.cpp:
(webkit_mime_info_ref):
(webkit_mime_info_unref):
(webkit_mime_info_get_mime_type):
(webkit_mime_info_get_description):
(webkit_mime_info_get_extensions):
* UIProcess/API/glib/WebKitMimeInfoPrivate.h: Removed.
* UIProcess/API/glib/WebKitPlugin.cpp:
(webkit_plugin_get_name):
(webkit_plugin_get_description):
(webkit_plugin_get_path):
(webkit_plugin_get_mime_info_list):
* UIProcess/API/glib/WebKitPluginPrivate.h: Removed.
* UIProcess/API/glib/WebKitSettings.cpp:
(webKitSettingsSetProperty):
(webKitSettingsGetProperty):
(webkit_settings_class_init):
(webkit_settings_get_enable_plugins):
(webkit_settings_set_enable_plugins):
* UIProcess/API/glib/WebKitWebContext.cpp:
(webkit_web_context_set_additional_plugins_directory):
(webkit_web_context_get_plugins):
(webkit_web_context_get_plugins_finish):
* UIProcess/API/glib/WebKitWebView.cpp:
(webkitWebViewLoadFail):
* UIProcess/API/glib/WebKitWebsiteData.cpp:
(recordContainsSupportedDataTypes):
(toWebKitWebsiteDataTypes):
* UIProcess/API/glib/WebKitWebsiteDataManager.cpp:
(toWebsiteDataTypes):
* UIProcess/API/gtk/WebKitError.h:
* UIProcess/API/gtk/WebKitMimeInfo.h:
* UIProcess/API/gtk/WebKitPlugin.h:
* UIProcess/API/gtk/WebKitSettings.h:
* UIProcess/API/gtk/WebKitWebContext.h:
* UIProcess/API/gtk/WebKitWebsiteData.h:
* UIProcess/API/wpe/WebKitError.h:
* UIProcess/API/wpe/WebKitMimeInfo.h:
* UIProcess/API/wpe/WebKitPlugin.h:
* UIProcess/API/wpe/WebKitSettings.h:
* UIProcess/API/wpe/WebKitWebContext.h:
* UIProcess/API/wpe/WebKitWebsiteData.h:
* UIProcess/Launcher/glib/BubblewrapLauncher.cpp:
(WebKit::bubblewrapSpawn):
* UIProcess/Launcher/glib/ProcessLauncherGLib.cpp:
(WebKit::ProcessLauncher::launchProcess):
* UIProcess/Plugins/gtk/PluginInfoCache.cpp: Removed.
* UIProcess/Plugins/gtk/PluginInfoCache.h: Removed.
* UIProcess/Plugins/unix/PluginInfoStoreUnix.cpp: Removed.
* UIProcess/Plugins/unix/PluginProcessProxyUnix.cpp: Removed.
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/gtk/WebPageProxyGtk.cpp:
* WebProcess/Plugins/Netscape/unix/NetscapePluginUnix.cpp: Removed.
* WebProcess/Plugins/Netscape/unix/NetscapePluginUnix.h: Removed.
* WebProcess/Plugins/Netscape/unix/PluginProxyUnix.cpp: Removed.
* WebProcess/Plugins/Netscape/x11/NetscapePluginX11.cpp: Removed.
* WebProcess/Plugins/Netscape/x11/NetscapePluginX11.h: Removed.
* WebProcess/Plugins/PluginController.h:
* WebProcess/Plugins/PluginView.cpp:
* WebProcess/Plugins/PluginView.h:
2020-08-17 Youenn Fablet <youenn@apple.com>
Set Sec-WebSocket-Protocol for WebSocket NSURLSession code path
https://bugs.webkit.org/show_bug.cgi?id=215456
<rdar://problem/66996369>
Reviewed by Alex Christensen.
Covered by existing protocol tests with NSURLSession code path enabled.
http/tests/websocket/tests/hybi/set-protocol.html, http/tests/websocket/tests/hybi/workers/multiple-subprotocols.html
Code change originating from Jiten Mehta.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::createWebSocketTask):
Set Sec-WebSocket-Protocol to support WebSocket protocol negotiation.
2020-08-17 Carlos Garcia Campos <cgarcia@igalia.com>
REGRESSION(r265725) [GTK][WPE] Several http/tests timeout/failures
https://bugs.webkit.org/show_bug.cgi?id=215542
Reviewed by Sergio Villar Senin.
The problem is that ShareableResource is now using the dataSize member of IPCHandle to encode/decode the
resource size, but SharedMemoryUnix is not encoding/decoding dataSize.
* Platform/unix/SharedMemoryUnix.cpp:
(WebKit::SharedMemory::IPCHandle::encode const):
(WebKit::SharedMemory::IPCHandle::decode):
2020-08-14 Simon Fraser <simon.fraser@apple.com>
Scrolling sync changes in r261985 regressed CPU usage by ~2 ms/s
https://bugs.webkit.org/show_bug.cgi?id=215529
<rdar://problem/66866163>
Reviewed by Geoff Garen.
r261985 added two code paths that wake up the scrolling thread on every
rendering update (triggered by displayDidRefresh()). One is a ping from
the EventDispatcher thread, the other is a wake-and-block from the main
thread. If the scrolling thread isn't active (no wheel events received recently),
we can avoid both of these to reduce the number of CPU core wakeups.
* WebProcess/WebPage/EventDispatcher.cpp:
(WebKit::EventDispatcher::wheelEvent):
2020-08-15 Tim Horton <timothy_horton@apple.com>
Live Web Content processes do not respect accent color if dynamically changed to "multicolor"
https://bugs.webkit.org/show_bug.cgi?id=215523
<rdar://problem/63941133>
Reviewed by Wenson Hsieh.
* UIProcess/Cocoa/PreferenceObserver.mm:
(-[WKUserDefaults _notifyObserversOfChangeFromValuesForKeys:toValuesForKeys:]):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::setPreferenceValue):
The NSUserDefaults syncing mechanism does not propagate defaults that are
deleted or changed to nil.
Allow nil, and fix up an assert that would fire for the same reason.
2020-08-14 Alex Christensen <achristensen@webkit.org>
Revert r263551
https://bugs.webkit.org/show_bug.cgi?id=215530
<rdar://problem/66961066>
Reviewed by Geoffrey Garen.
I was hoping to fix many developers' bugs, but it caused hangs and more hangs, so back to the status quo.
https://bugs.webkit.org/show_bug.cgi?id=203547 will fix this the right way in the hopefully-not-too-distant future.
* UIProcess/API/APIHTTPCookieStore.cpp:
(API::HTTPCookieStore::setCookies):
(API::HTTPCookieStore::registerForNewProcessPoolNotifications):
(API::HTTPCookieStore::flushDefaultUIProcessCookieStore):
* UIProcess/API/APIHTTPCookieStore.h:
* UIProcess/API/Cocoa/APIHTTPCookieStoreCocoa.mm:
(API::HTTPCookieStore::flushDefaultUIProcessCookieStore):
2020-08-14 Kate Cheney <katherine_cheney@apple.com>
Convert remaining SharedMemory::Handle to SharedMemory::IPCHandle
https://bugs.webkit.org/show_bug.cgi?id=215471
<rdar://problem/67024177>
Reviewed by David Kilzer.
Use the new struct for sending SharedMemory::Handle objects via IPC, so
we can include the exact size of data (SharedMemory::Handle::size()
rounds up to the nearest page), and verify these sizes match when decoding.
In many cases, the size is not sent as a separate argument via IPC, so
we can initialize the IPCHandle object with SharedMemory::Handle::size().
We can remove the additional size parameter when it exists, because
we are now passing IPCHandle which should include the size value.
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRenderer.cpp:
(WebKit::RemoteAudioMediaStreamTrackRenderer::audioSamplesStorageChanged):
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRenderer.h:
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRenderer.messages.in:
* GPUProcess/webrtc/RemoteMediaRecorder.cpp:
(WebKit::RemoteMediaRecorder::audioSamplesStorageChanged):
* GPUProcess/webrtc/RemoteMediaRecorder.h:
* GPUProcess/webrtc/RemoteMediaRecorder.messages.in:
Send SharedMemory::IPCHandle object instead of SharedMemory::Handle
over IPC.
* NetworkProcess/ServiceWorker/WebSWOriginStore.cpp:
(WebKit::WebSWOriginStore::sendStoreHandle):
Since no additional size parameter is being sent in this function, we
can use handle.size() as the IPCHandle::dataSize parameter.
* Platform/SharedMemory.h:
Remove encode/decode functions in SharedMemory::Handle.
* Platform/cocoa/SharedMemoryCocoa.cpp:
(WebKit::SharedMemory::Handle::encode const): Deleted.
(WebKit::SharedMemory::Handle::decode): Deleted.
* Platform/unix/SharedMemoryUnix.cpp:
(WebKit::SharedMemory::IPCHandle::encode const):
(WebKit::SharedMemory::IPCHandle::decode):
* Platform/win/SharedMemoryWin.cpp:
(WebKit::SharedMemory::IPCHandle::encode const):
(WebKit::SharedMemory::IPCHandle::decode):
(WebKit::SharedMemory::Handle::encode const): Deleted.
(WebKit::SharedMemory::Handle::decode): Deleted.
Remove SharedMemory::Handle encode/decode functions so we will not
ever send a SharedMemory::Handle object over IPC accidentally in the future.
* Shared/ShareableBitmap.cpp:
(WebKit::ShareableBitmap::Handle::encode const):
(WebKit::ShareableBitmap::Handle::decode):
Keep m_size here because it is an IntSize object and is useful to
check width/height parameters in the SharableBitmap object.
* Shared/ShareableResource.cpp:
(WebKit::ShareableResource::Handle::encode const):
(WebKit::ShareableResource::Handle::decode):
* Shared/WebCompiledContentRuleListData.cpp:
(WebKit::WebCompiledContentRuleListData::encode const):
(WebKit::WebCompiledContentRuleListData::decode):
* Shared/WebCoreArgumentCoders.cpp:
(IPC::encodeSharedBuffer):
(IPC::decodeSharedBuffer):
* Shared/WebHitTestResultData.cpp:
(WebKit::WebHitTestResultData::encode const):
(WebKit::WebHitTestResultData::decode):
More cases of converting between SharedMemory::Handle and
SharedMemory::IPCHandle when encoding/decoding.
* UIProcess/Cocoa/UserMediaCaptureManagerProxy.cpp:
* UIProcess/VisitedLinkStore.cpp:
(WebKit::VisitedLinkStore::sendStoreHandleToProcess):
* WebProcess/GPU/media/RemoteAudioBusData.h:
(WebKit::RemoteAudioBusData::encode const):
(WebKit::RemoteAudioBusData::decode):
Since no additional size parameter is being sent in this function, we
can use handle.size() as the IPCHandle::dataSize parameter.
* WebProcess/GPU/webrtc/AudioMediaStreamTrackRenderer.cpp:
(WebKit::AudioMediaStreamTrackRenderer::storageChanged):
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
(WebKit::MediaRecorderPrivate::storageChanged):
* WebProcess/Storage/WebSWClientConnection.cpp:
(WebKit::WebSWClientConnection::setSWOriginTableSharedMemory):
* WebProcess/Storage/WebSWClientConnection.h:
* WebProcess/Storage/WebSWClientConnection.messages.in:
* WebProcess/WebPage/VisitedLinkTableController.cpp:
(WebKit::VisitedLinkTableController::setVisitedLinkTable):
* WebProcess/WebPage/VisitedLinkTableController.h:
* WebProcess/WebPage/VisitedLinkTableController.messages.in:
* WebProcess/cocoa/RemoteCaptureSampleManager.cpp:
(WebKit::RemoteCaptureSampleManager::audioStorageChanged):
* WebProcess/cocoa/RemoteCaptureSampleManager.h:
* WebProcess/cocoa/RemoteCaptureSampleManager.messages.in:
More cases of converting between SharedMemory::Handle and
SharedMemory::IPCHandle when encoding/decoding.
2020-08-14 Per Arne Vollan <pvollan@apple.com>
[Cocoa] Avoid changing XPC target queue inside XPC event handler
https://bugs.webkit.org/show_bug.cgi?id=215460
Reviewed by Darin Adler.
In WebProcess::handleXPCEndpointMessages we currently change the XPC target queue for the XPC bootstrap connection while
under the XPC event handler. This sometimes causes simulated crashes on iOS and should be avoided. According to the
documentation in https://developer.apple.com/documentation/xpc/1448786-xpc_connection_set_target_queue?language=objc,
there does not seem to be anything saying this is a programming error, but the simulated crash claims otherwise. This
patch addresses this issue by changing the initial target queue for the XPC bootstrap connection from the main thread
queue to a queue on a secondary thread. The WebKit initializer function needs to be called on the main thread, which
is done by synchronously dispatching the call on the main thread. This patch also stops changing the event handler for
the bootstrap connection, but instead adds the additional XPC handling to the current event handler.
No new tests, since this is covered by existing tests. The handling of the Launch Services XPC endpoint message that
this patch changes, is used to receive the Launch Services database from the Networking process. The Launch Services
database is used in MIME type mapping APIs, and many tests would fail if this database was not received in the
WebContent process.
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceMain.mm:
(WebKit::XPCServiceEventHandler):
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebPage/Cocoa/WebPageCocoa.mm:
(WebKit::WebPage::platformDidReceiveLoadParameters):
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::initializeConnection):
* WebProcess/WebProcess.h:
* WebProcess/cocoa/HandleXPCEndpointMessages.h: Added.
* WebProcess/cocoa/HandleXPCEndpointMessages.mm: Added.
(WebKit::handleXPCEndpointMessages):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::handleXPCEndpointMessages const): Deleted.
2020-08-14 Kate Cheney <katherine_cheney@apple.com>
Convert SharedMemory::Handle to SharedMemory::IPCHandle WebPage image/pasteboard functions
https://bugs.webkit.org/show_bug.cgi?id=215478
<rdar://problem/67024582>
Reviewed by Chris Dumez.
Convert SharedMemory::Handle objects to SharedMemory::IPCHandle
objects which automatically verify the size of data being sent.
We can remove the additional size parameter being sent in many IPC
messages along with the handle, now that both are stored in the
IPCHandle struct.
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::setPromisedDataForImage):
* UIProcess/Cocoa/WebPasteboardProxyCocoa.mm:
(WebKit::WebPasteboardProxy::getPasteboardBufferForType):
No longer need size check, this is done in the IPCHandle::decode()
function.
(WebKit::WebPasteboardProxy::setPasteboardBufferForType):
(WebKit::WebPasteboardProxy::readBufferFromPasteboard):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/WebPasteboardProxy.cpp:
(WebKit::WebPasteboardProxy::readBufferFromPasteboard):
* UIProcess/WebPasteboardProxy.h:
* UIProcess/WebPasteboardProxy.messages.in:
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::saveImageToLibrary):
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::dataSelectionForPasteboard):
No longer need size check, this is done in the IPCHandle::decode()
function.
(WebKit::WebPageProxy::setPromisedDataForImage):
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::writeItemsToPasteboard):
* WebProcess/WebCoreSupport/WebPlatformStrategies.cpp:
(WebKit::WebPlatformStrategies::bufferForType):
(WebKit::WebPlatformStrategies::setBufferForType):
(WebKit::WebPlatformStrategies::readBufferFromPasteboard):
No longer need size check, this is done in the IPCHandle::decode()
function.
* WebProcess/WebCoreSupport/mac/WebDragClientMac.mm:
(WebKit::WebDragClient::declareAndWriteDragImage):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::getDataSelectionForPasteboard):
(WebKit::WebPage::performActionOnElement):
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::getDataSelectionForPasteboard):
2020-08-14 Wenson Hsieh <wenson_hsieh@apple.com>
Unreviewed, revert r265213 and r265441
https://bugs.webkit.org/show_bug.cgi?id=215089
This logging is no longer needed, since the test seems to be passing reliably in internal automation now.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView selectAllForWebView:]):
2020-08-13 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Notify the web process on drag leave
https://bugs.webkit.org/show_bug.cgi?id=215373
Reviewed by Adrian Perez de Castro.
We missed this when implemented drag and drop support for GTK4.
* UIProcess/API/gtk/DropTargetGtk4.cpp:
(WebKit::DropTarget::leave): Call dragExited() and resetCurrentDragInformation() on WebPageProxy.
2020-08-13 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] Do not accept drag operations when the matched target list is empty
https://bugs.webkit.org/show_bug.cgi?id=215372
Reviewed by Adrian Perez de Castro.
Do not initialize m_selectionData if we don't have valid targets.
* UIProcess/API/gtk/DropTargetGtk3.cpp:
(WebKit::DropTarget::accept):
(WebKit::DropTarget::leaveTimerFired):
2020-08-13 Brady Eidson <beidson@apple.com>
(WK2 only) gamepad mapping stops being reported as “standard” in new tabs.
rdar://problem/66946505 and https://bugs.webkit.org/show_bug.cgi?id=215486
Reviewed by Tim Horton.
Covered by new API test.
Get rid of the premature optimization that was "full gamepaddata" vs "condensed gamepaddata"
In practice, the actual button and axis values are larger than the mapping and id, so
not much IPC was being saved.
* Shared/Gamepad/GamepadData.cpp:
* Shared/Gamepad/GamepadData.h:
* UIProcess/Gamepad/UIGamepad.cpp:
(WebKit::UIGamepad::gamepadData const):
(WebKit::UIGamepad::condensedGamepadData const): Deleted.
(WebKit::UIGamepad::fullGamepadData const): Deleted.
* UIProcess/Gamepad/UIGamepad.h:
* UIProcess/Gamepad/UIGamepadProvider.cpp:
(WebKit::UIGamepadProvider::snapshotGamepads):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::gamepadConnected):
2020-08-13 Kate Cheney <katherine_cheney@apple.com>
Create SharedMemory::IPCHandle object to validate the size of SharedMemory::Handle objects sent over IPC
https://bugs.webkit.org/show_bug.cgi?id=215288
<rdar://problem/60870795>
Reviewed by David Kilzer.
Part 1 of a multi-patch plan to convert all SharedMemory::Handle
objects being sent over IPC to use SharedMemory::IPCHandle objects instead.
* Platform/SharedMemory.h:
(WebKit::SharedMemory::IPCHandle::IPCHandle):
* Platform/cocoa/SharedMemoryCocoa.cpp:
(WebKit::SharedMemory::IPCHandle::encode const):
(WebKit::SharedMemory::IPCHandle::decode):
Validate the size of the data sent in the IPCHandle::decode()
function.
* Platform/unix/SharedMemoryUnix.cpp:
(WebKit::SharedMemory::IPCHandle::encode const):
(WebKit::SharedMemory::IPCHandle::decode):
* Platform/win/SharedMemoryWin.cpp:
Implement IPCHandle encode and decode for all platforms for when we remove
SharedMemory::Handle encode/decode functions.
2020-08-13 Adrian Perez de Castro <aperez@igalia.com>
Unreviewed non-unified build fix.
No new tests needed.
* WebProcess/FullScreen/WebFullScreenManager.cpp:
(WebKit::WebFullScreenManager::requestEnterFullScreen): Add missing WebCore:: namespace
prefix to usage of WebCore::FullscreenManager.
2020-08-13 Chris Dumez <cdumez@apple.com>
REGRESSION (r260684): Messages YouTube inline video: after Multitasking away and Back, Audio is heard but icon indicates "muted"
https://bugs.webkit.org/show_bug.cgi?id=215453
<rdar://problem/66136673>
Reviewed by Tim Horton.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::setIsDoingSnapshotSequence):
(WebKit::WebPageProxy::setShouldFireEvents): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/ios/WKApplicationStateTrackingView.mm:
(-[WKApplicationStateTrackingView _willBeginSnapshotSequence]):
(-[WKApplicationStateTrackingView _didCompleteSnapshotSequence]):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::setIsDoingSnapshotSequence):
(WebKit::WebPage::setShouldFireEvents): Deleted.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-08-13 Per Arne Vollan <pvollan@apple.com>
Unreviewed, reverting r265520.
Invalid sandbox change
Reverted changeset:
"[macOS] Deny access to directory for compiled WebKit
sandboxes"
https://bugs.webkit.org/show_bug.cgi?id=215384
https://trac.webkit.org/changeset/265520
2020-08-12 Keith Rollin <krollin@apple.com>
Remove the need for defining USE_NEW_BUILD_SYSTEM
https://bugs.webkit.org/show_bug.cgi?id=215439
Reviewed by Darin Adler.
When building WebKit for XCBuild, we currently require that the
external build system (such as the Makefile, build-webkit, etc.)
defines the USE_NEW_BUILD_SYSTEM=YES build setting. This build setting
controls parts of our build instructions that are sensitive to when
XCBuild or the Legacy build system are being used. Notably, we need to
know when to use our custom “copy and modify” scripts with copying
certain header files (used with the Legacy build system) vs. using the
enhanced Copy Headers build phase that’s enabled with
APPLY_RULES_IN_COPY_HEADERS=YES (introduced with and used by XCBuild).
The choice of which method to copy headers is used is controlled by
USE_NEW_BUILD_SYSTEM.
There is no built-in build setting that we can probe to help us
determine which approach to take when copying and modifying headers,
which is why we need to define USE_NEW_BUILD_SYSTEM ourselves. But it
turns out that we can *detect* which build system is being used by
taking advantage of a subtle difference between the two systems. As
noted in:
https://developer.apple.com/documentation/xcode-release-notes/build-system-release-notes-for-xcode-10
“When an .xcconfig file contains multiple assignments of the same
build setting, later assignments using $(inherited) or
$(<setting_name>) will inherit from earlier assignments in the
.xcconfig. The legacy build system caused every use of
$(inherited) or $(<setting_name>) skip any other values defined
within the .xcconfig.”
This difference can be exploited as follows:
WK_WHICH_BUILD_SYSTEM = not_
WK_WHICH_BUILD_SYSTEM = $(inherited)legacy
WK_USE_NEW_BUILD_SYSTEM = $(WK_USE_NEW_BUILD_SYSTEM_$(WK_WHICH_BUILD_SYSTEM))
WK_USE_NEW_BUILD_SYSTEM_legacy = NO
WK_USE_NEW_BUILD_SYSTEM_not_legacy = YES
We can then use WK_USE_NEW_BUILD_SYSTEM where we used to use the
externally-defined USE_NEW_BUILD_SYSTEM.
No new tests -- no new or changed functionality.
* Configurations/Base.xcconfig:
* Configurations/WebKit.xcconfig:
* WebKit.xcodeproj/project.pbxproj:
2020-08-12 Alex Christensen <achristensen@webkit.org>
Fail preconnect requests to deprecated TLS instead of allowing application to show warning
https://bugs.webkit.org/show_bug.cgi?id=215424
<rdar://problem/66784116>
Reviewed by Geoffrey Garen.
Preconnecting is just a suggestion, so if the content really wants to load something from the deprecated TLS server,
we will see the warning when the load actually happens.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(-[WKNetworkSessionDelegate URLSession:task:didReceiveChallenge:completionHandler:]):
2020-08-12 Brent Fulgham <bfulgham@apple.com>
[iOS] Allow additional iokit-get-property to support HEIF images
https://bugs.webkit.org/show_bug.cgi?id=215431
<rdar://problem/66760775>
Reviewed by Per Arne Vollan.
Allow the WebContent process to read the HEVCCanDecodeTileToCanvas property.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-08-12 Per Arne Vollan <pvollan@apple.com>
[macOS] Fix iokit get property sandbox violation
https://bugs.webkit.org/show_bug.cgi?id=215430
<rdar://problem/65290967>
Reviewed by Brent Fulgham.
The WebContent process' sandbox on macOS needs to allow querying the iokit property "acoustic-id".
* WebProcess/com.apple.WebProcess.sb.in:
2020-08-12 Peng Liu <peng.liu6@apple.com>
Add the support to return to element fullscreen from picture-in-picture
https://bugs.webkit.org/show_bug.cgi?id=215305
Reviewed by Jer Noble.
This patch replaces WKFullScreenViewControllerVideoFullscreenModelClient
with WKFullScreenWindowControllerVideoFullscreenModelClient. We need to do
that because the instance of WKFullScreenViewControllerVideoFullscreenModelClient
will be destroyed after the container element exits fullscreen while the video
element enters picture-in-picture.
The instance of WKFullScreenWindowControllerVideoFullscreenModelClient
will always exist when the WKFullScreenWindowController instance is alive,
so that it can receive callbacks from the VideoFullscreenInterfaceAVKit instance
to implement the "return to element fullscreen from picture-in-picture" feature.
This patch supports the following transitions:
element fullscreen -> picture-in-picture (through user gestures)
element fullscreen -> picture-in-picture (through the PiP button)
picture-in-picture -> element fullscreen (when the tab is visible)
picture-in-picture -> element fullscreen (when the tab is invisible)
picture-in-picture -> element fullscreen (when the browser is in background)
picture-in-picture -> inline (when the browser is in foreground)
exit picture-in-picture when the browser is in background
* UIProcess/Cocoa/VideoFullscreenManagerProxy.h:
* UIProcess/Cocoa/VideoFullscreenManagerProxy.mm:
(WebKit::VideoFullscreenModelContext::fullscreenWillReturnToInline):
(WebKit::VideoFullscreenModelContext::prepareToExitFullscreen):
Notify clients to prepare for the stop of fullscreen/picture-in-picture.
(WebKit::VideoFullscreenManagerProxy::setupFullscreenWithID):
Fix issues that increase the client count unnecessarily.
(WebKit::VideoFullscreenManagerProxy::preparedToReturnToInline):
Call m_page->fullscreenMayReturnToInline() earlier.
(WebKit::VideoFullscreenManagerProxy::fullscreenMayReturnToInline):
Ditto.
(WebKit::VideoFullscreenManagerProxy::fullscreenWillReturnToInline):
Notify VideoFullscreenManager to prepare for exiting picture-in-picture or video fullscreen
and report the destination rectangle of the exit picture-in-picture or video fullscreen animations.
* UIProcess/WebFullScreenManagerProxy.cpp:
(WebKit::WebFullScreenManagerProxy::requestEnterFullScreen):
* UIProcess/WebFullScreenManagerProxy.h:
Add a function requestEnterFullScreen() so that we can request an element to enter fullscreen
from the UI process side.
* UIProcess/ios/fullscreen/WKFullScreenViewController.h:
* UIProcess/ios/fullscreen/WKFullScreenViewController.mm:
(-[WKFullScreenViewController initWithWebView:]):
(-[WKFullScreenViewController dealloc]):
(-[WKFullScreenViewController videoControlsManagerDidChange]):
(-[WKFullScreenViewController setAnimatingViewAlpha:]):
(-[WKFullScreenViewController _cancelAction:]):
(WKFullScreenViewControllerVideoFullscreenModelClient::setParent): Deleted.
(WKFullScreenViewControllerVideoFullscreenModelClient::setInterface): Deleted.
(WKFullScreenViewControllerVideoFullscreenModelClient::interface const): Deleted.
(-[WKFullScreenViewController willEnterPictureInPicture]): Deleted.
(-[WKFullScreenViewController didEnterPictureInPicture]): Deleted.
(-[WKFullScreenViewController failedToEnterPictureInPicture]): Deleted.
Minor clean-ups and remove code related to WKFullScreenViewControllerVideoFullscreenModelClient.
* UIProcess/ios/fullscreen/WKFullScreenWindowControllerIOS.h:
* UIProcess/ios/fullscreen/WKFullScreenWindowControllerIOS.mm:
(WKFullScreenWindowControllerVideoFullscreenModelClient::setParent):
(WKFullScreenWindowControllerVideoFullscreenModelClient::setInterface):
(WKFullScreenWindowControllerVideoFullscreenModelClient::interface const):
(-[WKFullScreenWindowController initWithWebView:]):
(-[WKFullScreenWindowController dealloc]):
(-[WKFullScreenWindowController enterFullScreen]):
(-[WKFullScreenWindowController beganEnterFullScreenWithInitialFrame:finalFrame:]):
(-[WKFullScreenWindowController requestEnterFullScreen]):
(-[WKFullScreenWindowController requestExitFullScreen]):
(-[WKFullScreenWindowController _completedExitFullScreen]):
(-[WKFullScreenWindowController videoControlsManagerDidChange]):
(-[WKFullScreenWindowController willEnterPictureInPicture]):
(-[WKFullScreenWindowController didEnterPictureInPicture]):
(-[WKFullScreenWindowController failedToEnterPictureInPicture]):
(-[WKFullScreenWindowController prepareToExitPictureInPicture]):
(-[WKFullScreenWindowController didExitPictureInPicture]):
Add WKFullScreenWindowControllerVideoFullscreenModelClient and implement the support
to "restore fullscreen from picture-in-picture".
* WebProcess/FullScreen/WebFullScreenManager.cpp:
(WebKit::WebFullScreenManager::enterFullScreenForElement):
(WebKit::WebFullScreenManager::requestEnterFullScreen):
* WebProcess/FullScreen/WebFullScreenManager.h:
* WebProcess/FullScreen/WebFullScreenManager.messages.in:
Add the interface requestEnterFullScreen() and the corresponding IPC message.
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::prepareForVideoFullscreen):
* WebProcess/WebCoreSupport/WebChromeClient.h:
Add the interface to prepare for video fullscreen standby. The web process can
use this interface to create a VideoFullscreenManager instance to avoid the
scenario that an IPC message comes from the VideoFullscreenManagerProxy but
the VideoFullscreenManager instance is not constructed yet.
* WebProcess/cocoa/VideoFullscreenManager.h:
* WebProcess/cocoa/VideoFullscreenManager.messages.in:
* WebProcess/cocoa/VideoFullscreenManager.mm:
(WebKit::VideoFullscreenManager::fullscreenWillReturnToInline):
(WebKit::VideoFullscreenManager::fullscreenMayReturnToInline): Deleted.
Rename fullscreenMayReturnToInline() to fullscreenWillReturnToInline().
In the UI process side, fullscreenMayReturnToInline() is used by
VideoFullscreenManagerProxy to notify applications regarding UI
changes (e.g., switch browser tabs).
2020-08-12 Wenson Hsieh <wenson_hsieh@apple.com>
WebPageProxy::registerEditCommand should be robust against invalid undo step identifiers
https://bugs.webkit.org/show_bug.cgi?id=215412
<rdar://problem/66296820>
Reviewed by Tim Horton.
Add a `MESSAGE_CHECK` to avoid ever creating a `WebEditCommandProxy` with an invalid command identifier.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::registerEditCommand):
2020-08-12 Youenn Fablet <youenn@apple.com>
Refresh WritableStream up to spec
https://bugs.webkit.org/show_bug.cgi?id=215267
Reviewed by Geoff Garen.
* Shared/WebPreferences.yaml:
Add a dedicated runtime flag, off by default.
2020-08-12 Youenn Fablet <youenn@apple.com>
Enable H264 low latency code path by default for MacOS
https://bugs.webkit.org/show_bug.cgi?id=215371
Reviewed by Eric Carlson.
Covered by existing tests, in particular platform/mac/webrtc/captureCanvas-webrtc-software-encoder.html.
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultWebRTCH264LowLatencyEncoderEnabled):
2020-08-12 Carlos Garcia Campos <cgarcia@igalia.com>
REGRESSION(r261570): [GTK] Fails to send drop event to JavaScript
https://bugs.webkit.org/show_bug.cgi?id=215032
Reviewed by Darin Adler.
Handle custom data in drag and drop operations.
* UIProcess/API/gtk/DragSourceGtk3.cpp:
(WebKit::DragSource::DragSource):
(WebKit::DragSource::begin):
* UIProcess/API/gtk/DragSourceGtk4.cpp:
(WebKit::DragSource::begin):
* UIProcess/API/gtk/DropTargetGtk3.cpp:
(WebKit::DropTarget::DropTarget):
(WebKit::DropTarget::accept):
(WebKit::DropTarget::dataReceived):
* UIProcess/API/gtk/DropTargetGtk4.cpp:
(WebKit::DropTarget::DropTarget):
(WebKit::DropTarget::accept):
2020-08-11 Jer Noble <jer.noble@apple.com>
[Mac] Add Experimental Feature preference for SW VP9
https://bugs.webkit.org/show_bug.cgi?id=215043
<rdar://problem/66400034>
Reviewed by Beth Dakin.
* FeatureFlags/WebKit.plist:
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
* Shared/WebPageCreationParameters.h:
* Shared/WebPreferences.yaml:
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultVP9SWDecoderEnabledOnBattery):
* Shared/WebPreferencesDefaultValues.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::enableVP9Decoder):
(WebKit::WebProcess::enableVP9SWDecoder):
* WebProcess/WebProcess.h:
2020-08-11 Brady Eidson <beidson@apple.com>
Add a "use stored credentials" setting to WKWebView.
<rdar://problem/63308019> and https://bugs.webkit.org/show_bug.cgi?id=215388
Reviewed by Geoff Garen.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::preconnectTo):
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _canUseCredentialStorage]):
(-[WKWebView _setCanUseCredentialStorage:]):
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::preconnectTo):
(WebKit::WebPageProxy::setCanUseCredentialStorage):
(WebKit::WebPageProxy::creationParameters):
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::canUseCredentialStorage):
* WebProcess/Network/WebLoaderStrategy.cpp:
(WebKit::WebLoaderStrategy::scheduleLoadFromNetworkProcess):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
(WebKit::WebPage::setCanUseCredentialStorage):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-08-11 Per Arne Vollan <pvollan@apple.com>
[macOS] Deny access to directory for compiled WebKit sandboxes
https://bugs.webkit.org/show_bug.cgi?id=215384
Reviewed by Brent Fulgham.
A WebKit process on macOS should not be allowed access to the directory containing compiled sandboxes
after entering the sandbox itself.
* GPUProcess/mac/com.apple.WebKit.GPUProcess.sb.in:
* NetworkProcess/mac/com.apple.WebKit.NetworkProcess.sb.in:
* WebProcess/com.apple.WebProcess.sb.in:
2020-08-11 Tim Horton <timothy_horton@apple.com>
iOS: Scrolling and touch events sporadically stop working after navigating
https://bugs.webkit.org/show_bug.cgi?id=215368
<rdar://problem/65801531>
Reviewed by Wenson Hsieh.
Test: fast/events/touch/ios/touch-event-stall-after-navigating-with-pending-asynchronous-touch-start.html
WebPageProxy::handlePreventableTouchEvent keeps a counter of the number
of outstanding "preventable" touch events that it is waiting to hear
from the Web Content process about. This counter is incremented when the
event is dispatched to the Web Content process's EventHandler queue,
and decremented when the reply (whether it was handled or not) comes
back from the Web Content process. While the counter is non-zero, all
deferrable gestures remain deferred (and when it returns to zero, the
WKDeferringGestureRecognizer gate is lifted, and events are allowed to flow).
This means that it is very important that every event eventually reply
about its handling status.
Before this change, when the Web Content process is navigating, and reaches didCommitLoad,
it cleared all queued touch events, without replying to to the UI process.
Thus, there is a small window of time in which an incoming touch event will
end up in the queue, *not* get dispatched/replied, and then get dropped on
the floor in didCommitLoad. Most events do not meet this fate, because they
are handled promptly, and commitLoad/didCommitLoad tend to be very quick.
However, if the Web Content process ends up spending any significant amount
of time under commitLoad (say, in an unload handler, or in media frameworks),
any incoming touch events during that time will get lost, and the UI process'
count of outstanding events will end up getting stuck non-zero.
Fix this by always pretending that the page ate any events that were outstanding
when didCommitLoad occurs, instead of just dropping them on the floor.
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::doneWithTouchEvent):
* WebProcess/WebPage/EventDispatcher.cpp:
(WebKit::EventDispatcher::takeQueuedTouchEventsForPage):
(WebKit::EventDispatcher::clearQueuedTouchEventsForPage): Deleted.
(WebKit::EventDispatcher::getQueuedTouchEventsForPage): Deleted.
Also, rename getQueuedTouchEventsForPage to takeQueuedTouchEventsForPage,
since it removes the queue from EventDispatcher's set.
* WebProcess/WebPage/EventDispatcher.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::touchEventSync):
(WebKit::WebPage::didCommitLoad):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::cancelAsynchronousTouchEvents):
2020-08-11 Philippe Normand <pnormand@igalia.com>
[GStreamer] gst-full standalone library support
https://bugs.webkit.org/show_bug.cgi?id=215262
Reviewed by Xabier Rodriguez-Calvar.
Disable missing-plugins handling when gst-full is in use. The plugin installer doesn't make
much sense for this scenario.
* PlatformWPE.cmake:
* UIProcess/API/glib/WebKitInstallMissingMediaPluginsPermissionRequest.cpp:
(webkitInstallMissingMediaPluginsPermissionRequestAllow):
(webkitInstallMissingMediaPluginsPermissionRequestDeny):
(webkit_install_missing_media_plugins_permission_request_get_description):
* UIProcess/API/glib/WebKitInstallMissingMediaPluginsPermissionRequestPrivate.h:
* UIProcess/API/glib/WebKitWebView.cpp:
(webkitWebViewRequestInstallMissingMediaPlugins):
* UIProcess/gstreamer/InstallMissingMediaPluginsPermissionRequest.cpp:
* UIProcess/gstreamer/InstallMissingMediaPluginsPermissionRequest.h:
* UIProcess/gstreamer/WebPageProxyGStreamer.cpp:
(WebKit::WebPageProxy::requestInstallMissingMediaPlugins):
2020-08-10 Chris Dumez <cdumez@apple.com>
AuxiliaryProcess::didReceiveInvalidMessage() for WebPage::PerformDragControllerAction IPC
https://bugs.webkit.org/show_bug.cgi?id=215341
<rdar://problem/59344091>
Reviewed by Alex Christensen.
DragApplicationFlags is an enum containing flags. It was being sent over IPC as a DragApplicationFlags
instead of an OptionSet<DragApplicationFlags>, and thus would fail enum value validation when decoding
when more than one flag is set.
* Scripts/webkit/messages.py:
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<DragData>::decode):
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::applicationFlagsForDrag):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::performDragControllerAction):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-08-10 David Kilzer <ddkilzer@apple.com>
Use RefPtr for WebKit::WebOpenPanelResultListenerProxy
<https://webkit.org/b/215252>
<rdar://problem/65753821>
Reviewed by Geoffrey Garen.
Unable to create an API test for this.
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::UIClient::runOpenPanel):
- Use RefPtr<WebOpenPanelResultListenerProxy> when passing
`listener` into the block.
2020-08-10 Wenson Hsieh <wenson_hsieh@apple.com>
Add more logging to diagnose editing/selection/ios/select-all-non-editable-text-using-keyboard.html
https://bugs.webkit.org/show_bug.cgi?id=215089
<rdar://problem/66282806>
Reviewed by Tim Horton.
Remove some unnecessary declarations that were only used for debug logging in WebKitTestRunner, and add some
more logging to check whether the web view is even getting a call to select all from UIKit.
* Platform/spi/ios/UIKitSPI.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView selectAllForWebView:]):
2020-08-10 Sihui Liu <sihui_liu@apple.com>
Always suspend IDB work when network process is prepared to suspend
https://bugs.webkit.org/show_bug.cgi?id=215239
<rdar://problem/65690450>
Reviewed by Geoffrey Garen.
We do not suspend IDB work in the network process when there is an ongoing transaction because the network
process is going to ask the UI process to hold a background process assertion for it. However, it is possible
that the request from the network process does not reach the UI process in time: RunningBoard may already decide
to suspend the network process after app is backgrounded and UI process drops the foreground assertion for the
network process.
In this case, IDB in the network process would continue its transaction and the network process will be killed
when it becomes suspened for holding database file locks. A network process crash can lead to a worse result
than suspending IDB work, which aborts ongoing transactions, because it will destroy all database connections
and transaction. Therefore, let's just suspend IDB work when the network process receives prepareToSuspend
message.
* NetworkProcess/IndexedDB/WebIDBServer.cpp:
(WebKit::WebIDBServer::suspend):
* NetworkProcess/IndexedDB/WebIDBServer.h:
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::prepareToSuspend):
2020-08-10 Per Arne Vollan <pvollan@apple.com>
Add required entitlement for Catalyst
https://bugs.webkit.org/show_bug.cgi?id=215244
Reviewed by Darin Adler.
The entitlement 'com.apple.private.webkit.use-xpc-endpoint' should be added to the WebContent process on Catalyst as well,
since it is needed on all Apple platforms. This entitlement is needed to support direct XPC communication between WebKit
processes, which is used to send the Launch Services database to the WebContent process from the Networking process. Also,
add some more logging related to this, to detect if it takes a long time for the WebContent process to receive the database.
* Scripts/process-entitlements.sh:
* WebProcess/WebPage/Cocoa/WebPageCocoa.mm:
(WebKit::WebPage::platformDidReceiveLoadParameters):
2020-08-09 Said Abou-Hallawa <sabouhallawa@apple.com>
[macOS] Drag/drop an image of a unsupported format to an file input element should convert it to a supported format
https://bugs.webkit.org/show_bug.cgi?id=212482
<rdar://problem/63731672>
Reviewed by Darin Adler.
Move ImageUtilities.h and ImageUtilitiesCG.cpp from WebKit to WebCore.
Use the image transcoding functions and shared WorkQueue from WebCore.
* Platform/ImageUtilities.h: Removed.
* Platform/cg: Removed.
* SourcesCocoa.txt:
* UIProcess/WebPageProxy.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
(WebKit::WebPageProxy::didChooseFilesForOpenPanelWithImageTranscoding):
(WebKit::m_transcodingQueue): Deleted.
* UIProcess/WebPageProxy.h:
* WebKit.xcodeproj/project.pbxproj:
2020-08-08 Wenson Hsieh <wenson_hsieh@apple.com>
[ iOS wk2 ] editing/pasteboard/paste-without-nesting.html is a flaky failure
https://bugs.webkit.org/show_bug.cgi?id=215218
<rdar://problem/66628493>
Reviewed by Darin Adler.
This test writes to the system pasteboard using `document.execCommand("Copy")`, and then immediately reads from
the system pasteboard by triggering `document.execCommand("Paste")`. On rare occasions, this fails on iOS, where
IPC messages for writing content to the pasteboard (e.g. `WebPasteboardProxy::WriteWebContentToPasteboard`) are
asynchronous, but the IPC message to get the pasteboard change count before pasting (`GetPasteboardChangeCount`)
is synchronous. This means that `Connection` may end up dispatching the sync `GetPasteboardChangeCount` IPC
message before dispatching `WriteWebContentToPasteboard`, which causes the pasteboard read to fail, because the
contents on the pasteboard have changed after starting to read from the pasteboard.
Note that this is not a problem on macOS since all pasteboard writing IPC is synchronous. Instead of turning all
of the async iOS pasteboard writing messages synchronous as well, we can fix this by adding a mechanism in
`WebProcess` to keep track of outgoing async pasteboard write messages; then, before attempting to grab the
change count when we start reading, wait for any of these pending async writes to finish before we continue.
In a future patch, we could actually adopt this same mechanism in `SetPasteboardTypes` and neighboring IPC
messages to turn these all asynchronous.
* UIProcess/Cocoa/WebPasteboardProxyCocoa.mm:
(WebKit::WebPasteboardProxy::writeURLToPasteboard):
(WebKit::WebPasteboardProxy::writeWebContentToPasteboard):
(WebKit::WebPasteboardProxy::writeImageToPasteboard):
(WebKit::WebPasteboardProxy::writeStringToPasteboard):
Call `didWriteToPasteboardAsynchronously` after we finish writing to the pasteboard asynchronously.
* WebProcess/WebCoreSupport/WebPlatformStrategies.cpp:
(WebKit::WebPlatformStrategies::changeCount):
Before accessing changeCount, wait until any asynchronous calls to write to the system pasteboard have finished.
(WebKit::WebPlatformStrategies::writeToPasteboard):
Right before we send an async message to the UI process to write to the system pasteboard, notify `WebProcess`
so that it can keep track of pending clipboard writes.
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::willWriteToPasteboardAsynchronously):
(WebKit::WebProcess::didWriteToPasteboardAsynchronously):
(WebKit::WebProcess::waitForPendingPasteboardWritesToFinish):
Wait for a maximum of 1 second for any outgoing pasteboard writing messages to return to the web process.
2020-08-07 Chris Dumez <cdumez@apple.com>
baseLatency attribute is missing on AudioContext interface
https://bugs.webkit.org/show_bug.cgi?id=215277
Reviewed by Eric Carlson.
* GPUProcess/media/RemoteAudioDestinationManager.cpp:
(WebKit::RemoteAudioDestination::framesPerBuffer const):
(WebKit::RemoteAudioDestinationManager::createAudioDestination):
* GPUProcess/media/RemoteAudioDestinationManager.h:
* GPUProcess/media/RemoteAudioDestinationManager.messages.in:
* WebProcess/GPU/media/RemoteAudioDestinationProxy.cpp:
(WebKit::RemoteAudioDestinationProxy::RemoteAudioDestinationProxy):
* WebProcess/GPU/media/RemoteAudioDestinationProxy.h:
2020-08-07 John Wilander <wilander@apple.com>
Experimental: Cap the expiry of persistent cookies set in 3rd-party CNAME cloaked HTTP responses
https://bugs.webkit.org/show_bug.cgi?id=215201
<rdar://problem/57454633>
Reviewed by Brent Fulgham. Also reviewed and commented on by Chris Dumez, Jiten Mehta, Sam Weinig, and Alex Christensen.
This experimental feature is off by default.
CNAME cloaking means a host resolves to a a different domain, potentially a
third-party domain, as part of DNS resolution.
This patch makes WebKit::NetworkDataTaskCocoa capture any CNAME cloaking for
the first party host and stores it in a table in the WebKit::NetworkSession.
It then checks first party subresource loads to see if they resolve to a
different domain and if so, compare that domain to both the first party
domain and its CNAME cloaking domain, if there is one. If there's a
mismatch, it's deemed a case of third-party CNAME cloaking and any cookies
set in the response of the cloaked subresource load will have their expiry
capped to 7 days.
The cases for capping expiry look like this (and are backed by test cases):
First-party host | First-party subdomain | Capped expiry
------------------|----------------------------|--------------
No CNAME cloaking | No CNAME cloaking | No
No CNAME cloaking | First-party CNAME cloaking | No
No CNAME cloaking | Third-party CNAME cloaking | Yes
CNAME cloaking | No CNAME cloaking | No
CNAME cloaking | Matching CNAME cloaking | No
CNAME cloaking | First-party CNAME cloaking | No
CNAME cloaking | Third-party CNAME cloaking | Yes
This patch makes use of two new CFNetwork SPIs on NSURLSessionTask:
- _cookieTransformCallback
- _resolvedCNAMEChain
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::resetParametersToDefaultValues):
This reset now covers the new CNAME cloaking member variables.
(WebKit::NetworkProcess::setIsRunningResourceLoadStatisticsTest):
This function now also enables the CNAME cloaking code.
(WebKit::NetworkProcess::setFirstPartyHostCNAMEDomainForTesting):
(WebKit::NetworkProcess::setThirdPartyCNAMEDomainForTesting):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
New IPC to forward test configuration.
* NetworkProcess/NetworkSession.cpp:
(WebKit::NetworkSession::NetworkSession):
Now picks up the flag for this feature.
(WebKit::NetworkSession::setFirstPartyHostCNAMEDomain):
This is called from NetworkDataTaskCocoa::updateFirstPartyInfoForTaskAndSession()
when there is CNAME cloaking for the first party host. This is done to make it
possible to not cap the expiry of cookies if subsequent subresource loads have
CNAME cloaking that matches the first-party host's CNAME cloaking. This happens
when whole websites are hosted on edge networks.
This function is also used by the test infrastructure to mock the DNS resolution
for a first-party host.
(WebKit::NetworkSession::firstPartyHostCNAMEDomain):
This returns any captured CNAME cloaking for a host, if there is one.
(WebKit::NetworkSession::resetCNAMEDomainData):
* NetworkProcess/NetworkSession.h:
(WebKit::NetworkSession::setCNAMECloakingMitigationEnabled):
(WebKit::NetworkSession::cnameCloakingMitigationEnabled const):
(WebKit::NetworkSession::setThirdPartyCNAMEDomainForTesting):
This test functions allows us to mock the DNS resolution for a subresource.
(WebKit::NetworkSession::thirdPartyCNAMEDomainForTesting const):
* NetworkProcess/cocoa/NetworkDataTaskCocoa.h:
* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::hasCNAMEAndCookieTransformSPI):
(WebKit::lastCNAMEDomain):
(WebKit::NetworkDataTaskCocoa::updateFirstPartyInfoForSession):
This is called by NetworkDataTaskCocoa::didReceiveResponse() to capture any
CNAME cloaking for the first-party host.
(WebKit::NetworkDataTaskCocoa::applyCookiePolicyForThirdPartyCNAMECloaking):
This is called in the NetworkDataTaskCocoa constructor and in
NetworkDataTaskCocoa::willPerformHTTPRedirection() and sets the new
_cookieTransformCallback SPI property on the task which will check the
response for any third-party CNAME cloaking and cap the expiry of incoming
cookies accordingly.
(WebKit::NetworkDataTaskCocoa::NetworkDataTaskCocoa):
This now calls
NetworkDataTaskCocoa::applyCookiePolicyForThirdPartyCNAMECloaking().
(WebKit::NetworkDataTaskCocoa::didReceiveResponse):
This now calls
NetworkDataTaskCocoa::updateFirstPartyInfoForTaskAndSession().
(WebKit::NetworkDataTaskCocoa::willPerformHTTPRedirection):
This now calls
NetworkDataTaskCocoa::applyCookiePolicyForThirdPartyCNAMECloaking().
* Shared/ResourceLoadStatisticsParameters.h:
Now holds a WebCore::CNAMECloakingMitigationEnabled flag.
(WebKit::ResourceLoadStatisticsParameters::encode const):
(WebKit::ResourceLoadStatisticsParameters::decode):
* Shared/WebPreferences.yaml:
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreSetResourceLoadStatisticsFirstPartyHostCNAMEDomainForTesting):
Used to mock CNAME resolution data.
(WKWebsiteDataStoreSetResourceLoadStatisticsThirdPartyCNAMEDomainForTesting):
Used to mock CNAME resolution data.
* UIProcess/API/C/WKWebsiteDataStoreRef.h:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::setFirstPartyHostCNAMEDomainForTesting):
(WebKit::NetworkProcessProxy::setThirdPartyCNAMEDomainForTesting):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::ensureNetworkProcess):
Now handles the WebCore::CNAMECloakingMitigationEnabled flag.
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::WebsiteDataStore::platformSetNetworkParameters):
Now handles the WebCore::CNAMECloakingMitigationEnabled flag.
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::setResourceLoadStatisticsFirstPartyHostCNAMEDomainForTesting):
(WebKit::WebsiteDataStore::setResourceLoadStatisticsThirdPartyCNAMEDomainForTesting):
(WebKit::WebsiteDataStore::parameters):
* UIProcess/WebsiteData/WebsiteDataStore.h:
2020-08-07 Wenson Hsieh <wenson_hsieh@apple.com>
Remove UIScriptController.removeAllDynamicDictionaries()
https://bugs.webkit.org/show_bug.cgi?id=215207
Reviewed by Sam Weinig.
See Tools/ChangeLog for more detail.
* Platform/spi/ios/UIKitSPI.h:
2020-08-06 Kenneth Russell <kbr@chromium.org>
Implement createImageBitmap(ImageData)
https://bugs.webkit.org/show_bug.cgi?id=183438
Reviewed by Dean Jackson.
Add "AlphaPremultiplication destFormat" argument to putImageData,
to handle creation of non-premultiplied ImageBitmaps from
ImageData.
* GPUProcess/graphics/RemoteImageBufferProxy.h:
* WebProcess/GPU/graphics/ImageBufferShareableBitmapBackend.cpp:
(WebKit::ImageBufferShareableBitmapBackend::putImageData):
* WebProcess/GPU/graphics/ImageBufferShareableBitmapBackend.h:
* WebProcess/GPU/graphics/RemoteImageBuffer.h:
(WebKit::RemoteImageBuffer::putImageData):
2020-08-06 David Kilzer <ddkilzer@apple.com>
WTF::makeString() should handle enum values
<https://webkit.org/b/214906>
Reviewed by Sam Weinig.
* Platform/IPC/cocoa/ConnectionCocoa.mm:
(IPC::Connection::sendMessage):
* Shared/Cocoa/AuxiliaryProcessCocoa.mm:
(WebKit::AuxiliaryProcess::didReceiveInvalidMessage):
- Update to take advantage of enum support in WTF::makeString().
2020-08-06 Jer Noble <jer.noble@apple.com>
[Mac,WK2] REGRESSION(r262322): ScreenTime overlay is hidden in fullscreen mode
https://bugs.webkit.org/show_bug.cgi?id=215222
<rdar://problem/65871844>
Reviewed by Eric Carlson.
During a refactor, a call to -[NSWindow setAutodisplay:YES] was dropped (in addition to a call to
NSEnableScreenUpdates(), but that has a 1s timeout so its effects aren't persistent). This meant
all NSViews added to that window need -display called on them explicitly in order to paint, and
so subviews like the ScreenTime overlay is never drawn.
* UIProcess/mac/WKFullScreenWindowController.mm:
(-[WKFullScreenWindowController beganEnterFullScreenWithInitialFrame:finalFrame:]):
2020-08-06 Truitt Savell <tsavell@apple.com>
Unreviewed, reverting r265328.
Broke 17 MediaRecorder tests.
Reverted changeset:
"Add support for MediaRecorder bitrate options"
https://bugs.webkit.org/show_bug.cgi?id=214973
https://trac.webkit.org/changeset/265328
2020-08-06 Youenn Fablet <youenn@apple.com>
Add support for MediaRecorder bitrate options
https://bugs.webkit.org/show_bug.cgi?id=214973
Reviewed by Eric Carlson.
Serialize options when creating remote media recorder.
* GPUProcess/webrtc/RemoteMediaRecorder.cpp:
(WebKit::RemoteMediaRecorder::create):
* GPUProcess/webrtc/RemoteMediaRecorder.h:
* GPUProcess/webrtc/RemoteMediaRecorderManager.cpp:
(WebKit::RemoteMediaRecorderManager::createRecorder):
* GPUProcess/webrtc/RemoteMediaRecorderManager.h:
* GPUProcess/webrtc/RemoteMediaRecorderManager.messages.in:
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
(WebKit::MediaRecorderPrivate::MediaRecorderPrivate):
(WebKit::MediaRecorderPrivate::startRecording):
* WebProcess/GPU/webrtc/MediaRecorderPrivate.h:
* WebProcess/GPU/webrtc/MediaRecorderProvider.cpp:
(WebKit::MediaRecorderProvider::createMediaRecorderPrivate):
* WebProcess/GPU/webrtc/MediaRecorderProvider.h:
2020-08-06 Adrian Perez de Castro <aperez@igalia.com>
[WPE][GTK] Wrong argument order for clone syscall seccomp filter on s390x
https://bugs.webkit.org/show_bug.cgi?id=215212
Reviewed by Michael Catanzaro.
Patch based on this Flatpak pull request:
https://github.com/flatpak/flatpak/pull/3777
No new tests needed.
* UIProcess/Launcher/glib/BubblewrapLauncher.cpp:
(WebKit::setupSeccomp): Add preprocessor guard to choose the correct
clone() system call argument on S390.
2020-08-06 James Savage <james.savage@apple.com>
Could not find module 'WebKit' for target 'armv7-apple-ios'
<https://bugs.webkit.org/show_bug.cgi?id=215190>
<rdar://problem/65642049>
Reviewed by Brady Eidson.
* SwiftOverlay/Configurations/WebKitSwiftOverlayTests.xcconfig: Define
SWIFT_MODULE_ONLY_ARCHS to emit other architectures.
2020-08-05 Tim Horton <timothy_horton@apple.com>
Remove all references to non-existent 10.16
https://bugs.webkit.org/show_bug.cgi?id=215202
Reviewed by Wenson Hsieh.
* Configurations/Base.xcconfig:
* Configurations/DebugRelease.xcconfig:
* Configurations/Version.xcconfig:
* Configurations/WebKitTargetConditionals.xcconfig:
* Scripts/process-entitlements.sh:
* WebProcess/com.apple.WebProcess.sb.in:
2020-08-05 Per Arne Vollan <pvollan@apple.com>
[Cocoa] Sandbox extension token not cleared from memory
https://bugs.webkit.org/show_bug.cgi?id=215136
Reviewed by Geoffrey Garen.
As a security mitigation, an invalidated sandbox extension should have its token cleared from memory.
No new tests, covered by existing tests.
* Shared/Cocoa/SandboxExtensionCocoa.mm:
(WebKit::SandboxExtensionImpl::m_length):
(WebKit::SandboxExtensionImpl::~SandboxExtensionImpl):
2020-08-05 Per Arne Vollan <pvollan@apple.com>
[Cocoa] Remove obsolete sandbox extension after r264178
https://bugs.webkit.org/show_bug.cgi?id=215154
Reviewed by Brent Fulgham.
After r264178, the code related to issuing an extension to com.apple.lsd.mapdb is obsolete, and should be removed.
No new tests, covered by existing tests.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
* WebProcess/com.apple.WebProcess.sb.in:
2020-08-05 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] Keyboard shortcuts and arrow key scrolling stop working after navigating via swipe gesture
https://bugs.webkit.org/show_bug.cgi?id=215137
<rdar://problem/65082979>
Reviewed by Tim Horton.
The process of starting a navigation swipe gesture causes the first responder (in this case, WKContentView) to
resign. Subsequently, nothing makes the content view first responder again once the navigation gesture ends,
even if the gesture is cancelled and we don't end up navigating. This results in several symptoms, two of which
are that keyboard shortcuts stop working, and pressing arrow keys does not scroll the web view after ending the
swipe gesture.
To mitigate this, add a mechanism to have the web view remember that our content view was first responder before
calling `-startInteractiveTransition:`; after ending the interactive transition (i.e. swipe gesture), we then
restore the web view's content view as first responder if it is not already first responder (and it is also
parented, which is not the case when swiping back to close a newly opened tab in Safari).
Test: NavigationSwipeTests.RestoreFirstResponderAfterNavigationSwipe
NavigationSwipeTests.DoNotBecomeFirstResponderAfterNavigationSwipeIfWebViewIsUnparented
* UIProcess/API/Cocoa/WKWebViewInternal.h:
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _navigationGestureDidBegin]):
(-[WKWebView _navigationGestureDidEnd]):
2020-08-04 Brent Fulgham <bfulgham@apple.com>
Unreviewed build fix after r265263
https://bugs.webkit.org/show_bug.cgi?id=215139
<rdar://problem/66501746>
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
2020-08-04 Devin Rousso <drousso@apple.com>
"DoubleDown Casino" respin button stops working with trackpad
https://bugs.webkit.org/show_bug.cgi?id=215146
<rdar://problem/64668138>
Reviewed by Tim Horton.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView shouldUseMouseGestureRecognizer]):
2020-08-04 Per Arne Vollan <pvollan@apple.com>
[iOS] Issue a temporary extension to the MobileGestalt daemon when the MobileGestalt cache is invalid
https://bugs.webkit.org/show_bug.cgi?id=214965
Unreviewed, address review comments.
Add comment explaining why a set of MobileGestalt queries are performed when starting up the WebContent process on iOS.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-08-04 Per Arne Vollan <pvollan@apple.com>
[iOS] Check that Accessibility is enabled when receiving the enable Accessibility notification
https://bugs.webkit.org/show_bug.cgi?id=215112
<rdar://problem/66498397>
Reviewed by Youenn Fablet.
As a security mitigation, return early when handling this notification if Accessibility is not enabled.
API test: WebKit.AccessibilityHasNoPreferencesServiceAccessWhenPostingNotification
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::registerNotificationObservers):
2020-08-04 Brent Fulgham <bfulgham@apple.com>
Limit 'com.apple.webkit.microphone' use to macOS and MacCatalyst
https://bugs.webkit.org/show_bug.cgi?id=215139
<rdar://problem/66501746>
Reviewed by Eric Carlson.
We only need the 'com.apple.webkit.microphone' Sandbox Extension at startup on
macOS and MacCatalyst (and even on those platforms, only until <rdar://problem/29448368>
is fixed).
We should not issue this extension on the iOS family of targets.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
2020-08-04 Per Arne Vollan <pvollan@apple.com>
[iOS Simulator] Assert under WebPage::platformDidReceiveLoadParameters
https://bugs.webkit.org/show_bug.cgi?id=215108
Reviewed by Youenn Fablet.
This asserts because it appears that the WebContent and Networking process do not have the required entitlement
com.apple.private.webkit.use-xpc-endpoint on the iOS Simulator. Add the entitlement for simulator.
No new tests, covered by existing tests.
* Resources/ios/XPCService-embedded-simulator.entitlements:
2020-08-03 Wenson Hsieh <wenson_hsieh@apple.com>
Allow -accessoryDone to blur the focused element on iPad when AutoFilling strong passwords
https://bugs.webkit.org/show_bug.cgi?id=215105
<rdar://problem/65143984>
Reviewed by Tim Horton.
`-accessoryDone` is now used to dismiss the strong password AutoFill keyboard after choosing a password on iOS,
due to how it hides the keyboard without causing the content view to resign first responder; being stuck in a
state where the content view is not first responder causes several issues when choosing strong passwords on iOS,
such as keyboard shortcuts no longer working.
On iPad, to ensure that `-accessoryDone` actually dismisses the keyboard, we need to additionally teach
`-endEditingAndUpdateFocusAppearanceWithReason:` (when given `EndEditingReasonAccessoryDone`) to allow the
focused element to blur when dismissing the strong password input view.
Test: KeyboardInputTests.TestWebViewAccessoryDoneDuringStrongPasswordAssistance
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView endEditingAndUpdateFocusAppearanceWithReason:]):
Refactor this logic into a local helper lambda with early returns, instead of using a single if statement.
2020-08-03 Youenn Fablet <youenn@apple.com>
Move user gesture propagation over promise behind a feature flag
https://bugs.webkit.org/show_bug.cgi?id=215014
Reviewed by Eric Carlson.
Add a preference, on by default.
* Shared/WebPreferences.yaml:
2020-08-03 Alex Christensen <achristensen@webkit.org>
Null check parentProcessConnection when creating a NetworkDataTaskCocoa
https://bugs.webkit.org/show_bug.cgi?id=215109
<rdar://problem/64853922>
Reviewed by Chris Dumez.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::sessionWrapperForTask):
Null check parentProcessConnection. Otherwise, we can dereference null and crash between disconnecting with the parent process and terminating,
which isn't the worst time to crash because we are trying to terminate anyways. But we may as well not crash.
2020-08-03 Wenson Hsieh <wenson_hsieh@apple.com>
Remove the ENABLE_DATA_INTERACTION feature flag
https://bugs.webkit.org/show_bug.cgi?id=215091
Reviewed by Megan Gardner.
* UIProcess/PageClient.h:
* UIProcess/WebPageProxy.cpp:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/ios/DragDropInteractionState.h:
* UIProcess/ios/DragDropInteractionState.mm:
* UIProcess/ios/PageClientImplIOS.h:
* UIProcess/ios/PageClientImplIOS.mm:
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView setUpInteraction]):
(-[WKContentView cleanUpInteraction]):
(-[WKContentView actionSheetAssistant:showCustomSheetForElement:]):
* UIProcess/ios/WebPageProxyIOS.mm:
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/ios/WebPageIOS.mm:
2020-08-03 Tim Horton <timothy_horton@apple.com>
Fix the macCatalyst build
* UIProcess/EndowmentStateTracker.mm:
(WebKit::EndowmentStateTracker::stateFromEndowments):
2020-08-03 Per Arne Vollan <pvollan@apple.com>
Finalize the list of MobileGestalt queries needed to populate the in-memory cache in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=215095
<rdar://problem/66034080>
Reviewed by Geoffrey Garen.
When the MobileGestalt cache is invalid, we issue a temporary extension to the MobileGestalt daemon for the WebContent process,
which will populate the in-memory cache by doing a set of MobileGestalt queries before revoking the extension. This patch
finalizes the list of MobileGestalt queries needed to populate the in-memory cache in the WebContent process on iOS before
revoking the temporary extension to com.apple.mobilegestalt.xpc. This list was compiled by querying every possible MobileGestalt
key after the temporary extension was revoked, and make a note of all queries that were contacting the MobileGestalt daemon to
get the answer. The added queries in this patch should make the list complete, and ensures that all cachable values are in the
in-memory cache before revoking the extension, since every possible MobileGestalt query was tested. This again means that users
that have an invalid MobileGestalt cache, will have access to the exact same MobileGestalt values, as those users who have a
valid MobileGestalt cache.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-08-03 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS 14] editing/selection/ios/select-all-non-editable-text-using-keyboard.html sometimes times out
https://bugs.webkit.org/show_bug.cgi?id=215089
<rdar://problem/66282806>
Reviewed by Tim Horton.
* Platform/spi/ios/UIKitSPI.h:
2020-08-03 Devin Rousso <drousso@apple.com>
should be able to request an editing context for a given region of a given input
https://bugs.webkit.org/show_bug.cgi?id=215042
<rdar://problem/65561491>
Reviewed by Wenson Hsieh.
Right now, it's only possible to request an editing context for either an entire input
element or a particular rect of the page. It should be possible to request an editing
context for a particular rect within a given input element as well (e.g. get one paragraph
or the currently selected text in a `<textarea>` instead of the entire thing).
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::requestDocumentEditingContext):
2020-08-03 Kate Cheney <katherine_cheney@apple.com>
doesParentProcessHaveITPEnabled() check should match behavior in determineITPStateInternal() check
https://bugs.webkit.org/show_bug.cgi?id=215031
<rdar://problem/66384618>
Reviewed by Brent Fulgham.
Checking ITP state from a WebKit child process should have the same
behavior as checking it from the UI process. In order to do this,
we must pass the value of hasRequestedCrossWebsiteTrackingPermission
to the network process to know if an application has included the
NSCrossWebsiteTrackingUsageDescription value in their Info.plist.
* NetworkProcess/NetworkSessionCreationParameters.cpp:
(WebKit::NetworkSessionCreationParameters::encode const):
(WebKit::NetworkSessionCreationParameters::decode):
* NetworkProcess/NetworkSessionCreationParameters.h:
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::activateSessionCleanup):
* Shared/Cocoa/DefaultWebBrowserChecks.h:
* Shared/Cocoa/DefaultWebBrowserChecks.mm:
(WebKit::hasRequestedCrossWebsiteTrackingPermission):
(WebKit::doesParentProcessHaveITPEnabled):
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeNetworkProcess):
2020-08-03 Adrian Perez de Castro <aperez@igalia.com>
Non-unified build fixes, early August 20202 edition
https://bugs.webkit.org/show_bug.cgi?id=215082
Unreviewed build fix.
* WebProcess/WebPage/FindController.h: Add missing inclusion of WebCore/SimpleRange.h.
2020-08-02 Darin Adler <darin@apple.com>
Remove some member functions of Range and many calls to createLiveRange
https://bugs.webkit.org/show_bug.cgi?id=215070
Reviewed by Sam Weinig.
* WebProcess/InjectedBundle/API/Cocoa/WKWebProcessPlugInRangeHandle.mm:
(-[WKWebProcessPlugInRangeHandle detectDataWithTypes:context:]): Use makeSimpleRange.
* WebProcess/InjectedBundle/API/c/WKBundlePageOverlay.cpp:
(API::PageOverlayClientImpl::actionContextForResultAtPoint): Ditto.
* WebProcess/InjectedBundle/API/mac/WKDOMRange.mm:
(-[WKDOMRange textRects]): Ditto.
* WebProcess/InjectedBundle/API/mac/WKDOMTextIterator.mm:
(-[WKDOMTextIterator initWithRange:]): Ditto.
* WebProcess/InjectedBundle/DOM/InjectedBundleRangeHandle.cpp:
(WebKit::InjectedBundleRangeHandle::renderedImage): Ditto.
* WebProcess/WebPage/Cocoa/WebPageCocoa.mm:
(WebKit::WebPage::performDictionaryLookupAtLocation): Remove createLiveRange.
(WebKit::WebPage::performDictionaryLookupForSelection): Ditto.
* WebProcess/WebPage/FindController.cpp:
(WebKit::FindController::updateFindUIAfterPageScroll): Ditto.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::getPlatformEditorState const): Ditto.
(WebKit::WebPage::computeAndSendEditDragSnapshot): Ditto.
(WebKit::WebPage::getRectsForGranularityWithSelectionOffset): Ditto.
(WebKit::WebPage::getRectsAtSelectionOffsetWithText): Ditto.
(WebKit::WebPage::requestAutocorrectionData): Ditto.
2020-08-01 Darin Adler <darin@apple.com>
Fix and further refine makefile rule for WebKit messages generated files
https://bugs.webkit.org/show_bug.cgi?id=215064
Reviewed by Sam Weinig.
* DerivedSources-output.xcfilelist: Updated for a file that was omitted
because of a typo in the makefile.
* DerivedSources.make: Fix "addsufix" typo and reorganized things a bit.
2020-08-01 Keith Rollin <krollin@apple.com>
Prevent multiple invocations of generate-message-receiver.py
https://bugs.webkit.org/show_bug.cgi?id=215054
Reviewed by Darin Adler.
WebKit/DerivedSources.make is written such that
generate-message-receiver.py can be invoked N times, where N is the
value passed to -j. This results in the N invocations of
generate-message-receiver.py trying trying to generate the same set of
files at the same time, possibly leading to their being corrupted. We
have reports of generated files starting with a 4K block of NULs, and
this contention may be causing it. Address this by using a "pattern
rule" in the makefile, which is documented in GNU make's documentation
as the approach to take when there is one build rule that creates
multiple outputs, as is the case here.
See also <https://bugs.webkit.org/show_bug.cgi?id=166814> and
<http://trac.webkit.org/changeset/210507>.
No new tests -- no new or changed functionality.
* DerivedSources.make:
2020-08-01 Darin Adler <darin@apple.com>
Remove Range::create and many more uses of live ranges
https://bugs.webkit.org/show_bug.cgi?id=215004
Reviewed by Sam Weinig.
* WebProcess/InjectedBundle/API/c/WKBundlePageOverlay.cpp:
(API::PageOverlayClientImpl::actionContextForResultAtPoint): Return a
structure that includes a SimpleRange.
* WebProcess/InjectedBundle/DOM/InjectedBundleNodeHandle.cpp:
(WebKit::InjectedBundleNodeHandle::visibleRange): Use makeSimpleRange.
* WebProcess/WebPage/Cocoa/TextCheckingControllerProxy.h: Use SimpleRange.
* WebProcess/WebPage/Cocoa/TextCheckingControllerProxy.mm:
(WebKit::TextCheckingControllerProxy::rangeAndOffsetRelativeToSelection): Use SimpleRange.
(WebKit::TextCheckingControllerProxy::replaceRelativeToSelection): Ditto.
(WebKit::TextCheckingControllerProxy::removeAnnotationRelativeToSelection): Ditto.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::characterIndexForPointAsync): Remove makeSimpleRange.
(WebKit::WebPage::deleteSurrounding): Use makeSimpleRange.
* WebProcess/WebPage/WebPageOverlay.cpp:
(WebKit::WebPageOverlay::actionContextForResultAtPoint): Use return value instead of out argument.
* WebProcess/WebPage/WebPageOverlay.h: Update for above.
* WebProcess/WebPage/glib/WebPageGLib.cpp:
(WebKit::WebPage::getPlatformEditorState const): Use makeSimpleRange.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::computeEditableRootHasContentAndPlainText): Use makeSimpleRange.
(WebKit::WebPage::getRectsForGranularityWithSelectionOffset): Removed unneeded null checks that
are already done by makeSimpleRange.
(WebKit::WebPage::getRectsAtSelectionOffsetWithText): Use makeSimpleRange.
(WebKit::WebPage::requestDictationContext): Ditto.
(WebKit::WebPage::autocorrectionContext): Ditto.
(WebKit::linkIndicatorPositionInformation): Use makeRangeSelectingNodeContents.
(WebKit::dataDetectorLinkPositionInformation): Ditto.
(WebKit::WebPage::requestDocumentEditingContext): Use makeSimpleRange.
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::performImmediateActionHitTestAtLocation): Update for change to the
actionContextForResultAtPoint return value.
2020-08-01 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r265171.
https://bugs.webkit.org/show_bug.cgi?id=215060
broke 2 mac api tests
Reverted changeset:
"[Mac] Add Experimental Feature preference for SW VP9"
https://bugs.webkit.org/show_bug.cgi?id=215043
https://trac.webkit.org/changeset/265171
2020-08-01 Megan Gardner <megan_gardner@apple.com>
Unable to select multiple lines of vertical text correctly
https://bugs.webkit.org/show_bug.cgi?id=213671
<rdar://problem/53753636>
Reviewed by Darin Adler.
In order to make for a better text selection experience, we pulled the selection position
down to be on the last line selectable, rather than snap the selection to a single position.
This made for a better selection experience on small text, but we failed to take
vertical text into account, and a user is locked into only selecting vertical text that ends below the
other anchor point of the selection. We should have the same behavior for vertical text, but correctly
calculated for X instead of Y.
This does not fix all cases, but it fixes the most glaring cases.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::rangeForPointInRootViewCoordinates):
2020-07-31 Jer Noble <jer.noble@apple.com>
[Mac] Add Experimental Feature preference for SW VP9
https://bugs.webkit.org/show_bug.cgi?id=215043
<rdar://problem/66400034>
Reviewed by Beth Dakin.
* FeatureFlags/WebKit.plist:
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
* Shared/WebPageCreationParameters.h:
* Shared/WebPreferences.yaml:
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultVP9SWDecoderEnabledOnBattery):
* Shared/WebPreferencesDefaultValues.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::enableVP9Decoder):
(WebKit::WebProcess::enableVP9SWDecoder):
* WebProcess/WebProcess.h:
2020-07-31 Chris Dumez <cdumez@apple.com>
Launch time regression due to EndowmentStateTracker initialization in WebPageProxy constructor
https://bugs.webkit.org/show_bug.cgi?id=215029
<rdar://problem/66362333>
Reviewed by Geoffrey Garen.
EndowmentStateTracker was expensive and in a critical code path during MobileSafari app launch. Since
this code is new, this was a net app launch time regression. To address the regression, the following
changes were made:
1. We only need the EndowmentStateTracker to monitor if the app is user-facing or not, so that we can
suspend / resume media playback as needed. As a result, it is unnecessary to start monitoring
user-facing state changes as soon as the WebPageProxy is constructed. Instead, we now wait until
media plays on the page for the first time.
2. It turns out that over 70% of CPU time in the EndowmentStateTracker constructor was spent getting
the current endownment for the process. As a result, I updated the code to lazy populate those
until they are requested for the first time.
3. The remaining 30% of CPU time in the EndowmentStateTracker constructor was spent constructing the
RBSProcessMonitor. I moved the initialization of the RBSProcessMonitor from the constructor to
EndowmentStateTracker::addClient() so that we only construct a monitor when we actually have our
first client.
4. All this code was actually only useful for Catalyst app and was therefore doing work on iOS for no
reason. I properly #ifdef'd out this code so that it now only applies to Catalyst apps.
* UIProcess/EndowmentStateTracker.h:
(WebKit::EndowmentStateTracker::isVisible const):
(WebKit::EndowmentStateTracker::isUserFacing const):
* UIProcess/EndowmentStateTracker.mm:
(WebKit::EndowmentStateTracker::stateFromEndowments):
(WebKit::EndowmentStateTracker::registerMonitorIfNecessary):
(WebKit::EndowmentStateTracker::addClient):
(WebKit::EndowmentStateTracker::ensureState const const):
(WebKit::EndowmentStateTracker::setState):
(WebKit::EndowmentStateTracker::EndowmentStateTracker): Deleted.
(WebKit::EndowmentStateTracker::setIsUserFacing): Deleted.
(WebKit::EndowmentStateTracker::setIsVisible): Deleted.
* UIProcess/WebPageProxy.cpp:
(WebKit::m_transcodingQueue):
(WebKit::WebPageProxy::updatePlayingMediaDidChange):
* UIProcess/WebPageProxy.h:
2020-07-31 Chris Dumez <cdumez@apple.com>
Drop ProcessAssertion::Client and replace with a simple invalidation handler
https://bugs.webkit.org/show_bug.cgi?id=214976
Reviewed by Brady Eidson.
ProcessAssertion had a Client class with 2 functions:
(1) one to indicate that the assertion was invalidated
(2) one to indicate that the UI assertion is about to expire
While ProcessAssertion is used in several places, only the ProcessThrottler
ever sets itself as a client. The reason is that other call sites use
assertion types that do not expire. Also, (2) only makes sense for
ProcessAndUIAssertion, not for ProcessAssertion. Only the ProcessThrottler
is using a ProcessAndUIAssertion.
I think a better design is to have a simple invalidation handler on
the ProcessAssertion, that the call site can set if they are interested
in invalidation.
Similarly, I added a UIAssertion expiration handler on ProcessAndUIAssertion
so that the ProcessThrottler can know if the UIAssertion is about to expire.
This new design also matches more closely the system process assertion API
that ProcessAssertion / ProcessAndUIAssertion are wrapping.
This patch also fixes a bug found by Youenn Fablet where the ProcessThrottler
would only set its invalidation handler in didConnectToProcess(), instead of
doing it every time a new ProcessAssertion is created in setAssertionType().
The assertion type can change (e.g. from foreground to background) after the
process has launched.
* UIProcess/ProcessAssertion.h:
(WebKit::ProcessAssertion::setInvalidationHandler):
(WebKit::ProcessAssertion::Client::~Client): Deleted.
(WebKit::ProcessAssertion::setClient): Deleted.
(WebKit::ProcessAssertion::client): Deleted.
* UIProcess/ProcessThrottler.cpp:
(WebKit::ProcessThrottler::setAssertionType):
(WebKit::ProcessThrottler::didConnectToProcess):
* UIProcess/ProcessThrottler.h:
* UIProcess/ios/ProcessAssertionIOS.mm:
(WebKit::ProcessAssertion::processAssertionWasInvalidated):
(WebKit::ProcessAndUIAssertion::uiAssertionWillExpireImminently):
2020-07-31 Jiewen Tan <jiewen_tan@apple.com>
SOAuthorizationSession::dismissViewController could crash on calling WebPageProxy::platformWindow
https://bugs.webkit.org/show_bug.cgi?id=214816
<rdar://problem/65681530>
Reviewed by Youenn Fablet.
pageClient() called in WebPageProxy::platformWindow() could be null. Therefore, platformWindow()
needs to check m_pageClient instead.
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::platformWindow):
2020-07-31 Youenn Fablet <youenn@apple.com>
Update service worker if certificate info does not match when soft updating
https://bugs.webkit.org/show_bug.cgi?id=214988
Reviewed by Alex Christensen.
Make sure service worker soft updater gathers certificate info.
* NetworkProcess/ServiceWorker/ServiceWorkerSoftUpdateLoader.cpp:
(WebKit::ServiceWorkerSoftUpdateLoader::loadFromNetwork):
(WebKit::ServiceWorkerSoftUpdateLoader::didReceiveResponse):
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::allowsSpecificHTTPSCertificateForHost):
2020-07-31 Sihui Liu <sihui_liu@appe.com>
-[WKWebsiteDataStore _renameOrigin:] needs to support IndexedDB renames
https://bugs.webkit.org/show_bug.cgi?id=214905
<rdar://problem/66247978>
Reviewed by Alex Christensen.
* NetworkProcess/IndexedDB/WebIDBServer.cpp:
(WebKit::WebIDBServer::renameOrigin):
* NetworkProcess/IndexedDB/WebIDBServer.h:
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::renameOriginInWebsiteData):
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
(-[WKWebsiteDataStore _renameOrigin:to:forDataOfTypes:completionHandler:]):
2020-07-31 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed Catalyst build fix.
* WebProcess/WebPage/Cocoa/TextCheckingControllerProxy.h: I should have added the include here.
* WebProcess/WebPage/Cocoa/TextCheckingControllerProxy.mm:
2020-07-31 Per Arne Vollan <pvollan@apple.com>
[iOS] Issue a temporary extension to the MobileGestalt daemon when the MobileGestalt cache is invalid
https://bugs.webkit.org/show_bug.cgi?id=214965
Reviewed by Geoffrey Garen.
When the MobileGestalt cache is invalid, some MG queries will fail in the WebContent process, since it has no access to the daemon.
This can be fixed by issuing a temporary extension to the daemon, and having the WebContent process query all relevant MG values
while holding the extension. This will bring the values into the in-memory cache, which will be valid after the extension to the
daemon has been revoked. We will still attempt to rebuild the MG cache if invalid, but since we now have a workaround for when the
cache is invalid, we can do this on a secondary thread to avoid blocking the main thread. We also only attempt to rebuild once now,
instead of every time a WebContent process is started. Additionally, add IOKit property read access to "platform-name", since that
is needed by a MG query.
No new tests, since testing this depends on having an invalid cache, which is rare, and non-trivial to enforce in the test
environment.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitialize):
(WebKit::WebProcessPool::platformInitializeWebProcess):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-07-31 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] Mouse wheel events on horizontal scrollbar not correctly handled when AC mode is forced
https://bugs.webkit.org/show_bug.cgi?id=207113
Reviewed by Adrian Perez de Castro.
Move the direction swap to the UI process before passing the event to the web
process. WebHitTestResultData::isScrollbar is now an enum with values No, Verttical, Horizontal. When a wheel
event is received in the widget we check whether we should invert the direction or not.
* Shared/API/glib/WebKitHitTestResult.cpp:
(webkitHitTestResultCreate):
(webkitHitTestResultCompare):
* Shared/WebHitTestResultData.cpp:
(WebKit::WebHitTestResultData::WebHitTestResultData):
* Shared/WebHitTestResultData.h:
* UIProcess/API/APIHitTestResult.h:
(API::HitTestResult::isScrollbar const):
* UIProcess/API/glib/WebKitWebView.cpp:
(webkitWebViewMouseTargetChanged):
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(shouldInvertDirectionForScrollEvent):
(webkitWebViewBaseScrollEvent):
(webkitWebViewBaseScroll):
(webkitWebViewBaseSetMouseIsOverScrollbar):
* UIProcess/API/gtk/WebKitWebViewBasePrivate.h:
2020-07-30 Alex Christensen <achristensen@webkit.org>
Optimize WebsiteDataStoreConfiguration::copy
https://bugs.webkit.org/show_bug.cgi?id=215012
<rdar://problem/64263406>
Reviewed by Brady Eidson.
The WebsiteDataStoreConfiguration constructor does some read/write heavy things.
We only want to do these things if we are not just going to overwrite the strings with strings from another WebsiteDataStoreConfiguration.
No change in behavior, just making things faster.
* UIProcess/WebsiteData/WebsiteDataStoreConfiguration.cpp:
(WebKit::WebsiteDataStoreConfiguration::WebsiteDataStoreConfiguration):
(WebKit::WebsiteDataStoreConfiguration::copy const):
* UIProcess/WebsiteData/WebsiteDataStoreConfiguration.h:
(WebKit::WebsiteDataStoreConfiguration::create):
2020-07-30 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed second attempt to fix Catalyst build after r265084.
* WebProcess/WebPage/Cocoa/TextCheckingControllerProxy.mm:
2020-07-30 Wenson Hsieh <wenson_hsieh@apple.com>
[iPadOS] Two taps are required to choose a country/region when creating an account on ea.com
https://bugs.webkit.org/show_bug.cgi?id=214984
<rdar://problem/66261490>
Reviewed by Devin Rousso.
On the desktop version of the account creation page on ea.com, tapping a custom selection menu element
programmatically focuses a real `select` element that is positioned far outside the viewport. As a result, on
iPad, we still attempt to show a popover for this focused select at the element location. This popover is not
visible to the user, but on iOS 13, it still steals modal focus from the web view, which means that an
additional tap anywhere in the window is required to dismiss the (invisible) popover before the user can choose
one of the options in the custom in-page menu that is shown.
Furthermore, upon choosing one of these options, the page programmatically focuses the offscreen select again.
This causes us to re-present the offscreen popover, which requires yet another tap to dismiss on iOS 13.
However, on iOS 14, this actually results in a crash when attempting to scroll the popover's inner UITableView
to the selected row.
To fix both the crash on iOS 14 and the multiple redundant taps required to dismiss the popover on iOS 13, avoid
showing popovers altogether if the popover's presentation rect is outside the bounds of the content view, since
we cannot reasonably expect a user to interact with such a popover anyways.
Test: fast/forms/ios/repeatedly-focus-offscreen-select.html
* UIProcess/ios/forms/WKFormPopover.mm:
(-[WKRotatingPopover presentPopoverAnimated:]):
2020-07-30 Jer Noble <jer.noble@apple.com>
[Cocoa] Adopt -[AVContentKeyRequest willOutputBeObscuredDueToInsufficientExternalProtectionForDisplays:]
https://bugs.webkit.org/show_bug.cgi?id=214659
<rdar://problem/63555006>
Reviewed by Darin Adler.
Support clients in the GPU process asking for (and receiving updates to) the current displayID.
* GPUProcess/media/RemoteCDMInstanceSessionProxy.cpp:
(WebKit::RemoteCDMInstanceSessionProxy::displayIDChanged):
* GPUProcess/media/RemoteCDMInstanceSessionProxy.h:
* GPUProcess/media/RemoteCDMInstanceSessionProxy.messages.in:
2020-07-30 Antoine Quint <graouts@webkit.org>
[ iOS ] webanimations/accelerated-animation-with-easing.html is still flaky
https://bugs.webkit.org/show_bug.cgi?id=214327
<rdar://problem/65569237>
Reviewed by Dean Jackson.
When we fixed bug 213495 in r263506, we started calling PlatformCAAnimation::setTimingFunction()
to set the animation-wide timing function of CSS Animations and JS-originated animations alike.
However, we neglected to update PlatformCAAnimationRemote, used on iOS, to be able to handle
both animation-wide timing functions as well as keyframe-specific timing functions. This patch
adds a new "timingFunction" member to PlatformCAAnimationRemote::Properties to allow for this
and this new member is used in setTimingFunction() and copyTimingFunctionFrom(). Finally, after
the animation is decoded in the UI process, we call -[CAKeyframeAnimation setTimingFunction:]
inside addAnimationToLayer() to use this animation-wide timing function.
This fixes the test added originally in r263506 which would only work flakily, a symptom of this
missing implementation on iOS.
* WebProcess/WebPage/RemoteLayerTree/PlatformCAAnimationRemote.h:
* WebProcess/WebPage/RemoteLayerTree/PlatformCAAnimationRemote.mm:
(WebKit::encodeTimingFunction):
(WebKit::decodeTimingFunction):
(WebKit::PlatformCAAnimationRemote::Properties::encode const):
(WebKit::PlatformCAAnimationRemote::Properties::decode):
(WebKit::PlatformCAAnimationRemote::setTimingFunction):
(WebKit::PlatformCAAnimationRemote::copyTimingFunctionFrom):
(WebKit::addAnimationToLayer):
(WebKit::operator<<):
2020-07-30 Jer Noble <jer.noble@apple.com>
[Cocoa] Make DecoderTest run against .mp4 files; add rate-limiting.
https://bugs.webkit.org/show_bug.cgi?id=214807
Reviewed by Darin Adler.
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::enableVP9Decoder):
2020-07-30 Tim Horton <timothy_horton@apple.com>
Web content gets stuck in an inactive state (no cursor updates or text insertion caret) when activating a tab with a thumbnail visible
https://bugs.webkit.org/show_bug.cgi?id=214962
<rdar://problem/65670984>
Reviewed by Wenson Hsieh.
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::setThumbnailView):
When WKThumbnailView is unparented, invalidate all activity state bits.
We do this because many of the functions that compute activity state bits
use the WKThumbnailView's window while it is parented. When it is unparented,
and we switch back to using the WKWebView's window, we mustn't get stuck
with activity state bits from WKThumbnailView's window.
This was particularly problematic in the case of the WKThumbnailViews
used for tab hover previews, because that window is not key, so we'd get
stuck with the inactive-window state, resulting in a lack of cursor updates,
a hidden text caret, the wrong selection color, etc.
* UIProcess/mac/PageClientImplMac.mm:
(WebKit::PageClientImpl::isViewWindowActive):
While writing the API test for this, I noticed that isViewWindowActive
returns YES if there is no key window AND WKWebView is unparented
([NSApp keyWindow] == activeWindow() == nil).
This seems completely insane (how can the window be active if it doesn't
exist?), and breaks the test, so ensure that we only say YES if we actually
have a window.
2020-07-30 Peng Liu <peng.liu6@apple.com>
ASSERTION FAILED: !isInRoutingArbitrationForArbitrator(proxy) in WebKit::SharedArbitrator::beginRoutingArbitrationForArbitrator
https://bugs.webkit.org/show_bug.cgi?id=214904
Reviewed by Jer Noble.
Only add an AudioSessionRoutingArbitratorProxy to the WeakHashSet (m_arbitrators)
if the routing arbitration is completed successfully.
* UIProcess/Media/cocoa/AudioSessionRoutingArbitratorProxyCocoa.mm:
(WebKit::SharedArbitrator::beginRoutingArbitrationForArbitrator):
2020-07-30 Per Arne Vollan <pvollan@apple.com>
Remember to check entitlement before communicating over XPC
https://bugs.webkit.org/show_bug.cgi?id=214825
Reviewed by Brent Fulgham.
Remember to check entitlement before communicating over XPC with another WebKit process. This needs to be done
to make sure that it really is a WebKit process on the other end.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::m_messagePortChannelRegistry):
* Shared/Cocoa/XPCEndpoint.mm:
(WebKit::XPCEndpoint::XPCEndpoint):
* Shared/Cocoa/XPCEndpointClient.mm:
(WebKit::XPCEndpointClient::setEndpoint):
* WebProcess/WebPage/Cocoa/WebPageCocoa.mm:
(WebKit::WebPage::platformDidReceiveLoadParameters):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::handleXPCEndpointMessages const):
2020-07-30 Kate Cheney <katherine_cheney@apple.com>
REGRESSION (r264925): run-safari --debug no longer works
https://bugs.webkit.org/show_bug.cgi?id=214947
<rdar://problem/66297850>
Reviewed by Darin Adler.
Removing these APIs in WebKit caused older versions of Safari to
break. We should add back in empty functions to make this work.
No new tests. Manually verified this fix.
* Shared/API/c/WKRenderLayer.cpp: Added.
(WKRenderLayerGetTypeID):
(WKRenderLayerGetRenderer):
(WKRenderLayerCopyRendererName):
(WKRenderLayerCopyElementTagName):
(WKRenderLayerCopyElementID):
(WKRenderLayerGetElementClassNames):
(WKRenderLayerGetAbsoluteBounds):
(WKRenderLayerIsClipping):
(WKRenderLayerIsClipped):
(WKRenderLayerIsReflection):
(WKRenderLayerGetCompositingLayerType):
(WKRenderLayerGetBackingStoreMemoryEstimate):
(WKRenderLayerGetNegativeZOrderList):
(WKRenderLayerGetNormalFlowList):
(WKRenderLayerGetPositiveZOrderList):
(WKRenderLayerGetFrameContentsLayer):
* Shared/API/c/WKRenderLayer.h: Added.
* Shared/API/c/WKRenderObject.cpp: Added.
(WKRenderObjectGetTypeID):
(WKRenderObjectCopyName):
(WKRenderObjectCopyTextSnippet):
(WKRenderObjectGetTextLength):
(WKRenderObjectCopyElementTagName):
(WKRenderObjectCopyElementID):
(WKRenderObjectGetElementClassNames):
(WKRenderObjectGetAbsolutePosition):
(WKRenderObjectGetFrameRect):
(WKRenderObjectGetChildren):
* Shared/API/c/WKRenderObject.h: Added.
* Sources.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/InjectedBundle/API/c/WKBundlePage.cpp:
(WKBundlePageCopyRenderTree):
(WKBundlePageCopyRenderLayerTree):
* WebProcess/InjectedBundle/API/c/WKBundlePagePrivate.h:
2020-07-30 Darin Adler <darin@apple.com>
Further reduction in the use of live ranges, particularly in headers
https://bugs.webkit.org/show_bug.cgi?id=214793
Reviewed by Sam Weinig.
* Shared/API/c/mac/WKWebArchiveRef.cpp:
(WKWebArchiveCreateFromRange): Use makeSimpleRange.
* Shared/APIWebArchive.h: Use SimpleRange.
* Shared/APIWebArchive.mm:
(API::WebArchive::create): Take SimpleRange.
* Shared/EditingRange.cpp:
(WebKit::EditingRange::toRange): Don't call createLiveRange.
(WebKit::EditingRange::fromRange): Take SimpleRange.
* Shared/EditingRange.h: Updated for above.
* WebProcess/InjectedBundle/API/APIInjectedBundleEditorClient.h:
Take SimpleRange, String, StyleProperties instead of
Range, StringImpl, CSSStyleDeclaration.
* WebProcess/InjectedBundle/API/glib/WebKitWebEditor.cpp:
String instead of StringImpl.
* WebProcess/InjectedBundle/API/mac/WKDOMDocument.mm:
(-[WKDOMDocument createDocumentFragmentWithText:]): Use SimpleRange.
* WebProcess/InjectedBundle/API/mac/WKDOMRange.mm:
(-[WKDOMRange rangeByExpandingToWordBoundaryByCharacters:inDirection:]):
Use SimpleRange and createLiveRange.
* WebProcess/InjectedBundle/API/mac/WKWebProcessPlugInBrowserContextController.mm:
(-[WKWebProcessPlugInBrowserContextController _setEditingDelegate:]):
Updated for changes to bundle editor client.
* WebProcess/InjectedBundle/DOM/InjectedBundleRangeHandle.cpp:
(WebKit::InjectedBundleRangeHandle::create): Deleted. Moved the
code into InjectedBundleRangeHandle::getOrCreate.
(WebKit::createHandle): Added. Makes a handle for a SimpleRange.
* WebProcess/InjectedBundle/DOM/InjectedBundleRangeHandle.h:
Updated for above.
* WebProcess/InjectedBundle/InjectedBundlePageEditorClient.cpp:
(WebKit::createHandle): Added. Makes a handle for a CSSStyleDeclaration.
(WebKit::InjectedBundlePageEditorClient::shouldBeginEditing):
Updated for changes to bundle editor client.
(WebKit::InjectedBundlePageEditorClient::shouldEndEditing): Ditto.
(WebKit::InjectedBundlePageEditorClient::shouldInsertNode): Ditto.
(WebKit::InjectedBundlePageEditorClient::shouldInsertText): Ditto.
(WebKit::InjectedBundlePageEditorClient::shouldDeleteRange): Ditto.
(WebKit::InjectedBundlePageEditorClient::shouldChangeSelectedRange): Ditto.
(WebKit::InjectedBundlePageEditorClient::shouldApplyStyle): Ditto.
(WebKit::InjectedBundlePageEditorClient::didBeginEditing): Ditto.
(WebKit::InjectedBundlePageEditorClient::didEndEditing): Ditto.
(WebKit::InjectedBundlePageEditorClient::didChange): Ditto.
(WebKit::InjectedBundlePageEditorClient::didChangeSelection): Ditto.
(WebKit::InjectedBundlePageEditorClient::willWriteToPasteboard): Ditto.
(WebKit::InjectedBundlePageEditorClient::getPasteboardDataForRange): Ditto.
(WebKit::InjectedBundlePageEditorClient::performTwoStepDrop): Ditto.
* WebProcess/InjectedBundle/InjectedBundlePageEditorClient.h:
Updated for above.
* WebProcess/WebCoreSupport/WebEditorClient.cpp:
(WebKit::WebEditorClient::shouldDeleteRange): Updated to call new
revised bundle editor client.
(WebKit::WebEditorClient::shouldBeginEditing): Ditto.
(WebKit::WebEditorClient::shouldEndEditing): Ditto.
(WebKit::WebEditorClient::shouldInsertNode): Ditto.
(WebKit::WebEditorClient::shouldInsertText): Ditto.
(WebKit::WebEditorClient::shouldChangeSelectedRange): Ditto.
(WebKit::WebEditorClient::shouldApplyStyle): Ditto.
(WebKit::WebEditorClient::willWriteSelectionToPasteboard): Ditto.
(WebKit::WebEditorClient::getClientPasteboardData): Ditto.
(WebKit::WebEditorClient::performTwoStepDrop): Ditto.
(WebKit::insertionPointFromCurrentSelection): Use SimpleRange.
* WebProcess/WebPage/FindController.cpp:
(WebKit::FindController::replaceMatches): Use SimpleRange.
(WebKit::FindController::findString): Updated for change to
Page::findTextMatches function.
(WebKit::FindController::findStringMatches): Ditto.
(WebKit::FindController::getImageForFindMatch): Use SimpleRange.
(WebKit::FindController::selectFindMatch): Ditto.
* WebProcess/WebPage/FindController.h: Updated for above.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::getMarkedRangeAsync): Do not call createLiveRange.
(WebKit::WebPage::getSelectedRangeAsync): Ditto.
(WebKit::WebPage::characterIndexForPointAsync): Call makeSimpleRange.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::plainTextForContext): Changed overloads a bit to work
better with Optional<SimpleRange> and removed the ones that explicitly
take a Range, updating callers to compensate.
(WebKit::plainTextForDisplay): Ditto.
(WebKit::WebPage::getPlatformEditorState const): Use SimpleRange.
(WebKit::WebPage::getSelectionContext): Ditto.
(WebKit::WebPage::didConcludeEditDrag): Ditto.
(WebKit::WebPage::handleStylusSingleTapAtPoint): Ditto.
(WebKit::WebPage::selectWithGesture): Ditto.
(WebKit::rangeForPointInRootViewCoordinates): Return SimpleRange.
(WebKit::rangeAtWordBoundaryForPosition): Ditto.
(WebKit::WebPage::updateSelectionWithTouches): Use SimpleRange.
(WebKit::WebPage::selectWithTwoTouches): Ditto.
(WebKit::WebPage::extendSelection): Ditto.
(WebKit::WebPage::selectWordBackward): Ditto.
(WebKit::WebPage::moveSelectionByOffset): Ditto.
(WebKit::WebPage::getRectsForGranularityWithSelectionOffset): Ditto.
(WebKit::WebPage::selectPositionAtPoint): Ditto.
(WebKit::WebPage::selectPositionAtBoundaryWithDirection): Ditto.
(WebKit::WebPage::moveSelectionAtBoundaryWithDirection): Ditto.
(WebKit::WebPage::rangeForGranularityAtPoint): Ditto.
(WebKit::WebPage::selectTextWithGranularityAtPoint): Ditto.
(WebKit::WebPage::updateSelectionWithExtentPointAndBoundary): Ditto.
(WebKit::WebPage::updateSelectionWithExtentPoint): Ditto.
(WebKit::WebPage::replaceSelectedText): Ditto.
(WebKit::WebPage::replaceDictatedText): Ditto.
(WebKit::WebPage::requestAutocorrectionData): Ditto.
(WebKit::WebPage::applyAutocorrectionInternal): Ditto.
(WebKit::WebPage::autocorrectionContext): Ditto.
(WebKit::dataDetectorLinkPositionInformation): Ditto.
(WebKit::WebPage::updateSelectionWithDelta): Ditto.
(WebKit::WebPage::focusTextInputContextAndPlaceCaret): Ditto.
2020-07-30 Alex Christensen <achristensen@webkit.org>
Remove non-inclusive terminology from WebKit variable names and test-only SPI
https://bugs.webkit.org/show_bug.cgi?id=214941
Reviewed by Michael Catanzaro.
* NetworkProcess/NetworkConnectionToWebProcess.cpp:
(WebKit::NetworkConnectionToWebProcess::addOriginAccessAllowListEntry):
(WebKit::NetworkConnectionToWebProcess::removeOriginAccessAllowListEntry):
(WebKit::NetworkConnectionToWebProcess::resetOriginAccessAllowLists):
(WebKit::NetworkConnectionToWebProcess::addOriginAccessWhitelistEntry): Deleted.
(WebKit::NetworkConnectionToWebProcess::removeOriginAccessWhitelistEntry): Deleted.
(WebKit::NetworkConnectionToWebProcess::resetOriginAccessWhitelists): Deleted.
* NetworkProcess/NetworkConnectionToWebProcess.h:
* NetworkProcess/NetworkConnectionToWebProcess.messages.in:
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/API/glib/WebKitUserContent.cpp:
(_WebKitUserStyleSheet::_WebKitUserStyleSheet):
(webkit_user_style_sheet_new):
(webkit_user_style_sheet_new_for_world):
(_WebKitUserScript::_WebKitUserScript):
(webkit_user_script_new):
(webkit_user_script_new_for_world):
* UIProcess/API/glib/WebKitWebContext.cpp:
(pathIsBlocked):
(webkit_web_context_add_path_to_sandbox):
(pathIsBlacklisted): Deleted.
* UIProcess/API/gtk/WebKitUserContent.h:
* UIProcess/API/wpe/WebKitUserContent.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
* UIProcess/Launcher/glib/BubblewrapLauncher.cpp:
(WebKit::setupSeccomp):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::setFontAllowList):
* UIProcess/WebProcessPool.h:
* WebProcess/InjectedBundle/API/c/WKBundle.cpp:
(WKBundleAddOriginAccessAllowListEntry):
(WKBundleRemoveOriginAccessAllowListEntry):
(WKBundleResetOriginAccessAllowLists):
(WKBundleAddOriginAccessWhitelistEntry): Deleted.
(WKBundleRemoveOriginAccessWhitelistEntry): Deleted.
(WKBundleResetOriginAccessWhitelists): Deleted.
* WebProcess/InjectedBundle/API/c/WKBundlePrivate.h:
* WebProcess/InjectedBundle/InjectedBundle.cpp:
(WebKit::InjectedBundle::addOriginAccessAllowListEntry):
(WebKit::InjectedBundle::removeOriginAccessAllowListEntry):
(WebKit::InjectedBundle::resetOriginAccessAllowLists):
(WebKit::InjectedBundle::addOriginAccessWhitelistEntry): Deleted.
(WebKit::InjectedBundle::removeOriginAccessWhitelistEntry): Deleted.
(WebKit::InjectedBundle::resetOriginAccessWhitelists): Deleted.
* WebProcess/InjectedBundle/InjectedBundle.h:
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-07-15 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Enable API tests
https://bugs.webkit.org/show_bug.cgi?id=214344
Reviewed by Adrian Perez de Castro.
Use a custom KeyEvent struct to pass to the IM filter as a fake GdkEvent to be able to test input methods with
GTK4.
* UIProcess/API/glib/InputMethodFilter.h:
* UIProcess/API/gtk/InputMethodFilterGtk.cpp:
(WebKit::InputMethodFilter::platformEventKeyIsKeyPress const):
(WebKit::InputMethodFilter::filterKeyEvent):
* UIProcess/API/gtk/WebKitWebViewBaseInternal.h:
* UIProcess/CoordinatedGraphics/DrawingAreaProxyCoordinatedGraphics.cpp:
(WebKit::DrawingAreaProxyCoordinatedGraphics::DrawingMonitor::start):
(WebKit::DrawingAreaProxyCoordinatedGraphics::DrawingMonitor::stop):
2020-07-30 Brady Eidson <beidson@apple.com>
Refactor HID gamepad code to be much less fragile and much easier to hack on.
https://bugs.webkit.org/show_bug.cgi?id=214910
Reviewed by Darin Adler.
* Shared/Gamepad/GamepadData.cpp:
(WebKit::GamepadData::GamepadData):
* Shared/Gamepad/GamepadData.h:
* UIProcess/Gamepad/UIGamepad.h:
* WebProcess/Gamepad/WebGamepad.cpp:
(WebKit::WebGamepad::axisValues const):
(WebKit::WebGamepad::buttonValues const):
(WebKit::WebGamepad::updateValues):
* WebProcess/Gamepad/WebGamepad.h:
2020-07-30 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r265043.
https://bugs.webkit.org/show_bug.cgi?id=214959
Will re-land after Safari 14 branching
Reverted changeset:
"Enable paint timing by default"
https://bugs.webkit.org/show_bug.cgi?id=211736
https://trac.webkit.org/changeset/265043
2020-07-29 Megan Gardner <megan_gardner@apple.com>
Highlight color does not switch after being set it system preferences.
https://bugs.webkit.org/show_bug.cgi?id=214938
<rdar://problem/65270586>
Reviewed by Tim Horton.
After the changes to the sandbox and how defaults are being transferred to the web process
(https://trac.webkit.org/changeset/258064/webkit), the highlight color was not being updated
immediately. We need to send the proper notification after setting the defaults so that the
color cache is invalidated after it is actually set and we do not end up with a stale value.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::userHighlightColorPreferenceKey):
(WebKit::dispatchSimulatedNotificationsForPreferenceChange):
2020-07-29 Kate Cheney <katherine_cheney@apple.com>
Migrate App-Bound Domains code from WebKit Additions
https://bugs.webkit.org/show_bug.cgi?id=214889
<rdar://problem/66075857>
Reviewed by Brent Fulgham.
No new tests, this is just code migration and will be tested
by existing App-Bound Domains tests.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::activateSessionCleanup):
(WebKit::NetworkSessionCocoa::sessionWrapperForTask):
* Shared/Cocoa/DefaultWebBrowserChecks.h: Added.
* Shared/Cocoa/DefaultWebBrowserChecks.mm: Added.
Holds all functions related to full browser checks and
App-Bound Domains.
(WebKit::shouldBeTreatedAsFullBrowser):
(WebKit::getAppBoundDomainsTesting):
These functions allow us to treat our test runners as full browsers
while triggering non-browser mode for testing other App-Bound Domains
functionality.
(WebKit::isInWebKitChildProcess):
(WebKit::hasRequestedCrossWebsiteTrackingPermission):
(WebKit::determineITPStateInternal):
(WebKit::determineITPState):
(WebKit::doesAppHaveITPEnabled):
(WebKit::doesParentProcessHaveITPEnabled):
(WebKit::hasProhibitedUsageStrings):
(WebKit::isParentProcessAFullWebBrowser):
(WebKit::isFullWebBrowser):
* UIProcess/API/APIHTTPCookieStore.cpp:
(API::HTTPCookieStore::filterAppBoundCookies):
* UIProcess/API/APIPageConfiguration.cpp:
(API::PageConfiguration::PageConfiguration):
(): Deleted.
* UIProcess/WebPageProxy.cpp:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::WebProcessPool):
(WebKit::WebProcessPool::ensureNetworkProcess):
(WebKit::WebProcessPool::createNewWebProcess):
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::WebsiteDataStore::addTestDomains const):
(WebKit::WebsiteDataStore::ensureAppBoundDomains const):
* UIProcess/WebsiteData/WebsiteDataStore.h:
* WebKit.xcodeproj/project.pbxproj:
2020-07-29 Jer Noble <jer.noble@apple.com>
REGRESSION(r264476): Calling systemHasAC() regresses launch time performance
https://bugs.webkit.org/show_bug.cgi?id=214907
<rdar://problem/66191430>
Reviewed by Eric Carlson.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
2020-07-29 Darin Adler <darin@apple.com>
Improve range idioms and other changes to prepare the way for more reduction in live range use
https://bugs.webkit.org/show_bug.cgi?id=214882
Reviewed by Sam Weinig.
* WebProcess/InjectedBundle/API/c/WKBundleNodeHandle.cpp:
(WKBundleNodeHandleGetRenderRect): Emptied out this unused function and added
ASSERT_NOT_REACHED. Later, we can delete this once we deal with any link-time
dependencies. I believe Safari may link with this but never calls it.
(WKBundleNodeHandleCopyVisibleRange): Ditto.
(WKBundleNodeHandleSetHTMLInputElementSpellcheckEnabled): Ditto.
(WKBundleNodeHandleGetHTMLInputElementAutoFilled): Ditto.
(WKBundleNodeHandleGetHTMLInputElementAutoFillButtonEnabled): Ditto.
(WKBundleNodeHandleGetHTMLInputElementAutoFillAvailable): Ditto.
(WKBundleNodeHandleGetHTMLInputElementAutoFillButtonBounds): Ditto.
(WKBundleNodeHandleCopyHTMLTableCellElementCellAbove): Ditto.
(WKBundleNodeHandleCopyHTMLFrameElementContentFrame): Ditto.
(WKBundleNodeHandleGetHTMLInputElementAutofilled): Ditto.
(WKBundleNodeHandleSetHTMLInputElementAutofilled): Ditto.
(WKBundleNodeHandleSetHTMLInputElementAutoFillButtonEnabled): Ditto.
* WebProcess/WebPage/Cocoa/TextCheckingControllerProxy.mm:
(WebKit::TextCheckingControllerProxy::annotatedSubstringBetweenPositions):
Simplify with makeSimpleRange.
* WebProcess/WebPage/glib/WebPageGLib.cpp: Added include of "Range.h",
may not be needed right now, but highly likely to be needed after
future live range work.
* WebProcess/WebPage/ios/WebPageIOS.mm: Ditto.
(WebKit::WebPage::selectWithGesture): Simplify with makeSimpleRange.
(WebKit::WebPage::requestDocumentEditingContext): Ditto.
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::attributedSubstringForCharacterRangeAsync):
Use auto for range to prepare for possible change of type from live
range to SimpleRange.
2020-07-29 Noam Rosenthal <noam@webkit.org>
Enable paint timing by default
https://bugs.webkit.org/show_bug.cgi?id=211736
Reviewed by Simon Fraser.
* Shared/WebPreferences.yaml:
PaintTimingEnabled default to true.
2020-07-29 Adrian Perez de Castro <aperez@igalia.com>
Unreviewed. Update OptionsWPE.cmake and NEWS for the 2.29.2 release
* wpe/NEWS: Add release notes for 2.29.2.
2020-07-29 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r265018.
https://bugs.webkit.org/show_bug.cgi?id=214917
Broke various ios-wk2 tests.
Reverted changeset:
"Migrate App-Bound Domains code from WebKit Additions"
https://bugs.webkit.org/show_bug.cgi?id=214889
https://trac.webkit.org/changeset/265018
2020-07-29 Youenn Fablet <youenn@apple.com>
TestWebKitAPI.WebKit.QuotaDelegate flaky times out
https://bugs.webkit.org/show_bug.cgi?id=214842
Reviewed by Geoffrey Garen.
Add some release logging.
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::requestStorageSpace):
2020-07-29 Carlos Garcia Campos <cgarcia@igalia.com>
Unreviewed. Update OptionsGTK.cmake and NEWS for 2.29.4 release
* gtk/NEWS: Add release notes for 2.29.4.
2020-07-28 Kate Cheney <katherine_cheney@apple.com>
Migrate App-Bound Domains code from WebKit Additions
https://bugs.webkit.org/show_bug.cgi?id=214889
<rdar://problem/66075857>
Reviewed by Brent Fulgham.
No new tests, this is just code migration and will be tested
by existing App-Bound Domains tests.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::activateSessionCleanup):
(WebKit::NetworkSessionCocoa::sessionWrapperForTask):
* Shared/Cocoa/DefaultWebBrowserChecks.h: Added.
* Shared/Cocoa/DefaultWebBrowserChecks.mm: Added.
Holds all functions related to full browser checks and
App-Bound Domains.
(WebKit::shouldBeTreatedAsFullBrowser):
(WebKit::getAppBoundDomainsTesting):
These functions allow us to treat our test runners as full browsers
while triggering non-browser mode for testing other App-Bound Domains
functionality.
(WebKit::isInWebKitChildProcess):
(WebKit::hasRequestedCrossWebsiteTrackingPermission):
(WebKit::determineITPStateInternal):
(WebKit::determineITPState):
(WebKit::doesAppHaveITPEnabled):
(WebKit::doesParentProcessHaveITPEnabled):
(WebKit::hasProhibitedUsageStrings):
(WebKit::isParentProcessAFullWebBrowser):
(WebKit::isFullWebBrowser):
* UIProcess/API/APIHTTPCookieStore.cpp:
(API::HTTPCookieStore::filterAppBoundCookies):
* UIProcess/API/APIPageConfiguration.cpp:
(API::PageConfiguration::PageConfiguration):
(): Deleted.
* UIProcess/WebPageProxy.cpp:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::WebProcessPool):
(WebKit::WebProcessPool::ensureNetworkProcess):
(WebKit::WebProcessPool::createNewWebProcess):
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::WebsiteDataStore::addTestDomains const):
(WebKit::WebsiteDataStore::ensureAppBoundDomains const):
* UIProcess/WebsiteData/WebsiteDataStore.h:
* WebKit.xcodeproj/project.pbxproj:
2020-07-28 Tim Horton <timothy_horton@apple.com>
Web Share API Level 2: Can only perform one share action per share sheet invocation
https://bugs.webkit.org/show_bug.cgi?id=214894
Reviewed by Wenson Hsieh.
No new tests; our WKShareSheet testing mechanism intentionally bypasses
actually presenting the share sheet, which is the only way to test this.
WKShareSheet assumes that UIActivityViewController's completionWithItemsHandler
is only called once, when the share sheet is dismissed, and uses it to
clean up the temporary file written to disk, and send the success state
to the API.
However, it is actually called per share action, of which an arbitrary number
can be performed per UIActivityViewController invocation.
Defer cleanup until the share sheet is actually dismissed (which we observe
using UIAdaptivePresentationControllerDelegate's -presentationControllerDidDismiss: callback),
and respond to the Web Share API with success if *any* attempted share action
succeeded.
* UIProcess/Cocoa/WKShareSheet.mm:
(-[WKShareSheet presentWithParameters:inRect:completionHandler:]):
(-[WKShareSheet presentWithShareDataArray:inRect:]):
(-[WKShareSheet sharingServicePicker:didChooseSharingService:]):
(-[WKShareSheet sharingService:didFailToShareItems:error:]):
(-[WKShareSheet sharingService:didShareItems:]):
(-[WKShareSheet presentationControllerDidDismiss:]):
(-[WKShareSheet dismiss]):
Fold dispatchDidDismiss and _didCompleteWithSuccess into -dismiss,
so we only have one "dismiss if needed and also do the requisite cleanup"
path. -dismiss should only be called when it is no longer possible to do
additional sharing actions for the current share sheet invocation.
(-[WKShareSheet _didCompleteWithSuccess:]): Deleted.
(-[WKShareSheet dispatchDidDismiss]): Deleted.
2020-07-28 Timothy Hatcher <timothy@apple.com>
Web Inspector: Don't allow docking to a hidden web view
https://bugs.webkit.org/show_bug.cgi?id=214886
Reviewed by Devin Rousso.
* UIProcess/Inspector/WebInspectorProxy.cpp:
(WebKit::WebInspectorProxy::attach): Call platformCanAttach() since it can do a live check.
(WebKit::WebInspectorProxy::detach): Return early if already attached, otherwise asserts later.
* UIProcess/Inspector/mac/WebInspectorProxyMac.mm:
(WebKit::WebInspectorProxy::platformCanAttach): Return flase when the web view is hidden.
2020-07-28 Alex Christensen <achristensen@webkit.org>
Add null checks in ResourceLoader
https://bugs.webkit.org/show_bug.cgi?id=214786
<rdar://problem/61106685>
Reviewed by Youenn Fablet.
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::getResourceBytesAtPosition):
2020-07-27 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK][WPE] Support for backdrop-filter
https://bugs.webkit.org/show_bug.cgi?id=169988
Reviewed by Adrian Perez de Castro.
Set the backdrop layer if changed when updating the scene state.
* Shared/CoordinatedGraphics/CoordinatedGraphicsScene.cpp:
(WebKit::CoordinatedGraphicsScene::updateSceneState):
2020-07-28 Youenn Fablet <youenn@apple.com>
Disable low latency code path for H264 constrained baseline
https://bugs.webkit.org/show_bug.cgi?id=214830
Reviewed by Eric Carlson.
Enable low latency code path on iOS.
Manually tested.
* Shared/WebPreferences.yaml:
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultWebRTCH264LowLatencyEncoderEnabled):
* Shared/WebPreferencesDefaultValues.h:
2020-07-27 Kate Cheney <katherine_cheney@apple.com>
Handle full browser apps that try to use prohibited Info.plist strings
https://bugs.webkit.org/show_bug.cgi?id=214852
<rdar://problem/66023860>
Reviewed by Brent Fulgham.
Full-web browsing applications which also use prohibited Info.plist
strings should be restricted from using certain powerful APIs like
JavaScript injection. These apps will have the equivalent privileges
of an app that opts into App-Bound Domains with an empty
WKAppBoundDomains key.
No new tests. Writing a test for this would require being able to
change TestWebKitAPI's Info.plist between tests, which we do not
support. I manually tested this by adding a prohibited
string to the TestWebKitAPI Info.plist then verified that restricted APIs,
like JS injection, were blocked.
* UIProcess/WebPageProxy.cpp:
(isFullWebBrowser):
(hasProhibitedUsageStrings):
(WebKit::WebPageProxy::loadData):
(WebKit::WebPageProxy::setIsNavigatingToAppBoundDomainAndCheckIfPermitted):
We do not need to set m_isNavigatingToAppBoundDomain in the case where
a full web browser does not use prohibited strings. In this case, we
want m_isNavigatingToAppBoundDomain to stay as WTF::nullopt so no API
restriction is enabled.
2020-07-27 David Kilzer <ddkilzer@apple.com>
WebKit::setCrashReportApplicationSpecificInformation() should also log to WTFLogAlways()
<https://webkit.org/b/214756>
<rdar://problem/66085705>
Reviewed by Darin Adler.
This also makes the following changes:
- Renames setCrashReportApplicationSpecificInformation() to
logAndSetCrashLogMessage().
- Replaces _WKSetCrashReportApplicationSpecificInformation()
with WebKit::logAndSetCrashLogMessage().
- Calls WTF::setCrashLogMessage() in DumpRenderTree and
WebKitTestRunner to avoid new WTFLogAlways() calls, which
broke internal tests and caused the first patch to be rolled
out.
- Changes WebKit::logAndSetCrashLogMessage() to take a
const char* argument instead of CFStringRef.
* Platform/IPC/cocoa/ConnectionCocoa.mm:
(IPC::Connection::sendMessage):
(IPC::readFromMachPort):
- Update call to logAndSetCrashLogMessage() to pass const char*
value.
* Platform/cocoa/WKCrashReporter.h:
(WebKit::logAndSetCrashLogMessage):
- Rename from setCrashReportApplicationSpecificInformation().
- Update function signature to use `const char*`.
* Platform/cocoa/WKCrashReporter.mm:
(WebKit::logAndSetCrashLogMessage):
- Rename from setCrashReportApplicationSpecificInformation().
- Update function signature to use `const char*`.
- Call WTFLogAlways() in addition to setting the Application
Specific Information.
* Shared/Cocoa/AuxiliaryProcessCocoa.mm:
(WebKit::AuxiliaryProcess::didReceiveInvalidMessage):
- Update call to logAndSetCrashLogMessage() to pass const char*
value.
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceMain.mm:
(WebKit::XPCServiceMain):
- Update call to logAndSetCrashLogMessage() instead of both
WTFLogAlways() and
_WKSetCrashReportApplicationSpecificInformation().
* WebKit.xcodeproj/project.pbxproj:
- Remove WebProcessCocoa.h from the project.
* WebProcess/cocoa/WebProcessCocoa.h: Remove.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::_WKSetCrashReportApplicationSpecificInformation): Delete.
2020-07-27 Tim Horton <timothy_horton@apple.com>
Occasional crashes under PDFPlugin::createScrollbar and PDFPlugin::updateScrollbars
https://bugs.webkit.org/show_bug.cgi?id=214845
<rdar://problem/59293598>
Reviewed by Brady Eidson.
No new tests; timing is such that I can't reproduce without inserting
intentional delays into the main thread hops, which is further than
I'm willing to go for a test.
This is a speculative fix due to the aforementioned reproducibility issue.
* WebProcess/Plugins/PDF/PDFPlugin.h:
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::destroy):
Keep track of when destroy() has been called.
(WebKit::PDFPlugin::receivedNonLinearizedPDFSentinel):
(WebKit::PDFPlugin::threadEntry):
If the PDFPlugin is destroyed (explicit destroy(), not the destructor, which won't run since we're protecting `this`)
during the window in which the PDF thread is dispatching to the main thread, just ignore
the callback. We don't want to run it, because many of the things we depend on (like PluginView)
are disconnected at this point.
(WebKit::PDFPlugin::scrollbarStyleChanged):
It appears that in some much rarer case, even before the PDF thread existed, WebScrollerImpPairDelegate could
call us back after being destroy()'d but before being deallocated. Use the new bit to guard scrollbarStyleChanged().
2020-07-27 Megan Gardner <megan_gardner@apple.com>
Trying to lookup when WebView is in a popover causes process to hang.
https://bugs.webkit.org/show_bug.cgi?id=214773
<rdar://problem/52317762>
Reviewed by Devin Rousso.
The Lookup framework does not populate the menus that it vends with the option to 'lookup'
words that are selected in popovers. WebKit should follow the pattern and not put the
item in the menu as Lookup is not able to handle this situation. Also if Lookup is disabled
via defaults, we should not show it in the menu.
* UIProcess/mac/WebContextMenuProxyMac.mm:
(WebKit::WebContextMenuProxyMac::getContextMenuFromItems):
2020-07-27 Kate Cheney <katherine_cheney@apple.com>
Remove API::Object::Type::RenderLayer and API::Object::Type::RenderObject
https://bugs.webkit.org/show_bug.cgi?id=214772
<rdar://problem/52734803>
Reviewed by Simon Fraser.
These APIs are no longer used, so we should remove them.
* Shared/API/APIObject.h:
* Shared/API/c/WKBase.h:
* Shared/API/c/WKRenderLayer.cpp: Removed.
* Shared/API/c/WKRenderLayer.h: Removed.
* Shared/API/c/WKRenderObject.cpp: Removed.
* Shared/API/c/WKRenderObject.h: Removed.
* Shared/UserData.cpp:
(WebKit::UserData::encode):
(WebKit::UserData::decode):
* Shared/WebRenderLayer.cpp: Removed.
* Shared/WebRenderLayer.h: Removed.
* Shared/WebRenderObject.cpp: Removed.
* Shared/WebRenderObject.h: Removed.
* Sources.txt:
* UIProcess/API/C/WKAPICast.h:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/InjectedBundle/API/c/WKBundlePage.cpp:
(WKBundlePageCopyRenderTree): Deleted.
(WKBundlePageCopyRenderLayerTree): Deleted.
* WebProcess/InjectedBundle/API/c/WKBundlePagePrivate.h:
2020-07-27 Youenn Fablet <youenn@apple.com>
Check m_connection in NetworkRTCProvider::createResolver/stopResolver
https://bugs.webkit.org/show_bug.cgi?id=214818
Reviewed by Eric Carlson.
We need to check for m_connection when hopping to main thread as the NetworkRTCProvider might already be closed.
* NetworkProcess/webrtc/NetworkRTCProvider.cpp:
(WebKit::NetworkRTCProvider::createResolver):
(WebKit::NetworkRTCProvider::stopResolver):
2020-07-27 Adrian Perez de Castro <aperez@igalia.com>
[GTK] GeoclueGeolocationProvider uses very low accuracy setting (city) as default geolocation API
https://bugs.webkit.org/show_bug.cgi?id=214566
Reviewed by Carlos Garcia Campos.
No new tests needed.
* UIProcess/geoclue/GeoclueGeolocationProvider.cpp:
(WebKit::GeoclueGeolocationProvider::requestAccuracyLevel): Use
GeoclueAccuracySteetLevel when high accuracy was not requested.
2020-07-27 Carlos Garcia Campos <cgarcia@igalia.com>
[WPE][GTK] Add API to support "Privacy Report"
https://bugs.webkit.org/show_bug.cgi?id=213493
Reviewed by Adrian Perez de Castro.
Add WebKitITPFirstParty and WebKitITPThirdParty boxed types to expose the data returned by the ITP summary.
* UIProcess/API/glib/WebKitWebsiteDataManager.cpp:
(_WebKitITPFirstParty::_WebKitITPFirstParty):
(webkitITPFirstPartyCreate):
(webkit_itp_first_party_ref):
(webkit_itp_first_party_unref):
(webkit_itp_first_party_get_domain):
(webkit_itp_first_party_get_website_data_access_allowed):
(webkit_itp_first_party_get_last_update_time):
(_WebKitITPThirdParty::_WebKitITPThirdParty):
(_WebKitITPThirdParty::~_WebKitITPThirdParty):
(webkitITPThirdPartyCreate):
(webkit_itp_third_party_ref):
(webkit_itp_third_party_unref):
(webkit_itp_third_party_get_domain):
(webkit_itp_third_party_get_first_parties):
(webkit_website_data_manager_get_itp_summary):
(webkit_website_data_manager_get_itp_summary_finish):
* UIProcess/API/gtk/WebKitAutocleanups.h:
* UIProcess/API/gtk/WebKitWebsiteDataManager.h:
* UIProcess/API/gtk/docs/webkit2gtk-4.0-sections.txt:
* UIProcess/API/wpe/WebKitAutocleanups.h:
* UIProcess/API/wpe/WebKitWebsiteDataManager.h:
* UIProcess/API/wpe/docs/wpe-1.0-sections.txt:
2020-07-27 Youenn Fablet <youenn@apple.com>
Fix null pointer crash in NetworkRTCProvider::createServerTCPSocket
https://bugs.webkit.org/show_bug.cgi?id=214796
Reviewed by Darin Adler.
In case of creating a TCP socket, NetworkRTCProvider will hop to the main thread before creating the socket.
In that case, NetworkRTCProvider may actually be closed between the time of receiving the message and hoping to the main thread.
Protect from this by adding nullptr checks.
* NetworkProcess/webrtc/NetworkRTCProvider.cpp:
(WebKit::NetworkRTCProvider::createServerTCPSocket):
(WebKit::NetworkRTCProvider::createClientTCPSocket):
2020-07-25 Simon Fraser <simon.fraser@apple.com>
Scroll Snap broken when using RTL layout
https://bugs.webkit.org/show_bug.cgi?id=193671
<rdar://problem/47457471>
Reviewed by Wenson Hsieh.
There were various places in the scroll snapping code which confused scroll offsets (zero-based)
with scroll positions (relative to scroll origin, can be negative). This broke snapping in
`direction: rtl` content.
The computation of snap locations in updateSnapOffsetsForScrollableArea() was also broken
in RTL; we need to snap the right edge of boxes to the right edge of the scroller for the 'start'
alignment. (Snapping for LTR boxes in an RTL scroller is poorly defined; for now, follow Gecko.)
* Shared/RemoteLayerTree/RemoteScrollingCoordinatorTransaction.cpp:
(WebKit::dump):
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _updateVisibleContentRects]):
* UIProcess/RemoteLayerTree/RemoteScrollingCoordinatorProxy.h:
* UIProcess/RemoteLayerTree/ios/RemoteScrollingCoordinatorProxyIOS.mm:
(WebKit::RemoteScrollingCoordinatorProxy::shouldSnapForMainFrameScrolling const):
(WebKit::RemoteScrollingCoordinatorProxy::closestSnapOffsetForMainFrameScrolling const):
(WebKit::RemoteScrollingCoordinatorProxy::hasActiveSnapPoint const):
(WebKit::RemoteScrollingCoordinatorProxy::nearestActiveContentInsetAdjustedSnapOffset const):
(WebKit::RemoteScrollingCoordinatorProxy::nearestActiveContentInsetAdjustedSnapPoint const): Deleted.
2020-07-26 Wenson Hsieh <wenson_hsieh@apple.com>
REGRESSION (r264690): [ iOS wk2 ] editing/selection/shift-click-includes-existing-selection.html is a flaky failure
https://bugs.webkit.org/show_bug.cgi?id=214688
<rdar://problem/65993683>
Reviewed by Darin Adler.
Remove logging added in r264882, r264890, r264892, and r264896.
* Platform/spi/ios/UIKitSPI.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _singleTapIdentified:]):
(-[WKContentView _singleTapRecognized:]):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::commitPotentialTap):
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::handleSyntheticClick):
2020-07-22 Darin Adler <darin@apple.com>
Stop using live ranges in SpellChecker.h and TextCheckingHelper.h
https://bugs.webkit.org/show_bug.cgi?id=214648
Reviewed by Sam Weinig.
* WebProcess/WebCoreSupport/WebEditorClient.cpp: Added an include
of Range.h, no longer pulled in indirectly by another header.
2020-07-26 Wenson Hsieh <wenson_hsieh@apple.com>
Unreviewed, partially revert r264896.
https://bugs.webkit.org/show_bug.cgi?id=214688
This logging is no longer needed.
* Platform/spi/ios/UIKitSPI.h:
2020-07-26 Wenson Hsieh <wenson_hsieh@apple.com>
Unreviewed, add even more logging to try and diagnose a flaky test
https://bugs.webkit.org/show_bug.cgi?id=214688
<rdar://problem/65993683>
Add another IPI declaration on UIApplication.
* Platform/spi/ios/UIKitSPI.h:
2020-07-25 Wenson Hsieh <wenson_hsieh@apple.com>
REGRESSION (r264690): [ iOS wk2 ] editing/selection/shift-click-includes-existing-selection.html is a flaky failure
https://bugs.webkit.org/show_bug.cgi?id=214688
<rdar://problem/65993683>
Reviewed by Tim Horton.
Add an SPI declaration on UIApplication.
* Platform/spi/ios/UIKitSPI.h:
2020-07-24 Wenson Hsieh <wenson_hsieh@apple.com>
REGRESSION (r264690): [ iOS wk2 ] editing/selection/shift-click-includes-existing-selection.html is a flaky failure
https://bugs.webkit.org/show_bug.cgi?id=214688
<rdar://problem/65993683>
Reviewed by Megan Gardner.
Add more logging to try and see why this test is failing on some post-commit iOS bots. In particular, these logs
should test a theory that it's possible for `UIScriptControllerIOS::singleTapAtPointWithModifiers` to try and
synthesize and send HID events for the tap before GraphicsServices has registered the fact that a modifier key
is being held, which would lead to a regular tap instead of a shift-tap in this test.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _singleTapIdentified:]):
(-[WKContentView _singleTapRecognized:]):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::commitPotentialTap):
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::handleSyntheticClick):
2020-07-24 Brady Eidson <beidson@apple.com>
Hard code some MFi controller devices instead of dynamically managing HID vs GameController.framework.
rdar://problem/65961406 and https://bugs.webkit.org/show_bug.cgi?id=214664
Reviewed by Tim Horton.
Add more test-only SPI to monitor the state of connected gamepads.
* UIProcess/API/Cocoa/WKProcessPool.mm:
(-[WKProcessPool _numberOfConnectedGamepadsForTesting]):
(-[WKProcessPool _numberOfConnectedHIDGamepadsForTesting]):
(-[WKProcessPool _numberOfConnectedGameControllerFrameworkGamepadsForTesting]):
* UIProcess/API/Cocoa/WKProcessPoolPrivate.h:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::numberOfConnectedGamepadsForTesting):
* UIProcess/WebProcessPool.h:
2020-07-24 Devin Rousso <drousso@apple.com>
Followup, adopt more descriptive symbol image name for non-LTR
https://bugs.webkit.org/show_bug.cgi?id=214764
<rdar://problem/64000059>
Reviewed by Tim Horton.
* UIProcess/API/Cocoa/_WKElementAction.mm:
(+[_WKElementAction imageForElementActionType:]):
2020-07-24 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed, reverting r264858.
Caused spurious CRASHING TEST logging during layout test runs
Reverted changeset:
"WebKit::setCrashReportApplicationSpecificInformation() should
also log to WTFLogAlways()"
https://bugs.webkit.org/show_bug.cgi?id=214756
https://trac.webkit.org/changeset/264858
2020-07-24 Tadeu Zagallo <tzagallo@apple.com>
WKStringGetMaximumUTF8CStringSize should cast before arithmetic
https://bugs.webkit.org/show_bug.cgi?id=214750
<rdar://problem/65884455>
Reviewed by Mark Lam.
When the string is long enough, the multiplication can result in an integer overflow.
Test: js/throw-large-string-oom.html
* Shared/API/c/WKString.cpp:
(WKStringGetMaximumUTF8CStringSize):
2020-07-24 Devin Rousso <drousso@apple.com>
Adopt more descriptive symbol image name
https://bugs.webkit.org/show_bug.cgi?id=214764
<rdar://problem/64000059>
Reviewed by Dean Jackson.
* UIProcess/API/Cocoa/_WKElementAction.mm:
(+[_WKElementAction imageForElementActionType:]):
2020-07-24 Kate Cheney <katherine_cheney@apple.com>
CrashTracer: com.apple.WebKit.WebContent at com.apple.WebKit: WebKit::WebFrame::shouldEnableInAppBrowserPrivacyProtections
https://bugs.webkit.org/show_bug.cgi?id=214745
<rdar://66018965>
Reviewed by Chris Dumez.
We should stop iterating if the frame is null, even if it is not
the main frame, to avoid calling functions on a null frame.
* WebProcess/WebPage/WebFrame.cpp:
(WebKit::WebFrame::shouldEnableInAppBrowserPrivacyProtections):
2020-07-24 David Kilzer <ddkilzer@apple.com>
WebKit::setCrashReportApplicationSpecificInformation() should also log to WTFLogAlways()
<https://webkit.org/b/214756>
Reviewed by Chris Dumez.
* Platform/cocoa/WKCrashReporter.mm:
(WebKit::setCrashReportApplicationSpecificInformation):
- Also output the log information via WTFLogAlways() in case
Application Specific Information is overwritten or is not
included in the crash log.
2020-07-24 Chris Dumez <cdumez@apple.com>
Crash under PreconnectTask::start()
https://bugs.webkit.org/show_bug.cgi?id=214747
<rdar://problem/65989668>
Reviewed by Geoffrey Garen.
All user of PreconnectTask construct a PreconnectTask and then call start() on it.
start() dereferences m_networkLoad unconditionally. However, the PreconnectTask
constructor may return early and not initialize m_networkLoad if there is not
NetworkSession for the given sessionID.
To address the issue, update call sites to pass in a valid NetworkSession, so that
the PreconnectTask now always initializes m_networkLoad.
No new tests, we do not know how to reproduce, likely due to a race between
a preconnect request being issued and a network session getting destroyed.
* NetworkProcess/NetworkConnectionToWebProcess.cpp:
(WebKit::NetworkConnectionToWebProcess::sendH2Ping):
(WebKit::NetworkConnectionToWebProcess::preconnectTo):
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::preconnectTo):
* NetworkProcess/PreconnectTask.cpp:
(WebKit::PreconnectTask::PreconnectTask):
* NetworkProcess/PreconnectTask.h:
* NetworkProcess/cache/NetworkCacheSpeculativeLoadManager.cpp:
(WebKit::NetworkCache::SpeculativeLoadManager::preconnectForSubresource):
2020-07-24 David Kilzer <ddkilzer@apple.com>
[IPC hardening] Add nullptr checks for WebPage::m_activeColorChooser
<https://webkit.org/b/214699>
<rdar://problem/65850975>
Reviewed by Darin Adler.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::didEndColorPicker):
(WebKit::WebPage::didChooseColor):
- Add nullptr check before dereferencing m_activeColorChooser.
2020-07-24 David Kilzer <ddkilzer@apple.com>
Add/Update CRASH_WITH_INFO() for invalid IPC message logging
<https://webkit.org/b/214713>
Reviewed by Anders Carlsson.
* Platform/IPC/cocoa/ConnectionCocoa.mm:
(IPC::Connection::sendMessage):
- Simplify logging by removing CString and using
IPC::MessageName enum value instead of hash.
* Shared/Cocoa/AuxiliaryProcessCocoa.mm:
(WebKit::AuxiliaryProcess::didReceiveInvalidMessage):
- Add IPC::MessageName enum value to
setCrashReportApplicationSpecificInformation() output.
- Change CRASH() to CRASH_WITH_INFO() and pass IPC::MessageName
enum value.
2020-07-24 Youenn Fablet <youenn@apple.com>
Use one socket server for the whole network process instead of one socket server per web process connection
https://bugs.webkit.org/show_bug.cgi?id=214593
Reviewed by Alex Christensen.
Instead of a socket server per web process, we will only create one for all web processes.
Conversely, we do the same for the network manager.
This should allow to reduce our use of sockets in network process and hopefully reduce the occurence of sockets fd above FD_SETSIZE.
Covered by existing tests.
* NetworkProcess/webrtc/NetworkRTCMonitor.cpp:
(WebKit::networkManager):
(WebKit::NetworkManagerWrapper::addListener):
(WebKit::NetworkManagerWrapper::removeListener):
(WebKit::NetworkManagerWrapper::onNetworksChanged):
(WebKit::NetworkRTCMonitor::~NetworkRTCMonitor):
(WebKit::NetworkRTCMonitor::startUpdatingIfNeeded):
(WebKit::NetworkRTCMonitor::stopUpdating):
(WebKit::NetworkRTCMonitor::onNetworksChanged):
* NetworkProcess/webrtc/NetworkRTCMonitor.h:
* NetworkProcess/webrtc/NetworkRTCProvider.cpp:
(WebKit::rtcNetworkThread):
(WebKit::NetworkRTCProvider::NetworkRTCProvider):
(WebKit::NetworkRTCProvider::close):
(WebKit::NetworkRTCProvider::createSocket):
(WebKit::NetworkRTCProvider::createUDPSocket):
(WebKit::NetworkRTCProvider::createServerTCPSocket):
(WebKit::NetworkRTCProvider::wrapNewTCPConnection):
(WebKit::NetworkRTCProvider::sendToSocket):
(WebKit::NetworkRTCProvider::closeSocket):
(WebKit::NetworkRTCProvider::setSocketOption):
(WebKit::NetworkRTCProvider::addSocket):
(WebKit::NetworkRTCProvider::takeSocket):
(WebKit::NetworkRTCProvider::newConnection):
(WebKit::NetworkRTCProvider::callOnRTCNetworkThread):
* NetworkProcess/webrtc/NetworkRTCProvider.h:
(WebKit::NetworkRTCProvider::connection):
2020-07-23 Wenson Hsieh <wenson_hsieh@apple.com>
Tapping QuickType suggestions for a misspelled word does nothing in Mail compose
https://bugs.webkit.org/show_bug.cgi?id=214705
<rdar://problem/58320785>
Reviewed by Tim Horton.
In WebKit2-based editing contexts on iOS (such as Mail compose), tapping on a misspelled word will result in
QuickType suggestions in the input accessory row either showing up blank (i.e. with no suggestions), or will
show suggestions that are contextual to the text selection prior to selecting the misspelled word. Subsequently,
attempting to choose of these suggestions results in no text being inserted. This happens for a couple of
reasons:
1. UIKit doesn't attempt to re-request the autocorrection context after tapping on a misspelled word. UIKit
just uses `-selectWordForReplacement` to extend the selection over the misspelled word, which doesn't
currently fire any of the selection change callbacks on UIInputDelegate. As a result, UIKit will instead
believe that the selection is collapsed to the start or end of the misspelled word.
2. Upon attempting to compute text suggestions for QuickType, `-[UIKeyboardImpl replacementsFromSelectedText]`
uses `-textInRange:` to grab the currently selected text, which (in WKContentView) is currently hard-coded
to always return `nil`.
We address (1) by bookending `-selectWordForReplacement` with calls to `-beginSelectionChange` and
`-endSelectionChange`, and address (2) by teaching `-textInRange:` to return the current text selection, only in
the case where the given `UITextRange` is equal to the current ranged selection.
Test: SelectionTests.SelectedTextAfterSelectingWordForReplacement
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::extendSelection):
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView cleanUpInteraction]):
(-[WKContentView textInRange:]):
Special-case the scenario where the input text range is exactly the same as `self.selectedTextRange` by
returning the currently selected text that is cached on EditorState. In all other cases, continue to return
`nil`.
(-[WKContentView selectedTextRange]):
Cache the current selected text range on WKContentView, such that repeated calls to selectedTextRange when the
selection has not changed does not require a new WKTextRange to be constructed every time.
(-[WKContentView _selectionChanged]):
Invalidate the cached selected text range whenever the selection has changed.
(-[WKContentView selectWordForReplacement]):
Use the completion callback added below to bookend this method with `-(begin|end)SelectionChange`.
(-[WKContentView _updateChangedSelection:]):
Invalidate the cached selected text range whenever we're about to tell UIKit to repaint the selection.
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::extendSelection):
Add an optional completion callback to `extendSelection`.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::extendSelection):
2020-07-23 Peng Liu <peng.liu6@apple.com>
Safari Related Leaked Assertions after Playing YouTube Video
https://bugs.webkit.org/show_bug.cgi?id=214706
Reviewed by Jer Noble.
A 10 second grace period before removing the activity token was introduced in r252692.
However, some websites, e.g., youtube.com, requests the video element to play the
next video 10 seconds after completing the current one (when the Safari is in the background).
That can lead to a race condition between the web process and the UI process. They may send
IPC messages to each other at the same time:
1) The web process sends a WebPageProxy::IsPlayingMediaDidChange message to the UI process
to notify that video/audio is playing, so the UI process will take a MediaPlayback assertion.
2) The UI process sends a WebProcess::PrepareToSuspend message to the web process, which
will suspend the web process.
Eventually, the web process will suspend, which means no video/audio playback.
But the UI process believes that there is a web process playing video/audio, so it will
take a MediaPlayback assertion.
This patch fixes the race condition by ignoring the WebPageProxy::IsPlayingMediaDidChange
message in the UIProcess if the process throttler believes that the process should not be
running, so that the MediaPlayback assertion will not be created.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::isPlayingMediaDidChange):
2020-07-23 Per Arne Vollan <pvollan@apple.com>
Crash in block in WebKit::LaunchServicesDatabaseObserver::LaunchServicesDatabaseObserver
https://bugs.webkit.org/show_bug.cgi?id=214702
Reviewed by Chris Dumez.
Calling xpc_connection_send_message in the block in WebKit::LaunchServicesDatabaseObserver::LaunchServicesDatabaseObserver will
sometimes crash because the XPC connection object is invalid. The block is called on a secondary thread, and the m_connections
member variable needs to be protected by a lock. Also, make sure the XPC connections objects in m_connections are valid by
retaining them when captured in a block. This is done by redeclaring a local XPC connection variable as an OSObjectPtr instead
of a raw XPC connection object.
No new tests, since I have not been able to reproduce this issue.
* NetworkProcess/cocoa/LaunchServicesDatabaseObserver.h:
* NetworkProcess/cocoa/LaunchServicesDatabaseObserver.mm:
(WebKit::LaunchServicesDatabaseObserver::LaunchServicesDatabaseObserver):
(WebKit::LaunchServicesDatabaseObserver::startObserving):
(WebKit::LaunchServicesDatabaseObserver::handleEvent):
* Shared/Cocoa/XPCEndpoint.mm:
(WebKit::XPCEndpoint::XPCEndpoint):
2020-07-23 Brian Burg <bburg@apple.com>
Web Inspector: developerExtrasEnabled should be respected when opening local Web Inspector (part 2)
https://bugs.webkit.org/show_bug.cgi?id=214669
Reviewed by Devin Rousso and Joseph Pecoraro.
* UIProcess/Inspector/WebInspectorProxy.cpp:
(WebKit::WebInspectorProxy::save):
(WebKit::WebInspectorProxy::append): Add missing check for developerExtrasEnabled.
2020-07-23 Sihui Liu <sihui_liu@appe.com>
Allow IndexedDB in third-party frames
https://bugs.webkit.org/show_bug.cgi?id=214583
<rdar://problem/61278487>
Reviewed by Geoffrey Garen.
Do no append extra space to displayName if host is empty.
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::fetchDataAndApply):
2020-07-23 David Quesada <david_quesada@apple.com>
Unable to present file upload picker after interactively dismissing a previously opened picker
https://bugs.webkit.org/show_bug.cgi?id=209613
rdar://problem/60931009
Reviewed by Daniel Bates.
-[<UIDocumentPickerViewController> documentPickerWasCancelled:] doesn't get called if you
dismiss the view controller using the interactive swipe dismissal gesture added in iOS 13.
As a result, the web view will continue to think it's showing a file picker after dismissing
it this way, meaning you can't present any other file pickers from the same web view. Fix
this by implementing UIAdaptivePresentationControllerDelegate in WKFileUploadPanel to cancel
and tear down the file picker operation when the view controller is interactively dismised.
* UIProcess/ios/forms/WKFileUploadPanel.mm:
(-[WKFileUploadPanel showFilePickerMenu]):
(-[WKFileUploadPanel presentationControllerDidDismiss:]):
2020-07-23 Aditya Keerthi <akeerthi@apple.com>
[macOS] Datalist dropdown scrollbar position does not match visible region
https://bugs.webkit.org/show_bug.cgi?id=214656
Reviewed by Wenson Hsieh.
Currently, the content insets for the dropdown are set on an NSScrollView's contentView,
rather than the NSScrollView itself. On the latest macOS, this results in accurate state
within the NSScrollView, leading to an inaccurate scrollbar position. To fix, set the
content insets directly on the NSScrollView.
However, on older macOS, setting content insets on the NSScrollView causes the view to start
out slightly scrolled (by a value equal to the top inset). This issue is solved by preserving
the existing behavior (setting insets on the contentView) on older macOS.
Also, set the style of the embedded NSTableView to NSTableViewStyleFullWidth, to ensure the
view respects the insets we specify.
No new tests, as the change is purely visual and affects native UI.
* UIProcess/mac/WebDataListSuggestionsDropdownMac.mm:
(-[WKDataListSuggestionTableView initWithElementRect:]):
(-[WKDataListSuggestionsController initWithInformation:inView:]):
2020-07-23 David Kilzer <ddkilzer@apple.com>
[IPC hardening] WebKit::ArgumentCoder<BlobPart>::decode() and encode() should use enum BlobPart::Type
<https://webkit.org/b/214665>
<rdar://problem/65777948>
Reviewed by Darin Adler.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<BlobPart>::encode):
- Change BlobPart::Type enums to use fully qualified name.
- Encode using WebCore::BlobPart::Type value.
- Change break statements to early return statements.
- Add ASSERT_NOT_REACHED() to catch bugs.
(IPC::ArgumentCoder<BlobPart>::decode):
- Change BlobPart::Type enums to use fully qualified name.
- Decode using Optional<WebCore::BlobPart::Type> variable.
- Change break statements to early return statements.
- Remove default: label, move `return WTF::nullopt` to the end
of the method, and add ASSERT_NOT_REACHED() to catch bugs.
2020-07-23 David Kilzer <ddkilzer@apple.com>
[IPC hardening] FilterOperation decode/encode should use early returns
<https://webkit.org/b/214667>
<rdar://problem/65946400>
Reviewed by Darin Adler.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<FilterOperation>::encode):
- Change break statements to early return statements.
- Add ASSERT_NOT_REACHED() to catch bugs.
(IPC::decodeFilterOperation):
- Change break statements to early return statements.
- Remove call to `decoder.markInvalid()` since it is redundant.
- Return `false` from FilterOperation::APPLE_INVERT_LIGHTNESS
label to match FilterOperation::NONE and
FilterOperation::REFERENCE.
- Add ASSERT_NOT_REACHED() to catch bugs.
- Change final return statement from `true` to `false` since
valid decoding returns `true` earlier.
2020-07-23 Brady Eidson <beidson@apple.com>
Add Gamepad tests that exercise the native frameworks
<rdar://problem/65343674> and https://bugs.webkit.org/show_bug.cgi?id=214188
Reviewed by Tim Horton.
Add some testing-only SPI for TestWebKitAPI.
* UIProcess/API/Cocoa/WKProcessPool.mm:
(-[WKProcessPool _numberOfConnectedGamepadsForTesting]):
(-[WKProcessPool _setUsesOnlyHIDGamepadProviderForTesting:]):
* UIProcess/API/Cocoa/WKProcessPoolPrivate.h:
* UIProcess/Gamepad/UIGamepadProvider.h:
(WebKit::UIGamepadProvider::numberOfConnectedGamepads const):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::numberOfConnectedGamepadsForTesting):
(WebKit::WebProcessPool::setUsesOnlyHIDGamepadProviderForTesting):
* UIProcess/WebProcessPool.h:
2020-07-22 Jer Noble <jer.noble@apple.com>
[Cocoa] Add feature flag for WebM & VP9
https://bugs.webkit.org/show_bug.cgi?id=214672
<rdar://problem/65959506>
Reviewed by Beth Dakin.
* FeatureFlags/WebKit.plist:
* Shared/WebPreferences.yaml:
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultVP9DecoderEnabled):
(WebKit::defaultWebMParserEnabled):
* Shared/WebPreferencesDefaultValues.h:
2020-07-22 Conrad Shultz <conrad_shultz@apple.com>
Update macOS Version macros
https://bugs.webkit.org/show_bug.cgi?id=214653
Reviewed by Tim Horton.
* Configurations/Base.xcconfig:
* Configurations/DebugRelease.xcconfig:
* Configurations/Version.xcconfig:
* Configurations/WebKitTargetConditionals.xcconfig:
2020-07-22 Fujii Hironori <Hironori.Fujii@sony.com>
[WinCairo][32bit] ConnectionWin.cpp(327): error C2664: 'BOOL RegisterWaitForSingleObject(...)': cannot convert argument 3 from 'void (__cdecl *)(...)' to 'WAITORTIMERCALLBACK'
https://bugs.webkit.org/show_bug.cgi?id=214658
Reviewed by Darin Adler.
__cdecl and __stdcall calling conventions don't match.
<https://docs.microsoft.com/en-us/cpp/cpp/argument-passing-and-naming-conventions>
* Platform/IPC/Connection.h: Added WINAPI to Connection::EventListener::callback method.
2020-07-22 David Kilzer <ddkilzer@apple.com>
Follow-up: [IPC hardening] WebKit::ObjCObjectGraph::decode() and encode() should use enum ObjCType type
<https://webkit.org/b/214618>
<rdar://problem/65777899>
Unreviewed fix to get rid of extra namespace blocks.
* Shared/mac/ObjCObjectGraph.mm:
(enum class WebKit::ObjCType):
- Move enum class definition back to its original location.
(WTF::EnumTraits<WebKit::ObjCType>):
- Move to the end of the file. Noticed this worked in
ColorSpaceData.mm.
2020-07-22 David Kilzer <ddkilzer@apple.com>
[IPC hardening] WebKit::ObjCObjectGraph::decode() and encode() should use enum ObjCType type
<https://webkit.org/b/214618>
<rdar://problem/65777899>
Unreviewed fix to include the better header.
* Shared/mac/ObjCObjectGraph.mm:
- Switch #include from OptionSet.h to EnumTraits.h.
2020-07-22 Jer Noble <jer.noble@apple.com>
Unreviewed build fix after r264710; add a HAVE_AVPLAYER_VIDEORANGEOVERRIDE guard.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::updatePageScreenProperties):
2020-07-22 Jer Noble <jer.noble@apple.com>
[macOS] Adopt AVPlayer.videoRangeOverride
https://bugs.webkit.org/show_bug.cgi?id=213902
<rdar://problem/63953509>
Reviewed by Youenn Fablet.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::screenPropertiesDidChange):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::updatePageScreenProperties):
2020-07-21 Simon Fraser <simon.fraser@apple.com>
Scrolling tesla.com on iPad is not smooth
https://bugs.webkit.org/show_bug.cgi?id=214627
<rdar://problem/65568979>
Reviewed by Tim Horton.
This is the iOS equivalent of the fix in r264203; if the user is scrolling (or we're animating for
scroll snap), layout should not trigger scroll position changes via computation of the current snap
location.
To fix this, we need to send data from the UI process to the web process about which scrolling tree
nodes the user is scrolling. That's done via RemoteScrollingUIState, which will, in the long term,
become a more general UI -> WebProcess scrolling state updating mechanism. For now, it just keeps
track of nodes undergoing scroll snap and user scrolls (only the latter is hooked up for now).
* Shared/RemoteLayerTree/RemoteScrollingUIState.cpp: Added.
(WebKit::RemoteScrollingUIState::encode const):
(WebKit::RemoteScrollingUIState::decode):
(WebKit::RemoteScrollingUIState::addNodeWithActiveScrollSnap):
(WebKit::RemoteScrollingUIState::removeNodeWithActiveScrollSnap):
(WebKit::RemoteScrollingUIState::addNodeWithActiveUserScroll):
(WebKit::RemoteScrollingUIState::removeNodeWithActiveUserScroll):
* Shared/RemoteLayerTree/RemoteScrollingUIState.h: Added.
(WebKit::RemoteScrollingUIState::changes const):
(WebKit::RemoteScrollingUIState::clearChanges):
(WebKit::RemoteScrollingUIState::nodesWithActiveScrollSnap const):
(WebKit::RemoteScrollingUIState::nodesWithActiveUserScrolls const):
* SourcesCocoa.txt:
* UIProcess/RemoteLayerTree/RemoteScrollingCoordinatorProxy.cpp:
(WebKit::RemoteScrollingCoordinatorProxy::RemoteScrollingCoordinatorProxy):
(WebKit::RemoteScrollingCoordinatorProxy::sendUIStateChangedIfNecessary):
* UIProcess/RemoteLayerTree/RemoteScrollingCoordinatorProxy.h:
* UIProcess/RemoteLayerTree/RemoteScrollingTree.cpp:
(WebKit::RemoteScrollingTree::scrollingTreeNodeWillStartPanGesture):
(WebKit::RemoteScrollingTree::scrollingTreeNodeWillStartScroll):
(WebKit::RemoteScrollingTree::scrollingTreeNodeDidEndScroll):
* UIProcess/RemoteLayerTree/RemoteScrollingTree.h:
* UIProcess/RemoteLayerTree/ios/RemoteScrollingCoordinatorProxyIOS.mm:
(WebKit::RemoteScrollingCoordinatorProxy::scrollViewForScrollingNodeID const):
(WebKit::RemoteScrollingCoordinatorProxy::scrollingTreeNodeWillStartPanGesture):
(WebKit::RemoteScrollingCoordinatorProxy::scrollingTreeNodeWillStartScroll):
(WebKit::RemoteScrollingCoordinatorProxy::scrollingTreeNodeDidEndScroll):
* UIProcess/RemoteLayerTree/ios/ScrollingTreeScrollingNodeDelegateIOS.mm:
(WebKit::ScrollingTreeScrollingNodeDelegateIOS::scrollWillStart const):
(WebKit::ScrollingTreeScrollingNodeDelegateIOS::scrollDidEnd const):
(WebKit::ScrollingTreeScrollingNodeDelegateIOS::scrollViewWillStartPanGesture const):
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebPage/RemoteLayerTree/RemoteScrollingCoordinator.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteScrollingCoordinator.messages.in:
* WebProcess/WebPage/RemoteLayerTree/RemoteScrollingCoordinator.mm:
(WebKit::RemoteScrollingCoordinator::isUserScrollInProgress const):
(WebKit::RemoteScrollingCoordinator::isScrollSnapInProgress const):
(WebKit::RemoteScrollingCoordinator::scrollingStateInUIProcessChanged):
2020-07-22 David Kilzer <ddkilzer@apple.com>
[IPC hardening] WebKit::ObjCObjectGraph::decode() and encode() should use enum ObjCType type
<https://webkit.org/b/214618>
<rdar://problem/65777899>
Reviewed by Youenn Fablet.
* Shared/mac/ObjCObjectGraph.h:
- Drive-by fix to use #pragma once. This is included in some
plain C++ source files.
* Shared/mac/ObjCObjectGraph.mm:
(WebKit::ObjCType):
- Move enum definition to the top of the file so that the
EnumTraits<> can be defined after it, but before it's used.
(WTF::EnumTraits<WebKit::ObjCType>): Add.
- Define so that IPC::Decoder and IPC::Encoder can validate enum
values.
(WebKit::ObjCObjectGraph::encode):
- Encode using WebKit::ObjCType value.
- Add ObjCType::Null label so that default: label can be
removed. Change break statements to early return statements.
- Move ASSERT_NOT_REACHED() to the end of the method.
(WebKit::ObjCObjectGraph::decode):
- Decode using WebKit::ObjCType value.
- Change break statements to early return statements.
- Remove default: label, and move `return false` to the end of
the method. Add ASSERT_NOT_REACHED().
2020-07-22 Megan Gardner <megan_gardner@apple.com>
Tapped DataDetected links present sub-menus from the wrong location.
https://bugs.webkit.org/show_bug.cgi?id=214628
<rdar://problem/64808642>
When getting the positionInformation associated with presenting DataDetectors UI
we were not indicating that we need the linkIndicator data. This caused the
resulting CGRect to not be filled out, and we were returning a zeroed out CGRect
to the DataDetectors framework for the location that it should present the subMenu from.
Setting this bit ensures that all the necessary data is calculated and returned
when a DataDetector link is tapped.
Reviewed by Wenson Hsieh.
* WebProcess/WebCoreSupport/ios/WebChromeClientIOS.mm:
(WebKit::WebChromeClient::showDataDetectorsUIForElement):
2020-07-22 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK][WPE] ITP: use the database store unconditionally
https://bugs.webkit.org/show_bug.cgi?id=214590
Reviewed by Adrian Perez de Castro.
The memory store is now considered legacy.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::ResourceLoadStatisticsDatabaseStore): Use
FileSystem::pathByAppendingComponent() to create the database path.
(WebKit::ResourceLoadStatisticsDatabaseStore::openITPDatabase): Ensure the directory exists before trying to
open the database.
* UIProcess/soup/WebProcessPoolSoup.cpp:
(WebKit::WebProcessPool::platformInitializeNetworkProcess): Always enable the ITP databae by default.
2020-07-19 Darin Adler <darin@apple.com>
Remove live ranges from Editor.h and EditorClient.h
https://bugs.webkit.org/show_bug.cgi?id=214261
Reviewed by Sam Weinig.
* UIProcess/mac/WKPrintingView.mm:
(-[WKPrintingView rectForPage:]): Use NSZeroRect.
* WebProcess/WebCoreSupport/WebEditorClient.cpp:
(WebKit::WebEditorClient::shouldDeleteRange): Updated for live
range changes and also removed gratuitous call to notImplemented
since the function is implemented.
(WebKit::WebEditorClient::shouldBeginEditing): Ditto.
(WebKit::WebEditorClient::shouldEndEditing): Ditto.
(WebKit::WebEditorClient::shouldInsertNode): Ditto.
(WebKit::WebEditorClient::shouldInsertText): Ditto.
(WebKit::WebEditorClient::shouldChangeSelectedRange): Ditto.
(WebKit::WebEditorClient::shouldApplyStyle): Ditto.
(WebKit::WebEditorClient::shouldMoveRangeAfterDelete): Ditto.
(WebKit::WebEditorClient::didBeginEditing): Ditto.
(WebKit::WebEditorClient::didEndEditing): Ditto.
(WebKit::WebEditorClient::willWriteSelectionToPasteboard): Ditto.
(WebKit::WebEditorClient::getClientPasteboardData): Ditto.
(WebKit::WebEditorClient::performTwoStepDrop): Ditto.
(WebKit::WebEditorClient::overflowScrollPositionChanged): Ditto.
(WebKit::WebEditorClient::subFrameScrollPositionChanged): Ditto.
(WebKit::WebEditorClient::setInputMethodState): Ditto.
* WebProcess/WebCoreSupport/WebEditorClient.h: Updated
for changes to EditorClient.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::getMarkedRangeAsync): Updated to reduce
the use of local values and live ranges.
(WebKit::WebPage::getSelectedRangeAsync): Ditto.
(WebKit::WebPage::characterIndexForPointAsync): Ditto.
(WebKit::WebPage::firstRectForCharacterRangeAsync): Ditto.
(WebKit::WebPage::setCompositionAsync): Ditto.
* WebProcess/WebPage/glib/WebPageGLib.cpp:
(WebKit::WebPage::getPlatformEditorState const): Ditto.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::getPlatformEditorState const): Ditto.
(WebKit::WebPage::selectWithGesture): Ditto.
(WebKit::WebPage::autocorrectionContext): Ditto.
(WebKit::focusedElementPositionInformation): Ditto.
(WebKit::WebPage::requestDocumentEditingContext): Ditto.
2020-07-21 Jer Noble <jer.noble@apple.com>
[Cocoa] Add experimental MSE WebM parser
https://bugs.webkit.org/show_bug.cgi?id=214529
<rdar://problem/65782467>
Reviewed by Eric Carlson.
Add an Experimental setting to enable the MSE WebM parser (off by default).
* Shared/WebPreferences.yaml:
2020-07-21 Tim Horton <timothy_horton@apple.com>
watchOS: Web Content process crashes when it tries to open files
https://bugs.webkit.org/show_bug.cgi?id=214614
<rdar://problem/65900517>
Reviewed by Darin Adler.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
Allow reading kern.maxfilesperproc, as it is allowed on macOS and in the Network process sandbox, and libc needs it.
2020-07-21 Per Arne Vollan <pvollan@apple.com>
[macOS] Layout tests exiting early with crashes
https://bugs.webkit.org/show_bug.cgi?id=214612
<rdar://problem/65897905>
Reviewed by Brent Fulgham.
This is caused by the WebContent process making a syscall not allowed by the sandbox.
* WebProcess/com.apple.WebProcess.sb.in:
2020-07-21 David Kilzer <ddkilzer@apple.com>
WebKit::ColorSpaceData::decode() does not need default: case
<https://webkit.org/b/214587>
<rdar://problem/65777868>
Reviewed by Darin Adler.
* Shared/mac/ColorSpaceData.mm:
(WebKit::ColorSpaceData::decode):
- Remove default: case and add ASSERT_NOT_REACHED() and
move `return false;` to bottom of the method.
2020-07-21 James Savage <james.savage@apple.com>
Unable to call WKScriptMessageHandlerWithReply's replyHandler with a nil errorMessage in Swift
https://bugs.webkit.org/show_bug.cgi?id=214581
<rdar://problem/65639342>
Reviewed by Brady Eidson.
* UIProcess/API/Cocoa/WKScriptMessageHandlerWithReply.h: Modify the block
signature to allow nil for either argument.
2020-07-21 Kate Cheney <katherine_cheney@apple.com>
Frequent Network error messages: "ResourceLoadStatisticsDatabaseStore::topPrevalentResourceWithUserInteractionDaysSinceUserInteractionStatement query failed to step, error message: no more rows available"
https://bugs.webkit.org/show_bug.cgi?id=214597
<rdar://65884740>
Reviewed by John Wilander.
We should not log an error message when no rows are returned when running
telemetry queries, because this is not an error.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::getMedianOfPrevalentResourcesWithUserInteraction):
(WebKit::ResourceLoadStatisticsDatabaseStore::getTopPrevelentResourceDaysSinceUI const):
Remove unnecessary error logging and only check for telemetry results if the
query returns a row.
2020-07-21 Youenn Fablet <youenn@apple.com>
NetworkConnectionToWebProcess should not handle NetworkRTCProvider messages
https://bugs.webkit.org/show_bug.cgi?id=214547
Reviewed by Alex Christensen.
Normally NetworkRTCProvider is created when starting to monitor the network.
In case network process is crashing though, there might be requests to open RTC socket on the new network process which has not started to monitor the network.
In that case, we will go through NetworkConnectionToWebProcess as NetworkRTCProvider is not yet created.
To properly handle this case, on web process side, we send an IPC message to explictly create the NetworkRTCProvider.
Once done, we set the IPC connection to the socket factory which will start sending messages to open sockets if needed.
Until its connection is set, the socket factory will buffer messages to open sockets.
Covered by test no longer crashing in debug.
* NetworkProcess/NetworkConnectionToWebProcess.cpp:
(WebKit::NetworkConnectionToWebProcess::didReceiveMessage):
(WebKit::NetworkConnectionToWebProcess::createRTCProvider):
* NetworkProcess/NetworkConnectionToWebProcess.h:
* NetworkProcess/NetworkConnectionToWebProcess.messages.in:
* WebProcess/Network/webrtc/LibWebRTCNetwork.cpp:
(WebKit::LibWebRTCNetwork::setAsActive):
(WebKit::LibWebRTCNetwork::setConnection):
(WebKit::LibWebRTCNetwork::setSocketFactoryConnection):
* WebProcess/Network/webrtc/LibWebRTCNetwork.h:
* WebProcess/Network/webrtc/LibWebRTCSocketFactory.cpp:
(WebKit::LibWebRTCSocketFactory::setConnection):
(WebKit::LibWebRTCSocketFactory::createServerTcpSocket):
(WebKit::LibWebRTCSocketFactory::createUdpSocket):
(WebKit::LibWebRTCSocketFactory::createClientTcpSocket):
(WebKit::LibWebRTCSocketFactory::createNewConnectionSocket):
* WebProcess/Network/webrtc/LibWebRTCSocketFactory.h:
2020-07-21 Youenn Fablet <youenn@apple.com>
Fetch/XHR loads done by extensions should opt out of response sanitisation done in network process
https://bugs.webkit.org/show_bug.cgi?id=214588
<rdar://problem/65060560>
Reviewed by Alex Christensen.
Send to network process the CORS disabling patterns.
Disable sanitization for fetch/xhr loads whose pages have a matching pattern.
Covered by API test.
* NetworkProcess/NetworkConnectionToWebProcess.cpp:
(WebKit::NetworkConnectionToWebProcess::setCORSDisablingPatterns):
* NetworkProcess/NetworkConnectionToWebProcess.h:
* NetworkProcess/NetworkConnectionToWebProcess.messages.in:
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::shouldDisableCORSForRequestTo const):
(WebKit::NetworkProcess::setCORSDisablingPatterns):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkResourceLoader.cpp:
(WebKit::shouldSanitizeResponse):
(WebKit::NetworkResourceLoader::sanitizeResponseIfPossible):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::parseAndAllowAccessToCORSDisablingPatterns):
(WebKit::m_limitsNavigationsToAppBoundDomains):
(WebKit::WebPage::~WebPage):
(WebKit::WebPage::updateCORSDisablingPatterns):
(WebKit::WebPage::synchronizeCORSDisablingPatternsWithNetworkProcess):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::ensureNetworkProcessConnection):
2020-07-21 Eric Carlson <eric.carlson@apple.com>
Use AVRoutePickerView when available for choosing AirPlay devices
https://bugs.webkit.org/show_bug.cgi?id=213497
<rdar://problem/58610662>
Reviewed by Jer Noble.
* UIProcess/WebPageProxy.h:
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::platformView const): Add new override.
2020-07-21 Jer Noble <jer.noble@apple.com>
[Cocoa] Adopt VTRegisterSupplementalVideoDecoderIfAvailable
https://bugs.webkit.org/show_bug.cgi?id=214585
<rdar://problem/65863651>
Reviewed by Youenn Fablet.
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::enableVP9Decoder):
2020-07-21 Michael Catanzaro <mcatanzaro@gnome.org>
[GTK] WebKitSettings:enable-private-browsing missing G_PARAM_DEPRECATED
https://bugs.webkit.org/show_bug.cgi?id=214554
Reviewed by Carlos Garcia Campos.
* UIProcess/API/glib/WebKitSettings.cpp:
(webkit_settings_class_init):
2020-07-21 Adrian Perez de Castro <aperez@igalia.com>
Unreviewed. Update OptionsWPE.cmake and NEWS for the 2.29.1 release
* wpe/NEWS: Add release notes for 2.29.1.
2020-07-21 Carlos Garcia Campos <cgarcia@igalia.com>
Fix ITP data summary when using the memory store and blocking all third-party cookies
https://bugs.webkit.org/show_bug.cgi?id=214401
Reviewed by Michael Catanzaro.
The memory store behaves differently than the database one, it only includes prevalent resources in the list,
while the database store includes all resources when cooking blocking policy is All.
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:
(WebKit::ResourceLoadStatisticsMemoryStore::aggregatedThirdPartyData const): Only check isPrevalentResource if
third party cookie blocking mode is not All to decide whether to include the domain in the report.
2020-07-21 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] Prepare unit tests for GTK4
https://bugs.webkit.org/show_bug.cgi?id=214251
Reviewed by Adrian Perez de Castro.
Make it possible to test input methods.
* UIProcess/API/glib/InputMethodFilter.h:
* UIProcess/API/gtk/InputMethodFilterGtk.cpp:
(WebKit::InputMethodFilter::filterKeyEvent): Added to receive the key event properties instead of the GdkEvent.
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseSynthesizeMouseEvent): Create a GdkEvent for right clicks since it's needed by GTK3 API.
(webkitWebViewBaseSynthesizeKeyEvent): Send key press and release events to the IM filter.
2020-07-20 Megan Gardner <megan_gardner@apple.com>
Add OK button to Date/Time form controls.
https://bugs.webkit.org/show_bug.cgi?id=214195
<rdar://problem/65006000>
Reviewed by Darin Adler.
Add 'OK' button to add clarity to UI, and increase the default size of the control to
ensure space for both buttons. Also fixed a crash that happened when closing the control before
the animation completed.
* UIProcess/ios/forms/WKDateTimeInputControl.mm:
(-[WKDateTimeContextMenuViewController preferredContentSize]):
(-[WKDateTimePicker contextMenuInteraction:configurationForMenuAtLocation:]):
(-[WKDateTimePicker contextMenuInteraction:willDisplayMenuForConfiguration:animator:]):
(-[WKDateTimePicker contextMenuInteraction:willEndForConfiguration:animator:]):
(-[WKDateTimePicker ok:]):
2020-07-20 Adrian Perez de Castro <aperez@igalia.com>
Non unified build fixes, midsummer 2020 edition
https://bugs.webkit.org/show_bug.cgi?id=213616
Unreviewed build fix.
* NetworkProcess/soup/WebSocketTaskSoup.cpp: Add missing inclusion of NetworkProcess.h
2020-07-20 Brian Burg <bburg@apple.com>
Web Inspector: developerExtrasEnabled should be respected when opening local Web Inspector
https://bugs.webkit.org/show_bug.cgi?id=214573
<rdar://problem/65848496>
Reviewed by David Kilzer.
Ensure that WKPreferences.developerExtrasEnabled is true prior to opening a Web Inspector
window. This matches the check in WebCore when deciding to show "Inspect Element" menu item.
* UIProcess/Inspector/WebInspectorProxy.cpp:
(WebKit::WebInspectorProxy::connect):
(WebKit::WebInspectorProxy::openLocalInspectorFrontend):
2020-07-20 Adrian Perez de Castro <aperez@igalia.com>
[WPE] WebPopupItem should always be forward declared as struct
https://bugs.webkit.org/show_bug.cgi?id=214569
Unreviewed build fix.
No new tests needed.
* UIProcess/API/wpe/WebKitWebViewClient.h: Forward-declare WebPopupItem as a struct.
2020-07-20 Alex Christensen <achristensen@webkit.org>
REGRESSION(r263551) Main thread often hangs while saving cookies
https://bugs.webkit.org/show_bug.cgi?id=214564
<rdar://problem/65779712>
Reviewed by Geoffrey Garen.
I had assumed that CFNetwork wouldn't write cookies on the main thread, but I was wrong.
That change caused spins, so let's just do the work on the WebsiteDataStore's queue instead.
* UIProcess/API/APIHTTPCookieStore.h:
* UIProcess/API/Cocoa/APIHTTPCookieStoreCocoa.mm:
(API::HTTPCookieStore::flushDefaultUIProcessCookieStore):
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::dispatchOnQueue):
* UIProcess/WebsiteData/WebsiteDataStore.h:
2020-07-20 David Kilzer <ddkilzer@apple.com>
WebPageProxyIdentifier parameter of WebPageProxy::CreateNewPage() should not be Optional
<https://webkit.org/b/214542>
<rdar://problem/65768644>
Reviewed by Darin Adler.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::createNewPage):
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::createNewPage):
* UIProcess/WebPageProxy.messages.in:
- Remove Optional<> wrapper for WebPageProxyIdentifier
parameter since the value is used unconditionally.
2020-07-20 Adrian Perez de Castro <aperez@igalia.com>
Some classes have final destructors but are non-final
https://bugs.webkit.org/show_bug.cgi?id=214549
Reviewed by Youenn Fablet.
No new tests needed.
* NetworkProcess/ServiceWorker/WebSWServerConnection.h: Mark WebSWServerConnection
class as final.
* WebProcess/WebCoreSupport/WebMessagePortChannelProvider.h: Ditto.
2020-07-20 Carlos Garcia Campos <cgarcia@igalia.com>
valgrind claims uninitialized memory when opening inspector (WebKit::InspectorBrowserAgent::enable(WTF::String&))
https://bugs.webkit.org/show_bug.cgi?id=209468
Reviewed by Michael Catanzaro.
Initialize m_targetAgent.
* UIProcess/Inspector/WebPageInspectorController.h:
2020-07-19 Sam Weinig <weinig@apple.com>
Rename Color::transparent to Color::transparentBlack to more clearly state what it is
https://bugs.webkit.org/show_bug.cgi?id=214522
Reviewed by Darin Adler.
* Shared/RemoteLayerTree/RemoteLayerTreeTransaction.mm:
(WebKit::RemoteLayerTreeTransaction::LayerProperties::LayerProperties):
* UIProcess/API/ios/WKWebViewIOS.mm:
(scrollViewBackgroundColor):
(-[WKWebView _setOpaqueInternal:]):
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::setDrawsBackground):
* UIProcess/WebPageProxy.cpp:
(WebKit::m_transcodingQueue):
* WebProcess/InjectedBundle/API/glib/WebKitWebPage.cpp:
(webkitWebPageDidReceiveMessage):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::snapshotNode):
* WebProcess/cocoa/VideoFullscreenManager.mm:
(WebKit::VideoFullscreenManager::enterVideoFullscreenForVideoElement):
Color::transparent to Color::transparentBlack.
2020-07-19 Sam Weinig <weinig@apple.com>
Remove ColorBuilder
https://bugs.webkit.org/show_bug.cgi?id=214521
Reviewed by Darin Adler.
* WebProcess/Inspector/WebInspectorClient.cpp:
(WebKit::WebInspectorClient::showPaintRect):
* WebProcess/WebPage/FindController.cpp:
(WebKit::FindController::drawRect):
Update for rename from colorWithAlpha to colorWithAlphaByte.
2020-07-19 Michael Catanzaro <mcatanzaro@gnome.org>
Enable cross-platform release logging in ResourceLoadStatisticsDatabaseStore.cpp
https://bugs.webkit.org/show_bug.cgi?id=214538
Reviewed by Darin Adler.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
2020-07-18 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r264563.
https://bugs.webkit.org/show_bug.cgi?id=214518
Broke the watchOS build
Reverted changeset:
"Add OK button to Date/Time form controls."
https://bugs.webkit.org/show_bug.cgi?id=214195
https://trac.webkit.org/changeset/264563
2020-07-18 Rob Buis <rbuis@igalia.com>
Only use enum classes in HTTPParsers
https://bugs.webkit.org/show_bug.cgi?id=214451
Reviewed by Darin Adler.
Adjust to XFrameOptionsDisposition change.
* NetworkProcess/NetworkResourceLoader.cpp:
(WebKit::NetworkResourceLoader::shouldInterruptLoadForXFrameOptions):
2020-07-17 Sam Weinig <weinig@apple.com>
Remove final vestigates of SimpleColor
https://bugs.webkit.org/show_bug.cgi?id=214439
Reviewed by Simon Fraser.
* Shared/RemoteLayerTree/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::drawInContext):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _tapHighlightColorForFastClick:]):
* WebProcess/Inspector/WebInspectorClient.cpp:
(WebKit::WebInspectorClient::showPaintRect):
* WebProcess/WebPage/FindController.cpp:
(WebKit::FindController::drawRect):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::paintSnapshotAtSize):
* WebProcess/WebPage/ios/FindControllerIOS.mm:
Replace remaining makeSimpleColor() uses.
2020-07-17 Megan Gardner <megan_gardner@apple.com>
Add OK button to Date/Time form controls.
https://bugs.webkit.org/show_bug.cgi?id=214195
<rdar://problem/65006000>
Reviewed by Darin Adler.
Add 'OK' button to add clarity to UI, and increase the default size of the control to
ensure space for both buttons. Also fixed a crash that happened when closing the control before
the animation completed.
* UIProcess/ios/forms/WKDateTimeInputControl.mm:
(-[WKDateTimeContextMenuViewController preferredContentSize]):
(-[WKDateTimePicker contextMenuInteraction:configurationForMenuAtLocation:]):
(-[WKDateTimePicker contextMenuInteraction:willDisplayMenuForConfiguration:animator:]):
(-[WKDateTimePicker contextMenuInteraction:willEndForConfiguration:animator:]):
(-[WKDateTimePicker ok:]):
2020-07-17 Lauro Moura <lmoura@igalia.com>
[GTK] fast/events/context-activated-by-key-event.html is crashing
https://bugs.webkit.org/show_bug.cgi?id=199444
Reviewed by Michael Catanzaro.
GTK supports showing the context menu for keyboard events. In these
cases there may be no pending mouse events, causing the assert to
fail.
Covered by existing tests.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::showContextMenu):
2020-07-17 Per Arne Vollan <pvollan@apple.com>
Add Launch Services database update logging
https://bugs.webkit.org/show_bug.cgi?id=214500
Reviewed by Geoffrey Garen.
Log if the WebContent process is waiting a considerable amount of time for the initial update of the Launch Services database.
No new tests, since this patch is not introducing any behavior change.
* WebProcess/WebPage/Cocoa/WebPageCocoa.mm:
(WebKit::WebPage::platformDidReceiveLoadParameters):
2020-07-17 Brent Fulgham <bfulgham@apple.com>
Remove logging for necessary sysctl call
https://bugs.webkit.org/show_bug.cgi?id=214497
<rdar://problem/65737636>
Reviewed by Per Arne Vollan.
We discovered that kern.proc.pid is needed in SecTaskCopyDebugDescription, so
we can remove that telemetry and logging.
* NetworkProcess/mac/com.apple.WebKit.NetworkProcess.sb.in:
* Resources/SandboxProfiles/ios/com.apple.WebKit.Networking.sb:
2020-07-17 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Cache the PIN to improve NFC user experience
https://bugs.webkit.org/show_bug.cgi?id=213900
<rdar://problem/60073622>
Reviewed by Brent Fulgham.
Cache the PIN to improve NFC user experience. Users might not hold the NFC key against the NFC scanner all the time
while entering the PIN. Therefore, we cache the PIN from the previous connection and immediately return the cached PIN to
the current connection such that users don't need to enter the PIN again.
The only downside for this optimization is that a wrong PIN could be used if the user switch to another authenticator for
the new connection. Given there is no UUID to identify a particular authenticator, there is nothing we can do to resolve the
issue. The probability of the issue, however, should be rare.
Covered by new API tests.
* UIProcess/WebAuthentication/AuthenticatorManager.cpp:
(WebKit::AuthenticatorManager::authenticatorStatusUpdated):
(WebKit::AuthenticatorManager::requestPin):
Where the above logic is implemented.
* UIProcess/WebAuthentication/Cocoa/NfcConnection.mm:
(WebKit::NfcConnection::transact const):
Adds a comment.
* UIProcess/WebAuthentication/WebAuthenticationRequestData.h:
* UIProcess/WebAuthentication/WebAuthenticatorCoordinatorProxy.cpp:
(WebKit::WebAuthenticatorCoordinatorProxy::makeCredential):
(WebKit::WebAuthenticatorCoordinatorProxy::getAssertion):
Adds a field for the cached PIN.
2020-07-17 Per Arne Vollan <pvollan@apple.com>
[iOS] Crash under WebProcess::handleXPCEndpointMessages
https://bugs.webkit.org/show_bug.cgi?id=214490
Reviewed by Chris Dumez.
According to crash logs, it seems possible that the parent process XPC connection can be null. Add a null check,
and also check that the connection XPC object really is an XPC connection object.
No new tests, since I have not been able to reproduce this issue.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::handleXPCEndpointMessages const):
2020-07-17 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Use MediaOnly user gesture
https://bugs.webkit.org/show_bug.cgi?id=213595
<rdar://problem/64946612>
Reviewed by Brent Fulgham.
MediaOnly user gesture is more flexible than regular user gesture. For example, it can be propagated through XHR.
A common use case for WebAuthn is to call the API within XHR events, which could be triggered by user activated events initially.
By using MediaOnly user gesture, it allows developer to keep using this XHR model.
* WebProcess/WebAuthentication/WebAuthenticatorCoordinator.cpp:
(WebKit::WebAuthenticatorCoordinator::makeCredential):
(WebKit::WebAuthenticatorCoordinator::getAssertion):
2020-07-17 David Kilzer <ddkilzer@apple.com>
[IPC hardening] Add nullptr checks to WebKit::WebInspectorProxy::platformStartWindowDrag()
<https://webkit.org/b/214436>
<rdar://problem/59645910>
Reviewed by Brent Fulgham.
* UIProcess/Inspector/mac/WebInspectorProxyMac.mm:
(WebKit::WebInspectorProxy::platformStartWindowDrag):
- Add nullptr checks.
2020-07-17 Simon Fraser <simon.fraser@apple.com>
Make ScrollbarControlSize an enum class
https://bugs.webkit.org/show_bug.cgi?id=199325
Reviewed by Sam Weinig.
Make ScrollbarControlSize an enum class, and make some RenderScrollbar methods const.
* UIProcess/win/WebPopupMenuProxyWin.cpp:
(WebKit::WebPopupMenuProxyWin::showPopupMenu):
(WebKit::WebPopupMenuProxyWin::calculatePositionAndSize):
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::createScrollbar):
2020-07-17 Tim Horton <timothy_horton@apple.com>
Add a addition point for extra user media sandbox extensions
https://bugs.webkit.org/show_bug.cgi?id=214445
<rdar://problem/65148262>
Reviewed by Per Arne Vollan.
* UIProcess/UserMediaProcessManager.cpp:
(WebKit::UserMediaProcessManager::willCreateMediaStream):
(WebKit::UserMediaProcessManager::revokeSandboxExtensionsIfNeeded):
* WebProcess/com.apple.WebProcess.sb.in:
2020-07-17 Truitt Savell <tsavell@apple.com>
Unreviewed, reverting r264477.
Broke the internal build
Reverted changeset:
"Use AVRoutePickerView when available for choosing AirPlay
devices"
https://bugs.webkit.org/show_bug.cgi?id=213497
https://trac.webkit.org/changeset/264477
2020-07-16 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Add a console message: "User gesture is required to use the platform authenticator."
https://bugs.webkit.org/show_bug.cgi?id=214380
<rdar://problem/65300587>
Reviewed by Devin Rousso.
Covered by existing tests.
* WebProcess/WebAuthentication/WebAuthenticatorCoordinator.cpp:
(WebKit::WebAuthenticatorCoordinator::makeCredential):
(WebKit::WebAuthenticatorCoordinator::getAssertion):
2020-07-16 Per Arne Vollan <pvollan@apple.com>
XPC connection should be suspended before setting target queue
https://bugs.webkit.org/show_bug.cgi?id=214427
Reviewed by Darin Adler.
According to system logs, an active XPC connection should be suspended before setting a new target queue.
No new tests, since there should be no behavior change introduced by this patch.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::handleXPCEndpointMessages const):
2020-07-16 Fujii Hironori <Hironori.Fujii@sony.com>
[WTF] Remove the unnecessary inner class DefaultHash<T>::Hash
https://bugs.webkit.org/show_bug.cgi?id=214389
Reviewed by Darin Adler.
* NetworkProcess/Downloads/DownloadID.h:
* NetworkProcess/cache/NetworkCache.h:
* NetworkProcess/cache/NetworkCacheKey.h:
* Platform/IPC/MessageReceiverMap.h:
* Platform/IPC/StringReference.h:
* Shared/CallbackID.h:
* UIProcess/API/cpp/WKRetainPtr.h:
2020-07-16 Takashi Komori <Takashi.Komori@sony.com>
[WinCairo][CURL] NetworkProcess consumes CPU resource greedily
https://bugs.webkit.org/show_bug.cgi?id=214383
Reviewed by Fujii Hironori.
Add PolicyAction::Ignore handling routine in NetworkDataTaskCurl::invokeDidReceiveResponse()
If we don't cancel the ignored requests the worker thread for curl scheduler goes busy-looping because the fds of the requests continue to be monitored.
No tests needed as existing tests cover this change.
* NetworkProcess/curl/NetworkDataTaskCurl.cpp:
(WebKit::NetworkDataTaskCurl::invokeDidReceiveResponse):
2020-07-16 Tim Horton <timothy_horton@apple.com>
iPad cursor is sometimes slow to change to I-beam (e.g. on reddit.com)
https://bugs.webkit.org/show_bug.cgi?id=214424
<rdar://problem/59503572>
Reviewed by Wenson Hsieh.
No new tests, just an optimization.
* Shared/ios/InteractionInformationRequest.h:
* Shared/ios/InteractionInformationRequest.cpp:
(WebKit::InteractionInformationRequest::encode const):
(WebKit::InteractionInformationRequest::decode):
(WebKit::InteractionInformationRequest::isValidForRequest):
(WebKit::InteractionInformationRequest::isApproximatelyValidForRequest):
'const'-ify these getters
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::positionInformation):
Make it possible to disable computation of the 'nodeAtPositionHasDoubleClickHandler' property,
because it can be quite expensive for some web content. We will still compute it by default,
but high-rate interaction information requests (like those for the cursor) can optionally disable it.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView cleanUpInteraction]):
(-[WKContentView requestAsynchronousPositionInformationUpdate:]):
(-[WKContentView _hasValidOutstandingPositionInformationRequest:]):
(-[WKContentView _positionInformationDidChange:]):
When position information updates come in, only clear the outstanding request
if the reply is valid for that request; otherwise, the outstanding request is
definitely still outstanding, and we've probably just sent multiple requests.
A more formal mechanism that actually keeps track of all requests might be
more appropriate here in the future, but this helps us avoid re-requesting
position information for the same point repeatedly, just because old (stale)
updates are coming in.
(-[WKContentView _cursorInteraction:regionForLocation:defaultRegion:completion:]):
Instead of requesting an interaction information update for every cursor update,
and then dropping them on the floor when they return, if a new request was sent,
simply coalesce all requests that come in while we have an outstanding one.
This avoids a great deal of unnecessary (and ignored) work when the Web Content
process is responding to requests at a slower rate than the cursor interaction
is trying to retrieve new regions.
2020-07-16 Megan Gardner <megan_gardner@apple.com>
Selection is not always clearing when tapping.
https://bugs.webkit.org/show_bug.cgi?id=214326
<rdar://problem/65069201>
Reviewed by Wenson Hsieh.
In r262280 we stopped the UIWKGestureOneFingerTap gesture from starting if the tap was not inside
the current selection. That caused the selection to not always be cleared, especially when tapping
on an element that could create an overlay in which the selection was supposed to be obscured. We
short circuited this gesture because it used to cause a sync IPC message to be sent to the web process,
but that code has subsequently changed, so allowing the gesture to start and clear the selection in all
cases is no longer a performance concern, so changing the behavior back to allowing the gesture to always
start, even if the touch is not inside the selection rect.
Test: editing/selection/ios/hide-selection-after-tap-on-prevent-default-element.html
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView textInteractionGesture:shouldBeginAtPoint:]):
2020-07-16 Eric Carlson <eric.carlson@apple.com>
Use AVRoutePickerView when available for choosing AirPlay devices
https://bugs.webkit.org/show_bug.cgi?id=213497
<rdar://problem/58610662>
Reviewed by Jer Noble.
* UIProcess/WebPageProxy.h:
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::platformView const): Add new override.
2020-07-16 Jer Noble <jer.noble@apple.com>
[Cocoa] Add MediaCapabilities support for SW VP9 decoder.
https://bugs.webkit.org/show_bug.cgi?id=214316
Reviewed by Eric Carlson.
Pass the system AC state to the WebContent process at creation time, and also create
a power state listener so the WebProcess can be updated dynamically.
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
(WebKit::WebProcessPool::registerNotificationObservers):
* UIProcess/WebProcessPool.cpp:
* UIProcess/WebProcessPool.h:
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
(WebKit::WebProcess::powerSourceDidChange):
2020-07-16 Per Arne Vollan <pvollan@apple.com>
[macOS] Add syscall to WebContent sandbox
https://bugs.webkit.org/show_bug.cgi?id=214394
<rdar://problem/65638758>
Reviewed by Darin Adler.
Add required syscall to WebContent process on macOS.
* WebProcess/com.apple.WebProcess.sb.in:
2020-07-16 Chris Dumez <cdumez@apple.com>
Unreviewed, fix logging added in r264414 that was incorrectly marked as "error" logging.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::receivedNavigationPolicyDecision):
(WebKit::WebPageProxy::decidePolicyForNavigationAction):
2020-07-16 Tuomas Karkkainen <tuomas.webkit@apple.com>
ASAN builds of WebKit should not freeze when an EXC_RESOURCE is delivered
https://bugs.webkit.org/show_bug.cgi?id=214346
Reviewed by Mark Lam.
When the WebContent process allocates 1000 MB of memory or starts 32
threads, it receives an EXC_RESOURCE. This requires iterating all the
mappings of the process. For the case of a process instrumented with
AddressSanitizer, the mappings total 20 terabytes, and cause the
process to "freeze" for several minutes.
Setting the sysctl debug.toggle_address_reuse to 1 causes the
EXC_RESOURCE mechanism to skip iteration of all the mappings.
With this patch, setting the environment variable
DISABLE_FREEZING_ON_EXC_RESOURCE will set the sysctl mentioned above
for WebContent and Networking processes.
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceMain.mm:
2020-07-15 Per Arne Vollan <pvollan@apple.com>
Unreviewed, revert part of r264385.
The change set r264385 speculatively added two messages to the mach-kernel-endpoint filter on iOS.
These messages are not needed, and should be removed.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-07-15 Darin Adler <darin@apple.com>
Remove a few more uses of the terms black/white list
https://bugs.webkit.org/show_bug.cgi?id=214371
Reviewed by Alex Christensen.
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::shouldPlugInAutoStartFromOrigin): Reworded a comment.
2020-07-15 Devin Rousso <drousso@apple.com>
should represent `TextPlaceholderElement` as an `NSTextAttachmentCharacter` instead of a `\n`
https://bugs.webkit.org/show_bug.cgi?id=214331
<rdar://problem/64779558>
Reviewed by Wenson Hsieh.
* Platform/spi/ios/UIKitSPI.h:
Add missing SPI declarations.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView removeTextPlaceholder:willInsertText:completionHandler:]):
Drive-by: ensure that the `completionHandler` is always called.
2020-07-15 Simon Fraser <simon.fraser@apple.com>
itsnicethat.com page is sometimes non-scrollable
https://bugs.webkit.org/show_bug.cgi?id=214362
<rdar://problem/65431127>
Reviewed by Tim Horton.
This page uses 'clip-path: inset(1px); on a 1px x 1px element to hit descendants from
hit-testing and compositing code turns this into a layer mask using a CAShapeLayer.
Our layer/view hit-testing code failed to check for masking; if a layer has a mask,
it needs to be handled like -masksToBounds, in that points outside the mask should
never hit descendant layers.
Make a -_web_maskContainsPoint: to test whether the point is inside the mask layer,
and _web_maskMayIntersectRect: which does approximate testing using path bounds
intersection (this is used for editable regions).
Also convert some hard-coded winding rule strings to use the CA constants.
* Shared/RemoteLayerTree/RemoteLayerTreePropertyApplier.mm:
(WebKit::RemoteLayerTreePropertyApplier::applyPropertiesToLayer):
* UIProcess/RemoteLayerTree/ios/RemoteLayerTreeViews.mm:
(WebKit::collectDescendantViewsAtPoint):
(WebKit::collectDescendantViewsInRect):
2020-07-15 Chris Dumez <cdumez@apple.com>
Improve navigation policy decision release logging to help better diagnose issues
https://bugs.webkit.org/show_bug.cgi?id=214368
Reviewed by Alex Christensen.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::decidePolicyForNavigationAction):
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::dispatchDecidePolicyForNavigationAction):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::didReceivePolicyDecision):
2020-07-15 Chris Dumez <cdumez@apple.com>
[WK2] Make establishing a connection between the WebProcess and the Network process more robust
https://bugs.webkit.org/show_bug.cgi?id=214307
<rdar://problem/64592340>
Reviewed by Alex Christensen.
Make establishing a connection between the WebProcess and the Network process more robust.
When the network process is crashy, it is not uncommon for the WebProcess to fail to establish
a connection to the network process. Previously, we would try twice and then give up by calling
CRASH(). I have tweaked the logic so that we now try 10 times and wait 100 ms between each
attempt to make a best effort at establishing the connection. I also updated the logic to exit
cleanly instead of crashing if the IPC connection to the UIProcess becomes invalid (i.e. the
UIProcess has severed the connection because this WebProcess is no longer needed).
* WebProcess/WebProcess.cpp:
(WebKit::getNetworkProcessConnection):
(WebKit::WebProcess::ensureNetworkProcessConnection):
2020-07-15 Rob Buis <rbuis@igalia.com>
Building AccessibilitySupportSPI with the macOS Big Sur public SDK fails
https://bugs.webkit.org/show_bug.cgi?id=214237
Reviewed by Alex Christensen.
Need to include HIServicesSPI.h for _AXUIElementRequestServicedBySecondaryAXThread.
* WebProcess/WebPage/mac/WKAccessibilityWebPageObjectMac.mm:
2020-07-15 Brady Eidson <beidson@apple.com>
Update a WKPreferences API comment.
https://bugs.webkit.org/show_bug.cgi?id=214360
Rubberstamped by Tim Horton
* UIProcess/API/Cocoa/WKPreferences.h:
2020-07-15 Lauro Moura <lmoura@igalia.com>
[SOUP] Artificial delay to WebSocket connection to mitigate port scanning attacks
https://bugs.webkit.org/show_bug.cgi?id=214293
Reviewed by Carlos Garcia Campos.
r264306 added an artificial delay when NetworkSocketStream closed due
to closed ports but Soup-based ports use another code path and were
still returning immediately.
SOUP WebSocket errors do not distinguish closed ports separately, so
this commit checks for connections that finished with NOT_WEBSOCKET
errors.
Covered by existing tests.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::randomClosedPortDelay): Move the delay
calculation here to be shared between the different code paths.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkSocketStream.cpp:
(WebKit::NetworkSocketStream::didFailSocketStream): Use the shared
delay.
(WebKit::randomDelay): Deleted.
* NetworkProcess/soup/WebSocketTaskSoup.cpp:
(WebKit::WebSocketTask::WebSocketTask): Add a oneShot timer with
random duration when failing to connect to something that is not a
WebSocket.
(WebKit::WebSocketTask::delayFailTimerFired):
2020-07-14 Per Arne Vollan <pvollan@apple.com>
[iOS] Add missing messages to message filter in WebContent sandbox
https://bugs.webkit.org/show_bug.cgi?id=214334
<rdar://problem/65529952>
Reviewed by Brent Fulgham.
Allow some messages which are required in the message filter in the WebContent process on iOS.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-07-14 Wenson Hsieh <wenson_hsieh@apple.com>
REGRESSION (r264101): Sharing a link attaches an image instead of the URL
https://bugs.webkit.org/show_bug.cgi?id=214329
<rdar://problem/65513607>
Reviewed by Tim Horton and Devin Rousso.
r264101 added logic that attempts to share an activated (i.e. long-pressed) element as an image, if the image
URL scheme for the element is equal to "data" (ignoring case sensitivity). However, in the case where the image
URL is nil, the call to `[element.imageURL.scheme caseInsensitiveCompare:@"data"]` is 0, which (conveniently) is
equal to `NSOrderedSame`. This causes us to incorrectly proceed by sharing the element as an image.
Fix this by going through URL::protocolIsData(), so we don't end up with 0 (`NSOrderedSame`) when
`element.imageURL.scheme` is `nil`.
Test: ShareSheetTests.ShareAnchorElementAsURL
* UIProcess/ios/WKActionSheetAssistant.mm:
(-[WKActionSheetAssistant handleElementActionWithType:element:needsInteraction:]):
2020-07-14 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] The completion handler in -handleKeyWebEvent:withCompletionHandler: is sometimes never called
https://bugs.webkit.org/show_bug.cgi?id=214295
<rdar://problem/60539389>
Reviewed by Devin Rousso.
This is a speculative fix for <rdar://problem/60539389>, wherein hardware key commands seemingly stop working in
a web page that is (presumably) otherwise responsive. It's possible that the bug exercises a scenario in which
the completion handler in `-[WKContentView handleKeyWebEvent:withCompletionHandler:]` is never invoked, which
subsequently leads to the keyboard task queue being backed up with key events.
This can happen in several ways. For instance, if the web process is swapped or terminates in the middle of
handling a key event, the key event queue will be cleared, but the UI process will still retain the (uncalled)
completion handler for that key event. Additionally, `WebPageProxy::handleKeyboardEvent` may not even have
attempted to propagate the event to the web process, in which case we shouldn't be saving the completion handler
and waiting for a response.
Test: KeyboardInputTests.HandleKeyEventsInCrashedOrUninitializedWebProcess
KeyboardInputTests.HandleKeyEventsWhileSwappingWebProcess
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::handleKeyboardEvent):
Make this return a `bool` indicating whether the key event was sent to the web process. If not, then we
should immediately invoke the completion handler in -handleKeyWebEvent:withCompletionHandler: below, instead of
stashing the Objective-C block and waiting for a response from the web process (which is presumably not
running).
* UIProcess/WebPageProxy.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView cleanUpInteraction]):
(-[WKContentView _cancelPendingKeyEventHandler]):
When the web process terminates or swaps in the middle of handling a key event, go ahead and invoke the key
event completion handler early with the queued event, since we aren't going to receive a response from the web
process anyways.
(-[WKContentView handleKeyWebEvent:withCompletionHandler:]):
2020-07-14 Simon Fraser <simon.fraser@apple.com>
Flashes of incorrect scroll position when zooming on quip
https://bugs.webkit.org/show_bug.cgi?id=214273
<rdar://problem/59385421>
Reviewed by Tim Horton.
Quip uses a non-scrollable "body { overflow:hidden }")" page, but aggressively scrolls to 0,0 if it
gets a scroll event and top/left are non-zero. Panning is necessarily allowed on overflow:hidden pages,
so to avoid programmatic scrolls yanking the content to 0,0 during interaction, ignore them if
the page is non-scrollable yet zoomed.
When refusing such a scroll, we have to inform the web process to avoid mismatched ui/web process state,
and work around the fact that WebPageProxy::updateVisibleContentRects() would bail if the update
appeared to be the same as the last one.
Test: fast/scrolling/ios/programmatic-scroll-while-zoomed.html
* UIProcess/API/Cocoa/WKWebViewInternal.h:
* UIProcess/API/ios/WKWebViewIOS.h:
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _updateScrollViewForTransaction:]):
(-[WKWebView _scrollToContentScrollPosition:scrollOrigin:]):
(-[WKWebView _scheduleForcedVisibleContentRectUpdate]):
(-[WKWebView _updateVisibleContentRects]):
* UIProcess/RemoteLayerTree/RemoteScrollingCoordinatorProxy.cpp:
(WebKit::RemoteScrollingCoordinatorProxy::hasScrollableMainFrame const):
(WebKit::RemoteScrollingCoordinatorProxy::hasScrollableOrZoomedMainFrame const):
* UIProcess/RemoteLayerTree/RemoteScrollingCoordinatorProxy.h:
* UIProcess/WebPageProxy.h:
* UIProcess/ios/WKContentView.h:
* UIProcess/ios/WKContentView.mm:
(-[WKContentView didUpdateVisibleRect:unobscuredRect:contentInsets:unobscuredRectInScrollViewCoordinates:obscuredInsets:unobscuredSafeAreaInsets:inputViewBounds:scale:minimumScale:inStableState:isChangingObscuredInsetsInteractively:enclosedInScrollableAncestorView:sendEvenIfUnchanged:]):
(-[WKContentView didUpdateVisibleRect:unobscuredRect:contentInsets:unobscuredRectInScrollViewCoordinates:obscuredInsets:unobscuredSafeAreaInsets:inputViewBounds:scale:minimumScale:inStableState:isChangingObscuredInsetsInteractively:enclosedInScrollableAncestorView:]): Deleted.
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::updateVisibleContentRects):
2020-07-14 Kate Cheney <katherine_cheney@apple.com>
[ macOS iOS ] http/tests/resourceLoadStatistics/telemetry-generation-basic-functionality-database.html is a rare flaky failure
https://bugs.webkit.org/show_bug.cgi?id=212165
<rdar://problem/63462045>
Reviewed by Brent Fulgham.
Test was flaking because of multiple async calls to submit telemetry,
which happen when a new ITP store is created and again when the test
makes an SPI call. If the first telemetry call is reported after the test
sets notifyPagesWhenTelemetryWasCaptured but before the SPI call to capture
telemetry is submitted, the test will receive the wrong data.
To fix this, we should simplify by removing the separate function to
tell ITP to report telemetry, and instead always report when the testing
SPI call is made.
No new tests, will fix flaky tests.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::calculateAndSubmitTelemetry const):
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:
(WebKit::ResourceLoadStatisticsMemoryStore::calculateAndSubmitTelemetry const):
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.h:
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::submitTelemetry):
Pass in new enum value to indicate when the telemetry call is coming
from the testing SPI.
(WebKit::WebResourceLoadStatisticsStore::setNotifyPagesWhenTelemetryWasCaptured): Deleted.
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.h:
* NetworkProcess/Classifier/WebResourceLoadStatisticsTelemetry.cpp:
(WebKit::WebResourceLoadStatisticsTelemetry::calculateAndSubmit):
(WebKit::WebResourceLoadStatisticsTelemetry::submitTelemetry):
(WebKit::WebResourceLoadStatisticsTelemetry::setNotifyPagesWhenTelemetryWasCaptured): Deleted.
* NetworkProcess/Classifier/WebResourceLoadStatisticsTelemetry.h:
Remove static bool variable notifyPagesWhenTelemetryWasCaptured and
rely on the submitTelemetry function call to let us know when to report telemetry.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::setNotifyPagesWhenTelemetryWasCaptured): Deleted.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreSetStatisticsNotifyPagesWhenTelemetryWasCaptured): Deleted.
* UIProcess/API/C/WKWebsiteDataStoreRef.h:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::setNotifyPagesWhenTelemetryWasCaptured): Deleted.
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::setNotifyPagesWhenTelemetryWasCaptured): Deleted.
* UIProcess/WebsiteData/WebsiteDataStore.h:
Remove setNotifyPagesWhenTelemetryWasCaptured function.
2020-07-14 Daniel Bates <dabates@apple.com>
[iOS] Sometimes unable to type Japanese on docs.google.com
https://bugs.webkit.org/show_bug.cgi?id=214272
<rdar://problem/59449972>
Reviewed by Wenson Hsieh.
Tell the web process to restore the selection of the focused element when re-focusing it
on page activation.
When resigning first responder the text interaction assistant is deactivated and calls
back into WKContentView's -clearSelection, which clears the focused element and tells
the web process to clear the selection. Keep in mind that the focused element in the web
process is not changed. When the WKContentView subsequently becomes first responder the
web process messages the UI process to tell it to start assistance (aka focus) for focused
element. However the selection is not restored and thus the editor state the UI process
has, and is even sent up until there is a selection change, mismatches with the focused
element information it also has. To match Mac, key events are only passed to the input
manager if an editable element is focused (see <https://bugs.webkit.org/show_bug.cgi?id=199122>).
So, key events are not passed to the input manager when in this state. To prevent this,
restore the selection of the focused element when the element is re-focused because the
page became active again.
Selection restoration is driven by the UI process (not the web process) because element re-
focusing may be denied by the embedding client. In this case, selection should not be restored
because UI assistance was not started. For the same reason, I do not restore it in -becomeFirstResponderForWebView.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView cleanUpInteraction]):
(-[WKContentView _didCommitLoadForMainFrame]):
Clear out the flag. See below for more details.
(-[WKContentView handleKeyWebEvent:withCompletionHandler:]): Pass the key to the input manager
if the _treatAsContentEditableUntilNextEditorStateUpdate is YES.
(-[WKContentView _elementDidFocus:userIsInteracting:blurPreviousNode:activityStateChanges:userObject:]):
If this request is a result of the page activating (aka focus activity change) into an editable field
(i.e. mayContainSelectableText() resturns YES) and assistance is not being suppressed then tell the web
process to restore the selection. As part of this I tell the text interaction assistant to activate even
though -becomeFirstResponderForWebView was called beforehand because -becomeFirstResponderForWebView only
activates it if there is a non-empty selection. (Note the page having an empty selection could have been
a result of -clearSelection having been called though the page could have done it programmatically as well).
I also set a new flag to YES and schedule a full editor state update to ensure this flag is cleared.
The flag is so that key events that come in between now and the next editor state update are passed to the
input manager.
(-[WKContentView _didUpdateEditorState]): Clear out the flag. See above for more details.
2020-07-14 Fujii Hironori <Hironori.Fujii@sony.com>
Unreviewed non-unified build fixes
* NetworkProcess/NetworkSocketStream.cpp:
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreGetAllStorageAccessEntries):
* UIProcess/WebContextMenuProxy.cpp:
* UIProcess/WebOpenPanelResultListenerProxy.h:
2020-07-13 Chris Dumez <cdumez@apple.com>
Add more logging for navigation policy decisions on UIProcess side
https://bugs.webkit.org/show_bug.cgi?id=214277
Reviewed by Geoffrey Garen.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::receivedNavigationPolicyDecision):
(WebKit::WebPageProxy::decidePolicyForNavigationAction):
2020-07-13 Ryan Haddad <ryanhaddad@apple.com>
Remove live ranges from Document.h, AlternativeTextController.h, DictionaryLookup.h, and WebPage.h
https://bugs.webkit.org/show_bug.cgi?id=214109
Unreviewed Catalyst build fix.
* WebProcess/WebPage/Cocoa/WebPageCocoa.mm:
(WebKit::WebPage::dictionaryPopupInfoForRange):
2020-07-13 Kate Cheney <katherine_cheney@apple.com>
Followup to bug 213143: Add artificial delay to WebSocket connections to mitigate port scanning attacks
https://bugs.webkit.org/show_bug.cgi?id=213143
Reviewed by Darin Adler.
Convert min and max millisecond values from double/int to Seconds.
No new tests, behavior should not change.
* NetworkProcess/NetworkSocketStream.cpp:
(WebKit::randomDelay):
2020-07-13 Eric Carlson <eric.carlson@apple.com>
Impossible to pause playback of MediaStream video track
https://bugs.webkit.org/show_bug.cgi?id=214262
<rdar://problem/65483729>
Reviewed by Jer Noble.
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayer.cpp:
(WebKit::RemoteSampleBufferDisplayLayer::play):
(WebKit::RemoteSampleBufferDisplayLayer::pause):
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayer.h:
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayer.messages.in:
* WebProcess/GPU/webrtc/SampleBufferDisplayLayer.cpp:
(WebKit::SampleBufferDisplayLayer::play):
(WebKit::SampleBufferDisplayLayer::pause):
* WebProcess/GPU/webrtc/SampleBufferDisplayLayer.h:
2020-07-13 Sam Weinig <weinig@apple.com>
Replace single argument makeSimpleColor uses with their implementation
https://bugs.webkit.org/show_bug.cgi?id=214240
Reviewed by Darin Adler.
* UIProcess/API/wpe/WebKitColor.cpp:
(webkitColorToWebCoreColor):
Replace makeSimpleColor with convertToComponentBytes.
2020-07-13 Per Arne Vollan <pvollan@apple.com>
Crash under WebKit::LaunchServicesDatabaseManager::didConnect()
https://bugs.webkit.org/show_bug.cgi?id=214263
Reviewed by Darin Adler.
There is a crash under WebKit::LaunchServicesDatabaseManager::didConnect() when trying to send a message to an invalid connection.
No new tests, since I have not been able to reproduce.
* WebProcess/cocoa/LaunchServicesDatabaseManager.mm:
(WebKit::LaunchServicesDatabaseManager::didConnect):
2020-07-13 Brent Fulgham <bfulgham@apple.com>
REGRESSION (r262538): Nullptr Deref in WTF::Detail::CallableWrapper<WebKit::WebResourceLoadStatisticsStore::performDailyTasks()::$_66, void>::call
https://bugs.webkit.org/show_bug.cgi?id=214264
<rdar://problem/65093949>
Reviewed by Darin Adler.
A new type check was added outside of the protecting nullptr check. This introduced
a new nullptr dereference bug.
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::performDailyTasks):
2020-07-13 Kate Cheney <katherine_cheney@apple.com>
Add artificial delay to WebSocket connections to mitigate port scanning attacks
https://bugs.webkit.org/show_bug.cgi?id=213143
<rdar://problem/64308927>
Reviewed by Alex Christensen.
When establishing web socket connections, closed ports typically
return an error more quickly than open ports due to the additional
time it takes open ports to perform a TLS handshake. This patch
adds a delay to the closed-port case to prevent distinguishing these
cases.
Test: http/tests/websocket/tests/hybi/closed-port-delay.html
* NetworkProcess/NetworkSocketStream.cpp:
(WebKit::NetworkSocketStream::NetworkSocketStream):
(WebKit::randomDelay):
Add a random delay between 10 and 100ms before sending IPC.
(WebKit::NetworkSocketStream::sendDelayedFailMessage):
(WebKit::NetworkSocketStream::didFailSocketStream):
Only delay for the proper error code indicating the connection
was refused for a closed port.
* NetworkProcess/NetworkSocketStream.h:
2020-07-13 Zan Dobersek <zdobersek@igalia.com>
[GTK][WPE] WTR and API-tests timing out after r264283
https://bugs.webkit.org/show_bug.cgi?id=214244
Reviewed by Carlos Garcia Campos.
r264283 broke local file loads at least for GTK and WPE ports due to
additionally-introduced ENABLE(SANDBOX_EXTENSIONS) build guards.
This patch reverts to using the sandboxing code without these guards,
relying instead on no-op SandboxExtension implementations for ports not
yet providing the system-specific sandboxing functionality. Uses of the
audit token functionality (only available on Cocoa platforms) gain
HAVE(AUDIT_TOKEN) guards. Overall the changes keep the r264283-enabled
code running where viable while unbreaking local file loading for other
ports.
* NetworkProcess/NetworkResourceLoadParameters.cpp:
(WebKit::NetworkResourceLoadParameters::encode const):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::maybeInitializeSandboxExtensionHandle):
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::MediaPlayerPrivateRemote::load):
2020-07-12 Said Abou-Hallawa <sabouhallawa@apple.com>
[macOS]: A HEIF image, selected from the OpenPanel, should be converted to an accepted MIME type
https://bugs.webkit.org/show_bug.cgi?id=213347
<rdar://problem/57258464>
Reviewed by Darin Adler.
Transcode the images given the list of the selected file and the allowed
MIME types. Only selected files whose MIME types do not exist in the allowed
MIME types will be transcoded. Right now, this only will apply to the HEIF
images on macOS.
* Platform/ImageUtilities.h: Renamed from Source/WebKit/Platform/cg/CGUtilities.h.
* Platform/Logging.h:
* Platform/cg/CGUtilities.cpp: Removed.
* Platform/cg/ImageUtilitiesCG.cpp: Added.
(WebKit::transcodeImage):
(WebKit::findImagesForTranscoding):
(WebKit::transcodeImages):
Implement a function that transcodes images among a set of files given
the allowedMIMETypes. The destination MIME type of these images is the
first MIME type in the allowedMIMETypes which CG supports encoding to.
* Shared/cg/ShareableBitmapCG.cpp:
(WebKit::ShareableBitmap::paint):
Call drawNativeImage() after renaming paintImage() to drawNativeImage().
* SourcesCocoa.txt:
* UIProcess/API/C/WKOpenPanelParametersRef.cpp:
(WKOpenPanelParametersCopyAllowedMIMETypes):
* UIProcess/API/C/WKOpenPanelParametersRef.h:
* UIProcess/API/C/WKOpenPanelResultListener.cpp:
(WKOpenPanelResultListenerChooseFiles):
* UIProcess/API/C/WKOpenPanelResultListener.h:
Transfer the list of the allowed MIME types from the OpenPanel to the
file chooser listener.
* UIProcess/WebPageProxy.cpp:
(WebKit::m_transcodingQueue):
(WebKit::WebPageProxy::didChooseFilesForOpenPanelWithImageTranscoding):
Transcode the images in a work queue so the UI process is not blocked.
Once the transcoding finishes, the message will be sent to the WebPage.
(WebKit::WebPageProxy::didChooseFilesForOpenPanel):
* UIProcess/WebPageProxy.h:
* WebKit.xcodeproj/project.pbxproj:
2020-07-12 Darin Adler <darin@apple.com>
REGRESSION (r261152): 9 blocks of critical sandboxing code disabled because they are guarded with HAVE(SANDBOX_ISSUE_READ_EXTENSION_TO_PROCESS_BY_AUDIT_TOKEN)
https://bugs.webkit.org/show_bug.cgi?id=214233
Reviewed by Per Arne Vollan.
* NetworkProcess/NetworkResourceLoadParameters.cpp:
(WebKit::NetworkResourceLoadParameters::encode const): Remove #if HAVE.
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::createSandboxExtensionsIfNeeded): Ditto.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::maybeInitializeSandboxExtensionHandle): Ditto.
(WebKit::WebPageProxy::loadRequestWithNavigationShared): Ditto.
(WebKit::WebPageProxy::loadFile): Ditto.
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::shouldSendPendingMessage): Ditto.
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::MediaPlayerPrivateRemote::load): Ditto.
2020-07-11 Darin Adler <darin@apple.com>
Remove live ranges from AccessibilityObject.h, AccessibilityObjectInterface.h, AccessibilityRenderObject.h, AXIsolatedObject.h
https://bugs.webkit.org/show_bug.cgi?id=214215
Reviewed by Sam Weinig.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::requestDocumentEditingContext): Use nested if for start and end
so we don't do the work on end if the test fails on start.
2020-07-10 Brent Fulgham <bfulgham@apple.com>
[iOS] Prompt for requestStorageAccess is inverted in WKWebView
https://bugs.webkit.org/show_bug.cgi?id=214198
<rdar://problem/65369725>
Reviewed by Darin Adler.
The labels for "allow" and "deny" StorageAccess API were inverted in the WKWebView
implementation.
* UIProcess/Cocoa/WKStorageAccessAlert.mm:
(WebKit::presentStorageAccessAlert):
2020-07-10 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] Prevent SPI clients from making the minimum layout height or width negative
https://bugs.webkit.org/show_bug.cgi?id=214199
<rdar://problem/64109242>
Reviewed by Tim Horton.
It's an error in the embedding client to override the minimum layout size with a negative width or height (in
this case, Mail on iOS). This happens when double tapping on a contact pill in the "To:" line in Mail compose,
which causes the top content inset of the web view's scroll view to increase such that Mail's logic for
overriding layout width ends up passing a negative height.
After a recent change (<https://trac.webkit.org/r261874>), this now results in the web view immediately
scrolling to an incorrect content offset when rotating into landscape mode after double tapping a contact. Fix
this by clamping the overridden layout size to a minimum of (0, 0), to prevent SPI clients from triggering
subtle bugs like this in the future.
Test: WebKit.OverrideMinimumLayoutSizeWithNegativeHeight
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _overrideLayoutParametersWithMinimumLayoutSize:maximumUnobscuredSizeOverride:]):
2020-07-08 Darin Adler <darin@apple.com>
Remove live ranges from Document.h, AlternativeTextController.h, DictionaryLookup.h, and WebPage.h
https://bugs.webkit.org/show_bug.cgi?id=214109
Reviewed by Sam Weinig.
* WebProcess/InjectedBundle/API/mac/WKWebProcessPlugInBrowserContextController.mm:
(-[WKWebProcessPlugInBrowserContextController selectedRange]): Updated since
WebPage::currentSelectionAsRange returns SimpleRange.
* WebProcess/WebPage/Cocoa/WebPageCocoa.mm:
(WebKit::WebPage::performDictionaryLookupAtLocation): Updated since
DictionaryLookup::rangeAtHitTestResult returns SimpleRange.
(WebKit::WebPage::performDictionaryLookupForSelection): Updated since
DictionaryLookup::rangeForSelection returns SimpleRange.
(WebKit::WebPage::performDictionaryLookupForRange): Take SimpleRange.
(WebKit::WebPage::dictionaryPopupInfoForRange): Ditto.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::currentSelectionAsRange): Return SimpleRange.
* WebProcess/WebPage/WebPage.h: Use SimpleRange instead of live ranges.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::selectWithGesture): Updated since m_startingGestureRange
is SimpleRange.
(WebKit::WebPage::clearSelection): Ditto.
(WebKit::WebPage::rangeForGranularityAtPoint): Return SimpleRange.
(WebKit::WebPage::selectTextWithGranularityAtPoint): Update since
rangeForGranularityAtPoint returns SimpleRange.
(WebKit::WebPage::updateSelectionWithExtentPointAndBoundary): Use
SimpleRange.
(WebKit::WebPage::updateSelectionWithExtentPoint): Use auto.
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::performImmediateActionHitTestAtLocation): Update
since lookupTextAtLocation returns SimpleRange.
(WebKit::WebPage::lookupTextAtLocation): Return SimpleRange.
2020-07-10 Brady Eidson <beidson@apple.com>
Limit MultiGamepadProvider to Catalina and Big Sur.
https://bugs.webkit.org/show_bug.cgi?id=214190
Unreviewed build fix.
* UIProcess/Gamepad/cocoa/UIGamepadProviderCocoa.mm:
(WebKit::UIGamepadProvider::platformSetDefaultGamepadProvider):
2020-07-10 Brent Fulgham <bfulgham@apple.com>
[IPC hardening] Use MESSAGE_CHECK in WebPageProxy::setColorPickerColor()
https://bugs.webkit.org/show_bug.cgi?id=214168
<rdar://problem/58632672>
Reviewed by David Kilzer.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::setColorPickerColor): Replace ASSERT with
MESSAGE_CHECK.
2020-07-10 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] Flush the cairo surface after painting when using software
https://bugs.webkit.org/show_bug.cgi?id=214177
Reviewed by Adrian Perez de Castro.
This is similar to bug #210636 but it happens under wayland when using GTK4. It's good idea to flush always
after painting in any case, so we do ity in case of GTK3 as well.
* UIProcess/gtk/AcceleratedBackingStoreWayland.cpp:
(WebKit::AcceleratedBackingStoreWayland::snapshot): Call cairo_surface_flush() after drawing.
(WebKit::AcceleratedBackingStoreWayland::paint): Ditto.
2020-07-10 Carlos Garcia Campos <cgarcia@igalia.com>
Unreviewed. Fix GTK4 build
* UIProcess/API/gtk/WebKitWebViewGtk.cpp:
(fileChooserDialogResponseCallback):
(webkitWebViewRunFileChooser):
* UIProcess/gtk/AcceleratedBackingStoreWayland.cpp:
(WebKit::AcceleratedBackingStoreWayland::snapshot):
2020-07-09 Brady Eidson <beidson@apple.com>
Multiplex the HID and GameController gamepad providers on Mac.
<rdar://problem/63192532> and https://bugs.webkit.org/show_bug.cgi?id=214126
Reviewed by Tim Horton.
* UIProcess/Gamepad/cocoa/UIGamepadProviderCocoa.mm:
(WebKit::UIGamepadProvider::platformSetDefaultGamepadProvider):
2020-07-09 Tim Horton <timothy_horton@apple.com>
"ESPN Fantasy Sports" does not respond to mouse events, only touch events
https://bugs.webkit.org/show_bug.cgi?id=214165
<rdar://problem/64671543>
Reviewed by Wenson Hsieh.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView shouldUseMouseGestureRecognizer]):
Add a bundle check.
2020-07-09 Brent Fulgham <bfulgham@apple.com>
[macOS] Adopt the WebKit-specific PAC key and entitlements
https://bugs.webkit.org/show_bug.cgi?id=214162
<rdar://problem/65099687>
Reviewed by Geoffrey Garen.
Add the relevant entitlements for PAC use and exceptions on macOS. This will be
ignored on Intel, but will be a great thing on Apple Silicon!
* Scripts/process-entitlements.sh:
2020-07-09 Per Arne Vollan <pvollan@apple.com>
Move XPCEndpoint and XPCEndpointClient classes from WebCore to WebKit
https://bugs.webkit.org/show_bug.cgi?id=214099
Reviewed by Alex Christensen.
* NetworkProcess/cocoa/LaunchServicesDatabaseObserver.h:
* Shared/Cocoa/XPCEndpoint.h: Copied from Source/WebCore/platform/cocoa/XPCEndpoint.h.
* Shared/Cocoa/XPCEndpoint.mm: Copied from Source/WebCore/platform/cocoa/XPCEndpoint.mm.
(WebCore::XPCEndpoint::XPCEndpoint): Deleted.
(WebCore::XPCEndpoint::sendEndpointToConnection): Deleted.
(WebCore::XPCEndpoint::endpoint const): Deleted.
* Shared/Cocoa/XPCEndpointClient.h: Copied from Source/WebCore/platform/cocoa/XPCEndpointClient.h.
(WebCore::XPCEndpointClient::~XPCEndpointClient): Deleted.
* Shared/Cocoa/XPCEndpointClient.mm: Copied from Source/WebCore/platform/cocoa/XPCEndpointClient.mm.
(WebCore::XPCEndpointClient::setEndpoint): Deleted.
(WebCore::XPCEndpointClient::connection): Deleted.
* UIProcess/Network/NetworkProcessProxyCocoa.mm:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/cocoa/LaunchServicesDatabaseManager.h:
* WebProcess/cocoa/LaunchServicesDatabaseManager.mm:
* WebProcess/cocoa/WebProcessCocoa.mm:
2020-07-09 Zalan Bujtas <zalan@apple.com>
[iOS] Conversation view messages in Mail are too small and scrollable
https://bugs.webkit.org/show_bug.cgi?id=214157
<rdar://problem/64788069>
Reviewed by Wenson Hsieh.
Revert the drive-by contentSize set optimization introduced in r262721.
iOS Mail relies on this delayed/late setContentSize call on WKScrollView (see rdar://problem/65277759).
* UIProcess/ios/WKScrollView.mm:
(-[WKScrollView _setContentSizePreservingContentOffsetDuringRubberband:]):
2020-07-09 Lauro Moura <lmoura@igalia.com>
[GTK] Remaining build fix after r264179
Like last revision, avoid "missing binary operator before(" errors
when WPE libraries are not installed.
* UIProcess/gtk/AcceleratedBackingStoreWayland.cpp:
(WebKit::AcceleratedBackingStoreWayland::paint):
2020-07-09 Wenson Hsieh <wenson_hsieh@apple.com>
REGRESSION (r259840): Pressing Esc or ⌘+"." does not dismiss time picker on iOS
https://bugs.webkit.org/show_bug.cgi?id=214122
<rdar://problem/64940487>
Reviewed by Tim Horton.
Prior to r259840, when presenting context menus for certain input types (i.e. "time" and "datetime-local") on
iOS, UIKit would tell the presented view controller's view (a `UIDatePicker` in this case) to become first
responder. As a result, `WKContentView` would resign first responder, which (by default) blurs the focused
element and dismisses the context menu. The result is thats time pickers would immediately dismiss upon
presentation.
To mitigate this, r259840 adopted the existing active focus retaining mechanism to temporarily decouple
WKContentView's first responder status from the currently focused form element. However, this also causes
`-endEditingAndUpdateFocusAppearanceWithReason:` to bail, due to `self.webView._retainingActiveFocusedState`
returning `YES`. This means that codepaths meant to immediately dismiss UI for the focused element (for example,
when pressing Escape on a hardware keyboard on iOS) will not be able to dismiss UI.
To fix this, push the flag that keeps track of whether we're relinquishing first responder to the focused
element down into `WKContentView`, and have `WKContentView` automatically stop relinquishing first responder to
the focused element in `-[WKContentView accessoryDone]`.
Fixes an existing test: fast/forms/ios/dismiss-picker-using-keyboard.html
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView accessoryDone]):
Stop relinquishing focus to the focused element, so that form elements that require first responder can be
dismissed when triggering `-accessoryDone` (e.g. when pressing the Esc key on a hardware keyboard).
(-[WKContentView startRelinquishingFirstResponderToFocusedElement]):
(-[WKContentView stopRelinquishingFirstResponderToFocusedElement]):
Renamed from `-preserveFocus` and `-releaseFocus`, respectively. Also, made these methods idempotent using a new
`BOOL` flag, `_isRelinquishingFirstResponderToFocusedElement`.
(-[WKContentView preserveFocus]): Deleted.
(-[WKContentView releaseFocus]): Deleted.
* UIProcess/ios/forms/WKDateTimeInputControl.mm:
(-[WKDateTimePicker controlBeginEditing]):
(-[WKDateTimePicker controlEndEditing]):
Use the renamed WKContentView methods, and also remove the `_preservingFocus` flag.
* UIProcess/ios/forms/WKFormPopover.mm:
(-[WKRotatingPopover presentPopoverAnimated:]):
(-[WKRotatingPopover dismissPopoverAnimated:]):
Use the renamed WKContentView methods.
2020-07-09 Brent Fulgham <bfulgham@apple.com>
[iOS] Stop telemetry collection for the 'hw.cpu64bit_capable' sysctl
https://bugs.webkit.org/show_bug.cgi?id=214148
<rdar://problem/65222086>
Reviewed by Per Arne Vollan.
Telemetry shows that SQLite3 on iOS-based builds needs read access to the
'hw.cpu64bit_capable' sysctl. We can stop logging this.
* Resources/SandboxProfiles/ios/com.apple.WebKit.Networking.sb:
2020-07-09 Lauro Moura <lmoura@igalia.com>
[GTK] Buildfix when WPEBackend-fdo is not available
Using the && operator for WPE_FDO_CHECK_VERSION was giving the
"missing binary operator before(" error as the macro was not defined
when the backend is not installed, like in the LTS/Debian vanilla
bots.
Unreviewed build fix.
* UIProcess/gtk/AcceleratedBackingStoreWayland.cpp:
(WebKit::AcceleratedBackingStoreWayland::snapshot):
2020-07-09 Per Arne Vollan <pvollan@apple.com>
[Cocoa] Update Launch Services database in the WebContent process from the Network process
https://bugs.webkit.org/show_bug.cgi?id=213794
Reviewed by Brent Fulgham.
To be able to deny mach lookup access to the Launch Services database mapping service in the WebContent process, the database needs to be
sent from the Networking process to the WebContent process, represented by an xpc object. The Networking process is chosen instead of the UI
process, since WebKit is controlling the sandbox of the Networking process. There is already a direct connection between the Networking
process and the WebContent process, but this is a CoreIPC connection. Since the database is represented by an xpc object, an xpc connection
is requiered to transfer the database. This patch creates a direct xpc connection between the Networking process and the WebContent process,
by creating an xpc endpoint in the Networking process, sending this to the UI process over the bootstrap xpc connection, and then providing
each WebContent process with this endpoint by sending it over the bootstrap xpc connection between the UI process and the WebContent process.
This endpoint has to be sent via the xpc connections to the UI process, since the already existing direct connection between the Networking
process and the WebContent process is a lower level CoreIPC connection, which does not support transferring xpc objects like the xpc endpoint.
When the WebContent process receives the endpoint, it will create an xpc connection from the endpoint, which will create a direct connection
with the Networking process. A message is then sent over the connection to request the database xpc object. When the Networking process
receives this message, it checks the entitlement of the remote end to make sure it is communicating with a WebContent process, and then sends
the database if the entitlement check succeeds. When the WebContent process receives this message, it performs an entitlement check as well,
and if successful, the database object is consumed. Consuming this database object, will make sure UTI to MIME type mapping is working
correctly in the WebContent process.
For the UI process to be able to pick up xpc messages on the bootstrap connection, a new virtual method, handleXPCEvent, is added to the
ProcessLauncher Client class. This way, a ProcessLauncher client will be able to pick up xpc messages from the xpc bootstrap connection.
Previously the event handler for the bootstrap connection in the UI process would handle only xpc error messages. It has now been expanded
to handle normal messages, which are being forwarded to the ProcessLauncher client. In this case the client is the Network process proxy,
which will check the message received, and if it turns out to be a message containing an xpc endpoint, the message will be forwarded to
all Web processes. The ProcessLauncher client can be expanded to handle any xpc message, but right not the Network process proxy only
handles the xpc endpoint message. The handleXPCEvent virtual method return a RefPtr to an XPCEventHandler, which is captured in the event
handler of the bootstrap xpc connection.
No new tests, since there already exist tests for this. UTI to MIME type mapping is being used in WebKit and underlying framework, and many
tests would fail if this mapping was not working.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::initializeConnection):
* NetworkProcess/cocoa/LaunchServicesDatabaseObserver.h: Added.
* NetworkProcess/cocoa/LaunchServicesDatabaseObserver.mm: Added.
(WebKit::LaunchServicesDatabaseObserver::singleton):
(WebKit::LaunchServicesDatabaseObserver::LaunchServicesDatabaseObserver):
(WebKit::LaunchServicesDatabaseObserver::startObserving):
(WebKit::LaunchServicesDatabaseObserver::~LaunchServicesDatabaseObserver):
(WebKit::LaunchServicesDatabaseObserver::xpcEndpointMessageNameKey const):
(WebKit::LaunchServicesDatabaseObserver::xpcEndpointMessageName const):
(WebKit::LaunchServicesDatabaseObserver::xpcEndpointNameKey const):
(WebKit::LaunchServicesDatabaseObserver::handleEvent):
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* Scripts/process-entitlements.sh:
* Shared/Cocoa/LaunchServicesDatabaseXPCConstants.h: Added.
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceMain.mm:
* SourcesCocoa.txt:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::xpcEndpointMessage const):
(WebKit::WebProcessPool::sendNetworkProcessXPCEndpointToWebProcess):
* UIProcess/Cocoa/XPCEventHandler.h: Added.
(WebKit::XPCEventHandler::~XPCEventHandler):
* UIProcess/Launcher/ProcessLauncher.h:
(WebKit::ProcessLauncher::Client::xpcEventHandler const):
* UIProcess/Launcher/mac/ProcessLauncherMac.mm:
(WebKit::ProcessLauncher::launchProcess):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/Network/NetworkProcessProxyCocoa.mm: Added.
(WebKit::NetworkProcessProxy::xpcEventHandler const):
(WebKit::NetworkProcessProxy::XPCEventHandler::handleXPCEvent const):
(WebKit::NetworkProcessProxy::XPCEventHandler::XPCEventHandler):
(WebKit::NetworkProcessProxy::XPCEventHandler::~XPCEventHandler):
* UIProcess/WebProcessPool.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::didFinishLaunching):
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebPage/Cocoa/WebPageCocoa.mm:
(WebKit::WebPage::platformDidReceiveLoadParameters):
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::initializeConnection):
* WebProcess/WebProcess.h:
* WebProcess/cocoa/LaunchServicesDatabaseManager.h: Added.
* WebProcess/cocoa/LaunchServicesDatabaseManager.mm: Added.
(WebKit::LaunchServicesDatabaseManager::singleton):
(WebKit::LaunchServicesDatabaseManager::handleEvent):
(WebKit::LaunchServicesDatabaseManager::didConnect):
(WebKit::LaunchServicesDatabaseManager::waitForDatabaseUpdate):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::handleXPCEndpointMessages const):
(WebKit::WebProcess::platformInitializeWebProcess):
2020-07-09 Eric Carlson <eric.carlson@apple.com>
Reset AVSystemController_PIDToInheritApplicationStateFrom when mediaserverd dies
https://bugs.webkit.org/show_bug.cgi?id=214112
<rdar://problem/65229214>
Reviewed by Jer Noble.
* GPUProcess/media/ios/RemoteMediaSessionHelperProxy.cpp:
(WebKit::RemoteMediaSessionHelperProxy::providePresentingApplicationPID): Don't bother
tracking whether or not providePresentingApplicationPID has been called, the session
helper already does that.
* GPUProcess/media/ios/RemoteMediaSessionHelperProxy.h:
2020-07-09 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] Occasional crash under -[UIView _setViewDelegate:] when presenting date and time pickers
https://bugs.webkit.org/show_bug.cgi?id=214120
<rdar://problem/65246918>
Reviewed by Darin Adler.
This crash happens when attempting to present a date picker (`<input type=date>`) immediately after dismissing
it. We encounter an Objective-C exception thrown by UIKit, due to the `WKDateTimeContextMenuViewController`'s
view (the `UIDatePicker`) being presented while it is still owned by the preview view controller. We often avoid
this crash because the `WKDateTimeContextMenuViewController` is usually only owned by `WKDateTimePicker`, so
when we set `_viewController` to a new instance of `WKDateTimeContextMenuViewController`, the old view
controller is destroyed, and thus no longer owns the `UIDatePicker` view.
However, it's possible for anything (e.g. animation blocks in UIKit) to cause the old view controller to live
past the creation of the new view controller. If this happens, when we go and call `-setView:` on the new view
controller with the date picker view, the date picker view may still be the view of the old controller, and we
end up crashing.
To fix this, explicitly unload the old view controller's view before attempting to create the new view
controller.
Test: fast/forms/ios/show-and-dismiss-date-input.html
* UIProcess/ios/forms/WKDateTimeInputControl.mm:
(-[WKDateTimePicker contextMenuInteraction:configurationForMenuAtLocation:]):
(-[WKDateTimePicker contextMenuInteraction:willDisplayMenuForConfiguration:animator:]):
(-[WKDateTimePicker contextMenuInteraction:willEndForConfiguration:animator:]):
Make it possible to test date and time picker presentation and dismissal by calling into the private testing-
only subclassing hooks on `WKWebView` when we finish presenting and dismissing the date picker context menu.
(-[WKDateTimePicker removeContextMenuInteraction]):
2020-07-09 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] Add support for shm buffers in AcceleratedBackingStoreWayland when using the WPE renderer
https://bugs.webkit.org/show_bug.cgi?id=214135
Reviewed by Žan Doberšek.
It's available since wpe-fdo 1.7.0. This makes accelerated compositing work in platforms where
EGL_WL_bind_wayland_display is not available. It can also be forced with the environment variable
LIBGL_ALWAYS_SOFTWARE.
* UIProcess/glib/WebProcessPoolGLib.cpp:
(WebKit::WebProcessPool::platformInitializeWebProcess): Use AcceleratedBackingStoreWayland::checkRequirements()
instead of calling wpe_fdo_initialize_for_egl_display() again.
* UIProcess/gtk/AcceleratedBackingStoreWayland.cpp:
(WebKit::tryInitializeEGL): Moved EGL specific initialization here.
(WebKit::tryInitializeSHM): Added SHM initialization.
(WebKit::AcceleratedBackingStoreWayland::checkRequirements): Try to initialize EGL first and then SHM if EGL failed.
(WebKit::AcceleratedBackingStoreWayland::AcceleratedBackingStoreWayland): Create a EGL exportable only when
using EGL implementation.
(WebKit::AcceleratedBackingStoreWayland::~AcceleratedBackingStoreWayland): Release EGL resources only when using
EGL implementation.
(WebKit::AcceleratedBackingStoreWayland::unrealize): Ditto.
(WebKit::AcceleratedBackingStoreWayland::tryEnsureGLContext): Return early when not using EGL implementation.
(WebKit::AcceleratedBackingStoreWayland::update): Add implementation for the SHM case.
(WebKit::AcceleratedBackingStoreWayland::displayImage): Renamed.
(WebKit::AcceleratedBackingStoreWayland::displayBuffer): SHM implementation, copy the shm buffer into a cairo
image surface.
(WebKit::AcceleratedBackingStoreWayland::tryEnsureTexture):
(WebKit::AcceleratedBackingStoreWayland::downloadTexture):
(WebKit::AcceleratedBackingStoreWayland::snapshot): Handle the SHM case.
(WebKit::AcceleratedBackingStoreWayland::paint): Ditto.
* UIProcess/gtk/AcceleratedBackingStoreWayland.h:
2020-07-09 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r264148.
https://bugs.webkit.org/show_bug.cgi?id=214136
Introduced layout test failures
Reverted changeset:
"[Cocoa] Update Launch Services database in the WebContent
process from the Network process"
https://bugs.webkit.org/show_bug.cgi?id=213794
https://trac.webkit.org/changeset/264148
2020-07-09 Carlos Garcia Campos <cgarcia@igalia.com>
[SOUP] Add support for HTTPCookieAcceptPolicy::OnlyFromMainDocumentDomain
https://bugs.webkit.org/show_bug.cgi?id=213954
Reviewed by Michael Catanzaro.
Use ExclusivelyFromMainDocumentDomain for WEBKIT_COOKIE_POLICY_ACCEPT_NO_THIRD_PARTY instead of
OnlyFromMainDocumentDomain.
* UIProcess/API/glib/WebKitCookieManager.cpp:
(toWebKitCookieAcceptPolicy):
(toHTTPCookieAcceptPolicy):
* UIProcess/WebProcessPool.h:
2020-07-08 Per Arne Vollan <pvollan@apple.com>
[Cocoa] Update Launch Services database in the WebContent process from the Network process
https://bugs.webkit.org/show_bug.cgi?id=213794
Reviewed by Brent Fulgham.
To be able to deny mach lookup access to the Launch Services database mapping service in the WebContent process, the database needs to be
sent from the Networking process to the WebContent process, represented by an xpc object. The Networking process is chosen instead of the UI
process, since WebKit is controlling the sandbox of the Networking process. There is already a direct connection between the Networking
process and the WebContent process, but this is a CoreIPC connection. Since the database is represented by an xpc object, an xpc connection
is requiered to transfer the database. This patch creates a direct xpc connection between the Networking process and the WebContent process,
by creating an xpc endpoint in the Networking process, sending this to the UI process over the bootstrap xpc connection, and then providing
each WebContent process with this endpoint by sending it over the bootstrap xpc connection between the UI process and the WebContent process.
This endpoint has to be sent via the xpc connections to the UI process, since the already existing direct connection between the Networking
process and the WebContent process is a lower level CoreIPC connection, which does not support transferring xpc objects like the xpc endpoint.
When the WebContent process receives the endpoint, it will create an xpc connection from the endpoint, which will create a direct connection
with the Networking process. A message is then sent over the connection to request the database xpc object. When the Networking process
receives this message, it checks the entitlement of the remote end to make sure it is communicating with a WebContent process, and then sends
the database if the entitlement check succeeds. When the WebContent process receives this message, it performs an entitlement check as well,
and if successful, the database object is consumed. Consuming this database object, will make sure UTI to MIME type mapping is working
correctly in the WebContent process.
For the UI process to be able to pick up xpc messages on the bootstrap connection, a new virtual method, handleXPCEvent, is added to the
ProcessLauncher Client class. This way, a ProcessLauncher client will be able to pick up xpc messages from the xpc bootstrap connection.
Previously the event handler for the bootstrap connection in the UI process would handle only xpc error messages. It has now been expanded
to handle normal messages, which are being forwarded to the ProcessLauncher client. In this case the client is the Network process proxy,
which will check the message received, and if it turns out to be a message containing an xpc endpoint, the message will be forwarded to
all Web processes. The ProcessLauncher client can be expanded to handle any xpc message, but right not the Network process proxy only
handles the xpc endpoint message. The handleXPCEvent virtual method return a RefPtr to an XPCEventHandler, which is captured in the event
handler of the bootstrap xpc connection.
No new tests, since there already exist tests for this. UTI to MIME type mapping is being used in WebKit and underlying framework, and many
tests would fail if this mapping was not working.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::initializeConnection):
* NetworkProcess/cocoa/LaunchServicesDatabaseObserver.h: Added.
* NetworkProcess/cocoa/LaunchServicesDatabaseObserver.mm: Added.
(WebKit::LaunchServicesDatabaseObserver::singleton):
(WebKit::LaunchServicesDatabaseObserver::LaunchServicesDatabaseObserver):
(WebKit::LaunchServicesDatabaseObserver::startObserving):
(WebKit::LaunchServicesDatabaseObserver::~LaunchServicesDatabaseObserver):
(WebKit::LaunchServicesDatabaseObserver::xpcEndpointMessageNameKey const):
(WebKit::LaunchServicesDatabaseObserver::xpcEndpointMessageName const):
(WebKit::LaunchServicesDatabaseObserver::xpcEndpointNameKey const):
(WebKit::LaunchServicesDatabaseObserver::handleEvent):
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* Scripts/process-entitlements.sh:
* Shared/Cocoa/LaunchServicesDatabaseXPCConstants.h: Added.
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceMain.mm:
* SourcesCocoa.txt:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::xpcEndpointMessage const):
(WebKit::WebProcessPool::sendNetworkProcessXPCEndpointToWebProcess):
* UIProcess/Cocoa/XPCEventHandler.h: Added.
(WebKit::XPCEventHandler::~XPCEventHandler):
* UIProcess/Launcher/ProcessLauncher.h:
(WebKit::ProcessLauncher::Client::xpcEventHandler const):
* UIProcess/Launcher/mac/ProcessLauncherMac.mm:
(WebKit::ProcessLauncher::launchProcess):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/Network/NetworkProcessProxyCocoa.mm: Added.
(WebKit::NetworkProcessProxy::xpcEventHandler const):
(WebKit::NetworkProcessProxy::XPCEventHandler::handleXPCEvent const):
(WebKit::NetworkProcessProxy::XPCEventHandler::XPCEventHandler):
(WebKit::NetworkProcessProxy::XPCEventHandler::~XPCEventHandler):
* UIProcess/WebProcessPool.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::didFinishLaunching):
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebPage/Cocoa/WebPageCocoa.mm:
(WebKit::WebPage::platformDidReceiveLoadParameters):
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::initializeConnection):
* WebProcess/WebProcess.h:
* WebProcess/cocoa/LaunchServicesDatabaseManager.h: Added.
* WebProcess/cocoa/LaunchServicesDatabaseManager.mm: Added.
(WebKit::LaunchServicesDatabaseManager::singleton):
(WebKit::LaunchServicesDatabaseManager::handleEvent):
(WebKit::LaunchServicesDatabaseManager::didConnect):
(WebKit::LaunchServicesDatabaseManager::waitForDatabaseUpdate):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::handleXPCEndpointMessages const):
(WebKit::WebProcess::platformInitializeWebProcess):
2020-07-08 Brady Eidson <beidson@apple.com>
Aggressively prime GameController.framework gamepad provider.
https://bugs.webkit.org/show_bug.cgi?id=214094
Reviewed by Darin Adler.
* WebProcess/Gamepad/WebGamepadProvider.cpp:
(WebKit::WebGamepadProvider::gamepadConnected):
(WebKit::WebGamepadProvider::gamepadDisconnected):
(WebKit::WebGamepadProvider::gamepadActivity):
2020-07-08 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r264132 and r264139.
https://bugs.webkit.org/show_bug.cgi?id=214108
Introduced assert failure.
Reverted changesets:
"[Cocoa] Update Launch Services database in the WebContent
process from the Network process"
https://bugs.webkit.org/show_bug.cgi?id=213794
https://trac.webkit.org/changeset/264132
"Unreviewed build fix attempt after r264132."
https://trac.webkit.org/changeset/264139
2020-07-08 Chris Dumez <cdumez@apple.com>
ASSERTION FAILURE: Completion handlers not invalidated when WebPage::~WebPage() invoked navigating to docs.google.com and signing in
https://bugs.webkit.org/show_bug.cgi?id=214098
<rdar://problem/64848288>
Reviewed by Geoffrey Garen.
Make sure the WebPage destructor calls its "markLayersAsVolatile" CompletionHandlers before destroying
them.
Change is covered by new API test.
* UIProcess/API/Cocoa/WKWebViewPrivateForTesting.h:
* UIProcess/API/Cocoa/WKWebViewTesting.mm:
(-[WKWebView _processWillSuspendForTesting:]):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::~WebPage):
2020-07-08 Per Arne Vollan <pvollan@apple.com>
[Cocoa] Update Launch Services database in the WebContent process from the Network process
https://bugs.webkit.org/show_bug.cgi?id=213794
Reviewed by Brent Fulgham.
To be able to deny mach lookup access to the Launch Services database mapping service in the WebContent process, the database needs to be
sent from the Networking process to the WebContent process, represented by an xpc object. The Networking process is chosen instead of the UI
process, since WebKit is controlling the sandbox of the Networking process. There is already a direct connection between the Networking
process and the WebContent process, but this is a CoreIPC connection. Since the database is represented by an xpc object, an xpc connection
is requiered to transfer the database. This patch creates a direct xpc connection between the Networking process and the WebContent process,
by creating an xpc endpoint in the Networking process, sending this to the UI process over the bootstrap xpc connection, and then providing
each WebContent process with this endpoint by sending it over the bootstrap xpc connection between the UI process and the WebContent process.
This endpoint has to be sent via the xpc connections to the UI process, since the already existing direct connection between the Networking
process and the WebContent process is a lower level CoreIPC connection, which does not support transferring xpc objects like the xpc endpoint.
When the WebContent process receives the endpoint, it will create an xpc connection from the endpoint, which will create a direct connection
with the Networking process. A message is then sent over the connection to request the database xpc object. When the Networking process
receives this message, it checks the entitlement of the remote end to make sure it is communicating with a WebContent process, and then sends
the database if the entitlement check succeeds. When the WebContent process receives this message, it performs an entitlement check as well,
and if successful, the database object is consumed. Consuming this database object, will make sure UTI to MIME type mapping is working
correctly in the WebContent process.
For the UI process to be able to pick up xpc messages on the bootstrap connection, a new virtual method, handleXPCEvent, is added to the
ProcessLauncher Client class. This way, a ProcessLauncher client will be able to pick up xpc messages from the xpc bootstrap connection.
Previously the event handler for the bootstrap connection in the UI process would handle only xpc error messages. It has now been expanded
to handle normal messages, which are being forwarded to the ProcessLauncher client. In this case the client is the Network process proxy,
which will check the message received, and if it turns out to be a message containing an xpc endpoint, the message will be forwarded to
all Web processes. The ProcessLauncher client can be expanded to handle any xpc message, but right not the Network process proxy only
handles the xpc endpoint message. The handleXPCEvent virtual method return a RefPtr to an XPCEventHandler, which is captured in the event
handler of the bootstrap xpc connection.
No new tests, since there already exist tests for this. UTI to MIME type mapping is being used in WebKit and underlying framework, and many
tests would fail if this mapping was not working.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::initializeConnection):
* NetworkProcess/cocoa/LaunchServicesDatabaseObserver.h: Added.
* NetworkProcess/cocoa/LaunchServicesDatabaseObserver.mm: Added.
(WebKit::LaunchServicesDatabaseObserver::singleton):
(WebKit::LaunchServicesDatabaseObserver::LaunchServicesDatabaseObserver):
(WebKit::LaunchServicesDatabaseObserver::startObserving):
(WebKit::LaunchServicesDatabaseObserver::~LaunchServicesDatabaseObserver):
(WebKit::LaunchServicesDatabaseObserver::xpcEndpointMessageNameKey const):
(WebKit::LaunchServicesDatabaseObserver::xpcEndpointMessageName const):
(WebKit::LaunchServicesDatabaseObserver::xpcEndpointNameKey const):
(WebKit::LaunchServicesDatabaseObserver::handleEvent):
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* Scripts/process-entitlements.sh:
* Shared/Cocoa/LaunchServicesDatabaseXPCConstants.h: Added.
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceMain.mm:
* SourcesCocoa.txt:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::xpcEndpointMessage const):
(WebKit::WebProcessPool::sendNetworkProcessXPCEndpointToWebProcess):
* UIProcess/Cocoa/XPCEventHandler.h: Added.
(WebKit::XPCEventHandler::~XPCEventHandler):
* UIProcess/Launcher/ProcessLauncher.h:
(WebKit::ProcessLauncher::Client::xpcEventHandler const):
* UIProcess/Launcher/mac/ProcessLauncherMac.mm:
(WebKit::ProcessLauncher::launchProcess):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/Network/NetworkProcessProxyCocoa.mm: Added.
(WebKit::NetworkProcessProxy::xpcEventHandler const):
(WebKit::NetworkProcessProxy::XPCEventHandler::handleXPCEvent const):
(WebKit::NetworkProcessProxy::XPCEventHandler::XPCEventHandler):
(WebKit::NetworkProcessProxy::XPCEventHandler::~XPCEventHandler):
* UIProcess/WebProcessPool.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::didFinishLaunching):
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebPage/Cocoa/WebPageCocoa.mm:
(WebKit::WebPage::platformDidReceiveLoadParameters):
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::initializeConnection):
* WebProcess/WebProcess.h:
* WebProcess/cocoa/LaunchServicesDatabaseManager.h: Added.
* WebProcess/cocoa/LaunchServicesDatabaseManager.mm: Added.
(WebKit::LaunchServicesDatabaseManager::singleton):
(WebKit::LaunchServicesDatabaseManager::handleEvent):
(WebKit::LaunchServicesDatabaseManager::didConnect):
(WebKit::LaunchServicesDatabaseManager::waitForDatabaseUpdate):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::handleXPCEndpointMessages const):
(WebKit::WebProcess::platformInitializeWebProcess):
2020-07-08 Brent Fulgham <bfulgham@apple.com>
[macOS] Silence spam reports about deprecated logging system
https://bugs.webkit.org/show_bug.cgi?id=214090
<rdar://problem/63943836>
Reviewed by Per Arne Vollan.
Change the default 'block' behavior for 'com.apple.system.logger' to
block "with no-report" so that we don't spend time generating backtraces
when non-WebKit code attempts to use this deprecated logging system.
* WebProcess/com.apple.WebProcess.sb.in:
2020-07-08 Sihui Liu <sihui_liu@appe.com>
Network process crashes in WebKit::WebIDBServer::removeConnection
https://bugs.webkit.org/show_bug.cgi?id=214071
<rdar://problem/64853929>
Reviewed by Chris Dumez.
In network process, WebIDBServer may be removed and destroyed before web process using the same session
disconnects. Then a new WebIDBServer can be added before disconnection of web process and the new WebIDBServer
does not have information about the web process connection.
To solve this issue, now we only remove WebIDBServer when session is destroyed and there is no web process
connection of that session.
* NetworkProcess/IndexedDB/WebIDBServer.h:
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::createNetworkConnectionToWebProcess):
(WebKit::NetworkProcess::destroySession):
(WebKit::NetworkProcess::removeWebIDBServerIfPossible):
(WebKit::NetworkProcess::connectionToWebProcessClosed):
* NetworkProcess/NetworkProcess.h:
2020-07-08 Andy Estes <aestes@apple.com>
REGRESSION (r207558): [Mac] Context clicks on links sometimes result in navigations
https://bugs.webkit.org/show_bug.cgi?id=213793
<rdar://problem/64294015>
Reviewed by Tim Horton.
WebPageProxy::showContextMenu was written with the assumption that WebContextMenuProxy::show
would present a context menu and spin a nested run loop, blocking until the menu dismisses.
Once WebContextMenuProxy::show returns, showContextMenu would send the
WebPage::ContextMenuHidden message, telling the WebContent process that it's OK to resume
processing mouse events.
This assumption was invalidated in r207558 (and again in several subsequent changes), which
added asynchronous APIs to WKPageContextMenuClient. If the context menu client provides
context menu items asynchronously, WebContextMenuProxy::show would return -- and the
WebContent process would resume processing mouse events -- before the context menu was even
presented. Since context menu presentation is triggered by a mouse press event, it's
possible that the corresponding mouse release event would be delivered after
getContextMenuFromProposedMenuAsync was dispatched but before the client responded with menu
items to present. If this happened, since mouse event processing was not disabled, the
WebContent process would dispatch a DOM click event to the page, possibly triggering a
navigation.
Fix this by teaching WebPageProxy to defer sending WebPage::ContextMenuHidden until the
context menu has been dismissed, even in cases where the context menu client provides menu
items asynchronously. Also, take the opportunity to modernize WebContextMenuProxy and its
subclasses by, e.g., moving duplicated subclass code into the base class, using the client
pattern, and using WeakPtrs instead of raw pointers/references.
New API tests: ContextMenuMouseEvents.*
* UIProcess/API/APIContextMenuClient.h:
(API::ContextMenuClient::showContextMenu):
(API::ContextMenuClient::canShowContextMenu const):
* UIProcess/API/C/WKPage.cpp:
(WKPageSetPageContextMenuClient): Added canShowContextMenu to determine whether the client
can show a context menu without actually showing it.
* UIProcess/API/wpe/PageClientImpl.cpp:
(WebKit::PageClientImpl::createContextMenuProxy): Passed page to
WebContextMenuProxyWPE::create.
* UIProcess/WebContextMenuListenerProxy.cpp:
(WebKit::WebContextMenuListenerProxy::WebContextMenuListenerProxy):
(WebKit::WebContextMenuListenerProxy::useContextMenuItems):
(WebKit::WebContextMenuListenerProxy::invalidate): Deleted. Not needed now that m_client is
a WeakPtr.
* UIProcess/WebContextMenuListenerProxy.h:
(WebKit::WebContextMenuListenerProxy::create): Rather than store a raw pointer to a
WebContextMenuProxy, store a WeakPtr to a WebContextMenuListenerProxy::Client (which
WebContextMenuProxy implements).
* UIProcess/WebContextMenuProxy.cpp:
(WebKit::WebContextMenuProxy::WebContextMenuProxy): Stored a WeakPtr to a WebPageProxy
rather than storing a raw pointer to the same in each subclass.
(WebKit::WebContextMenuProxy::proposedItems const): Added a virtual function to create
proposed WebContextMenuItems from m_context's menu items. WebContextMenuProxyGtk overrides
this to do something slightly different.
(WebKit::WebContextMenuProxy::show): Added a base class implementation that creates a
WebContextMenuListenerProxy and calls ContextMenuClient::getContextMenuFromProposedMenu.
(WebKit::WebContextMenuProxy::useContextMenuItems): Added a base class implementation that
performs the work previously done in WebPageProxy::showContextMenu. Calls
showContextMenuWithItems, which now always synchronously presents the menu.
* UIProcess/WebContextMenuProxy.h: Inherited from WebContextMenuListenerProxy::Client.
(WebKit::WebContextMenuProxy::page const): Added for use by WKSharingServicePickerDelegate.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::showContextMenu): Moved code that assumes synchronous menu
presentation to WebContextMenuProxy::useContextMenuItems.
* UIProcess/gtk/WebContextMenuProxyGtk.cpp:
(WebKit::WebContextMenuProxyGtk::proposedItems const): Implemented GTK's special version
that excludes ContextMenuItemTagShareMenu.
(WebKit::WebContextMenuProxyGtk::show): Deleted. WebContextMenuProxy::show handles this now.
(WebKit::WebContextMenuProxyGtk::append):
(WebKit::WebContextMenuProxyGtk::showContextMenuWithItems):
(WebKit::WebContextMenuProxyGtk::WebContextMenuProxyGtk):
* UIProcess/gtk/WebContextMenuProxyGtk.h: Moved m_page to WebContextMenuProxy.
* UIProcess/mac/WKSharingServicePickerDelegate.mm:
(-[WKSharingServicePickerDelegate sharingService:didShareItems:]): Changed now that
WebContextMenuProxy::page returns a pointer.
* UIProcess/mac/WebContextMenuProxyMac.h:
(WebKit::WebContextMenuProxyMac::page const): Deleted.
* UIProcess/mac/WebContextMenuProxyMac.mm:
(WebKit::WebContextMenuProxyMac::WebContextMenuProxyMac):
(WebKit::WebContextMenuProxyMac::~WebContextMenuProxyMac): No need to invalidate
m_contextMenuListener now that it stores a WeakPtr to its WebContextMenuProxy.
(WebKit::WebContextMenuProxyMac::contextMenuItemSelected):
(WebKit::WebContextMenuProxyMac::showServicesMenu):
(WebKit::WebContextMenuProxyMac::show):
(WebKit::WebContextMenuProxyMac::showContextMenuWithItems): Overrode to handle service menus.
(WebKit::WebContextMenuProxyMac::useContextMenuItems): Overrode to get the standard Mac menu
items and dispatch ContextMenuClient::menuFromProposedMenu before showing the menu.
(WebKit::WebContextMenuProxyMac::showContextMenu): Deleted. WebContextMenuProxy::show now
handles this.
* UIProcess/win/WebContextMenuProxyWin.cpp:
(WebKit::WebContextMenuProxyWin::showContextMenuWithItems):
(WebKit::WebContextMenuProxyWin::WebContextMenuProxyWin):
(WebKit::WebContextMenuProxyWin::show): Deleted. WebContextMenuProxy::show now handles this.
* UIProcess/win/WebContextMenuProxyWin.h:
* UIProcess/wpe/WebContextMenuProxyWPE.h: Passed page to the WebContextMenuProxy ctor.
2020-07-08 Wenson Hsieh <wenson_hsieh@apple.com>
autocapitalize="words" capitalizes every word's second character
https://bugs.webkit.org/show_bug.cgi?id=148504
<rdar://problem/57814304>
Reviewed by Tim Horton.
This bug resurfaced in iOS 13 due to timing changes that caused a task added to UIKeyboardTaskQueue to no longer
get dequeued after the next editor state update is received in the UI process; as a result, UIKit asks us for
the context before the caret selection (i.e. `characterBeforeSelection`) before we have a chance to update it,
which means that UIKeyboardImpl's autoshift state will always be off by one character whie typing.
Note that solely deferring this update until after the next editor state update is received is insufficient to
fix this bug, since the keyboard's shift state will still be incorrect when typing characters very quickly (or
if the web process is unresponsive).
Instead, completely address this bug by adding a mechanism for WKContentView to cache the last character in the
string that is currently being inserted using `-insertText:`, and return it in `-_characterBeforeCaretSelection`
until the next editor state update comes in with the real character before the selection. This way, we're able
to respond to UIKit's requests immediately with a (reasonable) guess of what the character before the selection
should be, and then follow it up with the actual up-to-date value from the web process during the next editor
state update (invalidating the stale shift state if necessary). This cached last character should only be
incorrect in the case where the page uses script to influence what was typed, or the typed character was never
inserted at all (e.g. typing a newline character in a single line input element).
Test: fast/forms/ios/autocapitalize-words.html
* Platform/spi/ios/UIKitSPI.h:
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView cleanUpInteraction]):
(-[WKContentView _characterBeforeCaretSelection]):
Use `_lastInsertedCharacterToOverrideCharacterBeforeSelection`, if it is set.
(-[WKContentView _characterInRelationToCaretSelection:]):
(-[WKContentView insertText:]):
Update `_lastInsertedCharacterToOverrideCharacterBeforeSelection` and ensure an editor state update immediately
after text is inserted, if we're currently editing an element with `autocapitalize="words"`. The editor state
update could be unnecessarily expensive, so we don't want to do this unconditionally. Also,
`autocapitalize="words"` is special in the sense that it is the only autocapitalization type where the platform
(UIKit) will ask for the last character immediately and use it to unshift the keyboard. Even in the case of
`autocapitalize="sentences"`, UIKit makes the decision to unshift using information from `kbd` instead.
(-[WKContentView _elementDidBlur]):
(-[WKContentView _selectionChanged]):
Clear out `_lastInsertedCharacterToOverrideCharacterBeforeSelection` when the next editor state update is
received, and invalidate the current keyboard shift state by calling `-clearShiftState`. This SPI method
automatically schedules a shift state update, as well.
Note the extra `_usingGestureForSelection` check here, which prevents the keyboard from autoshifting and
unshifting while changing the selection via loupe gesture.
2020-07-07 Antoine Quint <graouts@webkit.org>
[iOS] Sharing an <img> element with a base64-encoded URL shares the URL as raw text instead of an image
https://bugs.webkit.org/show_bug.cgi?id=214042
<rdar://problem/56669102>
Reviewed by Wenson Hsieh.
In the case where the <img> element that was the target when the share sheet was brought has its data encoded
as a base64 URL, we need to share the UIImage that was created as its snapshot as a file rather than the URL
to avoid a base64 string to be shared instead of the image data.
To support testing of this, we add a new optional WKShareSheet delegate to indicate that the share sheet will
show and the provided activity items, which we forward to a new optional WKUIDelegatePrivate delegate method.
API tests can thus be written to check the data provided to the share sheet.
* UIProcess/API/Cocoa/WKUIDelegatePrivate.h:
* UIProcess/Cocoa/WKShareSheet.h:
* UIProcess/Cocoa/WKShareSheet.mm:
(-[WKShareSheet presentWithShareDataArray:inRect:]):
* UIProcess/ios/WKActionSheetAssistant.h:
* UIProcess/ios/WKActionSheetAssistant.mm:
(-[WKActionSheetAssistant handleElementActionWithType:element:needsInteraction:]):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView shareSheet:willShowActivityItems:]):
(-[WKContentView actionSheetAssistant:shareElementWithURL:rect:]):
(-[WKContentView actionSheetAssistant:shareElementWithImage:rect:]):
2020-07-08 Philippe Normand <pnormand@igalia.com>
[WPE][Qt] Deprecation warnings
https://bugs.webkit.org/show_bug.cgi?id=214074
Reviewed by Carlos Garcia Campos.
* UIProcess/API/wpe/qt/WPEQtView.cpp:
(WPEQtView::updatePaintNode): Switch to new createTextureFromNativeObject API when building against Qt 5.15.
* UIProcess/API/wpe/qt/WPEQtViewBackend.cpp:
(WPEQtViewBackend::dispatchWheelEvent): Switch to new WPE axis 2D event API.
2020-07-08 Carlos Garcia Campos <cgarcia@igalia.com>
Unreviewed. Update OptionsGTK.cmake and NEWS for 2.29.3 release
* gtk/NEWS: Add release notes for 2.29.3.
2020-07-07 Alex Christensen <achristensen@webkit.org>
Modernize FindOptions IPC code
https://bugs.webkit.org/show_bug.cgi?id=214063
Reviewed by Chris Dumez.
Use callWithAsyncReply instead of GenericCallback.
Use OptionSet<FindOptions> instead of FindOptions or uint32_t.
No change in behavior.
* Shared/API/c/WKSharedAPICast.h:
(WebKit::toFindOptions):
* Shared/WebFindOptions.h:
(): Deleted.
* UIProcess/API/Cocoa/WKWebView.mm:
(toFindOptions):
(-[WKWebView findString:withConfiguration:completionHandler:]):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::findStringMatches):
(WebKit::WebPageProxy::findString):
(WebKit::WebPageProxy::countStringMatches):
(WebKit::WebPageProxy::findStringCallback): Deleted.
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::findString):
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebPage/FindController.cpp:
(WebKit::core):
(WebKit::FindController::countStringMatches):
(WebKit::FindController::updateFindUIAfterPageScroll):
(WebKit::FindController::findString):
(WebKit::FindController::findStringMatches):
* WebProcess/WebPage/FindController.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-07-07 Sam Weinig <weinig@apple.com>
Part 2 of SimpleColor and SRGBA<uint8_t> are essentially the same - let's converge them
https://bugs.webkit.org/show_bug.cgi?id=213981
Reviewed by Darin Adler.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::paintSnapshotAtSize):
* WebProcess/WebPage/ios/FindControllerIOS.mm:
Make calls to makeSimpleColor more consistent. Use decimal rather than hex, don't pass unncessary
255 as the 4th argument.
2020-07-07 Alex Christensen <achristensen@webkit.org>
Use sendWithAsyncReply instead of EditingRangeCallback
https://bugs.webkit.org/show_bug.cgi?id=214056
Reviewed by Chris Dumez.
No change in behavior.
* UIProcess/Cocoa/WKReloadFrameErrorRecoveryAttempter.mm:
(-[WKReloadFrameErrorRecoveryAttempter initWithWebView:frameHandle:urlString:]):
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::selectedRangeWithCompletionHandler):
(WebKit::WebViewImpl::markedRangeWithCompletionHandler):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::getMarkedRangeAsync):
(WebKit::WebPageProxy::getSelectedRangeAsync):
(WebKit::WebPageProxy::editingRangeCallback): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-07-07 Brent Fulgham <bfulgham@apple.com>
[macOS] Allow additional IOKit classes for Apple Silicon
https://bugs.webkit.org/show_bug.cgi?id=214021
Reviewed by Sam Weinig.
<rdar://problem/63696732>
Update two of the arm64-specific sandbox rules to include additional
IOKit classes (and one IOKit property).
* WebProcess/com.apple.WebProcess.sb.in:
2020-07-07 Sihui Liu <sihui_liu@appe.com>
Network process crashes in WebKit::StorageManagerSet::deleteSessionStorageForOrigins
https://bugs.webkit.org/show_bug.cgi?id=214050
Reviewed by Chris Dumez.
In NetworkProcess::deleteAndRestrictWebsiteDataForRegistrableDomains, deleteSessionStorageForOrigins is called
in a callback without checking if session still exists.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::deleteAndRestrictWebsiteDataForRegistrableDomains):
2020-07-07 Chris Dumez <cdumez@apple.com>
WebContent process sometimes kills itself because it is receiving too much IPC from the UIProcess
https://bugs.webkit.org/show_bug.cgi?id=214048
<rdar://problem/64844770>
Reviewed by Darin Adler.
There was logic in IPC::Connection::enqueueIncomingMessage() that would kill the XPC connection
if too many IPC messages had been queued without the main thread processing them. The intention
was to protect the UIProcess again a badly behavior WebProcess sending too much IPC and it would
in effect terminate that WebProcess. However, the logic would actually apply too ALL IPC
connections. In <rdar://problem/64844770>, we see that the WebProcess is stuck on a sync IPC and
thus not processing other incoming IPCs. The UIProcess is sending it a lot of IPC so the
WebProcess decides to kill its IPC connection, thus terminating itself, which does not make much
sense. To address the issue, we now enable to killing on too much IPC behavior only if IPC message
throttling is enabled for the IPC connection, which means it only impacts the UIProcess connections
to its WebProcesses.
* Platform/IPC/Connection.cpp:
(IPC::Connection::enableIncomingMessagesThrottling):
(IPC::Connection::enqueueIncomingMessage):
* Platform/IPC/Connection.h:
(IPC::Connection::isIncomingMessagesThrottlingEnabled const):
2020-07-07 Simon Fraser <simon.fraser@apple.com>
Unable to scroll elcomerico.pe page until the page finishes loading
https://bugs.webkit.org/show_bug.cgi?id=214027
<rdar://problem/64646259>
Reviewed by Antti Koivisto.
The same issue with non-invertible transforms existed on iOS. r248015 attempted to
fix some instances of this, but failed to take into account the fact that a non-invertible
transfrom should also prevent hit-testing of descendant views.
* UIProcess/RemoteLayerTree/ios/RemoteLayerTreeViews.mm:
(WebKit::collectDescendantViewsAtPoint):
2020-07-07 Alex Christensen <achristensen@webkit.org>
Add WebCoreNSURLSession SPI to send HTTP/2 ping
https://bugs.webkit.org/show_bug.cgi?id=214030
<rdar://problem/64495827>
Reviewed by Jer Noble.
The SPI to do this through CFNetwork requires setting a block on the task
between construction and resumption. To allow this, I made PreconnectTask
and NetworkLoad require a start() call, and I added start() calls everywhere
they were implied with construction before.
* NetworkProcess/Downloads/PendingDownload.cpp:
(WebKit::PendingDownload::PendingDownload):
* NetworkProcess/NetworkConnectionToWebProcess.cpp:
(WebKit::NetworkConnectionToWebProcess::sendH2Ping):
(WebKit::NetworkConnectionToWebProcess::preconnectTo):
* NetworkProcess/NetworkConnectionToWebProcess.h:
* NetworkProcess/NetworkConnectionToWebProcess.messages.in:
* NetworkProcess/NetworkDataTask.cpp:
(WebKit::NetworkDataTask::setH2PingCallback):
* NetworkProcess/NetworkDataTask.h:
* NetworkProcess/NetworkLoad.cpp:
(WebKit::NetworkLoad::NetworkLoad):
(WebKit::NetworkLoad::start):
(WebKit::NetworkLoad::setH2PingCallback):
(WebKit::NetworkLoad::initialize): Deleted.
* NetworkProcess/NetworkLoad.h:
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::preconnectTo):
* NetworkProcess/NetworkResourceLoader.cpp:
(WebKit::NetworkResourceLoader::startNetworkLoad):
* NetworkProcess/PreconnectTask.cpp:
(WebKit::PreconnectTask::setH2PingCallback):
(WebKit::PreconnectTask::start):
(WebKit::PreconnectTask::~PreconnectTask): Deleted.
* NetworkProcess/PreconnectTask.h:
* NetworkProcess/ServiceWorker/ServiceWorkerSoftUpdateLoader.cpp:
(WebKit::ServiceWorkerSoftUpdateLoader::loadFromNetwork):
* NetworkProcess/cache/NetworkCacheSpeculativeLoad.cpp:
(WebKit::NetworkCache::SpeculativeLoad::SpeculativeLoad):
* NetworkProcess/cache/NetworkCacheSpeculativeLoadManager.cpp:
(WebKit::NetworkCache::SpeculativeLoadManager::preconnectForSubresource):
* NetworkProcess/cocoa/NetworkDataTaskCocoa.h:
* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::NetworkDataTaskCocoa::setH2PingCallback):
* Shared/WebErrors.cpp:
(WebKit::internalError): Deleted.
* Shared/WebErrors.h:
* WebProcess/Network/WebLoaderStrategy.cpp:
(WebKit::WebLoaderStrategy::generateLoadIdentifier):
(WebKit::WebLoaderStrategy::startPingLoad):
(WebKit::WebLoaderStrategy::preconnectTo):
(WebKit::generateLoadIdentifier): Deleted.
* WebProcess/Network/WebLoaderStrategy.h:
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::sendH2Ping):
* WebProcess/WebCoreSupport/WebFrameLoaderClient.h:
2020-07-07 Brady Eidson <beidson@apple.com>
Forward declare WKContentWorld in WKUserScript.h
<rdar://problem/65164826> and https://bugs.webkit.org/show_bug.cgi?id=214046
Unreviewed.
* UIProcess/API/Cocoa/WKUserScript.h:
2020-07-07 Brady Eidson <beidson@apple.com>
Fix CrashTracer reported in PDFPlugin::ByteRangeRequest::maybeComplete.
<rdar://problem/64884982> and https://bugs.webkit.org/show_bug.cgi?id=214026
Reviewed by Tim Horton.
No new tests (CrashTracer with no reproduction)
Due to the racy-ness of how PDFKit calls us on the background thread vs. what might be
happening with the main thread, sometimes the main thread is asked to handle a data
request either:
- Before m_data is allocated
or
- After it is deallocated
In either case it's okay to just not handle it.
Despite understanding how to reproduce in theory, writing a test case has been elusive.
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::ByteRangeRequest::maybeComplete): Null check m_data before doing anything with it.
2020-07-07 Carlos Garcia Campos <cgarcia@igalia.com>
REGRESSION(r262680): [GTK] Crash in WebKit::DropTarget::didPerformAction
https://bugs.webkit.org/show_bug.cgi?id=213885
Reviewed by Michael Catanzaro.
Just compare if new operation is the same as the current one, the == operator of Optional already takes care of
values being nullopt.
* UIProcess/API/gtk/DropTargetGtk3.cpp:
(WebKit::DropTarget::didPerformAction):
* UIProcess/API/gtk/DropTargetGtk4.cpp:
(WebKit::DropTarget::didPerformAction):
2020-07-07 Tomoki Imai <Tomoki.Imai@sony.com>
[Win] Implement Pasteboard::writeCustomData for Web Inspector Console tab
https://bugs.webkit.org/show_bug.cgi?id=213986
Reviewed by Fujii Hironori.
* Shared/WebPreferencesDefaultValues.h: Turn DEFAULT_CUSTOM_PASTEBOARD_DATA_ENABLED on for Windows
2020-07-06 Simon Fraser <simon.fraser@apple.com>
High CPU usage on Stash search results pages
https://bugs.webkit.org/show_bug.cgi?id=214018
<rdar://problem/64832917>
Reviewed by Tim Horton.
Stash search results pages can contain a lot of overflow:scroll areas, so scrolling thread CA commits
would take a long time because ScrollingTreeScrollingNodeDelegateMac::updateScrollbarPainters() would
open and close a CA commit for each scroller.
Fix by not explicitly starting a CA commit, and only entering this code if a gesture is active and in
the momentum phase.
Also sprinkle BEGIN_BLOCK_OBJC_EXCEPTIONS/END_BLOCK_OBJC_EXCEPTIONS in more places that call into Core
Animation.
* UIProcess/RemoteLayerTree/ios/ScrollingTreeFrameScrollingNodeRemoteIOS.mm:
(WebKit::ScrollingTreeFrameScrollingNodeRemoteIOS::repositionRelatedLayers):
2020-07-06 Simon Fraser <simon.fraser@apple.com>
BEGIN_BLOCK_OBJC_EXCEPTIONS/END_BLOCK_OBJC_EXCEPTIONS should not have trailing semicolons
https://bugs.webkit.org/show_bug.cgi?id=214019
Reviewed by Tim Horton.
These macros are defined as:
#define BEGIN_BLOCK_OBJC_EXCEPTIONS @try {
#define END_BLOCK_OBJC_EXCEPTIONS } @catch(NSException *localException) { ReportBlockedObjCException(localException); }
so they should not be used with trailing semicolons.
* Shared/Cocoa/ArgumentCodersCocoa.mm:
(IPC::decodeFontInternal):
* Shared/Cocoa/CoreTextHelpers.mm:
(WebKit::fontWithAttributes):
* Shared/RemoteLayerTree/RemoteLayerTreePropertyApplier.mm:
(WebKit::RemoteLayerTreePropertyApplier::applyProperties):
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::updateFontManagerIfNeeded):
* UIProcess/RemoteLayerTree/mac/ScrollerMac.mm:
(-[WKScrollbarPartAnimation invalidate]):
(-[WKScrollerImpDelegate cancelAnimations]):
(-[WKScrollerImpDelegate invalidate]):
(WebKit::ScrollerMac::updateValues):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView assignLegacyDataForContextMenuInteraction]):
(-[WKContentView continueContextMenuInteractionWithDataDetectors:]):
* UIProcess/mac/WebPopupMenuProxyMac.mm:
(WebKit::WebPopupMenuProxyMac::showPopupMenu):
* WebProcess/WebPage/RemoteLayerTree/PlatformCAAnimationRemote.mm:
(WebKit::PlatformCAAnimationRemote::updateLayerAnimations):
2020-07-06 Brady Eidson <beidson@apple.com>
Get rid of concept of "initial connected gamepads"
https://bugs.webkit.org/show_bug.cgi?id=214010
Reviewed by Tim Horton.
* Scripts/webkit/messages.py:
* UIProcess/Gamepad/UIGamepadProvider.cpp:
(WebKit::UIGamepadProvider::gamepadSyncTimerFired):
(WebKit::UIGamepadProvider::platformGamepadConnected):
(WebKit::UIGamepadProvider::platformGamepadInputActivity):
(WebKit::UIGamepadProvider::setInitialConnectedGamepads): Deleted.
* UIProcess/Gamepad/UIGamepadProvider.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::gamepadActivity):
* UIProcess/WebPageProxy.h:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::gamepadConnected):
(WebKit::WebProcessPool::setInitialConnectedGamepads): Deleted.
* UIProcess/WebProcessPool.h:
* WebProcess/Gamepad/WebGamepadProvider.cpp:
(WebKit::WebGamepadProvider::gamepadConnected):
(WebKit::WebGamepadProvider::gamepadActivity):
* WebProcess/Gamepad/WebGamepadProvider.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::gamepadActivity):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::gamepadConnected):
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
2020-07-06 Chris Dumez <cdumez@apple.com>
Regression(r249303) Crash under NetworkLoad::NetworkLoad()
https://bugs.webkit.org/show_bug.cgi?id=214008
<rdar://problem/64853936>
Reviewed by Alex Christensen.
* NetworkProcess/cache/NetworkCacheSpeculativeLoad.cpp:
(WebKit::NetworkCache::SpeculativeLoad::SpeculativeLoad):
Do some hardening and fail the SpeculativeLoad if the network session is null, instead of dereferencing
the network session unconditionally. The NetworkCache owns the NetworkCacheSpeculativeLoadManager and
the NetworkCache is RefCounted so it may outlive its NetworkSession in theory and schedule speculative
loads for a session that was just destroyed.
* NetworkProcess/cache/NetworkCacheSpeculativeLoadManager.cpp:
(WebKit::NetworkCache::SpeculativeLoadManager::registerLoad):
(WebKit::NetworkCache::SpeculativeLoadManager::preloadEntry):
* NetworkProcess/cache/NetworkCacheSpeculativeLoadManager.h:
Capture weakThis in a few lambda and check it when the lambda gets called. It looked unsafe so I
decided to do some hardening.
2020-07-06 Peng Liu <peng.liu6@apple.com>
REGRESSION (r261001): ASSERTION FAILED: m_clientCounts.contains(contextId) in WebKit::VideoFullscreenManagerProxy::removeClientForContext
https://bugs.webkit.org/show_bug.cgi?id=212986
Reviewed by Jer Noble.
* UIProcess/Cocoa/VideoFullscreenManagerProxy.mm:
(WebKit::VideoFullscreenManagerProxy::removeClientForContext):
Function removeClientForContext() could be called multiple times with the same contextId.
We should bail out if m_clientCounts does not contain the contextId.
2020-07-06 Kate Cheney <katherine_cheney@apple.com>
update-webkit-localizable-strings exits without writing localized strings because of an R"" string
https://bugs.webkit.org/show_bug.cgi?id=213274
<rdar://problem/64427891>
Reviewed by Darin Adler.
Remove R"" string, use adjacent string literals instead.
No new tests, functionality tested by existing tests, specifically
(http/tests/resourceLoadStatistics/telemetry-generation-basic-functionality-database.html
and http/tests/resourceLoadStatistics/telemetry-generation-advanced-functionality-database.html).
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::joinSubStatisticsForSorting):
(WebKit::CompletionHandler<void): Deleted.
2020-07-06 Wenson Hsieh <wenson_hsieh@apple.com>
[iPadOS] Unable to change focus between Google Docs windows by tapping
https://bugs.webkit.org/show_bug.cgi?id=213985
<rdar://problem/57083267>
Reviewed by Darin Adler.
When putting two Google Docs windows side-by-side on iPad, it's currently not possible to change the window to
which keyboard input is routed. In native views (e.g. two side-by-side Notes windows), this is normally handled
by `UITextMultiTapRecognizer`, which is part of the `UITextSelectionInteraction`; tapping to place the text
selection calls into `-[UITextInteractionAssistant setFirstResponderIfNecessaryActivatingSelection:]`, which
updates the key window if needed. This doesn't apply to Google Docs because they instead use touch events to
drive their own "text interaction"-like behaviors instead of relying on system gesture recognizers, which we
suppress due to the fact that the selection is within a hidden contenteditable.
But even in non-hidden editable areas, the initial tap to focus an editable element still doesn't automatically
make the window key, since the editable text interaction gestures are still inactive when tapping to focus an
editable element. This means two taps are currently required to change the key window when focusing two Safari
windows side-by-side: the first tap to focus the element (via the synthetic click gesture), and the second tap
to trigger the editable text interaction tap gesture that is used to set the selection.
To fix both of these issues, make some minor adjustments to call `-makeKeyWindow` from WebKit when focusing
editable elements. See below for more details.
Test: editing/selection/ios/become-key-window-when-focusing-editable-area.html
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _webTouchEventsRecognized:]):
For the case in Google Docs where a hidden editable element is used and tapping in the page does not result in
the element being refocused, we additionally need to make sure that we make our window key anyways. Limit this
hack to tap gestures, and only when there is a hidden focused editable element to emulate platform behavior of
updating the key window based on the text tap gesture.
(-[WKContentView _elementDidFocus:userIsInteracting:blurPreviousNode:activityStateChanges:userObject:]):
Make the content view's window key when focusing an editable element.
(-[WKContentView hasHiddenContentEditable]):
Consider the selection to be inside a hidden editable area if the `WebKit::FocusedElementIsTooSmall` flag is set
as well. While this doesn't affect Google Docs, it does affect some other custom editors, such as Quip.
(-[WKContentView mouseGestureRecognizerChanged:]):
Additionally make the current window key when clicking in a hidden editable area with a trackpad on iOS.
2020-07-06 Sam Weinig <weinig@apple.com>
Compile-time enable (but leave disabled at runtime by default) date/time input types on macOS to allow testing of cross platform (e.g. DOM) aspects of the controls
https://bugs.webkit.org/show_bug.cgi?id=213949
Reviewed by Darin Adler.
* Shared/FocusedElementInformation.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKFocusedElementInfo initWithFocusedElementInformation:isUserInitiated:userObject:]):
(-[WKContentView requiresAccessoryView]):
(-[WKContentView textInputTraits]):
(mayContainSelectableText):
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::getFocusedElementInformation):
Remove DateTime from the InputType enum. This should have been removed when datetime was removed.
* Shared/WebPreferences.yaml:
Add preferences for all the date/time related input types, and have them automatically
link up with the newly added WebCore Settings of the same names (this is the default
behavior if no explicit webcoreBinding: is provided).
* Shared/WebPreferencesDefaultValues.h:
Disable date/time input elements by default for non iOS/watchOS cocoa platforms. For non-Cocoa platforms
(and iOS/watchOS), they are enabled by default, but only actually get enabled if the appropriate ENABLE_*
macro is also set.
2020-07-06 Haelwenn (lanodan) Monnier <contact@hacktivis.me>
BubblewrapLauncher.cpp: Add paths required for sndio
https://bugs.webkit.org/show_bug.cgi?id=212524
Information about the paths currently ins't into the sndio
documentation but it can be found in this email:
http://www.sndio.org/arch/0077.html
Reviewed by Adrian Perez de Castro.
Manually tested, there is currently no automatic tests for it
* UIProcess/Launcher/glib/BubblewrapLauncher.cpp:
(WebKit::bindSndio):
(WebKit::bubblewrapSpawn):
2020-07-06 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK][WPE] Change the cookies accept policy when ITP is enabled
https://bugs.webkit.org/show_bug.cgi?id=213502
Reviewed by Michael Catanzaro.
* NetworkProcess/Cookies/soup/WebCookieManagerSoup.cpp:
(WebKit::WebCookieManager::platformSetHTTPCookieAcceptPolicy): Use NetworkStorageSession::setCookieAcceptPolicy().
* NetworkProcess/soup/NetworkSessionSoup.cpp:
(WebKit::NetworkSessionSoup::setCookiePersistentStorage): Do not change the accept policy, it's now done by
NetworkStorageSession when the new cookie jar is set.
* UIProcess/API/glib/WebKitCookieManager.cpp: Update API docs.
* UIProcess/API/glib/WebKitWebsiteDataManager.cpp: Ditto.
2020-07-05 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r263960.
https://bugs.webkit.org/show_bug.cgi?id=213980
Re-land, because r263959 somehow fixed the build issue caused
by r263953
Reverted changeset:
"Unreviewed, reverting r263953 and r263959."
https://bugs.webkit.org/show_bug.cgi?id=213979
https://trac.webkit.org/changeset/263960
2020-07-05 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r263953 and r263959.
https://bugs.webkit.org/show_bug.cgi?id=213979
Broke internal build
Reverted changesets:
"[Cocoa] Move almost all features from FeatureDefines.xcconfig
to PlatformEnableCocoa.h"
https://bugs.webkit.org/show_bug.cgi?id=212542
https://trac.webkit.org/changeset/263953
"[Cocoa] Remove FEATURE_DEFINES from the Cocoa/Xcode build
system"
https://bugs.webkit.org/show_bug.cgi?id=213976
https://trac.webkit.org/changeset/263959
2020-07-05 Darin Adler <darin@apple.com>
[Cocoa] Remove FEATURE_DEFINES from the Cocoa/Xcode build system
https://bugs.webkit.org/show_bug.cgi?id=213976
Reviewed by Sam Weinig.
* Configurations/BaseTarget.xcconfig: Removed FEATURE_DEFINES, and added
WK_ENABLE_EXPERIMENTAL_FEATURES,
* Configurations/FeatureDefines.xcconfig: Removed everything except for ENABLE_WEB_RTC.
2020-07-05 Darin Adler <darin@apple.com>
[Cocoa] Move almost all features from FeatureDefines.xcconfig to PlatformEnableCocoa.h
https://bugs.webkit.org/show_bug.cgi?id=212542
Reviewed by Sam Weinig.
* Configurations/FeatureDefines.xcconfig: Delete everything except
ENABLE_EXPERIMENTAL_FEATURES and ENABLE_WEBRTC.
* Configurations/WebKit.xcconfig: Added include of WebKitTargetConditionals.xcconfig.
2020-07-05 Said Abou-Hallawa <sabouhallawa@apple.com>
[macOS] Allow selecting HEIF images if the 'accept' attribute includes an image MIME type that the platform can transcode the HEIF to
https://bugs.webkit.org/show_bug.cgi?id=212489
Reviewed by Darin Adler.
Make WKOpenPanelParameters call allowedFileExtensions() and allowedMIMETypes()
of MIMETypeRegistry via OpenPanelParameters.
UIDelegate will get allowedMIMETypes() and pass it to WebOpenPanelResultListenerProxy
which will pass it to WebPageProxy. In a following patch, WebPageProxy
will use allowedMIMETypes to transcode the selected images before sending
them to WebPage.
* UIProcess/API/APIOpenPanelParameters.cpp:
(API::OpenPanelParameters::allowedMIMETypes const):
(API::OpenPanelParameters::allowedFileExtensions const):
(API::OpenPanelParameters::mediaCaptureType const): Deleted.
* UIProcess/API/APIOpenPanelParameters.h:
(API::OpenPanelParameters::mediaCaptureType const):
* UIProcess/API/Cocoa/WKOpenPanelParameters.mm:
(-[WKOpenPanelParameters _allowedFileExtensions]):
(-[WKOpenPanelParameters _allowedFileExtensionsTitles]): Deleted.
* UIProcess/API/Cocoa/WKOpenPanelParametersPrivate.h:
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::UIClient::runOpenPanel):
* UIProcess/WebOpenPanelResultListenerProxy.cpp:
(WebKit::WebOpenPanelResultListenerProxy::chooseFiles):
* UIProcess/WebOpenPanelResultListenerProxy.h:
(WebKit::WebOpenPanelResultListenerProxy::chooseFiles):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didChooseFilesForOpenPanel):
* UIProcess/WebPageProxy.h:
2020-07-04 Darin Adler <darin@apple.com>
[Cocoa] Remove all features from FeatureDefines.xcconfig that are already mentioned in PlatformEnableCocoa.h
https://bugs.webkit.org/show_bug.cgi?id=213962
Reviewed by Sam Weinig.
* Configurations/FeatureDefines.xcconfig: Removed all features that were mentioned
in PlatformEnableCocoa.h; the rules in that file now define whether they are enabled.
2020-07-04 Darin Adler <darin@apple.com>
[Cocoa] Remove unconditional features from FeatureDefines.xcconfig, making sure they are covered in PlatformEnableCocoa.h
https://bugs.webkit.org/show_bug.cgi?id=212418
Reviewed by Sam Weinig.
* Configurations/FeatureDefines.xcconfig: Removed features that are either unconditionally not enabled,
or unconditionally enabled. Double checked that all the enabled ones are either in PlatformEnable.h or
PlatformEnableCocoa.h.
2020-07-04 Sam Weinig <weinig@apple.com>
Part 1 of SimpleColor and SRGBA<uint8_t> are essentially the same - let's converge them
https://bugs.webkit.org/show_bug.cgi?id=213948
Reviewed by Darin Adler.
* UIProcess/API/wpe/WebKitColor.cpp:
(webkitColorFillFromWebCoreColor):
* UIProcess/gtk/ViewGestureControllerGtk.cpp:
(WebKit::ViewGestureController::beginSwipeGesture):
* WebProcess/WebPage/RemoteLayerTree/PlatformCAAnimationRemote.mm:
(WebKit::animationValueFromKeyframeValue):
* WebProcess/WebPage/WebFrame.cpp:
(WebKit::WebFrame::getDocumentBackgroundColor):
Adapt to use templatized toSRGBALossy<float/uint8_t>.
2020-07-03 Darin Adler <darin@apple.com>
Make generate-unified-sources.sh not depend on features being listed in FEATURE_DEFINES environment variable
https://bugs.webkit.org/show_bug.cgi?id=212420
Reviewed by Don Olmstead.
* Scripts/generate-unified-sources.sh: Removed --feature-flags from the
arguments passed to generate-unified-source-bundles.rb. Also removed many
unneeded quotes.
2020-07-04 Darin Adler <darin@apple.com>
Update comment in FeatureDefines.xcconfig since PlatformEnableCocoa.h should be used instead
https://bugs.webkit.org/show_bug.cgi?id=213952
Reviewed by Sam Weinig.
* Configurations/FeatureDefines.xcconfig: Updated comment.
2020-07-04 Youenn Fablet <youenn@apple.com>
Enable VTB required low latency code path
https://bugs.webkit.org/show_bug.cgi?id=210609
<rdar://problem/61890332>
Reviewed by Darin Adler.
Add a runtime flag to enable/disable the low latency code path.
* Shared/WebPreferences.yaml:
2020-07-03 Youenn Fablet <youenn@apple.com>
MediaRecorder should support peer connection remote video tracks
https://bugs.webkit.org/show_bug.cgi?id=213924
Reviewed by Darin Adler.
Pass a boolean instead of width/height for the video track.
* GPUProcess/webrtc/RemoteMediaRecorder.cpp:
(WebKit::RemoteMediaRecorder::create):
* GPUProcess/webrtc/RemoteMediaRecorder.h:
* GPUProcess/webrtc/RemoteMediaRecorderManager.cpp:
(WebKit::RemoteMediaRecorderManager::createRecorder):
* GPUProcess/webrtc/RemoteMediaRecorderManager.h:
* GPUProcess/webrtc/RemoteMediaRecorderManager.messages.in:
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
(WebKit::MediaRecorderPrivate::startRecording):
* WebProcess/GPU/webrtc/MediaRecorderPrivate.h:
2020-07-03 Sam Weinig <weinig@apple.com>
Remove support for ENABLE_INPUT_TYPE_DATETIME_INCOMPLETE
https://bugs.webkit.org/show_bug.cgi?id=213932
Reviewed by Darin Adler.
Removes support for non-standard <input type="datetime">, currently being
guarded by the macro ENABLE_INPUT_TYPE_DATETIME_INCOMPLETE. This macro, was
added back in 2013 as a temporary measure to support some engines who shipped
support for <input type="datetime">. It is currently not enabled for any
ports so now seems like as good a time as any to remove it.
* Configurations/FeatureDefines.xcconfig:
2020-07-03 Sam Weinig <weinig@apple.com>
Add "-Wliteral-conversion" warning to Xcode based builds and fix the issues it finds
https://bugs.webkit.org/show_bug.cgi?id=213931
Reviewed by Darin Adler.
* Configurations/Base.xcconfig:
Add -Wliteral-conversion.
2020-07-03 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r263882.
https://bugs.webkit.org/show_bug.cgi?id=213933
Broke the build
Reverted changeset:
"Make _WKWebsiteDataStoreConfiguration SPI for HSTS storage to
replace _WKProcessPoolConfiguration.hstsStorageDirectory"
https://bugs.webkit.org/show_bug.cgi?id=213048
https://trac.webkit.org/changeset/263882
2020-07-03 Michael Catanzaro <mcatanzaro@gnome.org>
[WPE][GTK] Bubblewrap sandbox should not attempt to bind empty paths
https://bugs.webkit.org/show_bug.cgi?id=202880
Reviewed by Carlos Garcia Campos.
If an empty path is set in LD_LIBRARY_PATH (e.g. /foo::/bar), we wind up passing
--ro-bind-try --ro-bind-try to bwrap, where the second --ro-bind-try is treated as the
source path to mount, and the next flag we pass gets treated as the destination path. The
web process crashes with this confusing error:
bwrap: Can't create file at /.flatpak-info: Read-only file system
* UIProcess/Launcher/glib/BubblewrapLauncher.cpp:
(WebKit::bindIfExists):
2020-07-03 Youenn Fablet <youenn@apple.com>
Allow registering VP9 as a VT decoder
https://bugs.webkit.org/show_bug.cgi?id=213837
<rdar://problem/64984881>
Reviewed by Jer Noble.
Add a runtime flag to enable registration of VP9 as a VT decoder.
Register conditionally this decoder on creation of page based on this flag.
Test: media/vp9.html
* Scripts/GeneratePreferences.rb:
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
* Shared/WebPreferences.yaml:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::enableVP9Decoder):
* WebProcess/WebProcess.h:
2020-07-03 Youenn Fablet <youenn@apple.com>
MediaRecorder stopRecorder() returns empty Blob after first use
https://bugs.webkit.org/show_bug.cgi?id=212274
<rdar://problem/63601298>
Reviewed by Eric Carlson.
Update implementation to do initialization as part of startRecording.
* GPUProcess/webrtc/RemoteMediaRecorderManager.cpp:
(WebKit::RemoteMediaRecorderManager::releaseRecorder):
Remove ASSERT as recorder creation in WebProcess is always ok while creation in GPUProcess may fail and m_recorders may not be populated.
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
(WebKit::MediaRecorderPrivate::MediaRecorderPrivate):
(WebKit::MediaRecorderPrivate::startRecording):
* WebProcess/GPU/webrtc/MediaRecorderPrivate.h:
2020-07-03 Said Abou-Hallawa <sabouhallawa@apple.com>
Change the names of MIMETypeRegistry methods to comply with webkit naming style
https://bugs.webkit.org/show_bug.cgi?id=213912
Reviewed by Youenn Fablet.
* NetworkProcess/soup/NetworkDataTaskSoup.cpp:
(WebKit::NetworkDataTaskSoup::didSendRequest):
* UIProcess/API/Cocoa/APIAttachmentCocoa.mm:
(API::mimeTypeInferredFromFileExtension):
(API::Attachment::setFileWrapperAndUpdateContentType):
* UIProcess/API/Cocoa/WKOpenPanelParameters.mm:
(-[WKOpenPanelParameters _allowedFileExtensions]):
* UIProcess/API/glib/WebKitURISchemeRequest.cpp:
(webkitURISchemeRequestReadCallback):
* UIProcess/Automation/WebAutomationSession.cpp:
(WebKit::fileCanBeAcceptedForUpload):
* UIProcess/Plugins/PluginInfoStore.cpp:
(WebKit::PluginInfoStore::findPlugin):
* UIProcess/ios/forms/WKFileUploadPanel.mm:
(-[WKFileUploadPanel presentWithParameters:resultListener:]):
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::objectContentType):
2020-07-03 Alexander Mikhaylenko <alexm@gnome.org>
[GTK4] Support kinetic scrolling
https://bugs.webkit.org/show_bug.cgi?id=213791
Reviewed by Carlos Garcia Campos.
GtkEventControllerScroll skips the last event and exposes it in a separate signal.
Normally one would add the KINETIC flag and listen to decelerate signal, but since
WebKit already has code for calculating the velocity for GTK3, just use the same
code path as scroll signal handler.
* Shared/gtk/NativeWebWheelEventGtk.cpp:
(WebKit::NativeWebWheelEvent::NativeWebWheelEvent):
* Shared/gtk/WebEventFactory.cpp:
(WebKit::WebEventFactory::createWebWheelEvent):
* Shared/gtk/WebEventFactory.h:
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseScrollEnd):
(webkitWebViewBaseConstructed):
2020-07-02 Alex Christensen <achristensen@webkit.org>
Make _WKWebsiteDataStoreConfiguration SPI for HSTS storage to replace _WKProcessPoolConfiguration.hstsStorageDirectory
https://bugs.webkit.org/show_bug.cgi?id=213048
Reviewed by Youenn Fablet.
This uses CFNetwork SPI introduced in rdar://problem/50109631 to allow HSTS storage per NSURLSession.
To be complete, I also deprecated our UI process HSTS state removal attempt SPIs, WKContextResetHSTSHosts and
WKContextResetHSTSHostsAddedAfterDate, which had their last use removed in rdar://problem/64220838.
I manually verified that this new SPI puts HSTS data in the specified location, and I also verified that HSTS
state querying and removal works with the new CFNetwork SPI as it did with the old one.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::fetchWebsiteData):
(WebKit::NetworkProcess::deleteWebsiteData):
(WebKit::NetworkProcess::deleteWebsiteDataForOrigins):
(WebKit::NetworkProcess::deleteAndRestrictWebsiteDataForRegistrableDomains):
(WebKit::NetworkProcess::registrableDomainsWithWebsiteData):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkSessionCreationParameters.cpp:
(WebKit::NetworkSessionCreationParameters::encode const):
(WebKit::NetworkSessionCreationParameters::decode):
* NetworkProcess/NetworkSessionCreationParameters.h:
* NetworkProcess/cocoa/NetworkProcessCocoa.mm:
(WebKit::NetworkProcess::hostNamesWithHSTSCache const):
(WebKit::NetworkProcess::deleteHSTSCacheForHostNames):
(WebKit::NetworkProcess::clearHSTSCache):
(WebKit::NetworkProcess::getHostNamesWithHSTSCache): Deleted.
* NetworkProcess/cocoa/NetworkSessionCocoa.h:
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::hstsStorage const):
(WebKit::NetworkSessionCocoa::NetworkSessionCocoa):
* NetworkProcess/soup/NetworkProcessSoup.cpp:
(WebKit::NetworkProcess::hostNamesWithHSTSCache const):
(WebKit::NetworkProcess::deleteHSTSCacheForHostNames):
(WebKit::NetworkProcess::clearHSTSCache):
(WebKit::NetworkProcess::getHostNamesWithHSTSCache): Deleted.
* UIProcess/API/C/mac/WKContextPrivateMac.h:
* UIProcess/API/C/mac/WKContextPrivateMac.mm:
(WKContextResetHSTSHosts):
(WKContextResetHSTSHostsAddedAfterDate):
* UIProcess/API/Cocoa/_WKProcessPoolConfiguration.h:
* UIProcess/API/Cocoa/_WKWebsiteDataStoreConfiguration.h:
* UIProcess/API/Cocoa/_WKWebsiteDataStoreConfiguration.mm:
(-[_WKWebsiteDataStoreConfiguration hstsStorageDirectory]):
(-[_WKWebsiteDataStoreConfiguration setHSTSStorageDirectory:]):
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::privateBrowsingSession): Deleted.
(WebKit::WebProcessPool::resetHSTSHosts): Deleted.
(WebKit::WebProcessPool::resetHSTSHostsAddedAfterDate): Deleted.
* UIProcess/WebProcessPool.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::resolveDirectoriesIfNecessary):
(WebKit::WebsiteDataStore::parameters):
* UIProcess/WebsiteData/WebsiteDataStore.h:
(WebKit::WebsiteDataStore::resolvedHSTSStorageDirectory const):
* UIProcess/WebsiteData/WebsiteDataStoreConfiguration.cpp:
(WebKit::WebsiteDataStoreConfiguration::copy const):
* UIProcess/WebsiteData/WebsiteDataStoreConfiguration.h:
2020-07-02 Alex Christensen <achristensen@webkit.org>
Update Mac CMake build
* PlatformMac.cmake:
2020-07-02 Kate Cheney <katherine_cheney@apple.com>
Custom URL schemes should be treated as app-bound
https://bugs.webkit.org/show_bug.cgi?id=213889
<rdar://problem/64804671>
Reviewed by Brent Fulgham.
For applications which opt-in to App-Bound Domains, allow
specification of app-bound custom URL schemes. All content loaded
using an app-bound scheme will have access to otherwise restricted
APIs. A custom scheme is specified by a colon at the end of a string
in the WKAppBoundDomains list. Custom schemes are included in the
count of 10 app-bound domains.
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
(-[WKWebsiteDataStore _appBoundSchemes:]):
* UIProcess/API/Cocoa/WKWebsiteDataStorePrivate.h:
SPI for testing.
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::appBoundSchemes):
(WebKit::WebsiteDataStore::initializeAppBoundDomains):
Change variable name from appBoundDomains to appBoundData now that
more than domains can be specified in the Info.plist.
(WebKit::WebsiteDataStore::ensureAppBoundDomains const):
Return both domains and schemes to avoid code duplication, at the expense
of occasionally returning an unused parameter if only domains or only
schemes are needed.
(WebKit::WebsiteDataStore::beginAppBoundDomainCheck):
(WebKit::WebsiteDataStore::getAppBoundDomains const):
(WebKit::WebsiteDataStore::getAppBoundSchemes const):
* UIProcess/WebsiteData/WebsiteDataStore.h:
2020-07-02 Austin Blackwood <ablackwoood@apple.com>
Crash in +[UIViewController _viewControllerForFullScreenPresentationFromView:] when WKContentView is deallocated
https://bugs.webkit.org/show_bug.cgi?id=213867
Reviewed by Darin Adler.
+[UIViewController _viewControllerForFullScreenPresentationFromView:]
asserts that the view parameter is non-nil. WKFileUploadPanel
passes its view (a weak pointer to the WKContentView), which may
have already been nil'd out when the content view tries to dismiss
the panel in its own dealloc.
* UIProcess/ios/forms/WKFileUploadPanel.mm:
(-[WKFileUploadPanel dismiss]):
Check for nil.
2020-07-02 Chris Dumez <cdumez@apple.com>
FloatingPointEnvironment::saveMainThreadEnvironment() is not called for macOS Catalyst apps on Apple Silicon
https://bugs.webkit.org/show_bug.cgi?id=213897
<rdar://problem/64607098>
Reviewed by Darin Adler.
FloatingPointEnvironment::saveMainThreadEnvironment() is not called for macOS Catalyst apps on Apple Silicon
This causes crashes because FloatingPointEnvironment::propagateMainThreadEnvironment() is called for worker
threads for such apps on this platform.
* Shared/ios/AuxiliaryProcessIOS.mm:
(WebKit::AuxiliaryProcess::initializeSandbox):
Drop some dead code for PLATFORM(MACCATALYST) since the whole file is protected with !PLATFORM(MACCATALYST).
Mac Catalyst uses AuxiliaryProcessMac.mm instead.
* Shared/mac/AuxiliaryProcessMac.mm:
(WebKit::AuxiliaryProcess::platformInitialize):
Call FloatingPointEnvironment::saveMainThreadEnvironment() on AuxiliaryProcessMac.mm since this
platformInitialize() is used for Mac Catalyst and Catalyst apps may now run on Apple Silicon / ARM.
Note that FloatingPointEnvironment::saveMainThreadEnvironment() is a no-op on platforms where it is
not needed.
2020-07-02 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed, reverting r263633, r263651, and r263671.
Still seeing MediaRecorder test crashes after re-landing
r263633
Reverted changesets:
"MediaRecorder stopRecorder() returns empty Blob after first
use"
https://bugs.webkit.org/show_bug.cgi?id=212274
https://trac.webkit.org/changeset/263633
"MediaRecorder.start() Method is Ignoring the "timeslice"
Parameter"
https://bugs.webkit.org/show_bug.cgi?id=202233
https://trac.webkit.org/changeset/263651
"Support MediaRecorder.onstart"
https://bugs.webkit.org/show_bug.cgi?id=213720
https://trac.webkit.org/changeset/263671
2020-07-02 Chris Dumez <cdumez@apple.com>
Crash under WebKit::NetworkProcessProxy::updateProcessAssertion()
https://bugs.webkit.org/show_bug.cgi?id=213891
<rdar://problem/65017909>
Reviewed by Alex Christensen.
The crash was due to NetworkProcessProxy::updateProcessAssertion() re-entering while
in the middle of the `m_activityFromWebProcesses = nullptr;` assignment. Calling
the ProcessThrottler::BackgroundActivity destructor, could cause updateProcessAssertion()
to get called again, in which case we may dereference m_activityFromWebProcesses and
crash. To address the issue, use `std::exchange(m_activityFromWebProcesses, nullptr);`
instead, so that m_activityFromWebProcesses becomes null BEFORE the BackgroundActivity
destructor gets called. updateProcessAssertion() will still re-enter but
m_activityFromWebProcesses will be nullptr and updateProcessAssertion() will do the
right thing.
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::updateProcessAssertion):
2020-07-02 Carlos Garcia Campos <cgarcia@igalia.com>
Unreviewed. Fix GTK4 build
* UIProcess/API/gtk/WebKitScriptDialogImpl.cpp:
(webkitScriptDialogImplConstructed):
* UIProcess/gtk/WebDataListSuggestionsDropdownGtk.cpp:
(WebKit::WebDataListSuggestionsDropdownGtk::WebDataListSuggestionsDropdownGtk):
* UIProcess/gtk/WebPopupMenuProxyGtk.cpp:
(WebKit::WebPopupMenuProxyGtk::createPopupMenu):
2020-07-01 Said Abou-Hallawa <sabouhallawa@apple.com>
MIMETypeRegistry::getExtensionsForMIMEType() needs to handle wildcard MIME types
https://bugs.webkit.org/show_bug.cgi?id=213826
Reviewed by Darin Adler.
Replace extensionsForMIMEType() with MIMETypeRegistry::getExtensionsForMIMEType().
* UIProcess/API/Cocoa/WKOpenPanelParameters.mm:
(-[WKOpenPanelParameters _allowedFileExtensions]):
2020-07-01 Lauro Moura <lmoura@igalia.com>
[SOUP] Build fix after r263797 for older soup versions.
Build was failing in LTS distributions due to -fpermissive
Unreviewed build fix.
* NetworkProcess/soup/WebSocketTaskSoup.cpp:
(WebKit::WebSocketTask::sendString):
2020-07-01 Said Abou-Hallawa <sabouhallawa@apple.com>
Allow the File object to be created with a replacement file
https://bugs.webkit.org/show_bug.cgi?id=213825
Reviewed by Darin Adler.
The UIProcess passes a list of strings which represents the replacement
paths along with a list to the original paths to the WebProcess. The
WebProcess passes these two list to FileChooser which creates the File
objects and register the Blobs.
The WebProcess registers the Blobs in the NetworkProcess which creates
BlobDataFileReference objects with both the original path and the
replacement path.
The WebProcess unregisters the Blobs from the NetworkProcess which deletes
the corresponding BlobDataFileReference from its registry. Upon destroying
the BlobDataFileReference, the replacement file should be deleted.
* NetworkProcess/NetworkConnectionToWebProcess.cpp:
(WebKit::NetworkConnectionToWebProcess::registerFileBlobURL):
(WebKit::NetworkConnectionToWebProcess::registerBlobURLOptionallyFileBacked):
* NetworkProcess/NetworkConnectionToWebProcess.h:
* NetworkProcess/NetworkConnectionToWebProcess.messages.in:
* NetworkProcess/NetworkProcessPlatformStrategies.cpp:
(WebKit::NetworkProcessPlatformStrategies::createBlobRegistry):
* Shared/BlobDataFileReferenceWithSandboxExtension.cpp:
(WebKit::BlobDataFileReferenceWithSandboxExtension::BlobDataFileReferenceWithSandboxExtension):
* Shared/BlobDataFileReferenceWithSandboxExtension.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didChooseFilesForOpenPanel):
* WebProcess/FileAPI/BlobRegistryProxy.cpp:
(WebKit::BlobRegistryProxy::registerFileBlobURL):
* WebProcess/FileAPI/BlobRegistryProxy.h:
* WebProcess/WebPage/WebOpenPanelResultListener.cpp:
(WebKit::WebOpenPanelResultListener::didChooseFiles):
* WebProcess/WebPage/WebOpenPanelResultListener.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::didChooseFilesForOpenPanel):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-07-01 Tim Horton <timothy_horton@apple.com>
Swipe snapshot is removed too early when swiping away from a page that is still loading
https://bugs.webkit.org/show_bug.cgi?id=213763
<rdar://problem/64576811>
Reviewed by Darin Adler.
Test: http/tests/swipe/swipe-back-with-outstanding-load-cancellation.html
When swiping back from a page that is still loading, when the swipe commits,
and we start navigating back to the previous page, the unfinished forward navigation
is cancelled.
This cancellation resulted in a "didFailLoadForMainFrame" making it to ViewGestureController,
which took it as an indication that the navigation *it* was interested in (the back navigation)
had failed, and immediately removed the snapshot.
Instead of listening to any random navigation's load notifications, keep track of the navigation
started by goToBackForwardItem, and only listen to notifications from it, ignoring the others.
This requires a bunch of plumbing to get the navigation from WebPageProxy to ViewGestureController,
but is otherwise fairly trivial.
* UIProcess/API/gtk/PageClientImpl.cpp:
(WebKit::PageClientImpl::didFinishNavigation):
(WebKit::PageClientImpl::didFailNavigation):
(WebKit::PageClientImpl::didFinishLoadForMainFrame): Deleted.
(WebKit::PageClientImpl::didFailLoadForMainFrame): Deleted.
* UIProcess/API/gtk/PageClientImpl.h:
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseDidFinishLoadForMainFrame):
(webkitWebViewBaseDidFailLoadForMainFrame):
* UIProcess/API/gtk/WebKitWebViewBasePrivate.h:
* UIProcess/API/ios/WKWebViewIOS.h:
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _didFinishNavigation:]):
(-[WKWebView _didFailNavigation:]):
(-[WKWebView _didFinishLoadForMainFrame]): Deleted.
(-[WKWebView _didFailLoadForMainFrame]): Deleted.
* UIProcess/API/wpe/PageClientImpl.cpp:
(WebKit::PageClientImpl::didFinishNavigation):
(WebKit::PageClientImpl::didFailNavigation):
(WebKit::PageClientImpl::didFinishLoadForMainFrame): Deleted.
(WebKit::PageClientImpl::didFailLoadForMainFrame): Deleted.
* UIProcess/API/wpe/PageClientImpl.h:
* UIProcess/PageClient.h:
* UIProcess/ViewGestureController.cpp:
(WebKit::ViewGestureController::didReachNavigationTerminalState):
(WebKit::ViewGestureController::willEndSwipeGesture):
(WebKit::ViewGestureController::didReachMainFrameLoadTerminalState): Deleted.
* UIProcess/ViewGestureController.h:
(WebKit::ViewGestureController::didFinishNavigation):
(WebKit::ViewGestureController::didFailNavigation):
(WebKit::ViewGestureController::didFinishLoadForMainFrame): Deleted.
(WebKit::ViewGestureController::didFailLoadForMainFrame): Deleted.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didFinishLoadForFrame):
(WebKit::WebPageProxy::didFailLoadForFrame):
(WebKit::WebPageProxy::didSameDocumentNavigationForFrame):
* UIProcess/ios/PageClientImplIOS.h:
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::didFinishNavigation):
(WebKit::PageClientImpl::didFailNavigation):
(WebKit::PageClientImpl::didFinishLoadForMainFrame): Deleted.
(WebKit::PageClientImpl::didFailLoadForMainFrame): Deleted.
* UIProcess/ios/ViewGestureControllerIOS.mm:
(WebKit::ViewGestureController::willEndSwipeGesture):
(WebKit::ViewGestureController::resetState):
* UIProcess/mac/PageClientImplMac.h:
* UIProcess/mac/PageClientImplMac.mm:
(WebKit::PageClientImpl::didFinishNavigation):
(WebKit::PageClientImpl::didFailNavigation):
(WebKit::PageClientImpl::didFinishLoadForMainFrame): Deleted.
(WebKit::PageClientImpl::didFailLoadForMainFrame): Deleted.
* UIProcess/mac/ViewGestureControllerMac.mm:
(WebKit::ViewGestureController::resetState):
* UIProcess/playstation/PageClientImpl.cpp:
(WebKit::PageClientImpl::didFinishNavigation):
(WebKit::PageClientImpl::didFailNavigation):
(WebKit::PageClientImpl::didFinishLoadForMainFrame): Deleted.
(WebKit::PageClientImpl::didFailLoadForMainFrame): Deleted.
* UIProcess/playstation/PageClientImpl.h:
* UIProcess/win/PageClientImpl.cpp:
(WebKit::PageClientImpl::didFinishNavigation):
(WebKit::PageClientImpl::didFinishLoadForMainFrame): Deleted.
* UIProcess/win/PageClientImpl.h:
2020-07-01 Brent Fulgham <bfulgham@apple.com>
[iOS] Allow the WebContent sandbox to check the 'hw.product' sysctl
https://bugs.webkit.org/show_bug.cgi?id=213861
<rdar://problem/64988957>
Reviewed by Per Arne Vollan.
Telemetry shows that we need to read the 'hw.product' sysctl in the font subsystem,
primarily to enable certain font features (for various products) in different locales.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-07-01 Tim Horton <timothy_horton@apple.com>
Upstream application accent color support
https://bugs.webkit.org/show_bug.cgi?id=213859
Reviewed by Wenson Hsieh.
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
* UIProcess/Cocoa/PageClientImplCocoa.h:
* UIProcess/PageClient.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
* UIProcess/mac/PageClientImplMac.h:
* UIProcess/mac/PageClientImplMac.mm:
(WebKit::PageClientImpl::accentColor):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::setAccentColor):
Plumb accent color to the Web Content process so that it is used when
deciding the colors of selection and form controls.
2020-07-01 Keith Miller <keith_miller@apple.com>
Add WebXR to WebKit.plist preferences, off by default
https://bugs.webkit.org/show_bug.cgi?id=213860
<rdar://64411568>
Reviewed by Tim Horton.
Right now we look for "WebXR" in the plist, however, it's not in
the file. It so happens that, *I think*, we fall back to false for
missing keys. We should have an entry for readability and
consistency, though.
No tests needed because this shouldn't change behavior.
* FeatureFlags/WebKit.plist:
2020-07-01 Chris Dumez <cdumez@apple.com>
Regression(r259242) WebKit no longer does client navigations at foreground priority for extensions
https://bugs.webkit.org/show_bug.cgi?id=213852
<rdar://problem/63300428>
Reviewed by Alex Christensen.
r259242 tried to address a ASSERT(window) debug assertion hit in applicationType()
by only calling applicationType() when the window is non-null and assuming the
process is an application when the window is null. However, this is not correct
for extensions. The first thing isApplicationVisible() does is call
_UIApplicationIsExtension() to determine if we are an extension. It does not need
the window to make this determination. As a result, I have reverted r259242 and
simply dropped the ASSERT(window) assertion in isApplicationVisible() instead.
This fixes the regression tracked by <rdar://problem/63300428> by reverting the
change that caused it (r259242) and drops the assertion that was hit before r259242.
* UIProcess/ApplicationStateTracker.mm:
(WebKit::applicationType):
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::isApplicationVisible):
2020-07-01 Daniel Bates <dabates@apple.com>
[iOS] Implement support for UIWKDocumentRequestSpatialAndCurrentSelection
https://bugs.webkit.org/show_bug.cgi?id=213704
<rdar://problem/59738878>
Reviewed by Wenson Hsieh.
Implement support for the new request type UIWKDocumentRequestSpatialAndCurrentSelection.
Requests of this type return the contents inside the range that covers both the specified
rect and the current selection, if there is one.
The flag UIWKDocumentRequestSpatialAndCurrentSelection was added in <rdar://problem/64867540>.
* Platform/spi/ios/UIKitSPI.h: Same thing I did in TestWebKitAPI/ios/UIKitSPI.h. See Tools ChangeLog for why.
* Shared/DocumentEditingContext.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(toWebDocumentRequestOptions):
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::requestDocumentEditingContext):
2020-07-01 Chris Dumez <cdumez@apple.com>
Crash under IPC::isValidMessageName()
https://bugs.webkit.org/show_bug.cgi?id=213751
<rdar://problem/64773416>
Reviewed by Geoffrey Garen.
This is a speculative fix for the crashes tracked by <rdar://problem/64773416>.
The suspicion is that the compiler optimizes the code for IPC::isValidMessageName()
because its parameter is of type MessageName. It may therefore assume the parameter
is a valid MessageName enum value, which is ironic because the purpose of this
function is to check whether or not the enum value is valid.
To address the issue, I got rid of the #ifdefs in MessageName enum so that the
enum values have an underlying value in the range [0, MessageName::Last]. I then
updated the isValidEnum template specialization to verify that the input integer
is indeed in the range [0, MessageName::Last] before casting it to a MessageName
and calling IPC::isValidMessageName().
* Scripts/webkit/messages.py:
2020-07-01 Kate Cheney <katherine_cheney@apple.com>
WKUserScripts injecting into all frames seem to violate app-bound domains
https://bugs.webkit.org/show_bug.cgi?id=213816
<rdar://problem/64872296>
Reviewed by Brady Eidson.
Patch to make app-bound domain checks on a per-frame basis instead
of per-page-based-on-main-frame. This will prevent WKUserScripts from
being injected in subframes when the main frame is app-bound. Now, all
ancestors of a frame must be app-bound in order for a frame to
evaluate user script.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::setIsNavigatingToAppBoundDomainAndCheckIfPermitted):
Move 'isMainFrame' check to be inside of the check for
m_limitsNavigationsToAppBoundDomains. The navigation should fail if
the main frame is not app-bound and the WKWebView is marked with the
limitsNavigationToAppBoundDomains flag, but non-app-bound subframe
loads are allowed to continue.
(WebKit::WebPageProxy::decidePolicyForNavigationAction):
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::isTopFrameNavigatingToAppBoundDomain const):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::download):
* WebProcess/Network/WebLoaderStrategy.cpp:
(WebKit::WebLoaderStrategy::scheduleLoadFromNetworkProcess):
(WebKit::WebLoaderStrategy::loadResourceSynchronously):
(WebKit::WebLoaderStrategy::startPingLoad):
(WebKit::WebLoaderStrategy::preconnectTo):
For reporting loads to the Network Process, we only care if the main
frame is app bound, so we need to store that info in the WebPage and
WebPageProxy.
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::shouldEnableInAppBrowserPrivacyProtections const):
* WebProcess/WebPage/WebFrame.cpp:
(WebKit::WebFrame::startDownload):
(WebKit::WebFrame::convertMainResourceLoadToDownload):
(WebKit::WebFrame::shouldEnableInAppBrowserPrivacyProtections):
(WebKit::WebFrame::isTopFrameNavigatingToAppBoundDomain const):
* WebProcess/WebPage/WebFrame.h:
(WebKit::WebFrame::setIsNavigatingToAppBoundDomain):
(WebKit::WebFrame::isNavigatingToAppBoundDomain const):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::loadRequest):
(WebKit::WebPage::loadDataImpl):
(WebKit::WebPage::didReceivePolicyDecision):
(WebKit::WebPage::runJavaScript):
(WebKit::WebPage::setIsNavigatingToAppBoundDomain):
(WebKit::WebPage::shouldEnableInAppBrowserPrivacyProtections): Deleted.
* WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::needsInAppBrowserPrivacyQuirks):
(WebKit::WebPage::isNavigatingToAppBoundDomain const): Deleted.
Move all app-bound domain logic from the WebPage to the WebFrame now
that checks are on a per-frame basis.
2020-07-01 Youenn Fablet <youenn@apple.com>
Make NetworkSendQueue use CString instead of String for UTF-8 data
https://bugs.webkit.org/show_bug.cgi?id=213714
Reviewed by Darin Adler.
Update code to use CString/DataReference to transmit WebSocket text messages.
* NetworkProcess/NetworkSocketChannel.cpp:
(WebKit::NetworkSocketChannel::sendString):
* NetworkProcess/NetworkSocketChannel.h:
* NetworkProcess/NetworkSocketChannel.messages.in:
* NetworkProcess/WebSocketTask.h:
(WebKit::WebSocketTask::sendString):
* NetworkProcess/cocoa/WebSocketTaskCocoa.h:
* NetworkProcess/cocoa/WebSocketTaskCocoa.mm:
(WebKit::WebSocketTask::sendString):
* NetworkProcess/soup/WebSocketTaskSoup.h:
* NetworkProcess/soup/WebSocketTaskSoup.cpp:
(WebKit::WebSocketTask::sendString):
* WebProcess/Network/WebSocketChannel.cpp:
(WebKit::WebSocketChannel::createMessageQueue):
(WebKit::WebSocketChannel::send):
2020-07-01 James Savage <james.savage@apple.com>
Fix Swift overlay build after r263727.
https://bugs.webkit.org/show_bug.cgi?id=213823.
Unreviewed, follow up build fix.
* SwiftOverlay/Tests/WebKitTests.swift:
(WebKitTests.testAPI): Fix tests to refer to new parameter name.
* UIProcess/API/Cocoa/WebKitSwiftOverlay.swift:
(callAsyncJavaScript(_:arguments:in:in:completionHandler:Error:)): Use
the imported ObjC name that the compiler expects.
(evaluateJavaScript(_:in:in:completionHandler:Error:)): Ditto.
2020-07-01 Youenn Fablet <youenn@apple.com>
Make NetworkRTCProvider receive IPC messages from a background thread
https://bugs.webkit.org/show_bug.cgi?id=213557
Reviewed by Geoffrey Garen.
To ensure webrtc packets do not get blocked on the main thread,
we process WebProcess IPC message directly in the socket thread.
For new sockets, we still go through the main thread, to check suspension or proxy values.
For sending packets/setting options/closing sockets, we stay in the background thread.
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* NetworkProcess/NetworkConnectionToWebProcess.cpp:
(WebKit::NetworkConnectionToWebProcess::didReceiveMessage):
* NetworkProcess/webrtc/LibWebRTCSocketClient.cpp:
(WebKit::LibWebRTCSocketClient::sendTo):
* NetworkProcess/webrtc/LibWebRTCSocketClient.h:
* NetworkProcess/webrtc/NetworkRTCProvider.cpp:
(WebKit::NetworkRTCProvider::NetworkRTCProvider):
(WebKit::NetworkRTCProvider::close):
(WebKit::NetworkRTCProvider::createSocket):
(WebKit::NetworkRTCProvider::createUDPSocket):
(WebKit::NetworkRTCProvider::createServerTCPSocket):
(WebKit::NetworkRTCProvider::createClientTCPSocket):
(WebKit::NetworkRTCProvider::wrapNewTCPConnection):
(WebKit::NetworkRTCProvider::sendToSocket):
(WebKit::NetworkRTCProvider::closeSocket):
(WebKit::NetworkRTCProvider::setSocketOption):
(WebKit::NetworkRTCProvider::addSocket):
(WebKit::NetworkRTCProvider::takeSocket):
(WebKit::NetworkRTCProvider::newConnection):
(WebKit::NetworkRTCProvider::dispatchToThread):
* NetworkProcess/webrtc/NetworkRTCProvider.h:
(WebKit::NetworkRTCProvider::didReceiveNetworkRTCMonitorMessage):
* NetworkProcess/webrtc/NetworkRTCProvider.messages.in:
* NetworkProcess/webrtc/NetworkRTCSocket.cpp: Removed.
* NetworkProcess/webrtc/NetworkRTCSocket.h: Removed.
* NetworkProcess/webrtc/NetworkRTCSocket.messages.in: Removed.
* SourcesCocoa.txt:
* SourcesGTK.txt:
* SourcesWPE.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/Network/webrtc/LibWebRTCSocket.cpp:
(WebKit::LibWebRTCSocket::SendTo):
(WebKit::LibWebRTCSocket::Close):
(WebKit::LibWebRTCSocket::SetOption):
(WebKit::LibWebRTCSocket::suspend):
2020-06-30 Sam Weinig <weinig@apple.com>
Split Color serialization out of Color classes
https://bugs.webkit.org/show_bug.cgi?id=213820
Reviewed by Darin Adler.
Adopt ColorSerialization.h serialization functions.
* UIProcess/API/Cocoa/WKWebView.mm:
(dictionaryRepresentationForEditorState):
* UIProcess/Cocoa/WebViewImpl.mm:
(-[WKTextTouchBarItemController _wkChangeColor:]):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _setTextColorForWebView:sender:]):
* UIProcess/ios/forms/WKFormColorPicker.mm:
(-[WKColorPicker setControlValueFromUIColor:]):
* WebProcess/Plugins/PDF/PDFPluginChoiceAnnotation.mm:
(WebKit::PDFPluginChoiceAnnotation::createAnnotationElement):
* WebProcess/Plugins/PDF/PDFPluginTextAnnotation.mm:
(WebKit::PDFPluginTextAnnotation::createAnnotationElement):
2020-06-30 Geoffrey Garen <ggaren@apple.com>
[Cocoa] [GTK] RunLoop::Timer::isActive() is incorrect for timers while they are firing
https://bugs.webkit.org/show_bug.cgi?id=213771
Unreviewed assertion fix.
* UIProcess/WebAuthentication/AuthenticatorManager.cpp:
(WebKit::AuthenticatorManager::timeOutTimerFired): Removed this ASSERT.
One-shot timers are not active while they are firing.
2020-06-30 Wenson Hsieh <wenson_hsieh@apple.com>
Several key event tests in fast/events/ios are failing after <rdar://problem/62197116>
https://bugs.webkit.org/show_bug.cgi?id=213821
<rdar://problem/64821309>
Reviewed by Megan Gardner.
Add an SPI method declaration.
* Platform/spi/ios/UIKitSPI.h:
2020-06-30 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Remove whitelistedRpId
https://bugs.webkit.org/show_bug.cgi?id=213817
<rdar://problem/60108131>
Reviewed by Geoffrey Garen.
Remove the whitelist such that we can test the attestation service in a wider range.
* UIProcess/WebAuthentication/Cocoa/LocalAuthenticator.mm:
(WebKit::LocalAuthenticator::continueMakeCredentialAfterUserVerification):
(WebKit::LocalAuthenticatorInternal::whitelistedRpId): Deleted.
2020-06-30 Brady Eidson <beidson@apple.com>
Fix Swift overlay build after r263727.
<rdar://problem/64962370> and https://bugs.webkit.org/show_bug.cgi?id=213823
Reviewed by Darin Adler.
No new tests (No behavior change).
Adapt the Swift overlay refinemnt to handle the inFrame: argument for these methods.
* UIProcess/API/Cocoa/WebKitSwiftOverlay.swift:
(callAsyncJavaScript(_:arguments:in:in:completionHandler:Error:)):
(evaluateJavaScript(_:in:in:completionHandler:Error:)):
(callAsyncJavaScript(_:arguments:in:completionHandler:Error:)): Deleted.
(evaluateJavaScript(_:in:completionHandler:Error:)): Deleted.
2020-06-30 Brady Eidson <beidson@apple.com>
App-bound JavaScript and Navigation failures should have specific error codes.
<rdar://problem/64940268> and https://bugs.webkit.org/show_bug.cgi?id=213808
Reviewed by Tim Hatcher.
(Informally by Kate Cheney)
* UIProcess/API/Cocoa/WKError.h:
* UIProcess/API/Cocoa/WKError.mm:
(localizedDescriptionForErrorCode):
* UIProcess/API/Cocoa/WKWebView.mm:
(nsErrorFromExceptionDetails):
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::errorForUnpermittedAppBoundDomainNavigation):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::decidePolicyForNavigationAction):
* UIProcess/WebPageProxy.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::runJavaScript):
2020-06-30 Per Arne Vollan <pvollan@apple.com>
[macOS] Connections to the preference daemon are established before entering the sandbox
https://bugs.webkit.org/show_bug.cgi?id=213379
Reviewed by Darin Adler.
On macOS, connections to the preference daemon are established before entering the sandbox. These connections also persist
after entering the sandbox and denying access to the preference daemon. There should not be attempts to connect to the
preference daemon before entering the sandbox, since these attempts will not be stopped by the sandbox. This patch moves
code that connects to the preference daemon to be executed after the sandbox has been entered. That includes code to
prevent connections to the Dock and code to initialize WebKit logging. Also, instead of calling [NSBundle bundleForClass:],
call [NSBundle bundleWithIdentifier:], since calling [NSBundle bundleForClass:] will connect to the preference daemon.
Finally, allow the syscall SYS_gethostuuid, since that is needed by CoreFoundation when there is no access to the
preference daemon.
No new tests. This should be covered by existing tests. It would be nice to have a test to make sure that there are no
connections to the preference daemon just before entering the sandbox, but I am not aware of how to implement this.
* NetworkProcess/mac/NetworkProcessMac.mm:
(WebKit::NetworkProcess::initializeSandbox):
* Shared/AuxiliaryProcess.cpp:
(WebKit::AuxiliaryProcess::initialize):
* Shared/Cocoa/WebKit2InitializeCocoa.mm:
(WebKit::runInitializationCode):
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceMain.mm:
(WebKit::XPCServiceMain):
* Shared/mac/AuxiliaryProcessMac.mm:
(WebKit::webKit2Bundle):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
(WebKit::WebProcess::initializeSandbox):
* WebProcess/com.apple.WebProcess.sb.in:
2020-06-30 Brent Fulgham <bfulgham@apple.com>
Partially revert r257135 now that the underlying bug is resolved
https://bugs.webkit.org/show_bug.cgi?id=213814
<rdar://problem/59859573>
Reviewed by Per Arne Vollan.
This patch is a partial revert of the code change in Bug 208033, where a temporary
workaround was added for a bug in an underlying framework. Now that the bug has been
fixed, we can remove this workaround.
* UIProcess/GPU/GPUProcessProxy.cpp:
(WebKit::gpuProcessSessionParameters):
2020-06-30 Brent Fulgham <bfulgham@apple.com>
[macOS] Stop blocking two audio-related user client classes
https://bugs.webkit.org/show_bug.cgi?id=213807
<rdar://problem/64874902>
Reviewed by Per Arne Vollan.
Restore IOAudioControlUserClient and IOAudioEngineUserClient since we find them being
actively used on macOS for some media playback purposes.
* WebProcess/com.apple.WebProcess.sb.in:
2020-06-30 Peng Liu <peng.liu6@apple.com>
Enable the support of FULLSCREEN_API in WebKitTestRunner
https://bugs.webkit.org/show_bug.cgi?id=213774
Reviewed by Youenn Fablet.
Replace the definition of ENABLE_FULLSCREEN_API in FeatureDefines.xcconfig with
the one in PlatformEnableCocoa.h. We have to do that because WebKitTestRunner
does not have a FeatureDefines.xcconfig but it uses "ENABLE(FULLSCREEN_API)"
to conditionally compile code to test the element fullscreen API.
WebKitTestRunner can use the macro defined in PlatformEnableCocoa.h.
* Configurations/FeatureDefines.xcconfig:
2020-06-30 Daniel Bates <dabates@apple.com>
[iOS] Editable regions causes ~1% slowdown in PLT5
https://bugs.webkit.org/show_bug.cgi?id=213659
<rdar://problem/64361390>
Reviewed by Simon Fraser.
Fix up RemoteLayerTreeViews now that the editable region is an Optional<>. Have
the UI process message the web process to enable editable region on the first
invocation of _requestTextInputContextsInRect, which is good indicator that there
would be benefit to computing it as a typical client that calls it will call it
many times.
* UIProcess/RemoteLayerTree/ios/RemoteLayerTreeViews.mm:
(WebKit::mayContainEditableElementsInRect): Fix up the logic now that there may
be not be an editable region sent over in the last event region update. If there
isn't one then we don't know if there are editable elements in the rect of not.
So, return true, which could turn out to be a false positive if there aren't any
editable elements in the rect. The caller will have to message the web process
to find out the real answer if they want it. Just to clarify, it's OK for this
function to have false positives, but it must never have false negatives.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::textInputContextsInRect): Enable editable region. If it's already
enabled then doing so again does nothing.
2020-06-30 Peng Liu <peng.liu6@apple.com>
Scrunching a video to PiP can result in broken animation and leave Safari in a bad state
https://bugs.webkit.org/show_bug.cgi?id=213175
Reviewed by Jer Noble.
* UIProcess/ios/fullscreen/WKFullScreenWindowControllerIOS.mm:
(-[WKFullScreenWindowController _completedExitFullScreen]):
Update _exitRequested after exiting fullscreen to make sure the following
enter fullscreen request can be processed.
(-[WKFullScreenWindowController _dismissFullscreenViewController]):
Make sure _completedExitFullScreen function will always execute.
(-[WKFullScreenWindowController _interactivePinchDismissChanged:]):
Remove a function call which corrupts the state machine under stress test.
2020-06-30 Antoine Quint <graouts@webkit.org>
[iOS] Crash under WebKit::WebPage::getFocusedElementInformation()
https://bugs.webkit.org/show_bug.cgi?id=213798
<rdar://problem/62754594>
Reviewed by Wenson Hsieh.
Since LabelableElement::labels() returns a RefPtr<NodeList>, ensure we get a value before attempting to read data from it.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::getFocusedElementInformation):
2020-06-30 Andy Estes <aestes@apple.com>
[Xcode] Enable the "My Mac (Mac Catalyst)" destination in WebKit Xcode projects
https://bugs.webkit.org/show_bug.cgi?id=213740
Reviewed by Darin Adler.
* Configurations/Base.xcconfig: Set SUPPORTS_MACCATALYST to YES to tell Xcode that this
project supports building for Mac Catalyst.
2020-06-30 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r263724.
https://bugs.webkit.org/show_bug.cgi?id=213781
Reverted changeset:
"Make _WKWebsiteDataStoreConfiguration SPI for HSTS storage to
replace _WKProcessPoolConfiguration.hstsStorageDirectory"
https://bugs.webkit.org/show_bug.cgi?id=213048
https://trac.webkit.org/changeset/263724
2020-06-30 Youenn Fablet <youenn@apple.com>
Add VP9 WebRTC codec runtime flag
https://bugs.webkit.org/show_bug.cgi?id=213724
Reviewed by Eric Carlson.
* Shared/WebPreferences.yaml:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences):
2020-06-29 Brady Eidson <beidson@apple.com>
JavaScript cannot be injected into iframes
<rdar://problem/54168946> and https://bugs.webkit.org/show_bug.cgi?id=213556
Reviewed by Geoff Garen.
This adds a few mechanisms:
- Allows for WKUserScripts to have a target content world
- Allows "evaluateJavaScript" and "callAsyncJavaScript" to target a specific frame
- Allows for the completion handlers of those methods to be able to distinguish
failure-due-to-missing-frame.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<ExceptionDetails>::encode):
(IPC::ArgumentCoder<ExceptionDetails>::decode):
* Shared/WebCoreArgumentCoders.h:
* UIProcess/API/Cocoa/WKError.h:
* UIProcess/API/Cocoa/WKError.mm:
(localizedDescriptionForErrorCode):
* UIProcess/API/Cocoa/WKUserScript.h:
* UIProcess/API/Cocoa/WKUserScript.mm:
(-[WKUserScript initWithSource:injectionTime:forMainFrameOnly:]):
(-[WKUserScript initWithSource:injectionTime:forMainFrameOnly:inContentWorld:]):
* UIProcess/API/Cocoa/WKWebView.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView evaluateJavaScript:inFrame:inContentWorld:completionHandler:]):
(-[WKWebView callAsyncJavaScript:arguments:inFrame:inContentWorld:completionHandler:]):
(nsErrorFromExceptionDetails):
(-[WKWebView _evaluateJavaScript:asAsyncFunction:withSourceURL:withArguments:forceUserGesture:inFrame:inWorld:completionHandler:]):
(-[WKWebView callAsyncJavaScript:arguments:inContentWorld:completionHandler:]): Deleted.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::runJavaScript):
2020-06-29 Alex Christensen <achristensen@webkit.org>
Make _WKWebsiteDataStoreConfiguration SPI for HSTS storage to replace _WKProcessPoolConfiguration.hstsStorageDirectory
https://bugs.webkit.org/show_bug.cgi?id=213048
Reviewed by Youenn Fablet.
This uses CFNetwork SPI introduced in rdar://problem/50109631 to allow HSTS storage per NSURLSession.
To be complete, I also deprecated our UI process HSTS state removal attempt SPIs, WKContextResetHSTSHosts and
WKContextResetHSTSHostsAddedAfterDate, which had their last use removed in rdar://problem/64220838.
I manually verified that this new SPI puts HSTS data in the specified location, and I also verified that HSTS
state querying and removal works with the new CFNetwork SPI as it did with the old one.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::fetchWebsiteData):
(WebKit::NetworkProcess::deleteWebsiteData):
(WebKit::NetworkProcess::deleteWebsiteDataForOrigins):
(WebKit::NetworkProcess::deleteAndRestrictWebsiteDataForRegistrableDomains):
(WebKit::NetworkProcess::registrableDomainsWithWebsiteData):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkSessionCreationParameters.cpp:
(WebKit::NetworkSessionCreationParameters::encode const):
(WebKit::NetworkSessionCreationParameters::decode):
* NetworkProcess/NetworkSessionCreationParameters.h:
* NetworkProcess/cocoa/NetworkProcessCocoa.mm:
(WebKit::NetworkProcess::hostNamesWithHSTSCache const):
(WebKit::NetworkProcess::deleteHSTSCacheForHostNames):
(WebKit::NetworkProcess::clearHSTSCache):
(WebKit::NetworkProcess::getHostNamesWithHSTSCache): Deleted.
* NetworkProcess/cocoa/NetworkSessionCocoa.h:
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::hstsStorage const):
(WebKit::NetworkSessionCocoa::NetworkSessionCocoa):
* NetworkProcess/soup/NetworkProcessSoup.cpp:
(WebKit::NetworkProcess::hostNamesWithHSTSCache const):
(WebKit::NetworkProcess::deleteHSTSCacheForHostNames):
(WebKit::NetworkProcess::clearHSTSCache):
(WebKit::NetworkProcess::getHostNamesWithHSTSCache): Deleted.
* UIProcess/API/C/mac/WKContextPrivateMac.h:
* UIProcess/API/C/mac/WKContextPrivateMac.mm:
(WKContextResetHSTSHosts):
(WKContextResetHSTSHostsAddedAfterDate):
* UIProcess/API/Cocoa/_WKProcessPoolConfiguration.h:
* UIProcess/API/Cocoa/_WKWebsiteDataStoreConfiguration.h:
* UIProcess/API/Cocoa/_WKWebsiteDataStoreConfiguration.mm:
(-[_WKWebsiteDataStoreConfiguration hstsStorageDirectory]):
(-[_WKWebsiteDataStoreConfiguration setHSTSStorageDirectory:]):
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::privateBrowsingSession): Deleted.
(WebKit::WebProcessPool::resetHSTSHosts): Deleted.
(WebKit::WebProcessPool::resetHSTSHostsAddedAfterDate): Deleted.
* UIProcess/WebProcessPool.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::resolveDirectoriesIfNecessary):
(WebKit::WebsiteDataStore::parameters):
* UIProcess/WebsiteData/WebsiteDataStore.h:
(WebKit::WebsiteDataStore::resolvedHSTSStorageDirectory const):
* UIProcess/WebsiteData/WebsiteDataStoreConfiguration.cpp:
(WebKit::WebsiteDataStoreConfiguration::copy const):
* UIProcess/WebsiteData/WebsiteDataStoreConfiguration.h:
2020-06-29 Tim Horton <timothy_horton@apple.com>
Adopt adjusted symbol image names
https://bugs.webkit.org/show_bug.cgi?id=213768
<rdar://problem/64472257>
Reviewed by Andy Estes.
* UIProcess/API/Cocoa/_WKElementAction.mm:
(+[_WKElementAction imageForElementActionType:]):
2020-06-29 Guowei Yang <guowei_yang@apple.com>
Adding Experimental Feature Flags for CoreImage backed SVG/CSS Filters
https://bugs.webkit.org/show_bug.cgi?id=213578
Reviewed by Darin Adler, Simon Fraser, Myles C. Maxfield.
Preparing to implement CoreImage backed filter rendering
Needs Compiler guards and experimental feature guard.
* Shared/WebPreferences.yaml: added feature flag definition
2020-06-29 Peng Liu <peng.liu6@apple.com>
Video spills over PiP screen a little when using Picture in Picture
https://bugs.webkit.org/show_bug.cgi?id=213658
Reviewed by Eric Carlson.
Add the video content dimensions to the IPC message VideoFullscreenManagerProxy::SetupFullscreenWithID.
* UIProcess/Cocoa/VideoFullscreenManagerProxy.h:
* UIProcess/Cocoa/VideoFullscreenManagerProxy.messages.in:
* UIProcess/Cocoa/VideoFullscreenManagerProxy.mm:
(WebKit::VideoFullscreenManagerProxy::setupFullscreenWithID):
(WebKit::VideoFullscreenManagerProxy::setVideoDimensions):
(WebKit::VideoFullscreenManagerProxy::enterFullscreen):
* WebProcess/cocoa/VideoFullscreenManager.mm:
(WebKit::VideoFullscreenManager::enterVideoFullscreenForVideoElement):
2020-06-29 Chris Dumez <cdumez@apple.com>
Unreviewed, reverting r262004.
<rdar://problem/64540215>
Revert r262004 as it is not OK to have a data vault in
TMPDIR.
Reverted changeset:
"[WK2] WebKit abandons compiled sandbox profiles"
https://bugs.webkit.org/show_bug.cgi?id=212177
https://trac.webkit.org/changeset/262004
2020-06-29 Tetsuharu Ohzeki <tetsuharu.ohzeki@gmail.com>
Remove ENABLE_STREAMS_API compilation flag
https://bugs.webkit.org/show_bug.cgi?id=213728
Reviewed by Sam Weinig.
* Configurations/FeatureDefines.xcconfig:
* Shared/WebPreferences.yaml:
* WebProcess/InjectedBundle/InjectedBundle.cpp:
(WebKit::InjectedBundle::overrideBoolPreferenceForTestRunner):
2020-06-29 Per Arne Vollan <pvollan@apple.com>
[iOS] Temporarily open up sandbox for mapping of Launch Services database
https://bugs.webkit.org/show_bug.cgi?id=213756
<rdar://problem/64707000>
Reviewed by Brent Fulgham.
Address review comments.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-06-29 Per Arne Vollan <pvollan@apple.com>
[iOS] Temporarily open up sandbox for mapping of Launch Services database
https://bugs.webkit.org/show_bug.cgi?id=213756
<rdar://problem/64707000>
Reviewed by Brent Fulgham.
Temporarily open up sandbox for mapping of Launch Services database in the WebContent process on iOS.
This is just a short term fix, and access to the service will be closed again soon.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-06-29 Chris Dumez <cdumez@apple.com>
Unreviewed, rebaseline messages.py tests.
* Scripts/MessageNames.cpp:
(IPC::description):
(IPC::receiverName):
(IPC::isValidMessageName):
* Scripts/MessageNames.h:
(WTF::isValidEnum):
* Scripts/test-superclassMessagesReplies.h:
2020-06-29 David Kilzer <ddkilzer@apple.com>
Crash in NetworkProcessProxy::getNetworkProcessConnection() lambda due to missing Optional<> value check
<https://webkit.org/b/213700>
<rdar://problem/64852903>
Reviewed by Darin Adler.
* UIProcess/GPU/GPUProcessProxy.cpp:
(WebKit::GPUProcessProxy::getGPUProcessConnection):
- Rename `connectionIdentifier` to `identifier`.
- Return early from the lamba if `identifier` does not contain a
value.
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::getNetworkProcessConnection):
- Ditto.
2020-06-29 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] editing/selection/ios/select-text-after-changing-focus.html sometimes fails
https://bugs.webkit.org/show_bug.cgi?id=213745
Work towards <rdar://problem/64808138>
Reviewed by Tim Horton.
This test first taps an input field to focus it, taps a button below the input field to focus the button and
dismiss the keyboard, and finally long presses to select a word below the button. On recent iOS 14 builds, this
test occasionally fails due to recent changes in UIKit that may cause the callout bar appearance callback to
fire after focusing the input field. If this happens, we end up showing the callout bar with a single option to
"Select All" (despite the field being empty), and the subsequent tap that's intended to hit the button instead
hit-tests to this callout bar item. The tests subsequently times out waiting for the keyboard to dismiss, which
never happens because the input field remains focused.
Showing the "Select All" callout bar option in empty text fields is inconsistent with the rest of the platform
(iOS), and appears to have been unintentionally introduced in iOS 12 by <https://trac.webkit.org/r231726>. While
it's still unknown which exact system changes caused this test to begin timing out, we can at least fix it by
addressing this existing regression in behavior.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView canPerformActionForWebView:withSender:]):
Check `-hasContent` here, instead of just checking if the selection is a caret. Note that we don't need to check
whether the selection is none separately, since `-hasContent` will always return `NO` if there is no selection.
2020-06-29 Sam Weinig <weinig@apple.com>
Remove remaining makeSimpleColorFrom* variants
https://bugs.webkit.org/show_bug.cgi?id=213706
Reviewed by Darin Adler.
* UIProcess/API/wpe/WebKitColor.cpp:
(webkitColorToWebCoreColor):
Use makeSimpleColor(SRGBA { ... }) rather than makeSimpleColorFromFloats.
* WebProcess/Inspector/WebInspectorClient.cpp:
(WebKit::WebInspectorClient::showPaintRect):
* WebProcess/WebPage/FindController.cpp:
(WebKit::FindController::drawRect):
Use makeSimpleColor rather than makeSimpleColorFromFloats, allowing for
constant expression creation of SimpleColors.
2020-06-29 Youenn Fablet <youenn@apple.com>
Send WebRTC packets received from network process to web process from a background thread
https://bugs.webkit.org/show_bug.cgi?id=213548
Reviewed by Eric Carlson.
Make LibWebRTCSocketClient take a ref to an IPC connection at creation time.
Use this connection from the RTC network thread to send received packets to improve performances.
Covered by existing tests.
* NetworkProcess/webrtc/LibWebRTCSocketClient.cpp:
(WebKit::LibWebRTCSocketClient::LibWebRTCSocketClient):
(WebKit::LibWebRTCSocketClient::signalReadPacket):
(WebKit::LibWebRTCSocketClient::signalSentPacket):
(WebKit::LibWebRTCSocketClient::signalAddressReady):
(WebKit::LibWebRTCSocketClient::signalConnect):
(WebKit::LibWebRTCSocketClient::signalClose):
* NetworkProcess/webrtc/LibWebRTCSocketClient.h:
* NetworkProcess/webrtc/NetworkRTCProvider.cpp:
(WebKit::NetworkRTCProvider::createSocket):
(WebKit::NetworkRTCProvider::createUDPSocket):
(WebKit::NetworkRTCProvider::createServerTCPSocket):
(WebKit::NetworkRTCProvider::createClientTCPSocket):
(WebKit::NetworkRTCProvider::wrapNewTCPConnection):
* NetworkProcess/webrtc/NetworkRTCProvider.h:
2020-06-29 Elliot <CheeseEBoi@mailo.com>
[GTK] Dark mode for GTK themes that end with -Dark
https://bugs.webkit.org/show_bug.cgi?id=213465
Reviewed by Carlos Garcia Campos.
WebKitGtk has support for automatic detection of dark mode when it comes to GTK themes with the "-dark" and
":dark" suffixes. However, when using GTK themes that end with "-Dark" or ":Dark", this is not the case. This
affects many themes like "Arc" and "Flat-Remix".
* UIProcess/API/gtk/PageClientImpl.cpp:
(WebKit::PageClientImpl::effectiveAppearanceIsDark const):
(WebKit::PageClientImpl::themeName const):
2020-06-29 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK][WPE] Add webkit_authentication_request_get_security_origin
https://bugs.webkit.org/show_bug.cgi?id=213596
Reviewed by Michael Catanzaro.
Returns the security origin for the authentication protection space.
* UIProcess/API/glib/WebKitAuthenticationRequest.cpp:
(webkit_authentication_request_get_security_origin):
* UIProcess/API/gtk/WebKitAuthenticationRequest.h:
* UIProcess/API/gtk/docs/webkit2gtk-4.0-sections.txt:
* UIProcess/API/wpe/WebKitAuthenticationRequest.h:
* UIProcess/API/wpe/docs/wpe-1.0-sections.txt:
2020-06-29 Alexander Mikhaylenko <alexm@gnome.org>
[GTK4] Context menu is misaligned
https://bugs.webkit.org/show_bug.cgi?id=213703
Reviewed by Carlos Garcia Campos.
Set halign on the context menu popover to the same value other context menus in GTK use,
like in GtkEntry.
* UIProcess/gtk/WebContextMenuProxyGtk.cpp:
(WebKit::createMenuWidget): Set halign for the context menu to 'start'.
2020-06-28 Chris Dumez <cdumez@apple.com>
Fix thread-safety issue in webProcessPoolHighDynamicRangeDidChangeCallback()
https://bugs.webkit.org/show_bug.cgi?id=213689
<rdar://problem/64443996>
Reviewed by Darin Adler.
webProcessPoolHighDynamicRangeDidChangeCallback() gets called on a background thread so we
need to make sure we dispatch to the main thread before we iterate over the process pools.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::webProcessPoolHighDynamicRangeDidChangeCallback):
2020-06-28 Geoffrey Garen <ggaren@apple.com>
Rename initializeThreading to initialize
https://bugs.webkit.org/show_bug.cgi?id=213674
Reviewed by Mark Lam.
* Shared/API/c/WKString.cpp:
(WKStringCopyJSString):
* Shared/Cocoa/WebKit2InitializeCocoa.mm:
(WebKit::runInitializationCode):
* Shared/WebKit2Initialize.cpp:
(WebKit::InitializeWebKit2):
2020-06-28 Daniel Bates <dabates@apple.com>
[iOS] Clean up visible position comparisons in WebPage::requestDocumentEditingContext()
https://bugs.webkit.org/show_bug.cgi?id=213701
Reviewed by Geoffrey Garen.
Use std::min and std::max when the result needs to be copied instead of doing the comparison
by hand to make the code a tiny bit more clear. Also, don't copy a VisiblePosition when doing
the min/max in-place. Both of these things are unlikely to effect code generation. If they do,
the latter makes things a tiny bit more efficient.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::requestDocumentEditingContext):
2020-06-28 Youenn Fablet <youenn@apple.com>
MediaRecorder stopRecorder() returns empty Blob after first use
https://bugs.webkit.org/show_bug.cgi?id=212274
<rdar://problem/63601298>
Reviewed by Eric Carlson.
Update implementation to do initialization as part of startRecording.
* GPUProcess/webrtc/RemoteMediaRecorderManager.cpp:
(WebKit::RemoteMediaRecorderManager::releaseRecorder):
Remove ASSERT as recorder creation in WebProcess is always ok while creation in GPUProcess may fail and m_recorders may not be populated.
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
(WebKit::MediaRecorderPrivate::MediaRecorderPrivate):
(WebKit::MediaRecorderPrivate::startRecording):
* WebProcess/GPU/webrtc/MediaRecorderPrivate.h:
2020-06-28 Philippe Normand <pnormand@igalia.com>
Unreviewed GTK/WPE Debug build fix after r263589.
* Platform/Logging.h: Define a log category for Memory Pressure.
2020-06-27 Daniel Bates <dabates@apple.com>
Revert r262654 because it caused <rdar://problem/64664156>.
Moreover it is no longer necessary as the embedding client
changes selection through a different SPI after focusing.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView cleanUpInteraction]):
(-[WKContentView _didCommitLoadForMainFrame]):
(-[WKContentView _focusTextInputContext:placeCaretAt:completionHandler:]):
2020-06-26 David Kilzer <ddkilzer@apple.com>
Crash at WebKit: __62-[WKContentView(WKInteraction) resignFirstResponderForWebView]_block_invoke
<https://webkit.org/b/213666>
<rdar://problem/64238661>
Reviewed by Darin Adler.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView resignFirstResponderForWebView]):
- Speculative fix to return early if there are no queued key
events available by the time the block runs.
2020-06-26 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] Layout tests sometimes assert under -[WKContentView hasContent]
https://bugs.webkit.org/show_bug.cgi?id=213656
<rdar://problem/64818648>
Reviewed by Tim Horton.
Make it safe to invoke `-[WKContentView hasContent]` when post-layout data is not present, and there is no
ranged or caret selection. Some editing layout tests on iOS tend to hit the post-layout data assertion due to
the callout bar presentation callback firing while we reset the test harness to a consistent state (and spin the
runloop while doing so).
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView hasContent]):
2020-06-26 Pablo Saavedra <psaavedra@igalia.com>
[GTK][WPE] Fix the matching of an empty value in getCgroupControllerPath() when only cgroupsV2 hierarchy is found
https://bugs.webkit.org/show_bug.cgi?id=213646
Reviewed by Adrian Perez de Castro.
* UIProcess/linux/MemoryPressureMonitor.cpp:
(WebKit::getCgroupFile):
(WebKit::getCgroupControllerPath):
(WebKit::systemMemoryUsedAsPercentage):
2020-06-26 Jason Lawrence <lawrence.j@apple.com>
Unreviewed, reverting r263511, r263514, and r263565.
r263511 caused MediaRecorder test crashes on internal testers.
Reverted changesets:
"MediaRecorder stopRecorder() returns empty Blob after first
use"
https://bugs.webkit.org/show_bug.cgi?id=212274
https://trac.webkit.org/changeset/263511
"Unreviewed iOS build fix after r263511."
https://trac.webkit.org/changeset/263514
"MediaRecorder.start() Method is Ignoring the "timeslice"
Parameter"
https://bugs.webkit.org/show_bug.cgi?id=202233
https://trac.webkit.org/changeset/263565
2020-06-26 Geoffrey Garen <ggaren@apple.com>
Initializing the main thread should initialize the main run loop
https://bugs.webkit.org/show_bug.cgi?id=213637
Reviewed by Anders Carlsson.
* Shared/Cocoa/WebKit2InitializeCocoa.mm:
(WebKit::runInitializationCode):
(WebKit::InitializeWebKit2):
* Shared/WebKit2Initialize.cpp:
(WebKit::InitializeWebKit2):
2020-06-26 Chris Dumez <cdumez@apple.com>
[iOS] Network process is crashing when launching TJMaxx app due to invalid NetworkProcess::DestroySession IPC message
https://bugs.webkit.org/show_bug.cgi?id=213625
<rdar://problem/64737890>
Reviewed by Alex Christensen.
The app is calling [WKWebsiteDataStore init] despite the method being marked as unavailable in
WKWebsiteDataStore.h. As a result, they end up with a WKWebsiteDataStore object whose internal
_websiteDataStore is bad because its constructor was never called. When [WKWebsiteDataStore dealloc]
gets called later own, it calls the ~WebsiteDataStore() destructor for _websiteDataStore but its
m_sessionID is 0 because we never called the constructor. This causes us to send a
NetworkProcess::DestroySession IPC with a sessionID that is 0, which is not valid so the
NetworkProcess crashes.
To address the issue, we now provide an implementation of [WKWebsiteDataStore init] which raises an
exception, behind a linked-on-after check. To keep the app working, [WKWebsiteDataStore init] returns
a new ephemeral data store until rebuilt with the new SDK.
* UIProcess/API/Cocoa/WKWebsiteDataStore.h:
Mark "new" as unavailable, otherwise [WKWebsiteDataStore new] builds.
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
(-[WKWebsiteDataStore init]):
Raise an exception with latest SDK, a new ephemeral data store otherwise.
* UIProcess/Cocoa/VersionChecks.h:
Add linked-on-after check.
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::~WebsiteDataStore):
Add a release assertion to make sure that m_sessionID is always valid when the destructor is called.
2020-06-26 Stephan Szabo <stephan.szabo@sony.com>
[WinCairo] Cannot build without resource load statistics
https://bugs.webkit.org/show_bug.cgi?id=213534
Reviewed by Don Olmstead.
Build fix for !ENABLE(RESOURCE_LOAD_STATISTICS)
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
2020-06-25 Alex Christensen <achristensen@webkit.org>
WKHTTPCookieStore.setCookie should flush the cookies to disk before calling completion handler when we are using the UIProcess's default cookie storage
https://bugs.webkit.org/show_bug.cgi?id=213636
Reviewed by Brady Eidson.
This fixes a race condition that many, many third party developers run into, causing them to not be logged in when they thought they did.
* NetworkProcess/cocoa/NetworkProcessCocoa.mm:
(WebKit::saveCookies):
* UIProcess/API/APIHTTPCookieStore.cpp:
(API::HTTPCookieStore::setCookies):
(API::HTTPCookieStore::registerForNewProcessPoolNotifications):
(API::HTTPCookieStore::flushDefaultUIProcessCookieStore):
* UIProcess/API/APIHTTPCookieStore.h:
* UIProcess/API/Cocoa/APIHTTPCookieStoreCocoa.mm:
(API::HTTPCookieStore::flushDefaultUIProcessCookieStore):
2020-06-25 Megan Gardner <megan_gardner@apple.com>
Upstream date/time picker style
https://bugs.webkit.org/show_bug.cgi?id=213630
Reviewed by Wenson Hsieh.
* UIProcess/ios/forms/WKDateTimeInputControl.mm:
(-[WKDateTimePicker datePickerStyle]):
2020-06-25 Chris Dumez <cdumez@apple.com>
Unreviewed, reverting r263545.
Patch will need refining
Reverted changeset:
"[iOS] Network process is crashing when launching TJMaxx app
due to invalid NetworkProcess::DestroySession IPC message"
https://bugs.webkit.org/show_bug.cgi?id=213625
https://trac.webkit.org/changeset/263545
2020-06-25 Chris Dumez <cdumez@apple.com>
[iOS] Network process is crashing when launching TJMaxx app due to invalid NetworkProcess::DestroySession IPC message
https://bugs.webkit.org/show_bug.cgi?id=213625
<rdar://problem/64737890>
Reviewed by Geoffrey Garen.
The app is calling [WKWebsiteDataStore init] despite the method being marked as unavailable in
WKWebsiteDataStore.h. As a result, they end up with a WKWebsiteDataStore object whose internal
_websiteDataStore is bad because its constructor was never called. When [WKWebsiteDataStore dealloc]
gets called later own, it calls the ~WebsiteDataStore() destructor for _websiteDataStore but its
m_sessionID is 0 because we never called the constructor. This causes us to send a
NetworkProcess::DestroySession IPC with a sessionID that is 0, which is not valid so the
NetworkProcess crashes.
To address the issue, we now provide an implementation of [WKWebsiteDataStore init] which returns nil
behind a linked-on-after check (since this crashes the app). To keep the app working,
[WKWebsiteDataStore init] returns the default data store until rebuilt with the new SDK. Note that
I tried returning a new ephemeral data store instead but this was getting the app in a bad state.
* UIProcess/API/Cocoa/WKWebsiteDataStore.h:
Mark "new" as unavailable, otherwise [WKWebsiteDataStore new] builds.
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
(-[WKWebsiteDataStore init]):
Return nil with latest SDK, the default data store otherwise.
* UIProcess/Cocoa/VersionChecks.h:
Add linked-on-after check.
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::~WebsiteDataStore):
Add a release assertion to make sure that m_sessionID is always valid when the destructor is called.
2020-06-25 Daniel Bates <dabates@apple.com>
[iOS] -_requestTextInputContextsInRect cannot find empty Quip spreadsheet title
https://bugs.webkit.org/show_bug.cgi?id=213564
<rdar://problem/59355847>
Reviewed by Simon Fraser.
Check if the search rect intersects the interaction rect, even if it's empty, so long as
there is an assisted element. If there isn't an assisted element just do what we do now.
Otherwise, if the interaction rect is empty then this means the focused element is an
empty inline and it's not tracked in the editable region (see WebCore ChangeLog entry for
more details). In this case, just ask the web process directly to find the contexts.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _requestTextInputContextsInRect:completionHandler:]):
2020-06-25 Kate Cheney <katherine_cheney@apple.com>
Allow service workers for web browsers
https://bugs.webkit.org/show_bug.cgi?id=213573
<rdar://problem/64712630>
Reviewed by Brent Fulgham.
Allows applications with the full web-browser entitlement to use
service workers.
* NetworkProcess/ios/NetworkProcessIOS.mm:
(WebKit::NetworkProcess::parentProcessHasServiceWorkerEntitlement const):
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _setupPageConfiguration:]):
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::parentProcessHasServiceWorkerEntitlement const):
2020-06-25 Youenn Fablet <youenn@apple.com>
MediaRecorder stopRecorder() returns empty Blob after first use
https://bugs.webkit.org/show_bug.cgi?id=212274
<rdar://problem/63601298>
Reviewed by Eric Carlson.
Update implementation to do initialization as part of startRecording.
* GPUProcess/webrtc/RemoteMediaRecorderManager.cpp:
(WebKit::RemoteMediaRecorderManager::releaseRecorder):
Remove ASSERT as recorder creation in WebProcess is always ok while creation in GPUProcess may fail and m_recorders may not be populated.
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
(WebKit::MediaRecorderPrivate::MediaRecorderPrivate):
(WebKit::MediaRecorderPrivate::startRecording):
* WebProcess/GPU/webrtc/MediaRecorderPrivate.h:
2020-06-24 James Savage <james.savage@apple.com>
Upstream iPadOS 13.0 multi-window support
https://bugs.webkit.org/show_bug.cgi?id=213590
Reviewed by Alexey Proskuryakov.
* UIProcess/ios/fullscreen/WKFullScreenWindowControllerIOS.mm:
(-[WKFullScreenWindowController enterFullScreen]): Always construct a
UIWindow using the expected UIWindowScene.
(makeWindowFromView): Deleted.
2020-06-24 David Kilzer <ddkilzer@apple.com>
Use ObjectIdentifier<>instead of WebCore::nextPlaybackTargetClientContextId() in Document.cpp
<https://webkit.org/b/213546>
<rdar://problem/61803576>
Reviewed by Youenn Fablet.
Switch from uint64_t to WebCore::PlaybackTargetClientContextIdentifier
for contextId values.
* Scripts/webkit/messages.py:
(types_that_cannot_be_forward_declared):
- Add WebCore::PlaybackTargetClientContextIdentifier to list.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::addPlaybackTargetPickerClient):
(WebKit::WebPageProxy::removePlaybackTargetPickerClient):
(WebKit::WebPageProxy::showPlaybackTargetPicker):
(WebKit::WebPageProxy::playbackTargetPickerClientStateDidChange):
(WebKit::WebPageProxy::setPlaybackTarget):
(WebKit::WebPageProxy::externalOutputDeviceAvailableDidChange):
(WebKit::WebPageProxy::setShouldPlayToPlaybackTarget):
(WebKit::WebPageProxy::playbackTargetPickerWasDismissed):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::addPlaybackTargetPickerClient):
(WebKit::WebChromeClient::removePlaybackTargetPickerClient):
(WebKit::WebChromeClient::showPlaybackTargetPicker):
(WebKit::WebChromeClient::playbackTargetPickerClientStateDidChange):
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::playbackTargetSelected const):
(WebKit::WebPage::playbackTargetAvailabilityDidChange):
(WebKit::WebPage::setShouldPlayToPlaybackTarget):
(WebKit::WebPage::playbackTargetPickerWasDismissed):
2020-06-24 David Kilzer <ddkilzer@apple.com>
[IPC] Use WebKit::LayerHostingContextID instead of uint32_t for videoLayerID
<https://webkit.org/b/213576>
<rdar://problem/61806253>
Reviewed by Darin Adler.
Use WebKit::LayerHostingContextID instead of uint32_t as type
for videoLayerID since the value comes from
LayerHostingContext::contextID() in
VideoFullscreenManager::enterVideoFullscreenForVideoElement().
* UIProcess/Cocoa/VideoFullscreenManagerProxy.h:
* UIProcess/Cocoa/VideoFullscreenManagerProxy.messages.in:
* UIProcess/Cocoa/VideoFullscreenManagerProxy.mm:
(WebKit::VideoFullscreenManagerProxy::setupFullscreenWithID):
2020-06-24 Tim Horton <timothy_horton@apple.com>
Cannot play back any video on Apple Silicon devices
https://bugs.webkit.org/show_bug.cgi?id=213584
<rdar://problem/64714596>
Reviewed by Per Arne Vollan.
* WebProcess/com.apple.WebProcess.sb.in:
An unfortunate and incorrect refactoring caused some of the sandbox exceptions to get dropped on the floor.
2020-06-24 Simon Fraser <simon.fraser@apple.com>
REGRESSION (r260276): Scrolling through shelves on music.apple.com is not smooth
https://bugs.webkit.org/show_bug.cgi?id=213572
<rdar://problem/64306590>
Reviewed by Wenson Hsieh.
The scroll position for an overflow:scroll element with scroll-snap could jump around
during scrolling, if layout triggered a call to ScrollableArea::updateScrollSnapState().
The crux of the issue is that isScrollSnapInProgress() returned false for overflow:scrollers
which are scrolling asynchronously.
Fix by extending the existing ScrollingTree::isScrollSnapInProgress() to track all scrolling
nodes by storing a HashSet<ScrollingNodeID> rather than just a flag for the main frame.
RenderLayer::isScrollSnapInProgress() then consults the ScrollingCoordinator, which consults
the scrolling tree for this state.
Add the ability to test via internals.isScrollSnapInProgress().
* WebProcess/WebPage/RemoteLayerTree/RemoteScrollingCoordinator.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteScrollingCoordinator.mm:
(WebKit::RemoteScrollingCoordinator::isScrollSnapInProgress const):
2020-06-24 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r263295.
https://bugs.webkit.org/show_bug.cgi?id=213577
Introduced logging regression
Reverted changeset:
"[macOS] Connections to the preference daemon are established
before entering the sandbox"
https://bugs.webkit.org/show_bug.cgi?id=213379
https://trac.webkit.org/changeset/263295
2020-06-24 David Kilzer <ddkilzer@apple.com>
Use ObjectIdentifier<> instead of uint64_t for context IDs in VideoFullscreenManagerProxy
<https://webkit.org/b/212392>
<rdar://problem/61799040>
Reviewed by Youenn Fablet.
Switch from uint64_t to WebKit::PlaybackSessionContextIdentifier
for contextId values.
* Scripts/webkit/messages.py:
(types_that_cannot_be_forward_declared):
- Add WebKit::PlaybackSessionContextIdentifier to list.
* UIProcess/Cocoa/PlaybackSessionManagerProxy.h:
(WebKit::PlaybackSessionManagerProxy::controlsManagerContextId const):
* UIProcess/Cocoa/PlaybackSessionManagerProxy.messages.in:
* UIProcess/Cocoa/PlaybackSessionManagerProxy.mm:
(WebKit::PlaybackSessionManagerProxy::createModelAndInterface):
(WebKit::PlaybackSessionManagerProxy::ensureModelAndInterface):
(WebKit::PlaybackSessionManagerProxy::ensureModel):
(WebKit::PlaybackSessionManagerProxy::ensureInterface):
(WebKit::PlaybackSessionManagerProxy::addClientForContext):
(WebKit::PlaybackSessionManagerProxy::removeClientForContext):
(WebKit::PlaybackSessionManagerProxy::setUpPlaybackControlsManagerWithID):
(WebKit::PlaybackSessionManagerProxy::clearPlaybackControlsManager):
(WebKit::PlaybackSessionManagerProxy::currentTimeChanged):
(WebKit::PlaybackSessionManagerProxy::bufferedTimeChanged):
(WebKit::PlaybackSessionManagerProxy::seekableRangesVectorChanged):
(WebKit::PlaybackSessionManagerProxy::canPlayFastReverseChanged):
(WebKit::PlaybackSessionManagerProxy::audioMediaSelectionOptionsChanged):
(WebKit::PlaybackSessionManagerProxy::legibleMediaSelectionOptionsChanged):
(WebKit::PlaybackSessionManagerProxy::audioMediaSelectionIndexChanged):
(WebKit::PlaybackSessionManagerProxy::legibleMediaSelectionIndexChanged):
(WebKit::PlaybackSessionManagerProxy::externalPlaybackPropertiesChanged):
(WebKit::PlaybackSessionManagerProxy::wirelessVideoPlaybackDisabledChanged):
(WebKit::PlaybackSessionManagerProxy::mutedChanged):
(WebKit::PlaybackSessionManagerProxy::volumeChanged):
(WebKit::PlaybackSessionManagerProxy::durationChanged):
(WebKit::PlaybackSessionManagerProxy::playbackStartedTimeChanged):
(WebKit::PlaybackSessionManagerProxy::rateChanged):
(WebKit::PlaybackSessionManagerProxy::pictureInPictureSupportedChanged):
(WebKit::PlaybackSessionManagerProxy::handleControlledElementIDResponse const):
(WebKit::PlaybackSessionManagerProxy::play):
(WebKit::PlaybackSessionManagerProxy::pause):
(WebKit::PlaybackSessionManagerProxy::togglePlayState):
(WebKit::PlaybackSessionManagerProxy::beginScrubbing):
(WebKit::PlaybackSessionManagerProxy::endScrubbing):
(WebKit::PlaybackSessionManagerProxy::seekToTime):
(WebKit::PlaybackSessionManagerProxy::fastSeek):
(WebKit::PlaybackSessionManagerProxy::beginScanningForward):
(WebKit::PlaybackSessionManagerProxy::beginScanningBackward):
(WebKit::PlaybackSessionManagerProxy::endScanning):
(WebKit::PlaybackSessionManagerProxy::selectAudioMediaOption):
(WebKit::PlaybackSessionManagerProxy::selectLegibleMediaOption):
(WebKit::PlaybackSessionManagerProxy::togglePictureInPicture):
(WebKit::PlaybackSessionManagerProxy::toggleMuted):
(WebKit::PlaybackSessionManagerProxy::setMuted):
(WebKit::PlaybackSessionManagerProxy::setVolume):
(WebKit::PlaybackSessionManagerProxy::setPlayingOnSecondScreen):
- Remove MESSAGE_CHECK_CONTEXTID() since it is now redundant.
* UIProcess/Cocoa/VideoFullscreenManagerProxy.h:
* UIProcess/Cocoa/VideoFullscreenManagerProxy.messages.in:
* UIProcess/Cocoa/VideoFullscreenManagerProxy.mm:
(WebKit::VideoFullscreenModelContext::VideoFullscreenModelContext):
(WebKit::VideoFullscreenManagerProxy::requestRouteSharingPolicyAndContextUID):
(WebKit::VideoFullscreenManagerProxy::createModelAndInterface):
(WebKit::VideoFullscreenManagerProxy::ensureModelAndInterface):
(WebKit::VideoFullscreenManagerProxy::ensureModel):
(WebKit::VideoFullscreenManagerProxy::ensureInterface):
(WebKit::VideoFullscreenManagerProxy::findInterface):
(WebKit::VideoFullscreenManagerProxy::addClientForContext):
(WebKit::VideoFullscreenManagerProxy::removeClientForContext):
(WebKit::VideoFullscreenManagerProxy::setupFullscreenWithID):
(WebKit::VideoFullscreenManagerProxy::setHasVideo):
(WebKit::VideoFullscreenManagerProxy::setVideoDimensions):
(WebKit::VideoFullscreenManagerProxy::enterFullscreen):
(WebKit::VideoFullscreenManagerProxy::exitFullscreen):
(WebKit::VideoFullscreenManagerProxy::exitFullscreenWithoutAnimationToMode):
(WebKit::VideoFullscreenManagerProxy::setInlineRect):
(WebKit::VideoFullscreenManagerProxy::setHasVideoContentLayer):
(WebKit::VideoFullscreenManagerProxy::cleanupFullscreen):
(WebKit::VideoFullscreenManagerProxy::preparedToReturnToInline):
(WebKit::VideoFullscreenManagerProxy::preparedToExitFullscreen):
(WebKit::VideoFullscreenManagerProxy::requestFullscreenMode):
(WebKit::VideoFullscreenManagerProxy::requestUpdateInlineRect):
(WebKit::VideoFullscreenManagerProxy::requestVideoContentLayer):
(WebKit::VideoFullscreenManagerProxy::returnVideoContentLayer):
(WebKit::VideoFullscreenManagerProxy::didSetupFullscreen):
(WebKit::VideoFullscreenManagerProxy::willExitFullscreen):
(WebKit::VideoFullscreenManagerProxy::didExitFullscreen):
(WebKit::VideoFullscreenManagerProxy::didEnterFullscreen):
(WebKit::VideoFullscreenManagerProxy::didCleanupFullscreen):
(WebKit::VideoFullscreenManagerProxy::setVideoLayerFrame):
(WebKit::VideoFullscreenManagerProxy::setVideoLayerGravity):
(WebKit::VideoFullscreenManagerProxy::fullscreenModeChanged):
(WebKit::VideoFullscreenManagerProxy::fullscreenMayReturnToInline):
- Remove MESSAGE_CHECK_CONTEXTID() since it is now redundant.
* WebKit.xcodeproj/project.pbxproj:
- Add PlaybackSessionContextIdentifier.h to the project.
* WebProcess/cocoa/PlaybackSessionContextIdentifier.h: Add.
* WebProcess/cocoa/PlaybackSessionManager.h:
* WebProcess/cocoa/PlaybackSessionManager.messages.in:
* WebProcess/cocoa/PlaybackSessionManager.mm:
(WebKit::PlaybackSessionInterfaceContext::PlaybackSessionInterfaceContext):
(WebKit::PlaybackSessionManager::createModelAndInterface):
(WebKit::PlaybackSessionManager::ensureModelAndInterface):
(WebKit::PlaybackSessionManager::ensureModel):
(WebKit::PlaybackSessionManager::ensureInterface):
(WebKit::PlaybackSessionManager::removeContext):
(WebKit::PlaybackSessionManager::addClientForContext):
(WebKit::PlaybackSessionManager::removeClientForContext):
(WebKit::PlaybackSessionManager::setUpPlaybackControlsManager):
(WebKit::PlaybackSessionManager::clearPlaybackControlsManager):
(WebKit::PlaybackSessionManager::contextIdForMediaElement):
(WebKit::PlaybackSessionManager::durationChanged):
(WebKit::PlaybackSessionManager::currentTimeChanged):
(WebKit::PlaybackSessionManager::bufferedTimeChanged):
(WebKit::PlaybackSessionManager::playbackStartedTimeChanged):
(WebKit::PlaybackSessionManager::rateChanged):
(WebKit::PlaybackSessionManager::seekableRangesChanged):
(WebKit::PlaybackSessionManager::canPlayFastReverseChanged):
(WebKit::PlaybackSessionManager::audioMediaSelectionOptionsChanged):
(WebKit::PlaybackSessionManager::legibleMediaSelectionOptionsChanged):
(WebKit::PlaybackSessionManager::externalPlaybackChanged):
(WebKit::PlaybackSessionManager::audioMediaSelectionIndexChanged):
(WebKit::PlaybackSessionManager::legibleMediaSelectionIndexChanged):
(WebKit::PlaybackSessionManager::wirelessVideoPlaybackDisabledChanged):
(WebKit::PlaybackSessionManager::mutedChanged):
(WebKit::PlaybackSessionManager::volumeChanged):
(WebKit::PlaybackSessionManager::isPictureInPictureSupportedChanged):
(WebKit::PlaybackSessionManager::play):
(WebKit::PlaybackSessionManager::pause):
(WebKit::PlaybackSessionManager::togglePlayState):
(WebKit::PlaybackSessionManager::beginScrubbing):
(WebKit::PlaybackSessionManager::endScrubbing):
(WebKit::PlaybackSessionManager::seekToTime):
(WebKit::PlaybackSessionManager::fastSeek):
(WebKit::PlaybackSessionManager::beginScanningForward):
(WebKit::PlaybackSessionManager::beginScanningBackward):
(WebKit::PlaybackSessionManager::endScanning):
(WebKit::PlaybackSessionManager::selectAudioMediaOption):
(WebKit::PlaybackSessionManager::selectLegibleMediaOption):
(WebKit::PlaybackSessionManager::handleControlledElementIDRequest):
(WebKit::PlaybackSessionManager::togglePictureInPicture):
(WebKit::PlaybackSessionManager::toggleMuted):
(WebKit::PlaybackSessionManager::setMuted):
(WebKit::PlaybackSessionManager::setVolume):
(WebKit::PlaybackSessionManager::setPlayingOnSecondScreen):
(WebKit::nextContextId): Delete.
- Replace with PlaybackSessionContextIdentifier::generate().
* WebProcess/cocoa/VideoFullscreenManager.h:
(WebKit::VideoFullscreenInterfaceContext::create):
* WebProcess/cocoa/VideoFullscreenManager.messages.in:
* WebProcess/cocoa/VideoFullscreenManager.mm:
(WebKit::VideoFullscreenInterfaceContext::VideoFullscreenInterfaceContext):
(WebKit::VideoFullscreenManager::createModelAndInterface):
(WebKit::VideoFullscreenManager::ensureModelAndInterface):
(WebKit::VideoFullscreenManager::ensureModel):
(WebKit::VideoFullscreenManager::ensureInterface):
(WebKit::VideoFullscreenManager::removeContext):
(WebKit::VideoFullscreenManager::addClientForContext):
(WebKit::VideoFullscreenManager::removeClientForContext):
(WebKit::VideoFullscreenManager::enterVideoFullscreenForVideoElement):
(WebKit::VideoFullscreenManager::exitVideoFullscreenForVideoElement):
(WebKit::VideoFullscreenManager::exitVideoFullscreenToModeWithoutAnimation):
(WebKit::VideoFullscreenManager::hasVideoChanged):
(WebKit::VideoFullscreenManager::videoDimensionsChanged):
(WebKit::VideoFullscreenManager::requestFullscreenMode):
(WebKit::VideoFullscreenManager::fullscreenModeChanged):
(WebKit::VideoFullscreenManager::requestUpdateInlineRect):
(WebKit::VideoFullscreenManager::requestVideoContentLayer):
(WebKit::VideoFullscreenManager::returnVideoContentLayer):
(WebKit::VideoFullscreenManager::didSetupFullscreen):
(WebKit::VideoFullscreenManager::willExitFullscreen):
(WebKit::VideoFullscreenManager::didEnterFullscreen):
(WebKit::VideoFullscreenManager::didExitFullscreen):
(WebKit::VideoFullscreenManager::didCleanupFullscreen):
(WebKit::VideoFullscreenManager::setVideoLayerGravityEnum):
(WebKit::VideoFullscreenManager::fullscreenMayReturnToInline):
(WebKit::VideoFullscreenManager::requestRouteSharingPolicyAndContextUID):
(WebKit::VideoFullscreenManager::setVideoLayerFrameFenced):
2020-06-24 Alex Christensen <achristensen@webkit.org>
REGRESSION(r262341) Use UTF-8 to decode CFURLRefs from IPC
https://bugs.webkit.org/show_bug.cgi?id=213565
<rdar://problem/64408924>
Reviewed by Darin Adler.
My assumption that all valid CFURLRefs are ASCII was incorrect.
Some of them are UTF-8 encoded non-ASCII.
Using kCFStringEncodingUTF8 covers both.
Test: fast/url/navigate-non-ascii.html
* Shared/cf/ArgumentCodersCF.cpp:
(IPC::decode):
2020-06-24 Umar Iqbal <uiqbal@apple.com>
We should resurrect the older patch that collects some statistics of web API calls
https://bugs.webkit.org/show_bug.cgi?id=213319
Reviewed by Brent Fulgham.
+ Enabled ENABLE_WEB_API_STATISTICS flag
+ Removed .string() from mainFrameRegistrableDomain to append it to topFrameRegistrableDomainsWhichAccessedWebAPIs
because topFrameRegistrableDomainsWhichAccessedWebAPIs is a HashSet of WebCore::RegistrableDomain
* Configurations/FeatureDefines.xcconfig:
* WebProcess/WebCoreSupport/WebResourceLoadObserver.cpp:
(WebKit::WebResourceLoadObserver::logFontLoad):
(WebKit::WebResourceLoadObserver::logCanvasRead):
(WebKit::WebResourceLoadObserver::logCanvasWriteOrMeasure):
(WebKit::WebResourceLoadObserver::logNavigatorAPIAccessed):
(WebKit::WebResourceLoadObserver::logScreenAPIAccessed):
2020-06-24 Andy Estes <aestes@apple.com>
[iOS] Open element actions sometimes result in a hover instead of a click
https://bugs.webkit.org/show_bug.cgi?id=213530
<rdar://problem/64176707>
Reviewed by Darin Adler.
_WKElementAction implements a standard action handler for _WKElementActionTypeOpen that
attempts a synthetic click at the interaction location, expecting this click to start a
navigation to the interaction node's target URL. Synthetic clicks are subject to content
change observing, however, and if ContentChangeObserver detects a meaningful visible change
after dispatching a mousemove event at the interaction location then WebPage will refrain
from dispatching the mousedown and mouseup events that would have started a navigation.
While this behavior is necessary to support hover menus, it doesn't make sense when the user
is interacting with a context menu. Since the system might display a preview of the target
URL found at the interaction location, users expect the "Open" action to navigate to that
same URL rather than merely display a hover menu.
This change opts synthetic clicks from the open action out of content change observing by
ensuring that an interaction is started when handling _WKElementActionTypeOpen (like we do
for "Copy" and "Save Image"), then checking if WebPage::m_interactionNode matches the node
found in WebPage::attemptSyntheticClick. If it does, WebPage now bypasses content change
observing and completes a synthetic click.
New API test: ElementActionTests.OpenLinkWithHoverMenu
* UIProcess/API/Cocoa/_WKElementAction.mm:
(+[_WKElementAction _elementActionWithType:customTitle:assistant:]): Changed to call
-[WKActionSheetAssistant handleElementActionWithType:element:needsInteraction:] in the
handlers for _WKElementActionTypeCopy, _WKElementActionTypeOpen,
_WKElementActionTypeSaveImage, and _WKElementActionTypeShare.
* UIProcess/ios/WKActionSheetAssistant.h:
* UIProcess/ios/WKActionSheetAssistant.mm:
(-[WKActionSheetAssistant handleElementActionWithType:element:needsInteraction:]): Added to
handle the subset of standard element actions that flow through
WKActionSheetAssistantDelegate. Used a boolean parameter to decide whether the delegate
should start and stop an interaction when handling the action.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _highlightLongPressRecognized:]):
(-[WKContentView actionSheetAssistant:openElementAtLocation:]): Renamed
-_attemptClickAtLocation:modifierFlags: to -_attemptSyntheticClickAtLocation:modifierFlags:.
(-[WKContentView _attemptSyntheticClickAtLocation:modifierFlags:]): Changed to call
WebPageProxy::attemptSyntheticClick instead of WebPageProxy::handleTap.
(-[WKContentView _simulateElementAction:atLocation:]): Set
_layerTreeTransactionIdAtLastInteractionStart to better simulate context menu interactions.
* UIProcess/WebPageProxy.h:
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::attemptSyntheticClick):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::attemptSyntheticClick): Renamed handleTap to attemptSyntheticClick. If the
node responding to clicks matches the node found on interaction start (m_interactionNode),
call WebPage::completeSyntheticClick directly.
2020-06-24 Tim Horton <timothy_horton@apple.com>
Fix the build once RBSProcessLimitations.h is introduced
https://bugs.webkit.org/show_bug.cgi?id=213562
Reviewed by Wenson Hsieh.
* Platform/spi/ios/RunningBoardServicesSPI.h:
We can't forward-declare Objective-C SPI if it actually exists
without inspiring duplicate definition errors.
2020-06-24 Alex Christensen <achristensen@webkit.org>
Remove now-invalid assertion after r263441
https://bugs.webkit.org/show_bug.cgi?id=213527
The alternative service storage directory is now null when HTTP/3 is disabled, not just in private browsing.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::NetworkSessionCocoa):
2020-06-24 Jer Noble <jer.noble@apple.com>
REGRESSION (r263328): Crash in bool WTF::WeakHashSet<WebKit::AudioSessionRoutingArbitratorProxy, WTF::EmptyCounter>::contains<WebKit::AudioSessionRoutingArbitratorProxy> const
https://bugs.webkit.org/show_bug.cgi?id=213526
<rdar://problem/64591777>
Reviewed by Eric Carlson.
Clear the m_routingArbitrator before WebProcessPool::disconnectProcess(), not after.
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::shutDown):
2020-06-24 Carlos Garcia Campos <cgarcia@igalia.com>
Unreviewed. Update OptionsGTK.cmake and NEWS for 2.29.2 release
* gtk/NEWS: Add release notes for 2.29.2.
2020-06-24 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK][WPE] Add API to allow applications to handle the HTTP authentication credential storage
https://bugs.webkit.org/show_bug.cgi?id=213177
Reviewed by Michael Catanzaro.
Add API to disable the internal credential storage support when building with libsecret. And add new API to
WebKitAuthenticationRequest to allow applications to use their own credentials storage:
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/NetworkSessionCreationParameters.cpp:
(WebKit::NetworkSessionCreationParameters::encode const):
(WebKit::NetworkSessionCreationParameters::decode):
* NetworkProcess/NetworkSessionCreationParameters.h:
* NetworkProcess/soup/NetworkDataTaskSoup.cpp:
(WebKit::NetworkDataTaskSoup::persistentCredentialStorageEnabled const):
(WebKit::NetworkDataTaskSoup::authenticate):
(WebKit::NetworkDataTaskSoup::continueAuthenticate):
(WebKit::NetworkDataTaskSoup::didGetHeaders):
* NetworkProcess/soup/NetworkDataTaskSoup.h:
* NetworkProcess/soup/NetworkProcessSoup.cpp:
(WebKit::NetworkProcess::setPersistentCredentialStorageEnabled):
* NetworkProcess/soup/NetworkSessionSoup.cpp:
(WebKit::NetworkSessionSoup::NetworkSessionSoup):
* NetworkProcess/soup/NetworkSessionSoup.h:
* UIProcess/API/glib/WebKitAuthenticationRequest.cpp:
(webkitAuthenticationRequestDispose):
(webkit_authentication_request_class_init):
(webkitAuthenticationRequestCreate):
(webkitAuthenticationRequestDidAuthenticate):
(webkitAuthenticationRequestGetProposedCredential):
(webkit_authentication_request_can_save_credentials):
(webkit_authentication_request_set_can_save_credentials):
(webkit_authentication_request_get_proposed_credential):
(webkit_authentication_request_set_proposed_credential):
(webkit_authentication_request_authenticate):
(webkit_authentication_request_cancel):
* UIProcess/API/glib/WebKitAuthenticationRequestPrivate.h:
* UIProcess/API/glib/WebKitWebView.cpp:
(webkitWebViewCompleteAuthenticationRequest):
(webkitWebViewLoadChanged):
(webkitWebViewLoadFailed):
(webkitWebViewLoadFailedWithTLSErrors):
(webkitWebViewHandleAuthenticationChallenge):
* UIProcess/API/glib/WebKitWebsiteDataManager.cpp:
(webkit_website_data_manager_set_persistent_credential_storage_enabled):
(webkit_website_data_manager_get_persistent_credential_storage_enabled):
* UIProcess/API/gtk/WebKitAuthenticationDialog.cpp:
(webkitAuthenticationDialogInitialize):
* UIProcess/API/gtk/WebKitAuthenticationRequest.h:
* UIProcess/API/gtk/WebKitWebsiteDataManager.h:
* UIProcess/API/gtk/docs/webkit2gtk-4.0-sections.txt:
* UIProcess/API/wpe/WebKitAuthenticationRequest.h:
* UIProcess/API/wpe/WebKitWebsiteDataManager.h:
* UIProcess/API/wpe/docs/wpe-1.0-sections.txt:
* UIProcess/WebsiteData/WebsiteDataStore.h:
(WebKit::WebsiteDataStore::persistentCredentialStorageEnabled const):
* UIProcess/WebsiteData/soup/WebsiteDataStoreSoup.cpp:
(WebKit::WebsiteDataStore::platformSetNetworkParameters):
(WebKit::WebsiteDataStore::setPersistentCredentialStorageEnabled):
* UIProcess/soup/WebProcessPoolSoup.cpp:
(WebKit::WebProcessPool::platformInitializeNetworkProcess):
2020-06-24 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK][WPE] Do not set the default configuration again when creating the WebsiteDataStore
https://bugs.webkit.org/show_bug.cgi?id=213340
Reviewed by Michael Catanzaro.
WebsiteDataStoreConfiguration already initializes on construction with the default paths, so we only need to
override the ones set by the user.
* UIProcess/API/glib/WebKitWebsiteDataManager.cpp:
(webkitWebsiteDataManagerGetDataStore):
2020-06-23 Alex Christensen <achristensen@webkit.org>
Make HTTP/3 experimental feature work on iOS and only create storage directory if enabled
https://bugs.webkit.org/show_bug.cgi?id=213527
Reviewed by Geoffrey Garen.
Reading the experimental feature wasn't working on iOS because the user defaults prefix is different.
Also, only resolve and create the directory if it is enabled. This should help with startup time, which
has to wait for all directories to be resolved and created. See rdar://problem/64263228 which this should help with.
We also respect the system default unless explicitly turned on, which should make it easier for CFNetwork to change things.
I added an API test that verifies that the default alt-svc storage directory is not created when we don't want it to be.
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::ensureNetworkProcess):
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::WebsiteDataStore::platformSetNetworkParameters):
(WebKit::WebsiteDataStore::http3Enabled):
2020-06-23 Wenson Hsieh <wenson_hsieh@apple.com>
REGRESSION (r244633): Mail flashes when copying text in an email
https://bugs.webkit.org/show_bug.cgi?id=213529
<rdar://problem/55025522>
Reviewed by Tim Horton.
When copying a text selection in Mail conversation view, Mail uses injected bundle hooks to supply additional
metadata before writing to the pasteboard. Grabbing this data involves temporarily expanding any collapsed
blockquotes that appear in the mail message, before re-collapsing them after we finish gathering data to be
written to the pasteboard.
Prior to r244633, the size of the document while expanding collapsed content would not be propagated to the
WKWebView, since intrinsic content size changes were batched and sent ahead of the next drawing area flush.
After r244633, however, updates are now sent every time the layout size changes. This was done to ensure that
the intrinsic content size was always up to date before sending the first layout milestone to the UI process.
To preserve the intent of r244633 (with respect to the first layout milestone) as well as intrinsic content size
behavior prior to r244633 after performing layout, we reintroduce the mechanism for batching intrinsic size
updates until the next flush, but keep the out-of-band mechanism for immediately sending an up-to-date intrinsic
size to the UI process before firing the first layout milestone.
Test: WebKit.AutoLayoutBatchesUpdatesWhenInvalidatingIntrinsicContentSize
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::intrinsicContentsSizeChanged const):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updateIntrinsicContentSizeIfNeeded):
If we have a pending intrinsic content size update, clear it out here so that an out-of-band call to
updateIntrinsicContentSizeIfNeeded doesn't get overridden by an older intrinsic content size that was previously
scheduled.
(WebKit::WebPage::flushPendingIntrinsicContentSizeUpdate):
Add a helper method to send any pending intrinsic content size update.
(WebKit::WebPage::scheduleIntrinsicContentSizeUpdate):
Add a helper method to schedule an intrinsic content size update for the next tiled drawing area flush.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::updateRendering):
2020-06-23 Andy Estes <aestes@apple.com>
[Apple Pay] WebPaymentCoordinatorProxy can be destroyed without dismissing its authorization presenter
https://bugs.webkit.org/show_bug.cgi?id=213423
<rdar://problem/60416221>
Reviewed by Youenn Fablet.
When a website completes a payment session, WebPaymentCoordinatorProxy would set its m_state
to Idle by calling didReachFinalState, but would wait to call hidePaymentUI until the
payment authorization presenter calls its presenterDidFinish callback. This allows the
presenter to display a final message to the user before dismissing. As a catch-all,
~WebPaymentCoordinatorProxy would call hidePaymentUI, but only if m_state wasn't Idle. If a
WebPaymentCoordinatorProxy was destroyed after a website completed a payment session but
before the presenter called presenterDidFinish, hidePaymentUI would never be called.
This patch addresses this by simplifying the logic for completing or canceling a payment
session. It's based on the following observations:
1. With the exception of the case described above, every caller of didReachFinalState would
also call hidePaymentUI immediately beforehand. That means that we can move the calls to
hidePaymentUI inside didReachFinalState. For the completePaymentSession case, rather than
call didReachFinalState without dismissing the UI, we enter a new Completing state, then
call didReachFinalState either in ~WebPaymentCoordinatorProxy or presenterDidFinish.
2. didCancelPaymentSession is really a special case of didReachFinalState that also sends
the DidCancelPaymentSession message to cancel active sessions in the web process. We can
teach didReachFinalState to send this message when necessary and remove
didCancelPaymentSession.
3. PaymentAuthorizationPresenter stored a "did reach final state" boolean on payment
completion and passed this value back to the client in the presenterDidFinish callback.
WebPaymentCoordinatorProxy used this to determine whether to cancel an active session by
calling didCancelPaymentSession or clean up a completed session by calling hidePaymentUI.
With the addition of the Completing state and the merging of didCancelPaymentSession and
hidePaymentUI into didReachFinalState, we can get rid of this final state tracking in
PaymentAuthorizationPresenter and unconditionally call didReachFinalState in the
presenterDidFinish callback.
No new tests possible. A manual test case is described in <rdar://problem/60416221>.
* Platform/cocoa/PaymentAuthorizationPresenter.h: Removed the didReachFinalState argument
from Client::presenterDidFinish.
* Platform/cocoa/PaymentAuthorizationPresenter.mm:
(WebKit::PaymentAuthorizationPresenter::completePaymentSession): Changed to call
-[WKPaymentAuthorizationDelegate completePaymentSession:errors:].
* Platform/cocoa/WKPaymentAuthorizationDelegate.h:
* Platform/cocoa/WKPaymentAuthorizationDelegate.mm: Removed the _didReachFinalState ivar.
(-[WKPaymentAuthorizationDelegate completePaymentSession:errors:]): Renamed from
-completePaymentSession:errors:didReachFinalState:. Stopped setting _didReachFinalState.
(-[WKPaymentAuthorizationDelegate invalidate]): Changed to call
-completePaymentSession:errors:.
(-[WKPaymentAuthorizationDelegate _didAuthorizePayment:completion:]): Ditto.
(-[WKPaymentAuthorizationDelegate _didFinish]): Stopped passing the value of
_didReachFinalState to PaymentAuthorizationPresenter::Client::presenterDidFinish.
(-[WKPaymentAuthorizationDelegate completePaymentSession:errors:didReachFinalState:]):
Renamed to -completePaymentSession:errors:.
* Shared/ApplePay/WebPaymentCoordinatorProxy.cpp:
(WebKit::WebPaymentCoordinatorProxy::showPaymentUI): Replaced calls to
didCancelPaymentSession with didReachFinalState.
(WebKit::WebPaymentCoordinatorProxy::completePaymentSession): Set m_state to Completing
rather than call didReachFinalState.
(WebKit::WebPaymentCoordinatorProxy::abortPaymentSession): Removed call to hidePaymentUI
since didReachFinalState was already being called.
(WebKit::WebPaymentCoordinatorProxy::cancelPaymentSession): Replaced call to
didCancelPaymentSession with didReachFinalState.
(WebKit::WebPaymentCoordinatorProxy::presenterDidFinish): Called didReachFinalState
unconditionally now that the didReachFinalState boolean has been removed.
(WebKit::WebPaymentCoordinatorProxy::canBegin const): Returned false for State::Completing.
(WebKit::WebPaymentCoordinatorProxy::canCancel const): Ditto.
(WebKit::WebPaymentCoordinatorProxy::canCompletePayment const): Ditto.
(WebKit::WebPaymentCoordinatorProxy::canAbort const): Ditto.
(WebKit::WebPaymentCoordinatorProxy::didReachFinalState): Added an early return if m_state
is already Idle. Sent the DidCancelPaymentSession message if canCancel is true. Called
platformHidePaymentUI before clearing m_authorizationPresenter and setting m_state to Idle.
(WebKit::WebPaymentCoordinatorProxy::didCancelPaymentSession): Deleted.
* Shared/ApplePay/WebPaymentCoordinatorProxy.h:
(WebKit::WebPaymentCoordinatorProxy::didReachFinalState): Added a PaymentSessionError&& arg.
(WebKit::WebPaymentCoordinatorProxy::didCancelPaymentSession): Deleted.
* Shared/ApplePay/cocoa/WebPaymentCoordinatorProxyCocoa.mm:
(WebKit::WebPaymentCoordinatorProxy::~WebPaymentCoordinatorProxy): Replaced call to
hidePaymentUI with didReachFinalState.
* Shared/ApplePay/ios/WebPaymentCoordinatorProxyIOS.mm:
(WebKit::WebPaymentCoordinatorProxy::platformHidePaymentUI): Renamed from hidePaymentUI.
Stopped clearing m_authorizationPresenter since didReachFinalState now does that.
(WebKit::WebPaymentCoordinatorProxy::hidePaymentUI): Renamed to platformHidePaymentUI.
* Shared/ApplePay/mac/WebPaymentCoordinatorProxyMac.mm:
(WebKit::WebPaymentCoordinatorProxy::platformShowPaymentUI): Removed call to hidePaymentUI
since didReachFinalState was already being called.
(WebKit::WebPaymentCoordinatorProxy::platformHidePaymentUI): Renamed from hidePaymentUI.
Stopped clearing m_authorizationPresenter since didReachFinalState now does that.
(WebKit::WebPaymentCoordinatorProxy::hidePaymentUI): Renamed to platformHidePaymentUI.
2020-06-23 Chris Dumez <cdumez@apple.com>
Remove a lot of unnecessary calls to Ref::copyRef()
https://bugs.webkit.org/show_bug.cgi?id=213533
Reviewed by Darin Adler.
Remove a lot of unnecessary calls to Ref::copyRef() now that Ref is copyable.
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
(WebKit::RemoteMediaPlayerProxy::requestResource):
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::CompletionHandler<void):
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:
(WebKit::ResourceLoadStatisticsMemoryStore::hasStorageAccess):
(WebKit::ResourceLoadStatisticsMemoryStore::grantStorageAccessInternal):
(WebKit::ResourceLoadStatisticsMemoryStore::clear):
(WebKit::ResourceLoadStatisticsMemoryStore::updateCookieBlocking):
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.cpp:
(WebKit::ResourceLoadStatisticsStore::removeDataRecords):
(WebKit::ResourceLoadStatisticsStore::grandfatherExistingWebsiteData):
(WebKit::ResourceLoadStatisticsStore::updateCookieBlockingForDomains):
(WebKit::ResourceLoadStatisticsStore::removeAllStorageAccess):
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::resourceLoadStatisticsUpdated):
(WebKit::WebResourceLoadStatisticsStore::requestStorageAccess):
(WebKit::WebResourceLoadStatisticsStore::scheduleClearInMemoryAndPersistent):
* NetworkProcess/NetworkConnectionToWebProcess.cpp:
(WebKit::NetworkConnectionToWebProcess::loadPing):
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::didClose):
(WebKit::NetworkProcess::fetchWebsiteData):
(WebKit::NetworkProcess::deleteWebsiteData):
(WebKit::NetworkProcess::deleteWebsiteDataForOrigins):
(WebKit::NetworkProcess::deleteAndRestrictWebsiteDataForRegistrableDomains):
(WebKit::NetworkProcess::registrableDomainsWithWebsiteData):
(WebKit::NetworkProcess::renameOriginInWebsiteData):
* NetworkProcess/cache/CacheStorageEngine.cpp:
(WebKit::CacheStorage::Engine::readCachesFromDisk):
(WebKit::CacheStorage::Engine::fetchDirectoryEntries):
(WebKit::CacheStorage::Engine::clearAllCaches):
(WebKit::CacheStorage::Engine::clearCachesForOrigin):
(WebKit::CacheStorage::Engine::clearCachesForOriginFromDirectories):
* NetworkProcess/cocoa/NetworkProcessCocoa.mm:
(WebKit::NetworkProcess::clearDiskCache):
* NetworkProcess/soup/NetworkProcessSoup.cpp:
(WebKit::NetworkProcess::clearDiskCache):
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreStatisticsResetToConsistentState):
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _takeViewSnapshot]):
* UIProcess/Automation/WebAutomationSession.cpp:
(WebKit::WebAutomationSession::resolveChildFrameHandle):
(WebKit::WebAutomationSession::resolveParentFrameHandle):
(WebKit::WebAutomationSession::computeElementLayout):
(WebKit::WebAutomationSession::selectOptionElement):
(WebKit::WebAutomationSession::setFilesForInputFileUpload):
(WebKit::WebAutomationSession::getAllCookies):
(WebKit::WebAutomationSession::deleteSingleCookie):
(WebKit::WebAutomationSession::addSingleCookie):
(WebKit::WebAutomationSession::performMouseInteraction):
(WebKit::WebAutomationSession::takeScreenshot):
* UIProcess/Cocoa/NavigationState.mm:
(WebKit::NavigationState::NavigationClient::decidePolicyForNavigationAction):
* UIProcess/UserMediaPermissionRequestManagerProxy.cpp:
(WebKit::UserMediaPermissionRequestManagerProxy::enumerateMediaDevicesForFrame):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::handleSynchronousMessage):
(WebKit::WebPageProxy::clearServiceWorkerEntitlementOverride):
(WebKit::WebPageProxy::decidePolicyForNavigationActionAsyncShared):
(WebKit::WebPageProxy::decidePolicyForNavigationAction):
(WebKit::WebPageProxy::decidePolicyForResponseShared):
(WebKit::WebPageProxy::didReceiveAuthenticationChallengeProxy):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::processForNavigationInternal):
(WebKit::WebProcessPool::seedResourceLoadStatisticsForTesting):
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::setMaxStatisticsEntries):
(WebKit::WebsiteDataStore::setPruneEntriesDownTo):
(WebKit::WebsiteDataStore::setGrandfatheringTime):
(WebKit::WebsiteDataStore::setMinimumTimeBetweenDataRecordsRemoval):
(WebKit::WebsiteDataStore::setPrevalentResource):
(WebKit::WebsiteDataStore::setPrevalentResourceForDebugMode):
(WebKit::WebsiteDataStore::setVeryPrevalentResource):
(WebKit::WebsiteDataStore::setShouldClassifyResourcesBeforeDataRecordsRemoval):
(WebKit::WebsiteDataStore::setSubframeUnderTopFrameDomain):
(WebKit::WebsiteDataStore::setSubresourceUnderTopFrameDomain):
(WebKit::WebsiteDataStore::setSubresourceUniqueRedirectTo):
(WebKit::WebsiteDataStore::setSubresourceUniqueRedirectFrom):
(WebKit::WebsiteDataStore::setTopFrameUniqueRedirectTo):
(WebKit::WebsiteDataStore::setTopFrameUniqueRedirectFrom):
(WebKit::WebsiteDataStore::clearPrevalentResource):
(WebKit::WebsiteDataStore::resetParametersToDefaultValues):
(WebKit::WebsiteDataStore::scheduleClearInMemoryAndPersistent):
(WebKit::WebsiteDataStore::getResourceLoadStatisticsDataSummary):
(WebKit::WebsiteDataStore::scheduleCookieBlockingUpdate):
(WebKit::WebsiteDataStore::scheduleStatisticsAndDataRecordsProcessing):
(WebKit::WebsiteDataStore::setLastSeen):
(WebKit::WebsiteDataStore::mergeStatisticForTesting):
(WebKit::WebsiteDataStore::insertExpiredStatisticForTesting):
(WebKit::WebsiteDataStore::setNotifyPagesWhenDataRecordsWereScanned):
(WebKit::WebsiteDataStore::setIsRunningResourceLoadStatisticsTest):
(WebKit::WebsiteDataStore::setNotifyPagesWhenTelemetryWasCaptured):
(WebKit::WebsiteDataStore::setTimeToLiveUserInteraction):
(WebKit::WebsiteDataStore::logUserInteraction):
(WebKit::WebsiteDataStore::clearUserInteraction):
(WebKit::WebsiteDataStore::setGrandfathered):
(WebKit::WebsiteDataStore::setUseITPDatabase):
(WebKit::WebsiteDataStore::setCrossSiteLoadWithLinkDecorationForTesting):
(WebKit::WebsiteDataStore::resetCrossSiteLoadsWithLinkDecorationForTesting):
(WebKit::WebsiteDataStore::deleteCookiesForTesting):
(WebKit::WebsiteDataStore::setResourceLoadStatisticsShouldDowngradeReferrerForTesting):
(WebKit::WebsiteDataStore::setThirdPartyCookieBlockingMode):
(WebKit::WebsiteDataStore::setResourceLoadStatisticsShouldEnbleSameSiteStrictEnforcementForTesting):
(WebKit::WebsiteDataStore::setResourceLoadStatisticsFirstPartyWebsiteDataRemovalModeForTesting):
(WebKit::WebsiteDataStore::setResourceLoadStatisticsToSameSiteStrictCookiesForTesting):
(WebKit::WebsiteDataStore::setCacheMaxAgeCapForPrevalentResources):
(WebKit::WebsiteDataStore::resetCacheMaxAgeCapForPrevalentResources):
(WebKit::WebsiteDataStore::setResourceLoadStatisticsDebugMode):
(WebKit::WebsiteDataStore::resetQuota):
(WebKit::WebsiteDataStore::clearAppBoundSession):
(WebKit::WebsiteDataStore::renameOriginInWebsiteData):
(WebKit::WebsiteDataStore::forwardAppBoundDomainsToITPIfInitialized):
(WebKit::WebsiteDataStore::setAppBoundDomainsForITP):
(WebKit::WebsiteDataStore::updateBundleIdentifierInNetworkProcess):
(WebKit::WebsiteDataStore::clearBundleIdentifierInNetworkProcess):
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::markAllLayersVolatile):
2020-06-23 Fujii Hironori <Hironori.Fujii@sony.com>
[curl] Unreviewed build fix for r263407
https://bugs.webkit.org/show_bug.cgi?id=213409
<rdar://problem/64654646>
* NetworkProcess/curl/NetworkDataTaskCurl.h: Include <wtf/MonotonicTime.h>.
2020-06-23 Alex Christensen <achristensen@webkit.org>
Expose already-available WKNavigationAction.request.HTTPBody through the API
https://bugs.webkit.org/show_bug.cgi?id=179077
Reviewed by Brady Eidson.
When this bug was written, we needed to serialize the HTTP body to the UI process to expose it through the API.
The controversial part of this was done in r237639, so this just exposes the data that is already available.
Verified by an API test.
* UIProcess/API/Cocoa/WKNavigationAction.mm:
(-[WKNavigationAction request]):
2020-06-23 Alex Christensen <achristensen@webkit.org>
Remove unnecessary Timer from NetworkDataTask
https://bugs.webkit.org/show_bug.cgi?id=213409
Reviewed by Youenn Fablet.
* NetworkProcess/NetworkDataTask.cpp:
(WebKit::NetworkDataTask::NetworkDataTask):
(WebKit::NetworkDataTask::scheduleFailure):
(WebKit::NetworkDataTask::failureTimerFired): Deleted.
* NetworkProcess/NetworkDataTask.h:
* NetworkProcess/NetworkDataTaskBlob.cpp:
(WebKit::NetworkDataTaskBlob::resume):
* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::NetworkDataTaskCocoa::NetworkDataTaskCocoa):
(WebKit::NetworkDataTaskCocoa::resume):
2020-06-22 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK][WPE] Add API to configure and handle DOM cache to WebKitWebsiteDataManager
https://bugs.webkit.org/show_bug.cgi?id=213337
Reviewed by Adrian Perez de Castro.
The default path is always used even for apps setting a base data directory. We should handle
WebsiteDataType::DOMCache to configure the directory and allow to fetch and clear it.
* UIProcess/API/glib/WebKitWebsiteData.cpp:
(recordContainsSupportedDataTypes):
(toWebKitWebsiteDataTypes):
* UIProcess/API/glib/WebKitWebsiteDataManager.cpp:
(webkitWebsiteDataManagerGetProperty):
(webkitWebsiteDataManagerSetProperty):
(webkitWebsiteDataManagerConstructed):
(webkit_website_data_manager_class_init):
(webkitWebsiteDataManagerGetDataStore):
(webkit_website_data_manager_get_dom_cache_directory):
(toWebsiteDataTypes):
* UIProcess/API/gtk/WebKitWebsiteData.h:
* UIProcess/API/gtk/WebKitWebsiteDataManager.h:
* UIProcess/API/gtk/docs/webkit2gtk-4.0-sections.txt:
* UIProcess/API/wpe/WebKitWebsiteData.h:
* UIProcess/API/wpe/WebKitWebsiteDataManager.h:
* UIProcess/API/wpe/docs/wpe-1.0-sections.txt:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::websiteDataStoreFromSessionID): GTK and WPE ports don't use the default website
data store, so only use it when it already exists. Use the process pool website data store instead.
2020-06-22 John Wilander <wilander@apple.com>
Storage Access API: Add the capability to call the Storage Access API as a quirk, on behalf of websites that should be doing it themselves
https://bugs.webkit.org/show_bug.cgi?id=213418
<rdar://problem/64549429>
Reviewed by Alex Christensen.
These changes are for forwarding ITP's knowledge of user interaction
for specific quirks domains to the WebKit::WebProcessPool where it in
turn can be distributed to all existing and new Web Content processes.
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::needsUserInteractionQuirk const):
(WebKit::WebResourceLoadStatisticsStore::callUpdatePrevalentDomainsToBlockCookiesForHandler):
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.h:
* Shared/WebProcessDataStoreParameters.h:
(WebKit::WebProcessDataStoreParameters::encode const):
(WebKit::WebProcessDataStoreParameters::decode):
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::setDomainsWithUserInteraction):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/Network/NetworkProcessProxy.messages.in:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::webProcessDataStoreParameters):
(WebKit::WebProcessPool::setDomainsWithUserInteraction):
* UIProcess/WebProcessPool.h:
* WebProcess/WebCoreSupport/WebResourceLoadObserver.cpp:
(WebKit::WebResourceLoadObserver::hasHadUserInteraction const):
* WebProcess/WebCoreSupport/WebResourceLoadObserver.h:
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::setWebsiteDataStoreParameters):
(WebKit::WebProcess::setDomainsWithUserInteraction):
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
2020-06-22 Wenson Hsieh <wenson_hsieh@apple.com>
[macOS] WebPageProxy::setPromisedDataForImage should sanitize its filename
https://bugs.webkit.org/show_bug.cgi?id=213486
<rdar://problem/56522262>
Reviewed by Megan Gardner.
Ensure that the filename passed into WebPageProxy::setPromisedDataForImage is sanitized in the UI process prior
to being used as a suggested filename when writing a dragged image to the pasteboard.
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::setPromisedDataForImage):
2020-06-22 Tim Horton <timothy_horton@apple.com>
WebEx can not share video from Safari on arm64
https://bugs.webkit.org/show_bug.cgi?id=213481
Reviewed by Dean Jackson.
* UIProcess/UserMediaProcessManager.cpp:
(WebKit::needsAppleCameraService):
(WebKit::UserMediaProcessManager::willCreateMediaStream):
(WebKit::UserMediaProcessManager::revokeSandboxExtensionsIfNeeded):
Pass sandbox extension for "com.apple.applecamerad" to the Web Content
process after the user allows camera access, when running on hardware which requires it.
* WebProcess/com.apple.WebProcess.sb.in:
Make it possible to dynamically extend the sandbox as above.
2020-06-22 Tim Horton <timothy_horton@apple.com>
Cannot play back encrypted media on arm64
https://bugs.webkit.org/show_bug.cgi?id=213483
Reviewed by Jer Noble and Per Arne Vollan.
* WebProcess/com.apple.WebProcess.sb.in:
Extend the sandbox as necessary for encrypted media support.
2020-06-22 Michael Catanzaro <mcatanzaro@gnome.org>
[WPE][GTK] Public API should not allow trying to register a special URI scheme
https://bugs.webkit.org/show_bug.cgi?id=209900
<rdar://problem/61200217>
Reviewed by Adrian Perez de Castro
Epiphany prior to 3.36 registers a scheme handler for ftp. This introduced criticals. Let's
use g_warning() instead, and avoid hardcoding the list of disallowed schemes.
* UIProcess/API/glib/WebKitWebContext.cpp:
(webkit_web_context_register_uri_scheme):
2020-06-22 Tim Horton <timothy_horton@apple.com>
dlopen() always fails on arm64, cannot load soft-linked libraries
https://bugs.webkit.org/show_bug.cgi?id=213480
Reviewed by Alexey Proskuryakov and Per Arne Vollan.
* Shared/mac/AuxiliaryProcessMac.mm:
(WebKit::initializeSandboxParameters):
Expose the auxiliary process' CPU family as a parameter to the sandbox profile,
so that it can be used to predicate sandbox expressions.
* WebProcess/com.apple.WebProcess.sb.in:
Allow mremap_encrypted from the Web Content process, because the arm64 dlopen() needs it.
2020-06-22 Tim Horton <timothy_horton@apple.com>
Update macOS version macros
https://bugs.webkit.org/show_bug.cgi?id=213484
Reviewed by Alexey Proskuryakov.
* Configurations/Base.xcconfig:
* Configurations/DebugRelease.xcconfig:
* Configurations/Version.xcconfig:
* Configurations/WebKitTargetConditionals.xcconfig:
2020-06-22 Tim Horton <timothy_horton@apple.com>
sysctl() fails due to sandbox violation
https://bugs.webkit.org/show_bug.cgi?id=213482
Reviewed by Maciej Stachowiak.
* NetworkProcess/mac/com.apple.WebKit.NetworkProcess.sb.in:
* WebProcess/com.apple.WebProcess.sb.in:
Extend the sandbox as necessary to make sysctl calls work.
2020-06-22 Antoine Quint <graouts@webkit.org>
WebKit fails to leave audio routing arbitration during navigation, closing.
https://bugs.webkit.org/show_bug.cgi?id=213426
<rdar://problem/64395051>
Unreviewed build fix for an unannotated switch fall-through between switch labels.
* UIProcess/API/Cocoa/WKWebViewTesting.mm:
(-[WKWebView _audioRoutingArbitrationStatus]):
2020-06-21 Jer Noble <jer.noble@apple.com>
WebKit fails to leave audio routing arbitration during navigation, closing.
https://bugs.webkit.org/show_bug.cgi?id=213426
<rdar://problem/64395051>
Reviewed by Eric Carlson.
Add testing SPIs to verify whether a WebPage successfully entered or left audio routing
arbitration. Notify the Arbitration proxy when the page is shut down, which ensures arbitration
will end if the client closes the WKWebView.
* UIProcess/API/Cocoa/WKWebViewPrivateForTesting.h:
* UIProcess/API/Cocoa/WKWebViewTesting.mm:
(-[WKWebView _audioRoutingArbitrationStatus]):
* UIProcess/API/mac/WKWebViewTestingMac.mm:
* UIProcess/Media/AudioSessionRoutingArbitratorProxy.h:
(WebKit::AudioSessionRoutingArbitratorProxy::arbitrationStatus const):
* UIProcess/Media/cocoa/AudioSessionRoutingArbitratorProxyCocoa.mm:
(WebKit::AudioSessionRoutingArbitratorProxy::processDidTerminate):
(WebKit::AudioSessionRoutingArbitratorProxy::beginRoutingArbitrationWithCategory):
(WebKit::AudioSessionRoutingArbitratorProxy::endRoutingArbitration):
* UIProcess/WebPageProxy.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::shutDown):
* UIProcess/WebProcessProxy.h:
(WebKit::WebProcessProxy::audioSessionRoutingArbitrator):
2020-06-21 Michael Catanzaro <mcatanzaro@gnome.org>
[WPE][GTK] Specify underlying storage type for InputMethodState::Hint
https://bugs.webkit.org/show_bug.cgi?id=213401
Reviewed by Carlos Garcia Campos.
Using unsigned types for flags enums seems nice to do.
* Shared/glib/InputMethodState.h:
2020-06-21 Michael Catanzaro <mcatanzaro@gnome.org>
[WPE][GTK] Add autocleanup for WebKitWebsitePolicies
https://bugs.webkit.org/show_bug.cgi?id=213399
Reviewed by Carlos Garcia Campos.
Add missing autocleanups.
* UIProcess/API/gtk/WebKitAutocleanups.h:
* UIProcess/API/wpe/WebKitAutocleanups.h:
2020-06-20 Jiewen Tan <jiewen_tan@apple.com>
[AppSSO] Should do null check on m_page in dismissViewController()
https://bugs.webkit.org/show_bug.cgi?id=213415
<rdar://problem/59672418>
Reviewed by Alexey Proskuryakov.
* UIProcess/Cocoa/SOAuthorization/SOAuthorizationSession.mm:
(WebKit::SOAuthorizationSession::dismissViewController):
2020-06-20 Jer Noble <jer.noble@apple.com>
RecoveryOS: PAL::getAVPlayerLayerClass() will crash when AVFoundation is missing
https://bugs.webkit.org/show_bug.cgi?id=213437
<rdar://problem/64563064>
Reviewed by Eric Carlson.
Check PAL::isAVFoundationAvailable() before calling PAL::getAVPlayerLayerClass();
* WebProcess/WebPage/RemoteLayerTree/PlatformCALayerRemoteCustom.mm:
(WebKit::PlatformCALayerRemoteCustom::clone const):
2020-06-20 Jer Noble <jer.noble@apple.com>
REGRESSION(r259219): Sleep assertion remains active if WKWebView is closed or WebContent process crashes
https://bugs.webkit.org/show_bug.cgi?id=213434
<rdar://problem/57541662>
Reviewed by Eric Carlson.
Tests: TestWebKitAPI tests:
SleepDisabler.Basic
SleepDisabler.Pause
SleepDisabler.Mute
SleepDisabler.Unmute
SleepDisabler.DisableAudioTrack
SleepDisabler.Loop
SleepDisabler.ChangeSrc
SleepDisabler.Load
SleepDisabler.Unload
SleepDisabler.Navigate
SleepDisabler.NavigateBack
SleepDisabler.Reload
SleepDisabler.Close
SleepDisabler.Crash
Ensure the SleepDisablers are cleared when the WebProcess closes or crashes.
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::shutDown):
(WebKit::WebProcessProxy::processDidTerminateOrFailedToLaunch):
2020-06-20 Brent Fulgham <bfulgham@apple.com>
[iOS, macOS] Allow access to the container manager to support Mail InjectedBundle
https://bugs.webkit.org/show_bug.cgi?id=213357
<rdar://problem/63837247>
Reviewed by Darin Adler.
The Mail Injected Bundle requires access to the container manager to support certain OS operations. We do not need
this access for web browsing, and should limit this access to this one case.
This patch creates a dynamic mach extension to the container manager for this single use case. It also denies the
non-extension access case with a backtrace so we can see if any other clients are hitting this.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
* WebProcess/com.apple.WebProcess.sb.in:
2020-06-19 Chris Dumez <cdumez@apple.com>
[Cocoa] Delay issuing ManagedSession & Network Extension sandbox extensions until a load is actually issued
https://bugs.webkit.org/show_bug.cgi?id=213414
<rdar://problem/64548684>
Reviewed by Per Arne Vollan.
Delay issuing ManagedSession & Network Extension sandbox extensions until a load is actually issued.
This is a Safari launch time optimization since the checks needed to decide whether or not to issue
the extensions are expensive and there is no reason to issue them as soon as the process launches
(especially in the case of a prewarmed process).
* Shared/Cocoa/LoadParametersCocoa.mm:
(WebKit::LoadParameters::platformEncode const):
(WebKit::LoadParameters::platformDecode):
* Shared/LoadParameters.h:
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::addPlatformLoadParameters):
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::addPlatformLoadParameters):
(WebKit::WebPageProxy::loadRequestWithNavigationShared):
(WebKit::WebPageProxy::loadFile):
(WebKit::WebPageProxy::loadDataWithNavigationShared):
(WebKit::WebPageProxy::loadAlternateHTML):
(WebKit::WebPageProxy::loadWebArchiveData):
* UIProcess/WebPageProxy.h:
* UIProcess/WebProcessProxy.h:
(WebKit::WebProcessProxy::hasNetworkExtensionSandboxAccess const):
(WebKit::WebProcessProxy::markHasNetworkExtensionSandboxAccess):
(WebKit::WebProcessProxy::hasManagedSessionSandboxAccess const):
(WebKit::WebProcessProxy::markHasManagedSessionSandboxAccess):
* WebProcess/WebPage/Cocoa/WebPageCocoa.mm:
(WebKit::WebPage::platformDidReceiveLoadParameters):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-06-19 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed, reverting r263287.
Caused tests to exit early with crashes on Catalina.
Reverted changeset:
"[iOS, macOS] Allow access to the container manager to support Mail InjectedBundle"
https://bugs.webkit.org/show_bug.cgi?id=213357
https://trac.webkit.org/changeset/263287
2020-06-19 Chris Dumez <cdumez@apple.com>
Crash under WebKit::ProcessAndUIAssertion::updateRunInBackgroundCount()
https://bugs.webkit.org/show_bug.cgi?id=213417
<rdar://problem/63477676>
Reviewed by Darin Adler.
Make sure |this| is still alive in ProcessAndUIAssertion::processAssertionWasInvalidated()
after calling ProcessAssertion::processAssertionWasInvalidated() and before calling
updateRunInBackgroundCount(). Calling ProcessAssertion::processAssertionWasInvalidated()
notifies the client and the client may destroy the assertion as a result.
* UIProcess/ios/ProcessAssertionIOS.mm:
(WebKit::ProcessAndUIAssertion::processAssertionWasInvalidated):
2020-06-19 Kate Cheney <katherine_cheney@apple.com>
com.apple.WebKit.Networking crash: suspended with locked system files (observations.db)
https://bugs.webkit.org/show_bug.cgi?id=213391
<rdar://problem/64494167>
Reviewed by Chris Dumez.
ITP is the only database that attempts to close when
NetworkProcess::didClose() is called. This causes the network process
to sometimes crash if the database is still closing when checking for
locked system files. Instead, we should only flush the
memory store to disk when calling NetworkProcess::didClose (in the
memory store case), and only close the database when the network session is destroyed.
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::didDestroyNetworkSession):
(WebKit::WebResourceLoadStatisticsStore::destroyResourceLoadStatisticsStore):
(WebKit::WebResourceLoadStatisticsStore::flushAndDestroyPersistentStore):
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.h:
Split into two functions, one to flush and destroy the persistent
store, and one to destroy the memory or database store.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::didClose):
Only flush the persistent store here, as described above.
* NetworkProcess/NetworkSession.cpp:
(WebKit::NetworkSession::flushAndDestroyPersistentStore):
* NetworkProcess/NetworkSession.h:
2020-06-19 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Provide a _WKWebAuthenticationPanelUpdatePINInvalid update to UI clients if the returned PIN from the client is not valid
https://bugs.webkit.org/show_bug.cgi?id=213404
<rdar://problem/64543894>
Reviewed by Brent Fulgham.
Provide a _WKWebAuthenticationPanelUpdatePINInvalid update to UI clients if the returned PIN from the client is not valid such that clients can
reuse the same logic to handle invalid pin from the authenticator. This change makes their life easier.
Covered by API tests.
* UIProcess/API/APIWebAuthenticationPanelClient.h:
(API::WebAuthenticationPanelClient::requestPin const):
* UIProcess/WebAuthentication/Cocoa/WebAuthenticationPanelClient.mm:
(WebKit::WebAuthenticationPanelClient::requestPin const):
Now, only null strings are intepreted as cancels.
* UIProcess/WebAuthentication/fido/CtapAuthenticator.cpp:
(WebKit::CtapAuthenticator::continueMakeCredentialAfterResponseReceived):
(WebKit::CtapAuthenticator::continueGetAssertionAfterResponseReceived):
(WebKit::CtapAuthenticator::continueGetPinTokenAfterRequestPin):
(WebKit::CtapAuthenticator::continueRequestAfterGetPinToken):
This patch also removes potential null pointer dereferences.
2020-06-19 Per Arne Vollan <pvollan@apple.com>
[macOS] Connections to the preference daemon are established before entering the sandbox
https://bugs.webkit.org/show_bug.cgi?id=213379
Reviewed by Darin Adler.
On macOS, connections to the preference daemon are established before entering the sandbox. These connections also persist
after entering the sandbox and denying access to the preference daemon. There should not be attempts to connect to the
preference daemon before entering the sandbox, since these attempts will not be stopped by the sandbox. This patch moves
code that connects to the preference daemon to be executed after the sandbox has been entered. That includes code to
prevent connections to the Dock and code to initialize WebKit logging. Also, instead of calling [NSBundle bundleForClass:],
call [NSBundle bundleWithIdentifier:], since calling [NSBundle bundleForClass:] will connect to the preference daemon.
Finally, allow the syscall SYS_gethostuuid, since that is needed by CoreFoundation when there is no access to the
preference daemon.
No new tests. This should be covered by existing tests. It would be nice to have a test to make sure that there are no
connections to the preference daemon just before entering the sandbox, but I am not aware of how to implement this.
* NetworkProcess/mac/NetworkProcessMac.mm:
(WebKit::NetworkProcess::initializeSandbox):
* Shared/AuxiliaryProcess.cpp:
(WebKit::AuxiliaryProcess::initialize):
* Shared/Cocoa/WebKit2InitializeCocoa.mm:
(WebKit::runInitializationCode):
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceMain.mm:
(WebKit::XPCServiceMain):
* Shared/mac/AuxiliaryProcessMac.mm:
(WebKit::webKit2Bundle):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
(WebKit::WebProcess::initializeSandbox):
* WebProcess/com.apple.WebProcess.sb.in:
2020-06-19 Chris Dumez <cdumez@apple.com>
Avoid initializing RenderTheme singleton unnecessarily in the UIProcess
https://bugs.webkit.org/show_bug.cgi?id=213406
Reviewed by Per Arne Vollan.
Avoid initializing RenderTheme singleton unnecessarily in the UIProcess. Instead, introduce
a static function to get the focus ring color on iOS.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
2020-06-19 Chris Dumez <cdumez@apple.com>
Unreviewed build fix after r263288.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
2020-06-19 Chris Dumez <cdumez@apple.com>
Use ASCIILiteral more for SandboxExtension functions
https://bugs.webkit.org/show_bug.cgi?id=213400
Reviewed by Per Arne Vollan.
Use ASCIILiteral more for SandboxExtension functions. Functions like createHandleForMachLookup() / createHandleForIOKitClassExtension()
don't need to take in Strings, ASCIILiteral is enough.
* Shared/Cocoa/SandboxExtensionCocoa.mm:
(WebKit::createHandlesForResources):
(WebKit::SandboxExtension::createReadOnlyHandlesForFiles):
(WebKit::SandboxExtension::createHandleForGenericExtension):
(WebKit::SandboxExtension::createHandleForMachLookup):
(WebKit::SandboxExtension::createHandlesForMachLookup):
(WebKit::SandboxExtension::createHandleForIOKitClassExtension):
(WebKit::SandboxExtension::createHandlesForIOKitClassExtensions):
* Shared/SandboxExtension.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::mediaRelatedMachServices):
(WebKit::nonBrowserServices):
(WebKit::diagnosticServices):
(WebKit::agxCompilerClasses):
(WebKit::WebProcessPool::platformInitializeWebProcess):
* UIProcess/Cocoa/WebProcessProxyCocoa.mm:
(WebKit::WebProcessProxy::enableRemoteInspectorIfNeeded):
* UIProcess/GPU/GPUProcessProxy.cpp:
(WebKit::GPUProcessProxy::singleton):
* UIProcess/UserMediaPermissionRequestManagerProxy.cpp:
(WebKit::UserMediaPermissionRequestManagerProxy::finishGrantingRequest):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didChooseFilesForOpenPanelWithDisplayStringAndIcon):
(WebKit::WebPageProxy::creationParameters):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::grantAccessToAssetServices):
2020-06-19 Brent Fulgham <bfulgham@apple.com>
[iOS, macOS] Allow access to the container manager to support Mail InjectedBundle
https://bugs.webkit.org/show_bug.cgi?id=213357
<rdar://problem/63837247>
Reviewed by Darin Adler.
The Mail Injected Bundle requires access to the container manager to support certain OS operations. We do not need
this access for web browsing, and should limit this access to this one case.
This patch creates a dynamic mach extension to the container manager for this single use case. It also denies the
non-extension access case with a backtrace so we can see if any other clients are hitting this.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
* WebProcess/com.apple.WebProcess.sb.in:
2020-06-19 Andres Gonzalez <andresg_22@apple.com>
AX: web process crash in AXObjectCache::postNotification.
https://bugs.webkit.org/show_bug.cgi?id=213398
Reviewed by Chris Fleizach.
AXObjectCache was being instantiated on the AX secondary thread.
Therefore the timers for the different delayed notifications where
initialized with the secondary thread. When postNotification was triggered
on the main thread as it should, and the timer was accessed, the timer
would assert/crash for being accessed in a thread different than where
it was created. This change guaranties that AXObjectCache is always
created on the main thread.
* WebProcess/WebPage/mac/WKAccessibilityWebPageObjectBase.mm:
(-[WKAccessibilityWebPageObjectBase axObjectCache]):
(-[WKAccessibilityWebPageObjectBase accessibilityPluginObject]):
(-[WKAccessibilityWebPageObjectBase accessibilityRootObjectWrapper]):
(-[WKAccessibilityWebPageObjectBase setWebPage:]):
(-[WKAccessibilityWebPageObjectBase setHasMainFramePlugin:]):
(-[WKAccessibilityWebPageObjectBase setRemoteParent:]):
2020-06-19 Chris Fleizach <cfleizach@apple.com>
AX: Make isolated tree enablement status dependent on client preference
https://bugs.webkit.org/show_bug.cgi?id=213355
<rdar://problem/64506577>
Reviewed by Zalan Bujtas.
We don't want the isolated tree mode in all clients (like Mail or Dictionary).
As a result, we can set this setting to off and allow safari and mini browser turn on more directly.
* Shared/WebPreferences.yaml:
* UIProcess/API/Cocoa/WKPreferences.mm:
(-[WKPreferences _setAccessibilityIsolatedTreeEnabled:]):
(-[WKPreferences _accessibilityIsolatedTreeEnabled]):
* UIProcess/API/Cocoa/WKPreferencesPrivate.h:
* WebProcess/InjectedBundle/InjectedBundle.cpp:
(WebKit::InjectedBundle::overrideBoolPreferenceForTestRunner):
(WebKit::InjectedBundle::setAccessibilityIsolatedTreeEnabled):
* WebProcess/InjectedBundle/InjectedBundle.h:
2020-06-19 Rob Buis <rbuis@igalia.com>
Enable referrer policy attribute support by default
https://bugs.webkit.org/show_bug.cgi?id=213285
Reviewed by Youenn Fablet.
Enable referrer policy attribute support by default by flipping the switch.
* Shared/WebPreferences.yaml:
2020-06-19 Youenn Fablet <youenn@apple.com>
LibWebRTCSocketClient::sendTo is too verbose in case of error
https://bugs.webkit.org/show_bug.cgi?id=213380
Reviewed by Eric Carlson.
Only log error message if the error code is different from the previous one.
* NetworkProcess/webrtc/LibWebRTCSocketClient.cpp:
(WebKit::LibWebRTCSocketClient::sendTo):
* NetworkProcess/webrtc/LibWebRTCSocketClient.h:
2020-06-19 Chris Dumez <cdumez@apple.com>
Web Inspector: RemoteInspector::singleton() slows down MobileSafari launch
https://bugs.webkit.org/show_bug.cgi?id=213381
<rdar://problem/64533003>
Reviewed by Darin Adler.
RemoteInspector::singleton() is expensive according to traces. It was called in WebProcessPool::initializeNewWebProcess()
which happens on MobileSafari launch because we prewarm a WebProcess. However, there is no reason to initialize remote
Web Inspector at this point. Per associated comment, Remote Web Inspector needs to be initialized once there is a
sub process hosting one of our Web View. As a result, I moved the initialization call to WebPageProxy::initializeWebPage()
which is when we actually send the IPC to the WebProcess to create the WebPage that is backing the UI-side WebView.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::initializeWebPage):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::initializeNewWebProcess):
2020-06-19 Chris Dumez <cdumez@apple.com>
Move Prefixed WebAudio interfaces behind their own feature flag
https://bugs.webkit.org/show_bug.cgi?id=213356
Reviewed by Darin Adler.
* Shared/WebPreferences.yaml:
2020-06-19 Rob Buis <rbuis@igalia.com>
Enable stale-while-revalidate support by default
https://bugs.webkit.org/show_bug.cgi?id=213286
Reviewed by Youenn Fablet.
Enable stale-while-revalidate support for Mac/GTK/WPE by changing the default
to true. Note that platforms that do not support speculative loads, which s-w-r
relies on, will not see any change in bahavior.
* UIProcess/WebsiteData/WebsiteDataStoreConfiguration.h:
2020-06-19 Alexander Mikhaylenko <alexm@gnome.org>
REGRESSION(r253360): [GTK] Page starts loading during animation in back/forward gesture
https://bugs.webkit.org/show_bug.cgi?id=205650
Reviewed by Carlos Garcia Campos.
r253360 changed the swipe gesture to start loading the page during the gesture, relying on
freezing the layer tree state to prevent unwanted redraws. However, it was implemented for
AC mode, but didn't prevent redraws in non-accelerated mode. Add a simple check to skip
redraws in this case.
* WebProcess/WebPage/CoordinatedGraphics/DrawingAreaCoordinatedGraphics.cpp:
(WebKit::DrawingAreaCoordinatedGraphics::display): Skip drawing if layer tree state is frozen.
2020-06-19 Carlos Garcia Campos <cgarcia@igalia.com>
Add support for fetching registrable domains with resource load statistics
https://bugs.webkit.org/show_bug.cgi?id=213291
Reviewed by Adrian Perez de Castro and Youenn Fablet.
WebsiteDataStore::fetchData() doesn't return anything for resource load statistics because
NetworkProcess::fetchWebsiteData() doesn't handle WebsiteDataType::ResourceLoadStatistics.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::allDomains const): Query all registrable domains from database.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:
(WebKit::ResourceLoadStatisticsMemoryStore::allDomains const): Return all registrable domains in memory map.
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.h:
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::registrableDomains): Get the list of registrable domains.
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.h:
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::fetchWebsiteData): Handle WebsiteDataType::ResourceLoadStatistics.
* Shared/WebsiteData/WebsiteData.cpp:
(WebKit::WebsiteData::encode const): Encode registrableDomainsWithResourceLoadStatistics.
(WebKit::WebsiteData::decode): Decode registrableDomainsWithResourceLoadStatistics.
* Shared/WebsiteData/WebsiteData.h:
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreRemoveITPDataForDomain): Use WebsiteDataRecord::addResourceLoadStatisticsRegistrableDomain()
instead of the display name.
* UIProcess/WebsiteData/WebsiteDataRecord.cpp:
(WebKit::WebsiteDataRecord::addResourceLoadStatisticsRegistrableDomain): Add the given registrable domain to the list.
* UIProcess/WebsiteData/WebsiteDataRecord.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::fetchDataAndApply): Handle registrable domains with resource load statistics.
(WebKit::WebsiteDataStore::removeData): Use resourceLoadStatisticsRegistrableDomains instead of the display name.
2020-06-18 David Kilzer <ddkilzer@apple.com>
Use fastMalloc/fastFree in WebCoreArgumentCodersMac.mm
<https://webkit.org/b/213325>
Reviewed by Darin Adler.
* Shared/mac/WebCoreArgumentCodersMac.mm:
(IPC::createArchiveList):
(IPC::createCFURLRequestFromSerializableRepresentation):
2020-06-18 Tim Horton <timothy_horton@apple.com>
Remove some vestiges of the 32-bit Plugin Process
https://bugs.webkit.org/show_bug.cgi?id=213361
Reviewed by Dan Bernstein.
We haven't launched 32-bit plugins for a while, but there are some
build system complications and random references in the code that we can get rid of.
* WebKit.xcodeproj/project.pbxproj:
* Configurations/PluginService.xcconfig: Renamed from Source/WebKit/Configurations/PluginService.64.xcconfig.
No need for .64 in the name anymore, since it's the only one.
We do have to leave the .64 in the product name because there are other things in the system that depend on this.
* PluginProcess/EntryPoint/Cocoa/XPCService/PluginService.Info.plist: Renamed from Source/WebKit/PluginProcess/EntryPoint/Cocoa/XPCService/PluginService.64.Info.plist.
No need for .64 in the name anymore.
* UIProcess/AuxiliaryProcessProxy.cpp:
(WebKit::AuxiliaryProcessProxy::getLaunchOptions):
* UIProcess/Launcher/ProcessLauncher.h:
* UIProcess/Launcher/mac/ProcessLauncherMac.mm:
(WebKit::serviceName):
* UIProcess/Plugins/mac/PluginProcessProxyMac.mm:
(WebKit::PluginProcessProxy::platformGetLaunchOptionsWithAttributes):
There need only be one ProcessType for the plugin process now!
* UIProcess/Plugins/unix/PluginProcessProxyUnix.cpp:
(WebKit::PluginProcessProxy::platformGetLaunchOptionsWithAttributes):
* UIProcess/Launcher/glib/BubblewrapLauncher.cpp:
(WebKit::bubblewrapSpawn):
* UIProcess/Launcher/glib/ProcessLauncherGLib.cpp:
(WebKit::ProcessLauncher::launchProcess):
GTK WebKit appears to also only launch 64-bit plugin processes.
2020-06-18 Megan Gardner <megan_gardner@apple.com>
Workaround for UIKit layout bug in time picker.
https://bugs.webkit.org/show_bug.cgi?id=213346
<rdar://problem/64042277>
Reviewed by Tim Horton.
There is a bug in the layout of UIKit that needs to be fixed.
Until then, this workaround should give us a view that is viable.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView resignFirstResponderForWebView]):
* UIProcess/ios/forms/WKDateTimeInputControl.mm:
(-[WKDateTimeContextMenuViewController preferredContentSize]):
2020-06-18 David Kilzer <ddkilzer@apple.com>
[IPC hardening] OptionSet<> values should be validated
<https://webkit.org/b/213199>
<rdar://problem/64369963>
Reviewed by Anders Carlsson.
Summary:
- Add WTF::EnumTraits<> for all OptionSet<> enums.
- Specify unsigned backing types for enum classes.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkSession.h:
* Platform/IPC/ArgumentCoders.h:
(IPC::ArgumentCoder<OptionSet<T>>::encode):
(IPC::ArgumentCoder<OptionSet<T>>::decode):
- Add WTF::isValidOptionSet() checks.
* Platform/IPC/Decoder.h:
* Platform/IPC/Encoder.h:
- Replace <wtf/EnumTraits.h> with <wtf/OptionSet.h> since the
latter now includes the former.
* Platform/IPC/MessageFlags.h:
* Shared/DocumentEditingContext.h:
* Shared/RemoteLayerTree/RemoteLayerTreeTransaction.h:
* Shared/WebEvent.h:
* Shared/WebsiteAutoplayQuirk.h:
* Shared/WebsiteData/WebsiteData.h:
* Shared/WebsiteData/WebsiteDataFetchOption.h:
* Shared/WebsiteData/WebsiteDataType.h:
* Shared/ios/GestureTypes.h:
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/Plugins/PluginProcessManager.h:
* UIProcess/WebProcessProxy.h:
* UIProcess/WebsiteData/WebsiteDataStore.h:
* WebProcess/WebProcess.h:
2020-06-18 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK][WPE] Add API to configure and handle service worker registrations to WebKitWebsiteDataManager
https://bugs.webkit.org/show_bug.cgi?id=213290
Reviewed by Michael Catanzaro.
The default path is always used even for apps setting a base data directory. We should handle
WebsiteDataType::ServiceWorkerRegistrations to configure the directory and allow to fetch and clear them.
* UIProcess/API/glib/WebKitWebsiteData.cpp:
(recordContainsSupportedDataTypes):
(toWebKitWebsiteDataTypes):
* UIProcess/API/glib/WebKitWebsiteDataManager.cpp:
(webkitWebsiteDataManagerGetProperty):
(webkitWebsiteDataManagerSetProperty):
(webkitWebsiteDataManagerConstructed):
(webkit_website_data_manager_class_init):
(webkitWebsiteDataManagerGetDataStore):
(webkit_website_data_manager_get_service_worker_registrations_directory):
(toWebsiteDataTypes):
* UIProcess/API/gtk/WebKitWebsiteData.h:
* UIProcess/API/gtk/WebKitWebsiteDataManager.h:
* UIProcess/API/gtk/docs/webkit2gtk-4.0-sections.txt:
* UIProcess/API/wpe/WebKitWebsiteData.h:
* UIProcess/API/wpe/WebKitWebsiteDataManager.h:
* UIProcess/API/wpe/docs/wpe-1.0-sections.txt:
2020-06-18 Yuri Chornoivan <yurchor@ukr.net>
Unreviewed. Fix a typo introduced in r263085
* UIProcess/API/glib/WebKitWebsiteDataManager.cpp:
(webkit_website_data_manager_class_init):
2020-06-17 Chris Dumez <cdumez@apple.com>
Add experimental feature flag for modern & unprefixed WebAudio API
https://bugs.webkit.org/show_bug.cgi?id=213268
Reviewed by Jer Noble.
Add experimental feature flag for modern & unprefixed WebAudio API,
off by default.
* Shared/WebPreferences.yaml:
2020-06-17 Wenson Hsieh <wenson_hsieh@apple.com>
[macOS] Shift-tab in a bullet list in Mail Compose jumps back to Subject field
https://bugs.webkit.org/show_bug.cgi?id=213320
<rdar://problem/63831962>
Reviewed by Tim Horton.
See WebCore/ChangeLog for more detail.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::handleKeyEventByRelinquishingFocusToChrome):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::handleEditingKeyboardEvent):
Hoist logic for relinquishing focus as a result of shift+tab from EventHandler to WebPage, so that it is the
default behavior when processing a "keypress" event that corresponds to shift+tab in an editable web view on
Cocoa platforms. If we do end up relinquishing focus, then we consider the keypress event to be handled.
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::handleEditingKeyboardEvent):
2020-06-17 Beth Dakin <bdakin@apple.com>
Remove references to blacklist/whitelist in the old C API
https://bugs.webkit.org/show_bug.cgi?id=213322
Reviewed by Tim Horton.
* UIProcess/API/C/WKPageGroup.cpp:
(WKPageGroupAddUserStyleSheet):
(WKPageGroupAddUserScript):
* UIProcess/API/C/WKPageGroup.h:
* WebProcess/InjectedBundle/API/c/WKBundlePrivate.h:
2020-06-17 Alex Christensen <achristensen@webkit.org>
Null check plugin.m_data before using it in PDFPlugin::ByteRangeRequest::completeUnconditionally
https://bugs.webkit.org/show_bug.cgi?id=213305
<rdar://problem/60349279>
Reviewed by Brady Eidson.
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::ByteRangeRequest::completeUnconditionally):
2020-06-17 Darryl Pogue <darryl@dpogue.ca>
IndexedDB: Support IDBFactory databases method
https://bugs.webkit.org/show_bug.cgi?id=211043
Reviewed by Youenn Fablet.
Plumbing for returning a list of IDB databases and versions.
* NetworkProcess/IndexedDB/WebIDBConnectionToClient.cpp:
(WebKit::WebIDBConnectionToClient::didGetAllDatabaseNamesAndVersions):
(WebKit::WebIDBConnectionToClient::didGetAllDatabaseNames): Deleted.
* NetworkProcess/IndexedDB/WebIDBConnectionToClient.h:
* NetworkProcess/IndexedDB/WebIDBServer.cpp:
(WebKit::WebIDBServer::getAllDatabaseNamesAndVersions):
(WebKit::WebIDBServer::getAllDatabaseNames): Deleted.
* NetworkProcess/IndexedDB/WebIDBServer.h:
* NetworkProcess/IndexedDB/WebIDBServer.messages.in:
* WebProcess/Databases/IndexedDB/WebIDBConnectionToServer.cpp:
(WebKit::WebIDBConnectionToServer::getAllDatabaseNamesAndVersions):
(WebKit::WebIDBConnectionToServer::didGetAllDatabaseNamesAndVersions):
(WebKit::WebIDBConnectionToServer::getAllDatabaseNames): Deleted.
(WebKit::WebIDBConnectionToServer::didGetAllDatabaseNames): Deleted.
* WebProcess/Databases/IndexedDB/WebIDBConnectionToServer.h:
* WebProcess/Databases/IndexedDB/WebIDBConnectionToServer.messages.in:
2020-06-16 Simon Fraser <simon.fraser@apple.com>
REGRESSION (r255037): Broken position while comparing watch bands on www.apple.com/shop/studio/apple-watch
https://bugs.webkit.org/show_bug.cgi?id=213282
<rdar://problem/63862940>
Reviewed by Antti Koivisto.
If a scrolling tree commit has a requested scroll position update for a node, we need
to call applyLayerPositons() to set the UIScrollView's contentOffset.
Often the layer tree commit will contain a boundsOrigin change for the layer in question
(which is equivalent to setting the UIScrollView's contentOffset), but some combinations of
user and programmatic scrolling result in no boundsOrigin change, revealing the bug.
* UIProcess/RemoteLayerTree/ios/ScrollingTreeScrollingNodeDelegateIOS.mm:
(WebKit::ScrollingTreeScrollingNodeDelegateIOS::commitStateAfterChildren):
2020-06-16 Kate Cheney <katherine_cheney@apple.com>
Check for inAppBrowserPrivacyEnabled flag in WebsiteDataStore::initializeAppBoundDomains is expensive and calls a non-static function WebsiteDataStore::parameters().
https://bugs.webkit.org/show_bug.cgi?id=213261
<rdar://problem/64317084>
Reviewed by Brent Fulgham.
There are two bugs here. First, WebsiteDataStore::parameters()
does a lot of work aside from returning the parameters, making this an
expensive call just to check the value of one flag. Second,
appBoundDomains() is a static hashset, and initializing the hashset
should not rely on non-static functions.
* NetworkProcess/NetworkSessionCreationParameters.cpp:
(WebKit::NetworkSessionCreationParameters::encode const):
(WebKit::NetworkSessionCreationParameters::decode):
* NetworkProcess/NetworkSessionCreationParameters.h:
* NetworkProcess/cocoa/NetworkSessionCocoa.h:
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::NetworkSessionCocoa):
Remove flag, it is no longer used here.
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::WebsiteDataStore::platformSetNetworkParameters):
Save the flag value in a member variable so we can use that later
when checking app-bound domains. Rename to something more descriptive.
(WebKit::WebsiteDataStore::initializeAppBoundDomains):
Remove the check for the non-static runtime flag.
(WebKit::WebsiteDataStore::ensureAppBoundDomains const):
Enable test mode quirks. We must do it in two places in case the
initialization hasn't happened by the time we call
ensureAppBoundDomains.
* UIProcess/WebsiteData/WebsiteDataStore.h:
2020-06-16 Mark Lam <mark.lam@apple.com>
Make Options::useJIT() be the canonical source of truth on whether we should use the JIT.
https://bugs.webkit.org/show_bug.cgi?id=212556
<rdar://problem/63780436>
Reviewed by Saam Barati.
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::isJITEnabled):
2020-06-16 David Kilzer <ddkilzer@apple.com>
REGRESSION (r262994): Web Inspector is killed when context-clicking anywhere
<https://webkit.org/b/213224>
<rdar://problem/64383320>
Reviewed by Darin Adler.
* Platform/IPC/Decoder.h:
* Platform/IPC/Encoder.h:
- Include <WebCore/ContextMenuItem.h> since the definition of
the custom WTF::isValidEnum<WebCore::ContextMenuAction>
function is needed.
* Scripts/webkit/messages.py:
(generate_message_names_header):
- Change the third typename in the template for
WTF::isValidEnum<IPC::MessageName>() to avoid partial template
specialization errors from other implementations like
WTF::isValidEnum<WebCore::ContextMenuAction>().
- Also replace hard-coded "IPC::MessageName" names with template
values.
2020-06-16 Chris Dumez <cdumez@apple.com>
Stop calling userPreferredLanguages() in the UIProcess
https://bugs.webkit.org/show_bug.cgi?id=213214
<rdar://problem/64317593>
Reviewed by Darin Adler.
Follow-up to r263094 to address post-landing feedback from Darin.
Use makeVector() functionality in WTF to convert the NSArray into a Vector instead
of doing it by hand. Also mark the NeverDestroyed variable as static const to
avoid leaking.
* UIProcess/Cocoa/WebProcessProxyCocoa.mm:
(WebKit::WebProcessProxy::platformOverrideLanguages const):
2020-06-16 Alex Christensen <achristensen@webkit.org>
Provide alternatively-named WKBrowsingContextGroup SPI
https://bugs.webkit.org/show_bug.cgi?id=213229
Reviewed by Tim Horton.
This is unfortunately still used. In order to try to convince the user to make the smallest change possible, make newly named SPI that behaves identically.
Doubly-deprecate the old SPI and implement it in terms of the new during the transition period.
* UIProcess/API/Cocoa/WKBrowsingContextGroup.h:
* UIProcess/API/Cocoa/WKBrowsingContextGroup.mm:
(-[WKBrowsingContextGroup addUserStyleSheet:baseURL:whitelistedURLPatterns:blacklistedURLPatterns:mainFrameOnly:]):
(-[WKBrowsingContextGroup addUserStyleSheet:baseURL:includeMatchPatternStrings:excludeMatchPatternStrings:mainFrameOnly:]):
(-[WKBrowsingContextGroup addUserScript:baseURL:whitelistedURLPatterns:blacklistedURLPatterns:injectionTime:mainFrameOnly:]):
(-[WKBrowsingContextGroup addUserScript:baseURL:includeMatchPatternStrings:excludeMatchPatternStrings:injectionTime:mainFrameOnly:]):
2020-06-16 Youenn Fablet <youenn@apple.com>
Loads triggered by an opaque stylesheet should not be visible to service workers
https://bugs.webkit.org/show_bug.cgi?id=213246
Reviewed by Alex Christensen.
* WebProcess/Network/WebLoaderStrategy.cpp:
(WebKit::WebLoaderStrategy::scheduleLoadFromNetworkProcess):
In case the load is triggered from an opaque response, set service worker mode to none.
2020-06-16 Charlie Turner <cturner@igalia.com>
[GTK] Disable video autoplay
https://bugs.webkit.org/show_bug.cgi?id=184845
Reviewed by Carlos Garcia Campos.
* PlatformGTK.cmake: Add new source files to the build.
* PlatformWPE.cmake: Ditto.
* SourcesGTK.txt: Ditto.
* SourcesWPE.txt: Ditto.
* UIProcess/API/glib/WebKitPolicyDecision.cpp:
(webkit_policy_decision_use_with_policies): : Add an API to
respond to navigation decisions with policies.
* UIProcess/API/glib/WebKitWebContext.cpp:
(webkitWebContextCreatePageForWebView): Plumb in the default
website policies that are associated with the web view.
* UIProcess/API/glib/WebKitWebContextPrivate.h:
* UIProcess/API/glib/WebKitWebView.cpp:
(webkitWebViewConstructed):
(webkitWebViewSetProperty):
(webkitWebViewGetProperty):
(webkit_web_view_class_init):
(webkit_web_view_get_website_policies):
* UIProcess/API/glib/WebKitWebsitePolicies.cpp: Added.
(_WebKitWebsitePoliciesPrivate::_WebKitWebsitePoliciesPrivate):
(webkitWebsitePoliciesGetWebsitePolicies):
(webkitWebsitePoliciesGetPoliciesData):
(webkitWebsitePoliciesGetProperty):
(webkitWebsitePoliciesSetAutoplayPolicy):
(webkitWebsitePoliciesSetProperty):
(webkit_website_policies_class_init):
(webkit_website_policies_new):
(webkit_website_policies_new_with_policies):
(webkit_website_policies_get_autoplay_policy):
* UIProcess/API/glib/WebKitWebsitePoliciesPrivate.h: Added.
* UIProcess/API/gtk/WebKitPolicyDecision.h:
* UIProcess/API/gtk/WebKitWebView.h:
* UIProcess/API/gtk/WebKitWebViewGtk.cpp:
(webkit_web_view_new_with_related_view):
* UIProcess/API/gtk/WebKitWebsitePolicies.h: Added.
* UIProcess/API/gtk/docs/webkit2gtk-4.0-sections.txt:
* UIProcess/API/wpe/WebKitPolicyDecision.h:
* UIProcess/API/wpe/WebKitWebView.h:
* UIProcess/API/wpe/WebKitWebsitePolicies.h: Added.
* UIProcess/API/wpe/docs/wpe-1.0-sections.txt:
2020-06-16 Chris Dumez <cdumez@apple.com>
Stop calling userPreferredLanguages() in the UIProcess
https://bugs.webkit.org/show_bug.cgi?id=213214
<rdar://problem/64317593>
Reviewed by Per Arne Vollan.
Stop calling userPreferredLanguages() in the UIProcess since this can be slow and may keep
the main thread busy. Nowadays, the WebProcess is able to call it by itself anyway. The
sandbox is allowing it.
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::languageChanged):
(WebKit::WebProcessPool::initializeNewWebProcess):
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::initializeWebProcess):
(WebKit::WebProcess::userPreferredLanguagesChanged const):
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
2020-06-16 Truitt Savell <tsavell@apple.com>
Unreviewed, reverting r263079.
Broke Internal builds.
Reverted changeset:
"IndexedDB: Support IDBFactory databases method"
https://bugs.webkit.org/show_bug.cgi?id=211043
https://trac.webkit.org/changeset/263079
2020-06-16 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK][WPE] Add API to configure and enable resource load statistics
https://bugs.webkit.org/show_bug.cgi?id=177943
Reviewed by Michael Catanzaro.
* UIProcess/API/glib/WebKitWebsiteData.cpp:
(recordContainsSupportedDataTypes):
(toWebKitWebsiteDataTypes):
* UIProcess/API/glib/WebKitWebsiteDataManager.cpp:
(webkitWebsiteDataManagerGetProperty):
(webkitWebsiteDataManagerSetProperty):
(webkitWebsiteDataManagerConstructed):
(webkit_website_data_manager_class_init):
(webkitWebsiteDataManagerGetDataStore):
(webkit_website_data_manager_get_itp_directory):
(webkit_website_data_manager_set_itp_enabled):
(webkit_website_data_manager_get_itp_enabled):
(toWebsiteDataTypes):
* UIProcess/API/gtk/WebKitWebsiteData.h:
* UIProcess/API/gtk/WebKitWebsiteDataManager.h:
* UIProcess/API/gtk/docs/webkit2gtk-4.0-sections.txt:
* UIProcess/API/wpe/WebKitWebsiteData.h:
* UIProcess/API/wpe/WebKitWebsiteDataManager.h:
* UIProcess/API/wpe/docs/wpe-1.0-sections.txt:
* UIProcess/glib/WebsiteDataStoreGLib.cpp:
(WebKit::WebsiteDataStore::defaultResourceLoadStatisticsDirectory):
2020-06-16 Darryl Pogue <darryl@dpogue.ca>
IndexedDB: Support IDBFactory databases method
https://bugs.webkit.org/show_bug.cgi?id=211043
Reviewed by Youenn Fablet.
Plumbing for returning a list of IDB databases and versions.
* NetworkProcess/IndexedDB/WebIDBConnectionToClient.cpp:
(WebKit::WebIDBConnectionToClient::didGetAllDatabaseNamesAndVersions):
(WebKit::WebIDBConnectionToClient::didGetAllDatabaseNames): Deleted.
* NetworkProcess/IndexedDB/WebIDBConnectionToClient.h:
* NetworkProcess/IndexedDB/WebIDBServer.cpp:
(WebKit::WebIDBServer::getAllDatabaseNamesAndVersions):
(WebKit::WebIDBServer::getAllDatabaseNames): Deleted.
* NetworkProcess/IndexedDB/WebIDBServer.h:
* NetworkProcess/IndexedDB/WebIDBServer.messages.in:
* WebProcess/Databases/IndexedDB/WebIDBConnectionToServer.cpp:
(WebKit::WebIDBConnectionToServer::getAllDatabaseNamesAndVersions):
(WebKit::WebIDBConnectionToServer::didGetAllDatabaseNamesAndVersions):
(WebKit::WebIDBConnectionToServer::getAllDatabaseNames): Deleted.
(WebKit::WebIDBConnectionToServer::didGetAllDatabaseNames): Deleted.
* WebProcess/Databases/IndexedDB/WebIDBConnectionToServer.h:
* WebProcess/Databases/IndexedDB/WebIDBConnectionToServer.messages.in:
2020-06-15 Pavel Feldman <pavel.feldman@gmail.com>
Web Inspector: introduce request interception
https://bugs.webkit.org/show_bug.cgi?id=207446
Reviewed by Devin Rousso.
This change introduces network request interception to the Network
protocol domain. It adds Network.interceptWithRequest notification that
can be continued, modified or fulfilled. NetworkStage enum can now have
'request' and 'response' values.
* WebProcess/Network/WebLoaderStrategy.cpp:
(WebKit::WebLoaderStrategy::scheduleLoad):
2020-06-15 Brent Fulgham <bfulgham@apple.com>
[macOS] Update sandboxes to reduce logging for media-related operations
https://bugs.webkit.org/show_bug.cgi?id=213210
<rdar://problem/64376237>
Reviewed by Per Arne Vollan.
Ongoing testing has uncovered a set of additional services and IOKit
properties that we should allow without logging.
* NetworkProcess/mac/com.apple.WebKit.NetworkProcess.sb.in:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::mediaRelatedMachServices):
* WebProcess/com.apple.WebProcess.sb.in:
2020-06-15 Alex Christensen <achristensen@webkit.org>
Provide alternatively-named SPI for user style sheets and scripts
https://bugs.webkit.org/show_bug.cgi?id=213206
Reviewed by Brady Eidson.
Test-only SPI was just renamed.
SPI used by other projects was deprecated with a replacement so we can remove the old names soon.
_WKUserContentWorld/WKContentWorld was tied up in this, and rather than provide replacement SPI with deprecated _WKUserContentWorld,
this will move some SPI from _WKUserContentWorld to WKContentWorld too. The most heavy user of _WKUserContentWorld is TestWebKitAPI and we'll remove it soon.
* UIProcess/API/C/WKContext.cpp:
(WKContextSetFontAllowList):
(WKContextSetFontWhitelist): Deleted.
* UIProcess/API/C/WKContextPrivate.h:
* UIProcess/API/Cocoa/WKUserContentController.mm:
(-[WKUserContentController _removeAllUserScriptsAssociatedWithContentWorld:]):
(-[WKUserContentController _removeAllUserStyleSheetsAssociatedWithContentWorld:]):
(-[WKUserContentController _addScriptMessageHandler:name:contentWorld:]):
(-[WKUserContentController _removeAllUserScriptsAssociatedWithUserContentWorld:]): Deleted.
(-[WKUserContentController _removeAllUserStyleSheetsAssociatedWithUserContentWorld:]): Deleted.
* UIProcess/API/Cocoa/WKUserContentControllerPrivate.h:
* UIProcess/API/Cocoa/WKUserScript.mm:
(-[WKUserScript _initWithSource:injectionTime:forMainFrameOnly:includeMatchPatternStrings:excludeMatchPatternStrings:associatedURL:contentWorld:deferRunningUntilNotification:]):
* UIProcess/API/Cocoa/WKUserScriptPrivate.h:
* UIProcess/API/Cocoa/_WKUserStyleSheet.h:
* UIProcess/API/Cocoa/_WKUserStyleSheet.mm:
(-[_WKUserStyleSheet initWithSource:forWKWebView:forMainFrameOnly:includeMatchPatternStrings:excludeMatchPatternStrings:baseURL:level:contentWorld:]):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::setFontAllowList):
(WebKit::WebProcessPool::setFontWhitelist): Deleted.
* UIProcess/WebProcessPool.h:
2020-06-15 Alex Christensen <achristensen@webkit.org>
Add SPI to preconnect to a server
https://bugs.webkit.org/show_bug.cgi?id=213109
<rdar://problem/64184412>
Reviewed by Geoff Garen.
Covered by API tests.
* UIProcess/API/Cocoa/WKProcessPoolPrivate.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _preconnectToServer:]):
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::preconnectTo):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::preconnectTo):
* UIProcess/WebPageProxy.h:
2020-06-15 Keith Miller <keith_miller@apple.com>
Signal handlers should have a two phase installation.
https://bugs.webkit.org/show_bug.cgi?id=213160
Reviewed by Mark Lam.
Put back old WASM fast memory installation.
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::initializeWebProcess):
2020-06-15 Víctor Manuel Jáquez Leal <vjaquez@igalia.com>
Suppress compiler warnings
https://bugs.webkit.org/show_bug.cgi?id=213034
Reviewed by Darin Adler.
Convert to std:size_t to compare equal data types.
* NetworkProcess/cache/NetworkCacheData.cpp:
(WebKit::NetworkCache::readOrMakeSalt):
2020-06-15 Chris Dumez <cdumez@apple.com>
Revert r260856 as it seems to have regressed PLT on iOS
https://bugs.webkit.org/show_bug.cgi?id=213194
<rdar://problem/64331458>
Unreviewed, revert r260856 due to regression.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::NetworkProcess):
(WebKit::NetworkProcess::prepareToSuspend):
(WebKit::NetworkProcess::resume):
* NetworkProcess/NetworkProcess.h:
* Shared/WebSQLiteDatabaseTracker.cpp: Added.
(WebKit::WebSQLiteDatabaseTracker::WebSQLiteDatabaseTracker):
(WebKit::WebSQLiteDatabaseTracker::~WebSQLiteDatabaseTracker):
(WebKit::WebSQLiteDatabaseTracker::setIsSuspended):
(WebKit::WebSQLiteDatabaseTracker::willBeginFirstTransaction):
(WebKit::WebSQLiteDatabaseTracker::didFinishLastTransaction):
(WebKit::WebSQLiteDatabaseTracker::setIsHoldingLockedFiles):
* Shared/WebSQLiteDatabaseTracker.h: Added.
* SourcesCocoa.txt:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::didClose):
(WebKit::NetworkProcessProxy::setIsHoldingLockedFiles):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/Network/NetworkProcessProxy.messages.in:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::shutDown):
(WebKit::WebProcessProxy::setIsHoldingLockedFiles):
* UIProcess/WebProcessProxy.h:
* UIProcess/WebProcessProxy.messages.in:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebProcess.cpp:
(WebKit::m_webSQLiteDatabaseTracker):
(WebKit::WebProcess::prepareToSuspend):
(WebKit::WebProcess::processDidResume):
* WebProcess/WebProcess.h:
2020-06-15 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK][WPE] Add API to expose UIClient::requestStorageAccessConfirm
https://bugs.webkit.org/show_bug.cgi?id=210422
Reviewed by Michael Catanzaro.
Add WebKitWebsiteDataAccessPermissionRequest to ask for permission to the user using the existing permission
request API.
* PlatformGTK.cmake:
* PlatformWPE.cmake:
* SourcesGTK.txt:
* SourcesWPE.txt:
* UIProcess/API/glib/WebKitUIClient.cpp:
* UIProcess/API/glib/WebKitWebsiteDataAccessPermissionRequest.cpp: Added.
(webkitWebsiteDataAccessPermissionRequestAllow):
(webkitWebsiteDataAccessPermissionRequestDeny):
(webkit_permission_request_interface_init):
(webkitWebsiteDataAccessPermissionRequestDispose):
(webkit_website_data_access_permission_request_class_init):
(webkitWebsiteDataAccessPermissionRequestCreate):
(webkit_website_data_access_permission_request_get_requesting_domain):
(webkit_website_data_access_permission_request_get_current_domain):
* UIProcess/API/glib/WebKitWebsiteDataAccessPermissionRequestPrivate.h: Added.
* UIProcess/API/gtk/WebKitWebsiteDataAccessPermissionRequest.h: Added.
* UIProcess/API/gtk/docs/webkit2gtk-4.0-sections.txt:
* UIProcess/API/gtk/docs/webkit2gtk-docs.sgml:
* UIProcess/API/gtk/webkit2.h:
* UIProcess/API/wpe/WebKitWebsiteDataAccessPermissionRequest.h: Added.
* UIProcess/API/wpe/docs/wpe-1.0-sections.txt:
* UIProcess/API/wpe/docs/wpe-docs.sgml:
* UIProcess/API/wpe/webkit.h:
2020-06-14 Sam Weinig <weinig@apple.com>
[WPT] html/webappapis/system-state-and-capabilities/the-navigator-object/navigator-pluginarray.html fails due to lack of caching of Plugin objects
https://bugs.webkit.org/show_bug.cgi?id=213185
Reviewed by Darin Adler.
* UIProcess/Plugins/PluginInfoStore.h:
Add missing #include that is now needed due to pruning of unnecessary #includes
in WebCore.
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::pluginSupportsExtension):
Update to use new webVisibleMimeTypes() rather than the clunky getWebVisibleMimesAndPluginIndices().
2020-06-14 David Kilzer <ddkilzer@apple.com>
[IPC hardening] Return type of -[NSCoder validateClassSupportsSecureCoding:] is incorrect
<https://webkit.org/b/213161>
<rdar://problem/64050085>
Reviewed by Darin Adler.
* Shared/API/Cocoa/WKRemoteObjectCoder.mm:
(validateClass):
- Fix return type of
-[NSCoder validateClassSupportsSecureCoding:]. Throw an
NSException in case the method ever returns NO without
throwing its own NSException. (Currently this method does
throw an NSException instead of returning NO.)
2020-06-14 Carlos Garcia Campos <cgarcia@igalia.com>
[SOUP] Disable HSTS for requests when cookies will be blocked by ITP
https://bugs.webkit.org/show_bug.cgi?id=210739
Reviewed by Michael Catanzaro.
* NetworkProcess/soup/NetworkDataTaskSoup.cpp:
(WebKit::NetworkDataTaskSoup::shouldAllowHSTSProtocolUpgrade const):
2020-06-13 Richard Houle <rhoule@apple.com>
Fix Overrelease in makeFailureSetForAllTextManipulationItems
https://bugs.webkit.org/show_bug.cgi?id=213165
<rdar://problem/61389164>
Reviewed by Wenson Hsieh.
* UIProcess/API/Cocoa/WKWebView.mm:
(makeFailureSetForAllTextManipulationItems):
2020-06-13 Wenson Hsieh <wenson_hsieh@apple.com>
[Mac Catalyst] Color inputs and selects fail to display popovers when clicked
https://bugs.webkit.org/show_bug.cgi?id=213157
<rdar://problem/64004135>
Reviewed by Tim Horton.
In Mac Catalyst, presenting popovers using `UIPopoverController` causes the first responder to change. This
means that the WKContentView will resign first responder, which in turn causes the currently presented popover
to be dismissed. This means that popovers for color inputs and select elements will be dismissed immediately
after presentation.
To mitigate this, adapt the approach taken in r259840 to `WKColorPopover` and `WKSelectPopover` by teaching the
base class (`WKRotatingPopover`) to temporarily preserve the focused element in light of first responder changes
while presenting a popover on Mac Catalyst.
* UIProcess/ios/forms/WKFormPopover.mm:
(-[WKRotatingPopover presentPopoverAnimated:]):
(-[WKRotatingPopover dismissPopoverAnimated:]):
2020-06-13 Sam Weinig <weinig@apple.com>
Extended Color: Experiment with strongly typed ColorComponents
https://bugs.webkit.org/show_bug.cgi?id=212396
Reviewed by Darin Adler.
* UIProcess/API/wpe/WebKitColor.cpp:
(webkitColorFillFromWebCoreColor):
* UIProcess/gtk/ViewGestureControllerGtk.cpp:
(WebKit::ViewGestureController::beginSwipeGesture):
* WebProcess/WebPage/WebFrame.cpp:
(WebKit::WebFrame::getDocumentBackgroundColor):
Update to call toSRGBALossy() rather than toSRGBAComponentsLossy().
2020-06-13 Tetsuharu Ohzeki <tetsuharu.ohzeki@gmail.com>
Remove FileError.h
https://bugs.webkit.org/show_bug.cgi?id=213119
Reviewed by Chris Dumez.
* WebProcess/Network/WebSocketChannel.cpp:
(WebKit::WebSocketChannel::createMessageQueue):
2020-06-12 Brian Burg <bburg@apple.com>
Automation.computeElementLayout should return iframe-relative element rects (when coordinate system is 'Page')
https://bugs.webkit.org/show_bug.cgi?id=213139
<rdar://problem/55042445>
Reviewed by Devin Rousso.
Automation.computeElementLayout has two coordinate systems it can use, LayoutViewport (used for hit testing)
and Page (used for Get Element Rect). Since Get Element Rect expects coordinates relative to the current
browsing context, make the mode simpler by reporting the bounds from Element::boundingClientRect() which
are frame-relative.
Covered by existing WPT suite. The semantics of this command are under review, as
the remote end steps seem to describe a different result than what the non-normative text does.
* WebProcess/Automation/WebAutomationSessionProxy.cpp:
(WebKit::WebAutomationSessionProxy::computeElementLayout):
2020-06-12 David Kilzer <ddkilzer@apple.com>
[IPC hardening] Check enum values in IPC::Decoder::decodeEnum() an IPC::Encoder::encodeEnum()
<https://webkit.org/b/211988>
<rdar://problem/63137695>
Reviewed by Darin Adler.
Replace decodeEnum() with decode() and encodeEnum() with
operator<<().
* GPUProcess/media/TextTrackPrivateRemoteConfiguration.h:
(WebKit::TextTrackPrivateRemoteConfiguration::encode const):
* GPUProcess/media/TrackPrivateRemoteConfiguration.h:
(WebKit::TrackPrivateRemoteConfiguration::encode const):
* NetworkProcess/NetworkProcessCreationParameters.cpp:
(WebKit::NetworkProcessCreationParameters::encode const):
(WebKit::NetworkProcessCreationParameters::decode):
* NetworkProcess/NetworkResourceLoadParameters.cpp:
(WebKit::NetworkResourceLoadParameters::encode const):
(WebKit::NetworkResourceLoadParameters::decode):
* Platform/IPC/Decoder.h:
(IPC::Decoder::decode):
(IPC::Decoder::operator>>):
- Make use of std::underlying_type_t<>.
(IPC::Decoder::decodeEnum): Delete.
- Replace with calls to decode().
* Platform/IPC/Encoder.h:
(IPC::Encoder::operator<<):
(IPC::Encoder::encode):
- Make use of WTF::enumToUnderlyingType<>.
(IPC::Encoder::encodeEnum): Delete.
- Replace with calls to operator<<().
* Shared/Cocoa/ArgumentCodersCocoa.mm:
(IPC::decodeObject):
* Shared/Cocoa/WebCoreArgumentCodersCocoa.mm:
(IPC::ArgumentCoder<ApplePaySessionPaymentRequest>::encode):
(IPC::ArgumentCoder<ApplePaySessionPaymentRequest>::decode):
(IPC::ArgumentCoder<ApplePaySessionPaymentRequest::LineItem>::encode):
(IPC::ArgumentCoder<ApplePaySessionPaymentRequest::LineItem>::decode):
* Shared/ContextMenuContextData.cpp:
(WebKit::ContextMenuContextData::encode const):
(WebKit::ContextMenuContextData::decode):
* Shared/FocusedElementInformation.cpp:
(WebKit::FocusedElementInformation::encode const):
(WebKit::FocusedElementInformation::decode):
* Shared/LoadParameters.cpp:
(WebKit::LoadParameters::encode const):
(WebKit::LoadParameters::decode):
* Shared/NavigationActionData.cpp:
(WebKit::NavigationActionData::encode const):
(WebKit::NavigationActionData::decode):
* Shared/PlatformPopupMenuData.cpp:
(WebKit::PlatformPopupMenuData::encode const):
(WebKit::PlatformPopupMenuData::decode):
* Shared/Plugins/PluginProcessCreationParameters.cpp:
(WebKit::PluginProcessCreationParameters::encode const):
(WebKit::PluginProcessCreationParameters::decode):
* Shared/PrintInfo.cpp:
(WebKit::PrintInfo::encode const):
(WebKit::PrintInfo::decode):
* Shared/RTCPacketOptions.cpp:
(WebKit::RTCPacketOptions::encode const):
(WebKit::RTCPacketOptions::decode):
* Shared/RemoteLayerTree/RemoteLayerTreeTransaction.mm:
(WebKit::RemoteLayerTreeTransaction::LayerCreationProperties::encode const):
(WebKit::RemoteLayerTreeTransaction::LayerCreationProperties::decode):
(WebKit::RemoteLayerTreeTransaction::LayerProperties::encode const):
(WebKit::RemoteLayerTreeTransaction::LayerProperties::decode):
* Shared/RemoteLayerTree/RemoteScrollingCoordinatorTransaction.cpp:
(ArgumentCoder<ScrollingStateNode>::encode):
(ArgumentCoder<RequestedScrollData>::encode):
(ArgumentCoder<RequestedScrollData>::decode):
(WebKit::RemoteScrollingCoordinatorTransaction::decode):
* Shared/SessionState.cpp:
(WebKit::HTTPBody::Element::encode const):
(WebKit::HTTPBody::Element::decode):
(WebKit::PageState::encode const):
(WebKit::PageState::decode):
* Shared/TouchBarMenuItemData.cpp:
(WebKit::TouchBarMenuItemData::encode const):
(WebKit::TouchBarMenuItemData::decode):
* Shared/UserData.cpp:
(WebKit::UserData::encode):
(WebKit::UserData::decode):
* Shared/WebContextMenuItemData.cpp:
(WebKit::WebContextMenuItemData::encode const):
(WebKit::WebContextMenuItemData::decode):
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<LinearTimingFunction>::encode):
(IPC::ArgumentCoder<CubicBezierTimingFunction>::encode):
(IPC::ArgumentCoder<CubicBezierTimingFunction>::decode):
(IPC::ArgumentCoder<StepsTimingFunction>::encode):
(IPC::ArgumentCoder<SpringTimingFunction>::encode):
(IPC::ArgumentCoder<PluginInfo>::encode):
(IPC::ArgumentCoder<PluginInfo>::decode):
(IPC::ArgumentCoder<ProtectionSpace>::encode):
(IPC::ArgumentCoder<ProtectionSpace>::decode):
(IPC::ArgumentCoder<Credential>::encode):
(IPC::ArgumentCoder<Credential>::decode):
(IPC::ArgumentCoder<Cursor>::encode):
(IPC::ArgumentCoder<Cursor>::decode):
(IPC::ArgumentCoder<ResourceError>::encode):
(IPC::ArgumentCoder<ResourceError>::decode):
(IPC::ArgumentCoder<DragData>::encode):
(IPC::ArgumentCoder<DragData>::decode):
(IPC::ArgumentCoder<CompositionUnderline>::encode):
(IPC::ArgumentCoder<CompositionUnderline>::decode):
(IPC::ArgumentCoder<FileChooserSettings>::encode):
(IPC::ArgumentCoder<FileChooserSettings>::decode):
(IPC::ArgumentCoder<TextCheckingRequestData>::encode):
(IPC::ArgumentCoder<TextCheckingRequestData>::decode):
(IPC::ArgumentCoder<TextCheckingResult>::encode):
(IPC::ArgumentCoder<TextCheckingResult>::decode):
(IPC::ArgumentCoder<UserStyleSheet>::encode):
(IPC::ArgumentCoder<UserStyleSheet>::decode):
(IPC::ArgumentCoder<ScrollableAreaParameters>::encode):
(IPC::ArgumentCoder<ScrollableAreaParameters>::decode):
(IPC::ArgumentCoder<FilterOperation>::encode):
(IPC::decodeFilterOperation):
(IPC::ArgumentCoder<FontAttributes>::encode):
(IPC::ArgumentCoder<FontAttributes>::decode):
(IPC::ArgumentCoder<WebCore::SerializedPlatformDataCueValue>::encode):
(IPC::ArgumentCoder<WebCore::SerializedPlatformDataCueValue>::decode):
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPopupItem.cpp:
(WebKit::WebPopupItem::encode const):
(WebKit::WebPopupItem::decode):
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebsiteData/WebsiteData.cpp:
(WebKit::WebsiteData::Entry::encode const):
(WebKit::WebsiteData::Entry::decode):
* Shared/curl/WebCoreArgumentCodersCurl.cpp:
(IPC::ArgumentCoder<ResourceError>::encodePlatformData):
(IPC::ArgumentCoder<ResourceError>::decodePlatformData):
(IPC::ArgumentCoder<ProtectionSpace>::encodePlatformData):
(IPC::ArgumentCoder<ProtectionSpace>::decodePlatformData):
* Shared/glib/InputMethodState.cpp:
(WebKit::InputMethodState::encode const):
(WebKit::InputMethodState::decode):
* Shared/glib/UserMessage.cpp:
(WebKit::UserMessage::encode const):
(WebKit::UserMessage::decode):
* Shared/mac/ColorSpaceData.mm:
(WebKit::ColorSpaceData::encode const):
(WebKit::ColorSpaceData::decode):
* Shared/mac/SecItemRequestData.cpp:
(WebKit::SecItemRequestData::encode const):
(WebKit::SecItemRequestData::decode):
* Shared/mac/WebCoreArgumentCodersMac.mm:
(IPC::ArgumentCoder<WebCore::ResourceRequest>::encodePlatformData):
(IPC::ArgumentCoder<WebCore::ResourceRequest>::decodePlatformData):
(IPC::ArgumentCoder<WebCore::CertificateInfo>::encode):
(IPC::ArgumentCoder<WebCore::CertificateInfo>::decode):
* Shared/soup/WebCoreArgumentCodersSoup.cpp:
(IPC::ArgumentCoder<SoupNetworkProxySettings>::encode):
(IPC::ArgumentCoder<SoupNetworkProxySettings>::decode):
* WebProcess/GPU/media/RemoteMediaPlayerState.h:
(WebKit::RemoteMediaPlayerState::encode const):
(WebKit::RemoteMediaPlayerState::decode):
* WebProcess/Network/NetworkProcessConnectionInfo.h:
(WebKit::NetworkProcessConnectionInfo::decode):
* WebProcess/Plugins/Plugin.cpp:
(WebKit::Plugin::Parameters::encode const):
(WebKit::Plugin::Parameters::decode):
* WebProcess/WebPage/RemoteLayerTree/PlatformCAAnimationRemote.mm:
(WebKit::PlatformCAAnimationRemote::KeyframeValue::encode const):
(WebKit::PlatformCAAnimationRemote::KeyframeValue::decode):
(WebKit::PlatformCAAnimationRemote::Properties::encode const):
(WebKit::PlatformCAAnimationRemote::Properties::decode):
2020-06-12 Wenson Hsieh <wenson_hsieh@apple.com>
[iPadOS] Focusing selects and color inputs should not bring up the software keyboard
https://bugs.webkit.org/show_bug.cgi?id=213149
<rdar://problem/64312450>
Reviewed by Darin Adler.
After <trac.webkit.org/r261658>, we now opt into bringing up the software keyboard when focusing color inputs
and select elements. This is due to the changes in `-_shouldShowAutomaticKeyboardUIIgnoringInputMode`, which
were intended to make us return `NO` for date and time inputs, but also unintentionally makes us return `YES`
on iPadOS.
This patch fixes the bug by returning `!WebKit::currentUserInterfaceIdiomIsPad()` instead, and additionally
preserves the existing behavior of now showing a color picker for `<input type="color" readonly>` on iPhone.
Test: fast/forms/ios/ipad/select-should-not-bring-up-software-keyboard.html
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView becomeFirstResponderForWebView]):
(-[WKContentView resignFirstResponderForWebView]):
(-[WKContentView _shouldShowAutomaticKeyboardUIIgnoringInputMode]):
(-[WKContentView _elementDidFocus:userIsInteracting:blurPreviousNode:activityStateChanges:userObject:]):
2020-06-12 Alex Christensen <achristensen@webkit.org>
Make API tests tolerant of our relatively new use of WebPageProxy::preconnectTo
https://bugs.webkit.org/show_bug.cgi?id=213144
Reviewed by Geofferey Garen.
* NetworkProcess/cocoa/NetworkSessionCocoa.h:
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(-[WKNetworkSessionDelegate URLSession:task:willPerformHTTPRedirection:newRequest:completionHandler:]):
(-[WKNetworkSessionDelegate URLSession:task:didCompleteWithError:]):
(-[WKNetworkSessionDelegate URLSession:dataTask:didReceiveResponse:completionHandler:]):
(WebKit::NetworkSessionCocoa::taskServerConnectionSucceeded):
(WebKit::NetworkSessionCocoa::taskReceivedBytes): Deleted.
Fix our logic that remembers successful client certificate connections.
If a task completes with no error (like a preconnect task does), consider that a successful connection.
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::preconnectTo):
Remove the check to not preconnect to loopback addresses.
This was making our tests different than the real internet and prevented me from landing a perfectly good test
in https://bugs.webkit.org/show_bug.cgi?id=213109
2020-06-12 Per Arne Vollan <pvollan@apple.com>
[iOS] Preferences are not being observed if the process pool is being created after activation
https://bugs.webkit.org/show_bug.cgi?id=213145
Reviewed by Brent Fulgham.
If the app is creating the Web process pool after being activated, preferences will not be observed until the app is being backgrounded
and then foregrounded again, since the preference observer is initialized when handling the app activation notification.
No new tests, since an API test would make sense in this case, but API tests aren't run on iOS.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
2020-06-12 Takashi Komori <Takashi.Komori@sony.com>
[Curl] Implement functions to use ResourceLoadStatistics.
https://bugs.webkit.org/show_bug.cgi?id=207692
Reviewed by Don Olmstead.
Implement functions which are required to implement ResourceLoadStatistics for Curl port.
In NetworkDataTaskCurl.cpp we check if we should block cookies and block if needed.
Tests: http/tests/resourceLoadStatistics/
* NetworkProcess/NetworkDataTask.cpp:
(WebKit::NetworkDataTask::create):
* NetworkProcess/curl/NetworkDataTaskCurl.cpp:
(WebKit::NetworkDataTaskCurl::NetworkDataTaskCurl):
(WebKit::NetworkDataTaskCurl::createCurlRequest):
(WebKit::NetworkDataTaskCurl::willPerformHTTPRedirection):
(WebKit::NetworkDataTaskCurl::blockCookies):
(WebKit::NetworkDataTaskCurl::unblockCookies):
(WebKit::NetworkDataTaskCurl::shouldBlockCookies):
(WebKit::NetworkDataTaskCurl::isThirdPartyRequest):
* NetworkProcess/curl/NetworkDataTaskCurl.h:
* NetworkProcess/curl/NetworkSessionCurl.cpp:
(WebKit::NetworkSessionCurl::NetworkSessionCurl):
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreGetAllStorageAccessEntries):
* UIProcess/API/C/WKWebsiteDataStoreRef.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::parameters):
2020-06-12 Antoine Quint <graouts@webkit.org>
Double tap to zoom out doesn't work after upgrading to iOS 13
https://bugs.webkit.org/show_bug.cgi?id=205158
<rdar://problem/205158>
Reviewed by Wenson Hsieh.
Test: fast/events/ios/fast-click-double-tap-to-zoom-in-on-text-and-then-again-to-zoom-out.html
In order to determine whether a significant zoom to happen, we should use the current page scale
factor and not the initial page scale factor.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _handleSmartMagnificationInformationForPotentialTap:renderRect:fitEntireRect:viewportMinimumScale:viewportMaximumScale:nodeIsRootLevel:]):
2020-06-12 Alex Christensen <achristensen@webkit.org>
Add WKUserScript initializer SPI with alternative names
https://bugs.webkit.org/show_bug.cgi?id=213096
Reviewed by Youenn Fablet.
And deprecate the existing ones so we can remove them soon.
* UIProcess/API/Cocoa/WKUserScript.mm:
(-[WKUserScript _initWithSource:injectionTime:forMainFrameOnly:includeMatchPatternStrings:excludeMatchPatternStrings:associatedURL:contentWorld:deferRunningUntilNotification:]):
* UIProcess/API/Cocoa/WKUserScriptPrivate.h:
2020-06-12 Carlos Garcia Campos <cgarcia@igalia.com>
Unreviewed. Fix GTK4 build
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseSynthesizeKeyEvent):
2020-06-11 Sam Weinig <weinig@apple.com>
Document.currentScript does not work for SVGScriptElements
https://bugs.webkit.org/show_bug.cgi?id=213104
Reviewed by Yusuke Suzuki.
* WebProcess/InjectedBundle/API/gtk/DOM/WebKitDOMDocumentGtk.cpp:
(webkit_dom_document_get_current_script):
Update to account for change in Document::currentScript() now returning an
Element* that can be either an HTMLScriptElement or an SVGScriptElement. To
keep API compatibility, only return non-null for HTMLScriptElements.
2020-06-12 Charlie Turner <cturner@igalia.com>
[GTK] Add an internal API to run javascript without forced user gestures
https://bugs.webkit.org/show_bug.cgi?id=212969
Reviewed by Carlos Garcia Campos.
To be used by tests in a follow-up commit.
* UIProcess/API/glib/WebKitWebView.cpp:
(webkitWebViewRunJavaScriptWithParams): Factor out the glue to run
JavaScript in the page.
(webkitWebViewRunJavascriptWithoutForcedUserGestures): Add a new
internal API for use by tests.
(webkit_web_view_run_javascript): Modified to use factored out
glue.
* UIProcess/API/glib/WebKitWebViewInternal.h: Added. Place to keep
API test-specific internal APIs.
2020-06-12 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Make WebDriver work
https://bugs.webkit.org/show_bug.cgi?id=212316
Reviewed by Adrian Perez de Castro.
Fix UIClient::setWindowFrame to wait for surface size-changed signal after resizing the window and implement
maximize, minimize and restore windows for GTK4.
* UIProcess/API/glib/WebKitUIClient.cpp:
* UIProcess/API/gtk/WebKitWebViewGtk.cpp:
(surfaceStateChangedCallback):
(webkitWebViewMonitorWindowState):
(webkitWebViewMaximizeWindow):
(webkitWebViewMinimizeWindow):
(webkitWebViewRestoreWindow):
2020-06-11 David Kilzer <ddkilzer@apple.com>
[IPC] Add WTF::EnumTraits<> for every enum type used in IPC
<https://webkit.org/b/213093>
Reviewed by Darin Adler.
Summary:
- Change underlying type of enum class to `bool` when there are
only two values. In some cases, reorder the two values so the
mapping to 0 and 1 makes more sense. Converting every enum to
an enum class is not a goal of this patch, so some two-value
enums stil have WTF::EnumTraits<> defined as noted below.
- Add WTF::EnumTraits<> for the remaining enum types that are
used by IPC::Encoder::encodeEnum() and
IPC::Decoder::decodeEnum() so that WTF::isValidEnum<>() checks
may be added next.
- Add #include <WebCore/LibWebRTCEnumTraits.h> as needed.
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/GPUProcess.h:
* NetworkProcess/NetworkLoadParameters.h:
* Shared/API/APIObject.h:
* Shared/ContextMenuContextData.h:
* Shared/DrawingAreaInfo.h:
* Shared/FocusedElementInformation.h:
* Shared/PrintInfo.h:
(WTF::EnumTraits<WebKit::PrintInfo::PrintMode>):
- Define this since PrintMode is not an enum class.
* Shared/RTCPacketOptions.cpp:
* Shared/SessionState.h:
* Shared/TouchBarMenuItemData.h:
* Shared/WebEvent.h:
* Shared/WebPopupItem.cpp:
(WebKit::WebPopupItem::WebPopupItem):
- Fix reference to WebPopupItem::Type enum.
* Shared/WebPopupItem.h:
(WebKit::WebPopupItem::Type):
- Convert from enum to enum class.
* Shared/WebsiteData/WebsiteDataType.h:
* Shared/glib/InputMethodState.h:
* Shared/glib/UserMessage.h:
* Shared/mac/ColorSpaceData.mm:
* Shared/mac/SecItemRequestData.h:
* UIProcess/mac/WebPopupMenuProxyMac.mm:
(WebKit::WebPopupMenuProxyMac::populate):
- Fix reference to WebPopupItem::Type enum.
* WebProcess/Plugins/Plugin.cpp:
* WebProcess/WebCoreSupport/WebPopupMenu.cpp:
(WebKit::WebPopupMenu::populateItems):
- Fix reference to WebPopupItem::Type enum.
* WebProcess/WebPage/RemoteLayerTree/PlatformCAAnimationRemote.h:
2020-06-11 Alex Christensen <achristensen@webkit.org>
Re-enable download resume tests
https://bugs.webkit.org/show_bug.cgi?id=213098
<rdar://problem/63512518>
Reviewed by Geoffrey Garen.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(-[WKNetworkSessionDelegate URLSession:downloadTask:didResumeAtOffset:expectedTotalBytes:]):
2020-06-11 Beth Dakin <bdakin@apple.com>
Replace instances of whitelist in WebCore with allowlist
https://bugs.webkit.org/show_bug.cgi?id=213068
Reviewed by Tim Horton.
* NetworkProcess/NetworkConnectionToWebProcess.cpp:
(WebKit::NetworkConnectionToWebProcess::addOriginAccessWhitelistEntry):
(WebKit::NetworkConnectionToWebProcess::removeOriginAccessWhitelistEntry):
(WebKit::NetworkConnectionToWebProcess::resetOriginAccessWhitelists):
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<UserStyleSheet>::encode):
(IPC::ArgumentCoder<UserStyleSheet>::decode):
* WebProcess/InjectedBundle/InjectedBundle.cpp:
(WebKit::InjectedBundle::addOriginAccessWhitelistEntry):
(WebKit::InjectedBundle::removeOriginAccessWhitelistEntry):
(WebKit::InjectedBundle::resetOriginAccessWhitelists):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-06-11 Jonathan Bedard <jbedard@apple.com>
WebKit: Guard requestDocumentContext with HAVE(UI_WK_DOCUMENT_CONTEXT)
https://bugs.webkit.org/show_bug.cgi?id=213076
<rdar://problem/64254698>
Reviewed by Tim Horton.
No new tests, behavior unchanged.
* Platform/spi/ios/UIKitSPI.h:
* UIProcess/API/ios/WKWebViewTestingIOS.mm:
(-[WKWebView _requestDocumentContext:completionHandler:]): Guard requestDocumentContext with HAVE(UI_WK_DOCUMENT_CONTEXT).
(-[WKWebView _adjustSelectionWithDelta:completionHandler:]): Ditto.
2020-06-11 Jonathan Bedard <jbedard@apple.com>
WebKit: Add UIKit SPI for watchOS and tvOS
https://bugs.webkit.org/show_bug.cgi?id=213077
<rdar://problem/64255131>
Reviewed by Tim Horton.
No new tests, behavior unchanged.
* Platform/spi/ios/UIKitSPI.h:
2020-06-11 David Kilzer <ddkilzer@apple.com>
[IPC] Adopt enum class for DragSourceAction
<https://webkit.org/b/212885>
<rdar://problem/64094134>
Reviewed by Darin Adler.
Summary:
- Convert argument parameters from uint64_t to
OptionSet<DragSourceAction>.
- Convert DragSourceAction to OptionSet<DragSourceAction>.
* Scripts/webkit/messages.py:
- Map DragSourceAction enum to DragActions.h header.
* UIProcess/WebPageProxy.h:
* UIProcess/ios/DragDropInteractionState.h:
* UIProcess/ios/DragDropInteractionState.mm:
(WebKit::shouldUseDragImageToCreatePreviewForDragSource):
(WebKit::shouldUseVisiblePathToCreatePreviewForDragSource):
(WebKit::shouldUseTextIndicatorToCreatePreviewForDragSource):
(WebKit::canUpdatePreviewForActiveDragSource):
(WebKit::DragDropInteractionState::anyActiveDragSourceIs const):
(WebKit::DragDropInteractionState::stageDragItem):
(WebKit::DragDropInteractionState::hasStagedDragSource const):
(WebKit::DragDropInteractionState::updatePreviewsForActiveDragSources):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _startDrag:item:]):
(-[WKContentView _allowedDragSourceActions]):
(-[WKContentView dragInteraction:willAnimateLiftWithAnimator:session:]):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::requestDragStart):
(WebKit::WebPageProxy::requestAdditionalItemsForDragSession):
* WebProcess/WebCoreSupport/WebDragClient.cpp:
(WebKit::WebDragClient::dragSourceActionMaskForPoint):
* WebProcess/WebCoreSupport/WebDragClient.h:
* WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::allowedDragSourceActions const):
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::requestDragStart):
(WebKit::WebPage::requestAdditionalItemsForDragSession):
2020-06-11 Eric Carlson <eric.carlson@apple.com>
[Cocoa] MediaUsageManagerCocoa should use a different USVideoUsage initializer
https://bugs.webkit.org/show_bug.cgi?id=213058
Reviewed by Jer Noble.
No new tests, tested manually.
* UIProcess/Media/cocoa/MediaUsageManagerCocoa.mm:
(WebKit::MediaUsageManagerCocoa::reset): Drive-by fix: don't call -stop unless the
element was playing.
(WebKit::MediaUsageManagerCocoa::updateMediaUsage): Use a different initializer so
we can pass in the state.
2020-06-11 ChangSeok Oh <changseok@webkit.org>
[GTK] Implement button-press-event, button-release-event, and absolute-axis-event of GAMEPAD API.
https://bugs.webkit.org/show_bug.cgi?id=133850
Reviewed by Carlos Garcia Campos.
* UIProcess/Gamepad/UIGamepadProvider.cpp: Add the OS(LINUX) guard.
* UIProcess/Gamepad/gtk/UIGamepadProviderGtk.cpp:
(WebKit::getWebPageProxy): Build fix. WebKitWebViewBase is used, not WebKitWebViewPrivate.
* UIProcess/Gamepad/manette/UIGamepadProviderManette.cpp: Add the OS(LINUX) guard.
2020-06-10 Said Abou-Hallawa <sabouhallawa@apple.com>
[macOS]: The File Picker of the <input> file element should show the selection filter
https://bugs.webkit.org/show_bug.cgi?id=212485
Reviewed by Darin Adler.
Add two private helpers to WKOpenPanelParameters.
* UIProcess/API/Cocoa/WKOpenPanelParameters.mm:
(-[WKOpenPanelParameters _allowedFileExtensions]):
Returns an array of strings representing the allowed extensions.
(-[WKOpenPanelParameters _allowedFileExtensionsDescription]):
Returns a string representing a description for the allowed extensions.
* UIProcess/API/Cocoa/WKOpenPanelParametersPrivate.h:
2020-06-10 Beth Dakin <bdakin@apple.com>
Replace instances of blacklist in WebCore with blocklist
https://bugs.webkit.org/show_bug.cgi?id=213064
Reviewed by Tim Horton.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<UserStyleSheet>::encode):
(IPC::ArgumentCoder<UserStyleSheet>::decode):
2020-06-10 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] editing/pasteboard/dom-paste tests consistently time out
https://bugs.webkit.org/show_bug.cgi?id=213061
<rdar://problem/64211161>
Reviewed by Darin Adler.
On recent versions of iOS, system changes have moved the timing of the call to
`-requestAutocorrectionContextWithCompletionHandler:` when focusing an editable element. This results in a
consistently reproducible deadlock between the UI process and web process if a page tries to request
programmatic clipboard access during the click event (importantly, in the same runloop as element focus).
What ends up happening is that the autocorrection context is requested by the UI process shortly after the
(unbounded) sync IPC message for the DOM paste arrives, so we end up waiting forever for the sync-IPC-blocked
web process to respond. Normally this just results in a 1 second hang, and the DOM paste proceeds as normal.
However these layout tests purposefully opt out of sync IPC timeouts (precisely so that we can catch bugs like
this). See below for more details.
* UIProcess/ios/WKContentViewInteraction.h:
Keep track of the most recent autocorrection context that arrived in the UI process.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView cleanUpInteraction]):
Clear out the autocorrection context, in case the web process crashes.
(-[WKContentView requestAutocorrectionContextWithCompletionHandler:]):
If the autocorrection context is requested while handling a DOM paste, we know that the latest autocorrection
context must be up to date, since the autocorrection context is computed and sent immediately before a DOM paste
request. Use this to immediately invoke the completion handler, without any sync IPC back to the web process.
(-[WKContentView _handleAutocorrectionContext:]):
2020-06-10 Hiro (mzp) Mizuno <mzp@apple.com>
[iOS] Option + Forward Delete should delete next word
https://bugs.webkit.org/show_bug.cgi?id=213062
rdar://64225458
Reviewed by Daniel Bates.
Implement support for pressing Option + Forward Delete to delete the next word. This makes text editing
in WebKit more closely match the platform conventions.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView canPerformAction:withSender:]): Handle _deleteForwardByWord.
(-[WKContentView _deleteForwardByWord]): Added.
2020-06-10 Geoffrey Garen <ggaren@apple.com>
Some style improvements to main thread code
https://bugs.webkit.org/show_bug.cgi?id=213051
Reviewed by Darin Adler.
Updated for rename.
* Shared/Cocoa/WebKit2InitializeCocoa.mm:
(WebKit::runInitializationCode):
(WebKit::InitializeWebKit2):
* Shared/WebKit2Initialize.cpp:
(WebKit::InitializeWebKit2):
2020-06-10 Jonathan Bedard <jbedard@apple.com>
WebKit:Add PepperUICoreSPI.h (Follow-up fix 2)
https://bugs.webkit.org/show_bug.cgi?id=212996
<rdar://problem/64182627>
Unreviewed build fix.
* UIProcess/ios/forms/WKNumberPadViewController.mm:
(-[WKNumberPadViewController addContentViewAnimations:]): Correct APPLE_INTERNAL_SDK macro.
* UIProcess/ios/forms/WKQuickboardListViewController.mm:
(-[WKQuickboardListViewController initWithDelegate:]): Ditto.
2020-06-10 Jonathan Bedard <jbedard@apple.com>
WebKit:Add PepperUICoreSPI.h (Follow-up fix)
https://bugs.webkit.org/show_bug.cgi?id=212996
<rdar://problem/64182627>
Unreviewed build fix.
* Platform/spi/watchos/PepperUICoreSPI.h: Add PUICPickerView headers.
2020-06-10 David Kilzer <ddkilzer@apple.com>
REGRESSION (r262858): Fix macOS build when building EndowmentStateTracker.mm
* UIProcess/EndowmentStateTracker.mm:
- Move #if PLATFORM(IOS_FAMILY) to fix macOS build.
2020-06-10 Per Arne Vollan <pvollan@apple.com>
[Cocoa] Return early from preference change notification handler if there is no observer
https://bugs.webkit.org/show_bug.cgi?id=213050
Reviewed by Darin Adler.
Currently, the observer member is checked inside the loop iterating over all the keys,
but this check should be moved outside of the loop.
No new tests, since this is not a change in behavior.
* UIProcess/Cocoa/PreferenceObserver.mm:
(-[WKUserDefaults _notifyObserversOfChangeFromValuesForKeys:toValuesForKeys:]):
2020-06-10 Brady Eidson <beidson@apple.com>
Crash growing a CFData with incremental PDF loading.
<rdar://problem/63670403> and https://bugs.webkit.org/show_bug.cgi?id=213035
Reviewed by Alex Christensen.
No test - No reliable way to trigger.
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::ensureDataBufferLength): When you call CFDataCreateMutable with a size other than 0,
your data object can never grow beyond that size.
And, in fact, CFData will crash on purpose when this is attempted.
So always create our mutable CFDatas with size 0 even if we immediately plan to grow them.
2020-06-10 Jonathan Bedard <jbedard@apple.com>
WebKit: Add ClockKitSPI.h
https://bugs.webkit.org/show_bug.cgi?id=212994
<rdar://problem/64181092>
Reviewed by Wenson Hsieh.
No new tests, behavior unchanged.
* Platform/spi/watchos: Added.
* Platform/spi/watchos/ClockKitSPI.h: Added.
* UIProcess/ios/forms/WKTimePickerViewController.mm: Import ClockKitSPI.h and UIKitSPI.h.
* WebKit.xcodeproj/project.pbxproj: Include ClockKitSPI.
2020-06-10 Brian Burg <bburg@apple.com>
WebDriver on non-iOS ports cannot perform ActionChain which has scrolling down to the element and click it
https://bugs.webkit.org/show_bug.cgi?id=208232
<rdar://problem/59859491>
Reviewed by Devin Rousso.
* UIProcess/Automation/mac/WebAutomationSessionMac.mm:
(WebKit::WebAutomationSession::platformSimulateMouseInteraction):
The provided coordinates are in LayoutViewport coordinate system, which does
not take topContentInset into account. Add back in the topContentInset
when translating to flipped window coordinates.
* WebProcess/Automation/WebAutomationSessionProxy.cpp:
(WebKit::WebAutomationSessionProxy::computeElementLayout):
The calculations of element bounds and IVCP have an incorrect handling of
root coordinates as contents/absolute coordinates. Add the missing conversion.
2020-06-10 Brent Fulgham <bfulgham@apple.com>
[iOS] Update sandbox rules for correct sanitizer paths in current OS releases
https://bugs.webkit.org/show_bug.cgi?id=213040
<rdar://problem/64187043>
Reviewed by Per Arne Vollan.
Update the sandbox rules on iOS platforms to allow access to the
new system locations for Asan libraries.
* Resources/SandboxProfiles/ios/com.apple.WebKit.GPU.sb:
* Resources/SandboxProfiles/ios/com.apple.WebKit.Networking.sb:
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-06-10 Jer Noble <jer.noble@apple.com>
Catalyst WebKit apps continue to play audio after quitting
https://bugs.webkit.org/show_bug.cgi?id=212981
<rdar://problem/57089471>
Reviewed by Chris Dumez.
RunningBoard will track whether a process (and it's associated child processes) are "user-visible" by
granting those processes an "endowment", similar to the existing endowment for application visibility.
Track changes to these endowments using a RBSProcessMonitor, and for Catalyst apps, react to losing the
"user-visible" endowment by suspending all media playback.
Move all the endowment-related code into a new class, EndowmentStateTracker, which is a singleton object
that can have mulitple WebPageProxy clients. Update references to the previously global static
isApplicationForeground() function to refer to EndowmentStateTracker instead.
Drive-by fix: move some iOS-only functions in WebPageProxy.cpp into WebPageProxyIOS.mm.
* UIProcess/ApplicationStateTracker.h:
* UIProcess/ApplicationStateTracker.mm:
(WebKit::ApplicationStateTracker::ApplicationStateTracker):
(WebKit::isApplicationForeground): Deleted.
* UIProcess/EndowmentStateTracker.h: Added.
(WebKit::EndowmentStateTracker::isVisible const):
(WebKit::EndowmentStateTracker::isUserFacing const):
* UIProcess/EndowmentStateTracker.mm: Added.
(WebKit::handleForPID):
(WebKit::EndowmentStateTracker::isApplicationForeground):
(WebKit::EndowmentStateTracker::isApplicationUserFacing):
(WebKit::EndowmentStateTracker::singleton):
(WebKit::EndowmentStateTracker::EndowmentStateTracker):
(WebKit::EndowmentStateTracker::addClient):
(WebKit::EndowmentStateTracker::removeClient):
(WebKit::EndowmentStateTracker::setIsUserFacing):
(WebKit::EndowmentStateTracker::setIsVisible):
* UIProcess/WebPageProxy.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
(WebKit::WebPageProxy::~WebPageProxy):
(WebKit::WebPageProxy::processWillBecomeSuspended): Deleted.
(WebKit::WebPageProxy::processWillBecomeForeground): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::isApplicationVisible):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::processWillBecomeSuspended):
(WebKit::WebPageProxy::processWillBecomeForeground):
(WebKit::WebPageProxy::isUserFacingChanged):
(WebKit::WebPageProxy::isVisibleChanged):
* WebKit.xcodeproj/project.pbxproj:
2020-06-10 Antoine Quint <graouts@webkit.org>
Subframes should not autosize independently
https://bugs.webkit.org/show_bug.cgi?id=212984
<rdar://problem/64175493>
Reviewed by Simon Fraser.
Ensure we only call FrameView::enableSizeToContentAutoSizeMode() for the main frame.
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::transitionToCommittedForNewPage):
2020-06-10 Jonathan Bedard <jbedard@apple.com>
WebKit:Add PepperUICoreSPI.h
https://bugs.webkit.org/show_bug.cgi?id=212996
<rdar://problem/64182627>
Reviewed by Tim Horton.
No new tests, behavior unchanged.
* Platform/spi/watchos: Added.
* Platform/spi/watchos/PepperUICoreSPI.h: Added.
* UIProcess/Cocoa/WKSafeBrowsingWarning.mm: Include PepperUICoreSPI header.
* UIProcess/ios/WKScrollView.mm: Ditto.
* UIProcess/ios/forms/WKDatePickerViewController.mm: Remove headers duplicated in PepperUICoreSPI.h.
* UIProcess/ios/forms/WKFocusedFormControlView.mm: Ditto.
* UIProcess/ios/forms/WKNumberPadView.mm: Include PepperUICoreSPI header.
* UIProcess/ios/forms/WKNumberPadViewController.h: Remove headers duplicated in PepperUICoreSPI.h.
* UIProcess/ios/forms/WKNumberPadViewController.mm: Ditto.
(-[WKNumberPadViewController addContentViewAnimations:]): Limit addAnimation to Internal builds.
* UIProcess/ios/forms/WKQuickboardListViewController.h: Include PepperUICoreSPI header.
* UIProcess/ios/forms/WKQuickboardListViewController.mm: Remove headers duplicated in PepperUICoreSPI.h.
(-[WKQuickboardListViewController initWithDelegate:]):
* UIProcess/ios/forms/WKSelectMenuListViewController.mm: Remove headers duplicated in PepperUICoreSPI.h.
* UIProcess/ios/forms/WKTextInputListViewController.mm: Ditto.
* WebKit.xcodeproj/project.pbxproj: Add PepperUICoreSPI.h.
2020-06-10 Per Arne Vollan <pvollan@apple.com>
[iOS] Add entitlement for PAC exceptions
https://bugs.webkit.org/show_bug.cgi?id=213027
<rdar://problem/63985737>
Reviewed by Brent Fulgham.
On iOS, add entitlement for PAC exceptions.
* Scripts/process-entitlements.sh:
2020-06-10 James Savage <james.savage@apple.com>
Swift Overlay API refinements
https://bugs.webkit.org/show_bug.cgi?id=212939
<rdar://problem/64140013>
Reviewed by Darin Adler.
This patch cleans up some inconsistencies in the various Swift projections,
and modifies JavaScript evaluation callbacks to be optional to account for
scripts which may not evaluate to a useful result (such as ones that simply
modify the DOM without reporting a status). It also standardizes on using
completionHandler as an argument label to match the rest of WebKit API, and
the majority of the Swift API refinements.
It also fixes a module issue where the WebKit.h umbrella was missing an
import of WKScriptReplyWithMessageHandler.h, making the type unusable from
Swift source files/
* Shared/API/Cocoa/WebKit.h: Add missing header.
* UIProcess/API/Cocoa/WebKitSwiftOverlay.swift:
(callAsyncJavaScript(_:arguments:in:completionHandler:)): Rename completion
argument, and make optional. Note: Optional closures are implicitly
@escaping, making the annotation redundant and incompatible, so it must
be removed. In Swift, Optional.map()'s closure is invoked iff the Optional
contains some value, and the result of that transform is returned, else
nil is propagated. This allows us to apply the closure transform if the
client value is non-nil, without using intermediary variables or if/let.
(evaluateJavaScript(_:in:completionHandler:)): Ditto.
* SwiftOverlay/Tests/WebKitTests.swift:
(WebKitTests.testAPI): Update names to reflect changes.
* SwiftOverlay/WebKitSwiftOverlay.xcodeproj/xcshareddata/xcschemes/Everything (Catalyst).xcscheme:
Ensure that the unit test files get built as part of Build & Run, so that
any changes to names get flagged without even running the tests.
* SwiftOverlay/WebKitSwiftOverlay.xcodeproj/xcshareddata/xcschemes/Everything.xcscheme:
Ditto.
2020-06-10 Kate Cheney <katherine_cheney@apple.com>
ASSERTION FAILED: suspendedState == State::Running in WebKit::WebResourceLoadStatisticsStore::suspend
https://bugs.webkit.org/show_bug.cgi?id=213031
<rdar://problem/64131092>
Reviewed by Chris Dumez.
There is a race condition if two calls are made to
WebResourceLoadStatisticsStore::suspend() simultanously.
WebResourceLoadStatisticsStore::resume() notifies the background
thread to continue and releases a lock, then sometimes the second call
to suspend obtains the lock and updates the suspendedState variable
before the background thread resumes and checks the result. This patch allows the
suspendedState to be either State::Running or State::WillSuspend to
account for this.
No new tests, this should fix http/wpt/service-workers/service-worker-spinning-fetch.https.html.
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::resourceLoadStatisticsUpdated):
(WebKit::WebResourceLoadStatisticsStore::suspend):
2020-06-10 Alex Christensen <achristensen@webkit.org>
Revert r259770
https://bugs.webkit.org/show_bug.cgi?id=210097
<rdar://problem/64175992>
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::decidePolicyForNavigationAction):
2020-06-10 Alex Christensen <achristensen@webkit.org>
Add some null checks in PDFPlugin::updatePageAndDeviceScaleFactors
https://bugs.webkit.org/show_bug.cgi?id=213009
<rdar://problem/63756100>
Reviewed by Chris Dumez.
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::updatePageAndDeviceScaleFactors):
2020-06-09 Fujii Hironori <Hironori.Fujii@sony.com>
Unreviewed, reverting r262791.
WinCairo WebKit1 is crashing.
Reverted changeset:
"[Curl] Implement functions to use ResourceLoadStatistics."
https://bugs.webkit.org/show_bug.cgi?id=207692
https://trac.webkit.org/changeset/262791
2020-06-09 Tim Horton <timothy_horton@apple.com>
Stop using the wrong LaunchServices SPI for getting an app's localized name
https://bugs.webkit.org/show_bug.cgi?id=213003
<rdar://problem/64169000>
Reviewed by Sam Weinig.
* UIProcess/ios/WKActionSheetAssistant.mm:
(-[WKActionSheetAssistant _appendAppLinkOpenActionsForURL:actions:elementInfo:]):
2020-06-09 Mark Lam <mark.lam@apple.com>
Disambiguate the OverridesGetPropertyNames structure flag
https://bugs.webkit.org/show_bug.cgi?id=212909
<rdar://problem/63823557>
Reviewed by Saam Barati.
* WebProcess/Plugins/Netscape/JSNPObject.h:
2020-06-09 Dean Jackson <dino@apple.com>
Stop using discriminatory names for WebGL and Plugin blocking
https://bugs.webkit.org/show_bug.cgi?id=213000
Reviewed by Simon Fraser.
* Shared/Plugins/Netscape/mac/PluginInformationMac.mm:
(WebKit::getPlatformPluginModuleInformation):
* UIProcess/API/C/mac/WKContextPrivateMac.mm:
(WKContextIsPlugInUpdateAvailable):
(WKContextShouldBlockWebGL):
(WKContextShouldSuggestBlockWebGL):
* UIProcess/Plugins/mac/PluginInfoStoreMac.mm:
(WebKit::PluginInfoStore::defaultLoadPolicyForPlugin):
2020-06-09 Brian Burg <bburg@apple.com>
[Cocoa] Element Send Keys can hang when an automation window becomes unfocused
https://bugs.webkit.org/show_bug.cgi?id=212985
<rdar://problem/56354471>
Reviewed by Devin Rousso.
* UIProcess/Automation/mac/WebAutomationSessionMac.mm:
(WebKit::WebAutomationSession::sendSynthesizedEventsToPage):
Make sure to focus the window and set first responder so sending keys doesn't hang.
2020-06-09 Jonathan Bedard <jbedard@apple.com>
WebKit: Import NSURLConnectionSPI.h instead of CFNSURLConnection.h
https://bugs.webkit.org/show_bug.cgi?id=212980
<rdar://problem/64172694>
Unreviewed build fix.
No new tests, behavior unchanged.
* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
2020-06-09 Dean Jackson <dino@apple.com>
REGRESSION: [Safari Mojave for High Sierra] Accessing some of the featured pages on apple.com causes the webpage to crash
https://bugs.webkit.org/show_bug.cgi?id=212940
Reviewed by Tim Horton.
The code to use the singleton for a SwitchingGPUClient was assuming it
has always been set, which was not the case when
ENABLE(WEBPROCESS_WINDOWSERVER_BLOCKING) was not true.
* WebProcess/cocoa/WebProcessCocoa.mm: Set the singleton unconditionally.
(WebKit::WebProcess::platformInitializeProcess):
2020-06-09 Wenson Hsieh <wenson_hsieh@apple.com>
REGRESSION (r260820): [macCatalyst] Web process crashes when uploading a file
https://bugs.webkit.org/show_bug.cgi?id=212976
<rdar://problem/64033186>
Reviewed by Tim Horton.
On Mac Catalyst, we attempt to take a sandbox extension for `com.apple.frontboard.systemappservices` when
uploading a file; this service does not exist in Catalyst, and we consequently encounter a release assertion
under `WebPage::didChooseFilesForOpenPanelWithDisplayStringAndIcon`.
Fix this by introducing `HAVE(FRONTBOARD_SYSTEM_APP_SERVICES)`, and using it to guard codepaths that attempt to
grab a mach extension to this service. (As a followup, we should consider adopting this in other places that
currently reference `com.apple.frontboard.systemappservices`).
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didChooseFilesForOpenPanelWithDisplayStringAndIcon):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::didChooseFilesForOpenPanelWithDisplayStringAndIcon):
2020-06-09 Jonathan Bedard <jbedard@apple.com>
WebKit: Support watchOS and tvOS in xcconfigs
https://bugs.webkit.org/show_bug.cgi?id=212977
<rdar://problem/64170532>
Reviewed by Tim Horton.
No new tests, behavior unchanged.
* Configurations/Base.xcconfig: Add tvOS and watchOS major version macros.
* Configurations/BaseTarget.xcconfig: Include tvOS and watchOS framework stubs.
* Configurations/BaseXPCService.xcconfig: Generalize excluded files for all embedded platforms.
* Configurations/GPUService.xcconfig: Generalize INFOPLIST for embedded platforms.
* Configurations/NetworkService.xcconfig: Ditto.
* Configurations/WebContentService.xcconfig: Ditto.
2020-06-09 Alex Christensen <achristensen@webkit.org>
Expose PDF file URL in shouldAllowPDFToOpenFromFrame SPI
https://bugs.webkit.org/show_bug.cgi?id=212974
Reviewed by Tim Horton.
* UIProcess/API/APIUIClient.h:
(API::UIClient::confirmPDFOpening):
* UIProcess/API/Cocoa/WKUIDelegatePrivate.h:
* UIProcess/Cocoa/UIDelegate.h:
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::setDelegate):
(WebKit::UIDelegate::UIClient::confirmPDFOpening):
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::savePDFToTemporaryFolderAndOpenWithNativeApplication):
(WebKit::WebPageProxy::openPDFFromTemporaryFolderWithNativeApplication):
2020-06-09 Takashi Komori <Takashi.Komori@sony.com>
[Curl] Implement functions to use ResourceLoadStatistics.
https://bugs.webkit.org/show_bug.cgi?id=207692
Reviewed by Don Olmstead.
Implement functions which are required to implement ResourceLoadStatistics for Curl port.
In NetworkDataTaskCurl.cpp we check if we should block cookies and block if needed.
Tests: http/tests/resourceLoadStatistics/
* NetworkProcess/NetworkDataTask.cpp:
(WebKit::NetworkDataTask::create):
* NetworkProcess/curl/NetworkDataTaskCurl.cpp:
(WebKit::NetworkDataTaskCurl::NetworkDataTaskCurl):
(WebKit::NetworkDataTaskCurl::createCurlRequest):
(WebKit::NetworkDataTaskCurl::willPerformHTTPRedirection):
(WebKit::NetworkDataTaskCurl::blockCookies):
(WebKit::NetworkDataTaskCurl::unblockCookies):
(WebKit::NetworkDataTaskCurl::shouldBlockCookies):
(WebKit::NetworkDataTaskCurl::isThirdPartyRequest):
* NetworkProcess/curl/NetworkDataTaskCurl.h:
* NetworkProcess/curl/NetworkSessionCurl.cpp:
(WebKit::NetworkSessionCurl::NetworkSessionCurl):
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreGetAllStorageAccessEntries):
* UIProcess/API/C/WKWebsiteDataStoreRef.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::parameters):
2020-06-09 Lauro Moura <lmoura@igalia.com>
[WebDriver][WPE] Follow up WPE fix after r262703
https://bugs.webkit.org/show_bug.cgi?id=212948
Reviewed by Carlos Garcia Campos.
Covered by existing tests.
* UIProcess/Automation/wpe/WebAutomationSessionWPE.cpp:
(WebKit::WebAutomationSession::platformWebModifiersFromRaw):
2020-06-08 David Kilzer <ddkilzer@apple.com>
[IPC] Adopt enum class for IPC::CFType
<https://webkit.org/b/212921>
<rdar://problem/64134506>
Reviewed by Darin Adler.
Summary:
- Convert IPC::CFType to enum class.
- Rule for new enum names is to remove "Ref" suffix from the
actual type name.
- Special enum CFType::Null renamed to CFType::Nullptr to
differentiate from CFType::CFNull objects.
* Shared/cf/ArgumentCodersCF.cpp:
(IPC::tokenNullTypeRef): Delete.
(IPC::tokenNullptrTypeRef): Add.
- Rename tokenNullTypeRef() to tokenNullptrTypeRef().
(IPC::CFType):
- Convert to enum class.
(IPC::typeFromCFTypeRef):
(IPC::encode):
- Switch from encodeEnum() to operator<<().
(IPC::decode):
- Switch from decodeEnum() to decode().
(WTF::EnumTraits<IPC::CFType>): Add.
* Shared/cf/ArgumentCodersCF.h:
(IPC::tokenNullTypeRef): Delete.
(IPC::tokenNullptrTypeRef): Add.
- Rename tokenNullTypeRef() to tokenNullptrTypeRef().
* Shared/mac/WebCoreArgumentCodersMac.mm:
(IPC::ArgumentCoder<WebCore::ResourceRequest>::encodePlatformData):
(IPC::ArgumentCoder<WebCore::ResourceRequest>::decodePlatformData):
- Switch to use tokenNullptrTypeRef().
* mac/WebKit2.order:
- Update symbol name after renaming tokenNullTypeRef().
2020-06-08 Kate Cheney <katherine_cheney@apple.com>
ResourceLoadStatisticsDatabaseStore::domainIDFromString failed, error message: bad parameter or other API misuse
https://bugs.webkit.org/show_bug.cgi?id=212916
<rdar://problem/64127238>
Reviewed by Sihui Liu.
Make sure m_domainIDFromStringStatement is accessed in a separate
scope in ensureResourceStatisticsForRegistrableDomain() so it will be
reset before being used again.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::domainID const):
Add debug assert to ensure that tests catch this case.
(WebKit::ResourceLoadStatisticsDatabaseStore::ensureResourceStatisticsForRegistrableDomain const):
2020-06-08 Chris Dumez <cdumez@apple.com>
Turn some WebPageProxy logging into error logging
https://bugs.webkit.org/show_bug.cgi?id=212922
<rdar://problem/63894722>
Reviewed by Geoffrey Garen.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didFailProvisionalLoadForFrameShared):
(WebKit::WebPageProxy::didFailLoadForFrame):
(WebKit::WebPageProxy::processDidTerminate):
(WebKit::WebPageProxy::dispatchProcessDidTerminate):
(WebKit::WebPageProxy::tryReloadAfterProcessTermination):
(WebKit::WebPageProxy::didExceedInactiveMemoryLimitWhileActive):
(WebKit::WebPageProxy::didExceedBackgroundCPULimitWhileInForeground):
2020-06-08 David Kilzer <ddkilzer@apple.com>
[IPC] Adopt enum class for DragOperation
<https://webkit.org/b/212870>
<rdar://problem/64069940>
Reviewed by Darin Adler.
* Scripts/webkit/messages.py:
- Remove FIXME now that it's not needed.
* Shared/API/Cocoa/_WKDragActionsInternal.h:
(WebKit::coreDragOperationMask):
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::kit):
* UIProcess/ios/WKContentViewInteraction.mm:
(dropOperationForWebCoreDragOperation):
(coreDragOperationForUIDropOperation):
(-[WKContentView dragDataForDropSession:dragDestinationAction:]):
2020-06-08 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r262685.
https://bugs.webkit.org/show_bug.cgi?id=212914
caused flakey crashes in an iOS test
Reverted changeset:
"Clear WebPageProxy event state between tests"
https://bugs.webkit.org/show_bug.cgi?id=212864
https://trac.webkit.org/changeset/262685
2020-06-08 Sam Weinig <weinig@apple.com>
Extended Color: Unify rounding / clamping conversions between 0-1 float components and 0-255 byte components
https://bugs.webkit.org/show_bug.cgi?id=212871
Reviewed by Simon Fraser.
* UIProcess/API/ios/WKWebViewIOS.mm:
(scrollViewBackgroundColor):
Replaces uses of colorWithAlphaUsingAlternativeRounding() with colorWithAlpha().
2020-06-08 Richard Houle <rhoule@apple.com>
Links in NSAttributedString passed to -[WKWebView _showSafeBrowsingWarningWithURL:title:warning:details:completionHandler:] are ignored
https://bugs.webkit.org/show_bug.cgi?id=212801
<rdar://problem/63027631>
Reviewed by Alex Christensen.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _showSafeBrowsingWarningWithTitle:warning:details:completionHandler:]): Call the new detailsWithLinks
implementation, asserting if an URL is non nil.
(-[WKWebView _showSafeBrowsingWarningWithURL:title:warning:details:completionHandler:]): Ditto.
(-[WKWebView _showSafeBrowsingWarningWithURL:title:warning:detailsWithLinks:completionHandler:]): Pass the URL to
the completion handler when appropriate.
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
2020-06-08 Per Arne Vollan <pvollan@apple.com>
[Cocoa] Adopt read-only mode for preferences in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=212411
Reviewed by Darin Adler.
Change name of HAVE define to match SPI name.
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceMain.mm:
(WebKit::XPCServiceMain):
2020-06-08 Per Arne Vollan <pvollan@apple.com>
[Cocoa] CFPrefs synchronization does not always work for global preference changes
https://bugs.webkit.org/show_bug.cgi?id=212783
Reviewed by Tim Horton.
In CFPrefs direct mode, synchronization of global preference changes from the UI process to the WebContent process does not always work.
This is caused by the KVO notification being sent to all or some of the NSUserDefault objects, which leads us to believe a non-global
preference was changed. Since a global preference change always leads to some NSUserDefaults object receiving the notification, we can
work around this by checking if the preference being changed really belongs to the domain of the NSUserDefaults object, or if it is a
global preference. Also, stop creating NSUserDefaults object for the global domain, since this is an error according to the docs.
API test: WebKit.GlobalPreferenceChangesUsingDefaultsWrite
* UIProcess/Cocoa/PreferenceObserver.mm:
(-[WKUserDefaults _notifyObserversOfChangeFromValuesForKeys:toValuesForKeys:]):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
(WebKit::WebProcess::notifyPreferencesChanged):
2020-06-08 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] Hit-testing on icloud.com is offset after closing a tab
https://bugs.webkit.org/show_bug.cgi?id=212890
<rdar://problem/58775297>
Reviewed by Simon Fraser.
After tapping to create a new tab on the icloud.com settings page (with no other tabs open) and then closing the
new tab, all hit-testing on the page is offset by -33px until the page is reloaded. This bug appears to be a
corner case triggered by a combination of the following three changes:
- <https://trac.webkit.org/r170463>
- <https://trac.webkit.org/r245006>
- <rdar://problem/53660458> (which is an iOS-specific Safari change)
We start on icloud.com's settings page. When closing the newly created tab, Safari hides the tab bar, causing
both the top content inset and top obscured inset to decrease by 33 points. When applying the top content inset
change, Safari (after <rdar://problem/53660458>) temporarily sets `_automaticContentOffsetAdjustmentEnabled` on
`WKScrollView` to `NO`, which means that the scroll view doesn't automatically adjust its content offset to -70;
instead, it remains at -103 (the content offset when the tab bar is shown).
Because of this, during the next remote layer tree commit where the content size changes,
`-_setContentSizePreservingContentOffsetDuringRubberband:` will believe that we're currently rubber-banding by
33px, and therefore try to adjust the content offset of the scroll view to be -103 instead of allowing it to be
adjusted back to -70 (see r170463 and `-_restoreContentOffsetWithRubberbandAmount:`).
This results in the scroll position on the page (`document.scrollingElement.scrollTop`) reporting -33, even
though the top of the page is flush with the bottom of Safari's browser chrome. Finally, because `WKScrollView`
is made unscrollable due to `overflow: hidden;` (r245006), we end up permanently stuck in this state until the
next page load, rather than the next remote layer tree commit that causes any scrolling.
To fix this, add an additional restriction on the rubberbanding offset restoration code in
`-_setContentSizePreservingContentOffsetDuringRubberband:`, such that we only try to preserve the current
rubberbanded offset if we're actually rubberbanding (that is, dragging or bouncing against the edge of the
scroll view). Note that when rubberbanding against the top of the scroll view on iPad using a trackpad, the
`-isDragging` property is also true, which matches behavior when panning via touch.
Test: ScrollViewInsetTests.ChangeInsetWithoutAutomaticAdjustmentWhileWebProcessIsUnresponsive
* Platform/spi/ios/UIKitSPI.h:
* UIProcess/ios/WKScrollView.mm:
(-[WKScrollView _setContentSizePreservingContentOffsetDuringRubberband:]):
Also, split out the `CGSizeEqualToSize(currentContentSize, contentSize)` case into a separate early return that
skips the call to `-setContentSize:` altogether.
2020-06-08 Andy Estes <aestes@apple.com>
[Apple Pay] Remove -respondsToSelector: check before calling -[PKPaymentRequest setBoundInterfaceIdentifier:]
https://bugs.webkit.org/show_bug.cgi?id=212884
<rdar://problem/64090963>
Reviewed by Youenn Fablet.
* Shared/ApplePay/cocoa/WebPaymentCoordinatorProxyCocoa.mm:
(WebKit::WebPaymentCoordinatorProxy::platformPaymentRequest): Relied on the definition of
HAVE_PASSKIT_BOUND_INTERFACE_IDENTIFIER to ensure that PKPaymentRequest will respond to
-setBoundInterfaceIdentifier:.
2020-06-08 Andy Estes <aestes@apple.com>
[Apple Pay] Remove ENABLE_APPLE_PAY_SETUP, ENABLE_APPLE_PAY_SESSION_V7, and HAVE_PASSKIT_PAYMENT_SETUP
https://bugs.webkit.org/show_bug.cgi?id=212883
<rdar://problem/64090763>
Reviewed by Youenn Fablet.
These macros evaluate to true whenever ENABLE(APPLE_PAY) is true on platforms supported by
trunk WebKit, so we can either remove them or replace them with ENABLE(APPLE_PAY).
* Shared/ApplePay/ApplePayPaymentSetupFeatures.mm:
* Shared/ApplePay/ApplePayPaymentSetupFeaturesWebKit.h:
* Shared/ApplePay/PaymentSetupConfiguration.mm:
* Shared/ApplePay/PaymentSetupConfigurationWebKit.h:
* Shared/ApplePay/WebPaymentCoordinatorProxy.h:
* Shared/ApplePay/WebPaymentCoordinatorProxy.messages.in:
* Shared/ApplePay/cocoa/WebPaymentCoordinatorProxyCocoa.mm:
* WebProcess/ApplePay/WebPaymentCoordinator.cpp:
* WebProcess/ApplePay/WebPaymentCoordinator.h:
2020-06-08 youenn fablet <youenn@apple.com>
[Cocoa] Use AVAssetWriterDelegate to implement MediaRecorder
https://bugs.webkit.org/show_bug.cgi?id=206582
<rdar://problem/58985368>
Reviewed by Eric Carlson.
Enable RemoteMediaRecorder only for systems supporting AVAssetWriterDelegate.
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::didReceiveMessage):
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/webrtc/RemoteMediaRecorder.cpp:
* GPUProcess/webrtc/RemoteMediaRecorder.h:
* GPUProcess/webrtc/RemoteMediaRecorder.messages.in:
* GPUProcess/webrtc/RemoteMediaRecorderManager.cpp:
* GPUProcess/webrtc/RemoteMediaRecorderManager.h:
* GPUProcess/webrtc/RemoteMediaRecorderManager.messages.in:
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayerManager.h:
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
* WebProcess/GPU/webrtc/MediaRecorderPrivate.h:
* WebProcess/GPU/webrtc/MediaRecorderProvider.cpp:
(WebKit::MediaRecorderProvider::createMediaRecorderPrivate):
2020-06-07 Lauro Moura <lmoura@igalia.com>
[WebDriver] Allow platform-dependent modifier keys conversion to WebEvent::Modifier
https://bugs.webkit.org/show_bug.cgi?id=212848
Reviewed by Brian Burg.
WebEvent::Modifier flags does not map directly to the platform flags,
so we can't directly convert the modifier flags to the OptionSet.
Covered by existing WebDriver tests.
* UIProcess/Automation/WebAutomationSession.cpp:
(WebKit::WebAutomationSession::simulateMouseInteraction): Use
platform-specific method to convert the modifiers.
* UIProcess/Automation/WebAutomationSession.h:
* UIProcess/Automation/gtk/WebAutomationSessionGtk.cpp: Added
conversion from GDK modifier mask to WebEvent::Modifier.
(WebKit::WebAutomationSession::platformWebModifiersFromRaw):
* UIProcess/Automation/mac/WebAutomationSessionMac.mm: Added
conversion from NSModifierFlag to WebEvent::Modifier.
(WebKit::WebAutomationSession::platformWebModifiersFromRaw):
* UIProcess/Automation/wpe/WebAutomationSessionWPE.cpp: Added
conversion from wpe_input_keyboard_modifier to WebEvent::Modifier.
(WebKit::WebAutomationSession::platformWebModifiersFromRaw):
2020-06-07 Lauro Moura <lmoura@igalia.com>
[WPE] Behave like GDK_CURRENT_TIME when the timestamp is zero
https://bugs.webkit.org/show_bug.cgi?id=212797
WebDriver tests come with zero timestamp. So behave like
GDK_CURRENT_TIME and assume zero timestamp == WallTime::now().
Reviewed by Carlos Garcia Campos.
Covered by existing tests.
* Shared/libwpe/WebEventFactory.cpp:
(WebKit::wallTimeForEventTime):
2020-06-07 Philippe Normand <pnormand@igalia.com>
Remove ENABLE_VIDEO_TRACK ifdef guards
https://bugs.webkit.org/show_bug.cgi?id=212568
Reviewed by Youenn Fablet.
* Configurations/FeatureDefines.xcconfig:
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::dispatchMessage):
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
(WebKit::RemoteMediaPlayerProxy::textTrackSetMode):
(WebKit::RemoteMediaPlayerProxy::notifyTrackModeChanged):
* GPUProcess/media/RemoteMediaPlayerProxy.h:
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRenderer.cpp:
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRenderer.h:
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRenderer.messages.in:
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRendererManager.cpp:
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRendererManager.h:
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRendererManager.messages.in:
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayer.cpp:
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayer.h:
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayer.messages.in:
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayerManager.cpp:
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayerManager.h:
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayerManager.messages.in:
* Shared/WebPreferences.yaml:
* WebProcess/GPU/GPUProcessConnection.cpp:
(WebKit::GPUProcessConnection::dispatchMessage):
* WebProcess/GPU/GPUProcessConnection.h:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::tracksChanged):
* WebProcess/GPU/media/MediaPlayerPrivateRemote.h:
* WebProcess/GPU/media/RemoteMediaPlayerManager.cpp:
(WebKit::RemoteMediaPlayerManager::updatePreferences):
* WebProcess/GPU/webrtc/AudioMediaStreamTrackRenderer.cpp:
* WebProcess/GPU/webrtc/AudioMediaStreamTrackRenderer.h:
* WebProcess/GPU/webrtc/SampleBufferDisplayLayer.cpp:
* WebProcess/GPU/webrtc/SampleBufferDisplayLayer.h:
* WebProcess/GPU/webrtc/SampleBufferDisplayLayer.messages.in:
* WebProcess/GPU/webrtc/SampleBufferDisplayLayerManager.cpp:
* WebProcess/GPU/webrtc/SampleBufferDisplayLayerManager.h:
* WebProcess/InjectedBundle/InjectedBundle.cpp:
(WebKit::InjectedBundle::overrideBoolPreferenceForTestRunner):
* WebProcess/Network/WebLoaderStrategy.cpp:
(WebKit::maximumBufferingTime):
2020-06-06 Simon Fraser <simon.fraser@apple.com>
Clear WebPageProxy event state between tests
https://bugs.webkit.org/show_bug.cgi?id=212864
Reviewed by Anders Carlsson.
Make sure that event state doesn't leak between tests via events queued in WebPageProxy.
* UIProcess/API/C/WKPage.cpp:
(WKPageClearEventState):
* UIProcess/API/C/WKPage.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::resetStateAfterProcessExited):
(WebKit::WebPageProxy::clearEventState):
* UIProcess/WebPageProxy.h:
2020-06-06 Rob Buis <rbuis@igalia.com>
Move URL scheme check to NetworkLoadChecker
https://bugs.webkit.org/show_bug.cgi?id=191924
Reviewed by Joseph Pecoraro.
Reduce includes for CustomHeaderFields, I guess these were needed
at some point but not anymore.
* Shared/WebsitePoliciesData.cpp:
2020-06-06 David Kilzer <ddkilzer@apple.com>
Use OptionSet<DragOperation> for mask values
<https://webkit.org/b/212605>
Reviewed by Darin Adler.
In broad strokes:
- Replace use of DragOperation with OptionSet<DragOperation> or
Optional<DragOperation>.
- Rename function parameters and local variables to denote use
of mask values.
- Remove DragOperationNone enum value.
- Replace DragOperationEvery enum value with anyDragOperation().
* Platform/spi/ios/UIKitSPI.h:
(UIDragOperation):
- Remove unused (and misnamed) UIDragOperation enum typedef.
* Scripts/webkit/messages.py:
(types_that_cannot_be_forward_declared):
(headers_for_type):
- Add support for WebCore::DragOperation enum.
* Shared/API/Cocoa/_WKDragActionsInternal.h:
(WebKit::coreDragOperationMask): Add conversion function.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<DragData>::encode):
(IPC::ArgumentCoder<DragData>::decode):
- Update to encode/decode OptionSet<DragOperation>.
* UIProcess/API/gtk/DragSource.h:
(WebKit::DragSource::begin):
* UIProcess/API/gtk/DragSourceGtk3.cpp:
(WebKit::DragSource::begin):
* UIProcess/API/gtk/DragSourceGtk4.cpp:
(WebKit::DragSource::begin):
* UIProcess/API/gtk/DropTarget.h:
* UIProcess/API/gtk/DropTargetGtk3.cpp:
(WebKit::DropTarget::didPerformAction):
(WebKit::DropTarget::leaveTimerFired):
* UIProcess/API/gtk/DropTargetGtk4.cpp:
(WebKit::DropTarget::DropTarget):
(WebKit::DropTarget::didPerformAction):
* UIProcess/API/gtk/PageClientImpl.cpp:
(WebKit::PageClientImpl::startDrag):
* UIProcess/API/gtk/PageClientImpl.h:
(WebKit::PageClientImpl::startDrag):
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseStartDrag):
* UIProcess/API/gtk/WebKitWebViewBasePrivate.h:
(webkitWebViewBaseStartDrag):
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::sendDragEndToPage):
(WebKit::WebViewImpl::draggingEntered):
(WebKit::kit): Add conversion function.
(WebKit::WebViewImpl::draggingUpdated):
(WebKit::WebViewImpl::draggingExited):
(WebKit::WebViewImpl::performDragOperation):
- Make use of coreDragOperationMask() conversion function.
* UIProcess/PageClient.h:
(WebKit::WebViewImpl::startDrag):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didPerformDragControllerAction):
- Remove MESSAGE_CHECK now that we're using the actual type.
(WebKit::WebPageProxy::startDrag):
(WebKit::WebPageProxy::dragEnded):
(WebKit::WebPageProxy::resetCurrentDragInformation):
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::didPerformDragControllerAction):
(WebKit::WebPageProxy::dragEnded):
(WebKit::WebPageProxy::startDrag):
(WebKit::WebPageProxy::currentDragOperation const):
* UIProcess/WebPageProxy.messages.in:
(WebPageProxy::DidPerformDragControllerAction):
(WebPageProxy::StartDrag):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _didHandleDragStartRequest:]):
(dropOperationForWebCoreDragOperation):
(coreDragOperationForUIDropOperation): Add conversion function.
- Note that this was not being converted from UIDropOperation to
WebCore::DragOperation previously, and UIDropOperationMove
(0x3) was not being properly converted to DragOperationMove
(0x16).
(-[WKContentView dragDataForDropSession:dragDestinationAction:]):
(-[WKContentView _didPerformDragOperation:]):
(-[WKContentView dragInteraction:willAnimateLiftWithAnimator:session:]):
(-[WKContentView dragInteraction:session:didEndWithOperation:]):
(-[WKContentView dropInteraction:sessionDidEnd:]):
* WebProcess/WebCoreSupport/gtk/WebDragClientGtk.cpp:
(WebKit::WebDragClient::startDrag):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::performDragControllerAction):
(WebKit::WebPage::dragEnded):
* WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::performDragControllerAction):
(WebKit::WebPage::dragEnded):
* WebProcess/WebPage/WebPage.messages.in:
(WebPage::PerformDragControllerAction):
(WebPage::DragEnded):
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::requestAdditionalItemsForDragSession):
2020-06-06 Devin Rousso <drousso@apple.com>
Web Inspector: unify the naming scheme for agents used by instrumentation
https://bugs.webkit.org/show_bug.cgi?id=212859
Reviewed by Timothy Hatcher.
Inspector agents fall into one of three categories:
- "persistent" when Web Inspector is connected
- "enabled" when that agent is `enable`d, such as if the corresponding tab is visible
- "tracking" when that agent is part of a timeline recording.
The only exception to this is the Console agent, as that exists regardless of whether Web
Inspector is connected as it needs to preserve messages logged before Web Inspector connects.
Also remove the "Inspector" prefix from getter/setter methods as it adds confusion if that
agent also has subclasses (e.g. `InspectorRuntimeAgent` and `PageRuntimeAgent`).
* UIProcess/Inspector/Agents/InspectorBrowserAgent.cpp:
* UIProcess/Inspector/WebInspectorProxy.cpp:
* UIProcess/Inspector/WebPageInspectorController.h:
Simple naming changes elided to avoid a long ChangeLog.
2020-06-05 David Kilzer <ddkilzer@apple.com>
[IPC] Adopt enum class for AutocapitalizeType
<https://webkit.org/b/212846>
<rdar://problem/64042825>
Reviewed by Darin Adler.
Summary:
- Move AutocapitalizeType into WebCore namespace.
- Convert AutocapitalizeType to an enum class.
- Add WTF::EnumTraits<AutocapitalizeType> for IPC.
* Shared/FocusedElementInformation.cpp:
(WebKit::FocusedElementInformation::encode const):
(WebKit::FocusedElementInformation::decode):
- Stop using decodeEnum() and encodeEnum() now that
WTF::EnumTraits<AutocapitalizeType> exists.
* Shared/FocusedElementInformation.h:
* Shared/glib/InputMethodState.cpp:
(WebKit::InputMethodState::addHintsForAutocapitalizeType):
* Shared/glib/InputMethodState.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(toUITextAutocapitalize):
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::getFocusedElementInformation):
2020-06-05 Brent Fulgham <bfulgham@apple.com>
[iOS] Silence telemetry generated during download operations
https://bugs.webkit.org/show_bug.cgi?id=212854
<rdar://problem/63909414>
Reviewed by Per Arne Vollan.
Remove reporting and telemetry for the 'com.apple.ProgressReporting' service
from the Network Process sandbox, since this is a critical part of supplying
download status to the WebKit UI.
* Resources/SandboxProfiles/ios/com.apple.WebKit.Networking.sb:
2020-06-05 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed, reverting r262619, r262625, and r262641.
Caused mediarecorder layout test crashes.
Reverted changesets:
"[Cocoa] Use AVAssetWriterDelegate to implement MediaRecorder"
https://bugs.webkit.org/show_bug.cgi?id=206582
https://trac.webkit.org/changeset/262619
"[Cocoa] Use AVAssetWriterDelegate to implement MediaRecorder"
https://bugs.webkit.org/show_bug.cgi?id=206582
https://trac.webkit.org/changeset/262625
"Unreviewed, silence deprecation warning to fix build with
latest SDK."
https://trac.webkit.org/changeset/262641
2020-06-05 Kate Cheney <katherine_cheney@apple.com>
ITP SQLite Database should only vacuum once per day
https://bugs.webkit.org/show_bug.cgi?id=212712
<rdar://problem/63939711>
Reviewed by Brent Fulgham.
Followup to https://trac.webkit.org/changeset/262538/webkit. We should
run incremental auto vacuum, not a full vacuum, once per-day, because
it is less expensive.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.h:
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::performDailyTasks):
2020-06-05 Daniel Bates <dabates@apple.com>
-_focusTextInputContext should reposition the caret if called for an already focused text input context
https://bugs.webkit.org/show_bug.cgi?id=212830
<rdar://problem/64032956>
Reviewed by Wenson Hsieh.
Update the code based on new feedback: the caret should now be re-positioned when -_focusTextInputContext
is called on an already assisted text input context.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView cleanUpInteraction]):
(-[WKContentView _didCommitLoadForMainFrame]): Reset state. See below for why.
(-[WKContentView _focusTextInputContext:placeCaretAt:completionHandler:]): I cannot simply remove the early
return in -_focusTextInputContext that detects if the specified text input context is already focused because
it could effect user-visible responsiveness due to an Apple internal client invoking this function more times
than necessary with the same point during a text interaction. See <rdar://problem/59430806>. That radar
represents a correctness issue. I didn't actually evaluate whether responsiveness is affected by this patch,
but I would feel more confident that there would be 0 impact when removing the early branch after that radar
is fixed without the need to even perform a responsiveness analysis. Note that there is already a comment about
that radar at the top of this function.
2020-06-05 Alex Christensen <achristensen@webkit.org>
Fix PDF opening after r262592
https://bugs.webkit.org/show_bug.cgi?id=212795
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::savePDFToTemporaryFolderAndOpenWithNativeApplication):
(WebKit::WebPageProxy::openPDFFromTemporaryFolderWithNativeApplication):
As Darin mentioned and I was about to discover, my last-minute switching things around got the boolean condition backwards.
We want to do nothing if we're not allowed to open the PDF, and we want to open the PDF if we are allowed to.s
2020-06-05 David Kilzer <ddkilzer@apple.com>
[IPC] Adopt enum class for PluginLoadClientPolicy
<https://webkit.org/b/212827>
<rdar://problem/64030431>
Reviewed by Alex Christensen.
Summary:
- Update for renamed PluginLoadClientPolicy enums.
- Update to make PluginLoadClientPolicy type work with IPC.
- Replace uint8_t with strongly-typed PluginLoadClientPolicy.
* Scripts/webkit/messages.py:
* Shared/Plugins/Netscape/mac/NetscapePluginModuleMac.mm:
(WebKit::NetscapePluginModule::getPluginInfo):
* Shared/WebProcessCreationParameters.h:
* UIProcess/API/C/WKAPICast.h:
(WebKit::toWKPluginLoadClientPolicy):
(WebKit::toPluginLoadClientPolicy):
* UIProcess/API/Cocoa/WKProcessPool.mm:
(toPluginLoadClientPoliciesHashMap):
(policiesHashMapToDictionary):
(isPluginLoadClientPolicyAcceptable): Delete.
- Replace with
WTF::isValidEnum<WebCore::PluginLoadClientPolicy>().
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::setPluginLoadClientPolicy):
(WebKit::WebProcessPool::resetPluginLoadClientPolicies):
* UIProcess/WebProcessPool.h:
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::pluginInfo):
* WebProcess/Plugins/WebPluginInfoProvider.cpp:
(WebKit::WebPluginInfoProvider::webVisiblePluginInfo):
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::setPluginLoadClientPolicy):
(WebKit::WebProcess::resetPluginLoadClientPolicies):
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
2020-06-05 Alex Christensen <achristensen@webkit.org>
Update header documentation for WKPreferences.fraudulentWebsiteWarningEnabled
https://bugs.webkit.org/show_bug.cgi?id=212843
<rdar://problem/64010164>
Reviewed by Geoffrey Garen.
* UIProcess/API/Cocoa/WKPreferences.h:
2020-06-05 Brent Fulgham <bfulgham@apple.com>
[iOS] Remove message filtering telemetry for expected case
https://bugs.webkit.org/show_bug.cgi?id=212841
<rdar://problem/63893460>
Reviewed by Per Arne Vollan.
Remove unintentional telemetry being captured for expected behavior.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-06-05 Per Arne Vollan <pvollan@apple.com>
[Cocoa] Adopt read-only mode for preferences in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=212411
<rdar://problem/63640788>
Reviewed by Darin Adler.
The WebContent process should never write preferences. Adopt read-only mode for preferences.
This will fix some sandbox violations when the UI process is notifying the WebContent process
about preference changes. When notified, the WebContent process will use the preferences API
to write those preferences in-process, but there should be no attempt at writing these new
values to disk, which this patch will ensure.
No new tests, since this is fixing a benign sandbox violation.
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceMain.mm:
(WebKit::XPCServiceMain):
2020-06-05 Per Arne Vollan <pvollan@apple.com>
[iOS] Fix message filter sandbox violation
https://bugs.webkit.org/show_bug.cgi?id=212837
<rdar://problem/63830368>
Reviewed by Darin Adler.
Add required message to message filter in the iOS WebContent sandbox.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-06-05 Peng Liu <peng.liu6@apple.com>
REGRESSION (r262456?): [macOS] media/picture-in-picture/picture-in-picture-api tests are flaky timeouts
https://bugs.webkit.org/show_bug.cgi?id=212762
Reviewed by Eric Carlson.
The Picture-in-Picture implementation expects the two callbacks HTMLVideoElement::didBecomeFullscreenElement()
and HTMLVideoElement::setVideoFullscreenFrame() to come in order. So we need to remove the dispatch_async()
and call HTMLVideoElement::didBecomeFullscreenElement() directly.
* WebProcess/cocoa/VideoFullscreenManager.mm:
(WebKit::VideoFullscreenManager::didEnterFullscreen):
2020-06-05 Adrian Perez de Castro <aperez@igalia.com>
Non-unified build fixes, early summer 2020 edition
https://bugs.webkit.org/show_bug.cgi?id=212819
Unreviewed build fix.
* Shared/gtk/NativeWebMouseEventGtk.cpp:
(WebKit::NativeWebMouseEvent::NativeWebMouseEvent): Add missing WebCore:: namespace to
a couple of uses of WebCore::FloatSize.
* WebProcess/Storage/WebServiceWorkerFetchTaskClient.cpp: Add missing Logging.h header.
2020-06-05 Andy Estes <aestes@apple.com>
[Apple Pay] Remove conditionals for ENABLE_APPLE_PAY_SESSION_V(3|4)
https://bugs.webkit.org/show_bug.cgi?id=212541
<rdar://problem/63781452>
Reviewed by Darin Adler.
APPLE_PAY_SESSION_V(3|4) is now enabled whenever APPLE_PAY itself is enabled.
* Configurations/FeatureDefines.xcconfig:
2020-06-05 youenn fablet <youenn@apple.com>
[Cocoa] Use AVAssetWriterDelegate to implement MediaRecorder
https://bugs.webkit.org/show_bug.cgi?id=206582
<rdar://problem/58985368>
Reviewed by Eric Carlson.
Enable RemoteMediaRecorder only for systems supporting AVAssetWriterDelegate.
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::didReceiveMessage):
* GPUProcess/GPUConnectionToWebProcess.h:
* GPUProcess/webrtc/RemoteMediaRecorder.cpp:
* GPUProcess/webrtc/RemoteMediaRecorder.h:
* GPUProcess/webrtc/RemoteMediaRecorder.messages.in:
* GPUProcess/webrtc/RemoteMediaRecorderManager.cpp:
* GPUProcess/webrtc/RemoteMediaRecorderManager.h:
* GPUProcess/webrtc/RemoteMediaRecorderManager.messages.in:
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayerManager.h:
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
* WebProcess/GPU/webrtc/MediaRecorderPrivate.h:
* WebProcess/GPU/webrtc/MediaRecorderProvider.cpp:
(WebKit::MediaRecorderProvider::createMediaRecorderPrivate):
2020-06-05 Carlos Garcia Campos <cgarcia@igalia.com>
REGRESSION(r262184): [GTK] Two fast/forms tests are crashing
https://bugs.webkit.org/show_bug.cgi?id=212747
Reviewed by Adrian Perez de Castro.
* UIProcess/gtk/WebPopupMenuProxyGtk.cpp:
(WebKit::WebPopupMenuProxyGtk::handleKeyPress): Return early if the menu has already been closed.
2020-06-05 Youenn Fablet <youenn@apple.com>
Generate code to set the runtime settings for a service worker process
https://bugs.webkit.org/show_bug.cgi?id=212467
Reviewed by Alex Christensen.
Split the code generated to handle web preference changes to have one method for process wide flags and one for page settings.
Make the process wide flag method static and reuse it in WebSWContextManagerConnection.
Covered by existing tests.
* Scripts/PreferencesTemplates/WebPageUpdatePreferences.cpp.erb:
* WebProcess/Storage/WebSWContextManagerConnection.cpp:
(WebKit::WebSWContextManagerConnection::updatePreferencesStore):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences):
* WebProcess/WebPage/WebPage.h:
2020-06-04 David Kilzer <ddkilzer@apple.com>
[IPC] Adopt enum class for PluginProcessType and PluginProcessSandboxPolicy
<https://webkit.org/b/212758>
Reviewed by Darin Adler.
* PluginProcess/mac/PluginProcessMac.mm:
(WebKit::PluginProcess::platformInitializePluginProcess):
* Scripts/webkit/messages.py:
* Shared/Plugins/PluginProcessAttributes.h:
(WebKit::PluginProcessType): Convert to enum class.
(WebKit::PluginProcessSandboxPolicy): Ditto.
* UIProcess/Plugins/PluginProcessManager.cpp:
(WebKit::PluginProcessManager::fetchWebsiteData):
(WebKit::PluginProcessManager::deleteWebsiteData):
(WebKit::PluginProcessManager::deleteWebsiteDataForHostNames):
* UIProcess/Plugins/PluginProcessProxy.cpp:
(WebKit::PluginProcessProxy::didFinishLaunching):
* UIProcess/Plugins/mac/PluginProcessProxyMac.mm:
(WebKit::PluginProcessProxy::platformGetLaunchOptionsWithAttributes):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::findPlugin):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::createPlugin):
(WebKit::WebPage::canPluginHandleResponse):
2020-06-04 Alex Christensen <achristensen@webkit.org>
Add WKUIDelegate SPI to confirm before opening a PDF
https://bugs.webkit.org/show_bug.cgi?id=212795
<rdar://problem/58715847>
Reviewed by Tim Horton.
* UIProcess/API/APIUIClient.h:
(API::UIClient::confirmPDFOpening):
* UIProcess/API/Cocoa/WKUIDelegatePrivate.h:
* UIProcess/Cocoa/UIDelegate.h:
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::setDelegate):
(WebKit::UIDelegate::UIClient::confirmPDFOpening):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::savePDFToTemporaryFolderAndOpenWithNativeApplication):
(WebKit::WebPageProxy::openPDFFromTemporaryFolderWithNativeApplication):
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::savePDFToTemporaryFolderAndOpenWithNativeApplication):
(WebKit::WebPageProxy::openPDFFromTemporaryFolderWithNativeApplication):
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::openWithNativeApplication):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::savePDFToTemporaryFolderAndOpenWithNativeApplication):
* WebProcess/WebPage/WebPage.h:
2020-06-04 Chris Dumez <cdumez@apple.com>
UIProcess may crash after its prewarmed WebProcess gets terminated
https://bugs.webkit.org/show_bug.cgi?id=212790
<rdar://problem/63696191>
Reviewed by Geoffrey Garen.
Query WebProcessProxy::isDummyProcessProxy() because trying to look up the process
in m_dummyProcessProxies. The key for m_dummyProcessProxies is a sessionID but
not all processes have a sessionID. In particular, prewarmed web processes have
no associated WebsiteDataStore and thus not sessionID.
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::disconnectProcess):
2020-06-04 Brent Fulgham <bfulgham@apple.com>
Reduce telemetry and report generation for WebContent Process sandbox
https://bugs.webkit.org/show_bug.cgi?id=212779
<rdar://problem/63769879>
Reviewed by Per Arne Vollan.
This patch cleans up a set of telemetry and reports generated during normal
WebKit operation. Now that we understand the uses of these XPC services, we
can reduce the logging traffic.
This is important to do now, since generating the telemetry and reports has
a non-zero power cost.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* WebProcess/com.apple.WebProcess.sb.in:
2020-06-04 Jiewen Tan <jiewen_tan@apple.com>
Disable CFNetwork AppSSO interception for Mac Catalyst
https://bugs.webkit.org/show_bug.cgi?id=212787
<rdar://problem/63738783>
Reviewed by Chris Dumez.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::NetworkSessionCocoa):
2020-06-04 Brent Fulgham <bfulgham@apple.com>
Reduce telemetry and report generation for Network Process sandbox
https://bugs.webkit.org/show_bug.cgi?id=212778
<rdar://problem/63891745>
Reviewed by Per Arne Vollan.
This patch cleans up a set of telemetry and reports generated during normal
WebKit operation. Now that we understand the uses of these XPC services, we
can reduce the logging traffic.
This is important to do now, since telemetry and report generation has
battery life and performance impact.
* NetworkProcess/mac/com.apple.WebKit.NetworkProcess.sb.in:
* Resources/SandboxProfiles/ios/com.apple.WebKit.Networking.sb:
2020-06-04 Jacob Uphoff <jacob_uphoff@apple.com>
Unreviewed, reverting r262546.
This commit caused internal build failures
Reverted changeset:
"[Cocoa] Adopt read-only mode for preferences in the
WebContent process"
https://bugs.webkit.org/show_bug.cgi?id=212411
https://trac.webkit.org/changeset/262546
2020-06-04 Chris Dumez <cdumez@apple.com>
[iOS] Drop DependentProcessLink ProcessAssertion as it creates power leaks
https://bugs.webkit.org/show_bug.cgi?id=212768
<rdar://problem/62408828>
Reviewed by Alex Christensen.
Drop DependentProcessLink ProcessAssertion since it creates a process assertion cycle between
the UIProcess and the WebContent process and leads to power leaks.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
* UIProcess/ProcessAssertion.h:
* UIProcess/ProcessThrottler.cpp:
(WebKit::ProcessThrottler::assertionName const):
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::didSetAssertionType):
* UIProcess/ios/ProcessAssertionIOS.mm:
(WebKit::runningBoardNameForAssertionType):
(WebKit::flagsForAssertionType):
(WebKit::toBKSProcessAssertionReason):
* WebProcess/WebProcess.h:
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-06-04 Brent Fulgham <bfulgham@apple.com>
Revise sandbox to support network stack telemetry without sandbox violation reports
https://bugs.webkit.org/show_bug.cgi?id=212720
<rdar://problem/63895783>
Reviewed by Per Arne Vollan.
Telemetry captured for various networking configurations has revealed a handful of mach service
connections that should be allowed in the Network Process to support troubleshooting and field diagnostics.
This patch revises the sandbox to avoid logging these accesses during normal operation.
* NetworkProcess/mac/com.apple.WebKit.NetworkProcess.sb.in: Do not generate traces for 'com.apple.diagnosticd'.
* Resources/SandboxProfiles/ios/com.apple.WebKit.Networking.sb: Ditto.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb: Allow 'com.apple.osanalytics.osanalyticshelper' to
be used on internal development builds.
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const): Update to handle both 'diagnosticd' and 'osanalyticshelper'
as optional services for internal builds.
(WebKit::WebProcessCreationParameters::decode): Ditto.
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::diagnosticServices): Ditto.
(WebKit::WebProcessPool::platformInitializeWebProcess): Ditto.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess): Ditto.
2020-06-04 Per Arne Vollan <pvollan@apple.com>
[Cocoa] Adopt read-only mode for preferences in the WebContent process
https://bugs.webkit.org/show_bug.cgi?id=212411
<rdar://problem/63640788>
Reviewed by Darin Adler.
The WebContent process should never write preferences. Adopt read-only mode for preferences.
This will fix some sandbox violations when the UI process is notifying the WebContent process
about preference changes. When notified, the WebContent process will use the preferences API
to write those preferences in-process, but there should be no attempt at writing these new
values to disk, which this patch will ensure.
No new tests, since this is fixing a benign sandbox violation.
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceMain.mm:
(WebKit::XPCServiceMain):
2020-06-04 Per Arne Vollan <pvollan@apple.com>
[iOS] The WebContent process should be allowed to consume an extension to the runningboard service
https://bugs.webkit.org/show_bug.cgi?id=212753
<rdar://problem/63948859>
Reviewed by Chris Dumez.
The WebContent process on iOS is failing to consume an extension to the runningboard service, because it is not allowed to do so.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-06-04 Tim Horton <timothy_horton@apple.com>
Work around broken system version macro
https://bugs.webkit.org/show_bug.cgi?id=212726
Reviewed by Dan Bernstein.
* Configurations/DebugRelease.xcconfig:
2020-06-04 Andy Estes <aestes@apple.com>
[watchOS] Re-enable content filtering in the simulator build
https://bugs.webkit.org/show_bug.cgi?id=212711
<rdar://problem/63938350>
Reviewed by Wenson Hsieh.
* Configurations/FeatureDefines.xcconfig:
2020-06-04 Kate Cheney <katherine_cheney@apple.com>
ITP SQLite Database should only vacuum once per day
https://bugs.webkit.org/show_bug.cgi?id=212712
<rdar://problem/63939711>
Reviewed by Brent Fulgham.
Currently the SQLite Database runs a vacuum command each time statistics
are processed. This is probably too frequent, because the database
rarely needs to delete large amounts of data. We can cut this down to once a day.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.h:
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::performDailyTasks):
2020-06-04 Youenn Fablet <youenn@apple.com>
Minor optimisation to RemoteAudioBusData::decode
https://bugs.webkit.org/show_bug.cgi?id=212692
Reviewed by Eric Carlson.
* WebProcess/GPU/media/RemoteAudioBusData.h:
(WebKit::RemoteAudioBusData::decode):
Use releaseNonNull to remove some count churn.
2020-06-03 Wenson Hsieh <wenson_hsieh@apple.com>
[watchOS] fast/events/watchos/change-focus-during-change-event.html crashes
https://bugs.webkit.org/show_bug.cgi?id=212722
<rdar://problem/60633339>
Reviewed by Andy Estes.
When HAVE(QUICKBOARD_COLLECTION_VIEWS) is enabled, the codepath that's used to simulate interacting with the
select menu view controller on watchOS will crash, since it tries to pretend that the user has selected an item
in a table view section (`PUICQuickboardListSectionTextOptions`) that only exists when Quickboard collection
views are disabled. Fix the crash by adjusting the testing codepath to select the item in the first section when
HAVE(QUICKBOARD_COLLECTION_VIEWS) is defined.
* UIProcess/ios/forms/WKSelectMenuListViewController.mm:
(-[WKSelectMenuListViewController selectItemAtIndex:]):
2020-06-03 Jer Noble <jer.noble@apple.com>
[iOS] Hide the PiP button in fullscreen mode if PiP is disabled in preferences
https://bugs.webkit.org/show_bug.cgi?id=212699
<rdar://problem/60391437>
Reviewed by Eric Carlson.
In r260474, we hid the PiP button in fullscreen when PiP was unsupported on the
current device; we must also hide the PiP button when clients have disabled PiP
via preferences.
* UIProcess/ios/fullscreen/WKFullScreenViewController.mm:
(-[WKFullScreenViewController videoControlsManagerDidChange]):
2020-06-03 Kate Cheney <katherine_cheney@apple.com>
Any active sqlite transactions for the ITP database should be aborted when the network process suspends.
https://bugs.webkit.org/show_bug.cgi?id=212608
<rdar://problem/60540768>
Reviewed by Chris Dumez.
Calls to WebResourceLoadStatisticsStore::suspend() should abort any
active SQLite transactions. Unfinished transactions will hold the lock
of a database file and could cause the network process to crash when
suspending.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::ResourceLoadStatisticsDatabaseStore):
(WebKit::ResourceLoadStatisticsDatabaseStore::~ResourceLoadStatisticsDatabaseStore):
Store all ResourceLoadStatisticsDatabaseStore instances in a static
HashSet so existing transactions can be aborted when the process is
preparing to suspend.
(WebKit::ResourceLoadStatisticsDatabaseStore::interrupt):
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.h:
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::suspend):
2020-06-03 Chris Dumez <cdumez@apple.com>
WebProcessPool::notifyPreferencesChanged() is unsafely called on a background thread
https://bugs.webkit.org/show_bug.cgi?id=212697
<rdar://problem/63876652>
Reviewed by Per Arne Vollan.
Make sure we dispatch to the main thread in [WKPreferenceObserver preferenceDidChange] before
we iterate over the process pools and call WebProcessPool::notifyPreferencesChanged() on each
one.
* UIProcess/Cocoa/PreferenceObserver.mm:
(-[WKPreferenceObserver preferenceDidChange:key:encodedValue:]):
2020-06-03 Brent Fulgham <bfulgham@apple.com>
[iOS] Reduce sandbox logging by permitting necessary sysctl
https://bugs.webkit.org/show_bug.cgi?id=212710
<rdar://problem/63900239>
Reviewed by Per Arne Vollan.
Telemetry indicates that we need to access "sysctl.name2oid" in the
Network process during error handling for HTTP connections. Allow this
access and avoid generating unnecessary logging for this case.
* Resources/SandboxProfiles/ios/com.apple.WebKit.Networking.sb:
2020-06-03 Youenn Fablet <youenn@apple.com>
Add more logging related to service worker fetch event handling
https://bugs.webkit.org/show_bug.cgi?id=212632
<rdar://problem/63913271>
Unreviewed.
Build fix.
* WebProcess/Storage/WebServiceWorkerFetchTaskClient.cpp:
(WebKit::WebServiceWorkerFetchTaskClient::continueDidReceiveResponse):
2020-06-03 Alex Christensen <achristensen@webkit.org>
Pass WebPageProxyIdentifier to PreconnectTask
https://bugs.webkit.org/show_bug.cgi?id=212648
Reviewed by Chris Dumez.
Otherwise, AuthenticationManager::didReceiveAuthenticationChallenge sometimes returned early because there was no pageID.
Observed connecting to real websites from my home wifi. Unable to test using local servers because connecting happens too fast.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::preconnectTo):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::preconnectTo):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::preconnectTo):
2020-06-03 Rob Buis <rbuis@igalia.com>
Make generated C++ code use modern C++
https://bugs.webkit.org/show_bug.cgi?id=190714
Reviewed by Jonathan Bedard.
Replace typedef usage by alias-declaration.
* Scripts/test-legacyMessages.h:
* Scripts/test-superclassMessages.h:
* Scripts/testMessages.h:
* Scripts/webkit/messages.py:
2020-06-03 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] WTR: Forward key events to the active popup menu
https://bugs.webkit.org/show_bug.cgi?id=212690
Reviewed by Adrian Perez de Castro.
I forgot to remove the ifdefs when added support for popup menus in r262414.
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseSynthesizeKeyEvent):
2020-06-03 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Fix input methods
https://bugs.webkit.org/show_bug.cgi?id=212317
Reviewed by Adrian Perez de Castro.
Stop using GdkEvent to synthesize compositing events, add a specific NativeWebKeyboardEvent constructor for them.
* Shared/NativeWebKeyboardEvent.h:
* Shared/gtk/NativeWebKeyboardEventGtk.cpp:
(WebKit::NativeWebKeyboardEvent::NativeWebKeyboardEvent):
* UIProcess/API/gtk/WebKitInputMethodContextImplGtk.cpp:
(webkitInputMethodContextImplGtkSetClientWidget):
(webkitInputMethodContextImplGtkSetClientWindow): Deleted.
* UIProcess/API/gtk/WebKitInputMethodContextImplGtk.h:
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseRealize):
(webkitWebViewBaseUnrealize):
(webkitWebViewBaseKeyPressEvent):
(webkitWebViewBaseKeyReleaseEvent):
(webkitWebViewBaseFocusEnter):
(webkitWebViewBaseFocusLeave):
(webkitWebViewBaseKeyPressed):
(webkitWebViewBaseKeyReleased):
(toplevelWindowIsActiveChanged):
(webkitWebViewBaseSynthesizeCompositionKeyPress):
2020-06-03 Santosh Mahto <santosh.mahto@collabora.com>
[GTK4] Implement file chooser
https://bugs.webkit.org/show_bug.cgi?id=212322
Reviewed by Carlos Garcia Campos.
Port FileChooser to gtk4, Now FileChooser uses undeprecated api's.
Covered by existing tests.
* UIProcess/API/gtk/WebKitWebViewGtk.cpp:
(fileChooserDialogResponseCallback):
(webkitWebViewRunFileChooser):
2020-06-03 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Implement emoji chooser
https://bugs.webkit.org/show_bug.cgi?id=212330
Reviewed by Adrian Perez de Castro.
GtkEmojiChooser is public API in GTK4, so just use that instead of our own.
* UIProcess/API/gtk/WebKitEmojiChooser.cpp:
(emojiActivated): Close the popover before emitting emoji-picked for consistency with GTK.
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(_WebKitWebViewBasePrivate::_WebKitWebViewBasePrivate):
(_WebKitWebViewBasePrivate::releaseEmojiChooserTimerFired):
(webkitWebViewBaseDispose):
(emojiChooserClosed):
(webkitWebViewBaseShowEmojiChooser):
2020-06-03 Youenn Fablet <youenn@apple.com>
Add more logging related to service worker fetch event handling
https://bugs.webkit.org/show_bug.cgi?id=212632
Reviewed by Chris Dumez.
Add logging to identify the page ID and frame ID used by a service worker.
Add logging to identify what is happening in the case of a response being buffered for navigation purposes.
* WebProcess/Storage/WebSWContextManagerConnection.cpp:
(WebKit::ServiceWorkerFrameLoaderClient::ServiceWorkerFrameLoaderClient):
(WebKit::WebSWContextManagerConnection::installServiceWorker):
(WebKit::WebSWContextManagerConnection::continueDidReceiveFetchResponse):
* WebProcess/Storage/WebServiceWorkerFetchTaskClient.cpp:
(WebKit::WebServiceWorkerFetchTaskClient::didFail):
(WebKit::WebServiceWorkerFetchTaskClient::didFinish):
(WebKit::WebServiceWorkerFetchTaskClient::continueDidReceiveResponse):
2020-06-02 Chris Dumez <cdumez@apple.com>
[iOS] WKProcessAssertionBackgroundTaskManager incorrectly ignores expiration notifications for daemons
https://bugs.webkit.org/show_bug.cgi?id=212619
Reviewed by Alex Christensen.
WKProcessAssertionBackgroundTaskManager was incorrectly ignoring process assertion expiration notifications
for daemons, because it was relying on visibility to make decisions. For daemons (or ViewServices), we would
not get application visibility notifications and would therefore always assume the app is visible.
As a result, _handleBackgroundTaskExpiration would think it received an outdated expiration notification for
a visible app and would simply re-take the assertion right away.
To address the issue, we now rely on a RunningBoard API that lets us know if the suspension timer has been
started or not (and how much remains on the timer). If the suspension timer is not active when get receive
the expiration notification, we know it is an outdated notification and we ignore it (release and re-take
assertion right away). This can happen if we did not have time to process the expiration notification before
suspended and thus only get it upon resuming. It can also happen if the user re-activates the app right after
the expiration notice has been sent.
* Platform/spi/ios/RunningBoardServicesSPI.h:
* UIProcess/ios/ProcessAssertionIOS.mm:
(processHasActiveRunTimeLimitation):
(-[WKProcessAssertionBackgroundTaskManager init]):
(-[WKProcessAssertionBackgroundTaskManager _updateBackgroundTask]):
(-[WKProcessAssertionBackgroundTaskManager _handleBackgroundTaskExpiration]):
(-[WKProcessAssertionBackgroundTaskManager _handleBackgroundTaskExpirationOnMainThread]):
(-[WKProcessAssertionBackgroundTaskManager _releaseBackgroundTask]):
2020-06-02 Christopher Reid <chris.reid@sony.com>
REGRESSION[r260844]: [GTK][WPE] Inspector GResource no longer updated after WebInspectorUI file updates
https://bugs.webkit.org/show_bug.cgi?id=212583
Reviewed by Adrian Perez de Castro.
Add a cmake dependency on inspector resource stamp so the Glib
Inspector resources rebuild when the stamp is updated.
* InspectorGResources.cmake:
2020-06-02 Keith Rollin <krollin@apple.com>
Revert FEATURES_DEFINES related changes
https://bugs.webkit.org/show_bug.cgi?id=212664
<rdar://problem/63893033>
Reviewed by Andy Estes.
Bug 262310, Bug 262311, Bug 262318, and Bug 262331 involve changes to
FEATURE_DEFINES and how the values there relate to those found in the
Platform*.h files. Those changes break XCBuild (by removing the
.xcfilelist related to UnifiedSources and the process for generating
them), and so are being reverted.
* Configurations/FeatureDefines.xcconfig:
2020-06-02 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed, reverting r262424.
Caused webkitpy test failure
Reverted changeset:
"Make generated C++ code use modern C++"
https://bugs.webkit.org/show_bug.cgi?id=190714
https://trac.webkit.org/changeset/262424
2020-06-02 Peng Liu <peng.liu6@apple.com>
Stressing webkitSetPresentationMode leads to wrong inline video dimensions
https://bugs.webkit.org/show_bug.cgi?id=202425
Reviewed by Eric Carlson.
* WebProcess/cocoa/VideoFullscreenManager.mm:
(WebKit::VideoFullscreenManager::didEnterFullscreen):
(WebKit::VideoFullscreenManager::didCleanupFullscreen):
2020-06-02 Kate Cheney <katherine_cheney@apple.com>
Assert network process is not suspended when trying to merge new ITP data
https://bugs.webkit.org/show_bug.cgi?id=212663
<rdar://problem/63892899>
Reviewed by Chris Dumez.
Add a debug assert to check if statistics are being merged into the
ITP database after prepareToSuspend() has been called. If this assert
is hit, it means the ITP database thread is not being stopped when
the process is suspended.
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::resourceLoadStatisticsUpdated):
2020-06-02 Per Arne Vollan <pvollan@apple.com>
REGRESSION(r261387): Introduced sandbox violations
https://bugs.webkit.org/show_bug.cgi?id=212634
Reviewed by Darin Adler.
The change set r261387 introduced some new mach lookup sandbox violations, when attempting to close Launch Services connections that were not open.
Fix this by only attempting to close open connections, and do it while holding the extension to the database mapping service.
No new tests, since this patch is fixing benign sandbox violations.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-06-02 Youenn Fablet <youenn@apple.com>
Allow using web processes for service workers even though they loaded about URLs
https://bugs.webkit.org/show_bug.cgi?id=212464
Reviewed by Chris Dumez.
Before the patch, a web process that loads about:blank will not be able to run service workers.
A web process that is running service workers and is loading an about:blank page will also kill its service workers.
To prevent this, do not take into account about:blank URLs for the computation of the registrable domain of a process.
Covered by API tests.
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::didStartProvisionalLoadForMainFrame):
2020-06-02 Tim Horton <timothy_horton@apple.com>
macCatalyst: WKWebView's tint color is not propagated to Web Content process
https://bugs.webkit.org/show_bug.cgi?id=212630
Reviewed by Anders Carlsson.
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
* UIProcess/Cocoa/PageClientImplCocoa.h:
* UIProcess/PageClient.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
* WebProcess/WebPage/WebPage.cpp:
* WebProcess/WebPage/WebPage.h:
2020-06-02 Tim Horton <timothy_horton@apple.com>
REGRESSION (r258949): Dynamic user accent color changes are not respected by WKWebView
https://bugs.webkit.org/show_bug.cgi?id=212628
<rdar://problem/63436275>
Reviewed by Per Arne Vollan.
Two issues with the "CFPrefs Direct Mode" mechanism caused dynamic changes to the
user accent color preference to not be respected anymore:
- the default change, when synchronized to the Web Content process, was
written into the wrong domain, because of bug 212627.
- AppKit caches the resolved accent color; the cache is invalidated when the
a CoreUI notification is posted. Since we synchronize preferences ourselves,
the ordering of the default change and notification are no longer guaranteed,
so AppKit sometimes reads and re-caches the old accent color.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::userAccentColorPreferenceKey):
(WebKit::shouldWriteToAppDomainForPreferenceKey):
(WebKit::WebProcess::notifyPreferencesChanged):
To fix the first problem, write the default into the standardUserDefaults.
This is just a workaround, we need a more general solution to this problem
for all global preferences. We know it's OK in this case, because AppKit
happens to read from standardUserDefaults.
(WebKit::dispatchSimulatedNotificationsForPreferenceChange):
To fix the second problem, re-emit the notification once we have synchronized
the preference to the Web Content process.
2020-06-02 Rob Buis <rbuis@igalia.com>
Make generated C++ code use modern C++
https://bugs.webkit.org/show_bug.cgi?id=190714
Reviewed by Sam Weinig.
Replace typedef usage by alias-declaration.
* Scripts/webkit/messages.py:
2020-06-02 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Implement data list drop down menu
https://bugs.webkit.org/show_bug.cgi?id=212333
Reviewed by Adrian Perez de Castro.
Use a GtkPopover for GTK4 instead of a GtkMenu.
* UIProcess/API/gtk/PageClientImpl.cpp:
(WebKit::PageClientImpl::createDataListSuggestionsDropdown):
* UIProcess/gtk/WebDataListSuggestionsDropdownGtk.cpp:
(WebKit::WebDataListSuggestionsDropdownGtk::WebDataListSuggestionsDropdownGtk):
(WebKit::WebDataListSuggestionsDropdownGtk::~WebDataListSuggestionsDropdownGtk):
(WebKit::WebDataListSuggestionsDropdownGtk::show):
* UIProcess/gtk/WebDataListSuggestionsDropdownGtk.h:
2020-06-02 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Make popup menus work
https://bugs.webkit.org/show_bug.cgi?id=211178
Reviewed by Adrian Perez de Castro.
Use a GtkPopover for GTK4 instead of a GtkMenu.
* UIProcess/API/gtk/PageClientImpl.cpp:
(WebKit::PageClientImpl::createPopupMenuProxy):
* UIProcess/API/gtk/WebKitPopupMenu.cpp:
(WebKit::WebKitPopupMenu::WebKitPopupMenu):
(WebKit::menuCloseCallback):
(WebKit::WebKitPopupMenu::showPopupMenu):
(WebKit::WebKitPopupMenu::hidePopupMenu):
(WebKit::WebKitPopupMenu::cancelTracking):
(WebKit::WebKitPopupMenu::activateItem):
* UIProcess/API/gtk/WebKitPopupMenu.h:
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseSizeAllocate): Iterate children and call check_resize on every native one.
* UIProcess/gtk/WebPopupMenuProxyGtk.cpp:
(WebKit::WebPopupMenuProxyGtk::treeViewButtonReleaseEventCallback):
(WebKit::WebPopupMenuProxyGtk::createPopupMenu):
(WebKit::WebPopupMenuProxyGtk::showPopupMenu):
(WebKit::WebPopupMenuProxyGtk::hidePopupMenu):
(WebKit::WebPopupMenuProxyGtk::treeViewRowActivatedCallback): Deleted.
* UIProcess/gtk/WebPopupMenuProxyGtk.h:
2020-06-01 Chris Dumez <cdumez@apple.com>
Fix thread-safety issue in [WKProcessAssertionBackgroundTaskManager _handleBackgroundTaskExpiration]
https://bugs.webkit.org/show_bug.cgi?id=212615
<rdar://problem/63569049>
Reviewed by Geoffrey Garen.
[WKProcessAssertionBackgroundTaskManager _handleBackgroundTaskExpiration] may get called on a background thread
by RunningBoard. When this happened it would check the _applicationIsBackgrounded flag on a background thread,
which was not safe, it would also call _releaseBackgroundTask on the background thread which definitely was not
safe and could lead to crashes such as the one in the radar.
To address the issue, we now call callOnMainRunLoopAndWait() as soon as _handleBackgroundTaskExpiration gets
called and run the whole implementation on the main thread. Previously, we were already doing a dispatch_sync()
to the main thread but only for part of the implementation.
* UIProcess/ios/ProcessAssertionIOS.mm:
(-[WKProcessAssertionBackgroundTaskManager _handleBackgroundTaskExpiration]):
(-[WKProcessAssertionBackgroundTaskManager _handleBackgroundTaskExpirationOnMainThread]):
2020-06-01 Brent Fulgham <bfulgham@apple.com>
Remove unused WebContent process sandbox syscall
https://bugs.webkit.org/show_bug.cgi?id=212609
<rdar://problem/63764590>
Reviewed by Per Arne Vollan.
After further testing we have determined that the syscall added in
Bug 206260 is not always enabled. Check if the Sandbox recognizes
the rule before using it.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* WebProcess/com.apple.WebProcess.sb.in:
2020-06-01 David Kilzer <ddkilzer@apple.com>
Don't use casts to convert between WebCore::DragDestinationAction and {Web,WK}DragDestinationAction types
<https://webkit.org/b/212507>
Reviewed by Darin Adler.
* Shared/API/Cocoa/_WKDragActionsInternal.h: Add.
(WebKit::coreDragDestinationActionMask):
- Add conversion function to avoid casting.
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::draggingEntered):
(WebKit::WebViewImpl::draggingUpdated):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView dragDataForDropSession:dragDestinationAction:]):
- Make use of coreDragDestinationActionMask() conversion
function.
- Extract `dragOperationMask` varible for future updates.
* WebKit.xcodeproj/project.pbxproj:
- Add _WKDragActionsInternal.h to project.
2020-06-01 Alex Christensen <achristensen@webkit.org>
Fix Glib build after r262391
https://bugs.webkit.org/show_bug.cgi?id=212604
* UIProcess/API/glib/WebKitWebContext.cpp:
(webkitWebContextConstructed):
2020-06-01 Alex Christensen <achristensen@webkit.org>
Revert r250312
https://bugs.webkit.org/show_bug.cgi?id=212604
<rdar://problem/63779241>
Reviewed by Tim Horton.
r250312 was a step towards a network process singleton.
We aren't there yet, so revert it until we are. There was something trying to keep HSTS storages separate, so let them keep trying.
* UIProcess/API/APIProcessPoolConfiguration.cpp:
(API::ProcessPoolConfiguration::copy):
* UIProcess/API/APIProcessPoolConfiguration.h:
* UIProcess/API/Cocoa/_WKProcessPoolConfiguration.mm:
(-[_WKProcessPoolConfiguration setHSTSStorageDirectory:]):
(-[_WKProcessPoolConfiguration hstsStorageDirectory]):
* UIProcess/LegacyGlobalSettings.h:
(WebKit::LegacyGlobalSettings::setHSTSStorageDirectory): Deleted.
(WebKit::LegacyGlobalSettings::hstsStorageDirectory const): Deleted.
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::ensureNetworkProcess):
2020-06-01 Chris Dumez <cdumez@apple.com>
ASSERTION FAILURE (r220931): !m_function in ~CompletionHandler() after switch tabs
https://bugs.webkit.org/show_bug.cgi?id=212537
<rdar://problem/63766838>
Reviewed by Alex Christensen.
When WebPage::markAllLayersVolatile(), it would destroy m_pageMarkingLayersAsVolatileCounter,
which may not have called its completion handler yet. As a result, we would hit an assertion
in the CompletionHandler destructor.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::callVolatilityCompletionHandlers):
(WebKit::WebPage::markLayersVolatile):
(WebKit::WebPage::cancelMarkLayersVolatile):
* WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::markLayersVolatile):
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::prepareToSuspend):
(WebKit::WebProcess::markAllLayersVolatile):
(WebKit::WebProcess::cancelMarkAllLayersVolatile):
* WebProcess/WebProcess.h:
2020-06-01 Sam Weinig <weinig@apple.com>
Extended Color: Replace Color constructors taking numeric values with type specific factory functions
https://bugs.webkit.org/show_bug.cgi?id=212576
Reviewed by Tim Horton.
Replaces all remaining implicit and explicit uses of the Color constructors taking numeric
values with explicit calls to makeSimpleColor/makeSimpleColorFromFloats/makeExtendedColor,
giving us a consistent way to create colors. Also addes use constexpr SimpleColors where possible.
* Shared/RemoteLayerTree/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::drawInContext):
* UIProcess/API/wpe/WebKitColor.cpp:
(webkitColorToWebCoreColor):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _tapHighlightColorForFastClick:]):
* WebProcess/Inspector/WebInspectorClient.cpp:
(WebKit::WebInspectorClient::showPaintRect):
* WebProcess/WebPage/FindController.cpp:
(WebKit::FindController::drawRect):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::paintSnapshotAtSize):
* WebProcess/WebPage/ios/FindControllerIOS.mm:
(WebKit::FindIndicatorOverlayClientIOS::drawRect):
(WebKit::highlightColor): Deleted.
2020-06-01 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Add printing support
https://bugs.webkit.org/show_bug.cgi?id=212320
Reviewed by Adrian Perez de Castro.
Printing API hasn't changed, the only problem was that we were using gtk_dialog_run() and gdk threads deprecated API.
* UIProcess/API/gtk/WebKitPrintOperation.cpp:
(webkitPrintOperationRunDialog): Remove GTK4 ifdefs.
* WebProcess/WebPage/gtk/WebPrintOperationGtk.cpp:
(WebKit::WebPrintOperationGtk::print): Use g_idle_add_full() instead of gdk threads deprecated API.
2020-06-01 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Wheel events are always handled by the main frame view
https://bugs.webkit.org/show_bug.cgi?id=212593
Reviewed by Adrian Perez de Castro.
It's not possible scroll other scrollable areas with the mouse wheel. This is because we are using the given x,
y delta in the scroll signal as the position. In GTK4 the scroll event doesn't include a position, so we have to
use the last motion event.
* Shared/NativeWebWheelEvent.h:
* Shared/gtk/NativeWebWheelEventGtk.cpp:
(WebKit::NativeWebWheelEvent::NativeWebWheelEvent): Add wheelTicks parameter.
* Shared/gtk/WebEventFactory.cpp:
(WebKit::WebEventFactory::createWebWheelEvent): New create function that receives the wheelTicks.
* Shared/gtk/WebEventFactory.h:
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseScroll): Use the x, y parameters as the delta and take the position from the last motion event.
2020-06-01 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Monitor root window to update activity state
https://bugs.webkit.org/show_bug.cgi?id=212581
Reviewed by Adrian Perez de Castro.
To update ActivityState::IsInWindow and ActivityState::WindowIsActive.
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(toplevelWindowStateEvent):
(webkitWebViewBaseSetToplevelOnScreenWindow):
(webkitWebViewBaseDispose):
(toplevelWindowIsActiveChanged):
(toplevelWindowStateChanged):
(toplevelWindowRealized):
(toplevelWindowUnrealized):
(webkitWebViewBaseRoot):
(webkitWebViewBaseUnroot):
(webkit_web_view_base_class_init):
2020-06-01 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Fix pointer lock in X11
https://bugs.webkit.org/show_bug.cgi?id=212592
Reviewed by Adrian Perez de Castro.
I forgot to forward the motion events for X11 in GTK4.
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(MotionEvent::MotionEvent): Add a constructor that receives the positions and state.
(webkitWebViewBaseMotion): Notify the pointer lock manager about the event. Also save the last motion event and
compute the movement delta.
2020-06-01 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Make inspector work
https://bugs.webkit.org/show_bug.cgi?id=212321
Reviewed by Adrian Perez de Castro.
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseRemoveDialog): Added instead of the generic webkitWebViewBaseRemoveChild that has been removed.
(webkitWebViewBaseRemoveWebInspector): Just unparent the inspector view instead of caling webkitWebViewBaseRemoveChild.
(webkitWebViewBaseDispose): Use specific functions to remove the children.
(webkitWebViewBaseSnapshot): Call gtk_widget_snapshot_child() on inspector view if present.
(webkitWebViewBaseRemoveChild): Deleted.
* UIProcess/Inspector/gtk/RemoteWebInspectorProxyGtk.cpp:
(WebKit::RemoteWebInspectorProxy::platformSave): Remove GTK4 ifdefs.
* UIProcess/Inspector/gtk/WebInspectorProxyGtk.cpp:
(WebKit::WebInspectorProxy::platformSave): Ditto.
* UIProcess/Inspector/gtk/WebKitInspectorWindow.cpp:
(webkit_inspector_window_init): Show title buttons in inspector window.
* UIProcess/gtk/AcceleratedBackingStoreWayland.cpp:
(WebKit::AcceleratedBackingStoreWayland::snapshot): Use the texture size instead of the view size, since it's
different when the inspector is attached.
2020-06-01 Carlos Garcia Campos <cgarcia@igalia.com>
REGRESSION(r262184): fast/forms/implicit-submission.html is crashing
https://bugs.webkit.org/show_bug.cgi?id=212520
Reviewed by Adrian Perez de Castro.
* UIProcess/gtk/WebPopupMenuProxyGtk.cpp:
(WebKit::WebPopupMenuProxyGtk::activateSelectedItem): Return early if the popup has already been closed.
2020-06-01 Rob Buis <rbuis@igalia.com>
Rename ResourceResponseBase::isHTTP to isInHTTPFamily
https://bugs.webkit.org/show_bug.cgi?id=208782
Reviewed by Sam Weinig.
Adapt to API change.
* NetworkProcess/cache/NetworkCache.cpp:
(WebKit::NetworkCache::makeStoreDecision):
* WebProcess/Plugins/PluginView.cpp:
(WebKit::buildHTTPHeaders):
2020-06-01 Michael Catanzaro <mcatanzaro@gnome.org>
[WPE][GTK] pthread_create() fails with EPERM in the second WebKitWebProcess with sandbox on
https://bugs.webkit.org/show_bug.cgi?id=212380
Reviewed by Adrian Perez de Castro.
Dereference scmp_arg_cmp before use.
* UIProcess/Launcher/glib/BubblewrapLauncher.cpp:
(WebKit::setupSeccomp):
2020-05-30 Alex Christensen <achristensen@webkit.org>
Deprecate WKBundlePostSynchronousMessage
https://bugs.webkit.org/show_bug.cgi?id=212536
Reviewed by Anders Carlsson.
* WebProcess/InjectedBundle/API/c/WKBundle.h:
2020-05-30 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] WebDriver: stop using GdkEvent API in preparation for GTK4
https://bugs.webkit.org/show_bug.cgi?id=212465
Reviewed by Adrian Perez de Castro.
Use the new internal API to synthesize events.
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseSynthesizeKeyEvent): Add type parameter, since WebDriver can send press and release events
separately. Also add a parameter to indicate that keyval should be translated according to current state that is
required by WebDriver.
* UIProcess/API/gtk/WebKitWebViewBaseInternal.h:
* UIProcess/Automation/WebAutomationSession.cpp:
(WebKit::WebAutomationSession::documentLoadedForFrame): Call resetClickCount().
(WebKit::WebAutomationSession::updateClickCount): Update the click count for the given button position and
maximum time and distance.
(WebKit::WebAutomationSession::resetClickCount): Reset the click count.
* UIProcess/Automation/WebAutomationSession.h:
* UIProcess/Automation/gtk/WebAutomationSessionGtk.cpp:
(WebKit::WebAutomationSession::platformSimulateMouseInteraction): Use webkitWebViewBaseSynthesizeMouseEvent().
(WebKit::WebAutomationSession::platformSimulateKeyboardInteraction): Use webkitWebViewBaseSynthesizeKeyEvent().
(WebKit::WebAutomationSession::platformSimulateKeySequence): Ditto.
(WebKit::doMouseEvent): Deleted.
(WebKit::doMotionEvent): Deleted.
(WebKit::doKeyStrokeEvent): Deleted.
* UIProcess/gtk/KeyBindingTranslator.cpp: Add missing key shortcuts to predefined list.
2020-05-29 Alex Christensen <achristensen@webkit.org>
Use correct encoding when converting a WTF::URL to CFURLRef
https://bugs.webkit.org/show_bug.cgi?id=212486
Reviewed by Darin Adler.
* Shared/API/c/cf/WKURLCF.mm:
(WKURLCopyCFURL):
* Shared/cf/ArgumentCodersCF.cpp:
(IPC::decode):
2020-05-29 Peng Liu <peng.liu6@apple.com>
Replace callOnMainThread() with callOnMainRunLoop() in AudioSessionRoutingArbitratorProxyCocoa.mm
https://bugs.webkit.org/show_bug.cgi?id=212553
Reviewed by Chris Dumez.
Use callOnMainRunLoop() instead of callOnMainThread() in the UIProcess. Also move
m_setupArbitrationOngoing flag to the end of the completion handler.
No new tests, no functional change.
* UIProcess/Media/cocoa/AudioSessionRoutingArbitratorProxyCocoa.mm:
(WebKit::SharedArbitrator::beginRoutingArbitrationForArbitrator):
2020-05-29 Brent Fulgham <bfulgham@apple.com>
[Cocoa] Improve logging quality for non-ephemeral sessions
https://bugs.webkit.org/show_bug.cgi?id=212551
<rdar://problem/62461099>
Reviewed by David Kilzer.
In Bug 209522 I switched normal mode logging to use the same privacy-protecting mode we use for ephemeral sessions.
This had the unintended consequence of removing network load data used to investigate networking issues.
This patch adopts the more fine-grained logging provided by the low-level 'nw_context_privacy_level' setting.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::configurationForSessionID): Adopt 'nw_context_privacy_level' setting.
2020-05-29 Andy Estes <aestes@apple.com>
[Apple Pay] Remove conditionals for ENABLE_APPLE_PAY_SESSION_V(3|4)
https://bugs.webkit.org/show_bug.cgi?id=212541
Reviewed by Darin Adler.
APPLE_PAY_SESSION_V(3|4) is now enabled whenever APPLE_PAY itself is enabled.
* Configurations/FeatureDefines.xcconfig:
2020-05-29 Dean Jackson <dino@apple.com>
iOS build fix.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::configurationForSessionID):
2020-05-29 Jer Noble <jer.noble@apple.com>
[Mac,WK2] Fullscreen animation missing a few frames at beginning
https://bugs.webkit.org/show_bug.cgi?id=212156
<rdar://problem/54799415>
Reviewed by Eric Carlson.
When starting the enter fullscreen animation, ensure that the fullscreen window is ordered front, and on top
of the content, as well as having all the animations configured so that their starting state is in place before
calling -[NSWindow enterFullscreenMode:]. Move all the window creation code into
-beganEnterFullScreenWithInitialFrame:finalFrame: from _startEnterFullScreenAnimationWithDuration:. Re-use the
existing zoomAnimation() and maskAnimation() utility methods, but give the animations a very long duration
(since there is no explicit way to start and stop a CAAnimation). This initial animation will be replaced with
the final one inside -_startEnterFullScreenAnimationWithDuration:. Separately, explictly disable implicit
animations of the fullscreen window during -orderIn: and -orderOut:.
* UIProcess/mac/WKFullScreenWindowController.mm:
(-[WKFullScreenWindowController initWithWindow:webView:page:impl:]):
(-[WKFullScreenWindowController beganEnterFullScreenWithInitialFrame:finalFrame:]):
(-[WKFullScreenWindowController _startEnterFullScreenAnimationWithDuration:]):
2020-05-29 Jan-Michael Brummer <jan.brummer@tabos.org>
[GTK][WPE] API for WebView audio mute support
https://bugs.webkit.org/show_bug.cgi?id=176119
Reviewed by Michael Catanzaro.
Test implemented in TestWebKitWebView.
* UIProcess/API/glib/WebKitWebView.cpp:
(webkitWebViewSetProperty):
(webkitWebViewGetProperty):
(webkit_web_view_class_init):
(webkit_web_view_set_is_muted):
(webkit_web_view_is_muted):
* UIProcess/API/gtk/WebKitWebView.h:
* UIProcess/API/gtk/docs/webkit2gtk-4.0-sections.txt:
* UIProcess/API/wpe/WebKitWebView.h:
* UIProcess/API/wpe/docs/wpe-1.0-sections.txt:
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::isAudioMuted const):
2020-05-29 Darin Adler <darin@apple.com>
Remove things from FeatureDefines.xcconfig that are covered by PlatformEnableCocoa.h
https://bugs.webkit.org/show_bug.cgi?id=212418
Rubber-stamped by Simon Fraser.
* Configurations/FeatureDefines.xcconfig: Add back ENABLE_CSS_CONIC_GRADIENTS, removed
by accident.
2020-05-29 Peng Liu <peng.liu6@apple.com>
ASSERT NOT REACHED in IPC::takeAsyncReplyHandler under WebKit::AudioSessionRoutingArbitrator::beginRoutingArbitrationWithCategory
https://bugs.webkit.org/show_bug.cgi?id=212533
Reviewed by Chris Dumez.
When the m_inDispatchMessageMarkedToUseFullySynchronousModeForTesting flag is set,
sendWithAsyncReply() may actually send the message synchronously. Therefore, we have
to set the async reply handler before actually sending the message.
No new tests, fixing test crashes due to assertion failures.
* Platform/IPC/Connection.h:
(IPC::Connection::sendWithAsyncReply):
2020-05-27 Darin Adler <darin@apple.com>
Remove things from FeatureDefines.xcconfig that are covered by PlatformEnableCocoa.h
https://bugs.webkit.org/show_bug.cgi?id=212418
Reviewed by Andy Estes.
* Configurations/FeatureDefines.xcconfig: Removed 83 of the 119 things defined in
this file. There are 36 more that are slightly more complex that we can remove
carefully later.
2020-05-29 Darin Adler <darin@apple.com>
[Cocoa] Pass all defines from Platform.h to various scripts, not just the ones from .xcconfig
https://bugs.webkit.org/show_bug.cgi?id=212451
Reviewed by Sam Weinig.
* DerivedSources.make: Use ":=" when calling $(shell) to make sure the same shell
command is not invoked over and over again.
2020-05-29 Keith Rollin <krollin@apple.com>
Revert switch to XCBuild
https://bugs.webkit.org/show_bug.cgi?id=212530
<rdar://problem/63764632>
Unreviewed build fix.
Bug 209890 enabled the use of XCBuild by default. Since then, some
build issues have shown up. While addressing them, temporarily turn
off the use of XCBuild by default.
* Configurations/WebKit.xcconfig:
* WebKit.xcodeproj/project.pbxproj:
2020-05-29 Devin Rousso <drousso@apple.com>
Web Inspector: add ITML debuggable/target type
https://bugs.webkit.org/show_bug.cgi?id=203300
<rdar://problem/56545896>
Reviewed by Joseph Pecoraro and Brian Burg.
* UIProcess/API/Cocoa/_WKInspectorDebuggableInfo.h:
* UIProcess/API/Cocoa/_WKInspectorDebuggableInfoInternal.h:
(fromWKInspectorDebuggableType):
(toWKInspectorDebuggableType):
* UIProcess/API/Cocoa/_WKRemoteWebInspectorViewController.h:
* UIProcess/API/Cocoa/_WKRemoteWebInspectorViewController.mm:
(legacyDebuggableTypeToModernDebuggableType):
* UIProcess/socket/RemoteInspectorProtocolHandler.cpp:
(WebKit::parseDebuggableTypeFromString):
2020-05-29 Simon Fraser <simon.fraser@apple.com>
Prepare for async scrolling in passive wheel event handler regions
https://bugs.webkit.org/show_bug.cgi?id=212455
Reviewed by Tim Horton.
Clarify the processing for wheel events by adding OptionSet<WheelEventProcessingSteps>,
which will, in future, allow us to describe the processing for an event in the passive
event handler region which does scrolling on the scrolling thread, and is then sent
to the main thread for DOM event dispatch.
Removed ScrollingEventResult, which conflated "handled" with "send to another thread".
The thread sending behavior is now encoded in the WheelEventProcessingSteps, and we can just
use a bool for handled.
Scrolling tree and node handleWheelEvent() functions return a WheelEventHandlingResult, which
is a tuple of OptionSet<WheelEventProcessingSteps> and 'handled', allowing for a node with
background-attachment:fixed to add the "send to main thread" processing step.
* UIProcess/RemoteLayerTree/RemoteScrollingCoordinatorProxy.cpp:
(WebKit::RemoteScrollingCoordinatorProxy::handleWheelEvent):
* UIProcess/RemoteLayerTree/mac/ScrollingTreeFrameScrollingNodeRemoteMac.cpp:
(WebKit::ScrollingTreeFrameScrollingNodeRemoteMac::handleWheelEvent):
* UIProcess/RemoteLayerTree/mac/ScrollingTreeFrameScrollingNodeRemoteMac.h:
* UIProcess/RemoteLayerTree/mac/ScrollingTreeOverflowScrollingNodeRemoteMac.cpp:
(WebKit::ScrollingTreeOverflowScrollingNodeRemoteMac::handleWheelEvent):
* UIProcess/RemoteLayerTree/mac/ScrollingTreeOverflowScrollingNodeRemoteMac.h:
* WebProcess/WebPage/EventDispatcher.cpp:
(WebKit::EventDispatcher::wheelEvent):
2020-05-29 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Implement HTTP auth dialog
https://bugs.webkit.org/show_bug.cgi?id=212319
Reviewed by Sergio Villar Senin.
* UIProcess/API/gtk/WebKitAuthenticationDialog.cpp:
(webkitAuthenticationDialogDestroy):
(okButtonClicked):
(cancelButtonClicked):
(authenticationCancelled):
(webkitAuthenticationDialogInitialize):
(webkitAuthenticationDialogMap):
(webkitAuthenticationDialogDispose):
(webkitAuthenticationDialogNew):
* UIProcess/API/gtk/WebKitAuthenticationDialog.h:
* UIProcess/API/gtk/WebKitWebViewGtk.cpp:
(webkitWebViewAuthenticate):
2020-05-29 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Implement script dialogs
https://bugs.webkit.org/show_bug.cgi?id=212318
Reviewed by Adrian Perez de Castro.
Adapt to the GTK4 API and theme changes.
* UIProcess/API/gtk/WebKitScriptDialogImpl.cpp:
(webkitScriptDialogImplClose):
(webkitScriptDialogImplKeyPressed):
(webkitScriptDialogImplMap):
(webkitScriptDialogImplConstructed):
(webkitScriptDialogImplDispose):
(webkit_script_dialog_impl_class_init):
(webkitScriptDialogImplAddButton):
(webkitScriptDialogImplNew):
(webkitScriptDialogImplSetEntryText):
* UIProcess/API/gtk/WebKitScriptDialogImpl.h:
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseAddDialog):
(webkitWebViewBaseRemoveChild):
(webkitWebViewBaseSnapshot):
* UIProcess/API/gtk/WebKitWebViewDialog.cpp:
(webkitWebViewDialogSnapshot):
(webkitWebViewDialogSizeAllocate):
(webkitWebViewDialogConstructed):
(webkit_web_view_dialog_class_init):
(webkitWebViewDialogSetChild):
(webkitWebViewDialogGetChild):
* UIProcess/API/gtk/WebKitWebViewDialog.h:
* UIProcess/API/gtk/WebKitWebViewGtk.cpp:
(webkitWebViewScriptDialog):
2020-05-29 Wenson Hsieh <wenson_hsieh@apple.com>
REGRESSION (r261812): editing/async-clipboard/clipboard-item-get-type-basic.html is flaky
https://bugs.webkit.org/show_bug.cgi?id=212281
<rdar://problem/63554912>
Reviewed by Tim Horton.
Declare `-delegateSupportsImagePaste` on UIKeyboardImpl.
* Platform/spi/ios/UIKitSPI.h:
2020-05-28 Peng Liu <peng.liu6@apple.com>
Clean up WebKit.xcodeproj/project.pbxproj
https://bugs.webkit.org/show_bug.cgi?id=212491
Reviewed by Tim Horton.
A follow-up patch to add AudioSessionRoutingArbitratorProxyCocoa.mm to the unified build.
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
Add AudioSessionRoutingArbitratorProxyCocoa.mm to the unified build.
* UIProcess/WebAuthentication/Cocoa/WebAuthenticationPanelClient.h:
Fix a unified build failure.
2020-05-28 Megan Gardner <megan_gardner@apple.com>
Avoid unnecessary sync IPC messages when togging the callout bar for selections.
https://bugs.webkit.org/show_bug.cgi?id=212508
The loupe gesture only needs to be activated and evaluated if the tap is inside
an existing selectionView. We can do that test in the UIProcess without resorting to a sync IPC message.
Doing that evaluation locally will eliminate unnecessary hangs in the UIProcess.
Reviewed by Wenson Hsieh.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _pointIsInsideSelectionRect:outBoundingRect:]):
(-[WKContentView _shouldToggleSelectionCommandsAfterTapAt:]):
(-[WKContentView textInteractionGesture:shouldBeginAtPoint:]):
2020-05-28 Tim Horton <timothy_horton@apple.com>
Fix the macOS build
* WebKit.xcodeproj/project.pbxproj:
Can't have this file in both the Xcode target and SourcesCocoa.txt...
2020-05-28 Lauro Moura <lmoura@igalia.com>
[GTK][WPE] Buildfix after r262242
Unreviewed build fix.
* Shared/glib/ArgumentCodersGLib.cpp:
(IPC::decode):
2020-05-28 Jonathan Bedard <jbedard@apple.com>
watchOS: Upstream WKNumberPadView
https://bugs.webkit.org/show_bug.cgi?id=212499
<rdar://problem/63736073>
Reviewed by Wenson Hsieh.
* UIProcess/ios/forms/WKNumberPadView.h: Moved from WebKitAdditions.
* UIProcess/ios/forms/WKNumberPadView.mm: Ditto.
2020-05-28 Jonathan Bedard <jbedard@apple.com>
watchOS: Upstream WKTextInputListViewController
https://bugs.webkit.org/show_bug.cgi?id=212495
<rdar://problem/63733949>
Reviewed by Wenson Hsieh.
* UIProcess/ios/forms/WKTextInputListViewController.h: Moved from WebKitAdditions.
* UIProcess/ios/forms/WKTextInputListViewController.mm: Ditto.
2020-05-28 Simon Fraser <simon.fraser@apple.com>
Simplify EventDispatcher wheel event dispatch
https://bugs.webkit.org/show_bug.cgi?id=212490
Reviewed by Tim Horton.
The various cross-thread bounces and completion lambdas in EventDispatcher::wheelEvent()
and ScrollingTree code made the logic very hard to follow.
Moving the ScrollingThread::dispatch() into EventHandler code simplifies things a little,
and allows for removal of the hokey "try to handle" ScrollingTree function, as well
as the clunky completion function.
Now, EventHandler call shouldHandleWheelEventSynchronously(), then does the
ScrollingThread::dispatch() which allows the lambda to easily call back into
EventHandler for the main thread dispatch.
RemoteScrollingCoordinatorProxy/RemoteLayerTree code is unused at present, and
will need work.
* UIProcess/RemoteLayerTree/RemoteScrollingCoordinatorProxy.cpp:
(WebKit::RemoteScrollingCoordinatorProxy::handleWheelEvent):
* UIProcess/RemoteLayerTree/RemoteScrollingTree.cpp:
(WebKit::RemoteScrollingTree::tryToHandleWheelEvent): Deleted.
* UIProcess/RemoteLayerTree/RemoteScrollingTree.h:
* WebProcess/WebPage/EventDispatcher.cpp:
(WebKit::EventDispatcher::wheelEvent):
(WebKit::EventDispatcher::dispatchWheelEventViaMainThread):
* WebProcess/WebPage/EventDispatcher.h:
2020-05-28 Peng Liu <peng.liu6@apple.com>
Clean up WebKit.xcodeproj/project.pbxproj
https://bugs.webkit.org/show_bug.cgi?id=212491
Reviewed by Simon Fraser.
Remove the references to deleted files and merge duplicated folders.
* WebKit.xcodeproj/project.pbxproj:
2020-05-28 Kate Cheney <katherine_cheney@apple.com>
Minimum user interaction time in ResourceLoadStatistics should handle the case of -1
https://bugs.webkit.org/show_bug.cgi?id=212445
<rdar://problem/63696470>
Reviewed by John Wilander.
Tests: http/tests/resourceLoadStatistics/operating-dates-all-but-cookies-removed-statistics-with-no-user-interaction-database.html
http/tests/resourceLoadStatistics/operating-dates-all-but-cookies-removed-statistics-with-no-user-interaction.html
Adds a getter for mostRecentUserInteractionTime which returns WTF::nullopt if the
timestamp is -1. Then does not consider this case in calculating the
minimum timestamp.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::getMostRecentUserInteractionTime):
(WebKit::ResourceLoadStatisticsDatabaseStore::registrableDomainsToDeleteOrRestrictWebsiteDataFor):
Flip sign to be less than, so we hold off on deleting data if the
oldest interaction was less than the minimum time between removal.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:
(WebKit::ResourceLoadStatisticsMemoryStore::getMostRecentUserInteractionTime):
(WebKit::ResourceLoadStatisticsMemoryStore::registrableDomainsToDeleteOrRestrictWebsiteDataFor):
Flip sign to be less than, so we hold off on deleting data if the
oldest interaction was less than the minimum time between removal.
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.h:
2020-05-28 Megan Gardner <megan_gardner@apple.com>
Do not send a second sync request for positition information to the web process if we have not recieved information since the previous sync request.
https://bugs.webkit.org/show_bug.cgi?id=212289
<rdar://problem/58494578>
Reviewed by Tim Horton.
If we have sent a sync requests to the web process for position information, and timed out, and have not
received a message with position information in the interim, do not send another sync request. The web
process is likely still hung, and there is no reason to hang the UIProcess again if we suspect that it
is unlikely that we will receive a reply.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView cleanUpInteraction]):
(-[WKContentView ensurePositionInformationIsUpToDate:]):
(-[WKContentView _positionInformationDidChange:]):
2020-05-28 Youenn Fablet <youenn@apple.com>
ReadableByteStream should be enabled/disabled in service workers as done in pages
https://bugs.webkit.org/show_bug.cgi?id=212466
Reviewed by Alex Christensen.
Introduce a WebPreference for readable byte stream.
Use it for enabling it in web processes running service workers as per the given store.
* Shared/WebPreferences.yaml:
* WebProcess/Storage/WebSWContextManagerConnection.cpp:
(WebKit::WebSWContextManagerConnection::updatePreferencesStore):
2020-05-28 Michael Catanzaro <mcatanzaro@gnome.org>
[WPE][GTK] GVariant decoding must copy the serialized data
https://bugs.webkit.org/show_bug.cgi?id=212441
Reviewed by Carlos Garcia Campos.
I tracked this down to ArgumentCodersGLib.cpp. The problem is that we construct a GVariant
using g_variant_new_from_data(), which does not copy or take ownership of the data, so here
we accidentally create the GVariant using data we don't own. (Here, the data is owned by the
Decoder itself in its internal m_buffer.) Anyway, this is fixable by manually copying and
freeing it with the GDestroyNotify parameter, but it's easier to switch to
g_variant_new_from_bytes() because GBytes takes ownership when constructed.
* Shared/glib/ArgumentCodersGLib.cpp:
(IPC::decode):
2020-05-28 Youenn Fablet <youenn@apple.com>
RemoteAudio::audioSamplesAvailable should check for m_buffer to be null
https://bugs.webkit.org/show_bug.cgi?id=212462
<rdar://problem/63627642>
Reviewed by Eric Carlson.
* WebProcess/cocoa/RemoteCaptureSampleManager.cpp:
(WebKit::RemoteCaptureSampleManager::RemoteAudio::audioSamplesAvailable):
m_buffer is initialized by a StorageChanged IPC message which might not always be successful, for instance if the shared memory mapping fails.
Return early if m_buffer is not yet initialized properly.
2020-05-28 Youenn Fablet <youenn@apple.com>
RealtimeIncomingVideoSourceCocoa::OnFrame should use video frame timestamp
https://bugs.webkit.org/show_bug.cgi?id=212402
Reviewed by Eric Carlson.
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayer.cpp:
(WebKit::RemoteSampleBufferDisplayLayer::enqueueSample):
* GPUProcess/webrtc/RemoteSampleBufferDisplayLayer.h:
2020-05-27 Keith Rollin <krollin@apple.com>
Add output file specifications to "Copy Custom WebContent Resources to Framework Private Headers" build phase in WebKit/WebContent
https://bugs.webkit.org/show_bug.cgi?id=212452
<rdar://problem/63700546>
Reviewed by Timothy Hatcher.
The "Copy Custom WebContent Resources to Framework Private Headers"
build phase in the WebContent target of the WebKit project doesn't
advertise the files that it produces. This has consequences on later
build phases that consume those files. In particular, XCBuild is left
not knowing where those files are coming from, and it fails saying
that those files don't exist. Fix this by adding the output files to
the "output files" section of the custom build phase.
No new tests -- this is just a build fix.
* WebKit.xcodeproj/project.pbxproj:
2020-05-27 Tim Horton <timothy_horton@apple.com>
REGRESSION (PSON): Resizing WKWebView during cross-site navigation results in the page laying out at the wrong size
https://bugs.webkit.org/show_bug.cgi?id=212434
<rdar://problem/51621621>
Reviewed by Chris Dumez.
New API test: ProcessSwap.ResizeWebViewDuringCrossSiteProvisionalNavigation.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::swapToProvisionalPage):
Set m_hasRunningProcess before calling setDrawingArea, instead of immediately after,
so that setDrawingArea can call methods that check m_hasRunningProcess.
(WebKit::WebPageProxy::setDrawingArea):
When swapping DrawingAreas, push the current view size down. DrawingArea
is initially created with a size, but the size can change while the DrawingArea
is still provisional.
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::viewSize):
Simplify this implementation now that we don't ever need to support WKContentView
outside of WKWebView (since 2014 or so).
2020-05-27 Michael Catanzaro <mcatanzaro@gnome.org>
REGRESSION(r260318): [WPE][GTK] Uninitialized memory read in MemoryPressureMonitor
https://bugs.webkit.org/show_bug.cgi?id=212412
Reviewed by Adrian Perez de Castro.
I think this is a false-positive, but let's suppress the warning by zero-initializing this
buffer.
* UIProcess/linux/MemoryPressureMonitor.cpp:
(WebKit::systemMemoryUsedAsPercentage):
2020-05-27 Kate Cheney <katherine_cheney@apple.com>
App-bound sessions should be for iOS only
https://bugs.webkit.org/show_bug.cgi?id=212426
<rdar://problem/63678147>
Reviewed by Darin Adler.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::SessionWrapper::initialize):
(WebKit::NetworkSessionCocoa::sessionWrapperForTask):
2020-05-27 Lauro Moura <lmoura@igalia.com>
[WPE] Avoid possible WTR event timing going back in time and fix conversion from s to ms
https://bugs.webkit.org/show_bug.cgi?id=212338
Reviewed by Michael Catanzaro.
When converting the input timestamps to WallTime, we must somehow
adjust them to give meaningful timestamps (due to Dom::Event
conversion from timestamp wall time to the time origin from document
loading).
As WPE timestamps are just intended to be compared among themselves,
we use the first processed event as the origin, comparing the
following events to it to get an actual time.
Test: fast/events/monotonic-event-time.html
* Shared/libwpe/WebEventFactory.cpp:
(WebKit::wallTimeForEventTime):
2020-05-27 Kate Cheney <katherine_cheney@apple.com>
Make sure bundle identifier testing override is set in the network process
https://bugs.webkit.org/show_bug.cgi?id=212288
<rdar://problem/63539061>
Reviewed by Chris Dumez.
Set the bundle identifier in the network process if a layout test
updates it. This value usually only gets updated when a new network
process is initialized, which does not happen for every test. So it
should be set/cleared as needed.
No new tests, will fix http/tests/in-app-browser-privacy/ tests.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::updateBundleIdentifierInNetworkProcess):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreUpdateBundleIdentifierInNetworkProcess):
* UIProcess/API/C/WKWebsiteDataStoreRef.h:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::updateBundleIdentifierInNetworkProcess):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::updateBundleIdentifierInNetworkProcess):
* UIProcess/WebsiteData/WebsiteDataStore.h:
2020-05-27 Per Arne Vollan <pvollan@apple.com>
Unreviewed, reverting r260840.
This patch is no longer needed after preferences root cause
has been fixed.
Reverted changeset:
"[Cocoa] Global preferences are not accessible in the
WebContent process when CFPrefs direct mode is enabled"
https://bugs.webkit.org/show_bug.cgi?id=211075
https://trac.webkit.org/changeset/260840
2020-05-27 Michael Catanzaro <mcatanzaro@gnome.org>
Unreviewed, fix -Wcomment build warning after r261986
https://bugs.webkit.org/show_bug.cgi?id=211950
<rdar://problem/63482718>
If we replace the /* with /&ast; then GCC won't complain that it looks like a comment.
* UIProcess/API/glib/WebKitSettings.cpp:
(webkit_settings_class_init):
2020-05-27 Brent Fulgham <bfulgham@apple.com>
Adjust HTTP3 label to the more correct HTTP/3
https://bugs.webkit.org/show_bug.cgi?id=212416
<rdar://problem/63514072>
Reviewed by Darin Adler.
No new tests. No change in behavior.
* Shared/WebPreferences.yaml: Correct label.
2020-05-27 Devin Rousso <drousso@apple.com>
Web Inspector: replace `featureGuard` and `availability` with a combined `condition` that accepts any macro
https://bugs.webkit.org/show_bug.cgi?id=210014
Reviewed by Brian Burg.
Previously, the generated InspectorBackendCommands.js would include code for things that the
backend doesn't actually support. By using actual macros and preprocessing that file, we can
ensure that the frontend doesn't incorrectly think that something is supported by the page
being inspected:
- the `Canvas` commands and events related to shader programs/pipelines should only exist
when the corresponding context type exists, namely `ENABLE(WEBGL)` and `ENABLE(WEBGPU)`.
- iOS doesn't support showing rulers, so create a variant of `DOM.setInspectModeEnabled`
that only exists for `PLATFORM(IOS_FAMILY)` that doesn't have the `showRulers` optional
parameter, as well as removing `Page.setShowRulers` entirely.
- setting the forced appearance should only be possible if dark mode is supported.
- web archives only exist if CF is used.
* Configurations/FeatureDefines.xcconfig:
Add `ENABLE_WEB_ARCHIVE` since it's always enabled in wtf/PlatformEnableCocoa.h.
* DerivedSources.make:
* PlatformMac.cmake:
Remove `platform` as it is handled by `condition`.
* UIProcess/Automation/Automation.json:
2020-05-27 Zalan Bujtas <zalan@apple.com>
[Quirk] Remove stackoverflow.com desktop content quirk.
https://bugs.webkit.org/show_bug.cgi?id=212414
Reviewed by Antti Koivisto.
Mobile content is fixed now.
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::desktopClassBrowsingRecommendedForRequest):
2020-05-27 Chris Dumez <cdumez@apple.com>
Crash under WebNotificationManager::removeNotificationFromContextMap()
https://bugs.webkit.org/show_bug.cgi?id=212410
<rdar://63496726>
Reviewed by Alex Christensen.
Protect the notification object in WebNotificationManager::didDestroyNotification() since
removing it from m_notificationIDMap will deref it and may end up destroying the notification
object otherwise.
Test: http/tests/notifications/crash-on-navigation.html
* WebProcess/Notifications/WebNotificationManager.cpp:
(WebKit::WebNotificationManager::didDestroyNotification):
2020-05-27 Carlos Garcia Campos <cgarcia@igalia.com>
Unreviewed. [GTK] Make webkitWebViewBaseSnapshot static
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseSnapshot):
2020-05-27 David Kilzer <ddkilzer@apple.com>
Use OptionSet<DragDestinationAction> for mask values
<https://webkit.org/b/212115>
<rdar://problem/63423380>
Reviewed by Alex Christensen.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<DragData>::encode):
(IPC::ArgumentCoder<DragData>::decode):
- Switch from decodeEnum() and encodeEnum() to modern
equivalents that check for valid enum values.
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::draggingEntered):
(WebKit::WebViewImpl::draggingUpdated):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView dragDataForDropSession:dragDestinationAction:]):
* WebProcess/WebCoreSupport/WebDragClient.cpp:
(WebKit::WebDragClient::willPerformDragDestinationAction):
2020-05-27 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Make PointerLock work
https://bugs.webkit.org/show_bug.cgi?id=212314
Reviewed by Adrian Perez de Castro.
Use XLib API for warp and grab, since they are no longer available in GTK4. The grab was not actuially needed
under Wayland. Also stop using GdkEvent to keep track of motion events and create the PointerLockManager to be
able to test pointer lock and movement delta.
* Shared/NativeWebMouseEvent.h:
* Shared/gtk/NativeWebMouseEventGtk.cpp:
(WebKit::NativeWebMouseEvent::NativeWebMouseEvent): Use FloatSize instead of FloatPoint for movement delta.
* Shared/gtk/WebEventFactory.cpp:
(WebKit::WebEventFactory::createWebMouseEvent): Ditto.
* Shared/gtk/WebEventFactory.h:
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(MotionEvent::MotionEvent): Helper struct to keep track of motion events without using GdkEvent.
(MotionEvent::delta): Return the movement delta for the given current event.
(webkitWebViewBaseHandleMouseEvent): Use MotionEvent instead of GdkEvent.
(webkitWebViewBaseMotionNotifyEvent): Pass the event coordinates to the pointer lock manager since that's the
only thing needed.
(webkitWebViewBaseRequestPointerLock): Use MotionEvent instead of GdkEvent.
(webkitWebViewBaseDidLosePointerLock): Remove GTK4 ifdefs.
(webkitWebViewBaseSynthesizeMouseEvent): Forward the event to the pointer lock manager if present. Also keep
track of the motion event and set the movement delta.
* UIProcess/gtk/PointerLockManager.cpp:
(WebKit::PointerLockManager::create): Receive event information as parameters instead of using GdkEvent.
(WebKit::PointerLockManager::PointerLockManager): Ditto.
(WebKit::PointerLockManager::~PointerLockManager):
(WebKit::PointerLockManager::lock): Remove the grab.
(WebKit::PointerLockManager::unlock): Remove the ungrab.
(WebKit::PointerLockManager::handleMotion): Use NativeWebMouseEvent constructor taking all information as parameters.
* UIProcess/gtk/PointerLockManager.h:
* UIProcess/gtk/PointerLockManagerWayland.cpp:
(WebKit::PointerLockManagerWayland::PointerLockManagerWayland): Receive event information as parameters instead of using GdkEvent.
(WebKit::PointerLockManagerWayland::lock): Hide the cursor.
(WebKit::PointerLockManagerWayland::unlock): Unhide the cursor.
* UIProcess/gtk/PointerLockManagerWayland.h:
* UIProcess/gtk/PointerLockManagerX11.cpp:
(WebKit::PointerLockManagerX11::PointerLockManagerX11): Receive event information as parameters instead of using GdkEvent.
(WebKit::PointerLockManagerX11::lock): Use XGrabPointer.
(WebKit::PointerLockManagerX11::unlock): Use XUngrabPointer.
(WebKit::PointerLockManagerX11::didReceiveMotionEvent): Use XWarpPointer.
* UIProcess/gtk/PointerLockManagerX11.h:
2020-05-27 Philippe Normand <pnormand@igalia.com>
Unreviewed, build warning fix after r261986
* UIProcess/API/glib/WebKitSettings.cpp: A colon is needed after the transfer annotation.
2020-05-27 Peng Liu <peng.liu6@apple.com>
VideoFullscreenInterfaceAVKit is leaking when a video element enters and exits fullscreen/picture-in-picture
https://bugs.webkit.org/show_bug.cgi?id=212293
VideoFullscreenManagerProxy::ensureInterface() makes sure a fullscreen interface object exists (an object
will be created if it does not exit). That means an extra object will be created by an IPC message from
the Web process after the original video fullscreen interface object has been destroyed (thats happens
when a video element is returning to inline from fullscreen or picture-in-picture).
Reviewed by Youenn Fablet.
* UIProcess/Cocoa/VideoFullscreenManagerProxy.mm:
(WebKit::VideoFullscreenManagerProxy::setHasVideo):
Don't call VideoFullscreenInterface[AVKit|Mac]::hasVideoChanged() before the instance of
VideoFullscreenInterface[AVKit|Mac] is ready.
(WebKit::VideoFullscreenManagerProxy::setVideoDimensions):
Don't call VideoFullscreenInterface[AVKit|Mac]::videoDimensionsChanged() after the instance of
VideoFullscreenInterface[AVKit|Mac] is destroyed.
2020-05-27 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK][WTR] EventSender: stop using GdkEvent API in preparation for GTK4
https://bugs.webkit.org/show_bug.cgi?id=212298
Reviewed by Adrian Perez de Castro.
Add internal API to WebKitWebViewBase to synthesize events, because with GTK4 it's not possible to
create events. In the case of layout tests, the web view is always the target of the events, so we don't really
need to send the events to GTK to process them.
* Shared/NativeWebKeyboardEvent.h: Add new constructor that doesn't receive a GdkEvent.
* Shared/NativeWebMouseEvent.h: Ditto.
* Shared/NativeWebWheelEvent.h: Ditto.
* Shared/gtk/NativeWebKeyboardEventGtk.cpp:
(WebKit::NativeWebKeyboardEvent::NativeWebKeyboardEvent): Create without a GdkEvent.
* Shared/gtk/NativeWebMouseEventGtk.cpp:
(WebKit::NativeWebMouseEvent::NativeWebMouseEvent): Ditto.
* Shared/gtk/NativeWebWheelEventGtk.cpp:
(WebKit::NativeWebWheelEvent::NativeWebWheelEvent): Ditto.
* UIProcess/API/gtk/PageClientImpl.cpp:
(WebKit::PageClientImpl::doneWithKeyEvent): Return early if event doesn't have a GdkEvent.
(WebKit::PageClientImpl::doneWithTouchEvent): Ditto.
(WebKit::PageClientImpl::wheelEventWasNotHandledByWebCore): Ditto.
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(toWebKitModifiers):
(webkitWebViewBaseSynthesizeMouseEvent):
(webkitWebViewBaseSynthesizeKeyEvent):
(toWebKitWheelEventPhase):
(webkitWebViewBaseSynthesizeWheelEvent):
* UIProcess/API/gtk/WebKitWebViewBaseInternal.h: Added.
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::activePopupMenu const): Return the active popup menu proxy.
* UIProcess/gtk/KeyBindingTranslator.cpp:
(WebKit::handleKeyBindingsForMap): Make it generic by receiving the map as parameter.
(WebKit::handleCustomKeyBindings): Call handleKeyBindingsForMap() passing the custom key bindings map.
(WebKit::KeyBindingTranslator::commandsForKeyval): Get commands for the predefined key bindings. This is used
only when synthesizing key events since we don't have a GdkEVent to forward to the GtkTextView.
* UIProcess/gtk/KeyBindingTranslator.h:
* UIProcess/gtk/WebContextMenuProxyGtk.cpp:
(WebKit::popupMenuWidget): Take a ref of the GdkWindow when creating the GdkEvent, since the event will unref it
on destroy.
* UIProcess/gtk/WebPopupMenuProxyGtk.cpp:
(WebKit::WebPopupMenuProxyGtk::handleKeyPress): Handle a key press event from its keyval and timestamp.
(WebKit::WebPopupMenuProxyGtk::activateSelectedItem): Activate the currently selected item.
(WebKit::WebPopupMenuProxyGtk::keyPressEventCallback): Call handleKeyPress().
(WebKit::WebPopupMenuProxyGtk::typeAheadFindIndex): Receive a keyval and timestamp instead of a GdkEvent.
(WebKit::WebPopupMenuProxyGtk::typeAheadFind): Ditto.
* UIProcess/gtk/WebPopupMenuProxyGtk.h:
2020-05-27 Keith Rollin <krollin@apple.com>
Enable the use of XCBuild by default in Apple builds
https://bugs.webkit.org/show_bug.cgi?id=209890
<rdar://problem/44182078>
Unreviewed build fix. Check the value of XCODE_VERSION_ACTUAL rather
than XCODE_VERSION_MAJOR when determining whether to use the XCBuild
or non-XCBuild method of running header post-processing scripts.
No new tests -- this is a build fix.
* WebKit.xcodeproj/project.pbxproj:
2020-05-26 Zan Dobersek <zdobersek@igalia.com>
Unreviewed WPE build fix after r262155.
* UIProcess/API/wpe/WebKitColor.cpp:
(webkit_color_parse): Use CSSParser::parseColor() to do the color parsing.
2020-05-26 Fujii Hironori <Hironori.Fujii@sony.com>
[WinCairo] Unreviewed clang-cl build fix after r262158
https://bugs.webkit.org/show_bug.cgi?id=211883
std::unique_ptr<WebCore::TextureMapper> can't delete an incomplete type TextureMapper.
* WebProcess/WebPage/CoordinatedGraphics/LayerTreeHostTextureMapper.cpp:
* WebProcess/WebPage/CoordinatedGraphics/LayerTreeHostTextureMapper.h:
2020-05-26 Andy Estes <aestes@apple.com>
Fix the engineering build on internal SDKs
* Shared/ApplePay/cocoa/WebPaymentCoordinatorProxyCocoa.mm:
(WebKit::WebPaymentCoordinatorProxy::platformPaymentRequest): Silenced a deprecation warning.
2020-05-26 Fujii Hironori <Hironori.Fujii@sony.com>
[WinCairo][WK2] Use GraphicsLayerTextureMapper for Accelerated Compositing mode
https://bugs.webkit.org/show_bug.cgi?id=211883
Reviewed by Don Olmstead.
GraphicsLayerTextureMapper is a simple GraphicsLayer
implementation using TextureMapper which is used for WinCairo WebKit1.
Add a new LayerTreeHost implementation for
GraphicsLayerTextureMapper by copying WinCairo WebKit1's
AcceleratedCompositingContext.
Windows WebKitTestRunner can't capture the window content for
pixel dump tests yet.
* PlatformWin.cmake:
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
* UIProcess/win/WebView.cpp:
(WebKit::WebView::WebView):
* WebProcess/WebPage/CoordinatedGraphics/DrawingAreaCoordinatedGraphics.cpp:
(WebKit::DrawingAreaCoordinatedGraphics::setNeedsDisplayInRect):
(WebKit::DrawingAreaCoordinatedGraphics::enterAcceleratedCompositingMode):
* WebProcess/WebPage/CoordinatedGraphics/LayerTreeHost.h:
* WebProcess/WebPage/CoordinatedGraphics/LayerTreeHostTextureMapper.cpp: Added.
(WebKit::LayerTreeHost::prepareForRendering):
(WebKit::LayerTreeHost::compositeLayersToContext):
(WebKit::LayerTreeHost::flushPendingLayerChanges):
(WebKit::LayerTreeHost::layerFlushTimerFired):
(WebKit::LayerTreeHost::LayerTreeHost):
(WebKit::LayerTreeHost::setLayerFlushSchedulingEnabled):
(WebKit::LayerTreeHost::setShouldNotifyAfterNextScheduledLayerFlush):
(WebKit::LayerTreeHost::scheduleLayerFlush):
(WebKit::LayerTreeHost::cancelPendingLayerFlush):
(WebKit::LayerTreeHost::setRootCompositingLayer):
(WebKit::LayerTreeHost::setViewOverlayRootLayer):
(WebKit::LayerTreeHost::setNonCompositedContentsNeedDisplay):
(WebKit::LayerTreeHost::scrollNonCompositedContents):
(WebKit::LayerTreeHost::flushAndRenderLayers):
(WebKit::LayerTreeHost::forceRepaint):
(WebKit::LayerTreeHost::forceRepaintAsync):
(WebKit::LayerTreeHost::sizeDidChange):
(WebKit::LayerTreeHost::pauseRendering):
(WebKit::LayerTreeHost::resumeRendering):
(WebKit::LayerTreeHost::graphicsLayerFactory):
(WebKit::LayerTreeHost::contentsSizeChanged):
(WebKit::LayerTreeHost::didChangeViewportAttributes):
(WebKit::LayerTreeHost::setIsDiscardable):
(WebKit::LayerTreeHost::deviceOrPageScaleFactorChanged):
(WebKit::LayerTreeHost::createDisplayRefreshMonitor):
(WebKit::LayerTreeHost::window):
(WebKit::LayerTreeHost::enabled):
(WebKit::LayerTreeHost::paintContents):
(WebKit::LayerTreeHost::deviceScaleFactor const):
(WebKit::LayerTreeHost::applyDeviceScaleFactor):
* WebProcess/WebPage/CoordinatedGraphics/LayerTreeHostTextureMapper.h: Added.
(WebKit::LayerTreeHost::layerTreeContext const):
(WebKit::LayerTreeHost::displayID const):
* WebProcess/WebPage/WebPage.cpp:
* WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::nativeWindowHandle):
* WebProcess/win/WebProcessMainWin.cpp:
2020-05-25 Darin Adler <darin@apple.com>
Eliminate Color constructors that take strings, moving color parsing entirely into the CSS parser
https://bugs.webkit.org/show_bug.cgi?id=212296
Reviewed by Sam Weinig.
* WebProcess/InjectedBundle/API/c/WKBundlePage.cpp:
(WKBundlePageSetComposition): Use WebCore::CSSParser::parseColor.
2020-05-26 Peng Liu <peng.liu6@apple.com>
ASSERTION FAILED: m_clientCounts.contains(contextId) - WebKit::VideoFullscreenManagerProxy::removeClientForContext()
https://bugs.webkit.org/show_bug.cgi?id=212308
Refactor the implementation of the callback hasVideoInPictureInPictureDidChange().
Reviewed by Jer Noble.
* UIProcess/Cocoa/VideoFullscreenManagerProxy.h:
* UIProcess/Cocoa/VideoFullscreenManagerProxy.mm:
(WebKit::VideoFullscreenModelContext::didEnterPictureInPicture):
(WebKit::VideoFullscreenModelContext::didExitPictureInPicture):
(WebKit::VideoFullscreenManagerProxy::hasVideoInPictureInPictureDidChange):
Add this function to wrap m_page->uiClient().hasVideoInPictureInPictureDidChange().
(WebKit::VideoFullscreenManagerProxy::setupFullscreenWithID):
Remove the call to hasVideoInPictureInPictureDidChange() because it is too early to do so.
(WebKit::VideoFullscreenManagerProxy::exitFullscreenWithoutAnimationToMode):
Add a call to hasVideoInPictureInPictureDidChange() on Mac as well. This fixes the timeout of
TestWebKitAPI.PictureInPicture.ExitPiPOnSuspendVideoElement on Mac.
(WebKit::VideoFullscreenManagerProxy::fullscreenModeChanged):
Remove the call to hasVideoInPictureInPictureDidChange() because it is not necessary.
2020-05-26 Keith Rollin <krollin@apple.com>
Enable the use of XCBuild by default in Apple builds
https://bugs.webkit.org/show_bug.cgi?id=209890
<rdar://problem/44182078>
Reviewed by Darin Adler.
Switch from the "legacy" Xcode build system to the "new" build system
(also known as "XCBuild"). Switching to the new system speeds up
builds by a small percentage, better validates projects for
build-related issues (such as dependency cycles), lets WebKit benefit
from future improvements in XCBuild such as those coming from the
underlying llbuild open source project, and prepares us for any other
tools built for this new ecosystem.
Specific changes:
- Remove Xcode project and workspace settings that selected the Build
system, allowing the default to take hold (which is currently the
New build system).
- Updated webkitdirs.pm with a terser check for Xcode version.
- Update build-webkit and Makefile.shared to be explicit when using
the old build system (no longer treat it as a default or fall-back
configuration).
- Update various xcconfig files similarly to treat the default as
using the new build system.
- Update various post-processing build steps to check for Xcode 11.4
and to no longer treat the default as using the old build system.
* Configurations/WebKit.xcconfig:
* WebKit.xcodeproj/project.pbxproj:
2020-05-26 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Fix context menu
https://bugs.webkit.org/show_bug.cgi?id=212344
Reviewed by Adrian Perez de Castro.
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseSizeAllocate): Call gtk_native_check_resize() on the popover to make it appear.
* UIProcess/gtk/WebContextMenuProxyGtk.cpp:
(WebKit::destroyMenuWidget): There's no gtk_widget_destroy() in GTK4 so we can just unparent it.
(WebKit::bindModelToMenuWidget): Do not set the action namespace because that's not possible in GTK4.
(WebKit::WebContextMenuProxyGtk::append): Use group.name again for the action name passed to
g_menu_item_set_action_and_target_value().
(WebKit::WebContextMenuProxyGtk::~WebContextMenuProxyGtk): Use destroyMenuWidget().
2020-05-25 Alex Christensen <achristensen@webkit.org>
Expose more network metrics to WebCoreNSURLSession
https://bugs.webkit.org/show_bug.cgi?id=212359
<rdar://problem/62909440>
Reviewed by Darin Adler.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(-[WKNetworkSessionDelegate URLSession:task:didFinishCollectingMetrics:]):
2020-05-24 Sam Weinig <weinig@apple.com>
Extended Color Cleanup: Use the name SimpleColor consistently
https://bugs.webkit.org/show_bug.cgi?id=212337
Reviewed by Anders Carlsson.
* UIProcess/ios/WKContentViewInteraction.mm:
(compositionHighlights):
Update for new name of CompositionHighlight::defaultCompositionFillColor.
2020-05-24 Rob Buis <rbuis@igalia.com>
Remove NetworkProcess/FileAPI
https://bugs.webkit.org/show_bug.cgi?id=212335
Reviewed by Sam Weinig.
Remove NetworkProcess/FileAPI, the only file it contains is unused.
It seems this could have been removed in r248593.
* NetworkProcess/FileAPI/NetworkBlobRegistry.cpp: Removed.
2020-05-24 Sam Weinig <weinig@apple.com>
Extended Color Cleanup: Stop allowing direct access to the underlying SimpleColor (it is almost always incorrect with respect to extended colors)
https://bugs.webkit.org/show_bug.cgi?id=212184
Reviewed by Dean Jackson.
* WebProcess/WebPage/RemoteLayerTree/PlatformCAAnimationRemote.mm:
(WebKit::animationValueFromKeyframeValue):
Use toSRGBASimpleColorLossy() to get access to color components.
2020-05-24 Carlos Garcia Campos <cgarcia@igalia.com>
Unreviewed. Remove unused file since r261570
I forgot to remove DragAndDropHandler.h.
* UIProcess/API/gtk/WebKitWebViewBasePrivate.h:
* UIProcess/gtk/DragAndDropHandler.h: Removed.
2020-05-23 Charlie Turner <cturner@igalia.com>
Error sending IPC message: Broken pipe
https://bugs.webkit.org/show_bug.cgi?id=206052
Reviewed by Adrian Perez de Castro.
The Connection implementation on *nix was not handling socket
shutdowns gracefully, this was leading to many IPC error messages
being printed. It looks like the "errors" were actually
application issues, trying to send messages during socket shutdown
sequences, rather than more serious message drop problems.
Unclear how to autotest this, but manual testing opening lots of tabs
definitely hits these codepaths.
* Platform/IPC/unix/ConnectionUnix.cpp:
(IPC::Connection::readyReadHandler): Handle ECONNRESET by cleanly
closing the connection.
(IPC::Connection::sendOutputMessage): Ditto, also handle the
wrinkle that for "reasons", Linux can return EPIPE when it means
ECONNRESET.
2020-05-22 James Savage <james.savage@apple.com>
Create Swift overlay for WebKit framework
https://bugs.webkit.org/show_bug.cgi?id=211270
<rdar://problem/60549174>
Reviewed by Brady Eidson.
Provides Swift-only overlay methods for our NS_REFINED_FOR_SWIFT API.
The overlay does not provide different API than its Objective-C counterparts,
so overlay testing focuses on ensuring that the overlay produces the
expected interface, and provides the expected mappings to and from ObjC.
The overlay is defined in a separate Xcode project (from WebKit) by
recommendation from the Swift team. Due to build system constraints this
project cannot be a sibbling of WebKit.xcodeproj, and must live in a
subdirectory. However, because it generally acts as part of the WebKit
framework, it still uses the same Version.xcconfig and source file
locations as standard WebKit sources.
* SwiftOverlay/Configurations/Base.xcconfig: Added. Set some default
warnings and product configuration.
* SwiftOverlay/Configurations/WebKitSwiftOverlay-maccatalyst.xcconfig: Added.
Override SDK configurations to build for Catalyst.
* SwiftOverlay/Configurations/WebKitSwiftOverlay.xcconfig: Added.
These flags were all provided by the Xcode template.
* SwiftOverlay/Configurations/WebKitSwiftOverlayTests-maccatalyst.xcconfig: Added.
See above.
* SwiftOverlay/Configurations/WebKitSwiftOverlayTests.xcconfig: Added.
See above.
* SwiftOverlay/SwiftOverlay/install-swiftmodules.sh: Added. Provided by the
Swift team.
* SwiftOverlay/Tests/WebKitSwiftOverlayTests-Info.plist: Added.
* SwiftOverlay/Tests/WebKitTests.swift: Added.
(WebKitTests.testAPI): This ensures that the function names come through as we
expect them to. It also ensures that the functions were correctly set to
public visiblity.
(WebKitTests.testWKPDFConfigurationRect): This does the same as the above, but
also exercises the null <-> nil conversion, and that regular values are
correctly round tripped to Objective-C.
* SwiftOverlay/WebKitSwiftOverlay.xcodeproj/project.pbxproj: Added.
* SwiftOverlay/WebKitSwiftOverlay.xcodeproj/xcshareddata/xcschemes/Everything (Catalyst).xcscheme: Added.
* SwiftOverlay/WebKitSwiftOverlay.xcodeproj/xcshareddata/xcschemes/Everything.xcscheme: Added.
* UIProcess/API/Cocoa/WebKitSwiftOverlay.swift: Added.
(WKPDFConfiguration.rect): Provide mapping of CGRect.null to nil for Swift.
(WKWebView.callAsyncJavaScript(_:arguments:in:completion:)): Provide a Result
in the completion handler, and default value for arguments.
(WKWebView.evaluateJavaScript(_:in:completion:)): Ditto.
(WKWebView.createPDF(configuration:completionHandler:)): Provide a Swift function
that takes a Result<Data, Error> instead of (Data?, Error?) tuple. Provide
a default value for the configuration.
(WKWebView.createWebArchiveData(completionHandler:)): See above.
(WKWebView.find(_:configuration:completionHandler:)): See above.
2020-05-22 Chris Dumez <cdumez@apple.com>
RELEASE_ASSERT() that InitializeWebKit2() is getting called on the main thread
https://bugs.webkit.org/show_bug.cgi?id=212283
Reviewed by Alex Christensen.
RELEASE_ASSERT() that InitializeWebKit2() is getting called on the main thread, behind
a linked-on-after check. In r217137, we tried to add a dispatch_sync() to the main thread to
work around the issue but it was a hack and it can cause deadlocks. It is best to force
developers to fix their code by crashing.
* Shared/Cocoa/WebKit2InitializeCocoa.mm:
(WebKit::runInitializationCode):
(WebKit::InitializeWebKit2):
* UIProcess/Cocoa/VersionChecks.h:
2020-05-22 Alex Christensen <achristensen@webkit.org>
Make download resume workaround forgiving of changes in CFNetwork
https://bugs.webkit.org/show_bug.cgi?id=212273
Reviewed by Geoffrey Garen.
They broke my workaround. See rdar://problem/63512518
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(-[WKNetworkSessionDelegate URLSession:downloadTask:didResumeAtOffset:expectedTotalBytes:]):
2020-05-22 Wenson Hsieh <wenson_hsieh@apple.com>
[IPC] Add support for specifying `Async WantsConnection` in message files
https://bugs.webkit.org/show_bug.cgi?id=212276
Reviewed by Alex Christensen.
Augments the IPC message receiver generation script to allow for "Async WantsConnection" in `.message.in` files.
Currently, specifying this in a message causes the `connection` argument to be passed twice when handling the
IPC message. This is because normal async IPC messages without replies normally don't have the `IPC::Connection`
argument, and use the overloaded `handleMessage(Connection& connection, ...)` version of `handleMessage` when
`WantsConnection` is specified.
However, in the `Async` reply case, we already pass in the `IPC::Connection`. Instead of overloading the method
signature, we introduce a different method instead, named `handleMessageAsyncWantsConnection`, which forwards
the given `IPC::Connection` along to the member function.
Test: TestAsyncMessageWithConnection
* Platform/IPC/HandleMessage.h:
(IPC::handleMessageAsyncWantsConnection):
Add another variant of the message receiver template, for the case where the message receiver wants a connection.
This is similar to handleMessageSynchronousWantsConnection, above.
* Scripts/test-superclassMessageReceiver.cpp:
(Messages::WebPage::TestAsyncMessageWithConnection::callReply):
(Messages::WebPage::TestAsyncMessageWithConnection::cancelReply):
(Messages::WebPage::TestAsyncMessageWithConnection::send):
(WebKit::WebPage::didReceiveMessage):
* Scripts/test-superclassMessages.h:
(Messages::WebPage::TestAsyncMessageWithConnection::name):
(Messages::WebPage::TestAsyncMessageWithConnection::asyncMessageReplyName):
(Messages::WebPage::TestAsyncMessageWithConnection::TestAsyncMessageWithConnection):
(Messages::WebPage::TestAsyncMessageWithConnection::arguments const):
* Scripts/webkit/messages.py:
* Scripts/webkit/messages_unittest.py:
* Scripts/webkit/test-superclass.messages.in:
2020-05-22 Tim Horton <timothy_horton@apple.com>
Excessive hang time in iOS Safari under waitForDidUpdateActivityState
https://bugs.webkit.org/show_bug.cgi?id=212272
<rdar://problem/62787789>
Reviewed by Andy Estes.
* UIProcess/RemoteLayerTree/RemoteLayerTreeDrawingAreaProxy.mm:
(WebKit::RemoteLayerTreeDrawingAreaProxy::waitForDidUpdateActivityState):
Make the view state change timeout equivalent on iOS and macOS.
We believe that it is preferable to paint white than block the UI
process main thread for half a second. Reduce the timeout on iOS
to be the same as it has been for years on macOS.
2020-05-22 Alex Christensen <achristensen@webkit.org>
Add SPI to unblock third party cookies from WKWebViews with ResourceLoadStatistics turned on
https://bugs.webkit.org/show_bug.cgi?id=212058
<rdar://problem/60595539>
Reviewed by John Wilander.
This value makes NetworkDataTaskCocoa::blockCookies() not do anything for certain WKWebViews.
* NetworkProcess/NetworkDataTask.cpp:
(WebKit::NetworkDataTask::create):
* NetworkProcess/NetworkLoadParameters.h:
* NetworkProcess/NetworkResourceLoadParameters.cpp:
(WebKit::NetworkResourceLoadParameters::encode const):
(WebKit::NetworkResourceLoadParameters::decode):
* NetworkProcess/cocoa/NetworkDataTaskCocoa.h:
* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::NetworkDataTaskCocoa::blockCookies):
(WebKit::NetworkDataTaskCocoa::unblockCookies):
(WebKit::NetworkDataTaskCocoa::NetworkDataTaskCocoa):
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
* UIProcess/API/APIPageConfiguration.cpp:
(API::PageConfiguration::copy const):
* UIProcess/API/APIPageConfiguration.h:
(API::PageConfiguration::setAllowAllThirdPartyCookies):
(API::PageConfiguration::allowAllThirdPartyCookies const):
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration _allowAllThirdPartyCookies]):
(-[WKWebViewConfiguration _setAllowAllThirdPartyCookies:]):
* UIProcess/API/Cocoa/WKWebViewConfigurationPrivate.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
* WebProcess/Network/WebLoaderStrategy.cpp:
(WebKit::addParametersShared):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
2020-05-22 Chris Dumez <cdumez@apple.com>
Regression(r254859) DOM storage event gets fired at the frame that caused the storage modification
https://bugs.webkit.org/show_bug.cgi?id=211503
<rdar://problem/62983284>
Reviewed by Maciej Stachowiak.
r254859 refactored StorageAreaMap's dispatchSessionStorageEvent() &
dispatchLocalStorageEvent() to share more code by moving that code to
a new framesForEventDispatching() static function. However,
framesForEventDispatching() was always using the session storage no
matter the call site. It should be using the local storage when called
from dispatchLocalStorageEvent().
Test: storage/domstorage/events/storage-event-not-in-originator.html
* WebProcess/WebStorage/StorageAreaMap.cpp:
(WebKit::framesForEventDispatching):
(WebKit::StorageAreaMap::dispatchSessionStorageEvent):
(WebKit::StorageAreaMap::dispatchLocalStorageEvent):
2020-05-22 Tim Horton <timothy_horton@apple.com>
REGRESSION (r261978): Cannot click on links with trackpad on iPad
https://bugs.webkit.org/show_bug.cgi?id=212262
Reviewed by Andy Estes.
No new tests; our current WKMouseGestureRecognizer testing strategy
unfortunately bypasses this particular problem.
* UIProcess/ios/WKMouseGestureRecognizer.mm:
(-[WKMouseGestureRecognizer _hoverEntered:withEvent:]):
(-[WKMouseGestureRecognizer _hoverMoved:withEvent:]):
(-[WKMouseGestureRecognizer _hoverExited:withEvent:]):
(-[WKMouseGestureRecognizer _hoverCancelled:withEvent:]):
Don't call super in any of the hover callbacks now that we derive from
UIHoverGestureRecognizer. We want to wholly own the state of the gesture,
including differences from UIHoverGestureRecognizer behavior such as
remaining in the recognizing state while the trackpad button is pressed.
We don't need any of the code in the default implementation, so just
override it entirely. This returns behavior to what it was like before r261978.
2020-05-21 Sam Weinig <weinig@apple.com>
Extended Color Cleanup: Move Color coder definitions to Color to allow for future encaspulation improvements
https://bugs.webkit.org/show_bug.cgi?id=212247
Reviewed by Simon Fraser.
Move IPC encoder/decoder definitions from WebKit down into Color itself to move closer
to making Color::rgb() private.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<Color>::encode): Deleted.
(IPC::ArgumentCoder<Color>::decode): Deleted.
* Shared/WebCoreArgumentCoders.h:
2020-05-21 Peng Liu <peng.liu6@apple.com>
Fix issues of the Picture-in-Picture API under stress tests
https://bugs.webkit.org/show_bug.cgi?id=212191
Reviewed by Eric Carlson.
* UIProcess/Cocoa/VideoFullscreenManagerProxy.mm:
(WebKit::VideoFullscreenManagerProxy::enterFullscreen):
Mock the behavior of the VideoFullscreenInterface[AVKit|Mac] regarding
the Picture-in-Picture mode support.
* WebProcess/cocoa/VideoFullscreenManager.mm:
(WebKit::VideoFullscreenManager::didEnterFullscreen):
(WebKit::VideoFullscreenManager::didCleanupFullscreen):
Add callbacks for the end of entering/exiting fullscreen.
2020-05-21 Tyler Wilcock <twilco.o@protonmail.com>
Fix misspelling -- m_releaseNetwrokActivityTimer --> m_releaseNetworkActivityTimer
https://bugs.webkit.org/show_bug.cgi?id=212237
Reviewed by Simon Fraser.
No new tests required -- simply fixing a misspelling.
* UIProcess/Cocoa/NavigationState.h: Rename m_releaseNetwrokActivityTimer to m_releaseNetworkActivityTimer
* UIProcess/Cocoa/NavigationState.mm: Rename m_releaseNetwrokActivityTimer to m_releaseNetworkActivityTimer
(WebKit::NavigationState::NavigationState):
(WebKit::NavigationState::releaseNetworkActivity):
(WebKit::NavigationState::didChangeIsLoading):
2020-05-21 Sam Weinig <weinig@apple.com>
Extended Color Cleanup: Remove trivial uses of Color::rgb()
https://bugs.webkit.org/show_bug.cgi?id=212231
Reviewed by Darin Adler.
* UIProcess/API/ios/WKWebViewIOS.mm:
(scrollViewBackgroundColor):
Replace colorWithOverrideAlpha() with Color::colorWithAlphaUsingAlternativeRounding() to avoid
unnecessary use of Color::rgb()
2020-05-21 Alex Christensen <achristensen@webkit.org>
Use an OptionSet instead of uint8_t for MessageFlags
https://bugs.webkit.org/show_bug.cgi?id=212230
<rdar://problem/63496543>
Reviewed by Darin Adler.
I also reduce the encoding size of OptionSet.
* Platform/IPC/ArgumentCoders.h:
(IPC::ArgumentCoder<OptionSet<T>>::encode):
(IPC::ArgumentCoder<OptionSet<T>>::decode):
* Platform/IPC/Decoder.cpp:
(IPC::Decoder::isSyncMessage const):
(IPC::Decoder::shouldDispatchMessageWhenWaitingForSyncReply const):
(IPC::Decoder::shouldUseFullySynchronousModeForTesting const):
* Platform/IPC/Decoder.h:
* Platform/IPC/Encoder.cpp:
(IPC::Encoder::isSyncMessage const):
(IPC::Encoder::shouldDispatchMessageWhenWaitingForSyncReply const):
(IPC::Encoder::setIsSyncMessage):
(IPC::Encoder::setShouldDispatchMessageWhenWaitingForSyncReply):
(IPC::Encoder::setFullySynchronousModeForTesting):
(IPC::Encoder::messageFlags):
(IPC::Encoder::messageFlags const):
* Platform/IPC/Encoder.h:
* Platform/IPC/MessageFlags.h:
(): Deleted.
2020-05-21 John Wilander <wilander@apple.com>
Storage Access API: Allow configurable storage access scope
https://bugs.webkit.org/show_bug.cgi?id=212114
<rdar://problem/63423063>
Reviewed by Alex Christensen.
The scope of storage access as per-frame or per-page was discussed in the
standards process here: https://github.com/privacycg/storage-access/issues/3
The decision was to have per-page storage access by default. Recent feedback
from Google and conversation with Mozilla suggest that we might want to
support the caller choosing the scope.
This patch adds support for different scope configurations while keeping the
existing default as per-frame. A later patch will switch the default and add
test cases for per-page scope.
Most of the changes is piping through the new enum
WebCore::StorageAccessScope from the call side to encode per-frame and
per-page access, and piping through the new struct
WebCore::RequestStorageAccessResult on the return side which carries full
information about the storage access request result.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::CompletionHandler<void):
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:
(WebKit::ResourceLoadStatisticsMemoryStore::requestStorageAccess):
(WebKit::ResourceLoadStatisticsMemoryStore::requestStorageAccessUnderOpener):
(WebKit::ResourceLoadStatisticsMemoryStore::grantStorageAccess):
(WebKit::ResourceLoadStatisticsMemoryStore::grantStorageAccessInternal):
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.h:
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::requestStorageAccess):
(WebKit::WebResourceLoadStatisticsStore::requestStorageAccessEphemeral):
(WebKit::WebResourceLoadStatisticsStore::grantStorageAccess):
(WebKit::WebResourceLoadStatisticsStore::grantStorageAccessEphemeral):
(WebKit::WebResourceLoadStatisticsStore::grantStorageAccessInStorageSession):
Renamed from WebResourceLoadStatisticsStore::grantStorageAccess()
to reduce confusion since we already have another function called
WebResourceLoadStatisticsStore::grantStorageAccess().
(WebKit::WebResourceLoadStatisticsStore::callGrantStorageAccessHandler):
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.h:
* NetworkProcess/NetworkConnectionToWebProcess.cpp:
(WebKit::NetworkConnectionToWebProcess::requestStorageAccess):
* NetworkProcess/NetworkConnectionToWebProcess.h:
* NetworkProcess/NetworkConnectionToWebProcess.messages.in:
* Scripts/webkit/messages.py:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::requestStorageAccess):
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::requestStorageAccess):
* WebProcess/WebPage/WebPage.h:
2020-05-21 Yoshiaki Jitsukawa <yoshiaki.jitsukawa@sony.com>
[PlayStation] Add minimal WKView API to enable TestWebKitAPI
https://bugs.webkit.org/show_bug.cgi?id=211868
Reviewed by Alex Christensen.
Enable TestWebKitAPI
* Platform/IPC/Connection.h:
* Platform/IPC/unix/ConnectionUnix.cpp:
(IPC::Connection::platformInvalidate):
(IPC::Connection::open):
Add SocketMonitor to monitor received messages.
* UIProcess/API/APIProcessPoolConfiguration.cpp:
(API::ProcessPoolConfiguration::copy):
* UIProcess/API/APIProcessPoolConfiguration.h:
Add PlayStation specific members: WebProcess and NetworkProcess paths, and user ID.
* UIProcess/Launcher/ProcessLauncher.h:
Add PlayStation specific members: Process path and user ID.
* NetworkProcess/EntryPoint/playstation/NetworkProcessMain.cpp: Copied from Tools\TestWebKitAPI\playstation\main.cpp.
(loadLibraryOrExit):
(main):
* PlatformPlayStation.cmake:
* Shared/API/c/playstation/WKEventPlayStation.cpp: Added.
(WKKeyboardEventMake):
(WKMouseEventMake):
(WKWheelEventMake):
* Shared/API/c/playstation/WKEventPlayStation.h: Added.
* UIProcess/API/C/playstation/WKAPICastPlayStation.h:
(WebKit::toAPI<>):
(WebKit::toImpl<>):
* UIProcess/API/C/playstation/WKContextConfigurationPlayStation.cpp: Copied from Tools\TestWebKitAPI\playstation\PlatformUtilitiesPlayStation.cpp.
(WKContextConfigurationSetWebProcessPath):
(WKContextConfigurationCopyWebProcessPath):
(WKContextConfigurationSetNetworkProcessPath):
(WKContextConfigurationCopyNetworkProcessPath):
(WKContextConfigurationSetUserId):
(WKContextConfigurationGetUserId):
* UIProcess/API/C/playstation/WKContextConfigurationPlayStation.h: Copied from Source\WebKit\UIProcess\API\C\playstation\WKView.h.
* UIProcess/API/C/playstation/WKPagePrivatePlayStation.cpp: Added.
(WKPageHandleKeyboardEvent):
(WKPageHandleMouseEvent):
(WKPageHandleWheelEvent):
* UIProcess/API/C/playstation/WKPagePrivatePlayStation.h: Copied from Source\WebKit\UIProcess\API\C\playstation\WKView.h.
* UIProcess/API/C/playstation/WKView.cpp: Copied from Source\WebKit\WebProcess\InjectedBundle\playstation\InjectedBundlePlayStation.cpp.
(WKViewCreate):
(WKViewGetPage):
* UIProcess/API/C/playstation/WKView.h:
* UIProcess/Launcher/playstation/ProcessLauncherPlayStation.cpp:
(WebKit::defaultProcessPath):
(WebKit::ProcessLauncher::launchProcess):
(WebKit::ProcessLauncher::terminateProcess):
(WebKit::ProcessLauncher::platformInvalidate):
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::getLaunchOptions):
* UIProcess/WebProcessPool.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::getLaunchOptions):
* UIProcess/playstation/PageClientImpl.cpp: Added.
(WebKit::PageClientImpl::PageClientImpl):
(WebKit::PageClientImpl::createDrawingAreaProxy):
(WebKit::PageClientImpl::setViewNeedsDisplay):
(WebKit::PageClientImpl::requestScroll):
(WebKit::PageClientImpl::viewScrollPosition):
(WebKit::PageClientImpl::viewSize):
(WebKit::PageClientImpl::isViewWindowActive):
(WebKit::PageClientImpl::isViewFocused):
(WebKit::PageClientImpl::isViewVisible):
(WebKit::PageClientImpl::isViewInWindow):
(WebKit::PageClientImpl::processDidExit):
(WebKit::PageClientImpl::didRelaunchProcess):
(WebKit::PageClientImpl::pageClosed):
(WebKit::PageClientImpl::preferencesDidChange):
(WebKit::PageClientImpl::toolTipChanged):
(WebKit::PageClientImpl::didCommitLoadForMainFrame):
(WebKit::PageClientImpl::handleDownloadRequest):
(WebKit::PageClientImpl::didChangeContentSize):
(WebKit::PageClientImpl::setCursor):
(WebKit::PageClientImpl::setCursorHiddenUntilMouseMoves):
(WebKit::PageClientImpl::didChangeViewportProperties):
(WebKit::PageClientImpl::registerEditCommand):
(WebKit::PageClientImpl::clearAllEditCommands):
(WebKit::PageClientImpl::canUndoRedo):
(WebKit::PageClientImpl::executeUndoRedo):
(WebKit::PageClientImpl::wheelEventWasNotHandledByWebCore):
(WebKit::PageClientImpl::convertToDeviceSpace):
(WebKit::PageClientImpl::convertToUserSpace):
(WebKit::PageClientImpl::screenToRootView):
(WebKit::PageClientImpl::rootViewToScreen):
(WebKit::PageClientImpl::accessibilityScreenToRootView):
(WebKit::PageClientImpl::rootViewToAccessibilityScreen):
(WebKit::PageClientImpl::doneWithKeyEvent):
(WebKit::PageClientImpl::createPopupMenuProxy):
(WebKit::PageClientImpl::enterAcceleratedCompositingMode):
(WebKit::PageClientImpl::exitAcceleratedCompositingMode):
(WebKit::PageClientImpl::updateAcceleratedCompositingMode):
(WebKit::PageClientImpl::fullScreenManagerProxyClient):
(WebKit::PageClientImpl::didFinishLoadingDataForCustomContentProvider):
(WebKit::PageClientImpl::navigationGestureDidBegin):
(WebKit::PageClientImpl::navigationGestureWillEnd):
(WebKit::PageClientImpl::navigationGestureDidEnd):
(WebKit::PageClientImpl::willRecordNavigationSnapshot):
(WebKit::PageClientImpl::didRemoveNavigationGestureSnapshot):
(WebKit::PageClientImpl::didFirstVisuallyNonEmptyLayoutForMainFrame):
(WebKit::PageClientImpl::didFinishLoadForMainFrame):
(WebKit::PageClientImpl::didFailLoadForMainFrame):
(WebKit::PageClientImpl::didSameDocumentNavigationForMainFrame):
(WebKit::PageClientImpl::didChangeBackgroundColor):
(WebKit::PageClientImpl::isPlayingAudioWillChange):
(WebKit::PageClientImpl::isPlayingAudioDidChange):
(WebKit::PageClientImpl::refView):
(WebKit::PageClientImpl::derefView):
(WebKit::PageClientImpl::didRestoreScrollPosition):
(WebKit::PageClientImpl::userInterfaceLayoutDirection):
(WebKit::PageClientImpl::requestDOMPasteAccess):
* UIProcess/playstation/PageClientImpl.h: Added.
* UIProcess/playstation/WebProcessPoolPlayStation.cpp:
(WebKit::WebProcessPool::platformInitialize):
* UIProcess/playstation/PlayStationWebView.cpp: Copied from Source\WebKit\UIProcess\playstation\WebProcessPoolPlayStation.cpp.
(WebKit::PlayStationWebView::create):
(WebKit::PlayStationWebView::PlayStationWebView):
(WebKit::PlayStationWebView::~PlayStationWebView):
(WebKit::PlayStationWebView::isActive const):
(WebKit::PlayStationWebView::isFocused const):
(WebKit::PlayStationWebView::isVisible const):
* UIProcess/playstation/PlayStationWebView.h: Copied from Source\WebKit\WebProcess\InjectedBundle\playstation\InjectedBundlePlayStation.cpp.
(WebKit::PlayStationWebView::page):
* WebProcess/EntryPoint/playstation/WebProcessMain.cpp: Copied from Tools\TestWebKitAPI\playstation\main.cpp.
(loadLibraryOrExit):
(main):
* WebProcess/InjectedBundle/playstation/InjectedBundlePlayStation.cpp:
(WebKit::InjectedBundle::initialize):
(WebKit::InjectedBundle::setBundleParameter):
* WebProcess/WebPage/AcceleratedSurface.cpp:
(WebKit::AcceleratedSurface::create):
2020-05-21 Chris Dumez <cdumez@apple.com>
Address post review comments from Mitz after r262004.
* Shared/mac/AuxiliaryProcessMac.mm:
(WebKit::sandboxDataVaultParentDirectory):
2020-05-21 Chris Dumez <cdumez@apple.com>
[WK2] WebKit abandons compiled sandbox profiles
https://bugs.webkit.org/show_bug.cgi?id=212177
<rdar://problem/54613619>
Reviewed by Saam Barati.
WebKit generates new compiled sandbox profiles whenever sandbox profiles are updated or
whenever webkit directories change (which is super common for WebKitTestRunner). Previous
compiled sandbox profiles do not get deleted and therefore may accumulate.
To address the issue, we now save compiled sandbox profiles under _CS_DARWIN_USER_TEMP_DIR
instead of _CS_DARWIN_USER_CACHE_DIR. The system takes care of deleting the files under
_CS_DARWIN_USER_TEMP_DIR upon reboot or if they have not been accessed for 3 days.
* Shared/mac/AuxiliaryProcessMac.mm:
(WebKit::sandboxDataVaultParentDirectory):
2020-05-21 Wenson Hsieh <wenson_hsieh@apple.com>
Unreviewed, fix the iOS 13.4 build after r261978
* Platform/spi/ios/UIKitSPI.h:
Only declare `_UIHoverEventRespondable` if HAVE(UI_HOVER_EVENT_RESPONDABLE) is defined.
* UIProcess/ios/WKMouseGestureRecognizer.mm:
If `_UIHoverEventRespondable` is not present, fall back to declaring internal methods on
`UIHoverGestureRecognizer`.
2020-05-21 Yoshiaki JITSUKAWA <yoshiaki.jitsukawa@sony.com>
[ConnectionUnix] Set MSG_NOSIGNAL to avoid SIGPIPE
https://bugs.webkit.org/show_bug.cgi?id=211719
Reviewed by Don Olmstead.
Set MSG_NOSIGNAL to sendmsg()/resvmsg() in order to avoid
SIGPIPE when the other process disappears.
* Platform/IPC/unix/ConnectionUnix.cpp:
(IPC::readBytesFromSocket):
(IPC::Connection::sendOutputMessage):
2020-05-21 Enrique Ocaña González <eocanha@igalia.com>
[GStreamer][GTK][WPE] Expose and honor the media content types requiring hardware support setting
https://bugs.webkit.org/show_bug.cgi?id=211950
Reviewed by Adrian Perez de Castro.
Expose mediaContentTypesRequiringHardwareSupport as a setting on WPE and WebKitGTK.
* UIProcess/API/glib/WebKitSettings.cpp:
(_WebKitSettingsPrivate::_WebKitSettingsPrivate):
(webKitSettingsSetProperty):
(webKitSettingsGetProperty):
(webkit_settings_class_init):
(webkit_settings_get_media_content_types_requiring_hardware_support):
(webkit_settings_set_media_content_types_requiring_hardware_support):
* UIProcess/API/gtk/WebKitSettings.h:
* UIProcess/API/gtk/docs/webkit2gtk-4.0-sections.txt:
* UIProcess/API/gtk/docs/webkit2gtk-docs.sgml:
* UIProcess/API/wpe/WebKitSettings.h:
* UIProcess/API/wpe/docs/wpe-1.0-sections.txt:
2020-05-20 Darin Adler <darin@apple.com>
Dictation context should be an object identifier, not a type-punned pointer
https://bugs.webkit.org/show_bug.cgi?id=212174
Reviewed by Anders Carlsson.
* Scripts/webkit/messages.py: Added DictationContext to the list of types that are
not C++ classes.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<DictationAlternative>::encode): Updated for change to
DictationAlternative structure.
(IPC::ArgumentCoder<DictationAlternative>::decode): Ditto.
* UIProcess/Cocoa/PageClientImplCocoa.h: Use DictationContext instead of uint64_t.
* UIProcess/Cocoa/PageClientImplCocoa.mm:
(WebKit::PageClientImplCocoa::addDictationAlternatives): Ditto.
(WebKit::PageClientImplCocoa::removeDictationAlternatives): Ditto.
(WebKit::PageClientImplCocoa::dictationAlternatives): Ditto.
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::insertDictatedTextAsync): Ditto.
* UIProcess/PageClient.h: Ditto.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::showDictationAlternativeUI): Ditto.
(WebKit::WebPageProxy::removeDictationAlternatives): Ditto.
(WebKit::WebPageProxy::dictationAlternatives): Ditto.
* UIProcess/WebPageProxy.h: Ditto.
* UIProcess/WebPageProxy.messages.in: Ditto.
* UIProcess/ios/PageClientImplIOS.h: Ditto.
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::showDictationAlternativeUI): Ditto.
* UIProcess/mac/PageClientImplMac.h: Ditto.
* UIProcess/mac/PageClientImplMac.mm:
(WebKit::PageClientImpl::showDictationAlternativeUI): Ditto.
* UIProcess/mac/WebPageProxyMac.mm: Removed unneeded include of
DictationAlternative.h.
* WebProcess/WebCoreSupport/WebAlternativeTextClient.h: Use DictationContext
instead of uint64_t. Also marked class final and made everything private.
* WebProcess/WebCoreSupport/mac/WebAlternativeTextClient.cpp:
(WebKit::WebAlternativeTextClient::removeDictationAlternatives): Ditto.
(WebKit::WebAlternativeTextClient::showDictationAlternativeUI): Ditto.
(WebKit::WebAlternativeTextClient::dictationAlternatives): Ditto.
2020-05-20 Tim Horton <timothy_horton@apple.com>
WKMouseGestureRecognizer should be implemented without using UIKit internals
https://bugs.webkit.org/show_bug.cgi?id=212173
<rdar://problem/61163726>
Reviewed by Wenson Hsieh.
* Platform/spi/ios/UIKitSPI.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView gestureRecognizer:shouldReceiveTouch:]):
* UIProcess/ios/WKMouseGestureRecognizer.h:
* UIProcess/ios/WKMouseGestureRecognizer.mm:
(-[WKMouseGestureRecognizer _hoverEntered:withEvent:]):
(-[WKMouseGestureRecognizer initWithTarget:action:]): Deleted.
(-[WKMouseGestureRecognizer setView:]): Deleted.
(-[WKMouseGestureRecognizer _wantsHoverEvents]): Deleted.
(-[WKMouseGestureRecognizer canPreventGestureRecognizer:]): Deleted.
(-[WKMouseGestureRecognizer canBePreventedByGestureRecognizer:]): Deleted.
Reimplement WKMouseGestureRecognizer as a UIHoverGestureRecognizer subclass,
so we can decouple ourselves a bit from UIKit internals.
2020-05-20 Per Arne Vollan <pvollan@apple.com>
[iOS] Fix message filter sandbox violation
https://bugs.webkit.org/show_bug.cgi?id=212100
<rdar://problem/63407155>
Reviewed by Brent Fulgham.
A message needs to be added to the message filter in the WebContent sandbox on iOS.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-05-20 Sam Weinig <weinig@apple.com>
Replace Color::getHSL() with sRGBToHSL to ensure it at least gives somewhat sensible results for ExtendedColors and reduce code duplication
https://bugs.webkit.org/show_bug.cgi?id=212143
Reviewed by Simon Fraser.
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _updateScrollViewBackground]):
Switch to using Color::lightness().
2020-05-20 Megan Gardner <megan_gardner@apple.com>
Hide password echo when screen is being captured.
https://bugs.webkit.org/show_bug.cgi?id=212060
<rdar://problem/47653578>
Reviewed by Wenson Hsieh.
Use the UIScreen state and notification to determine if the
screen is being shared or captured. This flag is for all
capture methods, Air Play, Screen Recording, etc. If the screen is
being captured, turn off the password echo to prevent the password from
being leaked.
* UIProcess/WebPageProxy.h:
* UIProcess/ios/WKContentView.mm:
(-[WKContentView _commonInitializationWithProcessPool:configuration:]):
(-[WKContentView _screenCapturedDidChange:]):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::setIsScreenCaptured):
* WebProcess/WebCoreSupport/WebEditorClient.h:
* WebProcess/WebCoreSupport/ios/WebEditorClientIOS.mm:
(WebKit::WebEditorClient::isScreenCaptured const):
* WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::isScreenCaptured const):
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::setIsScreenCaptured):
2020-05-20 ChangSeok Oh <changseok@webkit.org>
[GTK] Implement connected and disconnected events of GAMEPAD API with libmanette
https://bugs.webkit.org/show_bug.cgi?id=133854
Reviewed by Carlos Garcia Campos.
This patch brings initial GAMEPAD API support to the gtk port. We use libmanette,
a simple GObject game controller library to handle gamepad connection and input.
* SourcesGTK.txt:
* UIProcess/Gamepad/UIGamepadProvider.cpp: Add a build flag to void default interface
* UIProcess/Gamepad/gtk/UIGamepadProviderGtk.cpp: Added to retrieve WebPageProxy to which
a gamepad connects. The returned WebPageProxy handles input events from the gamepad.
(WebKit::getWebPageProxy):
(WebKit::UIGamepadProvider::platformWebPageProxyForGamepadInput):
* UIProcess/Gamepad/manette/UIGamepadProviderManette.cpp: Added.
(WebKit::UIGamepadProvider::platformSetDefaultGamepadProvider):
(WebKit::UIGamepadProvider::platformWebPageProxyForGamepadInput):
(WebKit::UIGamepadProvider::platformStopMonitoringInput):
(WebKit::UIGamepadProvider::platformStartMonitoringInput):
2020-05-20 Kate Cheney <katherine_cheney@apple.com>
Support operating dates in ResourceLoadStatisticsDatabaseStore
https://bugs.webkit.org/show_bug.cgi?id=212034
<rdar://problem/63349242>
Reviewed by Brent Fulgham.
Tests: http/tests/resourceLoadStatistics/operating-dates-all-but-cookies-removed-database.html
http/tests/resourceLoadStatistics/operating-dates-all-but-cookies-removed.html
http/tests/resourceLoadStatistics/operating-dates-all-website-data-removed-database.html
http/tests/resourceLoadStatistics/operating-dates-all-website-data-removed.html
Patch to support Operating Dates in the ResourceLoadStatisticsDatabaseStore.
Moves functions out of ResourceLoadStatisticsStore now that each store
needs its own implementation.
Adds a new function for testing which inserts an expired statistic
and makes sure the proper behavior occurs depending on additional
parameters.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::createTableQueries):
(WebKit::ResourceLoadStatisticsDatabaseStore::ResourceLoadStatisticsDatabaseStore):
(WebKit::ResourceLoadStatisticsDatabaseStore::checkForMissingTablesInSchema):
(WebKit::ResourceLoadStatisticsDatabaseStore::createUniqueIndices):
(WebKit::ResourceLoadStatisticsDatabaseStore::createSchema):
(WebKit::ResourceLoadStatisticsDatabaseStore::getLeastRecentOperatingDate):
(WebKit::ResourceLoadStatisticsDatabaseStore::mergeOperatingDates):
(WebKit::ResourceLoadStatisticsDatabaseStore::includeTodayAsOperatingDateIfNecessary):
(WebKit::ResourceLoadStatisticsDatabaseStore::hasStatisticsExpired const):
(WebKit::ResourceLoadStatisticsDatabaseStore::insertExpiredStatisticForTesting):
Migrated Operating Dates functions from ResourceLoadStatisticsStore and implemented
them using SQLite instead of a vector. Functionality is
identical to the original in-memory implementation.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:
(WebKit::ResourceLoadStatisticsMemoryStore::ResourceLoadStatisticsMemoryStore):
(WebKit::ResourceLoadStatisticsMemoryStore::mergeOperatingDates):
(WebKit::ResourceLoadStatisticsMemoryStore::includeTodayAsOperatingDateIfNecessary):
(WebKit::ResourceLoadStatisticsMemoryStore::hasStatisticsExpired const):
(WebKit::ResourceLoadStatisticsMemoryStore::insertExpiredStatisticForTesting):
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.h:
Migrated Operating Dates functions from ResourceLoadStatisticsStore
now that the database store does not use them.
* NetworkProcess/Classifier/ResourceLoadStatisticsPersistentStorage.cpp:
(WebKit::ResourceLoadStatisticsPersistentStorage::populateMemoryStoreFromDisk):
Spelling error.
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.cpp:
(WebKit::ResourceLoadStatisticsStore::ResourceLoadStatisticsStore):
(WebKit::ResourceLoadStatisticsStore::updateCookieBlockingForDomains):
(WebKit::ResourceLoadStatisticsStore::statisticsEpirationTime const): Deleted.
(WebKit::ResourceLoadStatisticsStore::mergeOperatingDates): Deleted.
(WebKit::ResourceLoadStatisticsStore::includeTodayAsOperatingDateIfNecessary): Deleted.
(WebKit::ResourceLoadStatisticsStore::hasStatisticsExpired const): Deleted.
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.h:
(WebKit::OperatingDate::year const):
(WebKit::OperatingDate::month const):
(WebKit::OperatingDate::monthDay const):
These functions should be public now that they are needed outside
of this file after migrating OperatingDates functions.
(WebKit::ResourceLoadStatisticsStore::parameters const):
(WebKit::ResourceLoadStatisticsStore::operatingDates const): Deleted.
(WebKit::ResourceLoadStatisticsStore::clearOperatingDates): Deleted.
Moved OperatingDates functions to the memory store. Moved
operatingDatesWindow constexpr variables to the header file so that
the database and memory stores can both use them.
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::insertExpiredStatisticForTesting):
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.h:
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::insertExpiredStatisticForTesting):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreSetStatisticsExpiredStatistic):
* UIProcess/API/C/WKWebsiteDataStoreRef.h:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::insertExpiredStatisticForTesting):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::mergeStatisticForTesting):
(WebKit::WebsiteDataStore::insertExpiredStatisticForTesting):
* UIProcess/WebsiteData/WebsiteDataStore.h:
Function to add expired statistic to the ITP store.
2020-05-20 Chris Dumez <cdumez@apple.com>
Deprecate WKBundlePageSetPolicyClient() Injected bundle C API
https://bugs.webkit.org/show_bug.cgi?id=212166
Reviewed by Alex Christensen.
Deprecate WKBundlePageSetPolicyClient() Injected bundle C API. Approving navigations from
the injected bundle can lead to WebProcess terminations like in <rdar://problem/60340647>
because the WebProcess is not in a position to do sandbox checks when navigating to file
URLs.
* WebProcess/InjectedBundle/API/c/WKBundlePage.h:
2020-05-20 Simon Fraser <simon.fraser@apple.com>
Plumb the display's nominal refresh rate down to ScrollingTree for use in scroll synchronization
https://bugs.webkit.org/show_bug.cgi?id=212159
Reviewed by Tim Horton.
Plumb an Optional<unsigned> down windowScreenDidChange, which contains the nominal
display refresh rate (as frames per second) if available. On macOS, we get this
from CVDisplayLinkGetNominalOutputVideoRefreshPeriod().
To read it, WebProcessPool::nominalFramesPerSecondForDisplay() makes a DisplayLink
that doesn't get any observers, but that DisplayLink will very likely get used
as soon as we schedule a rendering update.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::nominalFramesPerSecondForDisplay):
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::windowDidChangeScreen):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::windowScreenDidChange):
* UIProcess/WebPageProxy.h:
* UIProcess/WebProcessPool.h:
* UIProcess/mac/DisplayLink.cpp:
(WebKit::DisplayLink::DisplayLink):
(WebKit::DisplayLink::nominalFramesPerSecond const):
* UIProcess/mac/DisplayLink.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.mm:
(WebKit::RemoteLayerTreeDrawingArea::RemoteLayerTreeDrawingArea):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::windowScreenDidChange):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-05-20 Myles C. Maxfield <mmaxfield@apple.com>
[iPadOS] -webkit-text-size-adjust:percentage doesn't work in native apps
https://bugs.webkit.org/show_bug.cgi?id=212122
<rdar://problem/54560875>
Reviewed by Wenson Hsieh.
* Shared/WebsitePoliciesData.cpp:
(WebKit::WebsitePoliciesData::encode const):
(WebKit::WebsitePoliciesData::decode):
(WebKit::WebsitePoliciesData::applyToDocumentLoader):
* Shared/WebsitePoliciesData.h:
* UIProcess/API/APIWebsitePolicies.cpp:
(API::WebsitePolicies::copy const):
(API::WebsitePolicies::data):
* UIProcess/API/APIWebsitePolicies.h:
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::effectiveContentModeAfterAdjustingPolicies):
2020-05-20 Kate Cheney <katherine_cheney@apple.com>
REGRESSION (r261892) [ Mac Debug ] 4 API tests failing related to ResourceLoadStatistics
https://bugs.webkit.org/show_bug.cgi?id=212153
<rdar://problem/63452541>
Reviewed by Sihui Liu.
Accidentally deleted necessary reset() call for a SQLite statement
when converting stored statements to SQLiteStatementAutoResetScope in
https://trac.webkit.org/changeset/261892/webkit.
This statement is used in a loop with multiple bind() calls so it must be reset.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::checkForMissingTablesInSchema):
2020-05-20 Eric Carlson <eric.carlson@apple.com>
Update some media logging
https://bugs.webkit.org/show_bug.cgi?id=212109
<rdar://problem/63424816>
Unreviewed, remove a change unintentionally included in r261899.
* UIProcess/ios/fullscreen/WKFullScreenWindowControllerIOS.mm:
(-[WKFullScreenWindowController _interactivePinchDismissChanged:]):
2020-05-20 Andy Estes <aestes@apple.com>
[Mac] UI processes spin when creating the "Share" context menu item
https://bugs.webkit.org/show_bug.cgi?id=212137
<rdar://problem/54498394>
Reviewed by Wenson Hsieh.
On Mac, WebKit has been using the +[NSMenuItem standardShareMenuItemForItems:] SPI to create
the "Share" context submenu and menu item. This call performs synchronous IPC and can result
in UI process spins.
Where available, switch to using the asynchronous
-[NSSharingServicePicker getMenuWithCompletion:] SPI instead. This method only creates the
submenu, not the menu item, so when using this new method we must create the "Share" menu
item ourselves.
* UIProcess/mac/WebContextMenuProxyMac.h:
* UIProcess/mac/WebContextMenuProxyMac.mm:
(WebKit::getStandardShareMenuItem):
(WebKit::WebContextMenuProxyMac::getShareMenuItem):
(WebKit::WebContextMenuProxyMac::getContextMenuFromItems):
(WebKit::WebContextMenuProxyMac::getContextMenuItem):
(WebKit::WebContextMenuProxyMac::showContextMenuWithItems):
(WebKit::WebContextMenuProxyMac::createShareMenuItem): Deleted.
(WebKit::WebContextMenuProxyMac::createContextMenuFromItems): Deleted.
(WebKit::WebContextMenuProxyMac::createContextMenuItem): Deleted.
2020-05-20 Carlos Garcia Campos <cgarcia@igalia.com>
Unreviewed. Fix GTK4 build with GTK 3.98.4
* PlatformGTK.cmake:
* UIProcess/API/gtk/WebKitPrintOperation.cpp:
(webkitPrintOperationRunDialog):
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseAddDialog):
(webkitWebViewBaseRemoveChild):
(webkitWebViewBaseContainerForall):
(webkitWebViewBaseAddWebInspector):
(webkitWebViewBaseRemoveWebInspector):
(webkitWebViewBaseDispose):
(webkitWebViewBaseSizeAllocate):
(webkitWebViewBaseConstructed):
(webkit_web_view_base_class_init):
* UIProcess/API/gtk/WebKitWebViewBase.h: Removed.
* UIProcess/API/gtk/WebKitWebViewBasePrivate.h:
* UIProcess/Inspector/gtk/RemoteWebInspectorProxyGtk.cpp:
(WebKit::RemoteWebInspectorProxy::platformCreateFrontendPageAndWindow):
(WebKit::RemoteWebInspectorProxy::platformSave):
* UIProcess/Inspector/gtk/WebInspectorProxyGtk.cpp:
(WebKit::WebInspectorProxy::platformCreateFrontendWindow):
(WebKit::WebInspectorProxy::platformAttach):
(WebKit::WebInspectorProxy::platformDetach):
(WebKit::WebInspectorProxy::platformSave):
* UIProcess/gtk/WebContextMenuProxyGtk.cpp:
(WebKit::WebContextMenuProxyGtk::~WebContextMenuProxyGtk):
2020-05-12 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Add support for drag and drop operations
https://bugs.webkit.org/show_bug.cgi?id=211779
Reviewed by Adrian Perez de Castro.
* UIProcess/API/gtk/DragSource.h:
* UIProcess/API/gtk/DragSourceGtk4.cpp:
(WebKit::DragSource::begin):
* UIProcess/API/gtk/DropTarget.h:
* UIProcess/API/gtk/DropTargetGtk4.cpp:
(WebKit::DropTarget::DropTarget):
(WebKit::DropTarget::~DropTarget):
(WebKit::DropTarget::accept):
(WebKit::DropReadAsyncData::DropReadAsyncData):
(WebKit::DropTarget::loadData):
(WebKit::DropTarget::didLoadData):
(WebKit::DropTarget::enter):
(WebKit::DropTarget::update):
(WebKit::DropTarget::didPerformAction):
(WebKit::DropTarget::leave):
(WebKit::DropTarget::drop):
2020-05-19 Wenson Hsieh <wenson_hsieh@apple.com>
REGRESSION (r258919): fast/events/ios/key-events-comprehensive/key-events-control tests are failing
https://bugs.webkit.org/show_bug.cgi?id=212124
<rdar://problem/60893744>
Reviewed by Tim Horton.
After the changes in r258919, WKContentView's text input traits report that we're in a single line input, even
when there is no editable focused element. This prevents UIKit from allowing WebKit to handle the `CTRL + M`
key command, which is interpreted at the system level as inserting a line break in a multi-line editable area.
It seems odd to consider ourselves in a single line input when there isn't even an editable element focused
element; for now, we can fix these failing tests by just going with the default value for the text input traits'
`setIsSingleLineDocument` property, but not explicitly setting it to either `NO` or `YES` when no element is
focused.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView textInputTraits]):
2020-05-19 Sam Weinig <weinig@apple.com>
Remove almost always incorrect Color::getRGBA
https://bugs.webkit.org/show_bug.cgi?id=212059
Reviewed by Darin Adler and Simon Fraser.
* UIProcess/API/wpe/WebKitColor.cpp:
(webkitColorFillFromWebCoreColor):
* UIProcess/gtk/ViewGestureControllerGtk.cpp:
(WebKit::ViewGestureController::beginSwipeGesture):
* WebProcess/WebPage/WebFrame.cpp:
(WebKit::WebFrame::getDocumentBackgroundColor):
Replace Color::getRGBA with Color::toSRGBAComponentsLossy()
2020-05-19 Simon Fraser <simon.fraser@apple.com>
Use an ObjectIdentifier<> for DisplayLink observer IDs
https://bugs.webkit.org/show_bug.cgi?id=212112
Reviewed by Per Arne Vollan.
Replace unsigned observerID with DisplayLinkObserverID, which is a typesafe ObjectIdentifier<>.
Use PlatformDisplayID in more places.
* Scripts/webkit/messages.py:
* Shared/DisplayLinkObserverID.h: Copied from Source/WebKit/UIProcess/mac/DisplayLink.h.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::startDisplayLink):
(WebKit::WebProcessPool::stopDisplayLink):
* UIProcess/WebProcessPool.h:
* UIProcess/WebProcessProxy.h:
* UIProcess/WebProcessProxy.messages.in:
* UIProcess/mac/DisplayLink.cpp:
(WebKit::DisplayLink::addObserver):
(WebKit::DisplayLink::removeObserver):
* UIProcess/mac/DisplayLink.h:
* UIProcess/mac/WebProcessProxyMac.mm:
(WebKit::WebProcessProxy::startDisplayLink):
(WebKit::WebProcessProxy::stopDisplayLink):
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebPage/mac/DrawingAreaMac.cpp:
(WebKit::DisplayRefreshMonitorMac::DisplayRefreshMonitorMac):
2020-05-19 Alex Christensen <achristensen@webkit.org>
Add _WKDownloadDelegate callback including totalBytesWritten
https://bugs.webkit.org/show_bug.cgi?id=212110
<rdar://problem/63358981>
Reviewed by Geoffrey Garen.
Without this new callback, after resuming a download, a client has no way to tell whether the download was successfully
resumed by a server that has proper etag and range request support or whether the download began at the beginning again.
A client was guessing that the download did not restart, causing incorrect reported download sizes when the download did restart.
Luckily, the data on disk was not corrupted, just the UI. This allows us to fix the UI.
Testing covered by expanding the API test for resuming downloads.
* NetworkProcess/Downloads/Download.cpp:
(WebKit::Download::didReceiveData):
* NetworkProcess/Downloads/Download.h:
* NetworkProcess/NetworkDataTaskBlob.cpp:
(WebKit::NetworkDataTaskBlob::writeDownload):
* NetworkProcess/NetworkDataTaskBlob.h:
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(-[WKNetworkSessionDelegate URLSession:downloadTask:didWriteData:totalBytesWritten:totalBytesExpectedToWrite:]):
* UIProcess/API/APIDownloadClient.h:
(API::DownloadClient::didReceiveData):
* UIProcess/API/C/WKContext.cpp:
(WKContextSetDownloadClient):
* UIProcess/API/Cocoa/_WKDownloadDelegate.h:
* UIProcess/Cocoa/DownloadClient.h:
* UIProcess/Cocoa/DownloadClient.mm:
(WebKit::DownloadClient::DownloadClient):
(WebKit::DownloadClient::didReceiveResponse):
(WebKit::DownloadClient::didReceiveData):
* UIProcess/Downloads/DownloadProxy.cpp:
(WebKit::DownloadProxy::didReceiveData):
* UIProcess/Downloads/DownloadProxy.h:
(WebKit::DownloadProxy::expectedContentLength const): Deleted.
(WebKit::DownloadProxy::setExpectedContentLength): Deleted.
(WebKit::DownloadProxy::bytesLoaded const): Deleted.
(WebKit::DownloadProxy::setBytesLoaded): Deleted.
* UIProcess/Downloads/DownloadProxy.messages.in:
2020-05-19 Kate Cheney <katherine_cheney@apple.com>
ITP database should finalize all prepared statements before closing
https://bugs.webkit.org/show_bug.cgi?id=211929
<rdar://problem/63246945>
Reviewed by John Wilander.
No new tests, behavior confirmed by existing tests.
Convert all SQLite statement objects to be unique pointers to better
manage the lifetime of each statement object. Only initialize and
prepare an object when it is needed instead of preparing all objects
at once. Set each statement to null before closing the database
because ~SQLiteStatement() will finalize the statement.
This patch removes all reset() commands now that we initialize
each statement using SQLiteStatementAutoResetScope.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::ResourceLoadStatisticsDatabaseStore):
(WebKit::ResourceLoadStatisticsDatabaseStore::close):
(WebKit::ResourceLoadStatisticsDatabaseStore::getScopedStatement const):
Function to initialize and prepare a statement.
(WebKit::ResourceLoadStatisticsDatabaseStore::insertDomainRelationship):
This function is not used and can be deleted.
(WebKit::ResourceLoadStatisticsDatabaseStore::checkForMissingTablesInSchema):
(WebKit::ResourceLoadStatisticsDatabaseStore::isMigrationNecessary):
(WebKit::ResourceLoadStatisticsDatabaseStore::openAndUpdateSchemaIfNecessary):
(WebKit::ResourceLoadStatisticsDatabaseStore::isEmpty const):
(WebKit::ResourceLoadStatisticsDatabaseStore::destroyStatements):
(WebKit::ResourceLoadStatisticsDatabaseStore::insertObservedDomain):
(WebKit::ResourceLoadStatisticsDatabaseStore::relationshipExists const):
(WebKit::ResourceLoadStatisticsDatabaseStore::domainID const):
(WebKit::ResourceLoadStatisticsDatabaseStore::merge):
(WebKit::ResourceLoadStatisticsDatabaseStore::mergeStatistic):
(WebKit::CompletionHandler<void):
(WebKit::ResourceLoadStatisticsDatabaseStore::prepareStatements): Deleted.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.h:
Updated all statement variables to end with "Statement" to follow the
established pattern.
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::flushAndDestroyPersistentStore):
2020-05-19 Per Arne Vollan <pvollan@apple.com>
[iOS] Fix message filter sandbox violation
https://bugs.webkit.org/show_bug.cgi?id=212100
<rdar://problem/63407155>
Reviewed by Andy Estes.
A message needs to be added to the message filter in the WebContent sandbox on iOS.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-05-19 Kate Cheney <katherine_cheney@apple.com>
Revert WebKitTestRunner to reset ITP to the memory store between all tests.
https://bugs.webkit.org/show_bug.cgi?id=212094
<rdar://problem/63405806>
Reviewed by Chris Dumez.
Changing this in https://bugs.webkit.org/show_bug.cgi?id=211637 caused
test flakiness.
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreStatisticsResetToConsistentState):
2020-05-19 Simon Fraser <simon.fraser@apple.com>
Push a PlatformDisplayID to scrolling trees, and allow the scrolling thread to get displayDidRefresh notifications
https://bugs.webkit.org/show_bug.cgi?id=211034
Reviewed by Sam Weinig.
Prep work for webkit.org/b210884.
Have EventDispatcher call displayDidRefresh() on each scrolling tree it knows about. It calls this
on the EventDispatcher thread; the scrolling tree is responsible for bouncing that to the scrolling thread.
* WebProcess/WebPage/EventDispatcher.cpp:
(WebKit::EventDispatcher::displayDidRefreshOnScrollingThread):
(WebKit::EventDispatcher::displayWasRefreshed):
* WebProcess/WebPage/EventDispatcher.h:
2020-05-19 Wenson Hsieh <wenson_hsieh@apple.com>
[macOS] Drag and drop within a contenteditable area duplicates attachment-backed images
https://bugs.webkit.org/show_bug.cgi?id=212075
<rdar://problem/62434146>
Reviewed by Andy Estes.
When the attachment API is enabled, starting a drag on an attachment-element-backed image within a
contenteditable area sometimes causes the ranged selection to collapse. This is because `WebViewImpl::startDrag`
has a separate codepath for starting a drag with attachment metadata that doesn't invoke `NSView`'s deprecated
`-dragImage:at:offset:event:pasteboar:source:slideBack:` method, but instead uses
`-beginDraggingSessionWithItems:event:source:`. While the former spins the runloop until the end of the drag
session, the latter does not and returns immediately after beginning the drag. As a result,
`WebPageProxy::didStartDrag()` is actually invoked after the drag ends when dragging non-attachment content,
whereas `WebPageProxy::didStartDrag()` is invoked immediately after beginning the drag when starting a drag in
an attachment or attachment-backed image element.
This call to `didStartDrag()` is used to inform the web process that dragging has begun, and sets the
`m_isStartingDrag` flag on `WebPage` to false. As a result, `m_isStartingDrag` is true over the course of the
drag session when dragging non-attachment elements, but it's only true while starting the drag when dragging
attachments. This flag is consulted in `WebPage::mouseEvent()` to determine whether we should avoid handling the
incoming mouse event when beginning a drag. This prevents these incoming mouse events from triggering selection
changes while dragging (for more details, refer to <http://trac.webkit.org/r176687>).
Now, if we combine this with the fact that mouse events can be queued up in the UI process (refer to
<https://trac.webkit.org/r230817>), it's possible to end up handling these queued mouse events in the web
process after `m_isStartingDrag` has been set back to false when dragging an attachment. This doesn't happen
when dragging non-attachment content, since we don't call `didStartDrag()` until the drag has already ended, so
any queued mouse events are sent to the web process and immediately dropped on the floor since
`m_isStartingDrag` is true.
To address this, we make a couple of minor adjustments:
1. Invoke `didStartDrag()` right before calling into `-dragImage:at:offset:event:pasteboar:source:slideBack:`,
so that we don't wait until the drag ends to set `m_isStartingDrag` to false. This makes the behavior of
both codepaths for starting a drag (attachment and non-attachment) consistent, and also makes it consistent
with iOS.
2. In `didStartDrag()`, discard any queued mouse events to prevent them from being handled by the page (e.g.
causing the selection to change immediately after starting the drag). We already have identical logic to do
this when showing a context menu, so we can factor this out into a private `WebPageProxy` helper and call it
from both places.
Test: WKAttachmentTestsMac.DraggingAttachmentBackedImagePreservesRangedSelection
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::startDrag):
Invoke `didStartDrag()` earlier when starting a drag using the deprecated method, such that we don't wait until
the drag ends to set `m_isStartingDrag` to false in the web process.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::discardQueuedMouseEvents):
Refactor logic to discard all queued mouse events into a separate private helper.
(WebKit::WebPageProxy::didStartDrag):
Use the above helper to empty queued mouse events when starting a drag.
(WebKit::WebPageProxy::showContextMenu):
* UIProcess/WebPageProxy.h:
2020-05-19 Andy Estes <aestes@apple.com>
[Apple Pay] Add testing and logging for ApplePaySetup
https://bugs.webkit.org/show_bug.cgi?id=211972
<rdar://problem/63291965>
Reviewed by Alex Christensen.
* Shared/ApplePay/PaymentSetupConfiguration.mm:
(WebKit::toPlatformConfiguration):
(WebKit::PaymentSetupConfiguration::PaymentSetupConfiguration):
* Shared/ApplePay/PaymentSetupConfigurationWebKit.h:
* WebProcess/ApplePay/WebPaymentCoordinator.cpp:
(WebKit::WebPaymentCoordinator::getSetupFeatures):
(WebKit::WebPaymentCoordinator::beginApplePaySetup):
* WebProcess/ApplePay/WebPaymentCoordinator.h:
2020-05-19 Brent Fulgham <bfulgham@apple.com>
REGRESSION(r243149): createMediaElementSource not working
https://bugs.webkit.org/show_bug.cgi?id=211394
<rdar://problem/62866132>
Reviewed by Per Arne Vollan.
I removed access to some Audio services in r243149, because we believed they were already
blocked by a global deny command. However, Sandbox rules override generic rules with
explicit rules, so the imported sandbox rules were not being blocked.
This patch re-adds the rules needed to support these audio routines on iOS hardware.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::mediaRelatedMachServices):
2020-05-19 Chris Dumez <cdumez@apple.com>
[iOS] Port Foreground process assertion to RunningBoard
https://bugs.webkit.org/show_bug.cgi?id=211795
<rdar://problem/61831863>
Reviewed by Geoffrey Garen.
Port Foreground process assertion to RunningBoard now that <rdar://problem/62614429>
has been fixed.
* Configurations/WebKit.xcconfig:
* UIProcess/ProcessAssertion.h:
* UIProcess/ios/ProcessAssertionIOS.mm:
(WebKit::runningBoardNameForAssertionType):
(WebKit::ProcessAssertion::ProcessAssertion):
(WebKit::ProcessAssertion::~ProcessAssertion):
(WebKit::ProcessAssertion::isValid const):
2020-05-19 Andy Estes <aestes@apple.com>
[Apple Pay] Add testing and logging for ApplePaySetup
https://bugs.webkit.org/show_bug.cgi?id=211972
<rdar://problem/63291965>
Reviewed by Alex Christensen.
* Shared/ApplePay/PaymentSetupConfiguration.mm:
(WebKit::toPlatformConfiguration):
(WebKit::PaymentSetupConfiguration::PaymentSetupConfiguration):
* Shared/ApplePay/PaymentSetupConfigurationWebKit.h:
* WebProcess/ApplePay/WebPaymentCoordinator.cpp:
(WebKit::WebPaymentCoordinator::getSetupFeatures):
(WebKit::WebPaymentCoordinator::beginApplePaySetup):
* WebProcess/ApplePay/WebPaymentCoordinator.h:
2020-05-19 Michael Catanzaro <mcatanzaro@gnome.org>
[GTK] Add WEBKIT_EDITING_COMMAND_PASTE_AS_PLAIN_TEXT
https://bugs.webkit.org/show_bug.cgi?id=212038
Reviewed by Carlos Garcia Campos.
* UIProcess/API/gtk/WebKitEditingCommands.h:
* UIProcess/API/gtk/docs/webkit2gtk-4.0-sections.txt:
* UIProcess/API/wpe/WebKitEditingCommands.h:
* UIProcess/API/wpe/docs/wpe-1.0-sections.txt:
2020-05-18 David Kilzer <ddkilzer@apple.com>
Replace TextIndicatorOptions with OptionSet<TextIndicatorOption>
<https://webkit.org/b/212051>
<rdar://problem/63368556>
Reviewed by Simon Fraser.
Use OptionSet<TextIndicatorOption> everywhere
TextIndicatorOptions was previously used, plus:
- Make TextIndicatorOption an enum class. Remove
"TextIndicatorOption" prefix so TextIndicatorOptionBar becomes
TextIndicatorOption::Bar.
- Remove TextIndicatorOptionDefault because OptionSet<>
initializes to zero.
- Replace static variables (including two globals in WebCore)
with constexpr variables.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<TextIndicatorData>::encode):
(IPC::ArgumentCoder<TextIndicatorData>::decode):
* WebProcess/WebPage/Cocoa/WebPageCocoa.mm:
(WebKit::WebPage::dictionaryPopupInfoForRange):
* WebProcess/WebPage/FindController.cpp:
(WebKit::FindController::updateFindIndicator):
* WebProcess/WebPage/ios/FindControllerIOS.mm:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::insertDroppedImagePlaceholders):
(WebKit::WebPage::computeAndSendEditDragSnapshot):
(WebKit::linkIndicatorPositionInformation):
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::performImmediateActionHitTestAtLocation):
2020-05-18 Andy Estes <aestes@apple.com>
http/tests/ssl/applepay/ApplePayInstallmentConfiguration.https.html fails in public SDK builds
https://bugs.webkit.org/show_bug.cgi?id=212000
<rdar://problem/63323082>
Reviewed by Youenn Fablet.
* Configurations/FeatureDefines.xcconfig:
* Shared/ApplePay/cocoa/WebPaymentCoordinatorProxyCocoa.h:
* Shared/WebCoreArgumentCoders.h:
2020-05-18 Tim Horton <timothy_horton@apple.com>
Rare crash under -[WKContentView resignFirstResponderForWebView]
https://bugs.webkit.org/show_bug.cgi?id=212050
<rdar://problem/60187111>
Reviewed by Wenson Hsieh.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView resignFirstResponderForWebView]):
Reorganize this block so that it always takes a strong reference /before/
null checking and using the pointer. Also, add a _page null check.
This is a speculative fix for an unreproducible low-rate crash.
In theory the existing time-of-check race here should not be a problem,
since WKWebView is supposed to be freed on the main thread, and this code
runs on the main thread, but we have ample evidence of WKWebView
being freed off the main thread in various clients.
2020-05-18 David Kilzer <ddkilzer@apple.com>
Use default initializers in TextIndicatorData
<https://webkit.org/b/212039>
<rdar://problem/63355619>
Reviewed by Alex Christensen.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<TextIndicatorData>::encode):
(IPC::ArgumentCoder<TextIndicatorData>::decode):
- Switch from decodeEnum() and encodeEnum() to modern
equivalents that check for valid enum values.
2020-05-18 Wenson Hsieh <wenson_hsieh@apple.com>
Replace uses of +self with +class
https://bugs.webkit.org/show_bug.cgi?id=212041
Reviewed by Darin Adler.
* Shared/API/Cocoa/_WKFrameHandle.mm:
(-[_WKFrameHandle isEqual:]):
(-[_WKFrameHandle initWithCoder:]):
* Shared/Cocoa/APIObject.mm:
(API::Object::newObject):
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration _setAttachmentFileWrapperClass:]):
* UIProcess/Cocoa/NavigationState.mm:
(WebKit::NavigationState::NavigationClient::decidePolicyForNavigationAction):
* UIProcess/Cocoa/PageClientImplCocoa.mm:
(WebKit::PageClientImplCocoa::allocFileWrapperInstance const):
(WebKit::PageClientImplCocoa::serializableFileWrapperClasses const):
* UIProcess/RemoteLayerTree/ios/RemoteLayerTreeViews.mm:
(+[WKTransformView layerClass]):
(+[WKSimpleBackdropView layerClass]): Deleted.
(+[WKShapeView layerClass]): Deleted.
(-[WKRemoteView initWithFrame:contextID:]): Deleted.
* UIProcess/RemoteLayerTree/ios/ScrollingTreeScrollingNodeDelegateIOS.mm:
(WebKit::ScrollingTreeScrollingNodeDelegateIOS::~ScrollingTreeScrollingNodeDelegateIOS):
(WebKit::ScrollingTreeScrollingNodeDelegateIOS::resetScrollViewDelegate):
(WebKit::ScrollingTreeScrollingNodeDelegateIOS::scrollView const):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView selectFormAccessoryPickerRow:]):
(-[WKContentView selectFormPopoverTitle]):
2020-05-18 Alex Christensen <achristensen@webkit.org>
WebPageProxy::DidCommitLoadForFrame should use FrameLoadType instead of casting to/from uint32_t for IPC
https://bugs.webkit.org/show_bug.cgi?id=212028
<rdar://problem/63343021>
Reviewed by David Kilzer.
* UIProcess/ProvisionalPageProxy.cpp:
(WebKit::ProvisionalPageProxy::didCommitLoadForFrame):
* UIProcess/ProvisionalPageProxy.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::commitProvisionalPage):
(WebKit::WebPageProxy::didCommitLoadForFrame):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::dispatchDidCommitLoad):
2020-05-18 Wenson Hsieh <wenson_hsieh@apple.com>
Single selection <select> with <optgroups> shows multiple selected options
https://bugs.webkit.org/show_bug.cgi?id=199485
<rdar://problem/52757531>
Reviewed by Megan Gardner.
Fixes a long-standing bug in WKMultipleSelectPicker. Prior to this patch, we rely on the delegate method
`-pickerView:row:column:checked:` to be called twice whenever an item is selected: one time for the item that is
no longer checked, and another for the newly checked item. This method is responsible for updating the cached
`FocusedElementInformation` that determines the data model for the select menu, with the expectation that the
unchecked item would be updated to have `isSelected = false;`, and the new checked item would have `isSelected`
`= true;`.
However, `-pickerView:row:column:checked:` is only called for visible item cells. This means that if the user
checks an item, scrolls the select menu items down so that the checked item is offscreen, and then checks a
different item, we only get notified that the new item is checked, and as a result, fail to uncheck the previous
item.
To address this, tweak our logic for handling a single select so that when an item is checked, we additionally
update the previously checked item to not be selected. Also, fix what seems to be a bug in the logic for
updating `_singleSelectionIndex`, which is currently updated even when the item is unchecked. It seems to work
out at the moment, because `-pickerView:row:column:checked:` seems to be called with `checked := YES` after the
previous item was unchecked (assuming that it was visible).
Test: fast/forms/ios/no-stale-checked-items-in-select-picker.html
* UIProcess/API/ios/WKWebViewPrivateForTestingIOS.h:
* UIProcess/API/ios/WKWebViewTestingIOS.mm:
(-[WKWebView selectFormAccessoryHasCheckedItemAtRow:]):
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView selectFormAccessoryHasCheckedItemAtRow:]):
Add plumbing for a new testing hook.
* UIProcess/ios/forms/WKFormSelectControl.h:
* UIProcess/ios/forms/WKFormSelectControl.mm:
(-[WKFormSelectControl selectFormAccessoryHasCheckedItemAtRow:]):
* UIProcess/ios/forms/WKFormSelectPicker.mm:
(-[WKMultipleSelectPicker pickerView:viewForRow:forComponent:reusingView:]):
(-[WKMultipleSelectPicker pickerView:row:column:checked:]):
(-[WKMultipleSelectPicker selectRow:inComponent:extendingSelection:]):
Also, fix an existing bug in this testing helper method that crashed the test runner due to calling an
unimplemented selector. Instead of trying to invoke `-pickerView:didSelectRow:inComponent:`, we should be using
`-pickerView:row:column:checked:` instead for multiple select pickers (which, somewhat confusingly, are still
used for single select elements that have `optgroup`s.)
(-[WKMultipleSelectPicker selectFormAccessoryHasCheckedItemAtRow:]):
2020-05-18 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] Unexpected capitalization of next word after repositioning caret
https://bugs.webkit.org/show_bug.cgi?id=211969
<rdar://problem/62605526>
Reviewed by Alex Christensen.
The changes in r242551 refactored synchronous autocorrection context request logic such that it uses
`waitForAndDispatchImmediately` instead of `sendSync`, in order to make it interruptible by unbounded sync IPC
sent from the web process. If the UI process receives sync IPC, it will immediately cancel the autocorrection
context request (returning an empty context), before proceeding to handle the incoming sync IPC.
In a more recent version of iOS, other changes around spellchecking have caused the synchronous message
`WebPageProxy::checkTextOfParagraph` to be sent from the web process in such a way that it now frequently
coincides with the synchronous autocorrection context request being sent from the UI process. The result is that
we now frequently end up cancelling autocorrection requests early by responding with empty contexts. This
manifests in the keyboard sometimes losing information about its autocapitalization context and believing that
it is in an empty text field, which reverts to default autocorrection suggestions and autocapitalizes the
software keyboard.
To fix this, instead of using the `InterruptWaitingIfSyncMessageArrives` option when waiting for the IPC
response, add and use a new flag that allows us to process an incoming sync IPC message if we're waiting for
the sync message response. We use this new IPC flag when waiting synchronously for HandleAutocorrectionContext.
Test: editing/selection/ios/changing-selection-does-not-trigger-autocapitalization.html
* Platform/IPC/Connection.cpp:
(IPC::Connection::processIncomingMessage):
If the new IPC flag is set and the incoming message is synchronous, allow it to immediately dispatch the sync
message by enqueueing it and then waking up `m_waitForMessageCondition` so that it can process the message.
* Platform/IPC/Connection.h:
* Platform/spi/ios/UIKitSPI.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView requestAutocorrectionContextWithCompletionHandler:]):
Use `DispatchIncomingSyncMessagesWhileWaiting` instead of `InterruptWaitingIfSyncMessageArrives`.
2020-05-18 Lauro Moura <lmoura@igalia.com>
webkitpy: Update test data after r261776
https://bugs.webkit.org/show_bug.cgi?id=212008
Reviewed by Alex Christensen.
Covered by existing tests.
* Scripts/test-legacyMessageReceiver.cpp:
* Scripts/test-legacyMessages.h:
(Messages::WebPage::DidCreateWebProcessConnection::DidCreateWebProcessConnection):
* Scripts/webkit/messages_unittest.py:
* Scripts/webkit/test-legacy.messages.in:
2020-05-18 Carlos Garcia Campos <cgarcia@igalia.com>
Unreviewed. Update OptionsGTK.cmake and NEWS for 2.29.1 release
* gtk/NEWS: Add release notes for 2.29.1.
2020-05-18 Carlos Garcia Campos <cgarcia@igalia.com>
Move iOS files from Source/WebKit/Sources.txt to Source/WebKit/SourcesCocoa.txt
https://bugs.webkit.org/show_bug.cgi?id=212014
Unreviewed. Move platform specific files to the platform sources file.
* Sources.txt:
* SourcesCocoa.txt:
2020-05-18 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] "ASSERTION FAILED: !m_adoptionIsRequired" when double clicking on a word
https://bugs.webkit.org/show_bug.cgi?id=211957
Reviewed by Adrian Perez de Castro.
Remove WebSelectionData wrapper since we can just encode/decode WebCore::SelectionData.
* Shared/WebCoreArgumentCoders.h:
* Shared/gtk/ArgumentCodersGtk.cpp:
(IPC::ArgumentCoder<SelectionData>::decode): Use modern decoder.
* Shared/gtk/ArgumentCodersGtk.h:
* Shared/gtk/WebSelectionData.cpp: Removed.
* Shared/gtk/WebSelectionData.h: Removed.
* SourcesGTK.txt:
* UIProcess/API/gtk/DragSource.h:
* UIProcess/API/gtk/DragSourceGtk3.cpp:
(WebKit::DragSource::DragSource):
(WebKit::DragSource::begin):
* UIProcess/API/gtk/DragSourceGtk4.cpp:
(WebKit::DragSource::begin):
* UIProcess/API/gtk/DropTarget.h:
* UIProcess/API/gtk/DropTargetGtk3.cpp:
(WebKit::DropTarget::accept):
(WebKit::DropTarget::enter):
(WebKit::DropTarget::update):
(WebKit::DropTarget::leaveTimerFired):
(WebKit::DropTarget::drop):
* UIProcess/API/gtk/PageClientImpl.cpp:
(WebKit::PageClientImpl::startDrag):
* UIProcess/API/gtk/PageClientImpl.h:
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseStartDrag):
* UIProcess/API/gtk/WebKitWebViewBasePrivate.h:
* UIProcess/PageClient.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::performDragControllerAction):
(WebKit::WebPageProxy::startDrag):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/WebPasteboardProxy.h:
* UIProcess/WebPasteboardProxy.messages.in:
* UIProcess/gtk/Clipboard.h:
* UIProcess/gtk/ClipboardGtk3.cpp:
(WebKit::WriteAsyncData::WriteAsyncData):
(WebKit::Clipboard::write):
* UIProcess/gtk/ClipboardGtk4.cpp:
(WebKit::Clipboard::write):
* UIProcess/gtk/WebPasteboardProxyGtk.cpp:
(WebKit::WebPasteboardProxy::writeToClipboard):
(WebKit::WebPasteboardProxy::writeCustomData):
* WebProcess/WebCoreSupport/WebPlatformStrategies.cpp:
(WebKit::WebPlatformStrategies::writeToClipboard):
* WebProcess/WebCoreSupport/gtk/WebDragClientGtk.cpp:
(WebKit::WebDragClient::startDrag):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::performDragControllerAction):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-05-18 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] Add WebKitContextMenuItemType for paste as plaintext
https://bugs.webkit.org/show_bug.cgi?id=177638
Reviewed by Michael Catanzaro.
Add WEBKIT_CONTEXT_MENU_ACTION_PASTE_AS_PLAIN_TEXT.
* Shared/API/glib/WebKitContextMenuActions.cpp:
(webkitContextMenuActionGetActionTag):
(webkitContextMenuActionGetForContextMenuItem):
(webkitContextMenuActionGetLabel):
* UIProcess/API/gtk/WebKitContextMenuActions.h:
* UIProcess/gtk/KeyBindingTranslator.cpp:
2020-05-17 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] Move to new Pasteboard API
https://bugs.webkit.org/show_bug.cgi?id=177633
Reviewed by Adrian Perez de Castro.
Enable custom pasteboard data.
* Shared/WebPreferencesDefaultValues.h:
* Shared/gtk/ArgumentCodersGtk.cpp:
(IPC::ArgumentCoder<SelectionData>::encode):
(IPC::ArgumentCoder<SelectionData>::decode):
* UIProcess/WebPasteboardProxy.cpp:
* UIProcess/gtk/ClipboardGtk3.cpp:
(WebKit::Clipboard::write): Add support for writing URI list and custom data.
* UIProcess/gtk/ClipboardGtk4.cpp:
(WebKit::Clipboard::write): Ditto.
* UIProcess/gtk/WebPasteboardProxyGtk.cpp:
(WebKit::WebPasteboardProxy::typesSafeForDOMToReadAndWrite):
(WebKit::WebPasteboardProxy::writeCustomData):
2020-05-16 Andy Estes <aestes@apple.com>
REGRESSION (r260717): installmentConfiguration member is no longer available on ApplePayPaymentRequest
https://bugs.webkit.org/show_bug.cgi?id=211911
<rdar://problem/63236367>
Reviewed by Tim Horton.
* Shared/Cocoa/WebCoreArgumentCodersCocoa.mm:
(IPC::ArgumentCoder<ApplePaySessionPaymentRequest>::encode):
(IPC::ArgumentCoder<ApplePaySessionPaymentRequest>::decode):
(IPC::finishDecoding): Deleted.
(IPC::finishEncoding): Deleted.
2020-05-16 Yury Semikhatsky <yurys@chromium.org>
[GTK] Do not leak pages created by window.open
https://bugs.webkit.org/show_bug.cgi?id=211970
Reviewed by Carlos Garcia Campos.
Previously webkitWebViewCreateNewPage created unbalanced refence to WebPageProxy and returned
it as a raw pointer. That raw pointer was later wrapped in a RefPtr in WebKitUIClient.
This change ensures that all refs are balanced by returnin RefPtr instead of
a raw pointer.
* UIProcess/API/glib/WebKitWebView.cpp:
(webkitWebViewCreateNewPage):
* UIProcess/API/glib/WebKitWebViewPrivate.h:
2020-05-15 Alex Christensen <achristensen@webkit.org>
Use enum serialization instead of casting to/from uint32_t
https://bugs.webkit.org/show_bug.cgi?id=211885
Reviewed by Geoffrey Garen.
* Scripts/webkit/messages.py:
* Shared/DocumentEditingContext.h:
* Shared/ios/GestureTypes.h:
* UIProcess/API/C/WKAPICast.h:
(WebKit::toWebGLLoadPolicy):
(WebKit::toAPI):
* UIProcess/Cocoa/NavigationState.mm:
(WebKit::toWebCoreWebGLLoadPolicy):
(WebKit::NavigationState::NavigationClient::webGLLoadPolicy const):
(WebKit::NavigationState::NavigationClient::resolveWebGLLoadPolicy const):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::webGLPolicyForURL):
(WebKit::WebPageProxy::resolveWebGLPolicyForURL):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView selectForWebView:]):
(toUIWKSelectionFlags):
(toSelectionFlags):
(toWKTextGranularity):
(toWKSelectionDirection):
(selectionChangedWithGesture):
(selectionChangedWithTouch):
(-[WKContentView changeSelectionWithGestureAt:withGesture:withState:withFlags:]):
(-[WKContentView changeSelectionWithTouchAt:withSelectionTouch:baseIsStart:withFlags:]):
(-[WKContentView changeSelectionWithTouchesFrom:to:withGesture:withState:]):
(-[WKContentView selectWordForReplacement]):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::gestureCallback):
(WebKit::WebPageProxy::touchesCallback):
(WebKit::WebPageProxy::selectWithGesture):
(WebKit::WebPageProxy::updateSelectionWithTouches):
(WebKit::WebPageProxy::selectTextWithGranularityAtPoint):
(WebKit::WebPageProxy::selectPositionAtBoundaryWithDirection):
(WebKit::WebPageProxy::moveSelectionAtBoundaryWithDirection):
(WebKit::WebPageProxy::selectWithTwoTouches):
(WebKit::WebPageProxy::extendSelection):
(WebKit::WebPageProxy::requestRectsForGranularityWithSelectionOffset):
* WebProcess/InjectedBundle/API/mac/WKDOMRange.mm:
(-[WKDOMRange rangeByExpandingToWordBoundaryByCharacters:inDirection:]):
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::webGLPolicyForURL const):
(WebKit::WebFrameLoaderClient::resolveWebGLPolicyForURL const):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::webGLPolicyForURL):
(WebKit::WebPage::resolveWebGLPolicyForURL):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::getPlatformEditorState const):
(WebKit::WebPage::getSelectionContext):
(WebKit::WebPage::selectWithGesture):
(WebKit::rangeAtWordBoundaryForPosition):
(WebKit::WebPage::updateSelectionWithTouches):
(WebKit::WebPage::selectWithTwoTouches):
(WebKit::WebPage::extendSelection):
(WebKit::WebPage::selectWordBackward):
(WebKit::WebPage::moveSelectionByOffset):
(WebKit::WebPage::getRectsForGranularityWithSelectionOffset):
(WebKit::WebPage::selectPositionAtBoundaryWithDirection):
(WebKit::WebPage::moveSelectionAtBoundaryWithDirection):
(WebKit::WebPage::rangeForGranularityAtPoint):
(WebKit::WebPage::selectTextWithGranularityAtPoint):
(WebKit::WebPage::beginSelectionInDirection):
(WebKit::WebPage::updateSelectionWithExtentPointAndBoundary):
(WebKit::WebPage::requestDictationContext):
(WebKit::WebPage::applyAutocorrectionInternal):
(WebKit::WebPage::autocorrectionContext):
(WebKit::dataDetectorLinkPositionInformation):
(WebKit::moveByGranularityRespectingWordBoundary):
(WebKit::WebPage::requestDocumentEditingContext):
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::webGLPolicyForURL):
(WebKit::WebPage::resolveWebGLPolicyForURL):
2020-05-15 Alex Christensen <achristensen@webkit.org>
[iOS Simulator] TestWebKitAPI._WKDownload.Resume is a flaky crash
https://bugs.webkit.org/show_bug.cgi?id=211786
Reviewed by Geoffrey Garen.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(-[WKNetworkSessionDelegate URLSession:downloadTask:didResumeAtOffset:expectedTotalBytes:]):
2020-05-15 Alex Christensen <achristensen@webkit.org>
Regression (iOS 13.4, r249142): WKWebView.loading is never reset after a Back navigation in an iframe
https://bugs.webkit.org/show_bug.cgi?id=211449
Reviewed by Geoffrey Garen.
Always call clearPendingAPIRequest in WebPageProxy::didStartProvisionalLoadForFrameShared.
r249142 regressed this and caused more "loading appears to continue forever" bugs fixed in r249890.
This should fix both, by removing the problematic half of r249142.
Covered by an API test.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didStartProvisionalLoadForFrameShared):
2020-05-15 Adrian Perez de Castro <aperez@igalia.com>
[GTK3] Bring back usage of GtkMenu for context menus
https://bugs.webkit.org/show_bug.cgi?id=211557
Reviewed by Carlos Garcia Campos.
* UIProcess/API/glib/WebKitWebView.cpp:
(contextMenuDismissed): Removed.
(webkitWebViewPopulateContextMenu): Use WebContextMenuProxyGtk::widgetDismissedSignal
and move the code from contextMenuDismissed() into an inline lambda.
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(activeContextMenuClosed): Removed.
(webkitWebViewBaseSetActiveContextMenuProxy): Use WebContextMenuProxyGtk::widgetDismissedSignal
and move the code from activeContextMenuClosed() into an inline lambda.
* UIProcess/gtk/WebContextMenuProxyGtk.cpp: Add helper functions to manipulate the actual
widget used for context menus depending on whether GTK3 or GTK4 is being used. The GTK3
versions bring back bits of code removed in r260889.
(WebKit::createMenuWidget): Added.
(WebKit::popupMenuWidget): Added.
(WebKit::popdownMenuWidget): Added.
(WebKit::menuWidgetHasItems): Added.
(WebKit::bindModelToMenuWidget): Added.
(WebKit::WebContextMenuProxyGtk::populate): Use context menu widget helper functions.
(WebKit::WebContextMenuProxyGtk::showContextMenuWithItems): Ditto.
(WebKit::WebContextMenuProxyGtk::WebContextMenuProxyGtk): Ditto. Also handle the
signal triggered when the context menu is dismissed and arrange to call
webkitWebViewContextMenuDismissed() and webkitWebViewBaseContextMenuDismissed().
(WebKit::WebContextMenuProxyGtk::~WebContextMenuProxyGtk): Use helper functions.
* UIProcess/gtk/WebContextMenuProxyGtk.h: Add WebContextMenuProxyGtk::widgetDismissedSignal
constant.
2020-05-14 Kate Cheney <katherine_cheney@apple.com>
Attribute non-network loads and loads with html strings as automatically app-bound
https://bugs.webkit.org/show_bug.cgi?id=211913
<rdar://problem/63157801
Reviewed by Brent Fulgham.
Move logic from WebsiteDataStoreCocoa to WebPageProxy to check for
special app-bound protocols and set m_limitsNavigationToAppBoundDomains
to force these WebViews into app-bound mode.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _isForcedIntoAppBoundMode:]):
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
Testing SPI to see whether m_limitsNavigationsToAppBoundDomains was
set.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::loadData):
Check for html string loads and force WebView into app-bound mode if needed.
No need to worry about setting this variable even if the app hasn't
opted in because setIsNavigatingToAppBoundDomainAndCheckIfPermitted
does an early return before checking this variable if
WKAppBoundDomains does not exist.
(WebKit::shouldTreatURLProtocolAsAppBound):
Adds javascript protocols to this check.
(WebKit::WebPageProxy::setIsNavigatingToAppBoundDomainAndCheckIfPermitted):
Check for special protocols here and force WebView into app-bound mode.
(WebKit::WebPageProxy::isForcedIntoAppBoundModeTesting):
* UIProcess/WebPageProxy.h:
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::WebsiteDataStore::beginAppBoundDomainCheck):
(WebKit::shouldTreatURLProtocolAsAppBound): Deleted.
2020-05-14 Timothy Hatcher <timothy@apple.com>
Add baseURL version of _WKUserStyleSheet forWKWebView.
https://bugs.webkit.org/show_bug.cgi?id=211926
rdar://problem/62074675
Reviewed by Devin Rousso.
* UIProcess/API/Cocoa/_WKUserStyleSheet.h:
* UIProcess/API/Cocoa/_WKUserStyleSheet.mm:
(-[_WKUserStyleSheet initWithSource:forWKWebView:forMainFrameOnly:baseURL:level:userContentWorld:]): Added.
(-[_WKUserStyleSheet initWithSource:forWKWebView:forMainFrameOnly:level:userContentWorld:]): Clean up WebCore::UserStyleSheet initializer.
(-[_WKUserStyleSheet initWithSource:forMainFrameOnly:]): Ditto.
(-[_WKUserStyleSheet initWithSource:forMainFrameOnly:legacyWhitelist:legacyBlacklist:userContentWorld:]): Ditto.
(-[_WKUserStyleSheet initWithSource:forMainFrameOnly:legacyWhitelist:legacyBlacklist:baseURL:userContentWorld:]): Ditto.
(-[_WKUserStyleSheet initWithSource:forMainFrameOnly:legacyWhitelist:legacyBlacklist:baseURL:level:userContentWorld:]): Ditto.
2020-05-14 John Wilander <wilander@apple.com>
Add quirk for cookie blocking latch mode aolmail.com redirecting to aol.com under aol.com
https://bugs.webkit.org/show_bug.cgi?id=211925
<rdar://problem/61901041>
Reviewed by Brent Fulgham.
No new tests. Just a new quirk for a pair of domains added by request.
* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::NetworkDataTaskCocoa::needsFirstPartyCookieBlockingLatchModeQuirk const):
2020-05-14 Chris Dumez <cdumez@apple.com>
Regression(r261163) Causes crashes on builds without libwebrtc.dylib
https://bugs.webkit.org/show_bug.cgi?id=211924
<rdar://problem/63206792>
Reviewed by Jer Noble.
* WebProcess/Network/NetworkProcessConnection.cpp:
(WebKit::NetworkProcessConnection::NetworkProcessConnection):
2020-05-14 Timothy Hatcher <timothy@apple.com>
Add sourceURL to _evaluateJavaScript: so the scripts appear in Web Inspector.
https://bugs.webkit.org/show_bug.cgi?id=211904
rdar://problem/62074376
Reviewed by Devin Rousso.
Added sourceURL version of _evaluateJavaScript: that passes the sourceURL to RunJavaScriptParameters.
If the sourceURL is invalid, generate a unique user script URL so the source code errors are not showing
up with the frame's document URL and bogus locations.
* UIProcess/API/C/WKPage.cpp:
(WKPageRunJavaScriptInMainFrame): pass API::UserScript::generateUniqueURL() for the sourceURL.
* UIProcess/API/Cocoa/WKUserScript.mm:
(-[WKUserScript initWithSource:injectionTime:forMainFrameOnly:]): Clean up WebCore::UserScript initializer.
(-[WKUserScript _initWithSource:injectionTime:forMainFrameOnly:legacyWhitelist:legacyBlacklist:userContentWorld:]): Ditto.
(-[WKUserScript _initWithSource:injectionTime:forMainFrameOnly:legacyWhitelist:legacyBlacklist:associatedURL:userContentWorld:]): Ditto.
(-[WKUserScript _initWithSource:injectionTime:forMainFrameOnly:legacyWhitelist:legacyBlacklist:contentWorld:]): Ditto.
(-[WKUserScript _initWithSource:injectionTime:forMainFrameOnly:legacyWhitelist:legacyBlacklist:associatedURL:contentWorld:]): Ditto.
(-[WKUserScript _initWithSource:injectionTime:forMainFrameOnly:legacyWhitelist:legacyBlacklist:associatedURL:contentWorld:deferRunningUntilNotification:]): Ditto.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _evaluateJavaScript:asAsyncFunction:withArguments:forceUserGesture:inFrame:inWorld:completionHandler:]): Renamed to add sourceURL.
(-[WKWebView _evaluateJavaScript:asAsyncFunction:withSourceURL:withArguments:forceUserGesture:inFrame:inWorld:completionHandler:]): Added sourceURL.
(-[WKWebView _evaluateJavaScript:withSourceURL:inFrame:inContentWorld:completionHandler:]): Added. Pass sourceURL through.
(-[WKWebView evaluateJavaScript:completionHandler:]): Call new method with sourceURL as nil.
(-[WKWebView evaluateJavaScript:inContentWorld:completionHandler:]): Ditto.
(-[WKWebView callAsyncJavaScript:arguments:inContentWorld:completionHandler:]): Ditto.
(-[WKWebView _callAsyncJavaScript:arguments:inFrame:inContentWorld:completionHandler:]): Ditto.
(-[WKWebView _evaluateJavaScript:inFrame:inContentWorld:completionHandler:]): Ditto.
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
2020-05-14 Chris Dumez <cdumez@apple.com>
WebsiteDataStore functions should only send IPC to WebProcesses that are associated with its session
https://bugs.webkit.org/show_bug.cgi?id=211917
Reviewed by Alex Christensen.
WebsiteDataStore functions should only send IPC to WebProcesses that are associated with its session.
setResourceLoadStatisticsEnabled() / clearResourceLoadStatisticsInWebProcesses() were talking to all
WebProcesses, even those associated with other sessions, which was wrong.
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::clearResourceLoadStatistics): Deleted.
* UIProcess/WebProcessPool.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::setResourceLoadStatisticsEnabled):
(WebKit::WebsiteDataStore::clearResourceLoadStatisticsInWebProcesses):
2020-05-14 Daniel Bates <dabates@apple.com>
[iOS] Remove dead dictation code
https://bugs.webkit.org/show_bug.cgi?id=211871
Reviewed by Wenson Hsieh.
UIKit has long removed all callers that would call -nextUnperturbedDictationResultBoundaryFromPosition
and -previousUnperturbedDictationResultBoundaryFromPosition. So, remove the implementations.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView previousUnperturbedDictationResultBoundaryFromPosition:]): Deleted.
(-[WKContentView nextUnperturbedDictationResultBoundaryFromPosition:]): Deleted.
2020-05-14 Brady Eidson <beidson@apple.com>
Crash in PDFPlugin::ensureDataBufferLength
<rdar://problem/62932155> and https://bugs.webkit.org/show_bug.cgi?id=211818
Reviewed by Tim Horton.
There's some currently unreproducible case(s) where a range request finishes
while there's no m_data to append it to.
It's fair hardening to handle that case.
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::getResourceBytesAtPositionMainThread): Handle null m_data.
(WebKit::PDFPlugin::ensureDataBufferLength): Ditto.
2020-05-14 Per Arne Vollan <pvollan@apple.com>
[iOS] Update message filtering rules in the WebContent process' sandbox
https://bugs.webkit.org/show_bug.cgi?id=211188
Reviewed by Brent Fulgham.
Based on telemetry and local testing, update the message filtering rules in the WebContent process' sandbox on iOS.
Messages that have not been observed being in use, should be denied.
No new tests, covered by existing tests.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-05-14 Adrian Perez de Castro <aperez@igalia.com>
Non-unified build fixed, mid May 2020 edition
https://bugs.webkit.org/show_bug.cgi?id=211859
Unreviewed build fix.
* UIProcess/gtk/Clipboard.cpp: Add missing includes for <mutex>, wtf/NeverDestroyed.h, and
wtf/text/WTFString.h.
* WebProcess/WebCoreSupport/WebPlatformStrategies.cpp: Add missing
SharedBufferDataReference.h include.
* WebProcess/WebCoreSupport/WebPlatformStrategies.h: Add missing namespace to usage of
the WebCore::SharedBuffer type.
2020-05-13 David Kilzer <ddkilzer@apple.com>
[IPC] Use templates to reduce duplicate code in IPC::Decoder and IPC::Encoder classes
<https://webkit.org/b/211861>
<rdar://problem/62360390>
Reviewed by Alex Christensen.
In broad strokes, this change fixes the following issues:
- Change `unsigned` type to `size_t` for `alignment` parameters
for consistency on 32-bit platforms.
- Templatize Encoder::encode(), Decoder::decode() and
Decoder::operator>>() methods to reduce duplicate code. This
deleted code in source files, added template methods to header
files, and required a few changes to existing templates in
header files to disambiguate std::is_arithmetic parameter
types.
- Use `typename E` for all template methods that handle enum
types.
- Use `typename T` for all other template method types.
- Move WARN_UNUSED_RETURN to same line as `bool` return type.
- Add FIXME comments to validate enum values.
* Platform/IPC/Decoder.cpp:
(IPC::roundUpToAlignment):
(IPC::Decoder::alignBufferPosition):
(IPC::Decoder::bufferIsLargeEnoughToContain const):
(IPC::Decoder::decodeFixedLengthData):
(IPC::decodeValueFromBuffer): Delete.
(IPC::Decoder::getOptional): Delete.
(IPC::Decoder::operator>>): Delete.
(IPC::Decoder::decode): Delete.
* Platform/IPC/Decoder.h:
(IPC::Decoder::decode):
(IPC::Decoder::operator>>):
(IPC::Decoder::decodeEnum):
* Platform/IPC/Encoder.cpp:
(IPC::roundUpToAlignment):
(IPC::Encoder::grow):
(IPC::Encoder::encodeFixedLengthData):
(IPC::copyValueToBuffer): Delete.
(IPC::Encoder::encode): Delete.
* Platform/IPC/Encoder.h:
2020-05-13 Devin Rousso <drousso@apple.com>
Web Inspector: `RuntimeEnabledFeatures` should also be enabled when remotely inspecting
https://bugs.webkit.org/show_bug.cgi?id=211874
Reviewed by Timothy Hatcher.
* WebProcess/Inspector/WebInspectorUI.h:
* WebProcess/Inspector/WebInspectorUI.cpp:
(WebKit::WebInspectorUI::enableFrontendFeatures): Added.
(WebKit::WebInspectorUI::WebInspectorUI):
* WebProcess/Inspector/RemoteWebInspectorUI.cpp:
(WebKit::RemoteWebInspectorUI::RemoteWebInspectorUI):
2020-05-13 Megan Gardner <megan_gardner@apple.com>
Unreviewed Build Fix.
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::setMouseEventPolicy):
* UIProcess/ios/forms/WKDateTimeInputControl.mm:
(-[WKDateTimePicker dealloc]):
(-[WKDateTimePicker controlBeginEditing]):
(-[WKDateTimePicker controlEndEditing]):
2020-05-13 Kate Cheney <katherine_cheney@apple.com>
SHOULD NEVER BE REACHED in WebKit::ResourceLoadStatisticsDatabaseStore::openAndUpdateSchemaIfNecessary
https://bugs.webkit.org/show_bug.cgi?id=211637
<rdar://problem/63031185>
Reviewed by John Wilander.
No new tests, behavior confirmed by existing tests.
Stop reseting ITP to use the memory store between tests. This results
in consecutive calls to NetworkSession::recreateResourceLoadStatisticStore
if two database tests occur in a row, which can set the m_resourceLoadStatistics
object before a previous call to WebResourceLoadStatisticsStore::flushAndDestroyPersistentStore
is done.
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreStatisticsResetToConsistentState):
2020-05-13 Megan Gardner <megan_gardner@apple.com>
Change Date/Time popovers to contextMenus.
https://bugs.webkit.org/show_bug.cgi?id=211825
<rdar://problem/63102524>
Reviewed by Wenson Hsieh.
Update the Date/Time picker to use UIContextMenus instead of popovers.
Removed all the unneeded popover code.
Renamed WKFormInputControl to WKDateTimeInputControl because it was only used
for date and time controls, unneccessary abstraction.
We need to ensure the poisition information is up to date before presenting
the context menu, or the position it will be presented from can be stale.
* SourcesCocoa.txt:
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _shouldShowAutomaticKeyboardUIIgnoringInputMode]):
(-[WKContentView requiresAccessoryView]):
(-[WKContentView _updateAccessory]):
(mayContainSelectableText):
(createInputPeripheralWithView):
(-[WKContentView inputLabelText]):
(-[WKContentView _removeContextMenuViewIfPossible]):
(-[WKContentView inputLabelTextForViewController:]):
(-[WKContentView formInputControl]):
(-[WKContentView setTimePickerValueToHour:minute:]):
(-[WKContentView timePickerValueHour]):
(-[WKContentView timePickerValueMinute]):
* UIProcess/ios/forms/WKDateTimeInputControl.h: Renamed from Source/WebKit/UIProcess/ios/forms/WKFormInputControl.h.
* UIProcess/ios/forms/WKDateTimeInputControl.mm: Renamed from Source/WebKit/UIProcess/ios/forms/WKFormInputControl.mm.
(-[WKDateTimeContextMenuViewController init]):
(-[WKDateTimeContextMenuViewController preferredContentSize]):
(-[WKDateTimePicker initWithView:datePickerMode:]):
(-[WKDateTimePicker contextMenuInteraction:previewForHighlightingMenuWithConfiguration:]):
(-[WKDateTimePicker _contextMenuInteraction:styleForMenuWithConfiguration:]):
(-[WKDateTimePicker contextMenuInteraction:configurationForMenuAtLocation:]):
(-[WKDateTimePicker contextMenuInteraction:willEndForConfiguration:animator:]):
(-[WKDateTimePicker removeContextMenuInteraction]):
(-[WKDateTimePicker ensureContextMenuInteraction]):
(-[WKDateTimePicker showDateTimePicker]):
(-[WKDateTimePicker reset:]):
(-[WKDateTimePicker calendarType]):
(-[WKDateTimePicker dealloc]):
(-[WKDateTimePicker setDateTimePickerToInitialValue]):
(-[WKDateTimePicker controlBeginEditing]):
(-[WKDateTimePicker viewController]):
(-[WKDateTimePicker controlEndEditing]):
(-[WKDateTimeInputControl initWithView:]):
(-[WKDateTimeInputControl setTimePickerHour:minute:]):
(-[WKDateTimeInputControl dateTimePickerCalendarType]):
(-[WKDateTimeInputControl timePickerValueHour]):
(-[WKDateTimeInputControl timePickerValueMinute]):
* WebKit.xcodeproj/project.pbxproj:
2020-05-13 Wenson Hsieh <wenson_hsieh@apple.com>
Unreviewed, fix tvOS and watchOS builds after r261638
`-[UIPasteboard itemProviders]` is not present on watchOS and tvOS.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _writePromisedAttachmentToPasteboard:]):
2020-05-13 Timothy Horton <timothy_horton@apple.com>
Build fix.
* UIProcess/API/Cocoa/WKWebpagePreferencesPrivate.h:
2020-05-13 Jer Noble <jer.noble@apple.com>
Replace isNullFunctionPointer with real weak-linking support
https://bugs.webkit.org/show_bug.cgi?id=211751
Reviewed by Sam Weinig.
Use the new WTF_WEAK_LINK_FORCE_IMPORT macro.
* Platform/classifier/cocoa/ResourceLoadStatisticsClassifierCocoa.cpp:
(WebKit::ResourceLoadStatisticsClassifierCocoa::canUseCorePrediction):
2020-05-13 Tim Horton <timothy_horton@apple.com>
Add SPI for reverting to touch events for iPad trackpad interactions
https://bugs.webkit.org/show_bug.cgi?id=211824
<rdar://problem/61363084>
Reviewed by Megan Gardner.
New Test: IOSMouseSupport.WebsiteMouseEventPolicies
Plumb a "mouse event policy" WKWebpagePreference, which has two values:
- Default, which means that WebKit decides whether or not indirect
pointing devices send mouse or touch events.
- SynthesizeTouchEvents, which ensures that WebKit will always send
touch events from indirect pointing devices.
This value is only available on platforms that implement touch events.
We use the WKWebpagePreferences mechanism in order to make it easy to
implement it as a "per-site" quirk, but unlike most WKWebpagePreferences,
it turns around and changes UI process state (the enablement of the
WKMouseGestureRecgonizer).
* Scripts/webkit/messages.py:
* Shared/WebsitePoliciesData.cpp:
(WebKit::WebsitePoliciesData::encode const):
(WebKit::WebsitePoliciesData::decode):
(WebKit::WebsitePoliciesData::applyToDocumentLoader):
* Shared/WebsitePoliciesData.h:
* UIProcess/API/APIWebsitePolicies.cpp:
(API::WebsitePolicies::copy const):
(API::WebsitePolicies::data):
* UIProcess/API/APIWebsitePolicies.h:
* UIProcess/API/Cocoa/WKWebpagePreferences.mm:
(WebKit::mouseEventPolicy):
(WebKit::coreMouseEventPolicy):
(-[WKWebpagePreferences _setMouseEventPolicy:]):
(-[WKWebpagePreferences _mouseEventPolicy]):
* UIProcess/API/Cocoa/WKWebpagePreferencesPrivate.h:
* UIProcess/PageClient.h:
(WebKit::PageClient::setMouseEventPolicy):
* UIProcess/ProvisionalPageProxy.cpp:
(WebKit::ProvisionalPageProxy::didCommitLoadForFrame):
* UIProcess/ProvisionalPageProxy.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::commitProvisionalPage):
(WebKit::WebPageProxy::didCommitLoadForFrame):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/ios/PageClientImplIOS.h:
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::setMouseEventPolicy):
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView shouldUseMouseGestureRecognizer]):
(-[WKContentView setUpMouseGestureRecognizer]):
(-[WKContentView _configureMouseGestureRecognizer]):
(-[WKContentView _setMouseEventPolicy:]):
* UIProcess/ios/WKMouseGestureRecognizer.mm:
(-[WKMouseGestureRecognizer setEnabled:]):
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::dispatchDidCommitLoad):
2020-05-13 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] "Copy" context menu action for attachment element does not work in Mail
https://bugs.webkit.org/show_bug.cgi?id=211817
<rdar://problem/58043110>
Reviewed by Tim Horton.
Implements support for copying an attachment element, via context menu actions. To achieve this, we first
refactor code in WebCore that is currently responsible for converting an attachment element into a
PromisedAttachmentInfo, which provides a handle to an UI-process-side API attachment object. We then use this
helper in WebKit to send PromisedAttachmentInfo back to the UI process when handling the copy action from the
context menu on iOS.
On iOS, we then take this promised attachment info in the UI process, map it to an API::Attachment object, and
use the file handle to create and write an NSItemProvider to the general pasteboard. (In the future, similar
logic could be implemented on macOS as well to handle copying attachment elements by implementing a version of
writePromisedAttachmentToPasteboard in PageClientImplMac. Currently, we fall back to treating this case as if
we're copying a text selection containing a single attachment element).
* UIProcess/API/ios/WKWebViewPrivateForTestingIOS.h:
Add a new helper function to simulate an element action (_WKElementActionType) for an element at the given
location. This is used by the new API test.
* UIProcess/API/ios/WKWebViewTestingIOS.mm:
(-[WKWebView _simulateElementAction:atLocation:]):
* UIProcess/PageClient.h:
(WebKit::PageClient::writePromisedAttachmentToPasteboard):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::writePromisedAttachmentToPasteboard):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
Add an IPC message for the UI process to receive a PromisedAttachmentInfo, and copy it to the system pasteboard
if it maps to an actual API::Attachment. For now, we only support copying data if the attachment is backed by
a file wrapper in the UI process (as opposed to blob data written by the web process).
* UIProcess/ios/PageClientImplIOS.h:
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::writePromisedAttachmentToPasteboard):
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(createItemProvider):
Create a helper function to convert PromisedAttachmentInfo into an NSItemProvider. This is similar to
`-_prepareToDragPromisedAttachment:`, but with a couple differences: (1) we only create an NSItemProvider
instead of going through `WebItemProviderRegistrationInfoList`; (2) since there's no opportunity to clean up
temporary after pasting, serialize the attachment data using `-serializedRepresentation`, and provide the
attachment data instead of a file URL.
Ideally, we would offer the data directly to the pasteboard so that even if Mail terminates, the contents of the
pasteboard can still be provided when pasting; however, it doesn't seem like we can do this while being able to
offer a suggested name for the item we're writing. Some other apps work around this limitation by writing the
file name as plain text on the pasteboard, but this doesn't work then attempting to copy a plain text file that
has a file name.
(-[WKContentView _writePromisedAttachmentToPasteboard:]):
(-[WKContentView _simulateElementAction:atLocation:]):
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::performActionOnElement):
2020-05-13 Per Arne Vollan <pvollan@apple.com>
[iOS] Update message filtering rules in the WebContent process' sandbox
https://bugs.webkit.org/show_bug.cgi?id=211188
Reviewed by Brent Fulgham.
Based on telemetry and local testing, update the message filtering rules in the WebContent process' sandbox on iOS.
Messages that have not been observed being in use, should be denied.
No new tests, covered by existing tests.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-05-13 Per Arne Vollan <pvollan@apple.com>
[iOS] In CFPrefs direct mode, the WebContent sandbox needs to allow access to plist files
https://bugs.webkit.org/show_bug.cgi?id=211782
Reviewed by Brent Fulgham.
Since preference files will be read directly in CFPrefs direct mode, the WebContent sandbox on iOS needs to allow reading these files.
No new tests. There is API test coverage for this in PreferenceChanges.mm, but these tests are only run on simulator, not device.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-05-13 Peng Liu <peng.liu6@apple.com>
Add the feature flags plist file back
https://bugs.webkit.org/show_bug.cgi?id=211816
Reviewed by Jer Noble.
Add the WebKit.plist file to the WebKit project and use it instead of the one from the WebKitAdditions.
* FeatureFlags/WebKit.plist: Added.
* WebKit.xcodeproj/project.pbxproj:
2020-05-13 Youenn Fablet <youenn@apple.com>
Allow WebAudioBufferList to dynamically change its number of frames
https://bugs.webkit.org/show_bug.cgi?id=211720
Reviewed by Eric Carlson.
* WebProcess/cocoa/RemoteCaptureSampleManager.cpp:
(WebKit::RemoteCaptureSampleManager::RemoteAudio::setStorage):
(WebKit::RemoteCaptureSampleManager::RemoteAudio::audioSamplesAvailable):
* WebProcess/cocoa/RemoteCaptureSampleManager.h:
2020-05-13 Per Arne Vollan <pvollan@apple.com>
Unreviewed compile fix after r261584.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
2020-05-13 Chris Dumez <cdumez@apple.com>
REGRESSION (r261254): Broke Apple Pay on macOS
https://bugs.webkit.org/show_bug.cgi?id=211826
<rdar://problem/63161750>
Reviewed by Alex Christensen.
r261254 dropped the code needed to add the WebPaymentCoordinatorProxy as an IPC message receiver.
This patch re-introduces the code in order to address the regression.
No new tests, tested manually on https://applepaydemo.apple.com.
* NetworkProcess/NetworkConnectionToWebProcess.h:
* NetworkProcess/ios/NetworkConnectionToWebProcessIOS.mm:
(WebKit::NetworkConnectionToWebProcess::paymentCoordinatorAddMessageReceiver):
(WebKit::NetworkConnectionToWebProcess::paymentCoordinatorRemoveMessageReceiver):
* Shared/ApplePay/WebPaymentCoordinatorProxy.h:
* Shared/ApplePay/cocoa/WebPaymentCoordinatorProxyCocoa.mm:
(WebKit::WebPaymentCoordinatorProxy::WebPaymentCoordinatorProxy):
(WebKit::WebPaymentCoordinatorProxy::~WebPaymentCoordinatorProxy):
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::paymentCoordinatorAddMessageReceiver):
(WebKit::WebPageProxy::paymentCoordinatorRemoveMessageReceiver):
* UIProcess/WebPageProxy.h:
2020-05-13 Zan Dobersek <zdobersek@igalia.com>
REGRESSION(r260571): RequestedScrollPosition doesn't have its scroll change applied to layers anymore
https://bugs.webkit.org/show_bug.cgi?id=211374
Reviewed by Carlos Garcia Campos.
Use WebPage::finalizeRenderingUpdate() to perform the compositing state flush
and the scrolling tree layer positioning, instead of doing things half-baked
by manually invoking the FrameView method.
Additional WebPage calls (updateRendering(), flushPendingEditorStateUpdate())
are moved into CompositingCoordinator::flushPendingLayerChanges() from the
LayerTreeHost::layerFlushTimerFired() method for consistency. The additional
CompositingCoordinator::syncDisplayState() method call is removed -- this call
triggered any necessary layout, but that's already done through WebPage::updateRendering().
CompositingCoordinator::flushPendingLayerChanges() doesn't rely on the
FrameView flush return value since that's not available anymore -- instead,
true is returned always.
* WebProcess/WebPage/CoordinatedGraphics/CompositingCoordinator.cpp:
(WebKit::CompositingCoordinator::flushPendingLayerChanges):
* WebProcess/WebPage/CoordinatedGraphics/LayerTreeHost.cpp:
(WebKit::LayerTreeHost::layerFlushTimerFired):
2020-05-13 Per Arne Vollan <pvollan@apple.com>
Unreviewed compile fix after r261584.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
2020-05-13 Tomoki Imai <Tomoki.Imai@sony.com>
Selected element on Web Inspector is not highlighted with CPU Rendering.
https://bugs.webkit.org/show_bug.cgi?id=195933
Reviewed by Devin Rousso.
Since WebInspectorClient::drawHighlight was not called while non accelerated compositing mode, we cannot see element highlight on web inspector.
This patch adds WebInspectorClient::drawHighlight call in WebPage::drawRect, which draws webpage image while non accelerated compositing mode, to overlay the highlight.
WebInspectorClient::highlight and WebInspectorClient::hideHighlight currently requests re-paint whole web page, but it should be able to optimize by only updating dirty rects in the future.
* WebProcess/Inspector/WebInspectorClient.cpp:
(WebKit::WebInspectorClient::highlight): If it's not acceleratedCompositingEnabled, then just request repaint.
(WebKit::WebInspectorClient::hideHighlight): If it's not acceleratedCompositingEnabled, then just request repaint.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::drawRect): Draw inspector overlay here if it's not accelerated compositing mode.
2020-05-13 Carlos Garcia Campos <cgarcia@igalia.com>
Unreviewed. Fix GTK debug build after r261554
Remove writeToClipboard that receives a const SelectionData& that is no longer used.
* WebProcess/WebCoreSupport/WebPlatformStrategies.cpp:
* WebProcess/WebCoreSupport/WebPlatformStrategies.h:
2020-05-12 Peng Liu <peng.liu6@apple.com>
Add the feature flag plist file parser
https://bugs.webkit.org/show_bug.cgi?id=211457
Reviewed by Tim Horton.
WebKit needs to parse the feature flags plist by itself in some cases.
* Shared/Cocoa/WebPreferencesDefaultValuesCocoa.mm: Added.
(WebKit::featureFlagEnabled):
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultAsyncFrameAndOverflowScrollingEnabled):
(WebKit::defaultUseGPUProcessForMedia):
(WebKit::defaultRenderCanvasInGPUProcessEnabled):
(WebKit::defaultCaptureAudioInGPUProcessEnabled):
(WebKit::defaultCaptureVideoInGPUProcessEnabled):
(WebKit::defaultWebRTCCodecsInGPUProcess):
(WebKit::defaultWebGL2Enabled):
(WebKit::defaultWebGPUEnabled):
(WebKit::defaultInAppBrowserPrivacy):
(WebKit::defaultIncrementalPDFEnabled):
(WebKit::defaultWebXREnabled):
* Shared/WebPreferencesDefaultValues.h:
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
2020-05-12 Alex Christensen <achristensen@webkit.org>
Give some NetworkLoadMetrics to WebCoreNSURLSession's delegate
https://bugs.webkit.org/show_bug.cgi?id=211759
<rdar://problem/62909440>
Reviewed by Jer Noble.
This also reduces duplicate lookups in RemoteMediaResourceManager
* GPUProcess/media/RemoteMediaResource.cpp:
(WebKit::RemoteMediaResource::loadFinished):
* GPUProcess/media/RemoteMediaResource.h:
* GPUProcess/media/RemoteMediaResourceManager.cpp:
(WebKit::RemoteMediaResourceManager::responseReceived):
(WebKit::RemoteMediaResourceManager::redirectReceived):
(WebKit::RemoteMediaResourceManager::dataSent):
(WebKit::RemoteMediaResourceManager::dataReceived):
(WebKit::RemoteMediaResourceManager::accessControlCheckFailed):
(WebKit::RemoteMediaResourceManager::loadFailed):
(WebKit::RemoteMediaResourceManager::loadFinished):
* GPUProcess/media/RemoteMediaResourceManager.h:
* GPUProcess/media/RemoteMediaResourceManager.messages.in:
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(-[WKNetworkSessionDelegate URLSession:task:didFinishCollectingMetrics:]):
* WebProcess/GPU/media/RemoteMediaResourceProxy.cpp:
(WebKit::RemoteMediaResourceProxy::loadFinished):
* WebProcess/GPU/media/RemoteMediaResourceProxy.h:
2020-05-12 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Don't assume extensions always exist
https://bugs.webkit.org/show_bug.cgi?id=211760
<rdar://problem/61217642>
Reviewed by Brent Fulgham.
* UIProcess/WebAuthentication/AuthenticatorManager.cpp:
(WebKit::WebCore::processGoogleLegacyAppIdSupportExtension):
* UIProcess/WebAuthentication/fido/CtapAuthenticator.cpp:
(WebKit::CtapAuthenticator::processGoogleLegacyAppIdSupportExtension):
2020-05-12 Chris Dumez <cdumez@apple.com>
[WK2] Neuter WKFrameIsFrameSet() / WKPageGetFrameSetLargestFrame() C API
https://bugs.webkit.org/show_bug.cgi?id=211808
Reviewed by Darin Adler.
Neuter WKFrameIsFrameSet() / WKPageGetFrameSetLargestFrame() C API. This is only SPI and is only used for slightly
different printing behavior in Safari. Framesets are no longer supported in HTML5 and are now super rare. Support
for this C API adds quite a bit of code complexity and crashes such as <rdar://problem/60322282>, it just does not
seem worth it anymore.
* UIProcess/API/C/WKFrame.cpp:
(WKFrameIsFrameSet):
* UIProcess/API/C/WKPage.cpp:
(WKPageGetFrameSetLargestFrame):
* UIProcess/WebFrameProxy.cpp:
(WebKit::WebFrameProxy::didCommitLoad):
* UIProcess/WebFrameProxy.h:
(WebKit::WebFrameProxy::setIsFrameSet): Deleted.
(WebKit::WebFrameProxy::isFrameSet const): Deleted.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::resetState):
(WebKit::WebPageProxy::frameDidBecomeFrameSet): Deleted.
(WebKit::WebPageProxy::frameSetLargestFrameChanged): Deleted.
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::focusedFrame const):
(WebKit::WebPageProxy::frameSetLargestFrame const): Deleted.
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::contentsSizeChanged const):
(WebKit::findLargestFrameInFrameSet): Deleted.
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::dispatchDidBecomeFrameset): Deleted.
* WebProcess/WebCoreSupport/WebFrameLoaderClient.h:
2020-05-12 Per Arne Vollan <pvollan@apple.com>
[iOS] Rebuild MobileGestalt cache if needed
https://bugs.webkit.org/show_bug.cgi?id=211739
Reviewed by Darin Adler.
This patch was written by Geoff Garen and myself. Rebuild the MobileGestalt cache in the UI process if the
MobileGestalt cache is stale. This should be done before starting a new WebContent process. If the cache
is not rebuilt, every MobileGestalt query in the WebContent process will fail, since access to the daemon
is denied.
No new tests, since testing this is not trivial. It would involve deleting the MobileGestalt cache and
making sure it is rebuilt afterwards. This could be an API test, but API tests are not supported on iOS
devices.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
2020-05-12 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] Rework drag and drop handling in preparation for GTK4
https://bugs.webkit.org/show_bug.cgi?id=211723
Reviewed by Adrian Perez de Castro.
Split DragAndDropHelper class into DragSource and DropTarget classes. This separates the source and destination
parts making it easier to follow. This patch also adds PageClient::didPerformDragControllerAction() to notify
back the view when an operation is done to update the result.
* SourcesGTK.txt:
* UIProcess/API/gtk/DragSource.h: Added.
* UIProcess/API/gtk/DragSourceGtk3.cpp: Added.
(WebKit::DragSource::DragSource):
(WebKit::DragSource::~DragSource):
(WebKit::DragSource::begin):
* UIProcess/API/gtk/DropTarget.h: Added.
* UIProcess/API/gtk/DropTargetGtk3.cpp: Added.
(WebKit::DropTarget::DropTarget):
(WebKit::DropTarget::~DropTarget):
(WebKit::DropTarget::accept):
(WebKit::DropTarget::enter):
(WebKit::DropTarget::update):
(WebKit::DropTarget::dataReceived):
(WebKit::DropTarget::didPerformAction):
(WebKit::DropTarget::leaveTimerFired):
(WebKit::DropTarget::leave):
(WebKit::DropTarget::drop):
* UIProcess/API/gtk/PageClientImpl.cpp:
(WebKit::PageClientImpl::startDrag):
(WebKit::PageClientImpl::didPerformDragControllerAction):
* UIProcess/API/gtk/PageClientImpl.h:
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseConstructed):
(webkit_web_view_base_class_init):
(webkitWebViewBaseStartDrag):
(webkitWebViewBaseDidPerformDragControllerAction):
* UIProcess/API/gtk/WebKitWebViewBasePrivate.h:
* UIProcess/PageClient.h:
(WebKit::PageClient::didPerformDragControllerAction):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didPerformDragControllerAction):
* UIProcess/gtk/DragAndDropHandler.cpp: Removed.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::performDragControllerAction):
* WebProcess/WebPage/gtk/WebPageGtk.cpp:
2020-05-12 Jacob Uphoff <jacob_uphoff@apple.com>
Unreviewed, reverting r261557.
This commit caused testing to exit early due to too many
crashes on macOS Catalina Asan
Reverted changeset:
"Allow WebAudioBufferList to dynamically change its number of
frames"
https://bugs.webkit.org/show_bug.cgi?id=211720
https://trac.webkit.org/changeset/261557
2020-05-12 Per Arne Vollan <pvollan@apple.com>
Unreviewed crash fix.
Use correct global name for icon service in RELEASE_ASSERT.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::didChooseFilesForOpenPanelWithDisplayStringAndIcon):
2020-05-12 Youenn Fablet <youenn@apple.com>
Allow WebAudioBufferList to dynamically change its number of frames
https://bugs.webkit.org/show_bug.cgi?id=211720
Reviewed by Eric Carlson.
* WebProcess/cocoa/RemoteCaptureSampleManager.cpp:
(WebKit::RemoteCaptureSampleManager::RemoteAudio::setStorage):
(WebKit::RemoteCaptureSampleManager::RemoteAudio::audioSamplesAvailable):
* WebProcess/cocoa/RemoteCaptureSampleManager.h:
2020-05-12 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Add support for clipboard operations
https://bugs.webkit.org/show_bug.cgi?id=211561
Reviewed by Adrian Perez de Castro.
* UIProcess/gtk/Clipboard.h:
* UIProcess/gtk/ClipboardGtk4.cpp:
(WebKit::Clipboard::Clipboard):
(WebKit::Clipboard::type const):
(WebKit::Clipboard::formats):
(WebKit::ReadTextAsyncData::ReadTextAsyncData):
(WebKit::Clipboard::readText):
(WebKit::ReadFilePathsAsyncData::ReadFilePathsAsyncData):
(WebKit::Clipboard::readFilePaths):
(WebKit::ReadBufferAsyncData::ReadBufferAsyncData):
(WebKit::Clipboard::readBuffer):
(WebKit::Clipboard::write):
(WebKit::Clipboard::clear):
2020-05-12 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] Rework clipboard handling in preparation for GTK4
https://bugs.webkit.org/show_bug.cgi?id=211511
Reviewed by Adrian Perez de Castro.
Use async APIs to avoid blocking the UI process during the communication with the clipboard. In GTK4 only async
APIs are available. Add Clipboard class to handle the communication with the clipboard with the GTK
implementation in ClipboardGtk3 and ClipboardGtk4 files.
* Shared/gtk/WebSelectionData.cpp:
(WebKit::WebSelectionData::WebSelectionData): Add constructor receivnig an rvalue reference.
* Shared/gtk/WebSelectionData.h:
* SourcesGTK.txt:
* UIProcess/WebPasteboardProxy.h:
* UIProcess/WebPasteboardProxy.messages.in:
* UIProcess/gtk/Clipboard.cpp: Added.
(WebKit::clipboard):
(WebKit::primary):
(WebKit::Clipboard::get):
* UIProcess/gtk/Clipboard.h: Added.
* UIProcess/gtk/ClipboardGtk3.cpp: Added.
(WebKit::Clipboard::Clipboard):
(WebKit::isPrimaryClipboard):
(WebKit::Clipboard::type const):
(WebKit::FormatsAsyncData::FormatsAsyncData):
(WebKit::Clipboard::formats):
(WebKit::ReadTextAsyncData::ReadTextAsyncData):
(WebKit::Clipboard::readText):
(WebKit::ReadFilePathsAsyncData::ReadFilePathsAsyncData):
(WebKit::Clipboard::readFilePaths):
(WebKit::ReadBufferAsyncData::ReadBufferAsyncData):
(WebKit::Clipboard::readBuffer):
(WebKit::Clipboard::write):
(WebKit::Clipboard::clear):
* UIProcess/gtk/ClipboardGtk4.cpp: Added.
(WebKit::Clipboard::Clipboard):
(WebKit::Clipboard::type const):
(WebKit::Clipboard::formats):
(WebKit::Clipboard::readText):
(WebKit::Clipboard::readFilePaths):
(WebKit::Clipboard::readBuffer):
(WebKit::Clipboard::write):
(WebKit::Clipboard::clear):
* UIProcess/gtk/WebPasteboardProxyGtk.cpp:
(WebKit::WebPasteboardProxy::getTypes):
(WebKit::WebPasteboardProxy::readText):
(WebKit::WebPasteboardProxy::readFilePaths):
(WebKit::WebPasteboardProxy::readBuffer):
(WebKit::WebPasteboardProxy::writeToClipboard):
(WebKit::WebPasteboardProxy::clearClipboard):
* WebProcess/WebCoreSupport/WebPlatformStrategies.cpp:
(WebKit::WebPlatformStrategies::types):
(WebKit::WebPlatformStrategies::readTextFromClipboard):
(WebKit::WebPlatformStrategies::readFilePathsFromClipboard):
(WebKit::WebPlatformStrategies::readBufferFromClipboard):
(WebKit::WebPlatformStrategies::writeToClipboard):
(WebKit::WebPlatformStrategies::clearClipboard):
* WebProcess/WebCoreSupport/WebPlatformStrategies.h:
2020-05-12 Youenn Fablet <youenn@apple.com>
Introduce a RealtimeMediaSource video sample observer
https://bugs.webkit.org/show_bug.cgi?id=211718
Reviewed by Eric Carlson.
* UIProcess/Cocoa/UserMediaCaptureManagerProxy.cpp:
(WebKit::UserMediaCaptureManagerProxy::SourceProxy::SourceProxy):
(WebKit::UserMediaCaptureManagerProxy::SourceProxy::~SourceProxy):
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
(WebKit::MediaRecorderPrivate::MediaRecorderPrivate):
(WebKit::MediaRecorderPrivate::~MediaRecorderPrivate):
(WebKit::MediaRecorderPrivate::videoSampleAvailable):
(WebKit::MediaRecorderPrivate::stopRecording):
(WebKit::MediaRecorderPrivate::sampleBufferUpdated): Deleted.
* WebProcess/GPU/webrtc/MediaRecorderPrivate.h:
2020-05-12 Mark Lam <mark.lam@apple.com>
Wasm::enableFastMemory() was called too late.
https://bugs.webkit.org/show_bug.cgi?id=211773
Reviewed by Yusuke Suzuki.
Previously, we were calling Wasm::enableFastMemory() in WebProcess::initializeWebProcess().
By then, it is too late because the JSC::Config and WTF::Config have already been
frozen. There is nothing special about Wasm::enableFastMemory() that it needs to
be called this late. This patch moves it into JSC::initializeThreading() where
it belongs.
This change also make the initialization of Wasm fast memory consistent with how
we have always done it for the jsc shell i.e. it is initialized at the end of
JSC::initializeThreading().
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::initializeWebProcess):
2020-05-11 Darin Adler <darin@apple.com>
Fix problems caught by replacing WTF::Optional with std::optional
https://bugs.webkit.org/show_bug.cgi?id=211703
Reviewed by Chris Dumez.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::plainTextForContext): Clarified ambiguous overload resolution by writing
the conversion from a live range to a SimpleRange out explicitly. Also fixed a typo
where there was a missing "&" in a "const&" argument.
2020-05-11 Simon Fraser <simon.fraser@apple.com>
[ macOS ] scrollingcoordinator/mac/latching/scrolling-select-should-not-latch-mainframe.html is a flaky failure
https://bugs.webkit.org/show_bug.cgi?id=211747
Reviewed by Tim Horton.
Add an option to monitorWheelEvents to reset latching.
* WebProcess/InjectedBundle/API/c/WKBundlePage.cpp:
(WKBundlePageStartMonitoringScrollOperations):
* WebProcess/InjectedBundle/API/c/WKBundlePage.h:
2020-05-11 Kate Cheney <katherine_cheney@apple.com>
Fail navigations to non app-bound domains after use of app-bound APIs
https://bugs.webkit.org/show_bug.cgi?id=211647
<rdar://problem/62978159>
Reviewed by Brent Fulgham.
A lot of this patch is deleting the unnecessary variable/function
hasNavigatedAwayFromAppBoundDomain now that navigating away from an
app-bound domain is not possible.
To address the bug, this sets the default isNavigatingToAppBoundDomain
value to be WTF::nullopt. This will allow app-bound behaviors until
a navigation has been attempted, in which case it will fail or will
update isNavigatingToAppBoundDomain to the correct value.
* Shared/LoadParameters.cpp:
(WebKit::LoadParameters::encode const):
(WebKit::LoadParameters::decode):
* Shared/LoadParameters.h:
* Shared/PolicyDecision.h:
(WebKit::PolicyDecision::encode const):
(WebKit::PolicyDecision::decode):
* UIProcess/ProvisionalPageProxy.cpp:
(WebKit::ProvisionalPageProxy::loadData):
(WebKit::ProvisionalPageProxy::loadRequest):
(WebKit::ProvisionalPageProxy::decidePolicyForNavigationActionSync):
* UIProcess/ProvisionalPageProxy.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::loadRequest):
(WebKit::WebPageProxy::loadRequestWithNavigationShared):
(WebKit::WebPageProxy::loadData):
(WebKit::WebPageProxy::loadDataWithNavigationShared):
(WebKit::WebPageProxy::receivedPolicyDecision):
(WebKit::WebPageProxy::continueNavigationInNewProcess):
(WebKit::WebPageProxy::decidePolicyForNavigationActionSyncShared):
(WebKit::WebPageProxy::hasNavigatedAwayFromAppBoundDomain const): Deleted.
Remove hasNavigatedAwayFromAppBoundDomain.
(WebKit::WebPageProxy::setIsNavigatingToAppBoundDomainAndCheckIfPermitted):
Check for new m_hasExecutedAppBoundBehaviorBeforeNavigation parameter
and fail the navigation if a WebView has used app-bound behavior then
tries to navigate to a non app-bound domain.
(WebKit::WebPageProxy::decidePolicyForNavigationAction):
Update error message to be more general now that more than one error
can occur.
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::setHasExecutedAppBoundBehaviorBeforeNavigation):
* UIProcess/WebPageProxy.messages.in:
New parameter to indicate a WebView has used app-bound APIs before
navigating.
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::dispatchDecidePolicyForResponse):
(WebKit::WebFrameLoaderClient::dispatchDecidePolicyForNavigationAction):
(WebKit::WebFrameLoaderClient::shouldEnableInAppBrowserPrivacyProtections):
(WebKit::WebFrameLoaderClient::hasNavigatedAwayFromAppBoundDomain): Deleted.
(WebKit::WebFrameLoaderClient::needsInAppBrowserPrivacyQuirks const): Deleted.
* WebProcess/WebCoreSupport/WebFrameLoaderClient.h:
Combined two functions into a simpler function:
shouldEnableInAppBrowserPrivacyProtections().
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
(WebKit::WebPage::updatePreferences):
Store the special app-bound domain flag to gate the service worker
API. This should be stored separately from
m_isNavigatingToAppBoundDomain, because in the WebPage constructor we
don't yet know whether the WKAppBoundDomains key exists.
(WebKit::WebPage::loadRequest):
(WebKit::WebPage::loadDataImpl):
(WebKit::WebPage::loadData):
(WebKit::WebPage::loadAlternateHTML):
(WebKit::WebPage::didReceivePolicyDecision):
Remove hasNavigatedAwayFromAppBoundDomain.
(WebKit::WebPage::runJavaScript):
(WebKit::WebPage::setIsNavigatingToAppBoundDomain):
(WebKit::WebPage::shouldEnableInAppBrowserPrivacyProtections):
If m_needsInAppBrowserPrivacyQuirks are on for testing, don't enable
protections. Only notify the WebPageProxy of app-bound behavior if
a navigation has not occured and we know the webView is not app-bound
(in order to limit IPC).
(WebKit::m_isNavigatingToAppBoundDomain): Deleted.
* WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::isNavigatingToAppBoundDomain const):
(WebKit::WebPage::setIsNavigatingToAppBoundDomain): Deleted.
(WebKit::WebPage::hasNavigatedAwayFromAppBoundDomain const): Deleted.
(WebKit::WebPage::setHasNavigatedAwayFromAppBoundDomain): Deleted.
(WebKit::WebPage::needsInAppBrowserPrivacyQuirks const): Deleted.
2020-05-11 Wenson Hsieh <wenson_hsieh@apple.com>
Remove some unnecessary indirection when getting Document’s Editor
https://bugs.webkit.org/show_bug.cgi?id=211744
Reviewed by Geoffrey Garen.
See WebCore ChangeLog for more information. No change in behavior.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::performActionOnElement):
2020-05-11 Simon Fraser <simon.fraser@apple.com>
Fix possible crash in ViewGestureController on macOS
https://bugs.webkit.org/show_bug.cgi?id=211745
Reviewed by Tim Horton.
When ViewGestureController::resetState() clears m_swipeCancellationTracker, it should set it to canceled so that
any pending -trackSwipeEventWithOptions:... which is holding a reference to the WKSwipeCancellationTracker knows
to stop.
* UIProcess/mac/ViewGestureControllerMac.mm:
(WebKit::ViewGestureController::resetState):
(WebKit::ViewGestureController::reset):
2020-05-11 Peng Liu <peng.liu6@apple.com>
Enable the mock video presentation mode in related layout tests and fix test failures
https://bugs.webkit.org/show_bug.cgi?id=211645
Reviewed by Darin Adler.
Fix the mistake in VideoFullscreenManager::supportsVideoFullscreen() introduced in r202274.
* WebProcess/cocoa/VideoFullscreenManager.mm:
(WebKit::VideoFullscreenManager::supportsVideoFullscreen const):
2020-05-11 Per Arne Vollan <pvollan@apple.com>
[iOS] Update message filtering rules in the WebContent process' sandbox
https://bugs.webkit.org/show_bug.cgi?id=211188
<rdar://problem/60922910>
Reviewed by Brent Fulgham.
Based on telemetry and local testing, update the message filtering rules in the WebContent process' sandbox on iOS.
Messages that have not been observed being in use, should be denied.
No new tests, covered by existing tests.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-05-11 Chris Dumez <cdumez@apple.com>
[iOS] WKWebView.evaluateJavaScript should run at foreground priority when app is foreground
https://bugs.webkit.org/show_bug.cgi?id=211734
Reviewed by Darin Adler.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::runJavaScriptInFrameInScriptWorld):
2020-05-11 Wenson Hsieh <wenson_hsieh@apple.com>
REGRESSION (r253267): issues on touchstart/touchend/touchmove (pointerdown/pointerup/pointermove) events
https://bugs.webkit.org/show_bug.cgi?id=211521
<rdar://problem/62942374>
Reviewed by Darin Adler.
As a brief refresher, deferring gesture recognizers allow us to handle otherwise blocking touch events
asynchronously by having all preventable native gesture recognizers require the deferring gesture recognizer to
fail; we only fail the deferring gesture recognizer once the web process has handled the touch event, and did
not call `preventDefault()`.
These additional failure requirements can cause preventable gestures to be linked together in the same gesture
dependency subgraph; since each subgraph is reset only once all gestures in the subgraph have failed or ended,
this might cause some gestures to be reset after a delay (rather than being reset immediately). To mitigate
this, we divide the set of preventable gestures into multiple (currently, 2) subgraphs: one for gestures that
are reset after a delay, and another for gestures that are immediately resettable. This way, immediately
resettable gestures are able to reset and recognize again, without having to wait for other slower preventable
gesture recognizers to reset.
When fast-clicking is disabled (e.g. when loading a desktop web page on a mobile form factor, or when the
viewport has been zoomed in), the blocking synthetic double tap gesture recognizer (that is, `WKContentView`'s
`_doubleTapGestureRecognizer`) is enabled, and adds itself as a dynamic failure requirement to the content
view's synthetic single tap gesture recognizer (`_singleTapGestureRecognizer`). In terms of the gesture
dependency graph, this causes the single tap gesture to form an edge with the double tap gesture, which ends up
uniting both deferring gesture recognizers under the same subgraph. This means UIWebTouchEventsGestureRecognizer,
which should be one of the gestures in the immediately resettable subgraph, is now connected to the rest of the
delayed resettable gestures, meaning that it cannot recognize until "slowly resettable" gestures such as the
tap-and-half text selection gesture have also been reset. This delay causes touch events to be dropped, as is
the case in this bug.
To fix this, simply quarantine the single tap and double tap gestures inside their own subgraph by introducing a
separate deferring gesture recognizer for them. When fast-clicking is enabled, this does not hinder the ability
for the single tap gesture to fire in rapid succession, since the double tap gesture is disabled (and thus, not
a part of the graph at all). when fast-clicking is disabled, then the double tap gesture will prevent the single
tap gesture from being immediately reset anyways, due to the direct failure requirement between the double and
single tap gesture.
Doing this ensures that no other immediately resettable gesture (`UIWebTouchEventsGestureRecognizer` included)
is accidentally blocked from immediately resetting due to being linked to the delayed resettable gestures by way
of the synthetic single and double tap gestures.
Test: fast/events/touch/ios/tap-and-half-when-viewport-is-not-responsive.html
* UIProcess/ios/WKContentViewInteraction.h:
Add a dedicated deferring gesture recognizer for the synthetic single tap and double tap gesture recognizers.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView setUpInteraction]):
(-[WKContentView cleanUpInteraction]):
Use -_deferringGestureRecognizers instead of hard-coding logic for each deferring gesture.
(-[WKContentView _removeDefaultGestureRecognizers]): Ditto.
(-[WKContentView _addDefaultGestureRecognizers]): Ditto.
(-[WKContentView _deferringGestureRecognizers]):
We now have 3 distinct deferring gestures; instead of handling the three deferring gestures individually in
various places in this file, group them all behind a getter that returns an array of deferring gestures, and use
this instead.
(-[WKContentView _doneDeferringNativeGestures:]): Ditto.
(-[WKContentView gestureRecognizer:shouldRecognizeSimultaneouslyWithGestureRecognizer:]): Ditto.
(-[WKContentView deferringGestureRecognizer:shouldDeferOtherGestureRecognizer:]):
Partition the synthetic single tap and double tap gestures into their own subgraph.
2020-05-11 Per Arne Vollan <pvollan@apple.com>
Unreviewed, reverting r261296.
Rolling r260769 back in, since this was not causing a
regression.
Reverted changeset:
"Unreviewed, reverting r260769."
https://bugs.webkit.org/show_bug.cgi?id=211578
https://trac.webkit.org/changeset/261296
2020-05-11 Wenson Hsieh <wenson_hsieh@apple.com>
Unreviewed, fix the watchOS build after r261457
Put -[WKActionSheetAssistant suggestedActionsForContextMenuWithPositionInformation:] behind a USE(UICONTEXTMENU)
compiler guard; currently, it is declared everywhere, but only implemented if ENABLE(DATA_DETECTION) &&
USE(UICONTEXTMENU). However, WKContentView only invokes it behind code that is guarded by USE(UICONTEXTMENU).
* UIProcess/ios/WKActionSheetAssistant.h:
* UIProcess/ios/WKActionSheetAssistant.mm:
(-[WKActionSheetAssistant suggestedActionsForContextMenuWithPositionInformation:]):
2020-05-11 Philippe Normand <pnormand@igalia.com>
[GTK] Media and audio permission request unit tests are timing out
https://bugs.webkit.org/show_bug.cgi?id=158257
Reviewed by Carlos Garcia Campos.
Add new private API to disable secure capture device contexts.
This is needed for the API tests.
* UIProcess/API/glib/WebKitWebView.cpp:
(webkitWebViewSetMediaCaptureRequiresSecureConnection):
* UIProcess/API/glib/WebKitWebViewPrivate.h:
2020-05-10 Basuke Suzuki <basuke.suzuki@sony.com>
Add ENABLE_PERIODIC_MEMORY_MONITOR flag.
https://bugs.webkit.org/show_bug.cgi?id=211704
Reviewed by Yusuke Suzuki.
No new tests because there's no behavior change.
Replace PLATFORM() macros with ENABLE() macro.
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::initializeWebProcess):
2020-05-10 Michael Catanzaro <mcatanzaro@gnome.org>
REGRESSION(r261270): Broke build with python3
https://bugs.webkit.org/show_bug.cgi?id=211699
Reviewed by Yusuke Suzuki.
Remove the problematic print statement. It causes the script to fail. If fixed, it spams the
build output. We only print messages during the build when there are problems to be fixed.
* Scripts/generate-message-receiver.py:
(main):
2020-05-10 Tim Horton <timothy_horton@apple.com>
Clicking a tel:// link on iPad with a trackpad presents different UI than tapping on it
https://bugs.webkit.org/show_bug.cgi?id=211686
<rdar://problem/57941589>
Reviewed by Wenson Hsieh.
Previously, iOS WebKit had code in handleTap, handleTwoFingerTapAtPoint, and commitPotentialTap,
which introspected the hit element /before/ dispatching the tap (or synthetic click) to the DOM,
short circuiting all normal event handling machinery in order to prevent the default action
and present the Data Detectors UI for the given link.
There's one immediate problem with this, which is that there is no similar code in the direct
click event handler path, which we now use for trackpad/mouse input on iPad. There's also a
secondary problem (which we have not seen in practice), which is that pages cannot preventDefault
to avoid launching the Data Detectors UI on a link, like they could to prevent an ordinary
link from being followed.
In order to resolve both issues and reduce duplication of logic, move the code that introspects
the element and invokes Data Detectors UI from high-level WebKit event dispatch code (where it
doesn't really belong) down to WebCore's default click event handler for anchors, right next to
the code that would ordinarily cause a navigation. This way, no matter how we get here, whether
by synthetic click or real click, we'll invoke the Data Detectors UI.
One alternative considered at length: we could actually let the navigation happen, and instead
handle the switch in navigation policy code, but there are a few reasons that did not seem ideal:
1) We'd need to plumb platform-specific data structures through a ton of FrameLoader code.
2) Clients often block non-HTTP(S) navigations, so in order to keep the feature functional,
we'd have to disregard the client's policy decision, defeating much of the purpose of making
use of the navigation mechanism in the first place.
* UIProcess/PageClient.h:
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/ios/PageClientImplIOS.h:
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::showDataDetectorsUIForPositionInformation):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::showDataDetectorsUIForPositionInformation):
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebCoreSupport/ios/WebChromeClientIOS.mm:
(WebKit::WebChromeClient::showDataDetectorsUIForElement):
Plumb showDataDetectorsUIForPositionInformation from iOS WebKit's ChromeClient
implementation all the way to WKContentView.
* UIProcess/ios/WKActionSheetAssistant.h:
* UIProcess/ios/WKActionSheetAssistant.mm:
(-[WKActionSheetAssistant showDataDetectorsUIForPositionInformation:]):
(-[WKActionSheetAssistant contextMenuInteraction:configurationForMenuAtLocation:]):
(-[WKActionSheetAssistant contextMenuInteraction:previewForHighlightingMenuWithConfiguration:]):
(-[WKActionSheetAssistant showDataDetectorsSheet]): Deleted.
Make it possible to push InteractionInformationAtPosition into WKActionSheetAssistant,
instead of having it always pull a fresh one upon invocation, so that we can use the
position information passed along with the showDataDetectorsUIForPositionInformation message.
(-[WKActionSheetAssistant _contextMenuInteraction:overrideSuggestedActionsForConfiguration:]):
(-[WKActionSheetAssistant suggestedActionsForContextMenuWithPositionInformation:]):
Fix an existing bug where Data Detectors context menus would show extra default actions,
by moving our implementation of the -_contextMenuInteraction:overrideSuggestedActionsForConfiguration:
delegate here, sharing it with WKContentViewInteraction.
In a future patch, we should reconsider having three separate UIContextMenuInteractions.
Also, remove the contextMenuPresentationLocationForActionSheetAssistant: delegate,
since it will be wrong in the case of pushed-in position information, and instead
use the position information's original hit test point.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _showDataDetectorsUI]):
(-[WKContentView _showDataDetectorsUIForPositionInformation:]):
(-[WKContentView _actionForLongPressFromPositionInformation:]):
(-[WKContentView _didNotHandleTapAsClick:]):
(-[WKContentView dataDetectionContextForPositionInformation:]):
(-[WKContentView dataDetectionContextForActionSheetAssistant:positionInformation:]):
(-[WKContentView _showDataDetectorsSheet]): Deleted.
(-[WKContentView currentPositionInformation]): Deleted.
(-[WKContentView dataDetectionContextForActionSheetAssistant:]): Deleted.
Instead of exposing an unnecessary -currentPositionInformation getter on WKActionSheetAssistant,
provide it as an argument to the WKActionSheetAssistantDelegate methods that actually need it.
(-[WKContentView contextMenuPresentationLocationForActionSheetAssistant:]): Deleted.
(-[WKContentView _contextMenuInteraction:overrideSuggestedActionsForConfiguration:]):
Adopt WKActionSheetAssistant's suggestedActions mechanism so we can share code.
* UIProcess/ios/WKPDFView.mm:
(-[WKPDFView _showActionSheetForURL:atLocation:withAnnotationRect:]):
(-[WKPDFView dataDetectionContextForActionSheetAssistant:positionInformation:]):
(-[WKPDFView dataDetectionContextForActionSheetAssistant:]): Deleted.
Adapt to some WKActionSheetAssistantDelegate changes.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::handleTap):
(WebKit::WebPage::handleTwoFingerTapAtPoint):
(WebKit::WebPage::commitPotentialTap):
Remove the aforementioned short-circuiting paths for Data Detectors links.
2020-05-09 David Kilzer <ddkilzer@apple.com>
Attempt #2 to fix internal iOS build
* UIProcess/ios/WKMouseGestureRecognizer.mm:
- Use ALLOW_DEPRECATED_IMPLEMENTATIONS_{BEGIN,END}.
2020-05-09 David Kilzer <ddkilzer@apple.com>
Attemp to fix internal iOS build
* Platform/spi/ios/UIKitSPI.h:
- Define @class UIHoverEvent using header if available, else
fall back to local class definition.
- Define methods separately from class definition.
* UIProcess/ios/WKMouseGestureRecognizer.mm:
(-[WKMouseGestureRecognizer setView:]):
- Use new method if it is available.
2020-05-09 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r261389.
https://bugs.webkit.org/show_bug.cgi?id=211678
Introduced page loading regression. (Requested by perarne on
#webkit).
Reverted changeset:
"[iOS] Update message filtering rules in the WebContent
process' sandbox"
https://bugs.webkit.org/show_bug.cgi?id=211188
https://trac.webkit.org/changeset/261389
2020-05-09 David Kilzer <ddkilzer@apple.com>
Adapt LocalCurrentGraphicsContext for iOS
<https://webkit.org/b/211660>
Reviewed by Darin Adler.
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::convertPlatformImageToBitmap):
- Use LocalCurrentGraphicsContext for both iOS and Mac code
paths.
* WebProcess/WebCoreSupport/mac/WebDragClientMac.mm:
- Switch from PLATFORM(MAC) and PLATFORM(IOS_FAMIY) to
USE(APPKIT) to separate platform code.
(WebKit::convertDragImageToBitmap):
(WebKit::convertImageToBitmap): Delete.
(WebKit::convertCGImageToBitmap): Delete.
- Combine Mac-specific convertImageToBitmap() and iOS-specific
convertCGImageToBitmap() into a single method using DragImage
type and special-casing one line of code.
(WebKit::WebDragClient::startDrag):
(WebKit::WebDragClient::didConcludeEditDrag):
- Combine iOS and Mac methods into a singe method.
2020-05-09 David Quesada <david_quesada@apple.com>
Remove HAVE_UI_SCROLL_VIEW_INDICATOR_FLASHING_SPI
https://bugs.webkit.org/show_bug.cgi?id=211662
rdar://problem/63048713
Reviewed by Darin Adler.
Remove this platform conditional and unconditionally use -[UIScrolView _flashScrollIndicatorsForAxes:
persistingPreviousFlashes:]. All supported platforms with UIScrollView have this SPI, and the conditional
was excluding Mac Catalyst, making WKKeyboardScrollViewAnimator send a message to its scroll view that
is no longer implemented.
* Platform/spi/ios/UIKitSPI.h:
Remove the declaration of the now unused -_flashScrollIndicatorsPersistingPreviousFlashes:.
* UIProcess/ios/WKKeyboardScrollingAnimator.mm:
(axesForDelta):
(-[WKKeyboardScrollViewAnimator scrollToContentOffset:animated:]):
2020-05-08 David Kilzer <ddkilzer@apple.com>
Remove empty directories from from svn.webkit.org repository
<https://webkit.org/b/211644>
Reviewed by Darin Adler.
* UIProcess/API/win: Removed.
2020-05-08 Chris Dumez <cdumez@apple.com>
Regression(r210945): No attachment when sharing to mail
https://bugs.webkit.org/show_bug.cgi?id=211651
<rdar://problem/63040170>
Reviewed by Tim Horton.
When failing to determine foreground state, the code before r210945 was assuming
'foreground', while the code in r210945 assume 'background'. This causes things
to not render properly when we fail to determine the foreground state.
This patch merely restores pre-r210945 behavior. I will investigate why we are
unable to determinate the host application PID for the MailCompositionService.
* UIProcess/ApplicationStateTracker.mm:
(WebKit::isApplicationForeground):
2020-05-08 Nikos Mouchtaris <nmouchtaris@apple.com>
Implement web-share v2 for files
https://bugs.webkit.org/show_bug.cgi?id=209265
Reviewed by Andy Estes.
Added support for passing file objects over IPC and
sharing of files to share sheet.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<Vector<RawFile>>::encode):
(IPC::ArgumentCoder<Vector<RawFile>>::decode):
(IPC::ArgumentCoder<ShareDataWithParsedURL>::encode):
(IPC::ArgumentCoder<ShareDataWithParsedURL>::decode):
* Shared/WebCoreArgumentCoders.h:
* UIProcess/Cocoa/ShareableFileWrite.h: Added.
* UIProcess/Cocoa/ShareableFileWrite.mm: Added.
(+[WKShareableFileWrite getSharingDirectoryPath]):
(+[WKShareableFileWrite getFileDirectoryForSharing]):
(+[WKShareableFileWrite removeFileDirectoryForSharing]):
(+[WKShareableFileWrite setQuarantineInformationForFilePath:]):
(+[WKShareableFileWrite applyQuarantineSandboxAndDownloadFlagsToFileAtPath:]):
(+[WKShareableFileWrite createFilename:]):
(+[WKShareableFileWrite writeFileToShareableURL:data:]):
* UIProcess/Cocoa/WKShareSheet.mm:
(-[WKShareSheet presentWithParameters:inRect:completionHandler:]):
(-[WKShareSheet _didCompleteWithSuccess:]):
* WebKit.xcodeproj/project.pbxproj:
2020-05-08 Brent Fulgham <bfulgham@apple.com>
Revise sandbox to stop logging some commonly-used properties
https://bugs.webkit.org/show_bug.cgi?id=211636
<rdar://problem/63014244>
Reviewed by Per Arne Vollan.
Further living on and testing has shown a handful of sysctl entries that are in regular use, and do not need
to be logged or considered for removal. This patch has no change in behavior, just reduces logging.
* NetworkProcess/mac/com.apple.WebKit.NetworkProcess.sb.in:
* Resources/SandboxProfiles/ios/com.apple.WebKit.Networking.sb:
2020-05-08 Chris Dumez <cdumez@apple.com>
Make WKWebViewConfiguration._alwaysRunsAtForegroundPriority an alias for _clientNavigationsRunAtForegroundPriority
https://bugs.webkit.org/show_bug.cgi?id=211632
Reviewed by Darin Adler.
Make WKWebViewConfiguration._alwaysRunsAtForegroundPriority an alias for _clientNavigationsRunAtForegroundPriority.
_alwaysRunsAtForegroundPriority is a frequent source of process assertion leaks that get blamed on WebKit. Clients
are using this in order to be able to do loads in background views so _clientNavigationsRunAtForegroundPriority
should suffice and is not leak prone since the process assertion is released after the navigation has completed.
* UIProcess/API/APIPageConfiguration.cpp:
(API::PageConfiguration::copy const):
* UIProcess/API/APIPageConfiguration.h:
(API::PageConfiguration::alwaysRunsAtForegroundPriority const): Deleted.
(API::PageConfiguration::setAlwaysRunsAtForegroundPriority): Deleted.
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration _alwaysRunsAtForegroundPriority]):
(-[WKWebViewConfiguration _setAlwaysRunsAtForegroundPriority:]):
* UIProcess/API/Cocoa/WKWebViewConfigurationPrivate.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::close):
(WebKit::WebPageProxy::updateThrottleState):
(WebKit::WebPageProxy::resetStateAfterProcessExited):
* UIProcess/WebPageProxy.h:
2020-05-08 Alex Christensen <achristensen@webkit.org>
Revert r260390
https://bugs.webkit.org/show_bug.cgi?id=210759
<rdar://problem/62608140>
r260390 was intended to stop hitting some assertions in ~PendingFrameLoad, but it seems to have caused a regression.
Reverting it and removing the invalid assertions.
* NetworkProcess/cache/NetworkCacheSpeculativeLoadManager.cpp:
(WebKit::NetworkCache::SpeculativeLoadManager::PendingFrameLoad::~PendingFrameLoad):
(WebKit::NetworkCache::SpeculativeLoadManager::~SpeculativeLoadManager):
2020-05-08 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] caret appears in the middle of a search field when field is focused on agoda.com
https://bugs.webkit.org/show_bug.cgi?id=211591
<rdar://problem/60605873>
Reviewed by Antoine Quint.
The main search field on the mobile version of this website begins offscreen, with a CSS transform that moves it
to the far right; tapping the button element that (visually) has a search-field-like appearance on the front
page programmatically focuses the real offscreen search field, and animates it onscreen by changing the CSS
transform attribute to remove the x-axis translation.
On iOS, the caret rect is computed and sent to the UI process via editor state updates; however, the editor
state is computed immediately after focusing the input field. As such, the caret rect at this moment is computed
in the middle of the animation, leaving it stuck in an unpredictable location.
To fix this, add plumbing to call into the WebKit client layer when an animation has ended. On iOS, if the
selection is visible (i.e. a ranged selection, or editable caret), then check to see whether the element that
has finished animating contains either endpoint of the selection. If so, then schedule a followup editor state
update to push updated selection information to the UI process.
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::animationDidFinishForElement):
* WebProcess/WebCoreSupport/WebChromeClient.h:
Add a new client hook for when animations end.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::animationDidFinishForElement):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/ios/WebPageIOS.mm:
Add logic to schedule a new editor state update if needed, after an animation ends that might affect either
the start or end of the selection.
(WebKit::WebPage::animationDidFinishForElement):
2020-05-08 David Kilzer <ddkilzer@apple.com>
REGRESSION (r260228): Linker warning about limitsNavigationsToAppBoundDomains property overriding instance methods from class
<https://webkit.org/b/211393>
Reviewed by Brent Fulgham.
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration limitsNavigationsToAppBoundDomains]):
(-[WKWebViewConfiguration setLimitsNavigationsToAppBoundDomains:]):
- Move from WKPrivate category to main category to fix linker
warnings.
2020-05-08 Chris Dumez <cdumez@apple.com>
REGRESSION(r259209) Webview's pending URL is null after restoring session state
https://bugs.webkit.org/show_bug.cgi?id=211626
<rdar://problem/62992262>
Reviewed by Alex Christensen.
The issue was that WebPageProxy::goToBackForwardItem() would behave differently whether
the page has a running process or not. In particular, when the page did not have a
running process, goToBackForwardItem() would return early and call launchProcessWithItem()
instead. Unlike goToBackForwardItem(), launchProcessWithItem() would fail to set the
pending API request.
To address the issue, I am getting rid of launchProcessWithItem() and merging its logic
into goToBackForwardItem() instead. Both methods shared a lot of code anyway and having
2 separate code paths that may diverge is error prone.
Change is covered by new API test.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::goToBackForwardItem):
(WebKit::WebPageProxy::launchProcessWithItem): Deleted.
* UIProcess/WebPageProxy.h:
2020-05-08 Alex Christensen <achristensen@webkit.org>
WKWebView.title should be safe browsing warning's title during a safe browsing warning
https://bugs.webkit.org/show_bug.cgi?id=211403
Reviewed by Tim Horton.
Add a new property to PageLoadState, which allows the title to be set underneath it.
That way, once the safe browsing warning is closed, we can update the title to what it would've been,
including any changes that happened while the warning was open.
Covered by an API test. I also manually verified this fixes rdar://problem/56201982
* UIProcess/PageLoadState.cpp:
(WebKit::PageLoadState::commitChanges):
(WebKit::PageLoadState::reset):
(WebKit::PageLoadState::didCommitLoad):
(WebKit::PageLoadState::title const):
(WebKit::PageLoadState::setTitleFromSafeBrowsingWarning):
* UIProcess/PageLoadState.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::decidePolicyForNavigationAction):
2020-05-08 Per Arne Vollan <pvollan@apple.com>
[iOS] Update message filtering rules in the WebContent process' sandbox
https://bugs.webkit.org/show_bug.cgi?id=211188
Reviewed by Brent Fulgham.
Based on telemetry and local testing, update the message filtering rules in the WebContent process' sandbox on iOS.
Messages that have not been observed being in use should be denied.
No new tests, covered by existing tests.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-05-08 Chris Dumez <cdumez@apple.com>
Turn processDidBecomeUnresponsive logging into error logging
https://bugs.webkit.org/show_bug.cgi?id=211629
Reviewed by Youenn Fablet.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::processDidBecomeUnresponsive):
2020-05-08 Per Arne Vollan <pvollan@apple.com>
[Cocoa] Close connections to Launch Services after revoking extension
https://bugs.webkit.org/show_bug.cgi?id=211594
Reviewed by Brent Fulgham.
In <https://trac.webkit.org/changeset/258915/webkit>, connections to the Launch Services database mapping
services were denied, by creating a temporary extension. After revoking this extension in the WebContent
process, it is important to close any open connections to Launch Services initiated by the mapping.
No new tests, covered by existing UTI/MIME type tests.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-05-08 Youenn Fablet <youenn@apple.com>
Handle remote audio capture IPC messages in a background thread
https://bugs.webkit.org/show_bug.cgi?id=211583
Reviewed by Eric Carlson.
We introduce RemoteCaptureSampleManager which is responsible to get remote capture sample IPC messages and treat them from a background thread.
It is currently used to handle audio capture IPC messages.
We move UserMediaCaptureManager::Source in its own file and rename it to RemoteRealtimeMediaSource.
Covered by existing tests.
* DerivedSources-input.xcfilelist:
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* PlatformMac.cmake:
* SourcesCocoa.txt:
* UIProcess/Cocoa/UserMediaCaptureManagerProxy.cpp:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/cocoa/RemoteCaptureSampleManager.cpp: Added.
(WebKit::RemoteCaptureSampleManager::RemoteCaptureSampleManager):
(WebKit::RemoteCaptureSampleManager::~RemoteCaptureSampleManager):
(WebKit::RemoteCaptureSampleManager::setConnection):
(WebKit::RemoteCaptureSampleManager::addSource):
(WebKit::RemoteCaptureSampleManager::removeSource):
(WebKit::RemoteCaptureSampleManager::dispatchToThread):
(WebKit::RemoteCaptureSampleManager::storageChanged):
(WebKit::RemoteCaptureSampleManager::audioSamplesAvailable):
* WebProcess/cocoa/RemoteCaptureSampleManager.h: Added.
* WebProcess/cocoa/RemoteCaptureSampleManager.messages.in: Added.
* WebProcess/cocoa/RemoteRealtimeMediaSource.cpp: Added.
(WebKit::RemoteRealtimeMediaSource::RemoteRealtimeMediaSource):
(WebKit::RemoteRealtimeMediaSource::~RemoteRealtimeMediaSource):
(WebKit::RemoteRealtimeMediaSource::whenReady):
(WebKit::RemoteRealtimeMediaSource::didFail):
(WebKit::RemoteRealtimeMediaSource::setAsReady):
(WebKit::RemoteRealtimeMediaSource::clone):
(WebKit::RemoteRealtimeMediaSource::setCapabilities):
(WebKit::RemoteRealtimeMediaSource::setSettings):
(WebKit::RemoteRealtimeMediaSource::remoteAudioSamplesAvailable):
(WebKit::RemoteRealtimeMediaSource::remoteVideoSampleAvailable):
(WebKit::RemoteRealtimeMediaSource::connection):
(WebKit::RemoteRealtimeMediaSource::startProducingData):
(WebKit::RemoteRealtimeMediaSource::stopProducingData):
(WebKit::RemoteRealtimeMediaSource::setShouldApplyRotation):
(WebKit::RemoteRealtimeMediaSource::capabilities):
(WebKit::RemoteRealtimeMediaSource::applyConstraints):
(WebKit::RemoteRealtimeMediaSource::applyConstraintsSucceeded):
(WebKit::RemoteRealtimeMediaSource::applyConstraintsFailed):
(WebKit::RemoteRealtimeMediaSource::hasEnded):
(WebKit::RemoteRealtimeMediaSource::stopBeingObserved):
(WebKit::RemoteRealtimeMediaSource::requestToEnd):
* WebProcess/cocoa/RemoteRealtimeMediaSource.h: Added.
(WebKit::RemoteRealtimeMediaSource::identifier const):
(WebKit::RemoteRealtimeMediaSource::settings const):
(WebKit::RemoteRealtimeMediaSource::setShouldCaptureInGPUProcess):
(WebKit::RemoteRealtimeMediaSource::shouldCaptureInGPUProcess const):
* WebProcess/cocoa/UserMediaCaptureManager.cpp:
(WebKit::UserMediaCaptureManager::createCaptureSource):
(WebKit::UserMediaCaptureManager::cloneSource):
(WebKit::UserMediaCaptureManager::cloneVideoSource):
(WebKit::UserMediaCaptureManager::AudioFactory::createAudioCaptureSource):
(WebKit::UserMediaCaptureManager::AudioFactory::setShouldCaptureInGPUProcess):
(WebKit::UserMediaCaptureManager::VideoFactory::createVideoCaptureSource):
(WebKit::UserMediaCaptureManager::DisplayFactory::createDisplayCaptureSource):
* WebProcess/cocoa/UserMediaCaptureManager.h:
* WebProcess/cocoa/UserMediaCaptureManager.messages.in:
2020-05-08 Youenn Fablet <youenn@apple.com>
Video capture does not get unmuted in case of tab switch on iOS
https://bugs.webkit.org/show_bug.cgi?id=211509
Reviewed by Eric Carlson.
Remove no longer needed code.
* WebProcess/cocoa/UserMediaCaptureManager.cpp:
(WebKit::UserMediaCaptureManager::VideoFactory::setActiveSource):
* WebProcess/cocoa/UserMediaCaptureManager.h:
2020-05-08 Carlos Garcia Campos <cgarcia@igalia.com>
Unreviewed. Fix GTK4 build after r261370
* UIProcess/API/gtk/WebKitScriptDialogGtk.cpp:
(webkitScriptDialogIsUserHandled):
2020-05-08 Lauro Moura <lmoura@igalia.com>
[WPE][WebDriver] Implement user prompt support
https://bugs.webkit.org/show_bug.cgi?id=207372
Reviewed by Carlos Garcia Campos.
Provide a default implementation for WebDriver dialog commands.
This implementation will mimic the expected browser behavior of
accepting/dismissing the dialogs, while not exposing any new API.
* UIProcess/API/glib/WebKitScriptDialogPrivate.h:
- Expose new private function to check whether the dialog is using a
default implementation or is handled by the user.
* UIProcess/API/glib/WebKitWebView.cpp:
- Avoid default WebDriver behavior if the dialog is user-handled.
(webkitWebViewSetCurrentScriptDialogUserInput):
(webkitWebViewAcceptCurrentScriptDialog):
(webkitWebViewDismissCurrentScriptDialog):
* UIProcess/API/gtk/WebKitScriptDialogGtk.cpp:
(webkitScriptDialogIsUserHandled):
- Return whether the dialog is user handled.
* UIProcess/API/wpe/WebKitScriptDialogWPE.cpp:
- Added basic behavior for the WebDriver dialog callbacks for WPE.
(webkitScriptDialogAccept):
(webkitScriptDialogDismiss):
(webkitScriptDialogSetUserInput):
(webkitScriptDialogIsUserHandled):
* UIProcess/API/wpe/WebKitWebViewWPE.cpp:
(webkitWebViewScriptDialog):
- Hold a dialog ref if the webview is controlled by automation and ends
up calling the default "script-dialog" event handler.
2020-05-07 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Roll back newly created credentials if an error occurs
https://bugs.webkit.org/show_bug.cgi?id=183530
<rdar://problem/43357305>
Reviewed by Brent Fulgham.
We should clean up any newly created credentials if an error occurs before the relying party
registers the identity. Otherwise we are left with a dangling credential.
Covered by API tests.
* UIProcess/WebAuthentication/Cocoa/LocalAuthenticator.h:
* UIProcess/WebAuthentication/Cocoa/LocalAuthenticator.mm:
(WebKit::LocalAuthenticator::continueMakeCredentialAfterUserVerification):
(WebKit::LocalAuthenticator::continueGetAssertionAfterUserVerification):
(WebKit::LocalAuthenticator::receiveException const):
2020-05-07 Chris Dumez <cdumez@apple.com>
Unreviewed, reverting r261252.
Reland r260684 now that a proper fix has landed in Reader
Reverted changeset:
"REGRESSION (r260684): Reader background is lost after
multitasking"
https://bugs.webkit.org/show_bug.cgi?id=211533
https://trac.webkit.org/changeset/261252
2020-05-07 Megan Gardner <megan_gardner@apple.com>
Build Fix for MacCatalyst
https://bugs.webkit.org/show_bug.cgi?id=211601
Reviewed by Tim Horton.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _removeContextMenuViewIfPossible]):
2020-05-07 Chris Dumez <cdumez@apple.com>
Crash under WebKit::XPCServiceMain
https://bugs.webkit.org/show_bug.cgi?id=211600
<rdar://problem/62875458>
Reviewed by Darin Adler.
Made the following changes:
- Crash under strcmp() could in theory happen if expectedBundleVersion.UTF8String was null, which could
happen if expectedBundleVersion was null. I now use higher level String types for the versions, make
sure they are not null and use String comparison to compare them.
- Call CRASH() instead of __builtin_trap(), to guarantee that we crash with a SIGTRAP (see comment next
to CRASH() macro).
- Move CRASH() call to a crashDueWebKitFrameworkVersionMismatch() never inlined function to make it
clear when it crashes due to a framework version mismatch and distinguish from proper crashes.
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceMain.mm:
(WebKit::crashDueWebKitFrameworkVersionMismatch):
(WebKit::XPCServiceMain):
2020-05-07 Youenn Fablet <youenn@apple.com>
Remove AudioMediaStreamTrackRenderer::muted
https://bugs.webkit.org/show_bug.cgi?id=211289
Reviewed by Eric Carlson.
muted is unnecessary since we are using start/stop instead.
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRenderer.cpp:
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRenderer.h:
* GPUProcess/webrtc/RemoteAudioMediaStreamTrackRenderer.messages.in:
* WebProcess/GPU/webrtc/AudioMediaStreamTrackRenderer.cpp:
* WebProcess/GPU/webrtc/AudioMediaStreamTrackRenderer.h:
2020-05-07 Don Olmstead <don.olmstead@sony.com>
Remove unused USE(REQUEST_ANIMATION_FRAME_DISPLAY_MONITOR)
https://bugs.webkit.org/show_bug.cgi?id=211582
Reviewed by Fujii Hironori.
After r261264 all ports implemented USE_REQUEST_ANIMATION_FRAME_DISPLAY_MONITOR.
* Shared/CoordinatedGraphics/threadedcompositor/ThreadedCompositor.cpp:
(WebKit::ThreadedCompositor::ThreadedCompositor):
(WebKit::m_displayRefreshMonitor):
(WebKit::ThreadedCompositor::invalidate):
(WebKit::ThreadedCompositor::sceneUpdateFinished):
(WebKit::ThreadedCompositor::displayRefreshMonitor):
* Shared/CoordinatedGraphics/threadedcompositor/ThreadedCompositor.h:
* Shared/CoordinatedGraphics/threadedcompositor/ThreadedDisplayRefreshMonitor.cpp:
* Shared/CoordinatedGraphics/threadedcompositor/ThreadedDisplayRefreshMonitor.h:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::createDisplayRefreshMonitor const):
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebPage/CoordinatedGraphics/DrawingAreaCoordinatedGraphics.cpp:
(WebKit::DrawingAreaCoordinatedGraphics::createDisplayRefreshMonitor):
* WebProcess/WebPage/CoordinatedGraphics/DrawingAreaCoordinatedGraphics.h:
* WebProcess/WebPage/CoordinatedGraphics/LayerTreeHost.cpp:
(WebKit::LayerTreeHost::createDisplayRefreshMonitor):
* WebProcess/WebPage/CoordinatedGraphics/LayerTreeHost.h:
(WebKit::LayerTreeHost::deviceOrPageScaleFactorChanged):
* WebProcess/WebPage/DrawingArea.cpp:
* WebProcess/WebPage/DrawingArea.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDisplayRefreshMonitor.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDisplayRefreshMonitor.mm:
* WebProcess/WebPage/mac/DrawingAreaMac.cpp:
2020-05-07 Darin Adler <darin@apple.com>
Add some missing null checks for DocumentLoader
https://bugs.webkit.org/show_bug.cgi?id=211544
rdar://62843516
Reviewed by Anders Carlsson.
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::transitionToCommittedForNewPage):
Add a null check before calling DocumentLoader::response.
2020-05-07 Darin Adler <darin@apple.com>
Remove USE(INSERTION_UNDO_GROUPING) checks in macOS platform code
https://bugs.webkit.org/show_bug.cgi?id=211525
Reviewed by Anders Carlsson.
* UIProcess/Cocoa/WebViewImpl.mm: Use NSTextInputContextSPI.h.
(WebKit::WebViewImpl::validAttributesForMarkedText): Removed
USE(INSERTION_UNDO_GROUPING) from the macOS-specific code here.
(WebKit::WebViewImpl::insertText): Ditto.
* UIProcess/PageClient.h: Ditto.
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::registerInsertionUndoGrouping): Deleted.
This function is not needed at all on iOS.
* UIProcess/mac/PageClientImplMac.h: Removed USE(INSERTION_UNDO_GROUPING).
* UIProcess/mac/PageClientImplMac.mm:
(WebKit::PageClientImpl::registerInsertionUndoGrouping): Ditto.
2020-05-07 Chris Dumez <cdumez@apple.com>
Drop dead platformPrepareToSuspend / platformProcessDidResume in NetworkProcess
https://bugs.webkit.org/show_bug.cgi?id=211579
Reviewed by Geoffrey Garen.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::prepareToSuspend):
(WebKit::NetworkProcess::resume):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/cocoa/NetworkProcessCocoa.mm:
(WebKit::NetworkProcess::platformPrepareToSuspend): Deleted.
(WebKit::NetworkProcess::platformProcessDidResume): Deleted.
* NetworkProcess/curl/NetworkProcessCurl.cpp:
(WebKit::NetworkProcess::platformPrepareToSuspend): Deleted.
(WebKit::NetworkProcess::platformProcessDidResume): Deleted.
* NetworkProcess/soup/NetworkProcessSoup.cpp:
(WebKit::NetworkProcess::platformPrepareToSuspend): Deleted.
(WebKit::NetworkProcess::platformProcessDidResume): Deleted.
2020-05-07 Megan Gardner <megan_gardner@apple.com>
ContextMenu: entire webpage pops and disappears to generate preview of embedded image
https://bugs.webkit.org/show_bug.cgi?id=211272
<rdar://problem/62482013>
Reviewed by Tim Horton.
The ActionSheetAssistant is always around, so using the presence of one
to determine if we need to remove the _contextMenuHintContainerView is incorrect
and causes it to be around on the next invocation of the context menu, and
causing the failure of the creation of a new preview, which then will default to
popping the whole page. We need to ask the ActionSheetAssistant if it is actively showing
anything, and then correctly clean up the _contextMenuHintContainerView after the
end of the interactions for FileUploadMenu and ActionSheetAssistant are done with their
interactions.
* UIProcess/ios/WKActionSheetAssistant.h:
* UIProcess/ios/WKActionSheetAssistant.mm:
(-[WKActionSheetAssistant removeContextMenuInteraction]):
(-[WKActionSheetAssistant hasContextMenuInteraction]):
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView removeContextMenuViewIfPossibleForActionSheetAssistant:]):
(-[WKContentView _canRemoveContextMenuView]):
(-[WKContentView _removeContextMenuViewIfPossible]):
(-[WKContentView contextMenuInteraction:willEndForConfiguration:animator:]):
* UIProcess/ios/forms/WKFileUploadPanel.mm:
(-[WKFileUploadPanel removeContextMenuInteraction]):
2020-05-07 Chris Dumez <cdumez@apple.com>
Simplify several IPC sending call sites in the UIProcess
https://bugs.webkit.org/show_bug.cgi?id=211575
Reviewed by Alex Christensen.
Simplify several IPC sending call sites in the UIProcess by leveraging the fact
that WebPageProxy subclasses IPC::MessageSender.
* UIProcess/Cocoa/TextCheckingController.mm:
(WebKit::TextCheckingController::replaceRelativeToSelection):
(WebKit::TextCheckingController::removeAnnotationRelativeToSelection):
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::performDictionaryLookupAtLocation):
(WebKit::WebPageProxy::performDictionaryLookupOfCurrentSelection):
(WebKit::WebPageProxy::insertDictatedTextAsync):
(WebKit::WebPageProxy::speakingErrorOccurred):
(WebKit::WebPageProxy::boundaryEventOccurred):
(WebKit::WebPageProxy::voicesDidChange):
* UIProcess/GeolocationPermissionRequestManagerProxy.cpp:
(WebKit::GeolocationPermissionRequestManagerProxy::didReceiveGeolocationPermissionDecision):
* UIProcess/Inspector/RemoteWebInspectorProxy.cpp:
(WebKit::RemoteWebInspectorProxy::setDiagnosticLoggingAvailable):
* UIProcess/Inspector/WebInspectorProxy.cpp:
(WebKit::WebInspectorProxy::setDiagnosticLoggingAvailable):
* UIProcess/Inspector/gtk/RemoteWebInspectorProxyGtk.cpp:
(WebKit::remoteFileReplaceContentsCallback):
* UIProcess/Inspector/gtk/WebInspectorProxyGtk.cpp:
(WebKit::fileReplaceContentsCallback):
* UIProcess/Inspector/mac/RemoteWebInspectorProxyMac.mm:
(WebKit::RemoteWebInspectorProxy::platformSave):
(WebKit::RemoteWebInspectorProxy::platformAppend):
* UIProcess/Inspector/mac/WebInspectorProxyMac.mm:
(WebKit::WebInspectorProxy::platformSave):
(WebKit::WebInspectorProxy::platformAppend):
* UIProcess/Notifications/NotificationPermissionRequestManagerProxy.cpp:
(WebKit::NotificationPermissionRequestManagerProxy::createRequest):
* UIProcess/UserMediaPermissionRequestManagerProxy.cpp:
(WebKit::UserMediaPermissionRequestManagerProxy::captureDevicesChanged):
(WebKit::UserMediaPermissionRequestManagerProxy::denyRequest):
* UIProcess/WebFrameProxy.cpp:
(WebKit::WebFrameProxy::collapseSelection):
* UIProcess/WebFullScreenManagerProxy.cpp:
(WebKit::WebFullScreenManagerProxy::willEnterFullScreen):
(WebKit::WebFullScreenManagerProxy::didEnterFullScreen):
(WebKit::WebFullScreenManagerProxy::willExitFullScreen):
(WebKit::WebFullScreenManagerProxy::didExitFullScreen):
(WebKit::WebFullScreenManagerProxy::setAnimatingFullScreen):
(WebKit::WebFullScreenManagerProxy::requestExitFullScreen):
(WebKit::WebFullScreenManagerProxy::saveScrollPosition):
(WebKit::WebFullScreenManagerProxy::restoreScrollPosition):
(WebKit::WebFullScreenManagerProxy::setFullscreenInsets):
(WebKit::WebFullScreenManagerProxy::setFullscreenAutoHideDuration):
(WebKit::WebFullScreenManagerProxy::setFullscreenControlsHidden):
* UIProcess/gtk/WebPageProxyGtk.cpp:
(WebKit::WebPageProxy::getCenterForZoomGesture):
* UIProcess/ios/SmartMagnificationController.mm:
(WebKit::SmartMagnificationController::handleSmartMagnificationGesture):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView ensurePositionInformationIsUpToDate:]):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::handleTwoFingerTapAtPoint):
(WebKit::WebPageProxy::handleStylusSingleTapAtPoint):
(WebKit::WebPageProxy::registerUIProcessAccessibilityTokens):
(WebKit::WebPageProxy::willStartUserTriggeredZooming):
(WebKit::WebPageProxy::potentialTapAtPosition):
(WebKit::WebPageProxy::commitPotentialTap):
(WebKit::WebPageProxy::cancelPotentialTap):
(WebKit::WebPageProxy::tapHighlightAtPosition):
(WebKit::WebPageProxy::handleTap):
(WebKit::WebPageProxy::didRecognizeLongPress):
(WebKit::WebPageProxy::handleDoubleTapForDoubleClickAtPoint):
(WebKit::WebPageProxy::inspectorNodeSearchMovedToPosition):
(WebKit::WebPageProxy::inspectorNodeSearchEndedAtPosition):
(WebKit::WebPageProxy::blurFocusedElement):
(WebKit::WebPageProxy::setIsShowingInputViewForFocusedElement):
(WebKit::WebPageProxy::setFocusedElementValue):
(WebKit::WebPageProxy::setFocusedElementValueAsNumber):
(WebKit::WebPageProxy::setFocusedElementSelectedIndex):
(WebKit::WebPageProxy::computePagesForPrintingAndDrawToPDF):
(WebKit::WebPageProxy::contentSizeCategoryDidChange):
(WebKit::WebPageProxy::generateSyntheticEditingCommand):
* UIProcess/mac/ViewGestureControllerMac.mm:
(WebKit::ViewGestureController::handleMagnificationGestureEvent):
(WebKit::ViewGestureController::handleSmartMagnificationGesture):
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::windowAndViewFramesChanged):
(WebKit::WebPageProxy::setMainFrameIsScrollable):
(WebKit::WebPageProxy::attributedSubstringForCharacterRangeAsync):
(WebKit::WebPageProxy::fontAtSelection):
(WebKit::WebPageProxy::stringSelectionForPasteboard):
(WebKit::WebPageProxy::dataSelectionForPasteboard):
(WebKit::WebPageProxy::readSelectionFromPasteboard):
(WebKit::WebPageProxy::replaceSelectionWithPasteboardData):
(WebKit::WebPageProxy::sendComplexTextInputToPlugin):
(WebKit::WebPageProxy::uppercaseWord):
(WebKit::WebPageProxy::lowercaseWord):
(WebKit::WebPageProxy::capitalizeWord):
(WebKit::WebPageProxy::setSmartInsertDeleteEnabled):
(WebKit::WebPageProxy::registerUIProcessAccessibilityTokens):
(WebKit::WebPageProxy::shouldDelayWindowOrderingForEvent):
(WebKit::WebPageProxy::acceptsFirstMouse):
2020-05-07 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r260769.
https://bugs.webkit.org/show_bug.cgi?id=211578
Introduced regressions related to sharing (Requested by
perarne on #webkit).
Reverted changeset:
"[Cocoa] After r258891, r255119 can be reverted"
https://bugs.webkit.org/show_bug.cgi?id=211083
https://trac.webkit.org/changeset/260769
2020-05-07 Chris Dumez <cdumez@apple.com>
[iOS] AuxiliaryProcessProxy::sendWithAsyncReply() should prevent auxiliary process suspension while processing the IPC
https://bugs.webkit.org/show_bug.cgi?id=209928
Reviewed by Youenn Fablet.
AuxiliaryProcessProxy::sendWithAsyncReply() should prevent auxiliary process suspension
on iOS while processing the IPC. If the process is suspended either before or during
the IPC, this could result in hangs. This is a speculative fix for the flaky timeouts
we see on many tests on iOS only. We know that missing activities with
sendWithAsyncReply has caused test timeouts in the past (Bug 211421).
* UIProcess/AuxiliaryProcessProxy.cpp:
(WebKit::AuxiliaryProcessProxy::sendMessage):
* UIProcess/AuxiliaryProcessProxy.h:
(WebKit::AuxiliaryProcessProxy::sendWithAsyncReply):
* UIProcess/GPU/GPUProcessProxy.cpp:
(WebKit::GPUProcessProxy::getGPUProcessConnection):
* UIProcess/GPU/GPUProcessProxy.h:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::getNetworkProcessConnection):
(WebKit::NetworkProcessProxy::setUseITPDatabase):
(WebKit::NetworkProcessProxy::resetCacheMaxAgeCapForPrevalentResources):
(WebKit::NetworkProcessProxy::resetParametersToDefaultValues):
(WebKit::NetworkProcessProxy::scheduleClearInMemoryAndPersistent):
(WebKit::NetworkProcessProxy::resetCrossSiteLoadsWithLinkDecorationForTesting):
(WebKit::NetworkProcessProxy::setAppBoundDomainsForResourceLoadStatistics):
(WebKit::NetworkProcessProxy::setShouldDowngradeReferrerForTesting):
(WebKit::NetworkProcessProxy::setThirdPartyCookieBlockingMode):
(WebKit::NetworkProcessProxy::setShouldEnbleSameSiteStrictEnforcementForTesting):
(WebKit::NetworkProcessProxy::setFirstPartyWebsiteDataRemovalModeForTesting):
(WebKit::NetworkProcessProxy::sendPrepareToSuspend):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/Plugins/PluginProcessProxy.cpp:
(WebKit::PluginProcessProxy::PluginProcessProxy):
* UIProcess/Plugins/PluginProcessProxy.h:
* UIProcess/ProcessThrottler.h:
(WebKit::ProcessThrottler::Activity::Activity):
(WebKit::ProcessThrottler::Activity::isQuietActivity const):
(WebKit::ProcessThrottler::Activity::invalidate):
* UIProcess/UserMediaPermissionRequestManagerProxy.cpp:
(WebKit::UserMediaPermissionRequestManagerProxy::finishGrantingRequest):
* UIProcess/WebBackForwardCacheEntry.cpp:
(WebKit::WebBackForwardCacheEntry::~WebBackForwardCacheEntry):
* UIProcess/WebCookieManagerProxy.cpp:
(WebKit::WebCookieManagerProxy::getHostnamesWithCookies):
(WebKit::WebCookieManagerProxy::deleteCookie):
(WebKit::WebCookieManagerProxy::deleteAllCookiesModifiedSince):
(WebKit::WebCookieManagerProxy::setCookies):
(WebKit::WebCookieManagerProxy::getAllCookies):
(WebKit::WebCookieManagerProxy::getCookies):
(WebKit::WebCookieManagerProxy::setHTTPCookieAcceptPolicy):
(WebKit::WebCookieManagerProxy::getHTTPCookieAcceptPolicy):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::getProcessDisplayName):
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::setThirdPartyCookieBlockingMode):
(WebKit::WebProcessProxy::fetchWebsiteData):
(WebKit::WebProcessProxy::deleteWebsiteData):
(WebKit::WebProcessProxy::deleteWebsiteDataForOrigins):
(WebKit::WebProcessProxy::sendPrepareToSuspend):
(WebKit::WebProcessProxy::activePagesDomainsForTesting):
(WebKit::WebProcessProxy::establishServiceWorkerContext):
* UIProcess/WebProcessProxy.h:
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::insertTextPlaceholder):
(WebKit::WebPageProxy::removeTextPlaceholder):
(WebKit::WebPageProxy::requestAutocorrectionData):
(WebKit::WebPageProxy::requestEvasionRectsAboveSelection):
(WebKit::WebPageProxy::updateSelectionWithDelta):
(WebKit::WebPageProxy::requestDocumentEditingContext):
(WebKit::WebPageProxy::insertDroppedImagePlaceholders):
2020-05-07 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] REGRESSION(r260996): Pressing <Tab> in an HTTP authentication dialog makes the dialog lose focus
https://bugs.webkit.org/show_bug.cgi?id=211555
Reviewed by Philippe Normand.
Bring back webkitWebViewBaseFocus to GTK3. I made it GTK4 only by mistake in r260996.
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseFocus):
(webkit_web_view_base_class_init):
2020-05-07 Philippe Normand <pnormand@igalia.com>
[GTK] Build warning fixes
Rubber-stamped by Žan Doberšek.
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseGrabFocus):
* UIProcess/GPU/GPUProcessProxy.cpp:
(WebKit::GPUProcessProxy::getGPUProcessConnection):
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::getNetworkProcessConnection):
2020-05-07 Youenn Fablet <youenn@apple.com>
Sending WebRTC network packets should not go through the main thread
https://bugs.webkit.org/show_bug.cgi?id=211291
Reviewed by Eric Carlson.
Following on receiving RTC packets from a background thread, we also send RTC packets from a background thread.
Creation of the sockets also happens in a background thread.
LibWebRTCNetwork is getting the connection whenever a new connection to network process is created.
It will then hop to the RTC network thread to set the IPC connection to the libwebrtc socket factory.
At creation of the socket, we get the IPC connection to the network process and keep a ref in the RTC socket.
In case network process crashed and the IPC connection of the RTC network is null, we hop to the main thread to create a new IPC connection.
This will fail the creation of the socket (as well as new ones as well) as long as the IPC connection to network process is not valid again.
Covered by existing tests.
* WebProcess/Network/webrtc/LibWebRTCNetwork.cpp:
(WebKit::LibWebRTCNetwork::setAsActive):
(WebKit::LibWebRTCNetwork::setConnection):
(WebKit::LibWebRTCNetwork::dispatchToThread):
* WebProcess/Network/webrtc/LibWebRTCNetwork.h:
(WebKit::LibWebRTCNetwork::connection):
(WebKit::LibWebRTCNetwork::isActive const):
* WebProcess/Network/webrtc/LibWebRTCProvider.cpp:
(WebKit::LibWebRTCProvider::startedNetworkThread):
* WebProcess/Network/webrtc/LibWebRTCProvider.h:
* WebProcess/Network/webrtc/LibWebRTCSocket.cpp:
(WebKit::LibWebRTCSocket::SendTo):
(WebKit::LibWebRTCSocket::Close):
(WebKit::LibWebRTCSocket::SetOption):
(WebKit::LibWebRTCSocket::suspend):
* WebProcess/Network/webrtc/LibWebRTCSocket.h:
* WebProcess/Network/webrtc/LibWebRTCSocketFactory.cpp:
(WebKit::LibWebRTCSocketFactory::setConnection):
(WebKit::LibWebRTCSocketFactory::connection):
(WebKit::LibWebRTCSocketFactory::createServerTcpSocket):
(WebKit::LibWebRTCSocketFactory::createUdpSocket):
(WebKit::LibWebRTCSocketFactory::createClientTcpSocket):
(WebKit::LibWebRTCSocketFactory::createNewConnectionSocket):
(WebKit::LibWebRTCSocketFactory::addSocket):
(WebKit::LibWebRTCSocketFactory::removeSocket):
(WebKit::LibWebRTCSocketFactory::forSocketInGroup):
* WebProcess/Network/webrtc/LibWebRTCSocketFactory.h:
2020-05-07 Adrian Perez de Castro <aperez@igalia.com>
[GTK4] Stop using the GtkWidget.destroy vfunc
https://bugs.webkit.org/show_bug.cgi?id=211553
Reviewed by Carlos Garcia Campos.
No new tests needed.
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseDestroy): Removed.
(webkitWebViewBaseDispose): Bring in the code from the removed function, adapt
to use g_clear_pointer() to destroy the nested dialog.
(webkit_web_view_base_class_init): Do not set the GtkWidget.destroy vfunc.
2020-05-07 Philippe Normand <pnormand@igalia.com>
[WPE][Qt] Enable by default
https://bugs.webkit.org/show_bug.cgi?id=211517
Reviewed by Adrian Perez de Castro.
* PlatformWPE.cmake: Fix WPEQt build.
2020-05-06 Alex Christensen <achristensen@webkit.org>
Fix webkitpy tests after r261254
https://bugs.webkit.org/show_bug.cgi?id=211112
test.messages.in was accidentally deleted instead of moved, and messages_unittest.py needed to be told of the new locations of results.
The rest is just running the generator to update results.
* Scripts/MessageNames.cpp:
(IPC::description):
(IPC::receiverName):
(IPC::isValidMessageName):
* Scripts/MessageNames.h:
* Scripts/testMessageReceiver.cpp:
(Messages::WebPage::GetPluginProcessConnection::send):
(Messages::WebPage::TestMultipleAttributes::send):
(WebKit::WebPage::didReceiveMessage):
(WebKit::WebPage::didReceiveSyncMessage):
* Scripts/testMessages.h:
(Messages::WebPage::messageReceiverName):
(Messages::WebPage::LoadURL::name):
(Messages::WebPage::LoadURL::LoadURL):
(Messages::WebPage::LoadURL::arguments const):
(Messages::WebPage::LoadSomething::name):
(Messages::WebPage::LoadSomething::LoadSomething):
(Messages::WebPage::LoadSomething::arguments const):
(Messages::WebPage::TouchEvent::name):
(Messages::WebPage::TouchEvent::TouchEvent):
(Messages::WebPage::TouchEvent::arguments const):
(Messages::WebPage::AddEvent::name):
(Messages::WebPage::AddEvent::AddEvent):
(Messages::WebPage::AddEvent::arguments const):
(Messages::WebPage::LoadSomethingElse::name):
(Messages::WebPage::LoadSomethingElse::LoadSomethingElse):
(Messages::WebPage::LoadSomethingElse::arguments const):
(Messages::WebPage::DidReceivePolicyDecision::name):
(Messages::WebPage::DidReceivePolicyDecision::DidReceivePolicyDecision):
(Messages::WebPage::DidReceivePolicyDecision::arguments const):
(Messages::WebPage::Close::name):
(Messages::WebPage::Close::arguments const):
(Messages::WebPage::PreferencesDidChange::name):
(Messages::WebPage::PreferencesDidChange::PreferencesDidChange):
(Messages::WebPage::PreferencesDidChange::arguments const):
(Messages::WebPage::SendDoubleAndFloat::name):
(Messages::WebPage::SendDoubleAndFloat::SendDoubleAndFloat):
(Messages::WebPage::SendDoubleAndFloat::arguments const):
(Messages::WebPage::SendInts::name):
(Messages::WebPage::SendInts::SendInts):
(Messages::WebPage::SendInts::arguments const):
(Messages::WebPage::CreatePlugin::name):
(Messages::WebPage::CreatePlugin::CreatePlugin):
(Messages::WebPage::CreatePlugin::arguments const):
(Messages::WebPage::RunJavaScriptAlert::name):
(Messages::WebPage::RunJavaScriptAlert::RunJavaScriptAlert):
(Messages::WebPage::RunJavaScriptAlert::arguments const):
(Messages::WebPage::GetPlugins::name):
(Messages::WebPage::GetPlugins::GetPlugins):
(Messages::WebPage::GetPlugins::arguments const):
(Messages::WebPage::GetPluginProcessConnection::name):
(Messages::WebPage::GetPluginProcessConnection::GetPluginProcessConnection):
(Messages::WebPage::GetPluginProcessConnection::arguments const):
(Messages::WebPage::TestMultipleAttributes::name):
(Messages::WebPage::TestMultipleAttributes::arguments const):
(Messages::WebPage::TestParameterAttributes::name):
(Messages::WebPage::TestParameterAttributes::TestParameterAttributes):
(Messages::WebPage::TestParameterAttributes::arguments const):
(Messages::WebPage::TemplateTest::name):
(Messages::WebPage::TemplateTest::TemplateTest):
(Messages::WebPage::TemplateTest::arguments const):
(Messages::WebPage::SetVideoLayerID::name):
(Messages::WebPage::SetVideoLayerID::SetVideoLayerID):
(Messages::WebPage::SetVideoLayerID::arguments const):
(Messages::WebPage::DidCreateWebProcessConnection::name):
(Messages::WebPage::DidCreateWebProcessConnection::DidCreateWebProcessConnection):
(Messages::WebPage::DidCreateWebProcessConnection::arguments const):
(Messages::WebPage::InterpretKeyEvent::name):
(Messages::WebPage::InterpretKeyEvent::InterpretKeyEvent):
(Messages::WebPage::InterpretKeyEvent::arguments const):
(Messages::WebPage::DeprecatedOperation::name):
(Messages::WebPage::DeprecatedOperation::DeprecatedOperation):
(Messages::WebPage::DeprecatedOperation::arguments const):
(Messages::WebPage::ExperimentalOperation::name):
(Messages::WebPage::ExperimentalOperation::ExperimentalOperation):
(Messages::WebPage::ExperimentalOperation::arguments const):
(Messages::None::messageReceiverName): Deleted.
* Scripts/testMessagesReplies.h:
* Scripts/webkit/messages_unittest.py:
* Scripts/webkit/test.messages.in: Added.
2020-05-06 Megan Gardner <megan_gardner@apple.com>
Context Menus presented from the wrong location when presented from sub-frames
https://bugs.webkit.org/show_bug.cgi?id=211537
<rdar://problem/60390846>
Reviewed by Tim Horton.
The calculation for bounds for selection Information did not take frames into account,
so if a frame was used, the location would not be correct in the context of the whole
page, so the presentation location ended up being way off.
* UIProcess/ios/forms/WKFileUploadPanel.mm:
(-[WKFileUploadPanel showDocumentPickerMenu]):
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::boundsPositionInformation):
(WebKit::selectionPositionInformation):
2020-05-06 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r260689.
https://bugs.webkit.org/show_bug.cgi?id=211542
"Caused regressions deleting text and filling password fields"
(Requested by ddkilzer on #webkit).
Reverted changeset:
"WebPasteboardProxy::getPasteboardStringsForType() and
WebPasteboardProxy::readURLFromPasteboard() should check
return value of SharedMemory::createHandle()"
https://bugs.webkit.org/show_bug.cgi?id=211002
https://trac.webkit.org/changeset/260689
2020-05-06 Zalan Bujtas <zalan@apple.com>
[ContentObservation] Shutterstock search bar is not activated on the first tap
https://bugs.webkit.org/show_bug.cgi?id=211529
<rdar://problem/58843932>
Reviewed by Simon Fraser.
Shutterstock has a “initial click” action which converts some “actionable” elements (<a>) to some other “actionable” elements (<button>).
If this initial click also happens to be on an element that triggers content observation (input), we see those changes as actionable content and not proceed with the click event (stay at hover).
Any subsequent click works as expected.
It’s very difficult to eliminate such false positives since all we see is that some "actionable" content is going away while some "actionable" content is being created.
This quirk ensures that the first tap on the page does not trigger content observation. (It also means that any hover menu gets submitted on the first tap, but apparently
the Shutterstock top menu bar works fine with click events.)
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::didStartPageTransition):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::handleSyntheticClick):
2020-05-06 Wenson Hsieh <wenson_hsieh@apple.com>
Use CocoaColor in more places instead of platform defines
https://bugs.webkit.org/show_bug.cgi?id=211527
Reviewed by Darin Adler and Tim Horton.
Use CocoaColor (added in r261247) in more places, and also introduce `CocoaFont.h` and use it to simplify some
existing code that uses platform ifdefs for UIFont, NSFont, UIFontDescriptor and NSFontDescriptor.
No change in behavior.
* Platform/cocoa/CocoaColor.h:
* Platform/cocoa/CocoaFont.h: Copied from Source/WebKit/Platform/cocoa/CocoaColor.h.
* Platform/cocoa/CocoaImage.h:
Change `@class`es to `OBJC_CLASS`, so that this header can be safely imported in C++ code.
* Shared/Cocoa/ArgumentCodersCocoa.mm:
(IPC::typeFromObject):
(IPC::isSerializableValue):
(IPC::encodeFontInternal):
(IPC::encodeObject):
(IPC::platformColorClass): Deleted.
(IPC::platformFontClass): Deleted.
* Shared/Cocoa/CoreTextHelpers.h:
* Shared/Cocoa/CoreTextHelpers.mm:
(WebKit::fontWithAttributes):
* Shared/Cocoa/WebCoreArgumentCodersCocoa.mm:
* UIProcess/Cocoa/WKSafeBrowsingWarning.h:
* UIProcess/Cocoa/WKSafeBrowsingWarning.mm:
(fontOfSize):
(colorForItem):
(-[WKSafeBrowsingBox setSafeBrowsingBackgroundColor:]):
(-[WKSafeBrowsingWarning showDetailsClicked]):
(-[WKSafeBrowsingTextView initWithAttributedString:forWarning:]):
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::updateFontManagerIfNeeded):
* WebKit.xcodeproj/project.pbxproj:
2020-05-06 Alex Christensen <achristensen@webkit.org>
Reduce IPC overhead for message receiver name and message name to 2 bytes
https://bugs.webkit.org/show_bug.cgi?id=211112
Reviewed by Chris Dumez.
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* GPUProcess/GPUConnectionToWebProcess.cpp:
(WebKit::GPUConnectionToWebProcess::didReceiveInvalidMessage):
* GPUProcess/GPUConnectionToWebProcess.h:
* NetworkProcess/NetworkConnectionToWebProcess.cpp:
(WebKit::NetworkConnectionToWebProcess::didReceiveMessage):
(WebKit::NetworkConnectionToWebProcess::didReceiveSyncMessage):
(WebKit::NetworkConnectionToWebProcess::didReceiveInvalidMessage):
* NetworkProcess/NetworkConnectionToWebProcess.h:
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::didReceiveMessage):
(WebKit::NetworkProcess::didReceiveSyncMessage):
* NetworkProcess/ios/NetworkConnectionToWebProcessIOS.mm:
(WebKit::NetworkConnectionToWebProcess::paymentCoordinatorAddMessageReceiver): Deleted.
(WebKit::NetworkConnectionToWebProcess::paymentCoordinatorRemoveMessageReceiver): Deleted.
* Platform/IPC/Connection.cpp:
(IPC::Connection::WaitForMessageState::WaitForMessageState):
(IPC::Connection::addWorkQueueMessageReceiver):
(IPC::Connection::removeWorkQueueMessageReceiver):
(IPC::Connection::dispatchWorkQueueMessageReceiverMessage):
(IPC::Connection::addThreadMessageReceiver):
(IPC::Connection::removeThreadMessageReceiver):
(IPC::Connection::dispatchThreadMessageReceiverMessage):
(IPC::Connection::createSyncMessageEncoder):
(IPC::Connection::sendMessage):
(IPC::Connection::waitForMessage):
(IPC::Connection::sendSyncMessage):
(IPC::Connection::waitForSyncReply):
(IPC::Connection::processIncomingMessage):
(IPC::Connection::dispatchSyncMessage):
(IPC::Connection::dispatchDidReceiveInvalidMessage):
(IPC::Connection::dispatchMessage):
(IPC::Connection::dispatchMessageToWorkQueueReceiver):
(IPC::Connection::dispatchMessageToThreadReceiver):
(IPC::Connection::dispatchIncomingMessages):
* Platform/IPC/Connection.h:
(IPC::Connection::send):
(IPC::Connection::sendWithAsyncReply):
(IPC::Connection::sendSync):
(IPC::Connection::waitForAndDispatchImmediately):
* Platform/IPC/Decoder.cpp:
* Platform/IPC/Decoder.h:
(IPC::Decoder::messageReceiverName const):
(IPC::Decoder::messageName const):
(IPC::Decoder::decode):
(IPC::Decoder::operator>>):
* Platform/IPC/Encoder.cpp:
(IPC::Encoder::Encoder):
(IPC::Encoder::encodeHeader):
* Platform/IPC/Encoder.h:
* Platform/IPC/HandleMessage.h:
(IPC::handleMessageAsync):
* Platform/IPC/MessageReceiverMap.cpp:
(IPC::MessageReceiverMap::addMessageReceiver):
(IPC::MessageReceiverMap::removeMessageReceiver):
(IPC::MessageReceiverMap::dispatchMessage):
(IPC::MessageReceiverMap::dispatchSyncMessage):
* Platform/IPC/MessageReceiverMap.h:
* Platform/IPC/MessageSender.h:
* Platform/IPC/cocoa/ConnectionCocoa.mm:
(IPC::Connection::open):
(IPC::Connection::sendMessage):
(IPC::Connection::sendOutgoingMessage):
(IPC::Connection::receiveSourceEventHandler):
* Platform/IPC/cocoa/MachMessage.cpp:
(IPC::MachMessage::create):
(IPC::MachMessage::MachMessage):
* Platform/IPC/cocoa/MachMessage.h:
(IPC::MachMessage::messageReceiverName const):
(IPC::MachMessage::messageName const):
* PluginProcess/WebProcessConnection.cpp:
(WebKit::WebProcessConnection::didReceiveInvalidMessage):
* PluginProcess/WebProcessConnection.h:
* Scripts/Makefile:
* Scripts/generate-message-receiver.py:
(main):
* Scripts/webkit/LegacyMessages-expected.h: Removed.
* Scripts/webkit/LegacyMessagesReplies-expected.h: Removed.
* Scripts/webkit/MessageNames.cpp: Added.
(IPC::description):
* Scripts/webkit/MessageNames.h: Added.
(IPC::receiverName):
(WTF::isValidEnum):
* Scripts/webkit/MessageReceiver-expected.cpp: Removed.
* Scripts/webkit/MessageReceiverSuperclass-expected.cpp: Removed.
* Scripts/webkit/Messages-expected.h: Removed.
* Scripts/webkit/MessagesReplies-expected.h: Removed.
* Scripts/webkit/MessagesRepliesSuperclassReplies-expected.h: Removed.
* Scripts/webkit/MessagesSuperclass-expected.h: Removed.
* Scripts/webkit/messages.py:
* Scripts/webkit/test-legacy-messages.in: Removed.
* Scripts/webkit/test-legacy.messages.in: Copied from Source/WebKit/Scripts/webkit/test-legacy-messages.in.
* Scripts/webkit/test-legacyMessageReceiver.cpp: Added.
(Messages::WebPage::GetPluginProcessConnection::send):
(Messages::WebPage::TestMultipleAttributes::send):
(WebKit::WebPage::didReceiveWebPageMessage):
(WebKit::WebPage::didReceiveSyncWebPageMessage):
* Scripts/webkit/test-legacyMessages.h: Added.
(Messages::WebPage::messageReceiverName):
(Messages::WebPage::LoadURL::name):
(Messages::WebPage::LoadURL::LoadURL):
(Messages::WebPage::LoadURL::arguments const):
(Messages::WebPage::LoadSomething::name):
(Messages::WebPage::LoadSomething::LoadSomething):
(Messages::WebPage::LoadSomething::arguments const):
(Messages::WebPage::TouchEvent::name):
(Messages::WebPage::TouchEvent::TouchEvent):
(Messages::WebPage::TouchEvent::arguments const):
(Messages::WebPage::AddEvent::name):
(Messages::WebPage::AddEvent::AddEvent):
(Messages::WebPage::AddEvent::arguments const):
(Messages::WebPage::LoadSomethingElse::name):
(Messages::WebPage::LoadSomethingElse::LoadSomethingElse):
(Messages::WebPage::LoadSomethingElse::arguments const):
(Messages::WebPage::DidReceivePolicyDecision::name):
(Messages::WebPage::DidReceivePolicyDecision::DidReceivePolicyDecision):
(Messages::WebPage::DidReceivePolicyDecision::arguments const):
(Messages::WebPage::Close::name):
(Messages::WebPage::Close::arguments const):
(Messages::WebPage::PreferencesDidChange::name):
(Messages::WebPage::PreferencesDidChange::PreferencesDidChange):
(Messages::WebPage::PreferencesDidChange::arguments const):
(Messages::WebPage::SendDoubleAndFloat::name):
(Messages::WebPage::SendDoubleAndFloat::SendDoubleAndFloat):
(Messages::WebPage::SendDoubleAndFloat::arguments const):
(Messages::WebPage::SendInts::name):
(Messages::WebPage::SendInts::SendInts):
(Messages::WebPage::SendInts::arguments const):
(Messages::WebPage::CreatePlugin::name):
(Messages::WebPage::CreatePlugin::CreatePlugin):
(Messages::WebPage::CreatePlugin::arguments const):
(Messages::WebPage::RunJavaScriptAlert::name):
(Messages::WebPage::RunJavaScriptAlert::RunJavaScriptAlert):
(Messages::WebPage::RunJavaScriptAlert::arguments const):
(Messages::WebPage::GetPlugins::name):
(Messages::WebPage::GetPlugins::GetPlugins):
(Messages::WebPage::GetPlugins::arguments const):
(Messages::WebPage::GetPluginProcessConnection::name):
(Messages::WebPage::GetPluginProcessConnection::GetPluginProcessConnection):
(Messages::WebPage::GetPluginProcessConnection::arguments const):
(Messages::WebPage::TestMultipleAttributes::name):
(Messages::WebPage::TestMultipleAttributes::arguments const):
(Messages::WebPage::TestParameterAttributes::name):
(Messages::WebPage::TestParameterAttributes::TestParameterAttributes):
(Messages::WebPage::TestParameterAttributes::arguments const):
(Messages::WebPage::TemplateTest::name):
(Messages::WebPage::TemplateTest::TemplateTest):
(Messages::WebPage::TemplateTest::arguments const):
(Messages::WebPage::SetVideoLayerID::name):
(Messages::WebPage::SetVideoLayerID::SetVideoLayerID):
(Messages::WebPage::SetVideoLayerID::arguments const):
(Messages::WebPage::DidCreateWebProcessConnection::name):
(Messages::WebPage::DidCreateWebProcessConnection::DidCreateWebProcessConnection):
(Messages::WebPage::DidCreateWebProcessConnection::arguments const):
(Messages::WebPage::InterpretKeyEvent::name):
(Messages::WebPage::InterpretKeyEvent::InterpretKeyEvent):
(Messages::WebPage::InterpretKeyEvent::arguments const):
(Messages::WebPage::DeprecatedOperation::name):
(Messages::WebPage::DeprecatedOperation::DeprecatedOperation):
(Messages::WebPage::DeprecatedOperation::arguments const):
(Messages::WebPage::ExperimentalOperation::name):
(Messages::WebPage::ExperimentalOperation::ExperimentalOperation):
(Messages::WebPage::ExperimentalOperation::arguments const):
* Scripts/webkit/test-legacyMessagesReplies.h: Added.
* Scripts/webkit/test-messages.in: Removed.
* Scripts/webkit/test-superclass-messages.in: Removed.
* Scripts/webkit/test-superclass.messages.in: Copied from Source/WebKit/Scripts/webkit/test-superclass-messages.in.
* Scripts/webkit/test-superclassMessageReceiver.cpp: Added.
(Messages::WebPage::TestAsyncMessage::callReply):
(Messages::WebPage::TestAsyncMessage::cancelReply):
(Messages::WebPage::TestAsyncMessage::send):
(Messages::WebPage::TestAsyncMessageWithNoArguments::callReply):
(Messages::WebPage::TestAsyncMessageWithNoArguments::cancelReply):
(Messages::WebPage::TestAsyncMessageWithNoArguments::send):
(Messages::WebPage::TestAsyncMessageWithMultipleArguments::callReply):
(Messages::WebPage::TestAsyncMessageWithMultipleArguments::cancelReply):
(Messages::WebPage::TestAsyncMessageWithMultipleArguments::send):
(Messages::WebPage::TestSyncMessage::send):
(Messages::WebPage::TestSynchronousMessage::send):
(WebKit::WebPage::didReceiveMessage):
(WebKit::WebPage::didReceiveSyncMessage):
* Scripts/webkit/test-superclassMessages.h: Added.
(Messages::WebPage::messageReceiverName):
(Messages::WebPage::LoadURL::name):
(Messages::WebPage::LoadURL::LoadURL):
(Messages::WebPage::LoadURL::arguments const):
(Messages::WebPage::TestAsyncMessage::name):
(Messages::WebPage::TestAsyncMessage::asyncMessageReplyName):
(Messages::WebPage::TestAsyncMessage::TestAsyncMessage):
(Messages::WebPage::TestAsyncMessage::arguments const):
(Messages::WebPage::TestAsyncMessageWithNoArguments::name):
(Messages::WebPage::TestAsyncMessageWithNoArguments::asyncMessageReplyName):
(Messages::WebPage::TestAsyncMessageWithNoArguments::arguments const):
(Messages::WebPage::TestAsyncMessageWithMultipleArguments::name):
(Messages::WebPage::TestAsyncMessageWithMultipleArguments::asyncMessageReplyName):
(Messages::WebPage::TestAsyncMessageWithMultipleArguments::arguments const):
(Messages::WebPage::TestSyncMessage::name):
(Messages::WebPage::TestSyncMessage::TestSyncMessage):
(Messages::WebPage::TestSyncMessage::arguments const):
(Messages::WebPage::TestSynchronousMessage::name):
(Messages::WebPage::TestSynchronousMessage::TestSynchronousMessage):
(Messages::WebPage::TestSynchronousMessage::arguments const):
* Scripts/webkit/test-superclassMessagesReplies.h: Added.
* Scripts/webkit/test.messages.in: Added.
* Scripts/webkit/testMessageReceiver.cpp: Added.
* Scripts/webkit/testMessages.h: Added.
(Messages::None::messageReceiverName):
* Scripts/webkit/testMessagesReplies.h: Added.
* Shared/ApplePay/WebPaymentCoordinatorProxy.h:
* Shared/ApplePay/cocoa/WebPaymentCoordinatorProxyCocoa.mm:
(WebKit::WebPaymentCoordinatorProxy::WebPaymentCoordinatorProxy):
(WebKit::WebPaymentCoordinatorProxy::~WebPaymentCoordinatorProxy):
* Shared/AuxiliaryProcess.cpp:
(WebKit::AuxiliaryProcess::addMessageReceiver):
(WebKit::AuxiliaryProcess::removeMessageReceiver):
(WebKit::AuxiliaryProcess::didReceiveInvalidMessage):
* Shared/AuxiliaryProcess.h:
(WebKit::AuxiliaryProcess::addMessageReceiver):
(WebKit::AuxiliaryProcess::removeMessageReceiver):
* Shared/Cocoa/AuxiliaryProcessCocoa.mm:
(WebKit::AuxiliaryProcess::didReceiveInvalidMessage):
* SourcesCocoa.txt:
* UIProcess/AuxiliaryProcessProxy.cpp:
(WebKit::AuxiliaryProcessProxy::addMessageReceiver):
(WebKit::AuxiliaryProcessProxy::removeMessageReceiver):
(WebKit::AuxiliaryProcessProxy::logInvalidMessage):
* UIProcess/AuxiliaryProcessProxy.h:
(WebKit::AuxiliaryProcessProxy::addMessageReceiver):
(WebKit::AuxiliaryProcessProxy::removeMessageReceiver):
(WebKit::AuxiliaryProcessProxy::send):
(WebKit::AuxiliaryProcessProxy::sendWithAsyncReply):
* UIProcess/Cocoa/UserMediaCaptureManagerProxy.h:
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::paymentCoordinatorAddMessageReceiver): Deleted.
(WebKit::WebPageProxy::paymentCoordinatorRemoveMessageReceiver): Deleted.
* UIProcess/GPU/GPUProcessProxy.cpp:
(WebKit::GPUProcessProxy::didReceiveInvalidMessage):
* UIProcess/GPU/GPUProcessProxy.h:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::didReceiveInvalidMessage):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/Plugins/PluginProcessProxy.cpp:
(WebKit::PluginProcessProxy::didReceiveInvalidMessage):
* UIProcess/Plugins/PluginProcessProxy.h:
* UIProcess/ProvisionalPageProxy.cpp:
(WebKit::ProvisionalPageProxy::didReceiveMessage):
* UIProcess/SuspendedPageProxy.cpp:
(WebKit::messageNamesToIgnoreWhileSuspended):
(WebKit::SuspendedPageProxy::didReceiveMessage):
* UIProcess/WebPageProxy.h:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::didReceiveInvalidMessage):
(WebKit::WebProcessPool::addMessageReceiver):
(WebKit::WebProcessPool::removeMessageReceiver):
* UIProcess/WebProcessPool.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::shouldSendPendingMessage):
(WebKit::WebProcessProxy::didReceiveInvalidMessage):
* UIProcess/WebProcessProxy.h:
* WebProcess/GPU/GPUProcessConnection.cpp:
(WebKit::GPUProcessConnection::didReceiveInvalidMessage):
* WebProcess/GPU/GPUProcessConnection.h:
* WebProcess/Inspector/WebInspector.h:
* WebProcess/Inspector/WebInspectorUI.h:
* WebProcess/Network/NetworkProcessConnection.cpp:
(WebKit::NetworkProcessConnection::didReceiveInvalidMessage):
* WebProcess/Network/NetworkProcessConnection.h:
* WebProcess/Plugins/PluginProcessConnection.cpp:
(WebKit::PluginProcessConnection::didReceiveInvalidMessage):
* WebProcess/Plugins/PluginProcessConnection.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.mm:
(WebKit::RemoteLayerTreeDrawingArea::updateRendering):
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::didReceiveMessage):
2020-05-06 Chris Dumez <cdumez@apple.com>
REGRESSION (r260684): Reader background is lost after multitasking
https://bugs.webkit.org/show_bug.cgi?id=211533
<rdar://problem/62941837>
Unreviewed, revert r260684 due to regression.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::setShouldFireResizeEvents):
* UIProcess/WebPageProxy.h:
* UIProcess/ios/WKApplicationStateTrackingView.mm:
(-[WKApplicationStateTrackingView _willBeginSnapshotSequence]):
(-[WKApplicationStateTrackingView _didCompleteSnapshotSequence]):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::setShouldFireResizeEvents):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-05-06 Kate Cheney <katherine_cheney@apple.com>
ResourceLoadStatistics data summary call should create a web process pool if one doesn't exist
https://bugs.webkit.org/show_bug.cgi?id=211520
<rdar://problem/59869619>
Reviewed by Chris Dumez.
Ensures that a process pool and network process exist when requesting
resource load statistics data.
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::getResourceLoadStatisticsDataSummary):
2020-05-06 Keith Rollin <krollin@apple.com>
Better surfacing of the presenting parent PID in the Network process
https://bugs.webkit.org/show_bug.cgi?id=211495
<rdar://problem/62917205>
Reviewed by Youenn Fablet.
In Bug 205295, NetworkResourceLoader logging was improved, with one of
the changes being that the PID of parent process that invoked the
Network process was logged at the start of
NetworkResourceLoader::start(). However, there was an execution path
that skipped calling start()
(NetworkConnectionToWebProcess::scheduleResourceLoad ->
NetworkResourceLoader::startWithServiceWorker ->
serviceWorkerDidNotHandle -> restartNetworkLoad -> startNetworkLoad),
and so the the logging would not show the parent PID. This logging is
needed for diagnostic purposes, so tweak the logging a little bit more
in order to ensure we emit it. In this change, logging is added to
various loading-related NetworkConnectionToWebProcess entry points
(not just scheduleResourceLoad) in order to (a) ensure we emit the
logging we want and (b) to give a clearer picture of what operations
are being invoked in the Network process.
No new tests -- no new or changed functionality.
* NetworkProcess/NetworkConnectionToWebProcess.cpp:
(WebKit::NetworkConnectionToWebProcess::hasUploadStateChanged):
(WebKit::NetworkConnectionToWebProcess::resolveBlobReferences):
(WebKit::NetworkConnectionToWebProcess::scheduleResourceLoad):
(WebKit::NetworkConnectionToWebProcess::performSynchronousLoad):
(WebKit::NetworkConnectionToWebProcess::loadPing):
(WebKit::NetworkConnectionToWebProcess::preconnectTo):
(WebKit::NetworkConnectionToWebProcess::serverToContextConnectionNoLongerNeeded):
* NetworkProcess/NetworkResourceLoader.cpp:
(WebKit::NetworkResourceLoader::start):
2020-05-06 Wenson Hsieh <wenson_hsieh@apple.com>
Cut and paste from Google Doc to Notes in several (non-Latin) languages doesn't work
https://bugs.webkit.org/show_bug.cgi?id=211498
<rdar://problem/56675345>
Reviewed by Darin Adler.
Add a new header to allow Cocoa code to reason about UIColors and NSColors on iOS and macOS (respectively)
without requiring platform ifdefs. A followup patch will adopt this in several places in WebKit, where we
currently need ifdefs for iOS and macOS.
* Platform/cocoa/CocoaColor.h: Added.
* WebKit.xcodeproj/project.pbxproj:
2020-05-06 Antoine Quint <graouts@apple.com>
pointermove event sometimes has incorrect pointerType of 'mouse' on touch interactions
https://bugs.webkit.org/show_bug.cgi?id=210716
<rdar://problem/62084817>
Reviewed by Darin Adler.
Set the SyntheticClickType to OneFingerTap when creating a synthetic "mousemove" event for a synthetic "click" event as part of a tap.
On top of being more indicative of the actual action that triggered the synthetic event, it also has the added quality that it will
not yield a "pointermove" event.
Test: pointerevents/ios/click-no-pointermove.html
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::dispatchSyntheticMouseMove):
2020-05-06 John Wilander <wilander@apple.com>
Exempt app-bound domains from ITP's website data deletion and third-party cookie blocking between themselves
https://bugs.webkit.org/show_bug.cgi?id=210674
<rdar://problem/61950767>
Reviewed by Chris Dumez.
This change forwards information about app-bound domains to ITP and web
processes so that they can be exempt from website data deletion and
third-party cookie blocking between themselves.
App-bound domains are configured statically and apply to all website
data stores. Therefore the setting needs to be forwarded to all
website data stores and ITP functionality in all network and web
content processes. This is done through the new static function
WebsiteDataStore::setAppBoundDomainsForITPIfInitialized().
Since app-bound domains are loaded lazily from disk and on a background
thread, this patch forwards them in ResourceLoadStatisticsParameters if
they've already been loaded. Then every time app-bound domains are
updated, they are forwarded to ITP. This ensures that ITP will have them
as soon as possible.
Setting app-bound domains for the purposes of ITP automatically switches
ITP's cookie blocking policy to the new
WebCore::ThirdPartyCookieBlockingMode::AllExceptBetweenAppBoundDomains.
This is done in WebResourceLoadStatisticsStore::setAppBoundDomains().
The C API changes are for test purposes.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:
(WebKit::ResourceLoadStatisticsMemoryStore::registrableDomainsToDeleteOrRestrictWebsiteDataFor):
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.cpp:
(WebKit::ResourceLoadStatisticsStore::setAppBoundDomains):
(WebKit::ResourceLoadStatisticsStore::resetParametersToDefaultValues):
(WebKit::ResourceLoadStatisticsStore::shouldExemptFromWebsiteDataDeletion const):
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.h:
(WebKit::ResourceLoadStatisticsStore::standaloneApplicationDomain const): Deleted.
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::setAppBoundDomains):
(WebKit::WebResourceLoadStatisticsStore::resetParametersToDefaultValues):
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.h:
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::setAppBoundDomainsForResourceLoadStatistics):
(WebKit::NetworkProcess::setThirdPartyCookieBlockingMode):
(WebKit::NetworkProcess::setShouldBlockThirdPartyCookiesForTesting): Deleted.
Renamed setThirdPartyCookieBlockingMode.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::NetworkSessionCocoa):
* Shared/ResourceLoadStatisticsParameters.h:
(WebKit::ResourceLoadStatisticsParameters::encode const):
(WebKit::ResourceLoadStatisticsParameters::decode):
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreSetAppBoundDomainsForTesting):
* UIProcess/API/C/WKWebsiteDataStoreRef.h:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::setAppBoundDomainsForResourceLoadStatistics):
(WebKit::NetworkProcessProxy::setThirdPartyCookieBlockingMode):
(WebKit::NetworkProcessProxy::setShouldBlockThirdPartyCookiesForTesting): Deleted.
Renamed setThirdPartyCookieBlockingMode.
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::ensureNetworkProcess):
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::setThirdPartyCookieBlockingMode):
(WebKit::WebProcessProxy::setShouldBlockThirdPartyCookiesForTesting): Deleted.
Renamed setThirdPartyCookieBlockingMode.
* UIProcess/WebProcessProxy.h:
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::WebsiteDataStore::initializeAppBoundDomains):
(WebKit::WebsiteDataStore::ensureAppBoundDomains const):
(WebKit::WebsiteDataStore::appBoundDomainsIfInitialized):
This function allows fetching of app-bound domains without triggering
the lazy loading. This is just to allow speculative configuration of ITP
right when it's created — if any app-bound domains are already configured,
forward them to ITP via ResourceLoadStatisticsParameters.
(WebKit::WebsiteDataStore::setAppBoundDomainsForTesting):
This function is Cocoa-specific and only accepts localhost and 127.0.0.1
to be configured as app-bound domains.
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::setResourceLoadStatisticsShouldBlockThirdPartyCookiesForTesting):
(WebKit::WebsiteDataStore::setThirdPartyCookieBlockingMode):
(WebKit::WebsiteDataStore::parameters):
(WebKit::WebsiteDataStore::forwardAppBoundDomainsToITPIfInitialized):
(WebKit::WebsiteDataStore::setAppBoundDomainsForITP):
* UIProcess/WebsiteData/WebsiteDataStore.h:
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::setThirdPartyCookieBlockingMode):
(WebKit::WebProcess::setShouldBlockThirdPartyCookiesForTesting): Deleted.
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
2020-05-06 Daniel Bates <dabates@apple.com>
[iOS] ASSERTION FAILED: !(_keyboardFlags & WebEventKeyboardInputModifierFlagsChanged) in -[WebEvent charactersIgnoringModifiers] when pressing modifier on PDF
https://bugs.webkit.org/show_bug.cgi?id=211472
Reviewed by Darin Adler.
Do not ask for event character string for a flags changed event: they don't have one.
Test: fast/events/ios/pdf-modifer-key-down-crash.html
* UIProcess/ios/WKKeyboardScrollingAnimator.mm:
(-[WKKeyboardScrollingAnimator keyboardScrollForEvent:]):
2020-05-06 Brent Fulgham <bfulgham@apple.com>
[MacCatalyst] Processes should check for network entitlement as we do for macOS apps
https://bugs.webkit.org/show_bug.cgi?id=211474
<rdar://problem/61182060>
Reviewed by Alexey Proskuryakov.
We should make the same entitlement checks in MacCatalyst as we do for macOS.
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceEntryPoint.mm:
(WebKit::XPCServiceInitializerDelegate::checkEntitlements):
2020-05-06 Darin Adler <darin@apple.com>
Eliminate checks of USE(DICTATION_ALTERNATIVES) in Cocoa-specific code
https://bugs.webkit.org/show_bug.cgi?id=211460
Reviewed by Anders Carlsson.
* UIProcess/Cocoa/PageClientImplCocoa.h: Remove USE(DICTATION_ALTERNATIVES).
Also remove unnecessary use of RetainPtr.
* UIProcess/Cocoa/PageClientImplCocoa.mm:
(WebKit::PageClientImplCocoa::PageClientImplCocoa): Ditto.
(WebKit::PageClientImplCocoa::pageClosed): Ditto.
(WebKit::PageClientImplCocoa::dictationAlternatives): Ditto.
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::insertDictatedTextAsync): Ditto.
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::validAttributesForMarkedText): Ditto.
(WebKit::WebViewImpl::insertText): Ditto.
* UIProcess/PageClient.h: Ditto.
* UIProcess/ios/PageClientImplIOS.h: Ditto.
* UIProcess/ios/PageClientImplIOS.mm: Ditto.
* UIProcess/ios/WKContentViewInteraction.mm: Ditto.
* UIProcess/mac/PageClientImplMac.h: Ditto.
* UIProcess/mac/PageClientImplMac.mm: Ditto.
* WebProcess/WebCoreSupport/mac/WebAlternativeTextClient.cpp: Ditto.
2020-05-06 Lauro Moura <lmoura@igalia.com>
[GTK] Cleanup KeyBindingTranslator on WebView disposal
https://bugs.webkit.org/show_bug.cgi?id=211465
Reviewed by Carlos Garcia Campos.
Fixes assertion in debug mode when a test failed and the widget
followed a destruction pattern where KeyBindingTranslator would be
destroyed before the container widget is cleaned up.
Also renamed the method from 'destroyed' to 'invalidate' to try to
better convey the action being performed on the KeyBindingTranslator.
Covered by existing tests.
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseContainerRemove): Update method name.
(webkitWebViewBaseDispose): Cleanup keyBindingTranslator when
disposing this widget.
* UIProcess/gtk/KeyBindingTranslator.h:
(WebKit::KeyBindingTranslator::invalidate): Renamed from destroyed.
(WebKit::KeyBindingTranslator::destroyed): Deleted.
2020-05-06 Darin Adler <darin@apple.com>
Reduce HAVE(HOSTED_CORE_ANIMATION)
https://bugs.webkit.org/show_bug.cgi?id=211423
Reviewed by Anders Carlsson.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess): Removed redundant #if
checking both HAVE(HOSTED_CORE_ANIMATION) and !PLATFORM(IOS_FAMILY).
* UIProcess/Plugins/mac/PluginProcessProxyMac.mm:
(WebKit::PluginProcessProxy::platformInitializePluginProcess): Removed unnecessary
HAVE(HOSTED_CORE_ANIMATION) check in code that is only compiled for macOS.
2020-05-05 David Kilzer <ddkilzer@apple.com>
Fix deprecated NSGraphicsContext methods using 'graphicsPort'
<https://webkit.org/b/211481>
Reviewed by Darin Adler.
- Replace uses of -graphicsPort with -CGContext.
- Replace uses of -graphicsContextWithGraphicsPort:flipped: with
-graphicsContextWithCGContext:flipped:.
- Remove ALLOW_DEPRECATED_DECLARATIONS_{BEGIN,END} if possible.
* UIProcess/mac/WKPrintingView.mm:
(-[WKPrintingView _drawPDFDocument:page:atPoint:]):
(-[WKPrintingView _drawPreview:]):
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::drawPDFPage):
2020-05-05 Zalan Bujtas <zalan@apple.com>
[Quirk] Job listing page goes blank when scrolling on stackoverflow.com
https://bugs.webkit.org/show_bug.cgi?id=211480
<rdar://problem/62558405>
Reviewed by Wenson Hsieh.
Add the option of returning "use desktop web content" on any configuration when the host application asks for content recommendation.
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::desktopClassBrowsingRecommendedForRequest):
(WebKit::desktopClassBrowsingRecommended):
(WebKit::WebPageProxy::effectiveContentModeAfterAdjustingPolicies):
2020-05-05 Chris Dumez <cdumez@apple.com>
Crash under _LSSetApplicationInformationItem()
https://bugs.webkit.org/show_bug.cgi?id=211478
<rdar://problem/62201314>
Reviewed by Alex Christensen.
Given the crashes, I suspect it is not actually safe to call _LSSetApplicationInformationItem()
from a non main-thread like it was done in r238289. We still run the code asynchronously to
address the issue that r238289 was trying to fix but we now call _LSSetApplicationInformationItem()
on the main thread.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::updateProcessName):
(WebKit::WebProcess::updateActivePages):
(WebKit::setProcessNameQueue): Deleted.
2020-05-05 Peng Liu <peng.liu6@apple.com>
Update WebKitTestRunner to support running multiple video fullscreen and Picture-in-Picture tests simultaneously
https://bugs.webkit.org/show_bug.cgi?id=203723
Reviewed by Jer Noble.
When the flag MockVideoPresentationModeEnabled is true, the VideoFullscreenManagerProxy in the UI process
will mock the behavior of the VideoFullscreenInterface[Mac|AVKit].
* UIProcess/Cocoa/VideoFullscreenManagerProxy.h:
(WebKit::VideoFullscreenManagerProxy::setMockVideoPresentationModeEnabled):
* UIProcess/Cocoa/VideoFullscreenManagerProxy.mm:
(WebKit::VideoFullscreenManagerProxy::requestHideAndExitFullscreen):
(WebKit::VideoFullscreenManagerProxy::applicationDidBecomeActive):
(WebKit::VideoFullscreenManagerProxy::setupFullscreenWithID):
(WebKit::VideoFullscreenManagerProxy::setHasVideo):
(WebKit::VideoFullscreenManagerProxy::setVideoDimensions):
(WebKit::VideoFullscreenManagerProxy::enterFullscreen):
(WebKit::VideoFullscreenManagerProxy::exitFullscreen):
(WebKit::VideoFullscreenManagerProxy::exitFullscreenWithoutAnimationToMode):
(WebKit::VideoFullscreenManagerProxy::setInlineRect):
(WebKit::VideoFullscreenManagerProxy::setHasVideoContentLayer):
(WebKit::VideoFullscreenManagerProxy::cleanupFullscreen):
(WebKit::VideoFullscreenManagerProxy::preparedToReturnToInline):
(WebKit::VideoFullscreenManagerProxy::preparedToExitFullscreen):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didAttachToRunningProcess):
(WebKit::WebPageProxy::setMockVideoPresentationModeEnabled):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::setMockVideoPresentationModeEnabled):
(WebKit::WebChromeClient::setUpPlaybackControlsManager):
(WebKit::WebChromeClient::clearPlaybackControlsManager):
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/cocoa/VideoFullscreenManager.h:
2020-05-05 Per Arne Vollan <pvollan@apple.com>
Unreviewed fix for MotionMark regression.
Revert minor part of r260017. Remove telemetry from IOKit filtering allow rules. This telemetry is causing higher
CPU usage on the system, and is not needed, since we have already confirmed these messages are being used.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-05-05 Daniel Bates <dabates@apple.com>
Should show selection highlight during text interaction
https://bugs.webkit.org/show_bug.cgi?id=211362
<rdar://problem/59191873>
Reviewed by Wenson Hsieh.
Break out text interaction tracking into its own ivar instead of using the
WebKit::InteractionIsHappening selection suppression reason so that I can
key off it to only disallow zooming to reveal the focused element during
a text interaction. The selection suppression machinery does more than this.
It also deactivates selection assistance, painting of highlights, etc, which
I don't want.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView cleanUpInteraction]): Reset state.
(-[WKContentView becomeFirstResponderForWebView]): Activate selection assistant
during a text interaction if the selection assistant is not being suppressed.
(-[WKContentView _zoomToRevealFocusedElement]): Early return if called during a
text interaction. The function -_didFinishTextInteractionInTextInputContext will
trigger the zoom after the interaction completes.
(-[WKContentView _didCommitLoadForMainFrame]): Reset state.
(-[WKContentView _willBeginTextInteractionInTextInputContext:]):
(-[WKContentView _didFinishTextInteractionInTextInputContext:]):
Update state. While I am here, text interaction is considered a gesture so
also update _usingGestureForSelection. This means that the selection view
will be updated immediately on a selection change during a text interaction
instead of waiting until the next layer tree commit occurs.
2020-05-05 Kate Cheney <katherine_cheney@apple.com>
Check for app-bound domains should confirm WKAppBoundDomains key exists when checking for default app-bound protocols.
https://bugs.webkit.org/show_bug.cgi?id=211451
<rdar://problem/62715316
Reviewed by Brent Fulgham.
Checks for WKAppBoundDomains key before treating a protocol as
app-bound. This is a regression in expected behavior after making
In-App Browser privacy opt-in based on the presence of the key.
Also moves the check for special protocols inside of the completion handler for
ensureAppBoundDomains to make sure the check for the key has finished.
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::WebsiteDataStore::beginAppBoundDomainCheck):
Added a comment to clarify why we need to check for both an empty
app-bound domains list and the presence of the key for testing
purposes.
2020-05-05 Timothy Horton <timothy_horton@apple.com>
"Essential Skeleton" does not respond to mouse events, only touch events
https://bugs.webkit.org/show_bug.cgi?id=211439
<rdar://problem/62694519>
Reviewed by Wenson Hsieh.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView shouldUseMouseGestureRecognizer]):
Add a quirk.
2020-05-05 Víctor Manuel Jáquez Leal <vjaquez@igalia.com>
Compile GPUProcess in WPE port as experimental feature
https://bugs.webkit.org/show_bug.cgi?id=211442
Reviewed by Don Olmstead.
Enable GPU Process in WPE.
This is compilation only. No new tests required.
* PlatformWPE.cmake: Add Platform/generic include directory.
* SourcesWPE.txt: Add related source files.
2020-05-05 Youenn Fablet <youenn@apple.com>
Remove LegacySchemeRegistry::canServiceWorkersHandleURLScheme
https://bugs.webkit.org/show_bug.cgi?id=211170
Reviewed by Alex Christensen.
Remove unused parameters since they are no longer being set in UIProcess.
* NetworkProcess/NetworkProcessCreationParameters.cpp:
(WebKit::NetworkProcessCreationParameters::encode const):
* NetworkProcess/NetworkProcessCreationParameters.h:
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::initializeWebProcess):
2020-05-05 Darin Adler <darin@apple.com>
Remove now-uneeded HAVE macros related to PDF
https://bugs.webkit.org/show_bug.cgi?id=211435
Reviewed by Anders Carlsson.
* Platform/spi/ios/PDFKitSPI.h: Removed HAVE(PDFHOSTVIEWCONTROLLER_SNAPSHOTTING).
* UIProcess/ios/WKPDFView.mm:
(-[WKPDFView web_initWithFrame:webView:mimeType:]): Removed
HAVE(PDF_HOST_VIEW_CONTROLLER_WITH_BACKGROUND_COLOR).
(+[WKPDFView web_requiresCustomSnapshotting]): Removed HAVE(PDFHOSTVIEWCONTROLLER_SNAPSHOTTING).
(-[WKPDFView web_snapshotRectInContentViewCoordinates:snapshotWidth:completionHandler:]):
Ditto.
2020-05-05 Youenn Fablet <youenn@apple.com>
Receiving WebRTC network packets should not go through the main thread
https://bugs.webkit.org/show_bug.cgi?id=211290
Reviewed by Alex Christensen.
Instead of going to main thread before going to rtc network thread, we register a message receiver and dispatch directly to rtc network thread.
This ensures rtc packets are not blocked if main thread is busy.
* DerivedSources-input.xcfilelist:
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* NetworkProcess/webrtc/LibWebRTCSocketClient.cpp:
(WebKit::LibWebRTCSocketClient::signalReadPacket):
(WebKit::LibWebRTCSocketClient::signalSentPacket):
(WebKit::LibWebRTCSocketClient::signalAddressReady):
(WebKit::LibWebRTCSocketClient::signalConnect):
(WebKit::LibWebRTCSocketClient::signalClose):
* NetworkProcess/webrtc/NetworkRTCProvider.cpp:
(WebKit::NetworkRTCProvider::createSocket):
(WebKit::NetworkRTCProvider::createServerTCPSocket):
(WebKit::NetworkRTCProvider::createClientTCPSocket):
(WebKit::NetworkRTCProvider::newConnection):
(WebKit::NetworkRTCProvider::closeListeningSockets):
* Sources.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/Network/NetworkProcessConnection.cpp:
(WebKit::NetworkProcessConnection::NetworkProcessConnection):
(WebKit::NetworkProcessConnection::didReceiveMessage):
* WebProcess/Network/webrtc/LibWebRTCNetwork.cpp: Added
* WebProcess/Network/webrtc/LibWebRTCNetwork.h:
(WebKit::LibWebRTCNetwork::connection):
(WebKit::LibWebRTCNetwork::~LibWebRTCNetwork):
(WebKit::LibWebRTCNetwork::networkProcessCrashed):
(WebKit::LibWebRTCNetwork::setConnection):
(WebKit::LibWebRTCNetwork::dispatchToThread):
(WebKit::LibWebRTCNetwork::signalAddressReady):
(WebKit::LibWebRTCNetwork::signalReadPacket):
(WebKit::LibWebRTCNetwork::signalSentPacket):
(WebKit::LibWebRTCNetwork::signalConnect):
(WebKit::LibWebRTCNetwork::signalClose):
(WebKit::LibWebRTCNetwork::signalNewConnection):
* WebProcess/Network/webrtc/LibWebRTCNetwork.messages.in: Added.
* WebProcess/Network/webrtc/LibWebRTCSocket.h:
* WebProcess/Network/webrtc/LibWebRTCSocketFactory.cpp:
* WebProcess/Network/webrtc/WebRTCSocket.cpp: Removed.
* WebProcess/Network/webrtc/WebRTCSocket.h: Removed.
2020-05-05 Tim Horton <timothy_horton@apple.com>
Excessive error logging from daemons trying to use WebKit, under -[UIDevice currentDevice]
https://bugs.webkit.org/show_bug.cgi?id=211397
<rdar://problem/61635403>
Reviewed by Simon Fraser.
* Shared/UserInterfaceIdiom.mm:
(WebKit::userInterfaceIdiomIsPad):
Adjust userInterfaceIdiomIsPad so that in daemons, it consults only MobileGestalt,
which returns the actual hardware model, and does not try to use UIDevice.
UIDevice is more accurate for applications because it will report that
the device is an iPhone when called inside an iPhone app running on iPad,
but it cannot be used in daemons that do not have a UIApplication.
For the behaviors we gate on this bit, it makes sense to use iPhone
behaviors on iPad in the iPhone app jail, so we continue using
UIDevice if possible.
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration init]):
Make use of the new mechanism instead of going straight to MobileGestalt,
for the aforementioned reasons.
2020-05-04 Tim Horton <timothy_horton@apple.com>
Relax WKRemoteObjectRegistry signature validation for bool-equivalent types
https://bugs.webkit.org/show_bug.cgi?id=211419
Reviewed by Saam Barati.
* Shared/API/Cocoa/_WKRemoteObjectRegistry.mm:
(blockSignaturesAreCompatible):
(replyBlockSignature):
(-[_WKRemoteObjectRegistry _invokeMethod:]):
(validateReplyBlockSignature): Deleted.
Relax reply block signature validation slightly, considering signed char and BOOL,
which are equivalent and sometimes substituted for each other, to be equal.
This is still stricter than NSXPC's validation, but this is one of the
exceptions that they make.
2020-05-04 Darin Adler <darin@apple.com>
[Mac] Remove __MAC_OS_X_VERSION_MIN_REQUIRED checks for versions older than 10.14
https://bugs.webkit.org/show_bug.cgi?id=211420
Reviewed by Alex Christensen.
* NetworkProcess/Downloads/cocoa/DownloadCocoa.mm:
(WebKit::Download::resume): Remove __MAC_OS_X_VERSION_MIN_REQUIRED >= 101400.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::NetworkSessionCocoa): Ditto.
* PluginProcess/mac/PluginProcessMac.mm:
(WebKit::PluginProcess::platformInitializePluginProcess): Ditto.
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceMain.mm:
(WebKit::XPCServiceMain): Ditto.
* UIProcess/mac/ServicesController.mm:
(WebKit::hasCompatibleServicesForItems): Ditto.
* UIProcess/mac/WebProcessProxyMac.mm:
(WebKit::WebProcessProxy::shouldAllowNonValidInjectedCode const): Ditto.
2020-05-04 Darin Adler <darin@apple.com>
Remove now-unneeded HAVE(SANDBOX_ISSUE_MACH/READ_EXTENSION_TO_PROCESS_BY_AUDIT_TOKEN)
https://bugs.webkit.org/show_bug.cgi?id=211427
Reviewed by Alex Christensen.
* Shared/Cocoa/SandboxExtensionCocoa.mm:
(WebKit::SandboxExtensionImpl::sandboxExtensionForType): Remove #if HAVE.
2020-05-04 Darin Adler <darin@apple.com>
Remove now-unneeded HAVE(AUTHORIZATION_STATUS_FOR_MEDIA_TYPE)
https://bugs.webkit.org/show_bug.cgi?id=211426
Reviewed by Alex Christensen.
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::UIClient::decidePolicyForUserMediaPermissionRequest):
Remove HAVE(AUTHORIZATION_STATUS_FOR_MEDIA_TYPE).
2020-05-04 Darin Adler <darin@apple.com>
Remove now-unneeded HAVE(ACCESSIBILITY_SUPPORT)
https://bugs.webkit.org/show_bug.cgi?id=211425
Reviewed by Alex Christensen.
* Platform/spi/Cocoa/AccessibilitySupportSPI.h: Remove HAVE(ACCESSIBILITY_SUPPORT)
and now-unnecessary C include guards.
2020-05-04 Darin Adler <darin@apple.com>
Remove now-unneeded HAVE(SEC_TRUST_EVALUATE_WITH_ERROR)
https://bugs.webkit.org/show_bug.cgi?id=211429
Reviewed by Alex Christensen.
* UIProcess/ios/fullscreen/WKFullScreenWindowControllerIOS.mm:
(-[WKFullScreenWindowController _EVOrganizationName]):
Remove HAVE(SEC_TRUST_EVALUATE_WITH_ERROR).
2020-05-04 Kate Cheney <katherine_cheney@apple.com>
REGRESSION (r260791?): Assert not reached in ResourceLoadStatisticsDatabaseStore::openAndUpdateSchemaIfNecessary() and ResourceLoadStatisticsDatabaseStore::addMissingTablesIfNecessary()
https://bugs.webkit.org/show_bug.cgi?id=211305
<rdar://problem/62737871>
Reviewed by Alex Christensen.
This patch adds a busy timeout for the ITP database to allow for
some concurrency flexibility when running many WebKitTestRunner tests, which
recreate the ITP database store very often to test both the memory and
database stores.
With the timeout, an attempt to write to the database when another
thread is holding the lock will keep attempting to insert for 5
seconds as long as the statement execution returns SQLITE_BUSY.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::openITPDatabase):
2020-05-04 Alex Christensen <achristensen@webkit.org>
TestWebKitAPI.WebKit.CustomDisplayName is a flaky timeout
https://bugs.webkit.org/show_bug.cgi?id=211299
Reviewed by Chris Dumez.
Use a serial queue instead of a concurrent queue to set the web process's display name to prevent race conditions.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::setProcessNameQueue):
(WebKit::WebProcess::updateProcessName):
(WebKit::WebProcess::updateActivePages):
2020-05-04 Adrian Perez de Castro <aperez@igalia.com>
[GTK4] Header bars no longer have title and subtitle properties
https://bugs.webkit.org/show_bug.cgi?id=211412
Reviewed by Carlos Garcia Campos.
No new tests needed.
* UIProcess/Inspector/gtk/WebKitInspectorWindow.cpp: For GTK4 use a vertical
box as title widget, with two labels acting as title and subtitle.
(webkit_inspector_window_init):
(webkitInspectorWindowSetSubtitle):
2020-05-04 Chris Dumez <cdumez@apple.com>
[iOS] Make sure TestController::statisticsResetToConsistentState() does not hang due to process suspension
https://bugs.webkit.org/show_bug.cgi?id=211421
Reviewed by Maciej Stachowiak.
We have evidence of hangs in TestController::statisticsResetToConsistentState() on iOS and we suspect it
is due to process suspension. To address the issue, this patch updates the IPCs being sent as a result of
a call to statisticsResetToConsistentState() to take a background assertion while waiting for the IPC
response.
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::setUseITPDatabase):
(WebKit::NetworkProcessProxy::resetCacheMaxAgeCapForPrevalentResources):
(WebKit::NetworkProcessProxy::resetParametersToDefaultValues):
(WebKit::NetworkProcessProxy::scheduleClearInMemoryAndPersistent):
(WebKit::NetworkProcessProxy::resetCrossSiteLoadsWithLinkDecorationForTesting):
(WebKit::NetworkProcessProxy::setShouldDowngradeReferrerForTesting):
(WebKit::NetworkProcessProxy::setShouldBlockThirdPartyCookiesForTesting):
(WebKit::NetworkProcessProxy::setShouldEnbleSameSiteStrictEnforcementForTesting):
(WebKit::NetworkProcessProxy::setFirstPartyWebsiteDataRemovalModeForTesting):
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::setShouldBlockThirdPartyCookiesForTesting):
2020-05-04 Wenson Hsieh <wenson_hsieh@apple.com>
Tapping to focus editable elements should start caret selection at word boundary
https://bugs.webkit.org/show_bug.cgi?id=211409
<rdar://problem/62869098>
Reviewed by Megan Gardner.
See WebCore/ChangeLog for more details.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::selectWithGesture):
Rewrite this using the new VisibleUnits helper function, `wordBoundaryForPositionWithoutCrossingLine`.
2020-05-04 Darin Adler <darin@apple.com>
Make __IPHONE_OS_VERSION_MIN_REQUIRED checks against old versions explicit about watchOS and tvOS
https://bugs.webkit.org/show_bug.cgi?id=211402
Reviewed by Alexey Proskuryakov.
* NetworkProcess/NetworkActivityTracker.h: Move from __IPHONE_OS_VERSION_MIN_REQUIRED
to !PLATFORM(APPLETV). There was already an explicit watchOS case here.
* Platform/spi/ios/UIKitSPI.h: Move from __IPHONE_OS_VERSION_MIN_REQUIRED to
PLATFORM(IOS) in one place, and in the other removed the conditional entirely.
THe one place here that does PLATFORM(IOS) is for a property that is not on
watchOS and tvOS at this time, related to dragging. Decided not to add a comment
about it.
* Shared/WebPreferencesDefaultValues.h: Move from __IPHONE_OS_VERSION_MIN_REQUIRED
to !PLATFORM(APPLETV). There was already an explicit watchOS case here.
* UIProcess/API/ios/WKWebViewIOS.mm:
(contentOffsetBoundedInValidRange): Move from __IPHONE_OS_VERSION_MIN_REQUIRED
to !PLATFORM(WATCHOS) && !PLATFORM(APPLETV).
(-[WKWebView activeViewLayoutSize:]): Ditto.
(-[WKWebView _updateScrollViewInsetAdjustmentBehavior]): Ditto.
* UIProcess/RemoteLayerTree/ios/RemoteLayerTreeViews.mm:
(-[WKChildScrollView initWithFrame:]): Ditto.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView gestureRecognizer:canBePreventedByGestureRecognizer:]): Ditto.
(-[WKContentView _showPlaybackTargetPicker:fromRect:routeSharingPolicy:routingContextUID:]): Ditto.
* UIProcess/ios/WKScrollView.h: Ditto.
* UIProcess/ios/WKScrollView.mm:
(-[WKScrollView initWithFrame:]): Ditto.
2020-05-04 Peng Liu <peng.liu6@apple.com>
Fix the mistake on macro HAVE_SYSTEM_FEATURE_FLAGS
https://bugs.webkit.org/show_bug.cgi?id=211416
Reviewed by Tim Horton.
We should use "#if HAVE(SYSTEM_FEATURE_FLAGS)" instead of "#if HAVE(HAVE_SYSTEM_FEATURE_FLAGS)".
And we have to use the function _os_feature_enabled_impl() instead of the macro os_feature_enabled()
because WebKit is a namespace and cannot be used in os_feature_enabled().
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultAsyncFrameAndOverflowScrollingEnabled):
(WebKit::defaultUseGPUProcessForMedia):
(WebKit::defaultRenderCanvasInGPUProcessEnabled):
(WebKit::defaultCaptureAudioInGPUProcessEnabled):
(WebKit::defaultCaptureVideoInGPUProcessEnabled):
(WebKit::defaultWebRTCCodecsInGPUProcess):
(WebKit::defaultWebGL2Enabled):
(WebKit::defaultWebGPUEnabled):
(WebKit::defaultInAppBrowserPrivacy):
(WebKit::defaultIncrementalPDFEnabled):
(WebKit::defaultWebXREnabled):
2020-05-04 Brent Fulgham <bfulgham@apple.com>
[macOS] Eliminate SecurityServer and OCSPD from the WebContent sandbox
https://bugs.webkit.org/show_bug.cgi?id=211417
<rdar://problem/46235370>
Reviewed by Per Arne Vollan.
We can now re-block access to these services.
* WebProcess/com.apple.WebProcess.sb.in:
2020-05-04 Timothy Horton <timothy_horton@apple.com>
Unreviewed, reverting r261117.
Broke lots of API tests
Reverted changeset:
"Excessive error logging from daemons trying to use WebKit,
under -[UIDevice currentDevice]"
https://bugs.webkit.org/show_bug.cgi?id=211397
https://trac.webkit.org/changeset/261117
2020-05-04 Darin Adler <darin@apple.com>
[Cocoa] Remove a few unneeded __IPHONE_OS_VERSION_MIN_REQUIRED checks
https://bugs.webkit.org/show_bug.cgi?id=211383
Reviewed by Alexey Proskuryakov.
* UIProcess/ios/forms/WKAirPlayRoutePicker.h: Removed a check against
iOS 11, in code that already explicitly checks for watchOS and tvOS.
2020-05-04 Darin Adler <darin@apple.com>
Remove HAVE(IOSURFACE) checks in Cocoa-platform-specific code
https://bugs.webkit.org/show_bug.cgi?id=211389
Reviewed by Alexey Proskuryakov.
* GPUProcess/webrtc/LibWebRTCCodecsProxy.mm:
(WebKit::LibWebRTCCodecsProxy::encodeFrame): Remove HAVE(IOSURFACE) since
this is Cocoa-specific code.
* Shared/RemoteLayerTree/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::ensureBackingStore): Ditto.
(WebKit::RemoteLayerBackingStore::clearBackingStore): Ditto.
(WebKit::RemoteLayerBackingStore::encode const): Ditto.
(WebKit::RemoteLayerBackingStore::decode): Ditto.
(WebKit::RemoteLayerBackingStore::bytesPerPixel const): Ditto.
(WebKit::RemoteLayerBackingStore::swapToValidFrontBuffer): Ditto.
(WebKit::RemoteLayerBackingStore::display): Ditto.
(WebKit::RemoteLayerBackingStore::applyBackingStoreToLayer): Ditto.
(WebKit::RemoteLayerBackingStore::setBufferVolatility): Ditto.
(WebKit::RemoteLayerBackingStore::Buffer::discard): Ditto.
(WebKit::RemoteLayerBackingStore::surfaceBufferFormat const): Ditto.
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _takeViewSnapshot]): Ditto.
(-[WKWebView _snapshotRect:intoImageOfWidth:completionHandler:]): Ditto.
* UIProcess/Cocoa/UserMediaCaptureManagerProxy.cpp: Ditto.
* UIProcess/RemoteLayerTree/RemoteLayerTreeDrawingAreaProxy.mm:
(WebKit::RemoteLayerTreeDrawingAreaProxy::RemoteLayerTreeDrawingAreaProxy): Ditto.
* UIProcess/RemoteLayerTree/RemoteLayerTreeHost.mm:
(WebKit::recursivelyMapIOSurfaceBackingStore): Ditto.
(WebKit::RemoteLayerTreeHost::mapAllIOSurfaceBackingStore): Ditto.
* WebProcess/GPU/graphics/cocoa/ImageBufferShareableIOSurfaceBackend.cpp: Ditto.
* WebProcess/GPU/graphics/cocoa/ImageBufferShareableIOSurfaceBackend.h: Ditto.
* WebProcess/GPU/webrtc/MediaRecorderPrivate.cpp:
(WebKit::MediaRecorderPrivate::sampleBufferUpdated): Ditto.
* WebProcess/Network/webrtc/LibWebRTCProvider.cpp:
(WebKit::LibWebRTCProvider::createDecoderFactory): Ditto.
* WebProcess/cocoa/UserMediaCaptureManager.cpp:
(WebKit::UserMediaCaptureManager::Source::remoteVideoSampleAvailable): Ditto.
(WebKit::UserMediaCaptureManager::remoteVideoSampleAvailable): Ditto.
2020-05-04 David Kilzer <ddkilzer@apple.com>
Use default constructor and default initializer for SecItemRequestData
<https://webkit.org/b/211399>
<rdar://problem/62861551>
Reviewed by Darin Adler.
* Shared/mac/SecItemRequestData.cpp:
- Remove default constructor implementation.
* Shared/mac/SecItemRequestData.h:
- Use default constructor for class.
- Use default initializer for m_type.
2020-05-04 Brent Fulgham <bfulgham@apple.com>
REGRESSION (r259470): Accessibility Stereo to Mono is not working
https://bugs.webkit.org/show_bug.cgi?id=211408
<rdar://problem/62847299>
Reviewed by Per Arne Vollan.
The changes in r259470 regressed our Accessibility features. We need to do a rollout of that
change to restore proper behavior.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* WebProcess/com.apple.WebProcess.sb.in:
2020-05-04 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Implement +[_WKWebAuthenticationPanel clearAllLocalAuthenticatorCredentials]
https://bugs.webkit.org/show_bug.cgi?id=211369
<rdar://problem/60246635>
Reviewed by Brent Fulgham.
Adds a new SPI for UI clients to clear all existing credentials from the keychain.
* UIProcess/API/Cocoa/_WKWebAuthenticationPanel.h:
* UIProcess/API/Cocoa/_WKWebAuthenticationPanel.mm:
(+[_WKWebAuthenticationPanel clearAllLocalAuthenticatorCredentials]):
* UIProcess/WebAuthentication/Cocoa/LocalAuthenticator.h:
* UIProcess/WebAuthentication/Cocoa/LocalAuthenticator.mm:
(WebKit::LocalAuthenticator::clearAllCredentials):
* UIProcess/WebAuthentication/Cocoa/LocalConnection.mm:
(WebKit::LocalConnection::createCredentialPrivateKey const):
2020-05-04 Tim Horton <timothy_horton@apple.com>
Excessive error logging from daemons trying to use WebKit, under -[UIDevice currentDevice]
https://bugs.webkit.org/show_bug.cgi?id=211397
<rdar://problem/61635403>
Reviewed by Simon Fraser.
* Shared/UserInterfaceIdiom.mm:
(WebKit::userInterfaceIdiomIsPad):
Adjust userInterfaceIdiomIsPad so that in daemons, it consults only MobileGestalt,
which returns the actual hardware model, and does not try to use UIDevice.
UIDevice is more accurate for applications because it will report that
the device is an iPhone when called inside an iPhone app running on iPad,
but it cannot be used in daemons that do not have a UIApplication.
For the behaviors we gate on this bit, it makes sense to use iPhone
behaviors on iPad in the iPhone app jail, so we continue using
UIDevice if possible.
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration init]):
Make use of the new mechanism instead of going straight to MobileGestalt,
for the aforementioned reasons.
2020-05-04 Chris Dumez <cdumez@apple.com>
Stop hard-coding get-task-allow entitlement for simulator builds
https://bugs.webkit.org/show_bug.cgi?id=211392
Reviewed by Geoffrey Garen.
Stop hard-coding get-task-allow entitlement for simulator builds and let Xcode inject it as needed.
* Configurations/BaseXPCService.xcconfig:
* Resources/ios/XPCService-embedded-simulator.entitlements:
2020-05-04 Brent Fulgham <bfulgham@apple.com>
REGRESSION (JazzkonE): ResourceLoadStatisticsDatabaseStore checks 'hasHadUserInteraction' without ensuring the domain has been added to the ITP database (211388)
https://bugs.webkit.org/show_bug.cgi?id=211388
<rdar://problem/62849919>
Reviewed by John Wilander.
A spurious log entry is created if we check 'hasHadUserInteraction' before inserting a row for the domain
being checked (if that domain had not previosly been observed). This patch changes the order of the two
commands to avoid this possibility. This makes the database implementation behave like the in-memory version.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::logUserInteraction):
2020-05-04 Alex Christensen <achristensen@webkit.org>
Reduce AlternativeService assertion to debug-only assert
https://bugs.webkit.org/show_bug.cgi?id=211263
<rdar://problem/61354813>
Reviewed by Chris Dumez.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::NetworkSessionCocoa):
This is asserting sometimes. No need to crash. Reduce to debug-only assert.
2020-05-04 Said Abou-Hallawa <sabouhallawa@apple.com>
Throttling requestAnimationFrame should be controlled by RenderingUpdateScheduler
https://bugs.webkit.org/show_bug.cgi?id=204713
Reviewed by Simon Fraser.
Create an IPC message on the DrawingArea to send a message from the
WebProcess to the UIProcess to setPreferredFramesPerSecond of the
DisplayRefreshMonitor.
* UIProcess/RemoteLayerTree/RemoteLayerTreeDrawingAreaProxy.h:
* UIProcess/RemoteLayerTree/RemoteLayerTreeDrawingAreaProxy.messages.in:
* UIProcess/RemoteLayerTree/RemoteLayerTreeDrawingAreaProxy.mm:
(-[WKOneShotDisplayLinkHandler setPreferredFramesPerSecond:]):
(WebKit::RemoteLayerTreeDrawingAreaProxy::setPreferredFramesPerSecond):
Set the preferredFramesPerSecond of the CADisplayLink.
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDisplayRefreshMonitor.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDisplayRefreshMonitor.mm:
(WebKit::RemoteLayerTreeDisplayRefreshMonitor::setPreferredFramesPerSecond):
Forward the call to RemoteLayerTreeDrawingArea.
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.mm:
(WebKit::RemoteLayerTreeDrawingArea::setPreferredFramesPerSecond):
Send the IPC message from the WebProcess to the UIProcess.
2020-05-04 Alex Christensen <achristensen@webkit.org>
TestWebKitAPI.WebKit.CustomDisplayName is a flaky timeout
https://bugs.webkit.org/show_bug.cgi?id=211299
Reviewed by Chris Dumez.
Sometimes the task queued in WebProcess::updateProcessName happens after the one in WebProcess::updateActivePages.
Schedule them on the same queue so they happen in deterministic order so we get reliable process names eventually.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::updateActivePages):
2020-05-04 Brent Fulgham <bfulgham@apple.com>
Remove unneeded check for Apple Media Services
https://bugs.webkit.org/show_bug.cgi?id=211391
<rdar://problem/62853199>
Reviewed by Per Arne Vollan.
Earlier in the development cycle we added an entitlement check for AppleMediaServices, that we now realize
we do not need. Since this is a non-zero performance cost, remove this check.
* UIProcess/WebPageProxy.cpp:
2020-05-04 Devin Rousso <drousso@apple.com>
Web Inspector: provide a way for inspector to turn on/off ITP debug mode and AdClickAttribution debug mode
https://bugs.webkit.org/show_bug.cgi?id=209763
Reviewed by Brian Burg.
Generalize the `setMockCaptureDevicesEnabledOverride` to be `setDeveloperPreferenceOverride`
that uses an enum `WebCore::InspectorClient::DeveloperPreference` to know what to do.
Communicate with the NetworkProcess (from the WebProcess via the UIProcess) in the case of:
- `AdClickAttributionDebugModeEnabled`
- `ITPDebugModeEnabled`
* WebProcess/Inspector/WebInspectorClient.h:
* WebProcess/Inspector/WebInspectorClient.cpp:
(WebKit::WebInspectorClient::setDeveloperPreferenceOverride): Added.
(WebKit::WebInspectorClient::setMockCaptureDevicesEnabledOverride): Deleted.
* WebProcess/Inspector/WebInspector.h:
* WebProcess/Inspector/WebInspector.cpp:
(WebKit::WebInspector::setDeveloperPreferenceOverride): Added.
(WebKit::WebInspector::setMockCaptureDevicesEnabledOverride): Deleted.
* UIProcess/Inspector/WebInspectorProxy.messages.in:
* UIProcess/Inspector/WebInspectorProxy.h:
* UIProcess/Inspector/WebInspectorProxy.cpp:
(WebKit::WebInspectorProxy::setDeveloperPreferenceOverride): Added.
(WebKit::WebInspectorProxy::setMockCaptureDevicesEnabledOverride): Deleted.
* UIProcess/WebsiteData/WebsiteDataStore.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::setAdClickAttributionDebugMode): Added.
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::setAdClickAttributionDebugMode): Added.
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::initializeNetworkProcess):
(WebKit::NetworkProcess::setAdClickAttributionDebugMode): Added.
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.cpp:
(WebKit::ResourceLoadStatisticsStore::setResourceLoadStatisticsDebugMode):
2020-05-04 Chris Dumez <cdumez@apple.com>
[iOS] Unable to take RunningBoard process assertions in the iOS Simulator
https://bugs.webkit.org/show_bug.cgi?id=211254
<rdar://problem/62674074>
Reviewed by Geoffrey Garen.
Add com.apple.runningboard.assertions.webkit entitlement to iOS Simulator builds so that
we can take RunningBoard assertions on those platforms too.
While doing this, I found that we were not adding the entitlements properly for simulator
builds. For simulator builds, entitlements go into an __entitlements section of the binary,
rather than in the signature. Use CODE_SIGN_ENTITLEMENTS directive in xcconfig so that
XCode does the right thing for simulator builds.
* Configurations/BaseXPCService.xcconfig:
* Resources/ios/XPCService-embedded-simulator.entitlements: Added.
* Scripts/process-entitlements.sh:
2020-05-04 Darin Adler <darin@apple.com>
Remove now-unneded HAVE(WINDOW_SERVER_OCCLUSION_NOTIFICATIONS)
https://bugs.webkit.org/show_bug.cgi?id=211380
Reviewed by Sam Weinig.
* UIProcess/mac/WindowServerConnection.h: Don't check
HAVE(WINDOW_SERVER_OCCLUSION_NOTIFICATIONS) in this Mac-only header.
Also removed #pragma once in this header that is only included from .mm files.
* UIProcess/mac/WindowServerConnection.mm: Wrapped the whole file in
PLATFORM(MAC) because this is macOS-specific code that may be compiled on
iOS family platforms. Could wrap it in HAVE(WINDOW_SERVER) instead, but we
don't have anything like that and PageClientImplMac.mm, the client of this
code, uses PLATFORM(MAC).
(WebKit::registerOcclusionNotificationHandler): Ditto.
(WebKit::WindowServerConnection::WindowServerConnection): Ditto.
2020-05-04 Darin Adler <darin@apple.com>
Remove now-unneeded HAVE(UI_REMOTE_VIEW)
https://bugs.webkit.org/show_bug.cgi?id=211382
Reviewed by Alex Christensen.
* UIProcess/RemoteLayerTree/ios/RemoteLayerTreeHostIOS.mm:
(WebKit::createRemoteView): Deleted.
(WebKit::RemoteLayerTreeHost::makeNode): After researching to be sure the method
is present in the relevant versions of iOS 13, changed this code to use the
initWithFrame:pid:contextID: method without doing a selector check, and also
without a HAVE(UI_REMOTE_VIEW) conditional.
* UIProcess/RemoteLayerTree/ios/RemoteLayerTreeViews.mm: Removed the
HAVE(UI_REMOTE_VIEW) conditional.
2020-05-04 Darin Adler <darin@apple.com>
Remove now-unneeded HAVE(MENU_CONTROLLER_SHOW_HIDE_API)
https://bugs.webkit.org/show_bug.cgi?id=211381
Reviewed by Alex Christensen.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _handleDOMPasteRequestWithResult:]): Use hideMenuFromView: directly.
(-[WKContentView _requestDOMPasteAccessWithElementRect:originIdentifier:completionHandler:]):
Use showMenuFromView: directly.
(-[WKContentView showGlobalMenuControllerInRect:]): Deleted.
(-[WKContentView hideGlobalMenuController]): Deleted.
2020-05-04 Darin Adler <darin@apple.com>
Remove now-unneeded HAVE(VOUCHERS)
https://bugs.webkit.org/show_bug.cgi?id=211379
Reviewed by Alex Christensen.
* Shared/EntryPointUtilities/Cocoa/XPCService/XPCServiceEntryPoint.h:
(WebKit::XPCServiceInitializer): Remove HAVE(VOUCHERS) checks from this
Cocoa-only source file.
2020-05-04 Alex Christensen <achristensen@webkit.org>
TestWebKitAPI.WebKit.CustomDisplayName is a flaky timeout
https://bugs.webkit.org/show_bug.cgi?id=211299
Reviewed by Darin Adler.
The application name is being set on a background queue, but there's nothing actually keeping the NSString alive.
This may fix some timeouts we were seeing. I wonder why ASAN or guard malloc didn't catch this yet.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::updateProcessName):
2020-05-04 Darin Adler <darin@apple.com>
Remove no-longer-needed HAVE(UISCENE)
https://bugs.webkit.org/show_bug.cgi?id=211376
Reviewed by Chris Dumez.
* Platform/spi/ios/UIKitSPI.h: Don't check HAVE(UISCENE) any more in this
iOS-family-only source file; it's always true.
* UIProcess/ApplicationStateTracker.h: Ditto.
* UIProcess/ApplicationStateTracker.mm:
(WebKit::ApplicationStateTracker::ApplicationStateTracker): Ditto.
(WebKit::ApplicationStateTracker::~ApplicationStateTracker): Ditto.
* UIProcess/ios/WKContentView.mm:
(-[WKContentView interfaceOrientation]): Ditto.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView textEffectsWindow]): Ditto.
2020-05-04 Yusuke Suzuki <ysuzuki@apple.com>
Unreviewed, speculative build fix for r261071
https://bugs.webkit.org/show_bug.cgi?id=211274
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
2020-05-04 Emilio Cobos Álvarez <emilio@crisal.io>
Put lh / rlh units behind a flag until bug 211351 is sorted out.
https://bugs.webkit.org/show_bug.cgi?id=211356
Reviewed by Antti Koivisto.
* Shared/WebPreferences.yaml: Define the runtime flag.
2020-05-03 David Kilzer <ddkilzer@apple.com>
Use LocalCurrentGraphicsContext in WebKit::convertPlatformImageToBitmap()
<https://webkit.org/b/211274>
Reviewed by Darin Adler.
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::convertPlatformImageToBitmap):
- Use LocalCurrentGraphicsContext to replace code.
2020-05-03 David Kilzer <ddkilzer@apple.com>
Use default initializers and default constructors in WebEvent.h
<https://webkit.org/b/211354>
Reviewed by Daniel Bates.
* Shared/WebEvent.h:
(WebKit::WebWheelEvent):
(WebKit::WebKeyboardEvent):
(WebKit::WebPlatformTouchPoint):
(WebKit::WebTouchEvent):
- Change empty constructors to use `default`.
- Use default initializers to make sure all fields are
initialized.
2020-05-02 Simon Fraser <simon.fraser@apple.com>
handleWheelEventPhase() should include the relevant ScrollingNodeID
https://bugs.webkit.org/show_bug.cgi?id=211315
Reviewed by Tim Horton.
handleWheelEventPhase() is used to send information about wheel event phases
to the main thread, which make their way to ScrollAnimatorMac::handleWheelEventPhase()
and are used to update the state of overlay scrollbars. In order to talk to the
correct set of scrollbars with overflow:scroll, we need to send along the ScrollingNodeID
and map that to the appropriate ScrollableArea.
Will be tested by future overlay scrollbar tests.
* UIProcess/RemoteLayerTree/RemoteScrollingTree.cpp:
(WebKit::RemoteScrollingTree::handleWheelEventPhase):
* UIProcess/RemoteLayerTree/RemoteScrollingTree.h:
2020-05-02 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Re-inject key press events not handled by the web process
https://bugs.webkit.org/show_bug.cgi?id=211286
Reviewed by Adrian Perez de Castro.
Use gdk_display_put_event() since gtk_main_do_event() is no longer available in GTK4. Also stop forwarding key
release events in GTK3 case, this only makes sense for key press events and we were re-injecting every release
event since they are not handled by web elements in most of the cases.
* UIProcess/API/gtk/PageClientImpl.cpp:
(WebKit::PageClientImpl::doneWithKeyEvent):
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseKeyPressEvent):
(webkitWebViewBaseKeyReleaseEvent):
(webkitWebViewBaseKeyPressed):
2020-05-02 Adrian Perez de Castro <aperez@igalia.com>
[GTK] Specify action group name when binding context menu models
https://bugs.webkit.org/show_bug.cgi?id=211288
Reviewed by Carlos Garcia Campos.
No new tests needed.
* UIProcess/gtk/WebContextMenuProxyGtk.cpp:
(WebKit::WebContextMenuProxyGtk::append): Use the action name directly for the menu item.
(WebKit::WebContextMenuProxyGtk::populate): Indicate the action group name when binding the
menu model, which automatically adds the group name as action name prefix without needing
to specify it by hand when adding menu items.
2020-05-02 Devin Rousso <drousso@apple.com>
[CSS Easing 1] implement `jump-*` step positions
https://bugs.webkit.org/show_bug.cgi?id=211271
Reviewed by Dean Jackson.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<StepsTimingFunction>::encode):
(IPC::ArgumentCoder<StepsTimingFunction>::decode):
2020-05-01 Tim Horton <timothy_horton@apple.com>
Books sometimes ends up with blank pages, especially after adjusting font size
https://bugs.webkit.org/show_bug.cgi?id=211265
<rdar://problem/59898144>
Reviewed by Darin Adler.
A few problems:
- There is short time during page creation where a WKWebView created
with _clipsToVisibleRect=YES would not yet have sent its viewExposedRect
to the Web Content process, and if we end up constructing tiles during
that time, we can make way too many, bogging down the process (or crashing).
Fix this by always keeping track of the viewExposedRect (on WebPageProxy,
instead of the somewhat-more-transient DrawingAreaProxy) and sending it
to the Web Content process in the WebPage creation parameters, to entirely
remove this window.
- Even when the viewExposedRect successfully gets to the Web Content
process, it can still end up wildly wrong: the DrawingArea was tasked with
watching scrolling changes, applying the scroll offset to the viewExposedRect,
and pushing it to FrameView in content coordinates.
It turns out that this was all unnecessary, as we need viewExposedRect
in root view coordinates (same space as visibleContentRect, which we
intersect it with), and we just didn't notice because all clients of
_clipsToVisibleRect: expand the view to its layout size and insert
a scrolling view outside the web view, and so don't use our scrolling.
Avoid this conversion and complexity entirely; I tested Mail and Books
where there is no impact (other than fixing the original bug), and also
a custom test app with a scrollable WKWebView inside a NSScrollView,
which improved significantly.
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
Plumb view exposed rect via WebPage creation parameters.
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::updateViewExposedRect):
Send viewExposedRect changes to WebPageProxy instead of DrawingAreaProxy.
* UIProcess/DrawingAreaProxy.cpp:
(WebKit::DrawingAreaProxy::didChangeViewExposedRect):
(WebKit::DrawingAreaProxy::viewExposedRectChangedTimerFired):
(WebKit::DrawingAreaProxy::setViewExposedRect): Deleted.
* UIProcess/DrawingAreaProxy.h:
(WebKit::DrawingAreaProxy::viewExposedRect const): Deleted.
* UIProcess/RemoteLayerTree/RemoteLayerTreeDrawingAreaProxy.h:
* UIProcess/RemoteLayerTree/RemoteLayerTreeDrawingAreaProxy.mm:
(WebKit::RemoteLayerTreeDrawingAreaProxy::didChangeViewExposedRect):
(WebKit::RemoteLayerTreeDrawingAreaProxy::indicatorLocation const):
(WebKit::RemoteLayerTreeDrawingAreaProxy::updateDebugIndicator):
(WebKit::RemoteLayerTreeDrawingAreaProxy::setViewExposedRect): Deleted.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::setViewExposedRect):
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::viewExposedRect const):
Maintain viewExposedRect on WebPageProxy instead of DrawingAreaProxy,
so that we can always store it even if we don't have a DrawingAreaProxy yet
(or change DrawingAreaProxies) and can send it in WebPage creation parameters.
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.mm:
(WebKit::RemoteLayerTreeDrawingArea::RemoteLayerTreeDrawingArea):
(WebKit::RemoteLayerTreeDrawingArea::setViewExposedRect):
(WebKit::RemoteLayerTreeDrawingArea::updateRendering):
(WebKit::RemoteLayerTreeDrawingArea::updateScrolledExposedRect): Deleted.
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.h:
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::TiledCoreAnimationDrawingArea):
(WebKit::TiledCoreAnimationDrawingArea::updateRendering):
(WebKit::TiledCoreAnimationDrawingArea::setViewExposedRect):
(WebKit::TiledCoreAnimationDrawingArea::scroll): Deleted.
(WebKit::TiledCoreAnimationDrawingArea::updateScrolledExposedRect): Deleted.
Plumb viewExposedRect directly to FrameView, instead of trying to apply
the root view -> contents mapping ourselves and pushing updates.
2020-05-01 David Kilzer <ddkilzer@apple.com>
[IPC hardening] Refactor createMessageDecoder() for clarity
<https://webkit.org/b/211322>
Reviewed by Darin Adler.
* Platform/IPC/cocoa/ConnectionCocoa.mm:
(IPC::createMessageDecoder):
- Rename `numDescriptors` to `numberOfPortDescriptors` to match
variable name in sendOutgoingMessage().
- Add new `numberOfAttachments` variable to make it clear that
one port descriptor is left for an out-of-line message body.
- Add FIXME about another issue.
2020-05-01 Alex Christensen <achristensen@webkit.org>
Add SPI to move localStorage to a different domain
https://bugs.webkit.org/show_bug.cgi?id=209260
<rdar://problem/60285683>
Reviewed by Brady Eidson.
Covered by an API test.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::renameDomainInWebsiteData):
(WebKit::NetworkProcess::getLocalStorageOriginDetails):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/WebStorage/LocalStorageNamespace.cpp:
(WebKit::LocalStorageNamespace::renameDomain):
* NetworkProcess/WebStorage/LocalStorageNamespace.h:
* NetworkProcess/WebStorage/StorageArea.cpp:
(WebKit::StorageArea::close):
* NetworkProcess/WebStorage/StorageArea.h:
* NetworkProcess/WebStorage/StorageManager.cpp:
(WebKit::StorageManager::renameDomain):
* NetworkProcess/WebStorage/StorageManager.h:
* NetworkProcess/WebStorage/StorageManagerSet.cpp:
(WebKit::StorageManagerSet::renameDomain):
* NetworkProcess/WebStorage/StorageManagerSet.h:
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
(-[WKWebsiteDataStore _renameDomain:to:forDataOfTypes:completionHandler:]):
* UIProcess/API/Cocoa/WKWebsiteDataStorePrivate.h:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::renameDomainInWebsiteData):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::renameDomainInWebsiteData):
* UIProcess/WebsiteData/WebsiteDataStore.h:
2020-05-01 Chris Dumez <cdumez@apple.com>
[iOS] ProcessThrottler fails to re-take ProcessAssertion if the previous one was invalidated
https://bugs.webkit.org/show_bug.cgi?id=211297
<rdar://problem/62542463>
Reviewed by Jer Noble.
Our ProcessAssertions may get invalidated upon backgrounding of the app. When the app becomes
foreground and the ProcessThrottler tries to take a Foreground assertion as a result, it would
incorrectly think it already had such assertion and not do anything, even though the previous
one is no longer valid. As a result, the child processes would stay suspended even though the
app was foregrounded.
To address the issue, add a isValid() method to ProcessAssertion() and check it in
ProcessThrottler::setAssertionType() to determine if we need to re-take an assertion or not.
We also invalidate all pending ProcessThrottler activities upon ProcessAssertion invalidation
for good measure. This way, the holders of these activities will be able to rely on
Activity::isValid() to determine if they need to re-take their activities or not.
* Platform/spi/ios/AssertionServicesSPI.h:
* Platform/spi/ios/RunningBoardServicesSPI.h:
* UIProcess/ProcessAssertion.cpp:
(WebKit::ProcessAssertion::isValid const):
* UIProcess/ProcessAssertion.h:
(WebKit::ProcessAssertion::validity const): Deleted.
* UIProcess/ProcessThrottler.cpp:
(WebKit::ProcessThrottler::setAssertionType):
(WebKit::ProcessThrottler::assertionWasInvalidated):
* UIProcess/ProcessThrottler.h:
* UIProcess/ios/ProcessAssertionIOS.mm:
(WebKit::ProcessAssertion::processAssertionWasInvalidated):
(WebKit::ProcessAssertion::isValid const):
(WebKit::ProcessAndUIAssertion::updateRunInBackgroundCount):
2020-05-01 Chris Dumez <cdumez@apple.com>
Unreviewed, reverting r261015.
Seems to have broken clean builds
Reverted changeset:
"[iOS] Unable to take RunningBoard process assertions in the
iOS Simulator"
https://bugs.webkit.org/show_bug.cgi?id=211254
https://trac.webkit.org/changeset/261015
2020-05-01 David Kilzer <ddkilzer@apple.com>
[IPC hardening] createMessageDecoder() needs a validity check
<https://webkit.org/b/211260>
<rdar://problem/61914087>
Reviewed by Darin Adler.
* Platform/IPC/cocoa/ConnectionCocoa.mm:
(IPC::createMessageDecoder):
- Add validity check and early return for
`numDescriptors`.
2020-05-01 Devin Rousso <drousso@apple.com>
Web Inspector: Browser: crash when handling `disable`
https://bugs.webkit.org/show_bug.cgi?id=211251
Reviewed by Daniel Bates.
* UIProcess/Inspector/Agents/InspectorBrowserAgent.cpp:
(WebKit::InspectorBrowserAgent::disable):
If the inspected page crashed, we won't have a `WebInspectorProxy` anymore.
2020-05-01 Per Arne Vollan <pvollan@apple.com>
[iOS] Every running WebContent process should be granted access to frontboard services when Accessibility is enabled
https://bugs.webkit.org/show_bug.cgi?id=211238
Reviewed by Daniel Bates.
Currently, every WebContent process is granted access to frontboard services if Accessibility is enabled at the
time of startup. However, WebContent processes running at the time when Accessibility is enabled are not granted
access, which is a bug.
API test: WebKit.AccessibilityHasFrontboardServiceAccess
* UIProcess/AuxiliaryProcessProxy.cpp:
(WebKit::AuxiliaryProcessProxy::sendMessage):
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
* UIProcess/Cocoa/WebProcessProxyCocoa.mm:
(WebKit::WebProcessProxy::unblockAccessibilityServerIfNeeded):
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::unblockServicesRequiredByAccessibility):
(WebKit::WebProcess::unblockAccessibilityServer): Deleted.
2020-05-01 Chris Dumez <cdumez@apple.com>
[iOS] Unable to take RunningBoard process assertions in the iOS Simulator
https://bugs.webkit.org/show_bug.cgi?id=211254
<rdar://problem/62674074>
Reviewed by Tim Horton.
* Configurations/BaseXPCService.xcconfig:
Use this directive to add our entitlements:
CODE_SIGN_ENTITLEMENTS = $(WK_PROCESSED_XCENT_FILE);
instead of using:
OTHER_CODE_SIGN_FLAGS = --entitlements $(WK_PROCESSED_XCENT_FILE);
This is important because entitlements are added differently for iOS Simulator builds,
so that they only apply inside the simulator (and not for the host system). If we use
CODE_SIGN_ENTITLEMENTS, then XCode does the right thing on all platforms.
* Scripts/process-entitlements.sh:
Add com.apple.runningboard.assertions.webkit entitlement to iOS Simulator builds.
2020-05-01 Don Olmstead <don.olmstead@sony.com>
[GTK] Add additional exports to support hidden visibility
https://bugs.webkit.org/show_bug.cgi?id=211246
Reviewed by Michael Catanzaro.
* WebProcess/InjectedBundle/API/glib/WebKitWebExtensionPrivate.h:
2020-05-01 Per Arne Vollan <pvollan@apple.com>
[iOS] Update message filtering rules in the WebContent process' sandbox
https://bugs.webkit.org/show_bug.cgi?id=211188
<rdar://problem/60922910>
Reviewed by Brent Fulgham.
Based on telemetry and local testing, update the message filtering rules in the WebContent process' sandbox on iOS.
Messages that have not been observed being in use, should be denied.
No new tests, covered by existing tests.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* UIProcess/AuxiliaryProcessProxy.cpp:
(WebKit::AuxiliaryProcessProxy::sendMessage):
2020-05-01 Peng Liu <peng.liu6@apple.com>
A PiP window doesn’t actually dismiss after the browser navigates to a different page within the same domain
https://bugs.webkit.org/show_bug.cgi?id=211257
Reviewed by Jer Noble.
Add the support of exitVideoFullscreenToModeWithoutAnimation() in iOS,
so that the Web process can request the UI process to close the PiP window
without exchanging IPC messages back and forth.
* UIProcess/Cocoa/VideoFullscreenManagerProxy.h:
* UIProcess/Cocoa/VideoFullscreenManagerProxy.messages.in:
* UIProcess/Cocoa/VideoFullscreenManagerProxy.mm:
(WebKit::VideoFullscreenManagerProxy::exitFullscreenWithoutAnimationToMode):
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
* WebProcess/WebCoreSupport/WebChromeClient.h:
Only stop the watchdog timer if it is active.
* WebProcess/cocoa/VideoFullscreenManager.mm:
(WebKit::VideoFullscreenManager::exitVideoFullscreenToModeWithoutAnimation):
2020-05-01 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r260920.
https://bugs.webkit.org/show_bug.cgi?id=211292
Sandbox compile error is no longer an issue (Requested by
perarne on #webkit).
Reverted changeset:
"Unreviewed sandbox compile fix."
https://trac.webkit.org/changeset/260920
2020-05-01 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] Move focus out of the web view when tab is pressed and there isn't a next element to focus
https://bugs.webkit.org/show_bug.cgi?id=211283
Reviewed by Adrian Perez de Castro.
Implement API::UIClient::takeFocus() that is called when WebCore wants to focus the chrome.
* UIProcess/API/glib/WebKitUIClient.cpp:
2020-05-01 David Kilzer <ddkilzer@apple.com>
REGRESSION (r260214): Leak of _WKTextInputContext in -[WKContentView _isTextInputContextFocused:]
<https://webkit.org/b/211276>
<rdar://problem/62696724>
Reviewed by Daniel Bates.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _focusTextInputContext:placeCaretAt:completionHandler:]):
- Use RetainPtr<> to pass the copy of _WKTextInputContext into
the block to fix the leak.
2020-05-01 Saam Barati <sbarati@apple.com>
We can't cast toLength result to unsigned
https://bugs.webkit.org/show_bug.cgi?id=211205
<rdar://problem/62625562>
Reviewed by Yusuke Suzuki.
* WebProcess/Plugins/Netscape/NPJSObject.cpp:
(WebKit::NPJSObject::hasProperty):
(WebKit::NPJSObject::getProperty):
(WebKit::NPJSObject::removeProperty):
2020-05-01 Adrian Perez de Castro <aperez@igalia.com>
[GTK4] Disable arrow on context menu popover
https://bugs.webkit.org/show_bug.cgi?id=211241
Reviewed by Carlos Garcia Campos.
* UIProcess/gtk/WebContextMenuProxyGtk.cpp:
(WebKit::WebContextMenuProxyGtk::WebContextMenuProxyGtk): Disable the menu popover arrow.
2020-04-30 Ross Kirsling <ross.kirsling@sony.com>
TriState should be an enum class and use "Indeterminate" instead of "Mixed"
https://bugs.webkit.org/show_bug.cgi?id=211268
Reviewed by Mark Lam.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::validateCommand):
* WebProcess/WebPage/glib/WebPageGLib.cpp:
(WebKit::WebPage::getPlatformEditorState const):
2020-04-30 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Require user gestures for LocalAuthenticator
https://bugs.webkit.org/show_bug.cgi?id=182893
<rdar://problem/43357293>
Reviewed by Brent Fulgham.
Covered by API tests.
* UIProcess/WebAuthentication/AuthenticatorManager.cpp:
(WebKit::AuthenticatorManager::filterTransports const):
(WebKit::AuthenticatorManager::getTransports const):
Checks UserGesture here. Since filterTransports will be overrided in mock testings,
all mock tests don't require user gestures.
* UIProcess/WebAuthentication/WebAuthenticationRequestData.h:
* UIProcess/WebAuthentication/WebAuthenticatorCoordinatorProxy.cpp:
(WebKit::WebAuthenticatorCoordinatorProxy::makeCredential):
(WebKit::WebAuthenticatorCoordinatorProxy::getAssertion):
* UIProcess/WebAuthentication/WebAuthenticatorCoordinatorProxy.h:
* UIProcess/WebAuthentication/WebAuthenticatorCoordinatorProxy.messages.in:
* WebProcess/WebAuthentication/WebAuthenticatorCoordinator.cpp:
(WebKit::WebAuthenticatorCoordinator::makeCredential):
(WebKit::WebAuthenticatorCoordinator::getAssertion):
Adds user gesture corresponding fields.
2020-04-30 Wenson Hsieh <wenson_hsieh@apple.com>
Unreviewed, fix the Mac Catalyst build after r260979
The presence of `_contextMenuInteraction` is conditional not only on `USE(UICONTEXTMENU)`, but also
`HAVE(LINK_PREVIEW)`. Guard logic in `-deferringGestureRecognizer:shouldDeferOtherGestureRecognizer:` with the
latter as well.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView deferringGestureRecognizer:shouldDeferOtherGestureRecognizer:]):
2020-04-30 Simon Fraser <simon.fraser@apple.com>
Clean up some EventHandler coordinate-related naming and fix ScrollableArea::lastKnownMousePosition() conversions
https://bugs.webkit.org/show_bug.cgi?id=211259
Reviewed by Zalan Bujtas.
* Shared/WebEvent.h:
(WebKit::WebMouseEvent::position const):
* Shared/WebMouseEvent.cpp:
(WebKit::WebMouseEvent::WebMouseEvent):
* WebProcess/Plugins/PDF/PDFPlugin.h:
2020-04-30 David Kilzer <ddkilzer@apple.com>
[iOS] Fix confusing idiom for releasing _selectionRects in -[WKTextRange dealloc]
<https://webkit.org/b/211269>
Reviewed by Wenson Hsieh.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKTextRange dealloc]):
- Switch to more common idiom to release an instance variable.
This fixes a false positive warning from the clang static
analyzer about releasing a variable not owned by the caller.
2020-04-30 Wenson Hsieh <wenson_hsieh@apple.com>
REGRESSION: Double tap dispatches one click event on iOS 13.4 when it would dispatch two on iOS 13.3
https://bugs.webkit.org/show_bug.cgi?id=211179
<rdar://problem/62594779>
Reviewed by Tim Horton.
https://trac.webkit.org/r253267 introduced deferring gesture recognizers as a way to handle preventable (non-
passive) touchstart events without blocking the UI process. These deferring gesture recognizers work by having
other gesture recognizers at or below WKWebView (with few exceptions) require the failure of these deferring
gestures. These gestures transition to possible state when beginning a touch inside a non-passive touch event
handling region, and transition to either failed or ended state (depending on whether `preventDefault()` was
called) after the web content process finished handling the touch event.
However, this means that the resulting dependency graph now has an edge between each gesture under WKWebView and
one of the deferring gesture recognizers, due to these new failure requirements. Since gestures that have been
recognized or have failed don't get reset until all other gestures in the same dependency subgraph have also
recognized or failed, some gestures (such as the synthetic single tap gesture recognizer in this bug) might not
be resetting as soon after ending as they did before, since they may be connected to other gesture recognizers
that are still in possible state by way of the failure requirements added by the new deferring gestures.
I was already aware of this problem in r253267, and attempted to solve this by bisecting the gesture dependency
graph into two subgraphs: one containing all the one-finger multi-tap gestures that are reset after a lengthy
delay, and another containing everything else. To do this, I added two different deferring gesture recognizers:
one for immediately resettable gestures (meant for gestures in the latter subgraph), and another for gestures
that are reset after a delay (meant for gestures in the former subgraph).
Unfortunately, unrelated changes around text interactions in UIKit in iOS 13.4 caused the tap-and-a-half
gesture for selecting a range of text to now have a delayed reset; this means that gestures in the "immediately
resettable" gesture subgraph are all forced to wait on the tap-and-a-half text interaction gesture before
resetting, which causes the bug here, since the synthetic single tap gesture is in this "immediately resettable"
gesture subgraph.
To mitigate this, this patch pushes the tap-and-a-half text selection gesture, along with the loupe gesture,
context menu relationship gesture, and drag lift gestures (i.e. the other gestures that are connected to the
tap-and-a-half gesture via failure requirements) out of the "immediately resettable" subgraph, and into the
"delayed" subgraph.
Test: fast/events/touch/ios/two-click-events-after-double-tap-with-touch-handler.html
* Platform/spi/ios/UIKitSPI.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView deferringGestureRecognizer:shouldDeferOtherGestureRecognizer:]):
2020-04-30 Brent Fulgham <bfulgham@apple.com>
REGRESSION (r260932): ASSERTION FAILED: ok in webkit::SandboxExtension::consumePermanently for 3 tests
https://bugs.webkit.org/show_bug.cgi?id=211253
<rdar://problem/62673727>
Reviewed by Per Arne Vollan.
In r260932 I added an assertion to confirm that a sandbox extension was successfully consumed. This triggered three
test failures because those tests attempt to pass a sandbox extension from the WebContent process to the UIProcess.
After speaking with the Sandbox team, I learned that an unsandboxed program that attempts to consume a valid
sandbox extension will return 0. Sandboxed processes either receive -1 (for error), or a handle representing the
newly added extension (so that it can be removed when no longer needed).
Since the TestWebKitAPI (and WebKitTestRunner) processes are not sandboxed they receive a zero return value here.
We should recognize this as a valid and expected result. We should also consider making some form of sandboxed
driver for WebKit.
* Shared/Cocoa/SandboxExtensionCocoa.mm:
(WebKit::SandboxExtensionImpl::consume): Recognize a 0 handle as valid.
2020-04-30 Jiewen Tan <jiewen_tan@apple.com>
[WebAuthn] Optimize LocalAuthenticator
https://bugs.webkit.org/show_bug.cgi?id=183534
<rdar://problem/43357408>
Reviewed by Brent Fulgham.
This patch implements the following small optimizations:
1. Replacing local constants with ones from FidoConstants.h;
2. Merging m_assertionResponses and m_existingCredentials by replacing HashSet with Vector in Authenticator::Observer::selectAssertionResponse;
3. Using Base64 encoded strings as the keys of the HashSet in produceHashSet() instead of the old casting hack;
4. Invaliding the LAContext in LocalConnection::~LocalConnection() such that any displaying LocalAuthentication dialogs can be dismissed after the object is destroyed;
5. Sorting existing credentials according to LRU before returning to UI clients;
6. Improving the LocalAuthentication dialog titles for iOS.
* UIProcess/WebAuthentication/Authenticator.h:
* UIProcess/WebAuthentication/AuthenticatorManager.cpp:
(WebKit::AuthenticatorManager::selectAssertionResponse):
* UIProcess/WebAuthentication/AuthenticatorManager.h:
* UIProcess/WebAuthentication/Cocoa/LocalAuthenticator.h:
* UIProcess/WebAuthentication/Cocoa/LocalAuthenticator.mm:
(WebKit::LocalAuthenticatorInternal::produceHashSet):
(WebKit::LocalAuthenticatorInternal::toNSData):
(WebKit::LocalAuthenticatorInternal::getExistingCredentials):
(WebKit::LocalAuthenticator::makeCredential):
(WebKit::LocalAuthenticator::continueMakeCredentialAfterUserVerification):
(WebKit::LocalAuthenticator::getAssertion):
(WebKit::LocalAuthenticator::continueGetAssertionAfterUserVerification):
* UIProcess/WebAuthentication/Cocoa/LocalConnection.h:
(WebKit::LocalConnection::filterResponses const):
* UIProcess/WebAuthentication/Cocoa/LocalConnection.mm:
(WebKit::WebCore::bundleName):
(WebKit::LocalConnection::~LocalConnection):
(WebKit::LocalConnection::verifyUser):
(WebKit::LocalConnection::verifyUser const): Deleted.
* UIProcess/WebAuthentication/Mock/MockLocalConnection.h:
* UIProcess/WebAuthentication/Mock/MockLocalConnection.mm:
(WebKit::MockLocalConnection::verifyUser):
(WebKit::MockLocalConnection::filterResponses const):
(WebKit::MockLocalConnection::verifyUser const): Deleted.
* UIProcess/WebAuthentication/fido/CtapAuthenticator.cpp:
(WebKit::CtapAuthenticator::continueGetAssertionAfterResponseReceived):
(WebKit::CtapAuthenticator::continueGetNextAssertionAfterResponseReceived):
* UIProcess/WebAuthentication/fido/CtapAuthenticator.h:
2020-04-30 Kate Cheney <katherine_cheney@apple.com>
Change default of "inAppBrowserPrivacyQuirks" testing toggle to use it for tests only
https://bugs.webkit.org/show_bug.cgi?id=211209
<rdar://problem/62619659>
Reviewed by Brent Fulgham.
No new tests, behavior confirmed by existing tests.
* Shared/WebPreferences.yaml:
2020-04-30 Devin Rousso <drousso@apple.com>
Unreviewed, add missing parenthesis
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::handleMediaEvent):
2020-04-30 Chris Dumez <cdumez@apple.com>
[iOS][WK2] Add timeout for "Client navigation" foreground assertion
https://bugs.webkit.org/show_bug.cgi?id=211202
Reviewed by Alex Christensen.
Add a 30s timeout for "Client navigation" foreground assertion. Foreground assertions do not timeout
at RunningBoard level and we thus need to make sure we release them to avoid power leaks.
* UIProcess/API/APINavigation.cpp:
(API::Navigation::Navigation):
* UIProcess/API/APINavigation.h:
(API::Navigation::setClientNavigationActivity):
(API::Navigation::setForegroundActivity): Deleted.
* UIProcess/ProcessThrottler.cpp:
(WebKit::ProcessThrottler::TimedActivity::TimedActivity):
(WebKit::ProcessThrottler::TimedActivity::operator=):
(WebKit::ProcessThrottler::TimedActivity::activityTimedOut):
(WebKit::ProcessThrottler::TimedActivity::updateTimer):
* UIProcess/ProcessThrottler.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::loadRequest):
(WebKit::WebPageProxy::loadFile):
(WebKit::WebPageProxy::loadData):
(WebKit::WebPageProxy::didFailProvisionalLoadForFrameShared):
(WebKit::WebPageProxy::didFinishLoadForFrame):
(WebKit::WebPageProxy::didFailLoadForFrame):
2020-04-30 Alex Christensen <achristensen@webkit.org>
Add SPI to change a WKWebView's CORS disabling pattern after initialization
https://bugs.webkit.org/show_bug.cgi?id=211211
<rdar://problem/61837474>
Reviewed by Chris Dumez.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _corsDisablingPatterns]):
(-[WKWebView _setCORSDisablingPatterns:]):
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/WebPageProxy.cpp:
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::corsDisablingPatterns const):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::parseAndAllowAccessToCORSDisablingPatterns):
(WebKit::m_isNavigatingToAppBoundDomain):
(WebKit::WebPage::updateCORSDisablingPatterns):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-04-30 Daniel Bates <dabates@apple.com>
[iOS] Implement -markedTextRange
https://bugs.webkit.org/show_bug.cgi?id=211148
<rdar://problem/57865890>
Reviewed by Wenson Hsieh.
Return the UITextRange-like object for the marked text range. Clients can use this range to
access the selection rects or query for the caret rect at the start or end of the range,
if desired.
* Shared/EditorState.cpp:
(WebKit::EditorState::PostLayoutData::encode const):
(WebKit::EditorState::PostLayoutData::decode):
Encode and decode the marked text selection rects and caret rects at the state and end
of the marked text range. This is the same information that we compute in order to
implement -selectedTextRange. I thought about collecting these details into class and
having editor state hold two instances of it: one for selected text and one for marked text.
However I decided against it because this class is unlikely to find use outside of what is
needed to implement -selectedTextRange and -markedTextRange. This is because only those
functions require such details so as to return an opaque UITextRange-derived object that
can be passed to- or its sub-objects can be passed to WKContentView's -caretRectForPosition
and -selectionRectsForRange. Literally, WKContentView does not support other UITextInput
functions that operate on UITextRange or UITextPosition. I didn't pursue fixing this in
general because 1) I don't need it to solve my problem and 2) UIWKDocumentContext SPI seems
to be the new general-purpose solution.
(WebKit::operator<<): Pretty print all the new marked text details we collect. While I am here,
use Vector::isEmpty() instead of Vector::size() to check if there are selection rects.
* Shared/EditorState.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView textFirstRect]):
(-[WKContentView textLastRect]):
Update code now that the marked text rects are in the post layout data sub-object.
(-[WKContentView _scaledCaretRectForSelectionStart:]): Extracted from -selectedTextRange.
(-[WKContentView _scaledCaretRectForSelectionEnd:]): Ditto.
(-[WKContentView selectedTextRange]): Write in terms of -_scaledCaretRectForSelectionStart
and -_scaledCaretRectForSelectionEnd. Also modernize the code while I am here.
(-[WKContentView markedTextRange]): Added.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::getPlatformEditorState const): Collect the marked text rects and the caret
rects at the start and end of the range.
2020-04-30 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4][X11] Add support for rendering web view contents
https://bugs.webkit.org/show_bug.cgi?id=211189
Reviewed by Adrian Perez de Castro.
Use GdkDisplayX11::xevent signal instead of filtering events that is no longer available in GTK4. GdkVisual no
longer exists either, so we get the X visual from the PlatformDisplay.
* UIProcess/gtk/AcceleratedBackingStoreX11.cpp:
(WebKit::XDamageNotifier::add):
(WebKit::XDamageNotifier::remove):
(WebKit::XDamageNotifier::filterXDamageEvent):
(WebKit::AcceleratedBackingStoreX11::~AcceleratedBackingStoreX11):
(WebKit::AcceleratedBackingStoreX11::update):
(WebKit::AcceleratedBackingStoreX11::snapshot):
* WebProcess/WebPage/gtk/AcceleratedSurfaceX11.cpp:
(WebKit::AcceleratedSurfaceX11::AcceleratedSurfaceX11):
(WebKit::AcceleratedSurfaceX11::createPixmap):
2020-04-30 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Add support click events
https://bugs.webkit.org/show_bug.cgi?id=211175
Reviewed by Adrian Perez de Castro.
Handle button pressed and released events using a GtkGestureClick.
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(ClickCounter::currentClickCountForGdkButtonEvent):
(webkitWebViewBaseButtonPressed):
(webkitWebViewBaseButtonReleased):
(webkitWebViewBaseConstructed):
(webkitWebViewBaseResetClickCounter):
2020-04-29 Simon Fraser <simon.fraser@apple.com>
Use initializers in PlatformMouseEvent and WebEvent
https://bugs.webkit.org/show_bug.cgi?id=211217
Reviewed by Tim Horton.
Use initializers im WebEvent.
* Shared/WebEvent.h:
* Shared/WebMouseEvent.cpp:
2020-04-29 Brent Fulgham <bfulgham@apple.com>
Improve SandboxExtension::HandleArray to reduce boilerplate
https://bugs.webkit.org/show_bug.cgi?id=211103
<rdar://problem/62533632>
Reviewed by Per Arne Vollan.
There are a number of boilerplate patterns needed when using SandboxExtension::HandleArray.
We could make these simpler and less error prone by improving the class.
This patch:
1. Adds convenience methods to create and consume HandleArrays.
2. Updates uses of HandleArray to use the convenience methods.
Tested by existing tests.
* Platform/IPC/FormDataReference.h:
(IPC::FormDataReference::decode):
* Shared/Cocoa/SandboxExtensionCocoa.mm:
(WebKit::SandboxExtension::createHandlesForFiles):
(WebKit::SandboxExtension::createHandlesForMachLookup):
(WebKit::SandboxExtension::consumePermanently):
* Shared/SandboxExtension.h:
(WebKit::createHandlesForFiles):
(WebKit::SandboxExtension::consumePermanently):
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
* UIProcess/Cocoa/WebProcessProxyCocoa.mm:
(WebKit::WebProcessProxy::unblockPreferenceServiceIfNeeded):
* UIProcess/WebPageProxy.cpp:
* WebProcess/WebCoreSupport/WebPlatformStrategies.cpp:
(WebKit::WebPlatformStrategies::getPathnamesForType):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::extendSandboxForFilesFromOpenPanel):
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::initializeWebProcess):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
(WebKit::WebProcess::unblockPreferenceService):
2020-04-29 Chris Dumez <cdumez@apple.com>
REGRESSION(r260791) Network process fails to suspend promptly
https://bugs.webkit.org/show_bug.cgi?id=211207
<rdar://problem/62620454>
Reviewed by Alex Christensen.
After r260791, all WebResourceLoadStatisticsStore instances share a single WorkQueue.
As a result, the logic to suspend WebResourceLoadStatisticsStore's WorkQueues in
NetworkProcess::prepareToSuspend() needs to get updated to reflect this.
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::suspend):
(WebKit::WebResourceLoadStatisticsStore::resume):
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.h:
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::prepareToSuspend):
(WebKit::NetworkProcess::resume):
2020-04-29 Adrian Perez de Castro <aperez@igalia.com>
REGRESSION(r260889): TestContextMenu:/webkit/WebKitWebView/populate-menu no longer passes
https://bugs.webkit.org/show_bug.cgi?id=211203
Reviewed by Aakash Jain.
No new tests needed.
* UIProcess/gtk/WebContextMenuProxyGtk.cpp:
(WebKit::WebContextMenuProxyGtk::WebContextMenuProxyGtk): Attach the correct action
group to the popover widget, instead of a temporary which will always be empty.
2020-04-29 David Kilzer <ddkilzer@apple.com>
REGRESSION (r246395): Leak of ARQuickLookPreviewItem and ARQuickLookWebKitItem in -[_WKPreviewControllerDataSource previewController:previewItemAtIndex:]
<https://webkit.org/b/211197>
<rdar://problem/62612483>
Reviewed by Darin Adler.
* UIProcess/Cocoa/SystemPreviewControllerCocoa.mm:
(-[_WKPreviewControllerDataSource previewController:previewItemAtIndex:]):
- Use RetainPtr<> and adoptNS() to fix leaks.
2020-04-29 Per Arne Vollan <pvollan@apple.com>
Unreviewed sandbox compile fix.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-04-29 Alex Christensen <achristensen@webkit.org>
Add WKNavigationDelegate API shouldAllowDeprecatedTLS
https://bugs.webkit.org/show_bug.cgi?id=210981
<rdar://problem/61742976>
Reviewed by Geoffrey Garen.
* UIProcess/API/Cocoa/WKNavigationDelegate.h:
* UIProcess/Cocoa/NavigationState.h:
* UIProcess/Cocoa/NavigationState.mm:
(WebKit::NavigationState::setNavigationDelegate):
(WebKit::NavigationState::NavigationClient::shouldAllowLegacyTLS):
2020-04-29 Chris Dumez <cdumez@apple.com>
[iOS][WK2] Temporarily stop using RunningBoard for the foreground process assertion
https://bugs.webkit.org/show_bug.cgi?id=211196
<rdar://problem/62535822>
Reviewed by Tim Horton.
Temporarily stop using RunningBoard for the foreground process assertion. On recent iOS builds, there
are issues where this assertion can time out after 30 seconds, even though the process is still
foreground.
* Configurations/WebKit.xcconfig:
* UIProcess/ProcessAssertion.h:
* UIProcess/ios/ProcessAssertionIOS.mm:
(WebKit::runningBoardNameForAssertionType):
(WebKit::ProcessAssertion::ProcessAssertion):
(WebKit::ProcessAssertion::~ProcessAssertion):
2020-04-29 Youenn Fablet <youenn@apple.com>
Set _STAttributionDisplayName to iOS WebContent Info plist
https://bugs.webkit.org/show_bug.cgi?id=210772
<rdar://problem/62075201>
Reviewed by Eric Carlson.
Set _STAttributionDisplayName to 'Website' value.
* WebProcess/EntryPoint/Cocoa/XPCService/WebContentService/Info-iOS.plist:
2020-04-29 Kate Cheney <katherine_cheney@apple.com>
Refactor layout tests after updates to In-App Browser Privacy
https://bugs.webkit.org/show_bug.cgi?id=211146
<rdar://problem/62524127>
Reviewed by Brent Fulgham.
Tests: http/tests/in-app-browser-privacy/app-bound-domain-gets-app-bound-session.html
http/tests/in-app-browser-privacy/non-app-bound-domain-does-not-get-app-bound-session.html
Removes setInAppBrowserPrivacyEnabled, which is no longer needed by
layout tests. Setting this in TestOptions is enough.
getWebViewCategory is not needed because this patch deletes the only test
which uses it. Lastly, adds a new function to clear app-bound sessions
now that multiple tests are checking for them.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::clearAppBoundSession):
(WebKit::NetworkProcess::setInAppBrowserPrivacyEnabled): Deleted.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/NetworkSession.h:
(WebKit::NetworkSession::clearAppBoundSession):
(WebKit::NetworkSession::setInAppBrowserPrivacyEnabled): Deleted.
* NetworkProcess/cocoa/NetworkSessionCocoa.h:
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::appBoundSession):
(WebKit::NetworkSessionCocoa::clearAppBoundSession):
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreClearAppBoundSession):
(WKWebsiteDataStoreSetInAppBrowserPrivacyEnabled): Deleted.
* UIProcess/API/C/WKWebsiteDataStoreRef.h:
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
(-[WKWebsiteDataStore _getWebViewCategoryFor:completionHandler:]): Deleted.
* UIProcess/API/Cocoa/WKWebsiteDataStorePrivate.h:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::clearAppBoundSession):
(WebKit::NetworkProcessProxy::setInAppBrowserPrivacyEnabled): Deleted.
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::clearAppBoundSession):
(WebKit::WebsiteDataStore::setInAppBrowserPrivacyEnabled): Deleted.
* UIProcess/WebsiteData/WebsiteDataStore.h:
2020-04-29 Víctor Manuel Jáquez Leal <vjaquez@igalia.com>
[GPUP][GTK] compile GPUProcess in GTK port
https://bugs.webkit.org/show_bug.cgi?id=208814
Reviewed by Don Olmstead.
Enable GPUProcess compilation in WebKitGTK as an experimental
feature
No new tests required.
* GPUProcess/GPUConnectionToWebProcess.cpp: guard
UserMediaCaptureManagerProxy for Cocoa platform only.
(WebKit::GPUConnectionToWebProcess::dispatchMessage): guard
messages from UserMediaCaptureManagerProxy for Cocoa platform,
moving up this guard from video track messages.
(WebKit::GPUConnectionToWebProcess::mediaKeyStorageDirectory):
removed guard for this function of ENCRYPTED_MEDIA leaving only
LEGACY_ENCRYPTED_MEDIA, since it's were it is defined.
(WebKit::GPUConnectionToWebProcess::setOrientationForMediaCapture):
guard the content of this function for Cocoa platform, since it is
calling a UserMediaCaptureManagerProxy object.
* GPUProcess/GPUProcess.cpp: Replace the included header to match
with the used symbols in file.
* GPUProcess/gstreamer/GPUProcessGStreamer.cpp: Added.
(WebKit::GPUProcess::initializeProcess): Empty function.
(WebKit::GPUProcess::initializeProcessName): Empty function.
(WebKit::GPUProcess::initializeSandbox): Empty function.
* GPUProcess/gstreamer/GPUProcessMainGStreamer.cpp: Added.
(WebKit::initializeAuxiliaryProcess<GPUProcess>): New function.
(WebKit::GPUProcessMain): New function.
* GPUProcess/media/RemoteLegacyCDMFactoryProxy.cpp: changed guard
for this file from ENCRYPTED_MEDIA to LEGACY_ENCRYPTED_MEDIA to
match with its header file.
* GPUProcess/media/gstreamer/RemoteMediaPlayerProxyGStreamer.cpp: Added.
(WebKit::RemoteMediaPlayerProxy::prepareForPlayback):
New function that raises noImplemented()
(WebKit::RemoteMediaPlayerProxy::mediaPlayerFirstVideoFrameAvailable):
New function that raises noImplemented()
(WebKit::RemoteMediaPlayerProxy::mediaPlayerRenderingModeChanged):
New function that raises noImplemented()
(WebKit::RemoteMediaPlayerProxy::enterFullscreen): New function
that raises noImplemented()
(WebKit::RemoteMediaPlayerProxy::exitFullscreen): New function
that raises noImplemented()
* PlatformGTK.cmake: add Platform/generic header directory.
* SourcesGTK.txt: Add the new files and
GPUProcess/media/RemoteAudioDestinationManager.cpp
* WebProcess/GPU/GPUProcessConnection.cpp: guard
UserMediaCaptureManager for Cocoa platform only.
(WebKit::GPUProcessConnection::dispatchMessage): guard messages
from UserMediaCaptureManager for Cocoa platform, moving up this
guard applied also for video track.
* WebProcess/GPU/media/RemoteCDM.cpp: include missing header
GPUProcessConnection.h
* WebProcess/GPU/media/gstreamer/VideoLayerRemoteGStreamer.cpp: Added.
(WebKit::createVideoLayerRemote): New function
that raises noImplemented()
2020-04-29 Claudio Saavedra <csaavedra@igalia.com>
[GTK4] Unreviewed compile-warning fixes
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseSetEnableBackForwardNavigationGesture):
(webkitWebViewBaseWillSwapWebProcess):
2020-04-29 Rob Buis <rbuis@igalia.com>
Make PolicyChecker an inner class of FrameLoader
https://bugs.webkit.org/show_bug.cgi?id=211138
Reviewed by Alex Christensen.
Adapt to PolicyChecker::ShouldContinue change.
* GPUProcess/media/RemoteMediaResource.cpp:
(WebKit::RemoteMediaResource::responseReceived):
* GPUProcess/media/RemoteMediaResource.h:
* GPUProcess/media/RemoteMediaResourceManager.cpp:
(WebKit::RemoteMediaResourceManager::responseReceived):
* GPUProcess/media/RemoteMediaResourceManager.h:
* GPUProcess/media/RemoteMediaResourceManager.messages.in:
* Scripts/webkit/messages.py:
* WebProcess/GPU/media/RemoteMediaResourceProxy.cpp:
(WebKit::RemoteMediaResourceProxy::responseReceived):
* WebProcess/GPU/media/RemoteMediaResourceProxy.h:
2020-04-29 Adrian Perez de Castro <aperez@igalia.com>
[GTK] Misplaced right click menu on web page due to deprecated gtk_menu_popup()
https://bugs.webkit.org/show_bug.cgi?id=170553
Reviewed by Carlos Garcia Campos.
Replace GtkMenuShell with a GtkPopoverMenu for context menus. The former is not available
at all in GTK4, and the later allows for simplifying the positioning code: it is enough
to provide a point in WebKitWebView widget where to place the popup, and GTK takes care
of everything. This removes the custom positioning code (as it is not needed anymore),
which did GdkScreen-relative calculations that GTK4 does not support.
No new tests needed.
* Shared/glib/WebContextMenuItemGlib.h:
* UIProcess/API/glib/WebKitWebView.cpp:
(contextMenuDismissed): Change parameter from GtkMenuShell to GtkMenuShell to GtkWidget.
(webkitWebViewPopulateContextMenu): Connect to the GtkPopover::closed signal instead of
GtkMenuShell::deactivate.
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(activeContextMenuClosed): Renamed from activeContextMenuUnmapped(), changed parameter
from GtkMenuShell to GtkWidget, and compare with WebContextMenuProxyGtk::gtkWidget().
(activeContextMenuUnmapped): Renamed to activeContextMenuClosed().
(webkitWebViewBaseSetActiveContextMenuProxy): Connect to the GtkPopover::closed signal
instead of GtkMenuShell::deactivate.
(webkitWebViewBaseGetActiveContextMenuProxy):
* UIProcess/API/gtk/WebKitWebViewBasePrivate.h:
* UIProcess/API/gtk/WebKitWebViewGtk.cpp:
* UIProcess/gtk/WebContextMenuProxyGtk.cpp: Arrange to use GtkPopoverMenu instead of
GtkMenuShell.
* UIProcess/gtk/WebContextMenuProxyGtk.cpp:
(WebKit::WebContextMenuProxyGtk::populate):
(WebKit::WebContextMenuProxyGtk::showContextMenuWithItems): Simplify using
m_context.menuLocation() to obtain the location where to make the context menu popup
next to, which allows removing the ::menuPositionFunction() callback as well.
(WebKit::WebContextMenuProxyGtk::WebContextMenuProxyGtk):
(WebKit::WebContextMenuProxyGtk::~WebContextMenuProxyGtk):
* UIProcess/gtk/WebContextMenuProxyGtk.h: Remove declarations for ::menuPositionFunction()
and ::m_popupPosition, which are now unneeded.
(WebKit::WebContextMenuProxyGtk::gtkWidget const): Renamed from ::gtkMenu(), and made it
return a GtkWidget.
2020-04-29 Commit Queue <commit-queue@webkit.org>
Unreviewed, reverting r260650.
https://bugs.webkit.org/show_bug.cgi?id=211172
It is breaking internal bots (Requested by youenn on #webkit).
Reverted changeset:
"Call STDynamicActivityAttributionPublisher in the WebProcess"
https://bugs.webkit.org/show_bug.cgi?id=210772
https://trac.webkit.org/changeset/260650
2020-04-29 Devin Rousso <drousso@apple.com>
Web Inspector: Uncaught Exception: SyntaxError: Unexpected identifier 'E'. Expected either a closing ']' or a ',' following an array element.
https://bugs.webkit.org/show_bug.cgi?id=211163
Reviewed by Joseph Pecoraro.
* WebProcess/Inspector/WebInspectorUI.cpp:
(WebKit::WebInspectorUI::updateFindString):
* WebProcess/Inspector/RemoteWebInspectorUI.cpp:
(WebKit::RemoteWebInspectorUI::updateFindString):
Use the newly exposed `JSON::Value::escapeString` to ensure that the `findString` will not
throw a JavaScript exception when placed in the `InspectorFrontendAPI` call.
2020-04-28 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Add support for key events
https://bugs.webkit.org/show_bug.cgi?id=211128
Reviewed by Adrian Perez de Castro.
Handle key-pressed and key-released events using a GtkEventControllerKey. To receive key events the widget needs
to be focused, so also implement GtkWidgetClass::grab_focus and handle enter and leave focus events. GTK4 needs
the widget used by the key bindings translator to be added to a parent to be able to forward key events, so it's
now added as a child internal widget of the WebKitWeViewBase.
* Shared/gtk/WebEventFactory.cpp:
(WebKit::WebEventFactory::createWebKeyboardEvent):
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewChildIsInternalWidget):
(webkitWebViewBaseContainerRemove):
(webkitWebViewBaseContainerForall):
(webkitWebViewBaseFocusEnter):
(webkitWebViewBaseFocusLeave):
(webkitWebViewBaseKeyPressed):
(webkitWebViewBaseKeyReleased):
(webkitWebViewBaseGrabFocus):
(webkitWebViewBaseConstructed):
(webkit_web_view_base_class_init):
* UIProcess/gtk/KeyBindingTranslator.cpp:
(WebKit::KeyBindingTranslator::KeyBindingTranslator):
(WebKit::KeyBindingTranslator::~KeyBindingTranslator):
(WebKit::handleCustomKeyBindings):
(WebKit::KeyBindingTranslator::commandsForKeyEvent):
* UIProcess/gtk/KeyBindingTranslator.h:
(WebKit::KeyBindingTranslator::widget const):
(WebKit::KeyBindingTranslator::destroyed):
2020-04-28 Wenson Hsieh <wenson_hsieh@apple.com>
[Text manipulation] Add a userInfo dictionary to _WKTextManipulationToken
https://bugs.webkit.org/show_bug.cgi?id=211151
<rdar://problem/62329534>
Reviewed by Darin Adler.
* UIProcess/API/Cocoa/WKWebView.mm:
(createUserInfo):
(-[WKWebView _startTextManipulationsWithConfiguration:completion:]):
(-[WKWebView _completeTextManipulation:completion:]):
(-[WKWebView _completeTextManipulationForItems:completion:]):
* UIProcess/API/Cocoa/_WKTextManipulationToken.h:
* UIProcess/API/Cocoa/_WKTextManipulationToken.mm:
(-[_WKTextManipulationToken setUserInfo:]):
(-[_WKTextManipulationToken userInfo]):
Add a new `userInfo` dictionary to `_WKTextManipulationToken`, which contains several named
`NSString` keys.
(-[_WKTextManipulationToken isEqualToTextManipulationToken:includingContentEquality:]):
(-[_WKTextManipulationToken _descriptionPreservingPrivacy:]):
2020-04-28 David Kilzer <ddkilzer@apple.com>
REGRESSION (r260407): Over-release of NSGraphicsContext in WebKit::convertPlatformImageToBitmap()
<https://webkit.org/b/211160>
Reviewed by Darin Adler.
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::convertPlatformImageToBitmap): Use retainPtr() instead
of adoptNS() to fix the over-release.
2020-04-28 David Kilzer <ddkilzer@apple.com>
IPC::Decoder::isValid() should just be a nullptr check
<https://webkit.org/b/211152>
<rdar://problem/62552699>
Reviewed by Darin Adler.
* Platform/IPC/Decoder.h:
(IPC::Decoder::isValid const): Change to nullptr check. I
reviewed the code in Decoder.{cpp,h}, and I didn't see
m_bufferPos being set without a bounds check. Also added
WARN_UNUSED_RETURN to make sure the result is always checked.
2020-04-28 Daniel Bates <dabates@apple.com>
Move WebPage::textInputContextsInRect() to WebPageIOS.mm
https://bugs.webkit.org/show_bug.cgi?id=211136
Reviewed by Eric Carlson.
The function WebPage::textInputContextsInRect is very specific to a client on iOS.
Although it's tempting to keep this cross-platform because its implementation does
not make use of iOS-specific technologies it does a very specific operation, the
result of which is a list of contexts to editable elements on the page. The contexts
provide enough info for the specific iOS client, but not enough to ever be useful
to any other client. Therefore move it to WebPageIOS.mm.
Only Mac and iOS were ever using this function. The former only for testing purposes
that have since been removed.
A side effect of this change is that I also move WebPageProxy::textInputContextsInRect()
from WebPageProxy.cpp to WebPageProxyIOS.mm.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::textInputContextsInRect): Deleted; moved to WebPageProxyIOS.mm.
* UIProcess/WebPageProxy.h:
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::textInputContextsInRect): Moved from WebPageProxy.cpp.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::textInputContextsInRect): Deleted; moved to WebPageIOS.mm.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::textInputContextsInRect): Moved from WebPage.cpp.
2020-04-28 Chris Dumez <cdumez@apple.com>
[iOS][WK2] Drop process assertion logic for holding locked files
https://bugs.webkit.org/show_bug.cgi?id=206910
Reviewed by Alex Christensen.
Drop process assertion logic for holding locked files as it is causing issues and
should no longer be needed.
The intention of this code was that the Network Process would send a
SetIsHoldingLockedFiles(bool) IPC to the UIProcess whenever there are pending
SQLite transactions and the UIProcess would then take a process assertion on behalf
of the WebProcess (or release it) to prevent / allow suspension.
This approach has some issues:
1. It was noisy in terms of IPC
2. Because it relies on an IPC to the UIProcess to keep running, it was inherently
racy
3. We already have logic to deal with suspension in the network / web processes.
The UIProcess sends a PrepareForSuspension IPC to its child processes. In turn,
those processes will make sure to suspend their database threads to prevent any
further database work. The child processes would then respond to the
PrepareForSuspension IPC to let the UIProcess know they are ready to suspend.
The 2 approaches were conflicting, because after the PrepareForSuspension IPC is
sent, the child process could process a transaction and send the
SetIsHoldingLockedFiles() IPC to the UIProcess which would incorrectly cancel the
suspension.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::NetworkProcess):
(WebKit::m_messagePortChannelRegistry): Deleted.
* NetworkProcess/NetworkProcess.h:
* Shared/WebSQLiteDatabaseTracker.cpp: Removed.
* Shared/WebSQLiteDatabaseTracker.h: Removed.
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::didClose):
(WebKit::NetworkProcessProxy::didSetAssertionState):
(WebKit::NetworkProcessProxy::setIsHoldingLockedFiles): Deleted.
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/Network/NetworkProcessProxy.messages.in:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::shutDown):
(WebKit::WebProcessProxy::setIsHoldingLockedFiles): Deleted.
* UIProcess/WebProcessProxy.h:
* UIProcess/WebProcessProxy.messages.in:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebProcess.cpp:
(WebKit::m_nonVisibleProcessCleanupTimer):
(WebKit::m_webSQLiteDatabaseTracker): Deleted.
* WebProcess/WebProcess.h:
2020-04-28 Noam Rosenthal <noam@webkit.org>
Implement FCP (first contentful paint)
https://bugs.webkit.org/show_bug.cgi?id=208499
Reviewed by Simon Fraser.
Add an experimental runtime flag for paint-timing (paintTimingEnabled).
* Shared/WebPreferences.yaml:
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetPaintTimingEnabled):
(WKPreferencesGetPaintTimingEnabled):
* UIProcess/API/C/WKPreferencesRefPrivate.h:
* WebProcess/Storage/WebSWContextManagerConnection.cpp:
(WebKit::WebSWContextManagerConnection::updatePreferencesStore):
2020-04-28 Ross Kirsling <ross.kirsling@sony.com>
[JSC] Align upon the name isCallable instead of isFunction
https://bugs.webkit.org/show_bug.cgi?id=211140
Reviewed by Darin Adler.
* WebProcess/Plugins/Netscape/NPJSObject.cpp:
(WebKit::NPJSObject::hasMethod):
2020-04-28 Devin Rousso <drousso@apple.com>
Web Inspector: find dialog does not populate search string from system find pasteboard
https://bugs.webkit.org/show_bug.cgi?id=113588
<rdar://problem/19281466>
Reviewed by Brian Burg.
* UIProcess/Inspector/mac/WKInspectorWKWebView.h:
* UIProcess/Inspector/mac/WKInspectorWKWebView.mm:
(-[WKInspectorWKWebView dealloc]): Added.
(-[WKInspectorWKWebView setInspectorWKWebViewDelegate:]):
(-[WKInspectorWKWebView becomeFirstResponder]): Added.
(-[WKInspectorWKWebView _handleWindowDidBecomeKey:]): Added.
* UIProcess/Inspector/mac/WKInspectorViewController.h:
* UIProcess/Inspector/mac/WKInspectorViewController.mm:
(-[WKInspectorViewController inspectorWKWebViewDidBecomeActive:]): Added.
* UIProcess/Inspector/WebInspectorProxy.h:
* UIProcess/Inspector/mac/WebInspectorProxyMac.mm:
(-[WKWebInspectorProxyObjCAdapter inspectorViewControllerDidBecomeActive:]): Added.
(WebKit::WebInspectorProxy::didBecomeActive): Added.
* UIProcess/Inspector/RemoteWebInspectorProxy.h:
* UIProcess/Inspector/mac/RemoteWebInspectorProxyMac.mm:
(-[WKRemoteWebInspectorProxyObjCAdapter inspectorWKWebViewDidBecomeActive:]): Added.
(WebKit::RemoteWebInspectorProxy::didBecomeActive): Added.
Add a new delegate call chain that notifies the inspector proxy about when the inspector web
view becomes active, at which point the inspector proxy can get the current find string and
sent it to the inspector page.
* WebProcess/Inspector/WebInspectorUI.messages.in:
* WebProcess/Inspector/WebInspectorUI.h:
* WebProcess/Inspector/WebInspectorUI.cpp:
(WebKit::WebInspectorUI::updateFindString): Added.
* WebProcess/Inspector/RemoteWebInspectorUI.messages.in:
* WebProcess/Inspector/RemoteWebInspectorUI.h:
* WebProcess/Inspector/RemoteWebInspectorUI.cpp:
(WebKit::RemoteWebInspectorUI::updateFindString): Added.
Pass along the new find string to the frontend via `InspectorFrontendAPI.updateFindString`.
2020-04-28 Daniel Bates <dabates@apple.com>
Remove unused WebPage::focusTextInputContext()
https://bugs.webkit.org/show_bug.cgi?id=211135
Reviewed by Anders Carlsson.
I accidentally forgot to remove the WebPage message and WebPage implementation that
backed -_focusTextInputContext when I removed it in r260225. Now the implementation
is unused. Let's remove it.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::focusTextInputContext): Deleted.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-04-28 Christopher Reid <chris.reid@sony.com>
[Win] Bundle Inspector Resources in Release builds
https://bugs.webkit.org/show_bug.cgi?id=210942
Reviewed by Fujii Hironori.
Move CMake logic for the inspector resource copy script to
Source/WebInspectorUI so it can be shared with Win and other platforms.
* InspectorGResources.cmake:
* PlatformWPE.cmake:
* PlatformWin.cmake:
2020-04-28 David Kilzer <ddkilzer@apple.com>
[MacCatalyst] Add missing symlinks to WebKit.frameworks
<https://webkit.org/b/211102>
<rdar://problem/62137758>
Reviewed by Brent Fulgham.
* WebKit.xcodeproj/project.pbxproj:
(Make Frameworks Symbolic Link):
- Update logic to run for macOS and macCatalyst builds.
- Add output path.
(Check For Inappropriate Objective-C Class Names):
- Let Xcode have its way with the project file by adding a
newline to the end of the script.
(Check For Weak VTables and Externals): Ditto.
(Add XPCServices symlink):
- Update logic to run for macOS and macCatalyst builds.
2020-04-28 Kate Cheney <katherine_cheney@apple.com>
Create a mechanism to add missing ITP Database tables when the schema is updated
https://bugs.webkit.org/show_bug.cgi?id=211004
<rdar://problem/62261187>
Reviewed by John Wilander.
This patch updates the ITP database to better handle added tables to
the schema. Currently the database store deletes and re-creates the
database when any schema change is encountered. This is simple but
would result in ITP data loss during schema updates, so this patch
searches the schema for missing tables and adds any it finds.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::createTableQueries):
Convert the tables array to a mapping of table names to their
respective CREATE TABLE queries so we can add missing tables on the
fly.
(WebKit::ResourceLoadStatisticsDatabaseStore::ResourceLoadStatisticsDatabaseStore):
(WebKit::ResourceLoadStatisticsDatabaseStore::openITPDatabase):
Remove call to createSchema() from the constructor. It makes more
sense to base this on whether the file is new as opposed to whether
one table exists.
(WebKit::ResourceLoadStatisticsDatabaseStore::checkForMissingTablesInSchema):
(WebKit::ResourceLoadStatisticsDatabaseStore::migrateDataToNewTablesIfNecessary):
(WebKit::ResourceLoadStatisticsDatabaseStore::addMissingTablesIfNecessary):
Checks for missing tables and adds them using a best-effort approach.
Call createUniqueIndices() to call all CREATE UNIQUE INDEX queries,
even though we may not need them all. It is much simpler than to
associate each query with its table, and it doesn't hurt to call
again.
(WebKit::ResourceLoadStatisticsDatabaseStore::openAndUpdateSchemaIfNecessary):
Add missing tables instead of deleting the old database file. Changed
the function name to reflect this. Deleted a FIXME which this patch
addresses but added a new FIXME for migrating data when adding new
columns, which this patch does not address.
(WebKit::ResourceLoadStatisticsDatabaseStore::isCorrectTableSchema): Deleted.
(WebKit::ResourceLoadStatisticsDatabaseStore::openAndDropOldDatabaseIfNecessary): Deleted.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.h:
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::statisticsDatabaseHasAllTables):
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.h:
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::statisticsDatabaseHasAllTables):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
(-[WKWebsiteDataStore _statisticsDatabaseHasAllTables:]):
* UIProcess/API/Cocoa/WKWebsiteDataStorePrivate.h:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::statisticsDatabaseHasAllTables):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::statisticsDatabaseHasAllTables):
* UIProcess/WebsiteData/WebsiteDataStore.h:
Added an SPI call to test that the database schema includes all
tables. This tests the database is not dropped when a new
table is added upon opening the database.
2020-04-28 Per Arne Vollan <pvollan@apple.com>
[Cocoa] Global preferences are not accessible in the WebContent process when CFPrefs direct mode is enabled
https://bugs.webkit.org/show_bug.cgi?id=211075
<rdar://problem/61866711>
Reviewed by Brent Fulgham.
Global preferences in the domain 'kCFPreferencesAnyApplication' are not readable in the WebContent process when CFPrefs direct mode
is enabled. Fix this by transferring a select set of global preferences to the WebContent process on startup.
API test: WebKit.GlobalPreferences
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::setGlobalPreferences):
(WebKit::WebProcess::platformInitializeWebProcess):
2020-04-28 Peng Liu <peng.liu6@apple.com>
Remove the WebKit.plist for Feature Flags
https://bugs.webkit.org/show_bug.cgi?id=210534
Unreviewed project file clean-up.
A follow-up patch for r260599. Remove the reference to the WebKit.plist from project.pbxproj.
* WebKit.xcodeproj/project.pbxproj:
2020-04-28 Alex Christensen <achristensen@webkit.org>
Fix iOS API tests after r260790
https://bugs.webkit.org/show_bug.cgi?id=211093
* UIProcess/AuxiliaryProcessProxy.cpp:
(WebKit::AuxiliaryProcessProxy::sendMessage):
Only call the completion handler once.
2020-04-28 Jer Noble <jer.noble@apple.com>
[Mac] Adopt kMTSupportNotification_ShouldPlayHDRVideoChanged notification
https://bugs.webkit.org/show_bug.cgi?id=211028
Reviewed by Eric Carlson.
Adopt a new notification which signals that the receiver should re-query for HDR support.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::webProcessPoolHighDynamicRangeDidChangeCallback):
(WebKit::WebProcessPool::registerHighDynamicRangeChangeCallback):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::registerHighDynamicRangeChangeCallback):
(WebKit::WebProcessPool::initializeNewWebProcess):
* UIProcess/WebProcessPool.h:
2020-04-28 Adrian Perez de Castro <aperez@igalia.com>
[GTK] Unreviewed build fix for Clang 10
* UIProcess/API/gtk/WebKitInputMethodContextImplGtk.cpp:
(webkitInputMethodContextImplGtkGetPreedit): Use the "f" suffix to make constants float to
prevent a narrowing double-to-float narrowing issue which Clang 10 considers an error. This
only happens with GTK4 because GdkRGBA now has float members. Using float constants is fine
with GTK3, because promoting floats to doubles never loses precision.
2020-04-28 Daniel Bates <dabates@apple.com>
[iOS] Marked text editor state only needs to be set when layout is up-to-date
https://bugs.webkit.org/show_bug.cgi?id=211087
Reviewed by Darin Adler.
Organize the code to reflect reality: editor state only has marked text rects if layout is
up-to-date. So, move these fields to from EditorState to EditorState::PostLayoutData.
While I am in this area of the code I also moved the editor state originIdentifierForPasteboard
to the top of the struct as it is aesthetically more pleasing to me to have the post layout
data last. It likely also produces better bit packing though this wasn't my primary motivation.
* Shared/EditorState.cpp:
(WebKit::EditorState::encode const):
(WebKit::EditorState::decode):
(WebKit::EditorState::PostLayoutData::encode const):
(WebKit::EditorState::PostLayoutData::decode):
Move encoding and decoding of marked test fields from EditorState to EditorState::PostLayoutData.
(WebKit::operator<<): Fix up code now that marked text fields are in PostLayoutData.
* Shared/EditorState.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView textFirstRect]): Update code now that this data is in the post layout data struct.
(-[WKContentView textLastRect]): Ditto.
(-[WKContentView _updateChangedSelection:]): Ditto.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::convertContentToRootViewSelectionRects): Added.
(WebKit::WebPage::getPlatformEditorState const): Update now that marked text fields are in the
post layout data struct. Simplify the early return as well: only continue if we're not missing post
layout data. If we're missing post layout data then we either do not have a frame view or layout is
not up-to-data. Either way we don't want to do anything. I also stored the FrameView in a Ref<> to
future proof the code should future code cause view detachment and destruction. Note that layout
could cause this, but the code currently covers this case. So, taking a Ref<> is about future proofing.
(WebKit::WebPage::getRectsForGranularityWithSelectionOffset): Update to use convertContentToRootViewSelectionRects().
(WebKit::WebPage::getRectsAtSelectionOffsetWithText): Ditto.
(WebKit::WebPage::requestAutocorrectionData): Use Vector::map() to convert the selection rects to
from content to root view FloatRects.
(WebKit::WebPage::convertSelectionRectsToRootView): Deleted.
2020-04-28 Per Arne Vollan <pvollan@apple.com>
[iOS] Fix sandbox violation when uploading a file
https://bugs.webkit.org/show_bug.cgi?id=210937
Reviewed by Darin Adler.
On iOS, the file chooser needs access to frontboard and icon services in the WebContent process. Create and
consume extensions for these services when choosing files. When done, the extensions should be revoked. This
patch also fixes an out-of-bounds array exception when running the test created for this patch. Additionally,
the function thumbnailSizedImageForImage should return a RetainPtr<UIImage>, since it seems unsafe to not
retain the UIImage after the image context is released in that function.
Test: fast/forms/file/open-file-panel-crash.html
* Shared/ios/WebIconUtilities.h:
* Shared/ios/WebIconUtilities.mm:
(WebKit::thumbnailSizedImageForImage):
(WebKit::fallbackIconForFile):
(WebKit::iconForImageFile):
(WebKit::iconForVideoFile):
(WebKit::iconForFile):
* UIProcess/WebPageProxy.cpp:
* UIProcess/ios/forms/WKFileUploadPanel.mm:
(-[_WKFileUploadItem displayImage]):
(-[_WKImageFileUploadItem displayImage]):
(-[_WKVideoFileUploadItem displayImage]):
(-[WKFileUploadPanel documentPicker:didPickDocumentsAtURLs:]):
(-[WKFileUploadPanel imagePickerController:didFinishPickingMediaWithInfo:]):
(-[WKFileUploadPanel imagePickerController:didFinishPickingMultipleMediaWithInfo:]):
* WebProcess/WebCoreSupport/ios/WebChromeClientIOS.mm:
(WebKit::WebChromeClient::createIconForFiles):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::didChooseFilesForOpenPanelWithDisplayStringAndIcon):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-04-27 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Add support for motion events
https://bugs.webkit.org/show_bug.cgi?id=211073
Reviewed by Adrian Perez de Castro.
Handle enter, leave and motion events using a GtkEventControllerMotion.
* Shared/NativeWebMouseEvent.h:
* Shared/gtk/NativeWebMouseEventGtk.cpp:
(WebKit::NativeWebMouseEvent::NativeWebMouseEvent):
* Shared/gtk/WebEventFactory.cpp:
(WebKit::WebEventFactory::createWebMouseEvent):
* Shared/gtk/WebEventFactory.h:
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseEnter):
(webkitWebViewBaseMotion):
(webkitWebViewBaseLeave):
(webkitWebViewBaseConstructed):
2020-04-28 Adrian Perez de Castro <aperez@igalia.com>
Non-unified build fixes late April 2020 edition
https://bugs.webkit.org/show_bug.cgi?id=211099
Unreviewed build fix.
* NetworkProcess/NetworkLoad.cpp: Add mising WebCoreArgumentCoders.h header.
* WebProcess/Automation/WebAutomationDOMWindowObserver.cpp:
(WebKit::WebAutomationDOMWindowObserver::WebAutomationDOMWindowObserver): Add namespace to
WebCore::DOMWindow usage.
2020-04-26 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Add support for scroll events
https://bugs.webkit.org/show_bug.cgi?id=211045
Reviewed by Adrian Perez de Castro.
Use an event controller to connect to scroll signal and get the GdkEvent.
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseHandleWheelEvent):
(webkitWebViewBaseScroll):
(webkitWebViewBaseConstructed):
2020-04-28 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4][Wayland] Add support for rendering web view contents
https://bugs.webkit.org/show_bug.cgi?id=211021
Reviewed by Adrian Perez de Castro.
Implement GtkWidgetClass::snapshot and GtkWidgetClass::measure instead of GtkWidgetClass::draw and
GtkWidgetClass::get_preferred_width|height. Add AcceleratedBackingStore::snapshot() pure virtual to be used with
GTK4 instead of AcceleratedBackingStore::paint(), and implement it for Wayland.
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseSnapshot): Call AcceleratedBackingStore::snapshot().
(webkitWebViewBaseMeasure): Return the natural width/height for the WebView.
(webkit_web_view_base_class_init): Add implementations for snapshot and measure vfuncs.
* UIProcess/gtk/AcceleratedBackingStore.h:
* UIProcess/gtk/AcceleratedBackingStoreWayland.cpp:
(WebKit::AcceleratedBackingStoreWayland::tryEnsureGLContext): Always try to realize the context here, since that
can fail too.
(WebKit::AcceleratedBackingStoreWayland::tryEnsureTexture): Helper to share the code to prepare the texture.
(WebKit::AcceleratedBackingStoreWayland::downloadTexture): Helper to share the code to download the texture.
(WebKit::AcceleratedBackingStoreWayland::snapshot): Use gtk_snapshot_append_texture().
(WebKit::AcceleratedBackingStoreWayland::paint): Use new helpers.
* UIProcess/gtk/AcceleratedBackingStoreWayland.h:
* UIProcess/gtk/AcceleratedBackingStoreX11.cpp:
(WebKit::AcceleratedBackingStoreX11::snapshot):
* UIProcess/gtk/AcceleratedBackingStoreX11.h:
* UIProcess/gtk/HardwareAccelerationManager.cpp:
(WebKit::HardwareAccelerationManager::HardwareAccelerationManager): Force accelerated compositing mode for GTK4.
2020-04-27 Devin Rousso <drousso@apple.com>
Web Inspector: Storage: can see third-party cookies
https://bugs.webkit.org/show_bug.cgi?id=211092
<rdar://problem/62469078>
Reviewed by Chris Dumez.
Test: http/tests/inspector/page/get-cookies.html
After r259649, Web Inspector no longer incorrectly bails when attempting to get cookies if
the last resource loaded by a given frame does not have cookie access. We also need to check
whether the resource we're attempting to get cookies for is first-party or third-party so as
to reflect the current cookie policy.
* WebProcess/WebPage/WebCookieJar.cpp:
(WebKit::shouldBlockCookies):
2020-04-27 Brent Fulgham <bfulgham@apple.com>
Follow-up change after r260798.
Allow iOS to consume the 'com.apple.cfprefsd.agent' endpoint if issued by the UIProcess.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-04-27 Brent Fulgham <bfulgham@apple.com>
[MacCatalyst] Dynamic access to accessibility services is incomplete
https://bugs.webkit.org/show_bug.cgi?id=211100
<rdar://problem/62133491>
Reviewed by Darin Adler.
Testing of MacCatalyst-based apps shows that using dynamic sandbox extensions to support accessibility
features is incomplete. In addition to the 'com.apple.cfprefsd.daemon' process needed on macOS and iOS,
we also need access to 'com.apple.cfprefsd.agent'.
This can only be tested manually in a MacCatlyst build, though existing accessibility tests will
confirm no regressions in iOS and macOS.
The changes are as follows:
1. Change WebProcessCreationParameters::preferencesExtensionHandle to a handle array.
2. Register (and unregister) for the accessibility enabled notification in MacCatalyst.
3. Open a connection 'com.apple.cfprefsd.agent' (in addition to the existing 'com.apple.cfprefsd.daemon')
in MacCatalyst builds.
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
(WebKit::WebProcessPool::registerNotificationObservers):
(WebKit::WebProcessPool::unregisterNotificationObservers):
* UIProcess/Cocoa/WebProcessProxyCocoa.mm:
(WebKit::WebProcessProxy::unblockPreferenceServiceIfNeeded):
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
(WebKit::WebProcess::unblockPreferenceService):
* WebProcess/com.apple.WebProcess.sb.in:
2020-04-27 Wenson Hsieh <wenson_hsieh@apple.com>
Prevent WebPasteboardProxy IPC messages from unconditionally requesting pasteboard types
https://bugs.webkit.org/show_bug.cgi?id=211061
<rdar://problem/62088185>
Reviewed by Tim Horton.
Tighten IPC message handling in WebPasteboardProxy, such that a web content process is unable to arbitrarily
request pasteboard types (and other metadata that depends on pasteboard types, such as the number of files
on the pasteboard).
This extends the mechanisms added in r259151 by distinguishing between allowing the web process to request
type information from the pasteboard, and allowing the web process to request both types and data. In the
case of drag and drop, the former is required by the web process from the moment the drag session enters
the platform web view; however, the latter is required only after a drop is performed.
We utilize this new check, `canAccessPasteboardTypes`, in seven existing IPC message handlers that either
directly return the list of types, or return metadata from which pasteboard types may be deduced.
See below for more details; covered by existing API and layout tests.
* UIProcess/Cocoa/WebPasteboardProxyCocoa.mm:
(WebKit::WebPasteboardProxy::grantAccessToCurrentTypes):
(WebKit::WebPasteboardProxy::grantAccessToCurrentData):
Split the pasteboard access check into two helpers: one to determine whether the web process is allowed to
request pasteboard types, and another to determine whether the web process is allowed to request pasteboard
data (in addition to types).
(WebKit::WebPasteboardProxy::grantAccess):
(WebKit::WebPasteboardProxy::revokeAccess):
Refactor these to use `PasteboardAccessInformation`; see below for more details.
(WebKit::WebPasteboardProxy::canAccessPasteboardTypes const):
Return true as long as the accessType is nonnull; this is because either `Types` or `TypesAndData` is
sufficient for access to pasteboard types.
(WebKit::WebPasteboardProxy::canAccessPasteboardData const):
(WebKit::WebPasteboardProxy::accessType const):
Add a private helper method to return the access type (`nullopt` for no access, `PasteboardAccessType::Types`
for access to types only, and `PasteboardAccessType::TypesAndData` for access to data in addition to types)
for the given named pasteboard.
(WebKit::WebPasteboardProxy::didModifyContentsOfPasteboard):
Refactor this to use `PasteboardAccessInformation`.
(WebKit::WebPasteboardProxy::getPasteboardTypes):
(WebKit::WebPasteboardProxy::getPasteboardPathnamesForType):
Adopt canAccessPasteboardTypes here, and update the FIXME to state that we should really be using
canAccessPasteboardData instead. This is blocked on refactoring some more drag and drop code, such
that accessing the list of file names on the pasteboard isn't necessary before the drop is handled.
(WebKit::WebPasteboardProxy::containsURLStringSuitableForLoading):
(WebKit::WebPasteboardProxy::getNumberOfFiles):
(WebKit::WebPasteboardProxy::typesSafeForDOMToReadAndWrite):
(WebKit::WebPasteboardProxy::allPasteboardItemInfo):
(WebKit::WebPasteboardProxy::informationForItemAtIndex):
(WebKit::WebPasteboardProxy::getPasteboardItemsCount):
(WebKit::WebPasteboardProxy::containsStringSafeForDOMToReadForType):
Adopt the canAccessPasteboardTypes check in various places.
(WebKit::WebPasteboardProxy::PasteboardAccessInformation::grantAccess):
(WebKit::WebPasteboardProxy::PasteboardAccessInformation::revokeAccess):
(WebKit::WebPasteboardProxy::PasteboardAccessInformation::accessType const):
Introduce PasteboardAccessInformation, which replaces the current `std::pair<uint64_t, WeakHashSet<WebProcessProxy>>`
that is used to represent which web processes are allowed to access certain changeCounts, for a given pasteboard name.
Instead of maintaining a changeCount and set of processes, we now maintain a changeCount and list of (process, access
type) pairs. Each item in this list tracks whether the process that has been granted access should be allowed to read
only pasteboard types, or both types and data.
(WebKit::WebPasteboardProxy::revokeAccessToAllData): Deleted.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::dragEntered):
(WebKit::WebPageProxy::dragUpdated):
When updating the drag session (i.e. as the user drags over the web view), ensure that the web content
process is able to request pasteboard types.
* UIProcess/WebPasteboardProxy.cpp:
(WebKit::WebPasteboardProxy::allPasteboardItemInfo):
(WebKit::WebPasteboardProxy::informationForItemAtIndex):
(WebKit::WebPasteboardProxy::getPasteboardItemsCount):
(WebKit::WebPasteboardProxy::containsStringSafeForDOMToReadForType):
(WebKit::WebPasteboardProxy::containsURLStringSuitableForLoading):
Update these method signatures.
* UIProcess/WebPasteboardProxy.h:
* UIProcess/WebPasteboardProxy.messages.in:
Mark more IPC messages as `WantsConnection` so that we can use the `IPC::Connection` argument in the message handlers.
2020-04-27 Alex Christensen <achristensen@webkit.org>
Stop waiting for a BinarySemaphore on the main thread in the NetworkProcess
https://bugs.webkit.org/show_bug.cgi?id=211080
<rdar://problem/62377357>
Reviewed by Darin Adler and Chris Dumez.
There was an out-of-date comment suggesting we needed to use a semaphore, but we've since added these in the destructor:
RELEASE_ASSERT(!m_statisticsStore);
RELEASE_ASSERT(!m_persistentStorage);
This indicates that flushAndDestroyPersistentStore is called before the destructor, at which time it is safe to add a reference to keep it alive.
WebResourceLoadStatisticsStore is also marked as WTF::DestructionThread::Main so this should do everything we need.
We also flush these databases to disk before closing like we did cookies.
In order to keep tests working as they are, I needed to make recreateResourceLoadStatisticStore have a CompletionHandler and have all
WebResourceLoadStatisticsStores share the same queue to serialize background tasks between WebResourceLoadStatisticsStores with and without database stores
sequentially to avoid opening a SQLiteDatabase before the previous WebResourceLoadStatisticsStore had closed it.
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::sharedStatisticsQueue):
(WebKit::WebResourceLoadStatisticsStore::WebResourceLoadStatisticsStore):
(WebKit::WebResourceLoadStatisticsStore::didDestroyNetworkSession):
(WebKit::WebResourceLoadStatisticsStore::flushAndDestroyPersistentStore):
(WebKit::WebResourceLoadStatisticsStore::applicationWillTerminate): Deleted.
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.h:
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::didClose):
* NetworkProcess/NetworkSession.cpp:
(WebKit::NetworkSession::~NetworkSession):
(WebKit::NetworkSession::destroyResourceLoadStatistics):
(WebKit::NetworkSession::setResourceLoadStatisticsEnabled):
(WebKit::NetworkSession::recreateResourceLoadStatisticStore):
* NetworkProcess/NetworkSession.h:
2020-04-27 Alex Christensen <achristensen@webkit.org>
Fix tests after r260764
https://bugs.webkit.org/show_bug.cgi?id=211093
Some tests were asserting in takeAsyncReplyHandler.
Calling addAsyncReplyHandler before sending the message fixes it.
* UIProcess/AuxiliaryProcessProxy.cpp:
(WebKit::AuxiliaryProcessProxy::sendMessage):
2020-04-27 Chris Dumez <cdumez@apple.com>
[iOS][WK2] Drop "com.apple.WebKit.DatabaseActivity" UIKit background activity
https://bugs.webkit.org/show_bug.cgi?id=211084
Reviewed by Geoffrey Garen.
Drop "com.apple.WebKit.DatabaseActivity" UIKit background activity in the UIProcess now that local storage
has been moved to the network process.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _initializeWithConfiguration:]):
(-[WKWebView _setUpSQLiteDatabaseTrackerClient]): Deleted.
2020-04-27 Ryan Haddad <ryanhaddad@apple.com>
[Cocoa] stop using out arguments for document attributes when converting to attributed strings
https://bugs.webkit.org/show_bug.cgi?id=211048
Unreviewed partial revert of r260739 to remove unexpected logging that broke Catalina-Release-WK2-Perf tests.
* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::updateTaskWithFirstPartyForSameSiteCookies):
2020-04-27 David Kilzer <ddkilzer@apple.com>
REGRESSION (r219050): Remove more code only needed for building with OS X Yosemite
<https://webkit.org/b/211088>
<rdar://problem/62452487>
Reviewed by Chris Dumez.
* Platform/IPC/cocoa/ConnectionCocoa.mm:
(IPC::Connection::receiveSourceEventHandler):
* Platform/mac/MachUtilities.cpp:
(setMachPortQueueLength):
(machExceptionPort): Deleted.
(setMachExceptionPort): Deleted.
* Platform/mac/MachUtilities.h:
- Delete unused code.
2020-04-27 Per Arne Vollan <pvollan@apple.com>
[Cocoa] After r258891, r255119 can be reverted
https://bugs.webkit.org/show_bug.cgi?id=211083
Unreviewed revert of r255119.
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-04-27 Brent Fulgham <bfulgham@apple.com>
Unreviewed MacCatalyst build fix after r260739.
* UIProcess/Cocoa/TextCheckingController.mm:
(WebKit::TextCheckingController::replaceRelativeToSelection):
* WebProcess/WebPage/Cocoa/TextCheckingControllerProxy.mm:
2020-04-27 Darin Adler <darin@apple.com>
Another try at fixing the ENABLE(PLATFORM_DRIVEN_TEXT_CHECKING) build
* UIProcess/Cocoa/TextCheckingController.mm:
(WebKit::TextCheckingController::replaceRelativeToSelection):
Convert an NSAttributedString to a WebCore::AttributedString using
argument list syntax.
2020-04-27 Alex Christensen <achristensen@webkit.org>
Fix assertion after r260709
https://bugs.webkit.org/show_bug.cgi?id=211007
<rdar://problem/62457303>
* WebProcess/Plugins/PluginView.cpp:
(WebKit::PluginView::performJavaScriptURLRequest):
Not only did I spell javascript wrong, I should've used the special protocolIsJavaScript() method.
2020-04-27 Alex Christensen <achristensen@webkit.org>
Reduce use of WebPageProxy::VoidCallback
https://bugs.webkit.org/show_bug.cgi?id=210944
Reviewed by Darin Adler.
Use sendWithAsyncReply and CompletionHandler<void()> instead.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _executeEditCommand:argument:completion:]):
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::becomeFirstResponder):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::setInitialFocus):
(WebKit::WebPageProxy::executeEditCommand):
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::focusNextFocusedElement):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView pasteWithCompletionHandler:]):
(-[WKContentView moveByOffset:]):
(-[WKContentView _selectPositionAtPoint:stayingWithinFocusedElement:completionHandler:]):
(-[WKContentView selectPositionAtBoundary:inDirection:fromPoint:completionHandler:]):
(-[WKContentView moveSelectionAtBoundary:inDirection:completionHandler:]):
(-[WKContentView selectTextWithGranularity:atPoint:completionHandler:]):
(-[WKContentView _becomeFirstResponderWithSelectionMovingForward:completionHandler:]):
(-[WKContentView accessoryTab:]):
(-[WKContentView executeEditCommandWithCallback:]):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::selectTextWithGranularityAtPoint):
(WebKit::WebPageProxy::selectPositionAtBoundaryWithDirection):
(WebKit::WebPageProxy::moveSelectionAtBoundaryWithDirection):
(WebKit::WebPageProxy::selectPositionAtPoint):
(WebKit::WebPageProxy::moveSelectionByOffset):
(WebKit::WebPageProxy::focusNextFocusedElement):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::executeEditCommandWithCallback):
(WebKit::WebPage::setInitialFocus):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::moveSelectionByOffset):
(WebKit::WebPage::selectPositionAtPoint):
(WebKit::WebPage::selectPositionAtBoundaryWithDirection):
(WebKit::WebPage::moveSelectionAtBoundaryWithDirection):
(WebKit::WebPage::selectTextWithGranularityAtPoint):
(WebKit::WebPage::focusNextFocusedElement):
2020-04-27 Per Arne Vollan <pvollan@apple.com>
Unreviewed sandbox compile fix.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-04-27 Philippe Normand <pnormand@igalia.com>
[GTK] ENABLE(OPENGL) remmants...
https://bugs.webkit.org/show_bug.cgi?id=211077
Reviewed by Adrian Perez de Castro.
ENABLE(OPENGL) was renamed to ENABLE(GRAPHICS_CONTEXT_GL) in
r254064 but not in all the code base...
* UIProcess/gtk/HardwareAccelerationManager.cpp:
(WebKit::HardwareAccelerationManager::HardwareAccelerationManager):
2020-04-27 Daniel Bates <dabates@apple.com>
Caret may be placed in the wrong spot for text input context that is a form control
https://bugs.webkit.org/show_bug.cgi?id=210939
<rdar://problem/61943089>
Reviewed by Darin Adler.
Find the closest editable position in the element for the point using the
newly introduced closestEditablePositionInElementForAbsolutePoint().
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::focusTextInputContextAndPlaceCaret):
2020-04-27 Darin Adler <darin@apple.com>
Fix ENABLE(PLATFORM_DRIVEN_TEXT_CHECKING) build
* WebProcess/WebPage/Cocoa/TextCheckingControllerProxy.h:
Change "class AttributedString" to "struct AttributedString".
2020-04-27 Philippe Normand <pnormand@igalia.com>
[WPE] about:gpu page is incomplete
https://bugs.webkit.org/show_bug.cgi?id=211076
Reviewed by Adrian Perez de Castro.
ENABLE(OPENGL) was renamed to ENABLE(GRAPHICS_CONTEXT_GL) in
r254064 but not in all the code base...
* UIProcess/API/glib/WebKitProtocolHandler.cpp:
(WebKit::WebKitProtocolHandler::handleGPU):
2020-04-26 Darin Adler <darin@apple.com>
Replace more uses of live ranges with SimpleRange
https://bugs.webkit.org/show_bug.cgi?id=211058
Reviewed by Antti Koivisto.
* Shared/API/Cocoa/WKDataDetectorTypesInternal.h:
(fromWKDataDetectorTypes): Updated since DataDetectorTypes
is now an enum class. Also got rid of special "none" and "all" values.
* WebProcess/InjectedBundle/API/Cocoa/WKWebProcessPlugInRangeHandle.mm:
(-[WKWebProcessPlugInRangeHandle detectDataWithTypes:context:]): Updated
since DataDetection now takes a SimpleRange.
* WebProcess/WebPage/ViewGestureGeometryCollector.cpp:
(WebKit::ViewGestureGeometryCollector::computeTextLegibilityScales):
Use SimpleRange.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::detectDataInAllFrames): Ditto.
* WebProcess/WebPage/WebPage.h: Removed unneeded include of DataDetection.h.
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::performImmediateActionHitTestAtLocation): Use SimpleRange.
2020-04-27 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Make it possible to build with GTK4 without errors
https://bugs.webkit.org/show_bug.cgi?id=210967
Reviewed by Adrian Perez de Castro.
Add ifdefs when needed to make it possible to complete a build with GTK4. This way we can continue working on
every feature individually and checking it actually works. Move public headers containing API specific to GTK3
and GTK4 into their own gtk3 and gtk4 directory.
* PlatformGTK.cmake:
* Shared/API/glib/WebKitContextMenuItem.cpp:
* Shared/NativeWebKeyboardEvent.h:
* Shared/NativeWebMouseEvent.h:
* Shared/NativeWebWheelEvent.h:
* Shared/glib/WebContextMenuItemGlib.cpp:
(WebKit::WebContextMenuItemGlib::WebContextMenuItemGlib):
(WebKit::WebContextMenuItemGlib::buildActionName const):
(WebKit::WebContextMenuItemGlib::createActionIfNeeded):
* Shared/glib/WebContextMenuItemGlib.h:
* Shared/gtk/NativeWebKeyboardEventGtk.cpp:
* Shared/gtk/NativeWebMouseEventGtk.cpp:
(WebKit::NativeWebMouseEvent::NativeWebMouseEvent):
* Shared/gtk/NativeWebTouchEventGtk.cpp:
(WebKit::NativeWebTouchEvent::NativeWebTouchEvent):
* Shared/gtk/NativeWebWheelEventGtk.cpp:
* Shared/gtk/WebEventFactory.cpp:
(WebKit::buttonForEvent):
(WebKit::WebEventFactory::createWebMouseEvent):
(WebKit::WebEventFactory::createWebKeyboardEvent):
(WebKit::WebEventFactory::createWebTouchEvent):
* Shared/gtk/WebEventFactory.h:
* SourcesGTK.txt:
* UIProcess/API/C/WKNativeEvent.h:
* UIProcess/API/glib/InputMethodFilter.cpp:
(WebKit::InputMethodFilter::filterKeyEvent):
* UIProcess/API/glib/InputMethodFilter.h:
* UIProcess/API/glib/WebKitInstallMissingMediaPluginsPermissionRequest.cpp:
(createGstInstallPluginsContext):
* UIProcess/API/glib/WebKitPrivate.cpp:
* UIProcess/API/glib/WebKitUIClient.cpp:
(UIClient::setWindowFrameTimerFired):
* UIProcess/API/glib/WebKitWebContext.cpp:
(webkitWebContextGetProperty):
(webkitWebContextSetProperty):
(webkit_web_context_class_init):
* UIProcess/API/glib/WebKitWebView.cpp:
(webkitWebViewRunAsModal):
(webkitWebViewPopulateContextMenu):
* UIProcess/API/gtk/InputMethodFilterGtk.cpp:
(WebKit::InputMethodFilter::platformEventKeyIsKeyPress const):
* UIProcess/API/gtk/PageClientImpl.cpp:
(WebKit::PageClientImpl::setViewNeedsDisplay):
(WebKit::PageClientImpl::setCursor):
(WebKit::PageClientImpl::doneWithKeyEvent):
(WebKit::PageClientImpl::createPopupMenuProxy):
(WebKit::PageClientImpl::createDataListSuggestionsDropdown):
(WebKit::PageClientImpl::startDrag):
(WebKit::PageClientImpl::doneWithTouchEvent):
(WebKit::PageClientImpl::wheelEventWasNotHandledByWebCore):
* UIProcess/API/gtk/WebKitAuthenticationDialog.cpp:
* UIProcess/API/gtk/WebKitAuthenticationDialog.h:
* UIProcess/API/gtk/WebKitEmojiChooser.cpp:
* UIProcess/API/gtk/WebKitEmojiChooser.h:
* UIProcess/API/gtk/WebKitInputMethodContextGtk.cpp:
(webkit_input_method_context_filter_key_event):
* UIProcess/API/gtk/WebKitInputMethodContextImplGtk.cpp:
(webkitInputMethodContextImplGtkFilterKeyEvent):
(webkitInputMethodContextImplGtkSetClientWindow):
* UIProcess/API/gtk/WebKitPopupMenu.cpp:
(WebKit::WebKitPopupMenu::WebKitPopupMenu):
(WebKit::WebKitPopupMenu::showPopupMenu):
(WebKit::WebKitPopupMenu::hidePopupMenu):
(WebKit::WebKitPopupMenu::cancelTracking):
* UIProcess/API/gtk/WebKitPopupMenu.h:
* UIProcess/API/gtk/WebKitPrintOperation.cpp:
* UIProcess/API/gtk/WebKitScriptDialogGtk.cpp:
(webkitScriptDialogAccept):
(webkitScriptDialogDismiss):
(webkitScriptDialogSetUserInput):
* UIProcess/API/gtk/WebKitScriptDialogImpl.cpp:
* UIProcess/API/gtk/WebKitScriptDialogImpl.h:
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(ClickCounter::currentClickCountForGdkButtonEvent):
(_WebKitWebViewBasePrivate::_WebKitWebViewBasePrivate):
(webkitWebViewBaseSetToplevelOnScreenWindow):
(webkitWebViewBaseRealize):
(webkitWebViewBaseContainerForall):
(webkitWebViewBaseDispose):
(webkitWebViewBaseConstructed):
(webkitWebViewBaseSizeAllocate):
(webkitWebViewBaseKeyPressEvent):
(webkitWebViewBaseKeyReleaseEvent):
(webkitWebViewBaseSetEnableBackForwardNavigationGesture):
(webkitWebViewBaseBeginBackSwipeForTesting):
(webkitWebViewBaseCompleteBackSwipeForTesting):
(webkit_web_view_base_class_init):
(webkitWebViewBaseWillSwapWebProcess):
(webkitWebViewBaseDidExitWebProcess):
(webkitWebViewBaseDidRelaunchWebProcess):
(webkitWebViewBaseTakeViewSnapshot):
(webkitWebViewBaseDidStartProvisionalLoadForMainFrame):
(webkitWebViewBaseDidFirstVisuallyNonEmptyLayoutForMainFrame):
(webkitWebViewBaseDidFinishLoadForMainFrame):
(webkitWebViewBaseDidFailLoadForMainFrame):
(webkitWebViewBaseDidSameDocumentNavigationForMainFrame):
(webkitWebViewBaseDidRestoreScrollPosition):
(webkitWebViewBaseShowEmojiChooser):
(webkitWebViewBaseRequestPointerLock):
(webkitWebViewBaseDidLosePointerLock):
(webkitWebViewBaseSynthesizeCompositionKeyPress):
* UIProcess/API/gtk/WebKitWebViewBasePrivate.h:
* UIProcess/API/gtk/WebKitWebViewDialog.cpp:
* UIProcess/API/gtk/WebKitWebViewDialog.h:
* UIProcess/API/gtk/WebKitWebViewGtk.cpp:
(webkitWebViewAuthenticate):
(webkitWebViewScriptDialog):
(webkitWebViewRunFileChooser):
(webkitWebViewMaximizeWindow):
(webkitWebViewMinimizeWindow):
(webkitWebViewRestoreWindow):
* UIProcess/API/gtk3/WebKitContextMenuItem.h: Copied from Source/WebKit/UIProcess/API/gtk/WebKitContextMenuItem.h.
* UIProcess/API/gtk3/WebKitInputMethodContext.h: Copied from Source/WebKit/UIProcess/API/gtk/WebKitInputMethodContext.h.
* UIProcess/API/gtk4/WebKitContextMenuItem.h: Renamed from Source/WebKit/UIProcess/API/gtk/WebKitContextMenuItem.h.
* UIProcess/API/gtk4/WebKitInputMethodContext.h: Renamed from Source/WebKit/UIProcess/API/gtk/WebKitInputMethodContext.h.
* UIProcess/Automation/gtk/WebAutomationSessionGtk.cpp:
(WebKit::WebAutomationSession::platformSimulateMouseInteraction):
(WebKit::WebAutomationSession::platformSimulateKeyboardInteraction):
(WebKit::WebAutomationSession::platformSimulateKeySequence):
* UIProcess/Inspector/gtk/RemoteWebInspectorProxyGtk.cpp:
(WebKit::RemoteWebInspectorProxy::platformSave):
* UIProcess/Inspector/gtk/WebInspectorProxyGtk.cpp:
(WebKit::WebInspectorProxy::platformSave):
* UIProcess/Inspector/gtk/WebKitInspectorWindow.cpp:
(webkit_inspector_window_init):
(webkitInspectorWindowNew):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/cairo/BackingStoreCairo.cpp:
(WebKit::BackingStore::createBackend):
* UIProcess/glib/WebProcessPoolGLib.cpp:
(WebKit::WebProcessPool::platformInitializeWebProcess):
* UIProcess/gtk/AcceleratedBackingStoreWayland.cpp:
(WebKit::AcceleratedBackingStoreWayland::tryEnsureGLContext):
(WebKit::AcceleratedBackingStoreWayland::paint):
* UIProcess/gtk/AcceleratedBackingStoreX11.cpp:
(WebKit::AcceleratedBackingStoreX11::~AcceleratedBackingStoreX11):
(WebKit::AcceleratedBackingStoreX11::update):
* UIProcess/gtk/DragAndDropHandler.cpp:
* UIProcess/gtk/DragAndDropHandler.h:
* UIProcess/gtk/GestureController.cpp:
* UIProcess/gtk/GestureController.h:
* UIProcess/gtk/KeyBindingTranslator.cpp:
(WebKit::KeyBindingTranslator::KeyBindingTranslator):
(WebKit::KeyBindingTranslator::commandsForKeyEvent):
* UIProcess/gtk/KeyBindingTranslator.h:
* UIProcess/gtk/PointerLockManager.cpp:
(WebKit::PointerLockManager::lock):
(WebKit::PointerLockManager::unlock):
* UIProcess/gtk/PointerLockManager.h:
* UIProcess/gtk/PointerLockManagerWayland.cpp:
(WebKit::PointerLockManagerWayland::lock):
* UIProcess/gtk/PointerLockManagerX11.cpp:
(WebKit::PointerLockManagerX11::didReceiveMotionEvent):
* UIProcess/gtk/WebColorPickerGtk.cpp:
* UIProcess/gtk/WebContextMenuProxyGtk.cpp:
(WebKit::WebContextMenuProxyGtk::show):
(WebKit::WebContextMenuProxyGtk::showContextMenuWithItems):
(WebKit::WebContextMenuProxyGtk::WebContextMenuProxyGtk):
(WebKit::WebContextMenuProxyGtk::~WebContextMenuProxyGtk):
* UIProcess/gtk/WebContextMenuProxyGtk.h:
* UIProcess/gtk/WebDataListSuggestionsDropdownGtk.cpp:
* UIProcess/gtk/WebDataListSuggestionsDropdownGtk.h:
* UIProcess/gtk/WebPageProxyGtk.cpp:
* UIProcess/gtk/WebPopupMenuProxyGtk.cpp:
* UIProcess/gtk/WebPopupMenuProxyGtk.h:
* WebProcess/Plugins/PluginController.h:
* WebProcess/Plugins/PluginView.cpp:
* WebProcess/Plugins/PluginView.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::close):
(WebKit::WebPage::mainFrameDidLayout):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/gtk/AcceleratedSurfaceX11.cpp:
(WebKit::AcceleratedSurfaceX11::AcceleratedSurfaceX11):
(WebKit::AcceleratedSurfaceX11::createPixmap):
* WebProcess/WebPage/gtk/WebPrintOperationGtk.cpp:
(WebKit::WebPrintOperationGtk::print):
2020-04-27 Rob Buis <rbuis@igalia.com>
Make loadURLIntoChildFrame private and non-exported
https://bugs.webkit.org/show_bug.cgi?id=211051
Reviewed by Darin Adler.
Adapt createFrame to strictly create a subframe and
not load anything.
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::createFrame):
* WebProcess/WebCoreSupport/WebFrameLoaderClient.h:
2020-04-27 Ross Kirsling <ross.kirsling@sony.com>
[JSC] CallData/ConstructData should include CallType/ConstructType
https://bugs.webkit.org/show_bug.cgi?id=211059
Reviewed by Darin Adler.
* WebProcess/Plugins/Netscape/JSNPObject.cpp:
(WebKit::JSNPObject::getCallData):
(WebKit::JSNPObject::getConstructData):
* WebProcess/Plugins/Netscape/JSNPObject.h:
* WebProcess/Plugins/Netscape/NPJSObject.cpp:
(WebKit::NPJSObject::hasMethod):
(WebKit::NPJSObject::construct):
(WebKit::NPJSObject::invoke):
2020-04-27 Diego Pino Garcia <dpino@igalia.com>
Unreviewed, reverting r260672.
[GTK] WebInspector tests are timing out after r260672
Reverted changeset:
"[Win] Bundle Inspector Resources in Release builds"
https://bugs.webkit.org/show_bug.cgi?id=210942
https://trac.webkit.org/changeset/260672
2020-04-27 Diego Pino Garcia <dpino@igalia.com>
Unreviewed, reverting r260696.
Build fix after r260672, no needed since r260672 is also being reverted.
Reverted changeset:
"REGRESSION(210942): [GTK][WPE] Unreviewed, EWS build bots
fail in compile-webkit step"
https://bugs.webkit.org/show_bug.cgi?id=211014
https://trac.webkit.org/changeset/260696
2020-04-26 Darin Adler <darin@apple.com>
[Cocoa] stop using out arguments for document attributes when converting to attributed strings
https://bugs.webkit.org/show_bug.cgi?id=211048
Reviewed by Sam Weinig.
* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::updateTaskWithFirstPartyForSameSiteCookies): Use init instead of initWithString:@"".
* Shared/Cocoa/WebCoreArgumentCodersCocoa.mm:
(IPC::ArgumentCoder<WebCore::AttributedString>::encode): Moved here from AttributedString.mm.
(IPC::ArgumentCoder<WebCore::AttributedString>::decode): Ditto.
* Shared/DocumentEditingContext.h: Updated to use WebCore::AttributeString.
* Shared/DocumentEditingContext.mm:
(IPC::ArgumentCoder<WebKit::DocumentEditingContext>::decode): Ditto.
* Shared/WebCoreArgumentCoders.h: Added coder for WebCore::AttributedString.
* Shared/mac/AttributedString.h: Removed.
* Shared/mac/AttributedString.mm: Removed.
* SourcesCocoa.txt: Removed AttributedString.mm.
* UIProcess/API/Cocoa/WKWebView.mm: Updated to use WebCore::AttributedString.
* UIProcess/Cocoa/TextCheckingController.mm: Ditto.
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::attributedSubstringForProposedRange): Ditto.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::getContentsAsAttributedString): Ditto.
* UIProcess/WebPageProxy.h: Ditto.
* UIProcess/WebPageProxy.messages.in: Ditto.
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::attributedSubstringForCharacterRangeAsync): Ditto.
(WebKit::WebPageProxy::attributedStringForCharacterRangeCallback): Ditto.
* WebKit.xcodeproj/project.pbxproj: Removed AttributedString.h and AttributedString.mm.
* WebProcess/WebPage/Cocoa/TextCheckingControllerProxy.h: Updated to use WebCore::AttributedString.
* WebProcess/WebPage/Cocoa/TextCheckingControllerProxy.messages.in: Ditto.
* WebProcess/WebPage/Cocoa/TextCheckingControllerProxy.mm:
(WebKit::TextCheckingControllerProxy::replaceRelativeToSelection): Ditto.
(WebKit::TextCheckingControllerProxy::annotatedSubstringBetweenPositions): Ditto.
* WebProcess/WebPage/Cocoa/WebPageCocoa.mm:
(WebKit::WebPage::dictionaryPopupInfoForRange): Updated since editingAttributedString
returns a struct.
(WebKit::WebPage::getContentsAsAttributedString): Updated to use SimpleRange and
since attributedString returns a struct.
* WebProcess/WebPage/WebPage.h: Updated to use WebCore::AttributedString.
* WebProcess/WebPage/WebPage.messages.in: Ditto.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::requestDocumentEditingContext): Ditto.
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::attributedSubstringForCharacterRangeAsync): Ditto.
(WebKit::WebPage::handleSelectionServiceClick): Ditto.
2020-04-26 Said Abou-Hallawa <sabouhallawa@apple.com>
Rendering update steps should use Seconds for the timestamps
https://bugs.webkit.org/show_bug.cgi?id=210990
Reviewed by Daniel Bates.
* WebProcess/WebPage/CoordinatedGraphics/CompositingCoordinator.cpp:
(WebKit::CompositingCoordinator::timestamp const):
2020-04-26 Peng Liu <peng.liu6@apple.com>
Remove unused class PlaybackSessionInterface
https://bugs.webkit.org/show_bug.cgi?id=211031
Reviewed by Daniel Bates.
* WebProcess/cocoa/PlaybackSessionManager.h:
2020-04-11 Darin Adler <darin@apple.com>
Stop using live ranges in functions that return range of the selection
https://bugs.webkit.org/show_bug.cgi?id=210396
Reviewed by Sam Weinig.
* WebProcess/WebCoreSupport/gtk/WebEditorClientGtk.cpp:
(WebKit::WebEditorClient::updateGlobalSelection): Remove use of live range.
* WebProcess/WebPage/Cocoa/WebPageCocoa.mm:
(WebKit::WebPage::dictionaryPopupInfoForRange): Updated for changes
to attributedString functions.
(WebKit::WebPage::getContentsAsAttributedString): Ditto.
* WebProcess/WebPage/FindController.cpp:
(WebKit::FindController::updateFindUIAfterPageScroll): Use createLiveRange.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::getSelectedRangeAsync): Ditto.
(WebKit::WebPage::currentSelectionAsRange): Ditto.
* WebProcess/WebPage/WebPage.h: Change m_rangeForDropSnapshot
to use a SimpleRange instead of a live range.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::plainTextForContext): Overloaded for SimpleRange and
Optional<SimpleRange> as needed by code below.
(WebKit::plainTextForDisplay): Ditto.
(WebKit::WebPage::getPlatformEditorState const): Use createLiveRange.
(WebKit::WebPage::getSelectionContext): Remove use of live range.
(WebKit::WebPage::didConcludeDrop): Ditto.
(WebKit::WebPage::didConcludeEditDrag): Use createLiveRange.
(WebKit::WebPage::computeAndSendEditDragSnapshot): Ditto.
(WebKit::WebPage::startAutoscrollAtPosition): Remove use of
live range.
(WebKit::WebPage::requestEvasionRectsAboveSelection): Use
createLiveRange.
(WebKit::WebPage::requestDictationContext): Remove use of live range.
(WebKit::WebPage::replaceSelectedText): Use createLiveRange.
(WebKit::WebPage::applyAutocorrectionInternal): Ditto.
(WebKit::WebPage::autocorrectionContext): Remove use of live range.
(WebKit::WebPage::requestDocumentEditingContext): Use createLiveRange.
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::getPlatformEditorState const): Remove use of
live range.
(WebKit::WebPage::attributedSubstringForCharacterRangeAsync):
Updated for change to HTMLConverter functions.
(WebKit::WebPage::handleSelectionServiceClick): Ditto.
(WebKit::WebPage::performImmediateActionHitTestAtLocation):
Remove use of live range.
2020-04-25 Alex Christensen <achristensen@webkit.org>
Fix internal iOS build.
https://bugs.webkit.org/show_bug.cgi?id=210521
* Shared/ApplePay/cocoa/WebPaymentCoordinatorProxyCocoa.mm:
(WebKit::WebPaymentCoordinatorProxy::platformBeginApplePaySetup):
2020-04-25 Alex Christensen <achristensen@webkit.org>
Move ApplePay code from WebKitAdditions to WebCore and WebKit
https://bugs.webkit.org/show_bug.cgi?id=210521
2020-04-25 Alex Christensen <achristensen@webkit.org>
Move ApplePay code from WebKitAdditions to WebCore and WebKit
https://bugs.webkit.org/show_bug.cgi?id=210521
Reviewed by Andy Estes.
* DerivedSources.make:
* Scripts/webkit/messages.py:
* Shared/ApplePay/ApplePayPaymentSetupFeatures.h: Added.
(WebKit::PaymentSetupFeatures::platformFeatures const):
* Shared/ApplePay/ApplePayPaymentSetupFeatures.mm: Added.
(WebKit::PaymentSetupFeatures::PaymentSetupFeatures):
(WebKit::PaymentSetupFeatures::encode const):
(WebKit::PaymentSetupFeatures::decode):
(WebKit::PaymentSetupFeatures::operator Vector<Ref<WebCore::ApplePaySetupFeature>> const):
* Shared/ApplePay/PaymentSetupConfiguration.h: Added.
(WebKit::PaymentSetupConfiguration::platformConfiguration const):
* Shared/ApplePay/PaymentSetupConfiguration.mm: Added.
(WebKit::toPlatformConfiguration):
(WebKit::PaymentSetupConfiguration::PaymentSetupConfiguration):
(WebKit::PaymentSetupConfiguration::encode const):
(WebKit::PaymentSetupConfiguration::decode):
* Shared/ApplePay/WebPaymentCoordinatorProxy.cpp:
* Shared/ApplePay/WebPaymentCoordinatorProxy.h:
(WebKit::WebPaymentCoordinatorProxy::finishConstruction): Deleted.
* Shared/ApplePay/WebPaymentCoordinatorProxy.messages.in:
* Shared/ApplePay/cocoa/PaymentSetupConfiguration.mm: Added.
(WebKitAdditions::toPlatformConfiguration):
(WebKitAdditions::PaymentSetupConfiguration::PaymentSetupConfiguration):
(WebKitAdditions::PaymentSetupConfiguration::encode const):
(WebKitAdditions::PaymentSetupConfiguration::decode):
* Shared/ApplePay/cocoa/WebPaymentCoordinatorProxyCocoa.mm:
(WebKit::finishCreating):
(WebKit::WebPaymentCoordinatorProxy::getSetupFeatures):
(WebKit::WebPaymentCoordinatorProxy::beginApplePaySetup):
(WebKit::WebPaymentCoordinatorProxy::endApplePaySetup):
(WebKit::WebPaymentCoordinatorProxy::platformBeginApplePaySetup):
(WebKit::WebPaymentCoordinatorProxy::platformEndApplePaySetup):
* Shared/Cocoa/WebCoreArgumentCodersCocoa.mm:
(IPC::finishDecoding):
(IPC::finishEncoding):
(IPC::ArgumentCoder<WebCore::PaymentInstallmentConfiguration>::encode):
(IPC::ArgumentCoder<WebCore::PaymentInstallmentConfiguration>::decode):
* SourcesCocoa.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/ApplePay/WebPaymentCoordinator.cpp:
(WebKit::WebPaymentCoordinator::getSetupFeatures):
(WebKit::WebPaymentCoordinator::beginApplePaySetup):
(WebKit::WebPaymentCoordinator::endApplePaySetup):
* WebProcess/ApplePay/WebPaymentCoordinator.h:
2020-04-25 Darin Adler <darin@apple.com>
[Cocoa] Deal with another round of Xcode upgrade checks
https://bugs.webkit.org/show_bug.cgi?id=211027
Reviewed by Alexey Proskuryakov.
* WebKit.xcodeproj/project.pbxproj: Bump the upgrade check version.
Add a base localization.
2020-04-25 Alex Christensen <achristensen@webkit.org>
Prepare to remove automatic URL->String conversion operators
https://bugs.webkit.org/show_bug.cgi?id=211007
Reviewed by Darin Adler.
* Shared/API/APIError.h:
(API::Error::failingURL const):
* Shared/API/c/WKSharedAPICast.h:
(WebKit::toCopiedURLAPI):
* UIProcess/API/Cocoa/WKBrowsingContextController.mm:
(setUpPagePolicyClient):
* UIProcess/Cocoa/SafeBrowsingWarningCocoa.mm:
(WebKit::reportAnErrorURL):
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::requestUserMediaAuthorizationForFrame):
(WebKit::UIDelegate::UIClient::checkUserMediaPermissionForOrigin):
* UIProcess/Cocoa/WebPasteboardProxyCocoa.mm:
(WebKit::WebPasteboardProxy::readURLFromPasteboard):
* UIProcess/ProvisionalPageProxy.cpp:
(WebKit::ProvisionalPageProxy::cancel):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::maybeInitializeSandboxExtensionHandle):
(WebKit::WebPageProxy::loadRequestWithNavigationShared):
(WebKit::WebPageProxy::loadAlternateHTML):
(WebKit::WebPageProxy::didStartProvisionalLoadForFrameShared):
(WebKit::WebPageProxy::didExplicitOpenForFrame):
(WebKit::WebPageProxy::didReceiveServerRedirectForProvisionalLoadForFrameShared):
(WebKit::WebPageProxy::didChangeProvisionalURLForFrameShared):
(WebKit::WebPageProxy::didSameDocumentNavigationForFrame):
(WebKit::WebPageProxy::decidePolicyForNavigationAction):
(WebKit::WebPageProxy::createNewPage):
* WebProcess/InjectedBundle/API/Cocoa/WKWebProcessPlugInFrame.mm:
(-[WKWebProcessPlugInFrame URL]):
* WebProcess/Plugins/Netscape/NetscapePluginStream.cpp:
(WebKit::NetscapePluginStream::didReceiveResponse):
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::clickedLink):
(WebKit::PDFPlugin::openWithNativeApplication):
* WebProcess/Plugins/PluginView.cpp:
(WebKit::PluginView::performURLRequest):
(WebKit::PluginView::performJavaScriptURLRequest):
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::dispatchWillPerformClientRedirect):
(WebKit::WebFrameLoaderClient::dispatchDidFailProvisionalLoad):
(WebKit::WebFrameLoaderClient::shouldForceUniversalAccessFromLocalURL):
* WebProcess/WebPage/WebFrame.cpp:
(WebKit::WebFrame::info const):
* WebProcess/WebStorage/StorageAreaMap.cpp:
(WebKit::StorageAreaMap::setItem):
(WebKit::StorageAreaMap::removeItem):
(WebKit::StorageAreaMap::clear):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::origin):
2020-04-25 Darin Adler <darin@apple.com>
Move URL to use StringView when returning substrings of the URL
https://bugs.webkit.org/show_bug.cgi?id=210431
Reviewed by Anders Carlsson.
* NetworkProcess/NetworkLoadChecker.cpp:
(WebKit::NetworkLoadChecker::checkRedirection): Use hasCredentials.
(WebKit::NetworkLoadChecker::applyHTTPSUpgradeIfNeeded const):
Remove use of ASCIILiteral for setProtocol.
* NetworkProcess/curl/NetworkDataTaskCurl.cpp:
(WebKit::NetworkDataTaskCurl::NetworkDataTaskCurl): Updated for
password name change.
(WebKit::NetworkDataTaskCurl::willPerformHTTPRedirection): Ditto.
* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::NetworkDataTaskCocoa::NetworkDataTaskCocoa): Updated
for password name change.
(WebKit::NetworkDataTaskCocoa::willPerformHTTPRedirection): Ditto.
* NetworkProcess/soup/NetworkDataTaskSoup.cpp:
(WebKit::NetworkDataTaskSoup::NetworkDataTaskSoup): Updated for
password name change.
(WebKit::NetworkDataTaskSoup::applyAuthenticationToRequest): Ditto.
(WebKit::NetworkDataTaskSoup::continueHTTPRedirection): Ditto.
(WebKit::NetworkDataTaskSoup::shouldAllowHSTSPolicySetting const):
Use != to compare hosts, rather than hostsAreEqual.
(WebKit::NetworkDataTaskSoup::shouldAllowHSTSProtocolUpgrade const):
Refactored to use || to match the function above.
* Shared/API/APIURL.h:
(API::URL::host const): Removed validity check; WTF::URL::host
returns null if the URL is invalid.
(API::URL::protocol const): Ditto.
(API::URL::path const): Ditto.
(API::URL::lastPathComponent const): Ditto. Also added toString
since WTF::URL::lastPathComponent now returns a StringView.
* UIProcess/Cocoa/DownloadClient.mm:
(WebKit::DownloadClient::didFinish): Use hasFragmentIdentifier.
* UIProcess/DeviceIdHashSaltStorage.cpp:
(WebKit::DeviceIdHashSaltStorage::loadStorageFromDisk):
Refactored a bit and use fileURLWithFileSystemPath and
updated since lastPathComponent returns a StringView.
* UIProcess/Plugins/PluginInfoStore.cpp:
(WebKit::pathExtension): Updated since lastPathComponent
returns a StringView. Refactored a little.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::maybeInitializeSandboxExtensionHandle):
Use truncatedForUseAsBase.
(WebKit::WebPageProxy::decidePolicyForNavigationAction):
Use != to compare hosts, rather than hostsAreEqual.
(WebKit::WebPageProxy::decidePolicyForNewWindowAction): Ditto.
(WebKit::WebPageProxy::createNewPage): Ditto.
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::assumeReadAccessToBaseURL):
Use truncatedForUseAsBase.
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::WebsiteDataStore::initializeAppBoundDomains):
Don't use ASCIILiteral for argument to setProtocol.
* WebProcess/MediaCache/WebMediaKeyStorageManager.cpp:
(WebKit::WebMediaKeyStorageManager::getMediaKeyOrigins):
Use fileURLWithFileSystemPath, also refactored a bit.
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::installPDFDocument): Updated since
fragmentIdentifier returns a StringView.
* WebProcess/WebCoreSupport/WebContextMenuClient.cpp:
(WebKit::WebContextMenuClient::searchWithGoogle): Streamlined
the implementation a bit.
* WebProcess/WebCoreSupport/mac/WebDragClientMac.mm:
(WebKit::WebDragClient::declareAndWriteDragImage):
Updated since lastPathComponent returns a StringView.
2020-04-25 David Kilzer <ddkilzer@apple.com>
IPC::Decoder::isInvalid() should be renamed to isValid()
<https://webkit.org/b/211000>
Reviewed by Darin Adler.
Negative logic is more difficult to reason about than positive
logic.
* Platform/IPC/Connection.cpp:
(IPC::Connection::dispatchMessage):
(IPC::Connection::dispatchWorkQueueMessageReceiverMessage):
(IPC::Connection::dispatchThreadMessageReceiverMessage):
(IPC::Connection::dispatchSyncMessage):
* Platform/IPC/Connection.h:
(IPC::Connection::sendWithAsyncReply):
* Platform/IPC/Decoder.cpp:
(IPC::Decoder::create):
* Platform/IPC/Decoder.h:
(IPC::Decoder::isValid const): Rename from isInvalid()
and invert logic.
(IPC::Decoder::isInvalid const): Rename to isValid().
* Platform/IPC/MessageSender.h:
* UIProcess/AuxiliaryProcessProxy.h:
(WebKit::AuxiliaryProcessProxy::sendWithAsyncReply):
- Change Decoder::isInvalid() to Decoder::isValid() and reverse
the Boolean logic.
2020-04-25 Diego Pino Garcia <dpino@igalia.com>
REGRESSION(210942): [GTK][WPE] EWS build bots fail in compile-webkit step
https://bugs.webkit.org/show_bug.cgi?id=211014
Make command that generates 'InspectorGResourceBundle.xml' depend on target
WebInspectorUI, instead of file 'inspector-resources.stamp'.
* InspectorGResources.cmake:
2020-04-24 Diego Pino Garcia <dpino@igalia.com>
[WPE][Debug] Unreviewed, fix build after r260063
https://bugs.webkit.org/show_bug.cgi?id=211009
r260063 removed include of WebProcess.h in WebSocketProvider.cpp,
which broke unified builds in WPE Debug.
* WebProcess/Network/WebSocketProvider.cpp:
2020-04-24 David Kilzer <ddkilzer@apple.com>
WebPasteboardProxy::getPasteboardStringsForType() and WebPasteboardProxy::readURLFromPasteboard() should check return value of SharedMemory::createHandle()
<https://webkit.org/b/211002>
Reviewed by Wenson Hsieh.
* UIProcess/Cocoa/WebPasteboardProxyCocoa.mm:
(WebKit::WebPasteboardProxy::getPasteboardBufferForType):
(WebKit::WebPasteboardProxy::readBufferFromPasteboard):
- Check result of SharedMemory::createHandle() and return early
on failure.
2020-04-24 Chris Dumez <cdumez@apple.com>
[iOS] Unable to sign up on twitter.com
https://bugs.webkit.org/show_bug.cgi?id=211003
<rdar://problem/58804852>
Reviewed by Darin Adler.
* UIProcess/WebPageProxy.cpp:
* UIProcess/WebPageProxy.h:
* UIProcess/ios/WKApplicationStateTrackingView.mm:
(-[WKApplicationStateTrackingView _willBeginSnapshotSequence]):
(-[WKApplicationStateTrackingView _didCompleteSnapshotSequence]):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::setShouldFireEvents):
(WebKit::WebPage::setShouldFireResizeEvents): Deleted.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-04-24 Christopher Reid <chris.reid@sony.com>
[Win] Bundle Inspector Resources in Release builds
https://bugs.webkit.org/show_bug.cgi?id=210942
Reviewed by Fujii Hironori.
Move CMake logic for the inspector resource copy script to
Source/WebInspectorUI so it can be shared with Win and other platforms.
* InspectorGResources.cmake:
* PlatformWPE.cmake:
* PlatformWin.cmake:
2020-04-24 Kate Cheney <katherine_cheney@apple.com>
Removing website data for a domain should delete corresponding ITP entry
https://bugs.webkit.org/show_bug.cgi?id=210864
<rdar://problem/59473193>
Reviewed by John Wilander.
Tests: http/tests/resourceLoadStatistics/remove-website-data-for-origin-deletes-link-decoration-database.html
http/tests/resourceLoadStatistics/remove-website-data-for-origin-deletes-link-decoration.html
http/tests/resourceLoadStatistics/remove-website-data-for-origin-deletes-mixed-statistics-entries-database.html
http/tests/resourceLoadStatistics/remove-website-data-for-origin-deletes-mixed-statistics-entries.html
http/tests/resourceLoadStatistics/remove-website-data-for-origin-deletes-third-party-script-loads-database.html
http/tests/resourceLoadStatistics/remove-website-data-for-origin-deletes-third-party-script-loads.html
In the database store: deletes domain from the ObservedDomains table
when website data is deleted for that domain. This deletes every
instance of the domainID in the database due to cascading deletions.
In the memory store: deletes every instance of the domain in the
statistics map, which will update the plist.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::removeDataForDomain):
(WebKit::ResourceLoadStatisticsDatabaseStore::domainIDExistsInDatabase):
Needed a new function to check for the existence of domainID in any
table. Existing queries rely on the ObservedDomains entry, which means
testing using those could result in a false positive if the domainID
was deleted from ObservedDomains but is floating around in another
table.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:
(WebKit::ResourceLoadStatisticsMemoryStore::removeDataForDomain):
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.h:
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::domainIDExistsInDatabase):
(WebKit::WebResourceLoadStatisticsStore::removeDataForDomain):
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.h:
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::domainIDExistsInDatabase):
(WebKit::NetworkProcess::deleteWebsiteDataForOrigins):
Pass RegistrableDomains vector to deleteWebsiteDataForOrigins,
captured based on the WebsiteData display name.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreRemoveITPDataForDomain):
(WKWebsiteDataStoreDoesStatisticsDomainIDExistInDatabase):
* UIProcess/API/C/WKWebsiteDataStoreRef.h:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::deleteWebsiteDataForOrigins):
(WebKit::NetworkProcessProxy::domainIDExistsInDatabase):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::removeData):
(WebKit::WebsiteDataStore::domainIDExistsInDatabase):
* UIProcess/WebsiteData/WebsiteDataStore.h:
2020-04-24 David Kilzer <ddkilzer@apple.com>
Use CocoaImage platform abstraction for NSImage/UIImage
<https://webkit.org/b/210974>
Reviewed by Darin Adler.
* Platform/cocoa/CocoaImage.h: Add.
- Define CocoaImage here for cross-platform use. Don't use
OBJC_CLASS() here because this is an Objective-C header.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView takeSnapshotWithConfiguration:completionHandler:]):
- Combine separate platform-specific methods into one method.
* UIProcess/API/Cocoa/_WKActivatedElementInfo.mm:
- Combine separate platform-specific instance variables into
one instance variable.
(-[_WKActivatedElementInfo image]):
- Combine separate methods into one platform-specific method.
* UIProcess/QuickLookThumbnailLoader.h:
- Move cross-platform definition to CocoaImge.h.
* UIProcess/QuickLookThumbnailLoader.mm:
- Drive-by fix of soft-linking header include order.
* WebKit.xcodeproj/project.pbxproj:
- Add CocoaImage.h to the project.
* WebProcess/InjectedBundle/API/Cocoa/WKWebProcessPlugInNodeHandle.mm:
(-[WKWebProcessPlugInNodeHandle renderedImageWithOptions:]):
(-[WKWebProcessPlugInNodeHandle renderedImageWithOptions:width:]):
- Combine separate platform-specific methods into one method.
2020-04-24 David Kilzer <ddkilzer@apple.com>
IPC::Decoder should use create() pattern
<https://webkit.org/b/210949>
<rdar://problem/62144409>
Reviewed by Geoffrey Garen.
* Platform/IPC/Decoder.cpp:
(IPC::Decoder::create): Add implementation. Returns nullptr if
Decoder constructor returns an invalid object.
(IPC::Decoder::Decoder): Mark invalid if m_buffer is not 64-bit
aligned.
(IPC::Decoder::unwrapForTesting): Switch to Decoder::create().
* Platform/IPC/Decoder.h:
(IPC::Decoder::create): Add declaration.
(IPC::Decoder::Decoder): Make explicit. (Can't be made private
since we use std::unique_ptr<Decoder>.)
* Platform/IPC/cocoa/ConnectionCocoa.mm:
(IPC::createMessageDecoder): Switch to Decoder::create().
2020-04-24 Tim Horton <timothy_horton@apple.com>
iPad: "Pocket City" interaction does not work with trackpad
https://bugs.webkit.org/show_bug.cgi?id=210985
<rdar://problem/62273077>
Reviewed by Wenson Hsieh.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView shouldUseMouseGestureRecognizer]):
Add another app to the list who fall back to touch event synthesis until rebuilt.
2020-04-24 Alex Christensen <achristensen@webkit.org>
SPI clients using fastServerTrustEvaluationEnabled need SPI to inform them of modern TLS negotiation
https://bugs.webkit.org/show_bug.cgi?id=210533
Reviewed by Brady Eidson.
* NetworkProcess/NetworkDataTask.h:
(WebKit::NetworkDataTaskClient::didNegotiateModernTLS):
* NetworkProcess/NetworkLoad.cpp:
(WebKit::NetworkLoad::didNegotiateModernTLS):
* NetworkProcess/NetworkLoad.h:
* NetworkProcess/cocoa/NetworkDataTaskCocoa.h:
* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::NetworkDataTaskCocoa::didNegotiateModernTLS):
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(-[WKNetworkSessionDelegate URLSession:task:didReceiveChallenge:completionHandler:]):
* UIProcess/API/APINavigationClient.h:
(API::NavigationClient::didNegotiateModernTLS):
* UIProcess/API/Cocoa/WKNavigationDelegatePrivate.h:
* UIProcess/Cocoa/NavigationState.h:
* UIProcess/Cocoa/NavigationState.mm:
(WebKit::NavigationState::setNavigationDelegate):
(WebKit::NavigationState::NavigationClient::didNegotiateModernTLS):
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::didNegotiateModernTLS):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/Network/NetworkProcessProxy.messages.in:
* UIProcess/WebPageProxy.cpp:
* UIProcess/WebPageProxy.h:
2020-04-24 Brian Burg <bburg@apple.com>
Web Automation: timeout underneath Automation.evaluateJavaScriptFunction in Selenium test frame_switching_tests.py::testShouldNotBeAbleToDoAnythingTheFrameIsDeletedFromUnderUs[Safari]
https://bugs.webkit.org/show_bug.cgi?id=210162
<rdar://problem/60561009>
Reviewed by Devin Rousso.
When an iframe is detached from the DOM, it is no longer exposed as a browsing context
and it's not possible to Evaluate JavaScript or perform other commands with it. This
patch adds frame lifecycle monitoring so that pending script evaluations are cancelled
with FrameNotFound as soon as the iframe is detached from the DOM. This change also avoids
running more commands with the frame if it's detached from its DOMWindow and ready to be GC'd.
* Sources.txt:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/Automation/WebAutomationDOMWindowObserver.h: Added.
* WebProcess/Automation/WebAutomationDOMWindowObserver.cpp: Added.
(WebKit::WebAutomationDOMWindowObserver::WebAutomationDOMWindowObserver):
(WebKit::WebAutomationDOMWindowObserver::~WebAutomationDOMWindowObserver):
(WebKit::WebAutomationDOMWindowObserver::frame const):
(WebKit::WebAutomationDOMWindowObserver::willDestroyGlobalObjectInCachedFrame):
(WebKit::WebAutomationDOMWindowObserver::willDestroyGlobalObjectInFrame):
(WebKit::WebAutomationDOMWindowObserver::willDetachGlobalObjectFromFrame):
This class is a stripped-down copy of DOMWindowExtension, which is the only other
client of DOMWindow::Observer interface. When a frame is detached, destroyed, or
navigates (global object detached), then call the callback and unregister.
* WebProcess/Automation/WebAutomationSessionProxy.h:
* WebProcess/Automation/WebAutomationSessionProxy.cpp:
(WebKit::WebAutomationSessionProxy::~WebAutomationSessionProxy):
(WebKit::WebAutomationSessionProxy::didClearWindowObjectForFrame):
(WebKit::WebAutomationSessionProxy::willDestroyGlobalObjectForFrame):
(WebKit::WebAutomationSessionProxy::evaluateJavaScriptFunction):
(WebKit::WebAutomationSessionProxy::ensureObserverForFrame): For non-main frames,
ensure we add a frame observer if we are about to evaluate JavaScript upon the frame.
This acts as a watchdog in case the frame becomes detached while waiting for pending
JS evaluations. When a frame is detached, the JS evaluation may or may not complete.
(WebKit::WebAutomationSessionProxy::selectOptionElement): Fix hilarious typo.
* WebProcess/GPU/graphics/cocoa/ImageBufferShareableIOSurfaceBackend.cpp:
* WebProcess/GPU/media/WebMediaStrategy.cpp:
(WebKit::WebMediaStrategy::clearNowPlayingInfo):
(WebKit::WebMediaStrategy::setNowPlayingInfo):
Adding a new file seems to have exposed a few missing includes and namespace qualifiers.
This is due to unified sources chunking.
2020-04-24 Alex Christensen <achristensen@webkit.org>
Use sendWithAsyncReply for ShareSheet related messages
https://bugs.webkit.org/show_bug.cgi?id=210828
<rdar://problem/61800730>
Reviewed by Brent Fulgham.
* Platform/IPC/MessageSender.cpp:
We need to call addAsyncReplyHandler before sendMessage in case this is the first async message from this process.
Otherwise the reply from the first message is dropped sometimes.
* Shared/ShareSheetCallbackID.h: Removed.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::showShareSheet):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::showShareSheet):
(WebKit::nextShareSheetCallbackID): Deleted.
(WebKit::WebPage::didCompleteShareSheet): Deleted.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-04-24 Youenn Fablet <youenn@apple.com> and Luming Yin <luming_yin@apple.com>
Call STDynamicActivityAttributionPublisher in the WebProcess
https://bugs.webkit.org/show_bug.cgi?id=210772
<rdar://problem/62075201>
Reviewed by Geoffrey Garen.
Call STDynamicActivityAttributionPublisher in WebProcess to make use of the newly added plist entry.
Use of a sandbox extension to protect this call.
* Configurations/WebKit.xcconfig:
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* Scripts/process-entitlements.sh:
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/EntryPoint/Cocoa/XPCService/WebContentService/Info-iOS.plist:
* WebProcess/EntryPoint/Cocoa/XPCService/WebContentService/InfoPlist.strings: Added.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-04-24 Simon Fraser <simon.fraser@apple.com>
Move some post-renderingUpdate code into WebCore
https://bugs.webkit.org/show_bug.cgi?id=210952
Reviewed by Antti Koivisto.
Factor some code called by the various DrawingArea subclasses into Page::finalizeRenderingUpdate(),
with some flags to control behavior that differs between drawing areas.
ScrollingCoordinator::commitTreeStateIfNeeded() is a no-op for RemoteScrollingCoordinator so
it's fine to always call it.
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.mm:
(WebKit::RemoteLayerTreeDrawingArea::updateRendering):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::finalizeRenderingUpdate):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::updateRendering):
2020-04-24 Chris Dumez <cdumez@apple.com>
[iOS] Stop using legacy BKSApplicationStateMonitor
https://bugs.webkit.org/show_bug.cgi?id=210945
Reviewed by Tim Horton.
Stop using legacy BKSApplicationStateMonitor and use RunningBoard API instead.
* Configurations/WebKit.xcconfig:
Stop linking against ApplicationServices when using iOS 14 SDK now that we are
fully transitioned to RunningBoard.
* Platform/spi/ios/RunningBoardServicesSPI.h:
* UIProcess/ApplicationStateTracker.h:
* UIProcess/ApplicationStateTracker.mm:
(WebKit::ApplicationStateTracker::ApplicationStateTracker):
(WebKit::isApplicationForeground):
(WebKit::ApplicationStateTracker::~ApplicationStateTracker):
(WebKit::isBackgroundState): Deleted.
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::isApplicationVisible):
2020-04-23 Simon Fraser <simon.fraser@apple.com>
Bounce displayWasRefreshed() via EventDispatcher
https://bugs.webkit.org/show_bug.cgi?id=208778
Reviewed by Antti Koivisto.
Allow the WebContent process to be notified of display refresh off the main thread, so other
threads like the scrolling thread can respond without being blocked. Achieve this
by having EventDispatcher receive the IPC message, then bounce it to the main thread.
* UIProcess/mac/DisplayLink.cpp:
(WebKit::DisplayLink::displayLinkCallback):
* WebProcess/WebPage/EventDispatcher.cpp:
(WebKit::EventDispatcher::displayWasRefreshed):
* WebProcess/WebPage/EventDispatcher.h:
* WebProcess/WebPage/EventDispatcher.messages.in:
* WebProcess/WebProcess.h:
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::displayWasRefreshed):
2020-04-23 David Kilzer <ddkilzer@apple.com>
Clean up QuickLookThumbnailLoader
<https://webkit.org/b/210814>
Reviewed by Darin Adler.
The following items are cleaned up:
- Extract `using PlatformImage` into QuickLookThumbnailLoader.h,
rename to `CocoaImage` and use to get rid of duplicate
code.
- Change `id` to `instancetype` for -init methods.
- Add `atomic` keyword to @property definitions that were using
it as the default. (Use of atomic properties is rare in
WebKit, so being explicit avoids a scenario where it looks
like `nonatomic` was left off by accident.)
- Change @property definitions to `readonly` that are never
written to outside of QuickLookThumbnailLoader.mm.
- Delete unused @property definitions.
- Change method declarations into read-only @property
definitions.
- Re-declare atomic read-only @property definitions in
QuickLookThumbnailLoader.h as read-write definitions in
QuickLookThumbnailLoader.mm if they are written to.
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::convertPlatformImageToBitmap):
* UIProcess/QuickLookThumbnailLoader.h:
- Rename qlThumbnailGenerationQueue @property to just `queue`.
- Remove `contentType` @property. It is not used anywhere.
This also fixes a theoretical leak found by the clang static
analyzer.
- Remove `shouldWrite` @property. It is only used within
QuickLookThumbnailLoader.mm.
- Change `identifier` and `thumbnail` to @property declarations.
* UIProcess/QuickLookThumbnailLoader.mm:
- Change WKQLThumbnailLoadOperation._identifier type from
NSMutableString to NSString. There was no reason for it to
be mutable.
(-[WKQLThumbnailQueueManager init]):
(-[WKQLThumbnailQueueManager dealloc]):
- Release `_queue` to fix theoretical leak found by the clang
static analyzer.
(-[WKQLThumbnailLoadOperation initWithAttachment:identifier:]):
(-[WKQLThumbnailLoadOperation initWithURL:identifier:]):
(-[WKQLThumbnailLoadOperation start]):
- Rename `req` to `request`.
- Change separate #if macros to #if/#else since only one version
of this code can be used at a time.
(-[WKQLThumbnailLoadOperation thumbnail]):
- Use CocoaImage to use one copy of the method.
2020-04-23 Megan Gardner <megan_gardner@apple.com>
Long pressing attachments in Notes does not activate Context Menu.
https://bugs.webkit.org/show_bug.cgi?id=210936
<rdar://problem/61171576>
Reviewed by Tim Horton.
Not having the ID available for notes makes their code to create the context
menu and its items fail. We should probably be passing this information on
for any element if we have it.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::selectionPositionInformation):
2020-04-23 Brent Fulgham <bfulgham@apple.com>
Allow "kern.osversion" sysctl read
https://bugs.webkit.org/show_bug.cgi?id=210929
<rdar://problem/62256013>
Reviewed by Per Arne Vollan.
Libdispatch checks "kern.osversion" as part of the code path we use for launching our XPC services.
We already allow it for some services (Network Process), and should be doing so for our other
helper processes.
* GPUProcess/mac/com.apple.WebKit.GPUProcess.sb.in:
* Resources/SandboxProfiles/ios/com.apple.WebKit.GPU.sb:
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* WebProcess/com.apple.WebProcess.sb.in:
2020-04-23 Peng Liu <peng.liu6@apple.com>
Remove the WebKit.plist for Feature Flags
https://bugs.webkit.org/show_bug.cgi?id=210534
Reviewed by Simon Fraser.
* FeatureFlags/WebKit.plist: Removed.
* WebKit.xcodeproj/project.pbxproj:
2020-04-23 Chris Dumez <cdumez@apple.com>
[iOS] Crash on RunningBoard process assertion invalidation
https://bugs.webkit.org/show_bug.cgi?id=210873
<rdar://problem/62194917>
Unreviewed, nil out the observer only after we've removed it to fix crashes on the bots.
* UIProcess/ios/ProcessAssertionIOS.mm:
(WebKit::ProcessAssertion::~ProcessAssertion):
2020-04-23 Kate Cheney <katherine_cheney@apple.com>
All ITP database tables should reference the ObservedDomains table on DELETE CASCADE
https://bugs.webkit.org/show_bug.cgi?id=210874
<rdar://problem/62209438>
Reviewed by Brady Eidson.
This makes it easier to delete an ITP entry from the entire database.
An entry shouldn't really exist outside of ObservedDomains anyways,
because that table holds all data about the domain, so this is a good
change to make regardless of the deletion use case.
(WebKit::needsNewCreateTableSchema):
Function to check for old CREATE TABLE queries.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::ResourceLoadStatisticsDatabaseStore):
(WebKit::ResourceLoadStatisticsDatabaseStore::enableForeignKeys):
Must enable foreign keys to use DELETE CASCADE.
(WebKit::ResourceLoadStatisticsDatabaseStore::isMigrationNecessary):
(WebKit::ResourceLoadStatisticsDatabaseStore::migrateDataToNewTablesIfNecessary):
Since this is a schema change that is executed when a table is created, we
need to check if the current schema is out of date. If so, we can
create new tables using the proper reference and migrate the data.
(WebKit::ResourceLoadStatisticsDatabaseStore::openAndDropOldDatabaseIfNecessary):
We only need to migrate the data if there were no issues with the
schema (otherwise the entire database will already have been created
with the correct CREATE TABLE queries).
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.h:
2020-04-23 Don Olmstead <don.olmstead@sony.com>
[CMake] CMAKE_BINARY_DIR should always be a PRIVATE include directory
https://bugs.webkit.org/show_bug.cgi?id=196717
Reviewed by Michael Catanzaro.
Include CMAKE_BINARY_DIR either directly or through WebKit_PRIVATE_INCLUDE_DIRECTORIES.
* PlatformGTK.cmake:
* PlatformWPE.cmake:
2020-04-23 Adrian Perez de Castro <aperez@igalia.com>
Regression after r260359 ([GTK][WPE] lowWatermarkPages() in MemoryPressureMonitor.cpp only searches the "low" value inside the first "Node" section)
https://bugs.webkit.org/show_bug.cgi?id=210916
Reviewed by Carlos Alberto Lopez Perez.
Switch over to using a FileHandle type based on std::unique_ptr with a custom deleter and an
utility function for opening files. This makes easier to follow the logic inside the polling
loop for opening files and retrying when needed. This also fixes exiting the thread when the
systemMemoryUsedAsPercentage() function would return -1, to loop restart instead to keep
trying.
Thanks to Pablo Saavedra for his help in making this patch.
No new tests needed.
* UIProcess/linux/MemoryPressureMonitor.cpp:
(WebKit::FileHandleDeleter::operator()): Add deleter to use with std::unique_ptr<>.
(WebKit::tryOpeningForUnbufferedReading): Add utility function to open a file handle if
needed and configuring its buffering upon opening.
(WebKit::MemoryPressureMonitor::start): Use FileHandle to ensure that handles are always
closed properly, and fix logic retry opening files on failure.
2020-04-23 Nikos Mouchtaris <nmouchtaris@apple.com>
Soft link QuickLookThumbnailing framework
https://bugs.webkit.org/show_bug.cgi?id=210894
Reviewed by Megan Gardner.
Soft link QuickLookThumbnailing framework to solve build error.
No new tests. Unnecessary.
* Configurations/WebKit.xcconfig:
* UIProcess/QuickLookThumbnailLoader.mm:
(-[WKQLThumbnailLoadOperation initWithURL:identifier:]):
(-[WKQLThumbnailLoadOperation start]):
* UIProcess/QuickLookThumbnailingSPI.h: Added.
* UIProcess/QuickLookThumbnailingSoftLink.h: Added.
* UIProcess/QuickLookThumbnailingSoftLink.mm: Added.
* WebKit.xcodeproj/project.pbxproj:
2020-04-23 Chris Dumez <cdumez@apple.com>
[iOS] Port MediaPlayback process assertion to RunningBoard
https://bugs.webkit.org/show_bug.cgi?id=210212
<rdar://problem/61476951>
Reviewed by Geoff Garen.
Port MediaPlayback process assertion to RunningBoard instead of the legacy BKSProcessAssertion.
We can now #ifdef out the legacy BKSProcessAssertion on recent iOS builds.
* UIProcess/ProcessAssertion.h:
* UIProcess/ios/ProcessAssertionIOS.mm:
(WebKit::runningBoardNameForAssertionType):
(WebKit::ProcessAssertion::ProcessAssertion):
(WebKit::ProcessAssertion::~ProcessAssertion):
2020-04-23 John Frankish <john.frankish@outlook.com>
[GTK] Crash in cairo_surface_mark_dirty_rectangle() in accelerated compositing mode under X11
https://bugs.webkit.org/show_bug.cgi?id=210636
Reviewed by Carlos Garcia Campos.
When cairo is configured to use xcb instead of xlib, it might use an image surface attached to the xlib one as
snapshot. In that case a flush is needed to detach that snapshot after we have drawn the surface in the
context.
* UIProcess/gtk/AcceleratedBackingStoreX11.cpp:
(WebKit::AcceleratedBackingStoreX11::paint): Call cairo_surface_flush() after drawing.
2020-04-23 Chris Dumez <cdumez@apple.com>
Unreviewed, reverting r260133.
We can use RunningBoard foreground assertion again now that
<rdar://problem/61830390> has been fixed
Reverted changeset:
"REGRESSION (r259610): WebGL does not work at all on iOS (was:
Google Maps tiles turn black after initial load)"
https://trac.webkit.org/changeset/260133
2020-04-22 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed, reverting r260553.
Broke iOS and Catalina builds.
Reverted changeset:
"Soft link QuickLookThumbnailing framework"
https://bugs.webkit.org/show_bug.cgi?id=210894
https://trac.webkit.org/changeset/260553
2020-04-22 Darin Adler <darin@apple.com>
[Cocoa] Build with UChar as char16_t even in builds that use Apple's internal SDK
https://bugs.webkit.org/show_bug.cgi?id=210845
Reviewed by Anders Carlsson.
* Configurations/BaseTarget.xcconfig: Move ICU-configuring macros to Platform.h.
* Platform/unix/EnvironmentUtilities.h: Removed unneeded includes.
This sidesteps a problem with Apple's internal InstallAPI tool that happens
if StringView.h is included and Platform.h is not included first.
* UIProcess/API/cpp/WKRetainPtr.h: Removed unneeded includes, and used
#pragma once.
* UIProcess/Cocoa/PreferenceObserver.h: Moved most things out of the header,
since they don't need to be there.
* UIProcess/Cocoa/PreferenceObserver.mm: Moved them here including using a
class extension for the data member of WKPreferenceObserver.
* UIProcess/ios/fullscreen/FullscreenTouchSecheuristicParameters.h:
Removed unneeded WTF prefix.
2020-04-22 Nikos Mouchtaris <nmouchtaris@apple.com>
Soft link QuickLookThumbnailing framework
https://bugs.webkit.org/show_bug.cgi?id=210894
Reviewed by Tim Horton.
Soft link QuickLookThumbnailing framework to solve build error.
No new tests. Unnecessary.
* Configurations/WebKit.xcconfig:
* UIProcess/QuickLookThumbnailLoader.mm:
(-[WKQLThumbnailLoadOperation initWithURL:identifier:]):
(-[WKQLThumbnailLoadOperation start]):
* UIProcess/QuickLookThumbnailingSPI.h: Added.
* UIProcess/QuickLookThumbnailingSoftLink.h: Added.
* UIProcess/QuickLookThumbnailingSoftLink.mm: Added.
* WebKit.xcodeproj/project.pbxproj:
2020-04-22 Daniel Bates <dabates@apple.com>
Support toggling debug overlay for touch action region and editable element region independent from non-fast scrollable region
https://bugs.webkit.org/show_bug.cgi?id=210774
Reviewed by Dean Jackson.
Expose two new enumerators to toggle touch action region and editable element region
overlay painting.
* UIProcess/API/C/WKPreferencesRef.h:
* UIProcess/API/Cocoa/WKPreferencesPrivate.h:
2020-04-22 Chris Dumez <cdumez@apple.com>
[iOS] Expose -_webView:willGoToBackForwardListItem:inPageCache:
https://bugs.webkit.org/show_bug.cgi?id=210878
<rdar://problem/62202276>
Reviewed by Alex Christensen.
Expose -_webView:willGoToBackForwardListItem:inPageCache: on iOS.
* UIProcess/API/Cocoa/WKNavigationDelegatePrivate.h:
* UIProcess/Cocoa/NavigationState.h:
* UIProcess/Cocoa/NavigationState.mm:
(WebKit::NavigationState::setNavigationDelegate):
(WebKit::NavigationState::NavigationClient::willGoToBackForwardListItem):
2020-04-22 Chris Dumez <cdumez@apple.com>
[iOS] Crash on RunningBoard process assertion invalidation
https://bugs.webkit.org/show_bug.cgi?id=210873
<rdar://problem/62194917>
Reviewed by Darin Adler.
* UIProcess/ios/ProcessAssertionIOS.mm:
(-[WKRBSAssertionDelegate assertion:didInvalidateWithError:]):
Capture a weak pointer to self and make sure we only access the invalidation handler on
the main thread if |self| is still alive.
(WebKit::ProcessAssertion::~ProcessAssertion):
Null out the WKRBSAssertionDelegate's observer in the ProcessAssertion destructor, to
make sure processAssertionWasInvalidated() cannot get called after the ProcessAssertion
has been destroyed.
2020-04-22 David Kilzer <ddkilzer@apple.com>
IPC::decodeSharedBuffer() should check the return value of SharedMemory::map()
<https://webkit.org/b/210844>
<rdar://problem/60773120>
Reviewed by Geoffrey Garen.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::decodeSharedBuffer):
- Add nullptr check and early return.
2020-04-22 Brent Fulgham <bfulgham@apple.com>
Switch from debug ASSERT to RELEASE_ASSERT in toNPJSObject
https://bugs.webkit.org/show_bug.cgi?id=210823
<rdar://problem/61774056>
Reviewed by David Kilzer.
* WebProcess/Plugins/Netscape/NPJSObject.h:
(WebKit::NPJSObject::toNPJSObject):
2020-04-22 Youenn Fablet <youenn@apple.com>
Simplify SWServerWorker::whenActivated logic
https://bugs.webkit.org/show_bug.cgi?id=210795
Reviewed by Alex Christensen.
Improve logging and update code according whenActivated implementation.
Add an early check for timeouts so that we return earlier in that case.
* NetworkProcess/ServiceWorker/WebSWServerConnection.cpp:
(WebKit::WebSWServerConnection::createFetchTask):
(WebKit::WebSWServerConnection::startFetch):
2020-04-21 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed, reverting r260478.
Caused TestWebKitAPI.WKAttachmentTests crashes on Catalina
Reverted changeset:
"Clean up QuickLookThumbnailLoader"
https://bugs.webkit.org/show_bug.cgi?id=210814
https://trac.webkit.org/changeset/260478
2020-04-21 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed, reverting r260410.
Caused crashes in Safari
Reverted changeset:
"SPI clients using fastServerTrustEvaluationEnabled need SPI
to inform them of modern TLS negotiation"
https://bugs.webkit.org/show_bug.cgi?id=210533
https://trac.webkit.org/changeset/260410
2020-04-21 Ryan Haddad <ryanhaddad@apple.com>
Unreviewed, reverting r260475.
New assert caused layout tests to exit early on iOS and macOS
Reverted changeset:
"REGRESSION(r260410) Frequent Network process crashes"
https://bugs.webkit.org/show_bug.cgi?id=210825
https://trac.webkit.org/changeset/260475
2020-04-21 David Kilzer <ddkilzer@apple.com>
IPC::Encoder::reserve() can be private
<https://webkit.org/b/210833>
<rdar://problem/62124193>
Reviewed by Darin Adler.
* Platform/IPC/Encoder.h:
(IPC::Encoder::reserve):
- Move from public to private.
2020-04-21 Stephan Szabo <stephan.szabo@sony.com>
[PlayStation] Fix GPU Process build for VIDEO_PRESENTATION_MODE changes
https://bugs.webkit.org/show_bug.cgi?id=210836
Reviewed by Darin Adler.
Build fix, no new tests.
* GPUProcess/media/playstation/RemoteMediaPlayerProxyPlayStation.cpp:
2020-04-21 Peng Liu <peng.liu6@apple.com>
Fix MACCATALYST build failures
https://bugs.webkit.org/show_bug.cgi?id=210815
Reviewed by Tim Horton.
* Configurations/FeatureDefines.xcconfig:
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _isShowingVideoPictureInPicture]):
(-[WKWebView _mayAutomaticallyShowVideoPictureInPicture]):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::applicationDidBecomeActive):
2020-04-19 Darin Adler <darin@apple.com>
[Cocoa] Use createNSArray in many more places that build NSArray objects from C++ collections
https://bugs.webkit.org/show_bug.cgi?id=210702
Reviewed by Alex Christensen.
* Platform/cocoa/PaymentAuthorizationPresenter.mm:
(WebKit::toNSErrors): Use createNSArray and return a RetainPtr to cut down
on autorelease.
(WebKit::toPKShippingMethods): Ditto.
(WebKit::PaymentAuthorizationPresenter::completePaymentSession): Updated for RetainPtr.
(WebKit::PaymentAuthorizationPresenter::completeShippingContactSelection): Ditto.
* Shared/ApplePay/cocoa/WebPaymentCoordinatorProxyCocoa.mm:
(WebKit::WebPaymentCoordinatorProxy::platformPaymentRequest): Use createNSArray.
Also removed code to support platforms without -setThumbnailURLs: since we don't
have those any more.
* Shared/RemoteLayerTree/RemoteLayerTreePropertyApplier.mm:
(WebKit::RemoteLayerTreePropertyApplier::updateChildren): Use createNSArray.
* UIProcess/API/Cocoa/WKHTTPCookieStore.mm:
(coreCookiesToNSCookies); Ditto.
* UIProcess/API/Cocoa/WKProcessPool.mm:
(+[WKProcessPool _allProcessPoolsForTesting]): Ditto.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _startTextManipulationsWithConfiguration:completion:]): Cut down on
repeated code by adding a lambda to create a _WKTextManipulationItem and by using
createNSArray.
(wkTextManipulationErrors): Use createNSArray. Also tighten up the code that maps
WebCore::TextManipulationController::ManipulationFailureType to the
_WKTextManipulationItemError code.
* UIProcess/API/Cocoa/WKWebpagePreferences.mm:
(-[WKWebpagePreferences _customHeaderFields]): Use createNSArray.
* UIProcess/API/Cocoa/WKWebsiteDataRecord.mm:
(-[WKWebsiteDataRecord _originsStrings]): Ditto.
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
(-[WKWebsiteDataStore _getResourceLoadStatisticsDataSummary:]): Ditto.
* UIProcess/API/Cocoa/_WKDownload.mm:
(-[_WKDownload redirectChain]): Ditto.
* UIProcess/API/Cocoa/_WKFrameTreeNode.mm:
(-[_WKFrameTreeNode childFrames]): Ditto.
* UIProcess/API/Cocoa/_WKProcessPoolConfiguration.mm:
(-[_WKProcessPoolConfiguration additionalReadAccessAllowedURLs]): Ditto.
* UIProcess/API/Cocoa/_WKResourceLoadStatisticsThirdParty.mm:
(-[_WKResourceLoadStatisticsThirdParty underFirstParties]): Ditto.
* UIProcess/API/ios/WKWebViewIOS.mm:
(-[WKWebView _accessibilityRetrieveRectsAtSelectionOffset:withText:completionHandler:]): Ditto.
* UIProcess/Cocoa/NavigationState.mm:
(WebKit::NavigationState::NavigationClient::didChangeBackForwardList): Ditto.
* UIProcess/RemoteLayerTree/RemoteLayerTreeScrollingPerformanceData.mm:
(WebKit::RemoteLayerTreeScrollingPerformanceData::data): Ditto.
* UIProcess/ios/DragDropInteractionState.mm:
(WebKit::createTargetedDragPreview): Ditto.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _updateTapHighlight]): Removed uneeded RetainPtr local variable.
Use createNSArray. Refactored to make the "all rectilinear" case more parallel
to the general case.
(-[WKContentView webSelectionRectsForSelectionRects:]): Use createNSArray.
(wkTextSelectionRects): Ditto.
(createTargetedPreview): Ditto.
(-[WKContentView _presentationRectsForPreviewItemController:]): Ditto.
* UIProcess/mac/WKTextFinderClient.mm:
(-[WKTextFinderClient didFindStringMatchesWithRects:didWrapAround:]): Use createNSArray.
* UIProcess/mac/WKTextInputWindowController.mm:
(-[WKTextInputView validAttributesForMarkedText]): Use +[NSArray arrayByAddingObject:].
* WebProcess/InjectedBundle/API/Cocoa/WKWebProcessPlugInFrame.mm:
(collectIcons): Use createNSArray.
* WebProcess/InjectedBundle/API/mac/WKWebProcessPlugInBrowserContextController.mm:
(-[WKWebProcessPlugInBrowserContextController _setFormDelegate:]): Marked class here
as final. Cut down on local variables a bit. Use createNSArray.
* WebProcess/WebPage/RemoteLayerTree/PlatformCAAnimationRemote.mm:
(WebKit::addAnimationToLayer): Removed unnecessary use of RetainPtr. Use createNSArray.
* WebProcess/WebPage/WebBackForwardListProxy.cpp:
(WebKit::WebBackForwardListProxy::WebBackForwardListProxy): Addd a FIXME about
a problem I noticed, unrelated to the rest of the changes.
* WebProcess/WebPage/mac/WKAccessibilityWebPageObjectMac.mm:
(-[WKAccessibilityWebPageObject accessibilityParameterizedAttributeNames]): Use createNSArray.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::registerWithStateDumper): Ditto.
(WebKit::activePagesOrigins): Ditto.
2020-04-21 David Kilzer <ddkilzer@apple.com>
Clean up QuickLookThumbnailLoader
<https://webkit.org/b/210814>
Reviewed by Darin Adler.
The following items are cleaned up:
- Extract `using PlatformImage` into QuickLookThumbnailLoader.h,
rename to `CocoaImage` and use to get rid of duplicate
code.
- Change `id` to `instancetype` for -init methods.
- Add `atomic` keyword to @property definitions that were using
it as the default. (Use of atomic properties is rare in
WebKit, so being explicit avoids a scenario where it looks
like `nonatomic` was left off by accident.)
- Change @property definitions to `readonly` that are never
written to outside of QuickLookThumbnailLoader.mm.
- Delete unused @property definitions.
- Change method declarations into read-only @property
definitions.
- Re-declare atomic read-only @property definitions in
QuickLookThumbnailLoader.h as read-write definitions in
QuickLookThumbnailLoader.mm if they are written to.
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::convertPlatformImageToBitmap):
* UIProcess/QuickLookThumbnailLoader.h:
- Rename qlThumbnailGenerationQueue @property to just `queue`.
- Remove `contentType` @property. It is not used anywhere.
This also fixes a theoretical leak found by the clang static
analyzer.
- Remove `shouldWrite` @property. It is only used within
QuickLookThumbnailLoader.mm.
- Change `identifier` and `thumbnail` to @property declarations.
* UIProcess/QuickLookThumbnailLoader.mm:
- Change WKQLThumbnailLoadOperation._identifier type from
NSMutableString to NSString. There was no reason for it to
be mutable.
(-[WKQLThumbnailQueueManager init]):
(-[WKQLThumbnailQueueManager dealloc]):
- Release `_queue` to fix theoretical leak found by the clang
static analyzer.
(-[WKQLThumbnailLoadOperation initWithAttachment:identifier:]):
(-[WKQLThumbnailLoadOperation initWithURL:identifier:]):
(-[WKQLThumbnailLoadOperation start]):
- Rename `req` to `request` and use RetainPtr<>.
- Change separate #if macros to #if/#else since only one version
of this code can be used at a time.
(-[WKQLThumbnailLoadOperation thumbnail]):
- Use CocoaImage to use one copy of the method.
2020-04-21 David Kilzer <ddkilzer@apple.com>
Fix extra/missing #undef statements
<https://webkit.org/b/210818>
<rdar://problem/61973329>
Reviewed by Chris Dumez.
* UIProcess/Network/NetworkProcessProxy.cpp:
- Remove #undef MESSAGE_CHECK_URL. It is no longer used.
* UIProcess/WebPageProxy.cpp:
- Add missing #undef MESSAGE_CHECK_COMPLETION.
- Sort #undef in reverse order of #define.
* UIProcess/mac/WebPageProxyMac.mm:
- Add missing #undef MESSAGE_CHECK_WITH_RETURN_VALUE.
- Sort #undef in reverse order of #define.
2020-04-21 Chris Dumez <cdumez@apple.com>
REGRESSION(r260410) Frequent Network process crashes
https://bugs.webkit.org/show_bug.cgi?id=210825
Reviewed by Alex Christensen.
The webPageProxyID in the NetworkProcessProxy::DidNegotiateModernTLS() IPC was invalid because
NetworkProcessProxy::preconnectTo() was failing to initialize NetworkLoadParameters::webPageProxyID.
This patch addresses the issue.
No new tests, added new assertion that would be tripped by existing tests.
* NetworkProcess/NetworkLoad.cpp:
(WebKit::NetworkLoad::NetworkLoad):
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::preconnectTo):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::preconnectTo):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::preconnectTo):
2020-04-21 Jer Noble <jer.noble@apple.com>
PiP buttons shows up in element fullscreen when PiP is disabled in WKWebViewConfiguration
https://bugs.webkit.org/show_bug.cgi?id=210813
Reviewed by Eric Carlson.
Not currently testable; entering fullscreen never completes in the iOS TestWebKitAPI app
because that process is not a UI Application.
* UIProcess/ios/fullscreen/WKFullScreenViewController.mm:
(-[WKFullScreenViewController videoControlsManagerDidChange]):
2020-04-21 Brent Fulgham <bfulgham@apple.com>
[Cocoa] Don't bother compiling UpdateInfo since it's not used
https://bugs.webkit.org/show_bug.cgi?id=210819
<rdar://problem/61774766>
Reviewed by Darin Adler.
Since we don't use the UpdateInfo class in Cocoa builds, let's not
bother building it.
* Shared/UpdateInfo.cpp:
* Shared/UpdateInfo.h:
* UIProcess/DrawingAreaProxy.h:
2020-04-21 Timothy Horton <timothy_horton@apple.com>
One more build fix after r260407
* Configurations/WebKit.xcconfig:
2020-04-21 Chris Dumez <cdumez@apple.com>
[iOS] Provide better names for ProcessThrottler process assertions
https://bugs.webkit.org/show_bug.cgi?id=210808
Reviewed by Geoffrey Garen.
Provide better names for ProcessThrottler process assertions, to facilitate debugging.
* UIProcess/GPU/GPUProcessProxy.h:
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/ProcessAssertion.cpp:
(WebKit::ProcessAssertion::ProcessAssertion):
* UIProcess/ProcessAssertion.h:
* UIProcess/ProcessThrottler.cpp:
(WebKit::ProcessThrottler::assertionName const):
(WebKit::ProcessThrottler::setAssertionType):
* UIProcess/ProcessThrottler.h:
* UIProcess/ProcessThrottlerClient.h:
* UIProcess/WebProcessProxy.h:
* UIProcess/ios/ProcessAssertionIOS.mm:
(WebKit::ProcessAssertion::ProcessAssertion):
(WebKit::ProcessAndUIAssertion::ProcessAndUIAssertion):
2020-04-21 Daniel Bates <dabates@apple.com>
[iOS] -_didFinishTextInteractionInTextInputContext should only zoom to reveal focused element if it changed
https://bugs.webkit.org/show_bug.cgi?id=210697
<rdar://problem/60997530>
Reviewed by Wenson Hsieh.
Add a new state variable to use to track whether -_focusTextInputContext invoked during a
text interaction actually focused an element. Then in -_didFinishTextInteractionInTextInputContext
condition marking the page to reveal the focused element on this state.
* UIProcess/API/ios/WKWebViewPrivateForTestingIOS.h:
* UIProcess/API/ios/WKWebViewTestingIOS.mm:
(-[WKWebView _willBeginTextInteractionInTextInputContext:]): Turn around and call the same
function on the content view.
(-[WKWebView _didFinishTextInteractionInTextInputContext:]): Ditto.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView cleanUpInteraction]): Reset state.
(-[WKContentView _didCommitLoadForMainFrame]): Ditto.
(-[WKContentView _focusTextInputContext:placeCaretAt:completionHandler:]): Update state to
indicate whether the focusd element did change or not. I use bitwise ORing instead of strict
assignment to update it because this function can be called multiple times by an internal
Apple client during a text interaction (e.g. <rdar://problem/59430806>). I thought about
making this a FIXME, but this function is otherwise capable of handling multiple invocations
so I didn't.
(-[WKContentView _willBeginTextInteractionInTextInputContext:]): Reset added state.
(-[WKContentView _didFinishTextInteractionInTextInputContext:]): Check if state was set to
indicate that a focus change happened. If so, do what we do now + reset state. Otherwise,
do everything we do now except for marking the page as needing to reveal the focused element
on the next editor state update.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::focusTextInputContextAndPlaceCaret): Check that the target is focusable
before calling FocusController::setFocusedElement() because it ultimately calls Document::setFocusedElement()
that can return true for a non-focusable element. I added a comment in Document.h to explain
this subtlety. I also decided not to check the return value of FocusController::setFocusedElement().
Instead I am checking the more important condition that the focused element is the target
element. Passing true to visiblePositionInFocusedNodeForPoint() will ultimately assert this
to be true. Note that setFocusedElement() may not have done anything as m_focusedElement
could have already been equal to the target element. This is OK and I still update the
caret position.
2020-04-21 Timothy Hatcher <timothy@apple.com>
Reset m_userScriptsNotified when web process crashes.
https://bugs.webkit.org/show_bug.cgi?id=210807
Reviewed by Alex Christensen.
* UIProcess/WebPageProxy.cpp:
(WebPageProxy::resetStateAfterProcessExited): Set m_userScriptsNotified to false.
2020-04-21 Timothy Horton <timothy_horton@apple.com>
Fix the build
* Configurations/WebKit.xcconfig:
2020-04-21 David Kilzer <ddkilzer@apple.com>
[IPC hardening] ShareableResource::create() should validate its parameters
<https://webkit.org/b/210779>
<rdar://problem/60887693>
Reviewed by Chris Dumez.
* NetworkProcess/cache/NetworkCache.cpp:
(WebKit::NetworkCache::Cache::store):
* NetworkProcess/cache/NetworkCacheEntry.cpp:
(WebKit::NetworkCache::Entry::initializeShareableResourceHandleFromStorageRecord const):
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::didFinishLoadForQuickLookDocumentInMainFrame):
- Handle nullptr returned from ShareableResource::create().
* Shared/ShareableResource.cpp:
(WebKit::ShareableResource::create):
- Change to return RefPtr<>.
- Validate parameters before calling constructor. Return
nullptr on failure.
(WebKit::ShareableResource::ShareableResource):
- Remove assert and comment since checks are now done in
ShareableResource::create().
* Shared/ShareableResource.h:
(WebKit::ShareableResource::create):
- Change to return RefPtr<>.
2020-04-21 Wenson Hsieh <wenson_hsieh@apple.com>
[Cocoa] Tighten up some more IPC message checks in WebPasteboardProxy
https://bugs.webkit.org/show_bug.cgi?id=210780
<rdar://problem/62088479>
Reviewed by Alex Christensen.
Adjust some existing IPC message checks in WebPasteboardProxy, and also introduces some new ones. See below for
more details.
* UIProcess/Cocoa/WebPasteboardProxyCocoa.mm:
(WebKit::WebPasteboardProxy::didModifyContentsOfPasteboard):
Move the `pasteboardName` check here out of `didModifyContentsOfPasteboard`, and into each call site, such that
we bail and kill the web process before invoking `didModifyContentsOfPasteboard` in the case where the name is
empty.
(WebKit::WebPasteboardProxy::getPasteboardPathnamesForType):
(WebKit::WebPasteboardProxy::getPasteboardStringForType):
(WebKit::WebPasteboardProxy::getPasteboardStringsForType):
(WebKit::WebPasteboardProxy::getPasteboardBufferForType):
(WebKit::WebPasteboardProxy::addPasteboardTypes):
(WebKit::WebPasteboardProxy::setPasteboardTypes):
(WebKit::WebPasteboardProxy::setPasteboardURL):
(WebKit::WebPasteboardProxy::setPasteboardColor):
(WebKit::WebPasteboardProxy::setPasteboardStringForType):
Add message checks for empty pasteboard types.
(WebKit::WebPasteboardProxy::setPasteboardBufferForType):
(WebKit::WebPasteboardProxy::writeCustomData):
(WebKit::WebPasteboardProxy::readStringFromPasteboard):
(WebKit::WebPasteboardProxy::readBufferFromPasteboard):
(WebKit::WebPasteboardProxy::writeURLToPasteboard):
(WebKit::WebPasteboardProxy::writeWebContentToPasteboard):
(WebKit::WebPasteboardProxy::writeImageToPasteboard):
(WebKit::WebPasteboardProxy::writeStringToPasteboard):
Add a message check for empty pasteboard types, but only when the string is also empty; this is because iOS
clears pasteboard data by writing an empty pasteboard string and type.
2020-04-21 Adrian Perez de Castro <aperez@igalia.com>
[GTK][CMake] Make gtk-unix-print a component of FindGTK.cmake
https://bugs.webkit.org/show_bug.cgi?id=210792
Reviewed by Carlos Garcia Campos.
No new tests needed.
* PlatformGTK.cmake: Use the GTK::UnixPrint imported target.
2020-04-21 Philippe Normand <pnormand@igalia.com>
[CMake] gst-build-backed build fails
https://bugs.webkit.org/show_bug.cgi?id=210742
Reviewed by Xabier Rodriguez-Calvar.
* PlatformWPE.cmake:
2020-04-21 Adrian Perez de Castro <aperez@igalia.com>
Non-unified build fixes late February 2020 edition
https://bugs.webkit.org/show_bug.cgi?id=210767
Unreviewed build fix.
* NetworkProcess/NetworkResourceLoadMap.h: Add missing wtf/Function.h header.
2020-04-20 Ross Kirsling <ross.kirsling@sony.com>
Classes marked final should not use protected access specifier
https://bugs.webkit.org/show_bug.cgi?id=210775
Reviewed by Daniel Bates.
* Shared/API/APIDictionary.h:
* WebProcess/GPU/media/TextTrackPrivateRemote.h:
* WebProcess/Plugins/Netscape/JSNPMethod.h:
* WebProcess/Plugins/Netscape/JSNPObject.h:
2020-04-20 Peng Liu <peng.liu6@apple.com>
Fix build failures when video fullscreen and picture-in-picture is disabled
https://bugs.webkit.org/show_bug.cgi?id=210777
Reviewed by Eric Carlson.
Wrap video fullscreen and picture-in-picture related code with "#if ENABLE(VIDEO_PRESENTATION_MODE)".
* Configurations/FeatureDefines.xcconfig:
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
(WebKit::RemoteMediaPlayerProxy::setVideoFullscreenGravity):
(WebKit::RemoteMediaPlayerProxy::updateVideoFullscreenInlineImage):
(WebKit::RemoteMediaPlayerProxy::setVideoFullscreenMode):
(WebKit::RemoteMediaPlayerProxy::videoFullscreenStandbyChanged):
(WebKit::RemoteMediaPlayerProxy::setBufferingPolicy):
* GPUProcess/media/RemoteMediaPlayerProxy.h:
* GPUProcess/media/RemoteMediaPlayerProxy.messages.in:
* GPUProcess/media/cocoa/RemoteMediaPlayerProxyCocoa.mm:
(WebKit::RemoteMediaPlayerProxy::prepareForPlayback):
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
(WebKit::MediaPlayerPrivateRemote::prepareForPlayback):
2020-04-20 Alex Christensen <achristensen@webkit.org>
SPI clients using fastServerTrustEvaluationEnabled need SPI to inform them of modern TLS negotiation
https://bugs.webkit.org/show_bug.cgi?id=210533
Reviewed by Brady Eidson.
* NetworkProcess/NetworkDataTask.h:
(WebKit::NetworkDataTaskClient::didNegotiateModernTLS):
* NetworkProcess/NetworkLoad.cpp:
(WebKit::NetworkLoad::didNegotiateModernTLS):
* NetworkProcess/NetworkLoad.h:
* NetworkProcess/cocoa/NetworkDataTaskCocoa.h:
* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::NetworkDataTaskCocoa::didNegotiateModernTLS):
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(-[WKNetworkSessionDelegate URLSession:task:didReceiveChallenge:completionHandler:]):
* UIProcess/API/APINavigationClient.h:
(API::NavigationClient::didNegotiateModernTLS):
* UIProcess/API/Cocoa/WKNavigationDelegatePrivate.h:
* UIProcess/Cocoa/NavigationState.h:
* UIProcess/Cocoa/NavigationState.mm:
(WebKit::NavigationState::setNavigationDelegate):
(WebKit::NavigationState::NavigationClient::didNegotiateModernTLS):
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::didNegotiateModernTLS):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/Network/NetworkProcessProxy.messages.in:
* UIProcess/WebPageProxy.cpp:
* UIProcess/WebPageProxy.h:
2020-04-20 Kate Cheney <katherine_cheney@apple.com>
App-bound domain behavior should abide by the limitsNavigationsToAppBoundDomains argument in WKWebView configuration
https://bugs.webkit.org/show_bug.cgi?id=210769
<rdar://problem/62065241>
Reviewed by Brent Fulgham.
Changes app-bound domain behavior to be triggered by the value of
limitsNavigationsToAppBoundDomains, a WebView configuration flag.
If the WebView has this parameter set and is currently navigating to
an app bound domain, then app-bound privileges will be granted.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::setIsNavigatingToAppBoundDomainAndCheckIfPermitted):
Update this function to return a boolean to indicate whether the
navigation should fail. A failure should occur if a WebView has set
the limitsNavigationsToAppBoundDomains flag and attempts to navigate
away from an app-bound domain.
If the limitsNavigationsToAppBoundDomains value has not been set to
true, maintain non-app bound behavior regardless of whether the domain
is app-bound or not.
(WebKit::WebPageProxy::decidePolicyForNavigationAction):
Check the result of setIsNavigatingToAppBoundDomainAndCheckIfPermitted and fail the
navigation if needed, with both RELEASE logging and an appropriate
error message.
(WebKit::WebPageProxy::setIsNavigatingToAppBoundDomain): Deleted.
* UIProcess/WebPageProxy.h:
Renamed for clarity.
2020-04-20 Nikos Mouchtaris <nmouchtaris@apple.com>
WK2 Quicklook for attachments
https://bugs.webkit.org/show_bug.cgi?id=208891
Reviewed by Darin Adler.
Allow attachment elements to render QuickLook thumbnail generated
from contents of the file as its icon, rather than the default
icons associated with each mime type.
* Configurations/WebKit.xcconfig:
Link QuickLook thumbnailing framework.
* UIProcess/API/APIAttachment.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _insertAttachmentWithFileWrapper:contentType:completion:]):
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
Create thumbnail request and add to queue.
(WebKit::convertNSImageToBitmap):
(WebKit::convertUIImageToBitmap):
(WebKit::WebPageProxy::getQLThumbnailForGenerationRequestion):
(WebKit::WebPageProxy::getQLThumbnailForFileWrapper):
(WebKit::WebPageProxy::getQLThumbnailForAttachment):
* UIProcess/QLThumbnailLoad.h: Added.
* UIProcess/QLThumbnailLoad.mm: Added.
Add code to request thumbnail for file data.
(-[WKQLThumbnailQueueManager init]):
(+[WKQLThumbnailQueueManager sharedInstance]):
(-[WKQLThumbnailLoadOperation initWithAttachment:identifier:]):
(-[WKQLThumbnailLoadOperation initWithURL:identifier:]):
(-[WKQLThumbnailLoadOperation start]):
(-[WKQLThumbnailLoadOperation isAsynchronous]):
(-[WKQLThumbnailLoadOperation isExecuting]):
(-[WKQLThumbnailLoadOperation setExecuting:]):
(-[WKQLThumbnailLoadOperation isFinished]):
(-[WKQLThumbnailLoadOperation setFinished:]):
* UIProcess/WebPageProxy.cpp:
* UIProcess/WebPageProxy.h:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updateAttachmentIcon):
Set thumbnail of attachment element
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
Thumbnail request message.
2020-04-20 Kate Cheney <katherine_cheney@apple.com>
Unreviewed macCatalyst build fix.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::disableServiceWorkerEntitlementInNetworkProcess):
(WebKit::WebPageProxy::clearServiceWorkerEntitlementOverride):
2020-04-20 Simon Fraser <simon.fraser@apple.com>
Scrolling with background-attachment: fixed needs to trigger repaints
https://bugs.webkit.org/show_bug.cgi?id=193893
<rdar://problem/47587017>
Reviewed by Dean Jackson.
When scrolling an overflow scroll which has "background-atttachment:fixed" in the content,
the node will have non-empty synchronous scrolling reasons. In this case we need to
send the scroll to the main thread, and trigger a repaint on scroll.
If handling the wheel event on the scrolling thread determines that the scroll must be sent
to the main thread, EventDispatcher::wheelEvent() does so in the callback function.
To trigger the repaint, RenderLayer::scrollTo() asks the composited layers backing whether
the node has synchronous scrolling reasons; this is implemented by asking the scrolling
coordinator.
* WebProcess/WebPage/EventDispatcher.cpp:
(WebKit::EventDispatcher::wheelEvent):
2020-04-20 Megan Gardner <megan_gardner@apple.com>
Date and Time form controls not showing correct initial values on immediate second invocation.
https://bugs.webkit.org/show_bug.cgi?id=210613
Reviewed by Wenson Hsieh.
Updating the webprocess with the newly picked time did not update the local store of that data in the
UI process, so when the control was activated a second time, and we skipped the round trip to the
webprocess to get the data we were certian we already had, we did not start the popover in the
correct state. We now update the webprocess and the local UI stored variable at the same time.
Test: fast/forms/ios/time-picker-value-change.html
* UIProcess/API/ios/WKWebViewPrivateForTestingIOS.h:
* UIProcess/API/ios/WKWebViewTestingIOS.mm:
(-[WKWebView timePickerValueHour]):
(-[WKWebView timePickerValueMinute]):
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView updateValueAsNumberForViewController:]):
(-[WKContentView updateValueForViewController:]):
(-[WKContentView setTimePickerValueToHour:minute:]):
(-[WKContentView timePickerValueHour]):
(-[WKContentView timePickerValueMinute]):
* UIProcess/ios/forms/WKFormInputControl.h:
* UIProcess/ios/forms/WKFormInputControl.mm:
(-[WKDateTimePicker hour]):
(-[WKDateTimePicker minute]):
(-[WKDateTimePicker _dateChangedSetAsNumber]):
(-[WKDateTimePicker _dateChangedSetAsString]):
(-[WKDateTimePicker setHour:minute:]):
(-[WKFormInputControl setTimePickerHour:minute:]):
(-[WKFormInputControl timePickerValueHour]):
(-[WKFormInputControl timePickerValueMinute]):
(-[WKDateTimePopover setHour:minute:]):
(-[WKDateTimePopover hour]):
(-[WKDateTimePopover minute]):
2020-04-20 Timothy Horton <timothy_horton@apple.com>
Try to fix the macCatalyst build after r260366
* UIProcess/API/Cocoa/_WKWebAuthenticationPanelInternal.h:
2020-04-20 Stephan Szabo <stephan.szabo@sony.com>
Fix build when !ENABLE(RESOURCE_LOAD_STATISTICS)
https://bugs.webkit.org/show_bug.cgi?id=210757
Reviewed by Ross Kirsling.
Build fix, no new tests
* NetworkProcess/NetworkDataTask.cpp:
Add ENABLE(RESOURCE_LOAD_STATISTICS) around block
using NetworkSession functions only declared when
that is enabled.
2020-04-20 Alex Christensen <achristensen@webkit.org>
When SpeculativeLoadManager is destroyed, properly clean up its PendingFrameLoads
https://bugs.webkit.org/show_bug.cgi?id=210759
<rdar://problem/62056856>
Reviewed by Darin Adler.
Recent work on the resourceLoadStatistics layout tests increased the amount we swap out the WebsiteDataStore.
When this happens, the NetworkSession is eventually destroyed in the NetworkProcess, sometimes when running the next test.
An assertion was firing in the PendingFrameLoad destructor because it hadn't been marked as complete when it was destroyed.
Rather than remove the assertion, when we destroy the SpeculativeLoadManager (which only happens when a WebsiteDataStore
is destroyed) during a speculative pending frame load, just mark the pending frame load as complete because it is being cancelled.
Marking the pending frame load as complete can tell the SpeculativeLoadManager to mutate m_pendingFrameLoads, which we don't want
to do while iterating, so copy the RefPtrs into a Vector first then iterate that to get them all.
This fixes an assertion that was sometimes hit in http/tests/resourceLoadStatistics/strip-referrer-to-origin-for-prevalent-subresource-redirects-database.html
but only after running other tests that had initiated speculative pending frame loads. This was ostensibly started by r260322 but is quite unrelated.
* NetworkProcess/cache/NetworkCacheSpeculativeLoadManager.cpp:
(WebKit::NetworkCache::SpeculativeLoadManager::~SpeculativeLoadManager):
2020-04-20 Stephan Szabo <stephan.szabo@sony.com>
[PlayStation] Fix build after r260277
https://bugs.webkit.org/show_bug.cgi?id=210756
Reviewed by Don Olmstead.
Build fix, no new tests.
* PlatformPlayStation.cmake:
2020-04-20 Chris Dumez <cdumez@apple.com>
Sending beacons when Fetch KeepAlive feature is disabled crashes the WebProcess
https://bugs.webkit.org/show_bug.cgi?id=210753
<rdar://problem/61896221>
Reviewed by Geoffrey Garen.
* WebProcess/Network/WebLoaderStrategy.cpp:
(WebKit::WebLoaderStrategy::startPingLoad):
Make sure NetworkResourceLoadParameters's webPageProxyID / webPageID / webFrameID are properly
initialized before sending the IPC or IPC decoding will fail.
2020-04-20 David Kilzer <ddkilzer@apple.com>
[IPC hardening] Use ObjectIdentifier<> for geolocationID
<https://webkit.org/b/210680>
<rdar://problem/58892340>
Reviewed by Chris Dumez.
Replace uses of uint64_t with GeolocationIdentifier.
* Scripts/webkit/messages.py:
* UIProcess/GeolocationIdentifier.h: Add.
* UIProcess/GeolocationPermissionRequestManagerProxy.cpp:
(WebKit::GeolocationPermissionRequestManagerProxy::createRequest):
(WebKit::GeolocationPermissionRequestManagerProxy::didReceiveGeolocationPermissionDecision):
* UIProcess/GeolocationPermissionRequestManagerProxy.h:
* UIProcess/GeolocationPermissionRequestProxy.cpp:
(WebKit::GeolocationPermissionRequestProxy::GeolocationPermissionRequestProxy):
* UIProcess/GeolocationPermissionRequestProxy.h:
(WebKit::GeolocationPermissionRequestProxy::create):
* UIProcess/WebPageProxy.cpp:
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/Geolocation/GeolocationPermissionRequestManager.cpp:
(WebKit::GeolocationPermissionRequestManager::startRequestForGeolocation):
(WebKit::GeolocationPermissionRequestManager::cancelRequestForGeolocation):
(WebKit::GeolocationPermissionRequestManager::didReceiveGeolocationPermissionDecision):
(WebKit::generateGeolocationID): Delete.
- Replace with GeolocationIdentifier::generate().
* WebProcess/Geolocation/GeolocationPermissionRequestManager.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::didReceiveGeolocationPermissionDecision):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-04-20 Chris Dumez <cdumez@apple.com>
[iOS] Refactor WebKit media playback process assertion logic to minimize chances of leaking them
https://bugs.webkit.org/show_bug.cgi?id=210670
Reviewed by Geoffrey Garen.
Refactor WebKit media playback process assertion logic to minimize chances of leaking them. In particular,
the following changes were made:
1. Instead of the WebProcessPool having a HashMap of media playback process assertions for
the WebProcess, we now store the assertion on the WebProcessProxy object itself. This is
less likely to get out of sync and leak.
2. Add a RefCounter to the WebProcessPool to count WebProcesses that have audible media.
Whenever a WebProcess starts or stops playing audible media, it merely grabs a RefCounter
token or releases it. The WebProcessPool relies on this counter to decide whether or not
to take a media playback assertion on behalf of the UIProcess. Since this is token-based
and the token is stored on the WebProcessProxy object, it makes it less likely to leak
the assertion.
3. The WebProcessProxy object now has a AudibleMediaActivity data structure wrapping both
its media playback assertion and its WebProcessWithAudibleMediaToken that it got from
the WebProcessPool. When the WebProcess shuts down (normally or due to crash/termination),
we make sure to clear this data structure.
4. Make sure that the WebProcessProxy updates its AudibleMediaActivity whenever a page is
removed from the WebProcess.
* UIProcess/WebPageProxy.cpp:
* UIProcess/WebProcessPool.cpp:
(WebKit::m_webProcessWithAudibleMediaCounter):
(WebKit::WebProcessPool::disconnectProcess):
(WebKit::WebProcessPool::webProcessWithAudibleMediaToken const):
(WebKit::WebProcessPool::updateAudibleMediaAssertions):
* UIProcess/WebProcessPool.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::shutDown):
(WebKit::WebProcessProxy::removeWebPage):
(WebKit::WebProcessProxy::updateAudibleMediaAssertions):
* UIProcess/WebProcessProxy.h:
2020-04-20 Kate Cheney <katherine_cheney@apple.com>
<rdar://problem/62059046>
Unreviewed macCatalyst build fix.
* NetworkProcess/NetworkProcess.messages.in:
2020-04-20 Darin Adler <darin@apple.com>
Use #import instead of #include in Objective-C and don't use #pragma once
https://bugs.webkit.org/show_bug.cgi?id=210724
Reviewed by David Kilzer.
* mac/MigrateHeadersFromWebKitLegacy.make: Delete stale MemoryMeasure.h
forwarding header from incremental WebKit builds and ignore when generating
new forwarding headers.
* Platform/foundation/LoggingFoundation.mm:
(WebKit::logLevelString): Use string directly instead of a named constant.
* UIProcess/mac/WebColorPickerMac.h: Mark class final.
* GPUProcess/webrtc/LibWebRTCCodecsProxy.mm:
* NetworkProcess/WebStorage/ios/LocalStorageDatabaseTrackerIOS.mm:
* NetworkProcess/cache/NetworkCacheDataCocoa.mm:
* NetworkProcess/cache/NetworkCacheIOChannelCocoa.mm:
* NetworkProcess/cocoa/NetworkActivityTrackerCocoa.mm:
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
* NetworkProcess/webrtc/NetworkRTCProvider.mm:
* Platform/spi/Cocoa/DeviceManagementSPI.h:
* Platform/spi/Cocoa/LocalAuthenticationSPI.h:
* Platform/spi/Cocoa/NearFieldSPI.h:
* Platform/spi/Cocoa/SafeBrowsingSPI.h:
* Platform/spi/mac/AppKitSPI.h:
* Shared/API/Cocoa/RemoteObjectInvocation.mm:
* Shared/API/Cocoa/WKMain.mm:
* Shared/APIWebArchive.mm:
* Shared/APIWebArchiveResource.mm:
* Shared/Authentication/cocoa/AuthenticationChallengeDispositionCocoa.mm:
* Shared/Cocoa/LoadParametersCocoa.mm:
* Shared/Cocoa/SandboxInitialiationParametersCocoa.mm:
* Shared/Cocoa/WKNSData.mm:
* Shared/DocumentEditingContext.mm:
* Shared/Plugins/mac/PluginSandboxProfile.mm:
* Shared/RemoteLayerTree/WKAnimationDelegate.h:
* Shared/UserInterfaceIdiom.mm:
* Shared/ios/WebIconUtilities.mm:
* Shared/mac/CodeSigning.mm:
* Shared/mac/ColorSpaceData.mm:
* Shared/mac/HangDetectionDisablerMac.mm:
* Shared/mac/PDFKitImports.mm:
* UIProcess/API/Cocoa/APIContentRuleListStoreCocoa.mm:
* UIProcess/API/Cocoa/APIHTTPCookieStoreCocoa.mm:
* UIProcess/API/Cocoa/APISerializedScriptValueCocoa.mm:
* UIProcess/API/Cocoa/WKContentRuleList.mm:
* UIProcess/API/Cocoa/WKMenuItemIdentifiers.mm:
* UIProcess/API/Cocoa/WKNavigationActionPrivate.h:
* UIProcess/API/Cocoa/WKPreviewActionItemIdentifiers.mm:
* UIProcess/API/Cocoa/WKWebViewInternal.h:
* UIProcess/API/Cocoa/WKWebpagePreferencesInternal.h:
* UIProcess/API/Cocoa/_WKApplicationManifest.h:
* UIProcess/API/Cocoa/_WKAttachment.h:
* UIProcess/API/Cocoa/_WKResourceLoadInfoInternal.h:
* UIProcess/API/Cocoa/_WKResourceLoadStatisticsFirstPartyInternal.h:
* UIProcess/API/Cocoa/_WKResourceLoadStatisticsThirdPartyInternal.h:
* UIProcess/API/Cocoa/_WKTextManipulationExclusionRule.mm:
* UIProcess/API/Cocoa/_WKUserContentFilter.mm:
* UIProcess/API/Cocoa/_WKWebAuthenticationAssertionResponseInternal.h:
* UIProcess/API/Cocoa/_WKWebAuthenticationPanelInternal.h:
* UIProcess/API/Cocoa/_WKWebsiteDataStoreDelegate.h:
* UIProcess/Authentication/mac/WebCredentialMac.mm:
* UIProcess/Automation/cocoa/WebAutomationSessionCocoa.mm:
* UIProcess/Cocoa/IconLoadingDelegate.mm:
* UIProcess/Cocoa/MediaUtilities.mm:
* UIProcess/Cocoa/PreferenceObserver.h:
* UIProcess/Cocoa/PreferenceObserver.mm:
* UIProcess/Cocoa/SOAuthorization/NavigationSOAuthorizationSession.mm:
* UIProcess/Cocoa/SOAuthorization/SOAuthorizationNSURLExtras.h:
* UIProcess/Cocoa/SOAuthorization/SOAuthorizationSession.mm:
* UIProcess/Cocoa/SOAuthorization/WKSOAuthorizationDelegate.h:
* UIProcess/Cocoa/WKEditCommand.h:
* UIProcess/Cocoa/WKFullKeyboardAccessWatcher.h:
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
* UIProcess/Gamepad/cocoa/UIGamepadProviderCocoa.mm:
* UIProcess/Inspector/mac/WKInspectorViewController.h:
* UIProcess/Inspector/mac/WKInspectorWKWebView.h:
* UIProcess/RemoteLayerTree/ios/RemoteLayerTreeViews.h:
* UIProcess/RemoteLayerTree/ios/ScrollingTreeScrollingNodeDelegateIOS.h:
* UIProcess/RemoteLayerTree/mac/ScrollerMac.mm:
* UIProcess/RemoteLayerTree/mac/ScrollerPairMac.mm:
* UIProcess/WebAuthentication/Cocoa/LocalAuthenticator.mm:
* UIProcess/WebAuthentication/Cocoa/WKNFReaderSessionDelegate.h:
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
* UIProcess/ios/WKDeferringGestureRecognizer.h:
* UIProcess/ios/WKHighlightLongPressGestureRecognizer.h:
* UIProcess/ios/WebDataListSuggestionsDropdownIOS.h:
* UIProcess/ios/WebDataListSuggestionsDropdownIOS.mm:
* UIProcess/ios/forms/WKDatePickerViewController.h:
* UIProcess/ios/forms/WKDatePickerViewController.mm:
* UIProcess/ios/forms/WKFocusedFormControlView.h:
* UIProcess/ios/forms/WKFocusedFormControlView.mm:
* UIProcess/ios/forms/WKFormColorPicker.mm:
* UIProcess/ios/forms/WKNumberPadViewController.h:
* UIProcess/ios/forms/WKNumberPadViewController.mm:
* UIProcess/ios/forms/WKQuickboardListViewController.h:
* UIProcess/ios/forms/WKQuickboardListViewController.mm:
* UIProcess/ios/forms/WKSelectMenuListViewController.h:
* UIProcess/ios/forms/WKTimePickerViewController.h:
* UIProcess/ios/forms/WKTimePickerViewController.mm:
* UIProcess/ios/fullscreen/WKFullScreenViewController.h:
* UIProcess/ios/fullscreen/WKFullscreenStackView.h:
* UIProcess/ios/fullscreen/WKFullscreenStackView.mm:
* UIProcess/mac/PageClientImplMac.mm:
* UIProcess/mac/WKSharingServicePickerDelegate.h:
* WebProcess/GPU/media/cocoa/MediaPlayerPrivateRemoteCocoa.mm:
* WebProcess/GPU/media/cocoa/VideoLayerRemoteCocoa.h:
* WebProcess/InjectedBundle/API/c/mac/WKBundlePageBannerMac.mm:
* WebProcess/WebCoreSupport/mac/WebFrameNetworkingContext.mm:
* WebProcess/WebPage/RemoteLayerTree/PlatformCAAnimationRemote.mm:
* WebProcess/WebPage/RemoteLayerTree/PlatformCALayerRemote.cpp:
* WebProcess/WebPage/RemoteLayerTree/PlatformCALayerRemoteCustom.mm:
* WebProcess/WebPage/RemoteLayerTree/PlatformCALayerRemoteTiledBacking.cpp:
* WebProcess/WebPage/WKAccessibilityWebPageObjectIOS.h:
* WebProcess/WebPage/mac/PageBannerMac.mm:
* WebProcess/WebPage/mac/WKAccessibilityWebPageObjectBase.h:
* WebProcess/WebPage/mac/WKAccessibilityWebPageObjectMac.h:
* WebProcess/cocoa/WebProcessCocoa.mm:
More #import, less #pragma once.
2020-04-20 Pablo Saavedra <psaavedra@igalia.com>
[GTK][WPE] lowWatermarkPages() in MemoryPressureMonitor.cpp only searches the "low" value inside the first "Node" section
https://bugs.webkit.org/show_bug.cgi?id=210345
Reviewed by Adrian Perez de Castro.
* UIProcess/linux/MemoryPressureMonitor.cpp:
(WebKit::lowWatermarkPages):
2020-04-20 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK][WPE] Enable resource load statistics
https://bugs.webkit.org/show_bug.cgi?id=210184
Reviewed by Žan Doberšek.
* NetworkProcess/NetworkDataTask.cpp:
(WebKit::NetworkDataTask::create): Pass page and frame identifiers to the constructor.
(WebKit::NetworkDataTask::isThirdPartyRequest const): Moved from NetworkDataTaskCocoa.mm.
(WebKit::NetworkDataTask::restrictRequestReferrerToOriginIfNeeded): Ditto.
* NetworkProcess/NetworkDataTask.h:
* NetworkProcess/cocoa/NetworkDataTaskCocoa.h:
* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
* NetworkProcess/soup/NetworkDataTaskSoup.cpp:
(WebKit::NetworkDataTaskSoup::NetworkDataTaskSoup): Initialize m_frameID and m_pageID.
(WebKit::NetworkDataTaskSoup::createRequest): Add WasBlockingCookies parameter. Call
restrictRequestReferrerToOriginIfNeeded() and disable cookies in the soup message if cookies should be blocked.
(WebKit::NetworkDataTaskSoup::clearRequest): Reset m_isBlockingCookies.
(WebKit::NetworkDataTaskSoup::continueHTTPRedirection): Pass WasBlockingCookies to createRequest().
* NetworkProcess/soup/NetworkDataTaskSoup.h:
* UIProcess/API/C/WKPage.cpp:
(WKPageLoadedThirdPartyDomains): Added for tests.
(WKPageClearLoadedThirdPartyDomains): Ditto.
* UIProcess/API/C/WKPagePrivate.h:
* UIProcess/glib/WebsiteDataStoreGLib.cpp:
(WebKit::WebsiteDataStore::defaultResourceLoadStatisticsDirectory): Use lowercase for consistency with other
default directories.
2020-04-20 David Kilzer <ddkilzer@apple.com>
REGRESSION (r259610): Leak of RBSAssertionInvalidationCallbackType due to missing -dealloc
<https://webkit.org/b/210711>
<rdar://problem/61993361>
Reviewed by Geoffrey Garen.
* UIProcess/ios/ProcessAssertionIOS.mm:
(-[WKRBSAssertionDelegate dealloc]): Add.
- Release _invalidationCallback to fix the leak.
2020-04-19 Brady Eidson <beidson@apple.com>
Add WKScriptMessageHandler API that asynchronously responds with a promise.
rdar://problem/57243492 and https://bugs.webkit.org/show_bug.cgi?id=206398
Reviewed by Andy Estes.
Change webkit.messageHandlers.<name>.postMessage() to return a promise instead of undefined.
Allow for that promise to be resolved by an asynchronous reply block up in the API client.
This is like the spiritual opposite version of [WKWebView callAsyncFunction:...]
And while we're adding a new script message handler variant, we're adding it sandboxed by WKContentWorld.
* Shared/API/APISerializedScriptValue.h:
* UIProcess/API/Cocoa/APISerializedScriptValueCocoa.mm:
(API::validateObject):
(API::coreValueFromNSObject):
(API::SerializedScriptValue::createFromNSObject):
(API::SerializedScriptValue::wireBytesFromNSObject): Deleted.
* UIProcess/API/Cocoa/WKScriptMessage.h: Now that script message handlers can be per-world, messages
declare which world they were posted from.
* UIProcess/API/Cocoa/WKScriptMessage.mm:
(-[WKScriptMessage _initWithBody:webView:frameInfo:name:world:]):
(-[WKScriptMessage world]):
(-[WKScriptMessage _initWithBody:webView:frameInfo:name:]): Deleted.
* UIProcess/API/Cocoa/WKScriptMessageInternal.h:
* UIProcess/API/Cocoa/WKScriptMessageHandlerWithReply.h: Added.
Declare the new protocol for a script message handler that can reply to messages asynchronously.
* UIProcess/API/Cocoa/WKUserContentController.h:
* UIProcess/API/Cocoa/WKUserContentController.mm:
(-[WKUserContentController _addScriptMessageHandler:]):
(-[WKUserContentController addScriptMessageHandler:name:]):
(-[WKUserContentController addScriptMessageHandler:contentWorld:name:]):
(-[WKUserContentController addScriptMessageHandlerWithReply:contentWorld:name:]):
(-[WKUserContentController removeScriptMessageHandlerForName:contentWorld:]):
(-[WKUserContentController removeAllScriptMessageHandlersFromContentWorld:]):
(-[WKUserContentController removeAllScriptMessageHandlers]):
* UIProcess/API/Cocoa/WKUserContentControllerInternal.h:
* UIProcess/API/Cocoa/WKUserContentControllerPrivate.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _evaluateJavaScript:asAsyncFunction:withArguments:forceUserGesture:inFrame:inWorld:completionHandler:]):
Update for new shared API::SerializedScriptValue initialization.
* UIProcess/API/glib/WebKitUserContentManager.cpp:
* UIProcess/API/gtk/WebKitRemoteInspectorProtocolHandler.cpp:
* UIProcess/Inspector/socket/RemoteInspectorProtocolHandler.cpp:
* UIProcess/UserContent/WebScriptMessageHandler.h:
* UIProcess/UserContent/WebUserContentControllerProxy.cpp:
(WebKit::WebUserContentControllerProxy::removeAllUserMessageHandlers):
(WebKit::WebUserContentControllerProxy::didPostMessage):
* UIProcess/UserContent/WebUserContentControllerProxy.h:
* UIProcess/UserContent/WebUserContentControllerProxy.messages.in:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/InjectedBundle/API/gtk/DOM/WebKitDOMDOMWindow.cpp:
(webkit_dom_dom_window_webkit_message_handlers_post_message):
* WebProcess/UserContent/WebUserContentController.cpp:
(WebKit::WebUserContentController::removeAllUserScriptMessageHandlers):
(WebKit::WebUserContentController::removeAllUserScriptMessageHandlersForWorlds):
* WebProcess/UserContent/WebUserContentController.h:
* WebProcess/UserContent/WebUserContentController.messages.in:
2020-04-19 David Kilzer <ddkilzer@apple.com>
REGRESSION (r244091): Leak of TaskStateChangedCallbackType due to missing -dealloc
<https://webkit.org/b/210710>
<rdar://problem/61992856>
Reviewed by Darin Adler.
* Shared/Cocoa/ProcessTaskStateObserver.mm:
(-[WKProcessTaskStateObserverDelegate dealloc]): Add.
- Release _taskStateChangedCallback to fix the leak.
2020-04-19 David Kilzer <ddkilzer@apple.com>
[IPC hardening] Use MESSAGE_CHECK in WebPageProxy::loadRecentSearches() and WebPageProxy::saveRecentSearches()
<https://webkit.org/b/210683>
<rdar://problem/59240446>
Reviewed by Geoffrey Garen.
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(MESSAGE_CHECK): Add.
(MESSAGE_CHECK_COMPLETION): Add.
(WebKit::WebPageProxy::saveRecentSearches):
(WebKit::WebPageProxy::loadRecentSearches):
- Replace boolean check of const String& with MESSAGE_CHECK.
2020-04-18 Darin Adler <darin@apple.com>
Update header postprocessing version cutoff to keep Apple internal builds working
https://bugs.webkit.org/show_bug.cgi?id=210708
Reviewed by Brady Eidson.
* Configurations/WebKit.xcconfig: Update versions to make building with older
Apple internal SDKs continue to work.
* Shared/API/Cocoa/WKFoundation.h:
* Shared/API/Cocoa/_WKFrameHandle.h:
* Shared/API/Cocoa/_WKRemoteObjectInterface.h:
* UIProcess/API/Cocoa/WKContentWorld.h:
* UIProcess/API/Cocoa/WKFindConfiguration.h:
* UIProcess/API/Cocoa/WKFindResult.h:
* UIProcess/API/Cocoa/WKFrameInfoPrivate.h:
* UIProcess/API/Cocoa/WKHTTPCookieStorePrivate.h:
* UIProcess/API/Cocoa/WKNavigationDelegatePrivate.h:
* UIProcess/API/Cocoa/WKPDFConfiguration.h:
* UIProcess/API/Cocoa/WKPreferences.h:
* UIProcess/API/Cocoa/WKPreferencesPrivate.h:
* UIProcess/API/Cocoa/WKProcessPoolPrivate.h:
* UIProcess/API/Cocoa/WKUIDelegatePrivate.h:
* UIProcess/API/Cocoa/WKURLSchemeTaskPrivate.h:
* UIProcess/API/Cocoa/WKUserScriptPrivate.h:
* UIProcess/API/Cocoa/WKWebView.h:
* UIProcess/API/Cocoa/WKWebViewConfiguration.h:
* UIProcess/API/Cocoa/WKWebViewConfigurationPrivate.h:
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/API/Cocoa/WKWebpagePreferences.h:
* UIProcess/API/Cocoa/WKWebpagePreferencesPrivate.h:
* UIProcess/API/Cocoa/WKWebsiteDataRecordPrivate.h:
* UIProcess/API/Cocoa/WKWebsiteDataStorePrivate.h:
* UIProcess/API/Cocoa/_WKDownload.h:
* UIProcess/API/Cocoa/_WKFrameTreeNode.h:
* UIProcess/API/Cocoa/_WKInputDelegate.h:
* UIProcess/API/Cocoa/_WKInspectorDebuggableInfo.h:
* UIProcess/API/Cocoa/_WKProcessPoolConfiguration.h:
* UIProcess/API/Cocoa/_WKRemoteWebInspectorViewController.h:
* UIProcess/API/Cocoa/_WKResourceLoadDelegate.h:
* UIProcess/API/Cocoa/_WKResourceLoadInfo.h:
* UIProcess/API/Cocoa/_WKResourceLoadStatisticsFirstParty.h:
* UIProcess/API/Cocoa/_WKResourceLoadStatisticsThirdParty.h:
* UIProcess/API/Cocoa/_WKTextManipulationConfiguration.h:
* UIProcess/API/Cocoa/_WKTextManipulationExclusionRule.h:
* UIProcess/API/Cocoa/_WKTextManipulationItem.h:
* UIProcess/API/Cocoa/_WKTextManipulationToken.h:
* UIProcess/API/Cocoa/_WKUserContentWorld.h:
* UIProcess/API/Cocoa/_WKUserStyleSheet.h:
* UIProcess/API/Cocoa/_WKWebAuthenticationAssertionResponse.h:
* UIProcess/API/Cocoa/_WKWebAuthenticationPanel.h:
* UIProcess/API/Cocoa/_WKWebsiteDataStoreConfiguration.h:
* UIProcess/API/Cocoa/_WKWebsitePolicies.h:
Because the way we do post-processing of headers in the legacy Xcode
build system won't work once the files have been processed once, touch
each file that has WK_MAC_TBA, WK_IOS_TBA, or
WK_FRAMEWORK_HEADER_POSTPROCESSING_ENABLED in it. Found something to
change in each file. This is likely unnecessary in the new Xcode build
system, but I wasn't able to test that locally.
* WebKit.xcodeproj/project.pbxproj:
(postprocess-header-rule): Added dependencies so that post-processing will be
redone if Configurations/WebKit.xcconfig is touched, since that is where
WK_FRAMEWORK_HEADER_POSTPROCESSING_DISABLED is set, or if
Scripts/postprocess-header-rule is touched, since that is where the
post-processing code is. If either of those files changes it could affect the
output of post-processing. This should make a change like this work in the
new Xcode build system without touching files as is done above.
2020-04-18 Brady Eidson <beidson@apple.com>
Fix WebUserContentControllerProxy vs ContentWorld lifetime
https://bugs.webkit.org/show_bug.cgi?id=210700
Reviewed by Alex Christensen.
Covered by API test.
WebUserContentControllerProxy currently keeps all of its associated API::ContentWorlds alive via RefPtrs.
This is despite the fact that all of the associated WebScriptMessageHandlers, UserScripts, and
UserStyleSheets already keep their associated API::ContentWorlds alive.
It then decideds to tell WebProcesses to forget a content world after all of its clients are removed.
Unfortunately, content worlds are used for more than just content controller stuff. They're used for direct
JavaScript evaluation as well.
So a client could:
- Add a script message handler in a content world.
- Evaluate JavaScript in that content world, setting up some persistent state.
- Remove the script message handler.
- Find that their persistent state from the JavaScript evaluation is gone from that world, even though they
still retain a usable handle to that world.
The only party who has any business managing the lifetime of an API::ContentWorld is the API::ContentWorld itself.
Making this change is:
1 - Nice cleanup
2 - Fixes the above mentioned bug
* UIProcess/API/APIContentWorld.cpp:
(API::ContentWorld::worldForIdentifier):
(API::ContentWorld::ContentWorld):
(API::ContentWorld::sharedWorldWithName):
(API::ContentWorld::~ContentWorld):
(API::ContentWorld::addAssociatedUserContentControllerProxy):
(API::ContentWorld::userContentControllerProxyDestroyed):
* UIProcess/API/APIContentWorld.h:
* UIProcess/UserContent/WebUserContentControllerProxy.cpp:
(WebKit::WebUserContentControllerProxy::parameters const):
(WebKit::WebUserContentControllerProxy::addContentWorld):
(WebKit::WebUserContentControllerProxy::contentWorldDestroyed):
(WebKit::WebUserContentControllerProxy::addUserScript):
(WebKit::WebUserContentControllerProxy::removeUserScript):
(WebKit::WebUserContentControllerProxy::removeAllUserScripts):
(WebKit::WebUserContentControllerProxy::addUserStyleSheet):
(WebKit::WebUserContentControllerProxy::removeUserStyleSheet):
(WebKit::WebUserContentControllerProxy::removeAllUserStyleSheets):
(WebKit::WebUserContentControllerProxy::addUserScriptMessageHandler):
(WebKit::WebUserContentControllerProxy::removeUserMessageHandlerForName):
(WebKit::WebUserContentControllerProxy::removeAllUserMessageHandlers):
(WebKit::WebUserContentControllerProxy::addContentWorldUse): Deleted.
(WebKit::WebUserContentControllerProxy::shouldSendRemoveContentWorldsMessage): Deleted.
(WebKit::WebUserContentControllerProxy::removeContentWorldUses): Deleted.
* UIProcess/UserContent/WebUserContentControllerProxy.h:
2020-04-18 David Kilzer <ddkilzer@apple.com>
Attempt #4 to fix tvOS build
Unreviewed.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
- Add #if ENABLE(VIDEO_PRESENTATION_MODE)/#endif to protect
function defined in WebCore/PictureInPictureSupport.h.
- See previous build fixes in r260307, r260308 and r260313.
2020-04-18 Alex Christensen <achristensen@webkit.org>
Fix client certificate authentication when using non-default WKWebsiteDataStores
https://bugs.webkit.org/show_bug.cgi?id=210681
Reviewed by Brady Eidson.
NetworkProcessProxy was trying to keep a map of WebsiteDataStores, but it wasn't as accurate as the one WebsiteDataStore was maintaining.
Use the latter map instead and client certificate authentication works. Otherwise, the credential isn't serialized correctly.
I found this while working on <rdar://problem/60340449> but this was unrelated so I put it in a different change.
This may fix <rdar://problem/60910392>.
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::websiteDataStoreFromSessionID):
2020-04-18 Pablo Saavedra <psaavedra@igalia.com>
[GTK][WPE] Replace evil strtok() calls with fscanf() in MemoryPressureMonitor.cpp
https://bugs.webkit.org/show_bug.cgi?id=210346
Reviewed by Adrian Perez de Castro.
* UIProcess/linux/MemoryPressureMonitor.cpp:
(WebKit::lowWatermarkPages):
(WebKit::getCgroupControllerPath):
(WebKit::systemMemoryUsedAsPercentage):
(WebKit::CGroupMemoryController::getCgroupFileValue):
2020-04-18 Rob Buis <rbuis@igalia.com>
Reduce parameter list of the FrameLoadRequest constructor
https://bugs.webkit.org/show_bug.cgi?id=210668
Reviewed by Darin Adler.
Adapt to API change.
* WebProcess/Inspector/WebInspector.cpp:
(WebKit::WebInspector::openInNewTab):
* WebProcess/Plugins/PluginView.cpp:
(WebKit::PluginView::performFrameLoadURLRequest):
(WebKit::PluginView::loadURL):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::loadURLInFrame):
(WebKit::WebPage::loadDataInFrame):
(WebKit::WebPage::loadRequest):
(WebKit::WebPage::loadDataImpl):
2020-04-18 David Kilzer <ddkilzer@apple.com>
[IPC hardening] Use MESSAGE_CHECK in WebPasteboardProxy
<https://webkit.org/b/210684>
<rdar://problem/59906721>
Reviewed by Wenson Hsieh.
* UIProcess/Cocoa/WebPasteboardProxyCocoa.mm:
- Add missing #undef of MESSAGE_CHECK_COMPLETION and
MESSAGE_CHECK_WITH_RETURN_VALUE.
(WebKit::WebPasteboardProxy::getPasteboardPathnamesForType):
(WebKit::WebPasteboardProxy::getPasteboardStringForType):
(WebKit::WebPasteboardProxy::getPasteboardStringsForType):
(WebKit::WebPasteboardProxy::getPasteboardBufferForType):
(WebKit::WebPasteboardProxy::setPasteboardStringForType):
(WebKit::WebPasteboardProxy::setPasteboardBufferForType):
(WebKit::WebPasteboardProxy::typesSafeForDOMToReadAndWrite):
(WebKit::WebPasteboardProxy::readStringFromPasteboard):
(WebKit::WebPasteboardProxy::readBufferFromPasteboard):
- Replace existing code with MESSAGE_CHECK_COMPLETION macros.
* UIProcess/WebPasteboardProxy.cpp:
(WebKit::WebPasteboardProxy::typesSafeForDOMToReadAndWrite):
* UIProcess/WebPasteboardProxy.h:
* UIProcess/WebPasteboardProxy.messages.in:
- Add IPC::Connection to TypesSafeForDOMToReadAndWrite.
2020-04-17 Kate Cheney <katherine_cheney@apple.com>
Enable service workers for app-bound domains
https://bugs.webkit.org/show_bug.cgi?id=210451
<rdar://problem/61479474>
Reviewed by Brent Fulgham.
* NetworkProcess/NetworkConnectionToWebProcess.cpp:
(WebKit::NetworkConnectionToWebProcess::createFetchTask):
(WebKit::NetworkConnectionToWebProcess::scheduleResourceLoad):
(WebKit::NetworkConnectionToWebProcess::swConnection):
(WebKit::NetworkConnectionToWebProcess::isServiceWorkerAllowed const): Deleted.
* NetworkProcess/NetworkConnectionToWebProcess.h:
Removed ASSERTS and service worker entitlement checks which are now done along with app-bound
domain checks in SWServer.cpp.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::initializeNetworkProcess):
(WebKit::NetworkProcess::addWebsiteDataStore):
Remove the check for parentProcessHasServiceWorkerEntitlement() before
adding a serviceWorkerSession. This doesn't do anything except add an
entry to a map, and simplifies the case where we need a session for
app-bound instances which should have service workers but have no entitlement.
Pass the entitlement and a new callback to request app-bound domains
from the UI Process to the SWServer. This will make sure the server
checks for the entitlement or list of app-bound domains before completing a load.
(WebKit::NetworkProcess::registerSWServerConnection):
Remove unnecessary ASSERT for entitlement.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/ios/NetworkProcessIOS.mm:
(WebKit::NetworkProcess::parentProcessHasServiceWorkerEntitlement const):
(WebKit::NetworkProcess::disableServiceWorkerEntitlement):
(WebKit::NetworkProcess::clearServiceWorkerEntitlementOverride):
In order to test this thoroughly, TestWebKitAPI needed a way to
temporarily disable the service worker entitlement. This function
overrides the check for the entitlement if the bool is set, disabling
the entitlement.
* Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode const):
(WebKit::WebPageCreationParameters::decode):
* Shared/WebPageCreationParameters.h:
WebPage needs to know if it is app-bound when created so it can update
the service worker preferences key accordingly.
* Shared/WebPreferences.yaml:
* UIProcess/API/Cocoa/WKPreferences.mm:
(-[WKPreferences _serviceWorkerEntitlementDisabledForTesting]):
(-[WKPreferences _setServiceWorkerEntitlementDisabledForTesting:]):
* UIProcess/API/Cocoa/WKPreferencesPrivate.h:
Disable entitlement for testing.
* UIProcess/API/APIPageConfiguration.cpp:
(API::PageConfiguration::copy const):
* UIProcess/API/APIPageConfiguration.h:
(API::PageConfiguration::limitsNavigationsToAppBoundDomains const):
(API::PageConfiguration::setLimitsNavigationsToAppBoundDomains):
(API::PageConfiguration::limitsNavigationToAppBoundDomains const): Deleted.
(API::PageConfiguration::setLimitsNavigationToAppBoundDomains): Deleted.
(-[WKWebViewConfiguration limitsNavigationsToAppBoundDomains]):
(-[WKWebViewConfiguration setLimitsNavigationsToAppBoundDomains:]):
(-[WKWebViewConfiguration limitsNavigationToAppBoundDomains]): Deleted.
(-[WKWebViewConfiguration setLimitsNavigationToAppBoundDomains:]): Deleted.
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
Inconsistent naming (Navigation vs Navigations).
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _setupPageConfiguration:]):
Check for entitlement or app-bound webview.
(-[WKWebView _serviceWorkersEnabled:]):
(-[WKWebView _clearServiceWorkerEntitlementOverride:]):
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
Testing SPIs.
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::getAppBoundDomains):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/Network/NetworkProcessProxy.messages.in:
Sends the app-bound domains to the Network Process.
* UIProcess/WebPageProxy.cpp:
(WebKit::m_limitsNavigationsToAppBoundDomains):
(WebKit::m_limitsNavigationToAppBoundDomains): Deleted.
Naming inconsistency.
(WebKit::WebPageProxy::disableServiceWorkerEntitlementInNetworkProcess):
(WebKit::WebPageProxy::clearServiceWorkerEntitlementOverride):
* UIProcess/WebPageProxy.h:
Disable entitlement for testing.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_isNavigatingToAppBoundDomain):
Set this parameter in the constructor so it will be accurate when
updating preferences.
(WebKit::WebPage::updatePreferences):
* WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::clearServiceWorkerEntitlementOverride):
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::parentProcessHasServiceWorkerEntitlement const):
(WebKit::WebPage::disableServiceWorkerEntitlement):
(WebKit::WebPage::clearServiceWorkerEntitlementOverride):
Disable entitlement for testing.
2020-04-17 Alex Christensen <achristensen@webkit.org>
NetworkSessionCocoa should request client certificate only once per host/port
https://bugs.webkit.org/show_bug.cgi?id=210626
<rdar://problem/60340449>
Reviewed by Geoffrey Garen.
NSURLSession creates more than one TCP connection to a server when using HTTP 1.1.
Each TCP connection with TLS generates a didReceiveChallenge to do the server trust evaluation of the certificate chain.
If the server requests a client certificate in the TLS handshake, it also generates a didReceiveChallenge to request client
certificates as well. This is an implementation detail of our networking. We should not actually ask the WKNavigationDelegate
for client certificates more than once per host/port. We should remember the credential and give it to NSURLSession immediately
if we have used this credential in the past for a task that has received bytes (either a response or a redirect). If the TLS
handshake fails, we should not reuse that same certificate automatically.
* NetworkProcess/cocoa/NetworkSessionCocoa.h:
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(-[WKNetworkSessionDelegate URLSession:task:willPerformHTTPRedirection:newRequest:completionHandler:]):
(-[WKNetworkSessionDelegate URLSession:task:didReceiveChallenge:completionHandler:]):
(-[WKNetworkSessionDelegate URLSession:task:didCompleteWithError:]):
(-[WKNetworkSessionDelegate URLSession:dataTask:didReceiveResponse:completionHandler:]):
(WebKit::NetworkSessionCocoa::clientCertificateSuggestedForHost):
(WebKit::NetworkSessionCocoa::taskReceivedBytes):
(WebKit::NetworkSessionCocoa::taskFailed):
(WebKit::NetworkSessionCocoa::successfulClientCertificateForHost const):
2020-04-17 David Kilzer <ddkilzer@apple.com>
Bug 210646: REGRESSION (r260112): createArchiveList() leaks malloc memory on early returns due to an error
<https://webkit.org/b/210646>
<rdar://problem/61928031>
Reviewed by Darin Adler.
* Shared/mac/WebCoreArgumentCodersMac.mm:
(IPC::createArchiveList):
- Move early returns for protocolProperties, protocolProperties
and mimeType above memory allocation to fix the memory leak.
2020-04-17 Brady Eidson <beidson@apple.com>
Pass sandbox extensions for back/forward list navigations after the policy is decided at process-swap time.
<rdar://problem/59535167> and https://bugs.webkit.org/show_bug.cgi?id=210623
Reviewed by Geoff Garen.
Covered by almost all existing tests, and a new API test.
Instead of granting a sandbox extension when updating the back/forward cursor for a pending
back/forward list traversal, do so after the client decides the policy.
(Which is also along with a process swap in interesting cases)
* Shared/PolicyDecision.h:
(WebKit::PolicyDecision::encode const):
(WebKit::PolicyDecision::decode):
* Shared/WebPageCreationParameters.h:
* UIProcess/API/APINavigation.cpp:
(API::Navigation::Navigation):
* UIProcess/API/APINavigation.h:
(API::Navigation::create):
(API::Navigation::reloadItem const):
* UIProcess/ProvisionalPageProxy.cpp:
(WebKit::ProvisionalPageProxy::backForwardGoToItem):
* UIProcess/ProvisionalPageProxy.h:
* UIProcess/WebNavigationState.cpp:
(WebKit::WebNavigationState::createReloadNavigation):
* UIProcess/WebNavigationState.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::launchProcessForReload):
(WebKit::WebPageProxy::reload):
(WebKit::WebPageProxy::receivedNavigationPolicyDecision):
(WebKit::WebPageProxy::receivedPolicyDecision):
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::receivedPolicyDecision):
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebPage/WebBackForwardListProxy.cpp:
(WebKit::WebBackForwardListProxy::goToItem):
* WebProcess/WebPage/WebFrame.cpp:
(WebKit::WebFrame::didReceivePolicyDecision):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::SandboxExtensionTracker::didStartProvisionalLoad):
2020-04-17 David Kilzer <ddkilzer@apple.com>
REGRESSION (r234105): [iOS] WKColorButton leaks a UIColor
<https://webkit.org/b/210658>
<rdar://problem/61938137>
Reviewed by Darin Adler.
* UIProcess/ios/forms/WKFormColorPicker.mm:
(-[WKColorButton dealloc]): Add.
- Release _color instance variable to fix leak.
2020-04-17 Per Arne Vollan <pvollan@apple.com>
[iOS] Deny iokit open access of unused class in the WebContent sandbox
https://bugs.webkit.org/show_bug.cgi?id=210669
Reviewed by Brent Fulgham.
Telemetry shows that iokit open access of an iokit class can be removed from the WebContent sandbox.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-04-17 Chris Dumez <cdumez@apple.com>
[iOS] Refactor WebKit uploads process assertion logic to minimize chances of leaking them
https://bugs.webkit.org/show_bug.cgi?id=210664
<rdar://problem/61751982>
Reviewed by Geoff Garen.
To minimize chances of leaking "WebKit uploads" process assertion, the following refactoring
was done:
1. WebKit upload assertions are now held by NetworkProcessProxy instead of the WebProcessPool.
This makes more sense since uploads are tied to a particular network process. In base of
a network process crash, we would not want those assertions to persist for example.
2. Store all upload assertions (UIProcess + NetworkProcess + WebProcesses) into a single
UploadActivity data structure that can easily be cleared. Make sure we clear it on network
process crash.
3. No longer on the WebProcess to send IPC whenever its has pending uploads or not. Instead,
have the network process send such IPC. This makes more sense, especially now that all
webkit upload process assertions are now stored on NetworkProcessProxy. It also makes it
so that we deal properly with WebProcess crashes. In case of a WebProcess crash, its
connection to the network process will get closed. NetworkConnectionToWebProcess::didClose()
gets called and causes all NetworkResourceLoader objects for this connection to get
destroyed. As a result, the network process will properly notify the UIProcess that this
process no longer has any upload in progress.
* NetworkProcess/NetworkConnectionToWebProcess.cpp:
(WebKit::NetworkConnectionToWebProcess::NetworkConnectionToWebProcess):
(WebKit::NetworkConnectionToWebProcess::~NetworkConnectionToWebProcess):
(WebKit::NetworkConnectionToWebProcess::hasUploadStateChanged):
* NetworkProcess/NetworkConnectionToWebProcess.h:
* NetworkProcess/NetworkResourceLoadMap.cpp:
(WebKit::NetworkResourceLoadMap::NetworkResourceLoadMap):
(WebKit::NetworkResourceLoadMap::~NetworkResourceLoadMap):
(WebKit::NetworkResourceLoadMap::add):
(WebKit::NetworkResourceLoadMap::clear):
(WebKit::NetworkResourceLoadMap::take):
(WebKit::NetworkResourceLoadMap::setHasUpload):
* NetworkProcess/NetworkResourceLoadMap.h:
(WebKit::NetworkResourceLoadMap::hasUpload const):
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::didClose):
(WebKit::NetworkProcessProxy::setWebProcessHasUploads):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/Network/NetworkProcessProxy.messages.in:
* UIProcess/ProcessAssertion.cpp:
(WebKit::ProcessAssertion::ProcessAssertion):
* UIProcess/ProcessAssertion.h:
(WebKit::ProcessAssertion::pid const):
* UIProcess/WebProcessPool.cpp:
* UIProcess/WebProcessPool.h:
* UIProcess/WebProcessPool.messages.in:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::~WebProcessProxy):
* UIProcess/WebProcessProxy.h:
* WebProcess/Network/WebLoaderStrategy.cpp:
(WebKit::WebLoaderStrategy::scheduleLoadFromNetworkProcess):
(WebKit::WebLoaderStrategy::remove):
(WebKit::WebLoaderStrategy::loadResourceSynchronously):
* WebProcess/Network/WebLoaderStrategy.h:
2020-04-17 Jer Noble <jer.noble@apple.com>
Bring Fullscreen API + alert() behavior from iOS to macOS
https://bugs.webkit.org/show_bug.cgi?id=210662
Reviewed by Eric Carlson.
When a page is showing an alert(), all non-synchronous messages will be blocked until the sync XPC for alert()
ends. And when exiting fullscreen, we block completing the fullscreen exit animation until a repaint XPC completes.
So a page that is showing an alert() will never fully exit element fullscreen mode.
On iOS, we fixed this behavior by exiting fullscreen mode whenever the page shows an alert() or prompt(). We should
bring this behavior to macOS as well.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::runJavaScriptAlert):
(WebKit::WebPageProxy::runJavaScriptConfirm):
(WebKit::WebPageProxy::runJavaScriptPrompt):
2020-04-17 Don Olmstead <don.olmstead@sony.com>
[CMake] Add WebKit::WebCore target
https://bugs.webkit.org/show_bug.cgi?id=210445
Reviewed by Michael Catanzaro.
Use the WebKit::WebCore target within the WebKit target.
* CMakeLists.txt:
* PlatformGTK.cmake:
* PlatformWPE.cmake:
* PlatformWin.cmake:
2020-04-17 Simon Fraser <simon.fraser@apple.com>
Turn async frame and overflow scrolling on by default on macOS
https://bugs.webkit.org/show_bug.cgi?id=210661
rdar://problem/60420037
Reviewed by Tim Horton.
Flip the plist value and the default value for macOS which is outside of the feature flag #ifdef.
Testing keeps async overflow scrolling off by default, so this should not affect testing.
This only impacts WebKit2.
* FeatureFlags/WebKit.plist:
* Shared/WebPreferencesDefaultValues.cpp:
(WebKit::defaultAsyncFrameAndOverflowScrollingEnabled):
2020-04-17 Per Arne Vollan <pvollan@apple.com>
[iOS] Add telemetry to some mach lookup sandbox extensions
https://bugs.webkit.org/show_bug.cgi?id=210657
Reviewed by Darin Adler.
Add telemetry in order to learn more about how some of the mach lookup extensions are being used in the WebContent process on iOS.
No new tests, since this patch does not change any behavior.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-04-17 Youenn Fablet <youenn@apple.com>
Validate socket in NetworkRTCProvider::wrapNewTCPConnection
https://bugs.webkit.org/show_bug.cgi?id=210594
Reviewed by Eric Carlson.
* NetworkProcess/webrtc/NetworkRTCProvider.cpp:
(WebKit::NetworkRTCProvider::wrapNewTCPConnection):
Add a null check in case the identifier does not relate to an existing pending socket.
2020-04-17 Peng Liu <peng.liu6@apple.com>
Cleanup the macros for video fullscreen and picture-in-picture
https://bugs.webkit.org/show_bug.cgi?id=210638
Reviewed by Eric Carlson.
Replace some "#if PLATFORM(IOS_FAMILY) || (PLATFORM(MAC) && ENABLE(VIDEO_PRESENTATION_MODE))"
and all "#if (PLATFORM(IOS_FAMILY) && HAVE(AVKIT)) || (PLATFORM(MAC) && ENABLE(VIDEO_PRESENTATION_MODE))"
with "#if ENABLE(VIDEO_PRESENTATION_MODE)".
* GPUProcess/media/RemoteMediaPlayerProxy.cpp:
(WebKit::RemoteMediaPlayerProxy::setVideoFullscreenGravity):
(WebKit::RemoteMediaPlayerProxy::updateVideoFullscreenInlineImage):
(WebKit::RemoteMediaPlayerProxy::setVideoFullscreenMode):
(WebKit::RemoteMediaPlayerProxy::videoFullscreenStandbyChanged):
* UIProcess/Cocoa/VideoFullscreenManagerProxy.h:
* UIProcess/Cocoa/VideoFullscreenManagerProxy.messages.in:
* UIProcess/Cocoa/VideoFullscreenManagerProxy.mm:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didAttachToRunningProcess):
(WebKit::WebPageProxy::viewDidLeaveWindow):
(WebKit::WebPageProxy::exitFullscreenImmediately):
* UIProcess/WebPageProxy.h:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.cpp:
* WebProcess/GPU/media/MediaPlayerPrivateRemote.h:
* WebProcess/GPU/media/cocoa/MediaPlayerPrivateRemoteCocoa.mm:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::~WebPage):
* WebProcess/WebPage/WebPage.h:
* WebProcess/cocoa/VideoFullscreenManager.h:
* WebProcess/cocoa/VideoFullscreenManager.messages.in:
* WebProcess/cocoa/VideoFullscreenManager.mm:
2020-04-17 Claudio Saavedra <csaavedra@igalia.com>
[GTK] Update for GdkKeymap API changes
https://bugs.webkit.org/show_bug.cgi?id=210642
Reviewed by Adrian Perez de Castro.
No new tests needed.
gdk_keymap_get_default() is deprecated in GTK+ 3.22, so use
gdk_keymap_get_for_display() instead. Since in GTK4 this method is
removed to gdk_display_get_keymap(), add a helper to
GtkVersioning.h to avoid cluttering with ifdefs all over the
place.
* UIProcess/Automation/gtk/WebAutomationSessionGtk.cpp:
(WebKit::doKeyStrokeEvent):
2020-04-17 Antoine Quint <graouts@apple.com>
Allow the UserStyleLevel to be set through _WKUserStyleSheet
https://bugs.webkit.org/show_bug.cgi?id=210653
<rdar://problem/60506645>
Reviewed by Tim Horton.
Add a new _WKUserStyleLevel enum such that we may provide the specificity level of the backing UserStyleSheet, represented by the UserStyleLevel enum in WebCore.
* UIProcess/API/Cocoa/_WKUserStyleSheet.h:
* UIProcess/API/Cocoa/_WKUserStyleSheet.mm:
(-[_WKUserStyleSheet initWithSource:forWKWebView:forMainFrameOnly:level:userContentWorld:]):
(-[_WKUserStyleSheet initWithSource:forMainFrameOnly:legacyWhitelist:legacyBlacklist:baseURL:level:userContentWorld:]):
(-[_WKUserStyleSheet initWithSource:forWKWebView:forMainFrameOnly:userContentWorld:]): Deleted.
* UIProcess/API/Cocoa/_WKUserStyleSheetInternal.h:
(API::toWebCoreUserStyleLevel):
2020-04-17 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK][X11] REGRESSION(r259944): Wrong position of select popup menu in X11
https://bugs.webkit.org/show_bug.cgi?id=210603
Reviewed by Michael Catanzaro.
gdk_window_move_to_rect expects the given rectangle in coordinates relative to the top-left corner of the window
that the popup window is transient for. We were using screen coordinates.
* UIProcess/gtk/WebPopupMenuProxyGtk.cpp:
(WebKit::WebPopupMenuProxyGtk::showPopupMenu): Translate widget coordinates to window coordinates before passing
the rectangle to gdk_window_move_to_rect().
2020-04-17 David Kilzer <ddkilzer@apple.com>
REGRESSION (r256756): -[WKUserDefaults initWithSuiteName:]: Instance variable used while 'self' is not set to the result of '[(super or self) init...]'
<https://webkit.org/b/210647>
<rdar://problem/61928289>
Reviewed by Anders Carlsson.
* UIProcess/Cocoa/PreferenceObserver.mm:
(-[WKUserDefaults initWithSuiteName:]):
- Call `[self initWithSuite:]` and set `self` to fix the clang
static analyzer warning.
- Fix camelCase of `suitename` parameter.
2020-04-17 Per Arne Vollan <pvollan@apple.com>
[iOS] Deny iokit open access to graphics related classes
https://bugs.webkit.org/show_bug.cgi?id=210616
Reviewed by Darin Adler.
Deny iokit open access to graphics related classes in the WebContent process on iOS, but issue
extensions for these for some devices which still need access to them.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* Shared/Cocoa/SandboxExtensionCocoa.mm:
(WebKit::SandboxExtensionImpl::sandboxExtensionForType):
(WebKit::SandboxExtension::createHandleForIOKitClassExtension):
* Shared/SandboxExtension.h:
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/API/ios/WKWebViewPrivateForTestingIOS.h:
* UIProcess/API/ios/WKWebViewTestingIOS.mm:
(-[WKWebView _setDeviceHasAGXCompilerServiceForTesting]):
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
(WebKit::deviceHasAGXCompilerService): Deleted.
* UIProcess/WebPageProxy.h:
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::setDeviceHasAGXCompilerServiceForTesting const):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-04-17 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] UI process crash when entering compositing mode when WPE_RENDERER is enabled
https://bugs.webkit.org/show_bug.cgi?id=209118
Reviewed by Michael Catanzaro.
Check if EGL_WL_bind_wayland_display extension is available when using WPE_RENDERER, since we don't suport the
SHM interface.
* UIProcess/gtk/AcceleratedBackingStoreWayland.cpp:
(WebKit::AcceleratedBackingStoreWayland::checkRequirements): Return false if EGL_WL_bind_wayland_display is not present.
2020-04-17 Rob Buis <rbuis@igalia.com>
Move allowPlugins to FrameLoader
https://bugs.webkit.org/show_bug.cgi?id=205876
Reviewed by Darin Adler.
Adapt to API change.
* WebProcess/Plugins/WebPluginInfoProvider.cpp:
(WebKit::WebPluginInfoProvider::pluginInfo):
(WebKit::WebPluginInfoProvider::populatePluginCache):
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::objectContentType):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::createPlugin):
(WebKit::WebPage::canPluginHandleResponse):
(WebKit::WebPage::canShowMIMEType const):
2020-04-16 Peng Liu <peng.liu6@apple.com>
Cleanup the usage of ENABLE_FULLSCREEN_API
https://bugs.webkit.org/show_bug.cgi?id=210630
Reviewed by Daniel Bates.
Replace "#if defined(ENABLE_FULLSCREEN_API) && ENABLE_FULLSCREEN_API" with
"#if ENABLE(FULLSCREEN_API)".
No new tests, no functional change
* WebProcess/InjectedBundle/API/c/WKBundlePage.cpp:
(WKBundlePageSetFullScreenClient):
(WKBundlePageWillEnterFullScreen):
(WKBundlePageDidEnterFullScreen):
(WKBundlePageWillExitFullScreen):
(WKBundlePageDidExitFullScreen):
2020-04-16 Kate Cheney <katherine_cheney@apple.com>
Distinguish between an empty WKAppBoundDomains array and a missing entry
https://bugs.webkit.org/show_bug.cgi?id=210628
<rdar://problem/61909669>
Reviewed by Brent Fulgham.
Currently we treat a missing WKAppBoundDomains key the same as an empty
key, which is incorrect. Having the key present, even with no entries,
should opt an app into app-bound domain behavior.
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::WebsiteDataStore::initializeAppBoundDomains):
(WebKit::WebsiteDataStore::beginAppBoundDomainCheck):
2020-04-16 Brent Fulgham <bfulgham@apple.com>
Add a new 'limitNavigationsToAppBoundDomains' property to WKWebViewConfiguration
https://bugs.webkit.org/show_bug.cgi?id=210620
<rdar://problem/61903225>
Reviewed by Andy Estes.
Follow-up fix. The property exposed by WKWebViewConfiguration should not
use a local BOOL; it simply exposes the APIPageConfiguration value.
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration init]):
(-[WKWebViewConfiguration copyWithZone:]):
(-[WKWebViewConfiguration limitsNavigationToAppBoundDomains]):
(-[WKWebViewConfiguration setLimitsNavigationToAppBoundDomains:]):
(-[WKWebViewConfiguration _setIgnoresAppBoundDomains:]):
2020-04-16 Brent Fulgham <bfulgham@apple.com>
Switch NetworkProcess and NetworkProcessProxy messages to use CallbackID class
https://bugs.webkit.org/show_bug.cgi?id=210566
<rdar://problem/61791784>
Reviewed by David Kilzer.
Instead of using the raw uint64_t type to represent callbacks, use the purpose
built WebKit::CallbackID type.
Tested by existing cases.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::fetchWebsiteData):
(WebKit::NetworkProcess::deleteWebsiteData):
(WebKit::NetworkProcess::deleteWebsiteDataForOrigins):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::fetchWebsiteData):
(WebKit::NetworkProcessProxy::deleteWebsiteData):
(WebKit::NetworkProcessProxy::deleteWebsiteDataForOrigins):
(WebKit::NetworkProcessProxy::didFetchWebsiteData):
(WebKit::NetworkProcessProxy::didDeleteWebsiteData):
(WebKit::NetworkProcessProxy::didDeleteWebsiteDataForOrigins):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/Network/NetworkProcessProxy.messages.in:
2020-04-16 Timothy Horton <timothy_horton@apple.com>
Try to fix the macCatalyst build
* UIProcess/ios/WKActionSheetAssistant.h:
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(uiImageForImage):
(createTargetedPreview):
(createFallbackTargetedPreview):
(-[WKContentView _createTargetedContextMenuHintPreviewIfPossible]):
2020-04-16 David Kilzer <ddkilzer@apple.com>
Re-land: [IPC Hardening] MachMessage::create() should use checked arithmetic
<https://webkit.org/b/210572>
<rdar://problem/61729947>
Reviewed by Chris Dumez.
* Platform/IPC/cocoa/ConnectionCocoa.mm:
(IPC::Connection::sendOutgoingMessage):
- Add nullptr check since MachMessage::create() can return an
empty std::unique_ptr<> now.
* Platform/IPC/cocoa/MachMessage.cpp:
(IPC::MachMessage::create):
- Use CheckedSize to compute the buffer size.
2020-04-16 Brent Fulgham <bfulgham@apple.com>
Add a new 'limitNavigationsToAppBoundDomains' property to WKWebViewConfiguration
https://bugs.webkit.org/show_bug.cgi?id=210620
<rdar://problem/61903225>
Reviewed by Andy Estes.
Provide a way for Developers to opt-in to AppBound domains through a new WKWebViewConfiguration
property.
This patch only creates the property and exposes the flag to the APIPageConfiguration and
WebPageProxy. A separate patch will take action based on its value.
* UIProcess/API/APIPageConfiguration.cpp:
(API::PageConfiguration::copy const):
* UIProcess/API/APIPageConfiguration.h:
(API::PageConfiguration::limitsNavigationToAppBoundDomains const):
(API::PageConfiguration::setLimitsNavigationToAppBoundDomains):
* UIProcess/API/Cocoa/WKWebViewConfiguration.h:
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration init]):
(-[WKWebViewConfiguration copyWithZone:]):
(-[WKWebViewConfiguration limitsNavigationToAppBoundDomains]):
(-[WKWebViewConfiguration setLimitsNavigationToAppBoundDomains:]):
(-[WKWebViewConfiguration _setIgnoresAppBoundDomains:]): Keep the new property in
sync with this internal flag.
* UIProcess/WebPageProxy.cpp:
(WebKit::m_limitsNavigationToAppBoundDomains):
* UIProcess/WebPageProxy.h:
2020-04-16 Sergio Villar Senin <svillar@igalia.com>
Unreviewed build fix for non unified builds.
* NetworkProcess/webrtc/LibWebRTCSocketClient.cpp: Added missing include.
2020-04-16 Daniel Bates <dabates@apple.com>
Remove unused -_focusTextInputContext
https://bugs.webkit.org/show_bug.cgi?id=210619
Reviewed by Darin Adler.
While this function was in the "SPI" header, there is no software at Apple that was using
this funciton and I will assume no non-Apple software was using it either.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _focusTextInputContext:completionHandler:]): Deleted.
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/WebPageProxy.cpp:
* UIProcess/WebPageProxy.h:
2020-04-16 Brent Fulgham <bfulgham@apple.com>
[macOS] Switch unused IOKit classes from allow-with-report to deny-with-report
https://bugs.webkit.org/show_bug.cgi?id=210136
<rdar://problem/61403216>
Reviewed by Per Arne Vollan.
After tracking telemetry for several weeks, we can now safely switch the default state
of a number of IOKit classes from 'allow with telemetry' to 'deny with telemetry'.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-04-16 Daniel Bates <dabates@apple.com>
Fix the iOS build when building without EDITABLE_REGION after r260192
(https://bugs.webkit.org/show_bug.cgi?id=210398)
Have -_mayContainEditableElementsInRect compile-time conditionally call WebKit::mayContainEditableElementsInRect().
If ENABLE(EDITABLE_REGION) is not enabled then return NO.
* UIProcess/API/ios/WKWebViewTestingIOS.mm:
(-[WKWebView _mayContainEditableElementsInRect:]):
2020-04-16 David Kilzer <ddkilzer@apple.com>
Rollout: [IPC Hardening] MachMessage::create() should use checked arithmetic
<https://webkit.org/b/210572>
<rdar://problem/61729947>
Unreviewed rollout of r260160.
Appears to have caused 8 API test failures on iOS.
* Platform/IPC/cocoa/ConnectionCocoa.mm:
(IPC::Connection::sendOutgoingMessage):
* Platform/IPC/cocoa/MachMessage.cpp:
(IPC::MachMessage::create):
2020-04-16 Brent Fulgham <bfulgham@apple.com>
REGRESSION(r260081) Broke iOS PLT due to InjectedBundle initialization (edit)
https://bugs.webkit.org/show_bug.cgi?id=210582
<rdar://problem/61838584>
Reviewed by Darin Adler.
The changes in r260081 began enforcing NSSecureCoding best practices, triggering
a bug in InjectedBundleMac.mm, which is used by iOS as well.
This patch does the following:
1. Delays parameter decoding until we use the appropriate InjectedBundle mechanism
for expressing the valid classes to use in the Unarchiver.
2. Removes a temporary workaround needed to get PLT running on iOS again.
This patch also moves the check of the CFBundleGetFunctionPointerForName slightly earlier
so we can use a single function pointer check, and only invoke the parameter decoding
if we are using the C API.
If not, we delay parameter decoding until we have constructed the WKWebProcessPlugIn
object so we can consume it's 'additionalClassesForParameterCoder' before calling
its initialization method.
* WebProcess/InjectedBundle/mac/InjectedBundleMac.mm:
(WebKit::InjectedBundle::decodeBundleParameters): Remove temporary workaround
needed to get PLT working again. Instead of using a debug assert, return as
a decoding failure if the resulting object is not an NSDictionary.
(WebKit::InjectedBundle::initialize): Delay decoding the bundle parameters on
iOS until the Plugin's additionalClassesForParameterCoder method can be called.
2020-04-16 Daniel Bates <dabates@apple.com>
[iOS] Add a way to focus a text input and place a caret
https://bugs.webkit.org/show_bug.cgi?id=210611
<rdar://problem/61893062>
Reviewed by Darin Adler.
Add some IPI that will be used by code in WebKitAdditions to focus a text input context
and place the caret in it. This will replace the existing -focusTextInput SPI, which I
will remove in a subsequent commit.
* UIProcess/API/ios/WKWebViewPrivateForTestingIOS.h:
* UIProcess/API/ios/WKWebViewTestingIOS.mm:
(-[WKWebView _requestTextInputContextsInRect:completionHandler:]): Fix up code style
of signature while I am here.
(-[WKWebView _focusTextInputContext:placeCaretAt:completionHandler:]): Added.
* UIProcess/WebPageProxy.h:
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _isTextInputContextFocused:]): Added.
(-[WKContentView _focusTextInputContext:placeCaretAt:completionHandler:]): Added.
(-[WKContentView _requestTextInputContextsInRect:completionHandler:]): Fix up code style
of signature while I am here.
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::focusTextInputContextAndPlaceCaret):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::focusTextInputContext): Use auto now that elementForContext() returns a RefPtr.
(WebKit::WebPage::elementForContext const): Have it return a RefPtr instead of a raw
pointer so callers don't have to remember to take out a ref of otherwise be mindful
of the element's lifetime.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::removeTextPlaceholder): Use auto now that elementForContext() returns a RefPtr.
(WebKit::WebPage::requestDocumentEditingContext): Ditto.
(WebKit::WebPage::focusTextInputContextAndPlaceCaret): Added.
2020-04-16 Chris Dumez <cdumez@apple.com>
Use safeRoundPage() instead of round_page() in SharedMemory
https://bugs.webkit.org/show_bug.cgi?id=210608
<rdar://problem/60600443>
Reviewed by Geoffrey Garen.
Use new safeRoundPage() instead of round_page() in SharedMemory, to deal safely with overflows.
* Platform/cocoa/SharedMemoryCocoa.cpp:
(WebKit::SharedMemory::allocate):
(WebKit::makeMemoryEntry):
(WebKit::SharedMemory::~SharedMemory):
(WebKit::SharedMemory::createHandle):
2020-04-16 Adrian Perez de Castro <aperez@igalia.com>
Non-unified build fixes mid April 2020 edition
https://bugs.webkit.org/show_bug.cgi?id=210599
Unreviewed build fix.
* UIProcess/DeviceIdHashSaltStorage.cpp: Add missing Logging.h header.
* UIProcess/Inspector/gtk/WebInspectorProxyGtk.cpp:
(WebKit::WebInspectorProxy::platformSetForcedAppearance): Add missing WebCore:: namespace
prefix.
* UIProcess/linux/MemoryPressureMonitor.h: Add missing wtf/text/CString.h header.
2020-04-16 Daniel Bates <dabates@apple.com>
REGRESSION (r259762): Should always hit test for editable elements if the WKWebView is fully editable
https://bugs.webkit.org/show_bug.cgi?id=210558
<rdar://problem/61798347>
Reviewed by Wenson Hsieh.
If the entire WKWebView is marked editable (via SPI -[WKWebView _setEditable]) then
always ask the web process to hit test the page because every element is considered
editable.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _requestTextInputContextsInRect:completionHandler:]):
2020-04-16 Daniel Bates <dabates@apple.com>
Move -_requestTextInputContextsInRect to WKContentView to simplify implementation
https://bugs.webkit.org/show_bug.cgi?id=210398
<rdar://problem/61656931>
Reviewed by Darin Adler.
-_requestTextInputContextsInRect is not needed on Mac, but supporting it complicates
its implementation: it has to deal with coordinate space differences due to differences
in what is the root view on Mac and iOS + it has to know about iOS's custom content views.
While this function was in the "SPI" header, there is no software at Apple that was using
these outside of WebKit, and I will assume no non-Apple software was using them either.
Moving this function to WKContentView lets me simplify the implementation.
* Platform/spi/ios/UIKitSPI.h: Add more forward declarations.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _convertRectFromRootViewCoordinates:]): Deleted.
(-[WKWebView _convertRectToRootViewCoordinates:]): Deleted.
(-[WKWebView _mayContainEditableElementsInRect:]): Deleted; moved to WKWebViewTestingIOS.mm.
(-[WKWebView _requestTextInputContextsInRect:completionHandler:]): Deleted; moved to WKContentViewInteraction.mm.
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/API/ios/WKWebViewPrivateForTestingIOS.h:
* UIProcess/API/ios/WKWebViewTestingIOS.mm:
(-[WKWebView _requestTextInputContextsInRect:completionHandler:]): Added. Turns around and
calls the function of the same name on the WKContentView. It takes care to convert the specified
rect from WKWebView coordinates to WKContentView coordinates and fixes up the bounding rects
for the returned contexts to be in WKWebView coordinates. This keeps existings tests passing
as mekes using this function intuitive since callers specify and get rects in WKWebView coordinates.
(-[WKWebView _requestDocumentContext:completionHandler:]): Added.
(-[WKWebView _adjustSelectionWithDelta:completionHandler:]): Added.
(-[WKWebView _mayContainEditableElementsInRect:]): Moved from WKWebView.mm.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _requestTextInputContextsInRect:completionHandler:]): Moved from WKWebView.mm and simplified.
2020-04-16 David Kilzer <ddkilzer@apple.com>
[IPC Hardening] Use ObjectIdentifier<> for LegacyCustomProtocol
<https://webkit.org/b/210580>
<rdar://problem/61791686>
Reviewed by Chris Dumez.
Replace uses of uint64_t with LegacyCustomProtocolID.
* NetworkProcess/CustomProtocols/Cocoa/LegacyCustomProtocolManagerCocoa.mm:
(WebKit::LegacyCustomProtocolManager::didFailWithError):
(WebKit::LegacyCustomProtocolManager::didLoadData):
(WebKit::LegacyCustomProtocolManager::didReceiveResponse):
(WebKit::LegacyCustomProtocolManager::didFinishLoading):
(WebKit::LegacyCustomProtocolManager::wasRedirectedToRequest):
(WebKit::LegacyCustomProtocolManager::protocolForID):
* NetworkProcess/CustomProtocols/LegacyCustomProtocolID.h: Add.
* NetworkProcess/CustomProtocols/LegacyCustomProtocolManager.cpp:
(WebKit::LegacyCustomProtocolManager::addCustomProtocol):
(WebKit::LegacyCustomProtocolManager::removeCustomProtocol):
(WebKit::LegacyCustomProtocolManager::startLoading):
(WebKit::LegacyCustomProtocolManager::stopLoading):
(WebKit::generateCustomProtocolID): Delete.
- Replace with LegacyCustomProtocolID::generate().
* NetworkProcess/CustomProtocols/LegacyCustomProtocolManager.h:
* NetworkProcess/CustomProtocols/LegacyCustomProtocolManager.messages.in:
* Scripts/webkit/messages.py:
* UIProcess/API/APICustomProtocolManagerClient.h:
(API::CustomProtocolManagerClient::startLoading):
(API::CustomProtocolManagerClient::stopLoading):
* UIProcess/Cocoa/LegacyCustomProtocolManagerClient.h:
* UIProcess/Cocoa/LegacyCustomProtocolManagerClient.mm:
(-[WKCustomProtocolLoader initWithLegacyCustomProtocolManagerProxy:customProtocolID:request:]):
(WebKit::LegacyCustomProtocolManagerClient::startLoading):
(WebKit::LegacyCustomProtocolManagerClient::stopLoading):
* UIProcess/Network/CustomProtocols/LegacyCustomProtocolManagerProxy.cpp:
(WebKit::LegacyCustomProtocolManagerProxy::startLoading):
(WebKit::LegacyCustomProtocolManagerProxy::stopLoading):
(WebKit::LegacyCustomProtocolManagerProxy::wasRedirectedToRequest):
(WebKit::LegacyCustomProtocolManagerProxy::didReceiveResponse):
(WebKit::LegacyCustomProtocolManagerProxy::didLoadData):
(WebKit::LegacyCustomProtocolManagerProxy::didFailWithError):
(WebKit::LegacyCustomProtocolManagerProxy::didFinishLoading):
* UIProcess/Network/CustomProtocols/LegacyCustomProtocolManagerProxy.h:
- Remove unused typedef and instance variable.
* UIProcess/Network/CustomProtocols/LegacyCustomProtocolManagerProxy.messages.in:
* WebKit.xcodeproj/project.pbxproj:
- Add LegacyCustomProtocolID.h to project.
2020-04-16 Tim Horton <timothy_horton@apple.com>
REGRESSION (r259898): WebKit-based Books views are all blank
https://bugs.webkit.org/show_bug.cgi?id=210590
<rdar://problem/61791109>
Reviewed by Chris Dumez.
* UIProcess/Cocoa/WebViewImpl.h:
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::enterAcceleratedCompositingWithRootLayer):
(WebKit::WebViewImpl::setAcceleratedCompositingRootLayer):
(WebKit::WebViewImpl::setAcceleratedCompositingRootLayerAfterFlush): Deleted.
* UIProcess/mac/PageClientImplMac.mm:
(WebKit::PageClientImpl::enterAcceleratedCompositingMode):
(WebKit::PageClientImpl::didFirstLayerFlush):
The changes to setAcceleratedCompositingRootLayer in r259898 proved
to be wrong in a second way (the first being fixed in r260104): when
setAcceleratedCompositingRootLayer is called from updateAcceleratedCompositingMode,
because the layer hosting mode changed (Books appears to use app-hosted layers),
it incorrectly identified the root layer change as a process swap, resulting
in the correct layer never being unhidden.
This is enough mistakes that I'm going to try a different approach:
put setAcceleratedCompositingRootLayer back to the way it was before, where
it immediately updates the layer without any smarts, remove
setAcceleratedCompositingRootLayerAfterFlush, because it's no longer necessary,
and add enterAcceleratedCompositingWithRootLayer, which is specifically only
called in the case where DrawingArea will for-sure send us a follow-up
(didFirstLayerFlush) that will unhide the root layer.
2020-04-16 Eric Carlson <eric.carlson@apple.com>
[macOS] Update ScreenTime as playback state changes
https://bugs.webkit.org/show_bug.cgi?id=210518
<rdar://problem/61181092>
Reviewed by Jer Noble.
Add a manager that gets media element state changes, and a Cocoa-specific
implementation that passes that state to ScreenTime.
* Scripts/webkit/messages.py:
* Sources.txt:
* SourcesCocoa.txt:
* UIProcess/Cocoa/WebPageProxyCocoa.mm:
(WebKit::WebPageProxy::mediaUsageManager):
(WebKit::WebPageProxy::addMediaUsageManagerSession):
(WebKit::WebPageProxy::updateMediaUsageManagerSessionState):
(WebKit::WebPageProxy::removeMediaUsageManagerSession):
* UIProcess/Media/MediaUsageManager.cpp: Added.
(WebKit::MediaUsageManager::create):
(WebKit::reset):
(WebKit::MediaUsageManager::addMediaSession):
(WebKit::MediaUsageManager::removeMediaSession):
(WebKit::MediaUsageManager::updateMediaUsage):
* UIProcess/Media/MediaUsageManager.h: Added.
* UIProcess/Media/cocoa/MediaUsageManagerCocoa.h: Added.
(WebKit::MediaUsageManagerCocoa::SessionMediaUsage::SessionMediaUsage):
* UIProcess/Media/cocoa/MediaUsageManagerCocoa.mm: Added.
(WebKit::usageTrackingAvailable):
(WebKit::MediaUsageManager::create):
(WebKit::MediaUsageManagerCocoa::~MediaUsageManagerCocoa):
(WebKit::MediaUsageManagerCocoa::reset):
(WebKit::MediaUsageManagerCocoa::addMediaSession):
(WebKit::MediaUsageManagerCocoa::removeMediaSession):
(WebKit::MediaUsageManagerCocoa::updateMediaUsage):
* UIProcess/WebAuthentication/AuthenticatorManager.cpp:
* UIProcess/WebAuthentication/Mock/MockLocalConnection.mm:
* UIProcess/WebPageProxy.cpp:
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::addMediaUsageManagerSession):
(WebKit::WebChromeClient::updateMediaUsageManagerSessionState):
(WebKit::WebChromeClient::removeMediaUsageManagerSession):
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::addMediaUsageManagerSession):
(WebKit::WebPage::updateMediaUsageManagerSessionState):
(WebKit::WebPage::removeMediaUsageManagerSession):
* WebProcess/WebPage/WebPage.h:
2020-04-16 David Kilzer <ddkilzer@apple.com>
[IPC hardening] DrawingArea::updateBackingStoreState() is not used on Apple platforms
<https://webkit.org/b/210585>
<rdar://problem/61802011>
Reviewed by Geoffrey Garen.
* UIProcess/DrawingAreaProxy.h:
* UIProcess/DrawingAreaProxy.messages.in:
* WebProcess/WebPage/DrawingArea.h:
(WebKit::DrawingArea::updateBackingStoreState):
* WebProcess/WebPage/DrawingArea.messages.in:
- Use USE(COORDINATED_GRAPHICS) || USE(TEXTURE_MAPPER) to
protect code not used on Apple platforms.
2020-04-16 Carlos Alberto Lopez Perez <clopez@igalia.com>
[GTK] MiniBrowser opens new windows too small causing failures on some WPT tests
https://bugs.webkit.org/show_bug.cgi?id=210206
Reviewed by Carlos Garcia Campos.
Provide the size of the default window (if is set) when windowRect is called
before the window is shown.
* UIProcess/API/glib/WebKitUIClient.cpp:
2020-04-16 Tomoki Imai <Tomoki.Imai@sony.com>
TextureMapper renders video element with "object-fit: cover" incorrectly
https://bugs.webkit.org/show_bug.cgi?id=210544
Reviewed by Žan Doberšek.
Propagate GraphicsLayer::contentsClippingRect information to TextureMapperLayer
to properly clip the outside of DOM element when the element has "object-fit: cover".
Unfortunately, the test is disabled on WebKitGTK due to bug 177536, bug 163528.
Test: compositing/video/video-object-fit.html
* Shared/CoordinatedGraphics/CoordinatedGraphicsScene.cpp:
(WebKit::CoordinatedGraphicsScene::updateSceneState):
2020-04-15 Jer Noble <jer.noble@apple.com>
REGRESSION (r260102): ASSERTION FAILED: m_arbitrators.contains(proxy) in WebKit::SharedArbitrator::endRoutingArbitrationForArbitrator
https://bugs.webkit.org/show_bug.cgi?id=210589
<rdar://problem/61844208>
Reviewed by Eric Carlson.
Protect against an assertion in the SharedArbitrator in case where a WebProcess is terminated before it can call
endRoutingArbitration().
* UIProcess/Media/AudioSessionRoutingArbitratorProxy.cpp:
(WebKit::AudioSessionRoutingArbitratorProxy::processDidTerminate):
* UIProcess/Media/AudioSessionRoutingArbitratorProxy.h:
* UIProcess/Media/cocoa/AudioSessionRoutingArbitratorProxyCocoa.mm:
(WebKit::SharedArbitrator::isInRoutingArbitrationForArbitrator):
(WebKit::SharedArbitrator::beginRoutingArbitrationForArbitrator):
(WebKit::SharedArbitrator::endRoutingArbitrationForArbitrator):
(WebKit::AudioSessionRoutingArbitratorProxy::processDidTerminate):
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::processDidTerminateOrFailedToLaunch):
2020-04-15 Tim Horton <timothy_horton@apple.com>
REGRESSION (r258337): Crash when right clicking on link that uses the system UI font with optimizeLegibility on Mojave
https://bugs.webkit.org/show_bug.cgi?id=210575
<rdar://problem/61646717>
Reviewed by Darin Adler.
* Shared/Cocoa/ArgumentCodersCocoa.mm:
(-[WKSecureCodingFontAttributeNormalizer archiver:willEncodeObject:]):
(IPC::encodeSecureCodingInternal):
The workaround introduced in r250640 and r251086 for archiving NSFonts
with optical sizing attributes is still incomplete, which was recently
revealed by making decoding failures more apparent (in r258337).
Specifically, the workaround was not applied to fonts encoded inside
NSAttributedStrings, which are encoded using a keyed archiver... only
to fonts encoded directly with CoreIPC.
Use a NSKeyedArchiver delegate to apply our fix-up to the optical sizing
attribute when any NSFont is being encoded, and only on platforms
that still exhibit the bug.
* Shared/Cocoa/CoreTextHelpers.h:
* Shared/Cocoa/CoreTextHelpers.mm:
(WebKit::fontWithAttributes):
(WebKit::fontDescriptorWithFontAttributes): Deleted.
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::updateFontManagerIfNeeded):
* UIProcess/mac/WebPopupMenuProxyMac.mm:
(WebKit::WebPopupMenuProxyMac::showPopupMenu):
Refactor fontDescriptorWithFontAttributes to just return a font since that's what all callers actually want.
2020-04-15 John Wilander <wilander@apple.com>
Add SPI to configure WebsiteDataStores with a URL for standalone web applications and use it to disable first-party website data removal in ITP
https://bugs.webkit.org/show_bug.cgi?id=209634
<rdar://problem/60943970>
Reviewed by Alex Christensen.
This change adds a new property to _WKWebsiteDataStoreConfiguration.h called
standaloneApplicationURL with which the hosting application can inform the
website data store that it's running as a standalone web application.
This change also forwards an existing standaloneApplicationURL as a
WebCore::RegistrableDomain into ITP so that explicit exemptions can be made
to first parties of standalone web applications. The exemptions made here
all for all of ITP's website data removal. This part of the change is
covered by the new layout tests.
Tests: http/tests/resourceLoadStatistics/standalone-web-application-exempt-from-website-data-deletion-database.html
http/tests/resourceLoadStatistics/standalone-web-application-exempt-from-website-data-deletion.html
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:
(WebKit::ResourceLoadStatisticsMemoryStore::registrableDomainsToDeleteOrRestrictWebsiteDataFor):
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.h:
(WebKit::ResourceLoadStatisticsStore::setStandaloneApplicationDomain):
(WebKit::ResourceLoadStatisticsStore::standaloneApplicationDomain const):
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::setStandaloneApplicationDomain):
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.h:
* NetworkProcess/NetworkSession.cpp:
(WebKit::NetworkSession::NetworkSession):
(WebKit::NetworkSession::forwardResourceLoadStatisticsSettings):
* NetworkProcess/NetworkSession.h:
* Shared/ResourceLoadStatisticsParameters.h:
(WebKit::ResourceLoadStatisticsParameters::encode const):
(WebKit::ResourceLoadStatisticsParameters::decode):
* UIProcess/API/Cocoa/_WKWebsiteDataStoreConfiguration.h:
* UIProcess/API/Cocoa/_WKWebsiteDataStoreConfiguration.mm:
(-[_WKWebsiteDataStoreConfiguration standaloneApplicationURL]):
(-[_WKWebsiteDataStoreConfiguration setStandaloneApplication:]):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::ensureNetworkProcess):
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::WebsiteDataStore::platformSetNetworkParameters):
* UIProcess/WebsiteData/WebsiteDataStoreConfiguration.cpp:
(WebKit::WebsiteDataStoreConfiguration::copy const):
* UIProcess/WebsiteData/WebsiteDataStoreConfiguration.h:
(WebKit::WebsiteDataStoreConfiguration::standaloneApplicationURL const):
(WebKit::WebsiteDataStoreConfiguration::setStandaloneApplicationURL):
2020-04-15 Chris Dumez <cdumez@apple.com>
[IPC Hardening] MachMessage::messageSize() should use checked arithmetic
https://bugs.webkit.org/show_bug.cgi?id=210567
<rdar://problem/61734355>
Reviewed by Geoffrey Garen.
Replace call to round_msg() call with a call to our new safeRoundMsg() which does
the same alignment up but using checked arithmetic to protect against overflows.
* Platform/IPC/cocoa/MachMessage.cpp:
(IPC::safeRoundMsg):
(IPC::MachMessage::messageSize):
2020-04-15 David Kilzer <ddkilzer@apple.com>
[IPC Hardening] MachMessage::create() should use checked arithmetic
<https://webkit.org/b/210572>
<rdar://problem/61729947>
Reviewed by Chris Dumez.
* Platform/IPC/cocoa/ConnectionCocoa.mm:
(IPC::Connection::sendOutgoingMessage):
- Add nullptr check since MachMessage::create() can return an
empty std::unique_ptr<> now.
* Platform/IPC/cocoa/MachMessage.cpp:
(IPC::MachMessage::create):
- Use CheckedSize to compute the buffer size.
2020-04-15 Brent Fulgham <bfulgham@apple.com>
Unreviewed PLT fix after r260081.
* WebProcess/InjectedBundle/mac/InjectedBundleMac.mm:
(WebKit::InjectedBundle::decodeBundleParameters):
2020-04-15 Ben Nham <nham@apple.com>
DidFirstLayerFlush IPC can be head-of-line-blocked and delay first paint by 200+ ms on process swap
https://bugs.webkit.org/show_bug.cgi?id=210571
Reviewed by Geoffrey Garen.
When we process swap, the layer swap/unhiding for the new hosted layer is gated by the
DidFirstLayerFlush IPC. But actually sending this IPC can be head-of-line blocked for an
arbitrarily long time because we append it to the queue of messages for the main run loop to
process. In some cases, e.g. for the NYT PLT content, this blocks first paint by 200 ms
while the WebProcess main thread is busy doing other things (like executing deferred
scripts).
To fix this, use a CATransaction commit handler to send the IPC right after we commit.
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::sendDidFirstLayerFlushIfNeeded):
2020-04-15 Wenson Hsieh <wenson_hsieh@apple.com>
Unreviewed, fix the watchOS and tvOS builds after r260146
Move some code for controlling the presentation of context menus on iOS behind USE(UICONTEXTMENU).
* UIProcess/ios/WKActionSheetAssistant.h:
* UIProcess/ios/WKContentViewInteraction.mm:
2020-04-15 Peng Liu <peng.liu6@apple.com>
Video elements don't return to the correct position when exiting fullscreen
https://bugs.webkit.org/show_bug.cgi?id=210529
Reviewed by Jer Noble.
Some web pages use the "resize" event handler to calculate the element size after
they exit fullscreen, and the calculation is based on the container element size which
might be affected by the fullscreen mode.
We need to call WebPageProxy::setNeedsDOMWindowResizeEvent() to fire the "resize" event
in the repaint callback after the exiting fullscreen process is completed and the
possible layout change due to exiting fullscreen is done. Otherwise the size calculation
might be wrong.
* UIProcess/WebPageProxy.cpp:
* UIProcess/WebPageProxy.h:
* UIProcess/ios/fullscreen/WKFullScreenWindowControllerIOS.mm:
(-[WKFullScreenWindowController _completedExitFullScreen]):
* UIProcess/mac/WKFullScreenWindowController.mm:
(-[WKFullScreenWindowController completeFinishExitFullScreenAnimationAfterRepaint]):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::setNeedsDOMWindowResizeEvent):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
2020-04-15 Wenson Hsieh <wenson_hsieh@apple.com>
[iPadOS] Some pages indefinitely zoom in and out due to idempotent text autosizing
https://bugs.webkit.org/show_bug.cgi?id=210551
<rdar://problem/56820674>
Reviewed by Tim Horton.
On a non-responsive web page with one or lines of non-wrapping text of a specific length (just under 1920px),
it's possible for the combination of idempotent text autosizing heuristics and viewport shrink-to-fit to cause
the single line of text to grow and shrink in size indefinitely, and additionally cause the initial scale to
thrash between multiple values indefinitely. This manifests in the entire page repeatedly zooming in and out
immediately after page load.
Consider the following scenario:
(1) A viewport configuration change (e.g. due to parsing the viewport meta tag) schedules the timer to reset
idempotent text autosizing. Let's suppose the page has a really long line of non-wrapping 12px text that is
below 1920px wide.
(2) The timer fires, invalidating styles and recomputing text autosizing given the current initial scale. The
current initial scale is below 1, since ViewportConfiguration will attempt to shrink to fit the page to
avoid horizontal scrolling. This causes text autosizing to boost the long line of text to a larger value
(let's say 17px).
(3) The next time we perform style recomputation and layout, we discover that the content width of the page is
now larger than 1920px, which is the maximum width which we'll attempt to shrink to fit; when computing
initial scale, we give up trying to shrink down to avoid making the inital scale too small, and instead just
keep it at 1.
(4) This change in viewport configuration then schedules another idempotent text autosizing reset. When this
timer fires, it sees that the initial scale is now 1, which means that the text is no longer boosted, so we
make the single line of text small again (12px).
(5) After the next style recomputation and layout, this causes the content width of the page to dip below the
1920px threshold, causing the initial scale to dip below 1 again. As detailed above, this schedules another
idempotent text autosizing update, which now boosts font size once again, and the cycle continues.
To fix this, instead of consulting the initial scale (`ViewportConfiguration::initialScale()`) when computing
the boosted font size for idempotent text autosizing, we can instead ask for the initial scale ignoring content
size (`ViewportConfiguration::initialScaleIgnoringContentSize()`). This prevents changes in content size due to
idempotent autosizing from affecting the idempotent autosizing heuristic (through the different initial scale),
and ensures that this method of text autosizing actually remains idempotent.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::resetIdempotentTextAutosizingIfNeeded):
(WebKit::WebPage::viewportConfigurationChanged):
2020-04-14 Megan Gardner <megan_gardner@apple.com>
Data Detected Actions sheets are presented from odd locations.
https://bugs.webkit.org/show_bug.cgi?id=210531
<rdar://problem/60941346>
Reviewed by Wenson Hsieh.
Use the existing context menu targeted preview code to ensure that the
data detected menus are presented in a good location.
* UIProcess/ios/WKActionSheetAssistant.h:
* UIProcess/ios/WKActionSheetAssistant.mm:
(-[WKActionSheetAssistant contextMenuInteraction:previewForHighlightingMenuWithConfiguration:]):
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView contextMenuInteraction:willEndForConfiguration:animator:]):
2020-04-15 Per Arne Vollan <pvollan@apple.com>
[iOS] Potential crash under WebProcess::platformInitializeWebProcess
https://bugs.webkit.org/show_bug.cgi?id=210547
Reviewed by Brent Fulgham.
If the UI process fails to issue an extension to the runningboard service, the WebContent process will crash
while trying to consume this extension.
No new tests, since the test runners are unsandboxed, so issuing the extension will always succeed.
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-04-15 Brent Fulgham <bfulgham@apple.com>
[Cocoa] Remove generic parameter serialization fallback
https://bugs.webkit.org/show_bug.cgi?id=210509
<rdar://problem/55522650>
Reviewed by Geoffrey Garen.
Tested by BundleFormDelegate parameter serialization tests...
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _setInputDelegate:]):
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::elementDidFocus):
2020-04-15 Chris Dumez <cdumez@apple.com>
REGRESSION (r259610): WebGL does not work at all on iOS (was: Google Maps tiles turn black after initial load)
<rdar://problem/61794480>
Unreviewed, partial revert of r259610 because it broke Google Maps.
Stop using RunningBoard for the "Foreground" assertion because it fails to
give the target GPU access.
* UIProcess/ios/ProcessAssertionIOS.mm:
(WebKit::runningBoardNameForAssertionType):
2020-04-15 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK4] Fix use of gtk init functions
https://bugs.webkit.org/show_bug.cgi?id=210550
Reviewed by Adrian Perez de Castro.
Include GtkVersioning.h instead of gtk.h
* WebProcess/gtk/WebProcessMainGtk.cpp:
2020-04-15 Peng Liu <peng.liu6@apple.com>
Adopt interface AVAudioRoutingArbiter for Mac
https://bugs.webkit.org/show_bug.cgi?id=210167
Reviewed by Jer Noble.
A follow-up patch to fix a build error.
* UIProcess/Media/cocoa/AudioSessionRoutingArbitratorProxyCocoa.mm:
(WebKit::SharedArbitrator::beginRoutingArbitrationForArbitrator):
2020-04-15 Adrian Perez de Castro <aperez@igalia.com>
[GTK4] Provide an alternative to gtk_widget_{get,is}_toplevel()
https://bugs.webkit.org/show_bug.cgi?id=210463
Reviewed by Carlos Garcia Campos.
* UIProcess/API/glib/WebKitUIClient.cpp: Adapt to take into account that GTK4 does not
provide the GtkWidget.configure-event signal.
2020-04-15 Said Abou-Hallawa <sabouhallawa@apple.com>
Unreviewed, reverting r259891.
Causes significant iOS MotionMark regression
Reverted changeset:
"Switch unused IOKit classes from allow-with-report to deny-
with-report"
https://bugs.webkit.org/show_bug.cgi?id=210136
https://trac.webkit.org/changeset/259891
2020-04-14 Megan Gardner <megan_gardner@apple.com>
File Upload Menus presented from odd locations.
https://bugs.webkit.org/show_bug.cgi?id=210523
<rdar://problem/60390846>
Use the existing context menu targeted preview code to ensure that the
file menus are presented in a good location.
Reviewed by Wenson Hsieh.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/forms/WKFileUploadPanel.mm:
(-[WKFileUploadPanel contextMenuInteraction:previewForHighlightingMenuWithConfiguration:]):
2020-04-14 David Kilzer <ddkilzer@apple.com>
dictionaryValueOfType() in WebCoreArgumentCodersMac.mm can be replaced with dynamic_cf_cast<>()
<https://webkit.org/b/210456>
Reviewed by Darin Adler.
* Shared/mac/WebCoreArgumentCodersMac.mm:
(IPC::dictionaryValueOfType): Delete.
(IPC::extractDictionaryValue): Add.
- Use dynamic_cf_cast<>() in place of manually checking the
CFTypeID of each object.
(IPC::createArchiveList):
- Call new extractDictionaryValue() template function to verify
values are the correct types in the dictionary and to set the
output variables.
2020-04-14 David Kilzer <ddkilzer@apple.com>
Use CFArrayGetValues() in createArchiveList() in WebCoreArgumentCodersMac.mm
<https://webkit.org/b/210519>
Reviewed by Darin Adler.
* Shared/mac/WebCoreArgumentCodersMac.mm:
(IPC::createArchiveList):
- Use CFArrayGetValues() to copy array more efficiently.
2020-04-14 Tim Horton <timothy_horton@apple.com>
REGRESSION (r259898): Tabs occasionally go blank and get stuck that way
https://bugs.webkit.org/show_bug.cgi?id=210520
<rdar://problem/61776830>
Reviewed by Simon Fraser.
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::setAcceleratedCompositingRootLayer):
setThumbnailView calls setAcceleratedCompositingRootLayer in order to reconnect the
existing root layer. After r259898, this would cause us to go down the
process-swap path, and hide the root layer, and nothing would ever
come along and fix it.
If setAcceleratedCompositingRootLayer is called with the existing root
layer, that cannot be a process swap, so ignore it and proceed as before.
2020-04-14 Jer Noble <jer.noble@apple.com>
Adopt interface AVAudioRoutingArbiter for Mac
https://bugs.webkit.org/show_bug.cgi?id=210167
<rdar://problem/59113994>
Reviewed by Eric Carlson.
Add a new cross-process object pair, AudioSessionRouterAbitrator/Proxy which passes routing
arbitration commands from WebContent -> UIProcess. Because the UIProcess can represent
multiple WebContent processes, it must track all outstanding arbitration requests globally.
* DerivedSources-input.xcfilelist:
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* GPUProcess/GPUProcess.cpp:
(WebKit::GPUProcess::initializeGPUProcess):
* Sources.txt:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::mediaRelatedMachServices):
* UIProcess/Media/AudioSessionRoutingArbitratorProxy.cpp: Added.
(WebKit::AudioSessionRoutingArbitratorProxy::AudioSessionRoutingArbitratorProxy):
(WebKit::AudioSessionRoutingArbitratorProxy::~AudioSessionRoutingArbitratorProxy):
(WebKit::AudioSessionRoutingArbitratorProxy::beginRoutingArbitrationWithCategory):
(WebKit::AudioSessionRoutingArbitratorProxy::leaveRoutingArbitration):
* UIProcess/Media/AudioSessionRoutingArbitratorProxy.h: Added.
(WebKit::AudioSessionRoutingArbitratorProxy::category const):
(WebKit::AudioSessionRoutingArbitratorProxy::destinationId):
* UIProcess/Media/AudioSessionRoutingArbitratorProxy.messages.in: Added.
* UIProcess/Media/cocoa/AudioSessionRoutingArbitratorProxyCocoa.mm: Added.
(WebKit::SharedArbitrator::sharedInstance):
(WebKit::SharedArbitrator::beginRoutingArbitrationForArbitrator):
(WebKit::SharedArbitrator::leaveRoutingArbitrationForArbitrator):
(WebKit::AudioSessionRoutingArbitratorProxy::AudioSessionRoutingArbitratorProxy):
(WebKit::AudioSessionRoutingArbitratorProxy::~AudioSessionRoutingArbitratorProxy):
(WebKit::AudioSessionRoutingArbitratorProxy::beginRoutingArbitrationWithCategory):
(WebKit::AudioSessionRoutingArbitratorProxy::leaveRoutingArbitration):
* UIProcess/WebPageProxy.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::WebProcessProxy):
* UIProcess/WebProcessProxy.h:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_processDisplayName):
* WebProcess/WebProcess.cpp:
* WebProcess/cocoa/AudioSessionRoutingArbitrator.cpp: Added.
(WebKit::AudioSessionRoutingArbitrator::AudioSessionRoutingArbitrator):
(WebKit::AudioSessionRoutingArbitrator::supplementName):
(WebKit::AudioSessionRoutingArbitrator::beginRoutingArbitrationWithCategory):
(WebKit::AudioSessionRoutingArbitrator::leaveRoutingAbritration):
* WebProcess/cocoa/AudioSessionRoutingArbitrator.h: Added.
* WebProcess/cocoa/UserMediaCaptureManager.cpp:
(WebKit::UserMediaCaptureManager::AudioFactory::createAudioCaptureSource):
* WebProcess/com.apple.WebProcess.sb.in:
2020-04-14 David Kilzer <ddkilzer@apple.com>
createArchiveList() in WebCoreArgumentCodersMac.mm should do more validity checks
<https://webkit.org/b/210448>
<rdar://problem/61677029>
Reviewed by Darin Adler.
* Shared/mac/WebCoreArgumentCodersMac.mm:
(IPC::createArchiveList): Add more validity checks.
2020-04-14 Brent Fulgham <bfulgham@apple.com>
Do more checking before reusing precompiled sandbox
https://bugs.webkit.org/show_bug.cgi?id=210304
<rdar://problem/61155623>
Reviewed by Darin Adler.
We recently discovered that the system sandbox framework version does not always change
when breaking changes in the sandbox format are made. This can lead to the precompiled
version of a sandbox not working properly when a sandbox version changes.
To guard against this, make the following changes:
1. Compare the SANDBOX_BUILD_ID in place when the sandbox was compiled to the value on
the current system. SANDBOX_BUILD_ID, which is a GUID, is always regenerated when the
sandbox framework is rebuilt.
2. Compare the OS version in place when the sandbox was compiled to the value on the current
system. This will trigger us recompiling sandboxes even when the sandbox framework did not
change, but this is a small performance cost that would only happen after a software update.
* Shared/mac/AuxiliaryProcessMac.mm:
(WebKit::compileAndCacheSandboxProfile):
(WebKit::tryApplyCachedSandbox):
2020-04-14 Per Arne Vollan <pvollan@apple.com>
[iOS] Content filtering needs access to the frontboard service
https://bugs.webkit.org/show_bug.cgi?id=210505
Reviewed by Geoffrey Garen.
Content filtering in the WebContent process needs access to the frontboard service on iOS. Issue an extension
to this service when content filtering is enabled.
API test: WebKit.WebFilterFeatureHasFrontboardServiceAccess
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-04-14 Wenson Hsieh <wenson_hsieh@apple.com>
[iPadOS] Wikipedia articles lay out incorrectly in 1/3 multitasking window
https://bugs.webkit.org/show_bug.cgi?id=210501
<rdar://problem/54856323>
Reviewed by Tim Horton.
Apply the viewport quirk if needed; see WebCore/ChangeLog for more details.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::updateVisibleContentRects):
2020-04-14 Brent Fulgham <bfulgham@apple.com>
InjectedBundle parameters often need initialization function called before unarchiving
https://bugs.webkit.org/show_bug.cgi?id=189709
<rdar://problem/44573653>
Reviewed by Ryosuke Niwa.
Handle the case where the InjectedBundle parameters do not successfully decode because they contain
an unexpected class from the embedding program. If this happens, try decoding the bundle parameters
after the bundle initialiation function runs, which gives the embedding program the opportunity to
register additional classes that are safe for serialization.
Extend WKWebProcessPlugIn with a method that returns the names of any custom classes that need
to be serialized by the InjectedBundle.
Create a new 'decodeBundleParameters' method that contains the logic that used to live in 'initialize'.
Revise 'initialize' to call this new method.
* WebProcess/InjectedBundle/InjectedBundle.h:
* WebProcess/InjectedBundle/mac/InjectedBundleMac.mm:
(WebKit::InjectedBundle::initialize): Use the new method.
(WebKit::InjectedBundle::decodeBundleParameters): Added.
(WebKit::InjectedBundle::setBundleParameters): Use 'decodeObjectOfClasses' with the more complete
'classesForCoder' method to unarchive the passed bundle parameters, rather than the
NSDictionary-specific method, since InjectedBundles often encode other types of objects, and the
NSDictionary object may itself hold other kinds of objects.
* WebProcess/InjectedBundle/API/mac/WKWebProcessPlugIn.h:
(WebKit::WKWebProcessPlugIn::additionalClassesForParameterCoder): Added.
2020-04-14 Wenson Hsieh <wenson_hsieh@apple.com>
Address review comments after r260035
https://bugs.webkit.org/show_bug.cgi?id=210440
Reviewed by David Kilzer.
Omit unused argument names.
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetIgnoreViewportScalingConstraints):
(WKPreferencesGetIgnoreViewportScalingConstraints):
2020-04-14 Megan Gardner <megan_gardner@apple.com>
YouTube sometimes navigates unexpectedly when invoking a context menu on a video link.
https://bugs.webkit.org/show_bug.cgi?id=210467
<rdar://problem/58841584>
Reviewed by Tim Horton.
Cancel any outstanding touches tracked by WebTouchEventsGestureRecognizer, so that we dispatch
`touchcancel` to the page instead of `touchend`, so that the page does not misinterpret
context menu invocation as a tap.
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView contextMenuInteraction:previewForHighlightingMenuWithConfiguration:]):
2020-04-14 Youenn Fablet <youenn@apple.com>
Do not process RTC Network messages coming from NetworkProcess if LibWebRTCNetwork is not active
https://bugs.webkit.org/show_bug.cgi?id=207376
<rdar://problem/59238611>
Unreviewed.
Post-commit review update, change variable name to network.
No change of behavior.
* WebProcess/Network/NetworkProcessConnection.cpp:
(WebKit::NetworkProcessConnection::didReceiveMessage):
2020-04-14 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] Bring back support for rendering scrollbars using the system appearance
https://bugs.webkit.org/show_bug.cgi?id=209805
Reviewed by Michael Catanzaro.
Add WebKitWebContext:use-system-appearance-for-scrollbars property. It's enabled by default to keep backwards
compatibility.
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const): Encode useSystemAppearanceForScrollbars.
(WebKit::WebProcessCreationParameters::decode): Decode useSystemAppearanceForScrollbars.
* Shared/WebProcessCreationParameters.h:
* UIProcess/API/APIProcessPoolConfiguration.cpp:
(API::ProcessPoolConfiguration::copy): Copy m_useSystemAppearanceForScrollbars;
* UIProcess/API/APIProcessPoolConfiguration.h:
* UIProcess/API/glib/WebKitWebContext.cpp:
(webkitWebContextGetProperty):
(webkitWebContextSetProperty):
(webkitWebContextConstructed):
(webkit_web_context_class_init):
(webkit_web_context_set_use_system_appearance_for_scrollbars):
(webkit_web_context_get_use_system_appearance_for_scrollbars):
* UIProcess/API/gtk/WebKitWebContext.h:
* UIProcess/API/gtk/docs/webkit2gtk-4.0-sections.txt:
* UIProcess/glib/WebProcessPoolGLib.cpp:
(WebKit::WebProcessPool::platformInitializeWebProcess): Initialize useSystemAppearanceForScrollbars parameter.
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:
* WebProcess/glib/WebProcessGLib.cpp:
(WebKit::WebProcess::platformInitializeWebProcess): Call setUseSystemAppearanceForScrollbars to set
useSystemAppearanceForScrollbars parameter.
(WebKit::WebProcess::setUseSystemAppearanceForScrollbars): Call ScrollbarThemeGtk::setUseSystemAppearance().
2020-04-14 Youenn Fablet <youenn@apple.com>
WebSocketChannel should remove itself from its manager map
https://bugs.webkit.org/show_bug.cgi?id=210424
Reviewed by Alex Christensen.
WebSocketChannelManager was never removing any entry from its map.
To fix this, the manager is now keeping a WeakPtr to each channel.
When the channel is destroyed, it will remove itself from its channel manager.
* WebProcess/Network/WebSocketChannel.cpp:
(WebKit::WebSocketChannel::WebSocketChannel):
(WebKit::WebSocketChannel::~WebSocketChannel):
* WebProcess/Network/WebSocketChannel.h:
* WebProcess/Network/WebSocketChannelManager.cpp:
(WebKit::WebSocketChannelManager::addChannel):
(WebKit::WebSocketChannelManager::createWebSocketChannel): Deleted.
* WebProcess/Network/WebSocketChannelManager.h:
(WebKit::WebSocketChannelManager::removeChannel):
* WebProcess/Network/WebSocketProvider.cpp:
(WebKit::WebSocketProvider::createWebSocketChannel):
2020-04-14 Youenn Fablet <youenn@apple.com>
Add logging in case of WebRTC socket error
https://bugs.webkit.org/show_bug.cgi?id=210428
Reviewed by Eric Carlson.
Add some release logging in case of error when sending, setting option or closing a socket.
* NetworkProcess/webrtc/LibWebRTCSocketClient.cpp:
(WebKit::LibWebRTCSocketClient::sendTo):
(WebKit::LibWebRTCSocketClient::close):
(WebKit::LibWebRTCSocketClient::setOption):
2020-04-13 David Kilzer <ddkilzer@apple.com>
Replace use of Checked<size_t, RecordOverflow> with CheckedSize
<https://webkit.org/b/210461>
Reviewed by Mark Lam.
* Platform/IPC/cocoa/MachMessage.cpp:
(IPC::MachMessage::messageSize):
* Platform/IPC/cocoa/MachMessage.h:
* WebProcess/Network/WebSocketChannel.cpp:
(WebKit::WebSocketChannel::increaseBufferedAmount):
2020-04-13 Per Arne Vollan <pvollan@apple.com>
[iOS] Update message filters in the WebContent sandbox
https://bugs.webkit.org/show_bug.cgi?id=210455
Reviewed by Geoffrey Garen.
Update message filters in the WebContent sandbox on iOS with required messages.
No new tests, since this patch should not introduce a behavior change.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-04-13 Brent Fulgham <bfulgham@apple.com>
Add roblox.com to the desktop class quirks list
https://bugs.webkit.org/show_bug.cgi?id=210452
<rdar://problem/58859657>
Reviewed by Zalan Bujtas.
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::desktopClassBrowsingRecommendedForRequest):
2020-04-13 Wenson Hsieh <wenson_hsieh@apple.com>
Remove the "IgnoreViewportScalingConstraints" debug WebKit feature
https://bugs.webkit.org/show_bug.cgi?id=210440
Reviewed by Tim Horton.
Remove the internal WebKit2 feature; this was introduced around iOS 9 to unconditionally force pages to shrink
to fit if they specified a responsive viewport, to avoid horizontal scrolling.
Nowadays, this behavior is instead controlled by the `-[WKWebView _allowsViewportShrinkToFit]` SPI, which
ensures that this shrink-to-fit preference is updated in sync with the visible content rect update lifecycle; it
is currently used by Safari to scale web pages down to fit in smaller multitasking viewports.
In contrast, this preference has no internal clients. Instead of maintaining two different ways for clients to
control shrink-to-fit behavior, remove support for the preference and make it a no-op.
* Shared/WebPreferences.yaml:
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetIgnoreViewportScalingConstraints):
(WKPreferencesGetIgnoreViewportScalingConstraints):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/ios/WebPageIOS.mm:
Remove this redundant mechanism for shrinking web pages down to fit.
(WebKit::WebPage::updateVisibleContentRects):
2020-04-13 Per Arne Vollan <pvollan@apple.com>
[iOS] Remove unused UTType swizzler code
https://bugs.webkit.org/show_bug.cgi?id=210435
Unreviewed rollout of r258120.
* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeWebProcess):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):
2020-04-13 Kate Cheney <katherine_cheney@apple.com>
Turn all instances of NavigatingToAppBoundDomain into Optionals
https://bugs.webkit.org/show_bug.cgi?id=210363
<rdar://problem/61612861>
Reviewed by Brent Fulgham.
NavigatingToAppBoundDomain should be an Optional when passed to the
Network Process to distinguish the case of an empty app-bound domain
list versus an actual app-bound domain navigation. An empty list and
an app-bound navigation should have the same behavior in NetworkSessionCocoa,
but we will need distinguished behavior in the future.
* NetworkProcess/Downloads/DownloadManager.cpp:
(WebKit::DownloadManager::startDownload):
* NetworkProcess/Downloads/DownloadManager.h:
(WebKit::DownloadManager::startDownload):
* NetworkProcess/NetworkConnectionToWebProcess.cpp:
(WebKit::NetworkConnectionToWebProcess::startDownload):
(WebKit::NetworkConnectionToWebProcess::convertMainResourceLoadToDownload):
* NetworkProcess/NetworkConnectionToWebProcess.h:
(WebKit::NetworkConnectionToWebProcess::startDownload):
* NetworkProcess/NetworkConnectionToWebProcess.messages.in:
* NetworkProcess/NetworkLoadParameters.h:
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::preconnectTo):
(WebKit::NetworkProcess::downloadRequest):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/NetworkResourceLoadParameters.cpp:
(WebKit::NetworkResourceLoadParameters::decode):
* NetworkProcess/NetworkResourceLoadParameters.h:
* NetworkProcess/cache/AsyncRevalidation.cpp:
(WebKit::NetworkCache::AsyncRevalidation::AsyncRevalidation):
* NetworkProcess/cache/AsyncRevalidation.h:
* NetworkProcess/cache/NetworkCache.cpp:
(WebKit::NetworkCache::Cache::startAsyncRevalidationIfNeeded):
(WebKit::NetworkCache::Cache::retrieve):
* NetworkProcess/cache/NetworkCache.h:
* NetworkProcess/cache/NetworkCacheSpeculativeLoad.cpp:
(WebKit::NetworkCache::SpeculativeLoad::SpeculativeLoad):
* NetworkProcess/cache/NetworkCacheSpeculativeLoad.h:
* NetworkProcess/cache/NetworkCacheSpeculativeLoadManager.cpp:
(WebKit::NetworkCache::SpeculativeLoadManager::registerLoad):
(WebKit::NetworkCache::SpeculativeLoadManager::preconnectForSubresource):
(WebKit::NetworkCache::SpeculativeLoadManager::revalidateSubresource):
(WebKit::NetworkCache::SpeculativeLoadManager::preloadEntry):
(WebKit::NetworkCache::SpeculativeLoadManager::startSpeculativeRevalidation):
* NetworkProcess/cache/NetworkCacheSpeculativeLoadManager.h:
* NetworkProcess/cocoa/NetworkDataTaskCocoa.h:
* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::NetworkDataTaskCocoa::NetworkDataTaskCocoa):
* NetworkProcess/cocoa/NetworkSessionCocoa.h:
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::sessionWrapperForTask):
We should treat an empty WKAppBoundDomains list like an app-bound domain
navigation in the network session, so this sets
NavigatingToAppBoundDomain to Yes if the Optional is WTF::nullopt.
(WebKit::NetworkSessionCocoa::appBoundSession): Deleted.
(WebKit::NetworkSessionCocoa::isolatedSession): Deleted.
(WebKit::NetworkSessionCocoa::hasIsolatedSession const): Deleted.
(WebKit::NetworkSessionCocoa::clearIsolatedSessions): Deleted.
(WebKit::NetworkSessionCocoa::invalidateAndCancel): Deleted.
(WebKit::NetworkSessionCocoa::clearCredentials): Deleted.
(WebKit::certificatesMatch): Deleted.
(WebKit::NetworkSessionCocoa::allowsSpecificHTTPSCertificateForHost): Deleted.
(WebKit::NetworkSessionCocoa::continueDidReceiveChallenge): Deleted.
(WebKit::NetworkSessionCocoa::deviceManagementPolicyMonitor): Deleted.
(WebKit::NetworkSessionCocoa::createWebSocketTask): Deleted.
(WebKit::NetworkSessionCocoa::addWebSocketTask): Deleted.
(WebKit::NetworkSessionCocoa::removeWebSocketTask): Deleted.
(WebKit::NetworkSessionCocoa::hostNamesWithAlternativeServices const): Deleted.
(WebKit::NetworkSessionCocoa::deleteAlternativeServicesForHostNames): Deleted.
(WebKit::NetworkSessionCocoa::clearAlternativeServices): Deleted.
* Shared/LoadParameters.h:
* Shared/PolicyDecision.h:
(WebKit::PolicyDecision::decode):
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::preconnectTo):
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/ProvisionalPageProxy.cpp:
(WebKit::ProvisionalPageProxy::loadData):
(WebKit::ProvisionalPageProxy::loadRequest):
* UIProcess/ProvisionalPageProxy.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::loadRequestWithNavigationShared):
(WebKit::WebPageProxy::loadDataWithNavigationShared):
(WebKit::WebPageProxy::setIsNavigatingToAppBoundDomain):
Set this to WTF::nullopt instead of NavigatingToAppBoundDomain::Yes
so the two cases are distinguished.
(WebKit::WebPageProxy::isNavigatingToAppBoundDomainTesting):
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::isNavigatingToAppBoundDomain const):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::download):
* WebProcess/WebPage/WebFrame.cpp:
(WebKit::WebFrame::startDownload):
(WebKit::WebFrame::convertMainResourceLoadToDownload):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::loadDataImpl):
* WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::setIsNavigatingToAppBoundDomain):
(WebKit::WebPage::isNavigatingToAppBoundDomain const):
2020-04-13 David Kilzer <ddkilzer@apple.com>
Fix clang static analyzer warnings about unused instance variables in WebIconDatabase, WKView
<https://webkit.org/b/210427>
Reviewed by Alex Christensen.
* UIProcess/API/Cocoa/WKView.h:
(WKView._unused):
- Use WK_UNUSED_INSTANCE_VARIABLE to ignore unused instance
variable.
2020-04-13 David Kilzer <ddkilzer@apple.com>
Add WARN_UNUSED_RETURN to decode methods in Source/WebKit
<https://webkit.org/b/210423>
<rdar://problem/61715582>
Reviewed by Alex Christensen.
* GPUProcess/GPUProcessCreationParameters.h:
* NetworkProcess/Downloads/DownloadID.h:
(IPC::ArgumentCoder<WebKit::DownloadID>::decode):
* NetworkProcess/NetworkProcessCreationParameters.h:
* NetworkProcess/cache/NetworkCacheKey.h:
(WebKit::NetworkCache::DataKey::decode):
* NetworkProcess/cache/NetworkCacheStorage.cpp:
(WebKit::NetworkCache::decodeRecordMetaData):
(WebKit::NetworkCache::decodeRecordHeader):
* Platform/SharedMemory.h:
* PluginProcess/PluginCreationParameters.h:
* Shared/API/APIData.h:
* Shared/API/APIError.h:
* Shared/API/APIFrameHandle.h:
* Shared/API/APIGeometry.h:
* Shared/API/APINumber.h:
(API::Number::decode):
* Shared/API/APIPageGroupHandle.h:
* Shared/API/APIPageHandle.h:
* Shared/API/APIURL.h:
(API::URL::decode):
* Shared/API/APIURLRequest.h:
* Shared/API/APIURLResponse.h:
* Shared/API/Cocoa/RemoteObjectInvocation.h:
* Shared/Cocoa/ArgumentCodersCocoa.h:
* Shared/ContextMenuContextData.h:
* Shared/Databases/IndexedDB/WebIDBResult.h:
* Shared/EditorState.h:
* Shared/FocusedElementInformation.h:
* Shared/FontInfo.h:
* Shared/LayerTreeContext.h:
* Shared/LoadParameters.h:
* Shared/OptionalCallbackID.h:
(WebKit::OptionalCallbackID::decode):
* Shared/PlatformPopupMenuData.h:
* Shared/Plugins/PluginProcessCreationParameters.h:
* Shared/PrintInfo.h:
* Shared/RemoteLayerTree/RemoteLayerBackingStore.h:
* Shared/RemoteLayerTree/RemoteLayerTreeTransaction.h:
* Shared/RemoteLayerTree/RemoteScrollingCoordinatorTransaction.cpp:
* Shared/RemoteLayerTree/RemoteScrollingCoordinatorTransaction.h:
* Shared/SessionState.h:
* Shared/ShareableBitmap.h:
* Shared/ShareableResource.h:
* Shared/TouchBarMenuData.h:
* Shared/UpdateInfo.h:
* Shared/VisibleContentRectUpdateInfo.h:
* Shared/WebEvent.h:
* Shared/WebHitTestResultData.h:
* Shared/WebNavigationDataStore.h:
(WebKit::WebNavigationDataStore::decode):
* Shared/WebPreferencesStore.h:
* Shared/WebProcessCreationParameters.h:
* Shared/WebsiteData/WebsiteData.h:
* Shared/cf/ArgumentCodersCF.h:
* Shared/gtk/ArgumentCodersGtk.cpp:
(IPC::decodeImage):
(IPC::decodeGKeyFile):
* Shared/gtk/ArgumentCodersGtk.h:
* Shared/gtk/WebSelectionData.h:
* Shared/ios/InteractionInformationAtPosition.h:
* Shared/ios/InteractionInformationRequest.h:
* Shared/mac/ColorSpaceData.h:
* Shared/mac/ObjCObjectGraph.h:
* Shared/mac/SecItemRequestData.h:
* Shared/mac/WebGestureEvent.h:
* UIProcess/Cocoa/SessionStateCoding.h:
* UIProcess/LegacySessionStateCoding.h:
* UIProcess/mac/LegacySessionStateCoding.cpp:
(WebKit::decodeSessionHistoryEntryData):
(WebKit::decodeSessionHistoryEntry):
(WebKit::decodeSessionHistoryEntries):
(WebKit::decodeV0SessionHistory):
(WebKit::decodeV1SessionHistory):
(WebKit::decodeSessionHistory):
* WebProcess/GPU/GPUProcessConnectionInfo.h:
(WebKit::GPUProcessConnectionInfo::decode):
* WebProcess/GPU/media/RemoteAudioBusData.h:
(WebKit::RemoteAudioBusData::decode):
* WebProcess/MediaStream/MediaDeviceSandboxExtensions.h:
* WebProcess/Network/NetworkProcessConnectionInfo.h:
(WebKit::NetworkProcessConnectionInfo::decode):
* WebProcess/Plugins/Plugin.h:
- Add WARN_UNUSED_RETURN to all decode functions.
2020-04-13 Per Arne Vollan <pvollan@apple.com>
[iOS] Add message filtering to all iokit open rules
https://bugs.webkit.org/show_bug.cgi?id=210352
Reviewed by Sam Weinig.
Apply message filtering with reports and telemetry to all iokit-open rules in the WebContent process' sandbox on iOS
to gain a better understanding of which iokit-open rules we can investigate.
No new tests, since adding telemetry will not introduce a change in behavior.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-04-13 Joonghun Park <jh718.park@samsung.com>
Unreviewed. Remove redundant move in return statement.
Return statement already returns rvalue,
so we don't need move here.
This patch removes the build warning below since r259922.
warning: redundant move in return statement [-Wredundant-move]
* NetworkProcess/cache/NetworkCacheCoders.cpp:
(WTF::Persistence::Coder<WebCore::HTTPHeaderMap>::decode):
2020-04-13 Michael Catanzaro <mcatanzaro@gnome.org>
Fix various build warnings
https://bugs.webkit.org/show_bug.cgi?id=210429
Reviewed by Mark Lam.
Use private-code instead of deprecated code command to build the Wayland protocol.
* PlatformGTK.cmake:
2020-04-12 Darin Adler <darin@apple.com>
Fix some strange uses of start/endOfDocument
https://bugs.webkit.org/show_bug.cgi?id=210408
Reviewed by Wenson Hsieh.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::replaceDictatedText): Pass the document to startOfDocument,
rather than getting the documentelement, because the function works on any
node, so there's no reason to write code to get a particular one.
(WebKit::WebPage::applyAutocorrectionInternal): Ditto.
2020-04-12 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] Use gdk_window_move_to_rect() to position popup menus
https://bugs.webkit.org/show_bug.cgi?id=210069
Reviewed by Michael Catanzaro.
* UIProcess/gtk/WebPopupMenuProxyGtk.cpp:
(WebKit::WebPopupMenuProxyGtk::showPopupMenu):
2020-04-12 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK] Page is not drawn with PSON, hardware acceleration and address typed in URL bar
https://bugs.webkit.org/show_bug.cgi?id=208091
Reviewed by Michael Catanzaro.
This happens only when accelerated compositing mode policy is set to always. In this case, the
bindSurfaceToWebPage wayland message is received before the WebPageProxy has updated the web page identifier. We
should always check if there's a provisional page and get its web page identifier.
* UIProcess/gtk/WaylandCompositor.cpp:
(WebKit::WaylandCompositor::bindSurfaceToWebPage):
2020-04-12 Carlos Garcia Campos <cgarcia@igalia.com>
[GTK][PSON] Crash in WaylandCompositor::Surface destructor with USE_WPE_RENDERER disabled
https://bugs.webkit.org/show_bug.cgi?id=209345
Reviewed by Michael Catanzaro.
In case of PSON the web page identifier can change fo the same web page proxy. When swapping processes the new
web page calls bindSurfaceToWebPage(), but we still have the web page proxy bound to the previous surface. We
just need to unbind the current surface before binding the new one.
* UIProcess/gtk/WaylandCompositor.cpp:
(WebKit::WaylandCompositor::bindSurfaceToWebPage):
2020-04-11 David Kilzer <ddkilzer@apple.com>
Follow-up: Add WARN_UNUSED_RETURN to decode methods in Source/WebKit/Platform/IPC
<https://webkit.org/b/210316>
<rdar://problem/61559305>
Changes based on feedback from Alex Christensen and Darin Adler:
- Remove WARN_UNUSED_RETURN from methods returning Optional<>.
- Place WARN_UNUSED_RETURN consistently before the return type.
* Platform/IPC/ArgumentCoder.h:
* Platform/IPC/ArgumentCoders.cpp:
* Platform/IPC/ArgumentCoders.h:
(IPC::ArgumentCoder<OptionSet<T>>::decode):
(IPC::ArgumentCoder<Optional<T>>::decode):
(IPC::ArgumentCoder<Box<T>>::decode):
(IPC::TupleDecoderImpl::decode):
(IPC::TupleDecoderImpl<Type>::decode):
(IPC::TupleDecoder::decode):
(IPC::TupleDecoder<0>::decode):
(IPC::VariantCoder::decode):
* Platform/IPC/Attachment.h:
* Platform/IPC/DataReference.h:
* Platform/IPC/Decoder.h:
* Platform/IPC/FormDataReference.h:
(IPC::FormDataReference::decode):
* Platform/IPC/ImageDataReference.h:
(IPC::ImageDataReference::decode):
* Platform/IPC/SharedBufferDataReference.h:
(IPC::SharedBufferDataReference::decode):
* Platform/IPC/StringReference.h:
2020-04-11 David Kilzer <ddkilzer@apple.com>
Follow-up: Add WARN_UNUSED_RETURN to decode methods in WebCoreArgumentCoders
<https://webkit.org/b/210322>
<rdar://problem/61565148>
Changes based on feedback from Alex Christensen and Darin Adler:
- Remove WARN_UNUSED_RETURN from methods returning Optional<>.
- Place WARN_UNUSED_RETURN consistently before the return type.
* Shared/WebCoreArgumentCoders.cpp:
* Shared/WebCoreArgumentCoders.h:
2020-04-11 Daniel Bates <dabates@apple.com>
Add text interaction delegate functions
https://bugs.webkit.org/show_bug.cgi?id=210395
<rdar://problem/61650549>
Reviewed by Wenson Hsieh.
Implement some delegate callbacks that are used to toggle selection reveal,
placeholder text, and selection assistance.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _willBeginTextInteractionInTextInputContext:]): Added.
(-[WKContentView _didFinishTextInteractionInTextInputContext:]): Added.
2020-04-10 Darin Adler <darin@apple.com>
Move more from live range to SimpleRange: callers of absoluteTextRects
https://bugs.webkit.org/show_bug.cgi?id=210369
Reviewed by Anders Carlsson.
* WebProcess/InjectedBundle/API/mac/WKDOMNode.mm:
(-[WKDOMNode textRects]): Moved implementation here. There's no reason for Node
to have a textRects function, since it's a layout/rendering operation and a
single node is also an unusual special case, but for now at least we will keep
this legacy method.
* WebProcess/InjectedBundle/API/mac/WKDOMRange.mm:
(-[WKDOMRange textRects]): Changed to call RenderObject::absoluteTextRects
instead of Range::absoluteTextRects.
* WebProcess/WebPage/FindController.cpp:
(WebKit::FindController::updateFindUIAfterPageScroll): Use
RenderObject::absoluteTextRects instead of Range::absoluteRectRects.
(WebKit::FindController::findStringMatches): Ditto.
2020-04-10 Pablo Saavedra <psaavedra@igalia.com>
[GTK][WPE] Replace fopen/fclose by fopen/fseek functions in MemoryPressureMonitor
https://bugs.webkit.org/show_bug.cgi?id=209942
Reviewed by Adrian Perez de Castro.
Replacing the open()/close() logic to get the data from /proc/meminfo,
/proc/self/cgroup, ... with a new logic where the files are just
opened once and we just perform a seek to the initial position of
of an unbuffered file.
Also refactoring of some logic:
- Added the CGroupMemoryController class what encapsulates the logic
to calculate the the total memory and the memory usage with
cgroups.
- CGroupMemoryController encpasulates the FILE pointers to the
relative files to cgroups.
- The memInfoFile, zoneInfoFile and cgroupControllerFile are now
opened once in the Thread::create() invoked by the
MemoryPressureMonitor::start().
* UIProcess/linux/MemoryPressureMonitor.cpp:
(WebKit::lowWatermarkPages):
(WebKit::calculateMemoryAvailable):
(WebKit::getCgroupFile):
(WebKit::getCgroupControllerPath):
(WebKit::systemMemoryUsedAsPercentage):
(WebKit::MemoryPressureMonitor::start):
(WebKit::CGroupMemoryController::setMemoryControllerPath):
(WebKit::CGroupMemoryController::disposeMemoryController):
(WebKit::CGroupMemoryController::getCgroupFileValue):
(WebKit::CGroupMemoryController::getMemoryTotalWithCgroup):
(WebKit::CGroupMemoryController::getMemoryUsageWithCgroup):
(WebKit::getCgroupFileValue): Deleted.
(WebKit::getMemoryTotalWithCgroup): Deleted.
(WebKit::getMemoryUsageWithCgroup): Deleted.
* UIProcess/linux/MemoryPressureMonitor.h:
(WebKit::CGroupMemoryController::isActive):
(WebKit::CGroupMemoryController::~CGroupMemoryController):
2020-04-10 Alex Christensen <achristensen@webkit.org>
PersistentCoders should use modern decoding syntax
https://bugs.webkit.org/show_bug.cgi?id=207497
Reviewed by Darin Adler.
* NetworkProcess/cache/CacheStorageEngineCache.cpp:
(WebKit::CacheStorage::decodeDOMCacheRecord):
(WebKit::CacheStorage::Cache::decodeRecordHeader):
* NetworkProcess/cache/CacheStorageEngineCaches.cpp:
(WebKit::CacheStorage::Caches::readOrigin):
(WebKit::CacheStorage::decodeCachesNames):
* NetworkProcess/cache/NetworkCacheCoders.cpp:
(WTF::Persistence::Coder<WebCore::HTTPHeaderMap>::decode):
* NetworkProcess/cache/NetworkCacheEntry.cpp:
(WebKit::NetworkCache::Entry::decodeStorageRecord):
* NetworkProcess/cache/NetworkCacheKey.cpp:
(WebKit::NetworkCache::Key::decode):
* NetworkProcess/cache/NetworkCacheKey.h:
* NetworkProcess/cache/NetworkCacheStorage.cpp:
(WebKit::NetworkCache::decodeRecordMetaData):
* NetworkProcess/cache/NetworkCacheSubresourcesEntry.cpp:
(WebKit::NetworkCache::SubresourceInfo::encode const):
(WebKit::NetworkCache::SubresourceInfo::decode):
(WebKit::NetworkCache::SubresourcesEntry::decodeStorageRecord):
* NetworkProcess/cache/NetworkCacheSubresourcesEntry.h:
* Shared/UserData.cpp:
(WebKit::UserData::decode):
* UIProcess/API/APIContentRuleListStore.cpp:
(API::decodeContentRuleListMetaData):
2020-04-10 Devin Rousso <drousso@apple.com>
Web Inspector: add `broadcastConsoleMessage` calls for new ITPDebug logs after r259275
https://bugs.webkit.org/show_bug.cgi?id=210362
Reviewed by John Wilander.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::logFrameNavigation):
(WebKit::ResourceLoadStatisticsDatabaseStore::registrableDomainsToDeleteOrRestrictWebsiteDataFor):
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:
(WebKit::ResourceLoadStatisticsMemoryStore::logFrameNavigation):
(WebKit::ResourceLoadStatisticsMemoryStore::registrableDomainsToDeleteOrRestrictWebsiteDataFor):
2020-04-10 Chris Dumez <cdumez@apple.com>
[IPC Hardening] WebHitTestResultData IPC decoder should validate imageSharedMemory / imageSize
https://bugs.webkit.org/show_bug.cgi?id=210361
<rdar://problem/60758280>
Reviewed by Geoffrey Garen.
WebHitTestResultData IPC decoder should validate imageSharedMemory / imageSize to make sure they are consistent.
* Shared/WebHitTestResultData.cpp:
(WebKit::WebHitTestResultData::decode):
2020-04-10 Alex Christensen <achristensen@webkit.org>
Add SPI to handle proxy authentication and require secure proxy connections
https://bugs.webkit.org/show_bug.cgi?id=210343
<rdar://problem/61546396> <rdar://problem/61546519> and <rdar://problem/61546658>
Reviewed by Geoffrey Garen.
I also use WKWebViewConfiguration.connectionProxyDictionary instead of CFURLRequestSetProxySettings
to set proxy settings. The equivalence of this behavior is covered by the API tests.
* NetworkProcess/NetworkSessionCreationParameters.cpp:
(WebKit::NetworkSessionCreationParameters::encode const):
(WebKit::NetworkSessionCreationParameters::decode):
* NetworkProcess/NetworkSessionCreationParameters.h:
* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::NetworkDataTaskCocoa::applySniffingPoliciesAndBindRequestToInferfaceIfNeeded):
* NetworkProcess/cocoa/NetworkSessionCocoa.h:
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(-[WKNetworkSessionDelegate URLSession:task:didReceiveChallenge:completionHandler:]):
(WebKit::NetworkSessionCocoa::NetworkSessionCocoa):
* UIProcess/API/Cocoa/_WKWebsiteDataStoreConfiguration.h:
* UIProcess/API/Cocoa/_WKWebsiteDataStoreConfiguration.mm:
(-[_WKWebsiteDataStoreConfiguration preventsSystemHTTPProxyAuthentication]):
(-[_WKWebsiteDataStoreConfiguration setPreventsSystemHTTPProxyAuthentication:]):
(-[_WKWebsiteDataStoreConfiguration requiresSecureHTTPSProxyConnection]):
(-[_WKWebsiteDataStoreConfiguration setRequiresSecureHTTPSProxyConnection:]):
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::parameters):
* UIProcess/WebsiteData/WebsiteDataStoreConfiguration.cpp:
(WebKit::WebsiteDataStoreConfiguration::copy const):
* UIProcess/WebsiteData/WebsiteDataStoreConfiguration.h:
(WebKit::WebsiteDataStoreConfiguration::preventsSystemHTTPProxyAuthentication const):
(WebKit::WebsiteDataStoreConfiguration::setPreventsSystemHTTPProxyAuthentication):
(WebKit::WebsiteDataStoreConfiguration::requiresSecureHTTPSProxyConnection const):
(WebKit::WebsiteDataStoreConfiguration::setRequiresSecureHTTPSProxyConnection):
2020-04-10 Chris Dumez <cdumez@apple.com>
[IPC Hardening] WebDeviceOrientationAndMotionAccessController::shouldAllowAccess() should not crash on invalid input
https://bugs.webkit.org/show_bug.cgi?id=210359
<rdar://problem/60092453>
Reviewed by Geoffrey Garen.
Make sure the security origin is a valid HashMap key before trying to do a HashMap lookup.
The security origin is coming from IPC and thus cannot be trusted.
* UIProcess/WebsiteData/WebDeviceOrientationAndMotionAccessController.cpp:
(WebKit::WebDeviceOrientationAndMotionAccessController::cachedDeviceOrientationPermission const):
2020-04-10 John Wilander <wilander@apple.com>
Add capability to opt in specific domains into SameSite=strict bounce tracking protection
https://bugs.webkit.org/show_bug.cgi?id=210347
<rdar://problem/61593323>
Reviewed by Brent Fulgham.
This change adds an additional opt-in condition to the shouldEnforceSameSiteStrictFor()
functions in both ResourceLoadStatisticsMemoryStore and ResourceLoadStatisticsDatabaseStore.
No new tests. This adds domain specific logic, quirk-style. The logFrameNavigation()
changes are covered by existing tests.
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::logFrameNavigation):
Now only captures real redirects for the purposes of the SameSite=strict rule.
(WebKit::ResourceLoadStatisticsDatabaseStore::shouldEnforceSameSiteStrictFor):
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:
(WebKit::ResourceLoadStatisticsMemoryStore::logFrameNavigation):
Now only captures real redirects for the purposes of the SameSite=strict rule.
(WebKit::ResourceLoadStatisticsMemoryStore::shouldEnforceSameSiteStrictFor):
(WebKit::ResourceLoadStatisticsMemoryStore::registrableDomainsToDeleteOrRestrictWebsiteDataFor):
Removed stray newline.
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.cpp:
(WebKit::ResourceLoadStatisticsStore::shouldEnforceSameSiteStrictForSpecificDomain const):
The new function with domains to subject to the new rule.
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.h:
2020-04-10 Chris Dumez <cdumez@apple.com>
First render after a process swap does not use accelerated rendering
https://bugs.webkit.org/show_bug.cgi?id=210297
<rdar://problem/61432515>
Reviewed by Antti Koivisto.
To prevent flashing on navigation when process-swapping, we were delaying the
DrawingAreaProxy::EnterAcceleratedCompositingMode IPC message from the WebProcess
to the UIProcess, until the new page is ready to display. The reason we did this
is that the EnterAcceleratedCompositingMode IPC message ends up setting the root
CALayer so the new content would become visible on screen. However, another side
effect of this is that we would also delay the point until which we get
accelerated rendering, which would negatively impact early page load performance.
To address the issue, the following changes were made:
1. We no longer delay the DrawingAreaProxy::EnterAcceleratedCompositingMode IPC.
In non process-swap cases, we send it as soon as the DrawingArea is constructed.
In process-swap cases, we send it as soon as we commit the load since this is
when the provisional drawing area gets committed and happens before we have
any content to paint.
2. In the UIProcess, when we get the DrawingAreaProxy::EnterAcceleratedCompositingMode
IPC and we already have a root layer, we add the new one behind the existing one
instead of replacing the existing one. As a result, the new layer will get
accelerated compositing but will not be visible on screen yet.
3. Introduce a new DrawingAreaProxy::DidFirstLayerFlush IPC which gets send at the
point where we used to send the EnterAcceleratedCompositingMode IPC, meaning where
the new content is ready to display. When the UIProcess receives this IPC, it
actually replaces the root layer with the new layer so that the new content visible.
After this patch, our CALayers get accelerated compositing right away after a
process-swap but we still do not see any blank-flashing on navigation.
* UIProcess/Cocoa/WebViewImpl.h:
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::setAcceleratedCompositingRootLayer):
(WebKit::WebViewImpl::setAcceleratedCompositingRootLayerAfterFlush):
* UIProcess/DrawingAreaProxy.h:
(WebKit::DrawingAreaProxy::didFirstLayerFlush):
* UIProcess/DrawingAreaProxy.messages.in:
* UIProcess/PageClient.h:
(WebKit::PageClient::didFirstLayerFlush):
* UIProcess/SuspendedPageProxy.cpp:
(WebKit::SuspendedPageProxy::SuspendedPageProxy):
(WebKit::SuspendedPageProxy::pageDidFirstLayerFlush):
(WebKit::SuspendedPageProxy::closeWithoutFlashing):
* UIProcess/SuspendedPageProxy.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::suspendCurrentPageIfPossible):
(WebKit::WebPageProxy::commitProvisionalPage):
* UIProcess/WebPageProxy.h:
* UIProcess/mac/PageClientImplMac.h:
* UIProcess/mac/PageClientImplMac.mm:
(WebKit::PageClientImpl::didFirstLayerFlush):
* UIProcess/mac/TiledCoreAnimationDrawingAreaProxy.h:
* UIProcess/mac/TiledCoreAnimationDrawingAreaProxy.mm:
(WebKit::TiledCoreAnimationDrawingAreaProxy::didFirstLayerFlush):
* WebProcess/WebPage/DrawingArea.h:
(WebKit::DrawingArea::sendEnterAcceleratedCompositingModeIfNeeded):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::didCommitLoad):
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.h:
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::TiledCoreAnimationDrawingArea):
(WebKit::TiledCoreAnimationDrawingArea::sendDidFirstLayerFlushIfNeeded):
(WebKit::TiledCoreAnimationDrawingArea::sendEnterAcceleratedCompositingModeIfNeeded):
(WebKit::TiledCoreAnimationDrawingArea::updateRendering):
2020-04-10 Wenson Hsieh <wenson_hsieh@apple.com>
[iOS] Unable to select text by tap-hold or double tap-hold when allowsLinkPreview property is set to NO
https://bugs.webkit.org/show_bug.cgi?id=210191
<rdar://problem/60454016>
Reviewed by Tim Horton.
In iOS 13.4, UIWKTextInteractionAssistant no longer uses the (misleadingly-named) "force press" gesture to drive
text selection when performing a long press. Instead, it now uses the UIVariableDelayLoupeGestureRecognizer,
i.e. the text interaction assistant's loupe gesture recognizer.
The highlight long press gesture, which is only enabled when `allowsLinkPreview` is set to `NO`, is currently
allowed to recognize alongside the force press gesture but not the loupe gesture. This means that when link
previews are disabled, we'll end up allowing the highlight long press gesture to prevent text selection. A
similar issue already exists for the tap-and-half gesture as well.
To fix this, we allow the highlight long press gesture to recognize alongside both gestures.
Test: editing/selection/ios/select-text-with-link-preview-disabled.html
* UIProcess/ios/WKContentViewInteraction.mm:
(tapAndAHalfRecognizerClass):
While UITapAndAHalfRecognizer is declared in a UIKit private header, it is (perhaps unintentionally) not
exported, so attempting to call +[UITapAndAHalfRecognizer class] leads to a linker error. Work around this by
dynamically grabbing the `Class` from the class name.
The tap and half gesture recognizer is private in UIKit and only used for text interaction gestures, so we
always want the highlight long press gesture to be able to recognize alongside it.
(-[WKContentView gestureRecognizer:shouldRecognizeSimultaneouslyWithGestureRecognizer:]):
Also clean up some existing code by replacing `.get().`s with Objective-C bracket operators.
2020-04-10 Brent Fulgham <bfulgham@apple.com>
Switch unused IOKit classes from allow-with-report to deny-with-report
https://bugs.webkit.org/show_bug.cgi?id=210136
<rdar://problem/61403216>
Reviewed by Per Arne Vollan.
After tracking telemetry for several weeks, we can now safely switch the default state
of a number of IOKit classes from 'allow with telemetry' to 'deny with telemetry'.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
* WebProcess/com.apple.WebProcess.sb.in:
2020-04-10 Daniel Bates <dabates@apple.com>
mayContainEditableElementsInRect() can terminate search once the search rect is wholly contained inside a layer
https://bugs.webkit.org/show_bug.cgi?id=210340
Reviewed by Simon Fraser.
Reduce false positives. This makes mayContainEditableElementsInRect() return false when the
search rect is wholly contained inside a higher paint order layer with no editable elements
that overlaps a layer with editable elements.
* UIProcess/RemoteLayerTree/ios/RemoteLayerTreeViews.mm:
(WebKit::mayContainEditableElementsInRect):
2020-04-10 Per Arne Vollan <pvollan@apple.com>
[iOS] Add routines to message filter in the WebContent sandbox
https://bugs.webkit.org/show_bug.cgi?id=210338
Reviewed by Geoffrey Garen.
Add required routines to the message filter in the WebContent sandbox on iOS. These have been observed
being in use during local testing.
* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
2020-04-10 Simon Fraser <simon.fraser@apple.com>
Reset view navigation gesture state between tests
https://bugs.webkit.org/show_bug.cgi?id=210283
Re-land r259818 with a null-check on m_webPageProxyForBackForwardListForCurrentSwipe.
* UIProcess/API/Cocoa/WKWebViewPrivateForTesting.h:
* UIProcess/API/Cocoa/WKWebViewTesting.mm:
(-[WKWebView _resetNavigationGestureStateForTesting]):
* UIProcess/ViewGestureController.cpp:
(WebKit::ViewGestureController::willBeginGesture):
(WebKit::ViewGestureController::didEndGesture):
(WebKit::ViewGestureController::PendingSwipeTracker::handleEvent):
(WebKit::ViewGestureController::PendingSwipeTracker::eventWasNotHandledByWebCore):
* UIProcess/ViewGestureController.h:
* UIProcess/ios/ViewGestureControllerIOS.mm:
(WebKit::ViewGestureController::removeSwipeSnapshot):
(WebKit::ViewGestureController::resetState):
(WebKit::ViewGestureController::reset):
* UIProcess/mac/ViewGestureControllerMac.mm:
(WebKit::ViewGestureController::removeSwipeSnapshot):
(WebKit::ViewGestureController::resetState):
(WebKit::ViewGestureController::reset):
2020-04-10 David Kilzer <ddkilzer@apple.com>
Add WARN_UNUSED_RETURN to decode methods in WebCoreArgumentCoders
<https://webkit.org/b/210322>
Reviewed by Darin Adler.
* Shared/WebCoreArgumentCoders.cpp:
(IPC::decodeSharedBuffer):
(IPC::decodeTypesAndData):
(IPC::decodeImage):
(IPC::decodeOptionalImage):
(IPC::decodeNativeImage):
(IPC::decodeOptionalNativeImage):
* Shared/WebCoreArgumentCoders.h:
* Shared/mac/WebCoreArgumentCodersMac.mm:
(IPC::decodeNSError):
2020-04-10 Peng Liu <peng.liu6@apple.com>
Add null pointer checkings in WebFullScreenManager
https://bugs.webkit.org/show_bug.cgi?id=210291
Reviewed by Daniel Bates.
Add null pointer checkings for "element" and "m_element" in WebFullScreenManager.
* WebProcess/FullScreen/WebFullScreenManager.cpp:
(WebKit::screenRectOfContents):
(WebKit::WebFullScreenManager::enterFullScreenForElement):
(WebKit::WebFullScreenManager::willEnterFullScreen):
(WebKit::WebFullScreenManager::didEnterFullScreen):
(WebKit::WebFullScreenManager::willExitFullScreen):
(WebKit::WebFullScreenManager::didExitFullScreen):
(WebKit::WebFullScreenManager::setAnimatingFullScreen):
(WebKit::WebFullScreenManager::requestExitFullScreen):
== Rolled over to ChangeLog-2020-04-10 ==
View Git Blame Copy Permalink