Automatic update from web-platform-tests
[TablesNG] wpt rowspan>1 redistribution test
Rowspan>1 height redistribution to rows is an area where there was little
consistency between browsers. This test tries to highlight the issues.
Change-Id: I9b26f472717767fa9d2d7d27deb3f7da2590056d
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2826844
Commit-Queue: Aleks Totic <atotic@chromium.org>
Reviewed-by: David Grogan <dgrogan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#873228}
--
wpt-commits: 90e0da651b47194ab03998cd50048aa40ef847fd
wpt-pr: 28487
Automatic update from web-platform-tests
[LayoutNG] Fixedpos offset in a nested multicol
When a fixedpos is nested in another OOF inside a nested multicol,
and the outer OOF is not positioned at top:0 inside the containing
block of the fixedpos, the offset of the fixedpos was incorrect. The
reason being that its static position was getting updated in
NGContainerFragmentBuilder::PropagateOOFPositionedInfo() to take the
offset of the OOF into account. This change makes sure that
the static position no longer adds in this offset and instead is
relative to its containing block.
Bug: 1198672
Change-Id: I79ec28a036637b4a985fe41752be323845b8c440
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2827007
Commit-Queue: Alison Maher <almaher@microsoft.com>
Reviewed-by: Morten Stenshorne <mstensho@chromium.org>
Cr-Commit-Position: refs/heads/master@{#873160}
--
wpt-commits: d69b6ed41a784586305b02fdafcc486acad20038
wpt-pr: 28497
Automatic update from web-platform-tests
[LayoutNG] OOF positioned offset in a nested multicol
OOF positioned elements (abspos or fixedpos) in a nested multicol
could get the wrong offset if the containing block started in a column
that was not the first in a given multicol fragment. The reason being
that we had adjusted the containing block offset in CL:2680070 to take
into account columns from previous fragments. However, if the containing
block was not in the first column in a multicol fragment, we would have
already adjusted for this in NGContainerFragmentBuilder (given that its
column's builder would have access to the previous column break token).
To fix this, utilize the previous_break_token_ stored in
NGBoxFragmentBuilder (which is available in both cases above) to
determine the correct offset for the containing block. Instead of
passing both this adjustment for the containing block and a bool to
determine if we should propagate OOFs, a new base::Optional<LayoutUnit>
variable is passed into PropagateOOFPositionedInfo() to determine both
(credit: CL:2827892). This containing block adjustment is determined
in NGBoxFragmentBuilder::BlockOffsetAdjustmentForFragmentainer().
Bug: 1198578
Change-Id: I3c183dd83ed45e240fb9f62347dd388e370b1510
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2822985
Commit-Queue: Alison Maher <almaher@microsoft.com>
Reviewed-by: Morten Stenshorne <mstensho@chromium.org>
Cr-Commit-Position: refs/heads/master@{#873126}
--
wpt-commits: 16e47e623385c2ff468cd10fcb9d42b78d2b8e80
wpt-pr: 28494
Automatic update from web-platform-tests
Revert "Expand synthesized clip layer by 1 pixel"
This reverts commit 8643a99de2538ac309252247a5fe90756d9014e9.
Reason for revert: Caused crbug.com/1199366. The underlying issue is
still unclear yet. Revert first, and we'll continue to investigate
which will be tracked by crbug.com/1171601.
Original change's description:
> Expand synthesized clip layer by 1 pixel
>
> Though the clip mask layer and the masked layer are perfectly aligned in
> blink and cc, the geometry of the quads may go through different mapping
> code paths, depending on the types of the quads, causing slight
> misalignment of the mask quad and masked quad and hairline align the
> edges. Expand the mask layer by 1 pixel to ensure its coverage on the
> masked layer in case of floating-point errors etc.
>
> Bug: 1171601
> Change-Id: I4b64e1b6147d4e003a1ad131a605ee4b9886d57f
> Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2786129
> Commit-Queue: Xianzhu Wang <wangxianzhu@chromium.org>
> Reviewed-by: Philip Rogers <pdr@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#869019}
Bug: 1171601,1199366
Change-Id: Ibd2c8206c28e494290b42f0c46cee86e277942a5
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2828563
Commit-Queue: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Auto-Submit: Xianzhu Wang <wangxianzhu@chromium.org>
Cr-Commit-Position: refs/heads/master@{#873117}
--
wpt-commits: 23ae23071d3c44ee4b6b5a625498229f6107a31c
wpt-pr: 28528
Automatic update from web-platform-tests
[webcodecs] Implement transfer semantics for VideoFrame
This CL adds paths for modules to implement transfer semantics:
- ExtractTransferables() is split into the outer loop (stays in
SerializedScriptValue) and the inner per-object processing which
moves into V8ScriptValueSerializer. Modules can add implementations
to V8ScriptValueSerializerForModules.
- Transferables gains a generic TransferList map modelled after
SerializedScriptValue::Attachment. The TransferList API includes a
method for finalizing transfers, which is called from
V8ScriptValueSerializer::FinalizeTransfer().
Using these APIs, transfer of VideoFrames is then implemented
straightforwardly.
Bug: 1178038
Change-Id: Iafdc02b368d09cdb04c6e48ef187a4733b14c0f5
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2719085
Commit-Queue: Dan Sanders <sandersd@chromium.org>
Reviewed-by: Jeremy Roman <jbroman@chromium.org>
Reviewed-by: Thomas Guilbert <tguilbert@chromium.org>
Cr-Commit-Position: refs/heads/master@{#873099}
--
wpt-commits: 9dfd4b723781cf12e10810e2b758d56792b008ab
wpt-pr: 27772
Automatic update from web-platform-tests
Always add consumed block-size when calculating legacy static pos.
We used to do it only when the parent of an OOF was a fragmentainer, but
this is always needed for any descendant as well. The legacy code in
LayoutBox::ComputeInlineStaticDistance() expects the block-offset to be
relative to the parent in the flow thread coordinate space.
Also initialize previous_break_token_ to nullptr. This was a mistake
during oilpan conversion, which suddenly became a problem (it may have
been safe-ish before, but not anymore).
The new test abspos-autopos-contained-by-viewport-001.html would fail
without this fix.
Bug: 1199258
Change-Id: I7520cad9112485e908df3be33b354aec58332db8
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2826460
Reviewed-by: Alison Maher <almaher@microsoft.com>
Reviewed-by: Ian Kilpatrick <ikilpatrick@chromium.org>
Commit-Queue: Morten Stenshorne <mstensho@chromium.org>
Cr-Commit-Position: refs/heads/master@{#873069}
--
wpt-commits: ac12ed34d2b923cac8af08055036679c3a080afb
wpt-pr: 28526
Automatic update from web-platform-tests
Revert "Include custom properties on computed CSSStyleDeclaration"
This reverts commit 3b8999adcdc905f13b22bbde6b198f158cb9d6c7.
Reason for revert: Huge performance regressions.
Bug: 949807
Fixed: 1199142
Original change's description:
> Include custom properties on computed CSSStyleDeclaration
>
> Since we need a "stable" list of custom properties for length()
> and item(), and we also need not-abysmal performance when calling
> those functions, this CL caches a vector with the variables names
> on ComputedStyle itself.
>
> In CSSComputedStyleDeclaration::item(), we check if the incoming
> index is in the range of the standard properties, and if so return
> the appropriate one. Otherwise, we're in the variable range (which
> is defined per spec to appear after the standard properties), and
> we'll fetch the list of variable names.
>
> Unfortunately the inclusion of custom properties requires a clean
> style, which means length()/item() now updates style/layout as needed.
> This might cause performance regressions, but I don't see a way around
> this.
>
> Note: ComputedStyle::StyleInheritedVariables/NonInheritedVariables
> were changed to return const pointers, as to not "leak mutability",
> which would have made it hard to invalidate the cache. (Nothing
> requires those return values to be non-const after StyleCascade
> anyway).
>
> Note: ComputedStyle::GetVariableNames() has pretty good test coverage
> in ComputedStyleTest already. Added a couple of new ones that target
> cache invalidation.
>
> Fixed: 949807
> Change-Id: I6b181af5c4f025c4fd433de0a6ca221055c16693
> Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2822260
> Commit-Queue: Anders Hartvoll Ruud <andruud@chromium.org>
> Reviewed-by: Rune Lillesveen <futhark@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#872446}
Change-Id: I071cc2e00838a17407fc0d1e5745db9894d59ceb
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2829151
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Commit-Queue: Anders Hartvoll Ruud <andruud@chromium.org>
Cr-Commit-Position: refs/heads/master@{#873035}
--
wpt-commits: f3cbe18b3432c3cd16d6e3ca35c5277cf4fb92d4
wpt-pr: 28525
Automatic update from web-platform-tests
Fix WPT transform-inline test by using Ahem
transform-inline-001.html fails on Edge and Chrome due to a slight
font mismatch. This issue is resolved by making the test use Ahem.
This change does not detract from the original purpose of the test,
which is CSS transforms.
Bug: 1181103
Change-Id: Iea4aa777c1630e874810ba6c9dd55d5bfeb082ab
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2828954
Reviewed-by: Philip Rogers <pdr@chromium.org>
Commit-Queue: Sahir Vellani <sahir.vellani@microsoft.com>
Cr-Commit-Position: refs/heads/master@{#873009}
--
wpt-commits: 505211ba4362d8bdb7431631b29a2923dfcb30b4
wpt-pr: 28524
Automatic update from web-platform-tests
[ResourceTiming] Prefer PerformanceObserver over getEntriesByType in WPTs
Older versions of the Resource Timing spec don't specify a strong enough
ordering constraint between an element's "onload" and the
PerformanceResourceTiming entry being added to the Performance Timeline.
As such, our strategy of calling "performance.getEntriesByType()" after
the "onload" event fires isn't guaranteed to return the expected entry.
This CL modifies our approach to use, and block on, a
PerformanceObserver being notified of entries being added to the
timeline. This eliminates a class of flakiness we were seeing during
manual testing.
Bug: 1171767
Change-Id: Ifa4995683f5ef4ff6e53f9780f6d485e6cfb3d81
GithubIssue: https://github.com/w3c/resource-timing/issues/254
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2826597
Reviewed-by: Yoav Weiss <yoavweiss@chromium.org>
Commit-Queue: Tom McKee <tommckee@chromium.org>
Cr-Commit-Position: refs/heads/master@{#872849}
--
wpt-commits: d0d5cb6819ec7f30424ff838186ee1769d01e14f
wpt-pr: 28506
Automatic update from web-platform-tests
[Credentialless]: Make fetch.tentative.https.html more reliable.
Previously, the test was assuming the window can be closed immediately
after sending the last fetch request.
This is not true, because closing the window might interrupt previously
scheduled fetch requests and fails the test.
We need to wait for the test completion before assuming we can close the
windows.
I did a local test with 100 wpt run:
- without fix: 16/100 failures.
- with fix: 0/100 failures.
Bug: 1175099,1198255
Change-Id: I70a56ba4821d46d12f45c38d8850c876c5957649
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2825904
Auto-Submit: Arthur Sonzogni <arthursonzogni@chromium.org>
Reviewed-by: Antonio Sartori <antoniosartori@chromium.org>
Commit-Queue: Antonio Sartori <antoniosartori@chromium.org>
Cr-Commit-Position: refs/heads/master@{#872837}
--
wpt-commits: 6d4fca70173368916259efd9d0d713a9c232bedf
wpt-pr: 28503
Automatic update from web-platform-tests
[selectors] Add tests related to script focus and :focus-visible
--
[selectors] Mark some script focus and :focus-visible tests as tentative
The tests reflect the current browser behavior,
but the user expectations might be different in these.
The spec text is still being worked on, so let's mark them as tentative
at this stage.
--
wpt-commits: d9b1b7f1be33ee72257970e4ad02991d2670c823, 70a9959076663507a5d115e06acc00aba1bdd2c3
wpt-pr: 27806
Automatic update from web-platform-tests
NativeIO: Enable 3rd party cookie block on capacity allocation
This change adds a check to all filesystem-like methods
(requestCapacity, releaseCapacity, getRemainingCapacity) that makes them
throw when 3rd party cookies are blocked and the call came from a 3rd
party context (e.g. from within an iframe).
This CL also adds a test that NativeIO does not crash the renderer when
accessed from destroyed execution contexts.
Bug: 1137788
Change-Id: Ifb0fcdae0dad0fc2eae5e061a6800520dffcef14
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2821323
Reviewed-by: Marijn Kruisselbrink <mek@chromium.org>
Reviewed-by: Christian Dullweber <dullweber@chromium.org>
Commit-Queue: Richard Stotz <rstz@chromium.org>
Cr-Commit-Position: refs/heads/master@{#872768}
--
wpt-commits: 1441bb6cc4af8eb81a614873456a7a9c8d885854
wpt-pr: 28492
Automatic update from web-platform-tests
XHR: restructure FormData tests
This commit restrucutres xhr/formdata tests that do not require a form element to be .any.js tests, and thus runnable outside the browser, and in workers (like in Deno). This commit only moves tests around (no changes or removals), with the exception of the "Passing a String object to FormData.set should work" and "Passing a String object to FormData.append should work." tests because the tests are exact replicas of the "formdata with string" test in xhr/formdata.html. Because of that they have been removed.
--
wpt-commits: e45a9f9e65d039c76817ee2a6a1ef02c9311a1cb
wpt-pr: 28460
Automatic update from web-platform-tests
implement crypto.randomUUID()
Implementation of crypto.randomUUID() using base/guid which indicates
"A cryptographically secure random source will be used".
Refs: https://wicg.github.io/uuid/
Bug: 1197594
Change-Id: I0f5402b4d83a5d89f02425d8050a7737264672dc
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2804758
Commit-Queue: Benjamin Coe <bencoe@google.com>
Reviewed-by: Domenic Denicola <domenic@chromium.org>
Reviewed-by: Mike West <mkwst@chromium.org>
Cr-Commit-Position: refs/heads/master@{#872711}
--
wpt-commits: 15874bf6200f929ecb91726aab4c8e4d1412fee0
wpt-pr: 28499
Automatic update from web-platform-tests
Update offsetParent behavior in shadow trees
New behavior for offsetParent in shadow trees was discussed in [1] and
[2]. This lead to a chromium patch [3] which changed the behavior. After
[3] landed, the desired behavior in the discussions in [1] and [2]
seemed to have changed. Unfortunately, the author of [3] was no longer
working on chromium. This new behavior was then added to the spec and
landed in webkit [4] and firefox [5], and a WPT was added for it [6].
This patch implements the new behavior to follow suit with webkit and
firefox based on the WPT in [6]. Unfortunately, there are several tests
which are either internal to chromium or are only passing in chromium
which appear to oppose this new behavior and will have to be updated or
removed:
- external/wpt/css/css-contain/content-visibility/content-visibility-035.html
- external/wpt/css/css-contain/content-visibility/content-visibility-044.html
- fast/dom/shadow/offset-parent-does-not-leak-ua-shadow.html
- shadow-dom/offsetParent.html
For shadow-dom/offsetParent.html, I verified that firefox and safari
both currently fail the same tests which this patch does.
[1] https://github.com/WICG/webcomponents/issues/497
[2] https://github.com/w3c/csswg-drafts/issues/159
[3] 18d455ee83
[4] https://trac.webkit.org/changeset/239313/webkit
[5] https://bugzilla.mozilla.org/show_bug.cgi?id=1514074
[6] https://wpt.fyi/results/shadow-dom/offsetParent-across-shadow-boundaries.html
Fixed: 920069
Change-Id: I168edc5ad0e4fcb92d0c4a440623f2424b14a988
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2775208
Reviewed-by: Mason Freed <masonf@chromium.org>
Reviewed-by: vmpstr <vmpstr@chromium.org>
Reviewed-by: Ian Kilpatrick <ikilpatrick@chromium.org>
Commit-Queue: Joey Arhar <jarhar@chromium.org>
Cr-Commit-Position: refs/heads/master@{#872658}
--
wpt-commits: 8d8b8c4e2e42d07398fd5e98b541ee239e1d35a4
wpt-pr: 28201
Automatic update from web-platform-tests
Background Fetch: use permission context directly when initiating fetch.
This reduces duplicated logic between BackgroundFetchDelegateImpl and
BackgroundFetchPermissionContext. It also forces the logic to align.
Specifically, the permission context returned ALLOW when automatic DLs
were allowed, but the implementation enforced ASK even if auto DLs
are allowed.
This temporarily regresses WebLayer because there's no
Background Fetch permission context there, but that will be addressed
soon.
Bug: 1189247, 1057770
Change-Id: I5441f00835c186826927d2ef38612c8e55eb049f
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2800512
Commit-Queue: Evan Stade <estade@chromium.org>
Reviewed-by: Nasko Oskov <nasko@chromium.org>
Reviewed-by: Mugdha Lakhani <nator@chromium.org>
Cr-Commit-Position: refs/heads/master@{#872641}
--
wpt-commits: c7646854131fe696fa38ae8b5e3e611dc6f52d19
wpt-pr: 28484
Automatic update from web-platform-tests
[ReplacedNG] Compute min-content/max-content within ComputeReplacedSize.
\o/
This computes the min/max content sizes for replaced elements directly
within ComputeReplacedSize.
This allows us to simply call:
ComputeReplacedSize(node, space)
and get the correct size for replaced elements with only those
arguments!
Bug: 1114280
Change-Id: Ieedacf5f0ee814668f1c2c4633946b0bbb27e89f
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2818210
Commit-Queue: Ian Kilpatrick <ikilpatrick@chromium.org>
Reviewed-by: David Grogan <dgrogan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#872558}
--
wpt-commits: 1c760edb722b39d5b69dece6b4ec463bf3dee9cb
wpt-pr: 28483
Automatic update from web-platform-tests
Add pointerevents' properties to gpubenchmarking.PointerActionSequence
I have added the pointerevents' properties to WebDriver Action API, and
some WPT tests. Because gpubenchmarking.PointerActionSequence does not
have these properties, these tests are failing on content shell. This CL
will add the properties to gpubenchmarking.PointerActionSequence as
well.
Bug: 1147126
Change-Id: I55d14a342d63461cdfab6a31389a7c0bb9757bbb
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2683954
Reviewed-by: David Bokan <bokan@chromium.org>
Reviewed-by: Nasko Oskov <nasko@chromium.org>
Reviewed-by: Philip Jägenstedt <foolip@chromium.org>
Reviewed-by: Mustaq Ahmed <mustaq@chromium.org>
Commit-Queue: Lan Wei <lanwei@chromium.org>
Cr-Commit-Position: refs/heads/master@{#872509}
--
wpt-commits: 51639933704c8a87a2ac7d728e84613cbef3c9e0
wpt-pr: 27681
Automatic update from web-platform-tests
Initial implementation of scrollbar-width
ComputedStyle::HasCustomScrollbarStyle() returns true if a custom
style for the scrollbar has been set via ::-webkit-scrollbar and
the value of scrollbar-width has not been changed from the
default. This replaces HasPseudoElementStyle(kPseudoIdScrollbar)
so the standard properties for styling scrollbars take
precedence over the non-standard method.
For convenience, the method Scrollbar::CSSScrollbarWidth()
returns the value of the scrollbar-width property in that
scrollbar's style.
Scrollbar themes take scrollbar-width into account when
calculating a scrollbar's thickness, margin, etc.
When scrollbar-width is "none", scrollbar themes simply return
a thickness of 0 to hide the scrollbar.
WebThemeEngine::ScrollbarStyle includes the thickness and margin
values for the thin variant.
ScrollbarThemeMac uses NSRegularControlSize and
NSSmallControlSize to size and display the appropriate type of
scrollbar for the default and thin variants.
ScrollbarThemeOverlay stores the thickness and margin for the
thin variant, although at the moment these values are the same
as the default (pending changes to WebThemeEngine). The same
applies to the ScrollbarThemeOverlayMobile subclass.
ScrollbarThemeAura simply scales the default scrollbar by a 2/3
ratio to paint the thin variant. In the future, the theme should
include specific width values for this variant. Perhaps it would
also be desirable to use a different set of graphic elements.
WPT tests are included.
Bug: 891944
Change-Id: I2d451f07de5a16e6cada8a68ae01da7358e8652a
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2681340
Reviewed-by: Xianzhu Wang <wangxianzhu@chromium.org>
Reviewed-by: Philip Rogers <pdr@chromium.org>
Reviewed-by: Rune Lillesveen <futhark@chromium.org>
Reviewed-by: Kentaro Hara <haraken@chromium.org>
Commit-Queue: Felipe Erias <felipeerias@igalia.com>
Cr-Commit-Position: refs/heads/master@{#872500}
--
wpt-commits: f68f2e6bbe1c50a30e690243b64906f2a490e515
wpt-pr: 28280
Automatic update from web-platform-tests
Add a test for crbug.com/1167352
Bug: 1167352
Change-Id: Id340d704e80c03cfc49a00a7f840c6866c450034
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2822813
Reviewed-by: Stephen Chenney <schenney@chromium.org>
Commit-Queue: Xianzhu Wang <wangxianzhu@chromium.org>
Cr-Commit-Position: refs/heads/master@{#872482}
--
wpt-commits: b3804b4ee1c96551ef3eccb7d4fb578db0ff3d21
wpt-pr: 28481
Automatic update from web-platform-tests
[css-color-4] Remove another test for color() fallback (#28479)
color() function fallback is no longer part of the spec.
--
wpt-commits: b6e84edf4a398cc48f4983cbd8398858816996ae
wpt-pr: 28479
Automatic update from web-platform-tests
Include custom properties on computed CSSStyleDeclaration
Since we need a "stable" list of custom properties for length()
and item(), and we also need not-abysmal performance when calling
those functions, this CL caches a vector with the variables names
on ComputedStyle itself.
In CSSComputedStyleDeclaration::item(), we check if the incoming
index is in the range of the standard properties, and if so return
the appropriate one. Otherwise, we're in the variable range (which
is defined per spec to appear after the standard properties), and
we'll fetch the list of variable names.
Unfortunately the inclusion of custom properties requires a clean
style, which means length()/item() now updates style/layout as needed.
This might cause performance regressions, but I don't see a way around
this.
Note: ComputedStyle::StyleInheritedVariables/NonInheritedVariables
were changed to return const pointers, as to not "leak mutability",
which would have made it hard to invalidate the cache. (Nothing
requires those return values to be non-const after StyleCascade
anyway).
Note: ComputedStyle::GetVariableNames() has pretty good test coverage
in ComputedStyleTest already. Added a couple of new ones that target
cache invalidation.
Fixed: 949807
Change-Id: I6b181af5c4f025c4fd433de0a6ca221055c16693
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2822260
Commit-Queue: Anders Hartvoll Ruud <andruud@chromium.org>
Reviewed-by: Rune Lillesveen <futhark@chromium.org>
Cr-Commit-Position: refs/heads/master@{#872446}
--
wpt-commits: 72e8d5c2463fb0109f53a3a4357a998cb66d391b
wpt-pr: 28476
Automatic update from web-platform-tests
content-visibility: Ensure to dirty continuations on force unlocks.
This patch ensures that we propagate a dirty bit from the continuations
up the ancestor chain when we do a force unlock. In "regular" cases
where we don't need to force unlock anything, we have layout containment
which should prevent us from having continuations.
R=ikilpatrick@chromium.org, kojii@chromium.org
Fixed: 1197492
Change-Id: I2d8386e480c48c97e77875d3f8d73ad3d419c18c
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2822899
Commit-Queue: vmpstr <vmpstr@chromium.org>
Reviewed-by: Koji Ishii <kojii@chromium.org>
Reviewed-by: Morten Stenshorne <mstensho@chromium.org>
Cr-Commit-Position: refs/heads/master@{#872454}
--
wpt-commits: 73eea9688017f191c7d68ca07049f968e543e022
wpt-pr: 28478
Automatic update from web-platform-tests
[LayoutNG] Nested fixedpos in a nested multicol
Previously, if a nested fixedpos was in a nested multicol, and the
containing block was the viewport, it was never laid out. To fix this,
propagate any OOF candidates found when performing layout of OOFs in an
inner multicol to ensure that any nested candidates bubble up to the
correct containing block.
Bug: 1188756
Change-Id: If5ff481d25c24f693575faf7309eae8cb5a37ba1
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2823871
Commit-Queue: Alison Maher <almaher@microsoft.com>
Reviewed-by: Morten Stenshorne <mstensho@chromium.org>
Cr-Commit-Position: refs/heads/master@{#872442}
--
wpt-commits: 7c912cad3fdef5f2c29d80390ba6b775ca348d98
wpt-pr: 28475
Automatic update from web-platform-tests
Revert "[Taskcluster] pin Chrome Dev to 91.0.4449.6" (#28464)
This reverts https://github.com/web-platform-tests/wpt/pull/28238.
Update expactations for websockets over h2, which now works.
--
wpt-commits: e2fe814ee728461bf4f14f5b8b705c565c48bd1c
wpt-pr: 28464
moz-wptsync-bot
Aleks Totic
Philip Jägenstedt
Alison Maher
Xianzhu Wang
Dan Sanders
Morten Stenshorne
Anders Hartvoll Ruud
Sahir Vellani
Aaron Krajeski
jgraham
Tom McKee
arthursonzogni
Manuel Rego Casasnovas
Richard Stotz
Luca Casonato
Benjamin Coe
Joey Arhar
Evan Stade
Eugene Zemtsov
Ian Kilpatrick
Lan Wei
Felipe Erias
Sam Weinig
Vladimir Levin
Yi Xu
Domenic Denicola
Daniel Vogelheim