Re: [PATCH x86/mm UPDATED] x86-64, NUMA: Fix distance tablehandling

[Tejun Heo]

Hello,

On Fri, Mar 11, 2011 at 10:02:41AM -0800, Yinghai Lu wrote:
> > No, NUMA implementation can skip numa_set_distance() entirely if the
> > distance is LOCAL_DISTANCE if nids are equal, REMOTE_DISTANCE
> > otherwise.  In fact, any amdtopology configuraiton would behave this
> > way, so it's incorrect to fill the table with LOCAL_DISTANCE.  You
> > have to check the physnid mapping and build new table whether physical
> > table exists or not.  Lack of physical distance table doesn't mean all
> > nodes are LOCAL_DISTANCE.
> 
> too bad. We should call numa_alloc_distance in amdtopology to set
> default value in that array.

I'm not following.  If there's no distance table, the distance is
assumed to be LOCAL between the same node and REMOTE if the nodes are
different, which is exactly the way it should be for those machines.
Why is this bad and why would you allocate distance table for such
configurations?

-- 
tejun
--
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/