Re: [PATCH] x66, UV: Enable 64-bit ACPI MFCG support for SGI UV2 platform

[Jack Steiner <steiner@sgi.com>]

On Fri, Jun 03, 2011 at 10:28:45AM +0200, Ingo Molnar wrote:
> 
> * Jack Steiner <steiner@sgi.com> wrote:
> 
> > Enable 64-bit ACPI MFCG support for SGI UV2 platform. The check is similar
> > to the check on UV1. UV2 has a different oem_id string.
> > 
> > Signed-off-by: Jack Steiner <steiner@sgi.com>
> > 
> > 
> > ---
> >  arch/x86/pci/mmconfig-shared.c |    3 ++-
> >  1 file changed, 2 insertions(+), 1 deletion(-)
> > 
> > Index: linux/arch/x86/pci/mmconfig-shared.c
> > ===================================================================
> > --- linux.orig/arch/x86/pci/mmconfig-shared.c	2011-05-31 12:13:19.635285219 -0500
> > +++ linux/arch/x86/pci/mmconfig-shared.c	2011-06-02 14:41:26.645034257 -0500
> > @@ -519,7 +519,8 @@ static int __init acpi_mcfg_check_entry(
> >  	if (cfg->address < 0xFFFFFFFF)
> >  		return 0;
> >  
> > -	if (!strcmp(mcfg->header.oem_id, "SGI"))
> > +	if (!strcmp(mcfg->header.oem_id, "SGI") ||
> > +	    		!strcmp(mcfg->header.oem_id, "SGI2"))
> 
> Wouldnt strncmp(,,3) cover any future SGI3, SGI4 strings as well?

I considered that but was not certain whether I could safely claim all
oemid's starting with "SGI".

There is a weak precedent in x2apic_uv_x.c - only SGI & SGI2 will be
recognized as an SGI platform. An id such as SGIABC will be treated as
a non SGI platform.

--- jack