[linux.git] / mm / slab.h

/* SPDX-License-Identifier: GPL-2.0 */
#ifndef MM_SLAB_H
#define MM_SLAB_H
/*
 * Internal slab definitions
 */

#ifdef CONFIG_SLOB
/*
 * Common fields provided in kmem_cache by all slab allocators
 * This struct is either used directly by the allocator (SLOB)
 * or the allocator must include definitions for all fields
 * provided in kmem_cache_common in their definition of kmem_cache.
 *
 * Once we can do anonymous structs (C11 standard) we could put a
 * anonymous struct definition in these allocators so that the
 * separate allocations in the kmem_cache structure of SLAB and
 * SLUB is no longer needed.
 */
struct kmem_cache {
	unsigned int object_size;/* The original size of the object */
	unsigned int size;	/* The aligned/padded/added on size  */
	unsigned int align;	/* Alignment as calculated */
	slab_flags_t flags;	/* Active flags on the slab */
	unsigned int useroffset;/* Usercopy region offset */
	unsigned int usersize;	/* Usercopy region size */
	const char *name;	/* Slab name for sysfs */
	int refcount;		/* Use counter */
	void (*ctor)(void *);	/* Called on object slot creation */
	struct list_head list;	/* List of all slab caches on the system */
};

#endif /* CONFIG_SLOB */

#ifdef CONFIG_SLAB
#include <linux/slab_def.h>
#endif

#ifdef CONFIG_SLUB
#include <linux/slub_def.h>
#endif

#include <linux/memcontrol.h>
#include <linux/fault-inject.h>
#include <linux/kasan.h>
#include <linux/kmemleak.h>
#include <linux/random.h>
#include <linux/sched/mm.h>

/*
 * State of the slab allocator.
 *
 * This is used to describe the states of the allocator during bootup.
 * Allocators use this to gradually bootstrap themselves. Most allocators
 * have the problem that the structures used for managing slab caches are
 * allocated from slab caches themselves.
 */
enum slab_state {
	DOWN,			/* No slab functionality yet */
	PARTIAL,		/* SLUB: kmem_cache_node available */
	PARTIAL_NODE,		/* SLAB: kmalloc size for node struct available */
	UP,			/* Slab caches usable but not all extras yet */
	FULL			/* Everything is working */
};

extern enum slab_state slab_state;

/* The slab cache mutex protects the management structures during changes */
extern struct mutex slab_mutex;

/* The list of all slab caches on the system */
extern struct list_head slab_caches;

/* The slab cache that manages slab cache information */
extern struct kmem_cache *kmem_cache;

/* A table of kmalloc cache names and sizes */
extern const struct kmalloc_info_struct {
	const char *name[NR_KMALLOC_TYPES];
	unsigned int size;
} kmalloc_info[];

#ifndef CONFIG_SLOB
/* Kmalloc array related functions */
void setup_kmalloc_cache_index_table(void);
void create_kmalloc_caches(slab_flags_t);

/* Find the kmalloc slab corresponding for a certain size */
struct kmem_cache *kmalloc_slab(size_t, gfp_t);
#endif

gfp_t kmalloc_fix_flags(gfp_t flags);

/* Functions provided by the slab allocators */
int __kmem_cache_create(struct kmem_cache *, slab_flags_t flags);

struct kmem_cache *create_kmalloc_cache(const char *name, unsigned int size,
			slab_flags_t flags, unsigned int useroffset,
			unsigned int usersize);
extern void create_boot_cache(struct kmem_cache *, const char *name,
			unsigned int size, slab_flags_t flags,
			unsigned int useroffset, unsigned int usersize);

int slab_unmergeable(struct kmem_cache *s);
struct kmem_cache *find_mergeable(unsigned size, unsigned align,
		slab_flags_t flags, const char *name, void (*ctor)(void *));
#ifndef CONFIG_SLOB
struct kmem_cache *
__kmem_cache_alias(const char *name, unsigned int size, unsigned int align,
		   slab_flags_t flags, void (*ctor)(void *));

slab_flags_t kmem_cache_flags(unsigned int object_size,
	slab_flags_t flags, const char *name,
	void (*ctor)(void *));
#else
static inline struct kmem_cache *
__kmem_cache_alias(const char *name, unsigned int size, unsigned int align,
		   slab_flags_t flags, void (*ctor)(void *))
{ return NULL; }

static inline slab_flags_t kmem_cache_flags(unsigned int object_size,
	slab_flags_t flags, const char *name,
	void (*ctor)(void *))
{
	return flags;
}
#endif


/* Legal flag mask for kmem_cache_create(), for various configurations */
#define SLAB_CORE_FLAGS (SLAB_HWCACHE_ALIGN | SLAB_CACHE_DMA | \
			 SLAB_CACHE_DMA32 | SLAB_PANIC | \
			 SLAB_TYPESAFE_BY_RCU | SLAB_DEBUG_OBJECTS )

#if defined(CONFIG_DEBUG_SLAB)
#define SLAB_DEBUG_FLAGS (SLAB_RED_ZONE | SLAB_POISON | SLAB_STORE_USER)
#elif defined(CONFIG_SLUB_DEBUG)
#define SLAB_DEBUG_FLAGS (SLAB_RED_ZONE | SLAB_POISON | SLAB_STORE_USER | \
			  SLAB_TRACE | SLAB_CONSISTENCY_CHECKS)
#else
#define SLAB_DEBUG_FLAGS (0)
#endif

#if defined(CONFIG_SLAB)
#define SLAB_CACHE_FLAGS (SLAB_MEM_SPREAD | SLAB_NOLEAKTRACE | \
			  SLAB_RECLAIM_ACCOUNT | SLAB_TEMPORARY | \
			  SLAB_ACCOUNT)
#elif defined(CONFIG_SLUB)
#define SLAB_CACHE_FLAGS (SLAB_NOLEAKTRACE | SLAB_RECLAIM_ACCOUNT | \
			  SLAB_TEMPORARY | SLAB_ACCOUNT)
#else
#define SLAB_CACHE_FLAGS (0)
#endif

/* Common flags available with current configuration */
#define CACHE_CREATE_MASK (SLAB_CORE_FLAGS | SLAB_DEBUG_FLAGS | SLAB_CACHE_FLAGS)

/* Common flags permitted for kmem_cache_create */
#define SLAB_FLAGS_PERMITTED (SLAB_CORE_FLAGS | \
			      SLAB_RED_ZONE | \
			      SLAB_POISON | \
			      SLAB_STORE_USER | \
			      SLAB_TRACE | \
			      SLAB_CONSISTENCY_CHECKS | \
			      SLAB_MEM_SPREAD | \
			      SLAB_NOLEAKTRACE | \
			      SLAB_RECLAIM_ACCOUNT | \
			      SLAB_TEMPORARY | \
			      SLAB_ACCOUNT)

bool __kmem_cache_empty(struct kmem_cache *);
int __kmem_cache_shutdown(struct kmem_cache *);
void __kmem_cache_release(struct kmem_cache *);
int __kmem_cache_shrink(struct kmem_cache *);
void slab_kmem_cache_release(struct kmem_cache *);

struct seq_file;
struct file;

struct slabinfo {
	unsigned long active_objs;
	unsigned long num_objs;
	unsigned long active_slabs;
	unsigned long num_slabs;
	unsigned long shared_avail;
	unsigned int limit;
	unsigned int batchcount;
	unsigned int shared;
	unsigned int objects_per_slab;
	unsigned int cache_order;
};

void get_slabinfo(struct kmem_cache *s, struct slabinfo *sinfo);
void slabinfo_show_stats(struct seq_file *m, struct kmem_cache *s);
ssize_t slabinfo_write(struct file *file, const char __user *buffer,
		       size_t count, loff_t *ppos);

/*
 * Generic implementation of bulk operations
 * These are useful for situations in which the allocator cannot
 * perform optimizations. In that case segments of the object listed
 * may be allocated or freed using these operations.
 */
void __kmem_cache_free_bulk(struct kmem_cache *, size_t, void **);
int __kmem_cache_alloc_bulk(struct kmem_cache *, gfp_t, size_t, void **);

static inline enum node_stat_item cache_vmstat_idx(struct kmem_cache *s)
{
	return (s->flags & SLAB_RECLAIM_ACCOUNT) ?
		NR_SLAB_RECLAIMABLE_B : NR_SLAB_UNRECLAIMABLE_B;
}

#ifdef CONFIG_SLUB_DEBUG
#ifdef CONFIG_SLUB_DEBUG_ON
DECLARE_STATIC_KEY_TRUE(slub_debug_enabled);
#else
DECLARE_STATIC_KEY_FALSE(slub_debug_enabled);
#endif
extern void print_tracking(struct kmem_cache *s, void *object);
#else
static inline void print_tracking(struct kmem_cache *s, void *object)
{
}
#endif

/*
 * Returns true if any of the specified slub_debug flags is enabled for the
 * cache. Use only for flags parsed by setup_slub_debug() as it also enables
 * the static key.
 */
static inline bool kmem_cache_debug_flags(struct kmem_cache *s, slab_flags_t flags)
{
#ifdef CONFIG_SLUB_DEBUG
	VM_WARN_ON_ONCE(!(flags & SLAB_DEBUG_FLAGS));
	if (static_branch_unlikely(&slub_debug_enabled))
		return s->flags & flags;
#endif
	return false;
}

#ifdef CONFIG_MEMCG_KMEM
int memcg_alloc_page_obj_cgroups(struct page *page, struct kmem_cache *s,
				 gfp_t gfp);

static inline void memcg_free_page_obj_cgroups(struct page *page)
{
	kfree(page_objcgs(page));
	page->memcg_data = 0;
}

static inline size_t obj_full_size(struct kmem_cache *s)
{
	/*
	 * For each accounted object there is an extra space which is used
	 * to store obj_cgroup membership. Charge it too.
	 */
	return s->size + sizeof(struct obj_cgroup *);
}

/*
 * Returns false if the allocation should fail.
 */
static inline bool memcg_slab_pre_alloc_hook(struct kmem_cache *s,
					     struct obj_cgroup **objcgp,
					     size_t objects, gfp_t flags)
{
	struct obj_cgroup *objcg;

	if (!memcg_kmem_enabled())
		return true;

	if (!(flags & __GFP_ACCOUNT) && !(s->flags & SLAB_ACCOUNT))
		return true;

	objcg = get_obj_cgroup_from_current();
	if (!objcg)
		return true;

	if (obj_cgroup_charge(objcg, flags, objects * obj_full_size(s))) {
		obj_cgroup_put(objcg);
		return false;
	}

	*objcgp = objcg;
	return true;
}

static inline void mod_objcg_state(struct obj_cgroup *objcg,
				   struct pglist_data *pgdat,
				   enum node_stat_item idx, int nr)
{
	struct mem_cgroup *memcg;
	struct lruvec *lruvec;

	rcu_read_lock();
	memcg = obj_cgroup_memcg(objcg);
	lruvec = mem_cgroup_lruvec(memcg, pgdat);
	mod_memcg_lruvec_state(lruvec, idx, nr);
	rcu_read_unlock();
}

static inline void memcg_slab_post_alloc_hook(struct kmem_cache *s,
					      struct obj_cgroup *objcg,
					      gfp_t flags, size_t size,
					      void **p)
{
	struct page *page;
	unsigned long off;
	size_t i;

	if (!memcg_kmem_enabled() || !objcg)
		return;

	flags &= ~__GFP_ACCOUNT;
	for (i = 0; i < size; i++) {
		if (likely(p[i])) {
			page = virt_to_head_page(p[i]);

			if (!page_objcgs(page) &&
			    memcg_alloc_page_obj_cgroups(page, s, flags)) {
				obj_cgroup_uncharge(objcg, obj_full_size(s));
				continue;
			}

			off = obj_to_index(s, page, p[i]);
			obj_cgroup_get(objcg);
			page_objcgs(page)[off] = objcg;
			mod_objcg_state(objcg, page_pgdat(page),
					cache_vmstat_idx(s), obj_full_size(s));
		} else {
			obj_cgroup_uncharge(objcg, obj_full_size(s));
		}
	}
	obj_cgroup_put(objcg);
}

static inline void memcg_slab_free_hook(struct kmem_cache *s_orig,
					void **p, int objects)
{
	struct kmem_cache *s;
	struct obj_cgroup **objcgs;
	struct obj_cgroup *objcg;
	struct page *page;
	unsigned int off;
	int i;

	if (!memcg_kmem_enabled())
		return;

	for (i = 0; i < objects; i++) {
		if (unlikely(!p[i]))
			continue;

		page = virt_to_head_page(p[i]);
		objcgs = page_objcgs(page);
		if (!objcgs)
			continue;

		if (!s_orig)
			s = page->slab_cache;
		else
			s = s_orig;

		off = obj_to_index(s, page, p[i]);
		objcg = objcgs[off];
		if (!objcg)
			continue;

		objcgs[off] = NULL;
		obj_cgroup_uncharge(objcg, obj_full_size(s));
		mod_objcg_state(objcg, page_pgdat(page), cache_vmstat_idx(s),
				-obj_full_size(s));
		obj_cgroup_put(objcg);
	}
}

#else /* CONFIG_MEMCG_KMEM */
static inline struct mem_cgroup *memcg_from_slab_obj(void *ptr)
{
	return NULL;
}

static inline int memcg_alloc_page_obj_cgroups(struct page *page,
					       struct kmem_cache *s, gfp_t gfp)
{
	return 0;
}

static inline void memcg_free_page_obj_cgroups(struct page *page)
{
}

static inline bool memcg_slab_pre_alloc_hook(struct kmem_cache *s,
					     struct obj_cgroup **objcgp,
					     size_t objects, gfp_t flags)
{
	return true;
}

static inline void memcg_slab_post_alloc_hook(struct kmem_cache *s,
					      struct obj_cgroup *objcg,
					      gfp_t flags, size_t size,
					      void **p)
{
}

static inline void memcg_slab_free_hook(struct kmem_cache *s,
					void **p, int objects)
{
}
#endif /* CONFIG_MEMCG_KMEM */

static inline struct kmem_cache *virt_to_cache(const void *obj)
{
	struct page *page;

	page = virt_to_head_page(obj);
	if (WARN_ONCE(!PageSlab(page), "%s: Object is not a Slab page!\n",
					__func__))
		return NULL;
	return page->slab_cache;
}

static __always_inline void account_slab_page(struct page *page, int order,
					      struct kmem_cache *s)
{
	mod_node_page_state(page_pgdat(page), cache_vmstat_idx(s),
			    PAGE_SIZE << order);
}

static __always_inline void unaccount_slab_page(struct page *page, int order,
						struct kmem_cache *s)
{
	if (memcg_kmem_enabled())
		memcg_free_page_obj_cgroups(page);

	mod_node_page_state(page_pgdat(page), cache_vmstat_idx(s),
			    -(PAGE_SIZE << order));
}

static inline struct kmem_cache *cache_from_obj(struct kmem_cache *s, void *x)
{
	struct kmem_cache *cachep;

	if (!IS_ENABLED(CONFIG_SLAB_FREELIST_HARDENED) &&
	    !kmem_cache_debug_flags(s, SLAB_CONSISTENCY_CHECKS))
		return s;

	cachep = virt_to_cache(x);
	if (WARN(cachep && cachep != s,
		  "%s: Wrong slab cache. %s but object is from %s\n",
		  __func__, s->name, cachep->name))
		print_tracking(cachep, x);
	return cachep;
}

static inline size_t slab_ksize(const struct kmem_cache *s)
{
#ifndef CONFIG_SLUB
	return s->object_size;

#else /* CONFIG_SLUB */
# ifdef CONFIG_SLUB_DEBUG
	/*
	 * Debugging requires use of the padding between object
	 * and whatever may come after it.
	 */
	if (s->flags & (SLAB_RED_ZONE | SLAB_POISON))
		return s->object_size;
# endif
	if (s->flags & SLAB_KASAN)
		return s->object_size;
	/*
	 * If we have the need to store the freelist pointer
	 * back there or track user information then we can
	 * only use the space before that information.
	 */
	if (s->flags & (SLAB_TYPESAFE_BY_RCU | SLAB_STORE_USER))
		return s->inuse;
	/*
	 * Else we can use all the padding etc for the allocation
	 */
	return s->size;
#endif
}

static inline struct kmem_cache *slab_pre_alloc_hook(struct kmem_cache *s,
						     struct obj_cgroup **objcgp,
						     size_t size, gfp_t flags)
{
	flags &= gfp_allowed_mask;

	might_alloc(flags);

	if (should_failslab(s, flags))
		return NULL;

	if (!memcg_slab_pre_alloc_hook(s, objcgp, size, flags))
		return NULL;

	return s;
}

static inline void slab_post_alloc_hook(struct kmem_cache *s,
					struct obj_cgroup *objcg,
					gfp_t flags, size_t size, void **p)
{
	size_t i;

	flags &= gfp_allowed_mask;
	for (i = 0; i < size; i++) {
		p[i] = kasan_slab_alloc(s, p[i], flags);
		/* As p[i] might get tagged, call kmemleak hook after KASAN. */
		kmemleak_alloc_recursive(p[i], s->object_size, 1,
					 s->flags, flags);
	}

	memcg_slab_post_alloc_hook(s, objcg, flags, size, p);
}

#ifndef CONFIG_SLOB
/*
 * The slab lists for all objects.
 */
struct kmem_cache_node {
	spinlock_t list_lock;

#ifdef CONFIG_SLAB
	struct list_head slabs_partial;	/* partial list first, better asm code */
	struct list_head slabs_full;
	struct list_head slabs_free;
	unsigned long total_slabs;	/* length of all slab lists */
	unsigned long free_slabs;	/* length of free slab list only */
	unsigned long free_objects;
	unsigned int free_limit;
	unsigned int colour_next;	/* Per-node cache coloring */
	struct array_cache *shared;	/* shared per node */
	struct alien_cache **alien;	/* on other nodes */
	unsigned long next_reap;	/* updated without locking */
	int free_touched;		/* updated without locking */
#endif

#ifdef CONFIG_SLUB
	unsigned long nr_partial;
	struct list_head partial;
#ifdef CONFIG_SLUB_DEBUG
	atomic_long_t nr_slabs;
	atomic_long_t total_objects;
	struct list_head full;
#endif
#endif

};

static inline struct kmem_cache_node *get_node(struct kmem_cache *s, int node)
{
	return s->node[node];
}

/*
 * Iterator over all nodes. The body will be executed for each node that has
 * a kmem_cache_node structure allocated (which is true for all online nodes)
 */
#define for_each_kmem_cache_node(__s, __node, __n) \
	for (__node = 0; __node < nr_node_ids; __node++) \
		 if ((__n = get_node(__s, __node)))

#endif

void *slab_start(struct seq_file *m, loff_t *pos);
void *slab_next(struct seq_file *m, void *p, loff_t *pos);
void slab_stop(struct seq_file *m, void *p);
int memcg_slab_show(struct seq_file *m, void *p);

#if defined(CONFIG_SLAB) || defined(CONFIG_SLUB_DEBUG)
void dump_unreclaimable_slab(void);
#else
static inline void dump_unreclaimable_slab(void)
{
}
#endif

void ___cache_free(struct kmem_cache *cache, void *x, unsigned long addr);

#ifdef CONFIG_SLAB_FREELIST_RANDOM
int cache_random_seq_create(struct kmem_cache *cachep, unsigned int count,
			gfp_t gfp);
void cache_random_seq_destroy(struct kmem_cache *cachep);
#else
static inline int cache_random_seq_create(struct kmem_cache *cachep,
					unsigned int count, gfp_t gfp)
{
	return 0;
}
static inline void cache_random_seq_destroy(struct kmem_cache *cachep) { }
#endif /* CONFIG_SLAB_FREELIST_RANDOM */

static inline bool slab_want_init_on_alloc(gfp_t flags, struct kmem_cache *c)
{
	if (static_branch_unlikely(&init_on_alloc)) {
		if (c->ctor)
			return false;
		if (c->flags & (SLAB_TYPESAFE_BY_RCU | SLAB_POISON))
			return flags & __GFP_ZERO;
		return true;
	}
	return flags & __GFP_ZERO;
}

static inline bool slab_want_init_on_free(struct kmem_cache *c)
{
	if (static_branch_unlikely(&init_on_free))
		return !(c->ctor ||
			 (c->flags & (SLAB_TYPESAFE_BY_RCU | SLAB_POISON)));
	return false;
}

#endif /* MM_SLAB_H */
Commit	Line	Data
b2441318	1	/* SPDX-License-Identifier: GPL-2.0 */
97d06609 CL	2	#ifndef MM_SLAB_H
	3	#define MM_SLAB_H
	4	/*
	5	* Internal slab definitions
	6	*/
	7
07f361b2 JK	8	#ifdef CONFIG_SLOB
	9	/*
	10	* Common fields provided in kmem_cache by all slab allocators
	11	* This struct is either used directly by the allocator (SLOB)
	12	* or the allocator must include definitions for all fields
	13	* provided in kmem_cache_common in their definition of kmem_cache.
	14	*
	15	* Once we can do anonymous structs (C11 standard) we could put a
	16	* anonymous struct definition in these allocators so that the
	17	* separate allocations in the kmem_cache structure of SLAB and
	18	* SLUB is no longer needed.
	19	*/
	20	struct kmem_cache {
	21	unsigned int object_size;/* The original size of the object */
	22	unsigned int size; /* The aligned/padded/added on size */
	23	unsigned int align; /* Alignment as calculated */
d50112ed	24	slab_flags_t flags; /* Active flags on the slab */
7bbdb81e AD	25	unsigned int useroffset;/* Usercopy region offset */
7bbdb81e AD	26	unsigned int usersize; /* Usercopy region size */
07f361b2 JK	27	const char name; / Slab name for sysfs */
	28	int refcount; /* Use counter */
	29	void (ctor)(void ); /* Called on object slot creation */
	30	struct list_head list; /* List of all slab caches on the system */
	31	};
	32
	33	#endif /* CONFIG_SLOB */
	34
	35	#ifdef CONFIG_SLAB
	36	#include <linux/slab_def.h>
	37	#endif
	38
	39	#ifdef CONFIG_SLUB
	40	#include <linux/slub_def.h>
	41	#endif
	42
	43	#include <linux/memcontrol.h>
11c7aec2	44	#include <linux/fault-inject.h>
11c7aec2 JDB	45	#include <linux/kasan.h>
11c7aec2 JDB	46	#include <linux/kmemleak.h>
7c00fce9	47	#include <linux/random.h>
d92a8cfc	48	#include <linux/sched/mm.h>
07f361b2	49
97d06609 CL	50	/*
	51	* State of the slab allocator.
	52	*
	53	* This is used to describe the states of the allocator during bootup.
	54	* Allocators use this to gradually bootstrap themselves. Most allocators
	55	* have the problem that the structures used for managing slab caches are
	56	* allocated from slab caches themselves.
	57	*/
	58	enum slab_state {
	59	DOWN, /* No slab functionality yet */
	60	PARTIAL, /* SLUB: kmem_cache_node available */
ce8eb6c4	61	PARTIAL_NODE, /* SLAB: kmalloc size for node struct available */
97d06609 CL	62	UP, /* Slab caches usable but not all extras yet */
	63	FULL /* Everything is working */
	64	};
	65
	66	extern enum slab_state slab_state;
	67
18004c5d CL	68	/* The slab cache mutex protects the management structures during changes */
18004c5d CL	69	extern struct mutex slab_mutex;
9b030cb8 CL	70
9b030cb8 CL	71	/* The list of all slab caches on the system */
18004c5d CL	72	extern struct list_head slab_caches;
18004c5d CL	73
9b030cb8 CL	74	/* The slab cache that manages slab cache information */
	75	extern struct kmem_cache *kmem_cache;
	76
af3b5f87 VB	77	/* A table of kmalloc cache names and sizes */
af3b5f87 VB	78	extern const struct kmalloc_info_struct {
cb5d9fb3	79	const char *name[NR_KMALLOC_TYPES];
55de8b9c	80	unsigned int size;
af3b5f87 VB	81	} kmalloc_info[];
af3b5f87 VB	82
f97d5f63 CL	83	#ifndef CONFIG_SLOB
f97d5f63 CL	84	/* Kmalloc array related functions */
34cc6990	85	void setup_kmalloc_cache_index_table(void);
d50112ed	86	void create_kmalloc_caches(slab_flags_t);
2c59dd65 CL	87
	88	/* Find the kmalloc slab corresponding for a certain size */
	89	struct kmem_cache *kmalloc_slab(size_t, gfp_t);
f97d5f63 CL	90	#endif
f97d5f63 CL	91
44405099	92	gfp_t kmalloc_fix_flags(gfp_t flags);
f97d5f63	93
9b030cb8	94	/* Functions provided by the slab allocators */
d50112ed	95	int __kmem_cache_create(struct kmem_cache *, slab_flags_t flags);
97d06609	96
55de8b9c AD	97	struct kmem_cache create_kmalloc_cache(const char name, unsigned int size,
	98	slab_flags_t flags, unsigned int useroffset,
	99	unsigned int usersize);
45530c44	100	extern void create_boot_cache(struct kmem_cache , const char name,
361d575e AD	101	unsigned int size, slab_flags_t flags,
361d575e AD	102	unsigned int useroffset, unsigned int usersize);
45530c44	103
423c929c	104	int slab_unmergeable(struct kmem_cache *s);
f4957d5b	105	struct kmem_cache *find_mergeable(unsigned size, unsigned align,
d50112ed	106	slab_flags_t flags, const char name, void (ctor)(void *));
12220dea	107	#ifndef CONFIG_SLOB
2633d7a0	108	struct kmem_cache *
f4957d5b	109	__kmem_cache_alias(const char *name, unsigned int size, unsigned int align,
d50112ed	110	slab_flags_t flags, void (ctor)(void ));
423c929c	111
0293d1fd	112	slab_flags_t kmem_cache_flags(unsigned int object_size,
d50112ed	113	slab_flags_t flags, const char *name,
423c929c	114	void (ctor)(void ));
cbb79694	115	#else
2633d7a0	116	static inline struct kmem_cache *
f4957d5b	117	__kmem_cache_alias(const char *name, unsigned int size, unsigned int align,
d50112ed	118	slab_flags_t flags, void (ctor)(void ))
cbb79694	119	{ return NULL; }
423c929c	120
0293d1fd	121	static inline slab_flags_t kmem_cache_flags(unsigned int object_size,
d50112ed	122	slab_flags_t flags, const char *name,
423c929c JK	123	void (ctor)(void ))
	124	{
	125	return flags;
	126	}
cbb79694 CL	127	#endif
	128
	129
d8843922	130	/* Legal flag mask for kmem_cache_create(), for various configurations */
6d6ea1e9 NB	131	#define SLAB_CORE_FLAGS (SLAB_HWCACHE_ALIGN \| SLAB_CACHE_DMA \| \
6d6ea1e9 NB	132	SLAB_CACHE_DMA32 \| SLAB_PANIC \| \
5f0d5a3a	133	SLAB_TYPESAFE_BY_RCU \| SLAB_DEBUG_OBJECTS )
d8843922 GC	134
	135	#if defined(CONFIG_DEBUG_SLAB)
	136	#define SLAB_DEBUG_FLAGS (SLAB_RED_ZONE \| SLAB_POISON \| SLAB_STORE_USER)
	137	#elif defined(CONFIG_SLUB_DEBUG)
	138	#define SLAB_DEBUG_FLAGS (SLAB_RED_ZONE \| SLAB_POISON \| SLAB_STORE_USER \| \
becfda68	139	SLAB_TRACE \| SLAB_CONSISTENCY_CHECKS)
d8843922 GC	140	#else
	141	#define SLAB_DEBUG_FLAGS (0)
	142	#endif
	143
	144	#if defined(CONFIG_SLAB)
	145	#define SLAB_CACHE_FLAGS (SLAB_MEM_SPREAD \| SLAB_NOLEAKTRACE \| \
230e9fc2	146	SLAB_RECLAIM_ACCOUNT \| SLAB_TEMPORARY \| \
75f296d9	147	SLAB_ACCOUNT)
d8843922 GC	148	#elif defined(CONFIG_SLUB)
d8843922 GC	149	#define SLAB_CACHE_FLAGS (SLAB_NOLEAKTRACE \| SLAB_RECLAIM_ACCOUNT \| \
75f296d9	150	SLAB_TEMPORARY \| SLAB_ACCOUNT)
d8843922 GC	151	#else
	152	#define SLAB_CACHE_FLAGS (0)
	153	#endif
	154
e70954fd	155	/* Common flags available with current configuration */
d8843922 GC	156	#define CACHE_CREATE_MASK (SLAB_CORE_FLAGS \| SLAB_DEBUG_FLAGS \| SLAB_CACHE_FLAGS)
d8843922 GC	157
e70954fd TG	158	/* Common flags permitted for kmem_cache_create */
	159	#define SLAB_FLAGS_PERMITTED (SLAB_CORE_FLAGS \| \
	160	SLAB_RED_ZONE \| \
	161	SLAB_POISON \| \
	162	SLAB_STORE_USER \| \
	163	SLAB_TRACE \| \
	164	SLAB_CONSISTENCY_CHECKS \| \
	165	SLAB_MEM_SPREAD \| \
	166	SLAB_NOLEAKTRACE \| \
	167	SLAB_RECLAIM_ACCOUNT \| \
	168	SLAB_TEMPORARY \| \
e70954fd TG	169	SLAB_ACCOUNT)
e70954fd TG	170
f9e13c0a	171	bool __kmem_cache_empty(struct kmem_cache *);
945cf2b6	172	int __kmem_cache_shutdown(struct kmem_cache *);
52b4b950	173	void __kmem_cache_release(struct kmem_cache *);
c9fc5864	174	int __kmem_cache_shrink(struct kmem_cache *);
41a21285	175	void slab_kmem_cache_release(struct kmem_cache *);
945cf2b6	176
b7454ad3 GC	177	struct seq_file;
b7454ad3 GC	178	struct file;
b7454ad3	179
0d7561c6 GC	180	struct slabinfo {
	181	unsigned long active_objs;
	182	unsigned long num_objs;
	183	unsigned long active_slabs;
	184	unsigned long num_slabs;
	185	unsigned long shared_avail;
	186	unsigned int limit;
	187	unsigned int batchcount;
	188	unsigned int shared;
	189	unsigned int objects_per_slab;
	190	unsigned int cache_order;
	191	};
	192
	193	void get_slabinfo(struct kmem_cache s, struct slabinfo sinfo);
	194	void slabinfo_show_stats(struct seq_file m, struct kmem_cache s);
b7454ad3 GC	195	ssize_t slabinfo_write(struct file file, const char __user buffer,
b7454ad3 GC	196	size_t count, loff_t *ppos);
ba6c496e	197
484748f0 CL	198	/*
	199	* Generic implementation of bulk operations
	200	* These are useful for situations in which the allocator cannot
9f706d68	201	* perform optimizations. In that case segments of the object listed
484748f0 CL	202	* may be allocated or freed using these operations.
	203	*/
	204	void __kmem_cache_free_bulk(struct kmem_cache , size_t, void *);
865762a8	205	int __kmem_cache_alloc_bulk(struct kmem_cache , gfp_t, size_t, void *);
484748f0	206
1a984c4e	207	static inline enum node_stat_item cache_vmstat_idx(struct kmem_cache *s)
6cea1d56 RG	208	{
6cea1d56 RG	209	return (s->flags & SLAB_RECLAIM_ACCOUNT) ?
d42f3245	210	NR_SLAB_RECLAIMABLE_B : NR_SLAB_UNRECLAIMABLE_B;
6cea1d56 RG	211	}
6cea1d56 RG	212
e42f174e VB	213	#ifdef CONFIG_SLUB_DEBUG
	214	#ifdef CONFIG_SLUB_DEBUG_ON
	215	DECLARE_STATIC_KEY_TRUE(slub_debug_enabled);
	216	#else
	217	DECLARE_STATIC_KEY_FALSE(slub_debug_enabled);
	218	#endif
	219	extern void print_tracking(struct kmem_cache s, void object);
	220	#else
	221	static inline void print_tracking(struct kmem_cache s, void object)
	222	{
	223	}
	224	#endif
	225
	226	/*
	227	* Returns true if any of the specified slub_debug flags is enabled for the
	228	* cache. Use only for flags parsed by setup_slub_debug() as it also enables
	229	* the static key.
	230	*/
	231	static inline bool kmem_cache_debug_flags(struct kmem_cache *s, slab_flags_t flags)
	232	{
	233	#ifdef CONFIG_SLUB_DEBUG
	234	VM_WARN_ON_ONCE(!(flags & SLAB_DEBUG_FLAGS));
	235	if (static_branch_unlikely(&slub_debug_enabled))
	236	return s->flags & flags;
	237	#endif
	238	return false;
	239	}
	240
84c07d11	241	#ifdef CONFIG_MEMCG_KMEM
10befea9 RG	242	int memcg_alloc_page_obj_cgroups(struct page page, struct kmem_cache s,
10befea9 RG	243	gfp_t gfp);
286e04b8 RG	244
	245	static inline void memcg_free_page_obj_cgroups(struct page *page)
	246	{
270c6a71	247	kfree(page_objcgs(page));
bcfe06bf	248	page->memcg_data = 0;
286e04b8 RG	249	}
286e04b8 RG	250
f2fe7b09 RG	251	static inline size_t obj_full_size(struct kmem_cache *s)
	252	{
	253	/*
	254	* For each accounted object there is an extra space which is used
	255	* to store obj_cgroup membership. Charge it too.
	256	*/
	257	return s->size + sizeof(struct obj_cgroup *);
	258	}
	259
becaba65 RG	260	/*
	261	* Returns false if the allocation should fail.
	262	*/
	263	static inline bool memcg_slab_pre_alloc_hook(struct kmem_cache *s,
	264	struct obj_cgroup **objcgp,
	265	size_t objects, gfp_t flags)
f2fe7b09	266	{
9855609b RG	267	struct obj_cgroup *objcg;
9855609b RG	268
becaba65 RG	269	if (!memcg_kmem_enabled())
	270	return true;
	271
	272	if (!(flags & __GFP_ACCOUNT) && !(s->flags & SLAB_ACCOUNT))
	273	return true;
	274
9855609b RG	275	objcg = get_obj_cgroup_from_current();
9855609b RG	276	if (!objcg)
becaba65	277	return true;
9855609b RG	278
	279	if (obj_cgroup_charge(objcg, flags, objects * obj_full_size(s))) {
	280	obj_cgroup_put(objcg);
becaba65	281	return false;
f2fe7b09 RG	282	}
f2fe7b09 RG	283
becaba65 RG	284	*objcgp = objcg;
becaba65 RG	285	return true;
f2fe7b09 RG	286	}
	287
	288	static inline void mod_objcg_state(struct obj_cgroup *objcg,
	289	struct pglist_data *pgdat,
1a984c4e	290	enum node_stat_item idx, int nr)
f2fe7b09 RG	291	{
	292	struct mem_cgroup *memcg;
	293	struct lruvec *lruvec;
	294
	295	rcu_read_lock();
	296	memcg = obj_cgroup_memcg(objcg);
	297	lruvec = mem_cgroup_lruvec(memcg, pgdat);
	298	mod_memcg_lruvec_state(lruvec, idx, nr);
	299	rcu_read_unlock();
	300	}
	301
964d4bd3 RG	302	static inline void memcg_slab_post_alloc_hook(struct kmem_cache *s,
964d4bd3 RG	303	struct obj_cgroup *objcg,
10befea9 RG	304	gfp_t flags, size_t size,
10befea9 RG	305	void **p)
964d4bd3 RG	306	{
	307	struct page *page;
	308	unsigned long off;
	309	size_t i;
	310
becaba65	311	if (!memcg_kmem_enabled() \|\| !objcg)
10befea9 RG	312	return;
	313
	314	flags &= ~__GFP_ACCOUNT;
964d4bd3 RG	315	for (i = 0; i < size; i++) {
	316	if (likely(p[i])) {
	317	page = virt_to_head_page(p[i]);
10befea9	318
270c6a71	319	if (!page_objcgs(page) &&
10befea9 RG	320	memcg_alloc_page_obj_cgroups(page, s, flags)) {
	321	obj_cgroup_uncharge(objcg, obj_full_size(s));
	322	continue;
	323	}
	324
964d4bd3 RG	325	off = obj_to_index(s, page, p[i]);
964d4bd3 RG	326	obj_cgroup_get(objcg);
270c6a71	327	page_objcgs(page)[off] = objcg;
f2fe7b09 RG	328	mod_objcg_state(objcg, page_pgdat(page),
	329	cache_vmstat_idx(s), obj_full_size(s));
	330	} else {
	331	obj_cgroup_uncharge(objcg, obj_full_size(s));
964d4bd3 RG	332	}
	333	}
	334	obj_cgroup_put(objcg);
964d4bd3 RG	335	}
964d4bd3 RG	336
d1b2cf6c BR	337	static inline void memcg_slab_free_hook(struct kmem_cache *s_orig,
d1b2cf6c BR	338	void **p, int objects)
964d4bd3	339	{
d1b2cf6c	340	struct kmem_cache *s;
270c6a71	341	struct obj_cgroup **objcgs;
964d4bd3	342	struct obj_cgroup *objcg;
d1b2cf6c	343	struct page *page;
964d4bd3	344	unsigned int off;
d1b2cf6c	345	int i;
964d4bd3	346
10befea9 RG	347	if (!memcg_kmem_enabled())
	348	return;
	349
d1b2cf6c BR	350	for (i = 0; i < objects; i++) {
	351	if (unlikely(!p[i]))
	352	continue;
964d4bd3	353
d1b2cf6c	354	page = virt_to_head_page(p[i]);
270c6a71 RG	355	objcgs = page_objcgs(page);
270c6a71 RG	356	if (!objcgs)
d1b2cf6c	357	continue;
f2fe7b09	358
d1b2cf6c BR	359	if (!s_orig)
	360	s = page->slab_cache;
	361	else
	362	s = s_orig;
10befea9	363
d1b2cf6c	364	off = obj_to_index(s, page, p[i]);
270c6a71	365	objcg = objcgs[off];
d1b2cf6c BR	366	if (!objcg)
d1b2cf6c BR	367	continue;
f2fe7b09	368
270c6a71	369	objcgs[off] = NULL;
d1b2cf6c BR	370	obj_cgroup_uncharge(objcg, obj_full_size(s));
	371	mod_objcg_state(objcg, page_pgdat(page), cache_vmstat_idx(s),
	372	-obj_full_size(s));
	373	obj_cgroup_put(objcg);
	374	}
964d4bd3 RG	375	}
964d4bd3 RG	376
84c07d11	377	#else /* CONFIG_MEMCG_KMEM */
9855609b	378	static inline struct mem_cgroup memcg_from_slab_obj(void ptr)
4d96ba35 RG	379	{
	380	return NULL;
	381	}
	382
286e04b8 RG	383	static inline int memcg_alloc_page_obj_cgroups(struct page *page,
	384	struct kmem_cache *s, gfp_t gfp)
	385	{
	386	return 0;
	387	}
	388
	389	static inline void memcg_free_page_obj_cgroups(struct page *page)
	390	{
	391	}
	392
becaba65 RG	393	static inline bool memcg_slab_pre_alloc_hook(struct kmem_cache *s,
	394	struct obj_cgroup **objcgp,
	395	size_t objects, gfp_t flags)
f2fe7b09	396	{
becaba65	397	return true;
f2fe7b09 RG	398	}
f2fe7b09 RG	399
964d4bd3 RG	400	static inline void memcg_slab_post_alloc_hook(struct kmem_cache *s,
964d4bd3 RG	401	struct obj_cgroup *objcg,
10befea9 RG	402	gfp_t flags, size_t size,
10befea9 RG	403	void **p)
964d4bd3 RG	404	{
	405	}
	406
d1b2cf6c BR	407	static inline void memcg_slab_free_hook(struct kmem_cache *s,
d1b2cf6c BR	408	void **p, int objects)
964d4bd3 RG	409	{
964d4bd3 RG	410	}
84c07d11	411	#endif /* CONFIG_MEMCG_KMEM */
b9ce5ef4	412
a64b5378 KC	413	static inline struct kmem_cache virt_to_cache(const void obj)
	414	{
	415	struct page *page;
	416
	417	page = virt_to_head_page(obj);
	418	if (WARN_ONCE(!PageSlab(page), "%s: Object is not a Slab page!\n",
	419	__func__))
	420	return NULL;
	421	return page->slab_cache;
	422	}
	423
74d555be RG	424	static __always_inline void account_slab_page(struct page *page, int order,
74d555be RG	425	struct kmem_cache *s)
6cea1d56	426	{
f2fe7b09 RG	427	mod_node_page_state(page_pgdat(page), cache_vmstat_idx(s),
f2fe7b09 RG	428	PAGE_SIZE << order);
6cea1d56 RG	429	}
6cea1d56 RG	430
74d555be RG	431	static __always_inline void unaccount_slab_page(struct page *page, int order,
74d555be RG	432	struct kmem_cache *s)
6cea1d56	433	{
10befea9	434	if (memcg_kmem_enabled())
f2fe7b09	435	memcg_free_page_obj_cgroups(page);
9855609b	436
f2fe7b09 RG	437	mod_node_page_state(page_pgdat(page), cache_vmstat_idx(s),
f2fe7b09 RG	438	-(PAGE_SIZE << order));
6cea1d56 RG	439	}
6cea1d56 RG	440
e42f174e VB	441	static inline struct kmem_cache cache_from_obj(struct kmem_cache s, void *x)
	442	{
	443	struct kmem_cache *cachep;
	444
	445	if (!IS_ENABLED(CONFIG_SLAB_FREELIST_HARDENED) &&
e42f174e VB	446	!kmem_cache_debug_flags(s, SLAB_CONSISTENCY_CHECKS))
	447	return s;
	448
	449	cachep = virt_to_cache(x);
10befea9	450	if (WARN(cachep && cachep != s,
e42f174e VB	451	"%s: Wrong slab cache. %s but object is from %s\n",
	452	__func__, s->name, cachep->name))
	453	print_tracking(cachep, x);
	454	return cachep;
	455	}
	456
11c7aec2 JDB	457	static inline size_t slab_ksize(const struct kmem_cache *s)
	458	{
	459	#ifndef CONFIG_SLUB
	460	return s->object_size;
	461
	462	#else /* CONFIG_SLUB */
	463	# ifdef CONFIG_SLUB_DEBUG
	464	/*
	465	* Debugging requires use of the padding between object
	466	* and whatever may come after it.
	467	*/
	468	if (s->flags & (SLAB_RED_ZONE \| SLAB_POISON))
	469	return s->object_size;
	470	# endif
80a9201a AP	471	if (s->flags & SLAB_KASAN)
80a9201a AP	472	return s->object_size;
11c7aec2 JDB	473	/*
	474	* If we have the need to store the freelist pointer
	475	* back there or track user information then we can
	476	* only use the space before that information.
	477	*/
5f0d5a3a	478	if (s->flags & (SLAB_TYPESAFE_BY_RCU \| SLAB_STORE_USER))
11c7aec2 JDB	479	return s->inuse;
	480	/*
	481	* Else we can use all the padding etc for the allocation
	482	*/
	483	return s->size;
	484	#endif
	485	}
	486
	487	static inline struct kmem_cache slab_pre_alloc_hook(struct kmem_cache s,
964d4bd3 RG	488	struct obj_cgroup **objcgp,
964d4bd3 RG	489	size_t size, gfp_t flags)
11c7aec2 JDB	490	{
11c7aec2 JDB	491	flags &= gfp_allowed_mask;
d92a8cfc	492
95d6c701	493	might_alloc(flags);
11c7aec2	494
fab9963a	495	if (should_failslab(s, flags))
11c7aec2 JDB	496	return NULL;
11c7aec2 JDB	497
becaba65 RG	498	if (!memcg_slab_pre_alloc_hook(s, objcgp, size, flags))
becaba65 RG	499	return NULL;
45264778 VD	500
45264778 VD	501	return s;
11c7aec2 JDB	502	}
11c7aec2 JDB	503
964d4bd3 RG	504	static inline void slab_post_alloc_hook(struct kmem_cache *s,
	505	struct obj_cgroup *objcg,
	506	gfp_t flags, size_t size, void **p)
11c7aec2 JDB	507	{
	508	size_t i;
	509
	510	flags &= gfp_allowed_mask;
	511	for (i = 0; i < size; i++) {
53128245	512	p[i] = kasan_slab_alloc(s, p[i], flags);
a2f77575	513	/* As p[i] might get tagged, call kmemleak hook after KASAN. */
53128245	514	kmemleak_alloc_recursive(p[i], s->object_size, 1,
11c7aec2	515	s->flags, flags);
11c7aec2	516	}
45264778	517
becaba65	518	memcg_slab_post_alloc_hook(s, objcg, flags, size, p);
11c7aec2 JDB	519	}
11c7aec2 JDB	520
44c5356f	521	#ifndef CONFIG_SLOB
ca34956b CL	522	/*
	523	* The slab lists for all objects.
	524	*/
	525	struct kmem_cache_node {
	526	spinlock_t list_lock;
	527
	528	#ifdef CONFIG_SLAB
	529	struct list_head slabs_partial; /* partial list first, better asm code */
	530	struct list_head slabs_full;
	531	struct list_head slabs_free;
bf00bd34 DR	532	unsigned long total_slabs; /* length of all slab lists */
bf00bd34 DR	533	unsigned long free_slabs; /* length of free slab list only */
ca34956b CL	534	unsigned long free_objects;
	535	unsigned int free_limit;
	536	unsigned int colour_next; /* Per-node cache coloring */
	537	struct array_cache shared; / shared per node */
c8522a3a	538	struct alien_cache *alien; / on other nodes */
ca34956b CL	539	unsigned long next_reap; /* updated without locking */
	540	int free_touched; /* updated without locking */
	541	#endif
	542
	543	#ifdef CONFIG_SLUB
	544	unsigned long nr_partial;
	545	struct list_head partial;
	546	#ifdef CONFIG_SLUB_DEBUG
	547	atomic_long_t nr_slabs;
	548	atomic_long_t total_objects;
	549	struct list_head full;
	550	#endif
	551	#endif
	552
	553	};
e25839f6	554
44c5356f CL	555	static inline struct kmem_cache_node get_node(struct kmem_cache s, int node)
	556	{
	557	return s->node[node];
	558	}
	559
	560	/*
	561	* Iterator over all nodes. The body will be executed for each node that has
	562	* a kmem_cache_node structure allocated (which is true for all online nodes)
	563	*/
	564	#define for_each_kmem_cache_node(__s, __node, __n) \
9163582c MP	565	for (__node = 0; __node < nr_node_ids; __node++) \
9163582c MP	566	if ((__n = get_node(__s, __node)))
44c5356f CL	567
	568	#endif
	569
1df3b26f	570	void slab_start(struct seq_file m, loff_t *pos);
276a2439 WL	571	void slab_next(struct seq_file m, void p, loff_t pos);
276a2439 WL	572	void slab_stop(struct seq_file m, void p);
b047501c	573	int memcg_slab_show(struct seq_file m, void p);
5240ab40	574
852d8be0 YS	575	#if defined(CONFIG_SLAB) \|\| defined(CONFIG_SLUB_DEBUG)
	576	void dump_unreclaimable_slab(void);
	577	#else
	578	static inline void dump_unreclaimable_slab(void)
	579	{
	580	}
	581	#endif
	582
55834c59 AP	583	void ___cache_free(struct kmem_cache cache, void x, unsigned long addr);
55834c59 AP	584
7c00fce9 TG	585	#ifdef CONFIG_SLAB_FREELIST_RANDOM
	586	int cache_random_seq_create(struct kmem_cache *cachep, unsigned int count,
	587	gfp_t gfp);
	588	void cache_random_seq_destroy(struct kmem_cache *cachep);
	589	#else
	590	static inline int cache_random_seq_create(struct kmem_cache *cachep,
	591	unsigned int count, gfp_t gfp)
	592	{
	593	return 0;
	594	}
	595	static inline void cache_random_seq_destroy(struct kmem_cache *cachep) { }
	596	#endif /* CONFIG_SLAB_FREELIST_RANDOM */
	597
6471384a AP	598	static inline bool slab_want_init_on_alloc(gfp_t flags, struct kmem_cache *c)
	599	{
	600	if (static_branch_unlikely(&init_on_alloc)) {
	601	if (c->ctor)
	602	return false;
	603	if (c->flags & (SLAB_TYPESAFE_BY_RCU \| SLAB_POISON))
	604	return flags & __GFP_ZERO;
	605	return true;
	606	}
	607	return flags & __GFP_ZERO;
	608	}
	609
	610	static inline bool slab_want_init_on_free(struct kmem_cache *c)
	611	{
	612	if (static_branch_unlikely(&init_on_free))
	613	return !(c->ctor \|\|
	614	(c->flags & (SLAB_TYPESAFE_BY_RCU \| SLAB_POISON)));
	615	return false;
	616	}
	617
5240ab40	618	#endif /* MM_SLAB_H */