All of lore.kernel.org
 help / color / mirror / Atom feed
From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
To: "linux-mm@kvack.org" <linux-mm@kvack.org>
Cc: "cgroups@vger.kernel.org" <cgroups@vger.kernel.org>,
	Michal Hocko <mhocko@suse.cz>,
	Johannes Weiner <hannes@cmpxchg.org>,
	Frederic Weisbecker <fweisbec@gmail.com>,
	Han Ying <yinghan@google.com>,
	Glauber Costa <glommer@parallels.com>, Tejun Heo <tj@kernel.org>,
	"Aneesh Kumar K.V" <aneesh.kumar@linux.vnet.ibm.com>,
	Andrew Morton <akpm@linux-foundation.org>,
	Hiroyuki Kamezawa <kamezawa.hiroyuki@gmail.com>,
	Linux Kernel <linux-kernel@vger.kernel.org>
Subject: [PATCH v3 3/6] memcg: use res_counter_uncharge_until in move_parent()
Date: Fri, 11 May 2012 18:48:12 +0900	[thread overview]
Message-ID: <4FACE05C.60608@jp.fujitsu.com> (raw)
In-Reply-To: <4FACDED0.3020400@jp.fujitsu.com>

By using res_counter_uncharge_until(), we can avoid race and
unnecessary charging.

Changelog since v2:
 - a coding style chanve in __mem_cgroup_cancel_local_charge()
 - fixed typos.

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
---
 mm/memcontrol.c |   63 ++++++++++++++++++++++++++++++++++++------------------
 1 files changed, 42 insertions(+), 21 deletions(-)

diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 09109be..cb90be1 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -2446,6 +2446,24 @@ static void __mem_cgroup_cancel_charge(struct mem_cgroup *memcg,
 }
 
 /*
+ * Cancel chrages in this cgroup....doesn't propagate to parent cgroup.
+ * This is useful when moving usage to parent cgroup.
+ */
+static void __mem_cgroup_cancel_local_charge(struct mem_cgroup *memcg,
+					unsigned int nr_pages)
+{
+	unsigned long bytes = nr_pages * PAGE_SIZE;
+
+	if (mem_cgroup_is_root(memcg))
+		return;
+
+	res_counter_uncharge_until(&memcg->res, memcg->res.parent, bytes);
+	if (do_swap_account)
+		res_counter_uncharge_until(&memcg->memsw,
+						memcg->memsw.parent, bytes);
+}
+
+/*
  * A helper function to get mem_cgroup from ID. must be called under
  * rcu_read_lock(). The caller must check css_is_removed() or some if
  * it's concern. (dropping refcnt from swap can be called against removed
@@ -2711,16 +2729,28 @@ static int mem_cgroup_move_parent(struct page *page,
 	nr_pages = hpage_nr_pages(page);
 
 	parent = mem_cgroup_from_cont(pcg);
-	ret = __mem_cgroup_try_charge(NULL, gfp_mask, nr_pages, &parent, false);
-	if (ret)
-		goto put_back;
+	if (!parent->use_hierarchy) {
+		ret = __mem_cgroup_try_charge(NULL,
+					gfp_mask, nr_pages, &parent, false);
+		if (ret)
+			goto put_back;
+	}
 
 	if (nr_pages > 1)
 		flags = compound_lock_irqsave(page);
 
-	ret = mem_cgroup_move_account(page, nr_pages, pc, child, parent, true);
-	if (ret)
-		__mem_cgroup_cancel_charge(parent, nr_pages);
+	if (parent->use_hierarchy) {
+		ret = mem_cgroup_move_account(page, nr_pages,
+					pc, child, parent, false);
+		if (!ret)
+			__mem_cgroup_cancel_local_charge(child, nr_pages);
+	} else {
+		ret = mem_cgroup_move_account(page, nr_pages,
+					pc, child, parent, true);
+
+		if (ret)
+			__mem_cgroup_cancel_charge(parent, nr_pages);
+	}
 
 	if (nr_pages > 1)
 		compound_unlock_irqrestore(page, flags);
@@ -3324,6 +3354,7 @@ int mem_cgroup_move_hugetlb_parent(int idx, struct cgroup *cgroup,
 	struct cgroup *pcgrp = cgroup->parent;
 	struct mem_cgroup *parent = mem_cgroup_from_cont(pcgrp);
 	struct mem_cgroup *memcg  = mem_cgroup_from_cont(cgroup);
+	struct res_counter *counter;
 
 	if (!get_page_unless_zero(page))
 		goto out;
@@ -3334,28 +3365,18 @@ int mem_cgroup_move_hugetlb_parent(int idx, struct cgroup *cgroup,
 		goto err_out;
 
 	csize = PAGE_SIZE << compound_order(page);
-	/*
-	 * If we have use_hierarchy set we can never fail here. So instead of
-	 * using res_counter_uncharge use the open-coded variant which just
-	 * uncharge the child res_counter. The parent will retain the charge.
-	 */
-	if (parent->use_hierarchy) {
-		unsigned long flags;
-		struct res_counter *counter;
-
-		counter = &memcg->hugepage[idx];
-		spin_lock_irqsave(&counter->lock, flags);
-		res_counter_uncharge_locked(counter, csize);
-		spin_unlock_irqrestore(&counter->lock, flags);
-	} else {
+	/* If parent->use_hierarchy == 0, we need to charge parent */
+	if (!parent->use_hierarchy) {
 		ret = res_counter_charge(&parent->hugepage[idx],
 					 csize, &fail_res);
 		if (ret) {
 			ret = -EBUSY;
 			goto err_out;
 		}
-		res_counter_uncharge(&memcg->hugepage[idx], csize);
 	}
+	counter = &memcg->hugepage[idx];
+	res_counter_uncharge_until(counter, counter->parent, csize);
+
 	pc->mem_cgroup = parent;
 err_out:
 	unlock_page_cgroup(pc);
-- 
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>

WARNING: multiple messages have this Message-ID (diff)
From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
To: "linux-mm@kvack.org" <linux-mm@kvack.org>
Cc: "cgroups@vger.kernel.org" <cgroups@vger.kernel.org>,
	Michal Hocko <mhocko@suse.cz>,
	Johannes Weiner <hannes@cmpxchg.org>,
	Frederic Weisbecker <fweisbec@gmail.com>,
	Han Ying <yinghan@google.com>,
	Glauber Costa <glommer@parallels.com>, Tejun Heo <tj@kernel.org>,
	"Aneesh Kumar K.V" <aneesh.kumar@linux.vnet.ibm.com>,
	Andrew Morton <akpm@linux-foundation.org>,
	Hiroyuki Kamezawa <kamezawa.hiroyuki@gmail.com>,
	Linux Kernel <linux-kernel@vger.kernel.org>
Subject: [PATCH v3 3/6] memcg: use res_counter_uncharge_until in move_parent()
Date: Fri, 11 May 2012 18:48:12 +0900	[thread overview]
Message-ID: <4FACE05C.60608@jp.fujitsu.com> (raw)
In-Reply-To: <4FACDED0.3020400@jp.fujitsu.com>

By using res_counter_uncharge_until(), we can avoid race and
unnecessary charging.

Changelog since v2:
 - a coding style chanve in __mem_cgroup_cancel_local_charge()
 - fixed typos.

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
---
 mm/memcontrol.c |   63 ++++++++++++++++++++++++++++++++++++------------------
 1 files changed, 42 insertions(+), 21 deletions(-)

diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 09109be..cb90be1 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -2446,6 +2446,24 @@ static void __mem_cgroup_cancel_charge(struct mem_cgroup *memcg,
 }
 
 /*
+ * Cancel chrages in this cgroup....doesn't propagate to parent cgroup.
+ * This is useful when moving usage to parent cgroup.
+ */
+static void __mem_cgroup_cancel_local_charge(struct mem_cgroup *memcg,
+					unsigned int nr_pages)
+{
+	unsigned long bytes = nr_pages * PAGE_SIZE;
+
+	if (mem_cgroup_is_root(memcg))
+		return;
+
+	res_counter_uncharge_until(&memcg->res, memcg->res.parent, bytes);
+	if (do_swap_account)
+		res_counter_uncharge_until(&memcg->memsw,
+						memcg->memsw.parent, bytes);
+}
+
+/*
  * A helper function to get mem_cgroup from ID. must be called under
  * rcu_read_lock(). The caller must check css_is_removed() or some if
  * it's concern. (dropping refcnt from swap can be called against removed
@@ -2711,16 +2729,28 @@ static int mem_cgroup_move_parent(struct page *page,
 	nr_pages = hpage_nr_pages(page);
 
 	parent = mem_cgroup_from_cont(pcg);
-	ret = __mem_cgroup_try_charge(NULL, gfp_mask, nr_pages, &parent, false);
-	if (ret)
-		goto put_back;
+	if (!parent->use_hierarchy) {
+		ret = __mem_cgroup_try_charge(NULL,
+					gfp_mask, nr_pages, &parent, false);
+		if (ret)
+			goto put_back;
+	}
 
 	if (nr_pages > 1)
 		flags = compound_lock_irqsave(page);
 
-	ret = mem_cgroup_move_account(page, nr_pages, pc, child, parent, true);
-	if (ret)
-		__mem_cgroup_cancel_charge(parent, nr_pages);
+	if (parent->use_hierarchy) {
+		ret = mem_cgroup_move_account(page, nr_pages,
+					pc, child, parent, false);
+		if (!ret)
+			__mem_cgroup_cancel_local_charge(child, nr_pages);
+	} else {
+		ret = mem_cgroup_move_account(page, nr_pages,
+					pc, child, parent, true);
+
+		if (ret)
+			__mem_cgroup_cancel_charge(parent, nr_pages);
+	}
 
 	if (nr_pages > 1)
 		compound_unlock_irqrestore(page, flags);
@@ -3324,6 +3354,7 @@ int mem_cgroup_move_hugetlb_parent(int idx, struct cgroup *cgroup,
 	struct cgroup *pcgrp = cgroup->parent;
 	struct mem_cgroup *parent = mem_cgroup_from_cont(pcgrp);
 	struct mem_cgroup *memcg  = mem_cgroup_from_cont(cgroup);
+	struct res_counter *counter;
 
 	if (!get_page_unless_zero(page))
 		goto out;
@@ -3334,28 +3365,18 @@ int mem_cgroup_move_hugetlb_parent(int idx, struct cgroup *cgroup,
 		goto err_out;
 
 	csize = PAGE_SIZE << compound_order(page);
-	/*
-	 * If we have use_hierarchy set we can never fail here. So instead of
-	 * using res_counter_uncharge use the open-coded variant which just
-	 * uncharge the child res_counter. The parent will retain the charge.
-	 */
-	if (parent->use_hierarchy) {
-		unsigned long flags;
-		struct res_counter *counter;
-
-		counter = &memcg->hugepage[idx];
-		spin_lock_irqsave(&counter->lock, flags);
-		res_counter_uncharge_locked(counter, csize);
-		spin_unlock_irqrestore(&counter->lock, flags);
-	} else {
+	/* If parent->use_hierarchy == 0, we need to charge parent */
+	if (!parent->use_hierarchy) {
 		ret = res_counter_charge(&parent->hugepage[idx],
 					 csize, &fail_res);
 		if (ret) {
 			ret = -EBUSY;
 			goto err_out;
 		}
-		res_counter_uncharge(&memcg->hugepage[idx], csize);
 	}
+	counter = &memcg->hugepage[idx];
+	res_counter_uncharge_until(counter, counter->parent, csize);
+
 	pc->mem_cgroup = parent;
 err_out:
 	unlock_page_cgroup(pc);
-- 
1.7.4.1



  parent reply	other threads:[~2012-05-11  9:48 UTC|newest]

Thread overview: 79+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-05-11  9:41 [PATCH v3][0/6] memcg: prevent -ENOMEM in pre_destroy() KAMEZAWA Hiroyuki
2012-05-11  9:41 ` KAMEZAWA Hiroyuki
2012-05-11  9:41 ` KAMEZAWA Hiroyuki
2012-05-11  9:45 ` [PATCH v3 1/6] memcg: fix error code in hugetlb_force_memcg_empty() KAMEZAWA Hiroyuki
2012-05-11  9:45   ` KAMEZAWA Hiroyuki
     [not found]   ` <4FACDFAE.5050808-+CUm20s59erQFUHtdCDX3A@public.gmane.org>
2012-05-11 21:17     ` Andrew Morton
2012-05-11 21:17       ` Andrew Morton
2012-05-11 21:17       ` Andrew Morton
     [not found]       ` <20120511141754.e0719c26.akpm-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b@public.gmane.org>
2012-05-14  1:07         ` KAMEZAWA Hiroyuki
2012-05-14  1:07           ` KAMEZAWA Hiroyuki
2012-05-14  1:07           ` KAMEZAWA Hiroyuki
2012-05-14 18:15     ` Tejun Heo
2012-05-14 18:15       ` Tejun Heo
2012-05-14 18:15       ` Tejun Heo
     [not found]       ` <20120514181556.GE2366-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org>
2012-05-14 18:32         ` Tejun Heo
2012-05-14 18:32           ` Tejun Heo
2012-05-14 18:32           ` Tejun Heo
     [not found]           ` <20120514183219.GG2366-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org>
2012-05-15  1:10             ` KAMEZAWA Hiroyuki
2012-05-15  1:10               ` KAMEZAWA Hiroyuki
2012-05-15  1:10               ` KAMEZAWA Hiroyuki
2012-05-15 15:12               ` Tejun Heo
2012-05-15 15:12                 ` Tejun Heo
2012-05-11  9:47 ` [PATCH 2/6] add res_counter_uncharge_until() KAMEZAWA Hiroyuki
2012-05-11  9:47   ` KAMEZAWA Hiroyuki
     [not found]   ` <4FACE01A.4040405-+CUm20s59erQFUHtdCDX3A@public.gmane.org>
2012-05-11 21:19     ` Andrew Morton
2012-05-11 21:19       ` Andrew Morton
2012-05-11 21:19       ` Andrew Morton
     [not found]       ` <20120511141945.c487e94c.akpm-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b@public.gmane.org>
2012-05-14  1:10         ` KAMEZAWA Hiroyuki
2012-05-14  1:10           ` KAMEZAWA Hiroyuki
2012-05-14  1:10           ` KAMEZAWA Hiroyuki
2012-05-14 10:08           ` Frederic Weisbecker
2012-05-14 10:08             ` Frederic Weisbecker
     [not found]             ` <CAFTL4hwGEhyxZO0sXx5gVyK_xjhMQEbHojJbHzQmVKafNyVWtw-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2012-05-14 10:32               ` KAMEZAWA Hiroyuki
2012-05-14 10:32                 ` KAMEZAWA Hiroyuki
2012-05-14 10:32                 ` KAMEZAWA Hiroyuki
     [not found]                 ` <4FB0DF4A.5010506-+CUm20s59erQFUHtdCDX3A@public.gmane.org>
2012-05-14 10:56                   ` Frederic Weisbecker
2012-05-14 10:56                     ` Frederic Weisbecker
2012-05-14 10:56                     ` Frederic Weisbecker
2012-05-14 18:17                   ` Tejun Heo
2012-05-14 18:17                     ` Tejun Heo
2012-05-14 18:17                     ` Tejun Heo
2012-05-11  9:48 ` KAMEZAWA Hiroyuki [this message]
2012-05-11  9:48   ` [PATCH v3 3/6] memcg: use res_counter_uncharge_until in move_parent() KAMEZAWA Hiroyuki
     [not found] ` <4FACDED0.3020400-+CUm20s59erQFUHtdCDX3A@public.gmane.org>
2012-05-11  9:49   ` [PATCH v3 4/6] memcg: move charges to root cgroup if use_hierarchy=0 KAMEZAWA Hiroyuki
2012-05-11  9:49     ` KAMEZAWA Hiroyuki
2012-05-11  9:49     ` KAMEZAWA Hiroyuki
     [not found]     ` <4FACE0A2.30608-+CUm20s59erQFUHtdCDX3A@public.gmane.org>
2012-05-14 20:14       ` Tejun Heo
2012-05-14 20:14         ` Tejun Heo
2012-05-14 20:14         ` Tejun Heo
     [not found]         ` <20120514201438.GI2366-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org>
2012-05-15  0:04           ` KAMEZAWA Hiroyuki
2012-05-15  0:04             ` KAMEZAWA Hiroyuki
2012-05-15  0:04             ` KAMEZAWA Hiroyuki
2012-05-11  9:50   ` [PATCH v3 5/6] memcg: don't uncharge in mem_cgroup_move_account KAMEZAWA Hiroyuki
2012-05-11  9:50     ` KAMEZAWA Hiroyuki
2012-05-11  9:50     ` KAMEZAWA Hiroyuki
2012-05-11  9:53   ` [PATCH v3 6/6] remove __must_check for res_counter_charge_nofail() KAMEZAWA Hiroyuki
2012-05-11  9:53     ` KAMEZAWA Hiroyuki
2012-05-11  9:53     ` KAMEZAWA Hiroyuki
     [not found]     ` <4FACE184.6020307-+CUm20s59erQFUHtdCDX3A@public.gmane.org>
2012-05-14 20:09       ` Tejun Heo
2012-05-14 20:09         ` Tejun Heo
2012-05-14 20:09         ` Tejun Heo
     [not found]         ` <20120514200925.GH2366-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org>
2012-05-15  0:02           ` KAMEZAWA Hiroyuki
2012-05-15  0:02             ` KAMEZAWA Hiroyuki
2012-05-15  0:02             ` KAMEZAWA Hiroyuki
2012-06-21 20:20   ` [PATCH v3][0/6] memcg: prevent -ENOMEM in pre_destroy() Tejun Heo
2012-06-21 20:20     ` Tejun Heo
2012-06-21 20:20     ` Tejun Heo
     [not found]     ` <20120621202043.GD4642-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org>
2012-06-21 23:27       ` Kamezawa Hiroyuki
2012-06-21 23:27         ` Kamezawa Hiroyuki
2012-06-21 23:27         ` Kamezawa Hiroyuki
     [not found]         ` <4FE3ADDD.9060908-+CUm20s59erQFUHtdCDX3A@public.gmane.org>
2012-06-27 17:58           ` Tejun Heo
2012-06-27 17:58             ` Tejun Heo
2012-06-27 17:58             ` Tejun Heo
     [not found]             ` <20120627175818.GM15811-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org>
2012-06-28  8:33               ` Kamezawa Hiroyuki
2012-06-28  8:33                 ` Kamezawa Hiroyuki
2012-06-28  8:33                 ` Kamezawa Hiroyuki
     [not found]                 ` <4FEC16EF.40408-+CUm20s59erQFUHtdCDX3A@public.gmane.org>
2012-06-28 16:06                   ` Tejun Heo
2012-06-28 16:06                     ` Tejun Heo
2012-06-28 16:06                     ` Tejun Heo

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=4FACE05C.60608@jp.fujitsu.com \
    --to=kamezawa.hiroyu@jp.fujitsu.com \
    --cc=akpm@linux-foundation.org \
    --cc=aneesh.kumar@linux.vnet.ibm.com \
    --cc=cgroups@vger.kernel.org \
    --cc=fweisbec@gmail.com \
    --cc=glommer@parallels.com \
    --cc=hannes@cmpxchg.org \
    --cc=kamezawa.hiroyuki@gmail.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=mhocko@suse.cz \
    --cc=tj@kernel.org \
    --cc=yinghan@google.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.