All of lore.kernel.org
 help / color / mirror / Atom feed
From: Yasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com>
To: "H. Peter Anvin" <hpa@zytor.com>, Tang Chen <tangchen@cn.fujitsu.com>
Cc: Bob Liu <lliubbo@gmail.com>, <akpm@linux-foundation.org>,
	<rob@landley.net>, <laijs@cn.fujitsu.com>, <wency@cn.fujitsu.com>,
	<linfeng@cn.fujitsu.com>, <jiang.liu@huawei.com>,
	<yinghai@kernel.org>, <kosaki.motohiro@jp.fujitsu.com>,
	<minchan.kim@gmail.com>, <mgorman@suse.de>, <rientjes@google.com>,
	<rusty@rustcorp.com.au>, <linux-kernel@vger.kernel.org>,
	<linux-mm@kvack.org>, <linux-doc@vger.kernel.org>
Subject: Re: [PATCH v2 0/5] Add movablecore_map boot option
Date: Tue, 27 Nov 2012 18:59:39 +0900	[thread overview]
Message-ID: <50B48F0B.60704@jp.fujitsu.com> (raw)
In-Reply-To: <50B47EB7.20000@zytor.com>

Hi HPA and Tang,

2012/11/27 17:49, H. Peter Anvin wrote:
> On 11/27/2012 12:29 AM, Tang Chen wrote:
>> Another approach is like the following:
>> movable_node = 1,3-5,8
>> This could set all the memory on the nodes to be movable. And the rest
>> of memory works as usual. But movablecore_map is more flexible.
>
> ... but *much* harder for users, so movable_node is better in most cases.

It seems that movable_node is easier to use than movablecore_map.
But I do not think movable_node is better because the node number is set
by OS and changed easily.


For exmaple:
If system has 4 nodes and we set moveble_node=2, we can hot remove node2.

    node0   node1   node2   node3
   +-----+ +-----+ +-----+ +-----+
   |     | |     | |/////| |     |
   |     | |     | |/////| |     |
   |     | |     | |/////| |     |
   |     | |     | |/////| |     |
   +-----+ +-----+ +-----+ +-----+
                   movable
                    node

But if we hot remove node2 and reboot the system, node3 is changed to node2
and set to movable node.

    node0   node1           node2
   +-----+ +-----+         +-----+
   |     | |     |         |/////|
   |     | |     |         |/////|
   |     | |     |         |/////|
   |     | |     |         |/////|
   +-----+ +-----+         +-----+
                           movable
                            node

Originally, node3 is not movable node. Changing the node attribution to
movable node is not intended. So if user uses movable_node,
user must confirm whether boot option is correctly set at hotplug.

But memory range is set by firmware and not changed. So if we set node2
as movable node by movablecore_map, the issue does not occur.

Thanks,
Yasuaki Ishimatsu

>
> 	-hpa
>



WARNING: multiple messages have this Message-ID (diff)
From: Yasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com>
To: "H. Peter Anvin" <hpa@zytor.com>, Tang Chen <tangchen@cn.fujitsu.com>
Cc: Bob Liu <lliubbo@gmail.com>,
	akpm@linux-foundation.org, rob@landley.net, laijs@cn.fujitsu.com,
	wency@cn.fujitsu.com, linfeng@cn.fujitsu.com,
	jiang.liu@huawei.com, yinghai@kernel.org,
	kosaki.motohiro@jp.fujitsu.com, minchan.kim@gmail.com,
	mgorman@suse.de, rientjes@google.com, rusty@rustcorp.com.au,
	linux-kernel@vger.kernel.org, linux-mm@kvack.org,
	linux-doc@vger.kernel.org
Subject: Re: [PATCH v2 0/5] Add movablecore_map boot option
Date: Tue, 27 Nov 2012 18:59:39 +0900	[thread overview]
Message-ID: <50B48F0B.60704@jp.fujitsu.com> (raw)
In-Reply-To: <50B47EB7.20000@zytor.com>

Hi HPA and Tang,

2012/11/27 17:49, H. Peter Anvin wrote:
> On 11/27/2012 12:29 AM, Tang Chen wrote:
>> Another approach is like the following:
>> movable_node = 1,3-5,8
>> This could set all the memory on the nodes to be movable. And the rest
>> of memory works as usual. But movablecore_map is more flexible.
>
> ... but *much* harder for users, so movable_node is better in most cases.

It seems that movable_node is easier to use than movablecore_map.
But I do not think movable_node is better because the node number is set
by OS and changed easily.


For exmaple:
If system has 4 nodes and we set moveble_node=2, we can hot remove node2.

    node0   node1   node2   node3
   +-----+ +-----+ +-----+ +-----+
   |     | |     | |/////| |     |
   |     | |     | |/////| |     |
   |     | |     | |/////| |     |
   |     | |     | |/////| |     |
   +-----+ +-----+ +-----+ +-----+
                   movable
                    node

But if we hot remove node2 and reboot the system, node3 is changed to node2
and set to movable node.

    node0   node1           node2
   +-----+ +-----+         +-----+
   |     | |     |         |/////|
   |     | |     |         |/////|
   |     | |     |         |/////|
   |     | |     |         |/////|
   +-----+ +-----+         +-----+
                           movable
                            node

Originally, node3 is not movable node. Changing the node attribution to
movable node is not intended. So if user uses movable_node,
user must confirm whether boot option is correctly set at hotplug.

But memory range is set by firmware and not changed. So if we set node2
as movable node by movablecore_map, the issue does not occur.

Thanks,
Yasuaki Ishimatsu

>
> 	-hpa
>


--
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:[~2012-11-27 10:00 UTC|newest]

Thread overview: 170+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-11-23 10:44 [PATCH v2 0/5] Add movablecore_map boot option Tang Chen
2012-11-23 10:44 ` Tang Chen
2012-11-23 10:44 ` [PATCH v2 1/5] x86: get pg_data_t's memory from other node Tang Chen
2012-11-23 10:44   ` Tang Chen
2012-11-24  1:19   ` Jiang Liu
2012-11-24  1:19     ` Jiang Liu
2012-11-26  1:19     ` Tang Chen
2012-11-26  1:19       ` Tang Chen
2012-12-02 15:11   ` Jiang Liu
2012-12-02 15:11     ` Jiang Liu
2012-11-23 10:44 ` [PATCH v2 2/5] page_alloc: add movable_memmap kernel parameter Tang Chen
2012-11-23 10:44   ` Tang Chen
2012-11-23 10:44 ` [PATCH v2 3/5] page_alloc: Introduce zone_movable_limit[] to keep movable limit for nodes Tang Chen
2012-11-23 10:44   ` Tang Chen
2012-12-05 15:46   ` Jiang Liu
2012-12-05 15:46     ` Jiang Liu
2012-12-06  1:20     ` Tang Chen
2012-12-06  1:20       ` Tang Chen
2012-11-23 10:44 ` [PATCH v2 4/5] page_alloc: Make movablecore_map has higher priority Tang Chen
2012-11-23 10:44   ` Tang Chen
2012-12-05 15:43   ` Jiang Liu
2012-12-05 15:43     ` Jiang Liu
2012-12-06  1:26     ` Tang Chen
2012-12-06  1:26       ` Tang Chen
2012-12-06  2:26       ` Jiang Liu
2012-12-06  2:26         ` Jiang Liu
2012-12-06  2:51         ` Jianguo Wu
2012-12-06  2:51           ` Jianguo Wu
2012-12-06  2:57           ` Tang Chen
2012-12-06  2:57             ` Tang Chen
2012-12-09  8:10         ` Tang Chen
2012-12-09  8:10           ` Tang Chen
2012-12-10  2:15           ` Jiang Liu
2012-12-10  2:15             ` Jiang Liu
2012-11-23 10:44 ` [PATCH v2 5/5] page_alloc: Bootmem limit with movablecore_map Tang Chen
2012-11-23 10:44   ` Tang Chen
2012-11-26 12:22   ` wujianguo
2012-11-26 12:22     ` wujianguo
2012-11-26 12:53     ` Tang Chen
2012-11-26 12:53       ` Tang Chen
2012-11-26 12:40   ` wujianguo
2012-11-26 12:40     ` wujianguo
2012-11-26 13:15     ` Tang Chen
2012-11-26 13:15       ` Tang Chen
2012-11-26 15:48       ` H. Peter Anvin
2012-11-26 15:48         ` H. Peter Anvin
2012-11-27  0:58         ` Jianguo Wu
2012-11-27  0:58           ` Jianguo Wu
2012-11-27  3:19           ` Wen Congyang
2012-11-27  3:19             ` Wen Congyang
2012-11-27  3:22             ` Jianguo Wu
2012-11-27  3:22               ` Jianguo Wu
2012-11-27  3:34               ` Wen Congyang
2012-11-27  3:34                 ` Wen Congyang
2012-11-27  1:12         ` Jiang Liu
2012-11-27  1:12           ` Jiang Liu
2012-11-27  1:20           ` H. Peter Anvin
2012-11-27  1:20             ` H. Peter Anvin
2012-11-27  3:15         ` Wen Congyang
2012-11-27  3:15           ` Wen Congyang
2012-11-27  5:31           ` H. Peter Anvin
2012-11-27  5:31             ` H. Peter Anvin
2012-12-06 17:28             ` Jiang Liu
2012-12-06 17:28               ` Jiang Liu
2012-12-06 17:41               ` H. Peter Anvin
2012-12-06 17:41                 ` H. Peter Anvin
2012-12-07  0:18                 ` Jiang Liu
2012-12-07  0:18                   ` Jiang Liu
2012-12-19  9:17     ` Tang Chen
2012-12-19  9:17       ` Tang Chen
2012-11-27  3:10 ` [PATCH v2 0/5] Add movablecore_map boot option wujianguo
2012-11-27  3:10   ` wujianguo
2012-11-27  5:43   ` Tang Chen
2012-11-27  5:43     ` Tang Chen
2012-11-27  6:20     ` H. Peter Anvin
2012-11-27  6:20       ` H. Peter Anvin
2012-11-27  6:47     ` Jianguo Wu
2012-11-27  6:47       ` Jianguo Wu
2012-11-28  3:47   ` Tang Chen
2012-11-28  3:47     ` Tang Chen
2012-11-28  4:01     ` Jiang Liu
2012-11-28  4:01       ` Jiang Liu
2012-11-28  5:21       ` Wen Congyang
2012-11-28  5:21         ` Wen Congyang
2012-11-28  5:17         ` Jiang Liu
2012-11-28  5:17           ` Jiang Liu
2012-11-28  4:53     ` Jianguo Wu
2012-11-28  4:53       ` Jianguo Wu
2012-11-27  8:00 ` Bob Liu
2012-11-27  8:00   ` Bob Liu
2012-11-27  8:29   ` Tang Chen
2012-11-27  8:29     ` Tang Chen
2012-11-27  8:49     ` H. Peter Anvin
2012-11-27  8:49       ` H. Peter Anvin
2012-11-27  9:47       ` Wen Congyang
2012-11-27  9:47         ` Wen Congyang
2012-11-27  9:53         ` H. Peter Anvin
2012-11-27  9:53           ` H. Peter Anvin
2012-11-27  9:59       ` Yasuaki Ishimatsu [this message]
2012-11-27  9:59         ` Yasuaki Ishimatsu
2012-11-27 12:09     ` Bob Liu
2012-11-27 12:09       ` Bob Liu
2012-11-27 12:49       ` Tang Chen
2012-11-27 12:49         ` Tang Chen
2012-11-28  3:24         ` Bob Liu
2012-11-28  3:24           ` Bob Liu
2012-11-28  4:08           ` Jiang Liu
2012-11-28  4:08             ` Jiang Liu
2012-11-28  6:16             ` Tang Chen
2012-11-28  6:16               ` Tang Chen
2012-11-28  7:03               ` Jiang Liu
2012-11-28  7:03                 ` Jiang Liu
2012-11-28  8:29             ` Wen Congyang
2012-11-28  8:29               ` Wen Congyang
2012-11-28  8:28               ` Jiang Liu
2012-11-28  8:28                 ` Jiang Liu
2012-11-28  8:38                 ` Wen Congyang
2012-11-28  8:38                   ` Wen Congyang
2012-11-29  0:43               ` Jaegeuk Hanse
2012-11-29  0:43                 ` Jaegeuk Hanse
2012-11-29  1:24                 ` Tang Chen
2012-11-29  1:24                   ` Tang Chen
2012-11-30  9:20             ` Lai Jiangshan
2012-11-30  9:20               ` Lai Jiangshan
2012-11-28  8:47 ` Jiang Liu
2012-11-28  8:47   ` Jiang Liu
2012-11-28 21:34   ` Luck, Tony
2012-11-28 21:34     ` Luck, Tony
2012-11-28 21:38     ` H. Peter Anvin
2012-11-28 21:38       ` H. Peter Anvin
2012-11-29 11:00       ` Mel Gorman
2012-11-29 11:00         ` Mel Gorman
2012-11-29 16:07         ` H. Peter Anvin
2012-11-29 16:07           ` H. Peter Anvin
2012-11-29 22:41           ` Luck, Tony
2012-11-29 22:41             ` Luck, Tony
2012-11-29 22:45             ` H. Peter Anvin
2012-11-29 22:45               ` H. Peter Anvin
2012-11-30  2:56         ` Jiang Liu
2012-11-30  2:56           ` Jiang Liu
2012-11-30  3:15           ` Yasuaki Ishimatsu
2012-11-30  3:15             ` Yasuaki Ishimatsu
2012-11-30 15:36             ` Jiang Liu
2012-11-30 15:36               ` Jiang Liu
2012-11-30  2:58         ` Luck, Tony
2012-11-30  2:58           ` Luck, Tony
2012-11-30  3:28           ` H. Peter Anvin
2012-11-30  3:28             ` H. Peter Anvin
2012-11-30 10:19           ` Glauber Costa
2012-11-30 10:19             ` Glauber Costa
2012-11-30 10:52           ` Mel Gorman
2012-11-30 10:52             ` Mel Gorman
2012-11-29 10:38     ` Yasuaki Ishimatsu
2012-11-29 10:38       ` Yasuaki Ishimatsu
2012-11-29 11:05       ` Mel Gorman
2012-11-29 11:05         ` Mel Gorman
2012-11-29 15:47       ` Jiang Liu
2012-11-29 15:47         ` Jiang Liu
2012-11-29 15:53       ` Jiang Liu
2012-11-29 15:53         ` Jiang Liu
2012-11-29  1:42   ` Jaegeuk Hanse
2012-11-29  1:42     ` Jaegeuk Hanse
2012-11-29  2:25     ` Jiang Liu
2012-11-29  2:25       ` Jiang Liu
2012-11-29  2:49       ` Wanpeng Li
2012-11-29  2:49       ` Wanpeng Li
2012-11-29  2:59         ` Jiang Liu
2012-11-29  2:59           ` Jiang Liu
2012-11-30 22:27       ` Toshi Kani
2012-11-30 22:27         ` Toshi Kani

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=50B48F0B.60704@jp.fujitsu.com \
    --to=isimatu.yasuaki@jp.fujitsu.com \
    --cc=akpm@linux-foundation.org \
    --cc=hpa@zytor.com \
    --cc=jiang.liu@huawei.com \
    --cc=kosaki.motohiro@jp.fujitsu.com \
    --cc=laijs@cn.fujitsu.com \
    --cc=linfeng@cn.fujitsu.com \
    --cc=linux-doc@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=lliubbo@gmail.com \
    --cc=mgorman@suse.de \
    --cc=minchan.kim@gmail.com \
    --cc=rientjes@google.com \
    --cc=rob@landley.net \
    --cc=rusty@rustcorp.com.au \
    --cc=tangchen@cn.fujitsu.com \
    --cc=wency@cn.fujitsu.com \
    --cc=yinghai@kernel.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 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.