On Thu, 21 Apr 2011, KOSAKI Motohiro wrote:
> > diff --git a/arch/parisc/mm/init.c b/arch/parisc/mm/init.c
> > --- a/arch/parisc/mm/init.c
> > +++ b/arch/parisc/mm/init.c
> > @@ -266,8 +266,10 @@ static void __init setup_bootmem(void)
> > }
> > memset(pfnnid_map, 0xff, sizeof(pfnnid_map));
> >
> > - for (i = 0; i < npmem_ranges; i++)
> > + for (i = 0; i < npmem_ranges; i++) {
> > + node_set_state(i, N_NORMAL_MEMORY);
> > node_set_online(i);
> > + }
> > #endif
>
>
> I'm surprised this one. If arch code doesn't initialized N_NORMAL_MEMORY,
> (or N_HIGH_MEMORY. N_HIGH_MEMORY == N_NORMAL_MEMORY if CONFIG_HIGHMEM=n)
> kswapd is created only at node0. wow.
>
> The initialization must be necessary even if !NUMA, I think. ;-)
> Probably we should have revisit all arch code when commit 9422ffba4a
> (Memoryless nodes: No need for kswapd) was introduced, at least.
>
I think we may want to just convert slub (and the memory controller) to
use N_HIGH_MEMORY rather than N_NORMAL_MEMORY since nothing else uses it
and the generic code seems to handle N_HIGH_MEMORY for all configs
appropriately.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxxx For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>