?  search:
summary | log | graphiclog1 | graphiclog2 | commit | commitdiff | tree | refs | edit | fork
blame | history | raw | HEAD
ext4: fix a potential fiemap/page fault deadlock w/ inline_data
[linux/fpc-iii.git] / fs / btrfs / delayed-inode.c
blob09fa5af9782e0dd03a8815d012d963b725934eb3
1 /*
2 * Copyright (C) 2011 Fujitsu. All rights reserved.
3 * Written by Miao Xie <miaox@cn.fujitsu.com>
4 *
5 * This program is free software; you can redistribute it and/or
6 * modify it under the terms of the GNU General Public
7 * License v2 as published by the Free Software Foundation.
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 GNU
12 * General Public License for more details.
13 *
14 * You should have received a copy of the GNU General Public
15 * License along with this program; if not, write to the
16 * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
17 * Boston, MA 021110-1307, USA.
18 */
19
20 #include <linux/slab.h>
21 #include "delayed-inode.h"
22 #include "disk-io.h"
23 #include "transaction.h"
24 #include "ctree.h"
25
26 #define BTRFS_DELAYED_WRITEBACK 512
27 #define BTRFS_DELAYED_BACKGROUND 128
28 #define BTRFS_DELAYED_BATCH 16
29
30 static struct kmem_cache *delayed_node_cache;
31
32 int __init btrfs_delayed_inode_init(void)
33 {
34 delayed_node_cache = kmem_cache_create("btrfs_delayed_node",
35 sizeof(struct btrfs_delayed_node),
36 0,
37 SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD,
38 NULL);
39 if (!delayed_node_cache)
40 return -ENOMEM;
41 return 0;
42 }
43
44 void btrfs_delayed_inode_exit(void)
45 {
46 if (delayed_node_cache)
47 kmem_cache_destroy(delayed_node_cache);
48 }
49
50 static inline void btrfs_init_delayed_node(
51 struct btrfs_delayed_node *delayed_node,
52 struct btrfs_root *root, u64 inode_id)
53 {
54 delayed_node->root = root;
55 delayed_node->inode_id = inode_id;
56 atomic_set(&delayed_node->refs, 0);
57 delayed_node->count = 0;
58 delayed_node->flags = 0;
59 delayed_node->ins_root = RB_ROOT;
60 delayed_node->del_root = RB_ROOT;
61 mutex_init(&delayed_node->mutex);
62 delayed_node->index_cnt = 0;
63 INIT_LIST_HEAD(&delayed_node->n_list);
64 INIT_LIST_HEAD(&delayed_node->p_list);
65 delayed_node->bytes_reserved = 0;
66 memset(&delayed_node->inode_item, 0, sizeof(delayed_node->inode_item));
67 }
68
69 static inline int btrfs_is_continuous_delayed_item(
70 struct btrfs_delayed_item *item1,
71 struct btrfs_delayed_item *item2)
72 {
73 if (item1->key.type == BTRFS_DIR_INDEX_KEY &&
74 item1->key.objectid == item2->key.objectid &&
75 item1->key.type == item2->key.type &&
76 item1->key.offset + 1 == item2->key.offset)
77 return 1;
78 return 0;
79 }
80
81 static inline struct btrfs_delayed_root *btrfs_get_delayed_root(
82 struct btrfs_root *root)
83 {
84 return root->fs_info->delayed_root;
85 }
86
87 static struct btrfs_delayed_node *btrfs_get_delayed_node(struct inode *inode)
88 {
89 struct btrfs_inode *btrfs_inode = BTRFS_I(inode);
90 struct btrfs_root *root = btrfs_inode->root;
91 u64 ino = btrfs_ino(inode);
92 struct btrfs_delayed_node *node;
93
94 node = ACCESS_ONCE(btrfs_inode->delayed_node);
95 if (node) {
96 atomic_inc(&node->refs);
97 return node;
98 }
99
100 spin_lock(&root->inode_lock);
101 node = radix_tree_lookup(&root->delayed_nodes_tree, ino);
102 if (node) {
103 if (btrfs_inode->delayed_node) {
104 atomic_inc(&node->refs); /* can be accessed */
105 BUG_ON(btrfs_inode->delayed_node != node);
106 spin_unlock(&root->inode_lock);
107 return node;
108 }
109 btrfs_inode->delayed_node = node;
110 /* can be accessed and cached in the inode */
111 atomic_add(2, &node->refs);
112 spin_unlock(&root->inode_lock);
113 return node;
114 }
115 spin_unlock(&root->inode_lock);
116
117 return NULL;
118 }
119
120 /* Will return either the node or PTR_ERR(-ENOMEM) */
121 static struct btrfs_delayed_node *btrfs_get_or_create_delayed_node(
122 struct inode *inode)
123 {
124 struct btrfs_delayed_node *node;
125 struct btrfs_inode *btrfs_inode = BTRFS_I(inode);
126 struct btrfs_root *root = btrfs_inode->root;
127 u64 ino = btrfs_ino(inode);
128 int ret;
129
130 again:
131 node = btrfs_get_delayed_node(inode);
132 if (node)
133 return node;
134
135 node = kmem_cache_alloc(delayed_node_cache, GFP_NOFS);
136 if (!node)
137 return ERR_PTR(-ENOMEM);
138 btrfs_init_delayed_node(node, root, ino);
139
140 /* cached in the btrfs inode and can be accessed */
141 atomic_add(2, &node->refs);
142
143 ret = radix_tree_preload(GFP_NOFS & ~__GFP_HIGHMEM);
144 if (ret) {
145 kmem_cache_free(delayed_node_cache, node);
146 return ERR_PTR(ret);
147 }
148
149 spin_lock(&root->inode_lock);
150 ret = radix_tree_insert(&root->delayed_nodes_tree, ino, node);
151 if (ret == -EEXIST) {
152 spin_unlock(&root->inode_lock);
153 kmem_cache_free(delayed_node_cache, node);
154 radix_tree_preload_end();
155 goto again;
156 }
157 btrfs_inode->delayed_node = node;
158 spin_unlock(&root->inode_lock);
159 radix_tree_preload_end();
160
161 return node;
162 }
163
164 /*
165 * Call it when holding delayed_node->mutex
166 *
167 * If mod = 1, add this node into the prepared list.
168 */
169 static void btrfs_queue_delayed_node(struct btrfs_delayed_root *root,
170 struct btrfs_delayed_node *node,
171 int mod)
172 {
173 spin_lock(&root->lock);
174 if (test_bit(BTRFS_DELAYED_NODE_IN_LIST, &node->flags)) {
175 if (!list_empty(&node->p_list))
176 list_move_tail(&node->p_list, &root->prepare_list);
177 else if (mod)
178 list_add_tail(&node->p_list, &root->prepare_list);
179 } else {
180 list_add_tail(&node->n_list, &root->node_list);
181 list_add_tail(&node->p_list, &root->prepare_list);
182 atomic_inc(&node->refs); /* inserted into list */
183 root->nodes++;
184 set_bit(BTRFS_DELAYED_NODE_IN_LIST, &node->flags);
185 }
186 spin_unlock(&root->lock);
187 }
188
189 /* Call it when holding delayed_node->mutex */
190 static void btrfs_dequeue_delayed_node(struct btrfs_delayed_root *root,
191 struct btrfs_delayed_node *node)
192 {
193 spin_lock(&root->lock);
194 if (test_bit(BTRFS_DELAYED_NODE_IN_LIST, &node->flags)) {
195 root->nodes--;
196 atomic_dec(&node->refs); /* not in the list */
197 list_del_init(&node->n_list);
198 if (!list_empty(&node->p_list))
199 list_del_init(&node->p_list);
200 clear_bit(BTRFS_DELAYED_NODE_IN_LIST, &node->flags);
201 }
202 spin_unlock(&root->lock);
203 }
204
205 static struct btrfs_delayed_node *btrfs_first_delayed_node(
206 struct btrfs_delayed_root *delayed_root)
207 {
208 struct list_head *p;
209 struct btrfs_delayed_node *node = NULL;
210
211 spin_lock(&delayed_root->lock);
212 if (list_empty(&delayed_root->node_list))
213 goto out;
214
215 p = delayed_root->node_list.next;
216 node = list_entry(p, struct btrfs_delayed_node, n_list);
217 atomic_inc(&node->refs);
218 out:
219 spin_unlock(&delayed_root->lock);
220
221 return node;
222 }
223
224 static struct btrfs_delayed_node *btrfs_next_delayed_node(
225 struct btrfs_delayed_node *node)
226 {
227 struct btrfs_delayed_root *delayed_root;
228 struct list_head *p;
229 struct btrfs_delayed_node *next = NULL;
230
231 delayed_root = node->root->fs_info->delayed_root;
232 spin_lock(&delayed_root->lock);
233 if (!test_bit(BTRFS_DELAYED_NODE_IN_LIST, &node->flags)) {
234 /* not in the list */
235 if (list_empty(&delayed_root->node_list))
236 goto out;
237 p = delayed_root->node_list.next;
238 } else if (list_is_last(&node->n_list, &delayed_root->node_list))
239 goto out;
240 else
241 p = node->n_list.next;
242
243 next = list_entry(p, struct btrfs_delayed_node, n_list);
244 atomic_inc(&next->refs);
245 out:
246 spin_unlock(&delayed_root->lock);
247
248 return next;
249 }
250
251 static void __btrfs_release_delayed_node(
252 struct btrfs_delayed_node *delayed_node,
253 int mod)
254 {
255 struct btrfs_delayed_root *delayed_root;
256
257 if (!delayed_node)
258 return;
259
260 delayed_root = delayed_node->root->fs_info->delayed_root;
261
262 mutex_lock(&delayed_node->mutex);
263 if (delayed_node->count)
264 btrfs_queue_delayed_node(delayed_root, delayed_node, mod);
265 else
266 btrfs_dequeue_delayed_node(delayed_root, delayed_node);
267 mutex_unlock(&delayed_node->mutex);
268
269 if (atomic_dec_and_test(&delayed_node->refs)) {
270 bool free = false;
271 struct btrfs_root *root = delayed_node->root;
272 spin_lock(&root->inode_lock);
273 if (atomic_read(&delayed_node->refs) == 0) {
274 radix_tree_delete(&root->delayed_nodes_tree,
275 delayed_node->inode_id);
276 free = true;
277 }
278 spin_unlock(&root->inode_lock);
279 if (free)
280 kmem_cache_free(delayed_node_cache, delayed_node);
281 }
282 }
283
284 static inline void btrfs_release_delayed_node(struct btrfs_delayed_node *node)
285 {
286 __btrfs_release_delayed_node(node, 0);
287 }
288
289 static struct btrfs_delayed_node *btrfs_first_prepared_delayed_node(
290 struct btrfs_delayed_root *delayed_root)
291 {
292 struct list_head *p;
293 struct btrfs_delayed_node *node = NULL;
294
295 spin_lock(&delayed_root->lock);
296 if (list_empty(&delayed_root->prepare_list))
297 goto out;
298
299 p = delayed_root->prepare_list.next;
300 list_del_init(p);
301 node = list_entry(p, struct btrfs_delayed_node, p_list);
302 atomic_inc(&node->refs);
303 out:
304 spin_unlock(&delayed_root->lock);
305
306 return node;
307 }
308
309 static inline void btrfs_release_prepared_delayed_node(
310 struct btrfs_delayed_node *node)
311 {
312 __btrfs_release_delayed_node(node, 1);
313 }
314
315 static struct btrfs_delayed_item *btrfs_alloc_delayed_item(u32 data_len)
316 {
317 struct btrfs_delayed_item *item;
318 item = kmalloc(sizeof(*item) + data_len, GFP_NOFS);
319 if (item) {
320 item->data_len = data_len;
321 item->ins_or_del = 0;
322 item->bytes_reserved = 0;
323 item->delayed_node = NULL;
324 atomic_set(&item->refs, 1);
325 }
326 return item;
327 }
328
329 /*
330 * __btrfs_lookup_delayed_item - look up the delayed item by key
331 * @delayed_node: pointer to the delayed node
332 * @key: the key to look up
333 * @prev: used to store the prev item if the right item isn't found
334 * @next: used to store the next item if the right item isn't found
335 *
336 * Note: if we don't find the right item, we will return the prev item and
337 * the next item.
338 */
339 static struct btrfs_delayed_item *__btrfs_lookup_delayed_item(
340 struct rb_root *root,
341 struct btrfs_key *key,
342 struct btrfs_delayed_item **prev,
343 struct btrfs_delayed_item **next)
344 {
345 struct rb_node *node, *prev_node = NULL;
346 struct btrfs_delayed_item *delayed_item = NULL;
347 int ret = 0;
348
349 node = root->rb_node;
350
351 while (node) {
352 delayed_item = rb_entry(node, struct btrfs_delayed_item,
353 rb_node);
354 prev_node = node;
355 ret = btrfs_comp_cpu_keys(&delayed_item->key, key);
356 if (ret < 0)
357 node = node->rb_right;
358 else if (ret > 0)
359 node = node->rb_left;
360 else
361 return delayed_item;
362 }
363
364 if (prev) {
365 if (!prev_node)
366 *prev = NULL;
367 else if (ret < 0)
368 *prev = delayed_item;
369 else if ((node = rb_prev(prev_node)) != NULL) {
370 *prev = rb_entry(node, struct btrfs_delayed_item,
371 rb_node);
372 } else
373 *prev = NULL;
374 }
375
376 if (next) {
377 if (!prev_node)
378 *next = NULL;
379 else if (ret > 0)
380 *next = delayed_item;
381 else if ((node = rb_next(prev_node)) != NULL) {
382 *next = rb_entry(node, struct btrfs_delayed_item,
383 rb_node);
384 } else
385 *next = NULL;
386 }
387 return NULL;
388 }
389
390 static struct btrfs_delayed_item *__btrfs_lookup_delayed_insertion_item(
391 struct btrfs_delayed_node *delayed_node,
392 struct btrfs_key *key)
393 {
394 struct btrfs_delayed_item *item;
395
396 item = __btrfs_lookup_delayed_item(&delayed_node->ins_root, key,
397 NULL, NULL);
398 return item;
399 }
400
401 static int __btrfs_add_delayed_item(struct btrfs_delayed_node *delayed_node,
402 struct btrfs_delayed_item *ins,
403 int action)
404 {
405 struct rb_node **p, *node;
406 struct rb_node *parent_node = NULL;
407 struct rb_root *root;
408 struct btrfs_delayed_item *item;
409 int cmp;
410
411 if (action == BTRFS_DELAYED_INSERTION_ITEM)
412 root = &delayed_node->ins_root;
413 else if (action == BTRFS_DELAYED_DELETION_ITEM)
414 root = &delayed_node->del_root;
415 else
416 BUG();
417 p = &root->rb_node;
418 node = &ins->rb_node;
419
420 while (*p) {
421 parent_node = *p;
422 item = rb_entry(parent_node, struct btrfs_delayed_item,
423 rb_node);
424
425 cmp = btrfs_comp_cpu_keys(&item->key, &ins->key);
426 if (cmp < 0)
427 p = &(*p)->rb_right;
428 else if (cmp > 0)
429 p = &(*p)->rb_left;
430 else
431 return -EEXIST;
432 }
433
434 rb_link_node(node, parent_node, p);
435 rb_insert_color(node, root);
436 ins->delayed_node = delayed_node;
437 ins->ins_or_del = action;
438
439 if (ins->key.type == BTRFS_DIR_INDEX_KEY &&
440 action == BTRFS_DELAYED_INSERTION_ITEM &&
441 ins->key.offset >= delayed_node->index_cnt)
442 delayed_node->index_cnt = ins->key.offset + 1;
443
444 delayed_node->count++;
445 atomic_inc(&delayed_node->root->fs_info->delayed_root->items);
446 return 0;
447 }
448
449 static int __btrfs_add_delayed_insertion_item(struct btrfs_delayed_node *node,
450 struct btrfs_delayed_item *item)
451 {
452 return __btrfs_add_delayed_item(node, item,
453 BTRFS_DELAYED_INSERTION_ITEM);
454 }
455
456 static int __btrfs_add_delayed_deletion_item(struct btrfs_delayed_node *node,
457 struct btrfs_delayed_item *item)
458 {
459 return __btrfs_add_delayed_item(node, item,
460 BTRFS_DELAYED_DELETION_ITEM);
461 }
462
463 static void finish_one_item(struct btrfs_delayed_root *delayed_root)
464 {
465 int seq = atomic_inc_return(&delayed_root->items_seq);
466
467 /*
468 * atomic_dec_return implies a barrier for waitqueue_active
469 */
470 if ((atomic_dec_return(&delayed_root->items) <
471 BTRFS_DELAYED_BACKGROUND || seq % BTRFS_DELAYED_BATCH == 0) &&
472 waitqueue_active(&delayed_root->wait))
473 wake_up(&delayed_root->wait);
474 }
475
476 static void __btrfs_remove_delayed_item(struct btrfs_delayed_item *delayed_item)
477 {
478 struct rb_root *root;
479 struct btrfs_delayed_root *delayed_root;
480
481 delayed_root = delayed_item->delayed_node->root->fs_info->delayed_root;
482
483 BUG_ON(!delayed_root);
484 BUG_ON(delayed_item->ins_or_del != BTRFS_DELAYED_DELETION_ITEM &&
485 delayed_item->ins_or_del != BTRFS_DELAYED_INSERTION_ITEM);
486
487 if (delayed_item->ins_or_del == BTRFS_DELAYED_INSERTION_ITEM)
488 root = &delayed_item->delayed_node->ins_root;
489 else
490 root = &delayed_item->delayed_node->del_root;
491
492 rb_erase(&delayed_item->rb_node, root);
493 delayed_item->delayed_node->count--;
494
495 finish_one_item(delayed_root);
496 }
497
498 static void btrfs_release_delayed_item(struct btrfs_delayed_item *item)
499 {
500 if (item) {
501 __btrfs_remove_delayed_item(item);
502 if (atomic_dec_and_test(&item->refs))
503 kfree(item);
504 }
505 }
506
507 static struct btrfs_delayed_item *__btrfs_first_delayed_insertion_item(
508 struct btrfs_delayed_node *delayed_node)
509 {
510 struct rb_node *p;
511 struct btrfs_delayed_item *item = NULL;
512
513 p = rb_first(&delayed_node->ins_root);
514 if (p)
515 item = rb_entry(p, struct btrfs_delayed_item, rb_node);
516
517 return item;
518 }
519
520 static struct btrfs_delayed_item *__btrfs_first_delayed_deletion_item(
521 struct btrfs_delayed_node *delayed_node)
522 {
523 struct rb_node *p;
524 struct btrfs_delayed_item *item = NULL;
525
526 p = rb_first(&delayed_node->del_root);
527 if (p)
528 item = rb_entry(p, struct btrfs_delayed_item, rb_node);
529
530 return item;
531 }
532
533 static struct btrfs_delayed_item *__btrfs_next_delayed_item(
534 struct btrfs_delayed_item *item)
535 {
536 struct rb_node *p;
537 struct btrfs_delayed_item *next = NULL;
538
539 p = rb_next(&item->rb_node);
540 if (p)
541 next = rb_entry(p, struct btrfs_delayed_item, rb_node);
542
543 return next;
544 }
545
546 static int btrfs_delayed_item_reserve_metadata(struct btrfs_trans_handle *trans,
547 struct btrfs_root *root,
548 struct btrfs_delayed_item *item)
549 {
550 struct btrfs_block_rsv *src_rsv;
551 struct btrfs_block_rsv *dst_rsv;
552 u64 num_bytes;
553 int ret;
554
555 if (!trans->bytes_reserved)
556 return 0;
557
558 src_rsv = trans->block_rsv;
559 dst_rsv = &root->fs_info->delayed_block_rsv;
560
561 num_bytes = btrfs_calc_trans_metadata_size(root, 1);
562 ret = btrfs_block_rsv_migrate(src_rsv, dst_rsv, num_bytes);
563 if (!ret) {
564 trace_btrfs_space_reservation(root->fs_info, "delayed_item",
565 item->key.objectid,
566 num_bytes, 1);
567 item->bytes_reserved = num_bytes;
568 }
569
570 return ret;
571 }
572
573 static void btrfs_delayed_item_release_metadata(struct btrfs_root *root,
574 struct btrfs_delayed_item *item)
575 {
576 struct btrfs_block_rsv *rsv;
577
578 if (!item->bytes_reserved)
579 return;
580
581 rsv = &root->fs_info->delayed_block_rsv;
582 trace_btrfs_space_reservation(root->fs_info, "delayed_item",
583 item->key.objectid, item->bytes_reserved,
584 0);
585 btrfs_block_rsv_release(root, rsv,
586 item->bytes_reserved);
587 }
588
589 static int btrfs_delayed_inode_reserve_metadata(
590 struct btrfs_trans_handle *trans,
591 struct btrfs_root *root,
592 struct inode *inode,
593 struct btrfs_delayed_node *node)
594 {
595 struct btrfs_block_rsv *src_rsv;
596 struct btrfs_block_rsv *dst_rsv;
597 u64 num_bytes;
598 int ret;
599 bool release = false;
600
601 src_rsv = trans->block_rsv;
602 dst_rsv = &root->fs_info->delayed_block_rsv;
603
604 num_bytes = btrfs_calc_trans_metadata_size(root, 1);
605
606 /*
607 * btrfs_dirty_inode will update the inode under btrfs_join_transaction
608 * which doesn't reserve space for speed. This is a problem since we
609 * still need to reserve space for this update, so try to reserve the
610 * space.
611 *
612 * Now if src_rsv == delalloc_block_rsv we'll let it just steal since
613 * we're accounted for.
614 */
615 if (!src_rsv || (!trans->bytes_reserved &&
616 src_rsv->type != BTRFS_BLOCK_RSV_DELALLOC)) {
617 ret = btrfs_block_rsv_add(root, dst_rsv, num_bytes,
618 BTRFS_RESERVE_NO_FLUSH);
619 /*
620 * Since we're under a transaction reserve_metadata_bytes could
621 * try to commit the transaction which will make it return
622 * EAGAIN to make us stop the transaction we have, so return
623 * ENOSPC instead so that btrfs_dirty_inode knows what to do.
624 */
625 if (ret == -EAGAIN)
626 ret = -ENOSPC;
627 if (!ret) {
628 node->bytes_reserved = num_bytes;
629 trace_btrfs_space_reservation(root->fs_info,
630 "delayed_inode",
631 btrfs_ino(inode),
632 num_bytes, 1);
633 }
634 return ret;
635 } else if (src_rsv->type == BTRFS_BLOCK_RSV_DELALLOC) {
636 spin_lock(&BTRFS_I(inode)->lock);
637 if (test_and_clear_bit(BTRFS_INODE_DELALLOC_META_RESERVED,
638 &BTRFS_I(inode)->runtime_flags)) {
639 spin_unlock(&BTRFS_I(inode)->lock);
640 release = true;
641 goto migrate;
642 }
643 spin_unlock(&BTRFS_I(inode)->lock);
644
645 /* Ok we didn't have space pre-reserved. This shouldn't happen
646 * too often but it can happen if we do delalloc to an existing
647 * inode which gets dirtied because of the time update, and then
648 * isn't touched again until after the transaction commits and
649 * then we try to write out the data. First try to be nice and
650 * reserve something strictly for us. If not be a pain and try
651 * to steal from the delalloc block rsv.
652 */
653 ret = btrfs_block_rsv_add(root, dst_rsv, num_bytes,
654 BTRFS_RESERVE_NO_FLUSH);
655 if (!ret)
656 goto out;
657
658 ret = btrfs_block_rsv_migrate(src_rsv, dst_rsv, num_bytes);
659 if (!WARN_ON(ret))
660 goto out;
661
662 /*
663 * Ok this is a problem, let's just steal from the global rsv
664 * since this really shouldn't happen that often.
665 */
666 ret = btrfs_block_rsv_migrate(&root->fs_info->global_block_rsv,
667 dst_rsv, num_bytes);
668 goto out;
669 }
670
671 migrate:
672 ret = btrfs_block_rsv_migrate(src_rsv, dst_rsv, num_bytes);
673
674 out:
675 /*
676 * Migrate only takes a reservation, it doesn't touch the size of the
677 * block_rsv. This is to simplify people who don't normally have things
678 * migrated from their block rsv. If they go to release their
679 * reservation, that will decrease the size as well, so if migrate
680 * reduced size we'd end up with a negative size. But for the
681 * delalloc_meta_reserved stuff we will only know to drop 1 reservation,
682 * but we could in fact do this reserve/migrate dance several times
683 * between the time we did the original reservation and we'd clean it
684 * up. So to take care of this, release the space for the meta
685 * reservation here. I think it may be time for a documentation page on
686 * how block rsvs. work.
687 */
688 if (!ret) {
689 trace_btrfs_space_reservation(root->fs_info, "delayed_inode",
690 btrfs_ino(inode), num_bytes, 1);
691 node->bytes_reserved = num_bytes;
692 }
693
694 if (release) {
695 trace_btrfs_space_reservation(root->fs_info, "delalloc",
696 btrfs_ino(inode), num_bytes, 0);
697 btrfs_block_rsv_release(root, src_rsv, num_bytes);
698 }
699
700 return ret;
701 }
702
703 static void btrfs_delayed_inode_release_metadata(struct btrfs_root *root,
704 struct btrfs_delayed_node *node)
705 {
706 struct btrfs_block_rsv *rsv;
707
708 if (!node->bytes_reserved)
709 return;
710
711 rsv = &root->fs_info->delayed_block_rsv;
712 trace_btrfs_space_reservation(root->fs_info, "delayed_inode",
713 node->inode_id, node->bytes_reserved, 0);
714 btrfs_block_rsv_release(root, rsv,
715 node->bytes_reserved);
716 node->bytes_reserved = 0;
717 }
718
719 /*
720 * This helper will insert some continuous items into the same leaf according
721 * to the free space of the leaf.
722 */
723 static int btrfs_batch_insert_items(struct btrfs_root *root,
724 struct btrfs_path *path,
725 struct btrfs_delayed_item *item)
726 {
727 struct btrfs_delayed_item *curr, *next;
728 int free_space;
729 int total_data_size = 0, total_size = 0;
730 struct extent_buffer *leaf;
731 char *data_ptr;
732 struct btrfs_key *keys;
733 u32 *data_size;
734 struct list_head head;
735 int slot;
736 int nitems;
737 int i;
738 int ret = 0;
739
740 BUG_ON(!path->nodes[0]);
741
742 leaf = path->nodes[0];
743 free_space = btrfs_leaf_free_space(root, leaf);
744 INIT_LIST_HEAD(&head);
745
746 next = item;
747 nitems = 0;
748
749 /*
750 * count the number of the continuous items that we can insert in batch
751 */
752 while (total_size + next->data_len + sizeof(struct btrfs_item) <=
753 free_space) {
754 total_data_size += next->data_len;
755 total_size += next->data_len + sizeof(struct btrfs_item);
756 list_add_tail(&next->tree_list, &head);
757 nitems++;
758
759 curr = next;
760 next = __btrfs_next_delayed_item(curr);
761 if (!next)
762 break;
763
764 if (!btrfs_is_continuous_delayed_item(curr, next))
765 break;
766 }
767
768 if (!nitems) {
769 ret = 0;
770 goto out;
771 }
772
773 /*
774 * we need allocate some memory space, but it might cause the task
775 * to sleep, so we set all locked nodes in the path to blocking locks
776 * first.
777 */
778 btrfs_set_path_blocking(path);
779
780 keys = kmalloc_array(nitems, sizeof(struct btrfs_key), GFP_NOFS);
781 if (!keys) {
782 ret = -ENOMEM;
783 goto out;
784 }
785
786 data_size = kmalloc_array(nitems, sizeof(u32), GFP_NOFS);
787 if (!data_size) {
788 ret = -ENOMEM;
789 goto error;
790 }
791
792 /* get keys of all the delayed items */
793 i = 0;
794 list_for_each_entry(next, &head, tree_list) {
795 keys[i] = next->key;
796 data_size[i] = next->data_len;
797 i++;
798 }
799
800 /* reset all the locked nodes in the patch to spinning locks. */
801 btrfs_clear_path_blocking(path, NULL, 0);
802
803 /* insert the keys of the items */
804 setup_items_for_insert(root, path, keys, data_size,
805 total_data_size, total_size, nitems);
806
807 /* insert the dir index items */
808 slot = path->slots[0];
809 list_for_each_entry_safe(curr, next, &head, tree_list) {
810 data_ptr = btrfs_item_ptr(leaf, slot, char);
811 write_extent_buffer(leaf, &curr->data,
812 (unsigned long)data_ptr,
813 curr->data_len);
814 slot++;
815
816 btrfs_delayed_item_release_metadata(root, curr);
817
818 list_del(&curr->tree_list);
819 btrfs_release_delayed_item(curr);
820 }
821
822 error:
823 kfree(data_size);
824 kfree(keys);
825 out:
826 return ret;
827 }
828
829 /*
830 * This helper can just do simple insertion that needn't extend item for new
831 * data, such as directory name index insertion, inode insertion.
832 */
833 static int btrfs_insert_delayed_item(struct btrfs_trans_handle *trans,
834 struct btrfs_root *root,
835 struct btrfs_path *path,
836 struct btrfs_delayed_item *delayed_item)
837 {
838 struct extent_buffer *leaf;
839 char *ptr;
840 int ret;
841
842 ret = btrfs_insert_empty_item(trans, root, path, &delayed_item->key,
843 delayed_item->data_len);
844 if (ret < 0 && ret != -EEXIST)
845 return ret;
846
847 leaf = path->nodes[0];
848
849 ptr = btrfs_item_ptr(leaf, path->slots[0], char);
850
851 write_extent_buffer(leaf, delayed_item->data, (unsigned long)ptr,
852 delayed_item->data_len);
853 btrfs_mark_buffer_dirty(leaf);
854
855 btrfs_delayed_item_release_metadata(root, delayed_item);
856 return 0;
857 }
858
859 /*
860 * we insert an item first, then if there are some continuous items, we try
861 * to insert those items into the same leaf.
862 */
863 static int btrfs_insert_delayed_items(struct btrfs_trans_handle *trans,
864 struct btrfs_path *path,
865 struct btrfs_root *root,
866 struct btrfs_delayed_node *node)
867 {
868 struct btrfs_delayed_item *curr, *prev;
869 int ret = 0;
870
871 do_again:
872 mutex_lock(&node->mutex);
873 curr = __btrfs_first_delayed_insertion_item(node);
874 if (!curr)
875 goto insert_end;
876
877 ret = btrfs_insert_delayed_item(trans, root, path, curr);
878 if (ret < 0) {
879 btrfs_release_path(path);
880 goto insert_end;
881 }
882
883 prev = curr;
884 curr = __btrfs_next_delayed_item(prev);
885 if (curr && btrfs_is_continuous_delayed_item(prev, curr)) {
886 /* insert the continuous items into the same leaf */
887 path->slots[0]++;
888 btrfs_batch_insert_items(root, path, curr);
889 }
890 btrfs_release_delayed_item(prev);
891 btrfs_mark_buffer_dirty(path->nodes[0]);
892
893 btrfs_release_path(path);
894 mutex_unlock(&node->mutex);
895 goto do_again;
896
897 insert_end:
898 mutex_unlock(&node->mutex);
899 return ret;
900 }
901
902 static int btrfs_batch_delete_items(struct btrfs_trans_handle *trans,
903 struct btrfs_root *root,
904 struct btrfs_path *path,
905 struct btrfs_delayed_item *item)
906 {
907 struct btrfs_delayed_item *curr, *next;
908 struct extent_buffer *leaf;
909 struct btrfs_key key;
910 struct list_head head;
911 int nitems, i, last_item;
912 int ret = 0;
913
914 BUG_ON(!path->nodes[0]);
915
916 leaf = path->nodes[0];
917
918 i = path->slots[0];
919 last_item = btrfs_header_nritems(leaf) - 1;
920 if (i > last_item)
921 return -ENOENT; /* FIXME: Is errno suitable? */
922
923 next = item;
924 INIT_LIST_HEAD(&head);
925 btrfs_item_key_to_cpu(leaf, &key, i);
926 nitems = 0;
927 /*
928 * count the number of the dir index items that we can delete in batch
929 */
930 while (btrfs_comp_cpu_keys(&next->key, &key) == 0) {
931 list_add_tail(&next->tree_list, &head);
932 nitems++;
933
934 curr = next;
935 next = __btrfs_next_delayed_item(curr);
936 if (!next)
937 break;
938
939 if (!btrfs_is_continuous_delayed_item(curr, next))
940 break;
941
942 i++;
943 if (i > last_item)
944 break;
945 btrfs_item_key_to_cpu(leaf, &key, i);
946 }
947
948 if (!nitems)
949 return 0;
950
951 ret = btrfs_del_items(trans, root, path, path->slots[0], nitems);
952 if (ret)
953 goto out;
954
955 list_for_each_entry_safe(curr, next, &head, tree_list) {
956 btrfs_delayed_item_release_metadata(root, curr);
957 list_del(&curr->tree_list);
958 btrfs_release_delayed_item(curr);
959 }
960
961 out:
962 return ret;
963 }
964
965 static int btrfs_delete_delayed_items(struct btrfs_trans_handle *trans,
966 struct btrfs_path *path,
967 struct btrfs_root *root,
968 struct btrfs_delayed_node *node)
969 {
970 struct btrfs_delayed_item *curr, *prev;
971 int ret = 0;
972
973 do_again:
974 mutex_lock(&node->mutex);
975 curr = __btrfs_first_delayed_deletion_item(node);
976 if (!curr)
977 goto delete_fail;
978
979 ret = btrfs_search_slot(trans, root, &curr->key, path, -1, 1);
980 if (ret < 0)
981 goto delete_fail;
982 else if (ret > 0) {
983 /*
984 * can't find the item which the node points to, so this node
985 * is invalid, just drop it.
986 */
987 prev = curr;
988 curr = __btrfs_next_delayed_item(prev);
989 btrfs_release_delayed_item(prev);
990 ret = 0;
991 btrfs_release_path(path);
992 if (curr) {
993 mutex_unlock(&node->mutex);
994 goto do_again;
995 } else
996 goto delete_fail;
997 }
998
999 btrfs_batch_delete_items(trans, root, path, curr);
1000 btrfs_release_path(path);
1001 mutex_unlock(&node->mutex);
1002 goto do_again;
1003
1004 delete_fail:
1005 btrfs_release_path(path);
1006 mutex_unlock(&node->mutex);
1007 return ret;
1008 }
1009
1010 static void btrfs_release_delayed_inode(struct btrfs_delayed_node *delayed_node)
1011 {
1012 struct btrfs_delayed_root *delayed_root;
1013
1014 if (delayed_node &&
1015 test_bit(BTRFS_DELAYED_NODE_INODE_DIRTY, &delayed_node->flags)) {
1016 BUG_ON(!delayed_node->root);
1017 clear_bit(BTRFS_DELAYED_NODE_INODE_DIRTY, &delayed_node->flags);
1018 delayed_node->count--;
1019
1020 delayed_root = delayed_node->root->fs_info->delayed_root;
1021 finish_one_item(delayed_root);
1022 }
1023 }
1024
1025 static void btrfs_release_delayed_iref(struct btrfs_delayed_node *delayed_node)
1026 {
1027 struct btrfs_delayed_root *delayed_root;
1028
1029 ASSERT(delayed_node->root);
1030 clear_bit(BTRFS_DELAYED_NODE_DEL_IREF, &delayed_node->flags);
1031 delayed_node->count--;
1032
1033 delayed_root = delayed_node->root->fs_info->delayed_root;
1034 finish_one_item(delayed_root);
1035 }
1036
1037 static int __btrfs_update_delayed_inode(struct btrfs_trans_handle *trans,
1038 struct btrfs_root *root,
1039 struct btrfs_path *path,
1040 struct btrfs_delayed_node *node)
1041 {
1042 struct btrfs_key key;
1043 struct btrfs_inode_item *inode_item;
1044 struct extent_buffer *leaf;
1045 int mod;
1046 int ret;
1047
1048 key.objectid = node->inode_id;
1049 key.type = BTRFS_INODE_ITEM_KEY;
1050 key.offset = 0;
1051
1052 if (test_bit(BTRFS_DELAYED_NODE_DEL_IREF, &node->flags))
1053 mod = -1;
1054 else
1055 mod = 1;
1056
1057 ret = btrfs_lookup_inode(trans, root, path, &key, mod);
1058 if (ret > 0) {
1059 btrfs_release_path(path);
1060 return -ENOENT;
1061 } else if (ret < 0) {
1062 return ret;
1063 }
1064
1065 leaf = path->nodes[0];
1066 inode_item = btrfs_item_ptr(leaf, path->slots[0],
1067 struct btrfs_inode_item);
1068 write_extent_buffer(leaf, &node->inode_item, (unsigned long)inode_item,
1069 sizeof(struct btrfs_inode_item));
1070 btrfs_mark_buffer_dirty(leaf);
1071
1072 if (!test_bit(BTRFS_DELAYED_NODE_DEL_IREF, &node->flags))
1073 goto no_iref;
1074
1075 path->slots[0]++;
1076 if (path->slots[0] >= btrfs_header_nritems(leaf))
1077 goto search;
1078 again:
1079 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
1080 if (key.objectid != node->inode_id)
1081 goto out;
1082
1083 if (key.type != BTRFS_INODE_REF_KEY &&
1084 key.type != BTRFS_INODE_EXTREF_KEY)
1085 goto out;
1086
1087 /*
1088 * Delayed iref deletion is for the inode who has only one link,
1089 * so there is only one iref. The case that several irefs are
1090 * in the same item doesn't exist.
1091 */
1092 btrfs_del_item(trans, root, path);
1093 out:
1094 btrfs_release_delayed_iref(node);
1095 no_iref:
1096 btrfs_release_path(path);
1097 err_out:
1098 btrfs_delayed_inode_release_metadata(root, node);
1099 btrfs_release_delayed_inode(node);
1100
1101 return ret;
1102
1103 search:
1104 btrfs_release_path(path);
1105
1106 key.type = BTRFS_INODE_EXTREF_KEY;
1107 key.offset = -1;
1108 ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
1109 if (ret < 0)
1110 goto err_out;
1111 ASSERT(ret);
1112
1113 ret = 0;
1114 leaf = path->nodes[0];
1115 path->slots[0]--;
1116 goto again;
1117 }
1118
1119 static inline int btrfs_update_delayed_inode(struct btrfs_trans_handle *trans,
1120 struct btrfs_root *root,
1121 struct btrfs_path *path,
1122 struct btrfs_delayed_node *node)
1123 {
1124 int ret;
1125
1126 mutex_lock(&node->mutex);
1127 if (!test_bit(BTRFS_DELAYED_NODE_INODE_DIRTY, &node->flags)) {
1128 mutex_unlock(&node->mutex);
1129 return 0;
1130 }
1131
1132 ret = __btrfs_update_delayed_inode(trans, root, path, node);
1133 mutex_unlock(&node->mutex);
1134 return ret;
1135 }
1136
1137 static inline int
1138 __btrfs_commit_inode_delayed_items(struct btrfs_trans_handle *trans,
1139 struct btrfs_path *path,
1140 struct btrfs_delayed_node *node)
1141 {
1142 int ret;
1143
1144 ret = btrfs_insert_delayed_items(trans, path, node->root, node);
1145 if (ret)
1146 return ret;
1147
1148 ret = btrfs_delete_delayed_items(trans, path, node->root, node);
1149 if (ret)
1150 return ret;
1151
1152 ret = btrfs_update_delayed_inode(trans, node->root, path, node);
1153 return ret;
1154 }
1155
1156 /*
1157 * Called when committing the transaction.
1158 * Returns 0 on success.
1159 * Returns < 0 on error and returns with an aborted transaction with any
1160 * outstanding delayed items cleaned up.
1161 */
1162 static int __btrfs_run_delayed_items(struct btrfs_trans_handle *trans,
1163 struct btrfs_root *root, int nr)
1164 {
1165 struct btrfs_delayed_root *delayed_root;
1166 struct btrfs_delayed_node *curr_node, *prev_node;
1167 struct btrfs_path *path;
1168 struct btrfs_block_rsv *block_rsv;
1169 int ret = 0;
1170 bool count = (nr > 0);
1171
1172 if (trans->aborted)
1173 return -EIO;
1174
1175 path = btrfs_alloc_path();
1176 if (!path)
1177 return -ENOMEM;
1178 path->leave_spinning = 1;
1179
1180 block_rsv = trans->block_rsv;
1181 trans->block_rsv = &root->fs_info->delayed_block_rsv;
1182
1183 delayed_root = btrfs_get_delayed_root(root);
1184
1185 curr_node = btrfs_first_delayed_node(delayed_root);
1186 while (curr_node && (!count || (count && nr--))) {
1187 ret = __btrfs_commit_inode_delayed_items(trans, path,
1188 curr_node);
1189 if (ret) {
1190 btrfs_release_delayed_node(curr_node);
1191 curr_node = NULL;
1192 btrfs_abort_transaction(trans, root, ret);
1193 break;
1194 }
1195
1196 prev_node = curr_node;
1197 curr_node = btrfs_next_delayed_node(curr_node);
1198 btrfs_release_delayed_node(prev_node);
1199 }
1200
1201 if (curr_node)
1202 btrfs_release_delayed_node(curr_node);
1203 btrfs_free_path(path);
1204 trans->block_rsv = block_rsv;
1205
1206 return ret;
1207 }
1208
1209 int btrfs_run_delayed_items(struct btrfs_trans_handle *trans,
1210 struct btrfs_root *root)
1211 {
1212 return __btrfs_run_delayed_items(trans, root, -1);
1213 }
1214
1215 int btrfs_run_delayed_items_nr(struct btrfs_trans_handle *trans,
1216 struct btrfs_root *root, int nr)
1217 {
1218 return __btrfs_run_delayed_items(trans, root, nr);
1219 }
1220
1221 int btrfs_commit_inode_delayed_items(struct btrfs_trans_handle *trans,
1222 struct inode *inode)
1223 {
1224 struct btrfs_delayed_node *delayed_node = btrfs_get_delayed_node(inode);
1225 struct btrfs_path *path;
1226 struct btrfs_block_rsv *block_rsv;
1227 int ret;
1228
1229 if (!delayed_node)
1230 return 0;
1231
1232 mutex_lock(&delayed_node->mutex);
1233 if (!delayed_node->count) {
1234 mutex_unlock(&delayed_node->mutex);
1235 btrfs_release_delayed_node(delayed_node);
1236 return 0;
1237 }
1238 mutex_unlock(&delayed_node->mutex);
1239
1240 path = btrfs_alloc_path();
1241 if (!path) {
1242 btrfs_release_delayed_node(delayed_node);
1243 return -ENOMEM;
1244 }
1245 path->leave_spinning = 1;
1246
1247 block_rsv = trans->block_rsv;
1248 trans->block_rsv = &delayed_node->root->fs_info->delayed_block_rsv;
1249
1250 ret = __btrfs_commit_inode_delayed_items(trans, path, delayed_node);
1251
1252 btrfs_release_delayed_node(delayed_node);
1253 btrfs_free_path(path);
1254 trans->block_rsv = block_rsv;
1255
1256 return ret;
1257 }
1258
1259 int btrfs_commit_inode_delayed_inode(struct inode *inode)
1260 {
1261 struct btrfs_trans_handle *trans;
1262 struct btrfs_delayed_node *delayed_node = btrfs_get_delayed_node(inode);
1263 struct btrfs_path *path;
1264 struct btrfs_block_rsv *block_rsv;
1265 int ret;
1266
1267 if (!delayed_node)
1268 return 0;
1269
1270 mutex_lock(&delayed_node->mutex);
1271 if (!test_bit(BTRFS_DELAYED_NODE_INODE_DIRTY, &delayed_node->flags)) {
1272 mutex_unlock(&delayed_node->mutex);
1273 btrfs_release_delayed_node(delayed_node);
1274 return 0;
1275 }
1276 mutex_unlock(&delayed_node->mutex);
1277
1278 trans = btrfs_join_transaction(delayed_node->root);
1279 if (IS_ERR(trans)) {
1280 ret = PTR_ERR(trans);
1281 goto out;
1282 }
1283
1284 path = btrfs_alloc_path();
1285 if (!path) {
1286 ret = -ENOMEM;
1287 goto trans_out;
1288 }
1289 path->leave_spinning = 1;
1290
1291 block_rsv = trans->block_rsv;
1292 trans->block_rsv = &delayed_node->root->fs_info->delayed_block_rsv;
1293
1294 mutex_lock(&delayed_node->mutex);
1295 if (test_bit(BTRFS_DELAYED_NODE_INODE_DIRTY, &delayed_node->flags))
1296 ret = __btrfs_update_delayed_inode(trans, delayed_node->root,
1297 path, delayed_node);
1298 else
1299 ret = 0;
1300 mutex_unlock(&delayed_node->mutex);
1301
1302 btrfs_free_path(path);
1303 trans->block_rsv = block_rsv;
1304 trans_out:
1305 btrfs_end_transaction(trans, delayed_node->root);
1306 btrfs_btree_balance_dirty(delayed_node->root);
1307 out:
1308 btrfs_release_delayed_node(delayed_node);
1309
1310 return ret;
1311 }
1312
1313 void btrfs_remove_delayed_node(struct inode *inode)
1314 {
1315 struct btrfs_delayed_node *delayed_node;
1316
1317 delayed_node = ACCESS_ONCE(BTRFS_I(inode)->delayed_node);
1318 if (!delayed_node)
1319 return;
1320
1321 BTRFS_I(inode)->delayed_node = NULL;
1322 btrfs_release_delayed_node(delayed_node);
1323 }
1324
1325 struct btrfs_async_delayed_work {
1326 struct btrfs_delayed_root *delayed_root;
1327 int nr;
1328 struct btrfs_work work;
1329 };
1330
1331 static void btrfs_async_run_delayed_root(struct btrfs_work *work)
1332 {
1333 struct btrfs_async_delayed_work *async_work;
1334 struct btrfs_delayed_root *delayed_root;
1335 struct btrfs_trans_handle *trans;
1336 struct btrfs_path *path;
1337 struct btrfs_delayed_node *delayed_node = NULL;
1338 struct btrfs_root *root;
1339 struct btrfs_block_rsv *block_rsv;
1340 int total_done = 0;
1341
1342 async_work = container_of(work, struct btrfs_async_delayed_work, work);
1343 delayed_root = async_work->delayed_root;
1344
1345 path = btrfs_alloc_path();
1346 if (!path)
1347 goto out;
1348
1349 again:
1350 if (atomic_read(&delayed_root->items) < BTRFS_DELAYED_BACKGROUND / 2)
1351 goto free_path;
1352
1353 delayed_node = btrfs_first_prepared_delayed_node(delayed_root);
1354 if (!delayed_node)
1355 goto free_path;
1356
1357 path->leave_spinning = 1;
1358 root = delayed_node->root;
1359
1360 trans = btrfs_join_transaction(root);
1361 if (IS_ERR(trans))
1362 goto release_path;
1363
1364 block_rsv = trans->block_rsv;
1365 trans->block_rsv = &root->fs_info->delayed_block_rsv;
1366
1367 __btrfs_commit_inode_delayed_items(trans, path, delayed_node);
1368
1369 trans->block_rsv = block_rsv;
1370 btrfs_end_transaction(trans, root);
1371 btrfs_btree_balance_dirty_nodelay(root);
1372
1373 release_path:
1374 btrfs_release_path(path);
1375 total_done++;
1376
1377 btrfs_release_prepared_delayed_node(delayed_node);
1378 if ((async_work->nr == 0 && total_done < BTRFS_DELAYED_WRITEBACK) ||
1379 total_done < async_work->nr)
1380 goto again;
1381
1382 free_path:
1383 btrfs_free_path(path);
1384 out:
1385 wake_up(&delayed_root->wait);
1386 kfree(async_work);
1387 }
1388
1389
1390 static int btrfs_wq_run_delayed_node(struct btrfs_delayed_root *delayed_root,
1391 struct btrfs_fs_info *fs_info, int nr)
1392 {
1393 struct btrfs_async_delayed_work *async_work;
1394
1395 if (atomic_read(&delayed_root->items) < BTRFS_DELAYED_BACKGROUND ||
1396 btrfs_workqueue_normal_congested(fs_info->delayed_workers))
1397 return 0;
1398
1399 async_work = kmalloc(sizeof(*async_work), GFP_NOFS);
1400 if (!async_work)
1401 return -ENOMEM;
1402
1403 async_work->delayed_root = delayed_root;
1404 btrfs_init_work(&async_work->work, btrfs_delayed_meta_helper,
1405 btrfs_async_run_delayed_root, NULL, NULL);
1406 async_work->nr = nr;
1407
1408 btrfs_queue_work(fs_info->delayed_workers, &async_work->work);
1409 return 0;
1410 }
1411
1412 void btrfs_assert_delayed_root_empty(struct btrfs_root *root)
1413 {
1414 struct btrfs_delayed_root *delayed_root;
1415 delayed_root = btrfs_get_delayed_root(root);
1416 WARN_ON(btrfs_first_delayed_node(delayed_root));
1417 }
1418
1419 static int could_end_wait(struct btrfs_delayed_root *delayed_root, int seq)
1420 {
1421 int val = atomic_read(&delayed_root->items_seq);
1422
1423 if (val < seq || val >= seq + BTRFS_DELAYED_BATCH)
1424 return 1;
1425
1426 if (atomic_read(&delayed_root->items) < BTRFS_DELAYED_BACKGROUND)
1427 return 1;
1428
1429 return 0;
1430 }
1431
1432 void btrfs_balance_delayed_items(struct btrfs_root *root)
1433 {
1434 struct btrfs_delayed_root *delayed_root;
1435 struct btrfs_fs_info *fs_info = root->fs_info;
1436
1437 delayed_root = btrfs_get_delayed_root(root);
1438
1439 if (atomic_read(&delayed_root->items) < BTRFS_DELAYED_BACKGROUND)
1440 return;
1441
1442 if (atomic_read(&delayed_root->items) >= BTRFS_DELAYED_WRITEBACK) {
1443 int seq;
1444 int ret;
1445
1446 seq = atomic_read(&delayed_root->items_seq);
1447
1448 ret = btrfs_wq_run_delayed_node(delayed_root, fs_info, 0);
1449 if (ret)
1450 return;
1451
1452 wait_event_interruptible(delayed_root->wait,
1453 could_end_wait(delayed_root, seq));
1454 return;
1455 }
1456
1457 btrfs_wq_run_delayed_node(delayed_root, fs_info, BTRFS_DELAYED_BATCH);
1458 }
1459
1460 /* Will return 0 or -ENOMEM */
1461 int btrfs_insert_delayed_dir_index(struct btrfs_trans_handle *trans,
1462 struct btrfs_root *root, const char *name,
1463 int name_len, struct inode *dir,
1464 struct btrfs_disk_key *disk_key, u8 type,
1465 u64 index)
1466 {
1467 struct btrfs_delayed_node *delayed_node;
1468 struct btrfs_delayed_item *delayed_item;
1469 struct btrfs_dir_item *dir_item;
1470 int ret;
1471
1472 delayed_node = btrfs_get_or_create_delayed_node(dir);
1473 if (IS_ERR(delayed_node))
1474 return PTR_ERR(delayed_node);
1475
1476 delayed_item = btrfs_alloc_delayed_item(sizeof(*dir_item) + name_len);
1477 if (!delayed_item) {
1478 ret = -ENOMEM;
1479 goto release_node;
1480 }
1481
1482 delayed_item->key.objectid = btrfs_ino(dir);
1483 delayed_item->key.type = BTRFS_DIR_INDEX_KEY;
1484 delayed_item->key.offset = index;
1485
1486 dir_item = (struct btrfs_dir_item *)delayed_item->data;
1487 dir_item->location = *disk_key;
1488 btrfs_set_stack_dir_transid(dir_item, trans->transid);
1489 btrfs_set_stack_dir_data_len(dir_item, 0);
1490 btrfs_set_stack_dir_name_len(dir_item, name_len);
1491 btrfs_set_stack_dir_type(dir_item, type);
1492 memcpy((char *)(dir_item + 1), name, name_len);
1493
1494 ret = btrfs_delayed_item_reserve_metadata(trans, root, delayed_item);
1495 /*
1496 * we have reserved enough space when we start a new transaction,
1497 * so reserving metadata failure is impossible
1498 */
1499 BUG_ON(ret);
1500
1501
1502 mutex_lock(&delayed_node->mutex);
1503 ret = __btrfs_add_delayed_insertion_item(delayed_node, delayed_item);
1504 if (unlikely(ret)) {
1505 btrfs_err(root->fs_info, "err add delayed dir index item(name: %.*s) "
1506 "into the insertion tree of the delayed node"
1507 "(root id: %llu, inode id: %llu, errno: %d)",
1508 name_len, name, delayed_node->root->objectid,
1509 delayed_node->inode_id, ret);
1510 BUG();
1511 }
1512 mutex_unlock(&delayed_node->mutex);
1513
1514 release_node:
1515 btrfs_release_delayed_node(delayed_node);
1516 return ret;
1517 }
1518
1519 static int btrfs_delete_delayed_insertion_item(struct btrfs_root *root,
1520 struct btrfs_delayed_node *node,
1521 struct btrfs_key *key)
1522 {
1523 struct btrfs_delayed_item *item;
1524
1525 mutex_lock(&node->mutex);
1526 item = __btrfs_lookup_delayed_insertion_item(node, key);
1527 if (!item) {
1528 mutex_unlock(&node->mutex);
1529 return 1;
1530 }
1531
1532 btrfs_delayed_item_release_metadata(root, item);
1533 btrfs_release_delayed_item(item);
1534 mutex_unlock(&node->mutex);
1535 return 0;
1536 }
1537
1538 int btrfs_delete_delayed_dir_index(struct btrfs_trans_handle *trans,
1539 struct btrfs_root *root, struct inode *dir,
1540 u64 index)
1541 {
1542 struct btrfs_delayed_node *node;
1543 struct btrfs_delayed_item *item;
1544 struct btrfs_key item_key;
1545 int ret;
1546
1547 node = btrfs_get_or_create_delayed_node(dir);
1548 if (IS_ERR(node))
1549 return PTR_ERR(node);
1550
1551 item_key.objectid = btrfs_ino(dir);
1552 item_key.type = BTRFS_DIR_INDEX_KEY;
1553 item_key.offset = index;
1554
1555 ret = btrfs_delete_delayed_insertion_item(root, node, &item_key);
1556 if (!ret)
1557 goto end;
1558
1559 item = btrfs_alloc_delayed_item(0);
1560 if (!item) {
1561 ret = -ENOMEM;
1562 goto end;
1563 }
1564
1565 item->key = item_key;
1566
1567 ret = btrfs_delayed_item_reserve_metadata(trans, root, item);
1568 /*
1569 * we have reserved enough space when we start a new transaction,
1570 * so reserving metadata failure is impossible.
1571 */
1572 BUG_ON(ret);
1573
1574 mutex_lock(&node->mutex);
1575 ret = __btrfs_add_delayed_deletion_item(node, item);
1576 if (unlikely(ret)) {
1577 btrfs_err(root->fs_info, "err add delayed dir index item(index: %llu) "
1578 "into the deletion tree of the delayed node"
1579 "(root id: %llu, inode id: %llu, errno: %d)",
1580 index, node->root->objectid, node->inode_id,
1581 ret);
1582 BUG();
1583 }
1584 mutex_unlock(&node->mutex);
1585 end:
1586 btrfs_release_delayed_node(node);
1587 return ret;
1588 }
1589
1590 int btrfs_inode_delayed_dir_index_count(struct inode *inode)
1591 {
1592 struct btrfs_delayed_node *delayed_node = btrfs_get_delayed_node(inode);
1593
1594 if (!delayed_node)
1595 return -ENOENT;
1596
1597 /*
1598 * Since we have held i_mutex of this directory, it is impossible that
1599 * a new directory index is added into the delayed node and index_cnt
1600 * is updated now. So we needn't lock the delayed node.
1601 */
1602 if (!delayed_node->index_cnt) {
1603 btrfs_release_delayed_node(delayed_node);
1604 return -EINVAL;
1605 }
1606
1607 BTRFS_I(inode)->index_cnt = delayed_node->index_cnt;
1608 btrfs_release_delayed_node(delayed_node);
1609 return 0;
1610 }
1611
1612 void btrfs_get_delayed_items(struct inode *inode, struct list_head *ins_list,
1613 struct list_head *del_list)
1614 {
1615 struct btrfs_delayed_node *delayed_node;
1616 struct btrfs_delayed_item *item;
1617
1618 delayed_node = btrfs_get_delayed_node(inode);
1619 if (!delayed_node)
1620 return;
1621
1622 mutex_lock(&delayed_node->mutex);
1623 item = __btrfs_first_delayed_insertion_item(delayed_node);
1624 while (item) {
1625 atomic_inc(&item->refs);
1626 list_add_tail(&item->readdir_list, ins_list);
1627 item = __btrfs_next_delayed_item(item);
1628 }
1629
1630 item = __btrfs_first_delayed_deletion_item(delayed_node);
1631 while (item) {
1632 atomic_inc(&item->refs);
1633 list_add_tail(&item->readdir_list, del_list);
1634 item = __btrfs_next_delayed_item(item);
1635 }
1636 mutex_unlock(&delayed_node->mutex);
1637 /*
1638 * This delayed node is still cached in the btrfs inode, so refs
1639 * must be > 1 now, and we needn't check it is going to be freed
1640 * or not.
1641 *
1642 * Besides that, this function is used to read dir, we do not
1643 * insert/delete delayed items in this period. So we also needn't
1644 * requeue or dequeue this delayed node.
1645 */
1646 atomic_dec(&delayed_node->refs);
1647 }
1648
1649 void btrfs_put_delayed_items(struct list_head *ins_list,
1650 struct list_head *del_list)
1651 {
1652 struct btrfs_delayed_item *curr, *next;
1653
1654 list_for_each_entry_safe(curr, next, ins_list, readdir_list) {
1655 list_del(&curr->readdir_list);
1656 if (atomic_dec_and_test(&curr->refs))
1657 kfree(curr);
1658 }
1659
1660 list_for_each_entry_safe(curr, next, del_list, readdir_list) {
1661 list_del(&curr->readdir_list);
1662 if (atomic_dec_and_test(&curr->refs))
1663 kfree(curr);
1664 }
1665 }
1666
1667 int btrfs_should_delete_dir_index(struct list_head *del_list,
1668 u64 index)
1669 {
1670 struct btrfs_delayed_item *curr, *next;
1671 int ret;
1672
1673 if (list_empty(del_list))
1674 return 0;
1675
1676 list_for_each_entry_safe(curr, next, del_list, readdir_list) {
1677 if (curr->key.offset > index)
1678 break;
1679
1680 list_del(&curr->readdir_list);
1681 ret = (curr->key.offset == index);
1682
1683 if (atomic_dec_and_test(&curr->refs))
1684 kfree(curr);
1685
1686 if (ret)
1687 return 1;
1688 else
1689 continue;
1690 }
1691 return 0;
1692 }
1693
1694 /*
1695 * btrfs_readdir_delayed_dir_index - read dir info stored in the delayed tree
1696 *
1697 */
1698 int btrfs_readdir_delayed_dir_index(struct dir_context *ctx,
1699 struct list_head *ins_list, bool *emitted)
1700 {
1701 struct btrfs_dir_item *di;
1702 struct btrfs_delayed_item *curr, *next;
1703 struct btrfs_key location;
1704 char *name;
1705 int name_len;
1706 int over = 0;
1707 unsigned char d_type;
1708
1709 if (list_empty(ins_list))
1710 return 0;
1711
1712 /*
1713 * Changing the data of the delayed item is impossible. So
1714 * we needn't lock them. And we have held i_mutex of the
1715 * directory, nobody can delete any directory indexes now.
1716 */
1717 list_for_each_entry_safe(curr, next, ins_list, readdir_list) {
1718 list_del(&curr->readdir_list);
1719
1720 if (curr->key.offset < ctx->pos) {
1721 if (atomic_dec_and_test(&curr->refs))
1722 kfree(curr);
1723 continue;
1724 }
1725
1726 ctx->pos = curr->key.offset;
1727
1728 di = (struct btrfs_dir_item *)curr->data;
1729 name = (char *)(di + 1);
1730 name_len = btrfs_stack_dir_name_len(di);
1731
1732 d_type = btrfs_filetype_table[di->type];
1733 btrfs_disk_key_to_cpu(&location, &di->location);
1734
1735 over = !dir_emit(ctx, name, name_len,
1736 location.objectid, d_type);
1737
1738 if (atomic_dec_and_test(&curr->refs))
1739 kfree(curr);
1740
1741 if (over)
1742 return 1;
1743 *emitted = true;
1744 }
1745 return 0;
1746 }
1747
1748 static void fill_stack_inode_item(struct btrfs_trans_handle *trans,
1749 struct btrfs_inode_item *inode_item,
1750 struct inode *inode)
1751 {
1752 btrfs_set_stack_inode_uid(inode_item, i_uid_read(inode));
1753 btrfs_set_stack_inode_gid(inode_item, i_gid_read(inode));
1754 btrfs_set_stack_inode_size(inode_item, BTRFS_I(inode)->disk_i_size);
1755 btrfs_set_stack_inode_mode(inode_item, inode->i_mode);
1756 btrfs_set_stack_inode_nlink(inode_item, inode->i_nlink);
1757 btrfs_set_stack_inode_nbytes(inode_item, inode_get_bytes(inode));
1758 btrfs_set_stack_inode_generation(inode_item,
1759 BTRFS_I(inode)->generation);
1760 btrfs_set_stack_inode_sequence(inode_item, inode->i_version);
1761 btrfs_set_stack_inode_transid(inode_item, trans->transid);
1762 btrfs_set_stack_inode_rdev(inode_item, inode->i_rdev);
1763 btrfs_set_stack_inode_flags(inode_item, BTRFS_I(inode)->flags);
1764 btrfs_set_stack_inode_block_group(inode_item, 0);
1765
1766 btrfs_set_stack_timespec_sec(&inode_item->atime,
1767 inode->i_atime.tv_sec);
1768 btrfs_set_stack_timespec_nsec(&inode_item->atime,
1769 inode->i_atime.tv_nsec);
1770
1771 btrfs_set_stack_timespec_sec(&inode_item->mtime,
1772 inode->i_mtime.tv_sec);
1773 btrfs_set_stack_timespec_nsec(&inode_item->mtime,
1774 inode->i_mtime.tv_nsec);
1775
1776 btrfs_set_stack_timespec_sec(&inode_item->ctime,
1777 inode->i_ctime.tv_sec);
1778 btrfs_set_stack_timespec_nsec(&inode_item->ctime,
1779 inode->i_ctime.tv_nsec);
1780
1781 btrfs_set_stack_timespec_sec(&inode_item->otime,
1782 BTRFS_I(inode)->i_otime.tv_sec);
1783 btrfs_set_stack_timespec_nsec(&inode_item->otime,
1784 BTRFS_I(inode)->i_otime.tv_nsec);
1785 }
1786
1787 int btrfs_fill_inode(struct inode *inode, u32 *rdev)
1788 {
1789 struct btrfs_delayed_node *delayed_node;
1790 struct btrfs_inode_item *inode_item;
1791
1792 delayed_node = btrfs_get_delayed_node(inode);
1793 if (!delayed_node)
1794 return -ENOENT;
1795
1796 mutex_lock(&delayed_node->mutex);
1797 if (!test_bit(BTRFS_DELAYED_NODE_INODE_DIRTY, &delayed_node->flags)) {
1798 mutex_unlock(&delayed_node->mutex);
1799 btrfs_release_delayed_node(delayed_node);
1800 return -ENOENT;
1801 }
1802
1803 inode_item = &delayed_node->inode_item;
1804
1805 i_uid_write(inode, btrfs_stack_inode_uid(inode_item));
1806 i_gid_write(inode, btrfs_stack_inode_gid(inode_item));
1807 btrfs_i_size_write(inode, btrfs_stack_inode_size(inode_item));
1808 inode->i_mode = btrfs_stack_inode_mode(inode_item);
1809 set_nlink(inode, btrfs_stack_inode_nlink(inode_item));
1810 inode_set_bytes(inode, btrfs_stack_inode_nbytes(inode_item));
1811 BTRFS_I(inode)->generation = btrfs_stack_inode_generation(inode_item);
1812 BTRFS_I(inode)->last_trans = btrfs_stack_inode_transid(inode_item);
1813
1814 inode->i_version = btrfs_stack_inode_sequence(inode_item);
1815 inode->i_rdev = 0;
1816 *rdev = btrfs_stack_inode_rdev(inode_item);
1817 BTRFS_I(inode)->flags = btrfs_stack_inode_flags(inode_item);
1818
1819 inode->i_atime.tv_sec = btrfs_stack_timespec_sec(&inode_item->atime);
1820 inode->i_atime.tv_nsec = btrfs_stack_timespec_nsec(&inode_item->atime);
1821
1822 inode->i_mtime.tv_sec = btrfs_stack_timespec_sec(&inode_item->mtime);
1823 inode->i_mtime.tv_nsec = btrfs_stack_timespec_nsec(&inode_item->mtime);
1824
1825 inode->i_ctime.tv_sec = btrfs_stack_timespec_sec(&inode_item->ctime);
1826 inode->i_ctime.tv_nsec = btrfs_stack_timespec_nsec(&inode_item->ctime);
1827
1828 BTRFS_I(inode)->i_otime.tv_sec =
1829 btrfs_stack_timespec_sec(&inode_item->otime);
1830 BTRFS_I(inode)->i_otime.tv_nsec =
1831 btrfs_stack_timespec_nsec(&inode_item->otime);
1832
1833 inode->i_generation = BTRFS_I(inode)->generation;
1834 BTRFS_I(inode)->index_cnt = (u64)-1;
1835
1836 mutex_unlock(&delayed_node->mutex);
1837 btrfs_release_delayed_node(delayed_node);
1838 return 0;
1839 }
1840
1841 int btrfs_delayed_update_inode(struct btrfs_trans_handle *trans,
1842 struct btrfs_root *root, struct inode *inode)
1843 {
1844 struct btrfs_delayed_node *delayed_node;
1845 int ret = 0;
1846
1847 delayed_node = btrfs_get_or_create_delayed_node(inode);
1848 if (IS_ERR(delayed_node))
1849 return PTR_ERR(delayed_node);
1850
1851 mutex_lock(&delayed_node->mutex);
1852 if (test_bit(BTRFS_DELAYED_NODE_INODE_DIRTY, &delayed_node->flags)) {
1853 fill_stack_inode_item(trans, &delayed_node->inode_item, inode);
1854 goto release_node;
1855 }
1856
1857 ret = btrfs_delayed_inode_reserve_metadata(trans, root, inode,
1858 delayed_node);
1859 if (ret)
1860 goto release_node;
1861
1862 fill_stack_inode_item(trans, &delayed_node->inode_item, inode);
1863 set_bit(BTRFS_DELAYED_NODE_INODE_DIRTY, &delayed_node->flags);
1864 delayed_node->count++;
1865 atomic_inc(&root->fs_info->delayed_root->items);
1866 release_node:
1867 mutex_unlock(&delayed_node->mutex);
1868 btrfs_release_delayed_node(delayed_node);
1869 return ret;
1870 }
1871
1872 int btrfs_delayed_delete_inode_ref(struct inode *inode)
1873 {
1874 struct btrfs_delayed_node *delayed_node;
1875
1876 /*
1877 * we don't do delayed inode updates during log recovery because it
1878 * leads to enospc problems. This means we also can't do
1879 * delayed inode refs
1880 */
1881 if (BTRFS_I(inode)->root->fs_info->log_root_recovering)
1882 return -EAGAIN;
1883
1884 delayed_node = btrfs_get_or_create_delayed_node(inode);
1885 if (IS_ERR(delayed_node))
1886 return PTR_ERR(delayed_node);
1887
1888 /*
1889 * We don't reserve space for inode ref deletion is because:
1890 * - We ONLY do async inode ref deletion for the inode who has only
1891 * one link(i_nlink == 1), it means there is only one inode ref.
1892 * And in most case, the inode ref and the inode item are in the
1893 * same leaf, and we will deal with them at the same time.
1894 * Since we are sure we will reserve the space for the inode item,
1895 * it is unnecessary to reserve space for inode ref deletion.
1896 * - If the inode ref and the inode item are not in the same leaf,
1897 * We also needn't worry about enospc problem, because we reserve
1898 * much more space for the inode update than it needs.
1899 * - At the worst, we can steal some space from the global reservation.
1900 * It is very rare.
1901 */
1902 mutex_lock(&delayed_node->mutex);
1903 if (test_bit(BTRFS_DELAYED_NODE_DEL_IREF, &delayed_node->flags))
1904 goto release_node;
1905
1906 set_bit(BTRFS_DELAYED_NODE_DEL_IREF, &delayed_node->flags);
1907 delayed_node->count++;
1908 atomic_inc(&BTRFS_I(inode)->root->fs_info->delayed_root->items);
1909 release_node:
1910 mutex_unlock(&delayed_node->mutex);
1911 btrfs_release_delayed_node(delayed_node);
1912 return 0;
1913 }
1914
1915 static void __btrfs_kill_delayed_node(struct btrfs_delayed_node *delayed_node)
1916 {
1917 struct btrfs_root *root = delayed_node->root;
1918 struct btrfs_delayed_item *curr_item, *prev_item;
1919
1920 mutex_lock(&delayed_node->mutex);
1921 curr_item = __btrfs_first_delayed_insertion_item(delayed_node);
1922 while (curr_item) {
1923 btrfs_delayed_item_release_metadata(root, curr_item);
1924 prev_item = curr_item;
1925 curr_item = __btrfs_next_delayed_item(prev_item);
1926 btrfs_release_delayed_item(prev_item);
1927 }
1928
1929 curr_item = __btrfs_first_delayed_deletion_item(delayed_node);
1930 while (curr_item) {
1931 btrfs_delayed_item_release_metadata(root, curr_item);
1932 prev_item = curr_item;
1933 curr_item = __btrfs_next_delayed_item(prev_item);
1934 btrfs_release_delayed_item(prev_item);
1935 }
1936
1937 if (test_bit(BTRFS_DELAYED_NODE_DEL_IREF, &delayed_node->flags))
1938 btrfs_release_delayed_iref(delayed_node);
1939
1940 if (test_bit(BTRFS_DELAYED_NODE_INODE_DIRTY, &delayed_node->flags)) {
1941 btrfs_delayed_inode_release_metadata(root, delayed_node);
1942 btrfs_release_delayed_inode(delayed_node);
1943 }
1944 mutex_unlock(&delayed_node->mutex);
1945 }
1946
1947 void btrfs_kill_delayed_inode_items(struct inode *inode)
1948 {
1949 struct btrfs_delayed_node *delayed_node;
1950
1951 delayed_node = btrfs_get_delayed_node(inode);
1952 if (!delayed_node)
1953 return;
1954
1955 __btrfs_kill_delayed_node(delayed_node);
1956 btrfs_release_delayed_node(delayed_node);
1957 }
1958
1959 void btrfs_kill_all_delayed_nodes(struct btrfs_root *root)
1960 {
1961 u64 inode_id = 0;
1962 struct btrfs_delayed_node *delayed_nodes[8];
1963 int i, n;
1964
1965 while (1) {
1966 spin_lock(&root->inode_lock);
1967 n = radix_tree_gang_lookup(&root->delayed_nodes_tree,
1968 (void **)delayed_nodes, inode_id,
1969 ARRAY_SIZE(delayed_nodes));
1970 if (!n) {
1971 spin_unlock(&root->inode_lock);
1972 break;
1973 }
1974
1975 inode_id = delayed_nodes[n - 1]->inode_id + 1;
1976
1977 for (i = 0; i < n; i++)
1978 atomic_inc(&delayed_nodes[i]->refs);
1979 spin_unlock(&root->inode_lock);
1980
1981 for (i = 0; i < n; i++) {
1982 __btrfs_kill_delayed_node(delayed_nodes[i]);
1983 btrfs_release_delayed_node(delayed_nodes[i]);
1984 }
1985 }
1986 }
1987
1988 void btrfs_destroy_delayed_inodes(struct btrfs_root *root)
1989 {
1990 struct btrfs_delayed_root *delayed_root;
1991 struct btrfs_delayed_node *curr_node, *prev_node;
1992
1993 delayed_root = btrfs_get_delayed_root(root);
1994
1995 curr_node = btrfs_first_delayed_node(delayed_root);
1996 while (curr_node) {
1997 __btrfs_kill_delayed_node(curr_node);
1998
1999 prev_node = curr_node;
2000 curr_node = btrfs_next_delayed_node(curr_node);
2001 btrfs_release_delayed_node(prev_node);
2002 }
2003 }
2004
Linux with added FPC-III support