31f42700e9ba5601bc7db181663ccb58acc0bbb3: Bug 1267918 - Add GMPCrashHelper for HTMLMediaElement. r=gerald draft
Chris Pearce <cpearce@mozilla.com> - Wed, 29 Jun 2016 11:42:07 +1200 - rev 382647
Push 21792 by cpearce@mozilla.com at Thu, 30 Jun 2016 03:07:44 +0000
Bug 1267918 - Add GMPCrashHelper for HTMLMediaElement. r=gerald This ensures that unencrypted GMP decoding crash reporting works. MozReview-Commit-ID: 84TAV5F9Ie0
f4350bf2035758eaf3a7824e5c1ee1f2f3e92710: Bug 1267918 - Add GMPCrashHelper for MediaKeys. r=gerald draft
Chris Pearce <cpearce@mozilla.com> - Wed, 29 Jun 2016 11:42:04 +1200 - rev 382646
Push 21792 by cpearce@mozilla.com at Thu, 30 Jun 2016 03:07:44 +0000
Bug 1267918 - Add GMPCrashHelper for MediaKeys. r=gerald So if a GMP crashes while doing EME, we'll get a crash report using the new mechanism. MozReview-Commit-ID: G8BlFI9jmiF
9871b0173275b211929d9edee27ad6c24ea8ad7b: Bug 1267918 - Add GMPCrashHelperHolder to automatically disconnect crash helpers on actor destory. r=gerald draft
Chris Pearce <cpearce@mozilla.com> - Wed, 29 Jun 2016 11:42:00 +1200 - rev 382645
Push 21792 by cpearce@mozilla.com at Thu, 30 Jun 2016 03:07:44 +0000
Bug 1267918 - Add GMPCrashHelperHolder to automatically disconnect crash helpers on actor destory. r=gerald Disconnecting the GMPCrashHelpers at the right time is hard, because in the crashing case we're all shutdown before the GMPCrashHelpers can be invoked to help handle the crash report. So add a class to help the helpers; GMPCrashHelperHolder. This composes into the GMP content protocol actors, to help them disconnect the crash helpers at the right time. See the comment in GMPCrashHelperHolder for the details. MozReview-Commit-ID: E5rE6e5Jues
440cc2848de4bdb921cfa5a1efe9ff58136fcac7: Bug 1267918 - Ensure GMPCrashHelper instances are destroyed on the main thread. r?karlt draft
Chris Pearce <cpearce@mozilla.com> - Thu, 30 Jun 2016 14:19:08 +1200 - rev 382644
Push 21792 by cpearce@mozilla.com at Thu, 30 Jun 2016 03:07:44 +0000
Bug 1267918 - Ensure GMPCrashHelper instances are destroyed on the main thread. r?karlt MozReview-Commit-ID: 7GFx6uoyAi9
676e1e805b23daed455df8066b2bebc6d07000e7: Bug 1267918 - Add GMPCrashHelper to GMPService::GetGMP* functions. r=gerald draft
Chris Pearce <cpearce@mozilla.com> - Wed, 29 Jun 2016 11:41:56 +1200 - rev 382643
Push 21792 by cpearce@mozilla.com at Thu, 30 Jun 2016 03:07:44 +0000
Bug 1267918 - Add GMPCrashHelper to GMPService::GetGMP* functions. r=gerald This enables callers to specify a way to determine the correct window to dispatch the PluginCrashed event to should the GMP actor crash. We need a way to determine the correct window at crash time, as the GMP's window can change at runtime. For example, if the GMP is being used for unencrypted decoding, the <video> element can be moved to a new browser window at runtime. Note: I don't handle disconnecting the GMPCrashHandlers in this patch; we do delete the GMPCrashHandlers in this patch when their associated GMP crashes, and in the next patch we handle disconnecting GMPCrashHandlers in the case where we don't crash. MozReview-Commit-ID: DrwcZAB6Ys0
bbbfa2fcc442bee488c3f21791620e77e992e84c: Bug 1267918 - Split LoadGMP message into select and load messages. r=gerald draft
Chris Pearce <cpearce@mozilla.com> - Wed, 29 Jun 2016 11:41:52 +1200 - rev 382642
Push 21792 by cpearce@mozilla.com at Thu, 30 Jun 2016 03:07:44 +0000
Bug 1267918 - Split LoadGMP message into select and load messages. r=gerald This will allow us to attach a crash handler to a GMP process after deciding which GMP to load but before actually loading it. MozReview-Commit-ID: HwBZU2Q4TX6
babf2601f53fd666bc11bc8ab3609c8897e3f976: Bug 1267918 - Add do_AddRef(const RefPtr<T>& aObj). r?froydnj draft
Chris Pearce <cpearce@mozilla.com> - Thu, 30 Jun 2016 10:06:10 +1200 - rev 382641
Push 21792 by cpearce@mozilla.com at Thu, 30 Jun 2016 03:07:44 +0000
Bug 1267918 - Add do_AddRef(const RefPtr<T>& aObj). r?froydnj This means we can return already_AddRefed<T> for any RefPtr<T>s being held as instance variables easier. MozReview-Commit-ID: HFHdkF8EUsK
881d4606e0bf4c7aa20b21d3d8b69f7ec91fa50c: Bug 1267918 - Remove obsolete GMP crash handling code. r=gerald draft
Chris Pearce <cpearce@mozilla.com> - Wed, 29 Jun 2016 11:42:14 +1200 - rev 382640
Push 21791 by cpearce@mozilla.com at Thu, 30 Jun 2016 03:02:56 +0000
Bug 1267918 - Remove obsolete GMP crash handling code. r=gerald MozReview-Commit-ID: EqzJagCHk7n
d9dc1909b3b24a8eeab36a0a5b5514cdb68e5f7e: Bug 1267918 - Add GMPCrashHelper for WebAudio buffer decoder. r?karlt draft
Chris Pearce <cpearce@mozilla.com> - Wed, 29 Jun 2016 11:42:10 +1200 - rev 382639
Push 21791 by cpearce@mozilla.com at Thu, 30 Jun 2016 03:02:56 +0000
Bug 1267918 - Add GMPCrashHelper for WebAudio buffer decoder. r?karlt This means if WebAudio is using the Adobe GMP for decoding and it crashes, we'll get a crash report for the GMP. MozReview-Commit-ID: FOZoPxvUwq5
2622551fd34389421e155313e849fede907d7e54: Bug 1267918 - Add GMPCrashHelper for HTMLMediaElement. r=gerald draft
Chris Pearce <cpearce@mozilla.com> - Wed, 29 Jun 2016 11:42:07 +1200 - rev 382638
Push 21791 by cpearce@mozilla.com at Thu, 30 Jun 2016 03:02:56 +0000
Bug 1267918 - Add GMPCrashHelper for HTMLMediaElement. r=gerald This ensures that unencrypted GMP decoding crash reporting works. MozReview-Commit-ID: 84TAV5F9Ie0
0f2828f92580612642c97686fc76fe40990dde22: Bug 1267918 - Add GMPCrashHelper for MediaKeys. r=gerald draft
Chris Pearce <cpearce@mozilla.com> - Wed, 29 Jun 2016 11:42:04 +1200 - rev 382637
Push 21791 by cpearce@mozilla.com at Thu, 30 Jun 2016 03:02:56 +0000
Bug 1267918 - Add GMPCrashHelper for MediaKeys. r=gerald So if a GMP crashes while doing EME, we'll get a crash report using the new mechanism. MozReview-Commit-ID: G8BlFI9jmiF
58f460b185e84205dd81ad229cd331b4d925f046: Bug 1267918 - Add GMPCrashHelperHolder to automatically disconnect crash helpers on actor destory. r=gerald draft
Chris Pearce <cpearce@mozilla.com> - Wed, 29 Jun 2016 11:42:00 +1200 - rev 382636
Push 21791 by cpearce@mozilla.com at Thu, 30 Jun 2016 03:02:56 +0000
Bug 1267918 - Add GMPCrashHelperHolder to automatically disconnect crash helpers on actor destory. r=gerald Disconnecting the GMPCrashHelpers at the right time is hard, because in the crashing case we're all shutdown before the GMPCrashHelpers can be invoked to help handle the crash report. So add a class to help the helpers; GMPCrashHelperHolder. This composes into the GMP content protocol actors, to help them disconnect the crash helpers at the right time. See the comment in GMPCrashHelperHolder for the details. MozReview-Commit-ID: E5rE6e5Jues
7b59d6c5639c6f7257235ec7821285b66590efb0: Bug 1267918 - Ensure GMPCrashHelper instances are destroyed on the main thread. r?karlt draft
Chris Pearce <cpearce@mozilla.com> - Thu, 30 Jun 2016 14:19:08 +1200 - rev 382635
Push 21791 by cpearce@mozilla.com at Thu, 30 Jun 2016 03:02:56 +0000
Bug 1267918 - Ensure GMPCrashHelper instances are destroyed on the main thread. r?karlt MozReview-Commit-ID: 7GFx6uoyAi9
bd50e52c2bc8c6bc137ac457484f4eea2fed2aee: Bug 1267918 - Add GMPCrashHelper to GMPService::GetGMP* functions. r=gerald draft
Chris Pearce <cpearce@mozilla.com> - Wed, 29 Jun 2016 11:41:56 +1200 - rev 382634
Push 21791 by cpearce@mozilla.com at Thu, 30 Jun 2016 03:02:56 +0000
Bug 1267918 - Add GMPCrashHelper to GMPService::GetGMP* functions. r=gerald This enables callers to specify a way to determine the correct window to dispatch the PluginCrashed event to should the GMP actor crash. We need a way to determine the correct window at crash time, as the GMP's window can change at runtime. For example, if the GMP is being used for unencrypted decoding, the <video> element can be moved to a new browser window at runtime. Note: I don't handle disconnecting the GMPCrashHandlers in this patch; we do delete the GMPCrashHandlers in this patch when their associated GMP crashes, and in the next patch we handle disconnecting GMPCrashHandlers in the case where we don't crash. MozReview-Commit-ID: DrwcZAB6Ys0
237176e6418ec133ee6686c527701a1a33d4bf30: Bug 1267918 - Split LoadGMP message into select and load messages. r=gerald draft
Chris Pearce <cpearce@mozilla.com> - Wed, 29 Jun 2016 11:41:52 +1200 - rev 382633
Push 21791 by cpearce@mozilla.com at Thu, 30 Jun 2016 03:02:56 +0000
Bug 1267918 - Split LoadGMP message into select and load messages. r=gerald This will allow us to attach a crash handler to a GMP process after deciding which GMP to load but before actually loading it. MozReview-Commit-ID: HwBZU2Q4TX6
60099a5ca6dfa574bf4f5188f9fbbd6b59769a30: Bug 1282585 - Add do_AddRef(const RefPtr<T>& aObj). r?froydnj draft
Chris Pearce <cpearce@mozilla.com> - Thu, 30 Jun 2016 10:06:10 +1200 - rev 382632
Push 21791 by cpearce@mozilla.com at Thu, 30 Jun 2016 03:02:56 +0000
Bug 1282585 - Add do_AddRef(const RefPtr<T>& aObj). r?froydnj MozReview-Commit-ID: HFHdkF8EUsK
65cb2192d0d92f516420600111f79c0d9d188eb2: Bug 1282658. Part 16 - remove checks for |mReader->IsWaiting{Audio,Video}Data()|. r=kaku draft
JW Wang <jwwang@mozilla.com> - Wed, 29 Jun 2016 17:26:13 +0800 - rev 382631
Push 21790 by jwwang@mozilla.com at Thu, 30 Jun 2016 02:34:24 +0000
Bug 1282658. Part 16 - remove checks for |mReader->IsWaiting{Audio,Video}Data()|. r=kaku EnsureAudioDecodeTaskQueued() is called from several places where mReader->IsWaitingAudioData() can be proven to be false: 1. OnAudioDecoded() - definitely false. 2. OnNotDecoded() - false because aReason is MediaDecoderReader::CANCELED. 3. OnSeekResolved() - false becuase we haven't requested any samples. 4. SetCallbacks() - false because AudioWaitCallback is resolved. MozReview-Commit-ID: 8ppYIQQw0uK
90a58d1adc83c4712a51b1c1612b2129aac9fd73: Bug 1282658. Part 15 - optimize checking of seek complete. r=kaku draft
JW Wang <jwwang@mozilla.com> - Wed, 29 Jun 2016 17:06:00 +0800 - rev 382630
Push 21790 by jwwang@mozilla.com at Thu, 30 Jun 2016 02:34:24 +0000
Bug 1282658. Part 15 - optimize checking of seek complete. r=kaku We need to request another audio sample in OnAudioDecoded() only when mDoneAudioSeeking is false which also applies OnVideoDecodd(). Therefore we remove calls to Ensure{Audio,Video}DecodeTaskQueued() from CheckIfSeekComplete() to prevent requesting video samples inside OnAudioDecoded() for they will be handled by OnVideoDecodd(). This also allows us to remove checking of mReader->IsRequesting{Audio,Video}Data() from Ensure{Audio,Video}DecodeTaskQueued(). MozReview-Commit-ID: LpXjiacp0D9
0ee33d5f34791935205aaf96ee764fb729e852d8: Bug 1282658. Part 14 - remove mDrop{Audio,Video}UntilNextDiscontinuity for MediaData::mDiscontinuity is guaranteed to be true for the 1st sample after seeking. r=kaku draft
JW Wang <jwwang@mozilla.com> - Wed, 29 Jun 2016 16:52:09 +0800 - rev 382629
Push 21790 by jwwang@mozilla.com at Thu, 30 Jun 2016 02:34:24 +0000
Bug 1282658. Part 14 - remove mDrop{Audio,Video}UntilNextDiscontinuity for MediaData::mDiscontinuity is guaranteed to be true for the 1st sample after seeking. r=kaku MozReview-Commit-ID: CttpYi6CV32
e23e2b16fa0fd031c643b5307e88b73a352cf79d: Bug 1282658. Part 13 - replace use of int64_t with media::TimeUnit. r=kaku draft
JW Wang <jwwang@mozilla.com> - Tue, 28 Jun 2016 17:16:31 +0800 - rev 382628
Push 21790 by jwwang@mozilla.com at Thu, 30 Jun 2016 02:34:24 +0000
Bug 1282658. Part 13 - replace use of int64_t with media::TimeUnit. r=kaku MozReview-Commit-ID: 6LVuqpgWGUg
(0) -300000 -100000 -30000 -10000 -3000 -1000 -300 -100 -50 -20 +20 +50 +100 +300 +1000 +3000 +10000 +30000 +100000 +300000 tip