linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Matthew Wilcox <willy@infradead.org>
To: Barry Song <baohua@kernel.org>
Cc: Anatoly Stepanov <stepanov.anatoly@huawei.com>,
	akpm@linux-foundation.org, david@kernel.org, ljs@kernel.org,
	Liam.Howlett@oracle.com, vbabka@kernel.org, rppt@kernel.org,
	surenb@google.com, mhocko@suse.com, wangkefeng.wang@huawei.com,
	yanquanmin1@huawei.com, zuoze1@huawei.com,
	artem.kuzin@huawei.com, gutierrez.asier@huawei-partners.com,
	linux-fsdevel@vger.kernel.org, linux-mm@kvack.org,
	linux-kernel@vger.kernel.org
Subject: Re: [RFC PATCH 2/2] filemap: use high-order folios in filemap sync RA
Date: Thu, 16 Apr 2026 18:58:30 +0100	[thread overview]
Message-ID: <aeEjRtNbQI71I0wA@casper.infradead.org> (raw)
In-Reply-To: <CAGsJ_4yBmrFDHLg8Cf7x0dCLELBp+kY6yoMuP_sHt4Z1wuVrBg@mail.gmail.com>

On Thu, Apr 16, 2026 at 01:30:32PM +0800, Barry Song wrote:
> > Or introduce some dedicated tunable like "sync_mmap_order".
> 
> I guess we could benefit from a small order, such as 1 or 2.
> Order 4 is really too large for many systems, such as Android.
> 
> But it seems Matthew never likes new control knobs?

Correct.  It moves the burden to the sysadmin.  And either there isn't
one (eg Android), or the sysadmin does not have enough information to
set it correctly, and doesn't have the time to know that it even exists.

We need to be self-tuning.  There's no other way.

Now, that control loop could run in user-space.  And if it does, then
we'd need the kernel to expose a control knob for the daemon to tweak.
But then the daemon has to come first, not the control knob with a vague
promise of a daemon to appear in future.  Preferably it's part of some
userspace collection of utilities that already exists like util-linux
or systemd, so we're not relying on people finding & installing some
new thing (that again they won't know they need).

THPs in particular have accumulated way too much cruft over the years,
and there's a huge amount of misinformation floating around on random
webpages about how to configure them.  I think we'd be much better off
just ignoring all of the knobs that we've put in.


      reply	other threads:[~2026-04-16 17:58 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-04-15 19:28 [RFC PATCH 0/2] Use high-order folios in mmap " Anatoly Stepanov
2026-04-15 13:18 ` Matthew Wilcox
2026-04-15 13:33   ` Stepanov Anatoly
2026-04-15 19:28 ` [RFC PATCH 1/2] procfs: add contpte info into smaps Anatoly Stepanov
2026-04-15 12:52   ` David Hildenbrand (Arm)
2026-04-15 19:28 ` [RFC PATCH 2/2] filemap: use high-order folios in filemap sync RA Anatoly Stepanov
2026-04-15 12:06   ` Pedro Falcato
2026-04-15 12:31     ` Stepanov Anatoly
2026-04-15 12:46     ` Stepanov Anatoly
2026-04-16  5:30   ` Barry Song
2026-04-16 17:58     ` Matthew Wilcox [this message]

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=aeEjRtNbQI71I0wA@casper.infradead.org \
    --to=willy@infradead.org \
    --cc=Liam.Howlett@oracle.com \
    --cc=akpm@linux-foundation.org \
    --cc=artem.kuzin@huawei.com \
    --cc=baohua@kernel.org \
    --cc=david@kernel.org \
    --cc=gutierrez.asier@huawei-partners.com \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=ljs@kernel.org \
    --cc=mhocko@suse.com \
    --cc=rppt@kernel.org \
    --cc=stepanov.anatoly@huawei.com \
    --cc=surenb@google.com \
    --cc=vbabka@kernel.org \
    --cc=wangkefeng.wang@huawei.com \
    --cc=yanquanmin1@huawei.com \
    --cc=zuoze1@huawei.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