LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: "KAMEZAWA Hiroyuki" <kamezawa.hiroyu@jp.fujitsu.com>
To: balbir@linux.vnet.ibm.com
Cc: "KAMEZAWA Hiroyuki" <kamezawa.hiroyu@jp.fujitsu.com>,
	linux-mm@kvack.org, "YAMAMOTO Takashi" <yamamoto@valinux.co.jp>,
	"Paul Menage" <menage@google.com>,
	lizf@cn.fujitsu.com, linux-kernel@vger.kernel.org,
	"Nick Piggin" <nickpiggin@yahoo.com.au>,
	"David Rientjes" <rientjes@google.com>,
	"Pavel Emelianov" <xemul@openvz.org>,
	"Dhaval Giani" <dhaval@linux.vnet.ibm.com>,
	"Andrew Morton" <akpm@linux-foundation.org>
Subject: Re: [mm] [PATCH 3/4] Memory cgroup hierarchical reclaim
Date: Thu, 6 Nov 2008 01:20:29 +0900 (JST)	[thread overview]
Message-ID: <45571.10.75.179.62.1225902029.squirrel@webmail-b.css.fujitsu.com> (raw)
In-Reply-To: <4911A0FC.9@linux.vnet.ibm.com>

Balbir Singh said:
>>>>> +	list_for_each_entry_safe_from(cgroup, cg,
>>>>> &cg_current->parent->children,
>>>>> +						 sibling) {
>>>>> +		mem_child = mem_cgroup_from_cont(cgroup);
>>>>> +
>>>>> +		/*
>>>>> +		 * Move beyond last scanned child
>>>>> +		 */
>>>>> +		if (mem_child == mem->last_scanned_child)
>>>>> +			continue;
>>>>> +
>>>>> +		ret = try_to_free_mem_cgroup_pages(mem_child, gfp_mask);
>>>>> +		mem->last_scanned_child = mem_child;
>>>>> +
>>>>> +		if (res_counter_check_under_limit(&mem->res)) {
>>>>> +			ret = 0;
>>>>> +			goto done;
>>>>> +		}
>>>>> +	}
>>>> Is this safe against cgroup create/remove ? cgroup_mutex is held ?
>>> Yes, I thought about it, but with the setup, each parent will be busy
>>> since they
>>> have children and hence cannot be removed. The leaf child itself has
>>> tasks, so
>>> it cannot be removed. IOW, it should be safe against removal.
>>>
>> I'm sorry if I misunderstand something. could you explain folloing ?
>>
>> In following tree,
>>
>>     level-1
>>          -  level-2
>>                 -  level-3
>>                        -  level-4
>> level-1's usage = level-1 + level-2 + level-3 + level-4
>> level-2's usage = level-2 + level-3 + level-4
>> level-3's usage = level-3 + level-4
>> level-4's usage = level-4
>>
>> Assume that a task in level-2 hits its limit. It has to reclaim memory
>> from
>> level-2 and level-3, level-4.
>>
>> How can we guarantee level-4 has a task in this case ?
>
> Good question. If there is no task, the LRU's will be empty and reclaim
> will
> return. We could also add other checks if needed.
>
If needed ?, yes, you need.
The problem is that you are walking a list in usual way without any lock
or guarantee that the list will never be modified.

My quick idea is following.
==
Before start reclaim.
 1. take lock_cgroup()
 2. scan the tree and create "private" list as snapshot of tree to be
    scanned.
 3. unlock_cgroup().
 4. start reclaim.

Adding refcnt to memcg to delay freeing memcg control area is necessary.
(mem+swap controller have function to do this and you may be able to
 reuse it.)

Thanks,
-Kame



  reply	other threads:[~2008-11-05 16:20 UTC|newest]

Thread overview: 34+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-11-01 18:48 [mm][PATCH 0/4] Memory cgroup hierarchy introduction Balbir Singh
2008-11-01 18:48 ` [mm] [PATCH 1/4] Memory cgroup hierarchy documentation Balbir Singh
2008-11-04  6:25   ` Paul Menage
2008-11-04  6:26     ` Paul Menage
2008-11-05 13:55       ` Balbir Singh
2008-11-01 18:48 ` [mm] [PATCH 2/4] Memory cgroup resource counters for hierarchy Balbir Singh
2008-11-02  5:42   ` KAMEZAWA Hiroyuki
2008-11-02  5:49     ` Balbir Singh
2008-11-02  5:56       ` KAMEZAWA Hiroyuki
2008-11-02 11:46         ` Balbir Singh
2008-11-01 18:48 ` [mm] [PATCH 3/4] Memory cgroup hierarchical reclaim Balbir Singh
2008-11-02  5:37   ` KAMEZAWA Hiroyuki
2008-11-02  5:44     ` Balbir Singh
2008-11-04  2:17       ` KAMEZAWA Hiroyuki
2008-11-05 13:34         ` Balbir Singh
2008-11-05 16:20           ` KAMEZAWA Hiroyuki [this message]
2008-11-06 14:00             ` Balbir Singh
2008-11-01 18:49 ` [mm] [PATCH 4/4] Memory cgroup hierarchy feature selector Balbir Singh
2008-11-02  5:38   ` KAMEZAWA Hiroyuki
2008-11-02  6:03     ` Balbir Singh
2008-11-02  6:24       ` KAMEZAWA Hiroyuki
2008-11-02 15:52         ` Balbir Singh
2008-11-04  6:37           ` Paul Menage
2008-11-06  7:00             ` Balbir Singh
2008-11-06  7:01               ` Balbir Singh
2008-11-06  6:56         ` Balbir Singh
2008-11-06  7:30           ` KAMEZAWA Hiroyuki
2008-11-04  0:15 ` [mm][PATCH 0/4] Memory cgroup hierarchy introduction KAMEZAWA Hiroyuki
2008-11-05 13:51   ` Balbir Singh
2008-11-05 16:33     ` KAMEZAWA Hiroyuki
2008-11-05 17:52       ` Balbir Singh
2008-11-06  0:22         ` KAMEZAWA Hiroyuki
2008-11-04  9:21 ` [patch 1/2] memcg: hierarchy, yet another one KAMEZAWA Hiroyuki
2008-11-04  9:25 ` KAMEZAWA Hiroyuki

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=45571.10.75.179.62.1225902029.squirrel@webmail-b.css.fujitsu.com \
    --to=kamezawa.hiroyu@jp.fujitsu.com \
    --cc=akpm@linux-foundation.org \
    --cc=balbir@linux.vnet.ibm.com \
    --cc=dhaval@linux.vnet.ibm.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=lizf@cn.fujitsu.com \
    --cc=menage@google.com \
    --cc=nickpiggin@yahoo.com.au \
    --cc=rientjes@google.com \
    --cc=xemul@openvz.org \
    --cc=yamamoto@valinux.co.jp \
    --subject='Re: [mm] [PATCH 3/4] Memory cgroup hierarchical reclaim' \
    /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

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).