For some reason, with clang >= 16 stack traces involving
"XPCWrappedNativeTearOff::AddTearOff" now show a non-namespaced
"AddTearOff" instead. Add that to the allowed list.
Differential Revision: https://phabricator.services.mozilla.com/D174196
Changes in inlining decision in clang >= 16 make it so that the stacks
that the mozilla::net::AddStaticElement exception is meant to catch
doesn't match anymore because mozilla::net::AddStaticElement is not in
the stack trace anymore. Use the caller, InitializeStaticHeaders,
instead.
Differential Revision: https://phabricator.services.mozilla.com/D174193
Changes in clang >= 16 change the stack trace for leaks that were
previously caught via "already_AddRefed", to one that starts with
"MakeAndAddRef". In both cases, though, the caller is
"NS_NewCancelableRunnableFunction", which is less broad to add an
exception on.
Differential Revision: https://phabricator.services.mozilla.com/D174195
On Linuxen without getrandom(), Rust (and Firefox broadly) uses /dev/urandom
as a fallback. Rust uses it for a few things, notably hashmaps... and will
panic if it can't use it.
Differential Revision: https://phabricator.services.mozilla.com/D174336
***
Bug 1821106 - SearchEngine.getResultDomain's responseType parameter is never used and the function needs renaming ,Comment is updated r=standard8
Differential Revision: https://phabricator.services.mozilla.com/D172293
Use PNGs instead of BMPs for Windows toast notification images.
BMPs were isolated as the cause of Windows toasts failing to display images when provided in the `src` attribute of the `<image>` tag for Windows 8. The same was not observed for PNGs.
This issue only affects Windows 8.
Differential Revision: https://phabricator.services.mozilla.com/D173787
`IToastNotification2` is not available in Windows 8, so querying it caused the toast to fail. When the content was only text, the toast would immediately fail triggering a fallback to XUL style alerts. When it contained an image, the async image load caused the toast to fail after the fallback branch had been decided.
The toast tag assigned and retrieved from `IToastNotification2` is used by Windows to track notifications in the Action Center and is only relevant in the context of Windows 10+.
Differential Revision: https://phabricator.services.mozilla.com/D173496
Adds "forced-color-adjust: none" to the search tooltip psuedo styles in preferences.css to disable any backplating in forced color mode.
High contrast color are now specified in a forced-colors media query.
Differential Revision: https://phabricator.services.mozilla.com/D173781
Tests failed:
- single-line-column-flex-fragmentation-062.html is adapted from
single-line-column-flex-fragmentation-061.html, with flex-direction value
changed from column to column-reverse.
- multi-line-row-flex-fragmentation-073.html is adapted from
multi-line-row-flex-fragmentation-072.html, with flex-wrap value
changed from wrap to wrap-reverse.
Tests passed with Part 2.
- single-line-column-flex-fragmentation-063.html tests `justify-content: end` so
that there are empty flex container fragments at the beginning of the columns.
- single-line-column-flex-fragmentation-064.html tests `justify-content: space-between` so
that there are empty flex container fragments in the middle columns.
- multi-line-row-flex-fragmentation-074.html tests `align-content: space-between;` so
that there are empty flex container fragments in the middle columns.
- multi-line-row-flex-fragmentation-075-print.html and
multi-line-row-flex-fragmentation-076-print.html are adapted from Daniel
Holbert's testcase in bug 1743890 comment 18, testing a item shift bug occurred
in Part 2 WIP.
- single-line-row-flex-fragmentation-043.html tests `align-self: start` and
`align-self: end` on items so that there are empty flex container fragments in
the middle columns.
Differential Revision: https://phabricator.services.mozilla.com/D167978
This patch deals with two things:
1. Push tall monolithic flex items to next-in-flow, and adjust their positions.
2. Grow flex container's block-size if its block-size is unconstrained.
This patch doesn't fix:
1. Item shifts in different lines in a multi-line column-oriented container
(bug 1806717).
2. Flex container block-size grow due to flex item's block-size grow in
fragmentation.
If a flex item has break-before:avoid, we don't want to push it to the
next-in-flow (in the computaion of `shouldPushItem`). Otherwise, we'll fail
web-platform/tests/css/css-break/flexbox/multi-line-column-flex-fragmentation-034.html
Differential Revision: https://phabricator.services.mozilla.com/D165192
MoveFlexItemToFinalPosition() already has a log printing flex item's position.
This patch adds a log in ReflowFlexItem() to print flex item's position, too.
Differential Revision: https://phabricator.services.mozilla.com/D165191
Emily McMinn
Daniel Holbert
Cristian Tuns
Mike Hommey
Chris H-C
Mark Banner
Abhishek Tiwari
Noemi Erli
Iain Ireland
Mozilla Releng Treescript
Kelsey Gilbert
elephanteatsapple
Heitor Neiva
Stephen A Pohl
Nicholas Rishel
Fred Chasen
Niklas Baumgardner
Stephanie Cunnane
Erich Gubler
serge-sans-paille
Ting-Yu Lin
Doug Thayer
steven wang
Florian Quèze