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 struct hash_table *
 114 _mesa_hash_table_create(void *mem_ctx,
 115                         uint32_t (*key_hash_function)(const void *key),
 116                         bool (*key_equals_function)(const void *a,
 117                                                     const void *b))
 118 {
 119    struct hash_table *ht;
 120
 121    ht = ralloc(mem_ctx, struct hash_table);
 122    if (ht == NULL)
 123       return NULL;
 124
 125    ht->size_index = 0;
 126    ht->size = hash_sizes[ht->size_index].size;
 127    ht->rehash = hash_sizes[ht->size_index].rehash;
 128    ht->max_entries = hash_sizes[ht->size_index].max_entries;
 129    ht->key_hash_function = key_hash_function;
 130    ht->key_equals_function = key_equals_function;
 131    ht->table = rzalloc_array(ht, struct hash_entry, ht->size);
 132    ht->entries = 0;
 133    ht->deleted_entries = 0;
 134    ht->deleted_key = &deleted_key_value;
 135
 136    if (ht->table == NULL) {
 137       ralloc_free(ht);
 138       return NULL;
 139    }
 140
 141    return ht;
 142 }
 143
 144 struct hash_table *
 145 _mesa_hash_table_clone(struct hash_table *src, void *dst_mem_ctx)
 146 {
 147    struct hash_table *ht;
 148
 149    ht = ralloc(dst_mem_ctx, struct hash_table);
 150    if (ht == NULL)
 151       return NULL;
 152
 153    memcpy(ht, src, sizeof(struct hash_table));
 154
 155    ht->table = ralloc_array(ht, struct hash_entry, ht->size);
 156    if (ht->table == NULL) {
 157       ralloc_free(ht);
 158       return NULL;
 159    }
 160
 161    memcpy(ht->table, src->table, ht->size * sizeof(struct hash_entry));
 162
 163    return ht;
 164 }
 165
 166 /**
 167  * Frees the given hash table.
 168  *
 169  * If delete_function is passed, it gets called on each entry present before
 170  * freeing.
 171  */
 172 void
 173 _mesa_hash_table_destroy(struct hash_table *ht,
 174                          void (*delete_function)(struct hash_entry *entry))
 175 {
 176    if (!ht)
 177       return;
 178
 179    if (delete_function) {
 180       struct hash_entry *entry;
 181
 182       hash_table_foreach(ht, entry) {
 183          delete_function(entry);
 184       }
 185    }
 186    ralloc_free(ht);
 187 }
 188
 189 /**
 190  * Deletes all entries of the given hash table without deleting the table
 191  * itself or changing its structure.
 192  *
 193  * If delete_function is passed, it gets called on each entry present.
 194  */
 195 void
 196 _mesa_hash_table_clear(struct hash_table *ht,
 197                        void (*delete_function)(struct hash_entry *entry))
 198 {
 199    struct hash_entry *entry;
 200
 201    for (entry = ht->table; entry != ht->table + ht->size; entry++) {
 202       if (entry->key == NULL)
 203          continue;
 204
 205       if (delete_function != NULL && entry->key != ht->deleted_key)
 206          delete_function(entry);
 207
 208       entry->key = NULL;
 209    }
 210
 211    ht->entries = 0;
 212    ht->deleted_entries = 0;
 213 }
 214
 215 /** Sets the value of the key pointer used for deleted entries in the table.
 216  *
 217  * The assumption is that usually keys are actual pointers, so we use a
 218  * default value of a pointer to an arbitrary piece of storage in the library.
 219  * But in some cases a consumer wants to store some other sort of value in the
 220  * table, like a uint32_t, in which case that pointer may conflict with one of
 221  * their valid keys.  This lets that user select a safe value.
 222  *
 223  * This must be called before any keys are actually deleted from the table.
 224  */
 225 void
 226 _mesa_hash_table_set_deleted_key(struct hash_table *ht, const void *deleted_key)
 227 {
 228    ht->deleted_key = deleted_key;
 229 }
 230
 231 static struct hash_entry *
 232 hash_table_search(struct hash_table *ht, uint32_t hash, const void *key)
 233 {
 234    uint32_t start_hash_address = hash % ht->size;
 235    uint32_t hash_address = start_hash_address;
 236
 237    do {
 238       uint32_t double_hash;
 239
 240       struct hash_entry *entry = ht->table + hash_address;
 241
 242       if (entry_is_free(entry)) {
 243          return NULL;
 244       } else if (entry_is_present(ht, entry) && entry->hash == hash) {
 245          if (ht->key_equals_function(key, entry->key)) {
 246             return entry;
 247          }
 248       }
 249
 250       double_hash = 1 + hash % ht->rehash;
 251
 252       hash_address = (hash_address + double_hash) % ht->size;
 253    } while (hash_address != start_hash_address);
 254
 255    return NULL;
 256 }
 257
 258 /**
 259  * Finds a hash table entry with the given key and hash of that key.
 260  *
 261  * Returns NULL if no entry is found.  Note that the data pointer may be
 262  * modified by the user.
 263  */
 264 struct hash_entry *
 265 _mesa_hash_table_search(struct hash_table *ht, const void *key)
 266 {
 267    assert(ht->key_hash_function);
 268    return hash_table_search(ht, ht->key_hash_function(key), key);
 269 }
 270
 271 struct hash_entry *
 272 _mesa_hash_table_search_pre_hashed(struct hash_table *ht, uint32_t hash,
 273                                   const void *key)
 274 {
 275    assert(ht->key_hash_function == NULL || hash == ht->key_hash_function(key));
 276    return hash_table_search(ht, hash, key);
 277 }
 278
 279 static struct hash_entry *
 280 hash_table_insert(struct hash_table *ht, uint32_t hash,
 281                   const void *key, void *data);
 282
 283 static void
 284 _mesa_hash_table_rehash(struct hash_table *ht, unsigned new_size_index)
 285 {
 286    struct hash_table old_ht;
 287    struct hash_entry *table, *entry;
 288
 289    if (new_size_index >= ARRAY_SIZE(hash_sizes))
 290       return;
 291
 292    table = rzalloc_array(ht, struct hash_entry,
 293                          hash_sizes[new_size_index].size);
 294    if (table == NULL)
 295       return;
 296
 297    old_ht = *ht;
 298
 299    ht->table = table;
 300    ht->size_index = new_size_index;
 301    ht->size = hash_sizes[ht->size_index].size;
 302    ht->rehash = hash_sizes[ht->size_index].rehash;
 303    ht->max_entries = hash_sizes[ht->size_index].max_entries;
 304    ht->entries = 0;
 305    ht->deleted_entries = 0;
 306
 307    hash_table_foreach(&old_ht, entry) {
 308       hash_table_insert(ht, entry->hash, entry->key, entry->data);
 309    }
 310
 311    ralloc_free(old_ht.table);
 312 }
 313
 314 static struct hash_entry *
 315 hash_table_insert(struct hash_table *ht, uint32_t hash,
 316                   const void *key, void *data)
 317 {
 318    uint32_t start_hash_address, hash_address;
 319    struct hash_entry *available_entry = NULL;
 320
 321    assert(key != NULL);
 322
 323    if (ht->entries >= ht->max_entries) {
 324       _mesa_hash_table_rehash(ht, ht->size_index + 1);
 325    } else if (ht->deleted_entries + ht->entries >= ht->max_entries) {
 326       _mesa_hash_table_rehash(ht, ht->size_index);
 327    }
 328
 329    start_hash_address = hash % ht->size;
 330    hash_address = start_hash_address;
 331    do {
 332       struct hash_entry *entry = ht->table + hash_address;
 333       uint32_t double_hash;
 334
 335       if (!entry_is_present(ht, entry)) {
 336          /* Stash the first available entry we find */
 337          if (available_entry == NULL)
 338             available_entry = entry;
 339          if (entry_is_free(entry))
 340             break;
 341       }
 342
 343       /* Implement replacement when another insert happens
 344        * with a matching key.  This is a relatively common
 345        * feature of hash tables, with the alternative
 346        * generally being "insert the new value as well, and
 347        * return it first when the key is searched for".
 348        *
 349        * Note that the hash table doesn't have a delete
 350        * callback.  If freeing of old data pointers is
 351        * required to avoid memory leaks, perform a search
 352        * before inserting.
 353        */
 354       if (!entry_is_deleted(ht, entry) &&
 355           entry->hash == hash &&
 356           ht->key_equals_function(key, entry->key)) {
 357          entry->key = key;
 358          entry->data = data;
 359          return entry;
 360       }
 361
 362
 363       double_hash = 1 + hash % ht->rehash;
 364
 365       hash_address = (hash_address + double_hash) % ht->size;
 366    } while (hash_address != start_hash_address);
 367
 368    if (available_entry) {
 369       if (entry_is_deleted(ht, available_entry))
 370          ht->deleted_entries--;
 371       available_entry->hash = hash;
 372       available_entry->key = key;
 373       available_entry->data = data;
 374       ht->entries++;
 375       return available_entry;
 376    }
 377
 378    /* We could hit here if a required resize failed. An unchecked-malloc
 379     * application could ignore this result.
 380     */
 381    return NULL;
 382 }
 383
 384 /**
 385  * Inserts the key with the given hash into the table.
 386  *
 387  * Note that insertion may rearrange the table on a resize or rehash,
 388  * so previously found hash_entries are no longer valid after this function.
 389  */
 390 struct hash_entry *
 391 _mesa_hash_table_insert(struct hash_table *ht, const void *key, void *data)
 392 {
 393    assert(ht->key_hash_function);
 394    return hash_table_insert(ht, ht->key_hash_function(key), key, data);
 395 }
 396
 397 struct hash_entry *
 398 _mesa_hash_table_insert_pre_hashed(struct hash_table *ht, uint32_t hash,
 399                                    const void *key, void *data)
 400 {
 401    assert(ht->key_hash_function == NULL || hash == ht->key_hash_function(key));
 402    return hash_table_insert(ht, hash, key, data);
 403 }
 404
 405 /**
 406  * This function deletes the given hash table entry.
 407  *
 408  * Note that deletion doesn't otherwise modify the table, so an iteration over
 409  * the table deleting entries is safe.
 410  */
 411 void
 412 _mesa_hash_table_remove(struct hash_table *ht,
 413                         struct hash_entry *entry)
 414 {
 415    if (!entry)
 416       return;
 417
 418    entry->key = ht->deleted_key;
 419    ht->entries--;
 420    ht->deleted_entries++;
 421 }
 422
 423 /**
 424  * This function is an iterator over the hash table.
 425  *
 426  * Pass in NULL for the first entry, as in the start of a for loop.  Note that
 427  * an iteration over the table is O(table_size) not O(entries).
 428  */
 429 struct hash_entry *
 430 _mesa_hash_table_next_entry(struct hash_table *ht,
 431                             struct hash_entry *entry)
 432 {
 433    if (entry == NULL)
 434       entry = ht->table;
 435    else
 436       entry = entry + 1;
 437
 438    for (; entry != ht->table + ht->size; entry++) {
 439       if (entry_is_present(ht, entry)) {
 440          return entry;
 441       }
 442    }
 443
 444    return NULL;
 445 }
 446
 447 /**
 448  * Returns a random entry from the hash table.
 449  *
 450  * This may be useful in implementing random replacement (as opposed
 451  * to just removing everything) in caches based on this hash table
 452  * implementation.  @predicate may be used to filter entries, or may
 453  * be set to NULL for no filtering.
 454  */
 455 struct hash_entry *
 456 _mesa_hash_table_random_entry(struct hash_table *ht,
 457                               bool (*predicate)(struct hash_entry *entry))
 458 {
 459    struct hash_entry *entry;
 460    uint32_t i = rand() % ht->size;
 461
 462    if (ht->entries == 0)
 463       return NULL;
 464
 465    for (entry = ht->table + i; entry != ht->table + ht->size; entry++) {
 466       if (entry_is_present(ht, entry) &&
 467           (!predicate || predicate(entry))) {
 468          return entry;
 469       }
 470    }
 471
 472    for (entry = ht->table; entry != ht->table + i; entry++) {
 473       if (entry_is_present(ht, entry) &&
 474           (!predicate || predicate(entry))) {
 475          return entry;
 476       }
 477    }
 478
 479    return NULL;
 480 }
 481
 482
 483 /**
 484  * Quick FNV-1a hash implementation based on:
 485  * http://www.isthe.com/chongo/tech/comp/fnv/
 486  *
 487  * FNV-1a is not be the best hash out there -- Jenkins's lookup3 is supposed
 488  * to be quite good, and it probably beats FNV.  But FNV has the advantage
 489  * that it involves almost no code.  For an improvement on both, see Paul
 490  * Hsieh's http://www.azillionmonkeys.com/qed/hash.html
 491  */
 492 uint32_t
 493 _mesa_hash_data(const void *data, size_t size)
 494 {
 495    return _mesa_fnv32_1a_accumulate_block(_mesa_fnv32_1a_offset_bias,
 496                                           data, size);
 497 }
 498
 499 /** FNV-1a string hash implementation */
 500 uint32_t
 501 _mesa_hash_string(const void *_key)
 502 {
 503    uint32_t hash = _mesa_fnv32_1a_offset_bias;
 504    const char *key = _key;
 505
 506    while (*key != 0) {
 507       hash = _mesa_fnv32_1a_accumulate(hash, *key);
 508       key++;
 509    }
 510
 511    return hash;
 512 }
 513
 514 /**
 515  * String compare function for use as the comparison callback in
 516  * _mesa_hash_table_create().
 517  */
 518 bool
 519 _mesa_key_string_equal(const void *a, const void *b)
 520 {
 521    return strcmp(a, b) == 0;
 522 }
 523
 524 bool
 525 _mesa_key_pointer_equal(const void *a, const void *b)
 526 {
 527    return a == b;
 528 }
 529
 530 /**
 531  * Hash table wrapper which supports 64-bit keys.
 532  *
 533  * TODO: unify all hash table implementations.
 534  */
 535
 536 struct hash_key_u64 {
 537    uint64_t value;
 538 };
 539
 540 static uint32_t
 541 key_u64_hash(const void *key)
 542 {
 543    return _mesa_hash_data(key, sizeof(struct hash_key_u64));
 544 }
 545
 546 static bool
 547 key_u64_equals(const void *a, const void *b)
 548 {
 549    const struct hash_key_u64 *aa = a;
 550    const struct hash_key_u64 *bb = b;
 551
 552    return aa->value == bb->value;
 553 }
 554
 555 struct hash_table_u64 *
 556 _mesa_hash_table_u64_create(void *mem_ctx)
 557 {
 558    struct hash_table_u64 *ht;
 559
 560    ht = CALLOC_STRUCT(hash_table_u64);
 561    if (!ht)
 562       return NULL;
 563
 564    if (sizeof(void *) == 8) {
 565       ht->table = _mesa_hash_table_create(mem_ctx, _mesa_hash_pointer,
 566                                           _mesa_key_pointer_equal);
 567    } else {
 568       ht->table = _mesa_hash_table_create(mem_ctx, key_u64_hash,
 569                                           key_u64_equals);
 570    }
 571
 572    if (ht->table)
 573       _mesa_hash_table_set_deleted_key(ht->table, uint_key(DELETED_KEY_VALUE));
 574
 575    return ht;
 576 }
 577
 578 void
 579 _mesa_hash_table_u64_destroy(struct hash_table_u64 *ht,
 580                              void (*delete_function)(struct hash_entry *entry))
 581 {
 582    if (!ht)
 583       return;
 584
 585    if (ht->deleted_key_data) {
 586       if (delete_function) {
 587          struct hash_table *table = ht->table;
 588          struct hash_entry deleted_entry;
 589
 590          /* Create a fake entry for the delete function. */
 591          deleted_entry.hash = table->key_hash_function(table->deleted_key);
 592          deleted_entry.key = table->deleted_key;
 593          deleted_entry.data = ht->deleted_key_data;
 594
 595          delete_function(&deleted_entry);
 596       }
 597       ht->deleted_key_data = NULL;
 598    }
 599
 600    _mesa_hash_table_destroy(ht->table, delete_function);
 601    free(ht);
 602 }
 603
 604 void
 605 _mesa_hash_table_u64_insert(struct hash_table_u64 *ht, uint64_t key,
 606                             void *data)
 607 {
 608    if (key == DELETED_KEY_VALUE) {
 609       ht->deleted_key_data = data;
 610       return;
 611    }
 612
 613    if (sizeof(void *) == 8) {
 614       _mesa_hash_table_insert(ht->table, (void *)(uintptr_t)key, data);
 615    } else {
 616       struct hash_key_u64 *_key = CALLOC_STRUCT(hash_key_u64);
 617
 618       if (!_key)
 619          return;
 620       _key->value = key;
 621
 622       _mesa_hash_table_insert(ht->table, _key, data);
 623    }
 624 }
 625
 626 static struct hash_entry *
 627 hash_table_u64_search(struct hash_table_u64 *ht, uint64_t key)
 628 {
 629    if (sizeof(void *) == 8) {
 630       return _mesa_hash_table_search(ht->table, (void *)(uintptr_t)key);
 631    } else {
 632       struct hash_key_u64 _key = { .value = key };
 633       return _mesa_hash_table_search(ht->table, &_key);
 634    }
 635 }
 636
 637 void *
 638 _mesa_hash_table_u64_search(struct hash_table_u64 *ht, uint64_t key)
 639 {
 640    struct hash_entry *entry;
 641
 642    if (key == DELETED_KEY_VALUE)
 643       return ht->deleted_key_data;
 644
 645    entry = hash_table_u64_search(ht, key);
 646    if (!entry)
 647       return NULL;
 648
 649    return entry->data;
 650 }
 651
 652 void
 653 _mesa_hash_table_u64_remove(struct hash_table_u64 *ht, uint64_t key)
 654 {
 655    struct hash_entry *entry;
 656
 657    if (key == DELETED_KEY_VALUE) {
 658       ht->deleted_key_data = NULL;
 659       return;
 660    }
 661
 662    entry = hash_table_u64_search(ht, key);
 663    if (!entry)
 664       return;
 665
 666    if (sizeof(void *) == 8) {
 667       _mesa_hash_table_remove(ht->table, entry);
 668    } else {
 669       struct hash_key *_key = (struct hash_key *)entry->key;
 670
 671       _mesa_hash_table_remove(ht->table, entry);
 672       free(_key);
 673    }
 674 }