Re: [PATCH/RFC v2] ARM: amba: Remove AMBA level regulator support

[Ulf Hansson <ulf.hansson-0IS4wlFg1OjSUeElwK9/Pw@public.gmane.org>]

On 04/13/2012 11:18 AM, Mark Brown wrote:
> On Fri, Apr 13, 2012 at 11:03:56AM +0200, Ulf Hansson wrote:
>
>> But, how should those amba drivers that implements runtime PM
>> support be able to switch of the vcore regulator during normal
>> suspend? In normal suspend case we can not use
>
> A generic AMBA driver should have no idea about the implementation of
> the particular SoC that it's integrated on to.  This applies even more
> to system suspend (where drivers can generally just assume that they
> will loose all power normally) than it does to runtime suspend.
>
>> pm_runtime_put/pm_runtime_put_sync to trigger the power domain
>> runtime functions to switch of vcore. This is kind of more generic
>> problem when dealing with power domains, but as said this patch will
>> have consequences.
>
> The power domain gets callbacks on the system suspend path too and can
> do whatever is sensible there.
>
>> As far as I can see, the power domain must then implement a
>> suspend_noirq function to make sure same things is done as for the
>> runtime_suspend function. Do you agree with this as well or is there
>> another option?
>
> Yes, the power domain should just be handling this transparently.

Alright, thanks for your confirmation. Let's see how this works out then.

Kind regards
Ulf Hansson

------------------------------------------------------------------------------
For Developers, A Lot Can Happen In A Second.
Boundary is the first to Know...and Tell You.
Monitor Your Applications in Ultra-Fine Resolution. Try it FREE!
http://p.sf.net/sfu/Boundary-d2dvs2