Re: [PATCH 01/11] rcar-vin: Refactor controls creation for video device

[Sakari Ailus <sakari.ailus@iki.fi>]

On Thu, Jul 08, 2021 at 04:05:43PM +0200, Niklas Söderlund wrote:
> Hi Sakari,
> 
> On 2021-07-08 16:40:40 +0300, Sakari Ailus wrote:
> > Hi Niklas, Jacopo,
> > 
> > On Tue, Jul 06, 2021 at 06:58:03PM +0200, Jacopo Mondi wrote:
> > > > > > @@ -522,10 +543,8 @@ static void rvin_parallel_subdevice_detach(struct rvin_dev *vin)
> > > > > >  	rvin_v4l2_unregister(vin);
> > > > > >  	vin->parallel.subdev = NULL;
> > > > > >
> > > > > > -	if (!vin->info->use_mc) {
> > > > > > -		v4l2_ctrl_handler_free(&vin->ctrl_handler);
> > > > > > -		vin->vdev.ctrl_handler = NULL;
> > > > > > -	}
> > > > > > +	if (!vin->info->use_mc)
> > > > >
> > > > > I know it was there already, but give that rvin_parallel_notify_unbind()
> > > > > is only registered for parallel, can this happen ?
> > > >
> > > > Yes, on Gen2 where we don't use a media-graph.
> > > >
> > > 
> > > Ah correct, for gen3 the controls are freed elsewhere, right!
> > > 
> > > Thanks for the clarification
> > 
> > I already had the set in my tree but I can throw it out if you'd prefer to
> > send v2 instead. At least I noticed only minor matters in the comments.
> 
> There is one small issue in a cleanup path in 4/11 that should be fixed 
> and it's always good to get the small things fixed. If I have a v2 out 
> before end of day tomorrow could you refresh what you have in your tree?

Certainly. There's no urgency --- we don't have even rc1 yet.

-- 
Sakari Ailus