mm: vmscan: count only dirty pages as congested
authorJamie Liu <jamieliu@google.com>
Wed, 10 Dec 2014 23:43:20 +0000 (15:43 -0800)
committerLinus Torvalds <torvalds@linux-foundation.org>
Thu, 11 Dec 2014 01:41:06 +0000 (17:41 -0800)
shrink_page_list() counts all pages with a mapping, including clean pages,
toward nr_congested if they're on a write-congested BDI.
shrink_inactive_list() then sets ZONE_CONGESTED if nr_dirty ==
nr_congested.  Fix this apples-to-oranges comparison by only counting
pages for nr_congested if they count for nr_dirty.

Signed-off-by: Jamie Liu <jamieliu@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Greg Thelen <gthelen@google.com>
Cc: Hugh Dickins <hughd@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mm/vmscan.c

index 59605b7c9970727cc184b784b9088d210008caf4..53157e157061b586a8317d6974b590cbc31cdc6c 100644 (file)
@@ -874,7 +874,8 @@ static unsigned long shrink_page_list(struct list_head *page_list,
                 * end of the LRU a second time.
                 */
                mapping = page_mapping(page);
-               if ((mapping && bdi_write_congested(mapping->backing_dev_info)) ||
+               if (((dirty || writeback) && mapping &&
+                    bdi_write_congested(mapping->backing_dev_info)) ||
                    (writeback && PageReclaim(page)))
                        nr_congested++;