[PATCH] [Cleanup] memcg: export memory cgroup's swappiness v2

[PATCH] [Cleanup] memcg: export memory cgroup's swappiness v2

[KAMEZAWA Hiroyuki]

This is onto 3 patches I posted yesterday.
I'm sorry I once got Acks in v1 but refleshed totally. 
==
>From fda72071ee473be1caee163920da0f5c397c95e8 Mon Sep 17 00:00:00 2001
From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Date: Wed, 29 Jun 2011 18:24:49 +0900
Subject: [PATCH] export memory cgroup's swappines

Each memory cgroup has 'swappiness' value and it can be accessed by
get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
and swappiness is passed by argument. It's propagated by scan_control.

get_swappiness is static function but some planned updates will need to
get swappiness from files other than memcontrol.c
This patch exports get_swappiness() as mem_cgroup_swappiness().
By this, we can remove the argument of swapiness from try_to_free...
and drop swappiness from scan_control. only memcg uses it.

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

Changelog:
  - drop swappiness from scan_control
---
 include/linux/swap.h |   10 +++++++---
 mm/memcontrol.c      |   13 ++++++-------
 mm/vmscan.c          |   23 ++++++++++-------------
 3 files changed, 23 insertions(+), 23 deletions(-)

diff --git a/include/linux/swap.h b/include/linux/swap.h
index a273468..aabcca4 100644
--- a/include/linux/swap.h
+++ b/include/linux/swap.h
@@ -252,11 +252,9 @@ static inline void lru_cache_add_file(struct page *page)
 extern unsigned long try_to_free_pages(struct zonelist *zonelist, int order,
 					gfp_t gfp_mask, nodemask_t *mask);
 extern unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem,
-						  gfp_t gfp_mask, bool noswap,
-						  unsigned int swappiness);
+						  gfp_t gfp_mask, bool noswap);
 extern unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
 						gfp_t gfp_mask, bool noswap,
-						unsigned int swappiness,
 						struct zone *zone,
 						unsigned long *nr_scanned);
 extern int __isolate_lru_page(struct page *page, int mode, int file);
@@ -365,11 +363,17 @@ static inline void put_swap_token(struct mm_struct *mm)
 extern void
 mem_cgroup_uncharge_swapcache(struct page *page, swp_entry_t ent, bool swapout);
 extern int mem_cgroup_count_swap_user(swp_entry_t ent, struct page **pagep);
+extern unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem);
 #else
 static inline void
 mem_cgroup_uncharge_swapcache(struct page *page, swp_entry_t ent, bool swapout)
 {
 }
+
+static inline unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem)
+{
+	return vm_swappiness;
+}
 #endif
 #ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP
 extern void mem_cgroup_uncharge_swap(swp_entry_t ent);
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 3e7d5e6..ae10796 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -1330,7 +1330,7 @@ static unsigned long mem_cgroup_margin(struct mem_cgroup *mem)
 	return margin >> PAGE_SHIFT;
 }
 
-static unsigned int get_swappiness(struct mem_cgroup *memcg)
+unsigned int mem_cgroup_swappiness(struct mem_cgroup *memcg)
 {
 	struct cgroup *cgrp = memcg->css.cgroup;
 
@@ -1776,12 +1776,11 @@ static int mem_cgroup_hierarchical_reclaim(struct mem_cgroup *root_mem,
 		/* we use swappiness of local cgroup */
 		if (check_soft) {
 			ret = mem_cgroup_shrink_node_zone(victim, gfp_mask,
-				noswap, get_swappiness(victim), zone,
-				&nr_scanned);
+				noswap, zone, &nr_scanned);
 			*total_scanned += nr_scanned;
 		} else
 			ret = try_to_free_mem_cgroup_pages(victim, gfp_mask,
-						noswap, get_swappiness(victim));
+						noswap);
 		css_put(&victim->css);
 		/*
 		 * At shrinking usage, we can't check we should stop here or
@@ -3826,7 +3825,7 @@ try_to_free:
 			goto out;
 		}
 		progress = try_to_free_mem_cgroup_pages(mem, GFP_KERNEL,
-						false, get_swappiness(mem));
+						false);
 		if (!progress) {
 			nr_retries--;
 			/* maybe some writeback is necessary */
@@ -4288,7 +4287,7 @@ static u64 mem_cgroup_swappiness_read(struct cgroup *cgrp, struct cftype *cft)
 {
 	struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp);
 
-	return get_swappiness(memcg);
+	return mem_cgroup_swappiness(memcg);
 }
 
 static int mem_cgroup_swappiness_write(struct cgroup *cgrp, struct cftype *cft,
@@ -4997,7 +4996,7 @@ mem_cgroup_create(struct cgroup_subsys *ss, struct cgroup *cont)
 	INIT_LIST_HEAD(&mem->oom_notify);
 
 	if (parent)
-		mem->swappiness = get_swappiness(parent);
+		mem->swappiness = mem_cgroup_swappiness(parent);
 	atomic_set(&mem->refcnt, 1);
 	mem->move_charge_at_immigrate = 0;
 	mutex_init(&mem->thresholds_lock);
diff --git a/mm/vmscan.c b/mm/vmscan.c
index 4f49535..fb37699 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -95,8 +95,6 @@ struct scan_control {
 	/* Can pages be swapped as part of reclaim? */
 	int may_swap;
 
-	int swappiness;
-
 	int order;
 
 	/*
@@ -1729,6 +1727,13 @@ static unsigned long shrink_list(enum lru_list lru, unsigned long nr_to_scan,
 	return shrink_inactive_list(nr_to_scan, zone, sc, priority, file);
 }
 
+static int vmscan_swappiness(struct scan_control *sc)
+{
+	if (scanning_global_lru(sc))
+		return vm_swappiness;
+	return mem_cgroup_swappiness(sc->mem_cgroup);
+}
+
 /*
  * Determine how aggressively the anon and file LRU lists should be
  * scanned.  The relative value of each set of LRU lists is determined
@@ -1789,8 +1794,8 @@ static void get_scan_count(struct zone *zone, struct scan_control *sc,
 	 * With swappiness at 100, anonymous and file have the same priority.
 	 * This scanning priority is essentially the inverse of IO cost.
 	 */
-	anon_prio = sc->swappiness;
-	file_prio = 200 - sc->swappiness;
+	anon_prio = vmscan_swappiness(sc);
+	file_prio = 200 - vmscan_swappiness(sc);
 
 	/*
 	 * OK, so we have swap space and a fair amount of page cache
@@ -2179,7 +2184,6 @@ unsigned long try_to_free_pages(struct zonelist *zonelist, int order,
 		.nr_to_reclaim = SWAP_CLUSTER_MAX,
 		.may_unmap = 1,
 		.may_swap = 1,
-		.swappiness = vm_swappiness,
 		.order = order,
 		.mem_cgroup = NULL,
 		.nodemask = nodemask,
@@ -2203,7 +2207,6 @@ 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,
 						unsigned long *nr_scanned)
 {
@@ -2213,7 +2216,6 @@ unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
 		.may_writepage = !laptop_mode,
 		.may_unmap = 1,
 		.may_swap = !noswap,
-		.swappiness = swappiness,
 		.order = 0,
 		.mem_cgroup = mem,
 	};
@@ -2242,8 +2244,7 @@ unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
 
 unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem_cont,
 					   gfp_t gfp_mask,
-					   bool noswap,
-					   unsigned int swappiness)
+					   bool noswap)
 {
 	struct zonelist *zonelist;
 	unsigned long nr_reclaimed;
@@ -2253,7 +2254,6 @@ unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem_cont,
 		.may_unmap = 1,
 		.may_swap = !noswap,
 		.nr_to_reclaim = SWAP_CLUSTER_MAX,
-		.swappiness = swappiness,
 		.order = 0,
 		.mem_cgroup = mem_cont,
 		.nodemask = NULL, /* we don't care the placement */
@@ -2403,7 +2403,6 @@ static unsigned long balance_pgdat(pg_data_t *pgdat, int order,
 		 * we want to put equal scanning pressure on each zone.
 		 */
 		.nr_to_reclaim = ULONG_MAX,
-		.swappiness = vm_swappiness,
 		.order = order,
 		.mem_cgroup = NULL,
 	};
@@ -2862,7 +2861,6 @@ unsigned long shrink_all_memory(unsigned long nr_to_reclaim)
 		.may_writepage = 1,
 		.nr_to_reclaim = nr_to_reclaim,
 		.hibernation_mode = 1,
-		.swappiness = vm_swappiness,
 		.order = 0,
 	};
 	struct shrink_control shrink = {
@@ -3049,7 +3047,6 @@ static int __zone_reclaim(struct zone *zone, gfp_t gfp_mask, unsigned int order)
 		.nr_to_reclaim = max_t(unsigned long, nr_pages,
 				       SWAP_CLUSTER_MAX),
 		.gfp_mask = gfp_mask,
-		.swappiness = vm_swappiness,
 		.order = order,
 	};
 	struct shrink_control shrink = {
-- 
1.7.4.1

[PATCH] [Cleanup] memcg: export memory cgroup's swappiness v2

[KAMEZAWA Hiroyuki]

This is onto 3 patches I posted yesterday.
I'm sorry I once got Acks in v1 but refleshed totally. 
==

Re: [PATCH] [Cleanup] memcg: export memory cgroup's swappiness v2

[Michal Hocko]

On Wed 29-06-11 19:03:25, KAMEZAWA Hiroyuki wrote:
> This is onto 3 patches I posted yesterday.
> I'm sorry I once got Acks in v1 but refleshed totally. 
> ==
> From fda72071ee473be1caee163920da0f5c397c95e8 Mon Sep 17 00:00:00 2001
> From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> Date: Wed, 29 Jun 2011 18:24:49 +0900
> Subject: [PATCH] export memory cgroup's swappines
> 
> Each memory cgroup has 'swappiness' value and it can be accessed by
> get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
> and swappiness is passed by argument. It's propagated by scan_control.
> 
> get_swappiness is static function but some planned updates will need to
> get swappiness from files other than memcontrol.c
> This patch exports get_swappiness() as mem_cgroup_swappiness().
> By this, we can remove the argument of swapiness from try_to_free...
> and drop swappiness from scan_control. only memcg uses it.
> 
> Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

Reviewed-by: Michal Hocko <mhocko@suse.cz>

> 
> Changelog:
>   - drop swappiness from scan_control

Yes, this is even better.

-- 
Michal Hocko
SUSE Labs
SUSE LINUX s.r.o.
Lihovarska 1060/12
190 00 Praha 9    
Czech Republic

Re: [PATCH] [Cleanup] memcg: export memory cgroup's swappiness v2

[Michal Hocko]

On Wed 29-06-11 19:03:25, KAMEZAWA Hiroyuki wrote:
> This is onto 3 patches I posted yesterday.
> I'm sorry I once got Acks in v1 but refleshed totally. 
> ==
> From fda72071ee473be1caee163920da0f5c397c95e8 Mon Sep 17 00:00:00 2001
> From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> Date: Wed, 29 Jun 2011 18:24:49 +0900
> Subject: [PATCH] export memory cgroup's swappines
> 
> Each memory cgroup has 'swappiness' value and it can be accessed by
> get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
> and swappiness is passed by argument. It's propagated by scan_control.
> 
> get_swappiness is static function but some planned updates will need to
> get swappiness from files other than memcontrol.c
> This patch exports get_swappiness() as mem_cgroup_swappiness().
> By this, we can remove the argument of swapiness from try_to_free...
> and drop swappiness from scan_control. only memcg uses it.
> 
> Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

Reviewed-by: Michal Hocko <mhocko@suse.cz>

> 
> Changelog:
>   - drop swappiness from scan_control

Yes, this is even better.

-- 
Michal Hocko
SUSE Labs
SUSE LINUX s.r.o.
Lihovarska 1060/12
190 00 Praha 9    
Czech Republic

--
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] [Cleanup] memcg: export memory cgroup's swappiness v2

[Balbir Singh]

On Wed, Jun 29, 2011 at 3:33 PM, KAMEZAWA Hiroyuki
<kamezawa.hiroyu@jp.fujitsu.com> wrote:
> This is onto 3 patches I posted yesterday.
> I'm sorry I once got Acks in v1 but refleshed totally.
> ==
> From fda72071ee473be1caee163920da0f5c397c95e8 Mon Sep 17 00:00:00 2001
> From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> Date: Wed, 29 Jun 2011 18:24:49 +0900
> Subject: [PATCH] export memory cgroup's swappines
>
> Each memory cgroup has 'swappiness' value and it can be accessed by
> get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
> and swappiness is passed by argument. It's propagated by scan_control.
>
> get_swappiness is static function but some planned updates will need to
> get swappiness from files other than memcontrol.c
> This patch exports get_swappiness() as mem_cgroup_swappiness().
> By this, we can remove the argument of swapiness from try_to_free...
> and drop swappiness from scan_control. only memcg uses it.
>
> Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Good refactoring

Acked-by: Balbir Singh <bsingharora@gmail.com>

Thanks,
Balbir Singh

Re: [PATCH] [Cleanup] memcg: export memory cgroup's swappiness v2

[Balbir Singh]

On Wed, Jun 29, 2011 at 3:33 PM, KAMEZAWA Hiroyuki
<kamezawa.hiroyu@jp.fujitsu.com> wrote:
> This is onto 3 patches I posted yesterday.
> I'm sorry I once got Acks in v1 but refleshed totally.
> ==
> From fda72071ee473be1caee163920da0f5c397c95e8 Mon Sep 17 00:00:00 2001
> From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> Date: Wed, 29 Jun 2011 18:24:49 +0900
> Subject: [PATCH] export memory cgroup's swappines
>
> Each memory cgroup has 'swappiness' value and it can be accessed by
> get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
> and swappiness is passed by argument. It's propagated by scan_control.
>
> get_swappiness is static function but some planned updates will need to
> get swappiness from files other than memcontrol.c
> This patch exports get_swappiness() as mem_cgroup_swappiness().
> By this, we can remove the argument of swapiness from try_to_free...
> and drop swappiness from scan_control. only memcg uses it.
>
> Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Good refactoring

Acked-by: Balbir Singh <bsingharora@gmail.com>

Thanks,
Balbir Singh

--
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] [Cleanup] memcg: export memory cgroup's swappiness v2

[Andrew Morton]

On Wed, 29 Jun 2011 19:03:25 +0900
KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:

> Each memory cgroup has 'swappiness' value and it can be accessed by
> get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
> and swappiness is passed by argument. It's propagated by scan_control.
> 
> get_swappiness is static function but some planned updates will need to
> get swappiness from files other than memcontrol.c
> This patch exports get_swappiness() as mem_cgroup_swappiness().
> By this, we can remove the argument of swapiness from try_to_free...
> and drop swappiness from scan_control. only memcg uses it.
> 

> +extern unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem);
> +unsigned int mem_cgroup_swappiness(struct mem_cgroup *memcg)
> +static int vmscan_swappiness(struct scan_control *sc)

The patch seems a bit confused about the signedness of swappiness.

Re: [PATCH] [Cleanup] memcg: export memory cgroup's swappiness v2

[Andrew Morton]

On Wed, 29 Jun 2011 19:03:25 +0900
KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:

> Each memory cgroup has 'swappiness' value and it can be accessed by
> get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
> and swappiness is passed by argument. It's propagated by scan_control.
> 
> get_swappiness is static function but some planned updates will need to
> get swappiness from files other than memcontrol.c
> This patch exports get_swappiness() as mem_cgroup_swappiness().
> By this, we can remove the argument of swapiness from try_to_free...
> and drop swappiness from scan_control. only memcg uses it.
> 

> +extern unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem);
> +unsigned int mem_cgroup_swappiness(struct mem_cgroup *memcg)
> +static int vmscan_swappiness(struct scan_control *sc)

The patch seems a bit confused about the signedness of swappiness.

--
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] [Cleanup] memcg: export memory cgroup's swappiness v2

[KAMEZAWA Hiroyuki]

On Wed, 29 Jun 2011 13:00:43 -0700
Andrew Morton <akpm@linux-foundation.org> wrote:

> On Wed, 29 Jun 2011 19:03:25 +0900
> KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> 
> > Each memory cgroup has 'swappiness' value and it can be accessed by
> > get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
> > and swappiness is passed by argument. It's propagated by scan_control.
> > 
> > get_swappiness is static function but some planned updates will need to
> > get swappiness from files other than memcontrol.c
> > This patch exports get_swappiness() as mem_cgroup_swappiness().
> > By this, we can remove the argument of swapiness from try_to_free...
> > and drop swappiness from scan_control. only memcg uses it.
> > 
> 
> > +extern unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem);
> > +unsigned int mem_cgroup_swappiness(struct mem_cgroup *memcg)
> > +static int vmscan_swappiness(struct scan_control *sc)
> 
> The patch seems a bit confused about the signedness of swappiness.
> 

ok, v3 here. Now, memcg's one use "int" because vm_swapiness is "int".
==

>From af1bae8f2c6a8dbff048222bb45c7162b505f38b Mon Sep 17 00:00:00 2001
From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Date: Wed, 29 Jun 2011 18:24:49 +0900
Subject: [PATCH] export memory cgroup's swappines by mem_cgroup_swappiness()

Each memory cgroup has 'swappiness' value and it can be accessed by
get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
and swappiness is passed by argument. It's propagated by scan_control.

get_swappiness is static function but some planned updates will need to
get swappiness from files other than memcontrol.c
This patch exports get_swappiness() as mem_cgroup_swappiness().
By this, we can remove the argument of swapiness from try_to_free...
and drop swappiness from scan_control. only memcg uses it.

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

Changelog:
  - adjusted signedness to vm_swappiness.
  - drop swappiness from scan_control
---
 include/linux/swap.h |   10 +++++++---
 mm/memcontrol.c      |   15 +++++++--------
 mm/vmscan.c          |   23 ++++++++++-------------
 3 files changed, 24 insertions(+), 24 deletions(-)

diff --git a/include/linux/swap.h b/include/linux/swap.h
index a273468..0299832 100644
--- a/include/linux/swap.h
+++ b/include/linux/swap.h
@@ -252,11 +252,9 @@ static inline void lru_cache_add_file(struct page *page)
 extern unsigned long try_to_free_pages(struct zonelist *zonelist, int order,
 					gfp_t gfp_mask, nodemask_t *mask);
 extern unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem,
-						  gfp_t gfp_mask, bool noswap,
-						  unsigned int swappiness);
+						  gfp_t gfp_mask, bool noswap);
 extern unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
 						gfp_t gfp_mask, bool noswap,
-						unsigned int swappiness,
 						struct zone *zone,
 						unsigned long *nr_scanned);
 extern int __isolate_lru_page(struct page *page, int mode, int file);
@@ -365,11 +363,17 @@ static inline void put_swap_token(struct mm_struct *mm)
 extern void
 mem_cgroup_uncharge_swapcache(struct page *page, swp_entry_t ent, bool swapout);
 extern int mem_cgroup_count_swap_user(swp_entry_t ent, struct page **pagep);
+extern int mem_cgroup_swappiness(struct mem_cgroup *mem);
 #else
 static inline void
 mem_cgroup_uncharge_swapcache(struct page *page, swp_entry_t ent, bool swapout)
 {
 }
+
+static inline int mem_cgroup_swappiness(struct mem_cgroup *mem)
+{
+	return vm_swappiness;
+}
 #endif
 #ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP
 extern void mem_cgroup_uncharge_swap(swp_entry_t ent);
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 3e7d5e6..db70176 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -249,7 +249,7 @@ struct mem_cgroup {
 	atomic_t	oom_lock;
 	atomic_t	refcnt;
 
-	unsigned int	swappiness;
+	int	swappiness;
 	/* OOM-Killer disable */
 	int		oom_kill_disable;
 
@@ -1330,7 +1330,7 @@ static unsigned long mem_cgroup_margin(struct mem_cgroup *mem)
 	return margin >> PAGE_SHIFT;
 }
 
-static unsigned int get_swappiness(struct mem_cgroup *memcg)
+int mem_cgroup_swappiness(struct mem_cgroup *memcg)
 {
 	struct cgroup *cgrp = memcg->css.cgroup;
 
@@ -1776,12 +1776,11 @@ static int mem_cgroup_hierarchical_reclaim(struct mem_cgroup *root_mem,
 		/* we use swappiness of local cgroup */
 		if (check_soft) {
 			ret = mem_cgroup_shrink_node_zone(victim, gfp_mask,
-				noswap, get_swappiness(victim), zone,
-				&nr_scanned);
+				noswap, zone, &nr_scanned);
 			*total_scanned += nr_scanned;
 		} else
 			ret = try_to_free_mem_cgroup_pages(victim, gfp_mask,
-						noswap, get_swappiness(victim));
+						noswap);
 		css_put(&victim->css);
 		/*
 		 * At shrinking usage, we can't check we should stop here or
@@ -3826,7 +3825,7 @@ try_to_free:
 			goto out;
 		}
 		progress = try_to_free_mem_cgroup_pages(mem, GFP_KERNEL,
-						false, get_swappiness(mem));
+						false);
 		if (!progress) {
 			nr_retries--;
 			/* maybe some writeback is necessary */
@@ -4288,7 +4287,7 @@ static u64 mem_cgroup_swappiness_read(struct cgroup *cgrp, struct cftype *cft)
 {
 	struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp);
 
-	return get_swappiness(memcg);
+	return mem_cgroup_swappiness(memcg);
 }
 
 static int mem_cgroup_swappiness_write(struct cgroup *cgrp, struct cftype *cft,
@@ -4997,7 +4996,7 @@ mem_cgroup_create(struct cgroup_subsys *ss, struct cgroup *cont)
 	INIT_LIST_HEAD(&mem->oom_notify);
 
 	if (parent)
-		mem->swappiness = get_swappiness(parent);
+		mem->swappiness = mem_cgroup_swappiness(parent);
 	atomic_set(&mem->refcnt, 1);
 	mem->move_charge_at_immigrate = 0;
 	mutex_init(&mem->thresholds_lock);
diff --git a/mm/vmscan.c b/mm/vmscan.c
index 4f49535..fb37699 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -95,8 +95,6 @@ struct scan_control {
 	/* Can pages be swapped as part of reclaim? */
 	int may_swap;
 
-	int swappiness;
-
 	int order;
 
 	/*
@@ -1729,6 +1727,13 @@ static unsigned long shrink_list(enum lru_list lru, unsigned long nr_to_scan,
 	return shrink_inactive_list(nr_to_scan, zone, sc, priority, file);
 }
 
+static int vmscan_swappiness(struct scan_control *sc)
+{
+	if (scanning_global_lru(sc))
+		return vm_swappiness;
+	return mem_cgroup_swappiness(sc->mem_cgroup);
+}
+
 /*
  * Determine how aggressively the anon and file LRU lists should be
  * scanned.  The relative value of each set of LRU lists is determined
@@ -1789,8 +1794,8 @@ static void get_scan_count(struct zone *zone, struct scan_control *sc,
 	 * With swappiness at 100, anonymous and file have the same priority.
 	 * This scanning priority is essentially the inverse of IO cost.
 	 */
-	anon_prio = sc->swappiness;
-	file_prio = 200 - sc->swappiness;
+	anon_prio = vmscan_swappiness(sc);
+	file_prio = 200 - vmscan_swappiness(sc);
 
 	/*
 	 * OK, so we have swap space and a fair amount of page cache
@@ -2179,7 +2184,6 @@ unsigned long try_to_free_pages(struct zonelist *zonelist, int order,
 		.nr_to_reclaim = SWAP_CLUSTER_MAX,
 		.may_unmap = 1,
 		.may_swap = 1,
-		.swappiness = vm_swappiness,
 		.order = order,
 		.mem_cgroup = NULL,
 		.nodemask = nodemask,
@@ -2203,7 +2207,6 @@ 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,
 						unsigned long *nr_scanned)
 {
@@ -2213,7 +2216,6 @@ unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
 		.may_writepage = !laptop_mode,
 		.may_unmap = 1,
 		.may_swap = !noswap,
-		.swappiness = swappiness,
 		.order = 0,
 		.mem_cgroup = mem,
 	};
@@ -2242,8 +2244,7 @@ unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
 
 unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem_cont,
 					   gfp_t gfp_mask,
-					   bool noswap,
-					   unsigned int swappiness)
+					   bool noswap)
 {
 	struct zonelist *zonelist;
 	unsigned long nr_reclaimed;
@@ -2253,7 +2254,6 @@ unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem_cont,
 		.may_unmap = 1,
 		.may_swap = !noswap,
 		.nr_to_reclaim = SWAP_CLUSTER_MAX,
-		.swappiness = swappiness,
 		.order = 0,
 		.mem_cgroup = mem_cont,
 		.nodemask = NULL, /* we don't care the placement */
@@ -2403,7 +2403,6 @@ static unsigned long balance_pgdat(pg_data_t *pgdat, int order,
 		 * we want to put equal scanning pressure on each zone.
 		 */
 		.nr_to_reclaim = ULONG_MAX,
-		.swappiness = vm_swappiness,
 		.order = order,
 		.mem_cgroup = NULL,
 	};
@@ -2862,7 +2861,6 @@ unsigned long shrink_all_memory(unsigned long nr_to_reclaim)
 		.may_writepage = 1,
 		.nr_to_reclaim = nr_to_reclaim,
 		.hibernation_mode = 1,
-		.swappiness = vm_swappiness,
 		.order = 0,
 	};
 	struct shrink_control shrink = {
@@ -3049,7 +3047,6 @@ static int __zone_reclaim(struct zone *zone, gfp_t gfp_mask, unsigned int order)
 		.nr_to_reclaim = max_t(unsigned long, nr_pages,
 				       SWAP_CLUSTER_MAX),
 		.gfp_mask = gfp_mask,
-		.swappiness = vm_swappiness,
 		.order = order,
 	};
 	struct shrink_control shrink = {
-- 
1.7.4.1

Re: [PATCH] [Cleanup] memcg: export memory cgroup's swappiness v2

[KAMEZAWA Hiroyuki]

On Wed, 29 Jun 2011 13:00:43 -0700
Andrew Morton <akpm@linux-foundation.org> wrote:

> On Wed, 29 Jun 2011 19:03:25 +0900
> KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> 
> > Each memory cgroup has 'swappiness' value and it can be accessed by
> > get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
> > and swappiness is passed by argument. It's propagated by scan_control.
> > 
> > get_swappiness is static function but some planned updates will need to
> > get swappiness from files other than memcontrol.c
> > This patch exports get_swappiness() as mem_cgroup_swappiness().
> > By this, we can remove the argument of swapiness from try_to_free...
> > and drop swappiness from scan_control. only memcg uses it.
> > 
> 
> > +extern unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem);
> > +unsigned int mem_cgroup_swappiness(struct mem_cgroup *memcg)
> > +static int vmscan_swappiness(struct scan_control *sc)
> 
> The patch seems a bit confused about the signedness of swappiness.
> 

ok, v3 here. Now, memcg's one use "int" because vm_swapiness is "int".
==

Re: [PATCH] [Cleanup] memcg: export memory cgroup's swappiness v2

[Michal Hocko]

On Thu 30-06-11 12:32:29, KAMEZAWA Hiroyuki wrote:
> On Wed, 29 Jun 2011 13:00:43 -0700
> Andrew Morton <akpm@linux-foundation.org> wrote:
> 
> > On Wed, 29 Jun 2011 19:03:25 +0900
> > KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> > 
> > > Each memory cgroup has 'swappiness' value and it can be accessed by
> > > get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
> > > and swappiness is passed by argument. It's propagated by scan_control.
> > > 
> > > get_swappiness is static function but some planned updates will need to
> > > get swappiness from files other than memcontrol.c
> > > This patch exports get_swappiness() as mem_cgroup_swappiness().
> > > By this, we can remove the argument of swapiness from try_to_free...
> > > and drop swappiness from scan_control. only memcg uses it.
> > > 
> > 
> > > +extern unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem);
> > > +unsigned int mem_cgroup_swappiness(struct mem_cgroup *memcg)
> > > +static int vmscan_swappiness(struct scan_control *sc)
> > 
> > The patch seems a bit confused about the signedness of swappiness.
> > 
> 
> ok, v3 here. Now, memcg's one use "int" because vm_swapiness is "int".
> ==
> 
> From af1bae8f2c6a8dbff048222bb45c7162b505f38b Mon Sep 17 00:00:00 2001
> From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> Date: Wed, 29 Jun 2011 18:24:49 +0900
> Subject: [PATCH] export memory cgroup's swappines by mem_cgroup_swappiness()
> 
> Each memory cgroup has 'swappiness' value and it can be accessed by
> get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
> and swappiness is passed by argument. It's propagated by scan_control.
> 
> get_swappiness is static function but some planned updates will need to
> get swappiness from files other than memcontrol.c
> This patch exports get_swappiness() as mem_cgroup_swappiness().
> By this, we can remove the argument of swapiness from try_to_free...
> and drop swappiness from scan_control. only memcg uses it.
> 
> Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

Reviewed-by: Michal Hocko <mhocko@suse.cz>

> 
> Changelog:
>   - adjusted signedness to vm_swappiness.
>   - drop swappiness from scan_control
> ---
>  include/linux/swap.h |   10 +++++++---
>  mm/memcontrol.c      |   15 +++++++--------
>  mm/vmscan.c          |   23 ++++++++++-------------
>  3 files changed, 24 insertions(+), 24 deletions(-)
> 
[...]
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index 3e7d5e6..db70176 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
[...]
> @@ -4288,7 +4287,7 @@ static u64 mem_cgroup_swappiness_read(struct cgroup *cgrp, struct cftype *cft)
>  {
>  	struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp);
>  
> -	return get_swappiness(memcg);
> +	return mem_cgroup_swappiness(memcg);
>  }

If you want to be type clean you should change this one as well. I
think it is worth it, though. The function is called only to return the
current value to userspace and mem_cgroup_swappiness_write guaranties
that it falls down into <0,100> interval. Additionally, cftype doesn't
have any read specialization for int values so you would need to use a
generic one. Finally if you changed read part you should change also
write part and add > 0 check which is a lot of code for not that good
reason.
-- 
Michal Hocko
SUSE Labs
SUSE LINUX s.r.o.
Lihovarska 1060/12
190 00 Praha 9    
Czech Republic

Re: [PATCH] [Cleanup] memcg: export memory cgroup's swappiness v2

[Michal Hocko]

On Thu 30-06-11 12:32:29, KAMEZAWA Hiroyuki wrote:
> On Wed, 29 Jun 2011 13:00:43 -0700
> Andrew Morton <akpm@linux-foundation.org> wrote:
> 
> > On Wed, 29 Jun 2011 19:03:25 +0900
> > KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> > 
> > > Each memory cgroup has 'swappiness' value and it can be accessed by
> > > get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
> > > and swappiness is passed by argument. It's propagated by scan_control.
> > > 
> > > get_swappiness is static function but some planned updates will need to
> > > get swappiness from files other than memcontrol.c
> > > This patch exports get_swappiness() as mem_cgroup_swappiness().
> > > By this, we can remove the argument of swapiness from try_to_free...
> > > and drop swappiness from scan_control. only memcg uses it.
> > > 
> > 
> > > +extern unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem);
> > > +unsigned int mem_cgroup_swappiness(struct mem_cgroup *memcg)
> > > +static int vmscan_swappiness(struct scan_control *sc)
> > 
> > The patch seems a bit confused about the signedness of swappiness.
> > 
> 
> ok, v3 here. Now, memcg's one use "int" because vm_swapiness is "int".
> ==
> 
> From af1bae8f2c6a8dbff048222bb45c7162b505f38b Mon Sep 17 00:00:00 2001
> From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> Date: Wed, 29 Jun 2011 18:24:49 +0900
> Subject: [PATCH] export memory cgroup's swappines by mem_cgroup_swappiness()
> 
> Each memory cgroup has 'swappiness' value and it can be accessed by
> get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
> and swappiness is passed by argument. It's propagated by scan_control.
> 
> get_swappiness is static function but some planned updates will need to
> get swappiness from files other than memcontrol.c
> This patch exports get_swappiness() as mem_cgroup_swappiness().
> By this, we can remove the argument of swapiness from try_to_free...
> and drop swappiness from scan_control. only memcg uses it.
> 
> Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

Reviewed-by: Michal Hocko <mhocko@suse.cz>

> 
> Changelog:
>   - adjusted signedness to vm_swappiness.
>   - drop swappiness from scan_control
> ---
>  include/linux/swap.h |   10 +++++++---
>  mm/memcontrol.c      |   15 +++++++--------
>  mm/vmscan.c          |   23 ++++++++++-------------
>  3 files changed, 24 insertions(+), 24 deletions(-)
> 
[...]
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index 3e7d5e6..db70176 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
[...]
> @@ -4288,7 +4287,7 @@ static u64 mem_cgroup_swappiness_read(struct cgroup *cgrp, struct cftype *cft)
>  {
>  	struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp);
>  
> -	return get_swappiness(memcg);
> +	return mem_cgroup_swappiness(memcg);
>  }

If you want to be type clean you should change this one as well. I
think it is worth it, though. The function is called only to return the
current value to userspace and mem_cgroup_swappiness_write guaranties
that it falls down into <0,100> interval. Additionally, cftype doesn't
have any read specialization for int values so you would need to use a
generic one. Finally if you changed read part you should change also
write part and add > 0 check which is a lot of code for not that good
reason.
-- 
Michal Hocko
SUSE Labs
SUSE LINUX s.r.o.
Lihovarska 1060/12
190 00 Praha 9    
Czech Republic

--
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] [Cleanup] memcg: export memory cgroup's swappiness v2

[KAMEZAWA Hiroyuki]

On Thu, 30 Jun 2011 08:32:32 +0200
Michal Hocko <mhocko@suse.cz> wrote:

> On Thu 30-06-11 12:32:29, KAMEZAWA Hiroyuki wrote:
> > On Wed, 29 Jun 2011 13:00:43 -0700
> > Andrew Morton <akpm@linux-foundation.org> wrote:
> > 
> > > On Wed, 29 Jun 2011 19:03:25 +0900
> > > KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> > > 
> > > > Each memory cgroup has 'swappiness' value and it can be accessed by
> > > > get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
> > > > and swappiness is passed by argument. It's propagated by scan_control.
> > > > 
> > > > get_swappiness is static function but some planned updates will need to
> > > > get swappiness from files other than memcontrol.c
> > > > This patch exports get_swappiness() as mem_cgroup_swappiness().
> > > > By this, we can remove the argument of swapiness from try_to_free...
> > > > and drop swappiness from scan_control. only memcg uses it.
> > > > 
> > > 
> > > > +extern unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem);
> > > > +unsigned int mem_cgroup_swappiness(struct mem_cgroup *memcg)
> > > > +static int vmscan_swappiness(struct scan_control *sc)
> > > 
> > > The patch seems a bit confused about the signedness of swappiness.
> > > 
> > 
> > ok, v3 here. Now, memcg's one use "int" because vm_swapiness is "int".
> > ==
> > 
> > From af1bae8f2c6a8dbff048222bb45c7162b505f38b Mon Sep 17 00:00:00 2001
> > From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> > Date: Wed, 29 Jun 2011 18:24:49 +0900
> > Subject: [PATCH] export memory cgroup's swappines by mem_cgroup_swappiness()
> > 
> > Each memory cgroup has 'swappiness' value and it can be accessed by
> > get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
> > and swappiness is passed by argument. It's propagated by scan_control.
> > 
> > get_swappiness is static function but some planned updates will need to
> > get swappiness from files other than memcontrol.c
> > This patch exports get_swappiness() as mem_cgroup_swappiness().
> > By this, we can remove the argument of swapiness from try_to_free...
> > and drop swappiness from scan_control. only memcg uses it.
> > 
> > Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> 
> Reviewed-by: Michal Hocko <mhocko@suse.cz>
> 
> > 
> > Changelog:
> >   - adjusted signedness to vm_swappiness.
> >   - drop swappiness from scan_control
> > ---
> >  include/linux/swap.h |   10 +++++++---
> >  mm/memcontrol.c      |   15 +++++++--------
> >  mm/vmscan.c          |   23 ++++++++++-------------
> >  3 files changed, 24 insertions(+), 24 deletions(-)
> > 
> [...]
> > diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> > index 3e7d5e6..db70176 100644
> > --- a/mm/memcontrol.c
> > +++ b/mm/memcontrol.c
> [...]
> > @@ -4288,7 +4287,7 @@ static u64 mem_cgroup_swappiness_read(struct cgroup *cgrp, struct cftype *cft)
> >  {
> >  	struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp);
> >  
> > -	return get_swappiness(memcg);
> > +	return mem_cgroup_swappiness(memcg);
> >  }
> 
> If you want to be type clean you should change this one as well. I
> think it is worth it, though. The function is called only to return the
> current value to userspace and mem_cgroup_swappiness_write guaranties
> that it falls down into <0,100> interval. Additionally, cftype doesn't
> have any read specialization for int values so you would need to use a
> generic one. Finally if you changed read part you should change also
> write part and add > 0 check which is a lot of code for not that good
> reason.

I don't want to make this type-clean. Should I add type casting as
==
 return (u64) mem_cgroup_swappiness(memcg);
==
?

Thanks,
-Kame

Re: [PATCH] [Cleanup] memcg: export memory cgroup's swappiness v2

[KAMEZAWA Hiroyuki]

On Thu, 30 Jun 2011 08:32:32 +0200
Michal Hocko <mhocko@suse.cz> wrote:

> On Thu 30-06-11 12:32:29, KAMEZAWA Hiroyuki wrote:
> > On Wed, 29 Jun 2011 13:00:43 -0700
> > Andrew Morton <akpm@linux-foundation.org> wrote:
> > 
> > > On Wed, 29 Jun 2011 19:03:25 +0900
> > > KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> > > 
> > > > Each memory cgroup has 'swappiness' value and it can be accessed by
> > > > get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
> > > > and swappiness is passed by argument. It's propagated by scan_control.
> > > > 
> > > > get_swappiness is static function but some planned updates will need to
> > > > get swappiness from files other than memcontrol.c
> > > > This patch exports get_swappiness() as mem_cgroup_swappiness().
> > > > By this, we can remove the argument of swapiness from try_to_free...
> > > > and drop swappiness from scan_control. only memcg uses it.
> > > > 
> > > 
> > > > +extern unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem);
> > > > +unsigned int mem_cgroup_swappiness(struct mem_cgroup *memcg)
> > > > +static int vmscan_swappiness(struct scan_control *sc)
> > > 
> > > The patch seems a bit confused about the signedness of swappiness.
> > > 
> > 
> > ok, v3 here. Now, memcg's one use "int" because vm_swapiness is "int".
> > ==
> > 
> > From af1bae8f2c6a8dbff048222bb45c7162b505f38b Mon Sep 17 00:00:00 2001
> > From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> > Date: Wed, 29 Jun 2011 18:24:49 +0900
> > Subject: [PATCH] export memory cgroup's swappines by mem_cgroup_swappiness()
> > 
> > Each memory cgroup has 'swappiness' value and it can be accessed by
> > get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
> > and swappiness is passed by argument. It's propagated by scan_control.
> > 
> > get_swappiness is static function but some planned updates will need to
> > get swappiness from files other than memcontrol.c
> > This patch exports get_swappiness() as mem_cgroup_swappiness().
> > By this, we can remove the argument of swapiness from try_to_free...
> > and drop swappiness from scan_control. only memcg uses it.
> > 
> > Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> 
> Reviewed-by: Michal Hocko <mhocko@suse.cz>
> 
> > 
> > Changelog:
> >   - adjusted signedness to vm_swappiness.
> >   - drop swappiness from scan_control
> > ---
> >  include/linux/swap.h |   10 +++++++---
> >  mm/memcontrol.c      |   15 +++++++--------
> >  mm/vmscan.c          |   23 ++++++++++-------------
> >  3 files changed, 24 insertions(+), 24 deletions(-)
> > 
> [...]
> > diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> > index 3e7d5e6..db70176 100644
> > --- a/mm/memcontrol.c
> > +++ b/mm/memcontrol.c
> [...]
> > @@ -4288,7 +4287,7 @@ static u64 mem_cgroup_swappiness_read(struct cgroup *cgrp, struct cftype *cft)
> >  {
> >  	struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp);
> >  
> > -	return get_swappiness(memcg);
> > +	return mem_cgroup_swappiness(memcg);
> >  }
> 
> If you want to be type clean you should change this one as well. I
> think it is worth it, though. The function is called only to return the
> current value to userspace and mem_cgroup_swappiness_write guaranties
> that it falls down into <0,100> interval. Additionally, cftype doesn't
> have any read specialization for int values so you would need to use a
> generic one. Finally if you changed read part you should change also
> write part and add > 0 check which is a lot of code for not that good
> reason.

I don't want to make this type-clean. Should I add type casting as
==
 return (u64) mem_cgroup_swappiness(memcg);
==
?

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] [Cleanup] memcg: export memory cgroup's swappiness v2

[Michal Hocko]

On Thu 30-06-11 16:10:39, KAMEZAWA Hiroyuki wrote:
> On Thu, 30 Jun 2011 08:32:32 +0200
> Michal Hocko <mhocko@suse.cz> wrote:
> 
> > On Thu 30-06-11 12:32:29, KAMEZAWA Hiroyuki wrote:
> > [...]
> > > @@ -4288,7 +4287,7 @@ static u64 mem_cgroup_swappiness_read(struct cgroup *cgrp, struct cftype *cft)
> > >  {
> > >  	struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp);
> > >  
> > > -	return get_swappiness(memcg);
> > > +	return mem_cgroup_swappiness(memcg);
> > >  }
> > 
> > If you want to be type clean you should change this one as well. I
> > think it is worth it, though. The function is called only to return the
> > current value to userspace and mem_cgroup_swappiness_write guaranties
> > that it falls down into <0,100> interval. Additionally, cftype doesn't
> > have any read specialization for int values so you would need to use a
> > generic one. Finally if you changed read part you should change also
> > write part and add > 0 check which is a lot of code for not that good
> > reason.
> 
> I don't want to make this type-clean. 

Agreed.

> Should I add type casting as
> ==
>  return (u64) mem_cgroup_swappiness(memcg);
> ==
> ?

I do not think it is necessary. The value is promoted to u64
automatically.

-- 
Michal Hocko
SUSE Labs
SUSE LINUX s.r.o.
Lihovarska 1060/12
190 00 Praha 9    
Czech Republic

Re: [PATCH] [Cleanup] memcg: export memory cgroup's swappiness v2

[Michal Hocko]

On Thu 30-06-11 16:10:39, KAMEZAWA Hiroyuki wrote:
> On Thu, 30 Jun 2011 08:32:32 +0200
> Michal Hocko <mhocko@suse.cz> wrote:
> 
> > On Thu 30-06-11 12:32:29, KAMEZAWA Hiroyuki wrote:
> > [...]
> > > @@ -4288,7 +4287,7 @@ static u64 mem_cgroup_swappiness_read(struct cgroup *cgrp, struct cftype *cft)
> > >  {
> > >  	struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp);
> > >  
> > > -	return get_swappiness(memcg);
> > > +	return mem_cgroup_swappiness(memcg);
> > >  }
> > 
> > If you want to be type clean you should change this one as well. I
> > think it is worth it, though. The function is called only to return the
> > current value to userspace and mem_cgroup_swappiness_write guaranties
> > that it falls down into <0,100> interval. Additionally, cftype doesn't
> > have any read specialization for int values so you would need to use a
> > generic one. Finally if you changed read part you should change also
> > write part and add > 0 check which is a lot of code for not that good
> > reason.
> 
> I don't want to make this type-clean. 

Agreed.

> Should I add type casting as
> ==
>  return (u64) mem_cgroup_swappiness(memcg);
> ==
> ?

I do not think it is necessary. The value is promoted to u64
automatically.

-- 
Michal Hocko
SUSE Labs
SUSE LINUX s.r.o.
Lihovarska 1060/12
190 00 Praha 9    
Czech Republic

--
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] [Cleanup] memcg: export memory cgroup's swappiness v2

[Andrew Morton]

On Wed, 29 Jun 2011 19:03:25 +0900
KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:

> Each memory cgroup has 'swappiness' value and it can be accessed by
> get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
> and swappiness is passed by argument. It's propagated by scan_control.
> 
> get_swappiness is static function but some planned updates will need to
> get swappiness from files other than memcontrol.c
> This patch exports get_swappiness() as mem_cgroup_swappiness().
> By this, we can remove the argument of swapiness from try_to_free...
> and drop swappiness from scan_control. only memcg uses it.

x86_64 allnoconfig (aka Documentation/SubmitChecklist, section 2b):

mm/vmscan.c: In function 'vmscan_swappiness':
mm/vmscan.c:1734: error: implicit declaration of function 'mem_cgroup_swappiness'

This is pretty broken.  I think we do want to implement this for
CONFIG_CGROUP_MEM_RES_CTLR=y, CONFIG_SWAP=n:

--- a/include/linux/swap.h~memcg-export-memory-cgroups-swappiness-fix
+++ a/include/linux/swap.h
@@ -365,17 +365,12 @@ static inline void put_swap_token(struct
 extern void
 mem_cgroup_uncharge_swapcache(struct page *page, swp_entry_t ent, bool swapout);
 extern int mem_cgroup_count_swap_user(swp_entry_t ent, struct page **pagep);
-extern unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem);
 #else
 static inline void
 mem_cgroup_uncharge_swapcache(struct page *page, swp_entry_t ent, bool swapout)
 {
 }
 
-static inline unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem)
-{
-	return vm_swappiness;
-}
 #endif
 #ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP
 extern void mem_cgroup_uncharge_swap(swp_entry_t ent);
@@ -510,5 +505,15 @@ mem_cgroup_count_swap_user(swp_entry_t e
 #endif
 
 #endif /* CONFIG_SWAP */
+
+#ifdef CONFIG_CGROUP_MEM_RES_CTLR
+extern unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem);
+#else
+static inline unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem)
+{
+	return vm_swappiness;
+}
+#endif
+
 #endif /* __KERNEL__*/
 #endif /* _LINUX_SWAP_H */

But what is the role of CONFIG_CGROUP_MEM_RES_CTLR_SWAP?

And in the above circumstances, vmscan_swappiness() devolves into

static int vmscan_swappiness(struct scan_control *sc)
{
       if (scanning_global_lru(sc))
               return vm_swappiness;
       return vm_swappiness;
}

which I guess makes sense but seems a bit odd.

Anyway, my confidence level is low so I think I'll drop this patch. 
Please have a think about the interplay between
CONFIG_CGROUP_MEM_RES_CTLR, CONFIG_CGROUP_MEM_RES_CTLR_SWAP and
CONFIG_SWAP.

Re: [PATCH] [Cleanup] memcg: export memory cgroup's swappiness v2

[Andrew Morton]

On Wed, 29 Jun 2011 19:03:25 +0900
KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:

> Each memory cgroup has 'swappiness' value and it can be accessed by
> get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
> and swappiness is passed by argument. It's propagated by scan_control.
> 
> get_swappiness is static function but some planned updates will need to
> get swappiness from files other than memcontrol.c
> This patch exports get_swappiness() as mem_cgroup_swappiness().
> By this, we can remove the argument of swapiness from try_to_free...
> and drop swappiness from scan_control. only memcg uses it.

x86_64 allnoconfig (aka Documentation/SubmitChecklist, section 2b):

mm/vmscan.c: In function 'vmscan_swappiness':
mm/vmscan.c:1734: error: implicit declaration of function 'mem_cgroup_swappiness'

This is pretty broken.  I think we do want to implement this for
CONFIG_CGROUP_MEM_RES_CTLR=y, CONFIG_SWAP=n:

--- a/include/linux/swap.h~memcg-export-memory-cgroups-swappiness-fix
+++ a/include/linux/swap.h
@@ -365,17 +365,12 @@ static inline void put_swap_token(struct
 extern void
 mem_cgroup_uncharge_swapcache(struct page *page, swp_entry_t ent, bool swapout);
 extern int mem_cgroup_count_swap_user(swp_entry_t ent, struct page **pagep);
-extern unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem);
 #else
 static inline void
 mem_cgroup_uncharge_swapcache(struct page *page, swp_entry_t ent, bool swapout)
 {
 }
 
-static inline unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem)
-{
-	return vm_swappiness;
-}
 #endif
 #ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP
 extern void mem_cgroup_uncharge_swap(swp_entry_t ent);
@@ -510,5 +505,15 @@ mem_cgroup_count_swap_user(swp_entry_t e
 #endif
 
 #endif /* CONFIG_SWAP */
+
+#ifdef CONFIG_CGROUP_MEM_RES_CTLR
+extern unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem);
+#else
+static inline unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem)
+{
+	return vm_swappiness;
+}
+#endif
+
 #endif /* __KERNEL__*/
 #endif /* _LINUX_SWAP_H */

But what is the role of CONFIG_CGROUP_MEM_RES_CTLR_SWAP?

And in the above circumstances, vmscan_swappiness() devolves into

static int vmscan_swappiness(struct scan_control *sc)
{
       if (scanning_global_lru(sc))
               return vm_swappiness;
       return vm_swappiness;
}

which I guess makes sense but seems a bit odd.

Anyway, my confidence level is low so I think I'll drop this patch. 
Please have a think about the interplay between
CONFIG_CGROUP_MEM_RES_CTLR, CONFIG_CGROUP_MEM_RES_CTLR_SWAP and
CONFIG_SWAP.


--
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] [Cleanup] memcg: export memory cgroup's swappiness v2

[KAMEZAWA Hiroyuki]

On Thu, 30 Jun 2011 13:01:34 -0700
Andrew Morton <akpm@linux-foundation.org> wrote:

> On Wed, 29 Jun 2011 19:03:25 +0900
> KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> 
> > Each memory cgroup has 'swappiness' value and it can be accessed by
> > get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
> > and swappiness is passed by argument. It's propagated by scan_control.
> > 
> > get_swappiness is static function but some planned updates will need to
> > get swappiness from files other than memcontrol.c
> > This patch exports get_swappiness() as mem_cgroup_swappiness().
> > By this, we can remove the argument of swapiness from try_to_free...
> > and drop swappiness from scan_control. only memcg uses it.
> 
> x86_64 allnoconfig (aka Documentation/SubmitChecklist, section 2b):
> 
> mm/vmscan.c: In function 'vmscan_swappiness':
> mm/vmscan.c:1734: error: implicit declaration of function 'mem_cgroup_swappiness'
> 
> This is pretty broken.  I think we do want to implement this for
> CONFIG_CGROUP_MEM_RES_CTLR=y, CONFIG_SWAP=n:
> 
> --- a/include/linux/swap.h~memcg-export-memory-cgroups-swappiness-fix
> +++ a/include/linux/swap.h
> @@ -365,17 +365,12 @@ static inline void put_swap_token(struct
>  extern void
>  mem_cgroup_uncharge_swapcache(struct page *page, swp_entry_t ent, bool swapout);
>  extern int mem_cgroup_count_swap_user(swp_entry_t ent, struct page **pagep);
> -extern unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem);
>  #else
>  static inline void
>  mem_cgroup_uncharge_swapcache(struct page *page, swp_entry_t ent, bool swapout)
>  {
>  }
>  
> -static inline unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem)
> -{
> -	return vm_swappiness;
> -}
>  #endif
>  #ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP
>  extern void mem_cgroup_uncharge_swap(swp_entry_t ent);
> @@ -510,5 +505,15 @@ mem_cgroup_count_swap_user(swp_entry_t e
>  #endif
>  
>  #endif /* CONFIG_SWAP */
> +
> +#ifdef CONFIG_CGROUP_MEM_RES_CTLR
> +extern unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem);
> +#else
> +static inline unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem)
> +{
> +	return vm_swappiness;
> +}
> +#endif
> +
>  #endif /* __KERNEL__*/
>  #endif /* _LINUX_SWAP_H */
> 
> But what is the role of CONFIG_CGROUP_MEM_RES_CTLR_SWAP?
> 
> And in the above circumstances, vmscan_swappiness() devolves into
> 
> static int vmscan_swappiness(struct scan_control *sc)
> {
>        if (scanning_global_lru(sc))
>                return vm_swappiness;
>        return vm_swappiness;
> }
> 
> which I guess makes sense but seems a bit odd.
> 
> Anyway, my confidence level is low so I think I'll drop this patch. 
> Please have a think about the interplay between
> CONFIG_CGROUP_MEM_RES_CTLR, CONFIG_CGROUP_MEM_RES_CTLR_SWAP and
> CONFIG_SWAP.
> 
> 

Ok, I'll check it. Maybe I miss !CONFIG_SWAP...

Thanks,
-Kame

Re: [PATCH] [Cleanup] memcg: export memory cgroup's swappiness v2

[KAMEZAWA Hiroyuki]

On Thu, 30 Jun 2011 13:01:34 -0700
Andrew Morton <akpm@linux-foundation.org> wrote:

> On Wed, 29 Jun 2011 19:03:25 +0900
> KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> 
> > Each memory cgroup has 'swappiness' value and it can be accessed by
> > get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
> > and swappiness is passed by argument. It's propagated by scan_control.
> > 
> > get_swappiness is static function but some planned updates will need to
> > get swappiness from files other than memcontrol.c
> > This patch exports get_swappiness() as mem_cgroup_swappiness().
> > By this, we can remove the argument of swapiness from try_to_free...
> > and drop swappiness from scan_control. only memcg uses it.
> 
> x86_64 allnoconfig (aka Documentation/SubmitChecklist, section 2b):
> 
> mm/vmscan.c: In function 'vmscan_swappiness':
> mm/vmscan.c:1734: error: implicit declaration of function 'mem_cgroup_swappiness'
> 
> This is pretty broken.  I think we do want to implement this for
> CONFIG_CGROUP_MEM_RES_CTLR=y, CONFIG_SWAP=n:
> 
> --- a/include/linux/swap.h~memcg-export-memory-cgroups-swappiness-fix
> +++ a/include/linux/swap.h
> @@ -365,17 +365,12 @@ static inline void put_swap_token(struct
>  extern void
>  mem_cgroup_uncharge_swapcache(struct page *page, swp_entry_t ent, bool swapout);
>  extern int mem_cgroup_count_swap_user(swp_entry_t ent, struct page **pagep);
> -extern unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem);
>  #else
>  static inline void
>  mem_cgroup_uncharge_swapcache(struct page *page, swp_entry_t ent, bool swapout)
>  {
>  }
>  
> -static inline unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem)
> -{
> -	return vm_swappiness;
> -}
>  #endif
>  #ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP
>  extern void mem_cgroup_uncharge_swap(swp_entry_t ent);
> @@ -510,5 +505,15 @@ mem_cgroup_count_swap_user(swp_entry_t e
>  #endif
>  
>  #endif /* CONFIG_SWAP */
> +
> +#ifdef CONFIG_CGROUP_MEM_RES_CTLR
> +extern unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem);
> +#else
> +static inline unsigned int mem_cgroup_swappiness(struct mem_cgroup *mem)
> +{
> +	return vm_swappiness;
> +}
> +#endif
> +
>  #endif /* __KERNEL__*/
>  #endif /* _LINUX_SWAP_H */
> 
> But what is the role of CONFIG_CGROUP_MEM_RES_CTLR_SWAP?
> 
> And in the above circumstances, vmscan_swappiness() devolves into
> 
> static int vmscan_swappiness(struct scan_control *sc)
> {
>        if (scanning_global_lru(sc))
>                return vm_swappiness;
>        return vm_swappiness;
> }
> 
> which I guess makes sense but seems a bit odd.
> 
> Anyway, my confidence level is low so I think I'll drop this patch. 
> Please have a think about the interplay between
> CONFIG_CGROUP_MEM_RES_CTLR, CONFIG_CGROUP_MEM_RES_CTLR_SWAP and
> CONFIG_SWAP.
> 
> 

Ok, I'll check it. Maybe I miss !CONFIG_SWAP...

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] [Cleanup] memcg: export memory cgroup's swappiness v2

[KAMEZAWA Hiroyuki]

On Fri, 1 Jul 2011 08:50:13 +0900
KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:

> On Thu, 30 Jun 2011 13:01:34 -0700
> Andrew Morton <akpm@linux-foundation.org> wrote:

> Ok, I'll check it. Maybe I miss !CONFIG_SWAP...
> 

v4 here. Thank you for pointing out. I could think of several ways but
maybe this one is good because using vm_swappines with !CONFIG_SWAP seems
to be a bug.
tested with allyesconfig/allnoconfig.
==
>From 0fb6905c288a16e99ee739f65cb486451e840637 Mon Sep 17 00:00:00 2001
From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Date: Wed, 29 Jun 2011 18:24:49 +0900
Subject: [PATCH] export memory cgroup's swappines by mem_cgroup_swappiness()

Each memory cgroup has 'swappiness' value and it can be accessed by
get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
and swappiness is passed by argument. It's propagated by scan_control.

get_swappiness is static function but some planned updates will need to
get swappiness from files other than memcontrol.c
This patch exports get_swappiness() as mem_cgroup_swappiness().
By this, we can remove the argument of swapiness from try_to_free...
and drop swappiness from scan_control. only memcg uses it.

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

Changelog:
  - fixed/tested allyesconfig/allnoconfig compile failure.
  - adjusted signedness to vm_swappiness.
  - drop swappiness from scan_control
---
 include/linux/swap.h |   10 +++++++---
 mm/memcontrol.c      |   15 +++++++--------
 mm/vmscan.c          |   33 ++++++++++++++++++++-------------
 3 files changed, 34 insertions(+), 24 deletions(-)

diff --git a/include/linux/swap.h b/include/linux/swap.h
index a273468..0299832 100644
--- a/include/linux/swap.h
+++ b/include/linux/swap.h
@@ -252,11 +252,9 @@ static inline void lru_cache_add_file(struct page *page)
 extern unsigned long try_to_free_pages(struct zonelist *zonelist, int order,
 					gfp_t gfp_mask, nodemask_t *mask);
 extern unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem,
-						  gfp_t gfp_mask, bool noswap,
-						  unsigned int swappiness);
+						  gfp_t gfp_mask, bool noswap);
 extern unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
 						gfp_t gfp_mask, bool noswap,
-						unsigned int swappiness,
 						struct zone *zone,
 						unsigned long *nr_scanned);
 extern int __isolate_lru_page(struct page *page, int mode, int file);
@@ -365,11 +363,17 @@ static inline void put_swap_token(struct mm_struct *mm)
 extern void
 mem_cgroup_uncharge_swapcache(struct page *page, swp_entry_t ent, bool swapout);
 extern int mem_cgroup_count_swap_user(swp_entry_t ent, struct page **pagep);
+extern int mem_cgroup_swappiness(struct mem_cgroup *mem);
 #else
 static inline void
 mem_cgroup_uncharge_swapcache(struct page *page, swp_entry_t ent, bool swapout)
 {
 }
+
+static inline int mem_cgroup_swappiness(struct mem_cgroup *mem)
+{
+	return vm_swappiness;
+}
 #endif
 #ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP
 extern void mem_cgroup_uncharge_swap(swp_entry_t ent);
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 3e7d5e6..db70176 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -249,7 +249,7 @@ struct mem_cgroup {
 	atomic_t	oom_lock;
 	atomic_t	refcnt;
 
-	unsigned int	swappiness;
+	int	swappiness;
 	/* OOM-Killer disable */
 	int		oom_kill_disable;
 
@@ -1330,7 +1330,7 @@ static unsigned long mem_cgroup_margin(struct mem_cgroup *mem)
 	return margin >> PAGE_SHIFT;
 }
 
-static unsigned int get_swappiness(struct mem_cgroup *memcg)
+int mem_cgroup_swappiness(struct mem_cgroup *memcg)
 {
 	struct cgroup *cgrp = memcg->css.cgroup;
 
@@ -1776,12 +1776,11 @@ static int mem_cgroup_hierarchical_reclaim(struct mem_cgroup *root_mem,
 		/* we use swappiness of local cgroup */
 		if (check_soft) {
 			ret = mem_cgroup_shrink_node_zone(victim, gfp_mask,
-				noswap, get_swappiness(victim), zone,
-				&nr_scanned);
+				noswap, zone, &nr_scanned);
 			*total_scanned += nr_scanned;
 		} else
 			ret = try_to_free_mem_cgroup_pages(victim, gfp_mask,
-						noswap, get_swappiness(victim));
+						noswap);
 		css_put(&victim->css);
 		/*
 		 * At shrinking usage, we can't check we should stop here or
@@ -3826,7 +3825,7 @@ try_to_free:
 			goto out;
 		}
 		progress = try_to_free_mem_cgroup_pages(mem, GFP_KERNEL,
-						false, get_swappiness(mem));
+						false);
 		if (!progress) {
 			nr_retries--;
 			/* maybe some writeback is necessary */
@@ -4288,7 +4287,7 @@ static u64 mem_cgroup_swappiness_read(struct cgroup *cgrp, struct cftype *cft)
 {
 	struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp);
 
-	return get_swappiness(memcg);
+	return mem_cgroup_swappiness(memcg);
 }
 
 static int mem_cgroup_swappiness_write(struct cgroup *cgrp, struct cftype *cft,
@@ -4997,7 +4996,7 @@ mem_cgroup_create(struct cgroup_subsys *ss, struct cgroup *cont)
 	INIT_LIST_HEAD(&mem->oom_notify);
 
 	if (parent)
-		mem->swappiness = get_swappiness(parent);
+		mem->swappiness = mem_cgroup_swappiness(parent);
 	atomic_set(&mem->refcnt, 1);
 	mem->move_charge_at_immigrate = 0;
 	mutex_init(&mem->thresholds_lock);
diff --git a/mm/vmscan.c b/mm/vmscan.c
index 4f49535..3d1d733 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -95,8 +95,6 @@ struct scan_control {
 	/* Can pages be swapped as part of reclaim? */
 	int may_swap;
 
-	int swappiness;
-
 	int order;
 
 	/*
@@ -1729,6 +1727,23 @@ static unsigned long shrink_list(enum lru_list lru, unsigned long nr_to_scan,
 	return shrink_inactive_list(nr_to_scan, zone, sc, priority, file);
 }
 
+#ifdef CONFIG_SWAP
+static int vmscan_swappiness(struct scan_control *sc)
+{
+	if (scanning_global_lru(sc))
+		return vm_swappiness;
+	return mem_cgroup_swappiness(sc->mem_cgroup);
+}
+#else
+static int vmscan_swappiness(struct scan_control *sc)
+{
+	/* Now, this function is never called with !CONFIG_SWAP */
+	BUG();
+	return 0;
+}
+#endif
+
+
 /*
  * Determine how aggressively the anon and file LRU lists should be
  * scanned.  The relative value of each set of LRU lists is determined
@@ -1789,8 +1804,8 @@ static void get_scan_count(struct zone *zone, struct scan_control *sc,
 	 * With swappiness at 100, anonymous and file have the same priority.
 	 * This scanning priority is essentially the inverse of IO cost.
 	 */
-	anon_prio = sc->swappiness;
-	file_prio = 200 - sc->swappiness;
+	anon_prio = vmscan_swappiness(sc);
+	file_prio = 200 - vmscan_swappiness(sc);
 
 	/*
 	 * OK, so we have swap space and a fair amount of page cache
@@ -2179,7 +2194,6 @@ unsigned long try_to_free_pages(struct zonelist *zonelist, int order,
 		.nr_to_reclaim = SWAP_CLUSTER_MAX,
 		.may_unmap = 1,
 		.may_swap = 1,
-		.swappiness = vm_swappiness,
 		.order = order,
 		.mem_cgroup = NULL,
 		.nodemask = nodemask,
@@ -2203,7 +2217,6 @@ 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,
 						unsigned long *nr_scanned)
 {
@@ -2213,7 +2226,6 @@ unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
 		.may_writepage = !laptop_mode,
 		.may_unmap = 1,
 		.may_swap = !noswap,
-		.swappiness = swappiness,
 		.order = 0,
 		.mem_cgroup = mem,
 	};
@@ -2242,8 +2254,7 @@ unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
 
 unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem_cont,
 					   gfp_t gfp_mask,
-					   bool noswap,
-					   unsigned int swappiness)
+					   bool noswap)
 {
 	struct zonelist *zonelist;
 	unsigned long nr_reclaimed;
@@ -2253,7 +2264,6 @@ unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem_cont,
 		.may_unmap = 1,
 		.may_swap = !noswap,
 		.nr_to_reclaim = SWAP_CLUSTER_MAX,
-		.swappiness = swappiness,
 		.order = 0,
 		.mem_cgroup = mem_cont,
 		.nodemask = NULL, /* we don't care the placement */
@@ -2403,7 +2413,6 @@ static unsigned long balance_pgdat(pg_data_t *pgdat, int order,
 		 * we want to put equal scanning pressure on each zone.
 		 */
 		.nr_to_reclaim = ULONG_MAX,
-		.swappiness = vm_swappiness,
 		.order = order,
 		.mem_cgroup = NULL,
 	};
@@ -2862,7 +2871,6 @@ unsigned long shrink_all_memory(unsigned long nr_to_reclaim)
 		.may_writepage = 1,
 		.nr_to_reclaim = nr_to_reclaim,
 		.hibernation_mode = 1,
-		.swappiness = vm_swappiness,
 		.order = 0,
 	};
 	struct shrink_control shrink = {
@@ -3049,7 +3057,6 @@ static int __zone_reclaim(struct zone *zone, gfp_t gfp_mask, unsigned int order)
 		.nr_to_reclaim = max_t(unsigned long, nr_pages,
 				       SWAP_CLUSTER_MAX),
 		.gfp_mask = gfp_mask,
-		.swappiness = vm_swappiness,
 		.order = order,
 	};
 	struct shrink_control shrink = {
-- 
1.7.4.1

Re: [PATCH] [Cleanup] memcg: export memory cgroup's swappiness v2

[KAMEZAWA Hiroyuki]

On Fri, 1 Jul 2011 08:50:13 +0900
KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:

> On Thu, 30 Jun 2011 13:01:34 -0700
> Andrew Morton <akpm@linux-foundation.org> wrote:

> Ok, I'll check it. Maybe I miss !CONFIG_SWAP...
> 

v4 here. Thank you for pointing out. I could think of several ways but
maybe this one is good because using vm_swappines with !CONFIG_SWAP seems
to be a bug.
tested with allyesconfig/allnoconfig.
==

Re: [PATCH] [Cleanup] memcg: export memory cgroup's swappiness v2

[Andrew Morton]

On Fri, 1 Jul 2011 09:20:59 +0900 KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:

> On Fri, 1 Jul 2011 08:50:13 +0900
> KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> 
> > On Thu, 30 Jun 2011 13:01:34 -0700
> > Andrew Morton <akpm@linux-foundation.org> wrote:
> 
> > Ok, I'll check it. Maybe I miss !CONFIG_SWAP...
> > 
> 
> v4 here. Thank you for pointing out. I could think of several ways but
> maybe this one is good because using vm_swappines with !CONFIG_SWAP seems
> to be a bug.

No, it isn't a bug - swappiness also controls the kernel's eagerness to
unmap and reclaim mmapped pagecache.

> tested with allyesconfig/allnoconfig.

Did it break the above?

> +#ifdef CONFIG_SWAP
> +static int vmscan_swappiness(struct scan_control *sc)
> +{
> +	if (scanning_global_lru(sc))
> +		return vm_swappiness;

Well that's a bit ugly - it assumes that all callers set
scan_control.swappiness to vm_swappiness then never change it.  That
may be true in the current code.

Ho hum, I guess that's a simplification we can make.

> +	return mem_cgroup_swappiness(sc->mem_cgroup);
> +}
> +#else
> +static int vmscan_swappiness(struct scan_control *sc)
> +{
> +	/* Now, this function is never called with !CONFIG_SWAP */
> +	BUG();
> +	return 0;
> +}
> +#endif
>
> ...
>
> @@ -1789,8 +1804,8 @@ static void get_scan_count(struct zone *zone, struct scan_control *sc,
>  	 * With swappiness at 100, anonymous and file have the same priority.
>  	 * This scanning priority is essentially the inverse of IO cost.
>  	 */
> -	anon_prio = sc->swappiness;
> -	file_prio = 200 - sc->swappiness;
> +	anon_prio = vmscan_swappiness(sc);
> +	file_prio = 200 - vmscan_swappiness(sc);

hah, this should go BUG if CONFIG_SWAP=n.  But it won't, because we
broke get_scan_count().  It fails to apply vm_swappiness to file-backed
pages if there's no available swap, which is daft.

I think this happened in 76a33fc380c9a ("vmscan: prevent
get_scan_ratio() rounding errors") which claims "this patch doesn't
really change logics, but just increase precision".

Re: [PATCH] [Cleanup] memcg: export memory cgroup's swappiness v2

[Andrew Morton]

On Fri, 1 Jul 2011 09:20:59 +0900 KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:

> On Fri, 1 Jul 2011 08:50:13 +0900
> KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> 
> > On Thu, 30 Jun 2011 13:01:34 -0700
> > Andrew Morton <akpm@linux-foundation.org> wrote:
> 
> > Ok, I'll check it. Maybe I miss !CONFIG_SWAP...
> > 
> 
> v4 here. Thank you for pointing out. I could think of several ways but
> maybe this one is good because using vm_swappines with !CONFIG_SWAP seems
> to be a bug.

No, it isn't a bug - swappiness also controls the kernel's eagerness to
unmap and reclaim mmapped pagecache.

> tested with allyesconfig/allnoconfig.

Did it break the above?

> +#ifdef CONFIG_SWAP
> +static int vmscan_swappiness(struct scan_control *sc)
> +{
> +	if (scanning_global_lru(sc))
> +		return vm_swappiness;

Well that's a bit ugly - it assumes that all callers set
scan_control.swappiness to vm_swappiness then never change it.  That
may be true in the current code.

Ho hum, I guess that's a simplification we can make.

> +	return mem_cgroup_swappiness(sc->mem_cgroup);
> +}
> +#else
> +static int vmscan_swappiness(struct scan_control *sc)
> +{
> +	/* Now, this function is never called with !CONFIG_SWAP */
> +	BUG();
> +	return 0;
> +}
> +#endif
>
> ...
>
> @@ -1789,8 +1804,8 @@ static void get_scan_count(struct zone *zone, struct scan_control *sc,
>  	 * With swappiness at 100, anonymous and file have the same priority.
>  	 * This scanning priority is essentially the inverse of IO cost.
>  	 */
> -	anon_prio = sc->swappiness;
> -	file_prio = 200 - sc->swappiness;
> +	anon_prio = vmscan_swappiness(sc);
> +	file_prio = 200 - vmscan_swappiness(sc);

hah, this should go BUG if CONFIG_SWAP=n.  But it won't, because we
broke get_scan_count().  It fails to apply vm_swappiness to file-backed
pages if there's no available swap, which is daft.

I think this happened in 76a33fc380c9a ("vmscan: prevent
get_scan_ratio() rounding errors") which claims "this patch doesn't
really change logics, but just increase precision".


--
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] [Cleanup] memcg: export memory cgroup's swappiness v2

[KAMEZAWA Hiroyuki]

On Thu, 30 Jun 2011 18:06:53 -0700
Andrew Morton <akpm@linux-foundation.org> wrote:

> On Fri, 1 Jul 2011 09:20:59 +0900 KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> 
> > On Fri, 1 Jul 2011 08:50:13 +0900
> > KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> > 
> > > On Thu, 30 Jun 2011 13:01:34 -0700
> > > Andrew Morton <akpm@linux-foundation.org> wrote:
> > 
> > > Ok, I'll check it. Maybe I miss !CONFIG_SWAP...
> > > 
> > 
> > v4 here. Thank you for pointing out. I could think of several ways but
> > maybe this one is good because using vm_swappines with !CONFIG_SWAP seems
> > to be a bug.
> 
> No, it isn't a bug - swappiness also controls the kernel's eagerness to
> unmap and reclaim mmapped pagecache.
> 

Oh, really ? I didn't understand that.


> > tested with allyesconfig/allnoconfig.
> 
> Did it break the above?
> 

Hmm. If !CONFIG_SWAP, get_scan_count() will see !nr_swap_pages and
set scan ratio as
  file: 100%
  anon: 0%



> > +#ifdef CONFIG_SWAP
> > +static int vmscan_swappiness(struct scan_control *sc)
> > +{
> > +	if (scanning_global_lru(sc))
> > +		return vm_swappiness;
> 
> Well that's a bit ugly - it assumes that all callers set
> scan_control.swappiness to vm_swappiness then never change it.  That
> may be true in the current code.
> 
> Ho hum, I guess that's a simplification we can make.
> 

We don't calculate kernel internal swappiness and just use vm_swappines
which the user specified. So, I thought it should not be in scan_control.



> > +	return mem_cgroup_swappiness(sc->mem_cgroup);
> > +}
> > +#else
> > +static int vmscan_swappiness(struct scan_control *sc)
> > +{
> > +	/* Now, this function is never called with !CONFIG_SWAP */
> > +	BUG();
> > +	return 0;
> > +}
> > +#endif
> >
> > ...
> >
> > @@ -1789,8 +1804,8 @@ static void get_scan_count(struct zone *zone, struct scan_control *sc,
> >  	 * With swappiness at 100, anonymous and file have the same priority.
> >  	 * This scanning priority is essentially the inverse of IO cost.
> >  	 */
> > -	anon_prio = sc->swappiness;
> > -	file_prio = 200 - sc->swappiness;
> > +	anon_prio = vmscan_swappiness(sc);
> > +	file_prio = 200 - vmscan_swappiness(sc);
> 
> hah, this should go BUG if CONFIG_SWAP=n.  But it won't, because we
> broke get_scan_count().  It fails to apply vm_swappiness to file-backed
> pages if there's no available swap, which is daft.
> 
> I think this happened in 76a33fc380c9a ("vmscan: prevent
> get_scan_ratio() rounding errors") which claims "this patch doesn't
> really change logics, but just increase precision".
> 

Hmm, IIUC.
  - the controller of unmapping file cache is now sc->may_unmap
    - may_unmap is always set 1 unless zone_reclaim_mode.
    - vm_swappiness doesn't affect it now. 
  - file LRU contains both mapped and unmapped pages in the same manner
    - get_scan_count() cannot help decisiion of "map or unmap"
    -  Active/Inactive scan ratio is now determined by reclaim_stat.

Hmm, swappiness should affect active/inactive scan ratio ?

Thanks,
-Kame

Re: [PATCH] [Cleanup] memcg: export memory cgroup's swappiness v2

[KAMEZAWA Hiroyuki]

On Thu, 30 Jun 2011 18:06:53 -0700
Andrew Morton <akpm@linux-foundation.org> wrote:

> On Fri, 1 Jul 2011 09:20:59 +0900 KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> 
> > On Fri, 1 Jul 2011 08:50:13 +0900
> > KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> > 
> > > On Thu, 30 Jun 2011 13:01:34 -0700
> > > Andrew Morton <akpm@linux-foundation.org> wrote:
> > 
> > > Ok, I'll check it. Maybe I miss !CONFIG_SWAP...
> > > 
> > 
> > v4 here. Thank you for pointing out. I could think of several ways but
> > maybe this one is good because using vm_swappines with !CONFIG_SWAP seems
> > to be a bug.
> 
> No, it isn't a bug - swappiness also controls the kernel's eagerness to
> unmap and reclaim mmapped pagecache.
> 

Oh, really ? I didn't understand that.


> > tested with allyesconfig/allnoconfig.
> 
> Did it break the above?
> 

Hmm. If !CONFIG_SWAP, get_scan_count() will see !nr_swap_pages and
set scan ratio as
  file: 100%
  anon: 0%



> > +#ifdef CONFIG_SWAP
> > +static int vmscan_swappiness(struct scan_control *sc)
> > +{
> > +	if (scanning_global_lru(sc))
> > +		return vm_swappiness;
> 
> Well that's a bit ugly - it assumes that all callers set
> scan_control.swappiness to vm_swappiness then never change it.  That
> may be true in the current code.
> 
> Ho hum, I guess that's a simplification we can make.
> 

We don't calculate kernel internal swappiness and just use vm_swappines
which the user specified. So, I thought it should not be in scan_control.



> > +	return mem_cgroup_swappiness(sc->mem_cgroup);
> > +}
> > +#else
> > +static int vmscan_swappiness(struct scan_control *sc)
> > +{
> > +	/* Now, this function is never called with !CONFIG_SWAP */
> > +	BUG();
> > +	return 0;
> > +}
> > +#endif
> >
> > ...
> >
> > @@ -1789,8 +1804,8 @@ static void get_scan_count(struct zone *zone, struct scan_control *sc,
> >  	 * With swappiness at 100, anonymous and file have the same priority.
> >  	 * This scanning priority is essentially the inverse of IO cost.
> >  	 */
> > -	anon_prio = sc->swappiness;
> > -	file_prio = 200 - sc->swappiness;
> > +	anon_prio = vmscan_swappiness(sc);
> > +	file_prio = 200 - vmscan_swappiness(sc);
> 
> hah, this should go BUG if CONFIG_SWAP=n.  But it won't, because we
> broke get_scan_count().  It fails to apply vm_swappiness to file-backed
> pages if there's no available swap, which is daft.
> 
> I think this happened in 76a33fc380c9a ("vmscan: prevent
> get_scan_ratio() rounding errors") which claims "this patch doesn't
> really change logics, but just increase precision".
> 

Hmm, IIUC.
  - the controller of unmapping file cache is now sc->may_unmap
    - may_unmap is always set 1 unless zone_reclaim_mode.
    - vm_swappiness doesn't affect it now. 
  - file LRU contains both mapped and unmapped pages in the same manner
    - get_scan_count() cannot help decisiion of "map or unmap"
    -  Active/Inactive scan ratio is now determined by reclaim_stat.

Hmm, swappiness should affect active/inactive scan ratio ?

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] [Cleanup] memcg: export memory cgroup's swappiness v2

[KAMEZAWA Hiroyuki]

On Fri, 1 Jul 2011 10:16:24 +0900
KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:

> On Thu, 30 Jun 2011 18:06:53 -0700
> Andrew Morton <akpm@linux-foundation.org> wrote:
> 
> > On Fri, 1 Jul 2011 09:20:59 +0900 KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> > 
> > > On Fri, 1 Jul 2011 08:50:13 +0900
> > > KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> > > 
> > > > On Thu, 30 Jun 2011 13:01:34 -0700
> > > > Andrew Morton <akpm@linux-foundation.org> wrote:
> > > 
> > > > Ok, I'll check it. Maybe I miss !CONFIG_SWAP...
> > > > 
> > > 
> > > v4 here. Thank you for pointing out. I could think of several ways but
> > > maybe this one is good because using vm_swappines with !CONFIG_SWAP seems
> > > to be a bug.
> > 
> > No, it isn't a bug - swappiness also controls the kernel's eagerness to
> > unmap and reclaim mmapped pagecache.
> > 
> 
> Oh, really ? I didn't understand that.
> 
Hmm, anyway, this new version of fix seems better.
==
>From 7daf93a277e19026bb6edef3e0ac01bbd31dcb5e Mon Sep 17 00:00:00 2001
From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Date: Fri, 1 Jul 2011 10:35:57 +0900
Subject: [PATCH] export memory cgroup's swappines by mem_cgroup_swappiness()

Each memory cgroup has 'swappiness' value and it can be accessed by
get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
and swappiness is passed by argument. It's propagated by scan_control.

get_swappiness is static function but some planned updates will need to
get swappiness from files other than memcontrol.c
This patch exports get_swappiness() as mem_cgroup_swappiness().
By this, we can remove the argument of swapiness from try_to_free...
and drop swappiness from scan_control. only memcg uses it.

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

Changelog:
  - move definistions out of CONFIG_SWAP.
  - fixed/tested allyesconfig/allnoconfig compile failure.
  - adjusted signedness to vm_swappiness.
  - drop swappiness from scan_control
---
 include/linux/swap.h |   13 +++++++++----
 mm/memcontrol.c      |   15 +++++++--------
 mm/vmscan.c          |   23 ++++++++++-------------
 3 files changed, 26 insertions(+), 25 deletions(-)

diff --git a/include/linux/swap.h b/include/linux/swap.h
index a273468..28f1490 100644
--- a/include/linux/swap.h
+++ b/include/linux/swap.h
@@ -252,11 +252,9 @@ static inline void lru_cache_add_file(struct page *page)
 extern unsigned long try_to_free_pages(struct zonelist *zonelist, int order,
 					gfp_t gfp_mask, nodemask_t *mask);
 extern unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem,
-						  gfp_t gfp_mask, bool noswap,
-						  unsigned int swappiness);
+						  gfp_t gfp_mask, bool noswap);
 extern unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
 						gfp_t gfp_mask, bool noswap,
-						unsigned int swappiness,
 						struct zone *zone,
 						unsigned long *nr_scanned);
 extern int __isolate_lru_page(struct page *page, int mode, int file);
@@ -299,7 +297,14 @@ static inline void scan_unevictable_unregister_node(struct node *node)
 
 extern int kswapd_run(int nid);
 extern void kswapd_stop(int nid);
-
+#ifdef CONFIG_MEM_CGROUP_MEM_RES_CTLR
+extern int mem_cgroup_swappiness(struct mem_cgroup *mem);
+#else
+static inline int mem_cgroup_swappiness(struct mem_cgroup *mem)
+{
+	return vm_swappiness;
+}
+#endif
 #ifdef CONFIG_SWAP
 /* linux/mm/page_io.c */
 extern int swap_readpage(struct page *);
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 3e7d5e6..db70176 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -249,7 +249,7 @@ struct mem_cgroup {
 	atomic_t	oom_lock;
 	atomic_t	refcnt;
 
-	unsigned int	swappiness;
+	int	swappiness;
 	/* OOM-Killer disable */
 	int		oom_kill_disable;
 
@@ -1330,7 +1330,7 @@ static unsigned long mem_cgroup_margin(struct mem_cgroup *mem)
 	return margin >> PAGE_SHIFT;
 }
 
-static unsigned int get_swappiness(struct mem_cgroup *memcg)
+int mem_cgroup_swappiness(struct mem_cgroup *memcg)
 {
 	struct cgroup *cgrp = memcg->css.cgroup;
 
@@ -1776,12 +1776,11 @@ static int mem_cgroup_hierarchical_reclaim(struct mem_cgroup *root_mem,
 		/* we use swappiness of local cgroup */
 		if (check_soft) {
 			ret = mem_cgroup_shrink_node_zone(victim, gfp_mask,
-				noswap, get_swappiness(victim), zone,
-				&nr_scanned);
+				noswap, zone, &nr_scanned);
 			*total_scanned += nr_scanned;
 		} else
 			ret = try_to_free_mem_cgroup_pages(victim, gfp_mask,
-						noswap, get_swappiness(victim));
+						noswap);
 		css_put(&victim->css);
 		/*
 		 * At shrinking usage, we can't check we should stop here or
@@ -3826,7 +3825,7 @@ try_to_free:
 			goto out;
 		}
 		progress = try_to_free_mem_cgroup_pages(mem, GFP_KERNEL,
-						false, get_swappiness(mem));
+						false);
 		if (!progress) {
 			nr_retries--;
 			/* maybe some writeback is necessary */
@@ -4288,7 +4287,7 @@ static u64 mem_cgroup_swappiness_read(struct cgroup *cgrp, struct cftype *cft)
 {
 	struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp);
 
-	return get_swappiness(memcg);
+	return mem_cgroup_swappiness(memcg);
 }
 
 static int mem_cgroup_swappiness_write(struct cgroup *cgrp, struct cftype *cft,
@@ -4997,7 +4996,7 @@ mem_cgroup_create(struct cgroup_subsys *ss, struct cgroup *cont)
 	INIT_LIST_HEAD(&mem->oom_notify);
 
 	if (parent)
-		mem->swappiness = get_swappiness(parent);
+		mem->swappiness = mem_cgroup_swappiness(parent);
 	atomic_set(&mem->refcnt, 1);
 	mem->move_charge_at_immigrate = 0;
 	mutex_init(&mem->thresholds_lock);
diff --git a/mm/vmscan.c b/mm/vmscan.c
index 4f49535..fb37699 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -95,8 +95,6 @@ struct scan_control {
 	/* Can pages be swapped as part of reclaim? */
 	int may_swap;
 
-	int swappiness;
-
 	int order;
 
 	/*
@@ -1729,6 +1727,13 @@ static unsigned long shrink_list(enum lru_list lru, unsigned long nr_to_scan,
 	return shrink_inactive_list(nr_to_scan, zone, sc, priority, file);
 }
 
+static int vmscan_swappiness(struct scan_control *sc)
+{
+	if (scanning_global_lru(sc))
+		return vm_swappiness;
+	return mem_cgroup_swappiness(sc->mem_cgroup);
+}
+
 /*
  * Determine how aggressively the anon and file LRU lists should be
  * scanned.  The relative value of each set of LRU lists is determined
@@ -1789,8 +1794,8 @@ static void get_scan_count(struct zone *zone, struct scan_control *sc,
 	 * With swappiness at 100, anonymous and file have the same priority.
 	 * This scanning priority is essentially the inverse of IO cost.
 	 */
-	anon_prio = sc->swappiness;
-	file_prio = 200 - sc->swappiness;
+	anon_prio = vmscan_swappiness(sc);
+	file_prio = 200 - vmscan_swappiness(sc);
 
 	/*
 	 * OK, so we have swap space and a fair amount of page cache
@@ -2179,7 +2184,6 @@ unsigned long try_to_free_pages(struct zonelist *zonelist, int order,
 		.nr_to_reclaim = SWAP_CLUSTER_MAX,
 		.may_unmap = 1,
 		.may_swap = 1,
-		.swappiness = vm_swappiness,
 		.order = order,
 		.mem_cgroup = NULL,
 		.nodemask = nodemask,
@@ -2203,7 +2207,6 @@ 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,
 						unsigned long *nr_scanned)
 {
@@ -2213,7 +2216,6 @@ unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
 		.may_writepage = !laptop_mode,
 		.may_unmap = 1,
 		.may_swap = !noswap,
-		.swappiness = swappiness,
 		.order = 0,
 		.mem_cgroup = mem,
 	};
@@ -2242,8 +2244,7 @@ unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
 
 unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem_cont,
 					   gfp_t gfp_mask,
-					   bool noswap,
-					   unsigned int swappiness)
+					   bool noswap)
 {
 	struct zonelist *zonelist;
 	unsigned long nr_reclaimed;
@@ -2253,7 +2254,6 @@ unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem_cont,
 		.may_unmap = 1,
 		.may_swap = !noswap,
 		.nr_to_reclaim = SWAP_CLUSTER_MAX,
-		.swappiness = swappiness,
 		.order = 0,
 		.mem_cgroup = mem_cont,
 		.nodemask = NULL, /* we don't care the placement */
@@ -2403,7 +2403,6 @@ static unsigned long balance_pgdat(pg_data_t *pgdat, int order,
 		 * we want to put equal scanning pressure on each zone.
 		 */
 		.nr_to_reclaim = ULONG_MAX,
-		.swappiness = vm_swappiness,
 		.order = order,
 		.mem_cgroup = NULL,
 	};
@@ -2862,7 +2861,6 @@ unsigned long shrink_all_memory(unsigned long nr_to_reclaim)
 		.may_writepage = 1,
 		.nr_to_reclaim = nr_to_reclaim,
 		.hibernation_mode = 1,
-		.swappiness = vm_swappiness,
 		.order = 0,
 	};
 	struct shrink_control shrink = {
@@ -3049,7 +3047,6 @@ static int __zone_reclaim(struct zone *zone, gfp_t gfp_mask, unsigned int order)
 		.nr_to_reclaim = max_t(unsigned long, nr_pages,
 				       SWAP_CLUSTER_MAX),
 		.gfp_mask = gfp_mask,
-		.swappiness = vm_swappiness,
 		.order = order,
 	};
 	struct shrink_control shrink = {
-- 
1.7.4.1

Re: [PATCH] [Cleanup] memcg: export memory cgroup's swappiness v2

[KAMEZAWA Hiroyuki]

On Fri, 1 Jul 2011 10:16:24 +0900
KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:

> On Thu, 30 Jun 2011 18:06:53 -0700
> Andrew Morton <akpm@linux-foundation.org> wrote:
> 
> > On Fri, 1 Jul 2011 09:20:59 +0900 KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> > 
> > > On Fri, 1 Jul 2011 08:50:13 +0900
> > > KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> > > 
> > > > On Thu, 30 Jun 2011 13:01:34 -0700
> > > > Andrew Morton <akpm@linux-foundation.org> wrote:
> > > 
> > > > Ok, I'll check it. Maybe I miss !CONFIG_SWAP...
> > > > 
> > > 
> > > v4 here. Thank you for pointing out. I could think of several ways but
> > > maybe this one is good because using vm_swappines with !CONFIG_SWAP seems
> > > to be a bug.
> > 
> > No, it isn't a bug - swappiness also controls the kernel's eagerness to
> > unmap and reclaim mmapped pagecache.
> > 
> 
> Oh, really ? I didn't understand that.
> 
Hmm, anyway, this new version of fix seems better.
==

Re: [PATCH] [Cleanup] memcg: export memory cgroup's swappiness v2

[KAMEZAWA Hiroyuki]

On Fri, 1 Jul 2011 10:30:07 +0900
KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:

> On Fri, 1 Jul 2011 10:16:24 +0900
> KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> 
> > On Thu, 30 Jun 2011 18:06:53 -0700
> > Andrew Morton <akpm@linux-foundation.org> wrote:
> > 
> > > On Fri, 1 Jul 2011 09:20:59 +0900 KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> > > 
> > > > On Fri, 1 Jul 2011 08:50:13 +0900
> > > > KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> > > > 
> > > > > On Thu, 30 Jun 2011 13:01:34 -0700
> > > > > Andrew Morton <akpm@linux-foundation.org> wrote:
> > > > 
> > > > > Ok, I'll check it. Maybe I miss !CONFIG_SWAP...
> > > > > 
> > > > 
> > > > v4 here. Thank you for pointing out. I could think of several ways but
> > > > maybe this one is good because using vm_swappines with !CONFIG_SWAP seems
> > > > to be a bug.
> > > 
> > > No, it isn't a bug - swappiness also controls the kernel's eagerness to
> > > unmap and reclaim mmapped pagecache.
> > > 
> > 
> > Oh, really ? I didn't understand that.
> > 
> Hmm, anyway, this new version of fix seems better.

Sorry, this seems still buggy. I'll send a new one in the next week :(

Thanks,
-Kame

> ==
> From 7daf93a277e19026bb6edef3e0ac01bbd31dcb5e Mon Sep 17 00:00:00 2001
> From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> Date: Fri, 1 Jul 2011 10:35:57 +0900
> Subject: [PATCH] export memory cgroup's swappines by mem_cgroup_swappiness()
> 
> Each memory cgroup has 'swappiness' value and it can be accessed by
> get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
> and swappiness is passed by argument. It's propagated by scan_control.
> 
> get_swappiness is static function but some planned updates will need to
> get swappiness from files other than memcontrol.c
> This patch exports get_swappiness() as mem_cgroup_swappiness().
> By this, we can remove the argument of swapiness from try_to_free...
> and drop swappiness from scan_control. only memcg uses it.
> 
> Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> 
> Changelog:
>   - move definistions out of CONFIG_SWAP.
>   - fixed/tested allyesconfig/allnoconfig compile failure.
>   - adjusted signedness to vm_swappiness.
>   - drop swappiness from scan_control
> ---
>  include/linux/swap.h |   13 +++++++++----
>  mm/memcontrol.c      |   15 +++++++--------
>  mm/vmscan.c          |   23 ++++++++++-------------
>  3 files changed, 26 insertions(+), 25 deletions(-)
> 
> diff --git a/include/linux/swap.h b/include/linux/swap.h
> index a273468..28f1490 100644
> --- a/include/linux/swap.h
> +++ b/include/linux/swap.h
> @@ -252,11 +252,9 @@ static inline void lru_cache_add_file(struct page *page)
>  extern unsigned long try_to_free_pages(struct zonelist *zonelist, int order,
>  					gfp_t gfp_mask, nodemask_t *mask);
>  extern unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem,
> -						  gfp_t gfp_mask, bool noswap,
> -						  unsigned int swappiness);
> +						  gfp_t gfp_mask, bool noswap);
>  extern unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
>  						gfp_t gfp_mask, bool noswap,
> -						unsigned int swappiness,
>  						struct zone *zone,
>  						unsigned long *nr_scanned);
>  extern int __isolate_lru_page(struct page *page, int mode, int file);
> @@ -299,7 +297,14 @@ static inline void scan_unevictable_unregister_node(struct node *node)
>  
>  extern int kswapd_run(int nid);
>  extern void kswapd_stop(int nid);
> -
> +#ifdef CONFIG_MEM_CGROUP_MEM_RES_CTLR
> +extern int mem_cgroup_swappiness(struct mem_cgroup *mem);
> +#else
> +static inline int mem_cgroup_swappiness(struct mem_cgroup *mem)
> +{
> +	return vm_swappiness;
> +}
> +#endif
>  #ifdef CONFIG_SWAP
>  /* linux/mm/page_io.c */
>  extern int swap_readpage(struct page *);
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index 3e7d5e6..db70176 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -249,7 +249,7 @@ struct mem_cgroup {
>  	atomic_t	oom_lock;
>  	atomic_t	refcnt;
>  
> -	unsigned int	swappiness;
> +	int	swappiness;
>  	/* OOM-Killer disable */
>  	int		oom_kill_disable;
>  
> @@ -1330,7 +1330,7 @@ static unsigned long mem_cgroup_margin(struct mem_cgroup *mem)
>  	return margin >> PAGE_SHIFT;
>  }
>  
> -static unsigned int get_swappiness(struct mem_cgroup *memcg)
> +int mem_cgroup_swappiness(struct mem_cgroup *memcg)
>  {
>  	struct cgroup *cgrp = memcg->css.cgroup;
>  
> @@ -1776,12 +1776,11 @@ static int mem_cgroup_hierarchical_reclaim(struct mem_cgroup *root_mem,
>  		/* we use swappiness of local cgroup */
>  		if (check_soft) {
>  			ret = mem_cgroup_shrink_node_zone(victim, gfp_mask,
> -				noswap, get_swappiness(victim), zone,
> -				&nr_scanned);
> +				noswap, zone, &nr_scanned);
>  			*total_scanned += nr_scanned;
>  		} else
>  			ret = try_to_free_mem_cgroup_pages(victim, gfp_mask,
> -						noswap, get_swappiness(victim));
> +						noswap);
>  		css_put(&victim->css);
>  		/*
>  		 * At shrinking usage, we can't check we should stop here or
> @@ -3826,7 +3825,7 @@ try_to_free:
>  			goto out;
>  		}
>  		progress = try_to_free_mem_cgroup_pages(mem, GFP_KERNEL,
> -						false, get_swappiness(mem));
> +						false);
>  		if (!progress) {
>  			nr_retries--;
>  			/* maybe some writeback is necessary */
> @@ -4288,7 +4287,7 @@ static u64 mem_cgroup_swappiness_read(struct cgroup *cgrp, struct cftype *cft)
>  {
>  	struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp);
>  
> -	return get_swappiness(memcg);
> +	return mem_cgroup_swappiness(memcg);
>  }
>  
>  static int mem_cgroup_swappiness_write(struct cgroup *cgrp, struct cftype *cft,
> @@ -4997,7 +4996,7 @@ mem_cgroup_create(struct cgroup_subsys *ss, struct cgroup *cont)
>  	INIT_LIST_HEAD(&mem->oom_notify);
>  
>  	if (parent)
> -		mem->swappiness = get_swappiness(parent);
> +		mem->swappiness = mem_cgroup_swappiness(parent);
>  	atomic_set(&mem->refcnt, 1);
>  	mem->move_charge_at_immigrate = 0;
>  	mutex_init(&mem->thresholds_lock);
> diff --git a/mm/vmscan.c b/mm/vmscan.c
> index 4f49535..fb37699 100644
> --- a/mm/vmscan.c
> +++ b/mm/vmscan.c
> @@ -95,8 +95,6 @@ struct scan_control {
>  	/* Can pages be swapped as part of reclaim? */
>  	int may_swap;
>  
> -	int swappiness;
> -
>  	int order;
>  
>  	/*
> @@ -1729,6 +1727,13 @@ static unsigned long shrink_list(enum lru_list lru, unsigned long nr_to_scan,
>  	return shrink_inactive_list(nr_to_scan, zone, sc, priority, file);
>  }
>  
> +static int vmscan_swappiness(struct scan_control *sc)
> +{
> +	if (scanning_global_lru(sc))
> +		return vm_swappiness;
> +	return mem_cgroup_swappiness(sc->mem_cgroup);
> +}
> +
>  /*
>   * Determine how aggressively the anon and file LRU lists should be
>   * scanned.  The relative value of each set of LRU lists is determined
> @@ -1789,8 +1794,8 @@ static void get_scan_count(struct zone *zone, struct scan_control *sc,
>  	 * With swappiness at 100, anonymous and file have the same priority.
>  	 * This scanning priority is essentially the inverse of IO cost.
>  	 */
> -	anon_prio = sc->swappiness;
> -	file_prio = 200 - sc->swappiness;
> +	anon_prio = vmscan_swappiness(sc);
> +	file_prio = 200 - vmscan_swappiness(sc);
>  
>  	/*
>  	 * OK, so we have swap space and a fair amount of page cache
> @@ -2179,7 +2184,6 @@ unsigned long try_to_free_pages(struct zonelist *zonelist, int order,
>  		.nr_to_reclaim = SWAP_CLUSTER_MAX,
>  		.may_unmap = 1,
>  		.may_swap = 1,
> -		.swappiness = vm_swappiness,
>  		.order = order,
>  		.mem_cgroup = NULL,
>  		.nodemask = nodemask,
> @@ -2203,7 +2207,6 @@ 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,
>  						unsigned long *nr_scanned)
>  {
> @@ -2213,7 +2216,6 @@ unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
>  		.may_writepage = !laptop_mode,
>  		.may_unmap = 1,
>  		.may_swap = !noswap,
> -		.swappiness = swappiness,
>  		.order = 0,
>  		.mem_cgroup = mem,
>  	};
> @@ -2242,8 +2244,7 @@ unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
>  
>  unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem_cont,
>  					   gfp_t gfp_mask,
> -					   bool noswap,
> -					   unsigned int swappiness)
> +					   bool noswap)
>  {
>  	struct zonelist *zonelist;
>  	unsigned long nr_reclaimed;
> @@ -2253,7 +2254,6 @@ unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem_cont,
>  		.may_unmap = 1,
>  		.may_swap = !noswap,
>  		.nr_to_reclaim = SWAP_CLUSTER_MAX,
> -		.swappiness = swappiness,
>  		.order = 0,
>  		.mem_cgroup = mem_cont,
>  		.nodemask = NULL, /* we don't care the placement */
> @@ -2403,7 +2403,6 @@ static unsigned long balance_pgdat(pg_data_t *pgdat, int order,
>  		 * we want to put equal scanning pressure on each zone.
>  		 */
>  		.nr_to_reclaim = ULONG_MAX,
> -		.swappiness = vm_swappiness,
>  		.order = order,
>  		.mem_cgroup = NULL,
>  	};
> @@ -2862,7 +2861,6 @@ unsigned long shrink_all_memory(unsigned long nr_to_reclaim)
>  		.may_writepage = 1,
>  		.nr_to_reclaim = nr_to_reclaim,
>  		.hibernation_mode = 1,
> -		.swappiness = vm_swappiness,
>  		.order = 0,
>  	};
>  	struct shrink_control shrink = {
> @@ -3049,7 +3047,6 @@ static int __zone_reclaim(struct zone *zone, gfp_t gfp_mask, unsigned int order)
>  		.nr_to_reclaim = max_t(unsigned long, nr_pages,
>  				       SWAP_CLUSTER_MAX),
>  		.gfp_mask = gfp_mask,
> -		.swappiness = vm_swappiness,
>  		.order = order,
>  	};
>  	struct shrink_control shrink = {
> -- 
> 1.7.4.1
> 
> 
> 

Re: [PATCH] [Cleanup] memcg: export memory cgroup's swappiness v2

[KAMEZAWA Hiroyuki]

On Fri, 1 Jul 2011 10:30:07 +0900
KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:

> On Fri, 1 Jul 2011 10:16:24 +0900
> KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> 
> > On Thu, 30 Jun 2011 18:06:53 -0700
> > Andrew Morton <akpm@linux-foundation.org> wrote:
> > 
> > > On Fri, 1 Jul 2011 09:20:59 +0900 KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> > > 
> > > > On Fri, 1 Jul 2011 08:50:13 +0900
> > > > KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> > > > 
> > > > > On Thu, 30 Jun 2011 13:01:34 -0700
> > > > > Andrew Morton <akpm@linux-foundation.org> wrote:
> > > > 
> > > > > Ok, I'll check it. Maybe I miss !CONFIG_SWAP...
> > > > > 
> > > > 
> > > > v4 here. Thank you for pointing out. I could think of several ways but
> > > > maybe this one is good because using vm_swappines with !CONFIG_SWAP seems
> > > > to be a bug.
> > > 
> > > No, it isn't a bug - swappiness also controls the kernel's eagerness to
> > > unmap and reclaim mmapped pagecache.
> > > 
> > 
> > Oh, really ? I didn't understand that.
> > 
> Hmm, anyway, this new version of fix seems better.

Sorry, this seems still buggy. I'll send a new one in the next week :(

Thanks,
-Kame

> ==
> From 7daf93a277e19026bb6edef3e0ac01bbd31dcb5e Mon Sep 17 00:00:00 2001
> From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> Date: Fri, 1 Jul 2011 10:35:57 +0900
> Subject: [PATCH] export memory cgroup's swappines by mem_cgroup_swappiness()
> 
> Each memory cgroup has 'swappiness' value and it can be accessed by
> get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
> and swappiness is passed by argument. It's propagated by scan_control.
> 
> get_swappiness is static function but some planned updates will need to
> get swappiness from files other than memcontrol.c
> This patch exports get_swappiness() as mem_cgroup_swappiness().
> By this, we can remove the argument of swapiness from try_to_free...
> and drop swappiness from scan_control. only memcg uses it.
> 
> Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> 
> Changelog:
>   - move definistions out of CONFIG_SWAP.
>   - fixed/tested allyesconfig/allnoconfig compile failure.
>   - adjusted signedness to vm_swappiness.
>   - drop swappiness from scan_control
> ---
>  include/linux/swap.h |   13 +++++++++----
>  mm/memcontrol.c      |   15 +++++++--------
>  mm/vmscan.c          |   23 ++++++++++-------------
>  3 files changed, 26 insertions(+), 25 deletions(-)
> 
> diff --git a/include/linux/swap.h b/include/linux/swap.h
> index a273468..28f1490 100644
> --- a/include/linux/swap.h
> +++ b/include/linux/swap.h
> @@ -252,11 +252,9 @@ static inline void lru_cache_add_file(struct page *page)
>  extern unsigned long try_to_free_pages(struct zonelist *zonelist, int order,
>  					gfp_t gfp_mask, nodemask_t *mask);
>  extern unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem,
> -						  gfp_t gfp_mask, bool noswap,
> -						  unsigned int swappiness);
> +						  gfp_t gfp_mask, bool noswap);
>  extern unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
>  						gfp_t gfp_mask, bool noswap,
> -						unsigned int swappiness,
>  						struct zone *zone,
>  						unsigned long *nr_scanned);
>  extern int __isolate_lru_page(struct page *page, int mode, int file);
> @@ -299,7 +297,14 @@ static inline void scan_unevictable_unregister_node(struct node *node)
>  
>  extern int kswapd_run(int nid);
>  extern void kswapd_stop(int nid);
> -
> +#ifdef CONFIG_MEM_CGROUP_MEM_RES_CTLR
> +extern int mem_cgroup_swappiness(struct mem_cgroup *mem);
> +#else
> +static inline int mem_cgroup_swappiness(struct mem_cgroup *mem)
> +{
> +	return vm_swappiness;
> +}
> +#endif
>  #ifdef CONFIG_SWAP
>  /* linux/mm/page_io.c */
>  extern int swap_readpage(struct page *);
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index 3e7d5e6..db70176 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -249,7 +249,7 @@ struct mem_cgroup {
>  	atomic_t	oom_lock;
>  	atomic_t	refcnt;
>  
> -	unsigned int	swappiness;
> +	int	swappiness;
>  	/* OOM-Killer disable */
>  	int		oom_kill_disable;
>  
> @@ -1330,7 +1330,7 @@ static unsigned long mem_cgroup_margin(struct mem_cgroup *mem)
>  	return margin >> PAGE_SHIFT;
>  }
>  
> -static unsigned int get_swappiness(struct mem_cgroup *memcg)
> +int mem_cgroup_swappiness(struct mem_cgroup *memcg)
>  {
>  	struct cgroup *cgrp = memcg->css.cgroup;
>  
> @@ -1776,12 +1776,11 @@ static int mem_cgroup_hierarchical_reclaim(struct mem_cgroup *root_mem,
>  		/* we use swappiness of local cgroup */
>  		if (check_soft) {
>  			ret = mem_cgroup_shrink_node_zone(victim, gfp_mask,
> -				noswap, get_swappiness(victim), zone,
> -				&nr_scanned);
> +				noswap, zone, &nr_scanned);
>  			*total_scanned += nr_scanned;
>  		} else
>  			ret = try_to_free_mem_cgroup_pages(victim, gfp_mask,
> -						noswap, get_swappiness(victim));
> +						noswap);
>  		css_put(&victim->css);
>  		/*
>  		 * At shrinking usage, we can't check we should stop here or
> @@ -3826,7 +3825,7 @@ try_to_free:
>  			goto out;
>  		}
>  		progress = try_to_free_mem_cgroup_pages(mem, GFP_KERNEL,
> -						false, get_swappiness(mem));
> +						false);
>  		if (!progress) {
>  			nr_retries--;
>  			/* maybe some writeback is necessary */
> @@ -4288,7 +4287,7 @@ static u64 mem_cgroup_swappiness_read(struct cgroup *cgrp, struct cftype *cft)
>  {
>  	struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp);
>  
> -	return get_swappiness(memcg);
> +	return mem_cgroup_swappiness(memcg);
>  }
>  
>  static int mem_cgroup_swappiness_write(struct cgroup *cgrp, struct cftype *cft,
> @@ -4997,7 +4996,7 @@ mem_cgroup_create(struct cgroup_subsys *ss, struct cgroup *cont)
>  	INIT_LIST_HEAD(&mem->oom_notify);
>  
>  	if (parent)
> -		mem->swappiness = get_swappiness(parent);
> +		mem->swappiness = mem_cgroup_swappiness(parent);
>  	atomic_set(&mem->refcnt, 1);
>  	mem->move_charge_at_immigrate = 0;
>  	mutex_init(&mem->thresholds_lock);
> diff --git a/mm/vmscan.c b/mm/vmscan.c
> index 4f49535..fb37699 100644
> --- a/mm/vmscan.c
> +++ b/mm/vmscan.c
> @@ -95,8 +95,6 @@ struct scan_control {
>  	/* Can pages be swapped as part of reclaim? */
>  	int may_swap;
>  
> -	int swappiness;
> -
>  	int order;
>  
>  	/*
> @@ -1729,6 +1727,13 @@ static unsigned long shrink_list(enum lru_list lru, unsigned long nr_to_scan,
>  	return shrink_inactive_list(nr_to_scan, zone, sc, priority, file);
>  }
>  
> +static int vmscan_swappiness(struct scan_control *sc)
> +{
> +	if (scanning_global_lru(sc))
> +		return vm_swappiness;
> +	return mem_cgroup_swappiness(sc->mem_cgroup);
> +}
> +
>  /*
>   * Determine how aggressively the anon and file LRU lists should be
>   * scanned.  The relative value of each set of LRU lists is determined
> @@ -1789,8 +1794,8 @@ static void get_scan_count(struct zone *zone, struct scan_control *sc,
>  	 * With swappiness at 100, anonymous and file have the same priority.
>  	 * This scanning priority is essentially the inverse of IO cost.
>  	 */
> -	anon_prio = sc->swappiness;
> -	file_prio = 200 - sc->swappiness;
> +	anon_prio = vmscan_swappiness(sc);
> +	file_prio = 200 - vmscan_swappiness(sc);
>  
>  	/*
>  	 * OK, so we have swap space and a fair amount of page cache
> @@ -2179,7 +2184,6 @@ unsigned long try_to_free_pages(struct zonelist *zonelist, int order,
>  		.nr_to_reclaim = SWAP_CLUSTER_MAX,
>  		.may_unmap = 1,
>  		.may_swap = 1,
> -		.swappiness = vm_swappiness,
>  		.order = order,
>  		.mem_cgroup = NULL,
>  		.nodemask = nodemask,
> @@ -2203,7 +2207,6 @@ 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,
>  						unsigned long *nr_scanned)
>  {
> @@ -2213,7 +2216,6 @@ unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
>  		.may_writepage = !laptop_mode,
>  		.may_unmap = 1,
>  		.may_swap = !noswap,
> -		.swappiness = swappiness,
>  		.order = 0,
>  		.mem_cgroup = mem,
>  	};
> @@ -2242,8 +2244,7 @@ unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
>  
>  unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem_cont,
>  					   gfp_t gfp_mask,
> -					   bool noswap,
> -					   unsigned int swappiness)
> +					   bool noswap)
>  {
>  	struct zonelist *zonelist;
>  	unsigned long nr_reclaimed;
> @@ -2253,7 +2254,6 @@ unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem_cont,
>  		.may_unmap = 1,
>  		.may_swap = !noswap,
>  		.nr_to_reclaim = SWAP_CLUSTER_MAX,
> -		.swappiness = swappiness,
>  		.order = 0,
>  		.mem_cgroup = mem_cont,
>  		.nodemask = NULL, /* we don't care the placement */
> @@ -2403,7 +2403,6 @@ static unsigned long balance_pgdat(pg_data_t *pgdat, int order,
>  		 * we want to put equal scanning pressure on each zone.
>  		 */
>  		.nr_to_reclaim = ULONG_MAX,
> -		.swappiness = vm_swappiness,
>  		.order = order,
>  		.mem_cgroup = NULL,
>  	};
> @@ -2862,7 +2861,6 @@ unsigned long shrink_all_memory(unsigned long nr_to_reclaim)
>  		.may_writepage = 1,
>  		.nr_to_reclaim = nr_to_reclaim,
>  		.hibernation_mode = 1,
> -		.swappiness = vm_swappiness,
>  		.order = 0,
>  	};
>  	struct shrink_control shrink = {
> @@ -3049,7 +3047,6 @@ static int __zone_reclaim(struct zone *zone, gfp_t gfp_mask, unsigned int order)
>  		.nr_to_reclaim = max_t(unsigned long, nr_pages,
>  				       SWAP_CLUSTER_MAX),
>  		.gfp_mask = gfp_mask,
> -		.swappiness = vm_swappiness,
>  		.order = order,
>  	};
>  	struct shrink_control shrink = {
> -- 
> 1.7.4.1
> 
> 
> 

--
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] [Cleanup] memcg: export memory cgroup's swappiness v5

[KAMEZAWA Hiroyuki]

On Fri, 1 Jul 2011 16:10:51 +0900
KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:

> Sorry, this seems still buggy. I'll send a new one in the next week :(
> 

tested with allnoconfig, allyesconfig, CONFIG_SWAP=y/n,
CONFIG_CGROUP_MEM_RES_CTLR=y/n

Patch is onto mmotm-0630.

==
>From 7daf93a277e19026bb6edef3e0ac01bbd31dcb5e Mon Sep 17 00:00:00 2001
From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Date: Fri, 1 Jul 2011 10:35:57 +0900
Subject: [PATCH] export memory cgroup's swappines by mem_cgroup_swappiness()

Each memory cgroup has 'swappiness' value and it can be accessed by
get_swappiness(memcg). The major user is try_to_free_mem_cgroup_pages()
and swappiness is passed by argument. It's propagated by scan_control.

get_swappiness is static function but some planned updates will need to
get swappiness from files other than memcontrol.c
This patch exports get_swappiness() as mem_cgroup_swappiness().
By this, we can remove the argument of swapiness from try_to_free...
and drop swappiness from scan_control. only memcg uses it.

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

Changelog:
  - move definistions out of CONFIG_SWAP.
  - fixed/tested allyesconfig/allnoconfig compile failure.
  - adjusted signedness to vm_swappiness.
  - drop swappiness from scan_control
---
 include/linux/swap.h |   13 +++++++++----
 mm/memcontrol.c      |   15 +++++++--------
 mm/vmscan.c          |   23 ++++++++++-------------
 3 files changed, 26 insertions(+), 25 deletions(-)

Index: mmotm-0701/include/linux/swap.h
===================================================================
--- mmotm-0701.orig/include/linux/swap.h
+++ mmotm-0701/include/linux/swap.h
@@ -254,11 +254,9 @@ static inline void lru_cache_add_file(st
 extern unsigned long try_to_free_pages(struct zonelist *zonelist, int order,
 					gfp_t gfp_mask, nodemask_t *mask);
 extern unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem,
-						  gfp_t gfp_mask, bool noswap,
-						  unsigned int swappiness);
+						  gfp_t gfp_mask, bool noswap);
 extern unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
 						gfp_t gfp_mask, bool noswap,
-						unsigned int swappiness,
 						struct zone *zone,
 						unsigned long *nr_scanned);
 extern int __isolate_lru_page(struct page *page, int mode, int file);
@@ -301,7 +299,14 @@ static inline void scan_unevictable_unre
 
 extern int kswapd_run(int nid);
 extern void kswapd_stop(int nid);
-
+#ifdef CONFIG_CGROUP_MEM_RES_CTLR
+extern int mem_cgroup_swappiness(struct mem_cgroup *mem);
+#else
+static inline int mem_cgroup_swappiness(struct mem_cgroup *mem)
+{
+	return vm_swappiness;
+}
+#endif
 #ifdef CONFIG_SWAP
 /* linux/mm/page_io.c */
 extern int swap_readpage(struct page *);
Index: mmotm-0701/mm/memcontrol.c
===================================================================
--- mmotm-0701.orig/mm/memcontrol.c
+++ mmotm-0701/mm/memcontrol.c
@@ -248,7 +248,7 @@ struct mem_cgroup {
 	atomic_t	oom_lock;
 	atomic_t	refcnt;
 
-	unsigned int	swappiness;
+	int	swappiness;
 	/* OOM-Killer disable */
 	int		oom_kill_disable;
 
@@ -1328,7 +1328,7 @@ static unsigned long mem_cgroup_margin(s
 	return margin >> PAGE_SHIFT;
 }
 
-static unsigned int get_swappiness(struct mem_cgroup *memcg)
+int mem_cgroup_swappiness(struct mem_cgroup *memcg)
 {
 	struct cgroup *cgrp = memcg->css.cgroup;
 
@@ -1775,12 +1775,11 @@ static int mem_cgroup_hierarchical_recla
 		/* we use swappiness of local cgroup */
 		if (check_soft) {
 			ret = mem_cgroup_shrink_node_zone(victim, gfp_mask,
-				noswap, get_swappiness(victim), zone,
-				&nr_scanned);
+				noswap, zone, &nr_scanned);
 			*total_scanned += nr_scanned;
 		} else
 			ret = try_to_free_mem_cgroup_pages(victim, gfp_mask,
-						noswap, get_swappiness(victim));
+						noswap);
 		css_put(&victim->css);
 		/*
 		 * At shrinking usage, we can't check we should stop here or
@@ -3776,7 +3775,7 @@ try_to_free:
 			goto out;
 		}
 		progress = try_to_free_mem_cgroup_pages(mem, GFP_KERNEL,
-						false, get_swappiness(mem));
+						false);
 		if (!progress) {
 			nr_retries--;
 			/* maybe some writeback is necessary */
@@ -4238,7 +4237,7 @@ static u64 mem_cgroup_swappiness_read(st
 {
 	struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp);
 
-	return get_swappiness(memcg);
+	return mem_cgroup_swappiness(memcg);
 }
 
 static int mem_cgroup_swappiness_write(struct cgroup *cgrp, struct cftype *cft,
@@ -4947,7 +4946,7 @@ mem_cgroup_create(struct cgroup_subsys *
 	INIT_LIST_HEAD(&mem->oom_notify);
 
 	if (parent)
-		mem->swappiness = get_swappiness(parent);
+		mem->swappiness = mem_cgroup_swappiness(parent);
 	atomic_set(&mem->refcnt, 1);
 	mem->move_charge_at_immigrate = 0;
 	mutex_init(&mem->thresholds_lock);
Index: mmotm-0701/mm/vmscan.c
===================================================================
--- mmotm-0701.orig/mm/vmscan.c
+++ mmotm-0701/mm/vmscan.c
@@ -95,8 +95,6 @@ struct scan_control {
 	/* Can pages be swapped as part of reclaim? */
 	int may_swap;
 
-	int swappiness;
-
 	int order;
 
 	/*
@@ -1729,6 +1727,13 @@ static unsigned long shrink_list(enum lr
 	return shrink_inactive_list(nr_to_scan, zone, sc, priority, file);
 }
 
+static int vmscan_swappiness(struct scan_control *sc)
+{
+	if (scanning_global_lru(sc))
+		return vm_swappiness;
+	return mem_cgroup_swappiness(sc->mem_cgroup);
+}
+
 /*
  * Determine how aggressively the anon and file LRU lists should be
  * scanned.  The relative value of each set of LRU lists is determined
@@ -1789,8 +1794,8 @@ static void get_scan_count(struct zone *
 	 * With swappiness at 100, anonymous and file have the same priority.
 	 * This scanning priority is essentially the inverse of IO cost.
 	 */
-	anon_prio = sc->swappiness;
-	file_prio = 200 - sc->swappiness;
+	anon_prio = vmscan_swappiness(sc);
+	file_prio = 200 - vmscan_swappiness(sc);
 
 	/*
 	 * OK, so we have swap space and a fair amount of page cache
@@ -2179,7 +2184,6 @@ unsigned long try_to_free_pages(struct z
 		.nr_to_reclaim = SWAP_CLUSTER_MAX,
 		.may_unmap = 1,
 		.may_swap = 1,
-		.swappiness = vm_swappiness,
 		.order = order,
 		.mem_cgroup = NULL,
 		.nodemask = nodemask,
@@ -2203,7 +2207,6 @@ unsigned long try_to_free_pages(struct z
 
 unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
 						gfp_t gfp_mask, bool noswap,
-						unsigned int swappiness,
 						struct zone *zone,
 						unsigned long *nr_scanned)
 {
@@ -2213,7 +2216,6 @@ unsigned long mem_cgroup_shrink_node_zon
 		.may_writepage = !laptop_mode,
 		.may_unmap = 1,
 		.may_swap = !noswap,
-		.swappiness = swappiness,
 		.order = 0,
 		.mem_cgroup = mem,
 	};
@@ -2242,8 +2244,7 @@ unsigned long mem_cgroup_shrink_node_zon
 
 unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem_cont,
 					   gfp_t gfp_mask,
-					   bool noswap,
-					   unsigned int swappiness)
+					   bool noswap)
 {
 	struct zonelist *zonelist;
 	unsigned long nr_reclaimed;
@@ -2253,7 +2254,6 @@ unsigned long try_to_free_mem_cgroup_pag
 		.may_unmap = 1,
 		.may_swap = !noswap,
 		.nr_to_reclaim = SWAP_CLUSTER_MAX,
-		.swappiness = swappiness,
 		.order = 0,
 		.mem_cgroup = mem_cont,
 		.nodemask = NULL, /* we don't care the placement */
@@ -2403,7 +2403,6 @@ static unsigned long balance_pgdat(pg_da
 		 * we want to put equal scanning pressure on each zone.
 		 */
 		.nr_to_reclaim = ULONG_MAX,
-		.swappiness = vm_swappiness,
 		.order = order,
 		.mem_cgroup = NULL,
 	};
@@ -2873,7 +2872,6 @@ unsigned long shrink_all_memory(unsigned
 		.may_writepage = 1,
 		.nr_to_reclaim = nr_to_reclaim,
 		.hibernation_mode = 1,
-		.swappiness = vm_swappiness,
 		.order = 0,
 	};
 	struct shrink_control shrink = {
@@ -3060,7 +3058,6 @@ static int __zone_reclaim(struct zone *z
 		.nr_to_reclaim = max_t(unsigned long, nr_pages,
 				       SWAP_CLUSTER_MAX),
 		.gfp_mask = gfp_mask,
-		.swappiness = vm_swappiness,
 		.order = order,
 	};
 	struct shrink_control shrink = {

[PATCH] [Cleanup] memcg: export memory cgroup's swappiness v5

[KAMEZAWA Hiroyuki]

On Fri, 1 Jul 2011 16:10:51 +0900
KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:

> Sorry, this seems still buggy. I'll send a new one in the next week :(
> 

tested with allnoconfig, allyesconfig, CONFIG_SWAP=y/n,
CONFIG_CGROUP_MEM_RES_CTLR=y/n

Patch is onto mmotm-0630.

==