Automatic update from web-platform-tests
Add test for resized image not being reconsidered for LCP (#42900)
This is to verify an image that is resized to be larger would not be
reconsidered as LCP.
Change-Id: I0a555bc418983baaddc405fd02e44e7ff85c8059
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/4995606
Reviewed-by: Ian Clelland <iclelland@chromium.org>
Commit-Queue: Hao Liu <haoliuk@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1244062}
Co-authored-by: Hao Liu <haoliuk@chromium.org>
--
wpt-commits: a12f1c70bab8f96a6040f64980921af575f69f74
wpt-pr: 42900
Automatic update from web-platform-tests
Update rendering state of automatic pull nodes before graph rendering
In rare cases, the rendering fan out count of automatic pull node
does not match the main thread fan out count after recreating
a platform destination followed by disconnection.
This CL forces the update of the rendering state of automatic
pull nodes before graph rendering to make sure that fan out counts
are synchronized before executing the audio processing function call.
NOTE: This change makes 2 WPTs fail. The follow-up work is planned
to address them once this patch is merged.
Bug: 1505080
Test: Locally confirmed that ASAN doesn't crash on all repro cases.
Change-Id: I6768cd8bc64525ea9d56a19b9c58439e9cdab9a8
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5131958
Reviewed-by: Michael Wilson <mjwilson@chromium.org>
Commit-Queue: Hongchan Choi <hongchan@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1246718}
--
wpt-commits: 9e0d5be01079de6c7d32474f2a26dfeaef2a791c
wpt-pr: 43978
Automatic update from web-platform-tests
Fix popover crash bug
It was previously possible to hit a CHECK (and I think some
underspecified behavior) with something like this:
<div popover id=p1>Popover 1
<div popover id=p2>Popover 2</div>
</div>
<script>
p1.showPopover();
p1.addEventListener('beforetoggle',() => p2.showPopover());
p1.hidePopover();
</script>
The problem is that "hide all popovers until" doesn't end up with
the desired "until" popover on the top of the stack in this case.
There is already a similar situation within the "hide all..."
algorithm itself, but that only handles the case where a popover
being hidden by "hide all..." has the beforetoggle listener. This
is the same problem, but for the case that the "until" popover
has that listener.
I believe this needs a spec update. WebKit doesn't crash (likely
it doesn't have the same CHECKs that Blink has), but also doesn't
result in both p1 and p2 being hidden in the above example.
Fixed: 1513282
Change-Id: I463888096db9a8983d355f246991bbbad596d1f2
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5141430
Auto-Submit: Mason Freed <masonf@chromium.org>
Commit-Queue: Mason Freed <masonf@chromium.org>
Reviewed-by: Joey Arhar <jarhar@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1246585}
--
wpt-commits: 138a1e36208d91293022211eb446a5b8c0aec8eb
wpt-pr: 43976
Automatic update from web-platform-tests
[IntersectionOptimization] Improve and fix cached rect invalidation
- Move invalidation from
ObjectPaintInvalidator::InvalidateDisplayItemClient to
PaintInvalidator::InvalidatePaint because the former doesn't
invalidate LayoutObjects that don't paint while these LayoutObjects
can still observe intersection based on layout.
- Rename InvalidateCachedRectsIfNeeded to
InvalidateCachedRectsIfPaintPropertiesChange() because it only
handles paint property changes but not local layout changes.
- Skip scroll optimization if the cached rect is invalidated for other
reasons.
Bug: 1400495, 1515283
Change-Id: Ib3015f13ac4280fc7a50e686982703eef3128812
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5154140
Reviewed-by: Stefan Zager <szager@chromium.org>
Commit-Queue: Xianzhu Wang <wangxianzhu@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1246582}
--
wpt-commits: 4f5b495f8c8b53bacb1009c81917f98c84ff9c0e
wpt-pr: 43890
Automatic update from web-platform-tests
Update popover=hint behavior to allow a stack of hints
The previous implementation only allowed one popover=hint to be open
at a time. Per conversation at [1], developers feel that it should be
possible to nest popover=hint popovers. This CL implements that
capability.
There are now two popover stacks in Document: PopoverAutoStack and
PopoverHintStack. Since it is possible to nest hints within autos,
the PopoverAutoStack can contain hints. However, there
are a few constraints:
- The PopoverHintStack only ever contains hints.
- Once the PopoverAutoStack contains a hint, all subsequent popovers
in the stack must also be hints.
- A popover=hint can never be the ancestor of a popover=auto.
The light dismiss behavior is roughly the same as before, with a
slight tweak that simplifies behavior: closing anything in the
PopoverAutoStack will always close everything in the PopoverHintStack.
That was not the case before, but it was a bit of a weird corner case.
Note that I found a crasher (happens in stable, with just auto
popovers) that I added a test for here. The bug for that is
crbug.com/1513282. I'll fix that in a followup.
[1] https://github.com/whatwg/html/issues/9776#issuecomment-1785814179
Bug: 1416284,1513282
Change-Id: Ic064ecf1377bb8abfc812654c85016e6d1cbbdaf
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5133909
Reviewed-by: Joey Arhar <jarhar@chromium.org>
Commit-Queue: Mason Freed <masonf@chromium.org>
Auto-Submit: Mason Freed <masonf@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1246573}
--
wpt-commits: 51c87dc4c5d4a61caef22344e4dba6f5f233ffc3
wpt-pr: 43737
Automatic update from web-platform-tests
Fenced frames: window.fence.disableUntrustedNetwork API surface
Add nonfunctional window.fence.disableUntrustedNetwork API surface
(behind feature flag), which is not available in urn iframes. Later CLs
will gradually make it functional.
This entails the following internal changes:
* New fields `can_disable_untrusted_network_` and
`has_disabled_untrusted_network_` in FencedFrameProperties.
`can_disable_untrusted_network_` is true for all fenced frame configs,
but disabled for urn iframes at the same time as partition nonce.
`can_disable_untrusted_network_` is sent to same-origin renderers to
gate availability of the window.fence.disableUntrustedNetwork API.
* Navigations through the FencedFrameConfig(url) constructor now store
`url` as the mapped url in the resulting FencedFrameProperties, so
that we can perform same- vs. cross-origin checks for availability of
window.fence.disableUntrustedNetwork.
I2P: https://groups.google.com/a/chromium.org/g/blink-dev/c/ofii__csdOY
Explainer: https://github.com/WICG/fenced-frame/blob/master/explainer/fenced_frames_with_local_unpartitioned_data_access.md#revoking-network-access
Bug: 1515599
Change-Id: I13ec7a570bd7d84b2bc3780b577e318b4d103e9a
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5124560
Reviewed-by: Dominic Farolino <dom@chromium.org>
Commit-Queue: Garrett Tanzer <gtanzer@chromium.org>
Reviewed-by: Arthur Sonzogni <arthursonzogni@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1246541}
--
wpt-commits: a09d86d617703f609473d4177543b0ffd05067ae
wpt-pr: 43710
Automatic update from web-platform-tests
[WebDriver BiDi] add `Storage.setCookie` test (#43968)
* Add `Storage` domain.
* Add `getCookies` command.
* Add `setCookie` command.
* Add a simple test for `setCookie`.
Out of scope: full test coverage of `Storage`.
---------
Co-authored-by: Maksim Sadym <sadym@google.com>
Co-authored-by: Alexandra Borovova <lutien@users.noreply.github.com>
--
wpt-commits: aee3194b908f9c39ca9923465316188aee8708b7
wpt-pr: 43968
Automatic update from web-platform-tests
[css-nesting] Fix @scope part of conditional-rules.html
This subtest has an inner rules with says "@scope (.test-12)", which
is equivalent to to "@scope (& .test-12)", i.e. with an implied "&"
in the front. This matches no elements, so I'm pretty sure the intent
was to just use "&" here instead.
Change-Id: I60b61f16f2012b10ba121c7e93e1134d2c79e218
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5192022
Auto-Submit: Anders Hartvoll Ruud <andruud@chromium.org>
Commit-Queue: Rune Lillesveen <futhark@chromium.org>
Reviewed-by: Rune Lillesveen <futhark@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1246468}
--
wpt-commits: a363c3c27f2794d96c8a092fae8e931fba0f1a36
wpt-pr: 43972
Automatic update from web-platform-tests
Add WPT for CSP with trailing dot in hostname
This CL adds a generic WPT test to make sure hosts with trailing dots
can be defined in CSP.
The CL targets connect-src directive as a proof of concept. More
tests can be added for other directives later on.
Bug: 1512075
Change-Id: Ia3d8ec4c4b6594d46ff6c65ff8275bfd6bdd70fb
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5185382
Reviewed-by: Antonio Sartori <antoniosartori@chromium.org>
Commit-Queue: Saeid Eid <saeideid@chromium.org>
Reviewed-by: Jonathan Hao <phao@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1246451}
--
wpt-commits: c48d657e34c181f1a96ba2d6a8d191239bcfc613
wpt-pr: 43939
Automatic update from web-platform-tests
Clarify instructions for testing on a local machine (#43948)
--
wpt-commits: e42493210e7b78d3c77043cb9fdb16e8eefafb9a
wpt-pr: 43948
Automatic update from web-platform-tests
[text-spacing-trim] Fix available width when line-start needs reshape
This patch moves the logic in `ShapingLineBreaker` to reshape
line-start to before computing the candidate break.
Candidate break is computed from available width, which
reshaping line-start may change.
The `text-spacing-trim` property may cause a large difference
when line-start reshaping occurs, large enough to find this
bug.
Bug: 1463891
Change-Id: Ie0f7c7cecafeaad38315bd1b9745c3073f9a8dc5
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5186201
Auto-Submit: Koji Ishii <kojii@chromium.org>
Commit-Queue: Kent Tamura <tkent@chromium.org>
Reviewed-by: Kent Tamura <tkent@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1246256}
--
wpt-commits: b0969422e3a66ff05675b8c84e6322ed962de60f
wpt-pr: 43959
Automatic update from web-platform-tests
[wptrunner] Rewrite `test_win_open_with_interaction.html`
This test didn't work as intended for `testdriver.js` implemented with
WebDriver because of crbug.com/1453780#c8. Fix the test page to also
include `testdriver.js`, and rewrite the test to perform interaction
like `infrastructure/testdriver/actions/crossOrigin.sub.html` [0].
The test now passes for `chrome` and other WebDriver-enabled browsers,
but `content_shell` on Linux or Windows still delivers click events to
the wrong window.
[0]: https://github.com/web-platform-tests/wpt/blob/master/infrastructure/testdriver/actions/crossOrigin.sub.html
Bug: 1453780
Change-Id: Iaaf33a7cbbccd8c63acfd5e68e6ea76954c520fd
Cq-Include-Trybots: luci.chromium.try:mac11.0-blink-rel
Cq-Include-Trybots: luci.chromium.try:mac11.0.arm64-blink-rel
Cq-Include-Trybots: luci.chromium.try:mac12.0-blink-rel
Cq-Include-Trybots: luci.chromium.try:mac12.0.arm64-blink-rel
Cq-Include-Trybots: luci.chromium.try:mac13-blink-rel
Cq-Include-Trybots: luci.chromium.try:mac13.arm64-blink-rel
Cq-Include-Trybots: luci.chromium.try:win11-blink-rel
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5171750
Commit-Queue: Jonathan Lee <jonathanjlee@google.com>
Reviewed-by: Weizhong Xia <weizhong@google.com>
Cr-Commit-Position: refs/heads/main@{#1246247}
--
wpt-commits: 2f29501192db686390788e2bbc29791c6021f571
wpt-pr: 43860
Automatic update from web-platform-tests
Infer a speculation rule's source from the other keys.
This is behind a RuntimeEnabledFeature, so we can enable it
once we're convinced it's reasonable, and disable it remotely
if anything surprising happens.
Bug: 1517696
Change-Id: Ie41529a13ab48671fc19c30003471e61fcc00781
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5186903
Reviewed-by: Kevin McNee <mcnee@chromium.org>
Auto-Submit: Jeremy Roman <jbroman@chromium.org>
Commit-Queue: Jeremy Roman <jbroman@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1246177}
--
wpt-commits: 9344126944631891175bee383db364749ed3fcc3
wpt-pr: 43947
Automatic update from web-platform-tests
Fix HTML directionality inheritance for empty <bdi> elements with no dir attribute.
This correctly initializes the flag indicating that the HTML
directionality is inherited from the parent for <bdi> elements, which
behaves like it has dir=auto by default. Even without this change, it
gets initialized correctly when a dir attribute is changed or when the
text content of the element changes.
This also clears the same flag when the dir attribute changes. This
should affect some other cases of dir attribute changes, but we somehow
didn't hit it until fixing the previous bug exposed the problem.
This fixes the failure of a test that I added in
https://github.com/web-platform-tests/wpt/pull/43896 in response to
noticing and fixing an error I made in the HTML spec in
https://github.com/whatwg/html/pull/10005 .
Bug: 576815
Change-Id: I3006d45079694d90dd178f8c439e22f6419bd992
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5186678
Commit-Queue: Mason Freed <masonf@chromium.org>
Reviewed-by: Di Zhang <dizhangg@chromium.org>
Auto-Submit: David Baron <dbaron@chromium.org>
Reviewed-by: Mason Freed <masonf@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1246176}
--
wpt-commits: 40b49b7273614349e83ad17076bc2c3c5531bb97
wpt-pr: 43933
André Bargull
Dão Gottwald
Cristina Horotan
Dimi
Manuel Bucher
Malte Juergens
Gabriele Svelto
Norisz Fay
Gabriel Bustamante
Jonathan Kew
moz-wptsync-bot
Blink WPT Bot
github-actions[bot]
Hongchan Choi
youennf
Jonathan Lee
Mason Freed
Xianzhu Wang
Yotam Hacohen
Garrett Tanzer
Maksim Sadym
Noam Rosenthal
Anders Hartvoll Ruud
Saeid Eid
dependabot[bot]
Nikolay Vitkov
Domenic Denicola
Chris Harrelson
Koji Ishii
Jonathan Lee
Jeremy Roman
L. David Baron