Re: swap storm since kernel 3.2.x

[Johannes Stezenbach <js@sig21.net>]

On Wed, Feb 08, 2012 at 08:34:14PM +0800, Hillf Danton wrote:
> And I want to ask kswapd to do less work, the attached diff is
> based on 3.2.5, mind to test it with CONFIG_DEBUG_OBJECTS enabled?

Sorry, for slow reply.  The patch does not apply to 3.2.4
(3.2.5 only has the ASPM change which I don't want to
try atm).  Is the patch below correct?

I'll let this run for a while and will report back.

Thanks
Johannes


--- mm/vmscan.c.orig	2012-02-03 21:39:51.000000000 +0100
+++ mm/vmscan.c	2012-02-09 12:30:42.000000000 +0100
@@ -2067,8 +2067,11 @@ restart:
 		 * with multiple processes reclaiming pages, the total
 		 * freeing target can get unreasonably large.
 		 */
-		if (nr_reclaimed >= nr_to_reclaim && priority < DEF_PRIORITY)
+		if (nr_reclaimed >= nr_to_reclaim) {
+			nr_to_reclaim = 0;
 			break;
+		}
+		nr_to_reclaim -= nr_reclaimed;
 	}
 	blk_finish_plug(&plug);
 	sc->nr_reclaimed += nr_reclaimed;
@@ -2535,12 +2538,12 @@ static unsigned long balance_pgdat(pg_da
 		 * we want to put equal scanning pressure on each zone.
 		 */
 		.nr_to_reclaim = ULONG_MAX,
-		.order = order,
 		.mem_cgroup = NULL,
 	};
 	struct shrink_control shrink = {
 		.gfp_mask = sc.gfp_mask,
 	};
+	sc.order = order = 0;
 loop_again:
 	total_scanned = 0;
 	sc.nr_reclaimed = 0;

--
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/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>