slab.c 106.5 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52
/*
 * linux/mm/slab.c
 * Written by Mark Hemment, 1996/97.
 * (markhe@nextd.demon.co.uk)
 *
 * kmem_cache_destroy() + some cleanup - 1999 Andrea Arcangeli
 *
 * Major cleanup, different bufctl logic, per-cpu arrays
 *	(c) 2000 Manfred Spraul
 *
 * Cleanup, make the head arrays unconditional, preparation for NUMA
 * 	(c) 2002 Manfred Spraul
 *
 * An implementation of the Slab Allocator as described in outline in;
 *	UNIX Internals: The New Frontiers by Uresh Vahalia
 *	Pub: Prentice Hall	ISBN 0-13-101908-2
 * or with a little more detail in;
 *	The Slab Allocator: An Object-Caching Kernel Memory Allocator
 *	Jeff Bonwick (Sun Microsystems).
 *	Presented at: USENIX Summer 1994 Technical Conference
 *
 * The memory is organized in caches, one cache for each object type.
 * (e.g. inode_cache, dentry_cache, buffer_head, vm_area_struct)
 * Each cache consists out of many slabs (they are small (usually one
 * page long) and always contiguous), and each slab contains multiple
 * initialized objects.
 *
 * This means, that your constructor is used only for newly allocated
 * slabs and you must pass objects with the same intializations to
 * kmem_cache_free.
 *
 * Each cache can only support one memory type (GFP_DMA, GFP_HIGHMEM,
 * normal). If you need a special memory type, then must create a new
 * cache for that memory type.
 *
 * In order to reduce fragmentation, the slabs are sorted in 3 groups:
 *   full slabs with 0 free objects
 *   partial slabs
 *   empty slabs with no allocated objects
 *
 * If partial slabs exist, then new allocations come from these slabs,
 * otherwise from empty slabs or new slabs are allocated.
 *
 * kmem_cache_destroy() CAN CRASH if you try to allocate from the cache
 * during kmem_cache_destroy(). The caller must prevent concurrent allocs.
 *
 * Each cache has a short per-cpu head array, most allocs
 * and frees go into that array, and if that array overflows, then 1/2
 * of the entries in the array are given back into the global cache.
 * The head array is strictly LIFO and should improve the cache hit rates.
 * On SMP, it additionally reduces the spinlock operations.
 *
A
Andrew Morton 已提交
53
 * The c_cpuarray may not be read with enabled local interrupts -
L
Linus Torvalds 已提交
54 55 56 57
 * it's changed with a smp_call_function().
 *
 * SMP synchronization:
 *  constructors and destructors are called without any locking.
58
 *  Several members in struct kmem_cache and struct slab never change, they
L
Linus Torvalds 已提交
59 60 61 62 63 64 65 66 67 68 69 70
 *	are accessed without any locking.
 *  The per-cpu arrays are never accessed from the wrong cpu, no locking,
 *  	and local interrupts are disabled so slab code is preempt-safe.
 *  The non-constant members are protected with a per-cache irq spinlock.
 *
 * Many thanks to Mark Hemment, who wrote another per-cpu slab patch
 * in 2000 - many ideas in the current implementation are derived from
 * his patch.
 *
 * Further notes from the original documentation:
 *
 * 11 April '97.  Started multi-threading - markhe
I
Ingo Molnar 已提交
71
 *	The global cache-chain is protected by the mutex 'cache_chain_mutex'.
L
Linus Torvalds 已提交
72 73 74 75 76 77
 *	The sem is only needed when accessing/extending the cache-chain, which
 *	can never happen inside an interrupt (kmem_cache_create(),
 *	kmem_cache_shrink() and kmem_cache_reap()).
 *
 *	At present, each engine can be growing a cache.  This should be blocked.
 *
78 79 80 81 82 83 84 85 86
 * 15 March 2005. NUMA slab allocator.
 *	Shai Fultheim <shai@scalex86.org>.
 *	Shobhit Dayal <shobhit@calsoftinc.com>
 *	Alok N Kataria <alokk@calsoftinc.com>
 *	Christoph Lameter <christoph@lameter.com>
 *
 *	Modified the slab allocator to be node aware on NUMA systems.
 *	Each node has its own list of partial, free and full slabs.
 *	All object allocations for a node occur from node specific slab lists.
L
Linus Torvalds 已提交
87 88 89 90 91 92 93 94 95 96
 */

#include	<linux/config.h>
#include	<linux/slab.h>
#include	<linux/mm.h>
#include	<linux/swap.h>
#include	<linux/cache.h>
#include	<linux/interrupt.h>
#include	<linux/init.h>
#include	<linux/compiler.h>
97
#include	<linux/cpuset.h>
L
Linus Torvalds 已提交
98 99 100 101 102 103 104
#include	<linux/seq_file.h>
#include	<linux/notifier.h>
#include	<linux/kallsyms.h>
#include	<linux/cpu.h>
#include	<linux/sysctl.h>
#include	<linux/module.h>
#include	<linux/rcupdate.h>
105
#include	<linux/string.h>
106
#include	<linux/nodemask.h>
107
#include	<linux/mempolicy.h>
I
Ingo Molnar 已提交
108
#include	<linux/mutex.h>
L
Linus Torvalds 已提交
109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173

#include	<asm/uaccess.h>
#include	<asm/cacheflush.h>
#include	<asm/tlbflush.h>
#include	<asm/page.h>

/*
 * DEBUG	- 1 for kmem_cache_create() to honour; SLAB_DEBUG_INITIAL,
 *		  SLAB_RED_ZONE & SLAB_POISON.
 *		  0 for faster, smaller code (especially in the critical paths).
 *
 * STATS	- 1 to collect stats for /proc/slabinfo.
 *		  0 for faster, smaller code (especially in the critical paths).
 *
 * FORCED_DEBUG	- 1 enables SLAB_RED_ZONE and SLAB_POISON (if possible)
 */

#ifdef CONFIG_DEBUG_SLAB
#define	DEBUG		1
#define	STATS		1
#define	FORCED_DEBUG	1
#else
#define	DEBUG		0
#define	STATS		0
#define	FORCED_DEBUG	0
#endif

/* Shouldn't this be in a header file somewhere? */
#define	BYTES_PER_WORD		sizeof(void *)

#ifndef cache_line_size
#define cache_line_size()	L1_CACHE_BYTES
#endif

#ifndef ARCH_KMALLOC_MINALIGN
/*
 * Enforce a minimum alignment for the kmalloc caches.
 * Usually, the kmalloc caches are cache_line_size() aligned, except when
 * DEBUG and FORCED_DEBUG are enabled, then they are BYTES_PER_WORD aligned.
 * Some archs want to perform DMA into kmalloc caches and need a guaranteed
 * alignment larger than BYTES_PER_WORD. ARCH_KMALLOC_MINALIGN allows that.
 * Note that this flag disables some debug features.
 */
#define ARCH_KMALLOC_MINALIGN 0
#endif

#ifndef ARCH_SLAB_MINALIGN
/*
 * Enforce a minimum alignment for all caches.
 * Intended for archs that get misalignment faults even for BYTES_PER_WORD
 * aligned buffers. Includes ARCH_KMALLOC_MINALIGN.
 * If possible: Do not enable this flag for CONFIG_DEBUG_SLAB, it disables
 * some debug features.
 */
#define ARCH_SLAB_MINALIGN 0
#endif

#ifndef ARCH_KMALLOC_FLAGS
#define ARCH_KMALLOC_FLAGS SLAB_HWCACHE_ALIGN
#endif

/* Legal flag mask for kmem_cache_create(). */
#if DEBUG
# define CREATE_MASK	(SLAB_DEBUG_INITIAL | SLAB_RED_ZONE | \
			 SLAB_POISON | SLAB_HWCACHE_ALIGN | \
174
			 SLAB_CACHE_DMA | \
L
Linus Torvalds 已提交
175 176
			 SLAB_MUST_HWCACHE_ALIGN | SLAB_STORE_USER | \
			 SLAB_RECLAIM_ACCOUNT | SLAB_PANIC | \
177
			 SLAB_DESTROY_BY_RCU | SLAB_MEM_SPREAD)
L
Linus Torvalds 已提交
178
#else
179
# define CREATE_MASK	(SLAB_HWCACHE_ALIGN | \
L
Linus Torvalds 已提交
180 181
			 SLAB_CACHE_DMA | SLAB_MUST_HWCACHE_ALIGN | \
			 SLAB_RECLAIM_ACCOUNT | SLAB_PANIC | \
182
			 SLAB_DESTROY_BY_RCU | SLAB_MEM_SPREAD)
L
Linus Torvalds 已提交
183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203
#endif

/*
 * kmem_bufctl_t:
 *
 * Bufctl's are used for linking objs within a slab
 * linked offsets.
 *
 * This implementation relies on "struct page" for locating the cache &
 * slab an object belongs to.
 * This allows the bufctl structure to be small (one int), but limits
 * the number of objects a slab (not a cache) can contain when off-slab
 * bufctls are used. The limit is the size of the largest general cache
 * that does not use off-slab slabs.
 * For 32bit archs with 4 kB pages, is this 56.
 * This is not serious, as it is only for large objects, when it is unwise
 * to have too many per slab.
 * Note: This limit can be raised by introducing a general cache whose size
 * is less than 512 (PAGE_SIZE<<3), but greater than 256.
 */

204
typedef unsigned int kmem_bufctl_t;
L
Linus Torvalds 已提交
205 206
#define BUFCTL_END	(((kmem_bufctl_t)(~0U))-0)
#define BUFCTL_FREE	(((kmem_bufctl_t)(~0U))-1)
207 208
#define	BUFCTL_ACTIVE	(((kmem_bufctl_t)(~0U))-2)
#define	SLAB_LIMIT	(((kmem_bufctl_t)(~0U))-3)
L
Linus Torvalds 已提交
209 210 211 212 213 214 215 216 217 218 219 220 221 222

/* Max number of objs-per-slab for caches which use off-slab slabs.
 * Needed to avoid a possible looping condition in cache_grow().
 */
static unsigned long offslab_limit;

/*
 * struct slab
 *
 * Manages the objs in a slab. Placed either at the beginning of mem allocated
 * for a slab, or allocated from an general cache.
 * Slabs are chained into three list: fully used, partial, fully free slabs.
 */
struct slab {
P
Pekka Enberg 已提交
223 224 225 226 227 228
	struct list_head list;
	unsigned long colouroff;
	void *s_mem;		/* including colour offset */
	unsigned int inuse;	/* num of objs active in slab */
	kmem_bufctl_t free;
	unsigned short nodeid;
L
Linus Torvalds 已提交
229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247
};

/*
 * struct slab_rcu
 *
 * slab_destroy on a SLAB_DESTROY_BY_RCU cache uses this structure to
 * arrange for kmem_freepages to be called via RCU.  This is useful if
 * we need to approach a kernel structure obliquely, from its address
 * obtained without the usual locking.  We can lock the structure to
 * stabilize it and check it's still at the given address, only if we
 * can be sure that the memory has not been meanwhile reused for some
 * other kind of object (which our subsystem's lock might corrupt).
 *
 * rcu_read_lock before reading the address, then rcu_read_unlock after
 * taking the spinlock within the structure expected at that address.
 *
 * We assume struct slab_rcu can overlay struct slab when destroying.
 */
struct slab_rcu {
P
Pekka Enberg 已提交
248
	struct rcu_head head;
249
	struct kmem_cache *cachep;
P
Pekka Enberg 已提交
250
	void *addr;
L
Linus Torvalds 已提交
251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269
};

/*
 * struct array_cache
 *
 * Purpose:
 * - LIFO ordering, to hand out cache-warm objects from _alloc
 * - reduce the number of linked list operations
 * - reduce spinlock operations
 *
 * The limit is stored in the per-cpu structure to reduce the data cache
 * footprint.
 *
 */
struct array_cache {
	unsigned int avail;
	unsigned int limit;
	unsigned int batchcount;
	unsigned int touched;
270
	spinlock_t lock;
A
Andrew Morton 已提交
271 272 273 274 275 276
	void *entry[0];	/*
			 * Must have this definition in here for the proper
			 * alignment of array_cache. Also simplifies accessing
			 * the entries.
			 * [0] is for gcc 2.95. It should really be [].
			 */
L
Linus Torvalds 已提交
277 278
};

A
Andrew Morton 已提交
279 280 281
/*
 * bootstrap: The caches do not work without cpuarrays anymore, but the
 * cpuarrays are allocated from the generic caches...
L
Linus Torvalds 已提交
282 283 284 285
 */
#define BOOT_CPUCACHE_ENTRIES	1
struct arraycache_init {
	struct array_cache cache;
P
Pekka Enberg 已提交
286
	void *entries[BOOT_CPUCACHE_ENTRIES];
L
Linus Torvalds 已提交
287 288 289
};

/*
290
 * The slab lists for all objects.
L
Linus Torvalds 已提交
291 292
 */
struct kmem_list3 {
P
Pekka Enberg 已提交
293 294 295 296 297
	struct list_head slabs_partial;	/* partial list first, better asm code */
	struct list_head slabs_full;
	struct list_head slabs_free;
	unsigned long free_objects;
	unsigned int free_limit;
298
	unsigned int colour_next;	/* Per-node cache coloring */
P
Pekka Enberg 已提交
299 300 301
	spinlock_t list_lock;
	struct array_cache *shared;	/* shared per node */
	struct array_cache **alien;	/* on other nodes */
302 303
	unsigned long next_reap;	/* updated without locking */
	int free_touched;		/* updated without locking */
L
Linus Torvalds 已提交
304 305
};

306 307 308 309 310 311 312 313 314 315
/*
 * Need this for bootstrapping a per node allocator.
 */
#define NUM_INIT_LISTS (2 * MAX_NUMNODES + 1)
struct kmem_list3 __initdata initkmem_list3[NUM_INIT_LISTS];
#define	CACHE_CACHE 0
#define	SIZE_AC 1
#define	SIZE_L3 (1 + MAX_NUMNODES)

/*
A
Andrew Morton 已提交
316 317
 * This function must be completely optimized away if a constant is passed to
 * it.  Mostly the same as what is in linux/slab.h except it returns an index.
318
 */
319
static __always_inline int index_of(const size_t size)
320
{
321 322
	extern void __bad_size(void);

323 324 325 326 327 328 329 330 331 332
	if (__builtin_constant_p(size)) {
		int i = 0;

#define CACHE(x) \
	if (size <=x) \
		return i; \
	else \
		i++;
#include "linux/kmalloc_sizes.h"
#undef CACHE
333
		__bad_size();
334
	} else
335
		__bad_size();
336 337 338 339 340
	return 0;
}

#define INDEX_AC index_of(sizeof(struct arraycache_init))
#define INDEX_L3 index_of(sizeof(struct kmem_list3))
L
Linus Torvalds 已提交
341

P
Pekka Enberg 已提交
342
static void kmem_list3_init(struct kmem_list3 *parent)
343 344 345 346 347 348
{
	INIT_LIST_HEAD(&parent->slabs_full);
	INIT_LIST_HEAD(&parent->slabs_partial);
	INIT_LIST_HEAD(&parent->slabs_free);
	parent->shared = NULL;
	parent->alien = NULL;
349
	parent->colour_next = 0;
350 351 352 353 354
	spin_lock_init(&parent->list_lock);
	parent->free_objects = 0;
	parent->free_touched = 0;
}

A
Andrew Morton 已提交
355 356 357 358
#define MAKE_LIST(cachep, listp, slab, nodeid)				\
	do {								\
		INIT_LIST_HEAD(listp);					\
		list_splice(&(cachep->nodelists[nodeid]->slab), listp);	\
359 360
	} while (0)

A
Andrew Morton 已提交
361 362
#define	MAKE_ALL_LISTS(cachep, ptr, nodeid)				\
	do {								\
363 364 365 366
	MAKE_LIST((cachep), (&(ptr)->slabs_full), slabs_full, nodeid);	\
	MAKE_LIST((cachep), (&(ptr)->slabs_partial), slabs_partial, nodeid); \
	MAKE_LIST((cachep), (&(ptr)->slabs_free), slabs_free, nodeid);	\
	} while (0)
L
Linus Torvalds 已提交
367 368

/*
369
 * struct kmem_cache
L
Linus Torvalds 已提交
370 371 372
 *
 * manages a cache.
 */
P
Pekka Enberg 已提交
373

374
struct kmem_cache {
L
Linus Torvalds 已提交
375
/* 1) per-cpu data, touched during every alloc/free */
P
Pekka Enberg 已提交
376
	struct array_cache *array[NR_CPUS];
377
/* 2) Cache tunables. Protected by cache_chain_mutex */
P
Pekka Enberg 已提交
378 379 380
	unsigned int batchcount;
	unsigned int limit;
	unsigned int shared;
381

382
	unsigned int buffer_size;
383
/* 3) touched by every alloc & free from the backend */
P
Pekka Enberg 已提交
384
	struct kmem_list3 *nodelists[MAX_NUMNODES];
385

A
Andrew Morton 已提交
386 387
	unsigned int flags;		/* constant flags */
	unsigned int num;		/* # of objs per slab */
L
Linus Torvalds 已提交
388

389
/* 4) cache_grow/shrink */
L
Linus Torvalds 已提交
390
	/* order of pgs per slab (2^n) */
P
Pekka Enberg 已提交
391
	unsigned int gfporder;
L
Linus Torvalds 已提交
392 393

	/* force GFP flags, e.g. GFP_DMA */
P
Pekka Enberg 已提交
394
	gfp_t gfpflags;
L
Linus Torvalds 已提交
395

A
Andrew Morton 已提交
396
	size_t colour;			/* cache colouring range */
P
Pekka Enberg 已提交
397
	unsigned int colour_off;	/* colour offset */
398
	struct kmem_cache *slabp_cache;
P
Pekka Enberg 已提交
399
	unsigned int slab_size;
A
Andrew Morton 已提交
400
	unsigned int dflags;		/* dynamic flags */
L
Linus Torvalds 已提交
401 402

	/* constructor func */
403
	void (*ctor) (void *, struct kmem_cache *, unsigned long);
L
Linus Torvalds 已提交
404 405

	/* de-constructor func */
406
	void (*dtor) (void *, struct kmem_cache *, unsigned long);
L
Linus Torvalds 已提交
407

408
/* 5) cache creation/removal */
P
Pekka Enberg 已提交
409 410
	const char *name;
	struct list_head next;
L
Linus Torvalds 已提交
411

412
/* 6) statistics */
L
Linus Torvalds 已提交
413
#if STATS
P
Pekka Enberg 已提交
414 415 416 417 418 419 420 421 422 423 424 425 426
	unsigned long num_active;
	unsigned long num_allocations;
	unsigned long high_mark;
	unsigned long grown;
	unsigned long reaped;
	unsigned long errors;
	unsigned long max_freeable;
	unsigned long node_allocs;
	unsigned long node_frees;
	atomic_t allochit;
	atomic_t allocmiss;
	atomic_t freehit;
	atomic_t freemiss;
L
Linus Torvalds 已提交
427 428
#endif
#if DEBUG
429 430 431 432 433 434 435 436
	/*
	 * If debugging is enabled, then the allocator can add additional
	 * fields and/or padding to every object. buffer_size contains the total
	 * object size including these internal fields, the following two
	 * variables contain the offset to the user object and its size.
	 */
	int obj_offset;
	int obj_size;
L
Linus Torvalds 已提交
437 438 439 440 441 442 443
#endif
};

#define CFLGS_OFF_SLAB		(0x80000000UL)
#define	OFF_SLAB(x)	((x)->flags & CFLGS_OFF_SLAB)

#define BATCHREFILL_LIMIT	16
A
Andrew Morton 已提交
444 445 446
/*
 * Optimization question: fewer reaps means less probability for unnessary
 * cpucache drain/refill cycles.
L
Linus Torvalds 已提交
447
 *
A
Adrian Bunk 已提交
448
 * OTOH the cpuarrays can contain lots of objects,
L
Linus Torvalds 已提交
449 450 451 452 453 454 455 456 457 458 459
 * which could lock up otherwise freeable slabs.
 */
#define REAPTIMEOUT_CPUC	(2*HZ)
#define REAPTIMEOUT_LIST3	(4*HZ)

#if STATS
#define	STATS_INC_ACTIVE(x)	((x)->num_active++)
#define	STATS_DEC_ACTIVE(x)	((x)->num_active--)
#define	STATS_INC_ALLOCED(x)	((x)->num_allocations++)
#define	STATS_INC_GROWN(x)	((x)->grown++)
#define	STATS_INC_REAPED(x)	((x)->reaped++)
A
Andrew Morton 已提交
460 461 462 463 464
#define	STATS_SET_HIGH(x)						\
	do {								\
		if ((x)->num_active > (x)->high_mark)			\
			(x)->high_mark = (x)->num_active;		\
	} while (0)
L
Linus Torvalds 已提交
465 466
#define	STATS_INC_ERR(x)	((x)->errors++)
#define	STATS_INC_NODEALLOCS(x)	((x)->node_allocs++)
467
#define	STATS_INC_NODEFREES(x)	((x)->node_frees++)
A
Andrew Morton 已提交
468 469 470 471 472
#define	STATS_SET_FREEABLE(x, i)					\
	do {								\
		if ((x)->max_freeable < i)				\
			(x)->max_freeable = i;				\
	} while (0)
L
Linus Torvalds 已提交
473 474 475 476 477 478 479 480 481 482 483 484 485
#define STATS_INC_ALLOCHIT(x)	atomic_inc(&(x)->allochit)
#define STATS_INC_ALLOCMISS(x)	atomic_inc(&(x)->allocmiss)
#define STATS_INC_FREEHIT(x)	atomic_inc(&(x)->freehit)
#define STATS_INC_FREEMISS(x)	atomic_inc(&(x)->freemiss)
#else
#define	STATS_INC_ACTIVE(x)	do { } while (0)
#define	STATS_DEC_ACTIVE(x)	do { } while (0)
#define	STATS_INC_ALLOCED(x)	do { } while (0)
#define	STATS_INC_GROWN(x)	do { } while (0)
#define	STATS_INC_REAPED(x)	do { } while (0)
#define	STATS_SET_HIGH(x)	do { } while (0)
#define	STATS_INC_ERR(x)	do { } while (0)
#define	STATS_INC_NODEALLOCS(x)	do { } while (0)
486
#define	STATS_INC_NODEFREES(x)	do { } while (0)
A
Andrew Morton 已提交
487
#define	STATS_SET_FREEABLE(x, i) do { } while (0)
L
Linus Torvalds 已提交
488 489 490 491 492 493 494
#define STATS_INC_ALLOCHIT(x)	do { } while (0)
#define STATS_INC_ALLOCMISS(x)	do { } while (0)
#define STATS_INC_FREEHIT(x)	do { } while (0)
#define STATS_INC_FREEMISS(x)	do { } while (0)
#endif

#if DEBUG
A
Andrew Morton 已提交
495 496
/*
 * Magic nums for obj red zoning.
L
Linus Torvalds 已提交
497 498 499 500 501 502 503 504 505 506
 * Placed in the first word before and the first word after an obj.
 */
#define	RED_INACTIVE	0x5A2CF071UL	/* when obj is inactive */
#define	RED_ACTIVE	0x170FC2A5UL	/* when obj is active */

/* ...and for poisoning */
#define	POISON_INUSE	0x5a	/* for use-uninitialised poisoning */
#define POISON_FREE	0x6b	/* for use-after-free poisoning */
#define	POISON_END	0xa5	/* end-byte of poisoning */

A
Andrew Morton 已提交
507 508
/*
 * memory layout of objects:
L
Linus Torvalds 已提交
509
 * 0		: objp
510
 * 0 .. cachep->obj_offset - BYTES_PER_WORD - 1: padding. This ensures that
L
Linus Torvalds 已提交
511 512
 * 		the end of an object is aligned with the end of the real
 * 		allocation. Catches writes behind the end of the allocation.
513
 * cachep->obj_offset - BYTES_PER_WORD .. cachep->obj_offset - 1:
L
Linus Torvalds 已提交
514
 * 		redzone word.
515 516
 * cachep->obj_offset: The real object.
 * cachep->buffer_size - 2* BYTES_PER_WORD: redzone word [BYTES_PER_WORD long]
A
Andrew Morton 已提交
517 518
 * cachep->buffer_size - 1* BYTES_PER_WORD: last caller address
 *					[BYTES_PER_WORD long]
L
Linus Torvalds 已提交
519
 */
520
static int obj_offset(struct kmem_cache *cachep)
L
Linus Torvalds 已提交
521
{
522
	return cachep->obj_offset;
L
Linus Torvalds 已提交
523 524
}

525
static int obj_size(struct kmem_cache *cachep)
L
Linus Torvalds 已提交
526
{
527
	return cachep->obj_size;
L
Linus Torvalds 已提交
528 529
}

530
static unsigned long *dbg_redzone1(struct kmem_cache *cachep, void *objp)
L
Linus Torvalds 已提交
531 532
{
	BUG_ON(!(cachep->flags & SLAB_RED_ZONE));
533
	return (unsigned long*) (objp+obj_offset(cachep)-BYTES_PER_WORD);
L
Linus Torvalds 已提交
534 535
}

536
static unsigned long *dbg_redzone2(struct kmem_cache *cachep, void *objp)
L
Linus Torvalds 已提交
537 538 539
{
	BUG_ON(!(cachep->flags & SLAB_RED_ZONE));
	if (cachep->flags & SLAB_STORE_USER)
540
		return (unsigned long *)(objp + cachep->buffer_size -
P
Pekka Enberg 已提交
541
					 2 * BYTES_PER_WORD);
542
	return (unsigned long *)(objp + cachep->buffer_size - BYTES_PER_WORD);
L
Linus Torvalds 已提交
543 544
}

545
static void **dbg_userword(struct kmem_cache *cachep, void *objp)
L
Linus Torvalds 已提交
546 547
{
	BUG_ON(!(cachep->flags & SLAB_STORE_USER));
548
	return (void **)(objp + cachep->buffer_size - BYTES_PER_WORD);
L
Linus Torvalds 已提交
549 550 551 552
}

#else

553 554
#define obj_offset(x)			0
#define obj_size(cachep)		(cachep->buffer_size)
L
Linus Torvalds 已提交
555 556 557 558 559 560 561
#define dbg_redzone1(cachep, objp)	({BUG(); (unsigned long *)NULL;})
#define dbg_redzone2(cachep, objp)	({BUG(); (unsigned long *)NULL;})
#define dbg_userword(cachep, objp)	({BUG(); (void **)NULL;})

#endif

/*
A
Andrew Morton 已提交
562 563
 * Maximum size of an obj (in 2^order pages) and absolute limit for the gfp
 * order.
L
Linus Torvalds 已提交
564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582
 */
#if defined(CONFIG_LARGE_ALLOCS)
#define	MAX_OBJ_ORDER	13	/* up to 32Mb */
#define	MAX_GFP_ORDER	13	/* up to 32Mb */
#elif defined(CONFIG_MMU)
#define	MAX_OBJ_ORDER	5	/* 32 pages */
#define	MAX_GFP_ORDER	5	/* 32 pages */
#else
#define	MAX_OBJ_ORDER	8	/* up to 1Mb */
#define	MAX_GFP_ORDER	8	/* up to 1Mb */
#endif

/*
 * Do not go above this order unless 0 objects fit into the slab.
 */
#define	BREAK_GFP_ORDER_HI	1
#define	BREAK_GFP_ORDER_LO	0
static int slab_break_gfp_order = BREAK_GFP_ORDER_LO;

A
Andrew Morton 已提交
583 584 585 586
/*
 * Functions for storing/retrieving the cachep and or slab from the page
 * allocator.  These are used to find the slab an obj belongs to.  With kfree(),
 * these are used to find the cache which an obj belongs to.
L
Linus Torvalds 已提交
587
 */
588 589 590 591 592 593 594
static inline void page_set_cache(struct page *page, struct kmem_cache *cache)
{
	page->lru.next = (struct list_head *)cache;
}

static inline struct kmem_cache *page_get_cache(struct page *page)
{
595 596
	if (unlikely(PageCompound(page)))
		page = (struct page *)page_private(page);
597 598 599 600 601 602 603 604 605 606
	return (struct kmem_cache *)page->lru.next;
}

static inline void page_set_slab(struct page *page, struct slab *slab)
{
	page->lru.prev = (struct list_head *)slab;
}

static inline struct slab *page_get_slab(struct page *page)
{
607 608
	if (unlikely(PageCompound(page)))
		page = (struct page *)page_private(page);
609 610
	return (struct slab *)page->lru.prev;
}
L
Linus Torvalds 已提交
611

612 613 614 615 616 617 618 619 620 621 622 623
static inline struct kmem_cache *virt_to_cache(const void *obj)
{
	struct page *page = virt_to_page(obj);
	return page_get_cache(page);
}

static inline struct slab *virt_to_slab(const void *obj)
{
	struct page *page = virt_to_page(obj);
	return page_get_slab(page);
}

624 625 626 627 628 629 630 631 632 633 634 635
static inline void *index_to_obj(struct kmem_cache *cache, struct slab *slab,
				 unsigned int idx)
{
	return slab->s_mem + cache->buffer_size * idx;
}

static inline unsigned int obj_to_index(struct kmem_cache *cache,
					struct slab *slab, void *obj)
{
	return (unsigned)(obj - slab->s_mem) / cache->buffer_size;
}

A
Andrew Morton 已提交
636 637 638
/*
 * These are the default caches for kmalloc. Custom caches can have other sizes.
 */
L
Linus Torvalds 已提交
639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655
struct cache_sizes malloc_sizes[] = {
#define CACHE(x) { .cs_size = (x) },
#include <linux/kmalloc_sizes.h>
	CACHE(ULONG_MAX)
#undef CACHE
};
EXPORT_SYMBOL(malloc_sizes);

/* Must match cache_sizes above. Out of line to keep cache footprint low. */
struct cache_names {
	char *name;
	char *name_dma;
};

static struct cache_names __initdata cache_names[] = {
#define CACHE(x) { .name = "size-" #x, .name_dma = "size-" #x "(DMA)" },
#include <linux/kmalloc_sizes.h>
P
Pekka Enberg 已提交
656
	{NULL,}
L
Linus Torvalds 已提交
657 658 659 660
#undef CACHE
};

static struct arraycache_init initarray_cache __initdata =
P
Pekka Enberg 已提交
661
    { {0, BOOT_CPUCACHE_ENTRIES, 1, 0} };
L
Linus Torvalds 已提交
662
static struct arraycache_init initarray_generic =
P
Pekka Enberg 已提交
663
    { {0, BOOT_CPUCACHE_ENTRIES, 1, 0} };
L
Linus Torvalds 已提交
664 665

/* internal cache of cache description objs */
666
static struct kmem_cache cache_cache = {
P
Pekka Enberg 已提交
667 668 669
	.batchcount = 1,
	.limit = BOOT_CPUCACHE_ENTRIES,
	.shared = 1,
670
	.buffer_size = sizeof(struct kmem_cache),
P
Pekka Enberg 已提交
671
	.name = "kmem_cache",
L
Linus Torvalds 已提交
672
#if DEBUG
673
	.obj_size = sizeof(struct kmem_cache),
L
Linus Torvalds 已提交
674 675 676 677
#endif
};

/* Guard access to the cache-chain. */
I
Ingo Molnar 已提交
678
static DEFINE_MUTEX(cache_chain_mutex);
L
Linus Torvalds 已提交
679 680 681
static struct list_head cache_chain;

/*
A
Andrew Morton 已提交
682 683
 * vm_enough_memory() looks at this to determine how many slab-allocated pages
 * are possibly freeable under pressure
L
Linus Torvalds 已提交
684 685 686 687 688 689 690 691 692 693 694
 *
 * SLAB_RECLAIM_ACCOUNT turns this on per-slab
 */
atomic_t slab_reclaim_pages;

/*
 * chicken and egg problem: delay the per-cpu array allocation
 * until the general caches are up.
 */
static enum {
	NONE,
695 696
	PARTIAL_AC,
	PARTIAL_L3,
L
Linus Torvalds 已提交
697 698 699 700 701
	FULL
} g_cpucache_up;

static DEFINE_PER_CPU(struct work_struct, reap_work);

A
Andrew Morton 已提交
702 703
static void free_block(struct kmem_cache *cachep, void **objpp, int len,
			int node);
704
static void enable_cpucache(struct kmem_cache *cachep);
P
Pekka Enberg 已提交
705
static void cache_reap(void *unused);
706
static int __node_shrink(struct kmem_cache *cachep, int node);
L
Linus Torvalds 已提交
707

708
static inline struct array_cache *cpu_cache_get(struct kmem_cache *cachep)
L
Linus Torvalds 已提交
709 710 711 712
{
	return cachep->array[smp_processor_id()];
}

A
Andrew Morton 已提交
713 714
static inline struct kmem_cache *__find_general_cachep(size_t size,
							gfp_t gfpflags)
L
Linus Torvalds 已提交
715 716 717 718 719
{
	struct cache_sizes *csizep = malloc_sizes;

#if DEBUG
	/* This happens if someone tries to call
P
Pekka Enberg 已提交
720 721 722
	 * kmem_cache_create(), or __kmalloc(), before
	 * the generic caches are initialized.
	 */
723
	BUG_ON(malloc_sizes[INDEX_AC].cs_cachep == NULL);
L
Linus Torvalds 已提交
724 725 726 727 728
#endif
	while (size > csizep->cs_size)
		csizep++;

	/*
729
	 * Really subtle: The last entry with cs->cs_size==ULONG_MAX
L
Linus Torvalds 已提交
730 731 732 733 734 735 736 737
	 * has cs_{dma,}cachep==NULL. Thus no special case
	 * for large kmalloc calls required.
	 */
	if (unlikely(gfpflags & GFP_DMA))
		return csizep->cs_dmacachep;
	return csizep->cs_cachep;
}

738
struct kmem_cache *kmem_find_general_cachep(size_t size, gfp_t gfpflags)
739 740 741 742 743
{
	return __find_general_cachep(size, gfpflags);
}
EXPORT_SYMBOL(kmem_find_general_cachep);

744
static size_t slab_mgmt_size(size_t nr_objs, size_t align)
L
Linus Torvalds 已提交
745
{
746 747
	return ALIGN(sizeof(struct slab)+nr_objs*sizeof(kmem_bufctl_t), align);
}
L
Linus Torvalds 已提交
748

A
Andrew Morton 已提交
749 750 751
/*
 * Calculate the number of objects and left-over bytes for a given buffer size.
 */
752 753 754 755 756 757 758
static void cache_estimate(unsigned long gfporder, size_t buffer_size,
			   size_t align, int flags, size_t *left_over,
			   unsigned int *num)
{
	int nr_objs;
	size_t mgmt_size;
	size_t slab_size = PAGE_SIZE << gfporder;
L
Linus Torvalds 已提交
759

760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807
	/*
	 * The slab management structure can be either off the slab or
	 * on it. For the latter case, the memory allocated for a
	 * slab is used for:
	 *
	 * - The struct slab
	 * - One kmem_bufctl_t for each object
	 * - Padding to respect alignment of @align
	 * - @buffer_size bytes for each object
	 *
	 * If the slab management structure is off the slab, then the
	 * alignment will already be calculated into the size. Because
	 * the slabs are all pages aligned, the objects will be at the
	 * correct alignment when allocated.
	 */
	if (flags & CFLGS_OFF_SLAB) {
		mgmt_size = 0;
		nr_objs = slab_size / buffer_size;

		if (nr_objs > SLAB_LIMIT)
			nr_objs = SLAB_LIMIT;
	} else {
		/*
		 * Ignore padding for the initial guess. The padding
		 * is at most @align-1 bytes, and @buffer_size is at
		 * least @align. In the worst case, this result will
		 * be one greater than the number of objects that fit
		 * into the memory allocation when taking the padding
		 * into account.
		 */
		nr_objs = (slab_size - sizeof(struct slab)) /
			  (buffer_size + sizeof(kmem_bufctl_t));

		/*
		 * This calculated number will be either the right
		 * amount, or one greater than what we want.
		 */
		if (slab_mgmt_size(nr_objs, align) + nr_objs*buffer_size
		       > slab_size)
			nr_objs--;

		if (nr_objs > SLAB_LIMIT)
			nr_objs = SLAB_LIMIT;

		mgmt_size = slab_mgmt_size(nr_objs, align);
	}
	*num = nr_objs;
	*left_over = slab_size - nr_objs*buffer_size - mgmt_size;
L
Linus Torvalds 已提交
808 809 810 811
}

#define slab_error(cachep, msg) __slab_error(__FUNCTION__, cachep, msg)

A
Andrew Morton 已提交
812 813
static void __slab_error(const char *function, struct kmem_cache *cachep,
			char *msg)
L
Linus Torvalds 已提交
814 815
{
	printk(KERN_ERR "slab error in %s(): cache `%s': %s\n",
P
Pekka Enberg 已提交
816
	       function, cachep->name, msg);
L
Linus Torvalds 已提交
817 818 819
	dump_stack();
}

820 821 822 823 824 825 826 827 828 829 830 831 832 833 834
#ifdef CONFIG_NUMA
/*
 * Special reaping functions for NUMA systems called from cache_reap().
 * These take care of doing round robin flushing of alien caches (containing
 * objects freed on different nodes from which they were allocated) and the
 * flushing of remote pcps by calling drain_node_pages.
 */
static DEFINE_PER_CPU(unsigned long, reap_node);

static void init_reap_node(int cpu)
{
	int node;

	node = next_node(cpu_to_node(cpu), node_online_map);
	if (node == MAX_NUMNODES)
835
		node = first_node(node_online_map);
836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860

	__get_cpu_var(reap_node) = node;
}

static void next_reap_node(void)
{
	int node = __get_cpu_var(reap_node);

	/*
	 * Also drain per cpu pages on remote zones
	 */
	if (node != numa_node_id())
		drain_node_pages(node);

	node = next_node(node, node_online_map);
	if (unlikely(node >= MAX_NUMNODES))
		node = first_node(node_online_map);
	__get_cpu_var(reap_node) = node;
}

#else
#define init_reap_node(cpu) do { } while (0)
#define next_reap_node(void) do { } while (0)
#endif

L
Linus Torvalds 已提交
861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877
/*
 * Initiate the reap timer running on the target CPU.  We run at around 1 to 2Hz
 * via the workqueue/eventd.
 * Add the CPU number into the expiration time to minimize the possibility of
 * the CPUs getting into lockstep and contending for the global cache chain
 * lock.
 */
static void __devinit start_cpu_timer(int cpu)
{
	struct work_struct *reap_work = &per_cpu(reap_work, cpu);

	/*
	 * When this gets called from do_initcalls via cpucache_init(),
	 * init_workqueues() has already run, so keventd will be setup
	 * at that time.
	 */
	if (keventd_up() && reap_work->func == NULL) {
878
		init_reap_node(cpu);
L
Linus Torvalds 已提交
879 880 881 882 883
		INIT_WORK(reap_work, cache_reap, NULL);
		schedule_delayed_work_on(cpu, reap_work, HZ + 3 * cpu);
	}
}

884
static struct array_cache *alloc_arraycache(int node, int entries,
P
Pekka Enberg 已提交
885
					    int batchcount)
L
Linus Torvalds 已提交
886
{
P
Pekka Enberg 已提交
887
	int memsize = sizeof(void *) * entries + sizeof(struct array_cache);
L
Linus Torvalds 已提交
888 889
	struct array_cache *nc = NULL;

890
	nc = kmalloc_node(memsize, GFP_KERNEL, node);
L
Linus Torvalds 已提交
891 892 893 894 895
	if (nc) {
		nc->avail = 0;
		nc->limit = entries;
		nc->batchcount = batchcount;
		nc->touched = 0;
896
		spin_lock_init(&nc->lock);
L
Linus Torvalds 已提交
897 898 899 900
	}
	return nc;
}

901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924
/*
 * Transfer objects in one arraycache to another.
 * Locking must be handled by the caller.
 *
 * Return the number of entries transferred.
 */
static int transfer_objects(struct array_cache *to,
		struct array_cache *from, unsigned int max)
{
	/* Figure out how many entries to transfer */
	int nr = min(min(from->avail, max), to->limit - to->avail);

	if (!nr)
		return 0;

	memcpy(to->entry + to->avail, from->entry + from->avail -nr,
			sizeof(void *) *nr);

	from->avail -= nr;
	to->avail += nr;
	to->touched = 1;
	return nr;
}

925
#ifdef CONFIG_NUMA
926
static void *__cache_alloc_node(struct kmem_cache *, gfp_t, int);
927
static void *alternate_node_alloc(struct kmem_cache *, gfp_t);
928

P
Pekka Enberg 已提交
929
static struct array_cache **alloc_alien_cache(int node, int limit)
930 931
{
	struct array_cache **ac_ptr;
P
Pekka Enberg 已提交
932
	int memsize = sizeof(void *) * MAX_NUMNODES;
933 934 935 936 937 938 939 940 941 942 943 944 945
	int i;

	if (limit > 1)
		limit = 12;
	ac_ptr = kmalloc_node(memsize, GFP_KERNEL, node);
	if (ac_ptr) {
		for_each_node(i) {
			if (i == node || !node_online(i)) {
				ac_ptr[i] = NULL;
				continue;
			}
			ac_ptr[i] = alloc_arraycache(node, limit, 0xbaadf00d);
			if (!ac_ptr[i]) {
P
Pekka Enberg 已提交
946
				for (i--; i <= 0; i--)
947 948 949 950 951 952 953 954 955
					kfree(ac_ptr[i]);
				kfree(ac_ptr);
				return NULL;
			}
		}
	}
	return ac_ptr;
}

P
Pekka Enberg 已提交
956
static void free_alien_cache(struct array_cache **ac_ptr)
957 958 959 960 961 962
{
	int i;

	if (!ac_ptr)
		return;
	for_each_node(i)
P
Pekka Enberg 已提交
963
	    kfree(ac_ptr[i]);
964 965 966
	kfree(ac_ptr);
}

967
static void __drain_alien_cache(struct kmem_cache *cachep,
P
Pekka Enberg 已提交
968
				struct array_cache *ac, int node)
969 970 971 972 973
{
	struct kmem_list3 *rl3 = cachep->nodelists[node];

	if (ac->avail) {
		spin_lock(&rl3->list_lock);
974
		free_block(cachep, ac->entry, ac->avail, node);
975 976 977 978 979
		ac->avail = 0;
		spin_unlock(&rl3->list_lock);
	}
}

980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996
/*
 * Called from cache_reap() to regularly drain alien caches round robin.
 */
static void reap_alien(struct kmem_cache *cachep, struct kmem_list3 *l3)
{
	int node = __get_cpu_var(reap_node);

	if (l3->alien) {
		struct array_cache *ac = l3->alien[node];
		if (ac && ac->avail) {
			spin_lock_irq(&ac->lock);
			__drain_alien_cache(cachep, ac, node);
			spin_unlock_irq(&ac->lock);
		}
	}
}

A
Andrew Morton 已提交
997 998
static void drain_alien_cache(struct kmem_cache *cachep,
				struct array_cache **alien)
999
{
P
Pekka Enberg 已提交
1000
	int i = 0;
1001 1002 1003 1004
	struct array_cache *ac;
	unsigned long flags;

	for_each_online_node(i) {
1005
		ac = alien[i];
1006 1007 1008 1009 1010 1011 1012 1013
		if (ac) {
			spin_lock_irqsave(&ac->lock, flags);
			__drain_alien_cache(cachep, ac, i);
			spin_unlock_irqrestore(&ac->lock, flags);
		}
	}
}
#else
1014

1015
#define drain_alien_cache(cachep, alien) do { } while (0)
1016
#define reap_alien(cachep, l3) do { } while (0)
1017

1018 1019 1020 1021 1022
static inline struct array_cache **alloc_alien_cache(int node, int limit)
{
	return (struct array_cache **) 0x01020304ul;
}

1023 1024 1025
static inline void free_alien_cache(struct array_cache **ac_ptr)
{
}
1026

1027 1028
#endif

L
Linus Torvalds 已提交
1029
static int __devinit cpuup_callback(struct notifier_block *nfb,
P
Pekka Enberg 已提交
1030
				    unsigned long action, void *hcpu)
L
Linus Torvalds 已提交
1031 1032
{
	long cpu = (long)hcpu;
1033
	struct kmem_cache *cachep;
1034 1035 1036
	struct kmem_list3 *l3 = NULL;
	int node = cpu_to_node(cpu);
	int memsize = sizeof(struct kmem_list3);
L
Linus Torvalds 已提交
1037 1038 1039

	switch (action) {
	case CPU_UP_PREPARE:
I
Ingo Molnar 已提交
1040
		mutex_lock(&cache_chain_mutex);
A
Andrew Morton 已提交
1041 1042
		/*
		 * We need to do this right in the beginning since
1043 1044 1045 1046 1047
		 * alloc_arraycache's are going to use this list.
		 * kmalloc_node allows us to add the slab to the right
		 * kmem_list3 and not this cpu's kmem_list3
		 */

L
Linus Torvalds 已提交
1048
		list_for_each_entry(cachep, &cache_chain, next) {
A
Andrew Morton 已提交
1049 1050
			/*
			 * Set up the size64 kmemlist for cpu before we can
1051 1052 1053 1054
			 * begin anything. Make sure some other cpu on this
			 * node has not already allocated this
			 */
			if (!cachep->nodelists[node]) {
A
Andrew Morton 已提交
1055 1056
				l3 = kmalloc_node(memsize, GFP_KERNEL, node);
				if (!l3)
1057 1058 1059
					goto bad;
				kmem_list3_init(l3);
				l3->next_reap = jiffies + REAPTIMEOUT_LIST3 +
P
Pekka Enberg 已提交
1060
				    ((unsigned long)cachep) % REAPTIMEOUT_LIST3;
1061

1062 1063 1064 1065 1066
				/*
				 * The l3s don't come and go as CPUs come and
				 * go.  cache_chain_mutex is sufficient
				 * protection here.
				 */
1067 1068
				cachep->nodelists[node] = l3;
			}
L
Linus Torvalds 已提交
1069

1070 1071
			spin_lock_irq(&cachep->nodelists[node]->list_lock);
			cachep->nodelists[node]->free_limit =
A
Andrew Morton 已提交
1072 1073
				(1 + nr_cpus_node(node)) *
				cachep->batchcount + cachep->num;
1074 1075 1076
			spin_unlock_irq(&cachep->nodelists[node]->list_lock);
		}

A
Andrew Morton 已提交
1077 1078 1079 1080
		/*
		 * Now we can go ahead with allocating the shared arrays and
		 * array caches
		 */
1081
		list_for_each_entry(cachep, &cache_chain, next) {
1082
			struct array_cache *nc;
1083 1084
			struct array_cache *shared;
			struct array_cache **alien;
1085

1086
			nc = alloc_arraycache(node, cachep->limit,
1087
						cachep->batchcount);
L
Linus Torvalds 已提交
1088 1089
			if (!nc)
				goto bad;
1090 1091 1092 1093 1094
			shared = alloc_arraycache(node,
					cachep->shared * cachep->batchcount,
					0xbaadf00d);
			if (!shared)
				goto bad;
1095

1096 1097 1098
			alien = alloc_alien_cache(node, cachep->limit);
			if (!alien)
				goto bad;
L
Linus Torvalds 已提交
1099
			cachep->array[cpu] = nc;
1100 1101 1102
			l3 = cachep->nodelists[node];
			BUG_ON(!l3);

1103 1104 1105 1106 1107 1108 1109 1110
			spin_lock_irq(&l3->list_lock);
			if (!l3->shared) {
				/*
				 * We are serialised from CPU_DEAD or
				 * CPU_UP_CANCELLED by the cpucontrol lock
				 */
				l3->shared = shared;
				shared = NULL;
1111
			}
1112 1113 1114 1115 1116 1117 1118 1119 1120
#ifdef CONFIG_NUMA
			if (!l3->alien) {
				l3->alien = alien;
				alien = NULL;
			}
#endif
			spin_unlock_irq(&l3->list_lock);
			kfree(shared);
			free_alien_cache(alien);
L
Linus Torvalds 已提交
1121
		}
I
Ingo Molnar 已提交
1122
		mutex_unlock(&cache_chain_mutex);
L
Linus Torvalds 已提交
1123 1124 1125 1126 1127 1128
		break;
	case CPU_ONLINE:
		start_cpu_timer(cpu);
		break;
#ifdef CONFIG_HOTPLUG_CPU
	case CPU_DEAD:
1129 1130 1131 1132 1133 1134 1135 1136
		/*
		 * Even if all the cpus of a node are down, we don't free the
		 * kmem_list3 of any cache. This to avoid a race between
		 * cpu_down, and a kmalloc allocation from another cpu for
		 * memory from the node of the cpu going down.  The list3
		 * structure is usually allocated from kmem_cache_create() and
		 * gets destroyed at kmem_cache_destroy().
		 */
L
Linus Torvalds 已提交
1137 1138
		/* fall thru */
	case CPU_UP_CANCELED:
I
Ingo Molnar 已提交
1139
		mutex_lock(&cache_chain_mutex);
L
Linus Torvalds 已提交
1140 1141
		list_for_each_entry(cachep, &cache_chain, next) {
			struct array_cache *nc;
1142 1143
			struct array_cache *shared;
			struct array_cache **alien;
1144
			cpumask_t mask;
L
Linus Torvalds 已提交
1145

1146
			mask = node_to_cpumask(node);
L
Linus Torvalds 已提交
1147 1148 1149
			/* cpu is dead; no one can alloc from it. */
			nc = cachep->array[cpu];
			cachep->array[cpu] = NULL;
1150 1151 1152
			l3 = cachep->nodelists[node];

			if (!l3)
1153
				goto free_array_cache;
1154

1155
			spin_lock_irq(&l3->list_lock);
1156 1157 1158 1159

			/* Free limit for this kmem_list3 */
			l3->free_limit -= cachep->batchcount;
			if (nc)
1160
				free_block(cachep, nc->entry, nc->avail, node);
1161 1162

			if (!cpus_empty(mask)) {
1163
				spin_unlock_irq(&l3->list_lock);
1164
				goto free_array_cache;
P
Pekka Enberg 已提交
1165
			}
1166

1167 1168
			shared = l3->shared;
			if (shared) {
1169
				free_block(cachep, l3->shared->entry,
P
Pekka Enberg 已提交
1170
					   l3->shared->avail, node);
1171 1172 1173
				l3->shared = NULL;
			}

1174 1175 1176 1177 1178 1179 1180 1181 1182
			alien = l3->alien;
			l3->alien = NULL;

			spin_unlock_irq(&l3->list_lock);

			kfree(shared);
			if (alien) {
				drain_alien_cache(cachep, alien);
				free_alien_cache(alien);
1183
			}
1184
free_array_cache:
L
Linus Torvalds 已提交
1185 1186
			kfree(nc);
		}
1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200
		/*
		 * In the previous loop, all the objects were freed to
		 * the respective cache's slabs,  now we can go ahead and
		 * shrink each nodelist to its limit.
		 */
		list_for_each_entry(cachep, &cache_chain, next) {
			l3 = cachep->nodelists[node];
			if (!l3)
				continue;
			spin_lock_irq(&l3->list_lock);
			/* free slabs belonging to this node */
			__node_shrink(cachep, node);
			spin_unlock_irq(&l3->list_lock);
		}
I
Ingo Molnar 已提交
1201
		mutex_unlock(&cache_chain_mutex);
L
Linus Torvalds 已提交
1202 1203 1204 1205
		break;
#endif
	}
	return NOTIFY_OK;
A
Andrew Morton 已提交
1206
bad:
I
Ingo Molnar 已提交
1207
	mutex_unlock(&cache_chain_mutex);
L
Linus Torvalds 已提交
1208 1209 1210 1211 1212
	return NOTIFY_BAD;
}

static struct notifier_block cpucache_notifier = { &cpuup_callback, NULL, 0 };

1213 1214 1215
/*
 * swap the static kmem_list3 with kmalloced memory
 */
A
Andrew Morton 已提交
1216 1217
static void init_list(struct kmem_cache *cachep, struct kmem_list3 *list,
			int nodeid)
1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231
{
	struct kmem_list3 *ptr;

	BUG_ON(cachep->nodelists[nodeid] != list);
	ptr = kmalloc_node(sizeof(struct kmem_list3), GFP_KERNEL, nodeid);
	BUG_ON(!ptr);

	local_irq_disable();
	memcpy(ptr, list, sizeof(struct kmem_list3));
	MAKE_ALL_LISTS(cachep, ptr, nodeid);
	cachep->nodelists[nodeid] = ptr;
	local_irq_enable();
}

A
Andrew Morton 已提交
1232 1233 1234
/*
 * Initialisation.  Called after the page allocator have been initialised and
 * before smp_init().
L
Linus Torvalds 已提交
1235 1236 1237 1238 1239 1240
 */
void __init kmem_cache_init(void)
{
	size_t left_over;
	struct cache_sizes *sizes;
	struct cache_names *names;
1241
	int i;
1242
	int order;
1243 1244 1245 1246 1247 1248

	for (i = 0; i < NUM_INIT_LISTS; i++) {
		kmem_list3_init(&initkmem_list3[i]);
		if (i < MAX_NUMNODES)
			cache_cache.nodelists[i] = NULL;
	}
L
Linus Torvalds 已提交
1249 1250 1251 1252 1253 1254 1255 1256 1257 1258

	/*
	 * Fragmentation resistance on low memory - only use bigger
	 * page orders on machines with more than 32MB of memory.
	 */
	if (num_physpages > (32 << 20) >> PAGE_SHIFT)
		slab_break_gfp_order = BREAK_GFP_ORDER_HI;

	/* Bootstrap is tricky, because several objects are allocated
	 * from caches that do not exist yet:
A
Andrew Morton 已提交
1259 1260 1261
	 * 1) initialize the cache_cache cache: it contains the struct
	 *    kmem_cache structures of all caches, except cache_cache itself:
	 *    cache_cache is statically allocated.
1262 1263 1264
	 *    Initially an __init data area is used for the head array and the
	 *    kmem_list3 structures, it's replaced with a kmalloc allocated
	 *    array at the end of the bootstrap.
L
Linus Torvalds 已提交
1265
	 * 2) Create the first kmalloc cache.
1266
	 *    The struct kmem_cache for the new cache is allocated normally.
1267 1268 1269
	 *    An __init data area is used for the head array.
	 * 3) Create the remaining kmalloc caches, with minimally sized
	 *    head arrays.
L
Linus Torvalds 已提交
1270 1271
	 * 4) Replace the __init data head arrays for cache_cache and the first
	 *    kmalloc cache with kmalloc allocated arrays.
1272 1273 1274
	 * 5) Replace the __init data for kmem_list3 for cache_cache and
	 *    the other cache's with kmalloc allocated memory.
	 * 6) Resize the head arrays of the kmalloc caches to their final sizes.
L
Linus Torvalds 已提交
1275 1276 1277 1278 1279 1280 1281
	 */

	/* 1) create the cache_cache */
	INIT_LIST_HEAD(&cache_chain);
	list_add(&cache_cache.next, &cache_chain);
	cache_cache.colour_off = cache_line_size();
	cache_cache.array[smp_processor_id()] = &initarray_cache.cache;
1282
	cache_cache.nodelists[numa_node_id()] = &initkmem_list3[CACHE_CACHE];
L
Linus Torvalds 已提交
1283

A
Andrew Morton 已提交
1284 1285
	cache_cache.buffer_size = ALIGN(cache_cache.buffer_size,
					cache_line_size());
L
Linus Torvalds 已提交
1286

1287 1288 1289 1290 1291 1292
	for (order = 0; order < MAX_ORDER; order++) {
		cache_estimate(order, cache_cache.buffer_size,
			cache_line_size(), 0, &left_over, &cache_cache.num);
		if (cache_cache.num)
			break;
	}
L
Linus Torvalds 已提交
1293 1294
	if (!cache_cache.num)
		BUG();
1295
	cache_cache.gfporder = order;
P
Pekka Enberg 已提交
1296 1297 1298
	cache_cache.colour = left_over / cache_cache.colour_off;
	cache_cache.slab_size = ALIGN(cache_cache.num * sizeof(kmem_bufctl_t) +
				      sizeof(struct slab), cache_line_size());
L
Linus Torvalds 已提交
1299 1300 1301 1302 1303

	/* 2+3) create the kmalloc caches */
	sizes = malloc_sizes;
	names = cache_names;

A
Andrew Morton 已提交
1304 1305 1306 1307
	/*
	 * Initialize the caches that provide memory for the array cache and the
	 * kmem_list3 structures first.  Without this, further allocations will
	 * bug.
1308 1309 1310
	 */

	sizes[INDEX_AC].cs_cachep = kmem_cache_create(names[INDEX_AC].name,
A
Andrew Morton 已提交
1311 1312 1313 1314
					sizes[INDEX_AC].cs_size,
					ARCH_KMALLOC_MINALIGN,
					ARCH_KMALLOC_FLAGS|SLAB_PANIC,
					NULL, NULL);
1315

A
Andrew Morton 已提交
1316
	if (INDEX_AC != INDEX_L3) {
1317
		sizes[INDEX_L3].cs_cachep =
A
Andrew Morton 已提交
1318 1319 1320 1321 1322 1323
			kmem_cache_create(names[INDEX_L3].name,
				sizes[INDEX_L3].cs_size,
				ARCH_KMALLOC_MINALIGN,
				ARCH_KMALLOC_FLAGS|SLAB_PANIC,
				NULL, NULL);
	}
1324

L
Linus Torvalds 已提交
1325
	while (sizes->cs_size != ULONG_MAX) {
1326 1327
		/*
		 * For performance, all the general caches are L1 aligned.
L
Linus Torvalds 已提交
1328 1329 1330
		 * This should be particularly beneficial on SMP boxes, as it
		 * eliminates "false sharing".
		 * Note for systems short on memory removing the alignment will
1331 1332
		 * allow tighter packing of the smaller caches.
		 */
A
Andrew Morton 已提交
1333
		if (!sizes->cs_cachep) {
1334
			sizes->cs_cachep = kmem_cache_create(names->name,
A
Andrew Morton 已提交
1335 1336 1337 1338 1339
					sizes->cs_size,
					ARCH_KMALLOC_MINALIGN,
					ARCH_KMALLOC_FLAGS|SLAB_PANIC,
					NULL, NULL);
		}
L
Linus Torvalds 已提交
1340 1341 1342

		/* Inc off-slab bufctl limit until the ceiling is hit. */
		if (!(OFF_SLAB(sizes->cs_cachep))) {
P
Pekka Enberg 已提交
1343
			offslab_limit = sizes->cs_size - sizeof(struct slab);
L
Linus Torvalds 已提交
1344 1345 1346 1347
			offslab_limit /= sizeof(kmem_bufctl_t);
		}

		sizes->cs_dmacachep = kmem_cache_create(names->name_dma,
A
Andrew Morton 已提交
1348 1349 1350 1351 1352
					sizes->cs_size,
					ARCH_KMALLOC_MINALIGN,
					ARCH_KMALLOC_FLAGS|SLAB_CACHE_DMA|
						SLAB_PANIC,
					NULL, NULL);
L
Linus Torvalds 已提交
1353 1354 1355 1356 1357
		sizes++;
		names++;
	}
	/* 4) Replace the bootstrap head arrays */
	{
P
Pekka Enberg 已提交
1358
		void *ptr;
1359

L
Linus Torvalds 已提交
1360
		ptr = kmalloc(sizeof(struct arraycache_init), GFP_KERNEL);
1361

L
Linus Torvalds 已提交
1362
		local_irq_disable();
1363 1364
		BUG_ON(cpu_cache_get(&cache_cache) != &initarray_cache.cache);
		memcpy(ptr, cpu_cache_get(&cache_cache),
P
Pekka Enberg 已提交
1365
		       sizeof(struct arraycache_init));
L
Linus Torvalds 已提交
1366 1367
		cache_cache.array[smp_processor_id()] = ptr;
		local_irq_enable();
1368

L
Linus Torvalds 已提交
1369
		ptr = kmalloc(sizeof(struct arraycache_init), GFP_KERNEL);
1370

L
Linus Torvalds 已提交
1371
		local_irq_disable();
1372
		BUG_ON(cpu_cache_get(malloc_sizes[INDEX_AC].cs_cachep)
P
Pekka Enberg 已提交
1373
		       != &initarray_generic.cache);
1374
		memcpy(ptr, cpu_cache_get(malloc_sizes[INDEX_AC].cs_cachep),
P
Pekka Enberg 已提交
1375
		       sizeof(struct arraycache_init));
1376
		malloc_sizes[INDEX_AC].cs_cachep->array[smp_processor_id()] =
P
Pekka Enberg 已提交
1377
		    ptr;
L
Linus Torvalds 已提交
1378 1379
		local_irq_enable();
	}
1380 1381 1382 1383 1384
	/* 5) Replace the bootstrap kmem_list3's */
	{
		int node;
		/* Replace the static kmem_list3 structures for the boot cpu */
		init_list(&cache_cache, &initkmem_list3[CACHE_CACHE],
P
Pekka Enberg 已提交
1385
			  numa_node_id());
1386 1387 1388

		for_each_online_node(node) {
			init_list(malloc_sizes[INDEX_AC].cs_cachep,
P
Pekka Enberg 已提交
1389
				  &initkmem_list3[SIZE_AC + node], node);
1390 1391 1392

			if (INDEX_AC != INDEX_L3) {
				init_list(malloc_sizes[INDEX_L3].cs_cachep,
P
Pekka Enberg 已提交
1393 1394
					  &initkmem_list3[SIZE_L3 + node],
					  node);
1395 1396 1397
			}
		}
	}
L
Linus Torvalds 已提交
1398

1399
	/* 6) resize the head arrays to their final sizes */
L
Linus Torvalds 已提交
1400
	{
1401
		struct kmem_cache *cachep;
I
Ingo Molnar 已提交
1402
		mutex_lock(&cache_chain_mutex);
L
Linus Torvalds 已提交
1403
		list_for_each_entry(cachep, &cache_chain, next)
A
Andrew Morton 已提交
1404
			enable_cpucache(cachep);
I
Ingo Molnar 已提交
1405
		mutex_unlock(&cache_chain_mutex);
L
Linus Torvalds 已提交
1406 1407 1408 1409 1410
	}

	/* Done! */
	g_cpucache_up = FULL;

A
Andrew Morton 已提交
1411 1412 1413
	/*
	 * Register a cpu startup notifier callback that initializes
	 * cpu_cache_get for all new cpus
L
Linus Torvalds 已提交
1414 1415 1416
	 */
	register_cpu_notifier(&cpucache_notifier);

A
Andrew Morton 已提交
1417 1418 1419
	/*
	 * The reap timers are started later, with a module init call: That part
	 * of the kernel is not yet operational.
L
Linus Torvalds 已提交
1420 1421 1422 1423 1424 1425 1426
	 */
}

static int __init cpucache_init(void)
{
	int cpu;

A
Andrew Morton 已提交
1427 1428
	/*
	 * Register the timers that return unneeded pages to the page allocator
L
Linus Torvalds 已提交
1429
	 */
1430
	for_each_online_cpu(cpu)
A
Andrew Morton 已提交
1431
		start_cpu_timer(cpu);
L
Linus Torvalds 已提交
1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442
	return 0;
}
__initcall(cpucache_init);

/*
 * Interface to system's page allocator. No need to hold the cache-lock.
 *
 * If we requested dmaable memory, we will get it. Even if we
 * did not request dmaable memory, we might get it, but that
 * would be relatively rare and ignorable.
 */
1443
static void *kmem_getpages(struct kmem_cache *cachep, gfp_t flags, int nodeid)
L
Linus Torvalds 已提交
1444 1445 1446 1447 1448 1449
{
	struct page *page;
	void *addr;
	int i;

	flags |= cachep->gfpflags;
1450
	page = alloc_pages_node(nodeid, flags, cachep->gfporder);
L
Linus Torvalds 已提交
1451 1452 1453 1454 1455 1456 1457 1458 1459
	if (!page)
		return NULL;
	addr = page_address(page);

	i = (1 << cachep->gfporder);
	if (cachep->flags & SLAB_RECLAIM_ACCOUNT)
		atomic_add(i, &slab_reclaim_pages);
	add_page_state(nr_slab, i);
	while (i--) {
N
Nick Piggin 已提交
1460
		__SetPageSlab(page);
L
Linus Torvalds 已提交
1461 1462 1463 1464 1465 1466 1467 1468
		page++;
	}
	return addr;
}

/*
 * Interface to system's page release.
 */
1469
static void kmem_freepages(struct kmem_cache *cachep, void *addr)
L
Linus Torvalds 已提交
1470
{
P
Pekka Enberg 已提交
1471
	unsigned long i = (1 << cachep->gfporder);
L
Linus Torvalds 已提交
1472 1473 1474 1475
	struct page *page = virt_to_page(addr);
	const unsigned long nr_freed = i;

	while (i--) {
N
Nick Piggin 已提交
1476 1477
		BUG_ON(!PageSlab(page));
		__ClearPageSlab(page);
L
Linus Torvalds 已提交
1478 1479 1480 1481 1482 1483
		page++;
	}
	sub_page_state(nr_slab, nr_freed);
	if (current->reclaim_state)
		current->reclaim_state->reclaimed_slab += nr_freed;
	free_pages((unsigned long)addr, cachep->gfporder);
P
Pekka Enberg 已提交
1484 1485
	if (cachep->flags & SLAB_RECLAIM_ACCOUNT)
		atomic_sub(1 << cachep->gfporder, &slab_reclaim_pages);
L
Linus Torvalds 已提交
1486 1487 1488 1489
}

static void kmem_rcu_free(struct rcu_head *head)
{
P
Pekka Enberg 已提交
1490
	struct slab_rcu *slab_rcu = (struct slab_rcu *)head;
1491
	struct kmem_cache *cachep = slab_rcu->cachep;
L
Linus Torvalds 已提交
1492 1493 1494 1495 1496 1497 1498 1499 1500

	kmem_freepages(cachep, slab_rcu->addr);
	if (OFF_SLAB(cachep))
		kmem_cache_free(cachep->slabp_cache, slab_rcu);
}

#if DEBUG

#ifdef CONFIG_DEBUG_PAGEALLOC
1501
static void store_stackinfo(struct kmem_cache *cachep, unsigned long *addr,
P
Pekka Enberg 已提交
1502
			    unsigned long caller)
L
Linus Torvalds 已提交
1503
{
1504
	int size = obj_size(cachep);
L
Linus Torvalds 已提交
1505

1506
	addr = (unsigned long *)&((char *)addr)[obj_offset(cachep)];
L
Linus Torvalds 已提交
1507

P
Pekka Enberg 已提交
1508
	if (size < 5 * sizeof(unsigned long))
L
Linus Torvalds 已提交
1509 1510
		return;

P
Pekka Enberg 已提交
1511 1512 1513 1514
	*addr++ = 0x12345678;
	*addr++ = caller;
	*addr++ = smp_processor_id();
	size -= 3 * sizeof(unsigned long);
L
Linus Torvalds 已提交
1515 1516 1517 1518 1519 1520 1521
	{
		unsigned long *sptr = &caller;
		unsigned long svalue;

		while (!kstack_end(sptr)) {
			svalue = *sptr++;
			if (kernel_text_address(svalue)) {
P
Pekka Enberg 已提交
1522
				*addr++ = svalue;
L
Linus Torvalds 已提交
1523 1524 1525 1526 1527 1528 1529
				size -= sizeof(unsigned long);
				if (size <= sizeof(unsigned long))
					break;
			}
		}

	}
P
Pekka Enberg 已提交
1530
	*addr++ = 0x87654321;
L
Linus Torvalds 已提交
1531 1532 1533
}
#endif

1534
static void poison_obj(struct kmem_cache *cachep, void *addr, unsigned char val)
L
Linus Torvalds 已提交
1535
{
1536 1537
	int size = obj_size(cachep);
	addr = &((char *)addr)[obj_offset(cachep)];
L
Linus Torvalds 已提交
1538 1539

	memset(addr, val, size);
P
Pekka Enberg 已提交
1540
	*(unsigned char *)(addr + size - 1) = POISON_END;
L
Linus Torvalds 已提交
1541 1542 1543 1544 1545 1546
}

static void dump_line(char *data, int offset, int limit)
{
	int i;
	printk(KERN_ERR "%03x:", offset);
A
Andrew Morton 已提交
1547
	for (i = 0; i < limit; i++)
P
Pekka Enberg 已提交
1548
		printk(" %02x", (unsigned char)data[offset + i]);
L
Linus Torvalds 已提交
1549 1550 1551 1552 1553 1554
	printk("\n");
}
#endif

#if DEBUG

1555
static void print_objinfo(struct kmem_cache *cachep, void *objp, int lines)
L
Linus Torvalds 已提交
1556 1557 1558 1559 1560 1561
{
	int i, size;
	char *realobj;

	if (cachep->flags & SLAB_RED_ZONE) {
		printk(KERN_ERR "Redzone: 0x%lx/0x%lx.\n",
A
Andrew Morton 已提交
1562 1563
			*dbg_redzone1(cachep, objp),
			*dbg_redzone2(cachep, objp));
L
Linus Torvalds 已提交
1564 1565 1566 1567
	}

	if (cachep->flags & SLAB_STORE_USER) {
		printk(KERN_ERR "Last user: [<%p>]",
A
Andrew Morton 已提交
1568
			*dbg_userword(cachep, objp));
L
Linus Torvalds 已提交
1569
		print_symbol("(%s)",
A
Andrew Morton 已提交
1570
				(unsigned long)*dbg_userword(cachep, objp));
L
Linus Torvalds 已提交
1571 1572
		printk("\n");
	}
1573 1574
	realobj = (char *)objp + obj_offset(cachep);
	size = obj_size(cachep);
P
Pekka Enberg 已提交
1575
	for (i = 0; i < size && lines; i += 16, lines--) {
L
Linus Torvalds 已提交
1576 1577
		int limit;
		limit = 16;
P
Pekka Enberg 已提交
1578 1579
		if (i + limit > size)
			limit = size - i;
L
Linus Torvalds 已提交
1580 1581 1582 1583
		dump_line(realobj, i, limit);
	}
}

1584
static void check_poison_obj(struct kmem_cache *cachep, void *objp)
L
Linus Torvalds 已提交
1585 1586 1587 1588 1589
{
	char *realobj;
	int size, i;
	int lines = 0;

1590 1591
	realobj = (char *)objp + obj_offset(cachep);
	size = obj_size(cachep);
L
Linus Torvalds 已提交
1592

P
Pekka Enberg 已提交
1593
	for (i = 0; i < size; i++) {
L
Linus Torvalds 已提交
1594
		char exp = POISON_FREE;
P
Pekka Enberg 已提交
1595
		if (i == size - 1)
L
Linus Torvalds 已提交
1596 1597 1598 1599 1600 1601
			exp = POISON_END;
		if (realobj[i] != exp) {
			int limit;
			/* Mismatch ! */
			/* Print header */
			if (lines == 0) {
P
Pekka Enberg 已提交
1602
				printk(KERN_ERR
A
Andrew Morton 已提交
1603 1604
					"Slab corruption: start=%p, len=%d\n",
					realobj, size);
L
Linus Torvalds 已提交
1605 1606 1607
				print_objinfo(cachep, objp, 0);
			}
			/* Hexdump the affected line */
P
Pekka Enberg 已提交
1608
			i = (i / 16) * 16;
L
Linus Torvalds 已提交
1609
			limit = 16;
P
Pekka Enberg 已提交
1610 1611
			if (i + limit > size)
				limit = size - i;
L
Linus Torvalds 已提交
1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623
			dump_line(realobj, i, limit);
			i += 16;
			lines++;
			/* Limit to 5 lines */
			if (lines > 5)
				break;
		}
	}
	if (lines != 0) {
		/* Print some data about the neighboring objects, if they
		 * exist:
		 */
1624
		struct slab *slabp = virt_to_slab(objp);
1625
		unsigned int objnr;
L
Linus Torvalds 已提交
1626

1627
		objnr = obj_to_index(cachep, slabp, objp);
L
Linus Torvalds 已提交
1628
		if (objnr) {
1629
			objp = index_to_obj(cachep, slabp, objnr - 1);
1630
			realobj = (char *)objp + obj_offset(cachep);
L
Linus Torvalds 已提交
1631
			printk(KERN_ERR "Prev obj: start=%p, len=%d\n",
P
Pekka Enberg 已提交
1632
			       realobj, size);
L
Linus Torvalds 已提交
1633 1634
			print_objinfo(cachep, objp, 2);
		}
P
Pekka Enberg 已提交
1635
		if (objnr + 1 < cachep->num) {
1636
			objp = index_to_obj(cachep, slabp, objnr + 1);
1637
			realobj = (char *)objp + obj_offset(cachep);
L
Linus Torvalds 已提交
1638
			printk(KERN_ERR "Next obj: start=%p, len=%d\n",
P
Pekka Enberg 已提交
1639
			       realobj, size);
L
Linus Torvalds 已提交
1640 1641 1642 1643 1644 1645
			print_objinfo(cachep, objp, 2);
		}
	}
}
#endif

1646 1647
#if DEBUG
/**
1648 1649 1650 1651 1652 1653
 * slab_destroy_objs - destroy a slab and its objects
 * @cachep: cache pointer being destroyed
 * @slabp: slab pointer being destroyed
 *
 * Call the registered destructor for each object in a slab that is being
 * destroyed.
L
Linus Torvalds 已提交
1654
 */
1655
static void slab_destroy_objs(struct kmem_cache *cachep, struct slab *slabp)
L
Linus Torvalds 已提交
1656 1657 1658
{
	int i;
	for (i = 0; i < cachep->num; i++) {
1659
		void *objp = index_to_obj(cachep, slabp, i);
L
Linus Torvalds 已提交
1660 1661 1662

		if (cachep->flags & SLAB_POISON) {
#ifdef CONFIG_DEBUG_PAGEALLOC
A
Andrew Morton 已提交
1663 1664
			if (cachep->buffer_size % PAGE_SIZE == 0 &&
					OFF_SLAB(cachep))
P
Pekka Enberg 已提交
1665
				kernel_map_pages(virt_to_page(objp),
A
Andrew Morton 已提交
1666
					cachep->buffer_size / PAGE_SIZE, 1);
L
Linus Torvalds 已提交
1667 1668 1669 1670 1671 1672 1673 1674 1675
			else
				check_poison_obj(cachep, objp);
#else
			check_poison_obj(cachep, objp);
#endif
		}
		if (cachep->flags & SLAB_RED_ZONE) {
			if (*dbg_redzone1(cachep, objp) != RED_INACTIVE)
				slab_error(cachep, "start of a freed object "
P
Pekka Enberg 已提交
1676
					   "was overwritten");
L
Linus Torvalds 已提交
1677 1678
			if (*dbg_redzone2(cachep, objp) != RED_INACTIVE)
				slab_error(cachep, "end of a freed object "
P
Pekka Enberg 已提交
1679
					   "was overwritten");
L
Linus Torvalds 已提交
1680 1681
		}
		if (cachep->dtor && !(cachep->flags & SLAB_POISON))
1682
			(cachep->dtor) (objp + obj_offset(cachep), cachep, 0);
L
Linus Torvalds 已提交
1683
	}
1684
}
L
Linus Torvalds 已提交
1685
#else
1686
static void slab_destroy_objs(struct kmem_cache *cachep, struct slab *slabp)
1687
{
L
Linus Torvalds 已提交
1688 1689 1690
	if (cachep->dtor) {
		int i;
		for (i = 0; i < cachep->num; i++) {
1691
			void *objp = index_to_obj(cachep, slabp, i);
P
Pekka Enberg 已提交
1692
			(cachep->dtor) (objp, cachep, 0);
L
Linus Torvalds 已提交
1693 1694
		}
	}
1695
}
L
Linus Torvalds 已提交
1696 1697
#endif

1698 1699 1700 1701 1702
/**
 * slab_destroy - destroy and release all objects in a slab
 * @cachep: cache pointer being destroyed
 * @slabp: slab pointer being destroyed
 *
1703
 * Destroy all the objs in a slab, and release the mem back to the system.
A
Andrew Morton 已提交
1704 1705
 * Before calling the slab must have been unlinked from the cache.  The
 * cache-lock is not held/needed.
1706
 */
1707
static void slab_destroy(struct kmem_cache *cachep, struct slab *slabp)
1708 1709 1710 1711
{
	void *addr = slabp->s_mem - slabp->colouroff;

	slab_destroy_objs(cachep, slabp);
L
Linus Torvalds 已提交
1712 1713 1714
	if (unlikely(cachep->flags & SLAB_DESTROY_BY_RCU)) {
		struct slab_rcu *slab_rcu;

P
Pekka Enberg 已提交
1715
		slab_rcu = (struct slab_rcu *)slabp;
L
Linus Torvalds 已提交
1716 1717 1718 1719 1720 1721 1722 1723 1724 1725
		slab_rcu->cachep = cachep;
		slab_rcu->addr = addr;
		call_rcu(&slab_rcu->head, kmem_rcu_free);
	} else {
		kmem_freepages(cachep, addr);
		if (OFF_SLAB(cachep))
			kmem_cache_free(cachep->slabp_cache, slabp);
	}
}

A
Andrew Morton 已提交
1726 1727 1728 1729
/*
 * For setting up all the kmem_list3s for cache whose buffer_size is same as
 * size of kmem_list3.
 */
1730
static void set_up_list3s(struct kmem_cache *cachep, int index)
1731 1732 1733 1734
{
	int node;

	for_each_online_node(node) {
P
Pekka Enberg 已提交
1735
		cachep->nodelists[node] = &initkmem_list3[index + node];
1736
		cachep->nodelists[node]->next_reap = jiffies +
P
Pekka Enberg 已提交
1737 1738
		    REAPTIMEOUT_LIST3 +
		    ((unsigned long)cachep) % REAPTIMEOUT_LIST3;
1739 1740 1741
	}
}

1742
/**
1743 1744 1745 1746 1747 1748 1749
 * calculate_slab_order - calculate size (page order) of slabs
 * @cachep: pointer to the cache that is being created
 * @size: size of objects to be created in this cache.
 * @align: required alignment for the objects.
 * @flags: slab allocation flags
 *
 * Also calculates the number of objects per slab.
1750 1751 1752 1753 1754
 *
 * This could be made much more intelligent.  For now, try to avoid using
 * high order pages for slabs.  When the gfp() functions are more friendly
 * towards high-order requests, this should be changed.
 */
A
Andrew Morton 已提交
1755
static size_t calculate_slab_order(struct kmem_cache *cachep,
R
Randy Dunlap 已提交
1756
			size_t size, size_t align, unsigned long flags)
1757 1758
{
	size_t left_over = 0;
1759
	int gfporder;
1760

A
Andrew Morton 已提交
1761
	for (gfporder = 0; gfporder <= MAX_GFP_ORDER; gfporder++) {
1762 1763 1764
		unsigned int num;
		size_t remainder;

1765
		cache_estimate(gfporder, size, align, flags, &remainder, &num);
1766 1767
		if (!num)
			continue;
1768

1769
		/* More than offslab_limit objects will cause problems */
1770
		if ((flags & CFLGS_OFF_SLAB) && num > offslab_limit)
1771 1772
			break;

1773
		/* Found something acceptable - save it away */
1774
		cachep->num = num;
1775
		cachep->gfporder = gfporder;
1776 1777
		left_over = remainder;

1778 1779 1780 1781 1782 1783 1784 1785
		/*
		 * A VFS-reclaimable slab tends to have most allocations
		 * as GFP_NOFS and we really don't want to have to be allocating
		 * higher-order pages when we are unable to shrink dcache.
		 */
		if (flags & SLAB_RECLAIM_ACCOUNT)
			break;

1786 1787 1788 1789
		/*
		 * Large number of objects is good, but very large slabs are
		 * currently bad for the gfp()s.
		 */
1790
		if (gfporder >= slab_break_gfp_order)
1791 1792
			break;

1793 1794 1795
		/*
		 * Acceptable internal fragmentation?
		 */
A
Andrew Morton 已提交
1796
		if (left_over * 8 <= (PAGE_SIZE << gfporder))
1797 1798 1799 1800 1801
			break;
	}
	return left_over;
}

1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855
static void setup_cpu_cache(struct kmem_cache *cachep)
{
	if (g_cpucache_up == FULL) {
		enable_cpucache(cachep);
		return;
	}
	if (g_cpucache_up == NONE) {
		/*
		 * Note: the first kmem_cache_create must create the cache
		 * that's used by kmalloc(24), otherwise the creation of
		 * further caches will BUG().
		 */
		cachep->array[smp_processor_id()] = &initarray_generic.cache;

		/*
		 * If the cache that's used by kmalloc(sizeof(kmem_list3)) is
		 * the first cache, then we need to set up all its list3s,
		 * otherwise the creation of further caches will BUG().
		 */
		set_up_list3s(cachep, SIZE_AC);
		if (INDEX_AC == INDEX_L3)
			g_cpucache_up = PARTIAL_L3;
		else
			g_cpucache_up = PARTIAL_AC;
	} else {
		cachep->array[smp_processor_id()] =
			kmalloc(sizeof(struct arraycache_init), GFP_KERNEL);

		if (g_cpucache_up == PARTIAL_AC) {
			set_up_list3s(cachep, SIZE_L3);
			g_cpucache_up = PARTIAL_L3;
		} else {
			int node;
			for_each_online_node(node) {
				cachep->nodelists[node] =
				    kmalloc_node(sizeof(struct kmem_list3),
						GFP_KERNEL, node);
				BUG_ON(!cachep->nodelists[node]);
				kmem_list3_init(cachep->nodelists[node]);
			}
		}
	}
	cachep->nodelists[numa_node_id()]->next_reap =
			jiffies + REAPTIMEOUT_LIST3 +
			((unsigned long)cachep) % REAPTIMEOUT_LIST3;

	cpu_cache_get(cachep)->avail = 0;
	cpu_cache_get(cachep)->limit = BOOT_CPUCACHE_ENTRIES;
	cpu_cache_get(cachep)->batchcount = 1;
	cpu_cache_get(cachep)->touched = 0;
	cachep->batchcount = 1;
	cachep->limit = BOOT_CPUCACHE_ENTRIES;
}

L
Linus Torvalds 已提交
1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870
/**
 * kmem_cache_create - Create a cache.
 * @name: A string which is used in /proc/slabinfo to identify this cache.
 * @size: The size of objects to be created in this cache.
 * @align: The required alignment for the objects.
 * @flags: SLAB flags
 * @ctor: A constructor for the objects.
 * @dtor: A destructor for the objects.
 *
 * Returns a ptr to the cache on success, NULL on failure.
 * Cannot be called within a int, but can be interrupted.
 * The @ctor is run when new pages are allocated by the cache
 * and the @dtor is run before the pages are handed back.
 *
 * @name must be valid until the cache is destroyed. This implies that
A
Andrew Morton 已提交
1871 1872
 * the module calling this has to destroy the cache before getting unloaded.
 *
L
Linus Torvalds 已提交
1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884
 * The flags are
 *
 * %SLAB_POISON - Poison the slab with a known test pattern (a5a5a5a5)
 * to catch references to uninitialised memory.
 *
 * %SLAB_RED_ZONE - Insert `Red' zones around the allocated memory to check
 * for buffer overruns.
 *
 * %SLAB_HWCACHE_ALIGN - Align the objects in this cache to a hardware
 * cacheline.  This can be beneficial if you're counting cycles as closely
 * as davem.
 */
1885
struct kmem_cache *
L
Linus Torvalds 已提交
1886
kmem_cache_create (const char *name, size_t size, size_t align,
A
Andrew Morton 已提交
1887 1888
	unsigned long flags,
	void (*ctor)(void*, struct kmem_cache *, unsigned long),
1889
	void (*dtor)(void*, struct kmem_cache *, unsigned long))
L
Linus Torvalds 已提交
1890 1891
{
	size_t left_over, slab_size, ralign;
1892
	struct kmem_cache *cachep = NULL;
1893
	struct list_head *p;
L
Linus Torvalds 已提交
1894 1895 1896 1897

	/*
	 * Sanity checks... these are all serious usage bugs.
	 */
A
Andrew Morton 已提交
1898
	if (!name || in_interrupt() || (size < BYTES_PER_WORD) ||
P
Pekka Enberg 已提交
1899
	    (size > (1 << MAX_OBJ_ORDER) * PAGE_SIZE) || (dtor && !ctor)) {
A
Andrew Morton 已提交
1900 1901
		printk(KERN_ERR "%s: Early error in slab %s\n", __FUNCTION__,
				name);
P
Pekka Enberg 已提交
1902 1903
		BUG();
	}
L
Linus Torvalds 已提交
1904

1905 1906 1907 1908 1909 1910
	/*
	 * Prevent CPUs from coming and going.
	 * lock_cpu_hotplug() nests outside cache_chain_mutex
	 */
	lock_cpu_hotplug();

I
Ingo Molnar 已提交
1911
	mutex_lock(&cache_chain_mutex);
1912 1913

	list_for_each(p, &cache_chain) {
1914
		struct kmem_cache *pc = list_entry(p, struct kmem_cache, next);
1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928
		mm_segment_t old_fs = get_fs();
		char tmp;
		int res;

		/*
		 * This happens when the module gets unloaded and doesn't
		 * destroy its slab cache and no-one else reuses the vmalloc
		 * area of the module.  Print a warning.
		 */
		set_fs(KERNEL_DS);
		res = __get_user(tmp, pc->name);
		set_fs(old_fs);
		if (res) {
			printk("SLAB: cache with size %d has lost its name\n",
1929
			       pc->buffer_size);
1930 1931 1932
			continue;
		}

P
Pekka Enberg 已提交
1933
		if (!strcmp(pc->name, name)) {
1934 1935 1936 1937 1938 1939
			printk("kmem_cache_create: duplicate cache %s\n", name);
			dump_stack();
			goto oops;
		}
	}

L
Linus Torvalds 已提交
1940 1941 1942 1943 1944
#if DEBUG
	WARN_ON(strchr(name, ' '));	/* It confuses parsers */
	if ((flags & SLAB_DEBUG_INITIAL) && !ctor) {
		/* No constructor, but inital state check requested */
		printk(KERN_ERR "%s: No con, but init state check "
P
Pekka Enberg 已提交
1945
		       "requested - %s\n", __FUNCTION__, name);
L
Linus Torvalds 已提交
1946 1947 1948 1949 1950 1951 1952 1953 1954
		flags &= ~SLAB_DEBUG_INITIAL;
	}
#if FORCED_DEBUG
	/*
	 * Enable redzoning and last user accounting, except for caches with
	 * large objects, if the increased size would increase the object size
	 * above the next power of two: caches with object sizes just above a
	 * power of two have a significant amount of internal fragmentation.
	 */
A
Andrew Morton 已提交
1955
	if (size < 4096 || fls(size - 1) == fls(size-1 + 3 * BYTES_PER_WORD))
P
Pekka Enberg 已提交
1956
		flags |= SLAB_RED_ZONE | SLAB_STORE_USER;
L
Linus Torvalds 已提交
1957 1958 1959 1960 1961 1962 1963 1964 1965 1966
	if (!(flags & SLAB_DESTROY_BY_RCU))
		flags |= SLAB_POISON;
#endif
	if (flags & SLAB_DESTROY_BY_RCU)
		BUG_ON(flags & SLAB_POISON);
#endif
	if (flags & SLAB_DESTROY_BY_RCU)
		BUG_ON(dtor);

	/*
A
Andrew Morton 已提交
1967 1968
	 * Always checks flags, a caller might be expecting debug support which
	 * isn't available.
L
Linus Torvalds 已提交
1969 1970 1971 1972
	 */
	if (flags & ~CREATE_MASK)
		BUG();

A
Andrew Morton 已提交
1973 1974
	/*
	 * Check that size is in terms of words.  This is needed to avoid
L
Linus Torvalds 已提交
1975 1976 1977
	 * unaligned accesses for some archs when redzoning is used, and makes
	 * sure any on-slab bufctl's are also correctly aligned.
	 */
P
Pekka Enberg 已提交
1978 1979 1980
	if (size & (BYTES_PER_WORD - 1)) {
		size += (BYTES_PER_WORD - 1);
		size &= ~(BYTES_PER_WORD - 1);
L
Linus Torvalds 已提交
1981 1982
	}

A
Andrew Morton 已提交
1983 1984
	/* calculate the final buffer alignment: */

L
Linus Torvalds 已提交
1985 1986
	/* 1) arch recommendation: can be overridden for debug */
	if (flags & SLAB_HWCACHE_ALIGN) {
A
Andrew Morton 已提交
1987 1988 1989 1990
		/*
		 * Default alignment: as specified by the arch code.  Except if
		 * an object is really small, then squeeze multiple objects into
		 * one cacheline.
L
Linus Torvalds 已提交
1991 1992
		 */
		ralign = cache_line_size();
P
Pekka Enberg 已提交
1993
		while (size <= ralign / 2)
L
Linus Torvalds 已提交
1994 1995 1996 1997 1998 1999 2000 2001
			ralign /= 2;
	} else {
		ralign = BYTES_PER_WORD;
	}
	/* 2) arch mandated alignment: disables debug if necessary */
	if (ralign < ARCH_SLAB_MINALIGN) {
		ralign = ARCH_SLAB_MINALIGN;
		if (ralign > BYTES_PER_WORD)
P
Pekka Enberg 已提交
2002
			flags &= ~(SLAB_RED_ZONE | SLAB_STORE_USER);
L
Linus Torvalds 已提交
2003 2004 2005 2006 2007
	}
	/* 3) caller mandated alignment: disables debug if necessary */
	if (ralign < align) {
		ralign = align;
		if (ralign > BYTES_PER_WORD)
P
Pekka Enberg 已提交
2008
			flags &= ~(SLAB_RED_ZONE | SLAB_STORE_USER);
L
Linus Torvalds 已提交
2009
	}
A
Andrew Morton 已提交
2010 2011
	/*
	 * 4) Store it. Note that the debug code below can reduce
L
Linus Torvalds 已提交
2012 2013 2014 2015 2016
	 *    the alignment to BYTES_PER_WORD.
	 */
	align = ralign;

	/* Get cache's description obj. */
P
Pekka Enberg 已提交
2017
	cachep = kmem_cache_zalloc(&cache_cache, SLAB_KERNEL);
L
Linus Torvalds 已提交
2018
	if (!cachep)
2019
		goto oops;
L
Linus Torvalds 已提交
2020 2021

#if DEBUG
2022
	cachep->obj_size = size;
L
Linus Torvalds 已提交
2023 2024 2025 2026 2027 2028

	if (flags & SLAB_RED_ZONE) {
		/* redzoning only works with word aligned caches */
		align = BYTES_PER_WORD;

		/* add space for red zone words */
2029
		cachep->obj_offset += BYTES_PER_WORD;
P
Pekka Enberg 已提交
2030
		size += 2 * BYTES_PER_WORD;
L
Linus Torvalds 已提交
2031 2032 2033 2034 2035 2036 2037 2038 2039 2040
	}
	if (flags & SLAB_STORE_USER) {
		/* user store requires word alignment and
		 * one word storage behind the end of the real
		 * object.
		 */
		align = BYTES_PER_WORD;
		size += BYTES_PER_WORD;
	}
#if FORCED_DEBUG && defined(CONFIG_DEBUG_PAGEALLOC)
P
Pekka Enberg 已提交
2041
	if (size >= malloc_sizes[INDEX_L3 + 1].cs_size
2042 2043
	    && cachep->obj_size > cache_line_size() && size < PAGE_SIZE) {
		cachep->obj_offset += PAGE_SIZE - size;
L
Linus Torvalds 已提交
2044 2045 2046 2047 2048 2049
		size = PAGE_SIZE;
	}
#endif
#endif

	/* Determine if the slab management is 'on' or 'off' slab. */
P
Pekka Enberg 已提交
2050
	if (size >= (PAGE_SIZE >> 3))
L
Linus Torvalds 已提交
2051 2052 2053 2054 2055 2056 2057 2058
		/*
		 * Size is large, assume best to place the slab management obj
		 * off-slab (should allow better packing of objs).
		 */
		flags |= CFLGS_OFF_SLAB;

	size = ALIGN(size, align);

2059
	left_over = calculate_slab_order(cachep, size, align, flags);
L
Linus Torvalds 已提交
2060 2061 2062 2063 2064

	if (!cachep->num) {
		printk("kmem_cache_create: couldn't create cache %s.\n", name);
		kmem_cache_free(&cache_cache, cachep);
		cachep = NULL;
2065
		goto oops;
L
Linus Torvalds 已提交
2066
	}
P
Pekka Enberg 已提交
2067 2068
	slab_size = ALIGN(cachep->num * sizeof(kmem_bufctl_t)
			  + sizeof(struct slab), align);
L
Linus Torvalds 已提交
2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080

	/*
	 * If the slab has been placed off-slab, and we have enough space then
	 * move it on-slab. This is at the expense of any extra colouring.
	 */
	if (flags & CFLGS_OFF_SLAB && left_over >= slab_size) {
		flags &= ~CFLGS_OFF_SLAB;
		left_over -= slab_size;
	}

	if (flags & CFLGS_OFF_SLAB) {
		/* really off slab. No need for manual alignment */
P
Pekka Enberg 已提交
2081 2082
		slab_size =
		    cachep->num * sizeof(kmem_bufctl_t) + sizeof(struct slab);
L
Linus Torvalds 已提交
2083 2084 2085 2086 2087 2088
	}

	cachep->colour_off = cache_line_size();
	/* Offset must be a multiple of the alignment. */
	if (cachep->colour_off < align)
		cachep->colour_off = align;
P
Pekka Enberg 已提交
2089
	cachep->colour = left_over / cachep->colour_off;
L
Linus Torvalds 已提交
2090 2091 2092 2093 2094
	cachep->slab_size = slab_size;
	cachep->flags = flags;
	cachep->gfpflags = 0;
	if (flags & SLAB_CACHE_DMA)
		cachep->gfpflags |= GFP_DMA;
2095
	cachep->buffer_size = size;
L
Linus Torvalds 已提交
2096 2097

	if (flags & CFLGS_OFF_SLAB)
2098
		cachep->slabp_cache = kmem_find_general_cachep(slab_size, 0u);
L
Linus Torvalds 已提交
2099 2100 2101 2102 2103
	cachep->ctor = ctor;
	cachep->dtor = dtor;
	cachep->name = name;


2104
	setup_cpu_cache(cachep);
L
Linus Torvalds 已提交
2105 2106 2107

	/* cache setup completed, link it into the list */
	list_add(&cachep->next, &cache_chain);
A
Andrew Morton 已提交
2108
oops:
L
Linus Torvalds 已提交
2109 2110
	if (!cachep && (flags & SLAB_PANIC))
		panic("kmem_cache_create(): failed to create slab `%s'\n",
P
Pekka Enberg 已提交
2111
		      name);
I
Ingo Molnar 已提交
2112
	mutex_unlock(&cache_chain_mutex);
2113
	unlock_cpu_hotplug();
L
Linus Torvalds 已提交
2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128
	return cachep;
}
EXPORT_SYMBOL(kmem_cache_create);

#if DEBUG
static void check_irq_off(void)
{
	BUG_ON(!irqs_disabled());
}

static void check_irq_on(void)
{
	BUG_ON(irqs_disabled());
}

2129
static void check_spinlock_acquired(struct kmem_cache *cachep)
L
Linus Torvalds 已提交
2130 2131 2132
{
#ifdef CONFIG_SMP
	check_irq_off();
2133
	assert_spin_locked(&cachep->nodelists[numa_node_id()]->list_lock);
L
Linus Torvalds 已提交
2134 2135
#endif
}
2136

2137
static void check_spinlock_acquired_node(struct kmem_cache *cachep, int node)
2138 2139 2140 2141 2142 2143 2144
{
#ifdef CONFIG_SMP
	check_irq_off();
	assert_spin_locked(&cachep->nodelists[node]->list_lock);
#endif
}

L
Linus Torvalds 已提交
2145 2146 2147 2148
#else
#define check_irq_off()	do { } while(0)
#define check_irq_on()	do { } while(0)
#define check_spinlock_acquired(x) do { } while(0)
2149
#define check_spinlock_acquired_node(x, y) do { } while(0)
L
Linus Torvalds 已提交
2150 2151
#endif

2152 2153 2154 2155
static void drain_array(struct kmem_cache *cachep, struct kmem_list3 *l3,
			struct array_cache *ac,
			int force, int node);

L
Linus Torvalds 已提交
2156 2157
static void do_drain(void *arg)
{
A
Andrew Morton 已提交
2158
	struct kmem_cache *cachep = arg;
L
Linus Torvalds 已提交
2159
	struct array_cache *ac;
2160
	int node = numa_node_id();
L
Linus Torvalds 已提交
2161 2162

	check_irq_off();
2163
	ac = cpu_cache_get(cachep);
2164 2165 2166
	spin_lock(&cachep->nodelists[node]->list_lock);
	free_block(cachep, ac->entry, ac->avail, node);
	spin_unlock(&cachep->nodelists[node]->list_lock);
L
Linus Torvalds 已提交
2167 2168 2169
	ac->avail = 0;
}

2170
static void drain_cpu_caches(struct kmem_cache *cachep)
L
Linus Torvalds 已提交
2171
{
2172 2173 2174
	struct kmem_list3 *l3;
	int node;

A
Andrew Morton 已提交
2175
	on_each_cpu(do_drain, cachep, 1, 1);
L
Linus Torvalds 已提交
2176
	check_irq_on();
P
Pekka Enberg 已提交
2177
	for_each_online_node(node) {
2178 2179
		l3 = cachep->nodelists[node];
		if (l3) {
2180
			drain_array(cachep, l3, l3->shared, 1, node);
2181
			if (l3->alien)
2182
				drain_alien_cache(cachep, l3->alien);
2183 2184
		}
	}
L
Linus Torvalds 已提交
2185 2186
}

2187
static int __node_shrink(struct kmem_cache *cachep, int node)
L
Linus Torvalds 已提交
2188 2189
{
	struct slab *slabp;
2190
	struct kmem_list3 *l3 = cachep->nodelists[node];
L
Linus Torvalds 已提交
2191 2192
	int ret;

2193
	for (;;) {
L
Linus Torvalds 已提交
2194 2195
		struct list_head *p;

2196 2197
		p = l3->slabs_free.prev;
		if (p == &l3->slabs_free)
L
Linus Torvalds 已提交
2198 2199
			break;

2200
		slabp = list_entry(l3->slabs_free.prev, struct slab, list);
L
Linus Torvalds 已提交
2201 2202 2203 2204 2205 2206
#if DEBUG
		if (slabp->inuse)
			BUG();
#endif
		list_del(&slabp->list);

2207 2208
		l3->free_objects -= cachep->num;
		spin_unlock_irq(&l3->list_lock);
L
Linus Torvalds 已提交
2209
		slab_destroy(cachep, slabp);
2210
		spin_lock_irq(&l3->list_lock);
L
Linus Torvalds 已提交
2211
	}
P
Pekka Enberg 已提交
2212
	ret = !list_empty(&l3->slabs_full) || !list_empty(&l3->slabs_partial);
L
Linus Torvalds 已提交
2213 2214 2215
	return ret;
}

2216
static int __cache_shrink(struct kmem_cache *cachep)
2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234
{
	int ret = 0, i = 0;
	struct kmem_list3 *l3;

	drain_cpu_caches(cachep);

	check_irq_on();
	for_each_online_node(i) {
		l3 = cachep->nodelists[i];
		if (l3) {
			spin_lock_irq(&l3->list_lock);
			ret += __node_shrink(cachep, i);
			spin_unlock_irq(&l3->list_lock);
		}
	}
	return (ret ? 1 : 0);
}

L
Linus Torvalds 已提交
2235 2236 2237 2238 2239 2240 2241
/**
 * kmem_cache_shrink - Shrink a cache.
 * @cachep: The cache to shrink.
 *
 * Releases as many slabs as possible for a cache.
 * To help debugging, a zero exit status indicates all slabs were released.
 */
2242
int kmem_cache_shrink(struct kmem_cache *cachep)
L
Linus Torvalds 已提交
2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254
{
	if (!cachep || in_interrupt())
		BUG();

	return __cache_shrink(cachep);
}
EXPORT_SYMBOL(kmem_cache_shrink);

/**
 * kmem_cache_destroy - delete a cache
 * @cachep: the cache to destroy
 *
2255
 * Remove a struct kmem_cache object from the slab cache.
L
Linus Torvalds 已提交
2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267
 * Returns 0 on success.
 *
 * It is expected this function will be called by a module when it is
 * unloaded.  This will remove the cache completely, and avoid a duplicate
 * cache being allocated each time a module is loaded and unloaded, if the
 * module doesn't have persistent in-kernel storage across loads and unloads.
 *
 * The cache must be empty before calling this function.
 *
 * The caller must guarantee that noone will allocate memory from the cache
 * during the kmem_cache_destroy().
 */
2268
int kmem_cache_destroy(struct kmem_cache *cachep)
L
Linus Torvalds 已提交
2269 2270
{
	int i;
2271
	struct kmem_list3 *l3;
L
Linus Torvalds 已提交
2272 2273 2274 2275 2276 2277 2278 2279

	if (!cachep || in_interrupt())
		BUG();

	/* Don't let CPUs to come and go */
	lock_cpu_hotplug();

	/* Find the cache in the chain of caches. */
I
Ingo Molnar 已提交
2280
	mutex_lock(&cache_chain_mutex);
L
Linus Torvalds 已提交
2281 2282 2283 2284
	/*
	 * the chain is never empty, cache_cache is never destroyed
	 */
	list_del(&cachep->next);
I
Ingo Molnar 已提交
2285
	mutex_unlock(&cache_chain_mutex);
L
Linus Torvalds 已提交
2286 2287 2288

	if (__cache_shrink(cachep)) {
		slab_error(cachep, "Can't free all objects");
I
Ingo Molnar 已提交
2289
		mutex_lock(&cache_chain_mutex);
P
Pekka Enberg 已提交
2290
		list_add(&cachep->next, &cache_chain);
I
Ingo Molnar 已提交
2291
		mutex_unlock(&cache_chain_mutex);
L
Linus Torvalds 已提交
2292 2293 2294 2295 2296
		unlock_cpu_hotplug();
		return 1;
	}

	if (unlikely(cachep->flags & SLAB_DESTROY_BY_RCU))
2297
		synchronize_rcu();
L
Linus Torvalds 已提交
2298

2299
	for_each_online_cpu(i)
P
Pekka Enberg 已提交
2300
	    kfree(cachep->array[i]);
L
Linus Torvalds 已提交
2301 2302

	/* NUMA: free the list3 structures */
2303
	for_each_online_node(i) {
A
Andrew Morton 已提交
2304 2305
		l3 = cachep->nodelists[i];
		if (l3) {
2306 2307 2308 2309 2310
			kfree(l3->shared);
			free_alien_cache(l3->alien);
			kfree(l3);
		}
	}
L
Linus Torvalds 已提交
2311 2312 2313 2314 2315 2316 2317
	kmem_cache_free(&cache_cache, cachep);
	unlock_cpu_hotplug();
	return 0;
}
EXPORT_SYMBOL(kmem_cache_destroy);

/* Get the memory for a slab management obj. */
2318
static struct slab *alloc_slabmgmt(struct kmem_cache *cachep, void *objp,
P
Pekka Enberg 已提交
2319
				   int colour_off, gfp_t local_flags)
L
Linus Torvalds 已提交
2320 2321
{
	struct slab *slabp;
P
Pekka Enberg 已提交
2322

L
Linus Torvalds 已提交
2323 2324 2325 2326 2327 2328
	if (OFF_SLAB(cachep)) {
		/* Slab management obj is off-slab. */
		slabp = kmem_cache_alloc(cachep->slabp_cache, local_flags);
		if (!slabp)
			return NULL;
	} else {
P
Pekka Enberg 已提交
2329
		slabp = objp + colour_off;
L
Linus Torvalds 已提交
2330 2331 2332 2333
		colour_off += cachep->slab_size;
	}
	slabp->inuse = 0;
	slabp->colouroff = colour_off;
P
Pekka Enberg 已提交
2334
	slabp->s_mem = objp + colour_off;
L
Linus Torvalds 已提交
2335 2336 2337 2338 2339
	return slabp;
}

static inline kmem_bufctl_t *slab_bufctl(struct slab *slabp)
{
P
Pekka Enberg 已提交
2340
	return (kmem_bufctl_t *) (slabp + 1);
L
Linus Torvalds 已提交
2341 2342
}

2343
static void cache_init_objs(struct kmem_cache *cachep,
P
Pekka Enberg 已提交
2344
			    struct slab *slabp, unsigned long ctor_flags)
L
Linus Torvalds 已提交
2345 2346 2347 2348
{
	int i;

	for (i = 0; i < cachep->num; i++) {
2349
		void *objp = index_to_obj(cachep, slabp, i);
L
Linus Torvalds 已提交
2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361
#if DEBUG
		/* need to poison the objs? */
		if (cachep->flags & SLAB_POISON)
			poison_obj(cachep, objp, POISON_FREE);
		if (cachep->flags & SLAB_STORE_USER)
			*dbg_userword(cachep, objp) = NULL;

		if (cachep->flags & SLAB_RED_ZONE) {
			*dbg_redzone1(cachep, objp) = RED_INACTIVE;
			*dbg_redzone2(cachep, objp) = RED_INACTIVE;
		}
		/*
A
Andrew Morton 已提交
2362 2363 2364
		 * Constructors are not allowed to allocate memory from the same
		 * cache which they are a constructor for.  Otherwise, deadlock.
		 * They must also be threaded.
L
Linus Torvalds 已提交
2365 2366
		 */
		if (cachep->ctor && !(cachep->flags & SLAB_POISON))
2367
			cachep->ctor(objp + obj_offset(cachep), cachep,
P
Pekka Enberg 已提交
2368
				     ctor_flags);
L
Linus Torvalds 已提交
2369 2370 2371 2372

		if (cachep->flags & SLAB_RED_ZONE) {
			if (*dbg_redzone2(cachep, objp) != RED_INACTIVE)
				slab_error(cachep, "constructor overwrote the"
P
Pekka Enberg 已提交
2373
					   " end of an object");
L
Linus Torvalds 已提交
2374 2375
			if (*dbg_redzone1(cachep, objp) != RED_INACTIVE)
				slab_error(cachep, "constructor overwrote the"
P
Pekka Enberg 已提交
2376
					   " start of an object");
L
Linus Torvalds 已提交
2377
		}
A
Andrew Morton 已提交
2378 2379
		if ((cachep->buffer_size % PAGE_SIZE) == 0 &&
			    OFF_SLAB(cachep) && cachep->flags & SLAB_POISON)
P
Pekka Enberg 已提交
2380
			kernel_map_pages(virt_to_page(objp),
2381
					 cachep->buffer_size / PAGE_SIZE, 0);
L
Linus Torvalds 已提交
2382 2383 2384 2385
#else
		if (cachep->ctor)
			cachep->ctor(objp, cachep, ctor_flags);
#endif
P
Pekka Enberg 已提交
2386
		slab_bufctl(slabp)[i] = i + 1;
L
Linus Torvalds 已提交
2387
	}
P
Pekka Enberg 已提交
2388
	slab_bufctl(slabp)[i - 1] = BUFCTL_END;
L
Linus Torvalds 已提交
2389 2390 2391
	slabp->free = 0;
}

2392
static void kmem_flagcheck(struct kmem_cache *cachep, gfp_t flags)
L
Linus Torvalds 已提交
2393
{
A
Andrew Morton 已提交
2394 2395 2396 2397
	if (flags & SLAB_DMA)
		BUG_ON(!(cachep->gfpflags & GFP_DMA));
	else
		BUG_ON(cachep->gfpflags & GFP_DMA);
L
Linus Torvalds 已提交
2398 2399
}

A
Andrew Morton 已提交
2400 2401
static void *slab_get_obj(struct kmem_cache *cachep, struct slab *slabp,
				int nodeid)
2402
{
2403
	void *objp = index_to_obj(cachep, slabp, slabp->free);
2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416
	kmem_bufctl_t next;

	slabp->inuse++;
	next = slab_bufctl(slabp)[slabp->free];
#if DEBUG
	slab_bufctl(slabp)[slabp->free] = BUFCTL_FREE;
	WARN_ON(slabp->nodeid != nodeid);
#endif
	slabp->free = next;

	return objp;
}

A
Andrew Morton 已提交
2417 2418
static void slab_put_obj(struct kmem_cache *cachep, struct slab *slabp,
				void *objp, int nodeid)
2419
{
2420
	unsigned int objnr = obj_to_index(cachep, slabp, objp);
2421 2422 2423 2424 2425

#if DEBUG
	/* Verify that the slab belongs to the intended node */
	WARN_ON(slabp->nodeid != nodeid);

2426
	if (slab_bufctl(slabp)[objnr] + 1 <= SLAB_LIMIT + 1) {
2427
		printk(KERN_ERR "slab: double free detected in cache "
A
Andrew Morton 已提交
2428
				"'%s', objp %p\n", cachep->name, objp);
2429 2430 2431 2432 2433 2434 2435 2436
		BUG();
	}
#endif
	slab_bufctl(slabp)[objnr] = slabp->free;
	slabp->free = objnr;
	slabp->inuse--;
}

A
Andrew Morton 已提交
2437 2438
static void set_slab_attr(struct kmem_cache *cachep, struct slab *slabp,
			void *objp)
L
Linus Torvalds 已提交
2439 2440 2441 2442 2443 2444
{
	int i;
	struct page *page;

	/* Nasty!!!!!! I hope this is OK. */
	page = virt_to_page(objp);
2445 2446 2447 2448

	i = 1;
	if (likely(!PageCompound(page)))
		i <<= cachep->gfporder;
L
Linus Torvalds 已提交
2449
	do {
2450 2451
		page_set_cache(page, cachep);
		page_set_slab(page, slabp);
L
Linus Torvalds 已提交
2452 2453 2454 2455 2456 2457 2458 2459
		page++;
	} while (--i);
}

/*
 * Grow (by 1) the number of slabs within a cache.  This is called by
 * kmem_cache_alloc() when there are no active objs left in a cache.
 */
2460
static int cache_grow(struct kmem_cache *cachep, gfp_t flags, int nodeid)
L
Linus Torvalds 已提交
2461
{
P
Pekka Enberg 已提交
2462 2463 2464 2465 2466
	struct slab *slabp;
	void *objp;
	size_t offset;
	gfp_t local_flags;
	unsigned long ctor_flags;
2467
	struct kmem_list3 *l3;
L
Linus Torvalds 已提交
2468

A
Andrew Morton 已提交
2469 2470 2471
	/*
	 * Be lazy and only check for valid flags here,  keeping it out of the
	 * critical path in kmem_cache_alloc().
L
Linus Torvalds 已提交
2472
	 */
P
Pekka Enberg 已提交
2473
	if (flags & ~(SLAB_DMA | SLAB_LEVEL_MASK | SLAB_NO_GROW))
L
Linus Torvalds 已提交
2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486
		BUG();
	if (flags & SLAB_NO_GROW)
		return 0;

	ctor_flags = SLAB_CTOR_CONSTRUCTOR;
	local_flags = (flags & SLAB_LEVEL_MASK);
	if (!(local_flags & __GFP_WAIT))
		/*
		 * Not allowed to sleep.  Need to tell a constructor about
		 * this - it might need to know...
		 */
		ctor_flags |= SLAB_CTOR_ATOMIC;

2487
	/* Take the l3 list lock to change the colour_next on this node */
L
Linus Torvalds 已提交
2488
	check_irq_off();
2489 2490
	l3 = cachep->nodelists[nodeid];
	spin_lock(&l3->list_lock);
L
Linus Torvalds 已提交
2491 2492

	/* Get colour for the slab, and cal the next value. */
2493 2494 2495 2496 2497
	offset = l3->colour_next;
	l3->colour_next++;
	if (l3->colour_next >= cachep->colour)
		l3->colour_next = 0;
	spin_unlock(&l3->list_lock);
L
Linus Torvalds 已提交
2498

2499
	offset *= cachep->colour_off;
L
Linus Torvalds 已提交
2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511

	if (local_flags & __GFP_WAIT)
		local_irq_enable();

	/*
	 * The test for missing atomic flag is performed here, rather than
	 * the more obvious place, simply to reduce the critical path length
	 * in kmem_cache_alloc(). If a caller is seriously mis-behaving they
	 * will eventually be caught here (where it matters).
	 */
	kmem_flagcheck(cachep, flags);

A
Andrew Morton 已提交
2512 2513 2514
	/*
	 * Get mem for the objs.  Attempt to allocate a physical page from
	 * 'nodeid'.
2515
	 */
A
Andrew Morton 已提交
2516 2517
	objp = kmem_getpages(cachep, flags, nodeid);
	if (!objp)
L
Linus Torvalds 已提交
2518 2519 2520
		goto failed;

	/* Get slab management. */
A
Andrew Morton 已提交
2521 2522
	slabp = alloc_slabmgmt(cachep, objp, offset, local_flags);
	if (!slabp)
L
Linus Torvalds 已提交
2523 2524
		goto opps1;

2525
	slabp->nodeid = nodeid;
L
Linus Torvalds 已提交
2526 2527 2528 2529 2530 2531 2532
	set_slab_attr(cachep, slabp, objp);

	cache_init_objs(cachep, slabp, ctor_flags);

	if (local_flags & __GFP_WAIT)
		local_irq_disable();
	check_irq_off();
2533
	spin_lock(&l3->list_lock);
L
Linus Torvalds 已提交
2534 2535

	/* Make slab active. */
2536
	list_add_tail(&slabp->list, &(l3->slabs_free));
L
Linus Torvalds 已提交
2537
	STATS_INC_GROWN(cachep);
2538 2539
	l3->free_objects += cachep->num;
	spin_unlock(&l3->list_lock);
L
Linus Torvalds 已提交
2540
	return 1;
A
Andrew Morton 已提交
2541
opps1:
L
Linus Torvalds 已提交
2542
	kmem_freepages(cachep, objp);
A
Andrew Morton 已提交
2543
failed:
L
Linus Torvalds 已提交
2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562
	if (local_flags & __GFP_WAIT)
		local_irq_disable();
	return 0;
}

#if DEBUG

/*
 * Perform extra freeing checks:
 * - detect bad pointers.
 * - POISON/RED_ZONE checking
 * - destructor calls, for caches with POISON+dtor
 */
static void kfree_debugcheck(const void *objp)
{
	struct page *page;

	if (!virt_addr_valid(objp)) {
		printk(KERN_ERR "kfree_debugcheck: out of range ptr %lxh.\n",
P
Pekka Enberg 已提交
2563 2564
		       (unsigned long)objp);
		BUG();
L
Linus Torvalds 已提交
2565 2566 2567
	}
	page = virt_to_page(objp);
	if (!PageSlab(page)) {
P
Pekka Enberg 已提交
2568 2569
		printk(KERN_ERR "kfree_debugcheck: bad ptr %lxh.\n",
		       (unsigned long)objp);
L
Linus Torvalds 已提交
2570 2571 2572 2573
		BUG();
	}
}

2574
static void *cache_free_debugcheck(struct kmem_cache *cachep, void *objp,
P
Pekka Enberg 已提交
2575
				   void *caller)
L
Linus Torvalds 已提交
2576 2577 2578 2579 2580
{
	struct page *page;
	unsigned int objnr;
	struct slab *slabp;

2581
	objp -= obj_offset(cachep);
L
Linus Torvalds 已提交
2582 2583 2584
	kfree_debugcheck(objp);
	page = virt_to_page(objp);

2585
	if (page_get_cache(page) != cachep) {
A
Andrew Morton 已提交
2586 2587
		printk(KERN_ERR "mismatch in kmem_cache_free: expected "
				"cache %p, got %p\n",
P
Pekka Enberg 已提交
2588
		       page_get_cache(page), cachep);
L
Linus Torvalds 已提交
2589
		printk(KERN_ERR "%p is %s.\n", cachep, cachep->name);
P
Pekka Enberg 已提交
2590 2591
		printk(KERN_ERR "%p is %s.\n", page_get_cache(page),
		       page_get_cache(page)->name);
L
Linus Torvalds 已提交
2592 2593
		WARN_ON(1);
	}
2594
	slabp = page_get_slab(page);
L
Linus Torvalds 已提交
2595 2596

	if (cachep->flags & SLAB_RED_ZONE) {
A
Andrew Morton 已提交
2597 2598 2599 2600 2601 2602
		if (*dbg_redzone1(cachep, objp) != RED_ACTIVE ||
				*dbg_redzone2(cachep, objp) != RED_ACTIVE) {
			slab_error(cachep, "double free, or memory outside"
						" object was overwritten");
			printk(KERN_ERR "%p: redzone 1:0x%lx, "
					"redzone 2:0x%lx.\n",
P
Pekka Enberg 已提交
2603 2604
			       objp, *dbg_redzone1(cachep, objp),
			       *dbg_redzone2(cachep, objp));
L
Linus Torvalds 已提交
2605 2606 2607 2608 2609 2610 2611
		}
		*dbg_redzone1(cachep, objp) = RED_INACTIVE;
		*dbg_redzone2(cachep, objp) = RED_INACTIVE;
	}
	if (cachep->flags & SLAB_STORE_USER)
		*dbg_userword(cachep, objp) = caller;

2612
	objnr = obj_to_index(cachep, slabp, objp);
L
Linus Torvalds 已提交
2613 2614

	BUG_ON(objnr >= cachep->num);
2615
	BUG_ON(objp != index_to_obj(cachep, slabp, objnr));
L
Linus Torvalds 已提交
2616 2617

	if (cachep->flags & SLAB_DEBUG_INITIAL) {
A
Andrew Morton 已提交
2618 2619 2620 2621
		/*
		 * Need to call the slab's constructor so the caller can
		 * perform a verify of its state (debugging).  Called without
		 * the cache-lock held.
L
Linus Torvalds 已提交
2622
		 */
2623
		cachep->ctor(objp + obj_offset(cachep),
P
Pekka Enberg 已提交
2624
			     cachep, SLAB_CTOR_CONSTRUCTOR | SLAB_CTOR_VERIFY);
L
Linus Torvalds 已提交
2625 2626 2627 2628 2629
	}
	if (cachep->flags & SLAB_POISON && cachep->dtor) {
		/* we want to cache poison the object,
		 * call the destruction callback
		 */
2630
		cachep->dtor(objp + obj_offset(cachep), cachep, 0);
L
Linus Torvalds 已提交
2631
	}
2632 2633 2634
#ifdef CONFIG_DEBUG_SLAB_LEAK
	slab_bufctl(slabp)[objnr] = BUFCTL_FREE;
#endif
L
Linus Torvalds 已提交
2635 2636
	if (cachep->flags & SLAB_POISON) {
#ifdef CONFIG_DEBUG_PAGEALLOC
A
Andrew Morton 已提交
2637
		if ((cachep->buffer_size % PAGE_SIZE)==0 && OFF_SLAB(cachep)) {
L
Linus Torvalds 已提交
2638
			store_stackinfo(cachep, objp, (unsigned long)caller);
P
Pekka Enberg 已提交
2639
			kernel_map_pages(virt_to_page(objp),
2640
					 cachep->buffer_size / PAGE_SIZE, 0);
L
Linus Torvalds 已提交
2641 2642 2643 2644 2645 2646 2647 2648 2649 2650
		} else {
			poison_obj(cachep, objp, POISON_FREE);
		}
#else
		poison_obj(cachep, objp, POISON_FREE);
#endif
	}
	return objp;
}

2651
static void check_slabp(struct kmem_cache *cachep, struct slab *slabp)
L
Linus Torvalds 已提交
2652 2653 2654
{
	kmem_bufctl_t i;
	int entries = 0;
P
Pekka Enberg 已提交
2655

L
Linus Torvalds 已提交
2656 2657 2658 2659 2660 2661 2662
	/* Check slab's freelist to see if this obj is there. */
	for (i = slabp->free; i != BUFCTL_END; i = slab_bufctl(slabp)[i]) {
		entries++;
		if (entries > cachep->num || i >= cachep->num)
			goto bad;
	}
	if (entries != cachep->num - slabp->inuse) {
A
Andrew Morton 已提交
2663 2664 2665 2666
bad:
		printk(KERN_ERR "slab: Internal list corruption detected in "
				"cache '%s'(%d), slabp %p(%d). Hexdump:\n",
			cachep->name, cachep->num, slabp, slabp->inuse);
P
Pekka Enberg 已提交
2667
		for (i = 0;
2668
		     i < sizeof(*slabp) + cachep->num * sizeof(kmem_bufctl_t);
P
Pekka Enberg 已提交
2669
		     i++) {
A
Andrew Morton 已提交
2670
			if (i % 16 == 0)
L
Linus Torvalds 已提交
2671
				printk("\n%03x:", i);
P
Pekka Enberg 已提交
2672
			printk(" %02x", ((unsigned char *)slabp)[i]);
L
Linus Torvalds 已提交
2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683
		}
		printk("\n");
		BUG();
	}
}
#else
#define kfree_debugcheck(x) do { } while(0)
#define cache_free_debugcheck(x,objp,z) (objp)
#define check_slabp(x,y) do { } while(0)
#endif

2684
static void *cache_alloc_refill(struct kmem_cache *cachep, gfp_t flags)
L
Linus Torvalds 已提交
2685 2686 2687 2688 2689 2690
{
	int batchcount;
	struct kmem_list3 *l3;
	struct array_cache *ac;

	check_irq_off();
2691
	ac = cpu_cache_get(cachep);
A
Andrew Morton 已提交
2692
retry:
L
Linus Torvalds 已提交
2693 2694
	batchcount = ac->batchcount;
	if (!ac->touched && batchcount > BATCHREFILL_LIMIT) {
A
Andrew Morton 已提交
2695 2696 2697 2698
		/*
		 * If there was little recent activity on this cache, then
		 * perform only a partial refill.  Otherwise we could generate
		 * refill bouncing.
L
Linus Torvalds 已提交
2699 2700 2701
		 */
		batchcount = BATCHREFILL_LIMIT;
	}
2702 2703 2704 2705
	l3 = cachep->nodelists[numa_node_id()];

	BUG_ON(ac->avail > 0 || !l3);
	spin_lock(&l3->list_lock);
L
Linus Torvalds 已提交
2706

2707 2708 2709 2710
	/* See if we can refill from the shared array */
	if (l3->shared && transfer_objects(ac, l3->shared, batchcount))
		goto alloc_done;

L
Linus Torvalds 已提交
2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730
	while (batchcount > 0) {
		struct list_head *entry;
		struct slab *slabp;
		/* Get slab alloc is to come from. */
		entry = l3->slabs_partial.next;
		if (entry == &l3->slabs_partial) {
			l3->free_touched = 1;
			entry = l3->slabs_free.next;
			if (entry == &l3->slabs_free)
				goto must_grow;
		}

		slabp = list_entry(entry, struct slab, list);
		check_slabp(cachep, slabp);
		check_spinlock_acquired(cachep);
		while (slabp->inuse < cachep->num && batchcount--) {
			STATS_INC_ALLOCED(cachep);
			STATS_INC_ACTIVE(cachep);
			STATS_SET_HIGH(cachep);

2731 2732
			ac->entry[ac->avail++] = slab_get_obj(cachep, slabp,
							    numa_node_id());
L
Linus Torvalds 已提交
2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743
		}
		check_slabp(cachep, slabp);

		/* move slabp to correct slabp list: */
		list_del(&slabp->list);
		if (slabp->free == BUFCTL_END)
			list_add(&slabp->list, &l3->slabs_full);
		else
			list_add(&slabp->list, &l3->slabs_partial);
	}

A
Andrew Morton 已提交
2744
must_grow:
L
Linus Torvalds 已提交
2745
	l3->free_objects -= ac->avail;
A
Andrew Morton 已提交
2746
alloc_done:
2747
	spin_unlock(&l3->list_lock);
L
Linus Torvalds 已提交
2748 2749 2750

	if (unlikely(!ac->avail)) {
		int x;
2751 2752
		x = cache_grow(cachep, flags, numa_node_id());

A
Andrew Morton 已提交
2753
		/* cache_grow can reenable interrupts, then ac could change. */
2754
		ac = cpu_cache_get(cachep);
A
Andrew Morton 已提交
2755
		if (!x && ac->avail == 0)	/* no objects in sight? abort */
L
Linus Torvalds 已提交
2756 2757
			return NULL;

A
Andrew Morton 已提交
2758
		if (!ac->avail)		/* objects refilled by interrupt? */
L
Linus Torvalds 已提交
2759 2760 2761
			goto retry;
	}
	ac->touched = 1;
2762
	return ac->entry[--ac->avail];
L
Linus Torvalds 已提交
2763 2764
}

A
Andrew Morton 已提交
2765 2766
static inline void cache_alloc_debugcheck_before(struct kmem_cache *cachep,
						gfp_t flags)
L
Linus Torvalds 已提交
2767 2768 2769 2770 2771 2772 2773 2774
{
	might_sleep_if(flags & __GFP_WAIT);
#if DEBUG
	kmem_flagcheck(cachep, flags);
#endif
}

#if DEBUG
A
Andrew Morton 已提交
2775 2776
static void *cache_alloc_debugcheck_after(struct kmem_cache *cachep,
				gfp_t flags, void *objp, void *caller)
L
Linus Torvalds 已提交
2777
{
P
Pekka Enberg 已提交
2778
	if (!objp)
L
Linus Torvalds 已提交
2779
		return objp;
P
Pekka Enberg 已提交
2780
	if (cachep->flags & SLAB_POISON) {
L
Linus Torvalds 已提交
2781
#ifdef CONFIG_DEBUG_PAGEALLOC
2782
		if ((cachep->buffer_size % PAGE_SIZE) == 0 && OFF_SLAB(cachep))
P
Pekka Enberg 已提交
2783
			kernel_map_pages(virt_to_page(objp),
2784
					 cachep->buffer_size / PAGE_SIZE, 1);
L
Linus Torvalds 已提交
2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795
		else
			check_poison_obj(cachep, objp);
#else
		check_poison_obj(cachep, objp);
#endif
		poison_obj(cachep, objp, POISON_INUSE);
	}
	if (cachep->flags & SLAB_STORE_USER)
		*dbg_userword(cachep, objp) = caller;

	if (cachep->flags & SLAB_RED_ZONE) {
A
Andrew Morton 已提交
2796 2797 2798 2799
		if (*dbg_redzone1(cachep, objp) != RED_INACTIVE ||
				*dbg_redzone2(cachep, objp) != RED_INACTIVE) {
			slab_error(cachep, "double free, or memory outside"
						" object was overwritten");
P
Pekka Enberg 已提交
2800
			printk(KERN_ERR
A
Andrew Morton 已提交
2801 2802 2803
				"%p: redzone 1:0x%lx, redzone 2:0x%lx\n",
				objp, *dbg_redzone1(cachep, objp),
				*dbg_redzone2(cachep, objp));
L
Linus Torvalds 已提交
2804 2805 2806 2807
		}
		*dbg_redzone1(cachep, objp) = RED_ACTIVE;
		*dbg_redzone2(cachep, objp) = RED_ACTIVE;
	}
2808 2809 2810 2811 2812 2813 2814 2815 2816 2817
#ifdef CONFIG_DEBUG_SLAB_LEAK
	{
		struct slab *slabp;
		unsigned objnr;

		slabp = page_get_slab(virt_to_page(objp));
		objnr = (unsigned)(objp - slabp->s_mem) / cachep->buffer_size;
		slab_bufctl(slabp)[objnr] = BUFCTL_ACTIVE;
	}
#endif
2818
	objp += obj_offset(cachep);
L
Linus Torvalds 已提交
2819
	if (cachep->ctor && cachep->flags & SLAB_POISON) {
P
Pekka Enberg 已提交
2820
		unsigned long ctor_flags = SLAB_CTOR_CONSTRUCTOR;
L
Linus Torvalds 已提交
2821 2822 2823 2824 2825

		if (!(flags & __GFP_WAIT))
			ctor_flags |= SLAB_CTOR_ATOMIC;

		cachep->ctor(objp, cachep, ctor_flags);
P
Pekka Enberg 已提交
2826
	}
L
Linus Torvalds 已提交
2827 2828 2829 2830 2831 2832
	return objp;
}
#else
#define cache_alloc_debugcheck_after(a,b,objp,d) (objp)
#endif

2833
static inline void *____cache_alloc(struct kmem_cache *cachep, gfp_t flags)
L
Linus Torvalds 已提交
2834
{
P
Pekka Enberg 已提交
2835
	void *objp;
L
Linus Torvalds 已提交
2836 2837
	struct array_cache *ac;

2838
#ifdef CONFIG_NUMA
2839
	if (unlikely(current->flags & (PF_SPREAD_SLAB | PF_MEMPOLICY))) {
2840 2841 2842
		objp = alternate_node_alloc(cachep, flags);
		if (objp != NULL)
			return objp;
2843 2844 2845
	}
#endif

2846
	check_irq_off();
2847
	ac = cpu_cache_get(cachep);
L
Linus Torvalds 已提交
2848 2849 2850
	if (likely(ac->avail)) {
		STATS_INC_ALLOCHIT(cachep);
		ac->touched = 1;
2851
		objp = ac->entry[--ac->avail];
L
Linus Torvalds 已提交
2852 2853 2854 2855
	} else {
		STATS_INC_ALLOCMISS(cachep);
		objp = cache_alloc_refill(cachep, flags);
	}
2856 2857 2858
	return objp;
}

A
Andrew Morton 已提交
2859 2860
static __always_inline void *__cache_alloc(struct kmem_cache *cachep,
						gfp_t flags, void *caller)
2861 2862
{
	unsigned long save_flags;
P
Pekka Enberg 已提交
2863
	void *objp;
2864 2865 2866 2867 2868

	cache_alloc_debugcheck_before(cachep, flags);

	local_irq_save(save_flags);
	objp = ____cache_alloc(cachep, flags);
L
Linus Torvalds 已提交
2869
	local_irq_restore(save_flags);
2870
	objp = cache_alloc_debugcheck_after(cachep, flags, objp,
2871
					    caller);
2872
	prefetchw(objp);
L
Linus Torvalds 已提交
2873 2874 2875
	return objp;
}

2876
#ifdef CONFIG_NUMA
2877
/*
2878
 * Try allocating on another node if PF_SPREAD_SLAB|PF_MEMPOLICY.
2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898
 *
 * If we are in_interrupt, then process context, including cpusets and
 * mempolicy, may not apply and should not be used for allocation policy.
 */
static void *alternate_node_alloc(struct kmem_cache *cachep, gfp_t flags)
{
	int nid_alloc, nid_here;

	if (in_interrupt())
		return NULL;
	nid_alloc = nid_here = numa_node_id();
	if (cpuset_do_slab_mem_spread() && (cachep->flags & SLAB_MEM_SPREAD))
		nid_alloc = cpuset_mem_spread_node();
	else if (current->mempolicy)
		nid_alloc = slab_node(current->mempolicy);
	if (nid_alloc != nid_here)
		return __cache_alloc_node(cachep, flags, nid_alloc);
	return NULL;
}

2899 2900
/*
 * A interface to enable slab creation on nodeid
L
Linus Torvalds 已提交
2901
 */
A
Andrew Morton 已提交
2902 2903
static void *__cache_alloc_node(struct kmem_cache *cachep, gfp_t flags,
				int nodeid)
2904 2905
{
	struct list_head *entry;
P
Pekka Enberg 已提交
2906 2907 2908 2909 2910 2911 2912 2913
	struct slab *slabp;
	struct kmem_list3 *l3;
	void *obj;
	int x;

	l3 = cachep->nodelists[nodeid];
	BUG_ON(!l3);

A
Andrew Morton 已提交
2914
retry:
2915
	check_irq_off();
P
Pekka Enberg 已提交
2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934
	spin_lock(&l3->list_lock);
	entry = l3->slabs_partial.next;
	if (entry == &l3->slabs_partial) {
		l3->free_touched = 1;
		entry = l3->slabs_free.next;
		if (entry == &l3->slabs_free)
			goto must_grow;
	}

	slabp = list_entry(entry, struct slab, list);
	check_spinlock_acquired_node(cachep, nodeid);
	check_slabp(cachep, slabp);

	STATS_INC_NODEALLOCS(cachep);
	STATS_INC_ACTIVE(cachep);
	STATS_SET_HIGH(cachep);

	BUG_ON(slabp->inuse == cachep->num);

2935
	obj = slab_get_obj(cachep, slabp, nodeid);
P
Pekka Enberg 已提交
2936 2937 2938 2939 2940
	check_slabp(cachep, slabp);
	l3->free_objects--;
	/* move slabp to correct slabp list: */
	list_del(&slabp->list);

A
Andrew Morton 已提交
2941
	if (slabp->free == BUFCTL_END)
P
Pekka Enberg 已提交
2942
		list_add(&slabp->list, &l3->slabs_full);
A
Andrew Morton 已提交
2943
	else
P
Pekka Enberg 已提交
2944
		list_add(&slabp->list, &l3->slabs_partial);
2945

P
Pekka Enberg 已提交
2946 2947
	spin_unlock(&l3->list_lock);
	goto done;
2948

A
Andrew Morton 已提交
2949
must_grow:
P
Pekka Enberg 已提交
2950 2951
	spin_unlock(&l3->list_lock);
	x = cache_grow(cachep, flags, nodeid);
L
Linus Torvalds 已提交
2952

P
Pekka Enberg 已提交
2953 2954
	if (!x)
		return NULL;
2955

P
Pekka Enberg 已提交
2956
	goto retry;
A
Andrew Morton 已提交
2957
done:
P
Pekka Enberg 已提交
2958
	return obj;
2959 2960 2961 2962 2963 2964
}
#endif

/*
 * Caller needs to acquire correct kmem_list's list_lock
 */
2965
static void free_block(struct kmem_cache *cachep, void **objpp, int nr_objects,
P
Pekka Enberg 已提交
2966
		       int node)
L
Linus Torvalds 已提交
2967 2968
{
	int i;
2969
	struct kmem_list3 *l3;
L
Linus Torvalds 已提交
2970 2971 2972 2973 2974

	for (i = 0; i < nr_objects; i++) {
		void *objp = objpp[i];
		struct slab *slabp;

2975
		slabp = virt_to_slab(objp);
2976
		l3 = cachep->nodelists[node];
L
Linus Torvalds 已提交
2977
		list_del(&slabp->list);
2978
		check_spinlock_acquired_node(cachep, node);
L
Linus Torvalds 已提交
2979
		check_slabp(cachep, slabp);
2980
		slab_put_obj(cachep, slabp, objp, node);
L
Linus Torvalds 已提交
2981
		STATS_DEC_ACTIVE(cachep);
2982
		l3->free_objects++;
L
Linus Torvalds 已提交
2983 2984 2985 2986
		check_slabp(cachep, slabp);

		/* fixup slab chains */
		if (slabp->inuse == 0) {
2987 2988
			if (l3->free_objects > l3->free_limit) {
				l3->free_objects -= cachep->num;
L
Linus Torvalds 已提交
2989 2990
				slab_destroy(cachep, slabp);
			} else {
2991
				list_add(&slabp->list, &l3->slabs_free);
L
Linus Torvalds 已提交
2992 2993 2994 2995 2996 2997
			}
		} else {
			/* Unconditionally move a slab to the end of the
			 * partial list on free - maximum time for the
			 * other objects to be freed, too.
			 */
2998
			list_add_tail(&slabp->list, &l3->slabs_partial);
L
Linus Torvalds 已提交
2999 3000 3001 3002
		}
	}
}

3003
static void cache_flusharray(struct kmem_cache *cachep, struct array_cache *ac)
L
Linus Torvalds 已提交
3004 3005
{
	int batchcount;
3006
	struct kmem_list3 *l3;
3007
	int node = numa_node_id();
L
Linus Torvalds 已提交
3008 3009 3010 3011 3012 3013

	batchcount = ac->batchcount;
#if DEBUG
	BUG_ON(!batchcount || batchcount > ac->avail);
#endif
	check_irq_off();
3014
	l3 = cachep->nodelists[node];
3015 3016 3017
	spin_lock(&l3->list_lock);
	if (l3->shared) {
		struct array_cache *shared_array = l3->shared;
P
Pekka Enberg 已提交
3018
		int max = shared_array->limit - shared_array->avail;
L
Linus Torvalds 已提交
3019 3020 3021
		if (max) {
			if (batchcount > max)
				batchcount = max;
3022
			memcpy(&(shared_array->entry[shared_array->avail]),
P
Pekka Enberg 已提交
3023
			       ac->entry, sizeof(void *) * batchcount);
L
Linus Torvalds 已提交
3024 3025 3026 3027 3028
			shared_array->avail += batchcount;
			goto free_done;
		}
	}

3029
	free_block(cachep, ac->entry, batchcount, node);
A
Andrew Morton 已提交
3030
free_done:
L
Linus Torvalds 已提交
3031 3032 3033 3034 3035
#if STATS
	{
		int i = 0;
		struct list_head *p;

3036 3037
		p = l3->slabs_free.next;
		while (p != &(l3->slabs_free)) {
L
Linus Torvalds 已提交
3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048
			struct slab *slabp;

			slabp = list_entry(p, struct slab, list);
			BUG_ON(slabp->inuse);

			i++;
			p = p->next;
		}
		STATS_SET_FREEABLE(cachep, i);
	}
#endif
3049
	spin_unlock(&l3->list_lock);
L
Linus Torvalds 已提交
3050
	ac->avail -= batchcount;
A
Andrew Morton 已提交
3051
	memmove(ac->entry, &(ac->entry[batchcount]), sizeof(void *)*ac->avail);
L
Linus Torvalds 已提交
3052 3053 3054
}

/*
A
Andrew Morton 已提交
3055 3056
 * Release an obj back to its cache. If the obj has a constructed state, it must
 * be in this state _before_ it is released.  Called with disabled ints.
L
Linus Torvalds 已提交
3057
 */
3058
static inline void __cache_free(struct kmem_cache *cachep, void *objp)
L
Linus Torvalds 已提交
3059
{
3060
	struct array_cache *ac = cpu_cache_get(cachep);
L
Linus Torvalds 已提交
3061 3062 3063 3064

	check_irq_off();
	objp = cache_free_debugcheck(cachep, objp, __builtin_return_address(0));

3065 3066 3067 3068 3069 3070
	/* Make sure we are not freeing a object from another
	 * node to the array cache on this cpu.
	 */
#ifdef CONFIG_NUMA
	{
		struct slab *slabp;
3071
		slabp = virt_to_slab(objp);
3072 3073 3074
		if (unlikely(slabp->nodeid != numa_node_id())) {
			struct array_cache *alien = NULL;
			int nodeid = slabp->nodeid;
A
Andrew Morton 已提交
3075
			struct kmem_list3 *l3;
3076

A
Andrew Morton 已提交
3077
			l3 = cachep->nodelists[numa_node_id()];
3078 3079 3080 3081 3082 3083
			STATS_INC_NODEFREES(cachep);
			if (l3->alien && l3->alien[nodeid]) {
				alien = l3->alien[nodeid];
				spin_lock(&alien->lock);
				if (unlikely(alien->avail == alien->limit))
					__drain_alien_cache(cachep,
P
Pekka Enberg 已提交
3084
							    alien, nodeid);
3085 3086 3087 3088
				alien->entry[alien->avail++] = objp;
				spin_unlock(&alien->lock);
			} else {
				spin_lock(&(cachep->nodelists[nodeid])->
P
Pekka Enberg 已提交
3089
					  list_lock);
3090
				free_block(cachep, &objp, 1, nodeid);
3091
				spin_unlock(&(cachep->nodelists[nodeid])->
P
Pekka Enberg 已提交
3092
					    list_lock);
3093 3094 3095 3096 3097
			}
			return;
		}
	}
#endif
L
Linus Torvalds 已提交
3098 3099
	if (likely(ac->avail < ac->limit)) {
		STATS_INC_FREEHIT(cachep);
3100
		ac->entry[ac->avail++] = objp;
L
Linus Torvalds 已提交
3101 3102 3103 3104
		return;
	} else {
		STATS_INC_FREEMISS(cachep);
		cache_flusharray(cachep, ac);
3105
		ac->entry[ac->avail++] = objp;
L
Linus Torvalds 已提交
3106 3107 3108 3109 3110 3111 3112 3113 3114 3115 3116
	}
}

/**
 * kmem_cache_alloc - Allocate an object
 * @cachep: The cache to allocate from.
 * @flags: See kmalloc().
 *
 * Allocate an object from this cache.  The flags are only relevant
 * if the cache has no available objects.
 */
3117
void *kmem_cache_alloc(struct kmem_cache *cachep, gfp_t flags)
L
Linus Torvalds 已提交
3118
{
3119
	return __cache_alloc(cachep, flags, __builtin_return_address(0));
L
Linus Torvalds 已提交
3120 3121 3122
}
EXPORT_SYMBOL(kmem_cache_alloc);

3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139
/**
 * kmem_cache_alloc - Allocate an object. The memory is set to zero.
 * @cache: The cache to allocate from.
 * @flags: See kmalloc().
 *
 * Allocate an object from this cache and set the allocated memory to zero.
 * The flags are only relevant if the cache has no available objects.
 */
void *kmem_cache_zalloc(struct kmem_cache *cache, gfp_t flags)
{
	void *ret = __cache_alloc(cache, flags, __builtin_return_address(0));
	if (ret)
		memset(ret, 0, obj_size(cache));
	return ret;
}
EXPORT_SYMBOL(kmem_cache_zalloc);

L
Linus Torvalds 已提交
3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153
/**
 * kmem_ptr_validate - check if an untrusted pointer might
 *	be a slab entry.
 * @cachep: the cache we're checking against
 * @ptr: pointer to validate
 *
 * This verifies that the untrusted pointer looks sane:
 * it is _not_ a guarantee that the pointer is actually
 * part of the slab cache in question, but it at least
 * validates that the pointer can be dereferenced and
 * looks half-way sane.
 *
 * Currently only used for dentry validation.
 */
3154
int fastcall kmem_ptr_validate(struct kmem_cache *cachep, void *ptr)
L
Linus Torvalds 已提交
3155
{
P
Pekka Enberg 已提交
3156
	unsigned long addr = (unsigned long)ptr;
L
Linus Torvalds 已提交
3157
	unsigned long min_addr = PAGE_OFFSET;
P
Pekka Enberg 已提交
3158
	unsigned long align_mask = BYTES_PER_WORD - 1;
3159
	unsigned long size = cachep->buffer_size;
L
Linus Torvalds 已提交
3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174
	struct page *page;

	if (unlikely(addr < min_addr))
		goto out;
	if (unlikely(addr > (unsigned long)high_memory - size))
		goto out;
	if (unlikely(addr & align_mask))
		goto out;
	if (unlikely(!kern_addr_valid(addr)))
		goto out;
	if (unlikely(!kern_addr_valid(addr + size - 1)))
		goto out;
	page = virt_to_page(ptr);
	if (unlikely(!PageSlab(page)))
		goto out;
3175
	if (unlikely(page_get_cache(page) != cachep))
L
Linus Torvalds 已提交
3176 3177
		goto out;
	return 1;
A
Andrew Morton 已提交
3178
out:
L
Linus Torvalds 已提交
3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191
	return 0;
}

#ifdef CONFIG_NUMA
/**
 * kmem_cache_alloc_node - Allocate an object on the specified node
 * @cachep: The cache to allocate from.
 * @flags: See kmalloc().
 * @nodeid: node number of the target node.
 *
 * Identical to kmem_cache_alloc, except that this function is slow
 * and can sleep. And it will allocate memory on the given node, which
 * can improve the performance for cpu bound structures.
3192 3193
 * New and improved: it will now make sure that the object gets
 * put on the correct node list so that there is no false sharing.
L
Linus Torvalds 已提交
3194
 */
3195
void *kmem_cache_alloc_node(struct kmem_cache *cachep, gfp_t flags, int nodeid)
L
Linus Torvalds 已提交
3196
{
3197 3198
	unsigned long save_flags;
	void *ptr;
L
Linus Torvalds 已提交
3199

3200 3201
	cache_alloc_debugcheck_before(cachep, flags);
	local_irq_save(save_flags);
3202 3203

	if (nodeid == -1 || nodeid == numa_node_id() ||
A
Andrew Morton 已提交
3204
			!cachep->nodelists[nodeid])
3205 3206 3207
		ptr = ____cache_alloc(cachep, flags);
	else
		ptr = __cache_alloc_node(cachep, flags, nodeid);
3208
	local_irq_restore(save_flags);
3209 3210 3211

	ptr = cache_alloc_debugcheck_after(cachep, flags, ptr,
					   __builtin_return_address(0));
L
Linus Torvalds 已提交
3212

3213
	return ptr;
L
Linus Torvalds 已提交
3214 3215 3216
}
EXPORT_SYMBOL(kmem_cache_alloc_node);

A
Al Viro 已提交
3217
void *kmalloc_node(size_t size, gfp_t flags, int node)
3218
{
3219
	struct kmem_cache *cachep;
3220 3221 3222 3223 3224 3225 3226

	cachep = kmem_find_general_cachep(size, flags);
	if (unlikely(cachep == NULL))
		return NULL;
	return kmem_cache_alloc_node(cachep, flags, node);
}
EXPORT_SYMBOL(kmalloc_node);
L
Linus Torvalds 已提交
3227 3228 3229 3230 3231 3232
#endif

/**
 * kmalloc - allocate memory
 * @size: how many bytes of memory are required.
 * @flags: the type of memory to allocate.
3233
 * @caller: function caller for debug tracking of the caller
L
Linus Torvalds 已提交
3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250
 *
 * kmalloc is the normal method of allocating memory
 * in the kernel.
 *
 * The @flags argument may be one of:
 *
 * %GFP_USER - Allocate memory on behalf of user.  May sleep.
 *
 * %GFP_KERNEL - Allocate normal kernel ram.  May sleep.
 *
 * %GFP_ATOMIC - Allocation will not sleep.  Use inside interrupt handlers.
 *
 * Additionally, the %GFP_DMA flag may be set to indicate the memory
 * must be suitable for DMA.  This can mean different things on different
 * platforms.  For example, on i386, it means that the memory must come
 * from the first 16MB.
 */
3251 3252
static __always_inline void *__do_kmalloc(size_t size, gfp_t flags,
					  void *caller)
L
Linus Torvalds 已提交
3253
{
3254
	struct kmem_cache *cachep;
L
Linus Torvalds 已提交
3255

3256 3257 3258 3259 3260 3261
	/* If you want to save a few bytes .text space: replace
	 * __ with kmem_.
	 * Then kmalloc uses the uninlined functions instead of the inline
	 * functions.
	 */
	cachep = __find_general_cachep(size, flags);
3262 3263
	if (unlikely(cachep == NULL))
		return NULL;
3264 3265 3266 3267 3268 3269
	return __cache_alloc(cachep, flags, caller);
}


void *__kmalloc(size_t size, gfp_t flags)
{
3270
#ifndef CONFIG_DEBUG_SLAB
3271
	return __do_kmalloc(size, flags, NULL);
3272 3273 3274
#else
	return __do_kmalloc(size, flags, __builtin_return_address(0));
#endif
L
Linus Torvalds 已提交
3275 3276 3277
}
EXPORT_SYMBOL(__kmalloc);

3278
#ifdef CONFIG_DEBUG_SLAB
3279 3280 3281 3282 3283 3284 3285
void *__kmalloc_track_caller(size_t size, gfp_t flags, void *caller)
{
	return __do_kmalloc(size, flags, caller);
}
EXPORT_SYMBOL(__kmalloc_track_caller);
#endif

L
Linus Torvalds 已提交
3286 3287 3288 3289 3290 3291 3292 3293
#ifdef CONFIG_SMP
/**
 * __alloc_percpu - allocate one copy of the object for every present
 * cpu in the system, zeroing them.
 * Objects should be dereferenced using the per_cpu_ptr macro only.
 *
 * @size: how many bytes of memory are required.
 */
3294
void *__alloc_percpu(size_t size)
L
Linus Torvalds 已提交
3295 3296
{
	int i;
P
Pekka Enberg 已提交
3297
	struct percpu_data *pdata = kmalloc(sizeof(*pdata), GFP_KERNEL);
L
Linus Torvalds 已提交
3298 3299 3300 3301

	if (!pdata)
		return NULL;

3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313
	/*
	 * Cannot use for_each_online_cpu since a cpu may come online
	 * and we have no way of figuring out how to fix the array
	 * that we have allocated then....
	 */
	for_each_cpu(i) {
		int node = cpu_to_node(i);

		if (node_online(node))
			pdata->ptrs[i] = kmalloc_node(size, GFP_KERNEL, node);
		else
			pdata->ptrs[i] = kmalloc(size, GFP_KERNEL);
L
Linus Torvalds 已提交
3314 3315 3316 3317 3318 3319 3320

		if (!pdata->ptrs[i])
			goto unwind_oom;
		memset(pdata->ptrs[i], 0, size);
	}

	/* Catch derefs w/o wrappers */
P
Pekka Enberg 已提交
3321
	return (void *)(~(unsigned long)pdata);
L
Linus Torvalds 已提交
3322

A
Andrew Morton 已提交
3323
unwind_oom:
L
Linus Torvalds 已提交
3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339 3340 3341 3342
	while (--i >= 0) {
		if (!cpu_possible(i))
			continue;
		kfree(pdata->ptrs[i]);
	}
	kfree(pdata);
	return NULL;
}
EXPORT_SYMBOL(__alloc_percpu);
#endif

/**
 * kmem_cache_free - Deallocate an object
 * @cachep: The cache the allocation was from.
 * @objp: The previously allocated object.
 *
 * Free an object which was previously allocated from this
 * cache.
 */
3343
void kmem_cache_free(struct kmem_cache *cachep, void *objp)
L
Linus Torvalds 已提交
3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356
{
	unsigned long flags;

	local_irq_save(flags);
	__cache_free(cachep, objp);
	local_irq_restore(flags);
}
EXPORT_SYMBOL(kmem_cache_free);

/**
 * kfree - free previously allocated memory
 * @objp: pointer returned by kmalloc.
 *
3357 3358
 * If @objp is NULL, no operation is performed.
 *
L
Linus Torvalds 已提交
3359 3360 3361 3362 3363
 * Don't free memory not originally allocated by kmalloc()
 * or you will run into trouble.
 */
void kfree(const void *objp)
{
3364
	struct kmem_cache *c;
L
Linus Torvalds 已提交
3365 3366 3367 3368 3369 3370
	unsigned long flags;

	if (unlikely(!objp))
		return;
	local_irq_save(flags);
	kfree_debugcheck(objp);
3371
	c = virt_to_cache(objp);
3372
	mutex_debug_check_no_locks_freed(objp, obj_size(c));
P
Pekka Enberg 已提交
3373
	__cache_free(c, (void *)objp);
L
Linus Torvalds 已提交
3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385
	local_irq_restore(flags);
}
EXPORT_SYMBOL(kfree);

#ifdef CONFIG_SMP
/**
 * free_percpu - free previously allocated percpu memory
 * @objp: pointer returned by alloc_percpu.
 *
 * Don't free memory not originally allocated by alloc_percpu()
 * The complemented objp is to check for that.
 */
P
Pekka Enberg 已提交
3386
void free_percpu(const void *objp)
L
Linus Torvalds 已提交
3387 3388
{
	int i;
P
Pekka Enberg 已提交
3389
	struct percpu_data *p = (struct percpu_data *)(~(unsigned long)objp);
L
Linus Torvalds 已提交
3390

3391 3392 3393 3394
	/*
	 * We allocate for all cpus so we cannot use for online cpu here.
	 */
	for_each_cpu(i)
P
Pekka Enberg 已提交
3395
	    kfree(p->ptrs[i]);
L
Linus Torvalds 已提交
3396 3397 3398 3399 3400
	kfree(p);
}
EXPORT_SYMBOL(free_percpu);
#endif

3401
unsigned int kmem_cache_size(struct kmem_cache *cachep)
L
Linus Torvalds 已提交
3402
{
3403
	return obj_size(cachep);
L
Linus Torvalds 已提交
3404 3405 3406
}
EXPORT_SYMBOL(kmem_cache_size);

3407
const char *kmem_cache_name(struct kmem_cache *cachep)
3408 3409 3410 3411 3412
{
	return cachep->name;
}
EXPORT_SYMBOL_GPL(kmem_cache_name);

3413 3414 3415
/*
 * This initializes kmem_list3 for all nodes.
 */
3416
static int alloc_kmemlist(struct kmem_cache *cachep)
3417 3418 3419 3420 3421 3422 3423 3424 3425
{
	int node;
	struct kmem_list3 *l3;
	int err = 0;

	for_each_online_node(node) {
		struct array_cache *nc = NULL, *new;
		struct array_cache **new_alien = NULL;
#ifdef CONFIG_NUMA
A
Andrew Morton 已提交
3426 3427
		new_alien = alloc_alien_cache(node, cachep->limit);
		if (!new_alien)
3428 3429
			goto fail;
#endif
A
Andrew Morton 已提交
3430 3431 3432
		new = alloc_arraycache(node, cachep->shared*cachep->batchcount,
					0xbaadf00d);
		if (!new)
3433
			goto fail;
A
Andrew Morton 已提交
3434 3435
		l3 = cachep->nodelists[node];
		if (l3) {
3436 3437
			spin_lock_irq(&l3->list_lock);

A
Andrew Morton 已提交
3438 3439
			nc = cachep->nodelists[node]->shared;
			if (nc)
P
Pekka Enberg 已提交
3440
				free_block(cachep, nc->entry, nc->avail, node);
3441 3442 3443 3444 3445 3446

			l3->shared = new;
			if (!cachep->nodelists[node]->alien) {
				l3->alien = new_alien;
				new_alien = NULL;
			}
P
Pekka Enberg 已提交
3447
			l3->free_limit = (1 + nr_cpus_node(node)) *
A
Andrew Morton 已提交
3448
					cachep->batchcount + cachep->num;
3449 3450 3451 3452 3453
			spin_unlock_irq(&l3->list_lock);
			kfree(nc);
			free_alien_cache(new_alien);
			continue;
		}
A
Andrew Morton 已提交
3454 3455
		l3 = kmalloc_node(sizeof(struct kmem_list3), GFP_KERNEL, node);
		if (!l3)
3456 3457 3458 3459
			goto fail;

		kmem_list3_init(l3);
		l3->next_reap = jiffies + REAPTIMEOUT_LIST3 +
A
Andrew Morton 已提交
3460
				((unsigned long)cachep) % REAPTIMEOUT_LIST3;
3461 3462
		l3->shared = new;
		l3->alien = new_alien;
P
Pekka Enberg 已提交
3463
		l3->free_limit = (1 + nr_cpus_node(node)) *
A
Andrew Morton 已提交
3464
					cachep->batchcount + cachep->num;
3465 3466 3467
		cachep->nodelists[node] = l3;
	}
	return err;
A
Andrew Morton 已提交
3468
fail:
3469 3470 3471 3472
	err = -ENOMEM;
	return err;
}

L
Linus Torvalds 已提交
3473
struct ccupdate_struct {
3474
	struct kmem_cache *cachep;
L
Linus Torvalds 已提交
3475 3476 3477 3478 3479
	struct array_cache *new[NR_CPUS];
};

static void do_ccupdate_local(void *info)
{
A
Andrew Morton 已提交
3480
	struct ccupdate_struct *new = info;
L
Linus Torvalds 已提交
3481 3482 3483
	struct array_cache *old;

	check_irq_off();
3484
	old = cpu_cache_get(new->cachep);
3485

L
Linus Torvalds 已提交
3486 3487 3488 3489
	new->cachep->array[smp_processor_id()] = new->new[smp_processor_id()];
	new->new[smp_processor_id()] = old;
}

3490
/* Always called with the cache_chain_mutex held */
A
Andrew Morton 已提交
3491 3492
static int do_tune_cpucache(struct kmem_cache *cachep, int limit,
				int batchcount, int shared)
L
Linus Torvalds 已提交
3493 3494
{
	struct ccupdate_struct new;
3495
	int i, err;
L
Linus Torvalds 已提交
3496

P
Pekka Enberg 已提交
3497
	memset(&new.new, 0, sizeof(new.new));
3498
	for_each_online_cpu(i) {
A
Andrew Morton 已提交
3499 3500
		new.new[i] = alloc_arraycache(cpu_to_node(i), limit,
						batchcount);
3501
		if (!new.new[i]) {
P
Pekka Enberg 已提交
3502 3503
			for (i--; i >= 0; i--)
				kfree(new.new[i]);
3504
			return -ENOMEM;
L
Linus Torvalds 已提交
3505 3506 3507 3508
		}
	}
	new.cachep = cachep;

A
Andrew Morton 已提交
3509
	on_each_cpu(do_ccupdate_local, (void *)&new, 1, 1);
3510

L
Linus Torvalds 已提交
3511 3512 3513
	check_irq_on();
	cachep->batchcount = batchcount;
	cachep->limit = limit;
3514
	cachep->shared = shared;
L
Linus Torvalds 已提交
3515

3516
	for_each_online_cpu(i) {
L
Linus Torvalds 已提交
3517 3518 3519
		struct array_cache *ccold = new.new[i];
		if (!ccold)
			continue;
3520
		spin_lock_irq(&cachep->nodelists[cpu_to_node(i)]->list_lock);
3521
		free_block(cachep, ccold->entry, ccold->avail, cpu_to_node(i));
3522
		spin_unlock_irq(&cachep->nodelists[cpu_to_node(i)]->list_lock);
L
Linus Torvalds 已提交
3523 3524 3525
		kfree(ccold);
	}

3526 3527 3528
	err = alloc_kmemlist(cachep);
	if (err) {
		printk(KERN_ERR "alloc_kmemlist failed for %s, error %d.\n",
P
Pekka Enberg 已提交
3529
		       cachep->name, -err);
3530
		BUG();
L
Linus Torvalds 已提交
3531 3532 3533 3534
	}
	return 0;
}

3535
/* Called with cache_chain_mutex held always */
3536
static void enable_cpucache(struct kmem_cache *cachep)
L
Linus Torvalds 已提交
3537 3538 3539 3540
{
	int err;
	int limit, shared;

A
Andrew Morton 已提交
3541 3542
	/*
	 * The head array serves three purposes:
L
Linus Torvalds 已提交
3543 3544
	 * - create a LIFO ordering, i.e. return objects that are cache-warm
	 * - reduce the number of spinlock operations.
A
Andrew Morton 已提交
3545
	 * - reduce the number of linked list operations on the slab and
L
Linus Torvalds 已提交
3546 3547 3548 3549
	 *   bufctl chains: array operations are cheaper.
	 * The numbers are guessed, we should auto-tune as described by
	 * Bonwick.
	 */
3550
	if (cachep->buffer_size > 131072)
L
Linus Torvalds 已提交
3551
		limit = 1;
3552
	else if (cachep->buffer_size > PAGE_SIZE)
L
Linus Torvalds 已提交
3553
		limit = 8;
3554
	else if (cachep->buffer_size > 1024)
L
Linus Torvalds 已提交
3555
		limit = 24;
3556
	else if (cachep->buffer_size > 256)
L
Linus Torvalds 已提交
3557 3558 3559 3560
		limit = 54;
	else
		limit = 120;

A
Andrew Morton 已提交
3561 3562
	/*
	 * CPU bound tasks (e.g. network routing) can exhibit cpu bound
L
Linus Torvalds 已提交
3563 3564 3565 3566 3567 3568 3569 3570 3571
	 * allocation behaviour: Most allocs on one cpu, most free operations
	 * on another cpu. For these cases, an efficient object passing between
	 * cpus is necessary. This is provided by a shared array. The array
	 * replaces Bonwick's magazine layer.
	 * On uniprocessor, it's functionally equivalent (but less efficient)
	 * to a larger limit. Thus disabled by default.
	 */
	shared = 0;
#ifdef CONFIG_SMP
3572
	if (cachep->buffer_size <= PAGE_SIZE)
L
Linus Torvalds 已提交
3573 3574 3575 3576
		shared = 8;
#endif

#if DEBUG
A
Andrew Morton 已提交
3577 3578 3579
	/*
	 * With debugging enabled, large batchcount lead to excessively long
	 * periods with disabled local interrupts. Limit the batchcount
L
Linus Torvalds 已提交
3580 3581 3582 3583
	 */
	if (limit > 32)
		limit = 32;
#endif
P
Pekka Enberg 已提交
3584
	err = do_tune_cpucache(cachep, limit, (limit + 1) / 2, shared);
L
Linus Torvalds 已提交
3585 3586
	if (err)
		printk(KERN_ERR "enable_cpucache failed for %s, error %d.\n",
P
Pekka Enberg 已提交
3587
		       cachep->name, -err);
L
Linus Torvalds 已提交
3588 3589
}

3590 3591
/*
 * Drain an array if it contains any elements taking the l3 lock only if
3592 3593
 * necessary. Note that the l3 listlock also protects the array_cache
 * if drain_array() is used on the shared array.
3594 3595 3596
 */
void drain_array(struct kmem_cache *cachep, struct kmem_list3 *l3,
			 struct array_cache *ac, int force, int node)
L
Linus Torvalds 已提交
3597 3598 3599
{
	int tofree;

3600 3601
	if (!ac || !ac->avail)
		return;
L
Linus Torvalds 已提交
3602 3603
	if (ac->touched && !force) {
		ac->touched = 0;
3604
	} else {
3605
		spin_lock_irq(&l3->list_lock);
3606 3607 3608 3609 3610 3611 3612 3613 3614
		if (ac->avail) {
			tofree = force ? ac->avail : (ac->limit + 4) / 5;
			if (tofree > ac->avail)
				tofree = (ac->avail + 1) / 2;
			free_block(cachep, ac->entry, tofree, node);
			ac->avail -= tofree;
			memmove(ac->entry, &(ac->entry[tofree]),
				sizeof(void *) * ac->avail);
		}
3615
		spin_unlock_irq(&l3->list_lock);
L
Linus Torvalds 已提交
3616 3617 3618 3619 3620
	}
}

/**
 * cache_reap - Reclaim memory from caches.
3621
 * @unused: unused parameter
L
Linus Torvalds 已提交
3622 3623 3624 3625 3626 3627
 *
 * Called from workqueue/eventd every few seconds.
 * Purpose:
 * - clear the per-cpu caches for this CPU.
 * - return freeable pages to the main free memory pool.
 *
A
Andrew Morton 已提交
3628 3629
 * If we cannot acquire the cache chain mutex then just give up - we'll try
 * again on the next iteration.
L
Linus Torvalds 已提交
3630 3631 3632 3633
 */
static void cache_reap(void *unused)
{
	struct list_head *walk;
3634
	struct kmem_list3 *l3;
3635
	int node = numa_node_id();
L
Linus Torvalds 已提交
3636

I
Ingo Molnar 已提交
3637
	if (!mutex_trylock(&cache_chain_mutex)) {
L
Linus Torvalds 已提交
3638
		/* Give up. Setup the next iteration. */
P
Pekka Enberg 已提交
3639 3640
		schedule_delayed_work(&__get_cpu_var(reap_work),
				      REAPTIMEOUT_CPUC);
L
Linus Torvalds 已提交
3641 3642 3643 3644
		return;
	}

	list_for_each(walk, &cache_chain) {
3645
		struct kmem_cache *searchp;
P
Pekka Enberg 已提交
3646
		struct list_head *p;
L
Linus Torvalds 已提交
3647 3648 3649
		int tofree;
		struct slab *slabp;

3650
		searchp = list_entry(walk, struct kmem_cache, next);
L
Linus Torvalds 已提交
3651 3652
		check_irq_on();

3653 3654 3655 3656 3657
		/*
		 * We only take the l3 lock if absolutely necessary and we
		 * have established with reasonable certainty that
		 * we can do some work if the lock was obtained.
		 */
3658
		l3 = searchp->nodelists[node];
3659

3660
		reap_alien(searchp, l3);
L
Linus Torvalds 已提交
3661

3662
		drain_array(searchp, l3, cpu_cache_get(searchp), 0, node);
L
Linus Torvalds 已提交
3663

3664 3665 3666 3667
		/*
		 * These are racy checks but it does not matter
		 * if we skip one check or scan twice.
		 */
3668
		if (time_after(l3->next_reap, jiffies))
3669
			goto next;
L
Linus Torvalds 已提交
3670

3671
		l3->next_reap = jiffies + REAPTIMEOUT_LIST3;
L
Linus Torvalds 已提交
3672

3673
		drain_array(searchp, l3, l3->shared, 0, node);
L
Linus Torvalds 已提交
3674

3675 3676
		if (l3->free_touched) {
			l3->free_touched = 0;
3677
			goto next;
L
Linus Torvalds 已提交
3678 3679
		}

A
Andrew Morton 已提交
3680 3681
		tofree = (l3->free_limit + 5 * searchp->num - 1) /
				(5 * searchp->num);
L
Linus Torvalds 已提交
3682
		do {
3683 3684 3685 3686 3687 3688 3689
			/*
			 * Do not lock if there are no free blocks.
			 */
			if (list_empty(&l3->slabs_free))
				break;

			spin_lock_irq(&l3->list_lock);
3690
			p = l3->slabs_free.next;
3691 3692
			if (p == &(l3->slabs_free)) {
				spin_unlock_irq(&l3->list_lock);
L
Linus Torvalds 已提交
3693
				break;
3694
			}
L
Linus Torvalds 已提交
3695 3696 3697 3698 3699 3700

			slabp = list_entry(p, struct slab, list);
			BUG_ON(slabp->inuse);
			list_del(&slabp->list);
			STATS_INC_REAPED(searchp);

A
Andrew Morton 已提交
3701 3702 3703
			/*
			 * Safe to drop the lock. The slab is no longer linked
			 * to the cache. searchp cannot disappear, we hold
L
Linus Torvalds 已提交
3704 3705
			 * cache_chain_lock
			 */
3706 3707
			l3->free_objects -= searchp->num;
			spin_unlock_irq(&l3->list_lock);
L
Linus Torvalds 已提交
3708
			slab_destroy(searchp, slabp);
P
Pekka Enberg 已提交
3709
		} while (--tofree > 0);
3710
next:
L
Linus Torvalds 已提交
3711 3712 3713
		cond_resched();
	}
	check_irq_on();
I
Ingo Molnar 已提交
3714
	mutex_unlock(&cache_chain_mutex);
3715
	next_reap_node();
A
Andrew Morton 已提交
3716
	/* Set up the next iteration */
3717
	schedule_delayed_work(&__get_cpu_var(reap_work), REAPTIMEOUT_CPUC);
L
Linus Torvalds 已提交
3718 3719 3720 3721
}

#ifdef CONFIG_PROC_FS

3722
static void print_slabinfo_header(struct seq_file *m)
L
Linus Torvalds 已提交
3723
{
3724 3725 3726 3727
	/*
	 * Output format version, so at least we can change it
	 * without _too_ many complaints.
	 */
L
Linus Torvalds 已提交
3728
#if STATS
3729
	seq_puts(m, "slabinfo - version: 2.1 (statistics)\n");
L
Linus Torvalds 已提交
3730
#else
3731
	seq_puts(m, "slabinfo - version: 2.1\n");
L
Linus Torvalds 已提交
3732
#endif
3733 3734 3735 3736
	seq_puts(m, "# name            <active_objs> <num_objs> <objsize> "
		 "<objperslab> <pagesperslab>");
	seq_puts(m, " : tunables <limit> <batchcount> <sharedfactor>");
	seq_puts(m, " : slabdata <active_slabs> <num_slabs> <sharedavail>");
L
Linus Torvalds 已提交
3737
#if STATS
3738 3739 3740
	seq_puts(m, " : globalstat <listallocs> <maxobjs> <grown> <reaped> "
		 "<error> <maxfreeable> <nodeallocs> <remotefrees>");
	seq_puts(m, " : cpustat <allochit> <allocmiss> <freehit> <freemiss>");
L
Linus Torvalds 已提交
3741
#endif
3742 3743 3744 3745 3746 3747 3748 3749
	seq_putc(m, '\n');
}

static void *s_start(struct seq_file *m, loff_t *pos)
{
	loff_t n = *pos;
	struct list_head *p;

I
Ingo Molnar 已提交
3750
	mutex_lock(&cache_chain_mutex);
3751 3752
	if (!n)
		print_slabinfo_header(m);
L
Linus Torvalds 已提交
3753 3754 3755 3756 3757 3758
	p = cache_chain.next;
	while (n--) {
		p = p->next;
		if (p == &cache_chain)
			return NULL;
	}
3759
	return list_entry(p, struct kmem_cache, next);
L
Linus Torvalds 已提交
3760 3761 3762 3763
}

static void *s_next(struct seq_file *m, void *p, loff_t *pos)
{
3764
	struct kmem_cache *cachep = p;
L
Linus Torvalds 已提交
3765
	++*pos;
A
Andrew Morton 已提交
3766 3767
	return cachep->next.next == &cache_chain ?
		NULL : list_entry(cachep->next.next, struct kmem_cache, next);
L
Linus Torvalds 已提交
3768 3769 3770 3771
}

static void s_stop(struct seq_file *m, void *p)
{
I
Ingo Molnar 已提交
3772
	mutex_unlock(&cache_chain_mutex);
L
Linus Torvalds 已提交
3773 3774 3775 3776
}

static int s_show(struct seq_file *m, void *p)
{
3777
	struct kmem_cache *cachep = p;
L
Linus Torvalds 已提交
3778
	struct list_head *q;
P
Pekka Enberg 已提交
3779 3780 3781 3782 3783
	struct slab *slabp;
	unsigned long active_objs;
	unsigned long num_objs;
	unsigned long active_slabs = 0;
	unsigned long num_slabs, free_objects = 0, shared_avail = 0;
3784
	const char *name;
L
Linus Torvalds 已提交
3785
	char *error = NULL;
3786 3787
	int node;
	struct kmem_list3 *l3;
L
Linus Torvalds 已提交
3788 3789 3790

	active_objs = 0;
	num_slabs = 0;
3791 3792 3793 3794 3795
	for_each_online_node(node) {
		l3 = cachep->nodelists[node];
		if (!l3)
			continue;

3796 3797
		check_irq_on();
		spin_lock_irq(&l3->list_lock);
3798

P
Pekka Enberg 已提交
3799
		list_for_each(q, &l3->slabs_full) {
3800 3801 3802 3803 3804 3805
			slabp = list_entry(q, struct slab, list);
			if (slabp->inuse != cachep->num && !error)
				error = "slabs_full accounting error";
			active_objs += cachep->num;
			active_slabs++;
		}
P
Pekka Enberg 已提交
3806
		list_for_each(q, &l3->slabs_partial) {
3807 3808 3809 3810 3811 3812 3813 3814
			slabp = list_entry(q, struct slab, list);
			if (slabp->inuse == cachep->num && !error)
				error = "slabs_partial inuse accounting error";
			if (!slabp->inuse && !error)
				error = "slabs_partial/inuse accounting error";
			active_objs += slabp->inuse;
			active_slabs++;
		}
P
Pekka Enberg 已提交
3815
		list_for_each(q, &l3->slabs_free) {
3816 3817 3818 3819 3820 3821
			slabp = list_entry(q, struct slab, list);
			if (slabp->inuse && !error)
				error = "slabs_free/inuse accounting error";
			num_slabs++;
		}
		free_objects += l3->free_objects;
3822 3823
		if (l3->shared)
			shared_avail += l3->shared->avail;
3824

3825
		spin_unlock_irq(&l3->list_lock);
L
Linus Torvalds 已提交
3826
	}
P
Pekka Enberg 已提交
3827 3828
	num_slabs += active_slabs;
	num_objs = num_slabs * cachep->num;
3829
	if (num_objs - active_objs != free_objects && !error)
L
Linus Torvalds 已提交
3830 3831
		error = "free_objects accounting error";

P
Pekka Enberg 已提交
3832
	name = cachep->name;
L
Linus Torvalds 已提交
3833 3834 3835 3836
	if (error)
		printk(KERN_ERR "slab: cache %s error: %s\n", name, error);

	seq_printf(m, "%-17s %6lu %6lu %6u %4u %4d",
3837
		   name, active_objs, num_objs, cachep->buffer_size,
P
Pekka Enberg 已提交
3838
		   cachep->num, (1 << cachep->gfporder));
L
Linus Torvalds 已提交
3839
	seq_printf(m, " : tunables %4u %4u %4u",
P
Pekka Enberg 已提交
3840
		   cachep->limit, cachep->batchcount, cachep->shared);
3841
	seq_printf(m, " : slabdata %6lu %6lu %6lu",
P
Pekka Enberg 已提交
3842
		   active_slabs, num_slabs, shared_avail);
L
Linus Torvalds 已提交
3843
#if STATS
P
Pekka Enberg 已提交
3844
	{			/* list3 stats */
L
Linus Torvalds 已提交
3845 3846 3847 3848 3849 3850 3851
		unsigned long high = cachep->high_mark;
		unsigned long allocs = cachep->num_allocations;
		unsigned long grown = cachep->grown;
		unsigned long reaped = cachep->reaped;
		unsigned long errors = cachep->errors;
		unsigned long max_freeable = cachep->max_freeable;
		unsigned long node_allocs = cachep->node_allocs;
3852
		unsigned long node_frees = cachep->node_frees;
L
Linus Torvalds 已提交
3853

3854
		seq_printf(m, " : globalstat %7lu %6lu %5lu %4lu \
A
Andrew Morton 已提交
3855 3856 3857
				%4lu %4lu %4lu %4lu", allocs, high, grown,
				reaped, errors, max_freeable, node_allocs,
				node_frees);
L
Linus Torvalds 已提交
3858 3859 3860 3861 3862 3863 3864 3865 3866
	}
	/* cpu stats */
	{
		unsigned long allochit = atomic_read(&cachep->allochit);
		unsigned long allocmiss = atomic_read(&cachep->allocmiss);
		unsigned long freehit = atomic_read(&cachep->freehit);
		unsigned long freemiss = atomic_read(&cachep->freemiss);

		seq_printf(m, " : cpustat %6lu %6lu %6lu %6lu",
P
Pekka Enberg 已提交
3867
			   allochit, allocmiss, freehit, freemiss);
L
Linus Torvalds 已提交
3868 3869 3870 3871 3872 3873 3874 3875 3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888
	}
#endif
	seq_putc(m, '\n');
	return 0;
}

/*
 * slabinfo_op - iterator that generates /proc/slabinfo
 *
 * Output layout:
 * cache-name
 * num-active-objs
 * total-objs
 * object size
 * num-active-slabs
 * total-slabs
 * num-pages-per-slab
 * + further values on SMP and with statistics enabled
 */

struct seq_operations slabinfo_op = {
P
Pekka Enberg 已提交
3889 3890 3891 3892
	.start = s_start,
	.next = s_next,
	.stop = s_stop,
	.show = s_show,
L
Linus Torvalds 已提交
3893 3894 3895 3896 3897 3898 3899 3900 3901 3902
};

#define MAX_SLABINFO_WRITE 128
/**
 * slabinfo_write - Tuning for the slab allocator
 * @file: unused
 * @buffer: user buffer
 * @count: data length
 * @ppos: unused
 */
P
Pekka Enberg 已提交
3903 3904
ssize_t slabinfo_write(struct file *file, const char __user * buffer,
		       size_t count, loff_t *ppos)
L
Linus Torvalds 已提交
3905
{
P
Pekka Enberg 已提交
3906
	char kbuf[MAX_SLABINFO_WRITE + 1], *tmp;
L
Linus Torvalds 已提交
3907 3908
	int limit, batchcount, shared, res;
	struct list_head *p;
P
Pekka Enberg 已提交
3909

L
Linus Torvalds 已提交
3910 3911 3912 3913
	if (count > MAX_SLABINFO_WRITE)
		return -EINVAL;
	if (copy_from_user(&kbuf, buffer, count))
		return -EFAULT;
P
Pekka Enberg 已提交
3914
	kbuf[MAX_SLABINFO_WRITE] = '\0';
L
Linus Torvalds 已提交
3915 3916 3917 3918 3919 3920 3921 3922 3923 3924

	tmp = strchr(kbuf, ' ');
	if (!tmp)
		return -EINVAL;
	*tmp = '\0';
	tmp++;
	if (sscanf(tmp, " %d %d %d", &limit, &batchcount, &shared) != 3)
		return -EINVAL;

	/* Find the cache in the chain of caches. */
I
Ingo Molnar 已提交
3925
	mutex_lock(&cache_chain_mutex);
L
Linus Torvalds 已提交
3926
	res = -EINVAL;
P
Pekka Enberg 已提交
3927
	list_for_each(p, &cache_chain) {
A
Andrew Morton 已提交
3928
		struct kmem_cache *cachep;
L
Linus Torvalds 已提交
3929

A
Andrew Morton 已提交
3930
		cachep = list_entry(p, struct kmem_cache, next);
L
Linus Torvalds 已提交
3931
		if (!strcmp(cachep->name, kbuf)) {
A
Andrew Morton 已提交
3932 3933
			if (limit < 1 || batchcount < 1 ||
					batchcount > limit || shared < 0) {
3934
				res = 0;
L
Linus Torvalds 已提交
3935
			} else {
3936
				res = do_tune_cpucache(cachep, limit,
P
Pekka Enberg 已提交
3937
						       batchcount, shared);
L
Linus Torvalds 已提交
3938 3939 3940 3941
			}
			break;
		}
	}
I
Ingo Molnar 已提交
3942
	mutex_unlock(&cache_chain_mutex);
L
Linus Torvalds 已提交
3943 3944 3945 3946
	if (res >= 0)
		res = count;
	return res;
}
3947 3948 3949 3950 3951 3952 3953 3954 3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965 3966 3967 3968 3969 3970 3971 3972 3973 3974 3975 3976 3977 3978 3979 3980 3981 3982 3983 3984 3985 3986 3987 3988 3989 3990 3991 3992 3993 3994 3995 3996 3997 3998 3999 4000 4001 4002 4003 4004 4005 4006 4007 4008 4009 4010 4011 4012 4013 4014 4015 4016 4017 4018 4019 4020 4021 4022 4023 4024 4025 4026 4027 4028 4029 4030 4031 4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 4043 4044 4045 4046 4047 4048 4049 4050 4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061 4062 4063 4064 4065 4066 4067 4068 4069 4070 4071 4072 4073 4074 4075 4076 4077 4078 4079 4080 4081 4082 4083 4084 4085 4086 4087 4088 4089 4090 4091 4092 4093 4094 4095 4096 4097 4098 4099

#ifdef CONFIG_DEBUG_SLAB_LEAK

static void *leaks_start(struct seq_file *m, loff_t *pos)
{
	loff_t n = *pos;
	struct list_head *p;

	mutex_lock(&cache_chain_mutex);
	p = cache_chain.next;
	while (n--) {
		p = p->next;
		if (p == &cache_chain)
			return NULL;
	}
	return list_entry(p, struct kmem_cache, next);
}

static inline int add_caller(unsigned long *n, unsigned long v)
{
	unsigned long *p;
	int l;
	if (!v)
		return 1;
	l = n[1];
	p = n + 2;
	while (l) {
		int i = l/2;
		unsigned long *q = p + 2 * i;
		if (*q == v) {
			q[1]++;
			return 1;
		}
		if (*q > v) {
			l = i;
		} else {
			p = q + 2;
			l -= i + 1;
		}
	}
	if (++n[1] == n[0])
		return 0;
	memmove(p + 2, p, n[1] * 2 * sizeof(unsigned long) - ((void *)p - (void *)n));
	p[0] = v;
	p[1] = 1;
	return 1;
}

static void handle_slab(unsigned long *n, struct kmem_cache *c, struct slab *s)
{
	void *p;
	int i;
	if (n[0] == n[1])
		return;
	for (i = 0, p = s->s_mem; i < c->num; i++, p += c->buffer_size) {
		if (slab_bufctl(s)[i] != BUFCTL_ACTIVE)
			continue;
		if (!add_caller(n, (unsigned long)*dbg_userword(c, p)))
			return;
	}
}

static void show_symbol(struct seq_file *m, unsigned long address)
{
#ifdef CONFIG_KALLSYMS
	char *modname;
	const char *name;
	unsigned long offset, size;
	char namebuf[KSYM_NAME_LEN+1];

	name = kallsyms_lookup(address, &size, &offset, &modname, namebuf);

	if (name) {
		seq_printf(m, "%s+%#lx/%#lx", name, offset, size);
		if (modname)
			seq_printf(m, " [%s]", modname);
		return;
	}
#endif
	seq_printf(m, "%p", (void *)address);
}

static int leaks_show(struct seq_file *m, void *p)
{
	struct kmem_cache *cachep = p;
	struct list_head *q;
	struct slab *slabp;
	struct kmem_list3 *l3;
	const char *name;
	unsigned long *n = m->private;
	int node;
	int i;

	if (!(cachep->flags & SLAB_STORE_USER))
		return 0;
	if (!(cachep->flags & SLAB_RED_ZONE))
		return 0;

	/* OK, we can do it */

	n[1] = 0;

	for_each_online_node(node) {
		l3 = cachep->nodelists[node];
		if (!l3)
			continue;

		check_irq_on();
		spin_lock_irq(&l3->list_lock);

		list_for_each(q, &l3->slabs_full) {
			slabp = list_entry(q, struct slab, list);
			handle_slab(n, cachep, slabp);
		}
		list_for_each(q, &l3->slabs_partial) {
			slabp = list_entry(q, struct slab, list);
			handle_slab(n, cachep, slabp);
		}
		spin_unlock_irq(&l3->list_lock);
	}
	name = cachep->name;
	if (n[0] == n[1]) {
		/* Increase the buffer size */
		mutex_unlock(&cache_chain_mutex);
		m->private = kzalloc(n[0] * 4 * sizeof(unsigned long), GFP_KERNEL);
		if (!m->private) {
			/* Too bad, we are really out */
			m->private = n;
			mutex_lock(&cache_chain_mutex);
			return -ENOMEM;
		}
		*(unsigned long *)m->private = n[0] * 2;
		kfree(n);
		mutex_lock(&cache_chain_mutex);
		/* Now make sure this entry will be retried */
		m->count = m->size;
		return 0;
	}
	for (i = 0; i < n[1]; i++) {
		seq_printf(m, "%s: %lu ", name, n[2*i+3]);
		show_symbol(m, n[2*i+2]);
		seq_putc(m, '\n');
	}
	return 0;
}

struct seq_operations slabstats_op = {
	.start = leaks_start,
	.next = s_next,
	.stop = s_stop,
	.show = leaks_show,
};
#endif
L
Linus Torvalds 已提交
4100 4101
#endif

4102 4103 4104 4105 4106 4107 4108 4109 4110 4111 4112 4113
/**
 * ksize - get the actual amount of memory allocated for a given object
 * @objp: Pointer to the object
 *
 * kmalloc may internally round up allocations and return more memory
 * than requested. ksize() can be used to determine the actual amount of
 * memory allocated. The caller may use this additional memory, even though
 * a smaller amount of memory was initially specified with the kmalloc call.
 * The caller must guarantee that objp points to a valid object previously
 * allocated with either kmalloc() or kmem_cache_alloc(). The object
 * must not be freed during the duration of the call.
 */
L
Linus Torvalds 已提交
4114 4115
unsigned int ksize(const void *objp)
{
4116 4117
	if (unlikely(objp == NULL))
		return 0;
L
Linus Torvalds 已提交
4118

4119
	return obj_size(virt_to_cache(objp));
L
Linus Torvalds 已提交
4120
}