1 /*
2 * GPL HEADER START
3 *
4 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
5 *
6 * This program is free software; you can redistribute it and/or modify
7 * it under the terms of the GNU General Public License version 2 only,
8 * as published by the Free Software Foundation.
9 *
10 * This program is distributed in the hope that it will be useful, but
11 * WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13 * General Public License version 2 for more details (a copy is included
14 * in the LICENSE file that accompanied this code).
15 *
16 * You should have received a copy of the GNU General Public License
17 * version 2 along with this program; If not, see
18 * http://www.sun.com/software/products/lustre/docs/GPLv2.pdf
19 *
20 * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara,
21 * CA 95054 USA or visit www.sun.com if you need additional information or
22 * have any questions.
23 *
24 * GPL HEADER END
25 */
26 /*
27 * Copyright (c) 2009, 2010, Oracle and/or its affiliates. All rights reserved.
28 * Use is subject to license terms.
29 *
30 * Copyright (c) 2011, 2012, Intel Corporation.
31 */
32 /*
33 * This file is part of Lustre, http://www.lustre.org/
34 * Lustre is a trademark of Sun Microsystems, Inc.
35 *
36 * libcfs/libcfs/hash.c
37 *
38 * Implement a hash class for hash process in lustre system.
39 *
40 * Author: YuZhangyong <yzy@clusterfs.com>
41 *
42 * 2008-08-15: Brian Behlendorf <behlendorf1@llnl.gov>
43 * - Simplified API and improved documentation
44 * - Added per-hash feature flags:
45 * * CFS_HASH_DEBUG additional validation
46 * * CFS_HASH_REHASH dynamic rehashing
47 * - Added per-hash statistics
48 * - General performance enhancements
49 *
50 * 2009-07-31: Liang Zhen <zhen.liang@sun.com>
51 * - move all stuff to libcfs
52 * - don't allow cur_bits != max_bits without setting of CFS_HASH_REHASH
53 * - ignore hs_rwlock if without CFS_HASH_REHASH setting
54 * - buckets are allocated one by one(instead of contiguous memory),
55 * to avoid unnecessary cacheline conflict
56 *
57 * 2010-03-01: Liang Zhen <zhen.liang@sun.com>
58 * - "bucket" is a group of hlist_head now, user can specify bucket size
59 * by bkt_bits of cfs_hash_create(), all hlist_heads in a bucket share
60 * one lock for reducing memory overhead.
61 *
62 * - support lockless hash, caller will take care of locks:
63 * avoid lock overhead for hash tables that are already protected
64 * by locking in the caller for another reason
65 *
66 * - support both spin_lock/rwlock for bucket:
67 * overhead of spinlock contention is lower than read/write
68 * contention of rwlock, so using spinlock to serialize operations on
69 * bucket is more reasonable for those frequently changed hash tables
70 *
71 * - support one-single lock mode:
72 * one lock to protect all hash operations to avoid overhead of
73 * multiple locks if hash table is always small
74 *
75 * - removed a lot of unnecessary addref & decref on hash element:
76 * addref & decref are atomic operations in many use-cases which
77 * are expensive.
78 *
79 * - support non-blocking cfs_hash_add() and cfs_hash_findadd():
80 * some lustre use-cases require these functions to be strictly
81 * non-blocking, we need to schedule required rehash on a different
82 * thread on those cases.
83 *
84 * - safer rehash on large hash table
85 * In old implementation, rehash function will exclusively lock the
86 * hash table and finish rehash in one batch, it's dangerous on SMP
87 * system because rehash millions of elements could take long time.
88 * New implemented rehash can release lock and relax CPU in middle
89 * of rehash, it's safe for another thread to search/change on the
90 * hash table even it's in rehasing.
91 *
92 * - support two different refcount modes
93 * . hash table has refcount on element
94 * . hash table doesn't change refcount on adding/removing element
95 *
96 * - support long name hash table (for param-tree)
97 *
98 * - fix a bug for cfs_hash_rehash_key:
99 * in old implementation, cfs_hash_rehash_key could screw up the
100 * hash-table because @key is overwritten without any protection.
101 * Now we need user to define hs_keycpy for those rehash enabled
102 * hash tables, cfs_hash_rehash_key will overwrite hash-key
103 * inside lock by calling hs_keycpy.
104 *
105 * - better hash iteration:
106 * Now we support both locked iteration & lockless iteration of hash
107 * table. Also, user can break the iteration by return 1 in callback.
108 */
109
110 #include "../../include/linux/libcfs/libcfs.h"
111 #include <linux/seq_file.h>
112
113 #if CFS_HASH_DEBUG_LEVEL >= CFS_HASH_DEBUG_1
114 static unsigned int warn_on_depth = 8;
115 module_param(warn_on_depth, uint, 0644);
116 MODULE_PARM_DESC(warn_on_depth, "warning when hash depth is high.");
117 #endif
118
119 struct cfs_wi_sched *cfs_sched_rehash;
120
121 static inline void
cfs_hash_nl_lock(union cfs_hash_lock * lock,int exclusive)122 cfs_hash_nl_lock(union cfs_hash_lock *lock, int exclusive) {}
123
124 static inline void
cfs_hash_nl_unlock(union cfs_hash_lock * lock,int exclusive)125 cfs_hash_nl_unlock(union cfs_hash_lock *lock, int exclusive) {}
126
127 static inline void
cfs_hash_spin_lock(union cfs_hash_lock * lock,int exclusive)128 cfs_hash_spin_lock(union cfs_hash_lock *lock, int exclusive)
129 __acquires(&lock->spin)
130 {
131 spin_lock(&lock->spin);
132 }
133
134 static inline void
cfs_hash_spin_unlock(union cfs_hash_lock * lock,int exclusive)135 cfs_hash_spin_unlock(union cfs_hash_lock *lock, int exclusive)
136 __releases(&lock->spin)
137 {
138 spin_unlock(&lock->spin);
139 }
140
141 static inline void
cfs_hash_rw_lock(union cfs_hash_lock * lock,int exclusive)142 cfs_hash_rw_lock(union cfs_hash_lock *lock, int exclusive)
143 __acquires(&lock->rw)
144 {
145 if (!exclusive)
146 read_lock(&lock->rw);
147 else
148 write_lock(&lock->rw);
149 }
150
151 static inline void
cfs_hash_rw_unlock(union cfs_hash_lock * lock,int exclusive)152 cfs_hash_rw_unlock(union cfs_hash_lock *lock, int exclusive)
153 __releases(&lock->rw)
154 {
155 if (!exclusive)
156 read_unlock(&lock->rw);
157 else
158 write_unlock(&lock->rw);
159 }
160
161 /** No lock hash */
162 static cfs_hash_lock_ops_t cfs_hash_nl_lops = {
163 .hs_lock = cfs_hash_nl_lock,
164 .hs_unlock = cfs_hash_nl_unlock,
165 .hs_bkt_lock = cfs_hash_nl_lock,
166 .hs_bkt_unlock = cfs_hash_nl_unlock,
167 };
168
169 /** no bucket lock, one spinlock to protect everything */
170 static cfs_hash_lock_ops_t cfs_hash_nbl_lops = {
171 .hs_lock = cfs_hash_spin_lock,
172 .hs_unlock = cfs_hash_spin_unlock,
173 .hs_bkt_lock = cfs_hash_nl_lock,
174 .hs_bkt_unlock = cfs_hash_nl_unlock,
175 };
176
177 /** spin bucket lock, rehash is enabled */
178 static cfs_hash_lock_ops_t cfs_hash_bkt_spin_lops = {
179 .hs_lock = cfs_hash_rw_lock,
180 .hs_unlock = cfs_hash_rw_unlock,
181 .hs_bkt_lock = cfs_hash_spin_lock,
182 .hs_bkt_unlock = cfs_hash_spin_unlock,
183 };
184
185 /** rw bucket lock, rehash is enabled */
186 static cfs_hash_lock_ops_t cfs_hash_bkt_rw_lops = {
187 .hs_lock = cfs_hash_rw_lock,
188 .hs_unlock = cfs_hash_rw_unlock,
189 .hs_bkt_lock = cfs_hash_rw_lock,
190 .hs_bkt_unlock = cfs_hash_rw_unlock,
191 };
192
193 /** spin bucket lock, rehash is disabled */
194 static cfs_hash_lock_ops_t cfs_hash_nr_bkt_spin_lops = {
195 .hs_lock = cfs_hash_nl_lock,
196 .hs_unlock = cfs_hash_nl_unlock,
197 .hs_bkt_lock = cfs_hash_spin_lock,
198 .hs_bkt_unlock = cfs_hash_spin_unlock,
199 };
200
201 /** rw bucket lock, rehash is disabled */
202 static cfs_hash_lock_ops_t cfs_hash_nr_bkt_rw_lops = {
203 .hs_lock = cfs_hash_nl_lock,
204 .hs_unlock = cfs_hash_nl_unlock,
205 .hs_bkt_lock = cfs_hash_rw_lock,
206 .hs_bkt_unlock = cfs_hash_rw_unlock,
207 };
208
209 static void
cfs_hash_lock_setup(struct cfs_hash * hs)210 cfs_hash_lock_setup(struct cfs_hash *hs)
211 {
212 if (cfs_hash_with_no_lock(hs)) {
213 hs->hs_lops = &cfs_hash_nl_lops;
214
215 } else if (cfs_hash_with_no_bktlock(hs)) {
216 hs->hs_lops = &cfs_hash_nbl_lops;
217 spin_lock_init(&hs->hs_lock.spin);
218
219 } else if (cfs_hash_with_rehash(hs)) {
220 rwlock_init(&hs->hs_lock.rw);
221
222 if (cfs_hash_with_rw_bktlock(hs))
223 hs->hs_lops = &cfs_hash_bkt_rw_lops;
224 else if (cfs_hash_with_spin_bktlock(hs))
225 hs->hs_lops = &cfs_hash_bkt_spin_lops;
226 else
227 LBUG();
228 } else {
229 if (cfs_hash_with_rw_bktlock(hs))
230 hs->hs_lops = &cfs_hash_nr_bkt_rw_lops;
231 else if (cfs_hash_with_spin_bktlock(hs))
232 hs->hs_lops = &cfs_hash_nr_bkt_spin_lops;
233 else
234 LBUG();
235 }
236 }
237
238 /**
239 * Simple hash head without depth tracking
240 * new element is always added to head of hlist
241 */
242 typedef struct {
243 struct hlist_head hh_head; /**< entries list */
244 } cfs_hash_head_t;
245
246 static int
cfs_hash_hh_hhead_size(struct cfs_hash * hs)247 cfs_hash_hh_hhead_size(struct cfs_hash *hs)
248 {
249 return sizeof(cfs_hash_head_t);
250 }
251
252 static struct hlist_head *
cfs_hash_hh_hhead(struct cfs_hash * hs,struct cfs_hash_bd * bd)253 cfs_hash_hh_hhead(struct cfs_hash *hs, struct cfs_hash_bd *bd)
254 {
255 cfs_hash_head_t *head = (cfs_hash_head_t *)&bd->bd_bucket->hsb_head[0];
256
257 return &head[bd->bd_offset].hh_head;
258 }
259
260 static int
cfs_hash_hh_hnode_add(struct cfs_hash * hs,struct cfs_hash_bd * bd,struct hlist_node * hnode)261 cfs_hash_hh_hnode_add(struct cfs_hash *hs, struct cfs_hash_bd *bd,
262 struct hlist_node *hnode)
263 {
264 hlist_add_head(hnode, cfs_hash_hh_hhead(hs, bd));
265 return -1; /* unknown depth */
266 }
267
268 static int
cfs_hash_hh_hnode_del(struct cfs_hash * hs,struct cfs_hash_bd * bd,struct hlist_node * hnode)269 cfs_hash_hh_hnode_del(struct cfs_hash *hs, struct cfs_hash_bd *bd,
270 struct hlist_node *hnode)
271 {
272 hlist_del_init(hnode);
273 return -1; /* unknown depth */
274 }
275
276 /**
277 * Simple hash head with depth tracking
278 * new element is always added to head of hlist
279 */
280 typedef struct {
281 struct hlist_head hd_head; /**< entries list */
282 unsigned int hd_depth; /**< list length */
283 } cfs_hash_head_dep_t;
284
285 static int
cfs_hash_hd_hhead_size(struct cfs_hash * hs)286 cfs_hash_hd_hhead_size(struct cfs_hash *hs)
287 {
288 return sizeof(cfs_hash_head_dep_t);
289 }
290
291 static struct hlist_head *
cfs_hash_hd_hhead(struct cfs_hash * hs,struct cfs_hash_bd * bd)292 cfs_hash_hd_hhead(struct cfs_hash *hs, struct cfs_hash_bd *bd)
293 {
294 cfs_hash_head_dep_t *head;
295
296 head = (cfs_hash_head_dep_t *)&bd->bd_bucket->hsb_head[0];
297 return &head[bd->bd_offset].hd_head;
298 }
299
300 static int
cfs_hash_hd_hnode_add(struct cfs_hash * hs,struct cfs_hash_bd * bd,struct hlist_node * hnode)301 cfs_hash_hd_hnode_add(struct cfs_hash *hs, struct cfs_hash_bd *bd,
302 struct hlist_node *hnode)
303 {
304 cfs_hash_head_dep_t *hh = container_of(cfs_hash_hd_hhead(hs, bd),
305 cfs_hash_head_dep_t, hd_head);
306 hlist_add_head(hnode, &hh->hd_head);
307 return ++hh->hd_depth;
308 }
309
310 static int
cfs_hash_hd_hnode_del(struct cfs_hash * hs,struct cfs_hash_bd * bd,struct hlist_node * hnode)311 cfs_hash_hd_hnode_del(struct cfs_hash *hs, struct cfs_hash_bd *bd,
312 struct hlist_node *hnode)
313 {
314 cfs_hash_head_dep_t *hh = container_of(cfs_hash_hd_hhead(hs, bd),
315 cfs_hash_head_dep_t, hd_head);
316 hlist_del_init(hnode);
317 return --hh->hd_depth;
318 }
319
320 /**
321 * double links hash head without depth tracking
322 * new element is always added to tail of hlist
323 */
324 typedef struct {
325 struct hlist_head dh_head; /**< entries list */
326 struct hlist_node *dh_tail; /**< the last entry */
327 } cfs_hash_dhead_t;
328
329 static int
cfs_hash_dh_hhead_size(struct cfs_hash * hs)330 cfs_hash_dh_hhead_size(struct cfs_hash *hs)
331 {
332 return sizeof(cfs_hash_dhead_t);
333 }
334
335 static struct hlist_head *
cfs_hash_dh_hhead(struct cfs_hash * hs,struct cfs_hash_bd * bd)336 cfs_hash_dh_hhead(struct cfs_hash *hs, struct cfs_hash_bd *bd)
337 {
338 cfs_hash_dhead_t *head;
339
340 head = (cfs_hash_dhead_t *)&bd->bd_bucket->hsb_head[0];
341 return &head[bd->bd_offset].dh_head;
342 }
343
344 static int
cfs_hash_dh_hnode_add(struct cfs_hash * hs,struct cfs_hash_bd * bd,struct hlist_node * hnode)345 cfs_hash_dh_hnode_add(struct cfs_hash *hs, struct cfs_hash_bd *bd,
346 struct hlist_node *hnode)
347 {
348 cfs_hash_dhead_t *dh = container_of(cfs_hash_dh_hhead(hs, bd),
349 cfs_hash_dhead_t, dh_head);
350
351 if (dh->dh_tail != NULL) /* not empty */
352 hlist_add_behind(hnode, dh->dh_tail);
353 else /* empty list */
354 hlist_add_head(hnode, &dh->dh_head);
355 dh->dh_tail = hnode;
356 return -1; /* unknown depth */
357 }
358
359 static int
cfs_hash_dh_hnode_del(struct cfs_hash * hs,struct cfs_hash_bd * bd,struct hlist_node * hnd)360 cfs_hash_dh_hnode_del(struct cfs_hash *hs, struct cfs_hash_bd *bd,
361 struct hlist_node *hnd)
362 {
363 cfs_hash_dhead_t *dh = container_of(cfs_hash_dh_hhead(hs, bd),
364 cfs_hash_dhead_t, dh_head);
365
366 if (hnd->next == NULL) { /* it's the tail */
367 dh->dh_tail = (hnd->pprev == &dh->dh_head.first) ? NULL :
368 container_of(hnd->pprev, struct hlist_node, next);
369 }
370 hlist_del_init(hnd);
371 return -1; /* unknown depth */
372 }
373
374 /**
375 * double links hash head with depth tracking
376 * new element is always added to tail of hlist
377 */
378 typedef struct {
379 struct hlist_head dd_head; /**< entries list */
380 struct hlist_node *dd_tail; /**< the last entry */
381 unsigned int dd_depth; /**< list length */
382 } cfs_hash_dhead_dep_t;
383
384 static int
cfs_hash_dd_hhead_size(struct cfs_hash * hs)385 cfs_hash_dd_hhead_size(struct cfs_hash *hs)
386 {
387 return sizeof(cfs_hash_dhead_dep_t);
388 }
389
390 static struct hlist_head *
cfs_hash_dd_hhead(struct cfs_hash * hs,struct cfs_hash_bd * bd)391 cfs_hash_dd_hhead(struct cfs_hash *hs, struct cfs_hash_bd *bd)
392 {
393 cfs_hash_dhead_dep_t *head;
394
395 head = (cfs_hash_dhead_dep_t *)&bd->bd_bucket->hsb_head[0];
396 return &head[bd->bd_offset].dd_head;
397 }
398
399 static int
cfs_hash_dd_hnode_add(struct cfs_hash * hs,struct cfs_hash_bd * bd,struct hlist_node * hnode)400 cfs_hash_dd_hnode_add(struct cfs_hash *hs, struct cfs_hash_bd *bd,
401 struct hlist_node *hnode)
402 {
403 cfs_hash_dhead_dep_t *dh = container_of(cfs_hash_dd_hhead(hs, bd),
404 cfs_hash_dhead_dep_t, dd_head);
405
406 if (dh->dd_tail != NULL) /* not empty */
407 hlist_add_behind(hnode, dh->dd_tail);
408 else /* empty list */
409 hlist_add_head(hnode, &dh->dd_head);
410 dh->dd_tail = hnode;
411 return ++dh->dd_depth;
412 }
413
414 static int
cfs_hash_dd_hnode_del(struct cfs_hash * hs,struct cfs_hash_bd * bd,struct hlist_node * hnd)415 cfs_hash_dd_hnode_del(struct cfs_hash *hs, struct cfs_hash_bd *bd,
416 struct hlist_node *hnd)
417 {
418 cfs_hash_dhead_dep_t *dh = container_of(cfs_hash_dd_hhead(hs, bd),
419 cfs_hash_dhead_dep_t, dd_head);
420
421 if (hnd->next == NULL) { /* it's the tail */
422 dh->dd_tail = (hnd->pprev == &dh->dd_head.first) ? NULL :
423 container_of(hnd->pprev, struct hlist_node, next);
424 }
425 hlist_del_init(hnd);
426 return --dh->dd_depth;
427 }
428
429 static cfs_hash_hlist_ops_t cfs_hash_hh_hops = {
430 .hop_hhead = cfs_hash_hh_hhead,
431 .hop_hhead_size = cfs_hash_hh_hhead_size,
432 .hop_hnode_add = cfs_hash_hh_hnode_add,
433 .hop_hnode_del = cfs_hash_hh_hnode_del,
434 };
435
436 static cfs_hash_hlist_ops_t cfs_hash_hd_hops = {
437 .hop_hhead = cfs_hash_hd_hhead,
438 .hop_hhead_size = cfs_hash_hd_hhead_size,
439 .hop_hnode_add = cfs_hash_hd_hnode_add,
440 .hop_hnode_del = cfs_hash_hd_hnode_del,
441 };
442
443 static cfs_hash_hlist_ops_t cfs_hash_dh_hops = {
444 .hop_hhead = cfs_hash_dh_hhead,
445 .hop_hhead_size = cfs_hash_dh_hhead_size,
446 .hop_hnode_add = cfs_hash_dh_hnode_add,
447 .hop_hnode_del = cfs_hash_dh_hnode_del,
448 };
449
450 static cfs_hash_hlist_ops_t cfs_hash_dd_hops = {
451 .hop_hhead = cfs_hash_dd_hhead,
452 .hop_hhead_size = cfs_hash_dd_hhead_size,
453 .hop_hnode_add = cfs_hash_dd_hnode_add,
454 .hop_hnode_del = cfs_hash_dd_hnode_del,
455 };
456
457 static void
cfs_hash_hlist_setup(struct cfs_hash * hs)458 cfs_hash_hlist_setup(struct cfs_hash *hs)
459 {
460 if (cfs_hash_with_add_tail(hs)) {
461 hs->hs_hops = cfs_hash_with_depth(hs) ?
462 &cfs_hash_dd_hops : &cfs_hash_dh_hops;
463 } else {
464 hs->hs_hops = cfs_hash_with_depth(hs) ?
465 &cfs_hash_hd_hops : &cfs_hash_hh_hops;
466 }
467 }
468
469 static void
cfs_hash_bd_from_key(struct cfs_hash * hs,struct cfs_hash_bucket ** bkts,unsigned int bits,const void * key,struct cfs_hash_bd * bd)470 cfs_hash_bd_from_key(struct cfs_hash *hs, struct cfs_hash_bucket **bkts,
471 unsigned int bits, const void *key, struct cfs_hash_bd *bd)
472 {
473 unsigned int index = cfs_hash_id(hs, key, (1U << bits) - 1);
474
475 LASSERT(bits == hs->hs_cur_bits || bits == hs->hs_rehash_bits);
476
477 bd->bd_bucket = bkts[index & ((1U << (bits - hs->hs_bkt_bits)) - 1)];
478 bd->bd_offset = index >> (bits - hs->hs_bkt_bits);
479 }
480
481 void
cfs_hash_bd_get(struct cfs_hash * hs,const void * key,struct cfs_hash_bd * bd)482 cfs_hash_bd_get(struct cfs_hash *hs, const void *key, struct cfs_hash_bd *bd)
483 {
484 /* NB: caller should hold hs->hs_rwlock if REHASH is set */
485 if (likely(hs->hs_rehash_buckets == NULL)) {
486 cfs_hash_bd_from_key(hs, hs->hs_buckets,
487 hs->hs_cur_bits, key, bd);
488 } else {
489 LASSERT(hs->hs_rehash_bits != 0);
490 cfs_hash_bd_from_key(hs, hs->hs_rehash_buckets,
491 hs->hs_rehash_bits, key, bd);
492 }
493 }
494 EXPORT_SYMBOL(cfs_hash_bd_get);
495
496 static inline void
cfs_hash_bd_dep_record(struct cfs_hash * hs,struct cfs_hash_bd * bd,int dep_cur)497 cfs_hash_bd_dep_record(struct cfs_hash *hs, struct cfs_hash_bd *bd, int dep_cur)
498 {
499 if (likely(dep_cur <= bd->bd_bucket->hsb_depmax))
500 return;
501
502 bd->bd_bucket->hsb_depmax = dep_cur;
503 # if CFS_HASH_DEBUG_LEVEL >= CFS_HASH_DEBUG_1
504 if (likely(warn_on_depth == 0 ||
505 max(warn_on_depth, hs->hs_dep_max) >= dep_cur))
506 return;
507
508 spin_lock(&hs->hs_dep_lock);
509 hs->hs_dep_max = dep_cur;
510 hs->hs_dep_bkt = bd->bd_bucket->hsb_index;
511 hs->hs_dep_off = bd->bd_offset;
512 hs->hs_dep_bits = hs->hs_cur_bits;
513 spin_unlock(&hs->hs_dep_lock);
514
515 cfs_wi_schedule(cfs_sched_rehash, &hs->hs_dep_wi);
516 # endif
517 }
518
519 void
cfs_hash_bd_add_locked(struct cfs_hash * hs,struct cfs_hash_bd * bd,struct hlist_node * hnode)520 cfs_hash_bd_add_locked(struct cfs_hash *hs, struct cfs_hash_bd *bd,
521 struct hlist_node *hnode)
522 {
523 int rc;
524
525 rc = hs->hs_hops->hop_hnode_add(hs, bd, hnode);
526 cfs_hash_bd_dep_record(hs, bd, rc);
527 bd->bd_bucket->hsb_version++;
528 if (unlikely(bd->bd_bucket->hsb_version == 0))
529 bd->bd_bucket->hsb_version++;
530 bd->bd_bucket->hsb_count++;
531
532 if (cfs_hash_with_counter(hs))
533 atomic_inc(&hs->hs_count);
534 if (!cfs_hash_with_no_itemref(hs))
535 cfs_hash_get(hs, hnode);
536 }
537 EXPORT_SYMBOL(cfs_hash_bd_add_locked);
538
539 void
cfs_hash_bd_del_locked(struct cfs_hash * hs,struct cfs_hash_bd * bd,struct hlist_node * hnode)540 cfs_hash_bd_del_locked(struct cfs_hash *hs, struct cfs_hash_bd *bd,
541 struct hlist_node *hnode)
542 {
543 hs->hs_hops->hop_hnode_del(hs, bd, hnode);
544
545 LASSERT(bd->bd_bucket->hsb_count > 0);
546 bd->bd_bucket->hsb_count--;
547 bd->bd_bucket->hsb_version++;
548 if (unlikely(bd->bd_bucket->hsb_version == 0))
549 bd->bd_bucket->hsb_version++;
550
551 if (cfs_hash_with_counter(hs)) {
552 LASSERT(atomic_read(&hs->hs_count) > 0);
553 atomic_dec(&hs->hs_count);
554 }
555 if (!cfs_hash_with_no_itemref(hs))
556 cfs_hash_put_locked(hs, hnode);
557 }
558 EXPORT_SYMBOL(cfs_hash_bd_del_locked);
559
560 void
cfs_hash_bd_move_locked(struct cfs_hash * hs,struct cfs_hash_bd * bd_old,struct cfs_hash_bd * bd_new,struct hlist_node * hnode)561 cfs_hash_bd_move_locked(struct cfs_hash *hs, struct cfs_hash_bd *bd_old,
562 struct cfs_hash_bd *bd_new, struct hlist_node *hnode)
563 {
564 struct cfs_hash_bucket *obkt = bd_old->bd_bucket;
565 struct cfs_hash_bucket *nbkt = bd_new->bd_bucket;
566 int rc;
567
568 if (cfs_hash_bd_compare(bd_old, bd_new) == 0)
569 return;
570
571 /* use cfs_hash_bd_hnode_add/del, to avoid atomic & refcount ops
572 * in cfs_hash_bd_del/add_locked */
573 hs->hs_hops->hop_hnode_del(hs, bd_old, hnode);
574 rc = hs->hs_hops->hop_hnode_add(hs, bd_new, hnode);
575 cfs_hash_bd_dep_record(hs, bd_new, rc);
576
577 LASSERT(obkt->hsb_count > 0);
578 obkt->hsb_count--;
579 obkt->hsb_version++;
580 if (unlikely(obkt->hsb_version == 0))
581 obkt->hsb_version++;
582 nbkt->hsb_count++;
583 nbkt->hsb_version++;
584 if (unlikely(nbkt->hsb_version == 0))
585 nbkt->hsb_version++;
586 }
587 EXPORT_SYMBOL(cfs_hash_bd_move_locked);
588
589 enum {
590 /** always set, for sanity (avoid ZERO intent) */
591 CFS_HS_LOOKUP_MASK_FIND = 1 << 0,
592 /** return entry with a ref */
593 CFS_HS_LOOKUP_MASK_REF = 1 << 1,
594 /** add entry if not existing */
595 CFS_HS_LOOKUP_MASK_ADD = 1 << 2,
596 /** delete entry, ignore other masks */
597 CFS_HS_LOOKUP_MASK_DEL = 1 << 3,
598 };
599
600 typedef enum cfs_hash_lookup_intent {
601 /** return item w/o refcount */
602 CFS_HS_LOOKUP_IT_PEEK = CFS_HS_LOOKUP_MASK_FIND,
603 /** return item with refcount */
604 CFS_HS_LOOKUP_IT_FIND = (CFS_HS_LOOKUP_MASK_FIND |
605 CFS_HS_LOOKUP_MASK_REF),
606 /** return item w/o refcount if existed, otherwise add */
607 CFS_HS_LOOKUP_IT_ADD = (CFS_HS_LOOKUP_MASK_FIND |
608 CFS_HS_LOOKUP_MASK_ADD),
609 /** return item with refcount if existed, otherwise add */
610 CFS_HS_LOOKUP_IT_FINDADD = (CFS_HS_LOOKUP_IT_FIND |
611 CFS_HS_LOOKUP_MASK_ADD),
612 /** delete if existed */
613 CFS_HS_LOOKUP_IT_FINDDEL = (CFS_HS_LOOKUP_MASK_FIND |
614 CFS_HS_LOOKUP_MASK_DEL)
615 } cfs_hash_lookup_intent_t;
616
617 static struct hlist_node *
cfs_hash_bd_lookup_intent(struct cfs_hash * hs,struct cfs_hash_bd * bd,const void * key,struct hlist_node * hnode,cfs_hash_lookup_intent_t intent)618 cfs_hash_bd_lookup_intent(struct cfs_hash *hs, struct cfs_hash_bd *bd,
619 const void *key, struct hlist_node *hnode,
620 cfs_hash_lookup_intent_t intent)
621
622 {
623 struct hlist_head *hhead = cfs_hash_bd_hhead(hs, bd);
624 struct hlist_node *ehnode;
625 struct hlist_node *match;
626 int intent_add = (intent & CFS_HS_LOOKUP_MASK_ADD) != 0;
627
628 /* with this function, we can avoid a lot of useless refcount ops,
629 * which are expensive atomic operations most time. */
630 match = intent_add ? NULL : hnode;
631 hlist_for_each(ehnode, hhead) {
632 if (!cfs_hash_keycmp(hs, key, ehnode))
633 continue;
634
635 if (match != NULL && match != ehnode) /* can't match */
636 continue;
637
638 /* match and ... */
639 if ((intent & CFS_HS_LOOKUP_MASK_DEL) != 0) {
640 cfs_hash_bd_del_locked(hs, bd, ehnode);
641 return ehnode;
642 }
643
644 /* caller wants refcount? */
645 if ((intent & CFS_HS_LOOKUP_MASK_REF) != 0)
646 cfs_hash_get(hs, ehnode);
647 return ehnode;
648 }
649 /* no match item */
650 if (!intent_add)
651 return NULL;
652
653 LASSERT(hnode != NULL);
654 cfs_hash_bd_add_locked(hs, bd, hnode);
655 return hnode;
656 }
657
658 struct hlist_node *
cfs_hash_bd_lookup_locked(struct cfs_hash * hs,struct cfs_hash_bd * bd,const void * key)659 cfs_hash_bd_lookup_locked(struct cfs_hash *hs, struct cfs_hash_bd *bd, const void *key)
660 {
661 return cfs_hash_bd_lookup_intent(hs, bd, key, NULL,
662 CFS_HS_LOOKUP_IT_FIND);
663 }
664 EXPORT_SYMBOL(cfs_hash_bd_lookup_locked);
665
666 struct hlist_node *
cfs_hash_bd_peek_locked(struct cfs_hash * hs,struct cfs_hash_bd * bd,const void * key)667 cfs_hash_bd_peek_locked(struct cfs_hash *hs, struct cfs_hash_bd *bd, const void *key)
668 {
669 return cfs_hash_bd_lookup_intent(hs, bd, key, NULL,
670 CFS_HS_LOOKUP_IT_PEEK);
671 }
672 EXPORT_SYMBOL(cfs_hash_bd_peek_locked);
673
674 struct hlist_node *
cfs_hash_bd_findadd_locked(struct cfs_hash * hs,struct cfs_hash_bd * bd,const void * key,struct hlist_node * hnode,int noref)675 cfs_hash_bd_findadd_locked(struct cfs_hash *hs, struct cfs_hash_bd *bd,
676 const void *key, struct hlist_node *hnode,
677 int noref)
678 {
679 return cfs_hash_bd_lookup_intent(hs, bd, key, hnode,
680 CFS_HS_LOOKUP_IT_ADD |
681 (!noref * CFS_HS_LOOKUP_MASK_REF));
682 }
683 EXPORT_SYMBOL(cfs_hash_bd_findadd_locked);
684
685 struct hlist_node *
cfs_hash_bd_finddel_locked(struct cfs_hash * hs,struct cfs_hash_bd * bd,const void * key,struct hlist_node * hnode)686 cfs_hash_bd_finddel_locked(struct cfs_hash *hs, struct cfs_hash_bd *bd,
687 const void *key, struct hlist_node *hnode)
688 {
689 /* hnode can be NULL, we find the first item with @key */
690 return cfs_hash_bd_lookup_intent(hs, bd, key, hnode,
691 CFS_HS_LOOKUP_IT_FINDDEL);
692 }
693 EXPORT_SYMBOL(cfs_hash_bd_finddel_locked);
694
695 static void
cfs_hash_multi_bd_lock(struct cfs_hash * hs,struct cfs_hash_bd * bds,unsigned n,int excl)696 cfs_hash_multi_bd_lock(struct cfs_hash *hs, struct cfs_hash_bd *bds,
697 unsigned n, int excl)
698 {
699 struct cfs_hash_bucket *prev = NULL;
700 int i;
701
702 /**
703 * bds must be ascendantly ordered by bd->bd_bucket->hsb_index.
704 * NB: it's possible that several bds point to the same bucket but
705 * have different bd::bd_offset, so need take care of deadlock.
706 */
707 cfs_hash_for_each_bd(bds, n, i) {
708 if (prev == bds[i].bd_bucket)
709 continue;
710
711 LASSERT(prev == NULL ||
712 prev->hsb_index < bds[i].bd_bucket->hsb_index);
713 cfs_hash_bd_lock(hs, &bds[i], excl);
714 prev = bds[i].bd_bucket;
715 }
716 }
717
718 static void
cfs_hash_multi_bd_unlock(struct cfs_hash * hs,struct cfs_hash_bd * bds,unsigned n,int excl)719 cfs_hash_multi_bd_unlock(struct cfs_hash *hs, struct cfs_hash_bd *bds,
720 unsigned n, int excl)
721 {
722 struct cfs_hash_bucket *prev = NULL;
723 int i;
724
725 cfs_hash_for_each_bd(bds, n, i) {
726 if (prev != bds[i].bd_bucket) {
727 cfs_hash_bd_unlock(hs, &bds[i], excl);
728 prev = bds[i].bd_bucket;
729 }
730 }
731 }
732
733 static struct hlist_node *
cfs_hash_multi_bd_lookup_locked(struct cfs_hash * hs,struct cfs_hash_bd * bds,unsigned n,const void * key)734 cfs_hash_multi_bd_lookup_locked(struct cfs_hash *hs, struct cfs_hash_bd *bds,
735 unsigned n, const void *key)
736 {
737 struct hlist_node *ehnode;
738 unsigned i;
739
740 cfs_hash_for_each_bd(bds, n, i) {
741 ehnode = cfs_hash_bd_lookup_intent(hs, &bds[i], key, NULL,
742 CFS_HS_LOOKUP_IT_FIND);
743 if (ehnode != NULL)
744 return ehnode;
745 }
746 return NULL;
747 }
748
749 static struct hlist_node *
cfs_hash_multi_bd_findadd_locked(struct cfs_hash * hs,struct cfs_hash_bd * bds,unsigned n,const void * key,struct hlist_node * hnode,int noref)750 cfs_hash_multi_bd_findadd_locked(struct cfs_hash *hs,
751 struct cfs_hash_bd *bds, unsigned n, const void *key,
752 struct hlist_node *hnode, int noref)
753 {
754 struct hlist_node *ehnode;
755 int intent;
756 unsigned i;
757
758 LASSERT(hnode != NULL);
759 intent = CFS_HS_LOOKUP_IT_PEEK | (!noref * CFS_HS_LOOKUP_MASK_REF);
760
761 cfs_hash_for_each_bd(bds, n, i) {
762 ehnode = cfs_hash_bd_lookup_intent(hs, &bds[i], key,
763 NULL, intent);
764 if (ehnode != NULL)
765 return ehnode;
766 }
767
768 if (i == 1) { /* only one bucket */
769 cfs_hash_bd_add_locked(hs, &bds[0], hnode);
770 } else {
771 struct cfs_hash_bd mybd;
772
773 cfs_hash_bd_get(hs, key, &mybd);
774 cfs_hash_bd_add_locked(hs, &mybd, hnode);
775 }
776
777 return hnode;
778 }
779
780 static struct hlist_node *
cfs_hash_multi_bd_finddel_locked(struct cfs_hash * hs,struct cfs_hash_bd * bds,unsigned n,const void * key,struct hlist_node * hnode)781 cfs_hash_multi_bd_finddel_locked(struct cfs_hash *hs, struct cfs_hash_bd *bds,
782 unsigned n, const void *key,
783 struct hlist_node *hnode)
784 {
785 struct hlist_node *ehnode;
786 unsigned i;
787
788 cfs_hash_for_each_bd(bds, n, i) {
789 ehnode = cfs_hash_bd_lookup_intent(hs, &bds[i], key, hnode,
790 CFS_HS_LOOKUP_IT_FINDDEL);
791 if (ehnode != NULL)
792 return ehnode;
793 }
794 return NULL;
795 }
796
797 static void
cfs_hash_bd_order(struct cfs_hash_bd * bd1,struct cfs_hash_bd * bd2)798 cfs_hash_bd_order(struct cfs_hash_bd *bd1, struct cfs_hash_bd *bd2)
799 {
800 int rc;
801
802 if (bd2->bd_bucket == NULL)
803 return;
804
805 if (bd1->bd_bucket == NULL) {
806 *bd1 = *bd2;
807 bd2->bd_bucket = NULL;
808 return;
809 }
810
811 rc = cfs_hash_bd_compare(bd1, bd2);
812 if (rc == 0) {
813 bd2->bd_bucket = NULL;
814
815 } else if (rc > 0) { /* swab bd1 and bd2 */
816 struct cfs_hash_bd tmp;
817
818 tmp = *bd2;
819 *bd2 = *bd1;
820 *bd1 = tmp;
821 }
822 }
823
824 void
cfs_hash_dual_bd_get(struct cfs_hash * hs,const void * key,struct cfs_hash_bd * bds)825 cfs_hash_dual_bd_get(struct cfs_hash *hs, const void *key, struct cfs_hash_bd *bds)
826 {
827 /* NB: caller should hold hs_lock.rw if REHASH is set */
828 cfs_hash_bd_from_key(hs, hs->hs_buckets,
829 hs->hs_cur_bits, key, &bds[0]);
830 if (likely(hs->hs_rehash_buckets == NULL)) {
831 /* no rehash or not rehashing */
832 bds[1].bd_bucket = NULL;
833 return;
834 }
835
836 LASSERT(hs->hs_rehash_bits != 0);
837 cfs_hash_bd_from_key(hs, hs->hs_rehash_buckets,
838 hs->hs_rehash_bits, key, &bds[1]);
839
840 cfs_hash_bd_order(&bds[0], &bds[1]);
841 }
842 EXPORT_SYMBOL(cfs_hash_dual_bd_get);
843
844 void
cfs_hash_dual_bd_lock(struct cfs_hash * hs,struct cfs_hash_bd * bds,int excl)845 cfs_hash_dual_bd_lock(struct cfs_hash *hs, struct cfs_hash_bd *bds, int excl)
846 {
847 cfs_hash_multi_bd_lock(hs, bds, 2, excl);
848 }
849 EXPORT_SYMBOL(cfs_hash_dual_bd_lock);
850
851 void
cfs_hash_dual_bd_unlock(struct cfs_hash * hs,struct cfs_hash_bd * bds,int excl)852 cfs_hash_dual_bd_unlock(struct cfs_hash *hs, struct cfs_hash_bd *bds, int excl)
853 {
854 cfs_hash_multi_bd_unlock(hs, bds, 2, excl);
855 }
856 EXPORT_SYMBOL(cfs_hash_dual_bd_unlock);
857
858 struct hlist_node *
cfs_hash_dual_bd_lookup_locked(struct cfs_hash * hs,struct cfs_hash_bd * bds,const void * key)859 cfs_hash_dual_bd_lookup_locked(struct cfs_hash *hs, struct cfs_hash_bd *bds,
860 const void *key)
861 {
862 return cfs_hash_multi_bd_lookup_locked(hs, bds, 2, key);
863 }
864 EXPORT_SYMBOL(cfs_hash_dual_bd_lookup_locked);
865
866 struct hlist_node *
cfs_hash_dual_bd_findadd_locked(struct cfs_hash * hs,struct cfs_hash_bd * bds,const void * key,struct hlist_node * hnode,int noref)867 cfs_hash_dual_bd_findadd_locked(struct cfs_hash *hs, struct cfs_hash_bd *bds,
868 const void *key, struct hlist_node *hnode,
869 int noref)
870 {
871 return cfs_hash_multi_bd_findadd_locked(hs, bds, 2, key,
872 hnode, noref);
873 }
874 EXPORT_SYMBOL(cfs_hash_dual_bd_findadd_locked);
875
876 struct hlist_node *
cfs_hash_dual_bd_finddel_locked(struct cfs_hash * hs,struct cfs_hash_bd * bds,const void * key,struct hlist_node * hnode)877 cfs_hash_dual_bd_finddel_locked(struct cfs_hash *hs, struct cfs_hash_bd *bds,
878 const void *key, struct hlist_node *hnode)
879 {
880 return cfs_hash_multi_bd_finddel_locked(hs, bds, 2, key, hnode);
881 }
882 EXPORT_SYMBOL(cfs_hash_dual_bd_finddel_locked);
883
884 static void
cfs_hash_buckets_free(struct cfs_hash_bucket ** buckets,int bkt_size,int prev_size,int size)885 cfs_hash_buckets_free(struct cfs_hash_bucket **buckets,
886 int bkt_size, int prev_size, int size)
887 {
888 int i;
889
890 for (i = prev_size; i < size; i++) {
891 if (buckets[i] != NULL)
892 LIBCFS_FREE(buckets[i], bkt_size);
893 }
894
895 LIBCFS_FREE(buckets, sizeof(buckets[0]) * size);
896 }
897
898 /*
899 * Create or grow bucket memory. Return old_buckets if no allocation was
900 * needed, the newly allocated buckets if allocation was needed and
901 * successful, and NULL on error.
902 */
903 static struct cfs_hash_bucket **
cfs_hash_buckets_realloc(struct cfs_hash * hs,struct cfs_hash_bucket ** old_bkts,unsigned int old_size,unsigned int new_size)904 cfs_hash_buckets_realloc(struct cfs_hash *hs, struct cfs_hash_bucket **old_bkts,
905 unsigned int old_size, unsigned int new_size)
906 {
907 struct cfs_hash_bucket **new_bkts;
908 int i;
909
910 LASSERT(old_size == 0 || old_bkts != NULL);
911
912 if (old_bkts != NULL && old_size == new_size)
913 return old_bkts;
914
915 LIBCFS_ALLOC(new_bkts, sizeof(new_bkts[0]) * new_size);
916 if (new_bkts == NULL)
917 return NULL;
918
919 if (old_bkts != NULL) {
920 memcpy(new_bkts, old_bkts,
921 min(old_size, new_size) * sizeof(*old_bkts));
922 }
923
924 for (i = old_size; i < new_size; i++) {
925 struct hlist_head *hhead;
926 struct cfs_hash_bd bd;
927
928 LIBCFS_ALLOC(new_bkts[i], cfs_hash_bkt_size(hs));
929 if (new_bkts[i] == NULL) {
930 cfs_hash_buckets_free(new_bkts, cfs_hash_bkt_size(hs),
931 old_size, new_size);
932 return NULL;
933 }
934
935 new_bkts[i]->hsb_index = i;
936 new_bkts[i]->hsb_version = 1; /* shouldn't be zero */
937 new_bkts[i]->hsb_depmax = -1; /* unknown */
938 bd.bd_bucket = new_bkts[i];
939 cfs_hash_bd_for_each_hlist(hs, &bd, hhead)
940 INIT_HLIST_HEAD(hhead);
941
942 if (cfs_hash_with_no_lock(hs) ||
943 cfs_hash_with_no_bktlock(hs))
944 continue;
945
946 if (cfs_hash_with_rw_bktlock(hs))
947 rwlock_init(&new_bkts[i]->hsb_lock.rw);
948 else if (cfs_hash_with_spin_bktlock(hs))
949 spin_lock_init(&new_bkts[i]->hsb_lock.spin);
950 else
951 LBUG(); /* invalid use-case */
952 }
953 return new_bkts;
954 }
955
956 /**
957 * Initialize new libcfs hash, where:
958 * @name - Descriptive hash name
959 * @cur_bits - Initial hash table size, in bits
960 * @max_bits - Maximum allowed hash table resize, in bits
961 * @ops - Registered hash table operations
962 * @flags - CFS_HASH_REHASH enable synamic hash resizing
963 * - CFS_HASH_SORT enable chained hash sort
964 */
965 static int cfs_hash_rehash_worker(cfs_workitem_t *wi);
966
967 #if CFS_HASH_DEBUG_LEVEL >= CFS_HASH_DEBUG_1
cfs_hash_dep_print(cfs_workitem_t * wi)968 static int cfs_hash_dep_print(cfs_workitem_t *wi)
969 {
970 struct cfs_hash *hs = container_of(wi, struct cfs_hash, hs_dep_wi);
971 int dep;
972 int bkt;
973 int off;
974 int bits;
975
976 spin_lock(&hs->hs_dep_lock);
977 dep = hs->hs_dep_max;
978 bkt = hs->hs_dep_bkt;
979 off = hs->hs_dep_off;
980 bits = hs->hs_dep_bits;
981 spin_unlock(&hs->hs_dep_lock);
982
983 LCONSOLE_WARN("#### HASH %s (bits: %d): max depth %d at bucket %d/%d\n",
984 hs->hs_name, bits, dep, bkt, off);
985 spin_lock(&hs->hs_dep_lock);
986 hs->hs_dep_bits = 0; /* mark as workitem done */
987 spin_unlock(&hs->hs_dep_lock);
988 return 0;
989 }
990
cfs_hash_depth_wi_init(struct cfs_hash * hs)991 static void cfs_hash_depth_wi_init(struct cfs_hash *hs)
992 {
993 spin_lock_init(&hs->hs_dep_lock);
994 cfs_wi_init(&hs->hs_dep_wi, hs, cfs_hash_dep_print);
995 }
996
cfs_hash_depth_wi_cancel(struct cfs_hash * hs)997 static void cfs_hash_depth_wi_cancel(struct cfs_hash *hs)
998 {
999 if (cfs_wi_deschedule(cfs_sched_rehash, &hs->hs_dep_wi))
1000 return;
1001
1002 spin_lock(&hs->hs_dep_lock);
1003 while (hs->hs_dep_bits != 0) {
1004 spin_unlock(&hs->hs_dep_lock);
1005 cond_resched();
1006 spin_lock(&hs->hs_dep_lock);
1007 }
1008 spin_unlock(&hs->hs_dep_lock);
1009 }
1010
1011 #else /* CFS_HASH_DEBUG_LEVEL < CFS_HASH_DEBUG_1 */
1012
cfs_hash_depth_wi_init(struct cfs_hash * hs)1013 static inline void cfs_hash_depth_wi_init(struct cfs_hash *hs) {}
cfs_hash_depth_wi_cancel(struct cfs_hash * hs)1014 static inline void cfs_hash_depth_wi_cancel(struct cfs_hash *hs) {}
1015
1016 #endif /* CFS_HASH_DEBUG_LEVEL >= CFS_HASH_DEBUG_1 */
1017
1018 struct cfs_hash *
cfs_hash_create(char * name,unsigned cur_bits,unsigned max_bits,unsigned bkt_bits,unsigned extra_bytes,unsigned min_theta,unsigned max_theta,cfs_hash_ops_t * ops,unsigned flags)1019 cfs_hash_create(char *name, unsigned cur_bits, unsigned max_bits,
1020 unsigned bkt_bits, unsigned extra_bytes,
1021 unsigned min_theta, unsigned max_theta,
1022 cfs_hash_ops_t *ops, unsigned flags)
1023 {
1024 struct cfs_hash *hs;
1025 int len;
1026
1027 CLASSERT(CFS_HASH_THETA_BITS < 15);
1028
1029 LASSERT(name != NULL);
1030 LASSERT(ops != NULL);
1031 LASSERT(ops->hs_key);
1032 LASSERT(ops->hs_hash);
1033 LASSERT(ops->hs_object);
1034 LASSERT(ops->hs_keycmp);
1035 LASSERT(ops->hs_get != NULL);
1036 LASSERT(ops->hs_put_locked != NULL);
1037
1038 if ((flags & CFS_HASH_REHASH) != 0)
1039 flags |= CFS_HASH_COUNTER; /* must have counter */
1040
1041 LASSERT(cur_bits > 0);
1042 LASSERT(cur_bits >= bkt_bits);
1043 LASSERT(max_bits >= cur_bits && max_bits < 31);
1044 LASSERT(ergo((flags & CFS_HASH_REHASH) == 0, cur_bits == max_bits));
1045 LASSERT(ergo((flags & CFS_HASH_REHASH) != 0,
1046 (flags & CFS_HASH_NO_LOCK) == 0));
1047 LASSERT(ergo((flags & CFS_HASH_REHASH_KEY) != 0,
1048 ops->hs_keycpy != NULL));
1049
1050 len = (flags & CFS_HASH_BIGNAME) == 0 ?
1051 CFS_HASH_NAME_LEN : CFS_HASH_BIGNAME_LEN;
1052 LIBCFS_ALLOC(hs, offsetof(struct cfs_hash, hs_name[len]));
1053 if (hs == NULL)
1054 return NULL;
1055
1056 strncpy(hs->hs_name, name, len);
1057 hs->hs_name[len - 1] = '\0';
1058 hs->hs_flags = flags;
1059
1060 atomic_set(&hs->hs_refcount, 1);
1061 atomic_set(&hs->hs_count, 0);
1062
1063 cfs_hash_lock_setup(hs);
1064 cfs_hash_hlist_setup(hs);
1065
1066 hs->hs_cur_bits = (__u8)cur_bits;
1067 hs->hs_min_bits = (__u8)cur_bits;
1068 hs->hs_max_bits = (__u8)max_bits;
1069 hs->hs_bkt_bits = (__u8)bkt_bits;
1070
1071 hs->hs_ops = ops;
1072 hs->hs_extra_bytes = extra_bytes;
1073 hs->hs_rehash_bits = 0;
1074 cfs_wi_init(&hs->hs_rehash_wi, hs, cfs_hash_rehash_worker);
1075 cfs_hash_depth_wi_init(hs);
1076
1077 if (cfs_hash_with_rehash(hs))
1078 __cfs_hash_set_theta(hs, min_theta, max_theta);
1079
1080 hs->hs_buckets = cfs_hash_buckets_realloc(hs, NULL, 0,
1081 CFS_HASH_NBKT(hs));
1082 if (hs->hs_buckets != NULL)
1083 return hs;
1084
1085 LIBCFS_FREE(hs, offsetof(struct cfs_hash, hs_name[len]));
1086 return NULL;
1087 }
1088 EXPORT_SYMBOL(cfs_hash_create);
1089
1090 /**
1091 * Cleanup libcfs hash @hs.
1092 */
1093 static void
cfs_hash_destroy(struct cfs_hash * hs)1094 cfs_hash_destroy(struct cfs_hash *hs)
1095 {
1096 struct hlist_node *hnode;
1097 struct hlist_node *pos;
1098 struct cfs_hash_bd bd;
1099 int i;
1100
1101 LASSERT(hs != NULL);
1102 LASSERT(!cfs_hash_is_exiting(hs) &&
1103 !cfs_hash_is_iterating(hs));
1104
1105 /**
1106 * prohibit further rehashes, don't need any lock because
1107 * I'm the only (last) one can change it.
1108 */
1109 hs->hs_exiting = 1;
1110 if (cfs_hash_with_rehash(hs))
1111 cfs_hash_rehash_cancel(hs);
1112
1113 cfs_hash_depth_wi_cancel(hs);
1114 /* rehash should be done/canceled */
1115 LASSERT(hs->hs_buckets != NULL &&
1116 hs->hs_rehash_buckets == NULL);
1117
1118 cfs_hash_for_each_bucket(hs, &bd, i) {
1119 struct hlist_head *hhead;
1120
1121 LASSERT(bd.bd_bucket != NULL);
1122 /* no need to take this lock, just for consistent code */
1123 cfs_hash_bd_lock(hs, &bd, 1);
1124
1125 cfs_hash_bd_for_each_hlist(hs, &bd, hhead) {
1126 hlist_for_each_safe(hnode, pos, hhead) {
1127 LASSERTF(!cfs_hash_with_assert_empty(hs),
1128 "hash %s bucket %u(%u) is not empty: %u items left\n",
1129 hs->hs_name, bd.bd_bucket->hsb_index,
1130 bd.bd_offset, bd.bd_bucket->hsb_count);
1131 /* can't assert key valicate, because we
1132 * can interrupt rehash */
1133 cfs_hash_bd_del_locked(hs, &bd, hnode);
1134 cfs_hash_exit(hs, hnode);
1135 }
1136 }
1137 LASSERT(bd.bd_bucket->hsb_count == 0);
1138 cfs_hash_bd_unlock(hs, &bd, 1);
1139 cond_resched();
1140 }
1141
1142 LASSERT(atomic_read(&hs->hs_count) == 0);
1143
1144 cfs_hash_buckets_free(hs->hs_buckets, cfs_hash_bkt_size(hs),
1145 0, CFS_HASH_NBKT(hs));
1146 i = cfs_hash_with_bigname(hs) ?
1147 CFS_HASH_BIGNAME_LEN : CFS_HASH_NAME_LEN;
1148 LIBCFS_FREE(hs, offsetof(struct cfs_hash, hs_name[i]));
1149 }
1150
cfs_hash_getref(struct cfs_hash * hs)1151 struct cfs_hash *cfs_hash_getref(struct cfs_hash *hs)
1152 {
1153 if (atomic_inc_not_zero(&hs->hs_refcount))
1154 return hs;
1155 return NULL;
1156 }
1157 EXPORT_SYMBOL(cfs_hash_getref);
1158
cfs_hash_putref(struct cfs_hash * hs)1159 void cfs_hash_putref(struct cfs_hash *hs)
1160 {
1161 if (atomic_dec_and_test(&hs->hs_refcount))
1162 cfs_hash_destroy(hs);
1163 }
1164 EXPORT_SYMBOL(cfs_hash_putref);
1165
1166 static inline int
cfs_hash_rehash_bits(struct cfs_hash * hs)1167 cfs_hash_rehash_bits(struct cfs_hash *hs)
1168 {
1169 if (cfs_hash_with_no_lock(hs) ||
1170 !cfs_hash_with_rehash(hs))
1171 return -EOPNOTSUPP;
1172
1173 if (unlikely(cfs_hash_is_exiting(hs)))
1174 return -ESRCH;
1175
1176 if (unlikely(cfs_hash_is_rehashing(hs)))
1177 return -EALREADY;
1178
1179 if (unlikely(cfs_hash_is_iterating(hs)))
1180 return -EAGAIN;
1181
1182 /* XXX: need to handle case with max_theta != 2.0
1183 * and the case with min_theta != 0.5 */
1184 if ((hs->hs_cur_bits < hs->hs_max_bits) &&
1185 (__cfs_hash_theta(hs) > hs->hs_max_theta))
1186 return hs->hs_cur_bits + 1;
1187
1188 if (!cfs_hash_with_shrink(hs))
1189 return 0;
1190
1191 if ((hs->hs_cur_bits > hs->hs_min_bits) &&
1192 (__cfs_hash_theta(hs) < hs->hs_min_theta))
1193 return hs->hs_cur_bits - 1;
1194
1195 return 0;
1196 }
1197
1198 /**
1199 * don't allow inline rehash if:
1200 * - user wants non-blocking change (add/del) on hash table
1201 * - too many elements
1202 */
1203 static inline int
cfs_hash_rehash_inline(struct cfs_hash * hs)1204 cfs_hash_rehash_inline(struct cfs_hash *hs)
1205 {
1206 return !cfs_hash_with_nblk_change(hs) &&
1207 atomic_read(&hs->hs_count) < CFS_HASH_LOOP_HOG;
1208 }
1209
1210 /**
1211 * Add item @hnode to libcfs hash @hs using @key. The registered
1212 * ops->hs_get function will be called when the item is added.
1213 */
1214 void
cfs_hash_add(struct cfs_hash * hs,const void * key,struct hlist_node * hnode)1215 cfs_hash_add(struct cfs_hash *hs, const void *key, struct hlist_node *hnode)
1216 {
1217 struct cfs_hash_bd bd;
1218 int bits;
1219
1220 LASSERT(hlist_unhashed(hnode));
1221
1222 cfs_hash_lock(hs, 0);
1223 cfs_hash_bd_get_and_lock(hs, key, &bd, 1);
1224
1225 cfs_hash_key_validate(hs, key, hnode);
1226 cfs_hash_bd_add_locked(hs, &bd, hnode);
1227
1228 cfs_hash_bd_unlock(hs, &bd, 1);
1229
1230 bits = cfs_hash_rehash_bits(hs);
1231 cfs_hash_unlock(hs, 0);
1232 if (bits > 0)
1233 cfs_hash_rehash(hs, cfs_hash_rehash_inline(hs));
1234 }
1235 EXPORT_SYMBOL(cfs_hash_add);
1236
1237 static struct hlist_node *
cfs_hash_find_or_add(struct cfs_hash * hs,const void * key,struct hlist_node * hnode,int noref)1238 cfs_hash_find_or_add(struct cfs_hash *hs, const void *key,
1239 struct hlist_node *hnode, int noref)
1240 {
1241 struct hlist_node *ehnode;
1242 struct cfs_hash_bd bds[2];
1243 int bits = 0;
1244
1245 LASSERT(hlist_unhashed(hnode));
1246
1247 cfs_hash_lock(hs, 0);
1248 cfs_hash_dual_bd_get_and_lock(hs, key, bds, 1);
1249
1250 cfs_hash_key_validate(hs, key, hnode);
1251 ehnode = cfs_hash_dual_bd_findadd_locked(hs, bds, key,
1252 hnode, noref);
1253 cfs_hash_dual_bd_unlock(hs, bds, 1);
1254
1255 if (ehnode == hnode) /* new item added */
1256 bits = cfs_hash_rehash_bits(hs);
1257 cfs_hash_unlock(hs, 0);
1258 if (bits > 0)
1259 cfs_hash_rehash(hs, cfs_hash_rehash_inline(hs));
1260
1261 return ehnode;
1262 }
1263
1264 /**
1265 * Add item @hnode to libcfs hash @hs using @key. The registered
1266 * ops->hs_get function will be called if the item was added.
1267 * Returns 0 on success or -EALREADY on key collisions.
1268 */
1269 int
cfs_hash_add_unique(struct cfs_hash * hs,const void * key,struct hlist_node * hnode)1270 cfs_hash_add_unique(struct cfs_hash *hs, const void *key, struct hlist_node *hnode)
1271 {
1272 return cfs_hash_find_or_add(hs, key, hnode, 1) != hnode ?
1273 -EALREADY : 0;
1274 }
1275 EXPORT_SYMBOL(cfs_hash_add_unique);
1276
1277 /**
1278 * Add item @hnode to libcfs hash @hs using @key. If this @key
1279 * already exists in the hash then ops->hs_get will be called on the
1280 * conflicting entry and that entry will be returned to the caller.
1281 * Otherwise ops->hs_get is called on the item which was added.
1282 */
1283 void *
cfs_hash_findadd_unique(struct cfs_hash * hs,const void * key,struct hlist_node * hnode)1284 cfs_hash_findadd_unique(struct cfs_hash *hs, const void *key,
1285 struct hlist_node *hnode)
1286 {
1287 hnode = cfs_hash_find_or_add(hs, key, hnode, 0);
1288
1289 return cfs_hash_object(hs, hnode);
1290 }
1291 EXPORT_SYMBOL(cfs_hash_findadd_unique);
1292
1293 /**
1294 * Delete item @hnode from the libcfs hash @hs using @key. The @key
1295 * is required to ensure the correct hash bucket is locked since there
1296 * is no direct linkage from the item to the bucket. The object
1297 * removed from the hash will be returned and obs->hs_put is called
1298 * on the removed object.
1299 */
1300 void *
cfs_hash_del(struct cfs_hash * hs,const void * key,struct hlist_node * hnode)1301 cfs_hash_del(struct cfs_hash *hs, const void *key, struct hlist_node *hnode)
1302 {
1303 void *obj = NULL;
1304 int bits = 0;
1305 struct cfs_hash_bd bds[2];
1306
1307 cfs_hash_lock(hs, 0);
1308 cfs_hash_dual_bd_get_and_lock(hs, key, bds, 1);
1309
1310 /* NB: do nothing if @hnode is not in hash table */
1311 if (hnode == NULL || !hlist_unhashed(hnode)) {
1312 if (bds[1].bd_bucket == NULL && hnode != NULL) {
1313 cfs_hash_bd_del_locked(hs, &bds[0], hnode);
1314 } else {
1315 hnode = cfs_hash_dual_bd_finddel_locked(hs, bds,
1316 key, hnode);
1317 }
1318 }
1319
1320 if (hnode != NULL) {
1321 obj = cfs_hash_object(hs, hnode);
1322 bits = cfs_hash_rehash_bits(hs);
1323 }
1324
1325 cfs_hash_dual_bd_unlock(hs, bds, 1);
1326 cfs_hash_unlock(hs, 0);
1327 if (bits > 0)
1328 cfs_hash_rehash(hs, cfs_hash_rehash_inline(hs));
1329
1330 return obj;
1331 }
1332 EXPORT_SYMBOL(cfs_hash_del);
1333
1334 /**
1335 * Delete item given @key in libcfs hash @hs. The first @key found in
1336 * the hash will be removed, if the key exists multiple times in the hash
1337 * @hs this function must be called once per key. The removed object
1338 * will be returned and ops->hs_put is called on the removed object.
1339 */
1340 void *
cfs_hash_del_key(struct cfs_hash * hs,const void * key)1341 cfs_hash_del_key(struct cfs_hash *hs, const void *key)
1342 {
1343 return cfs_hash_del(hs, key, NULL);
1344 }
1345 EXPORT_SYMBOL(cfs_hash_del_key);
1346
1347 /**
1348 * Lookup an item using @key in the libcfs hash @hs and return it.
1349 * If the @key is found in the hash hs->hs_get() is called and the
1350 * matching objects is returned. It is the callers responsibility
1351 * to call the counterpart ops->hs_put using the cfs_hash_put() macro
1352 * when when finished with the object. If the @key was not found
1353 * in the hash @hs NULL is returned.
1354 */
1355 void *
cfs_hash_lookup(struct cfs_hash * hs,const void * key)1356 cfs_hash_lookup(struct cfs_hash *hs, const void *key)
1357 {
1358 void *obj = NULL;
1359 struct hlist_node *hnode;
1360 struct cfs_hash_bd bds[2];
1361
1362 cfs_hash_lock(hs, 0);
1363 cfs_hash_dual_bd_get_and_lock(hs, key, bds, 0);
1364
1365 hnode = cfs_hash_dual_bd_lookup_locked(hs, bds, key);
1366 if (hnode != NULL)
1367 obj = cfs_hash_object(hs, hnode);
1368
1369 cfs_hash_dual_bd_unlock(hs, bds, 0);
1370 cfs_hash_unlock(hs, 0);
1371
1372 return obj;
1373 }
1374 EXPORT_SYMBOL(cfs_hash_lookup);
1375
1376 static void
cfs_hash_for_each_enter(struct cfs_hash * hs)1377 cfs_hash_for_each_enter(struct cfs_hash *hs) {
1378 LASSERT(!cfs_hash_is_exiting(hs));
1379
1380 if (!cfs_hash_with_rehash(hs))
1381 return;
1382 /*
1383 * NB: it's race on cfs_has_t::hs_iterating, but doesn't matter
1384 * because it's just an unreliable signal to rehash-thread,
1385 * rehash-thread will try to finish rehash ASAP when seeing this.
1386 */
1387 hs->hs_iterating = 1;
1388
1389 cfs_hash_lock(hs, 1);
1390 hs->hs_iterators++;
1391
1392 /* NB: iteration is mostly called by service thread,
1393 * we tend to cancel pending rehash-request, instead of
1394 * blocking service thread, we will relaunch rehash request
1395 * after iteration */
1396 if (cfs_hash_is_rehashing(hs))
1397 cfs_hash_rehash_cancel_locked(hs);
1398 cfs_hash_unlock(hs, 1);
1399 }
1400
1401 static void
cfs_hash_for_each_exit(struct cfs_hash * hs)1402 cfs_hash_for_each_exit(struct cfs_hash *hs) {
1403 int remained;
1404 int bits;
1405
1406 if (!cfs_hash_with_rehash(hs))
1407 return;
1408 cfs_hash_lock(hs, 1);
1409 remained = --hs->hs_iterators;
1410 bits = cfs_hash_rehash_bits(hs);
1411 cfs_hash_unlock(hs, 1);
1412 /* NB: it's race on cfs_has_t::hs_iterating, see above */
1413 if (remained == 0)
1414 hs->hs_iterating = 0;
1415 if (bits > 0) {
1416 cfs_hash_rehash(hs, atomic_read(&hs->hs_count) <
1417 CFS_HASH_LOOP_HOG);
1418 }
1419 }
1420
1421 /**
1422 * For each item in the libcfs hash @hs call the passed callback @func
1423 * and pass to it as an argument each hash item and the private @data.
1424 *
1425 * a) the function may sleep!
1426 * b) during the callback:
1427 * . the bucket lock is held so the callback must never sleep.
1428 * . if @removal_safe is true, use can remove current item by
1429 * cfs_hash_bd_del_locked
1430 */
1431 static __u64
cfs_hash_for_each_tight(struct cfs_hash * hs,cfs_hash_for_each_cb_t func,void * data,int remove_safe)1432 cfs_hash_for_each_tight(struct cfs_hash *hs, cfs_hash_for_each_cb_t func,
1433 void *data, int remove_safe) {
1434 struct hlist_node *hnode;
1435 struct hlist_node *pos;
1436 struct cfs_hash_bd bd;
1437 __u64 count = 0;
1438 int excl = !!remove_safe;
1439 int loop = 0;
1440 int i;
1441
1442 cfs_hash_for_each_enter(hs);
1443
1444 cfs_hash_lock(hs, 0);
1445 LASSERT(!cfs_hash_is_rehashing(hs));
1446
1447 cfs_hash_for_each_bucket(hs, &bd, i) {
1448 struct hlist_head *hhead;
1449
1450 cfs_hash_bd_lock(hs, &bd, excl);
1451 if (func == NULL) { /* only glimpse size */
1452 count += bd.bd_bucket->hsb_count;
1453 cfs_hash_bd_unlock(hs, &bd, excl);
1454 continue;
1455 }
1456
1457 cfs_hash_bd_for_each_hlist(hs, &bd, hhead) {
1458 hlist_for_each_safe(hnode, pos, hhead) {
1459 cfs_hash_bucket_validate(hs, &bd, hnode);
1460 count++;
1461 loop++;
1462 if (func(hs, &bd, hnode, data)) {
1463 cfs_hash_bd_unlock(hs, &bd, excl);
1464 goto out;
1465 }
1466 }
1467 }
1468 cfs_hash_bd_unlock(hs, &bd, excl);
1469 if (loop < CFS_HASH_LOOP_HOG)
1470 continue;
1471 loop = 0;
1472 cfs_hash_unlock(hs, 0);
1473 cond_resched();
1474 cfs_hash_lock(hs, 0);
1475 }
1476 out:
1477 cfs_hash_unlock(hs, 0);
1478
1479 cfs_hash_for_each_exit(hs);
1480 return count;
1481 }
1482
1483 typedef struct {
1484 cfs_hash_cond_opt_cb_t func;
1485 void *arg;
1486 } cfs_hash_cond_arg_t;
1487
1488 static int
cfs_hash_cond_del_locked(struct cfs_hash * hs,struct cfs_hash_bd * bd,struct hlist_node * hnode,void * data)1489 cfs_hash_cond_del_locked(struct cfs_hash *hs, struct cfs_hash_bd *bd,
1490 struct hlist_node *hnode, void *data)
1491 {
1492 cfs_hash_cond_arg_t *cond = data;
1493
1494 if (cond->func(cfs_hash_object(hs, hnode), cond->arg))
1495 cfs_hash_bd_del_locked(hs, bd, hnode);
1496 return 0;
1497 }
1498
1499 /**
1500 * Delete item from the libcfs hash @hs when @func return true.
1501 * The write lock being hold during loop for each bucket to avoid
1502 * any object be reference.
1503 */
1504 void
cfs_hash_cond_del(struct cfs_hash * hs,cfs_hash_cond_opt_cb_t func,void * data)1505 cfs_hash_cond_del(struct cfs_hash *hs, cfs_hash_cond_opt_cb_t func, void *data)
1506 {
1507 cfs_hash_cond_arg_t arg = {
1508 .func = func,
1509 .arg = data,
1510 };
1511
1512 cfs_hash_for_each_tight(hs, cfs_hash_cond_del_locked, &arg, 1);
1513 }
1514 EXPORT_SYMBOL(cfs_hash_cond_del);
1515
1516 void
cfs_hash_for_each(struct cfs_hash * hs,cfs_hash_for_each_cb_t func,void * data)1517 cfs_hash_for_each(struct cfs_hash *hs,
1518 cfs_hash_for_each_cb_t func, void *data)
1519 {
1520 cfs_hash_for_each_tight(hs, func, data, 0);
1521 }
1522 EXPORT_SYMBOL(cfs_hash_for_each);
1523
1524 void
cfs_hash_for_each_safe(struct cfs_hash * hs,cfs_hash_for_each_cb_t func,void * data)1525 cfs_hash_for_each_safe(struct cfs_hash *hs,
1526 cfs_hash_for_each_cb_t func, void *data) {
1527 cfs_hash_for_each_tight(hs, func, data, 1);
1528 }
1529 EXPORT_SYMBOL(cfs_hash_for_each_safe);
1530
1531 static int
cfs_hash_peek(struct cfs_hash * hs,struct cfs_hash_bd * bd,struct hlist_node * hnode,void * data)1532 cfs_hash_peek(struct cfs_hash *hs, struct cfs_hash_bd *bd,
1533 struct hlist_node *hnode, void *data)
1534 {
1535 *(int *)data = 0;
1536 return 1; /* return 1 to break the loop */
1537 }
1538
1539 int
cfs_hash_is_empty(struct cfs_hash * hs)1540 cfs_hash_is_empty(struct cfs_hash *hs)
1541 {
1542 int empty = 1;
1543
1544 cfs_hash_for_each_tight(hs, cfs_hash_peek, &empty, 0);
1545 return empty;
1546 }
1547 EXPORT_SYMBOL(cfs_hash_is_empty);
1548
1549 __u64
cfs_hash_size_get(struct cfs_hash * hs)1550 cfs_hash_size_get(struct cfs_hash *hs)
1551 {
1552 return cfs_hash_with_counter(hs) ?
1553 atomic_read(&hs->hs_count) :
1554 cfs_hash_for_each_tight(hs, NULL, NULL, 0);
1555 }
1556 EXPORT_SYMBOL(cfs_hash_size_get);
1557
1558 /*
1559 * cfs_hash_for_each_relax:
1560 * Iterate the hash table and call @func on each item without
1561 * any lock. This function can't guarantee to finish iteration
1562 * if these features are enabled:
1563 *
1564 * a. if rehash_key is enabled, an item can be moved from
1565 * one bucket to another bucket
1566 * b. user can remove non-zero-ref item from hash-table,
1567 * so the item can be removed from hash-table, even worse,
1568 * it's possible that user changed key and insert to another
1569 * hash bucket.
1570 * there's no way for us to finish iteration correctly on previous
1571 * two cases, so iteration has to be stopped on change.
1572 */
1573 static int
cfs_hash_for_each_relax(struct cfs_hash * hs,cfs_hash_for_each_cb_t func,void * data)1574 cfs_hash_for_each_relax(struct cfs_hash *hs, cfs_hash_for_each_cb_t func,
1575 void *data) {
1576 struct hlist_node *hnode;
1577 struct hlist_node *tmp;
1578 struct cfs_hash_bd bd;
1579 __u32 version;
1580 int count = 0;
1581 int stop_on_change;
1582 int rc;
1583 int i;
1584
1585 stop_on_change = cfs_hash_with_rehash_key(hs) ||
1586 !cfs_hash_with_no_itemref(hs) ||
1587 hs->hs_ops->hs_put_locked == NULL;
1588 cfs_hash_lock(hs, 0);
1589 LASSERT(!cfs_hash_is_rehashing(hs));
1590
1591 cfs_hash_for_each_bucket(hs, &bd, i) {
1592 struct hlist_head *hhead;
1593
1594 cfs_hash_bd_lock(hs, &bd, 0);
1595 version = cfs_hash_bd_version_get(&bd);
1596
1597 cfs_hash_bd_for_each_hlist(hs, &bd, hhead) {
1598 for (hnode = hhead->first; hnode != NULL;) {
1599 cfs_hash_bucket_validate(hs, &bd, hnode);
1600 cfs_hash_get(hs, hnode);
1601 cfs_hash_bd_unlock(hs, &bd, 0);
1602 cfs_hash_unlock(hs, 0);
1603
1604 rc = func(hs, &bd, hnode, data);
1605 if (stop_on_change)
1606 cfs_hash_put(hs, hnode);
1607 cond_resched();
1608 count++;
1609
1610 cfs_hash_lock(hs, 0);
1611 cfs_hash_bd_lock(hs, &bd, 0);
1612 if (!stop_on_change) {
1613 tmp = hnode->next;
1614 cfs_hash_put_locked(hs, hnode);
1615 hnode = tmp;
1616 } else { /* bucket changed? */
1617 if (version !=
1618 cfs_hash_bd_version_get(&bd))
1619 break;
1620 /* safe to continue because no change */
1621 hnode = hnode->next;
1622 }
1623 if (rc) /* callback wants to break iteration */
1624 break;
1625 }
1626 }
1627 cfs_hash_bd_unlock(hs, &bd, 0);
1628 }
1629 cfs_hash_unlock(hs, 0);
1630
1631 return count;
1632 }
1633
1634 int
cfs_hash_for_each_nolock(struct cfs_hash * hs,cfs_hash_for_each_cb_t func,void * data)1635 cfs_hash_for_each_nolock(struct cfs_hash *hs,
1636 cfs_hash_for_each_cb_t func, void *data) {
1637 if (cfs_hash_with_no_lock(hs) ||
1638 cfs_hash_with_rehash_key(hs) ||
1639 !cfs_hash_with_no_itemref(hs))
1640 return -EOPNOTSUPP;
1641
1642 if (hs->hs_ops->hs_get == NULL ||
1643 (hs->hs_ops->hs_put == NULL &&
1644 hs->hs_ops->hs_put_locked == NULL))
1645 return -EOPNOTSUPP;
1646
1647 cfs_hash_for_each_enter(hs);
1648 cfs_hash_for_each_relax(hs, func, data);
1649 cfs_hash_for_each_exit(hs);
1650
1651 return 0;
1652 }
1653 EXPORT_SYMBOL(cfs_hash_for_each_nolock);
1654
1655 /**
1656 * For each hash bucket in the libcfs hash @hs call the passed callback
1657 * @func until all the hash buckets are empty. The passed callback @func
1658 * or the previously registered callback hs->hs_put must remove the item
1659 * from the hash. You may either use the cfs_hash_del() or hlist_del()
1660 * functions. No rwlocks will be held during the callback @func it is
1661 * safe to sleep if needed. This function will not terminate until the
1662 * hash is empty. Note it is still possible to concurrently add new
1663 * items in to the hash. It is the callers responsibility to ensure
1664 * the required locking is in place to prevent concurrent insertions.
1665 */
1666 int
cfs_hash_for_each_empty(struct cfs_hash * hs,cfs_hash_for_each_cb_t func,void * data)1667 cfs_hash_for_each_empty(struct cfs_hash *hs,
1668 cfs_hash_for_each_cb_t func, void *data) {
1669 unsigned i = 0;
1670
1671 if (cfs_hash_with_no_lock(hs))
1672 return -EOPNOTSUPP;
1673
1674 if (hs->hs_ops->hs_get == NULL ||
1675 (hs->hs_ops->hs_put == NULL &&
1676 hs->hs_ops->hs_put_locked == NULL))
1677 return -EOPNOTSUPP;
1678
1679 cfs_hash_for_each_enter(hs);
1680 while (cfs_hash_for_each_relax(hs, func, data)) {
1681 CDEBUG(D_INFO, "Try to empty hash: %s, loop: %u\n",
1682 hs->hs_name, i++);
1683 }
1684 cfs_hash_for_each_exit(hs);
1685 return 0;
1686 }
1687 EXPORT_SYMBOL(cfs_hash_for_each_empty);
1688
1689 void
cfs_hash_hlist_for_each(struct cfs_hash * hs,unsigned hindex,cfs_hash_for_each_cb_t func,void * data)1690 cfs_hash_hlist_for_each(struct cfs_hash *hs, unsigned hindex,
1691 cfs_hash_for_each_cb_t func, void *data)
1692 {
1693 struct hlist_head *hhead;
1694 struct hlist_node *hnode;
1695 struct cfs_hash_bd bd;
1696
1697 cfs_hash_for_each_enter(hs);
1698 cfs_hash_lock(hs, 0);
1699 if (hindex >= CFS_HASH_NHLIST(hs))
1700 goto out;
1701
1702 cfs_hash_bd_index_set(hs, hindex, &bd);
1703
1704 cfs_hash_bd_lock(hs, &bd, 0);
1705 hhead = cfs_hash_bd_hhead(hs, &bd);
1706 hlist_for_each(hnode, hhead) {
1707 if (func(hs, &bd, hnode, data))
1708 break;
1709 }
1710 cfs_hash_bd_unlock(hs, &bd, 0);
1711 out:
1712 cfs_hash_unlock(hs, 0);
1713 cfs_hash_for_each_exit(hs);
1714 }
1715
1716 EXPORT_SYMBOL(cfs_hash_hlist_for_each);
1717
1718 /*
1719 * For each item in the libcfs hash @hs which matches the @key call
1720 * the passed callback @func and pass to it as an argument each hash
1721 * item and the private @data. During the callback the bucket lock
1722 * is held so the callback must never sleep.
1723 */
1724 void
cfs_hash_for_each_key(struct cfs_hash * hs,const void * key,cfs_hash_for_each_cb_t func,void * data)1725 cfs_hash_for_each_key(struct cfs_hash *hs, const void *key,
1726 cfs_hash_for_each_cb_t func, void *data) {
1727 struct hlist_node *hnode;
1728 struct cfs_hash_bd bds[2];
1729 unsigned i;
1730
1731 cfs_hash_lock(hs, 0);
1732
1733 cfs_hash_dual_bd_get_and_lock(hs, key, bds, 0);
1734
1735 cfs_hash_for_each_bd(bds, 2, i) {
1736 struct hlist_head *hlist = cfs_hash_bd_hhead(hs, &bds[i]);
1737
1738 hlist_for_each(hnode, hlist) {
1739 cfs_hash_bucket_validate(hs, &bds[i], hnode);
1740
1741 if (cfs_hash_keycmp(hs, key, hnode)) {
1742 if (func(hs, &bds[i], hnode, data))
1743 break;
1744 }
1745 }
1746 }
1747
1748 cfs_hash_dual_bd_unlock(hs, bds, 0);
1749 cfs_hash_unlock(hs, 0);
1750 }
1751 EXPORT_SYMBOL(cfs_hash_for_each_key);
1752
1753 /**
1754 * Rehash the libcfs hash @hs to the given @bits. This can be used
1755 * to grow the hash size when excessive chaining is detected, or to
1756 * shrink the hash when it is larger than needed. When the CFS_HASH_REHASH
1757 * flag is set in @hs the libcfs hash may be dynamically rehashed
1758 * during addition or removal if the hash's theta value exceeds
1759 * either the hs->hs_min_theta or hs->max_theta values. By default
1760 * these values are tuned to keep the chained hash depth small, and
1761 * this approach assumes a reasonably uniform hashing function. The
1762 * theta thresholds for @hs are tunable via cfs_hash_set_theta().
1763 */
1764 void
cfs_hash_rehash_cancel_locked(struct cfs_hash * hs)1765 cfs_hash_rehash_cancel_locked(struct cfs_hash *hs)
1766 {
1767 int i;
1768
1769 /* need hold cfs_hash_lock(hs, 1) */
1770 LASSERT(cfs_hash_with_rehash(hs) &&
1771 !cfs_hash_with_no_lock(hs));
1772
1773 if (!cfs_hash_is_rehashing(hs))
1774 return;
1775
1776 if (cfs_wi_deschedule(cfs_sched_rehash, &hs->hs_rehash_wi)) {
1777 hs->hs_rehash_bits = 0;
1778 return;
1779 }
1780
1781 for (i = 2; cfs_hash_is_rehashing(hs); i++) {
1782 cfs_hash_unlock(hs, 1);
1783 /* raise console warning while waiting too long */
1784 CDEBUG(IS_PO2(i >> 3) ? D_WARNING : D_INFO,
1785 "hash %s is still rehashing, rescheded %d\n",
1786 hs->hs_name, i - 1);
1787 cond_resched();
1788 cfs_hash_lock(hs, 1);
1789 }
1790 }
1791 EXPORT_SYMBOL(cfs_hash_rehash_cancel_locked);
1792
1793 void
cfs_hash_rehash_cancel(struct cfs_hash * hs)1794 cfs_hash_rehash_cancel(struct cfs_hash *hs)
1795 {
1796 cfs_hash_lock(hs, 1);
1797 cfs_hash_rehash_cancel_locked(hs);
1798 cfs_hash_unlock(hs, 1);
1799 }
1800 EXPORT_SYMBOL(cfs_hash_rehash_cancel);
1801
1802 int
cfs_hash_rehash(struct cfs_hash * hs,int do_rehash)1803 cfs_hash_rehash(struct cfs_hash *hs, int do_rehash)
1804 {
1805 int rc;
1806
1807 LASSERT(cfs_hash_with_rehash(hs) && !cfs_hash_with_no_lock(hs));
1808
1809 cfs_hash_lock(hs, 1);
1810
1811 rc = cfs_hash_rehash_bits(hs);
1812 if (rc <= 0) {
1813 cfs_hash_unlock(hs, 1);
1814 return rc;
1815 }
1816
1817 hs->hs_rehash_bits = rc;
1818 if (!do_rehash) {
1819 /* launch and return */
1820 cfs_wi_schedule(cfs_sched_rehash, &hs->hs_rehash_wi);
1821 cfs_hash_unlock(hs, 1);
1822 return 0;
1823 }
1824
1825 /* rehash right now */
1826 cfs_hash_unlock(hs, 1);
1827
1828 return cfs_hash_rehash_worker(&hs->hs_rehash_wi);
1829 }
1830 EXPORT_SYMBOL(cfs_hash_rehash);
1831
1832 static int
cfs_hash_rehash_bd(struct cfs_hash * hs,struct cfs_hash_bd * old)1833 cfs_hash_rehash_bd(struct cfs_hash *hs, struct cfs_hash_bd *old)
1834 {
1835 struct cfs_hash_bd new;
1836 struct hlist_head *hhead;
1837 struct hlist_node *hnode;
1838 struct hlist_node *pos;
1839 void *key;
1840 int c = 0;
1841
1842 /* hold cfs_hash_lock(hs, 1), so don't need any bucket lock */
1843 cfs_hash_bd_for_each_hlist(hs, old, hhead) {
1844 hlist_for_each_safe(hnode, pos, hhead) {
1845 key = cfs_hash_key(hs, hnode);
1846 LASSERT(key != NULL);
1847 /* Validate hnode is in the correct bucket. */
1848 cfs_hash_bucket_validate(hs, old, hnode);
1849 /*
1850 * Delete from old hash bucket; move to new bucket.
1851 * ops->hs_key must be defined.
1852 */
1853 cfs_hash_bd_from_key(hs, hs->hs_rehash_buckets,
1854 hs->hs_rehash_bits, key, &new);
1855 cfs_hash_bd_move_locked(hs, old, &new, hnode);
1856 c++;
1857 }
1858 }
1859
1860 return c;
1861 }
1862
1863 static int
cfs_hash_rehash_worker(cfs_workitem_t * wi)1864 cfs_hash_rehash_worker(cfs_workitem_t *wi)
1865 {
1866 struct cfs_hash *hs = container_of(wi, struct cfs_hash, hs_rehash_wi);
1867 struct cfs_hash_bucket **bkts;
1868 struct cfs_hash_bd bd;
1869 unsigned int old_size;
1870 unsigned int new_size;
1871 int bsize;
1872 int count = 0;
1873 int rc = 0;
1874 int i;
1875
1876 LASSERT (hs != NULL && cfs_hash_with_rehash(hs));
1877
1878 cfs_hash_lock(hs, 0);
1879 LASSERT(cfs_hash_is_rehashing(hs));
1880
1881 old_size = CFS_HASH_NBKT(hs);
1882 new_size = CFS_HASH_RH_NBKT(hs);
1883
1884 cfs_hash_unlock(hs, 0);
1885
1886 /*
1887 * don't need hs::hs_rwlock for hs::hs_buckets,
1888 * because nobody can change bkt-table except me.
1889 */
1890 bkts = cfs_hash_buckets_realloc(hs, hs->hs_buckets,
1891 old_size, new_size);
1892 cfs_hash_lock(hs, 1);
1893 if (bkts == NULL) {
1894 rc = -ENOMEM;
1895 goto out;
1896 }
1897
1898 if (bkts == hs->hs_buckets) {
1899 bkts = NULL; /* do nothing */
1900 goto out;
1901 }
1902
1903 rc = __cfs_hash_theta(hs);
1904 if ((rc >= hs->hs_min_theta) && (rc <= hs->hs_max_theta)) {
1905 /* free the new allocated bkt-table */
1906 old_size = new_size;
1907 new_size = CFS_HASH_NBKT(hs);
1908 rc = -EALREADY;
1909 goto out;
1910 }
1911
1912 LASSERT(hs->hs_rehash_buckets == NULL);
1913 hs->hs_rehash_buckets = bkts;
1914
1915 rc = 0;
1916 cfs_hash_for_each_bucket(hs, &bd, i) {
1917 if (cfs_hash_is_exiting(hs)) {
1918 rc = -ESRCH;
1919 /* someone wants to destroy the hash, abort now */
1920 if (old_size < new_size) /* OK to free old bkt-table */
1921 break;
1922 /* it's shrinking, need free new bkt-table */
1923 hs->hs_rehash_buckets = NULL;
1924 old_size = new_size;
1925 new_size = CFS_HASH_NBKT(hs);
1926 goto out;
1927 }
1928
1929 count += cfs_hash_rehash_bd(hs, &bd);
1930 if (count < CFS_HASH_LOOP_HOG ||
1931 cfs_hash_is_iterating(hs)) { /* need to finish ASAP */
1932 continue;
1933 }
1934
1935 count = 0;
1936 cfs_hash_unlock(hs, 1);
1937 cond_resched();
1938 cfs_hash_lock(hs, 1);
1939 }
1940
1941 hs->hs_rehash_count++;
1942
1943 bkts = hs->hs_buckets;
1944 hs->hs_buckets = hs->hs_rehash_buckets;
1945 hs->hs_rehash_buckets = NULL;
1946
1947 hs->hs_cur_bits = hs->hs_rehash_bits;
1948 out:
1949 hs->hs_rehash_bits = 0;
1950 if (rc == -ESRCH) /* never be scheduled again */
1951 cfs_wi_exit(cfs_sched_rehash, wi);
1952 bsize = cfs_hash_bkt_size(hs);
1953 cfs_hash_unlock(hs, 1);
1954 /* can't refer to @hs anymore because it could be destroyed */
1955 if (bkts != NULL)
1956 cfs_hash_buckets_free(bkts, bsize, new_size, old_size);
1957 if (rc != 0)
1958 CDEBUG(D_INFO, "early quit of rehashing: %d\n", rc);
1959 /* return 1 only if cfs_wi_exit is called */
1960 return rc == -ESRCH;
1961 }
1962
1963 /**
1964 * Rehash the object referenced by @hnode in the libcfs hash @hs. The
1965 * @old_key must be provided to locate the objects previous location
1966 * in the hash, and the @new_key will be used to reinsert the object.
1967 * Use this function instead of a cfs_hash_add() + cfs_hash_del()
1968 * combo when it is critical that there is no window in time where the
1969 * object is missing from the hash. When an object is being rehashed
1970 * the registered cfs_hash_get() and cfs_hash_put() functions will
1971 * not be called.
1972 */
cfs_hash_rehash_key(struct cfs_hash * hs,const void * old_key,void * new_key,struct hlist_node * hnode)1973 void cfs_hash_rehash_key(struct cfs_hash *hs, const void *old_key,
1974 void *new_key, struct hlist_node *hnode)
1975 {
1976 struct cfs_hash_bd bds[3];
1977 struct cfs_hash_bd old_bds[2];
1978 struct cfs_hash_bd new_bd;
1979
1980 LASSERT(!hlist_unhashed(hnode));
1981
1982 cfs_hash_lock(hs, 0);
1983
1984 cfs_hash_dual_bd_get(hs, old_key, old_bds);
1985 cfs_hash_bd_get(hs, new_key, &new_bd);
1986
1987 bds[0] = old_bds[0];
1988 bds[1] = old_bds[1];
1989 bds[2] = new_bd;
1990
1991 /* NB: bds[0] and bds[1] are ordered already */
1992 cfs_hash_bd_order(&bds[1], &bds[2]);
1993 cfs_hash_bd_order(&bds[0], &bds[1]);
1994
1995 cfs_hash_multi_bd_lock(hs, bds, 3, 1);
1996 if (likely(old_bds[1].bd_bucket == NULL)) {
1997 cfs_hash_bd_move_locked(hs, &old_bds[0], &new_bd, hnode);
1998 } else {
1999 cfs_hash_dual_bd_finddel_locked(hs, old_bds, old_key, hnode);
2000 cfs_hash_bd_add_locked(hs, &new_bd, hnode);
2001 }
2002 /* overwrite key inside locks, otherwise may screw up with
2003 * other operations, i.e: rehash */
2004 cfs_hash_keycpy(hs, new_key, hnode);
2005
2006 cfs_hash_multi_bd_unlock(hs, bds, 3, 1);
2007 cfs_hash_unlock(hs, 0);
2008 }
2009 EXPORT_SYMBOL(cfs_hash_rehash_key);
2010
cfs_hash_debug_header(struct seq_file * m)2011 void cfs_hash_debug_header(struct seq_file *m)
2012 {
2013 seq_printf(m, "%-*s cur min max theta t-min t-max flags rehash count maxdep maxdepb distribution\n",
2014 CFS_HASH_BIGNAME_LEN, "name");
2015 }
2016 EXPORT_SYMBOL(cfs_hash_debug_header);
2017
2018 static struct cfs_hash_bucket **
cfs_hash_full_bkts(struct cfs_hash * hs)2019 cfs_hash_full_bkts(struct cfs_hash *hs)
2020 {
2021 /* NB: caller should hold hs->hs_rwlock if REHASH is set */
2022 if (hs->hs_rehash_buckets == NULL)
2023 return hs->hs_buckets;
2024
2025 LASSERT(hs->hs_rehash_bits != 0);
2026 return hs->hs_rehash_bits > hs->hs_cur_bits ?
2027 hs->hs_rehash_buckets : hs->hs_buckets;
2028 }
2029
2030 static unsigned int
cfs_hash_full_nbkt(struct cfs_hash * hs)2031 cfs_hash_full_nbkt(struct cfs_hash *hs)
2032 {
2033 /* NB: caller should hold hs->hs_rwlock if REHASH is set */
2034 if (hs->hs_rehash_buckets == NULL)
2035 return CFS_HASH_NBKT(hs);
2036
2037 LASSERT(hs->hs_rehash_bits != 0);
2038 return hs->hs_rehash_bits > hs->hs_cur_bits ?
2039 CFS_HASH_RH_NBKT(hs) : CFS_HASH_NBKT(hs);
2040 }
2041
cfs_hash_debug_str(struct cfs_hash * hs,struct seq_file * m)2042 void cfs_hash_debug_str(struct cfs_hash *hs, struct seq_file *m)
2043 {
2044 int dist[8] = { 0, };
2045 int maxdep = -1;
2046 int maxdepb = -1;
2047 int total = 0;
2048 int theta;
2049 int i;
2050
2051 cfs_hash_lock(hs, 0);
2052 theta = __cfs_hash_theta(hs);
2053
2054 seq_printf(m, "%-*s %5d %5d %5d %d.%03d %d.%03d %d.%03d 0x%02x %6d ",
2055 CFS_HASH_BIGNAME_LEN, hs->hs_name,
2056 1 << hs->hs_cur_bits, 1 << hs->hs_min_bits,
2057 1 << hs->hs_max_bits,
2058 __cfs_hash_theta_int(theta), __cfs_hash_theta_frac(theta),
2059 __cfs_hash_theta_int(hs->hs_min_theta),
2060 __cfs_hash_theta_frac(hs->hs_min_theta),
2061 __cfs_hash_theta_int(hs->hs_max_theta),
2062 __cfs_hash_theta_frac(hs->hs_max_theta),
2063 hs->hs_flags, hs->hs_rehash_count);
2064
2065 /*
2066 * The distribution is a summary of the chained hash depth in
2067 * each of the libcfs hash buckets. Each buckets hsb_count is
2068 * divided by the hash theta value and used to generate a
2069 * histogram of the hash distribution. A uniform hash will
2070 * result in all hash buckets being close to the average thus
2071 * only the first few entries in the histogram will be non-zero.
2072 * If you hash function results in a non-uniform hash the will
2073 * be observable by outlier bucks in the distribution histogram.
2074 *
2075 * Uniform hash distribution: 128/128/0/0/0/0/0/0
2076 * Non-Uniform hash distribution: 128/125/0/0/0/0/2/1
2077 */
2078 for (i = 0; i < cfs_hash_full_nbkt(hs); i++) {
2079 struct cfs_hash_bd bd;
2080
2081 bd.bd_bucket = cfs_hash_full_bkts(hs)[i];
2082 cfs_hash_bd_lock(hs, &bd, 0);
2083 if (maxdep < bd.bd_bucket->hsb_depmax) {
2084 maxdep = bd.bd_bucket->hsb_depmax;
2085 maxdepb = ffz(~maxdep);
2086 }
2087 total += bd.bd_bucket->hsb_count;
2088 dist[min(fls(bd.bd_bucket->hsb_count / max(theta, 1)), 7)]++;
2089 cfs_hash_bd_unlock(hs, &bd, 0);
2090 }
2091
2092 seq_printf(m, "%7d %7d %7d ", total, maxdep, maxdepb);
2093 for (i = 0; i < 8; i++)
2094 seq_printf(m, "%d%c", dist[i], (i == 7) ? '\n' : '/');
2095
2096 cfs_hash_unlock(hs, 0);
2097 }
2098 EXPORT_SYMBOL(cfs_hash_debug_str);
2099