public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Alberto Panizzo <maramaopercheseimorto@gmail.com>
To: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Cc: "Mauro Carvalho Chehab" <mchehab@infradead.org>,
	"Hans Verkuil" <hverkuil@xs4all.nl>,
	"Laurent Pinchart" <laurent.pinchart@ideasonboard.com>,
	"Magnus Damm" <damm@opensource.se>,
	"Márton Németh" <nm127@freemail.hu>,
	linux-media@vger.kernel.org,
	linux-kernel <linux-kernel@vger.kernel.org>,
	"Uwe Kleine-König" <u.kleine-koenig@pengutronix.de>
Subject: Re: [PATCH 2/3] mx3_camera: Support correctly the YUV222 and BAYER configurations of CSI
Date: Tue, 30 Nov 2010 15:25:15 +0100	[thread overview]
Message-ID: <1291127115.6537.5.camel@realization> (raw)
In-Reply-To: <1290965045.3016.11.camel@realization>

On dom, 2010-11-28 at 18:24 +0100, Alberto Panizzo wrote:
> This patch is tested and works with the OV2640 camera that output
> YUV422 (UYVY) and RGB565 data.
> 
> The YUV422 format is managed to be converted in IPU internal YUV444 format
> so this stream could be used in the future to feed directly other IPU
> blocks.
> The RGB565 format is managed as GENERIC and can be moved only from CSI
> to memory.
> 
> Signed-off-by: Alberto Panizzo <maramaopercheseimorto@gmail.com>
> ---
> 
> Before applying, please give me feedback if this break in some way other
> pixel formats!
> 
> 

Hi Guennadi,
Have you got time to test this patch?
I think that this will solve the problems that lead to the need of
translating the 10 bits mbus codes to the corresponding 8 bit ones.

Best regards,
Alberto!

>  drivers/media/video/mx3_camera.c |  126 +++++++++++++++++++++++++++++++++-----
>  1 files changed, 110 insertions(+), 16 deletions(-)
> 
> diff --git a/drivers/media/video/mx3_camera.c b/drivers/media/video/mx3_camera.c
> index 29c5fc3..6811d6f 100644
> --- a/drivers/media/video/mx3_camera.c
> +++ b/drivers/media/video/mx3_camera.c
> @@ -55,6 +55,31 @@
>  #define CSI_SENS_CONF_EXT_VSYNC_SHIFT		15
>  #define CSI_SENS_CONF_DIVRATIO_SHIFT		16
>  
> +/*
> + * IPU support the following data formatting (44.1.1.3 Data Flows and Formats):
> + * 1 YUV 4:4:4 or RGB—8 bits per color component
> + * 2 YUV 4:4:4 or RGB—10 bits per color component
> + * 3 Generic data (from sensor to the system memory only)
> + * The formats 1 and 2 are aligned in words of 32 bits, 3 is free and not
> + * recognized by IPU blocks.
> + *
> + * Taking the value of SENS_DATA_FORMAT and DATA_WIDTH, the CSI tries to
> + * align (or rearrange) the sampled data to fit the IPU supported formats
> + * as follows:
> + * - CSI_SENS_CONF_DATA_FMT_RGB_YUV444: It consider the pixel as a sequence of
> + *	3 components of width DATA_WIDTH aligning these to a 32 bit word.
> + *	The CSI output in this case can feed other IPU blocks.
> + * - CSI_SENS_CONF_DATA_FMT_YUV422: It consider the pixel as a sequence of
> + *	2 components of width DATA_WIDTH were the first is the alternating U V
> + *	components and the second is Y. It construct the YUV444 word repeating
> + *	the previous U, V samples aligning the results to a 32 bit word.
> + *	The CSI output in this case can feed other IPU blocks.
> + * - CSI_SENS_CONF_DATA_FMT_BAYER: No rework is performed in this case.
> + *	The sensor data is given as is, considering _every sample_ as a pixel
> + *	data. This format (combined with the GENERIC IPU pixel formats) can
> + *	carry all the other sensor pixel formats to the system memory.
> + *	The CSI output in this case _can not_ feed other IPU blocks.
> + */
>  #define CSI_SENS_CONF_DATA_FMT_RGB_YUV444	(0UL << CSI_SENS_CONF_DATA_FMT_SHIFT)
>  #define CSI_SENS_CONF_DATA_FMT_YUV422		(2UL << CSI_SENS_CONF_DATA_FMT_SHIFT)
>  #define CSI_SENS_CONF_DATA_FMT_BAYER		(3UL << CSI_SENS_CONF_DATA_FMT_SHIFT)
> @@ -323,14 +348,12 @@ static enum pixel_fmt fourcc_to_ipu_pix(__u32 fourcc)
>  {
>  	/* Add more formats as need arises and test possibilities appear... */
>  	switch (fourcc) {
> -	case V4L2_PIX_FMT_RGB565:
> -		return IPU_PIX_FMT_RGB565;
>  	case V4L2_PIX_FMT_RGB24:
>  		return IPU_PIX_FMT_RGB24;
> +	case V4L2_PIX_FMT_UYVY:
> +		return IPU_PIX_FMT_UYVY;
> +	case V4L2_PIX_FMT_RGB565:
>  	case V4L2_PIX_FMT_RGB332:
> -		return IPU_PIX_FMT_RGB332;
> -	case V4L2_PIX_FMT_YUV422P:
> -		return IPU_PIX_FMT_YVU422P;
>  	default:
>  		return IPU_PIX_FMT_GENERIC;
>  	}
> @@ -358,9 +381,25 @@ static void mx3_videobuf_queue(struct videobuf_queue *vq,
>  
>  	/* This is the configuration of one sg-element */
>  	video->out_pixel_fmt	= fourcc_to_ipu_pix(fourcc);
> -	video->out_width	= icd->user_width;
> -	video->out_height	= icd->user_height;
> -	video->out_stride	= icd->user_width;
> +
> +	if (video->out_pixel_fmt == IPU_PIX_FMT_GENERIC) {
> +		/*
> +		 * IPU_PIX_FMT_GENERIC transport bytes, not pixels. So convert
> +		 * video->out_width and stride to the correct unit.
> +		 */
> +		int bytes_per_line = soc_mbus_bytes_per_line(icd->user_width,
> +						icd->current_fmt->host_fmt);
> +		BUG_ON(bytes_per_line <= 0);
> +
> +		video->out_width	= bytes_per_line;
> +		video->out_height	= icd->user_height;
> +		video->out_stride	= bytes_per_line;
> +	} else {
> +		/* For IPU known formats the pixel unit is OK */
> +		video->out_width	= icd->user_width;
> +		video->out_height	= icd->user_height;
> +		video->out_stride	= icd->user_width;
> +	}
>  
>  #ifdef DEBUG
>  	/* helps to see what DMA actually has written */
> @@ -730,18 +769,68 @@ static int mx3_camera_get_formats(struct soc_camera_device *icd, unsigned int id
>  	if (xlate) {
>  		xlate->host_fmt	= fmt;
>  		xlate->code	= code;
> +		dev_dbg(dev, "Providing format %c%c%c%c in pass-through mode\n",
> +			(xlate->host_fmt->fourcc >> (0*8)) & 0xFF,
> +			(xlate->host_fmt->fourcc >> (1*8)) & 0xFF,
> +			(xlate->host_fmt->fourcc >> (2*8)) & 0xFF,
> +			(xlate->host_fmt->fourcc >> (3*8)) & 0xFF);
>  		xlate++;
> -		dev_dbg(dev, "Providing format %x in pass-through mode\n",
> -			xlate->host_fmt->fourcc);
>  	}
>  
>  	return formats;
>  }
>  
> +static int samples_per_pixel(enum v4l2_mbus_pixelcode mcode)
> +{
> +	switch (mcode) {
> +	case V4L2_MBUS_FMT_YUYV8_2X8:
> +	case V4L2_MBUS_FMT_YVYU8_2X8:
> +	case V4L2_MBUS_FMT_VYUY8_2X8:
> +	case V4L2_MBUS_FMT_YVYU10_2X10:
> +	case V4L2_MBUS_FMT_YUYV10_2X10:
> +	case V4L2_MBUS_FMT_RGB444_2X8_PADHI_LE:
> +	case V4L2_MBUS_FMT_RGB444_2X8_PADHI_BE:
> +	case V4L2_MBUS_FMT_RGB555_2X8_PADHI_LE:
> +	case V4L2_MBUS_FMT_RGB555_2X8_PADHI_BE:
> +	case V4L2_MBUS_FMT_RGB565_2X8_LE:
> +	case V4L2_MBUS_FMT_RGB565_2X8_BE:
> +	case V4L2_MBUS_FMT_BGR565_2X8_LE:
> +	case V4L2_MBUS_FMT_BGR565_2X8_BE:
> +		return 2;
> +	case V4L2_MBUS_FMT_SBGGR8_1X8:
> +	case V4L2_MBUS_FMT_SBGGR10_1X10:
> +	case V4L2_MBUS_FMT_GREY8_1X8:
> +	case V4L2_MBUS_FMT_Y10_1X10:
> +	case V4L2_MBUS_FMT_SBGGR10_2X8_PADHI_LE:
> +	case V4L2_MBUS_FMT_SBGGR10_2X8_PADHI_BE:
> +	case V4L2_MBUS_FMT_SGRBG8_1X8:
> +		return 1;
> +	default:
> +		/* Add other pixel codes as needed */
> +		return 0;
> +	}
> +}
> +
>  static void configure_geometry(struct mx3_camera_dev *mx3_cam,
> -			       unsigned int width, unsigned int height)
> +			       unsigned int width, unsigned int height,
> +			       enum v4l2_mbus_pixelcode code)
>  {
>  	u32 ctrl, width_field, height_field;
> +	const struct soc_mbus_pixelfmt *fmt;
> +
> +	fmt = soc_mbus_get_fmtdesc(code);
> +	BUG_ON(!fmt);
> +
> +	if (fourcc_to_ipu_pix(fmt->fourcc) == IPU_PIX_FMT_GENERIC) {
> +		/*
> +		 * As we don't have an IPU native format, the CSI will be
> +		 * configured to output BAYER and here we need to convert
> +		 * geometry unit from pixels to samples.
> +		 * TODO: Support vertical down sampling (YUV420)
> +		 */
> +		width = width * samples_per_pixel(code);
> +		BUG_ON(!width);
> +	}
>  
>  	/* Setup frame size - this cannot be changed on-the-fly... */
>  	width_field = width - 1;
> @@ -850,7 +939,7 @@ static int mx3_camera_set_crop(struct soc_camera_device *icd,
>  				return ret;
>  		}
>  
> -		configure_geometry(mx3_cam, mf.width, mf.height);
> +		configure_geometry(mx3_cam, mf.width, mf.height, mf.code);
>  	}
>  
>  	dev_dbg(icd->dev.parent, "Sensor cropped %dx%d\n",
> @@ -893,7 +982,7 @@ static int mx3_camera_set_fmt(struct soc_camera_device *icd,
>  	 * mxc_v4l2_s_fmt()
>  	 */
>  
> -	configure_geometry(mx3_cam, pix->width, pix->height);
> +	configure_geometry(mx3_cam, pix->width, pix->height, xlate->code);
>  
>  	mf.width	= pix->width;
>  	mf.height	= pix->height;
> @@ -1112,10 +1201,15 @@ static int mx3_camera_set_bus_param(struct soc_camera_device *icd, __u32 pixfmt)
>  		  (3 << CSI_SENS_CONF_DATA_FMT_SHIFT) |
>  		  (3 << CSI_SENS_CONF_DATA_WIDTH_SHIFT));
>  
> -	/* TODO: Support RGB and YUV formats */
> +	/* TODO: Support RGB_YUV444 formats */
>  
> -	/* This has been set in mx3_camera_activate(), but we clear it above */
> -	sens_conf |= CSI_SENS_CONF_DATA_FMT_BAYER;
> +	switch (xlate->code) {
> +	case V4L2_MBUS_FMT_UYVY8_2X8:
> +		sens_conf |= CSI_SENS_CONF_DATA_FMT_YUV422;
> +		break;
> +	default:
> +		sens_conf |= CSI_SENS_CONF_DATA_FMT_BAYER;
> +	}
>  
>  	if (common_flags & SOCAM_PCLK_SAMPLE_FALLING)
>  		sens_conf |= 1 << CSI_SENS_CONF_PIX_CLK_POL_SHIFT;


-- 
Alberto!

        Be Persistent!
                - Greg Kroah-Hartman (FOSDEM 2010)


  parent reply	other threads:[~2010-11-30 14:25 UTC|newest]

Thread overview: 31+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-11-28 17:18 [PATCH 1/3] soc_camera: Add the ability to bind regulators to soc_camedra devices Alberto Panizzo
2010-11-28 17:24 ` [PATCH 2/3] mx3_camera: Support correctly the YUV222 and BAYER configurations of CSI Alberto Panizzo
2010-11-28 17:26   ` [PATCH 3/3] V4L2: Add a v4l2-subdev (soc-camera) driver for OmniVision OV2640 sensor Alberto Panizzo
2010-12-01 23:32     ` Guennadi Liakhovetski
2010-12-02 10:33       ` Alberto Panizzo
2010-12-02 14:53       ` [PATCH v2] " Alberto Panizzo
2010-11-30 14:25   ` Alberto Panizzo [this message]
2010-11-30 14:31     ` [PATCH 2/3] mx3_camera: Support correctly the YUV222 and BAYER configurations of CSI Guennadi Liakhovetski
2010-11-30 14:39       ` Alberto Panizzo
2010-12-01 18:54   ` Guennadi Liakhovetski
2010-12-18 16:24     ` Guennadi Liakhovetski
2010-12-30 19:38       ` Guennadi Liakhovetski
2011-01-03 11:46         ` Alberto Panizzo
2011-01-03 17:33         ` Alberto Panizzo
2011-01-03 19:37           ` Guennadi Liakhovetski
2011-01-03 22:07             ` Alberto Panizzo
2011-01-11 17:29               ` Alberto Panizzo
2011-01-12 11:13               ` [PATCH 0/2] Fix the way mx3_camera manage non 8-bpp pixel formats Alberto Panizzo
2011-01-12 11:16                 ` [PATCH 1/2] soc_mediabus: export a useful method to obtain the number of samples that makes up a pixel format Alberto Panizzo
2011-01-12 11:20                 ` [PATCH 2/2] Fix capture issues for non 8-bit per pixel formats Alberto Panizzo
2011-01-15 21:35                   ` Guennadi Liakhovetski
2011-01-17  9:41                     ` Alberto Panizzo
2011-01-17  9:52                     ` [PATCH 2/2 v2] " Alberto Panizzo
2011-01-03 16:06     ` [PATCH 2/3] mx3_camera: Support correctly the YUV222 and BAYER configurations of CSI Alberto Panizzo
2011-01-03 16:24       ` Guennadi Liakhovetski
2010-11-28 19:05 ` [PATCH 1/3] soc_camera: Add the ability to bind regulators to soc_camedra devices Guennadi Liakhovetski
2010-11-29  9:34   ` Alberto Panizzo
2010-11-29 15:51     ` Mark Brown
2010-11-30 10:45       ` Alberto Panizzo
2010-11-30 11:05         ` Mark Brown
2010-11-29 15:44   ` Mark Brown

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=1291127115.6537.5.camel@realization \
    --to=maramaopercheseimorto@gmail.com \
    --cc=damm@opensource.se \
    --cc=g.liakhovetski@gmx.de \
    --cc=hverkuil@xs4all.nl \
    --cc=laurent.pinchart@ideasonboard.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-media@vger.kernel.org \
    --cc=mchehab@infradead.org \
    --cc=nm127@freemail.hu \
    --cc=u.kleine-koenig@pengutronix.de \
    /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