825480 - crash when quitting firefox after completing the peer connection test page [@ nsTimerEvent::'scalar deleting destructor']

Status: RESOLVED DUPLICATE of bug 799142

(Core :: WebRTC, defect, P1)

Description

[Jason Smith [:jsmith]]

This bug was filed from the Socorro interface and is 
report bp-a1dea194-7992-4c21-acad-27e282121230 .
============================================================= 

Steps:

1. Go to http://mozilla.github.com/webrtc-landing/pc_test.html
2. Select Start
3. Wait for hip hip horray
4. Quit Firefox

Expected:

Firefox should shutdown without a crash.

Actual:

A crash occurs during shutdown. Reproducible too on Win 7.

Comment 1

[Jason Smith [:jsmith]]

0 	xul.dll 	nsTimerEvent::`scalar deleting destructor' 	
1 	xul.dll 	nsRunnable::Release 	obj-firefox/xpcom/build/nsThreadUtils.cpp:31
2 	xul.dll 	nsThread::ProcessNextEvent 	xpcom/threads/nsThread.cpp:633
3 	xul.dll 	nsThread::ThreadFunc 	xpcom/threads/nsThread.cpp:265
4 	nspr4.dll 	_PR_NativeRunThread 	nsprpub/pr/src/threads/combined/pruthr.c:395
5 	nspr4.dll 	pr_root 	nsprpub/pr/src/md/windows/w95thred.c:90
6 	msvcr100.dll 	_callthreadstartex 	f:\dd\vctools\crt_bld\self_x86\crt\src\threadex.c:314
7 	msvcr100.dll 	_threadstartex 	f:\dd\vctools\crt_bld\self_x86\crt\src\threadex.c:292
8 	kernel32.dll 	BaseThreadInitThunk 	
9 	ntdll.dll 	__RtlUserThreadStart 	
10 	ntdll.dll 	_RtlUserThreadStart

Comment 2

[Adam Roach [:abr]]

Attachment: Backtrace of OS X hang

I note that this test scenario does not crash under OS X, but does reliably lead to a hang on shutdown. The attached backtrace shows the stacks for all process threads after such a hang has occurred.

This may or may not be the same bug.

In the context of this error report, thread 41 seems particularly suspect.

Comment 3

[Adam Roach [:abr]]

This is ridiculously difficult to debug right now due to Bug 826115, which brings my whole machine to its knees when running a debug Windows build. I'm still trying, but this is taking much longer than it should...

Comment 4

[Randell Jesup [:jesup] (needinfo me)]

Lets try making the one-liner patch suggested there to speed testing it

Comment 5

[Adam Roach [:abr]]

Yeah, I was trying to chase down where this was introduced. The check-in is change set 1a2e77002690 from Monday Dec 31, but I can't track that back to a bug. I'll just make the change locally.

Comment 6

[Jason Smith [:jsmith]]

I think this is a dupe of bug 799142 after doing some stack comparison.