[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-ID: <HE1PR06MB4011D3B8C200D13829648D86AC660@HE1PR06MB4011.eurprd06.prod.outlook.com>
Date: Mon, 28 Oct 2019 23:59:20 +0000
From: Jonas Karlman <jonas@...boo.se>
To: Mauro Carvalho Chehab <mchehab+samsung@...nel.org>,
Ezequiel Garcia <ezequiel@...labora.com>,
Maxime Ripard <mripard@...nel.org>,
Paul Kocialkowski <paul.kocialkowski@...tlin.com>
CC: Jonas Karlman <jonas@...boo.se>, Hans Verkuil <hverkuil@...all.nl>,
Jernej Skrabec <jernej.skrabec@...l.net>,
Boris Brezillon <boris.brezillon@...labora.com>,
Tomasz Figa <tfiga@...omium.org>,
Alexandre Courbot <acourbot@...omium.org>,
Nicolas Dufresne <nicolas@...fresne.ca>,
Philipp Zabel <p.zabel@...gutronix.de>,
"linux-media@...r.kernel.org" <linux-media@...r.kernel.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: [PATCH 0/3] media: uapi: h264: clarify expected scaling list order
This is a suggested clarification on the expect order of lists and values
to use in the H264 scaling matrix ctrl, cedrus and hantro currently expect
values in two different orders.
The suggestion is to use the same value order that vaapi, vdpau and nvdec use,
i.e. have the scaling list values in "matrix order"/"raster order",
after the inverse scanning process has been applied.
This also contains a clarification on the order of the scaling lists.
Expected scaling list order will be,
for 4x4: Intra Y, Intra Cb, Intra Cr, Inter Y, Inter Cb, Inter Cr,
for 8x8: Intra Y, Inter Y, Intra Cb, Inter Cb, Intra Cr, Inter Cr.
This series is based on feedback and discussion of [1]
[1] https://patchwork.linuxtv.org/patch/58525/
Jonas Karlman (3):
media: uapi: h264: clarify expected scaling_list_4x4/8x8 order
media: cedrus: Use correct H264 8x8 scaling list
media: hantro: Do not reorder H264 scaling list
.../media/uapi/v4l/ext-ctrls-codec.rst | 8 ++-
drivers/staging/media/hantro/hantro_h264.c | 51 +++++--------------
.../staging/media/sunxi/cedrus/cedrus_h264.c | 4 +-
3 files changed, 20 insertions(+), 43 deletions(-)
--
2.17.1
Powered by blists - more mailing lists