[PATCH 00/20] arch_idle() cleanup and mach/system.h removal

[Jamie Iles]

Hi Nicolas,

On Mon, Dec 19, 2011 at 04:47:29AM -0500, Nicolas Pitre wrote:
> This is the logical continuation of the arch_reset() work that Russell did
> and therefore this must be applied on top of Russell's "reset" branch.
> 
> Most arch_idle() implementations are simply a call to cpu_do_idle() and
> can be eliminated in favor of a common call.  However, some platforms
> require special idle handling and the arm_pm_idle hook is made available
> for those cases.
> 
> Why "arm_pm_idle"?  Because there is already a generic "pm_idle", but its
> calling convention is rather awkward with regard to interrupts, so the
> default "pm_idle" handler will call arm_pm_idle if it is non null, or
> cpu_do_idle otherwise.  The generic pm_idle hook still can be overriden
> as before.
> 
> This series is therefore meant to be functionally a big no-op i.e.
> everything should just behave as before, despite having more than 1000
> lines of code and 55 files removed in the end.

Tested on picoxcell, works great.

Acked-and-tested-by: Jamie Iles <jamie at jamieiles.com>

btw,

> [PATCH 20/20] ARM: big removal of now unused arch_reset()

I guess this should be "ARM: big removal of now unused mach/system.h" or 
similar?

Jamie