Instead of creating and deleteing each thread, we use a nsThreadPool with a max of 8 resolver threads.
Whereas before each thread would run ThreadFunc exactly once then shut down, the threads may now remain active for a while. During this time we may post another task(runnable) to the thread.
MozReview-Commit-ID: FiE370ic1ah
--HG--
extra : rebase_source : 065bbf1f9867dfb38ac7b13816d4d06824d3a05d
This call was added in bug 929362, but the key factor to fix the bug was just
setting a flag representing that the target frame doesn't allow the animation
running on the compositor and checking the flag in the function whether the
animation can be run on the compositor or not in later ticks. So the call
wasn't necessary in the first place.
The test case here fails without this fix. The test case actually doesn't
observe animation restyle count at all, so it might look a bit awkward in
file_restyles.html, if we add other test cases checking SchedulePaint calls
in future, we will move the tests in a different file.
(The reason there is no animation restyles in this case is that we properly
throttle the animation in this case.)
MozReview-Commit-ID: AyHciRJHM0s
--HG--
extra : rebase_source : f3963336ea9165b0a9c1a662bdac5c645b209219
These cause us to load Console.jsm and create a ConsoleAPI instance in every
content process, which is expensive both in terms of memory and startup
performance.
Checking the log level before we touch the console object is much cheaper,
in terms of both memory and performance.
MozReview-Commit-ID: 19f0ggAda2J
--HG--
extra : rebase_source : 38077a14122d201d11079330565d82f6f963b428
Take AutoSaveLiveRegisters as an argument, and use that to ensure the registers were saved
--HG--
extra : rebase_source : bb54592bd009afc67aaa0ca9c705ab28f4a45d44
Automatic update from web-platform-testsClean up budget-api idl test (#11445)
Follow up for #9763 now that #10338 is resolved and we have #10381 too.
--
wpt-commits: 7b2cd8b78d938ea8879adc6419f2ea14b6af8bad
wpt-pr: 11445
Automatic update from web-platform-testsUpdate the IDL generated content headers to the content from reffy-reports (#11333)
--
wpt-commits: cdf41ff70d042eb78f461aeb44a5ad383964dbd3
wpt-pr: 11333
Automatic update from web-platform-tests[LayoutNG] Orthogonal flows need available inline size for min/max calculation.
Provide a suitable constraint space when calculating min/max inline
sizes for an orthogonal flow root. ComputeMinMaxSize() in NGBlockNode
typically ended up creating its zero-sized constraint space, which would
typically result in large block sizes (since there'd e.g. only be space
for one word per line).
Set percentage resolution size too, instead of leaving it at 0x0.
Percentages are often unresolvable (i.e. indefinite), in which case we
should of course refrain from resolving them, rather than resolving the
percentage against zero (which we used to do).
Add a DCHECK that we're always provided with a constraint space when
calculating min/max for orthogonal flows, as using the zero-size one
will not produce the correct result.
Bug: 848225
Cq-Include-Trybots: master.tryserver.chromium.linux:linux_layout_tests_layout_ng
Change-Id: Ic0aac380b2b733d3a55b25396c89584e4468899d
Reviewed-on: https://chromium-review.googlesource.com/1090845
Commit-Queue: Morten Stenshorne <mstensho@chromium.org>
Reviewed-by: Aleks Totic <atotic@chromium.org>
Reviewed-by: Koji Ishii <kojii@chromium.org>
Reviewed-by: Christian Biesinger <cbiesinger@chromium.org>
Cr-Commit-Position: refs/heads/master@{#565446}
--
wpt-commits: ce22baff58c81ab98f3cbd2ac390e51e755fd9f9
wpt-pr: 11417
Automatic update from web-platform-testsImprove RTCPeerConnection-setRemoteDescription-tracks.https.html tests.
Using async functions and promise_tests instead of then()-chains and
async_tests. This makes them more readable, easier to maintain and
shorter.
Replaced RTCPeerConnection-helper.js function performOffer() with
exchangeOffer()/exchangeAnswer()/exchangeOfferAnswer().
Updated tests that were performing multiple offers to perform a full
offer/answer exchange whenever there were future offers to be made.
Previously it would perform half of an offer/answer cycle before
starting the next offer/answer, which under Unified Plan could lead to
some unexpected failures.
TBR=guidou@chromium.org
Bug: 777617
Change-Id: Ib0305d55af5aef7f49b5418d4ba44ded9c84d5c6
Reviewed-on: https://chromium-review.googlesource.com/1082434
Reviewed-by: Henrik Boström <hbos@chromium.org>
Commit-Queue: Henrik Boström <hbos@chromium.org>
Cr-Commit-Position: refs/heads/master@{#565234}
--
wpt-commits: fb908f941185aff41514383293f029e21937a87d
wpt-pr: 11301
Automatic update from web-platform-testsAdd selector tests
* Test invalid pseudo-class starting with a valid one
* Test selector in :not() surrounded by spaces
* Test :nth-child used twice in selector
--
wpt-commits: 2436290affb197955acf8aa0115897ce5205dcc4
wpt-pr: 11449
Automatic update from web-platform-testswebdriver: update the webdriver WebIDL and fix test (#9851)
--
wpt-commits: 418538eb1cce3ecd931285db697736aa2066d05c
wpt-pr: 9851
Automatic update from web-platform-tests[idlharness.js] Clean up media-capabilities idl test
Follow up for #11255 which was unable to use add_dependency_idls because of #10338.
--
wpt-commits: c243b3c545f592a1635c770575dc6e20593ee392
wpt-pr: 11442
Automatic update from web-platform-tests[KeyboardMap] Blink module now handles iFrames and multiple calls
This CL adds the following:
- KeyboardMap now requires the caller to be a top-level, secure
browsing context
- Multiple calls to getKeyboardLayout() will now be given the same
promise instead of it being timing dependent (the initial promise
might get overwritten with the old logic)
- Sequential calls to getKeyboardLayout() still get their own promise
- All promise rejections will return a DOM exception (similar to
KeyboardLock)
- I've cleaned up KeyboardLock a bit as well so it has similar
logic/formatting to what I added to KeyboardMap
- Added several tests to KeyboardMap and fixed some failing cases
BUG=832811
Change-Id: I4a03f68c239cff42b8840e9513257384038f04b4
Reviewed-on: https://chromium-review.googlesource.com/1093474
Commit-Queue: Joe Downing <joedow@chromium.org>
Reviewed-by: Gary Kacmarcik <garykac@chromium.org>
Cr-Commit-Position: refs/heads/master@{#565833}
--
wpt-commits: 4565dc5875f3841b0884bdfd8ad6b9ad6f4a5944
wpt-pr: 11440
Automatic update from web-platform-testsCreated a test file to directly test the InputFile class in test_request.py, bypassing the server. Found and fixed a bug where passing a file larger than the buffer in InputFile caused the server to crash. Found and fixed a bug where seeking to a position before the current position and calling read() immediately after falsely tried to assert the positions of _buf_pos and _file_pos. Added tests for these bugs. (#11415)
--
wpt-commits: 99eb0f68f85f4d3c44d5c0d80021ed305e2f3613
wpt-pr: 11415
Automatic update from web-platform-testsservice worker: Add long timeout to registration-updateviacache.https.html.
The test is slow because it does a lot of registrations and updates.
Bug: 850842
Change-Id: I5782e4e426e571a604929da7369c8474ffda88e5
TBR: robertma
Reviewed-on: https://chromium-review.googlesource.com/1092958
Reviewed-by: Matt Falkenhagen <falken@chromium.org>
Reviewed-by: Robert Ma <robertma@chromium.org>
Commit-Queue: Matt Falkenhagen <falken@chromium.org>
Cr-Commit-Position: refs/heads/master@{#565641}
--
wpt-commits: 7bc2ca900c7f5d19040c5a1169d1257cb947459e
wpt-pr: 11434