]>
Commit | Line | Data |
---|---|---|
53e1b683 | 1 | /* SPDX-License-Identifier: LGPL-2.1+ */ |
a7334b09 | 2 | |
60918275 | 3 | #include <errno.h> |
11c3a366 | 4 | #include <stdint.h> |
d4510856 | 5 | #include <stdlib.h> |
11c3a366 | 6 | #include <string.h> |
60918275 | 7 | |
b5efdb8a | 8 | #include "alloc-util.h" |
b4f60743 | 9 | #include "env-util.h" |
556c7bae | 10 | #include "fileio.h" |
b4f60743 | 11 | #include "hashmap.h" |
60918275 | 12 | #include "macro.h" |
b3dcf58e | 13 | #include "mempool.h" |
d4510856 | 14 | #include "process-util.h" |
3df3e884 | 15 | #include "random-util.h" |
d4510856 LP |
16 | #include "set.h" |
17 | #include "siphash24.h" | |
556c7bae | 18 | #include "string-util.h" |
d4510856 LP |
19 | #include "strv.h" |
20 | #include "util.h" | |
60918275 | 21 | |
349cc4a5 | 22 | #if ENABLE_DEBUG_HASHMAP |
3d4db144 | 23 | #include <pthread.h> |
2eec67ac TA |
24 | #include "list.h" |
25 | #endif | |
26 | ||
89439d4f MS |
27 | /* |
28 | * Implementation of hashmaps. | |
29 | * Addressing: open | |
30 | * - uses less RAM compared to closed addressing (chaining), because | |
31 | * our entries are small (especially in Sets, which tend to contain | |
32 | * the majority of entries in systemd). | |
33 | * Collision resolution: Robin Hood | |
34 | * - tends to equalize displacement of entries from their optimal buckets. | |
35 | * Probe sequence: linear | |
36 | * - though theoretically worse than random probing/uniform hashing/double | |
37 | * hashing, it is good for cache locality. | |
38 | * | |
39 | * References: | |
40 | * Celis, P. 1986. Robin Hood Hashing. | |
41 | * Ph.D. Dissertation. University of Waterloo, Waterloo, Ont., Canada, Canada. | |
42 | * https://cs.uwaterloo.ca/research/tr/1986/CS-86-14.pdf | |
43 | * - The results are derived for random probing. Suggests deletion with | |
44 | * tombstones and two mean-centered search methods. None of that works | |
45 | * well for linear probing. | |
46 | * | |
47 | * Janson, S. 2005. Individual displacements for linear probing hashing with different insertion policies. | |
48 | * ACM Trans. Algorithms 1, 2 (October 2005), 177-213. | |
49 | * DOI=10.1145/1103963.1103964 http://doi.acm.org/10.1145/1103963.1103964 | |
50 | * http://www.math.uu.se/~svante/papers/sj157.pdf | |
51 | * - Applies to Robin Hood with linear probing. Contains remarks on | |
52 | * the unsuitability of mean-centered search with linear probing. | |
53 | * | |
54 | * Viola, A. 2005. Exact distribution of individual displacements in linear probing hashing. | |
55 | * ACM Trans. Algorithms 1, 2 (October 2005), 214-242. | |
56 | * DOI=10.1145/1103963.1103965 http://doi.acm.org/10.1145/1103963.1103965 | |
57 | * - Similar to Janson. Note that Viola writes about C_{m,n} (number of probes | |
58 | * in a successful search), and Janson writes about displacement. C = d + 1. | |
59 | * | |
60 | * Goossaert, E. 2013. Robin Hood hashing: backward shift deletion. | |
61 | * http://codecapsule.com/2013/11/17/robin-hood-hashing-backward-shift-deletion/ | |
62 | * - Explanation of backward shift deletion with pictures. | |
63 | * | |
64 | * Khuong, P. 2013. The Other Robin Hood Hashing. | |
65 | * http://www.pvk.ca/Blog/2013/11/26/the-other-robin-hood-hashing/ | |
66 | * - Short summary of random vs. linear probing, and tombstones vs. backward shift. | |
67 | */ | |
68 | ||
69 | /* | |
70 | * XXX Ideas for improvement: | |
71 | * For unordered hashmaps, randomize iteration order, similarly to Perl: | |
72 | * http://blog.booking.com/hardening-perls-hash-function.html | |
73 | */ | |
74 | ||
75 | /* INV_KEEP_FREE = 1 / (1 - max_load_factor) | |
76 | * e.g. 1 / (1 - 0.8) = 5 ... keep one fifth of the buckets free. */ | |
77 | #define INV_KEEP_FREE 5U | |
78 | ||
79 | /* Fields common to entries of all hashmap/set types */ | |
80 | struct hashmap_base_entry { | |
60918275 | 81 | const void *key; |
89439d4f MS |
82 | }; |
83 | ||
84 | /* Entry types for specific hashmap/set types | |
85 | * hashmap_base_entry must be at the beginning of each entry struct. */ | |
86 | ||
87 | struct plain_hashmap_entry { | |
88 | struct hashmap_base_entry b; | |
60918275 | 89 | void *value; |
60918275 LP |
90 | }; |
91 | ||
89439d4f MS |
92 | struct ordered_hashmap_entry { |
93 | struct plain_hashmap_entry p; | |
94 | unsigned iterate_next, iterate_previous; | |
95 | }; | |
60918275 | 96 | |
89439d4f MS |
97 | struct set_entry { |
98 | struct hashmap_base_entry b; | |
99 | }; | |
45fa9e29 | 100 | |
89439d4f MS |
101 | /* In several functions it is advantageous to have the hash table extended |
102 | * virtually by a couple of additional buckets. We reserve special index values | |
103 | * for these "swap" buckets. */ | |
104 | #define _IDX_SWAP_BEGIN (UINT_MAX - 3) | |
105 | #define IDX_PUT (_IDX_SWAP_BEGIN + 0) | |
106 | #define IDX_TMP (_IDX_SWAP_BEGIN + 1) | |
107 | #define _IDX_SWAP_END (_IDX_SWAP_BEGIN + 2) | |
39c2a6f1 | 108 | |
89439d4f MS |
109 | #define IDX_FIRST (UINT_MAX - 1) /* special index for freshly initialized iterators */ |
110 | #define IDX_NIL UINT_MAX /* special index value meaning "none" or "end" */ | |
111 | ||
112 | assert_cc(IDX_FIRST == _IDX_SWAP_END); | |
113 | assert_cc(IDX_FIRST == _IDX_ITERATOR_FIRST); | |
114 | ||
115 | /* Storage space for the "swap" buckets. | |
116 | * All entry types can fit into a ordered_hashmap_entry. */ | |
117 | struct swap_entries { | |
118 | struct ordered_hashmap_entry e[_IDX_SWAP_END - _IDX_SWAP_BEGIN]; | |
60918275 LP |
119 | }; |
120 | ||
89439d4f MS |
121 | /* Distance from Initial Bucket */ |
122 | typedef uint8_t dib_raw_t; | |
3ef11dcf ZJS |
123 | #define DIB_RAW_OVERFLOW ((dib_raw_t)0xfdU) /* indicates DIB value is greater than representable */ |
124 | #define DIB_RAW_REHASH ((dib_raw_t)0xfeU) /* entry yet to be rehashed during in-place resize */ | |
125 | #define DIB_RAW_FREE ((dib_raw_t)0xffU) /* a free bucket */ | |
126 | #define DIB_RAW_INIT ((char)DIB_RAW_FREE) /* a byte to memset a DIB store with when initializing */ | |
89439d4f MS |
127 | |
128 | #define DIB_FREE UINT_MAX | |
129 | ||
349cc4a5 | 130 | #if ENABLE_DEBUG_HASHMAP |
89439d4f MS |
131 | struct hashmap_debug_info { |
132 | LIST_FIELDS(struct hashmap_debug_info, debug_list); | |
133 | unsigned max_entries; /* high watermark of n_entries */ | |
134 | ||
135 | /* who allocated this hashmap */ | |
136 | int line; | |
137 | const char *file; | |
138 | const char *func; | |
139 | ||
140 | /* fields to detect modification while iterating */ | |
141 | unsigned put_count; /* counts puts into the hashmap */ | |
142 | unsigned rem_count; /* counts removals from hashmap */ | |
143 | unsigned last_rem_idx; /* remembers last removal index */ | |
39c2a6f1 LP |
144 | }; |
145 | ||
89439d4f MS |
146 | /* Tracks all existing hashmaps. Get at it from gdb. See sd_dump_hashmaps.py */ |
147 | static LIST_HEAD(struct hashmap_debug_info, hashmap_debug_list); | |
4f1b3061 | 148 | static pthread_mutex_t hashmap_debug_list_mutex = PTHREAD_MUTEX_INITIALIZER; |
39c2a6f1 | 149 | |
89439d4f | 150 | #define HASHMAP_DEBUG_FIELDS struct hashmap_debug_info debug; |
39c2a6f1 | 151 | |
fc86aa0e | 152 | #else /* !ENABLE_DEBUG_HASHMAP */ |
89439d4f | 153 | #define HASHMAP_DEBUG_FIELDS |
fc86aa0e | 154 | #endif /* ENABLE_DEBUG_HASHMAP */ |
39c2a6f1 | 155 | |
89439d4f MS |
156 | enum HashmapType { |
157 | HASHMAP_TYPE_PLAIN, | |
158 | HASHMAP_TYPE_ORDERED, | |
159 | HASHMAP_TYPE_SET, | |
160 | _HASHMAP_TYPE_MAX | |
161 | }; | |
39c2a6f1 | 162 | |
89439d4f | 163 | struct _packed_ indirect_storage { |
1a39bc8c | 164 | void *storage; /* where buckets and DIBs are stored */ |
89439d4f MS |
165 | uint8_t hash_key[HASH_KEY_SIZE]; /* hash key; changes during resize */ |
166 | ||
167 | unsigned n_entries; /* number of stored entries */ | |
168 | unsigned n_buckets; /* number of buckets */ | |
169 | ||
170 | unsigned idx_lowest_entry; /* Index below which all buckets are free. | |
171 | Makes "while(hashmap_steal_first())" loops | |
172 | O(n) instead of O(n^2) for unordered hashmaps. */ | |
173 | uint8_t _pad[3]; /* padding for the whole HashmapBase */ | |
174 | /* The bitfields in HashmapBase complete the alignment of the whole thing. */ | |
175 | }; | |
176 | ||
177 | struct direct_storage { | |
178 | /* This gives us 39 bytes on 64bit, or 35 bytes on 32bit. | |
179 | * That's room for 4 set_entries + 4 DIB bytes + 3 unused bytes on 64bit, | |
180 | * or 7 set_entries + 7 DIB bytes + 0 unused bytes on 32bit. */ | |
1a39bc8c | 181 | uint8_t storage[sizeof(struct indirect_storage)]; |
89439d4f MS |
182 | }; |
183 | ||
184 | #define DIRECT_BUCKETS(entry_t) \ | |
185 | (sizeof(struct direct_storage) / (sizeof(entry_t) + sizeof(dib_raw_t))) | |
186 | ||
187 | /* We should be able to store at least one entry directly. */ | |
188 | assert_cc(DIRECT_BUCKETS(struct ordered_hashmap_entry) >= 1); | |
189 | ||
190 | /* We have 3 bits for n_direct_entries. */ | |
191 | assert_cc(DIRECT_BUCKETS(struct set_entry) < (1 << 3)); | |
192 | ||
193 | /* Hashmaps with directly stored entries all use this shared hash key. | |
194 | * It's no big deal if the key is guessed, because there can be only | |
195 | * a handful of directly stored entries in a hashmap. When a hashmap | |
196 | * outgrows direct storage, it gets its own key for indirect storage. */ | |
197 | static uint8_t shared_hash_key[HASH_KEY_SIZE]; | |
198 | static bool shared_hash_key_initialized; | |
199 | ||
200 | /* Fields that all hashmap/set types must have */ | |
201 | struct HashmapBase { | |
202 | const struct hash_ops *hash_ops; /* hash and compare ops to use */ | |
203 | ||
204 | union _packed_ { | |
205 | struct indirect_storage indirect; /* if has_indirect */ | |
206 | struct direct_storage direct; /* if !has_indirect */ | |
207 | }; | |
208 | ||
209 | enum HashmapType type:2; /* HASHMAP_TYPE_* */ | |
210 | bool has_indirect:1; /* whether indirect storage is used */ | |
211 | unsigned n_direct_entries:3; /* Number of entries in direct storage. | |
212 | * Only valid if !has_indirect. */ | |
213 | bool from_pool:1; /* whether was allocated from mempool */ | |
45ea84d8 VC |
214 | bool dirty:1; /* whether dirtied since last iterated_cache_get() */ |
215 | bool cached:1; /* whether this hashmap is being cached */ | |
89439d4f MS |
216 | HASHMAP_DEBUG_FIELDS /* optional hashmap_debug_info */ |
217 | }; | |
218 | ||
219 | /* Specific hash types | |
220 | * HashmapBase must be at the beginning of each hashmap struct. */ | |
221 | ||
222 | struct Hashmap { | |
223 | struct HashmapBase b; | |
224 | }; | |
225 | ||
226 | struct OrderedHashmap { | |
227 | struct HashmapBase b; | |
228 | unsigned iterate_list_head, iterate_list_tail; | |
229 | }; | |
230 | ||
231 | struct Set { | |
232 | struct HashmapBase b; | |
233 | }; | |
234 | ||
45ea84d8 VC |
235 | typedef struct CacheMem { |
236 | const void **ptr; | |
237 | size_t n_populated, n_allocated; | |
238 | bool active:1; | |
239 | } CacheMem; | |
240 | ||
241 | struct IteratedCache { | |
242 | HashmapBase *hashmap; | |
243 | CacheMem keys, values; | |
244 | }; | |
245 | ||
89439d4f MS |
246 | DEFINE_MEMPOOL(hashmap_pool, Hashmap, 8); |
247 | DEFINE_MEMPOOL(ordered_hashmap_pool, OrderedHashmap, 8); | |
248 | /* No need for a separate Set pool */ | |
249 | assert_cc(sizeof(Hashmap) == sizeof(Set)); | |
250 | ||
251 | struct hashmap_type_info { | |
252 | size_t head_size; | |
253 | size_t entry_size; | |
254 | struct mempool *mempool; | |
255 | unsigned n_direct_buckets; | |
256 | }; | |
257 | ||
258 | static const struct hashmap_type_info hashmap_type_info[_HASHMAP_TYPE_MAX] = { | |
259 | [HASHMAP_TYPE_PLAIN] = { | |
260 | .head_size = sizeof(Hashmap), | |
261 | .entry_size = sizeof(struct plain_hashmap_entry), | |
262 | .mempool = &hashmap_pool, | |
263 | .n_direct_buckets = DIRECT_BUCKETS(struct plain_hashmap_entry), | |
264 | }, | |
265 | [HASHMAP_TYPE_ORDERED] = { | |
266 | .head_size = sizeof(OrderedHashmap), | |
267 | .entry_size = sizeof(struct ordered_hashmap_entry), | |
268 | .mempool = &ordered_hashmap_pool, | |
269 | .n_direct_buckets = DIRECT_BUCKETS(struct ordered_hashmap_entry), | |
270 | }, | |
271 | [HASHMAP_TYPE_SET] = { | |
272 | .head_size = sizeof(Set), | |
273 | .entry_size = sizeof(struct set_entry), | |
274 | .mempool = &hashmap_pool, | |
275 | .n_direct_buckets = DIRECT_BUCKETS(struct set_entry), | |
276 | }, | |
277 | }; | |
39c2a6f1 | 278 | |
d18cb393 | 279 | #if VALGRIND |
556c7bae ZJS |
280 | __attribute__((destructor)) static void cleanup_pools(void) { |
281 | _cleanup_free_ char *t = NULL; | |
282 | int r; | |
283 | ||
284 | /* Be nice to valgrind */ | |
285 | ||
286 | /* The pool is only allocated by the main thread, but the memory can | |
287 | * be passed to other threads. Let's clean up if we are the main thread | |
288 | * and no other threads are live. */ | |
289 | if (!is_main_thread()) | |
290 | return; | |
291 | ||
292 | r = get_proc_field("/proc/self/status", "Threads", WHITESPACE, &t); | |
293 | if (r < 0 || !streq(t, "1")) | |
294 | return; | |
295 | ||
296 | mempool_drop(&hashmap_pool); | |
297 | mempool_drop(&ordered_hashmap_pool); | |
298 | } | |
299 | #endif | |
300 | ||
89439d4f MS |
301 | static unsigned n_buckets(HashmapBase *h) { |
302 | return h->has_indirect ? h->indirect.n_buckets | |
303 | : hashmap_type_info[h->type].n_direct_buckets; | |
304 | } | |
305 | ||
306 | static unsigned n_entries(HashmapBase *h) { | |
307 | return h->has_indirect ? h->indirect.n_entries | |
308 | : h->n_direct_entries; | |
309 | } | |
310 | ||
311 | static void n_entries_inc(HashmapBase *h) { | |
312 | if (h->has_indirect) | |
313 | h->indirect.n_entries++; | |
314 | else | |
315 | h->n_direct_entries++; | |
316 | } | |
317 | ||
318 | static void n_entries_dec(HashmapBase *h) { | |
319 | if (h->has_indirect) | |
320 | h->indirect.n_entries--; | |
321 | else | |
322 | h->n_direct_entries--; | |
323 | } | |
324 | ||
1a39bc8c | 325 | static void *storage_ptr(HashmapBase *h) { |
89439d4f MS |
326 | return h->has_indirect ? h->indirect.storage |
327 | : h->direct.storage; | |
328 | } | |
329 | ||
330 | static uint8_t *hash_key(HashmapBase *h) { | |
331 | return h->has_indirect ? h->indirect.hash_key | |
332 | : shared_hash_key; | |
333 | } | |
334 | ||
335 | static unsigned base_bucket_hash(HashmapBase *h, const void *p) { | |
b826ab58 | 336 | struct siphash state; |
0cb3c286 | 337 | uint64_t hash; |
b826ab58 | 338 | |
0cb3c286 | 339 | siphash24_init(&state, hash_key(h)); |
b826ab58 TG |
340 | |
341 | h->hash_ops->hash(p, &state); | |
342 | ||
933f9cae | 343 | hash = siphash24_finalize(&state); |
0cb3c286 TG |
344 | |
345 | return (unsigned) (hash % n_buckets(h)); | |
9bf3b535 | 346 | } |
89439d4f | 347 | #define bucket_hash(h, p) base_bucket_hash(HASHMAP_BASE(h), p) |
9bf3b535 | 348 | |
84dcca75 VC |
349 | static inline void base_set_dirty(HashmapBase *h) { |
350 | h->dirty = true; | |
351 | } | |
352 | #define hashmap_set_dirty(h) base_set_dirty(HASHMAP_BASE(h)) | |
353 | ||
9bf3b535 LP |
354 | static void get_hash_key(uint8_t hash_key[HASH_KEY_SIZE], bool reuse_is_ok) { |
355 | static uint8_t current[HASH_KEY_SIZE]; | |
356 | static bool current_initialized = false; | |
357 | ||
358 | /* Returns a hash function key to use. In order to keep things | |
359 | * fast we will not generate a new key each time we allocate a | |
360 | * new hash table. Instead, we'll just reuse the most recently | |
361 | * generated one, except if we never generated one or when we | |
362 | * are rehashing an entire hash table because we reached a | |
363 | * fill level */ | |
364 | ||
365 | if (!current_initialized || !reuse_is_ok) { | |
366 | random_bytes(current, sizeof(current)); | |
367 | current_initialized = true; | |
368 | } | |
369 | ||
370 | memcpy(hash_key, current, sizeof(current)); | |
a3b6fafe LP |
371 | } |
372 | ||
89439d4f MS |
373 | static struct hashmap_base_entry *bucket_at(HashmapBase *h, unsigned idx) { |
374 | return (struct hashmap_base_entry*) | |
1a39bc8c | 375 | ((uint8_t*) storage_ptr(h) + idx * hashmap_type_info[h->type].entry_size); |
89439d4f MS |
376 | } |
377 | ||
378 | static struct plain_hashmap_entry *plain_bucket_at(Hashmap *h, unsigned idx) { | |
379 | return (struct plain_hashmap_entry*) bucket_at(HASHMAP_BASE(h), idx); | |
380 | } | |
381 | ||
382 | static struct ordered_hashmap_entry *ordered_bucket_at(OrderedHashmap *h, unsigned idx) { | |
383 | return (struct ordered_hashmap_entry*) bucket_at(HASHMAP_BASE(h), idx); | |
384 | } | |
39c2a6f1 | 385 | |
89439d4f MS |
386 | static struct set_entry *set_bucket_at(Set *h, unsigned idx) { |
387 | return (struct set_entry*) bucket_at(HASHMAP_BASE(h), idx); | |
388 | } | |
39c2a6f1 | 389 | |
89439d4f MS |
390 | static struct ordered_hashmap_entry *bucket_at_swap(struct swap_entries *swap, unsigned idx) { |
391 | return &swap->e[idx - _IDX_SWAP_BEGIN]; | |
392 | } | |
39c2a6f1 | 393 | |
89439d4f MS |
394 | /* Returns a pointer to the bucket at index idx. |
395 | * Understands real indexes and swap indexes, hence "_virtual". */ | |
396 | static struct hashmap_base_entry *bucket_at_virtual(HashmapBase *h, struct swap_entries *swap, | |
397 | unsigned idx) { | |
398 | if (idx < _IDX_SWAP_BEGIN) | |
399 | return bucket_at(h, idx); | |
400 | ||
401 | if (idx < _IDX_SWAP_END) | |
402 | return &bucket_at_swap(swap, idx)->p.b; | |
403 | ||
404 | assert_not_reached("Invalid index"); | |
405 | } | |
406 | ||
407 | static dib_raw_t *dib_raw_ptr(HashmapBase *h) { | |
408 | return (dib_raw_t*) | |
1a39bc8c | 409 | ((uint8_t*) storage_ptr(h) + hashmap_type_info[h->type].entry_size * n_buckets(h)); |
89439d4f MS |
410 | } |
411 | ||
412 | static unsigned bucket_distance(HashmapBase *h, unsigned idx, unsigned from) { | |
413 | return idx >= from ? idx - from | |
414 | : n_buckets(h) + idx - from; | |
415 | } | |
416 | ||
417 | static unsigned bucket_calculate_dib(HashmapBase *h, unsigned idx, dib_raw_t raw_dib) { | |
418 | unsigned initial_bucket; | |
419 | ||
420 | if (raw_dib == DIB_RAW_FREE) | |
421 | return DIB_FREE; | |
422 | ||
423 | if (_likely_(raw_dib < DIB_RAW_OVERFLOW)) | |
424 | return raw_dib; | |
425 | ||
426 | /* | |
427 | * Having an overflow DIB value is very unlikely. The hash function | |
428 | * would have to be bad. For example, in a table of size 2^24 filled | |
429 | * to load factor 0.9 the maximum observed DIB is only about 60. | |
430 | * In theory (assuming I used Maxima correctly), for an infinite size | |
431 | * hash table with load factor 0.8 the probability of a given entry | |
432 | * having DIB > 40 is 1.9e-8. | |
433 | * This returns the correct DIB value by recomputing the hash value in | |
434 | * the unlikely case. XXX Hitting this case could be a hint to rehash. | |
435 | */ | |
436 | initial_bucket = bucket_hash(h, bucket_at(h, idx)->key); | |
437 | return bucket_distance(h, idx, initial_bucket); | |
438 | } | |
439 | ||
440 | static void bucket_set_dib(HashmapBase *h, unsigned idx, unsigned dib) { | |
441 | dib_raw_ptr(h)[idx] = dib != DIB_FREE ? MIN(dib, DIB_RAW_OVERFLOW) : DIB_RAW_FREE; | |
442 | } | |
443 | ||
444 | static unsigned skip_free_buckets(HashmapBase *h, unsigned idx) { | |
445 | dib_raw_t *dibs; | |
446 | ||
447 | dibs = dib_raw_ptr(h); | |
448 | ||
449 | for ( ; idx < n_buckets(h); idx++) | |
450 | if (dibs[idx] != DIB_RAW_FREE) | |
451 | return idx; | |
452 | ||
453 | return IDX_NIL; | |
454 | } | |
455 | ||
456 | static void bucket_mark_free(HashmapBase *h, unsigned idx) { | |
eccaf899 | 457 | memzero(bucket_at(h, idx), hashmap_type_info[h->type].entry_size); |
89439d4f MS |
458 | bucket_set_dib(h, idx, DIB_FREE); |
459 | } | |
460 | ||
461 | static void bucket_move_entry(HashmapBase *h, struct swap_entries *swap, | |
462 | unsigned from, unsigned to) { | |
463 | struct hashmap_base_entry *e_from, *e_to; | |
464 | ||
465 | assert(from != to); | |
39c2a6f1 | 466 | |
89439d4f MS |
467 | e_from = bucket_at_virtual(h, swap, from); |
468 | e_to = bucket_at_virtual(h, swap, to); | |
469 | ||
470 | memcpy(e_to, e_from, hashmap_type_info[h->type].entry_size); | |
471 | ||
472 | if (h->type == HASHMAP_TYPE_ORDERED) { | |
473 | OrderedHashmap *lh = (OrderedHashmap*) h; | |
474 | struct ordered_hashmap_entry *le, *le_to; | |
475 | ||
476 | le_to = (struct ordered_hashmap_entry*) e_to; | |
477 | ||
478 | if (le_to->iterate_next != IDX_NIL) { | |
479 | le = (struct ordered_hashmap_entry*) | |
480 | bucket_at_virtual(h, swap, le_to->iterate_next); | |
481 | le->iterate_previous = to; | |
482 | } | |
483 | ||
484 | if (le_to->iterate_previous != IDX_NIL) { | |
485 | le = (struct ordered_hashmap_entry*) | |
486 | bucket_at_virtual(h, swap, le_to->iterate_previous); | |
487 | le->iterate_next = to; | |
488 | } | |
489 | ||
490 | if (lh->iterate_list_head == from) | |
491 | lh->iterate_list_head = to; | |
492 | if (lh->iterate_list_tail == from) | |
493 | lh->iterate_list_tail = to; | |
39c2a6f1 | 494 | } |
89439d4f | 495 | } |
60918275 | 496 | |
89439d4f MS |
497 | static unsigned next_idx(HashmapBase *h, unsigned idx) { |
498 | return (idx + 1U) % n_buckets(h); | |
499 | } | |
60918275 | 500 | |
89439d4f MS |
501 | static unsigned prev_idx(HashmapBase *h, unsigned idx) { |
502 | return (n_buckets(h) + idx - 1U) % n_buckets(h); | |
503 | } | |
60918275 | 504 | |
89439d4f MS |
505 | static void *entry_value(HashmapBase *h, struct hashmap_base_entry *e) { |
506 | switch (h->type) { | |
45fa9e29 | 507 | |
89439d4f MS |
508 | case HASHMAP_TYPE_PLAIN: |
509 | case HASHMAP_TYPE_ORDERED: | |
510 | return ((struct plain_hashmap_entry*)e)->value; | |
39c2a6f1 | 511 | |
89439d4f MS |
512 | case HASHMAP_TYPE_SET: |
513 | return (void*) e->key; | |
a3b6fafe | 514 | |
89439d4f MS |
515 | default: |
516 | assert_not_reached("Unknown hashmap type"); | |
517 | } | |
60918275 LP |
518 | } |
519 | ||
89439d4f MS |
520 | static void base_remove_entry(HashmapBase *h, unsigned idx) { |
521 | unsigned left, right, prev, dib; | |
522 | dib_raw_t raw_dib, *dibs; | |
45fa9e29 | 523 | |
89439d4f MS |
524 | dibs = dib_raw_ptr(h); |
525 | assert(dibs[idx] != DIB_RAW_FREE); | |
034c6ed7 | 526 | |
349cc4a5 | 527 | #if ENABLE_DEBUG_HASHMAP |
89439d4f MS |
528 | h->debug.rem_count++; |
529 | h->debug.last_rem_idx = idx; | |
530 | #endif | |
034c6ed7 | 531 | |
89439d4f MS |
532 | left = idx; |
533 | /* Find the stop bucket ("right"). It is either free or has DIB == 0. */ | |
534 | for (right = next_idx(h, left); ; right = next_idx(h, right)) { | |
535 | raw_dib = dibs[right]; | |
4c701096 | 536 | if (IN_SET(raw_dib, 0, DIB_RAW_FREE)) |
89439d4f MS |
537 | break; |
538 | ||
539 | /* The buckets are not supposed to be all occupied and with DIB > 0. | |
540 | * That would mean we could make everyone better off by shifting them | |
541 | * backward. This scenario is impossible. */ | |
542 | assert(left != right); | |
543 | } | |
034c6ed7 | 544 | |
89439d4f MS |
545 | if (h->type == HASHMAP_TYPE_ORDERED) { |
546 | OrderedHashmap *lh = (OrderedHashmap*) h; | |
547 | struct ordered_hashmap_entry *le = ordered_bucket_at(lh, idx); | |
548 | ||
549 | if (le->iterate_next != IDX_NIL) | |
550 | ordered_bucket_at(lh, le->iterate_next)->iterate_previous = le->iterate_previous; | |
551 | else | |
552 | lh->iterate_list_tail = le->iterate_previous; | |
553 | ||
554 | if (le->iterate_previous != IDX_NIL) | |
555 | ordered_bucket_at(lh, le->iterate_previous)->iterate_next = le->iterate_next; | |
556 | else | |
557 | lh->iterate_list_head = le->iterate_next; | |
558 | } | |
559 | ||
560 | /* Now shift all buckets in the interval (left, right) one step backwards */ | |
561 | for (prev = left, left = next_idx(h, left); left != right; | |
562 | prev = left, left = next_idx(h, left)) { | |
563 | dib = bucket_calculate_dib(h, left, dibs[left]); | |
564 | assert(dib != 0); | |
565 | bucket_move_entry(h, NULL, left, prev); | |
566 | bucket_set_dib(h, prev, dib - 1); | |
567 | } | |
568 | ||
569 | bucket_mark_free(h, prev); | |
570 | n_entries_dec(h); | |
84dcca75 | 571 | base_set_dirty(h); |
034c6ed7 | 572 | } |
89439d4f MS |
573 | #define remove_entry(h, idx) base_remove_entry(HASHMAP_BASE(h), idx) |
574 | ||
575 | static unsigned hashmap_iterate_in_insertion_order(OrderedHashmap *h, Iterator *i) { | |
576 | struct ordered_hashmap_entry *e; | |
577 | unsigned idx; | |
034c6ed7 | 578 | |
101d8e63 | 579 | assert(h); |
89439d4f MS |
580 | assert(i); |
581 | ||
582 | if (i->idx == IDX_NIL) | |
583 | goto at_end; | |
584 | ||
585 | if (i->idx == IDX_FIRST && h->iterate_list_head == IDX_NIL) | |
586 | goto at_end; | |
587 | ||
588 | if (i->idx == IDX_FIRST) { | |
589 | idx = h->iterate_list_head; | |
590 | e = ordered_bucket_at(h, idx); | |
101d8e63 | 591 | } else { |
89439d4f MS |
592 | idx = i->idx; |
593 | e = ordered_bucket_at(h, idx); | |
594 | /* | |
595 | * We allow removing the current entry while iterating, but removal may cause | |
596 | * a backward shift. The next entry may thus move one bucket to the left. | |
597 | * To detect when it happens, we remember the key pointer of the entry we were | |
598 | * going to iterate next. If it does not match, there was a backward shift. | |
599 | */ | |
600 | if (e->p.b.key != i->next_key) { | |
601 | idx = prev_idx(HASHMAP_BASE(h), idx); | |
602 | e = ordered_bucket_at(h, idx); | |
603 | } | |
604 | assert(e->p.b.key == i->next_key); | |
101d8e63 | 605 | } |
101d8e63 | 606 | |
349cc4a5 | 607 | #if ENABLE_DEBUG_HASHMAP |
89439d4f MS |
608 | i->prev_idx = idx; |
609 | #endif | |
610 | ||
611 | if (e->iterate_next != IDX_NIL) { | |
612 | struct ordered_hashmap_entry *n; | |
613 | i->idx = e->iterate_next; | |
614 | n = ordered_bucket_at(h, i->idx); | |
615 | i->next_key = n->p.b.key; | |
616 | } else | |
617 | i->idx = IDX_NIL; | |
618 | ||
619 | return idx; | |
620 | ||
621 | at_end: | |
622 | i->idx = IDX_NIL; | |
623 | return IDX_NIL; | |
101d8e63 LP |
624 | } |
625 | ||
89439d4f MS |
626 | static unsigned hashmap_iterate_in_internal_order(HashmapBase *h, Iterator *i) { |
627 | unsigned idx; | |
628 | ||
60918275 | 629 | assert(h); |
89439d4f | 630 | assert(i); |
60918275 | 631 | |
89439d4f MS |
632 | if (i->idx == IDX_NIL) |
633 | goto at_end; | |
60918275 | 634 | |
89439d4f MS |
635 | if (i->idx == IDX_FIRST) { |
636 | /* fast forward to the first occupied bucket */ | |
637 | if (h->has_indirect) { | |
638 | i->idx = skip_free_buckets(h, h->indirect.idx_lowest_entry); | |
639 | h->indirect.idx_lowest_entry = i->idx; | |
640 | } else | |
641 | i->idx = skip_free_buckets(h, 0); | |
642 | ||
643 | if (i->idx == IDX_NIL) | |
644 | goto at_end; | |
645 | } else { | |
646 | struct hashmap_base_entry *e; | |
647 | ||
648 | assert(i->idx > 0); | |
60918275 | 649 | |
89439d4f MS |
650 | e = bucket_at(h, i->idx); |
651 | /* | |
652 | * We allow removing the current entry while iterating, but removal may cause | |
653 | * a backward shift. The next entry may thus move one bucket to the left. | |
654 | * To detect when it happens, we remember the key pointer of the entry we were | |
655 | * going to iterate next. If it does not match, there was a backward shift. | |
656 | */ | |
657 | if (e->key != i->next_key) | |
658 | e = bucket_at(h, --i->idx); | |
60918275 | 659 | |
89439d4f MS |
660 | assert(e->key == i->next_key); |
661 | } | |
662 | ||
663 | idx = i->idx; | |
349cc4a5 | 664 | #if ENABLE_DEBUG_HASHMAP |
89439d4f MS |
665 | i->prev_idx = idx; |
666 | #endif | |
667 | ||
668 | i->idx = skip_free_buckets(h, i->idx + 1); | |
669 | if (i->idx != IDX_NIL) | |
670 | i->next_key = bucket_at(h, i->idx)->key; | |
101d8e63 | 671 | else |
89439d4f MS |
672 | i->idx = IDX_NIL; |
673 | ||
674 | return idx; | |
60918275 | 675 | |
89439d4f MS |
676 | at_end: |
677 | i->idx = IDX_NIL; | |
678 | return IDX_NIL; | |
60918275 LP |
679 | } |
680 | ||
89439d4f MS |
681 | static unsigned hashmap_iterate_entry(HashmapBase *h, Iterator *i) { |
682 | if (!h) { | |
683 | i->idx = IDX_NIL; | |
684 | return IDX_NIL; | |
685 | } | |
101d8e63 | 686 | |
349cc4a5 | 687 | #if ENABLE_DEBUG_HASHMAP |
89439d4f MS |
688 | if (i->idx == IDX_FIRST) { |
689 | i->put_count = h->debug.put_count; | |
690 | i->rem_count = h->debug.rem_count; | |
691 | } else { | |
692 | /* While iterating, must not add any new entries */ | |
693 | assert(i->put_count == h->debug.put_count); | |
694 | /* ... or remove entries other than the current one */ | |
695 | assert(i->rem_count == h->debug.rem_count || | |
696 | (i->rem_count == h->debug.rem_count - 1 && | |
697 | i->prev_idx == h->debug.last_rem_idx)); | |
698 | /* Reset our removals counter */ | |
699 | i->rem_count = h->debug.rem_count; | |
700 | } | |
701 | #endif | |
101d8e63 | 702 | |
89439d4f MS |
703 | return h->type == HASHMAP_TYPE_ORDERED ? hashmap_iterate_in_insertion_order((OrderedHashmap*) h, i) |
704 | : hashmap_iterate_in_internal_order(h, i); | |
705 | } | |
39c2a6f1 | 706 | |
8927b1da | 707 | bool internal_hashmap_iterate(HashmapBase *h, Iterator *i, void **value, const void **key) { |
89439d4f MS |
708 | struct hashmap_base_entry *e; |
709 | void *data; | |
710 | unsigned idx; | |
711 | ||
712 | idx = hashmap_iterate_entry(h, i); | |
713 | if (idx == IDX_NIL) { | |
8927b1da DH |
714 | if (value) |
715 | *value = NULL; | |
89439d4f MS |
716 | if (key) |
717 | *key = NULL; | |
718 | ||
8927b1da | 719 | return false; |
89439d4f MS |
720 | } |
721 | ||
722 | e = bucket_at(h, idx); | |
723 | data = entry_value(h, e); | |
8927b1da DH |
724 | if (value) |
725 | *value = data; | |
89439d4f MS |
726 | if (key) |
727 | *key = e->key; | |
728 | ||
8927b1da | 729 | return true; |
101d8e63 LP |
730 | } |
731 | ||
8927b1da DH |
732 | bool set_iterate(Set *s, Iterator *i, void **value) { |
733 | return internal_hashmap_iterate(HASHMAP_BASE(s), i, value, NULL); | |
89439d4f | 734 | } |
60918275 | 735 | |
89439d4f MS |
736 | #define HASHMAP_FOREACH_IDX(idx, h, i) \ |
737 | for ((i) = ITERATOR_FIRST, (idx) = hashmap_iterate_entry((h), &(i)); \ | |
738 | (idx != IDX_NIL); \ | |
739 | (idx) = hashmap_iterate_entry((h), &(i))) | |
740 | ||
45ea84d8 VC |
741 | IteratedCache *internal_hashmap_iterated_cache_new(HashmapBase *h) { |
742 | IteratedCache *cache; | |
743 | ||
744 | assert(h); | |
745 | assert(!h->cached); | |
746 | ||
747 | if (h->cached) | |
748 | return NULL; | |
749 | ||
750 | cache = new0(IteratedCache, 1); | |
751 | if (!cache) | |
752 | return NULL; | |
753 | ||
754 | cache->hashmap = h; | |
755 | h->cached = true; | |
756 | ||
757 | return cache; | |
758 | } | |
759 | ||
89439d4f MS |
760 | static void reset_direct_storage(HashmapBase *h) { |
761 | const struct hashmap_type_info *hi = &hashmap_type_info[h->type]; | |
762 | void *p; | |
763 | ||
764 | assert(!h->has_indirect); | |
765 | ||
766 | p = mempset(h->direct.storage, 0, hi->entry_size * hi->n_direct_buckets); | |
767 | memset(p, DIB_RAW_INIT, sizeof(dib_raw_t) * hi->n_direct_buckets); | |
768 | } | |
769 | ||
b4f60743 LP |
770 | static bool use_pool(void) { |
771 | static int b = -1; | |
772 | ||
773 | if (!is_main_thread()) | |
774 | return false; | |
775 | ||
776 | if (b < 0) | |
777 | b = getenv_bool("SYSTEMD_MEMPOOL") != 0; | |
778 | ||
779 | return b; | |
780 | } | |
781 | ||
3ef11dcf | 782 | static struct HashmapBase *hashmap_base_new(const struct hash_ops *hash_ops, enum HashmapType type HASHMAP_DEBUG_PARAMS) { |
89439d4f MS |
783 | HashmapBase *h; |
784 | const struct hashmap_type_info *hi = &hashmap_type_info[type]; | |
b4f60743 | 785 | bool up; |
89439d4f | 786 | |
b4f60743 | 787 | up = use_pool(); |
67f3c402 | 788 | |
b4f60743 | 789 | h = up ? mempool_alloc0_tile(hi->mempool) : malloc0(hi->head_size); |
60918275 | 790 | if (!h) |
89439d4f MS |
791 | return NULL; |
792 | ||
793 | h->type = type; | |
b4f60743 | 794 | h->from_pool = up; |
89439d4f MS |
795 | h->hash_ops = hash_ops ? hash_ops : &trivial_hash_ops; |
796 | ||
797 | if (type == HASHMAP_TYPE_ORDERED) { | |
798 | OrderedHashmap *lh = (OrderedHashmap*)h; | |
799 | lh->iterate_list_head = lh->iterate_list_tail = IDX_NIL; | |
800 | } | |
801 | ||
802 | reset_direct_storage(h); | |
60918275 | 803 | |
89439d4f MS |
804 | if (!shared_hash_key_initialized) { |
805 | random_bytes(shared_hash_key, sizeof(shared_hash_key)); | |
806 | shared_hash_key_initialized= true; | |
807 | } | |
808 | ||
349cc4a5 | 809 | #if ENABLE_DEBUG_HASHMAP |
89439d4f MS |
810 | h->debug.func = func; |
811 | h->debug.file = file; | |
812 | h->debug.line = line; | |
4f1b3061 TG |
813 | assert_se(pthread_mutex_lock(&hashmap_debug_list_mutex) == 0); |
814 | LIST_PREPEND(debug_list, hashmap_debug_list, &h->debug); | |
815 | assert_se(pthread_mutex_unlock(&hashmap_debug_list_mutex) == 0); | |
89439d4f MS |
816 | #endif |
817 | ||
818 | return h; | |
819 | } | |
60918275 | 820 | |
89439d4f | 821 | Hashmap *internal_hashmap_new(const struct hash_ops *hash_ops HASHMAP_DEBUG_PARAMS) { |
3ef11dcf | 822 | return (Hashmap*) hashmap_base_new(hash_ops, HASHMAP_TYPE_PLAIN HASHMAP_DEBUG_PASS_ARGS); |
89439d4f MS |
823 | } |
824 | ||
825 | OrderedHashmap *internal_ordered_hashmap_new(const struct hash_ops *hash_ops HASHMAP_DEBUG_PARAMS) { | |
3ef11dcf | 826 | return (OrderedHashmap*) hashmap_base_new(hash_ops, HASHMAP_TYPE_ORDERED HASHMAP_DEBUG_PASS_ARGS); |
89439d4f MS |
827 | } |
828 | ||
829 | Set *internal_set_new(const struct hash_ops *hash_ops HASHMAP_DEBUG_PARAMS) { | |
3ef11dcf | 830 | return (Set*) hashmap_base_new(hash_ops, HASHMAP_TYPE_SET HASHMAP_DEBUG_PASS_ARGS); |
89439d4f MS |
831 | } |
832 | ||
833 | static int hashmap_base_ensure_allocated(HashmapBase **h, const struct hash_ops *hash_ops, | |
3ef11dcf | 834 | enum HashmapType type HASHMAP_DEBUG_PARAMS) { |
89439d4f MS |
835 | HashmapBase *q; |
836 | ||
837 | assert(h); | |
838 | ||
839 | if (*h) | |
840 | return 0; | |
841 | ||
3ef11dcf | 842 | q = hashmap_base_new(hash_ops, type HASHMAP_DEBUG_PASS_ARGS); |
89439d4f MS |
843 | if (!q) |
844 | return -ENOMEM; | |
845 | ||
846 | *h = q; | |
847 | return 0; | |
848 | } | |
849 | ||
850 | int internal_hashmap_ensure_allocated(Hashmap **h, const struct hash_ops *hash_ops HASHMAP_DEBUG_PARAMS) { | |
3ef11dcf | 851 | return hashmap_base_ensure_allocated((HashmapBase**)h, hash_ops, HASHMAP_TYPE_PLAIN HASHMAP_DEBUG_PASS_ARGS); |
89439d4f MS |
852 | } |
853 | ||
854 | int internal_ordered_hashmap_ensure_allocated(OrderedHashmap **h, const struct hash_ops *hash_ops HASHMAP_DEBUG_PARAMS) { | |
3ef11dcf | 855 | return hashmap_base_ensure_allocated((HashmapBase**)h, hash_ops, HASHMAP_TYPE_ORDERED HASHMAP_DEBUG_PASS_ARGS); |
89439d4f MS |
856 | } |
857 | ||
858 | int internal_set_ensure_allocated(Set **s, const struct hash_ops *hash_ops HASHMAP_DEBUG_PARAMS) { | |
3ef11dcf | 859 | return hashmap_base_ensure_allocated((HashmapBase**)s, hash_ops, HASHMAP_TYPE_SET HASHMAP_DEBUG_PASS_ARGS); |
89439d4f MS |
860 | } |
861 | ||
862 | static void hashmap_free_no_clear(HashmapBase *h) { | |
863 | assert(!h->has_indirect); | |
864 | assert(!h->n_direct_entries); | |
865 | ||
349cc4a5 | 866 | #if ENABLE_DEBUG_HASHMAP |
4f1b3061 | 867 | assert_se(pthread_mutex_lock(&hashmap_debug_list_mutex) == 0); |
89439d4f | 868 | LIST_REMOVE(debug_list, hashmap_debug_list, &h->debug); |
4f1b3061 | 869 | assert_se(pthread_mutex_unlock(&hashmap_debug_list_mutex) == 0); |
89439d4f | 870 | #endif |
45fa9e29 | 871 | |
205c085b LP |
872 | if (h->from_pool) { |
873 | /* Ensure that the object didn't get migrated between threads. */ | |
874 | assert_se(is_main_thread()); | |
89439d4f | 875 | mempool_free_tile(hashmap_type_info[h->type].mempool, h); |
205c085b | 876 | } else |
39c2a6f1 | 877 | free(h); |
60918275 LP |
878 | } |
879 | ||
cfe561a4 | 880 | HashmapBase *internal_hashmap_free(HashmapBase *h) { |
89439d4f MS |
881 | |
882 | /* Free the hashmap, but nothing in it */ | |
883 | ||
cfe561a4 DH |
884 | if (h) { |
885 | internal_hashmap_clear(h); | |
886 | hashmap_free_no_clear(h); | |
887 | } | |
89439d4f | 888 | |
cfe561a4 | 889 | return NULL; |
89439d4f MS |
890 | } |
891 | ||
cfe561a4 | 892 | HashmapBase *internal_hashmap_free_free(HashmapBase *h) { |
67f3c402 LP |
893 | |
894 | /* Free the hashmap and all data objects in it, but not the | |
895 | * keys */ | |
896 | ||
cfe561a4 DH |
897 | if (h) { |
898 | internal_hashmap_clear_free(h); | |
899 | hashmap_free_no_clear(h); | |
900 | } | |
61b1477c | 901 | |
cfe561a4 | 902 | return NULL; |
449ddb2d LP |
903 | } |
904 | ||
cfe561a4 | 905 | Hashmap *hashmap_free_free_free(Hashmap *h) { |
fabe5c0e LP |
906 | |
907 | /* Free the hashmap and all data and key objects in it */ | |
908 | ||
cfe561a4 DH |
909 | if (h) { |
910 | hashmap_clear_free_free(h); | |
911 | hashmap_free_no_clear(HASHMAP_BASE(h)); | |
912 | } | |
fabe5c0e | 913 | |
cfe561a4 | 914 | return NULL; |
fabe5c0e LP |
915 | } |
916 | ||
89439d4f | 917 | void internal_hashmap_clear(HashmapBase *h) { |
11dd41ce LP |
918 | if (!h) |
919 | return; | |
920 | ||
89439d4f MS |
921 | if (h->has_indirect) { |
922 | free(h->indirect.storage); | |
923 | h->has_indirect = false; | |
924 | } | |
925 | ||
926 | h->n_direct_entries = 0; | |
927 | reset_direct_storage(h); | |
928 | ||
929 | if (h->type == HASHMAP_TYPE_ORDERED) { | |
930 | OrderedHashmap *lh = (OrderedHashmap*) h; | |
931 | lh->iterate_list_head = lh->iterate_list_tail = IDX_NIL; | |
932 | } | |
84dcca75 VC |
933 | |
934 | base_set_dirty(h); | |
11dd41ce LP |
935 | } |
936 | ||
89439d4f MS |
937 | void internal_hashmap_clear_free(HashmapBase *h) { |
938 | unsigned idx; | |
9946996c | 939 | |
61b1477c LP |
940 | if (!h) |
941 | return; | |
9946996c | 942 | |
89439d4f MS |
943 | for (idx = skip_free_buckets(h, 0); idx != IDX_NIL; |
944 | idx = skip_free_buckets(h, idx + 1)) | |
945 | free(entry_value(h, bucket_at(h, idx))); | |
946 | ||
947 | internal_hashmap_clear(h); | |
9946996c LP |
948 | } |
949 | ||
fabe5c0e | 950 | void hashmap_clear_free_free(Hashmap *h) { |
89439d4f MS |
951 | unsigned idx; |
952 | ||
fabe5c0e LP |
953 | if (!h) |
954 | return; | |
955 | ||
89439d4f MS |
956 | for (idx = skip_free_buckets(HASHMAP_BASE(h), 0); idx != IDX_NIL; |
957 | idx = skip_free_buckets(HASHMAP_BASE(h), idx + 1)) { | |
958 | struct plain_hashmap_entry *e = plain_bucket_at(h, idx); | |
959 | free((void*)e->b.key); | |
960 | free(e->value); | |
fabe5c0e | 961 | } |
89439d4f MS |
962 | |
963 | internal_hashmap_clear(HASHMAP_BASE(h)); | |
fabe5c0e LP |
964 | } |
965 | ||
89439d4f MS |
966 | static int resize_buckets(HashmapBase *h, unsigned entries_add); |
967 | ||
968 | /* | |
969 | * Finds an empty bucket to put an entry into, starting the scan at 'idx'. | |
970 | * Performs Robin Hood swaps as it goes. The entry to put must be placed | |
971 | * by the caller into swap slot IDX_PUT. | |
972 | * If used for in-place resizing, may leave a displaced entry in swap slot | |
973 | * IDX_PUT. Caller must rehash it next. | |
974 | * Returns: true if it left a displaced entry to rehash next in IDX_PUT, | |
975 | * false otherwise. | |
976 | */ | |
977 | static bool hashmap_put_robin_hood(HashmapBase *h, unsigned idx, | |
978 | struct swap_entries *swap) { | |
979 | dib_raw_t raw_dib, *dibs; | |
980 | unsigned dib, distance; | |
981 | ||
349cc4a5 | 982 | #if ENABLE_DEBUG_HASHMAP |
89439d4f MS |
983 | h->debug.put_count++; |
984 | #endif | |
985 | ||
986 | dibs = dib_raw_ptr(h); | |
987 | ||
988 | for (distance = 0; ; distance++) { | |
989 | raw_dib = dibs[idx]; | |
3742095b | 990 | if (IN_SET(raw_dib, DIB_RAW_FREE, DIB_RAW_REHASH)) { |
89439d4f MS |
991 | if (raw_dib == DIB_RAW_REHASH) |
992 | bucket_move_entry(h, swap, idx, IDX_TMP); | |
993 | ||
994 | if (h->has_indirect && h->indirect.idx_lowest_entry > idx) | |
995 | h->indirect.idx_lowest_entry = idx; | |
60918275 | 996 | |
89439d4f MS |
997 | bucket_set_dib(h, idx, distance); |
998 | bucket_move_entry(h, swap, IDX_PUT, idx); | |
999 | if (raw_dib == DIB_RAW_REHASH) { | |
1000 | bucket_move_entry(h, swap, IDX_TMP, IDX_PUT); | |
1001 | return true; | |
1002 | } | |
60918275 | 1003 | |
89439d4f MS |
1004 | return false; |
1005 | } | |
1006 | ||
1007 | dib = bucket_calculate_dib(h, idx, raw_dib); | |
1008 | ||
1009 | if (dib < distance) { | |
1010 | /* Found a wealthier entry. Go Robin Hood! */ | |
89439d4f MS |
1011 | bucket_set_dib(h, idx, distance); |
1012 | ||
1013 | /* swap the entries */ | |
1014 | bucket_move_entry(h, swap, idx, IDX_TMP); | |
1015 | bucket_move_entry(h, swap, IDX_PUT, idx); | |
1016 | bucket_move_entry(h, swap, IDX_TMP, IDX_PUT); | |
1017 | ||
1018 | distance = dib; | |
1019 | } | |
1020 | ||
1021 | idx = next_idx(h, idx); | |
1022 | } | |
60918275 LP |
1023 | } |
1024 | ||
89439d4f MS |
1025 | /* |
1026 | * Puts an entry into a hashmap, boldly - no check whether key already exists. | |
1027 | * The caller must place the entry (only its key and value, not link indexes) | |
1028 | * in swap slot IDX_PUT. | |
1029 | * Caller must ensure: the key does not exist yet in the hashmap. | |
1030 | * that resize is not needed if !may_resize. | |
1031 | * Returns: 1 if entry was put successfully. | |
1032 | * -ENOMEM if may_resize==true and resize failed with -ENOMEM. | |
1033 | * Cannot return -ENOMEM if !may_resize. | |
1034 | */ | |
1035 | static int hashmap_base_put_boldly(HashmapBase *h, unsigned idx, | |
1036 | struct swap_entries *swap, bool may_resize) { | |
1037 | struct ordered_hashmap_entry *new_entry; | |
1038 | int r; | |
1039 | ||
1040 | assert(idx < n_buckets(h)); | |
1041 | ||
1042 | new_entry = bucket_at_swap(swap, IDX_PUT); | |
1043 | ||
1044 | if (may_resize) { | |
1045 | r = resize_buckets(h, 1); | |
1046 | if (r < 0) | |
1047 | return r; | |
1048 | if (r > 0) | |
1049 | idx = bucket_hash(h, new_entry->p.b.key); | |
1050 | } | |
1051 | assert(n_entries(h) < n_buckets(h)); | |
1052 | ||
1053 | if (h->type == HASHMAP_TYPE_ORDERED) { | |
1054 | OrderedHashmap *lh = (OrderedHashmap*) h; | |
1055 | ||
1056 | new_entry->iterate_next = IDX_NIL; | |
1057 | new_entry->iterate_previous = lh->iterate_list_tail; | |
1058 | ||
1059 | if (lh->iterate_list_tail != IDX_NIL) { | |
1060 | struct ordered_hashmap_entry *old_tail; | |
1061 | ||
1062 | old_tail = ordered_bucket_at(lh, lh->iterate_list_tail); | |
1063 | assert(old_tail->iterate_next == IDX_NIL); | |
1064 | old_tail->iterate_next = IDX_PUT; | |
1065 | } | |
1066 | ||
1067 | lh->iterate_list_tail = IDX_PUT; | |
1068 | if (lh->iterate_list_head == IDX_NIL) | |
1069 | lh->iterate_list_head = IDX_PUT; | |
1070 | } | |
1071 | ||
1072 | assert_se(hashmap_put_robin_hood(h, idx, swap) == false); | |
1073 | ||
1074 | n_entries_inc(h); | |
349cc4a5 | 1075 | #if ENABLE_DEBUG_HASHMAP |
89439d4f MS |
1076 | h->debug.max_entries = MAX(h->debug.max_entries, n_entries(h)); |
1077 | #endif | |
1078 | ||
84dcca75 VC |
1079 | base_set_dirty(h); |
1080 | ||
89439d4f MS |
1081 | return 1; |
1082 | } | |
1083 | #define hashmap_put_boldly(h, idx, swap, may_resize) \ | |
1084 | hashmap_base_put_boldly(HASHMAP_BASE(h), idx, swap, may_resize) | |
1085 | ||
1086 | /* | |
1087 | * Returns 0 if resize is not needed. | |
f131770b | 1088 | * 1 if successfully resized. |
89439d4f MS |
1089 | * -ENOMEM on allocation failure. |
1090 | */ | |
1091 | static int resize_buckets(HashmapBase *h, unsigned entries_add) { | |
1092 | struct swap_entries swap; | |
1a39bc8c | 1093 | void *new_storage; |
89439d4f MS |
1094 | dib_raw_t *old_dibs, *new_dibs; |
1095 | const struct hashmap_type_info *hi; | |
1096 | unsigned idx, optimal_idx; | |
1097 | unsigned old_n_buckets, new_n_buckets, n_rehashed, new_n_entries; | |
1098 | uint8_t new_shift; | |
1099 | bool rehash_next; | |
45fa9e29 LP |
1100 | |
1101 | assert(h); | |
1102 | ||
89439d4f MS |
1103 | hi = &hashmap_type_info[h->type]; |
1104 | new_n_entries = n_entries(h) + entries_add; | |
e4c691b5 MS |
1105 | |
1106 | /* overflow? */ | |
89439d4f | 1107 | if (_unlikely_(new_n_entries < entries_add)) |
e4c691b5 MS |
1108 | return -ENOMEM; |
1109 | ||
89439d4f MS |
1110 | /* For direct storage we allow 100% load, because it's tiny. */ |
1111 | if (!h->has_indirect && new_n_entries <= hi->n_direct_buckets) | |
9700d698 | 1112 | return 0; |
45fa9e29 | 1113 | |
89439d4f MS |
1114 | /* |
1115 | * Load factor = n/m = 1 - (1/INV_KEEP_FREE). | |
1116 | * From it follows: m = n + n/(INV_KEEP_FREE - 1) | |
1117 | */ | |
1118 | new_n_buckets = new_n_entries + new_n_entries / (INV_KEEP_FREE - 1); | |
1119 | /* overflow? */ | |
1120 | if (_unlikely_(new_n_buckets < new_n_entries)) | |
9700d698 | 1121 | return -ENOMEM; |
45fa9e29 | 1122 | |
89439d4f MS |
1123 | if (_unlikely_(new_n_buckets > UINT_MAX / (hi->entry_size + sizeof(dib_raw_t)))) |
1124 | return -ENOMEM; | |
a3b6fafe | 1125 | |
89439d4f | 1126 | old_n_buckets = n_buckets(h); |
45fa9e29 | 1127 | |
89439d4f MS |
1128 | if (_likely_(new_n_buckets <= old_n_buckets)) |
1129 | return 0; | |
45fa9e29 | 1130 | |
89439d4f MS |
1131 | new_shift = log2u_round_up(MAX( |
1132 | new_n_buckets * (hi->entry_size + sizeof(dib_raw_t)), | |
1133 | 2 * sizeof(struct direct_storage))); | |
45fa9e29 | 1134 | |
89439d4f MS |
1135 | /* Realloc storage (buckets and DIB array). */ |
1136 | new_storage = realloc(h->has_indirect ? h->indirect.storage : NULL, | |
1137 | 1U << new_shift); | |
1138 | if (!new_storage) | |
1139 | return -ENOMEM; | |
45fa9e29 | 1140 | |
89439d4f MS |
1141 | /* Must upgrade direct to indirect storage. */ |
1142 | if (!h->has_indirect) { | |
1143 | memcpy(new_storage, h->direct.storage, | |
1144 | old_n_buckets * (hi->entry_size + sizeof(dib_raw_t))); | |
1145 | h->indirect.n_entries = h->n_direct_entries; | |
1146 | h->indirect.idx_lowest_entry = 0; | |
1147 | h->n_direct_entries = 0; | |
1148 | } | |
45fa9e29 | 1149 | |
89439d4f MS |
1150 | /* Get a new hash key. If we've just upgraded to indirect storage, |
1151 | * allow reusing a previously generated key. It's still a different key | |
1152 | * from the shared one that we used for direct storage. */ | |
1153 | get_hash_key(h->indirect.hash_key, !h->has_indirect); | |
1154 | ||
1155 | h->has_indirect = true; | |
1156 | h->indirect.storage = new_storage; | |
1157 | h->indirect.n_buckets = (1U << new_shift) / | |
1158 | (hi->entry_size + sizeof(dib_raw_t)); | |
1159 | ||
1a39bc8c | 1160 | old_dibs = (dib_raw_t*)((uint8_t*) new_storage + hi->entry_size * old_n_buckets); |
89439d4f MS |
1161 | new_dibs = dib_raw_ptr(h); |
1162 | ||
1163 | /* | |
1164 | * Move the DIB array to the new place, replacing valid DIB values with | |
1165 | * DIB_RAW_REHASH to indicate all of the used buckets need rehashing. | |
1166 | * Note: Overlap is not possible, because we have at least doubled the | |
1167 | * number of buckets and dib_raw_t is smaller than any entry type. | |
1168 | */ | |
1169 | for (idx = 0; idx < old_n_buckets; idx++) { | |
1170 | assert(old_dibs[idx] != DIB_RAW_REHASH); | |
1171 | new_dibs[idx] = old_dibs[idx] == DIB_RAW_FREE ? DIB_RAW_FREE | |
1172 | : DIB_RAW_REHASH; | |
45fa9e29 LP |
1173 | } |
1174 | ||
89439d4f | 1175 | /* Zero the area of newly added entries (including the old DIB area) */ |
eccaf899 | 1176 | memzero(bucket_at(h, old_n_buckets), |
89439d4f | 1177 | (n_buckets(h) - old_n_buckets) * hi->entry_size); |
45fa9e29 | 1178 | |
89439d4f MS |
1179 | /* The upper half of the new DIB array needs initialization */ |
1180 | memset(&new_dibs[old_n_buckets], DIB_RAW_INIT, | |
1181 | (n_buckets(h) - old_n_buckets) * sizeof(dib_raw_t)); | |
9bf3b535 | 1182 | |
89439d4f MS |
1183 | /* Rehash entries that need it */ |
1184 | n_rehashed = 0; | |
1185 | for (idx = 0; idx < old_n_buckets; idx++) { | |
1186 | if (new_dibs[idx] != DIB_RAW_REHASH) | |
1187 | continue; | |
45fa9e29 | 1188 | |
89439d4f | 1189 | optimal_idx = bucket_hash(h, bucket_at(h, idx)->key); |
45fa9e29 | 1190 | |
89439d4f MS |
1191 | /* |
1192 | * Not much to do if by luck the entry hashes to its current | |
1193 | * location. Just set its DIB. | |
1194 | */ | |
1195 | if (optimal_idx == idx) { | |
1196 | new_dibs[idx] = 0; | |
1197 | n_rehashed++; | |
1198 | continue; | |
1199 | } | |
1200 | ||
1201 | new_dibs[idx] = DIB_RAW_FREE; | |
1202 | bucket_move_entry(h, &swap, idx, IDX_PUT); | |
1203 | /* bucket_move_entry does not clear the source */ | |
eccaf899 | 1204 | memzero(bucket_at(h, idx), hi->entry_size); |
89439d4f MS |
1205 | |
1206 | do { | |
1207 | /* | |
1208 | * Find the new bucket for the current entry. This may make | |
1209 | * another entry homeless and load it into IDX_PUT. | |
1210 | */ | |
1211 | rehash_next = hashmap_put_robin_hood(h, optimal_idx, &swap); | |
1212 | n_rehashed++; | |
1213 | ||
1214 | /* Did the current entry displace another one? */ | |
1215 | if (rehash_next) | |
1216 | optimal_idx = bucket_hash(h, bucket_at_swap(&swap, IDX_PUT)->p.b.key); | |
1217 | } while (rehash_next); | |
1218 | } | |
60918275 | 1219 | |
89439d4f | 1220 | assert(n_rehashed == n_entries(h)); |
60918275 | 1221 | |
89439d4f MS |
1222 | return 1; |
1223 | } | |
45fa9e29 | 1224 | |
89439d4f MS |
1225 | /* |
1226 | * Finds an entry with a matching key | |
1227 | * Returns: index of the found entry, or IDX_NIL if not found. | |
1228 | */ | |
1229 | static unsigned base_bucket_scan(HashmapBase *h, unsigned idx, const void *key) { | |
1230 | struct hashmap_base_entry *e; | |
1231 | unsigned dib, distance; | |
1232 | dib_raw_t *dibs = dib_raw_ptr(h); | |
39c2a6f1 | 1233 | |
89439d4f | 1234 | assert(idx < n_buckets(h)); |
60918275 | 1235 | |
89439d4f MS |
1236 | for (distance = 0; ; distance++) { |
1237 | if (dibs[idx] == DIB_RAW_FREE) | |
1238 | return IDX_NIL; | |
60918275 | 1239 | |
89439d4f | 1240 | dib = bucket_calculate_dib(h, idx, dibs[idx]); |
60918275 | 1241 | |
89439d4f MS |
1242 | if (dib < distance) |
1243 | return IDX_NIL; | |
1244 | if (dib == distance) { | |
1245 | e = bucket_at(h, idx); | |
1246 | if (h->hash_ops->compare(e->key, key) == 0) | |
1247 | return idx; | |
1248 | } | |
1249 | ||
1250 | idx = next_idx(h, idx); | |
1251 | } | |
60918275 | 1252 | } |
89439d4f | 1253 | #define bucket_scan(h, idx, key) base_bucket_scan(HASHMAP_BASE(h), idx, key) |
60918275 | 1254 | |
923041cb | 1255 | int hashmap_put(Hashmap *h, const void *key, void *value) { |
89439d4f MS |
1256 | struct swap_entries swap; |
1257 | struct plain_hashmap_entry *e; | |
1258 | unsigned hash, idx; | |
923041cb MS |
1259 | |
1260 | assert(h); | |
1261 | ||
1262 | hash = bucket_hash(h, key); | |
89439d4f MS |
1263 | idx = bucket_scan(h, hash, key); |
1264 | if (idx != IDX_NIL) { | |
1265 | e = plain_bucket_at(h, idx); | |
923041cb MS |
1266 | if (e->value == value) |
1267 | return 0; | |
1268 | return -EEXIST; | |
1269 | } | |
1270 | ||
89439d4f MS |
1271 | e = &bucket_at_swap(&swap, IDX_PUT)->p; |
1272 | e->b.key = key; | |
1273 | e->value = value; | |
1274 | return hashmap_put_boldly(h, hash, &swap, true); | |
1275 | } | |
1276 | ||
1277 | int set_put(Set *s, const void *key) { | |
1278 | struct swap_entries swap; | |
1279 | struct hashmap_base_entry *e; | |
1280 | unsigned hash, idx; | |
1281 | ||
1282 | assert(s); | |
1283 | ||
1284 | hash = bucket_hash(s, key); | |
1285 | idx = bucket_scan(s, hash, key); | |
1286 | if (idx != IDX_NIL) | |
1287 | return 0; | |
1288 | ||
1289 | e = &bucket_at_swap(&swap, IDX_PUT)->p.b; | |
1290 | e->key = key; | |
1291 | return hashmap_put_boldly(s, hash, &swap, true); | |
923041cb MS |
1292 | } |
1293 | ||
3158713e | 1294 | int hashmap_replace(Hashmap *h, const void *key, void *value) { |
89439d4f MS |
1295 | struct swap_entries swap; |
1296 | struct plain_hashmap_entry *e; | |
1297 | unsigned hash, idx; | |
3158713e LP |
1298 | |
1299 | assert(h); | |
1300 | ||
a3b6fafe | 1301 | hash = bucket_hash(h, key); |
89439d4f MS |
1302 | idx = bucket_scan(h, hash, key); |
1303 | if (idx != IDX_NIL) { | |
1304 | e = plain_bucket_at(h, idx); | |
349cc4a5 | 1305 | #if ENABLE_DEBUG_HASHMAP |
89439d4f MS |
1306 | /* Although the key is equal, the key pointer may have changed, |
1307 | * and this would break our assumption for iterating. So count | |
1308 | * this operation as incompatible with iteration. */ | |
1309 | if (e->b.key != key) { | |
1310 | h->b.debug.put_count++; | |
1311 | h->b.debug.rem_count++; | |
1312 | h->b.debug.last_rem_idx = idx; | |
1313 | } | |
1314 | #endif | |
1315 | e->b.key = key; | |
3158713e | 1316 | e->value = value; |
84dcca75 VC |
1317 | hashmap_set_dirty(h); |
1318 | ||
3158713e LP |
1319 | return 0; |
1320 | } | |
1321 | ||
89439d4f MS |
1322 | e = &bucket_at_swap(&swap, IDX_PUT)->p; |
1323 | e->b.key = key; | |
1324 | e->value = value; | |
1325 | return hashmap_put_boldly(h, hash, &swap, true); | |
3158713e LP |
1326 | } |
1327 | ||
d99ae53a | 1328 | int hashmap_update(Hashmap *h, const void *key, void *value) { |
89439d4f MS |
1329 | struct plain_hashmap_entry *e; |
1330 | unsigned hash, idx; | |
d99ae53a LP |
1331 | |
1332 | assert(h); | |
1333 | ||
a3b6fafe | 1334 | hash = bucket_hash(h, key); |
89439d4f MS |
1335 | idx = bucket_scan(h, hash, key); |
1336 | if (idx == IDX_NIL) | |
d99ae53a LP |
1337 | return -ENOENT; |
1338 | ||
89439d4f | 1339 | e = plain_bucket_at(h, idx); |
d99ae53a | 1340 | e->value = value; |
84dcca75 VC |
1341 | hashmap_set_dirty(h); |
1342 | ||
d99ae53a LP |
1343 | return 0; |
1344 | } | |
1345 | ||
89439d4f MS |
1346 | void *internal_hashmap_get(HashmapBase *h, const void *key) { |
1347 | struct hashmap_base_entry *e; | |
1348 | unsigned hash, idx; | |
60918275 LP |
1349 | |
1350 | if (!h) | |
1351 | return NULL; | |
1352 | ||
a3b6fafe | 1353 | hash = bucket_hash(h, key); |
89439d4f MS |
1354 | idx = bucket_scan(h, hash, key); |
1355 | if (idx == IDX_NIL) | |
60918275 LP |
1356 | return NULL; |
1357 | ||
89439d4f MS |
1358 | e = bucket_at(h, idx); |
1359 | return entry_value(h, e); | |
60918275 LP |
1360 | } |
1361 | ||
89439d4f MS |
1362 | void *hashmap_get2(Hashmap *h, const void *key, void **key2) { |
1363 | struct plain_hashmap_entry *e; | |
1364 | unsigned hash, idx; | |
d99ae53a LP |
1365 | |
1366 | if (!h) | |
1367 | return NULL; | |
1368 | ||
a3b6fafe | 1369 | hash = bucket_hash(h, key); |
89439d4f MS |
1370 | idx = bucket_scan(h, hash, key); |
1371 | if (idx == IDX_NIL) | |
d99ae53a LP |
1372 | return NULL; |
1373 | ||
89439d4f | 1374 | e = plain_bucket_at(h, idx); |
d99ae53a | 1375 | if (key2) |
89439d4f | 1376 | *key2 = (void*) e->b.key; |
d99ae53a LP |
1377 | |
1378 | return e->value; | |
1379 | } | |
1380 | ||
89439d4f | 1381 | bool internal_hashmap_contains(HashmapBase *h, const void *key) { |
96342de6 | 1382 | unsigned hash; |
96342de6 LN |
1383 | |
1384 | if (!h) | |
1385 | return false; | |
1386 | ||
a3b6fafe | 1387 | hash = bucket_hash(h, key); |
89439d4f | 1388 | return bucket_scan(h, hash, key) != IDX_NIL; |
96342de6 LN |
1389 | } |
1390 | ||
89439d4f MS |
1391 | void *internal_hashmap_remove(HashmapBase *h, const void *key) { |
1392 | struct hashmap_base_entry *e; | |
1393 | unsigned hash, idx; | |
60918275 LP |
1394 | void *data; |
1395 | ||
1396 | if (!h) | |
1397 | return NULL; | |
1398 | ||
a3b6fafe | 1399 | hash = bucket_hash(h, key); |
89439d4f MS |
1400 | idx = bucket_scan(h, hash, key); |
1401 | if (idx == IDX_NIL) | |
60918275 LP |
1402 | return NULL; |
1403 | ||
89439d4f MS |
1404 | e = bucket_at(h, idx); |
1405 | data = entry_value(h, e); | |
1406 | remove_entry(h, idx); | |
60918275 LP |
1407 | |
1408 | return data; | |
1409 | } | |
1410 | ||
89439d4f MS |
1411 | void *hashmap_remove2(Hashmap *h, const void *key, void **rkey) { |
1412 | struct plain_hashmap_entry *e; | |
1413 | unsigned hash, idx; | |
c582a3b3 LP |
1414 | void *data; |
1415 | ||
1416 | if (!h) { | |
1417 | if (rkey) | |
1418 | *rkey = NULL; | |
1419 | return NULL; | |
1420 | } | |
1421 | ||
1422 | hash = bucket_hash(h, key); | |
89439d4f MS |
1423 | idx = bucket_scan(h, hash, key); |
1424 | if (idx == IDX_NIL) { | |
c582a3b3 LP |
1425 | if (rkey) |
1426 | *rkey = NULL; | |
1427 | return NULL; | |
1428 | } | |
1429 | ||
89439d4f | 1430 | e = plain_bucket_at(h, idx); |
c582a3b3 LP |
1431 | data = e->value; |
1432 | if (rkey) | |
89439d4f | 1433 | *rkey = (void*) e->b.key; |
c582a3b3 | 1434 | |
89439d4f | 1435 | remove_entry(h, idx); |
c582a3b3 LP |
1436 | |
1437 | return data; | |
1438 | } | |
1439 | ||
101d8e63 | 1440 | int hashmap_remove_and_put(Hashmap *h, const void *old_key, const void *new_key, void *value) { |
89439d4f MS |
1441 | struct swap_entries swap; |
1442 | struct plain_hashmap_entry *e; | |
1443 | unsigned old_hash, new_hash, idx; | |
101d8e63 LP |
1444 | |
1445 | if (!h) | |
1446 | return -ENOENT; | |
1447 | ||
a3b6fafe | 1448 | old_hash = bucket_hash(h, old_key); |
89439d4f MS |
1449 | idx = bucket_scan(h, old_hash, old_key); |
1450 | if (idx == IDX_NIL) | |
101d8e63 LP |
1451 | return -ENOENT; |
1452 | ||
a3b6fafe | 1453 | new_hash = bucket_hash(h, new_key); |
89439d4f | 1454 | if (bucket_scan(h, new_hash, new_key) != IDX_NIL) |
101d8e63 LP |
1455 | return -EEXIST; |
1456 | ||
89439d4f | 1457 | remove_entry(h, idx); |
101d8e63 | 1458 | |
89439d4f MS |
1459 | e = &bucket_at_swap(&swap, IDX_PUT)->p; |
1460 | e->b.key = new_key; | |
101d8e63 | 1461 | e->value = value; |
89439d4f MS |
1462 | assert_se(hashmap_put_boldly(h, new_hash, &swap, false) == 1); |
1463 | ||
1464 | return 0; | |
1465 | } | |
1466 | ||
1467 | int set_remove_and_put(Set *s, const void *old_key, const void *new_key) { | |
1468 | struct swap_entries swap; | |
1469 | struct hashmap_base_entry *e; | |
1470 | unsigned old_hash, new_hash, idx; | |
101d8e63 | 1471 | |
89439d4f MS |
1472 | if (!s) |
1473 | return -ENOENT; | |
1474 | ||
1475 | old_hash = bucket_hash(s, old_key); | |
1476 | idx = bucket_scan(s, old_hash, old_key); | |
1477 | if (idx == IDX_NIL) | |
1478 | return -ENOENT; | |
1479 | ||
1480 | new_hash = bucket_hash(s, new_key); | |
1481 | if (bucket_scan(s, new_hash, new_key) != IDX_NIL) | |
1482 | return -EEXIST; | |
1483 | ||
1484 | remove_entry(s, idx); | |
1485 | ||
1486 | e = &bucket_at_swap(&swap, IDX_PUT)->p.b; | |
1487 | e->key = new_key; | |
1488 | assert_se(hashmap_put_boldly(s, new_hash, &swap, false) == 1); | |
101d8e63 LP |
1489 | |
1490 | return 0; | |
1491 | } | |
1492 | ||
8fe914ec | 1493 | int hashmap_remove_and_replace(Hashmap *h, const void *old_key, const void *new_key, void *value) { |
89439d4f MS |
1494 | struct swap_entries swap; |
1495 | struct plain_hashmap_entry *e; | |
1496 | unsigned old_hash, new_hash, idx_old, idx_new; | |
8fe914ec LP |
1497 | |
1498 | if (!h) | |
1499 | return -ENOENT; | |
1500 | ||
a3b6fafe | 1501 | old_hash = bucket_hash(h, old_key); |
89439d4f MS |
1502 | idx_old = bucket_scan(h, old_hash, old_key); |
1503 | if (idx_old == IDX_NIL) | |
8fe914ec LP |
1504 | return -ENOENT; |
1505 | ||
89439d4f | 1506 | old_key = bucket_at(HASHMAP_BASE(h), idx_old)->key; |
8fe914ec | 1507 | |
89439d4f MS |
1508 | new_hash = bucket_hash(h, new_key); |
1509 | idx_new = bucket_scan(h, new_hash, new_key); | |
1510 | if (idx_new != IDX_NIL) | |
1511 | if (idx_old != idx_new) { | |
1512 | remove_entry(h, idx_new); | |
1513 | /* Compensate for a possible backward shift. */ | |
1514 | if (old_key != bucket_at(HASHMAP_BASE(h), idx_old)->key) | |
1515 | idx_old = prev_idx(HASHMAP_BASE(h), idx_old); | |
1516 | assert(old_key == bucket_at(HASHMAP_BASE(h), idx_old)->key); | |
1517 | } | |
1518 | ||
1519 | remove_entry(h, idx_old); | |
1520 | ||
1521 | e = &bucket_at_swap(&swap, IDX_PUT)->p; | |
1522 | e->b.key = new_key; | |
8fe914ec | 1523 | e->value = value; |
89439d4f | 1524 | assert_se(hashmap_put_boldly(h, new_hash, &swap, false) == 1); |
8fe914ec LP |
1525 | |
1526 | return 0; | |
1527 | } | |
1528 | ||
89439d4f MS |
1529 | void *hashmap_remove_value(Hashmap *h, const void *key, void *value) { |
1530 | struct plain_hashmap_entry *e; | |
1531 | unsigned hash, idx; | |
3158713e LP |
1532 | |
1533 | if (!h) | |
1534 | return NULL; | |
1535 | ||
a3b6fafe | 1536 | hash = bucket_hash(h, key); |
89439d4f MS |
1537 | idx = bucket_scan(h, hash, key); |
1538 | if (idx == IDX_NIL) | |
3158713e LP |
1539 | return NULL; |
1540 | ||
89439d4f | 1541 | e = plain_bucket_at(h, idx); |
3158713e LP |
1542 | if (e->value != value) |
1543 | return NULL; | |
1544 | ||
89439d4f | 1545 | remove_entry(h, idx); |
3158713e LP |
1546 | |
1547 | return value; | |
1548 | } | |
1549 | ||
89439d4f MS |
1550 | static unsigned find_first_entry(HashmapBase *h) { |
1551 | Iterator i = ITERATOR_FIRST; | |
60918275 | 1552 | |
89439d4f MS |
1553 | if (!h || !n_entries(h)) |
1554 | return IDX_NIL; | |
60918275 | 1555 | |
89439d4f | 1556 | return hashmap_iterate_entry(h, &i); |
60918275 LP |
1557 | } |
1558 | ||
89439d4f MS |
1559 | void *internal_hashmap_first(HashmapBase *h) { |
1560 | unsigned idx; | |
60918275 | 1561 | |
89439d4f MS |
1562 | idx = find_first_entry(h); |
1563 | if (idx == IDX_NIL) | |
60918275 LP |
1564 | return NULL; |
1565 | ||
89439d4f | 1566 | return entry_value(h, bucket_at(h, idx)); |
60918275 LP |
1567 | } |
1568 | ||
89439d4f MS |
1569 | void *internal_hashmap_first_key(HashmapBase *h) { |
1570 | struct hashmap_base_entry *e; | |
1571 | unsigned idx; | |
2e4a6ff4 | 1572 | |
89439d4f MS |
1573 | idx = find_first_entry(h); |
1574 | if (idx == IDX_NIL) | |
2e4a6ff4 LP |
1575 | return NULL; |
1576 | ||
89439d4f MS |
1577 | e = bucket_at(h, idx); |
1578 | return (void*) e->key; | |
2e4a6ff4 LP |
1579 | } |
1580 | ||
89439d4f MS |
1581 | void *internal_hashmap_steal_first(HashmapBase *h) { |
1582 | struct hashmap_base_entry *e; | |
60918275 | 1583 | void *data; |
89439d4f | 1584 | unsigned idx; |
60918275 | 1585 | |
89439d4f MS |
1586 | idx = find_first_entry(h); |
1587 | if (idx == IDX_NIL) | |
60918275 LP |
1588 | return NULL; |
1589 | ||
89439d4f MS |
1590 | e = bucket_at(h, idx); |
1591 | data = entry_value(h, e); | |
1592 | remove_entry(h, idx); | |
60918275 LP |
1593 | |
1594 | return data; | |
1595 | } | |
1596 | ||
89439d4f MS |
1597 | void *internal_hashmap_steal_first_key(HashmapBase *h) { |
1598 | struct hashmap_base_entry *e; | |
22be093f | 1599 | void *key; |
89439d4f | 1600 | unsigned idx; |
22be093f | 1601 | |
89439d4f MS |
1602 | idx = find_first_entry(h); |
1603 | if (idx == IDX_NIL) | |
22be093f LP |
1604 | return NULL; |
1605 | ||
89439d4f MS |
1606 | e = bucket_at(h, idx); |
1607 | key = (void*) e->key; | |
1608 | remove_entry(h, idx); | |
22be093f LP |
1609 | |
1610 | return key; | |
1611 | } | |
1612 | ||
89439d4f | 1613 | unsigned internal_hashmap_size(HashmapBase *h) { |
60918275 LP |
1614 | |
1615 | if (!h) | |
1616 | return 0; | |
1617 | ||
89439d4f | 1618 | return n_entries(h); |
60918275 LP |
1619 | } |
1620 | ||
89439d4f | 1621 | unsigned internal_hashmap_buckets(HashmapBase *h) { |
45fa9e29 LP |
1622 | |
1623 | if (!h) | |
1624 | return 0; | |
1625 | ||
89439d4f | 1626 | return n_buckets(h); |
45fa9e29 LP |
1627 | } |
1628 | ||
89439d4f MS |
1629 | int internal_hashmap_merge(Hashmap *h, Hashmap *other) { |
1630 | Iterator i; | |
1631 | unsigned idx; | |
60918275 | 1632 | |
89439d4f | 1633 | assert(h); |
60918275 | 1634 | |
89439d4f MS |
1635 | HASHMAP_FOREACH_IDX(idx, HASHMAP_BASE(other), i) { |
1636 | struct plain_hashmap_entry *pe = plain_bucket_at(other, idx); | |
1637 | int r; | |
91cdde8a | 1638 | |
89439d4f MS |
1639 | r = hashmap_put(h, pe->b.key, pe->value); |
1640 | if (r < 0 && r != -EEXIST) | |
1641 | return r; | |
1642 | } | |
91cdde8a | 1643 | |
89439d4f MS |
1644 | return 0; |
1645 | } | |
91cdde8a | 1646 | |
89439d4f MS |
1647 | int set_merge(Set *s, Set *other) { |
1648 | Iterator i; | |
1649 | unsigned idx; | |
91cdde8a | 1650 | |
89439d4f MS |
1651 | assert(s); |
1652 | ||
1653 | HASHMAP_FOREACH_IDX(idx, HASHMAP_BASE(other), i) { | |
1654 | struct set_entry *se = set_bucket_at(other, idx); | |
91cdde8a LP |
1655 | int r; |
1656 | ||
89439d4f MS |
1657 | r = set_put(s, se->b.key); |
1658 | if (r < 0) | |
a3b6fafe | 1659 | return r; |
91cdde8a LP |
1660 | } |
1661 | ||
1662 | return 0; | |
1663 | } | |
1664 | ||
89439d4f | 1665 | int internal_hashmap_reserve(HashmapBase *h, unsigned entries_add) { |
e4c691b5 MS |
1666 | int r; |
1667 | ||
1668 | assert(h); | |
1669 | ||
1670 | r = resize_buckets(h, entries_add); | |
1671 | if (r < 0) | |
1672 | return r; | |
1673 | ||
1674 | return 0; | |
1675 | } | |
1676 | ||
89439d4f MS |
1677 | /* |
1678 | * The same as hashmap_merge(), but every new item from other is moved to h. | |
1679 | * Keys already in h are skipped and stay in other. | |
1680 | * Returns: 0 on success. | |
1681 | * -ENOMEM on alloc failure, in which case no move has been done. | |
1682 | */ | |
1683 | int internal_hashmap_move(HashmapBase *h, HashmapBase *other) { | |
1684 | struct swap_entries swap; | |
1685 | struct hashmap_base_entry *e, *n; | |
1686 | Iterator i; | |
1687 | unsigned idx; | |
1688 | int r; | |
101d8e63 LP |
1689 | |
1690 | assert(h); | |
1691 | ||
101d8e63 | 1692 | if (!other) |
7ad63f57 | 1693 | return 0; |
101d8e63 | 1694 | |
89439d4f MS |
1695 | assert(other->type == h->type); |
1696 | ||
1697 | /* | |
1698 | * This reserves buckets for the worst case, where none of other's | |
1699 | * entries are yet present in h. This is preferable to risking | |
1700 | * an allocation failure in the middle of the moving and having to | |
1701 | * rollback or return a partial result. | |
1702 | */ | |
1703 | r = resize_buckets(h, n_entries(other)); | |
1704 | if (r < 0) | |
1705 | return r; | |
101d8e63 | 1706 | |
89439d4f MS |
1707 | HASHMAP_FOREACH_IDX(idx, other, i) { |
1708 | unsigned h_hash; | |
101d8e63 | 1709 | |
89439d4f | 1710 | e = bucket_at(other, idx); |
a3b6fafe | 1711 | h_hash = bucket_hash(h, e->key); |
89439d4f | 1712 | if (bucket_scan(h, h_hash, e->key) != IDX_NIL) |
101d8e63 LP |
1713 | continue; |
1714 | ||
89439d4f MS |
1715 | n = &bucket_at_swap(&swap, IDX_PUT)->p.b; |
1716 | n->key = e->key; | |
1717 | if (h->type != HASHMAP_TYPE_SET) | |
1718 | ((struct plain_hashmap_entry*) n)->value = | |
1719 | ((struct plain_hashmap_entry*) e)->value; | |
1720 | assert_se(hashmap_put_boldly(h, h_hash, &swap, false) == 1); | |
1721 | ||
1722 | remove_entry(other, idx); | |
101d8e63 | 1723 | } |
7ad63f57 MS |
1724 | |
1725 | return 0; | |
101d8e63 LP |
1726 | } |
1727 | ||
89439d4f MS |
1728 | int internal_hashmap_move_one(HashmapBase *h, HashmapBase *other, const void *key) { |
1729 | struct swap_entries swap; | |
1730 | unsigned h_hash, other_hash, idx; | |
1731 | struct hashmap_base_entry *e, *n; | |
1732 | int r; | |
101d8e63 | 1733 | |
101d8e63 LP |
1734 | assert(h); |
1735 | ||
a3b6fafe | 1736 | h_hash = bucket_hash(h, key); |
89439d4f | 1737 | if (bucket_scan(h, h_hash, key) != IDX_NIL) |
101d8e63 LP |
1738 | return -EEXIST; |
1739 | ||
bf3d3e2b MS |
1740 | if (!other) |
1741 | return -ENOENT; | |
1742 | ||
89439d4f MS |
1743 | assert(other->type == h->type); |
1744 | ||
a3b6fafe | 1745 | other_hash = bucket_hash(other, key); |
89439d4f MS |
1746 | idx = bucket_scan(other, other_hash, key); |
1747 | if (idx == IDX_NIL) | |
101d8e63 LP |
1748 | return -ENOENT; |
1749 | ||
89439d4f MS |
1750 | e = bucket_at(other, idx); |
1751 | ||
1752 | n = &bucket_at_swap(&swap, IDX_PUT)->p.b; | |
1753 | n->key = e->key; | |
1754 | if (h->type != HASHMAP_TYPE_SET) | |
1755 | ((struct plain_hashmap_entry*) n)->value = | |
1756 | ((struct plain_hashmap_entry*) e)->value; | |
1757 | r = hashmap_put_boldly(h, h_hash, &swap, true); | |
1758 | if (r < 0) | |
1759 | return r; | |
101d8e63 | 1760 | |
89439d4f | 1761 | remove_entry(other, idx); |
101d8e63 LP |
1762 | return 0; |
1763 | } | |
1764 | ||
89439d4f MS |
1765 | HashmapBase *internal_hashmap_copy(HashmapBase *h) { |
1766 | HashmapBase *copy; | |
1767 | int r; | |
91cdde8a LP |
1768 | |
1769 | assert(h); | |
1770 | ||
89439d4f | 1771 | copy = hashmap_base_new(h->hash_ops, h->type HASHMAP_DEBUG_SRC_ARGS); |
45fa9e29 | 1772 | if (!copy) |
91cdde8a LP |
1773 | return NULL; |
1774 | ||
89439d4f MS |
1775 | switch (h->type) { |
1776 | case HASHMAP_TYPE_PLAIN: | |
1777 | case HASHMAP_TYPE_ORDERED: | |
1778 | r = hashmap_merge((Hashmap*)copy, (Hashmap*)h); | |
1779 | break; | |
1780 | case HASHMAP_TYPE_SET: | |
1781 | r = set_merge((Set*)copy, (Set*)h); | |
1782 | break; | |
1783 | default: | |
1784 | assert_not_reached("Unknown hashmap type"); | |
1785 | } | |
1786 | ||
1787 | if (r < 0) { | |
1788 | internal_hashmap_free(copy); | |
91cdde8a LP |
1789 | return NULL; |
1790 | } | |
1791 | ||
1792 | return copy; | |
1793 | } | |
db1413d7 | 1794 | |
89439d4f | 1795 | char **internal_hashmap_get_strv(HashmapBase *h) { |
db1413d7 | 1796 | char **sv; |
89439d4f MS |
1797 | Iterator i; |
1798 | unsigned idx, n; | |
db1413d7 | 1799 | |
89439d4f | 1800 | sv = new(char*, n_entries(h)+1); |
729e3769 | 1801 | if (!sv) |
db1413d7 KS |
1802 | return NULL; |
1803 | ||
1804 | n = 0; | |
89439d4f MS |
1805 | HASHMAP_FOREACH_IDX(idx, h, i) |
1806 | sv[n++] = entry_value(h, bucket_at(h, idx)); | |
db1413d7 KS |
1807 | sv[n] = NULL; |
1808 | ||
1809 | return sv; | |
1810 | } | |
3c1668da | 1811 | |
89439d4f MS |
1812 | void *ordered_hashmap_next(OrderedHashmap *h, const void *key) { |
1813 | struct ordered_hashmap_entry *e; | |
1814 | unsigned hash, idx; | |
3c1668da | 1815 | |
3c1668da LP |
1816 | if (!h) |
1817 | return NULL; | |
1818 | ||
a3b6fafe | 1819 | hash = bucket_hash(h, key); |
89439d4f MS |
1820 | idx = bucket_scan(h, hash, key); |
1821 | if (idx == IDX_NIL) | |
3c1668da LP |
1822 | return NULL; |
1823 | ||
89439d4f MS |
1824 | e = ordered_bucket_at(h, idx); |
1825 | if (e->iterate_next == IDX_NIL) | |
3c1668da | 1826 | return NULL; |
89439d4f MS |
1827 | return ordered_bucket_at(h, e->iterate_next)->p.value; |
1828 | } | |
3c1668da | 1829 | |
89439d4f MS |
1830 | int set_consume(Set *s, void *value) { |
1831 | int r; | |
1832 | ||
d97c5aea LP |
1833 | assert(s); |
1834 | assert(value); | |
1835 | ||
89439d4f | 1836 | r = set_put(s, value); |
575ccc1b | 1837 | if (r <= 0) |
89439d4f MS |
1838 | free(value); |
1839 | ||
1840 | return r; | |
1841 | } | |
1842 | ||
1843 | int set_put_strdup(Set *s, const char *p) { | |
1844 | char *c; | |
89439d4f MS |
1845 | |
1846 | assert(s); | |
1847 | assert(p); | |
1848 | ||
454f0f86 LP |
1849 | if (set_contains(s, (char*) p)) |
1850 | return 0; | |
1851 | ||
89439d4f MS |
1852 | c = strdup(p); |
1853 | if (!c) | |
1854 | return -ENOMEM; | |
1855 | ||
454f0f86 | 1856 | return set_consume(s, c); |
89439d4f MS |
1857 | } |
1858 | ||
1859 | int set_put_strdupv(Set *s, char **l) { | |
1860 | int n = 0, r; | |
1861 | char **i; | |
1862 | ||
d97c5aea LP |
1863 | assert(s); |
1864 | ||
89439d4f MS |
1865 | STRV_FOREACH(i, l) { |
1866 | r = set_put_strdup(s, *i); | |
1867 | if (r < 0) | |
1868 | return r; | |
1869 | ||
1870 | n += r; | |
1871 | } | |
1872 | ||
1873 | return n; | |
3c1668da | 1874 | } |
d97c5aea LP |
1875 | |
1876 | int set_put_strsplit(Set *s, const char *v, const char *separators, ExtractFlags flags) { | |
1877 | const char *p = v; | |
1878 | int r; | |
1879 | ||
1880 | assert(s); | |
1881 | assert(v); | |
1882 | ||
1883 | for (;;) { | |
1884 | char *word; | |
1885 | ||
1886 | r = extract_first_word(&p, &word, separators, flags); | |
1887 | if (r <= 0) | |
1888 | return r; | |
1889 | ||
1890 | r = set_consume(s, word); | |
1891 | if (r < 0) | |
1892 | return r; | |
1893 | } | |
1894 | } | |
45ea84d8 VC |
1895 | |
1896 | /* expand the cachemem if needed, return true if newly (re)activated. */ | |
1897 | static int cachemem_maintain(CacheMem *mem, unsigned size) { | |
45ea84d8 VC |
1898 | assert(mem); |
1899 | ||
1900 | if (!GREEDY_REALLOC(mem->ptr, mem->n_allocated, size)) { | |
1901 | if (size > 0) | |
1902 | return -ENOMEM; | |
1903 | } | |
1904 | ||
afbbc068 ZJS |
1905 | if (!mem->active) { |
1906 | mem->active = true; | |
1907 | return true; | |
1908 | } | |
45ea84d8 | 1909 | |
afbbc068 | 1910 | return false; |
45ea84d8 VC |
1911 | } |
1912 | ||
1913 | int iterated_cache_get(IteratedCache *cache, const void ***res_keys, const void ***res_values, unsigned *res_n_entries) { | |
1914 | bool sync_keys = false, sync_values = false; | |
1915 | unsigned size; | |
1916 | int r; | |
1917 | ||
1918 | assert(cache); | |
1919 | assert(cache->hashmap); | |
1920 | ||
1921 | size = n_entries(cache->hashmap); | |
1922 | ||
1923 | if (res_keys) { | |
1924 | r = cachemem_maintain(&cache->keys, size); | |
1925 | if (r < 0) | |
1926 | return r; | |
1927 | ||
1928 | sync_keys = r; | |
1929 | } else | |
1930 | cache->keys.active = false; | |
1931 | ||
1932 | if (res_values) { | |
1933 | r = cachemem_maintain(&cache->values, size); | |
1934 | if (r < 0) | |
1935 | return r; | |
1936 | ||
1937 | sync_values = r; | |
1938 | } else | |
1939 | cache->values.active = false; | |
1940 | ||
1941 | if (cache->hashmap->dirty) { | |
1942 | if (cache->keys.active) | |
1943 | sync_keys = true; | |
1944 | if (cache->values.active) | |
1945 | sync_values = true; | |
1946 | ||
1947 | cache->hashmap->dirty = false; | |
1948 | } | |
1949 | ||
1950 | if (sync_keys || sync_values) { | |
1951 | unsigned i, idx; | |
1952 | Iterator iter; | |
1953 | ||
1954 | i = 0; | |
1955 | HASHMAP_FOREACH_IDX(idx, cache->hashmap, iter) { | |
1956 | struct hashmap_base_entry *e; | |
1957 | ||
1958 | e = bucket_at(cache->hashmap, idx); | |
1959 | ||
1960 | if (sync_keys) | |
1961 | cache->keys.ptr[i] = e->key; | |
1962 | if (sync_values) | |
1963 | cache->values.ptr[i] = entry_value(cache->hashmap, e); | |
1964 | i++; | |
1965 | } | |
1966 | } | |
1967 | ||
1968 | if (res_keys) | |
1969 | *res_keys = cache->keys.ptr; | |
1970 | if (res_values) | |
1971 | *res_values = cache->values.ptr; | |
1972 | if (res_n_entries) | |
1973 | *res_n_entries = size; | |
1974 | ||
1975 | return 0; | |
1976 | } | |
1977 | ||
1978 | IteratedCache *iterated_cache_free(IteratedCache *cache) { | |
1979 | if (cache) { | |
1980 | free(cache->keys.ptr); | |
1981 | free(cache->values.ptr); | |
1982 | free(cache); | |
1983 | } | |
1984 | ||
1985 | return NULL; | |
1986 | } |