Re: [discuss] [Patch 1/2] x86, x86_64: Intel HT, Multi core detection fixes
From: Siddha, Suresh B
Date: Wed Oct 12 2005 - 17:19:55 EST
On Wed, Oct 12, 2005 at 11:49:04PM +0200, Andi Kleen wrote:
> On Wednesday 12 October 2005 23:36, Siddha, Suresh B wrote:
> > Fields obtained through cpuid vector 0x1(ebx[16:23]) and
> > vector 0x4(eax[14:25], eax[26:31]) indicate the maximum values and might not
> > always be the same as what is available and what OS sees. So make sure
> > "siblings" and "cpu cores" values in /proc/cpuinfo reflect the values as seen
> > by OS instead of what cpuid instruction says. This will also fix the buggy BIOS
> > cases (for example where cpuid on a single core cpu says there are "2" siblings,
> > even when HT is disabled in the BIOS.
> > http://bugzilla.kernel.org/show_bug.cgi?id=4359)
> I'm not too fond of this new booted_core variable. How about
> you just put the true number of cores into x86_num_cores?
> What should x86_num_cores be in your setup anyways if not
> "booted cores"?
x86_num_cores (cpuid returned value) is required while right shifting the
apicid to get the core-id and the package-id. booted_cores will indicate
how many cores actually cameup. x86_num_cores can differ from "booted cores"
in these scenarios
a) x86_num_cores may not be '1' when multi-core is disabled in the BIOS
b) "maxcpus=" kernel boot param
c) logical hotplug
> Also I must admit the number of different variables to keep
> track of multicore and siblingness starts to become mindboggling,
> so I would recommend you add a fat overview comment somewhere
> that describes their definition and relationship. Or better put
> something into Documentation, it is probably as confusing for
> user space /proc/cpuinfo consumer too.
I will send a new documentation patch.
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/