Hi Sylwester On Fri, 23 Jul 2010, Sylwester Nawrocki wrote: > Hi Laurent, > > > -----Original Message----- > > From: linux-media-owner@xxxxxxxxxxxxxxx [mailto:linux-media- > > owner@xxxxxxxxxxxxxxx] On Behalf Of Laurent Pinchart > > Sent: Friday, July 23, 2010 10:35 AM > > To: Guennadi Liakhovetski > > Cc: Linux Media Mailing List; Hans Verkuil > > Subject: Re: [PATCH] mediabus: add MIPI CSI-2 pixel format codes > > > > Hi Guennadi, > > > > On Friday 23 July 2010 10:13:37 Guennadi Liakhovetski wrote: > > > Add pixel format codes, defined in the MIPI CSI-2 specification. > > > > > > Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@xxxxxx> > > > --- > > > > > > Even though it affects the same enum as my patch from yesterday, they > > are > > > independent, Hans and Laurent CCed just to avoid possible conflicts, > > when > > > further patching this file. > > > > > > include/media/v4l2-mediabus.h | 26 ++++++++++++++++++++++++++ > > > 1 files changed, 26 insertions(+), 0 deletions(-) > > > > > > diff --git a/include/media/v4l2-mediabus.h b/include/media/v4l2- > > mediabus.h > > > index a870965..b0dcace 100644 > > > --- a/include/media/v4l2-mediabus.h > > > +++ b/include/media/v4l2-mediabus.h > > > @@ -41,6 +41,32 @@ enum v4l2_mbus_pixelcode { > > > V4L2_MBUS_FMT_SBGGR10_2X8_PADHI_BE, > > > V4L2_MBUS_FMT_SBGGR10_2X8_PADLO_BE, > > > V4L2_MBUS_FMT_SGRBG8_1X8, > > > + /* MIPI CSI-2 codes */ > > > + V4L2_MBUS_FMT_MIPI_CSI2_YUV420_8_L, > > > + V4L2_MBUS_FMT_MIPI_CSI2_YUV420_8, > > > + V4L2_MBUS_FMT_MIPI_CSI2_YUV420_10, > > > + V4L2_MBUS_FMT_MIPI_CSI2_YUV420_8_CSPS, > > > + V4L2_MBUS_FMT_MIPI_CSI2_YUV420_10_CSPS, > > > + V4L2_MBUS_FMT_MIPI_CSI2_YUV422_8, > > > + V4L2_MBUS_FMT_MIPI_CSI2_YUV422_10, > > > + V4L2_MBUS_FMT_MIPI_CSI2_RGB888, > > > + V4L2_MBUS_FMT_MIPI_CSI2_RGB666, > > > + V4L2_MBUS_FMT_MIPI_CSI2_RGB565, > > > + V4L2_MBUS_FMT_MIPI_CSI2_RGB555, > > > + V4L2_MBUS_FMT_MIPI_CSI2_RGB444, > > > + V4L2_MBUS_FMT_MIPI_CSI2_RAW6, > > > + V4L2_MBUS_FMT_MIPI_CSI2_RAW7, > > > + V4L2_MBUS_FMT_MIPI_CSI2_RAW8, > > > + V4L2_MBUS_FMT_MIPI_CSI2_RAW10, > > > + V4L2_MBUS_FMT_MIPI_CSI2_RAW12, > > > + V4L2_MBUS_FMT_MIPI_CSI2_RAW14, > > > + V4L2_MBUS_FMT_MIPI_CSI2_GEN_NULL, > > > + V4L2_MBUS_FMT_MIPI_CSI2_GEN_BLANKING, > > > + V4L2_MBUS_FMT_MIPI_CSI2_GEN_EMBEDDED8, > > > + V4L2_MBUS_FMT_MIPI_CSI2_USER_1, > > > + V4L2_MBUS_FMT_MIPI_CSI2_USER_2, > > > + V4L2_MBUS_FMT_MIPI_CSI2_USER_3, > > > + V4L2_MBUS_FMT_MIPI_CSI2_USER_4, > > > > I don't think I like this. Take the raw formats for instance, they're > > used for > > Bayer RGB. V4L2_MBUS_FMT_MIPI_CSI2_RAW8 could map to any Bayer format > > (GRBG, > > RGGB, ...). Why don't we just use "standard" pixel codes ? > > As far as I understand on some media buses exact pixel formats are not > defined, > although we still need information to configure the bus. > MIPI CSI-2 seem an example of such to me, e.g. we do configure MIPI > interface > to "*_USER_1" format but over the bus is transferred JPEG data. > I guess we could try to use "standard" pixel codes but then we would > probably have > to map from "any" format to specific MIPI format code to configure the > hardware. > Moreover MIPI formats are quite specific, for instance for RAW12 in 32-bit > sample > (from MSb to LSb) we have dummy 8-bits, then 12-bit of actual data and > remaining > 12 dummy bits. I think, we have to approach this problem from the other side - from the user perspective. A "RAW8" format tells nothing to mplayer or gstreamer, whereas they shall understand 8-bit Bayer. Similarly, the sensor will know, that for sending of 8-bit Bayer data it'll use RAW8. So, on the receiver side, as you correctly point out, you will have to configure which data format to receive. If 8-bit Bayer is sent by the sensor, you will guess, that it should be RAW8. If RGB565 is expected you'll know what to set too. The problem arises with USER? formats. Only the sensor will know, that when requested JPEG, it will user USER1, for MPEG-4 it will use USER2. And there's currently no way to know this in the bridge / host driver... So, at latest, when we get such a sensor, we'll have to decide how to map those. Until then I would just propose to continue using the existing mediabus formats and hope, that their mapping to CSI formats is of a many-to-one nature... Thanks Guennadi --- Guennadi Liakhovetski, Ph.D. Freelance Open-Source Software Developer http://www.open-technology.de/ -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
