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/module.h>
22 #include "print-tree.h"
23 #include "transaction.h"
25 static int find_free_extent(struct btrfs_trans_handle *trans, struct btrfs_root
26 *orig_root, u64 num_blocks, u64 search_start,
27 u64 search_end, u64 hint_block,
28 struct btrfs_key *ins, int data);
29 static int finish_current_insert(struct btrfs_trans_handle *trans, struct
30 btrfs_root *extent_root);
31 static int del_pending_extents(struct btrfs_trans_handle *trans, struct
32 btrfs_root *extent_root);
34 static void reada_extent_leaves(struct btrfs_root *root,
35 struct btrfs_path *path, u64 limit)
37 struct btrfs_node *node;
47 node = btrfs_buffer_node(path->nodes[1]);
48 slot = path->slots[1] + 1;
49 nritems = btrfs_header_nritems(&node->header);
50 for (i = slot; i < nritems && i < slot + 8; i++) {
51 item_objectid = btrfs_disk_key_objectid(&node->ptrs[i].key);
52 if (item_objectid > limit)
54 blocknr = btrfs_node_blockptr(node, i);
55 ret = readahead_tree_block(root, blocknr);
61 static int cache_block_group(struct btrfs_root *root,
62 struct btrfs_block_group_cache *block_group)
64 struct btrfs_path *path;
67 struct btrfs_leaf *leaf;
68 struct radix_tree_root *extent_radix;
76 root = root->fs_info->extent_root;
77 extent_radix = &root->fs_info->extent_map_radix;
79 if (block_group->cached)
81 if (block_group->data)
83 path = btrfs_alloc_path();
86 key.objectid = block_group->key.objectid;
89 btrfs_set_key_type(&key, BTRFS_EXTENT_ITEM_KEY);
90 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
93 if (ret && path->slots[0] > 0)
95 limit = block_group->key.objectid + block_group->key.offset;
96 reada_extent_leaves(root, path, limit);
98 leaf = btrfs_buffer_leaf(path->nodes[0]);
99 slot = path->slots[0];
100 if (slot >= btrfs_header_nritems(&leaf->header)) {
101 reada_extent_leaves(root, path, limit);
102 ret = btrfs_next_leaf(root, path);
109 hole_size = block_group->key.objectid +
110 block_group->key.offset - last;
112 last = block_group->key.objectid;
113 hole_size = block_group->key.offset;
115 for (i = 0; i < hole_size; i++) {
116 set_radix_bit(extent_radix,
122 btrfs_disk_key_to_cpu(&key, &leaf->items[slot].key);
123 if (key.objectid >= block_group->key.objectid +
124 block_group->key.offset) {
126 hole_size = block_group->key.objectid +
127 block_group->key.offset - last;
129 last = block_group->key.objectid;
130 hole_size = block_group->key.offset;
132 for (i = 0; i < hole_size; i++) {
133 set_radix_bit(extent_radix, last + i);
137 if (btrfs_key_type(&key) == BTRFS_EXTENT_ITEM_KEY) {
139 last = key.objectid + key.offset;
142 hole_size = key.objectid - last;
143 for (i = 0; i < hole_size; i++) {
144 set_radix_bit(extent_radix, last + i);
146 last = key.objectid + key.offset;
152 block_group->cached = 1;
154 btrfs_free_path(path);
158 struct btrfs_block_group_cache *btrfs_lookup_block_group(struct
162 struct btrfs_block_group_cache *block_group;
165 ret = radix_tree_gang_lookup(&info->block_group_radix,
166 (void **)&block_group,
169 if (block_group->key.objectid <= blocknr && blocknr <=
170 block_group->key.objectid + block_group->key.offset)
173 ret = radix_tree_gang_lookup(&info->block_group_data_radix,
174 (void **)&block_group,
177 if (block_group->key.objectid <= blocknr && blocknr <=
178 block_group->key.objectid + block_group->key.offset)
184 static u64 leaf_range(struct btrfs_root *root)
186 u64 size = BTRFS_LEAF_DATA_SIZE(root);
187 do_div(size, sizeof(struct btrfs_extent_item) +
188 sizeof(struct btrfs_item));
192 static u64 find_search_start(struct btrfs_root *root,
193 struct btrfs_block_group_cache **cache_ret,
194 u64 search_start, int num)
196 unsigned long gang[8];
198 struct btrfs_block_group_cache *cache = *cache_ret;
199 u64 last = max(search_start, cache->key.objectid);
204 last = max(last, cache->last_prealloc);
207 ret = cache_block_group(root, cache);
211 ret = find_first_radix_bit(&root->fs_info->extent_map_radix,
212 gang, last, ARRAY_SIZE(gang));
215 last = gang[ret-1] + 1;
217 if (ret != ARRAY_SIZE(gang)) {
220 if (gang[ret-1] - gang[0] > leaf_range(root)) {
224 if (gang[0] >= cache->key.objectid + cache->key.offset) {
230 return max(cache->last_alloc, search_start);
233 cache = btrfs_lookup_block_group(root->fs_info,
234 last + cache->key.offset - 1);
236 return max((*cache_ret)->last_alloc, search_start);
238 cache = btrfs_find_block_group(root, cache,
239 last + cache->key.offset - 1, 0, 0);
244 static u64 div_factor(u64 num, int factor)
251 struct btrfs_block_group_cache *btrfs_find_block_group(struct btrfs_root *root,
252 struct btrfs_block_group_cache
253 *hint, u64 search_start,
256 struct btrfs_block_group_cache *cache[8];
257 struct btrfs_block_group_cache *found_group = NULL;
258 struct btrfs_fs_info *info = root->fs_info;
259 struct radix_tree_root *radix;
260 struct radix_tree_root *swap_radix;
274 radix = &info->block_group_data_radix;
275 swap_radix = &info->block_group_radix;
277 radix = &info->block_group_radix;
278 swap_radix = &info->block_group_data_radix;
282 struct btrfs_block_group_cache *shint;
283 shint = btrfs_lookup_block_group(info, search_start);
284 if (shint->data == data) {
285 used = btrfs_block_group_used(&shint->item);
286 if (used + shint->pinned <
287 div_factor(shint->key.offset, factor)) {
292 if (hint && hint->data == data) {
293 used = btrfs_block_group_used(&hint->item);
294 if (used + hint->pinned <
295 div_factor(hint->key.offset, factor)) {
298 if (used >= div_factor(hint->key.offset, 8)) {
299 radix_tree_tag_clear(radix,
301 hint->key.offset - 1,
302 BTRFS_BLOCK_GROUP_AVAIL);
304 last = hint->key.offset * 3;
305 if (hint->key.objectid >= last)
306 last = max(search_start + hint->key.offset - 1,
307 hint->key.objectid - last);
309 last = hint->key.objectid + hint->key.offset;
313 hint_last = max(hint->key.objectid, search_start);
315 hint_last = search_start;
320 ret = radix_tree_gang_lookup_tag(radix, (void **)cache,
321 last, ARRAY_SIZE(cache),
322 BTRFS_BLOCK_GROUP_AVAIL);
325 for (i = 0; i < ret; i++) {
326 last = cache[i]->key.objectid +
327 cache[i]->key.offset;
328 used = btrfs_block_group_used(&cache[i]->item);
329 if (used + cache[i]->pinned <
330 div_factor(cache[i]->key.offset, factor)) {
331 found_group = cache[i];
334 if (used >= div_factor(cache[i]->key.offset, 8)) {
335 radix_tree_tag_clear(radix,
336 cache[i]->key.objectid +
337 cache[i]->key.offset - 1,
338 BTRFS_BLOCK_GROUP_AVAIL);
346 ret = radix_tree_gang_lookup(radix, (void **)cache,
347 last, ARRAY_SIZE(cache));
350 for (i = 0; i < ret; i++) {
351 last = cache[i]->key.objectid +
352 cache[i]->key.offset;
353 used = btrfs_block_group_used(&cache[i]->item);
354 if (used + cache[i]->pinned < cache[i]->key.offset) {
355 found_group = cache[i];
358 if (used >= cache[i]->key.offset) {
359 radix_tree_tag_clear(radix,
360 cache[i]->key.objectid +
361 cache[i]->key.offset - 1,
362 BTRFS_BLOCK_GROUP_AVAIL);
373 struct radix_tree_root *tmp = radix;
381 ret = radix_tree_gang_lookup(radix,
382 (void **)&found_group, 0, 1);
384 ret = radix_tree_gang_lookup(swap_radix,
385 (void **)&found_group,
394 int btrfs_inc_extent_ref(struct btrfs_trans_handle *trans,
395 struct btrfs_root *root,
396 u64 blocknr, u64 num_blocks)
398 struct btrfs_path *path;
400 struct btrfs_key key;
401 struct btrfs_leaf *l;
402 struct btrfs_extent_item *item;
403 struct btrfs_key ins;
406 path = btrfs_alloc_path();
409 ret = find_free_extent(trans, root->fs_info->extent_root, 0, 0,
410 (u64)-1, 0, &ins, 0);
412 btrfs_free_path(path);
415 key.objectid = blocknr;
417 btrfs_set_key_type(&key, BTRFS_EXTENT_ITEM_KEY);
418 key.offset = num_blocks;
419 ret = btrfs_search_slot(trans, root->fs_info->extent_root, &key, path,
427 l = btrfs_buffer_leaf(path->nodes[0]);
428 item = btrfs_item_ptr(l, path->slots[0], struct btrfs_extent_item);
429 refs = btrfs_extent_refs(item);
430 btrfs_set_extent_refs(item, refs + 1);
431 btrfs_mark_buffer_dirty(path->nodes[0]);
433 btrfs_release_path(root->fs_info->extent_root, path);
434 btrfs_free_path(path);
435 finish_current_insert(trans, root->fs_info->extent_root);
436 del_pending_extents(trans, root->fs_info->extent_root);
440 static int lookup_extent_ref(struct btrfs_trans_handle *trans,
441 struct btrfs_root *root, u64 blocknr,
442 u64 num_blocks, u32 *refs)
444 struct btrfs_path *path;
446 struct btrfs_key key;
447 struct btrfs_leaf *l;
448 struct btrfs_extent_item *item;
450 path = btrfs_alloc_path();
451 key.objectid = blocknr;
452 key.offset = num_blocks;
454 btrfs_set_key_type(&key, BTRFS_EXTENT_ITEM_KEY);
455 ret = btrfs_search_slot(trans, root->fs_info->extent_root, &key, path,
461 l = btrfs_buffer_leaf(path->nodes[0]);
462 item = btrfs_item_ptr(l, path->slots[0], struct btrfs_extent_item);
463 *refs = btrfs_extent_refs(item);
465 btrfs_free_path(path);
469 int btrfs_inc_root_ref(struct btrfs_trans_handle *trans,
470 struct btrfs_root *root)
472 return btrfs_inc_extent_ref(trans, root, bh_blocknr(root->node), 1);
475 int btrfs_inc_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root,
476 struct buffer_head *buf)
479 struct btrfs_node *buf_node;
480 struct btrfs_leaf *buf_leaf;
481 struct btrfs_disk_key *key;
482 struct btrfs_file_extent_item *fi;
491 buf_node = btrfs_buffer_node(buf);
492 leaf = btrfs_is_leaf(buf_node);
493 buf_leaf = btrfs_buffer_leaf(buf);
494 for (i = 0; i < btrfs_header_nritems(&buf_node->header); i++) {
497 key = &buf_leaf->items[i].key;
498 if (btrfs_disk_key_type(key) != BTRFS_EXTENT_DATA_KEY)
500 fi = btrfs_item_ptr(buf_leaf, i,
501 struct btrfs_file_extent_item);
502 if (btrfs_file_extent_type(fi) ==
503 BTRFS_FILE_EXTENT_INLINE)
505 disk_blocknr = btrfs_file_extent_disk_blocknr(fi);
506 if (disk_blocknr == 0)
508 ret = btrfs_inc_extent_ref(trans, root, disk_blocknr,
509 btrfs_file_extent_disk_num_blocks(fi));
515 blocknr = btrfs_node_blockptr(buf_node, i);
516 ret = btrfs_inc_extent_ref(trans, root, blocknr, 1);
525 for (i =0; i < faili; i++) {
528 key = &buf_leaf->items[i].key;
529 if (btrfs_disk_key_type(key) != BTRFS_EXTENT_DATA_KEY)
531 fi = btrfs_item_ptr(buf_leaf, i,
532 struct btrfs_file_extent_item);
533 if (btrfs_file_extent_type(fi) ==
534 BTRFS_FILE_EXTENT_INLINE)
536 disk_blocknr = btrfs_file_extent_disk_blocknr(fi);
537 if (disk_blocknr == 0)
539 err = btrfs_free_extent(trans, root, disk_blocknr,
540 btrfs_file_extent_disk_num_blocks(fi), 0);
543 blocknr = btrfs_node_blockptr(buf_node, i);
544 err = btrfs_free_extent(trans, root, blocknr, 1, 0);
551 static int write_one_cache_group(struct btrfs_trans_handle *trans,
552 struct btrfs_root *root,
553 struct btrfs_path *path,
554 struct btrfs_block_group_cache *cache)
558 struct btrfs_root *extent_root = root->fs_info->extent_root;
559 struct btrfs_block_group_item *bi;
560 struct btrfs_key ins;
562 ret = find_free_extent(trans, extent_root, 0, 0, (u64)-1, 0, &ins, 0);
563 /* FIXME, set bit to recalc cache groups on next mount */
566 ret = btrfs_search_slot(trans, extent_root, &cache->key, path, 0, 1);
570 bi = btrfs_item_ptr(btrfs_buffer_leaf(path->nodes[0]), path->slots[0],
571 struct btrfs_block_group_item);
572 memcpy(bi, &cache->item, sizeof(*bi));
573 mark_buffer_dirty(path->nodes[0]);
574 btrfs_release_path(extent_root, path);
576 finish_current_insert(trans, extent_root);
577 pending_ret = del_pending_extents(trans, extent_root);
583 cache->last_alloc = cache->first_free;
588 static int write_dirty_block_radix(struct btrfs_trans_handle *trans,
589 struct btrfs_root *root,
590 struct radix_tree_root *radix)
592 struct btrfs_block_group_cache *cache[8];
597 struct btrfs_path *path;
598 unsigned long off = 0;
600 path = btrfs_alloc_path();
605 ret = radix_tree_gang_lookup_tag(radix, (void **)cache,
606 off, ARRAY_SIZE(cache),
607 BTRFS_BLOCK_GROUP_DIRTY);
610 for (i = 0; i < ret; i++) {
611 err = write_one_cache_group(trans, root,
614 * if we fail to write the cache group, we want
615 * to keep it marked dirty in hopes that a later
620 off = cache[i]->key.objectid +
621 cache[i]->key.offset;
625 radix_tree_tag_clear(radix, cache[i]->key.objectid +
626 cache[i]->key.offset - 1,
627 BTRFS_BLOCK_GROUP_DIRTY);
630 btrfs_free_path(path);
634 int btrfs_write_dirty_block_groups(struct btrfs_trans_handle *trans,
635 struct btrfs_root *root)
639 ret = write_dirty_block_radix(trans, root,
640 &root->fs_info->block_group_radix);
641 ret2 = write_dirty_block_radix(trans, root,
642 &root->fs_info->block_group_data_radix);
650 static int update_block_group(struct btrfs_trans_handle *trans,
651 struct btrfs_root *root,
652 u64 blocknr, u64 num, int alloc, int mark_free,
655 struct btrfs_block_group_cache *cache;
656 struct btrfs_fs_info *info = root->fs_info;
664 cache = btrfs_lookup_block_group(info, blocknr);
668 block_in_group = blocknr - cache->key.objectid;
669 WARN_ON(block_in_group > cache->key.offset);
670 radix_tree_tag_set(cache->radix, cache->key.objectid +
671 cache->key.offset - 1,
672 BTRFS_BLOCK_GROUP_DIRTY);
674 old_val = btrfs_block_group_used(&cache->item);
675 num = min(total, cache->key.offset - block_in_group);
677 if (blocknr > cache->last_alloc)
678 cache->last_alloc = blocknr;
680 for (i = 0; i < num; i++) {
681 clear_radix_bit(&info->extent_map_radix,
685 if (cache->data != data &&
686 old_val < (cache->key.offset >> 1)) {
688 radix_tree_delete(cache->radix,
689 cache->key.objectid +
690 cache->key.offset - 1);
694 &info->block_group_data_radix;
696 BTRFS_BLOCK_GROUP_DATA;
698 cache->radix = &info->block_group_radix;
700 ~BTRFS_BLOCK_GROUP_DATA;
702 ret = radix_tree_insert(cache->radix,
703 cache->key.objectid +
704 cache->key.offset - 1,
710 if (blocknr < cache->first_free)
711 cache->first_free = blocknr;
712 if (!cache->data && mark_free) {
713 for (i = 0; i < num; i++) {
714 set_radix_bit(&info->extent_map_radix,
718 if (old_val < (cache->key.offset >> 1) &&
719 old_val + num >= (cache->key.offset >> 1)) {
720 radix_tree_tag_set(cache->radix,
721 cache->key.objectid +
722 cache->key.offset - 1,
723 BTRFS_BLOCK_GROUP_AVAIL);
726 btrfs_set_block_group_used(&cache->item, old_val);
733 static int try_remove_page(struct address_space *mapping, unsigned long index)
736 ret = invalidate_mapping_pages(mapping, index, index);
740 int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans, struct
743 unsigned long gang[8];
744 struct inode *btree_inode = root->fs_info->btree_inode;
745 struct btrfs_block_group_cache *block_group;
749 struct radix_tree_root *pinned_radix = &root->fs_info->pinned_radix;
750 struct radix_tree_root *extent_radix = &root->fs_info->extent_map_radix;
753 ret = find_first_radix_bit(pinned_radix, gang, 0,
759 for (i = 0; i < ret; i++) {
760 clear_radix_bit(pinned_radix, gang[i]);
761 block_group = btrfs_lookup_block_group(root->fs_info,
764 WARN_ON(block_group->pinned == 0);
765 block_group->pinned--;
766 if (gang[i] < block_group->last_alloc)
767 block_group->last_alloc = gang[i];
768 if (gang[i] < block_group->last_prealloc)
769 block_group->last_prealloc = gang[i];
770 if (!block_group->data)
771 set_radix_bit(extent_radix, gang[i]);
773 try_remove_page(btree_inode->i_mapping,
774 gang[i] << (PAGE_CACHE_SHIFT -
775 btree_inode->i_blkbits));
781 static int finish_current_insert(struct btrfs_trans_handle *trans, struct
782 btrfs_root *extent_root)
784 struct btrfs_key ins;
785 struct btrfs_extent_item extent_item;
788 u64 super_blocks_used;
789 struct btrfs_fs_info *info = extent_root->fs_info;
791 btrfs_set_extent_refs(&extent_item, 1);
794 btrfs_set_key_type(&ins, BTRFS_EXTENT_ITEM_KEY);
795 btrfs_set_extent_owner(&extent_item, extent_root->root_key.objectid);
797 for (i = 0; i < extent_root->fs_info->extent_tree_insert_nr; i++) {
798 ins.objectid = extent_root->fs_info->extent_tree_insert[i];
799 super_blocks_used = btrfs_super_blocks_used(info->disk_super);
800 btrfs_set_super_blocks_used(info->disk_super,
801 super_blocks_used + 1);
802 ret = btrfs_insert_item(trans, extent_root, &ins, &extent_item,
803 sizeof(extent_item));
806 extent_root->fs_info->extent_tree_insert_nr = 0;
810 static int pin_down_block(struct btrfs_root *root, u64 blocknr, int pending)
813 struct btrfs_header *header;
814 struct buffer_head *bh;
817 bh = btrfs_find_tree_block(root, blocknr);
819 if (buffer_uptodate(bh)) {
821 root->fs_info->running_transaction->transid;
822 header = btrfs_buffer_header(bh);
823 if (btrfs_header_generation(header) ==
825 btrfs_block_release(root, bh);
829 btrfs_block_release(root, bh);
831 err = set_radix_bit(&root->fs_info->pinned_radix, blocknr);
833 struct btrfs_block_group_cache *cache;
834 cache = btrfs_lookup_block_group(root->fs_info,
840 err = set_radix_bit(&root->fs_info->pending_del_radix, blocknr);
847 * remove an extent from the root, returns 0 on success
849 static int __free_extent(struct btrfs_trans_handle *trans, struct btrfs_root
850 *root, u64 blocknr, u64 num_blocks, int pin,
853 struct btrfs_path *path;
854 struct btrfs_key key;
855 struct btrfs_fs_info *info = root->fs_info;
856 struct btrfs_root *extent_root = info->extent_root;
858 struct btrfs_extent_item *ei;
859 struct btrfs_key ins;
862 key.objectid = blocknr;
864 btrfs_set_key_type(&key, BTRFS_EXTENT_ITEM_KEY);
865 key.offset = num_blocks;
867 path = btrfs_alloc_path();
871 ret = find_free_extent(trans, root, 0, 0, (u64)-1, 0, &ins, 0);
873 btrfs_free_path(path);
877 ret = btrfs_search_slot(trans, extent_root, &key, path, -1, 1);
881 ei = btrfs_item_ptr(btrfs_buffer_leaf(path->nodes[0]), path->slots[0],
882 struct btrfs_extent_item);
883 BUG_ON(ei->refs == 0);
884 refs = btrfs_extent_refs(ei) - 1;
885 btrfs_set_extent_refs(ei, refs);
886 btrfs_mark_buffer_dirty(path->nodes[0]);
888 u64 super_blocks_used;
891 ret = pin_down_block(root, blocknr, 0);
895 super_blocks_used = btrfs_super_blocks_used(info->disk_super);
896 btrfs_set_super_blocks_used(info->disk_super,
897 super_blocks_used - num_blocks);
898 ret = btrfs_del_item(trans, extent_root, path);
902 ret = update_block_group(trans, root, blocknr, num_blocks, 0,
906 btrfs_free_path(path);
907 finish_current_insert(trans, extent_root);
912 * find all the blocks marked as pending in the radix tree and remove
913 * them from the extent map
915 static int del_pending_extents(struct btrfs_trans_handle *trans, struct
916 btrfs_root *extent_root)
921 unsigned long gang[4];
923 struct radix_tree_root *pending_radix;
924 struct radix_tree_root *pinned_radix;
925 struct btrfs_block_group_cache *cache;
927 pending_radix = &extent_root->fs_info->pending_del_radix;
928 pinned_radix = &extent_root->fs_info->pinned_radix;
931 ret = find_first_radix_bit(pending_radix, gang, 0,
935 for (i = 0; i < ret; i++) {
936 wret = set_radix_bit(pinned_radix, gang[i]);
939 btrfs_lookup_block_group(extent_root->fs_info,
945 printk(KERN_CRIT "set_radix_bit, err %d\n",
949 wret = clear_radix_bit(pending_radix, gang[i]);
951 wret = __free_extent(trans, extent_root,
961 * remove an extent from the root, returns 0 on success
963 int btrfs_free_extent(struct btrfs_trans_handle *trans, struct btrfs_root
964 *root, u64 blocknr, u64 num_blocks, int pin)
966 struct btrfs_root *extent_root = root->fs_info->extent_root;
970 if (root == extent_root) {
971 pin_down_block(root, blocknr, 1);
974 ret = __free_extent(trans, root, blocknr, num_blocks, pin, pin == 0);
975 pending_ret = del_pending_extents(trans, root->fs_info->extent_root);
976 return ret ? ret : pending_ret;
980 * walks the btree of allocated extents and find a hole of a given size.
981 * The key ins is changed to record the hole:
982 * ins->objectid == block start
983 * ins->flags = BTRFS_EXTENT_ITEM_KEY
984 * ins->offset == number of blocks
985 * Any available blocks before search_start are skipped.
987 static int find_free_extent(struct btrfs_trans_handle *trans, struct btrfs_root
988 *orig_root, u64 num_blocks, u64 search_start, u64
989 search_end, u64 hint_block,
990 struct btrfs_key *ins, int data)
992 struct btrfs_path *path;
993 struct btrfs_key key;
999 u64 orig_search_start = search_start;
1001 struct btrfs_leaf *l;
1002 struct btrfs_root * root = orig_root->fs_info->extent_root;
1003 struct btrfs_fs_info *info = root->fs_info;
1004 int total_needed = num_blocks;
1005 int total_found = 0;
1006 int fill_prealloc = 0;
1008 struct btrfs_block_group_cache *block_group;
1014 btrfs_set_key_type(ins, BTRFS_EXTENT_ITEM_KEY);
1016 level = btrfs_header_level(btrfs_buffer_header(root->node));
1017 if (num_blocks == 0) {
1020 total_needed = (min(level + 1, BTRFS_MAX_LEVEL) + 2) * 3;
1022 if (fill_prealloc) {
1024 int nr = info->extent_tree_prealloc_nr;
1025 first = info->extent_tree_prealloc[nr - 1];
1026 if (info->extent_tree_prealloc_nr >= total_needed &&
1027 first >= search_start) {
1028 ins->objectid = info->extent_tree_prealloc[0];
1032 info->extent_tree_prealloc_nr = 0;
1034 if (search_end == (u64)-1)
1035 search_end = btrfs_super_total_blocks(info->disk_super);
1037 block_group = btrfs_lookup_block_group(info, hint_block);
1038 block_group = btrfs_find_block_group(root, block_group,
1039 hint_block, data, 1);
1041 block_group = btrfs_find_block_group(root,
1042 trans->block_group, 0,
1046 path = btrfs_alloc_path();
1049 if (!block_group->data)
1050 search_start = find_search_start(root, &block_group,
1051 search_start, total_needed);
1052 else if (!full_scan)
1053 search_start = max(block_group->last_alloc, search_start);
1055 btrfs_init_path(path);
1056 ins->objectid = search_start;
1060 ret = btrfs_search_slot(trans, root, ins, path, 0, 0);
1064 if (path->slots[0] > 0) {
1068 l = btrfs_buffer_leaf(path->nodes[0]);
1069 btrfs_disk_key_to_cpu(&key, &l->items[path->slots[0]].key);
1071 * a rare case, go back one key if we hit a block group item
1072 * instead of an extent item
1074 if (btrfs_key_type(&key) != BTRFS_EXTENT_ITEM_KEY &&
1075 key.objectid + key.offset >= search_start) {
1076 ins->objectid = key.objectid;
1077 ins->offset = key.offset - 1;
1078 btrfs_release_path(root, path);
1079 ret = btrfs_search_slot(trans, root, ins, path, 0, 0);
1083 if (path->slots[0] > 0) {
1089 l = btrfs_buffer_leaf(path->nodes[0]);
1090 slot = path->slots[0];
1091 if (slot >= btrfs_header_nritems(&l->header)) {
1092 if (fill_prealloc) {
1093 info->extent_tree_prealloc_nr = 0;
1097 limit = last_block +
1098 (block_group->key.offset >> 1);
1100 limit = search_start +
1101 (block_group->key.offset >> 1);
1102 ret = btrfs_next_leaf(root, path);
1108 ins->objectid = search_start;
1109 ins->offset = search_end - search_start;
1113 ins->objectid = last_block > search_start ?
1114 last_block : search_start;
1115 ins->offset = search_end - ins->objectid;
1119 btrfs_disk_key_to_cpu(&key, &l->items[slot].key);
1120 if (key.objectid >= search_start && key.objectid > last_block &&
1122 if (last_block < search_start)
1123 last_block = search_start;
1124 hole_size = key.objectid - last_block;
1125 if (hole_size >= num_blocks) {
1126 ins->objectid = last_block;
1127 ins->offset = hole_size;
1132 if (btrfs_key_type(&key) != BTRFS_EXTENT_ITEM_KEY)
1136 last_block = key.objectid + key.offset;
1137 if (!full_scan && last_block >= block_group->key.objectid +
1138 block_group->key.offset) {
1139 btrfs_release_path(root, path);
1140 search_start = block_group->key.objectid +
1141 block_group->key.offset * 2;
1149 /* we have to make sure we didn't find an extent that has already
1150 * been allocated by the map tree or the original allocation
1152 btrfs_release_path(root, path);
1153 BUG_ON(ins->objectid < search_start);
1155 if (ins->objectid + num_blocks >= search_end) {
1160 search_start = orig_search_start;
1167 for (test_block = ins->objectid;
1168 test_block < ins->objectid + num_blocks; test_block++) {
1169 if (test_radix_bit(&info->pinned_radix, test_block)) {
1170 search_start = test_block + 1;
1174 if (!fill_prealloc && info->extent_tree_insert_nr) {
1176 info->extent_tree_insert[info->extent_tree_insert_nr - 1];
1177 if (ins->objectid + num_blocks >
1178 info->extent_tree_insert[0] &&
1179 ins->objectid <= last) {
1180 search_start = last + 1;
1181 WARN_ON(!full_scan);
1185 if (!fill_prealloc && info->extent_tree_prealloc_nr) {
1187 info->extent_tree_prealloc[info->extent_tree_prealloc_nr - 1];
1188 if (ins->objectid + num_blocks > first &&
1189 ins->objectid <= info->extent_tree_prealloc[0]) {
1190 search_start = info->extent_tree_prealloc[0] + 1;
1194 if (fill_prealloc) {
1196 test_block = ins->objectid;
1197 if (test_block - info->extent_tree_prealloc[total_needed - 1] >=
1200 info->extent_tree_prealloc_nr = total_found;
1202 while(test_block < ins->objectid + ins->offset &&
1203 total_found < total_needed) {
1204 nr = total_needed - total_found - 1;
1206 info->extent_tree_prealloc[nr] = test_block;
1210 if (total_found < total_needed) {
1211 search_start = test_block;
1214 info->extent_tree_prealloc_nr = total_found;
1217 block_group = btrfs_lookup_block_group(info, ins->objectid);
1220 block_group->last_prealloc =
1221 info->extent_tree_prealloc[total_needed-1];
1223 trans->block_group = block_group;
1226 ins->offset = num_blocks;
1227 btrfs_free_path(path);
1231 if (search_start + num_blocks >= search_end) {
1232 search_start = orig_search_start;
1242 block_group = btrfs_lookup_block_group(info, search_start);
1245 block_group = btrfs_find_block_group(root, block_group,
1246 search_start, data, 0);
1250 btrfs_release_path(root, path);
1251 btrfs_free_path(path);
1255 * finds a free extent and does all the dirty work required for allocation
1256 * returns the key for the extent through ins, and a tree buffer for
1257 * the first block of the extent through buf.
1259 * returns 0 if everything worked, non-zero otherwise.
1261 int btrfs_alloc_extent(struct btrfs_trans_handle *trans,
1262 struct btrfs_root *root, u64 owner,
1263 u64 num_blocks, u64 hint_block,
1264 u64 search_end, struct btrfs_key *ins, int data)
1268 u64 super_blocks_used;
1269 u64 search_start = 0;
1270 struct btrfs_fs_info *info = root->fs_info;
1271 struct btrfs_root *extent_root = info->extent_root;
1272 struct btrfs_extent_item extent_item;
1273 struct btrfs_key prealloc_key;
1275 btrfs_set_extent_refs(&extent_item, 1);
1276 btrfs_set_extent_owner(&extent_item, owner);
1278 if (root == extent_root) {
1280 BUG_ON(info->extent_tree_prealloc_nr == 0);
1281 BUG_ON(num_blocks != 1);
1283 info->extent_tree_prealloc_nr--;
1284 nr = info->extent_tree_prealloc_nr;
1285 ins->objectid = info->extent_tree_prealloc[nr];
1286 info->extent_tree_insert[info->extent_tree_insert_nr++] =
1288 ret = update_block_group(trans, root,
1289 ins->objectid, ins->offset, 1, 0, 0);
1295 * if we're doing a data allocation, preallocate room in the
1296 * extent tree first. This way the extent tree blocks end up
1297 * in the correct block group.
1300 ret = find_free_extent(trans, root, 0, 0,
1301 search_end, 0, &prealloc_key, 0);
1305 if (prealloc_key.objectid + prealloc_key.offset >= search_end) {
1306 int nr = info->extent_tree_prealloc_nr;
1307 search_end = info->extent_tree_prealloc[nr - 1] - 1;
1309 search_start = info->extent_tree_prealloc[0] + 1;
1312 if (hint_block < search_start)
1313 hint_block = search_start;
1314 /* do the real allocation */
1315 ret = find_free_extent(trans, root, num_blocks, search_start,
1316 search_end, hint_block, ins, data);
1318 if (search_start == 0)
1320 search_end = search_start - 1;
1322 hint_block = search_start;
1323 ret = find_free_extent(trans, root, num_blocks, search_start,
1324 search_end, hint_block, ins, data);
1330 * if we're doing a metadata allocation, preallocate space in the
1331 * extent tree second. This way, we don't create a tiny hole
1332 * in the allocation map between any unused preallocation blocks
1333 * and the metadata block we're actually allocating. On disk,
1335 * [block we've allocated], [used prealloc 1], [ unused prealloc ]
1336 * The unused prealloc will get reused the next time around.
1339 if (ins->objectid + ins->offset >= search_end)
1340 search_end = ins->objectid - 1;
1342 search_start = ins->objectid + ins->offset;
1344 if (hint_block < search_start)
1345 hint_block = search_start;
1347 ret = find_free_extent(trans, root, 0, search_start,
1348 search_end, hint_block,
1351 if (search_start == 0)
1353 search_end = search_start - 1;
1355 hint_block = search_start;
1356 ret = find_free_extent(trans, root, 0, search_start,
1357 search_end, hint_block,
1364 super_blocks_used = btrfs_super_blocks_used(info->disk_super);
1365 btrfs_set_super_blocks_used(info->disk_super, super_blocks_used +
1367 ret = btrfs_insert_item(trans, extent_root, ins, &extent_item,
1368 sizeof(extent_item));
1370 finish_current_insert(trans, extent_root);
1371 pending_ret = del_pending_extents(trans, extent_root);
1378 ret = update_block_group(trans, root, ins->objectid, ins->offset, 1, 0,
1385 * helper function to allocate a block for a given tree
1386 * returns the tree buffer or NULL.
1388 struct buffer_head *btrfs_alloc_free_block(struct btrfs_trans_handle *trans,
1389 struct btrfs_root *root, u64 hint)
1391 struct btrfs_key ins;
1393 struct buffer_head *buf;
1395 ret = btrfs_alloc_extent(trans, root, root->root_key.objectid,
1396 1, hint, (unsigned long)-1, &ins, 0);
1399 return ERR_PTR(ret);
1401 buf = btrfs_find_create_tree_block(root, ins.objectid);
1403 btrfs_free_extent(trans, root, ins.objectid, 1, 0);
1404 return ERR_PTR(-ENOMEM);
1406 set_buffer_uptodate(buf);
1407 set_buffer_checked(buf);
1408 set_radix_bit(&trans->transaction->dirty_pages, buf->b_page->index);
1412 static int drop_leaf_ref(struct btrfs_trans_handle *trans,
1413 struct btrfs_root *root, struct buffer_head *cur)
1415 struct btrfs_disk_key *key;
1416 struct btrfs_leaf *leaf;
1417 struct btrfs_file_extent_item *fi;
1422 BUG_ON(!btrfs_is_leaf(btrfs_buffer_node(cur)));
1423 leaf = btrfs_buffer_leaf(cur);
1424 nritems = btrfs_header_nritems(&leaf->header);
1425 for (i = 0; i < nritems; i++) {
1427 key = &leaf->items[i].key;
1428 if (btrfs_disk_key_type(key) != BTRFS_EXTENT_DATA_KEY)
1430 fi = btrfs_item_ptr(leaf, i, struct btrfs_file_extent_item);
1431 if (btrfs_file_extent_type(fi) == BTRFS_FILE_EXTENT_INLINE)
1434 * FIXME make sure to insert a trans record that
1435 * repeats the snapshot del on crash
1437 disk_blocknr = btrfs_file_extent_disk_blocknr(fi);
1438 if (disk_blocknr == 0)
1440 ret = btrfs_free_extent(trans, root, disk_blocknr,
1441 btrfs_file_extent_disk_num_blocks(fi),
1448 static void reada_walk_down(struct btrfs_root *root,
1449 struct btrfs_node *node)
1457 nritems = btrfs_header_nritems(&node->header);
1458 for (i = 0; i < nritems; i++) {
1459 blocknr = btrfs_node_blockptr(node, i);
1460 ret = lookup_extent_ref(NULL, root, blocknr, 1, &refs);
1464 ret = readahead_tree_block(root, blocknr);
1471 * helper function for drop_snapshot, this walks down the tree dropping ref
1472 * counts as it goes.
1474 static int walk_down_tree(struct btrfs_trans_handle *trans, struct btrfs_root
1475 *root, struct btrfs_path *path, int *level)
1477 struct buffer_head *next;
1478 struct buffer_head *cur;
1483 WARN_ON(*level < 0);
1484 WARN_ON(*level >= BTRFS_MAX_LEVEL);
1485 ret = lookup_extent_ref(trans, root, bh_blocknr(path->nodes[*level]),
1492 * walk down to the last node level and free all the leaves
1494 while(*level >= 0) {
1495 WARN_ON(*level < 0);
1496 WARN_ON(*level >= BTRFS_MAX_LEVEL);
1497 cur = path->nodes[*level];
1499 if (*level > 0 && path->slots[*level] == 0)
1500 reada_walk_down(root, btrfs_buffer_node(cur));
1502 if (btrfs_header_level(btrfs_buffer_header(cur)) != *level)
1505 if (path->slots[*level] >=
1506 btrfs_header_nritems(btrfs_buffer_header(cur)))
1509 ret = drop_leaf_ref(trans, root, cur);
1513 blocknr = btrfs_node_blockptr(btrfs_buffer_node(cur),
1514 path->slots[*level]);
1515 ret = lookup_extent_ref(trans, root, blocknr, 1, &refs);
1518 path->slots[*level]++;
1519 ret = btrfs_free_extent(trans, root, blocknr, 1, 1);
1523 next = read_tree_block(root, blocknr);
1524 WARN_ON(*level <= 0);
1525 if (path->nodes[*level-1])
1526 btrfs_block_release(root, path->nodes[*level-1]);
1527 path->nodes[*level-1] = next;
1528 *level = btrfs_header_level(btrfs_buffer_header(next));
1529 path->slots[*level] = 0;
1532 WARN_ON(*level < 0);
1533 WARN_ON(*level >= BTRFS_MAX_LEVEL);
1534 ret = btrfs_free_extent(trans, root,
1535 bh_blocknr(path->nodes[*level]), 1, 1);
1536 btrfs_block_release(root, path->nodes[*level]);
1537 path->nodes[*level] = NULL;
1544 * helper for dropping snapshots. This walks back up the tree in the path
1545 * to find the first node higher up where we haven't yet gone through
1548 static int walk_up_tree(struct btrfs_trans_handle *trans, struct btrfs_root
1549 *root, struct btrfs_path *path, int *level)
1554 for(i = *level; i < BTRFS_MAX_LEVEL - 1 && path->nodes[i]; i++) {
1555 slot = path->slots[i];
1556 if (slot < btrfs_header_nritems(
1557 btrfs_buffer_header(path->nodes[i])) - 1) {
1562 ret = btrfs_free_extent(trans, root,
1563 bh_blocknr(path->nodes[*level]),
1566 btrfs_block_release(root, path->nodes[*level]);
1567 path->nodes[*level] = NULL;
1575 * drop the reference count on the tree rooted at 'snap'. This traverses
1576 * the tree freeing any blocks that have a ref count of zero after being
1579 int btrfs_drop_snapshot(struct btrfs_trans_handle *trans, struct btrfs_root
1580 *root, struct buffer_head *snap)
1585 struct btrfs_path *path;
1589 path = btrfs_alloc_path();
1592 level = btrfs_header_level(btrfs_buffer_header(snap));
1594 path->nodes[level] = snap;
1595 path->slots[level] = 0;
1597 wret = walk_down_tree(trans, root, path, &level);
1603 wret = walk_up_tree(trans, root, path, &level);
1609 for (i = 0; i <= orig_level; i++) {
1610 if (path->nodes[i]) {
1611 btrfs_block_release(root, path->nodes[i]);
1614 btrfs_free_path(path);
1618 static int free_block_group_radix(struct radix_tree_root *radix)
1621 struct btrfs_block_group_cache *cache[8];
1625 ret = radix_tree_gang_lookup(radix, (void **)cache, 0,
1629 for (i = 0; i < ret; i++) {
1630 radix_tree_delete(radix, cache[i]->key.objectid +
1631 cache[i]->key.offset - 1);
1638 int btrfs_free_block_groups(struct btrfs_fs_info *info)
1642 unsigned long gang[16];
1645 ret = free_block_group_radix(&info->block_group_radix);
1646 ret2 = free_block_group_radix(&info->block_group_data_radix);
1653 ret = find_first_radix_bit(&info->extent_map_radix,
1654 gang, 0, ARRAY_SIZE(gang));
1657 for (i = 0; i < ret; i++) {
1658 clear_radix_bit(&info->extent_map_radix, gang[i]);
1664 int btrfs_read_block_groups(struct btrfs_root *root)
1666 struct btrfs_path *path;
1669 struct btrfs_block_group_item *bi;
1670 struct btrfs_block_group_cache *cache;
1671 struct btrfs_fs_info *info = root->fs_info;
1672 struct radix_tree_root *radix;
1673 struct btrfs_key key;
1674 struct btrfs_key found_key;
1675 struct btrfs_leaf *leaf;
1676 u64 group_size_blocks;
1679 group_size_blocks = BTRFS_BLOCK_GROUP_SIZE >>
1680 root->fs_info->sb->s_blocksize_bits;
1681 root = info->extent_root;
1683 key.offset = group_size_blocks;
1685 btrfs_set_key_type(&key, BTRFS_BLOCK_GROUP_ITEM_KEY);
1687 path = btrfs_alloc_path();
1692 ret = btrfs_search_slot(NULL, info->extent_root,
1698 leaf = btrfs_buffer_leaf(path->nodes[0]);
1699 btrfs_disk_key_to_cpu(&found_key,
1700 &leaf->items[path->slots[0]].key);
1701 cache = kmalloc(sizeof(*cache), GFP_NOFS);
1707 bi = btrfs_item_ptr(leaf, path->slots[0],
1708 struct btrfs_block_group_item);
1709 if (bi->flags & BTRFS_BLOCK_GROUP_DATA) {
1710 radix = &info->block_group_data_radix;
1713 radix = &info->block_group_radix;
1717 memcpy(&cache->item, bi, sizeof(*bi));
1718 memcpy(&cache->key, &found_key, sizeof(found_key));
1719 cache->last_alloc = cache->key.objectid;
1720 cache->first_free = cache->key.objectid;
1721 cache->last_prealloc = cache->key.objectid;
1725 cache->radix = radix;
1727 key.objectid = found_key.objectid + found_key.offset;
1728 btrfs_release_path(root, path);
1729 ret = radix_tree_insert(radix, found_key.objectid +
1730 found_key.offset - 1,
1733 used = btrfs_block_group_used(bi);
1734 if (used < div_factor(key.offset, 8)) {
1735 radix_tree_tag_set(radix, found_key.objectid +
1736 found_key.offset - 1,
1737 BTRFS_BLOCK_GROUP_AVAIL);
1740 btrfs_super_total_blocks(info->disk_super))
1744 btrfs_free_path(path);