On Fri, 2008-02-01 at 10:14 +0000, Simon Kenyon wrote: > Daniel Kristjansson wrote: > > I've attached a patch which I believe addresses the stop, change input, > > restart procedure within the driver in just a few lines of code. The > > application writer still has the option of putting ivtv specific code > > in their V4L2 handling code, but doesn't have to. If they do put the > > ivtv driver code in their application this patch doesn't change the > > behaviour at all. > > > > My only concern is whether this is atomic, i.e. can ivtv_try_or_set_fmt > > be called by the application via an ioctl on a different file descriptor > > while the input switching ioctl is operating? > > > the problem now is that 2.6.24 is out and therefore this patch will not > (even if accepted) get into the kernel for at least a couple of months.
It's already not a good idea to use a kernel > 2.6.17 anyway, since VBI is broken in later ivtv drivers for PVR-250 and PVR-350 cards. We'll just recommend Linux 2.6.17 or earlier in the MythTV 0.21 release notes. When we release 0.21.1 hopefully both problems will be resolved in a newer kernel. -- Daniel _______________________________________________ ivtv-devel mailing list [email protected] http://ivtvdriver.org/mailman/listinfo/ivtv-devel
