kernel/bpf/hashtab.c

   1 /* Copyright (c) 2011-2014 PLUMgrid, http://plumgrid.com
   2  * Copyright (c) 2016 Facebook
   3  *
   4  * This program is free software; you can redistribute it and/or
   5  * modify it under the terms of version 2 of the GNU General Public
   6  * License as published by the Free Software Foundation.
   7  *
   8  * This program is distributed in the hope that it will be useful, but
   9  * WITHOUT ANY WARRANTY; without even the implied warranty of
  10  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
  11  * General Public License for more details.
  12  */
  13 #include <linux/bpf.h>
  14 #include <linux/jhash.h>
  15 #include <linux/filter.h>
  16 #include <linux/vmalloc.h>
  17 #include "percpu_freelist.h"
  18
  19 struct bucket {
  20         struct hlist_head head;
  21         raw_spinlock_t lock;
  22 };
  23
  24 struct bpf_htab {
  25         struct bpf_map map;
  26         struct bucket *buckets;
  27         void *elems;
  28         struct pcpu_freelist freelist;
  29         void __percpu *extra_elems;
  30         atomic_t count; /* number of elements in this hashtable */
  31         u32 n_buckets;  /* number of hash buckets */
  32         u32 elem_size;  /* size of each element in bytes */
  33 };
  34
  35 enum extra_elem_state {
  36         HTAB_NOT_AN_EXTRA_ELEM = 0,
  37         HTAB_EXTRA_ELEM_FREE,
  38         HTAB_EXTRA_ELEM_USED
  39 };
  40
  41 /* each htab element is struct htab_elem + key + value */
  42 struct htab_elem {
  43         union {
  44                 struct hlist_node hash_node;
  45                 struct bpf_htab *htab;
  46                 struct pcpu_freelist_node fnode;
  47         };
  48         union {
  49                 struct rcu_head rcu;
  50                 enum extra_elem_state state;
  51         };
  52         u32 hash;
  53         char key[0] __aligned(8);
  54 };
  55
  56 static inline void htab_elem_set_ptr(struct htab_elem *l, u32 key_size,
  57                                      void __percpu *pptr)
  58 {
  59         *(void __percpu **)(l->key + key_size) = pptr;
  60 }
  61
  62 static inline void __percpu *htab_elem_get_ptr(struct htab_elem *l, u32 key_size)
  63 {
  64         return *(void __percpu **)(l->key + key_size);
  65 }
  66
  67 static struct htab_elem *get_htab_elem(struct bpf_htab *htab, int i)
  68 {
  69         return (struct htab_elem *) (htab->elems + i * htab->elem_size);
  70 }
  71
  72 static void htab_free_elems(struct bpf_htab *htab)
  73 {
  74         int i;
  75
  76         if (htab->map.map_type != BPF_MAP_TYPE_PERCPU_HASH)
  77                 goto free_elems;
  78
  79         for (i = 0; i < htab->map.max_entries; i++) {
  80                 void __percpu *pptr;
  81
  82                 pptr = htab_elem_get_ptr(get_htab_elem(htab, i),
  83                                          htab->map.key_size);
  84                 free_percpu(pptr);
  85         }
  86 free_elems:
  87         vfree(htab->elems);
  88 }
  89
  90 static int prealloc_elems_and_freelist(struct bpf_htab *htab)
  91 {
  92         int err = -ENOMEM, i;
  93
  94         htab->elems = vzalloc(htab->elem_size * htab->map.max_entries);
  95         if (!htab->elems)
  96                 return -ENOMEM;
  97
  98         if (htab->map.map_type != BPF_MAP_TYPE_PERCPU_HASH)
  99                 goto skip_percpu_elems;
 100
 101         for (i = 0; i < htab->map.max_entries; i++) {
 102                 u32 size = round_up(htab->map.value_size, 8);
 103                 void __percpu *pptr;
 104
 105                 pptr = __alloc_percpu_gfp(size, 8, GFP_USER | __GFP_NOWARN);
 106                 if (!pptr)
 107                         goto free_elems;
 108                 htab_elem_set_ptr(get_htab_elem(htab, i), htab->map.key_size,
 109                                   pptr);
 110         }
 111
 112 skip_percpu_elems:
 113         err = pcpu_freelist_init(&htab->freelist);
 114         if (err)
 115                 goto free_elems;
 116
 117         pcpu_freelist_populate(&htab->freelist, htab->elems, htab->elem_size,
 118                                htab->map.max_entries);
 119         return 0;
 120
 121 free_elems:
 122         htab_free_elems(htab);
 123         return err;
 124 }
 125
 126 static int alloc_extra_elems(struct bpf_htab *htab)
 127 {
 128         void __percpu *pptr;
 129         int cpu;
 130
 131         pptr = __alloc_percpu_gfp(htab->elem_size, 8, GFP_USER | __GFP_NOWARN);
 132         if (!pptr)
 133                 return -ENOMEM;
 134
 135         for_each_possible_cpu(cpu) {
 136                 ((struct htab_elem *)per_cpu_ptr(pptr, cpu))->state =
 137                         HTAB_EXTRA_ELEM_FREE;
 138         }
 139         htab->extra_elems = pptr;
 140         return 0;
 141 }
 142
 143 /* Called from syscall */
 144 static struct bpf_map *htab_map_alloc(union bpf_attr *attr)
 145 {
 146         bool percpu = attr->map_type == BPF_MAP_TYPE_PERCPU_HASH;
 147         struct bpf_htab *htab;
 148         int err, i;
 149         u64 cost;
 150
 151         if (attr->map_flags & ~BPF_F_NO_PREALLOC)
 152                 /* reserved bits should not be used */
 153                 return ERR_PTR(-EINVAL);
 154
 155         htab = kzalloc(sizeof(*htab), GFP_USER);
 156         if (!htab)
 157                 return ERR_PTR(-ENOMEM);
 158
 159         /* mandatory map attributes */
 160         htab->map.map_type = attr->map_type;
 161         htab->map.key_size = attr->key_size;
 162         htab->map.value_size = attr->value_size;
 163         htab->map.max_entries = attr->max_entries;
 164         htab->map.map_flags = attr->map_flags;
 165
 166         /* check sanity of attributes.
 167          * value_size == 0 may be allowed in the future to use map as a set
 168          */
 169         err = -EINVAL;
 170         if (htab->map.max_entries == 0 || htab->map.key_size == 0 ||
 171             htab->map.value_size == 0)
 172                 goto free_htab;
 173
 174         /* hash table size must be power of 2 */
 175         htab->n_buckets = roundup_pow_of_two(htab->map.max_entries);
 176
 177         err = -E2BIG;
 178         if (htab->map.key_size > MAX_BPF_STACK)
 179                 /* eBPF programs initialize keys on stack, so they cannot be
 180                  * larger than max stack size
 181                  */
 182                 goto free_htab;
 183
 184         if (htab->map.value_size >= (1 << (KMALLOC_SHIFT_MAX - 1)) -
 185             MAX_BPF_STACK - sizeof(struct htab_elem))
 186                 /* if value_size is bigger, the user space won't be able to
 187                  * access the elements via bpf syscall. This check also makes
 188                  * sure that the elem_size doesn't overflow and it's
 189                  * kmalloc-able later in htab_map_update_elem()
 190                  */
 191                 goto free_htab;
 192
 193         if (percpu && round_up(htab->map.value_size, 8) > PCPU_MIN_UNIT_SIZE)
 194                 /* make sure the size for pcpu_alloc() is reasonable */
 195                 goto free_htab;
 196
 197         htab->elem_size = sizeof(struct htab_elem) +
 198                           round_up(htab->map.key_size, 8);
 199         if (percpu)
 200                 htab->elem_size += sizeof(void *);
 201         else
 202                 htab->elem_size += round_up(htab->map.value_size, 8);
 203
 204         /* prevent zero size kmalloc and check for u32 overflow */
 205         if (htab->n_buckets == 0 ||
 206             htab->n_buckets > U32_MAX / sizeof(struct bucket))
 207                 goto free_htab;
 208
 209         cost = (u64) htab->n_buckets * sizeof(struct bucket) +
 210                (u64) htab->elem_size * htab->map.max_entries;
 211
 212         if (percpu)
 213                 cost += (u64) round_up(htab->map.value_size, 8) *
 214                         num_possible_cpus() * htab->map.max_entries;
 215         else
 216                cost += (u64) htab->elem_size * num_possible_cpus();
 217
 218         if (cost >= U32_MAX - PAGE_SIZE)
 219                 /* make sure page count doesn't overflow */
 220                 goto free_htab;
 221
 222         htab->map.pages = round_up(cost, PAGE_SIZE) >> PAGE_SHIFT;
 223
 224         /* if map size is larger than memlock limit, reject it early */
 225         err = bpf_map_precharge_memlock(htab->map.pages);
 226         if (err)
 227                 goto free_htab;
 228
 229         err = -ENOMEM;
 230         htab->buckets = kmalloc_array(htab->n_buckets, sizeof(struct bucket),
 231                                       GFP_USER | __GFP_NOWARN);
 232
 233         if (!htab->buckets) {
 234                 htab->buckets = vmalloc(htab->n_buckets * sizeof(struct bucket));
 235                 if (!htab->buckets)
 236                         goto free_htab;
 237         }
 238
 239         for (i = 0; i < htab->n_buckets; i++) {
 240                 INIT_HLIST_HEAD(&htab->buckets[i].head);
 241                 raw_spin_lock_init(&htab->buckets[i].lock);
 242         }
 243
 244         if (!percpu) {
 245                 err = alloc_extra_elems(htab);
 246                 if (err)
 247                         goto free_buckets;
 248         }
 249
 250         if (!(attr->map_flags & BPF_F_NO_PREALLOC)) {
 251                 err = prealloc_elems_and_freelist(htab);
 252                 if (err)
 253                         goto free_extra_elems;
 254         }
 255
 256         return &htab->map;
 257
 258 free_extra_elems:
 259         free_percpu(htab->extra_elems);
 260 free_buckets:
 261         kvfree(htab->buckets);
 262 free_htab:
 263         kfree(htab);
 264         return ERR_PTR(err);
 265 }
 266
 267 static inline u32 htab_map_hash(const void *key, u32 key_len)
 268 {
 269         return jhash(key, key_len, 0);
 270 }
 271
 272 static inline struct bucket *__select_bucket(struct bpf_htab *htab, u32 hash)
 273 {
 274         return &htab->buckets[hash & (htab->n_buckets - 1)];
 275 }
 276
 277 static inline struct hlist_head *select_bucket(struct bpf_htab *htab, u32 hash)
 278 {
 279         return &__select_bucket(htab, hash)->head;
 280 }
 281
 282 static struct htab_elem *lookup_elem_raw(struct hlist_head *head, u32 hash,
 283                                          void *key, u32 key_size)
 284 {
 285         struct htab_elem *l;
 286
 287         hlist_for_each_entry_rcu(l, head, hash_node)
 288                 if (l->hash == hash && !memcmp(&l->key, key, key_size))
 289                         return l;
 290
 291         return NULL;
 292 }
 293
 294 /* Called from syscall or from eBPF program */
 295 static void *__htab_map_lookup_elem(struct bpf_map *map, void *key)
 296 {
 297         struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
 298         struct hlist_head *head;
 299         struct htab_elem *l;
 300         u32 hash, key_size;
 301
 302         /* Must be called with rcu_read_lock. */
 303         WARN_ON_ONCE(!rcu_read_lock_held());
 304
 305         key_size = map->key_size;
 306
 307         hash = htab_map_hash(key, key_size);
 308
 309         head = select_bucket(htab, hash);
 310
 311         l = lookup_elem_raw(head, hash, key, key_size);
 312
 313         return l;
 314 }
 315
 316 static void *htab_map_lookup_elem(struct bpf_map *map, void *key)
 317 {
 318         struct htab_elem *l = __htab_map_lookup_elem(map, key);
 319
 320         if (l)
 321                 return l->key + round_up(map->key_size, 8);
 322
 323         return NULL;
 324 }
 325
 326 /* Called from syscall */
 327 static int htab_map_get_next_key(struct bpf_map *map, void *key, void *next_key)
 328 {
 329         struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
 330         struct hlist_head *head;
 331         struct htab_elem *l, *next_l;
 332         u32 hash, key_size;
 333         int i;
 334
 335         WARN_ON_ONCE(!rcu_read_lock_held());
 336
 337         key_size = map->key_size;
 338
 339         hash = htab_map_hash(key, key_size);
 340
 341         head = select_bucket(htab, hash);
 342
 343         /* lookup the key */
 344         l = lookup_elem_raw(head, hash, key, key_size);
 345
 346         if (!l) {
 347                 i = 0;
 348                 goto find_first_elem;
 349         }
 350
 351         /* key was found, get next key in the same bucket */
 352         next_l = hlist_entry_safe(rcu_dereference_raw(hlist_next_rcu(&l->hash_node)),
 353                                   struct htab_elem, hash_node);
 354
 355         if (next_l) {
 356                 /* if next elem in this hash list is non-zero, just return it */
 357                 memcpy(next_key, next_l->key, key_size);
 358                 return 0;
 359         }
 360
 361         /* no more elements in this hash list, go to the next bucket */
 362         i = hash & (htab->n_buckets - 1);
 363         i++;
 364
 365 find_first_elem:
 366         /* iterate over buckets */
 367         for (; i < htab->n_buckets; i++) {
 368                 head = select_bucket(htab, i);
 369
 370                 /* pick first element in the bucket */
 371                 next_l = hlist_entry_safe(rcu_dereference_raw(hlist_first_rcu(head)),
 372                                           struct htab_elem, hash_node);
 373                 if (next_l) {
 374                         /* if it's not empty, just return it */
 375                         memcpy(next_key, next_l->key, key_size);
 376                         return 0;
 377                 }
 378         }
 379
 380         /* iterated over all buckets and all elements */
 381         return -ENOENT;
 382 }
 383
 384 static void htab_elem_free(struct bpf_htab *htab, struct htab_elem *l)
 385 {
 386         if (htab->map.map_type == BPF_MAP_TYPE_PERCPU_HASH)
 387                 free_percpu(htab_elem_get_ptr(l, htab->map.key_size));
 388         kfree(l);
 389 }
 390
 391 static void htab_elem_free_rcu(struct rcu_head *head)
 392 {
 393         struct htab_elem *l = container_of(head, struct htab_elem, rcu);
 394         struct bpf_htab *htab = l->htab;
 395
 396         /* must increment bpf_prog_active to avoid kprobe+bpf triggering while
 397          * we're calling kfree, otherwise deadlock is possible if kprobes
 398          * are placed somewhere inside of slub
 399          */
 400         preempt_disable();
 401         __this_cpu_inc(bpf_prog_active);
 402         htab_elem_free(htab, l);
 403         __this_cpu_dec(bpf_prog_active);
 404         preempt_enable();
 405 }
 406
 407 static void free_htab_elem(struct bpf_htab *htab, struct htab_elem *l)
 408 {
 409         if (l->state == HTAB_EXTRA_ELEM_USED) {
 410                 l->state = HTAB_EXTRA_ELEM_FREE;
 411                 return;
 412         }
 413
 414         if (!(htab->map.map_flags & BPF_F_NO_PREALLOC)) {
 415                 pcpu_freelist_push(&htab->freelist, &l->fnode);
 416         } else {
 417                 atomic_dec(&htab->count);
 418                 l->htab = htab;
 419                 call_rcu(&l->rcu, htab_elem_free_rcu);
 420         }
 421 }
 422
 423 static struct htab_elem *alloc_htab_elem(struct bpf_htab *htab, void *key,
 424                                          void *value, u32 key_size, u32 hash,
 425                                          bool percpu, bool onallcpus,
 426                                          bool old_elem_exists)
 427 {
 428         u32 size = htab->map.value_size;
 429         bool prealloc = !(htab->map.map_flags & BPF_F_NO_PREALLOC);
 430         struct htab_elem *l_new;
 431         void __percpu *pptr;
 432         int err = 0;
 433
 434         if (prealloc) {
 435                 l_new = (struct htab_elem *)pcpu_freelist_pop(&htab->freelist);
 436                 if (!l_new)
 437                         err = -E2BIG;
 438         } else {
 439                 if (atomic_inc_return(&htab->count) > htab->map.max_entries) {
 440                         atomic_dec(&htab->count);
 441                         err = -E2BIG;
 442                 } else {
 443                         l_new = kmalloc(htab->elem_size,
 444                                         GFP_ATOMIC | __GFP_NOWARN);
 445                         if (!l_new)
 446                                 return ERR_PTR(-ENOMEM);
 447                 }
 448         }
 449
 450         if (err) {
 451                 if (!old_elem_exists)
 452                         return ERR_PTR(err);
 453
 454                 /* if we're updating the existing element and the hash table
 455                  * is full, use per-cpu extra elems
 456                  */
 457                 l_new = this_cpu_ptr(htab->extra_elems);
 458                 if (l_new->state != HTAB_EXTRA_ELEM_FREE)
 459                         return ERR_PTR(-E2BIG);
 460                 l_new->state = HTAB_EXTRA_ELEM_USED;
 461         } else {
 462                 l_new->state = HTAB_NOT_AN_EXTRA_ELEM;
 463         }
 464
 465         memcpy(l_new->key, key, key_size);
 466         if (percpu) {
 467                 /* round up value_size to 8 bytes */
 468                 size = round_up(size, 8);
 469
 470                 if (prealloc) {
 471                         pptr = htab_elem_get_ptr(l_new, key_size);
 472                 } else {
 473                         /* alloc_percpu zero-fills */
 474                         pptr = __alloc_percpu_gfp(size, 8,
 475                                                   GFP_ATOMIC | __GFP_NOWARN);
 476                         if (!pptr) {
 477                                 kfree(l_new);
 478                                 return ERR_PTR(-ENOMEM);
 479                         }
 480                 }
 481
 482                 if (!onallcpus) {
 483                         /* copy true value_size bytes */
 484                         memcpy(this_cpu_ptr(pptr), value, htab->map.value_size);
 485                 } else {
 486                         int off = 0, cpu;
 487
 488                         for_each_possible_cpu(cpu) {
 489                                 bpf_long_memcpy(per_cpu_ptr(pptr, cpu),
 490                                                 value + off, size);
 491                                 off += size;
 492                         }
 493                 }
 494                 if (!prealloc)
 495                         htab_elem_set_ptr(l_new, key_size, pptr);
 496         } else {
 497                 memcpy(l_new->key + round_up(key_size, 8), value, size);
 498         }
 499
 500         l_new->hash = hash;
 501         return l_new;
 502 }
 503
 504 static int check_flags(struct bpf_htab *htab, struct htab_elem *l_old,
 505                        u64 map_flags)
 506 {
 507         if (l_old && map_flags == BPF_NOEXIST)
 508                 /* elem already exists */
 509                 return -EEXIST;
 510
 511         if (!l_old && map_flags == BPF_EXIST)
 512                 /* elem doesn't exist, cannot update it */
 513                 return -ENOENT;
 514
 515         return 0;
 516 }
 517
 518 /* Called from syscall or from eBPF program */
 519 static int htab_map_update_elem(struct bpf_map *map, void *key, void *value,
 520                                 u64 map_flags)
 521 {
 522         struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
 523         struct htab_elem *l_new = NULL, *l_old;
 524         struct hlist_head *head;
 525         unsigned long flags;
 526         struct bucket *b;
 527         u32 key_size, hash;
 528         int ret;
 529
 530         if (unlikely(map_flags > BPF_EXIST))
 531                 /* unknown flags */
 532                 return -EINVAL;
 533
 534         WARN_ON_ONCE(!rcu_read_lock_held());
 535
 536         key_size = map->key_size;
 537
 538         hash = htab_map_hash(key, key_size);
 539
 540         b = __select_bucket(htab, hash);
 541         head = &b->head;
 542
 543         /* bpf_map_update_elem() can be called in_irq() */
 544         raw_spin_lock_irqsave(&b->lock, flags);
 545
 546         l_old = lookup_elem_raw(head, hash, key, key_size);
 547
 548         ret = check_flags(htab, l_old, map_flags);
 549         if (ret)
 550                 goto err;
 551
 552         l_new = alloc_htab_elem(htab, key, value, key_size, hash, false, false,
 553                                 !!l_old);
 554         if (IS_ERR(l_new)) {
 555                 /* all pre-allocated elements are in use or memory exhausted */
 556                 ret = PTR_ERR(l_new);
 557                 goto err;
 558         }
 559
 560         /* add new element to the head of the list, so that
 561          * concurrent search will find it before old elem
 562          */
 563         hlist_add_head_rcu(&l_new->hash_node, head);
 564         if (l_old) {
 565                 hlist_del_rcu(&l_old->hash_node);
 566                 free_htab_elem(htab, l_old);
 567         }
 568         ret = 0;
 569 err:
 570         raw_spin_unlock_irqrestore(&b->lock, flags);
 571         return ret;
 572 }
 573
 574 static int __htab_percpu_map_update_elem(struct bpf_map *map, void *key,
 575                                          void *value, u64 map_flags,
 576                                          bool onallcpus)
 577 {
 578         struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
 579         struct htab_elem *l_new = NULL, *l_old;
 580         struct hlist_head *head;
 581         unsigned long flags;
 582         struct bucket *b;
 583         u32 key_size, hash;
 584         int ret;
 585
 586         if (unlikely(map_flags > BPF_EXIST))
 587                 /* unknown flags */
 588                 return -EINVAL;
 589
 590         WARN_ON_ONCE(!rcu_read_lock_held());
 591
 592         key_size = map->key_size;
 593
 594         hash = htab_map_hash(key, key_size);
 595
 596         b = __select_bucket(htab, hash);
 597         head = &b->head;
 598
 599         /* bpf_map_update_elem() can be called in_irq() */
 600         raw_spin_lock_irqsave(&b->lock, flags);
 601
 602         l_old = lookup_elem_raw(head, hash, key, key_size);
 603
 604         ret = check_flags(htab, l_old, map_flags);
 605         if (ret)
 606                 goto err;
 607
 608         if (l_old) {
 609                 void __percpu *pptr = htab_elem_get_ptr(l_old, key_size);
 610                 u32 size = htab->map.value_size;
 611
 612                 /* per-cpu hash map can update value in-place */
 613                 if (!onallcpus) {
 614                         memcpy(this_cpu_ptr(pptr), value, size);
 615                 } else {
 616                         int off = 0, cpu;
 617
 618                         size = round_up(size, 8);
 619                         for_each_possible_cpu(cpu) {
 620                                 bpf_long_memcpy(per_cpu_ptr(pptr, cpu),
 621                                                 value + off, size);
 622                                 off += size;
 623                         }
 624                 }
 625         } else {
 626                 l_new = alloc_htab_elem(htab, key, value, key_size,
 627                                         hash, true, onallcpus, false);
 628                 if (IS_ERR(l_new)) {
 629                         ret = PTR_ERR(l_new);
 630                         goto err;
 631                 }
 632                 hlist_add_head_rcu(&l_new->hash_node, head);
 633         }
 634         ret = 0;
 635 err:
 636         raw_spin_unlock_irqrestore(&b->lock, flags);
 637         return ret;
 638 }
 639
 640 static int htab_percpu_map_update_elem(struct bpf_map *map, void *key,
 641                                        void *value, u64 map_flags)
 642 {
 643         return __htab_percpu_map_update_elem(map, key, value, map_flags, false);
 644 }
 645
 646 /* Called from syscall or from eBPF program */
 647 static int htab_map_delete_elem(struct bpf_map *map, void *key)
 648 {
 649         struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
 650         struct hlist_head *head;
 651         struct bucket *b;
 652         struct htab_elem *l;
 653         unsigned long flags;
 654         u32 hash, key_size;
 655         int ret = -ENOENT;
 656
 657         WARN_ON_ONCE(!rcu_read_lock_held());
 658
 659         key_size = map->key_size;
 660
 661         hash = htab_map_hash(key, key_size);
 662         b = __select_bucket(htab, hash);
 663         head = &b->head;
 664
 665         raw_spin_lock_irqsave(&b->lock, flags);
 666
 667         l = lookup_elem_raw(head, hash, key, key_size);
 668
 669         if (l) {
 670                 hlist_del_rcu(&l->hash_node);
 671                 free_htab_elem(htab, l);
 672                 ret = 0;
 673         }
 674
 675         raw_spin_unlock_irqrestore(&b->lock, flags);
 676         return ret;
 677 }
 678
 679 static void delete_all_elements(struct bpf_htab *htab)
 680 {
 681         int i;
 682
 683         for (i = 0; i < htab->n_buckets; i++) {
 684                 struct hlist_head *head = select_bucket(htab, i);
 685                 struct hlist_node *n;
 686                 struct htab_elem *l;
 687
 688                 hlist_for_each_entry_safe(l, n, head, hash_node) {
 689                         hlist_del_rcu(&l->hash_node);
 690                         htab_elem_free(htab, l);
 691                 }
 692         }
 693 }
 694 /* Called when map->refcnt goes to zero, either from workqueue or from syscall */
 695 static void htab_map_free(struct bpf_map *map)
 696 {
 697         struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
 698
 699         /* at this point bpf_prog->aux->refcnt == 0 and this map->refcnt == 0,
 700          * so the programs (can be more than one that used this map) were
 701          * disconnected from events. Wait for outstanding critical sections in
 702          * these programs to complete
 703          */
 704         synchronize_rcu();
 705
 706         /* some of free_htab_elem() callbacks for elements of this map may
 707          * not have executed. Wait for them.
 708          */
 709         rcu_barrier();
 710         if (htab->map.map_flags & BPF_F_NO_PREALLOC) {
 711                 delete_all_elements(htab);
 712         } else {
 713                 htab_free_elems(htab);
 714                 pcpu_freelist_destroy(&htab->freelist);
 715         }
 716         free_percpu(htab->extra_elems);
 717         kvfree(htab->buckets);
 718         kfree(htab);
 719 }
 720
 721 static const struct bpf_map_ops htab_ops = {
 722         .map_alloc = htab_map_alloc,
 723         .map_free = htab_map_free,
 724         .map_get_next_key = htab_map_get_next_key,
 725         .map_lookup_elem = htab_map_lookup_elem,
 726         .map_update_elem = htab_map_update_elem,
 727         .map_delete_elem = htab_map_delete_elem,
 728 };
 729
 730 static struct bpf_map_type_list htab_type __read_mostly = {
 731         .ops = &htab_ops,
 732         .type = BPF_MAP_TYPE_HASH,
 733 };
 734
 735 /* Called from eBPF program */
 736 static void *htab_percpu_map_lookup_elem(struct bpf_map *map, void *key)
 737 {
 738         struct htab_elem *l = __htab_map_lookup_elem(map, key);
 739
 740         if (l)
 741                 return this_cpu_ptr(htab_elem_get_ptr(l, map->key_size));
 742         else
 743                 return NULL;
 744 }
 745
 746 int bpf_percpu_hash_copy(struct bpf_map *map, void *key, void *value)
 747 {
 748         struct htab_elem *l;
 749         void __percpu *pptr;
 750         int ret = -ENOENT;
 751         int cpu, off = 0;
 752         u32 size;
 753
 754         /* per_cpu areas are zero-filled and bpf programs can only
 755          * access 'value_size' of them, so copying rounded areas
 756          * will not leak any kernel data
 757          */
 758         size = round_up(map->value_size, 8);
 759         rcu_read_lock();
 760         l = __htab_map_lookup_elem(map, key);
 761         if (!l)
 762                 goto out;
 763         pptr = htab_elem_get_ptr(l, map->key_size);
 764         for_each_possible_cpu(cpu) {
 765                 bpf_long_memcpy(value + off,
 766                                 per_cpu_ptr(pptr, cpu), size);
 767                 off += size;
 768         }
 769         ret = 0;
 770 out:
 771         rcu_read_unlock();
 772         return ret;
 773 }
 774
 775 int bpf_percpu_hash_update(struct bpf_map *map, void *key, void *value,
 776                            u64 map_flags)
 777 {
 778         int ret;
 779
 780         rcu_read_lock();
 781         ret = __htab_percpu_map_update_elem(map, key, value, map_flags, true);
 782         rcu_read_unlock();
 783
 784         return ret;
 785 }
 786
 787 static const struct bpf_map_ops htab_percpu_ops = {
 788         .map_alloc = htab_map_alloc,
 789         .map_free = htab_map_free,
 790         .map_get_next_key = htab_map_get_next_key,
 791         .map_lookup_elem = htab_percpu_map_lookup_elem,
 792         .map_update_elem = htab_percpu_map_update_elem,
 793         .map_delete_elem = htab_map_delete_elem,
 794 };
 795
 796 static struct bpf_map_type_list htab_percpu_type __read_mostly = {
 797         .ops = &htab_percpu_ops,
 798         .type = BPF_MAP_TYPE_PERCPU_HASH,
 799 };
 800
 801 static int __init register_htab_map(void)
 802 {
 803         bpf_register_map_type(&htab_type);
 804         bpf_register_map_type(&htab_percpu_type);
 805         return 0;
 806 }
 807 late_initcall(register_htab_map);