I have the following stack trace from valgrind. But it does not give me the full stack trace.
==2433== Invalid free() / delete / delete[] / realloc()
==2433== at 0x402B06C: free (in /usr/lib/valgrind/vgpreload_memcheck-x86-linux.so)
==2433== by 0x43F345B: av_freep (mem.c:172)
==2433== by 0x5A6F4D2: (below main) (libc-start.c:226)
In gdb, I think I get the same error:
#5 0xb7c1345c in av_free (ptr=<optimized out>) at libavutil/mem.c:172
#6 av_freep (arg=0x88a2e48) at libavutil/mem.c:181
#7 0xb7c165c8 in av_opt_free (obj=0x88a2ba0) at libavutil/opt.c:787
#8 0xb6b56efc in avcodec_close (avctx=0x88a2ba0) at libavcodec/utils.c:1675
#9 0x0808a3e0 in encode_lavc_finish (ctx=0x8343a40) at encode_lavc.c:288
#10 0x08077d0b in exit_player_with_rc (mpctx=0x8313058, how=EXIT_EOF, rc=0)
at mplayer.c:705
#11 0x0806ced0 in main (argc=8, argv=0xbffff374) at mplayer.c:4771
But with the gdb trace goes all the way up to main()
.
How do I get the full stack trace in valgrind, is at all possible?
In gdb, I think I get the same error
You didn't get an error in GDB. You got a breakpoint on av_free
, but surely av_free
is called a lot, and you've provided no evidence that this particular call is the one that triggers Valgrind error.
It's very likely that some other call to av_free
is in fact triggering the Valgrind error, and it's also very likely that that call is executing from atexit
handler, the Valgrind stack is in fact a full stack trace.
If you are using a recent version of Valgrind, you can actually debug the program running under Valgrind with --vgdb-error=1
, and attach GDB exactly where the problem happens. Documentation here.
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With