[FFmpeg-devel] [PATCH 1/2] avcodec/vp9: don't set AVFrame.pkt_dts

[wm4]

On Tue, 2 May 2017 14:50:32 -0300
James Almer <jamrial at gmail.com> wrote:

> On 5/2/2017 2:19 PM, wm4 wrote:
> > On Mon,  1 May 2017 20:52:14 -0300
> > James Almer <jamrial at gmail.com> wrote:
> >   
> >> decode.c already sets it to the default value.
> >>
> >> Signed-off-by: James Almer <jamrial at gmail.com>
> >> ---
> >>  libavcodec/vp9.c | 1 -
> >>  1 file changed, 1 deletion(-)
> >>
> >> diff --git a/libavcodec/vp9.c b/libavcodec/vp9.c
> >> index 4d7310f6d4..3147ffa0db 100644
> >> --- a/libavcodec/vp9.c
> >> +++ b/libavcodec/vp9.c
> >> @@ -1125,7 +1125,6 @@ FF_DISABLE_DEPRECATION_WARNINGS
> >>          ((AVFrame *)frame)->pkt_pts = pkt->pts;
> >>  FF_ENABLE_DEPRECATION_WARNINGS
> >>  #endif
> >> -        ((AVFrame *)frame)->pkt_dts = pkt->dts;
> >>          for (i = 0; i < 8; i++) {
> >>              if (s->next_refs[i].f->buf[0])
> >>                  ff_thread_release_buffer(avctx, &s->next_refs[i]);  
> > 
> > Is there any possibility the "default" value is actually wrong? It
> > assumes a fixed delay etc.  
> 
> The default is set after the AVCodec.decode() call (unless the
> FF_CODEC_CAP_SETS_PKT_DTS cap is set), so the above is overwritten by
> it, and considering it's also pkt->dts, I assume no?
> But if the above line is required, then perhaps the
> FF_CODEC_CAP_SETS_PKT_DTS cap should be set.
> 
> I may be missing something, though.

Yeah, you're right, it's going to be overwritten, and it always was
this way. So the patch can't change behavior.

Just wondering whether originally there was an intention to set this
directly, i.e. whether we should just set FF_CODEC_CAP_SETS_PKT_DTS.