[FFmpeg-devel] [PATCH 1/5] avcodec/vp56: Consider the alpha start as end of the prior header

[Michael Niedermayer]

On Sun, Aug 11, 2019 at 09:29:14AM +1000, Peter Ross wrote:
> On Tue, Aug 06, 2019 at 11:30:02PM +0200, Michael Niedermayer wrote:
> > Fixes: Timeout (23sec -> 71ms)
> > Fixes: 15661/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_VP6A_fuzzer-6257865947348992
> > 
> > Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
> > Signed-off-by: Michael Niedermayer <michael at niedermayer.cc>
> > ---
> >  libavcodec/vp56.c | 4 ++--
> >  1 file changed, 2 insertions(+), 2 deletions(-)
> > 
> > diff --git a/libavcodec/vp56.c b/libavcodec/vp56.c
> > index 72fea3780e..695f37e972 100644
> > --- a/libavcodec/vp56.c
> > +++ b/libavcodec/vp56.c
> > @@ -572,7 +572,7 @@ int ff_vp56_decode_frame(AVCodecContext *avctx, void *data, int *got_frame,
> >      VP56Context *s = avctx->priv_data;
> >      AVFrame *const p = s->frames[VP56_FRAME_CURRENT];
> >      int remaining_buf_size = avpkt->size;
> > -    int av_uninit(alpha_offset);
> > +    int alpha_offset = remaining_buf_size;
> >      int i, res;
> >      int ret;
> >  
> > @@ -585,7 +585,7 @@ int ff_vp56_decode_frame(AVCodecContext *avctx, void *data, int *got_frame,
> >              return AVERROR_INVALIDDATA;
> >      }
> >  
> > -    res = s->parse_header(s, buf, remaining_buf_size);
> > +    res = s->parse_header(s, buf, alpha_offset);
> >      if (res < 0)
> >          return res;
> >  
> 
> hi michael, i am strugling to see a problem with the existing logic.
> 
>     if (s->has_alpha) {
>         if (remaining_buf_size < 3)
>             return AVERROR_INVALIDDATA;
>         alpha_offset = bytestream_get_be24(&buf);
>         remaining_buf_size -= 3;
>         if (remaining_buf_size < alpha_offset)
>             return AVERROR_INVALIDDATA;
>     }
> 
>     res = s->parse_header(s, buf, remaining_buf_size);
>     if (res < 0)
>         return res;
> 
> 
> if the sample has alpha, remaining_buf_size is reduced in size.

> can you post the sanple that takes 23s to decode?

ill send the sample to you privatly

thanks

[...]

-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

it is not once nor twice but times without number that the same ideas make
their appearance in the world. -- Aristotle
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 181 bytes
Desc: not available
URL: <http://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20190811/26f497e6/attachment.sig>