From: Changsheng Liu <liuchangsheng@inspur.com>
To: Yasuaki Ishimatsu <yasu.isimatu@gmail.com>
Cc: Andrew Morton <akpm@linux-foundation.org>,
isimatu.yasuaki@jp.fujitsu.com, vbabka@suse.cz,
linux-mm@kvack.org, linux-kernel@vger.kernel.org,
yanxiaofeng@inspur.com, Changsheng Liu <liuchangcheng@inspur.com>
Subject: Re: [PATCH] Memory hot added,The memory can not been added to movable zone
Date: Tue, 25 Aug 2015 18:25:24 +0800 [thread overview]
Message-ID: <55DC4294.2020407@inspur.com> (raw)
In-Reply-To: <55db6d6d.82d1370a.dd0ff.6055@mx.google.com>
Thanks very much for your review, I can move the memory from normal zone
to movable zone succesfully.
And thank you for let me understand the memory mechanism better.
a?? 2015/8/25 3:15, Yasuaki Ishimatsu a??e??:
> Hi
> On Thu, 20 Aug 2015 14:15:13 +0800
> Changsheng Liu <liuchangsheng@inspur.com> wrote:
>
>> Hi Andrew Morton:
>> First, thanks very much for your review, I will update codes according
>> to your suggestio
>>
>> a?? 2015/8/20 7:50, Andrew Morton a??e??:
>>> On Wed, 19 Aug 2015 04:18:26 -0400 Changsheng Liu <liuchangsheng@inspur.com> wrote:
>>>
>>>> From: Changsheng Liu <liuchangcheng@inspur.com>
>>>>
>>>> When memory hot added, the function should_add_memory_movable
>>>> always return 0,because the movable zone is empty,
>>>> so the memory that hot added will add to normal zone even if
>>>> we want to remove the memory.
>>>> So we change the function should_add_memory_movable,if the user
>>>> config CONFIG_MOVABLE_NODE it will return 1 when
>>>> movable zone is empty
>>> I cleaned this up a bit:
>>>
>>> : Subject: mm: memory hot-add: memory can not been added to movable zone
>>> :
>>> : When memory is hot added, should_add_memory_movable() always returns 0
>>> : because the movable zone is empty, so the memory that was hot added will
>>> : add to the normal zone even if we want to remove the memory.
>>> :
>>> : So we change should_add_memory_movable(): if the user config
>>> : CONFIG_MOVABLE_NODE it will return 1 when the movable zone is empty.
>>>
>>> But I don't understand the "even if we want to remove the memory".
>>> This is hot-add, not hot-remove. What do you mean here?
>> After the system startup, we hot added one memory. After some time
>> we wanted to hot remove the memroy that was hot added,
>> but we could not offline some memory blocks successfully because
>> the memory was added to normal zone defaultly and the value of the file
>> named removable under some memory blocks is 0.
> For this, we prepared online_movable. When memory is onlined by online_movable,
> the memory move from ZONE_NORMAL to ZONE_MOVABLE.
>
> Ex.
> # echo online_movable > /sys/devices/system/memory/memoryXXX/state
>
> Thanks,
> Yasuaki Ishimatsu
>
>> we checked the value of the file under some memory blocks as follows:
>> "cat /sys/devices/system/memory/ memory***/removable"
>> When memory being hot added we let the memory be added to movable
>> zone,
>> so we will be able to hot remove the memory that have been hot added
>>>> --- a/mm/memory_hotplug.c
>>>> +++ b/mm/memory_hotplug.c
>>>> @@ -1198,9 +1198,13 @@ static int should_add_memory_movable(int nid, u64 start, u64 size)
>>>> pg_data_t *pgdat = NODE_DATA(nid);
>>>> struct zone *movable_zone = pgdat->node_zones + ZONE_MOVABLE;
>>>>
>>>> - if (zone_is_empty(movable_zone))
>>>> + if (zone_is_empty(movable_zone)) {
>>>> + #ifdef CONFIG_MOVABLE_NODE
>>>> + return 1;
>>>> + #else
>>>> return 0;
>>>> -
>>>> + #endif
>>>> + }
>>>> if (movable_zone->zone_start_pfn <= start_pfn)
>>>> return 1;
>>> Cleaner:
>>>
>>> --- a/mm/memory_hotplug.c~memory-hot-addedthe-memory-can-not-been-added-to-movable-zone-fix
>>> +++ a/mm/memory_hotplug.c
>>> @@ -1181,13 +1181,9 @@ static int should_add_memory_movable(int
>>> pg_data_t *pgdat = NODE_DATA(nid);
>>> struct zone *movable_zone = pgdat->node_zones + ZONE_MOVABLE;
>>>
>>> - if (zone_is_empty(movable_zone)) {
>>> - #ifdef CONFIG_MOVABLE_NODE
>>> - return 1;
>>> - #else
>>> - return 0;
>>> - #endif
>>> - }
>>> + if (zone_is_empty(movable_zone))
>>> + return IS_ENABLED(CONFIG_MOVABLE_NODE);
>>> +
>>> if (movable_zone->zone_start_pfn <= start_pfn)
>>> return 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>
next prev parent reply other threads:[~2015-08-25 10:26 UTC|newest]
Thread overview: 12+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-08-19 8:18 [PATCH] Memory hot added,The memory can not been added to movable zone Changsheng Liu
2015-08-19 23:50 ` Andrew Morton
2015-08-20 5:59 ` Changsheng Liu
2015-08-20 6:12 ` Changsheng Liu
2015-08-20 6:15 ` Changsheng Liu
2015-08-24 19:15 ` Yasuaki Ishimatsu
2015-08-25 10:25 ` Changsheng Liu [this message]
2015-08-25 11:13 ` Vlastimil Babka
2015-08-25 11:49 ` Changsheng Liu
2015-08-26 0:36 ` Changsheng Liu
2015-08-26 6:12 ` Vlastimil Babka
2015-08-26 7:53 ` Changsheng Liu
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=55DC4294.2020407@inspur.com \
--to=liuchangsheng@inspur.com \
--cc=akpm@linux-foundation.org \
--cc=isimatu.yasuaki@jp.fujitsu.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=liuchangcheng@inspur.com \
--cc=vbabka@suse.cz \
--cc=yanxiaofeng@inspur.com \
--cc=yasu.isimatu@gmail.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;
as well as URLs for NNTP newsgroup(s).