Branch: refs/heads/webkitglib/2.54
Home: https://github.com/WebKit/WebKit
Commit: 6d153629a66a4852cca65326a148c7a0303ea6f4
https://github.com/WebKit/WebKit/commit/6d153629a66a4852cca65326a148c7a0303ea6f4
Author: Przemyslaw Gorszkowski <[email protected]>
Date: 2026-09-10 (Thu, 10 Sep 2026)
Changed paths:
M Source/WebCore/platform/gamepad/manette/ManetteGamepad.cpp
M Source/WebKit/WPEPlatform/wpe/WPEGamepadManette.cpp
Log Message:
-----------
Cherry-pick 320820@main (9f1b3cc75649).
https://bugs.webkit.org/show_bug.cgi?id=323663
[WPE][GTK] rumble does not work in case of libmanette 0.2.13
https://bugs.webkit.org/show_bug.cgi?id=323663
Reviewed by Carlos Garcia Campos.
Macros from libmanette: LIBMANETTE_CHECK_VERSION(deprecated) and
MANETTE_CHECK_VERSION evaluate to TRUE in case of greater than or
equal to. This caused that in case of version 0.2.13 wrong values
were
passed
to
manette_device_rumble (normalized floating-point
rumble magnitudes) and it was rounded to 0 (guint16).
The normalized floating-point rumble magnitudes are demanded in
case of libmanette version >= 1.0.0.
Also deprecated macro is replaced with the new one.
No new tests.
* Source/WebCore/platform/gamepad/manette/ManetteGamepad.cpp:
(WebCore::ManetteGamepad::startRumble):
* Source/WebKit/WPEPlatform/wpe/WPEGamepadManette.cpp:
(wpeGamepadManetteRumble):
Canonical link:
https://flagged.apple.com:443/proxy?t2=dp9u2T7cb8&o=aHR0cHM6Ly9jb21taXRzLndlYmtpdC5vcmcvMzIwODIwQG1haW4=&emid=80847029-bc1b-41bf-8729-5a042957ea34&c=11
Canonical link:
https://flagged.apple.com:443/proxy?t2=Db1W9c4Os9&o=aHR0cHM6Ly9jb21taXRzLndlYmtpdC5vcmcvMzE3Njk1LjI1OEB3ZWJraXRnbGliLzIuNTQ=&emid=80847029-bc1b-41bf-8729-5a042957ea34&c=11
Commit: 1dcc8a1448b96429478363ad0dc6929d6dc76511
https://github.com/WebKit/WebKit/commit/1dcc8a1448b96429478363ad0dc6929d6dc76511
Author: Claudio Saavedra <[email protected]>
Date: 2026-09-10 (Thu, 10
Sep
2026)
Changed paths:
M Tools/Scripts/webkitpy/webdriver_tests/pytest_runner.py
Log Message:
-----------
Cherry-pick 320815@main
(92641830324a). https://bugs.webkit.org/show_bug.cgi?id=322171
[WebDriver] pytest-timeout's SIGALRM can wedge the asyncio loop and hang
the run
https://bugs.webkit.org/show_bug.cgi?id=322171
Reviewed by Carlos Alberto Lopez Perez.
pytest-timeout raises from its SIGALRM handler wherever the alarm lands.
When
that is inside asyncio's scheduler, for instance while Future.set_result()
is
queueing the wake-up of the task awaiting it, the future is marked done but
the
wake-up is never scheduled. asyncio swallows the exception as "Exception in
callback", the test task is parked for good, and the loop keeps servicing
the
websockets
keepalive until someone kills the bot. Every run of the BiDi tests
expected to time out goes through this path.
Wrap the handler pytest-timeout
installs so
that,
while an event loop is
running, the alarm is only delivered from the selector the idle loop is
blocked
in, where the exception
propagates cleanly out of run_until_complete(). Anywhere
else inside the loop it is re-armed a few milliseconds later instead.
Also match the timeout message pytest-timeout 2.4.0 produces, which has been
making unexpected timeouts show up as failures since the bump.
* Tools/Scripts/webkitpy/webdriver_tests/pytest_runner.py:
(SubtestResultRecorder._was_timeout):
(TimeoutSignalHandler):
(TimeoutSignalHandler.pytest_timeout_set_timer):
(TimeoutSignalHandler._should_defer_timeout):
(run):
Canonical
link:
https://flagged.apple.com:443/proxy?t2=de4s8x8tM7&o=aHR0cHM6Ly9jb21taXRzLndlYmtpdC5vcmcvMzIwODE1QG1haW4=&emid=80847029-bc1b-41bf-8729-5a042957ea34&c=11
Canonical
link:
https://flagged.apple.com:443/proxy?t2=Dx5Y7Z6FM6&o=aHR0cHM6Ly9jb21taXRzLndlYmtpdC5vcmcvMzE3Njk1LjI1OUB3ZWJraXRnbGliLzIuNTQ=&emid=80847029-bc1b-41bf-8729-5a042957ea34&c=11
Commit: 70bcd8b658c3a6bd99fbb411e9ec4ab4e0a73dd2
https://github.com/WebKit/WebKit/commit/70bcd8b658c3a6bd99fbb411e9ec4ab4e0a73dd2
Author: Devin Rousso <[email protected]>
Date:
2026-09-10 (Thu, 10 Sep 2026)
Changed paths:
A LayoutTests/compositing/filters/repeated-filter-transitions-expected.txt
A LayoutTests/compositing/filters/repeated-filter-transitions.html
M Source/WebCore/platform/graphics/skia/SkiaCompositingLayer.cpp
Log Message:
-----------
Cherry-pick 320824@main (361bf1bd5adf).
https://bugs.webkit.org/show_bug.cgi?id=323720
[WPE] SkiaCompositingLayer: SIGSEGV
in
isColorFilterNode
https://bugs.webkit.org/show_bug.cgi?id=323720
Reviewed by Carlos Garcia Campos.
Add a check for the underlying `sk_sp<SkImageFilter>` before
using
it.
* Source/WebCore/platform/graphics/skia/SkiaCompositingLayer.cpp:
(WebCore::SkiaCompositingLayer::paintWithFilterAndMask):
* LayoutTests/compositing/filters/repeated-filter-transitions.html: Added.
*
LayoutTests/compositing/filters/repeated-filter-transitions-expected.txt: Added.
Canonical link:
https://flagged.apple.com:443/proxy?t2=dV6K6g2jl4&o=aHR0cHM6Ly9jb21taXRzLndlYmtpdC5vcmcvMzIwODI0QG1haW4=&emid=80847029-bc1b-41bf-8729-5a042957ea34&c=11
Canonical link:
https://flagged.apple.com:443/proxy?t2=dt0s0w2ws3&o=aHR0cHM6Ly9jb21taXRzLndlYmtpdC5vcmcvMzE3Njk1LjI2MEB3ZWJraXRnbGliLzIuNTQ=&emid=80847029-bc1b-41bf-8729-5a042957ea34&c=11
Compare: https://github.com/WebKit/WebKit/compare/4d7525e156e1...70bcd8b658c3
To unsubscribe from these emails, change your
notification settings at https://github.com/WebKit/WebKit/settings/notifications
