projects / deliverable / linux.git / blame_incremental
? search:
summary | shortlog | log | commit | commitdiff | tree
blob | blame (non-incremental) | history | HEAD
Linux 3.19-rc5
[deliverable/linux.git] / fs / btrfs / ctree.c
... / ...
CommitLineData
1/*
2 * Copyright (C) 2007,2008 Oracle. All rights reserved.
3 *
4 * This program is free software; you can redistribute it and/or
5 * modify it under the terms of the GNU General Public
6 * License v2 as published by the Free Software Foundation.
7 *
8 * This program is distributed in the hope that it will be useful,
9 * but WITHOUT ANY WARRANTY; without even the implied warranty of
10 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
11 * General Public License for more details.
12 *
13 * You should have received a copy of the GNU General Public
14 * License along with this program; if not, write to the
15 * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
16 * Boston, MA 021110-1307, USA.
17 */
18
19#include <linux/sched.h>
20#include <linux/slab.h>
21#include <linux/rbtree.h>
22#include "ctree.h"
23#include "disk-io.h"
24#include "transaction.h"
25#include "print-tree.h"
26#include "locking.h"
27
28static int split_node(struct btrfs_trans_handle *trans, struct btrfs_root
29 *root, struct btrfs_path *path, int level);
30static int split_leaf(struct btrfs_trans_handle *trans, struct btrfs_root
31 *root, struct btrfs_key *ins_key,
32 struct btrfs_path *path, int data_size, int extend);
33static int push_node_left(struct btrfs_trans_handle *trans,
34 struct btrfs_root *root, struct extent_buffer *dst,
35 struct extent_buffer *src, int empty);
36static int balance_node_right(struct btrfs_trans_handle *trans,
37 struct btrfs_root *root,
38 struct extent_buffer *dst_buf,
39 struct extent_buffer *src_buf);
40static void del_ptr(struct btrfs_root *root, struct btrfs_path *path,
41 int level, int slot);
42static int tree_mod_log_free_eb(struct btrfs_fs_info *fs_info,
43 struct extent_buffer *eb);
44
45struct btrfs_path *btrfs_alloc_path(void)
46{
47 struct btrfs_path *path;
48 path = kmem_cache_zalloc(btrfs_path_cachep, GFP_NOFS);
49 return path;
50}
51
52/*
53 * set all locked nodes in the path to blocking locks. This should
54 * be done before scheduling
55 */
56noinline void btrfs_set_path_blocking(struct btrfs_path *p)
57{
58 int i;
59 for (i = 0; i < BTRFS_MAX_LEVEL; i++) {
60 if (!p->nodes[i] || !p->locks[i])
61 continue;
62 btrfs_set_lock_blocking_rw(p->nodes[i], p->locks[i]);
63 if (p->locks[i] == BTRFS_READ_LOCK)
64 p->locks[i] = BTRFS_READ_LOCK_BLOCKING;
65 else if (p->locks[i] == BTRFS_WRITE_LOCK)
66 p->locks[i] = BTRFS_WRITE_LOCK_BLOCKING;
67 }
68}
69
70/*
71 * reset all the locked nodes in the patch to spinning locks.
72 *
73 * held is used to keep lockdep happy, when lockdep is enabled
74 * we set held to a blocking lock before we go around and
75 * retake all the spinlocks in the path. You can safely use NULL
76 * for held
77 */
78noinline void btrfs_clear_path_blocking(struct btrfs_path *p,
79 struct extent_buffer *held, int held_rw)
80{
81 int i;
82
83 if (held) {
84 btrfs_set_lock_blocking_rw(held, held_rw);
85 if (held_rw == BTRFS_WRITE_LOCK)
86 held_rw = BTRFS_WRITE_LOCK_BLOCKING;
87 else if (held_rw == BTRFS_READ_LOCK)
88 held_rw = BTRFS_READ_LOCK_BLOCKING;
89 }
90 btrfs_set_path_blocking(p);
91
92 for (i = BTRFS_MAX_LEVEL - 1; i >= 0; i--) {
93 if (p->nodes[i] && p->locks[i]) {
94 btrfs_clear_lock_blocking_rw(p->nodes[i], p->locks[i]);
95 if (p->locks[i] == BTRFS_WRITE_LOCK_BLOCKING)
96 p->locks[i] = BTRFS_WRITE_LOCK;
97 else if (p->locks[i] == BTRFS_READ_LOCK_BLOCKING)
98 p->locks[i] = BTRFS_READ_LOCK;
99 }
100 }
101
102 if (held)
103 btrfs_clear_lock_blocking_rw(held, held_rw);
104}
105
106/* this also releases the path */
107void btrfs_free_path(struct btrfs_path *p)
108{
109 if (!p)
110 return;
111 btrfs_release_path(p);
112 kmem_cache_free(btrfs_path_cachep, p);
113}
114
115/*
116 * path release drops references on the extent buffers in the path
117 * and it drops any locks held by this path
118 *
119 * It is safe to call this on paths that no locks or extent buffers held.
120 */
121noinline void btrfs_release_path(struct btrfs_path *p)
122{
123 int i;
124
125 for (i = 0; i < BTRFS_MAX_LEVEL; i++) {
126 p->slots[i] = 0;
127 if (!p->nodes[i])
128 continue;
129 if (p->locks[i]) {
130 btrfs_tree_unlock_rw(p->nodes[i], p->locks[i]);
131 p->locks[i] = 0;
132 }
133 free_extent_buffer(p->nodes[i]);
134 p->nodes[i] = NULL;
135 }
136}
137
138/*
139 * safely gets a reference on the root node of a tree. A lock
140 * is not taken, so a concurrent writer may put a different node
141 * at the root of the tree. See btrfs_lock_root_node for the
142 * looping required.
143 *
144 * The extent buffer returned by this has a reference taken, so
145 * it won't disappear. It may stop being the root of the tree
146 * at any time because there are no locks held.
147 */
148struct extent_buffer *btrfs_root_node(struct btrfs_root *root)
149{
150 struct extent_buffer *eb;
151
152 while (1) {
153 rcu_read_lock();
154 eb = rcu_dereference(root->node);
155
156 /*
157 * RCU really hurts here, we could free up the root node because
158 * it was cow'ed but we may not get the new root node yet so do
159 * the inc_not_zero dance and if it doesn't work then
160 * synchronize_rcu and try again.
161 */
162 if (atomic_inc_not_zero(&eb->refs)) {
163 rcu_read_unlock();
164 break;
165 }
166 rcu_read_unlock();
167 synchronize_rcu();
168 }
169 return eb;
170}
171
172/* loop around taking references on and locking the root node of the
173 * tree until you end up with a lock on the root. A locked buffer
174 * is returned, with a reference held.
175 */
176struct extent_buffer *btrfs_lock_root_node(struct btrfs_root *root)
177{
178 struct extent_buffer *eb;
179
180 while (1) {
181 eb = btrfs_root_node(root);
182 btrfs_tree_lock(eb);
183 if (eb == root->node)
184 break;
185 btrfs_tree_unlock(eb);
186 free_extent_buffer(eb);
187 }
188 return eb;
189}
190
191/* loop around taking references on and locking the root node of the
192 * tree until you end up with a lock on the root. A locked buffer
193 * is returned, with a reference held.
194 */
195static struct extent_buffer *btrfs_read_lock_root_node(struct btrfs_root *root)
196{
197 struct extent_buffer *eb;
198
199 while (1) {
200 eb = btrfs_root_node(root);
201 btrfs_tree_read_lock(eb);
202 if (eb == root->node)
203 break;
204 btrfs_tree_read_unlock(eb);
205 free_extent_buffer(eb);
206 }
207 return eb;
208}
209
210/* cowonly root (everything not a reference counted cow subvolume), just get
211 * put onto a simple dirty list. transaction.c walks this to make sure they
212 * get properly updated on disk.
213 */
214static void add_root_to_dirty_list(struct btrfs_root *root)
215{
216 spin_lock(&root->fs_info->trans_lock);
217 if (test_bit(BTRFS_ROOT_TRACK_DIRTY, &root->state) &&
218 list_empty(&root->dirty_list)) {
219 list_add(&root->dirty_list,
220 &root->fs_info->dirty_cowonly_roots);
221 }
222 spin_unlock(&root->fs_info->trans_lock);
223}
224
225/*
226 * used by snapshot creation to make a copy of a root for a tree with
227 * a given objectid. The buffer with the new root node is returned in
228 * cow_ret, and this func returns zero on success or a negative error code.
229 */
230int btrfs_copy_root(struct btrfs_trans_handle *trans,
231 struct btrfs_root *root,
232 struct extent_buffer *buf,
233 struct extent_buffer **cow_ret, u64 new_root_objectid)
234{
235 struct extent_buffer *cow;
236 int ret = 0;
237 int level;
238 struct btrfs_disk_key disk_key;
239
240 WARN_ON(test_bit(BTRFS_ROOT_REF_COWS, &root->state) &&
241 trans->transid != root->fs_info->running_transaction->transid);
242 WARN_ON(test_bit(BTRFS_ROOT_REF_COWS, &root->state) &&
243 trans->transid != root->last_trans);
244
245 level = btrfs_header_level(buf);
246 if (level == 0)
247 btrfs_item_key(buf, &disk_key, 0);
248 else
249 btrfs_node_key(buf, &disk_key, 0);
250
251 cow = btrfs_alloc_tree_block(trans, root, 0, new_root_objectid,
252 &disk_key, level, buf->start, 0);
253 if (IS_ERR(cow))
254 return PTR_ERR(cow);
255
256 copy_extent_buffer(cow, buf, 0, 0, cow->len);
257 btrfs_set_header_bytenr(cow, cow->start);
258 btrfs_set_header_generation(cow, trans->transid);
259 btrfs_set_header_backref_rev(cow, BTRFS_MIXED_BACKREF_REV);
260 btrfs_clear_header_flag(cow, BTRFS_HEADER_FLAG_WRITTEN |
261 BTRFS_HEADER_FLAG_RELOC);
262 if (new_root_objectid == BTRFS_TREE_RELOC_OBJECTID)
263 btrfs_set_header_flag(cow, BTRFS_HEADER_FLAG_RELOC);
264 else
265 btrfs_set_header_owner(cow, new_root_objectid);
266
267 write_extent_buffer(cow, root->fs_info->fsid, btrfs_header_fsid(),
268 BTRFS_FSID_SIZE);
269
270 WARN_ON(btrfs_header_generation(buf) > trans->transid);
271 if (new_root_objectid == BTRFS_TREE_RELOC_OBJECTID)
272 ret = btrfs_inc_ref(trans, root, cow, 1);
273 else
274 ret = btrfs_inc_ref(trans, root, cow, 0);
275
276 if (ret)
277 return ret;
278
279 btrfs_mark_buffer_dirty(cow);
280 *cow_ret = cow;
281 return 0;
282}
283
284enum mod_log_op {
285 MOD_LOG_KEY_REPLACE,
286 MOD_LOG_KEY_ADD,
287 MOD_LOG_KEY_REMOVE,
288 MOD_LOG_KEY_REMOVE_WHILE_FREEING,
289 MOD_LOG_KEY_REMOVE_WHILE_MOVING,
290 MOD_LOG_MOVE_KEYS,
291 MOD_LOG_ROOT_REPLACE,
292};
293
294struct tree_mod_move {
295 int dst_slot;
296 int nr_items;
297};
298
299struct tree_mod_root {
300 u64 logical;
301 u8 level;
302};
303
304struct tree_mod_elem {
305 struct rb_node node;
306 u64 index; /* shifted logical */
307 u64 seq;
308 enum mod_log_op op;
309
310 /* this is used for MOD_LOG_KEY_* and MOD_LOG_MOVE_KEYS operations */
311 int slot;
312
313 /* this is used for MOD_LOG_KEY* and MOD_LOG_ROOT_REPLACE */
314 u64 generation;
315
316 /* those are used for op == MOD_LOG_KEY_{REPLACE,REMOVE} */
317 struct btrfs_disk_key key;
318 u64 blockptr;
319
320 /* this is used for op == MOD_LOG_MOVE_KEYS */
321 struct tree_mod_move move;
322
323 /* this is used for op == MOD_LOG_ROOT_REPLACE */
324 struct tree_mod_root old_root;
325};
326
327static inline void tree_mod_log_read_lock(struct btrfs_fs_info *fs_info)
328{
329 read_lock(&fs_info->tree_mod_log_lock);
330}
331
332static inline void tree_mod_log_read_unlock(struct btrfs_fs_info *fs_info)
333{
334 read_unlock(&fs_info->tree_mod_log_lock);
335}
336
337static inline void tree_mod_log_write_lock(struct btrfs_fs_info *fs_info)
338{
339 write_lock(&fs_info->tree_mod_log_lock);
340}
341
342static inline void tree_mod_log_write_unlock(struct btrfs_fs_info *fs_info)
343{
344 write_unlock(&fs_info->tree_mod_log_lock);
345}
346
347/*
348 * Pull a new tree mod seq number for our operation.
349 */
350static inline u64 btrfs_inc_tree_mod_seq(struct btrfs_fs_info *fs_info)
351{
352 return atomic64_inc_return(&fs_info->tree_mod_seq);
353}
354
355/*
356 * This adds a new blocker to the tree mod log's blocker list if the @elem
357 * passed does not already have a sequence number set. So when a caller expects
358 * to record tree modifications, it should ensure to set elem->seq to zero
359 * before calling btrfs_get_tree_mod_seq.
360 * Returns a fresh, unused tree log modification sequence number, even if no new
361 * blocker was added.
362 */
363u64 btrfs_get_tree_mod_seq(struct btrfs_fs_info *fs_info,
364 struct seq_list *elem)
365{
366 tree_mod_log_write_lock(fs_info);
367 spin_lock(&fs_info->tree_mod_seq_lock);
368 if (!elem->seq) {
369 elem->seq = btrfs_inc_tree_mod_seq(fs_info);
370 list_add_tail(&elem->list, &fs_info->tree_mod_seq_list);
371 }
372 spin_unlock(&fs_info->tree_mod_seq_lock);
373 tree_mod_log_write_unlock(fs_info);
374
375 return elem->seq;
376}
377
378void btrfs_put_tree_mod_seq(struct btrfs_fs_info *fs_info,
379 struct seq_list *elem)
380{
381 struct rb_root *tm_root;
382 struct rb_node *node;
383 struct rb_node *next;
384 struct seq_list *cur_elem;
385 struct tree_mod_elem *tm;
386 u64 min_seq = (u64)-1;
387 u64 seq_putting = elem->seq;
388
389 if (!seq_putting)
390 return;
391
392 spin_lock(&fs_info->tree_mod_seq_lock);
393 list_del(&elem->list);
394 elem->seq = 0;
395
396 list_for_each_entry(cur_elem, &fs_info->tree_mod_seq_list, list) {
397 if (cur_elem->seq < min_seq) {
398 if (seq_putting > cur_elem->seq) {
399 /*
400 * blocker with lower sequence number exists, we
401 * cannot remove anything from the log
402 */
403 spin_unlock(&fs_info->tree_mod_seq_lock);
404 return;
405 }
406 min_seq = cur_elem->seq;
407 }
408 }
409 spin_unlock(&fs_info->tree_mod_seq_lock);
410
411 /*
412 * anything that's lower than the lowest existing (read: blocked)
413 * sequence number can be removed from the tree.
414 */
415 tree_mod_log_write_lock(fs_info);
416 tm_root = &fs_info->tree_mod_log;
417 for (node = rb_first(tm_root); node; node = next) {
418 next = rb_next(node);
419 tm = container_of(node, struct tree_mod_elem, node);
420 if (tm->seq > min_seq)
421 continue;
422 rb_erase(node, tm_root);
423 kfree(tm);
424 }
425 tree_mod_log_write_unlock(fs_info);
426}
427
428/*
429 * key order of the log:
430 * index -> sequence
431 *
432 * the index is the shifted logical of the *new* root node for root replace
433 * operations, or the shifted logical of the affected block for all other
434 * operations.
435 *
436 * Note: must be called with write lock (tree_mod_log_write_lock).
437 */
438static noinline int
439__tree_mod_log_insert(struct btrfs_fs_info *fs_info, struct tree_mod_elem *tm)
440{
441 struct rb_root *tm_root;
442 struct rb_node **new;
443 struct rb_node *parent = NULL;
444 struct tree_mod_elem *cur;
445
446 BUG_ON(!tm);
447
448 tm->seq = btrfs_inc_tree_mod_seq(fs_info);
449
450 tm_root = &fs_info->tree_mod_log;
451 new = &tm_root->rb_node;
452 while (*new) {
453 cur = container_of(*new, struct tree_mod_elem, node);
454 parent = *new;
455 if (cur->index < tm->index)
456 new = &((*new)->rb_left);
457 else if (cur->index > tm->index)
458 new = &((*new)->rb_right);
459 else if (cur->seq < tm->seq)
460 new = &((*new)->rb_left);
461 else if (cur->seq > tm->seq)
462 new = &((*new)->rb_right);
463 else
464 return -EEXIST;
465 }
466
467 rb_link_node(&tm->node, parent, new);
468 rb_insert_color(&tm->node, tm_root);
469 return 0;
470}
471
472/*
473 * Determines if logging can be omitted. Returns 1 if it can. Otherwise, it
474 * returns zero with the tree_mod_log_lock acquired. The caller must hold
475 * this until all tree mod log insertions are recorded in the rb tree and then
476 * call tree_mod_log_write_unlock() to release.
477 */
478static inline int tree_mod_dont_log(struct btrfs_fs_info *fs_info,
479 struct extent_buffer *eb) {
480 smp_mb();
481 if (list_empty(&(fs_info)->tree_mod_seq_list))
482 return 1;
483 if (eb && btrfs_header_level(eb) == 0)
484 return 1;
485
486 tree_mod_log_write_lock(fs_info);
487 if (list_empty(&(fs_info)->tree_mod_seq_list)) {
488 tree_mod_log_write_unlock(fs_info);
489 return 1;
490 }
491
492 return 0;
493}
494
495/* Similar to tree_mod_dont_log, but doesn't acquire any locks. */
496static inline int tree_mod_need_log(const struct btrfs_fs_info *fs_info,
497 struct extent_buffer *eb)
498{
499 smp_mb();
500 if (list_empty(&(fs_info)->tree_mod_seq_list))
501 return 0;
502 if (eb && btrfs_header_level(eb) == 0)
503 return 0;
504
505 return 1;
506}
507
508static struct tree_mod_elem *
509alloc_tree_mod_elem(struct extent_buffer *eb, int slot,
510 enum mod_log_op op, gfp_t flags)
511{
512 struct tree_mod_elem *tm;
513
514 tm = kzalloc(sizeof(*tm), flags);
515 if (!tm)
516 return NULL;
517
518 tm->index = eb->start >> PAGE_CACHE_SHIFT;
519 if (op != MOD_LOG_KEY_ADD) {
520 btrfs_node_key(eb, &tm->key, slot);
521 tm->blockptr = btrfs_node_blockptr(eb, slot);
522 }
523 tm->op = op;
524 tm->slot = slot;
525 tm->generation = btrfs_node_ptr_generation(eb, slot);
526 RB_CLEAR_NODE(&tm->node);
527
528 return tm;
529}
530
531static noinline int
532tree_mod_log_insert_key(struct btrfs_fs_info *fs_info,
533 struct extent_buffer *eb, int slot,
534 enum mod_log_op op, gfp_t flags)
535{
536 struct tree_mod_elem *tm;
537 int ret;
538
539 if (!tree_mod_need_log(fs_info, eb))
540 return 0;
541
542 tm = alloc_tree_mod_elem(eb, slot, op, flags);
543 if (!tm)
544 return -ENOMEM;
545
546 if (tree_mod_dont_log(fs_info, eb)) {
547 kfree(tm);
548 return 0;
549 }
550
551 ret = __tree_mod_log_insert(fs_info, tm);
552 tree_mod_log_write_unlock(fs_info);
553 if (ret)
554 kfree(tm);
555
556 return ret;
557}
558
559static noinline int
560tree_mod_log_insert_move(struct btrfs_fs_info *fs_info,
561 struct extent_buffer *eb, int dst_slot, int src_slot,
562 int nr_items, gfp_t flags)
563{
564 struct tree_mod_elem *tm = NULL;
565 struct tree_mod_elem **tm_list = NULL;
566 int ret = 0;
567 int i;
568 int locked = 0;
569
570 if (!tree_mod_need_log(fs_info, eb))
571 return 0;
572
573 tm_list = kzalloc(nr_items * sizeof(struct tree_mod_elem *), flags);
574 if (!tm_list)
575 return -ENOMEM;
576
577 tm = kzalloc(sizeof(*tm), flags);
578 if (!tm) {
579 ret = -ENOMEM;
580 goto free_tms;
581 }
582
583 tm->index = eb->start >> PAGE_CACHE_SHIFT;
584 tm->slot = src_slot;
585 tm->move.dst_slot = dst_slot;
586 tm->move.nr_items = nr_items;
587 tm->op = MOD_LOG_MOVE_KEYS;
588
589 for (i = 0; i + dst_slot < src_slot && i < nr_items; i++) {
590 tm_list[i] = alloc_tree_mod_elem(eb, i + dst_slot,
591 MOD_LOG_KEY_REMOVE_WHILE_MOVING, flags);
592 if (!tm_list[i]) {
593 ret = -ENOMEM;
594 goto free_tms;
595 }
596 }
597
598 if (tree_mod_dont_log(fs_info, eb))
599 goto free_tms;
600 locked = 1;
601
602 /*
603 * When we override something during the move, we log these removals.
604 * This can only happen when we move towards the beginning of the
605 * buffer, i.e. dst_slot < src_slot.
606 */
607 for (i = 0; i + dst_slot < src_slot && i < nr_items; i++) {
608 ret = __tree_mod_log_insert(fs_info, tm_list[i]);
609 if (ret)
610 goto free_tms;
611 }
612
613 ret = __tree_mod_log_insert(fs_info, tm);
614 if (ret)
615 goto free_tms;
616 tree_mod_log_write_unlock(fs_info);
617 kfree(tm_list);
618
619 return 0;
620free_tms:
621 for (i = 0; i < nr_items; i++) {
622 if (tm_list[i] && !RB_EMPTY_NODE(&tm_list[i]->node))
623 rb_erase(&tm_list[i]->node, &fs_info->tree_mod_log);
624 kfree(tm_list[i]);
625 }
626 if (locked)
627 tree_mod_log_write_unlock(fs_info);
628 kfree(tm_list);
629 kfree(tm);
630
631 return ret;
632}
633
634static inline int
635__tree_mod_log_free_eb(struct btrfs_fs_info *fs_info,
636 struct tree_mod_elem **tm_list,
637 int nritems)
638{
639 int i, j;
640 int ret;
641
642 for (i = nritems - 1; i >= 0; i--) {
643 ret = __tree_mod_log_insert(fs_info, tm_list[i]);
644 if (ret) {
645 for (j = nritems - 1; j > i; j--)
646 rb_erase(&tm_list[j]->node,
647 &fs_info->tree_mod_log);
648 return ret;
649 }
650 }
651
652 return 0;
653}
654
655static noinline int
656tree_mod_log_insert_root(struct btrfs_fs_info *fs_info,
657 struct extent_buffer *old_root,
658 struct extent_buffer *new_root, gfp_t flags,
659 int log_removal)
660{
661 struct tree_mod_elem *tm = NULL;
662 struct tree_mod_elem **tm_list = NULL;
663 int nritems = 0;
664 int ret = 0;
665 int i;
666
667 if (!tree_mod_need_log(fs_info, NULL))
668 return 0;
669
670 if (log_removal && btrfs_header_level(old_root) > 0) {
671 nritems = btrfs_header_nritems(old_root);
672 tm_list = kzalloc(nritems * sizeof(struct tree_mod_elem *),
673 flags);
674 if (!tm_list) {
675 ret = -ENOMEM;
676 goto free_tms;
677 }
678 for (i = 0; i < nritems; i++) {
679 tm_list[i] = alloc_tree_mod_elem(old_root, i,
680 MOD_LOG_KEY_REMOVE_WHILE_FREEING, flags);
681 if (!tm_list[i]) {
682 ret = -ENOMEM;
683 goto free_tms;
684 }
685 }
686 }
687
688 tm = kzalloc(sizeof(*tm), flags);
689 if (!tm) {
690 ret = -ENOMEM;
691 goto free_tms;
692 }
693
694 tm->index = new_root->start >> PAGE_CACHE_SHIFT;
695 tm->old_root.logical = old_root->start;
696 tm->old_root.level = btrfs_header_level(old_root);
697 tm->generation = btrfs_header_generation(old_root);
698 tm->op = MOD_LOG_ROOT_REPLACE;
699
700 if (tree_mod_dont_log(fs_info, NULL))
701 goto free_tms;
702
703 if (tm_list)
704 ret = __tree_mod_log_free_eb(fs_info, tm_list, nritems);
705 if (!ret)
706 ret = __tree_mod_log_insert(fs_info, tm);
707
708 tree_mod_log_write_unlock(fs_info);
709 if (ret)
710 goto free_tms;
711 kfree(tm_list);
712
713 return ret;
714
715free_tms:
716 if (tm_list) {
717 for (i = 0; i < nritems; i++)
718 kfree(tm_list[i]);
719 kfree(tm_list);
720 }
721 kfree(tm);
722
723 return ret;
724}
725
726static struct tree_mod_elem *
727__tree_mod_log_search(struct btrfs_fs_info *fs_info, u64 start, u64 min_seq,
728 int smallest)
729{
730 struct rb_root *tm_root;
731 struct rb_node *node;
732 struct tree_mod_elem *cur = NULL;
733 struct tree_mod_elem *found = NULL;
734 u64 index = start >> PAGE_CACHE_SHIFT;
735
736 tree_mod_log_read_lock(fs_info);
737 tm_root = &fs_info->tree_mod_log;
738 node = tm_root->rb_node;
739 while (node) {
740 cur = container_of(node, struct tree_mod_elem, node);
741 if (cur->index < index) {
742 node = node->rb_left;
743 } else if (cur->index > index) {
744 node = node->rb_right;
745 } else if (cur->seq < min_seq) {
746 node = node->rb_left;
747 } else if (!smallest) {
748 /* we want the node with the highest seq */
749 if (found)
750 BUG_ON(found->seq > cur->seq);
751 found = cur;
752 node = node->rb_left;
753 } else if (cur->seq > min_seq) {
754 /* we want the node with the smallest seq */
755 if (found)
756 BUG_ON(found->seq < cur->seq);
757 found = cur;
758 node = node->rb_right;
759 } else {
760 found = cur;
761 break;
762 }
763 }
764 tree_mod_log_read_unlock(fs_info);
765
766 return found;
767}
768
769/*
770 * this returns the element from the log with the smallest time sequence
771 * value that's in the log (the oldest log item). any element with a time
772 * sequence lower than min_seq will be ignored.
773 */
774static struct tree_mod_elem *
775tree_mod_log_search_oldest(struct btrfs_fs_info *fs_info, u64 start,
776 u64 min_seq)
777{
778 return __tree_mod_log_search(fs_info, start, min_seq, 1);
779}
780
781/*
782 * this returns the element from the log with the largest time sequence
783 * value that's in the log (the most recent log item). any element with
784 * a time sequence lower than min_seq will be ignored.
785 */
786static struct tree_mod_elem *
787tree_mod_log_search(struct btrfs_fs_info *fs_info, u64 start, u64 min_seq)
788{
789 return __tree_mod_log_search(fs_info, start, min_seq, 0);
790}
791
792static noinline int
793tree_mod_log_eb_copy(struct btrfs_fs_info *fs_info, struct extent_buffer *dst,
794 struct extent_buffer *src, unsigned long dst_offset,
795 unsigned long src_offset, int nr_items)
796{
797 int ret = 0;
798 struct tree_mod_elem **tm_list = NULL;
799 struct tree_mod_elem **tm_list_add, **tm_list_rem;
800 int i;
801 int locked = 0;
802
803 if (!tree_mod_need_log(fs_info, NULL))
804 return 0;
805
806 if (btrfs_header_level(dst) == 0 && btrfs_header_level(src) == 0)
807 return 0;
808
809 tm_list = kzalloc(nr_items * 2 * sizeof(struct tree_mod_elem *),
810 GFP_NOFS);
811 if (!tm_list)
812 return -ENOMEM;
813
814 tm_list_add = tm_list;
815 tm_list_rem = tm_list + nr_items;
816 for (i = 0; i < nr_items; i++) {
817 tm_list_rem[i] = alloc_tree_mod_elem(src, i + src_offset,
818 MOD_LOG_KEY_REMOVE, GFP_NOFS);
819 if (!tm_list_rem[i]) {
820 ret = -ENOMEM;
821 goto free_tms;
822 }
823
824 tm_list_add[i] = alloc_tree_mod_elem(dst, i + dst_offset,
825 MOD_LOG_KEY_ADD, GFP_NOFS);
826 if (!tm_list_add[i]) {
827 ret = -ENOMEM;
828 goto free_tms;
829 }
830 }
831
832 if (tree_mod_dont_log(fs_info, NULL))
833 goto free_tms;
834 locked = 1;
835
836 for (i = 0; i < nr_items; i++) {
837 ret = __tree_mod_log_insert(fs_info, tm_list_rem[i]);
838 if (ret)
839 goto free_tms;
840 ret = __tree_mod_log_insert(fs_info, tm_list_add[i]);
841 if (ret)
842 goto free_tms;
843 }
844
845 tree_mod_log_write_unlock(fs_info);
846 kfree(tm_list);
847
848 return 0;
849
850free_tms:
851 for (i = 0; i < nr_items * 2; i++) {
852 if (tm_list[i] && !RB_EMPTY_NODE(&tm_list[i]->node))
853 rb_erase(&tm_list[i]->node, &fs_info->tree_mod_log);
854 kfree(tm_list[i]);
855 }
856 if (locked)
857 tree_mod_log_write_unlock(fs_info);
858 kfree(tm_list);
859
860 return ret;
861}
862
863static inline void
864tree_mod_log_eb_move(struct btrfs_fs_info *fs_info, struct extent_buffer *dst,
865 int dst_offset, int src_offset, int nr_items)
866{
867 int ret;
868 ret = tree_mod_log_insert_move(fs_info, dst, dst_offset, src_offset,
869 nr_items, GFP_NOFS);
870 BUG_ON(ret < 0);
871}
872
873static noinline void
874tree_mod_log_set_node_key(struct btrfs_fs_info *fs_info,
875 struct extent_buffer *eb, int slot, int atomic)
876{
877 int ret;
878
879 ret = tree_mod_log_insert_key(fs_info, eb, slot,
880 MOD_LOG_KEY_REPLACE,
881 atomic ? GFP_ATOMIC : GFP_NOFS);
882 BUG_ON(ret < 0);
883}
884
885static noinline int
886tree_mod_log_free_eb(struct btrfs_fs_info *fs_info, struct extent_buffer *eb)
887{
888 struct tree_mod_elem **tm_list = NULL;
889 int nritems = 0;
890 int i;
891 int ret = 0;
892
893 if (btrfs_header_level(eb) == 0)
894 return 0;
895
896 if (!tree_mod_need_log(fs_info, NULL))
897 return 0;
898
899 nritems = btrfs_header_nritems(eb);
900 tm_list = kzalloc(nritems * sizeof(struct tree_mod_elem *),
901 GFP_NOFS);
902 if (!tm_list)
903 return -ENOMEM;
904
905 for (i = 0; i < nritems; i++) {
906 tm_list[i] = alloc_tree_mod_elem(eb, i,
907 MOD_LOG_KEY_REMOVE_WHILE_FREEING, GFP_NOFS);
908 if (!tm_list[i]) {
909 ret = -ENOMEM;
910 goto free_tms;
911 }
912 }
913
914 if (tree_mod_dont_log(fs_info, eb))
915 goto free_tms;
916
917 ret = __tree_mod_log_free_eb(fs_info, tm_list, nritems);
918 tree_mod_log_write_unlock(fs_info);
919 if (ret)
920 goto free_tms;
921 kfree(tm_list);
922
923 return 0;
924
925free_tms:
926 for (i = 0; i < nritems; i++)
927 kfree(tm_list[i]);
928 kfree(tm_list);
929
930 return ret;
931}
932
933static noinline void
934tree_mod_log_set_root_pointer(struct btrfs_root *root,
935 struct extent_buffer *new_root_node,
936 int log_removal)
937{
938 int ret;
939 ret = tree_mod_log_insert_root(root->fs_info, root->node,
940 new_root_node, GFP_NOFS, log_removal);
941 BUG_ON(ret < 0);
942}
943
944/*
945 * check if the tree block can be shared by multiple trees
946 */
947int btrfs_block_can_be_shared(struct btrfs_root *root,
948 struct extent_buffer *buf)
949{
950 /*
951 * Tree blocks not in refernece counted trees and tree roots
952 * are never shared. If a block was allocated after the last
953 * snapshot and the block was not allocated by tree relocation,
954 * we know the block is not shared.
955 */
956 if (test_bit(BTRFS_ROOT_REF_COWS, &root->state) &&
957 buf != root->node && buf != root->commit_root &&
958 (btrfs_header_generation(buf) <=
959 btrfs_root_last_snapshot(&root->root_item) ||
960 btrfs_header_flag(buf, BTRFS_HEADER_FLAG_RELOC)))
961 return 1;
962#ifdef BTRFS_COMPAT_EXTENT_TREE_V0
963 if (test_bit(BTRFS_ROOT_REF_COWS, &root->state) &&
964 btrfs_header_backref_rev(buf) < BTRFS_MIXED_BACKREF_REV)
965 return 1;
966#endif
967 return 0;
968}
969
970static noinline int update_ref_for_cow(struct btrfs_trans_handle *trans,
971 struct btrfs_root *root,
972 struct extent_buffer *buf,
973 struct extent_buffer *cow,
974 int *last_ref)
975{
976 u64 refs;
977 u64 owner;
978 u64 flags;
979 u64 new_flags = 0;
980 int ret;
981
982 /*
983 * Backrefs update rules:
984 *
985 * Always use full backrefs for extent pointers in tree block
986 * allocated by tree relocation.
987 *
988 * If a shared tree block is no longer referenced by its owner
989 * tree (btrfs_header_owner(buf) == root->root_key.objectid),
990 * use full backrefs for extent pointers in tree block.
991 *
992 * If a tree block is been relocating
993 * (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID),
994 * use full backrefs for extent pointers in tree block.
995 * The reason for this is some operations (such as drop tree)
996 * are only allowed for blocks use full backrefs.
997 */
998
999 if (btrfs_block_can_be_shared(root, buf)) {
1000 ret = btrfs_lookup_extent_info(trans, root, buf->start,
1001 btrfs_header_level(buf), 1,
1002 &refs, &flags);
1003 if (ret)
1004 return ret;
1005 if (refs == 0) {
1006 ret = -EROFS;
1007 btrfs_std_error(root->fs_info, ret);
1008 return ret;
1009 }
1010 } else {
1011 refs = 1;
1012 if (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID ||
1013 btrfs_header_backref_rev(buf) < BTRFS_MIXED_BACKREF_REV)
1014 flags = BTRFS_BLOCK_FLAG_FULL_BACKREF;
1015 else
1016 flags = 0;
1017 }
1018
1019 owner = btrfs_header_owner(buf);
1020 BUG_ON(owner == BTRFS_TREE_RELOC_OBJECTID &&
1021 !(flags & BTRFS_BLOCK_FLAG_FULL_BACKREF));
1022
1023 if (refs > 1) {
1024 if ((owner == root->root_key.objectid ||
1025 root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID) &&
1026 !(flags & BTRFS_BLOCK_FLAG_FULL_BACKREF)) {
1027 ret = btrfs_inc_ref(trans, root, buf, 1);
1028 BUG_ON(ret); /* -ENOMEM */
1029
1030 if (root->root_key.objectid ==
1031 BTRFS_TREE_RELOC_OBJECTID) {
1032 ret = btrfs_dec_ref(trans, root, buf, 0);
1033 BUG_ON(ret); /* -ENOMEM */
1034 ret = btrfs_inc_ref(trans, root, cow, 1);
1035 BUG_ON(ret); /* -ENOMEM */
1036 }
1037 new_flags |= BTRFS_BLOCK_FLAG_FULL_BACKREF;
1038 } else {
1039
1040 if (root->root_key.objectid ==
1041 BTRFS_TREE_RELOC_OBJECTID)
1042 ret = btrfs_inc_ref(trans, root, cow, 1);
1043 else
1044 ret = btrfs_inc_ref(trans, root, cow, 0);
1045 BUG_ON(ret); /* -ENOMEM */
1046 }
1047 if (new_flags != 0) {
1048 int level = btrfs_header_level(buf);
1049
1050 ret = btrfs_set_disk_extent_flags(trans, root,
1051 buf->start,
1052 buf->len,
1053 new_flags, level, 0);
1054 if (ret)
1055 return ret;
1056 }
1057 } else {
1058 if (flags & BTRFS_BLOCK_FLAG_FULL_BACKREF) {
1059 if (root->root_key.objectid ==
1060 BTRFS_TREE_RELOC_OBJECTID)
1061 ret = btrfs_inc_ref(trans, root, cow, 1);
1062 else
1063 ret = btrfs_inc_ref(trans, root, cow, 0);
1064 BUG_ON(ret); /* -ENOMEM */
1065 ret = btrfs_dec_ref(trans, root, buf, 1);
1066 BUG_ON(ret); /* -ENOMEM */
1067 }
1068 clean_tree_block(trans, root, buf);
1069 *last_ref = 1;
1070 }
1071 return 0;
1072}
1073
1074/*
1075 * does the dirty work in cow of a single block. The parent block (if
1076 * supplied) is updated to point to the new cow copy. The new buffer is marked
1077 * dirty and returned locked. If you modify the block it needs to be marked
1078 * dirty again.
1079 *
1080 * search_start -- an allocation hint for the new block
1081 *
1082 * empty_size -- a hint that you plan on doing more cow. This is the size in
1083 * bytes the allocator should try to find free next to the block it returns.
1084 * This is just a hint and may be ignored by the allocator.
1085 */
1086static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans,
1087 struct btrfs_root *root,
1088 struct extent_buffer *buf,
1089 struct extent_buffer *parent, int parent_slot,
1090 struct extent_buffer **cow_ret,
1091 u64 search_start, u64 empty_size)
1092{
1093 struct btrfs_disk_key disk_key;
1094 struct extent_buffer *cow;
1095 int level, ret;
1096 int last_ref = 0;
1097 int unlock_orig = 0;
1098 u64 parent_start;
1099
1100 if (*cow_ret == buf)
1101 unlock_orig = 1;
1102
1103 btrfs_assert_tree_locked(buf);
1104
1105 WARN_ON(test_bit(BTRFS_ROOT_REF_COWS, &root->state) &&
1106 trans->transid != root->fs_info->running_transaction->transid);
1107 WARN_ON(test_bit(BTRFS_ROOT_REF_COWS, &root->state) &&
1108 trans->transid != root->last_trans);
1109
1110 level = btrfs_header_level(buf);
1111
1112 if (level == 0)
1113 btrfs_item_key(buf, &disk_key, 0);
1114 else
1115 btrfs_node_key(buf, &disk_key, 0);
1116
1117 if (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID) {
1118 if (parent)
1119 parent_start = parent->start;
1120 else
1121 parent_start = 0;
1122 } else
1123 parent_start = 0;
1124
1125 cow = btrfs_alloc_tree_block(trans, root, parent_start,
1126 root->root_key.objectid, &disk_key, level,
1127 search_start, empty_size);
1128 if (IS_ERR(cow))
1129 return PTR_ERR(cow);
1130
1131 /* cow is set to blocking by btrfs_init_new_buffer */
1132
1133 copy_extent_buffer(cow, buf, 0, 0, cow->len);
1134 btrfs_set_header_bytenr(cow, cow->start);
1135 btrfs_set_header_generation(cow, trans->transid);
1136 btrfs_set_header_backref_rev(cow, BTRFS_MIXED_BACKREF_REV);
1137 btrfs_clear_header_flag(cow, BTRFS_HEADER_FLAG_WRITTEN |
1138 BTRFS_HEADER_FLAG_RELOC);
1139 if (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID)
1140 btrfs_set_header_flag(cow, BTRFS_HEADER_FLAG_RELOC);
1141 else
1142 btrfs_set_header_owner(cow, root->root_key.objectid);
1143
1144 write_extent_buffer(cow, root->fs_info->fsid, btrfs_header_fsid(),
1145 BTRFS_FSID_SIZE);
1146
1147 ret = update_ref_for_cow(trans, root, buf, cow, &last_ref);
1148 if (ret) {
1149 btrfs_abort_transaction(trans, root, ret);
1150 return ret;
1151 }
1152
1153 if (test_bit(BTRFS_ROOT_REF_COWS, &root->state)) {
1154 ret = btrfs_reloc_cow_block(trans, root, buf, cow);
1155 if (ret)
1156 return ret;
1157 }
1158
1159 if (buf == root->node) {
1160 WARN_ON(parent && parent != buf);
1161 if (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID ||
1162 btrfs_header_backref_rev(buf) < BTRFS_MIXED_BACKREF_REV)
1163 parent_start = buf->start;
1164 else
1165 parent_start = 0;
1166
1167 extent_buffer_get(cow);
1168 tree_mod_log_set_root_pointer(root, cow, 1);
1169 rcu_assign_pointer(root->node, cow);
1170
1171 btrfs_free_tree_block(trans, root, buf, parent_start,
1172 last_ref);
1173 free_extent_buffer(buf);
1174 add_root_to_dirty_list(root);
1175 } else {
1176 if (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID)
1177 parent_start = parent->start;
1178 else
1179 parent_start = 0;
1180
1181 WARN_ON(trans->transid != btrfs_header_generation(parent));
1182 tree_mod_log_insert_key(root->fs_info, parent, parent_slot,
1183 MOD_LOG_KEY_REPLACE, GFP_NOFS);
1184 btrfs_set_node_blockptr(parent, parent_slot,
1185 cow->start);
1186 btrfs_set_node_ptr_generation(parent, parent_slot,
1187 trans->transid);
1188 btrfs_mark_buffer_dirty(parent);
1189 if (last_ref) {
1190 ret = tree_mod_log_free_eb(root->fs_info, buf);
1191 if (ret) {
1192 btrfs_abort_transaction(trans, root, ret);
1193 return ret;
1194 }
1195 }
1196 btrfs_free_tree_block(trans, root, buf, parent_start,
1197 last_ref);
1198 }
1199 if (unlock_orig)
1200 btrfs_tree_unlock(buf);
1201 free_extent_buffer_stale(buf);
1202 btrfs_mark_buffer_dirty(cow);
1203 *cow_ret = cow;
1204 return 0;
1205}
1206
1207/*
1208 * returns the logical address of the oldest predecessor of the given root.
1209 * entries older than time_seq are ignored.
1210 */
1211static struct tree_mod_elem *
1212__tree_mod_log_oldest_root(struct btrfs_fs_info *fs_info,
1213 struct extent_buffer *eb_root, u64 time_seq)
1214{
1215 struct tree_mod_elem *tm;
1216 struct tree_mod_elem *found = NULL;
1217 u64 root_logical = eb_root->start;
1218 int looped = 0;
1219
1220 if (!time_seq)
1221 return NULL;
1222
1223 /*
1224 * the very last operation that's logged for a root is the replacement
1225 * operation (if it is replaced at all). this has the index of the *new*
1226 * root, making it the very first operation that's logged for this root.
1227 */
1228 while (1) {
1229 tm = tree_mod_log_search_oldest(fs_info, root_logical,
1230 time_seq);
1231 if (!looped && !tm)
1232 return NULL;
1233 /*
1234 * if there are no tree operation for the oldest root, we simply
1235 * return it. this should only happen if that (old) root is at
1236 * level 0.
1237 */
1238 if (!tm)
1239 break;
1240
1241 /*
1242 * if there's an operation that's not a root replacement, we
1243 * found the oldest version of our root. normally, we'll find a
1244 * MOD_LOG_KEY_REMOVE_WHILE_FREEING operation here.
1245 */
1246 if (tm->op != MOD_LOG_ROOT_REPLACE)
1247 break;
1248
1249 found = tm;
1250 root_logical = tm->old_root.logical;
1251 looped = 1;
1252 }
1253
1254 /* if there's no old root to return, return what we found instead */
1255 if (!found)
1256 found = tm;
1257
1258 return found;
1259}
1260
1261/*
1262 * tm is a pointer to the first operation to rewind within eb. then, all
1263 * previous operations will be rewinded (until we reach something older than
1264 * time_seq).
1265 */
1266static void
1267__tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct extent_buffer *eb,
1268 u64 time_seq, struct tree_mod_elem *first_tm)
1269{
1270 u32 n;
1271 struct rb_node *next;
1272 struct tree_mod_elem *tm = first_tm;
1273 unsigned long o_dst;
1274 unsigned long o_src;
1275 unsigned long p_size = sizeof(struct btrfs_key_ptr);
1276
1277 n = btrfs_header_nritems(eb);
1278 tree_mod_log_read_lock(fs_info);
1279 while (tm && tm->seq >= time_seq) {
1280 /*
1281 * all the operations are recorded with the operator used for
1282 * the modification. as we're going backwards, we do the
1283 * opposite of each operation here.
1284 */
1285 switch (tm->op) {
1286 case MOD_LOG_KEY_REMOVE_WHILE_FREEING:
1287 BUG_ON(tm->slot < n);
1288 /* Fallthrough */
1289 case MOD_LOG_KEY_REMOVE_WHILE_MOVING:
1290 case MOD_LOG_KEY_REMOVE:
1291 btrfs_set_node_key(eb, &tm->key, tm->slot);
1292 btrfs_set_node_blockptr(eb, tm->slot, tm->blockptr);
1293 btrfs_set_node_ptr_generation(eb, tm->slot,
1294 tm->generation);
1295 n++;
1296 break;
1297 case MOD_LOG_KEY_REPLACE:
1298 BUG_ON(tm->slot >= n);
1299 btrfs_set_node_key(eb, &tm->key, tm->slot);
1300 btrfs_set_node_blockptr(eb, tm->slot, tm->blockptr);
1301 btrfs_set_node_ptr_generation(eb, tm->slot,
1302 tm->generation);
1303 break;
1304 case MOD_LOG_KEY_ADD:
1305 /* if a move operation is needed it's in the log */
1306 n--;
1307 break;
1308 case MOD_LOG_MOVE_KEYS:
1309 o_dst = btrfs_node_key_ptr_offset(tm->slot);
1310 o_src = btrfs_node_key_ptr_offset(tm->move.dst_slot);
1311 memmove_extent_buffer(eb, o_dst, o_src,
1312 tm->move.nr_items * p_size);
1313 break;
1314 case MOD_LOG_ROOT_REPLACE:
1315 /*
1316 * this operation is special. for roots, this must be
1317 * handled explicitly before rewinding.
1318 * for non-roots, this operation may exist if the node
1319 * was a root: root A -> child B; then A gets empty and
1320 * B is promoted to the new root. in the mod log, we'll
1321 * have a root-replace operation for B, a tree block
1322 * that is no root. we simply ignore that operation.
1323 */
1324 break;
1325 }
1326 next = rb_next(&tm->node);
1327 if (!next)
1328 break;
1329 tm = container_of(next, struct tree_mod_elem, node);
1330 if (tm->index != first_tm->index)
1331 break;
1332 }
1333 tree_mod_log_read_unlock(fs_info);
1334 btrfs_set_header_nritems(eb, n);
1335}
1336
1337/*
1338 * Called with eb read locked. If the buffer cannot be rewinded, the same buffer
1339 * is returned. If rewind operations happen, a fresh buffer is returned. The
1340 * returned buffer is always read-locked. If the returned buffer is not the
1341 * input buffer, the lock on the input buffer is released and the input buffer
1342 * is freed (its refcount is decremented).
1343 */
1344static struct extent_buffer *
1345tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct btrfs_path *path,
1346 struct extent_buffer *eb, u64 time_seq)
1347{
1348 struct extent_buffer *eb_rewin;
1349 struct tree_mod_elem *tm;
1350
1351 if (!time_seq)
1352 return eb;
1353
1354 if (btrfs_header_level(eb) == 0)
1355 return eb;
1356
1357 tm = tree_mod_log_search(fs_info, eb->start, time_seq);
1358 if (!tm)
1359 return eb;
1360
1361 btrfs_set_path_blocking(path);
1362 btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK);
1363
1364 if (tm->op == MOD_LOG_KEY_REMOVE_WHILE_FREEING) {
1365 BUG_ON(tm->slot != 0);
1366 eb_rewin = alloc_dummy_extent_buffer(eb->start,
1367 fs_info->tree_root->nodesize);
1368 if (!eb_rewin) {
1369 btrfs_tree_read_unlock_blocking(eb);
1370 free_extent_buffer(eb);
1371 return NULL;
1372 }
1373 btrfs_set_header_bytenr(eb_rewin, eb->start);
1374 btrfs_set_header_backref_rev(eb_rewin,
1375 btrfs_header_backref_rev(eb));
1376 btrfs_set_header_owner(eb_rewin, btrfs_header_owner(eb));
1377 btrfs_set_header_level(eb_rewin, btrfs_header_level(eb));
1378 } else {
1379 eb_rewin = btrfs_clone_extent_buffer(eb);
1380 if (!eb_rewin) {
1381 btrfs_tree_read_unlock_blocking(eb);
1382 free_extent_buffer(eb);
1383 return NULL;
1384 }
1385 }
1386
1387 btrfs_clear_path_blocking(path, NULL, BTRFS_READ_LOCK);
1388 btrfs_tree_read_unlock_blocking(eb);
1389 free_extent_buffer(eb);
1390
1391 extent_buffer_get(eb_rewin);
1392 btrfs_tree_read_lock(eb_rewin);
1393 __tree_mod_log_rewind(fs_info, eb_rewin, time_seq, tm);
1394 WARN_ON(btrfs_header_nritems(eb_rewin) >
1395 BTRFS_NODEPTRS_PER_BLOCK(fs_info->tree_root));
1396
1397 return eb_rewin;
1398}
1399
1400/*
1401 * get_old_root() rewinds the state of @root's root node to the given @time_seq
1402 * value. If there are no changes, the current root->root_node is returned. If
1403 * anything changed in between, there's a fresh buffer allocated on which the
1404 * rewind operations are done. In any case, the returned buffer is read locked.
1405 * Returns NULL on error (with no locks held).
1406 */
1407static inline struct extent_buffer *
1408get_old_root(struct btrfs_root *root, u64 time_seq)
1409{
1410 struct tree_mod_elem *tm;
1411 struct extent_buffer *eb = NULL;
1412 struct extent_buffer *eb_root;
1413 struct extent_buffer *old;
1414 struct tree_mod_root *old_root = NULL;
1415 u64 old_generation = 0;
1416 u64 logical;
1417
1418 eb_root = btrfs_read_lock_root_node(root);
1419 tm = __tree_mod_log_oldest_root(root->fs_info, eb_root, time_seq);
1420 if (!tm)
1421 return eb_root;
1422
1423 if (tm->op == MOD_LOG_ROOT_REPLACE) {
1424 old_root = &tm->old_root;
1425 old_generation = tm->generation;
1426 logical = old_root->logical;
1427 } else {
1428 logical = eb_root->start;
1429 }
1430
1431 tm = tree_mod_log_search(root->fs_info, logical, time_seq);
1432 if (old_root && tm && tm->op != MOD_LOG_KEY_REMOVE_WHILE_FREEING) {
1433 btrfs_tree_read_unlock(eb_root);
1434 free_extent_buffer(eb_root);
1435 old = read_tree_block(root, logical, 0);
1436 if (WARN_ON(!old || !extent_buffer_uptodate(old))) {
1437 free_extent_buffer(old);
1438 btrfs_warn(root->fs_info,
1439 "failed to read tree block %llu from get_old_root", logical);
1440 } else {
1441 eb = btrfs_clone_extent_buffer(old);
1442 free_extent_buffer(old);
1443 }
1444 } else if (old_root) {
1445 btrfs_tree_read_unlock(eb_root);
1446 free_extent_buffer(eb_root);
1447 eb = alloc_dummy_extent_buffer(logical, root->nodesize);
1448 } else {
1449 btrfs_set_lock_blocking_rw(eb_root, BTRFS_READ_LOCK);
1450 eb = btrfs_clone_extent_buffer(eb_root);
1451 btrfs_tree_read_unlock_blocking(eb_root);
1452 free_extent_buffer(eb_root);
1453 }
1454
1455 if (!eb)
1456 return NULL;
1457 extent_buffer_get(eb);
1458 btrfs_tree_read_lock(eb);
1459 if (old_root) {
1460 btrfs_set_header_bytenr(eb, eb->start);
1461 btrfs_set_header_backref_rev(eb, BTRFS_MIXED_BACKREF_REV);
1462 btrfs_set_header_owner(eb, btrfs_header_owner(eb_root));
1463 btrfs_set_header_level(eb, old_root->level);
1464 btrfs_set_header_generation(eb, old_generation);
1465 }
1466 if (tm)
1467 __tree_mod_log_rewind(root->fs_info, eb, time_seq, tm);
1468 else
1469 WARN_ON(btrfs_header_level(eb) != 0);
1470 WARN_ON(btrfs_header_nritems(eb) > BTRFS_NODEPTRS_PER_BLOCK(root));
1471
1472 return eb;
1473}
1474
1475int btrfs_old_root_level(struct btrfs_root *root, u64 time_seq)
1476{
1477 struct tree_mod_elem *tm;
1478 int level;
1479 struct extent_buffer *eb_root = btrfs_root_node(root);
1480
1481 tm = __tree_mod_log_oldest_root(root->fs_info, eb_root, time_seq);
1482 if (tm && tm->op == MOD_LOG_ROOT_REPLACE) {
1483 level = tm->old_root.level;
1484 } else {
1485 level = btrfs_header_level(eb_root);
1486 }
1487 free_extent_buffer(eb_root);
1488
1489 return level;
1490}
1491
1492static inline int should_cow_block(struct btrfs_trans_handle *trans,
1493 struct btrfs_root *root,
1494 struct extent_buffer *buf)
1495{
1496 if (btrfs_test_is_dummy_root(root))
1497 return 0;
1498
1499 /* ensure we can see the force_cow */
1500 smp_rmb();
1501
1502 /*
1503 * We do not need to cow a block if
1504 * 1) this block is not created or changed in this transaction;
1505 * 2) this block does not belong to TREE_RELOC tree;
1506 * 3) the root is not forced COW.
1507 *
1508 * What is forced COW:
1509 * when we create snapshot during commiting the transaction,
1510 * after we've finished coping src root, we must COW the shared
1511 * block to ensure the metadata consistency.
1512 */
1513 if (btrfs_header_generation(buf) == trans->transid &&
1514 !btrfs_header_flag(buf, BTRFS_HEADER_FLAG_WRITTEN) &&
1515 !(root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID &&
1516 btrfs_header_flag(buf, BTRFS_HEADER_FLAG_RELOC)) &&
1517 !test_bit(BTRFS_ROOT_FORCE_COW, &root->state))
1518 return 0;
1519 return 1;
1520}
1521
1522/*
1523 * cows a single block, see __btrfs_cow_block for the real work.
1524 * This version of it has extra checks so that a block isn't cow'd more than
1525 * once per transaction, as long as it hasn't been written yet
1526 */
1527noinline int btrfs_cow_block(struct btrfs_trans_handle *trans,
1528 struct btrfs_root *root, struct extent_buffer *buf,
1529 struct extent_buffer *parent, int parent_slot,
1530 struct extent_buffer **cow_ret)
1531{
1532 u64 search_start;
1533 int ret;
1534
1535 if (trans->transaction != root->fs_info->running_transaction)
1536 WARN(1, KERN_CRIT "trans %llu running %llu\n",
1537 trans->transid,
1538 root->fs_info->running_transaction->transid);
1539
1540 if (trans->transid != root->fs_info->generation)
1541 WARN(1, KERN_CRIT "trans %llu running %llu\n",
1542 trans->transid, root->fs_info->generation);
1543
1544 if (!should_cow_block(trans, root, buf)) {
1545 *cow_ret = buf;
1546 return 0;
1547 }
1548
1549 search_start = buf->start & ~((u64)(1024 * 1024 * 1024) - 1);
1550
1551 if (parent)
1552 btrfs_set_lock_blocking(parent);
1553 btrfs_set_lock_blocking(buf);
1554
1555 ret = __btrfs_cow_block(trans, root, buf, parent,
1556 parent_slot, cow_ret, search_start, 0);
1557
1558 trace_btrfs_cow_block(root, buf, *cow_ret);
1559
1560 return ret;
1561}
1562
1563/*
1564 * helper function for defrag to decide if two blocks pointed to by a
1565 * node are actually close by
1566 */
1567static int close_blocks(u64 blocknr, u64 other, u32 blocksize)
1568{
1569 if (blocknr < other && other - (blocknr + blocksize) < 32768)
1570 return 1;
1571 if (blocknr > other && blocknr - (other + blocksize) < 32768)
1572 return 1;
1573 return 0;
1574}
1575
1576/*
1577 * compare two keys in a memcmp fashion
1578 */
1579static int comp_keys(struct btrfs_disk_key *disk, struct btrfs_key *k2)
1580{
1581 struct btrfs_key k1;
1582
1583 btrfs_disk_key_to_cpu(&k1, disk);
1584
1585 return btrfs_comp_cpu_keys(&k1, k2);
1586}
1587
1588/*
1589 * same as comp_keys only with two btrfs_key's
1590 */
1591int btrfs_comp_cpu_keys(struct btrfs_key *k1, struct btrfs_key *k2)
1592{
1593 if (k1->objectid > k2->objectid)
1594 return 1;
1595 if (k1->objectid < k2->objectid)
1596 return -1;
1597 if (k1->type > k2->type)
1598 return 1;
1599 if (k1->type < k2->type)
1600 return -1;
1601 if (k1->offset > k2->offset)
1602 return 1;
1603 if (k1->offset < k2->offset)
1604 return -1;
1605 return 0;
1606}
1607
1608/*
1609 * this is used by the defrag code to go through all the
1610 * leaves pointed to by a node and reallocate them so that
1611 * disk order is close to key order
1612 */
1613int btrfs_realloc_node(struct btrfs_trans_handle *trans,
1614 struct btrfs_root *root, struct extent_buffer *parent,
1615 int start_slot, u64 *last_ret,
1616 struct btrfs_key *progress)
1617{
1618 struct extent_buffer *cur;
1619 u64 blocknr;
1620 u64 gen;
1621 u64 search_start = *last_ret;
1622 u64 last_block = 0;
1623 u64 other;
1624 u32 parent_nritems;
1625 int end_slot;
1626 int i;
1627 int err = 0;
1628 int parent_level;
1629 int uptodate;
1630 u32 blocksize;
1631 int progress_passed = 0;
1632 struct btrfs_disk_key disk_key;
1633
1634 parent_level = btrfs_header_level(parent);
1635
1636 WARN_ON(trans->transaction != root->fs_info->running_transaction);
1637 WARN_ON(trans->transid != root->fs_info->generation);
1638
1639 parent_nritems = btrfs_header_nritems(parent);
1640 blocksize = root->nodesize;
1641 end_slot = parent_nritems;
1642
1643 if (parent_nritems == 1)
1644 return 0;
1645
1646 btrfs_set_lock_blocking(parent);
1647
1648 for (i = start_slot; i < end_slot; i++) {
1649 int close = 1;
1650
1651 btrfs_node_key(parent, &disk_key, i);
1652 if (!progress_passed && comp_keys(&disk_key, progress) < 0)
1653 continue;
1654
1655 progress_passed = 1;
1656 blocknr = btrfs_node_blockptr(parent, i);
1657 gen = btrfs_node_ptr_generation(parent, i);
1658 if (last_block == 0)
1659 last_block = blocknr;
1660
1661 if (i > 0) {
1662 other = btrfs_node_blockptr(parent, i - 1);
1663 close = close_blocks(blocknr, other, blocksize);
1664 }
1665 if (!close && i < end_slot - 2) {
1666 other = btrfs_node_blockptr(parent, i + 1);
1667 close = close_blocks(blocknr, other, blocksize);
1668 }
1669 if (close) {
1670 last_block = blocknr;
1671 continue;
1672 }
1673
1674 cur = btrfs_find_tree_block(root, blocknr);
1675 if (cur)
1676 uptodate = btrfs_buffer_uptodate(cur, gen, 0);
1677 else
1678 uptodate = 0;
1679 if (!cur || !uptodate) {
1680 if (!cur) {
1681 cur = read_tree_block(root, blocknr, gen);
1682 if (!cur || !extent_buffer_uptodate(cur)) {
1683 free_extent_buffer(cur);
1684 return -EIO;
1685 }
1686 } else if (!uptodate) {
1687 err = btrfs_read_buffer(cur, gen);
1688 if (err) {
1689 free_extent_buffer(cur);
1690 return err;
1691 }
1692 }
1693 }
1694 if (search_start == 0)
1695 search_start = last_block;
1696
1697 btrfs_tree_lock(cur);
1698 btrfs_set_lock_blocking(cur);
1699 err = __btrfs_cow_block(trans, root, cur, parent, i,
1700 &cur, search_start,
1701 min(16 * blocksize,
1702 (end_slot - i) * blocksize));
1703 if (err) {
1704 btrfs_tree_unlock(cur);
1705 free_extent_buffer(cur);
1706 break;
1707 }
1708 search_start = cur->start;
1709 last_block = cur->start;
1710 *last_ret = search_start;
1711 btrfs_tree_unlock(cur);
1712 free_extent_buffer(cur);
1713 }
1714 return err;
1715}
1716
1717/*
1718 * The leaf data grows from end-to-front in the node.
1719 * this returns the address of the start of the last item,
1720 * which is the stop of the leaf data stack
1721 */
1722static inline unsigned int leaf_data_end(struct btrfs_root *root,
1723 struct extent_buffer *leaf)
1724{
1725 u32 nr = btrfs_header_nritems(leaf);
1726 if (nr == 0)
1727 return BTRFS_LEAF_DATA_SIZE(root);
1728 return btrfs_item_offset_nr(leaf, nr - 1);
1729}
1730
1731
1732/*
1733 * search for key in the extent_buffer. The items start at offset p,
1734 * and they are item_size apart. There are 'max' items in p.
1735 *
1736 * the slot in the array is returned via slot, and it points to
1737 * the place where you would insert key if it is not found in
1738 * the array.
1739 *
1740 * slot may point to max if the key is bigger than all of the keys
1741 */
1742static noinline int generic_bin_search(struct extent_buffer *eb,
1743 unsigned long p,
1744 int item_size, struct btrfs_key *key,
1745 int max, int *slot)
1746{
1747 int low = 0;
1748 int high = max;
1749 int mid;
1750 int ret;
1751 struct btrfs_disk_key *tmp = NULL;
1752 struct btrfs_disk_key unaligned;
1753 unsigned long offset;
1754 char *kaddr = NULL;
1755 unsigned long map_start = 0;
1756 unsigned long map_len = 0;
1757 int err;
1758
1759 while (low < high) {
1760 mid = (low + high) / 2;
1761 offset = p + mid * item_size;
1762
1763 if (!kaddr || offset < map_start ||
1764 (offset + sizeof(struct btrfs_disk_key)) >
1765 map_start + map_len) {
1766
1767 err = map_private_extent_buffer(eb, offset,
1768 sizeof(struct btrfs_disk_key),
1769 &kaddr, &map_start, &map_len);
1770
1771 if (!err) {
1772 tmp = (struct btrfs_disk_key *)(kaddr + offset -
1773 map_start);
1774 } else {
1775 read_extent_buffer(eb, &unaligned,
1776 offset, sizeof(unaligned));
1777 tmp = &unaligned;
1778 }
1779
1780 } else {
1781 tmp = (struct btrfs_disk_key *)(kaddr + offset -
1782 map_start);
1783 }
1784 ret = comp_keys(tmp, key);
1785
1786 if (ret < 0)
1787 low = mid + 1;
1788 else if (ret > 0)
1789 high = mid;
1790 else {
1791 *slot = mid;
1792 return 0;
1793 }
1794 }
1795 *slot = low;
1796 return 1;
1797}
1798
1799/*
1800 * simple bin_search frontend that does the right thing for
1801 * leaves vs nodes
1802 */
1803static int bin_search(struct extent_buffer *eb, struct btrfs_key *key,
1804 int level, int *slot)
1805{
1806 if (level == 0)
1807 return generic_bin_search(eb,
1808 offsetof(struct btrfs_leaf, items),
1809 sizeof(struct btrfs_item),
1810 key, btrfs_header_nritems(eb),
1811 slot);
1812 else
1813 return generic_bin_search(eb,
1814 offsetof(struct btrfs_node, ptrs),
1815 sizeof(struct btrfs_key_ptr),
1816 key, btrfs_header_nritems(eb),
1817 slot);
1818}
1819
1820int btrfs_bin_search(struct extent_buffer *eb, struct btrfs_key *key,
1821 int level, int *slot)
1822{
1823 return bin_search(eb, key, level, slot);
1824}
1825
1826static void root_add_used(struct btrfs_root *root, u32 size)
1827{
1828 spin_lock(&root->accounting_lock);
1829 btrfs_set_root_used(&root->root_item,
1830 btrfs_root_used(&root->root_item) + size);
1831 spin_unlock(&root->accounting_lock);
1832}
1833
1834static void root_sub_used(struct btrfs_root *root, u32 size)
1835{
1836 spin_lock(&root->accounting_lock);
1837 btrfs_set_root_used(&root->root_item,
1838 btrfs_root_used(&root->root_item) - size);
1839 spin_unlock(&root->accounting_lock);
1840}
1841
1842/* given a node and slot number, this reads the blocks it points to. The
1843 * extent buffer is returned with a reference taken (but unlocked).
1844 * NULL is returned on error.
1845 */
1846static noinline struct extent_buffer *read_node_slot(struct btrfs_root *root,
1847 struct extent_buffer *parent, int slot)
1848{
1849 int level = btrfs_header_level(parent);
1850 struct extent_buffer *eb;
1851
1852 if (slot < 0)
1853 return NULL;
1854 if (slot >= btrfs_header_nritems(parent))
1855 return NULL;
1856
1857 BUG_ON(level == 0);
1858
1859 eb = read_tree_block(root, btrfs_node_blockptr(parent, slot),
1860 btrfs_node_ptr_generation(parent, slot));
1861 if (eb && !extent_buffer_uptodate(eb)) {
1862 free_extent_buffer(eb);
1863 eb = NULL;
1864 }
1865
1866 return eb;
1867}
1868
1869/*
1870 * node level balancing, used to make sure nodes are in proper order for
1871 * item deletion. We balance from the top down, so we have to make sure
1872 * that a deletion won't leave an node completely empty later on.
1873 */
1874static noinline int balance_level(struct btrfs_trans_handle *trans,
1875 struct btrfs_root *root,
1876 struct btrfs_path *path, int level)
1877{
1878 struct extent_buffer *right = NULL;
1879 struct extent_buffer *mid;
1880 struct extent_buffer *left = NULL;
1881 struct extent_buffer *parent = NULL;
1882 int ret = 0;
1883 int wret;
1884 int pslot;
1885 int orig_slot = path->slots[level];
1886 u64 orig_ptr;
1887
1888 if (level == 0)
1889 return 0;
1890
1891 mid = path->nodes[level];
1892
1893 WARN_ON(path->locks[level] != BTRFS_WRITE_LOCK &&
1894 path->locks[level] != BTRFS_WRITE_LOCK_BLOCKING);
1895 WARN_ON(btrfs_header_generation(mid) != trans->transid);
1896
1897 orig_ptr = btrfs_node_blockptr(mid, orig_slot);
1898
1899 if (level < BTRFS_MAX_LEVEL - 1) {
1900 parent = path->nodes[level + 1];
1901 pslot = path->slots[level + 1];
1902 }
1903
1904 /*
1905 * deal with the case where there is only one pointer in the root
1906 * by promoting the node below to a root
1907 */
1908 if (!parent) {
1909 struct extent_buffer *child;
1910
1911 if (btrfs_header_nritems(mid) != 1)
1912 return 0;
1913
1914 /* promote the child to a root */
1915 child = read_node_slot(root, mid, 0);
1916 if (!child) {
1917 ret = -EROFS;
1918 btrfs_std_error(root->fs_info, ret);
1919 goto enospc;
1920 }
1921
1922 btrfs_tree_lock(child);
1923 btrfs_set_lock_blocking(child);
1924 ret = btrfs_cow_block(trans, root, child, mid, 0, &child);
1925 if (ret) {
1926 btrfs_tree_unlock(child);
1927 free_extent_buffer(child);
1928 goto enospc;
1929 }
1930
1931 tree_mod_log_set_root_pointer(root, child, 1);
1932 rcu_assign_pointer(root->node, child);
1933
1934 add_root_to_dirty_list(root);
1935 btrfs_tree_unlock(child);
1936
1937 path->locks[level] = 0;
1938 path->nodes[level] = NULL;
1939 clean_tree_block(trans, root, mid);
1940 btrfs_tree_unlock(mid);
1941 /* once for the path */
1942 free_extent_buffer(mid);
1943
1944 root_sub_used(root, mid->len);
1945 btrfs_free_tree_block(trans, root, mid, 0, 1);
1946 /* once for the root ptr */
1947 free_extent_buffer_stale(mid);
1948 return 0;
1949 }
1950 if (btrfs_header_nritems(mid) >
1951 BTRFS_NODEPTRS_PER_BLOCK(root) / 4)
1952 return 0;
1953
1954 left = read_node_slot(root, parent, pslot - 1);
1955 if (left) {
1956 btrfs_tree_lock(left);
1957 btrfs_set_lock_blocking(left);
1958 wret = btrfs_cow_block(trans, root, left,
1959 parent, pslot - 1, &left);
1960 if (wret) {
1961 ret = wret;
1962 goto enospc;
1963 }
1964 }
1965 right = read_node_slot(root, parent, pslot + 1);
1966 if (right) {
1967 btrfs_tree_lock(right);
1968 btrfs_set_lock_blocking(right);
1969 wret = btrfs_cow_block(trans, root, right,
1970 parent, pslot + 1, &right);
1971 if (wret) {
1972 ret = wret;
1973 goto enospc;
1974 }
1975 }
1976
1977 /* first, try to make some room in the middle buffer */
1978 if (left) {
1979 orig_slot += btrfs_header_nritems(left);
1980 wret = push_node_left(trans, root, left, mid, 1);
1981 if (wret < 0)
1982 ret = wret;
1983 }
1984
1985 /*
1986 * then try to empty the right most buffer into the middle
1987 */
1988 if (right) {
1989 wret = push_node_left(trans, root, mid, right, 1);
1990 if (wret < 0 && wret != -ENOSPC)
1991 ret = wret;
1992 if (btrfs_header_nritems(right) == 0) {
1993 clean_tree_block(trans, root, right);
1994 btrfs_tree_unlock(right);
1995 del_ptr(root, path, level + 1, pslot + 1);
1996 root_sub_used(root, right->len);
1997 btrfs_free_tree_block(trans, root, right, 0, 1);
1998 free_extent_buffer_stale(right);
1999 right = NULL;
2000 } else {
2001 struct btrfs_disk_key right_key;
2002 btrfs_node_key(right, &right_key, 0);
2003 tree_mod_log_set_node_key(root->fs_info, parent,
2004 pslot + 1, 0);
2005 btrfs_set_node_key(parent, &right_key, pslot + 1);
2006 btrfs_mark_buffer_dirty(parent);
2007 }
2008 }
2009 if (btrfs_header_nritems(mid) == 1) {
2010 /*
2011 * we're not allowed to leave a node with one item in the
2012 * tree during a delete. A deletion from lower in the tree
2013 * could try to delete the only pointer in this node.
2014 * So, pull some keys from the left.
2015 * There has to be a left pointer at this point because
2016 * otherwise we would have pulled some pointers from the
2017 * right
2018 */
2019 if (!left) {
2020 ret = -EROFS;
2021 btrfs_std_error(root->fs_info, ret);
2022 goto enospc;
2023 }
2024 wret = balance_node_right(trans, root, mid, left);
2025 if (wret < 0) {
2026 ret = wret;
2027 goto enospc;
2028 }
2029 if (wret == 1) {
2030 wret = push_node_left(trans, root, left, mid, 1);
2031 if (wret < 0)
2032 ret = wret;
2033 }
2034 BUG_ON(wret == 1);
2035 }
2036 if (btrfs_header_nritems(mid) == 0) {
2037 clean_tree_block(trans, root, mid);
2038 btrfs_tree_unlock(mid);
2039 del_ptr(root, path, level + 1, pslot);
2040 root_sub_used(root, mid->len);
2041 btrfs_free_tree_block(trans, root, mid, 0, 1);
2042 free_extent_buffer_stale(mid);
2043 mid = NULL;
2044 } else {
2045 /* update the parent key to reflect our changes */
2046 struct btrfs_disk_key mid_key;
2047 btrfs_node_key(mid, &mid_key, 0);
2048 tree_mod_log_set_node_key(root->fs_info, parent,
2049 pslot, 0);
2050 btrfs_set_node_key(parent, &mid_key, pslot);
2051 btrfs_mark_buffer_dirty(parent);
2052 }
2053
2054 /* update the path */
2055 if (left) {
2056 if (btrfs_header_nritems(left) > orig_slot) {
2057 extent_buffer_get(left);
2058 /* left was locked after cow */
2059 path->nodes[level] = left;
2060 path->slots[level + 1] -= 1;
2061 path->slots[level] = orig_slot;
2062 if (mid) {
2063 btrfs_tree_unlock(mid);
2064 free_extent_buffer(mid);
2065 }
2066 } else {
2067 orig_slot -= btrfs_header_nritems(left);
2068 path->slots[level] = orig_slot;
2069 }
2070 }
2071 /* double check we haven't messed things up */
2072 if (orig_ptr !=
2073 btrfs_node_blockptr(path->nodes[level], path->slots[level]))
2074 BUG();
2075enospc:
2076 if (right) {
2077 btrfs_tree_unlock(right);
2078 free_extent_buffer(right);
2079 }
2080 if (left) {
2081 if (path->nodes[level] != left)
2082 btrfs_tree_unlock(left);
2083 free_extent_buffer(left);
2084 }
2085 return ret;
2086}
2087
2088/* Node balancing for insertion. Here we only split or push nodes around
2089 * when they are completely full. This is also done top down, so we
2090 * have to be pessimistic.
2091 */
2092static noinline int push_nodes_for_insert(struct btrfs_trans_handle *trans,
2093 struct btrfs_root *root,
2094 struct btrfs_path *path, int level)
2095{
2096 struct extent_buffer *right = NULL;
2097 struct extent_buffer *mid;
2098 struct extent_buffer *left = NULL;
2099 struct extent_buffer *parent = NULL;
2100 int ret = 0;
2101 int wret;
2102 int pslot;
2103 int orig_slot = path->slots[level];
2104
2105 if (level == 0)
2106 return 1;
2107
2108 mid = path->nodes[level];
2109 WARN_ON(btrfs_header_generation(mid) != trans->transid);
2110
2111 if (level < BTRFS_MAX_LEVEL - 1) {
2112 parent = path->nodes[level + 1];
2113 pslot = path->slots[level + 1];
2114 }
2115
2116 if (!parent)
2117 return 1;
2118
2119 left = read_node_slot(root, parent, pslot - 1);
2120
2121 /* first, try to make some room in the middle buffer */
2122 if (left) {
2123 u32 left_nr;
2124
2125 btrfs_tree_lock(left);
2126 btrfs_set_lock_blocking(left);
2127
2128 left_nr = btrfs_header_nritems(left);
2129 if (left_nr >= BTRFS_NODEPTRS_PER_BLOCK(root) - 1) {
2130 wret = 1;
2131 } else {
2132 ret = btrfs_cow_block(trans, root, left, parent,
2133 pslot - 1, &left);
2134 if (ret)
2135 wret = 1;
2136 else {
2137 wret = push_node_left(trans, root,
2138 left, mid, 0);
2139 }
2140 }
2141 if (wret < 0)
2142 ret = wret;
2143 if (wret == 0) {
2144 struct btrfs_disk_key disk_key;
2145 orig_slot += left_nr;
2146 btrfs_node_key(mid, &disk_key, 0);
2147 tree_mod_log_set_node_key(root->fs_info, parent,
2148 pslot, 0);
2149 btrfs_set_node_key(parent, &disk_key, pslot);
2150 btrfs_mark_buffer_dirty(parent);
2151 if (btrfs_header_nritems(left) > orig_slot) {
2152 path->nodes[level] = left;
2153 path->slots[level + 1] -= 1;
2154 path->slots[level] = orig_slot;
2155 btrfs_tree_unlock(mid);
2156 free_extent_buffer(mid);
2157 } else {
2158 orig_slot -=
2159 btrfs_header_nritems(left);
2160 path->slots[level] = orig_slot;
2161 btrfs_tree_unlock(left);
2162 free_extent_buffer(left);
2163 }
2164 return 0;
2165 }
2166 btrfs_tree_unlock(left);
2167 free_extent_buffer(left);
2168 }
2169 right = read_node_slot(root, parent, pslot + 1);
2170
2171 /*
2172 * then try to empty the right most buffer into the middle
2173 */
2174 if (right) {
2175 u32 right_nr;
2176
2177 btrfs_tree_lock(right);
2178 btrfs_set_lock_blocking(right);
2179
2180 right_nr = btrfs_header_nritems(right);
2181 if (right_nr >= BTRFS_NODEPTRS_PER_BLOCK(root) - 1) {
2182 wret = 1;
2183 } else {
2184 ret = btrfs_cow_block(trans, root, right,
2185 parent, pslot + 1,
2186 &right);
2187 if (ret)
2188 wret = 1;
2189 else {
2190 wret = balance_node_right(trans, root,
2191 right, mid);
2192 }
2193 }
2194 if (wret < 0)
2195 ret = wret;
2196 if (wret == 0) {
2197 struct btrfs_disk_key disk_key;
2198
2199 btrfs_node_key(right, &disk_key, 0);
2200 tree_mod_log_set_node_key(root->fs_info, parent,
2201 pslot + 1, 0);
2202 btrfs_set_node_key(parent, &disk_key, pslot + 1);
2203 btrfs_mark_buffer_dirty(parent);
2204
2205 if (btrfs_header_nritems(mid) <= orig_slot) {
2206 path->nodes[level] = right;
2207 path->slots[level + 1] += 1;
2208 path->slots[level] = orig_slot -
2209 btrfs_header_nritems(mid);
2210 btrfs_tree_unlock(mid);
2211 free_extent_buffer(mid);
2212 } else {
2213 btrfs_tree_unlock(right);
2214 free_extent_buffer(right);
2215 }
2216 return 0;
2217 }
2218 btrfs_tree_unlock(right);
2219 free_extent_buffer(right);
2220 }
2221 return 1;
2222}
2223
2224/*
2225 * readahead one full node of leaves, finding things that are close
2226 * to the block in 'slot', and triggering ra on them.
2227 */
2228static void reada_for_search(struct btrfs_root *root,
2229 struct btrfs_path *path,
2230 int level, int slot, u64 objectid)
2231{
2232 struct extent_buffer *node;
2233 struct btrfs_disk_key disk_key;
2234 u32 nritems;
2235 u64 search;
2236 u64 target;
2237 u64 nread = 0;
2238 u64 gen;
2239 int direction = path->reada;
2240 struct extent_buffer *eb;
2241 u32 nr;
2242 u32 blocksize;
2243 u32 nscan = 0;
2244
2245 if (level != 1)
2246 return;
2247
2248 if (!path->nodes[level])
2249 return;
2250
2251 node = path->nodes[level];
2252
2253 search = btrfs_node_blockptr(node, slot);
2254 blocksize = root->nodesize;
2255 eb = btrfs_find_tree_block(root, search);
2256 if (eb) {
2257 free_extent_buffer(eb);
2258 return;
2259 }
2260
2261 target = search;
2262
2263 nritems = btrfs_header_nritems(node);
2264 nr = slot;
2265
2266 while (1) {
2267 if (direction < 0) {
2268 if (nr == 0)
2269 break;
2270 nr--;
2271 } else if (direction > 0) {
2272 nr++;
2273 if (nr >= nritems)
2274 break;
2275 }
2276 if (path->reada < 0 && objectid) {
2277 btrfs_node_key(node, &disk_key, nr);
2278 if (btrfs_disk_key_objectid(&disk_key) != objectid)
2279 break;
2280 }
2281 search = btrfs_node_blockptr(node, nr);
2282 if ((search <= target && target - search <= 65536) ||
2283 (search > target && search - target <= 65536)) {
2284 gen = btrfs_node_ptr_generation(node, nr);
2285 readahead_tree_block(root, search, blocksize);
2286 nread += blocksize;
2287 }
2288 nscan++;
2289 if ((nread > 65536 || nscan > 32))
2290 break;
2291 }
2292}
2293
2294static noinline void reada_for_balance(struct btrfs_root *root,
2295 struct btrfs_path *path, int level)
2296{
2297 int slot;
2298 int nritems;
2299 struct extent_buffer *parent;
2300 struct extent_buffer *eb;
2301 u64 gen;
2302 u64 block1 = 0;
2303 u64 block2 = 0;
2304 int blocksize;
2305
2306 parent = path->nodes[level + 1];
2307 if (!parent)
2308 return;
2309
2310 nritems = btrfs_header_nritems(parent);
2311 slot = path->slots[level + 1];
2312 blocksize = root->nodesize;
2313
2314 if (slot > 0) {
2315 block1 = btrfs_node_blockptr(parent, slot - 1);
2316 gen = btrfs_node_ptr_generation(parent, slot - 1);
2317 eb = btrfs_find_tree_block(root, block1);
2318 /*
2319 * if we get -eagain from btrfs_buffer_uptodate, we
2320 * don't want to return eagain here. That will loop
2321 * forever
2322 */
2323 if (eb && btrfs_buffer_uptodate(eb, gen, 1) != 0)
2324 block1 = 0;
2325 free_extent_buffer(eb);
2326 }
2327 if (slot + 1 < nritems) {
2328 block2 = btrfs_node_blockptr(parent, slot + 1);
2329 gen = btrfs_node_ptr_generation(parent, slot + 1);
2330 eb = btrfs_find_tree_block(root, block2);
2331 if (eb && btrfs_buffer_uptodate(eb, gen, 1) != 0)
2332 block2 = 0;
2333 free_extent_buffer(eb);
2334 }
2335
2336 if (block1)
2337 readahead_tree_block(root, block1, blocksize);
2338 if (block2)
2339 readahead_tree_block(root, block2, blocksize);
2340}
2341
2342
2343/*
2344 * when we walk down the tree, it is usually safe to unlock the higher layers
2345 * in the tree. The exceptions are when our path goes through slot 0, because
2346 * operations on the tree might require changing key pointers higher up in the
2347 * tree.
2348 *
2349 * callers might also have set path->keep_locks, which tells this code to keep
2350 * the lock if the path points to the last slot in the block. This is part of
2351 * walking through the tree, and selecting the next slot in the higher block.
2352 *
2353 * lowest_unlock sets the lowest level in the tree we're allowed to unlock. so
2354 * if lowest_unlock is 1, level 0 won't be unlocked
2355 */
2356static noinline void unlock_up(struct btrfs_path *path, int level,
2357 int lowest_unlock, int min_write_lock_level,
2358 int *write_lock_level)
2359{
2360 int i;
2361 int skip_level = level;
2362 int no_skips = 0;
2363 struct extent_buffer *t;
2364
2365 for (i = level; i < BTRFS_MAX_LEVEL; i++) {
2366 if (!path->nodes[i])
2367 break;
2368 if (!path->locks[i])
2369 break;
2370 if (!no_skips && path->slots[i] == 0) {
2371 skip_level = i + 1;
2372 continue;
2373 }
2374 if (!no_skips && path->keep_locks) {
2375 u32 nritems;
2376 t = path->nodes[i];
2377 nritems = btrfs_header_nritems(t);
2378 if (nritems < 1 || path->slots[i] >= nritems - 1) {
2379 skip_level = i + 1;
2380 continue;
2381 }
2382 }
2383 if (skip_level < i && i >= lowest_unlock)
2384 no_skips = 1;
2385
2386 t = path->nodes[i];
2387 if (i >= lowest_unlock && i > skip_level && path->locks[i]) {
2388 btrfs_tree_unlock_rw(t, path->locks[i]);
2389 path->locks[i] = 0;
2390 if (write_lock_level &&
2391 i > min_write_lock_level &&
2392 i <= *write_lock_level) {
2393 *write_lock_level = i - 1;
2394 }
2395 }
2396 }
2397}
2398
2399/*
2400 * This releases any locks held in the path starting at level and
2401 * going all the way up to the root.
2402 *
2403 * btrfs_search_slot will keep the lock held on higher nodes in a few
2404 * corner cases, such as COW of the block at slot zero in the node. This
2405 * ignores those rules, and it should only be called when there are no
2406 * more updates to be done higher up in the tree.
2407 */
2408noinline void btrfs_unlock_up_safe(struct btrfs_path *path, int level)
2409{
2410 int i;
2411
2412 if (path->keep_locks)
2413 return;
2414
2415 for (i = level; i < BTRFS_MAX_LEVEL; i++) {
2416 if (!path->nodes[i])
2417 continue;
2418 if (!path->locks[i])
2419 continue;
2420 btrfs_tree_unlock_rw(path->nodes[i], path->locks[i]);
2421 path->locks[i] = 0;
2422 }
2423}
2424
2425/*
2426 * helper function for btrfs_search_slot. The goal is to find a block
2427 * in cache without setting the path to blocking. If we find the block
2428 * we return zero and the path is unchanged.
2429 *
2430 * If we can't find the block, we set the path blocking and do some
2431 * reada. -EAGAIN is returned and the search must be repeated.
2432 */
2433static int
2434read_block_for_search(struct btrfs_trans_handle *trans,
2435 struct btrfs_root *root, struct btrfs_path *p,
2436 struct extent_buffer **eb_ret, int level, int slot,
2437 struct btrfs_key *key, u64 time_seq)
2438{
2439 u64 blocknr;
2440 u64 gen;
2441 struct extent_buffer *b = *eb_ret;
2442 struct extent_buffer *tmp;
2443 int ret;
2444
2445 blocknr = btrfs_node_blockptr(b, slot);
2446 gen = btrfs_node_ptr_generation(b, slot);
2447
2448 tmp = btrfs_find_tree_block(root, blocknr);
2449 if (tmp) {
2450 /* first we do an atomic uptodate check */
2451 if (btrfs_buffer_uptodate(tmp, gen, 1) > 0) {
2452 *eb_ret = tmp;
2453 return 0;
2454 }
2455
2456 /* the pages were up to date, but we failed
2457 * the generation number check. Do a full
2458 * read for the generation number that is correct.
2459 * We must do this without dropping locks so
2460 * we can trust our generation number
2461 */
2462 btrfs_set_path_blocking(p);
2463
2464 /* now we're allowed to do a blocking uptodate check */
2465 ret = btrfs_read_buffer(tmp, gen);
2466 if (!ret) {
2467 *eb_ret = tmp;
2468 return 0;
2469 }
2470 free_extent_buffer(tmp);
2471 btrfs_release_path(p);
2472 return -EIO;
2473 }
2474
2475 /*
2476 * reduce lock contention at high levels
2477 * of the btree by dropping locks before
2478 * we read. Don't release the lock on the current
2479 * level because we need to walk this node to figure
2480 * out which blocks to read.
2481 */
2482 btrfs_unlock_up_safe(p, level + 1);
2483 btrfs_set_path_blocking(p);
2484
2485 free_extent_buffer(tmp);
2486 if (p->reada)
2487 reada_for_search(root, p, level, slot, key->objectid);
2488
2489 btrfs_release_path(p);
2490
2491 ret = -EAGAIN;
2492 tmp = read_tree_block(root, blocknr, 0);
2493 if (tmp) {
2494 /*
2495 * If the read above didn't mark this buffer up to date,
2496 * it will never end up being up to date. Set ret to EIO now
2497 * and give up so that our caller doesn't loop forever
2498 * on our EAGAINs.
2499 */
2500 if (!btrfs_buffer_uptodate(tmp, 0, 0))
2501 ret = -EIO;
2502 free_extent_buffer(tmp);
2503 }
2504 return ret;
2505}
2506
2507/*
2508 * helper function for btrfs_search_slot. This does all of the checks
2509 * for node-level blocks and does any balancing required based on
2510 * the ins_len.
2511 *
2512 * If no extra work was required, zero is returned. If we had to
2513 * drop the path, -EAGAIN is returned and btrfs_search_slot must
2514 * start over
2515 */
2516static int
2517setup_nodes_for_search(struct btrfs_trans_handle *trans,
2518 struct btrfs_root *root, struct btrfs_path *p,
2519 struct extent_buffer *b, int level, int ins_len,
2520 int *write_lock_level)
2521{
2522 int ret;
2523 if ((p->search_for_split || ins_len > 0) && btrfs_header_nritems(b) >=
2524 BTRFS_NODEPTRS_PER_BLOCK(root) - 3) {
2525 int sret;
2526
2527 if (*write_lock_level < level + 1) {
2528 *write_lock_level = level + 1;
2529 btrfs_release_path(p);
2530 goto again;
2531 }
2532
2533 btrfs_set_path_blocking(p);
2534 reada_for_balance(root, p, level);
2535 sret = split_node(trans, root, p, level);
2536 btrfs_clear_path_blocking(p, NULL, 0);
2537
2538 BUG_ON(sret > 0);
2539 if (sret) {
2540 ret = sret;
2541 goto done;
2542 }
2543 b = p->nodes[level];
2544 } else if (ins_len < 0 && btrfs_header_nritems(b) <
2545 BTRFS_NODEPTRS_PER_BLOCK(root) / 2) {
2546 int sret;
2547
2548 if (*write_lock_level < level + 1) {
2549 *write_lock_level = level + 1;
2550 btrfs_release_path(p);
2551 goto again;
2552 }
2553
2554 btrfs_set_path_blocking(p);
2555 reada_for_balance(root, p, level);
2556 sret = balance_level(trans, root, p, level);
2557 btrfs_clear_path_blocking(p, NULL, 0);
2558
2559 if (sret) {
2560 ret = sret;
2561 goto done;
2562 }
2563 b = p->nodes[level];
2564 if (!b) {
2565 btrfs_release_path(p);
2566 goto again;
2567 }
2568 BUG_ON(btrfs_header_nritems(b) == 1);
2569 }
2570 return 0;
2571
2572again:
2573 ret = -EAGAIN;
2574done:
2575 return ret;
2576}
2577
2578static void key_search_validate(struct extent_buffer *b,
2579 struct btrfs_key *key,
2580 int level)
2581{
2582#ifdef CONFIG_BTRFS_ASSERT
2583 struct btrfs_disk_key disk_key;
2584
2585 btrfs_cpu_key_to_disk(&disk_key, key);
2586
2587 if (level == 0)
2588 ASSERT(!memcmp_extent_buffer(b, &disk_key,
2589 offsetof(struct btrfs_leaf, items[0].key),
2590 sizeof(disk_key)));
2591 else
2592 ASSERT(!memcmp_extent_buffer(b, &disk_key,
2593 offsetof(struct btrfs_node, ptrs[0].key),
2594 sizeof(disk_key)));
2595#endif
2596}
2597
2598static int key_search(struct extent_buffer *b, struct btrfs_key *key,
2599 int level, int *prev_cmp, int *slot)
2600{
2601 if (*prev_cmp != 0) {
2602 *prev_cmp = bin_search(b, key, level, slot);
2603 return *prev_cmp;
2604 }
2605
2606 key_search_validate(b, key, level);
2607 *slot = 0;
2608
2609 return 0;
2610}
2611
2612int btrfs_find_item(struct btrfs_root *fs_root, struct btrfs_path *found_path,
2613 u64 iobjectid, u64 ioff, u8 key_type,
2614 struct btrfs_key *found_key)
2615{
2616 int ret;
2617 struct btrfs_key key;
2618 struct extent_buffer *eb;
2619 struct btrfs_path *path;
2620
2621 key.type = key_type;
2622 key.objectid = iobjectid;
2623 key.offset = ioff;
2624
2625 if (found_path == NULL) {
2626 path = btrfs_alloc_path();
2627 if (!path)
2628 return -ENOMEM;
2629 } else
2630 path = found_path;
2631
2632 ret = btrfs_search_slot(NULL, fs_root, &key, path, 0, 0);
2633 if ((ret < 0) || (found_key == NULL)) {
2634 if (path != found_path)
2635 btrfs_free_path(path);
2636 return ret;
2637 }
2638
2639 eb = path->nodes[0];
2640 if (ret && path->slots[0] >= btrfs_header_nritems(eb)) {
2641 ret = btrfs_next_leaf(fs_root, path);
2642 if (ret)
2643 return ret;
2644 eb = path->nodes[0];
2645 }
2646
2647 btrfs_item_key_to_cpu(eb, found_key, path->slots[0]);
2648 if (found_key->type != key.type ||
2649 found_key->objectid != key.objectid)
2650 return 1;
2651
2652 return 0;
2653}
2654
2655/*
2656 * look for key in the tree. path is filled in with nodes along the way
2657 * if key is found, we return zero and you can find the item in the leaf
2658 * level of the path (level 0)
2659 *
2660 * If the key isn't found, the path points to the slot where it should
2661 * be inserted, and 1 is returned. If there are other errors during the
2662 * search a negative error number is returned.
2663 *
2664 * if ins_len > 0, nodes and leaves will be split as we walk down the
2665 * tree. if ins_len < 0, nodes will be merged as we walk down the tree (if
2666 * possible)
2667 */
2668int btrfs_search_slot(struct btrfs_trans_handle *trans, struct btrfs_root
2669 *root, struct btrfs_key *key, struct btrfs_path *p, int
2670 ins_len, int cow)
2671{
2672 struct extent_buffer *b;
2673 int slot;
2674 int ret;
2675 int err;
2676 int level;
2677 int lowest_unlock = 1;
2678 int root_lock;
2679 /* everything at write_lock_level or lower must be write locked */
2680 int write_lock_level = 0;
2681 u8 lowest_level = 0;
2682 int min_write_lock_level;
2683 int prev_cmp;
2684
2685 lowest_level = p->lowest_level;
2686 WARN_ON(lowest_level && ins_len > 0);
2687 WARN_ON(p->nodes[0] != NULL);
2688 BUG_ON(!cow && ins_len);
2689
2690 if (ins_len < 0) {
2691 lowest_unlock = 2;
2692
2693 /* when we are removing items, we might have to go up to level
2694 * two as we update tree pointers Make sure we keep write
2695 * for those levels as well
2696 */
2697 write_lock_level = 2;
2698 } else if (ins_len > 0) {
2699 /*
2700 * for inserting items, make sure we have a write lock on
2701 * level 1 so we can update keys
2702 */
2703 write_lock_level = 1;
2704 }
2705
2706 if (!cow)
2707 write_lock_level = -1;
2708
2709 if (cow && (p->keep_locks || p->lowest_level))
2710 write_lock_level = BTRFS_MAX_LEVEL;
2711
2712 min_write_lock_level = write_lock_level;
2713
2714again:
2715 prev_cmp = -1;
2716 /*
2717 * we try very hard to do read locks on the root
2718 */
2719 root_lock = BTRFS_READ_LOCK;
2720 level = 0;
2721 if (p->search_commit_root) {
2722 /*
2723 * the commit roots are read only
2724 * so we always do read locks
2725 */
2726 if (p->need_commit_sem)
2727 down_read(&root->fs_info->commit_root_sem);
2728 b = root->commit_root;
2729 extent_buffer_get(b);
2730 level = btrfs_header_level(b);
2731 if (p->need_commit_sem)
2732 up_read(&root->fs_info->commit_root_sem);
2733 if (!p->skip_locking)
2734 btrfs_tree_read_lock(b);
2735 } else {
2736 if (p->skip_locking) {
2737 b = btrfs_root_node(root);
2738 level = btrfs_header_level(b);
2739 } else {
2740 /* we don't know the level of the root node
2741 * until we actually have it read locked
2742 */
2743 b = btrfs_read_lock_root_node(root);
2744 level = btrfs_header_level(b);
2745 if (level <= write_lock_level) {
2746 /* whoops, must trade for write lock */
2747 btrfs_tree_read_unlock(b);
2748 free_extent_buffer(b);
2749 b = btrfs_lock_root_node(root);
2750 root_lock = BTRFS_WRITE_LOCK;
2751
2752 /* the level might have changed, check again */
2753 level = btrfs_header_level(b);
2754 }
2755 }
2756 }
2757 p->nodes[level] = b;
2758 if (!p->skip_locking)
2759 p->locks[level] = root_lock;
2760
2761 while (b) {
2762 level = btrfs_header_level(b);
2763
2764 /*
2765 * setup the path here so we can release it under lock
2766 * contention with the cow code
2767 */
2768 if (cow) {
2769 /*
2770 * if we don't really need to cow this block
2771 * then we don't want to set the path blocking,
2772 * so we test it here
2773 */
2774 if (!should_cow_block(trans, root, b))
2775 goto cow_done;
2776
2777 /*
2778 * must have write locks on this node and the
2779 * parent
2780 */
2781 if (level > write_lock_level ||
2782 (level + 1 > write_lock_level &&
2783 level + 1 < BTRFS_MAX_LEVEL &&
2784 p->nodes[level + 1])) {
2785 write_lock_level = level + 1;
2786 btrfs_release_path(p);
2787 goto again;
2788 }
2789
2790 btrfs_set_path_blocking(p);
2791 err = btrfs_cow_block(trans, root, b,
2792 p->nodes[level + 1],
2793 p->slots[level + 1], &b);
2794 if (err) {
2795 ret = err;
2796 goto done;
2797 }
2798 }
2799cow_done:
2800 p->nodes[level] = b;
2801 btrfs_clear_path_blocking(p, NULL, 0);
2802
2803 /*
2804 * we have a lock on b and as long as we aren't changing
2805 * the tree, there is no way to for the items in b to change.
2806 * It is safe to drop the lock on our parent before we
2807 * go through the expensive btree search on b.
2808 *
2809 * If we're inserting or deleting (ins_len != 0), then we might
2810 * be changing slot zero, which may require changing the parent.
2811 * So, we can't drop the lock until after we know which slot
2812 * we're operating on.
2813 */
2814 if (!ins_len && !p->keep_locks) {
2815 int u = level + 1;
2816
2817 if (u < BTRFS_MAX_LEVEL && p->locks[u]) {
2818 btrfs_tree_unlock_rw(p->nodes[u], p->locks[u]);
2819 p->locks[u] = 0;
2820 }
2821 }
2822
2823 ret = key_search(b, key, level, &prev_cmp, &slot);
2824
2825 if (level != 0) {
2826 int dec = 0;
2827 if (ret && slot > 0) {
2828 dec = 1;
2829 slot -= 1;
2830 }
2831 p->slots[level] = slot;
2832 err = setup_nodes_for_search(trans, root, p, b, level,
2833 ins_len, &write_lock_level);
2834 if (err == -EAGAIN)
2835 goto again;
2836 if (err) {
2837 ret = err;
2838 goto done;
2839 }
2840 b = p->nodes[level];
2841 slot = p->slots[level];
2842
2843 /*
2844 * slot 0 is special, if we change the key
2845 * we have to update the parent pointer
2846 * which means we must have a write lock
2847 * on the parent
2848 */
2849 if (slot == 0 && ins_len &&
2850 write_lock_level < level + 1) {
2851 write_lock_level = level + 1;
2852 btrfs_release_path(p);
2853 goto again;
2854 }
2855
2856 unlock_up(p, level, lowest_unlock,
2857 min_write_lock_level, &write_lock_level);
2858
2859 if (level == lowest_level) {
2860 if (dec)
2861 p->slots[level]++;
2862 goto done;
2863 }
2864
2865 err = read_block_for_search(trans, root, p,
2866 &b, level, slot, key, 0);
2867 if (err == -EAGAIN)
2868 goto again;
2869 if (err) {
2870 ret = err;
2871 goto done;
2872 }
2873
2874 if (!p->skip_locking) {
2875 level = btrfs_header_level(b);
2876 if (level <= write_lock_level) {
2877 err = btrfs_try_tree_write_lock(b);
2878 if (!err) {
2879 btrfs_set_path_blocking(p);
2880 btrfs_tree_lock(b);
2881 btrfs_clear_path_blocking(p, b,
2882 BTRFS_WRITE_LOCK);
2883 }
2884 p->locks[level] = BTRFS_WRITE_LOCK;
2885 } else {
2886 err = btrfs_tree_read_lock_atomic(b);
2887 if (!err) {
2888 btrfs_set_path_blocking(p);
2889 btrfs_tree_read_lock(b);
2890 btrfs_clear_path_blocking(p, b,
2891 BTRFS_READ_LOCK);
2892 }
2893 p->locks[level] = BTRFS_READ_LOCK;
2894 }
2895 p->nodes[level] = b;
2896 }
2897 } else {
2898 p->slots[level] = slot;
2899 if (ins_len > 0 &&
2900 btrfs_leaf_free_space(root, b) < ins_len) {
2901 if (write_lock_level < 1) {
2902 write_lock_level = 1;
2903 btrfs_release_path(p);
2904 goto again;
2905 }
2906
2907 btrfs_set_path_blocking(p);
2908 err = split_leaf(trans, root, key,
2909 p, ins_len, ret == 0);
2910 btrfs_clear_path_blocking(p, NULL, 0);
2911
2912 BUG_ON(err > 0);
2913 if (err) {
2914 ret = err;
2915 goto done;
2916 }
2917 }
2918 if (!p->search_for_split)
2919 unlock_up(p, level, lowest_unlock,
2920 min_write_lock_level, &write_lock_level);
2921 goto done;
2922 }
2923 }
2924 ret = 1;
2925done:
2926 /*
2927 * we don't really know what they plan on doing with the path
2928 * from here on, so for now just mark it as blocking
2929 */
2930 if (!p->leave_spinning)
2931 btrfs_set_path_blocking(p);
2932 if (ret < 0 && !p->skip_release_on_error)
2933 btrfs_release_path(p);
2934 return ret;
2935}
2936
2937/*
2938 * Like btrfs_search_slot, this looks for a key in the given tree. It uses the
2939 * current state of the tree together with the operations recorded in the tree
2940 * modification log to search for the key in a previous version of this tree, as
2941 * denoted by the time_seq parameter.
2942 *
2943 * Naturally, there is no support for insert, delete or cow operations.
2944 *
2945 * The resulting path and return value will be set up as if we called
2946 * btrfs_search_slot at that point in time with ins_len and cow both set to 0.
2947 */
2948int btrfs_search_old_slot(struct btrfs_root *root, struct btrfs_key *key,
2949 struct btrfs_path *p, u64 time_seq)
2950{
2951 struct extent_buffer *b;
2952 int slot;
2953 int ret;
2954 int err;
2955 int level;
2956 int lowest_unlock = 1;
2957 u8 lowest_level = 0;
2958 int prev_cmp = -1;
2959
2960 lowest_level = p->lowest_level;
2961 WARN_ON(p->nodes[0] != NULL);
2962
2963 if (p->search_commit_root) {
2964 BUG_ON(time_seq);
2965 return btrfs_search_slot(NULL, root, key, p, 0, 0);
2966 }
2967
2968again:
2969 b = get_old_root(root, time_seq);
2970 level = btrfs_header_level(b);
2971 p->locks[level] = BTRFS_READ_LOCK;
2972
2973 while (b) {
2974 level = btrfs_header_level(b);
2975 p->nodes[level] = b;
2976 btrfs_clear_path_blocking(p, NULL, 0);
2977
2978 /*
2979 * we have a lock on b and as long as we aren't changing
2980 * the tree, there is no way to for the items in b to change.
2981 * It is safe to drop the lock on our parent before we
2982 * go through the expensive btree search on b.
2983 */
2984 btrfs_unlock_up_safe(p, level + 1);
2985
2986 /*
2987 * Since we can unwind eb's we want to do a real search every
2988 * time.
2989 */
2990 prev_cmp = -1;
2991 ret = key_search(b, key, level, &prev_cmp, &slot);
2992
2993 if (level != 0) {
2994 int dec = 0;
2995 if (ret && slot > 0) {
2996 dec = 1;
2997 slot -= 1;
2998 }
2999 p->slots[level] = slot;
3000 unlock_up(p, level, lowest_unlock, 0, NULL);
3001
3002 if (level == lowest_level) {
3003 if (dec)
3004 p->slots[level]++;
3005 goto done;
3006 }
3007
3008 err = read_block_for_search(NULL, root, p, &b, level,
3009 slot, key, time_seq);
3010 if (err == -EAGAIN)
3011 goto again;
3012 if (err) {
3013 ret = err;
3014 goto done;
3015 }
3016
3017 level = btrfs_header_level(b);
3018 err = btrfs_tree_read_lock_atomic(b);
3019 if (!err) {
3020 btrfs_set_path_blocking(p);
3021 btrfs_tree_read_lock(b);
3022 btrfs_clear_path_blocking(p, b,
3023 BTRFS_READ_LOCK);
3024 }
3025 b = tree_mod_log_rewind(root->fs_info, p, b, time_seq);
3026 if (!b) {
3027 ret = -ENOMEM;
3028 goto done;
3029 }
3030 p->locks[level] = BTRFS_READ_LOCK;
3031 p->nodes[level] = b;
3032 } else {
3033 p->slots[level] = slot;
3034 unlock_up(p, level, lowest_unlock, 0, NULL);
3035 goto done;
3036 }
3037 }
3038 ret = 1;
3039done:
3040 if (!p->leave_spinning)
3041 btrfs_set_path_blocking(p);
3042 if (ret < 0)
3043 btrfs_release_path(p);
3044
3045 return ret;
3046}
3047
3048/*
3049 * helper to use instead of search slot if no exact match is needed but
3050 * instead the next or previous item should be returned.
3051 * When find_higher is true, the next higher item is returned, the next lower
3052 * otherwise.
3053 * When return_any and find_higher are both true, and no higher item is found,
3054 * return the next lower instead.
3055 * When return_any is true and find_higher is false, and no lower item is found,
3056 * return the next higher instead.
3057 * It returns 0 if any item is found, 1 if none is found (tree empty), and
3058 * < 0 on error
3059 */
3060int btrfs_search_slot_for_read(struct btrfs_root *root,
3061 struct btrfs_key *key, struct btrfs_path *p,
3062 int find_higher, int return_any)
3063{
3064 int ret;
3065 struct extent_buffer *leaf;
3066
3067again:
3068 ret = btrfs_search_slot(NULL, root, key, p, 0, 0);
3069 if (ret <= 0)
3070 return ret;
3071 /*
3072 * a return value of 1 means the path is at the position where the
3073 * item should be inserted. Normally this is the next bigger item,
3074 * but in case the previous item is the last in a leaf, path points
3075 * to the first free slot in the previous leaf, i.e. at an invalid
3076 * item.
3077 */
3078 leaf = p->nodes[0];
3079
3080 if (find_higher) {
3081 if (p->slots[0] >= btrfs_header_nritems(leaf)) {
3082 ret = btrfs_next_leaf(root, p);
3083 if (ret <= 0)
3084 return ret;
3085 if (!return_any)
3086 return 1;
3087 /*
3088 * no higher item found, return the next
3089 * lower instead
3090 */
3091 return_any = 0;
3092 find_higher = 0;
3093 btrfs_release_path(p);
3094 goto again;
3095 }
3096 } else {
3097 if (p->slots[0] == 0) {
3098 ret = btrfs_prev_leaf(root, p);
3099 if (ret < 0)
3100 return ret;
3101 if (!ret) {
3102 leaf = p->nodes[0];
3103 if (p->slots[0] == btrfs_header_nritems(leaf))
3104 p->slots[0]--;
3105 return 0;
3106 }
3107 if (!return_any)
3108 return 1;
3109 /*
3110 * no lower item found, return the next
3111 * higher instead
3112 */
3113 return_any = 0;
3114 find_higher = 1;
3115 btrfs_release_path(p);
3116 goto again;
3117 } else {
3118 --p->slots[0];
3119 }
3120 }
3121 return 0;
3122}
3123
3124/*
3125 * adjust the pointers going up the tree, starting at level
3126 * making sure the right key of each node is points to 'key'.
3127 * This is used after shifting pointers to the left, so it stops
3128 * fixing up pointers when a given leaf/node is not in slot 0 of the
3129 * higher levels
3130 *
3131 */
3132static void fixup_low_keys(struct btrfs_root *root, struct btrfs_path *path,
3133 struct btrfs_disk_key *key, int level)
3134{
3135 int i;
3136 struct extent_buffer *t;
3137
3138 for (i = level; i < BTRFS_MAX_LEVEL; i++) {
3139 int tslot = path->slots[i];
3140 if (!path->nodes[i])
3141 break;
3142 t = path->nodes[i];
3143 tree_mod_log_set_node_key(root->fs_info, t, tslot, 1);
3144 btrfs_set_node_key(t, key, tslot);
3145 btrfs_mark_buffer_dirty(path->nodes[i]);
3146 if (tslot != 0)
3147 break;
3148 }
3149}
3150
3151/*
3152 * update item key.
3153 *
3154 * This function isn't completely safe. It's the caller's responsibility
3155 * that the new key won't break the order
3156 */
3157void btrfs_set_item_key_safe(struct btrfs_root *root, struct btrfs_path *path,
3158 struct btrfs_key *new_key)
3159{
3160 struct btrfs_disk_key disk_key;
3161 struct extent_buffer *eb;
3162 int slot;
3163
3164 eb = path->nodes[0];
3165 slot = path->slots[0];
3166 if (slot > 0) {
3167 btrfs_item_key(eb, &disk_key, slot - 1);
3168 BUG_ON(comp_keys(&disk_key, new_key) >= 0);
3169 }
3170 if (slot < btrfs_header_nritems(eb) - 1) {
3171 btrfs_item_key(eb, &disk_key, slot + 1);
3172 BUG_ON(comp_keys(&disk_key, new_key) <= 0);
3173 }
3174
3175 btrfs_cpu_key_to_disk(&disk_key, new_key);
3176 btrfs_set_item_key(eb, &disk_key, slot);
3177 btrfs_mark_buffer_dirty(eb);
3178 if (slot == 0)
3179 fixup_low_keys(root, path, &disk_key, 1);
3180}
3181
3182/*
3183 * try to push data from one node into the next node left in the
3184 * tree.
3185 *
3186 * returns 0 if some ptrs were pushed left, < 0 if there was some horrible
3187 * error, and > 0 if there was no room in the left hand block.
3188 */
3189static int push_node_left(struct btrfs_trans_handle *trans,
3190 struct btrfs_root *root, struct extent_buffer *dst,
3191 struct extent_buffer *src, int empty)
3192{
3193 int push_items = 0;
3194 int src_nritems;
3195 int dst_nritems;
3196 int ret = 0;
3197
3198 src_nritems = btrfs_header_nritems(src);
3199 dst_nritems = btrfs_header_nritems(dst);
3200 push_items = BTRFS_NODEPTRS_PER_BLOCK(root) - dst_nritems;
3201 WARN_ON(btrfs_header_generation(src) != trans->transid);
3202 WARN_ON(btrfs_header_generation(dst) != trans->transid);
3203
3204 if (!empty && src_nritems <= 8)
3205 return 1;
3206
3207 if (push_items <= 0)
3208 return 1;
3209
3210 if (empty) {
3211 push_items = min(src_nritems, push_items);
3212 if (push_items < src_nritems) {
3213 /* leave at least 8 pointers in the node if
3214 * we aren't going to empty it
3215 */
3216 if (src_nritems - push_items < 8) {
3217 if (push_items <= 8)
3218 return 1;
3219 push_items -= 8;
3220 }
3221 }
3222 } else
3223 push_items = min(src_nritems - 8, push_items);
3224
3225 ret = tree_mod_log_eb_copy(root->fs_info, dst, src, dst_nritems, 0,
3226 push_items);
3227 if (ret) {
3228 btrfs_abort_transaction(trans, root, ret);
3229 return ret;
3230 }
3231 copy_extent_buffer(dst, src,
3232 btrfs_node_key_ptr_offset(dst_nritems),
3233 btrfs_node_key_ptr_offset(0),
3234 push_items * sizeof(struct btrfs_key_ptr));
3235
3236 if (push_items < src_nritems) {
3237 /*
3238 * don't call tree_mod_log_eb_move here, key removal was already
3239 * fully logged by tree_mod_log_eb_copy above.
3240 */
3241 memmove_extent_buffer(src, btrfs_node_key_ptr_offset(0),
3242 btrfs_node_key_ptr_offset(push_items),
3243 (src_nritems - push_items) *
3244 sizeof(struct btrfs_key_ptr));
3245 }
3246 btrfs_set_header_nritems(src, src_nritems - push_items);
3247 btrfs_set_header_nritems(dst, dst_nritems + push_items);
3248 btrfs_mark_buffer_dirty(src);
3249 btrfs_mark_buffer_dirty(dst);
3250
3251 return ret;
3252}
3253
3254/*
3255 * try to push data from one node into the next node right in the
3256 * tree.
3257 *
3258 * returns 0 if some ptrs were pushed, < 0 if there was some horrible
3259 * error, and > 0 if there was no room in the right hand block.
3260 *
3261 * this will only push up to 1/2 the contents of the left node over
3262 */
3263static int balance_node_right(struct btrfs_trans_handle *trans,
3264 struct btrfs_root *root,
3265 struct extent_buffer *dst,
3266 struct extent_buffer *src)
3267{
3268 int push_items = 0;
3269 int max_push;
3270 int src_nritems;
3271 int dst_nritems;
3272 int ret = 0;
3273
3274 WARN_ON(btrfs_header_generation(src) != trans->transid);
3275 WARN_ON(btrfs_header_generation(dst) != trans->transid);
3276
3277 src_nritems = btrfs_header_nritems(src);
3278 dst_nritems = btrfs_header_nritems(dst);
3279 push_items = BTRFS_NODEPTRS_PER_BLOCK(root) - dst_nritems;
3280 if (push_items <= 0)
3281 return 1;
3282
3283 if (src_nritems < 4)
3284 return 1;
3285
3286 max_push = src_nritems / 2 + 1;
3287 /* don't try to empty the node */
3288 if (max_push >= src_nritems)
3289 return 1;
3290
3291 if (max_push < push_items)
3292 push_items = max_push;
3293
3294 tree_mod_log_eb_move(root->fs_info, dst, push_items, 0, dst_nritems);
3295 memmove_extent_buffer(dst, btrfs_node_key_ptr_offset(push_items),
3296 btrfs_node_key_ptr_offset(0),
3297 (dst_nritems) *
3298 sizeof(struct btrfs_key_ptr));
3299
3300 ret = tree_mod_log_eb_copy(root->fs_info, dst, src, 0,
3301 src_nritems - push_items, push_items);
3302 if (ret) {
3303 btrfs_abort_transaction(trans, root, ret);
3304 return ret;
3305 }
3306 copy_extent_buffer(dst, src,
3307 btrfs_node_key_ptr_offset(0),
3308 btrfs_node_key_ptr_offset(src_nritems - push_items),
3309 push_items * sizeof(struct btrfs_key_ptr));
3310
3311 btrfs_set_header_nritems(src, src_nritems - push_items);
3312 btrfs_set_header_nritems(dst, dst_nritems + push_items);
3313
3314 btrfs_mark_buffer_dirty(src);
3315 btrfs_mark_buffer_dirty(dst);
3316
3317 return ret;
3318}
3319
3320/*
3321 * helper function to insert a new root level in the tree.
3322 * A new node is allocated, and a single item is inserted to
3323 * point to the existing root
3324 *
3325 * returns zero on success or < 0 on failure.
3326 */
3327static noinline int insert_new_root(struct btrfs_trans_handle *trans,
3328 struct btrfs_root *root,
3329 struct btrfs_path *path, int level)
3330{
3331 u64 lower_gen;
3332 struct extent_buffer *lower;
3333 struct extent_buffer *c;
3334 struct extent_buffer *old;
3335 struct btrfs_disk_key lower_key;
3336
3337 BUG_ON(path->nodes[level]);
3338 BUG_ON(path->nodes[level-1] != root->node);
3339
3340 lower = path->nodes[level-1];
3341 if (level == 1)
3342 btrfs_item_key(lower, &lower_key, 0);
3343 else
3344 btrfs_node_key(lower, &lower_key, 0);
3345
3346 c = btrfs_alloc_tree_block(trans, root, 0, root->root_key.objectid,
3347 &lower_key, level, root->node->start, 0);
3348 if (IS_ERR(c))
3349 return PTR_ERR(c);
3350
3351 root_add_used(root, root->nodesize);
3352
3353 memset_extent_buffer(c, 0, 0, sizeof(struct btrfs_header));
3354 btrfs_set_header_nritems(c, 1);
3355 btrfs_set_header_level(c, level);
3356 btrfs_set_header_bytenr(c, c->start);
3357 btrfs_set_header_generation(c, trans->transid);
3358 btrfs_set_header_backref_rev(c, BTRFS_MIXED_BACKREF_REV);
3359 btrfs_set_header_owner(c, root->root_key.objectid);
3360
3361 write_extent_buffer(c, root->fs_info->fsid, btrfs_header_fsid(),
3362 BTRFS_FSID_SIZE);
3363
3364 write_extent_buffer(c, root->fs_info->chunk_tree_uuid,
3365 btrfs_header_chunk_tree_uuid(c), BTRFS_UUID_SIZE);
3366
3367 btrfs_set_node_key(c, &lower_key, 0);
3368 btrfs_set_node_blockptr(c, 0, lower->start);
3369 lower_gen = btrfs_header_generation(lower);
3370 WARN_ON(lower_gen != trans->transid);
3371
3372 btrfs_set_node_ptr_generation(c, 0, lower_gen);
3373
3374 btrfs_mark_buffer_dirty(c);
3375
3376 old = root->node;
3377 tree_mod_log_set_root_pointer(root, c, 0);
3378 rcu_assign_pointer(root->node, c);
3379
3380 /* the super has an extra ref to root->node */
3381 free_extent_buffer(old);
3382
3383 add_root_to_dirty_list(root);
3384 extent_buffer_get(c);
3385 path->nodes[level] = c;
3386 path->locks[level] = BTRFS_WRITE_LOCK;
3387 path->slots[level] = 0;
3388 return 0;
3389}
3390
3391/*
3392 * worker function to insert a single pointer in a node.
3393 * the node should have enough room for the pointer already
3394 *
3395 * slot and level indicate where you want the key to go, and
3396 * blocknr is the block the key points to.
3397 */
3398static void insert_ptr(struct btrfs_trans_handle *trans,
3399 struct btrfs_root *root, struct btrfs_path *path,
3400 struct btrfs_disk_key *key, u64 bytenr,
3401 int slot, int level)
3402{
3403 struct extent_buffer *lower;
3404 int nritems;
3405 int ret;
3406
3407 BUG_ON(!path->nodes[level]);
3408 btrfs_assert_tree_locked(path->nodes[level]);
3409 lower = path->nodes[level];
3410 nritems = btrfs_header_nritems(lower);
3411 BUG_ON(slot > nritems);
3412 BUG_ON(nritems == BTRFS_NODEPTRS_PER_BLOCK(root));
3413 if (slot != nritems) {
3414 if (level)
3415 tree_mod_log_eb_move(root->fs_info, lower, slot + 1,
3416 slot, nritems - slot);
3417 memmove_extent_buffer(lower,
3418 btrfs_node_key_ptr_offset(slot + 1),
3419 btrfs_node_key_ptr_offset(slot),
3420 (nritems - slot) * sizeof(struct btrfs_key_ptr));
3421 }
3422 if (level) {
3423 ret = tree_mod_log_insert_key(root->fs_info, lower, slot,
3424 MOD_LOG_KEY_ADD, GFP_NOFS);
3425 BUG_ON(ret < 0);
3426 }
3427 btrfs_set_node_key(lower, key, slot);
3428 btrfs_set_node_blockptr(lower, slot, bytenr);
3429 WARN_ON(trans->transid == 0);
3430 btrfs_set_node_ptr_generation(lower, slot, trans->transid);
3431 btrfs_set_header_nritems(lower, nritems + 1);
3432 btrfs_mark_buffer_dirty(lower);
3433}
3434
3435/*
3436 * split the node at the specified level in path in two.
3437 * The path is corrected to point to the appropriate node after the split
3438 *
3439 * Before splitting this tries to make some room in the node by pushing
3440 * left and right, if either one works, it returns right away.
3441 *
3442 * returns 0 on success and < 0 on failure
3443 */
3444static noinline int split_node(struct btrfs_trans_handle *trans,
3445 struct btrfs_root *root,
3446 struct btrfs_path *path, int level)
3447{
3448 struct extent_buffer *c;
3449 struct extent_buffer *split;
3450 struct btrfs_disk_key disk_key;
3451 int mid;
3452 int ret;
3453 u32 c_nritems;
3454
3455 c = path->nodes[level];
3456 WARN_ON(btrfs_header_generation(c) != trans->transid);
3457 if (c == root->node) {
3458 /*
3459 * trying to split the root, lets make a new one
3460 *
3461 * tree mod log: We don't log_removal old root in
3462 * insert_new_root, because that root buffer will be kept as a
3463 * normal node. We are going to log removal of half of the
3464 * elements below with tree_mod_log_eb_copy. We're holding a
3465 * tree lock on the buffer, which is why we cannot race with
3466 * other tree_mod_log users.
3467 */
3468 ret = insert_new_root(trans, root, path, level + 1);
3469 if (ret)
3470 return ret;
3471 } else {
3472 ret = push_nodes_for_insert(trans, root, path, level);
3473 c = path->nodes[level];
3474 if (!ret && btrfs_header_nritems(c) <
3475 BTRFS_NODEPTRS_PER_BLOCK(root) - 3)
3476 return 0;
3477 if (ret < 0)
3478 return ret;
3479 }
3480
3481 c_nritems = btrfs_header_nritems(c);
3482 mid = (c_nritems + 1) / 2;
3483 btrfs_node_key(c, &disk_key, mid);
3484
3485 split = btrfs_alloc_tree_block(trans, root, 0, root->root_key.objectid,
3486 &disk_key, level, c->start, 0);
3487 if (IS_ERR(split))
3488 return PTR_ERR(split);
3489
3490 root_add_used(root, root->nodesize);
3491
3492 memset_extent_buffer(split, 0, 0, sizeof(struct btrfs_header));
3493 btrfs_set_header_level(split, btrfs_header_level(c));
3494 btrfs_set_header_bytenr(split, split->start);
3495 btrfs_set_header_generation(split, trans->transid);
3496 btrfs_set_header_backref_rev(split, BTRFS_MIXED_BACKREF_REV);
3497 btrfs_set_header_owner(split, root->root_key.objectid);
3498 write_extent_buffer(split, root->fs_info->fsid,
3499 btrfs_header_fsid(), BTRFS_FSID_SIZE);
3500 write_extent_buffer(split, root->fs_info->chunk_tree_uuid,
3501 btrfs_header_chunk_tree_uuid(split),
3502 BTRFS_UUID_SIZE);
3503
3504 ret = tree_mod_log_eb_copy(root->fs_info, split, c, 0,
3505 mid, c_nritems - mid);
3506 if (ret) {
3507 btrfs_abort_transaction(trans, root, ret);
3508 return ret;
3509 }
3510 copy_extent_buffer(split, c,
3511 btrfs_node_key_ptr_offset(0),
3512 btrfs_node_key_ptr_offset(mid),
3513 (c_nritems - mid) * sizeof(struct btrfs_key_ptr));
3514 btrfs_set_header_nritems(split, c_nritems - mid);
3515 btrfs_set_header_nritems(c, mid);
3516 ret = 0;
3517
3518 btrfs_mark_buffer_dirty(c);
3519 btrfs_mark_buffer_dirty(split);
3520
3521 insert_ptr(trans, root, path, &disk_key, split->start,
3522 path->slots[level + 1] + 1, level + 1);
3523
3524 if (path->slots[level] >= mid) {
3525 path->slots[level] -= mid;
3526 btrfs_tree_unlock(c);
3527 free_extent_buffer(c);
3528 path->nodes[level] = split;
3529 path->slots[level + 1] += 1;
3530 } else {
3531 btrfs_tree_unlock(split);
3532 free_extent_buffer(split);
3533 }
3534 return ret;
3535}
3536
3537/*
3538 * how many bytes are required to store the items in a leaf. start
3539 * and nr indicate which items in the leaf to check. This totals up the
3540 * space used both by the item structs and the item data
3541 */
3542static int leaf_space_used(struct extent_buffer *l, int start, int nr)
3543{
3544 struct btrfs_item *start_item;
3545 struct btrfs_item *end_item;
3546 struct btrfs_map_token token;
3547 int data_len;
3548 int nritems = btrfs_header_nritems(l);
3549 int end = min(nritems, start + nr) - 1;
3550
3551 if (!nr)
3552 return 0;
3553 btrfs_init_map_token(&token);
3554 start_item = btrfs_item_nr(start);
3555 end_item = btrfs_item_nr(end);
3556 data_len = btrfs_token_item_offset(l, start_item, &token) +
3557 btrfs_token_item_size(l, start_item, &token);
3558 data_len = data_len - btrfs_token_item_offset(l, end_item, &token);
3559 data_len += sizeof(struct btrfs_item) * nr;
3560 WARN_ON(data_len < 0);
3561 return data_len;
3562}
3563
3564/*
3565 * The space between the end of the leaf items and
3566 * the start of the leaf data. IOW, how much room
3567 * the leaf has left for both items and data
3568 */
3569noinline int btrfs_leaf_free_space(struct btrfs_root *root,
3570 struct extent_buffer *leaf)
3571{
3572 int nritems = btrfs_header_nritems(leaf);
3573 int ret;
3574 ret = BTRFS_LEAF_DATA_SIZE(root) - leaf_space_used(leaf, 0, nritems);
3575 if (ret < 0) {
3576 btrfs_crit(root->fs_info,
3577 "leaf free space ret %d, leaf data size %lu, used %d nritems %d",
3578 ret, (unsigned long) BTRFS_LEAF_DATA_SIZE(root),
3579 leaf_space_used(leaf, 0, nritems), nritems);
3580 }
3581 return ret;
3582}
3583
3584/*
3585 * min slot controls the lowest index we're willing to push to the
3586 * right. We'll push up to and including min_slot, but no lower
3587 */
3588static noinline int __push_leaf_right(struct btrfs_trans_handle *trans,
3589 struct btrfs_root *root,
3590 struct btrfs_path *path,
3591 int data_size, int empty,
3592 struct extent_buffer *right,
3593 int free_space, u32 left_nritems,
3594 u32 min_slot)
3595{
3596 struct extent_buffer *left = path->nodes[0];
3597 struct extent_buffer *upper = path->nodes[1];
3598 struct btrfs_map_token token;
3599 struct btrfs_disk_key disk_key;
3600 int slot;
3601 u32 i;
3602 int push_space = 0;
3603 int push_items = 0;
3604 struct btrfs_item *item;
3605 u32 nr;
3606 u32 right_nritems;
3607 u32 data_end;
3608 u32 this_item_size;
3609
3610 btrfs_init_map_token(&token);
3611
3612 if (empty)
3613 nr = 0;
3614 else
3615 nr = max_t(u32, 1, min_slot);
3616
3617 if (path->slots[0] >= left_nritems)
3618 push_space += data_size;
3619
3620 slot = path->slots[1];
3621 i = left_nritems - 1;
3622 while (i >= nr) {
3623 item = btrfs_item_nr(i);
3624
3625 if (!empty && push_items > 0) {
3626 if (path->slots[0] > i)
3627 break;
3628 if (path->slots[0] == i) {
3629 int space = btrfs_leaf_free_space(root, left);
3630 if (space + push_space * 2 > free_space)
3631 break;
3632 }
3633 }
3634
3635 if (path->slots[0] == i)
3636 push_space += data_size;
3637
3638 this_item_size = btrfs_item_size(left, item);
3639 if (this_item_size + sizeof(*item) + push_space > free_space)
3640 break;
3641
3642 push_items++;
3643 push_space += this_item_size + sizeof(*item);
3644 if (i == 0)
3645 break;
3646 i--;
3647 }
3648
3649 if (push_items == 0)
3650 goto out_unlock;
3651
3652 WARN_ON(!empty && push_items == left_nritems);
3653
3654 /* push left to right */
3655 right_nritems = btrfs_header_nritems(right);
3656
3657 push_space = btrfs_item_end_nr(left, left_nritems - push_items);
3658 push_space -= leaf_data_end(root, left);
3659
3660 /* make room in the right data area */
3661 data_end = leaf_data_end(root, right);
3662 memmove_extent_buffer(right,
3663 btrfs_leaf_data(right) + data_end - push_space,
3664 btrfs_leaf_data(right) + data_end,
3665 BTRFS_LEAF_DATA_SIZE(root) - data_end);
3666
3667 /* copy from the left data area */
3668 copy_extent_buffer(right, left, btrfs_leaf_data(right) +
3669 BTRFS_LEAF_DATA_SIZE(root) - push_space,
3670 btrfs_leaf_data(left) + leaf_data_end(root, left),
3671 push_space);
3672
3673 memmove_extent_buffer(right, btrfs_item_nr_offset(push_items),
3674 btrfs_item_nr_offset(0),
3675 right_nritems * sizeof(struct btrfs_item));
3676
3677 /* copy the items from left to right */
3678 copy_extent_buffer(right, left, btrfs_item_nr_offset(0),
3679 btrfs_item_nr_offset(left_nritems - push_items),
3680 push_items * sizeof(struct btrfs_item));
3681
3682 /* update the item pointers */
3683 right_nritems += push_items;
3684 btrfs_set_header_nritems(right, right_nritems);
3685 push_space = BTRFS_LEAF_DATA_SIZE(root);
3686 for (i = 0; i < right_nritems; i++) {
3687 item = btrfs_item_nr(i);
3688 push_space -= btrfs_token_item_size(right, item, &token);
3689 btrfs_set_token_item_offset(right, item, push_space, &token);
3690 }
3691
3692 left_nritems -= push_items;
3693 btrfs_set_header_nritems(left, left_nritems);
3694
3695 if (left_nritems)
3696 btrfs_mark_buffer_dirty(left);
3697 else
3698 clean_tree_block(trans, root, left);
3699
3700 btrfs_mark_buffer_dirty(right);
3701
3702 btrfs_item_key(right, &disk_key, 0);
3703 btrfs_set_node_key(upper, &disk_key, slot + 1);
3704 btrfs_mark_buffer_dirty(upper);
3705
3706 /* then fixup the leaf pointer in the path */
3707 if (path->slots[0] >= left_nritems) {
3708 path->slots[0] -= left_nritems;
3709 if (btrfs_header_nritems(path->nodes[0]) == 0)
3710 clean_tree_block(trans, root, path->nodes[0]);
3711 btrfs_tree_unlock(path->nodes[0]);
3712 free_extent_buffer(path->nodes[0]);
3713 path->nodes[0] = right;
3714 path->slots[1] += 1;
3715 } else {
3716 btrfs_tree_unlock(right);
3717 free_extent_buffer(right);
3718 }
3719 return 0;
3720
3721out_unlock:
3722 btrfs_tree_unlock(right);
3723 free_extent_buffer(right);
3724 return 1;
3725}
3726
3727/*
3728 * push some data in the path leaf to the right, trying to free up at
3729 * least data_size bytes. returns zero if the push worked, nonzero otherwise
3730 *
3731 * returns 1 if the push failed because the other node didn't have enough
3732 * room, 0 if everything worked out and < 0 if there were major errors.
3733 *
3734 * this will push starting from min_slot to the end of the leaf. It won't
3735 * push any slot lower than min_slot
3736 */
3737static int push_leaf_right(struct btrfs_trans_handle *trans, struct btrfs_root
3738 *root, struct btrfs_path *path,
3739 int min_data_size, int data_size,
3740 int empty, u32 min_slot)
3741{
3742 struct extent_buffer *left = path->nodes[0];
3743 struct extent_buffer *right;
3744 struct extent_buffer *upper;
3745 int slot;
3746 int free_space;
3747 u32 left_nritems;
3748 int ret;
3749
3750 if (!path->nodes[1])
3751 return 1;
3752
3753 slot = path->slots[1];
3754 upper = path->nodes[1];
3755 if (slot >= btrfs_header_nritems(upper) - 1)
3756 return 1;
3757
3758 btrfs_assert_tree_locked(path->nodes[1]);
3759
3760 right = read_node_slot(root, upper, slot + 1);
3761 if (right == NULL)
3762 return 1;
3763
3764 btrfs_tree_lock(right);
3765 btrfs_set_lock_blocking(right);
3766
3767 free_space = btrfs_leaf_free_space(root, right);
3768 if (free_space < data_size)
3769 goto out_unlock;
3770
3771 /* cow and double check */
3772 ret = btrfs_cow_block(trans, root, right, upper,
3773 slot + 1, &right);
3774 if (ret)
3775 goto out_unlock;
3776
3777 free_space = btrfs_leaf_free_space(root, right);
3778 if (free_space < data_size)
3779 goto out_unlock;
3780
3781 left_nritems = btrfs_header_nritems(left);
3782 if (left_nritems == 0)
3783 goto out_unlock;
3784
3785 if (path->slots[0] == left_nritems && !empty) {
3786 /* Key greater than all keys in the leaf, right neighbor has
3787 * enough room for it and we're not emptying our leaf to delete
3788 * it, therefore use right neighbor to insert the new item and
3789 * no need to touch/dirty our left leaft. */
3790 btrfs_tree_unlock(left);
3791 free_extent_buffer(left);
3792 path->nodes[0] = right;
3793 path->slots[0] = 0;
3794 path->slots[1]++;
3795 return 0;
3796 }
3797
3798 return __push_leaf_right(trans, root, path, min_data_size, empty,
3799 right, free_space, left_nritems, min_slot);
3800out_unlock:
3801 btrfs_tree_unlock(right);
3802 free_extent_buffer(right);
3803 return 1;
3804}
3805
3806/*
3807 * push some data in the path leaf to the left, trying to free up at
3808 * least data_size bytes. returns zero if the push worked, nonzero otherwise
3809 *
3810 * max_slot can put a limit on how far into the leaf we'll push items. The
3811 * item at 'max_slot' won't be touched. Use (u32)-1 to make us do all the
3812 * items
3813 */
3814static noinline int __push_leaf_left(struct btrfs_trans_handle *trans,
3815 struct btrfs_root *root,
3816 struct btrfs_path *path, int data_size,
3817 int empty, struct extent_buffer *left,
3818 int free_space, u32 right_nritems,
3819 u32 max_slot)
3820{
3821 struct btrfs_disk_key disk_key;
3822 struct extent_buffer *right = path->nodes[0];
3823 int i;
3824 int push_space = 0;
3825 int push_items = 0;
3826 struct btrfs_item *item;
3827 u32 old_left_nritems;
3828 u32 nr;
3829 int ret = 0;
3830 u32 this_item_size;
3831 u32 old_left_item_size;
3832 struct btrfs_map_token token;
3833
3834 btrfs_init_map_token(&token);
3835
3836 if (empty)
3837 nr = min(right_nritems, max_slot);
3838 else
3839 nr = min(right_nritems - 1, max_slot);
3840
3841 for (i = 0; i < nr; i++) {
3842 item = btrfs_item_nr(i);
3843
3844 if (!empty && push_items > 0) {
3845 if (path->slots[0] < i)
3846 break;
3847 if (path->slots[0] == i) {
3848 int space = btrfs_leaf_free_space(root, right);
3849 if (space + push_space * 2 > free_space)
3850 break;
3851 }
3852 }
3853
3854 if (path->slots[0] == i)
3855 push_space += data_size;
3856
3857 this_item_size = btrfs_item_size(right, item);
3858 if (this_item_size + sizeof(*item) + push_space > free_space)
3859 break;
3860
3861 push_items++;
3862 push_space += this_item_size + sizeof(*item);
3863 }
3864
3865 if (push_items == 0) {
3866 ret = 1;
3867 goto out;
3868 }
3869 WARN_ON(!empty && push_items == btrfs_header_nritems(right));
3870
3871 /* push data from right to left */
3872 copy_extent_buffer(left, right,
3873 btrfs_item_nr_offset(btrfs_header_nritems(left)),
3874 btrfs_item_nr_offset(0),
3875 push_items * sizeof(struct btrfs_item));
3876
3877 push_space = BTRFS_LEAF_DATA_SIZE(root) -
3878 btrfs_item_offset_nr(right, push_items - 1);
3879
3880 copy_extent_buffer(left, right, btrfs_leaf_data(left) +
3881 leaf_data_end(root, left) - push_space,
3882 btrfs_leaf_data(right) +
3883 btrfs_item_offset_nr(right, push_items - 1),
3884 push_space);
3885 old_left_nritems = btrfs_header_nritems(left);
3886 BUG_ON(old_left_nritems <= 0);
3887
3888 old_left_item_size = btrfs_item_offset_nr(left, old_left_nritems - 1);
3889 for (i = old_left_nritems; i < old_left_nritems + push_items; i++) {
3890 u32 ioff;
3891
3892 item = btrfs_item_nr(i);
3893
3894 ioff = btrfs_token_item_offset(left, item, &token);
3895 btrfs_set_token_item_offset(left, item,
3896 ioff - (BTRFS_LEAF_DATA_SIZE(root) - old_left_item_size),
3897 &token);
3898 }
3899 btrfs_set_header_nritems(left, old_left_nritems + push_items);
3900
3901 /* fixup right node */
3902 if (push_items > right_nritems)
3903 WARN(1, KERN_CRIT "push items %d nr %u\n", push_items,
3904 right_nritems);
3905
3906 if (push_items < right_nritems) {
3907 push_space = btrfs_item_offset_nr(right, push_items - 1) -
3908 leaf_data_end(root, right);
3909 memmove_extent_buffer(right, btrfs_leaf_data(right) +
3910 BTRFS_LEAF_DATA_SIZE(root) - push_space,
3911 btrfs_leaf_data(right) +
3912 leaf_data_end(root, right), push_space);
3913
3914 memmove_extent_buffer(right, btrfs_item_nr_offset(0),
3915 btrfs_item_nr_offset(push_items),
3916 (btrfs_header_nritems(right) - push_items) *
3917 sizeof(struct btrfs_item));
3918 }
3919 right_nritems -= push_items;
3920 btrfs_set_header_nritems(right, right_nritems);
3921 push_space = BTRFS_LEAF_DATA_SIZE(root);
3922 for (i = 0; i < right_nritems; i++) {
3923 item = btrfs_item_nr(i);
3924
3925 push_space = push_space - btrfs_token_item_size(right,
3926 item, &token);
3927 btrfs_set_token_item_offset(right, item, push_space, &token);
3928 }
3929
3930 btrfs_mark_buffer_dirty(left);
3931 if (right_nritems)
3932 btrfs_mark_buffer_dirty(right);
3933 else
3934 clean_tree_block(trans, root, right);
3935
3936 btrfs_item_key(right, &disk_key, 0);
3937 fixup_low_keys(root, path, &disk_key, 1);
3938
3939 /* then fixup the leaf pointer in the path */
3940 if (path->slots[0] < push_items) {
3941 path->slots[0] += old_left_nritems;
3942 btrfs_tree_unlock(path->nodes[0]);
3943 free_extent_buffer(path->nodes[0]);
3944 path->nodes[0] = left;
3945 path->slots[1] -= 1;
3946 } else {
3947 btrfs_tree_unlock(left);
3948 free_extent_buffer(left);
3949 path->slots[0] -= push_items;
3950 }
3951 BUG_ON(path->slots[0] < 0);
3952 return ret;
3953out:
3954 btrfs_tree_unlock(left);
3955 free_extent_buffer(left);
3956 return ret;
3957}
3958
3959/*
3960 * push some data in the path leaf to the left, trying to free up at
3961 * least data_size bytes. returns zero if the push worked, nonzero otherwise
3962 *
3963 * max_slot can put a limit on how far into the leaf we'll push items. The
3964 * item at 'max_slot' won't be touched. Use (u32)-1 to make us push all the
3965 * items
3966 */
3967static int push_leaf_left(struct btrfs_trans_handle *trans, struct btrfs_root
3968 *root, struct btrfs_path *path, int min_data_size,
3969 int data_size, int empty, u32 max_slot)
3970{
3971 struct extent_buffer *right = path->nodes[0];
3972 struct extent_buffer *left;
3973 int slot;
3974 int free_space;
3975 u32 right_nritems;
3976 int ret = 0;
3977
3978 slot = path->slots[1];
3979 if (slot == 0)
3980 return 1;
3981 if (!path->nodes[1])
3982 return 1;
3983
3984 right_nritems = btrfs_header_nritems(right);
3985 if (right_nritems == 0)
3986 return 1;
3987
3988 btrfs_assert_tree_locked(path->nodes[1]);
3989
3990 left = read_node_slot(root, path->nodes[1], slot - 1);
3991 if (left == NULL)
3992 return 1;
3993
3994 btrfs_tree_lock(left);
3995 btrfs_set_lock_blocking(left);
3996
3997 free_space = btrfs_leaf_free_space(root, left);
3998 if (free_space < data_size) {
3999 ret = 1;
4000 goto out;
4001 }
4002
4003 /* cow and double check */
4004 ret = btrfs_cow_block(trans, root, left,
4005 path->nodes[1], slot - 1, &left);
4006 if (ret) {
4007 /* we hit -ENOSPC, but it isn't fatal here */
4008 if (ret == -ENOSPC)
4009 ret = 1;
4010 goto out;
4011 }
4012
4013 free_space = btrfs_leaf_free_space(root, left);
4014 if (free_space < data_size) {
4015 ret = 1;
4016 goto out;
4017 }
4018
4019 return __push_leaf_left(trans, root, path, min_data_size,
4020 empty, left, free_space, right_nritems,
4021 max_slot);
4022out:
4023 btrfs_tree_unlock(left);
4024 free_extent_buffer(left);
4025 return ret;
4026}
4027
4028/*
4029 * split the path's leaf in two, making sure there is at least data_size
4030 * available for the resulting leaf level of the path.
4031 */
4032static noinline void copy_for_split(struct btrfs_trans_handle *trans,
4033 struct btrfs_root *root,
4034 struct btrfs_path *path,
4035 struct extent_buffer *l,
4036 struct extent_buffer *right,
4037 int slot, int mid, int nritems)
4038{
4039 int data_copy_size;
4040 int rt_data_off;
4041 int i;
4042 struct btrfs_disk_key disk_key;
4043 struct btrfs_map_token token;
4044
4045 btrfs_init_map_token(&token);
4046
4047 nritems = nritems - mid;
4048 btrfs_set_header_nritems(right, nritems);
4049 data_copy_size = btrfs_item_end_nr(l, mid) - leaf_data_end(root, l);
4050
4051 copy_extent_buffer(right, l, btrfs_item_nr_offset(0),
4052 btrfs_item_nr_offset(mid),
4053 nritems * sizeof(struct btrfs_item));
4054
4055 copy_extent_buffer(right, l,
4056 btrfs_leaf_data(right) + BTRFS_LEAF_DATA_SIZE(root) -
4057 data_copy_size, btrfs_leaf_data(l) +
4058 leaf_data_end(root, l), data_copy_size);
4059
4060 rt_data_off = BTRFS_LEAF_DATA_SIZE(root) -
4061 btrfs_item_end_nr(l, mid);
4062
4063 for (i = 0; i < nritems; i++) {
4064 struct btrfs_item *item = btrfs_item_nr(i);
4065 u32 ioff;
4066
4067 ioff = btrfs_token_item_offset(right, item, &token);
4068 btrfs_set_token_item_offset(right, item,
4069 ioff + rt_data_off, &token);
4070 }
4071
4072 btrfs_set_header_nritems(l, mid);
4073 btrfs_item_key(right, &disk_key, 0);
4074 insert_ptr(trans, root, path, &disk_key, right->start,
4075 path->slots[1] + 1, 1);
4076
4077 btrfs_mark_buffer_dirty(right);
4078 btrfs_mark_buffer_dirty(l);
4079 BUG_ON(path->slots[0] != slot);
4080
4081 if (mid <= slot) {
4082 btrfs_tree_unlock(path->nodes[0]);
4083 free_extent_buffer(path->nodes[0]);
4084 path->nodes[0] = right;
4085 path->slots[0] -= mid;
4086 path->slots[1] += 1;
4087 } else {
4088 btrfs_tree_unlock(right);
4089 free_extent_buffer(right);
4090 }
4091
4092 BUG_ON(path->slots[0] < 0);
4093}
4094
4095/*
4096 * double splits happen when we need to insert a big item in the middle
4097 * of a leaf. A double split can leave us with 3 mostly empty leaves:
4098 * leaf: [ slots 0 - N] [ our target ] [ N + 1 - total in leaf ]
4099 * A B C
4100 *
4101 * We avoid this by trying to push the items on either side of our target
4102 * into the adjacent leaves. If all goes well we can avoid the double split
4103 * completely.
4104 */
4105static noinline int push_for_double_split(struct btrfs_trans_handle *trans,
4106 struct btrfs_root *root,
4107 struct btrfs_path *path,
4108 int data_size)
4109{
4110 int ret;
4111 int progress = 0;
4112 int slot;
4113 u32 nritems;
4114 int space_needed = data_size;
4115
4116 slot = path->slots[0];
4117 if (slot < btrfs_header_nritems(path->nodes[0]))
4118 space_needed -= btrfs_leaf_free_space(root, path->nodes[0]);
4119
4120 /*
4121 * try to push all the items after our slot into the
4122 * right leaf
4123 */
4124 ret = push_leaf_right(trans, root, path, 1, space_needed, 0, slot);
4125 if (ret < 0)
4126 return ret;
4127
4128 if (ret == 0)
4129 progress++;
4130
4131 nritems = btrfs_header_nritems(path->nodes[0]);
4132 /*
4133 * our goal is to get our slot at the start or end of a leaf. If
4134 * we've done so we're done
4135 */
4136 if (path->slots[0] == 0 || path->slots[0] == nritems)
4137 return 0;
4138
4139 if (btrfs_leaf_free_space(root, path->nodes[0]) >= data_size)
4140 return 0;
4141
4142 /* try to push all the items before our slot into the next leaf */
4143 slot = path->slots[0];
4144 ret = push_leaf_left(trans, root, path, 1, space_needed, 0, slot);
4145 if (ret < 0)
4146 return ret;
4147
4148 if (ret == 0)
4149 progress++;
4150
4151 if (progress)
4152 return 0;
4153 return 1;
4154}
4155
4156/*
4157 * split the path's leaf in two, making sure there is at least data_size
4158 * available for the resulting leaf level of the path.
4159 *
4160 * returns 0 if all went well and < 0 on failure.
4161 */
4162static noinline int split_leaf(struct btrfs_trans_handle *trans,
4163 struct btrfs_root *root,
4164 struct btrfs_key *ins_key,
4165 struct btrfs_path *path, int data_size,
4166 int extend)
4167{
4168 struct btrfs_disk_key disk_key;
4169 struct extent_buffer *l;
4170 u32 nritems;
4171 int mid;
4172 int slot;
4173 struct extent_buffer *right;
4174 int ret = 0;
4175 int wret;
4176 int split;
4177 int num_doubles = 0;
4178 int tried_avoid_double = 0;
4179
4180 l = path->nodes[0];
4181 slot = path->slots[0];
4182 if (extend && data_size + btrfs_item_size_nr(l, slot) +
4183 sizeof(struct btrfs_item) > BTRFS_LEAF_DATA_SIZE(root))
4184 return -EOVERFLOW;
4185
4186 /* first try to make some room by pushing left and right */
4187 if (data_size && path->nodes[1]) {
4188 int space_needed = data_size;
4189
4190 if (slot < btrfs_header_nritems(l))
4191 space_needed -= btrfs_leaf_free_space(root, l);
4192
4193 wret = push_leaf_right(trans, root, path, space_needed,
4194 space_needed, 0, 0);
4195 if (wret < 0)
4196 return wret;
4197 if (wret) {
4198 wret = push_leaf_left(trans, root, path, space_needed,
4199 space_needed, 0, (u32)-1);
4200 if (wret < 0)
4201 return wret;
4202 }
4203 l = path->nodes[0];
4204
4205 /* did the pushes work? */
4206 if (btrfs_leaf_free_space(root, l) >= data_size)
4207 return 0;
4208 }
4209
4210 if (!path->nodes[1]) {
4211 ret = insert_new_root(trans, root, path, 1);
4212 if (ret)
4213 return ret;
4214 }
4215again:
4216 split = 1;
4217 l = path->nodes[0];
4218 slot = path->slots[0];
4219 nritems = btrfs_header_nritems(l);
4220 mid = (nritems + 1) / 2;
4221
4222 if (mid <= slot) {
4223 if (nritems == 1 ||
4224 leaf_space_used(l, mid, nritems - mid) + data_size >
4225 BTRFS_LEAF_DATA_SIZE(root)) {
4226 if (slot >= nritems) {
4227 split = 0;
4228 } else {
4229 mid = slot;
4230 if (mid != nritems &&
4231 leaf_space_used(l, mid, nritems - mid) +
4232 data_size > BTRFS_LEAF_DATA_SIZE(root)) {
4233 if (data_size && !tried_avoid_double)
4234 goto push_for_double;
4235 split = 2;
4236 }
4237 }
4238 }
4239 } else {
4240 if (leaf_space_used(l, 0, mid) + data_size >
4241 BTRFS_LEAF_DATA_SIZE(root)) {
4242 if (!extend && data_size && slot == 0) {
4243 split = 0;
4244 } else if ((extend || !data_size) && slot == 0) {
4245 mid = 1;
4246 } else {
4247 mid = slot;
4248 if (mid != nritems &&
4249 leaf_space_used(l, mid, nritems - mid) +
4250 data_size > BTRFS_LEAF_DATA_SIZE(root)) {
4251 if (data_size && !tried_avoid_double)
4252 goto push_for_double;
4253 split = 2;
4254 }
4255 }
4256 }
4257 }
4258
4259 if (split == 0)
4260 btrfs_cpu_key_to_disk(&disk_key, ins_key);
4261 else
4262 btrfs_item_key(l, &disk_key, mid);
4263
4264 right = btrfs_alloc_tree_block(trans, root, 0, root->root_key.objectid,
4265 &disk_key, 0, l->start, 0);
4266 if (IS_ERR(right))
4267 return PTR_ERR(right);
4268
4269 root_add_used(root, root->nodesize);
4270
4271 memset_extent_buffer(right, 0, 0, sizeof(struct btrfs_header));
4272 btrfs_set_header_bytenr(right, right->start);
4273 btrfs_set_header_generation(right, trans->transid);
4274 btrfs_set_header_backref_rev(right, BTRFS_MIXED_BACKREF_REV);
4275 btrfs_set_header_owner(right, root->root_key.objectid);
4276 btrfs_set_header_level(right, 0);
4277 write_extent_buffer(right, root->fs_info->fsid,
4278 btrfs_header_fsid(), BTRFS_FSID_SIZE);
4279
4280 write_extent_buffer(right, root->fs_info->chunk_tree_uuid,
4281 btrfs_header_chunk_tree_uuid(right),
4282 BTRFS_UUID_SIZE);
4283
4284 if (split == 0) {
4285 if (mid <= slot) {
4286 btrfs_set_header_nritems(right, 0);
4287 insert_ptr(trans, root, path, &disk_key, right->start,
4288 path->slots[1] + 1, 1);
4289 btrfs_tree_unlock(path->nodes[0]);
4290 free_extent_buffer(path->nodes[0]);
4291 path->nodes[0] = right;
4292 path->slots[0] = 0;
4293 path->slots[1] += 1;
4294 } else {
4295 btrfs_set_header_nritems(right, 0);
4296 insert_ptr(trans, root, path, &disk_key, right->start,
4297 path->slots[1], 1);
4298 btrfs_tree_unlock(path->nodes[0]);
4299 free_extent_buffer(path->nodes[0]);
4300 path->nodes[0] = right;
4301 path->slots[0] = 0;
4302 if (path->slots[1] == 0)
4303 fixup_low_keys(root, path, &disk_key, 1);
4304 }
4305 btrfs_mark_buffer_dirty(right);
4306 return ret;
4307 }
4308
4309 copy_for_split(trans, root, path, l, right, slot, mid, nritems);
4310
4311 if (split == 2) {
4312 BUG_ON(num_doubles != 0);
4313 num_doubles++;
4314 goto again;
4315 }
4316
4317 return 0;
4318
4319push_for_double:
4320 push_for_double_split(trans, root, path, data_size);
4321 tried_avoid_double = 1;
4322 if (btrfs_leaf_free_space(root, path->nodes[0]) >= data_size)
4323 return 0;
4324 goto again;
4325}
4326
4327static noinline int setup_leaf_for_split(struct btrfs_trans_handle *trans,
4328 struct btrfs_root *root,
4329 struct btrfs_path *path, int ins_len)
4330{
4331 struct btrfs_key key;
4332 struct extent_buffer *leaf;
4333 struct btrfs_file_extent_item *fi;
4334 u64 extent_len = 0;
4335 u32 item_size;
4336 int ret;
4337
4338 leaf = path->nodes[0];
4339 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
4340
4341 BUG_ON(key.type != BTRFS_EXTENT_DATA_KEY &&
4342 key.type != BTRFS_EXTENT_CSUM_KEY);
4343
4344 if (btrfs_leaf_free_space(root, leaf) >= ins_len)
4345 return 0;
4346
4347 item_size = btrfs_item_size_nr(leaf, path->slots[0]);
4348 if (key.type == BTRFS_EXTENT_DATA_KEY) {
4349 fi = btrfs_item_ptr(leaf, path->slots[0],
4350 struct btrfs_file_extent_item);
4351 extent_len = btrfs_file_extent_num_bytes(leaf, fi);
4352 }
4353 btrfs_release_path(path);
4354
4355 path->keep_locks = 1;
4356 path->search_for_split = 1;
4357 ret = btrfs_search_slot(trans, root, &key, path, 0, 1);
4358 path->search_for_split = 0;
4359 if (ret < 0)
4360 goto err;
4361
4362 ret = -EAGAIN;
4363 leaf = path->nodes[0];
4364 /* if our item isn't there or got smaller, return now */
4365 if (ret > 0 || item_size != btrfs_item_size_nr(leaf, path->slots[0]))
4366 goto err;
4367
4368 /* the leaf has changed, it now has room. return now */
4369 if (btrfs_leaf_free_space(root, path->nodes[0]) >= ins_len)
4370 goto err;
4371
4372 if (key.type == BTRFS_EXTENT_DATA_KEY) {
4373 fi = btrfs_item_ptr(leaf, path->slots[0],
4374 struct btrfs_file_extent_item);
4375 if (extent_len != btrfs_file_extent_num_bytes(leaf, fi))
4376 goto err;
4377 }
4378
4379 btrfs_set_path_blocking(path);
4380 ret = split_leaf(trans, root, &key, path, ins_len, 1);
4381 if (ret)
4382 goto err;
4383
4384 path->keep_locks = 0;
4385 btrfs_unlock_up_safe(path, 1);
4386 return 0;
4387err:
4388 path->keep_locks = 0;
4389 return ret;
4390}
4391
4392static noinline int split_item(struct btrfs_trans_handle *trans,
4393 struct btrfs_root *root,
4394 struct btrfs_path *path,
4395 struct btrfs_key *new_key,
4396 unsigned long split_offset)
4397{
4398 struct extent_buffer *leaf;
4399 struct btrfs_item *item;
4400 struct btrfs_item *new_item;
4401 int slot;
4402 char *buf;
4403 u32 nritems;
4404 u32 item_size;
4405 u32 orig_offset;
4406 struct btrfs_disk_key disk_key;
4407
4408 leaf = path->nodes[0];
4409 BUG_ON(btrfs_leaf_free_space(root, leaf) < sizeof(struct btrfs_item));
4410
4411 btrfs_set_path_blocking(path);
4412
4413 item = btrfs_item_nr(path->slots[0]);
4414 orig_offset = btrfs_item_offset(leaf, item);
4415 item_size = btrfs_item_size(leaf, item);
4416
4417 buf = kmalloc(item_size, GFP_NOFS);
4418 if (!buf)
4419 return -ENOMEM;
4420
4421 read_extent_buffer(leaf, buf, btrfs_item_ptr_offset(leaf,
4422 path->slots[0]), item_size);
4423
4424 slot = path->slots[0] + 1;
4425 nritems = btrfs_header_nritems(leaf);
4426 if (slot != nritems) {
4427 /* shift the items */
4428 memmove_extent_buffer(leaf, btrfs_item_nr_offset(slot + 1),
4429 btrfs_item_nr_offset(slot),
4430 (nritems - slot) * sizeof(struct btrfs_item));
4431 }
4432
4433 btrfs_cpu_key_to_disk(&disk_key, new_key);
4434 btrfs_set_item_key(leaf, &disk_key, slot);
4435
4436 new_item = btrfs_item_nr(slot);
4437
4438 btrfs_set_item_offset(leaf, new_item, orig_offset);
4439 btrfs_set_item_size(leaf, new_item, item_size - split_offset);
4440
4441 btrfs_set_item_offset(leaf, item,
4442 orig_offset + item_size - split_offset);
4443 btrfs_set_item_size(leaf, item, split_offset);
4444
4445 btrfs_set_header_nritems(leaf, nritems + 1);
4446
4447 /* write the data for the start of the original item */
4448 write_extent_buffer(leaf, buf,
4449 btrfs_item_ptr_offset(leaf, path->slots[0]),
4450 split_offset);
4451
4452 /* write the data for the new item */
4453 write_extent_buffer(leaf, buf + split_offset,
4454 btrfs_item_ptr_offset(leaf, slot),
4455 item_size - split_offset);
4456 btrfs_mark_buffer_dirty(leaf);
4457
4458 BUG_ON(btrfs_leaf_free_space(root, leaf) < 0);
4459 kfree(buf);
4460 return 0;
4461}
4462
4463/*
4464 * This function splits a single item into two items,
4465 * giving 'new_key' to the new item and splitting the
4466 * old one at split_offset (from the start of the item).
4467 *
4468 * The path may be released by this operation. After
4469 * the split, the path is pointing to the old item. The
4470 * new item is going to be in the same node as the old one.
4471 *
4472 * Note, the item being split must be smaller enough to live alone on
4473 * a tree block with room for one extra struct btrfs_item
4474 *
4475 * This allows us to split the item in place, keeping a lock on the
4476 * leaf the entire time.
4477 */
4478int btrfs_split_item(struct btrfs_trans_handle *trans,
4479 struct btrfs_root *root,
4480 struct btrfs_path *path,
4481 struct btrfs_key *new_key,
4482 unsigned long split_offset)
4483{
4484 int ret;
4485 ret = setup_leaf_for_split(trans, root, path,
4486 sizeof(struct btrfs_item));
4487 if (ret)
4488 return ret;
4489
4490 ret = split_item(trans, root, path, new_key, split_offset);
4491 return ret;
4492}
4493
4494/*
4495 * This function duplicate a item, giving 'new_key' to the new item.
4496 * It guarantees both items live in the same tree leaf and the new item
4497 * is contiguous with the original item.
4498 *
4499 * This allows us to split file extent in place, keeping a lock on the
4500 * leaf the entire time.
4501 */
4502int btrfs_duplicate_item(struct btrfs_trans_handle *trans,
4503 struct btrfs_root *root,
4504 struct btrfs_path *path,
4505 struct btrfs_key *new_key)
4506{
4507 struct extent_buffer *leaf;
4508 int ret;
4509 u32 item_size;
4510
4511 leaf = path->nodes[0];
4512 item_size = btrfs_item_size_nr(leaf, path->slots[0]);
4513 ret = setup_leaf_for_split(trans, root, path,
4514 item_size + sizeof(struct btrfs_item));
4515 if (ret)
4516 return ret;
4517
4518 path->slots[0]++;
4519 setup_items_for_insert(root, path, new_key, &item_size,
4520 item_size, item_size +
4521 sizeof(struct btrfs_item), 1);
4522 leaf = path->nodes[0];
4523 memcpy_extent_buffer(leaf,
4524 btrfs_item_ptr_offset(leaf, path->slots[0]),
4525 btrfs_item_ptr_offset(leaf, path->slots[0] - 1),
4526 item_size);
4527 return 0;
4528}
4529
4530/*
4531 * make the item pointed to by the path smaller. new_size indicates
4532 * how small to make it, and from_end tells us if we just chop bytes
4533 * off the end of the item or if we shift the item to chop bytes off
4534 * the front.
4535 */
4536void btrfs_truncate_item(struct btrfs_root *root, struct btrfs_path *path,
4537 u32 new_size, int from_end)
4538{
4539 int slot;
4540 struct extent_buffer *leaf;
4541 struct btrfs_item *item;
4542 u32 nritems;
4543 unsigned int data_end;
4544 unsigned int old_data_start;
4545 unsigned int old_size;
4546 unsigned int size_diff;
4547 int i;
4548 struct btrfs_map_token token;
4549
4550 btrfs_init_map_token(&token);
4551
4552 leaf = path->nodes[0];
4553 slot = path->slots[0];
4554
4555 old_size = btrfs_item_size_nr(leaf, slot);
4556 if (old_size == new_size)
4557 return;
4558
4559 nritems = btrfs_header_nritems(leaf);
4560 data_end = leaf_data_end(root, leaf);
4561
4562 old_data_start = btrfs_item_offset_nr(leaf, slot);
4563
4564 size_diff = old_size - new_size;
4565
4566 BUG_ON(slot < 0);
4567 BUG_ON(slot >= nritems);
4568
4569 /*
4570 * item0..itemN ... dataN.offset..dataN.size .. data0.size
4571 */
4572 /* first correct the data pointers */
4573 for (i = slot; i < nritems; i++) {
4574 u32 ioff;
4575 item = btrfs_item_nr(i);
4576
4577 ioff = btrfs_token_item_offset(leaf, item, &token);
4578 btrfs_set_token_item_offset(leaf, item,
4579 ioff + size_diff, &token);
4580 }
4581
4582 /* shift the data */
4583 if (from_end) {
4584 memmove_extent_buffer(leaf, btrfs_leaf_data(leaf) +
4585 data_end + size_diff, btrfs_leaf_data(leaf) +
4586 data_end, old_data_start + new_size - data_end);
4587 } else {
4588 struct btrfs_disk_key disk_key;
4589 u64 offset;
4590
4591 btrfs_item_key(leaf, &disk_key, slot);
4592
4593 if (btrfs_disk_key_type(&disk_key) == BTRFS_EXTENT_DATA_KEY) {
4594 unsigned long ptr;
4595 struct btrfs_file_extent_item *fi;
4596
4597 fi = btrfs_item_ptr(leaf, slot,
4598 struct btrfs_file_extent_item);
4599 fi = (struct btrfs_file_extent_item *)(
4600 (unsigned long)fi - size_diff);
4601
4602 if (btrfs_file_extent_type(leaf, fi) ==
4603 BTRFS_FILE_EXTENT_INLINE) {
4604 ptr = btrfs_item_ptr_offset(leaf, slot);
4605 memmove_extent_buffer(leaf, ptr,
4606 (unsigned long)fi,
4607 BTRFS_FILE_EXTENT_INLINE_DATA_START);
4608 }
4609 }
4610
4611 memmove_extent_buffer(leaf, btrfs_leaf_data(leaf) +
4612 data_end + size_diff, btrfs_leaf_data(leaf) +
4613 data_end, old_data_start - data_end);
4614
4615 offset = btrfs_disk_key_offset(&disk_key);
4616 btrfs_set_disk_key_offset(&disk_key, offset + size_diff);
4617 btrfs_set_item_key(leaf, &disk_key, slot);
4618 if (slot == 0)
4619 fixup_low_keys(root, path, &disk_key, 1);
4620 }
4621
4622 item = btrfs_item_nr(slot);
4623 btrfs_set_item_size(leaf, item, new_size);
4624 btrfs_mark_buffer_dirty(leaf);
4625
4626 if (btrfs_leaf_free_space(root, leaf) < 0) {
4627 btrfs_print_leaf(root, leaf);
4628 BUG();
4629 }
4630}
4631
4632/*
4633 * make the item pointed to by the path bigger, data_size is the added size.
4634 */
4635void btrfs_extend_item(struct btrfs_root *root, struct btrfs_path *path,
4636 u32 data_size)
4637{
4638 int slot;
4639 struct extent_buffer *leaf;
4640 struct btrfs_item *item;
4641 u32 nritems;
4642 unsigned int data_end;
4643 unsigned int old_data;
4644 unsigned int old_size;
4645 int i;
4646 struct btrfs_map_token token;
4647
4648 btrfs_init_map_token(&token);
4649
4650 leaf = path->nodes[0];
4651
4652 nritems = btrfs_header_nritems(leaf);
4653 data_end = leaf_data_end(root, leaf);
4654
4655 if (btrfs_leaf_free_space(root, leaf) < data_size) {
4656 btrfs_print_leaf(root, leaf);
4657 BUG();
4658 }
4659 slot = path->slots[0];
4660 old_data = btrfs_item_end_nr(leaf, slot);
4661
4662 BUG_ON(slot < 0);
4663 if (slot >= nritems) {
4664 btrfs_print_leaf(root, leaf);
4665 btrfs_crit(root->fs_info, "slot %d too large, nritems %d",
4666 slot, nritems);
4667 BUG_ON(1);
4668 }
4669
4670 /*
4671 * item0..itemN ... dataN.offset..dataN.size .. data0.size
4672 */
4673 /* first correct the data pointers */
4674 for (i = slot; i < nritems; i++) {
4675 u32 ioff;
4676 item = btrfs_item_nr(i);
4677
4678 ioff = btrfs_token_item_offset(leaf, item, &token);
4679 btrfs_set_token_item_offset(leaf, item,
4680 ioff - data_size, &token);
4681 }
4682
4683 /* shift the data */
4684 memmove_extent_buffer(leaf, btrfs_leaf_data(leaf) +
4685 data_end - data_size, btrfs_leaf_data(leaf) +
4686 data_end, old_data - data_end);
4687
4688 data_end = old_data;
4689 old_size = btrfs_item_size_nr(leaf, slot);
4690 item = btrfs_item_nr(slot);
4691 btrfs_set_item_size(leaf, item, old_size + data_size);
4692 btrfs_mark_buffer_dirty(leaf);
4693
4694 if (btrfs_leaf_free_space(root, leaf) < 0) {
4695 btrfs_print_leaf(root, leaf);
4696 BUG();
4697 }
4698}
4699
4700/*
4701 * this is a helper for btrfs_insert_empty_items, the main goal here is
4702 * to save stack depth by doing the bulk of the work in a function
4703 * that doesn't call btrfs_search_slot
4704 */
4705void setup_items_for_insert(struct btrfs_root *root, struct btrfs_path *path,
4706 struct btrfs_key *cpu_key, u32 *data_size,
4707 u32 total_data, u32 total_size, int nr)
4708{
4709 struct btrfs_item *item;
4710 int i;
4711 u32 nritems;
4712 unsigned int data_end;
4713 struct btrfs_disk_key disk_key;
4714 struct extent_buffer *leaf;
4715 int slot;
4716 struct btrfs_map_token token;
4717
4718 if (path->slots[0] == 0) {
4719 btrfs_cpu_key_to_disk(&disk_key, cpu_key);
4720 fixup_low_keys(root, path, &disk_key, 1);
4721 }
4722 btrfs_unlock_up_safe(path, 1);
4723
4724 btrfs_init_map_token(&token);
4725
4726 leaf = path->nodes[0];
4727 slot = path->slots[0];
4728
4729 nritems = btrfs_header_nritems(leaf);
4730 data_end = leaf_data_end(root, leaf);
4731
4732 if (btrfs_leaf_free_space(root, leaf) < total_size) {
4733 btrfs_print_leaf(root, leaf);
4734 btrfs_crit(root->fs_info, "not enough freespace need %u have %d",
4735 total_size, btrfs_leaf_free_space(root, leaf));
4736 BUG();
4737 }
4738
4739 if (slot != nritems) {
4740 unsigned int old_data = btrfs_item_end_nr(leaf, slot);
4741
4742 if (old_data < data_end) {
4743 btrfs_print_leaf(root, leaf);
4744 btrfs_crit(root->fs_info, "slot %d old_data %d data_end %d",
4745 slot, old_data, data_end);
4746 BUG_ON(1);
4747 }
4748 /*
4749 * item0..itemN ... dataN.offset..dataN.size .. data0.size
4750 */
4751 /* first correct the data pointers */
4752 for (i = slot; i < nritems; i++) {
4753 u32 ioff;
4754
4755 item = btrfs_item_nr( i);
4756 ioff = btrfs_token_item_offset(leaf, item, &token);
4757 btrfs_set_token_item_offset(leaf, item,
4758 ioff - total_data, &token);
4759 }
4760 /* shift the items */
4761 memmove_extent_buffer(leaf, btrfs_item_nr_offset(slot + nr),
4762 btrfs_item_nr_offset(slot),
4763 (nritems - slot) * sizeof(struct btrfs_item));
4764
4765 /* shift the data */
4766 memmove_extent_buffer(leaf, btrfs_leaf_data(leaf) +
4767 data_end - total_data, btrfs_leaf_data(leaf) +
4768 data_end, old_data - data_end);
4769 data_end = old_data;
4770 }
4771
4772 /* setup the item for the new data */
4773 for (i = 0; i < nr; i++) {
4774 btrfs_cpu_key_to_disk(&disk_key, cpu_key + i);
4775 btrfs_set_item_key(leaf, &disk_key, slot + i);
4776 item = btrfs_item_nr(slot + i);
4777 btrfs_set_token_item_offset(leaf, item,
4778 data_end - data_size[i], &token);
4779 data_end -= data_size[i];
4780 btrfs_set_token_item_size(leaf, item, data_size[i], &token);
4781 }
4782
4783 btrfs_set_header_nritems(leaf, nritems + nr);
4784 btrfs_mark_buffer_dirty(leaf);
4785
4786 if (btrfs_leaf_free_space(root, leaf) < 0) {
4787 btrfs_print_leaf(root, leaf);
4788 BUG();
4789 }
4790}
4791
4792/*
4793 * Given a key and some data, insert items into the tree.
4794 * This does all the path init required, making room in the tree if needed.
4795 */
4796int btrfs_insert_empty_items(struct btrfs_trans_handle *trans,
4797 struct btrfs_root *root,
4798 struct btrfs_path *path,
4799 struct btrfs_key *cpu_key, u32 *data_size,
4800 int nr)
4801{
4802 int ret = 0;
4803 int slot;
4804 int i;
4805 u32 total_size = 0;
4806 u32 total_data = 0;
4807
4808 for (i = 0; i < nr; i++)
4809 total_data += data_size[i];
4810
4811 total_size = total_data + (nr * sizeof(struct btrfs_item));
4812 ret = btrfs_search_slot(trans, root, cpu_key, path, total_size, 1);
4813 if (ret == 0)
4814 return -EEXIST;
4815 if (ret < 0)
4816 return ret;
4817
4818 slot = path->slots[0];
4819 BUG_ON(slot < 0);
4820
4821 setup_items_for_insert(root, path, cpu_key, data_size,
4822 total_data, total_size, nr);
4823 return 0;
4824}
4825
4826/*
4827 * Given a key and some data, insert an item into the tree.
4828 * This does all the path init required, making room in the tree if needed.
4829 */
4830int btrfs_insert_item(struct btrfs_trans_handle *trans, struct btrfs_root
4831 *root, struct btrfs_key *cpu_key, void *data, u32
4832 data_size)
4833{
4834 int ret = 0;
4835 struct btrfs_path *path;
4836 struct extent_buffer *leaf;
4837 unsigned long ptr;
4838
4839 path = btrfs_alloc_path();
4840 if (!path)
4841 return -ENOMEM;
4842 ret = btrfs_insert_empty_item(trans, root, path, cpu_key, data_size);
4843 if (!ret) {
4844 leaf = path->nodes[0];
4845 ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
4846 write_extent_buffer(leaf, data, ptr, data_size);
4847 btrfs_mark_buffer_dirty(leaf);
4848 }
4849 btrfs_free_path(path);
4850 return ret;
4851}
4852
4853/*
4854 * delete the pointer from a given node.
4855 *
4856 * the tree should have been previously balanced so the deletion does not
4857 * empty a node.
4858 */
4859static void del_ptr(struct btrfs_root *root, struct btrfs_path *path,
4860 int level, int slot)
4861{
4862 struct extent_buffer *parent = path->nodes[level];
4863 u32 nritems;
4864 int ret;
4865
4866 nritems = btrfs_header_nritems(parent);
4867 if (slot != nritems - 1) {
4868 if (level)
4869 tree_mod_log_eb_move(root->fs_info, parent, slot,
4870 slot + 1, nritems - slot - 1);
4871 memmove_extent_buffer(parent,
4872 btrfs_node_key_ptr_offset(slot),
4873 btrfs_node_key_ptr_offset(slot + 1),
4874 sizeof(struct btrfs_key_ptr) *
4875 (nritems - slot - 1));
4876 } else if (level) {
4877 ret = tree_mod_log_insert_key(root->fs_info, parent, slot,
4878 MOD_LOG_KEY_REMOVE, GFP_NOFS);
4879 BUG_ON(ret < 0);
4880 }
4881
4882 nritems--;
4883 btrfs_set_header_nritems(parent, nritems);
4884 if (nritems == 0 && parent == root->node) {
4885 BUG_ON(btrfs_header_level(root->node) != 1);
4886 /* just turn the root into a leaf and break */
4887 btrfs_set_header_level(root->node, 0);
4888 } else if (slot == 0) {
4889 struct btrfs_disk_key disk_key;
4890
4891 btrfs_node_key(parent, &disk_key, 0);
4892 fixup_low_keys(root, path, &disk_key, level + 1);
4893 }
4894 btrfs_mark_buffer_dirty(parent);
4895}
4896
4897/*
4898 * a helper function to delete the leaf pointed to by path->slots[1] and
4899 * path->nodes[1].
4900 *
4901 * This deletes the pointer in path->nodes[1] and frees the leaf
4902 * block extent. zero is returned if it all worked out, < 0 otherwise.
4903 *
4904 * The path must have already been setup for deleting the leaf, including
4905 * all the proper balancing. path->nodes[1] must be locked.
4906 */
4907static noinline void btrfs_del_leaf(struct btrfs_trans_handle *trans,
4908 struct btrfs_root *root,
4909 struct btrfs_path *path,
4910 struct extent_buffer *leaf)
4911{
4912 WARN_ON(btrfs_header_generation(leaf) != trans->transid);
4913 del_ptr(root, path, 1, path->slots[1]);
4914
4915 /*
4916 * btrfs_free_extent is expensive, we want to make sure we
4917 * aren't holding any locks when we call it
4918 */
4919 btrfs_unlock_up_safe(path, 0);
4920
4921 root_sub_used(root, leaf->len);
4922
4923 extent_buffer_get(leaf);
4924 btrfs_free_tree_block(trans, root, leaf, 0, 1);
4925 free_extent_buffer_stale(leaf);
4926}
4927/*
4928 * delete the item at the leaf level in path. If that empties
4929 * the leaf, remove it from the tree
4930 */
4931int btrfs_del_items(struct btrfs_trans_handle *trans, struct btrfs_root *root,
4932 struct btrfs_path *path, int slot, int nr)
4933{
4934 struct extent_buffer *leaf;
4935 struct btrfs_item *item;
4936 int last_off;
4937 int dsize = 0;
4938 int ret = 0;
4939 int wret;
4940 int i;
4941 u32 nritems;
4942 struct btrfs_map_token token;
4943
4944 btrfs_init_map_token(&token);
4945
4946 leaf = path->nodes[0];
4947 last_off = btrfs_item_offset_nr(leaf, slot + nr - 1);
4948
4949 for (i = 0; i < nr; i++)
4950 dsize += btrfs_item_size_nr(leaf, slot + i);
4951
4952 nritems = btrfs_header_nritems(leaf);
4953
4954 if (slot + nr != nritems) {
4955 int data_end = leaf_data_end(root, leaf);
4956
4957 memmove_extent_buffer(leaf, btrfs_leaf_data(leaf) +
4958 data_end + dsize,
4959 btrfs_leaf_data(leaf) + data_end,
4960 last_off - data_end);
4961
4962 for (i = slot + nr; i < nritems; i++) {
4963 u32 ioff;
4964
4965 item = btrfs_item_nr(i);
4966 ioff = btrfs_token_item_offset(leaf, item, &token);
4967 btrfs_set_token_item_offset(leaf, item,
4968 ioff + dsize, &token);
4969 }
4970
4971 memmove_extent_buffer(leaf, btrfs_item_nr_offset(slot),
4972 btrfs_item_nr_offset(slot + nr),
4973 sizeof(struct btrfs_item) *
4974 (nritems - slot - nr));
4975 }
4976 btrfs_set_header_nritems(leaf, nritems - nr);
4977 nritems -= nr;
4978
4979 /* delete the leaf if we've emptied it */
4980 if (nritems == 0) {
4981 if (leaf == root->node) {
4982 btrfs_set_header_level(leaf, 0);
4983 } else {
4984 btrfs_set_path_blocking(path);
4985 clean_tree_block(trans, root, leaf);
4986 btrfs_del_leaf(trans, root, path, leaf);
4987 }
4988 } else {
4989 int used = leaf_space_used(leaf, 0, nritems);
4990 if (slot == 0) {
4991 struct btrfs_disk_key disk_key;
4992
4993 btrfs_item_key(leaf, &disk_key, 0);
4994 fixup_low_keys(root, path, &disk_key, 1);
4995 }
4996
4997 /* delete the leaf if it is mostly empty */
4998 if (used < BTRFS_LEAF_DATA_SIZE(root) / 3) {
4999 /* push_leaf_left fixes the path.
5000 * make sure the path still points to our leaf
5001 * for possible call to del_ptr below
5002 */
5003 slot = path->slots[1];
5004 extent_buffer_get(leaf);
5005
5006 btrfs_set_path_blocking(path);
5007 wret = push_leaf_left(trans, root, path, 1, 1,
5008 1, (u32)-1);
5009 if (wret < 0 && wret != -ENOSPC)
5010 ret = wret;
5011
5012 if (path->nodes[0] == leaf &&
5013 btrfs_header_nritems(leaf)) {
5014 wret = push_leaf_right(trans, root, path, 1,
5015 1, 1, 0);
5016 if (wret < 0 && wret != -ENOSPC)
5017 ret = wret;
5018 }
5019
5020 if (btrfs_header_nritems(leaf) == 0) {
5021 path->slots[1] = slot;
5022 btrfs_del_leaf(trans, root, path, leaf);
5023 free_extent_buffer(leaf);
5024 ret = 0;
5025 } else {
5026 /* if we're still in the path, make sure
5027 * we're dirty. Otherwise, one of the
5028 * push_leaf functions must have already
5029 * dirtied this buffer
5030 */
5031 if (path->nodes[0] == leaf)
5032 btrfs_mark_buffer_dirty(leaf);
5033 free_extent_buffer(leaf);
5034 }
5035 } else {
5036 btrfs_mark_buffer_dirty(leaf);
5037 }
5038 }
5039 return ret;
5040}
5041
5042/*
5043 * search the tree again to find a leaf with lesser keys
5044 * returns 0 if it found something or 1 if there are no lesser leaves.
5045 * returns < 0 on io errors.
5046 *
5047 * This may release the path, and so you may lose any locks held at the
5048 * time you call it.
5049 */
5050int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path)
5051{
5052 struct btrfs_key key;
5053 struct btrfs_disk_key found_key;
5054 int ret;
5055
5056 btrfs_item_key_to_cpu(path->nodes[0], &key, 0);
5057
5058 if (key.offset > 0) {
5059 key.offset--;
5060 } else if (key.type > 0) {
5061 key.type--;
5062 key.offset = (u64)-1;
5063 } else if (key.objectid > 0) {
5064 key.objectid--;
5065 key.type = (u8)-1;
5066 key.offset = (u64)-1;
5067 } else {
5068 return 1;
5069 }
5070
5071 btrfs_release_path(path);
5072 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
5073 if (ret < 0)
5074 return ret;
5075 btrfs_item_key(path->nodes[0], &found_key, 0);
5076 ret = comp_keys(&found_key, &key);
5077 /*
5078 * We might have had an item with the previous key in the tree right
5079 * before we released our path. And after we released our path, that
5080 * item might have been pushed to the first slot (0) of the leaf we
5081 * were holding due to a tree balance. Alternatively, an item with the
5082 * previous key can exist as the only element of a leaf (big fat item).
5083 * Therefore account for these 2 cases, so that our callers (like
5084 * btrfs_previous_item) don't miss an existing item with a key matching
5085 * the previous key we computed above.
5086 */
5087 if (ret <= 0)
5088 return 0;
5089 return 1;
5090}
5091
5092/*
5093 * A helper function to walk down the tree starting at min_key, and looking
5094 * for nodes or leaves that are have a minimum transaction id.
5095 * This is used by the btree defrag code, and tree logging
5096 *
5097 * This does not cow, but it does stuff the starting key it finds back
5098 * into min_key, so you can call btrfs_search_slot with cow=1 on the
5099 * key and get a writable path.
5100 *
5101 * This does lock as it descends, and path->keep_locks should be set
5102 * to 1 by the caller.
5103 *
5104 * This honors path->lowest_level to prevent descent past a given level
5105 * of the tree.
5106 *
5107 * min_trans indicates the oldest transaction that you are interested
5108 * in walking through. Any nodes or leaves older than min_trans are
5109 * skipped over (without reading them).
5110 *
5111 * returns zero if something useful was found, < 0 on error and 1 if there
5112 * was nothing in the tree that matched the search criteria.
5113 */
5114int btrfs_search_forward(struct btrfs_root *root, struct btrfs_key *min_key,
5115 struct btrfs_path *path,
5116 u64 min_trans)
5117{
5118 struct extent_buffer *cur;
5119 struct btrfs_key found_key;
5120 int slot;
5121 int sret;
5122 u32 nritems;
5123 int level;
5124 int ret = 1;
5125 int keep_locks = path->keep_locks;
5126
5127 path->keep_locks = 1;
5128again:
5129 cur = btrfs_read_lock_root_node(root);
5130 level = btrfs_header_level(cur);
5131 WARN_ON(path->nodes[level]);
5132 path->nodes[level] = cur;
5133 path->locks[level] = BTRFS_READ_LOCK;
5134
5135 if (btrfs_header_generation(cur) < min_trans) {
5136 ret = 1;
5137 goto out;
5138 }
5139 while (1) {
5140 nritems = btrfs_header_nritems(cur);
5141 level = btrfs_header_level(cur);
5142 sret = bin_search(cur, min_key, level, &slot);
5143
5144 /* at the lowest level, we're done, setup the path and exit */
5145 if (level == path->lowest_level) {
5146 if (slot >= nritems)
5147 goto find_next_key;
5148 ret = 0;
5149 path->slots[level] = slot;
5150 btrfs_item_key_to_cpu(cur, &found_key, slot);
5151 goto out;
5152 }
5153 if (sret && slot > 0)
5154 slot--;
5155 /*
5156 * check this node pointer against the min_trans parameters.
5157 * If it is too old, old, skip to the next one.
5158 */
5159 while (slot < nritems) {
5160 u64 gen;
5161
5162 gen = btrfs_node_ptr_generation(cur, slot);
5163 if (gen < min_trans) {
5164 slot++;
5165 continue;
5166 }
5167 break;
5168 }
5169find_next_key:
5170 /*
5171 * we didn't find a candidate key in this node, walk forward
5172 * and find another one
5173 */
5174 if (slot >= nritems) {
5175 path->slots[level] = slot;
5176 btrfs_set_path_blocking(path);
5177 sret = btrfs_find_next_key(root, path, min_key, level,
5178 min_trans);
5179 if (sret == 0) {
5180 btrfs_release_path(path);
5181 goto again;
5182 } else {
5183 goto out;
5184 }
5185 }
5186 /* save our key for returning back */
5187 btrfs_node_key_to_cpu(cur, &found_key, slot);
5188 path->slots[level] = slot;
5189 if (level == path->lowest_level) {
5190 ret = 0;
5191 goto out;
5192 }
5193 btrfs_set_path_blocking(path);
5194 cur = read_node_slot(root, cur, slot);
5195 BUG_ON(!cur); /* -ENOMEM */
5196
5197 btrfs_tree_read_lock(cur);
5198
5199 path->locks[level - 1] = BTRFS_READ_LOCK;
5200 path->nodes[level - 1] = cur;
5201 unlock_up(path, level, 1, 0, NULL);
5202 btrfs_clear_path_blocking(path, NULL, 0);
5203 }
5204out:
5205 path->keep_locks = keep_locks;
5206 if (ret == 0) {
5207 btrfs_unlock_up_safe(path, path->lowest_level + 1);
5208 btrfs_set_path_blocking(path);
5209 memcpy(min_key, &found_key, sizeof(found_key));
5210 }
5211 return ret;
5212}
5213
5214static void tree_move_down(struct btrfs_root *root,
5215 struct btrfs_path *path,
5216 int *level, int root_level)
5217{
5218 BUG_ON(*level == 0);
5219 path->nodes[*level - 1] = read_node_slot(root, path->nodes[*level],
5220 path->slots[*level]);
5221 path->slots[*level - 1] = 0;
5222 (*level)--;
5223}
5224
5225static int tree_move_next_or_upnext(struct btrfs_root *root,
5226 struct btrfs_path *path,
5227 int *level, int root_level)
5228{
5229 int ret = 0;
5230 int nritems;
5231 nritems = btrfs_header_nritems(path->nodes[*level]);
5232
5233 path->slots[*level]++;
5234
5235 while (path->slots[*level] >= nritems) {
5236 if (*level == root_level)
5237 return -1;
5238
5239 /* move upnext */
5240 path->slots[*level] = 0;
5241 free_extent_buffer(path->nodes[*level]);
5242 path->nodes[*level] = NULL;
5243 (*level)++;
5244 path->slots[*level]++;
5245
5246 nritems = btrfs_header_nritems(path->nodes[*level]);
5247 ret = 1;
5248 }
5249 return ret;
5250}
5251
5252/*
5253 * Returns 1 if it had to move up and next. 0 is returned if it moved only next
5254 * or down.
5255 */
5256static int tree_advance(struct btrfs_root *root,
5257 struct btrfs_path *path,
5258 int *level, int root_level,
5259 int allow_down,
5260 struct btrfs_key *key)
5261{
5262 int ret;
5263
5264 if (*level == 0 || !allow_down) {
5265 ret = tree_move_next_or_upnext(root, path, level, root_level);
5266 } else {
5267 tree_move_down(root, path, level, root_level);
5268 ret = 0;
5269 }
5270 if (ret >= 0) {
5271 if (*level == 0)
5272 btrfs_item_key_to_cpu(path->nodes[*level], key,
5273 path->slots[*level]);
5274 else
5275 btrfs_node_key_to_cpu(path->nodes[*level], key,
5276 path->slots[*level]);
5277 }
5278 return ret;
5279}
5280
5281static int tree_compare_item(struct btrfs_root *left_root,
5282 struct btrfs_path *left_path,
5283 struct btrfs_path *right_path,
5284 char *tmp_buf)
5285{
5286 int cmp;
5287 int len1, len2;
5288 unsigned long off1, off2;
5289
5290 len1 = btrfs_item_size_nr(left_path->nodes[0], left_path->slots[0]);
5291 len2 = btrfs_item_size_nr(right_path->nodes[0], right_path->slots[0]);
5292 if (len1 != len2)
5293 return 1;
5294
5295 off1 = btrfs_item_ptr_offset(left_path->nodes[0], left_path->slots[0]);
5296 off2 = btrfs_item_ptr_offset(right_path->nodes[0],
5297 right_path->slots[0]);
5298
5299 read_extent_buffer(left_path->nodes[0], tmp_buf, off1, len1);
5300
5301 cmp = memcmp_extent_buffer(right_path->nodes[0], tmp_buf, off2, len1);
5302 if (cmp)
5303 return 1;
5304 return 0;
5305}
5306
5307#define ADVANCE 1
5308#define ADVANCE_ONLY_NEXT -1
5309
5310/*
5311 * This function compares two trees and calls the provided callback for
5312 * every changed/new/deleted item it finds.
5313 * If shared tree blocks are encountered, whole subtrees are skipped, making
5314 * the compare pretty fast on snapshotted subvolumes.
5315 *
5316 * This currently works on commit roots only. As commit roots are read only,
5317 * we don't do any locking. The commit roots are protected with transactions.
5318 * Transactions are ended and rejoined when a commit is tried in between.
5319 *
5320 * This function checks for modifications done to the trees while comparing.
5321 * If it detects a change, it aborts immediately.
5322 */
5323int btrfs_compare_trees(struct btrfs_root *left_root,
5324 struct btrfs_root *right_root,
5325 btrfs_changed_cb_t changed_cb, void *ctx)
5326{
5327 int ret;
5328 int cmp;
5329 struct btrfs_path *left_path = NULL;
5330 struct btrfs_path *right_path = NULL;
5331 struct btrfs_key left_key;
5332 struct btrfs_key right_key;
5333 char *tmp_buf = NULL;
5334 int left_root_level;
5335 int right_root_level;
5336 int left_level;
5337 int right_level;
5338 int left_end_reached;
5339 int right_end_reached;
5340 int advance_left;
5341 int advance_right;
5342 u64 left_blockptr;
5343 u64 right_blockptr;
5344 u64 left_gen;
5345 u64 right_gen;
5346
5347 left_path = btrfs_alloc_path();
5348 if (!left_path) {
5349 ret = -ENOMEM;
5350 goto out;
5351 }
5352 right_path = btrfs_alloc_path();
5353 if (!right_path) {
5354 ret = -ENOMEM;
5355 goto out;
5356 }
5357
5358 tmp_buf = kmalloc(left_root->nodesize, GFP_NOFS);
5359 if (!tmp_buf) {
5360 ret = -ENOMEM;
5361 goto out;
5362 }
5363
5364 left_path->search_commit_root = 1;
5365 left_path->skip_locking = 1;
5366 right_path->search_commit_root = 1;
5367 right_path->skip_locking = 1;
5368
5369 /*
5370 * Strategy: Go to the first items of both trees. Then do
5371 *
5372 * If both trees are at level 0
5373 * Compare keys of current items
5374 * If left < right treat left item as new, advance left tree
5375 * and repeat
5376 * If left > right treat right item as deleted, advance right tree
5377 * and repeat
5378 * If left == right do deep compare of items, treat as changed if
5379 * needed, advance both trees and repeat
5380 * If both trees are at the same level but not at level 0
5381 * Compare keys of current nodes/leafs
5382 * If left < right advance left tree and repeat
5383 * If left > right advance right tree and repeat
5384 * If left == right compare blockptrs of the next nodes/leafs
5385 * If they match advance both trees but stay at the same level
5386 * and repeat
5387 * If they don't match advance both trees while allowing to go
5388 * deeper and repeat
5389 * If tree levels are different
5390 * Advance the tree that needs it and repeat
5391 *
5392 * Advancing a tree means:
5393 * If we are at level 0, try to go to the next slot. If that's not
5394 * possible, go one level up and repeat. Stop when we found a level
5395 * where we could go to the next slot. We may at this point be on a
5396 * node or a leaf.
5397 *
5398 * If we are not at level 0 and not on shared tree blocks, go one
5399 * level deeper.
5400 *
5401 * If we are not at level 0 and on shared tree blocks, go one slot to
5402 * the right if possible or go up and right.
5403 */
5404
5405 down_read(&left_root->fs_info->commit_root_sem);
5406 left_level = btrfs_header_level(left_root->commit_root);
5407 left_root_level = left_level;
5408 left_path->nodes[left_level] = left_root->commit_root;
5409 extent_buffer_get(left_path->nodes[left_level]);
5410
5411 right_level = btrfs_header_level(right_root->commit_root);
5412 right_root_level = right_level;
5413 right_path->nodes[right_level] = right_root->commit_root;
5414 extent_buffer_get(right_path->nodes[right_level]);
5415 up_read(&left_root->fs_info->commit_root_sem);
5416
5417 if (left_level == 0)
5418 btrfs_item_key_to_cpu(left_path->nodes[left_level],
5419 &left_key, left_path->slots[left_level]);
5420 else
5421 btrfs_node_key_to_cpu(left_path->nodes[left_level],
5422 &left_key, left_path->slots[left_level]);
5423 if (right_level == 0)
5424 btrfs_item_key_to_cpu(right_path->nodes[right_level],
5425 &right_key, right_path->slots[right_level]);
5426 else
5427 btrfs_node_key_to_cpu(right_path->nodes[right_level],
5428 &right_key, right_path->slots[right_level]);
5429
5430 left_end_reached = right_end_reached = 0;
5431 advance_left = advance_right = 0;
5432
5433 while (1) {
5434 if (advance_left && !left_end_reached) {
5435 ret = tree_advance(left_root, left_path, &left_level,
5436 left_root_level,
5437 advance_left != ADVANCE_ONLY_NEXT,
5438 &left_key);
5439 if (ret < 0)
5440 left_end_reached = ADVANCE;
5441 advance_left = 0;
5442 }
5443 if (advance_right && !right_end_reached) {
5444 ret = tree_advance(right_root, right_path, &right_level,
5445 right_root_level,
5446 advance_right != ADVANCE_ONLY_NEXT,
5447 &right_key);
5448 if (ret < 0)
5449 right_end_reached = ADVANCE;
5450 advance_right = 0;
5451 }
5452
5453 if (left_end_reached && right_end_reached) {
5454 ret = 0;
5455 goto out;
5456 } else if (left_end_reached) {
5457 if (right_level == 0) {
5458 ret = changed_cb(left_root, right_root,
5459 left_path, right_path,
5460 &right_key,
5461 BTRFS_COMPARE_TREE_DELETED,
5462 ctx);
5463 if (ret < 0)
5464 goto out;
5465 }
5466 advance_right = ADVANCE;
5467 continue;
5468 } else if (right_end_reached) {
5469 if (left_level == 0) {
5470 ret = changed_cb(left_root, right_root,
5471 left_path, right_path,
5472 &left_key,
5473 BTRFS_COMPARE_TREE_NEW,
5474 ctx);
5475 if (ret < 0)
5476 goto out;
5477 }
5478 advance_left = ADVANCE;
5479 continue;
5480 }
5481
5482 if (left_level == 0 && right_level == 0) {
5483 cmp = btrfs_comp_cpu_keys(&left_key, &right_key);
5484 if (cmp < 0) {
5485 ret = changed_cb(left_root, right_root,
5486 left_path, right_path,
5487 &left_key,
5488 BTRFS_COMPARE_TREE_NEW,
5489 ctx);
5490 if (ret < 0)
5491 goto out;
5492 advance_left = ADVANCE;
5493 } else if (cmp > 0) {
5494 ret = changed_cb(left_root, right_root,
5495 left_path, right_path,
5496 &right_key,
5497 BTRFS_COMPARE_TREE_DELETED,
5498 ctx);
5499 if (ret < 0)
5500 goto out;
5501 advance_right = ADVANCE;
5502 } else {
5503 enum btrfs_compare_tree_result result;
5504
5505 WARN_ON(!extent_buffer_uptodate(left_path->nodes[0]));
5506 ret = tree_compare_item(left_root, left_path,
5507 right_path, tmp_buf);
5508 if (ret)
5509 result = BTRFS_COMPARE_TREE_CHANGED;
5510 else
5511 result = BTRFS_COMPARE_TREE_SAME;
5512 ret = changed_cb(left_root, right_root,
5513 left_path, right_path,
5514 &left_key, result, ctx);
5515 if (ret < 0)
5516 goto out;
5517 advance_left = ADVANCE;
5518 advance_right = ADVANCE;
5519 }
5520 } else if (left_level == right_level) {
5521 cmp = btrfs_comp_cpu_keys(&left_key, &right_key);
5522 if (cmp < 0) {
5523 advance_left = ADVANCE;
5524 } else if (cmp > 0) {
5525 advance_right = ADVANCE;
5526 } else {
5527 left_blockptr = btrfs_node_blockptr(
5528 left_path->nodes[left_level],
5529 left_path->slots[left_level]);
5530 right_blockptr = btrfs_node_blockptr(
5531 right_path->nodes[right_level],
5532 right_path->slots[right_level]);
5533 left_gen = btrfs_node_ptr_generation(
5534 left_path->nodes[left_level],
5535 left_path->slots[left_level]);
5536 right_gen = btrfs_node_ptr_generation(
5537 right_path->nodes[right_level],
5538 right_path->slots[right_level]);
5539 if (left_blockptr == right_blockptr &&
5540 left_gen == right_gen) {
5541 /*
5542 * As we're on a shared block, don't
5543 * allow to go deeper.
5544 */
5545 advance_left = ADVANCE_ONLY_NEXT;
5546 advance_right = ADVANCE_ONLY_NEXT;
5547 } else {
5548 advance_left = ADVANCE;
5549 advance_right = ADVANCE;
5550 }
5551 }
5552 } else if (left_level < right_level) {
5553 advance_right = ADVANCE;
5554 } else {
5555 advance_left = ADVANCE;
5556 }
5557 }
5558
5559out:
5560 btrfs_free_path(left_path);
5561 btrfs_free_path(right_path);
5562 kfree(tmp_buf);
5563 return ret;
5564}
5565
5566/*
5567 * this is similar to btrfs_next_leaf, but does not try to preserve
5568 * and fixup the path. It looks for and returns the next key in the
5569 * tree based on the current path and the min_trans parameters.
5570 *
5571 * 0 is returned if another key is found, < 0 if there are any errors
5572 * and 1 is returned if there are no higher keys in the tree
5573 *
5574 * path->keep_locks should be set to 1 on the search made before
5575 * calling this function.
5576 */
5577int btrfs_find_next_key(struct btrfs_root *root, struct btrfs_path *path,
5578 struct btrfs_key *key, int level, u64 min_trans)
5579{
5580 int slot;
5581 struct extent_buffer *c;
5582
5583 WARN_ON(!path->keep_locks);
5584 while (level < BTRFS_MAX_LEVEL) {
5585 if (!path->nodes[level])
5586 return 1;
5587
5588 slot = path->slots[level] + 1;
5589 c = path->nodes[level];
5590next:
5591 if (slot >= btrfs_header_nritems(c)) {
5592 int ret;
5593 int orig_lowest;
5594 struct btrfs_key cur_key;
5595 if (level + 1 >= BTRFS_MAX_LEVEL ||
5596 !path->nodes[level + 1])
5597 return 1;
5598
5599 if (path->locks[level + 1]) {
5600 level++;
5601 continue;
5602 }
5603
5604 slot = btrfs_header_nritems(c) - 1;
5605 if (level == 0)
5606 btrfs_item_key_to_cpu(c, &cur_key, slot);
5607 else
5608 btrfs_node_key_to_cpu(c, &cur_key, slot);
5609
5610 orig_lowest = path->lowest_level;
5611 btrfs_release_path(path);
5612 path->lowest_level = level;
5613 ret = btrfs_search_slot(NULL, root, &cur_key, path,
5614 0, 0);
5615 path->lowest_level = orig_lowest;
5616 if (ret < 0)
5617 return ret;
5618
5619 c = path->nodes[level];
5620 slot = path->slots[level];
5621 if (ret == 0)
5622 slot++;
5623 goto next;
5624 }
5625
5626 if (level == 0)
5627 btrfs_item_key_to_cpu(c, key, slot);
5628 else {
5629 u64 gen = btrfs_node_ptr_generation(c, slot);
5630
5631 if (gen < min_trans) {
5632 slot++;
5633 goto next;
5634 }
5635 btrfs_node_key_to_cpu(c, key, slot);
5636 }
5637 return 0;
5638 }
5639 return 1;
5640}
5641
5642/*
5643 * search the tree again to find a leaf with greater keys
5644 * returns 0 if it found something or 1 if there are no greater leaves.
5645 * returns < 0 on io errors.
5646 */
5647int btrfs_next_leaf(struct btrfs_root *root, struct btrfs_path *path)
5648{
5649 return btrfs_next_old_leaf(root, path, 0);
5650}
5651
5652int btrfs_next_old_leaf(struct btrfs_root *root, struct btrfs_path *path,
5653 u64 time_seq)
5654{
5655 int slot;
5656 int level;
5657 struct extent_buffer *c;
5658 struct extent_buffer *next;
5659 struct btrfs_key key;
5660 u32 nritems;
5661 int ret;
5662 int old_spinning = path->leave_spinning;
5663 int next_rw_lock = 0;
5664
5665 nritems = btrfs_header_nritems(path->nodes[0]);
5666 if (nritems == 0)
5667 return 1;
5668
5669 btrfs_item_key_to_cpu(path->nodes[0], &key, nritems - 1);
5670again:
5671 level = 1;
5672 next = NULL;
5673 next_rw_lock = 0;
5674 btrfs_release_path(path);
5675
5676 path->keep_locks = 1;
5677 path->leave_spinning = 1;
5678
5679 if (time_seq)
5680 ret = btrfs_search_old_slot(root, &key, path, time_seq);
5681 else
5682 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
5683 path->keep_locks = 0;
5684
5685 if (ret < 0)
5686 return ret;
5687
5688 nritems = btrfs_header_nritems(path->nodes[0]);
5689 /*
5690 * by releasing the path above we dropped all our locks. A balance
5691 * could have added more items next to the key that used to be
5692 * at the very end of the block. So, check again here and
5693 * advance the path if there are now more items available.
5694 */
5695 if (nritems > 0 && path->slots[0] < nritems - 1) {
5696 if (ret == 0)
5697 path->slots[0]++;
5698 ret = 0;
5699 goto done;
5700 }
5701 /*
5702 * So the above check misses one case:
5703 * - after releasing the path above, someone has removed the item that
5704 * used to be at the very end of the block, and balance between leafs
5705 * gets another one with bigger key.offset to replace it.
5706 *
5707 * This one should be returned as well, or we can get leaf corruption
5708 * later(esp. in __btrfs_drop_extents()).
5709 *
5710 * And a bit more explanation about this check,
5711 * with ret > 0, the key isn't found, the path points to the slot
5712 * where it should be inserted, so the path->slots[0] item must be the
5713 * bigger one.
5714 */
5715 if (nritems > 0 && ret > 0 && path->slots[0] == nritems - 1) {
5716 ret = 0;
5717 goto done;
5718 }
5719
5720 while (level < BTRFS_MAX_LEVEL) {
5721 if (!path->nodes[level]) {
5722 ret = 1;
5723 goto done;
5724 }
5725
5726 slot = path->slots[level] + 1;
5727 c = path->nodes[level];
5728 if (slot >= btrfs_header_nritems(c)) {
5729 level++;
5730 if (level == BTRFS_MAX_LEVEL) {
5731 ret = 1;
5732 goto done;
5733 }
5734 continue;
5735 }
5736
5737 if (next) {
5738 btrfs_tree_unlock_rw(next, next_rw_lock);
5739 free_extent_buffer(next);
5740 }
5741
5742 next = c;
5743 next_rw_lock = path->locks[level];
5744 ret = read_block_for_search(NULL, root, path, &next, level,
5745 slot, &key, 0);
5746 if (ret == -EAGAIN)
5747 goto again;
5748
5749 if (ret < 0) {
5750 btrfs_release_path(path);
5751 goto done;
5752 }
5753
5754 if (!path->skip_locking) {
5755 ret = btrfs_try_tree_read_lock(next);
5756 if (!ret && time_seq) {
5757 /*
5758 * If we don't get the lock, we may be racing
5759 * with push_leaf_left, holding that lock while
5760 * itself waiting for the leaf we've currently
5761 * locked. To solve this situation, we give up
5762 * on our lock and cycle.
5763 */
5764 free_extent_buffer(next);
5765 btrfs_release_path(path);
5766 cond_resched();
5767 goto again;
5768 }
5769 if (!ret) {
5770 btrfs_set_path_blocking(path);
5771 btrfs_tree_read_lock(next);
5772 btrfs_clear_path_blocking(path, next,
5773 BTRFS_READ_LOCK);
5774 }
5775 next_rw_lock = BTRFS_READ_LOCK;
5776 }
5777 break;
5778 }
5779 path->slots[level] = slot;
5780 while (1) {
5781 level--;
5782 c = path->nodes[level];
5783 if (path->locks[level])
5784 btrfs_tree_unlock_rw(c, path->locks[level]);
5785
5786 free_extent_buffer(c);
5787 path->nodes[level] = next;
5788 path->slots[level] = 0;
5789 if (!path->skip_locking)
5790 path->locks[level] = next_rw_lock;
5791 if (!level)
5792 break;
5793
5794 ret = read_block_for_search(NULL, root, path, &next, level,
5795 0, &key, 0);
5796 if (ret == -EAGAIN)
5797 goto again;
5798
5799 if (ret < 0) {
5800 btrfs_release_path(path);
5801 goto done;
5802 }
5803
5804 if (!path->skip_locking) {
5805 ret = btrfs_try_tree_read_lock(next);
5806 if (!ret) {
5807 btrfs_set_path_blocking(path);
5808 btrfs_tree_read_lock(next);
5809 btrfs_clear_path_blocking(path, next,
5810 BTRFS_READ_LOCK);
5811 }
5812 next_rw_lock = BTRFS_READ_LOCK;
5813 }
5814 }
5815 ret = 0;
5816done:
5817 unlock_up(path, 0, 1, 0, NULL);
5818 path->leave_spinning = old_spinning;
5819 if (!old_spinning)
5820 btrfs_set_path_blocking(path);
5821
5822 return ret;
5823}
5824
5825/*
5826 * this uses btrfs_prev_leaf to walk backwards in the tree, and keeps
5827 * searching until it gets past min_objectid or finds an item of 'type'
5828 *
5829 * returns 0 if something is found, 1 if nothing was found and < 0 on error
5830 */
5831int btrfs_previous_item(struct btrfs_root *root,
5832 struct btrfs_path *path, u64 min_objectid,
5833 int type)
5834{
5835 struct btrfs_key found_key;
5836 struct extent_buffer *leaf;
5837 u32 nritems;
5838 int ret;
5839
5840 while (1) {
5841 if (path->slots[0] == 0) {
5842 btrfs_set_path_blocking(path);
5843 ret = btrfs_prev_leaf(root, path);
5844 if (ret != 0)
5845 return ret;
5846 } else {
5847 path->slots[0]--;
5848 }
5849 leaf = path->nodes[0];
5850 nritems = btrfs_header_nritems(leaf);
5851 if (nritems == 0)
5852 return 1;
5853 if (path->slots[0] == nritems)
5854 path->slots[0]--;
5855
5856 btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
5857 if (found_key.objectid < min_objectid)
5858 break;
5859 if (found_key.type == type)
5860 return 0;
5861 if (found_key.objectid == min_objectid &&
5862 found_key.type < type)
5863 break;
5864 }
5865 return 1;
5866}
5867
5868/*
5869 * search in extent tree to find a previous Metadata/Data extent item with
5870 * min objecitd.
5871 *
5872 * returns 0 if something is found, 1 if nothing was found and < 0 on error
5873 */
5874int btrfs_previous_extent_item(struct btrfs_root *root,
5875 struct btrfs_path *path, u64 min_objectid)
5876{
5877 struct btrfs_key found_key;
5878 struct extent_buffer *leaf;
5879 u32 nritems;
5880 int ret;
5881
5882 while (1) {
5883 if (path->slots[0] == 0) {
5884 btrfs_set_path_blocking(path);
5885 ret = btrfs_prev_leaf(root, path);
5886 if (ret != 0)
5887 return ret;
5888 } else {
5889 path->slots[0]--;
5890 }
5891 leaf = path->nodes[0];
5892 nritems = btrfs_header_nritems(leaf);
5893 if (nritems == 0)
5894 return 1;
5895 if (path->slots[0] == nritems)
5896 path->slots[0]--;
5897
5898 btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
5899 if (found_key.objectid < min_objectid)
5900 break;
5901 if (found_key.type == BTRFS_EXTENT_ITEM_KEY ||
5902 found_key.type == BTRFS_METADATA_ITEM_KEY)
5903 return 0;
5904 if (found_key.objectid == min_objectid &&
5905 found_key.type < BTRFS_EXTENT_ITEM_KEY)
5906 break;
5907 }
5908 return 1;
5909}
Linux kernel - Deliverables
This page took 0.046011 seconds and 5 git commands to generate.