Commit 83e33a47 authored by Christoph Lameter's avatar Christoph Lameter Committed by Linus Torvalds

[PATCH] zone reclaim with slab: avoid unecessary off node allocations

Minor performance fix.

If we reclaimed enough slab pages from a zone then we can avoid going off
node with the current allocation.  Take care of updating nr_reclaimed when
reclaiming from the slab.
Signed-off-by: default avatarChristoph Lameter <clameter@sgi.com>
Signed-off-by: default avatarAndrew Morton <akpm@osdl.org>
Signed-off-by: default avatarLinus Torvalds <torvalds@osdl.org>
parent 0ff38490
......@@ -1566,6 +1566,7 @@ static int __zone_reclaim(struct zone *zone, gfp_t gfp_mask, unsigned int order)
.gfp_mask = gfp_mask,
.swappiness = vm_swappiness,
};
unsigned long slab_reclaimable;
disable_swap_token();
cond_resched();
......@@ -1592,7 +1593,8 @@ static int __zone_reclaim(struct zone *zone, gfp_t gfp_mask, unsigned int order)
} while (priority >= 0 && nr_reclaimed < nr_pages);
}
if (zone_page_state(zone, NR_SLAB_RECLAIMABLE) > zone->min_slab_pages) {
slab_reclaimable = zone_page_state(zone, NR_SLAB_RECLAIMABLE);
if (slab_reclaimable > zone->min_slab_pages) {
/*
* shrink_slab() does not currently allow us to determine how
* many pages were freed in this zone. So we take the current
......@@ -1603,12 +1605,17 @@ static int __zone_reclaim(struct zone *zone, gfp_t gfp_mask, unsigned int order)
* Note that shrink_slab will free memory on all zones and may
* take a long time.
*/
unsigned long limit = zone_page_state(zone,
NR_SLAB_RECLAIMABLE) - nr_pages;
while (shrink_slab(sc.nr_scanned, gfp_mask, order) &&
zone_page_state(zone, NR_SLAB_RECLAIMABLE) > limit)
zone_page_state(zone, NR_SLAB_RECLAIMABLE) >
slab_reclaimable - nr_pages)
;
/*
* Update nr_reclaimed by the number of slab pages we
* reclaimed from this zone.
*/
nr_reclaimed += slab_reclaimable -
zone_page_state(zone, NR_SLAB_RECLAIMABLE);
}
p->reclaim_state = NULL;
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment