On Thu, Aug 08, 2002 at 09:39:25PM +0200, Oliver Kurth wrote: > On Thu, Aug 08, 2002 at 09:13:49PM +0200, Zdenek Kabelac wrote: > > > Anyway could you recompile it with debug flags and don't stript > > the binary - from this backtrace it hard to see where the crash happens. > > I have built the debian package, and I have still the unstripeed version. I > just copied it to /usr/lib. It is currently running, it always needs some time until >it crashes.
Okay, this is the new bt, not much helpful as well: (gdb) bt #0 0x40446976 in malloc () from /lib/libc.so.6 #1 0x40446074 in malloc () from /lib/libc.so.6 #2 0x4011a989 in __builtin_new () from /usr/lib/libaviplay-0.7.so.0 #3 0x4011ab6f in __builtin_vec_new () from /usr/lib/libaviplay-0.7.so.0 #4 0x0804c280 in strcpy () #5 0x0804d9aa in avm::vector<avm::AttributeInfo>::vector () #6 0x402120ba in pthread_start_thread () from /lib/libpthread.so.0 #7 0x40212101 in pthread_start_thread_event () from /lib/libpthread.so.0 My impression is that this is some kind of buffer overflow, because the sequence above does not make much sense. pthread_start_thread does not call vector, I think. But I do not have much exprience in debugging multithreaded apps. I will now recompile without --enable-release. Greetings, Oliver -- Oliver Kurth mailto:[EMAIL PROTECTED] http://leinekanal.de
msg03865/pgp00000.pgp
Description: PGP signature
