Re: [PATCH] arm: Use HTPIDR to point to per-CPU state

[Tim Deegan <tim@xen.org>]

At 17:13 +0100 on 30 Mar (1333127612), Ian Campbell wrote:
> > +DECLARE_PER_CPU(unsigned int, cpu_id);
> > +#define get_processor_id()    (this_cpu(cpu_id))
> > +#define set_processor_id(id)  do {                      \
> > +    WRITE_CP32(__per_cpu_offset[cpuid], HTPIDR);        \
> > +    this_cpu(cpu_id) = (id);                            \
> 
> This macro uses id, cpu_id and cpuid all at the same time...
> 
> The one I'm not really sure about is cpuid, where did he come from?
> Looks like it might be a local variable in the two places which call
> set_processor_id, and also happens to be the id argument in one of those
> cases (the other using literal 0).

Yes, that's a silly mistake that just happens to be OK in all callers. 
Yet another reason to use static inlines rather than macros. :)
Should I resubmit with s/cpuid/id/ or can you fix up as you apply?

Tim.