From: Christoph Lameter nr_slab is accurate with the zoned VM stats. Remove the comment that states otherwise in swap_prefetch.c Signed-off-by: Christoph Lameter Signed-off-by: Andrew Morton --- mm/swap_prefetch.c | 4 ---- 1 file changed, 4 deletions(-) diff -puN mm/swap_prefetch.c~zoned-vm-stats-nr_slab-is-accurate-fix-comment mm/swap_prefetch.c --- devel/mm/swap_prefetch.c~zoned-vm-stats-nr_slab-is-accurate-fix-comment 2006-06-09 11:52:07.000000000 -0700 +++ devel-akpm/mm/swap_prefetch.c 2006-06-09 11:52:07.000000000 -0700 @@ -386,10 +386,6 @@ static int prefetch_suitable(void) /* * >2/3 of the ram on this node is mapped, slab, swapcache or * dirty, we need to leave some free for pagecache. - * Note that currently nr_slab is innacurate on numa because - * nr_slab is incremented on the node doing the accounting - * even if the slab is being allocated on a remote node. This - * would be expensive to fix and not of great significance. */ limit = global_page_state(NR_MAPPED) + global_page_state(NR_SLAB) + _