xref: /linux/fs/notify/mark.c (revision 4949009eb8d40a441dcddcd96e101e77d31cf1b2)
1 /*
2  *  Copyright (C) 2008 Red Hat, Inc., Eric Paris <eparis@redhat.com>
3  *
4  *  This program is free software; you can redistribute it and/or modify
5  *  it under the terms of the GNU General Public License as published by
6  *  the Free Software Foundation; either version 2, or (at your option)
7  *  any later version.
8  *
9  *  This program is distributed in the hope that it will be useful,
10  *  but WITHOUT ANY WARRANTY; without even the implied warranty of
11  *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
12  *  GNU General Public License for more details.
13  *
14  *  You should have received a copy of the GNU General Public License
15  *  along with this program; see the file COPYING.  If not, write to
16  *  the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
17  */
18 
19 /*
20  * fsnotify inode mark locking/lifetime/and refcnting
21  *
22  * REFCNT:
23  * The group->recnt and mark->refcnt tell how many "things" in the kernel
24  * currently are referencing the objects. Both kind of objects typically will
25  * live inside the kernel with a refcnt of 2, one for its creation and one for
26  * the reference a group and a mark hold to each other.
27  * If you are holding the appropriate locks, you can take a reference and the
28  * object itself is guaranteed to survive until the reference is dropped.
29  *
30  * LOCKING:
31  * There are 3 locks involved with fsnotify inode marks and they MUST be taken
32  * in order as follows:
33  *
34  * group->mark_mutex
35  * mark->lock
36  * inode->i_lock
37  *
38  * group->mark_mutex protects the marks_list anchored inside a given group and
39  * each mark is hooked via the g_list.  It also protects the groups private
40  * data (i.e group limits).
41 
42  * mark->lock protects the marks attributes like its masks and flags.
43  * Furthermore it protects the access to a reference of the group that the mark
44  * is assigned to as well as the access to a reference of the inode/vfsmount
45  * that is being watched by the mark.
46  *
47  * inode->i_lock protects the i_fsnotify_marks list anchored inside a
48  * given inode and each mark is hooked via the i_list. (and sorta the
49  * free_i_list)
50  *
51  *
52  * LIFETIME:
53  * Inode marks survive between when they are added to an inode and when their
54  * refcnt==0.
55  *
56  * The inode mark can be cleared for a number of different reasons including:
57  * - The inode is unlinked for the last time.  (fsnotify_inode_remove)
58  * - The inode is being evicted from cache. (fsnotify_inode_delete)
59  * - The fs the inode is on is unmounted.  (fsnotify_inode_delete/fsnotify_unmount_inodes)
60  * - Something explicitly requests that it be removed.  (fsnotify_destroy_mark)
61  * - The fsnotify_group associated with the mark is going away and all such marks
62  *   need to be cleaned up. (fsnotify_clear_marks_by_group)
63  *
64  * Worst case we are given an inode and need to clean up all the marks on that
65  * inode.  We take i_lock and walk the i_fsnotify_marks safely.  For each
66  * mark on the list we take a reference (so the mark can't disappear under us).
67  * We remove that mark form the inode's list of marks and we add this mark to a
68  * private list anchored on the stack using i_free_list; we walk i_free_list
69  * and before we destroy the mark we make sure that we dont race with a
70  * concurrent destroy_group by getting a ref to the marks group and taking the
71  * groups mutex.
72 
73  * Very similarly for freeing by group, except we use free_g_list.
74  *
75  * This has the very interesting property of being able to run concurrently with
76  * any (or all) other directions.
77  */
78 
79 #include <linux/fs.h>
80 #include <linux/init.h>
81 #include <linux/kernel.h>
82 #include <linux/kthread.h>
83 #include <linux/module.h>
84 #include <linux/mutex.h>
85 #include <linux/slab.h>
86 #include <linux/spinlock.h>
87 #include <linux/srcu.h>
88 
89 #include <linux/atomic.h>
90 
91 #include <linux/fsnotify_backend.h>
92 #include "fsnotify.h"
93 
94 struct srcu_struct fsnotify_mark_srcu;
95 static DEFINE_SPINLOCK(destroy_lock);
96 static LIST_HEAD(destroy_list);
97 static DECLARE_WAIT_QUEUE_HEAD(destroy_waitq);
98 
99 void fsnotify_get_mark(struct fsnotify_mark *mark)
100 {
101 	atomic_inc(&mark->refcnt);
102 }
103 
104 void fsnotify_put_mark(struct fsnotify_mark *mark)
105 {
106 	if (atomic_dec_and_test(&mark->refcnt)) {
107 		if (mark->group)
108 			fsnotify_put_group(mark->group);
109 		mark->free_mark(mark);
110 	}
111 }
112 
113 /* Calculate mask of events for a list of marks */
114 u32 fsnotify_recalc_mask(struct hlist_head *head)
115 {
116 	u32 new_mask = 0;
117 	struct fsnotify_mark *mark;
118 
119 	hlist_for_each_entry(mark, head, obj_list)
120 		new_mask |= mark->mask;
121 	return new_mask;
122 }
123 
124 /*
125  * Any time a mark is getting freed we end up here.
126  * The caller had better be holding a reference to this mark so we don't actually
127  * do the final put under the mark->lock
128  */
129 void fsnotify_destroy_mark_locked(struct fsnotify_mark *mark,
130 				  struct fsnotify_group *group)
131 {
132 	struct inode *inode = NULL;
133 
134 	BUG_ON(!mutex_is_locked(&group->mark_mutex));
135 
136 	spin_lock(&mark->lock);
137 
138 	/* something else already called this function on this mark */
139 	if (!(mark->flags & FSNOTIFY_MARK_FLAG_ALIVE)) {
140 		spin_unlock(&mark->lock);
141 		return;
142 	}
143 
144 	mark->flags &= ~FSNOTIFY_MARK_FLAG_ALIVE;
145 
146 	if (mark->flags & FSNOTIFY_MARK_FLAG_INODE) {
147 		inode = mark->inode;
148 		fsnotify_destroy_inode_mark(mark);
149 	} else if (mark->flags & FSNOTIFY_MARK_FLAG_VFSMOUNT)
150 		fsnotify_destroy_vfsmount_mark(mark);
151 	else
152 		BUG();
153 
154 	list_del_init(&mark->g_list);
155 
156 	spin_unlock(&mark->lock);
157 
158 	if (inode && (mark->flags & FSNOTIFY_MARK_FLAG_OBJECT_PINNED))
159 		iput(inode);
160 	/* release lock temporarily */
161 	mutex_unlock(&group->mark_mutex);
162 
163 	spin_lock(&destroy_lock);
164 	list_add(&mark->g_list, &destroy_list);
165 	spin_unlock(&destroy_lock);
166 	wake_up(&destroy_waitq);
167 	/*
168 	 * We don't necessarily have a ref on mark from caller so the above destroy
169 	 * may have actually freed it, unless this group provides a 'freeing_mark'
170 	 * function which must be holding a reference.
171 	 */
172 
173 	/*
174 	 * Some groups like to know that marks are being freed.  This is a
175 	 * callback to the group function to let it know that this mark
176 	 * is being freed.
177 	 */
178 	if (group->ops->freeing_mark)
179 		group->ops->freeing_mark(mark, group);
180 
181 	/*
182 	 * __fsnotify_update_child_dentry_flags(inode);
183 	 *
184 	 * I really want to call that, but we can't, we have no idea if the inode
185 	 * still exists the second we drop the mark->lock.
186 	 *
187 	 * The next time an event arrive to this inode from one of it's children
188 	 * __fsnotify_parent will see that the inode doesn't care about it's
189 	 * children and will update all of these flags then.  So really this
190 	 * is just a lazy update (and could be a perf win...)
191 	 */
192 
193 	atomic_dec(&group->num_marks);
194 
195 	mutex_lock_nested(&group->mark_mutex, SINGLE_DEPTH_NESTING);
196 }
197 
198 void fsnotify_destroy_mark(struct fsnotify_mark *mark,
199 			   struct fsnotify_group *group)
200 {
201 	mutex_lock_nested(&group->mark_mutex, SINGLE_DEPTH_NESTING);
202 	fsnotify_destroy_mark_locked(mark, group);
203 	mutex_unlock(&group->mark_mutex);
204 }
205 
206 /*
207  * Destroy all marks in the given list. The marks must be already detached from
208  * the original inode / vfsmount.
209  */
210 void fsnotify_destroy_marks(struct list_head *to_free)
211 {
212 	struct fsnotify_mark *mark, *lmark;
213 	struct fsnotify_group *group;
214 
215 	list_for_each_entry_safe(mark, lmark, to_free, free_list) {
216 		spin_lock(&mark->lock);
217 		fsnotify_get_group(mark->group);
218 		group = mark->group;
219 		spin_unlock(&mark->lock);
220 
221 		fsnotify_destroy_mark(mark, group);
222 		fsnotify_put_mark(mark);
223 		fsnotify_put_group(group);
224 	}
225 }
226 
227 void fsnotify_set_mark_mask_locked(struct fsnotify_mark *mark, __u32 mask)
228 {
229 	assert_spin_locked(&mark->lock);
230 
231 	mark->mask = mask;
232 
233 	if (mark->flags & FSNOTIFY_MARK_FLAG_INODE)
234 		fsnotify_set_inode_mark_mask_locked(mark, mask);
235 }
236 
237 void fsnotify_set_mark_ignored_mask_locked(struct fsnotify_mark *mark, __u32 mask)
238 {
239 	assert_spin_locked(&mark->lock);
240 
241 	mark->ignored_mask = mask;
242 }
243 
244 /*
245  * Sorting function for lists of fsnotify marks.
246  *
247  * Fanotify supports different notification classes (reflected as priority of
248  * notification group). Events shall be passed to notification groups in
249  * decreasing priority order. To achieve this marks in notification lists for
250  * inodes and vfsmounts are sorted so that priorities of corresponding groups
251  * are descending.
252  *
253  * Furthermore correct handling of the ignore mask requires processing inode
254  * and vfsmount marks of each group together. Using the group address as
255  * further sort criterion provides a unique sorting order and thus we can
256  * merge inode and vfsmount lists of marks in linear time and find groups
257  * present in both lists.
258  *
259  * A return value of 1 signifies that b has priority over a.
260  * A return value of 0 signifies that the two marks have to be handled together.
261  * A return value of -1 signifies that a has priority over b.
262  */
263 int fsnotify_compare_groups(struct fsnotify_group *a, struct fsnotify_group *b)
264 {
265 	if (a == b)
266 		return 0;
267 	if (!a)
268 		return 1;
269 	if (!b)
270 		return -1;
271 	if (a->priority < b->priority)
272 		return 1;
273 	if (a->priority > b->priority)
274 		return -1;
275 	if (a < b)
276 		return 1;
277 	return -1;
278 }
279 
280 /* Add mark into proper place in given list of marks */
281 int fsnotify_add_mark_list(struct hlist_head *head, struct fsnotify_mark *mark,
282 			   int allow_dups)
283 {
284 	struct fsnotify_mark *lmark, *last = NULL;
285 	int cmp;
286 
287 	/* is mark the first mark? */
288 	if (hlist_empty(head)) {
289 		hlist_add_head_rcu(&mark->obj_list, head);
290 		return 0;
291 	}
292 
293 	/* should mark be in the middle of the current list? */
294 	hlist_for_each_entry(lmark, head, obj_list) {
295 		last = lmark;
296 
297 		if ((lmark->group == mark->group) && !allow_dups)
298 			return -EEXIST;
299 
300 		cmp = fsnotify_compare_groups(lmark->group, mark->group);
301 		if (cmp >= 0) {
302 			hlist_add_before_rcu(&mark->obj_list, &lmark->obj_list);
303 			return 0;
304 		}
305 	}
306 
307 	BUG_ON(last == NULL);
308 	/* mark should be the last entry.  last is the current last entry */
309 	hlist_add_behind_rcu(&mark->obj_list, &last->obj_list);
310 	return 0;
311 }
312 
313 /*
314  * Attach an initialized mark to a given group and fs object.
315  * These marks may be used for the fsnotify backend to determine which
316  * event types should be delivered to which group.
317  */
318 int fsnotify_add_mark_locked(struct fsnotify_mark *mark,
319 			     struct fsnotify_group *group, struct inode *inode,
320 			     struct vfsmount *mnt, int allow_dups)
321 {
322 	int ret = 0;
323 
324 	BUG_ON(inode && mnt);
325 	BUG_ON(!inode && !mnt);
326 	BUG_ON(!mutex_is_locked(&group->mark_mutex));
327 
328 	/*
329 	 * LOCKING ORDER!!!!
330 	 * group->mark_mutex
331 	 * mark->lock
332 	 * inode->i_lock
333 	 */
334 	spin_lock(&mark->lock);
335 	mark->flags |= FSNOTIFY_MARK_FLAG_ALIVE;
336 
337 	fsnotify_get_group(group);
338 	mark->group = group;
339 	list_add(&mark->g_list, &group->marks_list);
340 	atomic_inc(&group->num_marks);
341 	fsnotify_get_mark(mark); /* for i_list and g_list */
342 
343 	if (inode) {
344 		ret = fsnotify_add_inode_mark(mark, group, inode, allow_dups);
345 		if (ret)
346 			goto err;
347 	} else if (mnt) {
348 		ret = fsnotify_add_vfsmount_mark(mark, group, mnt, allow_dups);
349 		if (ret)
350 			goto err;
351 	} else {
352 		BUG();
353 	}
354 
355 	/* this will pin the object if appropriate */
356 	fsnotify_set_mark_mask_locked(mark, mark->mask);
357 	spin_unlock(&mark->lock);
358 
359 	if (inode)
360 		__fsnotify_update_child_dentry_flags(inode);
361 
362 	return ret;
363 err:
364 	mark->flags &= ~FSNOTIFY_MARK_FLAG_ALIVE;
365 	list_del_init(&mark->g_list);
366 	fsnotify_put_group(group);
367 	mark->group = NULL;
368 	atomic_dec(&group->num_marks);
369 
370 	spin_unlock(&mark->lock);
371 
372 	spin_lock(&destroy_lock);
373 	list_add(&mark->g_list, &destroy_list);
374 	spin_unlock(&destroy_lock);
375 	wake_up(&destroy_waitq);
376 
377 	return ret;
378 }
379 
380 int fsnotify_add_mark(struct fsnotify_mark *mark, struct fsnotify_group *group,
381 		      struct inode *inode, struct vfsmount *mnt, int allow_dups)
382 {
383 	int ret;
384 	mutex_lock(&group->mark_mutex);
385 	ret = fsnotify_add_mark_locked(mark, group, inode, mnt, allow_dups);
386 	mutex_unlock(&group->mark_mutex);
387 	return ret;
388 }
389 
390 /*
391  * Given a list of marks, find the mark associated with given group. If found
392  * take a reference to that mark and return it, else return NULL.
393  */
394 struct fsnotify_mark *fsnotify_find_mark(struct hlist_head *head,
395 					 struct fsnotify_group *group)
396 {
397 	struct fsnotify_mark *mark;
398 
399 	hlist_for_each_entry(mark, head, obj_list) {
400 		if (mark->group == group) {
401 			fsnotify_get_mark(mark);
402 			return mark;
403 		}
404 	}
405 	return NULL;
406 }
407 
408 /*
409  * clear any marks in a group in which mark->flags & flags is true
410  */
411 void fsnotify_clear_marks_by_group_flags(struct fsnotify_group *group,
412 					 unsigned int flags)
413 {
414 	struct fsnotify_mark *lmark, *mark;
415 
416 	mutex_lock_nested(&group->mark_mutex, SINGLE_DEPTH_NESTING);
417 	list_for_each_entry_safe(mark, lmark, &group->marks_list, g_list) {
418 		if (mark->flags & flags) {
419 			fsnotify_get_mark(mark);
420 			fsnotify_destroy_mark_locked(mark, group);
421 			fsnotify_put_mark(mark);
422 		}
423 	}
424 	mutex_unlock(&group->mark_mutex);
425 }
426 
427 /*
428  * Given a group, destroy all of the marks associated with that group.
429  */
430 void fsnotify_clear_marks_by_group(struct fsnotify_group *group)
431 {
432 	fsnotify_clear_marks_by_group_flags(group, (unsigned int)-1);
433 }
434 
435 void fsnotify_duplicate_mark(struct fsnotify_mark *new, struct fsnotify_mark *old)
436 {
437 	assert_spin_locked(&old->lock);
438 	new->inode = old->inode;
439 	new->mnt = old->mnt;
440 	if (old->group)
441 		fsnotify_get_group(old->group);
442 	new->group = old->group;
443 	new->mask = old->mask;
444 	new->free_mark = old->free_mark;
445 }
446 
447 /*
448  * Nothing fancy, just initialize lists and locks and counters.
449  */
450 void fsnotify_init_mark(struct fsnotify_mark *mark,
451 			void (*free_mark)(struct fsnotify_mark *mark))
452 {
453 	memset(mark, 0, sizeof(*mark));
454 	spin_lock_init(&mark->lock);
455 	atomic_set(&mark->refcnt, 1);
456 	mark->free_mark = free_mark;
457 }
458 
459 static int fsnotify_mark_destroy(void *ignored)
460 {
461 	struct fsnotify_mark *mark, *next;
462 	struct list_head private_destroy_list;
463 
464 	for (;;) {
465 		spin_lock(&destroy_lock);
466 		/* exchange the list head */
467 		list_replace_init(&destroy_list, &private_destroy_list);
468 		spin_unlock(&destroy_lock);
469 
470 		synchronize_srcu(&fsnotify_mark_srcu);
471 
472 		list_for_each_entry_safe(mark, next, &private_destroy_list, g_list) {
473 			list_del_init(&mark->g_list);
474 			fsnotify_put_mark(mark);
475 		}
476 
477 		wait_event_interruptible(destroy_waitq, !list_empty(&destroy_list));
478 	}
479 
480 	return 0;
481 }
482 
483 static int __init fsnotify_mark_init(void)
484 {
485 	struct task_struct *thread;
486 
487 	thread = kthread_run(fsnotify_mark_destroy, NULL,
488 			     "fsnotify_mark");
489 	if (IS_ERR(thread))
490 		panic("unable to start fsnotify mark destruction thread.");
491 
492 	return 0;
493 }
494 device_initcall(fsnotify_mark_init);
495