linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Chris Li <chrisl@kernel.org>
To: Kairui Song <ryncsn@gmail.com>
Cc: Andrew Morton <akpm@linux-foundation.org>,
	Kemeng Shi <shikemeng@huaweicloud.com>,
	 Nhat Pham <nphamcs@gmail.com>, Baoquan He <bhe@redhat.com>,
	Barry Song <baohua@kernel.org>,
	 "Huang, Ying" <ying.huang@linux.alibaba.com>,
	linux-mm@kvack.org,  linux-kernel@vger.kernel.org
Subject: Re: [PATCH v2] mm/swapfile.c: introduce function alloc_swap_scan_list()
Date: Mon, 11 Aug 2025 22:58:25 -0700	[thread overview]
Message-ID: <CACePvbV65tjGnWr_zFs5ZHoAX2x=As=rMrwJKCho6c5-Dw7qaA@mail.gmail.com> (raw)
In-Reply-To: <CAMgjq7DnniGUdpQQA0YF+LuMuL9+FCLQJP2Q04G_Hsd1sztv9A@mail.gmail.com>

On Fri, Aug 8, 2025 at 12:52 PM Kairui Song <ryncsn@gmail.com> wrote:
>
> On Fri, Aug 8, 2025 at 3:48 PM Chris Li <chrisl@kernel.org> wrote:
> >
> > This the alloc_swap_scan_list() will scan the whole list or the first
> > cluster.
>
> Hi Chris,
>
> This sentence reads strange to me, but English is not my native
> language so I'm not very sure about it.

Ack, my bad, let me rephrase it.
>
> >
> > This reduces the repeat patterns of isolating a cluster then scanning
> > that cluster. As a result, cluster_alloc_swap_entry() is shorter and shallower.
> >
> > No functional change.
> >
> > Signed-off-by: Chris Li <chrisl@kernel.org>
> > ---
> > This patch goes on top of Kairui's swap improve cluster scan series:
> > https://lore.kernel.org/linux-mm/20250806161748.76651-1-ryncsn@gmail.com/
> > ---
> > Changes in v2:
> > - Adjust change base on Andrew's feedback on int type and break out of
> >   loop.
> > - Link to v1: https://lore.kernel.org/r/20250806-swap-scan-list-v1-1-a5fe2d9340a2@kernel.org
> > ---
> >  mm/swapfile.c | 86 ++++++++++++++++++++++++++++++++---------------------------
> >  1 file changed, 47 insertions(+), 39 deletions(-)
> >
> > diff --git a/mm/swapfile.c b/mm/swapfile.c
> > index 4a0cf4fb348d..f26678d68874 100644
> > --- a/mm/swapfile.c
> > +++ b/mm/swapfile.c
> > @@ -820,6 +820,29 @@ static unsigned int alloc_swap_scan_cluster(struct swap_info_struct *si,
> >         return found;
> >  }
> >
> > +static unsigned int alloc_swap_scan_list(struct swap_info_struct *si,
> > +                                        struct list_head *list,
> > +                                        unsigned int order,
> > +                                        unsigned char usage,
> > +                                        bool scan_all)
> > +{
> > +       unsigned int found = SWAP_ENTRY_INVALID;
> > +
> > +       do {
> > +               struct swap_cluster_info *ci = isolate_lock_cluster(si, list);
> > +               unsigned long offset;
> > +
> > +               if (!ci)
> > +                       break;
> > +               offset = cluster_offset(si, ci);
> > +               found = alloc_swap_scan_cluster(si, ci, offset, order, usage);
> > +               if (found)
> > +                       break;
> > +       } while (scan_all);
> > +
> > +       return found;
> > +}
> > +
> >  static void swap_reclaim_full_clusters(struct swap_info_struct *si, bool force)
> >  {
> >         long to_scan = 1;
> > @@ -913,32 +936,24 @@ static unsigned long cluster_alloc_swap_entry(struct swap_info_struct *si, int o
> >          * to spread out the writes.
> >          */
> >         if (si->flags & SWP_PAGE_DISCARD) {
> > -               ci = isolate_lock_cluster(si, &si->free_clusters);
> > -               if (ci) {
> > -                       found = alloc_swap_scan_cluster(si, ci, cluster_offset(si, ci),
> > -                                                       order, usage);
> > -                       if (found)
> > -                               goto done;
> > -               }
> > +               found = alloc_swap_scan_list(si, &si->free_clusters, order, usage,
> > +                                            false);
> > +               if (found)
> > +                       goto done;
> >         }
> >
> >         if (order < PMD_ORDER) {
> > -               while ((ci = isolate_lock_cluster(si, &si->nonfull_clusters[order]))) {
> > -                       found = alloc_swap_scan_cluster(si, ci, cluster_offset(si, ci),
> > -                                                       order, usage);
> > -                       if (found)
> > -                               goto done;
> > -               }
> > +               found = alloc_swap_scan_list(si, &si->nonfull_clusters[order],
> > +                                            order, usage, 0);
>
> All other alloc_swap_scan_list calls use `false`/`true` but this one
> uses `0` for `scan_all`.
> And it should be `true`, right?

Yes, good catch. Will send out a V3.

>
> > +               if (found)
> > +                       goto done;
> >         }
> >
> >         if (!(si->flags & SWP_PAGE_DISCARD)) {
> > -               ci = isolate_lock_cluster(si, &si->free_clusters);
> > -               if (ci) {
> > -                       found = alloc_swap_scan_cluster(si, ci, cluster_offset(si, ci),
> > -                                                       order, usage);
> > -                       if (found)
> > -                               goto done;
> > -               }
> > +               found = alloc_swap_scan_list(si, &si->free_clusters, order, usage,
> > +                                            false);
> > +               if (found)
> > +                       goto done;
> >         }
> >
> >         /* Try reclaim full clusters if free and nonfull lists are drained */
> > @@ -952,13 +967,10 @@ static unsigned long cluster_alloc_swap_entry(struct swap_info_struct *si, int o
> >                  * failure is not critical. Scanning one cluster still
> >                  * keeps the list rotated and reclaimed (for HAS_CACHE).
> >                  */
> > -               ci = isolate_lock_cluster(si, &si->frag_clusters[order]);
> > -               if (ci) {
> > -                       found = alloc_swap_scan_cluster(si, ci, cluster_offset(si, ci),
> > -                                                       order, usage);
> > -                       if (found)
> > -                               goto done;
> > -               }
> > +               found = alloc_swap_scan_list(si, &si->frag_clusters[order], order,
> > +                                            usage, true);
>
> And it should be `false` here.

Ack.

Chris


      parent reply	other threads:[~2025-08-12  5:58 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-08-08  7:48 Chris Li
2025-08-08 19:51 ` Kairui Song
2025-08-12  4:24   ` Andrew Morton
2025-08-12  4:47     ` Chris Li
2025-08-12  5:58   ` Chris Li [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='CACePvbV65tjGnWr_zFs5ZHoAX2x=As=rMrwJKCho6c5-Dw7qaA@mail.gmail.com' \
    --to=chrisl@kernel.org \
    --cc=akpm@linux-foundation.org \
    --cc=baohua@kernel.org \
    --cc=bhe@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=nphamcs@gmail.com \
    --cc=ryncsn@gmail.com \
    --cc=shikemeng@huaweicloud.com \
    --cc=ying.huang@linux.alibaba.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