Re: [PATCH v2 1/5] media: uapi: videodev2: Add support for AV1 stateful decoder

[Deepa Guthyappa Madivalara]

On 10/17/2025 1:01 PM, Nicolas Dufresne wrote:
> Le vendredi 17 octobre 2025 à 11:35 -0700, Deepa Guthyappa Madivalara a écrit :
> > Introduce a new pixel format, V4L2_PIX_FMT_AV1, to the
> > Video4Linux2(V4L2) API. This format is intended for AV1
> > bitstreams in stateful decoding/encoding workflows.
> > The fourcc code 'AV10' is used to distinguish
> > this format from the existing V4L2_PIX_FMT_AV1_FRAME,
> > which is used for stateless AV1 decoder implementation.
> >
> > Signed-off-by: Deepa Guthyappa Madivalara <deepa.madivalara@xxxxxxxxxxxxxxxx>
> > ---
> >   Documentation/userspace-api/media/v4l/pixfmt-compressed.rst | 8 ++++++++
> >   include/uapi/linux/videodev2.h                              | 1 +
> >   2 files changed, 9 insertions(+)
> >
> > diff --git a/Documentation/userspace-api/media/v4l/pixfmt-compressed.rst b/Documentation/userspace-api/media/v4l/pixfmt-compressed.rst
> > index 806ed73ac474ce0e6df00f902850db9fd0db240e..d82557a6ac30851b4fd432a5ceab969abaed2e25 100644
> > --- a/Documentation/userspace-api/media/v4l/pixfmt-compressed.rst
> > +++ b/Documentation/userspace-api/media/v4l/pixfmt-compressed.rst
> > @@ -274,6 +274,14 @@ Compressed Formats
> >           of macroblocks to decode a full corresponding frame to the matching
> >           capture buffer.
> >   
> > +    * .. _V4L2-PIX-FMT-AV1:
> > +
> > +      - ``V4L2_PIX_FMT_AV1``
> > +      - 'AV10'
> > +      - AV1 compressed video frame. This format is adapted for implementing AV1
> > +        pipeline. The decoder implements stateful video decoder and expects one
> > +        Temporal Unit per buffer from OBU-stream or AnnexB.
> > +        The encoder generates one Temporal Unit per buffer.
> >   .. raw:: latex
> >   
> >       \normalsize
> > diff --git a/include/uapi/linux/videodev2.h b/include/uapi/linux/videodev2.h
> > index becd08fdbddb857f8f2bf205d2164dc6e20e80b2..e03cfe0a5ae830595f0bf69b951ca16beaedab72 100644
> > --- a/include/uapi/linux/videodev2.h
> > +++ b/include/uapi/linux/videodev2.h
> > @@ -775,6 +775,7 @@ struct v4l2_pix_format {
> >   #define V4L2_PIX_FMT_H264_SLICE v4l2_fourcc('S', '2', '6', '4') /* H264 parsed slices */
> >   #define V4L2_PIX_FMT_HEVC_SLICE v4l2_fourcc('S', '2', '6', '5') /* HEVC parsed slices */
> >   #define V4L2_PIX_FMT_AV1_FRAME v4l2_fourcc('A', 'V', '1', 'F') /* AV1 parsed frame */
> > +#define V4L2_PIX_FMT_AV1      v4l2_fourcc('A', 'V', '1', '0') /* AV1 */
> An alternative fourcc could be to match the one used in the ISO specification,
> which is also what browsers and Android uses. It would be "av01", link to the
> spec:
>
> http://downloads.aomedia.org/assets/pdf/AV1-ISO-Base-Media-File-Format-Binding-Specification.pdf
>
> Though, we never had any requirement in this direction, its just an idea. With
> or without, I'm happy with the definition.

"av01" sounds good. It would keep it consistent with the spec.
I will update it in v3.

> Reviewed-by: Nicolas Dufresne <nicolas.dufresne@xxxxxxxxxxxxx>
>
> cheers,
> Nicolas
>
> >   #define V4L2_PIX_FMT_SPK      v4l2_fourcc('S', 'P', 'K', '0') /* Sorenson Spark */
> >   #define V4L2_PIX_FMT_RV30     v4l2_fourcc('R', 'V', '3', '0') /* RealVideo 8 */
> >   #define V4L2_PIX_FMT_RV40     v4l2_fourcc('R', 'V', '4', '0') /* RealVideo 9 & 10 */