[FFmpeg-devel] MP3 file that crashes ffmpeg

[Reimar Döffinger]

Hello,
On Sun, May 06, 2007 at 03:36:36PM +0200, Michael Niedermayer wrote:
> On Sun, May 06, 2007 at 11:03:01AM +0200, Diego Biurrun wrote:
> > On Sat, May 05, 2007 at 11:16:32PM +0200, Michael Niedermayer wrote:
> > > 
> > > also i wont look at crashes if the person reporting the crash is too lazy
> > > to provide gdb output
> > 
> > Note that MPlayer plays the sample fine, even with lavc decoder and lavf
> > demuxer ..
> > 
> > Here's your backtrace, let me know if you need anything else.
> > 
> > Diego
> > 
> > (gdb) run /var/tmp/bugs/11-i_need_your_love-daw.mp3
> > Starting program: /usr/src/ffmpeg/ffplay_g
> > /var/tmp/bugs/11-i_need_your_love-daw.mp3
> > Failed to read a valid object file image from memory.
> [...]
> > Program received signal SIGSEGV, Segmentation fault.
> > [Switching to Thread -1221788752 (LWP 4312)]
> > 0xfffffff8 in ?? ()
> > (gdb) bt
> > #0  0xfffffff8 in ?? ()
> > #1  0x0000000d in ?? ()
> > #2  0xffffffe5 in ?? ()
> > #3  0x00000002 in ?? ()
> > #4  0x00000000 in ?? ()
> 
> very usefull backtrace ...

Unfortunately there seems to be not much more info. Even valgrind only
spits out this useless thing:

==18651== Thread 3:
==18651== Jump to the invalid address stated on the next line
==18651==    at 0xFFFFFFE900000000: ???
==18651==  Address 0xFFFFFFE900000000 is not stack'd, malloc'd or (recently) free'd

I have no idea why there is no call stack even with valgrind, it does
not seem to be a newly created thread either, at least breakpoints on
pthread_create and SDL_CreateThread did not hit anywhere near this
crash...

Greetings,
Reimar D?ffinger