Re: [PATCH 2/6] x86: ifdef enable_IR_x2apic() out

From: Gleb Natapov
Date: Tue Feb 15 2011 - 09:16:21 EST


On Tue, Feb 15, 2011 at 02:47:50PM +0100, Henrik Kretzschmar wrote:
> Am 14.02.2011 12:03, schrieb Ingo Molnar:
> >
> >
> > * Henrik Kretzschmar <henne@xxxxxxxxxxxxxxxx> wrote:
> >
> >> +#ifdef CONFIG_X86_64
> >> extern void enable_IR_x2apic(void);
> >> +#endif
> >
> > Cannot we use the CONFIG_X86_X2APIC Kconfig switch here, instead of CONFIG_X86_64?
> >
> > enable_IR_x2apic() is not a 64-bit CPU feature.
> >
> > Thanks,
> >
> > Ingo
> >
> >
>
> Thats what I had liked also.
>
> At the time you moved the apic code from x86/kernel it was exactly that way.
> But after that, in commit 937582382c71b75b29fbb92615629494e1a05ac0, it was explicitely moved out of
> CONFIG_X86_X2APIC, which made that function also compile on 32bit machines, even if its not used there.
>
> Also commit ce69a784504222c3ab6f1b3c357d09ec5772127a enabled the x2apic without interrupt remapping,
> which means that CONFIG_X86_X2APIC depends on CONFIG_INTR_REMAP has been bypassed and so Kconfig is
> broken/not representing what happens to the code.
>
Intel asked to leave Kconfig dependency in place.

> So since I dont have an x2apic and therefore cannot test changes,
> I decided for CONFIG_X86_64 without touching code which seems to work.
>
> I wish I could offer something better here.
>
> Greets,
> Henrik Kretzschmar
>
>
>

--
Gleb.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/