Commit | Line | Data |
---|---|---|
1da177e4 LT |
1 | /* |
2 | * fs/dcache.c | |
3 | * | |
4 | * Complete reimplementation | |
5 | * (C) 1997 Thomas Schoebel-Theuer, | |
6 | * with heavy changes by Linus Torvalds | |
7 | */ | |
8 | ||
9 | /* | |
10 | * Notes on the allocation strategy: | |
11 | * | |
12 | * The dcache is a master of the icache - whenever a dcache entry | |
13 | * exists, the inode will always exist. "iput()" is done either when | |
14 | * the dcache entry is deleted or garbage collected. | |
15 | */ | |
16 | ||
1da177e4 LT |
17 | #include <linux/syscalls.h> |
18 | #include <linux/string.h> | |
19 | #include <linux/mm.h> | |
20 | #include <linux/fs.h> | |
7a91bf7f | 21 | #include <linux/fsnotify.h> |
1da177e4 LT |
22 | #include <linux/slab.h> |
23 | #include <linux/init.h> | |
1da177e4 LT |
24 | #include <linux/hash.h> |
25 | #include <linux/cache.h> | |
26 | #include <linux/module.h> | |
27 | #include <linux/mount.h> | |
28 | #include <linux/file.h> | |
29 | #include <asm/uaccess.h> | |
30 | #include <linux/security.h> | |
31 | #include <linux/seqlock.h> | |
32 | #include <linux/swap.h> | |
33 | #include <linux/bootmem.h> | |
5ad4e53b | 34 | #include <linux/fs_struct.h> |
613afbf8 | 35 | #include <linux/hardirq.h> |
ceb5bdc2 NP |
36 | #include <linux/bit_spinlock.h> |
37 | #include <linux/rculist_bl.h> | |
07f3f05c | 38 | #include "internal.h" |
1da177e4 | 39 | |
789680d1 NP |
40 | /* |
41 | * Usage: | |
873feea0 NP |
42 | * dcache->d_inode->i_lock protects: |
43 | * - i_dentry, d_alias, d_inode of aliases | |
ceb5bdc2 NP |
44 | * dcache_hash_bucket lock protects: |
45 | * - the dcache hash table | |
46 | * s_anon bl list spinlock protects: | |
47 | * - the s_anon list (see __d_drop) | |
23044507 NP |
48 | * dcache_lru_lock protects: |
49 | * - the dcache lru lists and counters | |
50 | * d_lock protects: | |
51 | * - d_flags | |
52 | * - d_name | |
53 | * - d_lru | |
b7ab39f6 | 54 | * - d_count |
da502956 | 55 | * - d_unhashed() |
2fd6b7f5 NP |
56 | * - d_parent and d_subdirs |
57 | * - childrens' d_child and d_parent | |
b23fb0a6 | 58 | * - d_alias, d_inode |
789680d1 NP |
59 | * |
60 | * Ordering: | |
873feea0 | 61 | * dentry->d_inode->i_lock |
b5c84bf6 NP |
62 | * dentry->d_lock |
63 | * dcache_lru_lock | |
ceb5bdc2 NP |
64 | * dcache_hash_bucket lock |
65 | * s_anon lock | |
789680d1 | 66 | * |
da502956 NP |
67 | * If there is an ancestor relationship: |
68 | * dentry->d_parent->...->d_parent->d_lock | |
69 | * ... | |
70 | * dentry->d_parent->d_lock | |
71 | * dentry->d_lock | |
72 | * | |
73 | * If no ancestor relationship: | |
789680d1 NP |
74 | * if (dentry1 < dentry2) |
75 | * dentry1->d_lock | |
76 | * dentry2->d_lock | |
77 | */ | |
fa3536cc | 78 | int sysctl_vfs_cache_pressure __read_mostly = 100; |
1da177e4 LT |
79 | EXPORT_SYMBOL_GPL(sysctl_vfs_cache_pressure); |
80 | ||
23044507 | 81 | static __cacheline_aligned_in_smp DEFINE_SPINLOCK(dcache_lru_lock); |
74c3cbe3 | 82 | __cacheline_aligned_in_smp DEFINE_SEQLOCK(rename_lock); |
1da177e4 | 83 | |
949854d0 | 84 | EXPORT_SYMBOL(rename_lock); |
1da177e4 | 85 | |
e18b890b | 86 | static struct kmem_cache *dentry_cache __read_mostly; |
1da177e4 | 87 | |
1da177e4 LT |
88 | /* |
89 | * This is the single most critical data structure when it comes | |
90 | * to the dcache: the hashtable for lookups. Somebody should try | |
91 | * to make this good - I've just made it work. | |
92 | * | |
93 | * This hash-function tries to avoid losing too many bits of hash | |
94 | * information, yet avoid using a prime hash-size or similar. | |
95 | */ | |
96 | #define D_HASHBITS d_hash_shift | |
97 | #define D_HASHMASK d_hash_mask | |
98 | ||
fa3536cc ED |
99 | static unsigned int d_hash_mask __read_mostly; |
100 | static unsigned int d_hash_shift __read_mostly; | |
ceb5bdc2 NP |
101 | |
102 | struct dcache_hash_bucket { | |
103 | struct hlist_bl_head head; | |
104 | }; | |
105 | static struct dcache_hash_bucket *dentry_hashtable __read_mostly; | |
106 | ||
107 | static inline struct dcache_hash_bucket *d_hash(struct dentry *parent, | |
108 | unsigned long hash) | |
109 | { | |
110 | hash += ((unsigned long) parent ^ GOLDEN_RATIO_PRIME) / L1_CACHE_BYTES; | |
111 | hash = hash ^ ((hash ^ GOLDEN_RATIO_PRIME) >> D_HASHBITS); | |
112 | return dentry_hashtable + (hash & D_HASHMASK); | |
113 | } | |
114 | ||
115 | static inline void spin_lock_bucket(struct dcache_hash_bucket *b) | |
116 | { | |
117 | bit_spin_lock(0, (unsigned long *)&b->head.first); | |
118 | } | |
119 | ||
120 | static inline void spin_unlock_bucket(struct dcache_hash_bucket *b) | |
121 | { | |
122 | __bit_spin_unlock(0, (unsigned long *)&b->head.first); | |
123 | } | |
1da177e4 LT |
124 | |
125 | /* Statistics gathering. */ | |
126 | struct dentry_stat_t dentry_stat = { | |
127 | .age_limit = 45, | |
128 | }; | |
129 | ||
3e880fb5 | 130 | static DEFINE_PER_CPU(unsigned int, nr_dentry); |
312d3ca8 CH |
131 | |
132 | #if defined(CONFIG_SYSCTL) && defined(CONFIG_PROC_FS) | |
3e880fb5 NP |
133 | static int get_nr_dentry(void) |
134 | { | |
135 | int i; | |
136 | int sum = 0; | |
137 | for_each_possible_cpu(i) | |
138 | sum += per_cpu(nr_dentry, i); | |
139 | return sum < 0 ? 0 : sum; | |
140 | } | |
141 | ||
312d3ca8 CH |
142 | int proc_nr_dentry(ctl_table *table, int write, void __user *buffer, |
143 | size_t *lenp, loff_t *ppos) | |
144 | { | |
3e880fb5 | 145 | dentry_stat.nr_dentry = get_nr_dentry(); |
312d3ca8 CH |
146 | return proc_dointvec(table, write, buffer, lenp, ppos); |
147 | } | |
148 | #endif | |
149 | ||
9c82ab9c | 150 | static void __d_free(struct rcu_head *head) |
1da177e4 | 151 | { |
9c82ab9c CH |
152 | struct dentry *dentry = container_of(head, struct dentry, d_u.d_rcu); |
153 | ||
fd217f4d | 154 | WARN_ON(!list_empty(&dentry->d_alias)); |
1da177e4 LT |
155 | if (dname_external(dentry)) |
156 | kfree(dentry->d_name.name); | |
157 | kmem_cache_free(dentry_cache, dentry); | |
158 | } | |
159 | ||
160 | /* | |
b5c84bf6 | 161 | * no locks, please. |
1da177e4 LT |
162 | */ |
163 | static void d_free(struct dentry *dentry) | |
164 | { | |
b7ab39f6 | 165 | BUG_ON(dentry->d_count); |
3e880fb5 | 166 | this_cpu_dec(nr_dentry); |
1da177e4 LT |
167 | if (dentry->d_op && dentry->d_op->d_release) |
168 | dentry->d_op->d_release(dentry); | |
312d3ca8 | 169 | |
b3423415 | 170 | /* if dentry was never inserted into hash, immediate free is OK */ |
ceb5bdc2 | 171 | if (hlist_bl_unhashed(&dentry->d_hash)) |
9c82ab9c | 172 | __d_free(&dentry->d_u.d_rcu); |
b3423415 | 173 | else |
9c82ab9c | 174 | call_rcu(&dentry->d_u.d_rcu, __d_free); |
1da177e4 LT |
175 | } |
176 | ||
31e6b01f NP |
177 | /** |
178 | * dentry_rcuwalk_barrier - invalidate in-progress rcu-walk lookups | |
179 | * After this call, in-progress rcu-walk path lookup will fail. This | |
180 | * should be called after unhashing, and after changing d_inode (if | |
181 | * the dentry has not already been unhashed). | |
182 | */ | |
183 | static inline void dentry_rcuwalk_barrier(struct dentry *dentry) | |
184 | { | |
185 | assert_spin_locked(&dentry->d_lock); | |
186 | /* Go through a barrier */ | |
187 | write_seqcount_barrier(&dentry->d_seq); | |
188 | } | |
189 | ||
1da177e4 LT |
190 | /* |
191 | * Release the dentry's inode, using the filesystem | |
31e6b01f NP |
192 | * d_iput() operation if defined. Dentry has no refcount |
193 | * and is unhashed. | |
1da177e4 | 194 | */ |
858119e1 | 195 | static void dentry_iput(struct dentry * dentry) |
31f3e0b3 | 196 | __releases(dentry->d_lock) |
873feea0 | 197 | __releases(dentry->d_inode->i_lock) |
1da177e4 LT |
198 | { |
199 | struct inode *inode = dentry->d_inode; | |
200 | if (inode) { | |
201 | dentry->d_inode = NULL; | |
202 | list_del_init(&dentry->d_alias); | |
203 | spin_unlock(&dentry->d_lock); | |
873feea0 | 204 | spin_unlock(&inode->i_lock); |
f805fbda LT |
205 | if (!inode->i_nlink) |
206 | fsnotify_inoderemove(inode); | |
1da177e4 LT |
207 | if (dentry->d_op && dentry->d_op->d_iput) |
208 | dentry->d_op->d_iput(dentry, inode); | |
209 | else | |
210 | iput(inode); | |
211 | } else { | |
212 | spin_unlock(&dentry->d_lock); | |
1da177e4 LT |
213 | } |
214 | } | |
215 | ||
31e6b01f NP |
216 | /* |
217 | * Release the dentry's inode, using the filesystem | |
218 | * d_iput() operation if defined. dentry remains in-use. | |
219 | */ | |
220 | static void dentry_unlink_inode(struct dentry * dentry) | |
221 | __releases(dentry->d_lock) | |
873feea0 | 222 | __releases(dentry->d_inode->i_lock) |
31e6b01f NP |
223 | { |
224 | struct inode *inode = dentry->d_inode; | |
225 | dentry->d_inode = NULL; | |
226 | list_del_init(&dentry->d_alias); | |
227 | dentry_rcuwalk_barrier(dentry); | |
228 | spin_unlock(&dentry->d_lock); | |
873feea0 | 229 | spin_unlock(&inode->i_lock); |
31e6b01f NP |
230 | if (!inode->i_nlink) |
231 | fsnotify_inoderemove(inode); | |
232 | if (dentry->d_op && dentry->d_op->d_iput) | |
233 | dentry->d_op->d_iput(dentry, inode); | |
234 | else | |
235 | iput(inode); | |
236 | } | |
237 | ||
da3bbdd4 | 238 | /* |
23044507 | 239 | * dentry_lru_(add|del|move_tail) must be called with d_lock held. |
da3bbdd4 KM |
240 | */ |
241 | static void dentry_lru_add(struct dentry *dentry) | |
242 | { | |
a4633357 | 243 | if (list_empty(&dentry->d_lru)) { |
23044507 | 244 | spin_lock(&dcache_lru_lock); |
a4633357 CH |
245 | list_add(&dentry->d_lru, &dentry->d_sb->s_dentry_lru); |
246 | dentry->d_sb->s_nr_dentry_unused++; | |
86c8749e | 247 | dentry_stat.nr_unused++; |
23044507 | 248 | spin_unlock(&dcache_lru_lock); |
a4633357 | 249 | } |
da3bbdd4 KM |
250 | } |
251 | ||
23044507 NP |
252 | static void __dentry_lru_del(struct dentry *dentry) |
253 | { | |
254 | list_del_init(&dentry->d_lru); | |
255 | dentry->d_sb->s_nr_dentry_unused--; | |
256 | dentry_stat.nr_unused--; | |
257 | } | |
258 | ||
da3bbdd4 KM |
259 | static void dentry_lru_del(struct dentry *dentry) |
260 | { | |
261 | if (!list_empty(&dentry->d_lru)) { | |
23044507 NP |
262 | spin_lock(&dcache_lru_lock); |
263 | __dentry_lru_del(dentry); | |
264 | spin_unlock(&dcache_lru_lock); | |
da3bbdd4 KM |
265 | } |
266 | } | |
267 | ||
a4633357 | 268 | static void dentry_lru_move_tail(struct dentry *dentry) |
da3bbdd4 | 269 | { |
23044507 | 270 | spin_lock(&dcache_lru_lock); |
a4633357 CH |
271 | if (list_empty(&dentry->d_lru)) { |
272 | list_add_tail(&dentry->d_lru, &dentry->d_sb->s_dentry_lru); | |
273 | dentry->d_sb->s_nr_dentry_unused++; | |
86c8749e | 274 | dentry_stat.nr_unused++; |
a4633357 CH |
275 | } else { |
276 | list_move_tail(&dentry->d_lru, &dentry->d_sb->s_dentry_lru); | |
da3bbdd4 | 277 | } |
23044507 | 278 | spin_unlock(&dcache_lru_lock); |
da3bbdd4 KM |
279 | } |
280 | ||
d52b9086 MS |
281 | /** |
282 | * d_kill - kill dentry and return parent | |
283 | * @dentry: dentry to kill | |
284 | * | |
31f3e0b3 | 285 | * The dentry must already be unhashed and removed from the LRU. |
d52b9086 MS |
286 | * |
287 | * If this is the root of the dentry tree, return NULL. | |
23044507 | 288 | * |
b5c84bf6 NP |
289 | * dentry->d_lock and parent->d_lock must be held by caller, and are dropped by |
290 | * d_kill. | |
d52b9086 | 291 | */ |
2fd6b7f5 | 292 | static struct dentry *d_kill(struct dentry *dentry, struct dentry *parent) |
31f3e0b3 | 293 | __releases(dentry->d_lock) |
2fd6b7f5 | 294 | __releases(parent->d_lock) |
873feea0 | 295 | __releases(dentry->d_inode->i_lock) |
d52b9086 | 296 | { |
949854d0 | 297 | dentry->d_parent = NULL; |
d52b9086 | 298 | list_del(&dentry->d_u.d_child); |
2fd6b7f5 NP |
299 | if (parent) |
300 | spin_unlock(&parent->d_lock); | |
d52b9086 | 301 | dentry_iput(dentry); |
b7ab39f6 NP |
302 | /* |
303 | * dentry_iput drops the locks, at which point nobody (except | |
304 | * transient RCU lookups) can reach this dentry. | |
305 | */ | |
d52b9086 | 306 | d_free(dentry); |
871c0067 | 307 | return parent; |
d52b9086 MS |
308 | } |
309 | ||
789680d1 NP |
310 | /** |
311 | * d_drop - drop a dentry | |
312 | * @dentry: dentry to drop | |
313 | * | |
314 | * d_drop() unhashes the entry from the parent dentry hashes, so that it won't | |
315 | * be found through a VFS lookup any more. Note that this is different from | |
316 | * deleting the dentry - d_delete will try to mark the dentry negative if | |
317 | * possible, giving a successful _negative_ lookup, while d_drop will | |
318 | * just make the cache lookup fail. | |
319 | * | |
320 | * d_drop() is used mainly for stuff that wants to invalidate a dentry for some | |
321 | * reason (NFS timeouts or autofs deletes). | |
322 | * | |
323 | * __d_drop requires dentry->d_lock. | |
324 | */ | |
325 | void __d_drop(struct dentry *dentry) | |
326 | { | |
327 | if (!(dentry->d_flags & DCACHE_UNHASHED)) { | |
ceb5bdc2 NP |
328 | if (unlikely(dentry->d_flags & DCACHE_DISCONNECTED)) { |
329 | bit_spin_lock(0, | |
330 | (unsigned long *)&dentry->d_sb->s_anon.first); | |
331 | dentry->d_flags |= DCACHE_UNHASHED; | |
332 | hlist_bl_del_init(&dentry->d_hash); | |
333 | __bit_spin_unlock(0, | |
334 | (unsigned long *)&dentry->d_sb->s_anon.first); | |
335 | } else { | |
336 | struct dcache_hash_bucket *b; | |
337 | b = d_hash(dentry->d_parent, dentry->d_name.hash); | |
338 | spin_lock_bucket(b); | |
339 | /* | |
340 | * We may not actually need to put DCACHE_UNHASHED | |
341 | * manipulations under the hash lock, but follow | |
342 | * the principle of least surprise. | |
343 | */ | |
344 | dentry->d_flags |= DCACHE_UNHASHED; | |
345 | hlist_bl_del_rcu(&dentry->d_hash); | |
346 | spin_unlock_bucket(b); | |
347 | dentry_rcuwalk_barrier(dentry); | |
348 | } | |
789680d1 NP |
349 | } |
350 | } | |
351 | EXPORT_SYMBOL(__d_drop); | |
352 | ||
353 | void d_drop(struct dentry *dentry) | |
354 | { | |
789680d1 NP |
355 | spin_lock(&dentry->d_lock); |
356 | __d_drop(dentry); | |
357 | spin_unlock(&dentry->d_lock); | |
789680d1 NP |
358 | } |
359 | EXPORT_SYMBOL(d_drop); | |
360 | ||
77812a1e NP |
361 | /* |
362 | * Finish off a dentry we've decided to kill. | |
363 | * dentry->d_lock must be held, returns with it unlocked. | |
364 | * If ref is non-zero, then decrement the refcount too. | |
365 | * Returns dentry requiring refcount drop, or NULL if we're done. | |
366 | */ | |
367 | static inline struct dentry *dentry_kill(struct dentry *dentry, int ref) | |
368 | __releases(dentry->d_lock) | |
369 | { | |
873feea0 | 370 | struct inode *inode; |
77812a1e NP |
371 | struct dentry *parent; |
372 | ||
873feea0 NP |
373 | inode = dentry->d_inode; |
374 | if (inode && !spin_trylock(&inode->i_lock)) { | |
77812a1e NP |
375 | relock: |
376 | spin_unlock(&dentry->d_lock); | |
377 | cpu_relax(); | |
378 | return dentry; /* try again with same dentry */ | |
379 | } | |
380 | if (IS_ROOT(dentry)) | |
381 | parent = NULL; | |
382 | else | |
383 | parent = dentry->d_parent; | |
384 | if (parent && !spin_trylock(&parent->d_lock)) { | |
873feea0 NP |
385 | if (inode) |
386 | spin_unlock(&inode->i_lock); | |
77812a1e NP |
387 | goto relock; |
388 | } | |
31e6b01f | 389 | |
77812a1e NP |
390 | if (ref) |
391 | dentry->d_count--; | |
392 | /* if dentry was on the d_lru list delete it from there */ | |
393 | dentry_lru_del(dentry); | |
394 | /* if it was on the hash then remove it */ | |
395 | __d_drop(dentry); | |
396 | return d_kill(dentry, parent); | |
397 | } | |
398 | ||
1da177e4 LT |
399 | /* |
400 | * This is dput | |
401 | * | |
402 | * This is complicated by the fact that we do not want to put | |
403 | * dentries that are no longer on any hash chain on the unused | |
404 | * list: we'd much rather just get rid of them immediately. | |
405 | * | |
406 | * However, that implies that we have to traverse the dentry | |
407 | * tree upwards to the parents which might _also_ now be | |
408 | * scheduled for deletion (it may have been only waiting for | |
409 | * its last child to go away). | |
410 | * | |
411 | * This tail recursion is done by hand as we don't want to depend | |
412 | * on the compiler to always get this right (gcc generally doesn't). | |
413 | * Real recursion would eat up our stack space. | |
414 | */ | |
415 | ||
416 | /* | |
417 | * dput - release a dentry | |
418 | * @dentry: dentry to release | |
419 | * | |
420 | * Release a dentry. This will drop the usage count and if appropriate | |
421 | * call the dentry unlink method as well as removing it from the queues and | |
422 | * releasing its resources. If the parent dentries were scheduled for release | |
423 | * they too may now get deleted. | |
1da177e4 | 424 | */ |
1da177e4 LT |
425 | void dput(struct dentry *dentry) |
426 | { | |
427 | if (!dentry) | |
428 | return; | |
429 | ||
430 | repeat: | |
b7ab39f6 | 431 | if (dentry->d_count == 1) |
1da177e4 | 432 | might_sleep(); |
1da177e4 | 433 | spin_lock(&dentry->d_lock); |
61f3dee4 NP |
434 | BUG_ON(!dentry->d_count); |
435 | if (dentry->d_count > 1) { | |
436 | dentry->d_count--; | |
1da177e4 | 437 | spin_unlock(&dentry->d_lock); |
1da177e4 LT |
438 | return; |
439 | } | |
440 | ||
fb045adb | 441 | if (dentry->d_flags & DCACHE_OP_DELETE) { |
1da177e4 | 442 | if (dentry->d_op->d_delete(dentry)) |
61f3dee4 | 443 | goto kill_it; |
1da177e4 | 444 | } |
265ac902 | 445 | |
1da177e4 LT |
446 | /* Unreachable? Get rid of it */ |
447 | if (d_unhashed(dentry)) | |
448 | goto kill_it; | |
265ac902 NP |
449 | |
450 | /* Otherwise leave it cached and ensure it's on the LRU */ | |
451 | dentry->d_flags |= DCACHE_REFERENCED; | |
a4633357 | 452 | dentry_lru_add(dentry); |
265ac902 | 453 | |
61f3dee4 NP |
454 | dentry->d_count--; |
455 | spin_unlock(&dentry->d_lock); | |
1da177e4 LT |
456 | return; |
457 | ||
d52b9086 | 458 | kill_it: |
77812a1e | 459 | dentry = dentry_kill(dentry, 1); |
d52b9086 MS |
460 | if (dentry) |
461 | goto repeat; | |
1da177e4 | 462 | } |
ec4f8605 | 463 | EXPORT_SYMBOL(dput); |
1da177e4 LT |
464 | |
465 | /** | |
466 | * d_invalidate - invalidate a dentry | |
467 | * @dentry: dentry to invalidate | |
468 | * | |
469 | * Try to invalidate the dentry if it turns out to be | |
470 | * possible. If there are other dentries that can be | |
471 | * reached through this one we can't delete it and we | |
472 | * return -EBUSY. On success we return 0. | |
473 | * | |
474 | * no dcache lock. | |
475 | */ | |
476 | ||
477 | int d_invalidate(struct dentry * dentry) | |
478 | { | |
479 | /* | |
480 | * If it's already been dropped, return OK. | |
481 | */ | |
da502956 | 482 | spin_lock(&dentry->d_lock); |
1da177e4 | 483 | if (d_unhashed(dentry)) { |
da502956 | 484 | spin_unlock(&dentry->d_lock); |
1da177e4 LT |
485 | return 0; |
486 | } | |
487 | /* | |
488 | * Check whether to do a partial shrink_dcache | |
489 | * to get rid of unused child entries. | |
490 | */ | |
491 | if (!list_empty(&dentry->d_subdirs)) { | |
da502956 | 492 | spin_unlock(&dentry->d_lock); |
1da177e4 | 493 | shrink_dcache_parent(dentry); |
da502956 | 494 | spin_lock(&dentry->d_lock); |
1da177e4 LT |
495 | } |
496 | ||
497 | /* | |
498 | * Somebody else still using it? | |
499 | * | |
500 | * If it's a directory, we can't drop it | |
501 | * for fear of somebody re-populating it | |
502 | * with children (even though dropping it | |
503 | * would make it unreachable from the root, | |
504 | * we might still populate it if it was a | |
505 | * working directory or similar). | |
506 | */ | |
b7ab39f6 | 507 | if (dentry->d_count > 1) { |
1da177e4 LT |
508 | if (dentry->d_inode && S_ISDIR(dentry->d_inode->i_mode)) { |
509 | spin_unlock(&dentry->d_lock); | |
1da177e4 LT |
510 | return -EBUSY; |
511 | } | |
512 | } | |
513 | ||
514 | __d_drop(dentry); | |
515 | spin_unlock(&dentry->d_lock); | |
1da177e4 LT |
516 | return 0; |
517 | } | |
ec4f8605 | 518 | EXPORT_SYMBOL(d_invalidate); |
1da177e4 | 519 | |
b5c84bf6 | 520 | /* This must be called with d_lock held */ |
dc0474be | 521 | static inline void __dget_dlock(struct dentry *dentry) |
23044507 | 522 | { |
b7ab39f6 | 523 | dentry->d_count++; |
23044507 NP |
524 | } |
525 | ||
dc0474be | 526 | static inline void __dget(struct dentry *dentry) |
1da177e4 | 527 | { |
23044507 | 528 | spin_lock(&dentry->d_lock); |
dc0474be | 529 | __dget_dlock(dentry); |
23044507 | 530 | spin_unlock(&dentry->d_lock); |
1da177e4 LT |
531 | } |
532 | ||
b7ab39f6 NP |
533 | struct dentry *dget_parent(struct dentry *dentry) |
534 | { | |
535 | struct dentry *ret; | |
536 | ||
537 | repeat: | |
a734eb45 NP |
538 | /* |
539 | * Don't need rcu_dereference because we re-check it was correct under | |
540 | * the lock. | |
541 | */ | |
542 | rcu_read_lock(); | |
b7ab39f6 | 543 | ret = dentry->d_parent; |
a734eb45 NP |
544 | if (!ret) { |
545 | rcu_read_unlock(); | |
b7ab39f6 NP |
546 | goto out; |
547 | } | |
a734eb45 NP |
548 | spin_lock(&ret->d_lock); |
549 | if (unlikely(ret != dentry->d_parent)) { | |
550 | spin_unlock(&ret->d_lock); | |
551 | rcu_read_unlock(); | |
b7ab39f6 NP |
552 | goto repeat; |
553 | } | |
a734eb45 | 554 | rcu_read_unlock(); |
b7ab39f6 NP |
555 | BUG_ON(!ret->d_count); |
556 | ret->d_count++; | |
557 | spin_unlock(&ret->d_lock); | |
558 | out: | |
b7ab39f6 NP |
559 | return ret; |
560 | } | |
561 | EXPORT_SYMBOL(dget_parent); | |
562 | ||
1da177e4 LT |
563 | /** |
564 | * d_find_alias - grab a hashed alias of inode | |
565 | * @inode: inode in question | |
566 | * @want_discon: flag, used by d_splice_alias, to request | |
567 | * that only a DISCONNECTED alias be returned. | |
568 | * | |
569 | * If inode has a hashed alias, or is a directory and has any alias, | |
570 | * acquire the reference to alias and return it. Otherwise return NULL. | |
571 | * Notice that if inode is a directory there can be only one alias and | |
572 | * it can be unhashed only if it has no children, or if it is the root | |
573 | * of a filesystem. | |
574 | * | |
21c0d8fd | 575 | * If the inode has an IS_ROOT, DCACHE_DISCONNECTED alias, then prefer |
1da177e4 | 576 | * any other hashed alias over that one unless @want_discon is set, |
21c0d8fd | 577 | * in which case only return an IS_ROOT, DCACHE_DISCONNECTED alias. |
1da177e4 | 578 | */ |
da502956 | 579 | static struct dentry *__d_find_alias(struct inode *inode, int want_discon) |
1da177e4 | 580 | { |
da502956 | 581 | struct dentry *alias, *discon_alias; |
1da177e4 | 582 | |
da502956 NP |
583 | again: |
584 | discon_alias = NULL; | |
585 | list_for_each_entry(alias, &inode->i_dentry, d_alias) { | |
586 | spin_lock(&alias->d_lock); | |
1da177e4 | 587 | if (S_ISDIR(inode->i_mode) || !d_unhashed(alias)) { |
21c0d8fd | 588 | if (IS_ROOT(alias) && |
da502956 | 589 | (alias->d_flags & DCACHE_DISCONNECTED)) { |
1da177e4 | 590 | discon_alias = alias; |
da502956 | 591 | } else if (!want_discon) { |
dc0474be | 592 | __dget_dlock(alias); |
da502956 NP |
593 | spin_unlock(&alias->d_lock); |
594 | return alias; | |
595 | } | |
596 | } | |
597 | spin_unlock(&alias->d_lock); | |
598 | } | |
599 | if (discon_alias) { | |
600 | alias = discon_alias; | |
601 | spin_lock(&alias->d_lock); | |
602 | if (S_ISDIR(inode->i_mode) || !d_unhashed(alias)) { | |
603 | if (IS_ROOT(alias) && | |
604 | (alias->d_flags & DCACHE_DISCONNECTED)) { | |
dc0474be | 605 | __dget_dlock(alias); |
da502956 | 606 | spin_unlock(&alias->d_lock); |
1da177e4 LT |
607 | return alias; |
608 | } | |
609 | } | |
da502956 NP |
610 | spin_unlock(&alias->d_lock); |
611 | goto again; | |
1da177e4 | 612 | } |
da502956 | 613 | return NULL; |
1da177e4 LT |
614 | } |
615 | ||
da502956 | 616 | struct dentry *d_find_alias(struct inode *inode) |
1da177e4 | 617 | { |
214fda1f DH |
618 | struct dentry *de = NULL; |
619 | ||
620 | if (!list_empty(&inode->i_dentry)) { | |
873feea0 | 621 | spin_lock(&inode->i_lock); |
214fda1f | 622 | de = __d_find_alias(inode, 0); |
873feea0 | 623 | spin_unlock(&inode->i_lock); |
214fda1f | 624 | } |
1da177e4 LT |
625 | return de; |
626 | } | |
ec4f8605 | 627 | EXPORT_SYMBOL(d_find_alias); |
1da177e4 LT |
628 | |
629 | /* | |
630 | * Try to kill dentries associated with this inode. | |
631 | * WARNING: you must own a reference to inode. | |
632 | */ | |
633 | void d_prune_aliases(struct inode *inode) | |
634 | { | |
0cdca3f9 | 635 | struct dentry *dentry; |
1da177e4 | 636 | restart: |
873feea0 | 637 | spin_lock(&inode->i_lock); |
0cdca3f9 | 638 | list_for_each_entry(dentry, &inode->i_dentry, d_alias) { |
1da177e4 | 639 | spin_lock(&dentry->d_lock); |
b7ab39f6 | 640 | if (!dentry->d_count) { |
dc0474be | 641 | __dget_dlock(dentry); |
1da177e4 LT |
642 | __d_drop(dentry); |
643 | spin_unlock(&dentry->d_lock); | |
873feea0 | 644 | spin_unlock(&inode->i_lock); |
1da177e4 LT |
645 | dput(dentry); |
646 | goto restart; | |
647 | } | |
648 | spin_unlock(&dentry->d_lock); | |
649 | } | |
873feea0 | 650 | spin_unlock(&inode->i_lock); |
1da177e4 | 651 | } |
ec4f8605 | 652 | EXPORT_SYMBOL(d_prune_aliases); |
1da177e4 LT |
653 | |
654 | /* | |
77812a1e NP |
655 | * Try to throw away a dentry - free the inode, dput the parent. |
656 | * Requires dentry->d_lock is held, and dentry->d_count == 0. | |
657 | * Releases dentry->d_lock. | |
d702ccb3 | 658 | * |
77812a1e | 659 | * This may fail if locks cannot be acquired no problem, just try again. |
1da177e4 | 660 | */ |
77812a1e | 661 | static void try_prune_one_dentry(struct dentry *dentry) |
31f3e0b3 | 662 | __releases(dentry->d_lock) |
1da177e4 | 663 | { |
77812a1e | 664 | struct dentry *parent; |
d52b9086 | 665 | |
77812a1e | 666 | parent = dentry_kill(dentry, 0); |
d52b9086 | 667 | /* |
77812a1e NP |
668 | * If dentry_kill returns NULL, we have nothing more to do. |
669 | * if it returns the same dentry, trylocks failed. In either | |
670 | * case, just loop again. | |
671 | * | |
672 | * Otherwise, we need to prune ancestors too. This is necessary | |
673 | * to prevent quadratic behavior of shrink_dcache_parent(), but | |
674 | * is also expected to be beneficial in reducing dentry cache | |
675 | * fragmentation. | |
d52b9086 | 676 | */ |
77812a1e NP |
677 | if (!parent) |
678 | return; | |
679 | if (parent == dentry) | |
680 | return; | |
681 | ||
682 | /* Prune ancestors. */ | |
683 | dentry = parent; | |
d52b9086 | 684 | while (dentry) { |
b7ab39f6 | 685 | spin_lock(&dentry->d_lock); |
89e60548 NP |
686 | if (dentry->d_count > 1) { |
687 | dentry->d_count--; | |
688 | spin_unlock(&dentry->d_lock); | |
689 | return; | |
690 | } | |
77812a1e | 691 | dentry = dentry_kill(dentry, 1); |
d52b9086 | 692 | } |
1da177e4 LT |
693 | } |
694 | ||
3049cfe2 | 695 | static void shrink_dentry_list(struct list_head *list) |
1da177e4 | 696 | { |
da3bbdd4 | 697 | struct dentry *dentry; |
da3bbdd4 | 698 | |
ec33679d NP |
699 | rcu_read_lock(); |
700 | for (;;) { | |
ec33679d NP |
701 | dentry = list_entry_rcu(list->prev, struct dentry, d_lru); |
702 | if (&dentry->d_lru == list) | |
703 | break; /* empty */ | |
704 | spin_lock(&dentry->d_lock); | |
705 | if (dentry != list_entry(list->prev, struct dentry, d_lru)) { | |
706 | spin_unlock(&dentry->d_lock); | |
23044507 NP |
707 | continue; |
708 | } | |
709 | ||
1da177e4 LT |
710 | /* |
711 | * We found an inuse dentry which was not removed from | |
da3bbdd4 KM |
712 | * the LRU because of laziness during lookup. Do not free |
713 | * it - just keep it off the LRU list. | |
1da177e4 | 714 | */ |
b7ab39f6 | 715 | if (dentry->d_count) { |
ec33679d | 716 | dentry_lru_del(dentry); |
da3bbdd4 | 717 | spin_unlock(&dentry->d_lock); |
1da177e4 LT |
718 | continue; |
719 | } | |
ec33679d | 720 | |
ec33679d | 721 | rcu_read_unlock(); |
77812a1e NP |
722 | |
723 | try_prune_one_dentry(dentry); | |
724 | ||
ec33679d | 725 | rcu_read_lock(); |
da3bbdd4 | 726 | } |
ec33679d | 727 | rcu_read_unlock(); |
3049cfe2 CH |
728 | } |
729 | ||
730 | /** | |
731 | * __shrink_dcache_sb - shrink the dentry LRU on a given superblock | |
732 | * @sb: superblock to shrink dentry LRU. | |
733 | * @count: number of entries to prune | |
734 | * @flags: flags to control the dentry processing | |
735 | * | |
736 | * If flags contains DCACHE_REFERENCED reference dentries will not be pruned. | |
737 | */ | |
738 | static void __shrink_dcache_sb(struct super_block *sb, int *count, int flags) | |
739 | { | |
740 | /* called from prune_dcache() and shrink_dcache_parent() */ | |
741 | struct dentry *dentry; | |
742 | LIST_HEAD(referenced); | |
743 | LIST_HEAD(tmp); | |
744 | int cnt = *count; | |
745 | ||
23044507 NP |
746 | relock: |
747 | spin_lock(&dcache_lru_lock); | |
3049cfe2 CH |
748 | while (!list_empty(&sb->s_dentry_lru)) { |
749 | dentry = list_entry(sb->s_dentry_lru.prev, | |
750 | struct dentry, d_lru); | |
751 | BUG_ON(dentry->d_sb != sb); | |
752 | ||
23044507 NP |
753 | if (!spin_trylock(&dentry->d_lock)) { |
754 | spin_unlock(&dcache_lru_lock); | |
755 | cpu_relax(); | |
756 | goto relock; | |
757 | } | |
758 | ||
3049cfe2 CH |
759 | /* |
760 | * If we are honouring the DCACHE_REFERENCED flag and the | |
761 | * dentry has this flag set, don't free it. Clear the flag | |
762 | * and put it back on the LRU. | |
763 | */ | |
23044507 NP |
764 | if (flags & DCACHE_REFERENCED && |
765 | dentry->d_flags & DCACHE_REFERENCED) { | |
766 | dentry->d_flags &= ~DCACHE_REFERENCED; | |
767 | list_move(&dentry->d_lru, &referenced); | |
3049cfe2 | 768 | spin_unlock(&dentry->d_lock); |
23044507 NP |
769 | } else { |
770 | list_move_tail(&dentry->d_lru, &tmp); | |
771 | spin_unlock(&dentry->d_lock); | |
772 | if (!--cnt) | |
773 | break; | |
3049cfe2 | 774 | } |
ec33679d | 775 | cond_resched_lock(&dcache_lru_lock); |
3049cfe2 | 776 | } |
da3bbdd4 KM |
777 | if (!list_empty(&referenced)) |
778 | list_splice(&referenced, &sb->s_dentry_lru); | |
23044507 | 779 | spin_unlock(&dcache_lru_lock); |
ec33679d NP |
780 | |
781 | shrink_dentry_list(&tmp); | |
782 | ||
783 | *count = cnt; | |
da3bbdd4 KM |
784 | } |
785 | ||
786 | /** | |
787 | * prune_dcache - shrink the dcache | |
788 | * @count: number of entries to try to free | |
789 | * | |
790 | * Shrink the dcache. This is done when we need more memory, or simply when we | |
791 | * need to unmount something (at which point we need to unuse all dentries). | |
792 | * | |
793 | * This function may fail to free any resources if all the dentries are in use. | |
794 | */ | |
795 | static void prune_dcache(int count) | |
796 | { | |
dca33252 | 797 | struct super_block *sb, *p = NULL; |
da3bbdd4 | 798 | int w_count; |
86c8749e | 799 | int unused = dentry_stat.nr_unused; |
da3bbdd4 KM |
800 | int prune_ratio; |
801 | int pruned; | |
802 | ||
803 | if (unused == 0 || count == 0) | |
804 | return; | |
da3bbdd4 KM |
805 | if (count >= unused) |
806 | prune_ratio = 1; | |
807 | else | |
808 | prune_ratio = unused / count; | |
809 | spin_lock(&sb_lock); | |
dca33252 | 810 | list_for_each_entry(sb, &super_blocks, s_list) { |
551de6f3 AV |
811 | if (list_empty(&sb->s_instances)) |
812 | continue; | |
da3bbdd4 | 813 | if (sb->s_nr_dentry_unused == 0) |
1da177e4 | 814 | continue; |
da3bbdd4 KM |
815 | sb->s_count++; |
816 | /* Now, we reclaim unused dentrins with fairness. | |
817 | * We reclaim them same percentage from each superblock. | |
818 | * We calculate number of dentries to scan on this sb | |
819 | * as follows, but the implementation is arranged to avoid | |
820 | * overflows: | |
821 | * number of dentries to scan on this sb = | |
822 | * count * (number of dentries on this sb / | |
823 | * number of dentries in the machine) | |
0feae5c4 | 824 | */ |
da3bbdd4 KM |
825 | spin_unlock(&sb_lock); |
826 | if (prune_ratio != 1) | |
827 | w_count = (sb->s_nr_dentry_unused / prune_ratio) + 1; | |
828 | else | |
829 | w_count = sb->s_nr_dentry_unused; | |
830 | pruned = w_count; | |
0feae5c4 | 831 | /* |
da3bbdd4 KM |
832 | * We need to be sure this filesystem isn't being unmounted, |
833 | * otherwise we could race with generic_shutdown_super(), and | |
834 | * end up holding a reference to an inode while the filesystem | |
835 | * is unmounted. So we try to get s_umount, and make sure | |
836 | * s_root isn't NULL. | |
0feae5c4 | 837 | */ |
da3bbdd4 KM |
838 | if (down_read_trylock(&sb->s_umount)) { |
839 | if ((sb->s_root != NULL) && | |
840 | (!list_empty(&sb->s_dentry_lru))) { | |
da3bbdd4 KM |
841 | __shrink_dcache_sb(sb, &w_count, |
842 | DCACHE_REFERENCED); | |
843 | pruned -= w_count; | |
0feae5c4 | 844 | } |
da3bbdd4 | 845 | up_read(&sb->s_umount); |
0feae5c4 | 846 | } |
da3bbdd4 | 847 | spin_lock(&sb_lock); |
dca33252 AV |
848 | if (p) |
849 | __put_super(p); | |
da3bbdd4 | 850 | count -= pruned; |
dca33252 | 851 | p = sb; |
79893c17 AV |
852 | /* more work left to do? */ |
853 | if (count <= 0) | |
854 | break; | |
1da177e4 | 855 | } |
dca33252 AV |
856 | if (p) |
857 | __put_super(p); | |
da3bbdd4 | 858 | spin_unlock(&sb_lock); |
1da177e4 LT |
859 | } |
860 | ||
1da177e4 LT |
861 | /** |
862 | * shrink_dcache_sb - shrink dcache for a superblock | |
863 | * @sb: superblock | |
864 | * | |
3049cfe2 CH |
865 | * Shrink the dcache for the specified super block. This is used to free |
866 | * the dcache before unmounting a file system. | |
1da177e4 | 867 | */ |
3049cfe2 | 868 | void shrink_dcache_sb(struct super_block *sb) |
1da177e4 | 869 | { |
3049cfe2 CH |
870 | LIST_HEAD(tmp); |
871 | ||
23044507 | 872 | spin_lock(&dcache_lru_lock); |
3049cfe2 CH |
873 | while (!list_empty(&sb->s_dentry_lru)) { |
874 | list_splice_init(&sb->s_dentry_lru, &tmp); | |
ec33679d | 875 | spin_unlock(&dcache_lru_lock); |
3049cfe2 | 876 | shrink_dentry_list(&tmp); |
ec33679d | 877 | spin_lock(&dcache_lru_lock); |
3049cfe2 | 878 | } |
23044507 | 879 | spin_unlock(&dcache_lru_lock); |
1da177e4 | 880 | } |
ec4f8605 | 881 | EXPORT_SYMBOL(shrink_dcache_sb); |
1da177e4 | 882 | |
c636ebdb DH |
883 | /* |
884 | * destroy a single subtree of dentries for unmount | |
885 | * - see the comments on shrink_dcache_for_umount() for a description of the | |
886 | * locking | |
887 | */ | |
888 | static void shrink_dcache_for_umount_subtree(struct dentry *dentry) | |
889 | { | |
890 | struct dentry *parent; | |
f8713576 | 891 | unsigned detached = 0; |
c636ebdb DH |
892 | |
893 | BUG_ON(!IS_ROOT(dentry)); | |
894 | ||
895 | /* detach this root from the system */ | |
23044507 | 896 | spin_lock(&dentry->d_lock); |
a4633357 | 897 | dentry_lru_del(dentry); |
c636ebdb | 898 | __d_drop(dentry); |
da502956 | 899 | spin_unlock(&dentry->d_lock); |
c636ebdb DH |
900 | |
901 | for (;;) { | |
902 | /* descend to the first leaf in the current subtree */ | |
903 | while (!list_empty(&dentry->d_subdirs)) { | |
904 | struct dentry *loop; | |
905 | ||
906 | /* this is a branch with children - detach all of them | |
907 | * from the system in one go */ | |
2fd6b7f5 | 908 | spin_lock(&dentry->d_lock); |
c636ebdb DH |
909 | list_for_each_entry(loop, &dentry->d_subdirs, |
910 | d_u.d_child) { | |
2fd6b7f5 NP |
911 | spin_lock_nested(&loop->d_lock, |
912 | DENTRY_D_LOCK_NESTED); | |
a4633357 | 913 | dentry_lru_del(loop); |
c636ebdb | 914 | __d_drop(loop); |
da502956 | 915 | spin_unlock(&loop->d_lock); |
c636ebdb | 916 | } |
2fd6b7f5 | 917 | spin_unlock(&dentry->d_lock); |
c636ebdb DH |
918 | |
919 | /* move to the first child */ | |
920 | dentry = list_entry(dentry->d_subdirs.next, | |
921 | struct dentry, d_u.d_child); | |
922 | } | |
923 | ||
924 | /* consume the dentries from this leaf up through its parents | |
925 | * until we find one with children or run out altogether */ | |
926 | do { | |
927 | struct inode *inode; | |
928 | ||
b7ab39f6 | 929 | if (dentry->d_count != 0) { |
c636ebdb DH |
930 | printk(KERN_ERR |
931 | "BUG: Dentry %p{i=%lx,n=%s}" | |
932 | " still in use (%d)" | |
933 | " [unmount of %s %s]\n", | |
934 | dentry, | |
935 | dentry->d_inode ? | |
936 | dentry->d_inode->i_ino : 0UL, | |
937 | dentry->d_name.name, | |
b7ab39f6 | 938 | dentry->d_count, |
c636ebdb DH |
939 | dentry->d_sb->s_type->name, |
940 | dentry->d_sb->s_id); | |
941 | BUG(); | |
942 | } | |
943 | ||
2fd6b7f5 | 944 | if (IS_ROOT(dentry)) { |
c636ebdb | 945 | parent = NULL; |
2fd6b7f5 NP |
946 | list_del(&dentry->d_u.d_child); |
947 | } else { | |
871c0067 | 948 | parent = dentry->d_parent; |
b7ab39f6 NP |
949 | spin_lock(&parent->d_lock); |
950 | parent->d_count--; | |
2fd6b7f5 | 951 | list_del(&dentry->d_u.d_child); |
b7ab39f6 | 952 | spin_unlock(&parent->d_lock); |
871c0067 | 953 | } |
c636ebdb | 954 | |
f8713576 | 955 | detached++; |
c636ebdb DH |
956 | |
957 | inode = dentry->d_inode; | |
958 | if (inode) { | |
959 | dentry->d_inode = NULL; | |
960 | list_del_init(&dentry->d_alias); | |
961 | if (dentry->d_op && dentry->d_op->d_iput) | |
962 | dentry->d_op->d_iput(dentry, inode); | |
963 | else | |
964 | iput(inode); | |
965 | } | |
966 | ||
967 | d_free(dentry); | |
968 | ||
969 | /* finished when we fall off the top of the tree, | |
970 | * otherwise we ascend to the parent and move to the | |
971 | * next sibling if there is one */ | |
972 | if (!parent) | |
312d3ca8 | 973 | return; |
c636ebdb | 974 | dentry = parent; |
c636ebdb DH |
975 | } while (list_empty(&dentry->d_subdirs)); |
976 | ||
977 | dentry = list_entry(dentry->d_subdirs.next, | |
978 | struct dentry, d_u.d_child); | |
979 | } | |
980 | } | |
981 | ||
982 | /* | |
983 | * destroy the dentries attached to a superblock on unmounting | |
b5c84bf6 | 984 | * - we don't need to use dentry->d_lock because: |
c636ebdb DH |
985 | * - the superblock is detached from all mountings and open files, so the |
986 | * dentry trees will not be rearranged by the VFS | |
987 | * - s_umount is write-locked, so the memory pressure shrinker will ignore | |
988 | * any dentries belonging to this superblock that it comes across | |
989 | * - the filesystem itself is no longer permitted to rearrange the dentries | |
990 | * in this superblock | |
991 | */ | |
992 | void shrink_dcache_for_umount(struct super_block *sb) | |
993 | { | |
994 | struct dentry *dentry; | |
995 | ||
996 | if (down_read_trylock(&sb->s_umount)) | |
997 | BUG(); | |
998 | ||
999 | dentry = sb->s_root; | |
1000 | sb->s_root = NULL; | |
b7ab39f6 NP |
1001 | spin_lock(&dentry->d_lock); |
1002 | dentry->d_count--; | |
1003 | spin_unlock(&dentry->d_lock); | |
c636ebdb DH |
1004 | shrink_dcache_for_umount_subtree(dentry); |
1005 | ||
ceb5bdc2 NP |
1006 | while (!hlist_bl_empty(&sb->s_anon)) { |
1007 | dentry = hlist_bl_entry(hlist_bl_first(&sb->s_anon), struct dentry, d_hash); | |
c636ebdb DH |
1008 | shrink_dcache_for_umount_subtree(dentry); |
1009 | } | |
1010 | } | |
1011 | ||
1da177e4 LT |
1012 | /* |
1013 | * Search for at least 1 mount point in the dentry's subdirs. | |
1014 | * We descend to the next level whenever the d_subdirs | |
1015 | * list is non-empty and continue searching. | |
1016 | */ | |
1017 | ||
1018 | /** | |
1019 | * have_submounts - check for mounts over a dentry | |
1020 | * @parent: dentry to check. | |
1021 | * | |
1022 | * Return true if the parent or its subdirectories contain | |
1023 | * a mount point | |
1024 | */ | |
1da177e4 LT |
1025 | int have_submounts(struct dentry *parent) |
1026 | { | |
949854d0 | 1027 | struct dentry *this_parent; |
1da177e4 | 1028 | struct list_head *next; |
949854d0 | 1029 | unsigned seq; |
58db63d0 | 1030 | int locked = 0; |
949854d0 | 1031 | |
949854d0 | 1032 | seq = read_seqbegin(&rename_lock); |
58db63d0 NP |
1033 | again: |
1034 | this_parent = parent; | |
1da177e4 | 1035 | |
1da177e4 LT |
1036 | if (d_mountpoint(parent)) |
1037 | goto positive; | |
2fd6b7f5 | 1038 | spin_lock(&this_parent->d_lock); |
1da177e4 LT |
1039 | repeat: |
1040 | next = this_parent->d_subdirs.next; | |
1041 | resume: | |
1042 | while (next != &this_parent->d_subdirs) { | |
1043 | struct list_head *tmp = next; | |
5160ee6f | 1044 | struct dentry *dentry = list_entry(tmp, struct dentry, d_u.d_child); |
1da177e4 | 1045 | next = tmp->next; |
2fd6b7f5 NP |
1046 | |
1047 | spin_lock_nested(&dentry->d_lock, DENTRY_D_LOCK_NESTED); | |
1da177e4 | 1048 | /* Have we found a mount point ? */ |
2fd6b7f5 NP |
1049 | if (d_mountpoint(dentry)) { |
1050 | spin_unlock(&dentry->d_lock); | |
1051 | spin_unlock(&this_parent->d_lock); | |
1da177e4 | 1052 | goto positive; |
2fd6b7f5 | 1053 | } |
1da177e4 | 1054 | if (!list_empty(&dentry->d_subdirs)) { |
2fd6b7f5 NP |
1055 | spin_unlock(&this_parent->d_lock); |
1056 | spin_release(&dentry->d_lock.dep_map, 1, _RET_IP_); | |
1da177e4 | 1057 | this_parent = dentry; |
2fd6b7f5 | 1058 | spin_acquire(&this_parent->d_lock.dep_map, 0, 1, _RET_IP_); |
1da177e4 LT |
1059 | goto repeat; |
1060 | } | |
2fd6b7f5 | 1061 | spin_unlock(&dentry->d_lock); |
1da177e4 LT |
1062 | } |
1063 | /* | |
1064 | * All done at this level ... ascend and resume the search. | |
1065 | */ | |
1066 | if (this_parent != parent) { | |
949854d0 NP |
1067 | struct dentry *tmp; |
1068 | struct dentry *child; | |
1069 | ||
1070 | tmp = this_parent->d_parent; | |
1071 | rcu_read_lock(); | |
2fd6b7f5 | 1072 | spin_unlock(&this_parent->d_lock); |
949854d0 NP |
1073 | child = this_parent; |
1074 | this_parent = tmp; | |
2fd6b7f5 | 1075 | spin_lock(&this_parent->d_lock); |
949854d0 NP |
1076 | /* might go back up the wrong parent if we have had a rename |
1077 | * or deletion */ | |
1078 | if (this_parent != child->d_parent || | |
58db63d0 | 1079 | (!locked && read_seqretry(&rename_lock, seq))) { |
949854d0 | 1080 | spin_unlock(&this_parent->d_lock); |
949854d0 NP |
1081 | rcu_read_unlock(); |
1082 | goto rename_retry; | |
1083 | } | |
1084 | rcu_read_unlock(); | |
1085 | next = child->d_u.d_child.next; | |
1da177e4 LT |
1086 | goto resume; |
1087 | } | |
2fd6b7f5 | 1088 | spin_unlock(&this_parent->d_lock); |
58db63d0 | 1089 | if (!locked && read_seqretry(&rename_lock, seq)) |
949854d0 | 1090 | goto rename_retry; |
58db63d0 NP |
1091 | if (locked) |
1092 | write_sequnlock(&rename_lock); | |
1da177e4 LT |
1093 | return 0; /* No mount points found in tree */ |
1094 | positive: | |
58db63d0 | 1095 | if (!locked && read_seqretry(&rename_lock, seq)) |
949854d0 | 1096 | goto rename_retry; |
58db63d0 NP |
1097 | if (locked) |
1098 | write_sequnlock(&rename_lock); | |
1da177e4 | 1099 | return 1; |
58db63d0 NP |
1100 | |
1101 | rename_retry: | |
1102 | locked = 1; | |
1103 | write_seqlock(&rename_lock); | |
1104 | goto again; | |
1da177e4 | 1105 | } |
ec4f8605 | 1106 | EXPORT_SYMBOL(have_submounts); |
1da177e4 LT |
1107 | |
1108 | /* | |
1109 | * Search the dentry child list for the specified parent, | |
1110 | * and move any unused dentries to the end of the unused | |
1111 | * list for prune_dcache(). We descend to the next level | |
1112 | * whenever the d_subdirs list is non-empty and continue | |
1113 | * searching. | |
1114 | * | |
1115 | * It returns zero iff there are no unused children, | |
1116 | * otherwise it returns the number of children moved to | |
1117 | * the end of the unused list. This may not be the total | |
1118 | * number of unused children, because select_parent can | |
1119 | * drop the lock and return early due to latency | |
1120 | * constraints. | |
1121 | */ | |
1122 | static int select_parent(struct dentry * parent) | |
1123 | { | |
949854d0 | 1124 | struct dentry *this_parent; |
1da177e4 | 1125 | struct list_head *next; |
949854d0 | 1126 | unsigned seq; |
1da177e4 | 1127 | int found = 0; |
58db63d0 | 1128 | int locked = 0; |
1da177e4 | 1129 | |
949854d0 | 1130 | seq = read_seqbegin(&rename_lock); |
58db63d0 NP |
1131 | again: |
1132 | this_parent = parent; | |
2fd6b7f5 | 1133 | spin_lock(&this_parent->d_lock); |
1da177e4 LT |
1134 | repeat: |
1135 | next = this_parent->d_subdirs.next; | |
1136 | resume: | |
1137 | while (next != &this_parent->d_subdirs) { | |
1138 | struct list_head *tmp = next; | |
5160ee6f | 1139 | struct dentry *dentry = list_entry(tmp, struct dentry, d_u.d_child); |
1da177e4 LT |
1140 | next = tmp->next; |
1141 | ||
2fd6b7f5 | 1142 | spin_lock_nested(&dentry->d_lock, DENTRY_D_LOCK_NESTED); |
23044507 | 1143 | |
1da177e4 LT |
1144 | /* |
1145 | * move only zero ref count dentries to the end | |
1146 | * of the unused list for prune_dcache | |
1147 | */ | |
b7ab39f6 | 1148 | if (!dentry->d_count) { |
a4633357 | 1149 | dentry_lru_move_tail(dentry); |
1da177e4 | 1150 | found++; |
a4633357 CH |
1151 | } else { |
1152 | dentry_lru_del(dentry); | |
1da177e4 LT |
1153 | } |
1154 | ||
1155 | /* | |
1156 | * We can return to the caller if we have found some (this | |
1157 | * ensures forward progress). We'll be coming back to find | |
1158 | * the rest. | |
1159 | */ | |
2fd6b7f5 NP |
1160 | if (found && need_resched()) { |
1161 | spin_unlock(&dentry->d_lock); | |
1da177e4 | 1162 | goto out; |
2fd6b7f5 | 1163 | } |
1da177e4 LT |
1164 | |
1165 | /* | |
1166 | * Descend a level if the d_subdirs list is non-empty. | |
1167 | */ | |
1168 | if (!list_empty(&dentry->d_subdirs)) { | |
2fd6b7f5 NP |
1169 | spin_unlock(&this_parent->d_lock); |
1170 | spin_release(&dentry->d_lock.dep_map, 1, _RET_IP_); | |
1da177e4 | 1171 | this_parent = dentry; |
2fd6b7f5 | 1172 | spin_acquire(&this_parent->d_lock.dep_map, 0, 1, _RET_IP_); |
1da177e4 LT |
1173 | goto repeat; |
1174 | } | |
2fd6b7f5 NP |
1175 | |
1176 | spin_unlock(&dentry->d_lock); | |
1da177e4 LT |
1177 | } |
1178 | /* | |
1179 | * All done at this level ... ascend and resume the search. | |
1180 | */ | |
1181 | if (this_parent != parent) { | |
2fd6b7f5 | 1182 | struct dentry *tmp; |
949854d0 NP |
1183 | struct dentry *child; |
1184 | ||
2fd6b7f5 | 1185 | tmp = this_parent->d_parent; |
949854d0 | 1186 | rcu_read_lock(); |
2fd6b7f5 | 1187 | spin_unlock(&this_parent->d_lock); |
949854d0 | 1188 | child = this_parent; |
2fd6b7f5 NP |
1189 | this_parent = tmp; |
1190 | spin_lock(&this_parent->d_lock); | |
949854d0 NP |
1191 | /* might go back up the wrong parent if we have had a rename |
1192 | * or deletion */ | |
1193 | if (this_parent != child->d_parent || | |
58db63d0 | 1194 | (!locked && read_seqretry(&rename_lock, seq))) { |
949854d0 | 1195 | spin_unlock(&this_parent->d_lock); |
949854d0 NP |
1196 | rcu_read_unlock(); |
1197 | goto rename_retry; | |
1198 | } | |
1199 | rcu_read_unlock(); | |
1200 | next = child->d_u.d_child.next; | |
1da177e4 LT |
1201 | goto resume; |
1202 | } | |
1203 | out: | |
2fd6b7f5 | 1204 | spin_unlock(&this_parent->d_lock); |
58db63d0 | 1205 | if (!locked && read_seqretry(&rename_lock, seq)) |
949854d0 | 1206 | goto rename_retry; |
58db63d0 NP |
1207 | if (locked) |
1208 | write_sequnlock(&rename_lock); | |
1da177e4 | 1209 | return found; |
58db63d0 NP |
1210 | |
1211 | rename_retry: | |
1212 | if (found) | |
1213 | return found; | |
1214 | locked = 1; | |
1215 | write_seqlock(&rename_lock); | |
1216 | goto again; | |
1da177e4 LT |
1217 | } |
1218 | ||
1219 | /** | |
1220 | * shrink_dcache_parent - prune dcache | |
1221 | * @parent: parent of entries to prune | |
1222 | * | |
1223 | * Prune the dcache to remove unused children of the parent dentry. | |
1224 | */ | |
1225 | ||
1226 | void shrink_dcache_parent(struct dentry * parent) | |
1227 | { | |
da3bbdd4 | 1228 | struct super_block *sb = parent->d_sb; |
1da177e4 LT |
1229 | int found; |
1230 | ||
1231 | while ((found = select_parent(parent)) != 0) | |
da3bbdd4 | 1232 | __shrink_dcache_sb(sb, &found, 0); |
1da177e4 | 1233 | } |
ec4f8605 | 1234 | EXPORT_SYMBOL(shrink_dcache_parent); |
1da177e4 | 1235 | |
1da177e4 LT |
1236 | /* |
1237 | * Scan `nr' dentries and return the number which remain. | |
1238 | * | |
1239 | * We need to avoid reentering the filesystem if the caller is performing a | |
1240 | * GFP_NOFS allocation attempt. One example deadlock is: | |
1241 | * | |
1242 | * ext2_new_block->getblk->GFP->shrink_dcache_memory->prune_dcache-> | |
1243 | * prune_one_dentry->dput->dentry_iput->iput->inode->i_sb->s_op->put_inode-> | |
1244 | * ext2_discard_prealloc->ext2_free_blocks->lock_super->DEADLOCK. | |
1245 | * | |
1246 | * In this case we return -1 to tell the caller that we baled. | |
1247 | */ | |
7f8275d0 | 1248 | static int shrink_dcache_memory(struct shrinker *shrink, int nr, gfp_t gfp_mask) |
1da177e4 LT |
1249 | { |
1250 | if (nr) { | |
1251 | if (!(gfp_mask & __GFP_FS)) | |
1252 | return -1; | |
da3bbdd4 | 1253 | prune_dcache(nr); |
1da177e4 | 1254 | } |
312d3ca8 | 1255 | |
86c8749e | 1256 | return (dentry_stat.nr_unused / 100) * sysctl_vfs_cache_pressure; |
1da177e4 LT |
1257 | } |
1258 | ||
8e1f936b RR |
1259 | static struct shrinker dcache_shrinker = { |
1260 | .shrink = shrink_dcache_memory, | |
1261 | .seeks = DEFAULT_SEEKS, | |
1262 | }; | |
1263 | ||
1da177e4 LT |
1264 | /** |
1265 | * d_alloc - allocate a dcache entry | |
1266 | * @parent: parent of entry to allocate | |
1267 | * @name: qstr of the name | |
1268 | * | |
1269 | * Allocates a dentry. It returns %NULL if there is insufficient memory | |
1270 | * available. On a success the dentry is returned. The name passed in is | |
1271 | * copied and the copy passed in may be reused after this call. | |
1272 | */ | |
1273 | ||
1274 | struct dentry *d_alloc(struct dentry * parent, const struct qstr *name) | |
1275 | { | |
1276 | struct dentry *dentry; | |
1277 | char *dname; | |
1278 | ||
e12ba74d | 1279 | dentry = kmem_cache_alloc(dentry_cache, GFP_KERNEL); |
1da177e4 LT |
1280 | if (!dentry) |
1281 | return NULL; | |
1282 | ||
1283 | if (name->len > DNAME_INLINE_LEN-1) { | |
1284 | dname = kmalloc(name->len + 1, GFP_KERNEL); | |
1285 | if (!dname) { | |
1286 | kmem_cache_free(dentry_cache, dentry); | |
1287 | return NULL; | |
1288 | } | |
1289 | } else { | |
1290 | dname = dentry->d_iname; | |
1291 | } | |
1292 | dentry->d_name.name = dname; | |
1293 | ||
1294 | dentry->d_name.len = name->len; | |
1295 | dentry->d_name.hash = name->hash; | |
1296 | memcpy(dname, name->name, name->len); | |
1297 | dname[name->len] = 0; | |
1298 | ||
b7ab39f6 | 1299 | dentry->d_count = 1; |
1da177e4 LT |
1300 | dentry->d_flags = DCACHE_UNHASHED; |
1301 | spin_lock_init(&dentry->d_lock); | |
31e6b01f | 1302 | seqcount_init(&dentry->d_seq); |
1da177e4 LT |
1303 | dentry->d_inode = NULL; |
1304 | dentry->d_parent = NULL; | |
1305 | dentry->d_sb = NULL; | |
1306 | dentry->d_op = NULL; | |
1307 | dentry->d_fsdata = NULL; | |
ceb5bdc2 | 1308 | INIT_HLIST_BL_NODE(&dentry->d_hash); |
1da177e4 LT |
1309 | INIT_LIST_HEAD(&dentry->d_lru); |
1310 | INIT_LIST_HEAD(&dentry->d_subdirs); | |
1311 | INIT_LIST_HEAD(&dentry->d_alias); | |
2fd6b7f5 | 1312 | INIT_LIST_HEAD(&dentry->d_u.d_child); |
1da177e4 LT |
1313 | |
1314 | if (parent) { | |
2fd6b7f5 | 1315 | spin_lock(&parent->d_lock); |
89ad485f NP |
1316 | /* |
1317 | * don't need child lock because it is not subject | |
1318 | * to concurrency here | |
1319 | */ | |
dc0474be NP |
1320 | __dget_dlock(parent); |
1321 | dentry->d_parent = parent; | |
1da177e4 | 1322 | dentry->d_sb = parent->d_sb; |
5160ee6f | 1323 | list_add(&dentry->d_u.d_child, &parent->d_subdirs); |
2fd6b7f5 | 1324 | spin_unlock(&parent->d_lock); |
2fd6b7f5 | 1325 | } |
1da177e4 | 1326 | |
3e880fb5 | 1327 | this_cpu_inc(nr_dentry); |
312d3ca8 | 1328 | |
1da177e4 LT |
1329 | return dentry; |
1330 | } | |
ec4f8605 | 1331 | EXPORT_SYMBOL(d_alloc); |
1da177e4 | 1332 | |
4b936885 NP |
1333 | struct dentry *d_alloc_pseudo(struct super_block *sb, const struct qstr *name) |
1334 | { | |
1335 | struct dentry *dentry = d_alloc(NULL, name); | |
1336 | if (dentry) { | |
1337 | dentry->d_sb = sb; | |
1338 | dentry->d_parent = dentry; | |
1339 | dentry->d_flags |= DCACHE_DISCONNECTED; | |
1340 | } | |
1341 | return dentry; | |
1342 | } | |
1343 | EXPORT_SYMBOL(d_alloc_pseudo); | |
1344 | ||
1da177e4 LT |
1345 | struct dentry *d_alloc_name(struct dentry *parent, const char *name) |
1346 | { | |
1347 | struct qstr q; | |
1348 | ||
1349 | q.name = name; | |
1350 | q.len = strlen(name); | |
1351 | q.hash = full_name_hash(q.name, q.len); | |
1352 | return d_alloc(parent, &q); | |
1353 | } | |
ef26ca97 | 1354 | EXPORT_SYMBOL(d_alloc_name); |
1da177e4 | 1355 | |
fb045adb NP |
1356 | void d_set_d_op(struct dentry *dentry, const struct dentry_operations *op) |
1357 | { | |
1358 | BUG_ON(dentry->d_op); | |
1359 | BUG_ON(dentry->d_flags & (DCACHE_OP_HASH | | |
1360 | DCACHE_OP_COMPARE | | |
1361 | DCACHE_OP_REVALIDATE | | |
1362 | DCACHE_OP_DELETE )); | |
1363 | dentry->d_op = op; | |
1364 | if (!op) | |
1365 | return; | |
1366 | if (op->d_hash) | |
1367 | dentry->d_flags |= DCACHE_OP_HASH; | |
1368 | if (op->d_compare) | |
1369 | dentry->d_flags |= DCACHE_OP_COMPARE; | |
1370 | if (op->d_revalidate) | |
1371 | dentry->d_flags |= DCACHE_OP_REVALIDATE; | |
1372 | if (op->d_delete) | |
1373 | dentry->d_flags |= DCACHE_OP_DELETE; | |
1374 | ||
1375 | } | |
1376 | EXPORT_SYMBOL(d_set_d_op); | |
1377 | ||
360da900 OH |
1378 | static void __d_instantiate(struct dentry *dentry, struct inode *inode) |
1379 | { | |
b23fb0a6 | 1380 | spin_lock(&dentry->d_lock); |
360da900 OH |
1381 | if (inode) |
1382 | list_add(&dentry->d_alias, &inode->i_dentry); | |
1383 | dentry->d_inode = inode; | |
31e6b01f | 1384 | dentry_rcuwalk_barrier(dentry); |
b23fb0a6 | 1385 | spin_unlock(&dentry->d_lock); |
360da900 OH |
1386 | fsnotify_d_instantiate(dentry, inode); |
1387 | } | |
1388 | ||
1da177e4 LT |
1389 | /** |
1390 | * d_instantiate - fill in inode information for a dentry | |
1391 | * @entry: dentry to complete | |
1392 | * @inode: inode to attach to this dentry | |
1393 | * | |
1394 | * Fill in inode information in the entry. | |
1395 | * | |
1396 | * This turns negative dentries into productive full members | |
1397 | * of society. | |
1398 | * | |
1399 | * NOTE! This assumes that the inode count has been incremented | |
1400 | * (or otherwise set) by the caller to indicate that it is now | |
1401 | * in use by the dcache. | |
1402 | */ | |
1403 | ||
1404 | void d_instantiate(struct dentry *entry, struct inode * inode) | |
1405 | { | |
28133c7b | 1406 | BUG_ON(!list_empty(&entry->d_alias)); |
873feea0 NP |
1407 | if (inode) |
1408 | spin_lock(&inode->i_lock); | |
360da900 | 1409 | __d_instantiate(entry, inode); |
873feea0 NP |
1410 | if (inode) |
1411 | spin_unlock(&inode->i_lock); | |
1da177e4 LT |
1412 | security_d_instantiate(entry, inode); |
1413 | } | |
ec4f8605 | 1414 | EXPORT_SYMBOL(d_instantiate); |
1da177e4 LT |
1415 | |
1416 | /** | |
1417 | * d_instantiate_unique - instantiate a non-aliased dentry | |
1418 | * @entry: dentry to instantiate | |
1419 | * @inode: inode to attach to this dentry | |
1420 | * | |
1421 | * Fill in inode information in the entry. On success, it returns NULL. | |
1422 | * If an unhashed alias of "entry" already exists, then we return the | |
e866cfa9 | 1423 | * aliased dentry instead and drop one reference to inode. |
1da177e4 LT |
1424 | * |
1425 | * Note that in order to avoid conflicts with rename() etc, the caller | |
1426 | * had better be holding the parent directory semaphore. | |
e866cfa9 OD |
1427 | * |
1428 | * This also assumes that the inode count has been incremented | |
1429 | * (or otherwise set) by the caller to indicate that it is now | |
1430 | * in use by the dcache. | |
1da177e4 | 1431 | */ |
770bfad8 DH |
1432 | static struct dentry *__d_instantiate_unique(struct dentry *entry, |
1433 | struct inode *inode) | |
1da177e4 LT |
1434 | { |
1435 | struct dentry *alias; | |
1436 | int len = entry->d_name.len; | |
1437 | const char *name = entry->d_name.name; | |
1438 | unsigned int hash = entry->d_name.hash; | |
1439 | ||
770bfad8 | 1440 | if (!inode) { |
360da900 | 1441 | __d_instantiate(entry, NULL); |
770bfad8 DH |
1442 | return NULL; |
1443 | } | |
1444 | ||
1da177e4 LT |
1445 | list_for_each_entry(alias, &inode->i_dentry, d_alias) { |
1446 | struct qstr *qstr = &alias->d_name; | |
1447 | ||
9abca360 NP |
1448 | /* |
1449 | * Don't need alias->d_lock here, because aliases with | |
1450 | * d_parent == entry->d_parent are not subject to name or | |
1451 | * parent changes, because the parent inode i_mutex is held. | |
1452 | */ | |
1da177e4 LT |
1453 | if (qstr->hash != hash) |
1454 | continue; | |
1455 | if (alias->d_parent != entry->d_parent) | |
1456 | continue; | |
9d55c369 | 1457 | if (dentry_cmp(qstr->name, qstr->len, name, len)) |
1da177e4 | 1458 | continue; |
dc0474be | 1459 | __dget(alias); |
1da177e4 LT |
1460 | return alias; |
1461 | } | |
770bfad8 | 1462 | |
360da900 | 1463 | __d_instantiate(entry, inode); |
1da177e4 LT |
1464 | return NULL; |
1465 | } | |
770bfad8 DH |
1466 | |
1467 | struct dentry *d_instantiate_unique(struct dentry *entry, struct inode *inode) | |
1468 | { | |
1469 | struct dentry *result; | |
1470 | ||
1471 | BUG_ON(!list_empty(&entry->d_alias)); | |
1472 | ||
873feea0 NP |
1473 | if (inode) |
1474 | spin_lock(&inode->i_lock); | |
770bfad8 | 1475 | result = __d_instantiate_unique(entry, inode); |
873feea0 NP |
1476 | if (inode) |
1477 | spin_unlock(&inode->i_lock); | |
770bfad8 DH |
1478 | |
1479 | if (!result) { | |
1480 | security_d_instantiate(entry, inode); | |
1481 | return NULL; | |
1482 | } | |
1483 | ||
1484 | BUG_ON(!d_unhashed(result)); | |
1485 | iput(inode); | |
1486 | return result; | |
1487 | } | |
1488 | ||
1da177e4 LT |
1489 | EXPORT_SYMBOL(d_instantiate_unique); |
1490 | ||
1491 | /** | |
1492 | * d_alloc_root - allocate root dentry | |
1493 | * @root_inode: inode to allocate the root for | |
1494 | * | |
1495 | * Allocate a root ("/") dentry for the inode given. The inode is | |
1496 | * instantiated and returned. %NULL is returned if there is insufficient | |
1497 | * memory or the inode passed is %NULL. | |
1498 | */ | |
1499 | ||
1500 | struct dentry * d_alloc_root(struct inode * root_inode) | |
1501 | { | |
1502 | struct dentry *res = NULL; | |
1503 | ||
1504 | if (root_inode) { | |
1505 | static const struct qstr name = { .name = "/", .len = 1 }; | |
1506 | ||
1507 | res = d_alloc(NULL, &name); | |
1508 | if (res) { | |
1509 | res->d_sb = root_inode->i_sb; | |
1510 | res->d_parent = res; | |
1511 | d_instantiate(res, root_inode); | |
1512 | } | |
1513 | } | |
1514 | return res; | |
1515 | } | |
ec4f8605 | 1516 | EXPORT_SYMBOL(d_alloc_root); |
1da177e4 | 1517 | |
4ea3ada2 CH |
1518 | /** |
1519 | * d_obtain_alias - find or allocate a dentry for a given inode | |
1520 | * @inode: inode to allocate the dentry for | |
1521 | * | |
1522 | * Obtain a dentry for an inode resulting from NFS filehandle conversion or | |
1523 | * similar open by handle operations. The returned dentry may be anonymous, | |
1524 | * or may have a full name (if the inode was already in the cache). | |
1525 | * | |
1526 | * When called on a directory inode, we must ensure that the inode only ever | |
1527 | * has one dentry. If a dentry is found, that is returned instead of | |
1528 | * allocating a new one. | |
1529 | * | |
1530 | * On successful return, the reference to the inode has been transferred | |
44003728 CH |
1531 | * to the dentry. In case of an error the reference on the inode is released. |
1532 | * To make it easier to use in export operations a %NULL or IS_ERR inode may | |
1533 | * be passed in and will be the error will be propagate to the return value, | |
1534 | * with a %NULL @inode replaced by ERR_PTR(-ESTALE). | |
4ea3ada2 CH |
1535 | */ |
1536 | struct dentry *d_obtain_alias(struct inode *inode) | |
1537 | { | |
9308a612 CH |
1538 | static const struct qstr anonstring = { .name = "" }; |
1539 | struct dentry *tmp; | |
1540 | struct dentry *res; | |
4ea3ada2 CH |
1541 | |
1542 | if (!inode) | |
44003728 | 1543 | return ERR_PTR(-ESTALE); |
4ea3ada2 CH |
1544 | if (IS_ERR(inode)) |
1545 | return ERR_CAST(inode); | |
1546 | ||
9308a612 CH |
1547 | res = d_find_alias(inode); |
1548 | if (res) | |
1549 | goto out_iput; | |
1550 | ||
1551 | tmp = d_alloc(NULL, &anonstring); | |
1552 | if (!tmp) { | |
1553 | res = ERR_PTR(-ENOMEM); | |
1554 | goto out_iput; | |
4ea3ada2 | 1555 | } |
9308a612 CH |
1556 | tmp->d_parent = tmp; /* make sure dput doesn't croak */ |
1557 | ||
b5c84bf6 | 1558 | |
873feea0 | 1559 | spin_lock(&inode->i_lock); |
9308a612 CH |
1560 | res = __d_find_alias(inode, 0); |
1561 | if (res) { | |
873feea0 | 1562 | spin_unlock(&inode->i_lock); |
9308a612 CH |
1563 | dput(tmp); |
1564 | goto out_iput; | |
1565 | } | |
1566 | ||
1567 | /* attach a disconnected dentry */ | |
1568 | spin_lock(&tmp->d_lock); | |
1569 | tmp->d_sb = inode->i_sb; | |
1570 | tmp->d_inode = inode; | |
1571 | tmp->d_flags |= DCACHE_DISCONNECTED; | |
9308a612 | 1572 | list_add(&tmp->d_alias, &inode->i_dentry); |
ceb5bdc2 NP |
1573 | bit_spin_lock(0, (unsigned long *)&tmp->d_sb->s_anon.first); |
1574 | tmp->d_flags &= ~DCACHE_UNHASHED; | |
1575 | hlist_bl_add_head(&tmp->d_hash, &tmp->d_sb->s_anon); | |
1576 | __bit_spin_unlock(0, (unsigned long *)&tmp->d_sb->s_anon.first); | |
9308a612 | 1577 | spin_unlock(&tmp->d_lock); |
873feea0 | 1578 | spin_unlock(&inode->i_lock); |
9308a612 | 1579 | |
9308a612 CH |
1580 | return tmp; |
1581 | ||
1582 | out_iput: | |
1583 | iput(inode); | |
1584 | return res; | |
4ea3ada2 | 1585 | } |
adc48720 | 1586 | EXPORT_SYMBOL(d_obtain_alias); |
1da177e4 LT |
1587 | |
1588 | /** | |
1589 | * d_splice_alias - splice a disconnected dentry into the tree if one exists | |
1590 | * @inode: the inode which may have a disconnected dentry | |
1591 | * @dentry: a negative dentry which we want to point to the inode. | |
1592 | * | |
1593 | * If inode is a directory and has a 'disconnected' dentry (i.e. IS_ROOT and | |
1594 | * DCACHE_DISCONNECTED), then d_move that in place of the given dentry | |
1595 | * and return it, else simply d_add the inode to the dentry and return NULL. | |
1596 | * | |
1597 | * This is needed in the lookup routine of any filesystem that is exportable | |
1598 | * (via knfsd) so that we can build dcache paths to directories effectively. | |
1599 | * | |
1600 | * If a dentry was found and moved, then it is returned. Otherwise NULL | |
1601 | * is returned. This matches the expected return value of ->lookup. | |
1602 | * | |
1603 | */ | |
1604 | struct dentry *d_splice_alias(struct inode *inode, struct dentry *dentry) | |
1605 | { | |
1606 | struct dentry *new = NULL; | |
1607 | ||
21c0d8fd | 1608 | if (inode && S_ISDIR(inode->i_mode)) { |
873feea0 | 1609 | spin_lock(&inode->i_lock); |
1da177e4 LT |
1610 | new = __d_find_alias(inode, 1); |
1611 | if (new) { | |
1612 | BUG_ON(!(new->d_flags & DCACHE_DISCONNECTED)); | |
873feea0 | 1613 | spin_unlock(&inode->i_lock); |
1da177e4 | 1614 | security_d_instantiate(new, inode); |
1da177e4 LT |
1615 | d_move(new, dentry); |
1616 | iput(inode); | |
1617 | } else { | |
873feea0 | 1618 | /* already taking inode->i_lock, so d_add() by hand */ |
360da900 | 1619 | __d_instantiate(dentry, inode); |
873feea0 | 1620 | spin_unlock(&inode->i_lock); |
1da177e4 LT |
1621 | security_d_instantiate(dentry, inode); |
1622 | d_rehash(dentry); | |
1623 | } | |
1624 | } else | |
1625 | d_add(dentry, inode); | |
1626 | return new; | |
1627 | } | |
ec4f8605 | 1628 | EXPORT_SYMBOL(d_splice_alias); |
1da177e4 | 1629 | |
9403540c BN |
1630 | /** |
1631 | * d_add_ci - lookup or allocate new dentry with case-exact name | |
1632 | * @inode: the inode case-insensitive lookup has found | |
1633 | * @dentry: the negative dentry that was passed to the parent's lookup func | |
1634 | * @name: the case-exact name to be associated with the returned dentry | |
1635 | * | |
1636 | * This is to avoid filling the dcache with case-insensitive names to the | |
1637 | * same inode, only the actual correct case is stored in the dcache for | |
1638 | * case-insensitive filesystems. | |
1639 | * | |
1640 | * For a case-insensitive lookup match and if the the case-exact dentry | |
1641 | * already exists in in the dcache, use it and return it. | |
1642 | * | |
1643 | * If no entry exists with the exact case name, allocate new dentry with | |
1644 | * the exact case, and return the spliced entry. | |
1645 | */ | |
e45b590b | 1646 | struct dentry *d_add_ci(struct dentry *dentry, struct inode *inode, |
9403540c BN |
1647 | struct qstr *name) |
1648 | { | |
1649 | int error; | |
1650 | struct dentry *found; | |
1651 | struct dentry *new; | |
1652 | ||
b6520c81 CH |
1653 | /* |
1654 | * First check if a dentry matching the name already exists, | |
1655 | * if not go ahead and create it now. | |
1656 | */ | |
9403540c | 1657 | found = d_hash_and_lookup(dentry->d_parent, name); |
9403540c BN |
1658 | if (!found) { |
1659 | new = d_alloc(dentry->d_parent, name); | |
1660 | if (!new) { | |
1661 | error = -ENOMEM; | |
1662 | goto err_out; | |
1663 | } | |
b6520c81 | 1664 | |
9403540c BN |
1665 | found = d_splice_alias(inode, new); |
1666 | if (found) { | |
1667 | dput(new); | |
1668 | return found; | |
1669 | } | |
1670 | return new; | |
1671 | } | |
b6520c81 CH |
1672 | |
1673 | /* | |
1674 | * If a matching dentry exists, and it's not negative use it. | |
1675 | * | |
1676 | * Decrement the reference count to balance the iget() done | |
1677 | * earlier on. | |
1678 | */ | |
9403540c BN |
1679 | if (found->d_inode) { |
1680 | if (unlikely(found->d_inode != inode)) { | |
1681 | /* This can't happen because bad inodes are unhashed. */ | |
1682 | BUG_ON(!is_bad_inode(inode)); | |
1683 | BUG_ON(!is_bad_inode(found->d_inode)); | |
1684 | } | |
9403540c BN |
1685 | iput(inode); |
1686 | return found; | |
1687 | } | |
b6520c81 | 1688 | |
9403540c BN |
1689 | /* |
1690 | * Negative dentry: instantiate it unless the inode is a directory and | |
b6520c81 | 1691 | * already has a dentry. |
9403540c | 1692 | */ |
873feea0 | 1693 | spin_lock(&inode->i_lock); |
b6520c81 | 1694 | if (!S_ISDIR(inode->i_mode) || list_empty(&inode->i_dentry)) { |
360da900 | 1695 | __d_instantiate(found, inode); |
873feea0 | 1696 | spin_unlock(&inode->i_lock); |
9403540c BN |
1697 | security_d_instantiate(found, inode); |
1698 | return found; | |
1699 | } | |
b6520c81 | 1700 | |
9403540c | 1701 | /* |
b6520c81 CH |
1702 | * In case a directory already has a (disconnected) entry grab a |
1703 | * reference to it, move it in place and use it. | |
9403540c BN |
1704 | */ |
1705 | new = list_entry(inode->i_dentry.next, struct dentry, d_alias); | |
dc0474be | 1706 | __dget(new); |
873feea0 | 1707 | spin_unlock(&inode->i_lock); |
9403540c | 1708 | security_d_instantiate(found, inode); |
9403540c | 1709 | d_move(new, found); |
9403540c | 1710 | iput(inode); |
9403540c | 1711 | dput(found); |
9403540c BN |
1712 | return new; |
1713 | ||
1714 | err_out: | |
1715 | iput(inode); | |
1716 | return ERR_PTR(error); | |
1717 | } | |
ec4f8605 | 1718 | EXPORT_SYMBOL(d_add_ci); |
1da177e4 | 1719 | |
31e6b01f NP |
1720 | /** |
1721 | * __d_lookup_rcu - search for a dentry (racy, store-free) | |
1722 | * @parent: parent dentry | |
1723 | * @name: qstr of name we wish to find | |
1724 | * @seq: returns d_seq value at the point where the dentry was found | |
1725 | * @inode: returns dentry->d_inode when the inode was found valid. | |
1726 | * Returns: dentry, or NULL | |
1727 | * | |
1728 | * __d_lookup_rcu is the dcache lookup function for rcu-walk name | |
1729 | * resolution (store-free path walking) design described in | |
1730 | * Documentation/filesystems/path-lookup.txt. | |
1731 | * | |
1732 | * This is not to be used outside core vfs. | |
1733 | * | |
1734 | * __d_lookup_rcu must only be used in rcu-walk mode, ie. with vfsmount lock | |
1735 | * held, and rcu_read_lock held. The returned dentry must not be stored into | |
1736 | * without taking d_lock and checking d_seq sequence count against @seq | |
1737 | * returned here. | |
1738 | * | |
1739 | * A refcount may be taken on the found dentry with the __d_rcu_to_refcount | |
1740 | * function. | |
1741 | * | |
1742 | * Alternatively, __d_lookup_rcu may be called again to look up the child of | |
1743 | * the returned dentry, so long as its parent's seqlock is checked after the | |
1744 | * child is looked up. Thus, an interlocking stepping of sequence lock checks | |
1745 | * is formed, giving integrity down the path walk. | |
1746 | */ | |
1747 | struct dentry *__d_lookup_rcu(struct dentry *parent, struct qstr *name, | |
1748 | unsigned *seq, struct inode **inode) | |
1749 | { | |
1750 | unsigned int len = name->len; | |
1751 | unsigned int hash = name->hash; | |
1752 | const unsigned char *str = name->name; | |
ceb5bdc2 NP |
1753 | struct dcache_hash_bucket *b = d_hash(parent, hash); |
1754 | struct hlist_bl_node *node; | |
31e6b01f NP |
1755 | struct dentry *dentry; |
1756 | ||
1757 | /* | |
1758 | * Note: There is significant duplication with __d_lookup_rcu which is | |
1759 | * required to prevent single threaded performance regressions | |
1760 | * especially on architectures where smp_rmb (in seqcounts) are costly. | |
1761 | * Keep the two functions in sync. | |
1762 | */ | |
1763 | ||
1764 | /* | |
1765 | * The hash list is protected using RCU. | |
1766 | * | |
1767 | * Carefully use d_seq when comparing a candidate dentry, to avoid | |
1768 | * races with d_move(). | |
1769 | * | |
1770 | * It is possible that concurrent renames can mess up our list | |
1771 | * walk here and result in missing our dentry, resulting in the | |
1772 | * false-negative result. d_lookup() protects against concurrent | |
1773 | * renames using rename_lock seqlock. | |
1774 | * | |
1775 | * See Documentation/vfs/dcache-locking.txt for more details. | |
1776 | */ | |
ceb5bdc2 | 1777 | hlist_bl_for_each_entry_rcu(dentry, node, &b->head, d_hash) { |
31e6b01f NP |
1778 | struct inode *i; |
1779 | const char *tname; | |
1780 | int tlen; | |
1781 | ||
1782 | if (dentry->d_name.hash != hash) | |
1783 | continue; | |
1784 | ||
1785 | seqretry: | |
1786 | *seq = read_seqcount_begin(&dentry->d_seq); | |
1787 | if (dentry->d_parent != parent) | |
1788 | continue; | |
1789 | if (d_unhashed(dentry)) | |
1790 | continue; | |
1791 | tlen = dentry->d_name.len; | |
1792 | tname = dentry->d_name.name; | |
1793 | i = dentry->d_inode; | |
e1bb5782 NP |
1794 | prefetch(tname); |
1795 | if (i) | |
1796 | prefetch(i); | |
31e6b01f NP |
1797 | /* |
1798 | * This seqcount check is required to ensure name and | |
1799 | * len are loaded atomically, so as not to walk off the | |
1800 | * edge of memory when walking. If we could load this | |
1801 | * atomically some other way, we could drop this check. | |
1802 | */ | |
1803 | if (read_seqcount_retry(&dentry->d_seq, *seq)) | |
1804 | goto seqretry; | |
fb045adb | 1805 | if (parent->d_flags & DCACHE_OP_COMPARE) { |
31e6b01f NP |
1806 | if (parent->d_op->d_compare(parent, *inode, |
1807 | dentry, i, | |
1808 | tlen, tname, name)) | |
1809 | continue; | |
1810 | } else { | |
9d55c369 | 1811 | if (dentry_cmp(tname, tlen, str, len)) |
31e6b01f NP |
1812 | continue; |
1813 | } | |
1814 | /* | |
1815 | * No extra seqcount check is required after the name | |
1816 | * compare. The caller must perform a seqcount check in | |
1817 | * order to do anything useful with the returned dentry | |
1818 | * anyway. | |
1819 | */ | |
1820 | *inode = i; | |
1821 | return dentry; | |
1822 | } | |
1823 | return NULL; | |
1824 | } | |
1825 | ||
1da177e4 LT |
1826 | /** |
1827 | * d_lookup - search for a dentry | |
1828 | * @parent: parent dentry | |
1829 | * @name: qstr of name we wish to find | |
b04f784e | 1830 | * Returns: dentry, or NULL |
1da177e4 | 1831 | * |
b04f784e NP |
1832 | * d_lookup searches the children of the parent dentry for the name in |
1833 | * question. If the dentry is found its reference count is incremented and the | |
1834 | * dentry is returned. The caller must use dput to free the entry when it has | |
1835 | * finished using it. %NULL is returned if the dentry does not exist. | |
1da177e4 | 1836 | */ |
31e6b01f | 1837 | struct dentry *d_lookup(struct dentry *parent, struct qstr *name) |
1da177e4 | 1838 | { |
31e6b01f | 1839 | struct dentry *dentry; |
949854d0 | 1840 | unsigned seq; |
1da177e4 LT |
1841 | |
1842 | do { | |
1843 | seq = read_seqbegin(&rename_lock); | |
1844 | dentry = __d_lookup(parent, name); | |
1845 | if (dentry) | |
1846 | break; | |
1847 | } while (read_seqretry(&rename_lock, seq)); | |
1848 | return dentry; | |
1849 | } | |
ec4f8605 | 1850 | EXPORT_SYMBOL(d_lookup); |
1da177e4 | 1851 | |
31e6b01f | 1852 | /** |
b04f784e NP |
1853 | * __d_lookup - search for a dentry (racy) |
1854 | * @parent: parent dentry | |
1855 | * @name: qstr of name we wish to find | |
1856 | * Returns: dentry, or NULL | |
1857 | * | |
1858 | * __d_lookup is like d_lookup, however it may (rarely) return a | |
1859 | * false-negative result due to unrelated rename activity. | |
1860 | * | |
1861 | * __d_lookup is slightly faster by avoiding rename_lock read seqlock, | |
1862 | * however it must be used carefully, eg. with a following d_lookup in | |
1863 | * the case of failure. | |
1864 | * | |
1865 | * __d_lookup callers must be commented. | |
1866 | */ | |
31e6b01f | 1867 | struct dentry *__d_lookup(struct dentry *parent, struct qstr *name) |
1da177e4 LT |
1868 | { |
1869 | unsigned int len = name->len; | |
1870 | unsigned int hash = name->hash; | |
1871 | const unsigned char *str = name->name; | |
ceb5bdc2 NP |
1872 | struct dcache_hash_bucket *b = d_hash(parent, hash); |
1873 | struct hlist_bl_node *node; | |
31e6b01f | 1874 | struct dentry *found = NULL; |
665a7583 | 1875 | struct dentry *dentry; |
1da177e4 | 1876 | |
31e6b01f NP |
1877 | /* |
1878 | * Note: There is significant duplication with __d_lookup_rcu which is | |
1879 | * required to prevent single threaded performance regressions | |
1880 | * especially on architectures where smp_rmb (in seqcounts) are costly. | |
1881 | * Keep the two functions in sync. | |
1882 | */ | |
1883 | ||
b04f784e NP |
1884 | /* |
1885 | * The hash list is protected using RCU. | |
1886 | * | |
1887 | * Take d_lock when comparing a candidate dentry, to avoid races | |
1888 | * with d_move(). | |
1889 | * | |
1890 | * It is possible that concurrent renames can mess up our list | |
1891 | * walk here and result in missing our dentry, resulting in the | |
1892 | * false-negative result. d_lookup() protects against concurrent | |
1893 | * renames using rename_lock seqlock. | |
1894 | * | |
1895 | * See Documentation/vfs/dcache-locking.txt for more details. | |
1896 | */ | |
1da177e4 LT |
1897 | rcu_read_lock(); |
1898 | ||
ceb5bdc2 | 1899 | hlist_bl_for_each_entry_rcu(dentry, node, &b->head, d_hash) { |
31e6b01f NP |
1900 | const char *tname; |
1901 | int tlen; | |
1da177e4 | 1902 | |
1da177e4 LT |
1903 | if (dentry->d_name.hash != hash) |
1904 | continue; | |
1da177e4 LT |
1905 | |
1906 | spin_lock(&dentry->d_lock); | |
1da177e4 LT |
1907 | if (dentry->d_parent != parent) |
1908 | goto next; | |
d0185c08 LT |
1909 | if (d_unhashed(dentry)) |
1910 | goto next; | |
1911 | ||
1da177e4 LT |
1912 | /* |
1913 | * It is safe to compare names since d_move() cannot | |
1914 | * change the qstr (protected by d_lock). | |
1915 | */ | |
31e6b01f NP |
1916 | tlen = dentry->d_name.len; |
1917 | tname = dentry->d_name.name; | |
fb045adb | 1918 | if (parent->d_flags & DCACHE_OP_COMPARE) { |
621e155a NP |
1919 | if (parent->d_op->d_compare(parent, parent->d_inode, |
1920 | dentry, dentry->d_inode, | |
31e6b01f | 1921 | tlen, tname, name)) |
1da177e4 LT |
1922 | goto next; |
1923 | } else { | |
9d55c369 | 1924 | if (dentry_cmp(tname, tlen, str, len)) |
1da177e4 LT |
1925 | goto next; |
1926 | } | |
1927 | ||
b7ab39f6 | 1928 | dentry->d_count++; |
d0185c08 | 1929 | found = dentry; |
1da177e4 LT |
1930 | spin_unlock(&dentry->d_lock); |
1931 | break; | |
1932 | next: | |
1933 | spin_unlock(&dentry->d_lock); | |
1934 | } | |
1935 | rcu_read_unlock(); | |
1936 | ||
1937 | return found; | |
1938 | } | |
1939 | ||
3e7e241f EB |
1940 | /** |
1941 | * d_hash_and_lookup - hash the qstr then search for a dentry | |
1942 | * @dir: Directory to search in | |
1943 | * @name: qstr of name we wish to find | |
1944 | * | |
1945 | * On hash failure or on lookup failure NULL is returned. | |
1946 | */ | |
1947 | struct dentry *d_hash_and_lookup(struct dentry *dir, struct qstr *name) | |
1948 | { | |
1949 | struct dentry *dentry = NULL; | |
1950 | ||
1951 | /* | |
1952 | * Check for a fs-specific hash function. Note that we must | |
1953 | * calculate the standard hash first, as the d_op->d_hash() | |
1954 | * routine may choose to leave the hash value unchanged. | |
1955 | */ | |
1956 | name->hash = full_name_hash(name->name, name->len); | |
fb045adb | 1957 | if (dir->d_flags & DCACHE_OP_HASH) { |
b1e6a015 | 1958 | if (dir->d_op->d_hash(dir, dir->d_inode, name) < 0) |
3e7e241f EB |
1959 | goto out; |
1960 | } | |
1961 | dentry = d_lookup(dir, name); | |
1962 | out: | |
1963 | return dentry; | |
1964 | } | |
1965 | ||
1da177e4 | 1966 | /** |
786a5e15 | 1967 | * d_validate - verify dentry provided from insecure source (deprecated) |
1da177e4 LT |
1968 | * @dentry: The dentry alleged to be valid child of @dparent |
1969 | * @dparent: The parent dentry (known to be valid) | |
1da177e4 LT |
1970 | * |
1971 | * An insecure source has sent us a dentry, here we verify it and dget() it. | |
1972 | * This is used by ncpfs in its readdir implementation. | |
1973 | * Zero is returned in the dentry is invalid. | |
786a5e15 NP |
1974 | * |
1975 | * This function is slow for big directories, and deprecated, do not use it. | |
1da177e4 | 1976 | */ |
d3a23e16 | 1977 | int d_validate(struct dentry *dentry, struct dentry *dparent) |
1da177e4 | 1978 | { |
786a5e15 | 1979 | struct dentry *child; |
d3a23e16 | 1980 | |
2fd6b7f5 | 1981 | spin_lock(&dparent->d_lock); |
786a5e15 NP |
1982 | list_for_each_entry(child, &dparent->d_subdirs, d_u.d_child) { |
1983 | if (dentry == child) { | |
2fd6b7f5 | 1984 | spin_lock_nested(&dentry->d_lock, DENTRY_D_LOCK_NESTED); |
dc0474be | 1985 | __dget_dlock(dentry); |
2fd6b7f5 NP |
1986 | spin_unlock(&dentry->d_lock); |
1987 | spin_unlock(&dparent->d_lock); | |
1da177e4 LT |
1988 | return 1; |
1989 | } | |
1990 | } | |
2fd6b7f5 | 1991 | spin_unlock(&dparent->d_lock); |
786a5e15 | 1992 | |
1da177e4 LT |
1993 | return 0; |
1994 | } | |
ec4f8605 | 1995 | EXPORT_SYMBOL(d_validate); |
1da177e4 LT |
1996 | |
1997 | /* | |
1998 | * When a file is deleted, we have two options: | |
1999 | * - turn this dentry into a negative dentry | |
2000 | * - unhash this dentry and free it. | |
2001 | * | |
2002 | * Usually, we want to just turn this into | |
2003 | * a negative dentry, but if anybody else is | |
2004 | * currently using the dentry or the inode | |
2005 | * we can't do that and we fall back on removing | |
2006 | * it from the hash queues and waiting for | |
2007 | * it to be deleted later when it has no users | |
2008 | */ | |
2009 | ||
2010 | /** | |
2011 | * d_delete - delete a dentry | |
2012 | * @dentry: The dentry to delete | |
2013 | * | |
2014 | * Turn the dentry into a negative dentry if possible, otherwise | |
2015 | * remove it from the hash queues so it can be deleted later | |
2016 | */ | |
2017 | ||
2018 | void d_delete(struct dentry * dentry) | |
2019 | { | |
873feea0 | 2020 | struct inode *inode; |
7a91bf7f | 2021 | int isdir = 0; |
1da177e4 LT |
2022 | /* |
2023 | * Are we the only user? | |
2024 | */ | |
357f8e65 | 2025 | again: |
1da177e4 | 2026 | spin_lock(&dentry->d_lock); |
873feea0 NP |
2027 | inode = dentry->d_inode; |
2028 | isdir = S_ISDIR(inode->i_mode); | |
b7ab39f6 | 2029 | if (dentry->d_count == 1) { |
873feea0 | 2030 | if (inode && !spin_trylock(&inode->i_lock)) { |
357f8e65 NP |
2031 | spin_unlock(&dentry->d_lock); |
2032 | cpu_relax(); | |
2033 | goto again; | |
2034 | } | |
13e3c5e5 | 2035 | dentry->d_flags &= ~DCACHE_CANT_MOUNT; |
31e6b01f | 2036 | dentry_unlink_inode(dentry); |
7a91bf7f | 2037 | fsnotify_nameremove(dentry, isdir); |
1da177e4 LT |
2038 | return; |
2039 | } | |
2040 | ||
2041 | if (!d_unhashed(dentry)) | |
2042 | __d_drop(dentry); | |
2043 | ||
2044 | spin_unlock(&dentry->d_lock); | |
7a91bf7f JM |
2045 | |
2046 | fsnotify_nameremove(dentry, isdir); | |
1da177e4 | 2047 | } |
ec4f8605 | 2048 | EXPORT_SYMBOL(d_delete); |
1da177e4 | 2049 | |
ceb5bdc2 | 2050 | static void __d_rehash(struct dentry * entry, struct dcache_hash_bucket *b) |
1da177e4 | 2051 | { |
ceb5bdc2 NP |
2052 | BUG_ON(!d_unhashed(entry)); |
2053 | spin_lock_bucket(b); | |
1da177e4 | 2054 | entry->d_flags &= ~DCACHE_UNHASHED; |
ceb5bdc2 NP |
2055 | hlist_bl_add_head_rcu(&entry->d_hash, &b->head); |
2056 | spin_unlock_bucket(b); | |
1da177e4 LT |
2057 | } |
2058 | ||
770bfad8 DH |
2059 | static void _d_rehash(struct dentry * entry) |
2060 | { | |
2061 | __d_rehash(entry, d_hash(entry->d_parent, entry->d_name.hash)); | |
2062 | } | |
2063 | ||
1da177e4 LT |
2064 | /** |
2065 | * d_rehash - add an entry back to the hash | |
2066 | * @entry: dentry to add to the hash | |
2067 | * | |
2068 | * Adds a dentry to the hash according to its name. | |
2069 | */ | |
2070 | ||
2071 | void d_rehash(struct dentry * entry) | |
2072 | { | |
1da177e4 | 2073 | spin_lock(&entry->d_lock); |
770bfad8 | 2074 | _d_rehash(entry); |
1da177e4 | 2075 | spin_unlock(&entry->d_lock); |
1da177e4 | 2076 | } |
ec4f8605 | 2077 | EXPORT_SYMBOL(d_rehash); |
1da177e4 | 2078 | |
fb2d5b86 NP |
2079 | /** |
2080 | * dentry_update_name_case - update case insensitive dentry with a new name | |
2081 | * @dentry: dentry to be updated | |
2082 | * @name: new name | |
2083 | * | |
2084 | * Update a case insensitive dentry with new case of name. | |
2085 | * | |
2086 | * dentry must have been returned by d_lookup with name @name. Old and new | |
2087 | * name lengths must match (ie. no d_compare which allows mismatched name | |
2088 | * lengths). | |
2089 | * | |
2090 | * Parent inode i_mutex must be held over d_lookup and into this call (to | |
2091 | * keep renames and concurrent inserts, and readdir(2) away). | |
2092 | */ | |
2093 | void dentry_update_name_case(struct dentry *dentry, struct qstr *name) | |
2094 | { | |
2095 | BUG_ON(!mutex_is_locked(&dentry->d_inode->i_mutex)); | |
2096 | BUG_ON(dentry->d_name.len != name->len); /* d_lookup gives this */ | |
2097 | ||
fb2d5b86 | 2098 | spin_lock(&dentry->d_lock); |
31e6b01f | 2099 | write_seqcount_begin(&dentry->d_seq); |
fb2d5b86 | 2100 | memcpy((unsigned char *)dentry->d_name.name, name->name, name->len); |
31e6b01f | 2101 | write_seqcount_end(&dentry->d_seq); |
fb2d5b86 | 2102 | spin_unlock(&dentry->d_lock); |
fb2d5b86 NP |
2103 | } |
2104 | EXPORT_SYMBOL(dentry_update_name_case); | |
2105 | ||
1da177e4 LT |
2106 | static void switch_names(struct dentry *dentry, struct dentry *target) |
2107 | { | |
2108 | if (dname_external(target)) { | |
2109 | if (dname_external(dentry)) { | |
2110 | /* | |
2111 | * Both external: swap the pointers | |
2112 | */ | |
9a8d5bb4 | 2113 | swap(target->d_name.name, dentry->d_name.name); |
1da177e4 LT |
2114 | } else { |
2115 | /* | |
2116 | * dentry:internal, target:external. Steal target's | |
2117 | * storage and make target internal. | |
2118 | */ | |
321bcf92 BF |
2119 | memcpy(target->d_iname, dentry->d_name.name, |
2120 | dentry->d_name.len + 1); | |
1da177e4 LT |
2121 | dentry->d_name.name = target->d_name.name; |
2122 | target->d_name.name = target->d_iname; | |
2123 | } | |
2124 | } else { | |
2125 | if (dname_external(dentry)) { | |
2126 | /* | |
2127 | * dentry:external, target:internal. Give dentry's | |
2128 | * storage to target and make dentry internal | |
2129 | */ | |
2130 | memcpy(dentry->d_iname, target->d_name.name, | |
2131 | target->d_name.len + 1); | |
2132 | target->d_name.name = dentry->d_name.name; | |
2133 | dentry->d_name.name = dentry->d_iname; | |
2134 | } else { | |
2135 | /* | |
2136 | * Both are internal. Just copy target to dentry | |
2137 | */ | |
2138 | memcpy(dentry->d_iname, target->d_name.name, | |
2139 | target->d_name.len + 1); | |
dc711ca3 AV |
2140 | dentry->d_name.len = target->d_name.len; |
2141 | return; | |
1da177e4 LT |
2142 | } |
2143 | } | |
9a8d5bb4 | 2144 | swap(dentry->d_name.len, target->d_name.len); |
1da177e4 LT |
2145 | } |
2146 | ||
2fd6b7f5 NP |
2147 | static void dentry_lock_for_move(struct dentry *dentry, struct dentry *target) |
2148 | { | |
2149 | /* | |
2150 | * XXXX: do we really need to take target->d_lock? | |
2151 | */ | |
2152 | if (IS_ROOT(dentry) || dentry->d_parent == target->d_parent) | |
2153 | spin_lock(&target->d_parent->d_lock); | |
2154 | else { | |
2155 | if (d_ancestor(dentry->d_parent, target->d_parent)) { | |
2156 | spin_lock(&dentry->d_parent->d_lock); | |
2157 | spin_lock_nested(&target->d_parent->d_lock, | |
2158 | DENTRY_D_LOCK_NESTED); | |
2159 | } else { | |
2160 | spin_lock(&target->d_parent->d_lock); | |
2161 | spin_lock_nested(&dentry->d_parent->d_lock, | |
2162 | DENTRY_D_LOCK_NESTED); | |
2163 | } | |
2164 | } | |
2165 | if (target < dentry) { | |
2166 | spin_lock_nested(&target->d_lock, 2); | |
2167 | spin_lock_nested(&dentry->d_lock, 3); | |
2168 | } else { | |
2169 | spin_lock_nested(&dentry->d_lock, 2); | |
2170 | spin_lock_nested(&target->d_lock, 3); | |
2171 | } | |
2172 | } | |
2173 | ||
2174 | static void dentry_unlock_parents_for_move(struct dentry *dentry, | |
2175 | struct dentry *target) | |
2176 | { | |
2177 | if (target->d_parent != dentry->d_parent) | |
2178 | spin_unlock(&dentry->d_parent->d_lock); | |
2179 | if (target->d_parent != target) | |
2180 | spin_unlock(&target->d_parent->d_lock); | |
2181 | } | |
2182 | ||
1da177e4 | 2183 | /* |
2fd6b7f5 NP |
2184 | * When switching names, the actual string doesn't strictly have to |
2185 | * be preserved in the target - because we're dropping the target | |
2186 | * anyway. As such, we can just do a simple memcpy() to copy over | |
2187 | * the new name before we switch. | |
2188 | * | |
2189 | * Note that we have to be a lot more careful about getting the hash | |
2190 | * switched - we have to switch the hash value properly even if it | |
2191 | * then no longer matches the actual (corrupted) string of the target. | |
2192 | * The hash value has to match the hash queue that the dentry is on.. | |
1da177e4 | 2193 | */ |
9eaef27b | 2194 | /* |
b5c84bf6 | 2195 | * d_move - move a dentry |
1da177e4 LT |
2196 | * @dentry: entry to move |
2197 | * @target: new dentry | |
2198 | * | |
2199 | * Update the dcache to reflect the move of a file name. Negative | |
2200 | * dcache entries should not be moved in this way. | |
2201 | */ | |
b5c84bf6 | 2202 | void d_move(struct dentry * dentry, struct dentry * target) |
1da177e4 | 2203 | { |
1da177e4 LT |
2204 | if (!dentry->d_inode) |
2205 | printk(KERN_WARNING "VFS: moving negative dcache entry\n"); | |
2206 | ||
2fd6b7f5 NP |
2207 | BUG_ON(d_ancestor(dentry, target)); |
2208 | BUG_ON(d_ancestor(target, dentry)); | |
2209 | ||
1da177e4 | 2210 | write_seqlock(&rename_lock); |
2fd6b7f5 NP |
2211 | |
2212 | dentry_lock_for_move(dentry, target); | |
1da177e4 | 2213 | |
31e6b01f NP |
2214 | write_seqcount_begin(&dentry->d_seq); |
2215 | write_seqcount_begin(&target->d_seq); | |
2216 | ||
ceb5bdc2 NP |
2217 | /* __d_drop does write_seqcount_barrier, but they're OK to nest. */ |
2218 | ||
2219 | /* | |
2220 | * Move the dentry to the target hash queue. Don't bother checking | |
2221 | * for the same hash queue because of how unlikely it is. | |
2222 | */ | |
2223 | __d_drop(dentry); | |
789680d1 | 2224 | __d_rehash(dentry, d_hash(target->d_parent, target->d_name.hash)); |
1da177e4 LT |
2225 | |
2226 | /* Unhash the target: dput() will then get rid of it */ | |
2227 | __d_drop(target); | |
2228 | ||
5160ee6f ED |
2229 | list_del(&dentry->d_u.d_child); |
2230 | list_del(&target->d_u.d_child); | |
1da177e4 LT |
2231 | |
2232 | /* Switch the names.. */ | |
2233 | switch_names(dentry, target); | |
9a8d5bb4 | 2234 | swap(dentry->d_name.hash, target->d_name.hash); |
1da177e4 LT |
2235 | |
2236 | /* ... and switch the parents */ | |
2237 | if (IS_ROOT(dentry)) { | |
2238 | dentry->d_parent = target->d_parent; | |
2239 | target->d_parent = target; | |
5160ee6f | 2240 | INIT_LIST_HEAD(&target->d_u.d_child); |
1da177e4 | 2241 | } else { |
9a8d5bb4 | 2242 | swap(dentry->d_parent, target->d_parent); |
1da177e4 LT |
2243 | |
2244 | /* And add them back to the (new) parent lists */ | |
5160ee6f | 2245 | list_add(&target->d_u.d_child, &target->d_parent->d_subdirs); |
1da177e4 LT |
2246 | } |
2247 | ||
5160ee6f | 2248 | list_add(&dentry->d_u.d_child, &dentry->d_parent->d_subdirs); |
2fd6b7f5 | 2249 | |
31e6b01f NP |
2250 | write_seqcount_end(&target->d_seq); |
2251 | write_seqcount_end(&dentry->d_seq); | |
2252 | ||
2fd6b7f5 | 2253 | dentry_unlock_parents_for_move(dentry, target); |
1da177e4 | 2254 | spin_unlock(&target->d_lock); |
c32ccd87 | 2255 | fsnotify_d_move(dentry); |
1da177e4 LT |
2256 | spin_unlock(&dentry->d_lock); |
2257 | write_sequnlock(&rename_lock); | |
9eaef27b | 2258 | } |
ec4f8605 | 2259 | EXPORT_SYMBOL(d_move); |
1da177e4 | 2260 | |
e2761a11 OH |
2261 | /** |
2262 | * d_ancestor - search for an ancestor | |
2263 | * @p1: ancestor dentry | |
2264 | * @p2: child dentry | |
2265 | * | |
2266 | * Returns the ancestor dentry of p2 which is a child of p1, if p1 is | |
2267 | * an ancestor of p2, else NULL. | |
9eaef27b | 2268 | */ |
e2761a11 | 2269 | struct dentry *d_ancestor(struct dentry *p1, struct dentry *p2) |
9eaef27b TM |
2270 | { |
2271 | struct dentry *p; | |
2272 | ||
871c0067 | 2273 | for (p = p2; !IS_ROOT(p); p = p->d_parent) { |
9eaef27b | 2274 | if (p->d_parent == p1) |
e2761a11 | 2275 | return p; |
9eaef27b | 2276 | } |
e2761a11 | 2277 | return NULL; |
9eaef27b TM |
2278 | } |
2279 | ||
2280 | /* | |
2281 | * This helper attempts to cope with remotely renamed directories | |
2282 | * | |
2283 | * It assumes that the caller is already holding | |
873feea0 | 2284 | * dentry->d_parent->d_inode->i_mutex and the inode->i_lock |
9eaef27b TM |
2285 | * |
2286 | * Note: If ever the locking in lock_rename() changes, then please | |
2287 | * remember to update this too... | |
9eaef27b | 2288 | */ |
873feea0 NP |
2289 | static struct dentry *__d_unalias(struct inode *inode, |
2290 | struct dentry *dentry, struct dentry *alias) | |
9eaef27b TM |
2291 | { |
2292 | struct mutex *m1 = NULL, *m2 = NULL; | |
2293 | struct dentry *ret; | |
2294 | ||
2295 | /* If alias and dentry share a parent, then no extra locks required */ | |
2296 | if (alias->d_parent == dentry->d_parent) | |
2297 | goto out_unalias; | |
2298 | ||
2299 | /* Check for loops */ | |
2300 | ret = ERR_PTR(-ELOOP); | |
e2761a11 | 2301 | if (d_ancestor(alias, dentry)) |
9eaef27b TM |
2302 | goto out_err; |
2303 | ||
2304 | /* See lock_rename() */ | |
2305 | ret = ERR_PTR(-EBUSY); | |
2306 | if (!mutex_trylock(&dentry->d_sb->s_vfs_rename_mutex)) | |
2307 | goto out_err; | |
2308 | m1 = &dentry->d_sb->s_vfs_rename_mutex; | |
2309 | if (!mutex_trylock(&alias->d_parent->d_inode->i_mutex)) | |
2310 | goto out_err; | |
2311 | m2 = &alias->d_parent->d_inode->i_mutex; | |
2312 | out_unalias: | |
b5c84bf6 | 2313 | d_move(alias, dentry); |
9eaef27b TM |
2314 | ret = alias; |
2315 | out_err: | |
873feea0 | 2316 | spin_unlock(&inode->i_lock); |
9eaef27b TM |
2317 | if (m2) |
2318 | mutex_unlock(m2); | |
2319 | if (m1) | |
2320 | mutex_unlock(m1); | |
2321 | return ret; | |
2322 | } | |
2323 | ||
770bfad8 DH |
2324 | /* |
2325 | * Prepare an anonymous dentry for life in the superblock's dentry tree as a | |
2326 | * named dentry in place of the dentry to be replaced. | |
2fd6b7f5 | 2327 | * returns with anon->d_lock held! |
770bfad8 DH |
2328 | */ |
2329 | static void __d_materialise_dentry(struct dentry *dentry, struct dentry *anon) | |
2330 | { | |
2331 | struct dentry *dparent, *aparent; | |
2332 | ||
2fd6b7f5 | 2333 | dentry_lock_for_move(anon, dentry); |
770bfad8 | 2334 | |
31e6b01f NP |
2335 | write_seqcount_begin(&dentry->d_seq); |
2336 | write_seqcount_begin(&anon->d_seq); | |
2337 | ||
770bfad8 DH |
2338 | dparent = dentry->d_parent; |
2339 | aparent = anon->d_parent; | |
2340 | ||
2fd6b7f5 NP |
2341 | switch_names(dentry, anon); |
2342 | swap(dentry->d_name.hash, anon->d_name.hash); | |
2343 | ||
770bfad8 DH |
2344 | dentry->d_parent = (aparent == anon) ? dentry : aparent; |
2345 | list_del(&dentry->d_u.d_child); | |
2346 | if (!IS_ROOT(dentry)) | |
2347 | list_add(&dentry->d_u.d_child, &dentry->d_parent->d_subdirs); | |
2348 | else | |
2349 | INIT_LIST_HEAD(&dentry->d_u.d_child); | |
2350 | ||
2351 | anon->d_parent = (dparent == dentry) ? anon : dparent; | |
2352 | list_del(&anon->d_u.d_child); | |
2353 | if (!IS_ROOT(anon)) | |
2354 | list_add(&anon->d_u.d_child, &anon->d_parent->d_subdirs); | |
2355 | else | |
2356 | INIT_LIST_HEAD(&anon->d_u.d_child); | |
2357 | ||
31e6b01f NP |
2358 | write_seqcount_end(&dentry->d_seq); |
2359 | write_seqcount_end(&anon->d_seq); | |
2360 | ||
2fd6b7f5 NP |
2361 | dentry_unlock_parents_for_move(anon, dentry); |
2362 | spin_unlock(&dentry->d_lock); | |
2363 | ||
2364 | /* anon->d_lock still locked, returns locked */ | |
770bfad8 DH |
2365 | anon->d_flags &= ~DCACHE_DISCONNECTED; |
2366 | } | |
2367 | ||
2368 | /** | |
2369 | * d_materialise_unique - introduce an inode into the tree | |
2370 | * @dentry: candidate dentry | |
2371 | * @inode: inode to bind to the dentry, to which aliases may be attached | |
2372 | * | |
2373 | * Introduces an dentry into the tree, substituting an extant disconnected | |
2374 | * root directory alias in its place if there is one | |
2375 | */ | |
2376 | struct dentry *d_materialise_unique(struct dentry *dentry, struct inode *inode) | |
2377 | { | |
9eaef27b | 2378 | struct dentry *actual; |
770bfad8 DH |
2379 | |
2380 | BUG_ON(!d_unhashed(dentry)); | |
2381 | ||
770bfad8 DH |
2382 | if (!inode) { |
2383 | actual = dentry; | |
360da900 | 2384 | __d_instantiate(dentry, NULL); |
357f8e65 NP |
2385 | d_rehash(actual); |
2386 | goto out_nolock; | |
770bfad8 DH |
2387 | } |
2388 | ||
873feea0 | 2389 | spin_lock(&inode->i_lock); |
357f8e65 | 2390 | |
9eaef27b TM |
2391 | if (S_ISDIR(inode->i_mode)) { |
2392 | struct dentry *alias; | |
2393 | ||
2394 | /* Does an aliased dentry already exist? */ | |
2395 | alias = __d_find_alias(inode, 0); | |
2396 | if (alias) { | |
2397 | actual = alias; | |
2398 | /* Is this an anonymous mountpoint that we could splice | |
2399 | * into our tree? */ | |
2400 | if (IS_ROOT(alias)) { | |
9eaef27b TM |
2401 | __d_materialise_dentry(dentry, alias); |
2402 | __d_drop(alias); | |
2403 | goto found; | |
2404 | } | |
2405 | /* Nope, but we must(!) avoid directory aliasing */ | |
873feea0 | 2406 | actual = __d_unalias(inode, dentry, alias); |
9eaef27b TM |
2407 | if (IS_ERR(actual)) |
2408 | dput(alias); | |
2409 | goto out_nolock; | |
2410 | } | |
770bfad8 DH |
2411 | } |
2412 | ||
2413 | /* Add a unique reference */ | |
2414 | actual = __d_instantiate_unique(dentry, inode); | |
2415 | if (!actual) | |
2416 | actual = dentry; | |
357f8e65 NP |
2417 | else |
2418 | BUG_ON(!d_unhashed(actual)); | |
770bfad8 | 2419 | |
770bfad8 DH |
2420 | spin_lock(&actual->d_lock); |
2421 | found: | |
2422 | _d_rehash(actual); | |
2423 | spin_unlock(&actual->d_lock); | |
873feea0 | 2424 | spin_unlock(&inode->i_lock); |
9eaef27b | 2425 | out_nolock: |
770bfad8 DH |
2426 | if (actual == dentry) { |
2427 | security_d_instantiate(dentry, inode); | |
2428 | return NULL; | |
2429 | } | |
2430 | ||
2431 | iput(inode); | |
2432 | return actual; | |
770bfad8 | 2433 | } |
ec4f8605 | 2434 | EXPORT_SYMBOL_GPL(d_materialise_unique); |
770bfad8 | 2435 | |
cdd16d02 | 2436 | static int prepend(char **buffer, int *buflen, const char *str, int namelen) |
6092d048 RP |
2437 | { |
2438 | *buflen -= namelen; | |
2439 | if (*buflen < 0) | |
2440 | return -ENAMETOOLONG; | |
2441 | *buffer -= namelen; | |
2442 | memcpy(*buffer, str, namelen); | |
2443 | return 0; | |
2444 | } | |
2445 | ||
cdd16d02 MS |
2446 | static int prepend_name(char **buffer, int *buflen, struct qstr *name) |
2447 | { | |
2448 | return prepend(buffer, buflen, name->name, name->len); | |
2449 | } | |
2450 | ||
1da177e4 | 2451 | /** |
f2eb6575 MS |
2452 | * Prepend path string to a buffer |
2453 | * | |
9d1bc601 MS |
2454 | * @path: the dentry/vfsmount to report |
2455 | * @root: root vfsmnt/dentry (may be modified by this function) | |
f2eb6575 MS |
2456 | * @buffer: pointer to the end of the buffer |
2457 | * @buflen: pointer to buffer length | |
552ce544 | 2458 | * |
949854d0 | 2459 | * Caller holds the rename_lock. |
9d1bc601 MS |
2460 | * |
2461 | * If path is not reachable from the supplied root, then the value of | |
2462 | * root is changed (without modifying refcounts). | |
1da177e4 | 2463 | */ |
f2eb6575 MS |
2464 | static int prepend_path(const struct path *path, struct path *root, |
2465 | char **buffer, int *buflen) | |
1da177e4 | 2466 | { |
9d1bc601 MS |
2467 | struct dentry *dentry = path->dentry; |
2468 | struct vfsmount *vfsmnt = path->mnt; | |
f2eb6575 MS |
2469 | bool slash = false; |
2470 | int error = 0; | |
6092d048 | 2471 | |
99b7db7b | 2472 | br_read_lock(vfsmount_lock); |
f2eb6575 | 2473 | while (dentry != root->dentry || vfsmnt != root->mnt) { |
1da177e4 LT |
2474 | struct dentry * parent; |
2475 | ||
1da177e4 | 2476 | if (dentry == vfsmnt->mnt_root || IS_ROOT(dentry)) { |
552ce544 | 2477 | /* Global root? */ |
1da177e4 | 2478 | if (vfsmnt->mnt_parent == vfsmnt) { |
1da177e4 LT |
2479 | goto global_root; |
2480 | } | |
2481 | dentry = vfsmnt->mnt_mountpoint; | |
2482 | vfsmnt = vfsmnt->mnt_parent; | |
1da177e4 LT |
2483 | continue; |
2484 | } | |
2485 | parent = dentry->d_parent; | |
2486 | prefetch(parent); | |
9abca360 | 2487 | spin_lock(&dentry->d_lock); |
f2eb6575 | 2488 | error = prepend_name(buffer, buflen, &dentry->d_name); |
9abca360 | 2489 | spin_unlock(&dentry->d_lock); |
f2eb6575 MS |
2490 | if (!error) |
2491 | error = prepend(buffer, buflen, "/", 1); | |
2492 | if (error) | |
2493 | break; | |
2494 | ||
2495 | slash = true; | |
1da177e4 LT |
2496 | dentry = parent; |
2497 | } | |
2498 | ||
be285c71 | 2499 | out: |
f2eb6575 MS |
2500 | if (!error && !slash) |
2501 | error = prepend(buffer, buflen, "/", 1); | |
2502 | ||
99b7db7b | 2503 | br_read_unlock(vfsmount_lock); |
f2eb6575 | 2504 | return error; |
1da177e4 LT |
2505 | |
2506 | global_root: | |
98dc568b MS |
2507 | /* |
2508 | * Filesystems needing to implement special "root names" | |
2509 | * should do so with ->d_dname() | |
2510 | */ | |
2511 | if (IS_ROOT(dentry) && | |
2512 | (dentry->d_name.len != 1 || dentry->d_name.name[0] != '/')) { | |
2513 | WARN(1, "Root dentry has weird name <%.*s>\n", | |
2514 | (int) dentry->d_name.len, dentry->d_name.name); | |
2515 | } | |
9d1bc601 MS |
2516 | root->mnt = vfsmnt; |
2517 | root->dentry = dentry; | |
be285c71 | 2518 | goto out; |
f2eb6575 | 2519 | } |
be285c71 | 2520 | |
f2eb6575 MS |
2521 | /** |
2522 | * __d_path - return the path of a dentry | |
2523 | * @path: the dentry/vfsmount to report | |
2524 | * @root: root vfsmnt/dentry (may be modified by this function) | |
cd956a1c | 2525 | * @buf: buffer to return value in |
f2eb6575 MS |
2526 | * @buflen: buffer length |
2527 | * | |
ffd1f4ed | 2528 | * Convert a dentry into an ASCII path name. |
f2eb6575 MS |
2529 | * |
2530 | * Returns a pointer into the buffer or an error code if the | |
2531 | * path was too long. | |
2532 | * | |
be148247 | 2533 | * "buflen" should be positive. |
f2eb6575 MS |
2534 | * |
2535 | * If path is not reachable from the supplied root, then the value of | |
2536 | * root is changed (without modifying refcounts). | |
2537 | */ | |
2538 | char *__d_path(const struct path *path, struct path *root, | |
2539 | char *buf, int buflen) | |
2540 | { | |
2541 | char *res = buf + buflen; | |
2542 | int error; | |
2543 | ||
2544 | prepend(&res, &buflen, "\0", 1); | |
949854d0 | 2545 | write_seqlock(&rename_lock); |
f2eb6575 | 2546 | error = prepend_path(path, root, &res, &buflen); |
949854d0 | 2547 | write_sequnlock(&rename_lock); |
be148247 | 2548 | |
f2eb6575 MS |
2549 | if (error) |
2550 | return ERR_PTR(error); | |
f2eb6575 | 2551 | return res; |
1da177e4 LT |
2552 | } |
2553 | ||
ffd1f4ed MS |
2554 | /* |
2555 | * same as __d_path but appends "(deleted)" for unlinked files. | |
2556 | */ | |
2557 | static int path_with_deleted(const struct path *path, struct path *root, | |
2558 | char **buf, int *buflen) | |
2559 | { | |
2560 | prepend(buf, buflen, "\0", 1); | |
2561 | if (d_unlinked(path->dentry)) { | |
2562 | int error = prepend(buf, buflen, " (deleted)", 10); | |
2563 | if (error) | |
2564 | return error; | |
2565 | } | |
2566 | ||
2567 | return prepend_path(path, root, buf, buflen); | |
2568 | } | |
2569 | ||
8df9d1a4 MS |
2570 | static int prepend_unreachable(char **buffer, int *buflen) |
2571 | { | |
2572 | return prepend(buffer, buflen, "(unreachable)", 13); | |
2573 | } | |
2574 | ||
a03a8a70 JB |
2575 | /** |
2576 | * d_path - return the path of a dentry | |
cf28b486 | 2577 | * @path: path to report |
a03a8a70 JB |
2578 | * @buf: buffer to return value in |
2579 | * @buflen: buffer length | |
2580 | * | |
2581 | * Convert a dentry into an ASCII path name. If the entry has been deleted | |
2582 | * the string " (deleted)" is appended. Note that this is ambiguous. | |
2583 | * | |
52afeefb AV |
2584 | * Returns a pointer into the buffer or an error code if the path was |
2585 | * too long. Note: Callers should use the returned pointer, not the passed | |
2586 | * in buffer, to use the name! The implementation often starts at an offset | |
2587 | * into the buffer, and may leave 0 bytes at the start. | |
a03a8a70 | 2588 | * |
31f3e0b3 | 2589 | * "buflen" should be positive. |
a03a8a70 | 2590 | */ |
20d4fdc1 | 2591 | char *d_path(const struct path *path, char *buf, int buflen) |
1da177e4 | 2592 | { |
ffd1f4ed | 2593 | char *res = buf + buflen; |
6ac08c39 | 2594 | struct path root; |
9d1bc601 | 2595 | struct path tmp; |
ffd1f4ed | 2596 | int error; |
1da177e4 | 2597 | |
c23fbb6b ED |
2598 | /* |
2599 | * We have various synthetic filesystems that never get mounted. On | |
2600 | * these filesystems dentries are never used for lookup purposes, and | |
2601 | * thus don't need to be hashed. They also don't need a name until a | |
2602 | * user wants to identify the object in /proc/pid/fd/. The little hack | |
2603 | * below allows us to generate a name for these objects on demand: | |
2604 | */ | |
cf28b486 JB |
2605 | if (path->dentry->d_op && path->dentry->d_op->d_dname) |
2606 | return path->dentry->d_op->d_dname(path->dentry, buf, buflen); | |
c23fbb6b | 2607 | |
f7ad3c6b | 2608 | get_fs_root(current->fs, &root); |
949854d0 | 2609 | write_seqlock(&rename_lock); |
9d1bc601 | 2610 | tmp = root; |
ffd1f4ed MS |
2611 | error = path_with_deleted(path, &tmp, &res, &buflen); |
2612 | if (error) | |
2613 | res = ERR_PTR(error); | |
949854d0 | 2614 | write_sequnlock(&rename_lock); |
6ac08c39 | 2615 | path_put(&root); |
1da177e4 LT |
2616 | return res; |
2617 | } | |
ec4f8605 | 2618 | EXPORT_SYMBOL(d_path); |
1da177e4 | 2619 | |
8df9d1a4 MS |
2620 | /** |
2621 | * d_path_with_unreachable - return the path of a dentry | |
2622 | * @path: path to report | |
2623 | * @buf: buffer to return value in | |
2624 | * @buflen: buffer length | |
2625 | * | |
2626 | * The difference from d_path() is that this prepends "(unreachable)" | |
2627 | * to paths which are unreachable from the current process' root. | |
2628 | */ | |
2629 | char *d_path_with_unreachable(const struct path *path, char *buf, int buflen) | |
2630 | { | |
2631 | char *res = buf + buflen; | |
2632 | struct path root; | |
2633 | struct path tmp; | |
2634 | int error; | |
2635 | ||
2636 | if (path->dentry->d_op && path->dentry->d_op->d_dname) | |
2637 | return path->dentry->d_op->d_dname(path->dentry, buf, buflen); | |
2638 | ||
2639 | get_fs_root(current->fs, &root); | |
949854d0 | 2640 | write_seqlock(&rename_lock); |
8df9d1a4 MS |
2641 | tmp = root; |
2642 | error = path_with_deleted(path, &tmp, &res, &buflen); | |
2643 | if (!error && !path_equal(&tmp, &root)) | |
2644 | error = prepend_unreachable(&res, &buflen); | |
949854d0 | 2645 | write_sequnlock(&rename_lock); |
8df9d1a4 MS |
2646 | path_put(&root); |
2647 | if (error) | |
2648 | res = ERR_PTR(error); | |
2649 | ||
2650 | return res; | |
2651 | } | |
2652 | ||
c23fbb6b ED |
2653 | /* |
2654 | * Helper function for dentry_operations.d_dname() members | |
2655 | */ | |
2656 | char *dynamic_dname(struct dentry *dentry, char *buffer, int buflen, | |
2657 | const char *fmt, ...) | |
2658 | { | |
2659 | va_list args; | |
2660 | char temp[64]; | |
2661 | int sz; | |
2662 | ||
2663 | va_start(args, fmt); | |
2664 | sz = vsnprintf(temp, sizeof(temp), fmt, args) + 1; | |
2665 | va_end(args); | |
2666 | ||
2667 | if (sz > sizeof(temp) || sz > buflen) | |
2668 | return ERR_PTR(-ENAMETOOLONG); | |
2669 | ||
2670 | buffer += buflen - sz; | |
2671 | return memcpy(buffer, temp, sz); | |
2672 | } | |
2673 | ||
6092d048 RP |
2674 | /* |
2675 | * Write full pathname from the root of the filesystem into the buffer. | |
2676 | */ | |
ec2447c2 | 2677 | static char *__dentry_path(struct dentry *dentry, char *buf, int buflen) |
6092d048 RP |
2678 | { |
2679 | char *end = buf + buflen; | |
2680 | char *retval; | |
2681 | ||
6092d048 | 2682 | prepend(&end, &buflen, "\0", 1); |
6092d048 RP |
2683 | if (buflen < 1) |
2684 | goto Elong; | |
2685 | /* Get '/' right */ | |
2686 | retval = end-1; | |
2687 | *retval = '/'; | |
2688 | ||
cdd16d02 MS |
2689 | while (!IS_ROOT(dentry)) { |
2690 | struct dentry *parent = dentry->d_parent; | |
9abca360 | 2691 | int error; |
6092d048 | 2692 | |
6092d048 | 2693 | prefetch(parent); |
9abca360 NP |
2694 | spin_lock(&dentry->d_lock); |
2695 | error = prepend_name(&end, &buflen, &dentry->d_name); | |
2696 | spin_unlock(&dentry->d_lock); | |
2697 | if (error != 0 || prepend(&end, &buflen, "/", 1) != 0) | |
6092d048 RP |
2698 | goto Elong; |
2699 | ||
2700 | retval = end; | |
2701 | dentry = parent; | |
2702 | } | |
c103135c AV |
2703 | return retval; |
2704 | Elong: | |
2705 | return ERR_PTR(-ENAMETOOLONG); | |
2706 | } | |
ec2447c2 NP |
2707 | |
2708 | char *dentry_path_raw(struct dentry *dentry, char *buf, int buflen) | |
2709 | { | |
2710 | char *retval; | |
2711 | ||
949854d0 | 2712 | write_seqlock(&rename_lock); |
ec2447c2 | 2713 | retval = __dentry_path(dentry, buf, buflen); |
949854d0 | 2714 | write_sequnlock(&rename_lock); |
ec2447c2 NP |
2715 | |
2716 | return retval; | |
2717 | } | |
2718 | EXPORT_SYMBOL(dentry_path_raw); | |
c103135c AV |
2719 | |
2720 | char *dentry_path(struct dentry *dentry, char *buf, int buflen) | |
2721 | { | |
2722 | char *p = NULL; | |
2723 | char *retval; | |
2724 | ||
949854d0 | 2725 | write_seqlock(&rename_lock); |
c103135c AV |
2726 | if (d_unlinked(dentry)) { |
2727 | p = buf + buflen; | |
2728 | if (prepend(&p, &buflen, "//deleted", 10) != 0) | |
2729 | goto Elong; | |
2730 | buflen++; | |
2731 | } | |
2732 | retval = __dentry_path(dentry, buf, buflen); | |
949854d0 | 2733 | write_sequnlock(&rename_lock); |
c103135c AV |
2734 | if (!IS_ERR(retval) && p) |
2735 | *p = '/'; /* restore '/' overriden with '\0' */ | |
6092d048 RP |
2736 | return retval; |
2737 | Elong: | |
6092d048 RP |
2738 | return ERR_PTR(-ENAMETOOLONG); |
2739 | } | |
2740 | ||
1da177e4 LT |
2741 | /* |
2742 | * NOTE! The user-level library version returns a | |
2743 | * character pointer. The kernel system call just | |
2744 | * returns the length of the buffer filled (which | |
2745 | * includes the ending '\0' character), or a negative | |
2746 | * error value. So libc would do something like | |
2747 | * | |
2748 | * char *getcwd(char * buf, size_t size) | |
2749 | * { | |
2750 | * int retval; | |
2751 | * | |
2752 | * retval = sys_getcwd(buf, size); | |
2753 | * if (retval >= 0) | |
2754 | * return buf; | |
2755 | * errno = -retval; | |
2756 | * return NULL; | |
2757 | * } | |
2758 | */ | |
3cdad428 | 2759 | SYSCALL_DEFINE2(getcwd, char __user *, buf, unsigned long, size) |
1da177e4 | 2760 | { |
552ce544 | 2761 | int error; |
6ac08c39 | 2762 | struct path pwd, root; |
552ce544 | 2763 | char *page = (char *) __get_free_page(GFP_USER); |
1da177e4 LT |
2764 | |
2765 | if (!page) | |
2766 | return -ENOMEM; | |
2767 | ||
f7ad3c6b | 2768 | get_fs_root_and_pwd(current->fs, &root, &pwd); |
1da177e4 | 2769 | |
552ce544 | 2770 | error = -ENOENT; |
949854d0 | 2771 | write_seqlock(&rename_lock); |
f3da392e | 2772 | if (!d_unlinked(pwd.dentry)) { |
552ce544 | 2773 | unsigned long len; |
9d1bc601 | 2774 | struct path tmp = root; |
8df9d1a4 MS |
2775 | char *cwd = page + PAGE_SIZE; |
2776 | int buflen = PAGE_SIZE; | |
1da177e4 | 2777 | |
8df9d1a4 MS |
2778 | prepend(&cwd, &buflen, "\0", 1); |
2779 | error = prepend_path(&pwd, &tmp, &cwd, &buflen); | |
949854d0 | 2780 | write_sequnlock(&rename_lock); |
552ce544 | 2781 | |
8df9d1a4 | 2782 | if (error) |
552ce544 LT |
2783 | goto out; |
2784 | ||
8df9d1a4 MS |
2785 | /* Unreachable from current root */ |
2786 | if (!path_equal(&tmp, &root)) { | |
2787 | error = prepend_unreachable(&cwd, &buflen); | |
2788 | if (error) | |
2789 | goto out; | |
2790 | } | |
2791 | ||
552ce544 LT |
2792 | error = -ERANGE; |
2793 | len = PAGE_SIZE + page - cwd; | |
2794 | if (len <= size) { | |
2795 | error = len; | |
2796 | if (copy_to_user(buf, cwd, len)) | |
2797 | error = -EFAULT; | |
2798 | } | |
949854d0 NP |
2799 | } else { |
2800 | write_sequnlock(&rename_lock); | |
949854d0 | 2801 | } |
1da177e4 LT |
2802 | |
2803 | out: | |
6ac08c39 JB |
2804 | path_put(&pwd); |
2805 | path_put(&root); | |
1da177e4 LT |
2806 | free_page((unsigned long) page); |
2807 | return error; | |
2808 | } | |
2809 | ||
2810 | /* | |
2811 | * Test whether new_dentry is a subdirectory of old_dentry. | |
2812 | * | |
2813 | * Trivially implemented using the dcache structure | |
2814 | */ | |
2815 | ||
2816 | /** | |
2817 | * is_subdir - is new dentry a subdirectory of old_dentry | |
2818 | * @new_dentry: new dentry | |
2819 | * @old_dentry: old dentry | |
2820 | * | |
2821 | * Returns 1 if new_dentry is a subdirectory of the parent (at any depth). | |
2822 | * Returns 0 otherwise. | |
2823 | * Caller must ensure that "new_dentry" is pinned before calling is_subdir() | |
2824 | */ | |
2825 | ||
e2761a11 | 2826 | int is_subdir(struct dentry *new_dentry, struct dentry *old_dentry) |
1da177e4 LT |
2827 | { |
2828 | int result; | |
949854d0 | 2829 | unsigned seq; |
1da177e4 | 2830 | |
e2761a11 OH |
2831 | if (new_dentry == old_dentry) |
2832 | return 1; | |
2833 | ||
e2761a11 | 2834 | do { |
1da177e4 | 2835 | /* for restarting inner loop in case of seq retry */ |
1da177e4 | 2836 | seq = read_seqbegin(&rename_lock); |
949854d0 NP |
2837 | /* |
2838 | * Need rcu_readlock to protect against the d_parent trashing | |
2839 | * due to d_move | |
2840 | */ | |
2841 | rcu_read_lock(); | |
e2761a11 | 2842 | if (d_ancestor(old_dentry, new_dentry)) |
1da177e4 | 2843 | result = 1; |
e2761a11 OH |
2844 | else |
2845 | result = 0; | |
949854d0 | 2846 | rcu_read_unlock(); |
1da177e4 | 2847 | } while (read_seqretry(&rename_lock, seq)); |
1da177e4 LT |
2848 | |
2849 | return result; | |
2850 | } | |
2851 | ||
2096f759 AV |
2852 | int path_is_under(struct path *path1, struct path *path2) |
2853 | { | |
2854 | struct vfsmount *mnt = path1->mnt; | |
2855 | struct dentry *dentry = path1->dentry; | |
2856 | int res; | |
99b7db7b NP |
2857 | |
2858 | br_read_lock(vfsmount_lock); | |
2096f759 AV |
2859 | if (mnt != path2->mnt) { |
2860 | for (;;) { | |
2861 | if (mnt->mnt_parent == mnt) { | |
99b7db7b | 2862 | br_read_unlock(vfsmount_lock); |
2096f759 AV |
2863 | return 0; |
2864 | } | |
2865 | if (mnt->mnt_parent == path2->mnt) | |
2866 | break; | |
2867 | mnt = mnt->mnt_parent; | |
2868 | } | |
2869 | dentry = mnt->mnt_mountpoint; | |
2870 | } | |
2871 | res = is_subdir(dentry, path2->dentry); | |
99b7db7b | 2872 | br_read_unlock(vfsmount_lock); |
2096f759 AV |
2873 | return res; |
2874 | } | |
2875 | EXPORT_SYMBOL(path_is_under); | |
2876 | ||
1da177e4 LT |
2877 | void d_genocide(struct dentry *root) |
2878 | { | |
949854d0 | 2879 | struct dentry *this_parent; |
1da177e4 | 2880 | struct list_head *next; |
949854d0 | 2881 | unsigned seq; |
58db63d0 | 2882 | int locked = 0; |
1da177e4 | 2883 | |
949854d0 | 2884 | seq = read_seqbegin(&rename_lock); |
58db63d0 NP |
2885 | again: |
2886 | this_parent = root; | |
2fd6b7f5 | 2887 | spin_lock(&this_parent->d_lock); |
1da177e4 LT |
2888 | repeat: |
2889 | next = this_parent->d_subdirs.next; | |
2890 | resume: | |
2891 | while (next != &this_parent->d_subdirs) { | |
2892 | struct list_head *tmp = next; | |
5160ee6f | 2893 | struct dentry *dentry = list_entry(tmp, struct dentry, d_u.d_child); |
1da177e4 | 2894 | next = tmp->next; |
949854d0 | 2895 | |
da502956 NP |
2896 | spin_lock_nested(&dentry->d_lock, DENTRY_D_LOCK_NESTED); |
2897 | if (d_unhashed(dentry) || !dentry->d_inode) { | |
2898 | spin_unlock(&dentry->d_lock); | |
1da177e4 | 2899 | continue; |
da502956 | 2900 | } |
1da177e4 | 2901 | if (!list_empty(&dentry->d_subdirs)) { |
2fd6b7f5 NP |
2902 | spin_unlock(&this_parent->d_lock); |
2903 | spin_release(&dentry->d_lock.dep_map, 1, _RET_IP_); | |
1da177e4 | 2904 | this_parent = dentry; |
2fd6b7f5 | 2905 | spin_acquire(&this_parent->d_lock.dep_map, 0, 1, _RET_IP_); |
1da177e4 LT |
2906 | goto repeat; |
2907 | } | |
949854d0 NP |
2908 | if (!(dentry->d_flags & DCACHE_GENOCIDE)) { |
2909 | dentry->d_flags |= DCACHE_GENOCIDE; | |
2910 | dentry->d_count--; | |
2911 | } | |
b7ab39f6 | 2912 | spin_unlock(&dentry->d_lock); |
1da177e4 LT |
2913 | } |
2914 | if (this_parent != root) { | |
949854d0 NP |
2915 | struct dentry *tmp; |
2916 | struct dentry *child; | |
2917 | ||
2918 | tmp = this_parent->d_parent; | |
2919 | if (!(this_parent->d_flags & DCACHE_GENOCIDE)) { | |
2920 | this_parent->d_flags |= DCACHE_GENOCIDE; | |
2921 | this_parent->d_count--; | |
2922 | } | |
2923 | rcu_read_lock(); | |
b7ab39f6 | 2924 | spin_unlock(&this_parent->d_lock); |
949854d0 NP |
2925 | child = this_parent; |
2926 | this_parent = tmp; | |
2fd6b7f5 | 2927 | spin_lock(&this_parent->d_lock); |
949854d0 NP |
2928 | /* might go back up the wrong parent if we have had a rename |
2929 | * or deletion */ | |
2930 | if (this_parent != child->d_parent || | |
58db63d0 | 2931 | (!locked && read_seqretry(&rename_lock, seq))) { |
949854d0 | 2932 | spin_unlock(&this_parent->d_lock); |
949854d0 NP |
2933 | rcu_read_unlock(); |
2934 | goto rename_retry; | |
2935 | } | |
2936 | rcu_read_unlock(); | |
2937 | next = child->d_u.d_child.next; | |
1da177e4 LT |
2938 | goto resume; |
2939 | } | |
2fd6b7f5 | 2940 | spin_unlock(&this_parent->d_lock); |
58db63d0 | 2941 | if (!locked && read_seqretry(&rename_lock, seq)) |
949854d0 | 2942 | goto rename_retry; |
58db63d0 NP |
2943 | if (locked) |
2944 | write_sequnlock(&rename_lock); | |
2945 | return; | |
2946 | ||
2947 | rename_retry: | |
2948 | locked = 1; | |
2949 | write_seqlock(&rename_lock); | |
2950 | goto again; | |
1da177e4 LT |
2951 | } |
2952 | ||
2953 | /** | |
2954 | * find_inode_number - check for dentry with name | |
2955 | * @dir: directory to check | |
2956 | * @name: Name to find. | |
2957 | * | |
2958 | * Check whether a dentry already exists for the given name, | |
2959 | * and return the inode number if it has an inode. Otherwise | |
2960 | * 0 is returned. | |
2961 | * | |
2962 | * This routine is used to post-process directory listings for | |
2963 | * filesystems using synthetic inode numbers, and is necessary | |
2964 | * to keep getcwd() working. | |
2965 | */ | |
2966 | ||
2967 | ino_t find_inode_number(struct dentry *dir, struct qstr *name) | |
2968 | { | |
2969 | struct dentry * dentry; | |
2970 | ino_t ino = 0; | |
2971 | ||
3e7e241f EB |
2972 | dentry = d_hash_and_lookup(dir, name); |
2973 | if (dentry) { | |
1da177e4 LT |
2974 | if (dentry->d_inode) |
2975 | ino = dentry->d_inode->i_ino; | |
2976 | dput(dentry); | |
2977 | } | |
1da177e4 LT |
2978 | return ino; |
2979 | } | |
ec4f8605 | 2980 | EXPORT_SYMBOL(find_inode_number); |
1da177e4 LT |
2981 | |
2982 | static __initdata unsigned long dhash_entries; | |
2983 | static int __init set_dhash_entries(char *str) | |
2984 | { | |
2985 | if (!str) | |
2986 | return 0; | |
2987 | dhash_entries = simple_strtoul(str, &str, 0); | |
2988 | return 1; | |
2989 | } | |
2990 | __setup("dhash_entries=", set_dhash_entries); | |
2991 | ||
2992 | static void __init dcache_init_early(void) | |
2993 | { | |
2994 | int loop; | |
2995 | ||
2996 | /* If hashes are distributed across NUMA nodes, defer | |
2997 | * hash allocation until vmalloc space is available. | |
2998 | */ | |
2999 | if (hashdist) | |
3000 | return; | |
3001 | ||
3002 | dentry_hashtable = | |
3003 | alloc_large_system_hash("Dentry cache", | |
ceb5bdc2 | 3004 | sizeof(struct dcache_hash_bucket), |
1da177e4 LT |
3005 | dhash_entries, |
3006 | 13, | |
3007 | HASH_EARLY, | |
3008 | &d_hash_shift, | |
3009 | &d_hash_mask, | |
3010 | 0); | |
3011 | ||
3012 | for (loop = 0; loop < (1 << d_hash_shift); loop++) | |
ceb5bdc2 | 3013 | INIT_HLIST_BL_HEAD(&dentry_hashtable[loop].head); |
1da177e4 LT |
3014 | } |
3015 | ||
74bf17cf | 3016 | static void __init dcache_init(void) |
1da177e4 LT |
3017 | { |
3018 | int loop; | |
3019 | ||
3020 | /* | |
3021 | * A constructor could be added for stable state like the lists, | |
3022 | * but it is probably not worth it because of the cache nature | |
3023 | * of the dcache. | |
3024 | */ | |
0a31bd5f CL |
3025 | dentry_cache = KMEM_CACHE(dentry, |
3026 | SLAB_RECLAIM_ACCOUNT|SLAB_PANIC|SLAB_MEM_SPREAD); | |
1da177e4 | 3027 | |
8e1f936b | 3028 | register_shrinker(&dcache_shrinker); |
1da177e4 LT |
3029 | |
3030 | /* Hash may have been set up in dcache_init_early */ | |
3031 | if (!hashdist) | |
3032 | return; | |
3033 | ||
3034 | dentry_hashtable = | |
3035 | alloc_large_system_hash("Dentry cache", | |
ceb5bdc2 | 3036 | sizeof(struct dcache_hash_bucket), |
1da177e4 LT |
3037 | dhash_entries, |
3038 | 13, | |
3039 | 0, | |
3040 | &d_hash_shift, | |
3041 | &d_hash_mask, | |
3042 | 0); | |
3043 | ||
3044 | for (loop = 0; loop < (1 << d_hash_shift); loop++) | |
ceb5bdc2 | 3045 | INIT_HLIST_BL_HEAD(&dentry_hashtable[loop].head); |
1da177e4 LT |
3046 | } |
3047 | ||
3048 | /* SLAB cache for __getname() consumers */ | |
e18b890b | 3049 | struct kmem_cache *names_cachep __read_mostly; |
ec4f8605 | 3050 | EXPORT_SYMBOL(names_cachep); |
1da177e4 | 3051 | |
1da177e4 LT |
3052 | EXPORT_SYMBOL(d_genocide); |
3053 | ||
1da177e4 LT |
3054 | void __init vfs_caches_init_early(void) |
3055 | { | |
3056 | dcache_init_early(); | |
3057 | inode_init_early(); | |
3058 | } | |
3059 | ||
3060 | void __init vfs_caches_init(unsigned long mempages) | |
3061 | { | |
3062 | unsigned long reserve; | |
3063 | ||
3064 | /* Base hash sizes on available memory, with a reserve equal to | |
3065 | 150% of current kernel size */ | |
3066 | ||
3067 | reserve = min((mempages - nr_free_pages()) * 3/2, mempages - 1); | |
3068 | mempages -= reserve; | |
3069 | ||
3070 | names_cachep = kmem_cache_create("names_cache", PATH_MAX, 0, | |
20c2df83 | 3071 | SLAB_HWCACHE_ALIGN|SLAB_PANIC, NULL); |
1da177e4 | 3072 | |
74bf17cf DC |
3073 | dcache_init(); |
3074 | inode_init(); | |
1da177e4 | 3075 | files_init(mempages); |
74bf17cf | 3076 | mnt_init(); |
1da177e4 LT |
3077 | bdev_cache_init(); |
3078 | chrdev_init(); | |
3079 | } |