memcg: add inactive_anon_is_low()
The inactive_anon_is_low() is key component of active/inactive anon
balancing on reclaim. However current inactive_anon_is_low() function
only consider global reclaim.
Therefore, we need following ugly scan_global_lru() condition.
if (lru == LRU_ACTIVE_ANON &&
(!scan_global_lru(sc) || inactive_anon_is_low(zone))) {
shrink_active_list(nr_to_scan, zone, sc, priority, file);
return 0;
it cause that memcg reclaim always deactivate pages when shrink_list() is
called. To make mem_cgroup_inactive_anon_is_low() improve active/inactive
anon balancing of memcgroup.
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Acked-by: Rik van Riel <riel@redhat.com>
Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Cyrill Gorcunov <gorcunov@gmail.com>
Cc: "Pekka Enberg" <penberg@cs.helsinki.fi>
Cc: Balbir Singh <balbir@in.ibm.com>
Cc: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
Cc: Hugh Dickins <hugh@veritas.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 457d671..66113284 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -156,6 +156,9 @@
unsigned long last_oom_jiffies;
int obsolete;
atomic_t refcnt;
+
+ unsigned int inactive_ratio;
+
/*
* statistics. This must be placed at the end of memcg.
*/
@@ -431,6 +434,20 @@
return (nr_pages >> priority);
}
+int mem_cgroup_inactive_anon_is_low(struct mem_cgroup *memcg, struct zone *zone)
+{
+ unsigned long active;
+ unsigned long inactive;
+
+ inactive = mem_cgroup_get_all_zonestat(memcg, LRU_INACTIVE_ANON);
+ active = mem_cgroup_get_all_zonestat(memcg, LRU_ACTIVE_ANON);
+
+ if (inactive * memcg->inactive_ratio < active)
+ return 1;
+
+ return 0;
+}
+
unsigned long mem_cgroup_isolate_pages(unsigned long nr_to_scan,
struct list_head *dst,
unsigned long *scanned, int order,
@@ -1360,6 +1377,29 @@
return 0;
}
+/*
+ * The inactive anon list should be small enough that the VM never has to
+ * do too much work, but large enough that each inactive page has a chance
+ * to be referenced again before it is swapped out.
+ *
+ * this calculation is straightforward porting from
+ * page_alloc.c::setup_per_zone_inactive_ratio().
+ * it describe more detail.
+ */
+static void mem_cgroup_set_inactive_ratio(struct mem_cgroup *memcg)
+{
+ unsigned int gb, ratio;
+
+ gb = res_counter_read_u64(&memcg->res, RES_LIMIT) >> 30;
+ if (gb)
+ ratio = int_sqrt(10 * gb);
+ else
+ ratio = 1;
+
+ memcg->inactive_ratio = ratio;
+
+}
+
static DEFINE_MUTEX(set_limit_mutex);
static int mem_cgroup_resize_limit(struct mem_cgroup *memcg,
@@ -1398,6 +1438,10 @@
GFP_KERNEL, false);
if (!progress) retry_count--;
}
+
+ if (!ret)
+ mem_cgroup_set_inactive_ratio(memcg);
+
return ret;
}
@@ -1982,7 +2026,7 @@
res_counter_init(&mem->res, NULL);
res_counter_init(&mem->memsw, NULL);
}
-
+ mem_cgroup_set_inactive_ratio(mem);
mem->last_scanned_child = NULL;
return &mem->css;