Re: build issue #380 for v2.6.26-rc1-279-g28a4acb : mach-voyager: multiple definition of `phys_cpu_present_map'

[James Bottomley <James.Bottomley@HansenPartnership.com>]

On Sat, 2008-05-10 at 09:01 -0500, James Bottomley wrote:
> On Sat, 2008-05-10 at 15:17 +0200, Ingo Molnar wrote:
> > * Ingo Molnar <mingo@elte.hu> wrote:
> > 
> > > > I already told you that this was incorrect.  This is what I sent as 
> > > > the replacement on 28 April and I didn't hear back from you.
> > > > 
> > > > http://marc.info/?l=linux-kernel&m=120941111400620
> > > 
> > > ok - picked it up for testing - could you please send a signoff line 
> > > as well?
> > 
> > Alexey, the patch from James failed in testing on visws:
> > 
> >  arch/x86/kernel/built-in.o: In function `init_apic_mappings':
> >  : undefined reference to `boot_cpu_physical_apicid'
> >  arch/x86/kernel/built-in.o: In function `init_apic_mappings':
> >  : undefined reference to `boot_cpu_physical_apicid'
> >  arch/x86/kernel/built-in.o: In function `APIC_init_uniprocessor':
> >  : undefined reference to `boot_cpu_physical_apicid'
> >  arch/x86/kernel/built-in.o: In function `APIC_init_uniprocessor':
> > 
> > config is at:
> > 
> >   http://redhat.com/~mingo/misc/config-Sat_May_10_15_05_47_CEST_2008.bad
> 
> OK, so these symbols are required on visws as well.  So that makes the
> gating config symbol X86_LOCAL_APIC.
> 
> Could you run this through the same config to verify; thanks!

By the way, if this patch survives all the randconfig testing, it shows
the symbols don't belong in setup.c; they should be in apic.c (the
beginnings of the unified apic files).  I'll redo the patch that way on
testing confirmation.

Thanks,

James