src/util/set.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 #include <stdlib.h>
  36 #include <assert.h>
  37 #include <string.h>
  38
  39 #include "macros.h"
  40 #include "ralloc.h"
  41 #include "set.h"
  42
  43 /*
  44  * From Knuth -- a good choice for hash/rehash values is p, p-2 where
  45  * p and p-2 are both prime.  These tables are sized to have an extra 10%
  46  * free to avoid exponential performance degradation as the hash table fills
  47  */
  48
  49 static const uint32_t deleted_key_value;
  50 static const void *deleted_key = &deleted_key_value;
  51
  52 static const struct {
  53    uint32_t max_entries, size, rehash;
  54 } hash_sizes[] = {
  55    { 2,            5,            3            },
  56    { 4,            7,            5            },
  57    { 8,            13,           11           },
  58    { 16,           19,           17           },
  59    { 32,           43,           41           },
  60    { 64,           73,           71           },
  61    { 128,          151,          149          },
  62    { 256,          283,          281          },
  63    { 512,          571,          569          },
  64    { 1024,         1153,         1151         },
  65    { 2048,         2269,         2267         },
  66    { 4096,         4519,         4517         },
  67    { 8192,         9013,         9011         },
  68    { 16384,        18043,        18041        },
  69    { 32768,        36109,        36107        },
  70    { 65536,        72091,        72089        },
  71    { 131072,       144409,       144407       },
  72    { 262144,       288361,       288359       },
  73    { 524288,       576883,       576881       },
  74    { 1048576,      1153459,      1153457      },
  75    { 2097152,      2307163,      2307161      },
  76    { 4194304,      4613893,      4613891      },
  77    { 8388608,      9227641,      9227639      },
  78    { 16777216,     18455029,     18455027     },
  79    { 33554432,     36911011,     36911009     },
  80    { 67108864,     73819861,     73819859     },
  81    { 134217728,    147639589,    147639587    },
  82    { 268435456,    295279081,    295279079    },
  83    { 536870912,    590559793,    590559791    },
  84    { 1073741824,   1181116273,   1181116271   },
  85    { 2147483648ul, 2362232233ul, 2362232231ul }
  86 };
  87
  88 static int
  89 entry_is_free(struct set_entry *entry)
  90 {
  91    return entry->key == NULL;
  92 }
  93
  94 static int
  95 entry_is_deleted(struct set_entry *entry)
  96 {
  97    return entry->key == deleted_key;
  98 }
  99
 100 static int
 101 entry_is_present(struct set_entry *entry)
 102 {
 103    return entry->key != NULL && entry->key != deleted_key;
 104 }
 105
 106 struct set *
 107 _mesa_set_create(void *mem_ctx,
 108                  uint32_t (*key_hash_function)(const void *key),
 109                  bool (*key_equals_function)(const void *a,
 110                                              const void *b))
 111 {
 112    struct set *ht;
 113
 114    ht = ralloc(mem_ctx, struct set);
 115    if (ht == NULL)
 116       return NULL;
 117
 118    ht->size_index = 0;
 119    ht->size = hash_sizes[ht->size_index].size;
 120    ht->rehash = hash_sizes[ht->size_index].rehash;
 121    ht->max_entries = hash_sizes[ht->size_index].max_entries;
 122    ht->key_hash_function = key_hash_function;
 123    ht->key_equals_function = key_equals_function;
 124    ht->table = rzalloc_array(ht, struct set_entry, ht->size);
 125    ht->entries = 0;
 126    ht->deleted_entries = 0;
 127
 128    if (ht->table == NULL) {
 129       ralloc_free(ht);
 130       return NULL;
 131    }
 132
 133    return ht;
 134 }
 135
 136 struct set *
 137 _mesa_set_clone(struct set *set, void *dst_mem_ctx)
 138 {
 139    struct set *clone;
 140
 141    clone = ralloc(dst_mem_ctx, struct set);
 142    if (clone == NULL)
 143       return NULL;
 144
 145    memcpy(clone, set, sizeof(struct set));
 146
 147    clone->table = ralloc_array(clone, struct set_entry, clone->size);
 148    if (clone->table == NULL) {
 149       ralloc_free(clone);
 150       return NULL;
 151    }
 152
 153    memcpy(clone->table, set->table, clone->size * sizeof(struct set_entry));
 154
 155    return clone;
 156 }
 157
 158 /**
 159  * Frees the given set.
 160  *
 161  * If delete_function is passed, it gets called on each entry present before
 162  * freeing.
 163  */
 164 void
 165 _mesa_set_destroy(struct set *ht, void (*delete_function)(struct set_entry *entry))
 166 {
 167    if (!ht)
 168       return;
 169
 170    if (delete_function) {
 171       set_foreach (ht, entry) {
 172          delete_function(entry);
 173       }
 174    }
 175    ralloc_free(ht->table);
 176    ralloc_free(ht);
 177 }
 178
 179 /**
 180  * Clears all values from the given set.
 181  *
 182  * If delete_function is passed, it gets called on each entry present before
 183  * the set is cleared.
 184  */
 185 void
 186 _mesa_set_clear(struct set *set, void (*delete_function)(struct set_entry *entry))
 187 {
 188    if (!set)
 189       return;
 190
 191    set_foreach (set, entry) {
 192       if (delete_function)
 193          delete_function(entry);
 194       entry->key = deleted_key;
 195    }
 196
 197    set->entries = set->deleted_entries = 0;
 198 }
 199
 200 /**
 201  * Finds a set entry with the given key and hash of that key.
 202  *
 203  * Returns NULL if no entry is found.
 204  */
 205 static struct set_entry *
 206 set_search(const struct set *ht, uint32_t hash, const void *key)
 207 {
 208    uint32_t hash_address;
 209
 210    hash_address = hash % ht->size;
 211    do {
 212       uint32_t double_hash;
 213
 214       struct set_entry *entry = ht->table + hash_address;
 215
 216       if (entry_is_free(entry)) {
 217          return NULL;
 218       } else if (entry_is_present(entry) && entry->hash == hash) {
 219          if (ht->key_equals_function(key, entry->key)) {
 220             return entry;
 221          }
 222       }
 223
 224       double_hash = 1 + hash % ht->rehash;
 225
 226       hash_address = (hash_address + double_hash) % ht->size;
 227    } while (hash_address != hash % ht->size);
 228
 229    return NULL;
 230 }
 231
 232 struct set_entry *
 233 _mesa_set_search(const struct set *set, const void *key)
 234 {
 235    assert(set->key_hash_function);
 236    return set_search(set, set->key_hash_function(key), key);
 237 }
 238
 239 struct set_entry *
 240 _mesa_set_search_pre_hashed(const struct set *set, uint32_t hash,
 241                             const void *key)
 242 {
 243    assert(set->key_hash_function == NULL ||
 244           hash == set->key_hash_function(key));
 245    return set_search(set, hash, key);
 246 }
 247
 248 static struct set_entry *
 249 set_add(struct set *ht, uint32_t hash, const void *key);
 250
 251 static void
 252 set_rehash(struct set *ht, unsigned new_size_index)
 253 {
 254    struct set old_ht;
 255    struct set_entry *table;
 256
 257    if (new_size_index >= ARRAY_SIZE(hash_sizes))
 258       return;
 259
 260    table = rzalloc_array(ht, struct set_entry,
 261                          hash_sizes[new_size_index].size);
 262    if (table == NULL)
 263       return;
 264
 265    old_ht = *ht;
 266
 267    ht->table = table;
 268    ht->size_index = new_size_index;
 269    ht->size = hash_sizes[ht->size_index].size;
 270    ht->rehash = hash_sizes[ht->size_index].rehash;
 271    ht->max_entries = hash_sizes[ht->size_index].max_entries;
 272    ht->entries = 0;
 273    ht->deleted_entries = 0;
 274
 275    set_foreach(&old_ht, entry) {
 276       set_add(ht, entry->hash, entry->key);
 277    }
 278
 279    ralloc_free(old_ht.table);
 280 }
 281
 282 /**
 283  * Inserts the key with the given hash into the table.
 284  *
 285  * Note that insertion may rearrange the table on a resize or rehash,
 286  * so previously found hash_entries are no longer valid after this function.
 287  */
 288 static struct set_entry *
 289 set_add(struct set *ht, uint32_t hash, const void *key)
 290 {
 291    uint32_t hash_address;
 292    struct set_entry *available_entry = NULL;
 293
 294    if (ht->entries >= ht->max_entries) {
 295       set_rehash(ht, ht->size_index + 1);
 296    } else if (ht->deleted_entries + ht->entries >= ht->max_entries) {
 297       set_rehash(ht, ht->size_index);
 298    }
 299
 300    hash_address = hash % ht->size;
 301    do {
 302       struct set_entry *entry = ht->table + hash_address;
 303       uint32_t double_hash;
 304
 305       if (!entry_is_present(entry)) {
 306          /* Stash the first available entry we find */
 307          if (available_entry == NULL)
 308             available_entry = entry;
 309          if (entry_is_free(entry))
 310             break;
 311       }
 312
 313       /* Implement replacement when another insert happens
 314        * with a matching key.  This is a relatively common
 315        * feature of hash tables, with the alternative
 316        * generally being "insert the new value as well, and
 317        * return it first when the key is searched for".
 318        *
 319        * Note that the hash table doesn't have a delete callback.
 320        * If freeing of old keys is required to avoid memory leaks,
 321        * perform a search before inserting.
 322        */
 323       if (!entry_is_deleted(entry) &&
 324           entry->hash == hash &&
 325           ht->key_equals_function(key, entry->key)) {
 326          entry->key = key;
 327          return entry;
 328       }
 329
 330       double_hash = 1 + hash % ht->rehash;
 331
 332       hash_address = (hash_address + double_hash) % ht->size;
 333    } while (hash_address != hash % ht->size);
 334
 335    if (available_entry) {
 336       if (entry_is_deleted(available_entry))
 337          ht->deleted_entries--;
 338       available_entry->hash = hash;
 339       available_entry->key = key;
 340       ht->entries++;
 341       return available_entry;
 342    }
 343
 344    /* We could hit here if a required resize failed. An unchecked-malloc
 345     * application could ignore this result.
 346     */
 347    return NULL;
 348 }
 349
 350 struct set_entry *
 351 _mesa_set_add(struct set *set, const void *key)
 352 {
 353    assert(set->key_hash_function);
 354    return set_add(set, set->key_hash_function(key), key);
 355 }
 356
 357 struct set_entry *
 358 _mesa_set_add_pre_hashed(struct set *set, uint32_t hash, const void *key)
 359 {
 360    assert(set->key_hash_function == NULL ||
 361           hash == set->key_hash_function(key));
 362    return set_add(set, hash, key);
 363 }
 364
 365 /**
 366  * This function deletes the given hash table entry.
 367  *
 368  * Note that deletion doesn't otherwise modify the table, so an iteration over
 369  * the table deleting entries is safe.
 370  */
 371 void
 372 _mesa_set_remove(struct set *ht, struct set_entry *entry)
 373 {
 374    if (!entry)
 375       return;
 376
 377    entry->key = deleted_key;
 378    ht->entries--;
 379    ht->deleted_entries++;
 380 }
 381
 382 /**
 383  * Removes the entry with the corresponding key, if exists.
 384  */
 385 void
 386 _mesa_set_remove_key(struct set *set, const void *key)
 387 {
 388    _mesa_set_remove(set, _mesa_set_search(set, key));
 389 }
 390
 391 /**
 392  * This function is an iterator over the hash table.
 393  *
 394  * Pass in NULL for the first entry, as in the start of a for loop.  Note that
 395  * an iteration over the table is O(table_size) not O(entries).
 396  */
 397 struct set_entry *
 398 _mesa_set_next_entry(const struct set *ht, struct set_entry *entry)
 399 {
 400    if (entry == NULL)
 401       entry = ht->table;
 402    else
 403       entry = entry + 1;
 404
 405    for (; entry != ht->table + ht->size; entry++) {
 406       if (entry_is_present(entry)) {
 407          return entry;
 408       }
 409    }
 410
 411    return NULL;
 412 }
 413
 414 struct set_entry *
 415 _mesa_set_random_entry(struct set *ht,
 416                        int (*predicate)(struct set_entry *entry))
 417 {
 418    struct set_entry *entry;
 419    uint32_t i = rand() % ht->size;
 420
 421    if (ht->entries == 0)
 422       return NULL;
 423
 424    for (entry = ht->table + i; entry != ht->table + ht->size; entry++) {
 425       if (entry_is_present(entry) &&
 426           (!predicate || predicate(entry))) {
 427          return entry;
 428       }
 429    }
 430
 431    for (entry = ht->table; entry != ht->table + i; entry++) {
 432       if (entry_is_present(entry) &&
 433           (!predicate || predicate(entry))) {
 434          return entry;
 435       }
 436    }
 437
 438    return NULL;
 439 }