[webkit-changes] [WebKit/WebKit] 767aeb: Cherry-pick 255673 at main (14bba72947fd). https://bu...
Adrian Perez
noreply at github.com
Wed Jan 25 02:20:19 PST 2023
Branch: refs/heads/webkitglib/2.38
Home: https://github.com/WebKit/WebKit
Commit: 767aeb7c306f1f485afaf7491852c37904857ac2
https://github.com/WebKit/WebKit/commit/767aeb7c306f1f485afaf7491852c37904857ac2
Author: Xabier Rodriguez-Calvar <calvaris at igalia.com>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
M Source/WebCore/platform/graphics/gstreamer/eme/WebKitCommonEncryptionDecryptorGStreamer.cpp
Log Message:
-----------
Cherry-pick 255673 at main (14bba72947fd). https://bugs.webkit.org/show_bug.cgi?id=246682
[GStreamer][EME] Prevent race conditions when flushing
https://bugs.webkit.org/show_bug.cgi?id=246682
Reviewed by Philippe Normand.
In some cases, decryption can be slower than flush processing and we might be setting and unsetting the flushing flag
without the decryption thread noticing. We add a small state machine to ensure that if we get a flush while decrypting,
we are going to ditch that buffer even if the result was correctly decrypted.
* Source/WebCore/platform/graphics/gstreamer/eme/WebKitCommonEncryptionDecryptorGStreamer.cpp:
(transformInPlace):
(sinkEventHandler):
Canonical link: https://commits.webkit.org/255673@main
Commit: c6781fcb5a125801e198e2b4631006b0859fd9ec
https://github.com/WebKit/WebKit/commit/c6781fcb5a125801e198e2b4631006b0859fd9ec
Author: Kimmo Kinnunen <kkinnunen at apple.com>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
M Source/WTF/wtf/text/StringConcatenate.h
M Source/WTF/wtf/text/StringView.cpp
M Tools/TestWebKitAPI/CMakeLists.txt
M Tools/TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj
M Tools/TestWebKitAPI/Tests/WTF/StringOperators.cpp
R Tools/TestWebKitAPI/WTFStringUtilities.cpp
M Tools/TestWebKitAPI/WTFStringUtilities.h
Log Message:
-----------
Cherry-pick 256131 at main (d4f69df7d5b1). https://bugs.webkit.org/show_bug.cgi?id=247121
StringOperations.cpp test re-defines WTF::StringTypeAdapter implementation, undefined behaviour
https://bugs.webkit.org/show_bug.cgi?id=247121
rdar://problem/101669240
Reviewed by Antti Koivisto.
In C++, one definition rule states that in one program each definition must consists out of
same tokens in all compilation units in the program.
https://en.cppreference.com/w/cpp/language/definition
...
There can be more than one definition in a program of each of the following: class type,
enumeration type, inline function, inline variable (since C++17), templated entity
(template or member of template, but not full template specialization), as long as all
of the following is true:
...
- each definition consists of the same sequence of tokens (typically, appears in the same header file)
...
if the definition is for a template, then all these requirements apply to both names at the point of
definition and dependent names at the point of instantiation
If all these requirements are satisfied, the program behaves as if there is only one definition in the
entire program. Otherwise, the program is ill-formed, no diagnostic required.
In other words: you cannot have two implementations of same template member function, one 100 bytes and
one 1000mbs in size and select between them based on an #define before #include in different .cpp files
in the same program.
The tests would alter the StringView operator+ used templates via custom define before include.
This results in undefined behavior, since libWTF.a would contain one implementation and use
of the templates, while the StringOperators.cpp test would intend to contain a different
implementation. Also, any other test in TestWTF would use the same version as StringOperators.cpp,
due to how the includes are currently written.
Using the hack only for StringOperators.cpp would preserve the UB but be the minimalistic change.
However, this does not work:
The requirement that WTFStringUtilities.h needs to be included before StringView.h
is messing up the include order in tests. When trying to fix the include order for all
other files except the StringOperators.cpp, the omission of the hack include will cause new, non-hacked
template definition for the same templates that are hacked for StringOperators.cpp. This in turn
causes non-deterministic selection of the template: the non-hacked, correct template is selected
for all instantiations, including the ones in StringOperators.cpp. This causes the test to fail, as
the hack is not present.
This means that this sort of UB hack cannot be used for StringOperators.cpp while using the correct
templates for other instantiations in the binary. This means that almost all WTFString-using tests
must use WTFStringUtilities.h to get the hack. This has the problem:
- UB
- Messed up includes
Fix this properly by not using the hack, rather putting the debug logic into the Debug build and
disabling the test for non-Debug builds.
Use atomic<int> to count the copies so that it doesn't show up in TSAN or similar tools when running
normally for other purposes. The test itself of course does not need the atomic.
This is work that is prerequisite to fix tests to obey normal WebKit include rules.
* Source/WTF/wtf/text/StringConcatenate.h:
* Source/WTF/wtf/text/StringView.cpp:
* Tools/TestWebKitAPI/CMakeLists.txt:
* Tools/TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* Tools/TestWebKitAPI/Tests/WTF/StringOperators.cpp:
(TestWebKitAPI::TEST):
* Tools/TestWebKitAPI/WTFStringUtilities.cpp: Removed.
* Tools/TestWebKitAPI/WTFStringUtilities.h:
Canonical link: https://commits.webkit.org/256131@main
Commit: 9e2603ce4f1f8ffc16413be48a2b784afe04c0fa
https://github.com/WebKit/WebKit/commit/9e2603ce4f1f8ffc16413be48a2b784afe04c0fa
Author: Lauro Moura <lmoura at igalia.com>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
M Tools/Scripts/webkitpy/webdriver_tests/webdriver_driver_wpe.py
Log Message:
-----------
Cherry-pick 255995 at main (ea3511fd9cd1). https://bugs.webkit.org/show_bug.cgi?id=247010
[WPE][WebDriver] Fix Cog path in capabilities after 252411 at main
https://bugs.webkit.org/show_bug.cgi?id=247010
Reviewed by Michael Catanzaro.
* Tools/Scripts/webkitpy/webdriver_tests/webdriver_driver_wpe.py:
(WebDriverWPE.browser_path): Adjust path
Canonical link: https://commits.webkit.org/255995@main
Commit: bb61f1ca939c3b7cb2158649abcbf3bfdd0687a8
https://github.com/WebKit/WebKit/commit/bb61f1ca939c3b7cb2158649abcbf3bfdd0687a8
Author: Chris Dumez <cdumez at apple.com>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
M LayoutTests/http/tests/navigation/resources/shared-worker-script.js
A LayoutTests/http/tests/workers/shared/shared-worker-with-service-workers-disabled-expected.txt
A LayoutTests/http/tests/workers/shared/shared-worker-with-service-workers-disabled.html
M LayoutTests/platform/mac-wk1/TestExpectations
M LayoutTests/platform/win/TestExpectations
M Source/WebCore/workers/WorkerScriptLoader.cpp
M Source/WebCore/workers/WorkerScriptLoader.h
Log Message:
-----------
Cherry-pick 256185 at main (f6816a3f3c66). https://bugs.webkit.org/show_bug.cgi?id=247147
Regression(iOS 16.1) Use of Shared Workers cause Web Content process crash in non-default-browser WKWebViews
https://bugs.webkit.org/show_bug.cgi?id=247147
rdar://101562445
Reviewed by Geoffrey Garen.
253592 at main made the initialization of WorkerScriptLoader::m_clientIdentifier
conditional on service workers being enabled. However, this wasn't correct
since shared workers rely on this data member too.
This clientIdentifier gets sent over IPC at some point and fails decoding
when not initialized, leading to the WebProcess getting killed.
* LayoutTests/http/tests/navigation/resources/shared-worker-script.js:
(self.onconnect):
* LayoutTests/http/tests/workers/shared/shared-worker-with-service-workers-disabled-expected.txt: Added.
* LayoutTests/http/tests/workers/shared/shared-worker-with-service-workers-disabled.html: Added.
* LayoutTests/platform/mac-wk1/TestExpectations:
* LayoutTests/platform/win/TestExpectations:
* Source/WebCore/workers/WorkerScriptLoader.cpp:
(WebCore::WorkerScriptLoader::~WorkerScriptLoader):
(WebCore::WorkerScriptLoader::loadAsynchronously):
* Source/WebCore/workers/WorkerScriptLoader.h:
Canonical link: https://commits.webkit.org/256185@main
Commit: 0ca760d5517b311e482f802525081b682c71ccdc
https://github.com/WebKit/WebKit/commit/0ca760d5517b311e482f802525081b682c71ccdc
Author: Devin Rousso <drousso at apple.com>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
M Source/WebCore/Modules/modern-media-controls/controls/adwaita-layout-traits.js
M Source/WebCore/Modules/modern-media-controls/controls/ios-inline-media-controls.css
M Source/WebCore/Modules/modern-media-controls/controls/ios-layout-traits.js
M Source/WebCore/Modules/modern-media-controls/controls/layout-traits.js
M Source/WebCore/Modules/modern-media-controls/controls/macos-layout-traits.js
M Source/WebCore/Modules/modern-media-controls/controls/media-controls.css
M Source/WebCore/Modules/modern-media-controls/controls/media-controls.js
M Source/WebCore/Modules/modern-media-controls/controls/watchos-layout-traits.js
Log Message:
-----------
Cherry-pick 256167 at main (2476493a0ed7). https://bugs.webkit.org/show_bug.cgi?id=247152
[Modern Media Controls] [macOS] REGRESSION(252362 at main): adding `border-radius: inherit` causes negative power perf impact
https://bugs.webkit.org/show_bug.cgi?id=247152
<rdar://problem/100234334>
Reviewed by Eric Carlson.
This is only used on iOS, so adding it to all platforms is unnecessary (and apparently costly).
* Source/WebCore/Modules/modern-media-controls/controls/layout-traits.js:
(LayoutTraits.prototype.inheritsBorderRadius): Added.
* Source/WebCore/Modules/modern-media-controls/controls/adwaita-layout-traits.js:
(AdwaitaLayoutTraits.prototype.inheritsBorderRadius): Added.
* Source/WebCore/Modules/modern-media-controls/controls/ios-layout-traits.js:
(IOSLayoutTraits.prototype.inheritsBorderRadius): Added.
* Source/WebCore/Modules/modern-media-controls/controls/macos-layout-traits.js:
(MacOSLayoutTraits.prototype.inheritsBorderRadius): Added.
* Source/WebCore/Modules/modern-media-controls/controls/watchos-layout-traits.js:
(WatchOSLayoutTraits.prototype.inheritsBorderRadius): Added.
* Source/WebCore/Modules/modern-media-controls/controls/media-controls.js:
(MediaControls):
Add a new `LayoutTraits` method that controls whether `border-radius: inherit` is used.
* Source/WebCore/Modules/modern-media-controls/controls/ios-inline-media-controls.css:
(.media-controls.inline.ios:not(.audio)):
* Source/WebCore/Modules/modern-media-controls/controls/media-controls.css:
(.media-controls-container):
(.media-controls-container:has(.media-controls.inherits-border-radius), .media-controls.inherits-border-radius): Added.
Move the `border-radius: inherit` to only if the CSS class is applied (which is controlled by the above).
Canonical link: https://commits.webkit.org/256167@main
Commit: a6ffc2b0df129c5df2a51af4d10d3ab6496b1631
https://github.com/WebKit/WebKit/commit/a6ffc2b0df129c5df2a51af4d10d3ab6496b1631
Author: Ahmad Saleem <ahmad.saleem792+github at gmail.com>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
A LayoutTests/editing/selection/iframe-select-expected.html
A LayoutTests/editing/selection/iframe-select.html
M Source/WebCore/rendering/RenderElement.h
M Source/WebCore/rendering/RenderObject.h
M Source/WebCore/rendering/RenderWidget.cpp
Log Message:
-----------
Cherry-pick 256191 at main (fe7889ca7da9). https://bugs.webkit.org/show_bug.cgi?id=246939
Use the correct paint methods for painting a section rect of RenderWidgets
Use the correct paint methods for painting a section rect of RenderWidgets
https://bugs.webkit.org/show_bug.cgi?id=246939
Reviewed by Simon Fraser.
Merge - https://src.chromium.org/viewvc/blink?revision=181068&view=revision
Previously, it was using a paint invalidation method. That is less
efficient, and also wrong if the renderer is a squashed layer.
* Source/WebCore/rendering/RenderElement.h: Update "paint" function to to have "paintOffset" and also add comment for explanation
* Source/WebCore/rendering/RenderObject.h: Update Comment and remove "selectionRect" function
* Source/WebCore/rendering/RenderWidget.cpp: Fix paint logic for selected widgets and remove "FIXME" about selectionRect
* LayoutTests/editing/selection/iframe-select.html: Added Test Case
* LayoutTests/editing/selection/iframe-select-expected.html: Added Test Case Expectations
Canonical link: https://commits.webkit.org/256191@main
Commit: d3c3c472eaee784ee1b406dfd2a2e4324d90e6ac
https://github.com/WebKit/WebKit/commit/d3c3c472eaee784ee1b406dfd2a2e4324d90e6ac
Author: Simon Fraser <simon.fraser at apple.com>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
M Source/WebKit/WebProcess/WebPage/EventDispatcher.cpp
Log Message:
-----------
Cherry-pick 256199 at main (5172b1f269aa). https://bugs.webkit.org/show_bug.cgi?id=247302
Crash trying to scroll a main frame PDF with UI-side compositing on macOS.
https://bugs.webkit.org/show_bug.cgi?id=247302
rdar://101786249
Reviewed by Tim Horton.
Null check the scrolling tree before stuffing it into a map (with UI-side compositing
and PDF it can be null).
* Source/WebKit/WebProcess/WebPage/EventDispatcher.cpp:
(WebKit::EventDispatcher::addScrollingTreeForPage):
Canonical link: https://commits.webkit.org/256199@main
Commit: f8c39b6bb4141fa6f77a8e6325d20f5f952b0fc9
https://github.com/WebKit/WebKit/commit/f8c39b6bb4141fa6f77a8e6325d20f5f952b0fc9
Author: Enrique Ocaña González <eocanha at igalia.com>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
M Source/WebCore/platform/graphics/gstreamer/GStreamerRegistryScanner.cpp
M Source/WebCore/platform/graphics/gstreamer/GStreamerRegistryScanner.h
M Source/WebCore/platform/graphics/gstreamer/mse/AppendPipeline.cpp
Log Message:
-----------
Cherry-pick 256222 at main (b84521e66ded). https://bugs.webkit.org/show_bug.cgi?id=247129
[MSE][GStreamer] Relax h264/h265 codec check for MSE
https://bugs.webkit.org/show_bug.cgi?id=247129
Reviewed by Alicia Boya Garcia.
Some platforms disable/remove software parsers because of the way
encrypted playback is implemented, but we still want working MSE
playback there.
Although software parsers are strongly recommended for MSE playback
(eg: to complete info from buffers coming from some js transmuxers),
they aren't strictly needed (at your own risk!).
However, some video decoders (eg: the OMX h264 decoder) can only
accept some of the available formats (only stream-format=byte-stream
in the case of OMX). In that case, a parser is still needed in order
to convert between formats before supplying buffers to the decoder.
This patch checks that the most efficient decoder (hardware if
available, software if not) supports both formats. If the most
efficient decoder doesn't support all the formats, a parser is
demanded to be available. If it's not available, the affected codec
(h264, h265) is declared as not supported.
Inspired by work by Eugene Mutavchi <Ievgen_Mutavchi at comcast.com>
See: https://github.com/WebPlatformForEmbedded/WPEWebKit/pull/948
* Source/WebCore/platform/graphics/gstreamer/GStreamerRegistryScanner.cpp:
(WebCore::GStreamerRegistryScanner::initializeDecoders): Ask for the corresponding parser if any of the h264 or h265 decoders support don't support any of the possible stream-format variants. If the parser isn't present, declare the corresponding codec as unsupported.
* Source/WebCore/platform/graphics/gstreamer/GStreamerRegistryScanner.h:
(WebCore::GStreamerRegistryScanner::RegistryLookupResult::merge): Added method to merge two results, supported when both are supported, and using hardware when both are using hardware.
(WebCore::GStreamerRegistryScanner::RegistryLookupResult::operator==): Added equality operator.
(WebCore::GStreamerRegistryScanner::RegistryLookupResult::operator!=): Added inequality operator.
* Source/WebCore/platform/graphics/gstreamer/mse/AppendPipeline.cpp:
(WebCore::createOptionalParserForFormat): Warn when a parser can't be created.
Canonical link: https://commits.webkit.org/256222@main
Commit: 433de2de47e6190ce64bb6707ee2533756a65a1c
https://github.com/WebKit/WebKit/commit/433de2de47e6190ce64bb6707ee2533756a65a1c
Author: Przemyslaw Gorszkowski <pgorszkowski at igalia.com>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
M Source/WebKit/UIProcess/WebFrameProxy.cpp
Log Message:
-----------
Cherry-pick 256233 at main (628f0ba433d8). https://bugs.webkit.org/show_bug.cgi?id=247362
Missing undef MESSAGE_CHECK in WebFrameProxy.cpp
https://bugs.webkit.org/show_bug.cgi?id=247362
Reviewed by Carlos Garcia Campos.
Add missing #undef MESSAGE_CHECK at the end of the cpp file
solves the problem with errors in unified builds.
* Source/WebKit/UIProcess/WebFrameProxy.cpp:
Canonical link: https://commits.webkit.org/256233@main
Commit: d84c5310395d81d7033d59f5e6f8abeaba00a4cc
https://github.com/WebKit/WebKit/commit/d84c5310395d81d7033d59f5e6f8abeaba00a4cc
Author: Ahmad Saleem <ahmad.saleem792+github at gmail.com>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
M Source/WebCore/rendering/RenderLayerCompositor.cpp
Log Message:
-----------
Cherry-pick 256247 at main (1ad126f47e2a). https://bugs.webkit.org/show_bug.cgi?id=247369
Typo error in RenderLayerCompositor::destroyRootLayer() for Horizontal Scrollbar
Typo error in RenderLayerCompositor::destroyRootLayer() for Horizontal Scrollbar
https://bugs.webkit.org/show_bug.cgi?id=247369
Reviewed by Alan Baradlay.
Merge - https://src.chromium.org/viewvc/blink?view=revision&revision=194036
This is to fix a copy paste error in "m_layerForHorizontalScrollbar", where the "verticalScrollbar" was mentioned instead of "horizontalScrollbar".
* Source/WebCore/rendering/RenderLayerCompositor.cpp:
(RenderLayerCompositor::destroyRootLayer): Fix reference to horizontalScrollbar"
Canonical link: https://commits.webkit.org/256247@main
Commit: c68692c2b5729a386d68166ce4d09aad9b6afddd
https://github.com/WebKit/WebKit/commit/c68692c2b5729a386d68166ce4d09aad9b6afddd
Author: Alex Christensen <achristensen at webkit.org>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
M Source/WTF/wtf/URLHelpers.cpp
M Tools/TestWebKitAPI/Tests/WTF/cocoa/URLExtras.mm
Log Message:
-----------
Cherry-pick 256267 at main (16904a9a85c8). https://bugs.webkit.org/show_bug.cgi?id=247289
Punycode all IPA extensions code points in URLs
https://bugs.webkit.org/show_bug.cgi?id=247289
rdar://101429376
Reviewed by Tim Horton.
* Source/WTF/wtf/URLHelpers.cpp:
(WTF::URLHelpers::isLookalikeCharacter):
* Tools/TestWebKitAPI/Tests/WTF/cocoa/URLExtras.mm:
(TestWebKitAPI::TEST):
Canonical link: https://commits.webkit.org/256267@main
Commit: 914f2dfb99284d7a47ea43529fad9099d31c4384
https://github.com/WebKit/WebKit/commit/914f2dfb99284d7a47ea43529fad9099d31c4384
Author: Alex Christensen <achristensen at webkit.org>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
M LayoutTests/fast/encoding/idn-security-expected.txt
M LayoutTests/fast/encoding/idn-security.html
M Source/WTF/wtf/URLHelpers.cpp
M Tools/TestWebKitAPI/Tests/WTF/cocoa/URLExtras.mm
Log Message:
-----------
Cherry-pick 256276 at main (a6c922ea6f05). https://bugs.webkit.org/show_bug.cgi?id=247287
Punycode U+0E01 when in the context of non-Thai characters
https://bugs.webkit.org/show_bug.cgi?id=247287
rdar://101434628
Reviewed by Tim Horton.
* Source/WTF/wtf/URLHelpers.cpp:
(WTF::URLHelpers::isLookalikeCharacterOfScriptType<USCRIPT_THAI>):
(WTF::URLHelpers::isLookalikeCharacter):
* Tools/TestWebKitAPI/Tests/WTF/cocoa/URLExtras.mm:
(TestWebKitAPI::TEST):
Canonical link: https://commits.webkit.org/256276@main
Commit: c13c77f6888eba9f8884cfc7274a27f1a1c9fb8a
https://github.com/WebKit/WebKit/commit/c13c77f6888eba9f8884cfc7274a27f1a1c9fb8a
Author: Matt Woodrow <mattwoodrow at apple.com>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
M Source/WebCore/rendering/RenderLayerBacking.cpp
M Source/WebCore/rendering/RenderLayerCompositor.cpp
Log Message:
-----------
Cherry-pick 256299 at main (8ed3d524525a). https://bugs.webkit.org/show_bug.cgi?id=247251
Use device pixel scaled backing store for <canvas> with image-rendering: pixelated.
https://bugs.webkit.org/show_bug.cgi?id=247251
<rdar://99895407>
Reviewed by Simon Fraser.
We need to do scaling during painting to apply the image-rendering effect. An alternative fix would
be to set the image-rendering value on the GraphicsLayer, but its complicated to make that change
dependent on having computed isUnscaledBitmapOnly.
* LayoutTests/platform/ios-wk2/TestExpectations:
* Source/WebCore/rendering/RenderLayerBacking.cpp:
(WebCore::RenderLayerBacking::isUnscaledBitmapOnly const):
* Source/WebCore/rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::layerStyleChanged):
Canonical link: https://commits.webkit.org/256299@main
Commit: f35376ae6d7372db9995775e99f732f0ab9c1317
https://github.com/WebKit/WebKit/commit/f35376ae6d7372db9995775e99f732f0ab9c1317
Author: Alex Christensen <achristensen at webkit.org>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
M Source/WTF/wtf/URLHelpers.cpp
M Tools/TestWebKitAPI/Tests/WTF/cocoa/URLExtras.mm
Log Message:
-----------
Cherry-pick 256332 at main (85e478aae63c). https://bugs.webkit.org/show_bug.cgi?id=247461
Punycode Arabic diacritics when not preceded by an Arabic code point in a URL host
https://bugs.webkit.org/show_bug.cgi?id=247461
rdar://100426863
Reviewed by Said Abou-Hallawa.
* Source/WTF/wtf/URLHelpers.cpp:
(WTF::URLHelpers::isArabicDiacritic):
(WTF::URLHelpers::isArabicCodePoint):
(WTF::URLHelpers::isLookalikeCharacter):
* Tools/TestWebKitAPI/Tests/WTF/cocoa/URLExtras.mm:
(TestWebKitAPI::TEST):
Canonical link: https://commits.webkit.org/256332@main
Commit: bfe9ffc9f4f9118704971424577700fbf61d20db
https://github.com/WebKit/WebKit/commit/bfe9ffc9f4f9118704971424577700fbf61d20db
Author: Richard Robinson <richard_robinson2 at apple.com>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
M Source/WebCore/dom/Document.cpp
M Tools/TestWebKitAPI/Tests/WebKitCocoa/SystemPreview.mm
M Tools/TestWebKitAPI/Tests/WebKitCocoa/system-preview-trigger.html
Log Message:
-----------
Cherry-pick 256462 at main (a203e214d6b8). https://bugs.webkit.org/show_bug.cgi?id=245959
REGRESSION (iOS 16): AR QuickLook banner tap event no longer propagated to anchor if not attached to DOM
https://bugs.webkit.org/show_bug.cgi?id=245959
rdar://100952298
Reviewed by Wenson Hsieh.
The commit https://commits.webkit.org/251042@main added a check `!element->isConnected()`
in `Document::dispatchSystemPreviewActionEvent` to ensure the element is part
of the DOM before sending the `_apple_ar_quicklook_button_tapped` event.
However, this broke behavior in the case of a detached element, and so this PR
reverts that check to maintain the existing behavior.
* Source/WebCore/dom/Document.cpp:
(WebCore::Document::dispatchSystemPreviewActionEvent):
* Tools/TestWebKitAPI/Tests/WebKitCocoa/SystemPreview.mm:
(-[TestSystemPreviewTriggeredHandler userContentController:didReceiveScriptMessage:]):
(-[TestSystemPreviewTriggeredOnDetachedElementHandler userContentController:didReceiveScriptMessage:]):
(TestWebKitAPI::TEST):
* Tools/TestWebKitAPI/Tests/WebKitCocoa/system-preview-trigger.html:
Canonical link: https://commits.webkit.org/256462@main
Commit: a7bf5b8286c96dfc1c01d13a010539f78e42e53e
https://github.com/WebKit/WebKit/commit/a7bf5b8286c96dfc1c01d13a010539f78e42e53e
Author: Jer Noble <jer.noble at apple.com>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
M Source/WebCore/html/HTMLMediaElement.cpp
Log Message:
-----------
Cherry-pick 256352 at main (e4f004bdfd57). https://bugs.webkit.org/show_bug.cgi?id=247333
CRASH: Assertion in WPT /webvtt/rendering/cues-with-video/processing-model/regions/viewportanchor_y_50_percent.html
https://bugs.webkit.org/show_bug.cgi?id=247333
rdar://100892574
Reviewed by Eric Carlson.
An existing check protects HTMLMediaElement::configureTextTrackDisplay() from making script-exposed changes while a
Document and it's ActiveDOMObjects has been stopped, but also needs to protect when those same objects are Suspended.
Re-use HTMLMediaElement::isSuspended() (which encompasses both those above checks) everywhere within HTMLMediaElement.
* Source/WebCore/html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::userCancelledLoad):
(WebCore::HTMLMediaElement::exitFullscreen):
(WebCore::HTMLMediaElement::configureTextTrackDisplay):
(WebCore::HTMLMediaElement::updateMediaControlsAfterPresentationModeChange):
Canonical link: https://commits.webkit.org/256352@main
Commit: 89ca1faff15da753964b82c81f5ebab7c1f77465
https://github.com/WebKit/WebKit/commit/89ca1faff15da753964b82c81f5ebab7c1f77465
Author: Darin Adler <darin at apple.com>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
M Source/WTF/wtf/Vector.h
M Tools/TestWebKitAPI/Tests/WTF/Vector.cpp
Log Message:
-----------
Cherry-pick 256373 at main (7ea0a5456640). https://bugs.webkit.org/show_bug.cgi?id=247533
map and compactMap functions sometimes copy the passed in collection before mapping it to a vector
https://bugs.webkit.org/show_bug.cgi?id=247533
rdar://problem/102000000
Reviewed by Ryosuke Niwa and Tim Horton.
* Source/WTF/wtf/Vector.h:
(WTF::Mapper::map): Take const SourceType& or SourceType&&, not SourceType.
(WTF::CompactMapper::compactMap): Ditto.
* Tools/TestWebKitAPI/Tests/WTF/Vector.cpp: Added a new test, WTF_Vector.MapMinimalCopy.
Canonical link: https://commits.webkit.org/256373@main
Commit: 40b8d82cbb584858a5315675f22e7a774bc12992
https://github.com/WebKit/WebKit/commit/40b8d82cbb584858a5315675f22e7a774bc12992
Author: Philippe Normand <philn at igalia.com>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
M Source/WebCore/platform/mediastream/gstreamer/RealtimeIncomingSourceGStreamer.cpp
M Source/WebCore/platform/mediastream/gstreamer/RealtimeIncomingSourceGStreamer.h
M Source/WebCore/platform/mediastream/gstreamer/RealtimeIncomingVideoSourceGStreamer.cpp
Log Message:
-----------
Cherry-pick 256393 at main (25fd4849d5a2). https://bugs.webkit.org/show_bug.cgi?id=247423
[GStreamer][WebRTC] Minor cleanups in IncomingSourceGStreamer
https://bugs.webkit.org/show_bug.cgi?id=247423
Reviewed by Xabier Rodriguez-Calvar.
There is no need for a queue between the valve and the tee. Utility methods for opening/closing now
un-needed were removed as well.
* Source/WebCore/platform/mediastream/gstreamer/RealtimeIncomingSourceGStreamer.cpp:
(WebCore::RealtimeIncomingSourceGStreamer::RealtimeIncomingSourceGStreamer):
(WebCore::RealtimeIncomingSourceGStreamer::startProducingData):
(WebCore::RealtimeIncomingSourceGStreamer::stopProducingData):
(WebCore::RealtimeIncomingSourceGStreamer::closeValve const): Deleted.
(WebCore::RealtimeIncomingSourceGStreamer::openValve const): Deleted.
* Source/WebCore/platform/mediastream/gstreamer/RealtimeIncomingSourceGStreamer.h:
* Source/WebCore/platform/mediastream/gstreamer/RealtimeIncomingVideoSourceGStreamer.cpp:
Canonical link: https://commits.webkit.org/256393@main
Commit: 38bf4801b044374fb4be4d6785495d02875f0171
https://github.com/WebKit/WebKit/commit/38bf4801b044374fb4be4d6785495d02875f0171
Author: Philippe Normand <philn at igalia.com>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
M Source/WebCore/platform/graphics/gstreamer/GStreamerCommon.cpp
M Source/WebCore/platform/graphics/gstreamer/GStreamerCommon.h
M Source/WebCore/platform/graphics/gstreamer/VideoTrackPrivateGStreamer.cpp
Log Message:
-----------
Cherry-pick 256396 at main (2a9d6e23045c). https://bugs.webkit.org/show_bug.cgi?id=247524
[GStreamer] Rebrand colorspace handling to common facilities
https://bugs.webkit.org/show_bug.cgi?id=247524
Reviewed by Xabier Rodriguez-Calvar.
We had some basic GstVideoInfo -> PlatformVideoColorSpace conversion in VideoTrackPrivateGStreamer.
This now lives in GStreamerCommon along with a GstCaps -> PlatformVideoColorSpace and
PlatformVideoColorSpace -> GstCaps utilities. Those will be useful to improve the
VideoFrameGStreamer implementation for the WebCodecs backend.
* Source/WebCore/platform/graphics/gstreamer/GStreamerCommon.cpp:
(WebCore::videoColorSpaceFromCaps):
(WebCore::videoColorSpaceFromInfo):
(WebCore::fillVideoInfoColorimetryFromColorSpace):
* Source/WebCore/platform/graphics/gstreamer/GStreamerCommon.h:
* Source/WebCore/platform/graphics/gstreamer/VideoTrackPrivateGStreamer.cpp:
(WebCore::VideoTrackPrivateGStreamer::updateConfigurationFromCaps):
Canonical link: https://commits.webkit.org/256396@main
Commit: 79f99def66c22098eb4ca0c9bcf7a39b54f29048
https://github.com/WebKit/WebKit/commit/79f99def66c22098eb4ca0c9bcf7a39b54f29048
Author: Chirag M Shah <chirag_m_shah at apple.com>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
A LayoutTests/js/frame-application-cache-with-listener-crash-expected.txt
A LayoutTests/js/frame-application-cache-with-listener-crash.html
M Source/WebCore/loader/appcache/DOMApplicationCache.cpp
Log Message:
-----------
Cherry-pick 256402 at main (1624156ba5f9). https://bugs.webkit.org/show_bug.cgi?id=247380
Make sure scriptExecutionContext stays around when invoking listeners
https://bugs.webkit.org/show_bug.cgi?id=247380
Reviewed by Ryosuke Niwa.
This change fixes DOMApplicationCache::scriptExecutionContext to get the
correct scriptExecutionContext by getting it from the window instead of
the frame because the frame can navigate when invoking event listeners, in which
case the scriptExecutionContext will become NULL.
* LayoutTests/js/frame-application-cache-with-listener-crash-expected.txt: Added.
* LayoutTests/js/frame-application-cache-with-listener-crash.html: Added.
* Source/WebCore/loader/appcache/DOMApplicationCache.cpp:
(WebCore::DOMApplicationCache::scriptExecutionContext const):
Canonical link: https://commits.webkit.org/256402@main
Commit: 7026e87a1b8e647125a63303da6bc3c66257cd6a
https://github.com/WebKit/WebKit/commit/7026e87a1b8e647125a63303da6bc3c66257cd6a
Author: Jer Noble <jer.noble at apple.com>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
A LayoutTests/media/track/video-track-add-visible-expected.txt
A LayoutTests/media/track/video-track-add-visible.html
M Source/WebCore/html/HTMLMediaElement.cpp
Log Message:
-----------
Cherry-pick 256408 at main (7ff585820d47). https://bugs.webkit.org/show_bug.cgi?id=247549
Apple.com/apple-events: Unable to enable subtitles on first watch of the keynote
https://bugs.webkit.org/show_bug.cgi?id=247549
rdar://101785175
Reviewed by Eric Carlson.
Two interlocking bugs prevent captions from being visible if tracks are added after the video is
loaded: 1) `m_processingPreferenceChange` is set to `true` in
`markCaptionAndSubtitleTracksAsUnconfigured()` and is only set to `false` if at least one text
track is present and therefore `configureTextTrackGroup()` is called. 2) the MediaControlsHost is
not notified that the video element's videoBox has changed, as it's only notified during style
updates, and will not necessarily have a RenderVideo attached at that point.
Clear `m_processingPreferenceChange` unconditionally at the end of `configureTextTracks()`. And call
the MediaControlsHost's `updateCaptionDisplaySizes()` when the media element is resized.
* LayoutTests/media/track/video-track-add-visible-expected.txt: Added.
* LayoutTests/media/track/video-track-add-visible.html: Added.
* Source/WebCore/html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::configureTextTrackGroup):
(WebCore::HTMLMediaElement::layoutSizeChanged):
(WebCore::HTMLMediaElement::configureTextTracks):
Canonical link: https://commits.webkit.org/256408@main
Commit: b1e8d7704ad19db74977a546b434f3f2be0ddd71
https://github.com/WebKit/WebKit/commit/b1e8d7704ad19db74977a546b434f3f2be0ddd71
Author: Xabier Rodriguez-Calvar <calvaris at igalia.com>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
M LayoutTests/media/media-source/media-source-append-before-last-range-no-quota-exceeded-expected.txt
M LayoutTests/media/media-source/media-source-append-before-last-range-no-quota-exceeded.html
M LayoutTests/media/media-source/media-source-append-buffer-full-quota-exceeded-error-expected.txt
M LayoutTests/media/media-source/media-source-append-buffer-full-quota-exceeded-error.html
M LayoutTests/media/media-source/media-source-evict-codedframe-after-seek-expected.txt
M Source/WebCore/Modules/mediasource/SourceBuffer.cpp
M Source/WebCore/platform/graphics/PlatformTimeRanges.cpp
M Source/WebCore/platform/graphics/PlatformTimeRanges.h
M Source/WebCore/platform/graphics/SourceBufferPrivate.cpp
M Source/WebCore/platform/graphics/SourceBufferPrivate.h
M Source/WebKit/GPUProcess/media/RemoteSourceBufferProxy.cpp
M Source/WebKit/GPUProcess/media/RemoteSourceBufferProxy.h
M Source/WebKit/GPUProcess/media/RemoteSourceBufferProxy.messages.in
M Source/WebKit/WebProcess/GPU/media/SourceBufferPrivateRemote.cpp
M Source/WebKit/WebProcess/GPU/media/SourceBufferPrivateRemote.h
Log Message:
-----------
Cherry-pick 256441 at main (48b51f0536c9). https://bugs.webkit.org/show_bug.cgi?id=246144
[MSE] Improved coded frame eviction algorithm
https://bugs.webkit.org/show_bug.cgi?id=246144
Reviewed by Alicia Boya Garcia.
I made it in two aspects, first is that now it does not begin at 0, it
begins at the first buffered range.
A fix would be that sometimes you need to evict frames but a window of
30s is too much and it can easily happen it does not evict anything,
so we need to reduce the partition to be able to effectively evict
something. We do it in windows from 30 to 3s.
This is tested by several tests that needed a rebase line or small tweaks.
* LayoutTests/media/media-source/media-source-append-before-last-range-no-quota-exceeded-expected.txt:
* LayoutTests/media/media-source/media-source-append-before-last-range-no-quota-exceeded.html:
* LayoutTests/media/media-source/media-source-append-buffer-full-quota-exceeded-error-expected.txt:
* LayoutTests/media/media-source/media-source-append-buffer-full-quota-exceeded-error.html:
* LayoutTests/media/media-source/media-source-evict-codedframe-after-seek-expected.txt:
* Source/WebCore/Modules/mediasource/SourceBuffer.cpp:
(WebCore::SourceBuffer::appendBufferInternal):
* Source/WebCore/platform/graphics/PlatformTimeRanges.cpp:
(WebCore::PlatformTimeRanges::minimumBufferedTime const):
* Source/WebCore/platform/graphics/PlatformTimeRanges.h:
* Source/WebCore/platform/graphics/SourceBufferPrivate.cpp:
(WebCore::SourceBufferPrivate::evictCodedFrames):
* Source/WebCore/platform/graphics/SourceBufferPrivate.h:
* Source/WebKit/GPUProcess/media/RemoteSourceBufferProxy.cpp:
(WebKit::RemoteSourceBufferProxy::evictCodedFrames):
* Source/WebKit/GPUProcess/media/RemoteSourceBufferProxy.h:
* Source/WebKit/GPUProcess/media/RemoteSourceBufferProxy.messages.in:
* Source/WebKit/WebProcess/GPU/media/SourceBufferPrivateRemote.cpp:
(WebKit::SourceBufferPrivateRemote::evictCodedFrames):
* Source/WebKit/WebProcess/GPU/media/SourceBufferPrivateRemote.h:
Canonical link: https://commits.webkit.org/256441@main
Commit: 0ca02b713bb773b2bf1e4806b36a25426eef7f85
https://github.com/WebKit/WebKit/commit/0ca02b713bb773b2bf1e4806b36a25426eef7f85
Author: Chris Dumez <cdumez at apple.com>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
M Source/WebCore/xml/XMLHttpRequest.cpp
M Source/WebCore/xml/XMLHttpRequest.h
Log Message:
-----------
Cherry-pick 256512 at main (7cfb84288e85). https://bugs.webkit.org/show_bug.cgi?id=244715
Test addition (247071 at main): [ macOS wk1 Debug ] Two imported/w3c/web-platform-tests/xhr/ tests are a flaky failure
https://bugs.webkit.org/show_bug.cgi?id=244715
rdar://99492054
Reviewed by Geoffrey Garen.
The issue is that CFNetwork doesn't reliably calls didSendData() before calling
didFinishLoad(). As a result, it is possible for a XMLHttpRequest upload to be
marked as DONE without getting any progress events on XMLHttpRequest.upload.
This was causing the test to fail flakily since it verified that at least
one `progress` event get fired at `XMLHttpRequest.upload` before the XHR is
marked as DONE.
To address the issue, we now explicitly call didSendData(bodySize) inside
didFinishLoading() if didSendData() wasn't called yet.
* Source/WebCore/xml/XMLHttpRequest.cpp:
(WebCore::XMLHttpRequest::didFinishLoading):
(WebCore::XMLHttpRequest::didSendData):
* Source/WebCore/xml/XMLHttpRequest.h:
Canonical link: https://commits.webkit.org/256512@main
Commit: 0da4c1aeda84cb9318ae32c1320a45fc790741c4
https://github.com/WebKit/WebKit/commit/0da4c1aeda84cb9318ae32c1320a45fc790741c4
Author: Olivier Blin <olivier.blin at softathome.com>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
M Source/cmake/GStreamerChecks.cmake
Log Message:
-----------
Cherry-pick 256487 at main (ee453765232d). https://bugs.webkit.org/show_bug.cgi?id=247675
[GStreamer][WebRTC] Fix cmake macro name
https://bugs.webkit.org/show_bug.cgi?id=247675
Reviewed by Philippe Normand.
* Source/cmake/GStreamerChecks.cmake:
A cmake block checked for USE_GSTREAMER_WEB_RTC, while the actual
option name is USE_GSTREAMER_WEBRTC.
Canonical link: https://commits.webkit.org/256487@main
Commit: 877300a81906c3c3ce238ce156e1701595e71e3c
https://github.com/WebKit/WebKit/commit/877300a81906c3c3ce238ce156e1701595e71e3c
Author: Olivier Blin <olivier.blin at softathome.com>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
M Source/WebCore/platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp
M Source/WebCore/platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.h
M Source/WebCore/platform/graphics/gstreamer/mse/MediaPlayerPrivateGStreamerMSE.cpp
Log Message:
-----------
Cherry-pick 256489 at main (a12a0c36a046). https://bugs.webkit.org/show_bug.cgi?id=247679
[GStreamer] Remove extraneous extendedSupportsType method
https://bugs.webkit.org/show_bug.cgi?id=247679
Reviewed by Philippe Normand.
Content type checks are done in
GStreamerRegistryScanner::isContentTypeSupported() or
MediaPlayerPrivateGStreamerMSE::supportsType().
The extraneous MediaPlayerPrivateGStreamer::extendedSupportsType()
method is not implemented and should be removed, like done for other
ports in r157834.
* Source/WebCore/platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
(WebCore::MediaPlayerPrivateGStreamer::supportsType):
(WebCore::MediaPlayerPrivateGStreamer::extendedSupportsType): Deleted.
* Source/WebCore/platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.h:
* Source/WebCore/platform/graphics/gstreamer/mse/MediaPlayerPrivateGStreamerMSE.cpp:
(WebCore::MediaPlayerPrivateGStreamerMSE::supportsType):
Canonical link: https://commits.webkit.org/256489@main
Commit: 7d55f311c387e84c750314bd04abdeae1a2b74b9
https://github.com/WebKit/WebKit/commit/7d55f311c387e84c750314bd04abdeae1a2b74b9
Author: David Kilzer <ddkilzer at apple.com>
Date: 2023-01-24 (Tue, 24 Jan 2023)
Changed paths:
M Source/WebCore/xml/XSLTUnicodeSort.cpp
Log Message:
-----------
Cherry-pick 256459 at main (6649227d017b). https://bugs.webkit.org/show_bug.cgi?id=247538
Update xsltUnicodeSortFunction()
https://bugs.webkit.org/show_bug.cgi?id=247538
<rdar://101996640>
Reviewed by Alex Christensen.
* Source/WebCore/xml/XSLTUnicodeSort.cpp:
(WebCore::xsltUnicodeSortFunction):
- Update with changes from examples/xsltICUSort.c in libxml2.
Canonical link: https://commits.webkit.org/256459@main
Commit: b8362ed756bb928683c8f6158a1d8bf318999b75
https://github.com/WebKit/WebKit/commit/b8362ed756bb928683c8f6158a1d8bf318999b75
Author: Karl Rackler <rackler at apple.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M LayoutTests/platform/ios/TestExpectations
M LayoutTests/platform/mac-wk2/TestExpectations
Log Message:
-----------
Cherry-pick 256333 at main (c2407ed41733). https://bugs.webkit.org/show_bug.cgi?id=247499
[ New Test ] (256112 at main): ASSERTION FAILED: isValidIdentifier(m_identifier)
https://bugs.webkit.org/show_bug.cgi?id=247499
rdar://101970988
Unreviewed test gardening.
* LayoutTests/platform/ios/TestExpectations:
* LayoutTests/platform/mac-wk2/TestExpectations:
Canonical link: https://commits.webkit.org/256333@main
Commit: a0960047f03ba1df9e20cd9631c7146e9be0e7c9
https://github.com/WebKit/WebKit/commit/a0960047f03ba1df9e20cd9631c7146e9be0e7c9
Author: Miguel Salinas <miguel_salinas at apple.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M Source/WebCore/Modules/indexeddb/IDBOpenDBRequest.cpp
M Source/WebCore/Modules/indexeddb/client/IDBConnectionProxy.cpp
M Source/WebCore/Modules/indexeddb/client/IDBConnectionProxy.h
M Source/WebCore/Modules/indexeddb/client/IDBConnectionToServer.cpp
M Source/WebCore/Modules/indexeddb/client/IDBConnectionToServer.h
M Source/WebCore/Modules/indexeddb/client/IDBConnectionToServerDelegate.h
M Source/WebCore/Modules/indexeddb/server/IDBServer.cpp
M Source/WebCore/Modules/indexeddb/server/IDBServer.h
M Source/WebCore/loader/EmptyClients.cpp
M Source/WebKit/NetworkProcess/storage/NetworkStorageManager.cpp
M Source/WebKit/NetworkProcess/storage/NetworkStorageManager.h
M Source/WebKit/NetworkProcess/storage/NetworkStorageManager.messages.in
M Source/WebKit/WebProcess/Databases/IndexedDB/WebIDBConnectionToServer.cpp
M Source/WebKit/WebProcess/Databases/IndexedDB/WebIDBConnectionToServer.h
M Source/WebKitLegacy/Storage/InProcessIDBServer.cpp
M Source/WebKitLegacy/Storage/InProcessIDBServer.h
Log Message:
-----------
Cherry-pick 256464 at main (b9ff9a4dad90). https://bugs.webkit.org/show_bug.cgi?id=247499
Fix flaky test crash in indexeddb/crash-on-getdatabases.html
https://bugs.webkit.org/show_bug.cgi?id=247499
rdar://101970988
Reviewed by Sihui Liu.
We are sending invalid IDBResourceIdentifiers with
AbortOpenAndUpgradeNeeded IPC messages to the indexeddb server.
* Source/WebCore/Modules/indexeddb/IDBOpenDBRequest.cpp:
(WebCore::IDBOpenDBRequest::requestCompleted):
* Source/WebCore/Modules/indexeddb/client/IDBConnectionProxy.cpp:
(WebCore::IDBClient::IDBConnectionProxy::abortOpenAndUpgradeNeeded):
* Source/WebCore/Modules/indexeddb/client/IDBConnectionProxy.h:
* Source/WebCore/Modules/indexeddb/client/IDBConnectionToServer.cpp:
(WebCore::IDBClient::IDBConnectionToServer::abortOpenAndUpgradeNeeded):
* Source/WebCore/Modules/indexeddb/client/IDBConnectionToServer.h:
* Source/WebCore/Modules/indexeddb/client/IDBConnectionToServerDelegate.h:
* Source/WebCore/Modules/indexeddb/server/IDBServer.cpp:
(WebCore::IDBServer::IDBServer::abortOpenAndUpgradeNeeded):
* Source/WebCore/Modules/indexeddb/server/IDBServer.h:
* Source/WebCore/loader/EmptyClients.cpp:
* Source/WebKit/NetworkProcess/storage/NetworkStorageManager.cpp:
(WebKit::NetworkStorageManager::abortOpenAndUpgradeNeeded):
* Source/WebKit/NetworkProcess/storage/NetworkStorageManager.h:
* Source/WebKit/NetworkProcess/storage/NetworkStorageManager.messages.in:
* Source/WebKit/WebProcess/Databases/IndexedDB/WebIDBConnectionToServer.cpp:
(WebKit::WebIDBConnectionToServer::abortOpenAndUpgradeNeeded):
* Source/WebKit/WebProcess/Databases/IndexedDB/WebIDBConnectionToServer.h:
* Source/WebKitLegacy/Storage/InProcessIDBServer.cpp:
(InProcessIDBServer::abortOpenAndUpgradeNeeded):
* Source/WebKitLegacy/Storage/InProcessIDBServer.h:
* LayoutTests/platform/ios/TestExpectations:
* LayoutTests/platform/mac-wk2/TestExpectations:
Canonical link: https://commits.webkit.org/256464@main
Commit: 52346629495ca46756f96ddafca990cc50750d28
https://github.com/WebKit/WebKit/commit/52346629495ca46756f96ddafca990cc50750d28
Author: David Degazio <d_degazio at apple.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
A JSTests/stress/bigdecimal-identifiers-fail-on-oom.js
M Source/JavaScriptCore/parser/Parser.cpp
M Source/JavaScriptCore/parser/ParserArena.cpp
M Source/JavaScriptCore/parser/ParserArena.h
Log Message:
-----------
Cherry-pick 256501 at main (32b972813979). https://bugs.webkit.org/show_bug.cgi?id=247644
Segfault in JSC::IdentifierArena::makeBigIntDecimalIdentifier
https://bugs.webkit.org/show_bug.cgi?id=247644
rdar://98566429
Reviewed by Mark Lam and Yusuke Suzuki.
We currently get a segfault because the parser for bigdecimal identifiers allocates a JSBigInt, which
might cause us to run out of memory. The parser doesn't throw arbitrary exceptions elsewhere, so instead
of throwing out-of-memory as an exception, it just produces an empty JSBigInt and crashes when using it.
This patch addresses the issue by making the result of makeBigIntDecimalIdentifier nullable, checking for
it in the parser, and failing with a SyntaxError if the identifier could not be created.
* JSTests/stress/bigdecimal-identifiers-fail-on-oom.js: Added.
(foo):
* Source/JavaScriptCore/parser/Lexer.cpp:
* Source/JavaScriptCore/parser/Parser.cpp:
(JSC::Parser<LexerType>::parseDestructuringPattern):
(JSC::Parser<LexerType>::parseClass):
(JSC::Parser<LexerType>::parseClassFieldInitializerSourceElements):
(JSC::Parser<LexerType>::parseProperty):
(JSC::Parser<LexerType>::parseGetterSetter):
* Source/JavaScriptCore/parser/ParserArena.cpp:
(JSC::IdentifierArena::makeBigIntDecimalIdentifier):
* Source/JavaScriptCore/parser/ParserArena.h:
Canonical link: https://commits.webkit.org/256501@main
Commit: d55586dddcfe9ef302f81112391067b497250b00
https://github.com/WebKit/WebKit/commit/d55586dddcfe9ef302f81112391067b497250b00
Author: Chris Dumez <cdumez at apple.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M Source/WebCore/dom/Document.cpp
Log Message:
-----------
Cherry-pick 256552 at main (9c08d49c54bf). https://bugs.webkit.org/show_bug.cgi?id=247761
Regression(254699 at main): fast/events/remove-iframe-during-toggle-crash.html is crashing
https://bugs.webkit.org/show_bug.cgi?id=247761
rdar://100558596
Reviewed by Geoffrey Garen.
254699 at main added a call to `policyChecker().stopCheck()` inside Document::open().
The stopCheck() call can clear `m_frame` so we need to null-check `m_frame` again
after the call.
* Source/WebCore/dom/Document.cpp:
(WebCore::Document::open):
Canonical link: https://commits.webkit.org/256552@main
Commit: 9016f59fb243c288cfe9c432c0cdb61c3b9e20f1
https://github.com/WebKit/WebKit/commit/9016f59fb243c288cfe9c432c0cdb61c3b9e20f1
Author: Karl Dubost <karlcow at apple.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M Source/WebCore/page/Quirks.cpp
Log Message:
-----------
Cherry-pick 256558 at main (8e0e88f4f3d2). https://bugs.webkit.org/show_bug.cgi?id=247445
Please remove a desmos.com quirks entry
https://bugs.webkit.org/show_bug.cgi?id=247445
rdar://49124313
Reviewed by Simon Fraser and Tim Horton.
As discussed in the bug, one of the website developers want
the quirk to be removed. It was anyway ineffective because
it was targeting the wrong path.
* Source/WebCore/page/Quirks.cpp:
(WebCore::Quirks::shouldDispatchSimulatedMouseEvents const):
Canonical link: https://commits.webkit.org/256558@main
Commit: e369a91d8748924cf63fc5aefb2c938a57145cad
https://github.com/WebKit/WebKit/commit/e369a91d8748924cf63fc5aefb2c938a57145cad
Author: Yusuke Suzuki <ysuzuki at apple.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M Source/JavaScriptCore/heap/BlockDirectory.cpp
M Source/JavaScriptCore/heap/MarkedBlock.h
Log Message:
-----------
Cherry-pick 256554 at main (292a16569829). https://bugs.webkit.org/show_bug.cgi?id=247727
[UNIX] Web process crash in websites using service workers while doing garbage collection
https://bugs.webkit.org/show_bug.cgi?id=247727
rdar://102209090
Reviewed by Mark Lam.
MarkedBlock::Handle::start() is not returning page aligned address, so this is not appropriate for mincore.
* Source/JavaScriptCore/heap/BlockDirectory.cpp:
(JSC::BlockDirectory::updatePercentageOfPagedOutPages):
Canonical link: https://commits.webkit.org/256554@main
Commit: 96e5738465bf2c444ceff16644bbdafb9b96adde
https://github.com/WebKit/WebKit/commit/96e5738465bf2c444ceff16644bbdafb9b96adde
Author: Mark Lam <mark.lam at apple.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M Source/WTF/wtf/URLHelpers.cpp
Log Message:
-----------
Cherry-pick 256560 at main (06f9978e52a8). https://bugs.webkit.org/show_bug.cgi?id=247770
Change the value of scriptCodeLimit to placate a strange Clang quirk.
https://bugs.webkit.org/show_bug.cgi?id=247770
<rdar://problem/102115540>
Reviewed by Alex Christensen.
Some versions of Clang appear to think that scriptCodeLimit should fit in a signed
9-bit integer, which is crazy. For our purposes, a value of 255 works just as well.
So, we're changing scriptCodeLimit to be 255 instead of 256 to work around this Clang
quirk.
* Source/WTF/wtf/URLHelpers.cpp:
Canonical link: https://commits.webkit.org/256560@main
Commit: d9e02562c11ad29a0132b6fdc160190c6a0ebb87
https://github.com/WebKit/WebKit/commit/d9e02562c11ad29a0132b6fdc160190c6a0ebb87
Author: Chris Dumez <cdumez at apple.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M Source/WebKit/UIProcess/WebPageProxy.cpp
M Tools/TestWebKitAPI/Tests/WebKitCocoa/WebContentProcessDidTerminate.mm
Log Message:
-----------
Cherry-pick 256575 at main (b94ca3b38a05). https://bugs.webkit.org/show_bug.cgi?id=247773
Improve WebKit's default crash handling behavior
https://bugs.webkit.org/show_bug.cgi?id=247773
Reviewed by Alex Christensen.
Improve WebKit's default crash handling behavior in the event where there is a
pending API request URL. In such case, we now attempt to load the pending API
request URL again instead of issuing a reload.
This is particularly important when there is no current HistoryItem (i.e. this
is the first load and the WebProcess just crashed), since the default behavior
was to issue a reload and reloading is a no-op when there is no current
HistoryItem.
* Source/WebKit/UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::tryReloadAfterProcessTermination):
* Tools/TestWebKitAPI/Tests/WebKitCocoa/WebContentProcessDidTerminate.mm:
(TEST):
Canonical link: https://commits.webkit.org/256575@main
Commit: 97f322304f0d101eac808e7587126d142f6cd3ad
https://github.com/WebKit/WebKit/commit/97f322304f0d101eac808e7587126d142f6cd3ad
Author: Patrick Griffis <pgriffis at igalia.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M Source/WebCore/platform/audio/glib/MediaSessionGLib.cpp
M Source/WebKit/UIProcess/Launcher/glib/XDGDBusProxy.cpp
Log Message:
-----------
Cherry-pick 256654 at main (c319637f7e9c). https://bugs.webkit.org/show_bug.cgi?id=245221
[GLIB] Fix MPRIS in flatpak sandbox
https://bugs.webkit.org/show_bug.cgi?id=245221
Reviewed by Michael Catanzaro.
Currently MPRIS works in our bubblewrap sandbox but not in flatpak.
As of https://github.com/flatpak/flatpak/commit/6540f85511739028c0f8eb496ae64dcc73d40a28
it can function inside of flatpak but needs to be named `org.mpris.MediaPlayer2.%s.Sandboxed.*`.
* Source/WebCore/platform/audio/glib/MediaSessionGLib.cpp:
(WebCore::MediaSessionGLib::ensureMprisSessionRegistered):
* Source/WebKit/UIProcess/Launcher/glib/XDGDBusProxy.cpp:
(WebKit::XDGDBusProxy::dbusSessionProxy):
Canonical link: https://commits.webkit.org/256654@main
Commit: 8f5edf980ff272ea4d3acb41eee97403b2aaf70e
https://github.com/WebKit/WebKit/commit/8f5edf980ff272ea4d3acb41eee97403b2aaf70e
Author: Jean-Yves Avenard <jya at apple.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M Source/WebCore/Modules/mediasession/MediaMetadata.cpp
M Source/WebCore/html/MediaElementSession.cpp
M Source/WebCore/platform/NowPlayingManager.cpp
M Source/WebCore/platform/NowPlayingManager.h
M Source/WebCore/platform/audio/NowPlayingInfo.h
M Source/WebCore/platform/audio/cocoa/MediaSessionManagerCocoa.mm
M Source/WebCore/platform/audio/glib/MediaSessionGLib.cpp
M Source/WebCore/platform/mac/MediaRemoteSoftLink.h
M Source/WebCore/platform/mac/MediaRemoteSoftLink.mm
M Tools/TestWebKitAPI/Tests/WebCore/NowPlayingInfoTests.cpp
Log Message:
-----------
Cherry-pick 256627 at main (8f70e6bf626c). https://bugs.webkit.org/show_bug.cgi?id=247043
mediaSession API not showing artwork
https://bugs.webkit.org/show_bug.cgi?id=247043
rdar://101597398
Reviewed by Eric Carlson
When we send an artwork image to MediaRemote, recent versions now check if the dimensions of
that image are valid by calling CGImageSourceCreateWithData with the provided image
and looking at the dimensions. However, in the GPU process, CoreGraphic's decoders are disabled by default and the creation of the CGImage fails.
Adopt new MediaRemote entitlement: com.apple.mediaremote.external-artwork-validation
which will skip this check provided the dimensions are also provided to MediaRemote via the kMRMediaRemoteNowPlayingInfoArtworkDataWidth and kMRMediaRemoteNowPlayingInfoArtworkDataHeight key.
Additionally, we will now properly decode the image in the content process to
ensure its validity. This is automatically done by the RefPtr<Image> IPC encoder.
We then reencode that decoded image into a TIFF as MediaRemote does require a valid comppressed image.
Manually tested by temporarily removing the CG's decoders restriction until a version with the MediaRemote fix is available.
* Source/WebCore/Modules/mediasession/MediaMetadata.cpp:
(WebCore::ArtworkImageLoader::notifyFinished):
* Source/WebCore/html/MediaElementSession.cpp:
(WebCore::MediaElementSession::nowPlayingInfo const):
* Source/WebCore/platform/NowPlayingManager.cpp:
(WebCore::NowPlayingManager::setNowPlayingInfo):
(WebCore::NowPlayingManager::setNowPlayingInfoPrivate):
* Source/WebCore/platform/NowPlayingManager.h:
* Source/WebCore/platform/audio/NowPlayingInfo.h:
(WebCore::NowPlayingInfoArtwork::encode const):
(WebCore::NowPlayingInfoArtwork::decode):
* Source/WebCore/platform/audio/cocoa/MediaSessionManagerCocoa.mm:
(WebCore::MediaSessionManagerCocoa::setNowPlayingInfo):
* Source/WebCore/platform/mac/MediaRemoteSoftLink.h:
* Source/WebCore/platform/mac/MediaRemoteSoftLink.mm:
* Source/WebKit/Scripts/process-entitlements.sh:
* Tools/TestWebKitAPI/Tests/WebCore/NowPlayingInfoTests.cpp:
(TestWebKitAPI::testEmptyArtwork):
Canonical link: https://commits.webkit.org/256627@main
Commit: cbc36a419819da7e33428576149e231a10bee8e2
https://github.com/WebKit/WebKit/commit/cbc36a419819da7e33428576149e231a10bee8e2
Author: Justin Michaud <justin_michaud at apple.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
A JSTests/stress/iterator-done-exception-check.js
M Source/JavaScriptCore/runtime/IteratorOperations.cpp
Log Message:
-----------
Cherry-pick 256593 at main (a084fca782e9). https://bugs.webkit.org/show_bug.cgi?id=247419.
Iterator completion should check for exceptions.
https://bugs.webkit.org/show_bug.cgi?id=247419.
Reviewed by Yusuke Suzuki.
Let's add a happy little exception check after calling iterator.done.
* JSTests/stress/iterator-done-exception-check.js: Added.
(main):
* Source/JavaScriptCore/runtime/IteratorOperations.cpp:
(JSC::iteratorComplete):
Canonical link: https://commits.webkit.org/256593@main
Commit: 57cb5d7f49b1f9f9d670be7a93ccdb612eaff82c
https://github.com/WebKit/WebKit/commit/57cb5d7f49b1f9f9d670be7a93ccdb612eaff82c
Author: Simon Fraser <simon.fraser at apple.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
A LayoutTests/compositing/clipping/sticky-inside-overflow-x-clip-expected.txt
A LayoutTests/compositing/clipping/sticky-inside-overflow-x-clip.html
A LayoutTests/platform/gtk/compositing/clipping/sticky-inside-overflow-x-clip-expected.txt
A LayoutTests/platform/ios-wk2/compositing/clipping/sticky-inside-overflow-x-clip-expected.txt
A LayoutTests/platform/mac-wk1/compositing/clipping/sticky-inside-overflow-x-clip-expected.txt
M Source/WebCore/platform/graphics/LayoutRect.h
M Source/WebCore/rendering/RenderLayerCompositor.cpp
Log Message:
-----------
Cherry-pick 256618 at main (854716970c07). https://bugs.webkit.org/show_bug.cgi?id=247130
Elements with `position: sticky` jitter inside a container with `overflow-x: clip;`
https://bugs.webkit.org/show_bug.cgi?id=247130
rdar://101645570
Reviewed by Alan Baradlay.
`overflow:clip` on just one axis can output a clip rect which is infinite in just
x or y. In that case, we'd create a clipping layer with large floating point
geometry values which, when moved around by the scrolling tree, had low precision
resulting in visible stutters.
Fix by shrinking the infinite rect values by a factor of 16, which is sufficient
to provide enough device pixel precision for 3x displays.
* LayoutTests/compositing/clipping/sticky-inside-overflow-x-clip-expected.txt: Added.
* LayoutTests/compositing/clipping/sticky-inside-overflow-x-clip.html: Added.
* LayoutTests/platform/gtk/compositing/clipping/sticky-inside-overflow-x-clip-expected.txt: Added.
* LayoutTests/platform/ios-wk2/compositing/clipping/sticky-inside-overflow-x-clip-expected.txt: Added.
* LayoutTests/platform/mac-wk1/compositing/clipping/sticky-inside-overflow-x-clip-expected.txt: Added.
* Source/WebCore/platform/graphics/LayoutRect.h:
* Source/WebCore/rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::computeAncestorClippingStack const):
Canonical link: https://commits.webkit.org/256618@main
Commit: 10a8613e183ee8c80d6653c55150060779de77fa
https://github.com/WebKit/WebKit/commit/10a8613e183ee8c80d6653c55150060779de77fa
Author: Simon Fraser <simon.fraser at apple.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
A LayoutTests/platform/ios-wk2/scrollingcoordinator/scrolling-tree/composited-in-offscreen-fixed-expected.txt
A LayoutTests/scrollingcoordinator/scrolling-tree/composited-in-offscreen-fixed-expected.txt
A LayoutTests/scrollingcoordinator/scrolling-tree/composited-in-offscreen-fixed.html
A LayoutTests/scrollingcoordinator/scrolling-tree/scroll-coordinated-inside-fixed-expected.txt
A LayoutTests/scrollingcoordinator/scrolling-tree/scroll-coordinated-inside-fixed.html
M Source/WebCore/rendering/RenderLayerCompositor.cpp
Log Message:
-----------
Cherry-pick 256619 at main (242980046171). https://bugs.webkit.org/show_bug.cgi?id=247837
REGRESSION (253809 at main): Comment field on Engadget.com exhibits stutters while scrolling
https://bugs.webkit.org/show_bug.cgi?id=247837
rdar://102066231
Reviewed by Alan Baradlay.
Various sites (khanacademy.org, comments on engadget.com) showed incorrectly scroll-coordinated elements,
typically positioned nodes inside overflow:scroll inside position:fixed. On some other sites, we'd hit an
assertion in RenderLayerCompositor::updateScrollingNodeForViewportConstrainedRole().
In all cases, we were trying to treat a layer that was some descendant of a position:fixed layer as if it
were the fixed layer itself, because of a logic error in 253809 at main. The code in
`RenderLayerCompositor::isViewportConstrainedFixedOrStickyLayer()` needs to return early if the layer is
not both position:fixed, and behaving as fixed (i.e. no transformed ancestor).
Added two layout tests, both of which hit the assertion, and one of which showed incorrect scrolling
behavior.
* LayoutTests/platform/ios-wk2/scrollingcoordinator/scrolling-tree/composited-in-offscreen-fixed-expected.txt: Added.
* LayoutTests/scrollingcoordinator/scrolling-tree/composited-in-offscreen-fixed-expected.txt: Added.
* LayoutTests/scrollingcoordinator/scrolling-tree/composited-in-offscreen-fixed.html: Added.
* LayoutTests/scrollingcoordinator/scrolling-tree/scroll-coordinated-inside-fixed-expected.txt: Added.
* LayoutTests/scrollingcoordinator/scrolling-tree/scroll-coordinated-inside-fixed.html: Added.
* Source/WebCore/rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::isViewportConstrainedFixedOrStickyLayer const):
Canonical link: https://commits.webkit.org/256619@main
Commit: bd57377ebe2092505d06df81c038bab660403608
https://github.com/WebKit/WebKit/commit/bd57377ebe2092505d06df81c038bab660403608
Author: Ryosuke Niwa <rniwa at webkit.org>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
A LayoutTests/fast/forms/submit-form-dialog-crash-expected.txt
A LayoutTests/fast/forms/submit-form-dialog-crash.html
M Source/WebCore/loader/FormSubmission.cpp
Log Message:
-----------
Cherry-pick 256613 at main (d44b5de990f9). https://bugs.webkit.org/show_bug.cgi?id=247652
Crash in FormData::flatten
https://bugs.webkit.org/show_bug.cgi?id=247652
Reviewed by Youenn Fablet.
The bug was caused by FormSubmission::requestURL assuming that the method is GET
if it's not POST, and trying to serialize the form data as the URL query string.
Fixed the bug by adding method type check.
* LayoutTests/fast/forms/submit-form-dialog-crash-expected.txt: Added.
* LayoutTests/fast/forms/submit-form-dialog-crash.html: Added.
* Source/WebCore/loader/FormSubmission.cpp:
(WebCore::FormSubmission::requestURL const):
Canonical link: https://commits.webkit.org/256613@main
Commit: d038baf3954b50eb4bb2976531469b55ca38ee7a
https://github.com/WebKit/WebKit/commit/d038baf3954b50eb4bb2976531469b55ca38ee7a
Author: Philippe Normand <philn at igalia.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M LayoutTests/http/tests/security/resources/reference-movie-cross-origin-allow.py
M Source/WebCore/platform/graphics/gstreamer/TrackPrivateBaseGStreamer.cpp
M Source/WebCore/platform/graphics/gstreamer/TrackPrivateBaseGStreamer.h
Log Message:
-----------
Cherry-pick 256649 at main (d6ad2e4228d0). https://bugs.webkit.org/show_bug.cgi?id=247800
[GStreamer] Monitor GstStream tags and minor test fix
https://bugs.webkit.org/show_bug.cgi?id=247800
Reviewed by Xabier Rodriguez-Calvar.
For playbin2 case we have some code to keep track of sticky tag events in TrackPrivateBase. For
playbin3 case we can simply monitor the tags property on the GstStream.
* LayoutTests/http/tests/security/resources/reference-movie-cross-origin-allow.py: Fix content-type
for mp4 resource.
* Source/WebCore/platform/graphics/gstreamer/TrackPrivateBaseGStreamer.cpp:
(WebCore::TrackPrivateBaseGStreamer::TrackPrivateBaseGStreamer):
(WebCore::TrackPrivateBaseGStreamer::setPad):
(WebCore::TrackPrivateBaseGStreamer::disconnect):
(WebCore::TrackPrivateBaseGStreamer::tagsChangedCallback): Deleted.
* Source/WebCore/platform/graphics/gstreamer/TrackPrivateBaseGStreamer.h:
Canonical link: https://commits.webkit.org/256649@main
Commit: 707dcec505ae4ad08f17443fa93a61b33fef413d
https://github.com/WebKit/WebKit/commit/707dcec505ae4ad08f17443fa93a61b33fef413d
Author: Xabier Rodriguez-Calvar <calvaris at igalia.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M Source/WebCore/platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp
Log Message:
-----------
Cherry-pick 256650 at main (1302f18a9706). https://bugs.webkit.org/show_bug.cgi?id=247740
[GStreamer] Video sink should discard late frames to avoid audio and video desynchronization
https://bugs.webkit.org/show_bug.cgi?id=247740
Reviewed by Philippe Normand.
In some cases and/or platforms video frames can arrive late and if are not discarded, they end up being painted and
there can be a/v desynchronization. Setting the max-lateness property forces the sink to discard late frames and solve
the problem.
* Source/WebCore/platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
(WebCore::MediaPlayerPrivateGStreamer::createVideoSink):
Canonical link: https://commits.webkit.org/256650@main
Commit: 441769b233363930acbaf0f6c4b18be1659e4aca
https://github.com/WebKit/WebKit/commit/441769b233363930acbaf0f6c4b18be1659e4aca
Author: Sihui Liu <sihui_liu at apple.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M Source/WebKit/NetworkProcess/storage/SQLiteStorageArea.cpp
M Source/WebKit/NetworkProcess/storage/SQLiteStorageArea.h
Log Message:
-----------
Cherry-pick 256670 at main (2cb930b2456b). https://bugs.webkit.org/show_bug.cgi?id=247767
Revert 255044 at main as it introduces crash in SQLiteStorageArea::removeItem
https://bugs.webkit.org/show_bug.cgi?id=247767
rdar://101927216
Reviewed by Ben Nham and Per Arne Vollan.
From 255044 at main we started to track LocalStorage size manually for quota check instead of setting hard limit on
database file. However, LocalStorage might convert 16-bit string to 8-bit string when reading from database for memory
optimization (246784 at main), and we didn't consider that in computation. This has let to overflow error in
SQLiteStorageArea::removeItem.
* Source/WebKit/NetworkProcess/storage/SQLiteStorageArea.cpp:
(WebKit::SQLiteStorageArea::isEmpty):
(WebKit::SQLiteStorageArea::prepareDatabase):
(WebKit::SQLiteStorageArea::getItem):
(WebKit::SQLiteStorageArea::allItems):
(WebKit::SQLiteStorageArea::setItem):
(WebKit::SQLiteStorageArea::removeItem):
(WebKit::SQLiteStorageArea::clear):
(WebKit::SQLiteStorageArea::getAllItemsFromDatabase): Deleted.
(WebKit::SQLiteStorageArea::initializeCache): Deleted.
* Source/WebKit/NetworkProcess/storage/SQLiteStorageArea.h:
Canonical link: https://commits.webkit.org/256670@main
Commit: 8da5ad7217ae3e5ee7e3fcff67d99694b9c70a7e
https://github.com/WebKit/WebKit/commit/8da5ad7217ae3e5ee7e3fcff67d99694b9c70a7e
Author: Ahmad Saleem <ahmad.saleem792+github at gmail.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
A LayoutTests/editing/inserting/replace-in-heading-001-expected.txt
A LayoutTests/editing/inserting/replace-in-heading-001.html
A LayoutTests/editing/inserting/replace-in-paragraph-001-expected.txt
A LayoutTests/editing/inserting/replace-in-paragraph-001.html
M Source/WebCore/editing/ReplaceSelectionCommand.cpp
Log Message:
-----------
Cherry-pick 256682 at main (531d3679e268). https://bugs.webkit.org/show_bug.cgi?id=247918
Change ReplaceSelectionCommand::moveNodeOutOfAncestor() to avoid hitting assertion while trying replace selection
Change ReplaceSelectionCommand::moveNodeOutOfAncestor() to avoid hitting assertion while trying replace selection
https://bugs.webkit.org/show_bug.cgi?id=247918
Reviewed by Ryosuke Niwa.
Merge - https://src.chromium.org/viewvc/blink?revision=153031&view=revision
Before this patch, replacing selection in heading/paragraph with heading/paragraph gets assertion at InsertNodeBeforeCommand constructor, or crash when assertion disabled. It tries to move replaced heading/paragraph out of ancestor, ReplaceSelectionCommand::moveNodeOutOfAncestor().
This patch changes ReplaceSelectionCommand::moveNodeOutOfAncestor() to avoid the assertion.
* Source/WebCore/editing/ReplaceSelectionCommand.cpp:
(ReplaceSelectionCommand::moveNodeOutOfAncestor): Add logic to avoid crash
* LayoutTests/editing/inserting/replace-in-paragraph-001.html: Added Test Case
* LayoutTests/editing/inserting/replace-in-paragraph-001-expected.txt: Added Test Case Expectations
* LayoutTests/editing/inserting/replace-in-heading-001.html: Added Test Case
* LayoutTests/editing/inserting/replace-in-heading-001-expected.txt: Added Test Case Expectations
Canonical link: https://commits.webkit.org/256682@main
Commit: f27e527f78c3906288c6db34590e90021661fd73
https://github.com/WebKit/WebKit/commit/f27e527f78c3906288c6db34590e90021661fd73
Author: Richard Robinson <richard_robinson2 at apple.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M Source/WebCore/platform/ScrollTypes.h
Log Message:
-----------
Cherry-pick 256683 at main (3825a94eb763). https://bugs.webkit.org/show_bug.cgi?id=247908
With UI-side compositing, there is an assertion failure in `fast/scrolling/keyboard-scrolling-distance-downArrow.html`
https://bugs.webkit.org/show_bug.cgi?id=247908
rdar://102329463
Reviewed by Simon Fraser.
The `ScrollDirection` enum was never given an `EnumTraits` specialization which
led to a non-valid `ScrollDirection` value being used in a switch statement
and not being handled, therefore leading to an assertion.
* Source/WebCore/platform/ScrollTypes.h:
Canonical link: https://commits.webkit.org/256683@main
Commit: 5e0d3d6091ec1009c2c08583bc5bee4a38d733af
https://github.com/WebKit/WebKit/commit/5e0d3d6091ec1009c2c08583bc5bee4a38d733af
Author: Yury Semikhatsky <yurys at chromium.org>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M Source/WebKit/UIProcess/API/glib/WebKitDownload.cpp
Log Message:
-----------
Cherry-pick 256737 at main (c985acc37157). https://bugs.webkit.org/show_bug.cgi?id=242217
[WPE][GTK] web process leak if webkit_download_set_destination is called with empty destination
https://bugs.webkit.org/show_bug.cgi?id=242217
Reviewed by Michael Catanzaro.
Fail if download destination is set to "".
* Source/WebKit/UIProcess/API/glib/WebKitDownload.cpp:
(webkit_download_set_destination):
Canonical link: https://commits.webkit.org/256737@main
Commit: f1ad89548911127d86ee1e2495f2be11a96dcac0
https://github.com/WebKit/WebKit/commit/f1ad89548911127d86ee1e2495f2be11a96dcac0
Author: Madhu <enometh at net.meer>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M Source/WebKit/UIProcess/API/gtk/WebKitAuthenticationDialog.cpp
M Source/WebKit/UIProcess/API/gtk/WebKitWebViewBase.cpp
Log Message:
-----------
Cherry-pick 256742 at main (267e69155f83). https://bugs.webkit.org/show_bug.cgi?id=246852
[GTK] Crash on authentication dialog with GTK 4
https://bugs.webkit.org/show_bug.cgi?id=246852
Reviewed by Michael Catanzaro
fix gtk4 auth dialog crashes. take three
* Source/WebKit/UIProcess/API/gtk/WebKitAuthenticationDialog.cpp:
(okButtonClicked): `rememberPassword' is a GtkCheckButton in gtk4,
doesn't inherit Button or ToggleButton. fix cast.
(webkitAuthenticationDialogInitialize): `rememberCheckButton' is a
GtkCheckButton which in Gtk4 apparently gets created with a label
property and no explicit label child. Do not set the the line wrap
property on a non-existent label.
(webkitAuthenticationDialogInitialize): `rememberCheckButton', adjust
for CheckButton changes in gtk4.
(webkitAuthenticationDialogUnmap): new. Implementation unsets the
default window widget which is set in
webkitAuthenticationDialogMap. Avoids the gtk4 crash reported in bug.
(webkit_authentication_dialog_class_init): hook in the above.
* Source/WebKit/UIProcess/API/gtk/WebKitWebViewBase.cpp:
"focus" is not a valid signal for gtk4's widgets such as the
WebKitAutheticationDialog. implement "move-focus" instead, to handle
the situation when a dialog is active and weforward focus events to
it, and we can tab between elements in the box. Presently there is a
problem: if you tab past the last element in the dialog box you cannot
get back in.
(webkitWebViewBaseMoveFocus): new. implementation which will be called
when someone emits a move-focus signal on us.
Canonical link: https://commits.webkit.org/256742@main
Commit: 976a4282a82acfd4b9edb7137f680101c499532a
https://github.com/WebKit/WebKit/commit/976a4282a82acfd4b9edb7137f680101c499532a
Author: Karl Dubost <karlcow at apple.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M Source/WebCore/css/MediaQueryEvaluator.cpp
M Source/WebCore/page/Quirks.cpp
M Source/WebCore/page/Quirks.h
Log Message:
-----------
Cherry-pick 256714 at main (02f013235584). https://bugs.webkit.org/show_bug.cgi?id=247940
Remove Quirks shouldDisableResolutionMediaQuery()
https://bugs.webkit.org/show_bug.cgi?id=247940
rdar://102355671
Reviewed by Tim Nguyen.
This PR removes the quirks which previously disabled the
resolution media feature on Expedia Group sites. When
disabling Site Specific Hacks, the site behaves correctly.
They probably have changed their widget. The patch was tested
for different window widths, and the issue didn't reproduce.
* Source/WebCore/page/Quirks.cpp:
(WebCore::Quirks::shouldDisableResolutionMediaQuery const): Deleted.
* Source/WebCore/css/LegacyMediaQueryEvaluator.cpp:
(WebCore::resolutionEvaluate):
* Source/WebCore/page/Quirks.h:
Canonical link: https://commits.webkit.org/256714@main
Commit: 283030b891c3f944158afd47cf45a6df2016d875
https://github.com/WebKit/WebKit/commit/283030b891c3f944158afd47cf45a6df2016d875
Author: Youenn Fablet <youennf at gmail.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M Source/WebCore/loader/DocumentLoader.cpp
Log Message:
-----------
Cherry-pick 256725 at main (30bd7667b181). https://bugs.webkit.org/show_bug.cgi?id=247793
Clicking a link to a download file served by a service-worker with a fetch event handler yields "Frame load interrupted" error
https://bugs.webkit.org/show_bug.cgi?id=247793
rdar://problem/102296700
Web Inspector shows the "Frame load interrupted" error based on the sub resource load failing error.
Mark the error type as Cancellation so that Web Inspector does not show this message.
Reviewed by Alex Christensen.
* Source/WebCore/loader/DocumentLoader.cpp:
(WebCore::DocumentLoader::interruptedForPolicyChangeError const):
(WebCore::DocumentLoader::stopLoadingForPolicyChange):
Canonical link: https://commits.webkit.org/256725@main
Commit: 892d593204a8c37d16f461975a6b9e7f1e588036
https://github.com/WebKit/WebKit/commit/892d593204a8c37d16f461975a6b9e7f1e588036
Author: Youenn Fablet <youennf at gmail.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
A LayoutTests/http/wpt/fetch/fetch-permanent-redirect-same-origin-authorization-expected.txt
A LayoutTests/http/wpt/fetch/fetch-permanent-redirect-same-origin-authorization.html
A LayoutTests/http/wpt/fetch/resources/dump-authorization-header.py
A LayoutTests/http/wpt/fetch/resources/redirect301.py
M Source/WebKit/NetworkProcess/NetworkResourceLoader.cpp
M Source/WebKit/NetworkProcess/cache/NetworkCache.cpp
M Source/WebKit/NetworkProcess/cocoa/NetworkDataTaskCocoa.mm
Log Message:
-----------
Cherry-pick 256726 at main (63145fa91bb8). https://bugs.webkit.org/show_bug.cgi?id=247418
Safari reuses Authorization header on second call to 301 redirects even if the header value changed when replaying the request
https://bugs.webkit.org/show_bug.cgi?id=247418
rdar://problem/101935060
Reviewed by Chris Dumez.
In case of serving a redirection from HTTP cache, we cannot reuse the Authorization header of the past redirect request.
Instead, we should reuse the latest request Authorization header.
To do this, we remove Authorization header from stored redirected requests in network cache.
This handles the case of a cached redirection stored with authorization and reused without authorization.
We also move the NetworkDataTaskCocoa code used to reuse the original request Authorization header to NetworkResourceLoader,
so that it handles both the cache code path as well as the regular network code path.
* LayoutTests/http/wpt/fetch/fetch-permanent-redirect-same-origin-authorization-expected.txt: Added.
* LayoutTests/http/wpt/fetch/fetch-permanent-redirect-same-origin-authorization.html: Added.
* LayoutTests/http/wpt/fetch/resources/dump-authorization-header.py: Added.
(main):
* LayoutTests/http/wpt/fetch/resources/redirect301.py: Added.
(main):
* Source/WebKit/NetworkProcess/NetworkResourceLoader.cpp:
(WebKit::NetworkResourceLoader::willSendRedirectedRequestInternal):
* Source/WebKit/NetworkProcess/cache/NetworkCache.cpp:
(WebKit::NetworkCache::Cache::makeRedirectEntry):
* Source/WebKit/NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::NetworkDataTaskCocoa::willPerformHTTPRedirection):
Canonical link: https://commits.webkit.org/256726@main
Commit: 217bcc4bcc170dcabfb83c346518cf7b6c2d4574
https://github.com/WebKit/WebKit/commit/217bcc4bcc170dcabfb83c346518cf7b6c2d4574
Author: Youenn Fablet <youennf at gmail.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
A LayoutTests/http/wpt/service-workers/fetch-service-worker-preload-use-download-and-clone.https-expected.txt
A LayoutTests/http/wpt/service-workers/fetch-service-worker-preload-use-download-and-clone.https.html
A LayoutTests/http/wpt/service-workers/fetch-service-worker-preload-use-download.https-expected.txt
A LayoutTests/http/wpt/service-workers/fetch-service-worker-preload-use-download.https.html
M LayoutTests/http/wpt/service-workers/fetch-service-worker-preload-worker.js
M LayoutTests/platform/glib/TestExpectations
A LayoutTests/platform/ios-wk2/http/wpt/service-workers/fetch-service-worker-preload-use-download-and-clone.https-expected.txt
A LayoutTests/platform/ios-wk2/http/wpt/service-workers/fetch-service-worker-preload-use-download.https-expected.txt
M Source/WebKit/NetworkProcess/ServiceWorker/ServiceWorkerFetchTask.cpp
Log Message:
-----------
Cherry-pick 256731 at main (daf8b0ba46a8). https://bugs.webkit.org/show_bug.cgi?id=247885
[Safari 16 regression] service worker download does not work in case of preloads
https://bugs.webkit.org/show_bug.cgi?id=247885
rdar://problem/102312723
Reviewed by Alex Christensen and Geoffrey Garen.
When a fetch is converted to download and is served from a service worker that uses a preload, we should use that preload.
* LayoutTests/http/wpt/service-workers/fetch-service-worker-preload-use-download.https-expected.txt: Added.
* LayoutTests/http/wpt/service-workers/fetch-service-worker-preload-use-download.https.html: Added.
* LayoutTests/http/wpt/service-workers/fetch-service-worker-preload-use-download-and-clone.https-expected.txt: Added.
* LayoutTests/http/wpt/service-workers/fetch-service-worker-preload-use-download-and-clone.https.html: Added.
* LayoutTests/http/wpt/service-workers/fetch-service-worker-preload-worker.js:
(event.event.request.url.includes): Deleted.
* LayoutTests/platform/glib/TestExpectations:
* LayoutTests/platform/ios-wk2/http/wpt/service-workers/fetch-service-worker-preload-use-download-and-clone.https-expected.txt: Added.
* LayoutTests/platform/ios-wk2/http/wpt/service-workers/fetch-service-worker-preload-use-download.https-expected.txt: Added.
* Source/WebKit/NetworkProcess/ServiceWorker/ServiceWorkerFetchTask.cpp:
(WebKit::ServiceWorkerFetchTask::convertToDownload):
Canonical link: https://commits.webkit.org/256731@main
Commit: c1ec14318fb5446b2ffc0a1031c42dbc23ef6f95
https://github.com/WebKit/WebKit/commit/c1ec14318fb5446b2ffc0a1031c42dbc23ef6f95
Author: Miguel Salinas <miguel_salinas at apple.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
A LayoutTests/editing/execCommand/indent-user-select-all-blockquotes-expected.txt
A LayoutTests/editing/execCommand/indent-user-select-all-blockquotes.html
M Source/WebCore/editing/IndentOutdentCommand.cpp
Log Message:
-----------
Cherry-pick 256749 at main (8a344c3387b2). https://bugs.webkit.org/show_bug.cgi?id=240377
Fix debug assertion failure in IndentOutdentCommand::indentIntoBlockquote
https://bugs.webkit.org/show_bug.cgi?id=240377
rdar://93236442
Reviewed by Ryosuke Niwa.
When indenting an element into a new uneditable blockquote we end up
replacing the element with an empty blockquote instead of indenting.
This also fails an assert in debug builds.
We should just bail out instead of trying to indent if the blockquote
is not editable.
* LayoutTests/editing/execCommand/indent-user-select-all-blockquotes-expected.txt: Added.
* LayoutTests/editing/execCommand/indent-user-select-all-blockquotes.html: Added.
* Source/WebCore/editing/IndentOutdentCommand.cpp:
(WebCore::IndentOutdentCommand::indentIntoBlockquote):
Canonical link: https://commits.webkit.org/256749@main
Commit: 91e6fd8e1031597f6d3c8fc38555f7cf36ddbbea
https://github.com/WebKit/WebKit/commit/91e6fd8e1031597f6d3c8fc38555f7cf36ddbbea
Author: Carlos Garcia Campos <cgarcia at igalia.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M Source/WebCore/platform/graphics/gstreamer/WebKitWebSourceGStreamer.cpp
Log Message:
-----------
Cherry-pick 256829 at main (10077398852e). https://bugs.webkit.org/show_bug.cgi?id=248033
[GStreamer] Wrong first party for cookies set on every media request
https://bugs.webkit.org/show_bug.cgi?id=248033
Reviewed by Philippe Normand.
We are always setting the requested url as the first party for cookies,
instead of the document URL. Nowadays we don't need to set the first
party for cookies in the request from the media backend.
* Source/WebCore/platform/graphics/gstreamer/WebKitWebSourceGStreamer.cpp:
(webKitWebSrcMakeRequest):
Canonical link: https://commits.webkit.org/256829@main
Commit: 3c63ca2eb6e4070a8cb442197c5d510b12846e71
https://github.com/WebKit/WebKit/commit/3c63ca2eb6e4070a8cb442197c5d510b12846e71
Author: Ahmad Saleem <ahmad.saleem792+github at gmail.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M Source/WebCore/accessibility/AccessibilityMenuList.cpp
Log Message:
-----------
Cherry-pick 256814 at main (c441a937ca2a). https://bugs.webkit.org/show_bug.cgi?id=143188
Add missing null checks to AccessibilityMenuList.cpp
https://bugs.webkit.org/show_bug.cgi?id=143188
Reviewed by Chris Fleizach.
Merge - https://src.chromium.org/viewvc/blink?revision=193917&view=revision
This patch adds null check to press() and isCollapsed() functions to avoid assertion failure crash.
* Source/WebCore/accessibility/AccessibilityMenuList.cpp:
(AccessibilityMenuList::press): Add null check for m_renderer
(AccessibilityMenuList::isCollapsed): Ditto and also comment to explain rationale
Canonical link: https://commits.webkit.org/256814@main
Commit: df1502e14f206edb972043bfb253d5a8a5a1f951
https://github.com/WebKit/WebKit/commit/df1502e14f206edb972043bfb253d5a8a5a1f951
Author: Ryan Reno <rreno at apple.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
A LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/gzip-body.any-expected.txt
A LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/gzip-body.any.html
A LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/gzip-body.any.js
A LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/gzip-body.any.serviceworker-expected.txt
A LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/gzip-body.any.serviceworker.html
A LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/gzip-body.any.sharedworker-expected.txt
A LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/gzip-body.any.sharedworker.html
A LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/gzip-body.any.worker-expected.txt
A LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/gzip-body.any.worker.html
A LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/resources/foo.octetstream.gz
A LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/resources/foo.octetstream.gz.headers
A LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/resources/foo.text.gz
A LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/resources/foo.text.gz.headers
M Source/WebCore/Modules/fetch/FetchLoader.cpp
M Source/WebCore/loader/ResourceLoader.cpp
M Source/WebCore/loader/ResourceLoader.h
M Source/WebCore/loader/ResourceLoaderOptions.h
M Source/WebCore/loader/ThreadableLoader.cpp
M Source/WebCore/platform/network/BlobResourceHandle.cpp
M Source/WebCore/platform/network/ResourceHandle.cpp
M Source/WebCore/platform/network/ResourceHandle.h
M Source/WebCore/platform/network/ResourceHandleInternal.h
M Source/WebCore/platform/network/cf/ResourceHandleCFNet.cpp
M Source/WebCore/platform/network/curl/ResourceHandleCurl.cpp
M Source/WebCore/platform/network/mac/ResourceHandleMac.mm
M Source/WebCore/xml/XMLHttpRequest.cpp
M Source/WebKit/NetworkProcess/NetworkLoadParameters.h
M Source/WebKit/NetworkProcess/ServiceWorker/ServiceWorkerSoftUpdateLoader.cpp
M Source/WebKit/NetworkProcess/cache/NetworkCacheSpeculativeLoad.cpp
M Source/WebKit/NetworkProcess/cache/NetworkCacheSpeculativeLoadManager.cpp
M Source/WebKit/NetworkProcess/cocoa/NetworkDataTaskCocoa.h
M Source/WebKit/NetworkProcess/cocoa/NetworkDataTaskCocoa.mm
M Source/WebKit/WebProcess/Network/WebLoaderStrategy.cpp
M Source/WebKitLegacy/WebCoreSupport/PingHandle.h
Log Message:
-----------
Cherry-pick 256755 at main (ec8ff55e6568). https://bugs.webkit.org/show_bug.cgi?id=247421
Content downloaded with fetch() API when Content-Encoding: gzip is set is not decompressed
https://bugs.webkit.org/show_bug.cgi?id=247421
rdar://101935292
Reviewed by Brent Fulgham.
In 195247 at main we adopted CFNetwork SPI to opt-out of content encoding sniffing for XMLHttpRequest.
Content encoding sniffing would cause responses with `Content-Encoding: gzip` and
`Content-Type: application/octet-stream` to be treated as `Content-Type: application/gzip` and
`Content-Encoding: identity` on macOS. The user and developer visible behavior is that the gzipped data
is not decompressed.
This brings that change forward to Fetch requests by setting the ContentEncodingSniffingPolicy flag
in the FetchLoader. As an additional step this also renames the ContentEncodingSniffingPolicy flags to
Default and Disable and replaces raw bools representing those flags with the enum itself.
On iOS the flag has no effect since the default behavior is to opt-out of sniffing.
* LayoutTests/platform/mac-wk1/TestExpectations:
Skip new worker tests since workers are unsupported in WK1.
* LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/gzip-body.any-expected.txt: Added.
* LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/gzip-body.any.html: Added.
* LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/gzip-body.any.js: Added.
(string_appeared_here.forEach.contentType.promise_test.async t):
* LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/gzip-body.any.serviceworker-expected.txt: Added.
* LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/gzip-body.any.serviceworker.html: Added.
* LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/gzip-body.any.sharedworker-expected.txt: Added.
* LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/gzip-body.any.sharedworker.html: Added.
* LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/gzip-body.any.worker-expected.txt: Added.
* LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/gzip-body.any.worker.html: Added.
* LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/resources/foo.octetstream.gz: Added.
* LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/resources/foo.octetstream.gz.headers: Added.
* LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/resources/foo.text.gz: Added.
* LayoutTests/imported/w3c/web-platform-tests/fetch/content-encoding/resources/foo.text.gz.headers: Added.
* Source/WebCore/Modules/fetch/FetchLoader.cpp:
(WebCore::FetchLoader::start):
* Source/WebCore/loader/ResourceLoader.cpp:
(WebCore::ResourceLoader::start):
* Source/WebCore/loader/ResourceLoader.h:
(WebCore::ResourceLoader::contentEncodingSniffingPolicy const):
(WebCore::ResourceLoader::shouldSniffContentEncoding const): Deleted.
* Source/WebCore/loader/ResourceLoaderOptions.h:
(WebCore::ResourceLoaderOptions::ResourceLoaderOptions):
* Source/WebCore/loader/ThreadableLoader.cpp:
(WebCore::ThreadableLoaderOptions::isolatedCopy const):
* Source/WebCore/platform/network/BlobResourceHandle.cpp:
(WebCore::BlobResourceHandle::BlobResourceHandle):
* Source/WebCore/platform/network/ResourceHandle.cpp:
(WebCore::ResourceHandle::ResourceHandle):
(WebCore::ResourceHandle::create):
(WebCore::ResourceHandle::contentEncodingSniffingPolicy const):
(WebCore::ResourceHandle::shouldContentEncodingSniff const): Deleted.
* Source/WebCore/platform/network/ResourceHandle.h:
* Source/WebCore/platform/network/ResourceHandleInternal.h:
(WebCore::ResourceHandleInternal::ResourceHandleInternal):
* Source/WebCore/platform/network/cf/ResourceHandleCFNet.cpp:
(WebCore::ResourceHandle::createCFURLConnection):
(WebCore::ResourceHandle::start):
(WebCore::ResourceHandle::platformLoadResourceSynchronously):
* Source/WebCore/platform/network/curl/ResourceHandleCurl.cpp:
(WebCore::ResourceHandle::platformLoadResourceSynchronously):
* Source/WebCore/platform/network/mac/ResourceHandleMac.mm:
(WebCore::ResourceHandle::applySniffingPoliciesIfNeeded):
(WebCore::ResourceHandle::createNSURLConnection):
(WebCore::ResourceHandle::start):
(WebCore::ResourceHandle::platformLoadResourceSynchronously):
* Source/WebCore/xml/XMLHttpRequest.cpp:
(WebCore::XMLHttpRequest::createRequest):
* Source/WebKit/NetworkProcess/NetworkLoadParameters.h:
* Source/WebKit/NetworkProcess/ServiceWorker/ServiceWorkerSoftUpdateLoader.cpp:
(WebKit::ServiceWorkerSoftUpdateLoader::loadFromNetwork):
* Source/WebKit/NetworkProcess/cache/NetworkCacheSpeculativeLoad.cpp:
(WebKit::NetworkCache::SpeculativeLoad::SpeculativeLoad):
* Source/WebKit/NetworkProcess/cache/NetworkCacheSpeculativeLoadManager.cpp:
(WebKit::NetworkCache::SpeculativeLoadManager::preconnectForSubresource):
* Source/WebKit/NetworkProcess/cocoa/NetworkDataTaskCocoa.h:
* Source/WebKit/NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::NetworkDataTaskCocoa::applySniffingPoliciesAndBindRequestToInferfaceIfNeeded):
(WebKit::NetworkDataTaskCocoa::NetworkDataTaskCocoa):
* Source/WebKit/WebProcess/Network/WebLoaderStrategy.cpp:
(WebKit::WebLoaderStrategy::scheduleLoadFromNetworkProcess):
(WebKit::WebLoaderStrategy::loadResourceSynchronously):
* Source/WebKitLegacy/WebCoreSupport/PingHandle.h:
Canonical link: https://commits.webkit.org/256755@main
Commit: 4c71bb11107451fa241a0e1abdc6074b64ba6bb6
https://github.com/WebKit/WebKit/commit/4c71bb11107451fa241a0e1abdc6074b64ba6bb6
Author: Yousuke Kimoto <yousuke.kimoto at sony.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
A LayoutTests/media/media-source/media-source-no-seek-with-infinite-duration-expected.txt
A LayoutTests/media/media-source/media-source-no-seek-with-infinite-duration.html
M Source/WebCore/html/HTMLMediaElement.cpp
Log Message:
-----------
Cherry-pick 256759 at main (ab539817a721). https://bugs.webkit.org/show_bug.cgi?id=247602
[MSE] Should not seek with no seekable range
https://bugs.webkit.org/show_bug.cgi?id=247602
Reviewed by Jer Noble.
According https://html.spec.whatwg.org/multipage/media.html#seeking,
"If there are no ranges given in the seekable attribute then set
the seeking IDL attribute to false and return."
For live streaming videos, their duration is infinity, then there
is no seekable range. So 'seek' should be aborted in this case.
* LayoutTests/media/media-source/media-source-no-seek-with-infinite-duration-expected.txt: Added.
* LayoutTests/media/media-source/media-source-no-seek-with-infinite-duration.html: Test seeking
status and currentTime with an infinite duration.
* Source/WebCore/html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::seekTask): Added a condtion to check if there is no seekable range
for for notifying the mdedia engine of a seek.
Canonical link: https://commits.webkit.org/256759@main
Commit: 38353ace4e78b2858db5472ee7ae229ecc291355
https://github.com/WebKit/WebKit/commit/38353ace4e78b2858db5472ee7ae229ecc291355
Author: Xabier Rodriguez-Calvar <calvaris at igalia.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M Source/WebCore/html/HTMLMediaElement.cpp
M Source/WebCore/platform/encryptedmedia/CDMProxy.h
M Source/WebCore/platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp
Log Message:
-----------
Cherry-pick 256823 at main (9ba628e9a29d). https://bugs.webkit.org/show_bug.cgi?id=247977
[EME] CDMInstanceProxy: Use weak ptr for keeping player
https://bugs.webkit.org/show_bug.cgi?id=247977
Reviewed by Philippe Normand.
Use WeakPtr for keeping media player ptr as it can be used after player is destroyed. The media player can be removed by
HTMLMediaElement::load() request while still waiting for decryption key. CDM is not aware of such case and tries to use
m_player ptr anyway that resuts with crash.
Also reset m_playbackBlockedWaitingForKey when calling load() as the player will be re-created.
Patch by Andrzej Surdej <Andrzej_Surdej at comcast.com> coming from
https://github.com/WebPlatformForEmbedded/WPEWebKit/pull/900.
* Source/WebCore/html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::prepareForLoad):
* Source/WebCore/platform/encryptedmedia/CDMProxy.h:
(WebCore::CDMInstanceProxy::setPlayer):
* Source/WebCore/platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
(WebCore::MediaPlayerPrivateGStreamer::cdmInstanceAttached):
Canonical link: https://commits.webkit.org/256823@main
Commit: 512f8ed9f32ee5791f8fbd18ec1f6c0a6437fad4
https://github.com/WebKit/WebKit/commit/512f8ed9f32ee5791f8fbd18ec1f6c0a6437fad4
Author: Brandon Stewart <brandonstewart at apple.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M Source/WebCore/dom/Document.cpp
M Source/WebCore/dom/Document.h
M Source/WebCore/html/HTMLImageElement.idl
M Source/WebCore/html/parser/HTMLPreloadScanner.cpp
M Source/WebCore/loader/ImageLoader.cpp
M Source/WebCore/page/Quirks.cpp
M Source/WebCore/page/Quirks.h
Log Message:
-----------
Cherry-pick 256854 at main (5e60d4e7a72a). https://bugs.webkit.org/show_bug.cgi?id=248048
Images fail to load periodically on ikea.com
https://bugs.webkit.org/show_bug.cgi?id=248048
rdar://101657305
Reviewed by Chris Dumez.
Adding a quirk to address images showing up as grey boxes on ikea.com.
Disabling lazy image loading addresses an issue that was found
in gatsby. This issue has been addressed in a future version of gatsby.
* Source/WebCore/dom/Document.cpp:
(WebCore::Document::lazyImageLoadingEnabled const):
* Source/WebCore/dom/Document.h:
* Source/WebCore/html/parser/HTMLPreloadScanner.cpp:
(WebCore::TokenPreloadScanner::StartTagScanner::processAttribute):
* Source/WebCore/loader/ImageLoader.cpp:
(WebCore::ImageLoader::updateFromElement):
* Source/WebCore/page/Quirks.cpp:
(WebCore::Quirks::shouldExposeShowModalDialog const):
(WebCore::Quirks::shouldDisableLazyImageLoading const):
* Source/WebCore/page/Quirks.h:
Canonical link: https://commits.webkit.org/256854@main
Commit: e1b9ac602d318c63af91e72931af3991cf33d0a6
https://github.com/WebKit/WebKit/commit/e1b9ac602d318c63af91e72931af3991cf33d0a6
Author: Sihui Liu <sihui_liu at apple.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M Source/WebKit/NetworkProcess/storage/NetworkStorageManager.cpp
M Source/WebKit/NetworkProcess/storage/NetworkStorageManager.h
M Source/WebKit/NetworkProcess/storage/NetworkStorageManager.messages.in
M Source/WebKit/WebProcess/WebStorage/StorageAreaMap.cpp
M Source/WebKit/WebProcess/WebStorage/StorageAreaMap.h
Log Message:
-----------
Cherry-pick 256863 at main (5bc131992783). https://bugs.webkit.org/show_bug.cgi?id=245479
REGRESSION (iOS 16): LocalStorage is cleared after quota error
https://bugs.webkit.org/show_bug.cgi?id=245479
rdar://problem/100252023
Reviewed by Chris Dumez.
StorageAreaMap should sync with backend again instead of clearing map when an error occurs in backend. Also, all
synchronization should not remove pending changes from map, because backend will handle the requests and StorageAreaMap
will receive reply for pending requests after that. Until StorageAreaMap receives error in reply, it should assume
the request succeeds and report value in pending requests to client.
* Source/WebKit/NetworkProcess/storage/NetworkStorageManager.cpp:
(WebKit::NetworkStorageManager::setItem):
* Source/WebKit/NetworkProcess/storage/NetworkStorageManager.h:
* Source/WebKit/NetworkProcess/storage/NetworkStorageManager.messages.in:
* Source/WebKit/WebProcess/WebStorage/StorageAreaMap.cpp:
(WebKit::StorageAreaMap::setItem):
(WebKit::StorageAreaMap::clear):
(WebKit::StorageAreaMap::ensureMap):
(WebKit::StorageAreaMap::didSetItem):
(WebKit::StorageAreaMap::applyChange):
(WebKit::StorageAreaMap::clearCache):
(WebKit::StorageAreaMap::disconnect):
(WebKit::StorageAreaMap::syncOneItem):
(WebKit::StorageAreaMap::syncItems):
(WebKit::StorageAreaMap::resetValues): Deleted.
(WebKit::StorageAreaMap::shouldApplyChangeForKey const): Deleted.
* Source/WebKit/WebProcess/WebStorage/StorageAreaMap.h:
Canonical link: https://commits.webkit.org/256863@main
Commit: 076cbda52cb8f58d526d953ca203446fae672fbb
https://github.com/WebKit/WebKit/commit/076cbda52cb8f58d526d953ca203446fae672fbb
Author: Simon Fraser <simon.fraser at apple.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
A LayoutTests/css3/filters/change-filter-style-expected.html
A LayoutTests/css3/filters/change-filter-style.html
M Source/WebCore/platform/graphics/filters/FilterEffect.cpp
M Source/WebCore/platform/graphics/filters/FilterFunction.cpp
M Source/WebCore/platform/graphics/filters/FilterFunction.h
M Source/WebCore/rendering/CSSFilter.cpp
M Source/WebCore/rendering/RenderLayer.cpp
M Source/WebCore/rendering/RenderLayer.h
Log Message:
-----------
Cherry-pick 256920 at main (a0b92719854d). https://bugs.webkit.org/show_bug.cgi?id=247233
REGRESSION (249229 at main): CSS filter does not update on hover
https://bugs.webkit.org/show_bug.cgi?id=247233
rdar://101836748
Reviewed by Darin Adler.
249229 at main made it so that we don't recreate CSSFilter on every paint,
but there was no code to update the filters when style changed.
So have `RenderLayer::updateFiltersAfterStyleChange()` clear the filters
when filter style changes; we'll recreate them on the next paint.
Also improve some logging output that helped when debugging this.
* LayoutTests/css3/filters/change-filter-style-expected.html: Added.
* LayoutTests/css3/filters/change-filter-style.html: Added.
* Source/WebCore/platform/graphics/filters/FilterEffect.cpp:
(WebCore::FilterEffect::apply):
* Source/WebCore/platform/graphics/filters/FilterFunction.cpp:
(WebCore::operator<<):
* Source/WebCore/platform/graphics/filters/FilterFunction.h:
* Source/WebCore/rendering/CSSFilter.cpp:
(WebCore::CSSFilter::create):
* Source/WebCore/rendering/RenderLayer.cpp:
(WebCore::RenderLayer::calculateClipRects const):
* Source/WebCore/rendering/RenderLayer.h:
Canonical link: https://commits.webkit.org/256920@main
Commit: b5e87e203f41c8eaa0d47925d4766f298000fd38
https://github.com/WebKit/WebKit/commit/b5e87e203f41c8eaa0d47925d4766f298000fd38
Author: Victor Jaquez <vjaquez at igalia.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M Source/WebCore/platform/gamepad/wpe/WPEGamepadProvider.cpp
M Source/WebCore/platform/gamepad/wpe/WPEGamepadProvider.h
Log Message:
-----------
Cherry-pick 256940 at main (a6099cec8ff2). https://bugs.webkit.org/show_bug.cgi?id=248210
[WPE] gamepad crashes on raspberry pi 4 (arm64)
https://bugs.webkit.org/show_bug.cgi?id=248210
Reviewed by Michael Catanzaro.
This is a breadcrumbs of the initial merge, because in libwpe uintptr_t is used for
gamepad id, while in webkit's platform code unsigned is used.
* Source/WebCore/platform/gamepad/libwpe/GamepadProviderLibWPE.cpp:
(WebCore::GamepadProviderLibWPE::gamepadConnected):
(WebCore::GamepadProviderLibWPE::gamepadDisconnected):
(WebCore::GamepadProviderLibWPE::removeGamepadForId):
* Source/WebCore/platform/gamepad/libwpe/GamepadProviderLibWPE.h:
Canonical link: https://commits.webkit.org/256940@main
Commit: 3f3eeb50d555875f8d41eeaff55ccce1f014821b
https://github.com/WebKit/WebKit/commit/3f3eeb50d555875f8d41eeaff55ccce1f014821b
Author: Xabier Rodriguez-Calvar <calvaris at igalia.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
A LayoutTests/media/media-source/media-source-append-buffer-full-evict-prior-to-end-expected.txt
A LayoutTests/media/media-source/media-source-append-buffer-full-evict-prior-to-end.html
M Source/WebCore/platform/graphics/SourceBufferPrivate.cpp
M Source/WebCore/platform/graphics/SourceBufferPrivate.h
Log Message:
-----------
Cherry-pick 256939 at main (6d7afcf0850f). https://bugs.webkit.org/show_bug.cgi?id=247052
[MSE] Ensure we are not removing the current time during automatic eviction
https://bugs.webkit.org/show_bug.cgi?id=247052
Reviewed by Alicia Boya Garcia.
In certain cases, depending on the sync frame positions, we could end up removing the current time, which would lead to
playback stopping. Now we ensure that the previous sync frame to current time is taking into account when calculating
the maximumRangeEnd.
* LayoutTests/media/media-source/media-source-append-buffer-full-evict-prior-to-end-expected.txt: Added.
* LayoutTests/media/media-source/media-source-append-buffer-full-evict-prior-to-end.html: Added.
* Source/WebCore/platform/graphics/SourceBufferPrivate.cpp:
(WebCore::SourceBufferPrivate::findPreviousSyncSamplePresentationTime):
(WebCore::SourceBufferPrivate::evictCodedFrames):
* Source/WebCore/platform/graphics/SourceBufferPrivate.h:
Canonical link: https://commits.webkit.org/256939@main
Commit: 63253483cd07f9fe6880dec63f2200b11ae407e2
https://github.com/WebKit/WebKit/commit/63253483cd07f9fe6880dec63f2200b11ae407e2
Author: Adrian Perez de Castro <aperez at igalia.com>
Date: 2023-01-25 (Wed, 25 Jan 2023)
Changed paths:
M Source/WebCore/platform/audio/glib/MediaSessionManagerGLib.cpp
Log Message:
-----------
Unreviewed non-uinified build fixes
* Source/WebCore/platform/audio/glib/MediaSessionManagerGLib.cpp: Add
missing wtf/glib/GUniquePtr.h header inclusion.
Compare: https://github.com/WebKit/WebKit/compare/b8f565c7a7c7...63253483cd07
More information about the webkit-changes
mailing list