1914844 - Crash in [@ mozilla::media::Interval<T>::Interval<T> | mozilla::TrackBuffersManager::RemoveFrames]

Status: ASSIGNED

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

Description

[Gabriele Svelto [:gsvelto]]

Crash report: https://crash-stats.mozilla.org/report/index/3c5447a0-6d6c-442c-8b2b-1dc740240824

MOZ_CRASH Reason: MOZ_DIAGNOSTIC_ASSERT(mStart <= mEnd) (Invalid Interval)

Top 10 frames:

0  xul.dll  mozilla::media::Interval<mozilla::media::TimeUnit>::Interval<mozilla::media::...  dom/media/Intervals.h:48
1  xul.dll  mozilla::TrackBuffersManager::RemoveFrames(mozilla::media::TimeIntervals cons...  dom/media/mediasource/TrackBuffersManager.cpp:2637
2  xul.dll  mozilla::TrackBuffersManager::InsertFrames(nsTArray<RefPtr<mozilla::MediaRawD...  dom/media/mediasource/TrackBuffersManager.cpp:2395
3  xul.dll  mozilla::TrackBuffersManager::ProcessFrames(nsTArray<RefPtr<mozilla::MediaRaw...  dom/media/mediasource/TrackBuffersManager.cpp:2284
4  xul.dll  mozilla::TrackBuffersManager::CompleteCodedFrameProcessing()  dom/media/mediasource/TrackBuffersManager.cpp:1829
5  xul.dll  RefPtr<mozilla::TrackBuffersManager>::assign_assuming_AddRef(mozilla::TrackBu...  mfbt/RefPtr.h:65
5  xul.dll  RefPtr<mozilla::TrackBuffersManager>::operator=(void*)  mfbt/RefPtr.h:180
5  xul.dll  mozilla::MozPromise<RefPtr<mozilla::MediaTrackDemuxer::SamplesHolder>, mozill...  xpcom/threads/MozPromise.h:746
6  xul.dll  mozilla::MozPromise<bool, mozilla::MediaResult, 1>::ThenValueBase::DoResolveO...  xpcom/threads/MozPromise.h:621
6  xul.dll  mozilla::MozPromise<bool, mozilla::MediaResult, 1>::ThenValueBase::ResolveOrR...  xpcom/threads/MozPromise.h:488

This is an assertion that was introduced in bug 1530322 but started triggering relatively recently. The oldest build IDs that appear in this crash signature are 20240710091735 for beta and 20240728093905 for nightly. It's odd that we started seeing the crash in beta slightly before nightly, but it's possible that the bug was triggered by a change that was quickly uplifted. It would be interesting to know what are the values of mStart and mEnd, I'll have a look if I can retrieve them from a minidump.

Comment 1

[Karl Tomlinson (:karlt)]

Thanks! Since the fix for bug 1906667, this is tracking the more recent and more frequent crash seen in bug 1906088.

Comment 2

[BugBot [:suhaib / :marco/ :calixte]]

Set release status flags based on info from the regressing bug 1903974

Comment 3

[BugBot [:suhaib / :marco/ :calixte]]

The bug is linked to a topcrash signature, which matches the following criterion:

• Top 10 content process crashes on beta

:karlt, could you consider increasing the severity of this top-crash bug?

For more information, please visit BugBot documentation.

Comment 4

[Karl Tomlinson (:karlt)]

Leaving S3 because this doesn't affect release builds, but i am actively working on dependencies that are likely to improve the situation here.

Comment 5

[BugBot [:suhaib / :marco/ :calixte]]

Based on the topcrash criteria, the crash signatures linked to this bug are not in the topcrash signatures anymore.

For more information, please visit BugBot documentation.

Comment 6

[Dianna Smith [:diannaS]]

:karlt seems the volume picked up during the early beta cycle starting in june. As suspected they stopped in b7, but did something recently land that made this worse for early beta users?

Comment 7

[Karl Tomlinson (:karlt)]

The early June Beta spike (repeated for subsequent Betas) would have been for Firefox 140.
Crashes in the Beta 140 period initially reported from May 29 at rates not too dissimilar to previous release times, but then increased from June 6.

Nightly crash report numbers are low, but increased around the same time as the Beta 140 release (May 29), but there was a spike on May 20, and an increase mid June.

I looked through jj log -r FIREFOX_NIGHTLY_139_END::FIREFOX_BETA_140_END dom/media/ and didn't find anything that looked likely to influence the rate of these reports.
It seems more likely that something external has changed.

I'll see if I can get back to resolving some of the inconsistencies that would lead to this kind of assertion failure, perhaps before the next Beta.