From: David Hildenbrand <david@redhat.com>
To: Jingbo Xu <jefflexu@linux.alibaba.com>,
Shakeel Butt <shakeel.butt@linux.dev>
Cc: Bernd Schubert <bernd.schubert@fastmail.fm>,
Joanne Koong <joannelkoong@gmail.com>, Zi Yan <ziy@nvidia.com>,
miklos@szeredi.hu, linux-fsdevel@vger.kernel.org,
josef@toxicpanda.com, linux-mm@kvack.org, kernel-team@meta.com,
Matthew Wilcox <willy@infradead.org>,
Oscar Salvador <osalvador@suse.de>,
Michal Hocko <mhocko@kernel.org>
Subject: Re: [PATCH v6 4/5] mm/migrate: skip migrating folios under writeback with AS_WRITEBACK_INDETERMINATE mappings
Date: Sat, 21 Dec 2024 17:23:10 +0100 [thread overview]
Message-ID: <67fec986-6a5d-4b1e-a86f-7ecccb1bccf5@redhat.com> (raw)
In-Reply-To: <c163c6ab-6121-427c-ab06-58db2eea671b@linux.alibaba.com>
On 21.12.24 03:28, Jingbo Xu wrote:
>
>
> On 12/21/24 2:01 AM, Shakeel Butt wrote:
>> On Fri, Dec 20, 2024 at 03:49:39PM +0100, David Hildenbrand wrote:
>>>>> I'm wondering if there would be a way to just "cancel" the writeback and
>>>>> mark the folio dirty again. That way it could be migrated, but not
>>>>> reclaimed. At least we could avoid the whole AS_WRITEBACK_INDETERMINATE
>>>>> thing.
>>>>>
>>>>
>>>> That is what I basically meant with short timeouts. Obviously it is not
>>>> that simple to cancel the request and to retry - it would add in quite
>>>> some complexity, if all the issues that arise can be solved at all.
>>>
>>> At least it would keep that out of core-mm.
>>>
>>> AS_WRITEBACK_INDETERMINATE really has weird smell to it ... we should try to
>>> improve such scenarios, not acknowledge and integrate them, then work around
>>> using timeouts that must be manually configured, and ca likely no be default
>>> enabled because it could hurt reasonable use cases :(
>>
>> Just to be clear AS_WRITEBACK_INDETERMINATE is being used in two core-mm
>> parts. First is reclaim and second is compaction/migration. For reclaim,
>> it is a must have as explained by Jingbo in [1] i.e. due to potential
>> self deadlock by fuse server. If I understand you correctly, the main
>> concern you have is its usage in the second case.
>>
>> The reason for adding AS_WRITEBACK_INDETERMINATE in the second case was
>> to avoid untrusted fuse server causing pain to unrelated jobs on the
>> machine (fuse folks please correct me if I am wrong here).
>
> Right, IIUC direct MIGRATE_SYNC migration won't be triggered on the
> memory allocation path, i.e. the fuse server itself won't stumble into
> MIGRATE_SYNC migration.
>
Maybe memory compaction (on higher-order allocations only) could trigger it?
gfp_compaction_allowed() checks __GFP_IO. GFP_KERNEL includes that.
--
Cheers,
David / dhildenb
next prev parent reply other threads:[~2024-12-21 16:23 UTC|newest]
Thread overview: 124+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-11-22 23:23 [PATCH v6 0/5] fuse: remove temp page copies in writeback Joanne Koong
2024-11-22 23:23 ` [PATCH v6 1/5] mm: add AS_WRITEBACK_INDETERMINATE mapping flag Joanne Koong
2024-11-22 23:23 ` [PATCH v6 2/5] mm: skip reclaiming folios in legacy memcg writeback indeterminate contexts Joanne Koong
2024-11-22 23:23 ` [PATCH v6 3/5] fs/writeback: in wait_sb_inodes(), skip wait for AS_WRITEBACK_INDETERMINATE mappings Joanne Koong
2024-11-22 23:23 ` [PATCH v6 4/5] mm/migrate: skip migrating folios under writeback with " Joanne Koong
2024-12-19 13:05 ` David Hildenbrand
2024-12-19 14:19 ` Zi Yan
2024-12-19 15:08 ` Zi Yan
2024-12-19 15:39 ` David Hildenbrand
2024-12-19 15:47 ` Zi Yan
2024-12-19 15:50 ` David Hildenbrand
2024-12-19 15:43 ` Shakeel Butt
2024-12-19 15:47 ` David Hildenbrand
2024-12-19 15:53 ` Shakeel Butt
2024-12-19 15:55 ` Zi Yan
2024-12-19 15:56 ` Bernd Schubert
2024-12-19 16:00 ` Zi Yan
2024-12-19 16:02 ` Zi Yan
2024-12-19 16:09 ` Bernd Schubert
2024-12-19 16:14 ` Zi Yan
2024-12-19 16:26 ` Shakeel Butt
2024-12-19 16:31 ` David Hildenbrand
2024-12-19 16:53 ` Shakeel Butt
2024-12-19 16:22 ` Shakeel Butt
2024-12-19 16:29 ` David Hildenbrand
2024-12-19 16:40 ` Shakeel Butt
2024-12-19 16:41 ` David Hildenbrand
2024-12-19 17:14 ` Shakeel Butt
2024-12-19 17:26 ` David Hildenbrand
2024-12-19 17:30 ` Bernd Schubert
2024-12-19 17:37 ` Shakeel Butt
2024-12-19 17:40 ` Bernd Schubert
2024-12-19 17:44 ` Joanne Koong
2024-12-19 17:54 ` Shakeel Butt
2024-12-20 11:44 ` David Hildenbrand
2024-12-20 12:15 ` Bernd Schubert
2024-12-20 14:49 ` David Hildenbrand
2024-12-20 15:26 ` Bernd Schubert
2024-12-20 18:01 ` Shakeel Butt
2024-12-21 2:28 ` Jingbo Xu
2024-12-21 16:23 ` David Hildenbrand [this message]
2024-12-22 2:47 ` Jingbo Xu
2024-12-24 11:32 ` David Hildenbrand
2024-12-21 16:18 ` David Hildenbrand
2024-12-23 22:14 ` Shakeel Butt
2024-12-24 12:37 ` David Hildenbrand
2024-12-26 15:11 ` Zi Yan
2024-12-26 20:13 ` Shakeel Butt
2024-12-26 22:02 ` Bernd Schubert
2024-12-27 20:08 ` Joanne Koong
2024-12-27 20:32 ` Bernd Schubert
2024-12-30 17:52 ` Joanne Koong
2024-12-30 10:16 ` David Hildenbrand
2024-12-30 18:38 ` Joanne Koong
2024-12-30 19:52 ` David Hildenbrand
2024-12-30 20:11 ` Shakeel Butt
2025-01-02 18:54 ` Joanne Koong
2025-01-03 20:31 ` David Hildenbrand
2025-01-06 10:19 ` Miklos Szeredi
2025-01-06 18:17 ` Shakeel Butt
2025-01-07 8:34 ` David Hildenbrand
2025-01-07 18:07 ` Shakeel Butt
2025-01-09 11:22 ` David Hildenbrand
2025-01-10 20:28 ` Jeff Layton
2025-01-10 21:13 ` David Hildenbrand
2025-01-10 22:00 ` Shakeel Butt
2025-01-13 15:27 ` David Hildenbrand
2025-01-13 21:44 ` Jeff Layton
2025-01-14 8:38 ` Miklos Szeredi
2025-01-14 9:40 ` Miklos Szeredi
2025-01-14 9:55 ` Bernd Schubert
2025-01-14 10:07 ` Miklos Szeredi
2025-01-14 18:07 ` Joanne Koong
2025-01-14 18:58 ` Miklos Szeredi
2025-01-14 19:12 ` Joanne Koong
2025-01-14 20:00 ` Miklos Szeredi
2025-01-14 20:29 ` Jeff Layton
2025-01-14 21:40 ` Bernd Schubert
2025-01-23 16:06 ` Pavel Begunkov
2025-01-14 20:51 ` Joanne Koong
2025-01-24 12:25 ` David Hildenbrand
2025-01-14 15:49 ` Jeff Layton
2025-01-24 12:29 ` David Hildenbrand
2025-01-28 10:16 ` Miklos Szeredi
2025-01-14 15:44 ` Jeff Layton
2025-01-14 18:58 ` Joanne Koong
2025-01-10 23:11 ` Jeff Layton
2025-01-10 20:16 ` Jeff Layton
2025-01-10 20:20 ` David Hildenbrand
2025-01-10 20:43 ` Jeff Layton
2025-01-10 21:00 ` David Hildenbrand
2025-01-10 21:07 ` Jeff Layton
2025-01-10 21:21 ` David Hildenbrand
2025-01-07 16:15 ` Miklos Szeredi
2025-01-08 1:40 ` Jingbo Xu
2024-12-30 20:04 ` Shakeel Butt
2025-01-02 19:59 ` Joanne Koong
2025-01-02 20:26 ` Zi Yan
2024-12-20 21:01 ` Joanne Koong
2024-12-21 16:25 ` David Hildenbrand
2024-12-21 21:59 ` Bernd Schubert
2024-12-23 19:00 ` Joanne Koong
2024-12-26 22:44 ` Bernd Schubert
2024-12-27 18:25 ` Joanne Koong
2024-12-19 17:55 ` Joanne Koong
2024-12-19 18:04 ` Bernd Schubert
2024-12-19 18:11 ` Shakeel Butt
2024-12-20 7:55 ` Jingbo Xu
2025-04-02 21:34 ` Joanne Koong
2025-04-03 3:31 ` Jingbo Xu
2025-04-03 9:18 ` David Hildenbrand
2025-04-03 9:25 ` Bernd Schubert
2025-04-03 9:35 ` Christian Brauner
2025-04-03 19:09 ` Joanne Koong
2025-04-03 20:44 ` David Hildenbrand
2025-04-03 22:04 ` Joanne Koong
2024-11-22 23:23 ` [PATCH v6 5/5] fuse: remove tmp folio for writebacks and internal rb tree Joanne Koong
2024-11-25 9:46 ` Jingbo Xu
2024-12-12 21:55 ` [PATCH v6 0/5] fuse: remove temp page copies in writeback Joanne Koong
2024-12-13 11:52 ` Miklos Szeredi
2024-12-13 16:47 ` Shakeel Butt
2024-12-18 17:37 ` Joanne Koong
2024-12-18 17:44 ` Shakeel Butt
2024-12-18 17:53 ` Joanne Koong
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=67fec986-6a5d-4b1e-a86f-7ecccb1bccf5@redhat.com \
--to=david@redhat.com \
--cc=bernd.schubert@fastmail.fm \
--cc=jefflexu@linux.alibaba.com \
--cc=joannelkoong@gmail.com \
--cc=josef@toxicpanda.com \
--cc=kernel-team@meta.com \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mhocko@kernel.org \
--cc=miklos@szeredi.hu \
--cc=osalvador@suse.de \
--cc=shakeel.butt@linux.dev \
--cc=willy@infradead.org \
--cc=ziy@nvidia.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).