Re: [PATCH 5/8] sched, numa, mm: Add adaptive NUMA affinity support

[Andi Kleen <andi@firstfloor.org>]

Peter Zijlstra <a.p.zijlstra@chello.nl> writes:

> +
> +		down_write(&mm->mmap_sem);
> +		for (vma = mm->mmap; vma; vma = vma->vm_next) {
> +			if (!vma_migratable(vma))
> +				continue;
> +			change_protection(vma, vma->vm_start, vma->vm_end, vma_prot_none(vma), 0);
> +		}

What happens if I have a 1TB process? Will you really unmap all of the
1TB in that timer?


>  
>  	case MPOL_PREFERRED:
>  		if (pol->flags & MPOL_F_LOCAL)
> -			polnid = numa_node_id();
> +			best_nid = numa_node_id();
>  		else
> -			polnid = pol->v.preferred_node;
> +			best_nid = pol->v.preferred_node;

So that's not the local node anymore?  That will change behaviour for
people using the NUMA affinity APIs explicitely.  I don't think that's a
good idea, if someone set the affinity explicitely the kernel better
follow that.

If you want to change behaviour for non DEFAULT like this
please use a new policy type.

-Andi

-- 
ak@linux.intel.com -- Speaking for myself only

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>