[PATCH V2 0/2] Reduce reclaim from per-zone LRU in global kswapd

[PATCH V2 0/2] Reduce reclaim from per-zone LRU in global kswapd

[Ying Han]

The global kswapd scans per-zone LRU and reclaims pages regardless of the
cgroup. It breaks memory isolation since one cgroup can end up reclaiming
pages from another cgroup. Instead we should rely on memcg-aware target
reclaim including per-memcg kswapd and soft_limit hierarchical reclaim under
memory pressure.

In the global background reclaim, we do soft reclaim before scanning the
per-zone LRU. However, the return value is ignored. This patch is the first
step to skip shrink_zone() if soft_limit reclaim does enough work.

This is part of the effort which tries to reduce reclaiming pages in global
LRU in memcg. The per-memcg background reclaim patchset further enhances the
per-cgroup targetting reclaim, which I should have V4 posted shortly.

Try running multiple memory intensive workloads within seperate memcgs. Watch
the counters of soft_steal in memory.stat.

$ egrep 'steal|scan' /dev/cgroup/1/memory.stat
soft_steal 259200
total_soft_steal 259200

Ying Han (2):
  count the soft_limit reclaim in global background reclaim
  add stats to monitor soft_limit reclaim

 Documentation/cgroups/memory.txt |    2 +
 include/linux/memcontrol.h       |   11 ++++++++-
 include/linux/swap.h             |    3 +-
 mm/memcontrol.c                  |   43 ++++++++++++++++++++++++++++++--------
 mm/vmscan.c                      |   16 +++++++++++--
 5 files changed, 60 insertions(+), 15 deletions(-)

-- 
1.7.3.1

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

[PATCH V2 1/2] count the soft_limit reclaim in global background reclaim

[Ying Han]

In the global background reclaim, we do soft reclaim before scanning the
per-zone LRU. However, the return value is ignored.

We would like to skip shrink_zone() if soft_limit reclaim does enough work.
Also, we need to make the memory pressure balanced across per-memcg zones,
like the logic vm-core. This patch is the first step where we start with
counting the nr_scanned and nr_reclaimed from soft_limit reclaim into the
global scan_control.

Change log v2...v1:
1. Not skipping the shrink_zone() but instead count the nr_scanned and
nr_reclaimed in the global scan_control.
2. Removed the stats into the next patch.

Signed-off-by: Ying Han <yinghan@google.com>
---
 include/linux/memcontrol.h |    6 ++++--
 include/linux/swap.h       |    3 ++-
 mm/memcontrol.c            |   29 ++++++++++++++++++++---------
 mm/vmscan.c                |   16 +++++++++++++---
 4 files changed, 39 insertions(+), 15 deletions(-)

diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index 5a5ce70..01281ac 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -144,7 +144,8 @@ static inline void mem_cgroup_dec_page_stat(struct page *page,
 }
 
 unsigned long mem_cgroup_soft_limit_reclaim(struct zone *zone, int order,
-						gfp_t gfp_mask);
+						gfp_t gfp_mask,
+						unsigned long *total_scanned);
 u64 mem_cgroup_get_limit(struct mem_cgroup *mem);
 
 #ifdef CONFIG_TRANSPARENT_HUGEPAGE
@@ -338,7 +339,8 @@ static inline void mem_cgroup_dec_page_stat(struct page *page,
 
 static inline
 unsigned long mem_cgroup_soft_limit_reclaim(struct zone *zone, int order,
-					    gfp_t gfp_mask)
+					    gfp_t gfp_mask,
+					    unsigned long *total_scanned)
 {
 	return 0;
 }
diff --git a/include/linux/swap.h b/include/linux/swap.h
index ed6ebe6..3c6a9cd 100644
--- a/include/linux/swap.h
+++ b/include/linux/swap.h
@@ -257,7 +257,8 @@ extern unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem,
 extern unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
 						gfp_t gfp_mask, bool noswap,
 						unsigned int swappiness,
-						struct zone *zone);
+						struct zone *zone,
+						unsigned long *nr_scanned);
 extern int __isolate_lru_page(struct page *page, int mode, int file);
 extern unsigned long shrink_all_memory(unsigned long nr_pages);
 extern int vm_swappiness;
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 4407dd0..67fff28 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -1433,7 +1433,8 @@ mem_cgroup_select_victim(struct mem_cgroup *root_mem)
 static int mem_cgroup_hierarchical_reclaim(struct mem_cgroup *root_mem,
 						struct zone *zone,
 						gfp_t gfp_mask,
-						unsigned long reclaim_options)
+						unsigned long reclaim_options,
+						unsigned long *total_scanned)
 {
 	struct mem_cgroup *victim;
 	int ret, total = 0;
@@ -1442,6 +1443,7 @@ static int mem_cgroup_hierarchical_reclaim(struct mem_cgroup *root_mem,
 	bool shrink = reclaim_options & MEM_CGROUP_RECLAIM_SHRINK;
 	bool check_soft = reclaim_options & MEM_CGROUP_RECLAIM_SOFT;
 	unsigned long excess;
+	unsigned long nr_scanned;
 
 	excess = res_counter_soft_limit_excess(&root_mem->res) >> PAGE_SHIFT;
 
@@ -1484,10 +1486,12 @@ static int mem_cgroup_hierarchical_reclaim(struct mem_cgroup *root_mem,
 			continue;
 		}
 		/* we use swappiness of local cgroup */
-		if (check_soft)
+		if (check_soft) {
 			ret = mem_cgroup_shrink_node_zone(victim, gfp_mask,
-				noswap, get_swappiness(victim), zone);
-		else
+				noswap, get_swappiness(victim), zone,
+				&nr_scanned);
+			*total_scanned += nr_scanned;
+		} else
 			ret = try_to_free_mem_cgroup_pages(victim, gfp_mask,
 						noswap, get_swappiness(victim));
 		css_put(&victim->css);
@@ -1928,7 +1932,7 @@ static int mem_cgroup_do_charge(struct mem_cgroup *mem, gfp_t gfp_mask,
 		return CHARGE_WOULDBLOCK;
 
 	ret = mem_cgroup_hierarchical_reclaim(mem_over_limit, NULL,
-					      gfp_mask, flags);
+					      gfp_mask, flags, NULL);
 	if (mem_cgroup_margin(mem_over_limit) >= nr_pages)
 		return CHARGE_RETRY;
 	/*
@@ -3211,7 +3215,8 @@ static int mem_cgroup_resize_limit(struct mem_cgroup *memcg,
 			break;
 
 		mem_cgroup_hierarchical_reclaim(memcg, NULL, GFP_KERNEL,
-						MEM_CGROUP_RECLAIM_SHRINK);
+						MEM_CGROUP_RECLAIM_SHRINK,
+						NULL);
 		curusage = res_counter_read_u64(&memcg->res, RES_USAGE);
 		/* Usage is reduced ? */
   		if (curusage >= oldusage)
@@ -3271,7 +3276,8 @@ static int mem_cgroup_resize_memsw_limit(struct mem_cgroup *memcg,
 
 		mem_cgroup_hierarchical_reclaim(memcg, NULL, GFP_KERNEL,
 						MEM_CGROUP_RECLAIM_NOSWAP |
-						MEM_CGROUP_RECLAIM_SHRINK);
+						MEM_CGROUP_RECLAIM_SHRINK,
+						NULL);
 		curusage = res_counter_read_u64(&memcg->memsw, RES_USAGE);
 		/* Usage is reduced ? */
 		if (curusage >= oldusage)
@@ -3285,7 +3291,8 @@ static int mem_cgroup_resize_memsw_limit(struct mem_cgroup *memcg,
 }
 
 unsigned long mem_cgroup_soft_limit_reclaim(struct zone *zone, int order,
-					    gfp_t gfp_mask)
+					    gfp_t gfp_mask,
+					    unsigned long *total_scanned)
 {
 	unsigned long nr_reclaimed = 0;
 	struct mem_cgroup_per_zone *mz, *next_mz = NULL;
@@ -3293,6 +3300,7 @@ unsigned long mem_cgroup_soft_limit_reclaim(struct zone *zone, int order,
 	int loop = 0;
 	struct mem_cgroup_tree_per_zone *mctz;
 	unsigned long long excess;
+	unsigned long nr_scanned;
 
 	if (order > 0)
 		return 0;
@@ -3311,10 +3319,13 @@ unsigned long mem_cgroup_soft_limit_reclaim(struct zone *zone, int order,
 		if (!mz)
 			break;
 
+		nr_scanned = 0;
 		reclaimed = mem_cgroup_hierarchical_reclaim(mz->mem, zone,
 						gfp_mask,
-						MEM_CGROUP_RECLAIM_SOFT);
+						MEM_CGROUP_RECLAIM_SOFT,
+						&nr_scanned);
 		nr_reclaimed += reclaimed;
+		*total_scanned += nr_scanned;
 		spin_lock(&mctz->lock);
 
 		/*
diff --git a/mm/vmscan.c b/mm/vmscan.c
index 060e4c1..3755ad5 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -2147,9 +2147,11 @@ unsigned long try_to_free_pages(struct zonelist *zonelist, int order,
 unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
 						gfp_t gfp_mask, bool noswap,
 						unsigned int swappiness,
-						struct zone *zone)
+						struct zone *zone,
+						unsigned long *nr_scanned)
 {
 	struct scan_control sc = {
+		.nr_scanned = 0,
 		.nr_to_reclaim = SWAP_CLUSTER_MAX,
 		.may_writepage = !laptop_mode,
 		.may_unmap = 1,
@@ -2158,6 +2160,7 @@ unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
 		.order = 0,
 		.mem_cgroup = mem,
 	};
+
 	sc.gfp_mask = (gfp_mask & GFP_RECLAIM_MASK) |
 			(GFP_HIGHUSER_MOVABLE & ~GFP_RECLAIM_MASK);
 
@@ -2176,6 +2179,7 @@ unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
 
 	trace_mm_vmscan_memcg_softlimit_reclaim_end(sc.nr_reclaimed);
 
+	*nr_scanned = sc.nr_scanned;
 	return sc.nr_reclaimed;
 }
 
@@ -2320,6 +2324,8 @@ static unsigned long balance_pgdat(pg_data_t *pgdat, int order,
 	int end_zone = 0;	/* Inclusive.  0 = ZONE_DMA */
 	unsigned long total_scanned;
 	struct reclaim_state *reclaim_state = current->reclaim_state;
+	unsigned long nr_soft_reclaimed;
+	unsigned long nr_soft_scanned;
 	struct scan_control sc = {
 		.gfp_mask = GFP_KERNEL,
 		.may_unmap = 1,
@@ -2409,11 +2415,15 @@ loop_again:
 
 			sc.nr_scanned = 0;
 
+			nr_soft_scanned = 0;
 			/*
 			 * Call soft limit reclaim before calling shrink_zone.
-			 * For now we ignore the return value
 			 */
-			mem_cgroup_soft_limit_reclaim(zone, order, sc.gfp_mask);
+			nr_soft_reclaimed = mem_cgroup_soft_limit_reclaim(zone,
+							order, sc.gfp_mask,
+							&nr_soft_scanned);
+			sc.nr_reclaimed += nr_soft_reclaimed;
+			total_scanned += nr_soft_scanned;
 
 			/*
 			 * We put equal pressure on every zone, unless
-- 
1.7.3.1

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [PATCH V2 1/2] count the soft_limit reclaim in global background reclaim

[KAMEZAWA Hiroyuki]

On Mon, 28 Mar 2011 16:51:09 -0700
Ying Han <yinghan@google.com> wrote:

> In the global background reclaim, we do soft reclaim before scanning the
> per-zone LRU. However, the return value is ignored.
> 
> We would like to skip shrink_zone() if soft_limit reclaim does enough work.
> Also, we need to make the memory pressure balanced across per-memcg zones,
> like the logic vm-core. This patch is the first step where we start with
> counting the nr_scanned and nr_reclaimed from soft_limit reclaim into the
> global scan_control.
> 
> Change log v2...v1:
> 1. Not skipping the shrink_zone() but instead count the nr_scanned and
> nr_reclaimed in the global scan_control.
> 2. Removed the stats into the next patch.
> 
> Signed-off-by: Ying Han <yinghan@google.com>

seems better.

Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujtisu.com>

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [PATCH V2 1/2] count the soft_limit reclaim in global background reclaim

[KOSAKI Motohiro]

> @@ -2320,6 +2324,8 @@ static unsigned long balance_pgdat(pg_data_t *pgdat, int order,
>  	int end_zone = 0;	/* Inclusive.  0 = ZONE_DMA */
>  	unsigned long total_scanned;
>  	struct reclaim_state *reclaim_state = current->reclaim_state;
> +	unsigned long nr_soft_reclaimed;
> +	unsigned long nr_soft_scanned;
>  	struct scan_control sc = {
>  		.gfp_mask = GFP_KERNEL,
>  		.may_unmap = 1,
> @@ -2409,11 +2415,15 @@ loop_again:
>  
>  			sc.nr_scanned = 0;
>  
> +			nr_soft_scanned = 0;
>  			/*
>  			 * Call soft limit reclaim before calling shrink_zone.
> -			 * For now we ignore the return value
>  			 */
> -			mem_cgroup_soft_limit_reclaim(zone, order, sc.gfp_mask);
> +			nr_soft_reclaimed = mem_cgroup_soft_limit_reclaim(zone,
> +							order, sc.gfp_mask,
> +							&nr_soft_scanned);
> +			sc.nr_reclaimed += nr_soft_reclaimed;
> +			total_scanned += nr_soft_scanned;
>  
>  			/*
>  			 * We put equal pressure on every zone, unless

Thank you.
	Reviewed-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>



--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

[PATCH V2 2/2] add stats to monitor soft_limit reclaim

[Ying Han]

The stat is added:

/dev/cgroup/*/memory.stat
soft_steal:        - # of pages reclaimed from soft_limit hierarchical reclaim
total_soft_steal:  - # sum of all children's "soft_steal"

Change log v2...v1
1. removed the counting on number of skips on shrink_zone. This is due to the
change on the previous patch.

Signed-off-by: Ying Han <yinghan@google.com>
---
 Documentation/cgroups/memory.txt |    2 ++
 include/linux/memcontrol.h       |    5 +++++
 mm/memcontrol.c                  |   14 ++++++++++++++
 3 files changed, 21 insertions(+), 0 deletions(-)

diff --git a/Documentation/cgroups/memory.txt b/Documentation/cgroups/memory.txt
index b6ed61c..dcda6c5 100644
--- a/Documentation/cgroups/memory.txt
+++ b/Documentation/cgroups/memory.txt
@@ -385,6 +385,7 @@ mapped_file	- # of bytes of mapped file (includes tmpfs/shmem)
 pgpgin		- # of pages paged in (equivalent to # of charging events).
 pgpgout		- # of pages paged out (equivalent to # of uncharging events).
 swap		- # of bytes of swap usage
+soft_steal	- # of pages reclaimed from global hierarchical reclaim
 inactive_anon	- # of bytes of anonymous memory and swap cache memory on
 		LRU list.
 active_anon	- # of bytes of anonymous and swap cache memory on active
@@ -406,6 +407,7 @@ total_mapped_file	- sum of all children's "cache"
 total_pgpgin		- sum of all children's "pgpgin"
 total_pgpgout		- sum of all children's "pgpgout"
 total_swap		- sum of all children's "swap"
+total_soft_steal	- sum of all children's "soft_steal"
 total_inactive_anon	- sum of all children's "inactive_anon"
 total_active_anon	- sum of all children's "active_anon"
 total_inactive_file	- sum of all children's "inactive_file"
diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index 01281ac..151ab40 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -115,6 +115,7 @@ struct zone_reclaim_stat*
 mem_cgroup_get_reclaim_stat_from_page(struct page *page);
 extern void mem_cgroup_print_oom_info(struct mem_cgroup *memcg,
 					struct task_struct *p);
+void mem_cgroup_soft_steal(struct mem_cgroup *memcg, int val);
 
 #ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP
 extern int do_swap_account;
@@ -356,6 +357,10 @@ static inline void mem_cgroup_split_huge_fixup(struct page *head,
 {
 }
 
+static inline void mem_cgroup_soft_steal(struct mem_cgroup *memcg,
+					 int val)
+{
+}
 #endif /* CONFIG_CGROUP_MEM_CONT */
 
 #if !defined(CONFIG_CGROUP_MEM_RES_CTLR) || !defined(CONFIG_DEBUG_VM)
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 67fff28..5e4aa41 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -94,6 +94,8 @@ enum mem_cgroup_events_index {
 	MEM_CGROUP_EVENTS_PGPGIN,	/* # of pages paged in */
 	MEM_CGROUP_EVENTS_PGPGOUT,	/* # of pages paged out */
 	MEM_CGROUP_EVENTS_COUNT,	/* # of pages paged in/out */
+	MEM_CGROUP_EVENTS_SOFT_STEAL,	/* # of pages reclaimed from */
+					/* oft reclaim               */
 	MEM_CGROUP_EVENTS_NSTATS,
 };
 /*
@@ -624,6 +626,11 @@ static void mem_cgroup_charge_statistics(struct mem_cgroup *mem,
 	preempt_enable();
 }
 
+void mem_cgroup_soft_steal(struct mem_cgroup *mem, int val)
+{
+	this_cpu_add(mem->stat->events[MEM_CGROUP_EVENTS_SOFT_STEAL], val);
+}
+
 static unsigned long mem_cgroup_get_local_zonestat(struct mem_cgroup *mem,
 					enum lru_list idx)
 {
@@ -3326,6 +3333,9 @@ unsigned long mem_cgroup_soft_limit_reclaim(struct zone *zone, int order,
 						&nr_scanned);
 		nr_reclaimed += reclaimed;
 		*total_scanned += nr_scanned;
+
+		mem_cgroup_soft_steal(mz->mem, reclaimed);
+
 		spin_lock(&mctz->lock);
 
 		/*
@@ -3783,6 +3793,7 @@ enum {
 	MCS_PGPGIN,
 	MCS_PGPGOUT,
 	MCS_SWAP,
+	MCS_SOFT_STEAL,
 	MCS_INACTIVE_ANON,
 	MCS_ACTIVE_ANON,
 	MCS_INACTIVE_FILE,
@@ -3805,6 +3816,7 @@ struct {
 	{"pgpgin", "total_pgpgin"},
 	{"pgpgout", "total_pgpgout"},
 	{"swap", "total_swap"},
+	{"soft_steal", "total_soft_steal"},
 	{"inactive_anon", "total_inactive_anon"},
 	{"active_anon", "total_active_anon"},
 	{"inactive_file", "total_inactive_file"},
@@ -3833,6 +3845,8 @@ mem_cgroup_get_local_stat(struct mem_cgroup *mem, struct mcs_total_stat *s)
 		val = mem_cgroup_read_stat(mem, MEM_CGROUP_STAT_SWAPOUT);
 		s->stat[MCS_SWAP] += val * PAGE_SIZE;
 	}
+	val = mem_cgroup_read_events(mem, MEM_CGROUP_EVENTS_SOFT_STEAL);
+	s->stat[MCS_SOFT_STEAL] += val;
 
 	/* per zone stat */
 	val = mem_cgroup_get_local_zonestat(mem, LRU_INACTIVE_ANON);
-- 
1.7.3.1

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [PATCH V2 2/2] add stats to monitor soft_limit reclaim

[KAMEZAWA Hiroyuki]

On Mon, 28 Mar 2011 16:51:10 -0700
Ying Han <yinghan@google.com> wrote:

> The stat is added:
> 
> /dev/cgroup/*/memory.stat
> soft_steal:        - # of pages reclaimed from soft_limit hierarchical reclaim
> total_soft_steal:  - # sum of all children's "soft_steal"
> 
> Change log v2...v1
> 1. removed the counting on number of skips on shrink_zone. This is due to the
> change on the previous patch.
> 
> Signed-off-by: Ying Han <yinghan@google.com>

Hmm...


> ---
>  Documentation/cgroups/memory.txt |    2 ++
>  include/linux/memcontrol.h       |    5 +++++
>  mm/memcontrol.c                  |   14 ++++++++++++++
>  3 files changed, 21 insertions(+), 0 deletions(-)
> 
> diff --git a/Documentation/cgroups/memory.txt b/Documentation/cgroups/memory.txt
> index b6ed61c..dcda6c5 100644
> --- a/Documentation/cgroups/memory.txt
> +++ b/Documentation/cgroups/memory.txt
> @@ -385,6 +385,7 @@ mapped_file	- # of bytes of mapped file (includes tmpfs/shmem)
>  pgpgin		- # of pages paged in (equivalent to # of charging events).
>  pgpgout		- # of pages paged out (equivalent to # of uncharging events).
>  swap		- # of bytes of swap usage
> +soft_steal	- # of pages reclaimed from global hierarchical reclaim
>  inactive_anon	- # of bytes of anonymous memory and swap cache memory on
>  		LRU list.
>  active_anon	- # of bytes of anonymous and swap cache memory on active
> @@ -406,6 +407,7 @@ total_mapped_file	- sum of all children's "cache"
>  total_pgpgin		- sum of all children's "pgpgin"
>  total_pgpgout		- sum of all children's "pgpgout"
>  total_swap		- sum of all children's "swap"
> +total_soft_steal	- sum of all children's "soft_steal"
>  total_inactive_anon	- sum of all children's "inactive_anon"
>  total_active_anon	- sum of all children's "active_anon"
>  total_inactive_file	- sum of all children's "inactive_file"
> diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
> index 01281ac..151ab40 100644
> --- a/include/linux/memcontrol.h
> +++ b/include/linux/memcontrol.h
> @@ -115,6 +115,7 @@ struct zone_reclaim_stat*
>  mem_cgroup_get_reclaim_stat_from_page(struct page *page);
>  extern void mem_cgroup_print_oom_info(struct mem_cgroup *memcg,
>  					struct task_struct *p);
> +void mem_cgroup_soft_steal(struct mem_cgroup *memcg, int val);
>  
>  #ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP
>  extern int do_swap_account;
> @@ -356,6 +357,10 @@ static inline void mem_cgroup_split_huge_fixup(struct page *head,
>  {
>  }
>  
> +static inline void mem_cgroup_soft_steal(struct mem_cgroup *memcg,
> +					 int val)
> +{
> +}
>  #endif /* CONFIG_CGROUP_MEM_CONT */
>  
>  #if !defined(CONFIG_CGROUP_MEM_RES_CTLR) || !defined(CONFIG_DEBUG_VM)
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index 67fff28..5e4aa41 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -94,6 +94,8 @@ enum mem_cgroup_events_index {
>  	MEM_CGROUP_EVENTS_PGPGIN,	/* # of pages paged in */
>  	MEM_CGROUP_EVENTS_PGPGOUT,	/* # of pages paged out */
>  	MEM_CGROUP_EVENTS_COUNT,	/* # of pages paged in/out */
> +	MEM_CGROUP_EVENTS_SOFT_STEAL,	/* # of pages reclaimed from */
> +					/* oft reclaim               */
>  	MEM_CGROUP_EVENTS_NSTATS,
>  };
>  /*
> @@ -624,6 +626,11 @@ static void mem_cgroup_charge_statistics(struct mem_cgroup *mem,
>  	preempt_enable();
>  }
>  
> +void mem_cgroup_soft_steal(struct mem_cgroup *mem, int val)
> +{
> +	this_cpu_add(mem->stat->events[MEM_CGROUP_EVENTS_SOFT_STEAL], val);
> +}
> +
>  static unsigned long mem_cgroup_get_local_zonestat(struct mem_cgroup *mem,
>  					enum lru_list idx)
>  {
> @@ -3326,6 +3333,9 @@ unsigned long mem_cgroup_soft_limit_reclaim(struct zone *zone, int order,
>  						&nr_scanned);
>  		nr_reclaimed += reclaimed;
>  		*total_scanned += nr_scanned;
> +
> +		mem_cgroup_soft_steal(mz->mem, reclaimed);
> +

Here, you add "the number of reclaimed pages from the all descendants under me".
Could you move this to mem_cgroup_hierarchical_reclaim() ? Then, you can report
the correct stats even with hierarchy enabled.

Even if the value is recorded into hierarchy, total_steal will show total.

BTW, soft_scan and soft_total_scan aren't necessary ?

Thanks,
-Kame

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [PATCH V2 2/2] add stats to monitor soft_limit reclaim

[Ying Han]

On Mon, Mar 28, 2011 at 6:22 PM, KAMEZAWA Hiroyuki
<kamezawa.hiroyu@jp.fujitsu.com> wrote:
> On Mon, 28 Mar 2011 16:51:10 -0700
> Ying Han <yinghan@google.com> wrote:
>
>> The stat is added:
>>
>> /dev/cgroup/*/memory.stat
>> soft_steal:        - # of pages reclaimed from soft_limit hierarchical reclaim
>> total_soft_steal:  - # sum of all children's "soft_steal"
>>
>> Change log v2...v1
>> 1. removed the counting on number of skips on shrink_zone. This is due to the
>> change on the previous patch.
>>
>> Signed-off-by: Ying Han <yinghan@google.com>
>
> Hmm...
>
>
>> ---
>>  Documentation/cgroups/memory.txt |    2 ++
>>  include/linux/memcontrol.h       |    5 +++++
>>  mm/memcontrol.c                  |   14 ++++++++++++++
>>  3 files changed, 21 insertions(+), 0 deletions(-)
>>
>> diff --git a/Documentation/cgroups/memory.txt b/Documentation/cgroups/memory.txt
>> index b6ed61c..dcda6c5 100644
>> --- a/Documentation/cgroups/memory.txt
>> +++ b/Documentation/cgroups/memory.txt
>> @@ -385,6 +385,7 @@ mapped_file       - # of bytes of mapped file (includes tmpfs/shmem)
>>  pgpgin               - # of pages paged in (equivalent to # of charging events).
>>  pgpgout              - # of pages paged out (equivalent to # of uncharging events).
>>  swap         - # of bytes of swap usage
>> +soft_steal   - # of pages reclaimed from global hierarchical reclaim
>>  inactive_anon        - # of bytes of anonymous memory and swap cache memory on
>>               LRU list.
>>  active_anon  - # of bytes of anonymous and swap cache memory on active
>> @@ -406,6 +407,7 @@ total_mapped_file - sum of all children's "cache"
>>  total_pgpgin         - sum of all children's "pgpgin"
>>  total_pgpgout                - sum of all children's "pgpgout"
>>  total_swap           - sum of all children's "swap"
>> +total_soft_steal     - sum of all children's "soft_steal"
>>  total_inactive_anon  - sum of all children's "inactive_anon"
>>  total_active_anon    - sum of all children's "active_anon"
>>  total_inactive_file  - sum of all children's "inactive_file"
>> diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
>> index 01281ac..151ab40 100644
>> --- a/include/linux/memcontrol.h
>> +++ b/include/linux/memcontrol.h
>> @@ -115,6 +115,7 @@ struct zone_reclaim_stat*
>>  mem_cgroup_get_reclaim_stat_from_page(struct page *page);
>>  extern void mem_cgroup_print_oom_info(struct mem_cgroup *memcg,
>>                                       struct task_struct *p);
>> +void mem_cgroup_soft_steal(struct mem_cgroup *memcg, int val);
>>
>>  #ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP
>>  extern int do_swap_account;
>> @@ -356,6 +357,10 @@ static inline void mem_cgroup_split_huge_fixup(struct page *head,
>>  {
>>  }
>>
>> +static inline void mem_cgroup_soft_steal(struct mem_cgroup *memcg,
>> +                                      int val)
>> +{
>> +}
>>  #endif /* CONFIG_CGROUP_MEM_CONT */
>>
>>  #if !defined(CONFIG_CGROUP_MEM_RES_CTLR) || !defined(CONFIG_DEBUG_VM)
>> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
>> index 67fff28..5e4aa41 100644
>> --- a/mm/memcontrol.c
>> +++ b/mm/memcontrol.c
>> @@ -94,6 +94,8 @@ enum mem_cgroup_events_index {
>>       MEM_CGROUP_EVENTS_PGPGIN,       /* # of pages paged in */
>>       MEM_CGROUP_EVENTS_PGPGOUT,      /* # of pages paged out */
>>       MEM_CGROUP_EVENTS_COUNT,        /* # of pages paged in/out */
>> +     MEM_CGROUP_EVENTS_SOFT_STEAL,   /* # of pages reclaimed from */
>> +                                     /* oft reclaim               */
>>       MEM_CGROUP_EVENTS_NSTATS,
>>  };
>>  /*
>> @@ -624,6 +626,11 @@ static void mem_cgroup_charge_statistics(struct mem_cgroup *mem,
>>       preempt_enable();
>>  }
>>
>> +void mem_cgroup_soft_steal(struct mem_cgroup *mem, int val)
>> +{
>> +     this_cpu_add(mem->stat->events[MEM_CGROUP_EVENTS_SOFT_STEAL], val);
>> +}
>> +
>>  static unsigned long mem_cgroup_get_local_zonestat(struct mem_cgroup *mem,
>>                                       enum lru_list idx)
>>  {
>> @@ -3326,6 +3333,9 @@ unsigned long mem_cgroup_soft_limit_reclaim(struct zone *zone, int order,
>>                                               &nr_scanned);
>>               nr_reclaimed += reclaimed;
>>               *total_scanned += nr_scanned;
>> +
>> +             mem_cgroup_soft_steal(mz->mem, reclaimed);
>> +
>
> Here, you add "the number of reclaimed pages from the all descendants under me".
> Could you move this to mem_cgroup_hierarchical_reclaim() ? Then, you can report
> the correct stats even with hierarchy enabled.
>
> Even if the value is recorded into hierarchy, total_steal will show total.

good point. I will make that change.

>
> BTW, soft_scan and soft_total_scan aren't necessary ?

Hmm, i can look into that.

--Ying
>
> Thanks,
> -Kame
>
>

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [PATCH V2 2/2] add stats to monitor soft_limit reclaim

[Daisuke Nishimura]

On Mon, 28 Mar 2011 16:51:10 -0700
Ying Han <yinghan@google.com> wrote:

> The stat is added:
> 
> /dev/cgroup/*/memory.stat
> soft_steal:        - # of pages reclaimed from soft_limit hierarchical reclaim
> total_soft_steal:  - # sum of all children's "soft_steal"
> 
> Change log v2...v1
> 1. removed the counting on number of skips on shrink_zone. This is due to the
> change on the previous patch.
> 
> Signed-off-by: Ying Han <yinghan@google.com>
> ---
>  Documentation/cgroups/memory.txt |    2 ++
>  include/linux/memcontrol.h       |    5 +++++
>  mm/memcontrol.c                  |   14 ++++++++++++++
>  3 files changed, 21 insertions(+), 0 deletions(-)
> 
> diff --git a/Documentation/cgroups/memory.txt b/Documentation/cgroups/memory.txt
> index b6ed61c..dcda6c5 100644
> --- a/Documentation/cgroups/memory.txt
> +++ b/Documentation/cgroups/memory.txt
> @@ -385,6 +385,7 @@ mapped_file	- # of bytes of mapped file (includes tmpfs/shmem)
>  pgpgin		- # of pages paged in (equivalent to # of charging events).
>  pgpgout		- # of pages paged out (equivalent to # of uncharging events).
>  swap		- # of bytes of swap usage
> +soft_steal	- # of pages reclaimed from global hierarchical reclaim
>  inactive_anon	- # of bytes of anonymous memory and swap cache memory on
>  		LRU list.
>  active_anon	- # of bytes of anonymous and swap cache memory on active
> @@ -406,6 +407,7 @@ total_mapped_file	- sum of all children's "cache"
>  total_pgpgin		- sum of all children's "pgpgin"
>  total_pgpgout		- sum of all children's "pgpgout"
>  total_swap		- sum of all children's "swap"
> +total_soft_steal	- sum of all children's "soft_steal"
>  total_inactive_anon	- sum of all children's "inactive_anon"
>  total_active_anon	- sum of all children's "active_anon"
>  total_inactive_file	- sum of all children's "inactive_file"
> diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
> index 01281ac..151ab40 100644
> --- a/include/linux/memcontrol.h
> +++ b/include/linux/memcontrol.h
> @@ -115,6 +115,7 @@ struct zone_reclaim_stat*
>  mem_cgroup_get_reclaim_stat_from_page(struct page *page);
>  extern void mem_cgroup_print_oom_info(struct mem_cgroup *memcg,
>  					struct task_struct *p);
> +void mem_cgroup_soft_steal(struct mem_cgroup *memcg, int val);
>  
>  #ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP
>  extern int do_swap_account;
> @@ -356,6 +357,10 @@ static inline void mem_cgroup_split_huge_fixup(struct page *head,
>  {
>  }
>  
> +static inline void mem_cgroup_soft_steal(struct mem_cgroup *memcg,
> +					 int val)
> +{
> +}
>  #endif /* CONFIG_CGROUP_MEM_CONT */
>  
Do you use this function outside of memcontrol.c in future, right ?
I'm asking just for clarification, and I'm sorry if I miss some past discussions.

Thanks,
Daisuke Nishimura.

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [PATCH V2 2/2] add stats to monitor soft_limit reclaim

[Ying Han]

On Mon, Mar 28, 2011 at 7:32 PM, Daisuke Nishimura
<nishimura@mxp.nes.nec.co.jp> wrote:
> On Mon, 28 Mar 2011 16:51:10 -0700
> Ying Han <yinghan@google.com> wrote:
>
>> The stat is added:
>>
>> /dev/cgroup/*/memory.stat
>> soft_steal:        - # of pages reclaimed from soft_limit hierarchical reclaim
>> total_soft_steal:  - # sum of all children's "soft_steal"
>>
>> Change log v2...v1
>> 1. removed the counting on number of skips on shrink_zone. This is due to the
>> change on the previous patch.
>>
>> Signed-off-by: Ying Han <yinghan@google.com>
>> ---
>>  Documentation/cgroups/memory.txt |    2 ++
>>  include/linux/memcontrol.h       |    5 +++++
>>  mm/memcontrol.c                  |   14 ++++++++++++++
>>  3 files changed, 21 insertions(+), 0 deletions(-)
>>
>> diff --git a/Documentation/cgroups/memory.txt b/Documentation/cgroups/memory.txt
>> index b6ed61c..dcda6c5 100644
>> --- a/Documentation/cgroups/memory.txt
>> +++ b/Documentation/cgroups/memory.txt
>> @@ -385,6 +385,7 @@ mapped_file       - # of bytes of mapped file (includes tmpfs/shmem)
>>  pgpgin               - # of pages paged in (equivalent to # of charging events).
>>  pgpgout              - # of pages paged out (equivalent to # of uncharging events).
>>  swap         - # of bytes of swap usage
>> +soft_steal   - # of pages reclaimed from global hierarchical reclaim
>>  inactive_anon        - # of bytes of anonymous memory and swap cache memory on
>>               LRU list.
>>  active_anon  - # of bytes of anonymous and swap cache memory on active
>> @@ -406,6 +407,7 @@ total_mapped_file - sum of all children's "cache"
>>  total_pgpgin         - sum of all children's "pgpgin"
>>  total_pgpgout                - sum of all children's "pgpgout"
>>  total_swap           - sum of all children's "swap"
>> +total_soft_steal     - sum of all children's "soft_steal"
>>  total_inactive_anon  - sum of all children's "inactive_anon"
>>  total_active_anon    - sum of all children's "active_anon"
>>  total_inactive_file  - sum of all children's "inactive_file"
>> diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
>> index 01281ac..151ab40 100644
>> --- a/include/linux/memcontrol.h
>> +++ b/include/linux/memcontrol.h
>> @@ -115,6 +115,7 @@ struct zone_reclaim_stat*
>>  mem_cgroup_get_reclaim_stat_from_page(struct page *page);
>>  extern void mem_cgroup_print_oom_info(struct mem_cgroup *memcg,
>>                                       struct task_struct *p);
>> +void mem_cgroup_soft_steal(struct mem_cgroup *memcg, int val);
>>
>>  #ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP
>>  extern int do_swap_account;
>> @@ -356,6 +357,10 @@ static inline void mem_cgroup_split_huge_fixup(struct page *head,
>>  {
>>  }
>>
>> +static inline void mem_cgroup_soft_steal(struct mem_cgroup *memcg,
>> +                                      int val)
>> +{
>> +}
>>  #endif /* CONFIG_CGROUP_MEM_CONT */
>>
> Do you use this function outside of memcontrol.c in future, right ?
> I'm asking just for clarification, and I'm sorry if I miss some past discussions.

No, you didn't miss the discussion. That is a reasonable change. thanks

--Ying
>
> Thanks,
> Daisuke Nishimura.
>

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>