Blame view
fs/super.c
38.8 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> |
6e4eab577
|
35 |
#include <linux/user_namespace.h> |
6d59e7f58
|
36 |
#include "internal.h" |
1da177e4c
|
37 |
|
15d0f5ea3
|
38 39 |
static LIST_HEAD(super_blocks); static DEFINE_SPINLOCK(sb_lock); |
1da177e4c
|
40 |
|
5accdf82b
|
41 42 43 44 45 |
static char *sb_writers_name[SB_FREEZE_LEVELS] = { "sb_writers", "sb_pagefaults", "sb_internal", }; |
b0d40c92a
|
46 47 48 49 50 51 52 |
/* * 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
|
53 54 |
static unsigned long super_cache_scan(struct shrinker *shrink, struct shrink_control *sc) |
b0d40c92a
|
55 56 |
{ struct super_block *sb; |
0a234c6dc
|
57 58 59 60 61 |
long fs_objects = 0; long total_objects; long freed = 0; long dentries; long inodes; |
b0d40c92a
|
62 63 64 65 66 67 68 |
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
|
69 70 |
if (!(sc->gfp_mask & __GFP_FS)) return SHRINK_STOP; |
b0d40c92a
|
71 |
|
eb6ef3df4
|
72 |
if (!trylock_super(sb)) |
0a234c6dc
|
73 |
return SHRINK_STOP; |
b0d40c92a
|
74 |
|
d04079039
|
75 |
if (sb->s_op->nr_cached_objects) |
4101b6243
|
76 |
fs_objects = sb->s_op->nr_cached_objects(sb, sc); |
0e1fdafd9
|
77 |
|
503c358cf
|
78 79 |
inodes = list_lru_shrink_count(&sb->s_inode_lru, sc); dentries = list_lru_shrink_count(&sb->s_dentry_lru, sc); |
f60415675
|
80 |
total_objects = dentries + inodes + fs_objects + 1; |
475d0db74
|
81 82 |
if (!total_objects) total_objects = 1; |
0e1fdafd9
|
83 |
|
0a234c6dc
|
84 |
/* proportion the scan between the caches */ |
f60415675
|
85 |
dentries = mult_frac(sc->nr_to_scan, dentries, total_objects); |
bc3b14cb2
|
86 |
inodes = mult_frac(sc->nr_to_scan, inodes, total_objects); |
503c358cf
|
87 |
fs_objects = mult_frac(sc->nr_to_scan, fs_objects, total_objects); |
b0d40c92a
|
88 |
|
0a234c6dc
|
89 90 91 |
/* * prune the dcache first as the icache is pinned by it, then * prune the icache, followed by the filesystem specific caches |
49e7e7ff8
|
92 93 94 |
* * Ensure that we always scan at least one object - memcg kmem * accounting uses this to fully empty the caches. |
0a234c6dc
|
95 |
*/ |
49e7e7ff8
|
96 |
sc->nr_to_scan = dentries + 1; |
503c358cf
|
97 |
freed = prune_dcache_sb(sb, sc); |
49e7e7ff8
|
98 |
sc->nr_to_scan = inodes + 1; |
503c358cf
|
99 |
freed += prune_icache_sb(sb, sc); |
0a234c6dc
|
100 101 |
if (fs_objects) { |
49e7e7ff8
|
102 |
sc->nr_to_scan = fs_objects + 1; |
4101b6243
|
103 |
freed += sb->s_op->free_cached_objects(sb, sc); |
b0d40c92a
|
104 |
} |
eb6ef3df4
|
105 |
up_read(&sb->s_umount); |
0a234c6dc
|
106 107 108 109 110 111 112 113 114 115 |
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
|
116 |
/* |
eb6ef3df4
|
117 |
* Don't call trylock_super as it is a potential |
d23da150a
|
118 119 120 |
* 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
|
121 |
* ensures the safety of call to list_lru_shrink_count() and |
d23da150a
|
122 123 |
* s_op->nr_cached_objects(). */ |
0a234c6dc
|
124 |
if (sb->s_op && sb->s_op->nr_cached_objects) |
4101b6243
|
125 |
total_objects = sb->s_op->nr_cached_objects(sb, sc); |
0a234c6dc
|
126 |
|
503c358cf
|
127 128 |
total_objects += list_lru_shrink_count(&sb->s_dentry_lru, sc); total_objects += list_lru_shrink_count(&sb->s_inode_lru, sc); |
0a234c6dc
|
129 |
|
55f841ce9
|
130 |
total_objects = vfs_pressure_ratio(total_objects); |
0e1fdafd9
|
131 |
return total_objects; |
b0d40c92a
|
132 |
} |
853b39a7c
|
133 134 135 136 137 138 139 |
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
|
140 |
percpu_free_rwsem(&s->s_writers.rw_sem[i]); |
853b39a7c
|
141 142 143 144 145 146 147 148 149 |
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
|
150 151 152 153 154 155 156 |
/** * destroy_super - frees a superblock * @s: superblock to free * * Frees a superblock. */ static void destroy_super(struct super_block *s) |
5accdf82b
|
157 |
{ |
7eb5e8826
|
158 159 |
list_lru_destroy(&s->s_dentry_lru); list_lru_destroy(&s->s_inode_lru); |
7eb5e8826
|
160 161 |
security_sb_free(s); WARN_ON(!list_empty(&s->s_mounts)); |
6e4eab577
|
162 |
put_user_ns(s->s_user_ns); |
7eb5e8826
|
163 |
kfree(s->s_subtype); |
853b39a7c
|
164 |
call_rcu(&s->rcu, destroy_super_rcu); |
5accdf82b
|
165 |
} |
1da177e4c
|
166 167 |
/** * alloc_super - create new superblock |
fe2bbc483
|
168 |
* @type: filesystem type superblock should belong to |
9249e17fe
|
169 |
* @flags: the mount flags |
6e4eab577
|
170 |
* @user_ns: User namespace for the super_block |
1da177e4c
|
171 172 173 174 |
* * Allocates and initializes a new &struct super_block. alloc_super() * returns a pointer new superblock or %NULL if allocation had failed. */ |
6e4eab577
|
175 176 |
static struct super_block *alloc_super(struct file_system_type *type, int flags, struct user_namespace *user_ns) |
1da177e4c
|
177 |
{ |
11b0b5abb
|
178 |
struct super_block *s = kzalloc(sizeof(struct super_block), GFP_USER); |
b87221de6
|
179 |
static const struct super_operations default_op; |
7eb5e8826
|
180 181 182 183 |
int i; if (!s) return NULL; |
1da177e4c
|
184 |
|
b5bd856a0
|
185 |
INIT_LIST_HEAD(&s->s_mounts); |
6e4eab577
|
186 |
s->s_user_ns = get_user_ns(user_ns); |
b5bd856a0
|
187 |
|
7eb5e8826
|
188 189 |
if (security_sb_alloc(s)) goto fail; |
7b7a8665e
|
190 |
|
7eb5e8826
|
191 |
for (i = 0; i < SB_FREEZE_LEVELS; i++) { |
8129ed296
|
192 193 194 |
if (__percpu_init_rwsem(&s->s_writers.rw_sem[i], sb_writers_name[i], &type->s_writers_key[i])) |
7eb5e8826
|
195 |
goto fail; |
1da177e4c
|
196 |
} |
7eb5e8826
|
197 |
init_waitqueue_head(&s->s_writers.wait_unfrozen); |
df0ce26cb
|
198 |
s->s_bdi = &noop_backing_dev_info; |
7eb5e8826
|
199 |
s->s_flags = flags; |
cc50a07a2
|
200 |
if (s->s_user_ns != &init_user_ns) |
67690f937
|
201 |
s->s_iflags |= SB_I_NODEV; |
7eb5e8826
|
202 203 |
INIT_HLIST_NODE(&s->s_instances); INIT_HLIST_BL_HEAD(&s->s_anon); |
e97fedb9e
|
204 |
mutex_init(&s->s_sync_lock); |
7eb5e8826
|
205 |
INIT_LIST_HEAD(&s->s_inodes); |
74278da9f
|
206 |
spin_lock_init(&s->s_inode_list_lock); |
6c60d2b57
|
207 208 |
INIT_LIST_HEAD(&s->s_inodes_wb); spin_lock_init(&s->s_inode_wblist_lock); |
7eb5e8826
|
209 |
|
2acb60a04
|
210 |
if (list_lru_init_memcg(&s->s_dentry_lru)) |
7eb5e8826
|
211 |
goto fail; |
2acb60a04
|
212 |
if (list_lru_init_memcg(&s->s_inode_lru)) |
7eb5e8826
|
213 |
goto fail; |
7eb5e8826
|
214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 |
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); |
bc8230ee8
|
236 |
init_rwsem(&s->s_dquot.dqio_sem); |
7eb5e8826
|
237 238 239 |
s->s_maxbytes = MAX_NON_LFS; s->s_op = &default_op; s->s_time_gran = 1000000000; |
3cb29d111
|
240 |
s->cleancache_poolid = CLEANCACHE_NO_POOL; |
7eb5e8826
|
241 242 243 244 245 |
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
|
246 |
s->s_shrink.flags = SHRINKER_NUMA_AWARE | SHRINKER_MEMCG_AWARE; |
1da177e4c
|
247 |
return s; |
5ca302c8e
|
248 |
|
7eb5e8826
|
249 250 251 |
fail: destroy_super(s); return NULL; |
1da177e4c
|
252 253 254 255 256 |
} /* Superblock refcounting */ /* |
35cf7ba0b
|
257 |
* Drop a superblock's refcount. The caller must hold sb_lock. |
1da177e4c
|
258 |
*/ |
f47ec3f28
|
259 |
static void __put_super(struct super_block *sb) |
1da177e4c
|
260 |
{ |
1da177e4c
|
261 |
if (!--sb->s_count) { |
551de6f34
|
262 |
list_del_init(&sb->s_list); |
1da177e4c
|
263 |
destroy_super(sb); |
1da177e4c
|
264 |
} |
1da177e4c
|
265 266 267 268 269 270 271 272 273 |
} /** * 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
|
274 |
static void put_super(struct super_block *sb) |
1da177e4c
|
275 276 277 278 279 280 281 282 |
{ spin_lock(&sb_lock); __put_super(sb); spin_unlock(&sb_lock); } /** |
1712ac8fd
|
283 |
* deactivate_locked_super - drop an active reference to superblock |
1da177e4c
|
284 285 |
* @s: superblock to deactivate * |
bd7ced988
|
286 |
* Drops an active reference to superblock, converting it into a temporary |
1712ac8fd
|
287 |
* one if there is no other active references left. In that case we |
1da177e4c
|
288 289 |
* tell fs driver to shut it down and drop the temporary reference we * had just acquired. |
1712ac8fd
|
290 291 |
* * Caller holds exclusive lock on superblock; that lock is released. |
1da177e4c
|
292 |
*/ |
1712ac8fd
|
293 |
void deactivate_locked_super(struct super_block *s) |
1da177e4c
|
294 295 |
{ struct file_system_type *fs = s->s_type; |
b20bd1a5e
|
296 |
if (atomic_dec_and_test(&s->s_active)) { |
3167760f8
|
297 |
cleancache_invalidate_fs(s); |
b0d40c92a
|
298 |
unregister_shrinker(&s->s_shrink); |
28f2cd4f6
|
299 |
fs->kill_sb(s); |
f5e1dd345
|
300 |
|
c0a5b5609
|
301 302 303 304 305 306 307 |
/* * 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
|
308 309 |
put_filesystem(fs); put_super(s); |
1712ac8fd
|
310 311 |
} else { up_write(&s->s_umount); |
1da177e4c
|
312 313 |
} } |
1712ac8fd
|
314 |
EXPORT_SYMBOL(deactivate_locked_super); |
1da177e4c
|
315 316 |
/** |
1712ac8fd
|
317 |
* deactivate_super - drop an active reference to superblock |
74dbbdd7f
|
318 319 |
* @s: superblock to deactivate * |
1712ac8fd
|
320 321 322 |
* 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
|
323 |
*/ |
1712ac8fd
|
324 |
void deactivate_super(struct super_block *s) |
74dbbdd7f
|
325 |
{ |
1712ac8fd
|
326 327 328 |
if (!atomic_add_unless(&s->s_active, -1, 1)) { down_write(&s->s_umount); deactivate_locked_super(s); |
74dbbdd7f
|
329 330 |
} } |
1712ac8fd
|
331 |
EXPORT_SYMBOL(deactivate_super); |
74dbbdd7f
|
332 333 |
/** |
1da177e4c
|
334 335 336 337 338 339 340 341 |
* 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
|
342 343 344 |
* 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
|
345 |
*/ |
9c4dbee79
|
346 |
static int grab_super(struct super_block *s) __releases(sb_lock) |
1da177e4c
|
347 348 349 350 |
{ s->s_count++; spin_unlock(&sb_lock); down_write(&s->s_umount); |
e462ec50c
|
351 |
if ((s->s_flags & SB_BORN) && atomic_inc_not_zero(&s->s_active)) { |
acfec9a5a
|
352 353 354 |
put_super(s); return 1; } |
1da177e4c
|
355 356 |
up_write(&s->s_umount); put_super(s); |
1da177e4c
|
357 358 |
return 0; } |
cf9a2ae8d
|
359 |
/* |
eb6ef3df4
|
360 |
* trylock_super - try to grab ->s_umount shared |
331cbdeed
|
361 |
* @sb: reference we are trying to grab |
12ad3ab66
|
362 |
* |
eb6ef3df4
|
363 |
* Try to prevent fs shutdown. This is used in places where we |
12ad3ab66
|
364 |
* cannot take an active reference but we need to ensure that the |
eb6ef3df4
|
365 366 367 368 369 370 371 372 373 374 |
* 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
|
375 |
*/ |
eb6ef3df4
|
376 |
bool trylock_super(struct super_block *sb) |
12ad3ab66
|
377 |
{ |
12ad3ab66
|
378 |
if (down_read_trylock(&sb->s_umount)) { |
eb6ef3df4
|
379 |
if (!hlist_unhashed(&sb->s_instances) && |
e462ec50c
|
380 |
sb->s_root && (sb->s_flags & SB_BORN)) |
12ad3ab66
|
381 382 383 |
return true; up_read(&sb->s_umount); } |
12ad3ab66
|
384 385 |
return false; } |
1da177e4c
|
386 387 388 389 390 391 392 393 394 |
/** * 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
|
395 396 397 398 |
* * 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
|
399 400 401 |
*/ void generic_shutdown_super(struct super_block *sb) { |
ee9b6d61a
|
402 |
const struct super_operations *sop = sb->s_op; |
1da177e4c
|
403 |
|
c636ebdb1
|
404 405 |
if (sb->s_root) { shrink_dcache_for_umount(sb); |
60b0680fa
|
406 |
sync_filesystem(sb); |
e462ec50c
|
407 |
sb->s_flags &= ~SB_ACTIVE; |
efaee1920
|
408 |
|
74278da9f
|
409 |
fsnotify_unmount_inodes(sb); |
a1a0e23e4
|
410 |
cgroup_writeback_umount(); |
63997e98a
|
411 412 |
evict_inodes(sb); |
1da177e4c
|
413 |
|
7b7a8665e
|
414 415 416 417 |
if (sb->s_dio_done_wq) { destroy_workqueue(sb->s_dio_done_wq); sb->s_dio_done_wq = NULL; } |
1da177e4c
|
418 419 |
if (sop->put_super) sop->put_super(sb); |
63997e98a
|
420 |
if (!list_empty(&sb->s_inodes)) { |
7b4fe29e0
|
421 422 423 424 |
printk("VFS: Busy inodes after unmount of %s. " "Self-destruct in 5 seconds. Have a nice day... ", sb->s_id); |
1da177e4c
|
425 |
} |
1da177e4c
|
426 427 428 |
} spin_lock(&sb_lock); /* should be initialized for __put_super_and_need_restart() */ |
a5166169f
|
429 |
hlist_del_init(&sb->s_instances); |
1da177e4c
|
430 431 |
spin_unlock(&sb_lock); up_write(&sb->s_umount); |
c1844d536
|
432 |
if (sb->s_bdi != &noop_backing_dev_info) { |
fca39346a
|
433 434 |
bdi_put(sb->s_bdi); sb->s_bdi = &noop_backing_dev_info; |
fca39346a
|
435 |
} |
1da177e4c
|
436 437 438 439 440 |
} EXPORT_SYMBOL(generic_shutdown_super); /** |
6e4eab577
|
441 |
* sget_userns - find or create a superblock |
1da177e4c
|
442 443 444 |
* @type: filesystem type superblock should belong to * @test: comparison callback * @set: setup callback |
9249e17fe
|
445 |
* @flags: mount flags |
6e4eab577
|
446 |
* @user_ns: User namespace for the super_block |
1da177e4c
|
447 448 |
* @data: argument to each of them */ |
6e4eab577
|
449 |
struct super_block *sget_userns(struct file_system_type *type, |
1da177e4c
|
450 451 |
int (*test)(struct super_block *,void *), int (*set)(struct super_block *,void *), |
6e4eab577
|
452 |
int flags, struct user_namespace *user_ns, |
1da177e4c
|
453 454 455 |
void *data) { struct super_block *s = NULL; |
d47301271
|
456 |
struct super_block *old; |
1da177e4c
|
457 |
int err; |
e462ec50c
|
458 |
if (!(flags & (SB_KERNMOUNT|SB_SUBMOUNT)) && |
a001e74ce
|
459 460 461 |
!(type->fs_flags & FS_USERNS_MOUNT) && !capable(CAP_SYS_ADMIN)) return ERR_PTR(-EPERM); |
1da177e4c
|
462 463 |
retry: spin_lock(&sb_lock); |
d47301271
|
464 |
if (test) { |
b67bfe0d4
|
465 |
hlist_for_each_entry(old, &type->fs_supers, s_instances) { |
d47301271
|
466 467 |
if (!test(old, data)) continue; |
6e4eab577
|
468 469 470 471 472 473 474 475 |
if (user_ns != old->s_user_ns) { spin_unlock(&sb_lock); if (s) { up_write(&s->s_umount); destroy_super(s); } return ERR_PTR(-EBUSY); } |
d47301271
|
476 477 |
if (!grab_super(old)) goto retry; |
a3cfbb53b
|
478 479 |
if (s) { up_write(&s->s_umount); |
d47301271
|
480 |
destroy_super(s); |
7a4dec538
|
481 |
s = NULL; |
a3cfbb53b
|
482 |
} |
d47301271
|
483 484 |
return old; } |
1da177e4c
|
485 486 487 |
} if (!s) { spin_unlock(&sb_lock); |
e462ec50c
|
488 |
s = alloc_super(type, (flags & ~SB_SUBMOUNT), user_ns); |
1da177e4c
|
489 490 491 492 |
if (!s) return ERR_PTR(-ENOMEM); goto retry; } |
dd111b31e
|
493 |
|
1da177e4c
|
494 495 496 |
err = set(s, data); if (err) { spin_unlock(&sb_lock); |
a3cfbb53b
|
497 |
up_write(&s->s_umount); |
1da177e4c
|
498 499 500 501 502 503 |
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
|
504 |
hlist_add_head(&s->s_instances, &type->fs_supers); |
1da177e4c
|
505 506 |
spin_unlock(&sb_lock); get_filesystem(type); |
b0d40c92a
|
507 |
register_shrinker(&s->s_shrink); |
1da177e4c
|
508 509 |
return s; } |
6e4eab577
|
510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 |
EXPORT_SYMBOL(sget_userns); /** * sget - find or create a superblock * @type: filesystem type superblock should belong to * @test: comparison callback * @set: setup callback * @flags: mount flags * @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 *), int flags, void *data) { struct user_namespace *user_ns = current_user_ns(); |
93faccbbf
|
527 528 529 530 |
/* We don't yet pass the user namespace of the parent * mount through to here so always use &init_user_ns * until that changes. */ |
e462ec50c
|
531 |
if (flags & SB_SUBMOUNT) |
93faccbbf
|
532 |
user_ns = &init_user_ns; |
6e4eab577
|
533 |
/* Ensure the requestor has permissions over the target filesystem */ |
e462ec50c
|
534 |
if (!(flags & (SB_KERNMOUNT|SB_SUBMOUNT)) && !ns_capable(user_ns, CAP_SYS_ADMIN)) |
6e4eab577
|
535 536 537 538 |
return ERR_PTR(-EPERM); return sget_userns(type, test, set, flags, user_ns, data); } |
1da177e4c
|
539 540 541 542 543 544 545 546 547 |
EXPORT_SYMBOL(sget); void drop_super(struct super_block *sb) { up_read(&sb->s_umount); put_super(sb); } EXPORT_SYMBOL(drop_super); |
ba6379f7e
|
548 549 550 551 552 553 |
void drop_super_exclusive(struct super_block *sb) { up_write(&sb->s_umount); put_super(sb); } EXPORT_SYMBOL(drop_super_exclusive); |
e50047533
|
554 |
/** |
01a05b337
|
555 556 557 558 559 560 561 562 563 |
* 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
|
564 |
struct super_block *sb, *p = NULL; |
01a05b337
|
565 566 |
spin_lock(&sb_lock); |
dca332528
|
567 |
list_for_each_entry(sb, &super_blocks, s_list) { |
a5166169f
|
568 |
if (hlist_unhashed(&sb->s_instances)) |
01a05b337
|
569 570 571 572 573 |
continue; sb->s_count++; spin_unlock(&sb_lock); down_read(&sb->s_umount); |
e462ec50c
|
574 |
if (sb->s_root && (sb->s_flags & SB_BORN)) |
01a05b337
|
575 576 577 578 |
f(sb, arg); up_read(&sb->s_umount); spin_lock(&sb_lock); |
dca332528
|
579 580 581 |
if (p) __put_super(p); p = sb; |
01a05b337
|
582 |
} |
dca332528
|
583 584 |
if (p) __put_super(p); |
01a05b337
|
585 586 587 588 |
spin_unlock(&sb_lock); } /** |
43e15cdbe
|
589 590 591 592 593 594 595 596 597 598 599 600 601 602 |
* 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
|
603 |
hlist_for_each_entry(sb, &type->fs_supers, s_instances) { |
43e15cdbe
|
604 605 606 607 |
sb->s_count++; spin_unlock(&sb_lock); down_read(&sb->s_umount); |
e462ec50c
|
608 |
if (sb->s_root && (sb->s_flags & SB_BORN)) |
43e15cdbe
|
609 610 611 612 613 614 615 616 617 618 619 620 621 622 |
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); |
ba6379f7e
|
623 |
static struct super_block *__get_super(struct block_device *bdev, bool excl) |
1da177e4c
|
624 |
{ |
618f06362
|
625 |
struct super_block *sb; |
1da177e4c
|
626 627 |
if (!bdev) return NULL; |
618f06362
|
628 |
|
1da177e4c
|
629 |
spin_lock(&sb_lock); |
618f06362
|
630 631 |
rescan: list_for_each_entry(sb, &super_blocks, s_list) { |
a5166169f
|
632 |
if (hlist_unhashed(&sb->s_instances)) |
551de6f34
|
633 |
continue; |
618f06362
|
634 635 |
if (sb->s_bdev == bdev) { sb->s_count++; |
1da177e4c
|
636 |
spin_unlock(&sb_lock); |
ba6379f7e
|
637 638 639 640 |
if (!excl) down_read(&sb->s_umount); else down_write(&sb->s_umount); |
df40c01a9
|
641 |
/* still alive? */ |
e462ec50c
|
642 |
if (sb->s_root && (sb->s_flags & SB_BORN)) |
618f06362
|
643 |
return sb; |
ba6379f7e
|
644 645 646 647 |
if (!excl) up_read(&sb->s_umount); else up_write(&sb->s_umount); |
df40c01a9
|
648 |
/* nope, got unmounted */ |
618f06362
|
649 |
spin_lock(&sb_lock); |
df40c01a9
|
650 651 |
__put_super(sb); goto rescan; |
1da177e4c
|
652 653 654 655 656 |
} } spin_unlock(&sb_lock); return NULL; } |
4504230a7
|
657 |
/** |
ba6379f7e
|
658 |
* get_super - get the superblock of a device |
6b6dc836a
|
659 660 661 |
* @bdev: device to get the superblock for * * Scans the superblock list and finds the superblock of the file system |
ba6379f7e
|
662 |
* mounted on the device given. %NULL is returned if no match is found. |
6b6dc836a
|
663 |
*/ |
ba6379f7e
|
664 665 666 667 668 669 670 671 |
struct super_block *get_super(struct block_device *bdev) { return __get_super(bdev, false); } EXPORT_SYMBOL(get_super); static struct super_block *__get_super_thawed(struct block_device *bdev, bool excl) |
6b6dc836a
|
672 673 |
{ while (1) { |
ba6379f7e
|
674 |
struct super_block *s = __get_super(bdev, excl); |
5accdf82b
|
675 |
if (!s || s->s_writers.frozen == SB_UNFROZEN) |
6b6dc836a
|
676 |
return s; |
ba6379f7e
|
677 678 679 680 |
if (!excl) up_read(&s->s_umount); else up_write(&s->s_umount); |
5accdf82b
|
681 682 |
wait_event(s->s_writers.wait_unfrozen, s->s_writers.frozen == SB_UNFROZEN); |
6b6dc836a
|
683 684 685 |
put_super(s); } } |
ba6379f7e
|
686 687 688 689 690 691 692 693 694 695 696 697 698 699 |
/** * 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) { return __get_super_thawed(bdev, false); } |
6b6dc836a
|
700 701 702 |
EXPORT_SYMBOL(get_super_thawed); /** |
ba6379f7e
|
703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 |
* get_super_exclusive_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) and s_umount semaphore is held * in exclusive mode. %NULL is returned if no match is found. */ struct super_block *get_super_exclusive_thawed(struct block_device *bdev) { return __get_super_thawed(bdev, true); } EXPORT_SYMBOL(get_super_exclusive_thawed); /** |
4504230a7
|
718 719 720 721 722 |
* 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
|
723 |
* reference or %NULL if none was found. |
4504230a7
|
724 725 726 727 728 729 730 |
*/ struct super_block *get_active_super(struct block_device *bdev) { struct super_block *sb; if (!bdev) return NULL; |
1494583de
|
731 |
restart: |
4504230a7
|
732 733 |
spin_lock(&sb_lock); list_for_each_entry(sb, &super_blocks, s_list) { |
a5166169f
|
734 |
if (hlist_unhashed(&sb->s_instances)) |
551de6f34
|
735 |
continue; |
1494583de
|
736 |
if (sb->s_bdev == bdev) { |
acfec9a5a
|
737 |
if (!grab_super(sb)) |
1494583de
|
738 |
goto restart; |
acfec9a5a
|
739 740 |
up_write(&sb->s_umount); return sb; |
1494583de
|
741 |
} |
4504230a7
|
742 743 744 745 |
} spin_unlock(&sb_lock); return NULL; } |
dd111b31e
|
746 |
|
df40c01a9
|
747 |
struct super_block *user_get_super(dev_t dev) |
1da177e4c
|
748 |
{ |
618f06362
|
749 |
struct super_block *sb; |
1da177e4c
|
750 |
|
1da177e4c
|
751 |
spin_lock(&sb_lock); |
618f06362
|
752 753 |
rescan: list_for_each_entry(sb, &super_blocks, s_list) { |
a5166169f
|
754 |
if (hlist_unhashed(&sb->s_instances)) |
551de6f34
|
755 |
continue; |
618f06362
|
756 757 |
if (sb->s_dev == dev) { sb->s_count++; |
1da177e4c
|
758 |
spin_unlock(&sb_lock); |
618f06362
|
759 |
down_read(&sb->s_umount); |
df40c01a9
|
760 |
/* still alive? */ |
e462ec50c
|
761 |
if (sb->s_root && (sb->s_flags & SB_BORN)) |
618f06362
|
762 763 |
return sb; up_read(&sb->s_umount); |
df40c01a9
|
764 |
/* nope, got unmounted */ |
618f06362
|
765 |
spin_lock(&sb_lock); |
df40c01a9
|
766 767 |
__put_super(sb); goto rescan; |
1da177e4c
|
768 769 770 771 772 |
} } spin_unlock(&sb_lock); return NULL; } |
1da177e4c
|
773 |
/** |
1da177e4c
|
774 775 |
* do_remount_sb - asks filesystem to change mount options. * @sb: superblock in question |
e462ec50c
|
776 |
* @sb_flags: revised superblock flags |
1da177e4c
|
777 778 779 780 781 |
* @data: the rest of options * @force: whether or not to force the change * * Alters the mount options of a mounted file system. */ |
e462ec50c
|
782 |
int do_remount_sb(struct super_block *sb, int sb_flags, void *data, int force) |
1da177e4c
|
783 784 |
{ int retval; |
c79d967de
|
785 |
int remount_ro; |
4504230a7
|
786 |
|
5accdf82b
|
787 |
if (sb->s_writers.frozen != SB_UNFROZEN) |
4504230a7
|
788 |
return -EBUSY; |
9361401eb
|
789 |
#ifdef CONFIG_BLOCK |
e462ec50c
|
790 |
if (!(sb_flags & SB_RDONLY) && bdev_read_only(sb->s_bdev)) |
1da177e4c
|
791 |
return -EACCES; |
9361401eb
|
792 |
#endif |
4504230a7
|
793 |
|
e462ec50c
|
794 |
remount_ro = (sb_flags & SB_RDONLY) && !sb_rdonly(sb); |
d208bbdda
|
795 |
|
0aec09d04
|
796 |
if (remount_ro) { |
fdab684d7
|
797 |
if (!hlist_empty(&sb->s_pins)) { |
0aec09d04
|
798 |
up_write(&sb->s_umount); |
fdab684d7
|
799 |
group_pin_kill(&sb->s_pins); |
0aec09d04
|
800 801 802 803 804 |
down_write(&sb->s_umount); if (!sb->s_root) return 0; if (sb->s_writers.frozen != SB_UNFROZEN) return -EBUSY; |
e462ec50c
|
805 |
remount_ro = (sb_flags & SB_RDONLY) && !sb_rdonly(sb); |
0aec09d04
|
806 807 808 |
} } shrink_dcache_sb(sb); |
1da177e4c
|
809 810 |
/* If we are remounting RDONLY and current sb is read/write, make sure there are no rw files opened */ |
d208bbdda
|
811 |
if (remount_ro) { |
4ed5e82fe
|
812 |
if (force) { |
eee5cc270
|
813 814 |
sb->s_readonly_remount = 1; smp_wmb(); |
4ed5e82fe
|
815 816 817 818 |
} else { retval = sb_prepare_remount_readonly(sb); if (retval) return retval; |
4ed5e82fe
|
819 |
} |
1da177e4c
|
820 821 822 |
} if (sb->s_op->remount_fs) { |
e462ec50c
|
823 |
retval = sb->s_op->remount_fs(sb, &sb_flags, data); |
2833eb2b4
|
824 825 |
if (retval) { if (!force) |
4ed5e82fe
|
826 |
goto cancel_readonly; |
2833eb2b4
|
827 828 829 830 831 |
/* If forced remount, go ahead despite any errors */ WARN(1, "forced remount of a %s fs returned %i ", sb->s_type->name, retval); } |
1da177e4c
|
832 |
} |
e462ec50c
|
833 |
sb->s_flags = (sb->s_flags & ~MS_RMT_MASK) | (sb_flags & MS_RMT_MASK); |
4ed5e82fe
|
834 835 836 |
/* Needs to be ordered wrt mnt_is_readonly() */ smp_wmb(); sb->s_readonly_remount = 0; |
c79d967de
|
837 |
|
d208bbdda
|
838 839 840 841 842 843 844 845 846 847 |
/* * 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
|
848 |
return 0; |
4ed5e82fe
|
849 850 851 852 |
cancel_readonly: sb->s_readonly_remount = 0; return retval; |
1da177e4c
|
853 |
} |
a2a9537ac
|
854 |
static void do_emergency_remount(struct work_struct *work) |
1da177e4c
|
855 |
{ |
dca332528
|
856 |
struct super_block *sb, *p = NULL; |
1da177e4c
|
857 858 |
spin_lock(&sb_lock); |
dca332528
|
859 |
list_for_each_entry(sb, &super_blocks, s_list) { |
a5166169f
|
860 |
if (hlist_unhashed(&sb->s_instances)) |
551de6f34
|
861 |
continue; |
1da177e4c
|
862 863 |
sb->s_count++; spin_unlock(&sb_lock); |
443b94baa
|
864 |
down_write(&sb->s_umount); |
e462ec50c
|
865 |
if (sb->s_root && sb->s_bdev && (sb->s_flags & SB_BORN) && |
bc98a42c1
|
866 |
!sb_rdonly(sb)) { |
1da177e4c
|
867 |
/* |
1da177e4c
|
868 869 |
* What lock protects sb->s_flags?? */ |
e462ec50c
|
870 |
do_remount_sb(sb, SB_RDONLY, NULL, 1); |
1da177e4c
|
871 |
} |
443b94baa
|
872 |
up_write(&sb->s_umount); |
1da177e4c
|
873 |
spin_lock(&sb_lock); |
dca332528
|
874 875 876 |
if (p) __put_super(p); p = sb; |
1da177e4c
|
877 |
} |
dca332528
|
878 879 |
if (p) __put_super(p); |
1da177e4c
|
880 |
spin_unlock(&sb_lock); |
a2a9537ac
|
881 |
kfree(work); |
1da177e4c
|
882 883 884 885 886 887 |
printk("Emergency Remount complete "); } void emergency_remount(void) { |
a2a9537ac
|
888 889 890 891 892 893 894 |
struct work_struct *work; work = kmalloc(sizeof(*work), GFP_ATOMIC); if (work) { INIT_WORK(work, do_emergency_remount); schedule_work(work); } |
1da177e4c
|
895 896 897 898 899 900 |
} /* * Unnamed block devices are dummy devices used by virtual * filesystems which don't use real block-devices. -- jrs */ |
ad76cbc63
|
901 |
static DEFINE_IDA(unnamed_dev_ida); |
1da177e4c
|
902 |
static DEFINE_SPINLOCK(unnamed_dev_lock);/* protects the above */ |
a2a4dc494
|
903 904 905 906 |
/* 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
|
907 |
|
0ee5dc676
|
908 |
int get_anon_bdev(dev_t *p) |
1da177e4c
|
909 910 911 912 913 |
{ int dev; int error; retry: |
ad76cbc63
|
914 |
if (ida_pre_get(&unnamed_dev_ida, GFP_ATOMIC) == 0) |
1da177e4c
|
915 916 |
return -ENOMEM; spin_lock(&unnamed_dev_lock); |
c63e09ecc
|
917 |
error = ida_get_new_above(&unnamed_dev_ida, unnamed_dev_start, &dev); |
f21f62208
|
918 919 |
if (!error) unnamed_dev_start = dev + 1; |
1da177e4c
|
920 921 922 923 924 925 |
spin_unlock(&unnamed_dev_lock); if (error == -EAGAIN) /* We raced and lost with another CPU. */ goto retry; else if (error) return -EAGAIN; |
1af95de6f
|
926 |
if (dev >= (1 << MINORBITS)) { |
1da177e4c
|
927 |
spin_lock(&unnamed_dev_lock); |
ad76cbc63
|
928 |
ida_remove(&unnamed_dev_ida, dev); |
f21f62208
|
929 930 |
if (unnamed_dev_start > dev) unnamed_dev_start = dev; |
1da177e4c
|
931 932 933 |
spin_unlock(&unnamed_dev_lock); return -EMFILE; } |
0ee5dc676
|
934 |
*p = MKDEV(0, dev & MINORMASK); |
1da177e4c
|
935 936 |
return 0; } |
0ee5dc676
|
937 |
EXPORT_SYMBOL(get_anon_bdev); |
1da177e4c
|
938 |
|
0ee5dc676
|
939 |
void free_anon_bdev(dev_t dev) |
1da177e4c
|
940 |
{ |
0ee5dc676
|
941 |
int slot = MINOR(dev); |
1da177e4c
|
942 |
spin_lock(&unnamed_dev_lock); |
ad76cbc63
|
943 |
ida_remove(&unnamed_dev_ida, slot); |
c63e09ecc
|
944 945 |
if (slot < unnamed_dev_start) unnamed_dev_start = slot; |
1da177e4c
|
946 947 |
spin_unlock(&unnamed_dev_lock); } |
0ee5dc676
|
948 949 950 951 |
EXPORT_SYMBOL(free_anon_bdev); int set_anon_super(struct super_block *s, void *data) { |
df0ce26cb
|
952 |
return get_anon_bdev(&s->s_dev); |
0ee5dc676
|
953 954 955 956 957 958 959 960 961 962 |
} 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
|
963 964 |
EXPORT_SYMBOL(kill_anon_super); |
1da177e4c
|
965 966 967 968 969 970 971 972 |
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
|
973 974 975 976 977 978 979 980 981 982 |
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); } |
d91ee87d8
|
983 984 985 |
struct dentry *mount_ns(struct file_system_type *fs_type, int flags, void *data, void *ns, struct user_namespace *user_ns, int (*fill_super)(struct super_block *, void *, int)) |
909e6d947
|
986 987 |
{ struct super_block *sb; |
d91ee87d8
|
988 989 990 |
/* Don't allow mounting unless the caller has CAP_SYS_ADMIN * over the namespace. */ |
e462ec50c
|
991 |
if (!(flags & SB_KERNMOUNT) && !ns_capable(user_ns, CAP_SYS_ADMIN)) |
d91ee87d8
|
992 |
return ERR_PTR(-EPERM); |
6e4eab577
|
993 994 |
sb = sget_userns(fs_type, ns_test_super, ns_set_super, flags, user_ns, ns); |
909e6d947
|
995 |
if (IS_ERR(sb)) |
ceefda693
|
996 |
return ERR_CAST(sb); |
909e6d947
|
997 998 999 |
if (!sb->s_root) { int err; |
e462ec50c
|
1000 |
err = fill_super(sb, data, flags & SB_SILENT ? 1 : 0); |
909e6d947
|
1001 |
if (err) { |
74dbbdd7f
|
1002 |
deactivate_locked_super(sb); |
ceefda693
|
1003 |
return ERR_PTR(err); |
909e6d947
|
1004 |
} |
e462ec50c
|
1005 |
sb->s_flags |= SB_ACTIVE; |
909e6d947
|
1006 |
} |
ceefda693
|
1007 |
return dget(sb->s_root); |
909e6d947
|
1008 |
} |
ceefda693
|
1009 |
EXPORT_SYMBOL(mount_ns); |
909e6d947
|
1010 |
|
9361401eb
|
1011 |
#ifdef CONFIG_BLOCK |
1da177e4c
|
1012 1013 1014 1015 |
static int set_bdev_super(struct super_block *s, void *data) { s->s_bdev = data; s->s_dev = s->s_bdev->bd_dev; |
13eec2363
|
1016 |
s->s_bdi = bdi_get(s->s_bdev->bd_bdi); |
32a88aa1b
|
1017 |
|
1da177e4c
|
1018 1019 1020 1021 1022 1023 1024 |
return 0; } static int test_bdev_super(struct super_block *s, void *data) { return (void *)s->s_bdev == data; } |
152a08366
|
1025 |
struct dentry *mount_bdev(struct file_system_type *fs_type, |
1da177e4c
|
1026 |
int flags, const char *dev_name, void *data, |
152a08366
|
1027 |
int (*fill_super)(struct super_block *, void *, int)) |
1da177e4c
|
1028 1029 1030 |
{ struct block_device *bdev; struct super_block *s; |
d4d776299
|
1031 |
fmode_t mode = FMODE_READ | FMODE_EXCL; |
1da177e4c
|
1032 |
int error = 0; |
e462ec50c
|
1033 |
if (!(flags & SB_RDONLY)) |
30c40d2c0
|
1034 |
mode |= FMODE_WRITE; |
d4d776299
|
1035 |
bdev = blkdev_get_by_path(dev_name, mode, fs_type); |
1da177e4c
|
1036 |
if (IS_ERR(bdev)) |
152a08366
|
1037 |
return ERR_CAST(bdev); |
1da177e4c
|
1038 1039 1040 1041 1042 1043 |
/* * 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
|
1044 1045 1046 1047 1048 1049 |
mutex_lock(&bdev->bd_fsfreeze_mutex); if (bdev->bd_fsfreeze_count > 0) { mutex_unlock(&bdev->bd_fsfreeze_mutex); error = -EBUSY; goto error_bdev; } |
e462ec50c
|
1050 |
s = sget(fs_type, test_bdev_super, set_bdev_super, flags | SB_NOSEC, |
9249e17fe
|
1051 |
bdev); |
4fadd7bb2
|
1052 |
mutex_unlock(&bdev->bd_fsfreeze_mutex); |
1da177e4c
|
1053 |
if (IS_ERR(s)) |
454e2398b
|
1054 |
goto error_s; |
1da177e4c
|
1055 1056 |
if (s->s_root) { |
e462ec50c
|
1057 |
if ((flags ^ s->s_flags) & SB_RDONLY) { |
74dbbdd7f
|
1058 |
deactivate_locked_super(s); |
454e2398b
|
1059 1060 |
error = -EBUSY; goto error_bdev; |
1da177e4c
|
1061 |
} |
454e2398b
|
1062 |
|
4f331f01b
|
1063 1064 |
/* * s_umount nests inside bd_mutex during |
e525fd89d
|
1065 1066 1067 1068 |
* __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
|
1069 1070 |
*/ up_write(&s->s_umount); |
d4d776299
|
1071 |
blkdev_put(bdev, mode); |
4f331f01b
|
1072 |
down_write(&s->s_umount); |
1da177e4c
|
1073 |
} else { |
30c40d2c0
|
1074 |
s->s_mode = mode; |
a1c6f0573
|
1075 |
snprintf(s->s_id, sizeof(s->s_id), "%pg", bdev); |
e78c9a004
|
1076 |
sb_set_blocksize(s, block_size(bdev)); |
e462ec50c
|
1077 |
error = fill_super(s, data, flags & SB_SILENT ? 1 : 0); |
1da177e4c
|
1078 |
if (error) { |
74dbbdd7f
|
1079 |
deactivate_locked_super(s); |
454e2398b
|
1080 |
goto error; |
fa675765a
|
1081 |
} |
454e2398b
|
1082 |
|
e462ec50c
|
1083 |
s->s_flags |= SB_ACTIVE; |
87d8fe1ee
|
1084 |
bdev->bd_super = s; |
1da177e4c
|
1085 |
} |
152a08366
|
1086 |
return dget(s->s_root); |
1da177e4c
|
1087 |
|
454e2398b
|
1088 1089 1090 |
error_s: error = PTR_ERR(s); error_bdev: |
d4d776299
|
1091 |
blkdev_put(bdev, mode); |
454e2398b
|
1092 |
error: |
152a08366
|
1093 1094 1095 |
return ERR_PTR(error); } EXPORT_SYMBOL(mount_bdev); |
1da177e4c
|
1096 1097 1098 |
void kill_block_super(struct super_block *sb) { struct block_device *bdev = sb->s_bdev; |
30c40d2c0
|
1099 |
fmode_t mode = sb->s_mode; |
1da177e4c
|
1100 |
|
ddbaaf302
|
1101 |
bdev->bd_super = NULL; |
1da177e4c
|
1102 1103 |
generic_shutdown_super(sb); sync_blockdev(bdev); |
d4d776299
|
1104 |
WARN_ON_ONCE(!(mode & FMODE_EXCL)); |
e525fd89d
|
1105 |
blkdev_put(bdev, mode | FMODE_EXCL); |
1da177e4c
|
1106 1107 1108 |
} EXPORT_SYMBOL(kill_block_super); |
9361401eb
|
1109 |
#endif |
1da177e4c
|
1110 |
|
3c26ff6e4
|
1111 |
struct dentry *mount_nodev(struct file_system_type *fs_type, |
1da177e4c
|
1112 |
int flags, void *data, |
3c26ff6e4
|
1113 |
int (*fill_super)(struct super_block *, void *, int)) |
1da177e4c
|
1114 1115 |
{ int error; |
9249e17fe
|
1116 |
struct super_block *s = sget(fs_type, NULL, set_anon_super, flags, NULL); |
1da177e4c
|
1117 1118 |
if (IS_ERR(s)) |
3c26ff6e4
|
1119 |
return ERR_CAST(s); |
1da177e4c
|
1120 |
|
e462ec50c
|
1121 |
error = fill_super(s, data, flags & SB_SILENT ? 1 : 0); |
1da177e4c
|
1122 |
if (error) { |
74dbbdd7f
|
1123 |
deactivate_locked_super(s); |
3c26ff6e4
|
1124 |
return ERR_PTR(error); |
1da177e4c
|
1125 |
} |
e462ec50c
|
1126 |
s->s_flags |= SB_ACTIVE; |
3c26ff6e4
|
1127 |
return dget(s->s_root); |
1da177e4c
|
1128 |
} |
3c26ff6e4
|
1129 |
EXPORT_SYMBOL(mount_nodev); |
1da177e4c
|
1130 1131 1132 1133 |
static int compare_single(struct super_block *s, void *p) { return 1; } |
fc14f2fef
|
1134 |
struct dentry *mount_single(struct file_system_type *fs_type, |
1da177e4c
|
1135 |
int flags, void *data, |
fc14f2fef
|
1136 |
int (*fill_super)(struct super_block *, void *, int)) |
1da177e4c
|
1137 1138 1139 |
{ struct super_block *s; int error; |
9249e17fe
|
1140 |
s = sget(fs_type, compare_single, set_anon_super, flags, NULL); |
1da177e4c
|
1141 |
if (IS_ERR(s)) |
fc14f2fef
|
1142 |
return ERR_CAST(s); |
1da177e4c
|
1143 |
if (!s->s_root) { |
e462ec50c
|
1144 |
error = fill_super(s, data, flags & SB_SILENT ? 1 : 0); |
1da177e4c
|
1145 |
if (error) { |
74dbbdd7f
|
1146 |
deactivate_locked_super(s); |
fc14f2fef
|
1147 |
return ERR_PTR(error); |
1da177e4c
|
1148 |
} |
e462ec50c
|
1149 |
s->s_flags |= SB_ACTIVE; |
9329d1bea
|
1150 1151 |
} else { do_remount_sb(s, flags, data, 0); |
1da177e4c
|
1152 |
} |
fc14f2fef
|
1153 1154 1155 |
return dget(s->s_root); } EXPORT_SYMBOL(mount_single); |
9d412a43c
|
1156 1157 |
struct dentry * mount_fs(struct file_system_type *type, int flags, const char *name, void *data) |
1da177e4c
|
1158 |
{ |
c96e41e92
|
1159 |
struct dentry *root; |
9d412a43c
|
1160 |
struct super_block *sb; |
1da177e4c
|
1161 |
char *secdata = NULL; |
9d412a43c
|
1162 |
int error = -ENOMEM; |
8089352a1
|
1163 |
|
e00075298
|
1164 |
if (data && !(type->fs_flags & FS_BINARY_MOUNTDATA)) { |
1da177e4c
|
1165 |
secdata = alloc_secdata(); |
454e2398b
|
1166 |
if (!secdata) |
9d412a43c
|
1167 |
goto out; |
1da177e4c
|
1168 |
|
e00075298
|
1169 |
error = security_sb_copy_data(data, secdata); |
454e2398b
|
1170 |
if (error) |
1da177e4c
|
1171 |
goto out_free_secdata; |
1da177e4c
|
1172 |
} |
1a102ff92
|
1173 1174 1175 1176 |
root = type->mount(type, flags, name, data); if (IS_ERR(root)) { error = PTR_ERR(root); goto out_free_secdata; |
c96e41e92
|
1177 |
} |
9d412a43c
|
1178 1179 1180 |
sb = root->d_sb; BUG_ON(!sb); WARN_ON(!sb->s_bdi); |
e462ec50c
|
1181 |
sb->s_flags |= SB_BORN; |
454e2398b
|
1182 |
|
9d412a43c
|
1183 |
error = security_sb_kern_mount(sb, flags, secdata); |
5129a469a
|
1184 1185 |
if (error) goto out_sb; |
454e2398b
|
1186 |
|
42cb56ae2
|
1187 1188 1189 1190 |
/* * 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
|
1191 |
* violate this rule. |
42cb56ae2
|
1192 |
*/ |
9d412a43c
|
1193 1194 1195 |
WARN((sb->s_maxbytes < 0), "%s set sb->s_maxbytes to " "negative value (%lld) ", type->name, sb->s_maxbytes); |
42cb56ae2
|
1196 |
|
9d412a43c
|
1197 |
up_write(&sb->s_umount); |
8680e22f2
|
1198 |
free_secdata(secdata); |
9d412a43c
|
1199 |
return root; |
1da177e4c
|
1200 |
out_sb: |
9d412a43c
|
1201 1202 |
dput(root); deactivate_locked_super(sb); |
1da177e4c
|
1203 1204 |
out_free_secdata: free_secdata(secdata); |
1da177e4c
|
1205 |
out: |
454e2398b
|
1206 |
return ERR_PTR(error); |
1da177e4c
|
1207 |
} |
5accdf82b
|
1208 |
/* |
fca39346a
|
1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 |
* Setup private BDI for given superblock. It gets automatically cleaned up * in generic_shutdown_super(). */ int super_setup_bdi_name(struct super_block *sb, char *fmt, ...) { struct backing_dev_info *bdi; int err; va_list args; bdi = bdi_alloc(GFP_KERNEL); if (!bdi) return -ENOMEM; bdi->name = sb->s_type->name; va_start(args, fmt); |
7c4cc3002
|
1225 |
err = bdi_register_va(bdi, fmt, args); |
fca39346a
|
1226 1227 1228 1229 1230 1231 1232 |
va_end(args); if (err) { bdi_put(bdi); return err; } WARN_ON(sb->s_bdi != &noop_backing_dev_info); sb->s_bdi = bdi; |
fca39346a
|
1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 |
return 0; } EXPORT_SYMBOL(super_setup_bdi_name); /* * Setup private BDI for given superblock. I gets automatically cleaned up * in generic_shutdown_super(). */ int super_setup_bdi(struct super_block *sb) { static atomic_long_t bdi_seq = ATOMIC_LONG_INIT(0); return super_setup_bdi_name(sb, "%.28s-%ld", sb->s_type->name, atomic_long_inc_return(&bdi_seq)); } EXPORT_SYMBOL(super_setup_bdi); /* |
5accdf82b
|
1252 1253 1254 1255 1256 |
* This is an internal function, please use sb_end_{write,pagefault,intwrite} * instead. */ void __sb_end_write(struct super_block *sb, int level) { |
8129ed296
|
1257 |
percpu_up_read(sb->s_writers.rw_sem + level-1); |
5accdf82b
|
1258 1259 |
} EXPORT_SYMBOL(__sb_end_write); |
f4b554af9
|
1260 1261 1262 1263 1264 1265 1266 |
/* * 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
|
1267 |
int ret = 1; |
f4b554af9
|
1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 |
#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
|
1283 |
if (percpu_rwsem_is_held(sb->s_writers.rw_sem + i)) { |
f4b554af9
|
1284 1285 1286 1287 1288 |
force_trylock = true; break; } } #endif |
8129ed296
|
1289 1290 1291 1292 |
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); |
22224a175
|
1293 |
WARN_ON(force_trylock && !ret); |
f4b554af9
|
1294 1295 |
return ret; } |
5accdf82b
|
1296 1297 1298 1299 1300 1301 1302 1303 |
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
|
1304 |
* system. |
5accdf82b
|
1305 1306 1307 |
*/ static void sb_wait_write(struct super_block *sb, int level) { |
8129ed296
|
1308 |
percpu_down_write(sb->s_writers.rw_sem + level-1); |
8129ed296
|
1309 |
} |
5accdf82b
|
1310 |
|
f1a962203
|
1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 |
/* * We are going to return to userspace and forget about these locks, the * ownership goes to the caller of thaw_super() which does unlock(). */ static void lockdep_sb_freeze_release(struct super_block *sb) { int level; for (level = SB_FREEZE_LEVELS - 1; level >= 0; level--) percpu_rwsem_release(sb->s_writers.rw_sem + level, 0, _THIS_IP_); } /* * Tell lockdep we are holding these locks before we call ->unfreeze_fs(sb). */ static void lockdep_sb_freeze_acquire(struct super_block *sb) |
8129ed296
|
1327 1328 |
{ int level; |
5accdf82b
|
1329 |
|
8129ed296
|
1330 1331 |
for (level = 0; level < SB_FREEZE_LEVELS; ++level) percpu_rwsem_acquire(sb->s_writers.rw_sem + level, 0, _THIS_IP_); |
f1a962203
|
1332 1333 1334 1335 1336 |
} static void sb_freeze_unlock(struct super_block *sb) { int level; |
5accdf82b
|
1337 |
|
8129ed296
|
1338 1339 |
for (level = SB_FREEZE_LEVELS - 1; level >= 0; level--) percpu_up_write(sb->s_writers.rw_sem + level); |
5accdf82b
|
1340 |
} |
18e9e5104
|
1341 |
/** |
7000d3c42
|
1342 1343 |
* freeze_super - lock the filesystem and force it into a consistent state * @sb: the super to lock |
18e9e5104
|
1344 1345 1346 1347 |
* * 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
|
1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 |
* * 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
|
1373 1374 1375 1376 1377 1378 1379 |
*/ int freeze_super(struct super_block *sb) { int ret; atomic_inc(&sb->s_active); down_write(&sb->s_umount); |
5accdf82b
|
1380 |
if (sb->s_writers.frozen != SB_UNFROZEN) { |
18e9e5104
|
1381 1382 1383 |
deactivate_locked_super(sb); return -EBUSY; } |
e462ec50c
|
1384 |
if (!(sb->s_flags & SB_BORN)) { |
dabe0dc19
|
1385 1386 1387 |
up_write(&sb->s_umount); return 0; /* sic - it's "nothing to do" */ } |
bc98a42c1
|
1388 |
if (sb_rdonly(sb)) { |
5accdf82b
|
1389 1390 |
/* Nothing to do really... */ sb->s_writers.frozen = SB_FREEZE_COMPLETE; |
18e9e5104
|
1391 1392 1393 |
up_write(&sb->s_umount); return 0; } |
5accdf82b
|
1394 |
sb->s_writers.frozen = SB_FREEZE_WRITE; |
5accdf82b
|
1395 1396 |
/* Release s_umount to preserve sb_start_write -> s_umount ordering */ up_write(&sb->s_umount); |
5accdf82b
|
1397 |
sb_wait_write(sb, SB_FREEZE_WRITE); |
8129ed296
|
1398 |
down_write(&sb->s_umount); |
5accdf82b
|
1399 1400 |
/* Now we go and block page faults... */ |
5accdf82b
|
1401 |
sb->s_writers.frozen = SB_FREEZE_PAGEFAULT; |
5accdf82b
|
1402 1403 1404 |
sb_wait_write(sb, SB_FREEZE_PAGEFAULT); /* All writers are done so after syncing there won't be dirty data */ |
18e9e5104
|
1405 |
sync_filesystem(sb); |
5accdf82b
|
1406 1407 |
/* Now wait for internal filesystem counter */ sb->s_writers.frozen = SB_FREEZE_FS; |
5accdf82b
|
1408 |
sb_wait_write(sb, SB_FREEZE_FS); |
18e9e5104
|
1409 |
|
18e9e5104
|
1410 1411 1412 1413 1414 1415 |
if (sb->s_op->freeze_fs) { ret = sb->s_op->freeze_fs(sb); if (ret) { printk(KERN_ERR "VFS:Filesystem freeze failed "); |
5accdf82b
|
1416 |
sb->s_writers.frozen = SB_UNFROZEN; |
8129ed296
|
1417 |
sb_freeze_unlock(sb); |
5accdf82b
|
1418 |
wake_up(&sb->s_writers.wait_unfrozen); |
18e9e5104
|
1419 1420 1421 1422 |
deactivate_locked_super(sb); return ret; } } |
5accdf82b
|
1423 |
/* |
89f39af12
|
1424 1425 |
* For debugging purposes so that fs can warn if it sees write activity * when frozen is set to SB_FREEZE_COMPLETE, and for thaw_super(). |
5accdf82b
|
1426 1427 |
*/ sb->s_writers.frozen = SB_FREEZE_COMPLETE; |
f1a962203
|
1428 |
lockdep_sb_freeze_release(sb); |
18e9e5104
|
1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 |
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); |
89f39af12
|
1445 |
if (sb->s_writers.frozen != SB_FREEZE_COMPLETE) { |
18e9e5104
|
1446 1447 1448 |
up_write(&sb->s_umount); return -EINVAL; } |
bc98a42c1
|
1449 |
if (sb_rdonly(sb)) { |
8129ed296
|
1450 |
sb->s_writers.frozen = SB_UNFROZEN; |
18e9e5104
|
1451 |
goto out; |
8129ed296
|
1452 |
} |
18e9e5104
|
1453 |
|
f1a962203
|
1454 |
lockdep_sb_freeze_acquire(sb); |
18e9e5104
|
1455 1456 1457 1458 1459 1460 |
if (sb->s_op->unfreeze_fs) { error = sb->s_op->unfreeze_fs(sb); if (error) { printk(KERN_ERR "VFS:Filesystem thaw failed "); |
f1a962203
|
1461 |
lockdep_sb_freeze_release(sb); |
18e9e5104
|
1462 1463 1464 1465 |
up_write(&sb->s_umount); return error; } } |
5accdf82b
|
1466 |
sb->s_writers.frozen = SB_UNFROZEN; |
8129ed296
|
1467 1468 |
sb_freeze_unlock(sb); out: |
5accdf82b
|
1469 |
wake_up(&sb->s_writers.wait_unfrozen); |
18e9e5104
|
1470 |
deactivate_locked_super(sb); |
18e9e5104
|
1471 1472 1473 |
return 0; } EXPORT_SYMBOL(thaw_super); |