slab.h 23.8 KB
Newer Older
1
/* SPDX-License-Identifier: GPL-2.0 */
L
Linus Torvalds 已提交
2
/*
3 4
 * Written by Mark Hemment, 1996 (markhe@nextd.demon.co.uk).
 *
C
Christoph Lameter 已提交
5
 * (C) SGI 2006, Christoph Lameter
6 7
 * 	Cleaned up and restructured to ease the addition of alternative
 * 	implementations of SLAB allocators.
8 9
 * (C) Linux Foundation 2008-2013
 *      Unified interface for all slab allocators
L
Linus Torvalds 已提交
10 11 12 13 14
 */

#ifndef _LINUX_SLAB_H
#define	_LINUX_SLAB_H

15
#include <linux/gfp.h>
16
#include <linux/overflow.h>
17
#include <linux/types.h>
G
Glauber Costa 已提交
18
#include <linux/workqueue.h>
19
#include <linux/percpu-refcount.h>
G
Glauber Costa 已提交
20

L
Linus Torvalds 已提交
21

22 23
/*
 * Flags to pass to kmem_cache_create().
24
 * The ones marked DEBUG are only valid if CONFIG_DEBUG_SLAB is set.
L
Linus Torvalds 已提交
25
 */
26
/* DEBUG: Perform (expensive) checks on alloc/free */
27
#define SLAB_CONSISTENCY_CHECKS	((slab_flags_t __force)0x00000100U)
28
/* DEBUG: Red zone objs in a cache */
29
#define SLAB_RED_ZONE		((slab_flags_t __force)0x00000400U)
30
/* DEBUG: Poison objects */
31
#define SLAB_POISON		((slab_flags_t __force)0x00000800U)
32
/* Align objs on cache lines */
33
#define SLAB_HWCACHE_ALIGN	((slab_flags_t __force)0x00002000U)
34
/* Use GFP_DMA memory */
35
#define SLAB_CACHE_DMA		((slab_flags_t __force)0x00004000U)
36 37
/* Use GFP_DMA32 memory */
#define SLAB_CACHE_DMA32	((slab_flags_t __force)0x00008000U)
38
/* DEBUG: Store the last owner for bug hunting */
39
#define SLAB_STORE_USER		((slab_flags_t __force)0x00010000U)
40
/* Panic if kmem_cache_create() fails */
41
#define SLAB_PANIC		((slab_flags_t __force)0x00040000U)
42
/*
43
 * SLAB_TYPESAFE_BY_RCU - **WARNING** READ THIS!
44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67
 *
 * This delays freeing the SLAB page by a grace period, it does _NOT_
 * delay object freeing. This means that if you do kmem_cache_free()
 * that memory location is free to be reused at any time. Thus it may
 * be possible to see another object there in the same RCU grace period.
 *
 * This feature only ensures the memory location backing the object
 * stays valid, the trick to using this is relying on an independent
 * object validation pass. Something like:
 *
 *  rcu_read_lock()
 * again:
 *  obj = lockless_lookup(key);
 *  if (obj) {
 *    if (!try_get_ref(obj)) // might fail for free objects
 *      goto again;
 *
 *    if (obj->key != key) { // not the object we expected
 *      put_ref(obj);
 *      goto again;
 *    }
 *  }
 *  rcu_read_unlock();
 *
68 69 70 71 72 73 74 75
 * 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.
76 77
 *
 * Note that SLAB_TYPESAFE_BY_RCU was originally named SLAB_DESTROY_BY_RCU.
78
 */
79
/* Defer freeing slabs to RCU */
80
#define SLAB_TYPESAFE_BY_RCU	((slab_flags_t __force)0x00080000U)
81
/* Spread some memory over cpuset */
82
#define SLAB_MEM_SPREAD		((slab_flags_t __force)0x00100000U)
83
/* Trace allocations and frees */
84
#define SLAB_TRACE		((slab_flags_t __force)0x00200000U)
L
Linus Torvalds 已提交
85

86 87
/* Flag to prevent checks on free */
#ifdef CONFIG_DEBUG_OBJECTS
88
# define SLAB_DEBUG_OBJECTS	((slab_flags_t __force)0x00400000U)
89
#else
90
# define SLAB_DEBUG_OBJECTS	0
91 92
#endif

93
/* Avoid kmemleak tracing */
94
#define SLAB_NOLEAKTRACE	((slab_flags_t __force)0x00800000U)
95

96
/* Fault injection mark */
97
#ifdef CONFIG_FAILSLAB
98
# define SLAB_FAILSLAB		((slab_flags_t __force)0x02000000U)
99
#else
100
# define SLAB_FAILSLAB		0
101
#endif
102
/* Account to memcg */
103
#ifdef CONFIG_MEMCG_KMEM
104
# define SLAB_ACCOUNT		((slab_flags_t __force)0x04000000U)
V
Vladimir Davydov 已提交
105
#else
106
# define SLAB_ACCOUNT		0
V
Vladimir Davydov 已提交
107
#endif
V
Vegard Nossum 已提交
108

A
Alexander Potapenko 已提交
109
#ifdef CONFIG_KASAN
110
#define SLAB_KASAN		((slab_flags_t __force)0x08000000U)
A
Alexander Potapenko 已提交
111
#else
112
#define SLAB_KASAN		0
A
Alexander Potapenko 已提交
113 114
#endif

115
/* The following flags affect the page allocator grouping pages by mobility */
116
/* Objects are reclaimable */
117
#define SLAB_RECLAIM_ACCOUNT	((slab_flags_t __force)0x00020000U)
118
#define SLAB_TEMPORARY		SLAB_RECLAIM_ACCOUNT	/* Objects are short-lived */
119 120 121 122

/* Slab deactivation flag */
#define SLAB_DEACTIVATED	((slab_flags_t __force)0x10000000U)

123 124 125 126 127 128 129 130 131 132
/*
 * ZERO_SIZE_PTR will be returned for zero sized kmalloc requests.
 *
 * Dereferencing ZERO_SIZE_PTR will lead to a distinct access fault.
 *
 * ZERO_SIZE_PTR can be passed to kfree though in the same way that NULL can.
 * Both make kfree a no-op.
 */
#define ZERO_SIZE_PTR ((void *)16)

133
#define ZERO_OR_NULL_PTR(x) ((unsigned long)(x) <= \
134 135
				(unsigned long)ZERO_SIZE_PTR)

136
#include <linux/kasan.h>
137

138
struct mem_cgroup;
139 140 141 142
/*
 * struct kmem_cache related prototypes
 */
void __init kmem_cache_init(void);
143
bool slab_is_available(void);
L
Linus Torvalds 已提交
144

145 146
extern bool usercopy_fallback;

147 148
struct kmem_cache *kmem_cache_create(const char *name, unsigned int size,
			unsigned int align, slab_flags_t flags,
149 150
			void (*ctor)(void *));
struct kmem_cache *kmem_cache_create_usercopy(const char *name,
151 152
			unsigned int size, unsigned int align,
			slab_flags_t flags,
153
			unsigned int useroffset, unsigned int usersize,
154
			void (*ctor)(void *));
K
Kees Cook 已提交
155 156
void kmem_cache_destroy(struct kmem_cache *s);
int kmem_cache_shrink(struct kmem_cache *s);
157

158 159 160 161 162 163 164 165
/*
 * Please use this macro to create slab caches. Simply specify the
 * name of the structure and maybe some flags that are listed above.
 *
 * The alignment of the struct determines object alignment. If you
 * f.e. add ____cacheline_aligned_in_smp to the struct declaration
 * then the objects will be properly aligned in SMP configurations.
 */
166 167 168 169 170 171 172 173 174 175 176 177 178 179
#define KMEM_CACHE(__struct, __flags)					\
		kmem_cache_create(#__struct, sizeof(struct __struct),	\
			__alignof__(struct __struct), (__flags), NULL)

/*
 * To whitelist a single field for copying to/from usercopy, use this
 * macro instead for KMEM_CACHE() above.
 */
#define KMEM_CACHE_USERCOPY(__struct, __flags, __field)			\
		kmem_cache_create_usercopy(#__struct,			\
			sizeof(struct __struct),			\
			__alignof__(struct __struct), (__flags),	\
			offsetof(struct __struct, __field),		\
			sizeof_field(struct __struct, __field), NULL)
180

181 182 183
/*
 * Common kmalloc functions provided by all allocators
 */
184
void * __must_check krealloc(const void *objp, size_t new_size, gfp_t flags) __alloc_size(2);
K
Kees Cook 已提交
185 186 187 188
void kfree(const void *objp);
void kfree_sensitive(const void *objp);
size_t __ksize(const void *objp);
size_t ksize(const void *objp);
189
#ifdef CONFIG_PRINTK
190 191
bool kmem_valid_obj(void *object);
void kmem_dump_obj(void *object);
192
#endif
193

K
Kees Cook 已提交
194
#ifdef CONFIG_HAVE_HARDENED_USERCOPY_ALLOCATOR
195 196
void __check_heap_object(const void *ptr, unsigned long n, struct page *page,
			bool to_user);
K
Kees Cook 已提交
197
#else
198 199
static inline void __check_heap_object(const void *ptr, unsigned long n,
				       struct page *page, bool to_user) { }
K
Kees Cook 已提交
200 201
#endif

202 203 204 205 206 207 208 209 210 211 212 213 214
/*
 * Some archs want to perform DMA into kmalloc caches and need a guaranteed
 * alignment larger than the alignment of a 64-bit integer.
 * Setting ARCH_KMALLOC_MINALIGN in arch headers allows that.
 */
#if defined(ARCH_DMA_MINALIGN) && ARCH_DMA_MINALIGN > 8
#define ARCH_KMALLOC_MINALIGN ARCH_DMA_MINALIGN
#define KMALLOC_MIN_SIZE ARCH_DMA_MINALIGN
#define KMALLOC_SHIFT_LOW ilog2(ARCH_DMA_MINALIGN)
#else
#define ARCH_KMALLOC_MINALIGN __alignof__(unsigned long long)
#endif

215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232
/*
 * Setting ARCH_SLAB_MINALIGN in arch headers allows a different alignment.
 * Intended for arches that get misalignment faults even for 64 bit integer
 * aligned buffers.
 */
#ifndef ARCH_SLAB_MINALIGN
#define ARCH_SLAB_MINALIGN __alignof__(unsigned long long)
#endif

/*
 * kmalloc and friends return ARCH_KMALLOC_MINALIGN aligned
 * pointers. kmem_cache_alloc and friends return ARCH_SLAB_MINALIGN
 * aligned pointers.
 */
#define __assume_kmalloc_alignment __assume_aligned(ARCH_KMALLOC_MINALIGN)
#define __assume_slab_alignment __assume_aligned(ARCH_SLAB_MINALIGN)
#define __assume_page_alignment __assume_aligned(PAGE_SIZE)

233
/*
234 235 236 237 238 239
 * Kmalloc array related definitions
 */

#ifdef CONFIG_SLAB
/*
 * The largest kmalloc size supported by the SLAB allocators is
240 241 242 243 244 245 246
 * 32 megabyte (2^25) or the maximum allocatable page order if that is
 * less than 32 MB.
 *
 * WARNING: Its not easy to increase this value since the allocators have
 * to do various tricks to work around compiler limitations in order to
 * ensure proper constant folding.
 */
247 248
#define KMALLOC_SHIFT_HIGH	((MAX_ORDER + PAGE_SHIFT - 1) <= 25 ? \
				(MAX_ORDER + PAGE_SHIFT - 1) : 25)
249
#define KMALLOC_SHIFT_MAX	KMALLOC_SHIFT_HIGH
250
#ifndef KMALLOC_SHIFT_LOW
251
#define KMALLOC_SHIFT_LOW	5
252
#endif
253 254 255
#endif

#ifdef CONFIG_SLUB
256
/*
257 258
 * SLUB directly allocates requests fitting in to an order-1 page
 * (PAGE_SIZE*2).  Larger requests are passed to the page allocator.
259 260
 */
#define KMALLOC_SHIFT_HIGH	(PAGE_SHIFT + 1)
261
#define KMALLOC_SHIFT_MAX	(MAX_ORDER + PAGE_SHIFT - 1)
262
#ifndef KMALLOC_SHIFT_LOW
263 264
#define KMALLOC_SHIFT_LOW	3
#endif
265
#endif
266

267 268
#ifdef CONFIG_SLOB
/*
269
 * SLOB passes all requests larger than one page to the page allocator.
270 271 272 273
 * No kmalloc array is necessary since objects of different sizes can
 * be allocated from the same page.
 */
#define KMALLOC_SHIFT_HIGH	PAGE_SHIFT
274
#define KMALLOC_SHIFT_MAX	(MAX_ORDER + PAGE_SHIFT - 1)
275 276 277 278 279
#ifndef KMALLOC_SHIFT_LOW
#define KMALLOC_SHIFT_LOW	3
#endif
#endif

280 281 282 283
/* Maximum allocatable size */
#define KMALLOC_MAX_SIZE	(1UL << KMALLOC_SHIFT_MAX)
/* Maximum size for which we actually use a slab cache */
#define KMALLOC_MAX_CACHE_SIZE	(1UL << KMALLOC_SHIFT_HIGH)
284
/* Maximum order allocatable via the slab allocator */
285
#define KMALLOC_MAX_ORDER	(KMALLOC_SHIFT_MAX - PAGE_SHIFT)
286

287 288 289
/*
 * Kmalloc subsystem.
 */
290
#ifndef KMALLOC_MIN_SIZE
291
#define KMALLOC_MIN_SIZE (1 << KMALLOC_SHIFT_LOW)
292 293
#endif

J
Joonsoo Kim 已提交
294 295 296 297 298 299 300 301 302 303 304
/*
 * This restriction comes from byte sized index implementation.
 * Page size is normally 2^12 bytes and, in this case, if we want to use
 * byte sized index which can represent 2^8 entries, the size of the object
 * should be equal or greater to 2^12 / 2^8 = 2^4 = 16.
 * If minimum size of kmalloc is less than 16, we use it as minimum object
 * size and give up to use byte sized index.
 */
#define SLAB_OBJ_MIN_SIZE      (KMALLOC_MIN_SIZE < 16 ? \
                               (KMALLOC_MIN_SIZE) : 16)

305 306 307
/*
 * Whenever changing this, take care of that kmalloc_type() and
 * create_kmalloc_caches() still work as intended.
308 309 310 311
 *
 * KMALLOC_NORMAL can contain only unaccounted objects whereas KMALLOC_CGROUP
 * is for accounted but unreclaimable and non-dma objects. All the other
 * kmem caches can have both accounted and unaccounted objects.
312
 */
313 314
enum kmalloc_cache_type {
	KMALLOC_NORMAL = 0,
315 316 317 318 319 320 321 322
#ifndef CONFIG_ZONE_DMA
	KMALLOC_DMA = KMALLOC_NORMAL,
#endif
#ifndef CONFIG_MEMCG_KMEM
	KMALLOC_CGROUP = KMALLOC_NORMAL,
#else
	KMALLOC_CGROUP,
#endif
323
	KMALLOC_RECLAIM,
324 325 326 327 328 329
#ifdef CONFIG_ZONE_DMA
	KMALLOC_DMA,
#endif
	NR_KMALLOC_TYPES
};

330
#ifndef CONFIG_SLOB
331 332 333
extern struct kmem_cache *
kmalloc_caches[NR_KMALLOC_TYPES][KMALLOC_SHIFT_HIGH + 1];

334 335 336 337 338 339 340 341
/*
 * Define gfp bits that should not be set for KMALLOC_NORMAL.
 */
#define KMALLOC_NOT_NORMAL_BITS					\
	(__GFP_RECLAIMABLE |					\
	(IS_ENABLED(CONFIG_ZONE_DMA)   ? __GFP_DMA : 0) |	\
	(IS_ENABLED(CONFIG_MEMCG_KMEM) ? __GFP_ACCOUNT : 0))

342 343
static __always_inline enum kmalloc_cache_type kmalloc_type(gfp_t flags)
{
344 345
	/*
	 * The most common case is KMALLOC_NORMAL, so test for it
346
	 * with a single branch for all the relevant flags.
347
	 */
348
	if (likely((flags & KMALLOC_NOT_NORMAL_BITS) == 0))
349
		return KMALLOC_NORMAL;
350 351

	/*
352 353 354 355 356
	 * At least one of the flags has to be set. Their priorities in
	 * decreasing order are:
	 *  1) __GFP_DMA
	 *  2) __GFP_RECLAIMABLE
	 *  3) __GFP_ACCOUNT
357
	 */
358 359 360 361 362 363
	if (IS_ENABLED(CONFIG_ZONE_DMA) && (flags & __GFP_DMA))
		return KMALLOC_DMA;
	if (!IS_ENABLED(CONFIG_MEMCG_KMEM) || (flags & __GFP_RECLAIMABLE))
		return KMALLOC_RECLAIM;
	else
		return KMALLOC_CGROUP;
364 365
}

366 367 368 369 370
/*
 * Figure out which kmalloc slab an allocation of a certain size
 * belongs to.
 * 0 = zero alloc
 * 1 =  65 .. 96 bytes
371 372
 * 2 = 129 .. 192 bytes
 * n = 2^(n-1)+1 .. 2^n
373 374 375 376 377
 *
 * Note: __kmalloc_index() is compile-time optimized, and not runtime optimized;
 * typical usage is via kmalloc_index() and therefore evaluated at compile-time.
 * Callers where !size_is_constant should only be test modules, where runtime
 * overheads of __kmalloc_index() can be tolerated.  Also see kmalloc_slab().
378
 */
379 380
static __always_inline unsigned int __kmalloc_index(size_t size,
						    bool size_is_constant)
381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414
{
	if (!size)
		return 0;

	if (size <= KMALLOC_MIN_SIZE)
		return KMALLOC_SHIFT_LOW;

	if (KMALLOC_MIN_SIZE <= 32 && size > 64 && size <= 96)
		return 1;
	if (KMALLOC_MIN_SIZE <= 64 && size > 128 && size <= 192)
		return 2;
	if (size <=          8) return 3;
	if (size <=         16) return 4;
	if (size <=         32) return 5;
	if (size <=         64) return 6;
	if (size <=        128) return 7;
	if (size <=        256) return 8;
	if (size <=        512) return 9;
	if (size <=       1024) return 10;
	if (size <=   2 * 1024) return 11;
	if (size <=   4 * 1024) return 12;
	if (size <=   8 * 1024) return 13;
	if (size <=  16 * 1024) return 14;
	if (size <=  32 * 1024) return 15;
	if (size <=  64 * 1024) return 16;
	if (size <= 128 * 1024) return 17;
	if (size <= 256 * 1024) return 18;
	if (size <= 512 * 1024) return 19;
	if (size <= 1024 * 1024) return 20;
	if (size <=  2 * 1024 * 1024) return 21;
	if (size <=  4 * 1024 * 1024) return 22;
	if (size <=  8 * 1024 * 1024) return 23;
	if (size <=  16 * 1024 * 1024) return 24;
	if (size <=  32 * 1024 * 1024) return 25;
415 416 417 418 419 420

	if ((IS_ENABLED(CONFIG_CC_IS_GCC) || CONFIG_CLANG_VERSION >= 110000)
	    && !IS_ENABLED(CONFIG_PROFILE_ALL_BRANCHES) && size_is_constant)
		BUILD_BUG_ON_MSG(1, "unexpected size in kmalloc_index()");
	else
		BUG();
421 422 423 424

	/* Will never be reached. Needed because the compiler may complain */
	return -1;
}
425
#define kmalloc_index(s) __kmalloc_index(s, true)
426
#endif /* !CONFIG_SLOB */
427

428
void *__kmalloc(size_t size, gfp_t flags) __assume_kmalloc_alignment __alloc_size(1);
K
Kees Cook 已提交
429 430
void *kmem_cache_alloc(struct kmem_cache *s, gfp_t flags) __assume_slab_alignment __malloc;
void kmem_cache_free(struct kmem_cache *s, void *objp);
431

432
/*
J
Jesper Dangaard Brouer 已提交
433
 * Bulk allocation and freeing operations. These are accelerated in an
434 435 436 437 438
 * allocator specific way to avoid taking locks repeatedly or building
 * metadata structures unnecessarily.
 *
 * Note that interrupts must be enabled when calling these functions.
 */
K
Kees Cook 已提交
439 440
void kmem_cache_free_bulk(struct kmem_cache *s, size_t size, void **p);
int kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t size, void **p);
441

442 443 444 445 446 447 448 449 450
/*
 * Caller must not use kfree_bulk() on memory not originally allocated
 * by kmalloc(), because the SLOB allocator cannot handle this.
 */
static __always_inline void kfree_bulk(size_t size, void **p)
{
	kmem_cache_free_bulk(NULL, size, p);
}

451
#ifdef CONFIG_NUMA
452 453
void *__kmalloc_node(size_t size, gfp_t flags, int node) __assume_kmalloc_alignment
							 __alloc_size(1);
K
Kees Cook 已提交
454 455
void *kmem_cache_alloc_node(struct kmem_cache *s, gfp_t flags, int node) __assume_slab_alignment
									 __malloc;
456
#else
457
static __always_inline __alloc_size(1) void *__kmalloc_node(size_t size, gfp_t flags, int node)
458 459 460 461 462 463 464 465 466 467 468
{
	return __kmalloc(size, flags);
}

static __always_inline void *kmem_cache_alloc_node(struct kmem_cache *s, gfp_t flags, int node)
{
	return kmem_cache_alloc(s, flags);
}
#endif

#ifdef CONFIG_TRACING
K
Kees Cook 已提交
469
extern void *kmem_cache_alloc_trace(struct kmem_cache *s, gfp_t flags, size_t size)
470
				   __assume_slab_alignment __alloc_size(3);
471 472

#ifdef CONFIG_NUMA
K
Kees Cook 已提交
473
extern void *kmem_cache_alloc_node_trace(struct kmem_cache *s, gfp_t gfpflags,
474 475
					 int node, size_t size) __assume_slab_alignment
								__alloc_size(4);
476
#else
477 478
static __always_inline __alloc_size(4) void *kmem_cache_alloc_node_trace(struct kmem_cache *s,
						 gfp_t gfpflags, int node, size_t size)
479 480 481 482 483 484
{
	return kmem_cache_alloc_trace(s, gfpflags, size);
}
#endif /* CONFIG_NUMA */

#else /* CONFIG_TRACING */
485 486
static __always_inline __alloc_size(3) void *kmem_cache_alloc_trace(struct kmem_cache *s,
								    gfp_t flags, size_t size)
487
{
488 489
	void *ret = kmem_cache_alloc(s, flags);

490
	ret = kasan_kmalloc(s, ret, size, flags);
491
	return ret;
492 493
}

K
Kees Cook 已提交
494 495
static __always_inline void *kmem_cache_alloc_node_trace(struct kmem_cache *s, gfp_t gfpflags,
							 int node, size_t size)
496
{
497 498
	void *ret = kmem_cache_alloc_node(s, gfpflags, node);

499
	ret = kasan_kmalloc(s, ret, size, gfpflags);
500
	return ret;
501 502 503
}
#endif /* CONFIG_TRACING */

K
Kees Cook 已提交
504
extern void *kmalloc_order(size_t size, gfp_t flags, unsigned int order) __assume_page_alignment
505
									 __alloc_size(1);
506 507

#ifdef CONFIG_TRACING
K
Kees Cook 已提交
508
extern void *kmalloc_order_trace(size_t size, gfp_t flags, unsigned int order)
509
				__assume_page_alignment __alloc_size(1);
510
#else
511 512
static __always_inline __alloc_size(1) void *kmalloc_order_trace(size_t size, gfp_t flags,
								 unsigned int order)
513 514 515
{
	return kmalloc_order(size, flags, order);
}
516 517
#endif

518
static __always_inline __alloc_size(1) void *kmalloc_large(size_t size, gfp_t flags)
519 520 521 522 523 524 525 526
{
	unsigned int order = get_order(size);
	return kmalloc_order_trace(size, flags, order);
}

/**
 * kmalloc - allocate memory
 * @size: how many bytes of memory are required.
527
 * @flags: the type of memory to allocate.
528 529 530
 *
 * kmalloc is the normal method of allocating memory
 * for objects smaller than page size in the kernel.
531
 *
532 533 534 535
 * The allocated object address is aligned to at least ARCH_KMALLOC_MINALIGN
 * bytes. For @size of power of two bytes, the alignment is also guaranteed
 * to be at least to the size.
 *
536 537 538
 * The @flags argument may be one of the GFP flags defined at
 * include/linux/gfp.h and described at
 * :ref:`Documentation/core-api/mm-api.rst <mm-api-gfp-flags>`
539
 *
540
 * The recommended usage of the @flags is described at
541
 * :ref:`Documentation/core-api/memory-allocation.rst <memory_allocation>`
542
 *
543
 * Below is a brief outline of the most useful GFP flags
544
 *
545 546
 * %GFP_KERNEL
 *	Allocate normal kernel ram. May sleep.
547
 *
548 549
 * %GFP_NOWAIT
 *	Allocation will not sleep.
550
 *
551 552
 * %GFP_ATOMIC
 *	Allocation will not sleep.  May use emergency pools.
553
 *
554 555
 * %GFP_HIGHUSER
 *	Allocate memory from high memory on behalf of user.
556 557 558 559
 *
 * Also it is possible to set different flags by OR'ing
 * in one or more of the following additional @flags:
 *
560 561
 * %__GFP_HIGH
 *	This allocation has high priority and may use emergency pools.
562
 *
563 564 565
 * %__GFP_NOFAIL
 *	Indicate that this allocation is in no way allowed to fail
 *	(think twice before using).
566
 *
567 568 569
 * %__GFP_NORETRY
 *	If memory is not immediately available,
 *	then give up at once.
570
 *
571 572
 * %__GFP_NOWARN
 *	If allocation fails, don't issue any warnings.
573
 *
574 575 576
 * %__GFP_RETRY_MAYFAIL
 *	Try really hard to succeed the allocation but fail
 *	eventually.
577
 */
578
static __always_inline __alloc_size(1) void *kmalloc(size_t size, gfp_t flags)
579 580
{
	if (__builtin_constant_p(size)) {
581 582 583
#ifndef CONFIG_SLOB
		unsigned int index;
#endif
584 585 586
		if (size > KMALLOC_MAX_CACHE_SIZE)
			return kmalloc_large(size, flags);
#ifndef CONFIG_SLOB
587
		index = kmalloc_index(size);
588

589 590
		if (!index)
			return ZERO_SIZE_PTR;
591

592 593 594
		return kmem_cache_alloc_trace(
				kmalloc_caches[kmalloc_type(flags)][index],
				flags, size);
595 596 597 598 599
#endif
	}
	return __kmalloc(size, flags);
}

600
static __always_inline __alloc_size(1) void *kmalloc_node(size_t size, gfp_t flags, int node)
601 602 603
{
#ifndef CONFIG_SLOB
	if (__builtin_constant_p(size) &&
604
		size <= KMALLOC_MAX_CACHE_SIZE) {
605
		unsigned int i = kmalloc_index(size);
606 607 608 609

		if (!i)
			return ZERO_SIZE_PTR;

610 611
		return kmem_cache_alloc_node_trace(
				kmalloc_caches[kmalloc_type(flags)][i],
612 613 614 615 616 617
						flags, node, size);
	}
#endif
	return __kmalloc_node(size, flags, node);
}

618 619 620 621 622
/**
 * kmalloc_array - allocate memory for an array.
 * @n: number of elements.
 * @size: element size.
 * @flags: the type of memory to allocate (see kmalloc).
623
 */
624
static inline __alloc_size(1, 2) void *kmalloc_array(size_t n, size_t size, gfp_t flags)
L
Linus Torvalds 已提交
625
{
626 627 628
	size_t bytes;

	if (unlikely(check_mul_overflow(n, size, &bytes)))
P
Paul Mundt 已提交
629
		return NULL;
630
	if (__builtin_constant_p(n) && __builtin_constant_p(size))
631 632
		return kmalloc(bytes, flags);
	return __kmalloc(bytes, flags);
X
Xi Wang 已提交
633 634
}

635 636 637 638 639 640 641
/**
 * krealloc_array - reallocate memory for an array.
 * @p: pointer to the memory chunk to reallocate
 * @new_n: new number of elements to alloc
 * @new_size: new size of a single member of the array
 * @flags: the type of memory to allocate (see kmalloc)
 */
642 643 644 645
static inline __alloc_size(2, 3) void * __must_check krealloc_array(void *p,
								    size_t new_n,
								    size_t new_size,
								    gfp_t flags)
646 647 648 649 650 651 652 653 654
{
	size_t bytes;

	if (unlikely(check_mul_overflow(new_n, new_size, &bytes)))
		return NULL;

	return krealloc(p, bytes, flags);
}

X
Xi Wang 已提交
655 656 657 658 659 660
/**
 * kcalloc - allocate memory for an array. The memory is set to zero.
 * @n: number of elements.
 * @size: element size.
 * @flags: the type of memory to allocate (see kmalloc).
 */
661
static inline __alloc_size(1, 2) void *kcalloc(size_t n, size_t size, gfp_t flags)
X
Xi Wang 已提交
662 663
{
	return kmalloc_array(n, size, flags | __GFP_ZERO);
L
Linus Torvalds 已提交
664 665
}

666 667 668 669 670 671 672 673
/*
 * kmalloc_track_caller is a special version of kmalloc that records the
 * calling function of the routine calling it for slab leak tracking instead
 * of just the calling function (confusing, eh?).
 * It's useful when the call to kmalloc comes from a widely-used standard
 * allocator where we care about the real place the memory allocation
 * request comes from.
 */
674 675
extern void *__kmalloc_track_caller(size_t size, gfp_t flags, unsigned long caller)
				   __alloc_size(1);
676
#define kmalloc_track_caller(size, flags) \
677
	__kmalloc_track_caller(size, flags, _RET_IP_)
L
Linus Torvalds 已提交
678

679 680
static inline __alloc_size(1, 2) void *kmalloc_array_node(size_t n, size_t size, gfp_t flags,
							  int node)
681
{
682 683 684
	size_t bytes;

	if (unlikely(check_mul_overflow(n, size, &bytes)))
685 686
		return NULL;
	if (__builtin_constant_p(n) && __builtin_constant_p(size))
687 688
		return kmalloc_node(bytes, flags, node);
	return __kmalloc_node(bytes, flags, node);
689 690
}

691
static inline __alloc_size(1, 2) void *kcalloc_node(size_t n, size_t size, gfp_t flags, int node)
692 693 694 695 696
{
	return kmalloc_array_node(n, size, flags | __GFP_ZERO, node);
}


697
#ifdef CONFIG_NUMA
K
Kees Cook 已提交
698
extern void *__kmalloc_node_track_caller(size_t size, gfp_t flags, int node,
699
					 unsigned long caller) __alloc_size(1);
700 701
#define kmalloc_node_track_caller(size, flags, node) \
	__kmalloc_node_track_caller(size, flags, node, \
702
			_RET_IP_)
703

704 705 706 707
#else /* CONFIG_NUMA */

#define kmalloc_node_track_caller(size, flags, node) \
	kmalloc_track_caller(size, flags)
708

P
Pascal Terjan 已提交
709
#endif /* CONFIG_NUMA */
710

711 712 713 714 715 716 717 718 719 720 721 722 723
/*
 * Shortcuts
 */
static inline void *kmem_cache_zalloc(struct kmem_cache *k, gfp_t flags)
{
	return kmem_cache_alloc(k, flags | __GFP_ZERO);
}

/**
 * kzalloc - allocate memory. The memory is set to zero.
 * @size: how many bytes of memory are required.
 * @flags: the type of memory to allocate (see kmalloc).
 */
724
static inline __alloc_size(1) void *kzalloc(size_t size, gfp_t flags)
725 726 727 728
{
	return kmalloc(size, flags | __GFP_ZERO);
}

J
Jeff Layton 已提交
729 730 731 732 733 734
/**
 * kzalloc_node - allocate zeroed memory from a particular memory node.
 * @size: how many bytes of memory are required.
 * @flags: the type of memory to allocate (see kmalloc).
 * @node: memory node from which to allocate
 */
735
static inline __alloc_size(1) void *kzalloc_node(size_t size, gfp_t flags, int node)
J
Jeff Layton 已提交
736 737 738 739
{
	return kmalloc_node(size, flags | __GFP_ZERO, node);
}

740 741
extern void *kvmalloc_node(size_t size, gfp_t flags, int node) __alloc_size(1);
static inline __alloc_size(1) void *kvmalloc(size_t size, gfp_t flags)
742 743 744
{
	return kvmalloc_node(size, flags, NUMA_NO_NODE);
}
745
static inline __alloc_size(1) void *kvzalloc_node(size_t size, gfp_t flags, int node)
746 747 748
{
	return kvmalloc_node(size, flags | __GFP_ZERO, node);
}
749
static inline __alloc_size(1) void *kvzalloc(size_t size, gfp_t flags)
750 751 752 753
{
	return kvmalloc(size, flags | __GFP_ZERO);
}

754
static inline __alloc_size(1, 2) void *kvmalloc_array(size_t n, size_t size, gfp_t flags)
755 756 757 758 759 760 761 762 763
{
	size_t bytes;

	if (unlikely(check_mul_overflow(n, size, &bytes)))
		return NULL;

	return kvmalloc(bytes, flags);
}

764
static inline __alloc_size(1, 2) void *kvcalloc(size_t n, size_t size, gfp_t flags)
765 766 767 768
{
	return kvmalloc_array(n, size, flags | __GFP_ZERO);
}

769 770
extern void *kvrealloc(const void *p, size_t oldsize, size_t newsize, gfp_t flags)
		      __alloc_size(3);
771 772 773
extern void kvfree(const void *addr);
extern void kvfree_sensitive(const void *addr, size_t len);

774
unsigned int kmem_cache_size(struct kmem_cache *s);
775 776
void __init kmem_cache_init_late(void);

777 778 779 780 781 782 783 784
#if defined(CONFIG_SMP) && defined(CONFIG_SLAB)
int slab_prepare_cpu(unsigned int cpu);
int slab_dead_cpu(unsigned int cpu);
#else
#define slab_prepare_cpu	NULL
#define slab_dead_cpu		NULL
#endif

L
Linus Torvalds 已提交
785
#endif	/* _LINUX_SLAB_H */