Re: [PATCH v5 03/13] media: Add NV12_10LE40_4L4 pixel format

From: Nicolas Dufresne
Date: Tue Apr 11 2023 - 12:01:58 EST


As we discussed in person:

Le mardi 11 avril 2023 à 11:07 -0400, Nicolas Dufresne a écrit :
> Le jeudi 30 mars 2023 à 17:40 +0200, Benjamin Gaignard a écrit :
> > NV12_10LE40_4L4 is the 10-bits per component version of
> > NV12_4L4 pixel format.

NV12_10LE40 is a name I made up for GStreamer, but this is originally named NV15
by Rockchip, and was accepted under this name as DRM_FORMAT_NV15 in the graphic
subsystem. So we agreed that for Linux, it will be more consistant to use the
NV15 naming, renaming this one to NV15_4L4, with the plus side that its shorter.
The linear form of this format was already proposed a while back along with the
pending RK HEVC driver. We will include the base format doc in here, and re-
factor it in later branches.

> >
> >
> >
> >
> > Signed-off-by: Benjamin Gaignard <benjamin.gaignard@xxxxxxxxxxxxx>
> > Signed-off-by: Nicolas Dufresne <nicolas.dufresne@xxxxxxxxxxxxx>
> > ---
> > Documentation/userspace-api/media/v4l/pixfmt-yuv-planar.rst | 4 ++++
> > drivers/media/v4l2-core/v4l2-common.c | 1 +
> > drivers/media/v4l2-core/v4l2-ioctl.c | 1 +
> > include/uapi/linux/videodev2.h | 1 +
> > 4 files changed, 7 insertions(+)
> >
> > diff --git a/Documentation/userspace-api/media/v4l/pixfmt-yuv-planar.rst b/Documentation/userspace-api/media/v4l/pixfmt-yuv-planar.rst
> > index f1d5bb7b806d..8805213a75f8 100644
> > --- a/Documentation/userspace-api/media/v4l/pixfmt-yuv-planar.rst
> > +++ b/Documentation/userspace-api/media/v4l/pixfmt-yuv-planar.rst
> > @@ -270,6 +270,7 @@ of the luma plane.
> > .. _V4L2-PIX-FMT-NV12MT:
> > .. _V4L2-PIX-FMT-NV12MT-16X16:
> > .. _V4L2-PIX-FMT-NV12-4L4:
> > +.. _V4L2-PIX-FMT-NV12-10LE40-4L4:
> > .. _V4L2-PIX-FMT-NV12-16L16:
> > .. _V4L2-PIX-FMT-NV12-32L32:
> > .. _V4L2-PIX-FMT-NV12M-8L128:
> > @@ -306,6 +307,9 @@ tiles linearly in memory. The line stride and image height must be
> > aligned to a multiple of 4. The layouts of the luma and chroma planes are
> > identical.
> >
> > +``V4L2_PIX_FMT_NV12_10LE40_4L4`` is similar to ``V4L2_PIX_FMT_NV12_4L4``
> > +and stores 10 bits pixels in 4x4 tiles.
>
> I think this is not enough, as it does not describe the NV12_10E40 pixel
> packing. We don't have that as a linear format yet in V4L2 (unlike
> DRM_FORMAT_NV15 on DRM side). So I'd suggest to drop the usual bit packing table
> here for now. Please ping Sebastien Fricke, as he got that for (along with
> NV12_10LE40 base format) documented in his RKVDEC HEVC decoder work.
>
> >
> > +
> > ``V4L2_PIX_FMT_NV12_16L16`` stores pixels in 16x16 tiles, and stores
> > tiles linearly in memory. The line stride and image height must be
> > aligned to a multiple of 16. The layouts of the luma and chroma planes are
> > diff --git a/drivers/media/v4l2-core/v4l2-common.c b/drivers/media/v4l2-core/v4l2-common.c
> > index 9cb0895dea1c..7f46c45982d6 100644
> > --- a/drivers/media/v4l2-core/v4l2-common.c
> > +++ b/drivers/media/v4l2-core/v4l2-common.c
> > @@ -278,6 +278,7 @@ const struct v4l2_format_info *v4l2_format_info(u32 format)
> >
> > /* Tiled YUV formats */
> > { .format = V4L2_PIX_FMT_NV12_4L4, .pixel_enc = V4L2_PIXEL_ENC_YUV, .mem_planes = 1, .comp_planes = 2, .bpp = { 1, 2, 0, 0 }, .bpp_div = { 1, 1, 1, 1 }, .hdiv = 2, .vdiv = 2 },
> > + { .format = V4L2_PIX_FMT_NV12_10LE40_4L4, .pixel_enc = V4L2_PIXEL_ENC_YUV, .mem_planes = 1, .comp_planes = 2, .bpp = { 5, 10, 0, 0 }, .bpp_div = { 4, 4, 1, 1 }, .hdiv = 2, .vdiv = 2 },
>
> As you will be doing v6, can you initiallized the forgotten "block" parameters ?
>
>
> > { .format = V4L2_PIX_FMT_P010_4L4, .pixel_enc = V4L2_PIXEL_ENC_YUV, .mem_planes = 1, .comp_planes = 2, .bpp = { 2, 4, 0, 0 }, .bpp_div = { 1, 1, 1, 1 }, .hdiv = 2, .vdiv = 2 },
> >
> > /* YUV planar formats, non contiguous variant */
> > diff --git a/drivers/media/v4l2-core/v4l2-ioctl.c b/drivers/media/v4l2-core/v4l2-ioctl.c
> > index d2b905ba5240..5a0c5b3a6b0f 100644
> > --- a/drivers/media/v4l2-core/v4l2-ioctl.c
> > +++ b/drivers/media/v4l2-core/v4l2-ioctl.c
> > @@ -1351,6 +1351,7 @@ static void v4l_fill_fmtdesc(struct v4l2_fmtdesc *fmt)
> > case V4L2_PIX_FMT_NV12_4L4: descr = "Y/UV 4:2:0 (4x4 Linear)"; break;
> > case V4L2_PIX_FMT_NV12_16L16: descr = "Y/UV 4:2:0 (16x16 Linear)"; break;
> > case V4L2_PIX_FMT_NV12_32L32: descr = "Y/UV 4:2:0 (32x32 Linear)"; break;
> > + case V4L2_PIX_FMT_NV12_10LE40_4L4: descr = "10-bit YUV 4:2:0 (4x4 Linear)"; break;
> > case V4L2_PIX_FMT_P010_4L4: descr = "10-bit Y/UV 4:2:0 (4x4 Linear)"; break;
> > case V4L2_PIX_FMT_NV12M: descr = "Y/UV 4:2:0 (N-C)"; break;
> > case V4L2_PIX_FMT_NV21M: descr = "Y/VU 4:2:0 (N-C)"; break;
> > diff --git a/include/uapi/linux/videodev2.h b/include/uapi/linux/videodev2.h
> > index 8cdc7967cdea..93e300111ca1 100644
> > --- a/include/uapi/linux/videodev2.h
> > +++ b/include/uapi/linux/videodev2.h
> > @@ -664,6 +664,7 @@ struct v4l2_pix_format {
> > #define V4L2_PIX_FMT_NV12_4L4 v4l2_fourcc('V', 'T', '1', '2') /* 12 Y/CbCr 4:2:0 4x4 tiles */
> > #define V4L2_PIX_FMT_NV12_16L16 v4l2_fourcc('H', 'M', '1', '2') /* 12 Y/CbCr 4:2:0 16x16 tiles */
> > #define V4L2_PIX_FMT_NV12_32L32 v4l2_fourcc('S', 'T', '1', '2') /* 12 Y/CbCr 4:2:0 32x32 tiles */
> > +#define V4L2_PIX_FMT_NV12_10LE40_4L4 v4l2_fourcc('V', 'T', '1', '5')
> > #define V4L2_PIX_FMT_P010_4L4 v4l2_fourcc('T', '0', '1', '0') /* 12 Y/CbCr 4:2:0 10-bit 4x4 macroblocks */
> > #define V4L2_PIX_FMT_NV12_8L128 v4l2_fourcc('A', 'T', '1', '2') /* Y/CbCr 4:2:0 8x128 tiles */
> > #define V4L2_PIX_FMT_NV12_10BE_8L128 v4l2_fourcc_be('A', 'X', '1', '2') /* Y/CbCr 4:2:0 10-bit 8x128 tiles */
>