Until recently, beta-YYYY-MM-DD dated channels was the only way to get
specific betas, but now it's possible to use x.y.z-beta.n, but
repack_rust.py doesn't handle that string as a version.
By considering only channels starting with beta- or nightly to be dated,
we allow x.y.z-beta.n to be considered as a version.
Differential Revision: https://phabricator.services.mozilla.com/D213367
When building cargo sets up the patches for the vendored dependencies of
libstd, when several of those dependencies are different versions of the
same crate, only one of the corresponding patches work because the
package name, derived from the directory name, doesn't match.
Instead, we always take the package name from Cargo.toml.
While we're here, we add the missing resources on the rust-dev tasks.
Differential Revision: https://phabricator.services.mozilla.com/D213380
We were trigerring one click leading to a console.info call a bit before the failing assertion.
It was most likely trigerring a late console info resource.
Differential Revision: https://phabricator.services.mozilla.com/D213238
Add a new fetchpriority-urgency-destination.https.h2.html test that
is essentially combining features for
/fetch/api/request/destination/fetch-destination.https.html
/fetch/fetchpriority/fetchpriority-urgency.h2.html
in order to test urgency of a global fetch with a destination. Also add
a comment in fetch-tests-data.js to explain why we can't do a direct
testing of the internal priority.
Differential Revision: https://phabricator.services.mozilla.com/D210286
According to step 15 of [1], request’s destination is used in an
implementation-defined manner to set the internal priority but we
currently always use the preferences corresponding to the adjustments
of the global fetch API.
This patch handle destinations corresponding to CSP directives
"font-src", "script-src", "style-src", "img-src" and "media-src" [2]
specially by respectively using the adjustment preferences named
"link-preload-font", "link-preload-script", "link-preload-style"
"images" and a new preference "media".
Special handling of these destinations is motivated by the fact that
they are mentioned explicitly on [3]. For "media" we choose preferences
following the suggestions from that article (i.e. media is low
priority). However, for now we don't change the default value and make
sure fetchpriority=high/low preserves the invariants tested by
fetchpriority-adjustments.html
[1] https://fetch.spec.whatwg.org/#concept-fetch
[2] https://fetch.spec.whatwg.org/#destination-table
[3] https://web.dev/articles/fetch-priority?hl=en#browser_priority_and_fetchpriority
Differential Revision: https://phabricator.services.mozilla.com/D208467
This patch makes sure the RequestPriority of the internal request is
propagated to the child process when using PFetch here:
https://searchfox.org/mozilla-central/rev/14ff5b1650123600698f4bb8998ee02b102b8d2f/dom/fetch/Fetch.cpp#632
Note that IPCInternalRequest is also used in ServiceWorkerPrivate when
intercepting a request. However, propagating the RequestPriority of the
intercepted request does not seem desirable, so we just initialize with
PRIORITY_AUTO.
One version of the constructor for InternalRequest is no longer used, so
we just remove it.
Differential Revision: https://phabricator.services.mozilla.com/D210281
runSingleTest() from mozilla/tests/fetch/fetchpriority/fetchpriority.js
waits for a ChildLoaded message for each test before verifying the
requests and priorities. However, the fetch-init.h2.html test sends
on ChildLoaded message for each request, which would make subsequent
tests (currently none) fail. This patch fixes that issue by waiting for
all the fetch requests to resolve/reject before sending a single message.
Differential Revision: https://phabricator.services.mozilla.com/D210142
Automatic update from web-platform-tests
webnn: add npu virtualtestsuite for Mac and Windows
Skip the tests for Windows on CQ until we have infra support. But still
include Windows on VirtualTestSuite to make it easy for local runs.
Change-Id: I56a910a0883694f875c82ab88c0cadd218f17c9e
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5583198
Auto-Submit: Phillis Tang <phillis@chromium.org>
Reviewed-by: Austin Sullivan <asully@chromium.org>
Commit-Queue: Phillis Tang <phillis@chromium.org>
Reviewed-by: Robert Flack <flackr@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1312213}
--
wpt-commits: af50ef9e4cf21f89639f8719601060f96077997d
wpt-pr: 46659
Automatic update from web-platform-tests
Remove tests for color-contrast() that use a syntax no longer in the spec (#46656)
Fixes#46640.
--
wpt-commits: be8e91ef743dfe1b85559d14ca8fa6f6b97f9974
wpt-pr: 46656
Automatic update from web-platform-tests
Specify a viewport for resolving userSpaceOnUse filterUnits
For non-SVG content, SVGViewportResolver won't be able to produce a
viewport, but will return 0x0 which turn means that every part of the
filter region that is percentages computes to zero. Instead allow
setting up a viewport in FilterEffectBuilder and pass that along when
resolving the filter region. For HTML elements use the border box size
as the viewport (matches Gecko) and for <canvas> use the canvas
dimensions.
Guard behind the flag SvgFilterUserSpaceViewportForNonSvg.
Fixed: 41189886
Change-Id: Ib385340bdc4d30d6d17c447fb99d6db3bc66e8b7
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5594220
Reviewed-by: Xianzhu Wang <wangxianzhu@chromium.org>
Commit-Queue: Fredrik Söderquist <fs@opera.com>
Cr-Commit-Position: refs/heads/main@{#1311889}
--
wpt-commits: 11518fa184cb824c60302c931182d6e97ded5bdb
wpt-pr: 46653
Automatic update from web-platform-tests
[text-box-trim] Fix when `line-clamp` is applied
This patch fixes the `text-box-trim` behavior when
`line-clamp` or `-webkit-line-clamp` is applied.
All changes are under a runtime flag enabled only for testing.
Bug: 40254880
Change-Id: Ib8ead4e083b727a3be3c6b36ae5cc42f85ba01f2
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5606785
Auto-Submit: Koji Ishii <kojii@chromium.org>
Reviewed-by: Kent Tamura <tkent@chromium.org>
Commit-Queue: Kent Tamura <tkent@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1311806}
--
wpt-commits: 921085c5e8d1ac91e60d2d46791d71ae2d247dfd
wpt-pr: 46652
Automatic update from web-platform-tests
Add internal WPT for SW static routing API
This CL adds internal WPT for SW static routing API to test the behavior
when the router evaluator could not parse the regular expression
provided by developers. Since this is due to the regex framework
re2, chrome raises a PromiseError when encountering this situation.
Bug: 41492364
Change-Id: I13573d5b7060ab55a51727a66a17acfd5ec34d0c
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5595296
Reviewed-by: Yoshisato Yanagisawa <yyanagisawa@chromium.org>
Reviewed-by: Yoshisato Yanagisawa <yyanagisawa@google.com>
Commit-Queue: Keita Suzuki <suzukikeita@chromium.org>
Reviewed-by: Shunya Shishido <sisidovski@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1311067}
--
wpt-commits: b604a3dfa7a2bfc0f7474784af5cc077f13a2ea3
wpt-pr: 46641
Automatic update from web-platform-tests
Fenced frame: Set referrer header for all event-level beacons. (#46611)
A previous CL set the "Referer" header for reportEvent() beacons to the
frame's origin. This CL sets that header for automatic beacons as well,
and puts the change behind a new feature flag. Automatic beacons from
component ads will not have their "Referer" header set to the frames
origin for privacy reasons, and instead will be set to the root ad
frame's origin. To track this, ad component configs will now store the
origin of the root ad frame's config on the browser side.
This CL also updates WPTs and other tests to account for the changes.
Previous CL:
https://chromium-review.googlesource.com/c/chromium/src/+/5551871
Design Doc:
https://docs.google.com/document/d/1gKRZ9g_X5HCZbW__GinViwcC8iWEai_kBXsiJCXrt80/edit?usp=sharing
Change-Id: Idfaeee92ad5bf471b34002a6a701290d1ae31cf1
Bug: 341884774
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5544804
Commit-Queue: Liam Brady <lbrady@google.com>
Reviewed-by: Garrett Tanzer <gtanzer@chromium.org>
Reviewed-by: Alex Moshchuk <alexmos@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1310677}
Co-authored-by: Liam Brady <lbrady@google.com>
--
wpt-commits: cfac7a30e755a4530812a091705f64dd5d97228f
wpt-pr: 46611
Automatic update from web-platform-tests
Expose InteractionID to Keydown Entries in FID (#46622)
Given EventTimingKeypressAndCompositionInteractionId is fully launched
(though slowly rollout on webview, but soon), interactionID is now
generated upfront on keydown entries in Event Timing. As a result, we
can now easily expose that to FID to reduce developer confusions when
matching same entries between FID and Event Timing.
Bug: 40840075
Change-Id: I25c5d7dd05ef61915779acd926005fc10aa62a77
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5597879
Commit-Queue: Aoyuan Zuo <zuoaoyuan@chromium.org>
Reviewed-by: Michal Mocny <mmocny@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1310208}
Co-authored-by: Aoyuan Zuo <zuoaoyuan@chromium.org>
--
wpt-commits: a4787fec82370958a8901b8d793ac175acd3d42d
wpt-pr: 46622
Automatic update from web-platform-tests
[wptrunner] Mark WebDriver protocol as not alive in other cases (#46621)
* [wptrunner] Mark WebDriver protocol as not alive if `OSError` raised
This is a follow-up improvement to #45833. `Protocol.is_alive()` should
handle a browser that's unable to even establish the low-level
transport. `OSError` is a sufficiently broad base exception, which
encompasses `socket.timeout`, `IOError`, and the `Connection*Error`
family.
* Coerce all `WebDriverException`s to not alive
* Update safari infrastructure expectation
--
wpt-commits: 965dd6e03f9eb4fda9a553b7950cb496b68fb396
wpt-pr: 46621
Automatic update from web-platform-tests
Fix the "attributes" key for testharness canvas element grid tests
No test currently uses custom context attributes in a canvas grid, so
this typo was left undetected.
Bug: 340005429
Change-Id: Ibab7f92cc43939e13700bbe5dc5846db98776ef0
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5598921
Reviewed-by: Andres Ricardo Perez <andresrperez@chromium.org>
Commit-Queue: Jean-Philippe Gravel <jpgravel@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1310850}
--
wpt-commits: 706749d8b964ffeda099492d1b4cc6b02a8d77bd
wpt-pr: 46637
Automatic update from web-platform-tests
Fenced frame: fix parseFromString() crash in sandboxed iframes.
If parseFromString() is called inside a sandboxed iframe's DOMParser on
an HTML string that contains a <fencedframe> element, a fenced frame
will be created behind, including running the sandbox flags check. That
will fail, and upon logging, it will attempt to check if a frame is the
main frame, which requires getting the document's frame. Since the
fenced frame isn't attached to any DOMWindow at this point, the frame
will be null.
The existing code dereferences the frame without checking its nullness,
causing a crash. The fix switches that call to
`Document::IsInMainFrame()`, which includes a nullness check on the
frame. This CL also adds a minimal reproduction case as a test.
Bug: 40277893
Bug: 344029039
Change-Id: I53ca200b405a3f60ba4f64962ddc1b19950824a9
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5601581
Commit-Queue: Liam Brady <lbrady@google.com>
Reviewed-by: Garrett Tanzer <gtanzer@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1310781}
--
wpt-commits: a0ed93d67828fdd54f03b6e8f38c065a75e165b5
wpt-pr: 46638
Automatic update from web-platform-tests
Add a variant_name key to the WPT canvas test Jinja template
The Jinja template for WPT canvas tests is currently provided with a
variant_names list which stores the name of each variant dimensions.
This is useful to get the name of a specific dimension in a
multi-dimensional variant test, for instance in a 2d variant grid:
https://crsrc.org/c/third_party/blink/web_tests/external/wpt/html/canvas/tools/yaml-new/filters.yaml;l=620;drc=869e38fe2f8c29a863c92b775a3549a1adae37f8
Most variant tests however use a single dimension, so having to use a
variant_names list is cumbersome and can look strange to someone
unfamiliar with multi-dimensional variants (having to do
variant_names[0] to access the variant name is surprising).
To simplify test implementation, this CL adds a "variant_name" string
which stores a concatenation of all variant dimension names (e.g.
"variantA1.variantB1"). For tests with a single variant dimension,
variant_name is a shorthand for variant_names[0].
Bug: 340005429
Change-Id: Ic04db7dd687dc145e8622356ea958f905c375c64
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5597104
Reviewed-by: Andres Ricardo Perez <andresrperez@chromium.org>
Commit-Queue: Jean-Philippe Gravel <jpgravel@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1310775}
--
wpt-commits: f743feccdcecf17d894266bd9e82c6a385cacfdb
wpt-pr: 46636
Automatic update from web-platform-tests
webnn: Remove support for passing MLActivations for op fusion
Remove the ability to specify op fusions from the web platform layer.
Ops can still be fused, but this is now the responsibility of the
respective WebNN backend
Corresponds to the spec PR merged here:
https://github.com/webmachinelearning/webnn/pull/664
Bug: 341518634
Cq-Include-Trybots: luci.chromium.try:win11-blink-rel,mac14.arm64-blink-rel,mac14-blink-rel
Change-Id: I8a4f2ab708e34087eeb10421596243b17bc8d7e2
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5494397
Reviewed-by: ningxin hu <ningxin.hu@intel.com>
Reviewed-by: Alex Gough <ajgo@chromium.org>
Commit-Queue: Austin Sullivan <asully@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1310724}
--
wpt-commits: e3235d9662ee81945c7ff6242fc8fe0628c509c8
wpt-pr: 45937
Automatic update from web-platform-tests
[webdriver] Fix WebDriver error dunder methods (#46623)
* `super()` alone retrieves the superclass type. Calling `__init__()`
on it actually runs the superclass constructor.
* `BidiException` doesn't have an `error` attribute. This probably
should have been `error_code`.
--
wpt-commits: 638346974f0af854dc502ad23964ea4ab2f99493
wpt-pr: 46623
Automatic update from web-platform-tests
Fuzzing wpt tests to deal with scrollbar anti-alias.
Follow-up to CL:5362506
Adding fuzziness to more tests that are failing due to anti-aliasing
differences between composited and main-threaded Fluent scrollbars in
Windows. This CL adds fuzzy acceptance to the tests so that they'll
not fail due to this reason, which is not the main purpose of the
tests in question.
These tests were obtained by filtering the failing wpt tests in a
test CL that enables Fluent scrollbars by default and makes them paint
rounded thumbs (instead of square as they are now): CL:5593606
Bug: 326116366, 329199184
Change-Id: I701637ce3c1be0864009731548c2b7d0b1ef6ef1
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5597461
Reviewed-by: Yaroslav Shalivskyy <yshalivskyy@microsoft.com>
Reviewed-by: Olga Gerchikov <gerchiko@microsoft.com>
Commit-Queue: Gaston Rodriguez <gastonr@microsoft.com>
Cr-Commit-Position: refs/heads/main@{#1310656}
--
wpt-commits: a822e3518c728412681b542c8783da77c774b3f2
wpt-pr: 46625
ComplexTranslateIter is leftover from bug 1716933.
ClipNodeIndex is leftover from bug 1779952.
BlobImageRasterizerEpoch is leftover from bug 1587713.
Differential Revision: https://phabricator.services.mozilla.com/D213358