1 /* $NetBSD: hash.c,v 1.31 2020/09/05 13:55:08 rillig Exp $ */ 2 3 /* 4 * Copyright (c) 1988, 1989, 1990 The Regents of the University of California. 5 * All rights reserved. 6 * 7 * This code is derived from software contributed to Berkeley by 8 * Adam de Boor. 9 * 10 * Redistribution and use in source and binary forms, with or without 11 * modification, are permitted provided that the following conditions 12 * are met: 13 * 1. Redistributions of source code must retain the above copyright 14 * notice, this list of conditions and the following disclaimer. 15 * 2. Redistributions in binary form must reproduce the above copyright 16 * notice, this list of conditions and the following disclaimer in the 17 * documentation and/or other materials provided with the distribution. 18 * 3. Neither the name of the University nor the names of its contributors 19 * may be used to endorse or promote products derived from this software 20 * without specific prior written permission. 21 * 22 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 23 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 24 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 25 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 26 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 27 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 28 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 29 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 30 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 31 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 32 * SUCH DAMAGE. 33 */ 34 35 /* 36 * Copyright (c) 1988, 1989 by Adam de Boor 37 * Copyright (c) 1989 by Berkeley Softworks 38 * All rights reserved. 39 * 40 * This code is derived from software contributed to Berkeley by 41 * Adam de Boor. 42 * 43 * Redistribution and use in source and binary forms, with or without 44 * modification, are permitted provided that the following conditions 45 * are met: 46 * 1. Redistributions of source code must retain the above copyright 47 * notice, this list of conditions and the following disclaimer. 48 * 2. Redistributions in binary form must reproduce the above copyright 49 * notice, this list of conditions and the following disclaimer in the 50 * documentation and/or other materials provided with the distribution. 51 * 3. All advertising materials mentioning features or use of this software 52 * must display the following acknowledgement: 53 * This product includes software developed by the University of 54 * California, Berkeley and its contributors. 55 * 4. Neither the name of the University nor the names of its contributors 56 * may be used to endorse or promote products derived from this software 57 * without specific prior written permission. 58 * 59 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 60 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 61 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 62 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 63 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 64 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 65 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 66 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 67 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 68 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 69 * SUCH DAMAGE. 70 */ 71 72 #ifndef MAKE_NATIVE 73 static char rcsid[] = "$NetBSD: hash.c,v 1.31 2020/09/05 13:55:08 rillig Exp $"; 74 #else 75 #include <sys/cdefs.h> 76 #ifndef lint 77 #if 0 78 static char sccsid[] = "@(#)hash.c 8.1 (Berkeley) 6/6/93"; 79 #else 80 __RCSID("$NetBSD: hash.c,v 1.31 2020/09/05 13:55:08 rillig Exp $"); 81 #endif 82 #endif /* not lint */ 83 #endif 84 85 /* hash.c -- 86 * 87 * This module contains routines to manipulate a hash table. 88 * See hash.h for a definition of the structure of the hash 89 * table. Hash tables grow automatically as the amount of 90 * information increases. 91 */ 92 #include "make.h" 93 94 /* 95 * Forward references to local procedures that are used before they're 96 * defined: 97 */ 98 99 static void RebuildTable(Hash_Table *); 100 101 /* 102 * The following defines the ratio of # entries to # buckets 103 * at which we rebuild the table to make it larger. 104 */ 105 106 #define rebuildLimit 3 107 108 /* The hash function(s) */ 109 110 #ifndef HASH 111 /* The default: this one matches Gosling's emacs */ 112 #define HASH(h, key, p) do { \ 113 for (h = 0, p = key; *p;) \ 114 h = (h << 5) - h + (unsigned char)*p++; \ 115 } while (0) 116 117 #endif 118 119 /* Sets up the hash table. 120 * 121 * Input: 122 * t Structure to to hold the table. 123 */ 124 void 125 Hash_InitTable(Hash_Table *t) 126 { 127 size_t n = 16, i; 128 struct Hash_Entry **hp; 129 130 t->numEntries = 0; 131 t->maxchain = 0; 132 t->bucketsSize = n; 133 t->bucketsMask = n - 1; 134 t->buckets = hp = bmake_malloc(sizeof(*hp) * n); 135 for (i = 0; i < n; i++) 136 hp[i] = NULL; 137 } 138 139 /* Removes everything from the hash table and frees up the memory space it 140 * occupied (except for the space in the Hash_Table structure). */ 141 void 142 Hash_DeleteTable(Hash_Table *t) 143 { 144 struct Hash_Entry **hp, *h, *nexth = NULL; 145 int i; 146 147 for (hp = t->buckets, i = t->bucketsSize; --i >= 0;) { 148 for (h = *hp++; h != NULL; h = nexth) { 149 nexth = h->next; 150 free(h); 151 } 152 } 153 free(t->buckets); 154 155 /* 156 * Set up the hash table to cause memory faults on any future access 157 * attempts until re-initialization. 158 */ 159 t->buckets = NULL; 160 } 161 162 /* Searches the hash table for an entry corresponding to the key. 163 * 164 * Input: 165 * t Hash table to search. 166 * key A hash key. 167 * 168 * Results: 169 * Returns a pointer to the entry for key, or NULL if the table contains 170 * no entry for the key. 171 */ 172 Hash_Entry * 173 Hash_FindEntry(Hash_Table *t, const char *key) 174 { 175 Hash_Entry *e; 176 unsigned h; 177 const char *p; 178 int chainlen; 179 180 if (t == NULL || t->buckets == NULL) { 181 return NULL; 182 } 183 HASH(h, key, p); 184 p = key; 185 chainlen = 0; 186 #ifdef DEBUG_HASH_LOOKUP 187 if (DEBUG(HASH)) 188 fprintf(debug_file, "%s: %p h=%x key=%s\n", __func__, 189 t, h, key); 190 #endif 191 for (e = t->buckets[h & t->bucketsMask]; e != NULL; e = e->next) { 192 chainlen++; 193 if (e->namehash == h && strcmp(e->name, p) == 0) 194 break; 195 } 196 if (chainlen > t->maxchain) 197 t->maxchain = chainlen; 198 return e; 199 } 200 201 /* Searches the hash table for an entry corresponding to the key. 202 * If no entry is found, then one is created. 203 * 204 * Input: 205 * t Hash table to search. 206 * key A hash key. 207 * newPtr Filled with TRUE if new entry created, 208 * FALSE otherwise. 209 */ 210 Hash_Entry * 211 Hash_CreateEntry(Hash_Table *t, const char *key, Boolean *newPtr) 212 { 213 Hash_Entry *e; 214 unsigned h; 215 const char *p; 216 int keylen; 217 int chainlen; 218 struct Hash_Entry **hp; 219 220 /* 221 * Hash the key. As a side effect, save the length (strlen) of the 222 * key in case we need to create the entry. 223 */ 224 HASH(h, key, p); 225 keylen = p - key; 226 p = key; 227 chainlen = 0; 228 #ifdef DEBUG_HASH_LOOKUP 229 if (DEBUG(HASH)) 230 fprintf(debug_file, "%s: %p h=%x key=%s\n", __func__, 231 t, h, key); 232 #endif 233 for (e = t->buckets[h & t->bucketsMask]; e != NULL; e = e->next) { 234 chainlen++; 235 if (e->namehash == h && strcmp(e->name, p) == 0) { 236 if (newPtr != NULL) 237 *newPtr = FALSE; 238 break; 239 } 240 } 241 if (chainlen > t->maxchain) 242 t->maxchain = chainlen; 243 if (e) 244 return e; 245 246 /* 247 * The desired entry isn't there. Before allocating a new entry, 248 * expand the table if necessary (and this changes the resulting 249 * bucket chain). 250 */ 251 if (t->numEntries >= rebuildLimit * t->bucketsSize) 252 RebuildTable(t); 253 e = bmake_malloc(sizeof(*e) + keylen); 254 hp = &t->buckets[h & t->bucketsMask]; 255 e->next = *hp; 256 *hp = e; 257 Hash_SetValue(e, NULL); 258 e->namehash = h; 259 (void)strcpy(e->name, p); 260 t->numEntries++; 261 262 if (newPtr != NULL) 263 *newPtr = TRUE; 264 return e; 265 } 266 267 /* Delete the given hash table entry and free memory associated with it. */ 268 void 269 Hash_DeleteEntry(Hash_Table *t, Hash_Entry *e) 270 { 271 Hash_Entry **hp, *p; 272 273 if (e == NULL) 274 return; 275 for (hp = &t->buckets[e->namehash & t->bucketsMask]; 276 (p = *hp) != NULL; hp = &p->next) { 277 if (p == e) { 278 *hp = p->next; 279 free(p); 280 t->numEntries--; 281 return; 282 } 283 } 284 (void)write(2, "bad call to Hash_DeleteEntry\n", 29); 285 abort(); 286 } 287 288 /* Sets things up for enumerating all entries in the hash table. 289 * 290 * Input: 291 * t Table to be searched. 292 * searchPtr Area in which to keep state about search. 293 * 294 * Results: 295 * The return value is the address of the first entry in 296 * the hash table, or NULL if the table is empty. 297 */ 298 Hash_Entry * 299 Hash_EnumFirst(Hash_Table *t, Hash_Search *searchPtr) 300 { 301 searchPtr->table = t; 302 searchPtr->nextBucket = 0; 303 searchPtr->entry = NULL; 304 return Hash_EnumNext(searchPtr); 305 } 306 307 /* Returns the next entry in the hash table, or NULL if the end of the table 308 * is reached. 309 * 310 * Input: 311 * searchPtr Area used to keep state about search. 312 */ 313 Hash_Entry * 314 Hash_EnumNext(Hash_Search *searchPtr) 315 { 316 Hash_Entry *e; 317 Hash_Table *t = searchPtr->table; 318 319 /* 320 * The entry field points to the most recently returned 321 * entry, or is NULL if we are starting up. If not NULL, we have 322 * to start at the next one in the chain. 323 */ 324 e = searchPtr->entry; 325 if (e != NULL) 326 e = e->next; 327 /* 328 * If the chain ran out, or if we are starting up, we need to 329 * find the next nonempty chain. 330 */ 331 while (e == NULL) { 332 if (searchPtr->nextBucket >= t->bucketsSize) 333 return NULL; 334 e = t->buckets[searchPtr->nextBucket++]; 335 } 336 searchPtr->entry = e; 337 return e; 338 } 339 340 /* Makes a new hash table that is larger than the old one. The entire hash 341 * table is moved, so any bucket numbers from the old table become invalid. */ 342 static void 343 RebuildTable(Hash_Table *t) 344 { 345 Hash_Entry *e, *next = NULL, **hp, **xp; 346 int i, mask; 347 Hash_Entry **oldhp; 348 int oldsize; 349 350 oldhp = t->buckets; 351 oldsize = i = t->bucketsSize; 352 i <<= 1; 353 t->bucketsSize = i; 354 t->bucketsMask = mask = i - 1; 355 t->buckets = hp = bmake_malloc(sizeof(*hp) * i); 356 while (--i >= 0) 357 *hp++ = NULL; 358 for (hp = oldhp, i = oldsize; --i >= 0;) { 359 for (e = *hp++; e != NULL; e = next) { 360 next = e->next; 361 xp = &t->buckets[e->namehash & mask]; 362 e->next = *xp; 363 *xp = e; 364 } 365 } 366 free(oldhp); 367 if (DEBUG(HASH)) 368 fprintf(debug_file, "%s: %p size=%d entries=%d maxchain=%d\n", 369 __func__, t, t->bucketsSize, t->numEntries, t->maxchain); 370 t->maxchain = 0; 371 } 372 373 void 374 Hash_ForEach(Hash_Table *t, void (*action)(void *, void *), void *data) 375 { 376 Hash_Search search; 377 Hash_Entry *e; 378 379 for (e = Hash_EnumFirst(t, &search); 380 e != NULL; 381 e = Hash_EnumNext(&search)) 382 action(Hash_GetValue(e), data); 383 } 384 385 void 386 Hash_DebugStats(Hash_Table *t, const char *name) 387 { 388 if (DEBUG(HASH)) 389 fprintf(debug_file, "Hash_Table %s: size=%d numEntries=%d maxchain=%d\n", 390 name, t->bucketsSize, t->numEntries, t->maxchain); 391 } 392