linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Balbir Singh <balbir@linux.vnet.ibm.com>
To: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: "linux-mm@kvack.org" <linux-mm@kvack.org>,
	"nishimura@mxp.nes.nec.co.jp" <nishimura@mxp.nes.nec.co.jp>,
	"xemul@openvz.org" <xemul@openvz.org>,
	LKML <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH 9/13] memcg: lookup page cgroup (and remove pointer from struct page)
Date: Wed, 24 Sep 2008 01:31:59 -0700	[thread overview]
Message-ID: <48D9FAFF.8070404@linux.vnet.ibm.com> (raw)
In-Reply-To: <20080924120940.5aea6907.kamezawa.hiroyu@jp.fujitsu.com>

KAMEZAWA Hiroyuki wrote:
> On Wed, 24 Sep 2008 07:39:58 +0530
> "Balbir Singh" <balbir@linux.vnet.ibm.com> wrote:
>>> I'll add FLATMEM/DISCONTIGMEM/SPARSEMEM support directly.
>>> I already have wasted a month on this not-interesting work and want to fix
>>> this soon.
>>>
>> Let's look at the basic requirement, make memory resource controller
>> not suck with 32 bit systems. I have been thinking of about removing
>> page_cgroup from struct page only for 32 bit systems (use radix tree),
>> 32 bit systems can have a maximum of 64GB if PAE is enabled, I suspect
>> radix tree should work there and let the 64 bit systems work as is. If
>> performance is an issue, I would recommend the 32 bit folks upgrade to
>> 64 bit :) Can we build consensus around this approach?
>>
> My thinking is below. (assume 64bit)
> 

assume 64 bit for the calculations below?

>   - remove page_cgroup pointer from struct page allows us to reduce
>     static memory usage at boot by 8bytes/4096bytes if memory cgroup is disabled.
>     This reaches 96MB on my 48 GB box. I think this is big.
>   - pre-allocation of page_cgroup gives us following.
>    Pros.
>       - We are not necessary to be afraid of "failure of kmalloc" and
>         "goes down to memory reclaim at kmalloc"
>         This makes memory resource controller much simpler and robust.
>       - We can know what amount of kernel memory will be used for
>         LRU pages management.
>    Cons.
>       - All page_cgroups are allocated at boot.
>         This reaches 480MB on my 48GB box.
> 
>   But I think we can ignore "Cons.". If we use up memory, we'll use tons of
>   page_cgroup. Considering memory fragmentation caused by allocating a lots of
>   very small object, pre-allocation makes memcg better.

This looks like a good patch. I'll review and test it.

-- 
	Balbir

--
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>

  reply	other threads:[~2008-09-24  8:33 UTC|newest]

Thread overview: 37+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-09-22 10:51 [PATCH 0/13] memory cgroup updates v4 KAMEZAWA Hiroyuki
2008-09-22 10:55 ` [PATCH 1/13] memcg: avoid accounting special mapping KAMEZAWA Hiroyuki
2008-09-22 10:57 ` [PATCH 2/13] memcg: account fault-in swap under lock KAMEZAWA Hiroyuki
2008-09-22 10:58 ` [PATCH 3/13] memcg: nolimit root cgroup KAMEZAWA Hiroyuki
2008-09-22 11:00 ` [PATCH 4/13] memcg: force_empty moving account KAMEZAWA Hiroyuki
2008-09-22 14:23   ` Peter Zijlstra
2008-09-22 14:50   ` kamezawa.hiroyu
2008-09-22 14:56     ` Peter Zijlstra
2008-09-22 15:06     ` kamezawa.hiroyu
2008-09-22 15:32       ` Peter Zijlstra
2008-09-22 15:43       ` kamezawa.hiroyu
2008-09-22 11:02 ` [PATCH 5/13] memcg: cleanup to make mapping null before unchage KAMEZAWA Hiroyuki
2008-09-22 11:03 ` [PATCH 6/13] memcg: optimze per cpu accounting for memcg KAMEZAWA Hiroyuki
2008-09-22 11:05 ` [PATCH 3.5/13] memcg: make page_cgroup flags to be atomic KAMEZAWA Hiroyuki
2008-09-22 11:09 ` [PATCH 3.6/13] memcg: add function to move account KAMEZAWA Hiroyuki
2008-09-24  6:50   ` Daisuke Nishimura
2008-09-24  7:11     ` KAMEZAWA Hiroyuki
2008-09-22 11:12 ` [PATCH 9/13] memcg: lookup page cgroup (and remove pointer from struct page) KAMEZAWA Hiroyuki
2008-09-22 14:52   ` Dave Hansen
2008-09-22 15:14   ` kamezawa.hiroyu
2008-09-22 15:47     ` Dave Hansen
2008-09-22 15:57     ` kamezawa.hiroyu
2008-09-22 16:10       ` Dave Hansen
2008-09-22 17:34       ` kamezawa.hiroyu
2008-09-22 15:47   ` Peter Zijlstra
2008-09-22 16:04   ` kamezawa.hiroyu
2008-09-22 16:06     ` Peter Zijlstra
2008-09-23 23:48   ` KAMEZAWA Hiroyuki
2008-09-24  2:09     ` Balbir Singh
2008-09-24  3:09       ` KAMEZAWA Hiroyuki
2008-09-24  8:31         ` Balbir Singh [this message]
2008-09-24  8:46           ` KAMEZAWA Hiroyuki
2008-09-22 11:13 ` [PATCH 10/13] memcg: page_cgroup look aside table KAMEZAWA Hiroyuki
2008-09-22 11:17 ` [PATCH 11/13] memcg: lazy LRU free (NEW) KAMEZAWA Hiroyuki
2008-09-22 11:22 ` [PATCH 12/13] memcg: lazy LRU add KAMEZAWA Hiroyuki
2008-09-22 11:24 ` [PATCH 13/13] memcg: swap accounting fix KAMEZAWA Hiroyuki
2008-09-22 11:28 ` [PATCH 0/13] memory cgroup updates v4 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=48D9FAFF.8070404@linux.vnet.ibm.com \
    --to=balbir@linux.vnet.ibm.com \
    --cc=kamezawa.hiroyu@jp.fujitsu.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=nishimura@mxp.nes.nec.co.jp \
    --cc=xemul@openvz.org \
    /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