b96ee3dad4d8eb59ba72e2b5735461ec7cfb2540: Bug 1298202 - Baldr: take advantage of guard page to simplify asm.js/wasm memory access (r=sunfish)
Luke Wagner <luke@mozilla.com> - Thu, 08 Sep 2016 00:49:46 -0500 - rev 411653
Push 28941 by bmo:cku@mozilla.com at Thu, 08 Sep 2016 10:12:28 +0000
Bug 1298202 - Baldr: take advantage of guard page to simplify asm.js/wasm memory access (r=sunfish) MozReview-Commit-ID: 1jlCJCGUf29
7a59052a5a5a313e1d8f66c6731c59385ffb2327: Bug 1298202 - Baldr: allocate a guard page on wasm 32-bit (r=sunfish)
Luke Wagner <luke@mozilla.com> - Thu, 08 Sep 2016 00:09:07 -0500 - rev 411652
Push 28941 by bmo:cku@mozilla.com at Thu, 08 Sep 2016 10:12:28 +0000
Bug 1298202 - Baldr: allocate a guard page on wasm 32-bit (r=sunfish) MozReview-Commit-ID: IpYhSxkB4Lb
40854f9eaeb5e47d55aae70b35600f7230d78695: Bug 1180684 - Part 2. Hook GetKeyState on plugin process. r=aklotz
Makoto Kato <m_kato@ga2.so-net.ne.jp> - Mon, 22 Aug 2016 16:06:50 +0900 - rev 411651
Push 28941 by bmo:cku@mozilla.com at Thu, 08 Sep 2016 10:12:28 +0000
Bug 1180684 - Part 2. Hook GetKeyState on plugin process. r=aklotz MozReview-Commit-ID: 3cCWMYYaTkn
f32b355726da23e8910de45b66382f8a2999e320: Bug 1180684 - Part 1. Analyze MOV with GS. r=aklotz
Makoto Kato <m_kato@ga2.so-net.ne.jp> - Wed, 07 Sep 2016 16:15:58 +0900 - rev 411650
Push 28941 by bmo:cku@mozilla.com at Thu, 08 Sep 2016 10:12:28 +0000
Bug 1180684 - Part 1. Analyze MOV with GS. r=aklotz Allow MOV with GS prefix on x64. Windows 10 uses this opcode on GetKeyState(). MozReview-Commit-ID: CqDSxmWdkiH
c8ddbd169db3c4d1b24e1597d40b00896c31f62b: Bug 1298101. Implement telemetry to take event handling time during page load. r=smaug
Timothy Nikkel <tnikkel@gmail.com> - Wed, 07 Sep 2016 23:30:24 -0500 - rev 411649
Push 28941 by bmo:cku@mozilla.com at Thu, 08 Sep 2016 10:12:28 +0000
Bug 1298101. Implement telemetry to take event handling time during page load. r=smaug
df7cb65bee395b7a28793a759d4fa1dca9acc4a6: Bug 1301138. Set the one allowed sandbox navigator even when popups are allowed to escape the sandbox, so the new popup can be navigated by the sandboxed document. r=smaug
Boris Zbarsky <bzbarsky@mit.edu> - Thu, 08 Sep 2016 00:17:11 -0400 - rev 411648
Push 28941 by bmo:cku@mozilla.com at Thu, 08 Sep 2016 10:12:28 +0000
Bug 1301138. Set the one allowed sandbox navigator even when popups are allowed to escape the sandbox, so the new popup can be navigated by the sandboxed document. r=smaug
23530934541a0d8c23232273f7900af27ef7c09c: Bug 1299908 - Make sure we set the correct AGR to reference frame offset when flattening layers. r=mstange
Matt Woodrow <mwoodrow@mozilla.com> - Thu, 08 Sep 2016 15:39:04 +1200 - rev 411647
Push 28941 by bmo:cku@mozilla.com at Thu, 08 Sep 2016 10:12:28 +0000
Bug 1299908 - Make sure we set the correct AGR to reference frame offset when flattening layers. r=mstange
331524df5cab6383f44d55d777b7cededdbc12c9: merge fx-team to mozilla-central a=merge
Carsten "Tomcat" Book <cbook@mozilla.com> - Thu, 08 Sep 2016 12:00:23 +0200 - rev 411646
Push 28941 by bmo:cku@mozilla.com at Thu, 08 Sep 2016 10:12:28 +0000
merge fx-team to mozilla-central a=merge
bd28be90aed86db495c8f0fdb6fc86c47c4484ea: merge autoland to mozilla-central a=merge
Carsten "Tomcat" Book <cbook@mozilla.com> - Thu, 08 Sep 2016 11:59:31 +0200 - rev 411645
Push 28941 by bmo:cku@mozilla.com at Thu, 08 Sep 2016 10:12:28 +0000
merge autoland to mozilla-central a=merge
afb8c6617510e7a8a0132e8412ca1aa8c049933a: Bug 1300497. Part 6 - remove the call to |SetState(DECODER_STATE_SHUTDOWN)| from DecodeError(). r=kaku
JW Wang <jwwang@mozilla.com> - Tue, 06 Sep 2016 15:33:20 +0800 - rev 411644
Push 28941 by bmo:cku@mozilla.com at Thu, 08 Sep 2016 10:12:28 +0000
Bug 1300497. Part 6 - remove the call to |SetState(DECODER_STATE_SHUTDOWN)| from DecodeError(). r=kaku So we have only one place to transition to the SHUTDOWN state which is Shutdown(). MozReview-Commit-ID: 6MNISCea94Q
da34e47f89b50a8dfc9827139cab693e1de62b97: Bug 1300497. Part 5 - run MDSM cycles immediately in the callback of mDelayedScheduler. r=kaku
JW Wang <jwwang@mozilla.com> - Mon, 05 Sep 2016 17:51:54 +0800 - rev 411643
Push 28941 by bmo:cku@mozilla.com at Thu, 08 Sep 2016 10:12:28 +0000
Bug 1300497. Part 5 - run MDSM cycles immediately in the callback of mDelayedScheduler. r=kaku There is no point in scheduling an addition cycle to do that. Also remove the annoying debugging message which is not helpful. MozReview-Commit-ID: BMjeTNg6HCY
78525890f801cb4da2ce1a87befaf9c86862232a: Bug 1300497. Part 4 - remove the call to ScheduleStateMachine() before |SetState(DECODER_STATE_SHUTDOWN)| and cancel mDelayedScheduler in Shutdown(). r=kaku
JW Wang <jwwang@mozilla.com> - Tue, 06 Sep 2016 10:26:45 +0800 - rev 411642
Push 28941 by bmo:cku@mozilla.com at Thu, 08 Sep 2016 10:12:28 +0000
Bug 1300497. Part 4 - remove the call to ScheduleStateMachine() before |SetState(DECODER_STATE_SHUTDOWN)| and cancel mDelayedScheduler in Shutdown(). r=kaku We don't want to run MDSM cycles anymore once shutdown begins. MozReview-Commit-ID: 7Py20oqWNBG
6fcdb8ab059d1e823be888c2bebde1d157f21475: Bug 1300497. Part 3 - move some code from FinishShutdown() to Shutdown(). r=kaku
JW Wang <jwwang@mozilla.com> - Tue, 06 Sep 2016 10:23:02 +0800 - rev 411641
Push 28941 by bmo:cku@mozilla.com at Thu, 08 Sep 2016 10:12:28 +0000
Bug 1300497. Part 3 - move some code from FinishShutdown() to Shutdown(). r=kaku We don't want to receive notifications from MediaEventSource or WatchManager once shutdown begins. MozReview-Commit-ID: LiKUjOGuxyX
85c7215cdeada9f23cd068d110c1bb5e705e28c3: Bug 1300497. Part 2 - constify and fix the comment of IsShutdown(). r=kaku
JW Wang <jwwang@mozilla.com> - Mon, 05 Sep 2016 16:50:06 +0800 - rev 411640
Push 28941 by bmo:cku@mozilla.com at Thu, 08 Sep 2016 10:12:28 +0000
Bug 1300497. Part 2 - constify and fix the comment of IsShutdown(). r=kaku MozReview-Commit-ID: AsZwNPAlq0d
8fcadb9544bb005b69a2c50d45fd85500d138332: Bug 1300497. Part 1 - assert we never escape the SHUTDOWN state. r=kaku
JW Wang <jwwang@mozilla.com> - Mon, 05 Sep 2016 16:38:09 +0800 - rev 411639
Push 28941 by bmo:cku@mozilla.com at Thu, 08 Sep 2016 10:12:28 +0000
Bug 1300497. Part 1 - assert we never escape the SHUTDOWN state. r=kaku MozReview-Commit-ID: JDWdFAGZijm
817e76d33b76a53521f0263819a45376092b4718: Bug 1264110 - fix timing issue in test cases. r=kershaw.
Shih-Chiang Chien <schien@mozilla.com> - Mon, 05 Sep 2016 18:18:11 +0800 - rev 411638
Push 28941 by bmo:cku@mozilla.com at Thu, 08 Sep 2016 10:12:28 +0000
Bug 1264110 - fix timing issue in test cases. r=kershaw. MozReview-Commit-ID: 2Ia4L7EizrA
92f041a22d0a2f11d95453b0934fee8d3e9b6529: Bug 1301008 - Pass safebrowsing v4 list state in base64 format to avoid truncation. r=francois
Henry Chang <hchang@mozilla.com> - Wed, 07 Sep 2016 17:45:15 +0800 - rev 411637
Push 28941 by bmo:cku@mozilla.com at Thu, 08 Sep 2016 10:12:28 +0000
Bug 1301008 - Pass safebrowsing v4 list state in base64 format to avoid truncation. r=francois MozReview-Commit-ID: 6oVdQvEoMm2
3e3c5cd9d31eaa0fed88b46297ace3590613589c: Bug 1301068: Ensure InputExhausted is called if H264Converter drops a frame. r=gerald
Jean-Yves Avenard <jyavenard@mozilla.com> - Thu, 08 Sep 2016 00:22:06 +1000 - rev 411636
Push 28941 by bmo:cku@mozilla.com at Thu, 08 Sep 2016 10:12:28 +0000
Bug 1301068: Ensure InputExhausted is called if H264Converter drops a frame. r=gerald MozReview-Commit-ID: LyZlzU4D58o
ba7d7fe0ea3cafb1b27e738dfa3697eb3f49dec7: Bug 1300296: Use valid and spec-compliant MP4 file for all tests. r=gerald,jgraham
Jean-Yves Avenard <jyavenard@mozilla.com> - Tue, 06 Sep 2016 14:55:25 +1000 - rev 411635
Push 28941 by bmo:cku@mozilla.com at Thu, 08 Sep 2016 10:12:28 +0000
Bug 1300296: Use valid and spec-compliant MP4 file for all tests. r=gerald,jgraham The MP4 used in the tests was invalid: 1- It contains out of order and overlapping frames 2- The first frame for each of the first 9 media segments were marked as keyframe, even though they weren't The new file was encoded such that: - The total bitrate remains the same as in the previous files - The total duration is similar - Each media segment starts with a valid keyframe - Frames are properly ordered - GOP is made of 24 frames - The video track doesn't start at 0, which helps testing the proper behaviour regarding buffered range calculation. - Like in the previous video, the audio track is present but made of silence (which is puzzling for a video initially designed to test A/V sync) The webm file presents the same characteristics. MozReview-Commit-ID: 6AKKmInaenE
d7e38eda81dba94dd26713db46de15e924d5d3eb: Bug 1300296: P3. Ensure that a new or flushed decoder first h264 frame is always a keyframe. r=gerald
Jean-Yves Avenard <jyavenard@mozilla.com> - Tue, 06 Sep 2016 11:47:20 +1000 - rev 411634
Push 28941 by bmo:cku@mozilla.com at Thu, 08 Sep 2016 10:12:28 +0000
Bug 1300296: P3. Ensure that a new or flushed decoder first h264 frame is always a keyframe. r=gerald MozReview-Commit-ID: s0NEB8hesj
(0) -300000 -100000 -30000 -10000 -3000 -1000 -300 -100 -50 -20 +20 +50 +100 +300 +1000 +3000 +10000 +30000 +100000 +300000 tip