Branch: refs/heads/webkitglib/2.50
  Home:   https://github.com/WebKit/WebKit
  Commit: 3b6d5b7a3691f2c78279de70af2d9832fcbc9b90
      
https://github.com/WebKit/WebKit/commit/3b6d5b7a3691f2c78279de70af2d9832fcbc9b90
  Author: Antoine Quint <[email protected]>
  Date:   2026-01-20 (Tue, 20 Jan 2026)

  Changed paths:
    A 
LayoutTests/webanimations/threaded-animations/filter-animation-crash-expected.txt
    A LayoutTests/webanimations/threaded-animations/filter-animation-crash.html
    M Source/WebCore/platform/animation/AcceleratedEffect.cpp
    M Source/WebCore/platform/animation/AcceleratedEffect.h
    M Source/WebCore/rendering/RenderLayerBacking.cpp

  Log Message:
  -----------
  Cherry-pick 305530@main (b7aad8b68633). 
https://bugs.webkit.org/show_bug.cgi?id=305396

    [threaded-animations] merlyndesignworks.co.uk page crashes repeatedly 
(invalid message: RemoteLayerTreeDrawingAreaProxy_CommitLayerTree)
    https://bugs.webkit.org/show_bug.cgi?id=305396
    rdar://167992456

    Reviewed by Simon Fraser.

    There were several issues with our filter validation approach under 
`AcceleratedEffect::validateFilters()`:

    - while it would correctly catch the case where a `drop-shadow()` filter 
was used anywhere but
      as the last filter and mark that property as disallowed, it would not 
clear the `filter` or
      `backdropFilter` value in the keyframes,
    - disallowing a property also did not ensure that we would clear the base 
values,
    - it wouldn't either clear base values of any previous effect,
    - we wouldn't catch the case where `drop-shadow()` contained a color which 
is entirely not supported.

    We now correctly catch such situations.

    Test: webanimations/threaded-animations/filter-animation-crash.html

    * 
LayoutTests/webanimations/threaded-animations/filter-animation-crash-expected.txt:
 Added.
    * 
LayoutTests/webanimations/threaded-animations/filter-animation-crash.html: 
Added.
    * Source/WebCore/platform/animation/AcceleratedEffect.cpp:
    (WebCore::AcceleratedEffect::Keyframe::clearProperty):
    (WebCore::AcceleratedEffect::validateFilters):
    (WebCore::AcceleratedEffect::clearProperty):
    * Source/WebCore/platform/animation/AcceleratedEffect.h:
    * Source/WebCore/rendering/RenderLayerBacking.cpp:
    (WebCore::RenderLayerBacking::updateAcceleratedEffectsAndBaseValues):

    Canonical link: https://commits.webkit.org/305530@main

Canonical link: https://commits.webkit.org/298234.400@webkitglib/2.50


  Commit: 96c76b8fe5e237eaa64fedaea340b7fe58d60542
      
https://github.com/WebKit/WebKit/commit/96c76b8fe5e237eaa64fedaea340b7fe58d60542
  Author: Sihui Liu <[email protected]>
  Date:   2026-01-20 (Tue, 20 Jan 2026)

  Changed paths:
    M Source/WebCore/workers/service/server/SWServer.cpp
    M Source/WebCore/workers/service/server/SWServer.h

  Log Message:
  -----------
  Cherry-pick 305669@main (eb4391211130). 
https://bugs.webkit.org/show_bug.cgi?id=305537

    Ensure callbacks are invoked when SWServer is closed
    https://bugs.webkit.org/show_bug.cgi?id=305537
    rdar://168195990

    Reviewed by Per Arne Vollan.

    SWServer might be destroyed without invoking stored callbacks. The patch 
fixes that issue and switches to use
    CompletionHandler for callback.

    * Source/WebCore/workers/service/server/SWServer.cpp:
    (WebCore::SWServer::close):
    (WebCore::SWServer::getOriginsWithRegistrations):
    * Source/WebCore/workers/service/server/SWServer.h:

    Canonical link: https://commits.webkit.org/305669@main

Canonical link: https://commits.webkit.org/298234.401@webkitglib/2.50


  Commit: 4d937cdeb2f7bf0617953538f7e3926a77470af7
      
https://github.com/WebKit/WebKit/commit/4d937cdeb2f7bf0617953538f7e3926a77470af7
  Author: Sammy Gill <[email protected]>
  Date:   2026-01-20 (Tue, 20 Jan 2026)

  Changed paths:
    A 
LayoutTests/imported/w3c/web-platform-tests/css/css-grid/subgrid/subgrid-item-with-margin-left-auto-expected.html
    A 
LayoutTests/imported/w3c/web-platform-tests/css/css-grid/subgrid/subgrid-item-with-margin-left-auto.html
    M Source/WebCore/rendering/RenderBox.cpp
    M Source/WebCore/rendering/RenderGrid.cpp
    M Source/WebCore/rendering/RenderGrid.h

  Log Message:
  -----------
  Cherry-pick 300422@main (4a61ed0984fb). 
https://bugs.webkit.org/show_bug.cgi?id=296988

    proxmox.com: Large amounts of extra vertical spacing between content.
    https://bugs.webkit.org/show_bug.cgi?id=296988
    rdar://157638931

    Reviewed by Alan Baradlay.

    During track sizing we are supposed to treat any auto margins on grid
    items as 0. These would then be resolved after we have computed the grid
    area for the item. However, the attached testcase demonstrates that we
    are not doing this correctly in all scenarios for at least the inline
    margins.

    The content is along the lines of
    <grid>
      <subgrid columns style="margin-left: auto">
        <div> X X X X X X </div>
      </subrid columns>
    </grid>

    When we resolve the row sizes for the grid we end up calling layout() on
    the subgrid via GridTrackSizingAlgorithmStrategy::logicalHeightForGridItem.
    During the subgrid's layout while we are in updateLogicalWidth(), we end up
    computing a logical width of 0 and a resulting margin-left value of
    whatever the available space is. Then, we end up setting the subgrid's
    columns to the column sizes of the parent grid since it is subgridded in
    that dimension, but we also factor out the subgrid's incorrectly
    computed margins which are the same size as the columns from the parent.
    This ends up with column sizes of 0 for the content of the subgrid.
    Later on when we compute the rows of the subgrid, we end up setting an
    available space of 0 for its content which results in really tall
    content and really tall rows in the parent as a result.

    This patch corrects this behavior for the inline margins of grid item's
    by keeping track of a new member on RenderGrid which keeps track of
    whether or not we are in the middle of track sizing via SetForScope.
    When we are resolving inline margins for grid items we set any auto
    margins to 0 if we are in the middle of track sizing by checking for
    this state.

    Canonical link: https://commits.webkit.org/300422@main

Canonical link: https://commits.webkit.org/298234.402@webkitglib/2.50


Compare: https://github.com/WebKit/WebKit/compare/dcd7db602ed6...4d937cdeb2f7

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

Reply via email to