[mirror_ubuntu-jammy-kernel.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>
#include <linux/kmemleak.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 int 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
static inline struct obj_cgroup **page_obj_cgroups(struct page *page)
{
	/*
	 * page->mem_cgroup and page->obj_cgroups are sharing the same
	 * space. To distinguish between them in case we don't know for sure
	 * that the page is a slab page (e.g. page_cgroup_ino()), let's
	 * always set the lowest bit of obj_cgroups.
	 */
	return (struct obj_cgroup **)
		((unsigned long)page->obj_cgroups & ~0x1UL);
}

static inline bool page_has_obj_cgroups(struct page *page)
{
	return ((unsigned long)page->obj_cgroups & 0x1UL);
}

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_obj_cgroups(page));
	page->obj_cgroups = NULL;
}

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 *);
}

static inline struct obj_cgroup *memcg_slab_pre_alloc_hook(struct kmem_cache *s,
							   size_t objects,
							   gfp_t flags)
{
	struct obj_cgroup *objcg;

	if (memcg_kmem_bypass())
		return NULL;

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

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

	return objcg;
}

static inline void mod_objcg_state(struct obj_cgroup *objcg,
				   struct pglist_data *pgdat,
				   int 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 (!objcg)
		return;

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

			if (!page_has_obj_cgroups(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_obj_cgroups(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, struct page *page,
					void *p)
{
	struct obj_cgroup *objcg;
	unsigned int off;

	if (!memcg_kmem_enabled())
		return;

	if (!page_has_obj_cgroups(page))
		return;

	off = obj_to_index(s, page, p);
	objcg = page_obj_cgroups(page)[off];
	page_obj_cgroups(page)[off] = NULL;

	if (!objcg)
		return;

	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 bool page_has_obj_cgroups(struct page *page)
{
	return false;
}

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 struct obj_cgroup *memcg_slab_pre_alloc_hook(struct kmem_cache *s,
							   size_t objects,
							   gfp_t flags)
{
	return NULL;
}

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, struct page *page,
					void *p)
{
}
#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;

	fs_reclaim_acquire(flags);
	fs_reclaim_release(flags);

	might_sleep_if(gfpflags_allow_blocking(flags));

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

	if (memcg_kmem_enabled() &&
	    ((flags & __GFP_ACCOUNT) || (s->flags & SLAB_ACCOUNT)))
		*objcgp = memcg_slab_pre_alloc_hook(s, size, flags);

	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);
	}

	if (memcg_kmem_enabled())
		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>
286e04b8	49	#include <linux/kmemleak.h>
07f361b2	50
97d06609 CL	51	/*
	52	* State of the slab allocator.
	53	*
	54	* This is used to describe the states of the allocator during bootup.
	55	* Allocators use this to gradually bootstrap themselves. Most allocators
	56	* have the problem that the structures used for managing slab caches are
	57	* allocated from slab caches themselves.
	58	*/
	59	enum slab_state {
	60	DOWN, /* No slab functionality yet */
	61	PARTIAL, /* SLUB: kmem_cache_node available */
ce8eb6c4	62	PARTIAL_NODE, /* SLAB: kmalloc size for node struct available */
97d06609 CL	63	UP, /* Slab caches usable but not all extras yet */
	64	FULL /* Everything is working */
	65	};
	66
	67	extern enum slab_state slab_state;
	68
18004c5d CL	69	/* The slab cache mutex protects the management structures during changes */
18004c5d CL	70	extern struct mutex slab_mutex;
9b030cb8 CL	71
9b030cb8 CL	72	/* The list of all slab caches on the system */
18004c5d CL	73	extern struct list_head slab_caches;
18004c5d CL	74
9b030cb8 CL	75	/* The slab cache that manages slab cache information */
	76	extern struct kmem_cache *kmem_cache;
	77
af3b5f87 VB	78	/* A table of kmalloc cache names and sizes */
af3b5f87 VB	79	extern const struct kmalloc_info_struct {
cb5d9fb3	80	const char *name[NR_KMALLOC_TYPES];
55de8b9c	81	unsigned int size;
af3b5f87 VB	82	} kmalloc_info[];
af3b5f87 VB	83
f97d5f63 CL	84	#ifndef CONFIG_SLOB
f97d5f63 CL	85	/* Kmalloc array related functions */
34cc6990	86	void setup_kmalloc_cache_index_table(void);
d50112ed	87	void create_kmalloc_caches(slab_flags_t);
2c59dd65 CL	88
	89	/* Find the kmalloc slab corresponding for a certain size */
	90	struct kmem_cache *kmalloc_slab(size_t, gfp_t);
f97d5f63 CL	91	#endif
f97d5f63 CL	92
44405099	93	gfp_t kmalloc_fix_flags(gfp_t flags);
f97d5f63	94
9b030cb8	95	/* Functions provided by the slab allocators */
d50112ed	96	int __kmem_cache_create(struct kmem_cache *, slab_flags_t flags);
97d06609	97
55de8b9c AD	98	struct kmem_cache create_kmalloc_cache(const char name, unsigned int size,
	99	slab_flags_t flags, unsigned int useroffset,
	100	unsigned int usersize);
45530c44	101	extern void create_boot_cache(struct kmem_cache , const char name,
361d575e AD	102	unsigned int size, slab_flags_t flags,
361d575e AD	103	unsigned int useroffset, unsigned int usersize);
45530c44	104
423c929c	105	int slab_unmergeable(struct kmem_cache *s);
f4957d5b	106	struct kmem_cache *find_mergeable(unsigned size, unsigned align,
d50112ed	107	slab_flags_t flags, const char name, void (ctor)(void *));
12220dea	108	#ifndef CONFIG_SLOB
2633d7a0	109	struct kmem_cache *
f4957d5b	110	__kmem_cache_alias(const char *name, unsigned int size, unsigned int align,
d50112ed	111	slab_flags_t flags, void (ctor)(void ));
423c929c	112
0293d1fd	113	slab_flags_t kmem_cache_flags(unsigned int object_size,
d50112ed	114	slab_flags_t flags, const char *name,
423c929c	115	void (ctor)(void ));
cbb79694	116	#else
2633d7a0	117	static inline struct kmem_cache *
f4957d5b	118	__kmem_cache_alias(const char *name, unsigned int size, unsigned int align,
d50112ed	119	slab_flags_t flags, void (ctor)(void ))
cbb79694	120	{ return NULL; }
423c929c	121
0293d1fd	122	static inline slab_flags_t kmem_cache_flags(unsigned int object_size,
d50112ed	123	slab_flags_t flags, const char *name,
423c929c JK	124	void (ctor)(void ))
	125	{
	126	return flags;
	127	}
cbb79694 CL	128	#endif
	129
	130
d8843922	131	/* Legal flag mask for kmem_cache_create(), for various configurations */
6d6ea1e9 NB	132	#define SLAB_CORE_FLAGS (SLAB_HWCACHE_ALIGN \| SLAB_CACHE_DMA \| \
6d6ea1e9 NB	133	SLAB_CACHE_DMA32 \| SLAB_PANIC \| \
5f0d5a3a	134	SLAB_TYPESAFE_BY_RCU \| SLAB_DEBUG_OBJECTS )
d8843922 GC	135
	136	#if defined(CONFIG_DEBUG_SLAB)
	137	#define SLAB_DEBUG_FLAGS (SLAB_RED_ZONE \| SLAB_POISON \| SLAB_STORE_USER)
	138	#elif defined(CONFIG_SLUB_DEBUG)
	139	#define SLAB_DEBUG_FLAGS (SLAB_RED_ZONE \| SLAB_POISON \| SLAB_STORE_USER \| \
becfda68	140	SLAB_TRACE \| SLAB_CONSISTENCY_CHECKS)
d8843922 GC	141	#else
	142	#define SLAB_DEBUG_FLAGS (0)
	143	#endif
	144
	145	#if defined(CONFIG_SLAB)
	146	#define SLAB_CACHE_FLAGS (SLAB_MEM_SPREAD \| SLAB_NOLEAKTRACE \| \
230e9fc2	147	SLAB_RECLAIM_ACCOUNT \| SLAB_TEMPORARY \| \
75f296d9	148	SLAB_ACCOUNT)
d8843922 GC	149	#elif defined(CONFIG_SLUB)
d8843922 GC	150	#define SLAB_CACHE_FLAGS (SLAB_NOLEAKTRACE \| SLAB_RECLAIM_ACCOUNT \| \
75f296d9	151	SLAB_TEMPORARY \| SLAB_ACCOUNT)
d8843922 GC	152	#else
	153	#define SLAB_CACHE_FLAGS (0)
	154	#endif
	155
e70954fd	156	/* Common flags available with current configuration */
d8843922 GC	157	#define CACHE_CREATE_MASK (SLAB_CORE_FLAGS \| SLAB_DEBUG_FLAGS \| SLAB_CACHE_FLAGS)
d8843922 GC	158
e70954fd TG	159	/* Common flags permitted for kmem_cache_create */
	160	#define SLAB_FLAGS_PERMITTED (SLAB_CORE_FLAGS \| \
	161	SLAB_RED_ZONE \| \
	162	SLAB_POISON \| \
	163	SLAB_STORE_USER \| \
	164	SLAB_TRACE \| \
	165	SLAB_CONSISTENCY_CHECKS \| \
	166	SLAB_MEM_SPREAD \| \
	167	SLAB_NOLEAKTRACE \| \
	168	SLAB_RECLAIM_ACCOUNT \| \
	169	SLAB_TEMPORARY \| \
e70954fd TG	170	SLAB_ACCOUNT)
e70954fd TG	171
f9e13c0a	172	bool __kmem_cache_empty(struct kmem_cache *);
945cf2b6	173	int __kmem_cache_shutdown(struct kmem_cache *);
52b4b950	174	void __kmem_cache_release(struct kmem_cache *);
c9fc5864	175	int __kmem_cache_shrink(struct kmem_cache *);
41a21285	176	void slab_kmem_cache_release(struct kmem_cache *);
945cf2b6	177
b7454ad3 GC	178	struct seq_file;
b7454ad3 GC	179	struct file;
b7454ad3	180
0d7561c6 GC	181	struct slabinfo {
	182	unsigned long active_objs;
	183	unsigned long num_objs;
	184	unsigned long active_slabs;
	185	unsigned long num_slabs;
	186	unsigned long shared_avail;
	187	unsigned int limit;
	188	unsigned int batchcount;
	189	unsigned int shared;
	190	unsigned int objects_per_slab;
	191	unsigned int cache_order;
	192	};
	193
	194	void get_slabinfo(struct kmem_cache s, struct slabinfo sinfo);
	195	void slabinfo_show_stats(struct seq_file m, struct kmem_cache s);
b7454ad3 GC	196	ssize_t slabinfo_write(struct file file, const char __user buffer,
b7454ad3 GC	197	size_t count, loff_t *ppos);
ba6c496e	198
484748f0 CL	199	/*
	200	* Generic implementation of bulk operations
	201	* These are useful for situations in which the allocator cannot
9f706d68	202	* perform optimizations. In that case segments of the object listed
484748f0 CL	203	* may be allocated or freed using these operations.
	204	*/
	205	void __kmem_cache_free_bulk(struct kmem_cache , size_t, void *);
865762a8	206	int __kmem_cache_alloc_bulk(struct kmem_cache , gfp_t, size_t, void *);
484748f0	207
6cea1d56 RG	208	static inline int cache_vmstat_idx(struct kmem_cache *s)
	209	{
	210	return (s->flags & SLAB_RECLAIM_ACCOUNT) ?
d42f3245	211	NR_SLAB_RECLAIMABLE_B : NR_SLAB_UNRECLAIMABLE_B;
6cea1d56 RG	212	}
6cea1d56 RG	213
e42f174e VB	214	#ifdef CONFIG_SLUB_DEBUG
	215	#ifdef CONFIG_SLUB_DEBUG_ON
	216	DECLARE_STATIC_KEY_TRUE(slub_debug_enabled);
	217	#else
	218	DECLARE_STATIC_KEY_FALSE(slub_debug_enabled);
	219	#endif
	220	extern void print_tracking(struct kmem_cache s, void object);
	221	#else
	222	static inline void print_tracking(struct kmem_cache s, void object)
	223	{
	224	}
	225	#endif
	226
	227	/*
	228	* Returns true if any of the specified slub_debug flags is enabled for the
	229	* cache. Use only for flags parsed by setup_slub_debug() as it also enables
	230	* the static key.
	231	*/
	232	static inline bool kmem_cache_debug_flags(struct kmem_cache *s, slab_flags_t flags)
	233	{
	234	#ifdef CONFIG_SLUB_DEBUG
	235	VM_WARN_ON_ONCE(!(flags & SLAB_DEBUG_FLAGS));
	236	if (static_branch_unlikely(&slub_debug_enabled))
	237	return s->flags & flags;
	238	#endif
	239	return false;
	240	}
	241
84c07d11	242	#ifdef CONFIG_MEMCG_KMEM
286e04b8 RG	243	static inline struct obj_cgroup *page_obj_cgroups(struct page page)
	244	{
	245	/*
	246	* page->mem_cgroup and page->obj_cgroups are sharing the same
	247	* space. To distinguish between them in case we don't know for sure
	248	* that the page is a slab page (e.g. page_cgroup_ino()), let's
	249	* always set the lowest bit of obj_cgroups.
	250	*/
	251	return (struct obj_cgroup **)
	252	((unsigned long)page->obj_cgroups & ~0x1UL);
	253	}
	254
9855609b	255	static inline bool page_has_obj_cgroups(struct page *page)
4d96ba35	256	{
9855609b	257	return ((unsigned long)page->obj_cgroups & 0x1UL);
4d96ba35 RG	258	}
4d96ba35 RG	259
10befea9 RG	260	int memcg_alloc_page_obj_cgroups(struct page page, struct kmem_cache s,
10befea9 RG	261	gfp_t gfp);
286e04b8 RG	262
	263	static inline void memcg_free_page_obj_cgroups(struct page *page)
	264	{
	265	kfree(page_obj_cgroups(page));
	266	page->obj_cgroups = NULL;
	267	}
	268
f2fe7b09 RG	269	static inline size_t obj_full_size(struct kmem_cache *s)
	270	{
	271	/*
	272	* For each accounted object there is an extra space which is used
	273	* to store obj_cgroup membership. Charge it too.
	274	*/
	275	return s->size + sizeof(struct obj_cgroup *);
	276	}
	277
10befea9 RG	278	static inline struct obj_cgroup memcg_slab_pre_alloc_hook(struct kmem_cache s,
	279	size_t objects,
	280	gfp_t flags)
f2fe7b09	281	{
9855609b RG	282	struct obj_cgroup *objcg;
	283
	284	if (memcg_kmem_bypass())
10befea9	285	return NULL;
f2fe7b09	286
9855609b RG	287	objcg = get_obj_cgroup_from_current();
9855609b RG	288	if (!objcg)
10befea9	289	return NULL;
9855609b RG	290
	291	if (obj_cgroup_charge(objcg, flags, objects * obj_full_size(s))) {
	292	obj_cgroup_put(objcg);
10befea9	293	return NULL;
f2fe7b09 RG	294	}
f2fe7b09 RG	295
10befea9	296	return objcg;
f2fe7b09 RG	297	}
	298
	299	static inline void mod_objcg_state(struct obj_cgroup *objcg,
	300	struct pglist_data *pgdat,
	301	int idx, int nr)
	302	{
	303	struct mem_cgroup *memcg;
	304	struct lruvec *lruvec;
	305
	306	rcu_read_lock();
	307	memcg = obj_cgroup_memcg(objcg);
	308	lruvec = mem_cgroup_lruvec(memcg, pgdat);
	309	mod_memcg_lruvec_state(lruvec, idx, nr);
	310	rcu_read_unlock();
	311	}
	312
964d4bd3 RG	313	static inline void memcg_slab_post_alloc_hook(struct kmem_cache *s,
964d4bd3 RG	314	struct obj_cgroup *objcg,
10befea9 RG	315	gfp_t flags, size_t size,
10befea9 RG	316	void **p)
964d4bd3 RG	317	{
	318	struct page *page;
	319	unsigned long off;
	320	size_t i;
	321
10befea9 RG	322	if (!objcg)
	323	return;
	324
	325	flags &= ~__GFP_ACCOUNT;
964d4bd3 RG	326	for (i = 0; i < size; i++) {
	327	if (likely(p[i])) {
	328	page = virt_to_head_page(p[i]);
10befea9 RG	329
	330	if (!page_has_obj_cgroups(page) &&
	331	memcg_alloc_page_obj_cgroups(page, s, flags)) {
	332	obj_cgroup_uncharge(objcg, obj_full_size(s));
	333	continue;
	334	}
	335
964d4bd3 RG	336	off = obj_to_index(s, page, p[i]);
	337	obj_cgroup_get(objcg);
	338	page_obj_cgroups(page)[off] = objcg;
f2fe7b09 RG	339	mod_objcg_state(objcg, page_pgdat(page),
	340	cache_vmstat_idx(s), obj_full_size(s));
	341	} else {
	342	obj_cgroup_uncharge(objcg, obj_full_size(s));
964d4bd3 RG	343	}
	344	}
	345	obj_cgroup_put(objcg);
964d4bd3 RG	346	}
	347
	348	static inline void memcg_slab_free_hook(struct kmem_cache s, struct page page,
	349	void *p)
	350	{
	351	struct obj_cgroup *objcg;
	352	unsigned int off;
	353
10befea9 RG	354	if (!memcg_kmem_enabled())
	355	return;
	356
	357	if (!page_has_obj_cgroups(page))
964d4bd3 RG	358	return;
	359
	360	off = obj_to_index(s, page, p);
	361	objcg = page_obj_cgroups(page)[off];
	362	page_obj_cgroups(page)[off] = NULL;
f2fe7b09	363
10befea9 RG	364	if (!objcg)
	365	return;
	366
f2fe7b09 RG	367	obj_cgroup_uncharge(objcg, obj_full_size(s));
	368	mod_objcg_state(objcg, page_pgdat(page), cache_vmstat_idx(s),
	369	-obj_full_size(s));
	370
964d4bd3 RG	371	obj_cgroup_put(objcg);
	372	}
	373
84c07d11	374	#else /* CONFIG_MEMCG_KMEM */
9855609b RG	375	static inline bool page_has_obj_cgroups(struct page *page)
	376	{
	377	return false;
	378	}
	379
	380	static inline struct mem_cgroup memcg_from_slab_obj(void ptr)
4d96ba35 RG	381	{
	382	return NULL;
	383	}
	384
286e04b8 RG	385	static inline int memcg_alloc_page_obj_cgroups(struct page *page,
	386	struct kmem_cache *s, gfp_t gfp)
	387	{
	388	return 0;
	389	}
	390
	391	static inline void memcg_free_page_obj_cgroups(struct page *page)
	392	{
	393	}
	394
10befea9 RG	395	static inline struct obj_cgroup memcg_slab_pre_alloc_hook(struct kmem_cache s,
	396	size_t objects,
	397	gfp_t flags)
f2fe7b09 RG	398	{
	399	return NULL;
	400	}
	401
964d4bd3 RG	402	static inline void memcg_slab_post_alloc_hook(struct kmem_cache *s,
964d4bd3 RG	403	struct obj_cgroup *objcg,
10befea9 RG	404	gfp_t flags, size_t size,
10befea9 RG	405	void **p)
964d4bd3 RG	406	{
	407	}
	408
	409	static inline void memcg_slab_free_hook(struct kmem_cache s, struct page page,
	410	void *p)
	411	{
	412	}
84c07d11	413	#endif /* CONFIG_MEMCG_KMEM */
b9ce5ef4	414
a64b5378 KC	415	static inline struct kmem_cache virt_to_cache(const void obj)
	416	{
	417	struct page *page;
	418
	419	page = virt_to_head_page(obj);
	420	if (WARN_ONCE(!PageSlab(page), "%s: Object is not a Slab page!\n",
	421	__func__))
	422	return NULL;
	423	return page->slab_cache;
	424	}
	425
74d555be RG	426	static __always_inline void account_slab_page(struct page *page, int order,
74d555be RG	427	struct kmem_cache *s)
6cea1d56	428	{
f2fe7b09 RG	429	mod_node_page_state(page_pgdat(page), cache_vmstat_idx(s),
f2fe7b09 RG	430	PAGE_SIZE << order);
6cea1d56 RG	431	}
6cea1d56 RG	432
74d555be RG	433	static __always_inline void unaccount_slab_page(struct page *page, int order,
74d555be RG	434	struct kmem_cache *s)
6cea1d56	435	{
10befea9	436	if (memcg_kmem_enabled())
f2fe7b09	437	memcg_free_page_obj_cgroups(page);
9855609b	438
f2fe7b09 RG	439	mod_node_page_state(page_pgdat(page), cache_vmstat_idx(s),
f2fe7b09 RG	440	-(PAGE_SIZE << order));
6cea1d56 RG	441	}
6cea1d56 RG	442
e42f174e VB	443	static inline struct kmem_cache cache_from_obj(struct kmem_cache s, void *x)
	444	{
	445	struct kmem_cache *cachep;
	446
	447	if (!IS_ENABLED(CONFIG_SLAB_FREELIST_HARDENED) &&
e42f174e VB	448	!kmem_cache_debug_flags(s, SLAB_CONSISTENCY_CHECKS))
	449	return s;
	450
	451	cachep = virt_to_cache(x);
10befea9	452	if (WARN(cachep && cachep != s,
e42f174e VB	453	"%s: Wrong slab cache. %s but object is from %s\n",
	454	__func__, s->name, cachep->name))
	455	print_tracking(cachep, x);
	456	return cachep;
	457	}
	458
11c7aec2 JDB	459	static inline size_t slab_ksize(const struct kmem_cache *s)
	460	{
	461	#ifndef CONFIG_SLUB
	462	return s->object_size;
	463
	464	#else /* CONFIG_SLUB */
	465	# ifdef CONFIG_SLUB_DEBUG
	466	/*
	467	* Debugging requires use of the padding between object
	468	* and whatever may come after it.
	469	*/
	470	if (s->flags & (SLAB_RED_ZONE \| SLAB_POISON))
	471	return s->object_size;
	472	# endif
80a9201a AP	473	if (s->flags & SLAB_KASAN)
80a9201a AP	474	return s->object_size;
11c7aec2 JDB	475	/*
	476	* If we have the need to store the freelist pointer
	477	* back there or track user information then we can
	478	* only use the space before that information.
	479	*/
5f0d5a3a	480	if (s->flags & (SLAB_TYPESAFE_BY_RCU \| SLAB_STORE_USER))
11c7aec2 JDB	481	return s->inuse;
	482	/*
	483	* Else we can use all the padding etc for the allocation
	484	*/
	485	return s->size;
	486	#endif
	487	}
	488
	489	static inline struct kmem_cache slab_pre_alloc_hook(struct kmem_cache s,
964d4bd3 RG	490	struct obj_cgroup **objcgp,
964d4bd3 RG	491	size_t size, gfp_t flags)
11c7aec2 JDB	492	{
11c7aec2 JDB	493	flags &= gfp_allowed_mask;
d92a8cfc PZ	494
	495	fs_reclaim_acquire(flags);
	496	fs_reclaim_release(flags);
	497
11c7aec2 JDB	498	might_sleep_if(gfpflags_allow_blocking(flags));
11c7aec2 JDB	499
fab9963a	500	if (should_failslab(s, flags))
11c7aec2 JDB	501	return NULL;
11c7aec2 JDB	502
45264778 VD	503	if (memcg_kmem_enabled() &&
45264778 VD	504	((flags & __GFP_ACCOUNT) \|\| (s->flags & SLAB_ACCOUNT)))
10befea9	505	*objcgp = memcg_slab_pre_alloc_hook(s, size, flags);
45264778 VD	506
45264778 VD	507	return s;
11c7aec2 JDB	508	}
11c7aec2 JDB	509
964d4bd3 RG	510	static inline void slab_post_alloc_hook(struct kmem_cache *s,
	511	struct obj_cgroup *objcg,
	512	gfp_t flags, size_t size, void **p)
11c7aec2 JDB	513	{
	514	size_t i;
	515
	516	flags &= gfp_allowed_mask;
	517	for (i = 0; i < size; i++) {
53128245	518	p[i] = kasan_slab_alloc(s, p[i], flags);
a2f77575	519	/* As p[i] might get tagged, call kmemleak hook after KASAN. */
53128245	520	kmemleak_alloc_recursive(p[i], s->object_size, 1,
11c7aec2	521	s->flags, flags);
11c7aec2	522	}
45264778	523
10befea9 RG	524	if (memcg_kmem_enabled())
10befea9 RG	525	memcg_slab_post_alloc_hook(s, objcg, flags, size, p);
11c7aec2 JDB	526	}
11c7aec2 JDB	527
44c5356f	528	#ifndef CONFIG_SLOB
ca34956b CL	529	/*
	530	* The slab lists for all objects.
	531	*/
	532	struct kmem_cache_node {
	533	spinlock_t list_lock;
	534
	535	#ifdef CONFIG_SLAB
	536	struct list_head slabs_partial; /* partial list first, better asm code */
	537	struct list_head slabs_full;
	538	struct list_head slabs_free;
bf00bd34 DR	539	unsigned long total_slabs; /* length of all slab lists */
bf00bd34 DR	540	unsigned long free_slabs; /* length of free slab list only */
ca34956b CL	541	unsigned long free_objects;
	542	unsigned int free_limit;
	543	unsigned int colour_next; /* Per-node cache coloring */
	544	struct array_cache shared; / shared per node */
c8522a3a	545	struct alien_cache *alien; / on other nodes */
ca34956b CL	546	unsigned long next_reap; /* updated without locking */
	547	int free_touched; /* updated without locking */
	548	#endif
	549
	550	#ifdef CONFIG_SLUB
	551	unsigned long nr_partial;
	552	struct list_head partial;
	553	#ifdef CONFIG_SLUB_DEBUG
	554	atomic_long_t nr_slabs;
	555	atomic_long_t total_objects;
	556	struct list_head full;
	557	#endif
	558	#endif
	559
	560	};
e25839f6	561
44c5356f CL	562	static inline struct kmem_cache_node get_node(struct kmem_cache s, int node)
	563	{
	564	return s->node[node];
	565	}
	566
	567	/*
	568	* Iterator over all nodes. The body will be executed for each node that has
	569	* a kmem_cache_node structure allocated (which is true for all online nodes)
	570	*/
	571	#define for_each_kmem_cache_node(__s, __node, __n) \
9163582c MP	572	for (__node = 0; __node < nr_node_ids; __node++) \
9163582c MP	573	if ((__n = get_node(__s, __node)))
44c5356f CL	574
	575	#endif
	576
1df3b26f	577	void slab_start(struct seq_file m, loff_t *pos);
276a2439 WL	578	void slab_next(struct seq_file m, void p, loff_t pos);
276a2439 WL	579	void slab_stop(struct seq_file m, void p);
b047501c	580	int memcg_slab_show(struct seq_file m, void p);
5240ab40	581
852d8be0 YS	582	#if defined(CONFIG_SLAB) \|\| defined(CONFIG_SLUB_DEBUG)
	583	void dump_unreclaimable_slab(void);
	584	#else
	585	static inline void dump_unreclaimable_slab(void)
	586	{
	587	}
	588	#endif
	589
55834c59 AP	590	void ___cache_free(struct kmem_cache cache, void x, unsigned long addr);
55834c59 AP	591
7c00fce9 TG	592	#ifdef CONFIG_SLAB_FREELIST_RANDOM
	593	int cache_random_seq_create(struct kmem_cache *cachep, unsigned int count,
	594	gfp_t gfp);
	595	void cache_random_seq_destroy(struct kmem_cache *cachep);
	596	#else
	597	static inline int cache_random_seq_create(struct kmem_cache *cachep,
	598	unsigned int count, gfp_t gfp)
	599	{
	600	return 0;
	601	}
	602	static inline void cache_random_seq_destroy(struct kmem_cache *cachep) { }
	603	#endif /* CONFIG_SLAB_FREELIST_RANDOM */
	604
6471384a AP	605	static inline bool slab_want_init_on_alloc(gfp_t flags, struct kmem_cache *c)
	606	{
	607	if (static_branch_unlikely(&init_on_alloc)) {
	608	if (c->ctor)
	609	return false;
	610	if (c->flags & (SLAB_TYPESAFE_BY_RCU \| SLAB_POISON))
	611	return flags & __GFP_ZERO;
	612	return true;
	613	}
	614	return flags & __GFP_ZERO;
	615	}
	616
	617	static inline bool slab_want_init_on_free(struct kmem_cache *c)
	618	{
	619	if (static_branch_unlikely(&init_on_free))
	620	return !(c->ctor \|\|
	621	(c->flags & (SLAB_TYPESAFE_BY_RCU \| SLAB_POISON)));
	622	return false;
	623	}
	624
5240ab40	625	#endif /* MM_SLAB_H */