hashtab.c 35.2 KB
Newer Older
1
/* Copyright (c) 2011-2014 PLUMgrid, http://plumgrid.com
2
 * Copyright (c) 2016 Facebook
3 4 5 6 7 8 9 10 11 12 13 14 15
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of version 2 of the GNU General Public
 * License as published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful, but
 * WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
 * General Public License for more details.
 */
#include <linux/bpf.h>
#include <linux/jhash.h>
#include <linux/filter.h>
16
#include <linux/rculist_nulls.h>
17
#include "percpu_freelist.h"
18
#include "bpf_lru_list.h"
19
#include "map_in_map.h"
20

21 22 23
#define HTAB_CREATE_FLAG_MASK						\
	(BPF_F_NO_PREALLOC | BPF_F_NO_COMMON_LRU | BPF_F_NUMA_NODE |	\
	 BPF_F_RDONLY | BPF_F_WRONLY)
24

25
struct bucket {
26
	struct hlist_nulls_head head;
27 28 29
	raw_spinlock_t lock;
};

30 31
struct bpf_htab {
	struct bpf_map map;
32
	struct bucket *buckets;
33
	void *elems;
34 35 36 37
	union {
		struct pcpu_freelist freelist;
		struct bpf_lru lru;
	};
38
	struct htab_elem *__percpu *extra_elems;
39
	atomic_t count;	/* number of elements in this hashtable */
40 41 42 43 44 45
	u32 n_buckets;	/* number of hash buckets */
	u32 elem_size;	/* size of each element in bytes */
};

/* each htab element is struct htab_elem + key + value */
struct htab_elem {
46
	union {
47
		struct hlist_nulls_node hash_node;
48 49 50 51 52 53 54
		struct {
			void *padding;
			union {
				struct bpf_htab *htab;
				struct pcpu_freelist_node fnode;
			};
		};
55
	};
56 57
	union {
		struct rcu_head rcu;
58
		struct bpf_lru_node lru_node;
59
	};
60
	u32 hash;
61 62 63
	char key[0] __aligned(8);
};

64 65 66 67
static bool htab_lru_map_delete_node(void *arg, struct bpf_lru_node *node);

static bool htab_is_lru(const struct bpf_htab *htab)
{
68 69 70 71 72 73 74 75
	return htab->map.map_type == BPF_MAP_TYPE_LRU_HASH ||
		htab->map.map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH;
}

static bool htab_is_percpu(const struct bpf_htab *htab)
{
	return htab->map.map_type == BPF_MAP_TYPE_PERCPU_HASH ||
		htab->map.map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH;
76 77
}

78 79 80 81 82
static bool htab_is_prealloc(const struct bpf_htab *htab)
{
	return !(htab->map.map_flags & BPF_F_NO_PREALLOC);
}

83 84 85 86 87 88 89 90 91 92 93
static inline void htab_elem_set_ptr(struct htab_elem *l, u32 key_size,
				     void __percpu *pptr)
{
	*(void __percpu **)(l->key + key_size) = pptr;
}

static inline void __percpu *htab_elem_get_ptr(struct htab_elem *l, u32 key_size)
{
	return *(void __percpu **)(l->key + key_size);
}

94 95 96 97 98
static void *fd_htab_map_get_ptr(const struct bpf_map *map, struct htab_elem *l)
{
	return *(void **)(l->key + roundup(map->key_size, 8));
}

99 100 101 102 103 104 105 106 107
static struct htab_elem *get_htab_elem(struct bpf_htab *htab, int i)
{
	return (struct htab_elem *) (htab->elems + i * htab->elem_size);
}

static void htab_free_elems(struct bpf_htab *htab)
{
	int i;

108
	if (!htab_is_percpu(htab))
109 110 111 112 113 114 115 116
		goto free_elems;

	for (i = 0; i < htab->map.max_entries; i++) {
		void __percpu *pptr;

		pptr = htab_elem_get_ptr(get_htab_elem(htab, i),
					 htab->map.key_size);
		free_percpu(pptr);
117
		cond_resched();
118 119
	}
free_elems:
120
	bpf_map_area_free(htab->elems);
121 122
}

123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138
static struct htab_elem *prealloc_lru_pop(struct bpf_htab *htab, void *key,
					  u32 hash)
{
	struct bpf_lru_node *node = bpf_lru_pop_free(&htab->lru, hash);
	struct htab_elem *l;

	if (node) {
		l = container_of(node, struct htab_elem, lru_node);
		memcpy(l->key, key, htab->map.key_size);
		return l;
	}

	return NULL;
}

static int prealloc_init(struct bpf_htab *htab)
139
{
140
	u32 num_entries = htab->map.max_entries;
141 142
	int err = -ENOMEM, i;

143 144 145
	if (!htab_is_percpu(htab) && !htab_is_lru(htab))
		num_entries += num_possible_cpus();

146 147
	htab->elems = bpf_map_area_alloc(htab->elem_size * num_entries,
					 htab->map.numa_node);
148 149 150
	if (!htab->elems)
		return -ENOMEM;

151
	if (!htab_is_percpu(htab))
152 153
		goto skip_percpu_elems;

154
	for (i = 0; i < num_entries; i++) {
155 156 157 158 159 160 161 162
		u32 size = round_up(htab->map.value_size, 8);
		void __percpu *pptr;

		pptr = __alloc_percpu_gfp(size, 8, GFP_USER | __GFP_NOWARN);
		if (!pptr)
			goto free_elems;
		htab_elem_set_ptr(get_htab_elem(htab, i), htab->map.key_size,
				  pptr);
163
		cond_resched();
164 165 166
	}

skip_percpu_elems:
167 168 169 170 171 172 173 174 175 176
	if (htab_is_lru(htab))
		err = bpf_lru_init(&htab->lru,
				   htab->map.map_flags & BPF_F_NO_COMMON_LRU,
				   offsetof(struct htab_elem, hash) -
				   offsetof(struct htab_elem, lru_node),
				   htab_lru_map_delete_node,
				   htab);
	else
		err = pcpu_freelist_init(&htab->freelist);

177 178 179
	if (err)
		goto free_elems;

180 181 182
	if (htab_is_lru(htab))
		bpf_lru_populate(&htab->lru, htab->elems,
				 offsetof(struct htab_elem, lru_node),
183
				 htab->elem_size, num_entries);
184
	else
185 186
		pcpu_freelist_populate(&htab->freelist,
				       htab->elems + offsetof(struct htab_elem, fnode),
187
				       htab->elem_size, num_entries);
188

189 190 191 192 193 194 195
	return 0;

free_elems:
	htab_free_elems(htab);
	return err;
}

196 197 198 199 200 201 202 203 204 205
static void prealloc_destroy(struct bpf_htab *htab)
{
	htab_free_elems(htab);

	if (htab_is_lru(htab))
		bpf_lru_destroy(&htab->lru);
	else
		pcpu_freelist_destroy(&htab->freelist);
}

206 207
static int alloc_extra_elems(struct bpf_htab *htab)
{
208 209
	struct htab_elem *__percpu *pptr, *l_new;
	struct pcpu_freelist_node *l;
210 211
	int cpu;

212 213
	pptr = __alloc_percpu_gfp(sizeof(struct htab_elem *), 8,
				  GFP_USER | __GFP_NOWARN);
214 215 216 217
	if (!pptr)
		return -ENOMEM;

	for_each_possible_cpu(cpu) {
218 219 220 221 222 223
		l = pcpu_freelist_pop(&htab->freelist);
		/* pop will succeed, since prealloc_init()
		 * preallocated extra num_possible_cpus elements
		 */
		l_new = container_of(l, struct htab_elem, fnode);
		*per_cpu_ptr(pptr, cpu) = l_new;
224 225 226 227 228
	}
	htab->extra_elems = pptr;
	return 0;
}

229
/* Called from syscall */
230
static int htab_map_alloc_check(union bpf_attr *attr)
231
{
232 233 234 235
	bool percpu = (attr->map_type == BPF_MAP_TYPE_PERCPU_HASH ||
		       attr->map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH);
	bool lru = (attr->map_type == BPF_MAP_TYPE_LRU_HASH ||
		    attr->map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH);
236 237 238 239 240 241 242
	/* percpu_lru means each cpu has its own LRU list.
	 * it is different from BPF_MAP_TYPE_PERCPU_HASH where
	 * the map's value itself is percpu.  percpu_lru has
	 * nothing to do with the map's value.
	 */
	bool percpu_lru = (attr->map_flags & BPF_F_NO_COMMON_LRU);
	bool prealloc = !(attr->map_flags & BPF_F_NO_PREALLOC);
243
	int numa_node = bpf_map_attr_numa_node(attr);
244

245 246 247 248 249
	BUILD_BUG_ON(offsetof(struct htab_elem, htab) !=
		     offsetof(struct htab_elem, hash_node.pprev));
	BUILD_BUG_ON(offsetof(struct htab_elem, fnode.next) !=
		     offsetof(struct htab_elem, hash_node.pprev));

250 251 252 253
	if (lru && !capable(CAP_SYS_ADMIN))
		/* LRU implementation is much complicated than other
		 * maps.  Hence, limit to CAP_SYS_ADMIN for now.
		 */
254
		return -EPERM;
255

256
	if (attr->map_flags & ~HTAB_CREATE_FLAG_MASK)
257
		/* reserved bits should not be used */
258
		return -EINVAL;
259

260
	if (!lru && percpu_lru)
261
		return -EINVAL;
262 263

	if (lru && !prealloc)
264
		return -ENOTSUPP;
265

266
	if (numa_node != NUMA_NO_NODE && (percpu || percpu_lru))
267
		return -EINVAL;
268

269 270 271 272 273
	/* check sanity of attributes.
	 * value_size == 0 may be allowed in the future to use map as a set
	 */
	if (attr->max_entries == 0 || attr->key_size == 0 ||
	    attr->value_size == 0)
274
		return -EINVAL;
275 276 277 278 279

	if (attr->key_size > MAX_BPF_STACK)
		/* eBPF programs initialize keys on stack, so they cannot be
		 * larger than max stack size
		 */
280
		return -E2BIG;
281 282 283 284 285 286 287 288

	if (attr->value_size >= KMALLOC_MAX_SIZE -
	    MAX_BPF_STACK - sizeof(struct htab_elem))
		/* if value_size is bigger, the user space won't be able to
		 * access the elements via bpf syscall. This check also makes
		 * sure that the elem_size doesn't overflow and it's
		 * kmalloc-able later in htab_map_update_elem()
		 */
289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309
		return -E2BIG;

	return 0;
}

static struct bpf_map *htab_map_alloc(union bpf_attr *attr)
{
	bool percpu = (attr->map_type == BPF_MAP_TYPE_PERCPU_HASH ||
		       attr->map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH);
	bool lru = (attr->map_type == BPF_MAP_TYPE_LRU_HASH ||
		    attr->map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH);
	/* percpu_lru means each cpu has its own LRU list.
	 * it is different from BPF_MAP_TYPE_PERCPU_HASH where
	 * the map's value itself is percpu.  percpu_lru has
	 * nothing to do with the map's value.
	 */
	bool percpu_lru = (attr->map_flags & BPF_F_NO_COMMON_LRU);
	bool prealloc = !(attr->map_flags & BPF_F_NO_PREALLOC);
	struct bpf_htab *htab;
	int err, i;
	u64 cost;
310

311 312 313 314
	htab = kzalloc(sizeof(*htab), GFP_USER);
	if (!htab)
		return ERR_PTR(-ENOMEM);

315
	bpf_map_init_from_attr(&htab->map, attr);
316

317 318 319 320 321 322 323 324 325 326 327 328
	if (percpu_lru) {
		/* ensure each CPU's lru list has >=1 elements.
		 * since we are at it, make each lru list has the same
		 * number of elements.
		 */
		htab->map.max_entries = roundup(attr->max_entries,
						num_possible_cpus());
		if (htab->map.max_entries < attr->max_entries)
			htab->map.max_entries = rounddown(attr->max_entries,
							  num_possible_cpus());
	}

329 330 331
	/* hash table size must be power of 2 */
	htab->n_buckets = roundup_pow_of_two(htab->map.max_entries);

332
	htab->elem_size = sizeof(struct htab_elem) +
333 334 335 336
			  round_up(htab->map.key_size, 8);
	if (percpu)
		htab->elem_size += sizeof(void *);
	else
337
		htab->elem_size += round_up(htab->map.value_size, 8);
338

339
	err = -E2BIG;
340 341
	/* prevent zero size kmalloc and check for u32 overflow */
	if (htab->n_buckets == 0 ||
342
	    htab->n_buckets > U32_MAX / sizeof(struct bucket))
343 344
		goto free_htab;

345 346 347 348 349 350
	cost = (u64) htab->n_buckets * sizeof(struct bucket) +
	       (u64) htab->elem_size * htab->map.max_entries;

	if (percpu)
		cost += (u64) round_up(htab->map.value_size, 8) *
			num_possible_cpus() * htab->map.max_entries;
351 352
	else
	       cost += (u64) htab->elem_size * num_possible_cpus();
353 354

	if (cost >= U32_MAX - PAGE_SIZE)
355 356 357
		/* make sure page count doesn't overflow */
		goto free_htab;

358
	htab->map.pages = round_up(cost, PAGE_SIZE) >> PAGE_SHIFT;
359

360 361 362 363 364
	/* if map size is larger than memlock limit, reject it early */
	err = bpf_map_precharge_memlock(htab->map.pages);
	if (err)
		goto free_htab;

365
	err = -ENOMEM;
366
	htab->buckets = bpf_map_area_alloc(htab->n_buckets *
367 368
					   sizeof(struct bucket),
					   htab->map.numa_node);
369 370
	if (!htab->buckets)
		goto free_htab;
371

372
	for (i = 0; i < htab->n_buckets; i++) {
373
		INIT_HLIST_NULLS_HEAD(&htab->buckets[i].head, i);
374 375
		raw_spin_lock_init(&htab->buckets[i].lock);
	}
376

377 378
	if (prealloc) {
		err = prealloc_init(htab);
379
		if (err)
380 381 382 383 384 385 386 387 388 389
			goto free_buckets;

		if (!percpu && !lru) {
			/* lru itself can remove the least used element, so
			 * there is no need for an extra elem during map_update.
			 */
			err = alloc_extra_elems(htab);
			if (err)
				goto free_prealloc;
		}
390
	}
391 392 393

	return &htab->map;

394 395
free_prealloc:
	prealloc_destroy(htab);
396
free_buckets:
397
	bpf_map_area_free(htab->buckets);
398 399 400 401 402 403 404 405 406 407
free_htab:
	kfree(htab);
	return ERR_PTR(err);
}

static inline u32 htab_map_hash(const void *key, u32 key_len)
{
	return jhash(key, key_len, 0);
}

408
static inline struct bucket *__select_bucket(struct bpf_htab *htab, u32 hash)
409 410 411 412
{
	return &htab->buckets[hash & (htab->n_buckets - 1)];
}

413
static inline struct hlist_nulls_head *select_bucket(struct bpf_htab *htab, u32 hash)
414 415 416 417
{
	return &__select_bucket(htab, hash)->head;
}

418 419
/* this lookup function can only be called with bucket lock taken */
static struct htab_elem *lookup_elem_raw(struct hlist_nulls_head *head, u32 hash,
420 421
					 void *key, u32 key_size)
{
422
	struct hlist_nulls_node *n;
423 424
	struct htab_elem *l;

425
	hlist_nulls_for_each_entry_rcu(l, n, head, hash_node)
426 427 428 429 430 431
		if (l->hash == hash && !memcmp(&l->key, key, key_size))
			return l;

	return NULL;
}

432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453
/* can be called without bucket lock. it will repeat the loop in
 * the unlikely event when elements moved from one bucket into another
 * while link list is being walked
 */
static struct htab_elem *lookup_nulls_elem_raw(struct hlist_nulls_head *head,
					       u32 hash, void *key,
					       u32 key_size, u32 n_buckets)
{
	struct hlist_nulls_node *n;
	struct htab_elem *l;

again:
	hlist_nulls_for_each_entry_rcu(l, n, head, hash_node)
		if (l->hash == hash && !memcmp(&l->key, key, key_size))
			return l;

	if (unlikely(get_nulls_value(n) != (hash & (n_buckets - 1))))
		goto again;

	return NULL;
}

454 455 456 457 458
/* Called from syscall or from eBPF program directly, so
 * arguments have to match bpf_map_lookup_elem() exactly.
 * The return value is adjusted by BPF instructions
 * in htab_map_gen_lookup().
 */
459
static void *__htab_map_lookup_elem(struct bpf_map *map, void *key)
460 461
{
	struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
462
	struct hlist_nulls_head *head;
463 464 465 466 467 468 469 470 471 472 473 474
	struct htab_elem *l;
	u32 hash, key_size;

	/* Must be called with rcu_read_lock. */
	WARN_ON_ONCE(!rcu_read_lock_held());

	key_size = map->key_size;

	hash = htab_map_hash(key, key_size);

	head = select_bucket(htab, hash);

475
	l = lookup_nulls_elem_raw(head, hash, key, key_size, htab->n_buckets);
476

477 478 479 480 481 482 483
	return l;
}

static void *htab_map_lookup_elem(struct bpf_map *map, void *key)
{
	struct htab_elem *l = __htab_map_lookup_elem(map, key);

484 485 486 487 488 489
	if (l)
		return l->key + round_up(map->key_size, 8);

	return NULL;
}

490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505
/* inline bpf_map_lookup_elem() call.
 * Instead of:
 * bpf_prog
 *   bpf_map_lookup_elem
 *     map->ops->map_lookup_elem
 *       htab_map_lookup_elem
 *         __htab_map_lookup_elem
 * do:
 * bpf_prog
 *   __htab_map_lookup_elem
 */
static u32 htab_map_gen_lookup(struct bpf_map *map, struct bpf_insn *insn_buf)
{
	struct bpf_insn *insn = insn_buf;
	const int ret = BPF_REG_0;

506 507 508
	BUILD_BUG_ON(!__same_type(&__htab_map_lookup_elem,
		     (void *(*)(struct bpf_map *map, void *key))NULL));
	*insn++ = BPF_EMIT_CALL(BPF_CAST_CALL(__htab_map_lookup_elem));
509 510 511 512 513 514 515
	*insn++ = BPF_JMP_IMM(BPF_JEQ, ret, 0, 1);
	*insn++ = BPF_ALU64_IMM(BPF_ADD, ret,
				offsetof(struct htab_elem, key) +
				round_up(map->key_size, 8));
	return insn - insn_buf;
}

516 517 518 519 520 521 522 523 524 525 526 527
static void *htab_lru_map_lookup_elem(struct bpf_map *map, void *key)
{
	struct htab_elem *l = __htab_map_lookup_elem(map, key);

	if (l) {
		bpf_lru_node_set_ref(&l->lru_node);
		return l->key + round_up(map->key_size, 8);
	}

	return NULL;
}

528 529 530 531 532
static u32 htab_lru_map_gen_lookup(struct bpf_map *map,
				   struct bpf_insn *insn_buf)
{
	struct bpf_insn *insn = insn_buf;
	const int ret = BPF_REG_0;
533
	const int ref_reg = BPF_REG_1;
534

535 536 537
	BUILD_BUG_ON(!__same_type(&__htab_map_lookup_elem,
		     (void *(*)(struct bpf_map *map, void *key))NULL));
	*insn++ = BPF_EMIT_CALL(BPF_CAST_CALL(__htab_map_lookup_elem));
538 539 540 541 542
	*insn++ = BPF_JMP_IMM(BPF_JEQ, ret, 0, 4);
	*insn++ = BPF_LDX_MEM(BPF_B, ref_reg, ret,
			      offsetof(struct htab_elem, lru_node) +
			      offsetof(struct bpf_lru_node, ref));
	*insn++ = BPF_JMP_IMM(BPF_JNE, ref_reg, 0, 1);
543 544 545 546 547 548 549 550 551 552
	*insn++ = BPF_ST_MEM(BPF_B, ret,
			     offsetof(struct htab_elem, lru_node) +
			     offsetof(struct bpf_lru_node, ref),
			     1);
	*insn++ = BPF_ALU64_IMM(BPF_ADD, ret,
				offsetof(struct htab_elem, key) +
				round_up(map->key_size, 8));
	return insn - insn_buf;
}

553 554 555 556 557 558
/* It is called from the bpf_lru_list when the LRU needs to delete
 * older elements from the htab.
 */
static bool htab_lru_map_delete_node(void *arg, struct bpf_lru_node *node)
{
	struct bpf_htab *htab = (struct bpf_htab *)arg;
559 560 561
	struct htab_elem *l = NULL, *tgt_l;
	struct hlist_nulls_head *head;
	struct hlist_nulls_node *n;
562 563 564 565 566 567 568 569 570
	unsigned long flags;
	struct bucket *b;

	tgt_l = container_of(node, struct htab_elem, lru_node);
	b = __select_bucket(htab, tgt_l->hash);
	head = &b->head;

	raw_spin_lock_irqsave(&b->lock, flags);

571
	hlist_nulls_for_each_entry_rcu(l, n, head, hash_node)
572
		if (l == tgt_l) {
573
			hlist_nulls_del_rcu(&l->hash_node);
574 575 576 577 578 579 580 581
			break;
		}

	raw_spin_unlock_irqrestore(&b->lock, flags);

	return l == tgt_l;
}

582 583 584 585
/* Called from syscall */
static int htab_map_get_next_key(struct bpf_map *map, void *key, void *next_key)
{
	struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
586
	struct hlist_nulls_head *head;
587 588
	struct htab_elem *l, *next_l;
	u32 hash, key_size;
589
	int i = 0;
590 591 592 593 594

	WARN_ON_ONCE(!rcu_read_lock_held());

	key_size = map->key_size;

595 596 597
	if (!key)
		goto find_first_elem;

598 599 600 601 602
	hash = htab_map_hash(key, key_size);

	head = select_bucket(htab, hash);

	/* lookup the key */
603
	l = lookup_nulls_elem_raw(head, hash, key, key_size, htab->n_buckets);
604

605
	if (!l)
606 607 608
		goto find_first_elem;

	/* key was found, get next key in the same bucket */
609
	next_l = hlist_nulls_entry_safe(rcu_dereference_raw(hlist_nulls_next_rcu(&l->hash_node)),
610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627
				  struct htab_elem, hash_node);

	if (next_l) {
		/* if next elem in this hash list is non-zero, just return it */
		memcpy(next_key, next_l->key, key_size);
		return 0;
	}

	/* no more elements in this hash list, go to the next bucket */
	i = hash & (htab->n_buckets - 1);
	i++;

find_first_elem:
	/* iterate over buckets */
	for (; i < htab->n_buckets; i++) {
		head = select_bucket(htab, i);

		/* pick first element in the bucket */
628
		next_l = hlist_nulls_entry_safe(rcu_dereference_raw(hlist_nulls_first_rcu(head)),
629 630 631 632 633 634 635 636
					  struct htab_elem, hash_node);
		if (next_l) {
			/* if it's not empty, just return it */
			memcpy(next_key, next_l->key, key_size);
			return 0;
		}
	}

637
	/* iterated over all buckets and all elements */
638 639 640
	return -ENOENT;
}

641
static void htab_elem_free(struct bpf_htab *htab, struct htab_elem *l)
642
{
643 644
	if (htab->map.map_type == BPF_MAP_TYPE_PERCPU_HASH)
		free_percpu(htab_elem_get_ptr(l, htab->map.key_size));
645 646 647
	kfree(l);
}

648
static void htab_elem_free_rcu(struct rcu_head *head)
649 650
{
	struct htab_elem *l = container_of(head, struct htab_elem, rcu);
651
	struct bpf_htab *htab = l->htab;
652

653 654 655 656 657 658 659 660 661
	/* must increment bpf_prog_active to avoid kprobe+bpf triggering while
	 * we're calling kfree, otherwise deadlock is possible if kprobes
	 * are placed somewhere inside of slub
	 */
	preempt_disable();
	__this_cpu_inc(bpf_prog_active);
	htab_elem_free(htab, l);
	__this_cpu_dec(bpf_prog_active);
	preempt_enable();
662 663
}

664
static void free_htab_elem(struct bpf_htab *htab, struct htab_elem *l)
665
{
666 667 668 669 670 671 672 673
	struct bpf_map *map = &htab->map;

	if (map->ops->map_fd_put_ptr) {
		void *ptr = fd_htab_map_get_ptr(map, l);

		map->ops->map_fd_put_ptr(ptr);
	}

674
	if (htab_is_prealloc(htab)) {
675
		pcpu_freelist_push(&htab->freelist, &l->fnode);
676
	} else {
677 678 679
		atomic_dec(&htab->count);
		l->htab = htab;
		call_rcu(&l->rcu, htab_elem_free_rcu);
680 681 682
	}
}

683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700
static void pcpu_copy_value(struct bpf_htab *htab, void __percpu *pptr,
			    void *value, bool onallcpus)
{
	if (!onallcpus) {
		/* copy true value_size bytes */
		memcpy(this_cpu_ptr(pptr), value, htab->map.value_size);
	} else {
		u32 size = round_up(htab->map.value_size, 8);
		int off = 0, cpu;

		for_each_possible_cpu(cpu) {
			bpf_long_memcpy(per_cpu_ptr(pptr, cpu),
					value + off, size);
			off += size;
		}
	}
}

701 702 703 704 705 706 707 708 709 710 711 712 713 714 715
static bool fd_htab_map_needs_adjust(const struct bpf_htab *htab)
{
	return htab->map.map_type == BPF_MAP_TYPE_HASH_OF_MAPS &&
	       BITS_PER_LONG == 64;
}

static u32 htab_size_value(const struct bpf_htab *htab, bool percpu)
{
	u32 size = htab->map.value_size;

	if (percpu || fd_htab_map_needs_adjust(htab))
		size = round_up(size, 8);
	return size;
}

716 717
static struct htab_elem *alloc_htab_elem(struct bpf_htab *htab, void *key,
					 void *value, u32 key_size, u32 hash,
718
					 bool percpu, bool onallcpus,
719
					 struct htab_elem *old_elem)
720
{
721
	u32 size = htab_size_value(htab, percpu);
722 723
	bool prealloc = htab_is_prealloc(htab);
	struct htab_elem *l_new, **pl_new;
724 725
	void __percpu *pptr;

726
	if (prealloc) {
727 728 729 730 731 732 733 734 735
		if (old_elem) {
			/* if we're updating the existing element,
			 * use per-cpu extra elems to avoid freelist_pop/push
			 */
			pl_new = this_cpu_ptr(htab->extra_elems);
			l_new = *pl_new;
			*pl_new = old_elem;
		} else {
			struct pcpu_freelist_node *l;
736

737 738 739
			l = pcpu_freelist_pop(&htab->freelist);
			if (!l)
				return ERR_PTR(-E2BIG);
740
			l_new = container_of(l, struct htab_elem, fnode);
741
		}
742
	} else {
743 744 745 746 747 748 749
		if (atomic_inc_return(&htab->count) > htab->map.max_entries)
			if (!old_elem) {
				/* when map is full and update() is replacing
				 * old element, it's ok to allocate, since
				 * old element will be freed immediately.
				 * Otherwise return an error
				 */
750 751
				l_new = ERR_PTR(-E2BIG);
				goto dec_count;
752
			}
753 754
		l_new = kmalloc_node(htab->elem_size, GFP_ATOMIC | __GFP_NOWARN,
				     htab->map.numa_node);
755 756 757 758
		if (!l_new) {
			l_new = ERR_PTR(-ENOMEM);
			goto dec_count;
		}
759
	}
760 761 762

	memcpy(l_new->key, key, key_size);
	if (percpu) {
763 764 765 766 767 768 769 770
		if (prealloc) {
			pptr = htab_elem_get_ptr(l_new, key_size);
		} else {
			/* alloc_percpu zero-fills */
			pptr = __alloc_percpu_gfp(size, 8,
						  GFP_ATOMIC | __GFP_NOWARN);
			if (!pptr) {
				kfree(l_new);
771 772
				l_new = ERR_PTR(-ENOMEM);
				goto dec_count;
773
			}
774 775
		}

776
		pcpu_copy_value(htab, pptr, value, onallcpus);
777

778 779
		if (!prealloc)
			htab_elem_set_ptr(l_new, key_size, pptr);
780 781 782 783 784 785
	} else {
		memcpy(l_new->key + round_up(key_size, 8), value, size);
	}

	l_new->hash = hash;
	return l_new;
786 787 788
dec_count:
	atomic_dec(&htab->count);
	return l_new;
789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804
}

static int check_flags(struct bpf_htab *htab, struct htab_elem *l_old,
		       u64 map_flags)
{
	if (l_old && map_flags == BPF_NOEXIST)
		/* elem already exists */
		return -EEXIST;

	if (!l_old && map_flags == BPF_EXIST)
		/* elem doesn't exist, cannot update it */
		return -ENOENT;

	return 0;
}

805 806 807 808 809
/* Called from syscall or from eBPF program */
static int htab_map_update_elem(struct bpf_map *map, void *key, void *value,
				u64 map_flags)
{
	struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
810
	struct htab_elem *l_new = NULL, *l_old;
811
	struct hlist_nulls_head *head;
812
	unsigned long flags;
813 814
	struct bucket *b;
	u32 key_size, hash;
815 816
	int ret;

817
	if (unlikely(map_flags > BPF_EXIST))
818 819 820 821 822 823 824
		/* unknown flags */
		return -EINVAL;

	WARN_ON_ONCE(!rcu_read_lock_held());

	key_size = map->key_size;

825 826 827
	hash = htab_map_hash(key, key_size);

	b = __select_bucket(htab, hash);
828
	head = &b->head;
829 830

	/* bpf_map_update_elem() can be called in_irq() */
831
	raw_spin_lock_irqsave(&b->lock, flags);
832

833
	l_old = lookup_elem_raw(head, hash, key, key_size);
834

835 836
	ret = check_flags(htab, l_old, map_flags);
	if (ret)
837 838
		goto err;

839
	l_new = alloc_htab_elem(htab, key, value, key_size, hash, false, false,
840
				l_old);
841 842 843 844 845 846
	if (IS_ERR(l_new)) {
		/* all pre-allocated elements are in use or memory exhausted */
		ret = PTR_ERR(l_new);
		goto err;
	}

847 848
	/* add new element to the head of the list, so that
	 * concurrent search will find it before old elem
849
	 */
850
	hlist_nulls_add_head_rcu(&l_new->hash_node, head);
851
	if (l_old) {
852
		hlist_nulls_del_rcu(&l_old->hash_node);
853 854
		if (!htab_is_prealloc(htab))
			free_htab_elem(htab, l_old);
855
	}
856
	ret = 0;
857
err:
858
	raw_spin_unlock_irqrestore(&b->lock, flags);
859 860 861
	return ret;
}

862 863 864 865 866
static int htab_lru_map_update_elem(struct bpf_map *map, void *key, void *value,
				    u64 map_flags)
{
	struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
	struct htab_elem *l_new, *l_old = NULL;
867
	struct hlist_nulls_head *head;
868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907
	unsigned long flags;
	struct bucket *b;
	u32 key_size, hash;
	int ret;

	if (unlikely(map_flags > BPF_EXIST))
		/* unknown flags */
		return -EINVAL;

	WARN_ON_ONCE(!rcu_read_lock_held());

	key_size = map->key_size;

	hash = htab_map_hash(key, key_size);

	b = __select_bucket(htab, hash);
	head = &b->head;

	/* For LRU, we need to alloc before taking bucket's
	 * spinlock because getting free nodes from LRU may need
	 * to remove older elements from htab and this removal
	 * operation will need a bucket lock.
	 */
	l_new = prealloc_lru_pop(htab, key, hash);
	if (!l_new)
		return -ENOMEM;
	memcpy(l_new->key + round_up(map->key_size, 8), value, map->value_size);

	/* bpf_map_update_elem() can be called in_irq() */
	raw_spin_lock_irqsave(&b->lock, flags);

	l_old = lookup_elem_raw(head, hash, key, key_size);

	ret = check_flags(htab, l_old, map_flags);
	if (ret)
		goto err;

	/* add new element to the head of the list, so that
	 * concurrent search will find it before old elem
	 */
908
	hlist_nulls_add_head_rcu(&l_new->hash_node, head);
909 910
	if (l_old) {
		bpf_lru_node_set_ref(&l_new->lru_node);
911
		hlist_nulls_del_rcu(&l_old->hash_node);
912 913 914 915 916 917 918 919 920 921 922 923 924 925
	}
	ret = 0;

err:
	raw_spin_unlock_irqrestore(&b->lock, flags);

	if (ret)
		bpf_lru_push_free(&htab->lru, &l_new->lru_node);
	else if (l_old)
		bpf_lru_push_free(&htab->lru, &l_old->lru_node);

	return ret;
}

926 927 928
static int __htab_percpu_map_update_elem(struct bpf_map *map, void *key,
					 void *value, u64 map_flags,
					 bool onallcpus)
929 930 931
{
	struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
	struct htab_elem *l_new = NULL, *l_old;
932
	struct hlist_nulls_head *head;
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
	unsigned long flags;
	struct bucket *b;
	u32 key_size, hash;
	int ret;

	if (unlikely(map_flags > BPF_EXIST))
		/* unknown flags */
		return -EINVAL;

	WARN_ON_ONCE(!rcu_read_lock_held());

	key_size = map->key_size;

	hash = htab_map_hash(key, key_size);

	b = __select_bucket(htab, hash);
	head = &b->head;

	/* bpf_map_update_elem() can be called in_irq() */
	raw_spin_lock_irqsave(&b->lock, flags);

	l_old = lookup_elem_raw(head, hash, key, key_size);

	ret = check_flags(htab, l_old, map_flags);
	if (ret)
		goto err;

	if (l_old) {
		/* per-cpu hash map can update value in-place */
962 963
		pcpu_copy_value(htab, htab_elem_get_ptr(l_old, key_size),
				value, onallcpus);
964 965
	} else {
		l_new = alloc_htab_elem(htab, key, value, key_size,
966
					hash, true, onallcpus, NULL);
967 968
		if (IS_ERR(l_new)) {
			ret = PTR_ERR(l_new);
969 970
			goto err;
		}
971
		hlist_nulls_add_head_rcu(&l_new->hash_node, head);
972 973 974 975 976 977 978
	}
	ret = 0;
err:
	raw_spin_unlock_irqrestore(&b->lock, flags);
	return ret;
}

979 980 981 982 983 984
static int __htab_lru_percpu_map_update_elem(struct bpf_map *map, void *key,
					     void *value, u64 map_flags,
					     bool onallcpus)
{
	struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
	struct htab_elem *l_new = NULL, *l_old;
985
	struct hlist_nulls_head *head;
986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032
	unsigned long flags;
	struct bucket *b;
	u32 key_size, hash;
	int ret;

	if (unlikely(map_flags > BPF_EXIST))
		/* unknown flags */
		return -EINVAL;

	WARN_ON_ONCE(!rcu_read_lock_held());

	key_size = map->key_size;

	hash = htab_map_hash(key, key_size);

	b = __select_bucket(htab, hash);
	head = &b->head;

	/* For LRU, we need to alloc before taking bucket's
	 * spinlock because LRU's elem alloc may need
	 * to remove older elem from htab and this removal
	 * operation will need a bucket lock.
	 */
	if (map_flags != BPF_EXIST) {
		l_new = prealloc_lru_pop(htab, key, hash);
		if (!l_new)
			return -ENOMEM;
	}

	/* bpf_map_update_elem() can be called in_irq() */
	raw_spin_lock_irqsave(&b->lock, flags);

	l_old = lookup_elem_raw(head, hash, key, key_size);

	ret = check_flags(htab, l_old, map_flags);
	if (ret)
		goto err;

	if (l_old) {
		bpf_lru_node_set_ref(&l_old->lru_node);

		/* per-cpu hash map can update value in-place */
		pcpu_copy_value(htab, htab_elem_get_ptr(l_old, key_size),
				value, onallcpus);
	} else {
		pcpu_copy_value(htab, htab_elem_get_ptr(l_new, key_size),
				value, onallcpus);
1033
		hlist_nulls_add_head_rcu(&l_new->hash_node, head);
1034 1035 1036 1037 1038 1039 1040 1041 1042 1043
		l_new = NULL;
	}
	ret = 0;
err:
	raw_spin_unlock_irqrestore(&b->lock, flags);
	if (l_new)
		bpf_lru_push_free(&htab->lru, &l_new->lru_node);
	return ret;
}

1044 1045 1046 1047 1048 1049
static int htab_percpu_map_update_elem(struct bpf_map *map, void *key,
				       void *value, u64 map_flags)
{
	return __htab_percpu_map_update_elem(map, key, value, map_flags, false);
}

1050 1051 1052 1053 1054 1055 1056
static int htab_lru_percpu_map_update_elem(struct bpf_map *map, void *key,
					   void *value, u64 map_flags)
{
	return __htab_lru_percpu_map_update_elem(map, key, value, map_flags,
						 false);
}

1057 1058 1059 1060
/* Called from syscall or from eBPF program */
static int htab_map_delete_elem(struct bpf_map *map, void *key)
{
	struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
1061
	struct hlist_nulls_head *head;
1062
	struct bucket *b;
1063 1064 1065 1066 1067 1068 1069 1070 1071 1072
	struct htab_elem *l;
	unsigned long flags;
	u32 hash, key_size;
	int ret = -ENOENT;

	WARN_ON_ONCE(!rcu_read_lock_held());

	key_size = map->key_size;

	hash = htab_map_hash(key, key_size);
1073 1074
	b = __select_bucket(htab, hash);
	head = &b->head;
1075

1076
	raw_spin_lock_irqsave(&b->lock, flags);
1077 1078 1079 1080

	l = lookup_elem_raw(head, hash, key, key_size);

	if (l) {
1081
		hlist_nulls_del_rcu(&l->hash_node);
1082
		free_htab_elem(htab, l);
1083 1084 1085
		ret = 0;
	}

1086
	raw_spin_unlock_irqrestore(&b->lock, flags);
1087 1088 1089
	return ret;
}

1090 1091 1092
static int htab_lru_map_delete_elem(struct bpf_map *map, void *key)
{
	struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
1093
	struct hlist_nulls_head *head;
1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112
	struct bucket *b;
	struct htab_elem *l;
	unsigned long flags;
	u32 hash, key_size;
	int ret = -ENOENT;

	WARN_ON_ONCE(!rcu_read_lock_held());

	key_size = map->key_size;

	hash = htab_map_hash(key, key_size);
	b = __select_bucket(htab, hash);
	head = &b->head;

	raw_spin_lock_irqsave(&b->lock, flags);

	l = lookup_elem_raw(head, hash, key, key_size);

	if (l) {
1113
		hlist_nulls_del_rcu(&l->hash_node);
1114 1115 1116 1117 1118 1119 1120 1121 1122
		ret = 0;
	}

	raw_spin_unlock_irqrestore(&b->lock, flags);
	if (l)
		bpf_lru_push_free(&htab->lru, &l->lru_node);
	return ret;
}

1123 1124 1125 1126 1127
static void delete_all_elements(struct bpf_htab *htab)
{
	int i;

	for (i = 0; i < htab->n_buckets; i++) {
1128 1129
		struct hlist_nulls_head *head = select_bucket(htab, i);
		struct hlist_nulls_node *n;
1130 1131
		struct htab_elem *l;

1132 1133
		hlist_nulls_for_each_entry_safe(l, n, head, hash_node) {
			hlist_nulls_del_rcu(&l->hash_node);
1134
			htab_elem_free(htab, l);
1135 1136 1137
		}
	}
}
1138

1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150
/* Called when map->refcnt goes to zero, either from workqueue or from syscall */
static void htab_map_free(struct bpf_map *map)
{
	struct bpf_htab *htab = container_of(map, struct bpf_htab, map);

	/* at this point bpf_prog->aux->refcnt == 0 and this map->refcnt == 0,
	 * so the programs (can be more than one that used this map) were
	 * disconnected from events. Wait for outstanding critical sections in
	 * these programs to complete
	 */
	synchronize_rcu();

1151 1152
	/* some of free_htab_elem() callbacks for elements of this map may
	 * not have executed. Wait for them.
1153
	 */
1154
	rcu_barrier();
1155
	if (!htab_is_prealloc(htab))
1156
		delete_all_elements(htab);
1157 1158 1159
	else
		prealloc_destroy(htab);

1160
	free_percpu(htab->extra_elems);
1161
	bpf_map_area_free(htab->buckets);
1162 1163 1164
	kfree(htab);
}

1165
const struct bpf_map_ops htab_map_ops = {
1166
	.map_alloc_check = htab_map_alloc_check,
1167 1168 1169 1170 1171 1172
	.map_alloc = htab_map_alloc,
	.map_free = htab_map_free,
	.map_get_next_key = htab_map_get_next_key,
	.map_lookup_elem = htab_map_lookup_elem,
	.map_update_elem = htab_map_update_elem,
	.map_delete_elem = htab_map_delete_elem,
1173
	.map_gen_lookup = htab_map_gen_lookup,
1174 1175
};

1176
const struct bpf_map_ops htab_lru_map_ops = {
1177
	.map_alloc_check = htab_map_alloc_check,
1178 1179 1180 1181 1182 1183
	.map_alloc = htab_map_alloc,
	.map_free = htab_map_free,
	.map_get_next_key = htab_map_get_next_key,
	.map_lookup_elem = htab_lru_map_lookup_elem,
	.map_update_elem = htab_lru_map_update_elem,
	.map_delete_elem = htab_lru_map_delete_elem,
1184
	.map_gen_lookup = htab_lru_map_gen_lookup,
1185 1186
};

1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197
/* Called from eBPF program */
static void *htab_percpu_map_lookup_elem(struct bpf_map *map, void *key)
{
	struct htab_elem *l = __htab_map_lookup_elem(map, key);

	if (l)
		return this_cpu_ptr(htab_elem_get_ptr(l, map->key_size));
	else
		return NULL;
}

1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209
static void *htab_lru_percpu_map_lookup_elem(struct bpf_map *map, void *key)
{
	struct htab_elem *l = __htab_map_lookup_elem(map, key);

	if (l) {
		bpf_lru_node_set_ref(&l->lru_node);
		return this_cpu_ptr(htab_elem_get_ptr(l, map->key_size));
	}

	return NULL;
}

1210 1211
int bpf_percpu_hash_copy(struct bpf_map *map, void *key, void *value)
{
1212
	struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227
	struct htab_elem *l;
	void __percpu *pptr;
	int ret = -ENOENT;
	int cpu, off = 0;
	u32 size;

	/* per_cpu areas are zero-filled and bpf programs can only
	 * access 'value_size' of them, so copying rounded areas
	 * will not leak any kernel data
	 */
	size = round_up(map->value_size, 8);
	rcu_read_lock();
	l = __htab_map_lookup_elem(map, key);
	if (!l)
		goto out;
1228 1229
	if (htab_is_lru(htab))
		bpf_lru_node_set_ref(&l->lru_node);
1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244
	pptr = htab_elem_get_ptr(l, map->key_size);
	for_each_possible_cpu(cpu) {
		bpf_long_memcpy(value + off,
				per_cpu_ptr(pptr, cpu), size);
		off += size;
	}
	ret = 0;
out:
	rcu_read_unlock();
	return ret;
}

int bpf_percpu_hash_update(struct bpf_map *map, void *key, void *value,
			   u64 map_flags)
{
1245
	struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
1246 1247 1248
	int ret;

	rcu_read_lock();
1249 1250 1251 1252 1253 1254
	if (htab_is_lru(htab))
		ret = __htab_lru_percpu_map_update_elem(map, key, value,
							map_flags, true);
	else
		ret = __htab_percpu_map_update_elem(map, key, value, map_flags,
						    true);
1255 1256 1257
	rcu_read_unlock();

	return ret;
1258 1259
}

1260
const struct bpf_map_ops htab_percpu_map_ops = {
1261
	.map_alloc_check = htab_map_alloc_check,
1262 1263 1264 1265 1266 1267 1268 1269
	.map_alloc = htab_map_alloc,
	.map_free = htab_map_free,
	.map_get_next_key = htab_map_get_next_key,
	.map_lookup_elem = htab_percpu_map_lookup_elem,
	.map_update_elem = htab_percpu_map_update_elem,
	.map_delete_elem = htab_map_delete_elem,
};

1270
const struct bpf_map_ops htab_lru_percpu_map_ops = {
1271
	.map_alloc_check = htab_map_alloc_check,
1272 1273 1274 1275 1276 1277 1278 1279
	.map_alloc = htab_map_alloc,
	.map_free = htab_map_free,
	.map_get_next_key = htab_map_get_next_key,
	.map_lookup_elem = htab_lru_percpu_map_lookup_elem,
	.map_update_elem = htab_lru_percpu_map_update_elem,
	.map_delete_elem = htab_lru_map_delete_elem,
};

1280
static int fd_htab_map_alloc_check(union bpf_attr *attr)
1281 1282
{
	if (attr->value_size != sizeof(u32))
1283 1284
		return -EINVAL;
	return htab_map_alloc_check(attr);
1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307
}

static void fd_htab_map_free(struct bpf_map *map)
{
	struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
	struct hlist_nulls_node *n;
	struct hlist_nulls_head *head;
	struct htab_elem *l;
	int i;

	for (i = 0; i < htab->n_buckets; i++) {
		head = select_bucket(htab, i);

		hlist_nulls_for_each_entry_safe(l, n, head, hash_node) {
			void *ptr = fd_htab_map_get_ptr(map, l);

			map->ops->map_fd_put_ptr(ptr);
		}
	}

	htab_map_free(map);
}

1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327
/* only called from syscall */
int bpf_fd_htab_map_lookup_elem(struct bpf_map *map, void *key, u32 *value)
{
	void **ptr;
	int ret = 0;

	if (!map->ops->map_fd_sys_lookup_elem)
		return -ENOTSUPP;

	rcu_read_lock();
	ptr = htab_map_lookup_elem(map, key);
	if (ptr)
		*value = map->ops->map_fd_sys_lookup_elem(READ_ONCE(*ptr));
	else
		ret = -ENOENT;
	rcu_read_unlock();

	return ret;
}

1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354
/* only called from syscall */
int bpf_fd_htab_map_update_elem(struct bpf_map *map, struct file *map_file,
				void *key, void *value, u64 map_flags)
{
	void *ptr;
	int ret;
	u32 ufd = *(u32 *)value;

	ptr = map->ops->map_fd_get_ptr(map, map_file, ufd);
	if (IS_ERR(ptr))
		return PTR_ERR(ptr);

	ret = htab_map_update_elem(map, key, &ptr, map_flags);
	if (ret)
		map->ops->map_fd_put_ptr(ptr);

	return ret;
}

static struct bpf_map *htab_of_map_alloc(union bpf_attr *attr)
{
	struct bpf_map *map, *inner_map_meta;

	inner_map_meta = bpf_map_meta_alloc(attr->inner_map_fd);
	if (IS_ERR(inner_map_meta))
		return inner_map_meta;

1355
	map = htab_map_alloc(attr);
1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375
	if (IS_ERR(map)) {
		bpf_map_meta_free(inner_map_meta);
		return map;
	}

	map->inner_map_meta = inner_map_meta;

	return map;
}

static void *htab_of_map_lookup_elem(struct bpf_map *map, void *key)
{
	struct bpf_map **inner_map  = htab_map_lookup_elem(map, key);

	if (!inner_map)
		return NULL;

	return READ_ONCE(*inner_map);
}

1376 1377 1378 1379 1380 1381
static u32 htab_of_map_gen_lookup(struct bpf_map *map,
				  struct bpf_insn *insn_buf)
{
	struct bpf_insn *insn = insn_buf;
	const int ret = BPF_REG_0;

1382 1383 1384
	BUILD_BUG_ON(!__same_type(&__htab_map_lookup_elem,
		     (void *(*)(struct bpf_map *map, void *key))NULL));
	*insn++ = BPF_EMIT_CALL(BPF_CAST_CALL(__htab_map_lookup_elem));
1385 1386 1387 1388 1389 1390 1391 1392 1393
	*insn++ = BPF_JMP_IMM(BPF_JEQ, ret, 0, 2);
	*insn++ = BPF_ALU64_IMM(BPF_ADD, ret,
				offsetof(struct htab_elem, key) +
				round_up(map->key_size, 8));
	*insn++ = BPF_LDX_MEM(BPF_DW, ret, ret, 0);

	return insn - insn_buf;
}

1394 1395 1396 1397 1398 1399
static void htab_of_map_free(struct bpf_map *map)
{
	bpf_map_meta_free(map->inner_map_meta);
	fd_htab_map_free(map);
}

1400
const struct bpf_map_ops htab_of_maps_map_ops = {
1401
	.map_alloc_check = fd_htab_map_alloc_check,
1402 1403 1404 1405 1406 1407 1408
	.map_alloc = htab_of_map_alloc,
	.map_free = htab_of_map_free,
	.map_get_next_key = htab_map_get_next_key,
	.map_lookup_elem = htab_of_map_lookup_elem,
	.map_delete_elem = htab_map_delete_elem,
	.map_fd_get_ptr = bpf_map_fd_get_ptr,
	.map_fd_put_ptr = bpf_map_fd_put_ptr,
1409
	.map_fd_sys_lookup_elem = bpf_map_fd_sys_lookup_elem,
1410
	.map_gen_lookup = htab_of_map_gen_lookup,
1411
};