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       struct set_entry *entry;
 172
 173       set_foreach (ht, entry) {
 174          delete_function(entry);
 175       }
 176    }
 177    ralloc_free(ht->table);
 178    ralloc_free(ht);
 179 }
 180
 181 /**
 182  * Clears all values from the given set.
 183  *
 184  * If delete_function is passed, it gets called on each entry present before
 185  * the set is cleared.
 186  */
 187 void
 188 _mesa_set_clear(struct set *set, void (*delete_function)(struct set_entry *entry))
 189 {
 190    struct set_entry *entry;
 191
 192    if (!set)
 193       return;
 194
 195    set_foreach (set, entry) {
 196       if (delete_function)
 197          delete_function(entry);
 198       entry->key = deleted_key;
 199    }
 200
 201    set->entries = set->deleted_entries = 0;
 202 }
 203
 204 /**
 205  * Finds a set entry with the given key and hash of that key.
 206  *
 207  * Returns NULL if no entry is found.
 208  */
 209 static struct set_entry *
 210 set_search(const struct set *ht, uint32_t hash, const void *key)
 211 {
 212    uint32_t hash_address;
 213
 214    hash_address = hash % ht->size;
 215    do {
 216       uint32_t double_hash;
 217
 218       struct set_entry *entry = ht->table + hash_address;
 219
 220       if (entry_is_free(entry)) {
 221          return NULL;
 222       } else if (entry_is_present(entry) && entry->hash == hash) {
 223          if (ht->key_equals_function(key, entry->key)) {
 224             return entry;
 225          }
 226       }
 227
 228       double_hash = 1 + hash % ht->rehash;
 229
 230       hash_address = (hash_address + double_hash) % ht->size;
 231    } while (hash_address != hash % ht->size);
 232
 233    return NULL;
 234 }
 235
 236 struct set_entry *
 237 _mesa_set_search(const struct set *set, const void *key)
 238 {
 239    assert(set->key_hash_function);
 240    return set_search(set, set->key_hash_function(key), key);
 241 }
 242
 243 struct set_entry *
 244 _mesa_set_search_pre_hashed(const struct set *set, uint32_t hash,
 245                             const void *key)
 246 {
 247    assert(set->key_hash_function == NULL ||
 248           hash == set->key_hash_function(key));
 249    return set_search(set, hash, key);
 250 }
 251
 252 static struct set_entry *
 253 set_add(struct set *ht, uint32_t hash, const void *key);
 254
 255 static void
 256 set_rehash(struct set *ht, unsigned new_size_index)
 257 {
 258    struct set old_ht;
 259    struct set_entry *table, *entry;
 260
 261    if (new_size_index >= ARRAY_SIZE(hash_sizes))
 262       return;
 263
 264    table = rzalloc_array(ht, struct set_entry,
 265                          hash_sizes[new_size_index].size);
 266    if (table == NULL)
 267       return;
 268
 269    old_ht = *ht;
 270
 271    ht->table = table;
 272    ht->size_index = new_size_index;
 273    ht->size = hash_sizes[ht->size_index].size;
 274    ht->rehash = hash_sizes[ht->size_index].rehash;
 275    ht->max_entries = hash_sizes[ht->size_index].max_entries;
 276    ht->entries = 0;
 277    ht->deleted_entries = 0;
 278
 279    set_foreach(&old_ht, entry) {
 280       set_add(ht, entry->hash, entry->key);
 281    }
 282
 283    ralloc_free(old_ht.table);
 284 }
 285
 286 /**
 287  * Inserts the key with the given hash into the table.
 288  *
 289  * Note that insertion may rearrange the table on a resize or rehash,
 290  * so previously found hash_entries are no longer valid after this function.
 291  */
 292 static struct set_entry *
 293 set_add(struct set *ht, uint32_t hash, const void *key)
 294 {
 295    uint32_t hash_address;
 296    struct set_entry *available_entry = NULL;
 297
 298    if (ht->entries >= ht->max_entries) {
 299       set_rehash(ht, ht->size_index + 1);
 300    } else if (ht->deleted_entries + ht->entries >= ht->max_entries) {
 301       set_rehash(ht, ht->size_index);
 302    }
 303
 304    hash_address = hash % ht->size;
 305    do {
 306       struct set_entry *entry = ht->table + hash_address;
 307       uint32_t double_hash;
 308
 309       if (!entry_is_present(entry)) {
 310          /* Stash the first available entry we find */
 311          if (available_entry == NULL)
 312             available_entry = entry;
 313          if (entry_is_free(entry))
 314             break;
 315       }
 316
 317       /* Implement replacement when another insert happens
 318        * with a matching key.  This is a relatively common
 319        * feature of hash tables, with the alternative
 320        * generally being "insert the new value as well, and
 321        * return it first when the key is searched for".
 322        *
 323        * Note that the hash table doesn't have a delete callback.
 324        * If freeing of old keys is required to avoid memory leaks,
 325        * perform a search before inserting.
 326        */
 327       if (!entry_is_deleted(entry) &&
 328           entry->hash == hash &&
 329           ht->key_equals_function(key, entry->key)) {
 330          entry->key = key;
 331          return entry;
 332       }
 333
 334       double_hash = 1 + hash % ht->rehash;
 335
 336       hash_address = (hash_address + double_hash) % ht->size;
 337    } while (hash_address != hash % ht->size);
 338
 339    if (available_entry) {
 340       if (entry_is_deleted(available_entry))
 341          ht->deleted_entries--;
 342       available_entry->hash = hash;
 343       available_entry->key = key;
 344       ht->entries++;
 345       return available_entry;
 346    }
 347
 348    /* We could hit here if a required resize failed. An unchecked-malloc
 349     * application could ignore this result.
 350     */
 351    return NULL;
 352 }
 353
 354 struct set_entry *
 355 _mesa_set_add(struct set *set, const void *key)
 356 {
 357    assert(set->key_hash_function);
 358    return set_add(set, set->key_hash_function(key), key);
 359 }
 360
 361 struct set_entry *
 362 _mesa_set_add_pre_hashed(struct set *set, uint32_t hash, const void *key)
 363 {
 364    assert(set->key_hash_function == NULL ||
 365           hash == set->key_hash_function(key));
 366    return set_add(set, hash, key);
 367 }
 368
 369 /**
 370  * This function deletes the given hash table entry.
 371  *
 372  * Note that deletion doesn't otherwise modify the table, so an iteration over
 373  * the table deleting entries is safe.
 374  */
 375 void
 376 _mesa_set_remove(struct set *ht, struct set_entry *entry)
 377 {
 378    if (!entry)
 379       return;
 380
 381    entry->key = deleted_key;
 382    ht->entries--;
 383    ht->deleted_entries++;
 384 }
 385
 386 /**
 387  * Removes the entry with the corresponding key, if exists.
 388  */
 389 void
 390 _mesa_set_remove_key(struct set *set, const void *key)
 391 {
 392    _mesa_set_remove(set, _mesa_set_search(set, key));
 393 }
 394
 395 /**
 396  * This function is an iterator over the hash table.
 397  *
 398  * Pass in NULL for the first entry, as in the start of a for loop.  Note that
 399  * an iteration over the table is O(table_size) not O(entries).
 400  */
 401 struct set_entry *
 402 _mesa_set_next_entry(const struct set *ht, struct set_entry *entry)
 403 {
 404    if (entry == NULL)
 405       entry = ht->table;
 406    else
 407       entry = entry + 1;
 408
 409    for (; entry != ht->table + ht->size; entry++) {
 410       if (entry_is_present(entry)) {
 411          return entry;
 412       }
 413    }
 414
 415    return NULL;
 416 }
 417
 418 struct set_entry *
 419 _mesa_set_random_entry(struct set *ht,
 420                        int (*predicate)(struct set_entry *entry))
 421 {
 422    struct set_entry *entry;
 423    uint32_t i = rand() % ht->size;
 424
 425    if (ht->entries == 0)
 426       return NULL;
 427
 428    for (entry = ht->table + i; entry != ht->table + ht->size; entry++) {
 429       if (entry_is_present(entry) &&
 430           (!predicate || predicate(entry))) {
 431          return entry;
 432       }
 433    }
 434
 435    for (entry = ht->table; entry != ht->table + i; entry++) {
 436       if (entry_is_present(entry) &&
 437           (!predicate || predicate(entry))) {
 438          return entry;
 439       }
 440    }
 441
 442    return NULL;
 443 }