1367702 - Crashes in __pthread_kill | abort | __assert_rtn (when plugging/unplugging headphones on OSX)

Status: RESOLVED FIXED

(Core :: Audio/Video: cubeb, defect, P1)

Description

[[:philipp]]

This bug was filed from the Socorro interface and is 
report bp-81f71cad-5b37-4cd1-814d-676360170523.
=============================================================
Crashing Thread (65)
Frame 	Module 	Signature 	Source
0 	libsystem_kernel.dylib 	__pthread_kill 	
1 	libsystem_c.dylib 	abort 	
2 	libsystem_c.dylib 	__assert_rtn 	
3 	XUL 	audiounit_property_listener_callback 	media/libcubeb/src/cubeb_utils_unix.h:58
4 	CoreAudio 	HALPropertyListener::Call(unsigned int, unsigned int, AudioObjectPropertyAddress const*) 	
5 	CoreAudio 	HALObject::PropertiesChanged(unsigned int, AudioObjectPropertyAddress const*) 	
6 	CoreAudio 	HALSystem::AudioObjectPropertiesChanged(AudioHardwarePlugInInterface**, unsigned int, unsigned int, AudioObjectPropertyAddress const*) 	
7 	CoreAudio 	_SendPropertiesChanged(void*) 	
8 	libdispatch.dylib 	_dispatch_client_callout 	
9 	libdispatch.dylib 	_dispatch_queue_drain 	
10 	libdispatch.dylib 	_dispatch_queue_invoke 	
11 	libdispatch.dylib 	_dispatch_root_queue_drain 	
12 	libdispatch.dylib 	_dispatch_worker_thread3 	
13 	libsystem_pthread.dylib 	_pthread_wqthread 	
14 	libsystem_pthread.dylib 	start_wqthread 	
15 	libdispatch.dylib 	dispatch_source_set_cancel_handler

crashes with this signature have become more frequent since firefox 52 - currently such reports are accounting for 2.2% of browser crashes of mac users on 53.0.3. i think bug 1351728 tried to address the same or a very similar issue, but the crashes are still going on in all channels including 55.0a1.

some user comments about the issues:
https://crash-stats.mozilla.com/signature/?platform=Mac%20OS%20X&proto_signature=~audiounit_property_listener_callback%20%7C%20HALPropertyListener%3A%3ACall&signature=__pthread_kill%20%7C%20abort%20%7C%20__assert_rtn&date=%3E%3D2017-04-25T09%3A18%3A12.000Z#comments

(due to changes in signature generation these reports would have shown up as parts of the [@ __pthread_kill] or [@ 	libsystem_kernel.dylib@0x16f06] signatures before may 10).

Comment 1

[Ryan VanderMeulen [:RyanVM]]

Is this fixed now by bug 1345049?

Comment 2

[Alex Chronopoulos [:achronop]]

Yeah, forgot to close it.