Re: kswapd0: excessive CPU usage

[Mel Gorman <mgorman@suse.de>]

On Fri, Oct 12, 2012 at 02:37:58PM +0200, Jiri Slaby wrote:
> On 10/12/2012 12:08 AM, Jiri Slaby wrote:
> > (It's an effective revert of "mm: vmscan: scale number of pages
> > reclaimed by reclaim/compaction based on failures".)
> 
> Given kswapd had hours of runtime in ps/top output yesterday in the
> morning and after the revert it's now 2 minutes in sum for the last 24h,
> I would say, it's gone.
> 
> Mel, you wrote me it's unlikely the patch, but not impossible in the
> end. Can you take a look, please? If you need some trace-cmd output or
> anything, just let us know.
> 
> This is x86_64, 6G of RAM, no swap. FWIW EXT4, SLUB, COMPACTION all
> enabled/used.
> 

Can you monitor the behaviour of this patch please? Please keep a particular
eye on kswapd activity and the amount of free memory. If free memory is
spiking it might indicate that kswapd is still too aggressive with the loss
of the __GFP_NO_KSWAPD flag. One way to tell is to record /proc/vmstat over
time and see what the pgsteal_* figures look like. If they are climbing
aggressively during what should be normal usage then it might show that
kswapd is still too aggressive when asked to reclaim for THP.

Thanks very much.

---8<---
mm: vmscan: scale number of pages reclaimed by reclaim/compaction only in direct reclaim

Jiri Slaby reported the following:

	(It's an effective revert of "mm: vmscan: scale number of pages
	reclaimed by reclaim/compaction based on failures".)
	Given kswapd had hours of runtime in ps/top output yesterday in the
	morning and after the revert it's now 2 minutes in sum for the last 24h,
	I would say, it's gone.

The intention of the patch in question was to compensate for the loss of
lumpy reclaim. Part of the reason lumpy reclaim worked is because it
aggressively reclaimed pages and this patch was meant to be a
sane compromise.

When compaction fails, it gets deferred and both compaction and
reclaim/compaction is deferred avoid excessive reclaim. However, since
commit c6543459 (mm: remove __GFP_NO_KSWAPD), kswapd is woken up each time
and continues reclaiming which was not taken into account when the patch
was developed.

As it is not taking deferred compaction into account in this path it scans
aggressively before falling out and making the compaction_deferred check in
compaction_ready. This patch avoids kswapd scaling pages for reclaim and
leaves the aggressive reclaim to the process attempting the THP
allocation.

Signed-off-by: Mel Gorman <mgorman@suse.de>
---
 mm/vmscan.c |   10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)

diff --git a/mm/vmscan.c b/mm/vmscan.c
index 2624edc..2b7edfa 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -1763,14 +1763,20 @@ static bool in_reclaim_compaction(struct scan_control *sc)
 #ifdef CONFIG_COMPACTION
 /*
  * If compaction is deferred for sc->order then scale the number of pages
- * reclaimed based on the number of consecutive allocation failures
+ * reclaimed based on the number of consecutive allocation failures. This
+ * scaling only happens for direct reclaim as it is about to attempt
+ * compaction. If compaction fails, future allocations will be deferred
+ * and reclaim avoided. On the other hand, kswapd does not take compaction
+ * deferral into account so if it scaled, it could scan excessively even
+ * though allocations are temporarily not being attempted.
  */
 static unsigned long scale_for_compaction(unsigned long pages_for_compaction,
 			struct lruvec *lruvec, struct scan_control *sc)
 {
 	struct zone *zone = lruvec_zone(lruvec);
 
-	if (zone->compact_order_failed <= sc->order)
+	if (zone->compact_order_failed <= sc->order &&
+	    !current_is_kswapd())
 		pages_for_compaction <<= zone->compact_defer_shift;
 	return pages_for_compaction;
 }

--
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/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>