[PATCH v3 00/11] mm: memcontrol: constify the read side of the memcg API

[PATCH v3 00/11] mm: memcontrol: constify the read side of the memcg API

[Tal Zussman]

The memcg accessors, lruvec helpers, and stat readers only read from
the memcg, folio, or lruvec they are given, but take non-const
pointers. Constify them, along with the page_counter readers and the
swap I/O blkg helpers along the way.

This started as a follow-up to [1] that quickly grew in proportions :)

[1]: https://lore.kernel.org/linux-mm/anYqSDnC2BbJXv1h@casper.infradead.org/

---
Changes in v3:
- Patch 1: Constify get_obj_cgroup_from_folio() as well, per Sashiko
- Patch 3: Use container_of_const() in mem_cgroup_get_zone_lru_size(),
  per Sashiko
- Patch 3: Note that lruvec_pgdat() still returns a non-const pgdat and
  does not use container_of_const()
- Patch 8: Constify the local memcg stat readers as well, per Sashiko
- Link to v2: https://patch.msgid.link/20260908-folio_memcg-const-v2-0-811e9f443409@columbia.edu

Changes in v2:
- Patch 3: Constify the folio argument of folio_lruvec_relock_irq() and
  folio_lruvec_relock_irqsave() as well, per Sashiko
- Patch 3: Note that mem_cgroup_lruvec() updates lruvec->pgdat in the
  per-node structure, not the memcg, per Sashiko
- Patches 3 and 8: Use container_of_const() in lruvec_memcg() and the
  lruvec_page_state() family, per Sashiko
- Link to v1: https://patch.msgid.link/20260902-folio_memcg-const-v1-0-e2c1da22246d@columbia.edu

---
Tal Zussman (11):
      mm: memcontrol: take a const folio in folio_memcg() and friends
      mm: memcontrol: constify obj_cgroup_memcg() and friends
      mm: memcontrol: constify the lruvec helpers
      mm/page_io: take a const folio in bio_associate_blkg_from_folio()
      mm: memcontrol: constify the mem_cgroup accessors
      mm: page_counter: constify page_counter_read() and page_counter_margin()
      mm: memcontrol: constify the reclaim protection helpers
      mm: memcontrol: constify the memcg and lruvec stat readers
      mm: memcontrol: constify the swap accounting helpers
      mm: memcontrol: constify mem_cgroup_swappiness() and mem_cgroup_get_max()
      mm: memcontrol: constify the zswap and socket pressure helpers

 include/linux/memcontrol.h   | 212 ++++++++++++++++++++++---------------------
 include/linux/mmzone.h       |   2 +-
 include/linux/page_counter.h |   4 +-
 include/linux/swap.h         |  12 +--
 mm/memcontrol-v1.h           |   6 +-
 mm/memcontrol.c              |  57 ++++++------
 mm/page_counter.c            |   2 +-
 mm/page_io.c                 |  14 ++-
 mm/swap.h                    |   2 +-
 9 files changed, 159 insertions(+), 152 deletions(-)
---
base-commit: b08a65b93426d86e3f354d655d6225397b591877
change-id: 20260807-folio_memcg-const-4a67e40d485b

Best regards,
--  
Tal Zussman <tz2294@columbia.edu>

[PATCH v3 01/11] mm: memcontrol: take a const folio in folio_memcg() and friends

[Tal Zussman]

The folio_memcg() family only reads from the folio, and everything it
calls already takes a const folio. Constify it, along with the page
wrappers built on top of it and get_obj_cgroup_from_folio(), which only
reads the folio's objcg.

Signed-off-by: Tal Zussman <tz2294@columbia.edu>
---
 include/linux/memcontrol.h | 50 +++++++++++++++++++++++-----------------------
 mm/memcontrol.c            |  8 ++++----
 2 files changed, 29 insertions(+), 29 deletions(-)

diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index 46bf724cae7a..5a5ca6814782 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -410,7 +410,7 @@ static inline struct mem_cgroup *obj_cgroup_memcg(struct obj_cgroup *objcg)
  * or NULL. This function assumes that the folio is known to have a
  * proper object cgroup pointer.
  */
-static inline struct obj_cgroup *folio_objcg(struct folio *folio)
+static inline struct obj_cgroup *folio_objcg(const struct folio *folio)
 {
 	unsigned long memcg_data = folio->memcg_data;
 
@@ -448,7 +448,7 @@ static inline struct obj_cgroup *folio_objcg(struct folio *folio)
  * Note: The caller should hold an rcu read lock or cgroup_mutex to protect
  * memcg associated with a folio from being released.
  */
-static inline struct mem_cgroup *folio_memcg(struct folio *folio)
+static inline struct mem_cgroup *folio_memcg(const struct folio *folio)
 {
 	struct obj_cgroup *objcg = folio_objcg(folio);
 
@@ -461,7 +461,7 @@ static inline struct mem_cgroup *folio_memcg(struct folio *folio)
  *
  * Returns true if folio is charged to a memory cgroup, otherwise returns false.
  */
-static inline bool folio_memcg_charged(struct folio *folio)
+static inline bool folio_memcg_charged(const struct folio *folio)
 {
 	return folio->memcg_data != 0;
 }
@@ -481,7 +481,7 @@ static inline bool folio_memcg_charged(struct folio *folio)
  * A caller should hold an rcu read lock to protect memcg associated with a
  * page from being released.
  */
-static inline struct mem_cgroup *folio_memcg_check(struct folio *folio)
+static inline struct mem_cgroup *folio_memcg_check(const struct folio *folio)
 {
 	/*
 	 * Because folio->memcg_data might be changed asynchronously
@@ -498,11 +498,11 @@ static inline struct mem_cgroup *folio_memcg_check(struct folio *folio)
 	return obj_cgroup_memcg(objcg);
 }
 
-static inline struct mem_cgroup *page_memcg_check(struct page *page)
+static inline struct mem_cgroup *page_memcg_check(const struct page *page)
 {
 	if (PageTail(page))
 		return NULL;
-	return folio_memcg_check((struct folio *)page);
+	return folio_memcg_check((const struct folio *)page);
 }
 
 static inline struct mem_cgroup *get_mem_cgroup_from_objcg(struct obj_cgroup *objcg)
@@ -527,14 +527,14 @@ static inline struct mem_cgroup *get_mem_cgroup_from_objcg(struct obj_cgroup *ob
  * that the folio has an associated memory cgroup. It's not safe to call
  * this function against some types of folios, e.g. slab folios.
  */
-static inline bool folio_memcg_kmem(struct folio *folio)
+static inline bool folio_memcg_kmem(const struct folio *folio)
 {
 	VM_BUG_ON_PGFLAGS(PageTail(&folio->page), &folio->page);
 	VM_BUG_ON_FOLIO(folio->memcg_data & MEMCG_DATA_OBJEXTS, folio);
 	return folio->memcg_data & MEMCG_DATA_KMEM;
 }
 
-static inline bool PageMemcgKmem(struct page *page)
+static inline bool PageMemcgKmem(const struct page *page)
 {
 	return folio_memcg_kmem(page_folio(page));
 }
@@ -777,7 +777,7 @@ struct mem_cgroup *get_mem_cgroup_from_mm(struct mm_struct *mm);
 
 struct mem_cgroup *get_mem_cgroup_from_current(void);
 
-struct mem_cgroup *get_mem_cgroup_from_folio(struct folio *folio);
+struct mem_cgroup *get_mem_cgroup_from_folio(const struct folio *folio);
 
 struct lruvec *folio_lruvec_lock(struct folio *folio);
 struct lruvec *folio_lruvec_lock_irq(struct folio *folio);
@@ -905,8 +905,8 @@ static inline bool mm_match_cgroup(struct mm_struct *mm,
 	return match;
 }
 
-struct cgroup_subsys_state *get_mem_cgroup_css_from_folio(struct folio *folio);
-ino_t page_cgroup_ino(struct page *page);
+struct cgroup_subsys_state *get_mem_cgroup_css_from_folio(const struct folio *folio);
+ino_t page_cgroup_ino(const struct page *page);
 
 static inline bool mem_cgroup_online(struct mem_cgroup *memcg)
 {
@@ -956,7 +956,7 @@ void mem_cgroup_print_oom_group(struct mem_cgroup *memcg);
 void mod_memcg_state(struct mem_cgroup *memcg,
 		     enum memcg_stat_item idx, int val);
 
-static inline void mod_memcg_page_state(struct page *page,
+static inline void mod_memcg_page_state(const struct page *page,
 					enum memcg_stat_item idx, int val)
 {
 	struct mem_cgroup *memcg;
@@ -990,7 +990,7 @@ void mod_lruvec_kmem_state(void *p, enum node_stat_item idx, int val);
 void count_memcg_events(struct mem_cgroup *memcg, enum vm_event_item idx,
 			unsigned long count);
 
-static inline void count_memcg_folio_events(struct folio *folio,
+static inline void count_memcg_folio_events(const struct folio *folio,
 		enum vm_event_item idx, unsigned long nr)
 {
 	struct mem_cgroup *memcg;
@@ -1084,22 +1084,22 @@ static inline struct mem_cgroup *obj_cgroup_memcg(struct obj_cgroup *objcg)
 
 #define root_mem_cgroup		(NULL)
 
-static inline struct mem_cgroup *folio_memcg(struct folio *folio)
+static inline struct mem_cgroup *folio_memcg(const struct folio *folio)
 {
 	return NULL;
 }
 
-static inline bool folio_memcg_charged(struct folio *folio)
+static inline bool folio_memcg_charged(const struct folio *folio)
 {
 	return false;
 }
 
-static inline struct mem_cgroup *folio_memcg_check(struct folio *folio)
+static inline struct mem_cgroup *folio_memcg_check(const struct folio *folio)
 {
 	return NULL;
 }
 
-static inline struct mem_cgroup *page_memcg_check(struct page *page)
+static inline struct mem_cgroup *page_memcg_check(const struct page *page)
 {
 	return NULL;
 }
@@ -1109,12 +1109,12 @@ static inline struct mem_cgroup *get_mem_cgroup_from_objcg(struct obj_cgroup *ob
 	return NULL;
 }
 
-static inline bool folio_memcg_kmem(struct folio *folio)
+static inline bool folio_memcg_kmem(const struct folio *folio)
 {
 	return false;
 }
 
-static inline bool PageMemcgKmem(struct page *page)
+static inline bool PageMemcgKmem(const struct page *page)
 {
 	return false;
 }
@@ -1248,7 +1248,7 @@ static inline struct mem_cgroup *get_mem_cgroup_from_current(void)
 	return NULL;
 }
 
-static inline struct mem_cgroup *get_mem_cgroup_from_folio(struct folio *folio)
+static inline struct mem_cgroup *get_mem_cgroup_from_folio(const struct folio *folio)
 {
 	return NULL;
 }
@@ -1406,7 +1406,7 @@ static inline void mod_memcg_state(struct mem_cgroup *memcg,
 {
 }
 
-static inline void mod_memcg_page_state(struct page *page,
+static inline void mod_memcg_page_state(const struct page *page,
 					enum memcg_stat_item idx, int val)
 {
 }
@@ -1471,7 +1471,7 @@ static inline void count_memcg_events(struct mem_cgroup *memcg,
 {
 }
 
-static inline void count_memcg_folio_events(struct folio *folio,
+static inline void count_memcg_folio_events(const struct folio *folio,
 		enum vm_event_item idx, unsigned long nr)
 {
 }
@@ -1767,7 +1767,7 @@ void __memcg_kmem_uncharge_page(struct page *page, int order);
  * needs to be used outside of the local scope.
  */
 struct obj_cgroup *current_obj_cgroup(void);
-struct obj_cgroup *get_obj_cgroup_from_folio(struct folio *folio);
+struct obj_cgroup *get_obj_cgroup_from_folio(const struct folio *folio);
 
 static inline struct obj_cgroup *get_obj_cgroup_from_current(void)
 {
@@ -1870,7 +1870,7 @@ static inline void __memcg_kmem_uncharge_page(struct page *page, int order)
 {
 }
 
-static inline struct obj_cgroup *get_obj_cgroup_from_folio(struct folio *folio)
+static inline struct obj_cgroup *get_obj_cgroup_from_folio(const struct folio *folio)
 {
 	return NULL;
 }
@@ -1901,7 +1901,7 @@ static inline void count_objcg_events(struct obj_cgroup *objcg,
 {
 }
 
-static inline ino_t page_cgroup_ino(struct page *page)
+static inline ino_t page_cgroup_ino(const struct page *page)
 {
 	return 0;
 }
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 1460cba53588..54937efa075d 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -332,7 +332,7 @@ EXPORT_SYMBOL(memcg_bpf_enabled_key);
  * If memcg is bound to a traditional hierarchy, the css of root_mem_cgroup
  * is returned.
  */
-struct cgroup_subsys_state *get_mem_cgroup_css_from_folio(struct folio *folio)
+struct cgroup_subsys_state *get_mem_cgroup_css_from_folio(const struct folio *folio)
 {
 	struct mem_cgroup *memcg;
 
@@ -357,7 +357,7 @@ struct cgroup_subsys_state *get_mem_cgroup_css_from_folio(struct folio *folio)
  * after page_cgroup_ino() returns, so it only should be used by callers that
  * do not care (such as procfs interfaces).
  */
-ino_t page_cgroup_ino(struct page *page)
+ino_t page_cgroup_ino(const struct page *page)
 {
 	struct mem_cgroup *memcg;
 	unsigned long ino = 0;
@@ -1242,7 +1242,7 @@ struct mem_cgroup *get_mem_cgroup_from_current(void)
  *
  * See folio_memcg() for folio->objcg/memcg binding rules.
  */
-struct mem_cgroup *get_mem_cgroup_from_folio(struct folio *folio)
+struct mem_cgroup *get_mem_cgroup_from_folio(const struct folio *folio)
 {
 	struct mem_cgroup *memcg;
 
@@ -3140,7 +3140,7 @@ __always_inline struct obj_cgroup *current_obj_cgroup(void)
 	return rcu_dereference_check(root_mem_cgroup->nodeinfo[nid]->objcg, 1);
 }
 
-struct obj_cgroup *get_obj_cgroup_from_folio(struct folio *folio)
+struct obj_cgroup *get_obj_cgroup_from_folio(const struct folio *folio)
 {
 	struct obj_cgroup *objcg;
 

-- 
2.39.5

Re: [PATCH v3 01/11] mm: memcontrol: take a const folio in folio_memcg() and friends

[Muchun Song]

> On Sep 16, 2026, at 07:20, Tal Zussman <tz2294@columbia.edu> wrote:
> 
> The folio_memcg() family only reads from the folio, and everything it
> calls already takes a const folio. Constify it, along with the page
> wrappers built on top of it and get_obj_cgroup_from_folio(), which only
> reads the folio's objcg.
> 
> Signed-off-by: Tal Zussman <tz2294@columbia.edu>

Reviewed-by: Muchun Song <muchun.song@linux.dev>

Thanks.

[PATCH v3 02/11] mm: memcontrol: constify obj_cgroup_memcg() and friends

[Tal Zussman]

obj_cgroup_memcg() and get_mem_cgroup_from_objcg() only read from the
objcg. Constify them.

Signed-off-by: Tal Zussman <tz2294@columbia.edu>
---
 include/linux/memcontrol.h | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index 5a5ca6814782..65cb45f4be32 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -396,7 +396,7 @@ enum objext_flags {
  *
  * The caller must ensure that the returned memcg won't be released.
  */
-static inline struct mem_cgroup *obj_cgroup_memcg(struct obj_cgroup *objcg)
+static inline struct mem_cgroup *obj_cgroup_memcg(const struct obj_cgroup *objcg)
 {
 	lockdep_assert_once(rcu_read_lock_held() || lockdep_is_held(&cgroup_mutex));
 	return objcg ? READ_ONCE(objcg->memcg) : NULL;
@@ -505,7 +505,7 @@ static inline struct mem_cgroup *page_memcg_check(const struct page *page)
 	return folio_memcg_check((const struct folio *)page);
 }
 
-static inline struct mem_cgroup *get_mem_cgroup_from_objcg(struct obj_cgroup *objcg)
+static inline struct mem_cgroup *get_mem_cgroup_from_objcg(const struct obj_cgroup *objcg)
 {
 	struct mem_cgroup *memcg;
 
@@ -1075,7 +1075,7 @@ void mem_cgroup_flush_workqueue(void);
 extern int mem_cgroup_init(void);
 #else /* CONFIG_MEMCG */
 
-static inline struct mem_cgroup *obj_cgroup_memcg(struct obj_cgroup *objcg)
+static inline struct mem_cgroup *obj_cgroup_memcg(const struct obj_cgroup *objcg)
 {
 	return NULL;
 }
@@ -1104,7 +1104,7 @@ static inline struct mem_cgroup *page_memcg_check(const struct page *page)
 	return NULL;
 }
 
-static inline struct mem_cgroup *get_mem_cgroup_from_objcg(struct obj_cgroup *objcg)
+static inline struct mem_cgroup *get_mem_cgroup_from_objcg(const struct obj_cgroup *objcg)
 {
 	return NULL;
 }

-- 
2.39.5

Re: [PATCH v3 02/11] mm: memcontrol: constify obj_cgroup_memcg() and friends

[Muchun Song]

> On Sep 16, 2026, at 07:20, Tal Zussman <tz2294@columbia.edu> wrote:
> 
> obj_cgroup_memcg() and get_mem_cgroup_from_objcg() only read from the
> objcg. Constify them.
> 
> Signed-off-by: Tal Zussman <tz2294@columbia.edu>

Reviewed-by: Muchun Song <muchun.song@linux.dev>

Thanks.

[PATCH v3 03/11] mm: memcontrol: constify the lruvec helpers

[Tal Zussman]

The lruvec lookup helpers only read from the memcg, folio, or lruvec
they are given. Constify them, along with lruvec_pgdat() and the folio
argument of the folio_lruvec_relock_irq() helpers. mem_cgroup_lruvec()
may update lruvec->pgdat for a newly onlined node, but that lives in
the per-node structure, not the memcg. Similarly, lruvec_pgdat() still
returns a non-const pgdat and does not use container_of_const().

Use container_of_const() in lruvec_memcg() and
mem_cgroup_get_zone_lru_size(), so the const isn't silently cast away.

Signed-off-by: Tal Zussman <tz2294@columbia.edu>
---
 include/linux/memcontrol.h | 46 +++++++++++++++++++++++-----------------------
 include/linux/mmzone.h     |  2 +-
 mm/memcontrol.c            |  6 +++---
 3 files changed, 27 insertions(+), 27 deletions(-)

diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index 65cb45f4be32..436934030d79 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -722,7 +722,7 @@ void mem_cgroup_migrate(struct folio *old, struct folio *new);
  * @pgdat combination. This can be the node lruvec, if the memory
  * controller is disabled.
  */
-static inline struct lruvec *mem_cgroup_lruvec(struct mem_cgroup *memcg,
+static inline struct lruvec *mem_cgroup_lruvec(const struct mem_cgroup *memcg,
 					       struct pglist_data *pgdat)
 {
 	struct mem_cgroup_per_node *mz;
@@ -763,7 +763,7 @@ static inline struct lruvec *mem_cgroup_lruvec(struct mem_cgroup *memcg,
  * their binding is stable if the returned lruvec matches the one the caller has
  * locked. Useful for lock batching.
  */
-static inline struct lruvec *folio_lruvec(struct folio *folio)
+static inline struct lruvec *folio_lruvec(const struct folio *folio)
 {
 	struct mem_cgroup *memcg = folio_memcg(folio);
 
@@ -779,9 +779,9 @@ struct mem_cgroup *get_mem_cgroup_from_current(void);
 
 struct mem_cgroup *get_mem_cgroup_from_folio(const struct folio *folio);
 
-struct lruvec *folio_lruvec_lock(struct folio *folio);
-struct lruvec *folio_lruvec_lock_irq(struct folio *folio);
-struct lruvec *folio_lruvec_lock_irqsave(struct folio *folio,
+struct lruvec *folio_lruvec_lock(const struct folio *folio);
+struct lruvec *folio_lruvec_lock_irq(const struct folio *folio);
+struct lruvec *folio_lruvec_lock_irqsave(const struct folio *folio,
 						unsigned long *flags);
 
 static inline
@@ -861,14 +861,14 @@ static inline struct mem_cgroup *mem_cgroup_from_seq(struct seq_file *m)
 	return mem_cgroup_from_css(seq_css(m));
 }
 
-static inline struct mem_cgroup *lruvec_memcg(struct lruvec *lruvec)
+static inline struct mem_cgroup *lruvec_memcg(const struct lruvec *lruvec)
 {
-	struct mem_cgroup_per_node *mz;
+	const struct mem_cgroup_per_node *mz;
 
 	if (mem_cgroup_disabled())
 		return NULL;
 
-	mz = container_of(lruvec, struct mem_cgroup_per_node, lruvec);
+	mz = container_of_const(lruvec, struct mem_cgroup_per_node, lruvec);
 	return mz->memcg;
 }
 
@@ -919,13 +919,13 @@ void mem_cgroup_update_lru_size(struct lruvec *lruvec, enum lru_list lru,
 		int zid, long nr_pages);
 
 static inline
-unsigned long mem_cgroup_get_zone_lru_size(struct lruvec *lruvec,
+unsigned long mem_cgroup_get_zone_lru_size(const struct lruvec *lruvec,
 		enum lru_list lru, int zone_idx)
 {
 	long val;
-	struct mem_cgroup_per_node *mz;
+	const struct mem_cgroup_per_node *mz;
 
-	mz = container_of(lruvec, struct mem_cgroup_per_node, lruvec);
+	mz = container_of_const(lruvec, struct mem_cgroup_per_node, lruvec);
 	val = READ_ONCE(mz->lru_zone_size[zone_idx][lru]);
 	if (WARN_ON_ONCE(val < 0))
 		return 0;
@@ -1215,13 +1215,13 @@ static inline void mem_cgroup_migrate(struct folio *old, struct folio *new)
 {
 }
 
-static inline struct lruvec *mem_cgroup_lruvec(struct mem_cgroup *memcg,
+static inline struct lruvec *mem_cgroup_lruvec(const struct mem_cgroup *memcg,
 					       struct pglist_data *pgdat)
 {
 	return &pgdat->__lruvec;
 }
 
-static inline struct lruvec *folio_lruvec(struct folio *folio)
+static inline struct lruvec *folio_lruvec(const struct folio *folio)
 {
 	struct pglist_data *pgdat = folio_pgdat(folio);
 	return &pgdat->__lruvec;
@@ -1281,7 +1281,7 @@ static inline void mem_cgroup_put(struct mem_cgroup *memcg)
 {
 }
 
-static inline struct lruvec *folio_lruvec_lock(struct folio *folio)
+static inline struct lruvec *folio_lruvec_lock(const struct folio *folio)
 {
 	struct pglist_data *pgdat = folio_pgdat(folio);
 
@@ -1290,7 +1290,7 @@ static inline struct lruvec *folio_lruvec_lock(struct folio *folio)
 	return &pgdat->__lruvec;
 }
 
-static inline struct lruvec *folio_lruvec_lock_irq(struct folio *folio)
+static inline struct lruvec *folio_lruvec_lock_irq(const struct folio *folio)
 {
 	struct pglist_data *pgdat = folio_pgdat(folio);
 
@@ -1299,7 +1299,7 @@ static inline struct lruvec *folio_lruvec_lock_irq(struct folio *folio)
 	return &pgdat->__lruvec;
 }
 
-static inline struct lruvec *folio_lruvec_lock_irqsave(struct folio *folio,
+static inline struct lruvec *folio_lruvec_lock_irqsave(const struct folio *folio,
 		unsigned long *flagsp)
 {
 	struct pglist_data *pgdat = folio_pgdat(folio);
@@ -1354,7 +1354,7 @@ static inline struct mem_cgroup *mem_cgroup_from_seq(struct seq_file *m)
 	return NULL;
 }
 
-static inline struct mem_cgroup *lruvec_memcg(struct lruvec *lruvec)
+static inline struct mem_cgroup *lruvec_memcg(const struct lruvec *lruvec)
 {
 	return NULL;
 }
@@ -1365,7 +1365,7 @@ static inline bool mem_cgroup_online(struct mem_cgroup *memcg)
 }
 
 static inline
-unsigned long mem_cgroup_get_zone_lru_size(struct lruvec *lruvec,
+unsigned long mem_cgroup_get_zone_lru_size(const struct lruvec *lruvec,
 		enum lru_list lru, int zone_idx)
 {
 	return 0;
@@ -1505,7 +1505,7 @@ static inline void mem_cgroup_flush_workqueue(void) { }
 static inline int mem_cgroup_init(void) { return 0; }
 #endif /* CONFIG_MEMCG */
 
-static inline struct lruvec *parent_lruvec(struct lruvec *lruvec)
+static inline struct lruvec *parent_lruvec(const struct lruvec *lruvec)
 {
 	struct mem_cgroup *memcg;
 
@@ -1568,15 +1568,15 @@ static inline void lruvec_unlock_irqrestore(struct lruvec *lruvec, unsigned long
 }
 
 /* Test requires a stable folio->memcg binding, see folio_memcg() */
-static inline bool folio_matches_lruvec(struct folio *folio,
-		struct lruvec *lruvec)
+static inline bool folio_matches_lruvec(const struct folio *folio,
+		const struct lruvec *lruvec)
 {
 	return lruvec_pgdat(lruvec) == folio_pgdat(folio) &&
 	       lruvec_memcg(lruvec) == folio_memcg(folio);
 }
 
 /* Don't lock again iff page's lruvec locked */
-static inline struct lruvec *folio_lruvec_relock_irq(struct folio *folio,
+static inline struct lruvec *folio_lruvec_relock_irq(const struct folio *folio,
 		struct lruvec *locked_lruvec)
 {
 	if (locked_lruvec) {
@@ -1590,7 +1590,7 @@ static inline struct lruvec *folio_lruvec_relock_irq(struct folio *folio,
 }
 
 /* Don't lock again iff folio's lruvec locked */
-static inline void folio_lruvec_relock_irqsave(struct folio *folio,
+static inline void folio_lruvec_relock_irqsave(const struct folio *folio,
 		struct lruvec **lruvecp, unsigned long *flags)
 {
 	if (*lruvecp) {
diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h
index f8a2d823f7de..9c9c3119277d 100644
--- a/include/linux/mmzone.h
+++ b/include/linux/mmzone.h
@@ -1661,7 +1661,7 @@ extern void init_currently_empty_zone(struct zone *zone, unsigned long start_pfn
 
 extern void lruvec_init(struct lruvec *lruvec);
 
-static inline struct pglist_data *lruvec_pgdat(struct lruvec *lruvec)
+static inline struct pglist_data *lruvec_pgdat(const struct lruvec *lruvec)
 {
 #ifdef CONFIG_MEMCG
 	return lruvec->pgdat;
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 54937efa075d..1f660af3bc3a 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -1463,7 +1463,7 @@ void mem_cgroup_scan_tasks(struct mem_cgroup *memcg,
  *
  * Return: The lruvec this folio is on with its lock held and rcu read lock held.
  */
-struct lruvec *folio_lruvec_lock(struct folio *folio)
+struct lruvec *folio_lruvec_lock(const struct folio *folio)
 {
 	struct lruvec *lruvec;
 
@@ -1491,7 +1491,7 @@ struct lruvec *folio_lruvec_lock(struct folio *folio)
  * Return: The lruvec this folio is on with its lock held and interrupts
  * disabled and rcu read lock held.
  */
-struct lruvec *folio_lruvec_lock_irq(struct folio *folio)
+struct lruvec *folio_lruvec_lock_irq(const struct folio *folio)
 {
 	struct lruvec *lruvec;
 
@@ -1520,7 +1520,7 @@ struct lruvec *folio_lruvec_lock_irq(struct folio *folio)
  * Return: The lruvec this folio is on with its lock held and interrupts
  * disabled and rcu read lock held.
  */
-struct lruvec *folio_lruvec_lock_irqsave(struct folio *folio,
+struct lruvec *folio_lruvec_lock_irqsave(const struct folio *folio,
 		unsigned long *flags)
 {
 	struct lruvec *lruvec;

-- 
2.39.5

Re: [PATCH v3 03/11] mm: memcontrol: constify the lruvec helpers

[Muchun Song]

> On Sep 16, 2026, at 07:20, Tal Zussman <tz2294@columbia.edu> wrote:
> 
> The lruvec lookup helpers only read from the memcg, folio, or lruvec
> they are given. Constify them, along with lruvec_pgdat() and the folio
> argument of the folio_lruvec_relock_irq() helpers. mem_cgroup_lruvec()
> may update lruvec->pgdat for a newly onlined node, but that lives in
> the per-node structure, not the memcg. Similarly, lruvec_pgdat() still
> returns a non-const pgdat and does not use container_of_const().
> 
> Use container_of_const() in lruvec_memcg() and
> mem_cgroup_get_zone_lru_size(), so the const isn't silently cast away.
> 
> Signed-off-by: Tal Zussman <tz2294@columbia.edu>

Reviewed-by: Muchun Song <muchun.song@linux.dev>

Thanks

[PATCH v3 04/11] mm/page_io: take a const folio in bio_associate_blkg_from_folio()

[Tal Zussman]

bio_associate_blkg_from_folio() and its helpers only read from the
folio. Constify their folio arguments.

Signed-off-by: Tal Zussman <tz2294@columbia.edu>
---
 mm/page_io.c | 14 +++++++++-----
 1 file changed, 9 insertions(+), 5 deletions(-)

diff --git a/mm/page_io.c b/mm/page_io.c
index 1da4ff484f09..5f7756e370f7 100644
--- a/mm/page_io.c
+++ b/mm/page_io.c
@@ -256,12 +256,13 @@ int swap_writeout(struct swap_io_ctx *ctx, struct folio *folio)
 }
 
 #if defined(CONFIG_MEMCG) && defined(CONFIG_BLK_CGROUP)
-static struct cgroup_subsys_state *folio_memcg_blkg_css(struct folio *folio)
+static struct cgroup_subsys_state *folio_memcg_blkg_css(const struct folio *folio)
 {
 	return cgroup_e_css(folio_memcg(folio)->css.cgroup, &io_cgrp_subsys);
 }
 
-static bool folio_blkg_can_merge(struct folio *folio, struct folio *prev_folio)
+static bool folio_blkg_can_merge(const struct folio *folio,
+				 const struct folio *prev_folio)
 {
 	bool can_merge = true;
 
@@ -277,7 +278,8 @@ static bool folio_blkg_can_merge(struct folio *folio, struct folio *prev_folio)
 	return can_merge;
 }
 
-static void bio_associate_blkg_from_folio(struct bio *bio, struct folio *folio)
+static void bio_associate_blkg_from_folio(struct bio *bio,
+					  const struct folio *folio)
 {
 	struct cgroup_subsys_state *css;
 
@@ -294,11 +296,13 @@ static void bio_associate_blkg_from_folio(struct bio *bio, struct folio *folio)
 		css_put(css);
 }
 #else
-static bool folio_blkg_can_merge(struct folio *folio, struct folio *prev_folio)
+static bool folio_blkg_can_merge(const struct folio *folio,
+				 const struct folio *prev_folio)
 {
 	return true;
 }
-static void bio_associate_blkg_from_folio(struct bio *bio, struct folio *folio)
+static void bio_associate_blkg_from_folio(struct bio *bio,
+					  const struct folio *folio)
 {
 }
 #endif /* CONFIG_MEMCG && CONFIG_BLK_CGROUP */

-- 
2.39.5

Re: [PATCH v3 04/11] mm/page_io: take a const folio in bio_associate_blkg_from_folio()

[Muchun Song]

> On Sep 16, 2026, at 07:20, Tal Zussman <tz2294@columbia.edu> wrote:
> 
> bio_associate_blkg_from_folio() and its helpers only read from the
> folio. Constify their folio arguments.
> 
> Signed-off-by: Tal Zussman <tz2294@columbia.edu>

Reviewed-by: Muchun Song <muchun.song@linux.dev>

Thanks

[PATCH v3 05/11] mm: memcontrol: constify the mem_cgroup accessors

[Tal Zussman]

mem_cgroup_id(), parent_mem_cgroup(), mem_cgroup_is_root(),
mem_cgroup_is_descendant(), memcg_kmem_id(), and the other memcg
accessors only read from the memcg. Constify them, along with
mem_cgroup_shrink_is_root()'s shrink_control.

mem_cgroup_print_oom_context()'s task stays non-const, as
task_cgroup() takes a non-const task.

Signed-off-by: Tal Zussman <tz2294@columbia.edu>
---
 include/linux/memcontrol.h | 41 +++++++++++++++++++++--------------------
 mm/memcontrol.c            |  3 ++-
 2 files changed, 23 insertions(+), 21 deletions(-)

diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index 436934030d79..01f74413fc79 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -539,7 +539,7 @@ static inline bool PageMemcgKmem(const struct page *page)
 	return folio_memcg_kmem(page_folio(page));
 }
 
-static inline bool mem_cgroup_is_root(struct mem_cgroup *memcg)
+static inline bool mem_cgroup_is_root(const struct mem_cgroup *memcg)
 {
 	return (memcg == root_mem_cgroup);
 }
@@ -555,7 +555,7 @@ static inline bool mem_cgroup_is_root(struct mem_cgroup *memcg)
  * and do not honour sc->memcg can use this to early-return 0 in per-memcg
  * contexts.
  */
-static inline bool mem_cgroup_shrink_is_root(struct shrink_control *sc)
+static inline bool mem_cgroup_shrink_is_root(const struct shrink_control *sc)
 {
 	return !sc->memcg || mem_cgroup_is_root(sc->memcg);
 }
@@ -840,7 +840,7 @@ void mem_cgroup_iter_break(struct mem_cgroup *, struct mem_cgroup *);
 void mem_cgroup_scan_tasks(struct mem_cgroup *memcg,
 			   int (*)(struct task_struct *, void *), void *arg);
 
-static inline unsigned short mem_cgroup_private_id(struct mem_cgroup *memcg)
+static inline unsigned short mem_cgroup_private_id(const struct mem_cgroup *memcg)
 {
 	if (mem_cgroup_disabled())
 		return 0;
@@ -849,7 +849,7 @@ static inline unsigned short mem_cgroup_private_id(struct mem_cgroup *memcg)
 }
 struct mem_cgroup *mem_cgroup_from_private_id(unsigned short id);
 
-static inline u64 mem_cgroup_id(struct mem_cgroup *memcg)
+static inline u64 mem_cgroup_id(const struct mem_cgroup *memcg)
 {
 	return memcg ? cgroup_id(memcg->css.cgroup) : 0;
 }
@@ -878,13 +878,13 @@ static inline struct mem_cgroup *lruvec_memcg(const struct lruvec *lruvec)
  *
  * Returns the parent memcg, or NULL if this is the root.
  */
-static inline struct mem_cgroup *parent_mem_cgroup(struct mem_cgroup *memcg)
+static inline struct mem_cgroup *parent_mem_cgroup(const struct mem_cgroup *memcg)
 {
 	return mem_cgroup_from_css(memcg->css.parent);
 }
 
-static inline bool mem_cgroup_is_descendant(struct mem_cgroup *memcg,
-			      struct mem_cgroup *root)
+static inline bool mem_cgroup_is_descendant(const struct mem_cgroup *memcg,
+			      const struct mem_cgroup *root)
 {
 	if (root == memcg)
 		return true;
@@ -892,7 +892,7 @@ static inline bool mem_cgroup_is_descendant(struct mem_cgroup *memcg,
 }
 
 static inline bool mm_match_cgroup(struct mm_struct *mm,
-				   struct mem_cgroup *memcg)
+				   const struct mem_cgroup *memcg)
 {
 	struct mem_cgroup *task_memcg;
 	bool match = false;
@@ -943,7 +943,7 @@ static inline void mem_cgroup_handle_over_high(gfp_t gfp_mask)
 
 unsigned long mem_cgroup_get_max(struct mem_cgroup *memcg);
 
-void mem_cgroup_print_oom_context(struct mem_cgroup *memcg,
+void mem_cgroup_print_oom_context(const struct mem_cgroup *memcg,
 				struct task_struct *p);
 
 void mem_cgroup_print_oom_meminfo(struct mem_cgroup *memcg);
@@ -1119,12 +1119,12 @@ static inline bool PageMemcgKmem(const struct page *page)
 	return false;
 }
 
-static inline bool mem_cgroup_is_root(struct mem_cgroup *memcg)
+static inline bool mem_cgroup_is_root(const struct mem_cgroup *memcg)
 {
 	return true;
 }
 
-static inline bool mem_cgroup_shrink_is_root(struct shrink_control *sc)
+static inline bool mem_cgroup_shrink_is_root(const struct shrink_control *sc)
 {
 	return true;
 }
@@ -1227,13 +1227,13 @@ static inline struct lruvec *folio_lruvec(const struct folio *folio)
 	return &pgdat->__lruvec;
 }
 
-static inline struct mem_cgroup *parent_mem_cgroup(struct mem_cgroup *memcg)
+static inline struct mem_cgroup *parent_mem_cgroup(const struct mem_cgroup *memcg)
 {
 	return NULL;
 }
 
 static inline bool mm_match_cgroup(struct mm_struct *mm,
-		struct mem_cgroup *memcg)
+		const struct mem_cgroup *memcg)
 {
 	return true;
 }
@@ -1327,7 +1327,7 @@ static inline void mem_cgroup_scan_tasks(struct mem_cgroup *memcg,
 {
 }
 
-static inline unsigned short mem_cgroup_private_id(struct mem_cgroup *memcg)
+static inline unsigned short mem_cgroup_private_id(const struct mem_cgroup *memcg)
 {
 	return 0;
 }
@@ -1339,7 +1339,7 @@ static inline struct mem_cgroup *mem_cgroup_from_private_id(unsigned short id)
 	return NULL;
 }
 
-static inline u64 mem_cgroup_id(struct mem_cgroup *memcg)
+static inline u64 mem_cgroup_id(const struct mem_cgroup *memcg)
 {
 	return 0;
 }
@@ -1377,7 +1377,8 @@ static inline unsigned long mem_cgroup_get_max(struct mem_cgroup *memcg)
 }
 
 static inline void
-mem_cgroup_print_oom_context(struct mem_cgroup *memcg, struct task_struct *p)
+mem_cgroup_print_oom_context(const struct mem_cgroup *memcg,
+			     struct task_struct *p)
 {
 }
 
@@ -1813,7 +1814,7 @@ static inline void memcg_kmem_uncharge_page(struct page *page, int order)
  * A helper for accessing memcg's kmem_id, used for getting
  * corresponding LRU lists.
  */
-static inline int memcg_kmem_id(struct mem_cgroup *memcg)
+static inline int memcg_kmem_id(const struct mem_cgroup *memcg)
 {
 	return memcg ? memcg->kmemcg_id : -1;
 }
@@ -1885,7 +1886,7 @@ static inline bool memcg_kmem_online(void)
 	return false;
 }
 
-static inline int memcg_kmem_id(struct mem_cgroup *memcg)
+static inline int memcg_kmem_id(const struct mem_cgroup *memcg)
 {
 	return -1;
 }
@@ -1962,7 +1963,7 @@ static inline bool mem_cgroup_zswap_writeback_enabled(struct mem_cgroup *memcg)
 #ifdef CONFIG_MEMCG_V1
 bool mem_cgroup_oom_synchronize(bool wait);
 
-static inline bool task_in_memcg_oom(struct task_struct *p)
+static inline bool task_in_memcg_oom(const struct task_struct *p)
 {
 	return p->memcg_in_oom;
 }
@@ -1980,7 +1981,7 @@ static inline void mem_cgroup_exit_user_fault(void)
 }
 
 #else /* CONFIG_MEMCG_V1 */
-static inline bool task_in_memcg_oom(struct task_struct *p)
+static inline bool task_in_memcg_oom(const struct task_struct *p)
 {
 	return false;
 }
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 1f660af3bc3a..4b7c01daebce 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -1818,7 +1818,8 @@ static void memory_stat_format(struct mem_cgroup *memcg, struct seq_buf *s)
  * NOTE: @memcg and @p's mem_cgroup can be different when hierarchy is
  * enabled
  */
-void mem_cgroup_print_oom_context(struct mem_cgroup *memcg, struct task_struct *p)
+void mem_cgroup_print_oom_context(const struct mem_cgroup *memcg,
+				  struct task_struct *p)
 {
 	rcu_read_lock();
 

-- 
2.39.5

Re: [PATCH v3 05/11] mm: memcontrol: constify the mem_cgroup accessors

[Muchun Song]

> On Sep 16, 2026, at 07:20, Tal Zussman <tz2294@columbia.edu> wrote:
> 
> mem_cgroup_id(), parent_mem_cgroup(), mem_cgroup_is_root(),
> mem_cgroup_is_descendant(), memcg_kmem_id(), and the other memcg
> accessors only read from the memcg. Constify them, along with
> mem_cgroup_shrink_is_root()'s shrink_control.
> 
> mem_cgroup_print_oom_context()'s task stays non-const, as
> task_cgroup() takes a non-const task.
> 
> Signed-off-by: Tal Zussman <tz2294@columbia.edu>

Acked-by: Muchun Song <muchun.song@linux.dev>

Thanks.

[PATCH v3 06/11] mm: page_counter: constify page_counter_read() and page_counter_margin()

[Tal Zussman]

Both only read the counter. Constify them so that users can read
counters from const memcgs.

Signed-off-by: Tal Zussman <tz2294@columbia.edu>
---
 include/linux/page_counter.h | 4 ++--
 mm/page_counter.c            | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/include/linux/page_counter.h b/include/linux/page_counter.h
index 07b7cb12249c..2baf7a2b29b2 100644
--- a/include/linux/page_counter.h
+++ b/include/linux/page_counter.h
@@ -63,12 +63,12 @@ static inline void page_counter_init(struct page_counter *counter,
 	counter->track_failcnt = false;
 }
 
-static inline unsigned long page_counter_read(struct page_counter *counter)
+static inline unsigned long page_counter_read(const struct page_counter *counter)
 {
 	return atomic_long_read(&counter->usage);
 }
 
-long page_counter_margin(struct page_counter *counter);
+long page_counter_margin(const struct page_counter *counter);
 void page_counter_cancel(struct page_counter *counter, unsigned long nr_pages);
 void page_counter_charge(struct page_counter *counter, unsigned long nr_pages);
 bool page_counter_try_charge(struct page_counter *counter,
diff --git a/mm/page_counter.c b/mm/page_counter.c
index 98322803941a..9167ffd1380c 100644
--- a/mm/page_counter.c
+++ b/mm/page_counter.c
@@ -54,7 +54,7 @@ static void propagate_protected_usage(struct page_counter *c,
  * Return: The minimum value of max minus usage across @counter and all of
  * its ancestors. The value may be negative during a concurrent charge.
  */
-long page_counter_margin(struct page_counter *counter)
+long page_counter_margin(const struct page_counter *counter)
 {
 	long margin = PAGE_COUNTER_MAX;
 

-- 
2.39.5

Re: [PATCH v3 06/11] mm: page_counter: constify page_counter_read() and page_counter_margin()

[Muchun Song]

> On Sep 16, 2026, at 07:20, Tal Zussman <tz2294@columbia.edu> wrote:
> 
> Both only read the counter. Constify them so that users can read
> counters from const memcgs.
> 
> Signed-off-by: Tal Zussman <tz2294@columbia.edu>

Reviewed-by: Muchun Song <muchun.song@linux.dev>

[PATCH v3 07/11] mm: memcontrol: constify the reclaim protection helpers

[Tal Zussman]

mem_cgroup_protection(), mem_cgroup_unprotected(),
mem_cgroup_below_low(), and mem_cgroup_below_min() only read the
protection state. Constify them.

Signed-off-by: Tal Zussman <tz2294@columbia.edu>
---
 include/linux/memcontrol.h | 32 ++++++++++++++++----------------
 1 file changed, 16 insertions(+), 16 deletions(-)

diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index 01f74413fc79..22067899eb6c 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -570,8 +570,8 @@ static inline bool mem_cgroup_disabled(void)
 	return !cgroup_subsys_enabled(memory_cgrp_subsys);
 }
 
-static inline void mem_cgroup_protection(struct mem_cgroup *root,
-					 struct mem_cgroup *memcg,
+static inline void mem_cgroup_protection(const struct mem_cgroup *root,
+					 const struct mem_cgroup *memcg,
 					 unsigned long *min,
 					 unsigned long *low,
 					 unsigned long *usage)
@@ -625,8 +625,8 @@ static inline void mem_cgroup_protection(struct mem_cgroup *root,
 void mem_cgroup_calculate_protection(struct mem_cgroup *root,
 				     struct mem_cgroup *memcg);
 
-static inline bool mem_cgroup_unprotected(struct mem_cgroup *target,
-					  struct mem_cgroup *memcg)
+static inline bool mem_cgroup_unprotected(const struct mem_cgroup *target,
+					  const struct mem_cgroup *memcg)
 {
 	/*
 	 * The root memcg doesn't account charges, and doesn't support
@@ -637,8 +637,8 @@ static inline bool mem_cgroup_unprotected(struct mem_cgroup *target,
 		memcg == target;
 }
 
-static inline bool mem_cgroup_below_low(struct mem_cgroup *target,
-					struct mem_cgroup *memcg)
+static inline bool mem_cgroup_below_low(const struct mem_cgroup *target,
+					const struct mem_cgroup *memcg)
 {
 	if (mem_cgroup_unprotected(target, memcg))
 		return false;
@@ -647,8 +647,8 @@ static inline bool mem_cgroup_below_low(struct mem_cgroup *target,
 		page_counter_read(&memcg->memory);
 }
 
-static inline bool mem_cgroup_below_min(struct mem_cgroup *target,
-					struct mem_cgroup *memcg)
+static inline bool mem_cgroup_below_min(const struct mem_cgroup *target,
+					const struct mem_cgroup *memcg)
 {
 	if (mem_cgroup_unprotected(target, memcg))
 		return false;
@@ -1149,8 +1149,8 @@ static inline void memcg_memory_event_mm(struct mm_struct *mm,
 {
 }
 
-static inline void mem_cgroup_protection(struct mem_cgroup *root,
-					 struct mem_cgroup *memcg,
+static inline void mem_cgroup_protection(const struct mem_cgroup *root,
+					 const struct mem_cgroup *memcg,
 					 unsigned long *min,
 					 unsigned long *low,
 					 unsigned long *usage)
@@ -1163,19 +1163,19 @@ static inline void mem_cgroup_calculate_protection(struct mem_cgroup *root,
 {
 }
 
-static inline bool mem_cgroup_unprotected(struct mem_cgroup *target,
-					  struct mem_cgroup *memcg)
+static inline bool mem_cgroup_unprotected(const struct mem_cgroup *target,
+					  const struct mem_cgroup *memcg)
 {
 	return true;
 }
-static inline bool mem_cgroup_below_low(struct mem_cgroup *target,
-					struct mem_cgroup *memcg)
+static inline bool mem_cgroup_below_low(const struct mem_cgroup *target,
+					const struct mem_cgroup *memcg)
 {
 	return false;
 }
 
-static inline bool mem_cgroup_below_min(struct mem_cgroup *target,
-					struct mem_cgroup *memcg)
+static inline bool mem_cgroup_below_min(const struct mem_cgroup *target,
+					const struct mem_cgroup *memcg)
 {
 	return false;
 }

-- 
2.39.5

Re: [PATCH v3 07/11] mm: memcontrol: constify the reclaim protection helpers

[Muchun Song]

> On Sep 16, 2026, at 07:20, Tal Zussman <tz2294@columbia.edu> wrote:
> 
> mem_cgroup_protection(), mem_cgroup_unprotected(),
> mem_cgroup_below_low(), and mem_cgroup_below_min() only read the
> protection state. Constify them.
> 
> Signed-off-by: Tal Zussman <tz2294@columbia.edu>

Acked-by: Muchun Song <muchun.song@linux.dev>

Thanks.

[PATCH v3 08/11] mm: memcontrol: constify the memcg and lruvec stat readers

[Tal Zussman]

The memcg_page_state(), memcg_events(), and lruvec_page_state()
families only read counters. Constify them.

Use container_of_const() in the lruvec_page_state() family while at it,
so the const isn't silently cast away.

Signed-off-by: Tal Zussman <tz2294@columbia.edu>
---
 include/linux/memcontrol.h | 23 ++++++++++++-----------
 mm/memcontrol-v1.h         |  6 +++---
 mm/memcontrol.c            | 30 +++++++++++++++---------------
 3 files changed, 30 insertions(+), 29 deletions(-)

diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index 22067899eb6c..9beb065c0879 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -971,15 +971,16 @@ static inline void mod_memcg_page_state(const struct page *page,
 	rcu_read_unlock();
 }
 
-unsigned long memcg_events(struct mem_cgroup *memcg, int event);
-unsigned long memcg_page_state(struct mem_cgroup *memcg, int idx);
-unsigned long memcg_page_state_output(struct mem_cgroup *memcg, int item);
+unsigned long memcg_events(const struct mem_cgroup *memcg, int event);
+unsigned long memcg_page_state(const struct mem_cgroup *memcg, int idx);
+unsigned long memcg_page_state_output(const struct mem_cgroup *memcg, int item);
 bool memcg_stat_item_valid(int idx);
 bool memcg_vm_event_item_valid(enum vm_event_item idx);
-unsigned long lruvec_page_state(struct lruvec *lruvec, enum node_stat_item idx);
-unsigned long lruvec_page_state_monotonic(struct lruvec *lruvec,
+unsigned long lruvec_page_state(const struct lruvec *lruvec,
+				enum node_stat_item idx);
+unsigned long lruvec_page_state_monotonic(const struct lruvec *lruvec,
 					  enum node_stat_item idx);
-unsigned long lruvec_page_state_local(struct lruvec *lruvec,
+unsigned long lruvec_page_state_local(const struct lruvec *lruvec,
 				      enum node_stat_item idx);
 
 void mem_cgroup_flush_stats(struct mem_cgroup *memcg);
@@ -1412,12 +1413,12 @@ static inline void mod_memcg_page_state(const struct page *page,
 {
 }
 
-static inline unsigned long memcg_page_state(struct mem_cgroup *memcg, int idx)
+static inline unsigned long memcg_page_state(const struct mem_cgroup *memcg, int idx)
 {
 	return 0;
 }
 
-static inline unsigned long memcg_page_state_output(struct mem_cgroup *memcg, int item)
+static inline unsigned long memcg_page_state_output(const struct mem_cgroup *memcg, int item)
 {
 	return 0;
 }
@@ -1432,19 +1433,19 @@ static inline bool memcg_vm_event_item_valid(enum vm_event_item idx)
 	return false;
 }
 
-static inline unsigned long lruvec_page_state(struct lruvec *lruvec,
+static inline unsigned long lruvec_page_state(const struct lruvec *lruvec,
 					      enum node_stat_item idx)
 {
 	return node_page_state(lruvec_pgdat(lruvec), idx);
 }
 
-static inline unsigned long lruvec_page_state_monotonic(struct lruvec *lruvec,
+static inline unsigned long lruvec_page_state_monotonic(const struct lruvec *lruvec,
 							enum node_stat_item idx)
 {
 	return node_page_state_monotonic(lruvec_pgdat(lruvec), idx);
 }
 
-static inline unsigned long lruvec_page_state_local(struct lruvec *lruvec,
+static inline unsigned long lruvec_page_state_local(const struct lruvec *lruvec,
 						    enum node_stat_item idx)
 {
 	return node_page_state(lruvec_pgdat(lruvec), idx);
diff --git a/mm/memcontrol-v1.h b/mm/memcontrol-v1.h
index b9a21f0fd2c3..f6d128da443e 100644
--- a/mm/memcontrol-v1.h
+++ b/mm/memcontrol-v1.h
@@ -34,9 +34,9 @@ static inline bool do_memsw_account(void)
 	return !cgroup_subsys_on_dfl(memory_cgrp_subsys);
 }
 
-unsigned long memcg_events_local(struct mem_cgroup *memcg, int event);
-unsigned long memcg_page_state_local(struct mem_cgroup *memcg, int idx);
-unsigned long memcg_page_state_local_output(struct mem_cgroup *memcg, int item);
+unsigned long memcg_events_local(const struct mem_cgroup *memcg, int event);
+unsigned long memcg_page_state_local(const struct mem_cgroup *memcg, int idx);
+unsigned long memcg_page_state_local_output(const struct mem_cgroup *memcg, int item);
 bool memcg1_alloc_events(struct mem_cgroup *memcg);
 void memcg1_free_events(struct mem_cgroup *memcg);
 
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 4b7c01daebce..b13f5ed74a5a 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -490,9 +490,9 @@ struct lruvec_stats {
 	long state_pending[NR_MEMCG_NODE_STAT_ITEMS];
 };
 
-unsigned long lruvec_page_state(struct lruvec *lruvec, enum node_stat_item idx)
+unsigned long lruvec_page_state(const struct lruvec *lruvec, enum node_stat_item idx)
 {
-	struct mem_cgroup_per_node *pn;
+	const struct mem_cgroup_per_node *pn;
 	long x;
 	int i;
 
@@ -503,7 +503,7 @@ unsigned long lruvec_page_state(struct lruvec *lruvec, enum node_stat_item idx)
 	if (WARN_ONCE(BAD_STAT_IDX(i), "%s: missing stat item %d\n", __func__, idx))
 		return 0;
 
-	pn = container_of(lruvec, struct mem_cgroup_per_node, lruvec);
+	pn = container_of_const(lruvec, struct mem_cgroup_per_node, lruvec);
 	x = READ_ONCE(pn->lruvec_stats->state[i]);
 #ifdef CONFIG_SMP
 	if (x < 0)
@@ -531,10 +531,10 @@ unsigned long lruvec_page_state(struct lruvec *lruvec, enum node_stat_item idx)
  * monotonically-incremented event counters are stored in
  * enum node_stat_item.
  */
-unsigned long lruvec_page_state_monotonic(struct lruvec *lruvec,
+unsigned long lruvec_page_state_monotonic(const struct lruvec *lruvec,
 					  enum node_stat_item idx)
 {
-	struct mem_cgroup_per_node *pn;
+	const struct mem_cgroup_per_node *pn;
 	int i;
 
 	if (mem_cgroup_disabled())
@@ -544,14 +544,14 @@ unsigned long lruvec_page_state_monotonic(struct lruvec *lruvec,
 	if (WARN_ONCE(BAD_STAT_IDX(i), "%s: missing stat item %d\n", __func__, idx))
 		return 0;
 
-	pn = container_of(lruvec, struct mem_cgroup_per_node, lruvec);
+	pn = container_of_const(lruvec, struct mem_cgroup_per_node, lruvec);
 	return (unsigned long)READ_ONCE(pn->lruvec_stats->state[i]);
 }
 
-unsigned long lruvec_page_state_local(struct lruvec *lruvec,
+unsigned long lruvec_page_state_local(const struct lruvec *lruvec,
 				      enum node_stat_item idx)
 {
-	struct mem_cgroup_per_node *pn;
+	const struct mem_cgroup_per_node *pn;
 	long x;
 	int i;
 
@@ -562,7 +562,7 @@ unsigned long lruvec_page_state_local(struct lruvec *lruvec,
 	if (WARN_ONCE(BAD_STAT_IDX(i), "%s: missing stat item %d\n", __func__, idx))
 		return 0;
 
-	pn = container_of(lruvec, struct mem_cgroup_per_node, lruvec);
+	pn = container_of_const(lruvec, struct mem_cgroup_per_node, lruvec);
 	x = READ_ONCE(pn->lruvec_stats->state_local[i]);
 #ifdef CONFIG_SMP
 	if (x < 0)
@@ -827,7 +827,7 @@ static void flush_memcg_stats_dwork(struct work_struct *w)
 	queue_delayed_work(system_dfl_wq, &stats_flush_dwork, FLUSH_TIME);
 }
 
-unsigned long memcg_page_state(struct mem_cgroup *memcg, int idx)
+unsigned long memcg_page_state(const struct mem_cgroup *memcg, int idx)
 {
 	long x;
 	int i = memcg_stats_index(idx);
@@ -950,7 +950,7 @@ void mod_memcg_state(struct mem_cgroup *memcg, enum memcg_stat_item idx,
 
 #ifdef CONFIG_MEMCG_V1
 /* idx can be of type enum memcg_stat_item or node_stat_item. */
-unsigned long memcg_page_state_local(struct mem_cgroup *memcg, int idx)
+unsigned long memcg_page_state_local(const struct mem_cgroup *memcg, int idx)
 {
 	long x;
 	int i = memcg_stats_index(idx);
@@ -1113,7 +1113,7 @@ void count_memcg_events(struct mem_cgroup *memcg, enum vm_event_item idx,
 	put_cpu();
 }
 
-unsigned long memcg_events(struct mem_cgroup *memcg, int event)
+unsigned long memcg_events(const struct mem_cgroup *memcg, int event)
 {
 	int i = memcg_events_index(event);
 
@@ -1132,7 +1132,7 @@ bool memcg_vm_event_item_valid(enum vm_event_item idx)
 }
 
 #ifdef CONFIG_MEMCG_V1
-unsigned long memcg_events_local(struct mem_cgroup *memcg, int event)
+unsigned long memcg_events_local(const struct mem_cgroup *memcg, int event)
 {
 	int i = memcg_events_index(event);
 
@@ -1715,14 +1715,14 @@ static int memcg_page_state_output_unit(int item)
 	}
 }
 
-unsigned long memcg_page_state_output(struct mem_cgroup *memcg, int item)
+unsigned long memcg_page_state_output(const struct mem_cgroup *memcg, int item)
 {
 	return memcg_page_state(memcg, item) *
 		memcg_page_state_output_unit(item);
 }
 
 #ifdef CONFIG_MEMCG_V1
-unsigned long memcg_page_state_local_output(struct mem_cgroup *memcg, int item)
+unsigned long memcg_page_state_local_output(const struct mem_cgroup *memcg, int item)
 {
 	return memcg_page_state_local(memcg, item) *
 		memcg_page_state_output_unit(item);

-- 
2.39.5

Re: [PATCH v3 08/11] mm: memcontrol: constify the memcg and lruvec stat readers

[Muchun Song]

> On Sep 16, 2026, at 07:20, Tal Zussman <tz2294@columbia.edu> wrote:
> 
> The memcg_page_state(), memcg_events(), and lruvec_page_state()
> families only read counters. Constify them.
> 
> Use container_of_const() in the lruvec_page_state() family while at it,
> so the const isn't silently cast away.
> 
> Signed-off-by: Tal Zussman <tz2294@columbia.edu>

Acked-by: Muchun Song <muchun.song@linux.dev>

Thanks.

[PATCH v3 09/11] mm: memcontrol: constify the swap accounting helpers

[Tal Zussman]

mem_cgroup_get_nr_swap_pages(), mem_cgroup_get_folio_swap_margin(), and
mem_cgroup_swap_full() only read swap counters and limits. Constify
them.

Remove externs from function declarations while at it.

Signed-off-by: Tal Zussman <tz2294@columbia.edu>
---
 include/linux/swap.h | 12 ++++++------
 mm/memcontrol.c      |  6 +++---
 2 files changed, 9 insertions(+), 9 deletions(-)

diff --git a/include/linux/swap.h b/include/linux/swap.h
index 43155e122b5c..18f833a448fb 100644
--- a/include/linux/swap.h
+++ b/include/linux/swap.h
@@ -529,9 +529,9 @@ static inline void mem_cgroup_uncharge_swap(unsigned short id, unsigned int nr_p
 	__mem_cgroup_uncharge_swap(id, nr_pages);
 }
 
-long mem_cgroup_get_folio_swap_margin(struct folio *folio);
-extern long mem_cgroup_get_nr_swap_pages(struct mem_cgroup *memcg);
-extern bool mem_cgroup_swap_full(struct folio *folio);
+long mem_cgroup_get_folio_swap_margin(const struct folio *folio);
+long mem_cgroup_get_nr_swap_pages(const struct mem_cgroup *memcg);
+bool mem_cgroup_swap_full(const struct folio *folio);
 #else
 static inline int mem_cgroup_try_charge_swap(struct folio *folio)
 {
@@ -543,17 +543,17 @@ static inline void mem_cgroup_uncharge_swap(unsigned short id,
 {
 }
 
-static inline long mem_cgroup_get_folio_swap_margin(struct folio *folio)
+static inline long mem_cgroup_get_folio_swap_margin(const struct folio *folio)
 {
 	return PAGE_COUNTER_MAX;
 }
 
-static inline long mem_cgroup_get_nr_swap_pages(struct mem_cgroup *memcg)
+static inline long mem_cgroup_get_nr_swap_pages(const struct mem_cgroup *memcg)
 {
 	return get_nr_swap_pages();
 }
 
-static inline bool mem_cgroup_swap_full(struct folio *folio)
+static inline bool mem_cgroup_swap_full(const struct folio *folio)
 {
 	return vm_swap_full();
 }
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index b13f5ed74a5a..ebf325af7acc 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -6003,7 +6003,7 @@ void __mem_cgroup_uncharge_swap(unsigned short id, unsigned int nr_pages)
 	rcu_read_unlock();
 }
 
-long mem_cgroup_get_nr_swap_pages(struct mem_cgroup *memcg)
+long mem_cgroup_get_nr_swap_pages(const struct mem_cgroup *memcg)
 {
 	long nr_swap_pages = get_nr_swap_pages();
 
@@ -6019,7 +6019,7 @@ long mem_cgroup_get_nr_swap_pages(struct mem_cgroup *memcg)
  *
  * Return: Remaining chargeable pages in the folio's memcg hierarchy.
  */
-long mem_cgroup_get_folio_swap_margin(struct folio *folio)
+long mem_cgroup_get_folio_swap_margin(const struct folio *folio)
 {
 	struct mem_cgroup *memcg;
 	long margin;
@@ -6036,7 +6036,7 @@ long mem_cgroup_get_folio_swap_margin(struct folio *folio)
 	return margin;
 }
 
-bool mem_cgroup_swap_full(struct folio *folio)
+bool mem_cgroup_swap_full(const struct folio *folio)
 {
 	struct mem_cgroup *memcg;
 	bool ret = false;

-- 
2.39.5

Re: [PATCH v3 09/11] mm: memcontrol: constify the swap accounting helpers

[Muchun Song]

> On Sep 16, 2026, at 07:20, Tal Zussman <tz2294@columbia.edu> wrote:
> 
> mem_cgroup_get_nr_swap_pages(), mem_cgroup_get_folio_swap_margin(), and
> mem_cgroup_swap_full() only read swap counters and limits. Constify
> them.
> 
> Remove externs from function declarations while at it.
> 
> Signed-off-by: Tal Zussman <tz2294@columbia.edu>

Acked-by: Muchun Song <muchun.song@linux.dev>

Thanks.

[PATCH v3 10/11] mm: memcontrol: constify mem_cgroup_swappiness() and mem_cgroup_get_max()

[Tal Zussman]

Both only read swappiness and the memory limits. Constify them.

Signed-off-by: Tal Zussman <tz2294@columbia.edu>
---
 include/linux/memcontrol.h | 4 ++--
 mm/memcontrol.c            | 2 +-
 mm/swap.h                  | 2 +-
 3 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index 9beb065c0879..f11899085473 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -941,7 +941,7 @@ static inline void mem_cgroup_handle_over_high(gfp_t gfp_mask)
 		__mem_cgroup_handle_over_high(gfp_mask);
 }
 
-unsigned long mem_cgroup_get_max(struct mem_cgroup *memcg);
+unsigned long mem_cgroup_get_max(const struct mem_cgroup *memcg);
 
 void mem_cgroup_print_oom_context(const struct mem_cgroup *memcg,
 				struct task_struct *p);
@@ -1372,7 +1372,7 @@ unsigned long mem_cgroup_get_zone_lru_size(const struct lruvec *lruvec,
 	return 0;
 }
 
-static inline unsigned long mem_cgroup_get_max(struct mem_cgroup *memcg)
+static inline unsigned long mem_cgroup_get_max(const struct mem_cgroup *memcg)
 {
 	return 0;
 }
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index ebf325af7acc..6e3731b97690 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -1884,7 +1884,7 @@ void mem_cgroup_print_oom_meminfo(struct mem_cgroup *memcg)
 /*
  * Return the memory (and swap, if configured) limit for a memcg.
  */
-unsigned long mem_cgroup_get_max(struct mem_cgroup *memcg)
+unsigned long mem_cgroup_get_max(const struct mem_cgroup *memcg)
 {
 	unsigned long max = READ_ONCE(memcg->memory.max);
 
diff --git a/mm/swap.h b/mm/swap.h
index b3b54c28929a..1957960dc60d 100644
--- a/mm/swap.h
+++ b/mm/swap.h
@@ -82,7 +82,7 @@ enum swap_cluster_flags {
 
 extern int vm_swappiness;
 
-static inline int mem_cgroup_swappiness(struct mem_cgroup *memcg)
+static inline int mem_cgroup_swappiness(const struct mem_cgroup *memcg)
 {
 #ifdef CONFIG_MEMCG_V1
 	if (!cgroup_subsys_on_dfl(memory_cgrp_subsys) &&

-- 
2.39.5

Re: [PATCH v3 10/11] mm: memcontrol: constify mem_cgroup_swappiness() and mem_cgroup_get_max()

[Muchun Song]

> On Sep 16, 2026, at 07:20, Tal Zussman <tz2294@columbia.edu> wrote:
> 
> Both only read swappiness and the memory limits. Constify them.
> 
> Signed-off-by: Tal Zussman <tz2294@columbia.edu>

Acked-by: Muchun Song <muchun.song@linux.dev>

Thanks.

[PATCH v3 11/11] mm: memcontrol: constify the zswap and socket pressure helpers

[Tal Zussman]

mem_cgroup_zswap_writeback_enabled() only reads the zswap_writeback
flags, and mem_cgroup_get_socket_pressure() only reads the socket
pressure timestamp. Constify them.

Signed-off-by: Tal Zussman <tz2294@columbia.edu>
---
 include/linux/memcontrol.h | 8 ++++----
 mm/memcontrol.c            | 2 +-
 2 files changed, 5 insertions(+), 5 deletions(-)

diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index f11899085473..64c183be8cbf 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -1684,7 +1684,7 @@ static inline void mem_cgroup_set_socket_pressure(struct mem_cgroup *memcg)
 	write_sequnlock_irqrestore(&memcg->socket_pressure_seqlock, flags);
 }
 
-static inline u64 mem_cgroup_get_socket_pressure(struct mem_cgroup *memcg)
+static inline u64 mem_cgroup_get_socket_pressure(const struct mem_cgroup *memcg)
 {
 	unsigned int seq;
 	u64 val;
@@ -1702,7 +1702,7 @@ static inline void mem_cgroup_set_socket_pressure(struct mem_cgroup *memcg)
 	WRITE_ONCE(memcg->socket_pressure, jiffies + HZ);
 }
 
-static inline u64 mem_cgroup_get_socket_pressure(struct mem_cgroup *memcg)
+static inline u64 mem_cgroup_get_socket_pressure(const struct mem_cgroup *memcg)
 {
 	return READ_ONCE(memcg->socket_pressure);
 }
@@ -1937,7 +1937,7 @@ static inline void mem_cgroup_calculate_protection_path(struct mem_cgroup *root,
 bool obj_cgroup_may_zswap(struct obj_cgroup *objcg);
 void obj_cgroup_charge_zswap(struct obj_cgroup *objcg, size_t size);
 void obj_cgroup_uncharge_zswap(struct obj_cgroup *objcg, size_t size);
-bool mem_cgroup_zswap_writeback_enabled(struct mem_cgroup *memcg);
+bool mem_cgroup_zswap_writeback_enabled(const struct mem_cgroup *memcg);
 #else
 static inline bool obj_cgroup_may_zswap(struct obj_cgroup *objcg)
 {
@@ -1951,7 +1951,7 @@ static inline void obj_cgroup_uncharge_zswap(struct obj_cgroup *objcg,
 					     size_t size)
 {
 }
-static inline bool mem_cgroup_zswap_writeback_enabled(struct mem_cgroup *memcg)
+static inline bool mem_cgroup_zswap_writeback_enabled(const struct mem_cgroup *memcg)
 {
 	/* if zswap is disabled, do not block pages going to the swapping device */
 	return true;
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 6e3731b97690..d0f1b8782009 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -6304,7 +6304,7 @@ void obj_cgroup_uncharge_zswap(struct obj_cgroup *objcg, size_t size)
 	rcu_read_unlock();
 }
 
-bool mem_cgroup_zswap_writeback_enabled(struct mem_cgroup *memcg)
+bool mem_cgroup_zswap_writeback_enabled(const struct mem_cgroup *memcg)
 {
 	/* if zswap is disabled, do not block pages going to the swapping device */
 	if (!zswap_is_enabled())

-- 
2.39.5

Re: [PATCH v3 11/11] mm: memcontrol: constify the zswap and socket pressure helpers

[Muchun Song]

> On Sep 16, 2026, at 07:20, Tal Zussman <tz2294@columbia.edu> wrote:
> 
> mem_cgroup_zswap_writeback_enabled() only reads the zswap_writeback
> flags, and mem_cgroup_get_socket_pressure() only reads the socket
> pressure timestamp. Constify them.
> 
> Signed-off-by: Tal Zussman <tz2294@columbia.edu>

Acked-by: Muchun Song <muchun.song@linux.dev>

Thanks.

Re: [PATCH v3 00/11] mm: memcontrol: constify the read side of the memcg API

[Shakeel Butt]

On Tue, Sep 15, 2026 at 07:20:04PM -0400, Tal Zussman wrote:
> The memcg accessors, lruvec helpers, and stat readers only read from
> the memcg, folio, or lruvec they are given, but take non-const
> pointers. Constify them, along with the page_counter readers and the
> swap I/O blkg helpers along the way.
> 
> This started as a follow-up to [1] that quickly grew in proportions :)
> 
> [1]: https://lore.kernel.org/linux-mm/anYqSDnC2BbJXv1h@casper.infradead.org/

For the whole series:

Acked-by: Shakeel Butt <shakeel.butt@linux.dev>

Re: [PATCH v3 00/11] mm: memcontrol: constify the read side of the memcg API

[Lorenzo Stoakes (ARM)]

On Tue, Sep 15, 2026 at 08:53:10PM -0700, Shakeel Butt wrote:
> On Tue, Sep 15, 2026 at 07:20:04PM -0400, Tal Zussman wrote:
> > The memcg accessors, lruvec helpers, and stat readers only read from
> > the memcg, folio, or lruvec they are given, but take non-const
> > pointers. Constify them, along with the page_counter readers and the
> > swap I/O blkg helpers along the way.
> >
> > This started as a follow-up to [1] that quickly grew in proportions :)
> >
> > [1]: https://lore.kernel.org/linux-mm/anYqSDnC2BbJXv1h@casper.infradead.org/
>
> For the whole series:
>
> Acked-by: Shakeel Butt <shakeel.butt@linux.dev>

Oops, I replied on the v2 with the same:

https://lore.kernel.org/linux-mm/aqqaaDqTDahlZ7bR@gremlin/

but obviously, equally so for v3 :) thanks for doing this!

For whole series:

Acked-by: Lorenzo Stoakes (ARM) <ljs@kernel.org>

--
Cheers, Lorenzo