GNOME Bugzilla – Bug 97028
Rhythmbox crash on GStreamer thread malfunction.
Last modified: 2009-08-15 18:40:50 UTC
Package: GStreamer Severity: normal Version: 0.3.0.99 Synopsis: Rhythmbox crash on GStreamer thread malfunction. Bugzilla-Product: GStreamer Bugzilla-Component: gstreamer (core) BugBuddy-GnomeVersion: 2.0 (2.1.1) Description: Description of Problem: Random crash, looks like something to do with threading, which I can help little on. Looks like it tried to destroy a thread, failed because the device was busy, and then maybe the return value wasn't checked? This is a rhythmbox trace, but I think it has more to do with GStreamer. Debugging Information: Backtrace was generated from '/opt/gnome2/bin/rhythmbox' [New Thread 1024 (LWP 16587)] [New Thread 2049 (LWP 16588)] [New Thread 1026 (LWP 16589)] [New Thread 2051 (LWP 16590)] [New Thread 3076 (LWP 16591)] [New Thread 4101 (LWP 16592)] 0x40c881e9 in __wait4 () from /lib/libc.so.6
+ Trace 29522
Thread 5 (Thread 3076 (LWP 16591))
------- Bug moved to this database by unknown@bugzilla.gnome.org 2002-10-28 09:21 ------- Unknown version unspecified in product GStreamer. Setting version to "0.3.3". Reassigning to the default owner of the component, gstreamer-maint@bugzilla.gnome.org.
Version: 0.4.2.1 I'm seeing propably the same thing on this SMP. It's totally random when the crash occurs, sometimes only minutes, sometimes after hours of running. On the console I get: GThread-ERROR **: file gthread-posix.c: line 228 (g_cond_free_posix_impl): error 'Device or resource busy' during 'pthread_cond_destroy ((pthread_cond_t *) cond)' aborting... Xlib: unexpected async reply (sequence 0x296f3)!
Created attachment 12130 [details] Stack trace for the previous
this is 0.4.2 for me, and I also have SMP (dual celeron 366). Maybe this is something deeper like a pthread/SMP issue? (I also submitted this to rhythmbox as I haven't been able to reproduce it with gst-player yet. Hmm . . .) Also confirming since someone else reports they can see it:)
Just because I also have dual celeron 366 doesn't mean it's processors' fault... :) I'm posting here, because I'm sufficiently certain that rhythmbox has nothing to do with it. Now I got a crash with the basicwingo scheduler. Output to the console was identical to that in my previous comment. Boy is this fun or what :)
Created attachment 12137 [details] Stack trace for crash with basicwingo scheduler
*** Bug 97920 has been marked as a duplicate of this bug. ***
Ok, all thread related issues seems to be related to thread code for i686 in RH (and probably other i686 optimized glibc versions out there). glibc in RH8.1 solves this. Putting on needinfo for some days before closing in case my assumption about the systems involved is mistaken.
Nothing heard so closing.