If the graph leaks because it does not shut down its driver, then an nsThread
can be leaked and block shutdown at nsThreadManager::ShutdownNonMainThreads().
The shutdown blocker has better debug messages to identify the source of the
blockage than ShutdownNonMainThreads(), so I'd like to remove
MediaTrackGraph's code that removes it's shutdown blocker after a timeout.
This new assert is intended to give some visibility into how many problems
such removal might cause.
Depends on D198769
Differential Revision: https://phabricator.services.mozilla.com/D198770
When the scrollbar is on the right hand side, its size should be set in `right`
instead of in `left`.
This bug is not observable because all the callers are accessing the result
directly via `LeftRight()`, or they are transforming the result into a
`LogicalMargin` and access it via `IStartEnd()` or `BStartEnd()`. However, it is
still worth fixing for the future.
Differential Revision: https://phabricator.services.mozilla.com/D198907
This patch doesn't change the behavior (or pass/fail status) of the test at
all; it's just fixing some issues in the test that trigger parser errors and/or
make the test results hard to read.
(This is very similar to the patch that I landed earlier in
https://bugzilla.mozilla.org/show_bug.cgi?id=1874986 ; this one's just for
a different test that had similar issues.)
Specific fixes here:
- Removing the 'head' element since it's unnecessary and was being opened later
than appropriate (after head content had already begun).
- Wrapping the test content in an element that we can use to scope the rules,
so that we don't end up with a generic 'div' css selector that inadvertently
targets the test-results divs that the harness appends to the page.
- Remove a stray backtick at the end of the file, and add a final newline.
- Indentation changes to remove unnecessary indentation and to accommodate the
newly-added "test_content" wrapper div.
Differential Revision: https://phabricator.services.mozilla.com/D198906
This allows us to create a TextureClient on a different thread than the
actor without special effort on the part of the allocator. Similarly, it
also allows us to destroy a TextureClient on a different thread if it
has a readlock bound to it.
Differential Revision: https://phabricator.services.mozilla.com/D195123
This patch adds support for allocationg shmem sections for
ImageBridgeChild. The recording infrastructure depends on it.
Differential Revision: https://phabricator.services.mozilla.com/D189528
Startup of extension's background script is currently delayed until
after "browser-delayed-startup-finished". On desktop, this notification
is sent as soon as any browser window has been painted, which is usually
soon enough.
On Android, this notification is sent from geckoview.xhtml, which is
only loaded when any Gecko-hosted (web) content is shown. Network
requests can already be triggered from elsewhere (GeckoView) without
having loaded any Gecko-hosted "browser window" (geckoview.xhtml). This
is a problem when a `proxy.onRequest` handler has been registered,
because this can result in a request blocked indefinitely, as long as
the user has not opened an unrelated browser window.
To get around this problem, allow proxy request listeners to immediately
wake up the background page, on Android.
Differential Revision: https://phabricator.services.mozilla.com/D198842
If the platform code to discover third party certificates ever imports a
certificate that is already a built-in root, bad things can happen, so this
patch prevents that.
Bad thing 1: if the "third party" certificate is used as an intermediate when
it's really a built-in trust anchor, certificates issued from that root will
not verify correctly because the root's trust will be "inherits trust" and
there will usually be no valid issuer for that root (see e.g. bug 1873724).
Bad thing 2: if the "third party" certificate is used as a root and the user
has edited the default settings to not trust that root, it will unexpectedly be
trusted again.
Ideally, we'd check all loaded modules (in particular, the softoken) so as to
not override their settings as well, but that's slow and we don't want to do
that at startup.
Differential Revision: https://phabricator.services.mozilla.com/D198214
It doesn't hold if the load event is delayed a bit, but it breaking
shouldn't change any behavior / shouldn't matter in practice.
Differential Revision: https://phabricator.services.mozilla.com/D198785
We have previously updated the AboutRedirector from `about:firefoxview-next` to `about:firefoxview`.
`firefoxview-next` no longer exists, so has been removed from the static components manager.
Differential Revision: https://phabricator.services.mozilla.com/D198815