Bug#648843: ffmpeg: segmentation fault when transcoding mkv to mpg

[Reinhard Tartler]

tag 648843 moreinfo
stop

On Wed, Nov 16, 2011 at 00:26:34 (CET), Eric Cooper wrote:

> On Tue, Nov 15, 2011 at 05:07:19PM +0100, Reinhard Tartler wrote:
>> Can you please install the package libav-dbg and attach a stacktrace for
>> this crash?
>
> $ gdb ffmpeg
> [...]
> Press ctrl-c to stop encoding
> Input stream #0.1 frame changed from rate:48000 fmt:s16 ch:6 to rate:48000 fmt:flt ch:6
> [dvd @ 0x6480e0] buffer underflow i=1 bufi=1017 size=1792rate=4012.2kbits/s dup=4775 drop=0    
>     Last message repeated 3 times051400kB time=2597.80 bitrate=3315.5kbits/s dup=12910 drop=0    
> Input stream #0.1 frame changed from rate:48000 fmt:flt ch:6 to rate:48000 fmt:flt ch:2
> Warning, using s16 intermediate sample format for resampling
>
> Program received signal SIGSEGV, Segmentation fault.
> ac3_extract_exponents_c (exp=0x0, coef=0x6a68c0, nb_coefs=10752)
>     at /build/buildd-libav_0.7.2-1+b1-amd64-gyQUa8/libav-0.7.2/libavcodec/ac3dsp.c:181
> 181 /build/buildd-libav_0.7.2-1+b1-amd64-gyQUa8/libav-0.7.2/libavcodec/ac3dsp.c: No such file or directory.
>     in /build/buildd-libav_0.7.2-1+b1-amd64-gyQUa8/libav-0.7.2/libavcodec/ac3dsp.c
> (gdb) bt
> #0  ac3_extract_exponents_c (exp=0x0, coef=0x6a68c0, nb_coefs=10752)
>     at /build/buildd-libav_0.7.2-1+b1-amd64-gyQUa8/libav-0.7.2/libavcodec/ac3dsp.c:181
> #1  0x00007ffff6a24702 in extract_exponents (s=0x734d20)
>     at /build/buildd-libav_0.7.2-1+b1-amd64-gyQUa8/libav-0.7.2/libavcodec/ac3enc.c:318
> #2  process_exponents (s=0x734d20)
>     at /build/buildd-libav_0.7.2-1+b1-amd64-gyQUa8/libav-0.7.2/libavcodec/ac3enc.c:575

Just a wild guess, can you perhaps try to encode to some other audio
codec than AC3?  If the crash disappears, then we could at least narrow
down the issue.

>> Please try to reproduce the issue with a shortened version of the
>> file.
>
> I haven't been able to do this yet, since I don't know how to shorten
> the MKV file and still have it be well-formed.

Well, without a testcase, I fear there is little we can do about this
issue, then.

Cheers,
Reinhard

-- 
Gruesse/greetings,
Reinhard Tartler, KeyID 945348A4