[PATCH v2 0/5] mm, swap: misc cleanup and bugfix

[PATCH v2 0/5] mm, swap: misc cleanup and bugfix

[Kairui Song]

A few cleanups and a bugfix that are either suitable after the swap
table phase I or found during code review.

Patch 1 is a bugfix and needs to be included in the stable branch,
the rest have no behavior change.

---
Changes in v2:
- Update commit message for patch 1, it's a sub-optimal fix and a better
  fix can be done later. [ Chris Li ]
- Fix a lock balance issue in patch 1. [ YoungJun Park ]
- Add a trivial cleanup patch to remove an unused argument,
  no behavior change.
- Update kernel doc.
- Fix minor issue with commit message [ Nhat Pham ]
- Link to v1: https://lore.kernel.org/r/20251007-swap-clean-after-swap-table-p1-v1-0-74860ef8ba74@tencent.com

---
Kairui Song (5):
      mm, swap: do not perform synchronous discard during allocation
      mm, swap: rename helper for setup bad slots
      mm, swap: cleanup swap entry allocation parameter
      mm/migrate, swap: drop usage of folio_index
      mm, swap: remove redundant argument for isolating a cluster

 include/linux/swap.h |  4 +--
 mm/migrate.c         |  4 +--
 mm/shmem.c           |  2 +-
 mm/swap.h            | 21 ----------------
 mm/swapfile.c        | 71 +++++++++++++++++++++++++++++++++++-----------------
 mm/vmscan.c          |  4 +--
 6 files changed, 55 insertions(+), 51 deletions(-)
---
base-commit: 5b5c3e53c939318f6a0698c895c7ec40758bff6a
change-id: 20251007-swap-clean-after-swap-table-p1-b9a7635ee3fa

Best regards,
-- 
Kairui Song <kasong@tencent.com>

[PATCH v2 2/5] mm, swap: rename helper for setup bad slots

[Kairui Song]

From: Kairui Song <kasong@tencent.com>

The name inc_cluster_info_page is very confusing, as this helper is only
used during swapon to mark bad slots. Rename it properly and turn the
VM_BUG_ON in it into WARN_ON to expose more potential issues. Swapon is
a cold path, so adding more checks should be a good idea.

No feature change except new WARN_ON.

Acked-by: Chris Li <chrisl@kernel.org>
Acked-by: Nhat Pham <nphamcs@gmail.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Kairui Song <kasong@tencent.com>
---
 mm/swapfile.c | 22 +++++++++++-----------
 1 file changed, 11 insertions(+), 11 deletions(-)

diff --git a/mm/swapfile.c b/mm/swapfile.c
index 35038a0871e1..781a70dfcff1 100644
--- a/mm/swapfile.c
+++ b/mm/swapfile.c
@@ -751,14 +751,14 @@ static void relocate_cluster(struct swap_info_struct *si,
 }
 
 /*
- * The cluster corresponding to page_nr will be used. The cluster will not be
- * added to free cluster list and its usage counter will be increased by 1.
- * Only used for initialization.
+ * The cluster corresponding to @offset will be accounted as having one bad
+ * slot. The cluster will not be added to the free cluster list, and its
+ * usage counter will be increased by 1. Only used for initialization.
  */
-static int inc_cluster_info_page(struct swap_info_struct *si,
-	struct swap_cluster_info *cluster_info, unsigned long page_nr)
+static int swap_cluster_setup_bad_slot(struct swap_cluster_info *cluster_info,
+				       unsigned long offset)
 {
-	unsigned long idx = page_nr / SWAPFILE_CLUSTER;
+	unsigned long idx = offset / SWAPFILE_CLUSTER;
 	struct swap_table *table;
 	struct swap_cluster_info *ci;
 
@@ -772,8 +772,8 @@ static int inc_cluster_info_page(struct swap_info_struct *si,
 
 	ci->count++;
 
-	VM_BUG_ON(ci->count > SWAPFILE_CLUSTER);
-	VM_BUG_ON(ci->flags);
+	WARN_ON(ci->count > SWAPFILE_CLUSTER);
+	WARN_ON(ci->flags);
 
 	return 0;
 }
@@ -3396,7 +3396,7 @@ static struct swap_cluster_info *setup_clusters(struct swap_info_struct *si,
 	 * See setup_swap_map(): header page, bad pages,
 	 * and the EOF part of the last cluster.
 	 */
-	err = inc_cluster_info_page(si, cluster_info, 0);
+	err = swap_cluster_setup_bad_slot(cluster_info, 0);
 	if (err)
 		goto err;
 	for (i = 0; i < swap_header->info.nr_badpages; i++) {
@@ -3404,12 +3404,12 @@ static struct swap_cluster_info *setup_clusters(struct swap_info_struct *si,
 
 		if (page_nr >= maxpages)
 			continue;
-		err = inc_cluster_info_page(si, cluster_info, page_nr);
+		err = swap_cluster_setup_bad_slot(cluster_info, page_nr);
 		if (err)
 			goto err;
 	}
 	for (i = maxpages; i < round_up(maxpages, SWAPFILE_CLUSTER); i++) {
-		err = inc_cluster_info_page(si, cluster_info, i);
+		err = swap_cluster_setup_bad_slot(cluster_info, i);
 		if (err)
 			goto err;
 	}

-- 
2.51.0

[PATCH v2 3/5] mm, swap: cleanup swap entry allocation parameter

[Kairui Song]

From: Kairui Song <kasong@tencent.com>

We no longer need this GFP parameter after commit 8578e0c00dcf ("mm, swap:
use the swap table for the swap cache and switch API"). Before that
commit the GFP parameter is already almost identical for all callers, so
nothing changed by that commit. Swap table just moved the GFP to lower
layer and make it more defined and changes depend on atomic or sleep
allocation.

Now this parameter is no longer used, just remove it. No behavior
change.

Acked-by: Chris Li <chrisl@kernel.org>
Acked-by: Nhat Pham <nphamcs@gmail.com>
Reviewed-by: Baolin Wang <baolin.wang@linux.alibaba.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Kairui Song <kasong@tencent.com>
---
 include/linux/swap.h | 4 ++--
 mm/shmem.c           | 2 +-
 mm/swapfile.c        | 3 +--
 mm/vmscan.c          | 4 ++--
 4 files changed, 6 insertions(+), 7 deletions(-)

diff --git a/include/linux/swap.h b/include/linux/swap.h
index e818fbade1e2..a4b264817735 100644
--- a/include/linux/swap.h
+++ b/include/linux/swap.h
@@ -462,7 +462,7 @@ static inline long get_nr_swap_pages(void)
 }
 
 extern void si_swapinfo(struct sysinfo *);
-int folio_alloc_swap(struct folio *folio, gfp_t gfp_mask);
+int folio_alloc_swap(struct folio *folio);
 bool folio_free_swap(struct folio *folio);
 void put_swap_folio(struct folio *folio, swp_entry_t entry);
 extern swp_entry_t get_swap_page_of_type(int);
@@ -560,7 +560,7 @@ static inline int swp_swapcount(swp_entry_t entry)
 	return 0;
 }
 
-static inline int folio_alloc_swap(struct folio *folio, gfp_t gfp_mask)
+static inline int folio_alloc_swap(struct folio *folio)
 {
 	return -EINVAL;
 }
diff --git a/mm/shmem.c b/mm/shmem.c
index 10682328d54c..7559773ebb30 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -1617,7 +1617,7 @@ int shmem_writeout(struct folio *folio, struct swap_iocb **plug,
 		folio_mark_uptodate(folio);
 	}
 
-	if (!folio_alloc_swap(folio, __GFP_HIGH | __GFP_NOMEMALLOC | __GFP_NOWARN)) {
+	if (!folio_alloc_swap(folio)) {
 		bool first_swapped = shmem_recalc_inode(inode, 0, nr_pages);
 		int error;
 
diff --git a/mm/swapfile.c b/mm/swapfile.c
index 781a70dfcff1..42e2b2759240 100644
--- a/mm/swapfile.c
+++ b/mm/swapfile.c
@@ -1417,7 +1417,6 @@ static bool swap_sync_discard(void)
 /**
  * folio_alloc_swap - allocate swap space for a folio
  * @folio: folio we want to move to swap
- * @gfp: gfp mask for shadow nodes
  *
  * Allocate swap space for the folio and add the folio to the
  * swap cache.
@@ -1425,7 +1424,7 @@ static bool swap_sync_discard(void)
  * Context: Caller needs to hold the folio lock.
  * Return: Whether the folio was added to the swap cache.
  */
-int folio_alloc_swap(struct folio *folio, gfp_t gfp)
+int folio_alloc_swap(struct folio *folio)
 {
 	unsigned int order = folio_order(folio);
 	unsigned int size = 1 << order;
diff --git a/mm/vmscan.c b/mm/vmscan.c
index 92f4ca99b73c..c922bad2b8fd 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -1296,7 +1296,7 @@ static unsigned int shrink_folio_list(struct list_head *folio_list,
 					    split_folio_to_list(folio, folio_list))
 						goto activate_locked;
 				}
-				if (folio_alloc_swap(folio, __GFP_HIGH | __GFP_NOWARN)) {
+				if (folio_alloc_swap(folio)) {
 					int __maybe_unused order = folio_order(folio);
 
 					if (!folio_test_large(folio))
@@ -1312,7 +1312,7 @@ static unsigned int shrink_folio_list(struct list_head *folio_list,
 					}
 #endif
 					count_mthp_stat(order, MTHP_STAT_SWPOUT_FALLBACK);
-					if (folio_alloc_swap(folio, __GFP_HIGH | __GFP_NOWARN))
+					if (folio_alloc_swap(folio))
 						goto activate_locked_split;
 				}
 				/*

-- 
2.51.0

[PATCH v2 4/5] mm/migrate, swap: drop usage of folio_index

[Kairui Song]

From: Kairui Song <kasong@tencent.com>

This helper was used when swap cache was mixed with page cache. Now they
are completely separate from each other, access to the swap cache is all
wrapped by the swap_cache_* helpers, which expect the folio's swap entry
as a parameter.

This helper is no longer used, remove the last redundant user and drop it.

Acked-by: Chris Li <chrisl@kernel.org>
Acked-by: Nhat Pham <nphamcs@gmail.com>
Reviewed-by: Baolin Wang <baolin.wang@linux.alibaba.com>
Signed-off-by: Kairui Song <kasong@tencent.com>
---
 mm/migrate.c |  4 ++--
 mm/swap.h    | 21 ---------------------
 2 files changed, 2 insertions(+), 23 deletions(-)

diff --git a/mm/migrate.c b/mm/migrate.c
index 4324fc01bfce..ceee354ef215 100644
--- a/mm/migrate.c
+++ b/mm/migrate.c
@@ -563,7 +563,7 @@ void pmd_migration_entry_wait(struct mm_struct *mm, pmd_t *pmd)
 static int __folio_migrate_mapping(struct address_space *mapping,
 		struct folio *newfolio, struct folio *folio, int expected_count)
 {
-	XA_STATE(xas, &mapping->i_pages, folio_index(folio));
+	XA_STATE(xas, &mapping->i_pages, folio->index);
 	struct swap_cluster_info *ci = NULL;
 	struct zone *oldzone, *newzone;
 	int dirty;
@@ -716,7 +716,7 @@ EXPORT_SYMBOL(folio_migrate_mapping);
 int migrate_huge_page_move_mapping(struct address_space *mapping,
 				   struct folio *dst, struct folio *src)
 {
-	XA_STATE(xas, &mapping->i_pages, folio_index(src));
+	XA_STATE(xas, &mapping->i_pages, src->index);
 	int rc, expected_count = folio_expected_ref_count(src) + 1;
 
 	if (folio_ref_count(src) != expected_count)
diff --git a/mm/swap.h b/mm/swap.h
index 8d8efdf1297a..d034c13d8dd2 100644
--- a/mm/swap.h
+++ b/mm/swap.h
@@ -445,25 +445,4 @@ static inline int non_swapcache_batch(swp_entry_t entry, int max_nr)
 	return 0;
 }
 #endif /* CONFIG_SWAP */
-
-/**
- * folio_index - File index of a folio.
- * @folio: The folio.
- *
- * For a folio which is either in the page cache or the swap cache,
- * return its index within the address_space it belongs to.  If you know
- * the folio is definitely in the page cache, you can look at the folio's
- * index directly.
- *
- * Return: The index (offset in units of pages) of a folio in its file.
- */
-static inline pgoff_t folio_index(struct folio *folio)
-{
-#ifdef CONFIG_SWAP
-	if (unlikely(folio_test_swapcache(folio)))
-		return swp_offset(folio->swap);
-#endif
-	return folio->index;
-}
-
 #endif /* _MM_SWAP_H */

-- 
2.51.0

[PATCH v2 5/5] mm, swap: remove redundant argument for isolating a cluster

[Kairui Song]

From: Kairui Song <kasong@tencent.com>

The order argument was introduced by an intermediate commit and was then
never used, just remove it.

Signed-off-by: Kairui Song <kasong@tencent.com>
---
 mm/swapfile.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/mm/swapfile.c b/mm/swapfile.c
index 42e2b2759240..214ccd1f69cd 100644
--- a/mm/swapfile.c
+++ b/mm/swapfile.c
@@ -594,7 +594,7 @@ static void __free_cluster(struct swap_info_struct *si, struct swap_cluster_info
  * this returns NULL for an non-empty list.
  */
 static struct swap_cluster_info *isolate_lock_cluster(
-		struct swap_info_struct *si, struct list_head *list, int order)
+		struct swap_info_struct *si, struct list_head *list)
 {
 	struct swap_cluster_info *ci, *found = NULL;
 
@@ -957,7 +957,7 @@ static unsigned int alloc_swap_scan_list(struct swap_info_struct *si,
 	unsigned int found = SWAP_ENTRY_INVALID;
 
 	do {
-		struct swap_cluster_info *ci = isolate_lock_cluster(si, list, order);
+		struct swap_cluster_info *ci = isolate_lock_cluster(si, list);
 		unsigned long offset;
 
 		if (!ci)
@@ -982,7 +982,7 @@ static void swap_reclaim_full_clusters(struct swap_info_struct *si, bool force)
 	if (force)
 		to_scan = swap_usage_in_pages(si) / SWAPFILE_CLUSTER;
 
-	while ((ci = isolate_lock_cluster(si, &si->full_clusters, 0))) {
+	while ((ci = isolate_lock_cluster(si, &si->full_clusters))) {
 		offset = cluster_offset(si, ci);
 		end = min(si->max, offset + SWAPFILE_CLUSTER);
 		to_scan--;

-- 
2.51.0

Re: [PATCH v2 5/5] mm, swap: remove redundant argument for isolating a cluster

[Nhat Pham]

On Thu, Oct 23, 2025 at 11:02 AM Kairui Song <ryncsn@gmail.com> wrote:
>
> From: Kairui Song <kasong@tencent.com>
>
> The order argument was introduced by an intermediate commit and was then
> never used, just remove it.
>
> Signed-off-by: Kairui Song <kasong@tencent.com>

Acked-by: Nhat Pham <nphamcs@gmail.com>

Re: [PATCH v2 0/5] mm, swap: misc cleanup and bugfix

[Lorenzo Stoakes]

On Fri, Oct 24, 2025 at 02:00:38AM +0800, Kairui Song wrote:
> A few cleanups and a bugfix that are either suitable after the swap
> table phase I or found during code review.
>
> Patch 1 is a bugfix and needs to be included in the stable branch,
> the rest have no behavior change.
>
> ---
> Changes in v2:
> - Update commit message for patch 1, it's a sub-optimal fix and a better
>   fix can be done later. [ Chris Li ]
> - Fix a lock balance issue in patch 1. [ YoungJun Park ]
> - Add a trivial cleanup patch to remove an unused argument,
>   no behavior change.
> - Update kernel doc.
> - Fix minor issue with commit message [ Nhat Pham ]
> - Link to v1: https://lore.kernel.org/r/20251007-swap-clean-after-swap-table-p1-v1-0-74860ef8ba74@tencent.com
>
> ---
> Kairui Song (5):
>       mm, swap: do not perform synchronous discard during allocation

FYI For some reason this commit is not present on lore, see [0]

[0]: https://lore.kernel.org/all/20251024-swap-clean-after-swap-table-p1-v2-0-a709469052e7@tencent.com/

>       mm, swap: rename helper for setup bad slots
>       mm, swap: cleanup swap entry allocation parameter
>       mm/migrate, swap: drop usage of folio_index
>       mm, swap: remove redundant argument for isolating a cluster
>
>  include/linux/swap.h |  4 +--
>  mm/migrate.c         |  4 +--
>  mm/shmem.c           |  2 +-
>  mm/swap.h            | 21 ----------------
>  mm/swapfile.c        | 71 +++++++++++++++++++++++++++++++++++-----------------
>  mm/vmscan.c          |  4 +--
>  6 files changed, 55 insertions(+), 51 deletions(-)
> ---
> base-commit: 5b5c3e53c939318f6a0698c895c7ec40758bff6a
> change-id: 20251007-swap-clean-after-swap-table-p1-b9a7635ee3fa
>
> Best regards,
> --
> Kairui Song <kasong@tencent.com>
>

Re: [PATCH v2 0/5] mm, swap: misc cleanup and bugfix

[Kairui Song]

On Fri, Oct 24, 2025 at 9:18 PM Lorenzo Stoakes
<lorenzo.stoakes@oracle.com> wrote:
>
> On Fri, Oct 24, 2025 at 02:00:38AM +0800, Kairui Song wrote:
> > A few cleanups and a bugfix that are either suitable after the swap
> > table phase I or found during code review.
> >
> > Patch 1 is a bugfix and needs to be included in the stable branch,
> > the rest have no behavior change.
> >
> > ---
> > Changes in v2:
> > - Update commit message for patch 1, it's a sub-optimal fix and a better
> >   fix can be done later. [ Chris Li ]
> > - Fix a lock balance issue in patch 1. [ YoungJun Park ]
> > - Add a trivial cleanup patch to remove an unused argument,
> >   no behavior change.
> > - Update kernel doc.
> > - Fix minor issue with commit message [ Nhat Pham ]
> > - Link to v1: https://lore.kernel.org/r/20251007-swap-clean-after-swap-table-p1-v1-0-74860ef8ba74@tencent.com
> >
> > ---
> > Kairui Song (5):
> >       mm, swap: do not perform synchronous discard during allocation
>
> FYI For some reason this commit is not present on lore, see [0]
>
> [0]: https://lore.kernel.org/all/20251024-swap-clean-after-swap-table-p1-v2-0-a709469052e7@tencent.com/

Thanks for letting me know, strangely, it is here:
https://lkml.kernel.org/r/20251024-swap-clean-after-swap-table-p1-v2-1-c5b0e1092927@tencent.com

But the In-reply-to id is wrong. I'm using b4 and somehow patch 1 was
blocked by gmail's SMTP so I had to try to resend patch 1 again,
something went wrong with that part. I'll try to find out the problem
and avoid that from happening again.

I'm seeing that patch 1 is being merged into mm tree just fine, I
guess that should be OK.

If anyone is reading the threads, this url above should be helpful.

Re: [PATCH v2 0/5] mm, swap: misc cleanup and bugfix

[Baoquan He]

On 10/24/25 at 09:26pm, Kairui Song wrote:
> On Fri, Oct 24, 2025 at 9:18 PM Lorenzo Stoakes
> <lorenzo.stoakes@oracle.com> wrote:
> >
> > On Fri, Oct 24, 2025 at 02:00:38AM +0800, Kairui Song wrote:
> > > A few cleanups and a bugfix that are either suitable after the swap
> > > table phase I or found during code review.
> > >
> > > Patch 1 is a bugfix and needs to be included in the stable branch,
> > > the rest have no behavior change.
> > >
> > > ---
> > > Changes in v2:
> > > - Update commit message for patch 1, it's a sub-optimal fix and a better
> > >   fix can be done later. [ Chris Li ]
> > > - Fix a lock balance issue in patch 1. [ YoungJun Park ]
> > > - Add a trivial cleanup patch to remove an unused argument,
> > >   no behavior change.
> > > - Update kernel doc.
> > > - Fix minor issue with commit message [ Nhat Pham ]
> > > - Link to v1: https://lore.kernel.org/r/20251007-swap-clean-after-swap-table-p1-v1-0-74860ef8ba74@tencent.com
> > >
> > > ---
> > > Kairui Song (5):
> > >       mm, swap: do not perform synchronous discard during allocation
> >
> > FYI For some reason this commit is not present on lore, see [0]
> >
> > [0]: https://lore.kernel.org/all/20251024-swap-clean-after-swap-table-p1-v2-0-a709469052e7@tencent.com/
> 
> Thanks for letting me know, strangely, it is here:
> https://lkml.kernel.org/r/20251024-swap-clean-after-swap-table-p1-v2-1-c5b0e1092927@tencent.com

I don't receive the patch 1/5, thanks to the link.

> 
> But the In-reply-to id is wrong. I'm using b4 and somehow patch 1 was
> blocked by gmail's SMTP so I had to try to resend patch 1 again,
> something went wrong with that part. I'll try to find out the problem
> and avoid that from happening again.
> 
> I'm seeing that patch 1 is being merged into mm tree just fine, I
> guess that should be OK.
> 
> If anyone is reading the threads, this url above should be helpful.
>