2 * Copyright (C) 2007 Oracle. All rights reserved.
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.
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.
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.
19 #include <linux/sched.h>
22 #include "transaction.h"
23 #include "print-tree.h"
26 static int split_node(struct btrfs_trans_handle *trans, struct btrfs_root
27 *root, struct btrfs_path *path, int level);
28 static int split_leaf(struct btrfs_trans_handle *trans, struct btrfs_root
29 *root, struct btrfs_key *ins_key,
30 struct btrfs_path *path, int data_size, int extend);
31 static int push_node_left(struct btrfs_trans_handle *trans,
32 struct btrfs_root *root, struct extent_buffer *dst,
33 struct extent_buffer *src, int empty);
34 static int balance_node_right(struct btrfs_trans_handle *trans,
35 struct btrfs_root *root,
36 struct extent_buffer *dst_buf,
37 struct extent_buffer *src_buf);
38 static int del_ptr(struct btrfs_trans_handle *trans, struct btrfs_root *root,
39 struct btrfs_path *path, int level, int slot);
41 inline void btrfs_init_path(struct btrfs_path *p)
43 memset(p, 0, sizeof(*p));
46 struct btrfs_path *btrfs_alloc_path(void)
48 struct btrfs_path *path;
49 path = kmem_cache_alloc(btrfs_path_cachep, GFP_NOFS);
51 btrfs_init_path(path);
57 void btrfs_free_path(struct btrfs_path *p)
59 btrfs_release_path(NULL, p);
60 kmem_cache_free(btrfs_path_cachep, p);
63 void btrfs_release_path(struct btrfs_root *root, struct btrfs_path *p)
67 for (i = 0; i < BTRFS_MAX_LEVEL; i++) {
72 btrfs_tree_unlock(p->nodes[i]);
75 free_extent_buffer(p->nodes[i]);
80 struct extent_buffer *btrfs_root_node(struct btrfs_root *root)
82 struct extent_buffer *eb;
83 spin_lock(&root->node_lock);
85 extent_buffer_get(eb);
86 spin_unlock(&root->node_lock);
90 struct extent_buffer *btrfs_lock_root_node(struct btrfs_root *root)
92 struct extent_buffer *eb;
95 eb = btrfs_root_node(root);
98 spin_lock(&root->node_lock);
99 if (eb == root->node) {
100 spin_unlock(&root->node_lock);
103 spin_unlock(&root->node_lock);
105 btrfs_tree_unlock(eb);
106 free_extent_buffer(eb);
111 static void add_root_to_dirty_list(struct btrfs_root *root)
113 if (root->track_dirty && list_empty(&root->dirty_list)) {
114 list_add(&root->dirty_list,
115 &root->fs_info->dirty_cowonly_roots);
119 int btrfs_copy_root(struct btrfs_trans_handle *trans,
120 struct btrfs_root *root,
121 struct extent_buffer *buf,
122 struct extent_buffer **cow_ret, u64 new_root_objectid)
124 struct extent_buffer *cow;
128 struct btrfs_key first_key;
129 struct btrfs_root *new_root;
131 new_root = kmalloc(sizeof(*new_root), GFP_NOFS);
135 memcpy(new_root, root, sizeof(*new_root));
136 new_root->root_key.objectid = new_root_objectid;
138 WARN_ON(root->ref_cows && trans->transid !=
139 root->fs_info->running_transaction->transid);
140 WARN_ON(root->ref_cows && trans->transid != root->last_trans);
142 level = btrfs_header_level(buf);
143 nritems = btrfs_header_nritems(buf);
146 btrfs_item_key_to_cpu(buf, &first_key, 0);
148 btrfs_node_key_to_cpu(buf, &first_key, 0);
150 first_key.objectid = 0;
152 cow = btrfs_alloc_free_block(trans, new_root, buf->len,
154 trans->transid, first_key.objectid,
155 level, buf->start, 0);
161 copy_extent_buffer(cow, buf, 0, 0, cow->len);
162 btrfs_set_header_bytenr(cow, cow->start);
163 btrfs_set_header_generation(cow, trans->transid);
164 btrfs_set_header_owner(cow, new_root_objectid);
165 btrfs_clear_header_flag(cow, BTRFS_HEADER_FLAG_WRITTEN);
167 WARN_ON(btrfs_header_generation(buf) > trans->transid);
168 ret = btrfs_inc_ref(trans, new_root, buf);
174 btrfs_mark_buffer_dirty(cow);
179 int __btrfs_cow_block(struct btrfs_trans_handle *trans,
180 struct btrfs_root *root,
181 struct extent_buffer *buf,
182 struct extent_buffer *parent, int parent_slot,
183 struct extent_buffer **cow_ret,
184 u64 search_start, u64 empty_size)
187 struct extent_buffer *cow;
190 int different_trans = 0;
193 struct btrfs_key first_key;
198 WARN_ON(!btrfs_tree_locked(buf));
200 if (root->ref_cows) {
201 root_gen = trans->transid;
205 WARN_ON(root->ref_cows && trans->transid !=
206 root->fs_info->running_transaction->transid);
207 WARN_ON(root->ref_cows && trans->transid != root->last_trans);
209 level = btrfs_header_level(buf);
210 nritems = btrfs_header_nritems(buf);
213 btrfs_item_key_to_cpu(buf, &first_key, 0);
215 btrfs_node_key_to_cpu(buf, &first_key, 0);
217 first_key.objectid = 0;
219 cow = btrfs_alloc_free_block(trans, root, buf->len,
220 root->root_key.objectid,
221 root_gen, first_key.objectid, level,
222 search_start, empty_size);
226 copy_extent_buffer(cow, buf, 0, 0, cow->len);
227 btrfs_set_header_bytenr(cow, cow->start);
228 btrfs_set_header_generation(cow, trans->transid);
229 btrfs_set_header_owner(cow, root->root_key.objectid);
230 btrfs_clear_header_flag(cow, BTRFS_HEADER_FLAG_WRITTEN);
232 WARN_ON(btrfs_header_generation(buf) > trans->transid);
233 if (btrfs_header_generation(buf) != trans->transid) {
235 ret = btrfs_inc_ref(trans, root, buf);
239 clean_tree_block(trans, root, buf);
242 if (buf == root->node) {
243 WARN_ON(parent && parent != buf);
244 root_gen = btrfs_header_generation(buf);
246 spin_lock(&root->node_lock);
248 extent_buffer_get(cow);
249 spin_unlock(&root->node_lock);
251 if (buf != root->commit_root) {
252 btrfs_free_extent(trans, root, buf->start,
253 buf->len, root->root_key.objectid,
256 free_extent_buffer(buf);
257 add_root_to_dirty_list(root);
259 root_gen = btrfs_header_generation(parent);
260 btrfs_set_node_blockptr(parent, parent_slot,
262 WARN_ON(trans->transid == 0);
263 btrfs_set_node_ptr_generation(parent, parent_slot,
265 btrfs_mark_buffer_dirty(parent);
266 WARN_ON(btrfs_header_generation(parent) != trans->transid);
267 btrfs_free_extent(trans, root, buf->start, buf->len,
268 btrfs_header_owner(parent), root_gen,
272 btrfs_tree_unlock(buf);
273 free_extent_buffer(buf);
274 btrfs_mark_buffer_dirty(cow);
279 int btrfs_cow_block(struct btrfs_trans_handle *trans,
280 struct btrfs_root *root, struct extent_buffer *buf,
281 struct extent_buffer *parent, int parent_slot,
282 struct extent_buffer **cow_ret)
288 if (trans->transaction != root->fs_info->running_transaction) {
289 printk(KERN_CRIT "trans %Lu running %Lu\n", trans->transid,
290 root->fs_info->running_transaction->transid);
293 if (trans->transid != root->fs_info->generation) {
294 printk(KERN_CRIT "trans %Lu running %Lu\n", trans->transid,
295 root->fs_info->generation);
299 header_trans = btrfs_header_generation(buf);
300 spin_lock(&root->fs_info->hash_lock);
301 if (header_trans == trans->transid &&
302 !btrfs_header_flag(buf, BTRFS_HEADER_FLAG_WRITTEN)) {
304 spin_unlock(&root->fs_info->hash_lock);
307 spin_unlock(&root->fs_info->hash_lock);
308 search_start = buf->start & ~((u64)(1024 * 1024 * 1024) - 1);
309 ret = __btrfs_cow_block(trans, root, buf, parent,
310 parent_slot, cow_ret, search_start, 0);
314 static int close_blocks(u64 blocknr, u64 other, u32 blocksize)
316 if (blocknr < other && other - (blocknr + blocksize) < 32768)
318 if (blocknr > other && blocknr - (other + blocksize) < 32768)
324 * compare two keys in a memcmp fashion
326 static int comp_keys(struct btrfs_disk_key *disk, struct btrfs_key *k2)
330 btrfs_disk_key_to_cpu(&k1, disk);
332 if (k1.objectid > k2->objectid)
334 if (k1.objectid < k2->objectid)
336 if (k1.type > k2->type)
338 if (k1.type < k2->type)
340 if (k1.offset > k2->offset)
342 if (k1.offset < k2->offset)
348 int btrfs_realloc_node(struct btrfs_trans_handle *trans,
349 struct btrfs_root *root, struct extent_buffer *parent,
350 int start_slot, int cache_only, u64 *last_ret,
351 struct btrfs_key *progress)
353 struct extent_buffer *cur;
356 u64 search_start = *last_ret;
366 int progress_passed = 0;
367 struct btrfs_disk_key disk_key;
369 parent_level = btrfs_header_level(parent);
370 if (cache_only && parent_level != 1)
373 if (trans->transaction != root->fs_info->running_transaction) {
374 printk(KERN_CRIT "trans %Lu running %Lu\n", trans->transid,
375 root->fs_info->running_transaction->transid);
378 if (trans->transid != root->fs_info->generation) {
379 printk(KERN_CRIT "trans %Lu running %Lu\n", trans->transid,
380 root->fs_info->generation);
384 parent_nritems = btrfs_header_nritems(parent);
385 blocksize = btrfs_level_size(root, parent_level - 1);
386 end_slot = parent_nritems;
388 if (parent_nritems == 1)
391 for (i = start_slot; i < end_slot; i++) {
394 if (!parent->map_token) {
395 map_extent_buffer(parent,
396 btrfs_node_key_ptr_offset(i),
397 sizeof(struct btrfs_key_ptr),
398 &parent->map_token, &parent->kaddr,
399 &parent->map_start, &parent->map_len,
402 btrfs_node_key(parent, &disk_key, i);
403 if (!progress_passed && comp_keys(&disk_key, progress) < 0)
407 blocknr = btrfs_node_blockptr(parent, i);
408 gen = btrfs_node_ptr_generation(parent, i);
410 last_block = blocknr;
413 other = btrfs_node_blockptr(parent, i - 1);
414 close = close_blocks(blocknr, other, blocksize);
416 if (!close && i < end_slot - 2) {
417 other = btrfs_node_blockptr(parent, i + 1);
418 close = close_blocks(blocknr, other, blocksize);
421 last_block = blocknr;
424 if (parent->map_token) {
425 unmap_extent_buffer(parent, parent->map_token,
427 parent->map_token = NULL;
430 cur = btrfs_find_tree_block(root, blocknr, blocksize);
432 uptodate = btrfs_buffer_uptodate(cur, gen);
435 if (!cur || !uptodate) {
437 free_extent_buffer(cur);
441 cur = read_tree_block(root, blocknr,
443 } else if (!uptodate) {
444 btrfs_read_buffer(cur, gen);
447 if (search_start == 0)
448 search_start = last_block;
450 btrfs_tree_lock(cur);
451 err = __btrfs_cow_block(trans, root, cur, parent, i,
454 (end_slot - i) * blocksize));
456 btrfs_tree_unlock(cur);
457 free_extent_buffer(cur);
460 search_start = cur->start;
461 last_block = cur->start;
462 *last_ret = search_start;
463 btrfs_tree_unlock(cur);
464 free_extent_buffer(cur);
466 if (parent->map_token) {
467 unmap_extent_buffer(parent, parent->map_token,
469 parent->map_token = NULL;
475 * The leaf data grows from end-to-front in the node.
476 * this returns the address of the start of the last item,
477 * which is the stop of the leaf data stack
479 static inline unsigned int leaf_data_end(struct btrfs_root *root,
480 struct extent_buffer *leaf)
482 u32 nr = btrfs_header_nritems(leaf);
484 return BTRFS_LEAF_DATA_SIZE(root);
485 return btrfs_item_offset_nr(leaf, nr - 1);
488 static int check_node(struct btrfs_root *root, struct btrfs_path *path,
491 struct extent_buffer *parent = NULL;
492 struct extent_buffer *node = path->nodes[level];
493 struct btrfs_disk_key parent_key;
494 struct btrfs_disk_key node_key;
497 struct btrfs_key cpukey;
498 u32 nritems = btrfs_header_nritems(node);
500 if (path->nodes[level + 1])
501 parent = path->nodes[level + 1];
503 slot = path->slots[level];
504 BUG_ON(nritems == 0);
506 parent_slot = path->slots[level + 1];
507 btrfs_node_key(parent, &parent_key, parent_slot);
508 btrfs_node_key(node, &node_key, 0);
509 BUG_ON(memcmp(&parent_key, &node_key,
510 sizeof(struct btrfs_disk_key)));
511 BUG_ON(btrfs_node_blockptr(parent, parent_slot) !=
512 btrfs_header_bytenr(node));
514 BUG_ON(nritems > BTRFS_NODEPTRS_PER_BLOCK(root));
516 btrfs_node_key_to_cpu(node, &cpukey, slot - 1);
517 btrfs_node_key(node, &node_key, slot);
518 BUG_ON(comp_keys(&node_key, &cpukey) <= 0);
520 if (slot < nritems - 1) {
521 btrfs_node_key_to_cpu(node, &cpukey, slot + 1);
522 btrfs_node_key(node, &node_key, slot);
523 BUG_ON(comp_keys(&node_key, &cpukey) >= 0);
528 static int check_leaf(struct btrfs_root *root, struct btrfs_path *path,
531 struct extent_buffer *leaf = path->nodes[level];
532 struct extent_buffer *parent = NULL;
534 struct btrfs_key cpukey;
535 struct btrfs_disk_key parent_key;
536 struct btrfs_disk_key leaf_key;
537 int slot = path->slots[0];
539 u32 nritems = btrfs_header_nritems(leaf);
541 if (path->nodes[level + 1])
542 parent = path->nodes[level + 1];
548 parent_slot = path->slots[level + 1];
549 btrfs_node_key(parent, &parent_key, parent_slot);
550 btrfs_item_key(leaf, &leaf_key, 0);
552 BUG_ON(memcmp(&parent_key, &leaf_key,
553 sizeof(struct btrfs_disk_key)));
554 BUG_ON(btrfs_node_blockptr(parent, parent_slot) !=
555 btrfs_header_bytenr(leaf));
558 for (i = 0; nritems > 1 && i < nritems - 2; i++) {
559 btrfs_item_key_to_cpu(leaf, &cpukey, i + 1);
560 btrfs_item_key(leaf, &leaf_key, i);
561 if (comp_keys(&leaf_key, &cpukey) >= 0) {
562 btrfs_print_leaf(root, leaf);
563 printk("slot %d offset bad key\n", i);
566 if (btrfs_item_offset_nr(leaf, i) !=
567 btrfs_item_end_nr(leaf, i + 1)) {
568 btrfs_print_leaf(root, leaf);
569 printk("slot %d offset bad\n", i);
573 if (btrfs_item_offset_nr(leaf, i) +
574 btrfs_item_size_nr(leaf, i) !=
575 BTRFS_LEAF_DATA_SIZE(root)) {
576 btrfs_print_leaf(root, leaf);
577 printk("slot %d first offset bad\n", i);
583 if (btrfs_item_size_nr(leaf, nritems - 1) > 4096) {
584 btrfs_print_leaf(root, leaf);
585 printk("slot %d bad size \n", nritems - 1);
590 if (slot != 0 && slot < nritems - 1) {
591 btrfs_item_key(leaf, &leaf_key, slot);
592 btrfs_item_key_to_cpu(leaf, &cpukey, slot - 1);
593 if (comp_keys(&leaf_key, &cpukey) <= 0) {
594 btrfs_print_leaf(root, leaf);
595 printk("slot %d offset bad key\n", slot);
598 if (btrfs_item_offset_nr(leaf, slot - 1) !=
599 btrfs_item_end_nr(leaf, slot)) {
600 btrfs_print_leaf(root, leaf);
601 printk("slot %d offset bad\n", slot);
605 if (slot < nritems - 1) {
606 btrfs_item_key(leaf, &leaf_key, slot);
607 btrfs_item_key_to_cpu(leaf, &cpukey, slot + 1);
608 BUG_ON(comp_keys(&leaf_key, &cpukey) >= 0);
609 if (btrfs_item_offset_nr(leaf, slot) !=
610 btrfs_item_end_nr(leaf, slot + 1)) {
611 btrfs_print_leaf(root, leaf);
612 printk("slot %d offset bad\n", slot);
616 BUG_ON(btrfs_item_offset_nr(leaf, 0) +
617 btrfs_item_size_nr(leaf, 0) != BTRFS_LEAF_DATA_SIZE(root));
621 static int noinline check_block(struct btrfs_root *root,
622 struct btrfs_path *path, int level)
626 if (btrfs_header_level(path->nodes[level]) != level)
627 printk("warning: bad level %Lu wanted %d found %d\n",
628 path->nodes[level]->start, level,
629 btrfs_header_level(path->nodes[level]));
630 found_start = btrfs_header_bytenr(path->nodes[level]);
631 if (found_start != path->nodes[level]->start) {
632 printk("warning: bad bytentr %Lu found %Lu\n",
633 path->nodes[level]->start, found_start);
636 struct extent_buffer *buf = path->nodes[level];
638 if (memcmp_extent_buffer(buf, root->fs_info->fsid,
639 (unsigned long)btrfs_header_fsid(buf),
641 printk("warning bad block %Lu\n", buf->start);
646 return check_leaf(root, path, level);
647 return check_node(root, path, level);
651 * search for key in the extent_buffer. The items start at offset p,
652 * and they are item_size apart. There are 'max' items in p.
654 * the slot in the array is returned via slot, and it points to
655 * the place where you would insert key if it is not found in
658 * slot may point to max if the key is bigger than all of the keys
660 static int generic_bin_search(struct extent_buffer *eb, unsigned long p,
661 int item_size, struct btrfs_key *key,
668 struct btrfs_disk_key *tmp = NULL;
669 struct btrfs_disk_key unaligned;
670 unsigned long offset;
671 char *map_token = NULL;
673 unsigned long map_start = 0;
674 unsigned long map_len = 0;
678 mid = (low + high) / 2;
679 offset = p + mid * item_size;
681 if (!map_token || offset < map_start ||
682 (offset + sizeof(struct btrfs_disk_key)) >
683 map_start + map_len) {
685 unmap_extent_buffer(eb, map_token, KM_USER0);
688 err = map_extent_buffer(eb, offset,
689 sizeof(struct btrfs_disk_key),
691 &map_start, &map_len, KM_USER0);
694 tmp = (struct btrfs_disk_key *)(kaddr + offset -
697 read_extent_buffer(eb, &unaligned,
698 offset, sizeof(unaligned));
703 tmp = (struct btrfs_disk_key *)(kaddr + offset -
706 ret = comp_keys(tmp, key);
715 unmap_extent_buffer(eb, map_token, KM_USER0);
721 unmap_extent_buffer(eb, map_token, KM_USER0);
726 * simple bin_search frontend that does the right thing for
729 static int bin_search(struct extent_buffer *eb, struct btrfs_key *key,
730 int level, int *slot)
733 return generic_bin_search(eb,
734 offsetof(struct btrfs_leaf, items),
735 sizeof(struct btrfs_item),
736 key, btrfs_header_nritems(eb),
739 return generic_bin_search(eb,
740 offsetof(struct btrfs_node, ptrs),
741 sizeof(struct btrfs_key_ptr),
742 key, btrfs_header_nritems(eb),
748 static struct extent_buffer *read_node_slot(struct btrfs_root *root,
749 struct extent_buffer *parent, int slot)
751 int level = btrfs_header_level(parent);
754 if (slot >= btrfs_header_nritems(parent))
759 return read_tree_block(root, btrfs_node_blockptr(parent, slot),
760 btrfs_level_size(root, level - 1),
761 btrfs_node_ptr_generation(parent, slot));
764 static int balance_level(struct btrfs_trans_handle *trans,
765 struct btrfs_root *root,
766 struct btrfs_path *path, int level)
768 struct extent_buffer *right = NULL;
769 struct extent_buffer *mid;
770 struct extent_buffer *left = NULL;
771 struct extent_buffer *parent = NULL;
775 int orig_slot = path->slots[level];
776 int err_on_enospc = 0;
782 mid = path->nodes[level];
783 WARN_ON(!path->locks[level]);
784 WARN_ON(btrfs_header_generation(mid) != trans->transid);
786 orig_ptr = btrfs_node_blockptr(mid, orig_slot);
788 if (level < BTRFS_MAX_LEVEL - 1)
789 parent = path->nodes[level + 1];
790 pslot = path->slots[level + 1];
793 * deal with the case where there is only one pointer in the root
794 * by promoting the node below to a root
797 struct extent_buffer *child;
799 if (btrfs_header_nritems(mid) != 1)
802 /* promote the child to a root */
803 child = read_node_slot(root, mid, 0);
804 btrfs_tree_lock(child);
806 ret = btrfs_cow_block(trans, root, child, mid, 0, &child);
809 spin_lock(&root->node_lock);
811 spin_unlock(&root->node_lock);
813 add_root_to_dirty_list(root);
814 btrfs_tree_unlock(child);
815 path->locks[level] = 0;
816 path->nodes[level] = NULL;
817 clean_tree_block(trans, root, mid);
818 btrfs_tree_unlock(mid);
819 /* once for the path */
820 free_extent_buffer(mid);
821 ret = btrfs_free_extent(trans, root, mid->start, mid->len,
822 root->root_key.objectid,
823 btrfs_header_generation(mid), 0, 0, 1);
824 /* once for the root ptr */
825 free_extent_buffer(mid);
828 if (btrfs_header_nritems(mid) >
829 BTRFS_NODEPTRS_PER_BLOCK(root) / 4)
832 if (btrfs_header_nritems(mid) < 2)
835 left = read_node_slot(root, parent, pslot - 1);
837 btrfs_tree_lock(left);
838 wret = btrfs_cow_block(trans, root, left,
839 parent, pslot - 1, &left);
845 right = read_node_slot(root, parent, pslot + 1);
847 btrfs_tree_lock(right);
848 wret = btrfs_cow_block(trans, root, right,
849 parent, pslot + 1, &right);
856 /* first, try to make some room in the middle buffer */
858 orig_slot += btrfs_header_nritems(left);
859 wret = push_node_left(trans, root, left, mid, 1);
862 if (btrfs_header_nritems(mid) < 2)
867 * then try to empty the right most buffer into the middle
870 wret = push_node_left(trans, root, mid, right, 1);
871 if (wret < 0 && wret != -ENOSPC)
873 if (btrfs_header_nritems(right) == 0) {
874 u64 bytenr = right->start;
875 u64 generation = btrfs_header_generation(parent);
876 u32 blocksize = right->len;
878 clean_tree_block(trans, root, right);
879 btrfs_tree_unlock(right);
880 free_extent_buffer(right);
882 wret = del_ptr(trans, root, path, level + 1, pslot +
886 wret = btrfs_free_extent(trans, root, bytenr,
888 btrfs_header_owner(parent),
889 generation, 0, 0, 1);
893 struct btrfs_disk_key right_key;
894 btrfs_node_key(right, &right_key, 0);
895 btrfs_set_node_key(parent, &right_key, pslot + 1);
896 btrfs_mark_buffer_dirty(parent);
899 if (btrfs_header_nritems(mid) == 1) {
901 * we're not allowed to leave a node with one item in the
902 * tree during a delete. A deletion from lower in the tree
903 * could try to delete the only pointer in this node.
904 * So, pull some keys from the left.
905 * There has to be a left pointer at this point because
906 * otherwise we would have pulled some pointers from the
910 wret = balance_node_right(trans, root, mid, left);
916 wret = push_node_left(trans, root, left, mid, 1);
922 if (btrfs_header_nritems(mid) == 0) {
923 /* we've managed to empty the middle node, drop it */
924 u64 root_gen = btrfs_header_generation(parent);
925 u64 bytenr = mid->start;
926 u32 blocksize = mid->len;
928 clean_tree_block(trans, root, mid);
929 btrfs_tree_unlock(mid);
930 free_extent_buffer(mid);
932 wret = del_ptr(trans, root, path, level + 1, pslot);
935 wret = btrfs_free_extent(trans, root, bytenr, blocksize,
936 btrfs_header_owner(parent),
941 /* update the parent key to reflect our changes */
942 struct btrfs_disk_key mid_key;
943 btrfs_node_key(mid, &mid_key, 0);
944 btrfs_set_node_key(parent, &mid_key, pslot);
945 btrfs_mark_buffer_dirty(parent);
948 /* update the path */
950 if (btrfs_header_nritems(left) > orig_slot) {
951 extent_buffer_get(left);
952 /* left was locked after cow */
953 path->nodes[level] = left;
954 path->slots[level + 1] -= 1;
955 path->slots[level] = orig_slot;
957 btrfs_tree_unlock(mid);
958 free_extent_buffer(mid);
961 orig_slot -= btrfs_header_nritems(left);
962 path->slots[level] = orig_slot;
965 /* double check we haven't messed things up */
966 check_block(root, path, level);
968 btrfs_node_blockptr(path->nodes[level], path->slots[level]))
972 btrfs_tree_unlock(right);
973 free_extent_buffer(right);
976 if (path->nodes[level] != left)
977 btrfs_tree_unlock(left);
978 free_extent_buffer(left);
983 /* returns zero if the push worked, non-zero otherwise */
984 static int noinline push_nodes_for_insert(struct btrfs_trans_handle *trans,
985 struct btrfs_root *root,
986 struct btrfs_path *path, int level)
988 struct extent_buffer *right = NULL;
989 struct extent_buffer *mid;
990 struct extent_buffer *left = NULL;
991 struct extent_buffer *parent = NULL;
995 int orig_slot = path->slots[level];
1001 mid = path->nodes[level];
1002 WARN_ON(btrfs_header_generation(mid) != trans->transid);
1003 orig_ptr = btrfs_node_blockptr(mid, orig_slot);
1005 if (level < BTRFS_MAX_LEVEL - 1)
1006 parent = path->nodes[level + 1];
1007 pslot = path->slots[level + 1];
1012 left = read_node_slot(root, parent, pslot - 1);
1014 /* first, try to make some room in the middle buffer */
1018 btrfs_tree_lock(left);
1019 left_nr = btrfs_header_nritems(left);
1020 if (left_nr >= BTRFS_NODEPTRS_PER_BLOCK(root) - 1) {
1023 ret = btrfs_cow_block(trans, root, left, parent,
1028 wret = push_node_left(trans, root,
1035 struct btrfs_disk_key disk_key;
1036 orig_slot += left_nr;
1037 btrfs_node_key(mid, &disk_key, 0);
1038 btrfs_set_node_key(parent, &disk_key, pslot);
1039 btrfs_mark_buffer_dirty(parent);
1040 if (btrfs_header_nritems(left) > orig_slot) {
1041 path->nodes[level] = left;
1042 path->slots[level + 1] -= 1;
1043 path->slots[level] = orig_slot;
1044 btrfs_tree_unlock(mid);
1045 free_extent_buffer(mid);
1048 btrfs_header_nritems(left);
1049 path->slots[level] = orig_slot;
1050 btrfs_tree_unlock(left);
1051 free_extent_buffer(left);
1055 btrfs_tree_unlock(left);
1056 free_extent_buffer(left);
1058 right = read_node_slot(root, parent, pslot + 1);
1061 * then try to empty the right most buffer into the middle
1065 btrfs_tree_lock(right);
1066 right_nr = btrfs_header_nritems(right);
1067 if (right_nr >= BTRFS_NODEPTRS_PER_BLOCK(root) - 1) {
1070 ret = btrfs_cow_block(trans, root, right,
1076 wret = balance_node_right(trans, root,
1083 struct btrfs_disk_key disk_key;
1085 btrfs_node_key(right, &disk_key, 0);
1086 btrfs_set_node_key(parent, &disk_key, pslot + 1);
1087 btrfs_mark_buffer_dirty(parent);
1089 if (btrfs_header_nritems(mid) <= orig_slot) {
1090 path->nodes[level] = right;
1091 path->slots[level + 1] += 1;
1092 path->slots[level] = orig_slot -
1093 btrfs_header_nritems(mid);
1094 btrfs_tree_unlock(mid);
1095 free_extent_buffer(mid);
1097 btrfs_tree_unlock(right);
1098 free_extent_buffer(right);
1102 btrfs_tree_unlock(right);
1103 free_extent_buffer(right);
1109 * readahead one full node of leaves
1111 static void reada_for_search(struct btrfs_root *root, struct btrfs_path *path,
1112 int level, int slot, u64 objectid)
1114 struct extent_buffer *node;
1115 struct btrfs_disk_key disk_key;
1121 int direction = path->reada;
1122 struct extent_buffer *eb;
1130 if (!path->nodes[level])
1133 node = path->nodes[level];
1135 search = btrfs_node_blockptr(node, slot);
1136 blocksize = btrfs_level_size(root, level - 1);
1137 eb = btrfs_find_tree_block(root, search, blocksize);
1139 free_extent_buffer(eb);
1143 highest_read = search;
1144 lowest_read = search;
1146 nritems = btrfs_header_nritems(node);
1149 if (direction < 0) {
1153 } else if (direction > 0) {
1158 if (path->reada < 0 && objectid) {
1159 btrfs_node_key(node, &disk_key, nr);
1160 if (btrfs_disk_key_objectid(&disk_key) != objectid)
1163 search = btrfs_node_blockptr(node, nr);
1164 if ((search >= lowest_read && search <= highest_read) ||
1165 (search < lowest_read && lowest_read - search <= 32768) ||
1166 (search > highest_read && search - highest_read <= 32768)) {
1167 readahead_tree_block(root, search, blocksize,
1168 btrfs_node_ptr_generation(node, nr));
1172 if (path->reada < 2 && (nread > (256 * 1024) || nscan > 32))
1174 if(nread > (1024 * 1024) || nscan > 128)
1177 if (search < lowest_read)
1178 lowest_read = search;
1179 if (search > highest_read)
1180 highest_read = search;
1184 static void unlock_up(struct btrfs_path *path, int level, int lowest_unlock)
1187 int skip_level = level;
1189 struct extent_buffer *t;
1191 for (i = level; i < BTRFS_MAX_LEVEL; i++) {
1192 if (!path->nodes[i])
1194 if (!path->locks[i])
1196 if (!no_skips && path->slots[i] == 0) {
1200 if (!no_skips && path->keep_locks) {
1203 nritems = btrfs_header_nritems(t);
1204 if (nritems < 1 || path->slots[i] >= nritems - 1) {
1209 if (skip_level < i && i >= lowest_unlock)
1213 if (i >= lowest_unlock && i > skip_level && path->locks[i]) {
1214 btrfs_tree_unlock(t);
1221 * look for key in the tree. path is filled in with nodes along the way
1222 * if key is found, we return zero and you can find the item in the leaf
1223 * level of the path (level 0)
1225 * If the key isn't found, the path points to the slot where it should
1226 * be inserted, and 1 is returned. If there are other errors during the
1227 * search a negative error number is returned.
1229 * if ins_len > 0, nodes and leaves will be split as we walk down the
1230 * tree. if ins_len < 0, nodes will be merged as we walk down the tree (if
1233 int btrfs_search_slot(struct btrfs_trans_handle *trans, struct btrfs_root
1234 *root, struct btrfs_key *key, struct btrfs_path *p, int
1237 struct extent_buffer *b;
1238 struct extent_buffer *tmp;
1242 int should_reada = p->reada;
1243 int lowest_unlock = 1;
1245 u8 lowest_level = 0;
1249 lowest_level = p->lowest_level;
1250 WARN_ON(lowest_level && ins_len);
1251 WARN_ON(p->nodes[0] != NULL);
1252 WARN_ON(cow && root == root->fs_info->extent_root &&
1253 !mutex_is_locked(&root->fs_info->alloc_mutex));
1257 if (p->skip_locking)
1258 b = btrfs_root_node(root);
1260 b = btrfs_lock_root_node(root);
1263 level = btrfs_header_level(b);
1266 wret = btrfs_cow_block(trans, root, b,
1267 p->nodes[level + 1],
1268 p->slots[level + 1],
1271 free_extent_buffer(b);
1275 BUG_ON(!cow && ins_len);
1276 if (level != btrfs_header_level(b))
1278 level = btrfs_header_level(b);
1279 p->nodes[level] = b;
1280 if (!p->skip_locking)
1281 p->locks[level] = 1;
1282 ret = check_block(root, p, level);
1286 ret = bin_search(b, key, level, &slot);
1288 if (ret && slot > 0)
1290 p->slots[level] = slot;
1291 if (ins_len > 0 && btrfs_header_nritems(b) >=
1292 BTRFS_NODEPTRS_PER_BLOCK(root) - 3) {
1293 int sret = split_node(trans, root, p, level);
1297 b = p->nodes[level];
1298 slot = p->slots[level];
1299 } else if (ins_len < 0) {
1300 int sret = balance_level(trans, root, p,
1304 b = p->nodes[level];
1306 btrfs_release_path(NULL, p);
1309 slot = p->slots[level];
1310 BUG_ON(btrfs_header_nritems(b) == 1);
1312 unlock_up(p, level, lowest_unlock);
1314 /* this is only true while dropping a snapshot */
1315 if (level == lowest_level) {
1319 blocknr = btrfs_node_blockptr(b, slot);
1320 gen = btrfs_node_ptr_generation(b, slot);
1321 blocksize = btrfs_level_size(root, level - 1);
1323 tmp = btrfs_find_tree_block(root, blocknr, blocksize);
1324 if (tmp && btrfs_buffer_uptodate(tmp, gen)) {
1328 * reduce lock contention at high levels
1329 * of the btree by dropping locks before
1333 btrfs_release_path(NULL, p);
1335 free_extent_buffer(tmp);
1337 reada_for_search(root, p,
1341 tmp = read_tree_block(root, blocknr,
1344 free_extent_buffer(tmp);
1348 free_extent_buffer(tmp);
1350 reada_for_search(root, p,
1353 b = read_node_slot(root, b, slot);
1356 if (!p->skip_locking)
1359 p->slots[level] = slot;
1360 if (ins_len > 0 && btrfs_leaf_free_space(root, b) <
1361 sizeof(struct btrfs_item) + ins_len) {
1362 int sret = split_leaf(trans, root, key,
1363 p, ins_len, ret == 0);
1368 unlock_up(p, level, lowest_unlock);
1376 * adjust the pointers going up the tree, starting at level
1377 * making sure the right key of each node is points to 'key'.
1378 * This is used after shifting pointers to the left, so it stops
1379 * fixing up pointers when a given leaf/node is not in slot 0 of the
1382 * If this fails to write a tree block, it returns -1, but continues
1383 * fixing up the blocks in ram so the tree is consistent.
1385 static int fixup_low_keys(struct btrfs_trans_handle *trans,
1386 struct btrfs_root *root, struct btrfs_path *path,
1387 struct btrfs_disk_key *key, int level)
1391 struct extent_buffer *t;
1393 for (i = level; i < BTRFS_MAX_LEVEL; i++) {
1394 int tslot = path->slots[i];
1395 if (!path->nodes[i])
1398 btrfs_set_node_key(t, key, tslot);
1399 btrfs_mark_buffer_dirty(path->nodes[i]);
1407 * try to push data from one node into the next node left in the
1410 * returns 0 if some ptrs were pushed left, < 0 if there was some horrible
1411 * error, and > 0 if there was no room in the left hand block.
1413 static int push_node_left(struct btrfs_trans_handle *trans,
1414 struct btrfs_root *root, struct extent_buffer *dst,
1415 struct extent_buffer *src, int empty)
1422 src_nritems = btrfs_header_nritems(src);
1423 dst_nritems = btrfs_header_nritems(dst);
1424 push_items = BTRFS_NODEPTRS_PER_BLOCK(root) - dst_nritems;
1425 WARN_ON(btrfs_header_generation(src) != trans->transid);
1426 WARN_ON(btrfs_header_generation(dst) != trans->transid);
1428 if (!empty && src_nritems <= 8)
1431 if (push_items <= 0) {
1436 push_items = min(src_nritems, push_items);
1437 if (push_items < src_nritems) {
1438 /* leave at least 8 pointers in the node if
1439 * we aren't going to empty it
1441 if (src_nritems - push_items < 8) {
1442 if (push_items <= 8)
1448 push_items = min(src_nritems - 8, push_items);
1450 copy_extent_buffer(dst, src,
1451 btrfs_node_key_ptr_offset(dst_nritems),
1452 btrfs_node_key_ptr_offset(0),
1453 push_items * sizeof(struct btrfs_key_ptr));
1455 if (push_items < src_nritems) {
1456 memmove_extent_buffer(src, btrfs_node_key_ptr_offset(0),
1457 btrfs_node_key_ptr_offset(push_items),
1458 (src_nritems - push_items) *
1459 sizeof(struct btrfs_key_ptr));
1461 btrfs_set_header_nritems(src, src_nritems - push_items);
1462 btrfs_set_header_nritems(dst, dst_nritems + push_items);
1463 btrfs_mark_buffer_dirty(src);
1464 btrfs_mark_buffer_dirty(dst);
1469 * try to push data from one node into the next node right in the
1472 * returns 0 if some ptrs were pushed, < 0 if there was some horrible
1473 * error, and > 0 if there was no room in the right hand block.
1475 * this will only push up to 1/2 the contents of the left node over
1477 static int balance_node_right(struct btrfs_trans_handle *trans,
1478 struct btrfs_root *root,
1479 struct extent_buffer *dst,
1480 struct extent_buffer *src)
1488 WARN_ON(btrfs_header_generation(src) != trans->transid);
1489 WARN_ON(btrfs_header_generation(dst) != trans->transid);
1491 src_nritems = btrfs_header_nritems(src);
1492 dst_nritems = btrfs_header_nritems(dst);
1493 push_items = BTRFS_NODEPTRS_PER_BLOCK(root) - dst_nritems;
1494 if (push_items <= 0) {
1498 if (src_nritems < 4) {
1502 max_push = src_nritems / 2 + 1;
1503 /* don't try to empty the node */
1504 if (max_push >= src_nritems) {
1508 if (max_push < push_items)
1509 push_items = max_push;
1511 memmove_extent_buffer(dst, btrfs_node_key_ptr_offset(push_items),
1512 btrfs_node_key_ptr_offset(0),
1514 sizeof(struct btrfs_key_ptr));
1516 copy_extent_buffer(dst, src,
1517 btrfs_node_key_ptr_offset(0),
1518 btrfs_node_key_ptr_offset(src_nritems - push_items),
1519 push_items * sizeof(struct btrfs_key_ptr));
1521 btrfs_set_header_nritems(src, src_nritems - push_items);
1522 btrfs_set_header_nritems(dst, dst_nritems + push_items);
1524 btrfs_mark_buffer_dirty(src);
1525 btrfs_mark_buffer_dirty(dst);
1530 * helper function to insert a new root level in the tree.
1531 * A new node is allocated, and a single item is inserted to
1532 * point to the existing root
1534 * returns zero on success or < 0 on failure.
1536 static int noinline insert_new_root(struct btrfs_trans_handle *trans,
1537 struct btrfs_root *root,
1538 struct btrfs_path *path, int level)
1542 struct extent_buffer *lower;
1543 struct extent_buffer *c;
1544 struct extent_buffer *old;
1545 struct btrfs_disk_key lower_key;
1547 BUG_ON(path->nodes[level]);
1548 BUG_ON(path->nodes[level-1] != root->node);
1551 root_gen = trans->transid;
1555 lower = path->nodes[level-1];
1557 btrfs_item_key(lower, &lower_key, 0);
1559 btrfs_node_key(lower, &lower_key, 0);
1561 c = btrfs_alloc_free_block(trans, root, root->nodesize,
1562 root->root_key.objectid,
1563 root_gen, lower_key.objectid, level,
1564 root->node->start, 0);
1568 memset_extent_buffer(c, 0, 0, root->nodesize);
1569 btrfs_set_header_nritems(c, 1);
1570 btrfs_set_header_level(c, level);
1571 btrfs_set_header_bytenr(c, c->start);
1572 btrfs_set_header_generation(c, trans->transid);
1573 btrfs_set_header_owner(c, root->root_key.objectid);
1575 write_extent_buffer(c, root->fs_info->fsid,
1576 (unsigned long)btrfs_header_fsid(c),
1579 write_extent_buffer(c, root->fs_info->chunk_tree_uuid,
1580 (unsigned long)btrfs_header_chunk_tree_uuid(c),
1583 btrfs_set_node_key(c, &lower_key, 0);
1584 btrfs_set_node_blockptr(c, 0, lower->start);
1585 lower_gen = btrfs_header_generation(lower);
1586 WARN_ON(lower_gen == 0);
1588 btrfs_set_node_ptr_generation(c, 0, lower_gen);
1590 btrfs_mark_buffer_dirty(c);
1592 spin_lock(&root->node_lock);
1595 spin_unlock(&root->node_lock);
1597 /* the super has an extra ref to root->node */
1598 free_extent_buffer(old);
1600 add_root_to_dirty_list(root);
1601 extent_buffer_get(c);
1602 path->nodes[level] = c;
1603 path->locks[level] = 1;
1604 path->slots[level] = 0;
1606 if (root->ref_cows && lower_gen != trans->transid) {
1607 struct btrfs_path *back_path = btrfs_alloc_path();
1609 mutex_lock(&root->fs_info->alloc_mutex);
1610 ret = btrfs_insert_extent_backref(trans,
1611 root->fs_info->extent_root,
1613 root->root_key.objectid,
1614 trans->transid, 0, 0);
1616 mutex_unlock(&root->fs_info->alloc_mutex);
1617 btrfs_free_path(back_path);
1623 * worker function to insert a single pointer in a node.
1624 * the node should have enough room for the pointer already
1626 * slot and level indicate where you want the key to go, and
1627 * blocknr is the block the key points to.
1629 * returns zero on success and < 0 on any error
1631 static int insert_ptr(struct btrfs_trans_handle *trans, struct btrfs_root
1632 *root, struct btrfs_path *path, struct btrfs_disk_key
1633 *key, u64 bytenr, int slot, int level)
1635 struct extent_buffer *lower;
1638 BUG_ON(!path->nodes[level]);
1639 lower = path->nodes[level];
1640 nritems = btrfs_header_nritems(lower);
1643 if (nritems == BTRFS_NODEPTRS_PER_BLOCK(root))
1645 if (slot != nritems) {
1646 memmove_extent_buffer(lower,
1647 btrfs_node_key_ptr_offset(slot + 1),
1648 btrfs_node_key_ptr_offset(slot),
1649 (nritems - slot) * sizeof(struct btrfs_key_ptr));
1651 btrfs_set_node_key(lower, key, slot);
1652 btrfs_set_node_blockptr(lower, slot, bytenr);
1653 WARN_ON(trans->transid == 0);
1654 btrfs_set_node_ptr_generation(lower, slot, trans->transid);
1655 btrfs_set_header_nritems(lower, nritems + 1);
1656 btrfs_mark_buffer_dirty(lower);
1661 * split the node at the specified level in path in two.
1662 * The path is corrected to point to the appropriate node after the split
1664 * Before splitting this tries to make some room in the node by pushing
1665 * left and right, if either one works, it returns right away.
1667 * returns 0 on success and < 0 on failure
1669 static int split_node(struct btrfs_trans_handle *trans, struct btrfs_root
1670 *root, struct btrfs_path *path, int level)
1673 struct extent_buffer *c;
1674 struct extent_buffer *split;
1675 struct btrfs_disk_key disk_key;
1681 c = path->nodes[level];
1682 WARN_ON(btrfs_header_generation(c) != trans->transid);
1683 if (c == root->node) {
1684 /* trying to split the root, lets make a new one */
1685 ret = insert_new_root(trans, root, path, level + 1);
1689 ret = push_nodes_for_insert(trans, root, path, level);
1690 c = path->nodes[level];
1691 if (!ret && btrfs_header_nritems(c) <
1692 BTRFS_NODEPTRS_PER_BLOCK(root) - 3)
1698 c_nritems = btrfs_header_nritems(c);
1700 root_gen = trans->transid;
1704 btrfs_node_key(c, &disk_key, 0);
1705 split = btrfs_alloc_free_block(trans, root, root->nodesize,
1706 root->root_key.objectid,
1708 btrfs_disk_key_objectid(&disk_key),
1709 level, c->start, 0);
1711 return PTR_ERR(split);
1713 btrfs_set_header_flags(split, btrfs_header_flags(c));
1714 btrfs_set_header_level(split, btrfs_header_level(c));
1715 btrfs_set_header_bytenr(split, split->start);
1716 btrfs_set_header_generation(split, trans->transid);
1717 btrfs_set_header_owner(split, root->root_key.objectid);
1718 btrfs_set_header_flags(split, 0);
1719 write_extent_buffer(split, root->fs_info->fsid,
1720 (unsigned long)btrfs_header_fsid(split),
1722 write_extent_buffer(split, root->fs_info->chunk_tree_uuid,
1723 (unsigned long)btrfs_header_chunk_tree_uuid(split),
1726 mid = (c_nritems + 1) / 2;
1728 copy_extent_buffer(split, c,
1729 btrfs_node_key_ptr_offset(0),
1730 btrfs_node_key_ptr_offset(mid),
1731 (c_nritems - mid) * sizeof(struct btrfs_key_ptr));
1732 btrfs_set_header_nritems(split, c_nritems - mid);
1733 btrfs_set_header_nritems(c, mid);
1736 btrfs_mark_buffer_dirty(c);
1737 btrfs_mark_buffer_dirty(split);
1739 btrfs_node_key(split, &disk_key, 0);
1740 wret = insert_ptr(trans, root, path, &disk_key, split->start,
1741 path->slots[level + 1] + 1,
1746 if (path->slots[level] >= mid) {
1747 path->slots[level] -= mid;
1748 btrfs_tree_unlock(c);
1749 free_extent_buffer(c);
1750 path->nodes[level] = split;
1751 path->slots[level + 1] += 1;
1753 btrfs_tree_unlock(split);
1754 free_extent_buffer(split);
1760 * how many bytes are required to store the items in a leaf. start
1761 * and nr indicate which items in the leaf to check. This totals up the
1762 * space used both by the item structs and the item data
1764 static int leaf_space_used(struct extent_buffer *l, int start, int nr)
1767 int nritems = btrfs_header_nritems(l);
1768 int end = min(nritems, start + nr) - 1;
1772 data_len = btrfs_item_end_nr(l, start);
1773 data_len = data_len - btrfs_item_offset_nr(l, end);
1774 data_len += sizeof(struct btrfs_item) * nr;
1775 WARN_ON(data_len < 0);
1780 * The space between the end of the leaf items and
1781 * the start of the leaf data. IOW, how much room
1782 * the leaf has left for both items and data
1784 int btrfs_leaf_free_space(struct btrfs_root *root, struct extent_buffer *leaf)
1786 int nritems = btrfs_header_nritems(leaf);
1788 ret = BTRFS_LEAF_DATA_SIZE(root) - leaf_space_used(leaf, 0, nritems);
1790 printk("leaf free space ret %d, leaf data size %lu, used %d nritems %d\n",
1791 ret, (unsigned long) BTRFS_LEAF_DATA_SIZE(root),
1792 leaf_space_used(leaf, 0, nritems), nritems);
1798 * push some data in the path leaf to the right, trying to free up at
1799 * least data_size bytes. returns zero if the push worked, nonzero otherwise
1801 * returns 1 if the push failed because the other node didn't have enough
1802 * room, 0 if everything worked out and < 0 if there were major errors.
1804 static int push_leaf_right(struct btrfs_trans_handle *trans, struct btrfs_root
1805 *root, struct btrfs_path *path, int data_size,
1808 struct extent_buffer *left = path->nodes[0];
1809 struct extent_buffer *right;
1810 struct extent_buffer *upper;
1811 struct btrfs_disk_key disk_key;
1817 struct btrfs_item *item;
1825 slot = path->slots[1];
1826 if (!path->nodes[1]) {
1829 upper = path->nodes[1];
1830 if (slot >= btrfs_header_nritems(upper) - 1)
1833 WARN_ON(!btrfs_tree_locked(path->nodes[1]));
1835 right = read_node_slot(root, upper, slot + 1);
1836 btrfs_tree_lock(right);
1837 free_space = btrfs_leaf_free_space(root, right);
1838 if (free_space < data_size + sizeof(struct btrfs_item))
1841 /* cow and double check */
1842 ret = btrfs_cow_block(trans, root, right, upper,
1847 free_space = btrfs_leaf_free_space(root, right);
1848 if (free_space < data_size + sizeof(struct btrfs_item))
1851 left_nritems = btrfs_header_nritems(left);
1852 if (left_nritems == 0)
1860 i = left_nritems - 1;
1862 item = btrfs_item_nr(left, i);
1864 if (path->slots[0] == i)
1865 push_space += data_size + sizeof(*item);
1867 if (!left->map_token) {
1868 map_extent_buffer(left, (unsigned long)item,
1869 sizeof(struct btrfs_item),
1870 &left->map_token, &left->kaddr,
1871 &left->map_start, &left->map_len,
1875 this_item_size = btrfs_item_size(left, item);
1876 if (this_item_size + sizeof(*item) + push_space > free_space)
1879 push_space += this_item_size + sizeof(*item);
1884 if (left->map_token) {
1885 unmap_extent_buffer(left, left->map_token, KM_USER1);
1886 left->map_token = NULL;
1889 if (push_items == 0)
1892 if (!empty && push_items == left_nritems)
1895 /* push left to right */
1896 right_nritems = btrfs_header_nritems(right);
1898 push_space = btrfs_item_end_nr(left, left_nritems - push_items);
1899 push_space -= leaf_data_end(root, left);
1901 /* make room in the right data area */
1902 data_end = leaf_data_end(root, right);
1903 memmove_extent_buffer(right,
1904 btrfs_leaf_data(right) + data_end - push_space,
1905 btrfs_leaf_data(right) + data_end,
1906 BTRFS_LEAF_DATA_SIZE(root) - data_end);
1908 /* copy from the left data area */
1909 copy_extent_buffer(right, left, btrfs_leaf_data(right) +
1910 BTRFS_LEAF_DATA_SIZE(root) - push_space,
1911 btrfs_leaf_data(left) + leaf_data_end(root, left),
1914 memmove_extent_buffer(right, btrfs_item_nr_offset(push_items),
1915 btrfs_item_nr_offset(0),
1916 right_nritems * sizeof(struct btrfs_item));
1918 /* copy the items from left to right */
1919 copy_extent_buffer(right, left, btrfs_item_nr_offset(0),
1920 btrfs_item_nr_offset(left_nritems - push_items),
1921 push_items * sizeof(struct btrfs_item));
1923 /* update the item pointers */
1924 right_nritems += push_items;
1925 btrfs_set_header_nritems(right, right_nritems);
1926 push_space = BTRFS_LEAF_DATA_SIZE(root);
1927 for (i = 0; i < right_nritems; i++) {
1928 item = btrfs_item_nr(right, i);
1929 if (!right->map_token) {
1930 map_extent_buffer(right, (unsigned long)item,
1931 sizeof(struct btrfs_item),
1932 &right->map_token, &right->kaddr,
1933 &right->map_start, &right->map_len,
1936 push_space -= btrfs_item_size(right, item);
1937 btrfs_set_item_offset(right, item, push_space);
1940 if (right->map_token) {
1941 unmap_extent_buffer(right, right->map_token, KM_USER1);
1942 right->map_token = NULL;
1944 left_nritems -= push_items;
1945 btrfs_set_header_nritems(left, left_nritems);
1948 btrfs_mark_buffer_dirty(left);
1949 btrfs_mark_buffer_dirty(right);
1951 btrfs_item_key(right, &disk_key, 0);
1952 btrfs_set_node_key(upper, &disk_key, slot + 1);
1953 btrfs_mark_buffer_dirty(upper);
1955 /* then fixup the leaf pointer in the path */
1956 if (path->slots[0] >= left_nritems) {
1957 path->slots[0] -= left_nritems;
1958 if (btrfs_header_nritems(path->nodes[0]) == 0)
1959 clean_tree_block(trans, root, path->nodes[0]);
1960 btrfs_tree_unlock(path->nodes[0]);
1961 free_extent_buffer(path->nodes[0]);
1962 path->nodes[0] = right;
1963 path->slots[1] += 1;
1965 btrfs_tree_unlock(right);
1966 free_extent_buffer(right);
1971 btrfs_tree_unlock(right);
1972 free_extent_buffer(right);
1977 * push some data in the path leaf to the left, trying to free up at
1978 * least data_size bytes. returns zero if the push worked, nonzero otherwise
1980 static int push_leaf_left(struct btrfs_trans_handle *trans, struct btrfs_root
1981 *root, struct btrfs_path *path, int data_size,
1984 struct btrfs_disk_key disk_key;
1985 struct extent_buffer *right = path->nodes[0];
1986 struct extent_buffer *left;
1992 struct btrfs_item *item;
1993 u32 old_left_nritems;
1999 u32 old_left_item_size;
2001 slot = path->slots[1];
2004 if (!path->nodes[1])
2007 right_nritems = btrfs_header_nritems(right);
2008 if (right_nritems == 0) {
2012 WARN_ON(!btrfs_tree_locked(path->nodes[1]));
2014 left = read_node_slot(root, path->nodes[1], slot - 1);
2015 btrfs_tree_lock(left);
2016 free_space = btrfs_leaf_free_space(root, left);
2017 if (free_space < data_size + sizeof(struct btrfs_item)) {
2022 /* cow and double check */
2023 ret = btrfs_cow_block(trans, root, left,
2024 path->nodes[1], slot - 1, &left);
2026 /* we hit -ENOSPC, but it isn't fatal here */
2031 free_space = btrfs_leaf_free_space(root, left);
2032 if (free_space < data_size + sizeof(struct btrfs_item)) {
2040 nr = right_nritems - 1;
2042 for (i = 0; i < nr; i++) {
2043 item = btrfs_item_nr(right, i);
2044 if (!right->map_token) {
2045 map_extent_buffer(right, (unsigned long)item,
2046 sizeof(struct btrfs_item),
2047 &right->map_token, &right->kaddr,
2048 &right->map_start, &right->map_len,
2052 if (path->slots[0] == i)
2053 push_space += data_size + sizeof(*item);
2055 this_item_size = btrfs_item_size(right, item);
2056 if (this_item_size + sizeof(*item) + push_space > free_space)
2060 push_space += this_item_size + sizeof(*item);
2063 if (right->map_token) {
2064 unmap_extent_buffer(right, right->map_token, KM_USER1);
2065 right->map_token = NULL;
2068 if (push_items == 0) {
2072 if (!empty && push_items == btrfs_header_nritems(right))
2075 /* push data from right to left */
2076 copy_extent_buffer(left, right,
2077 btrfs_item_nr_offset(btrfs_header_nritems(left)),
2078 btrfs_item_nr_offset(0),
2079 push_items * sizeof(struct btrfs_item));
2081 push_space = BTRFS_LEAF_DATA_SIZE(root) -
2082 btrfs_item_offset_nr(right, push_items -1);
2084 copy_extent_buffer(left, right, btrfs_leaf_data(left) +
2085 leaf_data_end(root, left) - push_space,
2086 btrfs_leaf_data(right) +
2087 btrfs_item_offset_nr(right, push_items - 1),
2089 old_left_nritems = btrfs_header_nritems(left);
2090 BUG_ON(old_left_nritems < 0);
2092 old_left_item_size = btrfs_item_offset_nr(left, old_left_nritems - 1);
2093 for (i = old_left_nritems; i < old_left_nritems + push_items; i++) {
2096 item = btrfs_item_nr(left, i);
2097 if (!left->map_token) {
2098 map_extent_buffer(left, (unsigned long)item,
2099 sizeof(struct btrfs_item),
2100 &left->map_token, &left->kaddr,
2101 &left->map_start, &left->map_len,
2105 ioff = btrfs_item_offset(left, item);
2106 btrfs_set_item_offset(left, item,
2107 ioff - (BTRFS_LEAF_DATA_SIZE(root) - old_left_item_size));
2109 btrfs_set_header_nritems(left, old_left_nritems + push_items);
2110 if (left->map_token) {
2111 unmap_extent_buffer(left, left->map_token, KM_USER1);
2112 left->map_token = NULL;
2115 /* fixup right node */
2116 if (push_items > right_nritems) {
2117 printk("push items %d nr %u\n", push_items, right_nritems);
2121 if (push_items < right_nritems) {
2122 push_space = btrfs_item_offset_nr(right, push_items - 1) -
2123 leaf_data_end(root, right);
2124 memmove_extent_buffer(right, btrfs_leaf_data(right) +
2125 BTRFS_LEAF_DATA_SIZE(root) - push_space,
2126 btrfs_leaf_data(right) +
2127 leaf_data_end(root, right), push_space);
2129 memmove_extent_buffer(right, btrfs_item_nr_offset(0),
2130 btrfs_item_nr_offset(push_items),
2131 (btrfs_header_nritems(right) - push_items) *
2132 sizeof(struct btrfs_item));
2134 right_nritems -= push_items;
2135 btrfs_set_header_nritems(right, right_nritems);
2136 push_space = BTRFS_LEAF_DATA_SIZE(root);
2137 for (i = 0; i < right_nritems; i++) {
2138 item = btrfs_item_nr(right, i);
2140 if (!right->map_token) {
2141 map_extent_buffer(right, (unsigned long)item,
2142 sizeof(struct btrfs_item),
2143 &right->map_token, &right->kaddr,
2144 &right->map_start, &right->map_len,
2148 push_space = push_space - btrfs_item_size(right, item);
2149 btrfs_set_item_offset(right, item, push_space);
2151 if (right->map_token) {
2152 unmap_extent_buffer(right, right->map_token, KM_USER1);
2153 right->map_token = NULL;
2156 btrfs_mark_buffer_dirty(left);
2158 btrfs_mark_buffer_dirty(right);
2160 btrfs_item_key(right, &disk_key, 0);
2161 wret = fixup_low_keys(trans, root, path, &disk_key, 1);
2165 /* then fixup the leaf pointer in the path */
2166 if (path->slots[0] < push_items) {
2167 path->slots[0] += old_left_nritems;
2168 if (btrfs_header_nritems(path->nodes[0]) == 0)
2169 clean_tree_block(trans, root, path->nodes[0]);
2170 btrfs_tree_unlock(path->nodes[0]);
2171 free_extent_buffer(path->nodes[0]);
2172 path->nodes[0] = left;
2173 path->slots[1] -= 1;
2175 btrfs_tree_unlock(left);
2176 free_extent_buffer(left);
2177 path->slots[0] -= push_items;
2179 BUG_ON(path->slots[0] < 0);
2182 btrfs_tree_unlock(left);
2183 free_extent_buffer(left);
2188 * split the path's leaf in two, making sure there is at least data_size
2189 * available for the resulting leaf level of the path.
2191 * returns 0 if all went well and < 0 on failure.
2193 static int split_leaf(struct btrfs_trans_handle *trans, struct btrfs_root
2194 *root, struct btrfs_key *ins_key,
2195 struct btrfs_path *path, int data_size, int extend)
2198 struct extent_buffer *l;
2202 struct extent_buffer *right;
2203 int space_needed = data_size + sizeof(struct btrfs_item);
2210 int num_doubles = 0;
2211 struct btrfs_disk_key disk_key;
2214 space_needed = data_size;
2217 root_gen = trans->transid;
2221 /* first try to make some room by pushing left and right */
2222 if (ins_key->type != BTRFS_DIR_ITEM_KEY) {
2223 wret = push_leaf_right(trans, root, path, data_size, 0);
2228 wret = push_leaf_left(trans, root, path, data_size, 0);
2234 /* did the pushes work? */
2235 if (btrfs_leaf_free_space(root, l) >= space_needed)
2239 if (!path->nodes[1]) {
2240 ret = insert_new_root(trans, root, path, 1);
2247 slot = path->slots[0];
2248 nritems = btrfs_header_nritems(l);
2249 mid = (nritems + 1)/ 2;
2251 btrfs_item_key(l, &disk_key, 0);
2253 right = btrfs_alloc_free_block(trans, root, root->leafsize,
2254 root->root_key.objectid,
2255 root_gen, disk_key.objectid, 0,
2257 if (IS_ERR(right)) {
2259 return PTR_ERR(right);
2262 memset_extent_buffer(right, 0, 0, sizeof(struct btrfs_header));
2263 btrfs_set_header_bytenr(right, right->start);
2264 btrfs_set_header_generation(right, trans->transid);
2265 btrfs_set_header_owner(right, root->root_key.objectid);
2266 btrfs_set_header_level(right, 0);
2267 write_extent_buffer(right, root->fs_info->fsid,
2268 (unsigned long)btrfs_header_fsid(right),
2271 write_extent_buffer(right, root->fs_info->chunk_tree_uuid,
2272 (unsigned long)btrfs_header_chunk_tree_uuid(right),
2276 leaf_space_used(l, mid, nritems - mid) + space_needed >
2277 BTRFS_LEAF_DATA_SIZE(root)) {
2278 if (slot >= nritems) {
2279 btrfs_cpu_key_to_disk(&disk_key, ins_key);
2280 btrfs_set_header_nritems(right, 0);
2281 wret = insert_ptr(trans, root, path,
2282 &disk_key, right->start,
2283 path->slots[1] + 1, 1);
2287 btrfs_tree_unlock(path->nodes[0]);
2288 free_extent_buffer(path->nodes[0]);
2289 path->nodes[0] = right;
2291 path->slots[1] += 1;
2292 btrfs_mark_buffer_dirty(right);
2296 if (mid != nritems &&
2297 leaf_space_used(l, mid, nritems - mid) +
2298 space_needed > BTRFS_LEAF_DATA_SIZE(root)) {
2303 if (leaf_space_used(l, 0, mid + 1) + space_needed >
2304 BTRFS_LEAF_DATA_SIZE(root)) {
2305 if (!extend && slot == 0) {
2306 btrfs_cpu_key_to_disk(&disk_key, ins_key);
2307 btrfs_set_header_nritems(right, 0);
2308 wret = insert_ptr(trans, root, path,
2314 btrfs_tree_unlock(path->nodes[0]);
2315 free_extent_buffer(path->nodes[0]);
2316 path->nodes[0] = right;
2318 if (path->slots[1] == 0) {
2319 wret = fixup_low_keys(trans, root,
2320 path, &disk_key, 1);
2324 btrfs_mark_buffer_dirty(right);
2326 } else if (extend && slot == 0) {
2330 if (mid != nritems &&
2331 leaf_space_used(l, mid, nritems - mid) +
2332 space_needed > BTRFS_LEAF_DATA_SIZE(root)) {
2338 nritems = nritems - mid;
2339 btrfs_set_header_nritems(right, nritems);
2340 data_copy_size = btrfs_item_end_nr(l, mid) - leaf_data_end(root, l);
2342 copy_extent_buffer(right, l, btrfs_item_nr_offset(0),
2343 btrfs_item_nr_offset(mid),
2344 nritems * sizeof(struct btrfs_item));
2346 copy_extent_buffer(right, l,
2347 btrfs_leaf_data(right) + BTRFS_LEAF_DATA_SIZE(root) -
2348 data_copy_size, btrfs_leaf_data(l) +
2349 leaf_data_end(root, l), data_copy_size);
2351 rt_data_off = BTRFS_LEAF_DATA_SIZE(root) -
2352 btrfs_item_end_nr(l, mid);
2354 for (i = 0; i < nritems; i++) {
2355 struct btrfs_item *item = btrfs_item_nr(right, i);
2358 if (!right->map_token) {
2359 map_extent_buffer(right, (unsigned long)item,
2360 sizeof(struct btrfs_item),
2361 &right->map_token, &right->kaddr,
2362 &right->map_start, &right->map_len,
2366 ioff = btrfs_item_offset(right, item);
2367 btrfs_set_item_offset(right, item, ioff + rt_data_off);
2370 if (right->map_token) {
2371 unmap_extent_buffer(right, right->map_token, KM_USER1);
2372 right->map_token = NULL;
2375 btrfs_set_header_nritems(l, mid);
2377 btrfs_item_key(right, &disk_key, 0);
2378 wret = insert_ptr(trans, root, path, &disk_key, right->start,
2379 path->slots[1] + 1, 1);
2383 btrfs_mark_buffer_dirty(right);
2384 btrfs_mark_buffer_dirty(l);
2385 BUG_ON(path->slots[0] != slot);
2388 btrfs_tree_unlock(path->nodes[0]);
2389 free_extent_buffer(path->nodes[0]);
2390 path->nodes[0] = right;
2391 path->slots[0] -= mid;
2392 path->slots[1] += 1;
2394 btrfs_tree_unlock(right);
2395 free_extent_buffer(right);
2398 BUG_ON(path->slots[0] < 0);
2401 BUG_ON(num_doubles != 0);
2408 int btrfs_truncate_item(struct btrfs_trans_handle *trans,
2409 struct btrfs_root *root,
2410 struct btrfs_path *path,
2411 u32 new_size, int from_end)
2416 struct extent_buffer *leaf;
2417 struct btrfs_item *item;
2419 unsigned int data_end;
2420 unsigned int old_data_start;
2421 unsigned int old_size;
2422 unsigned int size_diff;
2425 slot_orig = path->slots[0];
2426 leaf = path->nodes[0];
2427 slot = path->slots[0];
2429 old_size = btrfs_item_size_nr(leaf, slot);
2430 if (old_size == new_size)
2433 nritems = btrfs_header_nritems(leaf);
2434 data_end = leaf_data_end(root, leaf);
2436 old_data_start = btrfs_item_offset_nr(leaf, slot);
2438 size_diff = old_size - new_size;
2441 BUG_ON(slot >= nritems);
2444 * item0..itemN ... dataN.offset..dataN.size .. data0.size
2446 /* first correct the data pointers */
2447 for (i = slot; i < nritems; i++) {
2449 item = btrfs_item_nr(leaf, i);
2451 if (!leaf->map_token) {
2452 map_extent_buffer(leaf, (unsigned long)item,
2453 sizeof(struct btrfs_item),
2454 &leaf->map_token, &leaf->kaddr,
2455 &leaf->map_start, &leaf->map_len,
2459 ioff = btrfs_item_offset(leaf, item);
2460 btrfs_set_item_offset(leaf, item, ioff + size_diff);
2463 if (leaf->map_token) {
2464 unmap_extent_buffer(leaf, leaf->map_token, KM_USER1);
2465 leaf->map_token = NULL;
2468 /* shift the data */
2470 memmove_extent_buffer(leaf, btrfs_leaf_data(leaf) +
2471 data_end + size_diff, btrfs_leaf_data(leaf) +
2472 data_end, old_data_start + new_size - data_end);
2474 struct btrfs_disk_key disk_key;
2477 btrfs_item_key(leaf, &disk_key, slot);
2479 if (btrfs_disk_key_type(&disk_key) == BTRFS_EXTENT_DATA_KEY) {
2481 struct btrfs_file_extent_item *fi;
2483 fi = btrfs_item_ptr(leaf, slot,
2484 struct btrfs_file_extent_item);
2485 fi = (struct btrfs_file_extent_item *)(
2486 (unsigned long)fi - size_diff);
2488 if (btrfs_file_extent_type(leaf, fi) ==
2489 BTRFS_FILE_EXTENT_INLINE) {
2490 ptr = btrfs_item_ptr_offset(leaf, slot);
2491 memmove_extent_buffer(leaf, ptr,
2493 offsetof(struct btrfs_file_extent_item,
2498 memmove_extent_buffer(leaf, btrfs_leaf_data(leaf) +
2499 data_end + size_diff, btrfs_leaf_data(leaf) +
2500 data_end, old_data_start - data_end);
2502 offset = btrfs_disk_key_offset(&disk_key);
2503 btrfs_set_disk_key_offset(&disk_key, offset + size_diff);
2504 btrfs_set_item_key(leaf, &disk_key, slot);
2506 fixup_low_keys(trans, root, path, &disk_key, 1);
2509 item = btrfs_item_nr(leaf, slot);
2510 btrfs_set_item_size(leaf, item, new_size);
2511 btrfs_mark_buffer_dirty(leaf);
2514 if (btrfs_leaf_free_space(root, leaf) < 0) {
2515 btrfs_print_leaf(root, leaf);
2521 int btrfs_extend_item(struct btrfs_trans_handle *trans,
2522 struct btrfs_root *root, struct btrfs_path *path,
2528 struct extent_buffer *leaf;
2529 struct btrfs_item *item;
2531 unsigned int data_end;
2532 unsigned int old_data;
2533 unsigned int old_size;
2536 slot_orig = path->slots[0];
2537 leaf = path->nodes[0];
2539 nritems = btrfs_header_nritems(leaf);
2540 data_end = leaf_data_end(root, leaf);
2542 if (btrfs_leaf_free_space(root, leaf) < data_size) {
2543 btrfs_print_leaf(root, leaf);
2546 slot = path->slots[0];
2547 old_data = btrfs_item_end_nr(leaf, slot);
2550 if (slot >= nritems) {
2551 btrfs_print_leaf(root, leaf);
2552 printk("slot %d too large, nritems %d\n", slot, nritems);
2557 * item0..itemN ... dataN.offset..dataN.size .. data0.size
2559 /* first correct the data pointers */
2560 for (i = slot; i < nritems; i++) {
2562 item = btrfs_item_nr(leaf, i);
2564 if (!leaf->map_token) {
2565 map_extent_buffer(leaf, (unsigned long)item,
2566 sizeof(struct btrfs_item),
2567 &leaf->map_token, &leaf->kaddr,
2568 &leaf->map_start, &leaf->map_len,
2571 ioff = btrfs_item_offset(leaf, item);
2572 btrfs_set_item_offset(leaf, item, ioff - data_size);
2575 if (leaf->map_token) {
2576 unmap_extent_buffer(leaf, leaf->map_token, KM_USER1);
2577 leaf->map_token = NULL;
2580 /* shift the data */
2581 memmove_extent_buffer(leaf, btrfs_leaf_data(leaf) +
2582 data_end - data_size, btrfs_leaf_data(leaf) +
2583 data_end, old_data - data_end);
2585 data_end = old_data;
2586 old_size = btrfs_item_size_nr(leaf, slot);
2587 item = btrfs_item_nr(leaf, slot);
2588 btrfs_set_item_size(leaf, item, old_size + data_size);
2589 btrfs_mark_buffer_dirty(leaf);
2592 if (btrfs_leaf_free_space(root, leaf) < 0) {
2593 btrfs_print_leaf(root, leaf);
2600 * Given a key and some data, insert an item into the tree.
2601 * This does all the path init required, making room in the tree if needed.
2603 int btrfs_insert_empty_items(struct btrfs_trans_handle *trans,
2604 struct btrfs_root *root,
2605 struct btrfs_path *path,
2606 struct btrfs_key *cpu_key, u32 *data_size,
2609 struct extent_buffer *leaf;
2610 struct btrfs_item *item;
2618 unsigned int data_end;
2619 struct btrfs_disk_key disk_key;
2621 for (i = 0; i < nr; i++) {
2622 total_data += data_size[i];
2625 total_size = total_data + (nr * sizeof(struct btrfs_item));
2626 ret = btrfs_search_slot(trans, root, cpu_key, path, total_size, 1);
2633 slot_orig = path->slots[0];
2634 leaf = path->nodes[0];
2636 nritems = btrfs_header_nritems(leaf);
2637 data_end = leaf_data_end(root, leaf);
2639 if (btrfs_leaf_free_space(root, leaf) <
2640 sizeof(struct btrfs_item) + total_size) {
2641 btrfs_print_leaf(root, leaf);
2642 printk("not enough freespace need %u have %d\n",
2643 total_size, btrfs_leaf_free_space(root, leaf));
2647 slot = path->slots[0];
2650 if (slot != nritems) {
2652 unsigned int old_data = btrfs_item_end_nr(leaf, slot);
2654 if (old_data < data_end) {
2655 btrfs_print_leaf(root, leaf);
2656 printk("slot %d old_data %d data_end %d\n",
2657 slot, old_data, data_end);
2661 * item0..itemN ... dataN.offset..dataN.size .. data0.size
2663 /* first correct the data pointers */
2664 WARN_ON(leaf->map_token);
2665 for (i = slot; i < nritems; i++) {
2668 item = btrfs_item_nr(leaf, i);
2669 if (!leaf->map_token) {
2670 map_extent_buffer(leaf, (unsigned long)item,
2671 sizeof(struct btrfs_item),
2672 &leaf->map_token, &leaf->kaddr,
2673 &leaf->map_start, &leaf->map_len,
2677 ioff = btrfs_item_offset(leaf, item);
2678 btrfs_set_item_offset(leaf, item, ioff - total_data);
2680 if (leaf->map_token) {
2681 unmap_extent_buffer(leaf, leaf->map_token, KM_USER1);
2682 leaf->map_token = NULL;
2685 /* shift the items */
2686 memmove_extent_buffer(leaf, btrfs_item_nr_offset(slot + nr),
2687 btrfs_item_nr_offset(slot),
2688 (nritems - slot) * sizeof(struct btrfs_item));
2690 /* shift the data */
2691 memmove_extent_buffer(leaf, btrfs_leaf_data(leaf) +
2692 data_end - total_data, btrfs_leaf_data(leaf) +
2693 data_end, old_data - data_end);
2694 data_end = old_data;
2697 /* setup the item for the new data */
2698 for (i = 0; i < nr; i++) {
2699 btrfs_cpu_key_to_disk(&disk_key, cpu_key + i);
2700 btrfs_set_item_key(leaf, &disk_key, slot + i);
2701 item = btrfs_item_nr(leaf, slot + i);
2702 btrfs_set_item_offset(leaf, item, data_end - data_size[i]);
2703 data_end -= data_size[i];
2704 btrfs_set_item_size(leaf, item, data_size[i]);
2706 btrfs_set_header_nritems(leaf, nritems + nr);
2707 btrfs_mark_buffer_dirty(leaf);
2711 btrfs_cpu_key_to_disk(&disk_key, cpu_key);
2712 ret = fixup_low_keys(trans, root, path, &disk_key, 1);
2715 if (btrfs_leaf_free_space(root, leaf) < 0) {
2716 btrfs_print_leaf(root, leaf);
2724 * Given a key and some data, insert an item into the tree.
2725 * This does all the path init required, making room in the tree if needed.
2727 int btrfs_insert_item(struct btrfs_trans_handle *trans, struct btrfs_root
2728 *root, struct btrfs_key *cpu_key, void *data, u32
2732 struct btrfs_path *path;
2733 struct extent_buffer *leaf;
2736 path = btrfs_alloc_path();
2738 ret = btrfs_insert_empty_item(trans, root, path, cpu_key, data_size);
2740 leaf = path->nodes[0];
2741 ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
2742 write_extent_buffer(leaf, data, ptr, data_size);
2743 btrfs_mark_buffer_dirty(leaf);
2745 btrfs_free_path(path);
2750 * delete the pointer from a given node.
2752 * If the delete empties a node, the node is removed from the tree,
2753 * continuing all the way the root if required. The root is converted into
2754 * a leaf if all the nodes are emptied.
2756 static int del_ptr(struct btrfs_trans_handle *trans, struct btrfs_root *root,
2757 struct btrfs_path *path, int level, int slot)
2759 struct extent_buffer *parent = path->nodes[level];
2764 nritems = btrfs_header_nritems(parent);
2765 if (slot != nritems -1) {
2766 memmove_extent_buffer(parent,
2767 btrfs_node_key_ptr_offset(slot),
2768 btrfs_node_key_ptr_offset(slot + 1),
2769 sizeof(struct btrfs_key_ptr) *
2770 (nritems - slot - 1));
2773 btrfs_set_header_nritems(parent, nritems);
2774 if (nritems == 0 && parent == root->node) {
2775 BUG_ON(btrfs_header_level(root->node) != 1);
2776 /* just turn the root into a leaf and break */
2777 btrfs_set_header_level(root->node, 0);
2778 } else if (slot == 0) {
2779 struct btrfs_disk_key disk_key;
2781 btrfs_node_key(parent, &disk_key, 0);
2782 wret = fixup_low_keys(trans, root, path, &disk_key, level + 1);
2786 btrfs_mark_buffer_dirty(parent);
2791 * delete the item at the leaf level in path. If that empties
2792 * the leaf, remove it from the tree
2794 int btrfs_del_items(struct btrfs_trans_handle *trans, struct btrfs_root *root,
2795 struct btrfs_path *path, int slot, int nr)
2797 struct extent_buffer *leaf;
2798 struct btrfs_item *item;
2806 leaf = path->nodes[0];
2807 last_off = btrfs_item_offset_nr(leaf, slot + nr - 1);
2809 for (i = 0; i < nr; i++)
2810 dsize += btrfs_item_size_nr(leaf, slot + i);
2812 nritems = btrfs_header_nritems(leaf);
2814 if (slot + nr != nritems) {
2816 int data_end = leaf_data_end(root, leaf);
2818 memmove_extent_buffer(leaf, btrfs_leaf_data(leaf) +
2820 btrfs_leaf_data(leaf) + data_end,
2821 last_off - data_end);
2823 for (i = slot + nr; i < nritems; i++) {
2826 item = btrfs_item_nr(leaf, i);
2827 if (!leaf->map_token) {
2828 map_extent_buffer(leaf, (unsigned long)item,
2829 sizeof(struct btrfs_item),
2830 &leaf->map_token, &leaf->kaddr,
2831 &leaf->map_start, &leaf->map_len,
2834 ioff = btrfs_item_offset(leaf, item);
2835 btrfs_set_item_offset(leaf, item, ioff + dsize);
2838 if (leaf->map_token) {
2839 unmap_extent_buffer(leaf, leaf->map_token, KM_USER1);
2840 leaf->map_token = NULL;
2843 memmove_extent_buffer(leaf, btrfs_item_nr_offset(slot),
2844 btrfs_item_nr_offset(slot + nr),
2845 sizeof(struct btrfs_item) *
2846 (nritems - slot - nr));
2848 btrfs_set_header_nritems(leaf, nritems - nr);
2851 /* delete the leaf if we've emptied it */
2853 if (leaf == root->node) {
2854 btrfs_set_header_level(leaf, 0);
2856 u64 root_gen = btrfs_header_generation(path->nodes[1]);
2857 wret = del_ptr(trans, root, path, 1, path->slots[1]);
2860 wret = btrfs_free_extent(trans, root,
2861 leaf->start, leaf->len,
2862 btrfs_header_owner(path->nodes[1]),
2868 int used = leaf_space_used(leaf, 0, nritems);
2870 struct btrfs_disk_key disk_key;
2872 btrfs_item_key(leaf, &disk_key, 0);
2873 wret = fixup_low_keys(trans, root, path,
2879 /* delete the leaf if it is mostly empty */
2880 if (used < BTRFS_LEAF_DATA_SIZE(root) / 4) {
2881 /* push_leaf_left fixes the path.
2882 * make sure the path still points to our leaf
2883 * for possible call to del_ptr below
2885 slot = path->slots[1];
2886 extent_buffer_get(leaf);
2888 wret = push_leaf_left(trans, root, path, 1, 1);
2889 if (wret < 0 && wret != -ENOSPC)
2892 if (path->nodes[0] == leaf &&
2893 btrfs_header_nritems(leaf)) {
2894 wret = push_leaf_right(trans, root, path, 1, 1);
2895 if (wret < 0 && wret != -ENOSPC)
2899 if (btrfs_header_nritems(leaf) == 0) {
2901 u64 bytenr = leaf->start;
2902 u32 blocksize = leaf->len;
2904 root_gen = btrfs_header_generation(
2907 wret = del_ptr(trans, root, path, 1, slot);
2911 free_extent_buffer(leaf);
2912 wret = btrfs_free_extent(trans, root, bytenr,
2914 btrfs_header_owner(path->nodes[1]),
2919 /* if we're still in the path, make sure
2920 * we're dirty. Otherwise, one of the
2921 * push_leaf functions must have already
2922 * dirtied this buffer
2924 if (path->nodes[0] == leaf)
2925 btrfs_mark_buffer_dirty(leaf);
2926 free_extent_buffer(leaf);
2929 btrfs_mark_buffer_dirty(leaf);
2936 * search the tree again to find a leaf with lesser keys
2937 * returns 0 if it found something or 1 if there are no lesser leaves.
2938 * returns < 0 on io errors.
2940 int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path)
2942 struct btrfs_key key;
2943 struct btrfs_disk_key found_key;
2946 btrfs_item_key_to_cpu(path->nodes[0], &key, 0);
2950 else if (key.type > 0)
2952 else if (key.objectid > 0)
2957 btrfs_release_path(root, path);
2958 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
2961 btrfs_item_key(path->nodes[0], &found_key, 0);
2962 ret = comp_keys(&found_key, &key);
2969 * A helper function to walk down the tree starting at min_key, and looking
2970 * for nodes or leaves that are either in cache or have a minimum
2971 * transaction id. This is used by the btree defrag code, but could
2972 * also be used to search for blocks that have changed since a given
2975 * This does not cow, but it does stuff the starting key it finds back
2976 * into min_key, so you can call btrfs_search_slot with cow=1 on the
2977 * key and get a writable path.
2979 * This does lock as it descends, and path->keep_locks should be set
2980 * to 1 by the caller.
2982 * This honors path->lowest_level to prevent descent past a given level
2985 * returns zero if something useful was found, < 0 on error and 1 if there
2986 * was nothing in the tree that matched the search criteria.
2988 int btrfs_search_forward(struct btrfs_root *root, struct btrfs_key *min_key,
2989 struct btrfs_path *path, int cache_only,
2992 struct extent_buffer *cur;
2993 struct btrfs_key found_key;
3000 cur = btrfs_lock_root_node(root);
3001 level = btrfs_header_level(cur);
3002 path->nodes[level] = cur;
3003 path->locks[level] = 1;
3005 if (btrfs_header_generation(cur) < min_trans) {
3010 nritems = btrfs_header_nritems(cur);
3011 level = btrfs_header_level(cur);
3012 bin_search(cur, min_key, level, &slot);
3014 /* at level = 0, we're done, setup the path and exit */
3017 path->slots[level] = slot;
3018 btrfs_item_key_to_cpu(cur, &found_key, slot);
3022 * check this node pointer against the cache_only and
3023 * min_trans parameters. If it isn't in cache or is too
3024 * old, skip to the next one.
3026 while(slot < nritems) {
3029 struct extent_buffer *tmp;
3030 blockptr = btrfs_node_blockptr(cur, slot);
3031 gen = btrfs_node_ptr_generation(cur, slot);
3032 if (gen < min_trans) {
3039 tmp = btrfs_find_tree_block(root, blockptr,
3040 btrfs_level_size(root, level - 1));
3042 if (tmp && btrfs_buffer_uptodate(tmp, gen)) {
3043 free_extent_buffer(tmp);
3047 free_extent_buffer(tmp);
3051 * we didn't find a candidate key in this node, walk forward
3052 * and find another one
3054 if (slot >= nritems) {
3055 ret = btrfs_find_next_key(root, path, min_key, level,
3056 cache_only, min_trans);
3058 btrfs_release_path(root, path);
3064 /* save our key for returning back */
3065 btrfs_node_key_to_cpu(cur, &found_key, slot);
3066 path->slots[level] = slot;
3067 if (level == path->lowest_level) {
3069 unlock_up(path, level, 1);
3072 cur = read_node_slot(root, cur, slot);
3074 btrfs_tree_lock(cur);
3075 path->locks[level - 1] = 1;
3076 path->nodes[level - 1] = cur;
3077 unlock_up(path, level, 1);
3081 memcpy(min_key, &found_key, sizeof(found_key));
3086 * this is similar to btrfs_next_leaf, but does not try to preserve
3087 * and fixup the path. It looks for and returns the next key in the
3088 * tree based on the current path and the cache_only and min_trans
3091 * 0 is returned if another key is found, < 0 if there are any errors
3092 * and 1 is returned if there are no higher keys in the tree
3094 * path->keep_locks should be set to 1 on the search made before
3095 * calling this function.
3097 int btrfs_find_next_key(struct btrfs_root *root, struct btrfs_path *path,
3098 struct btrfs_key *key, int lowest_level,
3099 int cache_only, u64 min_trans)
3101 int level = lowest_level;
3103 struct extent_buffer *c;
3105 while(level < BTRFS_MAX_LEVEL) {
3106 if (!path->nodes[level])
3109 slot = path->slots[level] + 1;
3110 c = path->nodes[level];
3112 if (slot >= btrfs_header_nritems(c)) {
3114 if (level == BTRFS_MAX_LEVEL) {
3120 btrfs_item_key_to_cpu(c, key, slot);
3122 u64 blockptr = btrfs_node_blockptr(c, slot);
3123 u64 gen = btrfs_node_ptr_generation(c, slot);
3126 struct extent_buffer *cur;
3127 cur = btrfs_find_tree_block(root, blockptr,
3128 btrfs_level_size(root, level - 1));
3129 if (!cur || !btrfs_buffer_uptodate(cur, gen)) {
3132 free_extent_buffer(cur);
3135 free_extent_buffer(cur);
3137 if (gen < min_trans) {
3141 btrfs_node_key_to_cpu(c, key, slot);
3149 * search the tree again to find a leaf with greater keys
3150 * returns 0 if it found something or 1 if there are no greater leaves.
3151 * returns < 0 on io errors.
3153 int btrfs_next_leaf(struct btrfs_root *root, struct btrfs_path *path)
3157 struct extent_buffer *c;
3158 struct extent_buffer *next = NULL;
3159 struct btrfs_key key;
3163 nritems = btrfs_header_nritems(path->nodes[0]);
3168 btrfs_item_key_to_cpu(path->nodes[0], &key, nritems - 1);
3170 btrfs_release_path(root, path);
3171 path->keep_locks = 1;
3172 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
3173 path->keep_locks = 0;
3178 nritems = btrfs_header_nritems(path->nodes[0]);
3180 * by releasing the path above we dropped all our locks. A balance
3181 * could have added more items next to the key that used to be
3182 * at the very end of the block. So, check again here and
3183 * advance the path if there are now more items available.
3185 if (nritems > 0 && path->slots[0] < nritems - 1) {
3190 while(level < BTRFS_MAX_LEVEL) {
3191 if (!path->nodes[level])
3194 slot = path->slots[level] + 1;
3195 c = path->nodes[level];
3196 if (slot >= btrfs_header_nritems(c)) {
3198 if (level == BTRFS_MAX_LEVEL) {
3205 btrfs_tree_unlock(next);
3206 free_extent_buffer(next);
3209 if (level == 1 && (path->locks[1] || path->skip_locking) &&
3211 reada_for_search(root, path, level, slot, 0);
3213 next = read_node_slot(root, c, slot);
3214 if (!path->skip_locking) {
3215 WARN_ON(!btrfs_tree_locked(c));
3216 btrfs_tree_lock(next);
3220 path->slots[level] = slot;
3223 c = path->nodes[level];
3224 if (path->locks[level])
3225 btrfs_tree_unlock(c);
3226 free_extent_buffer(c);
3227 path->nodes[level] = next;
3228 path->slots[level] = 0;
3229 if (!path->skip_locking)
3230 path->locks[level] = 1;
3233 if (level == 1 && path->locks[1] && path->reada)
3234 reada_for_search(root, path, level, slot, 0);
3235 next = read_node_slot(root, next, 0);
3236 if (!path->skip_locking) {
3237 WARN_ON(!btrfs_tree_locked(path->nodes[level]));
3238 btrfs_tree_lock(next);
3242 unlock_up(path, 0, 1);
3247 * this uses btrfs_prev_leaf to walk backwards in the tree, and keeps
3248 * searching until it gets past min_objectid or finds an item of 'type'
3250 * returns 0 if something is found, 1 if nothing was found and < 0 on error
3252 int btrfs_previous_item(struct btrfs_root *root,
3253 struct btrfs_path *path, u64 min_objectid,
3256 struct btrfs_key found_key;
3257 struct extent_buffer *leaf;
3261 if (path->slots[0] == 0) {
3262 ret = btrfs_prev_leaf(root, path);
3268 leaf = path->nodes[0];
3269 btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
3270 if (found_key.type == type)