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       hash_table_foreach(ht, entry) {
 181          delete_function(entry);
 182       }
 183    }
 184    ralloc_free(ht);
 185 }
 186
 187 /**
 188  * Deletes all entries of the given hash table without deleting the table
 189  * itself or changing its structure.
 190  *
 191  * If delete_function is passed, it gets called on each entry present.
 192  */
 193 void
 194 _mesa_hash_table_clear(struct hash_table *ht,
 195                        void (*delete_function)(struct hash_entry *entry))
 196 {
 197    struct hash_entry *entry;
 198
 199    for (entry = ht->table; entry != ht->table + ht->size; entry++) {
 200       if (entry->key == NULL)
 201          continue;
 202
 203       if (delete_function != NULL && entry->key != ht->deleted_key)
 204          delete_function(entry);
 205
 206       entry->key = NULL;
 207    }
 208
 209    ht->entries = 0;
 210    ht->deleted_entries = 0;
 211 }
 212
 213 /** Sets the value of the key pointer used for deleted entries in the table.
 214  *
 215  * The assumption is that usually keys are actual pointers, so we use a
 216  * default value of a pointer to an arbitrary piece of storage in the library.
 217  * But in some cases a consumer wants to store some other sort of value in the
 218  * table, like a uint32_t, in which case that pointer may conflict with one of
 219  * their valid keys.  This lets that user select a safe value.
 220  *
 221  * This must be called before any keys are actually deleted from the table.
 222  */
 223 void
 224 _mesa_hash_table_set_deleted_key(struct hash_table *ht, const void *deleted_key)
 225 {
 226    ht->deleted_key = deleted_key;
 227 }
 228
 229 static struct hash_entry *
 230 hash_table_search(struct hash_table *ht, uint32_t hash, const void *key)
 231 {
 232    uint32_t start_hash_address = hash % ht->size;
 233    uint32_t hash_address = start_hash_address;
 234
 235    do {
 236       uint32_t double_hash;
 237
 238       struct hash_entry *entry = ht->table + hash_address;
 239
 240       if (entry_is_free(entry)) {
 241          return NULL;
 242       } else if (entry_is_present(ht, entry) && entry->hash == hash) {
 243          if (ht->key_equals_function(key, entry->key)) {
 244             return entry;
 245          }
 246       }
 247
 248       double_hash = 1 + hash % ht->rehash;
 249
 250       hash_address = (hash_address + double_hash) % ht->size;
 251    } while (hash_address != start_hash_address);
 252
 253    return NULL;
 254 }
 255
 256 /**
 257  * Finds a hash table entry with the given key and hash of that key.
 258  *
 259  * Returns NULL if no entry is found.  Note that the data pointer may be
 260  * modified by the user.
 261  */
 262 struct hash_entry *
 263 _mesa_hash_table_search(struct hash_table *ht, const void *key)
 264 {
 265    assert(ht->key_hash_function);
 266    return hash_table_search(ht, ht->key_hash_function(key), key);
 267 }
 268
 269 struct hash_entry *
 270 _mesa_hash_table_search_pre_hashed(struct hash_table *ht, uint32_t hash,
 271                                   const void *key)
 272 {
 273    assert(ht->key_hash_function == NULL || hash == ht->key_hash_function(key));
 274    return hash_table_search(ht, hash, key);
 275 }
 276
 277 static struct hash_entry *
 278 hash_table_insert(struct hash_table *ht, uint32_t hash,
 279                   const void *key, void *data);
 280
 281 static void
 282 _mesa_hash_table_rehash(struct hash_table *ht, unsigned new_size_index)
 283 {
 284    struct hash_table old_ht;
 285    struct hash_entry *table;
 286
 287    if (new_size_index >= ARRAY_SIZE(hash_sizes))
 288       return;
 289
 290    table = rzalloc_array(ht, struct hash_entry,
 291                          hash_sizes[new_size_index].size);
 292    if (table == NULL)
 293       return;
 294
 295    old_ht = *ht;
 296
 297    ht->table = table;
 298    ht->size_index = new_size_index;
 299    ht->size = hash_sizes[ht->size_index].size;
 300    ht->rehash = hash_sizes[ht->size_index].rehash;
 301    ht->max_entries = hash_sizes[ht->size_index].max_entries;
 302    ht->entries = 0;
 303    ht->deleted_entries = 0;
 304
 305    hash_table_foreach(&old_ht, entry) {
 306       hash_table_insert(ht, entry->hash, entry->key, entry->data);
 307    }
 308
 309    ralloc_free(old_ht.table);
 310 }
 311
 312 static struct hash_entry *
 313 hash_table_insert(struct hash_table *ht, uint32_t hash,
 314                   const void *key, void *data)
 315 {
 316    uint32_t start_hash_address, hash_address;
 317    struct hash_entry *available_entry = NULL;
 318
 319    assert(key != NULL);
 320
 321    if (ht->entries >= ht->max_entries) {
 322       _mesa_hash_table_rehash(ht, ht->size_index + 1);
 323    } else if (ht->deleted_entries + ht->entries >= ht->max_entries) {
 324       _mesa_hash_table_rehash(ht, ht->size_index);
 325    }
 326
 327    start_hash_address = hash % ht->size;
 328    hash_address = start_hash_address;
 329    do {
 330       struct hash_entry *entry = ht->table + hash_address;
 331       uint32_t double_hash;
 332
 333       if (!entry_is_present(ht, entry)) {
 334          /* Stash the first available entry we find */
 335          if (available_entry == NULL)
 336             available_entry = entry;
 337          if (entry_is_free(entry))
 338             break;
 339       }
 340
 341       /* Implement replacement when another insert happens
 342        * with a matching key.  This is a relatively common
 343        * feature of hash tables, with the alternative
 344        * generally being "insert the new value as well, and
 345        * return it first when the key is searched for".
 346        *
 347        * Note that the hash table doesn't have a delete
 348        * callback.  If freeing of old data pointers is
 349        * required to avoid memory leaks, perform a search
 350        * before inserting.
 351        */
 352       if (!entry_is_deleted(ht, entry) &&
 353           entry->hash == hash &&
 354           ht->key_equals_function(key, entry->key)) {
 355          entry->key = key;
 356          entry->data = data;
 357          return entry;
 358       }
 359
 360
 361       double_hash = 1 + hash % ht->rehash;
 362
 363       hash_address = (hash_address + double_hash) % ht->size;
 364    } while (hash_address != start_hash_address);
 365
 366    if (available_entry) {
 367       if (entry_is_deleted(ht, available_entry))
 368          ht->deleted_entries--;
 369       available_entry->hash = hash;
 370       available_entry->key = key;
 371       available_entry->data = data;
 372       ht->entries++;
 373       return available_entry;
 374    }
 375
 376    /* We could hit here if a required resize failed. An unchecked-malloc
 377     * application could ignore this result.
 378     */
 379    return NULL;
 380 }
 381
 382 /**
 383  * Inserts the key with the given hash into the table.
 384  *
 385  * Note that insertion may rearrange the table on a resize or rehash,
 386  * so previously found hash_entries are no longer valid after this function.
 387  */
 388 struct hash_entry *
 389 _mesa_hash_table_insert(struct hash_table *ht, const void *key, void *data)
 390 {
 391    assert(ht->key_hash_function);
 392    return hash_table_insert(ht, ht->key_hash_function(key), key, data);
 393 }
 394
 395 struct hash_entry *
 396 _mesa_hash_table_insert_pre_hashed(struct hash_table *ht, uint32_t hash,
 397                                    const void *key, void *data)
 398 {
 399    assert(ht->key_hash_function == NULL || hash == ht->key_hash_function(key));
 400    return hash_table_insert(ht, hash, key, data);
 401 }
 402
 403 /**
 404  * This function deletes the given hash table entry.
 405  *
 406  * Note that deletion doesn't otherwise modify the table, so an iteration over
 407  * the table deleting entries is safe.
 408  */
 409 void
 410 _mesa_hash_table_remove(struct hash_table *ht,
 411                         struct hash_entry *entry)
 412 {
 413    if (!entry)
 414       return;
 415
 416    entry->key = ht->deleted_key;
 417    ht->entries--;
 418    ht->deleted_entries++;
 419 }
 420
 421 /**
 422  * Removes the entry with the corresponding key, if exists.
 423  */
 424 void _mesa_hash_table_remove_key(struct hash_table *ht,
 425                                  const void *key)
 426 {
 427    _mesa_hash_table_remove(ht, _mesa_hash_table_search(ht, key));
 428 }
 429
 430 /**
 431  * This function is an iterator over the hash table.
 432  *
 433  * Pass in NULL for the first entry, as in the start of a for loop.  Note that
 434  * an iteration over the table is O(table_size) not O(entries).
 435  */
 436 struct hash_entry *
 437 _mesa_hash_table_next_entry(struct hash_table *ht,
 438                             struct hash_entry *entry)
 439 {
 440    if (entry == NULL)
 441       entry = ht->table;
 442    else
 443       entry = entry + 1;
 444
 445    for (; entry != ht->table + ht->size; entry++) {
 446       if (entry_is_present(ht, entry)) {
 447          return entry;
 448       }
 449    }
 450
 451    return NULL;
 452 }
 453
 454 /**
 455  * Returns a random entry from the hash table.
 456  *
 457  * This may be useful in implementing random replacement (as opposed
 458  * to just removing everything) in caches based on this hash table
 459  * implementation.  @predicate may be used to filter entries, or may
 460  * be set to NULL for no filtering.
 461  */
 462 struct hash_entry *
 463 _mesa_hash_table_random_entry(struct hash_table *ht,
 464                               bool (*predicate)(struct hash_entry *entry))
 465 {
 466    struct hash_entry *entry;
 467    uint32_t i = rand() % ht->size;
 468
 469    if (ht->entries == 0)
 470       return NULL;
 471
 472    for (entry = ht->table + i; entry != ht->table + ht->size; entry++) {
 473       if (entry_is_present(ht, entry) &&
 474           (!predicate || predicate(entry))) {
 475          return entry;
 476       }
 477    }
 478
 479    for (entry = ht->table; entry != ht->table + i; entry++) {
 480       if (entry_is_present(ht, entry) &&
 481           (!predicate || predicate(entry))) {
 482          return entry;
 483       }
 484    }
 485
 486    return NULL;
 487 }
 488
 489
 490 /**
 491  * Quick FNV-1a hash implementation based on:
 492  * http://www.isthe.com/chongo/tech/comp/fnv/
 493  *
 494  * FNV-1a is not be the best hash out there -- Jenkins's lookup3 is supposed
 495  * to be quite good, and it probably beats FNV.  But FNV has the advantage
 496  * that it involves almost no code.  For an improvement on both, see Paul
 497  * Hsieh's http://www.azillionmonkeys.com/qed/hash.html
 498  */
 499 uint32_t
 500 _mesa_hash_data(const void *data, size_t size)
 501 {
 502    return _mesa_fnv32_1a_accumulate_block(_mesa_fnv32_1a_offset_bias,
 503                                           data, size);
 504 }
 505
 506 /** FNV-1a string hash implementation */
 507 uint32_t
 508 _mesa_hash_string(const void *_key)
 509 {
 510    uint32_t hash = _mesa_fnv32_1a_offset_bias;
 511    const char *key = _key;
 512
 513    while (*key != 0) {
 514       hash = _mesa_fnv32_1a_accumulate(hash, *key);
 515       key++;
 516    }
 517
 518    return hash;
 519 }
 520
 521 /**
 522  * String compare function for use as the comparison callback in
 523  * _mesa_hash_table_create().
 524  */
 525 bool
 526 _mesa_key_string_equal(const void *a, const void *b)
 527 {
 528    return strcmp(a, b) == 0;
 529 }
 530
 531 bool
 532 _mesa_key_pointer_equal(const void *a, const void *b)
 533 {
 534    return a == b;
 535 }
 536
 537 /**
 538  * Hash table wrapper which supports 64-bit keys.
 539  *
 540  * TODO: unify all hash table implementations.
 541  */
 542
 543 struct hash_key_u64 {
 544    uint64_t value;
 545 };
 546
 547 static uint32_t
 548 key_u64_hash(const void *key)
 549 {
 550    return _mesa_hash_data(key, sizeof(struct hash_key_u64));
 551 }
 552
 553 static bool
 554 key_u64_equals(const void *a, const void *b)
 555 {
 556    const struct hash_key_u64 *aa = a;
 557    const struct hash_key_u64 *bb = b;
 558
 559    return aa->value == bb->value;
 560 }
 561
 562 struct hash_table_u64 *
 563 _mesa_hash_table_u64_create(void *mem_ctx)
 564 {
 565    struct hash_table_u64 *ht;
 566
 567    ht = CALLOC_STRUCT(hash_table_u64);
 568    if (!ht)
 569       return NULL;
 570
 571    if (sizeof(void *) == 8) {
 572       ht->table = _mesa_hash_table_create(mem_ctx, _mesa_hash_pointer,
 573                                           _mesa_key_pointer_equal);
 574    } else {
 575       ht->table = _mesa_hash_table_create(mem_ctx, key_u64_hash,
 576                                           key_u64_equals);
 577    }
 578
 579    if (ht->table)
 580       _mesa_hash_table_set_deleted_key(ht->table, uint_key(DELETED_KEY_VALUE));
 581
 582    return ht;
 583 }
 584
 585 void
 586 _mesa_hash_table_u64_destroy(struct hash_table_u64 *ht,
 587                              void (*delete_function)(struct hash_entry *entry))
 588 {
 589    if (!ht)
 590       return;
 591
 592    if (ht->deleted_key_data) {
 593       if (delete_function) {
 594          struct hash_table *table = ht->table;
 595          struct hash_entry deleted_entry;
 596
 597          /* Create a fake entry for the delete function. */
 598          deleted_entry.hash = table->key_hash_function(table->deleted_key);
 599          deleted_entry.key = table->deleted_key;
 600          deleted_entry.data = ht->deleted_key_data;
 601
 602          delete_function(&deleted_entry);
 603       }
 604       ht->deleted_key_data = NULL;
 605    }
 606
 607    _mesa_hash_table_destroy(ht->table, delete_function);
 608    free(ht);
 609 }
 610
 611 void
 612 _mesa_hash_table_u64_insert(struct hash_table_u64 *ht, uint64_t key,
 613                             void *data)
 614 {
 615    if (key == DELETED_KEY_VALUE) {
 616       ht->deleted_key_data = data;
 617       return;
 618    }
 619
 620    if (sizeof(void *) == 8) {
 621       _mesa_hash_table_insert(ht->table, (void *)(uintptr_t)key, data);
 622    } else {
 623       struct hash_key_u64 *_key = CALLOC_STRUCT(hash_key_u64);
 624
 625       if (!_key)
 626          return;
 627       _key->value = key;
 628
 629       _mesa_hash_table_insert(ht->table, _key, data);
 630    }
 631 }
 632
 633 static struct hash_entry *
 634 hash_table_u64_search(struct hash_table_u64 *ht, uint64_t key)
 635 {
 636    if (sizeof(void *) == 8) {
 637       return _mesa_hash_table_search(ht->table, (void *)(uintptr_t)key);
 638    } else {
 639       struct hash_key_u64 _key = { .value = key };
 640       return _mesa_hash_table_search(ht->table, &_key);
 641    }
 642 }
 643
 644 void *
 645 _mesa_hash_table_u64_search(struct hash_table_u64 *ht, uint64_t key)
 646 {
 647    struct hash_entry *entry;
 648
 649    if (key == DELETED_KEY_VALUE)
 650       return ht->deleted_key_data;
 651
 652    entry = hash_table_u64_search(ht, key);
 653    if (!entry)
 654       return NULL;
 655
 656    return entry->data;
 657 }
 658
 659 void
 660 _mesa_hash_table_u64_remove(struct hash_table_u64 *ht, uint64_t key)
 661 {
 662    struct hash_entry *entry;
 663
 664    if (key == DELETED_KEY_VALUE) {
 665       ht->deleted_key_data = NULL;
 666       return;
 667    }
 668
 669    entry = hash_table_u64_search(ht, key);
 670    if (!entry)
 671       return;
 672
 673    if (sizeof(void *) == 8) {
 674       _mesa_hash_table_remove(ht->table, entry);
 675    } else {
 676       struct hash_key *_key = (struct hash_key *)entry->key;
 677
 678       _mesa_hash_table_remove(ht->table, entry);
 679       free(_key);
 680    }
 681 }