summaryrefslogtreecommitdiffstats
path: root/fs/btrfs/locking.c
blob: 9063072b399bd833423b7faeb84f523a54ef0f88 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
// SPDX-License-Identifier: GPL-2.0
/*
 * Copyright (C) 2008 Oracle.  All rights reserved.
 */

#include <linux/sched.h>
#include <linux/pagemap.h>
#include <linux/spinlock.h>
#include <linux/page-flags.h>
#include <asm/bug.h>
#include "misc.h"
#include "ctree.h"
#include "extent_io.h"
#include "locking.h"

/*
 * Lockdep class keys for extent_buffer->lock's in this root.  For a given
 * eb, the lockdep key is determined by the btrfs_root it belongs to and
 * the level the eb occupies in the tree.
 *
 * Different roots are used for different purposes and may nest inside each
 * other and they require separate keysets.  As lockdep keys should be
 * static, assign keysets according to the purpose of the root as indicated
 * by btrfs_root->root_key.objectid.  This ensures that all special purpose
 * roots have separate keysets.
 *
 * Lock-nesting across peer nodes is always done with the immediate parent
 * node locked thus preventing deadlock.  As lockdep doesn't know this, use
 * subclass to avoid triggering lockdep warning in such cases.
 *
 * The key is set by the readpage_end_io_hook after the buffer has passed
 * csum validation but before the pages are unlocked.  It is also set by
 * btrfs_init_new_buffer on freshly allocated blocks.
 *
 * We also add a check to make sure the highest level of the tree is the
 * same as our lockdep setup here.  If BTRFS_MAX_LEVEL changes, this code
 * needs update as well.
 */
#ifdef CONFIG_DEBUG_LOCK_ALLOC
#if BTRFS_MAX_LEVEL != 8
#error
#endif

#define DEFINE_LEVEL(stem, level)					\
	.names[level] = "btrfs-" stem "-0" #level,

#define DEFINE_NAME(stem)						\
	DEFINE_LEVEL(stem, 0)						\
	DEFINE_LEVEL(stem, 1)						\
	DEFINE_LEVEL(stem, 2)						\
	DEFINE_LEVEL(stem, 3)						\
	DEFINE_LEVEL(stem, 4)						\
	DEFINE_LEVEL(stem, 5)						\
	DEFINE_LEVEL(stem, 6)						\
	DEFINE_LEVEL(stem, 7)

static struct btrfs_lockdep_keyset {
	u64			id;		/* root objectid */
	/* Longest entry: btrfs-free-space-00 */
	char			names[BTRFS_MAX_LEVEL][20];
	struct lock_class_key	keys[BTRFS_MAX_LEVEL];
} btrfs_lockdep_keysets[] = {
	{ .id = BTRFS_ROOT_TREE_OBJECTID,	DEFINE_NAME("root")	},
	{ .id = BTRFS_EXTENT_TREE_OBJECTID,	DEFINE_NAME("extent")	},
	{ .id = BTRFS_CHUNK_TREE_OBJECTID,	DEFINE_NAME("chunk")	},
	{ .id = BTRFS_DEV_TREE_OBJECTID,	DEFINE_NAME("dev")	},
	{ .id = BTRFS_CSUM_TREE_OBJECTID,	DEFINE_NAME("csum")	},
	{ .id = BTRFS_QUOTA_TREE_OBJECTID,	DEFINE_NAME("quota")	},
	{ .id = BTRFS_TREE_LOG_OBJECTID,	DEFINE_NAME("log")	},
	{ .id = BTRFS_TREE_RELOC_OBJECTID,	DEFINE_NAME("treloc")	},
	{ .id = BTRFS_DATA_RELOC_TREE_OBJECTID,	DEFINE_NAME("dreloc")	},
	{ .id = BTRFS_UUID_TREE_OBJECTID,	DEFINE_NAME("uuid")	},
	{ .id = BTRFS_FREE_SPACE_TREE_OBJECTID,	DEFINE_NAME("free-space") },
	{ .id = 0,				DEFINE_NAME("tree")	},
};

#undef DEFINE_LEVEL
#undef DEFINE_NAME

void btrfs_set_buffer_lockdep_class(u64 objectid, struct extent_buffer *eb, int level)
{
	struct btrfs_lockdep_keyset *ks;

	BUG_ON(level >= ARRAY_SIZE(ks->keys));

	/* Find the matching keyset, id 0 is the default entry */
	for (ks = btrfs_lockdep_keysets; ks->id; ks++)
		if (ks->id == objectid)
			break;

	lockdep_set_class_and_name(&eb->lock, &ks->keys[level], ks->names[level]);
}

void btrfs_maybe_reset_lockdep_class(struct btrfs_root *root, struct extent_buffer *eb)
{
	if (test_bit(BTRFS_ROOT_RESET_LOCKDEP_CLASS, &root->state))
		btrfs_set_buffer_lockdep_class(root->root_key.objectid,
					       eb, btrfs_header_level(eb));
}

#endif

/*
 * Extent buffer locking
 * =====================
 *
 * We use a rw_semaphore for tree locking, and the semantics are exactly the
 * same:
 *
 * - reader/writer exclusion
 * - writer/writer exclusion
 * - reader/reader sharing
 * - try-lock semantics for readers and writers
 *
 * The rwsem implementation does opportunistic spinning which reduces number of
 * times the locking task needs to sleep.
 */

/*
 * __btrfs_tree_read_lock - lock extent buffer for read
 * @eb:		the eb to be locked
 * @nest:	the nesting level to be used for lockdep
 *
 * This takes the read lock on the extent buffer, using the specified nesting
 * level for lockdep purposes.
 */
void __btrfs_tree_read_lock(struct extent_buffer *eb, enum btrfs_lock_nesting nest)
{
	u64 start_ns = 0;

	if (trace_btrfs_tree_read_lock_enabled())
		start_ns = ktime_get_ns();

	down_read_nested(&eb->lock, nest);
	trace_btrfs_tree_read_lock(eb, start_ns);
}

void btrfs_tree_read_lock(struct extent_buffer *eb)
{
	__btrfs_tree_read_lock(eb, BTRFS_NESTING_NORMAL);
}

/*
 * Try-lock for read.
 *
 * Return 1 if the rwlock has been taken, 0 otherwise
 */
int btrfs_try_tree_read_lock(struct extent_buffer *eb)
{
	if (down_read_trylock(&eb->lock)) {
		trace_btrfs_try_tree_read_lock(eb);
		return 1;
	}
	return 0;
}

/*
 * Try-lock for write.
 *
 * Return 1 if the rwlock has been taken, 0 otherwise
 */
int btrfs_try_tree_write_lock(struct extent_buffer *eb)
{
	if (down_write_trylock(&eb->lock)) {
		eb->lock_owner = current->pid;
		trace_btrfs_try_tree_write_lock(eb);
		return 1;
	}
	return 0;
}

/*
 * Release read lock.
 */
void btrfs_tree_read_unlock(struct extent_buffer *eb)
{
	trace_btrfs_tree_read_unlock(eb);
	up_read(&eb->lock);
}

/*
 * __btrfs_tree_lock - lock eb for write
 * @eb:		the eb to lock
 * @nest:	the nesting to use for the lock
 *
 * Returns with the eb->lock write locked.
 */
void __btrfs_tree_lock(struct extent_buffer *eb, enum btrfs_lock_nesting nest)
	__acquires(&eb->lock)
{
	u64 start_ns = 0;

	if (trace_btrfs_tree_lock_enabled())
		start_ns = ktime_get_ns();

	down_write_nested(&eb->lock, nest);
	eb->lock_owner = current->pid;
	trace_btrfs_tree_lock(eb, start_ns);
}

void btrfs_tree_lock(struct extent_buffer *eb)
{
	__btrfs_tree_lock(eb, BTRFS_NESTING_NORMAL);
}

/*
 * Release the write lock.
 */
void btrfs_tree_unlock(struct extent_buffer *eb)
{
	trace_btrfs_tree_unlock(eb);
	eb->lock_owner = 0;
	up_write(&eb->lock);
}

/*
 * This releases any locks held in the path starting at level and going all the
 * way up to the root.
 *
 * btrfs_search_slot will keep the lock held on higher nodes in a few corner
 * cases, such as COW of the block at slot zero in the node.  This ignores
 * those rules, and it should only be called when there are no more updates to
 * be done higher up in the tree.
 */
void btrfs_unlock_up_safe(struct btrfs_path *path, int level)
{
	int i;

	if (path->keep_locks)
		return;

	for (i = level; i < BTRFS_MAX_LEVEL; i++) {
		if (!path->nodes[i])
			continue;
		if (!path->locks[i])
			continue;
		btrfs_tree_unlock_rw(path->nodes[i], path->locks[i]);
		path->locks[i] = 0;
	}
}

/*
 * Loop around taking references on and locking the root node of the tree until
 * we end up with a lock on the root node.
 *
 * Return: root extent buffer with write lock held
 */
struct extent_buffer *btrfs_lock_root_node(struct btrfs_root *root)
{
	struct extent_buffer *eb;

	while (1) {
		eb = btrfs_root_node(root);

		btrfs_maybe_reset_lockdep_class(root, eb);
		btrfs_tree_lock(eb);
		if (eb == root->node)
			break;
		btrfs_tree_unlock(eb);
		free_extent_buffer(eb);
	}
	return eb;
}

/*
 * Loop around taking references on and locking the root node of the tree until
 * we end up with a lock on the root node.
 *
 * Return: root extent buffer with read lock held
 */
struct extent_buffer *btrfs_read_lock_root_node(struct btrfs_root *root)
{
	struct extent_buffer *eb;

	while (1) {
		eb = btrfs_root_node(root);

		btrfs_maybe_reset_lockdep_class(root, eb);
		btrfs_tree_read_lock(eb);
		if (eb == root->node)
			break;
		btrfs_tree_read_unlock(eb);
		free_extent_buffer(eb);
	}
	return eb;
}

/*
 * DREW locks
 * ==========
 *
 * DREW stands for double-reader-writer-exclusion lock. It's used in situation
 * where you want to provide A-B exclusion but not AA or BB.
 *
 * Currently implementation gives more priority to reader. If a reader and a
 * writer both race to acquire their respective sides of the lock the writer
 * would yield its lock as soon as it detects a concurrent reader. Additionally
 * if there are pending readers no new writers would be allowed to come in and
 * acquire the lock.
 */

int btrfs_drew_lock_init(struct btrfs_drew_lock *lock)
{
	int ret;

	ret = percpu_counter_init(&lock->writers, 0, GFP_KERNEL);
	if (ret)
		return ret;

	atomic_set(&lock->readers, 0);
	init_waitqueue_head(&lock->pending_readers);
	init_waitqueue_head(&lock->pending_writers);

	return 0;
}

void btrfs_drew_lock_destroy(struct btrfs_drew_lock *lock)
{
	percpu_counter_destroy(&lock->writers);
}

/* Return true if acquisition is successful, false otherwise */
bool btrfs_drew_try_write_lock(struct btrfs_drew_lock *lock)
{
	if (atomic_read(&lock->readers))
		return false;

	percpu_counter_inc(&lock->writers);

	/* Ensure writers count is updated before we check for pending readers */
	smp_mb();
	if (atomic_read(&lock->readers)) {
		btrfs_drew_write_unlock(lock);
		return false;
	}

	return true;
}

void btrfs_drew_write_lock(struct btrfs_drew_lock *lock)
{
	while (true) {
		if (btrfs_drew_try_write_lock(lock))
			return;
		wait_event(lock->pending_writers, !atomic_read(&lock->readers));
	}
}

void btrfs_drew_write_unlock(struct btrfs_drew_lock *lock)
{
	percpu_counter_dec(&lock->writers);
	cond_wake_up(&lock->pending_readers);
}

void btrfs_drew_read_lock(struct btrfs_drew_lock *lock)
{
	atomic_inc(&lock->readers);

	/*
	 * Ensure the pending reader count is perceieved BEFORE this reader
	 * goes to sleep in case of active writers. This guarantees new writers
	 * won't be allowed and that the current reader will be woken up when
	 * the last active writer finishes its jobs.
	 */
	smp_mb__after_atomic();

	wait_event(lock->pending_readers,
		   percpu_counter_sum(&lock->writers) == 0);
}

void btrfs_drew_read_unlock(struct btrfs_drew_lock *lock)
{
	/*
	 * atomic_dec_and_test implies a full barrier, so woken up writers
	 * are guaranteed to see the decrement
	 */
	if (atomic_dec_and_test(&lock->readers))
		wake_up(&lock->pending_writers);
}