Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Message-ID: <6ace60b7da86e14b089bf2f5e8162288e1223ea5.camel@collabora.com>
Subject: Re: [PATCH 1/2] drm/fourcc: add a 10bits fully packed variant of
 NV12
From:   Nicolas Dufresne <nicolas.dufresne@collabora.com>
Reply-To: Nicolas Dufresne <nicolas.dufresne@collabora.com>
To:     Ville =?ISO-8859-1?Q?Syrj=E4l=E4?= 
        <ville.syrjala@linux.intel.com>, Randy Li <ayaka@soulik.info>
Cc:     dri-devel@lists.freedesktop.org, airlied@linux.ie,
        linux-kernel@vger.kernel.org, linux-rockchip@lists.infradead.org,
        linux-arm-kernel@lists.infradead.org
Date:   Mon, 21 May 2018 12:28:47 -0400
In-Reply-To: <20180521144948.GF23723@intel.com>
References: <20180520171705.29690-1-ayaka@soulik.info>
         <20180520171705.29690-2-ayaka@soulik.info>
         <20180521144948.GF23723@intel.com>
Organization: Collabora
Content-Type: multipart/signed; micalg="pgp-sha1"; protocol="application/pgp-signature";
        boundary="=-krAapsxR1YrYlSkpgxSk"
Mime-Version: 1.0
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk


--=-krAapsxR1YrYlSkpgxSk
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Le lundi 21 mai 2018 =C3=A0 17:49 +0300, Ville Syrj=C3=A4l=C3=A4 a =C3=A9cr=
it :
> On Mon, May 21, 2018 at 01:17:04AM +0800, Randy Li wrote:
> > This pixel format is a fully packed and 10bits variant of NV12.
> > A luma pixel would take 10bits in memory, without any
> > filled bits between pixels in a stride. The color gamut
> > follows the BT.2020 standard.
> >=20
> > Signed-off-by: Randy Li <ayaka@soulik.info>
> > ---
> >  drivers/gpu/drm/drm_fourcc.c  | 1 +
> >  include/uapi/drm/drm_fourcc.h | 3 +++
> >  2 files changed, 4 insertions(+)
> >=20
> > diff --git a/drivers/gpu/drm/drm_fourcc.c
> > b/drivers/gpu/drm/drm_fourcc.c
> > index 5ca6395cd4d3..1f43967c4013 100644
> > --- a/drivers/gpu/drm/drm_fourcc.c
> > +++ b/drivers/gpu/drm/drm_fourcc.c
> > @@ -173,6 +173,7 @@ const struct drm_format_info
> > *__drm_format_info(u32 format)
> >  		{ .format =3D DRM_FORMAT_UYVY,		.depth
> > =3D 0,  .num_planes =3D 1, .cpp =3D { 2, 0, 0 }, .hsub =3D 2, .vsub =3D=
 1 },
> >  		{ .format =3D DRM_FORMAT_VYUY,		.depth
> > =3D 0,  .num_planes =3D 1, .cpp =3D { 2, 0, 0 }, .hsub =3D 2, .vsub =3D=
 1 },
> >  		{ .format =3D DRM_FORMAT_AYUV,		.depth
> > =3D 0,  .num_planes =3D 1, .cpp =3D { 4, 0, 0 }, .hsub =3D 1, .vsub =3D=
 1,
> > .has_alpha =3D true },
> > +		{ .format =3D DRM_FORMAT_NV12_10LE40,	.depth
> > =3D 0,  .num_planes =3D 2, .cpp =3D { 1, 2, 0 }, .hsub =3D 2, .vsub =3D=
 2 },
> >  	};
> > =20
> >  	unsigned int i;
> > diff --git a/include/uapi/drm/drm_fourcc.h
> > b/include/uapi/drm/drm_fourcc.h
> > index e04613d30a13..8eabf01e966f 100644
> > --- a/include/uapi/drm/drm_fourcc.h
> > +++ b/include/uapi/drm/drm_fourcc.h
> > @@ -140,6 +140,9 @@ extern "C" {
> >  #define DRM_FORMAT_NV61		fourcc_code('N', 'V', '6',
> > '1') /* 2x1 subsampled Cb:Cr plane */
> >  #define DRM_FORMAT_NV24		fourcc_code('N', 'V', '2',
> > '4') /* non-subsampled Cr:Cb plane */
> >  #define DRM_FORMAT_NV42		fourcc_code('N', 'V', '4',
> > '2') /* non-subsampled Cb:Cr plane */
> > +/* A fully packed variant of NV12_10LE32 */
>=20
> What does "fully packed" mean? NV12_10LE32 doesn't even exist so
> referring to it makes no sense.

Fully packed means no padding bits at all, that's quite descriptive.
There is generally only one way to achieve this for a given layout and
format. Referring to NV12_10LE32 GStreamer format isn't very useful,
that I agree. I think Xilinx is submitting it as XV10.

In GStreamer, all the 10bit format naming started to be a mess, so I
encoded something, it's probably not great, but does the job. So when
we say NV12, it mean the YUV 4:2:0 with two planes, 10, means 10bit per
component, LE, for littlen endian, and 40 for 40bit packing length. If
you pack 10bit data over 40bit, you have basically 4 component per 5
bytes.

Unlike XV10 (aka NV12_10LE32), where you have 3 component per 4 bytes,
each 32bit have 3 components, and 2bit are ignored (padding).

>=20
> Please try to provide an unambiguous description of new formats like
> we
> have for everything else.
>=20
> > +#define DRM_FORMAT_NV12_10LE40	fourcc_code('R', 'K', '2',
> > '0') /* 2x2 subsampled Cr:Cb plane */
> > +
> > =20
> >  /*
> >   * 3 plane YCbCr
> > --=20
> > 2.14.3
> >=20
> > _______________________________________________
> > dri-devel mailing list
> > dri-devel@lists.freedesktop.org
> > https://lists.freedesktop.org/mailman/listinfo/dri-devel
>=20
>=20
--=-krAapsxR1YrYlSkpgxSk
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: This is a digitally signed message part
Content-Transfer-Encoding: 7bit

-----BEGIN PGP SIGNATURE-----

iF0EABECAB0WIQSScpfJiL+hb5vvd45xUwItrAaoHAUCWwLzvwAKCRBxUwItrAao
HIIyAKCf4quyHu7H3WtcJ5QQrFbz0NsIswCfZ1T1AdidUUiDeiGiAExACdNi+3E=
=gwHn
-----END PGP SIGNATURE-----

--=-krAapsxR1YrYlSkpgxSk--