Timeline

May 28, 2019:

11:00 PM Changeset in webkit [245848] by commit-queue@webkit.org

• 3 edits
  2 adds in trunk

[MSE][GStreamer] update the readyState correctly in MediaPlayerPrivateGStreamerMSE
​https://bugs.webkit.org/show_bug.cgi?id=197834

Patch by Yacine Bandou <​yacine.bandou@softathome.com> on 2019-05-28
Reviewed by Xabier Rodriguez-Calvar.

Source/WebCore:

The buffering state and the m_downloadFinished boolean aren't supported in the MSE case.
When the readyState is already "HaveEnoughData", we don't want to revert it to "HaveFutureData",
or else the MediaPlayer would send a "canplay" event instead of a "canplaythrough".

Test: media/media-source/media-source-canplaythrough-event.html

• platform/graphics/gstreamer/mse/MediaPlayerPrivateGStreamerMSE.cpp:

(WebCore::MediaPlayerPrivateGStreamerMSE::updateStates):

LayoutTests:

Add a new test that checks if the MediaElement receives the "canplaythrough"
event when the media content is entirely injected to MSE sourceBuffer.

• media/media-source/media-source-canplaythrough-event-expected.txt: Added.
• media/media-source/media-source-canplaythrough-event.html: Added.

9:40 PM Changeset in webkit [245847] by Fujii Hironori

• 2 edits in trunk/Source/WebKit

[WinCairo] REGRESSION(r245186) Crash in NetworkCache::IOChannel::read in http/tests/IndexedDB some tests
​https://bugs.webkit.org/show_bug.cgi?id=197941

Reviewed by Don Olmstead.

http/tests/IndexedDB some tests were crashing in
NetworkCache::IOChannel::read in order to allocate a buffer with
std::numeric_limits<size_t>::max() as the size.

IOChannel::read should check the file size, and calculate the read
size.

• NetworkProcess/cache/NetworkCacheIOChannelCurl.cpp:

(WebKit::NetworkCache::IOChannel::read): Limit the read buffer
size by calling FileSystem::getFileSize.

9:38 PM Changeset in webkit [245846] by Fujii Hironori

• 3 edits in trunk/Source/WebCore

[WinCairo] ASSERTION FAILED: !m_preparingToPlay in MediaPlayerPrivateMediaFoundation::prepareToPlay
​https://bugs.webkit.org/show_bug.cgi?id=190747

Reviewed by Alex Christensen.

HTMLMediaElement::prepareToPlay had a assertion ensuring that it
was not called twice. However, it was called twice. The first from
HTMLMediaElement::load, the second from
MediaPlayerPrivateMediaFoundation::onTopologySet.

prepareToPlay started loading. And, loading should be started
after onTopologySet is called back.

Covered by existing tests.

• platform/graphics/win/MediaPlayerPrivateMediaFoundation.cpp:

(WebCore::MediaPlayerPrivateMediaFoundation::onTopologySet): Moved code from prepareToPlay.
(WebCore::MediaPlayerPrivateMediaFoundation::prepareToPlay): Deleted and moved the code to onTopologySet.

• platform/graphics/win/MediaPlayerPrivateMediaFoundation.h: Removed prepareToPlay declaration.

9:33 PM Changeset in webkit [245845] by Fujii Hironori

• 2 edits in trunk/Source/WebCore

[WinCairo][MediaFoundation] Assertion failure in MediaPlayerPrivateMediaFoundation::Direct3DPresenter::presentSample
​https://bugs.webkit.org/show_bug.cgi?id=198290

Reviewed by Per Arne Vollan.

Covered by existing tests.

• platform/graphics/win/MediaPlayerPrivateMediaFoundation.cpp:

(WebCore::MediaPlayerPrivateMediaFoundation::Direct3DPresenter::presentSample):
Call clear() of m_memSurface before assigning new value.

9:13 PM Changeset in webkit [245844] by sbarati@apple.com

• 3 edits
  2 adds in trunk

[WHLSL] Type of dereference is the type of the thing we point to, not a pointer to that type
​https://bugs.webkit.org/show_bug.cgi?id=198321

Reviewed by Myles C. Maxfield.

Source/WebCore:

Consider this program:
`
thread int* x;
*x = 42
`

In the Checker, we were saying the type of "*x" was "int*" instead of "int".

Test: webgpu/whlsl-dereference-pointer-should-type-check.html

• Modules/webgpu/WHLSL/WHLSLChecker.cpp:

(WebCore::WHLSL::Checker::visit):

LayoutTests:

• webgpu/whlsl-dereference-pointer-should-type-check-expected.html: Added.
• webgpu/whlsl-dereference-pointer-should-type-check.html: Added.

8:27 PM Changeset in webkit [245843] by Justin Michaud

• 2 edits in trunk/Tools

Attempt to fix JSC test timeouts after adding collectContinuously to WASM tests.
​https://bugs.webkit.org/show_bug.cgi?id=198322

Reviewed by Saam Barati.

Increases the collection period from 1 to slightly higher to try to speed up the tests. Any higher and
the test runner does not detect the bug that required the memset in Wasm::Instance::Instance().

• Scripts/run-jsc-stress-tests:

8:08 PM Changeset in webkit [245842] by Alan Coon

• 7 edits in branches/safari-608.1.24.20-branch/Source

Versioning.

8:02 PM Changeset in webkit [245841] by Alan Coon

• 1 copy in tags/Safari-608.1.24.20.7

Tag Safari-608.1.24.20.7.

8:00 PM Changeset in webkit [245840] by Alan Coon

• 7 edits in branches/safari-608.1.24.20-branch/Source

Versioning.

7:56 PM Changeset in webkit [245839] by Wenson Hsieh

• 7 edits in trunk

[iOS] Respect NSItemProvider's registered types when dropping files that are loaded in-place
​https://bugs.webkit.org/show_bug.cgi?id=198315
<rdar://problem/51183762>

Reviewed by Tim Horton.

Source/WebCore:

Currently, logic in PasteboardIOS.mm and WebContentReaderCocoa.mm attempts to deduce the content type from the
file path when dropping attachments on iOS. Instead, we should be plumbing the content type through to the
reader.

Test: WKAttachmentTestsIOS.InsertDroppedImageWithNonImageFileExtension

• editing/WebContentReader.h:
• editing/cocoa/WebContentReaderCocoa.mm:

(WebCore::typeForAttachmentElement):

Add a helper method to determine which type to use in attachment elements. This makes the paste
(attachmentForData) and drop (attachmentForFilePaths) behave the same way, with respect to the type attribute
used to represent the attachment.

(WebCore::attachmentForFilePath):

Use the content type, if specified; otherwise, fall back to deducing it from the file path.

(WebCore::attachmentForData):
(WebCore::WebContentReader::readFilePath):

• platform/Pasteboard.h:

(WebCore::PasteboardWebContentReader::readFilePath):

Pass the highest fidelity representation's content type to the web content reader.

• platform/ios/PasteboardIOS.mm:

(WebCore::Pasteboard::readRespectingUTIFidelities):

Tools:

Adds a new API test to verify that when dropping a file that is loaded in-place with a file extension that is
not a .png (but was registered to the item provider as "public.png"), the resulting attachment is contained in
an image element, and the resulting attachment info indicates that the dropped attachment is a png file.

Additionally, rebaseline some existing tests.

• TestWebKitAPI/Tests/WebKitCocoa/WKAttachmentTests.mm:

(runTestWithTemporaryImageFile):
(TestWebKitAPI::TEST):

7:27 PM Changeset in webkit [245838] by mmaxfield@apple.com

• 16 edits
  5 adds
  2 deletes in trunk

Move idempotent text autosizing to StyleTreeResolver
​https://bugs.webkit.org/show_bug.cgi?id=197808
<rdar://problem/50283983>

Reviewed by Antti Koivisto.

Source/WebCore:

This patch migrates the idempotent text autosizing code to live inside style resolution. This is almost
the same as the algorithm that uses the result of layout to calculate autosizing, but this version only
operates on style (and thus doesn't require double layouts). Because it is being run in an environment
with less information, autosizing is occurring in more places, so the curves have been adjusted to make
autosizing not boost as much as the previous implementation did. The new algorithm is modelled after
text-decorations-in-effect. I've claimed 4 of the unused bits in RenderStyle to contain the state of the
autosizing algorithm. StyleResolver::adjustRenderStyle() is where the algorithm is implemented:

• Look at the inherited bits
• Interogate the element's RenderStyle
• Compute new bits for the element, and set them in its RenderStyle
• Based on the newly computed bits, determine whether we should increase the text size
• If so, determine how much using the specified font size, and apply the result to the computed font size

This works because StyleBuilderCustom::applyInheritFontSize() inherits from the specified font size, not
the computed font size.

This patch also will disable autosizing using the other methods (so there aren't two methods of autosizing
fighting each other) and will honor text-size-adjust:none. However, it won't honor text-size-adjust:100%.
If content says text-size-adjust:100%, we will disregard it and take this code path.

Tests: fast/text-autosizing/ios/idempotentmode/css-exposure.html

fast/text-autosizing/ios/idempotentmode/idempotent-autosizing-skip.html
fast/text-autosizing/ios/idempotentmode/idempotent-autosizing-identity.html
fast/text-autosizing/ios/idempotentmode/idempotent-autosizing.html

• Sources.txt:
• WebCore.xcodeproj/project.pbxproj:
• css/CSSComputedStyleDeclaration.cpp:

(WebCore::ComputedStyleExtractor::valueForPropertyinStyle):

• css/CSSProperties.json:
• css/StyleBuilderCustom.h:

(WebCore::computeBaseSpecifiedFontSize):
(WebCore::computeLineHeightMultiplierDueToFontSize):

• css/StyleResolver.cpp:

(WebCore::idempotentTextSize):
(WebCore::hasTextChildren):
(WebCore::StyleResolver::adjustRenderStyle):
(WebCore::StyleResolver::checkForTextSizeAdjust):

• page/FrameViewLayoutContext.cpp:

(WebCore::FrameViewLayoutContext::applyTextSizingIfNeeded):

• rendering/RenderBlockFlow.cpp:

(WebCore::RenderBlockFlow::adjustComputedFontSizes):
(WebCore::idempotentTextSize): Deleted.

• rendering/RenderBlockFlow.h:
• rendering/RenderElement.cpp:

(WebCore::includeNonFixedHeight):
(WebCore::RenderElement::adjustComputedFontSizesOnBlocks):
(WebCore::RenderElement::resetTextAutosizing):

• rendering/style/RenderStyle.cpp:

(WebCore::RenderStyle::RenderStyle):
(WebCore::RenderStyle::autosizeStatus const):
(WebCore::RenderStyle::setAutosizeStatus):

• rendering/style/RenderStyle.h:
• rendering/style/TextSizeAdjustment.cpp: Added.

(WebCore::AutosizeStatus::AutosizeStatus):
(WebCore::AutosizeStatus::contains const):
(WebCore::AutosizeStatus::modifiedStatus const):
(WebCore::AutosizeStatus::shouldSkipSubtree const):

• rendering/style/TextSizeAdjustment.h:

LayoutTests:

• fast/text-autosizing/ios/idempotentmode/css-exposure-expected.txt: Added.
• fast/text-autosizing/ios/idempotentmode/css-exposure.html: Added.
• fast/text-autosizing/ios/idempotentmode/idempotent-autosizing-skip-expected.html: Added.
• fast/text-autosizing/ios/idempotentmode/idempotent-autosizing-skip.html: Added.
• platform/ipad/fast/text-autosizing/text-size-adjust-inline-style-expected.html: Removed.
• platform/ipad/fast/text-autosizing/text-size-adjust-inline-style.html: Removed.

We're intentionally not honoring percentages, because this is the most common way that
text autosizing is disabled (by setting it to 100%) on the Web today. However, Web authors
that have done this did it without knowing the full extent of the behavior change, and
the new idempotent text autosizing code path seems to be a progression in most cases
we've seen.

6:07 PM Changeset in webkit [245837] by Simon Fraser

• 9 edits
  7 adds in trunk

Use scroll-velocity-based tile coverage for overflow:scroll
​https://bugs.webkit.org/show_bug.cgi?id=198294
rdar://problem/48942184

Reviewed by Tim Horton.

Source/WebCore:

Start using a velocity-based tile coverage computation on layers with Type::ScrolledContents,
which is the content layers for overflow:scroll when they get big enough to get tiled.

Move legacy macOS coverage code into adjustTileCoverageForDesktopPageScrolling() because
I don't want to change its behavior in this patch. Use TileController::adjustTileCoverageRectForScrolling()
for iOS and macOS overflow scrolling. Since only iOS page scrolling gets velocity data from the UI
process, compute velocity in TileController using the visible rect top-left.

For overflow scroll, we have to plumb horizontal and vertical coverage in from
RenderLayerBacking.

Tests: tiled-drawing/scrolling/overflow/overflow-scrolled-down-tile-coverage.html

tiled-drawing/scrolling/overflow/overflow-scrolled-up-tile-coverage.html
tiled-drawing/scrolling/overflow/overflow-tile-coverage.html

• platform/graphics/TiledBacking.h:
• platform/graphics/ca/GraphicsLayerCA.cpp:

(WebCore::GraphicsLayerCA::adjustCoverageRect const):

• platform/graphics/ca/PlatformCALayer.h:
• platform/graphics/ca/TileController.cpp:

(WebCore::TileController::setVelocity):
(WebCore::TileController::adjustTileCoverageRect):
(WebCore::TileController::adjustTileCoverageForDesktopPageScrolling const):
(WebCore::TileController::adjustTileCoverageWithScrollingVelocity const):
(WebCore::TileController::adjustTileCoverageRectForScrolling):
(WebCore::expandRectWithinRect): Deleted.
(WebCore::TileController::adjustTileCoverageRect const): Deleted.
(WebCore::TileController::adjustTileCoverageRectForScrolling const): Deleted.

• platform/graphics/ca/TileController.h:
• rendering/RenderLayer.h:
• rendering/RenderLayerBacking.cpp:

(WebCore::computePageTiledBackingCoverage):
(WebCore::computeOverflowTiledBackingCoverage):
(WebCore::RenderLayerBacking::adjustTiledBackingCoverage):
(WebCore::RenderLayerBacking::updateGeometry):

LayoutTests:

• tiled-drawing/scrolling/overflow/overflow-scrolled-down-tile-coverage-expected.txt: Added.
• tiled-drawing/scrolling/overflow/overflow-scrolled-down-tile-coverage.html: Added.
• tiled-drawing/scrolling/overflow/overflow-scrolled-up-tile-coverage-expected.txt: Added.
• tiled-drawing/scrolling/overflow/overflow-scrolled-up-tile-coverage.html: Added.
• tiled-drawing/scrolling/overflow/overflow-tile-coverage-expected.txt: Added.
• tiled-drawing/scrolling/overflow/overflow-tile-coverage.html: Added.

6:05 PM Changeset in webkit [245836] by ysuzuki@apple.com

• 2 edits in trunk/Tools

GCHeapInspector should accept weird filename
​https://bugs.webkit.org/show_bug.cgi?id=198314

Reviewed by Simon Fraser.

GCHeapInspector filenameForPath should have a fallback path if regexp does not match against the given path.

• GCHeapInspector/script/interface.js:

(filenameForPath):

5:34 PM Changeset in webkit [245835] by Brent Fulgham

• 2 edits in trunk/Source/WebKit

Fix sandbox violation when using QuickLook on iOS
​https://bugs.webkit.org/show_bug.cgi?id=198312
<rdar://problem/51134351>

Reviewed by Alexey Proskuryakov.

• Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:

5:20 PM Changeset in webkit [245834] by Alan Coon

• 1 copy in tags/Safari-608.1.24.20.6

Tag Safari-608.1.24.20.6.

5:14 PM Changeset in webkit [245833] by Devin Rousso

• 5 edits in trunk/Source/WebInspectorUI

Web Inspector: Timelines: spacing around pie chart is different between CPU and Memory
​https://bugs.webkit.org/show_bug.cgi?id=198299

Reviewed by Matt Baker.

"Style Resolution" is a much longer string than any of the legend strings in the Memory
timeline, and causes the CPU timeline legend to shift as a result.

Rename "Script" to "JavaScript" and "Style Resolution" to "Styles" so that the strings are
roughly the same length between the CPU and Memory timelines, meaning that they will appear
in the same spot with similar sizing.

• UserInterface/Views/CPUTimelineView.js:

(WI.CPUTimelineView.displayNameForSampleType):
(WI.CPUTimelineView.prototype.initialLayout):
(WI.CPUTimelineView.prototype._computeStatisticsData):
Drive-by: remove the WI prefix from all CPUTimelineView.SampleType since it's within the
same class.

• UserInterface/Views/CPUTimelineView.css:

(.timeline-view.cpu > .content > .overview .legend > .row > .swatch.sample-type-javascript): Added.
(.timeline-view.cpu > .content > .overview .legend > .row > .swatch.sample-type-script): Deleted.
Drive-by: move the .overview rules lower to be in the same area as the .overview * rules.

• UserInterface/Views/MemoryTimelineView.css:

(.timeline-view.memory > .content > .overview):
Drive-by: remove duplicate CSS rule.

• Localizations/en.lproj/localizedStrings.js:

4:57 PM Changeset in webkit [245832] by timothy_horton@apple.com

• 3 edits in trunk/Source/WebKit

Horizontal scrollbar flashes after scrolling vertically with keyboard
​https://bugs.webkit.org/show_bug.cgi?id=197942
<rdar://problem/46169578>

Reviewed by Dean Jackson.

• UIProcess/ios/WKKeyboardScrollingAnimator.mm:

(axesForDelta):
(-[WKKeyboardScrollViewAnimator scrollToContentOffset:animated:]):
Only flash relevant axes.

• Platform/spi/ios/UIKitSPI.h:

4:40 PM Changeset in webkit [245831] by achristensen@apple.com

• 2 edits in trunk/Source/WebKit

Build Fix.

• UIProcess/WKImagePreviewViewController.mm:

(-[WKImagePreviewViewController IGNORE_WARNINGS_END]):
(-[WKImagePreviewViewController previewActionItems]): Deleted.

4:27 PM Changeset in webkit [245830] by timothy_horton@apple.com

• 2 edits in trunk/Source/WebKit

REGRESSION(r245795): causing internal testers to exit early after 50 crashes.
​https://bugs.webkit.org/show_bug.cgi?id=198310
<rdar://problem/51192535>

Reviewed by Simon Fraser.

• UIProcess/RemoteLayerTree/ios/RemoteLayerTreeViews.mm:

(WebKit::touchActionsForPoint):
Initialize hitView to nil.

4:13 PM Changeset in webkit [245829] by youenn@apple.com

• 3 edits
  3 adds in trunk

createAnswer() SDP Rejected by setLocalDescription()
​https://bugs.webkit.org/show_bug.cgi?id=195930
<rdar://problem/49030489>

Reviewed by Eric Carlson.

Source/ThirdParty/libwebrtc:

Make sure to check packetization mode parameter when matching H264 video codec.

• Source/webrtc/media/base/codec.cc:
• WebKit/0001-fix-195930.patch: Added.

LayoutTests:

• webrtc/h264-packetization-mode-expected.txt: Added.
• webrtc/h264-packetization-mode.html: Added.

3:41 PM Changeset in webkit [245828] by eric.carlson@apple.com

• 2 edits in trunk/Tools

[MacOS] Filter GVA warning logged to stdout
​https://bugs.webkit.org/show_bug.cgi?id=198303
<rdar://problem/50098041>

Reviewed by Jer Noble.

• Scripts/webkitpy/port/mac.py:

(MacPort.logging_patterns_to_strip): Filter GVA warnings.

3:22 PM Changeset in webkit [245827] by Nikita Vasilyev

• 20 edits in trunk

Web Inspector: Provide UIString descriptions to improve localizations
​https://bugs.webkit.org/show_bug.cgi?id=195132
<rdar://problem/48457817>

Reviewed by Devin Rousso.

Source/WebInspectorUI:

• Localizations/en.lproj/localizedStrings.js:
• UserInterface/Base/LoadLocalizedStrings.js:

(WI.UIString):
(WI.repeatedUIString.timelineRecordLayout):
(WI.repeatedUIString.timelineRecordPaint):
(WI.repeatedUIString.timelineRecordComposite):
(WI.repeatedUIString.allExceptions):
(WI.repeatedUIString.uncaughtExceptions):
(WI.repeatedUIString.assertionFailures):
(WI.repeatedUIString.allRequests):
(WI.repeatedUIString.fetch):
(WI.repeatedUIString.revealInDOMTree):

• UserInterface/Models/LayoutTimelineRecord.js:

(WI.LayoutTimelineRecord.displayNameForEventType):

• UserInterface/Models/RenderingFrameTimelineRecord.js:

(WI.RenderingFrameTimelineRecord.displayNameForTaskType):

• UserInterface/Models/Resource.js:

(WI.Resource.displayNameForType):

• UserInterface/Views/AuditTestGroupContentView.js:

(WI.AuditTestGroupContentView.prototype.layout):

• UserInterface/Views/CPUTimelineView.js:

(WI.CPUTimelineView.displayNameForSampleType):

• UserInterface/Views/ContextMenuUtilities.js:
• UserInterface/Views/DOMBreakpointTreeElement.js:

(WI.DOMBreakpointTreeElement.displayNameForType):

• UserInterface/Views/DOMNodeTreeElement.js:

(WI.DOMNodeTreeElement.prototype.populateContextMenu):
(WI.DOMNodeTreeElement):

• UserInterface/Views/DOMTreeElement.js:

(WI.DOMTreeElement.prototype.populateDOMNodeContextMenu):

• UserInterface/Views/DebuggerSidebarPanel.js:

(WI.DebuggerSidebarPanel.prototype._addBreakpoint):
(WI.DebuggerSidebarPanel.prototype._handleCreateBreakpointMouseDown):
(WI.DebuggerSidebarPanel):

• UserInterface/Views/LayerTreeDetailsSidebarPanel.js:

(WI.LayerTreeDetailsSidebarPanel.prototype._buildDataGridSection):

• UserInterface/Views/NetworkTableContentView.js:

(WI.NetworkTableContentView.shortDisplayNameForResourceType):
(WI.NetworkTableContentView.prototype.initialLayout):

• UserInterface/Views/SourcesNavigationSidebarPanel.js:

(WI.SourcesNavigationSidebarPanel.prototype._addBreakpoint):
(WI.SourcesNavigationSidebarPanel.prototype._handleCreateBreakpointMouseDown):

• UserInterface/Views/SpreadsheetCSSStyleDeclarationSection.js:

(WI.SpreadsheetCSSStyleDeclarationSection.prototype._renderSelector):

• UserInterface/Views/SpreadsheetRulesStyleDetailsPanel.js:

(WI.SpreadsheetRulesStyleDetailsPanel.prototype.layout):

Tools:

Allow WI.UIString to take:

• WI.UIString(string, key, comment)
• WI.UIString(string, comment)
• WI.UIString(string)

• Scripts/extract-localizable-js-strings:

2:55 PM Changeset in webkit [245826] by Dewei Zhu

• 3 edits in trunk/Tools

Limit run-benchmark http server to specific interface.
​https://bugs.webkit.org/show_bug.cgi?id=198247

Reviewed by Ryosuke Niwa.

Add '--interface' option to 'twisted_http_server.py'.
'SimpleHTTPServerDriver' should specify interface for http server.
Update regex that determines http server port from 'lsof' output to support ipv6 address.

• Scripts/webkitpy/benchmark_runner/http_server_driver/http_server/twisted_http_server.py:

Added '--interface' argument.

• Scripts/webkitpy/benchmark_runner/http_server_driver/simple_http_server_driver.py:

Limited http server to a specific interfce.
Added '-P' option to 'lsof' invocation to inhibits the conversion of port numbers to port name so script won't
fail if it's listening on a 'well-known' port.
Abstracted '_check_http_server_is_running' to allow potential child class to have its own implementation.
(SimpleHTTPServerDriver.serve): Updated regex that determines http server port from 'lsof' output to support ipv6 address.
(SimpleHTTPServerDriver.kill_server): Added null check for 'self._server_process'.

2:10 PM Changeset in webkit [245825] by Shawn Roberts

• 15 edits
  29 deletes in trunk

Unreviewed, rolling out r245475.

Newly imported test is flaky. Features need flags.

Reverted changeset:

"Implement imagesrcset and imagesizes attributes on link
rel=preload"
​https://bugs.webkit.org/show_bug.cgi?id=192950
​https://trac.webkit.org/changeset/245475

2:04 PM Changeset in webkit [245824] by commit-queue@webkit.org

• 4 edits in trunk/Tools

webkitpy: Switch run-webkit-tests to tailspin
​https://bugs.webkit.org/show_bug.cgi?id=198144
<rdar://problem/32463212>

Patch by David Xiong <​w_xiong@apple.com> on 2019-05-28
Reviewed by Jonathan Bedard.

Changes run-webkit-tests to run tailspin on test time out
rather than spindump, and edited tests to look for tailspin logs
instead.

• Scripts/webkitpy/port/darwin.py:

(DarwinPort.sample_process): Replaced spindump with tailspin (+ symbolication)
(DarwinPort):
(DarwinPort.tailspin_file_path):
(DarwinTest.spindump_file_path): Deleted.

• Scripts/webkitpy/port/darwin_testcase.py:

(DarwinTest.test_tailspin): Changed spindump test (below) to test for tailspin instead
(DarwinTest.test_spindump): Deleted.
(DarwinTest.test_spindump.logging_run_command): Deleted.

• Scripts/webkitpy/port/ios_device_unittest.py:

(IOSDeviceTest.test_tailspin): Changed spindump tests (inc. below) to test for tailspin instead
(IOSDeviceTest.test_sample_process.logging_run_command):
(IOSDeviceTest.test_sample_process_exception.throwing_run_command):
(IOSDeviceTest.test_spindump): Deleted.
(IOSDeviceTest.test_spindump.logging_run_command): Deleted.

2:02 PM Changeset in webkit [245823] by Brent Fulgham

• 18 edits in trunk/Source

Protect frames during style and layout changes
​https://bugs.webkit.org/show_bug.cgi?id=198047
<rdar://problem/50954082>

Reviewed by Zalan Bujtas.

Be more careful about the scope and lifetime of objects that participate in layout or
style updates. If a method decides a layout or style update is needed, it needs to
confirm that the elements it was operating on are still valid and needed in the
current operation.

Source/WebCore:

• accessibility/AXObjectCache.cpp:

(WebCore::AXObjectCache::getOrCreate):

• accessibility/AccessibilityRenderObject.cpp:

(WebCore::AccessibilityRenderObject::accessibilityHitTest const):

• css/CSSComputedStyleDeclaration.cpp:

(WebCore::ComputedStyleExtractor::valueForPropertyinStyle):

• css/CSSComputedStyleDeclaration.h:
• css/SVGCSSComputedStyleDeclaration.cpp:

(WebCore::ComputedStyleExtractor::svgPropertyValue):

• dom/Document.cpp:

(WebCore::Document::setFocusedElement):

• editing/TypingCommand.cpp:

(WebCore::TypingCommand::insertTextRunWithoutNewlines):
(WebCore::TypingCommand::insertLineBreak):
(WebCore::TypingCommand::insertParagraphSeparator):
(WebCore::TypingCommand::insertParagraphSeparatorInQuotedContent):

• editing/ios/EditorIOS.mm:

(WebCore::Editor::setDictationPhrasesAsChildOfElement):

• html/HTMLLabelElement.cpp:

(WebCore::HTMLLabelElement::focus):

• html/HTMLTextAreaElement.cpp:

(WebCore::HTMLTextAreaElement::appendFormData):

• html/ImageDocument.cpp:

(WebCore::ImageDocument::imageClicked):

• html/ValidationMessage.cpp:

(WebCore::ValidationMessage::buildBubbleTree):

• page/FrameView.cpp:

(WebCore::FrameView::autoSizeIfEnabled):
(WebCore::FrameView::trackedRepaintRectsAsText const):

• page/PrintContext.cpp:

(WebCore::PrintContext::pageProperty):
(WebCore::PrintContext::numberOfPages):
(WebCore::PrintContext::spoolAllPagesWithBoundaries):

Source/WebKitLegacy/mac:

• DOM/DOM.mm:

(-[DOMRange renderedImageForcingBlackText:renderedImageForcingBlackText:]):

• WebView/WebHTMLView.mm:

(-[WebHTMLView _selectionDraggingImage]):
(-[WebHTMLView selectionImageForcingBlackText:selectionImageForcingBlackText:]):

12:29 PM Changeset in webkit [245822] by Brent Fulgham

• 4 edits in trunk/Source/WebKit

Remove dead code from sandboxes
​https://bugs.webkit.org/show_bug.cgi?id=198300

Reviewed by Alexey Proskuryakov.

SSIA.

• NetworkProcess/mac/com.apple.WebKit.NetworkProcess.sb.in:
• PluginProcess/mac/com.apple.WebKit.plugin-common.sb.in:
• WebProcess/com.apple.WebProcess.sb.in:

12:10 PM Changeset in webkit [245821] by Devin Rousso

• 4 edits in trunk/LayoutTests

REGRESSION (r244182): inspector/canvas/recording-webgl-snapshots.html became flaky on WK1
​https://bugs.webkit.org/show_bug.cgi?id=196875
<rdar://problem/49873252>

Reviewed by Said Abou-Hallawa.

Remove the frameLimit configuration on these tests, as they're short enough that when
running in debug, the timing between the InspectorCanvasAgent automatically stopping the
recording and the test page's LastFrame event (which will manually stop the recording) is
too close and can result in a race condition.

Instead, just wait for the test page to say "done" (LastFrame) before stopping the
recording, ensuring that the full "flow" of the recording is under the control of the test.

This isn't an issue when using Web Inspector "normally" (e.g. not in a test), because the
frontend UI will "ignore" these types of errors, not to mention it would be much harder for
a person to get the timing just right to even encounter this situation.

• inspector/canvas/recording-webgl-snapshots.html:
• inspector/canvas/recording-webgl2-snapshots.html:
• platform/mac/TestExpectations:

12:08 PM Changeset in webkit [245820] by commit-queue@webkit.org

• 5 edits in trunk/LayoutTests

[Pointer Events WPT] Unflake imported/w3c/web-platform-tests/pointerevents/pointerevent_suppress_compat_events_on_drag_mouse.html
​https://bugs.webkit.org/show_bug.cgi?id=197008

Patch by Antoine Quint <Antoine Quint> on 2019-05-28
Reviewed by Jon Lee.

We raised an issue on the WPT test which was testing the event timestamp in an invalid manner (see ​https://github.com/w3c/pointerevents/issues/284
and ​https://github.com/web-platform-tests/wpt/issues/170160). As a result this test now passes reliably, so we can update the expected out and
the TestExpectations.

LayoutTests/imported/w3c:

• web-platform-tests/pointerevents/pointerevent_suppress_compat_events_on_drag_mouse-expected.txt:
• web-platform-tests/pointerevents/pointerevent_suppress_compat_events_on_drag_mouse.html:

LayoutTests:

• platform/mac/TestExpectations:

12:06 PM Changeset in webkit [245819] by commit-queue@webkit.org

• 2 edits in trunk/Tools

webkitpy: Using sudo on iOS device during timeout spindumps
​https://bugs.webkit.org/show_bug.cgi?id=198142

Patch by David Xiong <​w_xiong@apple.com> on 2019-05-28
Reviewed by Jonathan Bedard.

• Scripts/webkitpy/port/darwin.py:

(DarwinPort.sample_process): Check target host instead of source host for sudo command.

11:11 AM Changeset in webkit [245818] by Antti Koivisto

• 20 edits
  4 adds in trunk

[async scrolling] Fixed positioning inside stacking context overflow scroll is jumpy
​https://bugs.webkit.org/show_bug.cgi?id=198292

Reviewed by Darin Adler.

Source/WebCore:

Tests: scrollingcoordinator/ios/fixed-in-stacking-context-overflow-scroll-2.html

scrollingcoordinator/ios/fixed-in-stacking-context-overflow-scroll.html

We were computing delta from the layout scroll position in ScrollingTree::notifyRelatedNodesAfterScrollPositionChange
based on the passed in node only. If any other node had deltas they were not taken into account at all. This would occur
frequently since the function is always invoked for the root node after layer tree commit.

Fix by moving the delta computation (and fetching layoutViewport) to ScrollingTreeFixedNode.

• page/scrolling/ScrollingTree.cpp:

(WebCore::ScrollingTree::applyLayerPositions):

No need to pass offset and layoutViewport around anymore.

(WebCore::ScrollingTree::applyLayerPositionsRecursive):
(WebCore::ScrollingTree::notifyRelatedNodesAfterScrollPositionChange):

Remove the offset and layoutViewport computations.

(WebCore::ScrollingTree::notifyRelatedNodesRecursive):

• page/scrolling/ScrollingTree.h:
• page/scrolling/ScrollingTreeFrameHostingNode.cpp:

(WebCore::ScrollingTreeFrameHostingNode::applyLayerPositions):

• page/scrolling/ScrollingTreeFrameHostingNode.h:
• page/scrolling/ScrollingTreeNode.cpp:

(WebCore::ScrollingTreeNode::relatedNodeScrollPositionDidChange):

• page/scrolling/ScrollingTreeNode.h:
• page/scrolling/ScrollingTreeScrollingNode.cpp:

(WebCore::ScrollingTreeScrollingNode::applyLayerPositions):

• page/scrolling/ScrollingTreeScrollingNode.h:
• page/scrolling/cocoa/ScrollingTreeFixedNode.h:
• page/scrolling/cocoa/ScrollingTreeFixedNode.mm:

(WebCore::ScrollingTreeFixedNode::applyLayerPositions):

Compute them here instead, always taking all overflow scrollers up to the closest frame into account.

• page/scrolling/cocoa/ScrollingTreePositionedNode.h:
• page/scrolling/cocoa/ScrollingTreePositionedNode.mm:

(WebCore::ScrollingTreePositionedNode::applyLayerPositions):
(WebCore::ScrollingTreePositionedNode::relatedNodeScrollPositionDidChange):

• page/scrolling/cocoa/ScrollingTreeStickyNode.h:
• page/scrolling/cocoa/ScrollingTreeStickyNode.mm:

(WebCore::ScrollingTreeStickyNode::applyLayerPositions):

LayoutTests:

• scrollingcoordinator/ios/fixed-in-stacking-context-overflow-scroll-2-expected.html: Added.
• scrollingcoordinator/ios/fixed-in-stacking-context-overflow-scroll-2.html: Added.
• scrollingcoordinator/ios/fixed-in-stacking-context-overflow-scroll-expected.html: Added.
• scrollingcoordinator/ios/fixed-in-stacking-context-overflow-scroll.html: Added.

10:33 AM Changeset in webkit [245817] by Brent Fulgham

• 3 edits in trunk/Source/WebKit

Update sandbox rules for more News use cases
​https://bugs.webkit.org/show_bug.cgi?id=198236
<rdar://problem/50054027>

Reviewed by Alexey Proskuryakov.

Update the WebContent and Network process sandboxes so that News has the same set of allowed
service access as regular WebKit views.

• NetworkProcess/mac/com.apple.WebKit.NetworkProcess.sb.in:
• WebProcess/com.apple.WebProcess.sb.in:

10:29 AM Changeset in webkit [245816] by Alan Bujtas

• 2 edits in trunk/Source/WebCore

[LFC][Verification] Add additional inline and block checks
​https://bugs.webkit.org/show_bug.cgi?id=198252
<rdar://problem/51140687>

Reviewed by Antti Koivisto.

Now we also test the geometry of the blocks with inline formatting contexts.

• layout/Verification.cpp:

(WebCore::Layout::checkForMatchingTextRuns):
(WebCore::Layout::verifyAndOutputSubtree):

10:19 AM Changeset in webkit [245815] by msaboff@apple.com

• 3 edits
  1 add in trunk

[YARR] Properly handle RegExp's that require large ParenContext space
​https://bugs.webkit.org/show_bug.cgi?id=198065

Reviewed by Keith Miller.

JSTests:

New test.

• stress/regexp-large-paren-context.js: Added.

(testLargeRegExp):

Source/JavaScriptCore:

Changed what happens when we exceed VM::patternContextBufferSize when compiling a RegExp
that needs ParenCOntextSpace to fail the RegExp JIT compilation and fall back to the YARR
interpreter. This can save large amounts of JIT memory for a
JIT'ed function that cannot ever succeed.

• yarr/YarrJIT.cpp:

(JSC::Yarr::YarrGenerator::initParenContextFreeList):
(JSC::Yarr::YarrGenerator::compile):

10:09 AM Changeset in webkit [245814] by aakash_jain@apple.com

• 2 edits in trunk/Tools

[ews-build] Remove unused buildbot tabs
​https://bugs.webkit.org/show_bug.cgi?id=198108

Reviewed by Jonathan Bedard.

• BuildSlaveSupport/ews-build/templates/build.jade: Removed unused 'Changes' and 'Responsible Users' tabs.

9:03 AM Changeset in webkit [245813] by Tadeu Zagallo

• 4 edits
  1 add in trunk

JITOperations putByVal should mark negative array indices as out-of-bounds
​https://bugs.webkit.org/show_bug.cgi?id=198271

Reviewed by Saam Barati.

JSTests:

• microbenchmarks/get-by-val-negative-array-index.js:

(foo):
Update the getByVal microbenchmark added in r245769. This now shows that r245769
is 4.2x faster than the previous commit.

• microbenchmarks/put-by-val-negative-array-index.js: Added.

(foo):

Source/JavaScriptCore:

Similar to what was done to getByVal in r245769, we should also mark put_by_val as out-of-bounds
when we exit from DFG for putting to a negative index. This avoids the same scenario where we keep
recompiling a CodeBlock with DFG and exiting at the same bytecode.

This is a 3.7x improvement in the microbenchmark being added: put-by-val-negative-array-index.js.

• jit/JITOperations.cpp:

8:36 AM Changeset in webkit [245812] by Alan Bujtas

• 5 edits in trunk/Source/WebCore

[LFC][IFC] Decouple line layout and processing inline runs.
​https://bugs.webkit.org/show_bug.cgi?id=198282
<rdar://problem/51167954>

Reviewed by Antti Koivisto.

This is in preparation for using "createInlineRunsForLine" logic when computing preferred width.

• layout/inlineformatting/InlineFormattingContext.h:
• layout/inlineformatting/InlineFormattingContextLineLayout.cpp:

(WebCore::Layout::UncommittedContent::size const):
(WebCore::Layout::InlineFormattingContext::LineLayout::createInlineRunsForLine const):
(WebCore::Layout::InlineFormattingContext::LineLayout::layout const):
(WebCore::Layout::InlineFormattingContext::LineLayout::processInlineRuns const):
(WebCore::Layout::InlineFormattingContext::LineLayout::closeLine const): Deleted.

• layout/inlineformatting/InlineFormattingState.h:

(WebCore::Layout::InlineFormattingState::addInlineItem):

• layout/inlineformatting/InlineTextItem.cpp:

(WebCore::Layout::InlineTextItem::createAndAppendTextItems):

7:51 AM Changeset in webkit [245811] by Alan Bujtas

• 4 edits in trunk/Source/WebCore

[LFC][IFC] Move intrinsic width computation from InlineFormattingContext to LineLayout
​https://bugs.webkit.org/show_bug.cgi?id=198258

Reviewed by Antti Koivisto.

This is in preparation for sharing even more code between line layout and preferred width computation.

• layout/inlineformatting/InlineFormattingContext.cpp:

(WebCore::Layout::InlineFormattingContext::layout const):
(WebCore::Layout::InlineFormattingContext::computeIntrinsicWidthConstraints const):

• layout/inlineformatting/InlineFormattingContext.h:
• layout/inlineformatting/InlineFormattingContextLineLayout.cpp:

(WebCore::Layout::InlineFormattingContext::LineLayout::initializeLine const):
(WebCore::Layout::InlineFormattingContext::LineLayout::layout const):
(WebCore::Layout::InlineFormattingContext::LineLayout::computedIntrinsicWidth const):
(WebCore::Layout::InlineFormattingContext::LineLayout::closeLine const):

7:48 AM Changeset in webkit [245810] by Alan Bujtas

• 2 edits in trunk/Source/WebCore

[LFC[IFC] Ignore the initial strut's height when the line does not have any content.
​https://bugs.webkit.org/show_bug.cgi?id=198268
<rdar://problem/51150057>

Reviewed by Antti Koivisto.

The strut (​https://www.w3.org/TR/CSS22/visudet.html#leading) defines the initial logical height
for the line. This height should be ignored though when the line does not have any content.

• layout/inlineformatting/InlineFormattingContextLineLayout.cpp:

(WebCore::Layout::InlineFormattingContext::LineLayout::closeLine const):

5:33 AM Changeset in webkit [245809] by graouts@webkit.org

• 4 edits in trunk/Source/WebCore

[Pointer Events] Check that capturing data managed by the PointerCaptureController gets cleared upon navigation
​https://bugs.webkit.org/show_bug.cgi?id=198191

Reviewed by Dean Jackson.

When the document of the page's main frame changes, make sure we clear all of the data accumulated for the previous document.
I don't think this particular change is testable as none of the data contained in the PointerIdToCapturingDataMap maintained by
the PointerCaptureController contains any data that could be inspected by the page due to other fixes landed to fix wkb.ug/198129,
but I've checked that removing those fixes and using this patch correctly fixes that bug.

• page/Page.cpp:

(WebCore::Page::didChangeMainDocument):

• page/PointerCaptureController.cpp:

(WebCore::PointerCaptureController::PointerCaptureController):
(WebCore::PointerCaptureController::reset):

• page/PointerCaptureController.h:

1:45 AM Changeset in webkit [245808] by ysuzuki@apple.com

• 6 edits in trunk/Source/JavaScriptCore

Unreviewed, revert r242070 due to Membuster regression
​https://bugs.webkit.org/show_bug.cgi?id=195013

Membuster shows ~0.3% regression.

• heap/Heap.cpp:

(JSC::Heap::Heap):
(JSC::Heap::runBeginPhase):

• heap/Heap.h:
• heap/HeapInlines.h:

(JSC::Heap::forEachSlotVisitor):
(JSC::Heap::numberOfSlotVisitors): Deleted.

• heap/MarkingConstraintSolver.cpp:

(JSC::MarkingConstraintSolver::didVisitSomething const):

• heap/SlotVisitor.h:

May 27, 2019:

11:36 PM Changeset in webkit [245807] by Carlos Garcia Campos

• 49 edits in trunk

[GTK] Use WPEBackend-fdo for accelerating compositing in Wayland instead of the nested compositor
​https://bugs.webkit.org/show_bug.cgi?id=197944

Reviewed by Michael Catanzaro.

.:

Add USE_WPE_RENDERER option to use WPEBackend-fdo.

• Source/cmake/OptionsGTK.cmake:
• Source/cmake/OptionsPlayStation.cmake:
• Source/cmake/OptionsWPE.cmake:

Source/WebCore:

• PlatformGTK.cmake:
• SourcesGTK.txt:
• platform/graphics/GLContext.cpp:

(WebCore::GLContext::createContextForWindow): Check current display is X11 before trying to create a GLX context.

• platform/graphics/PlatformDisplay.cpp:

(WebCore::PlatformDisplay::createPlatformDisplay): Use USE(WPE_RENDERER) instead of USE(LIBWPE).

• platform/graphics/PlatformDisplay.h:
• platform/graphics/egl/GLContextEGL.cpp:

(WebCore::GLContextEGL::createWindowContext): Use USE(WPE_RENDERER) instead of PLATFORM(WPE).
(WebCore::GLContextEGL::createContext): Ditto.
(WebCore::GLContextEGL::createSharingContext): Ditto.
(WebCore::GLContextEGL::~GLContextEGL): Ditto.

• platform/graphics/egl/GLContextEGL.h:
• platform/graphics/egl/GLContextEGLLibWPE.cpp:
• platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp:

(WebCore::MediaPlayerPrivateGStreamerBase::ensureGstGLContext): Ditto.

• platform/graphics/libwpe/PlatformDisplayLibWPE.cpp:

(WebCore::PlatformDisplayLibWPE::PlatformDisplayLibWPE): In GTK port set the display as the shared one for compositing.

• platform/graphics/libwpe/PlatformDisplayLibWPE.h:

Source/WebKit:

The WPEBackend-fdo implementation is quite similar to our Wayland nested compositor, but more efficient (using
linux dmabuf, resource caching, etc.). This allows us to share even more code with WPE port too.

• PlatformGTK.cmake:
• Shared/CoordinatedGraphics/threadedcompositor/ThreadedCompositor.cpp:

(WebKit::ThreadedCompositor::forceRepaint): WPE doesn't support force repaint yet.

• Shared/WebPageCreationParameters.cpp:

(WebKit::WebPageCreationParameters::encode const): Use USE(WPE_RENDERER) instead of PLATFORM(WPE).
(WebKit::WebPageCreationParameters::decode): Ditto.

• Shared/WebPageCreationParameters.h:
• Shared/WebProcessCreationParameters.cpp:

(WebKit::WebProcessCreationParameters::encode const): Ditto.
(WebKit::WebProcessCreationParameters::decode): Ditto.

• Shared/WebProcessCreationParameters.h:
• SourcesGTK.txt:
• UIProcess/API/gtk/PageClientImpl.cpp:

(WebKit::PageClientImpl::hostFileDescriptor): Implement it when using WPE just calling webkitWebViewBaseRenderHostFileDescriptor().

• UIProcess/API/gtk/PageClientImpl.h:
• UIProcess/API/gtk/WebKitWebViewBase.cpp:

(webkitWebViewBaseCreateWebPage): Initialize the web page after the accelerated backing store is created, since
it's used now during intialization.
(webkitWebViewBaseRenderHostFileDescriptor): Return AcceleratedBackingStore::renderHostFileDescriptor()

• UIProcess/API/gtk/WebKitWebViewBasePrivate.h:
• UIProcess/CoordinatedGraphics/DrawingAreaProxyCoordinatedGraphics.cpp:
• UIProcess/PageClient.h:
• UIProcess/WebPageProxy.cpp:

(WebKit::WebPageProxy::creationParameters): Use USE(WPE_RENDERER) instead of PLATFORM(WPE).

• UIProcess/WebProcessPool.cpp:

(WebKit::WebProcessPool::initializeNewWebProcess): Remove wayland display name initialization.

• UIProcess/glib/WebProcessPoolGLib.cpp:

(WebKit::WebProcessPool::platformInitializeWebProcess): Initialize wayland display name when not using WPE here
now. In case of using WPE, initialize hostClientFileDescriptor and implementationLibraryName.

• UIProcess/glib/WebProcessProxyGLib.cpp:

(WebKit::WebProcessProxy::platformGetLaunchOptions): Remove unused wayland socket command line process parameter.

• UIProcess/gtk/AcceleratedBackingStore.h:

(WebKit::AcceleratedBackingStore::renderHostFileDescriptor): Return -1 by default.

• UIProcess/gtk/AcceleratedBackingStoreWayland.cpp:

(WebKit::AcceleratedBackingStoreWayland::create): When using WPE call wpe_fdo_initialize_for_egl_display() and
check we have the required extensions.
(WebKit::AcceleratedBackingStoreWayland::AcceleratedBackingStoreWayland): When using WPE initialize the view backend.
(WebKit::AcceleratedBackingStoreWayland::~AcceleratedBackingStoreWayland): When using WPE release the EGL images
still in use and the view texture.
(WebKit::AcceleratedBackingStoreWayland::renderHostFileDescriptor): Call wpe_view_backend_get_renderer_host_fd().
(WebKit::AcceleratedBackingStoreWayland::displayBuffer): Create the view texture if needed, save the pending
image and schedule a redraw.
(WebKit::AcceleratedBackingStoreWayland::paint): When using WPE commit the pending image and use it as target of
the view texture.

• UIProcess/gtk/AcceleratedBackingStoreWayland.h:
• UIProcess/gtk/HardwareAccelerationManager.cpp:

(WebKit::HardwareAccelerationManager::HardwareAccelerationManager): When using WPE disable hardware acceleration
if wpe_fdo_initialize_for_egl_display() fails.

• UIProcess/gtk/WaylandCompositor.cpp:
• UIProcess/gtk/WaylandCompositor.h:
• WebProcess/WebPage/AcceleratedSurface.cpp:

(WebKit::AcceleratedSurface::create): Create an AcceleratedSurfaceLibWPE for GTK port when using WPE and shared
display is Wayland.

• WebProcess/WebPage/WebPage.cpp:
• WebProcess/WebPage/WebPage.h:
• WebProcess/WebPage/libwpe/AcceleratedSurfaceLibWPE.cpp:

(WebKit::AcceleratedSurfaceLibWPE::initialize): Use the shared display for compositing instead of shared display.

• WebProcess/WebPage/libwpe/AcceleratedSurfaceLibWPE.h:
• WebProcess/WebProcess.h:
• WebProcess/glib/WebProcessGLib.cpp:

(WebKit::WebProcess::platformInitializeWebProcess): Initialize the wpe display when using WPE.

Tools:

Add WPEBackend-fdo to jhbuild moduleset.

• gtk/jhbuild.modules:

9:06 PM Changeset in webkit [245806] by bshafiei@apple.com

• 7 edits in branches/safari-608.1.24.20-branch

Revert r245775. rdar://problem/35205373

9:06 PM Changeset in webkit [245805] by bshafiei@apple.com

• 13 edits in branches/safari-608.1.24.20-branch

Revert r245637. rdar://problem/35205373

6:36 PM Changeset in webkit [245804] by bshafiei@apple.com

• 8 edits in branches/safari-608.1.24.20-branch

Cherry-pick r245803. rdar://problem/51145977

[iOS] Dropping in an editable element should result in a ranged selection
​https://bugs.webkit.org/show_bug.cgi?id=198267
<rdar://problem/51145977>

Reviewed by Tim Horton.

Source/WebKit:

When drag and drop was first implemented for iOS in iOS 11, selection behavior when dropping into editable
elements matched that of macOS, by leaving the inserted content selected after performing the drop. However, in
other parts of the platform (e.g. Notes), both the keyboard and selection views are not shown after a drop.

Instead of matching macOS behavior, WebKit on iOS should match the rest of the platform. This is a little
tricky, since we use the selection range after a drop to create a TextIndicator snapshot when creating a drag
preview. To resolve this, we refactor some of the logic introduced in r245778 to remember the DOM range to
snapshot before collapsing the range to the end of the inserted content.

Tested by adjusting some existing API tests.

• UIProcess/ios/WKContentViewInteraction.mm: (-[WKContentView _elementDidFocus:userIsInteracting:blurPreviousNode:activityStateChanges:userObject:]):

Remove some logic that currently presents the keyboard while the user is performing a drop that focuses an
editable element.

• WebProcess/WebPage/WebPage.h:

Add a member variable to keep track of which range should be snapshotted when generating a drop preview.

• WebProcess/WebPage/ios/WebPageIOS.mm: (WebKit::WebPage::didConcludeDrop): (WebKit::WebPage::didConcludeEditDrag):

Collapse the selection range to the end after an edit drag (i.e., a drop in an editable area that inserted
content).

(WebKit::WebPage::computeAndSendEditDragSnapshot):

Tools:

Adjust some existing API tests that currently check for selection rects after a drop. Instead of checking for
visible selection rects, simply check for the start caret rect, as determined by WKContentView's
-selectionRange.

• TestWebKitAPI/Tests/ios/DragAndDropTestsIOS.mm: (TestWebKitAPI::TEST): (makeCGRectValue): Deleted. (checkSelectionRectsWithLogging): Deleted.
• TestWebKitAPI/cocoa/DragAndDropSimulator.h:

Replace finalSelectionRects with finalSelectionStartRect.

• TestWebKitAPI/ios/DragAndDropSimulatorIOS.mm: (-[DragAndDropSimulator _resetSimulatedState]): (-[DragAndDropSimulator runFrom:to:additionalItemRequestLocations:]): (-[DragAndDropSimulator finalSelectionRects]): Deleted.

git-svn-id: ​https://svn.webkit.org/repository/webkit/trunk@245803 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6:03 PM Changeset in webkit [245803] by Wenson Hsieh

• 8 edits in trunk

[iOS] Dropping in an editable element should result in a ranged selection
​https://bugs.webkit.org/show_bug.cgi?id=198267
<rdar://problem/51145977>

Reviewed by Tim Horton.

Source/WebKit:

When drag and drop was first implemented for iOS in iOS 11, selection behavior when dropping into editable
elements matched that of macOS, by leaving the inserted content selected after performing the drop. However, in
other parts of the platform (e.g. Notes), both the keyboard and selection views are not shown after a drop.

Instead of matching macOS behavior, WebKit on iOS should match the rest of the platform. This is a little
tricky, since we use the selection range after a drop to create a TextIndicator snapshot when creating a drag
preview. To resolve this, we refactor some of the logic introduced in r245778 to remember the DOM range to
snapshot before collapsing the range to the end of the inserted content.

Tested by adjusting some existing API tests.

• UIProcess/ios/WKContentViewInteraction.mm:

(-[WKContentView _elementDidFocus:userIsInteracting:blurPreviousNode:activityStateChanges:userObject:]):

Remove some logic that currently presents the keyboard while the user is performing a drop that focuses an
editable element.

• WebProcess/WebPage/WebPage.h:

Add a member variable to keep track of which range should be snapshotted when generating a drop preview.

• WebProcess/WebPage/ios/WebPageIOS.mm:

(WebKit::WebPage::didConcludeDrop):
(WebKit::WebPage::didConcludeEditDrag):

Collapse the selection range to the end after an edit drag (i.e., a drop in an editable area that inserted
content).

(WebKit::WebPage::computeAndSendEditDragSnapshot):

Tools:

Adjust some existing API tests that currently check for selection rects after a drop. Instead of checking for
visible selection rects, simply check for the start caret rect, as determined by WKContentView's
-selectionRange.

• TestWebKitAPI/Tests/ios/DragAndDropTestsIOS.mm:

(TestWebKitAPI::TEST):
(makeCGRectValue): Deleted.
(checkSelectionRectsWithLogging): Deleted.

• TestWebKitAPI/cocoa/DragAndDropSimulator.h:

Replace finalSelectionRects with finalSelectionStartRect.

• TestWebKitAPI/ios/DragAndDropSimulatorIOS.mm:

(-[DragAndDropSimulator _resetSimulatedState]):
(-[DragAndDropSimulator runFrom:to:additionalItemRequestLocations:]):
(-[DragAndDropSimulator finalSelectionRects]): Deleted.

5:10 PM Changeset in webkit [245802] by commit-queue@webkit.org

• 6 edits in trunk

[CURL] Fix crashing SocketStreamHandle.
​https://bugs.webkit.org/show_bug.cgi?id=197873

Patch by Takashi Komori <​Takashi.Komori@sony.com> on 2019-05-27
Reviewed by Fujii Hironori.

Source/WebCore:

When NetworkSocketStream was destructed SocketStreamHandleImple::platformClose was called wrongly times.
This is because closed state is not set.

Test: http/tests/websocket/tests/hybi/workers/close.html

• platform/network/curl/SocketStreamHandleImpl.h:
• platform/network/curl/SocketStreamHandleImplCurl.cpp:

(WebCore::SocketStreamHandleImpl::platformSendInternal):
(WebCore::SocketStreamHandleImpl::platformClose):
(WebCore::SocketStreamHandleImpl::threadEntryPoint):
(WebCore::SocketStreamHandleImpl::handleError):
(WebCore::SocketStreamHandleImpl::callOnWorkerThread):
(WebCore::SocketStreamHandleImpl::executeTasks):

LayoutTests:

• platform/wincairo-wk1/TestExpectations:
• platform/wincairo/TestExpectations:

3:00 PM Changeset in webkit [245801] by Chris Dumez

• 2 edits in trunk/Source/WebKit

Unreviewed Win Cairo build fix after r245796.

• UIProcess/WebsiteData/WebsiteDataStore.cpp:

(WebKit::WebsiteDataStore::getAllStorageAccessEntries):

2:51 PM Changeset in webkit [245800] by Tadeu Zagallo

• 3 edits in trunk/Source/JavaScriptCore

Fix opensource build of testapi
​https://bugs.webkit.org/show_bug.cgi?id=198256

Reviewed by Alexey Proskuryakov.

In r245564, we added custom entitlements to testapi to allow caching
bytecode in data vaults, but we should only use the entitlements for
internal builds. Otherwises, testapi gets killed on launch. Also fix the
formatting for the errors added in the same patch, according to comments
in the bug after the patch had already landed.

• API/JSScript.mm:

(validateBytecodeCachePath):

• Configurations/ToolExecutable.xcconfig:

2:50 PM Changeset in webkit [245799] by Chris Dumez

• 2 edits in trunk/Source/WebKit

Unreviewed Win Cairo build fix after r245796.

• Shared/WebPageCreationParameters.cpp:

(WebKit::WebPageCreationParameters::decode):

2:27 PM Changeset in webkit [245798] by Oriol Brufau

• 19 edits
  4 adds in trunk

[css-grid] Preserve repeat() notation when serializing declared values
​https://bugs.webkit.org/show_bug.cgi?id=197840

Reviewed by Manuel Rego Casasnovas.

LayoutTests/imported/w3c:

Import WPT test.

• web-platform-tests/css/css-grid/grid-definition/grid-support-repeat-002-expected.txt: Added.
• web-platform-tests/css/css-grid/grid-definition/grid-support-repeat-002.html: Added.
• web-platform-tests/css/css-grid/grid-definition/w3c-import.log:

Source/WebCore:

Tests: fast/css-grid-layout/grid-element-auto-repeat-get-set.html

fast/css-grid-layout/grid-repeat-calc.html
fast/css-grid-layout/named-grid-line-get-set.html
imported/w3c/web-platform-tests/css/css-grid/grid-definition/grid-support-repeat-002.html

Before this change, a repeat() notation with an integral number of
repetitions was expanded at parse time. This was observable when reading
declared values using JS APIs.

This patch makes the parser preserve that notation, like it was already
happening when the number of repetitions was automatic and not integral.

The resolved value in getComputedStyle() will still be expanded, though,
as required by the spec.

• Sources.txt:
• WebCore.xcodeproj/project.pbxproj:
• css/CSSGridIntegerRepeatValue.cpp: Added.

(WebCore::CSSGridIntegerRepeatValue::customCSSText const):
(WebCore::CSSGridIntegerRepeatValue::equals const):

• css/CSSGridIntegerRepeatValue.h: Added.
• css/CSSValue.cpp:

(WebCore::CSSValue::equals const):
(WebCore::CSSValue::cssText const):
(WebCore::CSSValue::destroy):

• css/CSSValue.h:

(WebCore::CSSValue::isGridIntegerRepeatValue const):

• css/StyleBuilderConverter.h:

(WebCore::StyleBuilderConverter::createGridTrackList):
(WebCore::StyleBuilderConverter::convertGridTrackSizeList):

• css/parser/CSSPropertyParser.cpp:

(WebCore::consumeGridTrackRepeatFunction):

Tools:

Adapt CSSPropertyParserTest to CSSGridIntegerRepeatValue.

• TestWebKitAPI/Tests/WebCore/CSSParser.cpp:

(TestWebKitAPI::computeNumberOfTracks):

LayoutTests:

Update test expectations.

• fast/css-grid-layout/grid-element-auto-repeat-get-set-expected.txt:
• fast/css-grid-layout/grid-element-auto-repeat-get-set.html:
• fast/css-grid-layout/grid-repeat-calc-expected.txt:
• fast/css-grid-layout/grid-repeat-calc.html:
• fast/css-grid-layout/named-grid-line-get-set-expected.txt:
• fast/css-grid-layout/named-grid-line-get-set.html:

2:21 PM Changeset in webkit [245797] by bshafiei@apple.com

• 7 edits in branches/safari-608.1.24.20-branch/Source

Versioning.

1:30 PM Changeset in webkit [245796] by Chris Dumez

• 185 edits
  1 copy in trunk/Source

Use a strongly-typed identifier for pages
​https://bugs.webkit.org/show_bug.cgi?id=198206

Reviewed by Youenn Fablet.

Source/WebCore:

• WebCore/*:

Source/WebKit:

• WebKit/*:

Source/WebKitLegacy:

• WebCoreSupport/WebResourceLoadScheduler.cpp:

(WebResourceLoadScheduler::pageLoadCompleted):

• WebCoreSupport/WebResourceLoadScheduler.h:

Source/WebKitLegacy/mac:

• WebCoreSupport/WebFrameLoaderClient.h:
• WebCoreSupport/WebFrameLoaderClient.mm:

(WebFrameLoaderClient::pageID const):

Source/WTF:

• wtf/ObjectIdentifier.h:

(WTF::operator<<):

9:19 AM Changeset in webkit [245795] by commit-queue@webkit.org

• 2 edits in trunk/Source/WebKit

[Pointer Events] Test at pointerevents/ios/touch-action-none-in-overflow-scrolling-touch.html times out
​https://bugs.webkit.org/show_bug.cgi?id=198275
<rdar://problem/50618496>

Patch by Antoine Quint <Antoine Quint> on 2019-05-27
Reviewed by Antti Koivisto.

Only WKCompositingView instances may have associated touch-action data, so we should filter out any hit-testing view
that isn't a WKCompositingView when finding the view that should hit-test when looking for touch-action information.

• UIProcess/RemoteLayerTree/ios/RemoteLayerTreeViews.mm:

(WebKit::touchActionsForPoint):

9:06 AM Changeset in webkit [245794] by Simon Fraser

• 6 edits in trunk/Source/WebCore

Move GraphicsLayerCA::adjustTiledLayerVisibleRect() to TileController
​https://bugs.webkit.org/show_bug.cgi?id=198266

Reviewed by Zalan Bujtas.

GraphicsLayerCA::adjustTiledLayerVisibleRect() was computing tile coverage for a
TiledBacking, just like TileController::adjustTileCoverageRect(), so move the code
into TileController as a first step to unifying more of this code. It's currently
used for tiled compositing layers, including overflow:scroll, and tiled layers
whose coverage may be affected by animations, so it's general-purpose.

TileController::adjustTileCoverageRect() is used for scrollable things, so rename
it to adjustTileCoverageRectForScrolling().

No behavior change.

• platform/graphics/TiledBacking.h:
• platform/graphics/ca/GraphicsLayerCA.cpp:

(WebCore::GraphicsLayerCA::adjustCoverageRect const):
(WebCore::GraphicsLayerCA::adjustTiledLayerVisibleRect): Deleted.

• platform/graphics/ca/GraphicsLayerCA.h:
• platform/graphics/ca/TileController.cpp:

(WebCore::TileController::adjustTileCoverageRect const):
(WebCore::TileController::adjustTileCoverageRectForScrolling const):

• platform/graphics/ca/TileController.h:

9:00 AM Changeset in webkit [245793] by Simon Fraser

• 2 edits in trunk/Source/WebCore

Fix Apple Internal builds after r245788.

• platform/PlatformScreen.h:

(WebCore::screenHasTouchDevice):
(WebCore::screenIsTouchPrimaryInputDevice):

7:15 AM Changeset in webkit [245792] by Alan Bujtas

• 2 edits in trunk/Source/WebCore

[LFC][IFC] Ignore collapsed runs when setting the width on the associated display boxes.
​https://bugs.webkit.org/show_bug.cgi?id=198260
<rdar://problem/51145704>

Reviewed by Antti Koivisto.

Collapsed runs don't contribute to the logical width of their containers.

Covered by existing tests.

• layout/inlineformatting/InlineFormattingContextLineLayout.cpp:

(WebCore::Layout::InlineFormattingContext::LineLayout::closeLine const):

2:55 AM Changeset in webkit [245791] by Carlos Garcia Campos

• 1 copy in releases/WebKitGTK/webkit-2.25.1

WebKitGTK 2.25.1

2:54 AM Changeset in webkit [245790] by Carlos Garcia Campos

• 4 edits in trunk

Unreviewed. Update OptionsGTK.cmake and NEWS for 2.25.1 release

.:

• Source/cmake/OptionsGTK.cmake: Bump version numbers

Source/WebKit:

• gtk/NEWS: Add release notes for 2.25.1.

2:52 AM Changeset in webkit [245789] by Carlos Garcia Campos

• 5 edits in trunk/Source

Unreviewed. Fix GTK distcheck

Source/WebCore:

Move mac headers to platform specific makefile.

• Headers.cmake:
• PlatformMac.cmake:

Source/WebKit:

Remove symbols that are not longer present in the library.

• webkitglib-symbols.map:

1:34 AM Changeset in webkit [245788] by Carlos Garcia Campos

• 11 edits in trunk

Touch support is reported even when the device doesn't have a touch screen
​https://bugs.webkit.org/show_bug.cgi?id=139681

Reviewed by Michael Catanzaro.

Source/WebCore:

Add screenHasTouchDevice() and screenIsTouchPrimaryInputDevice() to PlatformScreen and use it to decide whether
to expose touch events functionality or not.

• bindings/js/WebCoreBuiltinNames.h:
• css/MediaQueryEvaluator.cpp:

(WebCore::hoverEvaluate): Use screenIsTouchPrimaryInputDevice() when touch events are enabled at build time.
(WebCore::anyHoverEvaluate): Use screenHasTouchDevice() when touch events are enabled at build time.
(WebCore::pointerEvaluate): Use screenIsTouchPrimaryInputDevice() when touch events are enabled at build time.
(WebCore::anyPointerEvaluate): Use screenHasTouchDevice() when touch events are enabled at build time.

• dom/GlobalEventHandlers.idl: Make touch event attributes enabled at runtime.
• page/RuntimeEnabledFeatures.cpp:

(WebCore::RuntimeEnabledFeatures::touchEventsEnabled const): Return whether touch events are enabled.

• page/RuntimeEnabledFeatures.h:

(WebCore::RuntimeEnabledFeatures::setTouchEventsEnabled): Force touch events to be enabled or disabled.

• platform/PlatformScreen.h:

(WebCore::screenHasTouchDevice): Just return true when touch events are enabled at build time for non-gtk ports.
(WebCore::screenIsTouchPrimaryInputDevice): Ditto.

• platform/gtk/PlatformScreenGtk.cpp:

(WebCore::screenHasTouchDevice):
(WebCore::screenIsTouchPrimaryInputDevice):
(WebCore::isTouchDevice):
(WebCore::deviceAddedCallback):
(WebCore::deviceRemovedCallback):

LayoutTests:

Remove expectations for tests that are passing now.

• platform/gtk/TestExpectations: