Thank you Hans,
Now it is so clear.
Thank you for your clear explanation. I'll be posting a new camera
module (sensor+ISP) driver sooner or later.
Cheers,
Nate
On Mon, Mar 9, 2009 at 4:41 PM, Hans Verkuil <hverkuil@xxxxxxxxx> wrote:
> On Monday 09 March 2009 02:42:03 Dongsoo Nathaniel Kim wrote:
>> Hello Hans,
>>
>> I've been working on camera device based on v4l2-int-device until now,
>> but to follow latest work of yours I decided to make my driver in v4l2
>> subdev before I send patch to the list.
>> So I'm trying to find any passable driver to reference but I haven't
>> found any one yet.
>> Only thing I could find was v4l2-framework.txt for now. Am I missing
>> something? I've not look into every single repository in hg yet, but
>> which one could be the right one I am looking for?
>> Please give me a guideline if there are some further work besides
>> v4l2-framework.txt.
>> Cheers,
>>
>> Nate
>
> Hi Nate,
>
> In the master v4l-dvb (www.linuxtv.org/hg/v4l-dvb) almost all i2c modules
> have been converted. It can however be confusing due to the use of
> backwards compatibility headers v4l2-i2c-drv.h and v4l2-i2c-drv-legacy.h.
> The latter will disappear once all adapter drivers have been converted. The
> first is still needed since the v4l-dvb master also supports kernels <
> 2.6.22.
>
> There has been a lot of discussion about that and it looks like
> v4l2-i2c-drv.h will probably disappear as well in the future as we will
> attempt to do the backwards compatibility in a different way.
>
> Since you don't need such compatibility, you can ignore those headers and
> only concentrate on the subdev changes.
>
> Good examples are wm8739.c and saa7115.c. Note that there are no sensor
> drivers converted as these are currently all soc-camera or v4l2-int based.
> I'm sure that you will need to add new ops to v4l2-subdev.h for sensor
> drivers.
>
> The v4l-dvb master also contains several PCI and USB drivers that are
> converted to use v4l2_device/v4l2_subdev. Probably the best to look at is
> saa7134. Also interesting is cx18 since that models the analog digitizer
> part of the cx23418 device as a subdev, thus demonstrating that sub-devices
> do not have to be actual i2c devices.
>
> It would be great if at least one sensor driver could be implemented in
> v4l2_subdev before the 2.6.30 merge window opens. Any changes that need to
> be made to v4l2_subdev to support such drivers will then be in the 2.6.30
> tree which will make it easier for you to track compared to v4l-dvb.
>
> Regards,
>
> Hans
>
> --
> Hans Verkuil - video4linux developer - sponsored by TANDBERG
>
--
========================================================
DongSoo, Nathaniel Kim
Engineer
Mobile S/W Platform Lab.
Digital Media & Communications R&D Centre
Samsung Electronics CO., LTD.
e-mail : dongsoo.kim@xxxxxxxxx
dongsoo45.kim@xxxxxxxxxxx
========================================================
--
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
