From: Kairui Song <kasong@tencent.com>

swap_map had a static flexible size, so the last cluster won't be fully
covered, hence the allocator needs to check the scan border to avoid
OOB. But the swap table has a fixed-sized swap table for each cluster,
and the slots beyond the device size are marked as bad slots. The
allocator can simply scan all slots as usual, and any bad slots will be
skipped.

Signed-off-by: Kairui Song <kasong@tencent.com>
---
 mm/swap.h     | 2 +-
 mm/swapfile.c | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/mm/swap.h b/mm/swap.h
index 0a91e21e92b1..cc410b94e91a 100644
--- a/mm/swap.h
+++ b/mm/swap.h
@@ -85,7 +85,7 @@ static inline struct swap_cluster_info *__swap_offset_to_cluster(
 		struct swap_info_struct *si, pgoff_t offset)
 {
 	VM_WARN_ON_ONCE(percpu_ref_is_zero(&si->users)); /* race with swapoff */
-	VM_WARN_ON_ONCE(offset >= si->max);
+	VM_WARN_ON_ONCE(offset >= roundup(si->max, SWAPFILE_CLUSTER));
 	return &si->cluster_info[offset / SWAPFILE_CLUSTER];
 }
 
diff --git a/mm/swapfile.c b/mm/swapfile.c
index 801d8092be51..df2b88c6c67b 100644
--- a/mm/swapfile.c
+++ b/mm/swapfile.c
@@ -945,8 +945,8 @@ static unsigned int alloc_swap_scan_cluster(struct swap_info_struct *si,
 {
 	unsigned int next = SWAP_ENTRY_INVALID, found = SWAP_ENTRY_INVALID;
 	unsigned long start = ALIGN_DOWN(offset, SWAPFILE_CLUSTER);
-	unsigned long end = min(start + SWAPFILE_CLUSTER, si->max);
 	unsigned int order = likely(folio) ? folio_order(folio) : 0;
+	unsigned long end = start + SWAPFILE_CLUSTER;
 	unsigned int nr_pages = 1 << order;
 	bool need_reclaim, ret, usable;
 

-- 
2.52.0

On Tue, Feb 17, 2026 at 12:06 PM Kairui Song via B4 Relay
<devnull+kasong.tencent.com@kernel.org> wrote:
>
> From: Kairui Song <kasong@tencent.com>
>
> swap_map had a static flexible size, so the last cluster won't be fully
> covered, hence the allocator needs to check the scan border to avoid
> OOB. But the swap table has a fixed-sized swap table for each cluster,
> and the slots beyond the device size are marked as bad slots. The
> allocator can simply scan all slots as usual, and any bad slots will be
> skipped.
>
> Signed-off-by: Kairui Song <kasong@tencent.com>

Acked-by: Chris Li <chrisl@kernel.org>

Chris

> ---
>  mm/swap.h     | 2 +-
>  mm/swapfile.c | 2 +-
>  2 files changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/mm/swap.h b/mm/swap.h
> index 0a91e21e92b1..cc410b94e91a 100644
> --- a/mm/swap.h
> +++ b/mm/swap.h
> @@ -85,7 +85,7 @@ static inline struct swap_cluster_info *__swap_offset_to_cluster(
>                 struct swap_info_struct *si, pgoff_t offset)
>  {
>         VM_WARN_ON_ONCE(percpu_ref_is_zero(&si->users)); /* race with swapoff */
> -       VM_WARN_ON_ONCE(offset >= si->max);
> +       VM_WARN_ON_ONCE(offset >= roundup(si->max, SWAPFILE_CLUSTER));
>         return &si->cluster_info[offset / SWAPFILE_CLUSTER];
>  }
>
> diff --git a/mm/swapfile.c b/mm/swapfile.c
> index 801d8092be51..df2b88c6c67b 100644
> --- a/mm/swapfile.c
> +++ b/mm/swapfile.c
> @@ -945,8 +945,8 @@ static unsigned int alloc_swap_scan_cluster(struct swap_info_struct *si,
>  {
>         unsigned int next = SWAP_ENTRY_INVALID, found = SWAP_ENTRY_INVALID;
>         unsigned long start = ALIGN_DOWN(offset, SWAPFILE_CLUSTER);
> -       unsigned long end = min(start + SWAPFILE_CLUSTER, si->max);
>         unsigned int order = likely(folio) ? folio_order(folio) : 0;
> +       unsigned long end = start + SWAPFILE_CLUSTER;
>         unsigned int nr_pages = 1 << order;
>         bool need_reclaim, ret, usable;
>
>
> --
> 2.52.0
>
>