1828101 - Crash in [@ Resampler2::PushConvert]

Status: RESOLVED WORKSFORME

(Core :: Audio/Video, defect, P4)

Description

[Sebastian Hengst [:aryx] (needinfo me if it's about an intermittent or backout)]

52 crashes, only on macOS 13 but might have different signature on other OS versions because the stack is corrupted. 0 crashes for Firefox 111.0, 18 crashes for Firefox 111.0.1, already 25 for Firefox 112.0 release earlier this week.

Crash report: https://crash-stats.mozilla.org/report/index/d47dde62-9329-4741-9d4a-727990230414

MOZ_CRASH Reason: MOZ_DIAGNOSTIC_ASSERT(run->mMagic == 0x384adf93)

Top 10 frames of crashing thread:

0  AudioToolboxCore  Resampler2::PushConvert  
1  AudioToolboxCore  acv2::Resampler2Wrapper::ProduceOutput  
2  AudioToolboxCore  acv2::AudioConverterChain::ProduceOutput  
3  AudioToolboxCore  acv2::_AudioConverterFillComplexBuffer  
4  AudioToolboxCore  AudioConverterFillComplexBuffer  
5  CoreAudio  CoreAudio@0x8219  
6  CoreAudio  CoreAudio@0x1119b9  
7  CoreAudio  CoreAudio@0xc170  
8  CoreAudio  HALC_ProxyIOContext::IOWorkLoop  
9  CoreAudio  invocation function for block in HALC_ProxyIOContext::HALC_ProxyIOContext  

Comment 1

[Paul Adenot (:padenot)]

Matthew, I can't think of anything else but audioipc here, or the memory is trashed by something else and we're unlucky.

Comment 2

[Matthew Gregan [:kinetik]]

The common factor seems to be macOS 13.3.x and unsupported hardware (likely using something like Open Core Legacy Patcher, which was mentioned in one of the crash comments). The crashes show a pattern of Intel GPU adapter IDs corresponding to HD2000, HD3000, and HD4000 chipsets (including multiple with a Dell Latitude D500) - any Mac hardware with these chipsets appear to be from 2011-2012 and fall outside the supported hardware for macOS 13 (which only goes back as far as 2017 Macs).

macOS 13.3.0 was released 2023-03-27, 13.3.1 was released 2023-04-07. The earliest crashes with this signature appear 2023-04-07 and show up as early as Fx102.9 ESR.

AudioIPC is enabled from Fx113 onwards, which shows these crash stacks moving from content to parent process.

Comment 3

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

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

• Top 5 desktop browser crashes on Mac on release

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

For more information, please visit auto_nag documentation.

Comment 4

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

Closing because no crashes reported for 12 weeks.