summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorKOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>2009-01-07 18:08:25 -0800
committerLinus Torvalds <torvalds@linux-foundation.org>2009-01-08 08:31:09 -0800
commitc772be939e078afd2505ede7d596a30f8f61de95 (patch)
tree2cf3157fdd2b3570e105f2d91576585750526fb3
parenta7885eb8ad465ec9db99ac5b5e6680f0ca8e11c8 (diff)
memcg: fix calculation of active_ratio
Currently, inactive_ratio of memcg is calculated at setting limit. because page_alloc.c does so and current implementation is straightforward porting. However, memcg introduced hierarchy feature recently. In hierarchy restriction, memory limit is not only decided memory.limit_in_bytes of current cgroup, but also parent limit and sibling memory usage. Then, The optimal inactive_ratio is changed frequently. So, everytime calculation is better. Tested-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Balbir Singh <balbir@linux.vnet.ibm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
-rw-r--r--include/linux/memcontrol.h5
-rw-r--r--mm/memcontrol.c64
-rw-r--r--mm/vmscan.c2
3 files changed, 34 insertions, 37 deletions
diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index 8752052da8df..056cf82c0e86 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -97,8 +97,7 @@ extern void mem_cgroup_note_reclaim_priority(struct mem_cgroup *mem,
int priority);
extern void mem_cgroup_record_reclaim_priority(struct mem_cgroup *mem,
int priority);
-int mem_cgroup_inactive_anon_is_low(struct mem_cgroup *memcg,
- struct zone *zone);
+int mem_cgroup_inactive_anon_is_low(struct mem_cgroup *memcg);
unsigned long mem_cgroup_zone_nr_pages(struct mem_cgroup *memcg,
struct zone *zone,
enum lru_list lru);
@@ -252,7 +251,7 @@ static inline bool mem_cgroup_oom_called(struct task_struct *task)
}
static inline int
-mem_cgroup_inactive_anon_is_low(struct mem_cgroup *memcg, struct zone *zone)
+mem_cgroup_inactive_anon_is_low(struct mem_cgroup *memcg)
{
return 1;
}
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index ab2ecbb95b8d..c7d78eca8363 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -166,9 +166,6 @@ struct mem_cgroup {
unsigned int swappiness;
-
- unsigned int inactive_ratio;
-
/*
* statistics. This must be placed at the end of memcg.
*/
@@ -432,15 +429,43 @@ void mem_cgroup_record_reclaim_priority(struct mem_cgroup *mem, int priority)
spin_unlock(&mem->reclaim_param_lock);
}
-int mem_cgroup_inactive_anon_is_low(struct mem_cgroup *memcg, struct zone *zone)
+static int calc_inactive_ratio(struct mem_cgroup *memcg, unsigned long *present_pages)
{
unsigned long active;
unsigned long inactive;
+ unsigned long gb;
+ unsigned long inactive_ratio;
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)
+ gb = (inactive + active) >> (30 - PAGE_SHIFT);
+ if (gb)
+ inactive_ratio = int_sqrt(10 * gb);
+ else
+ inactive_ratio = 1;
+
+ if (present_pages) {
+ present_pages[0] = inactive;
+ present_pages[1] = active;
+ }
+
+ return inactive_ratio;
+}
+
+int mem_cgroup_inactive_anon_is_low(struct mem_cgroup *memcg)
+{
+ unsigned long active;
+ unsigned long inactive;
+ unsigned long present_pages[2];
+ unsigned long inactive_ratio;
+
+ inactive_ratio = calc_inactive_ratio(memcg, present_pages);
+
+ inactive = present_pages[0];
+ active = present_pages[1];
+
+ if (inactive * inactive_ratio < active)
return 1;
return 0;
@@ -1432,29 +1457,6 @@ int mem_cgroup_shrink_usage(struct mm_struct *mm, gfp_t gfp_mask)
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,
@@ -1496,9 +1498,6 @@ static int mem_cgroup_resize_limit(struct mem_cgroup *memcg,
if (!progress) retry_count--;
}
- if (!ret)
- mem_cgroup_set_inactive_ratio(memcg);
-
return ret;
}
@@ -1858,7 +1857,7 @@ static int mem_control_stat_show(struct cgroup *cont, struct cftype *cft,
}
#ifdef CONFIG_DEBUG_VM
- cb->fill(cb, "inactive_ratio", mem_cont->inactive_ratio);
+ cb->fill(cb, "inactive_ratio", calc_inactive_ratio(mem_cont, NULL));
{
int nid, zid;
@@ -2150,7 +2149,6 @@ mem_cgroup_create(struct cgroup_subsys *ss, struct cgroup *cont)
res_counter_init(&mem->res, NULL);
res_counter_init(&mem->memsw, NULL);
}
- mem_cgroup_set_inactive_ratio(mem);
mem->last_scanned_child = NULL;
spin_lock_init(&mem->reclaim_param_lock);
diff --git a/mm/vmscan.c b/mm/vmscan.c
index ece2f405187f..9a27c44aa327 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -1340,7 +1340,7 @@ static int inactive_anon_is_low(struct zone *zone, struct scan_control *sc)
if (scanning_global_lru(sc))
low = inactive_anon_is_low_global(zone);
else
- low = mem_cgroup_inactive_anon_is_low(sc->mem_cgroup, zone);
+ low = mem_cgroup_inactive_anon_is_low(sc->mem_cgroup);
return low;
}