| 1 |
/* SPDX-License-Identifier: LGPL-2.1-or-later */ |
| 2 |
|
| 3 |
#include <fnmatch.h> |
| 4 |
#include <pthread.h> |
| 5 |
#include <unistd.h> |
| 6 |
#if HAVE_VALGRIND_VALGRIND_H |
| 7 |
# include <valgrind/valgrind.h> |
| 8 |
#endif |
| 9 |
|
| 10 |
#include "alloc-util.h" |
| 11 |
#include "extract-word.h" |
| 12 |
#include "hashmap.h" |
| 13 |
#include "log.h" |
| 14 |
#include "logarithm.h" |
| 15 |
#include "memory-util.h" |
| 16 |
#include "mempool.h" |
| 17 |
#include "process-util.h" |
| 18 |
#include "random-util.h" |
| 19 |
#include "set.h" |
| 20 |
#include "siphash24.h" |
| 21 |
#include "sort-util.h" |
| 22 |
#include "string-util.h" |
| 23 |
#include "strv.h" |
| 24 |
|
| 25 |
#if ENABLE_DEBUG_HASHMAP |
| 26 |
#include "list.h" |
| 27 |
#endif |
| 28 |
|
| 29 |
/* |
| 30 |
* Implementation of hashmaps. |
| 31 |
* Addressing: open |
| 32 |
* - uses less RAM compared to closed addressing (chaining), because |
| 33 |
* our entries are small (especially in Sets, which tend to contain |
| 34 |
* the majority of entries in systemd). |
| 35 |
* Collision resolution: Robin Hood |
| 36 |
* - tends to equalize displacement of entries from their optimal buckets. |
| 37 |
* Probe sequence: linear |
| 38 |
* - though theoretically worse than random probing/uniform hashing/double |
| 39 |
* hashing, it is good for cache locality. |
| 40 |
* |
| 41 |
* References: |
| 42 |
* Celis, P. 1986. Robin Hood Hashing. |
| 43 |
* Ph.D. Dissertation. University of Waterloo, Waterloo, Ont., Canada, Canada. |
| 44 |
* https://cs.uwaterloo.ca/research/tr/1986/CS-86-14.pdf |
| 45 |
* - The results are derived for random probing. Suggests deletion with |
| 46 |
* tombstones and two mean-centered search methods. None of that works |
| 47 |
* well for linear probing. |
| 48 |
* |
| 49 |
* Janson, S. 2005. Individual displacements for linear probing hashing with different insertion policies. |
| 50 |
* ACM Trans. Algorithms 1, 2 (October 2005), 177-213. |
| 51 |
* DOI=10.1145/1103963.1103964 http://doi.acm.org/10.1145/1103963.1103964 |
| 52 |
* http://www.math.uu.se/~svante/papers/sj157.pdf |
| 53 |
* - Applies to Robin Hood with linear probing. Contains remarks on |
| 54 |
* the unsuitability of mean-centered search with linear probing. |
| 55 |
* |
| 56 |
* Viola, A. 2005. Exact distribution of individual displacements in linear probing hashing. |
| 57 |
* ACM Trans. Algorithms 1, 2 (October 2005), 214-242. |
| 58 |
* DOI=10.1145/1103963.1103965 http://doi.acm.org/10.1145/1103963.1103965 |
| 59 |
* - Similar to Janson. Note that Viola writes about C_{m,n} (number of probes |
| 60 |
* in a successful search), and Janson writes about displacement. C = d + 1. |
| 61 |
* |
| 62 |
* Goossaert, E. 2013. Robin Hood hashing: backward shift deletion. |
| 63 |
* http://codecapsule.com/2013/11/17/robin-hood-hashing-backward-shift-deletion/ |
| 64 |
* - Explanation of backward shift deletion with pictures. |
| 65 |
* |
| 66 |
* Khuong, P. 2013. The Other Robin Hood Hashing. |
| 67 |
* http://www.pvk.ca/Blog/2013/11/26/the-other-robin-hood-hashing/ |
| 68 |
* - Short summary of random vs. linear probing, and tombstones vs. backward shift. |
| 69 |
*/ |
| 70 |
|
| 71 |
/* |
| 72 |
* XXX Ideas for improvement: |
| 73 |
* For unordered hashmaps, randomize iteration order, similarly to Perl: |
| 74 |
* http://blog.booking.com/hardening-perls-hash-function.html |
| 75 |
*/ |
| 76 |
|
| 77 |
/* INV_KEEP_FREE = 1 / (1 - max_load_factor) |
| 78 |
* e.g. 1 / (1 - 0.8) = 5 ... keep one fifth of the buckets free. */ |
| 79 |
#define INV_KEEP_FREE 5U |
| 80 |
|
| 81 |
/* Fields common to entries of all hashmap/set types */ |
| 82 |
struct hashmap_base_entry { |
| 83 |
const void *key; |
| 84 |
}; |
| 85 |
|
| 86 |
/* Entry types for specific hashmap/set types |
| 87 |
* hashmap_base_entry must be at the beginning of each entry struct. */ |
| 88 |
|
| 89 |
struct plain_hashmap_entry { |
| 90 |
struct hashmap_base_entry b; |
| 91 |
void *value; |
| 92 |
}; |
| 93 |
|
| 94 |
struct ordered_hashmap_entry { |
| 95 |
struct plain_hashmap_entry p; |
| 96 |
unsigned iterate_next, iterate_previous; |
| 97 |
}; |
| 98 |
|
| 99 |
struct set_entry { |
| 100 |
struct hashmap_base_entry b; |
| 101 |
}; |
| 102 |
|
| 103 |
/* In several functions it is advantageous to have the hash table extended |
| 104 |
* virtually by a couple of additional buckets. We reserve special index values |
| 105 |
* for these "swap" buckets. */ |
| 106 |
#define _IDX_SWAP_BEGIN (UINT_MAX - 3) |
| 107 |
#define IDX_PUT (_IDX_SWAP_BEGIN + 0) |
| 108 |
#define IDX_TMP (_IDX_SWAP_BEGIN + 1) |
| 109 |
#define _IDX_SWAP_END (_IDX_SWAP_BEGIN + 2) |
| 110 |
|
| 111 |
#define IDX_FIRST (UINT_MAX - 1) /* special index for freshly initialized iterators */ |
| 112 |
#define IDX_NIL UINT_MAX /* special index value meaning "none" or "end" */ |
| 113 |
|
| 114 |
assert_cc(IDX_FIRST == _IDX_SWAP_END); |
| 115 |
assert_cc(IDX_FIRST == _IDX_ITERATOR_FIRST); |
| 116 |
|
| 117 |
/* Storage space for the "swap" buckets. |
| 118 |
* All entry types can fit into an ordered_hashmap_entry. */ |
| 119 |
struct swap_entries { |
| 120 |
struct ordered_hashmap_entry e[_IDX_SWAP_END - _IDX_SWAP_BEGIN]; |
| 121 |
}; |
| 122 |
|
| 123 |
/* Distance from Initial Bucket */ |
| 124 |
typedef uint8_t dib_raw_t; |
| 125 |
#define DIB_RAW_OVERFLOW ((dib_raw_t)0xfdU) /* indicates DIB value is greater than representable */ |
| 126 |
#define DIB_RAW_REHASH ((dib_raw_t)0xfeU) /* entry yet to be rehashed during in-place resize */ |
| 127 |
#define DIB_RAW_FREE ((dib_raw_t)0xffU) /* a free bucket */ |
| 128 |
#define DIB_RAW_INIT ((char)DIB_RAW_FREE) /* a byte to memset a DIB store with when initializing */ |
| 129 |
|
| 130 |
#define DIB_FREE UINT_MAX |
| 131 |
|
| 132 |
#if ENABLE_DEBUG_HASHMAP |
| 133 |
struct hashmap_debug_info { |
| 134 |
LIST_FIELDS(struct hashmap_debug_info, debug_list); |
| 135 |
unsigned max_entries; /* high watermark of n_entries */ |
| 136 |
|
| 137 |
/* fields to detect modification while iterating */ |
| 138 |
unsigned put_count; /* counts puts into the hashmap */ |
| 139 |
unsigned rem_count; /* counts removals from hashmap */ |
| 140 |
unsigned last_rem_idx; /* remembers last removal index */ |
| 141 |
}; |
| 142 |
|
| 143 |
/* Tracks all existing hashmaps. Get at it from gdb. See sd_dump_hashmaps.py */ |
| 144 |
static LIST_HEAD(struct hashmap_debug_info, hashmap_debug_list); |
| 145 |
static pthread_mutex_t hashmap_debug_list_mutex = PTHREAD_MUTEX_INITIALIZER; |
| 146 |
#endif |
| 147 |
|
| 148 |
enum HashmapType { |
| 149 |
HASHMAP_TYPE_PLAIN, |
| 150 |
HASHMAP_TYPE_ORDERED, |
| 151 |
HASHMAP_TYPE_SET, |
| 152 |
_HASHMAP_TYPE_MAX |
| 153 |
}; |
| 154 |
|
| 155 |
struct _packed_ indirect_storage { |
| 156 |
void *storage; /* where buckets and DIBs are stored */ |
| 157 |
uint8_t hash_key[HASH_KEY_SIZE]; /* hash key; changes during resize */ |
| 158 |
|
| 159 |
unsigned n_entries; /* number of stored entries */ |
| 160 |
unsigned n_buckets; /* number of buckets */ |
| 161 |
|
| 162 |
unsigned idx_lowest_entry; /* Index below which all buckets are free. |
| 163 |
Makes "while (hashmap_steal_first())" loops |
| 164 |
O(n) instead of O(n^2) for unordered hashmaps. */ |
| 165 |
uint8_t _pad[3]; /* padding for the whole HashmapBase */ |
| 166 |
/* The bitfields in HashmapBase complete the alignment of the whole thing. */ |
| 167 |
}; |
| 168 |
|
| 169 |
struct direct_storage { |
| 170 |
/* This gives us 39 bytes on 64-bit, or 35 bytes on 32-bit. |
| 171 |
* That's room for 4 set_entries + 4 DIB bytes + 3 unused bytes on 64-bit, |
| 172 |
* or 7 set_entries + 7 DIB bytes + 0 unused bytes on 32-bit. */ |
| 173 |
uint8_t storage[sizeof(struct indirect_storage)]; |
| 174 |
}; |
| 175 |
|
| 176 |
#define DIRECT_BUCKETS(entry_t) \ |
| 177 |
(sizeof(struct direct_storage) / (sizeof(entry_t) + sizeof(dib_raw_t))) |
| 178 |
|
| 179 |
/* We should be able to store at least one entry directly. */ |
| 180 |
assert_cc(DIRECT_BUCKETS(struct ordered_hashmap_entry) >= 1); |
| 181 |
|
| 182 |
/* We have 3 bits for n_direct_entries. */ |
| 183 |
assert_cc(DIRECT_BUCKETS(struct set_entry) < (1 << 3)); |
| 184 |
|
| 185 |
/* Hashmaps with directly stored entries all use this shared hash key. |
| 186 |
* It's no big deal if the key is guessed, because there can be only |
| 187 |
* a handful of directly stored entries in a hashmap. When a hashmap |
| 188 |
* outgrows direct storage, it gets its own key for indirect storage. */ |
| 189 |
static uint8_t shared_hash_key[HASH_KEY_SIZE]; |
| 190 |
|
| 191 |
/* Fields that all hashmap/set types must have */ |
| 192 |
struct HashmapBase { |
| 193 |
const struct hash_ops *hash_ops; /* hash and compare ops to use */ |
| 194 |
|
| 195 |
union _packed_ { |
| 196 |
struct indirect_storage indirect; /* if has_indirect */ |
| 197 |
struct direct_storage direct; /* if !has_indirect */ |
| 198 |
}; |
| 199 |
|
| 200 |
enum HashmapType type:2; /* HASHMAP_TYPE_* */ |
| 201 |
bool has_indirect:1; /* whether indirect storage is used */ |
| 202 |
unsigned n_direct_entries:3; /* Number of entries in direct storage. |
| 203 |
* Only valid if !has_indirect. */ |
| 204 |
bool from_pool:1; /* whether was allocated from mempool */ |
| 205 |
bool dirty:1; /* whether dirtied since last iterated_cache_get() */ |
| 206 |
bool cached:1; /* whether this hashmap is being cached */ |
| 207 |
|
| 208 |
#if ENABLE_DEBUG_HASHMAP |
| 209 |
struct hashmap_debug_info debug; |
| 210 |
#endif |
| 211 |
}; |
| 212 |
|
| 213 |
/* Specific hash types |
| 214 |
* HashmapBase must be at the beginning of each hashmap struct. */ |
| 215 |
|
| 216 |
struct Hashmap { |
| 217 |
struct HashmapBase b; |
| 218 |
}; |
| 219 |
|
| 220 |
struct OrderedHashmap { |
| 221 |
struct HashmapBase b; |
| 222 |
unsigned iterate_list_head, iterate_list_tail; |
| 223 |
}; |
| 224 |
|
| 225 |
struct Set { |
| 226 |
struct HashmapBase b; |
| 227 |
}; |
| 228 |
|
| 229 |
typedef struct CacheMem { |
| 230 |
const void **ptr; |
| 231 |
size_t n_populated; |
| 232 |
bool active:1; |
| 233 |
} CacheMem; |
| 234 |
|
| 235 |
struct IteratedCache { |
| 236 |
HashmapBase *hashmap; |
| 237 |
CacheMem keys, values; |
| 238 |
}; |
| 239 |
|
| 240 |
DEFINE_MEMPOOL(hashmap_pool, Hashmap, 8); |
| 241 |
DEFINE_MEMPOOL(ordered_hashmap_pool, OrderedHashmap, 8); |
| 242 |
/* No need for a separate Set pool */ |
| 243 |
assert_cc(sizeof(Hashmap) == sizeof(Set)); |
| 244 |
|
| 245 |
struct hashmap_type_info { |
| 246 |
size_t head_size; |
| 247 |
size_t entry_size; |
| 248 |
struct mempool *mempool; |
| 249 |
unsigned n_direct_buckets; |
| 250 |
}; |
| 251 |
|
| 252 |
static _used_ const struct hashmap_type_info hashmap_type_info[_HASHMAP_TYPE_MAX] = { |
| 253 |
[HASHMAP_TYPE_PLAIN] = { |
| 254 |
.head_size = sizeof(Hashmap), |
| 255 |
.entry_size = sizeof(struct plain_hashmap_entry), |
| 256 |
.mempool = &hashmap_pool, |
| 257 |
.n_direct_buckets = DIRECT_BUCKETS(struct plain_hashmap_entry), |
| 258 |
}, |
| 259 |
[HASHMAP_TYPE_ORDERED] = { |
| 260 |
.head_size = sizeof(OrderedHashmap), |
| 261 |
.entry_size = sizeof(struct ordered_hashmap_entry), |
| 262 |
.mempool = &ordered_hashmap_pool, |
| 263 |
.n_direct_buckets = DIRECT_BUCKETS(struct ordered_hashmap_entry), |
| 264 |
}, |
| 265 |
[HASHMAP_TYPE_SET] = { |
| 266 |
.head_size = sizeof(Set), |
| 267 |
.entry_size = sizeof(struct set_entry), |
| 268 |
.mempool = &hashmap_pool, |
| 269 |
.n_direct_buckets = DIRECT_BUCKETS(struct set_entry), |
| 270 |
}, |
| 271 |
}; |
| 272 |
|
| 273 |
void hashmap_trim_pools(void) { |
| 274 |
int r; |
| 275 |
|
| 276 |
/* The pool is only allocated by the main thread, but the memory can be passed to other |
| 277 |
* threads. Let's clean up if we are the main thread and no other threads are live. */ |
| 278 |
|
| 279 |
/* We build our own is_main_thread() here, which doesn't use C11 TLS based caching of the |
| 280 |
* result. That's because valgrind apparently doesn't like TLS to be used from a GCC destructor. */ |
| 281 |
if (getpid() != gettid()) |
| 282 |
return (void) log_debug("Not cleaning up memory pools, not in main thread."); |
| 283 |
|
| 284 |
r = get_process_threads(0); |
| 285 |
if (r < 0) |
| 286 |
return (void) log_debug_errno(r, "Failed to determine number of threads, not cleaning up memory pools: %m"); |
| 287 |
if (r != 1) |
| 288 |
return (void) log_debug("Not cleaning up memory pools, running in multi-threaded process."); |
| 289 |
|
| 290 |
mempool_trim(&hashmap_pool); |
| 291 |
mempool_trim(&ordered_hashmap_pool); |
| 292 |
} |
| 293 |
|
| 294 |
#if HAVE_VALGRIND_VALGRIND_H |
| 295 |
_destructor_ static void cleanup_pools(void) { |
| 296 |
/* Be nice to valgrind */ |
| 297 |
if (RUNNING_ON_VALGRIND) |
| 298 |
hashmap_trim_pools(); |
| 299 |
} |
| 300 |
#endif |
| 301 |
|
| 302 |
static unsigned n_buckets(HashmapBase *h) { |
| 303 |
return h->has_indirect ? h->indirect.n_buckets |
| 304 |
: hashmap_type_info[h->type].n_direct_buckets; |
| 305 |
} |
| 306 |
|
| 307 |
static unsigned n_entries(HashmapBase *h) { |
| 308 |
return h->has_indirect ? h->indirect.n_entries |
| 309 |
: h->n_direct_entries; |
| 310 |
} |
| 311 |
|
| 312 |
static void n_entries_inc(HashmapBase *h) { |
| 313 |
if (h->has_indirect) |
| 314 |
h->indirect.n_entries++; |
| 315 |
else |
| 316 |
h->n_direct_entries++; |
| 317 |
} |
| 318 |
|
| 319 |
static void n_entries_dec(HashmapBase *h) { |
| 320 |
if (h->has_indirect) |
| 321 |
h->indirect.n_entries--; |
| 322 |
else |
| 323 |
h->n_direct_entries--; |
| 324 |
} |
| 325 |
|
| 326 |
static void* storage_ptr(HashmapBase *h) { |
| 327 |
return h->has_indirect ? h->indirect.storage |
| 328 |
: h->direct.storage; |
| 329 |
} |
| 330 |
|
| 331 |
static uint8_t* hash_key(HashmapBase *h) { |
| 332 |
return h->has_indirect ? h->indirect.hash_key |
| 333 |
: shared_hash_key; |
| 334 |
} |
| 335 |
|
| 336 |
static unsigned base_bucket_hash(HashmapBase *h, const void *p) { |
| 337 |
struct siphash state; |
| 338 |
uint64_t hash; |
| 339 |
|
| 340 |
siphash24_init(&state, hash_key(h)); |
| 341 |
|
| 342 |
h->hash_ops->hash(p, &state); |
| 343 |
|
| 344 |
hash = siphash24_finalize(&state); |
| 345 |
|
| 346 |
return (unsigned) (hash % n_buckets(h)); |
| 347 |
} |
| 348 |
#define bucket_hash(h, p) base_bucket_hash(HASHMAP_BASE(h), p) |
| 349 |
|
| 350 |
static void base_set_dirty(HashmapBase *h) { |
| 351 |
h->dirty = true; |
| 352 |
} |
| 353 |
#define hashmap_set_dirty(h) base_set_dirty(HASHMAP_BASE(h)) |
| 354 |
|
| 355 |
static void get_hash_key(uint8_t hash_key[HASH_KEY_SIZE], bool reuse_is_ok) { |
| 356 |
static uint8_t current[HASH_KEY_SIZE]; |
| 357 |
static bool current_initialized = false; |
| 358 |
|
| 359 |
/* Returns a hash function key to use. In order to keep things |
| 360 |
* fast we will not generate a new key each time we allocate a |
| 361 |
* new hash table. Instead, we'll just reuse the most recently |
| 362 |
* generated one, except if we never generated one or when we |
| 363 |
* are rehashing an entire hash table because we reached a |
| 364 |
* fill level */ |
| 365 |
|
| 366 |
if (!current_initialized || !reuse_is_ok) { |
| 367 |
random_bytes(current, sizeof(current)); |
| 368 |
current_initialized = true; |
| 369 |
} |
| 370 |
|
| 371 |
memcpy(hash_key, current, sizeof(current)); |
| 372 |
} |
| 373 |
|
| 374 |
static struct hashmap_base_entry* bucket_at(HashmapBase *h, unsigned idx) { |
| 375 |
return CAST_ALIGN_PTR( |
| 376 |
struct hashmap_base_entry, |
| 377 |
(uint8_t *) storage_ptr(h) + idx * hashmap_type_info[h->type].entry_size); |
| 378 |
} |
| 379 |
|
| 380 |
static struct plain_hashmap_entry* plain_bucket_at(Hashmap *h, unsigned idx) { |
| 381 |
return (struct plain_hashmap_entry*) bucket_at(HASHMAP_BASE(h), idx); |
| 382 |
} |
| 383 |
|
| 384 |
static struct ordered_hashmap_entry* ordered_bucket_at(OrderedHashmap *h, unsigned idx) { |
| 385 |
return (struct ordered_hashmap_entry*) bucket_at(HASHMAP_BASE(h), idx); |
| 386 |
} |
| 387 |
|
| 388 |
static struct set_entry *set_bucket_at(Set *h, unsigned idx) { |
| 389 |
return (struct set_entry*) bucket_at(HASHMAP_BASE(h), idx); |
| 390 |
} |
| 391 |
|
| 392 |
static struct ordered_hashmap_entry* bucket_at_swap(struct swap_entries *swap, unsigned idx) { |
| 393 |
return &swap->e[idx - _IDX_SWAP_BEGIN]; |
| 394 |
} |
| 395 |
|
| 396 |
/* Returns a pointer to the bucket at index idx. |
| 397 |
* Understands real indexes and swap indexes, hence "_virtual". */ |
| 398 |
static struct hashmap_base_entry* bucket_at_virtual(HashmapBase *h, struct swap_entries *swap, |
| 399 |
unsigned idx) { |
| 400 |
if (idx < _IDX_SWAP_BEGIN) |
| 401 |
return bucket_at(h, idx); |
| 402 |
|
| 403 |
if (idx < _IDX_SWAP_END) |
| 404 |
return &bucket_at_swap(swap, idx)->p.b; |
| 405 |
|
| 406 |
assert_not_reached(); |
| 407 |
} |
| 408 |
|
| 409 |
static dib_raw_t* dib_raw_ptr(HashmapBase *h) { |
| 410 |
return (dib_raw_t*) |
| 411 |
((uint8_t*) storage_ptr(h) + hashmap_type_info[h->type].entry_size * n_buckets(h)); |
| 412 |
} |
| 413 |
|
| 414 |
static unsigned bucket_distance(HashmapBase *h, unsigned idx, unsigned from) { |
| 415 |
return idx >= from ? idx - from |
| 416 |
: n_buckets(h) + idx - from; |
| 417 |
} |
| 418 |
|
| 419 |
static unsigned bucket_calculate_dib(HashmapBase *h, unsigned idx, dib_raw_t raw_dib) { |
| 420 |
unsigned initial_bucket; |
| 421 |
|
| 422 |
if (raw_dib == DIB_RAW_FREE) |
| 423 |
return DIB_FREE; |
| 424 |
|
| 425 |
if (_likely_(raw_dib < DIB_RAW_OVERFLOW)) |
| 426 |
return raw_dib; |
| 427 |
|
| 428 |
/* |
| 429 |
* Having an overflow DIB value is very unlikely. The hash function |
| 430 |
* would have to be bad. For example, in a table of size 2^24 filled |
| 431 |
* to load factor 0.9 the maximum observed DIB is only about 60. |
| 432 |
* In theory (assuming I used Maxima correctly), for an infinite size |
| 433 |
* hash table with load factor 0.8 the probability of a given entry |
| 434 |
* having DIB > 40 is 1.9e-8. |
| 435 |
* This returns the correct DIB value by recomputing the hash value in |
| 436 |
* the unlikely case. XXX Hitting this case could be a hint to rehash. |
| 437 |
*/ |
| 438 |
initial_bucket = bucket_hash(h, bucket_at(h, idx)->key); |
| 439 |
return bucket_distance(h, idx, initial_bucket); |
| 440 |
} |
| 441 |
|
| 442 |
static void bucket_set_dib(HashmapBase *h, unsigned idx, unsigned dib) { |
| 443 |
dib_raw_ptr(h)[idx] = dib != DIB_FREE ? MIN(dib, DIB_RAW_OVERFLOW) : DIB_RAW_FREE; |
| 444 |
} |
| 445 |
|
| 446 |
static unsigned skip_free_buckets(HashmapBase *h, unsigned idx) { |
| 447 |
dib_raw_t *dibs; |
| 448 |
|
| 449 |
dibs = dib_raw_ptr(h); |
| 450 |
|
| 451 |
for ( ; idx < n_buckets(h); idx++) |
| 452 |
if (dibs[idx] != DIB_RAW_FREE) |
| 453 |
return idx; |
| 454 |
|
| 455 |
return IDX_NIL; |
| 456 |
} |
| 457 |
|
| 458 |
static void bucket_mark_free(HashmapBase *h, unsigned idx) { |
| 459 |
memzero(bucket_at(h, idx), hashmap_type_info[h->type].entry_size); |
| 460 |
bucket_set_dib(h, idx, DIB_FREE); |
| 461 |
} |
| 462 |
|
| 463 |
static void bucket_move_entry(HashmapBase *h, struct swap_entries *swap, |
| 464 |
unsigned from, unsigned to) { |
| 465 |
struct hashmap_base_entry *e_from, *e_to; |
| 466 |
|
| 467 |
assert(from != to); |
| 468 |
|
| 469 |
e_from = bucket_at_virtual(h, swap, from); |
| 470 |
e_to = bucket_at_virtual(h, swap, to); |
| 471 |
|
| 472 |
memcpy(e_to, e_from, hashmap_type_info[h->type].entry_size); |
| 473 |
|
| 474 |
if (h->type == HASHMAP_TYPE_ORDERED) { |
| 475 |
OrderedHashmap *lh = (OrderedHashmap*) h; |
| 476 |
struct ordered_hashmap_entry *le, *le_to; |
| 477 |
|
| 478 |
le_to = (struct ordered_hashmap_entry*) e_to; |
| 479 |
|
| 480 |
if (le_to->iterate_next != IDX_NIL) { |
| 481 |
le = (struct ordered_hashmap_entry*) |
| 482 |
bucket_at_virtual(h, swap, le_to->iterate_next); |
| 483 |
le->iterate_previous = to; |
| 484 |
} |
| 485 |
|
| 486 |
if (le_to->iterate_previous != IDX_NIL) { |
| 487 |
le = (struct ordered_hashmap_entry*) |
| 488 |
bucket_at_virtual(h, swap, le_to->iterate_previous); |
| 489 |
le->iterate_next = to; |
| 490 |
} |
| 491 |
|
| 492 |
if (lh->iterate_list_head == from) |
| 493 |
lh->iterate_list_head = to; |
| 494 |
if (lh->iterate_list_tail == from) |
| 495 |
lh->iterate_list_tail = to; |
| 496 |
} |
| 497 |
} |
| 498 |
|
| 499 |
static unsigned next_idx(HashmapBase *h, unsigned idx) { |
| 500 |
return (idx + 1U) % n_buckets(h); |
| 501 |
} |
| 502 |
|
| 503 |
static unsigned prev_idx(HashmapBase *h, unsigned idx) { |
| 504 |
return (n_buckets(h) + idx - 1U) % n_buckets(h); |
| 505 |
} |
| 506 |
|
| 507 |
static void* entry_value(HashmapBase *h, struct hashmap_base_entry *e) { |
| 508 |
switch (h->type) { |
| 509 |
|
| 510 |
case HASHMAP_TYPE_PLAIN: |
| 511 |
case HASHMAP_TYPE_ORDERED: |
| 512 |
return ((struct plain_hashmap_entry*)e)->value; |
| 513 |
|
| 514 |
case HASHMAP_TYPE_SET: |
| 515 |
return (void*) e->key; |
| 516 |
|
| 517 |
default: |
| 518 |
assert_not_reached(); |
| 519 |
} |
| 520 |
} |
| 521 |
|
| 522 |
static void base_remove_entry(HashmapBase *h, unsigned idx) { |
| 523 |
unsigned left, right, prev, dib; |
| 524 |
dib_raw_t raw_dib, *dibs; |
| 525 |
|
| 526 |
dibs = dib_raw_ptr(h); |
| 527 |
assert(dibs[idx] != DIB_RAW_FREE); |
| 528 |
|
| 529 |
#if ENABLE_DEBUG_HASHMAP |
| 530 |
h->debug.rem_count++; |
| 531 |
h->debug.last_rem_idx = idx; |
| 532 |
#endif |
| 533 |
|
| 534 |
left = idx; |
| 535 |
/* Find the stop bucket ("right"). It is either free or has DIB == 0. */ |
| 536 |
for (right = next_idx(h, left); ; right = next_idx(h, right)) { |
| 537 |
raw_dib = dibs[right]; |
| 538 |
if (IN_SET(raw_dib, 0, DIB_RAW_FREE)) |
| 539 |
break; |
| 540 |
|
| 541 |
/* The buckets are not supposed to be all occupied and with DIB > 0. |
| 542 |
* That would mean we could make everyone better off by shifting them |
| 543 |
* backward. This scenario is impossible. */ |
| 544 |
assert(left != right); |
| 545 |
} |
| 546 |
|
| 547 |
if (h->type == HASHMAP_TYPE_ORDERED) { |
| 548 |
OrderedHashmap *lh = (OrderedHashmap*) h; |
| 549 |
struct ordered_hashmap_entry *le = ordered_bucket_at(lh, idx); |
| 550 |
|
| 551 |
if (le->iterate_next != IDX_NIL) |
| 552 |
ordered_bucket_at(lh, le->iterate_next)->iterate_previous = le->iterate_previous; |
| 553 |
else |
| 554 |
lh->iterate_list_tail = le->iterate_previous; |
| 555 |
|
| 556 |
if (le->iterate_previous != IDX_NIL) |
| 557 |
ordered_bucket_at(lh, le->iterate_previous)->iterate_next = le->iterate_next; |
| 558 |
else |
| 559 |
lh->iterate_list_head = le->iterate_next; |
| 560 |
} |
| 561 |
|
| 562 |
/* Now shift all buckets in the interval (left, right) one step backwards */ |
| 563 |
for (prev = left, left = next_idx(h, left); left != right; |
| 564 |
prev = left, left = next_idx(h, left)) { |
| 565 |
dib = bucket_calculate_dib(h, left, dibs[left]); |
| 566 |
assert(dib != 0); |
| 567 |
bucket_move_entry(h, NULL, left, prev); |
| 568 |
bucket_set_dib(h, prev, dib - 1); |
| 569 |
} |
| 570 |
|
| 571 |
bucket_mark_free(h, prev); |
| 572 |
n_entries_dec(h); |
| 573 |
base_set_dirty(h); |
| 574 |
} |
| 575 |
#define remove_entry(h, idx) base_remove_entry(HASHMAP_BASE(h), idx) |
| 576 |
|
| 577 |
static unsigned hashmap_iterate_in_insertion_order(OrderedHashmap *h, Iterator *i) { |
| 578 |
struct ordered_hashmap_entry *e; |
| 579 |
unsigned idx; |
| 580 |
|
| 581 |
assert(h); |
| 582 |
assert(i); |
| 583 |
|
| 584 |
if (i->idx == IDX_NIL) |
| 585 |
goto at_end; |
| 586 |
|
| 587 |
if (i->idx == IDX_FIRST && h->iterate_list_head == IDX_NIL) |
| 588 |
goto at_end; |
| 589 |
|
| 590 |
if (i->idx == IDX_FIRST) { |
| 591 |
idx = h->iterate_list_head; |
| 592 |
e = ordered_bucket_at(h, idx); |
| 593 |
} else { |
| 594 |
idx = i->idx; |
| 595 |
e = ordered_bucket_at(h, idx); |
| 596 |
/* |
| 597 |
* We allow removing the current entry while iterating, but removal may cause |
| 598 |
* a backward shift. The next entry may thus move one bucket to the left. |
| 599 |
* To detect when it happens, we remember the key pointer of the entry we were |
| 600 |
* going to iterate next. If it does not match, there was a backward shift. |
| 601 |
*/ |
| 602 |
if (e->p.b.key != i->next_key) { |
| 603 |
idx = prev_idx(HASHMAP_BASE(h), idx); |
| 604 |
e = ordered_bucket_at(h, idx); |
| 605 |
} |
| 606 |
assert(e->p.b.key == i->next_key); |
| 607 |
} |
| 608 |
|
| 609 |
#if ENABLE_DEBUG_HASHMAP |
| 610 |
i->prev_idx = idx; |
| 611 |
#endif |
| 612 |
|
| 613 |
if (e->iterate_next != IDX_NIL) { |
| 614 |
struct ordered_hashmap_entry *n; |
| 615 |
i->idx = e->iterate_next; |
| 616 |
n = ordered_bucket_at(h, i->idx); |
| 617 |
i->next_key = n->p.b.key; |
| 618 |
} else |
| 619 |
i->idx = IDX_NIL; |
| 620 |
|
| 621 |
return idx; |
| 622 |
|
| 623 |
at_end: |
| 624 |
i->idx = IDX_NIL; |
| 625 |
return IDX_NIL; |
| 626 |
} |
| 627 |
|
| 628 |
static unsigned hashmap_iterate_in_internal_order(HashmapBase *h, Iterator *i) { |
| 629 |
unsigned idx; |
| 630 |
|
| 631 |
assert(h); |
| 632 |
assert(i); |
| 633 |
|
| 634 |
if (i->idx == IDX_NIL) |
| 635 |
goto at_end; |
| 636 |
|
| 637 |
if (i->idx == IDX_FIRST) { |
| 638 |
/* fast forward to the first occupied bucket */ |
| 639 |
if (h->has_indirect) { |
| 640 |
i->idx = skip_free_buckets(h, h->indirect.idx_lowest_entry); |
| 641 |
h->indirect.idx_lowest_entry = i->idx; |
| 642 |
} else |
| 643 |
i->idx = skip_free_buckets(h, 0); |
| 644 |
|
| 645 |
if (i->idx == IDX_NIL) |
| 646 |
goto at_end; |
| 647 |
} else { |
| 648 |
struct hashmap_base_entry *e; |
| 649 |
|
| 650 |
assert(i->idx > 0); |
| 651 |
|
| 652 |
e = bucket_at(h, i->idx); |
| 653 |
/* |
| 654 |
* We allow removing the current entry while iterating, but removal may cause |
| 655 |
* a backward shift. The next entry may thus move one bucket to the left. |
| 656 |
* To detect when it happens, we remember the key pointer of the entry we were |
| 657 |
* going to iterate next. If it does not match, there was a backward shift. |
| 658 |
*/ |
| 659 |
if (e->key != i->next_key) |
| 660 |
e = bucket_at(h, --i->idx); |
| 661 |
|
| 662 |
assert(e->key == i->next_key); |
| 663 |
} |
| 664 |
|
| 665 |
idx = i->idx; |
| 666 |
#if ENABLE_DEBUG_HASHMAP |
| 667 |
i->prev_idx = idx; |
| 668 |
#endif |
| 669 |
|
| 670 |
i->idx = skip_free_buckets(h, i->idx + 1); |
| 671 |
if (i->idx != IDX_NIL) |
| 672 |
i->next_key = bucket_at(h, i->idx)->key; |
| 673 |
else |
| 674 |
i->idx = IDX_NIL; |
| 675 |
|
| 676 |
return idx; |
| 677 |
|
| 678 |
at_end: |
| 679 |
i->idx = IDX_NIL; |
| 680 |
return IDX_NIL; |
| 681 |
} |
| 682 |
|
| 683 |
static unsigned hashmap_iterate_entry(HashmapBase *h, Iterator *i) { |
| 684 |
if (!h) { |
| 685 |
i->idx = IDX_NIL; |
| 686 |
return IDX_NIL; |
| 687 |
} |
| 688 |
|
| 689 |
#if ENABLE_DEBUG_HASHMAP |
| 690 |
if (i->idx == IDX_FIRST) { |
| 691 |
i->put_count = h->debug.put_count; |
| 692 |
i->rem_count = h->debug.rem_count; |
| 693 |
} else { |
| 694 |
/* While iterating, must not add any new entries */ |
| 695 |
assert(i->put_count == h->debug.put_count); |
| 696 |
/* ... or remove entries other than the current one */ |
| 697 |
assert(i->rem_count == h->debug.rem_count || |
| 698 |
(i->rem_count == h->debug.rem_count - 1 && |
| 699 |
i->prev_idx == h->debug.last_rem_idx)); |
| 700 |
/* Reset our removals counter */ |
| 701 |
i->rem_count = h->debug.rem_count; |
| 702 |
} |
| 703 |
#endif |
| 704 |
|
| 705 |
return h->type == HASHMAP_TYPE_ORDERED ? hashmap_iterate_in_insertion_order((OrderedHashmap*) h, i) |
| 706 |
: hashmap_iterate_in_internal_order(h, i); |
| 707 |
} |
| 708 |
|
| 709 |
bool _hashmap_iterate(HashmapBase *h, Iterator *i, void **value, const void **key) { |
| 710 |
struct hashmap_base_entry *e; |
| 711 |
void *data; |
| 712 |
unsigned idx; |
| 713 |
|
| 714 |
idx = hashmap_iterate_entry(h, i); |
| 715 |
if (idx == IDX_NIL) { |
| 716 |
if (value) |
| 717 |
*value = NULL; |
| 718 |
if (key) |
| 719 |
*key = NULL; |
| 720 |
|
| 721 |
return false; |
| 722 |
} |
| 723 |
|
| 724 |
e = bucket_at(h, idx); |
| 725 |
data = entry_value(h, e); |
| 726 |
if (value) |
| 727 |
*value = data; |
| 728 |
if (key) |
| 729 |
*key = e->key; |
| 730 |
|
| 731 |
return true; |
| 732 |
} |
| 733 |
|
| 734 |
#define HASHMAP_FOREACH_IDX(idx, h, i) \ |
| 735 |
for ((i) = ITERATOR_FIRST, (idx) = hashmap_iterate_entry((h), &(i)); \ |
| 736 |
(idx != IDX_NIL); \ |
| 737 |
(idx) = hashmap_iterate_entry((h), &(i))) |
| 738 |
|
| 739 |
IteratedCache* _hashmap_iterated_cache_new(HashmapBase *h) { |
| 740 |
IteratedCache *cache; |
| 741 |
|
| 742 |
assert(h); |
| 743 |
assert(!h->cached); |
| 744 |
|
| 745 |
if (h->cached) |
| 746 |
return NULL; |
| 747 |
|
| 748 |
cache = new0(IteratedCache, 1); |
| 749 |
if (!cache) |
| 750 |
return NULL; |
| 751 |
|
| 752 |
cache->hashmap = h; |
| 753 |
h->cached = true; |
| 754 |
|
| 755 |
return cache; |
| 756 |
} |
| 757 |
|
| 758 |
static void reset_direct_storage(HashmapBase *h) { |
| 759 |
const struct hashmap_type_info *hi = &hashmap_type_info[h->type]; |
| 760 |
void *p; |
| 761 |
|
| 762 |
assert(!h->has_indirect); |
| 763 |
|
| 764 |
p = mempset(h->direct.storage, 0, hi->entry_size * hi->n_direct_buckets); |
| 765 |
memset(p, DIB_RAW_INIT, sizeof(dib_raw_t) * hi->n_direct_buckets); |
| 766 |
} |
| 767 |
|
| 768 |
static void shared_hash_key_initialize(void) { |
| 769 |
random_bytes(shared_hash_key, sizeof(shared_hash_key)); |
| 770 |
} |
| 771 |
|
| 772 |
static struct HashmapBase* hashmap_base_new(const struct hash_ops *hash_ops, enum HashmapType type) { |
| 773 |
HashmapBase *h; |
| 774 |
const struct hashmap_type_info *hi = &hashmap_type_info[type]; |
| 775 |
|
| 776 |
bool use_pool = mempool_enabled && mempool_enabled(); /* mempool_enabled is a weak symbol */ |
| 777 |
|
| 778 |
h = use_pool ? mempool_alloc0_tile(hi->mempool) : malloc0(hi->head_size); |
| 779 |
if (!h) |
| 780 |
return NULL; |
| 781 |
|
| 782 |
h->type = type; |
| 783 |
h->from_pool = use_pool; |
| 784 |
h->hash_ops = hash_ops ?: &trivial_hash_ops; |
| 785 |
|
| 786 |
if (type == HASHMAP_TYPE_ORDERED) { |
| 787 |
OrderedHashmap *lh = (OrderedHashmap*)h; |
| 788 |
lh->iterate_list_head = lh->iterate_list_tail = IDX_NIL; |
| 789 |
} |
| 790 |
|
| 791 |
reset_direct_storage(h); |
| 792 |
|
| 793 |
static pthread_once_t once = PTHREAD_ONCE_INIT; |
| 794 |
assert_se(pthread_once(&once, shared_hash_key_initialize) == 0); |
| 795 |
|
| 796 |
#if ENABLE_DEBUG_HASHMAP |
| 797 |
assert_se(pthread_mutex_lock(&hashmap_debug_list_mutex) == 0); |
| 798 |
LIST_PREPEND(debug_list, hashmap_debug_list, &h->debug); |
| 799 |
assert_se(pthread_mutex_unlock(&hashmap_debug_list_mutex) == 0); |
| 800 |
#endif |
| 801 |
|
| 802 |
return h; |
| 803 |
} |
| 804 |
|
| 805 |
Hashmap *hashmap_new(const struct hash_ops *hash_ops) { |
| 806 |
return (Hashmap*) hashmap_base_new(hash_ops, HASHMAP_TYPE_PLAIN); |
| 807 |
} |
| 808 |
|
| 809 |
OrderedHashmap *ordered_hashmap_new(const struct hash_ops *hash_ops) { |
| 810 |
return (OrderedHashmap*) hashmap_base_new(hash_ops, HASHMAP_TYPE_ORDERED); |
| 811 |
} |
| 812 |
|
| 813 |
Set *set_new(const struct hash_ops *hash_ops) { |
| 814 |
return (Set*) hashmap_base_new(hash_ops, HASHMAP_TYPE_SET); |
| 815 |
} |
| 816 |
|
| 817 |
static int hashmap_base_ensure_allocated(HashmapBase **h, const struct hash_ops *hash_ops, |
| 818 |
enum HashmapType type) { |
| 819 |
HashmapBase *q; |
| 820 |
|
| 821 |
assert(h); |
| 822 |
|
| 823 |
if (*h) { |
| 824 |
assert((*h)->hash_ops == (hash_ops ?: &trivial_hash_ops)); |
| 825 |
return 0; |
| 826 |
} |
| 827 |
|
| 828 |
q = hashmap_base_new(hash_ops, type); |
| 829 |
if (!q) |
| 830 |
return -ENOMEM; |
| 831 |
|
| 832 |
*h = q; |
| 833 |
return 1; |
| 834 |
} |
| 835 |
|
| 836 |
int hashmap_ensure_allocated(Hashmap **h, const struct hash_ops *hash_ops) { |
| 837 |
return hashmap_base_ensure_allocated((HashmapBase**)h, hash_ops, HASHMAP_TYPE_PLAIN); |
| 838 |
} |
| 839 |
|
| 840 |
int ordered_hashmap_ensure_allocated(OrderedHashmap **h, const struct hash_ops *hash_ops) { |
| 841 |
return hashmap_base_ensure_allocated((HashmapBase**)h, hash_ops, HASHMAP_TYPE_ORDERED); |
| 842 |
} |
| 843 |
|
| 844 |
int set_ensure_allocated(Set **s, const struct hash_ops *hash_ops) { |
| 845 |
return hashmap_base_ensure_allocated((HashmapBase**)s, hash_ops, HASHMAP_TYPE_SET); |
| 846 |
} |
| 847 |
|
| 848 |
int hashmap_ensure_put(Hashmap **h, const struct hash_ops *hash_ops, const void *key, void *value) { |
| 849 |
int r; |
| 850 |
|
| 851 |
assert(h); |
| 852 |
|
| 853 |
r = hashmap_ensure_allocated(h, hash_ops); |
| 854 |
if (r < 0) |
| 855 |
return r; |
| 856 |
|
| 857 |
return hashmap_put(*h, key, value); |
| 858 |
} |
| 859 |
|
| 860 |
int ordered_hashmap_ensure_put(OrderedHashmap **h, const struct hash_ops *hash_ops, const void *key, void *value) { |
| 861 |
int r; |
| 862 |
|
| 863 |
assert(h); |
| 864 |
|
| 865 |
r = ordered_hashmap_ensure_allocated(h, hash_ops); |
| 866 |
if (r < 0) |
| 867 |
return r; |
| 868 |
|
| 869 |
return ordered_hashmap_put(*h, key, value); |
| 870 |
} |
| 871 |
|
| 872 |
int ordered_hashmap_ensure_replace(OrderedHashmap **h, const struct hash_ops *hash_ops, const void *key, void *value) { |
| 873 |
int r; |
| 874 |
|
| 875 |
assert(h); |
| 876 |
|
| 877 |
r = ordered_hashmap_ensure_allocated(h, hash_ops); |
| 878 |
if (r < 0) |
| 879 |
return r; |
| 880 |
|
| 881 |
return ordered_hashmap_replace(*h, key, value); |
| 882 |
} |
| 883 |
|
| 884 |
int hashmap_ensure_replace(Hashmap **h, const struct hash_ops *hash_ops, const void *key, void *value) { |
| 885 |
int r; |
| 886 |
|
| 887 |
assert(h); |
| 888 |
|
| 889 |
r = hashmap_ensure_allocated(h, hash_ops); |
| 890 |
if (r < 0) |
| 891 |
return r; |
| 892 |
|
| 893 |
return hashmap_replace(*h, key, value); |
| 894 |
} |
| 895 |
|
| 896 |
static void hashmap_free_no_clear(HashmapBase *h) { |
| 897 |
assert(!h->has_indirect); |
| 898 |
assert(h->n_direct_entries == 0); |
| 899 |
|
| 900 |
#if ENABLE_DEBUG_HASHMAP |
| 901 |
assert_se(pthread_mutex_lock(&hashmap_debug_list_mutex) == 0); |
| 902 |
LIST_REMOVE(debug_list, hashmap_debug_list, &h->debug); |
| 903 |
assert_se(pthread_mutex_unlock(&hashmap_debug_list_mutex) == 0); |
| 904 |
#endif |
| 905 |
|
| 906 |
if (h->from_pool) { |
| 907 |
/* Ensure that the object didn't get migrated between threads. */ |
| 908 |
assert_se(is_main_thread()); |
| 909 |
mempool_free_tile(hashmap_type_info[h->type].mempool, h); |
| 910 |
} else |
| 911 |
free(h); |
| 912 |
} |
| 913 |
|
| 914 |
HashmapBase* _hashmap_free(HashmapBase *h) { |
| 915 |
if (h) { |
| 916 |
_hashmap_clear(h); |
| 917 |
hashmap_free_no_clear(h); |
| 918 |
} |
| 919 |
|
| 920 |
return NULL; |
| 921 |
} |
| 922 |
|
| 923 |
void _hashmap_clear(HashmapBase *h) { |
| 924 |
if (!h) |
| 925 |
return; |
| 926 |
|
| 927 |
if (h->hash_ops->free_key || h->hash_ops->free_value) { |
| 928 |
|
| 929 |
/* If destructor calls are defined, let's destroy things defensively: let's take the item out of the |
| 930 |
* hash table, and only then call the destructor functions. If these destructors then try to unregister |
| 931 |
* themselves from our hash table a second time, the entry is already gone. */ |
| 932 |
|
| 933 |
while (_hashmap_size(h) > 0) { |
| 934 |
void *k = NULL; |
| 935 |
void *v; |
| 936 |
|
| 937 |
v = _hashmap_first_key_and_value(h, true, &k); |
| 938 |
|
| 939 |
if (h->hash_ops->free_key) |
| 940 |
h->hash_ops->free_key(k); |
| 941 |
|
| 942 |
if (h->hash_ops->free_value) |
| 943 |
h->hash_ops->free_value(v); |
| 944 |
} |
| 945 |
} |
| 946 |
|
| 947 |
if (h->has_indirect) { |
| 948 |
free(h->indirect.storage); |
| 949 |
h->has_indirect = false; |
| 950 |
} |
| 951 |
|
| 952 |
h->n_direct_entries = 0; |
| 953 |
reset_direct_storage(h); |
| 954 |
|
| 955 |
if (h->type == HASHMAP_TYPE_ORDERED) { |
| 956 |
OrderedHashmap *lh = (OrderedHashmap*) h; |
| 957 |
lh->iterate_list_head = lh->iterate_list_tail = IDX_NIL; |
| 958 |
} |
| 959 |
|
| 960 |
base_set_dirty(h); |
| 961 |
} |
| 962 |
|
| 963 |
static int resize_buckets(HashmapBase *h, unsigned entries_add); |
| 964 |
|
| 965 |
/* |
| 966 |
* Finds an empty bucket to put an entry into, starting the scan at 'idx'. |
| 967 |
* Performs Robin Hood swaps as it goes. The entry to put must be placed |
| 968 |
* by the caller into swap slot IDX_PUT. |
| 969 |
* If used for in-place resizing, may leave a displaced entry in swap slot |
| 970 |
* IDX_PUT. Caller must rehash it next. |
| 971 |
* Returns: true if it left a displaced entry to rehash next in IDX_PUT, |
| 972 |
* false otherwise. |
| 973 |
*/ |
| 974 |
static bool hashmap_put_robin_hood(HashmapBase *h, unsigned idx, |
| 975 |
struct swap_entries *swap) { |
| 976 |
dib_raw_t raw_dib, *dibs; |
| 977 |
unsigned dib, distance; |
| 978 |
|
| 979 |
#if ENABLE_DEBUG_HASHMAP |
| 980 |
h->debug.put_count++; |
| 981 |
#endif |
| 982 |
|
| 983 |
dibs = dib_raw_ptr(h); |
| 984 |
|
| 985 |
for (distance = 0; ; distance++) { |
| 986 |
raw_dib = dibs[idx]; |
| 987 |
if (IN_SET(raw_dib, DIB_RAW_FREE, DIB_RAW_REHASH)) { |
| 988 |
if (raw_dib == DIB_RAW_REHASH) |
| 989 |
bucket_move_entry(h, swap, idx, IDX_TMP); |
| 990 |
|
| 991 |
if (h->has_indirect && h->indirect.idx_lowest_entry > idx) |
| 992 |
h->indirect.idx_lowest_entry = idx; |
| 993 |
|
| 994 |
bucket_set_dib(h, idx, distance); |
| 995 |
bucket_move_entry(h, swap, IDX_PUT, idx); |
| 996 |
if (raw_dib == DIB_RAW_REHASH) { |
| 997 |
bucket_move_entry(h, swap, IDX_TMP, IDX_PUT); |
| 998 |
return true; |
| 999 |
} |
| 1000 |
|
| 1001 |
return false; |
| 1002 |
} |
| 1003 |
|
| 1004 |
dib = bucket_calculate_dib(h, idx, raw_dib); |
| 1005 |
|
| 1006 |
if (dib < distance) { |
| 1007 |
/* Found a wealthier entry. Go Robin Hood! */ |
| 1008 |
bucket_set_dib(h, idx, distance); |
| 1009 |
|
| 1010 |
/* swap the entries */ |
| 1011 |
bucket_move_entry(h, swap, idx, IDX_TMP); |
| 1012 |
bucket_move_entry(h, swap, IDX_PUT, idx); |
| 1013 |
bucket_move_entry(h, swap, IDX_TMP, IDX_PUT); |
| 1014 |
|
| 1015 |
distance = dib; |
| 1016 |
} |
| 1017 |
|
| 1018 |
idx = next_idx(h, idx); |
| 1019 |
} |
| 1020 |
} |
| 1021 |
|
| 1022 |
/* |
| 1023 |
* Puts an entry into a hashmap, boldly - no check whether key already exists. |
| 1024 |
* The caller must place the entry (only its key and value, not link indexes) |
| 1025 |
* in swap slot IDX_PUT. |
| 1026 |
* Caller must ensure: the key does not exist yet in the hashmap. |
| 1027 |
* that resize is not needed if !may_resize. |
| 1028 |
* Returns: 1 if entry was put successfully. |
| 1029 |
* -ENOMEM if may_resize==true and resize failed with -ENOMEM. |
| 1030 |
* Cannot return -ENOMEM if !may_resize. |
| 1031 |
*/ |
| 1032 |
static int hashmap_base_put_boldly(HashmapBase *h, unsigned idx, |
| 1033 |
struct swap_entries *swap, bool may_resize) { |
| 1034 |
struct ordered_hashmap_entry *new_entry; |
| 1035 |
int r; |
| 1036 |
|
| 1037 |
assert(idx < n_buckets(h)); |
| 1038 |
|
| 1039 |
new_entry = bucket_at_swap(swap, IDX_PUT); |
| 1040 |
|
| 1041 |
if (may_resize) { |
| 1042 |
r = resize_buckets(h, 1); |
| 1043 |
if (r < 0) |
| 1044 |
return r; |
| 1045 |
if (r > 0) |
| 1046 |
idx = bucket_hash(h, new_entry->p.b.key); |
| 1047 |
} |
| 1048 |
assert(n_entries(h) < n_buckets(h)); |
| 1049 |
|
| 1050 |
if (h->type == HASHMAP_TYPE_ORDERED) { |
| 1051 |
OrderedHashmap *lh = (OrderedHashmap*) h; |
| 1052 |
|
| 1053 |
new_entry->iterate_next = IDX_NIL; |
| 1054 |
new_entry->iterate_previous = lh->iterate_list_tail; |
| 1055 |
|
| 1056 |
if (lh->iterate_list_tail != IDX_NIL) { |
| 1057 |
struct ordered_hashmap_entry *old_tail; |
| 1058 |
|
| 1059 |
old_tail = ordered_bucket_at(lh, lh->iterate_list_tail); |
| 1060 |
assert(old_tail->iterate_next == IDX_NIL); |
| 1061 |
old_tail->iterate_next = IDX_PUT; |
| 1062 |
} |
| 1063 |
|
| 1064 |
lh->iterate_list_tail = IDX_PUT; |
| 1065 |
if (lh->iterate_list_head == IDX_NIL) |
| 1066 |
lh->iterate_list_head = IDX_PUT; |
| 1067 |
} |
| 1068 |
|
| 1069 |
assert_se(hashmap_put_robin_hood(h, idx, swap) == false); |
| 1070 |
|
| 1071 |
n_entries_inc(h); |
| 1072 |
#if ENABLE_DEBUG_HASHMAP |
| 1073 |
h->debug.max_entries = MAX(h->debug.max_entries, n_entries(h)); |
| 1074 |
#endif |
| 1075 |
|
| 1076 |
base_set_dirty(h); |
| 1077 |
|
| 1078 |
return 1; |
| 1079 |
} |
| 1080 |
#define hashmap_put_boldly(h, idx, swap, may_resize) \ |
| 1081 |
hashmap_base_put_boldly(HASHMAP_BASE(h), idx, swap, may_resize) |
| 1082 |
|
| 1083 |
/* |
| 1084 |
* Returns 0 if resize is not needed. |
| 1085 |
* 1 if successfully resized. |
| 1086 |
* -ENOMEM on allocation failure. |
| 1087 |
*/ |
| 1088 |
static int resize_buckets(HashmapBase *h, unsigned entries_add) { |
| 1089 |
struct swap_entries swap; |
| 1090 |
void *new_storage; |
| 1091 |
dib_raw_t *old_dibs, *new_dibs; |
| 1092 |
const struct hashmap_type_info *hi; |
| 1093 |
unsigned idx, optimal_idx; |
| 1094 |
unsigned old_n_buckets, new_n_buckets, n_rehashed, new_n_entries; |
| 1095 |
uint8_t new_shift; |
| 1096 |
bool rehash_next; |
| 1097 |
|
| 1098 |
assert(h); |
| 1099 |
|
| 1100 |
hi = &hashmap_type_info[h->type]; |
| 1101 |
new_n_entries = n_entries(h) + entries_add; |
| 1102 |
|
| 1103 |
/* overflow? */ |
| 1104 |
if (_unlikely_(new_n_entries < entries_add)) |
| 1105 |
return -ENOMEM; |
| 1106 |
|
| 1107 |
/* For direct storage we allow 100% load, because it's tiny. */ |
| 1108 |
if (!h->has_indirect && new_n_entries <= hi->n_direct_buckets) |
| 1109 |
return 0; |
| 1110 |
|
| 1111 |
/* |
| 1112 |
* Load factor = n/m = 1 - (1/INV_KEEP_FREE). |
| 1113 |
* From it follows: m = n + n/(INV_KEEP_FREE - 1) |
| 1114 |
*/ |
| 1115 |
new_n_buckets = new_n_entries + new_n_entries / (INV_KEEP_FREE - 1); |
| 1116 |
/* overflow? */ |
| 1117 |
if (_unlikely_(new_n_buckets < new_n_entries)) |
| 1118 |
return -ENOMEM; |
| 1119 |
|
| 1120 |
if (_unlikely_(new_n_buckets > UINT_MAX / (hi->entry_size + sizeof(dib_raw_t)))) |
| 1121 |
return -ENOMEM; |
| 1122 |
|
| 1123 |
old_n_buckets = n_buckets(h); |
| 1124 |
|
| 1125 |
if (_likely_(new_n_buckets <= old_n_buckets)) |
| 1126 |
return 0; |
| 1127 |
|
| 1128 |
new_shift = log2u_round_up(MAX( |
| 1129 |
new_n_buckets * (hi->entry_size + sizeof(dib_raw_t)), |
| 1130 |
2 * sizeof(struct direct_storage))); |
| 1131 |
|
| 1132 |
/* Realloc storage (buckets and DIB array). */ |
| 1133 |
new_storage = realloc(h->has_indirect ? h->indirect.storage : NULL, |
| 1134 |
1U << new_shift); |
| 1135 |
if (!new_storage) |
| 1136 |
return -ENOMEM; |
| 1137 |
|
| 1138 |
/* Must upgrade direct to indirect storage. */ |
| 1139 |
if (!h->has_indirect) { |
| 1140 |
memcpy(new_storage, h->direct.storage, |
| 1141 |
old_n_buckets * (hi->entry_size + sizeof(dib_raw_t))); |
| 1142 |
h->indirect.n_entries = h->n_direct_entries; |
| 1143 |
h->indirect.idx_lowest_entry = 0; |
| 1144 |
h->n_direct_entries = 0; |
| 1145 |
} |
| 1146 |
|
| 1147 |
/* Get a new hash key. If we've just upgraded to indirect storage, |
| 1148 |
* allow reusing a previously generated key. It's still a different key |
| 1149 |
* from the shared one that we used for direct storage. */ |
| 1150 |
get_hash_key(h->indirect.hash_key, !h->has_indirect); |
| 1151 |
|
| 1152 |
h->has_indirect = true; |
| 1153 |
h->indirect.storage = new_storage; |
| 1154 |
h->indirect.n_buckets = (1U << new_shift) / |
| 1155 |
(hi->entry_size + sizeof(dib_raw_t)); |
| 1156 |
|
| 1157 |
old_dibs = (dib_raw_t*)((uint8_t*) new_storage + hi->entry_size * old_n_buckets); |
| 1158 |
new_dibs = dib_raw_ptr(h); |
| 1159 |
|
| 1160 |
/* |
| 1161 |
* Move the DIB array to the new place, replacing valid DIB values with |
| 1162 |
* DIB_RAW_REHASH to indicate all of the used buckets need rehashing. |
| 1163 |
* Note: Overlap is not possible, because we have at least doubled the |
| 1164 |
* number of buckets and dib_raw_t is smaller than any entry type. |
| 1165 |
*/ |
| 1166 |
for (idx = 0; idx < old_n_buckets; idx++) { |
| 1167 |
assert(old_dibs[idx] != DIB_RAW_REHASH); |
| 1168 |
new_dibs[idx] = old_dibs[idx] == DIB_RAW_FREE ? DIB_RAW_FREE |
| 1169 |
: DIB_RAW_REHASH; |
| 1170 |
} |
| 1171 |
|
| 1172 |
/* Zero the area of newly added entries (including the old DIB area) */ |
| 1173 |
memzero(bucket_at(h, old_n_buckets), |
| 1174 |
(n_buckets(h) - old_n_buckets) * hi->entry_size); |
| 1175 |
|
| 1176 |
/* The upper half of the new DIB array needs initialization */ |
| 1177 |
memset(&new_dibs[old_n_buckets], DIB_RAW_INIT, |
| 1178 |
(n_buckets(h) - old_n_buckets) * sizeof(dib_raw_t)); |
| 1179 |
|
| 1180 |
/* Rehash entries that need it */ |
| 1181 |
n_rehashed = 0; |
| 1182 |
for (idx = 0; idx < old_n_buckets; idx++) { |
| 1183 |
if (new_dibs[idx] != DIB_RAW_REHASH) |
| 1184 |
continue; |
| 1185 |
|
| 1186 |
optimal_idx = bucket_hash(h, bucket_at(h, idx)->key); |
| 1187 |
|
| 1188 |
/* |
| 1189 |
* Not much to do if by luck the entry hashes to its current |
| 1190 |
* location. Just set its DIB. |
| 1191 |
*/ |
| 1192 |
if (optimal_idx == idx) { |
| 1193 |
new_dibs[idx] = 0; |
| 1194 |
n_rehashed++; |
| 1195 |
continue; |
| 1196 |
} |
| 1197 |
|
| 1198 |
new_dibs[idx] = DIB_RAW_FREE; |
| 1199 |
bucket_move_entry(h, &swap, idx, IDX_PUT); |
| 1200 |
/* bucket_move_entry does not clear the source */ |
| 1201 |
memzero(bucket_at(h, idx), hi->entry_size); |
| 1202 |
|
| 1203 |
do { |
| 1204 |
/* |
| 1205 |
* Find the new bucket for the current entry. This may make |
| 1206 |
* another entry homeless and load it into IDX_PUT. |
| 1207 |
*/ |
| 1208 |
rehash_next = hashmap_put_robin_hood(h, optimal_idx, &swap); |
| 1209 |
n_rehashed++; |
| 1210 |
|
| 1211 |
/* Did the current entry displace another one? */ |
| 1212 |
if (rehash_next) |
| 1213 |
optimal_idx = bucket_hash(h, bucket_at_swap(&swap, IDX_PUT)->p.b.key); |
| 1214 |
} while (rehash_next); |
| 1215 |
} |
| 1216 |
|
| 1217 |
assert_se(n_rehashed == n_entries(h)); |
| 1218 |
|
| 1219 |
return 1; |
| 1220 |
} |
| 1221 |
|
| 1222 |
/* |
| 1223 |
* Finds an entry with a matching key |
| 1224 |
* Returns: index of the found entry, or IDX_NIL if not found. |
| 1225 |
*/ |
| 1226 |
static unsigned base_bucket_scan(HashmapBase *h, unsigned idx, const void *key) { |
| 1227 |
struct hashmap_base_entry *e; |
| 1228 |
unsigned dib, distance; |
| 1229 |
dib_raw_t *dibs = dib_raw_ptr(h); |
| 1230 |
|
| 1231 |
assert(idx < n_buckets(h)); |
| 1232 |
|
| 1233 |
for (distance = 0; ; distance++) { |
| 1234 |
if (dibs[idx] == DIB_RAW_FREE) |
| 1235 |
return IDX_NIL; |
| 1236 |
|
| 1237 |
dib = bucket_calculate_dib(h, idx, dibs[idx]); |
| 1238 |
|
| 1239 |
if (dib < distance) |
| 1240 |
return IDX_NIL; |
| 1241 |
if (dib == distance) { |
| 1242 |
e = bucket_at(h, idx); |
| 1243 |
if (h->hash_ops->compare(e->key, key) == 0) |
| 1244 |
return idx; |
| 1245 |
} |
| 1246 |
|
| 1247 |
idx = next_idx(h, idx); |
| 1248 |
} |
| 1249 |
} |
| 1250 |
#define bucket_scan(h, idx, key) base_bucket_scan(HASHMAP_BASE(h), idx, key) |
| 1251 |
|
| 1252 |
int hashmap_put(Hashmap *h, const void *key, void *value) { |
| 1253 |
struct swap_entries swap; |
| 1254 |
struct plain_hashmap_entry *e; |
| 1255 |
unsigned hash, idx; |
| 1256 |
|
| 1257 |
assert(h); |
| 1258 |
|
| 1259 |
hash = bucket_hash(h, key); |
| 1260 |
idx = bucket_scan(h, hash, key); |
| 1261 |
if (idx != IDX_NIL) { |
| 1262 |
e = plain_bucket_at(h, idx); |
| 1263 |
if (e->value == value) |
| 1264 |
return 0; |
| 1265 |
return -EEXIST; |
| 1266 |
} |
| 1267 |
|
| 1268 |
e = &bucket_at_swap(&swap, IDX_PUT)->p; |
| 1269 |
e->b.key = key; |
| 1270 |
e->value = value; |
| 1271 |
return hashmap_put_boldly(h, hash, &swap, true); |
| 1272 |
} |
| 1273 |
|
| 1274 |
int set_put(Set *s, const void *key) { |
| 1275 |
struct swap_entries swap; |
| 1276 |
struct hashmap_base_entry *e; |
| 1277 |
unsigned hash, idx; |
| 1278 |
|
| 1279 |
assert(s); |
| 1280 |
|
| 1281 |
hash = bucket_hash(s, key); |
| 1282 |
idx = bucket_scan(s, hash, key); |
| 1283 |
if (idx != IDX_NIL) |
| 1284 |
return 0; |
| 1285 |
|
| 1286 |
e = &bucket_at_swap(&swap, IDX_PUT)->p.b; |
| 1287 |
e->key = key; |
| 1288 |
return hashmap_put_boldly(s, hash, &swap, true); |
| 1289 |
} |
| 1290 |
|
| 1291 |
int set_ensure_put(Set **s, const struct hash_ops *hash_ops, const void *key) { |
| 1292 |
int r; |
| 1293 |
|
| 1294 |
assert(s); |
| 1295 |
|
| 1296 |
r = set_ensure_allocated(s, hash_ops); |
| 1297 |
if (r < 0) |
| 1298 |
return r; |
| 1299 |
|
| 1300 |
return set_put(*s, key); |
| 1301 |
} |
| 1302 |
|
| 1303 |
int set_ensure_consume(Set **s, const struct hash_ops *hash_ops, void *key) { |
| 1304 |
int r; |
| 1305 |
|
| 1306 |
r = set_ensure_put(s, hash_ops, key); |
| 1307 |
if (r <= 0) { |
| 1308 |
if (hash_ops && hash_ops->free_key) |
| 1309 |
hash_ops->free_key(key); |
| 1310 |
else |
| 1311 |
free(key); |
| 1312 |
} |
| 1313 |
|
| 1314 |
return r; |
| 1315 |
} |
| 1316 |
|
| 1317 |
int hashmap_replace(Hashmap *h, const void *key, void *value) { |
| 1318 |
struct swap_entries swap; |
| 1319 |
struct plain_hashmap_entry *e; |
| 1320 |
unsigned hash, idx; |
| 1321 |
|
| 1322 |
assert(h); |
| 1323 |
|
| 1324 |
hash = bucket_hash(h, key); |
| 1325 |
idx = bucket_scan(h, hash, key); |
| 1326 |
if (idx != IDX_NIL) { |
| 1327 |
e = plain_bucket_at(h, idx); |
| 1328 |
#if ENABLE_DEBUG_HASHMAP |
| 1329 |
/* Although the key is equal, the key pointer may have changed, |
| 1330 |
* and this would break our assumption for iterating. So count |
| 1331 |
* this operation as incompatible with iteration. */ |
| 1332 |
if (e->b.key != key) { |
| 1333 |
h->b.debug.put_count++; |
| 1334 |
h->b.debug.rem_count++; |
| 1335 |
h->b.debug.last_rem_idx = idx; |
| 1336 |
} |
| 1337 |
#endif |
| 1338 |
e->b.key = key; |
| 1339 |
e->value = value; |
| 1340 |
hashmap_set_dirty(h); |
| 1341 |
|
| 1342 |
return 0; |
| 1343 |
} |
| 1344 |
|
| 1345 |
e = &bucket_at_swap(&swap, IDX_PUT)->p; |
| 1346 |
e->b.key = key; |
| 1347 |
e->value = value; |
| 1348 |
return hashmap_put_boldly(h, hash, &swap, true); |
| 1349 |
} |
| 1350 |
|
| 1351 |
int hashmap_update(Hashmap *h, const void *key, void *value) { |
| 1352 |
struct plain_hashmap_entry *e; |
| 1353 |
unsigned hash, idx; |
| 1354 |
|
| 1355 |
assert(h); |
| 1356 |
|
| 1357 |
hash = bucket_hash(h, key); |
| 1358 |
idx = bucket_scan(h, hash, key); |
| 1359 |
if (idx == IDX_NIL) |
| 1360 |
return -ENOENT; |
| 1361 |
|
| 1362 |
e = plain_bucket_at(h, idx); |
| 1363 |
e->value = value; |
| 1364 |
hashmap_set_dirty(h); |
| 1365 |
|
| 1366 |
return 0; |
| 1367 |
} |
| 1368 |
|
| 1369 |
void* _hashmap_get(HashmapBase *h, const void *key) { |
| 1370 |
struct hashmap_base_entry *e; |
| 1371 |
unsigned hash, idx; |
| 1372 |
|
| 1373 |
if (!h) |
| 1374 |
return NULL; |
| 1375 |
|
| 1376 |
hash = bucket_hash(h, key); |
| 1377 |
idx = bucket_scan(h, hash, key); |
| 1378 |
if (idx == IDX_NIL) |
| 1379 |
return NULL; |
| 1380 |
|
| 1381 |
e = bucket_at(h, idx); |
| 1382 |
return entry_value(h, e); |
| 1383 |
} |
| 1384 |
|
| 1385 |
void* hashmap_get2(Hashmap *h, const void *key, void **ret) { |
| 1386 |
struct plain_hashmap_entry *e; |
| 1387 |
unsigned hash, idx; |
| 1388 |
|
| 1389 |
if (!h) |
| 1390 |
return NULL; |
| 1391 |
|
| 1392 |
hash = bucket_hash(h, key); |
| 1393 |
idx = bucket_scan(h, hash, key); |
| 1394 |
if (idx == IDX_NIL) |
| 1395 |
return NULL; |
| 1396 |
|
| 1397 |
e = plain_bucket_at(h, idx); |
| 1398 |
if (ret) |
| 1399 |
*ret = (void*) e->b.key; |
| 1400 |
|
| 1401 |
return e->value; |
| 1402 |
} |
| 1403 |
|
| 1404 |
bool _hashmap_contains(HashmapBase *h, const void *key) { |
| 1405 |
unsigned hash; |
| 1406 |
|
| 1407 |
if (!h) |
| 1408 |
return false; |
| 1409 |
|
| 1410 |
hash = bucket_hash(h, key); |
| 1411 |
return bucket_scan(h, hash, key) != IDX_NIL; |
| 1412 |
} |
| 1413 |
|
| 1414 |
void* _hashmap_remove(HashmapBase *h, const void *key) { |
| 1415 |
struct hashmap_base_entry *e; |
| 1416 |
unsigned hash, idx; |
| 1417 |
void *data; |
| 1418 |
|
| 1419 |
if (!h) |
| 1420 |
return NULL; |
| 1421 |
|
| 1422 |
hash = bucket_hash(h, key); |
| 1423 |
idx = bucket_scan(h, hash, key); |
| 1424 |
if (idx == IDX_NIL) |
| 1425 |
return NULL; |
| 1426 |
|
| 1427 |
e = bucket_at(h, idx); |
| 1428 |
data = entry_value(h, e); |
| 1429 |
remove_entry(h, idx); |
| 1430 |
|
| 1431 |
return data; |
| 1432 |
} |
| 1433 |
|
| 1434 |
void* hashmap_remove2(Hashmap *h, const void *key, void **ret) { |
| 1435 |
struct plain_hashmap_entry *e; |
| 1436 |
unsigned hash, idx; |
| 1437 |
void *data; |
| 1438 |
|
| 1439 |
if (!h) { |
| 1440 |
if (ret) |
| 1441 |
*ret = NULL; |
| 1442 |
return NULL; |
| 1443 |
} |
| 1444 |
|
| 1445 |
hash = bucket_hash(h, key); |
| 1446 |
idx = bucket_scan(h, hash, key); |
| 1447 |
if (idx == IDX_NIL) { |
| 1448 |
if (ret) |
| 1449 |
*ret = NULL; |
| 1450 |
return NULL; |
| 1451 |
} |
| 1452 |
|
| 1453 |
e = plain_bucket_at(h, idx); |
| 1454 |
data = e->value; |
| 1455 |
if (ret) |
| 1456 |
*ret = (void*) e->b.key; |
| 1457 |
|
| 1458 |
remove_entry(h, idx); |
| 1459 |
|
| 1460 |
return data; |
| 1461 |
} |
| 1462 |
|
| 1463 |
int hashmap_remove_and_put(Hashmap *h, const void *old_key, const void *new_key, void *value) { |
| 1464 |
struct swap_entries swap; |
| 1465 |
struct plain_hashmap_entry *e; |
| 1466 |
unsigned old_hash, new_hash, idx; |
| 1467 |
|
| 1468 |
if (!h) |
| 1469 |
return -ENOENT; |
| 1470 |
|
| 1471 |
old_hash = bucket_hash(h, old_key); |
| 1472 |
idx = bucket_scan(h, old_hash, old_key); |
| 1473 |
if (idx == IDX_NIL) |
| 1474 |
return -ENOENT; |
| 1475 |
|
| 1476 |
new_hash = bucket_hash(h, new_key); |
| 1477 |
if (bucket_scan(h, new_hash, new_key) != IDX_NIL) |
| 1478 |
return -EEXIST; |
| 1479 |
|
| 1480 |
remove_entry(h, idx); |
| 1481 |
|
| 1482 |
e = &bucket_at_swap(&swap, IDX_PUT)->p; |
| 1483 |
e->b.key = new_key; |
| 1484 |
e->value = value; |
| 1485 |
assert_se(hashmap_put_boldly(h, new_hash, &swap, false) == 1); |
| 1486 |
|
| 1487 |
return 0; |
| 1488 |
} |
| 1489 |
|
| 1490 |
int set_remove_and_put(Set *s, const void *old_key, const void *new_key) { |
| 1491 |
struct swap_entries swap; |
| 1492 |
struct hashmap_base_entry *e; |
| 1493 |
unsigned old_hash, new_hash, idx; |
| 1494 |
|
| 1495 |
if (!s) |
| 1496 |
return -ENOENT; |
| 1497 |
|
| 1498 |
old_hash = bucket_hash(s, old_key); |
| 1499 |
idx = bucket_scan(s, old_hash, old_key); |
| 1500 |
if (idx == IDX_NIL) |
| 1501 |
return -ENOENT; |
| 1502 |
|
| 1503 |
new_hash = bucket_hash(s, new_key); |
| 1504 |
if (bucket_scan(s, new_hash, new_key) != IDX_NIL) |
| 1505 |
return -EEXIST; |
| 1506 |
|
| 1507 |
remove_entry(s, idx); |
| 1508 |
|
| 1509 |
e = &bucket_at_swap(&swap, IDX_PUT)->p.b; |
| 1510 |
e->key = new_key; |
| 1511 |
assert_se(hashmap_put_boldly(s, new_hash, &swap, false) == 1); |
| 1512 |
|
| 1513 |
return 0; |
| 1514 |
} |
| 1515 |
|
| 1516 |
int hashmap_remove_and_replace(Hashmap *h, const void *old_key, const void *new_key, void *value) { |
| 1517 |
struct swap_entries swap; |
| 1518 |
struct plain_hashmap_entry *e; |
| 1519 |
unsigned old_hash, new_hash, idx_old, idx_new; |
| 1520 |
|
| 1521 |
if (!h) |
| 1522 |
return -ENOENT; |
| 1523 |
|
| 1524 |
old_hash = bucket_hash(h, old_key); |
| 1525 |
idx_old = bucket_scan(h, old_hash, old_key); |
| 1526 |
if (idx_old == IDX_NIL) |
| 1527 |
return -ENOENT; |
| 1528 |
|
| 1529 |
old_key = bucket_at(HASHMAP_BASE(h), idx_old)->key; |
| 1530 |
|
| 1531 |
new_hash = bucket_hash(h, new_key); |
| 1532 |
idx_new = bucket_scan(h, new_hash, new_key); |
| 1533 |
if (idx_new != IDX_NIL) |
| 1534 |
if (idx_old != idx_new) { |
| 1535 |
remove_entry(h, idx_new); |
| 1536 |
/* Compensate for a possible backward shift. */ |
| 1537 |
if (old_key != bucket_at(HASHMAP_BASE(h), idx_old)->key) |
| 1538 |
idx_old = prev_idx(HASHMAP_BASE(h), idx_old); |
| 1539 |
assert(old_key == bucket_at(HASHMAP_BASE(h), idx_old)->key); |
| 1540 |
} |
| 1541 |
|
| 1542 |
remove_entry(h, idx_old); |
| 1543 |
|
| 1544 |
e = &bucket_at_swap(&swap, IDX_PUT)->p; |
| 1545 |
e->b.key = new_key; |
| 1546 |
e->value = value; |
| 1547 |
assert_se(hashmap_put_boldly(h, new_hash, &swap, false) == 1); |
| 1548 |
|
| 1549 |
return 0; |
| 1550 |
} |
| 1551 |
|
| 1552 |
void* _hashmap_remove_value(HashmapBase *h, const void *key, void *value) { |
| 1553 |
struct hashmap_base_entry *e; |
| 1554 |
unsigned hash, idx; |
| 1555 |
|
| 1556 |
if (!h) |
| 1557 |
return NULL; |
| 1558 |
|
| 1559 |
hash = bucket_hash(h, key); |
| 1560 |
idx = bucket_scan(h, hash, key); |
| 1561 |
if (idx == IDX_NIL) |
| 1562 |
return NULL; |
| 1563 |
|
| 1564 |
e = bucket_at(h, idx); |
| 1565 |
if (entry_value(h, e) != value) |
| 1566 |
return NULL; |
| 1567 |
|
| 1568 |
remove_entry(h, idx); |
| 1569 |
|
| 1570 |
return value; |
| 1571 |
} |
| 1572 |
|
| 1573 |
static unsigned find_first_entry(HashmapBase *h) { |
| 1574 |
Iterator i = ITERATOR_FIRST; |
| 1575 |
|
| 1576 |
if (!h || !n_entries(h)) |
| 1577 |
return IDX_NIL; |
| 1578 |
|
| 1579 |
return hashmap_iterate_entry(h, &i); |
| 1580 |
} |
| 1581 |
|
| 1582 |
void* _hashmap_first_key_and_value(HashmapBase *h, bool remove, void **ret_key) { |
| 1583 |
struct hashmap_base_entry *e; |
| 1584 |
void *key, *data; |
| 1585 |
unsigned idx; |
| 1586 |
|
| 1587 |
idx = find_first_entry(h); |
| 1588 |
if (idx == IDX_NIL) { |
| 1589 |
if (ret_key) |
| 1590 |
*ret_key = NULL; |
| 1591 |
return NULL; |
| 1592 |
} |
| 1593 |
|
| 1594 |
e = bucket_at(h, idx); |
| 1595 |
key = (void*) e->key; |
| 1596 |
data = entry_value(h, e); |
| 1597 |
|
| 1598 |
if (remove) |
| 1599 |
remove_entry(h, idx); |
| 1600 |
|
| 1601 |
if (ret_key) |
| 1602 |
*ret_key = key; |
| 1603 |
|
| 1604 |
return data; |
| 1605 |
} |
| 1606 |
|
| 1607 |
unsigned _hashmap_size(HashmapBase *h) { |
| 1608 |
if (!h) |
| 1609 |
return 0; |
| 1610 |
|
| 1611 |
return n_entries(h); |
| 1612 |
} |
| 1613 |
|
| 1614 |
unsigned _hashmap_buckets(HashmapBase *h) { |
| 1615 |
if (!h) |
| 1616 |
return 0; |
| 1617 |
|
| 1618 |
return n_buckets(h); |
| 1619 |
} |
| 1620 |
|
| 1621 |
int _hashmap_merge(Hashmap *h, Hashmap *other) { |
| 1622 |
Iterator i; |
| 1623 |
unsigned idx; |
| 1624 |
|
| 1625 |
assert(h); |
| 1626 |
|
| 1627 |
HASHMAP_FOREACH_IDX(idx, HASHMAP_BASE(other), i) { |
| 1628 |
struct plain_hashmap_entry *pe = plain_bucket_at(other, idx); |
| 1629 |
int r; |
| 1630 |
|
| 1631 |
r = hashmap_put(h, pe->b.key, pe->value); |
| 1632 |
if (r < 0 && r != -EEXIST) |
| 1633 |
return r; |
| 1634 |
} |
| 1635 |
|
| 1636 |
return 0; |
| 1637 |
} |
| 1638 |
|
| 1639 |
int set_merge(Set *s, Set *other) { |
| 1640 |
Iterator i; |
| 1641 |
unsigned idx; |
| 1642 |
|
| 1643 |
assert(s); |
| 1644 |
|
| 1645 |
HASHMAP_FOREACH_IDX(idx, HASHMAP_BASE(other), i) { |
| 1646 |
struct set_entry *se = set_bucket_at(other, idx); |
| 1647 |
int r; |
| 1648 |
|
| 1649 |
r = set_put(s, se->b.key); |
| 1650 |
if (r < 0) |
| 1651 |
return r; |
| 1652 |
} |
| 1653 |
|
| 1654 |
return 0; |
| 1655 |
} |
| 1656 |
|
| 1657 |
int _hashmap_reserve(HashmapBase *h, unsigned entries_add) { |
| 1658 |
int r; |
| 1659 |
|
| 1660 |
assert(h); |
| 1661 |
|
| 1662 |
r = resize_buckets(h, entries_add); |
| 1663 |
if (r < 0) |
| 1664 |
return r; |
| 1665 |
|
| 1666 |
return 0; |
| 1667 |
} |
| 1668 |
|
| 1669 |
/* |
| 1670 |
* The same as hashmap_merge(), but every new item from other is moved to h. |
| 1671 |
* Keys already in h are skipped and stay in other. |
| 1672 |
* Returns: 0 on success. |
| 1673 |
* -ENOMEM on alloc failure, in which case no move has been done. |
| 1674 |
*/ |
| 1675 |
int _hashmap_move(HashmapBase *h, HashmapBase *other) { |
| 1676 |
struct swap_entries swap; |
| 1677 |
struct hashmap_base_entry *e, *n; |
| 1678 |
Iterator i; |
| 1679 |
unsigned idx; |
| 1680 |
int r; |
| 1681 |
|
| 1682 |
assert(h); |
| 1683 |
|
| 1684 |
if (!other) |
| 1685 |
return 0; |
| 1686 |
|
| 1687 |
assert(other->type == h->type); |
| 1688 |
|
| 1689 |
/* |
| 1690 |
* This reserves buckets for the worst case, where none of other's |
| 1691 |
* entries are yet present in h. This is preferable to risking |
| 1692 |
* an allocation failure in the middle of the moving and having to |
| 1693 |
* rollback or return a partial result. |
| 1694 |
*/ |
| 1695 |
r = resize_buckets(h, n_entries(other)); |
| 1696 |
if (r < 0) |
| 1697 |
return r; |
| 1698 |
|
| 1699 |
HASHMAP_FOREACH_IDX(idx, other, i) { |
| 1700 |
unsigned h_hash; |
| 1701 |
|
| 1702 |
e = bucket_at(other, idx); |
| 1703 |
h_hash = bucket_hash(h, e->key); |
| 1704 |
if (bucket_scan(h, h_hash, e->key) != IDX_NIL) |
| 1705 |
continue; |
| 1706 |
|
| 1707 |
n = &bucket_at_swap(&swap, IDX_PUT)->p.b; |
| 1708 |
n->key = e->key; |
| 1709 |
if (h->type != HASHMAP_TYPE_SET) |
| 1710 |
((struct plain_hashmap_entry*) n)->value = |
| 1711 |
((struct plain_hashmap_entry*) e)->value; |
| 1712 |
assert_se(hashmap_put_boldly(h, h_hash, &swap, false) == 1); |
| 1713 |
|
| 1714 |
remove_entry(other, idx); |
| 1715 |
} |
| 1716 |
|
| 1717 |
return 0; |
| 1718 |
} |
| 1719 |
|
| 1720 |
int _hashmap_move_one(HashmapBase *h, HashmapBase *other, const void *key) { |
| 1721 |
struct swap_entries swap; |
| 1722 |
unsigned h_hash, other_hash, idx; |
| 1723 |
struct hashmap_base_entry *e, *n; |
| 1724 |
int r; |
| 1725 |
|
| 1726 |
assert(h); |
| 1727 |
|
| 1728 |
h_hash = bucket_hash(h, key); |
| 1729 |
if (bucket_scan(h, h_hash, key) != IDX_NIL) |
| 1730 |
return -EEXIST; |
| 1731 |
|
| 1732 |
if (!other) |
| 1733 |
return -ENOENT; |
| 1734 |
|
| 1735 |
assert(other->type == h->type); |
| 1736 |
|
| 1737 |
other_hash = bucket_hash(other, key); |
| 1738 |
idx = bucket_scan(other, other_hash, key); |
| 1739 |
if (idx == IDX_NIL) |
| 1740 |
return -ENOENT; |
| 1741 |
|
| 1742 |
e = bucket_at(other, idx); |
| 1743 |
|
| 1744 |
n = &bucket_at_swap(&swap, IDX_PUT)->p.b; |
| 1745 |
n->key = e->key; |
| 1746 |
if (h->type != HASHMAP_TYPE_SET) |
| 1747 |
((struct plain_hashmap_entry*) n)->value = |
| 1748 |
((struct plain_hashmap_entry*) e)->value; |
| 1749 |
r = hashmap_put_boldly(h, h_hash, &swap, true); |
| 1750 |
if (r < 0) |
| 1751 |
return r; |
| 1752 |
|
| 1753 |
remove_entry(other, idx); |
| 1754 |
return 0; |
| 1755 |
} |
| 1756 |
|
| 1757 |
HashmapBase* _hashmap_copy(HashmapBase *h) { |
| 1758 |
HashmapBase *copy; |
| 1759 |
int r; |
| 1760 |
|
| 1761 |
assert(h); |
| 1762 |
|
| 1763 |
copy = hashmap_base_new(h->hash_ops, h->type); |
| 1764 |
if (!copy) |
| 1765 |
return NULL; |
| 1766 |
|
| 1767 |
switch (h->type) { |
| 1768 |
case HASHMAP_TYPE_PLAIN: |
| 1769 |
case HASHMAP_TYPE_ORDERED: |
| 1770 |
r = hashmap_merge((Hashmap*)copy, (Hashmap*)h); |
| 1771 |
break; |
| 1772 |
case HASHMAP_TYPE_SET: |
| 1773 |
r = set_merge((Set*)copy, (Set*)h); |
| 1774 |
break; |
| 1775 |
default: |
| 1776 |
assert_not_reached(); |
| 1777 |
} |
| 1778 |
|
| 1779 |
if (r < 0) |
| 1780 |
return _hashmap_free(copy); |
| 1781 |
|
| 1782 |
return copy; |
| 1783 |
} |
| 1784 |
|
| 1785 |
char** _hashmap_get_strv(HashmapBase *h) { |
| 1786 |
char **sv; |
| 1787 |
Iterator i; |
| 1788 |
unsigned idx, n; |
| 1789 |
|
| 1790 |
if (!h) |
| 1791 |
return new0(char*, 1); |
| 1792 |
|
| 1793 |
sv = new(char*, n_entries(h)+1); |
| 1794 |
if (!sv) |
| 1795 |
return NULL; |
| 1796 |
|
| 1797 |
n = 0; |
| 1798 |
HASHMAP_FOREACH_IDX(idx, h, i) |
| 1799 |
sv[n++] = entry_value(h, bucket_at(h, idx)); |
| 1800 |
sv[n] = NULL; |
| 1801 |
|
| 1802 |
return sv; |
| 1803 |
} |
| 1804 |
|
| 1805 |
char** set_to_strv(Set **s) { |
| 1806 |
assert(s); |
| 1807 |
|
| 1808 |
/* This is similar to set_get_strv(), but invalidates the set on success. */ |
| 1809 |
|
| 1810 |
char **v = new(char*, set_size(*s) + 1); |
| 1811 |
if (!v) |
| 1812 |
return NULL; |
| 1813 |
|
| 1814 |
for (char **p = v; (*p = set_steal_first(*s)); p++) |
| 1815 |
; |
| 1816 |
|
| 1817 |
assert(set_isempty(*s)); |
| 1818 |
*s = set_free(*s); |
| 1819 |
return v; |
| 1820 |
} |
| 1821 |
|
| 1822 |
void* ordered_hashmap_next(OrderedHashmap *h, const void *key) { |
| 1823 |
struct ordered_hashmap_entry *e; |
| 1824 |
unsigned hash, idx; |
| 1825 |
|
| 1826 |
if (!h) |
| 1827 |
return NULL; |
| 1828 |
|
| 1829 |
hash = bucket_hash(h, key); |
| 1830 |
idx = bucket_scan(h, hash, key); |
| 1831 |
if (idx == IDX_NIL) |
| 1832 |
return NULL; |
| 1833 |
|
| 1834 |
e = ordered_bucket_at(h, idx); |
| 1835 |
if (e->iterate_next == IDX_NIL) |
| 1836 |
return NULL; |
| 1837 |
return ordered_bucket_at(h, e->iterate_next)->p.value; |
| 1838 |
} |
| 1839 |
|
| 1840 |
int set_consume(Set *s, void *value) { |
| 1841 |
int r; |
| 1842 |
|
| 1843 |
assert(s); |
| 1844 |
assert(value); |
| 1845 |
|
| 1846 |
r = set_put(s, value); |
| 1847 |
if (r <= 0) |
| 1848 |
free(value); |
| 1849 |
|
| 1850 |
return r; |
| 1851 |
} |
| 1852 |
|
| 1853 |
int hashmap_put_strdup_full(Hashmap **h, const struct hash_ops *hash_ops, const char *k, const char *v) { |
| 1854 |
int r; |
| 1855 |
|
| 1856 |
assert(h); |
| 1857 |
|
| 1858 |
r = hashmap_ensure_allocated(h, hash_ops); |
| 1859 |
if (r < 0) |
| 1860 |
return r; |
| 1861 |
|
| 1862 |
_cleanup_free_ char *kdup = NULL, *vdup = NULL; |
| 1863 |
|
| 1864 |
kdup = strdup(k); |
| 1865 |
if (!kdup) |
| 1866 |
return -ENOMEM; |
| 1867 |
|
| 1868 |
if (v) { |
| 1869 |
vdup = strdup(v); |
| 1870 |
if (!vdup) |
| 1871 |
return -ENOMEM; |
| 1872 |
} |
| 1873 |
|
| 1874 |
r = hashmap_put(*h, kdup, vdup); |
| 1875 |
if (r < 0) { |
| 1876 |
if (r == -EEXIST && streq_ptr(v, hashmap_get(*h, kdup))) |
| 1877 |
return 0; |
| 1878 |
return r; |
| 1879 |
} |
| 1880 |
|
| 1881 |
/* 0 with non-null vdup would mean vdup is already in the hashmap, which cannot be */ |
| 1882 |
assert(vdup == NULL || r > 0); |
| 1883 |
if (r > 0) |
| 1884 |
kdup = vdup = NULL; |
| 1885 |
|
| 1886 |
return r; |
| 1887 |
} |
| 1888 |
|
| 1889 |
int set_put_strndup_full(Set **s, const struct hash_ops *hash_ops, const char *p, size_t n) { |
| 1890 |
char *c; |
| 1891 |
int r; |
| 1892 |
|
| 1893 |
assert(s); |
| 1894 |
assert(p); |
| 1895 |
|
| 1896 |
r = set_ensure_allocated(s, hash_ops); |
| 1897 |
if (r < 0) |
| 1898 |
return r; |
| 1899 |
|
| 1900 |
if (n == SIZE_MAX) { |
| 1901 |
if (set_contains(*s, (char*) p)) |
| 1902 |
return 0; |
| 1903 |
|
| 1904 |
c = strdup(p); |
| 1905 |
} else |
| 1906 |
c = strndup(p, n); |
| 1907 |
if (!c) |
| 1908 |
return -ENOMEM; |
| 1909 |
|
| 1910 |
return set_consume(*s, c); |
| 1911 |
} |
| 1912 |
|
| 1913 |
int set_put_strdupv_full(Set **s, const struct hash_ops *hash_ops, char **l) { |
| 1914 |
int n = 0, r; |
| 1915 |
|
| 1916 |
assert(s); |
| 1917 |
|
| 1918 |
STRV_FOREACH(i, l) { |
| 1919 |
r = set_put_strndup_full(s, hash_ops, *i, SIZE_MAX); |
| 1920 |
if (r < 0) |
| 1921 |
return r; |
| 1922 |
|
| 1923 |
n += r; |
| 1924 |
} |
| 1925 |
|
| 1926 |
return n; |
| 1927 |
} |
| 1928 |
|
| 1929 |
int set_put_strsplit(Set *s, const char *v, const char *separators, ExtractFlags flags) { |
| 1930 |
const char *p = ASSERT_PTR(v); |
| 1931 |
int r; |
| 1932 |
|
| 1933 |
assert(s); |
| 1934 |
|
| 1935 |
for (;;) { |
| 1936 |
char *word; |
| 1937 |
|
| 1938 |
r = extract_first_word(&p, &word, separators, flags); |
| 1939 |
if (r <= 0) |
| 1940 |
return r; |
| 1941 |
|
| 1942 |
r = set_consume(s, word); |
| 1943 |
if (r < 0) |
| 1944 |
return r; |
| 1945 |
} |
| 1946 |
} |
| 1947 |
|
| 1948 |
/* expand the cachemem if needed, return true if newly (re)activated. */ |
| 1949 |
static int cachemem_maintain(CacheMem *mem, size_t size) { |
| 1950 |
assert(mem); |
| 1951 |
|
| 1952 |
if (!GREEDY_REALLOC(mem->ptr, size)) { |
| 1953 |
if (size > 0) |
| 1954 |
return -ENOMEM; |
| 1955 |
} |
| 1956 |
|
| 1957 |
if (!mem->active) { |
| 1958 |
mem->active = true; |
| 1959 |
return true; |
| 1960 |
} |
| 1961 |
|
| 1962 |
return false; |
| 1963 |
} |
| 1964 |
|
| 1965 |
int iterated_cache_get(IteratedCache *cache, const void ***res_keys, const void ***res_values, unsigned *res_n_entries) { |
| 1966 |
bool sync_keys = false, sync_values = false; |
| 1967 |
size_t size; |
| 1968 |
int r; |
| 1969 |
|
| 1970 |
assert(cache); |
| 1971 |
assert(cache->hashmap); |
| 1972 |
|
| 1973 |
size = n_entries(cache->hashmap); |
| 1974 |
|
| 1975 |
if (res_keys) { |
| 1976 |
r = cachemem_maintain(&cache->keys, size); |
| 1977 |
if (r < 0) |
| 1978 |
return r; |
| 1979 |
|
| 1980 |
sync_keys = r; |
| 1981 |
} else |
| 1982 |
cache->keys.active = false; |
| 1983 |
|
| 1984 |
if (res_values) { |
| 1985 |
r = cachemem_maintain(&cache->values, size); |
| 1986 |
if (r < 0) |
| 1987 |
return r; |
| 1988 |
|
| 1989 |
sync_values = r; |
| 1990 |
} else |
| 1991 |
cache->values.active = false; |
| 1992 |
|
| 1993 |
if (cache->hashmap->dirty) { |
| 1994 |
if (cache->keys.active) |
| 1995 |
sync_keys = true; |
| 1996 |
if (cache->values.active) |
| 1997 |
sync_values = true; |
| 1998 |
|
| 1999 |
cache->hashmap->dirty = false; |
| 2000 |
} |
| 2001 |
|
| 2002 |
if (sync_keys || sync_values) { |
| 2003 |
unsigned i, idx; |
| 2004 |
Iterator iter; |
| 2005 |
|
| 2006 |
i = 0; |
| 2007 |
HASHMAP_FOREACH_IDX(idx, cache->hashmap, iter) { |
| 2008 |
struct hashmap_base_entry *e; |
| 2009 |
|
| 2010 |
e = bucket_at(cache->hashmap, idx); |
| 2011 |
|
| 2012 |
if (sync_keys) |
| 2013 |
cache->keys.ptr[i] = e->key; |
| 2014 |
if (sync_values) |
| 2015 |
cache->values.ptr[i] = entry_value(cache->hashmap, e); |
| 2016 |
i++; |
| 2017 |
} |
| 2018 |
} |
| 2019 |
|
| 2020 |
if (res_keys) |
| 2021 |
*res_keys = cache->keys.ptr; |
| 2022 |
if (res_values) |
| 2023 |
*res_values = cache->values.ptr; |
| 2024 |
if (res_n_entries) |
| 2025 |
*res_n_entries = size; |
| 2026 |
|
| 2027 |
return 0; |
| 2028 |
} |
| 2029 |
|
| 2030 |
IteratedCache* iterated_cache_free(IteratedCache *cache) { |
| 2031 |
if (cache) { |
| 2032 |
free(cache->keys.ptr); |
| 2033 |
free(cache->values.ptr); |
| 2034 |
} |
| 2035 |
|
| 2036 |
return mfree(cache); |
| 2037 |
} |
| 2038 |
|
| 2039 |
int set_strjoin(Set *s, const char *separator, bool wrap_with_separator, char **ret) { |
| 2040 |
_cleanup_free_ char *str = NULL; |
| 2041 |
size_t separator_len, len = 0; |
| 2042 |
const char *value; |
| 2043 |
bool first; |
| 2044 |
|
| 2045 |
assert(ret); |
| 2046 |
|
| 2047 |
if (set_isempty(s)) { |
| 2048 |
*ret = NULL; |
| 2049 |
return 0; |
| 2050 |
} |
| 2051 |
|
| 2052 |
separator_len = strlen_ptr(separator); |
| 2053 |
|
| 2054 |
if (separator_len == 0) |
| 2055 |
wrap_with_separator = false; |
| 2056 |
|
| 2057 |
first = !wrap_with_separator; |
| 2058 |
|
| 2059 |
SET_FOREACH(value, s) { |
| 2060 |
size_t l = strlen_ptr(value); |
| 2061 |
|
| 2062 |
if (l == 0) |
| 2063 |
continue; |
| 2064 |
|
| 2065 |
if (!GREEDY_REALLOC(str, len + l + (first ? 0 : separator_len) + (wrap_with_separator ? separator_len : 0) + 1)) |
| 2066 |
return -ENOMEM; |
| 2067 |
|
| 2068 |
if (separator_len > 0 && !first) { |
| 2069 |
memcpy(str + len, separator, separator_len); |
| 2070 |
len += separator_len; |
| 2071 |
} |
| 2072 |
|
| 2073 |
memcpy(str + len, value, l); |
| 2074 |
len += l; |
| 2075 |
first = false; |
| 2076 |
} |
| 2077 |
|
| 2078 |
if (wrap_with_separator) { |
| 2079 |
memcpy(str + len, separator, separator_len); |
| 2080 |
len += separator_len; |
| 2081 |
} |
| 2082 |
|
| 2083 |
str[len] = '\0'; |
| 2084 |
|
| 2085 |
*ret = TAKE_PTR(str); |
| 2086 |
return 0; |
| 2087 |
} |
| 2088 |
|
| 2089 |
bool set_equal(Set *a, Set *b) { |
| 2090 |
void *p; |
| 2091 |
|
| 2092 |
/* Checks whether each entry of 'a' is also in 'b' and vice versa, i.e. the two sets contain the same |
| 2093 |
* entries */ |
| 2094 |
|
| 2095 |
if (a == b) |
| 2096 |
return true; |
| 2097 |
|
| 2098 |
if (set_isempty(a) && set_isempty(b)) |
| 2099 |
return true; |
| 2100 |
|
| 2101 |
if (set_size(a) != set_size(b)) /* Cheap check that hopefully catches a lot of inequality cases |
| 2102 |
* already */ |
| 2103 |
return false; |
| 2104 |
|
| 2105 |
SET_FOREACH(p, a) |
| 2106 |
if (!set_contains(b, p)) |
| 2107 |
return false; |
| 2108 |
|
| 2109 |
/* If we have the same hashops, then we don't need to check things backwards given we compared the |
| 2110 |
* size and that all of a is in b. */ |
| 2111 |
if (a->b.hash_ops == b->b.hash_ops) |
| 2112 |
return true; |
| 2113 |
|
| 2114 |
SET_FOREACH(p, b) |
| 2115 |
if (!set_contains(a, p)) |
| 2116 |
return false; |
| 2117 |
|
| 2118 |
return true; |
| 2119 |
} |
| 2120 |
|
| 2121 |
static bool set_fnmatch_one(Set *patterns, const char *needle) { |
| 2122 |
const char *p; |
| 2123 |
|
| 2124 |
assert(needle); |
| 2125 |
|
| 2126 |
/* Any failure of fnmatch() is treated as equivalent to FNM_NOMATCH, i.e. as non-matching pattern */ |
| 2127 |
|
| 2128 |
SET_FOREACH(p, patterns) |
| 2129 |
if (fnmatch(p, needle, 0) == 0) |
| 2130 |
return true; |
| 2131 |
|
| 2132 |
return false; |
| 2133 |
} |
| 2134 |
|
| 2135 |
bool set_fnmatch(Set *include_patterns, Set *exclude_patterns, const char *needle) { |
| 2136 |
assert(needle); |
| 2137 |
|
| 2138 |
if (set_fnmatch_one(exclude_patterns, needle)) |
| 2139 |
return false; |
| 2140 |
|
| 2141 |
if (set_isempty(include_patterns)) |
| 2142 |
return true; |
| 2143 |
|
| 2144 |
return set_fnmatch_one(include_patterns, needle); |
| 2145 |
} |
| 2146 |
|
| 2147 |
static int hashmap_entry_compare( |
| 2148 |
struct hashmap_base_entry * const *a, |
| 2149 |
struct hashmap_base_entry * const *b, |
| 2150 |
compare_func_t compare) { |
| 2151 |
|
| 2152 |
assert(a && *a); |
| 2153 |
assert(b && *b); |
| 2154 |
assert(compare); |
| 2155 |
|
| 2156 |
return compare((*a)->key, (*b)->key); |
| 2157 |
} |
| 2158 |
|
| 2159 |
static int _hashmap_dump_entries_sorted( |
| 2160 |
HashmapBase *h, |
| 2161 |
void ***ret, |
| 2162 |
size_t *ret_n) { |
| 2163 |
_cleanup_free_ void **entries = NULL; |
| 2164 |
Iterator iter; |
| 2165 |
unsigned idx; |
| 2166 |
size_t n = 0; |
| 2167 |
|
| 2168 |
assert(ret); |
| 2169 |
assert(ret_n); |
| 2170 |
|
| 2171 |
if (_hashmap_size(h) == 0) { |
| 2172 |
*ret = NULL; |
| 2173 |
*ret_n = 0; |
| 2174 |
return 0; |
| 2175 |
} |
| 2176 |
|
| 2177 |
/* We append one more element than needed so that the resulting array can be used as a strv. We |
| 2178 |
* don't count this entry in the returned size. */ |
| 2179 |
entries = new(void*, _hashmap_size(h) + 1); |
| 2180 |
if (!entries) |
| 2181 |
return -ENOMEM; |
| 2182 |
|
| 2183 |
HASHMAP_FOREACH_IDX(idx, h, iter) |
| 2184 |
entries[n++] = bucket_at(h, idx); |
| 2185 |
|
| 2186 |
assert(n == _hashmap_size(h)); |
| 2187 |
entries[n] = NULL; |
| 2188 |
|
| 2189 |
typesafe_qsort_r((struct hashmap_base_entry**) entries, n, |
| 2190 |
hashmap_entry_compare, h->hash_ops->compare); |
| 2191 |
|
| 2192 |
*ret = TAKE_PTR(entries); |
| 2193 |
*ret_n = n; |
| 2194 |
return 0; |
| 2195 |
} |
| 2196 |
|
| 2197 |
int _hashmap_dump_keys_sorted(HashmapBase *h, void ***ret, size_t *ret_n) { |
| 2198 |
_cleanup_free_ void **entries = NULL; |
| 2199 |
size_t n; |
| 2200 |
int r; |
| 2201 |
|
| 2202 |
assert(ret); |
| 2203 |
|
| 2204 |
r = _hashmap_dump_entries_sorted(h, &entries, &n); |
| 2205 |
if (r < 0) |
| 2206 |
return r; |
| 2207 |
|
| 2208 |
/* Reuse the array. */ |
| 2209 |
FOREACH_ARRAY(e, entries, n) |
| 2210 |
*e = (void*) (*(struct hashmap_base_entry**) e)->key; |
| 2211 |
|
| 2212 |
*ret = TAKE_PTR(entries); |
| 2213 |
if (ret_n) |
| 2214 |
*ret_n = n; |
| 2215 |
return 0; |
| 2216 |
} |
| 2217 |
|
| 2218 |
int _hashmap_dump_sorted(HashmapBase *h, void ***ret, size_t *ret_n) { |
| 2219 |
_cleanup_free_ void **entries = NULL; |
| 2220 |
size_t n; |
| 2221 |
int r; |
| 2222 |
|
| 2223 |
assert(ret); |
| 2224 |
|
| 2225 |
r = _hashmap_dump_entries_sorted(h, &entries, &n); |
| 2226 |
if (r < 0) |
| 2227 |
return r; |
| 2228 |
|
| 2229 |
/* Reuse the array. */ |
| 2230 |
FOREACH_ARRAY(e, entries, n) |
| 2231 |
*e = entry_value(h, *(struct hashmap_base_entry**) e); |
| 2232 |
|
| 2233 |
*ret = TAKE_PTR(entries); |
| 2234 |
if (ret_n) |
| 2235 |
*ret_n = n; |
| 2236 |
return 0; |
| 2237 |
} |