Re: [Mjpeg-developer] Fw: Driver differences

[Martin S. <sa...@ho...>]

On Sun, 30 Apr 2006 12:54:55 +0200
Bernhard Praschinger <sha...@ut...> wrote:

> > And I just want to get my New and Improved PVR box up and running. Waaah! :)
> Hmm, you might get more help if you are more concrete ;)

Oh, it was not a request for helt, per se.

I do have a working AVS6EYES driver, only it is in the -mm kernel, and I haven't had the opportunity (that's a lie; there's been lots of opportunities, hm, perhaps "the energy and devotion to the cause") to port it over to the CVS tree.

> Is there a way to test the thing ? I meant: how to test it to make it
> soure that you don't use by accident v4l ?

One way, I assume, would be ro replace all includes of video_decoder.h

[root@klon driver-cvs]# grep -r video_decoder.h .
./zoran_driver.c:#include <linux/video_decoder.h>
./saa7111.c:#include <linux/video_decoder.h>
./zoran_device.c:#include <linux/video_decoder.h>
./saa7110.c:#include <linux/video_decoder.h>
./zoran_card.c:#include <linux/video_decoder.h>
./saa7114.c:#include <linux/video_decoder.h>
./zoran.c:#include <linux/video_decoder.h>
./bt819.c:#include <linux/video_decoder.h>
./vpx3220.c:#include <linux/video_decoder.h>

with includes of v4l2-common.h and / or videodev2.h, see the following compile fail miserably, and fix all the errors. Some kind of brute force, or something. I'd expect that to be more than a day's work, but in the end, everything should be V4L2... :)

I could try to give it a stab with the avs6eyes, but it would most likely mean duplicated efforts, as the CVS driver (or the latest kernel driver, for that matter) differ a bit from the -mm driver I'd be working on in that case.

> Tomorrow we have one holiday. Maybe I can give it a try with the latest
> kernel 2.6.16.11.

I don't know what would be best. Besides syncing things up somewhat. :)

/Sam