Timeline
Jan 18, 2019:
- 11:53 PM Changeset in webkit [240199] by
-
- 19 edits2 adds in trunk
iOS: Updating input mode should update the software keyboard
https://bugs.webkit.org/show_bug.cgi?id=193565
<rdar://problem/47376334>
Reviewed by Wenson Hsieh.
Source/WebCore:
Let the chrome client know that the focused element's inputmode had changed.
Test: fast/forms/ios/inputmode-none-removed.html
- html/HTMLElement.cpp:
(WebCore::HTMLElement::parseAttribute):
- page/ChromeClient.h:
Source/WebKit:
Update the software keyboard when the inputmode content attribute on the focused element had been mutated.
- Scripts/webkit/messages.py:
- Shared/WebCoreArgumentCoders.h:
- UIProcess/PageClient.h:
- UIProcess/WebPageProxy.h:
- UIProcess/WebPageProxy.messages.in:
- UIProcess/ios/PageClientImplIOS.h:
- UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::focusedElementDidChangeInputMode):
- UIProcess/ios/WKContentViewInteraction.h:
- UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _didUpdateInputMode:]):
- UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::focusedElementDidChangeInputMode):
- WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::focusedElementDidChangeInputMode):
- WebProcess/WebCoreSupport/WebChromeClient.h:
- WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::focusedElementDidChangeInputMode):
- WebProcess/WebPage/WebPage.h:
LayoutTests:
Added a regression test for removing inputmode content attribute with the value of "none".
The test methodology is different between testRunner and in-browser since we don't force
software keyboard while running layout tests inside simulator which can elimiate
the visual viewport difference inside the test runner, and in-browser testing obviously
doesn't have access to the internal keyboard metrics.
- fast/forms/ios/inputmode-none-removed-expected.txt: Added.
- fast/forms/ios/inputmode-none-removed.html: Added.
- 9:42 PM Changeset in webkit [240198] by
-
- 2 edits in trunk/Source/WebKit
Adjust WKDrawingView protocol method name
https://bugs.webkit.org/show_bug.cgi?id=193604
Reviewed by Wenson Hsieh.
- UIProcess/ios/WKDrawingView.mm:
(-[WKDrawingView canvasViewDrawingDidChange:]):
(-[WKDrawingView drawingDidChange:]): Deleted.
- 9:28 PM Changeset in webkit [240197] by
-
- 2 edits in trunk/Source/WebKit
[WebKit] Change Xcode folders to map to folders on disk
- WebKit.xcodeproj/project.pbxproj: Update these folders:
- Shared/Authentication
- Shared/Authentication/cocoa
- 9:28 PM Changeset in webkit [240196] by
-
- 2 edits in trunk/Source/WebCore/PAL
[PAL] Change Xcode folders to map to folders on disk
Found by tidy-Xcode-project-file script (see Bug 188754).
- PAL.xcodeproj/project.pbxproj: Update these folders:
- pal/avfoundation
- pal/crypto/commoncrypto
- pal/ios
- 8:22 PM Changeset in webkit [240195] by
-
- 8 edits in trunk/Source
Automation.computeElementLayout should return visual viewport-aware coordinates
https://bugs.webkit.org/show_bug.cgi?id=193598
<rdar://problem/35325644>
Reviewed by Simon Fraser.
Source/WebCore:
- page/FrameView.h: export symbol to be usable from WebKit.
- page/FrameView.cpp:
(WebCore::FrameView::clientToLayoutViewportRect const): Added.
Do the same thing as clientToLayoutViewportPoint with a rect instead.
Source/WebKit:
Previously I added CoordinateSystem::VisualViewport to stub out this for iOS.
But I think that it's a mistake for safaridriver to care about VisualViewport
being enabled or not, because it is a runtime-switchable setting.
This patch removes CoordinateSystem::VisualViewport. Make the existing
CoordinateSystem::LayoutViewport use visual viewport semantics if needed.
This is tested by WebDriver element clicking tests. There should not be any
difference in behavior until it is possible to zoom with gestures via WebDriver.
- Shared/CoordinateSystem.h:
- UIProcess/Automation/Automation.json:
- UIProcess/Automation/WebAutomationSession.cpp:
(WebKit::protocolStringToCoordinateSystem):
- WebProcess/Automation/WebAutomationSessionProxy.cpp:
(WebKit::WebAutomationSessionProxy::computeElementLayout):
- 7:35 PM Changeset in webkit [240194] by
-
- 4 edits in trunk/Source/JavaScriptCore
JSScript API should only take ascii files.
https://bugs.webkit.org/show_bug.cgi?id=193420
Reviewed by Saam Barati.
This patch leaves the UTF8 method for binary compatablity, which
will be removed later.
- API/JSScript.h:
- API/JSScript.mm:
(fillBufferWithContentsOfFile):
(+[JSScript scriptFromASCIIFile:inVirtualMachine:withCodeSigning:andBytecodeCache:]):
(+[JSScript scriptFromUTF8File:inVirtualMachine:withCodeSigning:andBytecodeCache:]):
- API/tests/testapi.mm:
(-[JSContextFileLoaderDelegate context:fetchModuleForIdentifier:withResolveHandler:andRejectHandler:]):
- 7:27 PM Changeset in webkit [240193] by
-
- 4 edits in trunk/Source/bmalloc
gigacage slide should randomize both start and end
https://bugs.webkit.org/show_bug.cgi?id=193601
Reviewed by Yusuke Suzuki.
This patch makes it so that the gigacade slide has an arbitrary
distance from the end as well as the start. This is done by
picking a random size then based on that size picking an random
starting offset.
- bmalloc/Gigacage.h:
- bmalloc/Heap.cpp:
(bmalloc::Heap::Heap):
- 7:05 PM Changeset in webkit [240192] by
-
- 3 edits in trunk/PerformanceTests
Use scores everywhere in JetStream2's UI
https://bugs.webkit.org/show_bug.cgi?id=193593
Reviewed by Keith Miller.
- JetStream2/JetStreamDriver.js:
(DefaultBenchmark.prototype.processResults):
(DefaultBenchmark.prototype.get score):
(DefaultBenchmark.scoreDescription):
(WSLBenchmark.prototype.processResults):
(WSLBenchmark.prototype.get score):
(WSLBenchmark.scoreDescription):
(WasmBenchmark.prototype.processResults):
(WasmBenchmark.prototype.get score):
(WasmBenchmark.scoreDescription):
- JetStream2/index.html:
- 5:06 PM Changeset in webkit [240191] by
-
- 2 edits in trunk/Tools
Disable the run-jsc-stress-test remote host key check [second round].
https://bugs.webkit.org/show_bug.cgi?id=192988
Patch by Zhifei FANG <zhifei_fang@apple.com> on 2019-01-18
Reviewed by Dewei Zhu.
- Scripts/run-jsc-stress-tests:
- 4:22 PM Changeset in webkit [240190] by
-
- 7 edits4 adds in trunk/Source
Revert r238815, it broke WK1 video fullscreen on Mac
https://bugs.webkit.org/show_bug.cgi?id=193586
<rdar://problem/47358941>
Reviewed by Jer Noble.
Source/WebCore:
- PlatformMac.cmake:
- platform/graphics/avfoundation/MediaPlayerPrivateAVFoundation.cpp:
(WebCore::MediaPlayerPrivateAVFoundation::supportsFullscreen const):
Source/WebKitLegacy:
- WebKitLegacy.xcodeproj/project.pbxproj:
Source/WebKitLegacy/mac:
- WebView/WebVideoFullscreenController.h: Added.
- WebView/WebVideoFullscreenController.mm: Added.
(SOFT_LINK_CLASS):
(-[WebVideoFullscreenController init]):
(-[WebVideoFullscreenController dealloc]):
(-[WebVideoFullscreenController fullscreenWindow]):
(-[WebVideoFullscreenController windowDidLoad]):
(-[WebVideoFullscreenController videoElement]):
(-[WebVideoFullscreenController setVideoElement:]):
(-[WebVideoFullscreenController clearFadeAnimation]):
(-[WebVideoFullscreenController windowDidExitFullscreen]):
(-[WebVideoFullscreenController windowDidEnterFullscreen]):
(-[WebVideoFullscreenController videoElementRect]):
(-[WebVideoFullscreenController applicationDidResignActive:]):
(frameExpandedToRatioOfFrame):
(createBackgroundFullscreenWindow):
(-[WebVideoFullscreenController setupFadeAnimationIfNeededAndFadeIn:]):
(-[WebVideoFullscreenController enterFullscreen:]):
(-[WebVideoFullscreenController exitFullscreen]):
(-[WebVideoFullscreenController applicationDidChangeScreenParameters:]):
(-[WebVideoFullscreenController updateMenuAndDockForFullscreen]):
(-[WebVideoFullscreenController _requestExit]):
(-[WebVideoFullscreenController requestExitFullscreenWithAnimation:]):
(-[WebVideoFullscreenController requestExitFullscreen]):
(-[WebVideoFullscreenController fadeHUDIn]):
(-[WebVideoFullscreenController observeValueForKeyPath:ofObject:change:context:]):
(-[WebVideoFullscreenController rateChanged:]):
(-[WebVideoFullscreenWindow initWithContentRect:styleMask:backing:defer:]):
(-[WebVideoFullscreenWindow dealloc]):
(-[WebVideoFullscreenWindow resignFirstResponder]):
(-[WebVideoFullscreenWindow canBecomeKeyWindow]):
(-[WebVideoFullscreenWindow mouseDown:]):
(-[WebVideoFullscreenWindow cancelOperation:]):
(-[WebVideoFullscreenWindow animatedResizeDidEnd]):
(-[WebVideoFullscreenWindow animateFromRect:toRect:withSubAnimation:controllerAction:]):
(-[WebVideoFullscreenWindow animationDidEnd:]):
(-[WebVideoFullscreenWindow mouseMoved:]):
- WebView/WebVideoFullscreenHUDWindowController.h: Added.
- WebView/WebVideoFullscreenHUDWindowController.mm: Added.
(-[WebVideoFullscreenHUDWindow initWithContentRect:styleMask:backing:defer:]):
(-[WebVideoFullscreenHUDWindow canBecomeKeyWindow]):
(-[WebVideoFullscreenHUDWindow cancelOperation:]):
(-[WebVideoFullscreenHUDWindow center]):
(-[WebVideoFullscreenHUDWindow keyDown:]):
(-[WebVideoFullscreenHUDWindow resignFirstResponder]):
(-[WebVideoFullscreenHUDWindow performKeyEquivalent:]):
(-[WebVideoFullscreenHUDWindowController init]):
(-[WebVideoFullscreenHUDWindowController dealloc]):
(-[WebVideoFullscreenHUDWindowController setArea:]):
(-[WebVideoFullscreenHUDWindowController keyDown:]):
(-[WebVideoFullscreenHUDWindowController delegate]):
(-[WebVideoFullscreenHUDWindowController setDelegate:]):
(-[WebVideoFullscreenHUDWindowController scheduleTimeUpdate]):
(-[WebVideoFullscreenHUDWindowController unscheduleTimeUpdate]):
(-[WebVideoFullscreenHUDWindowController fadeWindowIn]):
(-[WebVideoFullscreenHUDWindowController fadeWindowOut]):
(-[WebVideoFullscreenHUDWindowController closeWindow]):
(createMediaUIControl):
(createControlWithMediaUIControlType):
(createTimeTextField):
(createMediaUIBackgroundView):
(-[WebVideoFullscreenHUDWindowController windowDidLoad]):
(-[WebVideoFullscreenHUDWindowController updateVolume]):
(-[WebVideoFullscreenHUDWindowController updateTime]):
(-[WebVideoFullscreenHUDWindowController endScrubbing]):
(-[WebVideoFullscreenHUDWindowController timelinePositionChanged:]):
(-[WebVideoFullscreenHUDWindowController currentTime]):
(-[WebVideoFullscreenHUDWindowController setCurrentTime:]):
(-[WebVideoFullscreenHUDWindowController duration]):
(-[WebVideoFullscreenHUDWindowController maxVolume]):
(-[WebVideoFullscreenHUDWindowController volumeChanged:]):
(-[WebVideoFullscreenHUDWindowController setVolumeToZero:]):
(-[WebVideoFullscreenHUDWindowController setVolumeToMaximum:]):
(-[WebVideoFullscreenHUDWindowController decrementVolume]):
(-[WebVideoFullscreenHUDWindowController incrementVolume]):
(-[WebVideoFullscreenHUDWindowController volume]):
(-[WebVideoFullscreenHUDWindowController setVolume:]):
(-[WebVideoFullscreenHUDWindowController updatePlayButton]):
(-[WebVideoFullscreenHUDWindowController updateRate]):
(-[WebVideoFullscreenHUDWindowController togglePlaying:]):
(-[WebVideoFullscreenHUDWindowController playing]):
(-[WebVideoFullscreenHUDWindowController setPlaying:]):
(timeToString):
(-[WebVideoFullscreenHUDWindowController remainingTimeText]):
(-[WebVideoFullscreenHUDWindowController elapsedTimeText]):
(-[WebVideoFullscreenHUDWindowController mouseEntered:]):
(-[WebVideoFullscreenHUDWindowController mouseExited:]):
(-[WebVideoFullscreenHUDWindowController rewind:]):
(-[WebVideoFullscreenHUDWindowController fastForward:]):
(-[WebVideoFullscreenHUDWindowController exitFullscreen:]):
(-[WebVideoFullscreenHUDWindowController windowDidExpose:]):
(-[WebVideoFullscreenHUDWindowController windowDidClose:]):
- WebView/WebView.mm:
(-[WebView _enterVideoFullscreenForVideoElement:mode:]):
(-[WebView _exitVideoFullscreen]):
- 4:18 PM Changeset in webkit [240189] by
-
- 2 edits in trunk/Tools
[build.webkit.org] unit-test failure Builder GTK Linux 32-bit Release doesn't exist
https://bugs.webkit.org/show_bug.cgi?id=193597
Reviewed by Lucas Forschler.
- BuildSlaveSupport/build.webkit.org-config/steps_unittest.py:
- 4:13 PM Changeset in webkit [240188] by
-
- 3 edits in trunk/Source/WebCore
Web Inspector: Audit: don't keep the injected sub-objects alive between runs
https://bugs.webkit.org/show_bug.cgi?id=193594
<rdar://problem/47398091>
Reviewed by Joseph Pecoraro.
Rather than save the
InspectorAuditAccessibilityObjectandInspectorAuditDOMObject
between individualAudit.setupinvocations, we should only keep them alive so long as the
injectedWebInspectorAuditobject is alive (e.g. fromAudit.setuptoAudit.teardown).
This change fixes inspector/audit/run-accessibility.html.
- inspector/agents/page/PageAuditAgent.h:
- inspector/agents/page/PageAuditAgent.cpp:
(WebCore::PageAuditAgent::populateAuditObject):
- 4:10 PM Changeset in webkit [240187] by
-
- 4 edits in trunk/Tools
[build.webkit.org] api tests failures are incorrectly named as unit test failures
https://bugs.webkit.org/show_bug.cgi?id=193595
Reviewed by Lucas Forschler.
- BuildSlaveSupport/build.webkit.org-config/factories.py: Renamed RunUnitTests to RunAPITests.
- BuildSlaveSupport/build.webkit.org-config/steps.py: Ditto.
- BuildSlaveSupport/build.webkit.org-config/steps_unittest.py: Ditto.
- 4:08 PM Changeset in webkit [240186] by
-
- 2 edits in trunk/Source/JavaScriptCore
Follow-up: Gigacages should start allocations from a slide
<https://bugs.webkit.org/show_bug.cgi?id=193523>
<rdar://problem/44958707>
- ftl/FTLLowerDFGToB3.cpp:
(JSC::FTL::DFG::LowerDFGToB3::caged): Add UNUSED_PARAM(kind) to
fix the build.
- 3:45 PM Changeset in webkit [240185] by
-
- 2 edits in trunk/Source/WebCore
Another attempt to fix the iOS build following <https://trac.webkit.org/changeset/240174>
(https://bugs.webkit.org/show_bug.cgi?id=193583)
Substitute ENABLE(FULL_KEYBOARD_ACCESS) for PLATFORM(MAC). On Mac, we always build with
ENABLE(FULL_KEYBOARD_ACCESS) enabled.
- rendering/RenderElement.cpp:
(WebCore::RenderElement::paintFocusRing):
- 3:44 PM Changeset in webkit [240184] by
-
- 2 edits in trunk/Source/WebCore/PAL
Follow-up: Override the session configuration for cookieAcceptPolicy
<https://bugs.webkit.org/show_bug.cgi?id=190925>
<rdar://problem/45497382>
Discussed with John Wilander.
- pal/spi/cf/CFNetworkSPI.h: Declare new SPI for the internal
macOS 10.12 Sierra SDK since it will never appear in those
headers. This fixes the build for trunk WebKit on Sierra.
- 3:34 PM Changeset in webkit [240183] by
-
- 5 edits in trunk/Source
Fix some build issues.
Source/WebCore:
Including UIKitSoftLinking.h is not compatible with unified builds.
- SourcesCocoa.txt:
- WebCore.xcodeproj/project.pbxproj:
Source/WebKit:
Include missing header WKDrawingCoordinator.
- UIProcess/ios/WKInkPickerView.mm:
- 3:26 PM Changeset in webkit [240182] by
-
- 3 edits in trunk/Websites/perf.webkit.org
Test group results notification should not say a build request to build had failed even when it had successfully completed.
https://bugs.webkit.org/show_bug.cgi?id=193064
Reviewed by Ryosuke Niwa.
Should show 'Build completed' or 'Build failed' for build type build requests.
- browser-tests/test-group-result-page-tests.js: Added a unit test to guard this bug.
- tools/js/test-group-result-page.js: Show 'Build completed' or 'Build failed' for build type build requests.
- 3:15 PM Changeset in webkit [240181] by
-
- 10 edits in trunk/Source
Get rid of ADVANCED_SPELL_CHECKING
https://bugs.webkit.org/show_bug.cgi?id=193592
Reviewed by Wenson Hsieh.
Source/WebCore/PAL:
- pal/spi/mac/NSSpellCheckerSPI.h:
It's always true on macOS, and all but one use of it is already inside PLATFORM(MAC)
Source/WebKit:
- UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::requestCandidatesForSelectionIfNeeded):
(WebKit::WebViewImpl::insertText):
- UIProcess/mac/TextCheckerMac.mm:
(WebKit::TextChecker::checkTextOfParagraph):
(WebKit::TextChecker::getGuessesForWord):
Source/WebKitLegacy/mac:
- WebCoreSupport/WebEditorClient.mm:
(insertionPointFromCurrentSelection):
(WebEditorClient::checkTextOfParagraph):
(WebEditorClient::getGuessesForWord):
(WebEditorClient::requestCandidatesForSelection):
(WebEditorClient::requestCheckingOfString):
- WebView/WebHTMLView.mm:
(-[WebHTMLView insertText:]):
Source/WTF:
- wtf/Platform.h:
- 3:09 PM Changeset in webkit [240180] by
-
- 19 edits2 adds in trunk
(WIP) [WebGPU] WebGPUProgrammablePassEncoder::setBindGroup prototype
https://bugs.webkit.org/show_bug.cgi?id=193457
<rdar://problem/47296678>
Reviewed by Dean Jackson.
Source/WebCore:
Enable WebGPU developers to assign buffer bind groups and access them via render pipeline.
Test: webgpu/buffer-resource-triangles.html
- Modules/webgpu/WebGPUBindGroup.h:
(WebCore::WebGPUBindGroup::bindGroup const): Added getter.
- Modules/webgpu/WebGPUProgrammablePassEncoder.cpp:
(WebCore::WebGPUProgrammablePassEncoder::setBindGroup const): Added. Delegates call to GPUProgrammablePassEncoder.
- Modules/webgpu/WebGPUProgrammablePassEncoder.h: Enable setBindGroup.
- Modules/webgpu/WebGPUProgrammablePassEncoder.idl: ditto.
- WebCore.xcodeproj/project.pbxproj:
- platform/graphics/gpu/GPUBindGroup.h:
(WebCore::GPUBindGroup::layout const): Added getter.
(WebCore::GPUBindGroup::bindings const): ditto.
- platform/graphics/gpu/GPUBindGroupLayout.h: Added ArgumentEncoderBuffer struct to retain ptr to both MTLArgumentEncoders and their argument MTLBuffers.
(WebCore::GPUBindGroupLayout::ArgumentEncoderBuffer::operator! const):
(WebCore::GPUBindGroupLayout::vertexArguments const): Added. Getter.
(WebCore::GPUBindGroupLayout::fragmentArguments const): ditto.
(WebCore::GPUBindGroupLayout::computeArguments const): ditto.
- platform/graphics/gpu/GPUDevice.cpp: Refactored unnecessary header include.
- platform/graphics/gpu/GPUDevice.h: ditto.
- platform/graphics/gpu/GPUProgrammablePassEncoder.h:
(WebCore::GPUProgrammablePassEncoder::setVertexBuffer): Added. Delegates to MTLCommandEncoder call.
(WebCore::GPUProgrammablePassEncoder::setFragmentBuffer): ditto.
- platform/graphics/gpu/GPURenderPassEncoder.h:
- platform/graphics/gpu/cocoa/GPUBindGroupLayoutMetal.mm:
(WebCore::tryCreateArgumentEncoderAndBuffer): Replace newEncoder(). Now create an ArgumentEncoderBuffer object from an array of MTLArgumentDescriptors, if able.
(WebCore::GPUBindGroupLayout::tryCreate): Refactor to support ArgumentEncoderBuffers, and allocate MTLBuffers for argument encoders.
(WebCore::GPUBindGroupLayout::GPUBindGroupLayout): Support ArgumentEncoderBuffers.
(WebCore::newEncoder): Deleted.
- platform/graphics/gpu/cocoa/GPUCommandBufferMetal.mm:
(WebCore::GPUCommandBuffer::GPUCommandBuffer):
- platform/graphics/gpu/cocoa/GPUProgrammablePassEncoderMetal.mm:
(WebCore::GPUProgrammablePassEncoder::setResourceAsBufferOnEncoder): Set a buffer resource on the MTLArgumentEncoder, and call useResource on the command encoder.
(WebCore::GPUProgrammablePassEncoder::setBindGroup): Parses the bind group to assign each resource on its matching MTLArgumentEncoder.
- platform/graphics/gpu/cocoa/GPURenderPassEncoderMetal.mm:
(WebCore::GPURenderPassEncoder::useResource): Resolves call from GPUProgrammablePassEncoder to call on MTLRenderCommandEncoder.
(WebCore::GPURenderPassEncoder::setVertexBuffer): ditto.
(WebCore::GPURenderPassEncoder::setFragmentBuffer): ditto.
LayoutTests:
Draw 3 triangles via WebGPU to create a green square canvas, with reference.
Each vertex is stored in its own WebGPUBuffer and they are mix/matched per bind group.
This tests multiple resource bindings in one bind group as well as setting multiple
bind groups per draw call and per command encoder.
- webgpu/buffer-resource-triangles-expected.html: Added. Reference generated using 2D canvas.
- webgpu/buffer-resource-triangles.html: Added.
1) Draw top two triangles using two bind groups in one draw call.
2) Swap first bind group and issue second draw call to draw middle triangle.
- webgpu/js/webgpu-functions.js:
(createBasicPipeline): Now handles new primitive topologies.
- 3:05 PM Changeset in webkit [240179] by
-
- 3 edits in trunk/Source/WebCore
Another attempt to fix the iOS build following <https://trac.webkit.org/changeset/240174>
(https://bugs.webkit.org/show_bug.cgi?id=193583)
Only override RenderTheme::platformFocusRingColor() when building with ENABLE(FULL_KEYBOARD_ACCESS)
enabled.
- rendering/RenderThemeIOS.h:
- rendering/RenderThemeIOS.mm:
- 3:02 PM Changeset in webkit [240178] by
-
- 4 edits in trunk
Regression(PSON) Content blockers are sometimes lost on back navigation cross-site
https://bugs.webkit.org/show_bug.cgi?id=193588
<rdar://problem/47131566>
Reviewed by Alex Christensen.
Source/WebKit:
When the WebPageProxy needs to create initialization parameters for its WebPage in the
WebContent process, it calls WebProcessProxy::addWebUserContentControllerProxy()
which calls WebUserContentControllerProxy::addProcess(). This last call is supposed to
register the WebProcessProxy with the WebUserContentControllerProxy and adding the
contentRuleLists to the WebPageCreationParameters. The issue is that if the
WebUserContentControllerProxy already knows about this WebProcessProxy, it would return
early and not populate the WebPageCreationParameters.
In PSON world, when navigating back to a page that failed to enter page cache, we reuse
the process where we previously loaded the page but re-create a new WebPage on the
WebContent process site. When this happens, WebUserContentControllerProxy would not
add the contentRuleLists to the WebPageCreationParameters and the new WebPage in the
previously-suspended process would be missing them.
- UIProcess/UserContent/WebUserContentControllerProxy.cpp:
(WebKit::WebUserContentControllerProxy::addProcess):
Tools:
Add layout test coverage.
- TestWebKitAPI/Tests/WebKitCocoa/ProcessSwapOnNavigation.mm:
- 2:55 PM Changeset in webkit [240177] by
-
- 2 edits in trunk/Source/WebCore
Attempt to fix the iOS build following <https://trac.webkit.org/changeset/240174>
(https://bugs.webkit.org/show_bug.cgi?id=193583)
Namespace qualify call to getUIColorClass().
- rendering/RenderThemeIOS.mm:
(WebCore::RenderThemeIOS::platformFocusRingColor const):
- 2:53 PM Changeset in webkit [240176] by
-
- 39 edits22 adds in trunk
SDK_VARIANT build destinations should be separate from non-SDK_VARIANT builds
https://bugs.webkit.org/show_bug.cgi?id=189553
Reviewed by Tim Horton.
.:
- Makefile.shared:
PerformanceTests:
- MediaTime/Configurations/Base.xcconfig:
- MediaTime/Configurations/SDKVariant.xcconfig: Added.
Source/bmalloc:
- Configurations/Base.xcconfig:
- Configurations/SDKVariant.xcconfig: Added.
Source/JavaScriptCore:
- Configurations/Base.xcconfig:
- Configurations/SDKVariant.xcconfig: Added.
Source/ThirdParty/ANGLE:
- Configurations/Base.xcconfig:
- Configurations/SDKVariant.xcconfig: Added.
Source/ThirdParty/libwebrtc:
- Configurations/Base.xcconfig:
- Configurations/SDKVariant.xcconfig: Added.
Source/WebCore:
- Configurations/Base.xcconfig:
- Configurations/SDKVariant.xcconfig: Added.
- Configurations/WebCore.xcconfig:
Source/WebCore/PAL:
- Configurations/Base.xcconfig:
- Configurations/PAL.xcconfig:
- Configurations/SDKVariant.xcconfig: Added.
Source/WebInspectorUI:
- Configurations/Base.xcconfig:
- Configurations/SDKVariant.xcconfig: Added.
Source/WebKit:
- Configurations/Base.xcconfig:
- Configurations/BaseTarget.xcconfig:
- Configurations/SDKVariant.xcconfig: Added.
Source/WebKitLegacy/mac:
- Configurations/Base.xcconfig:
- Configurations/SDKVariant.xcconfig: Added.
- Configurations/WebKitLegacy.xcconfig:
Source/WTF:
- Configurations/Base.xcconfig:
- Configurations/SDKVariant.xcconfig: Added.
Tools:
Add a check-webkit-style check that ensures all SDKVariant.xcconfig files match.
- ContentExtensionTester/Configurations/Base.xcconfig:
- ContentExtensionTester/Configurations/SDKVariant.xcconfig: Added.
- DumpRenderTree/mac/Configurations/Base.xcconfig:
- DumpRenderTree/mac/Configurations/SDKVariant.xcconfig: Added.
- ImageDiff/cg/Configurations/Base.xcconfig:
- ImageDiff/cg/Configurations/SDKVariant.xcconfig: Added.
- MiniBrowser/Configurations/Base.xcconfig:
- MiniBrowser/Configurations/SDKVariant.xcconfig: Added.
- MobileMiniBrowser/Configurations/Base.xcconfig:
- MobileMiniBrowser/Configurations/SDKVariant.xcconfig: Added.
- Scripts/webkitpy/style/checker.py:
(_all_categories):
(FileType):
(CheckerDispatcher._file_type):
(CheckerDispatcher._create_checker):
- Scripts/webkitpy/style/checkers/sdkvariant.py: Added.
(SDKVariantChecker):
(SDKVariantChecker.init):
(SDKVariantChecker.check):
- TestWebKitAPI/Configurations/Base.xcconfig:
- TestWebKitAPI/Configurations/SDKVariant.xcconfig: Added.
- WebEditingTester/Configurations/Base.xcconfig:
- WebEditingTester/Configurations/SDKVariant.xcconfig: Added.
- WebKitLauncher/Configurations/SDKVariant.xcconfig: Added.
- WebKitTestRunner/Configurations/Base.xcconfig:
- WebKitTestRunner/Configurations/SDKVariant.xcconfig: Added.
- lldb/lldbWebKitTester/Configurations/Base.xcconfig:
- lldb/lldbWebKitTester/Configurations/SDKVariant.xcconfig: Added.
- 2:48 PM Changeset in webkit [240175] by
-
- 11 edits in trunk/Source
Gigacages should start allocations from a slide
https://bugs.webkit.org/show_bug.cgi?id=193523
Reviewed by Mark Lam.
Source/bmalloc:
This patch makes it so that Gigacage Heaps slide the start of the
cage by some random amount. We still ensure that there is always
at least 4/2GB, on MacOS/iOS respectively, of VA space available
for allocation.
Also, this patch changes some macros into constants since macros
are the devil.
- bmalloc/Gigacage.cpp:
(Gigacage::bmalloc::protectGigacageBasePtrs):
(Gigacage::bmalloc::unprotectGigacageBasePtrs):
(Gigacage::bmalloc::runwaySize):
(Gigacage::ensureGigacage):
(Gigacage::shouldBeEnabled):
- bmalloc/Gigacage.h:
(Gigacage::name):
(Gigacage::gigacageSizeToMask):
(Gigacage::size):
(Gigacage::mask):
(Gigacage::basePtr):
(Gigacage::ensureGigacage):
(Gigacage::wasEnabled):
(Gigacage::isCaged):
(Gigacage::isEnabled):
(Gigacage::caged):
(Gigacage::disableDisablingPrimitiveGigacageIfShouldBeEnabled):
(Gigacage::canPrimitiveGigacageBeDisabled):
(Gigacage::disablePrimitiveGigacage):
(Gigacage::addPrimitiveDisableCallback):
(Gigacage::removePrimitiveDisableCallback):
- bmalloc/Heap.cpp:
(bmalloc::Heap::Heap):
- bmalloc/Sizes.h:
(bmalloc::Sizes::maskSizeClass):
(bmalloc::Sizes::maskObjectSize):
(bmalloc::Sizes::logSizeClass):
(bmalloc::Sizes::logObjectSize):
(bmalloc::Sizes::sizeClass):
(bmalloc::Sizes::objectSize):
(bmalloc::Sizes::pageSize):
Source/JavaScriptCore:
This patch changes some macros into constants since macros are the
devil.
- ftl/FTLLowerDFGToB3.cpp:
(JSC::FTL::DFG::LowerDFGToB3::caged):
- llint/LowLevelInterpreter64.asm:
Source/WTF:
This patch changes some macros into constants since macros are the
devil.
- wtf/Gigacage.cpp:
- wtf/Gigacage.h:
- 2:26 PM Changeset in webkit [240174] by
-
- 11 edits2 adds in trunk/Source/WebCore
CSS auto focus-ring outlines don't render on iOS
https://bugs.webkit.org/show_bug.cgi?id=193583
<rdar://problem/6508697>
Reviewed by Simon Fraser.
Source/WebCore:
Implement support for drawing focus rings on iOS when built with ENABLE(FULL_KEYBOARD_ACCESS)
enabled.
For now the focus ring drawing for iOS is tangled up into the Mac-specific code to draw
animated focus rings. I will fix this in <https://bugs.webkit.org/show_bug.cgi?id=193591>.
- SourcesCocoa.txt: Add file ColorIOS.mm.
- WebCore.xcodeproj/project.pbxproj: Add files ColorIOS.{h, mm}.
- platform/graphics/GraphicsContext.h:
- platform/graphics/cocoa/GraphicsContextCocoa.mm:
(WebCore::drawFocusRingAtTime):
(WebCore::GraphicsContext::drawFocusRing):
Compile focus ring drawing code when building with ENABLE(FULL_KEYBOARD_ACCESS) enabled.
This is always enabled on Mac.
- platform/ios/ColorIOS.h: Added.
- platform/ios/ColorIOS.mm: Added.
(WebCore::colorFromUIColor): Convert a UIColor to a WebCore::Color.
- rendering/RenderElement.cpp:
(WebCore::RenderElement::paintFocusRing):
- rendering/RenderImage.cpp:
(WebCore::RenderImage::paintAreaElementFocusRing):
Compile focus ring drawing code when building with ENABLE(FULL_KEYBOARD_ACCESS) enabled.
This is always enabled on Mac.
- rendering/RenderThemeIOS.h:
- rendering/RenderThemeIOS.mm:
(WebCore::RenderThemeIOS::platformFocusRingColor const): Implement this override for iOS.
(WebCore::RenderThemeIOS::supportsFocusRing const): Implement this override for iOS to always
return false - the iOS theme code does not support painting focus rings. By returning false we
will use the platform-independent, non-theme code path to draw focus rings.
Source/WebCore/PAL:
Forward declare some IPI.
- pal/spi/ios/UIKitSPI.h:
- 2:21 PM Changeset in webkit [240173] by
-
- 2 edits in trunk/Source/WebInspectorUI
Web Inspector: Uncaught Exception: TypeError: undefined is not an object (evaluating 'resource.finished')
https://bugs.webkit.org/show_bug.cgi?id=193589
Reviewed by Joseph Pecoraro.
- UserInterface/Views/NetworkTableContentView.js:
(WI.NetworkTableContentView.prototype.get supportsSave):
(WI.NetworkTableContentView.prototype._HARResources):
- 1:33 PM Changeset in webkit [240172] by
-
- 1 copy in tags/Safari-607.1.25
Tag Safari-607.1.25.
- 1:32 PM Changeset in webkit [240171] by
-
- 10 edits in trunk/Source
Unreviewed, rolling out r240160.
This broke multiple internal builds.
Reverted changeset:
"Gigacages should start allocations from a slide"
https://bugs.webkit.org/show_bug.cgi?id=193523
https://trac.webkit.org/changeset/240160
- 1:30 PM Changeset in webkit [240170] by
-
- 2 edits in branches/safari-607-branch
Cherry-pick r240118. rdar://problem/47394122
[Win] gperf command not found
https://bugs.webkit.org/show_bug.cgi?id=193538
<rdar://problem/47250549>
Reviewed by Brent Fulgham.
CMake only searches for gperf if ENABLE_WEBCORE is on. ENABLE_WEBCORE needs to be set to ON.
- Source/cmake/OptionsAppleWin.cmake:
git-svn-id: https://svn.webkit.org/repository/webkit/trunk@240118 268f45cc-cd09-0410-ab3c-d52691b4dbfc
- 1:16 PM Changeset in webkit [240169] by
-
- 2 edits in trunk/Source/WebCore
Follow-up: Switch remaining UIKit soft-linking in WebCore, WebKitLegacy over to UIKitSoftLink.{cpp,h}
<https://webkit.org/b/193568>
<rdar://problem/47381130>
- editing/cocoa/DictionaryLookup.mm:
(-[WebRevealHighlight drawHighlightContentForItem:context:]):
Fix the build by changing getUIApplicationClass() to
PAL::getUIApplicationClass().
- 12:41 PM Changeset in webkit [240168] by
-
- 30 edits in trunk
Switch remaining UIKit soft-linking in WebCore, WebKitLegacy over to UIKitSoftLink.{cpp,h}
<https://webkit.org/b/193568>
<rdar://problem/47381130>
Reviewed by Alex Christensen.
Source/WebCore:
This does the following:
- Removes local soft-linking of UIKit.
- Switches to using PAL functions for soft-linking.
- SourcesCocoa.txt: Add 4 source files to UnifiedSources.
- WebCore.xcodeproj/project.pbxproj: Remove 4 sources files
from WebCore target now that they are in UnifiedSources.
- editing/cocoa/DictionaryLookup.mm:
- editing/cocoa/HTMLConverter.mm: Move SPI to UIKitSPI.h. Remove
declarations in public headers.
(_fontForNameAndSize): Change use of getUIFontClass() to
PlatformFontClass.
(HTMLConverter::_processElement): Add cast here since compiler
finds the wrong selector for +clearColor and thinks it returns a
CIColor *object. The same cast already exists in
VideoFullscreenInterfaceAVKit.mm.
- page/cocoa/SettingsBaseCocoa.mm:
(WebCore::SettingsBase::defaultTextAutosizingEnabled):
- platform/audio/ios/MediaSessionManagerIOS.mm:
(-[WebMediaSessionHelper initWithCallback:]):
- platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
(WebCore::exernalDeviceDisplayNameForPlayer):
- platform/graphics/cocoa/FontCascadeCocoa.mm:
(WebCore::showLetterpressedGlyphsWithAdvances):
- platform/ios/DragImageIOS.mm:
(WebCore::scaleDragImage):
(WebCore::createDragImageFromImage):
(WebCore::cascadeForSystemFont):
(WebCore::createDragImageForLink):
(WebCore::createDragImageForSelection):
(WebCore::createDragImageForRange):
(WebCore::createDragImageForColor):
- platform/ios/PlatformPasteboardIOS.mm:
(WebCore::PlatformPasteboard::PlatformPasteboard):
(WebCore::PlatformPasteboard::color):
(WebCore::registerItemToPasteboard):
(WebCore::PlatformPasteboard::setColor):
- platform/ios/ThemeIOS.mm:
(WebCore::ThemeIOS::userPrefersReducedMotion const):
- platform/ios/UserAgentIOS.mm:
(WebCore::isClassic):
(WebCore::isClassicPad):
(WebCore::isClassicPhone):
- platform/ios/ValidationBubbleIOS.mm:
(-[WebValidationBubbleTapRecognizer initWithPopoverController:]):
(WebCore::ValidationBubble::ValidationBubble):
(WebCore::ValidationBubble::show):
(WebCore::fallbackViewController):
- platform/ios/VideoFullscreenInterfaceAVKit.mm:
(clearUIColor):
(allocWebAVPictureInPicturePlayerLayerViewInstance):
(WebAVPlayerLayerView_videoView):
(fallbackViewController):
(VideoFullscreenInterfaceAVKit::doSetup):
- platform/ios/WebEvent.mm:
(+[WebEvent modifierFlags]):
- platform/ios/WebItemProviderPasteboard.mm:
(allLoadableClasses):
- platform/ios/WebVideoFullscreenControllerAVKit.mm:
(VideoFullscreenControllerContext::setUpFullscreen):
- platform/network/mac/WebCoreURLResponse.mm: Add missing header
after unified sources were updated.
- rendering/RenderThemeIOS.mm:
(WebCore::contentSizeCategoryDidChange):
(WebCore::RenderThemeIOS::RenderThemeIOS):
(WebCore::RenderThemeIOS::contentSizeCategory):
(WebCore::RenderThemeIOS::systemColor const):
(WebCore::attachmentActionColor):
(WebCore::attachmentTitleColor):
(WebCore::attachmentSubtitleColor):
(WebCore::iconForAttachment):
- testing/Internals.mm:
(WebCore::Internals::userPrefersReducedMotion const):
Source/WebCore/PAL:
- pal/ios/UIKitSoftLink.h:
- pal/ios/UIKitSoftLink.mm:
- Add constants, classes and functions that were previously soft-linked in separate source files.
- UIAccessibilityAnnouncementNotification is declared as a
variable because this constant is not marked as
constin UIKit. - pal/spi/ios/UIKitSPI.h:
- Move SPI declarations from HTMLConverter.mm to here.
Source/WebKitLegacy/mac:
- WebView/WebView.mm:
- Remove unused soft-linking macros.
(-[WebUITextIndicatorData initWithImage:textIndicatorData:scale:]):
(-[WebUITextIndicatorData initWithImage:scale:]):
- Update to use exported PAL functions.
Tools:
Make sure no one re-introduces UIKit soft-linking by adding a
style rule.
- Scripts/webkitpy/style/checkers/cpp.py:
(check_language): Add softlink/uikit check.
(CppChecker): Enable softlink/uikit check.
- Scripts/webkitpy/style/checkers/cpp_unittest.py:
(CppStyleTest): Add test for softlink/uikit.
- 12:39 PM Changeset in webkit [240167] by
-
- 24 edits in trunk/Websites/bugs.webkit.org
Update Bugzilla running on bugs.webkit.org and bugs.webkit-uat.org to version 5.0.4
https://bugs.webkit.org/show_bug.cgi?id=193477
<rdar://problem/37618737>
Patch by Ling Ho <lingho@apple.com> on 2019-01-18
Reviewed by Lucas Forschler.
- Bugzilla/Config.pm:
(write_params):
(read_param_file):
- Bugzilla/Constants.pm:
- Bugzilla/DB/Sqlite.pm:
(sql_date_format):
- Bugzilla/Install/Filesystem.pm:
(update_filesystem):
(_update_old_mining_filenames):
- Bugzilla/Install/Requirements.pm:
(REQUIRED_MODULES):
- Bugzilla/JobQueue.pm:
(subprocess_worker):
(kill_worker):
- Bugzilla/Migrate.pm:
(parse_date):
- Bugzilla/Template.pm:
(_concatenate_css):
(_concatenate_js):
- Bugzilla/Util.pm:
(read_text):
(write_text):
(datetime_from):
- contrib/jb2bz.py:
(process_reply_file):
(maybe_add_attachment):
(process_text_plain):
(process_multi_part):
(process_jitterbug):
(process_mime_body): Deleted.
- docs/en/rst/conf.py:
- docs/en/rst/installing/linux.rst:
- docs/en/rst/installing/mac-os-x.rst:
- docs/en/rst/installing/migrating-from-2.inc.rst:
- docs/en/rst/installing/migrating.rst:
- docs/en/rst/installing/quick-start.rst:
- docs/en/rst/installing/upgrading-with-git.rst:
- docs/en/rst/installing/windows.rst:
- docs/en/rst/integrating/extensions.rst:
- docs/en/rst/integrating/skins.rst:
- editflagtypes.cgi:
(get_products_and_components):
- taskgraph.json:
- template/en/default/pages/release-notes.html.tmpl:
- 12:39 PM Changeset in webkit [240166] by
-
- 8 edits2 deletes in branches/safari-607-branch/Source/WebKit
REGRESSION (r236935): Holding down arrow key does not cause repeated DOM keydown events on device
https://bugs.webkit.org/show_bug.cgi?id=191408
<rdar://problem/45895135>
Rubber-stamped by Simon Fraser.
- SourcesCocoa.txt:
- UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _scrollByContentOffset:]):
- UIProcess/API/Cocoa/WKWebViewInternal.h:
- UIProcess/ios/WKContentViewInteraction.h:
- UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView setupInteraction]):
(-[WKContentView cleanupInteraction]):
(-[WKContentView scrollViewWillStartPanOrPinchGesture]):
(-[WKContentView canPerformActionForWebView:withSender:]):
(-[WKContentView keyCommands]):
(-[WKContentView _arrowKeyForWebView:]):
(-[WKContentView _didHandleKeyEvent:eventWasHandled:]):
(-[WKContentView _scrollOffsetForEvent:]):
(-[WKContentView _interpretKeyEvent:isCharEvent:]):
(-[WKContentView isScrollableForKeyboardScrollViewAnimator:]): Deleted.
(-[WKContentView keyboardScrollViewAnimator:distanceForIncrement:]): Deleted.
(-[WKContentView keyboardScrollViewAnimatorWillScroll:]): Deleted.
(-[WKContentView keyboardScrollViewAnimatorDidFinishScrolling:]): Deleted.
- UIProcess/ios/WKKeyboardScrollingAnimator.h: Removed.
- UIProcess/ios/WKKeyboardScrollingAnimator.mm: Removed.
- UIProcess/ios/WKPDFView.mm:
(-[WKPDFView dealloc]):
(-[WKPDFView web_handleKeyEvent:]):
(-[WKPDFView web_initWithFrame:webView:mimeType:]):
- WebKit.xcodeproj/project.pbxproj:
Revert WKKeyboardScrollingAnimator.
- 12:38 PM Changeset in webkit [240165] by
-
- 2 edits in trunk/Tools
[GTK] install-dependencies tool looks for libunistring-dev for dnf
https://bugs.webkit.org/show_bug.cgi?id=193584
Change "-dev" to "-devel".
Reviewed by Michael Catanzaro.
- gtk/install-dependencies:
- 12:28 PM Changeset in webkit [240164] by
-
- 5 edits in releases/WebKitGTK/webkit-2.22
Merge r240158 - FetchResponse::url should return the empty string for tainted responses
https://bugs.webkit.org/show_bug.cgi?id=193553
Reviewed by Youenn Fablet.
LayoutTests/imported/w3c:
- web-platform-tests/fetch/api/basic/mode-no-cors.sub.any.js:
(fetchNoCors):
Source/WebCore:
Check whether the response is tainted in FetchResponse::url, to match
the behavior described in https://fetch.spec.whatwg.org/#concept-filtered-response-opaque.
- Modules/fetch/FetchResponse.cpp:
(WebCore::FetchResponse::url const):
LayoutTests:
- http/wpt/fetch/response-opaque-clone.html:
- 12:12 PM Changeset in webkit [240163] by
-
- 4 edits in trunk/Tools
webkitpy: Add iPhone and iPad ports
https://bugs.webkit.org/show_bug.cgi?id=193537
<rdar://problem/47353390>
Reviewed by Lucas Forschler.
Add --iphone-simulator and --ipad-simulator commands to run-webkit-tests which separate iPhone and iPad into separate
ports. Note that this separation is optional, the --ios-simulator command remains.
- Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
(RunTest.test_device_type_test_division): Drive-by fix.
(RunTest.test_device_type_specific_listing): Ditto.
(RunTest.test_ipad_test_division):
(RunTest):
(RunTest.test_ipad_listing):
- Scripts/webkitpy/port/factory.py:
(platform_options): Add --iphone-simulator and --ipad-simulator flags.
(PortFactory):
- Scripts/webkitpy/port/ios_simulator.py:
(IPhoneSimulatorPort):
(IPadSimulatorPort):
- 12:07 PM Changeset in webkit [240162] by
-
- 12 edits2 adds in trunk
Implement asynchronous frame scrolling for iOS
https://bugs.webkit.org/show_bug.cgi?id=193539
<rdar://problem/47379873>
Reviewed by Simon Fraser.
Source/WebCore:
This patch implements UIScrollView based frame scrolling on iOS, enabled by the "Async Frame Scrolling"
internal setting (still off by default).
- page/scrolling/ios/ScrollingTreeFrameScrollingNodeIOS.h:
(WebCore::ScrollingTreeFrameScrollingNodeIOS::scrollLayer const): Deleted.
- page/scrolling/ios/ScrollingTreeFrameScrollingNodeIOS.mm:
(WebCore::ScrollingTreeFrameScrollingNodeIOS::setScrollPosition):
Export ScrollingTreeFrameScrollingNodeIOS.
- rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::visibleRectForLayerFlushing const):
(WebCore::RenderLayerCompositor::frameViewDidChangeSize):
(WebCore::RenderLayerCompositor::updateScrollLayerClipping):
If we don't have a separate clip layer, just resize and position the scroll layer.
(WebCore::RenderLayerCompositor::updateRootLayerPosition):
(WebCore::RenderLayerCompositor::updateOverflowControlsLayers):
(WebCore::RenderLayerCompositor::ensureRootLayer):
Use GraphicsLayer::Type::Scrolling as the scroll layer type when async frame scrolling is enabled.
Don't create a separate clip layer since the scroll layer will handle clipping.
- rendering/RenderLayerCompositor.h:
Source/WebKit:
Add a scrolling node type that can handle UIScrollView backed frames.
It basically just instantiates and forwards to the existing ScrollingTreeScrollingNodeDelegateIOS.
- UIProcess/RemoteLayerTree/RemoteScrollingTree.cpp:
(WebKit::RemoteScrollingTree::createScrollingTreeNode):
- UIProcess/RemoteLayerTree/ios/RemoteScrollingCoordinatorProxyIOS.mm:
(WebKit::RemoteScrollingCoordinatorProxy::connectStateNodeLayers):
- UIProcess/RemoteLayerTree/ios/ScrollingTreeFrameScrollingNodeRemoteIOS.h: Added.
- UIProcess/RemoteLayerTree/ios/ScrollingTreeFrameScrollingNodeRemoteIOS.mm: Added.
(WebKit::ScrollingTreeFrameScrollingNodeRemoteIOS::create):
(WebKit::ScrollingTreeFrameScrollingNodeRemoteIOS::ScrollingTreeFrameScrollingNodeRemoteIOS):
(WebKit::ScrollingTreeFrameScrollingNodeRemoteIOS::~ScrollingTreeFrameScrollingNodeRemoteIOS):
(WebKit::ScrollingTreeFrameScrollingNodeRemoteIOS::commitStateBeforeChildren):
(WebKit::ScrollingTreeFrameScrollingNodeRemoteIOS::commitStateAfterChildren):
(WebKit::ScrollingTreeFrameScrollingNodeRemoteIOS::updateLayersAfterAncestorChange):
(WebKit::ScrollingTreeFrameScrollingNodeRemoteIOS::scrollPosition const):
(WebKit::ScrollingTreeFrameScrollingNodeRemoteIOS::setScrollLayerPosition):
(WebKit::ScrollingTreeFrameScrollingNodeRemoteIOS::updateLayersAfterDelegatedScroll):
- WebKit.xcodeproj/project.pbxproj:
LayoutTests:
- platform/ios-wk2/compositing/tiling/tiled-drawing-async-frame-scrolling-expected.txt:
- 11:47 AM Changeset in webkit [240161] by
-
- 9 edits in trunk
Regression(PSON) Scroll position is not always restored properly when navigating back
https://bugs.webkit.org/show_bug.cgi?id=193578
<rdar://problem/47386331>
Reviewed by Tim Horton.
Source/WebKit:
Fix issues causing the scroll position to not be restored at all (or incorrectly) when
navigating back cross-site with PSON enabled. Also make sure that the swipe gesture
snapshot really stays up until we've restored the scroll position.
Note that even after those changes, I can still sometimes reproduce a white flash when
swiping back to Google search results (scroll position being correct now). This is
tracked by <rdar://problem/47071684> and happens even if I disable PSON entirely.
- Shared/SessionState.cpp:
(WebKit::FrameState::encode const):
(WebKit::FrameState::decode):
- Shared/SessionState.h:
- WebProcess/WebCoreSupport/SessionStateConversion.cpp:
(WebKit::toFrameState):
(WebKit::applyFrameState):
obscuredInsets is present on the HistoryItem in the WebProcess but was never passed to
or stored by the UIProcess on the WebBackForwardListItem. obscuredInsets is needed to
properly restore the scrollPosition (position was 70px off on my iPad without this).
With PSON enabled, if you swipe back cross-process and the previous page was not put
into PageCache, then the HistoryItem is gone on the WebProcess side. What happens is
that the UIProcess sends its WebBackForwardListItem to the WebProcess, which restores
the HistoryItem there, and then asks it to load it. The obscuredInsets was getting lost
in the process since the UIProcess never knew about it.
- UIProcess/Cocoa/ViewGestureController.cpp:
(WebKit::ViewGestureController::didReachMainFrameLoadTerminalState):
Drop logic that was causing the ViewGestureController to not wait for the scroll position
to be restored before taking down the snapshot, when UI-side compositing is enabled.
If you look at the comment above the code, you'll see that the code in question was meant
to impact only the non-UI side compositing code path. As a matter of fact, when the code
was reviewed at https://bugs.webkit.org/show_bug.cgi?id=151224, it was protected by a
#if PLATFORM(MAC), before getting modified the wrong way before landing. In practice, we
would have often restored the scroll position by the time the load is finished so it would
not cause a flash in most cases. However, with PSON enabled and the layer tree freezing we
do on process-swap, the first post-scroll restoration layer tree commit may now occur a
little bit later and we would lose the race more often.
- UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::updateBackForwardItem):
- UIProcess/WebProcessProxy.h:
When adding PageCache support to PSON, we used to navigate the "suspended" page to
about:blank. This would lead to unwanted WebProcessProxy::updateBackForwardItem()
calls from the WebProcess which we wanted to ignore. We thus added logic to ignore
updateBackForwardItem() IPC from the old WebProcess after a swap. The issue with this
is that we sometimes miss/ignore legit updates to the HistoryItem from the old process,
in particular with regards to the scroll position and the pageScaleFactor. So if you
swiped and then quickly enough did a cross-site navigation, the UIProcess'
WebBackForwardList would not get updated with the latest scroll position and we would
thus fail to restore it later on. To address the issue, we now stop ignoring updates
from the old WebProcess after a swap. This logic is no longer needed since we no longer
navigate the old page to about:blank after a swap, we merely suspend it "in place".
Tools:
Add API test coverage.
- TestWebKitAPI/Tests/WebKitCocoa/ProcessSwapOnNavigation.mm:
- 11:38 AM Changeset in webkit [240160] by
-
- 10 edits in trunk/Source
Gigacages should start allocations from a slide
https://bugs.webkit.org/show_bug.cgi?id=193523
Reviewed by Mark Lam.
Source/bmalloc:
This patch makes it so that Gigacage Heaps slide the start of the
cage by some random amount. We still ensure that there is always
at least 4/2GB, on MacOS/iOS respectively, of VA space available
for allocation.
Also, this patch changes some macros into constants since macros
are the devil.
- bmalloc/Gigacage.cpp:
(Gigacage::bmalloc::protectGigacageBasePtrs):
(Gigacage::bmalloc::unprotectGigacageBasePtrs):
(Gigacage::bmalloc::runwaySize):
(Gigacage::ensureGigacage):
(Gigacage::shouldBeEnabled):
- bmalloc/Gigacage.h:
(Gigacage::name):
(Gigacage::gigacageSizeToMask):
(Gigacage::size):
(Gigacage::mask):
(Gigacage::basePtr):
(Gigacage::ensureGigacage):
(Gigacage::wasEnabled):
(Gigacage::isCaged):
(Gigacage::caged):
(Gigacage::disableDisablingPrimitiveGigacageIfShouldBeEnabled):
(Gigacage::disablePrimitiveGigacage):
(Gigacage::addPrimitiveDisableCallback):
(Gigacage::removePrimitiveDisableCallback):
- bmalloc/Heap.cpp:
(bmalloc::Heap::Heap):
- bmalloc/Sizes.h:
(bmalloc::Sizes::maskSizeClass):
(bmalloc::Sizes::maskObjectSize):
(bmalloc::Sizes::logSizeClass):
(bmalloc::Sizes::logObjectSize):
(bmalloc::Sizes::sizeClass):
(bmalloc::Sizes::objectSize):
(bmalloc::Sizes::pageSize):
Source/JavaScriptCore:
This patch changes some macros into constants since macros are the
devil.
- llint/LowLevelInterpreter64.asm:
Source/WTF:
This patch changes some macros into constants since macros are the
devil.
- wtf/Gigacage.cpp:
- wtf/Gigacage.h:
- 11:10 AM Changeset in webkit [240159] by
-
- 4 edits in trunk/Source/WebKit
Remove some last vestiges of assisted node terminology in WebKit
https://bugs.webkit.org/show_bug.cgi?id=193572
Reviewed by Tim Horton.
- UIProcess/API/Cocoa/_WKFocusedElementInfo.h:
- UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView gestureRecognizerShouldBegin:]):
(-[WKContentView textInteractionGesture:shouldBeginAtPoint:]):
(-[WKContentView _showKeyboard]):
(-[WKContentView _hideKeyboard]):
Add a FIXME about invoking
-reloadInputViewson watchOS when the focused element is blurred.
(-[WKContentView _elementDidFocus:userIsInteracting:blurPreviousNode:changingActivityState:userObject:]):
(-[WKContentView _elementDidBlur]):
(-[WKContentView updateCurrentFocusedElementInformation:]):
(-[WKContentView _startAssistingKeyboard]): Deleted.
(-[WKContentView _stopAssistingKeyboard]): Deleted.
- WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::getFocusedElementInformation):
- 10:44 AM Changeset in webkit [240158] by
-
- 6 edits in trunk
FetchResponse::url should return the empty string for tainted responses
https://bugs.webkit.org/show_bug.cgi?id=193553
Reviewed by Youenn Fablet.
LayoutTests/imported/w3c:
- web-platform-tests/fetch/api/basic/mode-no-cors.sub.any.js:
(fetchNoCors):
Source/WebCore:
Check whether the response is tainted in FetchResponse::url, to match
the behavior described in https://fetch.spec.whatwg.org/#concept-filtered-response-opaque.
- Modules/fetch/FetchResponse.cpp:
(WebCore::FetchResponse::url const):
LayoutTests:
- http/wpt/fetch/response-opaque-clone.html:
- 10:37 AM Changeset in webkit [240157] by
-
- 1 edit1 delete in trunk/LayoutTests
webkitpy: Implement device type specific expected results (Follow-up fix)
https://bugs.webkit.org/show_bug.cgi?id=192162
<rdar://problem/46345449>
Unreviewed test gardening.
- tiled-drawing/ios/iphone7: Removed.
- 10:19 AM Changeset in webkit [240156] by
-
- 19 edits4 copies in trunk
Add a new SPI to request for cache storage quota increase
https://bugs.webkit.org/show_bug.cgi?id=193323
Reviewed by Alex Christensen.
Source/WebKit:
Add a delegate on the WebSiteDataStore for WebKit to ask for quota update.
The current SPI is currently CacheStorage specific but future work should
make it so that other storage like IDB use the same mechanism.
By default, quota remains unchanged if delegate is not implemented.
- NetworkProcess/NetworkProcess.cpp:
- UIProcess/API/Cocoa/WKStorageQuotaDelegatePrivate.h: Added.
- UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
(WebsiteDataStoreQuotaManager::WebsiteDataStoreQuotaManager):
(-[WKWebsiteDataStore _quotaDelegate]):
(-[WKWebsiteDataStore set_quotaDelegate:]):
- UIProcess/API/Cocoa/WKWebsiteDataStoreInternal.h:
- UIProcess/API/Cocoa/WKWebsiteDataStorePrivate.h:
- UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::requestCacheStorageSpace):
- UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::WebsiteDataStore):
- UIProcess/WebsiteData/WebsiteDataStore.h:
(WebKit::WebsiteDataStore::quotaManager):
(WebKit::WebsiteDataStore::setQuotaManager):
- UIProcess/WebsiteData/WebsiteDataStoreQuotaManager.h: Added.
(WebKit::WebsiteDataStoreQuotaManager::~WebsiteDataStoreQuotaManager):
(WebKit::WebsiteDataStoreQuotaManager::requestCacheStorageSpace):
- WebKit.xcodeproj/project.pbxproj:
Tools:
Implement WebsiteDataStore quota delegate to handle quota requests.
By default, do not update quota.
Update quota if test calls the new testRunner.allowCacheStorageQuotaIncrease method.
- WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
- WebKitTestRunner/InjectedBundle/TestRunner.cpp:
(WTR::TestRunner::allowCacheStorageQuotaIncrease):
- WebKitTestRunner/InjectedBundle/TestRunner.h:
- WebKitTestRunner/TestController.cpp:
(WTR::TestController::allowCacheStorageQuotaIncrease):
- WebKitTestRunner/TestController.h:
- WebKitTestRunner/TestInvocation.cpp:
(WTR::TestInvocation::didReceiveSynchronousMessageFromInjectedBundle):
- WebKitTestRunner/cocoa/TestControllerCocoa.mm:
(-[CacheStorageQuotaManager init]):
(-[CacheStorageQuotaManager _requestCacheStorageSpace:frameOrigin:quota:currentSize:spaceRequired:decisionHandler:]):
(WTR::initializeWebViewConfiguration):
(WTR::TestController::cocoaResetStateToConsistentValues):
(WTR::TestController::allowCacheStorageQuotaIncrease):
LayoutTests:
Use new testRunner method to bump the cache quota and verify adding a
cache entry works when bumping the cache quota.
- http/wpt/cache-storage/cache-quota.any.js:
(promise_test):
- 10:16 AM Changeset in webkit [240155] by
-
- 2 edits in trunk/Tools
[ews-build] Pass owner to the triggered builds
https://bugs.webkit.org/show_bug.cgi?id=193574
Reviewed by Lucas Forschler.
- BuildSlaveSupport/ews-build/factories.py:
(BuildFactory.propertiesToPassToTriggers):
- 10:14 AM Changeset in webkit [240154] by
-
- 2 edits in trunk/Tools
[ews-app] Send patch author as a build property
https://bugs.webkit.org/show_bug.cgi?id=193560
Reviewed by Lucas Forschler.
- BuildSlaveSupport/ews-app/ews/fetcher.py:
(BugzillaPatchFetcher.fetch):
- 9:54 AM Changeset in webkit [240153] by
-
- 2 edits in trunk/Tools
webkitpy: Implement device type specific expected results (Follow-up fix)
https://bugs.webkit.org/show_bug.cgi?id=192162
<rdar://problem/46345449>
Unreviewed infrastructure fix.
- Scripts/webkitpy/layout_tests/controllers/manager.py:
(Manager.run): Don't boot a device if no tests are available to run on it.
- 9:19 AM Changeset in webkit [240152] by
-
- 5 edits2 adds in trunk
A track source should be unmuted whenever reenabled after setDirection changes
https://bugs.webkit.org/show_bug.cgi?id=193554
<rdar://problem/47366196>
Reviewed by Eric Carlson.
Source/WebCore:
Ensure that track gets unmuted after being fired as part of track event.
Test is triggering some existing issues with MediaPlayerPrivateMediaStreamAVFObjC.
Given the enqueuing of samples happens in a different frame than the thread used to update media stream and the active video track,
some enqueued samples might not be from the right active video track or there might be no active video track.
Test: webrtc/video-setDirection.html
- Modules/mediastream/libwebrtc/LibWebRTCMediaEndpoint.cpp:
(WebCore::LibWebRTCMediaEndpoint::fireTrackEvent):
- Modules/mediastream/libwebrtc/LibWebRTCMediaEndpoint.h:
- platform/graphics/avfoundation/objc/MediaPlayerPrivateMediaStreamAVFObjC.mm:
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::enqueueVideoSample):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::requestNotificationWhenReadyForVideoData):
LayoutTests:
- webrtc/video-setDirection-expected.txt: Added.
- webrtc/video-setDirection.html: Added.
- 8:54 AM Changeset in webkit [240151] by
-
- 2 edits in trunk/Tools
webkitpy: Implement device type specific expected results (Follow-up fix)
https://bugs.webkit.org/show_bug.cgi?id=192162
<rdar://problem/46345449>
Unreviewed infrastructure fix.
- Scripts/webkitpy/api_tests/manager.py:
(Manager._initialize_devices): Use DEVICE_TYPE when running API tests.
- 8:37 AM Changeset in webkit [240150] by
-
- 24 edits62 adds43 deletes in trunk
webkitpy: Implement device type specific expected results (Part 2)
https://bugs.webkit.org/show_bug.cgi?id=192162
<rdar://problem/46345449>
Rubber-stamped by Aakash Jain.
Tools:
This change uses device_type to select a different set of TestExpectation files based on the type of device running
tests. This requires multiple baseline search paths to be used in a single instantiation of run-webkit-tests. Note
that while multiple baseline search paths are used, any single test will only ever be run on a single device type.
- Scripts/webkitpy/layout_tests/controllers/layout_test_finder.py:
(LayoutTestFinder.find_tests): Pass device_type through.
- Scripts/webkitpy/layout_tests/controllers/layout_test_finder_unittest.py:
(LayoutTestFinderTests.test_touched_but_skipped_test): Allow device_type to be passed in.
- Scripts/webkitpy/layout_tests/controllers/manager.py:
(Manager.init): Make _expectations an OrderedDict indexed by device_type, add current_device_type.
(Manager._collect_tests): Pass device type to test finder.
(Manager._needs_web_platform_test):
(Manager._prepare_lists): Index _expectations by device type.
(Manager._test_input_for_file):
(Manager._test_is_slow): Ditto.
(Manager._test_should_dump_jsconsolelog_in_stderr):
(Manager._test_should_dump_jsconsolelog_in_stderr): Ditto.
(Manager._get_test_inputs):
(Manager._update_worker_count):
(Manager._set_up_run): Use current_device_type instead of passing device_type in.
(Manager.run): Re-write function to gather and parse expectations from multiple baseline search paths. In the event
that a test can be run on multiple device types, the first device in the device type list which is eligible will be
the one which runs the test.
(Manager._run_test_subset):
(Manager._run_tests): Index _expectations by current_device_type.
(Manager._print_expectation_line_for_test): Ditto.
(Manager._print_expectations_for_subset): Ditto.
(Manager.print_expectations): Re-write function to gather and parse expectations from multiple baseline search paths.
Unlike run, this function will behave the same regardless of what devices are available.
(Manager._custom_device_for_test): Deleted.
- Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
(ManagerTest.test_uses_custom_device): Deleted.
(ManagerTest.test_uses_custom_device.get_manager): Deleted.
- Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
(JSONLayoutResultsGenerator.init): Pass expectations as a dictionary indexed by device type.
(JSONLayoutResultsGenerator._insert_failure_summaries): The number of fixable tests must take into consideration
all of the expectations.
- Scripts/webkitpy/layout_tests/lint_test_expectations_unittest.py:
(FakePort.skipped_layout_tests): Allow device_type to be passed in.
- Scripts/webkitpy/layout_tests/models/test_expectations.py:
(TestExpectations.init): Allow device_type to be passed through.
(TestExpectations.parse_all_expectations): Pick expectations_dict by device type.
- Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Allow device_type to be passed in.
- Scripts/webkitpy/layout_tests/models/test_run_results.py:
(summarize_results):
- Scripts/webkitpy/layout_tests/models/test_run_results_unittest.py:
(summarized_results): Expectations are stored in a dictionary indexed by device_type. When determining the expectation
a test ran with, one must check all expectations to find the first non-skip expectation.
- Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
(RunTest.test_device_type_test_division):
(RunTest):
(RunTest.test_device_type_specific_listing):
- Scripts/webkitpy/port/base.py:
(Port): Add default values for DEVICE_TYPE and DEFAULT_DEVICE_TYPES.
(Port.supported_device_types): The result of this function will change based on what devices are available.
- Scripts/webkitpy/port/device_port.py:
(DevicePort._device_type_with_version): Replace DEFAULT_DEVICE_TYPE with DEVICE_TYPE.
(DevicePort.default_child_processes): DEVICE_TYPE is a general type, so the built-in comparison of the DeviceType object
can be used.
(DevicePort.supported_device_types): Generates a list of device types determined by the available devices, if there are
any, or the DEFAULT_DEVICE_TYPES of the port.
- Scripts/webkitpy/port/ios.py:
(IOSPort): Change DEFAULT_DEVICE_TYPE to DEVICE_TYPE.
- Scripts/webkitpy/port/ios_simulator.py:
(IOSSimulatorPort): Set the DEFAULT_DEVICE_TYPES to the devices currently used to run layout tests.
- Scripts/webkitpy/port/watch.py:
(WatchPort): Change DEFAULT_DEVICE_TYPE to DEVICE_TYPE.
- Scripts/webkitpy/port/watch_simulator.py:
(WatchSimulatorPort): Set the DEFAULT_DEVICE_TYPES to the device currently used to run layout tests.
- Scripts/webkitpy/xcode/device_type.py:
(DeviceType):
(DeviceType.hash): Allow DeviceType to be used as a dictionary index.
LayoutTests:
Move iPad and iPhone 7 tests to their respective platform specific directories.
- TestExpectations:
- fast/events/touch/ios/iphone7: Removed.
- fast/forms/ios/ipad: Removed.
- fast/text-autosizing/ios/ipad: Removed.
- fast/viewport/ios/ipad: Removed.
- media/controls/ipad: Removed.
- media/modern-media-controls/media-documents/ipad: Removed.
- media/modern-media-controls/pip-support/ipad: Removed.
- platform/ios-wk1/TestExpectations:
- platform/ios/TestExpectations:
- platform/ipad: Added.
- platform/ipad/fast: Added.
- platform/ipad/fast/forms: Added.
- platform/ipad/fast/forms/choose-select-option-expected.txt: Added.
- platform/ipad/fast/forms/choose-select-option.html: Added.
- platform/ipad/fast/forms/focus-input-via-button-expected.txt: Added.
- platform/ipad/fast/forms/focus-input-via-button.html: Added.
- platform/ipad/fast/forms/multiple-select-updates-renderer-expected.txt: Added.
- platform/ipad/fast/forms/multiple-select-updates-renderer.html: Added.
- platform/ipad/fast/forms/select-form-run-twice-expected.txt: Added.
- platform/ipad/fast/forms/select-form-run-twice.html: Added.
- platform/ipad/fast/forms/select-with-title-expected.txt: Added.
- platform/ipad/fast/forms/select-with-title.html: Added.
- platform/ipad/fast/forms/unfocus-inside-fixed-hittest-expected.txt: Added.
- platform/ipad/fast/forms/unfocus-inside-fixed-hittest.html: Added.
- platform/ipad/fast/text-autosizing: Added.
- platform/ipad/fast/text-autosizing/programmatic-text-size-adjust-expected.txt: Added.
- platform/ipad/fast/text-autosizing/programmatic-text-size-adjust.html: Added.
- platform/ipad/fast/text-autosizing/text-size-adjust-inline-style-expected.html: Added.
- platform/ipad/fast/text-autosizing/text-size-adjust-inline-style.html: Added.
- platform/ipad/fast/viewport: Added.
- platform/ipad/fast/viewport/empty-meta-expected.txt: Added.
- platform/ipad/fast/viewport/empty-meta.html: Added.
- platform/ipad/fast/viewport/meta-viewport-ignored-expected.txt: Added.
- platform/ipad/fast/viewport/meta-viewport-ignored.html: Added.
- platform/ipad/fast/viewport/viewport-overriden-by-minimum-effective-width-if-ignore-meta-viewport-expected.txt: Added.
- platform/ipad/fast/viewport/viewport-overriden-by-minimum-effective-width-if-ignore-meta-viewport.html: Added.
- platform/ipad/fast/viewport/viewport-unchanged-by-minimum-effective-width-if-not-ignore-meta-viewport-expected.txt: Added.
- platform/ipad/fast/viewport/viewport-unchanged-by-minimum-effective-width-if-not-ignore-meta-viewport.html: Added.
- platform/ipad/fast/viewport/width-is-device-width-expected.txt: Added.
- platform/ipad/fast/viewport/width-is-device-width.html: Added.
- platform/ipad/media: Added.
- platform/ipad/media/controls: Added.
- platform/ipad/media/controls/close-page-with-picture-in-picture-video-assertion-failure-expected.txt: Added.
- platform/ipad/media/controls/close-page-with-picture-in-picture-video-assertion-failure.html: Added.
- platform/ipad/media/controls/resources: Added.
- platform/ipad/media/controls/resources/picture-in-picture.html: Added.
- platform/ipad/media/modern-media-controls: Added.
- platform/ipad/media/modern-media-controls/media-documents: Added.
- platform/ipad/media/modern-media-controls/media-documents/media-document-audio-ios-sizing-expected.txt: Added.
- platform/ipad/media/modern-media-controls/media-documents/media-document-audio-ios-sizing.html: Added.
- platform/ipad/media/modern-media-controls/media-documents/media-document-video-ios-sizing-expected.txt: Added.
- platform/ipad/media/modern-media-controls/media-documents/media-document-video-ios-sizing.html: Added.
- platform/ipad/media/modern-media-controls/pip-support: Added.
- platform/ipad/media/modern-media-controls/pip-support/pip-support-enabled-expected.txt: Added.
- platform/ipad/media/modern-media-controls/pip-support/pip-support-enabled.html: Added.
- platform/ipad/media/modern-media-controls/pip-support/pip-support-tap-expected.txt: Added.
- platform/ipad/media/modern-media-controls/pip-support/pip-support-tap.html: Added.
- platform/iphone-7: Added.
- platform/iphone-7/fast: Added.
- platform/iphone-7/fast/events: Added.
- platform/iphone-7/fast/events/touch: Added.
- platform/iphone-7/fast/events/touch/force-press-event-expected.txt: Added.
- platform/iphone-7/fast/events/touch/force-press-event.html: Added.
- platform/iphone-7/fast/events/touch/force-press-on-link-expected.txt: Added.
- platform/iphone-7/fast/events/touch/force-press-on-link.html: Added.
- platform/iphone-7/fast/events/touch/prevent-default-touchmove-prevents-scrolling-expected.txt: Added.
- platform/iphone-7/fast/events/touch/prevent-default-touchmove-prevents-scrolling.html: Added.
- platform/iphone-7/tiled-drawing: Added.
- platform/iphone-7/tiled-drawing/compositing-layers-deep-color-expected.txt: Added.
- platform/iphone-7/tiled-drawing/compositing-layers-deep-color.html: Added.
- platform/mac-wk2/TestExpectations:
- platform/mac/TestExpectations:
- tiled-drawing/ios: Removed.
- 8:12 AM Changeset in webkit [240149] by
-
- 2 edits in trunk/Tools
[WPE] run-qt-wpe-minibrowser needlessly expects a --wpe argument
https://bugs.webkit.org/show_bug.cgi?id=193575
Reviewed by Michael Catanzaro.
Allow the script to run without explicit port argument because
it's specific to the WPE port anyway.
- Scripts/run-qt-wpe-minibrowser: Manually insert --wpe in argv
and simplify the rest of the code accordingly.
- 7:59 AM Changeset in webkit [240148] by
-
- 11 edits in trunk
[GStreamer][EME][ClearKey] Request keys from CDMInstance rather than passing via bus messages
https://bugs.webkit.org/show_bug.cgi?id=192229
Reviewed by Xabier Rodriguez-Calvar.
Source/WebCore:
Covered by existing tests.
- platform/encryptedmedia/clearkey/CDMClearKey.cpp:
(WebCore::parseLicenseFormat): There is a defect in some C++11
compiles where they will copy this return value since the type
doesn't exactly match. Force a move with WTFMove.
- platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp:
(WebCore::MediaPlayerPrivateGStreamerBase::dispatchDecryptionKey):
Deleted. No longer used by anything.
- platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.h: Ditto.
- platform/graphics/gstreamer/eme/WebKitClearKeyDecryptorGStreamer.cpp:
Rename these methods to avoid "namespacing names".
(webkit_media_clear_key_decrypt_class_init):
(finalize):
(handleKeyResponse): This is a temporary fix, we need some more
reorganisation to be full driven by CDMInstance APIs for decryption.
(findAndSetKey):
(decrypt):
(webKitMediaClearKeyDecryptorFinalize): Deleted.
(webKitMediaClearKeyDecryptorHandleKeyResponse): Deleted.
(webKitMediaClearKeyDecryptorFindAndSetKey): Deleted.
(webKitMediaClearKeyDecryptorDecrypt): Deleted.
- platform/graphics/gstreamer/eme/WebKitCommonEncryptionDecryptorGStreamer.cpp: Ditto.
(webkit_media_common_encryption_decrypt_class_init):
(finalize):
(transformCaps):
(transformInPlace):
(isCDMInstanceAvailable):
(sinkEventHandler):
(queryHandler):
(changeState):
(setContext):
(webKitMediaCommonEncryptionDecryptorFinalize): Deleted.
(webkitMediaCommonEncryptionDecryptTransformCaps): Deleted.
(webkitMediaCommonEncryptionDecryptTransformInPlace): Deleted.
(webkitMediaCommonEncryptionDecryptIsCDMInstanceAvailable): Deleted.
(webkitMediaCommonEncryptionDecryptSinkEventHandler): Deleted.
(webkitMediaCommonEncryptionDecryptorQueryHandler): Deleted.
(webKitMediaCommonEncryptionDecryptorChangeState): Deleted.
(webKitMediaCommonEncryptionDecryptorSetContext): Deleted.
- platform/graphics/gstreamer/eme/WebKitCommonEncryptionDecryptorGStreamer.h:
- platform/graphics/gstreamer/mse/MediaPlayerPrivateGStreamerMSE.cpp:
(WebCore::MediaPlayerPrivateGStreamerMSE::attemptToDecryptWithLocalInstance):
Deleted. No longer passing key information over bus messages.
- platform/graphics/gstreamer/mse/MediaPlayerPrivateGStreamerMSE.h:
Tools:
- MiniBrowser/wpe/main.cpp:
(main): Enable encrypted-media by default.
- 7:40 AM Changeset in webkit [240147] by
-
- 2 edits in trunk/Tools
[Flatpak] Missing KeyboardInterrupt exception handler
https://bugs.webkit.org/show_bug.cgi?id=193570
Reviewed by Michael Catanzaro.
- flatpak/flatpakutils.py:
(WebkitFlatpak.run_in_sandbox): Basic KeyboardInterrupt handling.
- 6:42 AM Changeset in webkit [240146] by
-
- 5 edits2 adds in trunk
[LFC][BFC][MarginCollapsing] Collapsing through should not ignore floats.
https://bugs.webkit.org/show_bug.cgi?id=193564
Reviewed by Antti Koivisto.
Source/WebCore:
Float boxes prevent collapsing through.
Test: fast/block/float/float-in-descendant-formatting-context.html
- layout/blockformatting/BlockMarginCollapse.cpp:
(WebCore::Layout::BlockFormattingContext::MarginCollapse::marginsCollapseThrough):
Tools:
- LayoutReloaded/misc/LFC-passing-tests.txt:
LayoutTests:
- fast/block/float/float-in-descendant-formatting-context-expected.txt: Added.
- fast/block/float/float-in-descendant-formatting-context.html: Added.
- 6:39 AM Changeset in webkit [240145] by
-
- 5 edits in trunk
[LFC] Do not skip float boxes that are not part of the current formatting context when computing bottom.
https://bugs.webkit.org/show_bug.cgi?id=193562
Reviewed by Antti Koivisto.
Source/WebCore:
The current floating context's (float) boxes could belong to descendant formatting contexts.
We need to include them as well when computing height (bottom) (we essentially need to skip ancestor floats only).
<div id=container style="overflow: hidden"><div>foo<div style="float: left">bar</div></div></div>
While computing the height for "container", the float box needs to be taken into account even though
it is part of another (descendant) formatting context (the inline formatting context established by its parent div).
- layout/floats/FloatingState.cpp:
(WebCore::Layout::FloatingState::bottom const):
- layout/floats/FloatingState.h:
(WebCore::Layout::FloatingState::FloatItem::isDescendantOfFormattingRoot const):
(WebCore::Layout::FloatingState::FloatItem::inFormattingContext const): Deleted.
Tools:
- LayoutReloaded/misc/LFC-passing-tests.txt:
- 6:25 AM Changeset in webkit [240144] by
-
- 5 edits in trunk
[LFC][BFC] Check for inflow children while computing height for block formatting context roots.
https://bugs.webkit.org/show_bug.cgi?id=193555
Reviewed by Antti Koivisto.
Source/WebCore:
This patch also extends areEssentiallyEqual to 0.125px to be able to match (essentially equal) inline runs.
- layout/FormattingContextGeometry.cpp:
(WebCore::Layout::contentHeightForFormattingContextRoot):
- layout/Verification.cpp:
(WebCore::Layout::areEssentiallyEqual):
- page/FrameViewLayoutContext.cpp:
(WebCore::layoutUsingFormattingContext):
Tools:
- LayoutReloaded/misc/LFC-passing-tests.txt:
- 6:12 AM Changeset in webkit [240143] by
-
- 2 edits in trunk/Source/WebCore
[WebAudio] Release the AudioDestination when uninitializing DefaultAudioDestinationNode
https://bugs.webkit.org/show_bug.cgi?id=192590
Patch by Yacine Bandou <yacine.bandou@softathome.com> on 2019-01-18
Reviewed by Philippe Normand.
When we uninitialize DefaultAudioDestinationNode, the AudioDestination is stopped but not destroyed.
On some platforms the resources are allocated and released with the AudioDestination, thus when we uninitialize
DefaultAudioDestinationNode we don't release resources because the AudioDestination is not destroyed.
- Modules/webaudio/DefaultAudioDestinationNode.cpp:
(WebCore::DefaultAudioDestinationNode::uninitialize):
- 5:46 AM Changeset in webkit [240142] by
-
- 2 edits in trunk/Source/WebCore
[WebAudio] Call AudioContext::uninitialize() immediately when the AudioContext is stopped
https://bugs.webkit.org/show_bug.cgi?id=192586
Patch by Yacine Bandou <yacine.bandou@softathome.com> on 2019-01-18
Reviewed by Philippe Normand.
When WebProcess is killed, AudioContext::uninitialize() is not called immediately in the stop so
the AudioDestinationNode is not destroyed.
In my case, I have a resource device manager, the output audio device is reserved when AudioDestinationNode
is instantiated and it is released when AudioDestinationNode is destroyed, thus when the webprocess is killed,
the resources leak.
AudioContext::uninitialize() is not called immediately since r94608.
This modification can now be reverted without regression in WebAudio tests.
Test: webaudio/mediaelementaudiosourcenode-gc.html
- Modules/webaudio/AudioContext.cpp:
(WebCore::AudioContext::stop):
- 5:07 AM Changeset in webkit [240141] by
-
- 15 edits1 copy24 adds in trunk
[WPE] Add Qt extension
https://bugs.webkit.org/show_bug.cgi?id=191464
Reviewed by Carlos Garcia Campos.
.:
- Source/cmake/OptionsWPE.cmake: Add ENABLE_WPE_QT_API CMake option. Disabled by default.
Source/WebKit:
This new extension is a QML plugin embedding a WPE ViewBackend
implementation. It provides a public API very similar to Qt's
WebView module. It comes with a simple mini-browser implemented in
QML.
QtWPE is known to work with the Wayland-EGL (in GNOME and Weston
compositors) and EGLFS QPAs.
- PlatformWPE.cmake:
- UIProcess/API/wpe/qt/WPEQmlExtensionPlugin.cpp: Added.
(WPEQmlExtensionPlugin::registerTypes):
- UIProcess/API/wpe/qt/WPEQmlExtensionPlugin.h: Added.
- UIProcess/API/wpe/qt/WPEQtView.cpp: Added.
(WPEQtView::WPEQtView):
(WPEQtView::~WPEQtView):
(WPEQtView::geometryChanged):
(WPEQtView::configureWindow):
(WPEQtView::createOffscreen):
(WPEQtView::backendCreated):
(WPEQtView::notifyUrlChangedCallback):
(WPEQtView::notifyTitleChangedCallback):
(WPEQtView::notifyLoadProgressCallback):
(WPEQtView::notifyLoadChangedCallback):
(WPEQtView::notifyLoadFailedCallback):
(WPEQtView::updatePaintNode):
(WPEQtView::url const):
(WPEQtView::setUrl):
(WPEQtView::loadProgress const):
(WPEQtView::title const):
(WPEQtView::canGoBack const):
(WPEQtView::isLoading const):
(WPEQtView::canGoForward const):
(WPEQtView::goBack):
(WPEQtView::goForward):
(WPEQtView::reload):
(WPEQtView::stop):
(WPEQtView::loadHtml):
(WPEQtView::jsAsyncReadyCallback):
(WPEQtView::handleJsResult):
(WPEQtView::runJavaScript):
(WPEQtView::mousePressEvent):
(WPEQtView::mouseReleaseEvent):
(WPEQtView::hoverEnterEvent):
(WPEQtView::hoverLeaveEvent):
(WPEQtView::hoverMoveEvent):
(WPEQtView::wheelEvent):
(WPEQtView::keyPressEvent):
(WPEQtView::keyReleaseEvent):
(WPEQtView::touchEvent):
- UIProcess/API/wpe/qt/WPEQtView.h: Added.
- UIProcess/API/wpe/qt/WPEQtViewBackend.cpp: Added.
(configureCallback):
(WPEQtViewBackend::WPEQtViewBackend):
(WPEQtViewBackend::~WPEQtViewBackend):
(WPEQtViewBackend::configureGlibEglDisplay):
(WPEQtViewBackend::configureEglDisplay):
(WPEQtViewBackend::initialize):
(WPEQtViewBackend::backend const):
(WPEQtViewBackend::resize):
(WPEQtViewBackend::getTexture):
(WPEQtViewBackend::initSurface):
(WPEQtViewBackend::displayImage):
(WPEQtViewBackend::modifiers const):
(WPEQtViewBackend::dispatchHoverEnterEvent):
(WPEQtViewBackend::dispatchHoverLeaveEvent):
(WPEQtViewBackend::dispatchHoverMoveEvent):
(WPEQtViewBackend::dispatchMousePressEvent):
(WPEQtViewBackend::dispatchMouseReleaseEvent):
(WPEQtViewBackend::dispatchWheelEvent):
(WPEQtViewBackend::dispatchKeyEvent):
(WPEQtViewBackend::dispatchTouchEvent):
- UIProcess/API/wpe/qt/WPEQtViewBackend.h: Added.
- UIProcess/API/wpe/qt/WPEQtViewLoadRequest.cpp: Added.
(WPEQtViewLoadRequest::WPEQtViewLoadRequest):
(WPEQtViewLoadRequest::~WPEQtViewLoadRequest):
(WPEQtViewLoadRequest::url const):
(WPEQtViewLoadRequest::status const):
(WPEQtViewLoadRequest::errorString const):
- UIProcess/API/wpe/qt/WPEQtViewLoadRequest.h: Added.
- UIProcess/API/wpe/qt/WPEQtViewLoadRequestPrivate.h: Added.
(WPEQtViewLoadRequestPrivate::WPEQtViewLoadRequestPrivate):
(WPEQtViewLoadRequestPrivate::~WPEQtViewLoadRequestPrivate):
- UIProcess/API/wpe/qt/qmldir: Added.
Tools:
MiniBrowser and API tests for the WPE Qt API. To run the
MiniBrowser a new script is introduced. Example invocation:
$ run-qt-wpe-minibrowser -platform wayland https://webkit.org
Also note-worthy is the introduction of the python2-subprocess32
dependency to run the GLib API tests.
- MiniBrowser/wpe/CMakeLists.txt:
- MiniBrowser/wpe/qt/CMakeLists.txt: Added.
- MiniBrowser/wpe/qt/main.cpp: Added.
(main):
- MiniBrowser/wpe/qt/main.qml: Added.
- MiniBrowser/wpe/qt/qml.qrc: Added.
- Scripts/run-gtk-tests:
(GtkTestRunner.is_google_test):
(GtkTestRunner):
(GtkTestRunner.is_qt_test):
- Scripts/run-qt-wpe-minibrowser: Added.
- Scripts/run-wpe-tests:
(WPETestRunner):
(WPETestRunner.is_google_test):
(WPETestRunner.is_qt_test):
- TestWebKitAPI/Tests/WPEQt/TestLoad.cpp: Added.
(TestLoad::main):
- TestWebKitAPI/Tests/WPEQt/TestLoadHtml.cpp: Added.
(TestLoadHtml::main):
- TestWebKitAPI/Tests/WPEQt/TestLoadRequest.cpp: Added.
(TestLoadRequest::main):
- TestWebKitAPI/Tests/WPEQt/TestRunJavaScript.cpp: Added.
(TestRunJavaScript::main):
- TestWebKitAPI/Tests/WPEQt/WPEQtTest.cpp: Added.
(waitForSignal):
- TestWebKitAPI/Tests/WPEQt/WPEQtTest.h: Added.
(LoadSpy::LoadSpy):
(LoadSpy::~LoadSpy):
(LoadSpy::onLoadingChanged):
(LoadStartedCatcher::LoadStartedCatcher):
(LoadStartedCatcher::~LoadStartedCatcher):
(LoadStartedCatcher::onLoadingChanged):
(waitForLoadSucceeded):
(waitForLoadFailed):
- TestWebKitAPI/glib/CMakeLists.txt:
- flatpak/flatpakutils.py:
(WebkitFlatpak.load_from_args):
(WebkitFlatpak.init):
(WebkitFlatpak.clean_args):
(WebkitFlatpak.run_in_sandbox):
- flatpak/org.webkit.CommonModules.yaml:
- flatpak/org.webkit.WPE.yaml:
- flatpak/org.webkit.WPEModules.yaml:
- flatpak/org.webkit.WPEQT.yaml: Copied from Tools/flatpak/org.webkit.WebKit.yaml.
- flatpak/org.webkit.WebKit.yaml:
- glib/api_test_runner.py:
(TestRunner._run_test_qt):
(TestRunner.is_qt_test):
(TestRunner._run_test):
- 12:07 AM Changeset in webkit [240140] by
-
- 5 edits in trunk/Source/WebCore
ScrollingCoordinator::scrollableAreaScrollLayerDidChange() can be removed
https://bugs.webkit.org/show_bug.cgi?id=193559
Reviewed by Antti Koivisto.
ScrollingCoordinator::scrollableAreaScrollLayerDidChange() existed for CoordinatedGraphics,
but the code that used it was removed in webkit.org/r229318 so we can remove it and
code that calls it.
- page/scrolling/ScrollingCoordinator.h:
(WebCore::ScrollingCoordinator::willDestroyScrollableArea):
(WebCore::ScrollingCoordinator::scrollableAreaScrollLayerDidChange): Deleted.
- rendering/RenderLayerBacking.cpp:
(WebCore::RenderLayerBacking::updateGeometry):
- rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::willRemoveScrollingLayerWithBacking):
(WebCore::RenderLayerCompositor::didAddScrollingLayer):
(WebCore::RenderLayerCompositor::scrollingLayerDidChange): Deleted.
- rendering/RenderLayerCompositor.h:
Jan 17, 2019:
- 8:15 PM Changeset in webkit [240139] by
-
- 23 edits4 adds in trunk
[iOS] Content offset jumps erratically when autoscrolling near scroll view content inset areas
https://bugs.webkit.org/show_bug.cgi?id=193494
<rdar://problem/46859627>
Reviewed by Simon Fraser and Tim Horton.
Source/WebCore:
When computing the content offset to scroll to when revealing a given rect in content coordinates, we currently
just use the unobscured content rect. As a result, when scrolling to reveal a rect, we'll clamp the final scroll
position such that only content is visible. For example, when asked to reveal the rect(0, 0, 1, 1), we adjust
the scroll position to be the origin.
However, consider the case where a client (e.g. Mail on iOS) has added a content inset to the web view's scroll
view. If we're asked to reveal a rect that is outside the content area but within a content inset, we will still
end up clamping the scroll position to the unobscured rect. This manifests in a bug where selecting text and
autoscrolling in iOS Mail compose while the scroll view is scrolled all the way to the top to reveal the To/Cc/
Subject fields causes the content offset to jump to the origin, rather than staying at (0, -topContentInset).
To fix this, we teach
RenderLayer::scrollRectToVisibleabout content insets that are visible. Rather than use
the content rects as-is, expand to encompass visible content insets as well. This ensures that revealing a
position which is already visible won't cause us to scroll away the content inset area and only show the
unobscured rect.
Tests: editing/selection/ios/autoscroll-with-top-content-inset.html
fast/scrolling/ios/scroll-into-view-with-top-content-inset.html
- page/FrameView.cpp:
(WebCore::FrameView::unobscuredContentRectExpandedByContentInsets const):
Introduce a helper method that expands the unobscured content rect to include surrounding content insets.
- page/FrameView.h:
- page/Page.h:
(WebCore::Page::contentInsets const):
(WebCore::Page::setContentInsets):
- rendering/RenderLayer.cpp:
(WebCore::RenderLayer::scrollRectToVisible):
(WebCore::RenderLayer::getRectToExpose const):
Source/WebKit:
Adds
contentInsetstoVisibleContentRectUpdateInfo. This keeps track of the visible content insets
surrounding the unobscured content rect. See WebCore ChangeLog for more details.
- Shared/VisibleContentRectUpdateInfo.cpp:
(WebKit::VisibleContentRectUpdateInfo::encode const):
(WebKit::VisibleContentRectUpdateInfo::decode):
(WebKit::operator<<):
- Shared/VisibleContentRectUpdateInfo.h:
(WebKit::VisibleContentRectUpdateInfo::VisibleContentRectUpdateInfo):
(WebKit::VisibleContentRectUpdateInfo::contentInsets const):
(WebKit::operator==):
- UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _computedObscuredInset]):
(-[WKWebView _computedContentInset]):
(-[WKWebView _computedUnobscuredSafeAreaInset]):
We don't care about source compatibility with iOS 10 and below anymore, so we should change these >= iOS 11
target checks to simplyPLATFORM(IOS).
(-[WKWebView _updateVisibleContentRects]):
Compute the visible content insets on all sides of the unobscured content rect. These insets are scaled to
content coordinates.
- UIProcess/ios/WKContentView.h:
- UIProcess/ios/WKContentView.mm:
(floatBoxExtent):
Add a helper to convert
UIEdgeInsetstoWebCore::FloatBoxExtent, and use it in a few places below.
(-[WKContentView didUpdateVisibleRect:unobscuredRect:contentInsets:unobscuredRectInScrollViewCoordinates:obscuredInsets:unobscuredSafeAreaInsets:inputViewBounds:scale:minimumScale:inStableState:isChangingObscuredInsetsInteractively:enclosedInScrollableAncestorView:]):
(-[WKContentView didUpdateVisibleRect:unobscuredRect:unobscuredRectInScrollViewCoordinates:obscuredInsets:unobscuredSafeAreaInsets:inputViewBounds:scale:minimumScale:inStableState:isChangingObscuredInsetsInteractively:enclosedInScrollableAncestorView:]): Deleted.
- WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::updateVisibleContentRects):
Update the Page's content insets.
Tools:
Add a new test option to add a top content inset to the test runner's WKWebView's scroll view, and automatically
scroll to reveal this top content inset area when beginning the test (i.e., scroll to (0, -topContentInset)).
- DumpRenderTree/ios/UIScriptControllerIOS.mm:
(WTR::UIScriptController::contentOffsetX const):
(WTR::UIScriptController::contentOffsetY const):
- TestRunnerShared/UIScriptContext/Bindings/UIScriptController.idl:
- TestRunnerShared/UIScriptContext/UIScriptController.cpp:
(WTR::UIScriptController::contentOffsetX const):
(WTR::UIScriptController::contentOffsetY const):
- TestRunnerShared/UIScriptContext/UIScriptController.h:
Also add new UIScriptController methods to ask for the content offset of the platform scroll view.
- WebKitTestRunner/TestController.cpp:
(WTR::updateTestOptionsFromTestHeader):
- WebKitTestRunner/TestOptions.h:
(WTR::TestOptions::hasSameInitializationOptions const):
- WebKitTestRunner/ios/TestControllerIOS.mm:
(WTR::TestController::platformResetStateToConsistentValues):
- WebKitTestRunner/ios/UIScriptControllerIOS.mm:
(WTR::UIScriptController::contentOffsetX const):
(WTR::UIScriptController::contentOffsetY const):
LayoutTests:
- editing/selection/ios/autoscroll-with-top-content-inset-expected.txt: Added.
- editing/selection/ios/autoscroll-with-top-content-inset.html: Added.
Add a new test to verify that moving the selection by autoscrolling near the top content inset area does not
cause the scroll view's content offset to jump.
- fast/scrolling/ios/scroll-into-view-with-top-content-inset-expected.txt: Added.
- fast/scrolling/ios/scroll-into-view-with-top-content-inset.html: Added.
Add a new test to verify that programmatically scrolling an element that's already visible into view does not
scroll away the scroll view's content inset.
- resources/ui-helper.js:
(window.UIHelper.contentOffset):
(window.UIHelper):
- 6:11 PM Changeset in webkit [240138] by
-
- 24 edits in trunk/Source/JavaScriptCore
Audit bytecode fields and ensure that LLInt instructions for accessing them are appropriate.
https://bugs.webkit.org/show_bug.cgi?id=193557
<rdar://problem/47369125>
Reviewed by Yusuke Suzuki.
- Rename some bytecode fields so that it's easier to discern whether the LLInt
is accessing them the right way:
- distinguish between targetVirtualRegister and targetLabel.
- name all StructureID fields as structureID (oldStructureID, newStructureID) instead of structure (oldStructure, newStructure).
- Use bitwise_cast in struct Fits when sizeof(T) == size. This prevents potential undefined behavior issues arising from doing assignments with reinterpret_cast'ed pointers.
- Make Special::Pointer an unsigned type (previously int). Make ResolveType an unsigned type (previously int).
- In LowLevelInterpreter*.asm:
- rename the op macro argument to opcodeName or opcodeStruct respectively. This makes it clearer which argument type the macro is working with.
- rename the name macro argument to opcodeName.
- fix operator types to match the field type being accessed. The following may have resulted in bugs before:
- The following should be read with getu() instead of get() because they are unsigned ints:
OpSwitchImm::m_tableIndex
OpSwitchChar::m_tableIndex
OpGetFromArguments::m_index
OpPutToArguments::m_index
OpGetRestLength::m_numParametersToSkip
OpJneqPtr::m_specialPointer should also be read with getu() though this
wasn't a bug because it was previously an int by default, and is only
changed to an unsigned int in this patch.
2.The following should be read with loadi (not loadp) because they are of
unsigned type (not a pointer):
OpResolveScope::Metadata::m_resolveType
CodeBlock::m_numParameters (see prepareForTailCall)
- OpPutToScope::Metadata::m_operand should be read with loadp (not loadis) because it is a uintptr_t.
- The following should be read with loadi (not loadis) because they are unsigned ints:
OpNegate::Metadata::m_arithProfile + ArithProfile::m_bits
OpPutById::Metadata::m_oldStructureID
OpPutToScope::Metadata::m_getPutInfo + GetPutInfo::m_operand
These may not have manifested in bugs because the operations that follow
the load are 32-bit instructions which ignore the high word.
- Give class GetPutInfo a default constructor so that we can use bitwise_cast on it. Also befriend LLIntOffsetsExtractor so that we can take the offset of m_operand in it.
- bytecode/ArithProfile.h:
- bytecode/BytecodeList.rb:
- bytecode/BytecodeUseDef.h:
(JSC::computeUsesForBytecodeOffset):
(JSC::computeDefsForBytecodeOffset):
- bytecode/CodeBlock.cpp:
(JSC::CodeBlock::propagateTransitions):
(JSC::CodeBlock::finalizeLLIntInlineCaches):
- bytecode/Fits.h:
- bytecode/GetByIdMetadata.h:
- bytecode/GetByIdStatus.cpp:
(JSC::GetByIdStatus::computeFromLLInt):
- bytecode/LLIntPrototypeLoadAdaptiveStructureWatchpoint.cpp:
(JSC::LLIntPrototypeLoadAdaptiveStructureWatchpoint::clearLLIntGetByIdCache):
- bytecode/PreciseJumpTargetsInlines.h:
(JSC::jumpTargetForInstruction):
(JSC::updateStoredJumpTargetsForInstruction):
- bytecode/PutByIdStatus.cpp:
(JSC::PutByIdStatus::computeFromLLInt):
- bytecode/SpecialPointer.h:
- bytecompiler/BytecodeGenerator.cpp:
(JSC::Label::setLocation):
- dfg/DFGByteCodeParser.cpp:
(JSC::DFG::ByteCodeParser::parseBlock):
- jit/JITArithmetic.cpp:
(JSC::JIT::emit_compareAndJump):
(JSC::JIT::emit_compareUnsignedAndJump):
(JSC::JIT::emit_compareAndJumpSlow):
- jit/JITArithmetic32_64.cpp:
(JSC::JIT::emit_compareAndJump):
(JSC::JIT::emit_compareUnsignedAndJump):
(JSC::JIT::emit_compareAndJumpSlow):
(JSC::JIT::emitBinaryDoubleOp):
- jit/JITOpcodes.cpp:
(JSC::JIT::emit_op_jmp):
(JSC::JIT::emit_op_jfalse):
(JSC::JIT::emit_op_jeq_null):
(JSC::JIT::emit_op_jneq_null):
(JSC::JIT::emit_op_jneq_ptr):
(JSC::JIT::emit_op_jeq):
(JSC::JIT::emit_op_jtrue):
(JSC::JIT::emit_op_jneq):
(JSC::JIT::compileOpStrictEqJump):
(JSC::JIT::emitSlow_op_jstricteq):
(JSC::JIT::emitSlow_op_jnstricteq):
(JSC::JIT::emit_op_check_tdz):
(JSC::JIT::emitSlow_op_jeq):
(JSC::JIT::emitSlow_op_jneq):
(JSC::JIT::emit_op_profile_type):
- jit/JITOpcodes32_64.cpp:
(JSC::JIT::emit_op_jmp):
(JSC::JIT::emit_op_jfalse):
(JSC::JIT::emit_op_jtrue):
(JSC::JIT::emit_op_jeq_null):
(JSC::JIT::emit_op_jneq_null):
(JSC::JIT::emit_op_jneq_ptr):
(JSC::JIT::emit_op_jeq):
(JSC::JIT::emitSlow_op_jeq):
(JSC::JIT::emit_op_jneq):
(JSC::JIT::emitSlow_op_jneq):
(JSC::JIT::compileOpStrictEqJump):
(JSC::JIT::emitSlow_op_jstricteq):
(JSC::JIT::emitSlow_op_jnstricteq):
(JSC::JIT::emit_op_check_tdz):
(JSC::JIT::emit_op_profile_type):
- llint/LLIntSlowPaths.cpp:
(JSC::LLInt::LLINT_SLOW_PATH_DECL):
(JSC::LLInt::setupGetByIdPrototypeCache):
- llint/LowLevelInterpreter.asm:
- llint/LowLevelInterpreter32_64.asm:
- llint/LowLevelInterpreter64.asm:
- runtime/CommonSlowPaths.cpp:
- runtime/GetPutInfo.h:
- 4:53 PM Changeset in webkit [240137] by
-
- 7 edits in branches/safari-607-branch/Source
Versioning.
- 4:43 PM Changeset in webkit [240136] by
-
- 39 edits22 deletes in trunk
Unreviewed, rolling out r240124.
This commit broke an internal build.
Reverted changeset:
"SDK_VARIANT build destinations should be separate from non-
SDK_VARIANT builds"
https://bugs.webkit.org/show_bug.cgi?id=189553
https://trac.webkit.org/changeset/240124
- 4:19 PM Changeset in webkit [240135] by
-
- 2 edits in trunk/Source/WebCore
Web Inspector: fix Xcode project file list after r239976
https://bugs.webkit.org/show_bug.cgi?id=193474
Reviewed by Timothy Hatcher.
- WebCore.xcodeproj/project.pbxproj:
- 4:17 PM Changeset in webkit [240134] by
-
- 7 edits in trunk/Source
Versioning.
- 4:09 PM Changeset in webkit [240133] by
-
- 1 copy in tags/Safari-607.1.24
Tag Safari-607.1.24.
- 2:54 PM Changeset in webkit [240132] by
-
- 2 edits in trunk/Source/WebCore
Unreviewed WinCairo fix -- hundreds of tests crash after r240031.
- platform/network/curl/ResourceHandleCurl.cpp:
(WebCore::ResourceHandle::createCurlRequest):
(WebCore::ResourceHandle::didReceiveAuthenticationChallenge):
(WebCore::ResourceHandle::receivedCredential):
(WebCore::ResourceHandle::getCredential):
- 1:45 PM Changeset in webkit [240131] by
-
- 2 edits in trunk/Source/WebKitLegacy/win
[Win][HighDPI] Repaint glitches when scrolling.
https://bugs.webkit.org/show_bug.cgi?id=173152
<rdar://problem/45269953>
Reviewed by Brent Fulgham.
Non-integral device scale factors are causing repaint glitches, because the computation of the scroll
delta in pixel coordinates from the scroll delta in logical coordinates will not always be correct.
Instead of blitting the scroll rectangle, repaint the entire region affected by scrolling.
- WebView.cpp:
(WebView::scrollBackingStore):
- 1:38 PM Changeset in webkit [240130] by
-
- 2 edits in trunk/Tools
Fix string concatenation API test after r239920
https://bugs.webkit.org/show_bug.cgi?id=193551
<rdar://problem/47359304>
Reviewed by Youenn Fablet.
- TestWebKitAPI/Tests/WTF/StringConcatenate.cpp:
(TestWebKitAPI::TEST):
Cocoa platforms' interpretation of UChar/unsigned short equality is not affected by ICU's recent change of UChar's type
in WebKit's build configuration like it apparently is on Linux.
- 1:36 PM Changeset in webkit [240129] by
-
- 11 edits in trunk
Add infrastructure to enable/disable ITP Debug Mode through Preferences
https://bugs.webkit.org/show_bug.cgi?id=193510
<rdar://problem/47330969>
Reviewed by Dean Jackson.
Source/WebCore:
No new tests. These changes are purely for settings/preferences.
- page/RuntimeEnabledFeatures.h:
(WebCore::RuntimeEnabledFeatures::setItpDebugModeEnabled):
(WebCore::RuntimeEnabledFeatures::itpDebugModeEnabled const):
(WebCore::RuntimeEnabledFeatures::setResourceLoadStatisticsDebugMode): Deleted.
(WebCore::RuntimeEnabledFeatures::resourceLoadStatisticsDebugMode const): Deleted.
Renamed.
- page/Settings.yaml:
Removed since this particular setting should not be persisted for privacy
reasons.
Source/WebKit:
- NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:
(WebKit::ResourceLoadStatisticsMemoryStore::ensurePrevalentResourcesForDebugMode):
(WebKit::ResourceLoadStatisticsMemoryStore::setResourceLoadStatisticsDebugMode):
(WebKit::ResourceLoadStatisticsMemoryStore::setPrevalentResourceForDebugMode):
These functions are now suited for runtime switching of ITP Debug Mode.
- Shared/WebPreferences.yaml:
Changed name from ResourceLoadStatisticsDebugMode to ItpDebugModeEnabled.
- UIProcess/API/Cocoa/WKPreferences.mm:
(-[WKPreferences _setItpDebugModeEnabled:]):
(-[WKPreferences _itpDebugModeEnabled]):
New private preferences functions.
- UIProcess/API/Cocoa/WKPreferencesPrivate.h:
- UIProcess/Cocoa/ResourceLoadStatisticsMemoryStoreCocoa.mm:
(WebKit::ResourceLoadStatisticsMemoryStore::registerUserDefaultsIfNeeded):
Removed support for old experimental user default. Added a new one that's
not connected to any UI, i.e. merely for command line debugging where
ITP Debug Mode needs to be on all the time.
LayoutTests:
- http/tests/resourceLoadStatistics/set-custom-prevalent-resource-in-debug-mode.html:
Changed the order of function calls since testRunner.setStatisticsPrevalentResourceForDebugMode()
now has to be called before testRunner.setStatisticsDebugMode().
- 1:28 PM Changeset in webkit [240128] by
-
- 2 edits in trunk/Tools
[ews-build] validate-patch should handle the case when bug_id is None
https://bugs.webkit.org/show_bug.cgi?id=193547
Reviewed by Lucas Forschler.
- BuildSlaveSupport/ews-build/steps.py:
(ValidatePatch._is_bug_closed): Gracefully handle the case when bug_id is None.
(ValidatePatch.start): If bug_id is set as None in build properties, fetch it from the patch.
- 1:24 PM Changeset in webkit [240127] by
-
- 2 edits in trunk/Tools
[ews-build] Add patch author information in build
https://bugs.webkit.org/show_bug.cgi?id=193465
Reviewed by Lucas Forschler.
- BuildSlaveSupport/ews-build/steps.py:
(ValidatePatch._is_patch_obsolete):
- 1:14 PM Changeset in webkit [240126] by
-
- 2 edits in trunk/Tools
[ews-build] Pass bug_id to the triggered builds
https://bugs.webkit.org/show_bug.cgi?id=193548
Reviewed by Lucas Forschler.
- BuildSlaveSupport/ews-build/factories.py:
(BuildFactory.propertiesToPassToTriggers):
- 1:12 PM Changeset in webkit [240125] by
-
- 4 edits in trunk/Source
Adopt NSTextCheckingSuppressInitialCapitalizationKey
https://bugs.webkit.org/show_bug.cgi?id=193545
Reviewed by Dean Jackson.
Source/WebCore/PAL:
- pal/spi/mac/NSSpellCheckerSPI.h:
Source/WebKit:
- UIProcess/mac/TextCheckerMac.mm:
(WebKit::TextChecker::checkTextOfParagraph):
(WebKit::TextChecker::getGuessesForWord):
- 12:46 PM Changeset in webkit [240124] by
-
- 39 edits22 adds in trunk
SDK_VARIANT build destinations should be separate from non-SDK_VARIANT builds
https://bugs.webkit.org/show_bug.cgi?id=189553
Reviewed by Tim Horton.
.:
- Makefile.shared:
PerformanceTests:
- MediaTime/Configurations/Base.xcconfig:
- MediaTime/Configurations/SDKVariant.xcconfig: Added.
Source/bmalloc:
- Configurations/Base.xcconfig:
- Configurations/SDKVariant.xcconfig: Added.
Source/JavaScriptCore:
- Configurations/Base.xcconfig:
- Configurations/SDKVariant.xcconfig: Added.
Source/ThirdParty/ANGLE:
- Configurations/Base.xcconfig:
- Configurations/SDKVariant.xcconfig: Added.
Source/ThirdParty/libwebrtc:
- Configurations/Base.xcconfig:
- Configurations/SDKVariant.xcconfig: Added.
Source/WebCore:
- Configurations/Base.xcconfig:
- Configurations/SDKVariant.xcconfig: Added.
- Configurations/WebCore.xcconfig:
Source/WebCore/PAL:
- Configurations/Base.xcconfig:
- Configurations/PAL.xcconfig:
- Configurations/SDKVariant.xcconfig: Added.
Source/WebInspectorUI:
- Configurations/Base.xcconfig:
- Configurations/SDKVariant.xcconfig: Added.
Source/WebKit:
- Configurations/Base.xcconfig:
- Configurations/BaseTarget.xcconfig:
- Configurations/SDKVariant.xcconfig: Added.
Source/WebKitLegacy/mac:
- Configurations/Base.xcconfig:
- Configurations/SDKVariant.xcconfig: Added.
- Configurations/WebKitLegacy.xcconfig:
Source/WTF:
- Configurations/Base.xcconfig:
- Configurations/SDKVariant.xcconfig: Added.
Tools:
Add a check-webkit-style check that ensures all SDKVariant.xcconfig files match.
- ContentExtensionTester/Configurations/Base.xcconfig:
- ContentExtensionTester/Configurations/SDKVariant.xcconfig: Added.
- DumpRenderTree/mac/Configurations/Base.xcconfig:
- DumpRenderTree/mac/Configurations/SDKVariant.xcconfig: Added.
- ImageDiff/cg/Configurations/Base.xcconfig:
- ImageDiff/cg/Configurations/SDKVariant.xcconfig: Added.
- MiniBrowser/Configurations/Base.xcconfig:
- MiniBrowser/Configurations/SDKVariant.xcconfig: Added.
- MobileMiniBrowser/Configurations/Base.xcconfig:
- MobileMiniBrowser/Configurations/SDKVariant.xcconfig: Added.
- Scripts/webkitpy/style/checker.py:
(_all_categories):
(FileType):
(CheckerDispatcher._file_type):
(CheckerDispatcher._create_checker):
- Scripts/webkitpy/style/checkers/sdkvariant.py: Added.
(SDKVariantChecker):
(SDKVariantChecker.init):
(SDKVariantChecker.check):
- TestWebKitAPI/Configurations/Base.xcconfig:
- TestWebKitAPI/Configurations/SDKVariant.xcconfig: Added.
- WebEditingTester/Configurations/Base.xcconfig:
- WebEditingTester/Configurations/SDKVariant.xcconfig: Added.
- WebKitLauncher/Configurations/SDKVariant.xcconfig: Added.
- WebKitTestRunner/Configurations/Base.xcconfig:
- WebKitTestRunner/Configurations/SDKVariant.xcconfig: Added.
- lldb/lldbWebKitTester/Configurations/Base.xcconfig:
- lldb/lldbWebKitTester/Configurations/SDKVariant.xcconfig: Added.
- 12:24 PM Changeset in webkit [240123] by
-
- 7 edits in trunk/Source/WebCore
MediaPlayerPrivateAVFoundationObjC can return incorrect paused information
https://bugs.webkit.org/show_bug.cgi?id=193499
Reviewed by Eric Carlson.
MediaPlayerPrivateAVFoundation uses rate() as an indicator of whether the player
is paused or not. This is incorrect when playback is stalled waiting for more data.
For MPPAVFObjC, use the timeControlStatus as a more accurate indicator of whether
the player is playing.
Now that we have correct play state information, we can remove the handlePlaybackCommand()
path when playing remotely for a more direct approach of notifying the HTMLMediaElement
that the play state has changed.
Drive-by fix: Before throwing away the AVPlayer, clear its output context. This keeps
remote devices from keeping the AVPlayer alive.
Drive-by fix #2: The NullMediaPlayer should always return "true" for paused(), not "false",
since it can't possibly play anything.
- platform/graphics/MediaPlayer.cpp:
- platform/graphics/avfoundation/MediaPlayerPrivateAVFoundation.cpp:
(WebCore::MediaPlayerPrivateAVFoundation::paused const):
- platform/graphics/avfoundation/MediaPlayerPrivateAVFoundation.h:
(WebCore::MediaPlayerPrivateAVFoundation::platformPaused const):
- platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.h:
- platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
(WebCore::MediaPlayerPrivateAVFoundationObjC::cancelLoad):
(WebCore::MediaPlayerPrivateAVFoundationObjC::platformPaused const):
(WebCore::MediaPlayerPrivateAVFoundationObjC::timeControlStatusDidChange):
- 11:34 AM Changeset in webkit [240122] by
-
- 3 edits in trunk/Source/WebCore
[Mac] Add a new quirk to HTMLFormControlElement::isMouseFocusable
https://bugs.webkit.org/show_bug.cgi?id=193478
<rdar://problem/34368591>
Reviewed by Brent Fulgham.
By default in macOS, submit buttons (controls) are not focusable. WebKit follows this system convention
as suggested by the spec: https://html.spec.whatwg.org/multipage/interaction.html#focusable-area. This
is also the convention Firefox respects. However, Chrome doesn't. ceac.state.gov is by far the only
website that assumes submit buttons are focusable, and will prohibit users from completing immigration
forms, such as DS160 if buttons are not. To help immigrations, we decide to add a new quirk to
HTMLFormControlElement::isMouseFocusable such that submit buttons are mouse focusable.
This quirk is for ceac.state.gov specifically, and therefore no tests.
- html/HTMLFormControlElement.cpp:
(WebCore::HTMLFormControlElement::isMouseFocusable const):
(WebCore::HTMLFormControlElement::needsSiteSpecificQuirks const):
- html/HTMLFormControlElement.h:
- 11:32 AM Changeset in webkit [240121] by
-
- 3 edits in trunk/Source/WebCore
Fix WinCairo build after r240117
https://bugs.webkit.org/show_bug.cgi?id=193529
- PlatformWin.cmake:
- platform/network/curl/SocketStreamHandleImplCurl.cpp:
- 11:23 AM Changeset in webkit [240120] by
-
- 17 edits in trunk/Source/WebCore
Add release logging for incoming and outgoing webrtc audio tracks
https://bugs.webkit.org/show_bug.cgi?id=185545
Reviewed by Eric Carlson.
Add logging of audio tracks. When doing a WebRTC call,
one log line is added each second for each audio track.
Validated that logging is done through manual testing.
Refactored code to use LogHelper and apply it to video sources as well.
- Modules/mediastream/libwebrtc/LibWebRTCMediaEndpoint.cpp:
(WebCore::LibWebRTCMediaEndpoint::addTrack):
(WebCore::LibWebRTCMediaEndpoint::sourceFromNewReceiver):
(WebCore::sourceFromNewReceiver): Deleted.
- Modules/mediastream/libwebrtc/LibWebRTCMediaEndpoint.h:
- platform/mediastream/RealtimeIncomingAudioSource.cpp:
(WebCore::RealtimeIncomingAudioSource::RealtimeIncomingAudioSource):
(WebCore::RealtimeIncomingAudioSource::logChannel const):
(WebCore::RealtimeIncomingAudioSource::logger const):
- platform/mediastream/RealtimeIncomingAudioSource.h:
(WebCore::RealtimeIncomingAudioSource::setLogger):
- platform/mediastream/RealtimeIncomingVideoSource.cpp:
(WebCore::RealtimeIncomingVideoSource::RealtimeIncomingVideoSource):
(WebCore::RealtimeIncomingVideoSource::logChannel const):
(WebCore::RealtimeIncomingVideoSource::logger const):
- platform/mediastream/RealtimeIncomingVideoSource.h:
(WebCore::RealtimeIncomingVideoSource::setLogger):
- platform/mediastream/RealtimeOutgoingAudioSource.cpp:
(WebCore::RealtimeOutgoingAudioSource::RealtimeOutgoingAudioSource):
(WebCore::RealtimeOutgoingAudioSource::sendAudioFrames):
(WebCore::RealtimeOutgoingAudioSource::logChannel const):
(WebCore::RealtimeOutgoingAudioSource::logger const):
- platform/mediastream/RealtimeOutgoingAudioSource.h:
(WebCore::RealtimeOutgoingAudioSource::setLogger):
- platform/mediastream/RealtimeOutgoingVideoSource.cpp:
(WebCore::RealtimeOutgoingVideoSource::RealtimeOutgoingVideoSource):
(WebCore::RealtimeOutgoingVideoSource::sendBlackFramesIfNeeded):
(WebCore::RealtimeOutgoingVideoSource::sendOneBlackFrame):
(WebCore::RealtimeOutgoingVideoSource::logChannel const):
(WebCore::RealtimeOutgoingVideoSource::logger const):
- platform/mediastream/RealtimeOutgoingVideoSource.h:
(WebCore::RealtimeOutgoingVideoSource::setLogger):
- platform/mediastream/mac/RealtimeIncomingAudioSourceCocoa.cpp:
(WebCore::RealtimeIncomingAudioSourceCocoa::OnData):
- platform/mediastream/mac/RealtimeIncomingAudioSourceCocoa.h:
- platform/mediastream/mac/RealtimeIncomingVideoSourceCocoa.mm:
(WebCore::RealtimeIncomingVideoSourceCocoa::pixelBufferPool):
(WebCore::RealtimeIncomingVideoSourceCocoa::pixelBufferFromVideoFrame):
(WebCore::RealtimeIncomingVideoSourceCocoa::OnFrame):
- platform/mediastream/mac/RealtimeOutgoingAudioSourceCocoa.cpp:
- platform/mediastream/mac/RealtimeOutgoingVideoSourceCocoa.cpp:
(WebCore::RealtimeOutgoingVideoSourceCocoa::sampleBufferUpdated):
- platform/mediastream/mac/RealtimeOutgoingVideoSourceCocoa.mm:
(WebCore::RealtimeOutgoingVideoSourceCocoa::rotatePixelBuffer):
- 11:15 AM Changeset in webkit [240119] by
-
- 7 edits2 adds in trunk
(iPad) Link tapping is sluggish on many sites
https://bugs.webkit.org/show_bug.cgi?id=193522
<rdar://problem/47102987>
Reviewed by Wenson Hsieh.
Source/WebKit:
Some WKWebView clients might set the initial zoom scale of the page to
something other than 1, which disables the "fast tap" behaviour.
The fix is very simple -- just check against the initial scale rather
than 1.
The most likely regression from this would be pages designed for desktop,
but provide a viewport tag saying width=device-width and initial-scale.
They might stop allowing double-tap-to-zoom.
- UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _allowsDoubleTapGestures]): Check against initial page scale.
LayoutTests:
Add a test that checks a double tap will trigger a click
event on a page that is at initial scale.
Extra bonus: for some reason adding this test, or making this code
change, uncovered a couple of bugs in existing tests. The
viewport-zooms-from-element-to-initial-scale test was completely wrong
because it was expecting the incorrect result, which was triggered by
the zoom callback firing early at a forced scale value. The
viewport-no-width-value-allows-double-tap test was triggering a JS
error in its UI script. I modernised both of these to use UIHelper instead.
- fast/events/ios/fast-click-double-tap-sends-click-when-initial-scale-expected.txt: Added.
- fast/events/ios/fast-click-double-tap-sends-click-when-initial-scale.html: Added.
- fast/events/ios/viewport-no-width-value-allows-double-tap.html:
- fast/events/ios/viewport-zooms-from-element-to-initial-scale-expected.txt:
- fast/events/ios/viewport-zooms-from-element-to-initial-scale.html:
- resources/ui-helper.js: Add doubleTapAt and zoomByDoubleTapAt helpers. Remove the
unnecessary "Done" return value from many of the callbacks. Give zoomToScale a return
value.
(window.UIHelper.tapAt.return.new.Promise):
(window.UIHelper.tapAt):
(window.UIHelper.doubleTapAt.return.new.Promise):
(window.UIHelper.doubleTapAt):
(window.UIHelper.zoomByDoubleTappingAt):
(window.UIHelper.activateAt.return.new.Promise):
(window.UIHelper.activateAt):
(window.UIHelper.toggleCapsLock):
(window.UIHelper.ensurePresentationUpdate.return.new.Promise):
(window.UIHelper.ensurePresentationUpdate):
(window.UIHelper.activateAndWaitForInputSessionAt.return.new.Promise.):
(window.UIHelper.activateFormControl.return.new.Promise.):
(window.UIHelper.replaceTextAtRange):
(window.UIHelper.zoomToScale):
(window.UIHelper.stylusTapAt.return.new.Promise):
(window.UIHelper.stylusTapAt):
- 11:13 AM Changeset in webkit [240118] by
-
- 2 edits in trunk
[Win] gperf command not found
https://bugs.webkit.org/show_bug.cgi?id=193538
<rdar://problem/47250549>
Reviewed by Brent Fulgham.
CMake only searches for gperf if ENABLE_WEBCORE is on. ENABLE_WEBCORE needs to be set to ON.
- Source/cmake/OptionsAppleWin.cmake:
- 11:10 AM Changeset in webkit [240117] by
-
- 33 edits3 adds in trunk/Source
Stop using NetworkStorageSession::storageSession in WebCore
https://bugs.webkit.org/show_bug.cgi?id=193529
Reviewed by Tim Horton.
Source/WebCore:
Ownership of the map that NetworkStorageSession::storageSession searches needs to move to the WebKit/WebKitLegacy layer,
so when WebCore used to look in this map it needs to call a client function to ask the layer above it for the storage object.
I've called this client StorageSessionProvider::storageSession. Right now it just looks in NetworkStorageSession::storageSession,
but this added abstraction makes it possible to move the currently process-global map to be a member of the NetworkProcess object.
- Modules/websockets/WebSocketChannel.cpp:
(WebCore::WebSocketChannel::connect):
- WebCore.xcodeproj/project.pbxproj:
- loader/CookieJar.cpp:
(WebCore::CookieJar::create):
(WebCore::CookieJar::CookieJar):
(WebCore::CookieJar::cookies const):
(WebCore::CookieJar::setCookies):
(WebCore::CookieJar::cookiesEnabled const):
(WebCore::CookieJar::cookieRequestHeaderFieldValue const):
(WebCore::CookieJar::getRawCookies const):
(WebCore::CookieJar::deleteCookie):
- loader/CookieJar.h:
- loader/EmptyClients.cpp:
(WebCore::pageConfigurationWithEmptyClients):
- page/Page.h:
- page/SocketProvider.cpp:
(WebCore::SocketProvider::createSocketStreamHandle):
- page/SocketProvider.h:
- platform/network/NetworkingContext.h:
- platform/network/SocketStreamHandleImpl.cpp:
(WebCore::cookieDataForHandshake):
(WebCore::SocketStreamHandleImpl::platformSendHandshake):
- platform/network/StorageSessionProvider.h: Added.
(WebCore::StorageSessionProvider::~StorageSessionProvider):
- platform/network/cf/SocketStreamHandleImpl.h:
(WebCore::SocketStreamHandleImpl::create):
- platform/network/cf/SocketStreamHandleImplCFNet.cpp:
(WebCore::SocketStreamHandleImpl::SocketStreamHandleImpl):
(WebCore::SocketStreamHandleImpl::getStoredCONNECTProxyCredentials):
- platform/network/curl/SocketStreamHandleImpl.h:
(WebCore::SocketStreamHandleImpl::create):
- platform/network/soup/SocketStreamHandleImpl.h:
- platform/network/soup/SocketStreamHandleImplSoup.cpp:
(WebCore::SocketStreamHandleImpl::create):
Source/WebKit:
- NetworkProcess/NetworkConnectionToWebProcess.cpp:
(WebKit::NetworkConnectionToWebProcess::createSocketStream):
- NetworkProcess/NetworkProcess.h:
- NetworkProcess/NetworkSocketStream.cpp:
(WebKit::NetworkSocketStream::create):
(WebKit::NetworkSocketStream::NetworkSocketStream):
- NetworkProcess/NetworkSocketStream.h:
- NetworkProcess/NetworkStorageSessionProvider.h: Added.
- WebKit.xcodeproj/project.pbxproj:
- WebProcess/Network/WebSocketProvider.cpp:
(WebKit::WebSocketProvider::createSocketStreamHandle):
- WebProcess/Network/WebSocketProvider.h:
- WebProcess/WebPage/WebCookieJar.cpp:
(WebKit::WebCookieJar::WebCookieJar):
- WebProcess/WebPage/WebCookieJar.h:
Source/WebKitLegacy:
- WebCoreSupport/PageStorageSessionProvider.h: Added.
- WebKitLegacy.xcodeproj/project.pbxproj:
Source/WebKitLegacy/mac:
- Misc/WebCache.mm:
(+[WebCache addImageToCache:forURL:forFrame:]):
- WebView/WebView.mm:
(-[WebView _commonInitializationWithFrameName:groupName:]):
(-[WebView initSimpleHTMLDocumentWithStyle:frame:preferences:groupName:]):
Source/WebKitLegacy/win:
- WebView.cpp:
(WebView::initWithFrame):
- 11:03 AM Changeset in webkit [240116] by
-
- 2 edits in trunk/LayoutTests
Layout Test js/dfg-int-overflow-in-loop.html is failing
https://bugs.webkit.org/show_bug.cgi?id=193540
Unreviewed test gardening.
- platform/win/TestExpectations:
- 10:34 AM Changeset in webkit [240115] by
-
- 9 edits5 deletes in trunk/LayoutTests
Get rid of Sputnik tests.
https://bugs.webkit.org/show_bug.cgi?id=193464
Rubber stamped by Saam Barati.
- sputnik/: Removed.
- platform/gtk/sputnik/: Removed.
- platform/ios/sputnik/: Removed.
- platform/mac/sputnik/: Removed.
- platform/wpe/sputnik/: Removed.
- TestExpectations:
- platform/gtk/TestExpectations:
- platform/ios-wk2/TestExpectations:
- platform/ios/TestExpectations:
- platform/mac/TestExpectations:
- platform/win/TestExpectations:
- platform/wincairo/TestExpectations:
- platform/wpe/TestExpectations:
Remove all sputnik/ lines.
- 9:50 AM Changeset in webkit [240114] by
-
- 19 edits1 add1 delete in trunk
StringObjectUse should not be a structure check for the original string object structure
https://bugs.webkit.org/show_bug.cgi?id=193483
<rdar://problem/47280522>
Reviewed by Yusuke Suzuki.
JSTests:
- stress/cant-eliminate-string-object-structure-check-when-string-object-is-proven.js: Added.
(foo):
(a.valueOf.0):
Source/JavaScriptCore:
Prior to this patch, the use kind for StringObjectUse implied that we
do a StructureCheck on the input operand for the *original* StringObject
structure. This is generally not how we use UseKinds, so it's no surprise
that this is buggy. A UseKind should map to a set of SpeculatedTypes, not an
actual set of structures. This patch changes the meaning of StringObjectUse
to mean an object where jsDynamicCast<StringObject*> would succeed.
This patch also fixes a bug that was caused by the old and weird usage of the
UseKind to mean StructureCheck. Consider a program like this:
`
S1 = Original StringObject structure
S2 = Original StringObject structure with the field "f" added
a: GetLocal()
b: CheckStructure(@a, {S2})
c: ToString(StringObject:@a)
`
According to AI, in the above program, we would exit at @c, since
StringObject:@a implies a structure check of {S1}, and the intersection
of {S1} and {S2} is {}. So, we'd convert the program to be:
`
a: GetLocal()
b: CheckStructure(@a, {S2})
c: Check(StringObject:@a)
d: Unreachable
`
However, AI would set the proof status of the StringObject:@a edge
to be proven, since the SpeculatedType for @a is SpecStringObject.
This was incorrect of AI to do because the SpeculatedType itself
didn't capture the full power of StringObjectUse. However, having
a UseKind mean CheckStructure is weird precisely because what AI was
doing is a natural fit to how we typically we think about UseKinds.
So the above program would then incorrectly be converted to this, and
we'd crash when reaching the Unreachable node:
`
a: GetLocal()
b: CheckStructure(@a, {S2})
d: Unreachable
`
This patch makes it so that StringObjectUse just means that the object that
filters through a StringObjectUse check must !!jsDynamicCast<StringObject*>.
This is now in line with all other UseKinds. It also lets us simplify a bunch
of other code that had weird checks for the StringObjectUse UseKind.
This patch also makes it so that anywhere where we used to rely on
StringObjectUse implying a structure check we actually emit an explicit
CheckStructure node.
- JavaScriptCore.xcodeproj/project.pbxproj:
- bytecode/ExitKind.cpp:
(JSC::exitKindToString):
- bytecode/ExitKind.h:
- dfg/DFGAbstractInterpreterInlines.h:
(JSC::DFG::AbstractInterpreter<AbstractStateType>::executeEffects):
- dfg/DFGCSEPhase.cpp:
- dfg/DFGClobberize.h:
(JSC::DFG::clobberize):
- dfg/DFGEdgeUsesStructure.h: Removed.
- dfg/DFGFixupPhase.cpp:
(JSC::DFG::FixupPhase::attemptToForceStringArrayModeByToStringConversion):
(JSC::DFG::FixupPhase::addCheckStructureForOriginalStringObjectUse):
(JSC::DFG::FixupPhase::fixupToPrimitive):
(JSC::DFG::FixupPhase::fixupToStringOrCallStringConstructor):
(JSC::DFG::FixupPhase::attemptToMakeFastStringAdd):
(JSC::DFG::FixupPhase::isStringObjectUse): Deleted.
- dfg/DFGGraph.cpp:
(JSC::DFG::Graph::canOptimizeStringObjectAccess):
- dfg/DFGMayExit.cpp:
- dfg/DFGSpeculativeJIT.cpp:
(JSC::DFG::SpeculativeJIT::compileToStringOrCallStringConstructorOrStringValueOf):
(JSC::DFG::SpeculativeJIT::speculateStringObject):
(JSC::DFG::SpeculativeJIT::speculateStringOrStringObject):
- dfg/DFGSpeculativeJIT.h:
(JSC::DFG::SpeculativeJIT::speculateStringObjectForStructure): Deleted.
- dfg/DFGUseKind.h:
(JSC::DFG::alreadyChecked):
(JSC::DFG::usesStructure): Deleted.
- ftl/FTLLowerDFGToB3.cpp:
(JSC::FTL::DFG::LowerDFGToB3::compileToStringOrCallStringConstructorOrStringValueOf):
(JSC::FTL::DFG::LowerDFGToB3::speculateStringObject):
(JSC::FTL::DFG::LowerDFGToB3::speculateStringOrStringObject):
(JSC::FTL::DFG::LowerDFGToB3::speculateStringObjectForCell):
(JSC::FTL::DFG::LowerDFGToB3::speculateStringObjectForStructureID): Deleted.
- runtime/JSType.cpp:
(WTF::printInternal):
- runtime/JSType.h:
- runtime/StringObject.h:
(JSC::StringObject::createStructure):
- runtime/StringPrototype.h:
- 9:32 AM Changeset in webkit [240113] by
-
- 2 edits in trunk/Source/JavaScriptCore
[JSC] Add generateHeapSnapshotForGCDebugging function to dump GCDebugging data
https://bugs.webkit.org/show_bug.cgi?id=193526
Reviewed by Michael Saboff.
This patch adds generateHeapSnapshotForGCDebugging to JSC shell to dump heap snapshot JSON string with GCDebugging option.
GCDebuggingSnapshot mode is slightly different from InspectorSnapshot in terms of both the output data and the behavior.
It always takes full snapshot, and it reports internal data too. This is useful to view the live heap objects after running
the code. Also, generateHeapSnapshotForGCDebugging returns String instead of parsing it to JSObject internally by calling
JSON.parse. If we convert the String to bunch of objects by using JSON.parse, it is difficult to call generateHeapSnapshotForGCDebugging
multiple times for debugging. Currently, it only generates a large string, which is easily distinguishable in the heap inspector tool.
- jsc.cpp:
(GlobalObject::finishCreation):
(functionGenerateHeapSnapshotForGCDebugging):
- 8:36 AM Changeset in webkit [240112] by
-
- 2 edits in trunk/Source/WebCore
[EME] Remove Amazon Prime Video from quirks list
https://bugs.webkit.org/show_bug.cgi?id=193514
rdar://problem/47295330
Reviewed by Jer Noble.
- page/Quirks.cpp:
(WebCore::Quirks::hasBrokenEncryptedMediaAPISupportQuirk const):
- 8:24 AM Changeset in webkit [240111] by
-
- 4 edits in trunk
[LFC][BFC] An element with transform is a containing block for positioned descendants.
https://bugs.webkit.org/show_bug.cgi?id=193534
Reviewed by Antti Koivisto.
Source/WebCore:
"For elements whose layout is governed by the CSS box model, any value other than none for the transform
property also causes the element to establish a containing block for all descendants."
https://www.w3.org/TR/css-transforms-1/
- layout/layouttree/LayoutBox.cpp:
(WebCore::Layout::Box::containingBlock const):
Tools:
- LayoutReloaded/misc/LFC-passing-tests.txt:
- 8:07 AM Changeset in webkit [240110] by
-
- 6 edits in trunk/Source/WebCore
ScrollingCoordinator: separate updating node geometry from node layers
https://bugs.webkit.org/show_bug.cgi?id=193527
Reviewed by Antti Koivisto.
Updating scrolling tree node layers happens in a different code path from updating geometry;
the former has to be post-flush, when GraphicsLayer has made tiled or structural layers.
Geometry only needs to be updated after layout, and soon we'll do this via compositing updates.
We can also clean up the ScrollingCoordinator API and have a single function that updates
node layers (handling different node types), and separate functions for pushing viewport-constrained
and scrolling node geometry.
No observable behavior change.
- page/scrolling/AsyncScrollingCoordinator.cpp:
(WebCore::AsyncScrollingCoordinator::setNodeLayers):
(WebCore::AsyncScrollingCoordinator::setScrollingNodeGeometry):
(WebCore::AsyncScrollingCoordinator::setViewportConstraintedNodeGeometry):
(WebCore::AsyncScrollingCoordinator::updateFrameScrollingNode): Deleted.
(WebCore::AsyncScrollingCoordinator::updateOverflowScrollingNode): Deleted.
(WebCore::AsyncScrollingCoordinator::updateNodeLayer): Deleted.
(WebCore::AsyncScrollingCoordinator::updateNodeViewportConstraints): Deleted.
- page/scrolling/AsyncScrollingCoordinator.h:
- page/scrolling/ScrollingCoordinator.h:
(WebCore::ScrollingCoordinator::setNodeLayers):
(WebCore::ScrollingCoordinator::setScrollingNodeGeometry):
(WebCore::ScrollingCoordinator::setViewportConstraintedNodeGeometry):
(WebCore::ScrollingCoordinator::updateNodeLayer): Deleted.
(WebCore::ScrollingCoordinator::updateNodeViewportConstraints): Deleted.
(WebCore::ScrollingCoordinator::updateFrameScrollingNode): Deleted.
(WebCore::ScrollingCoordinator::updateOverflowScrollingNode): Deleted.
- rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::updateScrollCoordinationForThisFrame):
(WebCore::RenderLayerCompositor::updateScrollCoordinatedLayer):
- rendering/RenderLayerCompositor.h:
- 8:06 AM Changeset in webkit [240109] by
-
- 5 edits in trunk
[LFC][BFC][Quirk] Take body padding and border into account when stretching height.
https://bugs.webkit.org/show_bug.cgi?id=193528
Reviewed by Antti Koivisto.
Source/WebCore:
- layout/blockformatting/BlockFormattingContextQuirks.cpp:
(WebCore::Layout::BlockFormattingContext::Quirks::stretchedInFlowHeight):
- layout/displaytree/DisplayBox.h:
(WebCore::Display::Box::verticalBorder const):
(WebCore::Display::Box::horizontalBorder const):
(WebCore::Display::Box::verticalPadding const):
(WebCore::Display::Box::horizontalPadding const):
- page/FrameViewLayoutContext.cpp:
(WebCore::layoutUsingFormattingContext):
Tools:
- LayoutReloaded/misc/LFC-passing-tests.txt:
- 8:00 AM Changeset in webkit [240108] by
-
- 4 edits in trunk
Source/WebCore:
[LFC][BFC] For height computation, the bottom edge of the last line box value should not include top border/padding
https://bugs.webkit.org/show_bug.cgi?id=193520
Reviewed by Antti Koivisto.
This is similar to the other "10.6.3" cases. The bottom edge of the last inline box is in the coordinate systyem
of the containing block's border box (and for content height computation it needs to be mapped to the containing block's content box instead).
- layout/blockformatting/BlockFormattingContextGeometry.cpp:
(WebCore::Layout::BlockFormattingContext::Geometry::inFlowNonReplacedHeightAndMargin):
Tools:
[LFC][BFC] For height computation, the bottom edge of the last line box value should include top border/padding
https://bugs.webkit.org/show_bug.cgi?id=193520
Reviewed by Antti Koivisto.
- LayoutReloaded/misc/LFC-passing-tests.txt:
- 5:52 AM Changeset in webkit [240107] by
-
- 2 edits in trunk/Tools
[ews-build] unapply-patch step should not update source
https://bugs.webkit.org/show_bug.cgi?id=193234
Reviewed by Lucas Forschler.
- BuildSlaveSupport/ews-build/steps.py:
(CleanWorkingDirectory): Use clean-webkit script to clean the working directory.
(UnApplyPatchIfRequired): Use CleanWorkingDirectory as base class.
- 12:10 AM Changeset in webkit [240106] by
-
- 4 edits1 add in trunk
[JSC] ToThis omission in DFGByteCodeParser is wrong
https://bugs.webkit.org/show_bug.cgi?id=193513
<rdar://problem/45842236>
Reviewed by Saam Barati.
JSTests:
- stress/to-this-omission-with-different-strict-modes.js: Added.
(thisA):
(thisAStrictWrapper):
Source/JavaScriptCore:
DFGByteCodeParser omitted ToThis node when we have
ToThis(ToThis(value)). This semantics is wrong if ToThis has different semantics
in the sloppy mode and the strict mode. If we convertToThisInSloppyMode(ToThisInStrictMode(boolean))toToThisInStrictMode(boolean),
we get boolean instead of BooleanObject.
This optimization is introduced more than 7 years ago, and from that, we have several optimizations that can remove such ToThis nodes
in BytecodeParser, AI, and Fixup. Furthermore, this optimization is simply wrong sincetoThis()function of JSCell can be defined
as they want. Before ensuring all the toThis function is safe, we should not foldToThis(ToThis(value))=>ToThis(value).
This patch just removes the problematic optimization. The performance numbers look neutral.
- dfg/DFGAbstractInterpreterInlines.h:
(JSC::DFG::AbstractInterpreter<AbstractStateType>::executeEffects):
- dfg/DFGByteCodeParser.cpp:
(JSC::DFG::ByteCodeParser::parseBlock):
