From: Robert Jarzmik <robert.jarzmik@free.fr>
To: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Cc: video4linux-list@redhat.com
Subject: Re: [PATCH 2/2 v3] pxa-camera: pixel format negotiation
Date: Sun, 23 Nov 2008 11:46:57 +0100 [thread overview]
Message-ID: <873ahilnxa.fsf@free.fr> (raw)
In-Reply-To: <Pine.LNX.4.64.0811212256160.8956@axis700.grange> (Guennadi Liakhovetski's message of "Fri\, 21 Nov 2008 23\:16\:33 +0100 \(CET\)")
Guennadi Liakhovetski <g.liakhovetski@gmx.de> writes:
>> I think it is documented in Micron MT9M111 datasheet, table 6, page 14.
>> My understanding is that it has a buswidth=8, and depth=16. But I may be wrong,
>> have a look with your trained eye and tell me please.
>
> I think we shouldn't (and possibly cannot) process this data in
> pass-through mode on pxa270. In raw mode pxa270 expects each pixel to only
> occupy one pixel clock. And we use icd->width and icd->height to configure
> PXA registers in pxa_camera_set_bus_param(). Whereas in this case we would
> have to lie to the PXA and configure it with, for example, the double
> line-width. I think, this way it could work. Then your horizontal sync
> would stay valid. So, I think, we have three options with this format:
>
> 1. Refuse to support this configuration, as PXA doesn't support 2 pixel
> clocks per pixel in raw mode
>
> 2. Extend the API even further to allow for different geometries on the
> sensor and on the controller. This, in fact, will anyway be required once
> we support scaling on host...
>
> 3. Create a special translation entry for this mode and abuse some 16-bit
> preprocessed format, like, e.g., RGB565. I _think_ this would work too,
> because, in the end, PXA doesn't know what colour it should be:-)
Wow, that's clearly stated :)
I like both solutions 2 and 3. In solution 3, on YUV variant would be nicer I
think, because pxa reorders RGB while packing it, whereas VYUY is just passed
through.
I would even prefer a bit solution 3, just slightly, to split true host scaling
from "simulated" host scaling.
Now, it's up to you, make a choice. This would be anyway for the next patch
serie, not this one. I think we should finish that one ASAP (and I'm a bit late)
to stabilize the API.
Cheers.
--
Robert
PS: I forgot ... now I aggree we shouldn't use pass-through mode for 16bit
exotic formats :)
--
video4linux-list mailing list
Unsubscribe mailto:video4linux-list-request@redhat.com?subject=unsubscribe
https://www.redhat.com/mailman/listinfo/video4linux-list
next prev parent reply other threads:[~2008-11-23 10:57 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-11-18 19:25 [PATCH 0/2 v3] soc-camera: pixel format negotiation Guennadi Liakhovetski
2008-11-18 19:25 ` [PATCH 1/2 v3] soc-camera: pixel format negotiation - core support Guennadi Liakhovetski
2008-11-19 20:47 ` Robert Jarzmik
2008-11-20 19:53 ` Guennadi Liakhovetski
2008-11-18 19:25 ` [PATCH 2/2 v3] pxa-camera: pixel format negotiation Guennadi Liakhovetski
2008-11-19 21:29 ` Robert Jarzmik
2008-11-20 20:43 ` Guennadi Liakhovetski
2008-11-21 19:22 ` Robert Jarzmik
2008-11-21 20:03 ` Guennadi Liakhovetski
2008-11-21 20:53 ` Robert Jarzmik
2008-11-21 22:16 ` Guennadi Liakhovetski
2008-11-23 10:46 ` Robert Jarzmik [this message]
2008-11-23 15:26 ` Robert Jarzmik
2008-11-23 16:32 ` Guennadi Liakhovetski
2008-11-24 19:28 ` [PATCH 1/2] soc_camera: add format translation structure Robert Jarzmik
2008-11-24 19:28 ` [PATCH 2/2] pxa_camera: use the new " Robert Jarzmik
2008-11-27 23:00 ` Guennadi Liakhovetski
2008-11-28 23:27 ` Guennadi Liakhovetski
2008-11-29 12:30 ` Robert Jarzmik
2008-11-25 18:21 ` [PATCH 1/2] soc_camera: add format " Guennadi Liakhovetski
2008-11-27 21:27 ` Robert Jarzmik
2008-11-27 23:13 ` Guennadi Liakhovetski
2008-11-29 17:31 ` [PATCH v4 1/2] soc-camera: pixel format negotiation - core support Guennadi Liakhovetski
2008-11-30 17:33 ` Robert Jarzmik
2008-11-29 14:17 ` Robert Jarzmik
2008-11-29 14:17 ` [PATCH v4 2/2] pxa-camera: pixel format negotiation Robert Jarzmik
2008-11-29 15:49 ` Guennadi Liakhovetski
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=873ahilnxa.fsf@free.fr \
--to=robert.jarzmik@free.fr \
--cc=g.liakhovetski@gmx.de \
--cc=video4linux-list@redhat.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