[PATCH v6 07/11] mm, compaction: use correct watermark when checking compaction success

From: Vlastimil Babka
Date: Wed Aug 10 2016 - 16:42:03 EST


The __compact_finished() function uses low watermark in a check that has to
pass if the direct compaction is to finish and allocation should succeed. This
is too pessimistic, as the allocation will typically use min watermark. It may
happen that during compaction, we drop below the low watermark (due to parallel
activity), but still form the target high-order page. By checking against low
watermark, we might needlessly continue compaction.

Similarly, __compaction_suitable() uses low watermark in a check whether
allocation can succeed without compaction. Again, this is unnecessarily
pessimistic.

After this patch, these check will use direct compactor's alloc_flags to
determine the watermark, which is effectively the min watermark.

Signed-off-by: Vlastimil Babka <vbabka@xxxxxxx>
Acked-by: Michal Hocko <mhocko@xxxxxxxx>
---
mm/compaction.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/mm/compaction.c b/mm/compaction.c
index ae4f40afcca1..9bd788886b1a 100644
--- a/mm/compaction.c
+++ b/mm/compaction.c
@@ -1316,7 +1316,7 @@ static enum compact_result __compact_finished(struct zone *zone, struct compact_
return COMPACT_CONTINUE;

/* Compaction run is not finished if the watermark is not met */
- watermark = low_wmark_pages(zone);
+ watermark = zone->watermark[cc->alloc_flags & ALLOC_WMARK_MASK];

if (!zone_watermark_ok(zone, cc->order, watermark, cc->classzone_idx,
cc->alloc_flags))
@@ -1381,7 +1381,7 @@ static enum compact_result __compaction_suitable(struct zone *zone, int order,
if (is_via_compact_memory(order))
return COMPACT_CONTINUE;

- watermark = low_wmark_pages(zone);
+ watermark = zone->watermark[alloc_flags & ALLOC_WMARK_MASK];
/*
* If watermarks for high-order allocation are already met, there
* should be no need for compaction at all.
@@ -1395,7 +1395,7 @@ static enum compact_result __compaction_suitable(struct zone *zone, int order,
* This is because during migration, copies of pages need to be
* allocated and for a short time, the footprint is higher
*/
- watermark += (2UL << order);
+ watermark = low_wmark_pages(zone) + (2UL << order);
if (!__zone_watermark_ok(zone, 0, watermark, classzone_idx,
alloc_flags, wmark_target))
return COMPACT_SKIPPED;
--
2.9.2