[FFmpeg-devel] [PATCH 1/2] avformat/wtvdec: Check for EOF before potentially reseting the eof state

[Michael Niedermayer]

On Sun, Dec 13, 2020 at 07:12:29AM +1100, Peter Ross wrote:
> On Sat, Dec 12, 2020 at 12:38:43PM +0100, Michael Niedermayer wrote:
> > Fixes: infinite loop
> > Fixes: 28042/clusterfuzz-testcase-minimized-ffmpeg_dem_WTV_fuzzer-6311288967528448
> > 
> > Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
> > Signed-off-by: Michael Niedermayer <michael at niedermayer.cc>
> > ---
> >  libavformat/wtvdec.c | 3 +++
> >  1 file changed, 3 insertions(+)
> > 
> > diff --git a/libavformat/wtvdec.c b/libavformat/wtvdec.c
> > index 77908e6392..6c41e3c1a3 100644
> > --- a/libavformat/wtvdec.c
> > +++ b/libavformat/wtvdec.c
> > @@ -953,6 +953,9 @@ static int parse_chunks(AVFormatContext *s, int mode, int64_t seekts, int *len_p
> >          } else
> >              av_log(s, AV_LOG_WARNING, "unsupported chunk:"FF_PRI_GUID"\n", FF_ARG_GUID(g));
> >  
> > +        if (avio_feof(pb))
> > +            break;
> > +
> >          avio_skip(pb, WTV_PAD8(len) - consumed);
> >      }
> >      return AVERROR_EOF;
> > -- 
> > 2.17.1
> 
> curious how the infinte loop occurs here, as the while loop already tests avio_feof().
> 
> i suspect it is because the fuzzed case results in 'consumed > len', and therefore avio_skip is seeking backwards.

yes, WTV_PAD8(len) is 88 and consumed is 124 in the testcase, the eof flag
is reset from the backward seek and so the other test doesnt trigger

thx

[...]

-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Those who are best at talking, realize last or never when they are wrong.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 195 bytes
Desc: not available
URL: <https://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20210202/0b06592e/attachment.sig>