This bug became an unexpected PASS on tier-2 android platform after
landing the fixes for bug 1572197. The fixes in that bug were
specifically to fix this test on windows platforms, so it's not
surprising that it also fixes it on android.
Differential Revision: https://phabricator.services.mozilla.com/D41493
--HG--
extra : moz-landing-system : lando
Automatic update from web-platform-tests
Clean-up: Updated URLs in comments referencing web-share and web-share-target.
- Web Share spec has moved from WICG to W3C.
- Updated some Web Share Target URLs to point to the WICG spec, not the
explainer.
- Updated some Web Share Target URLs to point to the WICG spec, not a
draft on the pr-preview site.
- Removed some notes that say Web Share Target is part of an unlaunched
experiment.
Change-Id: I48132388192d773d962d13bc9361778ddeb66e4e
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1742258
Reviewed-by: Dominick Ng <dominickn@chromium.org>
Reviewed-by: Ted Choc <tedchoc@chromium.org>
Commit-Queue: Matt Giuca <mgiuca@chromium.org>
Cr-Commit-Position: refs/heads/master@{#685457}
--
wpt-commits: 40c12bca52d4a695eb36e2887d93675251a3e3ef
wpt-pr: 18331
Automatic update from web-platform-tests
[cookie-store] Deflake service worker registration
Instead of attempting to unregister previously installed service workers
at the beginning of a test, use `add_completion_callback` to unregister
them at test completion. This will deflake the tests when they are run
multiple times without restarts (`--rerun`).
The alternative is to do what many tests in service-workers do:
unregister service workers at the start, but wait for newly installed
service workers to become activated and use `registration.active`
instead of `registration.installing`, which is a bit more complex.
--
wpt-commits: cc9dc6bb113ac7e04aee6b8e4157fbe8eb2d731c
wpt-pr: 18324
Automatic update from web-platform-tests
Use correct URLLoaderFactory in an unload handler.
Context and behavior implemented + expected before and after the CL
===================================================================
A content::RenderFrameImpl stores |subresource_loader_factories| that
are used for all requests/XHRs/etc initiated by the frame. We don't
currently swap the RenderFrame when committing another same-process
document (this future work is tracked by https://crbug.com/936696).
The URLLoaderFactory for handling http/https requests contains a
|request_initiator_site_lock| in network::URLLoaderFactoryParams.
This lock limits what network::ResourceRequest::request_initiator may be
used. The lock is currently used in various, security-sensitive
features like CORB, CORP, Sec-Fetch-Site. In the future the lock may be
consulted for all requests handled by the NetworkService (see
https://crbug.com/920634 and also https://crbug.com/961614).
When a cross-origin, same-process navigation commits, then:
- The commit IPC carries a bundle of |subresource_loader_factories|.
The new |subresource_loader_factories| should be used by the newly
committed document.
- Before committing the new document, the old document's unload
handler needs to run. The unload handler needs to use the old
|subresource_loader_factories| (e.g. because otherwise
|URLLoaderFactoryParams::request_initiator_site_lock| might
not match the document origin).
The problematic behavior before this CL
=======================================
Before the CL, content::RenderFrameImpl::CommitNavigationWithParams
would start using the new |subresource_loader_factories| before running
the unload handler of the old document (i.e. before calling
blink::WebNavigationControl::CommitNavigation on |frame_|).
The CL adds WPT tests that fail when the old, incorrect behavior
happens. The new test initiates a beacon request from a frame's unload
handler and verifies the Sec-Fetch-Site request header associated with
this request. The header depends on the correct
request_initiator_site_lock / on using the correct URLLoaderFactory.
The fixes in this CL
====================
This CL introduces the |call_before_attaching_new_document| callback
that is taken by blink::WebNavigationControl::CommitNavigation and
called *after* finishing running the old document's unload handler
and *before* the new document can initiate any requests of its own.
This fix is a bit icky. In the long-term the callback can be avoided if
the |subresource_loader_factories| are stored in a per-document object
(i.e. if we swap RenderFrame on every document change, or if we replace
RenderFrame with a RenderDocument - see https://crbug.com/936696). The
CL adds a TODO to call this out.
The fix had to refactor BuildServiceWorkerNetworkProviderForNavigation
to make sure that it uses the |new_loader_factories| (even though they
have not yet been passes to SetLoaderFactoryBundle).
The fix also refactored GetLoaderFactoryBundleFromCreator to a separate
method, so that setting the creator-based factories can also be
postponed to |call_before_attaching_new_document|.
Bug: 986577
Change-Id: If0209df61b0305ec43b5179bfdc1b9f8668a88b5
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1716084
Reviewed-by: Wei Li <weili@chromium.org>
Reviewed-by: Tommy Li <tommycli@chromium.org>
Reviewed-by: Yutaka Hirano <yhirano@chromium.org>
Reviewed-by: Kinuko Yasuda <kinuko@chromium.org>
Commit-Queue: Łukasz Anforowicz <lukasza@chromium.org>
Cr-Commit-Position: refs/heads/master@{#685290}
--
wpt-commits: c0b0a0d72066faf8838420b601bcd99d94842b22
wpt-pr: 18061
Automatic update from web-platform-tests
Don't apply filter or clip-path in SVGForeignObjectPainter.
SVG foreign objects always have PaintLayers now, since they are
stacking contexts. The filter and clip-path are applied by
PaintLayerPainter.
Therefore we shouldn't try to apply them in SVGForeignObjectPainter,
nor should we early-out if they "fail" to apply.
Bug: 989673
Change-Id: I4a20da8f8ea85d377b25d8376d591f36f172d854
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1743256
Commit-Queue: Chris Harrelson <chrishtr@chromium.org>
Reviewed-by: Stephen Chenney <schenney@chromium.org>
Cr-Commit-Position: refs/heads/master@{#685285}
--
wpt-commits: 2a0b3d690c0b56aeb189dc3042f4b36e1acd0564
wpt-pr: 18332
Automatic update from web-platform-tests
Update tests to check XRInputSource profiles + gamepad.id
Per the latest WebXR spec, XRInputSource.gamepad.id must be the empty
string. Update the browser and layout tests to actually check this.
Also update the browser tests that run on Windows to check the profiles
array contents against expected values for WMR and OpenVR.
Bug: 989114, 942201
Change-Id: I36dff42cff55728d5d90384623623ea871d3a13e
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1742439
Commit-Queue: Jacob DeWitt <jacde@chromium.org>
Reviewed-by: Brian Sheedy <bsheedy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#685270}
--
wpt-commits: d328e5763f3d76e8b5f9e62030e9d4e578931c78
wpt-pr: 18330
Automatic update from web-platform-tests
CSS: Test height and width do not inherit
WPT to verify the width and height properties do
not inherit.
https://drafts.csswg.org/css-sizing-3/#propdef-width
Change-Id: Ib35db7d4d5b55a5bc1b56db38eff49ec0fb4c427
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1741785
Reviewed-by: Emil A Eklund <eae@chromium.org>
Commit-Queue: Emil A Eklund <eae@chromium.org>
Auto-Submit: Eric Willigers <ericwilligers@chromium.org>
Cr-Commit-Position: refs/heads/master@{#685251}
--
wpt-commits: 9c1737440fc7001ea43189e6a56a8e04905401e1
wpt-pr: 18327
Automatic update from web-platform-tests
CSS: Test dominant-baseline initial value 'auto'
The CSS Inline Layout spec has been updated to
show the initial value of dominant-baseline as
'auto', matching SVG 1.1 and browsers.
https://drafts.csswg.org/css-inline/#dominant-baseline-propertyhttps://github.com/w3c/csswg-drafts/issues/4115
Change-Id: Ib2dbdf91a9535c4d149aa773b61302c36c360305
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1741445
Auto-Submit: Eric Willigers <ericwilligers@chromium.org>
Commit-Queue: Emil A Eklund <eae@chromium.org>
Reviewed-by: Emil A Eklund <eae@chromium.org>
Cr-Commit-Position: refs/heads/master@{#685218}
--
wpt-commits: b1fd38d1568441ff1d6a93c04a7eedb5458cf30a
wpt-pr: 18325
Automatic update from web-platform-tests
[Contacts] Add contacts wpt tests with a mock contacts implementation.
- Create a mocked contacts interface that requires user agents to set
the contacts to return when queried.
- Implement the mocked contacts interface for chromium.
- Move chromium web_tests to the wpt folder. I've added one test and
removed one test.
Change-Id: I8fe518eef37eedf2908b4b0ea9bec9df9ea4938f
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1741915
Commit-Queue: Rayan Kanso <rayankans@chromium.org>
Reviewed-by: Peter Beverloo <peter@chromium.org>
Cr-Commit-Position: refs/heads/master@{#685182}
--
wpt-commits: 3c2d193651aad8eabad20b0bf870614f0df9c4c8
wpt-pr: 18319
This change adds a new export, SendUIMessageToVRWindow, from vrhost.dll
that allows the caller to forward a subset of UI messages to the VR
window in Firefox.
Differential Revision: https://phabricator.services.mozilla.com/D40828
--HG--
extra : moz-landing-system : lando
Automatic update from web-platform-tests
Catch all promise rejections so the test harness doesn't fail due to uncaught rejection
--
wpt-commits: 5c3c67adc2cfce655b6ec0e65b3e4b785590e821
wpt-pr: 18329
Automatic update from web-platform-tests
Test coverage for excluding application/x-www-form-urlencoded from CORB.
In addition to covering application/x-www-form-urlencoded, this CL
takes the opportunity to also cover application/octet-stream which
was identified long time ago in a Firefox bug.
This CL is a follow-up to https://crbug.com/990853, but won't need to be
merged to release branches, so it doesn't refer to this bug below.
Bug: 802836
Change-Id: Idc95a1b070f4b4ecce391eb4f173196b89d989fa
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1739749
Commit-Queue: Charlie Reis <creis@chromium.org>
Auto-Submit: Łukasz Anforowicz <lukasza@chromium.org>
Reviewed-by: Charlie Reis <creis@chromium.org>
Cr-Commit-Position: refs/heads/master@{#684988}
--
wpt-commits: 2a71f56c7f40ad03c66f9c1d0b9aea1f461f3e03
wpt-pr: 18301
Automatic update from web-platform-tests
[LayoutNG] Fix invalidation issue with grids.
with: https://chromium-review.googlesource.com/c/chromium/src/+/1554113
We introduced a new flag for SelfNeedsLayoutForAvailableSpace. For any
alignment change this was set to true for grid-items.
SelfNeedsLayout reports true if this flag is set, or if the normal
SelfNeedsLayoutForStyle is true.
This caused an issue where we weren't performing layout on the whole tree
however.
If a child of the grid-item (whose alignment changed) was marked for
needing style-layout, it would try and mark its container chain for
layout, hit the grid item, and abort, as it already "needed" layout.
Within NG we'd come across the grid-item, hit the cache, or perform
simplified layout, never clearing layout, having a bad ptr, etc.
This is a conservative fix so that this can be merged.
Bug: 991446
Change-Id: I75edebc864eb470ebfad757966ddc0e0248de153
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1741439
Commit-Queue: Ian Kilpatrick <ikilpatrick@chromium.org>
Commit-Queue: Emil A Eklund <eae@chromium.org>
Auto-Submit: Ian Kilpatrick <ikilpatrick@chromium.org>
Reviewed-by: Emil A Eklund <eae@chromium.org>
Cr-Commit-Position: refs/heads/master@{#684966}
--
wpt-commits: 65cb7eabbbc54c79dd01ae7e4acf458585f5d079
wpt-pr: 18316
Automatic update from web-platform-tests
[LayoutNG] First part of ShrinkToFit for flex items
Two places in the spec call for treating an item's inline size as
'fit-content' under certain conditions. This patch implements that
logic.
4 tests now pass and 2 "regress": FlexNG previously rendered the
scrollbars tests and their references so incorrectly that the first
800x600 pixels matched but most of rest of the output actually did not
match. Both the tests and their references are rendered better now, but
the differences now show up in the first 800x600 pixels.
Bug: 845235
Change-Id: I60de7ba4a26e54f6f171ba8a5e05299c3e953452
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1737742
Reviewed-by: Christian Biesinger <cbiesinger@chromium.org>
Commit-Queue: David Grogan <dgrogan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#684952}
--
wpt-commits: 99fda55bb52ae85162644dbeebce68de271c62da
wpt-pr: 18287
Automatic update from web-platform-tests
[Media Session] Implement position state API
Implement the new setPositionState API in Blink.
BUG=984620
Change-Id: I7a6eb1849f4e4afba2fb51771a6485aa58a7cdaf
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1730773
Reviewed-by: Daniel Cheng <dcheng@chromium.org>
Reviewed-by: Mounir Lamouri <mlamouri@chromium.org>
Commit-Queue: Becca Hughes <beccahughes@chromium.org>
Cr-Commit-Position: refs/heads/master@{#684943}
--
wpt-commits: b3624301279fb51242dc04e955e7c516b45a89d4
wpt-pr: 18240
Automatic update from web-platform-tests
[css-layout-api] Convert to promise based API.
As above! This switches the API from being generator based, to promise
based.
This introduces the specification's work queue, which is run until
exhaustion. After the work queue has been run the given promise should
be resolved, (if not we fallback to block layout).
This also introduces the CustomLayoutScope, and CustomLayoutToken
which are used together.
The CustomLayoutToken is a heap allocated class held onto objects which
should only be used within a certain layout pass. E.g. you cannot
hold onto a fragment from a previous layout pass, and use it within
the current layout pass.
This is managed by a stack allocated CustomLayoutScope, which once
destroyed marks the CustomLayoutToken as "detached". Any object which
has a "detached" token shouldn't be used.
The CustomLayoutScope is also used to hold onto the work queue.
Bug: 726125
Change-Id: Ic435c8c50fe3d3779f068b41eada815253d0b78b
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1716035
Commit-Queue: Ian Kilpatrick <ikilpatrick@chromium.org>
Reviewed-by: Koji Ishii <kojii@chromium.org>
Cr-Commit-Position: refs/heads/master@{#684927}
--
wpt-commits: dd38623d0edc9005cf87397feb2e011eb0679c52
wpt-pr: 18297
Automatic update from web-platform-tests
[docs] Correct errors in build process
Ensure that the contents of the top-level `certs/` directory are
available to the Sphinx process at build time.
Correct references to recently-relocated document from a new tutorial.
--
wpt-commits: 0a6beaa1610aa22683d2991811af98f587780764
wpt-pr: 18317
