slub.c 92.9 KB
Newer Older
Christoph Lameter's avatar
Christoph Lameter committed
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
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
/*
 * SLUB: A slab allocator that limits cache line use instead of queuing
 * objects in per cpu and per node lists.
 *
 * The allocator synchronizes using per slab locks and only
 * uses a centralized lock to manage a pool of partial slabs.
 *
 * (C) 2007 SGI, Christoph Lameter <clameter@sgi.com>
 */

#include <linux/mm.h>
#include <linux/module.h>
#include <linux/bit_spinlock.h>
#include <linux/interrupt.h>
#include <linux/bitops.h>
#include <linux/slab.h>
#include <linux/seq_file.h>
#include <linux/cpu.h>
#include <linux/cpuset.h>
#include <linux/mempolicy.h>
#include <linux/ctype.h>
#include <linux/kallsyms.h>

/*
 * Lock order:
 *   1. slab_lock(page)
 *   2. slab->list_lock
 *
 *   The slab_lock protects operations on the object of a particular
 *   slab and its metadata in the page struct. If the slab lock
 *   has been taken then no allocations nor frees can be performed
 *   on the objects in the slab nor can the slab be added or removed
 *   from the partial or full lists since this would mean modifying
 *   the page_struct of the slab.
 *
 *   The list_lock protects the partial and full list on each node and
 *   the partial slab counter. If taken then no new slabs may be added or
 *   removed from the lists nor make the number of partial slabs be modified.
 *   (Note that the total number of slabs is an atomic value that may be
 *   modified without taking the list lock).
 *
 *   The list_lock is a centralized lock and thus we avoid taking it as
 *   much as possible. As long as SLUB does not have to handle partial
 *   slabs, operations can continue without any centralized lock. F.e.
 *   allocating a long series of objects that fill up slabs does not require
 *   the list lock.
 *
 *   The lock order is sometimes inverted when we are trying to get a slab
 *   off a list. We take the list_lock and then look for a page on the list
 *   to use. While we do that objects in the slabs may be freed. We can
 *   only operate on the slab if we have also taken the slab_lock. So we use
 *   a slab_trylock() on the slab. If trylock was successful then no frees
 *   can occur anymore and we can use the slab for allocations etc. If the
 *   slab_trylock() does not succeed then frees are in progress in the slab and
 *   we must stay away from it for a while since we may cause a bouncing
 *   cacheline if we try to acquire the lock. So go onto the next slab.
 *   If all pages are busy then we may allocate a new slab instead of reusing
 *   a partial slab. A new slab has noone operating on it and thus there is
 *   no danger of cacheline contention.
 *
 *   Interrupts are disabled during allocation and deallocation in order to
 *   make the slab allocator safe to use in the context of an irq. In addition
 *   interrupts are disabled to ensure that the processor does not change
 *   while handling per_cpu slabs, due to kernel preemption.
 *
 * SLUB assigns one slab for allocation to each processor.
 * Allocations only occur from these slabs called cpu slabs.
 *
Christoph Lameter's avatar
Christoph Lameter committed
69
70
 * Slabs with free elements are kept on a partial list and during regular
 * operations no list for full slabs is used. If an object in a full slab is
Christoph Lameter's avatar
Christoph Lameter committed
71
 * freed then the slab will show up again on the partial lists.
Christoph Lameter's avatar
Christoph Lameter committed
72
73
 * We track full slabs for debugging purposes though because otherwise we
 * cannot scan all objects.
Christoph Lameter's avatar
Christoph Lameter committed
74
75
76
77
78
79
80
 *
 * Slabs are freed when they become empty. Teardown and setup is
 * minimal so we rely on the page allocators per cpu caches for
 * fast frees and allocs.
 *
 * Overloading of page flags that are otherwise used for LRU management.
 *
81
82
83
84
85
86
87
88
89
90
91
92
 * PageActive 		The slab is frozen and exempt from list processing.
 * 			This means that the slab is dedicated to a purpose
 * 			such as satisfying allocations for a specific
 * 			processor. Objects may be freed in the slab while
 * 			it is frozen but slab_free will then skip the usual
 * 			list operations. It is up to the processor holding
 * 			the slab to integrate the slab into the slab lists
 * 			when the slab is no longer needed.
 *
 * 			One use of this flag is to mark slabs that are
 * 			used for allocations. Then such a slab becomes a cpu
 * 			slab. The cpu slab may be equipped with an additional
93
 * 			freelist that allows lockless access to
94
95
 * 			free objects in addition to the regular freelist
 * 			that requires the slab lock.
Christoph Lameter's avatar
Christoph Lameter committed
96
97
98
 *
 * PageError		Slab requires special handling due to debug
 * 			options set. This moves	slab handling out of
99
 * 			the fast path and disables lockless freelists.
Christoph Lameter's avatar
Christoph Lameter committed
100
101
 */

102
103
104
105
106
107
108
109
#define FROZEN (1 << PG_active)

#ifdef CONFIG_SLUB_DEBUG
#define SLABDEBUG (1 << PG_error)
#else
#define SLABDEBUG 0
#endif

110
111
static inline int SlabFrozen(struct page *page)
{
112
	return page->flags & FROZEN;
113
114
115
116
}

static inline void SetSlabFrozen(struct page *page)
{
117
	page->flags |= FROZEN;
118
119
120
121
}

static inline void ClearSlabFrozen(struct page *page)
{
122
	page->flags &= ~FROZEN;
123
124
}

125
126
static inline int SlabDebug(struct page *page)
{
127
	return page->flags & SLABDEBUG;
128
129
130
131
}

static inline void SetSlabDebug(struct page *page)
{
132
	page->flags |= SLABDEBUG;
133
134
135
136
}

static inline void ClearSlabDebug(struct page *page)
{
137
	page->flags &= ~SLABDEBUG;
138
139
}

Christoph Lameter's avatar
Christoph Lameter committed
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
/*
 * Issues still to be resolved:
 *
 * - Support PAGE_ALLOC_DEBUG. Should be easy to do.
 *
 * - Variable sizing of the per node arrays
 */

/* Enable to test recovery from slab corruption on boot */
#undef SLUB_RESILIENCY_TEST

#if PAGE_SHIFT <= 12

/*
 * Small page size. Make sure that we do not fragment memory
 */
#define DEFAULT_MAX_ORDER 1
#define DEFAULT_MIN_OBJECTS 4

#else

/*
 * Large page machines are customarily able to handle larger
 * page orders.
 */
#define DEFAULT_MAX_ORDER 2
#define DEFAULT_MIN_OBJECTS 8

#endif

170
171
172
173
/*
 * Mininum number of partial slabs. These will be left on the partial
 * lists even if they are empty. kmem_cache_shrink may reclaim them.
 */
Christoph Lameter's avatar
Christoph Lameter committed
174
175
#define MIN_PARTIAL 2

176
177
178
179
180
181
182
/*
 * Maximum number of desirable partial slabs.
 * The existence of more partial slabs makes kmem_cache_shrink
 * sort the partial list by the number of objects in the.
 */
#define MAX_PARTIAL 10

Christoph Lameter's avatar
Christoph Lameter committed
183
184
#define DEBUG_DEFAULT_FLAGS (SLAB_DEBUG_FREE | SLAB_RED_ZONE | \
				SLAB_POISON | SLAB_STORE_USER)
Christoph Lameter's avatar
Christoph Lameter committed
185

Christoph Lameter's avatar
Christoph Lameter committed
186
187
188
189
190
191
192
193
194
195
/*
 * Set of flags that will prevent slab merging
 */
#define SLUB_NEVER_MERGE (SLAB_RED_ZONE | SLAB_POISON | SLAB_STORE_USER | \
		SLAB_TRACE | SLAB_DESTROY_BY_RCU)

#define SLUB_MERGE_SAME (SLAB_DEBUG_FREE | SLAB_RECLAIM_ACCOUNT | \
		SLAB_CACHE_DMA)

#ifndef ARCH_KMALLOC_MINALIGN
196
#define ARCH_KMALLOC_MINALIGN __alignof__(unsigned long long)
Christoph Lameter's avatar
Christoph Lameter committed
197
198
199
#endif

#ifndef ARCH_SLAB_MINALIGN
200
#define ARCH_SLAB_MINALIGN __alignof__(unsigned long long)
Christoph Lameter's avatar
Christoph Lameter committed
201
202
203
#endif

/* Internal SLUB flags */
204
205
#define __OBJECT_POISON		0x80000000 /* Poison object */
#define __SYSFS_ADD_DEFERRED	0x40000000 /* Not yet visible via sysfs */
Christoph Lameter's avatar
Christoph Lameter committed
206

207
208
209
210
211
/* Not all arches define cache_line_size */
#ifndef cache_line_size
#define cache_line_size()	L1_CACHE_BYTES
#endif

Christoph Lameter's avatar
Christoph Lameter committed
212
213
214
215
216
217
218
219
220
static int kmem_size = sizeof(struct kmem_cache);

#ifdef CONFIG_SMP
static struct notifier_block slab_notifier;
#endif

static enum {
	DOWN,		/* No slab functionality available */
	PARTIAL,	/* kmem_cache_open() works but kmalloc does not */
Christoph Lameter's avatar
Christoph Lameter committed
221
	UP,		/* Everything works but does not show up in sysfs */
Christoph Lameter's avatar
Christoph Lameter committed
222
223
224
225
226
	SYSFS		/* Sysfs up */
} slab_state = DOWN;

/* A list of all slab caches on the system */
static DECLARE_RWSEM(slub_lock);
Adrian Bunk's avatar
Adrian Bunk committed
227
static LIST_HEAD(slab_caches);
Christoph Lameter's avatar
Christoph Lameter committed
228

229
230
231
232
233
234
235
236
237
238
239
240
/*
 * Tracking user of a slab.
 */
struct track {
	void *addr;		/* Called from address */
	int cpu;		/* Was running on cpu */
	int pid;		/* Pid context */
	unsigned long when;	/* When did the operation occur */
};

enum track_item { TRACK_ALLOC, TRACK_FREE };

241
#if defined(CONFIG_SYSFS) && defined(CONFIG_SLUB_DEBUG)
Christoph Lameter's avatar
Christoph Lameter committed
242
243
244
245
static int sysfs_slab_add(struct kmem_cache *);
static int sysfs_slab_alias(struct kmem_cache *, const char *);
static void sysfs_slab_remove(struct kmem_cache *);
#else
246
247
248
249
static inline int sysfs_slab_add(struct kmem_cache *s) { return 0; }
static inline int sysfs_slab_alias(struct kmem_cache *s, const char *p)
							{ return 0; }
static inline void sysfs_slab_remove(struct kmem_cache *s) {}
Christoph Lameter's avatar
Christoph Lameter committed
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
#endif

/********************************************************************
 * 			Core slab cache functions
 *******************************************************************/

int slab_is_available(void)
{
	return slab_state >= UP;
}

static inline struct kmem_cache_node *get_node(struct kmem_cache *s, int node)
{
#ifdef CONFIG_NUMA
	return s->node[node];
#else
	return &s->local_node;
#endif
}

270
271
static inline struct kmem_cache_cpu *get_cpu_slab(struct kmem_cache *s, int cpu)
{
272
273
274
275
276
#ifdef CONFIG_SMP
	return s->cpu_slab[cpu];
#else
	return &s->cpu_slab;
#endif
277
278
}

279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
static inline int check_valid_pointer(struct kmem_cache *s,
				struct page *page, const void *object)
{
	void *base;

	if (!object)
		return 1;

	base = page_address(page);
	if (object < base || object >= base + s->objects * s->size ||
		(object - base) % s->size) {
		return 0;
	}

	return 1;
}

296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
/*
 * Slow version of get and set free pointer.
 *
 * This version requires touching the cache lines of kmem_cache which
 * we avoid to do in the fast alloc free paths. There we obtain the offset
 * from the page struct.
 */
static inline void *get_freepointer(struct kmem_cache *s, void *object)
{
	return *(void **)(object + s->offset);
}

static inline void set_freepointer(struct kmem_cache *s, void *object, void *fp)
{
	*(void **)(object + s->offset) = fp;
}

/* Loop over all objects in a slab */
#define for_each_object(__p, __s, __addr) \
	for (__p = (__addr); __p < (__addr) + (__s)->objects * (__s)->size;\
			__p += (__s)->size)

/* Scan freelist */
#define for_each_free_object(__p, __s, __free) \
	for (__p = (__free); __p; __p = get_freepointer((__s), __p))

/* Determine object index from a given position */
static inline int slab_index(void *p, struct kmem_cache *s, void *addr)
{
	return (p - addr) / s->size;
}

328
329
330
331
#ifdef CONFIG_SLUB_DEBUG
/*
 * Debug settings:
 */
332
333
334
#ifdef CONFIG_SLUB_DEBUG_ON
static int slub_debug = DEBUG_DEFAULT_FLAGS;
#else
335
static int slub_debug;
336
#endif
337
338
339

static char *slub_debug_slabs;

Christoph Lameter's avatar
Christoph Lameter committed
340
341
342
343
344
345
346
347
348
349
350
351
352
/*
 * Object debugging
 */
static void print_section(char *text, u8 *addr, unsigned int length)
{
	int i, offset;
	int newline = 1;
	char ascii[17];

	ascii[16] = 0;

	for (i = 0; i < length; i++) {
		if (newline) {
353
			printk(KERN_ERR "%8s 0x%p: ", text, addr + i);
Christoph Lameter's avatar
Christoph Lameter committed
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
			newline = 0;
		}
		printk(" %02x", addr[i]);
		offset = i % 16;
		ascii[offset] = isgraph(addr[i]) ? addr[i] : '.';
		if (offset == 15) {
			printk(" %s\n",ascii);
			newline = 1;
		}
	}
	if (!newline) {
		i %= 16;
		while (i < 16) {
			printk("   ");
			ascii[i] = ' ';
			i++;
		}
		printk(" %s\n", ascii);
	}
}

static struct track *get_track(struct kmem_cache *s, void *object,
	enum track_item alloc)
{
	struct track *p;

	if (s->offset)
		p = object + s->offset + sizeof(void *);
	else
		p = object + s->inuse;

	return p + alloc;
}

static void set_track(struct kmem_cache *s, void *object,
				enum track_item alloc, void *addr)
{
	struct track *p;

	if (s->offset)
		p = object + s->offset + sizeof(void *);
	else
		p = object + s->inuse;

	p += alloc;
	if (addr) {
		p->addr = addr;
		p->cpu = smp_processor_id();
		p->pid = current ? current->pid : -1;
		p->when = jiffies;
	} else
		memset(p, 0, sizeof(struct track));
}

static void init_tracking(struct kmem_cache *s, void *object)
{
410
411
412
413
414
	if (!(s->flags & SLAB_STORE_USER))
		return;

	set_track(s, object, TRACK_FREE, NULL);
	set_track(s, object, TRACK_ALLOC, NULL);
Christoph Lameter's avatar
Christoph Lameter committed
415
416
417
418
419
420
421
}

static void print_track(const char *s, struct track *t)
{
	if (!t->addr)
		return;

422
	printk(KERN_ERR "INFO: %s in ", s);
Christoph Lameter's avatar
Christoph Lameter committed
423
	__print_symbol("%s", (unsigned long)t->addr);
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
	printk(" age=%lu cpu=%u pid=%d\n", jiffies - t->when, t->cpu, t->pid);
}

static void print_tracking(struct kmem_cache *s, void *object)
{
	if (!(s->flags & SLAB_STORE_USER))
		return;

	print_track("Allocated", get_track(s, object, TRACK_ALLOC));
	print_track("Freed", get_track(s, object, TRACK_FREE));
}

static void print_page_info(struct page *page)
{
	printk(KERN_ERR "INFO: Slab 0x%p used=%u fp=0x%p flags=0x%04lx\n",
		page, page->inuse, page->freelist, page->flags);

}

static void slab_bug(struct kmem_cache *s, char *fmt, ...)
{
	va_list args;
	char buf[100];

	va_start(args, fmt);
	vsnprintf(buf, sizeof(buf), fmt, args);
	va_end(args);
	printk(KERN_ERR "========================================"
			"=====================================\n");
	printk(KERN_ERR "BUG %s: %s\n", s->name, buf);
	printk(KERN_ERR "----------------------------------------"
			"-------------------------------------\n\n");
Christoph Lameter's avatar
Christoph Lameter committed
456
457
}

458
459
460
461
462
463
464
465
466
467
468
469
static void slab_fix(struct kmem_cache *s, char *fmt, ...)
{
	va_list args;
	char buf[100];

	va_start(args, fmt);
	vsnprintf(buf, sizeof(buf), fmt, args);
	va_end(args);
	printk(KERN_ERR "FIX %s: %s\n", s->name, buf);
}

static void print_trailer(struct kmem_cache *s, struct page *page, u8 *p)
Christoph Lameter's avatar
Christoph Lameter committed
470
471
{
	unsigned int off;	/* Offset of last byte */
472
473
474
475
476
477
478
479
480
481
482
483
484
	u8 *addr = page_address(page);

	print_tracking(s, p);

	print_page_info(page);

	printk(KERN_ERR "INFO: Object 0x%p @offset=%tu fp=0x%p\n\n",
			p, p - addr, get_freepointer(s, p));

	if (p > addr + 16)
		print_section("Bytes b4", p - 16, 16);

	print_section("Object", p, min(s->objsize, 128));
Christoph Lameter's avatar
Christoph Lameter committed
485
486
487
488
489
490
491
492
493
494

	if (s->flags & SLAB_RED_ZONE)
		print_section("Redzone", p + s->objsize,
			s->inuse - s->objsize);

	if (s->offset)
		off = s->offset + sizeof(void *);
	else
		off = s->inuse;

495
	if (s->flags & SLAB_STORE_USER)
Christoph Lameter's avatar
Christoph Lameter committed
496
497
498
499
		off += 2 * sizeof(struct track);

	if (off != s->size)
		/* Beginning of the filler is the free pointer */
500
501
502
		print_section("Padding", p + off, s->size - off);

	dump_stack();
Christoph Lameter's avatar
Christoph Lameter committed
503
504
505
506
507
}

static void object_err(struct kmem_cache *s, struct page *page,
			u8 *object, char *reason)
{
508
509
	slab_bug(s, reason);
	print_trailer(s, page, object);
Christoph Lameter's avatar
Christoph Lameter committed
510
511
}

512
static void slab_err(struct kmem_cache *s, struct page *page, char *fmt, ...)
Christoph Lameter's avatar
Christoph Lameter committed
513
514
515
516
{
	va_list args;
	char buf[100];

517
518
	va_start(args, fmt);
	vsnprintf(buf, sizeof(buf), fmt, args);
Christoph Lameter's avatar
Christoph Lameter committed
519
	va_end(args);
520
521
	slab_bug(s, fmt);
	print_page_info(page);
Christoph Lameter's avatar
Christoph Lameter committed
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
	dump_stack();
}

static void init_object(struct kmem_cache *s, void *object, int active)
{
	u8 *p = object;

	if (s->flags & __OBJECT_POISON) {
		memset(p, POISON_FREE, s->objsize - 1);
		p[s->objsize -1] = POISON_END;
	}

	if (s->flags & SLAB_RED_ZONE)
		memset(p + s->objsize,
			active ? SLUB_RED_ACTIVE : SLUB_RED_INACTIVE,
			s->inuse - s->objsize);
}

540
static u8 *check_bytes(u8 *start, unsigned int value, unsigned int bytes)
Christoph Lameter's avatar
Christoph Lameter committed
541
542
543
{
	while (bytes) {
		if (*start != (u8)value)
544
			return start;
Christoph Lameter's avatar
Christoph Lameter committed
545
546
547
		start++;
		bytes--;
	}
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
	return NULL;
}

static void restore_bytes(struct kmem_cache *s, char *message, u8 data,
						void *from, void *to)
{
	slab_fix(s, "Restoring 0x%p-0x%p=0x%x\n", from, to - 1, data);
	memset(from, data, to - from);
}

static int check_bytes_and_report(struct kmem_cache *s, struct page *page,
			u8 *object, char *what,
			u8* start, unsigned int value, unsigned int bytes)
{
	u8 *fault;
	u8 *end;

	fault = check_bytes(start, value, bytes);
	if (!fault)
		return 1;

	end = start + bytes;
	while (end > fault && end[-1] == value)
		end--;

	slab_bug(s, "%s overwritten", what);
	printk(KERN_ERR "INFO: 0x%p-0x%p. First byte 0x%x instead of 0x%x\n",
					fault, end - 1, fault[0], value);
	print_trailer(s, page, object);

	restore_bytes(s, what, value, fault, end);
	return 0;
Christoph Lameter's avatar
Christoph Lameter committed
580
581
582
583
584
585
586
587
588
}

/*
 * Object layout:
 *
 * object address
 * 	Bytes of the object to be managed.
 * 	If the freepointer may overlay the object then the free
 * 	pointer is the first word of the object.
Christoph Lameter's avatar
Christoph Lameter committed
589
 *
Christoph Lameter's avatar
Christoph Lameter committed
590
591
592
593
594
 * 	Poisoning uses 0x6b (POISON_FREE) and the last byte is
 * 	0xa5 (POISON_END)
 *
 * object + s->objsize
 * 	Padding to reach word boundary. This is also used for Redzoning.
Christoph Lameter's avatar
Christoph Lameter committed
595
596
597
 * 	Padding is extended by another word if Redzoning is enabled and
 * 	objsize == inuse.
 *
Christoph Lameter's avatar
Christoph Lameter committed
598
599
600
601
 * 	We fill with 0xbb (RED_INACTIVE) for inactive objects and with
 * 	0xcc (RED_ACTIVE) for objects in use.
 *
 * object + s->inuse
Christoph Lameter's avatar
Christoph Lameter committed
602
603
 * 	Meta data starts here.
 *
Christoph Lameter's avatar
Christoph Lameter committed
604
605
 * 	A. Free pointer (if we cannot overwrite object on free)
 * 	B. Tracking data for SLAB_STORE_USER
Christoph Lameter's avatar
Christoph Lameter committed
606
607
608
609
610
 * 	C. Padding to reach required alignment boundary or at mininum
 * 		one word if debuggin is on to be able to detect writes
 * 		before the word boundary.
 *
 *	Padding is done using 0x5a (POISON_INUSE)
Christoph Lameter's avatar
Christoph Lameter committed
611
612
 *
 * object + s->size
Christoph Lameter's avatar
Christoph Lameter committed
613
 * 	Nothing is used beyond s->size.
Christoph Lameter's avatar
Christoph Lameter committed
614
 *
Christoph Lameter's avatar
Christoph Lameter committed
615
616
 * If slabcaches are merged then the objsize and inuse boundaries are mostly
 * ignored. And therefore no slab options that rely on these boundaries
Christoph Lameter's avatar
Christoph Lameter committed
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
 * may be used with merged slabcaches.
 */

static int check_pad_bytes(struct kmem_cache *s, struct page *page, u8 *p)
{
	unsigned long off = s->inuse;	/* The end of info */

	if (s->offset)
		/* Freepointer is placed after the object. */
		off += sizeof(void *);

	if (s->flags & SLAB_STORE_USER)
		/* We also have user information there */
		off += 2 * sizeof(struct track);

	if (s->size == off)
		return 1;

635
636
	return check_bytes_and_report(s, page, p, "Object padding",
				p + off, POISON_INUSE, s->size - off);
Christoph Lameter's avatar
Christoph Lameter committed
637
638
639
640
}

static int slab_pad_check(struct kmem_cache *s, struct page *page)
{
641
642
643
644
645
	u8 *start;
	u8 *fault;
	u8 *end;
	int length;
	int remainder;
Christoph Lameter's avatar
Christoph Lameter committed
646
647
648
649

	if (!(s->flags & SLAB_POISON))
		return 1;

650
651
	start = page_address(page);
	end = start + (PAGE_SIZE << s->order);
Christoph Lameter's avatar
Christoph Lameter committed
652
	length = s->objects * s->size;
653
	remainder = end - (start + length);
Christoph Lameter's avatar
Christoph Lameter committed
654
655
656
	if (!remainder)
		return 1;

657
658
659
660
661
662
663
664
665
666
667
	fault = check_bytes(start + length, POISON_INUSE, remainder);
	if (!fault)
		return 1;
	while (end > fault && end[-1] == POISON_INUSE)
		end--;

	slab_err(s, page, "Padding overwritten. 0x%p-0x%p", fault, end - 1);
	print_section("Padding", start, length);

	restore_bytes(s, "slab padding", POISON_INUSE, start, end);
	return 0;
Christoph Lameter's avatar
Christoph Lameter committed
668
669
670
671
672
673
674
675
676
677
678
679
}

static int check_object(struct kmem_cache *s, struct page *page,
					void *object, int active)
{
	u8 *p = object;
	u8 *endobject = object + s->objsize;

	if (s->flags & SLAB_RED_ZONE) {
		unsigned int red =
			active ? SLUB_RED_ACTIVE : SLUB_RED_INACTIVE;

680
681
		if (!check_bytes_and_report(s, page, object, "Redzone",
			endobject, red, s->inuse - s->objsize))
Christoph Lameter's avatar
Christoph Lameter committed
682
683
			return 0;
	} else {
684
685
686
		if ((s->flags & SLAB_POISON) && s->objsize < s->inuse)
			check_bytes_and_report(s, page, p, "Alignment padding", endobject,
				POISON_INUSE, s->inuse - s->objsize);
Christoph Lameter's avatar
Christoph Lameter committed
687
688
689
690
	}

	if (s->flags & SLAB_POISON) {
		if (!active && (s->flags & __OBJECT_POISON) &&
691
692
693
694
			(!check_bytes_and_report(s, page, p, "Poison", p,
					POISON_FREE, s->objsize - 1) ||
			 !check_bytes_and_report(s, page, p, "Poison",
			 	p + s->objsize -1, POISON_END, 1)))
Christoph Lameter's avatar
Christoph Lameter committed
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
			return 0;
		/*
		 * check_pad_bytes cleans up on its own.
		 */
		check_pad_bytes(s, page, p);
	}

	if (!s->offset && active)
		/*
		 * Object and freepointer overlap. Cannot check
		 * freepointer while object is allocated.
		 */
		return 1;

	/* Check free pointer validity */
	if (!check_valid_pointer(s, page, get_freepointer(s, p))) {
		object_err(s, page, p, "Freepointer corrupt");
		/*
		 * No choice but to zap it and thus loose the remainder
		 * of the free objects in this slab. May cause
Christoph Lameter's avatar
Christoph Lameter committed
715
		 * another error because the object count is now wrong.
Christoph Lameter's avatar
Christoph Lameter committed
716
717
718
719
720
721
722
723
724
725
726
727
		 */
		set_freepointer(s, p, NULL);
		return 0;
	}
	return 1;
}

static int check_slab(struct kmem_cache *s, struct page *page)
{
	VM_BUG_ON(!irqs_disabled());

	if (!PageSlab(page)) {
728
		slab_err(s, page, "Not a valid slab page");
Christoph Lameter's avatar
Christoph Lameter committed
729
730
731
		return 0;
	}
	if (page->inuse > s->objects) {
732
733
		slab_err(s, page, "inuse %u > max %u",
			s->name, page->inuse, s->objects);
Christoph Lameter's avatar
Christoph Lameter committed
734
735
736
737
738
739
740
741
		return 0;
	}
	/* Slab_pad_check fixes things up after itself */
	slab_pad_check(s, page);
	return 1;
}

/*
Christoph Lameter's avatar
Christoph Lameter committed
742
743
 * Determine if a certain object on a page is on the freelist. Must hold the
 * slab lock to guarantee that the chains are in a consistent state.
Christoph Lameter's avatar
Christoph Lameter committed
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
 */
static int on_freelist(struct kmem_cache *s, struct page *page, void *search)
{
	int nr = 0;
	void *fp = page->freelist;
	void *object = NULL;

	while (fp && nr <= s->objects) {
		if (fp == search)
			return 1;
		if (!check_valid_pointer(s, page, fp)) {
			if (object) {
				object_err(s, page, object,
					"Freechain corrupt");
				set_freepointer(s, object, NULL);
				break;
			} else {
761
				slab_err(s, page, "Freepointer corrupt");
Christoph Lameter's avatar
Christoph Lameter committed
762
763
				page->freelist = NULL;
				page->inuse = s->objects;
764
				slab_fix(s, "Freelist cleared");
Christoph Lameter's avatar
Christoph Lameter committed
765
766
767
768
769
770
771
772
773
774
				return 0;
			}
			break;
		}
		object = fp;
		fp = get_freepointer(s, object);
		nr++;
	}

	if (page->inuse != s->objects - nr) {
775
		slab_err(s, page, "Wrong object count. Counter is %d but "
776
			"counted were %d", page->inuse, s->objects - nr);
Christoph Lameter's avatar
Christoph Lameter committed
777
		page->inuse = s->objects - nr;
778
		slab_fix(s, "Object count adjusted.");
Christoph Lameter's avatar
Christoph Lameter committed
779
780
781
782
	}
	return search == NULL;
}

Christoph Lameter's avatar
Christoph Lameter committed
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
static void trace(struct kmem_cache *s, struct page *page, void *object, int alloc)
{
	if (s->flags & SLAB_TRACE) {
		printk(KERN_INFO "TRACE %s %s 0x%p inuse=%d fp=0x%p\n",
			s->name,
			alloc ? "alloc" : "free",
			object, page->inuse,
			page->freelist);

		if (!alloc)
			print_section("Object", (void *)object, s->objsize);

		dump_stack();
	}
}

799
/*
Christoph Lameter's avatar
Christoph Lameter committed
800
 * Tracking of fully allocated slabs for debugging purposes.
801
 */
Christoph Lameter's avatar
Christoph Lameter committed
802
static void add_full(struct kmem_cache_node *n, struct page *page)
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
{
	spin_lock(&n->list_lock);
	list_add(&page->lru, &n->full);
	spin_unlock(&n->list_lock);
}

static void remove_full(struct kmem_cache *s, struct page *page)
{
	struct kmem_cache_node *n;

	if (!(s->flags & SLAB_STORE_USER))
		return;

	n = get_node(s, page_to_nid(page));

	spin_lock(&n->list_lock);
	list_del(&page->lru);
	spin_unlock(&n->list_lock);
}

Christoph Lameter's avatar
Christoph Lameter committed
823
824
825
826
827
828
829
830
831
832
833
834
static void setup_object_debug(struct kmem_cache *s, struct page *page,
								void *object)
{
	if (!(s->flags & (SLAB_STORE_USER|SLAB_RED_ZONE|__OBJECT_POISON)))
		return;

	init_object(s, object, 0);
	init_tracking(s, object);
}

static int alloc_debug_processing(struct kmem_cache *s, struct page *page,
						void *object, void *addr)
Christoph Lameter's avatar
Christoph Lameter committed
835
836
837
838
839
{
	if (!check_slab(s, page))
		goto bad;

	if (object && !on_freelist(s, page, object)) {
840
		object_err(s, page, object, "Object already allocated");
841
		goto bad;
Christoph Lameter's avatar
Christoph Lameter committed
842
843
844
845
	}

	if (!check_valid_pointer(s, page, object)) {
		object_err(s, page, object, "Freelist Pointer check fails");
846
		goto bad;
Christoph Lameter's avatar
Christoph Lameter committed
847
848
	}

Christoph Lameter's avatar
Christoph Lameter committed
849
	if (object && !check_object(s, page, object, 0))
Christoph Lameter's avatar
Christoph Lameter committed
850
851
		goto bad;

Christoph Lameter's avatar
Christoph Lameter committed
852
853
854
855
856
	/* Success perform special debug activities for allocs */
	if (s->flags & SLAB_STORE_USER)
		set_track(s, object, TRACK_ALLOC, addr);
	trace(s, page, object, 1);
	init_object(s, object, 1);
Christoph Lameter's avatar
Christoph Lameter committed
857
	return 1;
Christoph Lameter's avatar
Christoph Lameter committed
858

Christoph Lameter's avatar
Christoph Lameter committed
859
860
861
862
863
bad:
	if (PageSlab(page)) {
		/*
		 * If this is a slab page then lets do the best we can
		 * to avoid issues in the future. Marking all objects
Christoph Lameter's avatar
Christoph Lameter committed
864
		 * as used avoids touching the remaining objects.
Christoph Lameter's avatar
Christoph Lameter committed
865
		 */
866
		slab_fix(s, "Marking all objects used");
Christoph Lameter's avatar
Christoph Lameter committed
867
868
869
870
871
872
		page->inuse = s->objects;
		page->freelist = NULL;
	}
	return 0;
}

Christoph Lameter's avatar
Christoph Lameter committed
873
874
static int free_debug_processing(struct kmem_cache *s, struct page *page,
						void *object, void *addr)
Christoph Lameter's avatar
Christoph Lameter committed
875
876
877
878
879
{
	if (!check_slab(s, page))
		goto fail;

	if (!check_valid_pointer(s, page, object)) {
880
		slab_err(s, page, "Invalid object pointer 0x%p", object);
Christoph Lameter's avatar
Christoph Lameter committed
881
882
883
884
		goto fail;
	}

	if (on_freelist(s, page, object)) {
885
		object_err(s, page, object, "Object already free");
Christoph Lameter's avatar
Christoph Lameter committed
886
887
888
889
890
891
892
893
		goto fail;
	}

	if (!check_object(s, page, object, 1))
		return 0;

	if (unlikely(s != page->slab)) {
		if (!PageSlab(page))
894
895
			slab_err(s, page, "Attempt to free object(0x%p) "
				"outside of slab", object);
Christoph Lameter's avatar
Christoph Lameter committed
896
		else
897
		if (!page->slab) {
Christoph Lameter's avatar
Christoph Lameter committed
898
			printk(KERN_ERR
899
				"SLUB <none>: no slab for object 0x%p.\n",
Christoph Lameter's avatar
Christoph Lameter committed
900
						object);
901
902
			dump_stack();
		}
Christoph Lameter's avatar
Christoph Lameter committed
903
		else
904
905
			object_err(s, page, object,
					"page slab pointer corrupt.");
Christoph Lameter's avatar
Christoph Lameter committed
906
907
		goto fail;
	}
Christoph Lameter's avatar
Christoph Lameter committed
908
909
910
911
912
913
914
915

	/* Special debug activities for freeing objects */
	if (!SlabFrozen(page) && !page->freelist)
		remove_full(s, page);
	if (s->flags & SLAB_STORE_USER)
		set_track(s, object, TRACK_FREE, addr);
	trace(s, page, object, 0);
	init_object(s, object, 0);
Christoph Lameter's avatar
Christoph Lameter committed
916
	return 1;
Christoph Lameter's avatar
Christoph Lameter committed
917

Christoph Lameter's avatar
Christoph Lameter committed
918
fail:
919
	slab_fix(s, "Object at 0x%p not freed", object);
Christoph Lameter's avatar
Christoph Lameter committed
920
921
922
	return 0;
}

923
924
static int __init setup_slub_debug(char *str)
{
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
	slub_debug = DEBUG_DEFAULT_FLAGS;
	if (*str++ != '=' || !*str)
		/*
		 * No options specified. Switch on full debugging.
		 */
		goto out;

	if (*str == ',')
		/*
		 * No options but restriction on slabs. This means full
		 * debugging for slabs matching a pattern.
		 */
		goto check_slabs;

	slub_debug = 0;
	if (*str == '-')
		/*
		 * Switch off all debugging measures.
		 */
		goto out;

	/*
	 * Determine which debug features should be switched on
	 */
	for ( ;*str && *str != ','; str++) {
		switch (tolower(*str)) {
		case 'f':
			slub_debug |= SLAB_DEBUG_FREE;
			break;
		case 'z':
			slub_debug |= SLAB_RED_ZONE;
			break;
		case 'p':
			slub_debug |= SLAB_POISON;
			break;
		case 'u':
			slub_debug |= SLAB_STORE_USER;
			break;
		case 't':
			slub_debug |= SLAB_TRACE;
			break;
		default:
			printk(KERN_ERR "slub_debug option '%c' "
				"unknown. skipped\n",*str);
		}
970
971
	}

972
check_slabs:
973
974
	if (*str == ',')
		slub_debug_slabs = str + 1;
975
out:
976
977
978
979
980
	return 1;
}

__setup("slub_debug", setup_slub_debug);

981
982
static unsigned long kmem_cache_flags(unsigned long objsize,
	unsigned long flags, const char *name,
983
	void (*ctor)(struct kmem_cache *, void *))
984
985
986
987
988
989
990
991
992
993
{
	/*
	 * The page->offset field is only 16 bit wide. This is an offset
	 * in units of words from the beginning of an object. If the slab
	 * size is bigger then we cannot move the free pointer behind the
	 * object anymore.
	 *
	 * On 32 bit platforms the limit is 256k. On 64bit platforms
	 * the limit is 512k.
	 *
994
	 * Debugging or ctor may create a need to move the free
995
996
	 * pointer. Fail if this happens.
	 */
997
998
	if (objsize >= 65535 * sizeof(void *)) {
		BUG_ON(flags & (SLAB_RED_ZONE | SLAB_POISON |
999
				SLAB_STORE_USER | SLAB_DESTROY_BY_RCU));
1000
1001
		BUG_ON(ctor);
	} else {
1002
1003
1004
1005
		/*
		 * Enable debugging if selected on the kernel commandline.
		 */
		if (slub_debug && (!slub_debug_slabs ||
1006
		    strncmp(slub_debug_slabs, name,
1007
		    	strlen(slub_debug_slabs)) == 0))
1008
1009
1010
1011
				flags |= slub_debug;
	}

	return flags;
1012
1013
}
#else
Christoph Lameter's avatar
Christoph Lameter committed
1014
1015
static inline void setup_object_debug(struct kmem_cache *s,
			struct page *page, void *object) {}
1016

Christoph Lameter's avatar
Christoph Lameter committed
1017
1018
static inline int alloc_debug_processing(struct kmem_cache *s,
	struct page *page, void *object, void *addr) { return 0; }
1019

Christoph Lameter's avatar
Christoph Lameter committed
1020
1021
static inline int free_debug_processing(struct kmem_cache *s,
	struct page *page, void *object, void *addr) { return 0; }
1022
1023
1024
1025
1026

static inline int slab_pad_check(struct kmem_cache *s, struct page *page)
			{ return 1; }
static inline int check_object(struct kmem_cache *s, struct page *page,
			void *object, int active) { return 1; }
Christoph Lameter's avatar
Christoph Lameter committed
1027
static inline void add_full(struct kmem_cache_node *n, struct page *page) {}
1028
1029
static inline unsigned long kmem_cache_flags(unsigned long objsize,
	unsigned long flags, const char *name,
1030
	void (*ctor)(struct kmem_cache *, void *))
1031
1032
1033
{
	return flags;
}
1034
1035
#define slub_debug 0
#endif
Christoph Lameter's avatar
Christoph Lameter committed
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
/*
 * Slab allocation and freeing
 */
static struct page *allocate_slab(struct kmem_cache *s, gfp_t flags, int node)
{
	struct page * page;
	int pages = 1 << s->order;

	if (s->order)
		flags |= __GFP_COMP;

	if (s->flags & SLAB_CACHE_DMA)
		flags |= SLUB_DMA;

1050
1051
1052
	if (s->flags & SLAB_RECLAIM_ACCOUNT)
		flags |= __GFP_RECLAIMABLE;

Christoph Lameter's avatar
Christoph Lameter committed
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
	if (node == -1)
		page = alloc_pages(flags, s->order);
	else
		page = alloc_pages_node(node, flags, s->order);

	if (!page)
		return NULL;

	mod_zone_page_state(page_zone(page),
		(s->flags & SLAB_RECLAIM_ACCOUNT) ?
		NR_SLAB_RECLAIMABLE : NR_SLAB_UNRECLAIMABLE,
		pages);

	return page;
}

static void setup_object(struct kmem_cache *s, struct page *page,
				void *object)
{
Christoph Lameter's avatar
Christoph Lameter committed
1072
	setup_object_debug(s, page, object);
1073
	if (unlikely(s->ctor))
1074
		s->ctor(s, object);
Christoph Lameter's avatar
Christoph Lameter committed
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
}

static struct page *new_slab(struct kmem_cache *s, gfp_t flags, int node)
{
	struct page *page;
	struct kmem_cache_node *n;
	void *start;
	void *end;
	void *last;
	void *p;

Christoph Lameter's avatar
Christoph Lameter committed
1086
	BUG_ON(flags & GFP_SLAB_BUG_MASK);
Christoph Lameter's avatar
Christoph Lameter committed
1087

Christoph Lameter's avatar
Christoph Lameter committed
1088
1089
	page = allocate_slab(s,
		flags & (GFP_RECLAIM_MASK | GFP_CONSTRAINT_MASK), node);
Christoph Lameter's avatar
Christoph Lameter committed
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
	if (!page)
		goto out;

	n = get_node(s, page_to_nid(page));
	if (n)
		atomic_long_inc(&n->nr_slabs);
	page->slab = s;
	page->flags |= 1 << PG_slab;
	if (s->flags & (SLAB_DEBUG_FREE | SLAB_RED_ZONE | SLAB_POISON |
			SLAB_STORE_USER | SLAB_TRACE))
1100
		SetSlabDebug(page);
Christoph Lameter's avatar
Christoph Lameter committed
1101
1102
1103
1104
1105
1106
1107
1108

	start = page_address(page);
	end = start + s->objects * s->size;

	if (unlikely(s->flags & SLAB_POISON))
		memset(start, POISON_INUSE, PAGE_SIZE << s->order);

	last = start;
1109
	for_each_object(p, s, start) {
Christoph Lameter's avatar
Christoph Lameter committed
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
		setup_object(s, page, last);
		set_freepointer(s, last, p);
		last = p;
	}
	setup_object(s, page, last);
	set_freepointer(s, last, NULL);

	page->freelist = start;
	page->inuse = 0;
out:
	return page;
}

static void __free_slab(struct kmem_cache *s, struct page *page)
{
	int pages = 1 << s->order;

1127
	if (unlikely(SlabDebug(page))) {
Christoph Lameter's avatar
Christoph Lameter committed
1128
1129
1130
		void *p;

		slab_pad_check(s, page);
1131
		for_each_object(p, s, page_address(page))
Christoph Lameter's avatar
Christoph Lameter committed
1132
			check_object(s, page, p, 0);
1133
		ClearSlabDebug(page);
Christoph Lameter's avatar
Christoph Lameter committed
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
	}

	mod_zone_page_state(page_zone(page),
		(s->flags & SLAB_RECLAIM_ACCOUNT) ?
		NR_SLAB_RECLAIMABLE : NR_SLAB_UNRECLAIMABLE,
		- pages);

	__free_pages(page, s->order);
}

static void rcu_free_slab(struct rcu_head *h)
{
	struct page *page;

	page = container_of((struct list_head *)h, struct page, lru);
	__free_slab(page->slab, page);
}

static void free_slab(struct kmem_cache *s, struct page *page)
{
	if (unlikely(s->flags & SLAB_DESTROY_BY_RCU)) {
		/*
		 * RCU free overloads the RCU head over the LRU
		 */
		struct rcu_head *head = (void *)&page->lru;

		call_rcu(head, rcu_free_slab);
	} else
		__free_slab(s, page);
}

static void discard_slab(struct kmem_cache *s, struct page *page)
{
	struct kmem_cache_node *n = get_node(s, page_to_nid(page));

	atomic_long_dec(&n->nr_slabs);
	reset_page_mapcount(page);
1171
	__ClearPageSlab(page);
Christoph Lameter's avatar
Christoph Lameter committed
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
	free_slab(s, page);
}

/*
 * Per slab locking using the pagelock
 */
static __always_inline void slab_lock(struct page *page)
{
	bit_spin_lock(PG_locked, &page->flags);
}

static __always_inline void slab_unlock(struct page *page)
{
	bit_spin_unlock(PG_locked, &page->flags);
}

static __always_inline int slab_trylock(struct page *page)
{
	int rc = 1;

	rc = bit_spin_trylock(PG_locked, &page->flags);
	return rc;
}

/*
 * Management of partially allocated slabs
 */
Christoph Lameter's avatar
Christoph Lameter committed
1199
static void add_partial_tail(struct kmem_cache_node *n, struct page *page)
Christoph Lameter's avatar
Christoph Lameter committed
1200
{
Christoph Lameter's avatar
Christoph Lameter committed
1201
1202
1203
1204
1205
	spin_lock(&n->list_lock);
	n->nr_partial++;
	list_add_tail(&page->lru, &n->partial);
	spin_unlock(&n->list_lock);
}
Christoph Lameter's avatar
Christoph Lameter committed
1206

Christoph Lameter's avatar
Christoph Lameter committed
1207
1208
static void add_partial(struct kmem_cache_node *n, struct page *page)
{
Christoph Lameter's avatar
Christoph Lameter committed
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
	spin_lock(&n->list_lock);
	n->nr_partial++;
	list_add(&page->lru, &n->partial);
	spin_unlock(&n->list_lock);
}

static void remove_partial(struct kmem_cache *s,
						struct page *page)
{
	struct kmem_cache_node *n = get_node(s, page_to_nid(page));

	spin_lock(&n->list_lock);
	list_del(&page->lru);
	n->nr_partial--;
	spin_unlock(&n->list_lock);
}

/*
Christoph Lameter's avatar
Christoph Lameter committed
1227
 * Lock slab and remove from the partial list.
Christoph Lameter's avatar
Christoph Lameter committed
1228
 *
Christoph Lameter's avatar
Christoph Lameter committed
1229
 * Must hold list_lock.
Christoph Lameter's avatar
Christoph Lameter committed
1230
 */
1231
static inline int lock_and_freeze_slab(struct kmem_cache_node *n, struct page *page)
Christoph Lameter's avatar
Christoph Lameter committed
1232
1233
1234
1235
{
	if (slab_trylock(page)) {
		list_del(&page->lru);
		n->nr_partial--;
1236
		SetSlabFrozen(page);
Christoph Lameter's avatar
Christoph Lameter committed
1237
1238
1239
1240
1241
1242
		return 1;
	}
	return 0;
}

/*
Christoph Lameter's avatar
Christoph Lameter committed
1243
 * Try to allocate a partial slab from a specific node.
Christoph Lameter's avatar
Christoph Lameter committed
1244
1245
1246
1247
1248
1249
1250
1251
 */
static struct page *get_partial_node(struct kmem_cache_node *n)
{
	struct page *page;

	/*
	 * Racy check. If we mistakenly see no partial slabs then we
	 * just allocate an empty slab. If we mistakenly try to get a
Christoph Lameter's avatar
Christoph Lameter committed
1252
1253
	 * partial slab and there is none available then get_partials()
	 * will return NULL.
Christoph Lameter's avatar
Christoph Lameter committed
1254
1255
1256
1257
1258
1259
	 */
	if (!n || !n->nr_partial)
		return NULL;

	spin_lock(&n->list_lock);
	list_for_each_entry(page, &n->partial, lru)
1260
		if (lock_and_freeze_slab(n, page))
Christoph Lameter's avatar
Christoph Lameter committed
1261
1262
1263
1264
1265
1266
1267
1268
			goto out;
	page = NULL;
out:
	spin_unlock(&n->list_lock);
	return page;
}

/*
Christoph Lameter's avatar
Christoph Lameter committed
1269
 * Get a page from somewhere. Search in increasing NUMA distances.
Christoph Lameter's avatar
Christoph Lameter committed
1270
1271
1272
1273
1274
1275
1276
1277
1278
 */
static struct page *get_any_partial(struct kmem_cache *s, gfp_t flags)
{
#ifdef CONFIG_NUMA
	struct zonelist *zonelist;
	struct zone **z;
	struct page *page;

	/*
Christoph Lameter's avatar
Christoph Lameter committed
1279
1280
1281
1282
	 * The defrag ratio allows a configuration of the tradeoffs between
	 * inter node defragmentation and node local allocations. A lower
	 * defrag_ratio increases the tendency to do local allocations
	 * instead of attempting to obtain partial slabs from other nodes.
Christoph Lameter's avatar
Christoph Lameter committed
1283
	 *
Christoph Lameter's avatar
Christoph Lameter committed
1284
1285
1286
1287
	 * If the defrag_ratio is set to 0 then kmalloc() always
	 * returns node local objects. If the ratio is higher then kmalloc()
	 * may return off node objects because partial slabs are obtained
	 * from other nodes and filled up.
Christoph Lameter's avatar
Christoph Lameter committed
1288
1289
	 *
	 * If /sys/slab/xx/defrag_ratio is set to 100 (which makes
Christoph Lameter's avatar
Christoph Lameter committed
1290
1291
1292
1293
1294
	 * defrag_ratio = 1000) then every (well almost) allocation will
	 * first attempt to defrag slab caches on other nodes. This means
	 * scanning over all nodes to look for partial slabs which may be
	 * expensive if we do it every time we are trying to find a slab
	 * with available objects.
Christoph Lameter's avatar
Christoph Lameter committed
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
	 */
	if (!s->defrag_ratio || get_cycles() % 1024 > s->defrag_ratio)
		return NULL;

	zonelist = &NODE_DATA(slab_node(current->mempolicy))
					->node_zonelists[gfp_zone(flags)];
	for (z = zonelist->zones; *z; z++) {
		struct kmem_cache_node *n;

		n = get_node(s, zone_to_nid(*z));

		if (n && cpuset_zone_allowed_hardwall(*z, flags) &&
Christoph Lameter's avatar
Christoph Lameter committed
1307
				n->nr_partial > MIN_PARTIAL) {
Christoph Lameter's avatar
Christoph Lameter committed
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
			page = get_partial_node(n);
			if (page)
				return page;
		}
	}
#endif
	return NULL;
}

/*
 * Get a partial page, lock it and return it.
 */
static struct page *get_partial(struct kmem_cache *s, gfp_t flags, int node)
{
	struct page *page;
	int searchnode = (node == -1) ? numa_node_id() : node;

	page = get_partial_node(get_node(s, searchnode));
	if (page || (flags & __GFP_THISNODE))
		return page;

	return get_any_partial(s, flags);
}

/*
 * Move a page back to the lists.
 *
 * Must be called with the slab lock held.
 *
 * On exit the slab lock will have been dropped.
 */
1339
static void unfreeze_slab(struct kmem_cache *s, struct page *page)
Christoph Lameter's avatar
Christoph Lameter committed
1340
{
Christoph Lameter's avatar
Christoph Lameter committed
1341
1342
	struct kmem_cache_node *n = get_node(s, page_to_nid(page));

1343
	ClearSlabFrozen(page);
Christoph Lameter's avatar
Christoph Lameter committed
1344
	if (page->inuse) {
Christoph Lameter's avatar
Christoph Lameter committed
1345

Christoph Lameter's avatar
Christoph Lameter committed
1346
		if (page->freelist)
Christoph Lameter's avatar
Christoph Lameter committed
1347
			add_partial(n, page);
1348
		else if (SlabDebug(page) && (s->flags & SLAB_STORE_USER))
Christoph Lameter's avatar
Christoph Lameter committed
1349
			add_full(n, page);
Christoph Lameter's avatar
Christoph Lameter committed
1350
		slab_unlock(page);
Christoph Lameter's avatar
Christoph Lameter committed
1351

Christoph Lameter's avatar
Christoph Lameter committed
1352
	} else {
Christoph Lameter's avatar
Christoph Lameter committed
1353
1354
		if (n->nr_partial < MIN_PARTIAL) {
			/*
Christoph Lameter's avatar
Christoph Lameter committed
1355
1356
1357
1358
1359
1360
			 * Adding an empty slab to the partial slabs in order
			 * to avoid page allocator overhead. This slab needs
			 * to come after the other slabs with objects in
			 * order to fill them up. That way the size of the
			 * partial list stays small. kmem_cache_shrink can
			 * reclaim empty slabs from the partial list.
Christoph Lameter's avatar
Christoph Lameter committed
1361
1362
1363
1364
1365
1366
1367
			 */
			add_partial_tail(n, page);
			slab_unlock(page);
		} else {
			slab_unlock(page);
			discard_slab(s, page);
		}
Christoph Lameter's avatar
Christoph Lameter committed
1368
1369
1370
1371
1372
1373
	}
}

/*
 * Remove the cpu slab
 */
1374
static void deactivate_slab(struct kmem_cache *s, struct kmem_cache_cpu *c)
Christoph Lameter's avatar
Christoph Lameter committed
1375
{
1376
	struct page *page = c->page;
1377
1378
1379
1380
1381
	/*
	 * Merge cpu freelist into freelist. Typically we get here
	 * because both freelists are empty. So this is unlikely
	 * to occur.
	 */
1382
	while (unlikely(c->freelist)) {
1383
1384
1385
		void **object;

		/* Retrieve object from cpu_freelist */
1386
		object = c->freelist;
1387
		c->freelist = c->freelist[c->offset];
1388
1389

		/* And put onto the regular freelist */
1390
		object[c->offset] = page->freelist;
1391
1392
1393
		page->freelist = object;
		page->inuse--;
	}
1394
	c->page = NULL;
1395
	unfreeze_slab(s, page);
Christoph Lameter's avatar
Christoph Lameter committed
1396
1397
}

1398
static inline void flush_slab(struct kmem_cache *s, struct kmem_cache_cpu *c)
Christoph Lameter's avatar
Christoph Lameter committed
1399
{
1400
1401
	slab_lock(c->page);
	deactivate_slab(s, c);
Christoph Lameter's avatar
Christoph Lameter committed
1402
1403
1404
1405
1406
1407
}

/*
 * Flush cpu slab.
 * Called from IPI handler with interrupts disabled.
 */
1408
static inline void __flush_cpu_slab(struct kmem_cache *s, int cpu)
Christoph Lameter's avatar
Christoph Lameter committed
1409
{
1410
	struct kmem_cache_cpu *c = get_cpu_slab(s, cpu);
Christoph Lameter's avatar
Christoph Lameter committed
1411

1412
1413
	if (likely(c && c->page))
		flush_slab(s, c);
Christoph Lameter's avatar
Christoph Lameter committed
1414
1415
1416
1417
1418
1419
}

static void flush_cpu_slab(void *d)
{
	struct kmem_cache *s = d;

1420
	__flush_cpu_slab(s, smp_processor_id());
Christoph Lameter's avatar
Christoph Lameter committed
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
}

static void flush_all(struct kmem_cache *s)
{
#ifdef CONFIG_SMP
	on_each_cpu(flush_cpu_slab, s, 1, 1);
#else
	unsigned long flags;

	local_irq_save(flags);
	flush_cpu_slab(s);
	local_irq_restore(flags);
#endif
}

1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
/*
 * Check if the objects in a per cpu structure fit numa
 * locality expectations.
 */
static inline int node_match(struct kmem_cache_cpu *c, int node)
{
#ifdef CONFIG_NUMA
	if (node != -1 && c->node != node)
		return 0;
#endif
	return 1;
}

Christoph Lameter's avatar
Christoph Lameter committed
1449
/*
1450
1451
1452
1453
 * Slow path. The lockless freelist is empty or we need to perform
 * debugging duties.
 *
 * Interrupts are disabled.
Christoph Lameter's avatar
Christoph Lameter committed
1454
 *
1455
1456
1457
 * Processing is still very fast if new objects have been freed to the
 * regular freelist. In that case we simply take over the regular freelist
 * as the lockless freelist and zap the regular freelist.
Christoph Lameter's avatar
Christoph Lameter committed
1458
 *
1459
1460
1461
 * If that is not working then we fall back to the partial lists. We take the
 * first element of the freelist as the object to allocate now and move the
 * rest of the freelist to the lockless freelist.
Christoph Lameter's avatar
Christoph Lameter committed
1462
 *
1463
1464
 * And if we were unable to get a new slab from the partial slab lists then
 * we need to allocate a new slab. This is slowest path since we may sleep.
Christoph Lameter's avatar
Christoph Lameter committed
1465
 */
1466
static void *__slab_alloc(struct kmem_cache *s,
1467
		gfp_t gfpflags, int node, void *addr, struct kmem_cache_cpu *c)
Christoph Lameter's avatar
Christoph Lameter committed
1468
1469
{
	void **object;
1470
	struct page *new;
Christoph Lameter's avatar
Christoph Lameter committed
1471

1472
	if (!c->page)
Christoph Lameter's avatar
Christoph Lameter committed
1473
1474
		goto new_slab;

1475
1476
	slab_lock(c->page);
	if (unlikely(!node_match(c, node)))
Christoph Lameter's avatar
Christoph Lameter committed
1477
		goto another_slab;
1478
load_freelist:
1479
	object = c->page->freelist;
Christoph Lameter's avatar
Christoph Lameter committed
1480
1481
	if (unlikely(!object))
		goto another_slab;
1482
	if (unlikely(SlabDebug(c->page)))
Christoph Lameter's avatar
Christoph Lameter committed
1483
1484
		goto debug;

1485
	object = c->page->freelist;
1486
	c->freelist = object[c->offset];
1487
1488
1489
1490
	c->page->inuse = s->objects;
	c->page->freelist = NULL;
	c->node = page_to_nid(c->page);
	slab_unlock(c->page);
Christoph Lameter's avatar
Christoph Lameter committed
1491
1492
1493
	return object;

another_slab:
1494
	deactivate_slab(s, c);
Christoph Lameter's avatar
Christoph Lameter committed
1495
1496

new_slab:
1497
1498
1499
	new = get_partial(s, gfpflags, node);
	if (new) {
		c->page = new;
1500
		goto load_freelist;
Christoph Lameter's avatar
Christoph Lameter committed
1501
1502
	}

1503
1504
1505
	if (gfpflags & __GFP_WAIT)
		local_irq_enable();

1506
	new = new_slab(s, gfpflags, node);
1507
1508
1509
1510

	if (gfpflags & __GFP_WAIT)
		local_irq_disable();

1511
1512
1513
	if (new) {
		c = get_cpu_slab(s, smp_processor_id());
		if (c->page) {
Christoph Lameter's avatar
Christoph Lameter committed
1514
			/*
Christoph Lameter's avatar
Christoph Lameter committed
1515
1516
1517
1518
1519
			 * Someone else populated the cpu_slab while we
			 * enabled interrupts, or we have gotten scheduled
			 * on another cpu. The page may not be on the
			 * requested node even if __GFP_THISNODE was
			 * specified. So we need to recheck.
Christoph Lameter's avatar
Christoph Lameter committed
1520
			 */
1521
			if (node_match(c, node)) {
Christoph Lameter's avatar
Christoph Lameter committed
1522
1523
1524
1525
				/*
				 * Current cpuslab is acceptable and we
				 * want the current one since its cache hot
				 */
1526
1527
				discard_slab(s, new);
				slab_lock(c->page);
1528
				goto load_freelist;
Christoph Lameter's avatar
Christoph Lameter committed
1529
			}