[uClibc]help : SIGSEGV in libpthread with xmms

[Simon Davis]

Erik, 

--- Erik Andersen <andersen at codepoet.org> wrote:
> On Tue Mar 25, 2003 at 09:05:43AM -0800, Simon Davis wrote:
> > Hi all,
> > 
> > I've successfully built xmms under the uclibc root fs.  However when running xmms it crashes
> > consistently when doing certain things.
> > 
> > It crashes with with a SIGSEGV in libpthread doing any of the following things ( See error
> below )
> >   o  Staring up with a mp3 file in the command line. 
> >   o  Startup reading the ~/.xmms/xmms.m3u.  It starts up fine if I delete the xmms.mp3 file.
> >   o  Selecting a new skin.
> >   o  Clicking on the close button.
> > 
> > This is the output of gdb -
> > 
> > Program received signal SIGSEGV, Segmentation fault.
> > 0x4000fcd6 in __pthread_unlock () from /lib/libpthread.so.0
> 
> Since it is consistant, and since you can catch it under gdb
> (which is very good), could you get a backtrace after the
> segfault?  Any chance of compiling uClibc with DODEBUG enabled
> and reproducing the segfault?  That would give even more detail,

Here's the backtrace.  As for compiling with DODEBUG I'll have to compile my own root fs and give
it a try.

#0  0x4000fcd6 in __pthread_unlock () from /lib/libpthread.so.0
#1  0x4000ca1b in pthread_mutex_unlock () from /lib/libpthread.so.0
#2  0x0808343d in playlistwin_update_sinfo ()
#3  0x0805d394 in util_read_menu_rc ()
#4  0x0805cf6a in util_read_menu_rc ()
#5  0x400ad15f in gtk_marshal_NONE__NONE (object=0x81cf860, 
    func=0x805cf48 <util_read_menu_rc+1360>, func_data=0x81c3900, 
    args=0xbfffee50) at gtkmarshal.c:312
#6  0x400dc722 in gtk_handlers_run (handlers=0x81c3900, 
    signal=0xbfffedf0, object=0x81cf860, params=0xbfffee50, after=0)
    at gtksignal.c:1917
#7  0x400dba32 in gtk_signal_real_emit (object=0x81cf860, signal_id=90, 
    params=0xbfffee50) at gtksignal.c:1477
#8  0x400d9afc in gtk_signal_emit (object=0x81cf860, signal_id=90)
    at gtksignal.c:552
#9  0x4004ce94 in gtk_button_clicked (button=0x81cf860) at gtkbutton.c:336
#10 0x4004e402 in gtk_real_button_released (button=0x81cf860)
    at gtkbutton.c:861
#11 0x400ad15f in gtk_marshal_NONE__NONE (object=0x81cf860, 
    func=0x4004e340 <gtk_real_button_released>, func_data=0x0, 
    args=0xbffff1e0) at gtkmarshal.c:312
#12 0x400db943 in gtk_signal_real_emit (object=0x81cf860, signal_id=89, 
    params=0xbffff1e0) at gtksignal.c:1440
#13 0x400d9afc in gtk_signal_emit (object=0x81cf860, signal_id=89)
    at gtksignal.c:552
#14 0x4004cdd8 in gtk_button_released (button=0x81cf860)
    at gtkbutton.c:327
#15 0x4004dd9e in gtk_button_button_release (widget=0x81cf860, 
    event=0x817d018) at gtkbutton.c:721
#16 0x400acf35 in gtk_marshal_BOOL__POINTER (object=0x81cf860, 
    func=0x4004dcf8 <gtk_button_button_release>, func_data=0x0, 
    args=0x400174a0) at gtkmarshal.c:28
#17 0x400dba67 in gtk_signal_real_emit (object=0x81cf860, signal_id=21, 
    params=0xbffff580) at gtksignal.c:1492
#18 0x400d9afc in gtk_signal_emit (object=0x81cf860, signal_id=21)
    at gtksignal.c:552
#19 0x4010f69d in gtk_widget_event (widget=0x81cf860, event=0x817d018)
    at gtkwidget.c:2864
#20 0x400acdb8 in gtk_propagate_event (widget=0x81cf860, event=0x817d018)
    at gtkmain.c:1385
#21 0x400abea9 in gtk_main_do_event (event=0x817d018) at gtkmain.c:818
#22 0x40154464 in gdk_event_dispatch (source_data=0x0, 
#23 0x40183999 in g_main_dispatch (dispatch_time=0xbffff950)
    at gmain.c:656
#24 0x40183e50 in g_main_iterate (block=0, dispatch=1) at gmain.c:877
#25 0x40184080 in g_main_run (loop=0x81a3418) at gmain.c:935
#26 0x400ab96b in gtk_main () at gtkmain.c:524
#27 0x0807843e in main ()
#28 0x4025c24a in __uClibc_start_main () from /lib/libc.so.0
#29 0x08055bd8 in _start ()


> 
> > I've look through the mailing lists but can't find any problems with __pthread_unlock. 
> 
> I'm not aware of any problems with __pthread_unlock, and I've
> tested it pretty extensively.  But maybe you have found a special
> corner case where it fails?
> 
>  -Erik
> 
> --
> Erik B. Andersen             http://codepoet-consulting.com/
> --This message was written using 73% post-consumer electrons--
> _______________________________________________
> uClibc mailing list
> uClibc at uclibc.org
> http://uclibc.org/mailman/listinfo/uclibc


__________________________________________________
Do you Yahoo!?
Yahoo! Platinum - Watch CBS' NCAA March Madness, live on your desktop!
http://platinum.yahoo.com