It should have been working on fis-xorig runs since bug 1714594.
Note that the try run used for filing bug 1716409 was pushed on June 7th, whereas
bug 1714594 landed on June 11th.
Differential Revision: https://phabricator.services.mozilla.com/D118767
Cached tasks help us reuse existing tasks for toolchain, docker-image,
and fetch tasks, among others. This saves us compute cycles and
end-to-end time; it also allows us to keep these artifacts pinned
until we rebuild, which is a way to keep non-deterministic
artifacts, like docker images, in a predictable state.
However, occasionally we need to rebuild these tasks, e.g. when we
rotate the CoT key and want to invalidate the old key that we signed
the existing cached_tasks with. Until now, this process was unwieldy.
Now we have an action to allow us to do this at will.
Differential Revision: https://phabricator.services.mozilla.com/D118749
This was added ing bug 1352595. It has only been used briefly, and we
since then have gone all the way to never compress omni.ja. Even if we
did go back to compressing, we'd probably go with zstd rather than
brotli.
The gecko-side support for this has only ever been nightly-only.
Differential Revision: https://phabricator.services.mozilla.com/D118655
This was added in bug 1355661, has been kept Nightly-only since then,
and was only used briefly for omni.ja, and eventually we switched
compression off entirely on omni.ja. If we ever switch compression on
again on omni.ja, it would likely be with zstd, which is much faster
on the compression side.
Differential Revision: https://phabricator.services.mozilla.com/D118654
This helps to fix an intermittent that occurs when the iframe that loads
the picker sometimes has an opportunity to fire a "load" event _before_
datepicker.xhtml is actually loaded (perhaps from the initial about:blank
document). This made it so that waitForPickerReady would attach the
PickerReady CustomEvent listener on the wrong document, and the test
would time out despite the PickerReady event properly firing.
The fix is to ensure that the load event that waitForPickerReady waits
for is for the datepicker.xhtml document.
Differential Revision: https://phabricator.services.mozilla.com/D118746
+ Begin to add video tests to ensure we ratchet towards correctness.
+ Test rec709 x (yuv420p, yuv420p10, gbrp) x (tv, pc) x codecs.
+ Just mark fuzziness for now. Better would be e.g. 16_127_233 'bad
references'.
Differential Revision: https://phabricator.services.mozilla.com/D115298
Bug 1215723 introduced a mechanism whereby DataStorage data would be propagated
to content processes to avoid the IPC calls involved in querying if hosts were
HSTS. With a low number of content processes, this was a reasonable approach.
However, with Fission and the proliferation of content processes, propagating
changes in DataStorage data to every content process wakes up unrelated
processes and is inefficient. This patch restores the behavior that
nsISiteSecurityService and DataStorage is not available to content processes.
Additionally, bug 1626076 made it possible to use DataStorage directly from the
socket process. However, as of bug 1650356, this is no longer necessary, so
this patch removes that behavior as well.
Differential Revision: https://phabricator.services.mozilla.com/D118207
Run TV and TC tasks with python 3. Several issues are addressed for TV:
- minor Android-only str/int type confusion for mozinfo parameter;
- the type of mozlog-generated log* options changes under py3, causing
difficulty when deepcopy'ing options; avoided by removing the need
for deepcopy;
- on Windows, with py 3.6, repeated process launches trigger a python
bug; avoided by running self-tests only once per TV session.
Differential Revision: https://phabricator.services.mozilla.com/D118769
This moves the logic of whether a pres shell should be active to a
single place to make it sane to reason about, and fixes the
subdocument propagation when a BrowserChild becomes visible.
Differential Revision: https://phabricator.services.mozilla.com/D118703
Updating raster_scale in post_update can mean that it is
updated one frame too late. This can cause reftest failures
in tests that set the scale factor (without an APZ) and then
capture that frame.
Differential Revision: https://phabricator.services.mozilla.com/D118672
This also removes code that handled the case where a Send Tab or Send Link to Device
menu needed to be populated when in the STATUS_NOT_CONFIGURED state. Since those menus
should be unavailable when in the STATUS_NOT_CONFIGURED state, we throw instead.
Depends on D118622
Differential Revision: https://phabricator.services.mozilla.com/D118623
We can't rely on the Children list since it may have been cleared on shutdown.
Since we don't clear parent edges, walking the parent chain and using
mChildOffset works.
Differential Revision: https://phabricator.services.mozilla.com/D118384
Bug 1215723 introduced a mechanism whereby DataStorage data would be propagated
to content processes to avoid the IPC calls involved in querying if hosts were
HSTS. With a low number of content processes, this was a reasonable approach.
However, with Fission and the proliferation of content processes, propagating
changes in DataStorage data to every content process wakes up unrelated
processes and is inefficient. This patch restores the behavior that
nsISiteSecurityService and DataStorage is not available to content processes.
Additionally, bug 1626076 made it possible to use DataStorage directly from the
socket process. However, as of bug 1650356, this is no longer necessary, so
this patch removes that behavior as well.
Differential Revision: https://phabricator.services.mozilla.com/D118207
`nsAvailableMemoryWatcher::mTimer` was nullptr when `LowMemoryCallback()` tried to
initialize a timer via `OnLowMemory()`. There are two possible reasons.
The first case is that NS_NewTimer() returned nullptr because the available memory was
already low when initializing `nsAvailableMemoryWatcher`. In this case, we should not
register the callback.
The second case is the low-memory callback was queued while shutting down
`nsAvailableMemoryWatcher` (just before we unregiter the callback). We do refcount
the object and use the mutex correctly, but we touch the nulled out member. We should
make an early return if the object was shut down.
Differential Revision: https://phabricator.services.mozilla.com/D118745
Tooru Fujisawa
Brad Werth
Hiroyuki Ikezoe
Aki Sasaki
Mike Hommey
Mike Conley
Sandor Molnar
Sam Foster
Jeff Gilbert
Dana Keeler
Geoff Brown
Kirk Steuber
Emilio Cobos Álvarez
Nihanth Subramanya
Iulian Moraru
Kagami Sascha Rosylight
Glenn Watson
Boris Chiou
Kashav Madan
dev
Toshihito Kikuchi
amy churchwell
Olivier Tilloy
Mitchell Hentges
Andi-Bogdan Postelnicu
Jon Coppeard
Ryan VanderMeulen
Gijs Kruitbosch
Andrew Osmond
Steve Fink