From: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
To: Greg Thelen <gthelen@google.com>
Cc: Andrew Morton <akpm@linux-foundation.org>,
linux-kernel@vger.kernel.org, linux-mm@kvack.org,
containers@lists.osdl.org, Andrea Righi <arighi@develer.com>,
Balbir Singh <balbir@linux.vnet.ibm.com>,
KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>,
Minchan Kim <minchan.kim@gmail.com>,
Ciju Rajan K <ciju@linux.vnet.ibm.com>,
David Rientjes <rientjes@google.com>,
Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
Subject: Re: [PATCH v3 04/11] memcg: add lock to synchronize page accounting and migration
Date: Tue, 19 Oct 2010 14:03:38 +0900 [thread overview]
Message-ID: <20101019140338.9fd664bc.nishimura@mxp.nes.nec.co.jp> (raw)
In-Reply-To: <1287448784-25684-5-git-send-email-gthelen@google.com>
On Mon, 18 Oct 2010 17:39:37 -0700
Greg Thelen <gthelen@google.com> wrote:
> From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
>
> From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
>
> Introduce a new bit spin lock, PCG_MOVE_LOCK, to synchronize
> the page accounting and migration code. This reworks the
> locking scheme of _update_stat() and _move_account() by
> adding new lock bit PCG_MOVE_LOCK, which is always taken
> under IRQ disable.
>
> 1. If pages are being migrated from a memcg, then updates to
> that memcg page statistics are protected by grabbing
> PCG_MOVE_LOCK using move_lock_page_cgroup(). In an
> upcoming commit, memcg dirty page accounting will be
> updating memcg page accounting (specifically: num
> writeback pages) from IRQ context (softirq). Avoid a
> deadlocking nested spin lock attempt by disabling irq on
> the local processor when grabbing the PCG_MOVE_LOCK.
>
> 2. lock for update_page_stat is used only for avoiding race
> with move_account(). So, IRQ awareness of
> lock_page_cgroup() itself is not a problem. The problem
> is between mem_cgroup_update_page_stat() and
> mem_cgroup_move_account_page().
>
> Trade-off:
> * Changing lock_page_cgroup() to always disable IRQ (or
> local_bh) has some impacts on performance and I think
> it's bad to disable IRQ when it's not necessary.
> * adding a new lock makes move_account() slower. Score is
> here.
>
> Performance Impact: moving a 8G anon process.
>
> Before:
> real 0m0.792s
> user 0m0.000s
> sys 0m0.780s
>
> After:
> real 0m0.854s
> user 0m0.000s
> sys 0m0.842s
>
> This score is bad but planned patches for optimization can reduce
> this impact.
>
> Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> Signed-off-by: Greg Thelen <gthelen@google.com>
This approach is more straightforward and easy to understand.
Acked-by: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
next prev parent reply other threads:[~2010-10-19 5:10 UTC|newest]
Thread overview: 64+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-10-19 0:39 [PATCH v3 00/11] memcg: per cgroup dirty page accounting Greg Thelen
2010-10-19 0:39 ` [PATCH v3 01/11] memcg: add page_cgroup flags for dirty page tracking Greg Thelen
2010-10-19 4:31 ` Daisuke Nishimura
2010-10-19 0:39 ` [PATCH v3 02/11] memcg: document cgroup dirty memory interfaces Greg Thelen
2010-10-19 0:46 ` KAMEZAWA Hiroyuki
2010-10-19 8:27 ` Daisuke Nishimura
2010-10-19 21:00 ` Greg Thelen
2010-10-20 0:11 ` KAMEZAWA Hiroyuki
2010-10-20 0:45 ` Greg Thelen
2010-10-20 4:06 ` KAMEZAWA Hiroyuki
2010-10-20 4:25 ` Greg Thelen
2010-10-20 4:26 ` KAMEZAWA Hiroyuki
2010-10-20 0:48 ` Daisuke Nishimura
2010-10-20 1:14 ` KAMEZAWA Hiroyuki
2010-10-20 2:24 ` KAMEZAWA Hiroyuki
2010-10-20 3:47 ` Daisuke Nishimura
2010-10-19 0:39 ` [PATCH v3 03/11] memcg: create extensible page stat update routines Greg Thelen
2010-10-19 0:47 ` KAMEZAWA Hiroyuki
2010-10-19 4:52 ` Daisuke Nishimura
2010-10-19 0:39 ` [PATCH v3 04/11] memcg: add lock to synchronize page accounting and migration Greg Thelen
2010-10-19 0:45 ` KAMEZAWA Hiroyuki
2010-10-19 4:43 ` [RFC][PATCH 1/2] memcg: move_account optimization by reduct put,get page (Re: " KAMEZAWA Hiroyuki
2010-10-19 4:45 ` [RFC][PATCH 2/2] memcg: move_account optimization by reduce locks " KAMEZAWA Hiroyuki
2010-10-19 1:17 ` Minchan Kim
2010-10-19 5:03 ` Daisuke Nishimura [this message]
2010-10-19 0:39 ` [PATCH v3 05/11] memcg: add dirty page accounting infrastructure Greg Thelen
2010-10-19 0:49 ` KAMEZAWA Hiroyuki
2010-10-20 0:53 ` Daisuke Nishimura
2010-10-19 0:39 ` [PATCH v3 06/11] memcg: add kernel calls for memcg dirty page stats Greg Thelen
2010-10-19 0:51 ` KAMEZAWA Hiroyuki
2010-10-19 7:03 ` Daisuke Nishimura
2010-10-19 0:39 ` [PATCH v3 07/11] memcg: add dirty limits to mem_cgroup Greg Thelen
2010-10-19 0:53 ` KAMEZAWA Hiroyuki
2010-10-20 0:50 ` Daisuke Nishimura
2010-10-20 4:08 ` Greg Thelen
2010-10-19 0:39 ` [PATCH v3 08/11] memcg: CPU hotplug lockdep warning fix Greg Thelen
2010-10-19 0:54 ` KAMEZAWA Hiroyuki
2010-10-20 3:47 ` Daisuke Nishimura
2010-10-19 0:39 ` [PATCH v3 09/11] memcg: add cgroupfs interface to memcg dirty limits Greg Thelen
2010-10-19 0:56 ` KAMEZAWA Hiroyuki
2010-10-20 3:31 ` Daisuke Nishimura
2010-10-20 3:44 ` KAMEZAWA Hiroyuki
2010-10-20 3:46 ` Daisuke Nishimura
2010-10-19 0:39 ` [PATCH v3 10/11] writeback: make determine_dirtyable_memory() static Greg Thelen
2010-10-19 0:57 ` KAMEZAWA Hiroyuki
2010-10-20 3:47 ` Daisuke Nishimura
2010-10-19 0:39 ` [PATCH v3 11/11] memcg: check memcg dirty limits in page writeback Greg Thelen
2010-10-19 1:00 ` KAMEZAWA Hiroyuki
2010-10-20 4:18 ` KAMEZAWA Hiroyuki
2010-10-20 4:33 ` Greg Thelen
2010-10-20 4:33 ` KAMEZAWA Hiroyuki
2010-10-20 4:34 ` Daisuke Nishimura
2010-10-20 5:25 ` Daisuke Nishimura
2010-10-20 3:21 ` [PATCH][memcg+dirtylimit] Fix overwriting global vm dirty limit setting by memcg (Re: [PATCH v3 00/11] memcg: per cgroup dirty page accounting KAMEZAWA Hiroyuki
2010-10-20 4:14 ` KAMEZAWA Hiroyuki
2010-10-20 5:02 ` [PATCH v2][memcg+dirtylimit] " KAMEZAWA Hiroyuki
2010-10-20 6:09 ` Daisuke Nishimura
2010-10-20 14:35 ` Minchan Kim
2010-10-21 0:10 ` KAMEZAWA Hiroyuki
2010-10-24 18:44 ` Greg Thelen
2010-10-25 0:24 ` KAMEZAWA Hiroyuki
2010-10-25 2:00 ` Daisuke Nishimura
2010-10-25 7:03 ` Ciju Rajan K
2010-10-25 7:08 ` 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=20101019140338.9fd664bc.nishimura@mxp.nes.nec.co.jp \
--to=nishimura@mxp.nes.nec.co.jp \
--cc=akpm@linux-foundation.org \
--cc=arighi@develer.com \
--cc=balbir@linux.vnet.ibm.com \
--cc=ciju@linux.vnet.ibm.com \
--cc=containers@lists.osdl.org \
--cc=gthelen@google.com \
--cc=kamezawa.hiroyu@jp.fujitsu.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=minchan.kim@gmail.com \
--cc=rientjes@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
all inboxes | Powered by JetHome®