On Thu, May 23, 2024 at 02:04:15PM +0300, Laurent Pinchart wrote: > On Thu, May 23, 2024 at 10:57:09AM +0000, Sakari Ailus wrote: > > On Thu, May 23, 2024 at 01:57:22PM +0300, Laurent Pinchart wrote: > > > On Thu, May 23, 2024 at 10:46:59AM +0000, Sakari Ailus wrote: > > > > On Thu, May 02, 2024 at 01:04:56PM +0200, Jacopo Mondi wrote: > > > > > The Y10P, Y12P and Y14P format variants are packed according to > > > > > the RAW10, RAW12 and RAW14 formats as defined by the MIPI CSI-2 > > > > > specification. Document it. > > > > > > > > > > Signed-off-by: Jacopo Mondi <jacopo.mondi@xxxxxxxxxxxxxxxx> > > > > > Reviewed-by: Laurent Pinchart <laurent.pinchart@xxxxxxxxxxxxxxxx> > > > > > Reviewed-by: Naushir Patuck <naush@xxxxxxxxxxxxxxx> > > > > > --- > > > > > Documentation/userspace-api/media/v4l/pixfmt-yuv-luma.rst | 4 ++++ > > > > > 1 file changed, 4 insertions(+) > > > > > > > > > > diff --git a/Documentation/userspace-api/media/v4l/pixfmt-yuv-luma.rst b/Documentation/userspace-api/media/v4l/pixfmt-yuv-luma.rst > > > > > index b3c5779521d8..f02e6cf3516a 100644 > > > > > --- a/Documentation/userspace-api/media/v4l/pixfmt-yuv-luma.rst > > > > > +++ b/Documentation/userspace-api/media/v4l/pixfmt-yuv-luma.rst > > > > > @@ -209,3 +209,7 @@ are often referred to as greyscale formats. > > > > > For Y012 and Y12 formats, Y012 places its data in the 12 high bits, with > > > > > padding zeros in the 4 low bits, in contrast to the Y12 format, which has > > > > > its padding located in the most significant bits of the 16 bit word. > > > > > + > > > > > + The 'P' variations of the Y10, Y12 and Y14 formats are packed according to > > > > > + the RAW10, RAW12 and RAW14 packing scheme as defined by the MIPI CSI-2 > > > > > + specification. > > > > > > > > The CSI-2 specification isn't publicly available. I think you'll need to > > > > add similar examples of the packing the packed raw formats have. > > > > > > Would referencing the CSI-2-packed bayer formats be acceptable ? > > > Otherwise we'll essentially duplicate documentation. > > > > I'm fine with that but the packing should be then documented independently > > of formats. I'd prefer referring to component numbers (1 to 4) so this > > could be re-used with Bayer formats. > > > > The documentation from the Bayer formats should be usable for this almost > > as-is. > > Sounds good to me. Oh, and obviously, the best way to address all this would be for the MIPI Alliance to publish an open version of the CSI-2 specification, covering at least the parts that are relevant from a software point of view. -- Regards, Laurent Pinchart
