Re: [PATCH v2] gfp.h: GFP_THISNODE can go to other nodes if some are unpopulated

[Nishanth Aravamudan <nacc@us.ibm.com>]

On 11.06.2007 [12:43:32 -0700], Christoph Lameter wrote:
> On Mon, 11 Jun 2007, Nishanth Aravamudan wrote:
> 
> > So, I'm splitting up the populated_map patch in two, so that these bits
> > or the hugetlbfs bits could be put on top of having that nodemask.
> 
> Well maybe just do a single populate_map patch first. We can easily review 
> that and get it in. And it will be useful for multiple other patchsets.

Right, sorry, that's what I meant -- I was moving ahead to the other
patches to make sure everything was sensible.

Will send out the populated_map patch ASAP.

> > *but*, if this change occurs in mempolicy.c, I think we still have a
> > problem, where me->il_next could be initialized in do_set_mempolicy() to
> > a memoryless node:
> 
> I thought that one misalloc would not be that problematic (hmmmm... unless 
> its a hugetlb page on smallist NUMA system...)

Right -- it all depends...

> > 	if (new && new->policy == MPOL_INTERLEAVE)
> > 		current->il_next = first_node(new->v.nodes);
> 
> Hmmmm... We could also switch off the nodes in v.nodes? Then we do not
> need any additional checks and the modifications to interleave() are
> not necessary?

Ah true, so that would happen at mpol_new() time. Makes sense.

Thanks,
Nish

--
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>