[MPlayer-users] Error occurring on q or >, not sure if mplayer or not

[stan]

On Sat, 29 Sep 2012 20:39:11 +0200
Reimar Döffinger <Reimar.Doeffinger at gmx.de> wrote:

> On Thu, Sep 27, 2012 at 01:38:02PM -0700, stan wrote:
> > It seems to be a problem with nouveau.  The output is large, but
> > here it is:
> > 
> > Program received signal SIGSEGV, Segmentation
> > fault. 0x00007fffebf4fb96 in nouveau_grobj_autobind ()
> > from /lib64/libdrm_nouveau.so.1 Missing separate debuginfos, use:
> > debuginfo-install libXcursor-1.1.13-1.fc17.x86_64
> > libXrender-0.9.7-1.fc17.x86_64 mesa-dri-drivers-8.0.4-1.fc17.x86_64
> > (gdb) bt #0  0x00007fffebf4fb96 in nouveau_grobj_autobind ()
> > from /lib64/libdrm_nouveau.so.1                               
> 
> Yes, I think it is likely that is a bug in nouveau code, though
> installing the debug packaged for it and maybe even running
> through valgrind might give more useful information.

This seems the most fruitful path.  The debug packages I've used before.
I don't think I've ever used valgrind.  Results could take a while since
I'll have the learning curve.

> 
> > #11 0x0000003b39e1d1e1 in glXDestroyContext ()
> > from /lib64/libGL.so.1 #12 0x00000000004c66a7 in
> > releaseGlContext_x11 (ctx=0x10b2f80) at
> > libvo/gl_common.c:2132                          
> 
> Obviously as a workaround you can comment out the call to
> glXDestroyContext at that line. The worst that should happen is that
> there will be a memleak, but if you don't play hundreds of files
> with the same MPlayer and without -fixed-vo it should not be a real
> issue.

I'll just use the -fixed-vo option.

> 
> > (gdb) disass $pc-32
> > $pc+32 A syntax error in expression, near
> > `$pc+32'.                                                                      
> 
> With newer gcc that should be
> disass $pc-32,$pc+32
> Yes, it is annoying that they changed syntax.

Since I was unfamiliar with the previous syntax, and was just using
rote cut and paste, it went right by me.

I'll have to recompile the debug version in order to do this again.  Is
there a way to use configuration flags in mplayer to tell it to turn off
optimization, -O0, so I can test if it is compiler optimization that is
causing the problem?  I can do it in the makefile if necessary, of
course.