Branch: refs/heads/webkitglib/2.52
  Home:   https://github.com/WebKit/WebKit
  Commit: 5d7be9bf16d02fdd249c3f3a1c4a382fc0b9750e
      
https://github.com/WebKit/WebKit/commit/5d7be9bf16d02fdd249c3f3a1c4a382fc0b9750e
  Author: Fujii Hironori <[email protected]>
  Date:   2026-05-12 (Tue, 12 May 2026)

  Changed paths:
    M LayoutTests/platform/glib/TestExpectations
    M 
Source/WebCore/platform/graphics/texmap/coordinated/GraphicsLayerCoordinated.cpp

  Log Message:
  -----------
  Cherry-pick 313063@main (ac54485f9d31). 
https://bugs.webkit.org/show_bug.cgi?id=307753

    [Coordinated Graphics] directly composited tile images were stretched 
unexpectedly
    https://bugs.webkit.org/show_bug.cgi?id=307753

    Reviewed by Carlos Garcia Campos.

    The tile phase was mistakenly set to the tile size.

    Tests: compositing/patterns/direct-pattern-compositing-position.html

    * LayoutTests/platform/glib/TestExpectations:
    * 
Source/WebCore/platform/graphics/texmap/coordinated/GraphicsLayerCoordinated.cpp:
    (WebCore::GraphicsLayerCoordinated::commitLayerChanges):

    Canonical link: https://commits.webkit.org/313063@main

Canonical link: https://commits.webkit.org/305877.491@webkitglib/2.52


  Commit: ca20ec86088c7d781ce0938edaa331f89f0dd61c
      
https://github.com/WebKit/WebKit/commit/ca20ec86088c7d781ce0938edaa331f89f0dd61c
  Author: Chris Dumez <[email protected]>
  Date:   2026-05-12 (Tue, 12 May 2026)

  Changed paths:
    M LayoutTests/platform/ios/TestExpectations
    M LayoutTests/platform/mac-wk2/TestExpectations
    M Source/WebKit/NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp

  Log Message:
  -----------
  Cherry-pick 313039@main (f08767a2f3c0). 
https://bugs.webkit.org/show_bug.cgi?id=305508

    REGRESSION(304892@main?): 
http/tests/resourceLoadStatistics/log-cross-site-load-with-link-decoration.html 
is a flakey text failure
    https://bugs.webkit.org/show_bug.cgi?id=305508
    rdar://168172161

    Reviewed by Brent Fulgham.

    304892@main defers old web process shutdown during commitProvisionalPage 
via a
    shutdownPreventingScope. This allows the old process to receive the Close 
message
    and flush its (potentially empty) resource load statistics to the network 
process.
    resourceLoadStatisticsUpdated() unconditionally calls 
processStatisticsAndDataRecords()
    even when there is nothing to merge. This processing iterates all domains 
and resets
    DataRemovalFrequency to Never for any domain that was recently flagged for 
removal,
    racing with logCrossSiteLoadWithLinkDecoration() which just set it to Short.

    The fix is to early-return when the statistics vector is empty since there 
is nothing
    to merge and no reason to trigger data records processing.

    No new tests, unskip existing test.

    * LayoutTests/platform/ios/TestExpectations:
    * LayoutTests/platform/mac-wk2/TestExpectations:
    * 
Source/WebKit/NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
    (WebKit::WebResourceLoadStatisticsStore::resourceLoadStatisticsUpdated):

    Canonical link: https://commits.webkit.org/313039@main

Canonical link: https://commits.webkit.org/305877.492@webkitglib/2.52


  Commit: 47bb36b39e7d1fc77c6142f9b290abe1b0226cba
      
https://github.com/WebKit/WebKit/commit/47bb36b39e7d1fc77c6142f9b290abe1b0226cba
  Author: Simon Fraser <[email protected]>
  Date:   2026-05-12 (Tue, 12 May 2026)

  Changed paths:
    M Source/WebCore/page/scrolling/ScrollingStateNode.h

  Log Message:
  -----------
  Cherry-pick 312950@main (bfb4c0bb6e54). 
https://bugs.webkit.org/show_bug.cgi?id=314358

    WKCompositingLayer leaks
    https://bugs.webkit.org/show_bug.cgi?id=314358
    rdar://160391523

    Reviewed by Tim Horton.

    Pages with subscrollers and some kinds of layer configuration changes
    triggered leaks of WKCompositingLayers (I was able to reproduce when
    playing YouTube shorts).

    The bug was that LayerRepresentation's assignment operator failed
    to release the old layer.

    The fix takes care to not release the old layer before retaining
    the new one, in case this is the only reference keeping the layer alive.

    A future PR will change LayerRepresentation to use a variant<>, which
    will replace this mess.

    * Source/WebCore/page/scrolling/ScrollingStateNode.h:
    (WebCore::LayerRepresentation::operator=):

    Canonical link: https://commits.webkit.org/312950@main

Canonical link: https://commits.webkit.org/305877.493@webkitglib/2.52


  Commit: 9b5c4731d8a323bf7e49ed319897fae9085b032f
      
https://github.com/WebKit/WebKit/commit/9b5c4731d8a323bf7e49ed319897fae9085b032f
  Author: Philippe Normand <[email protected]>
  Date:   2026-05-12 (Tue, 12 May 2026)

  Changed paths:
    M Source/WebCore/platform/mediastream/gstreamer/DesktopPortal.h

  Log Message:
  -----------
  Cherry-pick 312994@main (363ce7daefe0). 
https://bugs.webkit.org/show_bug.cgi?id=314511

    [GStreamer] Add a virtual DesktopPortal destructor
    https://bugs.webkit.org/show_bug.cgi?id=314511

    Reviewed by Claudio Saavedra.

    * Source/WebCore/platform/mediastream/gstreamer/DesktopPortal.h:

    Canonical link: https://commits.webkit.org/312994@main

Canonical link: https://commits.webkit.org/305877.494@webkitglib/2.52


  Commit: 6df9adc0d72ee63f56e79bada68dff1ddb7b9ca7
      
https://github.com/WebKit/WebKit/commit/6df9adc0d72ee63f56e79bada68dff1ddb7b9ca7
  Author: Philippe Normand <[email protected]>
  Date:   2026-05-12 (Tue, 12 May 2026)

  Changed paths:
    M LayoutTests/media/media-vp8-hiddenframes.html
    M Source/WebCore/platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp

  Log Message:
  -----------
  Cherry-pick 313064@main (65523b157f07). 
https://bugs.webkit.org/show_bug.cgi?id=314311

    [GStreamer] media/media-vp8-hiddenframes.html fails
    https://bugs.webkit.org/show_bug.cgi?id=314311

    Reviewed by Xabier Rodriguez-Calvar.

    Allow seeking to the exact duration. Manually verified by removing allowed 
fuzzyness on the test and
    checking the last video frame is rendered instead of the first one. Fuzzy 
parameters were then
    adjusted to accommodate with GStreamer's vpx decoder.

    The test was also flaky on mac-intel. Ensuring the video is rendered after 
the seeked event fixed
    it.

    The following test performs a seek at duration for an Ogg file, which is 
poorly supported by
    oggdemux, so keep the previous code, basically triggering EOS as was done 
before this patch, but
    specially for this scenario:

    
imported/w3c/web-platform-tests/html/semantics/embedded-content/media-elements/seeking/seek-to-max-value.htm

    * 
Source/WebCore/platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
    (WebCore::MediaPlayerPrivateGStreamer::doSeek):

    Canonical link: https://commits.webkit.org/313064@main

Canonical link: https://commits.webkit.org/305877.495@webkitglib/2.52


Compare: https://github.com/WebKit/WebKit/compare/d5043a96b5d0...6df9adc0d72e

To unsubscribe from these emails, change your notification settings at 
https://github.com/WebKit/WebKit/settings/notifications

Reply via email to