After we enabled support for large ArrayBuffers on 64-bit platforms, we could
also create larger ImageData objects. WebIDL bindings check for large ArrayBuffer{View}s
but not when they're wrapped in an ImageData.
It seems safest to limit ImageData arrays to the old 2 GB for now until we need larger
buffers.
Differential Revision: https://phabricator.services.mozilla.com/D121595
- `UDAT_TIMEZONE_GENERIC_FIELD` is used for the generic time zone name option.
- `UDAT_TIMEZONE_LOCALIZED_GMT_OFFSET_FIELD` is used for the time zone offset name option.
Differential Revision: https://phabricator.services.mozilla.com/D121873
This allows awsy-base to be run with Fission via the "variants" value.
It also makes us run it with Fission everywhere that we currently
run non-Fission, and changes it to be tier 1. For continuity, it
continues to run awsy-base with non-Fission on mozilla-central.
I'm not sure how much that matters, though in theory we could
have a regression there that doesn't show up with Fission.
This also disables bfcache in parent because it hits some process
selection issues that result in multiple about:blank ending up
in a single process.
I'm not sure if the bugbug-push-schedules.json changes are needed.
Differential Revision: https://phabricator.services.mozilla.com/D121682
During the pagehide event, we're checking that the target we're navigating away
from is the expected one.
We previously used the window property, but this is a getter that is updated when
a document is selected in the iframe picker.
Since the goal is to check the actual document the target was created for, we
can use the originalWindow property instead.
A test is added, which is failing without the fixes of this patch.
We take this as an opportunity to clear the iframe picker list on will-navigate,
(as the test was failing on Fission without this), which allows us to remove the
`_notifyDocShellDestroyAll` method of the `BrowsingContextTargetActor` that was
also called before emitting the `will-navigate` event.
Differential Revision: https://phabricator.services.mozilla.com/D121727
Automatic update from web-platform-tests
[webdriver]: element cloned into iframe (#29877)
Co-authored-by: Moshe Atlow <moshea@testim.io>
Co-authored-by: Henrik Skupin <mail@hskupin.info>
--
wpt-commits: f777c7b348c1e98d42a7c28415c8817d7fb79776
wpt-pr: 29877
Automatic update from web-platform-tests
MathML tools : upgrade scripts to Python 3 (#29912)
With this commit, the scripts in mathml/tools/ are upgraded to use
Python 3. A few more tweaks have been made.
- Use '#!/usr/bin/env python3' as header and make files executable.
- Use // for integer division.
- Don't import future print_function anymore.
- Use the latest Python3 API for urlopen.
- Fix url ofXML Entities's unicode.xml.
- Just use chr instead of unichr.
--
wpt-commits: 5a65cdf5c4f3ea2c637df92d837f613c75edf5b8
wpt-pr: 29912
Automatic update from web-platform-tests
[css-contain] Remove paint-containment-with-fixed-position.html
Remove paint-containment-with-fixed-position.html since
contain-paint-010.html is a similar test, it just needs to
also test inline.
Change-Id: Iaa3f7fe7fc830b4f7f53d4f41af5983bab69f29d
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/3071279
Reviewed-by: Anders Hartvoll Ruud <andruud@chromium.org>
Commit-Queue: Rob Buis <rbuis@igalia.com>
Cr-Commit-Position: refs/heads/master@{#908827}
--
wpt-commits: 3fb14d74ca0731c860d70c396041569e8cfd8b71
wpt-pr: 29910
Automatic update from web-platform-tests
[css-contain] Remove containment/contain-parse-and-apply.html
Remove containment/contain-parse-and-apply.html by adding the subtests
to WPT. Testing that contain computed values do not apply to the
children is tested by contain-computed-children.html.
The remaining tests for invalid contain values are added to
contain-invalid.html.
Change-Id: I283f6ce7de2b87f5cb23a244e9f83355a066972c
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/3065665
Commit-Queue: Rob Buis <rbuis@igalia.com>
Reviewed-by: Anders Hartvoll Ruud <andruud@chromium.org>
Cr-Commit-Position: refs/heads/master@{#908822}
--
wpt-commits: e27bdc34b98e75aaa08e5542f772935ea587684b
wpt-pr: 29892
Automatic update from web-platform-tests
Better monolithic content overflow handling.
Be ruby-aware when checking if we need a break before a line. Since ruby
annotations aren't part of the line box, just checking if the
block-offset + block-size of the line box is below the fragmentation
line won't do. Include annotation overflow as well.
Also make sure that it works correctly for blocks in vertical-rl
writing-mode. This was done incorrectly in CL:2696632.
Bug: 1236056, 1178742
Change-Id: Iaa137f1e45e437d31fe93305a29d80a2153d2d1e
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/3068937
Reviewed-by: Koji Ishii <kojii@chromium.org>
Reviewed-by: Kent Tamura <tkent@chromium.org>
Commit-Queue: Morten Stenshorne <mstensho@chromium.org>
Cr-Commit-Position: refs/heads/master@{#908801}
--
wpt-commits: 39eb0596a9f501131a5b6aa99b90963da0751b18
wpt-pr: 29890
Automatic update from web-platform-tests
[selectmenu] Add console warnings for missing button/listbox parts
Per the resolution of https://github.com/openui/open-ui/issues/291,
emit a console warning when a <selectmenu>'s default listbox or button
is removed without a replacement being provided.
In order to emit these warnings consistently, add a slotchange handler
to catch the case where the named slots are replaced at parse time
without a replacement part being provided. No mutation observer was
firing for this case, so the default parts were not being properly
swapped out.
Also, add a test that was missing for selected-value part in
selectmenu-parts-structure.tentative.html.
Bug: 1234899
Change-Id: If639d7eccf1db59518ef862e3beaaa8a2a09e6cc
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/3072218
Reviewed-by: Ionel Popescu <iopopesc@microsoft.com>
Reviewed-by: Mason Freed <masonf@chromium.org>
Commit-Queue: Dan Clark <daniec@microsoft.com>
Cr-Commit-Position: refs/heads/master@{#908669}
--
wpt-commits: 38ed3eab8b40cbcd875f2c0f773a52386ec6002d
wpt-pr: 29904
Automatic update from web-platform-tests
Add 'custom' value to prefers-contrast media query
As indicated in the spec, we are implementing the 'custom' value for
the prefers-contrast media query. This value corresponds to forced
colors that are neither high nor low contrast.
Spec: https://drafts.csswg.org/mediaqueries-5/#prefers-contrast
Bug: 1107431
Change-Id: I3af0a1965fbe57aaab2665c3c502f9b7555d2252
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/3061066
Reviewed-by: Chris Palmer <palmer@chromium.org>
Reviewed-by: Alison Maher <almaher@microsoft.com>
Reviewed-by: Peter Kasting <pkasting@chromium.org>
Commit-Queue: Sara Tang <sartang@microsoft.com>
Cr-Commit-Position: refs/heads/master@{#908632}
--
wpt-commits: 3790190e4421816fd4b334b1131a6a4f435c8e17
wpt-pr: 29845
Automatic update from web-platform-tests
Minor fixing to WPT css/css-fonts/fallback-remote-to-data-url.html
1. The font style is also applied to element non-target so that we can
verify the effect of unicode-range on data-font
2. Added 'font-display: swap' to remote-font so that we can observe the
fallback when loading remote-font
Bug: 897539
Change-Id: I9bbe0705bc7d3a9ebbaf92f2c8c26c557adec24a
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/3069529
Commit-Queue: Xiaocheng Hu <xiaochengh@chromium.org>
Auto-Submit: Xiaocheng Hu <xiaochengh@chromium.org>
Reviewed-by: Dominik Röttsches <drott@chromium.org>
Cr-Commit-Position: refs/heads/master@{#908594}
--
wpt-commits: ddaa08bb676adb2b173f85bf2dd88dcfb8e103e5
wpt-pr: 29885
Automatic update from web-platform-tests
Fix <input type=image> with CSS content
The LayoutObject creation was previously being performed via a call
to LayoutObjectFactory::CreateBlockFlow() and a conditional on the
style to LayoutInline(). This CL replaces all of that with a call to
LayoutObject::CreateObject(), which handles creating the correct
LayoutObject type, plus it also (now) handles CSS content pointing
to an image. Previous attempts to do this [1][2] fixed some of these
issues, but left a UAF in ImageResourceContent::RemoveObserver() [3].
To fix this, an unnecessary call to SetImageResource() in
ImageInputType::OnAttachWithLayoutObject() was removed.
Test coverage:
Fallback content display type: external/wpt/html/rendering/replaced-elements/attributes-for-embedded-content-and-images/input-image-inline-alt.html
UAF: external/wpt/html/rendering/replaced-elements/attributes-for-embedded-content-and-images/input-image-content-crash.html
CSS content (new test): external/wpt/html/rendering/replaced-elements/images/input-image-content.html
[1] crrev.com/779108
[2] crrev.com/780992
[3] https://bugs.chromium.org/p/chromium/issues/detail?id=1226558#c7
Fixed: 1226558
Change-Id: I64c61ebed91670b5fac665f1c4be9337f0814a9a
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/3064330
Commit-Queue: Ian Kilpatrick <ikilpatrick@chromium.org>
Auto-Submit: Mason Freed <masonf@chromium.org>
Reviewed-by: Ian Kilpatrick <ikilpatrick@chromium.org>
Cr-Commit-Position: refs/heads/master@{#908495}
--
wpt-commits: cdd20e3a6c7ea615f820cc28aa7c7029042c5f7a
wpt-pr: 29896
Automatic update from web-platform-tests
[ResourceTiming]: Refactor initiatorType tests (#29786)
This CL updates WPTs checking the behaviour of the initiatorType field
of PerformanceResourceTiming entries.
For resources corresponding to elements present in the initial page
response, the tests have been split into separate files grouped
(roughly) by the element type.
For resources corresponding to elements added to the page dynamically,
coverage has also been added for fonts, scripts and XMLHttpRequests.
Bug: 1171767
Change-Id: I547aab0db42d8f32bcbdfb310b16b4fddc103b18
GitHubIssue: https://github.com/w3c/resource-timing/issues/254
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2961075
Commit-Queue: Tom McKee <tommckee@chromium.org>
Reviewed-by: Yoav Weiss <yoavweiss@chromium.org>
Cr-Commit-Position: refs/heads/master@{#905732}
Co-authored-by: Tom McKee <tommckee@chromium.org>
--
wpt-commits: 0fd14d256e8e56863df9b23cf7f30137ddc1fc98
wpt-pr: 29786
Automatic update from web-platform-tests
Fix WPT linter error
In https://crrev.com/c/2961075 we migrated away from using
wpt/resource-timing/resources/all_resource_types.html but didn't remove the
file. We also updated wpt/lint.ignore to remove mention of
all_resource_types.html which caused downstream CI to fail
(https://github.com/web-platform-tests/wpt/pull/29786#issuecomment-892184957).
This change removes the unused file which, apart from cleaning up after
ourselves, should fix the downstream linter.
I've filed crbug.com/1236498 to follow up on getting chromium's presubmit to
keep us from making this mistake again.
Bug: 1236498
Change-Id: Id0556b08643b9c1994ce06877e4be6fb6e791fc4
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/3071040
Commit-Queue: Nicolás Peña Moreno <npm@chromium.org>
Auto-Submit: Tom McKee <tommckee@chromium.org>
Reviewed-by: Nicolás Peña Moreno <npm@chromium.org>
Cr-Commit-Position: refs/heads/master@{#908464}
--
wpt-commits: a44637c65403861a7756dd599d0cfb3de3947eaa
wpt-pr: 29891
Automatic update from web-platform-tests
Support relative selector to update the :has tests
Support the relative selector grammar starting with combinator.
- https://www.w3.org/TR/selectors-4/#typedef-relative-selector
To simplify matching operation, some relation types are added.
- kRelativeDescendant : Leftmost descendant combinator
- kRelativeChild : Leftmost child combinator
- kRelativeDirectAdjacent : Leftmost next-sibling combinator
- kRelativeIndirectAdjacent : Leftmost subsequent-sibling combinator
The ':scope' dependency in <relative-selector> definition creates
too much confusion especially with ':has' as the CSSWG issue
describes.
- https://github.com/w3c/csswg-drafts/issues/6399
1. ':scope' behavior in ':has' argument is different with usual
':scope' behavior.
2. Explicit ':scope' in a ':has' argument can create performance
issues or increase complexity when the ':scope' is not leftmost
or compounded with other simple selectors.
3. Absolutizing a relative selector with ':scope' doesn't make sense
when the ':has' argument already has explicit ':scope'
(e.g. ':has(~ .a :scope .b)' -> ':has(:scope ~ .a :scope .b)'
To skip those complexity and ambiguity, this CL removed some logic
related with the 'explicit :scope in :has argument', and added
TODO comment to handle it later separately.
As suggested in the CSSWG issue, this CL always absolutize the
<relative-selector> with a dummy pseudo class.
- kPseudoRelativeLeftmost
The added pseudo class represents any elements that is at the
relative position that matches with the leftmost combinator
of the relative selector.
This CL also includes tentative tests for some cases involving the
':scope' inside ':has' to show the result of the suggestion.
By removing the ':scope' dependency from the relative selector,
most of the ':scope' inside ':has' will be meaningless. (It will
not match or can be changed more simple/efficient expression)
Change-Id: I1e0ccf0c190d04b9636d86cb15e1bbb175b7cc30
Bug: 669058
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2972189
Reviewed-by: Rune Lillesveen <futhark@chromium.org>
Commit-Queue: Byungwoo Lee <blee@igalia.com>
Cr-Commit-Position: refs/heads/master@{#908421}
--
wpt-commits: 5800374cbcad335de936e9a30868b2e5e5340a9d
wpt-pr: 29718
Automatic update from web-platform-tests
Legacy block fragmentation support for ruby annotations.
Ruby annotations aren't part of any line box, so we need to include them
manually when checking whether a line fits in the current fragmentainer,
or whether it needs to be pushed to the next one.
This is already mostly working in LayoutNGBlockFragmentation, but one
of the new tests is failing there. This will be addressed in a follow-up
CL.
Bug: 631314
Change-Id: I9afd643a0d7b2ebc1b3053a111e1f10e256b3794
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/3065653
Commit-Queue: Morten Stenshorne <mstensho@chromium.org>
Reviewed-by: Koji Ishii <kojii@chromium.org>
Reviewed-by: Kent Tamura <tkent@chromium.org>
Cr-Commit-Position: refs/heads/master@{#908369}
--
wpt-commits: 0c37904c49048535d0a35a85dbc1db98d0a8acfe
wpt-pr: 29881
Automatic update from web-platform-tests
Make <base, <basefont>, and <embed> display:none
The spec [1] calls for these tags (along with several others) to
be display:none. Previously, these three (plus <area> - see below) were
missing from html.css. With this CL, I've updated html.css to more-
closely match the spec text. I also slightly changed the pre-
existing test to check for display:none only, and not display:inline
for <source> and <track>, because I could not find an explicit spec
for that. If I'm wrong about that, someone please point me to the spec
for those and I'll put the test back.
This CL leaves <area> unchanged for now, as it appears to break
hit-testing of <area> elements. This works, somehow, in WebKit,
while still having <area> elements display:none. TBD.
[1] https://html.spec.whatwg.org/multipage/rendering.html#hidden-elements
Bug: 1231263
Change-Id: Id67eed8a05a59174921b82783ced7958af8ff169
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/3050793
Commit-Queue: Mason Freed <masonf@chromium.org>
Auto-Submit: Mason Freed <masonf@chromium.org>
Reviewed-by: Joey Arhar <jarhar@chromium.org>
Cr-Commit-Position: refs/heads/master@{#908313}
--
wpt-commits: 7f84d2bfe8bc5f55e8fa4c9e05a7184e1bffc12a
wpt-pr: 29883
Jacek Kuzemczak
Olli Pettay
Jan de Mooij
Jonathan Kew
Emilio Cobos Álvarez
Geoff Brown
Ryan VanderMeulen
Julian Descottes
Dragana Damjanovic
Sebastian Hengst
Sebastian Streich
Kagami Sascha Rosylight
André Bargull
Paul Bone
Andrew McCreight
Andrew Osmond
Marian-Vasile Laza
Valentin Gosu
Nicolas Chevobbe
moz-wptsync-bot
Mathias Bynens
Moshe Atlow
Frédéric Wang
Rob Buis
Morten Stenshorne
Dan Clark
Sara Tang
Joey Arhar
Xiaocheng Hu
Mason Freed
Blink WPT Bot
Tom McKee
Byungwoo Lee
Marcos Cáceres