[FFmpeg-cvslog] r12434 - trunk/libavcodec/mpeg12.c

[Ivan Kalvachev]

On Fri, Mar 14, 2008 at 1:19 AM, Baptiste Coudurier
<baptiste.coudurier at smartjog.com> wrote:
>
> Baptiste Coudurier wrote:
>  > Hi,
>  >
>  > Ivan Kalvachev wrote:
>  >> On Fri, Mar 14, 2008 at 12:08 AM, Baptiste Coudurier
>  >> <baptiste.coudurier at smartjog.com> wrote:
>  >>> Hi,
>  >>>
>  >>>
>  >>>  M?ns Rullg?rd wrote:
>  >>>  > bcoudurier <subversion at mplayerhq.hu> writes:
>  >>>  >
>  >>>  >> Author: bcoudurier
>  >>>  >> Date: Thu Mar 13 19:26:53 2008
>  >>>  >> New Revision: 12434
>  >>>  >>
>  >>>  >> Log:
>  >>>  >> do not fail badly if aspect ratio is 0/forbidden, fix issue 369
>  >>>  >>
>  >>>  >> Modified:
>  >>>  >>    trunk/libavcodec/mpeg12.c
>  >>>  >>
>  >>>  >> Modified: trunk/libavcodec/mpeg12.c
>  >>>  >> ==============================================================================
>  >>>  >> --- trunk/libavcodec/mpeg12.c        (original)
>  >>>  >> +++ trunk/libavcodec/mpeg12.c        Thu Mar 13 19:26:53 2008
>  >>>  >> @@ -1970,8 +1970,6 @@ static int mpeg1_decode_sequence(AVCodec
>  >>>  >>          (width % 2) != 0 || (height % 2) != 0)
>  >>>  >>          return -1;
>  >>>  >>      s->aspect_ratio_info= get_bits(&s->gb, 4);
>  >>>  >> -    if (s->aspect_ratio_info == 0)
>  >>>  >> -        return -1;
>  >>>  >
>  >>>  > IMHO a warning should be printed in this case.  Users have right to
>  >>>  > know (or deserve to be annoyed).
>  >>>  >
>  >>>
>  >>>  Yes, what about this ?
>  >> Please make this "broken" behavior to be bound to
>  >> "error_resilience" or "strict_std_compliance" .
>  >
>  > I see no advantage in it, this does not slow code down, and cannot
>  > produce start code emulation since previous and next values are checked.
>  > Im against it.
>  >
>
>  Attached patch should not make it fail in the default case, but will if
>  user wants strict compliance, and log a warning in all cases. Ok for
>  everyone ?

Perfect.