On Fri, Apr 07, 2017 at 01:47:11PM +0300, Laurent Pinchart wrote:
> Hi Sakari,
>
> On Friday 07 Apr 2017 13:45:09 Sakari Ailus wrote:
> > On Fri, Apr 07, 2017 at 01:04:47PM +0300, Laurent Pinchart wrote:
> > > > @@ -58,6 +60,9 @@ struct v4l2_async_subdev {
> > > > const struct device_node *node;
> > > > } of;
> > > > struct {
> > > > + struct fwnode_handle *fwn;
> > >
> > > Shouldn't this be const ?
> >
> > I thought the same, but a lot of functions that operate on fwnode_handle
> > take a non-const argument. I attempted changing that, but it starts a
> > cascade of unavoidable changes elsewhere. That's not very well suitable for
> > this patchset.
>
> fwnode is young, we should try to fix it instead of propagating issues :-)
I'm not arguing this would be how I'd prefer things to be. Adding a
dependency to another kernel framework would mean postponing the set. This
can well be fixed later on as well.
--
Sakari Ailus
e-mail: sakari.ailus@xxxxxx XMPP: sailus@xxxxxxxxxxxxxx
