Bug 1791350 - Replace C++ uses of Services.startup.shuttingDown with appropriate AppShutdown::IsInOrBeyond. r=florian

Differential Revision: https://phabricator.services.mozilla.com/D157628

Differential Revision: https://phabricator.services.mozilla.com/D157617

Bug 1786571 - Add IsCancelled checks to NetworkLoadHandler::OnStreamComplete and have a GetCancelResult r=dom-worker-reviewers,asuth

Differential Revision: https://phabricator.services.mozilla.com/D157307

Consistently use `char32_t` when referring to code points instead of using
sometimes `char32_t` and in other places `uint32_t`.

Depends on D157622

Differential Revision: https://phabricator.services.mozilla.com/D157623

Bug 1791357 - Part 1: Remove unused out-params for TokenStreamSpecific::get{,FullAscii}CodePoint. r=arai

Differential Revision: https://phabricator.services.mozilla.com/D157622

Differential Revision: https://phabricator.services.mozilla.com/D157697

Differential Revision: https://phabricator.services.mozilla.com/D157686

Bug 1790451  - remove now-unnecessary QueryInterface(Ci.nsITransportSecurityInfo) calls r=jschanck,webdriver-reviewers,necko-reviewers,application-update-reviewers,nalexander,valentin

Differential Revision: https://phabricator.services.mozilla.com/D157166

Differential Revision: https://phabricator.services.mozilla.com/D157137

Bug 1790451 - make nsIChannel.securityInfo concrete as nsITransportSecurityInfo r=kershaw,necko-reviewers,valentin,geckoview-reviewers,owlish

Differential Revision: https://phabricator.services.mozilla.com/D157136

Bug 1790451 - use the concrete type nsITransportSecurityInfo for securityInfo in HttpTransaction r=kershaw,necko-reviewers

Differential Revision: https://phabricator.services.mozilla.com/D157135

The <mfenced> element [1] is not part of MathML Core [2] or MathML AAM
[3] and was removed in bug 1603773. It should be treated as
an unknown MathML element, that is like an `<mrow>`.

This commit performs some follow-up code cleanup:

* Remove a11y mapping, instead map it like an mrow.

* Remove obsolete deprecation message.

* Don't include mfenced in the sanitizer's default element allow list.
  The spec [4] does not mention an explicit list.

* Preference mathml.mfenced_element.disabled was removed, so no need
  to force it when running tests.

* Tweak some code comments.

[1] https://developer.mozilla.org/en-US/docs/Web/MathML/Element/mfenced
[2] https://w3c.github.io/mathml-core/#mathml-elements-and-attributes
[3] https://w3c.github.io/mathml-aam/
[4] https://wicg.github.io/sanitizer-api

Differential Revision: https://phabricator.services.mozilla.com/D157580

Make it a bit easier to read and less prone to race conditions. Remove a
setTimeout referencing bug 103197 which I'm pretty sure it's not an
issue.

Differential Revision: https://phabricator.services.mozilla.com/D156543

Differential Revision: https://phabricator.services.mozilla.com/D157545

Backed out changeset 149d45b6a85b (bug 1781434) for causing mochitest failures on test_bug398289.html CLOSED TREE

Bug 1791458 - Remove auto main-size workaround for replaced elements when computing flex base size. r=emilio

After bug 1340715, we use the same path [1] when computing `auto` and
`max-content` inline-sizes for replaced elements in
`nsContainerFrame::ComputeSizeWithIntrinsicDimensions()`. For block-sizes,
`max-content` and `auto` already have the same behavior.

This patch shouldn't change the behavior.

[1] https://searchfox.org/mozilla-central/rev/6d7704caae2a61b0101bdb67481a80af20376663/layout/generic/nsContainerFrame.cpp#2354-2355

Differential Revision: https://phabricator.services.mozilla.com/D157678

Differential Revision: https://phabricator.services.mozilla.com/D157402

Bug 1791471: Use RefPtr to manage lifetime of nsSplitterFrame::mInner, instead of raw pointer with manual AddRef/Release. r=hiro

This patch doesn't change behavior; it just formalizes the refcounting that
we're already doing for this variable by using the RefPtr type.

Differential Revision: https://phabricator.services.mozilla.com/D157683

Bug 1761370 - Fix bad formatting in FileSystemManagerParent::RecvGetAccessHandle and FileSystemManagerParent::RecvGetFile; r=dom-storage-reviewers,jesup

Differential Revision: https://phabricator.services.mozilla.com/D157397

Differential Revision: https://phabricator.services.mozilla.com/D157612

Make it a bit easier to read and less prone to race conditions. Remove a
setTimeout referencing bug 103197 which I'm pretty sure it's not an
issue.

Differential Revision: https://phabricator.services.mozilla.com/D156543

Differential Revision: https://phabricator.services.mozilla.com/D157662

The main issue here is that we have a dynamic change request from the
embed load. That would usually not get honored, because we're printing.
But we call into ReconstructFrames() and flush style _before_ asking to
reconstruct the document, which means that replicated fixed frames that
don't get properly cleaned-up.

This is not a correctness issue because we're about to reconstruct the
whole frame tree anyways, but the intermediate state is invalid and
caught by our assertions.

Differential Revision: https://phabricator.services.mozilla.com/D157674

Bug 1791454 - On the toolchains repo, run the same set of tasks as mozilla-central. r=taskgraph-reviewers,jmaher

Differential Revision: https://phabricator.services.mozilla.com/D157673

We previously annotated this test as needing a long timeout, since it can reach
the 10-second test-timeout threshold when run with "aggressive" chaos mode,
i.e. with MOZ_CHAOSMODE=0xfb.  However, when run normally, the test takes
~100-200ms, so it's silly that we're considering it as needing a "long
timeout".

Really, we should address the timeout by changing the thresholds in our
test-verify/chaos-mode test harness, rather than by doing one-off "long
timeout" annotations for every test that's slow in chaos mode (which turns out
to be quite a lot of tests).

Differential Revision: https://phabricator.services.mozilla.com/D157668

Differential Revision: https://phabricator.services.mozilla.com/D157667

Differential Revision: https://phabricator.services.mozilla.com/D157264

Differential Revision: https://phabricator.services.mozilla.com/D157263

Bug 1786910: Ignore 'align-content' when determining the aligned static position of abspos flex children. r=TYLin

Before this patch, we honored `align-content` (in combination with
`align-self`) for cross-axis alignment for abspos flex children **in cases
where the flex container was multi-line**. This was a bit weird, but was
required by the spec, and made some sense in the spirit of aligning the abspos
box as if it were the sole flex item in a flex container.

Now the CSSWG has resolved in [1] to simplify things by just ignoring
`align-content` for abspos flex children. So, this patch updates us in
accordance with this change. Such items now only have to look at `align-self`
for cross-axis alignment to determine their static position in their flex
container.

Before this commit, we had a bunch of WPT tests to check the impact of every
align-content value, with all of the various flavors of directionality.  Now
that align-content has *no effect* in any of these cases, all of these tests
become pretty trivial and redundant.  Rather than carrying them all forward
with trivial "no impact" expectations for each scenario, I've just updated the
first/simplest such test to expect no-behavioral-impact and I've removed the
rest.

[1] https://github.com/w3c/csswg-drafts/issues/7596#issuecomment-1225952646

Differential Revision: https://phabricator.services.mozilla.com/D157571

We keep clang 14 for webrender-wrench for now, because of bug 1789346.

Differential Revision: https://phabricator.services.mozilla.com/D155532

Bug 1781742 - Added nullptr check (Crash @ mozilla::dom::CanonicalBrowsingContext::SessionHistoryCommit). r=smaug

Differential Revision: https://phabricator.services.mozilla.com/D156978

Bug 1791239 - Annotate move-back-iframe-parse-error-external-classic.html as intermittent failing r=intermittent-reviewers,jmaher

Differential Revision: https://phabricator.services.mozilla.com/D157644

Differential Revision: https://phabricator.services.mozilla.com/D157642

Bug 1791336 - Extend command buffer infrastructure for future primitive types r=gfx-reviewers,lsalzman

Differential Revision: https://phabricator.services.mozilla.com/D157611

This inlines the megamorphic cache lookup, netting us about a 30-40%
improvement for the microbenchmark on top of the prior optimizations. There is
no measured performance losses for speedometer or matrix-react, and maybe a
1% win overall, though I need to do more runs with the cleaned up patch to be
confident.

We could do similar inlining for the CacheIRCompiler version and for
MegamorphicLoadSlotByValue - however, we wouldn't be able to include the id and
id hash as immediates, which would mean we'd need another register and several
more instructions to make it work, and it would be awkward to share the masm
code because of the different register requirements.

Differential Revision: https://phabricator.services.mozilla.com/D156615

Initially we explored caching this check on the shape, which still seems like
a decent idea. However, we can also just sidestep the problem for the most part
be moving the check after the cache lookup, since we can't get a cache hit if
it's not a NativeObject anyway, and we're hoping that a cache hit should be the
dominant path in most cases.

Differential Revision: https://phabricator.services.mozilla.com/D156614

The first three bits of the Shape pointer will always be 0, so we lose out on
some entropy here. I initially tried a "real" hashing algorithm, but noticed as
you likely already did that this can be incredibly hot and even adding a single
multiply in meaningfully degrades the performance of this. However, adding this
shift in just adds one mov and one shr, and I couldn't get it to regress perf
for a microbenchmark that tried to make it irrelevant. However, it did speed up
a microbenchmark that used the same key over and over again against 100
different shapes by a factor of about 25%, so I think it's worth it to avoid
potential pathological cases.

Further observations: we do hit this issue reasonably frequently in the matrix
react bench as well as in speedometer - this patch reduces cache misses where
only the shape was different quite substantially. However, those misses are
a small enough portion of overall cache misses that this patch doesn't seem to
observably affect those benchmarks' overall numbers.

Lastly, I wish I could have computed these shifts instead of just leaving a
comment explaining how they were derived. However, our tools for doing this in
C++ are not constexpr-friendly it seems. Also, I made these public because
they will be used from the MacroAssembler in later patches.

Differential Revision: https://phabricator.services.mozilla.com/D156613

Only show the Firefox View tab in the Ctrl Tab panel if it's currently selected
(and the same for any other pages that can be simultaneously hidden and
selected, if any are added in the future).

Differential Revision: https://phabricator.services.mozilla.com/D157595