linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Rei Yamamoto <yamamoto.rei@jp.fujitsu.com>
To: linmiaohe@huawei.com
Cc: akpm@linux-foundation.org, aquini@redhat.com, ddutile@redhat.com,
	linux-kernel@vger.kernel.org, linux-mm@kvack.org,
	mgorman@techsingularity.net, vvghjk1234@gmail.com,
	yamamoto.rei@jp.fujitsu.com
Subject: Re: [PATCH] mm, compaction: fast_find_migrateblock() should return pfn in the target zone
Date: Thu, 12 May 2022 10:47:36 +0900	[thread overview]
Message-ID: <20220512014736.16376-1-yamamoto.rei@jp.fujitsu.com> (raw)
In-Reply-To: <07e0ceb8-a637-b011-bbf6-ba760a0b12e4@huawei.com>

On Wed, 11 May 2022 17:26:16 Miaohe Lin wrote:
> On 2022/5/11 15:07, Rei Yamamoto wrote:
>> On Wed, 11 May 2022 14:25:34 Miaohe Lin wrote:
>>> On 2022/5/11 12:43, Rei Yamamoto wrote:
>>>> Prevent returning a pfn outside the target zone in case that not
>>>> aligned with pageblock boundary.
>>>> Otherwise isolate_migratepages_block() would handle pages not in
>>>> the target zone.
>>>>
>>>
>>> IIUC, the sole caller isolate_migratepages will ensure the pfn won't outside
>>> the target zone. So the below code change might not be necessary. Or am I miss
>>> something ?
>> 
>> While block_start_pfn is ensured, this variable is not used as the argument for 
>> isolate_migratepages_block():
>>   -----
>>   static isolate_migrate_t isolate_migratepages(struct compact_control *cc)
>>   {
>>   :
>>           low_pfn = fast_find_migrateblock(cc);
>>           block_start_pfn = pageblock_start_pfn(low_pfn);
>>           if (block_start_pfn < cc->zone->zone_start_pfn)
>>                   block_start_pfn = cc->zone->zone_start_pfn;  <--- block_start_pfn is ensured not outside 
>>                                                                     the target zone
>>   :
>>           block_end_pfn = pageblock_end_pfn(low_pfn);
>>   :
>>           for (; block_end_pfn <= cc->free_pfn;
>>                           fast_find_block = false,
>>                           cc->migrate_pfn = low_pfn = block_end_pfn,
>>                           block_start_pfn = block_end_pfn,
>>                           block_end_pfn += pageblock_nr_pages) {
>>   :
>>                   if (isolate_migratepages_block(cc, low_pfn, block_end_pfn,  <--- low_pfn is passed as 
>>                                                                                    the argument
>
> Sorry, I think you're right. And could you please add the runtime effect of this issue?
>
> Anyway, this patch looks good to me now. Thanks!

Thank you for your review.
The runtime effect is that compaction become unintended behavior.
For example, pages not in the target zone are added to cc->migratepages list in isolate_migratepages_block().
As a result, pages migrate between nodes unintentionally.

Thanks,
Rei


  reply	other threads:[~2022-05-12  2:08 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-05-11  4:43 Rei Yamamoto
2022-05-11  6:25 ` Miaohe Lin
2022-05-11  7:07   ` Rei Yamamoto
2022-05-11  9:26     ` Miaohe Lin
2022-05-12  1:47       ` Rei Yamamoto [this message]
2022-05-12  2:20         ` Andrew Morton
2022-05-12  2:27         ` Miaohe Lin
2022-05-12  4:27           ` Rei Yamamoto
2022-05-12 20:49             ` Andrew Morton
2022-05-13  4:11               ` Rei Yamamoto
2022-05-13 21:01                 ` Andrew Morton
2022-05-16  2:41                   ` Rei Yamamoto
2022-05-12  9:04 ` Mel Gorman
2022-05-13  7:54 ` Oscar Salvador

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=20220512014736.16376-1-yamamoto.rei@jp.fujitsu.com \
    --to=yamamoto.rei@jp.fujitsu.com \
    --cc=akpm@linux-foundation.org \
    --cc=aquini@redhat.com \
    --cc=ddutile@redhat.com \
    --cc=linmiaohe@huawei.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=mgorman@techsingularity.net \
    --cc=vvghjk1234@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