aboutsummaryrefslogtreecommitdiffstats
path: root/fs
diff options
context:
space:
mode:
authorDavid Howells <dhowells@redhat.com>2010-05-11 11:51:39 -0400
committerLinus Torvalds <torvalds@linux-foundation.org>2010-05-11 13:07:53 -0400
commitc61ea31dac0319ec64b33725917bda81fc293a25 (patch)
tree05a4f3011ea8b334795aae606d89bcf27e3e26c5 /fs
parent7d6fb7bd1919517937ec390f6ca2d7bcf4f89fb6 (diff)
CacheFiles: Fix occasional EIO on call to vfs_unlink()
Fix an occasional EIO returned by a call to vfs_unlink(): [ 4868.465413] CacheFiles: I/O Error: Unlink failed [ 4868.465444] FS-Cache: Cache cachefiles stopped due to I/O error [ 4947.320011] CacheFiles: File cache on md3 unregistering [ 4947.320041] FS-Cache: Withdrawing cache "mycache" [ 5127.348683] FS-Cache: Cache "mycache" added (type cachefiles) [ 5127.348716] CacheFiles: File cache on md3 registered [ 7076.871081] CacheFiles: I/O Error: Unlink failed [ 7076.871130] FS-Cache: Cache cachefiles stopped due to I/O error [ 7116.780891] CacheFiles: File cache on md3 unregistering [ 7116.780937] FS-Cache: Withdrawing cache "mycache" [ 7296.813394] FS-Cache: Cache "mycache" added (type cachefiles) [ 7296.813432] CacheFiles: File cache on md3 registered What happens is this: (1) A cached NFS file is seen to have become out of date, so NFS retires the object and immediately acquires a new object with the same key. (2) Retirement of the old object is done asynchronously - so the lookup/create to generate the new object may be done first. This can be a problem as the old object and the new object must exist at the same point in the backing filesystem (i.e. they must have the same pathname). (3) The lookup for the new object sees that a backing file already exists, checks to see whether it is valid and sees that it isn't. It then deletes that file and creates a new one on disk. (4) The retirement phase for the old file is then performed. It tries to delete the dentry it has, but ext4_unlink() returns -EIO because the inode attached to that dentry no longer matches the inode number associated with the filename in the parent directory. The trace below shows this quite well. [md5sum] ==> __fscache_relinquish_cookie(ffff88002d12fb58{NFS.fh,ffff88002ce62100},1) [md5sum] ==> __fscache_acquire_cookie({NFS.server},{NFS.fh},ffff88002ce62100) NFS has retired the old cookie and asked for a new one. [kslowd] ==> fscache_object_state_machine({OBJ52,OBJECT_ACTIVE,24}) [kslowd] <== fscache_object_state_machine() [->OBJECT_DYING] [kslowd] ==> fscache_object_state_machine({OBJ53,OBJECT_INIT,0}) [kslowd] <== fscache_object_state_machine() [->OBJECT_LOOKING_UP] [kslowd] ==> fscache_object_state_machine({OBJ52,OBJECT_DYING,24}) [kslowd] <== fscache_object_state_machine() [->OBJECT_RECYCLING] The old object (OBJ52) is going through the terminal states to get rid of it, whilst the new object - (OBJ53) - is coming into being. [kslowd] ==> fscache_object_state_machine({OBJ53,OBJECT_LOOKING_UP,0}) [kslowd] ==> cachefiles_walk_to_object({ffff88003029d8b8},OBJ53,@68,) [kslowd] lookup '@68' [kslowd] next -> ffff88002ce41bd0 positive [kslowd] advance [kslowd] lookup 'Es0g00og0_Nd_XCYe3BOzvXrsBLMlN6aw16M1htaA' [kslowd] next -> ffff8800369faac8 positive The new object has looked up the subdir in which the file would be in (getting dentry ffff88002ce41bd0) and then looked up the file itself (getting dentry ffff8800369faac8). [kslowd] validate 'Es0g00og0_Nd_XCYe3BOzvXrsBLMlN6aw16M1htaA' [kslowd] ==> cachefiles_bury_object(,'@68','Es0g00og0_Nd_XCYe3BOzvXrsBLMlN6aw16M1htaA') [kslowd] remove ffff8800369faac8 from ffff88002ce41bd0 [kslowd] unlink stale object [kslowd] <== cachefiles_bury_object() = 0 It then checks the file's xattrs to see if it's valid. NFS says that the auxiliary data indicate the file is out of date (obvious to us - that's why NFS ditched the old version and got a new one). CacheFiles then deletes the old file (dentry ffff8800369faac8). [kslowd] redo lookup [kslowd] lookup 'Es0g00og0_Nd_XCYe3BOzvXrsBLMlN6aw16M1htaA' [kslowd] next -> ffff88002cd94288 negative [kslowd] create -> ffff88002cd94288{ffff88002cdaf238{ino=148247}} CacheFiles then redoes the lookup and gets a negative result in a new dentry (ffff88002cd94288) which it then creates a file for. [kslowd] ==> cachefiles_mark_object_active(,OBJ53) [kslowd] <== cachefiles_mark_object_active() = 0 [kslowd] === OBTAINED_OBJECT === [kslowd] <== cachefiles_walk_to_object() = 0 [148247] [kslowd] <== fscache_object_state_machine() [->OBJECT_AVAILABLE] The new object is then marked active and the state machine moves to the available state - at which point NFS can start filling the object. [kslowd] ==> fscache_object_state_machine({OBJ52,OBJECT_RECYCLING,20}) [kslowd] ==> fscache_release_object() [kslowd] ==> cachefiles_drop_object({OBJ52,2}) [kslowd] ==> cachefiles_delete_object(,OBJ52{ffff8800369faac8}) The old object, meanwhile, goes on with being retired. If allocation occurs first, cachefiles_delete_object() has to wait for dir->d_inode->i_mutex to become available before it can continue. [kslowd] ==> cachefiles_bury_object(,'@68','Es0g00og0_Nd_XCYe3BOzvXrsBLMlN6aw16M1htaA') [kslowd] remove ffff8800369faac8 from ffff88002ce41bd0 [kslowd] unlink stale object EXT4-fs warning (device sda6): ext4_unlink: Inode number mismatch in unlink (148247!=148193) CacheFiles: I/O Error: Unlink failed FS-Cache: Cache cachefiles stopped due to I/O error CacheFiles then tries to delete the file for the old object, but the dentry it has (ffff8800369faac8) no longer points to a valid inode for that directory entry, and so ext4_unlink() returns -EIO when de->inode does not match i_ino. [kslowd] <== cachefiles_bury_object() = -5 [kslowd] <== cachefiles_delete_object() = -5 [kslowd] <== fscache_object_state_machine() [->OBJECT_DEAD] [kslowd] ==> fscache_object_state_machine({OBJ53,OBJECT_AVAILABLE,0}) [kslowd] <== fscache_object_state_machine() [->OBJECT_ACTIVE] (Note that the above trace includes extra information beyond that produced by the upstream code). The fix is to note when an object that is being retired has had its object deleted preemptively by a replacement object that is being created, and to skip the second removal attempt in such a case. Reported-by: Greg M <gregm@servu.net.au> Reported-by: Mark Moseley <moseleymark@gmail.com> Reported-by: Romain DEGEZ <romain.degez@smartjog.com> Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Diffstat (limited to 'fs')
-rw-r--r--fs/cachefiles/internal.h1
-rw-r--r--fs/cachefiles/namei.c98
2 files changed, 87 insertions, 12 deletions
diff --git a/fs/cachefiles/internal.h b/fs/cachefiles/internal.h
index f7c255f9c624..a8cd821226da 100644
--- a/fs/cachefiles/internal.h
+++ b/fs/cachefiles/internal.h
@@ -34,6 +34,7 @@ struct cachefiles_object {
34 loff_t i_size; /* object size */ 34 loff_t i_size; /* object size */
35 unsigned long flags; 35 unsigned long flags;
36#define CACHEFILES_OBJECT_ACTIVE 0 /* T if marked active */ 36#define CACHEFILES_OBJECT_ACTIVE 0 /* T if marked active */
37#define CACHEFILES_OBJECT_BURIED 1 /* T if preemptively buried */
37 atomic_t usage; /* object usage count */ 38 atomic_t usage; /* object usage count */
38 uint8_t type; /* object type */ 39 uint8_t type; /* object type */
39 uint8_t new; /* T if object new */ 40 uint8_t new; /* T if object new */
diff --git a/fs/cachefiles/namei.c b/fs/cachefiles/namei.c
index d5db84a1ee0d..f4a7840bf42c 100644
--- a/fs/cachefiles/namei.c
+++ b/fs/cachefiles/namei.c
@@ -93,6 +93,59 @@ static noinline void cachefiles_printk_object(struct cachefiles_object *object,
93} 93}
94 94
95/* 95/*
96 * mark the owner of a dentry, if there is one, to indicate that that dentry
97 * has been preemptively deleted
98 * - the caller must hold the i_mutex on the dentry's parent as required to
99 * call vfs_unlink(), vfs_rmdir() or vfs_rename()
100 */
101static void cachefiles_mark_object_buried(struct cachefiles_cache *cache,
102 struct dentry *dentry)
103{
104 struct cachefiles_object *object;
105 struct rb_node *p;
106
107 _enter(",'%*.*s'",
108 dentry->d_name.len, dentry->d_name.len, dentry->d_name.name);
109
110 write_lock(&cache->active_lock);
111
112 p = cache->active_nodes.rb_node;
113 while (p) {
114 object = rb_entry(p, struct cachefiles_object, active_node);
115 if (object->dentry > dentry)
116 p = p->rb_left;
117 else if (object->dentry < dentry)
118 p = p->rb_right;
119 else
120 goto found_dentry;
121 }
122
123 write_unlock(&cache->active_lock);
124 _leave(" [no owner]");
125 return;
126
127 /* found the dentry for */
128found_dentry:
129 kdebug("preemptive burial: OBJ%x [%s] %p",
130 object->fscache.debug_id,
131 fscache_object_states[object->fscache.state],
132 dentry);
133
134 if (object->fscache.state < FSCACHE_OBJECT_DYING) {
135 printk(KERN_ERR "\n");
136 printk(KERN_ERR "CacheFiles: Error:"
137 " Can't preemptively bury live object\n");
138 cachefiles_printk_object(object, NULL);
139 } else if (test_and_set_bit(CACHEFILES_OBJECT_BURIED, &object->flags)) {
140 printk(KERN_ERR "CacheFiles: Error:"
141 " Object already preemptively buried\n");
142 }
143
144 write_unlock(&cache->active_lock);
145 _leave(" [owner marked]");
146}
147
148/*
96 * record the fact that an object is now active 149 * record the fact that an object is now active
97 */ 150 */
98static int cachefiles_mark_object_active(struct cachefiles_cache *cache, 151static int cachefiles_mark_object_active(struct cachefiles_cache *cache,
@@ -219,7 +272,8 @@ requeue:
219 */ 272 */
220static int cachefiles_bury_object(struct cachefiles_cache *cache, 273static int cachefiles_bury_object(struct cachefiles_cache *cache,
221 struct dentry *dir, 274 struct dentry *dir,
222 struct dentry *rep) 275 struct dentry *rep,
276 bool preemptive)
223{ 277{
224 struct dentry *grave, *trap; 278 struct dentry *grave, *trap;
225 char nbuffer[8 + 8 + 1]; 279 char nbuffer[8 + 8 + 1];
@@ -229,11 +283,16 @@ static int cachefiles_bury_object(struct cachefiles_cache *cache,
229 dir->d_name.len, dir->d_name.len, dir->d_name.name, 283 dir->d_name.len, dir->d_name.len, dir->d_name.name,
230 rep->d_name.len, rep->d_name.len, rep->d_name.name); 284 rep->d_name.len, rep->d_name.len, rep->d_name.name);
231 285
286 _debug("remove %p from %p", rep, dir);
287
232 /* non-directories can just be unlinked */ 288 /* non-directories can just be unlinked */
233 if (!S_ISDIR(rep->d_inode->i_mode)) { 289 if (!S_ISDIR(rep->d_inode->i_mode)) {
234 _debug("unlink stale object"); 290 _debug("unlink stale object");
235 ret = vfs_unlink(dir->d_inode, rep); 291 ret = vfs_unlink(dir->d_inode, rep);
236 292
293 if (preemptive)
294 cachefiles_mark_object_buried(cache, rep);
295
237 mutex_unlock(&dir->d_inode->i_mutex); 296 mutex_unlock(&dir->d_inode->i_mutex);
238 297
239 if (ret == -EIO) 298 if (ret == -EIO)
@@ -325,6 +384,9 @@ try_again:
325 if (ret != 0 && ret != -ENOMEM) 384 if (ret != 0 && ret != -ENOMEM)
326 cachefiles_io_error(cache, "Rename failed with error %d", ret); 385 cachefiles_io_error(cache, "Rename failed with error %d", ret);
327 386
387 if (preemptive)
388 cachefiles_mark_object_buried(cache, rep);
389
328 unlock_rename(cache->graveyard, dir); 390 unlock_rename(cache->graveyard, dir);
329 dput(grave); 391 dput(grave);
330 _leave(" = 0"); 392 _leave(" = 0");
@@ -340,7 +402,7 @@ int cachefiles_delete_object(struct cachefiles_cache *cache,
340 struct dentry *dir; 402 struct dentry *dir;
341 int ret; 403 int ret;
342 404
343 _enter(",{%p}", object->dentry); 405 _enter(",OBJ%x{%p}", object->fscache.debug_id, object->dentry);
344 406
345 ASSERT(object->dentry); 407 ASSERT(object->dentry);
346 ASSERT(object->dentry->d_inode); 408 ASSERT(object->dentry->d_inode);
@@ -350,15 +412,25 @@ int cachefiles_delete_object(struct cachefiles_cache *cache,
350 412
351 mutex_lock_nested(&dir->d_inode->i_mutex, I_MUTEX_PARENT); 413 mutex_lock_nested(&dir->d_inode->i_mutex, I_MUTEX_PARENT);
352 414
353 /* we need to check that our parent is _still_ our parent - it may have 415 if (test_bit(CACHEFILES_OBJECT_BURIED, &object->flags)) {
354 * been renamed */ 416 /* object allocation for the same key preemptively deleted this
355 if (dir == object->dentry->d_parent) { 417 * object's file so that it could create its own file */
356 ret = cachefiles_bury_object(cache, dir, object->dentry); 418 _debug("object preemptively buried");
357 } else {
358 /* it got moved, presumably by cachefilesd culling it, so it's
359 * no longer in the key path and we can ignore it */
360 mutex_unlock(&dir->d_inode->i_mutex); 419 mutex_unlock(&dir->d_inode->i_mutex);
361 ret = 0; 420 ret = 0;
421 } else {
422 /* we need to check that our parent is _still_ our parent - it
423 * may have been renamed */
424 if (dir == object->dentry->d_parent) {
425 ret = cachefiles_bury_object(cache, dir,
426 object->dentry, false);
427 } else {
428 /* it got moved, presumably by cachefilesd culling it,
429 * so it's no longer in the key path and we can ignore
430 * it */
431 mutex_unlock(&dir->d_inode->i_mutex);
432 ret = 0;
433 }
362 } 434 }
363 435
364 dput(dir); 436 dput(dir);
@@ -381,7 +453,9 @@ int cachefiles_walk_to_object(struct cachefiles_object *parent,
381 const char *name; 453 const char *name;
382 int ret, nlen; 454 int ret, nlen;
383 455
384 _enter("{%p},,%s,", parent->dentry, key); 456 _enter("OBJ%x{%p},OBJ%x,%s,",
457 parent->fscache.debug_id, parent->dentry,
458 object->fscache.debug_id, key);
385 459
386 cache = container_of(parent->fscache.cache, 460 cache = container_of(parent->fscache.cache,
387 struct cachefiles_cache, cache); 461 struct cachefiles_cache, cache);
@@ -509,7 +583,7 @@ lookup_again:
509 * mutex) */ 583 * mutex) */
510 object->dentry = NULL; 584 object->dentry = NULL;
511 585
512 ret = cachefiles_bury_object(cache, dir, next); 586 ret = cachefiles_bury_object(cache, dir, next, true);
513 dput(next); 587 dput(next);
514 next = NULL; 588 next = NULL;
515 589
@@ -828,7 +902,7 @@ int cachefiles_cull(struct cachefiles_cache *cache, struct dentry *dir,
828 /* actually remove the victim (drops the dir mutex) */ 902 /* actually remove the victim (drops the dir mutex) */
829 _debug("bury"); 903 _debug("bury");
830 904
831 ret = cachefiles_bury_object(cache, dir, victim); 905 ret = cachefiles_bury_object(cache, dir, victim, false);
832 if (ret < 0) 906 if (ret < 0)
833 goto error; 907 goto error;
834 908