[PATCH 05/14] at91: use structure to store the current soc

[Russell King - ARM Linux]

On Tue, May 03, 2011 at 09:27:12AM +1200, Ryan Mallon wrote:
> Russell, what is your opinion on this? Should we use individual
> mach-types for the above boards and have them explicitly specify their
> cpu/soc type, or should we be aiming to have a single mach-type for all
> of them and determine the cpu/soc type in code?

I don't like answering these questions because it requires understanding
the differences between the various individual SoCs to determine what's
possible.

You are correct that CPU type is determined at run-time by the kernel,
and whether its an 920T or 926T CPU is neither here or there to the
kernel.  It's really the SoC type that's the problem, and we don't have
a very good way of specifying that (partly as there is no standard way
to tell what sort of SoC we have.)

So, really, it's a platform specific question.  If there is some way
that the SoC type can be detected from the hardware, that may be an
appropriate way to deal with that issue.  If not, then the mach-type
approach (which I assume is what AT91 currently does) is probably as
good as any other until we have something like DT.