Re: [PATCH] x86: fix build error when CONFIG_K8_NB=n

From: Ingo Molnar
Date: Thu Apr 02 2009 - 12:45:26 EST



* Randy Dunlap <randy.dunlap@xxxxxxxxxx> wrote:

> From: Randy Dunlap <randy.dunlap@xxxxxxxxxx>
>
> Fix build error when CONFIG_K8_NB=n:
>
> arch/x86/mm/built-in.o:(.bss+0x5c0): multiple definition of `k8_northbridges'
> arch/x86/kernel/built-in.o:(.bss+0x18c8): first defined here

Thanks for the report, this is a known bug.

> Signed-off-by: Randy Dunlap <randy.dunlap@xxxxxxxxxx>
> ---
> arch/x86/include/asm/k8.h | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> --- linux-next-20090402.orig/arch/x86/include/asm/k8.h
> +++ linux-next-20090402/arch/x86/include/asm/k8.h
> @@ -9,7 +9,7 @@ extern int early_is_k8_nb(u32 value);
> #ifdef CONFIG_K8_NB
> extern struct pci_dev **k8_northbridges;
> #else
> -struct pci_dev **k8_northbridges;
> +static struct pci_dev **k8_northbridges;

This might fix the build but the change does something rather
broken: there's now a new 'k8_northbridges' local variable array in
each file that includes k8.h - that's wrong. We dont put static
locals into include files that get included into several .c files,
ever.

The real fix would be to not export k8_northbridges[] from k8.c, but
to introduce a node_to_k8_northbridge() helper that returns a struct
pci_dev. In the !CONFIG_K8_NB case this would be an inlne in k8.h
that always returns NULL.

Anyway - just today i mailed about this this topic to Mark and
Andreas about this build breakage (i have triggered it too), so
there should be a resolution soon, either in form of a fix or in
form of a revert.

Thanks,

Ingo
--
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/