src/util/hash_table.c

   1 /*
   2  * Copyright © 2009,2012 Intel Corporation
   3  * Copyright © 1988-2004 Keith Packard and Bart Massey.
   4  *
   5  * Permission is hereby granted, free of charge, to any person obtaining a
   6  * copy of this software and associated documentation files (the "Software"),
   7  * to deal in the Software without restriction, including without limitation
   8  * the rights to use, copy, modify, merge, publish, distribute, sublicense,
   9  * and/or sell copies of the Software, and to permit persons to whom the
  10  * Software is furnished to do so, subject to the following conditions:
  11  *
  12  * The above copyright notice and this permission notice (including the next
  13  * paragraph) shall be included in all copies or substantial portions of the
  14  * Software.
  15  *
  16  * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
  17  * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
  18  * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
  19  * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
  20  * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
  21  * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
  22  * IN THE SOFTWARE.
  23  *
  24  * Except as contained in this notice, the names of the authors
  25  * or their institutions shall not be used in advertising or
  26  * otherwise to promote the sale, use or other dealings in this
  27  * Software without prior written authorization from the
  28  * authors.
  29  *
  30  * Authors:
  31  *    Eric Anholt <eric@anholt.net>
  32  *    Keith Packard <keithp@keithp.com>
  33  */
  34
  35 /**
  36  * Implements an open-addressing, linear-reprobing hash table.
  37  *
  38  * For more information, see:
  39  *
  40  * http://cgit.freedesktop.org/~anholt/hash_table/tree/README
  41  */
  42
  43 #include <stdlib.h>
  44 #include <string.h>
  45 #include <assert.h>
  46
  47 #include "hash_table.h"
  48 #include "ralloc.h"
  49 #include "macros.h"
  50 #include "main/hash.h"
  51
  52 static const uint32_t deleted_key_value;
  53
  54 /**
  55  * From Knuth -- a good choice for hash/rehash values is p, p-2 where
  56  * p and p-2 are both prime.  These tables are sized to have an extra 10%
  57  * free to avoid exponential performance degradation as the hash table fills
  58  */
  59 static const struct {
  60    uint32_t max_entries, size, rehash;
  61 } hash_sizes[] = {
  62    { 2,                 5,              3         },
  63    { 4,                 7,              5         },
  64    { 8,                 13,             11        },
  65    { 16,                19,             17        },
  66    { 32,                43,             41        },
  67    { 64,                73,             71        },
  68    { 128,               151,            149       },
  69    { 256,               283,            281       },
  70    { 512,               571,            569       },
  71    { 1024,              1153,           1151      },
  72    { 2048,              2269,           2267      },
  73    { 4096,              4519,           4517      },
  74    { 8192,              9013,           9011      },
  75    { 16384,             18043,          18041     },
  76    { 32768,             36109,          36107     },
  77    { 65536,             72091,          72089     },
  78    { 131072,            144409,         144407    },
  79    { 262144,            288361,         288359    },
  80    { 524288,            576883,         576881    },
  81    { 1048576,           1153459,        1153457   },
  82    { 2097152,           2307163,        2307161   },
  83    { 4194304,           4613893,        4613891   },
  84    { 8388608,           9227641,        9227639   },
  85    { 16777216,          18455029,       18455027  },
  86    { 33554432,          36911011,       36911009  },
  87    { 67108864,          73819861,       73819859  },
  88    { 134217728,         147639589,      147639587 },
  89    { 268435456,         295279081,      295279079 },
  90    { 536870912,         590559793,      590559791 },
  91    { 1073741824,        1181116273,     1181116271},
  92    { 2147483648ul,      2362232233ul,   2362232231ul}
  93 };
  94
  95 static int
  96 entry_is_free(const struct hash_entry *entry)
  97 {
  98    return entry->key == NULL;
  99 }
 100
 101 static int
 102 entry_is_deleted(const struct hash_table *ht, struct hash_entry *entry)
 103 {
 104    return entry->key == ht->deleted_key;
 105 }
 106
 107 static int
 108 entry_is_present(const struct hash_table *ht, struct hash_entry *entry)
 109 {
 110    return entry->key != NULL && entry->key != ht->deleted_key;
 111 }
 112
 113 bool
 114 _mesa_hash_table_init(struct hash_table *ht,
 115                       void *mem_ctx,
 116                       uint32_t (*key_hash_function)(const void *key),
 117                       bool (*key_equals_function)(const void *a,
 118                                                   const void *b))
 119 {
 120    ht->size_index = 0;
 121    ht->size = hash_sizes[ht->size_index].size;
 122    ht->rehash = hash_sizes[ht->size_index].rehash;
 123    ht->max_entries = hash_sizes[ht->size_index].max_entries;
 124    ht->key_hash_function = key_hash_function;
 125    ht->key_equals_function = key_equals_function;
 126    ht->table = rzalloc_array(mem_ctx, struct hash_entry, ht->size);
 127    ht->entries = 0;
 128    ht->deleted_entries = 0;
 129    ht->deleted_key = &deleted_key_value;
 130
 131    return ht->table != NULL;
 132 }
 133
 134 struct hash_table *
 135 _mesa_hash_table_create(void *mem_ctx,
 136                         uint32_t (*key_hash_function)(const void *key),
 137                         bool (*key_equals_function)(const void *a,
 138                                                     const void *b))
 139 {
 140    struct hash_table *ht;
 141
 142    /* mem_ctx is used to allocate the hash table, but the hash table is used
 143     * to allocate all of the suballocations.
 144     */
 145    ht = ralloc(mem_ctx, struct hash_table);
 146    if (ht == NULL)
 147       return NULL;
 148
 149    if (!_mesa_hash_table_init(ht, ht, key_hash_function, key_equals_function)) {
 150       ralloc_free(ht);
 151       return NULL;
 152    }
 153
 154    return ht;
 155 }
 156
 157 struct hash_table *
 158 _mesa_hash_table_clone(struct hash_table *src, void *dst_mem_ctx)
 159 {
 160    struct hash_table *ht;
 161
 162    ht = ralloc(dst_mem_ctx, struct hash_table);
 163    if (ht == NULL)
 164       return NULL;
 165
 166    memcpy(ht, src, sizeof(struct hash_table));
 167
 168    ht->table = ralloc_array(ht, struct hash_entry, ht->size);
 169    if (ht->table == NULL) {
 170       ralloc_free(ht);
 171       return NULL;
 172    }
 173
 174    memcpy(ht->table, src->table, ht->size * sizeof(struct hash_entry));
 175
 176    return ht;
 177 }
 178
 179 /**
 180  * Frees the given hash table.
 181  *
 182  * If delete_function is passed, it gets called on each entry present before
 183  * freeing.
 184  */
 185 void
 186 _mesa_hash_table_destroy(struct hash_table *ht,
 187                          void (*delete_function)(struct hash_entry *entry))
 188 {
 189    if (!ht)
 190       return;
 191
 192    if (delete_function) {
 193       hash_table_foreach(ht, entry) {
 194          delete_function(entry);
 195       }
 196    }
 197    ralloc_free(ht);
 198 }
 199
 200 /**
 201  * Deletes all entries of the given hash table without deleting the table
 202  * itself or changing its structure.
 203  *
 204  * If delete_function is passed, it gets called on each entry present.
 205  */
 206 void
 207 _mesa_hash_table_clear(struct hash_table *ht,
 208                        void (*delete_function)(struct hash_entry *entry))
 209 {
 210    struct hash_entry *entry;
 211
 212    for (entry = ht->table; entry != ht->table + ht->size; entry++) {
 213       if (entry->key == NULL)
 214          continue;
 215
 216       if (delete_function != NULL && entry->key != ht->deleted_key)
 217          delete_function(entry);
 218
 219       entry->key = NULL;
 220    }
 221
 222    ht->entries = 0;
 223    ht->deleted_entries = 0;
 224 }
 225
 226 /** Sets the value of the key pointer used for deleted entries in the table.
 227  *
 228  * The assumption is that usually keys are actual pointers, so we use a
 229  * default value of a pointer to an arbitrary piece of storage in the library.
 230  * But in some cases a consumer wants to store some other sort of value in the
 231  * table, like a uint32_t, in which case that pointer may conflict with one of
 232  * their valid keys.  This lets that user select a safe value.
 233  *
 234  * This must be called before any keys are actually deleted from the table.
 235  */
 236 void
 237 _mesa_hash_table_set_deleted_key(struct hash_table *ht, const void *deleted_key)
 238 {
 239    ht->deleted_key = deleted_key;
 240 }
 241
 242 static struct hash_entry *
 243 hash_table_search(struct hash_table *ht, uint32_t hash, const void *key)
 244 {
 245    uint32_t start_hash_address = hash % ht->size;
 246    uint32_t hash_address = start_hash_address;
 247
 248    do {
 249       uint32_t double_hash;
 250
 251       struct hash_entry *entry = ht->table + hash_address;
 252
 253       if (entry_is_free(entry)) {
 254          return NULL;
 255       } else if (entry_is_present(ht, entry) && entry->hash == hash) {
 256          if (ht->key_equals_function(key, entry->key)) {
 257             return entry;
 258          }
 259       }
 260
 261       double_hash = 1 + hash % ht->rehash;
 262
 263       hash_address = (hash_address + double_hash) % ht->size;
 264    } while (hash_address != start_hash_address);
 265
 266    return NULL;
 267 }
 268
 269 /**
 270  * Finds a hash table entry with the given key and hash of that key.
 271  *
 272  * Returns NULL if no entry is found.  Note that the data pointer may be
 273  * modified by the user.
 274  */
 275 struct hash_entry *
 276 _mesa_hash_table_search(struct hash_table *ht, const void *key)
 277 {
 278    assert(ht->key_hash_function);
 279    return hash_table_search(ht, ht->key_hash_function(key), key);
 280 }
 281
 282 struct hash_entry *
 283 _mesa_hash_table_search_pre_hashed(struct hash_table *ht, uint32_t hash,
 284                                   const void *key)
 285 {
 286    assert(ht->key_hash_function == NULL || hash == ht->key_hash_function(key));
 287    return hash_table_search(ht, hash, key);
 288 }
 289
 290 static struct hash_entry *
 291 hash_table_insert(struct hash_table *ht, uint32_t hash,
 292                   const void *key, void *data);
 293
 294 static void
 295 _mesa_hash_table_rehash(struct hash_table *ht, unsigned new_size_index)
 296 {
 297    struct hash_table old_ht;
 298    struct hash_entry *table;
 299
 300    if (new_size_index >= ARRAY_SIZE(hash_sizes))
 301       return;
 302
 303    table = rzalloc_array(ralloc_parent(ht->table), struct hash_entry,
 304                          hash_sizes[new_size_index].size);
 305    if (table == NULL)
 306       return;
 307
 308    old_ht = *ht;
 309
 310    ht->table = table;
 311    ht->size_index = new_size_index;
 312    ht->size = hash_sizes[ht->size_index].size;
 313    ht->rehash = hash_sizes[ht->size_index].rehash;
 314    ht->max_entries = hash_sizes[ht->size_index].max_entries;
 315    ht->entries = 0;
 316    ht->deleted_entries = 0;
 317
 318    hash_table_foreach(&old_ht, entry) {
 319       hash_table_insert(ht, entry->hash, entry->key, entry->data);
 320    }
 321
 322    ralloc_free(old_ht.table);
 323 }
 324
 325 static struct hash_entry *
 326 hash_table_insert(struct hash_table *ht, uint32_t hash,
 327                   const void *key, void *data)
 328 {
 329    uint32_t start_hash_address, hash_address;
 330    struct hash_entry *available_entry = NULL;
 331
 332    assert(key != NULL);
 333
 334    if (ht->entries >= ht->max_entries) {
 335       _mesa_hash_table_rehash(ht, ht->size_index + 1);
 336    } else if (ht->deleted_entries + ht->entries >= ht->max_entries) {
 337       _mesa_hash_table_rehash(ht, ht->size_index);
 338    }
 339
 340    start_hash_address = hash % ht->size;
 341    hash_address = start_hash_address;
 342    do {
 343       struct hash_entry *entry = ht->table + hash_address;
 344       uint32_t double_hash;
 345
 346       if (!entry_is_present(ht, entry)) {
 347          /* Stash the first available entry we find */
 348          if (available_entry == NULL)
 349             available_entry = entry;
 350          if (entry_is_free(entry))
 351             break;
 352       }
 353
 354       /* Implement replacement when another insert happens
 355        * with a matching key.  This is a relatively common
 356        * feature of hash tables, with the alternative
 357        * generally being "insert the new value as well, and
 358        * return it first when the key is searched for".
 359        *
 360        * Note that the hash table doesn't have a delete
 361        * callback.  If freeing of old data pointers is
 362        * required to avoid memory leaks, perform a search
 363        * before inserting.
 364        */
 365       if (!entry_is_deleted(ht, entry) &&
 366           entry->hash == hash &&
 367           ht->key_equals_function(key, entry->key)) {
 368          entry->key = key;
 369          entry->data = data;
 370          return entry;
 371       }
 372
 373
 374       double_hash = 1 + hash % ht->rehash;
 375
 376       hash_address = (hash_address + double_hash) % ht->size;
 377    } while (hash_address != start_hash_address);
 378
 379    if (available_entry) {
 380       if (entry_is_deleted(ht, available_entry))
 381          ht->deleted_entries--;
 382       available_entry->hash = hash;
 383       available_entry->key = key;
 384       available_entry->data = data;
 385       ht->entries++;
 386       return available_entry;
 387    }
 388
 389    /* We could hit here if a required resize failed. An unchecked-malloc
 390     * application could ignore this result.
 391     */
 392    return NULL;
 393 }
 394
 395 /**
 396  * Inserts the key with the given hash into the table.
 397  *
 398  * Note that insertion may rearrange the table on a resize or rehash,
 399  * so previously found hash_entries are no longer valid after this function.
 400  */
 401 struct hash_entry *
 402 _mesa_hash_table_insert(struct hash_table *ht, const void *key, void *data)
 403 {
 404    assert(ht->key_hash_function);
 405    return hash_table_insert(ht, ht->key_hash_function(key), key, data);
 406 }
 407
 408 struct hash_entry *
 409 _mesa_hash_table_insert_pre_hashed(struct hash_table *ht, uint32_t hash,
 410                                    const void *key, void *data)
 411 {
 412    assert(ht->key_hash_function == NULL || hash == ht->key_hash_function(key));
 413    return hash_table_insert(ht, hash, key, data);
 414 }
 415
 416 /**
 417  * This function deletes the given hash table entry.
 418  *
 419  * Note that deletion doesn't otherwise modify the table, so an iteration over
 420  * the table deleting entries is safe.
 421  */
 422 void
 423 _mesa_hash_table_remove(struct hash_table *ht,
 424                         struct hash_entry *entry)
 425 {
 426    if (!entry)
 427       return;
 428
 429    entry->key = ht->deleted_key;
 430    ht->entries--;
 431    ht->deleted_entries++;
 432 }
 433
 434 /**
 435  * Removes the entry with the corresponding key, if exists.
 436  */
 437 void _mesa_hash_table_remove_key(struct hash_table *ht,
 438                                  const void *key)
 439 {
 440    _mesa_hash_table_remove(ht, _mesa_hash_table_search(ht, key));
 441 }
 442
 443 /**
 444  * This function is an iterator over the hash table.
 445  *
 446  * Pass in NULL for the first entry, as in the start of a for loop.  Note that
 447  * an iteration over the table is O(table_size) not O(entries).
 448  */
 449 struct hash_entry *
 450 _mesa_hash_table_next_entry(struct hash_table *ht,
 451                             struct hash_entry *entry)
 452 {
 453    if (entry == NULL)
 454       entry = ht->table;
 455    else
 456       entry = entry + 1;
 457
 458    for (; entry != ht->table + ht->size; entry++) {
 459       if (entry_is_present(ht, entry)) {
 460          return entry;
 461       }
 462    }
 463
 464    return NULL;
 465 }
 466
 467 /**
 468  * Returns a random entry from the hash table.
 469  *
 470  * This may be useful in implementing random replacement (as opposed
 471  * to just removing everything) in caches based on this hash table
 472  * implementation.  @predicate may be used to filter entries, or may
 473  * be set to NULL for no filtering.
 474  */
 475 struct hash_entry *
 476 _mesa_hash_table_random_entry(struct hash_table *ht,
 477                               bool (*predicate)(struct hash_entry *entry))
 478 {
 479    struct hash_entry *entry;
 480    uint32_t i = rand() % ht->size;
 481
 482    if (ht->entries == 0)
 483       return NULL;
 484
 485    for (entry = ht->table + i; entry != ht->table + ht->size; entry++) {
 486       if (entry_is_present(ht, entry) &&
 487           (!predicate || predicate(entry))) {
 488          return entry;
 489       }
 490    }
 491
 492    for (entry = ht->table; entry != ht->table + i; entry++) {
 493       if (entry_is_present(ht, entry) &&
 494           (!predicate || predicate(entry))) {
 495          return entry;
 496       }
 497    }
 498
 499    return NULL;
 500 }
 501
 502
 503 /**
 504  * Quick FNV-1a hash implementation based on:
 505  * http://www.isthe.com/chongo/tech/comp/fnv/
 506  *
 507  * FNV-1a is not be the best hash out there -- Jenkins's lookup3 is supposed
 508  * to be quite good, and it probably beats FNV.  But FNV has the advantage
 509  * that it involves almost no code.  For an improvement on both, see Paul
 510  * Hsieh's http://www.azillionmonkeys.com/qed/hash.html
 511  */
 512 uint32_t
 513 _mesa_hash_data(const void *data, size_t size)
 514 {
 515    return _mesa_fnv32_1a_accumulate_block(_mesa_fnv32_1a_offset_bias,
 516                                           data, size);
 517 }
 518
 519 /** FNV-1a string hash implementation */
 520 uint32_t
 521 _mesa_hash_string(const void *_key)
 522 {
 523    uint32_t hash = _mesa_fnv32_1a_offset_bias;
 524    const char *key = _key;
 525
 526    while (*key != 0) {
 527       hash = _mesa_fnv32_1a_accumulate(hash, *key);
 528       key++;
 529    }
 530
 531    return hash;
 532 }
 533
 534 /**
 535  * String compare function for use as the comparison callback in
 536  * _mesa_hash_table_create().
 537  */
 538 bool
 539 _mesa_key_string_equal(const void *a, const void *b)
 540 {
 541    return strcmp(a, b) == 0;
 542 }
 543
 544 bool
 545 _mesa_key_pointer_equal(const void *a, const void *b)
 546 {
 547    return a == b;
 548 }
 549
 550 /**
 551  * Hash table wrapper which supports 64-bit keys.
 552  *
 553  * TODO: unify all hash table implementations.
 554  */
 555
 556 struct hash_key_u64 {
 557    uint64_t value;
 558 };
 559
 560 static uint32_t
 561 key_u64_hash(const void *key)
 562 {
 563    return _mesa_hash_data(key, sizeof(struct hash_key_u64));
 564 }
 565
 566 static bool
 567 key_u64_equals(const void *a, const void *b)
 568 {
 569    const struct hash_key_u64 *aa = a;
 570    const struct hash_key_u64 *bb = b;
 571
 572    return aa->value == bb->value;
 573 }
 574
 575 struct hash_table_u64 *
 576 _mesa_hash_table_u64_create(void *mem_ctx)
 577 {
 578    struct hash_table_u64 *ht;
 579
 580    ht = CALLOC_STRUCT(hash_table_u64);
 581    if (!ht)
 582       return NULL;
 583
 584    if (sizeof(void *) == 8) {
 585       ht->table = _mesa_hash_table_create(mem_ctx, _mesa_hash_pointer,
 586                                           _mesa_key_pointer_equal);
 587    } else {
 588       ht->table = _mesa_hash_table_create(mem_ctx, key_u64_hash,
 589                                           key_u64_equals);
 590    }
 591
 592    if (ht->table)
 593       _mesa_hash_table_set_deleted_key(ht->table, uint_key(DELETED_KEY_VALUE));
 594
 595    return ht;
 596 }
 597
 598 void
 599 _mesa_hash_table_u64_destroy(struct hash_table_u64 *ht,
 600                              void (*delete_function)(struct hash_entry *entry))
 601 {
 602    if (!ht)
 603       return;
 604
 605    if (ht->deleted_key_data) {
 606       if (delete_function) {
 607          struct hash_table *table = ht->table;
 608          struct hash_entry deleted_entry;
 609
 610          /* Create a fake entry for the delete function. */
 611          deleted_entry.hash = table->key_hash_function(table->deleted_key);
 612          deleted_entry.key = table->deleted_key;
 613          deleted_entry.data = ht->deleted_key_data;
 614
 615          delete_function(&deleted_entry);
 616       }
 617       ht->deleted_key_data = NULL;
 618    }
 619
 620    _mesa_hash_table_destroy(ht->table, delete_function);
 621    free(ht);
 622 }
 623
 624 void
 625 _mesa_hash_table_u64_insert(struct hash_table_u64 *ht, uint64_t key,
 626                             void *data)
 627 {
 628    if (key == DELETED_KEY_VALUE) {
 629       ht->deleted_key_data = data;
 630       return;
 631    }
 632
 633    if (sizeof(void *) == 8) {
 634       _mesa_hash_table_insert(ht->table, (void *)(uintptr_t)key, data);
 635    } else {
 636       struct hash_key_u64 *_key = CALLOC_STRUCT(hash_key_u64);
 637
 638       if (!_key)
 639          return;
 640       _key->value = key;
 641
 642       _mesa_hash_table_insert(ht->table, _key, data);
 643    }
 644 }
 645
 646 static struct hash_entry *
 647 hash_table_u64_search(struct hash_table_u64 *ht, uint64_t key)
 648 {
 649    if (sizeof(void *) == 8) {
 650       return _mesa_hash_table_search(ht->table, (void *)(uintptr_t)key);
 651    } else {
 652       struct hash_key_u64 _key = { .value = key };
 653       return _mesa_hash_table_search(ht->table, &_key);
 654    }
 655 }
 656
 657 void *
 658 _mesa_hash_table_u64_search(struct hash_table_u64 *ht, uint64_t key)
 659 {
 660    struct hash_entry *entry;
 661
 662    if (key == DELETED_KEY_VALUE)
 663       return ht->deleted_key_data;
 664
 665    entry = hash_table_u64_search(ht, key);
 666    if (!entry)
 667       return NULL;
 668
 669    return entry->data;
 670 }
 671
 672 void
 673 _mesa_hash_table_u64_remove(struct hash_table_u64 *ht, uint64_t key)
 674 {
 675    struct hash_entry *entry;
 676
 677    if (key == DELETED_KEY_VALUE) {
 678       ht->deleted_key_data = NULL;
 679       return;
 680    }
 681
 682    entry = hash_table_u64_search(ht, key);
 683    if (!entry)
 684       return;
 685
 686    if (sizeof(void *) == 8) {
 687       _mesa_hash_table_remove(ht->table, entry);
 688    } else {
 689       struct hash_key *_key = (struct hash_key *)entry->key;
 690
 691       _mesa_hash_table_remove(ht->table, entry);
 692       free(_key);
 693    }
 694 }