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 size = ht->size;
 246    uint32_t start_hash_address = hash % size;
 247    uint32_t hash_address = start_hash_address;
 248    uint32_t double_hash = 1 + hash % ht->rehash;
 249
 250    do {
 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       hash_address += double_hash;
 262       if (hash_address >= size)
 263          hash_address -= 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    struct hash_entry *available_entry = NULL;
 330
 331    assert(key != NULL);
 332
 333    if (ht->entries >= ht->max_entries) {
 334       _mesa_hash_table_rehash(ht, ht->size_index + 1);
 335    } else if (ht->deleted_entries + ht->entries >= ht->max_entries) {
 336       _mesa_hash_table_rehash(ht, ht->size_index);
 337    }
 338
 339    uint32_t size = ht->size;
 340    uint32_t start_hash_address = hash % size;
 341    uint32_t hash_address = start_hash_address;
 342    uint32_t double_hash = 1 + hash % ht->rehash;
 343    do {
 344       struct hash_entry *entry = ht->table + hash_address;
 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       hash_address += double_hash;
 374       if (hash_address >= size)
 375          hash_address -= size;
 376    } while (hash_address != start_hash_address);
 377
 378    if (available_entry) {
 379       if (entry_is_deleted(ht, available_entry))
 380          ht->deleted_entries--;
 381       available_entry->hash = hash;
 382       available_entry->key = key;
 383       available_entry->data = data;
 384       ht->entries++;
 385       return available_entry;
 386    }
 387
 388    /* We could hit here if a required resize failed. An unchecked-malloc
 389     * application could ignore this result.
 390     */
 391    return NULL;
 392 }
 393
 394 /**
 395  * Inserts the key with the given hash into the table.
 396  *
 397  * Note that insertion may rearrange the table on a resize or rehash,
 398  * so previously found hash_entries are no longer valid after this function.
 399  */
 400 struct hash_entry *
 401 _mesa_hash_table_insert(struct hash_table *ht, const void *key, void *data)
 402 {
 403    assert(ht->key_hash_function);
 404    return hash_table_insert(ht, ht->key_hash_function(key), key, data);
 405 }
 406
 407 struct hash_entry *
 408 _mesa_hash_table_insert_pre_hashed(struct hash_table *ht, uint32_t hash,
 409                                    const void *key, void *data)
 410 {
 411    assert(ht->key_hash_function == NULL || hash == ht->key_hash_function(key));
 412    return hash_table_insert(ht, hash, key, data);
 413 }
 414
 415 /**
 416  * This function deletes the given hash table entry.
 417  *
 418  * Note that deletion doesn't otherwise modify the table, so an iteration over
 419  * the table deleting entries is safe.
 420  */
 421 void
 422 _mesa_hash_table_remove(struct hash_table *ht,
 423                         struct hash_entry *entry)
 424 {
 425    if (!entry)
 426       return;
 427
 428    entry->key = ht->deleted_key;
 429    ht->entries--;
 430    ht->deleted_entries++;
 431 }
 432
 433 /**
 434  * Removes the entry with the corresponding key, if exists.
 435  */
 436 void _mesa_hash_table_remove_key(struct hash_table *ht,
 437                                  const void *key)
 438 {
 439    _mesa_hash_table_remove(ht, _mesa_hash_table_search(ht, key));
 440 }
 441
 442 /**
 443  * This function is an iterator over the hash table.
 444  *
 445  * Pass in NULL for the first entry, as in the start of a for loop.  Note that
 446  * an iteration over the table is O(table_size) not O(entries).
 447  */
 448 struct hash_entry *
 449 _mesa_hash_table_next_entry(struct hash_table *ht,
 450                             struct hash_entry *entry)
 451 {
 452    if (entry == NULL)
 453       entry = ht->table;
 454    else
 455       entry = entry + 1;
 456
 457    for (; entry != ht->table + ht->size; entry++) {
 458       if (entry_is_present(ht, entry)) {
 459          return entry;
 460       }
 461    }
 462
 463    return NULL;
 464 }
 465
 466 /**
 467  * Returns a random entry from the hash table.
 468  *
 469  * This may be useful in implementing random replacement (as opposed
 470  * to just removing everything) in caches based on this hash table
 471  * implementation.  @predicate may be used to filter entries, or may
 472  * be set to NULL for no filtering.
 473  */
 474 struct hash_entry *
 475 _mesa_hash_table_random_entry(struct hash_table *ht,
 476                               bool (*predicate)(struct hash_entry *entry))
 477 {
 478    struct hash_entry *entry;
 479    uint32_t i = rand() % ht->size;
 480
 481    if (ht->entries == 0)
 482       return NULL;
 483
 484    for (entry = ht->table + i; entry != ht->table + ht->size; entry++) {
 485       if (entry_is_present(ht, entry) &&
 486           (!predicate || predicate(entry))) {
 487          return entry;
 488       }
 489    }
 490
 491    for (entry = ht->table; entry != ht->table + i; entry++) {
 492       if (entry_is_present(ht, entry) &&
 493           (!predicate || predicate(entry))) {
 494          return entry;
 495       }
 496    }
 497
 498    return NULL;
 499 }
 500
 501
 502 /**
 503  * Quick FNV-1a hash implementation based on:
 504  * http://www.isthe.com/chongo/tech/comp/fnv/
 505  *
 506  * FNV-1a is not be the best hash out there -- Jenkins's lookup3 is supposed
 507  * to be quite good, and it probably beats FNV.  But FNV has the advantage
 508  * that it involves almost no code.  For an improvement on both, see Paul
 509  * Hsieh's http://www.azillionmonkeys.com/qed/hash.html
 510  */
 511 uint32_t
 512 _mesa_hash_data(const void *data, size_t size)
 513 {
 514    return _mesa_fnv32_1a_accumulate_block(_mesa_fnv32_1a_offset_bias,
 515                                           data, size);
 516 }
 517
 518 /** FNV-1a string hash implementation */
 519 uint32_t
 520 _mesa_hash_string(const void *_key)
 521 {
 522    uint32_t hash = _mesa_fnv32_1a_offset_bias;
 523    const char *key = _key;
 524
 525    while (*key != 0) {
 526       hash = _mesa_fnv32_1a_accumulate(hash, *key);
 527       key++;
 528    }
 529
 530    return hash;
 531 }
 532
 533 /**
 534  * String compare function for use as the comparison callback in
 535  * _mesa_hash_table_create().
 536  */
 537 bool
 538 _mesa_key_string_equal(const void *a, const void *b)
 539 {
 540    return strcmp(a, b) == 0;
 541 }
 542
 543 bool
 544 _mesa_key_pointer_equal(const void *a, const void *b)
 545 {
 546    return a == b;
 547 }
 548
 549 /**
 550  * Helper to create a hash table with pointer keys.
 551  */
 552 struct hash_table *
 553 _mesa_pointer_hash_table_create(void *mem_ctx)
 554 {
 555    return _mesa_hash_table_create(mem_ctx, _mesa_hash_pointer,
 556                                   _mesa_key_pointer_equal);
 557 }
 558
 559 /**
 560  * Hash table wrapper which supports 64-bit keys.
 561  *
 562  * TODO: unify all hash table implementations.
 563  */
 564
 565 struct hash_key_u64 {
 566    uint64_t value;
 567 };
 568
 569 static uint32_t
 570 key_u64_hash(const void *key)
 571 {
 572    return _mesa_hash_data(key, sizeof(struct hash_key_u64));
 573 }
 574
 575 static bool
 576 key_u64_equals(const void *a, const void *b)
 577 {
 578    const struct hash_key_u64 *aa = a;
 579    const struct hash_key_u64 *bb = b;
 580
 581    return aa->value == bb->value;
 582 }
 583
 584 struct hash_table_u64 *
 585 _mesa_hash_table_u64_create(void *mem_ctx)
 586 {
 587    struct hash_table_u64 *ht;
 588
 589    ht = CALLOC_STRUCT(hash_table_u64);
 590    if (!ht)
 591       return NULL;
 592
 593    if (sizeof(void *) == 8) {
 594       ht->table = _mesa_hash_table_create(mem_ctx, _mesa_hash_pointer,
 595                                           _mesa_key_pointer_equal);
 596    } else {
 597       ht->table = _mesa_hash_table_create(mem_ctx, key_u64_hash,
 598                                           key_u64_equals);
 599    }
 600
 601    if (ht->table)
 602       _mesa_hash_table_set_deleted_key(ht->table, uint_key(DELETED_KEY_VALUE));
 603
 604    return ht;
 605 }
 606
 607 void
 608 _mesa_hash_table_u64_destroy(struct hash_table_u64 *ht,
 609                              void (*delete_function)(struct hash_entry *entry))
 610 {
 611    if (!ht)
 612       return;
 613
 614    if (ht->deleted_key_data) {
 615       if (delete_function) {
 616          struct hash_table *table = ht->table;
 617          struct hash_entry deleted_entry;
 618
 619          /* Create a fake entry for the delete function. */
 620          deleted_entry.hash = table->key_hash_function(table->deleted_key);
 621          deleted_entry.key = table->deleted_key;
 622          deleted_entry.data = ht->deleted_key_data;
 623
 624          delete_function(&deleted_entry);
 625       }
 626       ht->deleted_key_data = NULL;
 627    }
 628
 629    _mesa_hash_table_destroy(ht->table, delete_function);
 630    free(ht);
 631 }
 632
 633 void
 634 _mesa_hash_table_u64_insert(struct hash_table_u64 *ht, uint64_t key,
 635                             void *data)
 636 {
 637    if (key == DELETED_KEY_VALUE) {
 638       ht->deleted_key_data = data;
 639       return;
 640    }
 641
 642    if (sizeof(void *) == 8) {
 643       _mesa_hash_table_insert(ht->table, (void *)(uintptr_t)key, data);
 644    } else {
 645       struct hash_key_u64 *_key = CALLOC_STRUCT(hash_key_u64);
 646
 647       if (!_key)
 648          return;
 649       _key->value = key;
 650
 651       _mesa_hash_table_insert(ht->table, _key, data);
 652    }
 653 }
 654
 655 static struct hash_entry *
 656 hash_table_u64_search(struct hash_table_u64 *ht, uint64_t key)
 657 {
 658    if (sizeof(void *) == 8) {
 659       return _mesa_hash_table_search(ht->table, (void *)(uintptr_t)key);
 660    } else {
 661       struct hash_key_u64 _key = { .value = key };
 662       return _mesa_hash_table_search(ht->table, &_key);
 663    }
 664 }
 665
 666 void *
 667 _mesa_hash_table_u64_search(struct hash_table_u64 *ht, uint64_t key)
 668 {
 669    struct hash_entry *entry;
 670
 671    if (key == DELETED_KEY_VALUE)
 672       return ht->deleted_key_data;
 673
 674    entry = hash_table_u64_search(ht, key);
 675    if (!entry)
 676       return NULL;
 677
 678    return entry->data;
 679 }
 680
 681 void
 682 _mesa_hash_table_u64_remove(struct hash_table_u64 *ht, uint64_t key)
 683 {
 684    struct hash_entry *entry;
 685
 686    if (key == DELETED_KEY_VALUE) {
 687       ht->deleted_key_data = NULL;
 688       return;
 689    }
 690
 691    entry = hash_table_u64_search(ht, key);
 692    if (!entry)
 693       return;
 694
 695    if (sizeof(void *) == 8) {
 696       _mesa_hash_table_remove(ht->table, entry);
 697    } else {
 698       struct hash_key *_key = (struct hash_key *)entry->key;
 699
 700       _mesa_hash_table_remove(ht->table, entry);
 701       free(_key);
 702    }
 703 }