This is needed because early in a content process's lifecycle, NeckoParent may
not have been created yet. This leads to issues when trying to redirect into a
fresh process which hasn't performed network loads yet. By sending the message
over PContent, we can be sure the APIs are available.
Differential Revision: https://phabricator.services.mozilla.com/D15608
--HG--
extra : moz-landing-system : lando
In the past we relied upon ViEEncoder::OnFrame to set the render time for
frames. With the branch 64 update, this code moved to
VideoStreamEncoder::OnFrame, and only sets the timestamp if it is greater than
the current time. This results in broken rtp timestamp estimates in the rtcp
sender report, which causes problems for Meet and possibly other services
that rewrite rtp timestamps based upon the sender report.
This patch makes VideoStreamEncoder::OnFrame always set the timestamp. In a
follow on bug, we'll move this behaviour to VideoConduit so we don't have to
maintain a local modification of the upstream code.
Differential Revision: https://phabricator.services.mozilla.com/D17413
--HG--
extra : moz-landing-system : lando
This also removes code duplication by making CallerSetterImpl call CallerGetterImpl.
Differential Revision: https://phabricator.services.mozilla.com/D17485
--HG--
extra : moz-landing-system : lando
Our normal Windows nightlies get LTO enabled implicitly by the PGO logic, but since we don't have arm64 PGO builds, let's enable LTO explicitly there.
Differential Revision: https://phabricator.services.mozilla.com/D17418
--HG--
extra : moz-landing-system : lando
r=ochameau,nchevobbe
This patch moves stopListeners to the actor, and removes the need for the callback in the
detach call on the client.
Differential Revision: https://phabricator.services.mozilla.com/D16178
--HG--
extra : moz-landing-system : lando
Several of our tests relied on the timing of the shutdown of the webconsole. I updated one
test that fired off an extra rdp request so that it is waiting until it resolves before shutting down.
For the rest, I started tracking the async behavior of initialization of the web-console-proxy, and
waiting for initialization to finish before destroying.
Differential Revision: https://phabricator.services.mozilla.com/D16177
--HG--
extra : moz-landing-system : lando
Also remove an assertion that doesn't really hold at all, and doesn't seem
relevant to the function anyway.
Differential Revision: https://phabricator.services.mozilla.com/D17359
--HG--
extra : moz-landing-system : lando
A lot of files include nsCOMPtr without really needing it in dom/svg. (And a
few files use the nsCOMPtr type without including its header.)
For this patch, I found candidate files by just searching for files that only
had a single line that contained "nsCOMPtr", and I took the following actions:
- If the match is an #include, then remove it (replacing it with the include
for RefPtr or AlreadyAddrefed if the file uses those types instead).
- If the match is a usage of the type, then add an #include for nsCOMPtr (as
well as one for RefPtr, if the file uses that type).
(And in a few cases, I moved an adjacent #include, to get the #include list
closer to sortedness.)
I verified that the dom/svg directory continues to build successfully in
non-unified mode (i.e. if I remove UNIFIED_ from its moz.build file). So, I'm
not removing any includes that files are now inadvertantly leaning on other
compilation units to provide.
Differential Revision: https://phabricator.services.mozilla.com/D17422
--HG--
extra : moz-landing-system : lando
This optimization was added in bug 864218 to optimize asm.js-like code in Ion.
We're removing it now because:
* WebAssembly happened and we've been focusing more on page load than on hot
loop performance.
* It complicates the interpreter work because JSOP_SETALIASEDVAR can't do the
fast thing on singleton call objects and needs to fall back to an IC.
* JSOP_SETALIASEDVAR and JSOP_INITALIASEDLEXICAL are currently marked as JOF_IC
so they always get a SetProp IC allocated for them. With this patch that no
longer happens.
* The patch removes a lot of complexity.
* It doesn't seem to affect Speedometer, Octane, other benchmarks.
Note that this also removes the EnvironmentCoordinateNameCache because
EnvironmentCoordinateName is now only called in debug-only or error-handling
code. I renamed it to EnvironmentCoordinateNameSlow to emphasize this.
Differential Revision: https://phabricator.services.mozilla.com/D17339
--HG--
extra : moz-landing-system : lando
This cannot actually happen in the real world because this path is specific to
when the compositor process is also the parent process, and thus is not
actually IPC. However, the fuzzer can trigger this case.
Depends on D14587
Differential Revision: https://phabricator.services.mozilla.com/D14588
--HG--
extra : moz-landing-system : lando
