On Fri, Aug 15, 2008 at 19:42, Jeff Squyres <jsquyres_at_[hidden]> wrote:
> On Aug 15, 2008, at 1:35 PM, Bert Wesarg wrote:
>
>>>> More specifically, it looks like different chips have different things
>>>> in
>>>> /sys. Ugh. Is there a way that PLPA can at least determine whether
>>>> it's on
>>>> a sys that it understands?
>>
>> Unfortunatly, it looks like these infos where just added to the kernel
>> sources:
>>
>> 2008-07-28 powerpc: Make core id information available to userspace [1]
>> 2008-07-28 powerpc: Make core sibling information available to userspace
>> [2]
>>
>> But the physical_package_id is still missing. Or, to be more precise:
>> its set to a default value (-1):
>>
>>
>> http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=e7849f16c13476288fe4fbd420975e8456c75aa0
>
>
> Ah. So should we ignore POWER until they properly support both core_id
> (which is done) and physical_package_id?
Have we a choice?
Its better to handle the default values, so that the code don't need
to changed again if powerpc is up-to-date.
This version is out-of-date:
http://www.mjmwired.net/kernel/Documentation/cputopology.txt
But this is the current (i.e. 2.6.27 stuff):
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=Documentation/cputopology.txt;h=bd699da246666a503195ed0d8240fb531b90f746;hb=e7849f16c13476288fe4fbd420975e8456c75aa0
Bert
>
> --
> Jeff Squyres
> Cisco Systems
>
> _______________________________________________
> plpa-users mailing list
> plpa-users_at_[hidden]
> http://www.open-mpi.org/mailman/listinfo.cgi/plpa-users
>
|