Re: [PATCH] - Improve drain pages performance on large systems

[Minchan Kim]

On Wed, Feb 16, 2011 at 7:38 AM, Jack Steiner <steiner@xxxxxxx> wrote:
>
> Heavy swapping within a cpuset causes frequent calls to drain_all_pages().
> This sends IPIs to all cpus to free PCP pages. In most cases, there are
> no pages to be freed on cpus outside of the swapping cpuset.
>
> Add checks to minimize locking and updates to potentially hot cachelines.
> Before acquiring locks, do a quick check to see if any pages are in the PCP
> queues. Exit if none.
>
> On a 128 node SGI UV system, this reduced the IPI overhead to cpus outside of the
> swapping cpuset by 38% and improved time to run a pass of the swaping test
> from 98 sec to 51 sec. These times are obviously test & configuration
> dependent but the improvements are significant.
>
>
> Signed-off-by: Jack Steiner <steiner@xxxxxxx>
>
> ---
> Âmm/page_alloc.c | Â 14 ++++++++++++++
> Â1 file changed, 14 insertions(+)
>
> Index: linux/mm/page_alloc.c
> ===================================================================
> --- linux.orig/mm/page_alloc.c Â2011-02-15 16:28:36.165921713 -0600
> +++ linux/mm/page_alloc.c    2011-02-15 16:29:43.085502487 -0600
> @@ -592,10 +592,24 @@ static void free_pcppages_bulk(struct zo
> Â Â Â Âint batch_free = 0;
> Â Â Â Âint to_free = count;
>
> + Â Â Â /*
> + Â Â Â Â* Quick scan of zones. If all are empty, there is nothing to do.
> + Â Â Â Â*/
> + Â Â Â for (migratetype = 0; migratetype < MIGRATE_PCPTYPES; migratetype++) {
> + Â Â Â Â Â Â Â struct list_head *list;
> +
> + Â Â Â Â Â Â Â list = &pcp->lists[migratetype];
> + Â Â Â Â Â Â Â if (!list_empty(list))
> + Â Â Â Â Â Â Â Â Â Â Â break;
> + Â Â Â }
> + Â Â Â if (migratetype == MIGRATE_PCPTYPES)
> + Â Â Â Â Â Â Â return;
> +
> Â Â Â Âspin_lock(&zone->lock);
> Â Â Â Âzone->all_unreclaimable = 0;
> Â Â Â Âzone->pages_scanned = 0;
>
> + Â Â Â migratetype = 0;
> Â Â Â Âwhile (to_free) {
> Â Â Â Â Â Â Â Âstruct page *page;
> Â Â Â Â Â Â Â Âstruct list_head *list;
>
> --
> To unsubscribe, send a message with 'unsubscribe linux-mm' in
> the body to majordomo@xxxxxxxxxx ÂFor 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>
>

It does make sense to me.
Although new code looks to be rather costly in small box, anyway we
use the same logic  in while loop so cache would be hot. so cost would
be little.

But how about this? This one never affect fast-critical path.

diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index ff7e158..2dfb61a 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -1095,8 +1095,10 @@ static void drain_pages(unsigned int cpu)
                pset = per_cpu_ptr(zone->pageset, cpu);

                pcp = &pset->pcp;
-               free_pcppages_bulk(zone, pcp->count, pcp);
-               pcp->count = 0;
+               if (pcp->count > 0) {
+                       free_pcppages_bulk(zone, pcp->count, pcp);
+                       pcp->count = 0;
+               }
                local_irq_restore(flags);
        }
 }



-- 
Kind regards,
Minchan Kim
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/