Re: [v0 PATCH 1/1] Use dev_name() for AMD8111/AMD8131 EDAC driver

From: Stephen Rothwell
Date: Sun May 10 2009 - 23:57:29 EST


Hi Harry,

On Mon, 11 May 2009 10:47:45 +0800 Harry Ciao <qingtao.cao@xxxxxxxxxxxxx> wrote:
>
> Andrew Morton wrote:
> >
> > The drivers break x86-64 allmodconfig:
> >
> > drivers/edac/amd8111_edac.c: In function 'amd8111_lpc_bridge_init':
> > drivers/edac/amd8111_edac.c:260: error: implicit declaration of function '__do_inb'
> > drivers/edac/amd8111_edac.c:274: error: implicit declaration of function '__do_outb'
>
> Ah, this is because __do_inb so far only supported on PPC arch:
>
> cao@cao-desktop:/work/community/linux-2.6/arch$ grep __do_inb * -r
> powerpc/include/asm/io.h:#define __do_inb(port) _rec_inb(port)
> powerpc/include/asm/io.h:#define __do_inb(port)
> readb((PCI_IO_ADDR)_IO_BASE + port);
> cao@cao-desktop:/work/community/linux-2.6/arch$
>
> So far the only machine that hosts AMD8111 chip is Maple, however,
> AMD8111 is of Hypertransport device and not necessarily depends on Maple
> nor PPC, so its Kconfig option is only depended on EDAC && EDAC_MC_EDAC
> && PCI.
>
> I think I could put some words into its Kconfig help section that
> AMD8111 should be selected on Maple now.
>
> What do you think? thanks!

You need to make them depend on CONFIG_PPC (or use a different i/o
interface) since the automated builds will enable them for
all{mod,yes}config builds. Some explanation of the dependency is, of
course, appropriate.

--
Cheers,
Stephen Rothwell sfr@xxxxxxxxxxxxxxxx
http://www.canb.auug.org.au/~sfr/

Attachment: pgp00000.pgp
Description: PGP signature