[linux-2.6-block.git] / include / linux / slub_def.h

#ifndef _LINUX_SLUB_DEF_H
#define _LINUX_SLUB_DEF_H

/*
 * SLUB : A Slab allocator without object queues.
 *
 * (C) 2007 SGI, Christoph Lameter
 */
#include <linux/types.h>
#include <linux/gfp.h>
#include <linux/workqueue.h>
#include <linux/kobject.h>
#include <linux/kmemtrace.h>
#include <linux/kmemleak.h>

enum stat_item {
	ALLOC_FASTPATH,		/* Allocation from cpu slab */
	ALLOC_SLOWPATH,		/* Allocation by getting a new cpu slab */
	FREE_FASTPATH,		/* Free to cpu slub */
	FREE_SLOWPATH,		/* Freeing not to cpu slab */
	FREE_FROZEN,		/* Freeing to frozen slab */
	FREE_ADD_PARTIAL,	/* Freeing moves slab to partial list */
	FREE_REMOVE_PARTIAL,	/* Freeing removes last object */
	ALLOC_FROM_PARTIAL,	/* Cpu slab acquired from partial list */
	ALLOC_SLAB,		/* Cpu slab acquired from page allocator */
	ALLOC_REFILL,		/* Refill cpu slab from slab freelist */
	FREE_SLAB,		/* Slab freed to the page allocator */
	CPUSLAB_FLUSH,		/* Abandoning of the cpu slab */
	DEACTIVATE_FULL,	/* Cpu slab was full when deactivated */
	DEACTIVATE_EMPTY,	/* Cpu slab was empty when deactivated */
	DEACTIVATE_TO_HEAD,	/* Cpu slab was moved to the head of partials */
	DEACTIVATE_TO_TAIL,	/* Cpu slab was moved to the tail of partials */
	DEACTIVATE_REMOTE_FREES,/* Slab contained remotely freed objects */
	ORDER_FALLBACK,		/* Number of times fallback was necessary */
	NR_SLUB_STAT_ITEMS };

struct kmem_cache_cpu {
	void **freelist;	/* Pointer to first free per cpu object */
	struct page *page;	/* The slab from which we are allocating */
	int node;		/* The node of the page (or -1 for debug) */
#ifdef CONFIG_SLUB_STATS
	unsigned stat[NR_SLUB_STAT_ITEMS];
#endif
};

struct kmem_cache_node {
	spinlock_t list_lock;	/* Protect partial list and nr_partial */
	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
};

/*
 * Word size structure that can be atomically updated or read and that
 * contains both the order and the number of objects that a slab of the
 * given order would contain.
 */
struct kmem_cache_order_objects {
	unsigned long x;
};

/*
 * Slab cache management.
 */
struct kmem_cache {
	struct kmem_cache_cpu *cpu_slab;
	/* Used for retriving partial slabs etc */
	unsigned long flags;
	int size;		/* The size of an object including meta data */
	int objsize;		/* The size of an object without meta data */
	int offset;		/* Free pointer offset. */
	struct kmem_cache_order_objects oo;

	/*
	 * Avoid an extra cache line for UP, SMP and for the node local to
	 * struct kmem_cache.
	 */
	struct kmem_cache_node local_node;

	/* Allocation and freeing of slabs */
	struct kmem_cache_order_objects max;
	struct kmem_cache_order_objects min;
	gfp_t allocflags;	/* gfp flags to use on each alloc */
	int refcount;		/* Refcount for slab cache destroy */
	void (*ctor)(void *);
	int inuse;		/* Offset to metadata */
	int align;		/* Alignment */
	unsigned long min_partial;
	const char *name;	/* Name (only for display!) */
	struct list_head list;	/* List of slab caches */
#ifdef CONFIG_SLUB_DEBUG
	struct kobject kobj;	/* For sysfs */
#endif

#ifdef CONFIG_NUMA
	/*
	 * Defragmentation by allocating from a remote node.
	 */
	int remote_node_defrag_ratio;
	struct kmem_cache_node *node[MAX_NUMNODES];
#endif
};

/*
 * Kmalloc subsystem.
 */
#if defined(ARCH_KMALLOC_MINALIGN) && ARCH_KMALLOC_MINALIGN > 8
#define KMALLOC_MIN_SIZE ARCH_KMALLOC_MINALIGN
#else
#define KMALLOC_MIN_SIZE 8
#endif

#define KMALLOC_SHIFT_LOW ilog2(KMALLOC_MIN_SIZE)

/*
 * Maximum kmalloc object size handled by SLUB. Larger object allocations
 * are passed through to the page allocator. The page allocator "fastpath"
 * is relatively slow so we need this value sufficiently high so that
 * performance critical objects are allocated through the SLUB fastpath.
 *
 * This should be dropped to PAGE_SIZE / 2 once the page allocator
 * "fastpath" becomes competitive with the slab allocator fastpaths.
 */
#define SLUB_MAX_SIZE (2 * PAGE_SIZE)

#define SLUB_PAGE_SHIFT (PAGE_SHIFT + 2)

#ifdef CONFIG_ZONE_DMA
#define SLUB_DMA __GFP_DMA
/* Reserve extra caches for potential DMA use */
#define KMALLOC_CACHES (2 * SLUB_PAGE_SHIFT - 6)
#else
/* Disable DMA functionality */
#define SLUB_DMA (__force gfp_t)0
#define KMALLOC_CACHES SLUB_PAGE_SHIFT
#endif

/*
 * We keep the general caches in an array of slab caches that are used for
 * 2^x bytes of allocations.
 */
extern struct kmem_cache kmalloc_caches[KMALLOC_CACHES];

/*
 * Sorry that the following has to be that ugly but some versions of GCC
 * have trouble with constant propagation and loops.
 */
static __always_inline int kmalloc_index(size_t size)
{
	if (!size)
		return 0;

	if (size <= KMALLOC_MIN_SIZE)
		return KMALLOC_SHIFT_LOW;

	if (KMALLOC_MIN_SIZE <= 32 && size > 64 && size <= 96)
		return 1;
	if (KMALLOC_MIN_SIZE <= 64 && size > 128 && size <= 192)
		return 2;
	if (size <=          8) return 3;
	if (size <=         16) return 4;
	if (size <=         32) return 5;
	if (size <=         64) return 6;
	if (size <=        128) return 7;
	if (size <=        256) return 8;
	if (size <=        512) return 9;
	if (size <=       1024) return 10;
	if (size <=   2 * 1024) return 11;
	if (size <=   4 * 1024) return 12;
/*
 * The following is only needed to support architectures with a larger page
 * size than 4k.
 */
	if (size <=   8 * 1024) return 13;
	if (size <=  16 * 1024) return 14;
	if (size <=  32 * 1024) return 15;
	if (size <=  64 * 1024) return 16;
	if (size <= 128 * 1024) return 17;
	if (size <= 256 * 1024) return 18;
	if (size <= 512 * 1024) return 19;
	if (size <= 1024 * 1024) return 20;
	if (size <=  2 * 1024 * 1024) return 21;
	return -1;

/*
 * What we really wanted to do and cannot do because of compiler issues is:
 *	int i;
 *	for (i = KMALLOC_SHIFT_LOW; i <= KMALLOC_SHIFT_HIGH; i++)
 *		if (size <= (1 << i))
 *			return i;
 */
}

/*
 * Find the slab cache for a given combination of allocation flags and size.
 *
 * This ought to end up with a global pointer to the right cache
 * in kmalloc_caches.
 */
static __always_inline struct kmem_cache *kmalloc_slab(size_t size)
{
	int index = kmalloc_index(size);

	if (index == 0)
		return NULL;

	return &kmalloc_caches[index];
}

void *kmem_cache_alloc(struct kmem_cache *, gfp_t);
void *__kmalloc(size_t size, gfp_t flags);

#ifdef CONFIG_TRACING
extern void *kmem_cache_alloc_notrace(struct kmem_cache *s, gfp_t gfpflags);
#else
static __always_inline void *
kmem_cache_alloc_notrace(struct kmem_cache *s, gfp_t gfpflags)
{
	return kmem_cache_alloc(s, gfpflags);
}
#endif

static __always_inline void *kmalloc_large(size_t size, gfp_t flags)
{
	unsigned int order = get_order(size);
	void *ret = (void *) __get_free_pages(flags | __GFP_COMP, order);

	kmemleak_alloc(ret, size, 1, flags);
	trace_kmalloc(_THIS_IP_, ret, size, PAGE_SIZE << order, flags);

	return ret;
}

static __always_inline void *kmalloc(size_t size, gfp_t flags)
{
	void *ret;

	if (__builtin_constant_p(size)) {
		if (size > SLUB_MAX_SIZE)
			return kmalloc_large(size, flags);

		if (!(flags & SLUB_DMA)) {
			struct kmem_cache *s = kmalloc_slab(size);

			if (!s)
				return ZERO_SIZE_PTR;

			ret = kmem_cache_alloc_notrace(s, flags);

			trace_kmalloc(_THIS_IP_, ret, size, s->size, flags);

			return ret;
		}
	}
	return __kmalloc(size, flags);
}

#ifdef CONFIG_NUMA
void *__kmalloc_node(size_t size, gfp_t flags, int node);
void *kmem_cache_alloc_node(struct kmem_cache *, gfp_t flags, int node);

#ifdef CONFIG_TRACING
extern void *kmem_cache_alloc_node_notrace(struct kmem_cache *s,
					   gfp_t gfpflags,
					   int node);
#else
static __always_inline void *
kmem_cache_alloc_node_notrace(struct kmem_cache *s,
			      gfp_t gfpflags,
			      int node)
{
	return kmem_cache_alloc_node(s, gfpflags, node);
}
#endif

static __always_inline void *kmalloc_node(size_t size, gfp_t flags, int node)
{
	void *ret;

	if (__builtin_constant_p(size) &&
		size <= SLUB_MAX_SIZE && !(flags & SLUB_DMA)) {
			struct kmem_cache *s = kmalloc_slab(size);

		if (!s)
			return ZERO_SIZE_PTR;

		ret = kmem_cache_alloc_node_notrace(s, flags, node);

		trace_kmalloc_node(_THIS_IP_, ret,
				   size, s->size, flags, node);

		return ret;
	}
	return __kmalloc_node(size, flags, node);
}
#endif

#endif /* _LINUX_SLUB_DEF_H */
Commit	Line	Data
81819f0f CL	1	#ifndef _LINUX_SLUB_DEF_H
	2	#define _LINUX_SLUB_DEF_H
	3
	4	/*
	5	* SLUB : A Slab allocator without object queues.
	6	*
cde53535	7	* (C) 2007 SGI, Christoph Lameter
81819f0f CL	8	*/
	9	#include <linux/types.h>
	10	#include <linux/gfp.h>
	11	#include <linux/workqueue.h>
	12	#include <linux/kobject.h>
02af61bb	13	#include <linux/kmemtrace.h>
e4f7c0b4	14	#include <linux/kmemleak.h>
81819f0f	15
8ff12cfc CL	16	enum stat_item {
	17	ALLOC_FASTPATH, /* Allocation from cpu slab */
	18	ALLOC_SLOWPATH, /* Allocation by getting a new cpu slab */
	19	FREE_FASTPATH, /* Free to cpu slub */
	20	FREE_SLOWPATH, /* Freeing not to cpu slab */
	21	FREE_FROZEN, /* Freeing to frozen slab */
	22	FREE_ADD_PARTIAL, /* Freeing moves slab to partial list */
	23	FREE_REMOVE_PARTIAL, /* Freeing removes last object */
	24	ALLOC_FROM_PARTIAL, /* Cpu slab acquired from partial list */
	25	ALLOC_SLAB, /* Cpu slab acquired from page allocator */
	26	ALLOC_REFILL, /* Refill cpu slab from slab freelist */
	27	FREE_SLAB, /* Slab freed to the page allocator */
	28	CPUSLAB_FLUSH, /* Abandoning of the cpu slab */
	29	DEACTIVATE_FULL, /* Cpu slab was full when deactivated */
	30	DEACTIVATE_EMPTY, /* Cpu slab was empty when deactivated */
	31	DEACTIVATE_TO_HEAD, /* Cpu slab was moved to the head of partials */
	32	DEACTIVATE_TO_TAIL, /* Cpu slab was moved to the tail of partials */
	33	DEACTIVATE_REMOTE_FREES,/* Slab contained remotely freed objects */
65c3376a	34	ORDER_FALLBACK, /* Number of times fallback was necessary */
8ff12cfc CL	35	NR_SLUB_STAT_ITEMS };
8ff12cfc CL	36
dfb4f096	37	struct kmem_cache_cpu {
da89b79e CL	38	void *freelist; / Pointer to first free per cpu object */
	39	struct page page; / The slab from which we are allocating */
	40	int node; /* The node of the page (or -1 for debug) */
8ff12cfc CL	41	#ifdef CONFIG_SLUB_STATS
	42	unsigned stat[NR_SLUB_STAT_ITEMS];
	43	#endif
4c93c355	44	};
dfb4f096	45
81819f0f CL	46	struct kmem_cache_node {
	47	spinlock_t list_lock; /* Protect partial list and nr_partial */
	48	unsigned long nr_partial;
81819f0f	49	struct list_head partial;
0c710013	50	#ifdef CONFIG_SLUB_DEBUG
0f389ec6	51	atomic_long_t nr_slabs;
205ab99d	52	atomic_long_t total_objects;
643b1138	53	struct list_head full;
0c710013	54	#endif
81819f0f CL	55	};
81819f0f CL	56
834f3d11 CL	57	/*
	58	* Word size structure that can be atomically updated or read and that
	59	* contains both the order and the number of objects that a slab of the
	60	* given order would contain.
	61	*/
	62	struct kmem_cache_order_objects {
	63	unsigned long x;
	64	};
	65
81819f0f CL	66	/*
	67	* Slab cache management.
	68	*/
	69	struct kmem_cache {
9dfc6e68	70	struct kmem_cache_cpu *cpu_slab;
81819f0f CL	71	/* Used for retriving partial slabs etc */
	72	unsigned long flags;
	73	int size; /* The size of an object including meta data */
	74	int objsize; /* The size of an object without meta data */
	75	int offset; /* Free pointer offset. */
834f3d11	76	struct kmem_cache_order_objects oo;
81819f0f CL	77
	78	/*
	79	* Avoid an extra cache line for UP, SMP and for the node local to
	80	* struct kmem_cache.
	81	*/
	82	struct kmem_cache_node local_node;
	83
	84	/* Allocation and freeing of slabs */
205ab99d	85	struct kmem_cache_order_objects max;
65c3376a	86	struct kmem_cache_order_objects min;
b7a49f0d	87	gfp_t allocflags; /* gfp flags to use on each alloc */
81819f0f	88	int refcount; /* Refcount for slab cache destroy */
51cc5068	89	void (ctor)(void );
81819f0f CL	90	int inuse; /* Offset to metadata */
81819f0f CL	91	int align; /* Alignment */
3b89d7d8	92	unsigned long min_partial;
81819f0f CL	93	const char name; / Name (only for display!) */
81819f0f CL	94	struct list_head list; /* List of slab caches */
0c710013	95	#ifdef CONFIG_SLUB_DEBUG
81819f0f	96	struct kobject kobj; /* For sysfs */
0c710013	97	#endif
81819f0f CL	98
81819f0f CL	99	#ifdef CONFIG_NUMA
9824601e CL	100	/*
	101	* Defragmentation by allocating from a remote node.
	102	*/
	103	int remote_node_defrag_ratio;
81819f0f CL	104	struct kmem_cache_node *node[MAX_NUMNODES];
81819f0f CL	105	#endif
81819f0f CL	106	};
	107
	108	/*
	109	* Kmalloc subsystem.
	110	*/
4b356be0 CL	111	#if defined(ARCH_KMALLOC_MINALIGN) && ARCH_KMALLOC_MINALIGN > 8
	112	#define KMALLOC_MIN_SIZE ARCH_KMALLOC_MINALIGN
	113	#else
	114	#define KMALLOC_MIN_SIZE 8
	115	#endif
	116
	117	#define KMALLOC_SHIFT_LOW ilog2(KMALLOC_MIN_SIZE)
81819f0f	118
ffadd4d0 CL	119	/*
	120	* Maximum kmalloc object size handled by SLUB. Larger object allocations
	121	* are passed through to the page allocator. The page allocator "fastpath"
	122	* is relatively slow so we need this value sufficiently high so that
	123	* performance critical objects are allocated through the SLUB fastpath.
	124	*
	125	* This should be dropped to PAGE_SIZE / 2 once the page allocator
	126	* "fastpath" becomes competitive with the slab allocator fastpaths.
	127	*/
51735a7c	128	#define SLUB_MAX_SIZE (2 * PAGE_SIZE)
ffadd4d0	129
51735a7c	130	#define SLUB_PAGE_SHIFT (PAGE_SHIFT + 2)
ffadd4d0	131
756dee75 CL	132	#ifdef CONFIG_ZONE_DMA
	133	#define SLUB_DMA __GFP_DMA
	134	/* Reserve extra caches for potential DMA use */
	135	#define KMALLOC_CACHES (2 * SLUB_PAGE_SHIFT - 6)
	136	#else
	137	/* Disable DMA functionality */
	138	#define SLUB_DMA (__force gfp_t)0
	139	#define KMALLOC_CACHES SLUB_PAGE_SHIFT
	140	#endif
	141
81819f0f CL	142	/*
	143	* We keep the general caches in an array of slab caches that are used for
	144	* 2^x bytes of allocations.
	145	*/
756dee75	146	extern struct kmem_cache kmalloc_caches[KMALLOC_CACHES];
81819f0f CL	147
	148	/*
	149	* Sorry that the following has to be that ugly but some versions of GCC
	150	* have trouble with constant propagation and loops.
	151	*/
aa137f9d	152	static __always_inline int kmalloc_index(size_t size)
81819f0f	153	{
272c1d21 CL	154	if (!size)
272c1d21 CL	155	return 0;
614410d5	156
4b356be0 CL	157	if (size <= KMALLOC_MIN_SIZE)
	158	return KMALLOC_SHIFT_LOW;
	159
acdfcd04	160	if (KMALLOC_MIN_SIZE <= 32 && size > 64 && size <= 96)
81819f0f	161	return 1;
acdfcd04	162	if (KMALLOC_MIN_SIZE <= 64 && size > 128 && size <= 192)
81819f0f CL	163	return 2;
	164	if (size <= 8) return 3;
	165	if (size <= 16) return 4;
	166	if (size <= 32) return 5;
	167	if (size <= 64) return 6;
	168	if (size <= 128) return 7;
	169	if (size <= 256) return 8;
	170	if (size <= 512) return 9;
	171	if (size <= 1024) return 10;
	172	if (size <= 2 * 1024) return 11;
6446faa2	173	if (size <= 4 * 1024) return 12;
aadb4bc4 CL	174	/*
	175	* The following is only needed to support architectures with a larger page
	176	* size than 4k.
	177	*/
81819f0f CL	178	if (size <= 8 * 1024) return 13;
	179	if (size <= 16 * 1024) return 14;
	180	if (size <= 32 * 1024) return 15;
	181	if (size <= 64 * 1024) return 16;
	182	if (size <= 128 * 1024) return 17;
	183	if (size <= 256 * 1024) return 18;
aadb4bc4	184	if (size <= 512 * 1024) return 19;
81819f0f	185	if (size <= 1024 * 1024) return 20;
81819f0f	186	if (size <= 2 * 1024 * 1024) return 21;
81819f0f CL	187	return -1;
	188
	189	/*
	190	* What we really wanted to do and cannot do because of compiler issues is:
	191	* int i;
	192	* for (i = KMALLOC_SHIFT_LOW; i <= KMALLOC_SHIFT_HIGH; i++)
	193	* if (size <= (1 << i))
	194	* return i;
	195	*/
	196	}
	197
	198	/*
	199	* Find the slab cache for a given combination of allocation flags and size.
	200	*
	201	* This ought to end up with a global pointer to the right cache
	202	* in kmalloc_caches.
	203	*/
aa137f9d	204	static __always_inline struct kmem_cache *kmalloc_slab(size_t size)
81819f0f CL	205	{
	206	int index = kmalloc_index(size);
	207
	208	if (index == 0)
	209	return NULL;
	210
81819f0f CL	211	return &kmalloc_caches[index];
	212	}
	213
6193a2ff PM	214	void kmem_cache_alloc(struct kmem_cache , gfp_t);
	215	void *__kmalloc(size_t size, gfp_t flags);
	216
0f24f128	217	#ifdef CONFIG_TRACING
5b882be4 EGM	218	extern void kmem_cache_alloc_notrace(struct kmem_cache s, gfp_t gfpflags);
	219	#else
	220	static __always_inline void *
	221	kmem_cache_alloc_notrace(struct kmem_cache *s, gfp_t gfpflags)
	222	{
	223	return kmem_cache_alloc(s, gfpflags);
	224	}
	225	#endif
	226
eada35ef PE	227	static __always_inline void *kmalloc_large(size_t size, gfp_t flags)
eada35ef PE	228	{
5b882be4 EGM	229	unsigned int order = get_order(size);
	230	void ret = (void ) __get_free_pages(flags \| __GFP_COMP, order);
	231
e4f7c0b4	232	kmemleak_alloc(ret, size, 1, flags);
ca2b84cb	233	trace_kmalloc(_THIS_IP_, ret, size, PAGE_SIZE << order, flags);
5b882be4 EGM	234
5b882be4 EGM	235	return ret;
eada35ef PE	236	}
eada35ef PE	237
aa137f9d	238	static __always_inline void *kmalloc(size_t size, gfp_t flags)
81819f0f	239	{
5b882be4 EGM	240	void *ret;
5b882be4 EGM	241
aadb4bc4	242	if (__builtin_constant_p(size)) {
ffadd4d0	243	if (size > SLUB_MAX_SIZE)
eada35ef	244	return kmalloc_large(size, flags);
81819f0f	245
aadb4bc4 CL	246	if (!(flags & SLUB_DMA)) {
	247	struct kmem_cache *s = kmalloc_slab(size);
	248
	249	if (!s)
	250	return ZERO_SIZE_PTR;
81819f0f	251
5b882be4 EGM	252	ret = kmem_cache_alloc_notrace(s, flags);
5b882be4 EGM	253
ca2b84cb	254	trace_kmalloc(_THIS_IP_, ret, size, s->size, flags);
5b882be4 EGM	255
5b882be4 EGM	256	return ret;
aadb4bc4 CL	257	}
	258	}
	259	return __kmalloc(size, flags);
81819f0f CL	260	}
81819f0f CL	261
81819f0f	262	#ifdef CONFIG_NUMA
6193a2ff PM	263	void *__kmalloc_node(size_t size, gfp_t flags, int node);
6193a2ff PM	264	void kmem_cache_alloc_node(struct kmem_cache , gfp_t flags, int node);
81819f0f	265
0f24f128	266	#ifdef CONFIG_TRACING
5b882be4 EGM	267	extern void kmem_cache_alloc_node_notrace(struct kmem_cache s,
	268	gfp_t gfpflags,
	269	int node);
	270	#else
	271	static __always_inline void *
	272	kmem_cache_alloc_node_notrace(struct kmem_cache *s,
	273	gfp_t gfpflags,
	274	int node)
	275	{
	276	return kmem_cache_alloc_node(s, gfpflags, node);
	277	}
	278	#endif
	279
aa137f9d	280	static __always_inline void *kmalloc_node(size_t size, gfp_t flags, int node)
81819f0f	281	{
5b882be4 EGM	282	void *ret;
5b882be4 EGM	283
aadb4bc4	284	if (__builtin_constant_p(size) &&
ffadd4d0	285	size <= SLUB_MAX_SIZE && !(flags & SLUB_DMA)) {
aadb4bc4	286	struct kmem_cache *s = kmalloc_slab(size);
81819f0f CL	287
81819f0f CL	288	if (!s)
272c1d21	289	return ZERO_SIZE_PTR;
81819f0f	290
5b882be4 EGM	291	ret = kmem_cache_alloc_node_notrace(s, flags, node);
5b882be4 EGM	292
ca2b84cb EGM	293	trace_kmalloc_node(_THIS_IP_, ret,
ca2b84cb EGM	294	size, s->size, flags, node);
5b882be4 EGM	295
5b882be4 EGM	296	return ret;
aadb4bc4 CL	297	}
aadb4bc4 CL	298	return __kmalloc_node(size, flags, node);
81819f0f CL	299	}
	300	#endif
	301
	302	#endif /* _LINUX_SLUB_DEF_H */