mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
To: paul.elder@ideasonboard.com
Cc: linux-media@vger.kernel.org,
	Dafna Hirschfeld <dafna@fastmail.com>,
	Mauro Carvalho Chehab <mchehab@kernel.org>,
	Heiko Stuebner <heiko@sntech.de>,
	linux-rockchip@lists.infradead.org,
	linux-arm-kernel@lists.infradead.org,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH 2/2] media: rkisp1: Add UYVY as an output format
Date: Thu, 28 Jul 2022 17:20:13 +0300	[thread overview]
Message-ID: <YuKbHdiZD5JGE/GY@pendragon.ideasonboard.com> (raw)
In-Reply-To: <20220728125259.GL3984498@pyrite.rasen.tech>

Hi Paul,

On Thu, Jul 28, 2022 at 09:52:59PM +0900, paul.elder@ideasonboard.com wrote:
> On Wed, Jul 20, 2022 at 01:29:57AM +0300, Laurent Pinchart wrote:
> > On Thu, Jul 14, 2022 at 08:26:03PM +0900, Paul Elder wrote:
> > > Add support for UYVY as an output format. The uv_swap bit in the
> > > MI_XTD_FORMAT_CTRL register that is used for the NV formats does not
> > > work for packed YUV formats. Thus, UYVY support is implemented via
> > > byte-swapping. This method clearly does not work for implementing
> > > support for YVYU and VYUY.
> > > 
> > > Signed-off-by: Paul Elder <paul.elder@ideasonboard.com>
> > > 
> > > ---
> > > UYVY for the self path has not been tested because no device supports
> > > it. The rk3399 has a self path, but does not have the
> > > MI_OUTPUT_ALIGN_FORMAT register and thus does not support UYVY. The
> > > i.MX8MP does support UYVY, but does not have a self path.
> > 
> > I'm tempted to drop it then, as the code below isn't correct given that
> > you use the same register for both MP and SP. Let's address MP only for
> > now.
> 
> The same register is used for both MP and SP. They just have different
> bits. Which is why we'd need the read-write cycle, assuming that there
> exists a device that has both an SP and the MI_OUTPUT_ALIGN_FORMAT
> register.

Indeed, I had missed that. The documentation is confusing, the register
is described as "Output align format for main path", has both
mp_byte_swap and sp_byte_swap, but no sp equivalent to mp_lsb_alignment
(maybe the self path doesn't support raw outputs though ?).

I'm OK keeping support for both paths, but I think the
MI_OUTPUT_ALIGN_FORMAT register should then be initialized to a default
value somewhere.

> > > ---
> > >  .../platform/rockchip/rkisp1/rkisp1-capture.c | 40 +++++++++++++++++++
> > >  1 file changed, 40 insertions(+)
> > > 
> > > diff --git a/drivers/media/platform/rockchip/rkisp1/rkisp1-capture.c b/drivers/media/platform/rockchip/rkisp1/rkisp1-capture.c
> > > index 85fd85fe208c..77496ccef7ec 100644
> > > --- a/drivers/media/platform/rockchip/rkisp1/rkisp1-capture.c
> > > +++ b/drivers/media/platform/rockchip/rkisp1/rkisp1-capture.c
> > > @@ -97,6 +97,12 @@ static const struct rkisp1_capture_fmt_cfg rkisp1_mp_fmts[] = {
> > >  		.uv_swap = 0,
> > >  		.write_format = RKISP1_MI_CTRL_MP_WRITE_YUVINT,
> > >  		.mbus = MEDIA_BUS_FMT_YUYV8_2X8,
> > > +	}, {
> > > +		.fourcc = V4L2_PIX_FMT_UYVY,
> > > +		.uv_swap = 0,
> > > +		.yc_swap = 1,
> > > +		.write_format = RKISP1_MI_CTRL_MP_WRITE_YUVINT,
> > > +		.mbus = MEDIA_BUS_FMT_YUYV8_2X8,
> > >  	}, {
> > >  		.fourcc = V4L2_PIX_FMT_YUV422P,
> > >  		.uv_swap = 0,
> > > @@ -231,6 +237,13 @@ static const struct rkisp1_capture_fmt_cfg rkisp1_sp_fmts[] = {
> > >  		.write_format = RKISP1_MI_CTRL_SP_WRITE_INT,
> > >  		.output_format = RKISP1_MI_CTRL_SP_OUTPUT_YUV422,
> > >  		.mbus = MEDIA_BUS_FMT_YUYV8_2X8,
> > > +	}, {
> > > +		.fourcc = V4L2_PIX_FMT_UYVY,
> > > +		.uv_swap = 0,
> > > +		.yc_swap = 1,
> > > +		.write_format = RKISP1_MI_CTRL_SP_WRITE_INT,
> > > +		.output_format = RKISP1_MI_CTRL_SP_OUTPUT_YUV422,
> > > +		.mbus = MEDIA_BUS_FMT_YUYV8_2X8,
> > >  	}, {
> > >  		.fourcc = V4L2_PIX_FMT_YUV422P,
> > >  		.uv_swap = 0,
> > > @@ -464,6 +477,20 @@ static void rkisp1_mp_config(struct rkisp1_capture *cap)
> > >  		rkisp1_write(rkisp1, RKISP1_CIF_MI_XTD_FORMAT_CTRL, reg);
> > >  	}
> > >  
> > > +	/*
> > > +	 * uv swapping with the MI_XTD_FORMAT_CTRL register only works for
> > 
> > s@uv@U/V@
> > 
> > > +	 * NV12/NV21 and NV16/NV61, so instead use byte swap to support UYVY.
> > > +	 * YVYU and VYUY cannot be supported with this method.
> > > +	 */
> > > +	if (rkisp1->info->features & RKISP1_FEATURE_MI_OUTPUT_ALIGN) {
> > > +		reg = rkisp1_read(rkisp1, RKISP1_CIF_MI_OUTPUT_ALIGN_FORMAT);
> > > +		if (cap->pix.cfg->yc_swap)
> > > +			reg |= RKISP1_CIF_OUTPUT_ALIGN_FORMAT_MP_BYTE_SWAP_BYTES;
> > > +		else
> > > +			reg &= ~RKISP1_CIF_OUTPUT_ALIGN_FORMAT_MP_BYTE_SWAP_BYTES;
> > > +		rkisp1_write(rkisp1, RKISP1_CIF_MI_OUTPUT_ALIGN_FORMAT, reg);
> > 
> > As the register is not initialized anywhere, it would be better to write
> > it fully here instead of a read-modify-write cycle. Same comments below.
> > 
> > > +	}
> > > +
> > >  	rkisp1_mi_config_ctrl(cap);
> > >  
> > >  	reg = rkisp1_read(rkisp1, RKISP1_CIF_MI_CTRL);
> > > @@ -505,6 +532,19 @@ static void rkisp1_sp_config(struct rkisp1_capture *cap)
> > >  		rkisp1_write(rkisp1, RKISP1_CIF_MI_XTD_FORMAT_CTRL, reg);
> > >  	}
> > >  
> > > +	/*
> > > +	 * uv swapping with the MI_XTD_FORMAT_CTRL register only works for
> > > +	 * NV12/NV21 and NV16/NV61, so instead use byte swap to support UYVY.
> > > +	 * YVYU and VYUY cannot be supported with this method.
> > > +	 */
> > > +	if (rkisp1->info->features & RKISP1_FEATURE_MI_OUTPUT_ALIGN) {
> > > +		reg = rkisp1_read(rkisp1, RKISP1_CIF_MI_OUTPUT_ALIGN_FORMAT);
> > > +		if (cap->pix.cfg->yc_swap)
> > > +			reg |= RKISP1_CIF_OUTPUT_ALIGN_FORMAT_SP_BYTE_SWAP_BYTES;
> > > +		else
> > > +			reg &= ~RKISP1_CIF_OUTPUT_ALIGN_FORMAT_SP_BYTE_SWAP_BYTES;
> > > +		rkisp1_write(rkisp1, RKISP1_CIF_MI_OUTPUT_ALIGN_FORMAT, reg);
> > > +	}
> > 
> > Missing blank line.
> > 
> > Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
> > 
> > >  	rkisp1_mi_config_ctrl(cap);
> > >  
> > >  	mi_ctrl = rkisp1_read(rkisp1, RKISP1_CIF_MI_CTRL);

-- 
Regards,

Laurent Pinchart

      reply	other threads:[~2022-07-28 14:20 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-07-14 11:26 [PATCH 0/2] media: rkisp1: Add support for UYVY output Paul Elder
2022-07-14 11:26 ` [PATCH 1/2] media: rkisp1: Add YC swap capability Paul Elder
2022-07-19 22:23   ` Laurent Pinchart
2022-07-14 11:26 ` [PATCH 2/2] media: rkisp1: Add UYVY as an output format Paul Elder
2022-07-19 22:29   ` Laurent Pinchart
2022-07-28 12:52     ` paul.elder
2022-07-28 14:20       ` Laurent Pinchart [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=YuKbHdiZD5JGE/GY@pendragon.ideasonboard.com \
    --to=laurent.pinchart@ideasonboard.com \
    --cc=dafna@fastmail.com \
    --cc=heiko@sntech.de \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-media@vger.kernel.org \
    --cc=linux-rockchip@lists.infradead.org \
    --cc=mchehab@kernel.org \
    --cc=paul.elder@ideasonboard.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®