Automatic update from web-platform-testsRely on git hashes for manifest update
We earlier hashed all in-tree files when updating the manifest.
Relying on git hashes instead gives us a speedup of ~50%
when running |wpt manifest|
With reference to https://github.com/web-platform-tests/wpt/issues/11388
--
wpt-commits: 29663d0151dcd630df7f48066443ab6c8a953466
wpt-pr: 12182
Automatic update from web-platform-testsLocked mouse event check ShouldGenerateAppCommand
When pointer is locked, we still need to check ShouldGenerateAppCommand
and set event handled so that won't generate unnecessary navigate
appcommand when kExtendedMouseButtons is enabled. This causes unable
to preventDefault on mouse back/forward buttons to prevent navigation
while pointer is locked.
Bug: 852709
Change-Id: I5f75a68e36037210c38310f8d83825c5efdf311e
Reviewed-on: https://chromium-review.googlesource.com/1148886
Commit-Queue: Ella Ge <eirage@chromium.org>
Reviewed-by: Dave Tapuska <dtapuska@chromium.org>
Reviewed-by: Timothy Dresser <tdresser@chromium.org>
Cr-Commit-Position: refs/heads/master@{#577889}
--
wpt-commits: 1778b7179d12143b26c0651200257eca39dbc601
wpt-pr: 12183
Automatic update from web-platform-testsRefactoring XR mojo interfaces
Merging presentation provider and magic window provider to get rid of
duplicate code that fetches data. Changing request session so that it
must always return at least one interface for getting data, allowing
some uniformity of getting sessions across devices.
Bug: 842025
Cq-Include-Trybots: luci.chromium.try:android_optional_gpu_tests_rel;luci.chromium.try:linux_layout_tests_slimming_paint_v2;luci.chromium.try:linux_optional_gpu_tests_rel;luci.chromium.try:linux_vr;luci.chromium.try:mac_optional_gpu_tests_rel;luci.chromium.try:win_optional_gpu_tests_rel;master.tryserver.blink:linux_trusty_blink_rel
Change-Id: I587b314daf177619440a82f12471ea567b346caf
Reviewed-on: https://chromium-review.googlesource.com/1135831
Reviewed-by: Ian Vollick <vollick@chromium.org>
Reviewed-by: Daniel Cheng <dcheng@chromium.org>
Reviewed-by: Klaus Weidner <klausw@chromium.org>
Reviewed-by: Bill Orr <billorr@chromium.org>
Commit-Queue: Anna Offenwanger <offenwanger@chromium.org>
Cr-Commit-Position: refs/heads/master@{#577572}
--
wpt-commits: f9cb1ddc13c35945ac40c6e93f9ae25887dbd811
wpt-pr: 12150
Automatic update from web-platform-testsFix resources/test suite
The suite was broken by PR #11976 where the config class was changed.
The suite uses a hack to get the default config of wptserve. This commit
fixes the hack, but unfortunately does not find a less hacky way to get
the default config.
--
Run resources_unittest on CI if tools are changed
To prevent changes in tools (especially `wpt serve` and its
dependencies) from breaking the resources/test suite again, this commit
adds "tools/" to the watchlist of the suite.
--
wpt-commits: 2877f572684541d6cd2f301d72f4114dd35ce885, d03becfbbe9df94fb06c2a00ff9ad549634f3660
wpt-pr: 12174
Automatic update from web-platform-testsConsolidated fixes to character encoders/decoders (#11707)
* Fix EUC-JP reference decoder
The relevant step fixed by this pull request says "Return error."; thus, the rest of the process should continue with the next iteration, rather than run the rest of the handler for the given byte.
* Fix ISO-2022-JP reference implementation
The relevant step in the Encoding Standard says "Prepend _lead_ and _byte_ to _stream_." However, the two bytes are prepended in the wrong order in the reference implementation. Note that under the Encoding Standard, "[w]hen one or more tokens are prepended to a stream, those tokens must be inserted, _in given order_, before the first token in the stream." (Note that the code, at the time of this request, moves _lead_ to the front of the array, then moves _byte_ to the front of the array.)
There may be other issues like this elsewhere in the multiple-byte encoder reference implementations.
* Fix bug in EUC-KR reference implementation
Makes one "if" statement conditional rather than unconditional
* Fix bug in Shift_JIS encoder
* Fix bug in Big5 reference implementation
* Fix Shift_JIS reference decoder
To conform to the most recent Encoding Standard.
--
wpt-commits: 0589266b84851660a1c70efb916f5d34fa5ebd7c
wpt-pr: 11707
Automatic update from web-platform-tests[wptrunner] Fix regression in Sauce Labs browser
A recent commit [1] updated the invocation of the `start` method for
"Browser" instances by introducing a new keyword argument. This
interfered with the use of the "Sauce Labs" browser because its
implementation of the `start` method did not accommodate any keyword
arguments.
Update the `start` method to support (but ignore) all keyword arguments.
[1] 91803e3a25afd8e0c201bc57cd05fffd5d543551
--
wpt-commits: addfa0d8e05da60dbc24f0f9affe2e3841ad33f6
wpt-pr: 12170
Automatic update from web-platform-testsUpdated h2 server start with ssl_config being packaged into config
--
wpt-commits: cc33a7eb4f84a9299db18309cfcc5c5e147c67fc
wpt-pr: 12173
Automatic update from web-platform-testsChange TextMetrics to return a dictionary of baselines instead of separate values
Some fonts do not set values for all baselines (hanging, ideographic), and in
those cases we're setting 0, which is a valid value. This new interface
getBaselines(), returns a dictionary with the set baselines and their values.
Bug: 277215
Change-Id: Ie94c74d1c95c7205ed47acaed589ae965f0e32f6
Reviewed-on: https://chromium-review.googlesource.com/1141084
Reviewed-by: Chris Harrelson <chrishtr@chromium.org>
Reviewed-by: Fernando Serboncini <fserb@chromium.org>
Commit-Queue: David Quiroz Marin <davidqu@chromium.org>
Cr-Commit-Position: refs/heads/master@{#577670}
--
wpt-commits: 38cfd261a78ba9b99bd22a84ee69c7e95c19adcc
wpt-pr: 12068
Automatic update from web-platform-testsDon't wait for Picture-in-Picture when expecting reject on second request (#12164)
--
wpt-commits: 2a8d821bd0a29656ff27eac35d9dd071d27be779
wpt-pr: 12164
Automatic update from web-platform-testsRename a WPT which has a duplicate basename
as another test in the same directory.
Blink test runner isn't able to handle such duplicate basenames because
it relies on filename mangling extensively.
Bug: 866520
Change-Id: I0c9671e83dcf9c61bc20f16a43dcace827145024
Reviewed-on: https://chromium-review.googlesource.com/1146989
Reviewed-by: Ned Nguyen <nednguyen@google.com>
Commit-Queue: Robert Ma <robertma@chromium.org>
Cr-Commit-Position: refs/heads/master@{#577297}
--
wpt-commits: 9a716de4dfa97ff62dc0f78db55966bf0e88d902
wpt-pr: 12146
--HG--
rename : testing/web-platform/tests/html/rendering/non-replaced-elements/tables/form-in-tables.xhtml => testing/web-platform/tests/html/rendering/non-replaced-elements/tables/form-in-tables-xhtml.xhtml
Automatic update from web-platform-tests[css-logical] Replace uses of webkit-prefixed logical properties with standard ones in /third_party/WebKit/LayoutTests
Bug 850000 added standard logical properties and aliased prefixed ones
to them. The prefixed properties are still used in various places, but
the standard ones should be used instead. This patch replaces all the
uses in /third_party/WebKit/LayoutTests, except that the original
fast/css/logical-property-resolution.html is preserved in a new test
logical-property-resolution-webkit-prefix.html to ensure that prefixed
properties continue working.
Spec: https://drafts.csswg.org/css-logical/#box
BUG=862141
Change-Id: I3090713fc6fbe507a19724cca4f6ca69bcaf5eb9
Reviewed-on: https://chromium-review.googlesource.com/1147523
Commit-Queue: Oriol Brufau <obrufau@igalia.com>
Reviewed-by: Manuel Rego <rego@igalia.com>
Cr-Commit-Position: refs/heads/master@{#577530}
--
wpt-commits: ebeb9290184795bd3a7c4987ffe62b120f531e75
wpt-pr: 12152
Automatic update from web-platform-tests[infra] Remove code for generating config file (#12159)
Taskcluster-github v1 configuration is now working properly.
This commit reverts a revert c597cee7678a37f623dace0cd03d652b992daa36,
i.e. relands e62afac4c9dac57d8b81d2bf1364694e3d653ff9 to remove the code
that generates v0 configuration.
--
wpt-commits: 7952b4cc9dee381604f0816e39cfd34aeca50320
wpt-pr: 12159
Automatic update from web-platform-testsFix sort order of hostnames when generating certificates.
The names are expected to be sorted shortest first, but had the order reversed.
--
Convert the config to a builder pattern.
We have two sets of problems with the config:
* It has to go to some lengths to be thread/process safe since several
attributes are lazilly computed, and cached, by whichever thread
happens to be first to access them.
* The OpenSSL-based certificate implementation is totally broken
because this class needs to be used as a context manager in order to
initialise some resources associated with the fake CA used to
generate the certificates.
This suggests that the design is not really correct. We really want to
generate the configuration once in a single process, and then share the
generated configuration across multiple processes. We also want to
ensure that the ssl environment is initialised once in the main
process using a context manager.
This patch introduces a ConfigBuilder object which allows
configuration properties to be set. When used as a context manager, it
computes the generated configuration, initalises the SSL environment,
and updates the configuration with that data, and returns a config
object with the final, immutable, configuration which may then be
shared across processes.
--
Don't pass the sslconfig through server startup
This is fully obtainable from the config object.
--
Update make_hosts_file for new config
--
Update wptrunner for new config
--
Reduce the output when fetching the git history in ci
--
wpt-commits: 2957a8316b2a8f473de7f7c2e5e7243e8e74b81a, 020e53ba4c7c35bbcfa80699d160b76c482880f2, a6e33105a8335e101ec8c5f1cd6d76775e3a64f9, 590cb1745653d1b55d9be5c761fa950f999428d0, 3b02e29929bb0062d2c7b5a4ccfaae242bbc4c75, 86543323c57653f13529c74297c11547820c3563
wpt-pr: 11976
Looks like these used this mechanism for no great reason, and actually doing
this exposes trivially whether fingerprinting-resistance is enabled, which looks
like an anti-goal (if a media query parses correctly, and doesn't match either 1
or 0, then fingerprinting-resistance is enabled).
Differential Revision: https://phabricator.services.mozilla.com/D2493
MozReview-Commit-ID: 76tIIkwlpeP
Always assume allowed-for-all-content. There are a couple callers which weren't
doing that:
* A unit test -> removed.
* ComputeAnimationDistance: Used for testing (in transitions_per_property), and
for the animation inspector. The animation inspector shouldn't show
non-enabled properties. The transitions_per_property test already relies on
getComputedStyle stuff which only uses eForAllContent.
* GetCSSImageURLs: I added this API for the context menu page and such. It
doesn't rely on non-enabled-everywhere properties, it was only using
eInChrome because it was a ChromeOnly API, but it doesn't really need this.
Differential Revision: https://phabricator.services.mozilla.com/D2514
MozReview-Commit-ID: 4VOi5Su3Bos
The loop was mutating the nsCSSPropertyID used to guard the exit, which is
obviously wrong.
This branch is pretty rarely taken, since people don't usually specify `all` as
a transition property other than the first, for which case we take the fast path
with `checkProperties = false`. Our test-suite failed to catch this.
Added a crashtest that hangs without this patch.
The reason bug 1478990 regressed this is because it changed the order of
nsCSSPropertyID so that `p` actually went backwards causing the infinite loop,
but the bug was introduced (by me, whoops) in bug 1309752.
Differential Revision: https://phabricator.services.mozilla.com/D2552
MozReview-Commit-ID: Ii3D1FaZ31R
The loop was mutating the nsCSSPropertyID used to guard the exit, which is
obviously wrong.
This branch is pretty rarely taken, since people don't usually specify `all` as
a transition property other than the first, for which case we take the fast path
with `checkProperties = false`. Our test-suite failed to catch this.
Added a crashtest that hangs without this patch.
The reason bug 1478990 regressed this is because it changed the order of
nsCSSPropertyID so that `p` actually went backwards causing the infinite loop,
but the bug was introduced (by me, whoops) in bug 1309752.
Differential Revision: https://phabricator.services.mozilla.com/D2552
MozReview-Commit-ID: Ii3D1FaZ31R
--HG--
extra : source : 78be4bbf4b050f6614bb9f4115f57fb61f4890df
It was accidentally changed from 524288 to 32768 on Android by bug 1448222.
This commit changes it back.
--HG--
extra : rebase_source : 9181928dccf0cfcd0f70c1c3840d43c378f38ca4
This is generally helpful. I'm also seeing much less than 50% of the
wrQualified people in the study getting webrender so this could help
determine why.
MozReview-Commit-ID: 2neTFBytzPz
--HG--
extra : rebase_source : 97b1b1fe80f1a2c574141eaef8ad23499699fb82
This again exposes the accelerator (e.g. Ctrl+T for New Tab) for menu items via IAccessible::accName on windows.
Bug 1438193 made Accessible::Name const.
On Windows, XULMenuitemAccessibleWrap::Name is overridden to append the accelerator to the name.
Unfortunately, this override was missed when converting Name to const (and wasn't a compile error because it didn't use the override keyword).
As such, this override stopped being called; we always called the const version, since that's the one inherited into AccessibleWrap.
Making this const causes this override to be called.
MozReview-Commit-ID: 7Gb199mzhHa
--HG--
extra : rebase_source : 41d15042a09e4743d127fb12378c775099aa09a0
Luke Bjerring
Anna Offenwanger
Ahilya Sinha
Ella Ge
Robert Ma
Peter Occil
Ilya Grigorik
Mike Pennisi
moz-wptsync-bot
David Quiroz Marin
Philipp Hancke
L. David Baron
Oriol Brufau
Robert Ma
Emilio Cobos Álvarez
Devika Sugathan
Jan Varga
shindli
Jon Coppeard
Paolo Amadini
Jan de Mooij
Andrew Swan
Nicholas Nethercote
Julian Descottes
Jeff Muizelaar
Coroiu Cristina
James Teh
Kearwood Gilbert
Cosmin Sabou