[RFC] ARM: OMAP2+: omap_device: add pinctrl handling

[Linus Walleij]

On Thu, Jun 27, 2013 at 4:45 PM, Tony Lindgren <tony at atomide.com> wrote:

> The off mode bits can be enabled continuously, the mux hardware
> automatically sets them. So sounds like you don't need any
> separate "idle" "sleep" and "off" states, the following should
> do:
>
> "default" (or "static") static pins that don't need to be touched
>                         after consumer driver probe

Please use "default" for this. We cannot shoehorn OMAP lingo
into the core kernel and expect everyone to accept that...

> "active"                dynamic pins that are not a subset of
>                         "default" needed for runtime; these pins
>                         are the same as "idle" below, but with
>                         different muxing or pinconf device
>                         runtime
>
> "idle"                  dynamic pins that are not a subset of
>                         "default" needed for various idle modes;
>                         these pins are the same as "active" above,
>                         but with different muxing or pinconf for
>                         various idle states

The misunderstanding earlier was that it was assumed that
"active" == "default", just with another name.

If the above is true with an actual driver patch calling
*both* pinctrl_pm_select_default_state() *and*
pinctrl_pm_select_active_state() in different runpaths I
see no reason to reject it.

But if it turns out that the drivers always use either "default"
or "active" and never both I consider it a pure naming
convention and will not accept the "active" state.

Now it's time for patches :-)

Yours,
Linus Walleij