1 // SPDX-License-Identifier: GPL-2.0-only 1 // SPDX-License-Identifier: GPL-2.0-only
2 #include <linux/spinlock.h> 2 #include <linux/spinlock.h>
3 #include <linux/slab.h> 3 #include <linux/slab.h>
4 #include <linux/list.h> 4 #include <linux/list.h>
5 #include <linux/list_bl.h> 5 #include <linux/list_bl.h>
6 #include <linux/module.h> 6 #include <linux/module.h>
7 #include <linux/sched.h> 7 #include <linux/sched.h>
8 #include <linux/workqueue.h> 8 #include <linux/workqueue.h>
9 #include <linux/mbcache.h> 9 #include <linux/mbcache.h>
10 10
11 /* 11 /*
12 * Mbcache is a simple key-value store. Keys n 12 * Mbcache is a simple key-value store. Keys need not be unique, however
13 * key-value pairs are expected to be unique ( 13 * key-value pairs are expected to be unique (we use this fact in
14 * mb_cache_entry_delete_or_get()). 14 * mb_cache_entry_delete_or_get()).
15 * 15 *
16 * Ext2 and ext4 use this cache for deduplicat 16 * Ext2 and ext4 use this cache for deduplication of extended attribute blocks.
17 * Ext4 also uses it for deduplication of xatt 17 * Ext4 also uses it for deduplication of xattr values stored in inodes.
18 * They use hash of data as a key and provide 18 * They use hash of data as a key and provide a value that may represent a
19 * block or inode number. That's why keys need 19 * block or inode number. That's why keys need not be unique (hash of different
20 * data may be the same). However user provide 20 * data may be the same). However user provided value always uniquely
21 * identifies a cache entry. 21 * identifies a cache entry.
22 * 22 *
23 * We provide functions for creation and remov 23 * We provide functions for creation and removal of entries, search by key,
24 * and a special "delete entry with given key- 24 * and a special "delete entry with given key-value pair" operation. Fixed
25 * size hash table is used for fast key lookup 25 * size hash table is used for fast key lookups.
26 */ 26 */
27 27
28 struct mb_cache { 28 struct mb_cache {
29 /* Hash table of entries */ 29 /* Hash table of entries */
30 struct hlist_bl_head *c_hash; 30 struct hlist_bl_head *c_hash;
31 /* log2 of hash table size */ 31 /* log2 of hash table size */
32 int c_bucket_bits; 32 int c_bucket_bits;
33 /* Maximum entries in cache to avoid d 33 /* Maximum entries in cache to avoid degrading hash too much */
34 unsigned long c_max_entries; 34 unsigned long c_max_entries;
35 /* Protects c_list, c_entry_count */ 35 /* Protects c_list, c_entry_count */
36 spinlock_t c_list_lock; 36 spinlock_t c_list_lock;
37 struct list_head c_list; 37 struct list_head c_list;
38 /* Number of entries in cache */ 38 /* Number of entries in cache */
39 unsigned long c_entry_count; 39 unsigned long c_entry_count;
40 struct shrinker *c_shrink; !! 40 struct shrinker c_shrink;
41 /* Work for shrinking when the cache h 41 /* Work for shrinking when the cache has too many entries */
42 struct work_struct c_shrink_work; 42 struct work_struct c_shrink_work;
43 }; 43 };
44 44
45 static struct kmem_cache *mb_entry_cache; 45 static struct kmem_cache *mb_entry_cache;
46 46
47 static unsigned long mb_cache_shrink(struct mb 47 static unsigned long mb_cache_shrink(struct mb_cache *cache,
48 unsigned 48 unsigned long nr_to_scan);
49 49
50 static inline struct hlist_bl_head *mb_cache_e 50 static inline struct hlist_bl_head *mb_cache_entry_head(struct mb_cache *cache,
51 51 u32 key)
52 { 52 {
53 return &cache->c_hash[hash_32(key, cac 53 return &cache->c_hash[hash_32(key, cache->c_bucket_bits)];
54 } 54 }
55 55
56 /* 56 /*
57 * Number of entries to reclaim synchronously 57 * Number of entries to reclaim synchronously when there are too many entries
58 * in cache 58 * in cache
59 */ 59 */
60 #define SYNC_SHRINK_BATCH 64 60 #define SYNC_SHRINK_BATCH 64
61 61
62 /* 62 /*
63 * mb_cache_entry_create - create entry in cac 63 * mb_cache_entry_create - create entry in cache
64 * @cache - cache where the entry should be cr 64 * @cache - cache where the entry should be created
65 * @mask - gfp mask with which the entry shoul 65 * @mask - gfp mask with which the entry should be allocated
66 * @key - key of the entry 66 * @key - key of the entry
67 * @value - value of the entry 67 * @value - value of the entry
68 * @reusable - is the entry reusable by others 68 * @reusable - is the entry reusable by others?
69 * 69 *
70 * Creates entry in @cache with key @key and v 70 * Creates entry in @cache with key @key and value @value. The function returns
71 * -EBUSY if entry with the same key and value 71 * -EBUSY if entry with the same key and value already exists in cache.
72 * Otherwise 0 is returned. 72 * Otherwise 0 is returned.
73 */ 73 */
74 int mb_cache_entry_create(struct mb_cache *cac 74 int mb_cache_entry_create(struct mb_cache *cache, gfp_t mask, u32 key,
75 u64 value, bool reus 75 u64 value, bool reusable)
76 { 76 {
77 struct mb_cache_entry *entry, *dup; 77 struct mb_cache_entry *entry, *dup;
78 struct hlist_bl_node *dup_node; 78 struct hlist_bl_node *dup_node;
79 struct hlist_bl_head *head; 79 struct hlist_bl_head *head;
80 80
81 /* Schedule background reclaim if ther 81 /* Schedule background reclaim if there are too many entries */
82 if (cache->c_entry_count >= cache->c_m 82 if (cache->c_entry_count >= cache->c_max_entries)
83 schedule_work(&cache->c_shrink 83 schedule_work(&cache->c_shrink_work);
84 /* Do some sync reclaim if background 84 /* Do some sync reclaim if background reclaim cannot keep up */
85 if (cache->c_entry_count >= 2*cache->c 85 if (cache->c_entry_count >= 2*cache->c_max_entries)
86 mb_cache_shrink(cache, SYNC_SH 86 mb_cache_shrink(cache, SYNC_SHRINK_BATCH);
87 87
88 entry = kmem_cache_alloc(mb_entry_cach 88 entry = kmem_cache_alloc(mb_entry_cache, mask);
89 if (!entry) 89 if (!entry)
90 return -ENOMEM; 90 return -ENOMEM;
91 91
92 INIT_LIST_HEAD(&entry->e_list); 92 INIT_LIST_HEAD(&entry->e_list);
93 /* 93 /*
94 * We create entry with two references 94 * We create entry with two references. One reference is kept by the
95 * hash table, the other reference is 95 * hash table, the other reference is used to protect us from
96 * mb_cache_entry_delete_or_get() unti 96 * mb_cache_entry_delete_or_get() until the entry is fully setup. This
97 * avoids nesting of cache->c_list_loc 97 * avoids nesting of cache->c_list_lock into hash table bit locks which
98 * is problematic for RT. 98 * is problematic for RT.
99 */ 99 */
100 atomic_set(&entry->e_refcnt, 2); 100 atomic_set(&entry->e_refcnt, 2);
101 entry->e_key = key; 101 entry->e_key = key;
102 entry->e_value = value; 102 entry->e_value = value;
103 entry->e_flags = 0; 103 entry->e_flags = 0;
104 if (reusable) 104 if (reusable)
105 set_bit(MBE_REUSABLE_B, &entry 105 set_bit(MBE_REUSABLE_B, &entry->e_flags);
106 head = mb_cache_entry_head(cache, key) 106 head = mb_cache_entry_head(cache, key);
107 hlist_bl_lock(head); 107 hlist_bl_lock(head);
108 hlist_bl_for_each_entry(dup, dup_node, 108 hlist_bl_for_each_entry(dup, dup_node, head, e_hash_list) {
109 if (dup->e_key == key && dup-> 109 if (dup->e_key == key && dup->e_value == value) {
110 hlist_bl_unlock(head); 110 hlist_bl_unlock(head);
111 kmem_cache_free(mb_ent 111 kmem_cache_free(mb_entry_cache, entry);
112 return -EBUSY; 112 return -EBUSY;
113 } 113 }
114 } 114 }
115 hlist_bl_add_head(&entry->e_hash_list, 115 hlist_bl_add_head(&entry->e_hash_list, head);
116 hlist_bl_unlock(head); 116 hlist_bl_unlock(head);
117 spin_lock(&cache->c_list_lock); 117 spin_lock(&cache->c_list_lock);
118 list_add_tail(&entry->e_list, &cache-> 118 list_add_tail(&entry->e_list, &cache->c_list);
119 cache->c_entry_count++; 119 cache->c_entry_count++;
120 spin_unlock(&cache->c_list_lock); 120 spin_unlock(&cache->c_list_lock);
121 mb_cache_entry_put(cache, entry); 121 mb_cache_entry_put(cache, entry);
122 122
123 return 0; 123 return 0;
124 } 124 }
125 EXPORT_SYMBOL(mb_cache_entry_create); 125 EXPORT_SYMBOL(mb_cache_entry_create);
126 126
127 void __mb_cache_entry_free(struct mb_cache *ca 127 void __mb_cache_entry_free(struct mb_cache *cache, struct mb_cache_entry *entry)
128 { 128 {
129 struct hlist_bl_head *head; 129 struct hlist_bl_head *head;
130 130
131 head = mb_cache_entry_head(cache, entr 131 head = mb_cache_entry_head(cache, entry->e_key);
132 hlist_bl_lock(head); 132 hlist_bl_lock(head);
133 hlist_bl_del(&entry->e_hash_list); 133 hlist_bl_del(&entry->e_hash_list);
134 hlist_bl_unlock(head); 134 hlist_bl_unlock(head);
135 kmem_cache_free(mb_entry_cache, entry) 135 kmem_cache_free(mb_entry_cache, entry);
136 } 136 }
137 EXPORT_SYMBOL(__mb_cache_entry_free); 137 EXPORT_SYMBOL(__mb_cache_entry_free);
138 138
139 /* 139 /*
140 * mb_cache_entry_wait_unused - wait to be the 140 * mb_cache_entry_wait_unused - wait to be the last user of the entry
141 * 141 *
142 * @entry - entry to work on 142 * @entry - entry to work on
143 * 143 *
144 * Wait to be the last user of the entry. 144 * Wait to be the last user of the entry.
145 */ 145 */
146 void mb_cache_entry_wait_unused(struct mb_cach 146 void mb_cache_entry_wait_unused(struct mb_cache_entry *entry)
147 { 147 {
148 wait_var_event(&entry->e_refcnt, atomi 148 wait_var_event(&entry->e_refcnt, atomic_read(&entry->e_refcnt) <= 2);
149 } 149 }
150 EXPORT_SYMBOL(mb_cache_entry_wait_unused); 150 EXPORT_SYMBOL(mb_cache_entry_wait_unused);
151 151
152 static struct mb_cache_entry *__entry_find(str 152 static struct mb_cache_entry *__entry_find(struct mb_cache *cache,
153 str 153 struct mb_cache_entry *entry,
154 u32 154 u32 key)
155 { 155 {
156 struct mb_cache_entry *old_entry = ent 156 struct mb_cache_entry *old_entry = entry;
157 struct hlist_bl_node *node; 157 struct hlist_bl_node *node;
158 struct hlist_bl_head *head; 158 struct hlist_bl_head *head;
159 159
160 head = mb_cache_entry_head(cache, key) 160 head = mb_cache_entry_head(cache, key);
161 hlist_bl_lock(head); 161 hlist_bl_lock(head);
162 if (entry && !hlist_bl_unhashed(&entry 162 if (entry && !hlist_bl_unhashed(&entry->e_hash_list))
163 node = entry->e_hash_list.next 163 node = entry->e_hash_list.next;
164 else 164 else
165 node = hlist_bl_first(head); 165 node = hlist_bl_first(head);
166 while (node) { 166 while (node) {
167 entry = hlist_bl_entry(node, s 167 entry = hlist_bl_entry(node, struct mb_cache_entry,
168 e_hash_ 168 e_hash_list);
169 if (entry->e_key == key && 169 if (entry->e_key == key &&
170 test_bit(MBE_REUSABLE_B, & 170 test_bit(MBE_REUSABLE_B, &entry->e_flags) &&
171 atomic_inc_not_zero(&entry 171 atomic_inc_not_zero(&entry->e_refcnt))
172 goto out; 172 goto out;
173 node = node->next; 173 node = node->next;
174 } 174 }
175 entry = NULL; 175 entry = NULL;
176 out: 176 out:
177 hlist_bl_unlock(head); 177 hlist_bl_unlock(head);
178 if (old_entry) 178 if (old_entry)
179 mb_cache_entry_put(cache, old_ 179 mb_cache_entry_put(cache, old_entry);
180 180
181 return entry; 181 return entry;
182 } 182 }
183 183
184 /* 184 /*
185 * mb_cache_entry_find_first - find the first 185 * mb_cache_entry_find_first - find the first reusable entry with the given key
186 * @cache: cache where we should search 186 * @cache: cache where we should search
187 * @key: key to look for 187 * @key: key to look for
188 * 188 *
189 * Search in @cache for a reusable entry with 189 * Search in @cache for a reusable entry with key @key. Grabs reference to the
190 * first reusable entry found and returns the 190 * first reusable entry found and returns the entry.
191 */ 191 */
192 struct mb_cache_entry *mb_cache_entry_find_fir 192 struct mb_cache_entry *mb_cache_entry_find_first(struct mb_cache *cache,
193 193 u32 key)
194 { 194 {
195 return __entry_find(cache, NULL, key); 195 return __entry_find(cache, NULL, key);
196 } 196 }
197 EXPORT_SYMBOL(mb_cache_entry_find_first); 197 EXPORT_SYMBOL(mb_cache_entry_find_first);
198 198
199 /* 199 /*
200 * mb_cache_entry_find_next - find next reusab 200 * mb_cache_entry_find_next - find next reusable entry with the same key
201 * @cache: cache where we should search 201 * @cache: cache where we should search
202 * @entry: entry to start search from 202 * @entry: entry to start search from
203 * 203 *
204 * Finds next reusable entry in the hash chain 204 * Finds next reusable entry in the hash chain which has the same key as @entry.
205 * If @entry is unhashed (which can happen whe 205 * If @entry is unhashed (which can happen when deletion of entry races with the
206 * search), finds the first reusable entry in 206 * search), finds the first reusable entry in the hash chain. The function drops
207 * reference to @entry and returns with a refe 207 * reference to @entry and returns with a reference to the found entry.
208 */ 208 */
209 struct mb_cache_entry *mb_cache_entry_find_nex 209 struct mb_cache_entry *mb_cache_entry_find_next(struct mb_cache *cache,
210 210 struct mb_cache_entry *entry)
211 { 211 {
212 return __entry_find(cache, entry, entr 212 return __entry_find(cache, entry, entry->e_key);
213 } 213 }
214 EXPORT_SYMBOL(mb_cache_entry_find_next); 214 EXPORT_SYMBOL(mb_cache_entry_find_next);
215 215
216 /* 216 /*
217 * mb_cache_entry_get - get a cache entry by v 217 * mb_cache_entry_get - get a cache entry by value (and key)
218 * @cache - cache we work with 218 * @cache - cache we work with
219 * @key - key 219 * @key - key
220 * @value - value 220 * @value - value
221 */ 221 */
222 struct mb_cache_entry *mb_cache_entry_get(stru 222 struct mb_cache_entry *mb_cache_entry_get(struct mb_cache *cache, u32 key,
223 u64 223 u64 value)
224 { 224 {
225 struct hlist_bl_node *node; 225 struct hlist_bl_node *node;
226 struct hlist_bl_head *head; 226 struct hlist_bl_head *head;
227 struct mb_cache_entry *entry; 227 struct mb_cache_entry *entry;
228 228
229 head = mb_cache_entry_head(cache, key) 229 head = mb_cache_entry_head(cache, key);
230 hlist_bl_lock(head); 230 hlist_bl_lock(head);
231 hlist_bl_for_each_entry(entry, node, h 231 hlist_bl_for_each_entry(entry, node, head, e_hash_list) {
232 if (entry->e_key == key && ent 232 if (entry->e_key == key && entry->e_value == value &&
233 atomic_inc_not_zero(&entry 233 atomic_inc_not_zero(&entry->e_refcnt))
234 goto out; 234 goto out;
235 } 235 }
236 entry = NULL; 236 entry = NULL;
237 out: 237 out:
238 hlist_bl_unlock(head); 238 hlist_bl_unlock(head);
239 return entry; 239 return entry;
240 } 240 }
241 EXPORT_SYMBOL(mb_cache_entry_get); 241 EXPORT_SYMBOL(mb_cache_entry_get);
242 242
243 /* mb_cache_entry_delete_or_get - remove a cac 243 /* mb_cache_entry_delete_or_get - remove a cache entry if it has no users
244 * @cache - cache we work with 244 * @cache - cache we work with
245 * @key - key 245 * @key - key
246 * @value - value 246 * @value - value
247 * 247 *
248 * Remove entry from cache @cache with key @ke 248 * Remove entry from cache @cache with key @key and value @value. The removal
249 * happens only if the entry is unused. The fu 249 * happens only if the entry is unused. The function returns NULL in case the
250 * entry was successfully removed or there's n 250 * entry was successfully removed or there's no entry in cache. Otherwise the
251 * function grabs reference of the entry that 251 * function grabs reference of the entry that we failed to delete because it
252 * still has users and return it. 252 * still has users and return it.
253 */ 253 */
254 struct mb_cache_entry *mb_cache_entry_delete_o 254 struct mb_cache_entry *mb_cache_entry_delete_or_get(struct mb_cache *cache,
255 255 u32 key, u64 value)
256 { 256 {
257 struct mb_cache_entry *entry; 257 struct mb_cache_entry *entry;
258 258
259 entry = mb_cache_entry_get(cache, key, 259 entry = mb_cache_entry_get(cache, key, value);
260 if (!entry) 260 if (!entry)
261 return NULL; 261 return NULL;
262 262
263 /* 263 /*
264 * Drop the ref we got from mb_cache_e 264 * Drop the ref we got from mb_cache_entry_get() and the initial hash
265 * ref if we are the last user 265 * ref if we are the last user
266 */ 266 */
267 if (atomic_cmpxchg(&entry->e_refcnt, 2 267 if (atomic_cmpxchg(&entry->e_refcnt, 2, 0) != 2)
268 return entry; 268 return entry;
269 269
270 spin_lock(&cache->c_list_lock); 270 spin_lock(&cache->c_list_lock);
271 if (!list_empty(&entry->e_list)) 271 if (!list_empty(&entry->e_list))
272 list_del_init(&entry->e_list); 272 list_del_init(&entry->e_list);
273 cache->c_entry_count--; 273 cache->c_entry_count--;
274 spin_unlock(&cache->c_list_lock); 274 spin_unlock(&cache->c_list_lock);
275 __mb_cache_entry_free(cache, entry); 275 __mb_cache_entry_free(cache, entry);
276 return NULL; 276 return NULL;
277 } 277 }
278 EXPORT_SYMBOL(mb_cache_entry_delete_or_get); 278 EXPORT_SYMBOL(mb_cache_entry_delete_or_get);
279 279
280 /* mb_cache_entry_touch - cache entry got used 280 /* mb_cache_entry_touch - cache entry got used
281 * @cache - cache the entry belongs to 281 * @cache - cache the entry belongs to
282 * @entry - entry that got used 282 * @entry - entry that got used
283 * 283 *
284 * Marks entry as used to give hit higher chan 284 * Marks entry as used to give hit higher chances of surviving in cache.
285 */ 285 */
286 void mb_cache_entry_touch(struct mb_cache *cac 286 void mb_cache_entry_touch(struct mb_cache *cache,
287 struct mb_cache_entr 287 struct mb_cache_entry *entry)
288 { 288 {
289 set_bit(MBE_REFERENCED_B, &entry->e_fl 289 set_bit(MBE_REFERENCED_B, &entry->e_flags);
290 } 290 }
291 EXPORT_SYMBOL(mb_cache_entry_touch); 291 EXPORT_SYMBOL(mb_cache_entry_touch);
292 292
293 static unsigned long mb_cache_count(struct shr 293 static unsigned long mb_cache_count(struct shrinker *shrink,
294 struct shr 294 struct shrink_control *sc)
295 { 295 {
296 struct mb_cache *cache = shrink->priva !! 296 struct mb_cache *cache = container_of(shrink, struct mb_cache,
>> 297 c_shrink);
297 298
298 return cache->c_entry_count; 299 return cache->c_entry_count;
299 } 300 }
300 301
301 /* Shrink number of entries in cache */ 302 /* Shrink number of entries in cache */
302 static unsigned long mb_cache_shrink(struct mb 303 static unsigned long mb_cache_shrink(struct mb_cache *cache,
303 unsigned 304 unsigned long nr_to_scan)
304 { 305 {
305 struct mb_cache_entry *entry; 306 struct mb_cache_entry *entry;
306 unsigned long shrunk = 0; 307 unsigned long shrunk = 0;
307 308
308 spin_lock(&cache->c_list_lock); 309 spin_lock(&cache->c_list_lock);
309 while (nr_to_scan-- && !list_empty(&ca 310 while (nr_to_scan-- && !list_empty(&cache->c_list)) {
310 entry = list_first_entry(&cach 311 entry = list_first_entry(&cache->c_list,
311 struc 312 struct mb_cache_entry, e_list);
312 /* Drop initial hash reference 313 /* Drop initial hash reference if there is no user */
313 if (test_bit(MBE_REFERENCED_B, 314 if (test_bit(MBE_REFERENCED_B, &entry->e_flags) ||
314 atomic_cmpxchg(&entry->e_r 315 atomic_cmpxchg(&entry->e_refcnt, 1, 0) != 1) {
315 clear_bit(MBE_REFERENC 316 clear_bit(MBE_REFERENCED_B, &entry->e_flags);
316 list_move_tail(&entry- 317 list_move_tail(&entry->e_list, &cache->c_list);
317 continue; 318 continue;
318 } 319 }
319 list_del_init(&entry->e_list); 320 list_del_init(&entry->e_list);
320 cache->c_entry_count--; 321 cache->c_entry_count--;
321 spin_unlock(&cache->c_list_loc 322 spin_unlock(&cache->c_list_lock);
322 __mb_cache_entry_free(cache, e 323 __mb_cache_entry_free(cache, entry);
323 shrunk++; 324 shrunk++;
324 cond_resched(); 325 cond_resched();
325 spin_lock(&cache->c_list_lock) 326 spin_lock(&cache->c_list_lock);
326 } 327 }
327 spin_unlock(&cache->c_list_lock); 328 spin_unlock(&cache->c_list_lock);
328 329
329 return shrunk; 330 return shrunk;
330 } 331 }
331 332
332 static unsigned long mb_cache_scan(struct shri 333 static unsigned long mb_cache_scan(struct shrinker *shrink,
333 struct shri 334 struct shrink_control *sc)
334 { 335 {
335 struct mb_cache *cache = shrink->priva !! 336 struct mb_cache *cache = container_of(shrink, struct mb_cache,
>> 337 c_shrink);
336 return mb_cache_shrink(cache, sc->nr_t 338 return mb_cache_shrink(cache, sc->nr_to_scan);
337 } 339 }
338 340
339 /* We shrink 1/X of the cache when we have too 341 /* We shrink 1/X of the cache when we have too many entries in it */
340 #define SHRINK_DIVISOR 16 342 #define SHRINK_DIVISOR 16
341 343
342 static void mb_cache_shrink_worker(struct work 344 static void mb_cache_shrink_worker(struct work_struct *work)
343 { 345 {
344 struct mb_cache *cache = container_of( 346 struct mb_cache *cache = container_of(work, struct mb_cache,
345 347 c_shrink_work);
346 mb_cache_shrink(cache, cache->c_max_en 348 mb_cache_shrink(cache, cache->c_max_entries / SHRINK_DIVISOR);
347 } 349 }
348 350
349 /* 351 /*
350 * mb_cache_create - create cache 352 * mb_cache_create - create cache
351 * @bucket_bits: log2 of the hash table size 353 * @bucket_bits: log2 of the hash table size
352 * 354 *
353 * Create cache for keys with 2^bucket_bits ha 355 * Create cache for keys with 2^bucket_bits hash entries.
354 */ 356 */
355 struct mb_cache *mb_cache_create(int bucket_bi 357 struct mb_cache *mb_cache_create(int bucket_bits)
356 { 358 {
357 struct mb_cache *cache; 359 struct mb_cache *cache;
358 unsigned long bucket_count = 1UL << bu 360 unsigned long bucket_count = 1UL << bucket_bits;
359 unsigned long i; 361 unsigned long i;
360 362
361 cache = kzalloc(sizeof(struct mb_cache 363 cache = kzalloc(sizeof(struct mb_cache), GFP_KERNEL);
362 if (!cache) 364 if (!cache)
363 goto err_out; 365 goto err_out;
364 cache->c_bucket_bits = bucket_bits; 366 cache->c_bucket_bits = bucket_bits;
365 cache->c_max_entries = bucket_count << 367 cache->c_max_entries = bucket_count << 4;
366 INIT_LIST_HEAD(&cache->c_list); 368 INIT_LIST_HEAD(&cache->c_list);
367 spin_lock_init(&cache->c_list_lock); 369 spin_lock_init(&cache->c_list_lock);
368 cache->c_hash = kmalloc_array(bucket_c 370 cache->c_hash = kmalloc_array(bucket_count,
369 sizeof(s 371 sizeof(struct hlist_bl_head),
370 GFP_KERN 372 GFP_KERNEL);
371 if (!cache->c_hash) { 373 if (!cache->c_hash) {
372 kfree(cache); 374 kfree(cache);
373 goto err_out; 375 goto err_out;
374 } 376 }
375 for (i = 0; i < bucket_count; i++) 377 for (i = 0; i < bucket_count; i++)
376 INIT_HLIST_BL_HEAD(&cache->c_h 378 INIT_HLIST_BL_HEAD(&cache->c_hash[i]);
377 379
378 cache->c_shrink = shrinker_alloc(0, "m !! 380 cache->c_shrink.count_objects = mb_cache_count;
379 if (!cache->c_shrink) { !! 381 cache->c_shrink.scan_objects = mb_cache_scan;
>> 382 cache->c_shrink.seeks = DEFAULT_SEEKS;
>> 383 if (register_shrinker(&cache->c_shrink, "mbcache-shrinker")) {
380 kfree(cache->c_hash); 384 kfree(cache->c_hash);
381 kfree(cache); 385 kfree(cache);
382 goto err_out; 386 goto err_out;
383 } 387 }
384 388
385 cache->c_shrink->count_objects = mb_ca <<
386 cache->c_shrink->scan_objects = mb_cac <<
387 cache->c_shrink->private_data = cache; <<
388 <<
389 shrinker_register(cache->c_shrink); <<
390 <<
391 INIT_WORK(&cache->c_shrink_work, mb_ca 389 INIT_WORK(&cache->c_shrink_work, mb_cache_shrink_worker);
392 390
393 return cache; 391 return cache;
394 392
395 err_out: 393 err_out:
396 return NULL; 394 return NULL;
397 } 395 }
398 EXPORT_SYMBOL(mb_cache_create); 396 EXPORT_SYMBOL(mb_cache_create);
399 397
400 /* 398 /*
401 * mb_cache_destroy - destroy cache 399 * mb_cache_destroy - destroy cache
402 * @cache: the cache to destroy 400 * @cache: the cache to destroy
403 * 401 *
404 * Free all entries in cache and cache itself. 402 * Free all entries in cache and cache itself. Caller must make sure nobody
405 * (except shrinker) can reach @cache when cal 403 * (except shrinker) can reach @cache when calling this.
406 */ 404 */
407 void mb_cache_destroy(struct mb_cache *cache) 405 void mb_cache_destroy(struct mb_cache *cache)
408 { 406 {
409 struct mb_cache_entry *entry, *next; 407 struct mb_cache_entry *entry, *next;
410 408
411 shrinker_free(cache->c_shrink); !! 409 unregister_shrinker(&cache->c_shrink);
412 410
413 /* 411 /*
414 * We don't bother with any locking. C 412 * We don't bother with any locking. Cache must not be used at this
415 * point. 413 * point.
416 */ 414 */
417 list_for_each_entry_safe(entry, next, 415 list_for_each_entry_safe(entry, next, &cache->c_list, e_list) {
418 list_del(&entry->e_list); 416 list_del(&entry->e_list);
419 WARN_ON(atomic_read(&entry->e_ 417 WARN_ON(atomic_read(&entry->e_refcnt) != 1);
420 mb_cache_entry_put(cache, entr 418 mb_cache_entry_put(cache, entry);
421 } 419 }
422 kfree(cache->c_hash); 420 kfree(cache->c_hash);
423 kfree(cache); 421 kfree(cache);
424 } 422 }
425 EXPORT_SYMBOL(mb_cache_destroy); 423 EXPORT_SYMBOL(mb_cache_destroy);
426 424
427 static int __init mbcache_init(void) 425 static int __init mbcache_init(void)
428 { 426 {
429 mb_entry_cache = KMEM_CACHE(mb_cache_e !! 427 mb_entry_cache = kmem_cache_create("mbcache",
>> 428 sizeof(struct mb_cache_entry), 0,
>> 429 SLAB_RECLAIM_ACCOUNT|SLAB_MEM_SPREAD, NULL);
430 if (!mb_entry_cache) 430 if (!mb_entry_cache)
431 return -ENOMEM; 431 return -ENOMEM;
432 return 0; 432 return 0;
433 } 433 }
434 434
435 static void __exit mbcache_exit(void) 435 static void __exit mbcache_exit(void)
436 { 436 {
437 kmem_cache_destroy(mb_entry_cache); 437 kmem_cache_destroy(mb_entry_cache);
438 } 438 }
439 439
440 module_init(mbcache_init) 440 module_init(mbcache_init)
441 module_exit(mbcache_exit) 441 module_exit(mbcache_exit)
442 442
443 MODULE_AUTHOR("Jan Kara <jack@suse.cz>"); 443 MODULE_AUTHOR("Jan Kara <jack@suse.cz>");
444 MODULE_DESCRIPTION("Meta block cache (for exte 444 MODULE_DESCRIPTION("Meta block cache (for extended attributes)");
445 MODULE_LICENSE("GPL"); 445 MODULE_LICENSE("GPL");
446 446
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.