On Mon, Feb 27, 2012 at 11:55 AM, Russell King - ARM Linux
[off-list ref] wrote:

On Mon, Feb 27, 2012 at 11:38:14AM +0100, Linus Walleij wrote:

quoted

Maybe I'm delusional but I think this could help centralizing the use
of the SP* cells, since they're obviously used in both SPEAr and
versatile express.

They can be used in any PrimeXsys based system, and I don't see that
there's anything which guarantees having both a SP810 and SP804
together.

There are platforms which have SP804 without SP810 - Integrator CP
and bcmring seem to fall into that category.

I was more thinking the other way around - I think that all systems
with an SP810 also have an SP804.

I'm led to thinking this because of these registers:

#define SCCTRL_TIMEREN0SEL_REFCLK       (0 << 15)
#define SCCTRL_TIMEREN0SEL_TIMCLK       (1 << 15)
#define SCCTRL_TIMEREN1SEL_REFCLK       (0 << 17)
#define SCCTRL_TIMEREN1SEL_TIMCLK       (1 << 17)

TIMER0 and TIMER1 seems to indicate TIMER0 and TIMER1
of the SP804 block.

In any case it seems to depend on CONFIG_ARM_TIMER_SP804.

But... the register concerned does a whole load of other things, and
I'm not sure we want to have accesses to it de-centralized.

Sure, just arch/arm/common/sc-810.c works fine as well.

Now I got incentive to finally write that letter to Greg about what
to do with system controllers.

Yours,
Linus Walleij