linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
To: linux-mm <linux-mm@kvack.org>
Cc: Balbir Singh <balbir@linux.vnet.ibm.com>,
	KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>,
	Paul Menage <menage@google.com>, Li Zefan <lizf@cn.fujitsu.com>,
	Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
Subject: [RFC][PATCH 6/8] memcg: avoid oom during charge migration
Date: Thu, 24 Sep 2009 14:49:02 +0900	[thread overview]
Message-ID: <20090924144902.f4e5854c.nishimura@mxp.nes.nec.co.jp> (raw)
In-Reply-To: <20090924144214.508469d1.nishimura@mxp.nes.nec.co.jp>

This charge migration feature has double charges on both "from" and "to"
mem_cgroup during charge migration.
This means unnecessary oom can happen because of charge migration.

This patch tries to avoid such oom.

Signed-off-by: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
---
 mm/memcontrol.c |   19 +++++++++++++++++++
 1 files changed, 19 insertions(+), 0 deletions(-)

diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index fbcc195..25de11c 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -287,6 +287,8 @@ struct migrate_charge {
 	unsigned long precharge;
 };
 static struct migrate_charge *mc;
+static struct task_struct *mc_task;
+static DECLARE_WAIT_QUEUE_HEAD(mc_waitq);
 
 static void mem_cgroup_get(struct mem_cgroup *mem);
 static void mem_cgroup_put(struct mem_cgroup *mem);
@@ -1317,6 +1319,7 @@ static int __mem_cgroup_try_charge(struct mm_struct *mm,
 	while (1) {
 		int ret = 0;
 		unsigned long flags = 0;
+		DEFINE_WAIT(wait);
 
 		if (mem_cgroup_is_root(mem))
 			goto done;
@@ -1358,6 +1361,17 @@ static int __mem_cgroup_try_charge(struct mm_struct *mm,
 		if (mem_cgroup_check_under_limit(mem_over_limit))
 			continue;
 
+		/* try to avoid oom while someone is migrating charge */
+		if (mc_task && current != mc_task) {
+			prepare_to_wait(&mc_waitq, &wait, TASK_INTERRUPTIBLE);
+			if (mc) {
+				schedule();
+				finish_wait(&mc_waitq, &wait);
+				continue;
+			}
+			finish_wait(&mc_waitq, &wait);
+		}
+
 		if (!nr_retries--) {
 			if (oom) {
 				mutex_lock(&memcg_tasklist);
@@ -3345,6 +3359,8 @@ static void mem_cgroup_clear_migrate_charge(void)
 		__mem_cgroup_cancel_charge(mc->to);
 	kfree(mc);
 	mc = NULL;
+	mc_task = NULL;
+	wake_up_all(&mc_waitq);
 }
 
 static int mem_cgroup_can_migrate_charge(struct mem_cgroup *mem,
@@ -3354,6 +3370,7 @@ static int mem_cgroup_can_migrate_charge(struct mem_cgroup *mem,
 	struct mem_cgroup *from = mem_cgroup_from_task(p);
 
 	VM_BUG_ON(mc);
+	VM_BUG_ON(mc_task);
 
 	if (from == mem)
 		return 0;
@@ -3367,6 +3384,8 @@ static int mem_cgroup_can_migrate_charge(struct mem_cgroup *mem,
 	mc->to = mem;
 	mc->precharge = 0;
 
+	mc_task = current;
+
 	ret = migrate_charge_prepare();
 
 	if (ret)
-- 
1.5.6.1

--
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/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

  parent reply	other threads:[~2009-09-24  5:52 UTC|newest]

Thread overview: 49+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-09-17  2:23 [RFC][EXPERIMENTAL][PATCH 0/8] memcg: migrate charge at task move Daisuke Nishimura
2009-09-17  2:24 ` [PATCH 1/8] memcg: introduce mem_cgroup_cancel_charge() Daisuke Nishimura
2009-09-17  4:12   ` KAMEZAWA Hiroyuki
2009-09-17  2:24 ` [PATCH 2/8] memcg: cleanup mem_cgroup_move_parent() Daisuke Nishimura
2009-09-17  4:15   ` KAMEZAWA Hiroyuki
2009-09-17  2:25 ` [PATCH 3/8] cgroup: introduce cancel_attach() Daisuke Nishimura
2009-09-17  2:26 ` [PATCH 4/8] memcg: add interface to migrate charge Daisuke Nishimura
2009-09-17  4:20   ` KAMEZAWA Hiroyuki
2009-09-17  4:40     ` Daisuke Nishimura
2009-09-17  2:26 ` [PATCH 5/8] memcg: migrate charge of anon Daisuke Nishimura
2009-09-17  4:57   ` KAMEZAWA Hiroyuki
2009-09-17  5:56     ` Daisuke Nishimura
2009-09-17  6:25       ` KAMEZAWA Hiroyuki
2009-09-17 23:52     ` KOSAKI Motohiro
2009-09-17  2:27 ` [PATCH 6/8] memcg: migrate charge of shmem Daisuke Nishimura
2009-09-17  5:02   ` KAMEZAWA Hiroyuki
2009-09-17  2:28 ` [PATCH 7/8] memcg: migrate charge of swap Daisuke Nishimura
2009-09-17  5:25   ` KAMEZAWA Hiroyuki
2009-09-17  6:17     ` Daisuke Nishimura
2009-09-17  6:28       ` KAMEZAWA Hiroyuki
2009-09-17  2:29 ` [PATCH 8/8] memcg: avoid oom during charge migration Daisuke Nishimura
2009-09-17  7:01 ` [RFC][EXPERIMENTAL][PATCH 0/8] memcg: migrate charge at task move Daisuke Nishimura
2009-09-24  5:42   ` [RFC][PATCH 0/8] memcg: migrate charge at task move (24/Sep) Daisuke Nishimura
2009-09-24  5:43     ` [RFC][PATCH 1/8] cgroup: introduce cancel_attach() Daisuke Nishimura
2009-09-24  6:33       ` KAMEZAWA Hiroyuki
2009-09-24 23:39         ` Daisuke Nishimura
2009-09-24  5:44     ` [RFC][PATCH 2/8] memcg: introduce mem_cgroup_cancel_charge() Daisuke Nishimura
2009-09-24  5:46     ` [RFC][PATCH 3/8] memcg: cleanup mem_cgroup_move_parent() Daisuke Nishimura
2009-09-24  6:37       ` KAMEZAWA Hiroyuki
2009-09-24  6:54         ` Daisuke Nishimura
2009-09-24  5:47     ` [RFC][PATCH 4/8] memcg: add interface to migrate charge Daisuke Nishimura
2009-09-24  6:54       ` KAMEZAWA Hiroyuki
2009-09-24 23:39         ` Daisuke Nishimura
2009-09-24  5:48     ` [RFC][PATCH 5/8] memcg: migrate charge of mapped page Daisuke Nishimura
2009-09-24  7:22       ` KAMEZAWA Hiroyuki
2009-09-24  8:00         ` Daisuke Nishimura
2009-09-25  0:28           ` Daisuke Nishimura
2009-09-25  0:49             ` KAMEZAWA Hiroyuki
2009-09-24  5:49     ` Daisuke Nishimura [this message]
2009-09-24  7:34       ` [RFC][PATCH 6/8] memcg: avoid oom during charge migration KAMEZAWA Hiroyuki
2009-09-25  1:44         ` Daisuke Nishimura
2009-09-25  1:55           ` KAMEZAWA Hiroyuki
2009-09-25  4:51             ` Daisuke Nishimura
2009-09-25  5:36               ` Daisuke Nishimura
2009-09-25  5:52                 ` KAMEZAWA Hiroyuki
2009-09-24  5:49     ` [RFC][PATCH 7/8] memcg: migrate charge of swap Daisuke Nishimura
2009-09-24  5:50     ` [RFC][PATCH 8/8] memcg: migrate charge of shmem swap Daisuke Nishimura
2009-09-24  7:41       ` KAMEZAWA Hiroyuki
2009-09-25  0:28         ` Daisuke Nishimura

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=20090924144902.f4e5854c.nishimura@mxp.nes.nec.co.jp \
    --to=nishimura@mxp.nes.nec.co.jp \
    --cc=balbir@linux.vnet.ibm.com \
    --cc=kamezawa.hiroyu@jp.fujitsu.com \
    --cc=linux-mm@kvack.org \
    --cc=lizf@cn.fujitsu.com \
    --cc=menage@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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox