[mythtv] Re:[PATCH] Re: ivtv settings patch (revisited)

Isaac Richards ijr at po.cwru.edu
Fri Sep 26 13:21:34 EDT 2003


On Friday 26 September 2003 03:24 am, Simon Kenyon wrote:
> Isaac Richards wrote:
> >On Wednesday 24 September 2003 09:23 am, Geoffrey Hausheer wrote:
> >>Then my suggestion would be 'don't apply the v4l2 patches to the kernel'
> >>(of course this won't help for people using 2.6).  It seems annoying that
> >>we have potentially 3 versions of the videodev2.h file around: myth's,
> >>ivtv's, and the kernel's.  This will not be easy to keep coordinated as
> >>stuff changes.  perhaps we should find a way to detect and use the
> >>kernel's version of videodev2.h before trying to use the
> >>videodev2_myth.h?
> >
> >Doesn't work, as the kernel videodev2.h cannot be used by userspace
> > programs, due to it including other kernel-only headers.
> >
> >Best solution would be for the v4l2 author to revert his completely
> >unnecessary binary-compatibility breaking changes.
>
> in the meantime is there a solution?
>
> using videodev2.h from v4l in:
>
> - a v4l2 patched 2.4.22
> - the decoder branch of the ivtv driver
> - mythtv by replacing videodev2_myth.h with it
>
> results in (after a struggle) a fully working mythtv for me

For people that can follow directions, sure.  Check out the mailing list 
archives from back when I first added support for the ivtv driver for what 
prompted me to add my own copy of the header in the first place.  

I've already said what needs to happen in another email -- the ivtv driver 
needs updated first so it can handle both the new and old ioctls, then I can 
(eventually, say in a month), update mythtv's copy of videodev2.h.

Isaac


More information about the mythtv-dev mailing list