[FFmpeg-devel] [PATCH] Port VDPAU to AVHWAccel infrastructure

[Reimar Döffinger]

On Mon, Jan 04, 2010 at 12:28:02PM +0100, Stefano Sabatini wrote:
> On date Monday 2010-01-04 11:28:47 +0100, Gwenole Beauchesne encoded:
> [...]
> > diff --git a/libavutil/pixfmt.h b/libavutil/pixfmt.h
> > index ddd17fb..1f17073 100644
> > --- a/libavutil/pixfmt.h
> > +++ b/libavutil/pixfmt.h
> > @@ -124,6 +124,7 @@ enum PixelFormat {
> >      PIX_FMT_YUV444P16LE,  ///< planar YUV 4:4:4, 48bpp, (1 Cr & Cb sample per 1x1 Y samples), little-endian
> >      PIX_FMT_YUV444P16BE,  ///< planar YUV 4:4:4, 48bpp, (1 Cr & Cb sample per 1x1 Y samples), big-endian
> >      PIX_FMT_VDPAU_MPEG4,  ///< MPEG4 HW decoding with VDPAU, data[0] contains a vdpau_render_state struct which contains the bitstream of the slices as well as various fields extracted from headers
> > +    PIX_FMT_VDPAU,     ///< HW decoding with VDPAU, data[3] contains a vdpau_render_state struct which contains the bitstream of the slices as well as various fields extracted from headers
> >      PIX_FMT_NB,        ///< number of pixel formats, DO NOT USE THIS if you want to link with shared libav* because the number of formats might differ between versions
> >  };
> 
> Missing libavutil/pixdesc.c entry addition.
> 
> Also maybe the already existing PIX_FMT_VDPAU_ defines should be
> marked for deletion at the next major bump.

Before that is done: is this new pixfmt scheme actually used anywhere
and proven to work?
Because I only know the MPlayer patch and that uses the codec id to
generate the "old" scheme.
If we get rid of codec-specific pixfmts just so that every single
application actually using them has to add their own hacks to get back
to the old API again that would just be really idiotic.