Re: [PATCH 5/5] Move definition of hard_smp_processor_id to asm/smp.h - alpha, m32r, powerpc, s390, sparc, sparc64, um

[Andrew Morton <akpm@linux-foundation.org>]

On Fri, 02 Mar 2007 17:31:21 +0900 Fernando Luis Vázquez Cao <fernando@oss.ntt.co.jp> wrote:

> On Thu, 2007-03-01 at 10:03 +0100, Heiko Carstens wrote:
> > On Thu, Mar 01, 2007 at 04:18:23PM +0900, Fernando Luis Vázquez Cao wrote:
> > > Move definition of hard_smp_processor_id to asm/smp.h on alpha, m32r,
> > > powerpc, s390, sparc, sparc64, and um architectures.
> > > 
> > > Signed-off-by: Fernando Luis Vazquez Cao <fernando@oss.ntt.co.jp>
> > > ---
> > > 
> > > diff -urNp linux-2.6.21-rc2/include/asm-alpha/smp.h linux-2.6.21-rc2-hwcpuid/include/asm-alpha/smp.h
> > > --- linux-2.6.21-rc2/include/asm-alpha/smp.h	2007-02-05 03:44:54.000000000 +0900
> > > +++ linux-2.6.21-rc2-hwcpuid/include/asm-alpha/smp.h	2007-03-07 13:34:14.000000000 +0900
> > > @@ -51,6 +51,7 @@ int smp_call_function_on_cpu(void (*func
> > > 
> > >  #else /* CONFIG_SMP */
> > > 
> > > +#define hard_smp_processor_id()		0
> > >  #define smp_call_function_on_cpu(func,info,retry,wait,cpu)    ({ 0; })
> > > 
> > >  #endif /* CONFIG_SMP */
> > 
> > /me thinks you should merge this patch with patch 1. Otherwise there is no
> > hard_smp_processor_id() defined on any architecture if only patch 1 is
> > applied. That would break a git bisect search.
> Hi Heiko,
> 
> Thank you for reviewing the patch-set. The patches have already been
> merged in Andrew's tree but I will take your advice into account when
> submitting new patches. I certainly would not like to break bisect
> searches.
> 

I swizzled your patches around so the bisect problem should be gone.

At least, the one which Heiko identified - there might be others.