[PATCH 22/34] mm, page_alloc: wake kswapd based on the highest eligible zone

From	Mel Gorman <mgorman@techsingularity.net>
Newsgroups	linux.kernel
Subject	[PATCH 22/34] mm, page_alloc: wake kswapd based on the highest eligible zone
Date	2016-07-08 11:50 +0200
Message-ID	<rSAA2-2Xn-29@gated-at.bofh.it> (permalink)
References	<rSAqm-2Tv-3@gated-at.bofh.it>
Organization	linux.* mail to news gateway

Show all headers | View raw

The ac_classzone_idx is used as the basis for waking kswapd and that is based
on the preferred zoneref. If the preferred zoneref's first zone is lower
than what is available on other nodes, it's possible that kswapd is woken
on a zone with only higher, but still eligible, zones. As classzone_idx
is strictly adhered to now, it causes a problem because eligible pages
are skipped.

For example, node 0 has only DMA32 and node 1 has only NORMAL. An allocating
context running on node 0 may wake kswapd on node 1 telling it to skip
all NORMAL pages.

Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Acked-by: Hillf Danton <hillf.zj@alibaba-inc.com>
---
 mm/page_alloc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index bb261885c121..e6ee52f1c15f 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -3415,7 +3415,7 @@ static void wake_all_kswapds(unsigned int order, const struct alloc_context *ac)
 	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist,
 					ac->high_zoneidx, ac->nodemask) {
 		if (last_pgdat != zone->zone_pgdat)
-			wakeup_kswapd(zone, order, ac_classzone_idx(ac));
+			wakeup_kswapd(zone, order, ac->high_zoneidx);
 		last_pgdat = zone->zone_pgdat;
 	}
 }
-- 
2.6.4

Thread

[PATCH 00/34] Move LRU page reclaim from zones to nodes v9 Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:40 +0200
  [PATCH 09/34] mm, vmscan: simplify the logic deciding whether kswapd sleeps Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:40 +0200
  [PATCH 01/34] mm, vmstat: add infrastructure for per-node vmstats Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:40 +0200
  [PATCH 13/34] mm, vmscan: make shrink_node decisions more node-centric Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:40 +0200
  [PATCH 16/34] mm, page_alloc: consider dirtyable memory in terms of nodes Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:40 +0200
  [PATCH 20/34] mm: move vmscan writes and file write accounting to the node Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:40 +0200
  [PATCH 12/34] mm: vmscan: do not reclaim from kswapd if there is any eligible zone Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:40 +0200
  [PATCH 24/34] mm, vmscan: avoid passing in classzone_idx unnecessarily to shrink_node Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:40 +0200
  [PATCH 14/34] mm, memcg: move memcg limit enforcement from zones to nodes Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:40 +0200
  [PATCH 04/34] mm, mmzone: clarify the usage of zone padding Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:40 +0200
  [PATCH 02/34] mm, vmscan: move lru_lock to the node Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:40 +0200
  [PATCH 15/34] mm, workingset: make working set detection node-aware Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:40 +0200
  [PATCH 23/34] mm: convert zone_reclaim to node_reclaim Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:40 +0200
  [PATCH 07/34] mm, vmscan: make kswapd reclaim in terms of nodes Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:50 +0200
  [PATCH 30/34] mm: page_alloc: cache the last node whose dirty limit is reached Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:50 +0200
  [PATCH 31/34] mm: vmstat: replace __count_zone_vm_events with a zone id equivalent Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:50 +0200
  [PATCH 34/34] mm, vmstat: remove zone and node double accounting by approximating retries Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:50 +0200
  [PATCH 10/34] mm, vmscan: by default have direct reclaim only shrink once per node Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:50 +0200
  [PATCH 32/34] mm: vmstat: account per-zone stalls and pages skipped during reclaim Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:50 +0200
  [PATCH 25/34] mm, vmscan: avoid passing in classzone_idx unnecessarily to compaction_ready Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:50 +0200
  [PATCH 28/34] mm, vmscan: add classzone information to tracepoints Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:50 +0200
  [PATCH 27/34] mm, vmscan: Have kswapd reclaim from all zones if reclaiming and buffer_heads_over_limit Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:50 +0200
  [PATCH 26/34] mm, vmscan: avoid passing in remaining unnecessarily to prepare_kswapd_sleep Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:50 +0200
  [PATCH 29/34] mm, page_alloc: remove fair zone allocation policy Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:50 +0200
  [PATCH 22/34] mm, page_alloc: wake kswapd based on the highest eligible zone Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:50 +0200
  [PATCH 33/34] mm, vmstat: print node-based stats in zoneinfo file Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:50 +0200
  [PATCH 19/34] mm: move most file-based accounting to the node Mel Gorman <mgorman@techsingularity.net> - 2016-07-08 11:50 +0200

csiph-web