- Mar 30, 2023
-
-
Hiroyuki Ikezoe authored
Differential Revision: https://phabricator.services.mozilla.com/D174042
-
Nathan LaPre authored
This revision adds logic to BoundsWithOffset to ensure that bounds, when calculated for hit testing, are constrained to the scroll areas that contain them. This ensures that we don't return an Accessible that's covered by another element when hit testing due to overflow: scroll situations. This revision also contains a fix for transforms and scroll: we now apply scroll before any transform, since transforms operate on scrolled content. This revision contains tests for both of the above changes. Differential Revision: https://phabricator.services.mozilla.com/D173392
-
- Mar 29, 2023
-
-
Emilio Cobos Álvarez authored
Differential Revision: https://phabricator.services.mozilla.com/D173862
-
Nico Grunbaum authored
Differential Revision: https://phabricator.services.mozilla.com/D174013
-
Emily McDonough authored
Bug 1824490 - Use the end page value rather than the start page value of the previous frame when performing named-page breakpoints r=dholbert Adding the test from this bug so that this can't happen next time. Differential Revision: https://phabricator.services.mozilla.com/D174037
-
Dan Baker authored
Depends on D173888 Differential Revision: https://phabricator.services.mozilla.com/D173987
-
Dan Baker authored
Depends on D173873 Differential Revision: https://phabricator.services.mozilla.com/D173888
-
Dan Baker authored
Depends on D173550 Differential Revision: https://phabricator.services.mozilla.com/D173873
-
Dan Baker authored
Depends on D173788 Differential Revision: https://phabricator.services.mozilla.com/D173550
-
Dan Baker authored
Differential Revision: https://phabricator.services.mozilla.com/D173788
-
Eitan Isaacson authored
Differential Revision: https://phabricator.services.mozilla.com/D173864
-
Shane Hughes authored
Bug 1813716 - Encapsulate aboutwelcome container and make its ID more unique. r=omc-reviewers,fxview-reviewers,sfoster,jprickett Differential Revision: https://phabricator.services.mozilla.com/D172493
-
Andrew Osmond authored
Bug 1825415 - Update GMP fallback downloader for OpenH264 1.8.1.2. r=media-playback-reviewers,azebrowski We never updated the fallback for 1.8.1.2, which was a false version bump to update the binaries for OSX. This patch updates the URLs, file sizes and hashes for the fallback path. Differential Revision: https://phabricator.services.mozilla.com/D174034
-
Emilio Cobos Álvarez authored
Bug 1824957 - Remove a bunch of now completely dead XUL layout code. r=layout-reviewers,desktop-theme-reviewers,devtools-reviewers,dao,TYLin And some related tests. A bunch of -moz-box* properties are web exposed, so I'll file a follow-up for hiding them. Differential Revision: https://phabricator.services.mozilla.com/D173819
-
Erich Gubler authored
As evidenced by a [recent Try push](https://treeherder.mozilla.org/jobs?repo=try&revision=7e1b556bed392abe2ba3bf7e461259421e44a728&selectedTaskRun=eDGDIk81RH6Yyz2-W8vAhg.0), my [workaround solution](https://phabricator.services.mozilla.com/D173484) for [bug 1824042](https://bugzilla.mozilla.org/show_bug.cgi?id=1824042) was wrong for a couple of reasons: 1. `os == "windows"` aggravatingly works for me locally, but _not_ in CI. 2. Squashing the crashes in the offending tests noted in the original workaround solution merely moved the crashing behavior's effects to other tests. I think it's safe to say that we have a more complicated issue than a single reftest here. So...let's just disable these reftests on ~~`windows`~~ `win` for now. Differential Revision: https://phabricator.services.mozilla.com/D173910
-
Tiaan Louw authored
This devtools test was assuming the color-mix flag is enabled, which it is on nightly, but not on beta/release, so checking the flag during the test. Differential Revision: https://phabricator.services.mozilla.com/D173876
-
James Teh authored
Bug 1821448: Improve handling of reentry in MsaaAccessible::VisitDocAccessibleParentDescendantsAtTopLevelInContentProcess. r=nlapre Differential Revision: https://phabricator.services.mozilla.com/D173652
-
Jim Blandy authored
The `about:support` page formatter for graphics information takes a JavaScript object (the `graphics` property of the snapshot returned by `Troubleshoot.snapshot`), and builds HTML covering each property of the object. It deletes properties as it goes, and at the end, any remaining properties get dumped into a "diagnostics" section. Fix the WebGPU adapter reporting code to delete the properties it reports, so that they don't end up generating stray rows in the diagnostics section. Differential Revision: https://phabricator.services.mozilla.com/D173751
-
Tom Schuster authored
Differential Revision: https://phabricator.services.mozilla.com/D173959
-
az authored
Bug 1822416 - Upgrade typed_ast package to 1.5.4 to fix compilation errors running ./mach lint with Python 3.11 r=sylvestre Please see comments in bug report for more details. I'm unfamiliar with this part of the codebase / typed_ast, so hopefully somebody with more experience can take a look at this. Differential Revision: https://phabricator.services.mozilla.com/D172611
-
Narcis Beleuzu authored
Backed out 2 changesets (bug 1746245, bug 1824447) for xpcshell failure on test_FeatureGate.js . CLOSED TREE Backed out changeset 5cc2daf4f55a (bug 1746245) Backed out changeset 754743cbc36b (bug 1824447)
-
Mozilla Releng Treescript authored
de -> 8b4f44c1036f1ff7be177dbaa09504f3e05a519c hsb -> e10e1c561e9e7173034ee140d1d65248a0db46ba oc -> b3cfc5fd4dae8975286dacdd1bebdbee1c451d3a skr -> 5828ceaeb70a1c54671a2cfe669c217db1eb989b
-
Emilio Cobos Álvarez authored
The spacers are unnecessary if there's a <label> or <description> that flexes. Since we allow labels/descriptions to shrink now they'd distribute the space along with the spacer, which caused unnecessarily small labels. Differential Revision: https://phabricator.services.mozilla.com/D173937
-
Hubert Boma Manilla authored
Bug 1811000 - [devtools] Move the project search to the side panel r=devtools-reviewers,nchevobbe,perftest-reviewers,sparky Differential Revision: https://phabricator.services.mozilla.com/D172985
-
Gavin Suntop authored
Differential Revision: https://phabricator.services.mozilla.com/D174001
-
Ebilite Uchenna authored
Differential Revision: https://phabricator.services.mozilla.com/D173926
-
Mike Conley authored
Bug 1801981 - Use the word "favorites" instead of "bookmarks" when importing from IE / Edge. r=niklas,fluent-reviewers,flod Differential Revision: https://phabricator.services.mozilla.com/D173567
-
Sammy Khamis authored
Differential Revision: https://phabricator.services.mozilla.com/D173369
-
Iain Ireland authored
Differential Revision: https://phabricator.services.mozilla.com/D173894
-
Iain Ireland authored
Differential Revision: https://phabricator.services.mozilla.com/D167106
-
Bilal authored
Differential Revision: https://phabricator.services.mozilla.com/D173571
-
CanadaHonk authored
Differential Revision: https://phabricator.services.mozilla.com/D173994
-
Julien Cristau authored
Differential Revision: https://phabricator.services.mozilla.com/D173979
-
Julien Cristau authored
oak was reset and moved to level2 in bug 1819778 so it shouldn't use prod signing now. Differential Revision: https://phabricator.services.mozilla.com/D173977
-
Ray Kraesig authored
Bug 1820268 - make nsI{Input,Output}Stream::Close() idempotent r=xpcom-reviewers,nika,necko-reviewers,valentin The `nsIAsync{Input,Output}Stream` base classes' Close() methods are already documented to be idempotent. Do the same for their synchronous counterparts, and make the necessary changes to ensure this. The only known functional change here (modulo the removal of some uninteresting logging) is to `nsFileInputStream`, which now avoids calling `Tell()` if the stream is already closed. (The other early-exits added here are formally redundant, but neither obviously nor robustly so.) This silences some console warning spam when running a debug build from the command line. Differential Revision: https://phabricator.services.mozilla.com/D171615
-
Jim Blandy authored
Change the default value of `dom.webgpu.enabled` from `false` to `true` in Nightly builds. Everything in `dom/webidl/WebGPU.webidl` that is `Exposed=(Window)` is also gated on the `dom.webgpu.enabled` pref, so it should not appear on `window` in beta or release builds. Beyond that, WebGPU is not usable in beta or release builds regardless of the pref setting, because of these lines of code in `dom/webgpu/Instance.cpp`: #ifdef RELEASE_OR_BETA if (true) { return "WebGPU is not yet available in Release or Beta builds."; } #endif This function is the only way to create a `GPUAdapter`, which is the only way to access any interesting parts of the WebGPU API. The mochitest `dom/webgpu/mochitest/test_disabled.html` checks that the value of the `dom.webgpu.enabled` pref is `false`. This is now only appropriate in beta and release builds, so change `dom/webgpu/mochitest/mochitest-no-pref.ini` accordingly. Differential Revision: https://phabricator.services.mozilla.com/D173475
-
Jim Blandy authored
Bug 1824447: Make WebGPU interfaces available only in secure contexts. r=webgpu-reviewers,webidl,saschanaz,jgilbert Add the [`[SecureContext]`](https://webidl.spec.whatwg.org/#SecureContext) extended attribute to all interfaces in `dom/webidl/WebGPU.webidl` and `dom/webidl/GPUUncapturedErrorEvent.webidl`, to match the specification. Update mochitests to run in secure contexts. It's standard practice now for newly introduced content-visible web APIs to be available only in secure contexts, to encourage the adoption of https. WebGPU's API was marked in this way in gpuweb#1363. Firefox's WebIDL needs to be updated to match. Differential Revision: https://phabricator.services.mozilla.com/D173584
-
Gijs Kruitbosch authored
Differential Revision: https://phabricator.services.mozilla.com/D173941
-
Tom Ritter authored
Differential Revision: https://phabricator.services.mozilla.com/D173566
-
Tom Ritter authored
Differential Revision: https://phabricator.services.mozilla.com/D173565
-