From: Kairui Song <ryncsn@gmail.com>
To: David Hildenbrand <david@redhat.com>
Cc: linux-mm@kvack.org, Andrew Morton <akpm@linux-foundation.org>,
Matthew Wilcox <willy@infradead.org>,
Hugh Dickins <hughd@google.com>, Chris Li <chrisl@kernel.org>,
Barry Song <baohua@kernel.org>, Baoquan He <bhe@redhat.com>,
Nhat Pham <nphamcs@gmail.com>,
Kemeng Shi <shikemeng@huaweicloud.com>,
Baolin Wang <baolin.wang@linux.alibaba.com>,
Ying Huang <ying.huang@linux.alibaba.com>,
Johannes Weiner <hannes@cmpxchg.org>,
Yosry Ahmed <yosryahmed@google.com>,
Lorenzo Stoakes <lorenzo.stoakes@oracle.com>,
Zi Yan <ziy@nvidia.com>,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH v2 11/15] mm, swap: use the swap table for the swap cache and switch API
Date: Mon, 8 Sep 2025 23:32:37 +0800 [thread overview]
Message-ID: <CAMgjq7DWOiz85i0aQZFKgdWcBvKLZDHkOT-FCiXwiax1kbSZng@mail.gmail.com> (raw)
In-Reply-To: <CAMgjq7DwOe_OJ4qyfY-2=kmaY4tfvQGaedie+6NvPPX7YD0YGA@mail.gmail.com>
On Mon, Sep 8, 2025 at 11:14 PM Kairui Song <ryncsn@gmail.com> wrote:
>
> On Mon, Sep 8, 2025 at 11:09 PM David Hildenbrand <david@redhat.com> wrote:
> > > +static inline struct swap_cluster_info *swap_cluster_lock(
> > > + struct swap_info_struct *si, pgoff_t offset, bool irq)
> > > +{
> > > + return NULL;
> > > +}
> > > +
> > > +static inline struct swap_cluster_info *swap_cluster_lock_by_folio(
> > > + struct folio *folio)
> >
> > I would probably call that "swap_cluster_get_and_lock" or sth like that ...
> >
> > > +{
> > > + return NULL;
> > > +}
> > > +
> > > +static inline struct swap_cluster_info *swap_cluster_lock_by_folio_irq(
> > > + struct folio *folio)
> > > +{
> >
> > ... then this would become "swap_cluster_get_and_lock_irq"
> >
> >
> > Alterantively, separate the lookup from the locking of the cluster.
> >
> > swap_cluster_from_folio() / folio_get_swap_cluster()
> > swap_cluster_lock()
> > swap_cluster_lock_irq()
> >
> > Which might look cleaner in the end.
>
> That's a very good suggestion.
Hmm, one problem here is swap_cluster_lock's args are the `si` and
`offset`. But swap_cluster_from_folio returns a pointer to the
cluster. Retrieving the offset / si pointer after having the pointer
to the cluster seems redundant and troublesome.
I think maybe swap_cluster_get_and_lock is better. (Not sure if it
looks strange to take a folio as argument).
>
> >
> > [...]
> >
> > > -struct address_space *swapper_spaces[MAX_SWAPFILES] __read_mostly;
> > > -static unsigned int nr_swapper_spaces[MAX_SWAPFILES] __read_mostly;
> > > +struct address_space swap_space __read_mostly = {
> > > + .a_ops = &swap_aops,
> > > +};
> > > +
> > > static bool enable_vma_readahead __read_mostly = true;
> > >
> > > #define SWAP_RA_ORDER_CEILING 5
> > > @@ -83,11 +86,21 @@ void show_swap_cache_info(void)
> > > */
> > > struct folio *swap_cache_get_folio(swp_entry_t entry)
> > > {
> > > - struct folio *folio = filemap_get_folio(swap_address_space(entry),
> > > - swap_cache_index(entry));
> > > - if (IS_ERR(folio))
> > > - return NULL;
> > > - return folio;
> > > +
> >
> > ^ superfluous empty line.
> >
> > [...]
> >
> > >
> > > @@ -420,6 +421,34 @@ static inline unsigned int cluster_offset(struct swap_info_struct *si,
> > > return cluster_index(si, ci) * SWAPFILE_CLUSTER;
> > > }
> > >
> > > +static int swap_table_alloc_table(struct swap_cluster_info *ci)
> >
> > swap_cluster_alloc_table ?
>
> Good idea.
>
> >
> > > +{
> > > + WARN_ON(ci->table);
> > > + ci->table = kzalloc(sizeof(unsigned long) * SWAPFILE_CLUSTER, GFP_KERNEL);
> > > + if (!ci->table)
> > > + return -ENOMEM;
> > > + return 0;
> > > +}
> > > +
> > > +static void swap_cluster_free_table(struct swap_cluster_info *ci)
> > > +{
> > > + unsigned int ci_off;
> > > + unsigned long swp_tb;
> > > +
> > > + if (!ci->table)
> > > + return;
> > > +
> > > + for (ci_off = 0; ci_off < SWAPFILE_CLUSTER; ci_off++) {
> > > + swp_tb = __swap_table_get(ci, ci_off);
> > > + if (!swp_tb_is_null(swp_tb))
> > > + pr_err_once("swap: unclean swap space on swapoff: 0x%lx",
> > > + swp_tb);
> > > + }
> > > +
> > > + kfree(ci->table);
> > > + ci->table = NULL;
> > > +}
> > > +
> > > static void move_cluster(struct swap_info_struct *si,
> > > struct swap_cluster_info *ci, struct list_head *list,
> > > enum swap_cluster_flags new_flags)
> > > @@ -702,6 +731,26 @@ static bool cluster_scan_range(struct swap_info_struct *si,
> > > return true;
> > > }
> > >
> > > +/*
> > > + * Currently, the swap table is not used for count tracking, just
> > > + * do a sanity check here to ensure nothing leaked, so the swap
> > > + * table should be empty upon freeing.
> > > + */
> > > +static void cluster_table_check(struct swap_cluster_info *ci,
> > > + unsigned int start, unsigned int nr)
> >
> > "swap_cluster_assert_table_empty()"
> >
> > or sth like that that makes it clearer what you are checking for.
>
> Agree.
>
> >
> > --
> > Cheers
> >
>
> Thanks!
next prev parent reply other threads:[~2025-09-08 15:33 UTC|newest]
Thread overview: 80+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-09-05 19:13 [PATCH v2 00/15] mm, swap: introduce swap table as swap cache (phase I) Kairui Song
2025-09-05 19:13 ` [PATCH v2 01/15] docs/mm: add document for swap table Kairui Song
2025-09-05 23:58 ` Chris Li
2025-09-06 13:31 ` Kairui Song
2025-09-08 12:35 ` Baoquan He
2025-09-08 14:27 ` Kairui Song
2025-09-08 15:06 ` Baoquan He
2025-09-08 15:01 ` Chris Li
2025-09-08 15:09 ` Baoquan He
2025-09-08 15:52 ` Chris Li
2025-09-05 19:13 ` [PATCH v2 02/15] mm, swap: use unified helper for swap cache look up Kairui Song
2025-09-05 23:59 ` Chris Li
2025-09-08 11:43 ` David Hildenbrand
2025-09-05 19:13 ` [PATCH v2 03/15] mm, swap: fix swap cahe index error when retrying reclaim Kairui Song
2025-09-05 22:40 ` Nhat Pham
2025-09-06 6:30 ` Kairui Song
2025-09-06 1:51 ` Chris Li
2025-09-06 6:28 ` Kairui Song
2025-09-06 11:58 ` Chris Li
2025-09-08 3:08 ` Baolin Wang
2025-09-08 11:45 ` David Hildenbrand
2025-09-05 19:13 ` [PATCH v2 04/15] mm, swap: check page poison flag after locking it Kairui Song
2025-09-06 2:00 ` Chris Li
2025-09-08 12:11 ` David Hildenbrand
2025-09-09 14:54 ` Kairui Song
2025-09-09 15:18 ` David Hildenbrand
2025-09-05 19:13 ` [PATCH v2 05/15] mm, swap: always lock and check the swap cache folio before use Kairui Song
2025-09-06 2:12 ` Chris Li
2025-09-06 6:32 ` Kairui Song
2025-09-08 12:18 ` David Hildenbrand
2025-09-09 14:58 ` Kairui Song
2025-09-09 15:19 ` David Hildenbrand
2025-09-10 12:56 ` Kairui Song
2025-09-05 19:13 ` [PATCH v2 06/15] mm, swap: rename and move some swap cluster definition and helpers Kairui Song
2025-09-06 2:13 ` Chris Li
2025-09-08 3:03 ` Baolin Wang
2025-09-05 19:13 ` [PATCH v2 07/15] mm, swap: tidy up swap device and cluster info helpers Kairui Song
2025-09-06 2:14 ` Chris Li
2025-09-08 12:21 ` David Hildenbrand
2025-09-08 15:01 ` Kairui Song
2025-09-05 19:13 ` [PATCH v2 08/15] mm/shmem, swap: remove redundant error handling for replacing folio Kairui Song
2025-09-08 3:17 ` Baolin Wang
2025-09-08 9:28 ` Kairui Song
2025-09-05 19:13 ` [PATCH v2 09/15] mm, swap: cleanup swap cache API and add kerneldoc Kairui Song
2025-09-06 5:45 ` Chris Li
2025-09-08 0:11 ` Barry Song
2025-09-08 3:23 ` Baolin Wang
2025-09-08 12:23 ` David Hildenbrand
2025-09-05 19:13 ` [PATCH v2 10/15] mm, swap: wrap swap cache replacement with a helper Kairui Song
2025-09-06 7:09 ` Chris Li
2025-09-08 3:41 ` Baolin Wang
2025-09-08 10:44 ` Kairui Song
2025-09-09 1:18 ` Baolin Wang
2025-09-08 12:30 ` David Hildenbrand
2025-09-08 14:20 ` Kairui Song
2025-09-08 14:39 ` David Hildenbrand
2025-09-08 14:49 ` Kairui Song
2025-09-05 19:13 ` [PATCH v2 11/15] mm, swap: use the swap table for the swap cache and switch API Kairui Song
2025-09-06 15:28 ` Chris Li
2025-09-08 15:38 ` Kairui Song
2025-09-07 12:55 ` Klara Modin
2025-09-08 14:34 ` Kairui Song
2025-09-08 15:00 ` Klara Modin
2025-09-08 15:10 ` Kairui Song
2025-09-08 13:45 ` David Hildenbrand
2025-09-08 15:14 ` Kairui Song
2025-09-08 15:32 ` Kairui Song [this message]
2025-09-10 2:53 ` SeongJae Park
2025-09-10 2:56 ` Kairui Song
2025-09-05 19:13 ` [PATCH v2 12/15] mm, swap: mark swap address space ro and add context debug check Kairui Song
2025-09-06 15:35 ` Chris Li
2025-09-08 13:10 ` David Hildenbrand
2025-09-05 19:13 ` [PATCH v2 13/15] mm, swap: remove contention workaround for swap cache Kairui Song
2025-09-06 15:30 ` Chris Li
2025-09-08 13:12 ` David Hildenbrand
2025-09-05 19:13 ` [PATCH v2 14/15] mm, swap: implement dynamic allocation of swap table Kairui Song
2025-09-06 15:45 ` Chris Li
2025-09-08 14:58 ` Kairui Song
2025-09-05 19:13 ` [PATCH v2 15/15] mm, swap: use a single page for swap table when the size fits Kairui Song
2025-09-06 15:48 ` Chris Li
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=CAMgjq7DWOiz85i0aQZFKgdWcBvKLZDHkOT-FCiXwiax1kbSZng@mail.gmail.com \
--to=ryncsn@gmail.com \
--cc=akpm@linux-foundation.org \
--cc=baohua@kernel.org \
--cc=baolin.wang@linux.alibaba.com \
--cc=bhe@redhat.com \
--cc=chrisl@kernel.org \
--cc=david@redhat.com \
--cc=hannes@cmpxchg.org \
--cc=hughd@google.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=lorenzo.stoakes@oracle.com \
--cc=nphamcs@gmail.com \
--cc=shikemeng@huaweicloud.com \
--cc=willy@infradead.org \
--cc=ying.huang@linux.alibaba.com \
--cc=yosryahmed@google.com \
--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