This was implemented in bug 1813554 to allow resizing tiled windows at
the top of the window, but since then KWin has changed behavior and no
longer allows resizing tiled windows, so this is now unnecessary.
Differential Revision: https://phabricator.services.mozilla.com/D226005
This correctly detects Adwaita if you don't have a valid gtk theme name.
I found this when testing some X11 environments for bug 1924616.
Differential Revision: https://phabricator.services.mozilla.com/D226008
This provides additional "flat/...." configurations which work with the flat config of ESLint. The changes in recommended.js make it easier to handle the two different types of configuration. The changes in index.js are to ensure that the resultant configurations are correct for legacy or flat configs. The index.js changes also inject the plugin object back into itself, as required by ESLint.
Once we have switched mozilla-central to flat configuration, we can drop support for the legacy config.
Differential Revision: https://phabricator.services.mozilla.com/D225705
This makes it easier for subscribers of eslint-plugin-mozilla to use it more generally as a set of recommended
rules. The only environment definitions are for mozilla-specific file types, where we should be safe to specify
them.
Differential Revision: https://phabricator.services.mozilla.com/D225704
Automatic update from web-platform-tests
IDB WPTs: Extend several IDB WPTs to run on workers (Part 2)
This set of IndexedDB WPTs currently only run in a window environment.
This change extends them to also run in dedicated, shared, and service
worker environments.
Bug: 41455766
Change-Id: Ib973b26a725d2b3c05e2696cee208c052e563632
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5926431
Commit-Queue: Rahul Singh <rahsin@microsoft.com>
Reviewed-by: Evan Stade <estade@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1369491}
--
wpt-commits: 477be4a28457a449c8505acede0ddbb31ab9e0e2
wpt-pr: 48629
Automatic update from web-platform-tests
[shared storage] Implement interestGroup()
Add interestGroups() to the shared storage worklet, to return the
Protected Audience interest groups associated with the
shared storage origin's owner, with some additional metadata.
Implement this behind a runtime feature, which is implicitly
controlled by a Finch flag.
Explainer PR: https://github.com/WICG/shared-storage/pull/180
Spec PR(s):
1) https://github.com/WICG/turtledove/pull/1299
2) https://github.com/WICG/shared-storage/pull/203
Bug: 367992703
Binary-Size: Size increase is unavoidable.
Fuchsia-Binary-Size: Size increase is unavoidable.
Change-Id: I5fc5767fa53a91f021d64a871a6dd9cb88f4431c
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5696046
Reviewed-by: Daniel Cheng <dcheng@chromium.org>
Commit-Queue: Yao Xiao <yaoxia@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1369483}
--
wpt-commits: eb87bbbacf996ab46607538a72ea0adaee229e7a
wpt-pr: 48650
Automatic update from web-platform-tests
Add a new canvas WPT template using pycairo-generated reference images
Using the "cairo_reference:" key in the YAML config, we can now express
the expected result of a test using pycairo code. Contrary to the
existing "expected:" key which produces an image that is only used for
"informational" purpose, "cairo_reference:" produces a reference test,
meaning that the test runner automatically compares the test result
with that generated image and fails the test if the result differs.
Both single variant tests (non-variant or tests with a single variant
per file) and variant grids are supported. For single variant, the
generated reference file is an HMTL page with a single <img> tag
pointing to the PNG generated with pycairo. To cut down on the number of
files generated for variant grids, all the reference images for all of
the cells of the grid are packed into a single PNG file,
3d-model-texture-style. The reference HTML file is a grid of <img> tags,
each framing a different portion of the PNG image.
Bug: 364549423
Change-Id: Icb3c246b22347054b7cc9b5e5cc40d21b30565bd
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5894774
Reviewed-by: Andres Ricardo Perez <andresrperez@chromium.org>
Commit-Queue: Jean-Philippe Gravel <jpgravel@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1369396}
--
wpt-commits: 2af17a84e98cd28940d2882c68ff824d6e98f4b7
wpt-pr: 48382
Automatic update from web-platform-tests
[@container] stuck queries with writing direction changes
- Fix a typo that caused stuck not to be dynamically changed with
writing direction changes.
- Add tests.
- Remove TODO as the current behavior is the spec'ed behavior.
Bug: 40268059
Change-Id: Icc0fb12eeec5735d623bc5272916b7fc92ad5e5c
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5938224
Reviewed-by: Morten Stenshorne <mstensho@chromium.org>
Commit-Queue: Rune Lillesveen <futhark@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1369380}
--
wpt-commits: d56d0a1180b71e512f87ef8fb1771298f5071f3c
wpt-pr: 48649
Automatic update from web-platform-tests
Support custom element reactions for moveBefore
- A new custom element reaction: connectedMoveCallback
- Called when the moved element is re-inserted
- Falls back to calling disconnectedMove/connectedMove if not provided.
Bug: 40150299
Change-Id: I0d3ed83ca03bbffcd2db79d1c421d8d155a4d9c8
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5920804
Reviewed-by: Dominic Farolino <dom@chromium.org>
Commit-Queue: Noam Rosenthal <nrosenthal@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1369383}
--
wpt-commits: c17fab297bdc750e97907281fabfbcbead8e993f
wpt-pr: 48648
Automatic update from web-platform-tests
Avoid wrapping line at the start of text run with `text-wrap-mode: nowrap`
When computing the min-content size of an inline formatting context,
we could allow a soft wrap opportunity at the start of a text run.
This shouldn't happen with `text-wrap-mode: nowrap`.
Signed-off-by: Oriol Brufau <obrufau@igalia.com>
--
wpt-commits: b8538c8c1cd7e6b5b7e42a618119b69b6521a17b
wpt-pr: 48627
Automatic update from web-platform-tests
Add view-transition-capture-mode CSS property
This property has the values of flat | layered, layered by
default.
- Flat is the current way we capture view transition content
- Layered is the new way, which captures box decorations and
effects as style instead of baking them into the image.
Bug: 347947051
Change-Id: If70c8436ef34d194002452447dbd2d5ebedfa387
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5935195
Commit-Queue: Noam Rosenthal <nrosenthal@chromium.org>
Reviewed-by: Khushal Sagar <khushalsagar@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1369285}
--
wpt-commits: c35d4f9d01157cbf8b49d1f9906dfbc36865a839
wpt-pr: 48643
Automatic update from web-platform-tests
sideways: Support scroll anchoring
* Store the anchor object offset as a LogicalOffset, not a block-
flipped LayoutPoint
* Add functions converting between LogicalOffset and gfx::Vector2d /
gfx::PointF / ScrollOffset.
Bug: 40501131
Change-Id: I8d32b1a2a8e9fc6ade68f12188430ac768dd14d2
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5933384
Reviewed-by: Koji Ishii <kojii@chromium.org>
Auto-Submit: Kent Tamura <tkent@chromium.org>
Commit-Queue: Koji Ishii <kojii@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1369217}
--
wpt-commits: 33cd7ece3a146fb5b2025a27ba8a6f6d3a72ac95
wpt-pr: 48639
Automatic update from web-platform-tests
webnn: Make MLOperandDescriptor.shape a required dictionary member
See https://github.com/webmachinelearning/webnn/issues/758
Also removes the temporary support for passing "dimensions" which was
added in https://crrev.com/e7e99aa5
Bug: 365813262
Cq-Include-Trybots: luci.chromium.try:mac14-blink-rel,win11-blink-rel
Change-Id: Ib714ae540da7fbd7d55365dc739bfb8dbf266406
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5850659
Reviewed-by: ningxin hu <ningxin.hu@intel.com>
Commit-Queue: ningxin hu <ningxin.hu@intel.com>
Auto-Submit: Austin Sullivan <asully@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1369192}
--
wpt-commits: c2d015181a4956b2f91061595ec7da93823ec3ec
wpt-pr: 48637
Automatic update from web-platform-tests
[Editing] Double-click selects wrong range for any element styled
as float.
The CL addresses the bug where double click on any list item
selects multiple list items when list item has floating style.
When there are multiple list items and anyone is selected using
double click, it selects all the list items along with the one
on which double click was performed. This happens because in
|ComputeInlineContentsAsBlockFlow|, float style is not considered
as text segment boundary and the selection is expanded for all.
Also, for other cases involving floating style on element, the
selection is not as expected as selecting a float element by
double click selects everything. This is fixed by removing the
condition which expands the selection when the element has a
floating style.
Firefox has the behavior similar to what we get after the fix
i.e it doesn't expand the selection for floating elements. Webkit
however, seems to have the same bug as chromium expanding the
selection for floating elements.
Bug: 40711666
Change-Id: I408b388cfd0cb5a8213ef2c433b547376ff83367
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5842631
Reviewed-by: Kent Tamura <tkent@chromium.org>
Commit-Queue: Pranav Modi <pranavmodi@microsoft.com>
Reviewed-by: Sanket Joshi <sajos@microsoft.com>
Reviewed-by: Siye Liu <siliu@microsoft.com>
Cr-Commit-Position: refs/heads/main@{#1369172}
--
wpt-commits: 721205f92ea832a0197975112a48f5355c8bfcbf
wpt-pr: 48634
Automatic update from web-platform-tests
Reland "Reland "webnn: add int4 and uint4 support for quantizeLinear and dequantizeLinear""
This is a reland of commit c80028acacc05ac6687d44d9af5c95851b88b11e
The original commit was reverted due to the crash of
WebNNGraphImplTest.Resample2dTest for android-oreo-x86-rel.
The root cause of this failure is due to the original CL introduces
GetBitsPerElement() which causes the intermediate bits number as
(34902 * 23243 * 8 * 4: the input operand dimensions [1, 1, 34902, 23343])
which is greater than maximum size_t on 32-bit system, and causes
OperandDescriptor::Create(type, dimensions) returning an unexpected
result.
The solution in this reland CL is using uint64_t instead of size_t for
intermediate calculation result. Besides, this reland CL also fixes a
minor data limited error of quantize_linear_zero_point for DirectML
backend.
Original change's description:
> Reland "webnn: add int4 and uint4 support for quantizeLinear and dequantizeLinear"
>
> This is a reland of commit 3a1707bcd20fa2fc4b1036755259ed01dc6632d5.
> The original commit was reverted due to the crash of WebNNGraphImplTest.Resample2dTest for android-oreo-x86-rel.
>
> However, I can't reproduce this crash locally and the
> android-oreo-x86-rel bot in chromium can pass now without
> changing any code.
>
> Original change's description:
> > webnn: add int4 and uint4 support for quantizeLinear and dequantizeLinear
> >
> > This CL also adds some WPT conformance tests to verify the
> > implementation.
> >
> > Bug: 40206287
> > Change-Id: Ieb8ce3ae2182388ae7cc98cd1fc8e3d2dcd9c7d2
> > Cq-Include-Trybots: luci.chromium.try:win11-blink-rel, mac14.arm64-blink-rel, mac14-blink-rel, mac15.arm64-blink-rel, mac15-blink-rel, linux-blink-rel
> > Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5861613
> > Commit-Queue: Lisha Guo <lisha.guo@intel.com>
> > Reviewed-by: ningxin hu <ningxin.hu@intel.com>
> > Reviewed-by: Alex Gough <ajgo@chromium.org>
> > Reviewed-by: Rafael Cintron <rafael.cintron@microsoft.com>
> > Cr-Commit-Position: refs/heads/main@{#1365343}
>
> Bug: 40206287
> Change-Id: I8c3b922d092214e4c56ffb5298b86c8d12baf8c0
> Cq-Include-Trybots: luci.chromium.try:win11-blink-rel, mac14.arm64-blink-rel, mac14-blink-rel, mac15.arm64-blink-rel, mac15-blink-rel, linux-blink-rel
> Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5918905
> Reviewed-by: Rafael Cintron <rafael.cintron@microsoft.com>
> Reviewed-by: ningxin hu <ningxin.hu@intel.com>
> Reviewed-by: Alex Gough <ajgo@chromium.org>
> Commit-Queue: Lisha Guo <lisha.guo@intel.com>
> Cr-Commit-Position: refs/heads/main@{#1366660}
Bug: 40206287
Change-Id: I1dab0a282f8579151387bb67b9045129186215d1
Cq-Include-Trybots: luci.chromium.try:win11-blink-rel, mac14.arm64-blink-rel, mac14-blink-rel, mac15.arm64-blink-rel, mac15-blink-rel, linux-blink-rel
Include-Ci-Only-Tests: true
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5922495
Reviewed-by: Alex Gough <ajgo@chromium.org>
Commit-Queue: Lisha Guo <lisha.guo@intel.com>
Reviewed-by: ningxin hu <ningxin.hu@intel.com>
Cr-Commit-Position: refs/heads/main@{#1369144}
--
wpt-commits: a7781e368543be6c8dea73a1bcba3d5065d29eec
wpt-pr: 48636
Automatic update from web-platform-tests
Fix use of color channel keywords in calc() division denominators
The calc expression parser was enforcing a restriction that the
denominator must be a number, and depending on this restriction to
simplify divisions at parse time. This restriction was specified in
css-values-3 but relaxed in css-values-4.
Prior to crrev.com/c/5825542, when a color channel keyword was used
in a calc() expression, we would eagerly substitute in the numeric value
as soon as we encountered the keyword token. As a result, the parsed
expression would satisfy the restriction and the expression would be
accepted. As of that change, we now do the value substitution later in
order to preserve the originally specified form.
The fix is to relax the type restriction in code, which necessitates
implementing support for Invert nodes in calculation expression trees
as specified in css-values-4.
This CL also re-enables CSSRelativeColorLateResolveAlways, which had
been disabled in crrev.com/c/5912180 due to this bug.
Bug: 371224775
Change-Id: Ia9b08030a08a64023b166abbd332e8c6db9043d0
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5909795
Reviewed-by: Rune Lillesveen <futhark@chromium.org>
Commit-Queue: Kevin Babbitt <kbabbitt@microsoft.com>
Cr-Commit-Position: refs/heads/main@{#1369108}
--
wpt-commits: 3585305e8365779fac42dd87e4f52c89c3e884a6
wpt-pr: 48563
Automatic update from web-platform-tests
Refactor visual overflow computation for view-transitions (part 2)
Use the new CC mechanism to compute the visual overflow.
We pass the layout-computed visual overflow as the max extent, and
then let CC correct it in order to map the content rect.
- We keep the replaced-content geometry projection as is, but also pass
the "max extents" to the CC laye (ViewTransitionContentLayerImpl),
and use it to map the border box of the original element to the
content box of the pseudo-element. Note that the border-box is now
in enclosing-layer space, as it needs to map correctly to the
max extents and actual extents. The border box sent over the wire
also has includes the offset from the enclosing layer, which would
only be different from zero for inline elements.
- When the captured rects from CC are delivered, they are used to map.
the contents of the old
pseudo-elements.
Bug: 347947051
Change-Id: I5202078666d016156c5abee99772726947e8a1d9
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5920431
Commit-Queue: Noam Rosenthal <nrosenthal@chromium.org>
Reviewed-by: Mike West <mkwst@chromium.org>
Reviewed-by: Khushal Sagar <khushalsagar@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1368988}
--
wpt-commits: ba4639ad64908ba29f6a9b9bf7a4d6b74dedfed2
wpt-pr: 48620
Automatic update from web-platform-tests
Preserve `none` in computed relative colors without a conversion
Per discussion at https://github.com/w3c/csswg-drafts/issues/10360,
there are a few test cases that incorrectly expect `none` to be
resolved to 0 in relative colors where no color space conversion was
needed. This CL updates both the test and the behavior in Blink.
Bug: 365818844
Change-Id: I14d6f83d792e4fb4e70171b833de308c66611fb4
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5905968
Reviewed-by: Fredrik Söderquist <fs@opera.com>
Commit-Queue: Kevin Babbitt <kbabbitt@microsoft.com>
Cr-Commit-Position: refs/heads/main@{#1368963}
--
wpt-commits: 766b106752e010363f30db798ba2ce2bceee25fb
wpt-pr: 48476
Automatic update from web-platform-tests
Fenced frame: Prevent file://localhost/* files from loading.
Fenced frames are meant to be able to load the following URLs with a
default constructor:
- about:blank URLs
- https://* URLs
- http://localhost URLs
The `net::IsLocalhost()` check that is currently being used for the last
case is only checking that the URL starts with "localhost", not that
the scheme is http/https. This allows the fenced frame to load URLs like file://localhost/path/to/file. This CL tightens the check to also
check that the scheme is HTTP. The "https://localhost/*" case will
still be covered by the "https://*" check.
Bug: 368628042
Change-Id: If53573f5701c28e5674ecb524191cddc12b950fe
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5902121
Reviewed-by: Shivani Sharma <shivanisha@chromium.org>
Commit-Queue: Liam Brady <lbrady@google.com>
Cr-Commit-Position: refs/heads/main@{#1368957}
--
wpt-commits: f4101379d6fbafa797c6e05a8e8d51f33cbb8602
wpt-pr: 48614
Automatic update from web-platform-tests
[Selection API] Use composed DOM tree for getComposedRanges
We update the implementation of getComposedRanges to use the cached
selection in DOM Tree to find start/end endpoints. These endpoints
are composed because they are not restricted to be within one TreeScope,
unlike the range associated with the selection.
We also remove the old logic of adding composed_range_ in Range to cache
flat tree endpoints.
Change-Id: Iee7807f74348d09a4b1e7c3dbb4ebe514bde874c
Bug: 356905057
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5921292
Commit-Queue: Di Zhang <dizhangg@chromium.org>
Reviewed-by: Mason Freed <masonf@chromium.org>
Reviewed-by: Siye Liu <siliu@microsoft.com>
Cr-Commit-Position: refs/heads/main@{#1368897}
--
wpt-commits: 94354b14d6c961a127691a205d2a7903c1b65412
wpt-pr: 48612
Automatic update from web-platform-tests
Mark iframe screenshot test as optional (#48494)
--
wpt-commits: 9c76757f5332678f9952f6ccb3824f62d30eca1f
wpt-pr: 48494
