Automatic update from web-platform-tests
Catch ProtocolError when trying to send h2 data
This looks very similar to
https://github.com/web-platform-tests/wpt/pull/27789 except we seem to
be getting a ProtocolError rather than a StreamClosedError. Handling
that by stopping trying to send more data seems pretty reasonable.
--
wpt-commits: fb5913ec08d888a05d3e21ba68106bdeb101138f
wpt-pr: 43753
Automatic update from web-platform-tests
Adjust syntax of text-spacing[-trim] to match CSSWG resolution
See https://github.com/w3c/csswg-drafts/issues/9511
--
Adjust behavior of existing tests to match CSS spec change
See https://github.com/w3c/csswg-drafts/issues/9511
--
wpt-commits: 028df8b0a1712346fe9cf95757d35b277031a167, b12ce07e54c644455dcf088584e586783e002871
wpt-pr: 43766
Automatic update from web-platform-tests
[A11y] Partial revert to fix crash when finding <option> parent
Partial revert of CL:5123413. Remove NOTREACHED() and restore comment.
This condition is rare, but can occur if a <select> is inside of a
<details> element. When the <select> is first created, there is
no layout object, and a generic AXObject (not AXMenuList) is
created. The autofill code can call in just at the right time to
request this parent, and hits the NOTREACHED().
Test: this crash occurs when loading clock.laplace.live, or by loading the included test in the browser. It's unclear why, but the test does
not crash when running as a web test.
Note that these AXMenuList* classes will be going away in favor
of a generic implementation based on the shadow DOM, using ordinary
AXNodeObjects.
Fixed: 1512999
Change-Id: I318d6198abc7b155e42030caf00f039c7b09e4ac
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5141647
Reviewed-by: Chris Harrelson <chrishtr@chromium.org>
Commit-Queue: Aaron Leventhal <aleventhal@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1239844}
--
wpt-commits: 8ac73b1966e968e9224a7869e5d1bf38b0a340cc
wpt-pr: 43759
Automatic update from web-platform-tests
[FedCM] Add more WPT tests for error API
These tests make use of the newly repurposed webdriver command to click
on buttons in the FedCM error dialog.
Bug: 1496474
Change-Id: I8cdbd271c73d123dd47f62d943148b604412f08a
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5021410
Reviewed-by: Christian Biesinger <cbiesinger@chromium.org>
Commit-Queue: Zachary Tan <tanzachary@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1239840}
--
wpt-commits: c84ca70108c0e0b26b731f16a4c2789040db66e4
wpt-pr: 43750
Automatic update from web-platform-tests
Lose pointer capture if capture target is removed by gotpointercapture.
Also check that capture target is still in DOM before getting pointer capture.
Bug: 1502031
Change-Id: Id7791dbd01fc8b688a23120dc4bdeec8e6bf62f8
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5053186
Auto-Submit: Robert Flack <flackr@chromium.org>
Reviewed-by: Mustaq Ahmed <mustaq@chromium.org>
Commit-Queue: Robert Flack <flackr@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1239810}
--
wpt-commits: a014594610d06c1f889fe25345639149a619dff5
wpt-pr: 43344
Automatic update from web-platform-tests
Notify the flow thread about style change for inlines as well.
We only did it for LayoutBox or better, but this is necessary for
LayoutInline as well.
When a layout object changes whether or not it's a containing block for
out-of-flow positioned descendants, the multicol flow thread machinery
may need to insert or remove LayoutMultiColumnSet objects - since
ouf-of-flow positioned descendants may become part of (or cease to be
part of) the multicol container because such a change.
This is detected in NeedsToReinsertIntoFlowThread().
Move the flow thread hooks in StyleWillChange() and StyleDidChange()
from LayoutBox to LayoutBoxModelObject, so that this also works for
inlines.
This was most likely a LayoutNG regression, since the legacy engine
didn't really let inlines act as containing blocks.
Bug: 1503298
Change-Id: I7afcf447f8d81bd35cb1f049d04c7cdae3d3abf1
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5137017
Reviewed-by: Ian Kilpatrick <ikilpatrick@chromium.org>
Commit-Queue: Morten Stenshorne <mstensho@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1239713}
--
wpt-commits: 71de0efe88ade73baf607fed62053ab9b4fbef8b
wpt-pr: 43744
Automatic update from web-platform-tests
Support *.test. in chrome (#43701)
Co-authored-by: Antonio Sartori <antoniosartori@google.com>
--
wpt-commits: 7024c1b11976f37bb24eca80a40699327b12334c
wpt-pr: 43701
Automatic update from web-platform-tests
Retry firefox_android package install
Instead of having a single very long timeout, retry the install from
the start up to 5 times.
It appears that when the install fails it's not running into a
timeout, but for some reason the emulator is in the wrong state. In
this case restarting the operation from scratch works. Locally three
attempts in total were sometimes required, but this allows up to 5
retries (six total attempts) since CI machines may be slower.
--
wpt-commits: ecf9566335f222a8a724edd3d19004cc5e1867f8
wpt-pr: 43752
Automatic update from web-platform-tests
Remove experimental ::details-summary pseudo-element.
Developers can already style the <summary> element. Being able to style
a pseudo-element that wraps only the <summary> element has little value
and appears to add substantial confusion. (It does slightly improve the
ability to style the default summary when no summary is provided, but
not necessarily in useful ways.)
Bug: 1469418
Change-Id: Ia0c9bfd61696e5ca46e4ab3113e7bc45fde88bf4
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5137042
Reviewed-by: Rune Lillesveen <futhark@chromium.org>
Auto-Submit: David Baron <dbaron@chromium.org>
Commit-Queue: David Baron <dbaron@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1239642}
--
wpt-commits: cf0e998b161ad83360079ed86a51cf3f18836711
wpt-pr: 43734
Automatic update from web-platform-tests
Re-enable infrastructure tests on Android for Firefox
--
Add android-specific requirements
--
Pass --no-incremental to `adb install`
This seems to make the install work reliably. Unfortunately using the
mozdevice command (which has better validation of success) will require a
new mozdevice release, so for now inline a simple adb call
--
Make the emulator log accessible in CI
Copy the log to the artifacts directory after the run is complete.
--
Add a timeout trying to start the emulator
--
Pin emulator version at 32.1.15 on linux
This is enough to ensure we get a working version in CI; the problem with the
newer version seems to mostly apply when running in the docker image, so
hopefully this doesn't disrupt local testing too much
--
Missing f-string prefix
--
Update infrastructure metadata
--
wpt-commits: b286792dcbdd5a367ebfe0996f857bbd09e1ecc1, 0f7ed2b24f23c2735a064972d2557e5df028e2cf, 029d84b1d78c83cd3d0dd14ff2e09103976795a1, 7984cebc691bd9e4dbb95ba0d6865a963e1ab69c, a572c11c115b2c5e4ee939972d17ace3c5733cbc, 683e49d460e9b622858dc894c949286bff0daaaf, df229d6a4144b32896b5990ce5c50cb6f6e7014c, dbc3df1f6c09c8cfe6c55855e5e0195cc7eb5a78
wpt-pr: 43689
Automatic update from web-platform-tests
Clone and update multicol fragments properly when adding OOFs.
If the original inner multicol container had descendant multicols with
pending OOFs, make sure that the cloned and updated multicol also
reflects this state.
For that to work, we also need to allow more OOF descendant propagation
when cloning and updating fragmentainers via SimplifiedLayoutAlgorithm.
Remove the disable_oof_descendants_propagation_ flag, and instead block
such propagation for OOF fragments that are inside a fragmentainer. Such
fragments are special, and the propagation is handled manually in
OutOfFlowLayoutPart::AddOOFToFragmentainer(). All other fragments should
propagate normally.
The problem here was that the HasNestedMulticolsWithOOFs() flag in
PhysicalFragment got lost when the fragments got cloned and updated, which in turn prevented us from missing the cache in
LayoutBox::CachedLayoutResult() when re-laying out.
Bug: 1501713
Change-Id: Ib56750bb63a0d0894eef4d2979d24b93bdf9a761
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5134653
Reviewed-by: Alison Maher <almaher@microsoft.com>
Commit-Queue: Morten Stenshorne <mstensho@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1239601}
--
wpt-commits: 2dbbdc316f24acc39f6d0cec22005b320459e798
wpt-pr: 43726
Automatic update from web-platform-tests
[css-text] balance and pretty are not values of white-space
related to https://bugs.chromium.org/p/chromium/issues/detail?id=1501748
--
wpt-commits: a41c93e155a26abe762bf93b9267b8ff4a7c1c5e
wpt-pr: 43741
Automatic update from web-platform-tests
Fix checkbox, radio baseline position when in vertical writing mode
We update LogicalBoxFragment::BaselineMetrics such that if all following
conditions are met, then the baseline should be centered:
- writing mode is vertical
- control part is radio or checkbox
- baseline type is not alphabetic
Further, for vertical writing mode, we map the correct margin values to
the ascent/descent baseline values.
Change-Id: Ib87b1c412181acb0170bb4ea017383a9eb73df1e
Bug: 1511382
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5123415
Reviewed-by: Di Zhang <dizhangg@chromium.org>
Commit-Queue: Ian Kilpatrick <ikilpatrick@chromium.org>
Reviewed-by: Ian Kilpatrick <ikilpatrick@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1239547}
--
wpt-commits: 715c01dd8ab90c9e402eddf7792f26f243ee843d
wpt-pr: 43682
Automatic update from web-platform-tests
Make showPicker() throw NotSupportedError if select isn't being rendered
If the select element has no layout box throw the correct dom exception.
Bug: 1485010
Change-Id: I7824c4b904ed4677faec6200f56c8055fefcfcf9
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5071276
Reviewed-by: Joey Arhar <jarhar@chromium.org>
Commit-Queue: Luke <lukewarlow156@gmail.com>
Reviewed-by: Vladimir Levin <vmpstr@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1239446}
--
wpt-commits: 67141e8e3dfca0d7073988cec7748b1d9b45f2e3
wpt-pr: 43648
Automatic update from web-platform-tests
[@position-fallback] Make sure LayoutObject style is correct
When trying to make room for absolute positioned with position-fallback
the style without animations applied is used. Since TryCalculateOffset
uses a mix of the passed in style and the style retrieved from the
LayoutObject, we need to set the base style on the LayoutObject to make
sure they are in sync.
Bug: 1512373
Change-Id: Id7bc35b11a5c9b2b7c0ac20883ff457f4d8bbfd3
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5134895
Reviewed-by: Ian Kilpatrick <ikilpatrick@chromium.org>
Commit-Queue: Rune Lillesveen <futhark@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1239412}
--
wpt-commits: 3d2d5d5bd237e705feba7568609c0261fc763720
wpt-pr: 43728
Automatic update from web-platform-tests
[@position-fallback] Avoid two-pass without fallback styles
@try fallbacks are done with base computed style without animations
applied first, with animations applied in a second step. If there are no
fallback styles there is no need to do the base computed style step.
This fixes the DCHECK triggered by TryCalculateOffset mixing styles from
the LayoutInputNode and the passed in style which may be the base style.
However, it just fixes the cases where there are no position fallback
styles present. That will be fixed in a separate CL.
Bug: 1512373
Change-Id: I72de67cc0b5ffbf4208f361a2947448d876df87c
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5132397
Reviewed-by: Ian Kilpatrick <ikilpatrick@chromium.org>
Commit-Queue: Rune Lillesveen <futhark@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1239353}
--
wpt-commits: 149fd81726ebaccf4d7bb9af5706a60ccfe93f73
wpt-pr: 43724
Automatic update from web-platform-tests
Reset softnav_painted_area_ when a new soft nav is about to happen
This is reset the softnav_painted_area_ when a new potential soft nav
is about to happen.
The image used in dropped-entries.tentative.html is changed for a
small one. That test adds the image 52 times to verify that entries
are dropped once the buffer is full. Small image is less like to
cause timeout which is seen in one dry run. The timeout is not related
to the change in this CL.
Change-Id: I6bff75e1b2e32d35fccc16b6b519eca66bf0ac59
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5097529
Reviewed-by: Ian Clelland <iclelland@chromium.org>
Commit-Queue: Hao Liu <haoliuk@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1239259}
--
wpt-commits: 2fc3cc60c14873819af55d59b949eced2d10864c
wpt-pr: 43564
Automatic update from web-platform-tests
Remove CSS Zoom Adjustment for GetBoundingClientRect
This change modifies the behavior of the call for elements with a CSS
property. Previously, the dimensions of the bounding box returned by
GetBoundingClientRect were adjusted to the effective zoom of an
element, which is the combination of the page zoom and the CSS zoom.
This meant that the dimensions were not affected by either of those
values.
This had the unintended consequence of causing two elements with the
same size before zoom, but with different zoom properties, to have the
same sized bounding box, even though they had different sizes on the
page. This also caused crbug.com/967097
The new behavior adjusts the bounding box to the page's zoom, but not to
the CSS zoom. This means that two elements with the same size before
zoom, but with different zoom properties, will now have different sized
bounding boxes.
Follow up patches will add the same logic to other calls in the file to
ensure consistency between the calls.
Bug: 967097, 1510969, 1511048
Change-Id: I66e8d4ca02f117c6133bb15cea29fc5c50c5d819
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5085708
Auto-Submit: Yotam Hacohen <yotha@chromium.org>
Reviewed-by: Chris Harrelson <chrishtr@chromium.org>
Commit-Queue: Chris Harrelson <chrishtr@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1239239}
--
wpt-commits: 55a2476161170898d9f19db51918353b2e2e9640
wpt-pr: 43507
This patch uses RootedDictionary around `EncodedVideoChunkMetadata`,
`VideoDecoderConfig` and `VideoColorSpaceInit` in `OutputEncodedData` to
work around the JS hazard (unrooted) issues.
Differential Revision: https://phabricator.services.mozilla.com/D197228
Nobody shares an instance of `nsFrameIterator`. Therefore, we can make it
non-refcountable. Additionally, `nsVisualIterator` is not required because
it overrides only 4 methods which are really simple. So, once we merge it
into `nsFrameIterator`, we can allocate it in the stack.
Differential Revision: https://phabricator.services.mozilla.com/D197144
The class is used only for creating `nsFrameIterator`, but it's unnecessary
so that we can get rid of it and its interface.
Differential Revision: https://phabricator.services.mozilla.com/D197143
It's inherited only by `nsFrameIterator`, and `nsFrameIterator` can be declared
in a header file. So, the interface is not required and removing it can avoid
virtual calls.
Differential Revision: https://phabricator.services.mozilla.com/D197142
MozTaggedMemoryIsSupported doesn't work under ASAN because it tries to prctl the
nullptr page. It should be harmless to simply call prctl and ignore the error.
Depends on D195064
Differential Revision: https://phabricator.services.mozilla.com/D195065
This patch makes CanvasDrawEventRecorder track what eventCount we
recorded an external surface reference. When the reader has increment
its processedCount above that, we will release our reference as it
should have acquired a strong reference to the data. This was previously
done when we forwarded the texture, but with remote textures, we no
longer have this event. Now we check when we start a new recording, or
attempt to clear cached resources.
Differential Revision: https://phabricator.services.mozilla.com/D197216
This patch doesn't change behavior.
Before this patch, we're appending to the array that we're iterating over,
which is risky, though it happens to be safe right now since we don't use
any of the newly-invalid-data after the array is mutated.
This patch just uses a less risky approach (appending to a temporary array)
to avoid this fragility.
Differential Revision: https://phabricator.services.mozilla.com/D197217
Mozilla Releng Treescript
moz-wptsync-bot
dependabot[bot]
James Graham
Florian Rivoal
Aaron Leventhal
Zachary Tan
Robert Flack
Gérard Talbot
Philip Jägenstedt
Morten Stenshorne
Antonio Sartori
L. David Baron
Di Zhang
Luke Warlow
Rune Lillesveen
Nicolás Peña
Hao Liu
Yotam Hacohen
Randell Jesup
Stephen A Pohl
Chun-Min Chang
Stanca Serban
Masayuki Nakano
Benjamin Peterson
Sandor Molnar
Andrew Osmond
Daniel Holbert
Dan Baker
Ting-Yu Lin