From: Kemeng Shi <shikemeng@huaweicloud.com>
To: akpm@linux-foundation.org
Cc: kasong@tencent.com, linux-mm@kvack.org, linux-kernel@vger.kernel.org
Subject: [PATCH 3/9] mm: swap: use __swap_entry_free() to free swap entry in swap_entry_put_locked()
Date: Fri, 14 Mar 2025 05:05:09 +0800 [thread overview]
Message-ID: <20250313210515.9920-4-shikemeng@huaweicloud.com> (raw)
In-Reply-To: <20250313210515.9920-1-shikemeng@huaweicloud.com>
In swap_entry_put_locked(), we will set slot to SWAP_HAS_CACHE before
using swap_entry_range_free to do actual swap entry freeing. This
introduce an unnecessary intermediate state.
By using __swap_entry_free() in swap_entry_put_locked(), we can eliminate
the need to set slot to SWAP_HAS_CACHE.
This change would make the behavior of swap_entry_put_locked() more
consistent with other put() operations which will do actual free work
after put last reference.
Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>
---
mm/swapfile.c | 28 ++++++++++++----------------
1 file changed, 12 insertions(+), 16 deletions(-)
diff --git a/mm/swapfile.c b/mm/swapfile.c
index 7c886f9dd6f9..ba37b9bff586 100644
--- a/mm/swapfile.c
+++ b/mm/swapfile.c
@@ -1367,9 +1367,11 @@ static inline void __swap_entries_free(struct swap_info_struct *si,
}
static unsigned char swap_entry_put_locked(struct swap_info_struct *si,
- unsigned long offset,
+ struct swap_cluster_info *ci,
+ swp_entry_t entry,
unsigned char usage)
{
+ unsigned long offset = swp_offset(entry);
unsigned char count;
unsigned char has_cache;
@@ -1398,10 +1400,9 @@ static unsigned char swap_entry_put_locked(struct swap_info_struct *si,
}
usage = count | has_cache;
- if (usage)
- WRITE_ONCE(si->swap_map[offset], usage);
- else
- WRITE_ONCE(si->swap_map[offset], SWAP_HAS_CACHE);
+ WRITE_ONCE(si->swap_map[offset], usage);
+ if (!usage)
+ __swap_entries_free(si, ci, entry, 1);
return usage;
}
@@ -1480,9 +1481,7 @@ static unsigned char swap_entry_put(struct swap_info_struct *si,
unsigned char usage;
ci = lock_cluster(si, offset);
- usage = swap_entry_put_locked(si, offset, 1);
- if (!usage)
- swap_entry_range_free(si, ci, swp_entry(si->type, offset), 1);
+ usage = swap_entry_put_locked(si, ci, entry, 1);
unlock_cluster(ci);
return usage;
@@ -1562,8 +1561,8 @@ static void cluster_swap_free_nr(struct swap_info_struct *si,
ci = lock_cluster(si, offset);
do {
- if (!swap_entry_put_locked(si, offset, usage))
- swap_entry_range_free(si, ci, swp_entry(si->type, offset), 1);
+ swap_entry_put_locked(si, ci, swp_entry(si->type, offset),
+ usage);
} while (++offset < end);
unlock_cluster(ci);
}
@@ -1607,12 +1606,9 @@ void put_swap_folio(struct folio *folio, swp_entry_t entry)
ci = lock_cluster(si, offset);
if (swap_only_has_cache(si, offset, size))
swap_entry_range_free(si, ci, entry, size);
- else {
- for (int i = 0; i < size; i++, entry.val++) {
- if (!swap_entry_put_locked(si, offset + i, SWAP_HAS_CACHE))
- swap_entry_range_free(si, ci, entry, 1);
- }
- }
+ else
+ for (int i = 0; i < size; i++, entry.val++)
+ swap_entry_put_locked(si, ci, entry, SWAP_HAS_CACHE);
unlock_cluster(ci);
}
--
2.30.0
next prev parent reply other threads:[~2025-03-13 12:09 UTC|newest]
Thread overview: 25+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-03-13 21:05 [PATCH 0/9] Minor cleanups and improvements to swap freeing code Kemeng Shi
2025-03-13 21:05 ` [PATCH 1/9] mm: swap: rename __swap_[entry/entries]_free[_locked] to swap_[entry/entries]_put[_locked] Kemeng Shi
2025-03-14 20:37 ` Tim Chen
2025-03-13 21:05 ` [PATCH 2/9] mm: swap: factor out the actual swap entry freeing logic to new helper Kemeng Shi
2025-03-13 17:42 ` Kairui Song
2025-03-14 7:32 ` Kemeng Shi
2025-03-14 7:47 ` Kairui Song
2025-03-14 8:39 ` Kemeng Shi
2025-03-13 21:05 ` Kemeng Shi [this message]
2025-03-14 20:59 ` [PATCH 3/9] mm: swap: use __swap_entry_free() to free swap entry in swap_entry_put_locked() Tim Chen
2025-03-13 21:05 ` [PATCH 4/9] mm: swap: remove unneeded VM_BUG_ON(*map != SWAP_HAS_CACHE) in swap_entry_range_free() Kemeng Shi
2025-03-14 21:09 ` Tim Chen
2025-03-13 21:05 ` [PATCH 5/9] mm: swap: use swap_entries_free() drop last 1 flag in swap_entries_put_nr() Kemeng Shi
2025-03-14 21:21 ` Tim Chen
2025-03-13 21:05 ` [PATCH 6/9] mm: swap: drop last SWAP_MAP_SHMEM flag in batch " Kemeng Shi
2025-03-14 21:34 ` Tim Chen
2025-03-13 21:05 ` [PATCH 7/9] mm: swap: free each cluster individually in swap_entries_put_map_nr() Kemeng Shi
2025-03-14 20:53 ` Tim Chen
2025-03-17 17:30 ` Tim Chen
2025-03-13 21:05 ` [PATCH 8/9] mm: swap: factor out helper to drop cache of entries within a single cluster Kemeng Shi
2025-03-15 0:24 ` Tim Chen
2025-03-13 21:05 ` [PATCH 9/9] mm: swap: replace cluster_swap_free_nr() with swap_entries_put_[map/cache]() Kemeng Shi
2025-03-17 18:23 ` Tim Chen
2025-03-14 20:27 ` [PATCH 0/9] Minor cleanups and improvements to swap freeing code Tim Chen
2025-03-18 1:29 ` Kemeng Shi
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=20250313210515.9920-4-shikemeng@huaweicloud.com \
--to=shikemeng@huaweicloud.com \
--cc=akpm@linux-foundation.org \
--cc=kasong@tencent.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
/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