mm: page_alloc: do not treat a zone that cannot be used for dirty pages as "full"
authorMel Gorman <mgorman@suse.de>
Wed, 4 Jun 2014 23:10:06 +0000 (16:10 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Wed, 4 Jun 2014 23:54:08 +0000 (16:54 -0700)
If a zone cannot be used for a dirty page then it gets marked "full" which
is cached in the zlc and later potentially skipped by allocation requests
that have nothing to do with dirty zones.

Signed-off-by: Mel Gorman <mgorman@suse.de>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Reviewed-by: Rik van Riel <riel@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mm/page_alloc.c

index 8e766241cf566cf6f8a72b94e560f546858ed0f2..b4381eaee71515cb291497c4fa7ce0d2ac8cf138 100644 (file)
@@ -1967,7 +1967,7 @@ zonelist_scan:
                 */
                if ((alloc_flags & ALLOC_WMARK_LOW) &&
                    (gfp_mask & __GFP_WRITE) && !zone_dirty_ok(zone))
-                       goto this_zone_full;
+                       continue;
 
                mark = zone->watermark[alloc_flags & ALLOC_WMARK_MASK];
                if (!zone_watermark_ok(zone, order, mark,