On Tue, 2009-03-03 at 21:50 -0300, Mauro Carvalho Chehab wrote:
> On Sun, 01 Mar 2009 22:04:55 -0500
> Andy Walls <[email protected]> wrote:
>
> > Mauro,
> >
> > Please pull from
> >
> > http://linuxtv.org/hg/~awalls/cx18
> >
> > for the following
> >
> > cx18: Add interlock so sliced VBI insertion only happens for an MPEG PS
>
>
> > cx18: Fix VPS service register code and ensure VBI consistentcy with ivtv
>
> Argh! This is really ugly!
>
> +
> +#include <linux/ivtv.h> /* For IVTV_SLICED_TYPE_* */
> +
Yes. It is.
> Why do you need to include a header for a device that has nothing to do with
> cx18?
I don't.
My rationale was that the cx18 driver supports a VBI data format type
enumerated in the V4L2 spec and in linux/include/linux/videodev2.h,
namely:
enum v4l2_mpeg_stream_vbi_fmt {
[...]
V4L2_MPEG_STREAM_VBI_FMT_IVTV = 1, /* VBI in private packets, IVTV
format */
};
And some values used in that data format are in no other internal nor
exported header than linux/include/linux/ivtv.h. The data format itself
isn't delared at all in any kernel header AFAIK, but only in
linux/Documentation/video4linux/README.vbi in prose instead of C
declarations.
> This doesn't belong here...
OK.
> If those VBI parameters should be global to all devices, then it should be,
> instead, at some subsystem header, and your patch should also touch on the
> other drivers.
Yes, they should be exported to userspace as well, so apps like MythTV
don't have to keep their own copies as well. I'm going to work on a
V4L2 spec change to add structures and defines for the packets indicated
by V4L2_MPEG_STREAM_VBI_FMT_IVTV, and then add it to some API header.
Maybe in linux/include/linux/videodev2.h with all the other VBI data
formats.
Unless someone really disagrees.
In the mean time I will issue a new PULL request omitting the portion of
the change you don't like.
Regards,
Andy
>
> Cheers,
> Mauro
_______________________________________________
ivtv-devel mailing list
[email protected]
http://ivtvdriver.org/mailman/listinfo/ivtv-devel
