1588198 - Crash in [@ mozilla::H264ChangeMonitor::PrepareSample]

Status: RESOLVED DUPLICATE of bug 1617929

(Core :: Audio/Video: Playback, defect, P2)

Description

[Marcia Knous [:marcia]]

This bug is for crash report bp-1fd2fc0f-498c-41df-9fae-6da5a0191011.

Seen while looking at nightly crash stats - this is a reproducible crash. Not sure if someone internal is testing and just forgot to add a note in the comment box?

URL is https://qa.unified-streaming.com/firefoxbug/test.html

https://bit.ly/2M8KRMg. There are some Linux crashes with that Moz crash reason as well, with the same URL.

MOZ_DIAGNOSTIC_ASSERT(aConversion == MediaDataDecoder::ConversionRequired::kNeedAnnexB || aConversion == MediaDataDecoder::ConversionRequired::kNeedAVCC) (Conversion must be either AVCC or AnnexB)

Top 10 frames of crashing thread:

0 XUL mozilla::H264ChangeMonitor::PrepareSample dom/media/platforms/wrappers/MediaChangeMonitor.cpp:111
1 XUL mozilla::MozPromise<mozilla::TrackInfo::TrackType, mozilla::MediaResult, true>::ThenValue<mozilla::MediaChangeMonitor::CreateDecoderAndInit xpcom/threads/MozPromise.h:726
2 XUL mozilla::MozPromise<mozilla::TrackInfo::TrackType, mozilla::MediaResult, true>::ThenValueBase::ResolveOrRejectRunnable::Run xpcom/threads/MozPromise.h:402
3 XUL mozilla::TaskQueue::Runner::Run xpcom/threads/TaskQueue.cpp:199
4 XUL nsThreadPool::Run xpcom/threads/nsThreadPool.cpp:256
5 XUL non-virtual thunk to nsThreadPool::Run xpcom/threads/nsThreadPool.cpp
6 XUL nsThread::ProcessNextEvent xpcom/threads/nsThread.cpp:1225
7 XUL NS_ProcessNextEvent xpcom/threads/nsThreadUtils.cpp:486
8 XUL mozilla::ipc::MessagePumpForNonMainThreads::Run ipc/glue/MessagePump.cpp:333
9 XUL nsThread::ThreadFunc xpcom/threads/nsThread.cpp:458

Comment 1

[Matthew Gregan [:kinetik]]

Looks like this is related to bug 1560092, see https://bugzilla.mozilla.org/show_bug.cgi?id=1560092#c18 onwards.

Comment 2

[Bryce Seager van Dyk [:bryce] (he/him) - Not reading bugmail]

QA have been looking at this, as have I, some of the recent spike may be at least in part our testing.

Comment 3

[Bob Clary [:bc] (inactive)]

fyi this also hits

Assertion failure: aConversion == MediaDataDecoder::ConversionRequired::kNeedAnnexB || aConversion == MediaDataDecoder::ConversionRequired::kNeedAVCC (Conversion must be either AVCC or AnnexB), at /builds/worker/workspace/build/src/dom/media/platforms/wrappers/MediaChangeMonitor.cpp:114
#01: mozilla::MediaChangeMonitor::DecodeFirstSample(mozilla::MediaRawData*) [dom/media/platforms/wrappers/MediaChangeMonitor.cpp:565]
#02: mozilla::MediaChangeMonitor::CreateDecoderAndInit(mozilla::MediaRawData*)::$_6::operator()(mozilla::TrackInfo::TrackType) const [dom/media/platforms/wrappers/MediaChangeMonitor.cpp:0]
#03: mozilla::MozPromise<mozilla::TrackInfo::TrackType, mozilla::MediaResult, true>::ThenValue<mozilla::MediaChangeMonitor::CreateDecoderAndInit(mozilla::MediaRawData*)::$_6, mozilla::MediaChangeMonitor::CreateDecoderAndInit(mozilla::MediaRawData*)::$_7>::DoResolveOrRejectInternal(mozilla::MozPromise<mozilla::TrackInfo::TrackType, mozilla::MediaResult, true>::ResolveOrRejectValue&) [xpcom/threads/MozPromise.h:726]
#04: mozilla::MozPromise<mozilla::TrackInfo::TrackType, mozilla::MediaResult, true>::ThenValueBase::ResolveOrRejectRunnable::Run() [xpcom/threads/MozPromise.h:403]

Comment 4

[Rares Doghi, Desktop QA]

Hi, This issue still occurs in our latest nightly builds on Mac and Ubuntu machines using the url: https://live.unified-streaming.com/firefoxbug/test.html , but it does not occur in Beta 74.0b4 or Release 73.0, I will update the flags.

Comment 5

[Bryce Seager van Dyk [:bryce] (he/him) - Not reading bugmail]

(In reply to Rares Doghi from comment #4)

Hi, This issue still occurs in our latest nightly builds on Mac and Ubuntu machines using the url: https://live.unified-streaming.com/firefoxbug/test.html , but it does not occur in Beta 74.0b4 or Release 73.0, I will update the flags.

I would have expected this to still fail on those older versions. Could you please double check?

Comment 6

[Bryce Seager van Dyk [:bryce] (he/him) - Not reading bugmail]

I expect the change in bug 1617929 to fix this. I'm going to wait for it to land and then test.

Comment 7

[Bryce Seager van Dyk [:bryce] (he/him) - Not reading bugmail]

This is fixed by bug 1617929. They're somewhat different issues, but the changes there remedy this also.