This matches how the `Dispatch(already_AddRefed<nsIRunnable>)`
overloads work in C++: `Dispatch` takes ownership of the runnable, and
leaks it if dispatch fails—because the thread manager is shutting down,
for instance. This avoids a race where a runnable can be released on
either the owning or target thread.
Rust doesn't allow arbitrary `Self` types yet (see
rust-lang/rust#44874), so we need to change `dispatch` and
`dispatch_with_options` to be associated methods.
Differential Revision: https://phabricator.services.mozilla.com/D75858
Some background information on why we want to flush writes: we're
currently working on a "fast shutdown" project, which will simply
call _exit(0) once a certain phase (controlled via pref) in
shutdown is hit. This will allow us to skip a lot of ceremony,
like freeing memory, removing listeners on observer notifications,
etc. However, we don't want to be cavalier about data loss, so
for the time being we want to ensure that all file writes which we
previously had some assurance would complete will continue to
complete. This means explicitly calling them immediately before a
fast shutdown. We also want to call these before we turn on late
write checking, which we control via another pref. This late write
checking surfaces writes via telemetry which we didn't notice with
local testing. However we want to ensure that we flush our writes
before this point so we don't have lots of noise in the telemetry.
Lina I'm requesting review from you since it looks like you did the
reviews on most of the initial code. However, please let me know if
there's someone more suitable.
Depends on D75389
Differential Revision: https://phabricator.services.mozilla.com/D75572
When performing the last GC, remaining state watchers would dispatch direct tasks ; this was exposed with P3. Ensure this never happens.
Differential Revision: https://phabricator.services.mozilla.com/D75499
During the last cycle collection, state watchers may attempt to dispatch tasks to the current AbstractThread, so we can't clear the TLS entry until that step has completed.
We want the object to be deleted last; ordering in call to ClearOnShutdown() can't guarantee it.
Differential Revision: https://phabricator.services.mozilla.com/D75498
In this bug we're moving away from monolithic JNI headers to class-specific
headers so that we don't have to rebuild the world every time we make a change
to a JNI interface.
Differential Revision: https://phabricator.services.mozilla.com/D75366
We already ensure that we finish our write before doing a fast
shutdown, but right now there's just too much noise in the late
write checking telemetry because of this.
Differential Revision: https://phabricator.services.mozilla.com/D75389
Gecko Idle detection relies heavily on RefreshDriver. GC/CC scheduling, including when to run GC or CC slices, and the length of the slices, is mostly based on idle time. As WebXR isn't using normal RefreshDriver, the content process thinks it's idle and GC and CC get basically up to 50 ms slices.
Not having WebXR idle detection is causing stutter frames during immersive presentation. This patch implements idle deadline hint and sets the correct GC slices budgets during WebXR presentation.
Differential Revision: https://phabricator.services.mozilla.com/D74426
Bug 1536796 introduces "\\?\" prefix (DOS device specifier) to Windows file
paths. At the moment, the prefix is only prepended to the file paths that start
with a disk designator and a backslash.
On the other hands, IsBlockedUNCPath blocks file paths that start with "\\" in
Windows and that includes DOS device paths (the cases mentioned above).
Thus, this patch prevents DOS device paths from being treating as UNC paths in
IsBlockedUNCPath.
Differential Revision: https://phabricator.services.mozilla.com/D73621
This wraps the current thread such that a call to Dispatch will be done as direct task dispatch instead so that the task will be run in the current event loop.
This allows for have similar (but not identical to) the microtask semantics of JS promises
Differential Revision: https://phabricator.services.mozilla.com/D69995
If set, the callback will be dispatched via a direct task.
Direct tasks are run via the current thread's tail dispatcher.
This mechanism is only available if both the caller and the target are on the same thread and an AbstractThread is available
Differential Revision: https://phabricator.services.mozilla.com/D71592
In Python 3, iteration order over the contents of a native `dict` is always unpredictable, which results in bugs like bug 1635755 where `xptdata.cpp` has irreproducible content. To avoid this, we be sure to always write out JSON in a fixed, deterministic order.
Differential Revision: https://phabricator.services.mozilla.com/D74443
already_AddRefed destructor assert that it's nullptr.
DelayedDispatch check that the value passed isn't 0 and return an error code, leaving the already_AddRefed untouched.
Differential Revision: https://phabricator.services.mozilla.com/D73821
already_AddRefed destructor assert that it's nullptr.
DelayedDispatch check that the value passed isn't 0 and return an error code, leaving the already_AddRefed untouched.
Differential Revision: https://phabricator.services.mozilla.com/D73821
As of bug 1621451 this argument was ignored, but it just silently runs your code with `python3` if you pass it anyway. Ensure this doesn't happen any more, and protect against any other unexpected arguments as well.
Differential Revision: https://phabricator.services.mozilla.com/D73485