We don't check for pkg-config on some platforms (Windows, OSX, Android).
On those platforms, calling 'pkg_check_modules' will not work. Adding
the same 'when' used for the 'pkg_config' check to all the options that
end up calling 'pkg_check_modules' effectively disables them, and prevents
'pkg_check_modules' from being called.
Differential Revision: https://phabricator.services.mozilla.com/D150649
This patch adds 4 new categories of tests to the perf selector: DAMP tests, Talos tests, Graphics, & Media Playback, and Resource Usage tests. Each of these use a different suite (talos), and one of them also makes use of AWSY. Some extra code needed to be added, and the existing setup that we use for categories had to change to handle the case of a category using multiple suites.
The Resource Usage category has the most going on in it. A few new fields are added to allow categories to specify/restrict where they can run (platform, variants, and app).
Differential Revision: https://phabricator.services.mozilla.com/D162871
'Lax' as value depends on the preference network.cookie.sameSite.laxByDefault
which is only set to true for Nightly and Early Beta (bug 1751435). See
bug 1617509 for the general development of the feature.
Differential Revision: https://phabricator.services.mozilla.com/D162720
Using ipc::Shmem causes unbounded shmem use growth until e.g. a Worker
yields to the event loop. If a Worker never yields, Shmems sent to WebGLParent
are never released.
Specifically the manager (PCanvasManager) for WebGLParent calls
DestroySharedMemory, which sends/enqueues for WebGLChild's manager a
matching call to ShmemDestroyed. However, while WebGLChild refuses to spin its
event loop (such as a no-return WASM Worker), the ShmemDestroyed events
will just pile up. Closing e.g. the tab frees the shmems, but they accumulate
unbounded until the Worker yields to the event loop.
This is true for other users of ipc::Shmem (or RaiiShmem) as well, but
entrypoints other than DispatchCommands are rarer and can be handled
later similarly.
Differential Revision: https://phabricator.services.mozilla.com/D162946
Some random thoughts:
* This is a minimal implementation - there were plans to expand this with autocomplete support that we never got around to
* I built this just from reading docs + poking around in the urlbar code, so there might be a more desireable way to do all of this
* We've gone back and forth between calling this "contextual search" and "site specific search," we should probably pick a name and stick to it
* There were never any design specs for this so the UI/UX was based off of existing urlbar results as they appeared in Flowstate
* There were some open questions around how we treat installing opensearch engines, since how they are handled here seems to differ from how we use them elsewhere
* I don't remember the reasoning for showing this result last in the urlbar, we may want to move it elsewhere
Differential Revision: https://phabricator.services.mozilla.com/D159052
Now hal uses AR_STATE for screen orientation, so we should move it from
WinUtils.cpp to WinUtils.h for newer Windows 11 SDK.
Also, newer Windows SDK has to set WINVER to WIN8 to use auto rotation API.
Differential Revision: https://phabricator.services.mozilla.com/D163099
By reproducing with increased test loops and in verify mode, by far the primary leak was from the iframe element's "src" attribute. So clearing it.
Differential Revision: https://phabricator.services.mozilla.com/D162911
And also making it and `BuildCSSDeclarations` return pairs of CSS property and
value with an array makes their callers simpler.
Depends on D162517
Differential Revision: https://phabricator.services.mozilla.com/D162518
It's unclear to me when they were disabled, but we do want to enable
these as otherwise there's no way to catch bindgen issues that can end
up in subtle bugs at best, or memory corruption at worst.
Differential Revision: https://phabricator.services.mozilla.com/D162997
To debug the idle trigger's interaction with OS sleep state, it's
helpful to know whether sleep/wake notifications were sent. This allows
most of the idle trigger activity to be logged to the console, if the
pref `messaging-system.log` is set to `debug`. The new logging will also
be used to debug the behavior reported in bug 1801301.
Differential Revision: https://phabricator.services.mozilla.com/D163010
When using images with drag & drop (or the clipboard) we use two or more different types:
- When dragging an image we put it into the DataTransfer object using the
type kNativeImageMime and the data as an imgIContainer object.
- When dropping an image we want to have some kind of image/ type,
but the data is already supposed to be a File or nsIInputStream when it comes from the OS.
This weird split usually works fine, because in our GTK, Windows and macOS code
we first convert imgIContainer to whatever native image format is required when
dragging something and then convert back to a file (or stream) when dropping.
What however doesn't work is when we never actually round-trip through the OS.
In that case we have the imgIContainer that we can't drop in a meaningful way.
(I actually already ran into this issue in bug 1396587 with the clipboard,
but we kind of resolved this by always doing the roundtrip through the OS's clipboard)
Differential Revision: https://phabricator.services.mozilla.com/D162493
