22 Sep
2012
22 Sep
'12
5:23 p.m.
On Fri, Sep 21, 2012 at 05:55:11PM +0300, Peter Ujfalusi wrote:
You're doing the odd word wrapping here too and it's not a commit message...
If we are going to do the pinctrl mux configuration in core we should also consider the machine driver as well. But they tend to request the needed GPIOs in their probe before registering the card. Same can be true for codec drivers where they might need the mux configured before they request the GPIOs they need.
I would really expect that the GPIO drivers would do the pinmuxing in the same way that other drivers do.