Blame view
fs/super.c
35 KB
1da177e4c
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 |
/* * linux/fs/super.c * * Copyright (C) 1991, 1992 Linus Torvalds * * super.c contains code to handle: - mount structures * - super-block tables * - filesystem drivers list * - mount system call * - umount system call * - ustat system call * * GK 2/5/95 - Changed to support mounting the root fs via NFS * * Added kerneld support: Jacques Gelinas and Bjorn Ekwall * Added change_root: Werner Almesberger & Hans Lermen, Feb '96 * Added options to /proc/mounts: |
96de0e252
|
18 |
* Torbjörn Lindh (torbjorn.lindh@gopta.se), April 14, 1996. |
1da177e4c
|
19 20 21 |
* Added devfs support: Richard Gooch <rgooch@atnf.csiro.au>, 13-JAN-1998 * Heavily rewritten for 'one fs - one tree' dcache architecture. AV, Mar 2000 */ |
630d9c472
|
22 |
#include <linux/export.h> |
1da177e4c
|
23 |
#include <linux/slab.h> |
1da177e4c
|
24 |
#include <linux/blkdev.h> |
1da177e4c
|
25 26 |
#include <linux/mount.h> #include <linux/security.h> |
1da177e4c
|
27 28 |
#include <linux/writeback.h> /* for the emergency remount stuff */ #include <linux/idr.h> |
353ab6e97
|
29 |
#include <linux/mutex.h> |
5477d0fac
|
30 |
#include <linux/backing-dev.h> |
ceb5bdc2d
|
31 |
#include <linux/rculist_bl.h> |
c515e1fd3
|
32 |
#include <linux/cleancache.h> |
404015308
|
33 |
#include <linux/fsnotify.h> |
5accdf82b
|
34 |
#include <linux/lockdep.h> |
6d59e7f58
|
35 |
#include "internal.h" |
1da177e4c
|
36 |
|
15d0f5ea3
|
37 38 |
static LIST_HEAD(super_blocks); static DEFINE_SPINLOCK(sb_lock); |
1da177e4c
|
39 |
|
5accdf82b
|
40 41 42 43 44 |
static char *sb_writers_name[SB_FREEZE_LEVELS] = { "sb_writers", "sb_pagefaults", "sb_internal", }; |
b0d40c92a
|
45 46 47 48 49 50 51 |
/* * One thing we have to be careful of with a per-sb shrinker is that we don't * drop the last active reference to the superblock from within the shrinker. * If that happens we could trigger unregistering the shrinker from within the * shrinker path and that leads to deadlock on the shrinker_rwsem. Hence we * take a passive reference to the superblock to avoid this from occurring. */ |
0a234c6dc
|
52 53 |
static unsigned long super_cache_scan(struct shrinker *shrink, struct shrink_control *sc) |
b0d40c92a
|
54 55 |
{ struct super_block *sb; |
0a234c6dc
|
56 57 58 59 60 |
long fs_objects = 0; long total_objects; long freed = 0; long dentries; long inodes; |
b0d40c92a
|
61 62 63 64 65 66 67 |
sb = container_of(shrink, struct super_block, s_shrink); /* * Deadlock avoidance. We may hold various FS locks, and we don't want * to recurse into the FS that called us in clear_inode() and friends.. */ |
0a234c6dc
|
68 69 |
if (!(sc->gfp_mask & __GFP_FS)) return SHRINK_STOP; |
b0d40c92a
|
70 |
|
eb6ef3df4
|
71 |
if (!trylock_super(sb)) |
0a234c6dc
|
72 |
return SHRINK_STOP; |
b0d40c92a
|
73 |
|
d04079039
|
74 |
if (sb->s_op->nr_cached_objects) |
4101b6243
|
75 |
fs_objects = sb->s_op->nr_cached_objects(sb, sc); |
0e1fdafd9
|
76 |
|
503c358cf
|
77 78 |
inodes = list_lru_shrink_count(&sb->s_inode_lru, sc); dentries = list_lru_shrink_count(&sb->s_dentry_lru, sc); |
f60415675
|
79 |
total_objects = dentries + inodes + fs_objects + 1; |
475d0db74
|
80 81 |
if (!total_objects) total_objects = 1; |
0e1fdafd9
|
82 |
|
0a234c6dc
|
83 |
/* proportion the scan between the caches */ |
f60415675
|
84 |
dentries = mult_frac(sc->nr_to_scan, dentries, total_objects); |
bc3b14cb2
|
85 |
inodes = mult_frac(sc->nr_to_scan, inodes, total_objects); |
503c358cf
|
86 |
fs_objects = mult_frac(sc->nr_to_scan, fs_objects, total_objects); |
b0d40c92a
|
87 |
|
0a234c6dc
|
88 89 90 |
/* * prune the dcache first as the icache is pinned by it, then * prune the icache, followed by the filesystem specific caches |
49e7e7ff8
|
91 92 93 |
* * Ensure that we always scan at least one object - memcg kmem * accounting uses this to fully empty the caches. |
0a234c6dc
|
94 |
*/ |
49e7e7ff8
|
95 |
sc->nr_to_scan = dentries + 1; |
503c358cf
|
96 |
freed = prune_dcache_sb(sb, sc); |
49e7e7ff8
|
97 |
sc->nr_to_scan = inodes + 1; |
503c358cf
|
98 |
freed += prune_icache_sb(sb, sc); |
0a234c6dc
|
99 100 |
if (fs_objects) { |
49e7e7ff8
|
101 |
sc->nr_to_scan = fs_objects + 1; |
4101b6243
|
102 |
freed += sb->s_op->free_cached_objects(sb, sc); |
b0d40c92a
|
103 |
} |
eb6ef3df4
|
104 |
up_read(&sb->s_umount); |
0a234c6dc
|
105 106 107 108 109 110 111 112 113 114 |
return freed; } static unsigned long super_cache_count(struct shrinker *shrink, struct shrink_control *sc) { struct super_block *sb; long total_objects = 0; sb = container_of(shrink, struct super_block, s_shrink); |
d23da150a
|
115 |
/* |
eb6ef3df4
|
116 |
* Don't call trylock_super as it is a potential |
d23da150a
|
117 118 119 |
* scalability bottleneck. The counts could get updated * between super_cache_count and super_cache_scan anyway. * Call to super_cache_count with shrinker_rwsem held |
503c358cf
|
120 |
* ensures the safety of call to list_lru_shrink_count() and |
d23da150a
|
121 122 |
* s_op->nr_cached_objects(). */ |
0a234c6dc
|
123 |
if (sb->s_op && sb->s_op->nr_cached_objects) |
4101b6243
|
124 |
total_objects = sb->s_op->nr_cached_objects(sb, sc); |
0a234c6dc
|
125 |
|
503c358cf
|
126 127 |
total_objects += list_lru_shrink_count(&sb->s_dentry_lru, sc); total_objects += list_lru_shrink_count(&sb->s_inode_lru, sc); |
0a234c6dc
|
128 |
|
55f841ce9
|
129 |
total_objects = vfs_pressure_ratio(total_objects); |
0e1fdafd9
|
130 |
return total_objects; |
b0d40c92a
|
131 |
} |
853b39a7c
|
132 133 134 135 136 137 138 |
static void destroy_super_work(struct work_struct *work) { struct super_block *s = container_of(work, struct super_block, destroy_work); int i; for (i = 0; i < SB_FREEZE_LEVELS; i++) |
8129ed296
|
139 |
percpu_free_rwsem(&s->s_writers.rw_sem[i]); |
853b39a7c
|
140 141 142 143 144 145 146 147 148 |
kfree(s); } static void destroy_super_rcu(struct rcu_head *head) { struct super_block *s = container_of(head, struct super_block, rcu); INIT_WORK(&s->destroy_work, destroy_super_work); schedule_work(&s->destroy_work); } |
7eb5e8826
|
149 150 151 152 153 154 155 |
/** * destroy_super - frees a superblock * @s: superblock to free * * Frees a superblock. */ static void destroy_super(struct super_block *s) |
5accdf82b
|
156 |
{ |
7eb5e8826
|
157 158 |
list_lru_destroy(&s->s_dentry_lru); list_lru_destroy(&s->s_inode_lru); |
7eb5e8826
|
159 160 161 162 |
security_sb_free(s); WARN_ON(!list_empty(&s->s_mounts)); kfree(s->s_subtype); kfree(s->s_options); |
853b39a7c
|
163 |
call_rcu(&s->rcu, destroy_super_rcu); |
5accdf82b
|
164 |
} |
1da177e4c
|
165 166 |
/** * alloc_super - create new superblock |
fe2bbc483
|
167 |
* @type: filesystem type superblock should belong to |
9249e17fe
|
168 |
* @flags: the mount flags |
1da177e4c
|
169 170 171 172 |
* * Allocates and initializes a new &struct super_block. alloc_super() * returns a pointer new superblock or %NULL if allocation had failed. */ |
9249e17fe
|
173 |
static struct super_block *alloc_super(struct file_system_type *type, int flags) |
1da177e4c
|
174 |
{ |
11b0b5abb
|
175 |
struct super_block *s = kzalloc(sizeof(struct super_block), GFP_USER); |
b87221de6
|
176 |
static const struct super_operations default_op; |
7eb5e8826
|
177 178 179 180 |
int i; if (!s) return NULL; |
1da177e4c
|
181 |
|
b5bd856a0
|
182 |
INIT_LIST_HEAD(&s->s_mounts); |
7eb5e8826
|
183 184 |
if (security_sb_alloc(s)) goto fail; |
7b7a8665e
|
185 |
|
7eb5e8826
|
186 |
for (i = 0; i < SB_FREEZE_LEVELS; i++) { |
8129ed296
|
187 188 189 |
if (__percpu_init_rwsem(&s->s_writers.rw_sem[i], sb_writers_name[i], &type->s_writers_key[i])) |
7eb5e8826
|
190 |
goto fail; |
1da177e4c
|
191 |
} |
7eb5e8826
|
192 |
init_waitqueue_head(&s->s_writers.wait_unfrozen); |
df0ce26cb
|
193 |
s->s_bdi = &noop_backing_dev_info; |
7eb5e8826
|
194 |
s->s_flags = flags; |
7eb5e8826
|
195 196 |
INIT_HLIST_NODE(&s->s_instances); INIT_HLIST_BL_HEAD(&s->s_anon); |
e97fedb9e
|
197 |
mutex_init(&s->s_sync_lock); |
7eb5e8826
|
198 |
INIT_LIST_HEAD(&s->s_inodes); |
74278da9f
|
199 |
spin_lock_init(&s->s_inode_list_lock); |
7eb5e8826
|
200 |
|
2acb60a04
|
201 |
if (list_lru_init_memcg(&s->s_dentry_lru)) |
7eb5e8826
|
202 |
goto fail; |
2acb60a04
|
203 |
if (list_lru_init_memcg(&s->s_inode_lru)) |
7eb5e8826
|
204 |
goto fail; |
7eb5e8826
|
205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 |
init_rwsem(&s->s_umount); lockdep_set_class(&s->s_umount, &type->s_umount_key); /* * sget() can have s_umount recursion. * * When it cannot find a suitable sb, it allocates a new * one (this one), and tries again to find a suitable old * one. * * In case that succeeds, it will acquire the s_umount * lock of the old one. Since these are clearly distrinct * locks, and this object isn't exposed yet, there's no * risk of deadlocks. * * Annotate this by putting this lock in a different * subclass. */ down_write_nested(&s->s_umount, SINGLE_DEPTH_NESTING); s->s_count = 1; atomic_set(&s->s_active, 1); mutex_init(&s->s_vfs_rename_mutex); lockdep_set_class(&s->s_vfs_rename_mutex, &type->s_vfs_rename_key); mutex_init(&s->s_dquot.dqio_mutex); mutex_init(&s->s_dquot.dqonoff_mutex); |
7eb5e8826
|
229 230 231 |
s->s_maxbytes = MAX_NON_LFS; s->s_op = &default_op; s->s_time_gran = 1000000000; |
3cb29d111
|
232 |
s->cleancache_poolid = CLEANCACHE_NO_POOL; |
7eb5e8826
|
233 234 235 236 237 |
s->s_shrink.seeks = DEFAULT_SEEKS; s->s_shrink.scan_objects = super_cache_scan; s->s_shrink.count_objects = super_cache_count; s->s_shrink.batch = 1024; |
2acb60a04
|
238 |
s->s_shrink.flags = SHRINKER_NUMA_AWARE | SHRINKER_MEMCG_AWARE; |
1da177e4c
|
239 |
return s; |
5ca302c8e
|
240 |
|
7eb5e8826
|
241 242 243 |
fail: destroy_super(s); return NULL; |
1da177e4c
|
244 245 246 247 248 |
} /* Superblock refcounting */ /* |
35cf7ba0b
|
249 |
* Drop a superblock's refcount. The caller must hold sb_lock. |
1da177e4c
|
250 |
*/ |
f47ec3f28
|
251 |
static void __put_super(struct super_block *sb) |
1da177e4c
|
252 |
{ |
1da177e4c
|
253 |
if (!--sb->s_count) { |
551de6f34
|
254 |
list_del_init(&sb->s_list); |
1da177e4c
|
255 |
destroy_super(sb); |
1da177e4c
|
256 |
} |
1da177e4c
|
257 258 259 260 261 262 263 264 265 |
} /** * put_super - drop a temporary reference to superblock * @sb: superblock in question * * Drops a temporary reference, frees superblock if there's no * references left. */ |
f47ec3f28
|
266 |
static void put_super(struct super_block *sb) |
1da177e4c
|
267 268 269 270 271 272 273 274 |
{ spin_lock(&sb_lock); __put_super(sb); spin_unlock(&sb_lock); } /** |
1712ac8fd
|
275 |
* deactivate_locked_super - drop an active reference to superblock |
1da177e4c
|
276 277 |
* @s: superblock to deactivate * |
1712ac8fd
|
278 279 |
* Drops an active reference to superblock, converting it into a temprory * one if there is no other active references left. In that case we |
1da177e4c
|
280 281 |
* tell fs driver to shut it down and drop the temporary reference we * had just acquired. |
1712ac8fd
|
282 283 |
* * Caller holds exclusive lock on superblock; that lock is released. |
1da177e4c
|
284 |
*/ |
1712ac8fd
|
285 |
void deactivate_locked_super(struct super_block *s) |
1da177e4c
|
286 287 |
{ struct file_system_type *fs = s->s_type; |
b20bd1a5e
|
288 |
if (atomic_dec_and_test(&s->s_active)) { |
3167760f8
|
289 |
cleancache_invalidate_fs(s); |
b0d40c92a
|
290 |
unregister_shrinker(&s->s_shrink); |
28f2cd4f6
|
291 |
fs->kill_sb(s); |
f5e1dd345
|
292 |
|
c0a5b5609
|
293 294 295 296 297 298 299 |
/* * Since list_lru_destroy() may sleep, we cannot call it from * put_super(), where we hold the sb_lock. Therefore we destroy * the lru lists right now. */ list_lru_destroy(&s->s_dentry_lru); list_lru_destroy(&s->s_inode_lru); |
1da177e4c
|
300 301 |
put_filesystem(fs); put_super(s); |
1712ac8fd
|
302 303 |
} else { up_write(&s->s_umount); |
1da177e4c
|
304 305 |
} } |
1712ac8fd
|
306 |
EXPORT_SYMBOL(deactivate_locked_super); |
1da177e4c
|
307 308 |
/** |
1712ac8fd
|
309 |
* deactivate_super - drop an active reference to superblock |
74dbbdd7f
|
310 311 |
* @s: superblock to deactivate * |
1712ac8fd
|
312 313 314 |
* Variant of deactivate_locked_super(), except that superblock is *not* * locked by caller. If we are going to drop the final active reference, * lock will be acquired prior to that. |
74dbbdd7f
|
315 |
*/ |
1712ac8fd
|
316 |
void deactivate_super(struct super_block *s) |
74dbbdd7f
|
317 |
{ |
1712ac8fd
|
318 319 320 |
if (!atomic_add_unless(&s->s_active, -1, 1)) { down_write(&s->s_umount); deactivate_locked_super(s); |
74dbbdd7f
|
321 322 |
} } |
1712ac8fd
|
323 |
EXPORT_SYMBOL(deactivate_super); |
74dbbdd7f
|
324 325 |
/** |
1da177e4c
|
326 327 328 329 330 331 332 333 |
* grab_super - acquire an active reference * @s: reference we are trying to make active * * Tries to acquire an active reference. grab_super() is used when we * had just found a superblock in super_blocks or fs_type->fs_supers * and want to turn it into a full-blown active reference. grab_super() * is called with sb_lock held and drops it. Returns 1 in case of * success, 0 if we had failed (superblock contents was already dead or |
acfec9a5a
|
334 335 336 |
* dying when grab_super() had been called). Note that this is only * called for superblocks not in rundown mode (== ones still on ->fs_supers * of their type), so increment of ->s_count is OK here. |
1da177e4c
|
337 |
*/ |
9c4dbee79
|
338 |
static int grab_super(struct super_block *s) __releases(sb_lock) |
1da177e4c
|
339 340 341 342 |
{ s->s_count++; spin_unlock(&sb_lock); down_write(&s->s_umount); |
acfec9a5a
|
343 344 345 346 |
if ((s->s_flags & MS_BORN) && atomic_inc_not_zero(&s->s_active)) { put_super(s); return 1; } |
1da177e4c
|
347 348 |
up_write(&s->s_umount); put_super(s); |
1da177e4c
|
349 350 |
return 0; } |
cf9a2ae8d
|
351 |
/* |
eb6ef3df4
|
352 |
* trylock_super - try to grab ->s_umount shared |
331cbdeed
|
353 |
* @sb: reference we are trying to grab |
12ad3ab66
|
354 |
* |
eb6ef3df4
|
355 |
* Try to prevent fs shutdown. This is used in places where we |
12ad3ab66
|
356 |
* cannot take an active reference but we need to ensure that the |
eb6ef3df4
|
357 358 359 360 361 362 363 364 365 366 |
* filesystem is not shut down while we are working on it. It returns * false if we cannot acquire s_umount or if we lose the race and * filesystem already got into shutdown, and returns true with the s_umount * lock held in read mode in case of success. On successful return, * the caller must drop the s_umount lock when done. * * Note that unlike get_super() et.al. this one does *not* bump ->s_count. * The reason why it's safe is that we are OK with doing trylock instead * of down_read(). There's a couple of places that are OK with that, but * it's very much not a general-purpose interface. |
12ad3ab66
|
367 |
*/ |
eb6ef3df4
|
368 |
bool trylock_super(struct super_block *sb) |
12ad3ab66
|
369 |
{ |
12ad3ab66
|
370 |
if (down_read_trylock(&sb->s_umount)) { |
eb6ef3df4
|
371 372 |
if (!hlist_unhashed(&sb->s_instances) && sb->s_root && (sb->s_flags & MS_BORN)) |
12ad3ab66
|
373 374 375 |
return true; up_read(&sb->s_umount); } |
12ad3ab66
|
376 377 |
return false; } |
1da177e4c
|
378 379 380 381 382 383 384 385 386 |
/** * generic_shutdown_super - common helper for ->kill_sb() * @sb: superblock to kill * * generic_shutdown_super() does all fs-independent work on superblock * shutdown. Typical ->kill_sb() should pick all fs-specific objects * that need destruction out of superblock, call generic_shutdown_super() * and release aforementioned objects. Note: dentries and inodes _are_ * taken care of and do not need specific handling. |
c636ebdb1
|
387 388 389 390 |
* * Upon calling this function, the filesystem may no longer alter or * rearrange the set of dentries belonging to this super_block, nor may it * change the attachments of dentries to inodes. |
1da177e4c
|
391 392 393 |
*/ void generic_shutdown_super(struct super_block *sb) { |
ee9b6d61a
|
394 |
const struct super_operations *sop = sb->s_op; |
1da177e4c
|
395 |
|
c636ebdb1
|
396 397 |
if (sb->s_root) { shrink_dcache_for_umount(sb); |
60b0680fa
|
398 |
sync_filesystem(sb); |
1da177e4c
|
399 |
sb->s_flags &= ~MS_ACTIVE; |
efaee1920
|
400 |
|
74278da9f
|
401 |
fsnotify_unmount_inodes(sb); |
63997e98a
|
402 403 |
evict_inodes(sb); |
1da177e4c
|
404 |
|
7b7a8665e
|
405 406 407 408 |
if (sb->s_dio_done_wq) { destroy_workqueue(sb->s_dio_done_wq); sb->s_dio_done_wq = NULL; } |
1da177e4c
|
409 410 |
if (sop->put_super) sop->put_super(sb); |
63997e98a
|
411 |
if (!list_empty(&sb->s_inodes)) { |
7b4fe29e0
|
412 413 414 415 |
printk("VFS: Busy inodes after unmount of %s. " "Self-destruct in 5 seconds. Have a nice day... ", sb->s_id); |
1da177e4c
|
416 |
} |
1da177e4c
|
417 418 419 |
} spin_lock(&sb_lock); /* should be initialized for __put_super_and_need_restart() */ |
a5166169f
|
420 |
hlist_del_init(&sb->s_instances); |
1da177e4c
|
421 422 423 424 425 426 427 428 429 430 431 |
spin_unlock(&sb_lock); up_write(&sb->s_umount); } EXPORT_SYMBOL(generic_shutdown_super); /** * sget - find or create a superblock * @type: filesystem type superblock should belong to * @test: comparison callback * @set: setup callback |
9249e17fe
|
432 |
* @flags: mount flags |
1da177e4c
|
433 434 435 436 437 |
* @data: argument to each of them */ struct super_block *sget(struct file_system_type *type, int (*test)(struct super_block *,void *), int (*set)(struct super_block *,void *), |
9249e17fe
|
438 |
int flags, |
1da177e4c
|
439 440 441 |
void *data) { struct super_block *s = NULL; |
d47301271
|
442 |
struct super_block *old; |
1da177e4c
|
443 444 445 446 |
int err; retry: spin_lock(&sb_lock); |
d47301271
|
447 |
if (test) { |
b67bfe0d4
|
448 |
hlist_for_each_entry(old, &type->fs_supers, s_instances) { |
d47301271
|
449 450 451 452 |
if (!test(old, data)) continue; if (!grab_super(old)) goto retry; |
a3cfbb53b
|
453 454 |
if (s) { up_write(&s->s_umount); |
d47301271
|
455 |
destroy_super(s); |
7a4dec538
|
456 |
s = NULL; |
a3cfbb53b
|
457 |
} |
d47301271
|
458 459 |
return old; } |
1da177e4c
|
460 461 462 |
} if (!s) { spin_unlock(&sb_lock); |
9249e17fe
|
463 |
s = alloc_super(type, flags); |
1da177e4c
|
464 465 466 467 468 469 470 471 |
if (!s) return ERR_PTR(-ENOMEM); goto retry; } err = set(s, data); if (err) { spin_unlock(&sb_lock); |
a3cfbb53b
|
472 |
up_write(&s->s_umount); |
1da177e4c
|
473 474 475 476 477 478 |
destroy_super(s); return ERR_PTR(err); } s->s_type = type; strlcpy(s->s_id, type->name, sizeof(s->s_id)); list_add_tail(&s->s_list, &super_blocks); |
a5166169f
|
479 |
hlist_add_head(&s->s_instances, &type->fs_supers); |
1da177e4c
|
480 481 |
spin_unlock(&sb_lock); get_filesystem(type); |
b0d40c92a
|
482 |
register_shrinker(&s->s_shrink); |
1da177e4c
|
483 484 485 486 487 488 489 490 491 492 493 494 |
return s; } EXPORT_SYMBOL(sget); void drop_super(struct super_block *sb) { up_read(&sb->s_umount); put_super(sb); } EXPORT_SYMBOL(drop_super); |
e50047533
|
495 |
/** |
01a05b337
|
496 497 498 499 500 501 502 503 504 |
* iterate_supers - call function for all active superblocks * @f: function to call * @arg: argument to pass to it * * Scans the superblock list and calls given function, passing it * locked superblock and given argument. */ void iterate_supers(void (*f)(struct super_block *, void *), void *arg) { |
dca332528
|
505 |
struct super_block *sb, *p = NULL; |
01a05b337
|
506 507 |
spin_lock(&sb_lock); |
dca332528
|
508 |
list_for_each_entry(sb, &super_blocks, s_list) { |
a5166169f
|
509 |
if (hlist_unhashed(&sb->s_instances)) |
01a05b337
|
510 511 512 513 514 |
continue; sb->s_count++; spin_unlock(&sb_lock); down_read(&sb->s_umount); |
dabe0dc19
|
515 |
if (sb->s_root && (sb->s_flags & MS_BORN)) |
01a05b337
|
516 517 518 519 |
f(sb, arg); up_read(&sb->s_umount); spin_lock(&sb_lock); |
dca332528
|
520 521 522 |
if (p) __put_super(p); p = sb; |
01a05b337
|
523 |
} |
dca332528
|
524 525 |
if (p) __put_super(p); |
01a05b337
|
526 527 528 529 |
spin_unlock(&sb_lock); } /** |
43e15cdbe
|
530 531 532 533 534 535 536 537 538 539 540 541 542 543 |
* iterate_supers_type - call function for superblocks of given type * @type: fs type * @f: function to call * @arg: argument to pass to it * * Scans the superblock list and calls given function, passing it * locked superblock and given argument. */ void iterate_supers_type(struct file_system_type *type, void (*f)(struct super_block *, void *), void *arg) { struct super_block *sb, *p = NULL; spin_lock(&sb_lock); |
b67bfe0d4
|
544 |
hlist_for_each_entry(sb, &type->fs_supers, s_instances) { |
43e15cdbe
|
545 546 547 548 |
sb->s_count++; spin_unlock(&sb_lock); down_read(&sb->s_umount); |
dabe0dc19
|
549 |
if (sb->s_root && (sb->s_flags & MS_BORN)) |
43e15cdbe
|
550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 |
f(sb, arg); up_read(&sb->s_umount); spin_lock(&sb_lock); if (p) __put_super(p); p = sb; } if (p) __put_super(p); spin_unlock(&sb_lock); } EXPORT_SYMBOL(iterate_supers_type); /** |
1da177e4c
|
566 567 568 569 570 571 |
* get_super - get the superblock of a device * @bdev: device to get the superblock for * * Scans the superblock list and finds the superblock of the file system * mounted on the device given. %NULL is returned if no match is found. */ |
df40c01a9
|
572 |
struct super_block *get_super(struct block_device *bdev) |
1da177e4c
|
573 |
{ |
618f06362
|
574 |
struct super_block *sb; |
1da177e4c
|
575 576 |
if (!bdev) return NULL; |
618f06362
|
577 |
|
1da177e4c
|
578 |
spin_lock(&sb_lock); |
618f06362
|
579 580 |
rescan: list_for_each_entry(sb, &super_blocks, s_list) { |
a5166169f
|
581 |
if (hlist_unhashed(&sb->s_instances)) |
551de6f34
|
582 |
continue; |
618f06362
|
583 584 |
if (sb->s_bdev == bdev) { sb->s_count++; |
1da177e4c
|
585 |
spin_unlock(&sb_lock); |
618f06362
|
586 |
down_read(&sb->s_umount); |
df40c01a9
|
587 |
/* still alive? */ |
dabe0dc19
|
588 |
if (sb->s_root && (sb->s_flags & MS_BORN)) |
618f06362
|
589 590 |
return sb; up_read(&sb->s_umount); |
df40c01a9
|
591 |
/* nope, got unmounted */ |
618f06362
|
592 |
spin_lock(&sb_lock); |
df40c01a9
|
593 594 |
__put_super(sb); goto rescan; |
1da177e4c
|
595 596 597 598 599 600 601 |
} } spin_unlock(&sb_lock); return NULL; } EXPORT_SYMBOL(get_super); |
4504230a7
|
602 603 |
/** |
6b6dc836a
|
604 605 606 607 608 609 610 611 612 613 614 615 |
* get_super_thawed - get thawed superblock of a device * @bdev: device to get the superblock for * * Scans the superblock list and finds the superblock of the file system * mounted on the device. The superblock is returned once it is thawed * (or immediately if it was not frozen). %NULL is returned if no match * is found. */ struct super_block *get_super_thawed(struct block_device *bdev) { while (1) { struct super_block *s = get_super(bdev); |
5accdf82b
|
616 |
if (!s || s->s_writers.frozen == SB_UNFROZEN) |
6b6dc836a
|
617 618 |
return s; up_read(&s->s_umount); |
5accdf82b
|
619 620 |
wait_event(s->s_writers.wait_unfrozen, s->s_writers.frozen == SB_UNFROZEN); |
6b6dc836a
|
621 622 623 624 625 626 |
put_super(s); } } EXPORT_SYMBOL(get_super_thawed); /** |
4504230a7
|
627 628 629 630 631 |
* get_active_super - get an active reference to the superblock of a device * @bdev: device to get the superblock for * * Scans the superblock list and finds the superblock of the file system * mounted on the device given. Returns the superblock with an active |
d3f214730
|
632 |
* reference or %NULL if none was found. |
4504230a7
|
633 634 635 636 637 638 639 |
*/ struct super_block *get_active_super(struct block_device *bdev) { struct super_block *sb; if (!bdev) return NULL; |
1494583de
|
640 |
restart: |
4504230a7
|
641 642 |
spin_lock(&sb_lock); list_for_each_entry(sb, &super_blocks, s_list) { |
a5166169f
|
643 |
if (hlist_unhashed(&sb->s_instances)) |
551de6f34
|
644 |
continue; |
1494583de
|
645 |
if (sb->s_bdev == bdev) { |
acfec9a5a
|
646 |
if (!grab_super(sb)) |
1494583de
|
647 |
goto restart; |
acfec9a5a
|
648 649 |
up_write(&sb->s_umount); return sb; |
1494583de
|
650 |
} |
4504230a7
|
651 652 653 654 |
} spin_unlock(&sb_lock); return NULL; } |
1da177e4c
|
655 |
|
df40c01a9
|
656 |
struct super_block *user_get_super(dev_t dev) |
1da177e4c
|
657 |
{ |
618f06362
|
658 |
struct super_block *sb; |
1da177e4c
|
659 |
|
1da177e4c
|
660 |
spin_lock(&sb_lock); |
618f06362
|
661 662 |
rescan: list_for_each_entry(sb, &super_blocks, s_list) { |
a5166169f
|
663 |
if (hlist_unhashed(&sb->s_instances)) |
551de6f34
|
664 |
continue; |
618f06362
|
665 666 |
if (sb->s_dev == dev) { sb->s_count++; |
1da177e4c
|
667 |
spin_unlock(&sb_lock); |
618f06362
|
668 |
down_read(&sb->s_umount); |
df40c01a9
|
669 |
/* still alive? */ |
dabe0dc19
|
670 |
if (sb->s_root && (sb->s_flags & MS_BORN)) |
618f06362
|
671 672 |
return sb; up_read(&sb->s_umount); |
df40c01a9
|
673 |
/* nope, got unmounted */ |
618f06362
|
674 |
spin_lock(&sb_lock); |
df40c01a9
|
675 676 |
__put_super(sb); goto rescan; |
1da177e4c
|
677 678 679 680 681 |
} } spin_unlock(&sb_lock); return NULL; } |
1da177e4c
|
682 |
/** |
1da177e4c
|
683 684 685 686 687 688 689 690 691 692 693 |
* do_remount_sb - asks filesystem to change mount options. * @sb: superblock in question * @flags: numeric part of options * @data: the rest of options * @force: whether or not to force the change * * Alters the mount options of a mounted file system. */ int do_remount_sb(struct super_block *sb, int flags, void *data, int force) { int retval; |
c79d967de
|
694 |
int remount_ro; |
4504230a7
|
695 |
|
5accdf82b
|
696 |
if (sb->s_writers.frozen != SB_UNFROZEN) |
4504230a7
|
697 |
return -EBUSY; |
9361401eb
|
698 |
#ifdef CONFIG_BLOCK |
1da177e4c
|
699 700 |
if (!(flags & MS_RDONLY) && bdev_read_only(sb->s_bdev)) return -EACCES; |
9361401eb
|
701 |
#endif |
4504230a7
|
702 |
|
d208bbdda
|
703 |
remount_ro = (flags & MS_RDONLY) && !(sb->s_flags & MS_RDONLY); |
d208bbdda
|
704 |
|
0aec09d04
|
705 |
if (remount_ro) { |
fdab684d7
|
706 |
if (!hlist_empty(&sb->s_pins)) { |
0aec09d04
|
707 |
up_write(&sb->s_umount); |
fdab684d7
|
708 |
group_pin_kill(&sb->s_pins); |
0aec09d04
|
709 710 711 712 713 714 715 716 717 |
down_write(&sb->s_umount); if (!sb->s_root) return 0; if (sb->s_writers.frozen != SB_UNFROZEN) return -EBUSY; remount_ro = (flags & MS_RDONLY) && !(sb->s_flags & MS_RDONLY); } } shrink_dcache_sb(sb); |
1da177e4c
|
718 719 |
/* If we are remounting RDONLY and current sb is read/write, make sure there are no rw files opened */ |
d208bbdda
|
720 |
if (remount_ro) { |
4ed5e82fe
|
721 |
if (force) { |
eee5cc270
|
722 723 |
sb->s_readonly_remount = 1; smp_wmb(); |
4ed5e82fe
|
724 725 726 727 |
} else { retval = sb_prepare_remount_readonly(sb); if (retval) return retval; |
4ed5e82fe
|
728 |
} |
1da177e4c
|
729 730 731 |
} if (sb->s_op->remount_fs) { |
1da177e4c
|
732 |
retval = sb->s_op->remount_fs(sb, &flags, data); |
2833eb2b4
|
733 734 |
if (retval) { if (!force) |
4ed5e82fe
|
735 |
goto cancel_readonly; |
2833eb2b4
|
736 737 738 739 740 |
/* If forced remount, go ahead despite any errors */ WARN(1, "forced remount of a %s fs returned %i ", sb->s_type->name, retval); } |
1da177e4c
|
741 742 |
} sb->s_flags = (sb->s_flags & ~MS_RMT_MASK) | (flags & MS_RMT_MASK); |
4ed5e82fe
|
743 744 745 |
/* Needs to be ordered wrt mnt_is_readonly() */ smp_wmb(); sb->s_readonly_remount = 0; |
c79d967de
|
746 |
|
d208bbdda
|
747 748 749 750 751 752 753 754 755 756 |
/* * Some filesystems modify their metadata via some other path than the * bdev buffer cache (eg. use a private mapping, or directories in * pagecache, etc). Also file data modifications go via their own * mappings. So If we try to mount readonly then copy the filesystem * from bdev, we could get stale data, so invalidate it to give a best * effort at coherency. */ if (remount_ro && sb->s_bdev) invalidate_bdev(sb->s_bdev); |
1da177e4c
|
757 |
return 0; |
4ed5e82fe
|
758 759 760 761 |
cancel_readonly: sb->s_readonly_remount = 0; return retval; |
1da177e4c
|
762 |
} |
a2a9537ac
|
763 |
static void do_emergency_remount(struct work_struct *work) |
1da177e4c
|
764 |
{ |
dca332528
|
765 |
struct super_block *sb, *p = NULL; |
1da177e4c
|
766 767 |
spin_lock(&sb_lock); |
dca332528
|
768 |
list_for_each_entry(sb, &super_blocks, s_list) { |
a5166169f
|
769 |
if (hlist_unhashed(&sb->s_instances)) |
551de6f34
|
770 |
continue; |
1da177e4c
|
771 772 |
sb->s_count++; spin_unlock(&sb_lock); |
443b94baa
|
773 |
down_write(&sb->s_umount); |
dabe0dc19
|
774 775 |
if (sb->s_root && sb->s_bdev && (sb->s_flags & MS_BORN) && !(sb->s_flags & MS_RDONLY)) { |
1da177e4c
|
776 |
/* |
1da177e4c
|
777 778 |
* What lock protects sb->s_flags?? */ |
1da177e4c
|
779 |
do_remount_sb(sb, MS_RDONLY, NULL, 1); |
1da177e4c
|
780 |
} |
443b94baa
|
781 |
up_write(&sb->s_umount); |
1da177e4c
|
782 |
spin_lock(&sb_lock); |
dca332528
|
783 784 785 |
if (p) __put_super(p); p = sb; |
1da177e4c
|
786 |
} |
dca332528
|
787 788 |
if (p) __put_super(p); |
1da177e4c
|
789 |
spin_unlock(&sb_lock); |
a2a9537ac
|
790 |
kfree(work); |
1da177e4c
|
791 792 793 794 795 796 |
printk("Emergency Remount complete "); } void emergency_remount(void) { |
a2a9537ac
|
797 798 799 800 801 802 803 |
struct work_struct *work; work = kmalloc(sizeof(*work), GFP_ATOMIC); if (work) { INIT_WORK(work, do_emergency_remount); schedule_work(work); } |
1da177e4c
|
804 805 806 807 808 809 |
} /* * Unnamed block devices are dummy devices used by virtual * filesystems which don't use real block-devices. -- jrs */ |
ad76cbc63
|
810 |
static DEFINE_IDA(unnamed_dev_ida); |
1da177e4c
|
811 |
static DEFINE_SPINLOCK(unnamed_dev_lock);/* protects the above */ |
a2a4dc494
|
812 813 814 815 |
/* Many userspace utilities consider an FSID of 0 invalid. * Always return at least 1 from get_anon_bdev. */ static int unnamed_dev_start = 1; |
1da177e4c
|
816 |
|
0ee5dc676
|
817 |
int get_anon_bdev(dev_t *p) |
1da177e4c
|
818 819 820 821 822 |
{ int dev; int error; retry: |
ad76cbc63
|
823 |
if (ida_pre_get(&unnamed_dev_ida, GFP_ATOMIC) == 0) |
1da177e4c
|
824 825 |
return -ENOMEM; spin_lock(&unnamed_dev_lock); |
c63e09ecc
|
826 |
error = ida_get_new_above(&unnamed_dev_ida, unnamed_dev_start, &dev); |
f21f62208
|
827 828 |
if (!error) unnamed_dev_start = dev + 1; |
1da177e4c
|
829 830 831 832 833 834 |
spin_unlock(&unnamed_dev_lock); if (error == -EAGAIN) /* We raced and lost with another CPU. */ goto retry; else if (error) return -EAGAIN; |
1af95de6f
|
835 |
if (dev >= (1 << MINORBITS)) { |
1da177e4c
|
836 |
spin_lock(&unnamed_dev_lock); |
ad76cbc63
|
837 |
ida_remove(&unnamed_dev_ida, dev); |
f21f62208
|
838 839 |
if (unnamed_dev_start > dev) unnamed_dev_start = dev; |
1da177e4c
|
840 841 842 |
spin_unlock(&unnamed_dev_lock); return -EMFILE; } |
0ee5dc676
|
843 |
*p = MKDEV(0, dev & MINORMASK); |
1da177e4c
|
844 845 |
return 0; } |
0ee5dc676
|
846 |
EXPORT_SYMBOL(get_anon_bdev); |
1da177e4c
|
847 |
|
0ee5dc676
|
848 |
void free_anon_bdev(dev_t dev) |
1da177e4c
|
849 |
{ |
0ee5dc676
|
850 |
int slot = MINOR(dev); |
1da177e4c
|
851 |
spin_lock(&unnamed_dev_lock); |
ad76cbc63
|
852 |
ida_remove(&unnamed_dev_ida, slot); |
c63e09ecc
|
853 854 |
if (slot < unnamed_dev_start) unnamed_dev_start = slot; |
1da177e4c
|
855 856 |
spin_unlock(&unnamed_dev_lock); } |
0ee5dc676
|
857 858 859 860 |
EXPORT_SYMBOL(free_anon_bdev); int set_anon_super(struct super_block *s, void *data) { |
df0ce26cb
|
861 |
return get_anon_bdev(&s->s_dev); |
0ee5dc676
|
862 863 864 865 866 867 868 869 870 871 |
} EXPORT_SYMBOL(set_anon_super); void kill_anon_super(struct super_block *sb) { dev_t dev = sb->s_dev; generic_shutdown_super(sb); free_anon_bdev(dev); } |
1da177e4c
|
872 873 |
EXPORT_SYMBOL(kill_anon_super); |
1da177e4c
|
874 875 876 877 878 879 880 881 |
void kill_litter_super(struct super_block *sb) { if (sb->s_root) d_genocide(sb->s_root); kill_anon_super(sb); } EXPORT_SYMBOL(kill_litter_super); |
909e6d947
|
882 883 884 885 886 887 888 889 890 891 |
static int ns_test_super(struct super_block *sb, void *data) { return sb->s_fs_info == data; } static int ns_set_super(struct super_block *sb, void *data) { sb->s_fs_info = data; return set_anon_super(sb, NULL); } |
ceefda693
|
892 893 |
struct dentry *mount_ns(struct file_system_type *fs_type, int flags, void *data, int (*fill_super)(struct super_block *, void *, int)) |
909e6d947
|
894 895 |
{ struct super_block *sb; |
9249e17fe
|
896 |
sb = sget(fs_type, ns_test_super, ns_set_super, flags, data); |
909e6d947
|
897 |
if (IS_ERR(sb)) |
ceefda693
|
898 |
return ERR_CAST(sb); |
909e6d947
|
899 900 901 |
if (!sb->s_root) { int err; |
909e6d947
|
902 903 |
err = fill_super(sb, data, flags & MS_SILENT ? 1 : 0); if (err) { |
74dbbdd7f
|
904 |
deactivate_locked_super(sb); |
ceefda693
|
905 |
return ERR_PTR(err); |
909e6d947
|
906 907 908 909 |
} sb->s_flags |= MS_ACTIVE; } |
ceefda693
|
910 |
return dget(sb->s_root); |
909e6d947
|
911 |
} |
ceefda693
|
912 |
EXPORT_SYMBOL(mount_ns); |
909e6d947
|
913 |
|
9361401eb
|
914 |
#ifdef CONFIG_BLOCK |
1da177e4c
|
915 916 917 918 |
static int set_bdev_super(struct super_block *s, void *data) { s->s_bdev = data; s->s_dev = s->s_bdev->bd_dev; |
32a88aa1b
|
919 920 921 922 923 924 |
/* * We set the bdi here to the queue backing, file systems can * overwrite this in ->fill_super() */ s->s_bdi = &bdev_get_queue(s->s_bdev)->backing_dev_info; |
1da177e4c
|
925 926 927 928 929 930 931 |
return 0; } static int test_bdev_super(struct super_block *s, void *data) { return (void *)s->s_bdev == data; } |
152a08366
|
932 |
struct dentry *mount_bdev(struct file_system_type *fs_type, |
1da177e4c
|
933 |
int flags, const char *dev_name, void *data, |
152a08366
|
934 |
int (*fill_super)(struct super_block *, void *, int)) |
1da177e4c
|
935 936 937 |
{ struct block_device *bdev; struct super_block *s; |
d4d776299
|
938 |
fmode_t mode = FMODE_READ | FMODE_EXCL; |
1da177e4c
|
939 |
int error = 0; |
30c40d2c0
|
940 941 |
if (!(flags & MS_RDONLY)) mode |= FMODE_WRITE; |
d4d776299
|
942 |
bdev = blkdev_get_by_path(dev_name, mode, fs_type); |
1da177e4c
|
943 |
if (IS_ERR(bdev)) |
152a08366
|
944 |
return ERR_CAST(bdev); |
1da177e4c
|
945 946 947 948 949 950 |
/* * once the super is inserted into the list by sget, s_umount * will protect the lockfs code from trying to start a snapshot * while we are mounting */ |
4fadd7bb2
|
951 952 953 954 955 956 |
mutex_lock(&bdev->bd_fsfreeze_mutex); if (bdev->bd_fsfreeze_count > 0) { mutex_unlock(&bdev->bd_fsfreeze_mutex); error = -EBUSY; goto error_bdev; } |
9249e17fe
|
957 958 |
s = sget(fs_type, test_bdev_super, set_bdev_super, flags | MS_NOSEC, bdev); |
4fadd7bb2
|
959 |
mutex_unlock(&bdev->bd_fsfreeze_mutex); |
1da177e4c
|
960 |
if (IS_ERR(s)) |
454e2398b
|
961 |
goto error_s; |
1da177e4c
|
962 963 964 |
if (s->s_root) { if ((flags ^ s->s_flags) & MS_RDONLY) { |
74dbbdd7f
|
965 |
deactivate_locked_super(s); |
454e2398b
|
966 967 |
error = -EBUSY; goto error_bdev; |
1da177e4c
|
968 |
} |
454e2398b
|
969 |
|
4f331f01b
|
970 971 |
/* * s_umount nests inside bd_mutex during |
e525fd89d
|
972 973 974 975 |
* __invalidate_device(). blkdev_put() acquires * bd_mutex and can't be called under s_umount. Drop * s_umount temporarily. This is safe as we're * holding an active reference. |
4f331f01b
|
976 977 |
*/ up_write(&s->s_umount); |
d4d776299
|
978 |
blkdev_put(bdev, mode); |
4f331f01b
|
979 |
down_write(&s->s_umount); |
1da177e4c
|
980 981 |
} else { char b[BDEVNAME_SIZE]; |
30c40d2c0
|
982 |
s->s_mode = mode; |
1da177e4c
|
983 |
strlcpy(s->s_id, bdevname(bdev, b), sizeof(s->s_id)); |
e78c9a004
|
984 |
sb_set_blocksize(s, block_size(bdev)); |
9b04c997b
|
985 |
error = fill_super(s, data, flags & MS_SILENT ? 1 : 0); |
1da177e4c
|
986 |
if (error) { |
74dbbdd7f
|
987 |
deactivate_locked_super(s); |
454e2398b
|
988 |
goto error; |
fa675765a
|
989 |
} |
454e2398b
|
990 991 |
s->s_flags |= MS_ACTIVE; |
87d8fe1ee
|
992 |
bdev->bd_super = s; |
1da177e4c
|
993 |
} |
152a08366
|
994 |
return dget(s->s_root); |
1da177e4c
|
995 |
|
454e2398b
|
996 997 998 |
error_s: error = PTR_ERR(s); error_bdev: |
d4d776299
|
999 |
blkdev_put(bdev, mode); |
454e2398b
|
1000 |
error: |
152a08366
|
1001 1002 1003 |
return ERR_PTR(error); } EXPORT_SYMBOL(mount_bdev); |
1da177e4c
|
1004 1005 1006 |
void kill_block_super(struct super_block *sb) { struct block_device *bdev = sb->s_bdev; |
30c40d2c0
|
1007 |
fmode_t mode = sb->s_mode; |
1da177e4c
|
1008 |
|
ddbaaf302
|
1009 |
bdev->bd_super = NULL; |
1da177e4c
|
1010 1011 |
generic_shutdown_super(sb); sync_blockdev(bdev); |
d4d776299
|
1012 |
WARN_ON_ONCE(!(mode & FMODE_EXCL)); |
e525fd89d
|
1013 |
blkdev_put(bdev, mode | FMODE_EXCL); |
1da177e4c
|
1014 1015 1016 |
} EXPORT_SYMBOL(kill_block_super); |
9361401eb
|
1017 |
#endif |
1da177e4c
|
1018 |
|
3c26ff6e4
|
1019 |
struct dentry *mount_nodev(struct file_system_type *fs_type, |
1da177e4c
|
1020 |
int flags, void *data, |
3c26ff6e4
|
1021 |
int (*fill_super)(struct super_block *, void *, int)) |
1da177e4c
|
1022 1023 |
{ int error; |
9249e17fe
|
1024 |
struct super_block *s = sget(fs_type, NULL, set_anon_super, flags, NULL); |
1da177e4c
|
1025 1026 |
if (IS_ERR(s)) |
3c26ff6e4
|
1027 |
return ERR_CAST(s); |
1da177e4c
|
1028 |
|
9b04c997b
|
1029 |
error = fill_super(s, data, flags & MS_SILENT ? 1 : 0); |
1da177e4c
|
1030 |
if (error) { |
74dbbdd7f
|
1031 |
deactivate_locked_super(s); |
3c26ff6e4
|
1032 |
return ERR_PTR(error); |
1da177e4c
|
1033 1034 |
} s->s_flags |= MS_ACTIVE; |
3c26ff6e4
|
1035 |
return dget(s->s_root); |
1da177e4c
|
1036 |
} |
3c26ff6e4
|
1037 |
EXPORT_SYMBOL(mount_nodev); |
1da177e4c
|
1038 1039 1040 1041 |
static int compare_single(struct super_block *s, void *p) { return 1; } |
fc14f2fef
|
1042 |
struct dentry *mount_single(struct file_system_type *fs_type, |
1da177e4c
|
1043 |
int flags, void *data, |
fc14f2fef
|
1044 |
int (*fill_super)(struct super_block *, void *, int)) |
1da177e4c
|
1045 1046 1047 |
{ struct super_block *s; int error; |
9249e17fe
|
1048 |
s = sget(fs_type, compare_single, set_anon_super, flags, NULL); |
1da177e4c
|
1049 |
if (IS_ERR(s)) |
fc14f2fef
|
1050 |
return ERR_CAST(s); |
1da177e4c
|
1051 |
if (!s->s_root) { |
9b04c997b
|
1052 |
error = fill_super(s, data, flags & MS_SILENT ? 1 : 0); |
1da177e4c
|
1053 |
if (error) { |
74dbbdd7f
|
1054 |
deactivate_locked_super(s); |
fc14f2fef
|
1055 |
return ERR_PTR(error); |
1da177e4c
|
1056 1057 |
} s->s_flags |= MS_ACTIVE; |
9329d1bea
|
1058 1059 |
} else { do_remount_sb(s, flags, data, 0); |
1da177e4c
|
1060 |
} |
fc14f2fef
|
1061 1062 1063 |
return dget(s->s_root); } EXPORT_SYMBOL(mount_single); |
9d412a43c
|
1064 1065 |
struct dentry * mount_fs(struct file_system_type *type, int flags, const char *name, void *data) |
1da177e4c
|
1066 |
{ |
c96e41e92
|
1067 |
struct dentry *root; |
9d412a43c
|
1068 |
struct super_block *sb; |
1da177e4c
|
1069 |
char *secdata = NULL; |
9d412a43c
|
1070 |
int error = -ENOMEM; |
8089352a1
|
1071 |
|
e00075298
|
1072 |
if (data && !(type->fs_flags & FS_BINARY_MOUNTDATA)) { |
1da177e4c
|
1073 |
secdata = alloc_secdata(); |
454e2398b
|
1074 |
if (!secdata) |
9d412a43c
|
1075 |
goto out; |
1da177e4c
|
1076 |
|
e00075298
|
1077 |
error = security_sb_copy_data(data, secdata); |
454e2398b
|
1078 |
if (error) |
1da177e4c
|
1079 |
goto out_free_secdata; |
1da177e4c
|
1080 |
} |
1a102ff92
|
1081 1082 1083 1084 |
root = type->mount(type, flags, name, data); if (IS_ERR(root)) { error = PTR_ERR(root); goto out_free_secdata; |
c96e41e92
|
1085 |
} |
9d412a43c
|
1086 1087 1088 1089 |
sb = root->d_sb; BUG_ON(!sb); WARN_ON(!sb->s_bdi); sb->s_flags |= MS_BORN; |
454e2398b
|
1090 |
|
9d412a43c
|
1091 |
error = security_sb_kern_mount(sb, flags, secdata); |
5129a469a
|
1092 1093 |
if (error) goto out_sb; |
454e2398b
|
1094 |
|
42cb56ae2
|
1095 1096 1097 1098 |
/* * filesystems should never set s_maxbytes larger than MAX_LFS_FILESIZE * but s_maxbytes was an unsigned long long for many releases. Throw * this warning for a little while to try and catch filesystems that |
4358b5678
|
1099 |
* violate this rule. |
42cb56ae2
|
1100 |
*/ |
9d412a43c
|
1101 1102 1103 |
WARN((sb->s_maxbytes < 0), "%s set sb->s_maxbytes to " "negative value (%lld) ", type->name, sb->s_maxbytes); |
42cb56ae2
|
1104 |
|
9d412a43c
|
1105 |
up_write(&sb->s_umount); |
8680e22f2
|
1106 |
free_secdata(secdata); |
9d412a43c
|
1107 |
return root; |
1da177e4c
|
1108 |
out_sb: |
9d412a43c
|
1109 1110 |
dput(root); deactivate_locked_super(sb); |
1da177e4c
|
1111 1112 |
out_free_secdata: free_secdata(secdata); |
1da177e4c
|
1113 |
out: |
454e2398b
|
1114 |
return ERR_PTR(error); |
1da177e4c
|
1115 |
} |
5accdf82b
|
1116 1117 1118 1119 1120 1121 |
/* * This is an internal function, please use sb_end_{write,pagefault,intwrite} * instead. */ void __sb_end_write(struct super_block *sb, int level) { |
8129ed296
|
1122 |
percpu_up_read(sb->s_writers.rw_sem + level-1); |
5accdf82b
|
1123 1124 |
} EXPORT_SYMBOL(__sb_end_write); |
f4b554af9
|
1125 1126 1127 1128 1129 1130 1131 |
/* * This is an internal function, please use sb_start_{write,pagefault,intwrite} * instead. */ int __sb_start_write(struct super_block *sb, int level, bool wait) { bool force_trylock = false; |
8129ed296
|
1132 |
int ret = 1; |
f4b554af9
|
1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 |
#ifdef CONFIG_LOCKDEP /* * We want lockdep to tell us about possible deadlocks with freezing * but it's it bit tricky to properly instrument it. Getting a freeze * protection works as getting a read lock but there are subtle * problems. XFS for example gets freeze protection on internal level * twice in some cases, which is OK only because we already hold a * freeze protection also on higher level. Due to these cases we have * to use wait == F (trylock mode) which must not fail. */ if (wait) { int i; for (i = 0; i < level - 1; i++) |
8129ed296
|
1148 |
if (percpu_rwsem_is_held(sb->s_writers.rw_sem + i)) { |
f4b554af9
|
1149 1150 1151 1152 1153 |
force_trylock = true; break; } } #endif |
8129ed296
|
1154 1155 1156 1157 |
if (wait && !force_trylock) percpu_down_read(sb->s_writers.rw_sem + level-1); else ret = percpu_down_read_trylock(sb->s_writers.rw_sem + level-1); |
f4b554af9
|
1158 1159 1160 |
WARN_ON(force_trylock & !ret); return ret; } |
5accdf82b
|
1161 1162 1163 1164 1165 1166 1167 1168 |
EXPORT_SYMBOL(__sb_start_write); /** * sb_wait_write - wait until all writers to given file system finish * @sb: the super for which we wait * @level: type of writers we wait for (normal vs page fault) * * This function waits until there are no writers of given type to given file |
8129ed296
|
1169 |
* system. |
5accdf82b
|
1170 1171 1172 |
*/ static void sb_wait_write(struct super_block *sb, int level) { |
8129ed296
|
1173 |
percpu_down_write(sb->s_writers.rw_sem + level-1); |
5accdf82b
|
1174 |
/* |
0e28e01f1
|
1175 1176 1177 1178 1179 1180 1181 1182 |
* We are going to return to userspace and forget about this lock, the * ownership goes to the caller of thaw_super() which does unlock. * * FIXME: we should do this before return from freeze_super() after we * called sync_filesystem(sb) and s_op->freeze_fs(sb), and thaw_super() * should re-acquire these locks before s_op->unfreeze_fs(sb). However * this leads to lockdep false-positives, so currently we do the early * release right after acquire. |
5accdf82b
|
1183 |
*/ |
8129ed296
|
1184 1185 |
percpu_rwsem_release(sb->s_writers.rw_sem + level-1, 0, _THIS_IP_); } |
5accdf82b
|
1186 |
|
8129ed296
|
1187 1188 1189 |
static void sb_freeze_unlock(struct super_block *sb) { int level; |
5accdf82b
|
1190 |
|
8129ed296
|
1191 1192 |
for (level = 0; level < SB_FREEZE_LEVELS; ++level) percpu_rwsem_acquire(sb->s_writers.rw_sem + level, 0, _THIS_IP_); |
5accdf82b
|
1193 |
|
8129ed296
|
1194 1195 |
for (level = SB_FREEZE_LEVELS - 1; level >= 0; level--) percpu_up_write(sb->s_writers.rw_sem + level); |
5accdf82b
|
1196 |
} |
18e9e5104
|
1197 |
/** |
7000d3c42
|
1198 1199 |
* freeze_super - lock the filesystem and force it into a consistent state * @sb: the super to lock |
18e9e5104
|
1200 1201 1202 1203 |
* * Syncs the super to make sure the filesystem is consistent and calls the fs's * freeze_fs. Subsequent calls to this without first thawing the fs will return * -EBUSY. |
5accdf82b
|
1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 |
* * During this function, sb->s_writers.frozen goes through these values: * * SB_UNFROZEN: File system is normal, all writes progress as usual. * * SB_FREEZE_WRITE: The file system is in the process of being frozen. New * writes should be blocked, though page faults are still allowed. We wait for * all writes to complete and then proceed to the next stage. * * SB_FREEZE_PAGEFAULT: Freezing continues. Now also page faults are blocked * but internal fs threads can still modify the filesystem (although they * should not dirty new pages or inodes), writeback can run etc. After waiting * for all running page faults we sync the filesystem which will clean all * dirty pages and inodes (no new dirty pages or inodes can be created when * sync is running). * * SB_FREEZE_FS: The file system is frozen. Now all internal sources of fs * modification are blocked (e.g. XFS preallocation truncation on inode * reclaim). This is usually implemented by blocking new transactions for * filesystems that have them and need this additional guard. After all * internal writers are finished we call ->freeze_fs() to finish filesystem * freezing. Then we transition to SB_FREEZE_COMPLETE state. This state is * mostly auxiliary for filesystems to verify they do not modify frozen fs. * * sb->s_writers.frozen is protected by sb->s_umount. |
18e9e5104
|
1229 1230 1231 1232 1233 1234 1235 |
*/ int freeze_super(struct super_block *sb) { int ret; atomic_inc(&sb->s_active); down_write(&sb->s_umount); |
5accdf82b
|
1236 |
if (sb->s_writers.frozen != SB_UNFROZEN) { |
18e9e5104
|
1237 1238 1239 |
deactivate_locked_super(sb); return -EBUSY; } |
dabe0dc19
|
1240 1241 1242 1243 |
if (!(sb->s_flags & MS_BORN)) { up_write(&sb->s_umount); return 0; /* sic - it's "nothing to do" */ } |
18e9e5104
|
1244 |
if (sb->s_flags & MS_RDONLY) { |
5accdf82b
|
1245 1246 |
/* Nothing to do really... */ sb->s_writers.frozen = SB_FREEZE_COMPLETE; |
18e9e5104
|
1247 1248 1249 |
up_write(&sb->s_umount); return 0; } |
5accdf82b
|
1250 |
sb->s_writers.frozen = SB_FREEZE_WRITE; |
5accdf82b
|
1251 1252 |
/* Release s_umount to preserve sb_start_write -> s_umount ordering */ up_write(&sb->s_umount); |
5accdf82b
|
1253 |
sb_wait_write(sb, SB_FREEZE_WRITE); |
8129ed296
|
1254 |
down_write(&sb->s_umount); |
5accdf82b
|
1255 1256 |
/* Now we go and block page faults... */ |
5accdf82b
|
1257 |
sb->s_writers.frozen = SB_FREEZE_PAGEFAULT; |
5accdf82b
|
1258 1259 1260 |
sb_wait_write(sb, SB_FREEZE_PAGEFAULT); /* All writers are done so after syncing there won't be dirty data */ |
18e9e5104
|
1261 |
sync_filesystem(sb); |
5accdf82b
|
1262 1263 |
/* Now wait for internal filesystem counter */ sb->s_writers.frozen = SB_FREEZE_FS; |
5accdf82b
|
1264 |
sb_wait_write(sb, SB_FREEZE_FS); |
18e9e5104
|
1265 |
|
18e9e5104
|
1266 1267 1268 1269 1270 1271 |
if (sb->s_op->freeze_fs) { ret = sb->s_op->freeze_fs(sb); if (ret) { printk(KERN_ERR "VFS:Filesystem freeze failed "); |
5accdf82b
|
1272 |
sb->s_writers.frozen = SB_UNFROZEN; |
8129ed296
|
1273 |
sb_freeze_unlock(sb); |
5accdf82b
|
1274 |
wake_up(&sb->s_writers.wait_unfrozen); |
18e9e5104
|
1275 1276 1277 1278 |
deactivate_locked_super(sb); return ret; } } |
5accdf82b
|
1279 1280 1281 1282 1283 |
/* * This is just for debugging purposes so that fs can warn if it * sees write activity when frozen is set to SB_FREEZE_COMPLETE. */ sb->s_writers.frozen = SB_FREEZE_COMPLETE; |
18e9e5104
|
1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 |
up_write(&sb->s_umount); return 0; } EXPORT_SYMBOL(freeze_super); /** * thaw_super -- unlock filesystem * @sb: the super to thaw * * Unlocks the filesystem and marks it writeable again after freeze_super(). */ int thaw_super(struct super_block *sb) { int error; down_write(&sb->s_umount); |
5accdf82b
|
1300 |
if (sb->s_writers.frozen == SB_UNFROZEN) { |
18e9e5104
|
1301 1302 1303 |
up_write(&sb->s_umount); return -EINVAL; } |
8129ed296
|
1304 1305 |
if (sb->s_flags & MS_RDONLY) { sb->s_writers.frozen = SB_UNFROZEN; |
18e9e5104
|
1306 |
goto out; |
8129ed296
|
1307 |
} |
18e9e5104
|
1308 1309 1310 1311 1312 1313 1314 |
if (sb->s_op->unfreeze_fs) { error = sb->s_op->unfreeze_fs(sb); if (error) { printk(KERN_ERR "VFS:Filesystem thaw failed "); |
18e9e5104
|
1315 1316 1317 1318 |
up_write(&sb->s_umount); return error; } } |
5accdf82b
|
1319 |
sb->s_writers.frozen = SB_UNFROZEN; |
8129ed296
|
1320 1321 |
sb_freeze_unlock(sb); out: |
5accdf82b
|
1322 |
wake_up(&sb->s_writers.wait_unfrozen); |
18e9e5104
|
1323 |
deactivate_locked_super(sb); |
18e9e5104
|
1324 1325 1326 |
return 0; } EXPORT_SYMBOL(thaw_super); |