Rebase - #72
Merged
Merged
Conversation
…ests/css/css-view-transitions/massive-element-below-viewport-offscreen-old.html (#60291) https://bugs.webkit.org/show_bug.cgi?id=305737 Co-authored-by: Cameron McCormack <heycam@apple.com>
This is an incremental step (1/2) towards full negative playbackRate support in AudioBufferSourceNode. Pre-calculating grain_duration_ assuming forward playback fails when playbackRate dynamically flips to negative, rendering the fixed boundary invalid. This CL instead sets the duration to infinity when unspecified, allowing the node to stop organically at the buffer boundaries regardless of the playback direction. It also enables negative clamping in ComputePlaybackRate. the-audio-api/the-audiobuffersourcenode-interface/ audiobuffersource-playbackrate-dynamic-direction.html Bug: 40719652 Test: third_party/blink/web_tests/external/wpt/webaudio/ Change-Id: I87c7dd404fa168d908374c7c13bc9a668a1281f7 Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7876495 Reviewed-by: Michael Wilson <mjwilson@chromium.org> Commit-Queue: Mahesh Kannan <kmaheshb@google.com> Cr-Commit-Position: refs/heads/main@{#1638840}
Add column and row versions of all the subgrid gap tests that exist in grid. This requires handling subgrid extra margins (due to margin, border and/or padding at subgrid edges, or due to subgrid gaps) in subgridded item contributions during track sizing. This will be implemented in a follow up change. A chunk of tests are marked as failing as a result. Note that some of the refs may not be 100% correct yet, but I will make sure to double check all of the expectations as part of implementing the actual fix (which will result in differing behavior from grid). Bug: 343257585 Change-Id: I34eec0170a99978d8df78f1b66fde15b08b120b7 Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7861306 Reviewed-by: Kurt Catti-Schmidt <kschmi@microsoft.com> Reviewed-by: Kevin Babbitt <kbabbitt@microsoft.com> Commit-Queue: Alison Maher <almaher@microsoft.com> Cr-Commit-Position: refs/heads/main@{#1638874}
Differential Revision: https://phabricator.services.mozilla.com/D296935 bugzilla-url: https://bugzilla.mozilla.org/show_bug.cgi?id=2005007 gecko-commit: 567230e612806099ad17c6fa9bb704bc0b7e6034 gecko-commit-git: 5f837dd775a3152d21f541d9f58b130987243368 gecko-reviewers: firefox-style-system-reviewers, emilio, dshin
Differential Revision: https://phabricator.services.mozilla.com/D302984 bugzilla-url: https://bugzilla.mozilla.org/show_bug.cgi?id=2043078 gecko-commit: 517f32cbaabce6a555e9bc6d96d8c996b48b346f gecko-commit-git: b0ad9c3d5cb27c0b0fa477ff6c4b3cd65902372c gecko-reviewers: sajidanwar
Thanks to the recent changes by Sajid and the changes in bug 2043078, which make color components properly return numbers for type-checking purposes, this is now trivial. We could add an "allowed but only numbers" or so, but type checking will basically deal with it (in fact we should probably allow parsing ~everything all the time and rely on unit() doing the type-checking). Differential Revision: https://phabricator.services.mozilla.com/D303375 bugzilla-url: https://bugzilla.mozilla.org/show_bug.cgi?id=1951206 gecko-commit: 84b5d08285154e0c7b1feb1604dc5302ca2ae024 gecko-commit-git: 8a7d2d52762439c5a429d6a73e0c272046d309e2 gecko-reviewers: sajidanwar
Canvas should be tainted when a pattern contains cross origin image content. Verify that this is the case. Also add a test for a canvas filter that uses an SVG feImage, and a drawElementImage case for a pattern that contains a feImage. Bug: 514609778 Change-Id: Iea19f30665f4cd114e678259cb18063fd3ae60bb Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7880261 Reviewed-by: Philip Rogers <pdr@chromium.org> Auto-Submit: Stephen Chenney <schenney@chromium.org> Commit-Queue: Stephen Chenney <schenney@chromium.org> Cr-Commit-Position: refs/heads/main@{#1639075}
This exposes some interop issues in <select>. Blink and WebKit hit the equivalent of https://issues.chromium.org/issues/40571175 so they fail the "no slot" tests, even though they indeed do not have a slot.
Add testing for additional display keyword combinations that are supported by the current CSS specification but reify as unsupported valued in Typed OM. Differential Revision: https://phabricator.services.mozilla.com/D301787 bugzilla-url: https://bugzilla.mozilla.org/show_bug.cgi?id=2041086 gecko-commit: 236706981b3a5c6ecaa47e5c34392d547ab049b7 gecko-reviewers: emilio, firefox-style-system-reviewers, Oriol
…ial about:blank. (#60003) Closes #60002
…tly for both fragments and feImageElements See https://www.w3.org/TR/media-frags/#valid-uri-spatial Differential Revision: https://phabricator.services.mozilla.com/D303672 bugzilla-url: https://bugzilla.mozilla.org/show_bug.cgi?id=2041697 gecko-commit: 8a14073c7a1bebaa3f738e3913bf08cea2d7d1b7 gecko-reviewers: firefox-svg-reviewers, jwatt
…ctly Also fix a mess in WebVTT due to e7c5d8b which didn't update the .dat file and instead percent-encoded manually. First step towards #27868.
Signed-off-by: Richard Tjokroutomo <richard.tjokro2@gmail.com>
Signed-off-by: Richard Tjokroutomo <richard.tjokro2@gmail.com>
This test demonstrates an interop bug with dialog.requestClose() and closedby.
This patch rewrites the flaky `test-analyser-output.html` WPT test to be 100% deterministic by replacing the legacy, real-time ScriptProcessorNode-based verification with OfflineAudioContext. - Replaced legacy `AudioContext` and `ScriptProcessorNode` with `OfflineAudioContext` running at both 44.1kHz and 48kHz. - Replaced `helpers.js` usage with direct WPT `promise_test` and `assert_approx_equals` element-by-element comparison. - Removed the flaky test expectation from TestExpectations. TAG=agy CONV=1685e534-e673-43a5-8c7a-8d8084e0ee42 Bug: 40170551 Test: external/wpt/webaudio/the-audio-api/the-analysernode-interface/test-analyser-output.html Change-Id: I90300a4443b4328c8330b80917cff4a4da76aec8 Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7886243 Reviewed-by: Mahesh Kannan <kmaheshb@google.com> Commit-Queue: Hongchan Choi <hongchan@chromium.org> Cr-Commit-Position: refs/heads/main@{#1639405}
Parsing of <grid-line> differs depending on if the span ident is present or not. Previously we'd parse an integer up front, but at this stage we don't know the correct range. Instead refactor the code (using savepoints) to match the specification more closely. Bug: 515750295 Change-Id: I036a88f6971c609f3131303f2cd875da51c3644c Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7879775 Reviewed-by: Anders Hartvoll Ruud <andruud@chromium.org> Commit-Queue: Ian Kilpatrick <ikilpatrick@chromium.org> Reviewed-by: Rune Lillesveen <futhark@chromium.org> Cr-Commit-Position: refs/heads/main@{#1639406}
In `GiveItemsFinalPositionAndSizeForFragmentation`, when a row flex item breaks across a fragmentainer, we subtract the already-advanced offset from `line_cross_size` to get the remaining cross-size for the current fragment. When the row is bounded by the parent (e.g. max-height: 0) but the row's stitched offset is bumped past its content size to align with a fragmentainer break, this subtraction makes the resulting cross size negative. The negative value is then passed as the block available_size to `BuildSpaceForLayout`, leading us to hit a DCHECK. Clamp the result to zero to ensure the cross size never goes negative. Bug: 515776659 Change-Id: Ibea3ffc5ebf2744d55bb3c10bf8a94cfa9c71a28 Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7884136 Commit-Queue: Alison Maher <almaher@microsoft.com> Reviewed-by: Morten Stenshorne <mstensho@chromium.org> Cr-Commit-Position: refs/heads/main@{#1639431}
The HTML parse will process the is= attribute and put the element into custom element state. When later on the is= attribute is removed by the Sanitizer, the element is still upgraded because it's been marked as a custom element. This change prevents the is=-related custom element to be processed, when the streaming Sanitizer will remove it. Bug: 492963096, 513844247, 517171036 Change-Id: I7b10f55b6dd9d568336c731ee8a8c2a52254371f Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7883146 Reviewed-by: Noam Rosenthal <nrosenthal@google.com> Commit-Queue: Daniel Vogelheim <vogelheim@chromium.org> Cr-Commit-Position: refs/heads/main@{#1639435}
For subgrids whose final position isn't known at sizing time (e.g. inside a grid-lanes ancestor that places after sizing), the accumulated edge extra margins propagated to a nested subgrid must account for the largest contribution the subgrid could pick up at any candidate position. Previously we only forwarded the parent's accumulated edge margins, which under-reserved space when a parent gap could land immediately outside the subgrid's boundary. Update `CreateSubgridTrackCollection` so that when the subgrid is auto-placed in its parent, the start/end accumulated extra margins are computed as follows: - If the subgrid spans every track of the parent, no parent gap can ever be adjacent to the subgrid's boundary; forward the parent's accumulated edge margins as-is. - If the parent has exactly one more track than the subgrid, the subgrid can only ever have a parent gap on one side. Reserve the half-gutter contribution on whichever side currently carries the smaller accumulated edge margin so we don't double-count. - Otherwise, take max(accumulated edge margin, parent gutter delta / 2) on both sides, since any combination of start/end positions is possible. Add a bit more test coverage, although a follow up change will be needed to get the true largest final accumulation numbers, which will be handled in a follow up change and will likely lead to some changes to existing test expectations in some cases. Bug: 343257585 Change-Id: I0742fef87af1008ba43a07f557f31281c24344c7 Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7897846 Reviewed-by: Ian Kilpatrick <ikilpatrick@chromium.org> Commit-Queue: Alison Maher <almaher@microsoft.com> Cr-Commit-Position: refs/heads/main@{#1644722}
This WPT was failing on Windows because of forced rounding from DWrite. Though certainly not ideal, that is not the space this test is attempting to cover, so this change updates the test to ensure the initial underline position expected from the font is a whole number, so that the test is only testing the text-underline-offset impact, and not the platform-specific font-reading code. For more details: The expected values for close_underline (UNDO=200) and far_underline (UNDO=1000) are 2.15625 and 9.375 respectively (if my math is correct), which when rounded to final pixels become 2 and 9. The starting underline position for the font (default UNDO=500) comes to 4.6875px, but DWrite returns this underlinePosition as an INT16, which in this case is rounded to 5. This means that on Windows, after applying the text-underline-offsets of -2.5px and 4.6px, the positions became 2.5px and 9.6px, which when rounded to final pixels became 3 and 10 (instead of the expected 2 and 9). Other platforms do not encounter this problem because they either return the underline position from the font in a float, or in more granular units. This change overrides the default UNDO=500 with UNDO=640, which comes to exactly 6px. Fix: 1140324 Change-Id: I5845eb8d258f0879427b9f203e063937afb9a035 Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7913911 Commit-Queue: Hoch Hochkeppel <mhochk@microsoft.com> Reviewed-by: Kurt Catti-Schmidt <kschmi@microsoft.com> Reviewed-by: Dominik Röttsches <drott@chromium.org> Cr-Commit-Position: refs/heads/main@{#1644739}
ServiceWorkerHost is responsible for brokering WebSocket and WebTransport connections within the worker. For proper integration with Connection Allowlist, the host now provides the ServiceWorkerVersion's network_restrictions_id when creating those connections. Note that once this CL is merged, WebSocket and WebTransport should have full coverage across documents, dedicated, shared, and service workers. Bug: 492462310 Fixed: 492462310 Change-Id: If9700eef886901e5ad20ea02981637f0c5c1c016 Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7902842 Reviewed-by: Shivani Sharma <shivanisha@chromium.org> Commit-Queue: Andrew Verge <averge@chromium.org> Reviewed-by: Hiroki Nakagawa <nhiroki@chromium.org> Cr-Commit-Position: refs/heads/main@{#1644755}
This change allows an accelerated video frame to be painted into a display list which is passed to another thread, and then rasterized from the other thread. The downside is that the PaintCanvasVideoRenderer cache is effectively disabled, so if a single video frame is painted multiple times it will be forced to use multiple GPU shared images. This seems like an unlikely circumstance and fixing it would be complicated, so this CL punts. Bug: 435230899 Change-Id: Ic51cac7aba434412293e0b72437dc0bf4a5edd31 Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7900160 Reviewed-by: Vasiliy Telezhnikov <vasilyt@chromium.org> Reviewed-by: Philip Rogers <pdr@chromium.org> Commit-Queue: Stefan Zager <szager@chromium.org> Reviewed-by: Dale Curtis <dalecurtis@chromium.org> Commit-Queue: Philip Rogers <pdr@chromium.org> Cr-Commit-Position: refs/heads/main@{#1644771}
This CL modifies blink::AnimationTrigger's play-forwards and play-backwards behaviors so that they do not auto-rewind if the animation is already finished in the target direction. CSSWG resolution: w3c/csswg-drafts#13643 (comment) We are getting this PSA ready to accompany this change: https://chromestatus.com/feature/5071640598806528 Bug: 390314945 Change-Id: Id9bcb06c8e0833141c5a61feb8dbec5a0397bfd6 Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7902733 Commit-Queue: David A <awogbemila@chromium.org> Reviewed-by: Kevin Ellis <kevers@chromium.org> Cr-Commit-Position: refs/heads/main@{#1644772}
Bug: 521861823 Change-Id: Idd76c57c027c2f7e2585c1525ab14932deea2c9e Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7919507 Commit-Queue: Kevin Ellis <kevers@chromium.org> Reviewed-by: David A <awogbemila@chromium.org> Cr-Commit-Position: refs/heads/main@{#1644815}
This is a reland of commit 6c3c7a1a61db01bcf82d2658a30b3632619c3b92 This reland updates the WPT expected results on Mac, as the mac15-x64-rel-tests bot caught some failures in the newly added tests. Original change's description: > WebNN: Fix Pool2d ceil_mode to check both spatial dimensions > > The previous code only compared the height dimension's floor output > size against the WebNN output height to determine ceil_mode. When > only the width dimension required ceiling rounding (e.g., input > [1,1,5,5], window [3,2], strides [2,2], roundingType:"ceil"), the > height check passed with floor, so ceil_mode=0 was emitted. > > This caused a shape mismatch between Chromium's WebNN operand > descriptors (e.g., [1,1,2,3]) and ORT's actual execution shape > (e.g., [1,1,2,2]). The mismatch invalidated gather/scatter index > sanitization: indices valid for the WebNN shape could be out of > bounds for ORT's smaller tensor, enabling OOB read/write on > execution providers that do not validate indices at dispatch time. > > Fix: compute CalculateConv2dOutputSize for both height and width, > and set ceil_mode=1 when either dimension needs ceiling rounding. > ONNX Pool's ceil_mode is a single global attribute, so setting it > when any dimension requires ceiling is both correct and safe (ceiling > of an integer is itself). > > This CL also added some conformance tests for this scenario. > > Bug: 519241091 > Change-Id: I9dc92f5f81687286eaf6aca8468032fe152360d7 > Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7901552 > Reviewed-by: Hu, Ningxin <ningxin.hu@intel.com> > Reviewed-by: Reilly Grant <reillyg@chromium.org> > Commit-Queue: Xu, Mingming1 <mingming1.xu@intel.com> > Cr-Commit-Position: refs/heads/main@{#1643553} Bug: 519241091,521419439 Change-Id: I4ec18fc7d0fb91d64a292178e2630c437d0bb76d Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7909424 Commit-Queue: Hu, Ningxin <ningxin.hu@intel.com> Reviewed-by: Reilly Grant <reillyg@chromium.org> Reviewed-by: Hu, Ningxin <ningxin.hu@intel.com> Cr-Commit-Position: refs/heads/main@{#1644829}
If the <reference-box> for overflow-clip-margin was content-box we didn't invalidate our clip if the padding changed. Mark the LayoutBox as needing a paint-property update for this case. Change-Id: I009e2ca39161e2d6aba8cb2390b78444376aa559 Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7914267 Reviewed-by: David Baron <dbaron@chromium.org> Commit-Queue: Ian Kilpatrick <ikilpatrick@chromium.org> Cr-Commit-Position: refs/heads/main@{#1644874}
This UA style rule was preventing overflow:hidden from doing anything when set from author styles. Fixed: 521434907 Change-Id: I2d41b122b76a82c02efda61bfdb9fdc524f91951 Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7919309 Reviewed-by: Ian Kilpatrick <ikilpatrick@chromium.org> Reviewed-by: Joey Arhar <jarhar@chromium.org> Commit-Queue: Joey Arhar <jarhar@chromium.org> Cr-Commit-Position: refs/heads/main@{#1644873}
This gives us better coverage. Differential Revision: https://phabricator.services.mozilla.com/D304961 bugzilla-url: https://bugzilla.mozilla.org/show_bug.cgi?id=2045230 gecko-commit: 4bca04155b988c0d98e01f10429567e89bb7db96 gecko-commit-git: 0c8c2d7a8ddb354e0c37f7024e77691d54ede11a gecko-reviewers: gfx-reviewers, lsalzman
This reduces the effect and frequency of #51981, where we frequently saw timeouts from the H2 server, because we no longer repeatedly create new threads and new queues when we receive further events after StreamEnded, or an event with a reference to a StreamEnded. See the state diagram in https://httpwg.org/specs/rfc9113.html#StreamStates, noting that receiving StreamEnded only closes the connection when the stream is in the half-closed (local) state. This also means we treat a stream as closed when we send/receive a RST_STREAM frame.
For simpler usage with Lightpanda WPT test runner, we override the default testharnessreport.
Some tests define their own format (and maybe name) function in the global namespace, which causes issues. If testharnessport.js is last, then those tests break. If it comes before, then the report gets invalid name/format.
Some tests define their own update function which causes a conflict, e.g. css/geometry/support/dommatrix-test-util.js.
Contributor
Author
|
@krichprollsch I suggest running |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
20 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.