searching for reviewer(gerald)
e52cb4ceec89c50c113a2787d018ea2783faf70c: Bug 1492365 - Prevent arithmetic overflow. r=gerald
Jean-Yves Avenard <jyavenard@mozilla.com> - Wed, 19 Sep 2018 08:34:15 +0000 - rev 485561
Push 241 by fmarier@mozilla.com at Mon, 24 Sep 2018 21:48:02 +0000
Bug 1492365 - Prevent arithmetic overflow. r=gerald Should mStreamLength be > 2^32, we could have overflowed leading to false positive test. Differential Revision: https://phabricator.services.mozilla.com/D6235
e0bd55a3e557e3312b44f82e2c279e42d4ac8df5: Bug 1491132 - Clamp cache reads to content size. r=gerald
Jean-Yves Avenard <jyavenard@mozilla.com> - Mon, 17 Sep 2018 17:41:31 +0000 - rev 485222
Push 241 by fmarier@mozilla.com at Mon, 24 Sep 2018 21:48:02 +0000
Bug 1491132 - Clamp cache reads to content size. r=gerald Prior bug 1416085, reads were clamped to the content's duration (if known). It appears that the new code relied on MediaCacheReadBlockFromCache to properly account for the end of content. However, this was not the case, the MediaCache always reads (and write) one full block at a time regardles of the size requested (a block is 32768 bytes). Rather than clamping in the Read() method as it used to be, we clamp in ReadBlockFromCache as such safety will benefit other callers that would have otherwise also returned garbage reads. Differential Revision: https://phabricator.services.mozilla.com/D5964
b2814d1bbbd7fc10f79c32daeaf67b4c6d527e7a: Bug 1484306 Part 4 - Use FunctionTypeTraits to yield better error message if aPredicate's signature is wrong. r=gerald
Ting-Yu Lin <tlin@mozilla.com> - Fri, 17 Aug 2018 17:02:58 -0700 - rev 480532
Push 229 by fmarier@mozilla.com at Sun, 19 Aug 2018 02:09:13 +0000
Bug 1484306 Part 4 - Use FunctionTypeTraits to yield better error message if aPredicate's signature is wrong. r=gerald Differential Revision: https://phabricator.services.mozilla.com/D3684
75ea1b10a4372754e6c6762c1512a617654c642e: Bug 1409664 - P8. Make explicit shutdown of TaskQueue optional. r=gerald
Jean-Yves Avenard <jyavenard@mozilla.com> - Thu, 07 Jun 2018 14:46:28 +0200 - rev 473068
Push 198 by fmarier@mozilla.com at Wed, 04 Jul 2018 00:35:04 +0000
Bug 1409664 - P8. Make explicit shutdown of TaskQueue optional. r=gerald Summary: With this change TaskQueue no longer requires an explicit call to BeginShutdown(). The taskqueue will automatically shutdown once it's now longer referenced. Depends on D1620 Tags: #secure-revision Bug #: 1409664 Differential Revision: https://phabricator.services.mozilla.com/D1621
779fdf8781bdee1776714a49a20e7fad44045b16: Bug 1409664 - P7. Remove unused TaskQueue method. r=gerald
Jean-Yves Avenard <jyavenard@mozilla.com> - Thu, 07 Jun 2018 14:26:11 +0200 - rev 473067
Push 198 by fmarier@mozilla.com at Wed, 04 Jul 2018 00:35:04 +0000
Bug 1409664 - P7. Remove unused TaskQueue method. r=gerald Summary: Depends on D1619 Tags: #secure-revision Bug #: 1409664 Differential Revision: https://phabricator.services.mozilla.com/D1620
3d039520dd5abfc686a47ec447624ee4f3072156: Bug 1468241 - P1. Add MozPromise::CreateAndResolveOrReject method r=gerald
Jean-Yves Avenard <jyavenard@mozilla.com> - Wed, 20 Jun 2018 22:14:05 +0000 - rev 471528
Push 198 by fmarier@mozilla.com at Wed, 04 Jul 2018 00:35:04 +0000
Bug 1468241 - P1. Add MozPromise::CreateAndResolveOrReject method r=gerald Useful when using promise chaining in combination with MozPromise::ResolveOrRejectValue parameter. MozReview-Commit-ID: F8qMh7yFnHQ Differential Revision: https://phabricator.services.mozilla.com/D1726
10631b9e6cbdf417ae61348141fedf13aa9c0a8d: Bug 1464967 - Part 2: Fix nsReflowStatus dumping when REALLY_NOISY_REFLOW is defined. r=gerald
Cameron McCormack <cam@mcc.id.au> - Tue, 29 May 2018 14:08:01 +1000 - rev 468368
Push 182 by fmarier@mozilla.com at Wed, 30 May 2018 01:07:46 +0000
Bug 1464967 - Part 2: Fix nsReflowStatus dumping when REALLY_NOISY_REFLOW is defined. r=gerald MozReview-Commit-ID: JPN7i5z5IeX
02f1c16d83afc65364a01b61a3f500b5ce390c38: Bug 1464967 - Part 1: Make nsLineBox::BreakTypeToString static. r=gerald
Cameron McCormack <cam@mcc.id.au> - Tue, 29 May 2018 14:06:50 +1000 - rev 468367
Push 182 by fmarier@mozilla.com at Wed, 30 May 2018 01:07:46 +0000
Bug 1464967 - Part 1: Make nsLineBox::BreakTypeToString static. r=gerald MozReview-Commit-ID: L97yf6oCXlq
f79ea1a6633dc7a3716b8a9f4716fc7ba3fe0e2e: Bug 1464947 - Fix ReflowInput::DisplayInitFrameTypeExit display type logging. r=gerald
Cameron McCormack <cam@mcc.id.au> - Tue, 29 May 2018 12:39:22 +1000 - rev 468360
Push 182 by fmarier@mozilla.com at Wed, 30 May 2018 01:07:46 +0000
Bug 1464947 - Fix ReflowInput::DisplayInitFrameTypeExit display type logging. r=gerald MozReview-Commit-ID: 4xPXJkys1so
4688aa455ae38eb0d359d96a925c5344376c398b: Bug 1450607 - P3. Remove unused argument. r=gerald
Jean-Yves Avenard <jyavenard@mozilla.com> - Thu, 24 May 2018 17:38:02 +0200 - rev 468015
Push 182 by fmarier@mozilla.com at Wed, 30 May 2018 01:07:46 +0000
Bug 1450607 - P3. Remove unused argument. r=gerald MozReview-Commit-ID: 6NU9BLJi6kl
2fce9c07ec2b56edc7ba969eea94287566d2c33e: Bug 1450607 - P2. Synchronously seek to prepare for resuming following stop request. r=gerald
Jean-Yves Avenard <jyavenard@mozilla.com> - Thu, 24 May 2018 14:36:58 +0200 - rev 468014
Push 182 by fmarier@mozilla.com at Wed, 30 May 2018 01:07:46 +0000
Bug 1450607 - P2. Synchronously seek to prepare for resuming following stop request. r=gerald Bug 1415090 attempted to remove the need to access from the MediaResource members of MediaCacheStream from the main thread. However, by doing so the logic flow for resuming the channel changed from a synchronous access to an asynchronous one. This changed some assumptions and allowed the ChannelMediaResource to be used before the Seek call completed. For now, re-add a cross thread access to the MediaCacheStream. A more elegant fix will be worked on in bug 1464045 MozReview-Commit-ID: 2xBTjDEqrkI
0ef665046a0627777f04209b363a25a94f73bf43: Bug 1450607 - P1. Fix constness. r=gerald
Jean-Yves Avenard <jyavenard@mozilla.com> - Wed, 23 May 2018 21:27:31 +0200 - rev 468013
Push 182 by fmarier@mozilla.com at Wed, 30 May 2018 01:07:46 +0000
Bug 1450607 - P1. Fix constness. r=gerald MozReview-Commit-ID: CeftB0qmzTf
08e9d3a3a6b13997cac9fa92231f1b17cb00c527: Bug 1433309 - Annotate createMediaKeys promise reject with whether failure occurred due to pending shutdown. r=gerald. a=lizzard
Chris Pearce <cpearce@mozilla.com> - Fri, 26 Jan 2018 12:20:35 +1300 - rev 463950
Push 177 by fmarier@mozilla.com at Thu, 24 May 2018 16:58:35 +0000
Bug 1433309 - Annotate createMediaKeys promise reject with whether failure occurred due to pending shutdown. r=gerald. a=lizzard Around every Firefox update, Netflix see a spike in MediaKeySystemAccess.createMediaKeys() promise rejects. I am wondering whether this is caused by the browser restarting to apply a Firefox update while Netflix's player is loading. So add more detail to the promise reject as to the state of the system, to try to validate that theory. MozReview-Commit-ID: 4IDPsFwKCtq
5a2fd74ee373d938c4c15b5a92f470fe813428bf: Bug 1420608. P1 - don't switch off blank decoder when seeking begins. r=gerald a=gchang
JW Wang <jwwang@mozilla.com> - Tue, 19 Dec 2017 15:43:55 +0800 - rev 463383
Push 177 by fmarier@mozilla.com at Thu, 24 May 2018 16:58:35 +0000
Bug 1420608. P1 - don't switch off blank decoder when seeking begins. r=gerald a=gchang When looping videos in a background tab, SeekingState::Enter() will switch off blank decoder and then DecodingState::Enter() will switch it on again. It is a waste of CPU cycles when the tab never goes to the foreground. The overhread is even more significant when looping short files. We should resume video decoding only when necessary that is we check in DecodingState::Enter() to see if mVideoDecodeSuspended matches mVideoDecodeMode. Bug 1420608. P2 - fix the test timeout. r=alwu See comment 50 for the cause. Since file_silentAudioTrack.html calls play() to start playback immediately, it is possible that 'mozentervideosuspend' has been fired before check_video_decoding_state() has a chance to register event handlers. We call load() and play() to start playback from the beginning so we won't miss any events. MozReview-Commit-ID: 9sKygfIxEtS
dcbdde8d683c1f2aedbdc1f484e3d1eaa9eaa3f5: Bug 1457137 - Move instead of copying strings returned from various GetDebugInfo functions for HTMLMediaElement::MozDumpDebugInfo. r=gerald
Paul Adenot <paul@paul.cx> - Thu, 26 Apr 2018 15:57:39 +0200 - rev 461032
Push 165 by fmarier@mozilla.com at Mon, 30 Apr 2018 23:50:51 +0000
Bug 1457137 - Move instead of copying strings returned from various GetDebugInfo functions for HTMLMediaElement::MozDumpDebugInfo. r=gerald MozReview-Commit-ID: LbZ0cWe7yb0
19f68300761ed2d550be280215da2103fa39a629: Bug 1445220 - Removed 'version' variable declaration as it was unused r=gerald
cubouyaka <victor.fontaine95@gmail.com> - Wed, 11 Apr 2018 16:05:40 +0200 - rev 460767
Push 165 by fmarier@mozilla.com at Mon, 30 Apr 2018 23:50:51 +0000
Bug 1445220 - Removed 'version' variable declaration as it was unused r=gerald
b49bd215b26e8831015572770ff8ec7dc91e4dad: Bug 1454052 - make MakeScopeExit a MOZ_MUST_USE function; r=gerald
Nathan Froyd <froydnj@mozilla.com> - Wed, 18 Apr 2018 12:14:18 -0400 - rev 459249
Push 163 by fmarier@mozilla.com at Fri, 20 Apr 2018 00:35:51 +0000
Bug 1454052 - make MakeScopeExit a MOZ_MUST_USE function; r=gerald Otherwise, one can do thinkos like: MakeScopeExit(...); and the scope exiting function will execute much earlier than the intended: auto guard = MakeScopeExit(...);
6a53c90b362b4eb2243e07592ef4aff23ef6166f: Bug 1453672 - Make MediaRawDataWriter's allocation methods MOZ_MUST_USE. r=gerald
Nathan Froyd <froydnj@gmail.com> - Thu, 12 Apr 2018 07:29:00 -0400 - rev 458133
Push 159 by fmarier@mozilla.com at Thu, 19 Apr 2018 00:14:50 +0000
Bug 1453672 - Make MediaRawDataWriter's allocation methods MOZ_MUST_USE. r=gerald The documentation says that the boolean return values from these methods must be checked. We might as well make the compiler check these things for us.
aa7373395767f7a58d5272067b6fd7770a9df2db: Bug 1453437 - return UniquePtr from MediaRawData::CreateWriter; r=gerald
Nathan Froyd <froydnj@mozilla.com> - Thu, 12 Apr 2018 15:28:21 -0400 - rev 457978
Push 159 by fmarier@mozilla.com at Thu, 19 Apr 2018 00:14:50 +0000
Bug 1453437 - return UniquePtr from MediaRawData::CreateWriter; r=gerald Returning UniquePtr is nicer than returning raw pointers, and has the nice side effect of forcing us to clean up the uses of nsAutoPtr that were hanging about.
a2217ad18058ac7c3167c4a39028fc4fb1e18e31: Bug 1444125 - don't poison Maybe on construction; r=gerald
Nathan Froyd <froydnj@mozilla.com> - Tue, 10 Apr 2018 13:46:00 -0400 - rev 457565
Push 155 by fmarier@mozilla.com at Thu, 12 Apr 2018 17:56:07 +0000
Bug 1444125 - don't poison Maybe on construction; r=gerald This behavior is pretty heavy-handed and degrades performance.
84c72a259b96bf116456351fbd8361cce50ef9c2: Bug 1444125 - work harder to generate good code for Maybe poisoning; r=gerald
Nathan Froyd <froydnj@mozilla.com> - Tue, 10 Apr 2018 13:46:00 -0400 - rev 457564
Push 155 by fmarier@mozilla.com at Thu, 12 Apr 2018 17:56:07 +0000
Bug 1444125 - work harder to generate good code for Maybe poisoning; r=gerald
8e0de009a63f8e4c05e16926d3dddbb1b63167e0: Bug 1446931 - P2. Refactor code flow in Benchmark. r=gerald
Jean-Yves Avenard <jyavenard@mozilla.com> - Fri, 30 Mar 2018 11:00:50 +0200 - rev 456644
Push 151 by fmarier@mozilla.com at Thu, 05 Apr 2018 22:59:31 +0000
Bug 1446931 - P2. Refactor code flow in Benchmark. r=gerald The code relied on a flag to be set to simply abort. However, that made the code workflow hard to read. We split each runs so that there's no ambiguity. MozReview-Commit-ID: LI7pL5p69zu
13b8bb0d5cc6d241af2b2fa64eabf94c4e137973: Bug 1446931 - P1. Handle errors in Benchmark. r=gerald
Jean-Yves Avenard <jyavenard@mozilla.com> - Thu, 29 Mar 2018 22:11:06 +0200 - rev 456643
Push 151 by fmarier@mozilla.com at Thu, 05 Apr 2018 22:59:31 +0000
Bug 1446931 - P1. Handle errors in Benchmark. r=gerald Benchmark had never been intended to be able to process invalid content. However over time Benchmark class has been used extensively by the fuzzing team. As such, it became necessary to handle errors of all kind. MozReview-Commit-ID: E0YulHuoiq2
aff1318be50a9ad35a9dba3198f599dccf725dc3: Bug 1447124 - Use int64_t for SaferMultDiv. r=gerald
Jean-Yves Avenard <jyavenard@mozilla.com> - Tue, 27 Mar 2018 11:27:08 +0200 - rev 455231
Push 151 by fmarier@mozilla.com at Thu, 05 Apr 2018 22:59:31 +0000
Bug 1447124 - Use int64_t for SaferMultDiv. r=gerald This prevent potential division by zero should the cast on the argument cause an overflow. We still limit the mul and div arguments to INT64_MAX. MozReview-Commit-ID: gHkv6m4zq0
e94142b7afbda68b1b4b396aa4cbfd5f89bebf8e: Bug 1448031 - make various audio/video/media constructors explicit; r=gerald
Nathan Froyd <froydnj@mozilla.com> - Tue, 27 Mar 2018 10:51:32 -0400 - rev 455128
Push 151 by fmarier@mozilla.com at Thu, 05 Apr 2018 22:59:31 +0000
Bug 1448031 - make various audio/video/media constructors explicit; r=gerald Attempting to stand up the static analysis for Android builds revealed that we weren't being explicit enough.
e56f781dbaf9761ad632aedea3ad847a746d056b: Bug 1447070: Give MediaKeySession.h the #include it needs to provide 'DecoderDoctorLifeLogger' type (to fix non-unified build error). r=gerald
Daniel Holbert <dholbert@cs.stanford.edu> - Mon, 19 Mar 2018 10:53:13 -0700 - rev 453826
Push 148 by fmarier@mozilla.com at Thu, 29 Mar 2018 23:06:47 +0000
Bug 1447070: Give MediaKeySession.h the #include it needs to provide 'DecoderDoctorLifeLogger' type (to fix non-unified build error). r=gerald MozReview-Commit-ID: An2FbvpI49I
3c4faed9b07dc64e89326db4bb83acba5e6c563b: Bug 1447091: Add #includes to fix non-unified bustage in dom/media. r=gerald
Daniel Holbert <dholbert@cs.stanford.edu> - Mon, 19 Mar 2018 11:48:38 -0700 - rev 453825
Push 148 by fmarier@mozilla.com at Thu, 29 Mar 2018 23:06:47 +0000
Bug 1447091: Add #includes to fix non-unified bustage in dom/media. r=gerald This fixes these two errors: FileMediaResource.cpp:30:36: error: invalid use of incomplete type ‘class mozilla::AbstractThread’ mCallback->AbstractMainThread()->Dispatch( MediaTimer.cpp:50:3: error: ‘Unused’ was not declared in this scope Unused << rv; MozReview-Commit-ID: WkPZc22dMF
68737a515062b449392822e9cd06292a0acf63d3: Bug 1445533 - Use for(:) in MozPromise code. r=gerald
Jan-Ivar Bruaroey <jib@mozilla.com> - Tue, 13 Mar 2018 19:13:18 -0400 - rev 453192
Push 148 by fmarier@mozilla.com at Thu, 29 Mar 2018 23:06:47 +0000
Bug 1445533 - Use for(:) in MozPromise code. r=gerald MozReview-Commit-ID: BsKR2o25hIQ
ad5ac59e30d1b27029e3877b7a5fa45a0a37809f: Bug 1445533 - Remove a refcount in MozPromise::Steal(), and use MakeRefPtr a few places. r=gerald
Jan-Ivar Bruaroey <jib@mozilla.com> - Tue, 13 Mar 2018 09:55:01 -0400 - rev 453191
Push 148 by fmarier@mozilla.com at Thu, 29 Mar 2018 23:06:47 +0000
Bug 1445533 - Remove a refcount in MozPromise::Steal(), and use MakeRefPtr a few places. r=gerald MozReview-Commit-ID: 2jFzxiqBVpc
cf267f31cfdf88c230bbfc35854c468f637b107c: Bug 1445584 - add a move assignment operator for nsCOMPtr; r=gerald
Nathan Froyd <froydnj@mozilla.com> - Wed, 14 Mar 2018 08:22:30 -0400 - rev 453053
Push 148 by fmarier@mozilla.com at Thu, 29 Mar 2018 23:06:47 +0000
Bug 1445584 - add a move assignment operator for nsCOMPtr; r=gerald Somehow we didn't have this before. The lack thereof was causing us to do unnecessary things during event dispatch.
83875d9859a4d5514e0bebe2169255e9213e7988: Bug 547707: replaced hardcoded strings by definitions in nsMimeTypes.h r=gerald
Videet <videetssinghai@gmail.com> - Tue, 06 Mar 2018 17:43:57 +0100 - rev 452328
Push 148 by fmarier@mozilla.com at Thu, 29 Mar 2018 23:06:47 +0000
Bug 547707: replaced hardcoded strings by definitions in nsMimeTypes.h r=gerald MozReview-Commit-ID: 6f85pRUe8Tg
757abead90e1058c94ab43320341c65c954900ac: Bug 1441779 - Part 1: Change nsTArray<SourceBufferTask> to nsTArray<RefPtr<SourceBufferTask>> r=gerald
Miko Mynttinen <mikokm@gmail.com> - Thu, 01 Mar 2018 11:55:18 +0100 - rev 451242
Push 148 by fmarier@mozilla.com at Thu, 29 Mar 2018 23:06:47 +0000
Bug 1441779 - Part 1: Change nsTArray<SourceBufferTask> to nsTArray<RefPtr<SourceBufferTask>> r=gerald MozReview-Commit-ID: 8lV4Pp9p7Ch
df08d5fbb370b56dd7451496d655bb260ac1c396: Bug 1430942: Tighten up handling codec specific config. r=gerald
Dan Glastonbury <dan.glastonbury@gmail.com> - Fri, 02 Mar 2018 01:33:53 +0200 - rev 451084
Push 148 by fmarier@mozilla.com at Thu, 29 Mar 2018 23:06:47 +0000
Bug 1430942: Tighten up handling codec specific config. r=gerald MozReview-Commit-ID: 8GO6z7U1kx8
43f6075d028ea136010a394fe35ab2306916d982: Bug 1439433 - Length check FileBlockCache::mBlockChanges access. r=gerald
Chris Pearce <cpearce@mozilla.com> - Thu, 22 Feb 2018 15:00:18 +1300 - rev 449421
Push 144 by fmarier@mozilla.com at Fri, 23 Feb 2018 02:24:57 +0000
Bug 1439433 - Length check FileBlockCache::mBlockChanges access. r=gerald I can't for the life of me figure out how we get into the situation where the block change list is empty here, or how we can get past some of the existing null checks in the code, but we can at least add some more checks to hopefully ensure we don't crash... MozReview-Commit-ID: 168G94IyrWt
c5285ff72ac4a20435415f7d484ac6c3b6057a5a: Bug 1439433 - Length check FileBlockCache::mBlockChanges access. r=gerald
Chris Pearce <cpearce@mozilla.com> - Thu, 22 Feb 2018 15:00:18 +1300 - rev 449347
Push 144 by fmarier@mozilla.com at Fri, 23 Feb 2018 02:24:57 +0000
Bug 1439433 - Length check FileBlockCache::mBlockChanges access. r=gerald I can't for the life of me figure out how we get into the situation where the block change list is empty here, or how we can get past some of the existing null checks in the code, but we can at least add some more checks to hopefully ensure we don't crash... MozReview-Commit-ID: 168G94IyrWt
6873d1a8a2adc31911ba6a2d0548921bdb24728a: Bug 1435262 - Remove the const type qualifiers to silent a gcc 8 warning r=gerald
Sylvestre Ledru <sledru@mozilla.com> - Fri, 02 Feb 2018 15:49:00 +0100 - rev 446895
Push 132 by fmarier@mozilla.com at Wed, 07 Feb 2018 00:44:43 +0000
Bug 1435262 - Remove the const type qualifiers to silent a gcc 8 warning r=gerald MozReview-Commit-ID: 28NSnyYMPQ0
eb87c9467a3598f9cf9fddfcb31fdcc42fc79241: Bug 1435262 - Remove the const type qualifiers to silent a gcc 8 warning r=gerald
Sylvestre Ledru <sledru@mozilla.com> - Fri, 02 Feb 2018 15:49:00 +0100 - rev 446890
Push 132 by fmarier@mozilla.com at Wed, 07 Feb 2018 00:44:43 +0000
Bug 1435262 - Remove the const type qualifiers to silent a gcc 8 warning r=gerald MozReview-Commit-ID: 28NSnyYMPQ0
f1373fe3d25c585cf234d877094126b2f7fef821: Bug 1433309 - Annotate createMediaKeys promise reject with whether failure occurred due to pending shutdown. r=gerald
Chris Pearce <cpearce@mozilla.com> - Fri, 26 Jan 2018 12:20:35 +1300 - rev 445940
Push 126 by fmarier@mozilla.com at Thu, 01 Feb 2018 01:20:50 +0000
Bug 1433309 - Annotate createMediaKeys promise reject with whether failure occurred due to pending shutdown. r=gerald Around every Firefox update, Netflix see a spike in MediaKeySystemAccess.createMediaKeys() promise rejects. I am wondering whether this is caused by the browser restarting to apply a Firefox update while Netflix's player is loading. So add more detail to the promise reject as to the state of the system, to try to validate that theory. MozReview-Commit-ID: 4IDPsFwKCtq
3af31d84214f7cf4617a6798917ed1a0154f569d: Bug 1428719 - Adjust decrypt throttle to smooth out decrypt bursting when seeking. r=gerald
Chris Pearce <cpearce@mozilla.com> - Fri, 19 Jan 2018 14:55:12 +1300 - rev 444515
Push 124 by fmarier@mozilla.com at Tue, 30 Jan 2018 01:38:36 +0000
Bug 1428719 - Adjust decrypt throttle to smooth out decrypt bursting when seeking. r=gerald We've had reports from a partner that their EME content is failing after a seek. This is caused by decrypting EME content too fast after a seek triggering bug 1338924. So we need to make our decryption less "bursty". So adjust our decrypt throttling to instead throttle the decrypt to 200ms of content every 100ms of wall-clock-time, rather than 2s of content every 1s of wall-clock-time. This makes us decode more smoothly, and appears to resolve the issue with the partner stream. MozReview-Commit-ID: 1nNuTJrreN
4ae35d2b108a7d98b98a6d6d3943b78b297dd35b: Bug 1416084. P2 - wake up readers when download is suspended. r=bechen,gerald
JW Wang <jwwang@mozilla.com> - Mon, 15 Jan 2018 10:25:11 +0800 - rev 443954
Push 124 by fmarier@mozilla.com at Tue, 30 Jan 2018 01:38:36 +0000
Bug 1416084. P2 - wake up readers when download is suspended. r=bechen,gerald It is possible that download is suspended after all bytes are received but before OnStopRequest is notified. In that case, we will fail to wake up the readers waiting to read the partial block and cause bug 1412737 comment 30. MozReview-Commit-ID: GUk4lXO6Upk
e3ca97a44f1c09771297071bb459d07e0ccf15b2: Bug 1416084. P1 - reapply bug 1412737 P4: wake up readers only when we have blocks committed to the cache. r=bechen,gerald
JW Wang <jwwang@mozilla.com> - Mon, 15 Jan 2018 10:09:02 +0800 - rev 443953
Push 124 by fmarier@mozilla.com at Tue, 30 Jan 2018 01:38:36 +0000
Bug 1416084. P1 - reapply bug 1412737 P4: wake up readers only when we have blocks committed to the cache. r=bechen,gerald Per bug 1412737 P2 changes, a reader will always read data from the cache or from the last block in the memory. NotifyDataReceived() will be slightly more efficient if we don't wake up readers unnecessarily when there are no new blocks committed to the cache. MozReview-Commit-ID: Afcy5OOeIk3
72b59b20a04812e8f0b80bc27df09ebe0f884982: Bug 1429009. P2 - remove unnecessary AbstractThread::AutoEnter. r=bechen,gerald
JW Wang <jwwang@mozilla.com> - Tue, 09 Jan 2018 17:49:53 +0800 - rev 443125
Push 124 by fmarier@mozilla.com at Tue, 30 Jan 2018 01:38:36 +0000
Bug 1429009. P2 - remove unnecessary AbstractThread::AutoEnter. r=bechen,gerald It is not needed since we are already in the right context. MozReview-Commit-ID: 6QXxLMQHavv
9a6c837be3b44d1fa175b0f00bb405d56ee2960a: Bug 1429009. P1 - notify 'DownloadProgressed' immediately to preserve the event order. r=bechen,gerald
JW Wang <jwwang@mozilla.com> - Mon, 08 Jan 2018 14:51:57 +0800 - rev 443124
Push 124 by fmarier@mozilla.com at Tue, 30 Jan 2018 01:38:36 +0000
Bug 1429009. P1 - notify 'DownloadProgressed' immediately to preserve the event order. r=bechen,gerald We also need to call GetOwner()->UpdateReadyState() since mCanPlayThrough might have changed. MozReview-Commit-ID: C5djzu1sXqV
1599fb3b361022c56638a1570272068dc69eb268: Bug 1428951 - add a getter for MediaCacheStream::GetDebugInfo() to take the cache monitor on the main thread. r=bechen,gerald
JW Wang <jwwang@mozilla.com> - Tue, 09 Jan 2018 10:29:10 +0800 - rev 443055
Push 124 by fmarier@mozilla.com at Tue, 30 Jan 2018 01:38:36 +0000
Bug 1428951 - add a getter for MediaCacheStream::GetDebugInfo() to take the cache monitor on the main thread. r=bechen,gerald MozReview-Commit-ID: 8qkZg0MJsGQ
8932f0853269bfdaf7cc043ae4b1167a1cf28ad0: Bug 1427699. P2 - rewrite InitAsCloneInternal() to mimic the case where data is downloaded from a channel. r=gerald
JW Wang <jwwang@mozilla.com> - Fri, 05 Jan 2018 16:31:39 +0800 - rev 442713
Push 124 by fmarier@mozilla.com at Tue, 30 Jan 2018 01:38:36 +0000
Bug 1427699. P2 - rewrite InitAsCloneInternal() to mimic the case where data is downloaded from a channel. r=gerald Note we add mClient->CacheClientSuspend() so the network state of the element is changed to IDLE because we have no channel to fetch data initially. MozReview-Commit-ID: DgJbMxvJBzH
15c31a08d6e3d530c0ed84bc162f5370d2e5ce29: Bug 1427699. P1 - allow element->DownloadSuspended() to be called when mChannel is null. r=gerald
JW Wang <jwwang@mozilla.com> - Fri, 05 Jan 2018 16:04:16 +0800 - rev 442712
Push 124 by fmarier@mozilla.com at Tue, 30 Jan 2018 01:38:36 +0000
Bug 1427699. P1 - allow element->DownloadSuspended() to be called when mChannel is null. r=gerald This is required by P2 where we want to notify the 'suspend' event for a cloned resource whose mChannel is initially null. MozReview-Commit-ID: 3znDl2TqlqK
c1bacce3317103a68433841841feb36813854c0b: Bug 1428242. P3 - use a non-reentrant monitor. r=bechen,gerald
JW Wang <jwwang@mozilla.com> - Fri, 05 Jan 2018 10:49:38 +0800 - rev 442534
Push 124 by fmarier@mozilla.com at Tue, 30 Jan 2018 01:38:36 +0000
Bug 1428242. P3 - use a non-reentrant monitor. r=bechen,gerald MozReview-Commit-ID: GCXBHugTLJV
90ffd059bdfa679780d7d4dd99a259e0f457546b: Bug 1428242. P2 - MediaCache::ReadCacheFile() doesn't need to drop the cache monitor. r=bechen,gerald
JW Wang <jwwang@mozilla.com> - Fri, 05 Jan 2018 10:48:14 +0800 - rev 442533
Push 124 by fmarier@mozilla.com at Tue, 30 Jan 2018 01:38:36 +0000
Bug 1428242. P2 - MediaCache::ReadCacheFile() doesn't need to drop the cache monitor. r=bechen,gerald Since we now never take the lock on the main thread, we can safely do file IO while holding the lock without blocking the main thread. This reverts the change of Bug 1354389 P1. MozReview-Commit-ID: EhEwTjINQIT
a5f886553c4beaea9816d24058a797a033b3e804: Bug 1428242. P1 - assert we always take the MediaCache monitor off the main thread. r=bechen,gerald
JW Wang <jwwang@mozilla.com> - Wed, 20 Dec 2017 15:28:52 +0800 - rev 442532
Push 124 by fmarier@mozilla.com at Tue, 30 Jan 2018 01:38:36 +0000
Bug 1428242. P1 - assert we always take the MediaCache monitor off the main thread. r=bechen,gerald MozReview-Commit-ID: 6fxhzTOFaWx
1f394da377b0250c21506645b5535dc4a8e7cdf0: Bug 1428184 - copy mStreams before iterating the array. r=bechen,gerald
JW Wang <jwwang@mozilla.com> - Fri, 05 Jan 2018 11:34:59 +0800 - rev 442459
Push 124 by fmarier@mozilla.com at Tue, 30 Jan 2018 01:38:36 +0000
Bug 1428184 - copy mStreams before iterating the array. r=bechen,gerald It is bad to modify the array while iterating it. MozReview-Commit-ID: JbpoRmM7GqP