[PATCH] pinctrl: document the pinctrl PM states

[Tony Lindgren]

* Linus Walleij <linus.walleij at linaro.org> [130624 05:43]:
> On Mon, Jun 17, 2013 at 8:02 PM, Tony Lindgren <tony at atomide.com> wrote:
> > * Linus Walleij <linus.walleij at linaro.org> [130617 09:11]:
> 
> >> Bear in mind that struct pinctl * forks effects in two paths,
> >> one is muxing the other is config, like pull-ups etc.
> >
> > I also thought the plan was to merge pinmux and pinconf and
> > do things based the named modes?
> 
> That is done from a consumer point of view.
> Consumers only care about pinctrl * handles
> and pinctrl_state * switches.
> 
> > The last time I tried using the pinconf functions it involved
> > knowing the name of the pin in the consumer driver. The name
> > may not be very descriptive in the device tree cases at least
> > for the pinctrl-single. So I did not pay much attention to
> > the pinconf functions.
> 
> Consumers should not use that interface, i.e.:
> 
> int pin_config_get(const char *dev_name, const char *name,
>                                 unsigned long *config);
> int pin_config_set(const char *dev_name, const char *name,
>                                  unsigned long config)
> 
> This needs to be deleted from <linux/pinctrl/consumer.h>
> I'll see if I can get rid of it pronto to avoid any more confusion
> and sorry for leaving that in place for too long.
> 
> The proper way to use it is to use the states.

OK thanks for clarifying that. Yes I think the named states
is a good way to handle the pins in a generic way.

Regards,

Tony