lkml.org 
[lkml]   [2012]   [Oct]   [30]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
SubjectRe: [PATCHv2] Input: omap4-keypad: Add pinctrl support
Date
On Tuesday, October 30, 2012 10:51:11 PM Linus Walleij wrote:
> On Tue, Oct 30, 2012 at 7:37 PM, Mark Brown
> <broonie@opensource.wolfsonmicro.com> wrote:
>
> > More seriously the amount of time we seem to have been spending recently
> > on changes which end up requiring us to go through essentially every
> > driver and add code to them (often several times) doesn't seem like
> > we're doing a good job here.
>
> If this is your main concern you should be made aware that there
> are people out there planning to supplant the existing DT probe paths
> that are now being added to each and every ARM-related driver
> with an ACPI probe path as ARM servers come into the picture.

That's correct.

> > pinctrl is really noticable because it's
> > new but it's not the only thing. As a subsystem maintainer this code
> > just makes me want to add new subsystem features to pull the code out of
> > drivers but obviously that's not something that should be being done at
> > the subsystem level.
>
> We did manage to drag the power/voltage domain per se out
> of the AMBA bus, and recommend that people (like us) do that
> business using the power domains.
>
> I think most people (including OMAP) have bought
> into the concept of using the runtime PM framework and power
> domains to control the power domain switches.
>
> It's this wider concept of using the loose concept "PM resource
> domains" to control also clocks and pins that is at stake, and so
> far the runtime PM core people (Rafael and Magnus) has not said
> much so I think we need some kind of indication from them as to
> what is to happen, long-term, with drivers handling their own clocks
> and pins. Should it be centralized or not? If it's to be centralized it
> needs to become a large piece of infrastructure refactoring and
> needs the attention of Linaro and the like to happen.

Well, I personally think it should be centralized somehow. I'm not quite
sure how to achieve that, though.

Thanks,
Rafael


--
I speak only for myself.
Rafael J. Wysocki, Intel Open Source Technology Center.


\
 
 \ /
  Last update: 2012-10-31 00:21    [W:0.133 / U:0.324 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site