[FFmpeg-devel] [PATCH] several EA demuxer fixes, issue 956

[Reimar Döffinger]

On Fri, Apr 10, 2009 at 08:17:12PM -0700, Mike Melanson wrote:
> Reimar D?ffinger wrote:
> > On Fri, Apr 10, 2009 at 10:57:25PM +0200, Michael Niedermayer wrote:
> >> On Thu, Apr 09, 2009 at 05:45:46PM +0200, Reimar D?ffinger wrote:
> >>> Hello,
> >>> attached patch fixes several issues:
> >>> 1) it returns partial packets, e.g. at the end of badly cut files. This
> >>> I think is the normal behaviour for FFmpeg, if not av_get_packet should
> >>> be changed instead.
> >>> 2) as an unavoidable side effect of 1) this fixes memleaks with partial
> >>> packets since pkt was not free if av_get_packet succeeded and we still
> >>> returned with an error (issue 956)
> >>> 3) since it is simple to do, along with 1) we also return the correct
> >>> error code we got from av_get_packet instead of always AVERROR_IO or
> >>> AVERROR(EIO).
> >>>
> >>> These IMO should be applied as one patch, splitting does not really make
> >>> sense.
> >>> The second patch is just another minor simplification after that.
> >> do we have a maintainer for electronicarts.c ?
> >> (noone is listed in MAINTAINERS)
> >> peter?
> > 
> > I was just going to do it the usual way for those "fringe" formats,
> > maintainer or not: wait a few days and then just change it.
> > This one would change the FATE test results though (only an additional
> > line), so I'd hope that Mike takes note of this so it can be resolved
> > quickly.
> 
> I will certainly take note when a FATE test changes.

Sorry for being needlessly unclear.
What I meant to say: If the ea tests change in a way that just adds an
additional audio and/or video frame at the end but otherwise remains
unchanged, please update the fate test specification directly, that
difference is expected and (almost certainly) correct.