Re: [RFC] pinctrl: generic: Add DT bindings

[Laurent Pinchart <laurent.pinchart@ideasonboard.com>]

Hi Heiko,

On Saturday 15 June 2013 22:16:13 Heiko Stübner wrote:
> Am Samstag, 15. Juni 2013, 21:56:05 schrieb Linus Walleij:
> > On Fri, Jun 14, 2013 at 12:39 AM, Laurent Pinchart wrote:
> > >         mmcif_pins: mmcif {
> > >                 mux {
> > >                         renesas,groups = "mmc0_data8_0", "mmc0_ctrl_0";
> > >                         renesas,function = "mmc0";
> > >                 };
> > >                 cfg {
> > >                         renesas,groups = "mmc0_data8_0";
> > >                         renesas,pins = "PORT279";
> > >                         bias-pull-up = <1>;
> > 
> > If I understood your code correctly that last statement can *optionally*
> > 
> > be written like just:
> >  bias-pull-up;
> > 
> > Without the parameter?
> > 
> > I think that Heiko's implementation does this anyway.
> 
> Yep, with the fixes-series from yesterday the bias-pull-* now have a better
> default value of <1>.
> 
> so you can do
> 	bias-pull-up;
> which is then identical to the
> 	bias-pull-up = <1>;
> above (both are valid of course).

My patch used bias-pull-up = <1>; as the current version of Heiko's DT parser 
didn't support bias-pull-up; correctly at that time. It's fixed now, and my 
prefered is then bias-pull-up;.

> Disable would the be either
> 	bias-disable;
> or
> 	bias-pull-up = <0>;
> 
> A driver should probably handle both, as both are valid pinconf options or
> this.

I feel a bit uneasy about that. Do we really need to support two different 
ways to achieve the same result ?

-- 
Regards,

Laurent Pinchart