1 // SPDX-License-Identifier: GPL-2.0
3 * Copyright (C) 2007 Oracle. All rights reserved.
6 #include <linux/sched.h>
7 #include <linux/sched/signal.h>
8 #include <linux/pagemap.h>
9 #include <linux/writeback.h>
10 #include <linux/blkdev.h>
11 #include <linux/sort.h>
12 #include <linux/rcupdate.h>
13 #include <linux/kthread.h>
14 #include <linux/slab.h>
15 #include <linux/ratelimit.h>
16 #include <linux/percpu_counter.h>
17 #include <linux/lockdep.h>
18 #include <linux/crc32c.h>
22 #include "print-tree.h"
26 #include "free-space-cache.h"
27 #include "free-space-tree.h"
30 #include "ref-verify.h"
31 #include "space-info.h"
32 #include "block-rsv.h"
33 #include "delalloc-space.h"
34 #include "block-group.h"
36 #include "rcu-string.h"
38 #undef SCRAMBLE_DELAYED_REFS
41 static int __btrfs_free_extent(struct btrfs_trans_handle *trans,
42 struct btrfs_delayed_ref_node *node, u64 parent,
43 u64 root_objectid, u64 owner_objectid,
44 u64 owner_offset, int refs_to_drop,
45 struct btrfs_delayed_extent_op *extra_op);
46 static void __run_delayed_extent_op(struct btrfs_delayed_extent_op *extent_op,
47 struct extent_buffer *leaf,
48 struct btrfs_extent_item *ei);
49 static int alloc_reserved_file_extent(struct btrfs_trans_handle *trans,
50 u64 parent, u64 root_objectid,
51 u64 flags, u64 owner, u64 offset,
52 struct btrfs_key *ins, int ref_mod);
53 static int alloc_reserved_tree_block(struct btrfs_trans_handle *trans,
54 struct btrfs_delayed_ref_node *node,
55 struct btrfs_delayed_extent_op *extent_op);
56 static int find_next_key(struct btrfs_path *path, int level,
57 struct btrfs_key *key);
59 static int block_group_bits(struct btrfs_block_group *cache, u64 bits)
61 return (cache->flags & bits) == bits;
64 int btrfs_add_excluded_extent(struct btrfs_fs_info *fs_info,
65 u64 start, u64 num_bytes)
67 u64 end = start + num_bytes - 1;
68 set_extent_bits(&fs_info->excluded_extents, start, end,
73 void btrfs_free_excluded_extents(struct btrfs_block_group *cache)
75 struct btrfs_fs_info *fs_info = cache->fs_info;
79 end = start + cache->length - 1;
81 clear_extent_bits(&fs_info->excluded_extents, start, end,
85 static u64 generic_ref_to_space_flags(struct btrfs_ref *ref)
87 if (ref->type == BTRFS_REF_METADATA) {
88 if (ref->tree_ref.root == BTRFS_CHUNK_TREE_OBJECTID)
89 return BTRFS_BLOCK_GROUP_SYSTEM;
91 return BTRFS_BLOCK_GROUP_METADATA;
93 return BTRFS_BLOCK_GROUP_DATA;
96 static void add_pinned_bytes(struct btrfs_fs_info *fs_info,
97 struct btrfs_ref *ref)
99 struct btrfs_space_info *space_info;
100 u64 flags = generic_ref_to_space_flags(ref);
102 space_info = btrfs_find_space_info(fs_info, flags);
104 percpu_counter_add_batch(&space_info->total_bytes_pinned, ref->len,
105 BTRFS_TOTAL_BYTES_PINNED_BATCH);
108 static void sub_pinned_bytes(struct btrfs_fs_info *fs_info,
109 struct btrfs_ref *ref)
111 struct btrfs_space_info *space_info;
112 u64 flags = generic_ref_to_space_flags(ref);
114 space_info = btrfs_find_space_info(fs_info, flags);
116 percpu_counter_add_batch(&space_info->total_bytes_pinned, -ref->len,
117 BTRFS_TOTAL_BYTES_PINNED_BATCH);
120 /* simple helper to search for an existing data extent at a given offset */
121 int btrfs_lookup_data_extent(struct btrfs_fs_info *fs_info, u64 start, u64 len)
124 struct btrfs_key key;
125 struct btrfs_path *path;
127 path = btrfs_alloc_path();
131 key.objectid = start;
133 key.type = BTRFS_EXTENT_ITEM_KEY;
134 ret = btrfs_search_slot(NULL, fs_info->extent_root, &key, path, 0, 0);
135 btrfs_free_path(path);
140 * helper function to lookup reference count and flags of a tree block.
142 * the head node for delayed ref is used to store the sum of all the
143 * reference count modifications queued up in the rbtree. the head
144 * node may also store the extent flags to set. This way you can check
145 * to see what the reference count and extent flags would be if all of
146 * the delayed refs are not processed.
148 int btrfs_lookup_extent_info(struct btrfs_trans_handle *trans,
149 struct btrfs_fs_info *fs_info, u64 bytenr,
150 u64 offset, int metadata, u64 *refs, u64 *flags)
152 struct btrfs_delayed_ref_head *head;
153 struct btrfs_delayed_ref_root *delayed_refs;
154 struct btrfs_path *path;
155 struct btrfs_extent_item *ei;
156 struct extent_buffer *leaf;
157 struct btrfs_key key;
164 * If we don't have skinny metadata, don't bother doing anything
167 if (metadata && !btrfs_fs_incompat(fs_info, SKINNY_METADATA)) {
168 offset = fs_info->nodesize;
172 path = btrfs_alloc_path();
177 path->skip_locking = 1;
178 path->search_commit_root = 1;
182 key.objectid = bytenr;
185 key.type = BTRFS_METADATA_ITEM_KEY;
187 key.type = BTRFS_EXTENT_ITEM_KEY;
189 ret = btrfs_search_slot(trans, fs_info->extent_root, &key, path, 0, 0);
193 if (ret > 0 && metadata && key.type == BTRFS_METADATA_ITEM_KEY) {
194 if (path->slots[0]) {
196 btrfs_item_key_to_cpu(path->nodes[0], &key,
198 if (key.objectid == bytenr &&
199 key.type == BTRFS_EXTENT_ITEM_KEY &&
200 key.offset == fs_info->nodesize)
206 leaf = path->nodes[0];
207 item_size = btrfs_item_size_nr(leaf, path->slots[0]);
208 if (item_size >= sizeof(*ei)) {
209 ei = btrfs_item_ptr(leaf, path->slots[0],
210 struct btrfs_extent_item);
211 num_refs = btrfs_extent_refs(leaf, ei);
212 extent_flags = btrfs_extent_flags(leaf, ei);
215 btrfs_print_v0_err(fs_info);
217 btrfs_abort_transaction(trans, ret);
219 btrfs_handle_fs_error(fs_info, ret, NULL);
224 BUG_ON(num_refs == 0);
234 delayed_refs = &trans->transaction->delayed_refs;
235 spin_lock(&delayed_refs->lock);
236 head = btrfs_find_delayed_ref_head(delayed_refs, bytenr);
238 if (!mutex_trylock(&head->mutex)) {
239 refcount_inc(&head->refs);
240 spin_unlock(&delayed_refs->lock);
242 btrfs_release_path(path);
245 * Mutex was contended, block until it's released and try
248 mutex_lock(&head->mutex);
249 mutex_unlock(&head->mutex);
250 btrfs_put_delayed_ref_head(head);
253 spin_lock(&head->lock);
254 if (head->extent_op && head->extent_op->update_flags)
255 extent_flags |= head->extent_op->flags_to_set;
257 BUG_ON(num_refs == 0);
259 num_refs += head->ref_mod;
260 spin_unlock(&head->lock);
261 mutex_unlock(&head->mutex);
263 spin_unlock(&delayed_refs->lock);
265 WARN_ON(num_refs == 0);
269 *flags = extent_flags;
271 btrfs_free_path(path);
276 * Back reference rules. Back refs have three main goals:
278 * 1) differentiate between all holders of references to an extent so that
279 * when a reference is dropped we can make sure it was a valid reference
280 * before freeing the extent.
282 * 2) Provide enough information to quickly find the holders of an extent
283 * if we notice a given block is corrupted or bad.
285 * 3) Make it easy to migrate blocks for FS shrinking or storage pool
286 * maintenance. This is actually the same as #2, but with a slightly
287 * different use case.
289 * There are two kinds of back refs. The implicit back refs is optimized
290 * for pointers in non-shared tree blocks. For a given pointer in a block,
291 * back refs of this kind provide information about the block's owner tree
292 * and the pointer's key. These information allow us to find the block by
293 * b-tree searching. The full back refs is for pointers in tree blocks not
294 * referenced by their owner trees. The location of tree block is recorded
295 * in the back refs. Actually the full back refs is generic, and can be
296 * used in all cases the implicit back refs is used. The major shortcoming
297 * of the full back refs is its overhead. Every time a tree block gets
298 * COWed, we have to update back refs entry for all pointers in it.
300 * For a newly allocated tree block, we use implicit back refs for
301 * pointers in it. This means most tree related operations only involve
302 * implicit back refs. For a tree block created in old transaction, the
303 * only way to drop a reference to it is COW it. So we can detect the
304 * event that tree block loses its owner tree's reference and do the
305 * back refs conversion.
307 * When a tree block is COWed through a tree, there are four cases:
309 * The reference count of the block is one and the tree is the block's
310 * owner tree. Nothing to do in this case.
312 * The reference count of the block is one and the tree is not the
313 * block's owner tree. In this case, full back refs is used for pointers
314 * in the block. Remove these full back refs, add implicit back refs for
315 * every pointers in the new block.
317 * The reference count of the block is greater than one and the tree is
318 * the block's owner tree. In this case, implicit back refs is used for
319 * pointers in the block. Add full back refs for every pointers in the
320 * block, increase lower level extents' reference counts. The original
321 * implicit back refs are entailed to the new block.
323 * The reference count of the block is greater than one and the tree is
324 * not the block's owner tree. Add implicit back refs for every pointer in
325 * the new block, increase lower level extents' reference count.
327 * Back Reference Key composing:
329 * The key objectid corresponds to the first byte in the extent,
330 * The key type is used to differentiate between types of back refs.
331 * There are different meanings of the key offset for different types
334 * File extents can be referenced by:
336 * - multiple snapshots, subvolumes, or different generations in one subvol
337 * - different files inside a single subvolume
338 * - different offsets inside a file (bookend extents in file.c)
340 * The extent ref structure for the implicit back refs has fields for:
342 * - Objectid of the subvolume root
343 * - objectid of the file holding the reference
344 * - original offset in the file
345 * - how many bookend extents
347 * The key offset for the implicit back refs is hash of the first
350 * The extent ref structure for the full back refs has field for:
352 * - number of pointers in the tree leaf
354 * The key offset for the implicit back refs is the first byte of
357 * When a file extent is allocated, The implicit back refs is used.
358 * the fields are filled in:
360 * (root_key.objectid, inode objectid, offset in file, 1)
362 * When a file extent is removed file truncation, we find the
363 * corresponding implicit back refs and check the following fields:
365 * (btrfs_header_owner(leaf), inode objectid, offset in file)
367 * Btree extents can be referenced by:
369 * - Different subvolumes
371 * Both the implicit back refs and the full back refs for tree blocks
372 * only consist of key. The key offset for the implicit back refs is
373 * objectid of block's owner tree. The key offset for the full back refs
374 * is the first byte of parent block.
376 * When implicit back refs is used, information about the lowest key and
377 * level of the tree block are required. These information are stored in
378 * tree block info structure.
382 * is_data == BTRFS_REF_TYPE_BLOCK, tree block type is required,
383 * is_data == BTRFS_REF_TYPE_DATA, data type is requiried,
384 * is_data == BTRFS_REF_TYPE_ANY, either type is OK.
386 int btrfs_get_extent_inline_ref_type(const struct extent_buffer *eb,
387 struct btrfs_extent_inline_ref *iref,
388 enum btrfs_inline_ref_type is_data)
390 int type = btrfs_extent_inline_ref_type(eb, iref);
391 u64 offset = btrfs_extent_inline_ref_offset(eb, iref);
393 if (type == BTRFS_TREE_BLOCK_REF_KEY ||
394 type == BTRFS_SHARED_BLOCK_REF_KEY ||
395 type == BTRFS_SHARED_DATA_REF_KEY ||
396 type == BTRFS_EXTENT_DATA_REF_KEY) {
397 if (is_data == BTRFS_REF_TYPE_BLOCK) {
398 if (type == BTRFS_TREE_BLOCK_REF_KEY)
400 if (type == BTRFS_SHARED_BLOCK_REF_KEY) {
403 * Every shared one has parent tree block,
404 * which must be aligned to sector size.
407 IS_ALIGNED(offset, eb->fs_info->sectorsize))
410 } else if (is_data == BTRFS_REF_TYPE_DATA) {
411 if (type == BTRFS_EXTENT_DATA_REF_KEY)
413 if (type == BTRFS_SHARED_DATA_REF_KEY) {
416 * Every shared one has parent tree block,
417 * which must be aligned to sector size.
420 IS_ALIGNED(offset, eb->fs_info->sectorsize))
424 ASSERT(is_data == BTRFS_REF_TYPE_ANY);
429 btrfs_print_leaf((struct extent_buffer *)eb);
430 btrfs_err(eb->fs_info,
431 "eb %llu iref 0x%lx invalid extent inline ref type %d",
432 eb->start, (unsigned long)iref, type);
435 return BTRFS_REF_TYPE_INVALID;
438 u64 hash_extent_data_ref(u64 root_objectid, u64 owner, u64 offset)
440 u32 high_crc = ~(u32)0;
441 u32 low_crc = ~(u32)0;
444 lenum = cpu_to_le64(root_objectid);
445 high_crc = btrfs_crc32c(high_crc, &lenum, sizeof(lenum));
446 lenum = cpu_to_le64(owner);
447 low_crc = btrfs_crc32c(low_crc, &lenum, sizeof(lenum));
448 lenum = cpu_to_le64(offset);
449 low_crc = btrfs_crc32c(low_crc, &lenum, sizeof(lenum));
451 return ((u64)high_crc << 31) ^ (u64)low_crc;
454 static u64 hash_extent_data_ref_item(struct extent_buffer *leaf,
455 struct btrfs_extent_data_ref *ref)
457 return hash_extent_data_ref(btrfs_extent_data_ref_root(leaf, ref),
458 btrfs_extent_data_ref_objectid(leaf, ref),
459 btrfs_extent_data_ref_offset(leaf, ref));
462 static int match_extent_data_ref(struct extent_buffer *leaf,
463 struct btrfs_extent_data_ref *ref,
464 u64 root_objectid, u64 owner, u64 offset)
466 if (btrfs_extent_data_ref_root(leaf, ref) != root_objectid ||
467 btrfs_extent_data_ref_objectid(leaf, ref) != owner ||
468 btrfs_extent_data_ref_offset(leaf, ref) != offset)
473 static noinline int lookup_extent_data_ref(struct btrfs_trans_handle *trans,
474 struct btrfs_path *path,
475 u64 bytenr, u64 parent,
477 u64 owner, u64 offset)
479 struct btrfs_root *root = trans->fs_info->extent_root;
480 struct btrfs_key key;
481 struct btrfs_extent_data_ref *ref;
482 struct extent_buffer *leaf;
488 key.objectid = bytenr;
490 key.type = BTRFS_SHARED_DATA_REF_KEY;
493 key.type = BTRFS_EXTENT_DATA_REF_KEY;
494 key.offset = hash_extent_data_ref(root_objectid,
499 ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
511 leaf = path->nodes[0];
512 nritems = btrfs_header_nritems(leaf);
514 if (path->slots[0] >= nritems) {
515 ret = btrfs_next_leaf(root, path);
521 leaf = path->nodes[0];
522 nritems = btrfs_header_nritems(leaf);
526 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
527 if (key.objectid != bytenr ||
528 key.type != BTRFS_EXTENT_DATA_REF_KEY)
531 ref = btrfs_item_ptr(leaf, path->slots[0],
532 struct btrfs_extent_data_ref);
534 if (match_extent_data_ref(leaf, ref, root_objectid,
537 btrfs_release_path(path);
549 static noinline int insert_extent_data_ref(struct btrfs_trans_handle *trans,
550 struct btrfs_path *path,
551 u64 bytenr, u64 parent,
552 u64 root_objectid, u64 owner,
553 u64 offset, int refs_to_add)
555 struct btrfs_root *root = trans->fs_info->extent_root;
556 struct btrfs_key key;
557 struct extent_buffer *leaf;
562 key.objectid = bytenr;
564 key.type = BTRFS_SHARED_DATA_REF_KEY;
566 size = sizeof(struct btrfs_shared_data_ref);
568 key.type = BTRFS_EXTENT_DATA_REF_KEY;
569 key.offset = hash_extent_data_ref(root_objectid,
571 size = sizeof(struct btrfs_extent_data_ref);
574 ret = btrfs_insert_empty_item(trans, root, path, &key, size);
575 if (ret && ret != -EEXIST)
578 leaf = path->nodes[0];
580 struct btrfs_shared_data_ref *ref;
581 ref = btrfs_item_ptr(leaf, path->slots[0],
582 struct btrfs_shared_data_ref);
584 btrfs_set_shared_data_ref_count(leaf, ref, refs_to_add);
586 num_refs = btrfs_shared_data_ref_count(leaf, ref);
587 num_refs += refs_to_add;
588 btrfs_set_shared_data_ref_count(leaf, ref, num_refs);
591 struct btrfs_extent_data_ref *ref;
592 while (ret == -EEXIST) {
593 ref = btrfs_item_ptr(leaf, path->slots[0],
594 struct btrfs_extent_data_ref);
595 if (match_extent_data_ref(leaf, ref, root_objectid,
598 btrfs_release_path(path);
600 ret = btrfs_insert_empty_item(trans, root, path, &key,
602 if (ret && ret != -EEXIST)
605 leaf = path->nodes[0];
607 ref = btrfs_item_ptr(leaf, path->slots[0],
608 struct btrfs_extent_data_ref);
610 btrfs_set_extent_data_ref_root(leaf, ref,
612 btrfs_set_extent_data_ref_objectid(leaf, ref, owner);
613 btrfs_set_extent_data_ref_offset(leaf, ref, offset);
614 btrfs_set_extent_data_ref_count(leaf, ref, refs_to_add);
616 num_refs = btrfs_extent_data_ref_count(leaf, ref);
617 num_refs += refs_to_add;
618 btrfs_set_extent_data_ref_count(leaf, ref, num_refs);
621 btrfs_mark_buffer_dirty(leaf);
624 btrfs_release_path(path);
628 static noinline int remove_extent_data_ref(struct btrfs_trans_handle *trans,
629 struct btrfs_path *path,
630 int refs_to_drop, int *last_ref)
632 struct btrfs_key key;
633 struct btrfs_extent_data_ref *ref1 = NULL;
634 struct btrfs_shared_data_ref *ref2 = NULL;
635 struct extent_buffer *leaf;
639 leaf = path->nodes[0];
640 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
642 if (key.type == BTRFS_EXTENT_DATA_REF_KEY) {
643 ref1 = btrfs_item_ptr(leaf, path->slots[0],
644 struct btrfs_extent_data_ref);
645 num_refs = btrfs_extent_data_ref_count(leaf, ref1);
646 } else if (key.type == BTRFS_SHARED_DATA_REF_KEY) {
647 ref2 = btrfs_item_ptr(leaf, path->slots[0],
648 struct btrfs_shared_data_ref);
649 num_refs = btrfs_shared_data_ref_count(leaf, ref2);
650 } else if (unlikely(key.type == BTRFS_EXTENT_REF_V0_KEY)) {
651 btrfs_print_v0_err(trans->fs_info);
652 btrfs_abort_transaction(trans, -EINVAL);
658 BUG_ON(num_refs < refs_to_drop);
659 num_refs -= refs_to_drop;
662 ret = btrfs_del_item(trans, trans->fs_info->extent_root, path);
665 if (key.type == BTRFS_EXTENT_DATA_REF_KEY)
666 btrfs_set_extent_data_ref_count(leaf, ref1, num_refs);
667 else if (key.type == BTRFS_SHARED_DATA_REF_KEY)
668 btrfs_set_shared_data_ref_count(leaf, ref2, num_refs);
669 btrfs_mark_buffer_dirty(leaf);
674 static noinline u32 extent_data_ref_count(struct btrfs_path *path,
675 struct btrfs_extent_inline_ref *iref)
677 struct btrfs_key key;
678 struct extent_buffer *leaf;
679 struct btrfs_extent_data_ref *ref1;
680 struct btrfs_shared_data_ref *ref2;
684 leaf = path->nodes[0];
685 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
687 BUG_ON(key.type == BTRFS_EXTENT_REF_V0_KEY);
690 * If type is invalid, we should have bailed out earlier than
693 type = btrfs_get_extent_inline_ref_type(leaf, iref, BTRFS_REF_TYPE_DATA);
694 ASSERT(type != BTRFS_REF_TYPE_INVALID);
695 if (type == BTRFS_EXTENT_DATA_REF_KEY) {
696 ref1 = (struct btrfs_extent_data_ref *)(&iref->offset);
697 num_refs = btrfs_extent_data_ref_count(leaf, ref1);
699 ref2 = (struct btrfs_shared_data_ref *)(iref + 1);
700 num_refs = btrfs_shared_data_ref_count(leaf, ref2);
702 } else if (key.type == BTRFS_EXTENT_DATA_REF_KEY) {
703 ref1 = btrfs_item_ptr(leaf, path->slots[0],
704 struct btrfs_extent_data_ref);
705 num_refs = btrfs_extent_data_ref_count(leaf, ref1);
706 } else if (key.type == BTRFS_SHARED_DATA_REF_KEY) {
707 ref2 = btrfs_item_ptr(leaf, path->slots[0],
708 struct btrfs_shared_data_ref);
709 num_refs = btrfs_shared_data_ref_count(leaf, ref2);
716 static noinline int lookup_tree_block_ref(struct btrfs_trans_handle *trans,
717 struct btrfs_path *path,
718 u64 bytenr, u64 parent,
721 struct btrfs_root *root = trans->fs_info->extent_root;
722 struct btrfs_key key;
725 key.objectid = bytenr;
727 key.type = BTRFS_SHARED_BLOCK_REF_KEY;
730 key.type = BTRFS_TREE_BLOCK_REF_KEY;
731 key.offset = root_objectid;
734 ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
740 static noinline int insert_tree_block_ref(struct btrfs_trans_handle *trans,
741 struct btrfs_path *path,
742 u64 bytenr, u64 parent,
745 struct btrfs_key key;
748 key.objectid = bytenr;
750 key.type = BTRFS_SHARED_BLOCK_REF_KEY;
753 key.type = BTRFS_TREE_BLOCK_REF_KEY;
754 key.offset = root_objectid;
757 ret = btrfs_insert_empty_item(trans, trans->fs_info->extent_root,
759 btrfs_release_path(path);
763 static inline int extent_ref_type(u64 parent, u64 owner)
766 if (owner < BTRFS_FIRST_FREE_OBJECTID) {
768 type = BTRFS_SHARED_BLOCK_REF_KEY;
770 type = BTRFS_TREE_BLOCK_REF_KEY;
773 type = BTRFS_SHARED_DATA_REF_KEY;
775 type = BTRFS_EXTENT_DATA_REF_KEY;
780 static int find_next_key(struct btrfs_path *path, int level,
781 struct btrfs_key *key)
784 for (; level < BTRFS_MAX_LEVEL; level++) {
785 if (!path->nodes[level])
787 if (path->slots[level] + 1 >=
788 btrfs_header_nritems(path->nodes[level]))
791 btrfs_item_key_to_cpu(path->nodes[level], key,
792 path->slots[level] + 1);
794 btrfs_node_key_to_cpu(path->nodes[level], key,
795 path->slots[level] + 1);
802 * look for inline back ref. if back ref is found, *ref_ret is set
803 * to the address of inline back ref, and 0 is returned.
805 * if back ref isn't found, *ref_ret is set to the address where it
806 * should be inserted, and -ENOENT is returned.
808 * if insert is true and there are too many inline back refs, the path
809 * points to the extent item, and -EAGAIN is returned.
811 * NOTE: inline back refs are ordered in the same way that back ref
812 * items in the tree are ordered.
814 static noinline_for_stack
815 int lookup_inline_extent_backref(struct btrfs_trans_handle *trans,
816 struct btrfs_path *path,
817 struct btrfs_extent_inline_ref **ref_ret,
818 u64 bytenr, u64 num_bytes,
819 u64 parent, u64 root_objectid,
820 u64 owner, u64 offset, int insert)
822 struct btrfs_fs_info *fs_info = trans->fs_info;
823 struct btrfs_root *root = fs_info->extent_root;
824 struct btrfs_key key;
825 struct extent_buffer *leaf;
826 struct btrfs_extent_item *ei;
827 struct btrfs_extent_inline_ref *iref;
837 bool skinny_metadata = btrfs_fs_incompat(fs_info, SKINNY_METADATA);
840 key.objectid = bytenr;
841 key.type = BTRFS_EXTENT_ITEM_KEY;
842 key.offset = num_bytes;
844 want = extent_ref_type(parent, owner);
846 extra_size = btrfs_extent_inline_ref_size(want);
847 path->keep_locks = 1;
852 * Owner is our level, so we can just add one to get the level for the
853 * block we are interested in.
855 if (skinny_metadata && owner < BTRFS_FIRST_FREE_OBJECTID) {
856 key.type = BTRFS_METADATA_ITEM_KEY;
861 ret = btrfs_search_slot(trans, root, &key, path, extra_size, 1);
868 * We may be a newly converted file system which still has the old fat
869 * extent entries for metadata, so try and see if we have one of those.
871 if (ret > 0 && skinny_metadata) {
872 skinny_metadata = false;
873 if (path->slots[0]) {
875 btrfs_item_key_to_cpu(path->nodes[0], &key,
877 if (key.objectid == bytenr &&
878 key.type == BTRFS_EXTENT_ITEM_KEY &&
879 key.offset == num_bytes)
883 key.objectid = bytenr;
884 key.type = BTRFS_EXTENT_ITEM_KEY;
885 key.offset = num_bytes;
886 btrfs_release_path(path);
891 if (ret && !insert) {
894 } else if (WARN_ON(ret)) {
899 leaf = path->nodes[0];
900 item_size = btrfs_item_size_nr(leaf, path->slots[0]);
901 if (unlikely(item_size < sizeof(*ei))) {
903 btrfs_print_v0_err(fs_info);
904 btrfs_abort_transaction(trans, err);
908 ei = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
909 flags = btrfs_extent_flags(leaf, ei);
911 ptr = (unsigned long)(ei + 1);
912 end = (unsigned long)ei + item_size;
914 if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK && !skinny_metadata) {
915 ptr += sizeof(struct btrfs_tree_block_info);
919 if (owner >= BTRFS_FIRST_FREE_OBJECTID)
920 needed = BTRFS_REF_TYPE_DATA;
922 needed = BTRFS_REF_TYPE_BLOCK;
930 iref = (struct btrfs_extent_inline_ref *)ptr;
931 type = btrfs_get_extent_inline_ref_type(leaf, iref, needed);
932 if (type == BTRFS_REF_TYPE_INVALID) {
940 ptr += btrfs_extent_inline_ref_size(type);
944 if (type == BTRFS_EXTENT_DATA_REF_KEY) {
945 struct btrfs_extent_data_ref *dref;
946 dref = (struct btrfs_extent_data_ref *)(&iref->offset);
947 if (match_extent_data_ref(leaf, dref, root_objectid,
952 if (hash_extent_data_ref_item(leaf, dref) <
953 hash_extent_data_ref(root_objectid, owner, offset))
957 ref_offset = btrfs_extent_inline_ref_offset(leaf, iref);
959 if (parent == ref_offset) {
963 if (ref_offset < parent)
966 if (root_objectid == ref_offset) {
970 if (ref_offset < root_objectid)
974 ptr += btrfs_extent_inline_ref_size(type);
976 if (err == -ENOENT && insert) {
977 if (item_size + extra_size >=
978 BTRFS_MAX_EXTENT_ITEM_SIZE(root)) {
983 * To add new inline back ref, we have to make sure
984 * there is no corresponding back ref item.
985 * For simplicity, we just do not add new inline back
986 * ref if there is any kind of item for this block
988 if (find_next_key(path, 0, &key) == 0 &&
989 key.objectid == bytenr &&
990 key.type < BTRFS_BLOCK_GROUP_ITEM_KEY) {
995 *ref_ret = (struct btrfs_extent_inline_ref *)ptr;
998 path->keep_locks = 0;
999 btrfs_unlock_up_safe(path, 1);
1005 * helper to add new inline back ref
1007 static noinline_for_stack
1008 void setup_inline_extent_backref(struct btrfs_fs_info *fs_info,
1009 struct btrfs_path *path,
1010 struct btrfs_extent_inline_ref *iref,
1011 u64 parent, u64 root_objectid,
1012 u64 owner, u64 offset, int refs_to_add,
1013 struct btrfs_delayed_extent_op *extent_op)
1015 struct extent_buffer *leaf;
1016 struct btrfs_extent_item *ei;
1019 unsigned long item_offset;
1024 leaf = path->nodes[0];
1025 ei = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
1026 item_offset = (unsigned long)iref - (unsigned long)ei;
1028 type = extent_ref_type(parent, owner);
1029 size = btrfs_extent_inline_ref_size(type);
1031 btrfs_extend_item(path, size);
1033 ei = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
1034 refs = btrfs_extent_refs(leaf, ei);
1035 refs += refs_to_add;
1036 btrfs_set_extent_refs(leaf, ei, refs);
1038 __run_delayed_extent_op(extent_op, leaf, ei);
1040 ptr = (unsigned long)ei + item_offset;
1041 end = (unsigned long)ei + btrfs_item_size_nr(leaf, path->slots[0]);
1042 if (ptr < end - size)
1043 memmove_extent_buffer(leaf, ptr + size, ptr,
1046 iref = (struct btrfs_extent_inline_ref *)ptr;
1047 btrfs_set_extent_inline_ref_type(leaf, iref, type);
1048 if (type == BTRFS_EXTENT_DATA_REF_KEY) {
1049 struct btrfs_extent_data_ref *dref;
1050 dref = (struct btrfs_extent_data_ref *)(&iref->offset);
1051 btrfs_set_extent_data_ref_root(leaf, dref, root_objectid);
1052 btrfs_set_extent_data_ref_objectid(leaf, dref, owner);
1053 btrfs_set_extent_data_ref_offset(leaf, dref, offset);
1054 btrfs_set_extent_data_ref_count(leaf, dref, refs_to_add);
1055 } else if (type == BTRFS_SHARED_DATA_REF_KEY) {
1056 struct btrfs_shared_data_ref *sref;
1057 sref = (struct btrfs_shared_data_ref *)(iref + 1);
1058 btrfs_set_shared_data_ref_count(leaf, sref, refs_to_add);
1059 btrfs_set_extent_inline_ref_offset(leaf, iref, parent);
1060 } else if (type == BTRFS_SHARED_BLOCK_REF_KEY) {
1061 btrfs_set_extent_inline_ref_offset(leaf, iref, parent);
1063 btrfs_set_extent_inline_ref_offset(leaf, iref, root_objectid);
1065 btrfs_mark_buffer_dirty(leaf);
1068 static int lookup_extent_backref(struct btrfs_trans_handle *trans,
1069 struct btrfs_path *path,
1070 struct btrfs_extent_inline_ref **ref_ret,
1071 u64 bytenr, u64 num_bytes, u64 parent,
1072 u64 root_objectid, u64 owner, u64 offset)
1076 ret = lookup_inline_extent_backref(trans, path, ref_ret, bytenr,
1077 num_bytes, parent, root_objectid,
1082 btrfs_release_path(path);
1085 if (owner < BTRFS_FIRST_FREE_OBJECTID) {
1086 ret = lookup_tree_block_ref(trans, path, bytenr, parent,
1089 ret = lookup_extent_data_ref(trans, path, bytenr, parent,
1090 root_objectid, owner, offset);
1096 * helper to update/remove inline back ref
1098 static noinline_for_stack
1099 void update_inline_extent_backref(struct btrfs_path *path,
1100 struct btrfs_extent_inline_ref *iref,
1102 struct btrfs_delayed_extent_op *extent_op,
1105 struct extent_buffer *leaf = path->nodes[0];
1106 struct btrfs_extent_item *ei;
1107 struct btrfs_extent_data_ref *dref = NULL;
1108 struct btrfs_shared_data_ref *sref = NULL;
1116 ei = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
1117 refs = btrfs_extent_refs(leaf, ei);
1118 WARN_ON(refs_to_mod < 0 && refs + refs_to_mod <= 0);
1119 refs += refs_to_mod;
1120 btrfs_set_extent_refs(leaf, ei, refs);
1122 __run_delayed_extent_op(extent_op, leaf, ei);
1125 * If type is invalid, we should have bailed out after
1126 * lookup_inline_extent_backref().
1128 type = btrfs_get_extent_inline_ref_type(leaf, iref, BTRFS_REF_TYPE_ANY);
1129 ASSERT(type != BTRFS_REF_TYPE_INVALID);
1131 if (type == BTRFS_EXTENT_DATA_REF_KEY) {
1132 dref = (struct btrfs_extent_data_ref *)(&iref->offset);
1133 refs = btrfs_extent_data_ref_count(leaf, dref);
1134 } else if (type == BTRFS_SHARED_DATA_REF_KEY) {
1135 sref = (struct btrfs_shared_data_ref *)(iref + 1);
1136 refs = btrfs_shared_data_ref_count(leaf, sref);
1139 BUG_ON(refs_to_mod != -1);
1142 BUG_ON(refs_to_mod < 0 && refs < -refs_to_mod);
1143 refs += refs_to_mod;
1146 if (type == BTRFS_EXTENT_DATA_REF_KEY)
1147 btrfs_set_extent_data_ref_count(leaf, dref, refs);
1149 btrfs_set_shared_data_ref_count(leaf, sref, refs);
1152 size = btrfs_extent_inline_ref_size(type);
1153 item_size = btrfs_item_size_nr(leaf, path->slots[0]);
1154 ptr = (unsigned long)iref;
1155 end = (unsigned long)ei + item_size;
1156 if (ptr + size < end)
1157 memmove_extent_buffer(leaf, ptr, ptr + size,
1160 btrfs_truncate_item(path, item_size, 1);
1162 btrfs_mark_buffer_dirty(leaf);
1165 static noinline_for_stack
1166 int insert_inline_extent_backref(struct btrfs_trans_handle *trans,
1167 struct btrfs_path *path,
1168 u64 bytenr, u64 num_bytes, u64 parent,
1169 u64 root_objectid, u64 owner,
1170 u64 offset, int refs_to_add,
1171 struct btrfs_delayed_extent_op *extent_op)
1173 struct btrfs_extent_inline_ref *iref;
1176 ret = lookup_inline_extent_backref(trans, path, &iref, bytenr,
1177 num_bytes, parent, root_objectid,
1181 * We're adding refs to a tree block we already own, this
1182 * should not happen at all.
1184 if (owner < BTRFS_FIRST_FREE_OBJECTID) {
1185 btrfs_crit(trans->fs_info,
1186 "adding refs to an existing tree ref, bytenr %llu num_bytes %llu root_objectid %llu",
1187 bytenr, num_bytes, root_objectid);
1188 if (IS_ENABLED(CONFIG_BTRFS_DEBUG)) {
1190 btrfs_crit(trans->fs_info,
1191 "path->slots[0]=%d path->nodes[0]:", path->slots[0]);
1192 btrfs_print_leaf(path->nodes[0]);
1196 update_inline_extent_backref(path, iref, refs_to_add,
1198 } else if (ret == -ENOENT) {
1199 setup_inline_extent_backref(trans->fs_info, path, iref, parent,
1200 root_objectid, owner, offset,
1201 refs_to_add, extent_op);
1207 static int remove_extent_backref(struct btrfs_trans_handle *trans,
1208 struct btrfs_path *path,
1209 struct btrfs_extent_inline_ref *iref,
1210 int refs_to_drop, int is_data, int *last_ref)
1214 BUG_ON(!is_data && refs_to_drop != 1);
1216 update_inline_extent_backref(path, iref, -refs_to_drop, NULL,
1218 } else if (is_data) {
1219 ret = remove_extent_data_ref(trans, path, refs_to_drop,
1223 ret = btrfs_del_item(trans, trans->fs_info->extent_root, path);
1228 static int btrfs_issue_discard(struct block_device *bdev, u64 start, u64 len,
1229 u64 *discarded_bytes)
1232 u64 bytes_left, end;
1233 u64 aligned_start = ALIGN(start, 1 << 9);
1235 if (WARN_ON(start != aligned_start)) {
1236 len -= aligned_start - start;
1237 len = round_down(len, 1 << 9);
1238 start = aligned_start;
1241 *discarded_bytes = 0;
1249 /* Skip any superblocks on this device. */
1250 for (j = 0; j < BTRFS_SUPER_MIRROR_MAX; j++) {
1251 u64 sb_start = btrfs_sb_offset(j);
1252 u64 sb_end = sb_start + BTRFS_SUPER_INFO_SIZE;
1253 u64 size = sb_start - start;
1255 if (!in_range(sb_start, start, bytes_left) &&
1256 !in_range(sb_end, start, bytes_left) &&
1257 !in_range(start, sb_start, BTRFS_SUPER_INFO_SIZE))
1261 * Superblock spans beginning of range. Adjust start and
1264 if (sb_start <= start) {
1265 start += sb_end - start;
1270 bytes_left = end - start;
1275 ret = blkdev_issue_discard(bdev, start >> 9, size >> 9,
1278 *discarded_bytes += size;
1279 else if (ret != -EOPNOTSUPP)
1288 bytes_left = end - start;
1292 ret = blkdev_issue_discard(bdev, start >> 9, bytes_left >> 9,
1295 *discarded_bytes += bytes_left;
1300 int btrfs_discard_extent(struct btrfs_fs_info *fs_info, u64 bytenr,
1301 u64 num_bytes, u64 *actual_bytes)
1304 u64 discarded_bytes = 0;
1305 u64 end = bytenr + num_bytes;
1307 struct btrfs_bio *bbio = NULL;
1311 * Avoid races with device replace and make sure our bbio has devices
1312 * associated to its stripes that don't go away while we are discarding.
1314 btrfs_bio_counter_inc_blocked(fs_info);
1316 struct btrfs_bio_stripe *stripe;
1319 num_bytes = end - cur;
1320 /* Tell the block device(s) that the sectors can be discarded */
1321 ret = btrfs_map_block(fs_info, BTRFS_MAP_DISCARD, cur,
1322 &num_bytes, &bbio, 0);
1324 * Error can be -ENOMEM, -ENOENT (no such chunk mapping) or
1325 * -EOPNOTSUPP. For any such error, @num_bytes is not updated,
1326 * thus we can't continue anyway.
1331 stripe = bbio->stripes;
1332 for (i = 0; i < bbio->num_stripes; i++, stripe++) {
1334 struct request_queue *req_q;
1336 if (!stripe->dev->bdev) {
1337 ASSERT(btrfs_test_opt(fs_info, DEGRADED));
1340 req_q = bdev_get_queue(stripe->dev->bdev);
1341 if (!blk_queue_discard(req_q))
1344 ret = btrfs_issue_discard(stripe->dev->bdev,
1349 discarded_bytes += bytes;
1350 } else if (ret != -EOPNOTSUPP) {
1352 * Logic errors or -ENOMEM, or -EIO, but
1353 * unlikely to happen.
1355 * And since there are two loops, explicitly
1356 * go to out to avoid confusion.
1358 btrfs_put_bbio(bbio);
1363 * Just in case we get back EOPNOTSUPP for some reason,
1364 * just ignore the return value so we don't screw up
1365 * people calling discard_extent.
1369 btrfs_put_bbio(bbio);
1373 btrfs_bio_counter_dec(fs_info);
1376 *actual_bytes = discarded_bytes;
1379 if (ret == -EOPNOTSUPP)
1384 /* Can return -ENOMEM */
1385 int btrfs_inc_extent_ref(struct btrfs_trans_handle *trans,
1386 struct btrfs_ref *generic_ref)
1388 struct btrfs_fs_info *fs_info = trans->fs_info;
1389 int old_ref_mod, new_ref_mod;
1392 ASSERT(generic_ref->type != BTRFS_REF_NOT_SET &&
1393 generic_ref->action);
1394 BUG_ON(generic_ref->type == BTRFS_REF_METADATA &&
1395 generic_ref->tree_ref.root == BTRFS_TREE_LOG_OBJECTID);
1397 if (generic_ref->type == BTRFS_REF_METADATA)
1398 ret = btrfs_add_delayed_tree_ref(trans, generic_ref,
1399 NULL, &old_ref_mod, &new_ref_mod);
1401 ret = btrfs_add_delayed_data_ref(trans, generic_ref, 0,
1402 &old_ref_mod, &new_ref_mod);
1404 btrfs_ref_tree_mod(fs_info, generic_ref);
1406 if (ret == 0 && old_ref_mod < 0 && new_ref_mod >= 0)
1407 sub_pinned_bytes(fs_info, generic_ref);
1413 * __btrfs_inc_extent_ref - insert backreference for a given extent
1415 * The counterpart is in __btrfs_free_extent(), with examples and more details
1418 * @trans: Handle of transaction
1420 * @node: The delayed ref node used to get the bytenr/length for
1421 * extent whose references are incremented.
1423 * @parent: If this is a shared extent (BTRFS_SHARED_DATA_REF_KEY/
1424 * BTRFS_SHARED_BLOCK_REF_KEY) then it holds the logical
1425 * bytenr of the parent block. Since new extents are always
1426 * created with indirect references, this will only be the case
1427 * when relocating a shared extent. In that case, root_objectid
1428 * will be BTRFS_TREE_RELOC_OBJECTID. Otheriwse, parent must
1431 * @root_objectid: The id of the root where this modification has originated,
1432 * this can be either one of the well-known metadata trees or
1433 * the subvolume id which references this extent.
1435 * @owner: For data extents it is the inode number of the owning file.
1436 * For metadata extents this parameter holds the level in the
1437 * tree of the extent.
1439 * @offset: For metadata extents the offset is ignored and is currently
1440 * always passed as 0. For data extents it is the fileoffset
1441 * this extent belongs to.
1443 * @refs_to_add Number of references to add
1445 * @extent_op Pointer to a structure, holding information necessary when
1446 * updating a tree block's flags
1449 static int __btrfs_inc_extent_ref(struct btrfs_trans_handle *trans,
1450 struct btrfs_delayed_ref_node *node,
1451 u64 parent, u64 root_objectid,
1452 u64 owner, u64 offset, int refs_to_add,
1453 struct btrfs_delayed_extent_op *extent_op)
1455 struct btrfs_path *path;
1456 struct extent_buffer *leaf;
1457 struct btrfs_extent_item *item;
1458 struct btrfs_key key;
1459 u64 bytenr = node->bytenr;
1460 u64 num_bytes = node->num_bytes;
1464 path = btrfs_alloc_path();
1468 /* this will setup the path even if it fails to insert the back ref */
1469 ret = insert_inline_extent_backref(trans, path, bytenr, num_bytes,
1470 parent, root_objectid, owner,
1471 offset, refs_to_add, extent_op);
1472 if ((ret < 0 && ret != -EAGAIN) || !ret)
1476 * Ok we had -EAGAIN which means we didn't have space to insert and
1477 * inline extent ref, so just update the reference count and add a
1480 leaf = path->nodes[0];
1481 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
1482 item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
1483 refs = btrfs_extent_refs(leaf, item);
1484 btrfs_set_extent_refs(leaf, item, refs + refs_to_add);
1486 __run_delayed_extent_op(extent_op, leaf, item);
1488 btrfs_mark_buffer_dirty(leaf);
1489 btrfs_release_path(path);
1491 /* now insert the actual backref */
1492 if (owner < BTRFS_FIRST_FREE_OBJECTID) {
1493 BUG_ON(refs_to_add != 1);
1494 ret = insert_tree_block_ref(trans, path, bytenr, parent,
1497 ret = insert_extent_data_ref(trans, path, bytenr, parent,
1498 root_objectid, owner, offset,
1502 btrfs_abort_transaction(trans, ret);
1504 btrfs_free_path(path);
1508 static int run_delayed_data_ref(struct btrfs_trans_handle *trans,
1509 struct btrfs_delayed_ref_node *node,
1510 struct btrfs_delayed_extent_op *extent_op,
1511 int insert_reserved)
1514 struct btrfs_delayed_data_ref *ref;
1515 struct btrfs_key ins;
1520 ins.objectid = node->bytenr;
1521 ins.offset = node->num_bytes;
1522 ins.type = BTRFS_EXTENT_ITEM_KEY;
1524 ref = btrfs_delayed_node_to_data_ref(node);
1525 trace_run_delayed_data_ref(trans->fs_info, node, ref, node->action);
1527 if (node->type == BTRFS_SHARED_DATA_REF_KEY)
1528 parent = ref->parent;
1529 ref_root = ref->root;
1531 if (node->action == BTRFS_ADD_DELAYED_REF && insert_reserved) {
1533 flags |= extent_op->flags_to_set;
1534 ret = alloc_reserved_file_extent(trans, parent, ref_root,
1535 flags, ref->objectid,
1538 } else if (node->action == BTRFS_ADD_DELAYED_REF) {
1539 ret = __btrfs_inc_extent_ref(trans, node, parent, ref_root,
1540 ref->objectid, ref->offset,
1541 node->ref_mod, extent_op);
1542 } else if (node->action == BTRFS_DROP_DELAYED_REF) {
1543 ret = __btrfs_free_extent(trans, node, parent,
1544 ref_root, ref->objectid,
1545 ref->offset, node->ref_mod,
1553 static void __run_delayed_extent_op(struct btrfs_delayed_extent_op *extent_op,
1554 struct extent_buffer *leaf,
1555 struct btrfs_extent_item *ei)
1557 u64 flags = btrfs_extent_flags(leaf, ei);
1558 if (extent_op->update_flags) {
1559 flags |= extent_op->flags_to_set;
1560 btrfs_set_extent_flags(leaf, ei, flags);
1563 if (extent_op->update_key) {
1564 struct btrfs_tree_block_info *bi;
1565 BUG_ON(!(flags & BTRFS_EXTENT_FLAG_TREE_BLOCK));
1566 bi = (struct btrfs_tree_block_info *)(ei + 1);
1567 btrfs_set_tree_block_key(leaf, bi, &extent_op->key);
1571 static int run_delayed_extent_op(struct btrfs_trans_handle *trans,
1572 struct btrfs_delayed_ref_head *head,
1573 struct btrfs_delayed_extent_op *extent_op)
1575 struct btrfs_fs_info *fs_info = trans->fs_info;
1576 struct btrfs_key key;
1577 struct btrfs_path *path;
1578 struct btrfs_extent_item *ei;
1579 struct extent_buffer *leaf;
1583 int metadata = !extent_op->is_data;
1585 if (TRANS_ABORTED(trans))
1588 if (metadata && !btrfs_fs_incompat(fs_info, SKINNY_METADATA))
1591 path = btrfs_alloc_path();
1595 key.objectid = head->bytenr;
1598 key.type = BTRFS_METADATA_ITEM_KEY;
1599 key.offset = extent_op->level;
1601 key.type = BTRFS_EXTENT_ITEM_KEY;
1602 key.offset = head->num_bytes;
1606 ret = btrfs_search_slot(trans, fs_info->extent_root, &key, path, 0, 1);
1613 if (path->slots[0] > 0) {
1615 btrfs_item_key_to_cpu(path->nodes[0], &key,
1617 if (key.objectid == head->bytenr &&
1618 key.type == BTRFS_EXTENT_ITEM_KEY &&
1619 key.offset == head->num_bytes)
1623 btrfs_release_path(path);
1626 key.objectid = head->bytenr;
1627 key.offset = head->num_bytes;
1628 key.type = BTRFS_EXTENT_ITEM_KEY;
1637 leaf = path->nodes[0];
1638 item_size = btrfs_item_size_nr(leaf, path->slots[0]);
1640 if (unlikely(item_size < sizeof(*ei))) {
1642 btrfs_print_v0_err(fs_info);
1643 btrfs_abort_transaction(trans, err);
1647 ei = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
1648 __run_delayed_extent_op(extent_op, leaf, ei);
1650 btrfs_mark_buffer_dirty(leaf);
1652 btrfs_free_path(path);
1656 static int run_delayed_tree_ref(struct btrfs_trans_handle *trans,
1657 struct btrfs_delayed_ref_node *node,
1658 struct btrfs_delayed_extent_op *extent_op,
1659 int insert_reserved)
1662 struct btrfs_delayed_tree_ref *ref;
1666 ref = btrfs_delayed_node_to_tree_ref(node);
1667 trace_run_delayed_tree_ref(trans->fs_info, node, ref, node->action);
1669 if (node->type == BTRFS_SHARED_BLOCK_REF_KEY)
1670 parent = ref->parent;
1671 ref_root = ref->root;
1673 if (node->ref_mod != 1) {
1674 btrfs_err(trans->fs_info,
1675 "btree block(%llu) has %d references rather than 1: action %d ref_root %llu parent %llu",
1676 node->bytenr, node->ref_mod, node->action, ref_root,
1680 if (node->action == BTRFS_ADD_DELAYED_REF && insert_reserved) {
1681 BUG_ON(!extent_op || !extent_op->update_flags);
1682 ret = alloc_reserved_tree_block(trans, node, extent_op);
1683 } else if (node->action == BTRFS_ADD_DELAYED_REF) {
1684 ret = __btrfs_inc_extent_ref(trans, node, parent, ref_root,
1685 ref->level, 0, 1, extent_op);
1686 } else if (node->action == BTRFS_DROP_DELAYED_REF) {
1687 ret = __btrfs_free_extent(trans, node, parent, ref_root,
1688 ref->level, 0, 1, extent_op);
1695 /* helper function to actually process a single delayed ref entry */
1696 static int run_one_delayed_ref(struct btrfs_trans_handle *trans,
1697 struct btrfs_delayed_ref_node *node,
1698 struct btrfs_delayed_extent_op *extent_op,
1699 int insert_reserved)
1703 if (TRANS_ABORTED(trans)) {
1704 if (insert_reserved)
1705 btrfs_pin_extent(trans, node->bytenr, node->num_bytes, 1);
1709 if (node->type == BTRFS_TREE_BLOCK_REF_KEY ||
1710 node->type == BTRFS_SHARED_BLOCK_REF_KEY)
1711 ret = run_delayed_tree_ref(trans, node, extent_op,
1713 else if (node->type == BTRFS_EXTENT_DATA_REF_KEY ||
1714 node->type == BTRFS_SHARED_DATA_REF_KEY)
1715 ret = run_delayed_data_ref(trans, node, extent_op,
1719 if (ret && insert_reserved)
1720 btrfs_pin_extent(trans, node->bytenr, node->num_bytes, 1);
1724 static inline struct btrfs_delayed_ref_node *
1725 select_delayed_ref(struct btrfs_delayed_ref_head *head)
1727 struct btrfs_delayed_ref_node *ref;
1729 if (RB_EMPTY_ROOT(&head->ref_tree.rb_root))
1733 * Select a delayed ref of type BTRFS_ADD_DELAYED_REF first.
1734 * This is to prevent a ref count from going down to zero, which deletes
1735 * the extent item from the extent tree, when there still are references
1736 * to add, which would fail because they would not find the extent item.
1738 if (!list_empty(&head->ref_add_list))
1739 return list_first_entry(&head->ref_add_list,
1740 struct btrfs_delayed_ref_node, add_list);
1742 ref = rb_entry(rb_first_cached(&head->ref_tree),
1743 struct btrfs_delayed_ref_node, ref_node);
1744 ASSERT(list_empty(&ref->add_list));
1748 static void unselect_delayed_ref_head(struct btrfs_delayed_ref_root *delayed_refs,
1749 struct btrfs_delayed_ref_head *head)
1751 spin_lock(&delayed_refs->lock);
1752 head->processing = 0;
1753 delayed_refs->num_heads_ready++;
1754 spin_unlock(&delayed_refs->lock);
1755 btrfs_delayed_ref_unlock(head);
1758 static struct btrfs_delayed_extent_op *cleanup_extent_op(
1759 struct btrfs_delayed_ref_head *head)
1761 struct btrfs_delayed_extent_op *extent_op = head->extent_op;
1766 if (head->must_insert_reserved) {
1767 head->extent_op = NULL;
1768 btrfs_free_delayed_extent_op(extent_op);
1774 static int run_and_cleanup_extent_op(struct btrfs_trans_handle *trans,
1775 struct btrfs_delayed_ref_head *head)
1777 struct btrfs_delayed_extent_op *extent_op;
1780 extent_op = cleanup_extent_op(head);
1783 head->extent_op = NULL;
1784 spin_unlock(&head->lock);
1785 ret = run_delayed_extent_op(trans, head, extent_op);
1786 btrfs_free_delayed_extent_op(extent_op);
1787 return ret ? ret : 1;
1790 void btrfs_cleanup_ref_head_accounting(struct btrfs_fs_info *fs_info,
1791 struct btrfs_delayed_ref_root *delayed_refs,
1792 struct btrfs_delayed_ref_head *head)
1794 int nr_items = 1; /* Dropping this ref head update. */
1796 if (head->total_ref_mod < 0) {
1797 struct btrfs_space_info *space_info;
1801 flags = BTRFS_BLOCK_GROUP_DATA;
1802 else if (head->is_system)
1803 flags = BTRFS_BLOCK_GROUP_SYSTEM;
1805 flags = BTRFS_BLOCK_GROUP_METADATA;
1806 space_info = btrfs_find_space_info(fs_info, flags);
1808 percpu_counter_add_batch(&space_info->total_bytes_pinned,
1810 BTRFS_TOTAL_BYTES_PINNED_BATCH);
1813 * We had csum deletions accounted for in our delayed refs rsv,
1814 * we need to drop the csum leaves for this update from our
1817 if (head->is_data) {
1818 spin_lock(&delayed_refs->lock);
1819 delayed_refs->pending_csums -= head->num_bytes;
1820 spin_unlock(&delayed_refs->lock);
1821 nr_items += btrfs_csum_bytes_to_leaves(fs_info,
1826 btrfs_delayed_refs_rsv_release(fs_info, nr_items);
1829 static int cleanup_ref_head(struct btrfs_trans_handle *trans,
1830 struct btrfs_delayed_ref_head *head)
1833 struct btrfs_fs_info *fs_info = trans->fs_info;
1834 struct btrfs_delayed_ref_root *delayed_refs;
1837 delayed_refs = &trans->transaction->delayed_refs;
1839 ret = run_and_cleanup_extent_op(trans, head);
1841 unselect_delayed_ref_head(delayed_refs, head);
1842 btrfs_debug(fs_info, "run_delayed_extent_op returned %d", ret);
1849 * Need to drop our head ref lock and re-acquire the delayed ref lock
1850 * and then re-check to make sure nobody got added.
1852 spin_unlock(&head->lock);
1853 spin_lock(&delayed_refs->lock);
1854 spin_lock(&head->lock);
1855 if (!RB_EMPTY_ROOT(&head->ref_tree.rb_root) || head->extent_op) {
1856 spin_unlock(&head->lock);
1857 spin_unlock(&delayed_refs->lock);
1860 btrfs_delete_ref_head(delayed_refs, head);
1861 spin_unlock(&head->lock);
1862 spin_unlock(&delayed_refs->lock);
1864 if (head->must_insert_reserved) {
1865 btrfs_pin_extent(trans, head->bytenr, head->num_bytes, 1);
1866 if (head->is_data) {
1867 ret = btrfs_del_csums(trans, fs_info->csum_root,
1868 head->bytenr, head->num_bytes);
1872 btrfs_cleanup_ref_head_accounting(fs_info, delayed_refs, head);
1874 trace_run_delayed_ref_head(fs_info, head, 0);
1875 btrfs_delayed_ref_unlock(head);
1876 btrfs_put_delayed_ref_head(head);
1880 static struct btrfs_delayed_ref_head *btrfs_obtain_ref_head(
1881 struct btrfs_trans_handle *trans)
1883 struct btrfs_delayed_ref_root *delayed_refs =
1884 &trans->transaction->delayed_refs;
1885 struct btrfs_delayed_ref_head *head = NULL;
1888 spin_lock(&delayed_refs->lock);
1889 head = btrfs_select_ref_head(delayed_refs);
1891 spin_unlock(&delayed_refs->lock);
1896 * Grab the lock that says we are going to process all the refs for
1899 ret = btrfs_delayed_ref_lock(delayed_refs, head);
1900 spin_unlock(&delayed_refs->lock);
1903 * We may have dropped the spin lock to get the head mutex lock, and
1904 * that might have given someone else time to free the head. If that's
1905 * true, it has been removed from our list and we can move on.
1908 head = ERR_PTR(-EAGAIN);
1913 static int btrfs_run_delayed_refs_for_head(struct btrfs_trans_handle *trans,
1914 struct btrfs_delayed_ref_head *locked_ref,
1915 unsigned long *run_refs)
1917 struct btrfs_fs_info *fs_info = trans->fs_info;
1918 struct btrfs_delayed_ref_root *delayed_refs;
1919 struct btrfs_delayed_extent_op *extent_op;
1920 struct btrfs_delayed_ref_node *ref;
1921 int must_insert_reserved = 0;
1924 delayed_refs = &trans->transaction->delayed_refs;
1926 lockdep_assert_held(&locked_ref->mutex);
1927 lockdep_assert_held(&locked_ref->lock);
1929 while ((ref = select_delayed_ref(locked_ref))) {
1931 btrfs_check_delayed_seq(fs_info, ref->seq)) {
1932 spin_unlock(&locked_ref->lock);
1933 unselect_delayed_ref_head(delayed_refs, locked_ref);
1939 rb_erase_cached(&ref->ref_node, &locked_ref->ref_tree);
1940 RB_CLEAR_NODE(&ref->ref_node);
1941 if (!list_empty(&ref->add_list))
1942 list_del(&ref->add_list);
1944 * When we play the delayed ref, also correct the ref_mod on
1947 switch (ref->action) {
1948 case BTRFS_ADD_DELAYED_REF:
1949 case BTRFS_ADD_DELAYED_EXTENT:
1950 locked_ref->ref_mod -= ref->ref_mod;
1952 case BTRFS_DROP_DELAYED_REF:
1953 locked_ref->ref_mod += ref->ref_mod;
1958 atomic_dec(&delayed_refs->num_entries);
1961 * Record the must_insert_reserved flag before we drop the
1964 must_insert_reserved = locked_ref->must_insert_reserved;
1965 locked_ref->must_insert_reserved = 0;
1967 extent_op = locked_ref->extent_op;
1968 locked_ref->extent_op = NULL;
1969 spin_unlock(&locked_ref->lock);
1971 ret = run_one_delayed_ref(trans, ref, extent_op,
1972 must_insert_reserved);
1974 btrfs_free_delayed_extent_op(extent_op);
1976 unselect_delayed_ref_head(delayed_refs, locked_ref);
1977 btrfs_put_delayed_ref(ref);
1978 btrfs_debug(fs_info, "run_one_delayed_ref returned %d",
1983 btrfs_put_delayed_ref(ref);
1986 spin_lock(&locked_ref->lock);
1987 btrfs_merge_delayed_refs(trans, delayed_refs, locked_ref);
1994 * Returns 0 on success or if called with an already aborted transaction.
1995 * Returns -ENOMEM or -EIO on failure and will abort the transaction.
1997 static noinline int __btrfs_run_delayed_refs(struct btrfs_trans_handle *trans,
2000 struct btrfs_fs_info *fs_info = trans->fs_info;
2001 struct btrfs_delayed_ref_root *delayed_refs;
2002 struct btrfs_delayed_ref_head *locked_ref = NULL;
2003 ktime_t start = ktime_get();
2005 unsigned long count = 0;
2006 unsigned long actual_count = 0;
2008 delayed_refs = &trans->transaction->delayed_refs;
2011 locked_ref = btrfs_obtain_ref_head(trans);
2012 if (IS_ERR_OR_NULL(locked_ref)) {
2013 if (PTR_ERR(locked_ref) == -EAGAIN) {
2022 * We need to try and merge add/drops of the same ref since we
2023 * can run into issues with relocate dropping the implicit ref
2024 * and then it being added back again before the drop can
2025 * finish. If we merged anything we need to re-loop so we can
2027 * Or we can get node references of the same type that weren't
2028 * merged when created due to bumps in the tree mod seq, and
2029 * we need to merge them to prevent adding an inline extent
2030 * backref before dropping it (triggering a BUG_ON at
2031 * insert_inline_extent_backref()).
2033 spin_lock(&locked_ref->lock);
2034 btrfs_merge_delayed_refs(trans, delayed_refs, locked_ref);
2036 ret = btrfs_run_delayed_refs_for_head(trans, locked_ref,
2038 if (ret < 0 && ret != -EAGAIN) {
2040 * Error, btrfs_run_delayed_refs_for_head already
2041 * unlocked everything so just bail out
2046 * Success, perform the usual cleanup of a processed
2049 ret = cleanup_ref_head(trans, locked_ref);
2051 /* We dropped our lock, we need to loop. */
2060 * Either success case or btrfs_run_delayed_refs_for_head
2061 * returned -EAGAIN, meaning we need to select another head
2066 } while ((nr != -1 && count < nr) || locked_ref);
2069 * We don't want to include ref heads since we can have empty ref heads
2070 * and those will drastically skew our runtime down since we just do
2071 * accounting, no actual extent tree updates.
2073 if (actual_count > 0) {
2074 u64 runtime = ktime_to_ns(ktime_sub(ktime_get(), start));
2078 * We weigh the current average higher than our current runtime
2079 * to avoid large swings in the average.
2081 spin_lock(&delayed_refs->lock);
2082 avg = fs_info->avg_delayed_ref_runtime * 3 + runtime;
2083 fs_info->avg_delayed_ref_runtime = avg >> 2; /* div by 4 */
2084 spin_unlock(&delayed_refs->lock);
2089 #ifdef SCRAMBLE_DELAYED_REFS
2091 * Normally delayed refs get processed in ascending bytenr order. This
2092 * correlates in most cases to the order added. To expose dependencies on this
2093 * order, we start to process the tree in the middle instead of the beginning
2095 static u64 find_middle(struct rb_root *root)
2097 struct rb_node *n = root->rb_node;
2098 struct btrfs_delayed_ref_node *entry;
2101 u64 first = 0, last = 0;
2105 entry = rb_entry(n, struct btrfs_delayed_ref_node, rb_node);
2106 first = entry->bytenr;
2110 entry = rb_entry(n, struct btrfs_delayed_ref_node, rb_node);
2111 last = entry->bytenr;
2116 entry = rb_entry(n, struct btrfs_delayed_ref_node, rb_node);
2117 WARN_ON(!entry->in_tree);
2119 middle = entry->bytenr;
2133 * this starts processing the delayed reference count updates and
2134 * extent insertions we have queued up so far. count can be
2135 * 0, which means to process everything in the tree at the start
2136 * of the run (but not newly added entries), or it can be some target
2137 * number you'd like to process.
2139 * Returns 0 on success or if called with an aborted transaction
2140 * Returns <0 on error and aborts the transaction
2142 int btrfs_run_delayed_refs(struct btrfs_trans_handle *trans,
2143 unsigned long count)
2145 struct btrfs_fs_info *fs_info = trans->fs_info;
2146 struct rb_node *node;
2147 struct btrfs_delayed_ref_root *delayed_refs;
2148 struct btrfs_delayed_ref_head *head;
2150 int run_all = count == (unsigned long)-1;
2152 /* We'll clean this up in btrfs_cleanup_transaction */
2153 if (TRANS_ABORTED(trans))
2156 if (test_bit(BTRFS_FS_CREATING_FREE_SPACE_TREE, &fs_info->flags))
2159 delayed_refs = &trans->transaction->delayed_refs;
2161 count = atomic_read(&delayed_refs->num_entries) * 2;
2164 #ifdef SCRAMBLE_DELAYED_REFS
2165 delayed_refs->run_delayed_start = find_middle(&delayed_refs->root);
2167 ret = __btrfs_run_delayed_refs(trans, count);
2169 btrfs_abort_transaction(trans, ret);
2174 btrfs_create_pending_block_groups(trans);
2176 spin_lock(&delayed_refs->lock);
2177 node = rb_first_cached(&delayed_refs->href_root);
2179 spin_unlock(&delayed_refs->lock);
2182 head = rb_entry(node, struct btrfs_delayed_ref_head,
2184 refcount_inc(&head->refs);
2185 spin_unlock(&delayed_refs->lock);
2187 /* Mutex was contended, block until it's released and retry. */
2188 mutex_lock(&head->mutex);
2189 mutex_unlock(&head->mutex);
2191 btrfs_put_delayed_ref_head(head);
2199 int btrfs_set_disk_extent_flags(struct btrfs_trans_handle *trans,
2200 struct extent_buffer *eb, u64 flags,
2201 int level, int is_data)
2203 struct btrfs_delayed_extent_op *extent_op;
2206 extent_op = btrfs_alloc_delayed_extent_op();
2210 extent_op->flags_to_set = flags;
2211 extent_op->update_flags = true;
2212 extent_op->update_key = false;
2213 extent_op->is_data = is_data ? true : false;
2214 extent_op->level = level;
2216 ret = btrfs_add_delayed_extent_op(trans, eb->start, eb->len, extent_op);
2218 btrfs_free_delayed_extent_op(extent_op);
2222 static noinline int check_delayed_ref(struct btrfs_root *root,
2223 struct btrfs_path *path,
2224 u64 objectid, u64 offset, u64 bytenr)
2226 struct btrfs_delayed_ref_head *head;
2227 struct btrfs_delayed_ref_node *ref;
2228 struct btrfs_delayed_data_ref *data_ref;
2229 struct btrfs_delayed_ref_root *delayed_refs;
2230 struct btrfs_transaction *cur_trans;
2231 struct rb_node *node;
2234 spin_lock(&root->fs_info->trans_lock);
2235 cur_trans = root->fs_info->running_transaction;
2237 refcount_inc(&cur_trans->use_count);
2238 spin_unlock(&root->fs_info->trans_lock);
2242 delayed_refs = &cur_trans->delayed_refs;
2243 spin_lock(&delayed_refs->lock);
2244 head = btrfs_find_delayed_ref_head(delayed_refs, bytenr);
2246 spin_unlock(&delayed_refs->lock);
2247 btrfs_put_transaction(cur_trans);
2251 if (!mutex_trylock(&head->mutex)) {
2252 refcount_inc(&head->refs);
2253 spin_unlock(&delayed_refs->lock);
2255 btrfs_release_path(path);
2258 * Mutex was contended, block until it's released and let
2261 mutex_lock(&head->mutex);
2262 mutex_unlock(&head->mutex);
2263 btrfs_put_delayed_ref_head(head);
2264 btrfs_put_transaction(cur_trans);
2267 spin_unlock(&delayed_refs->lock);
2269 spin_lock(&head->lock);
2271 * XXX: We should replace this with a proper search function in the
2274 for (node = rb_first_cached(&head->ref_tree); node;
2275 node = rb_next(node)) {
2276 ref = rb_entry(node, struct btrfs_delayed_ref_node, ref_node);
2277 /* If it's a shared ref we know a cross reference exists */
2278 if (ref->type != BTRFS_EXTENT_DATA_REF_KEY) {
2283 data_ref = btrfs_delayed_node_to_data_ref(ref);
2286 * If our ref doesn't match the one we're currently looking at
2287 * then we have a cross reference.
2289 if (data_ref->root != root->root_key.objectid ||
2290 data_ref->objectid != objectid ||
2291 data_ref->offset != offset) {
2296 spin_unlock(&head->lock);
2297 mutex_unlock(&head->mutex);
2298 btrfs_put_transaction(cur_trans);
2302 static noinline int check_committed_ref(struct btrfs_root *root,
2303 struct btrfs_path *path,
2304 u64 objectid, u64 offset, u64 bytenr,
2307 struct btrfs_fs_info *fs_info = root->fs_info;
2308 struct btrfs_root *extent_root = fs_info->extent_root;
2309 struct extent_buffer *leaf;
2310 struct btrfs_extent_data_ref *ref;
2311 struct btrfs_extent_inline_ref *iref;
2312 struct btrfs_extent_item *ei;
2313 struct btrfs_key key;
2318 key.objectid = bytenr;
2319 key.offset = (u64)-1;
2320 key.type = BTRFS_EXTENT_ITEM_KEY;
2322 ret = btrfs_search_slot(NULL, extent_root, &key, path, 0, 0);
2325 BUG_ON(ret == 0); /* Corruption */
2328 if (path->slots[0] == 0)
2332 leaf = path->nodes[0];
2333 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
2335 if (key.objectid != bytenr || key.type != BTRFS_EXTENT_ITEM_KEY)
2339 item_size = btrfs_item_size_nr(leaf, path->slots[0]);
2340 ei = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
2342 /* If extent item has more than 1 inline ref then it's shared */
2343 if (item_size != sizeof(*ei) +
2344 btrfs_extent_inline_ref_size(BTRFS_EXTENT_DATA_REF_KEY))
2348 * If extent created before last snapshot => it's shared unless the
2349 * snapshot has been deleted. Use the heuristic if strict is false.
2352 (btrfs_extent_generation(leaf, ei) <=
2353 btrfs_root_last_snapshot(&root->root_item)))
2356 iref = (struct btrfs_extent_inline_ref *)(ei + 1);
2358 /* If this extent has SHARED_DATA_REF then it's shared */
2359 type = btrfs_get_extent_inline_ref_type(leaf, iref, BTRFS_REF_TYPE_DATA);
2360 if (type != BTRFS_EXTENT_DATA_REF_KEY)
2363 ref = (struct btrfs_extent_data_ref *)(&iref->offset);
2364 if (btrfs_extent_refs(leaf, ei) !=
2365 btrfs_extent_data_ref_count(leaf, ref) ||
2366 btrfs_extent_data_ref_root(leaf, ref) !=
2367 root->root_key.objectid ||
2368 btrfs_extent_data_ref_objectid(leaf, ref) != objectid ||
2369 btrfs_extent_data_ref_offset(leaf, ref) != offset)
2377 int btrfs_cross_ref_exist(struct btrfs_root *root, u64 objectid, u64 offset,
2378 u64 bytenr, bool strict)
2380 struct btrfs_path *path;
2383 path = btrfs_alloc_path();
2388 ret = check_committed_ref(root, path, objectid,
2389 offset, bytenr, strict);
2390 if (ret && ret != -ENOENT)
2393 ret = check_delayed_ref(root, path, objectid, offset, bytenr);
2394 } while (ret == -EAGAIN);
2397 btrfs_free_path(path);
2398 if (root->root_key.objectid == BTRFS_DATA_RELOC_TREE_OBJECTID)
2403 static int __btrfs_mod_ref(struct btrfs_trans_handle *trans,
2404 struct btrfs_root *root,
2405 struct extent_buffer *buf,
2406 int full_backref, int inc)
2408 struct btrfs_fs_info *fs_info = root->fs_info;
2414 struct btrfs_key key;
2415 struct btrfs_file_extent_item *fi;
2416 struct btrfs_ref generic_ref = { 0 };
2417 bool for_reloc = btrfs_header_flag(buf, BTRFS_HEADER_FLAG_RELOC);
2423 if (btrfs_is_testing(fs_info))
2426 ref_root = btrfs_header_owner(buf);
2427 nritems = btrfs_header_nritems(buf);
2428 level = btrfs_header_level(buf);
2430 if (!test_bit(BTRFS_ROOT_SHAREABLE, &root->state) && level == 0)
2434 parent = buf->start;
2438 action = BTRFS_ADD_DELAYED_REF;
2440 action = BTRFS_DROP_DELAYED_REF;
2442 for (i = 0; i < nritems; i++) {
2444 btrfs_item_key_to_cpu(buf, &key, i);
2445 if (key.type != BTRFS_EXTENT_DATA_KEY)
2447 fi = btrfs_item_ptr(buf, i,
2448 struct btrfs_file_extent_item);
2449 if (btrfs_file_extent_type(buf, fi) ==
2450 BTRFS_FILE_EXTENT_INLINE)
2452 bytenr = btrfs_file_extent_disk_bytenr(buf, fi);
2456 num_bytes = btrfs_file_extent_disk_num_bytes(buf, fi);
2457 key.offset -= btrfs_file_extent_offset(buf, fi);
2458 btrfs_init_generic_ref(&generic_ref, action, bytenr,
2460 generic_ref.real_root = root->root_key.objectid;
2461 btrfs_init_data_ref(&generic_ref, ref_root, key.objectid,
2463 generic_ref.skip_qgroup = for_reloc;
2465 ret = btrfs_inc_extent_ref(trans, &generic_ref);
2467 ret = btrfs_free_extent(trans, &generic_ref);
2471 bytenr = btrfs_node_blockptr(buf, i);
2472 num_bytes = fs_info->nodesize;
2473 btrfs_init_generic_ref(&generic_ref, action, bytenr,
2475 generic_ref.real_root = root->root_key.objectid;
2476 btrfs_init_tree_ref(&generic_ref, level - 1, ref_root);
2477 generic_ref.skip_qgroup = for_reloc;
2479 ret = btrfs_inc_extent_ref(trans, &generic_ref);
2481 ret = btrfs_free_extent(trans, &generic_ref);
2491 int btrfs_inc_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root,
2492 struct extent_buffer *buf, int full_backref)
2494 return __btrfs_mod_ref(trans, root, buf, full_backref, 1);
2497 int btrfs_dec_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root,
2498 struct extent_buffer *buf, int full_backref)
2500 return __btrfs_mod_ref(trans, root, buf, full_backref, 0);
2503 int btrfs_extent_readonly(struct btrfs_fs_info *fs_info, u64 bytenr)
2505 struct btrfs_block_group *block_group;
2508 block_group = btrfs_lookup_block_group(fs_info, bytenr);
2509 if (!block_group || block_group->ro)
2512 btrfs_put_block_group(block_group);
2516 static u64 get_alloc_profile_by_root(struct btrfs_root *root, int data)
2518 struct btrfs_fs_info *fs_info = root->fs_info;
2523 flags = BTRFS_BLOCK_GROUP_DATA;
2524 else if (root == fs_info->chunk_root)
2525 flags = BTRFS_BLOCK_GROUP_SYSTEM;
2527 flags = BTRFS_BLOCK_GROUP_METADATA;
2529 ret = btrfs_get_alloc_profile(fs_info, flags);
2533 static u64 first_logical_byte(struct btrfs_fs_info *fs_info, u64 search_start)
2535 struct btrfs_block_group *cache;
2538 spin_lock(&fs_info->block_group_cache_lock);
2539 bytenr = fs_info->first_logical_byte;
2540 spin_unlock(&fs_info->block_group_cache_lock);
2542 if (bytenr < (u64)-1)
2545 cache = btrfs_lookup_first_block_group(fs_info, search_start);
2549 bytenr = cache->start;
2550 btrfs_put_block_group(cache);
2555 static int pin_down_extent(struct btrfs_trans_handle *trans,
2556 struct btrfs_block_group *cache,
2557 u64 bytenr, u64 num_bytes, int reserved)
2559 struct btrfs_fs_info *fs_info = cache->fs_info;
2561 spin_lock(&cache->space_info->lock);
2562 spin_lock(&cache->lock);
2563 cache->pinned += num_bytes;
2564 btrfs_space_info_update_bytes_pinned(fs_info, cache->space_info,
2567 cache->reserved -= num_bytes;
2568 cache->space_info->bytes_reserved -= num_bytes;
2570 spin_unlock(&cache->lock);
2571 spin_unlock(&cache->space_info->lock);
2573 percpu_counter_add_batch(&cache->space_info->total_bytes_pinned,
2574 num_bytes, BTRFS_TOTAL_BYTES_PINNED_BATCH);
2575 set_extent_dirty(&trans->transaction->pinned_extents, bytenr,
2576 bytenr + num_bytes - 1, GFP_NOFS | __GFP_NOFAIL);
2580 int btrfs_pin_extent(struct btrfs_trans_handle *trans,
2581 u64 bytenr, u64 num_bytes, int reserved)
2583 struct btrfs_block_group *cache;
2585 cache = btrfs_lookup_block_group(trans->fs_info, bytenr);
2586 BUG_ON(!cache); /* Logic error */
2588 pin_down_extent(trans, cache, bytenr, num_bytes, reserved);
2590 btrfs_put_block_group(cache);
2595 * this function must be called within transaction
2597 int btrfs_pin_extent_for_log_replay(struct btrfs_trans_handle *trans,
2598 u64 bytenr, u64 num_bytes)
2600 struct btrfs_block_group *cache;
2603 btrfs_add_excluded_extent(trans->fs_info, bytenr, num_bytes);
2605 cache = btrfs_lookup_block_group(trans->fs_info, bytenr);
2610 * pull in the free space cache (if any) so that our pin
2611 * removes the free space from the cache. We have load_only set
2612 * to one because the slow code to read in the free extents does check
2613 * the pinned extents.
2615 btrfs_cache_block_group(cache, 1);
2617 pin_down_extent(trans, cache, bytenr, num_bytes, 0);
2619 /* remove us from the free space cache (if we're there at all) */
2620 ret = btrfs_remove_free_space(cache, bytenr, num_bytes);
2621 btrfs_put_block_group(cache);
2625 static int __exclude_logged_extent(struct btrfs_fs_info *fs_info,
2626 u64 start, u64 num_bytes)
2629 struct btrfs_block_group *block_group;
2630 struct btrfs_caching_control *caching_ctl;
2632 block_group = btrfs_lookup_block_group(fs_info, start);
2636 btrfs_cache_block_group(block_group, 0);
2637 caching_ctl = btrfs_get_caching_control(block_group);
2641 BUG_ON(!btrfs_block_group_done(block_group));
2642 ret = btrfs_remove_free_space(block_group, start, num_bytes);
2645 * We must wait for v1 caching to finish, otherwise we may not
2648 btrfs_wait_space_cache_v1_finished(block_group, caching_ctl);
2649 mutex_lock(&caching_ctl->mutex);
2651 if (start >= caching_ctl->progress) {
2652 ret = btrfs_add_excluded_extent(fs_info, start,
2654 } else if (start + num_bytes <= caching_ctl->progress) {
2655 ret = btrfs_remove_free_space(block_group,
2658 num_bytes = caching_ctl->progress - start;
2659 ret = btrfs_remove_free_space(block_group,
2664 num_bytes = (start + num_bytes) -
2665 caching_ctl->progress;
2666 start = caching_ctl->progress;
2667 ret = btrfs_add_excluded_extent(fs_info, start,
2671 mutex_unlock(&caching_ctl->mutex);
2672 btrfs_put_caching_control(caching_ctl);
2674 btrfs_put_block_group(block_group);
2678 int btrfs_exclude_logged_extents(struct extent_buffer *eb)
2680 struct btrfs_fs_info *fs_info = eb->fs_info;
2681 struct btrfs_file_extent_item *item;
2682 struct btrfs_key key;
2687 if (!btrfs_fs_incompat(fs_info, MIXED_GROUPS))
2690 for (i = 0; i < btrfs_header_nritems(eb); i++) {
2691 btrfs_item_key_to_cpu(eb, &key, i);
2692 if (key.type != BTRFS_EXTENT_DATA_KEY)
2694 item = btrfs_item_ptr(eb, i, struct btrfs_file_extent_item);
2695 found_type = btrfs_file_extent_type(eb, item);
2696 if (found_type == BTRFS_FILE_EXTENT_INLINE)
2698 if (btrfs_file_extent_disk_bytenr(eb, item) == 0)
2700 key.objectid = btrfs_file_extent_disk_bytenr(eb, item);
2701 key.offset = btrfs_file_extent_disk_num_bytes(eb, item);
2702 ret = __exclude_logged_extent(fs_info, key.objectid, key.offset);
2711 btrfs_inc_block_group_reservations(struct btrfs_block_group *bg)
2713 atomic_inc(&bg->reservations);
2717 * Returns the free cluster for the given space info and sets empty_cluster to
2718 * what it should be based on the mount options.
2720 static struct btrfs_free_cluster *
2721 fetch_cluster_info(struct btrfs_fs_info *fs_info,
2722 struct btrfs_space_info *space_info, u64 *empty_cluster)
2724 struct btrfs_free_cluster *ret = NULL;
2727 if (btrfs_mixed_space_info(space_info))
2730 if (space_info->flags & BTRFS_BLOCK_GROUP_METADATA) {
2731 ret = &fs_info->meta_alloc_cluster;
2732 if (btrfs_test_opt(fs_info, SSD))
2733 *empty_cluster = SZ_2M;
2735 *empty_cluster = SZ_64K;
2736 } else if ((space_info->flags & BTRFS_BLOCK_GROUP_DATA) &&
2737 btrfs_test_opt(fs_info, SSD_SPREAD)) {
2738 *empty_cluster = SZ_2M;
2739 ret = &fs_info->data_alloc_cluster;
2745 static int unpin_extent_range(struct btrfs_fs_info *fs_info,
2747 const bool return_free_space)
2749 struct btrfs_block_group *cache = NULL;
2750 struct btrfs_space_info *space_info;
2751 struct btrfs_block_rsv *global_rsv = &fs_info->global_block_rsv;
2752 struct btrfs_free_cluster *cluster = NULL;
2754 u64 total_unpinned = 0;
2755 u64 empty_cluster = 0;
2758 while (start <= end) {
2761 start >= cache->start + cache->length) {
2763 btrfs_put_block_group(cache);
2765 cache = btrfs_lookup_block_group(fs_info, start);
2766 BUG_ON(!cache); /* Logic error */
2768 cluster = fetch_cluster_info(fs_info,
2771 empty_cluster <<= 1;
2774 len = cache->start + cache->length - start;
2775 len = min(len, end + 1 - start);
2777 down_read(&fs_info->commit_root_sem);
2778 if (start < cache->last_byte_to_unpin && return_free_space) {
2779 u64 add_len = min(len, cache->last_byte_to_unpin - start);
2781 btrfs_add_free_space(cache, start, add_len);
2783 up_read(&fs_info->commit_root_sem);
2786 total_unpinned += len;
2787 space_info = cache->space_info;
2790 * If this space cluster has been marked as fragmented and we've
2791 * unpinned enough in this block group to potentially allow a
2792 * cluster to be created inside of it go ahead and clear the
2795 if (cluster && cluster->fragmented &&
2796 total_unpinned > empty_cluster) {
2797 spin_lock(&cluster->lock);
2798 cluster->fragmented = 0;
2799 spin_unlock(&cluster->lock);
2802 spin_lock(&space_info->lock);
2803 spin_lock(&cache->lock);
2804 cache->pinned -= len;
2805 btrfs_space_info_update_bytes_pinned(fs_info, space_info, -len);
2806 space_info->max_extent_size = 0;
2807 percpu_counter_add_batch(&space_info->total_bytes_pinned,
2808 -len, BTRFS_TOTAL_BYTES_PINNED_BATCH);
2810 space_info->bytes_readonly += len;
2813 spin_unlock(&cache->lock);
2814 if (!readonly && return_free_space &&
2815 global_rsv->space_info == space_info) {
2818 spin_lock(&global_rsv->lock);
2819 if (!global_rsv->full) {
2820 to_add = min(len, global_rsv->size -
2821 global_rsv->reserved);
2822 global_rsv->reserved += to_add;
2823 btrfs_space_info_update_bytes_may_use(fs_info,
2824 space_info, to_add);
2825 if (global_rsv->reserved >= global_rsv->size)
2826 global_rsv->full = 1;
2829 spin_unlock(&global_rsv->lock);
2831 /* Add to any tickets we may have */
2832 if (!readonly && return_free_space && len)
2833 btrfs_try_granting_tickets(fs_info, space_info);
2834 spin_unlock(&space_info->lock);
2838 btrfs_put_block_group(cache);
2842 int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans)
2844 struct btrfs_fs_info *fs_info = trans->fs_info;
2845 struct btrfs_block_group *block_group, *tmp;
2846 struct list_head *deleted_bgs;
2847 struct extent_io_tree *unpin;
2852 unpin = &trans->transaction->pinned_extents;
2854 while (!TRANS_ABORTED(trans)) {
2855 struct extent_state *cached_state = NULL;
2857 mutex_lock(&fs_info->unused_bg_unpin_mutex);
2858 ret = find_first_extent_bit(unpin, 0, &start, &end,
2859 EXTENT_DIRTY, &cached_state);
2861 mutex_unlock(&fs_info->unused_bg_unpin_mutex);
2864 if (test_bit(BTRFS_FS_LOG_RECOVERING, &fs_info->flags))
2865 clear_extent_bits(&fs_info->excluded_extents, start,
2866 end, EXTENT_UPTODATE);
2868 if (btrfs_test_opt(fs_info, DISCARD_SYNC))
2869 ret = btrfs_discard_extent(fs_info, start,
2870 end + 1 - start, NULL);
2872 clear_extent_dirty(unpin, start, end, &cached_state);
2873 unpin_extent_range(fs_info, start, end, true);
2874 mutex_unlock(&fs_info->unused_bg_unpin_mutex);
2875 free_extent_state(cached_state);
2879 if (btrfs_test_opt(fs_info, DISCARD_ASYNC)) {
2880 btrfs_discard_calc_delay(&fs_info->discard_ctl);
2881 btrfs_discard_schedule_work(&fs_info->discard_ctl, true);
2885 * Transaction is finished. We don't need the lock anymore. We
2886 * do need to clean up the block groups in case of a transaction
2889 deleted_bgs = &trans->transaction->deleted_bgs;
2890 list_for_each_entry_safe(block_group, tmp, deleted_bgs, bg_list) {
2894 if (!TRANS_ABORTED(trans))
2895 ret = btrfs_discard_extent(fs_info,
2897 block_group->length,
2900 list_del_init(&block_group->bg_list);
2901 btrfs_unfreeze_block_group(block_group);
2902 btrfs_put_block_group(block_group);
2905 const char *errstr = btrfs_decode_error(ret);
2907 "discard failed while removing blockgroup: errno=%d %s",
2916 * Drop one or more refs of @node.
2918 * 1. Locate the extent refs.
2919 * It's either inline in EXTENT/METADATA_ITEM or in keyed SHARED_* item.
2920 * Locate it, then reduce the refs number or remove the ref line completely.
2922 * 2. Update the refs count in EXTENT/METADATA_ITEM
2924 * Inline backref case:
2926 * in extent tree we have:
2928 * item 0 key (13631488 EXTENT_ITEM 1048576) itemoff 16201 itemsize 82
2929 * refs 2 gen 6 flags DATA
2930 * extent data backref root FS_TREE objectid 258 offset 0 count 1
2931 * extent data backref root FS_TREE objectid 257 offset 0 count 1
2933 * This function gets called with:
2935 * node->bytenr = 13631488
2936 * node->num_bytes = 1048576
2937 * root_objectid = FS_TREE
2938 * owner_objectid = 257
2942 * Then we should get some like:
2944 * item 0 key (13631488 EXTENT_ITEM 1048576) itemoff 16201 itemsize 82
2945 * refs 1 gen 6 flags DATA
2946 * extent data backref root FS_TREE objectid 258 offset 0 count 1
2948 * Keyed backref case:
2950 * in extent tree we have:
2952 * item 0 key (13631488 EXTENT_ITEM 1048576) itemoff 3971 itemsize 24
2953 * refs 754 gen 6 flags DATA
2955 * item 2 key (13631488 EXTENT_DATA_REF <HASH>) itemoff 3915 itemsize 28
2956 * extent data backref root FS_TREE objectid 866 offset 0 count 1
2958 * This function get called with:
2960 * node->bytenr = 13631488
2961 * node->num_bytes = 1048576
2962 * root_objectid = FS_TREE
2963 * owner_objectid = 866
2967 * Then we should get some like:
2969 * item 0 key (13631488 EXTENT_ITEM 1048576) itemoff 3971 itemsize 24
2970 * refs 753 gen 6 flags DATA
2972 * And that (13631488 EXTENT_DATA_REF <HASH>) gets removed.
2974 static int __btrfs_free_extent(struct btrfs_trans_handle *trans,
2975 struct btrfs_delayed_ref_node *node, u64 parent,
2976 u64 root_objectid, u64 owner_objectid,
2977 u64 owner_offset, int refs_to_drop,
2978 struct btrfs_delayed_extent_op *extent_op)
2980 struct btrfs_fs_info *info = trans->fs_info;
2981 struct btrfs_key key;
2982 struct btrfs_path *path;
2983 struct btrfs_root *extent_root = info->extent_root;
2984 struct extent_buffer *leaf;
2985 struct btrfs_extent_item *ei;
2986 struct btrfs_extent_inline_ref *iref;
2989 int extent_slot = 0;
2990 int found_extent = 0;
2994 u64 bytenr = node->bytenr;
2995 u64 num_bytes = node->num_bytes;
2997 bool skinny_metadata = btrfs_fs_incompat(info, SKINNY_METADATA);
2999 path = btrfs_alloc_path();
3003 is_data = owner_objectid >= BTRFS_FIRST_FREE_OBJECTID;
3005 if (!is_data && refs_to_drop != 1) {
3007 "invalid refs_to_drop, dropping more than 1 refs for tree block %llu refs_to_drop %u",
3008 node->bytenr, refs_to_drop);
3010 btrfs_abort_transaction(trans, ret);
3015 skinny_metadata = false;
3017 ret = lookup_extent_backref(trans, path, &iref, bytenr, num_bytes,
3018 parent, root_objectid, owner_objectid,
3022 * Either the inline backref or the SHARED_DATA_REF/
3023 * SHARED_BLOCK_REF is found
3025 * Here is a quick path to locate EXTENT/METADATA_ITEM.
3026 * It's possible the EXTENT/METADATA_ITEM is near current slot.
3028 extent_slot = path->slots[0];
3029 while (extent_slot >= 0) {
3030 btrfs_item_key_to_cpu(path->nodes[0], &key,
3032 if (key.objectid != bytenr)
3034 if (key.type == BTRFS_EXTENT_ITEM_KEY &&
3035 key.offset == num_bytes) {
3039 if (key.type == BTRFS_METADATA_ITEM_KEY &&
3040 key.offset == owner_objectid) {
3045 /* Quick path didn't find the EXTEMT/METADATA_ITEM */
3046 if (path->slots[0] - extent_slot > 5)
3051 if (!found_extent) {
3054 "invalid iref, no EXTENT/METADATA_ITEM found but has inline extent ref");
3055 btrfs_abort_transaction(trans, -EUCLEAN);
3058 /* Must be SHARED_* item, remove the backref first */
3059 ret = remove_extent_backref(trans, path, NULL,
3061 is_data, &last_ref);
3063 btrfs_abort_transaction(trans, ret);
3066 btrfs_release_path(path);
3068 /* Slow path to locate EXTENT/METADATA_ITEM */
3069 key.objectid = bytenr;
3070 key.type = BTRFS_EXTENT_ITEM_KEY;
3071 key.offset = num_bytes;
3073 if (!is_data && skinny_metadata) {
3074 key.type = BTRFS_METADATA_ITEM_KEY;
3075 key.offset = owner_objectid;
3078 ret = btrfs_search_slot(trans, extent_root,
3080 if (ret > 0 && skinny_metadata && path->slots[0]) {
3082 * Couldn't find our skinny metadata item,
3083 * see if we have ye olde extent item.
3086 btrfs_item_key_to_cpu(path->nodes[0], &key,
3088 if (key.objectid == bytenr &&
3089 key.type == BTRFS_EXTENT_ITEM_KEY &&
3090 key.offset == num_bytes)
3094 if (ret > 0 && skinny_metadata) {
3095 skinny_metadata = false;
3096 key.objectid = bytenr;
3097 key.type = BTRFS_EXTENT_ITEM_KEY;
3098 key.offset = num_bytes;
3099 btrfs_release_path(path);
3100 ret = btrfs_search_slot(trans, extent_root,
3106 "umm, got %d back from search, was looking for %llu",
3109 btrfs_print_leaf(path->nodes[0]);
3112 btrfs_abort_transaction(trans, ret);
3115 extent_slot = path->slots[0];
3117 } else if (WARN_ON(ret == -ENOENT)) {
3118 btrfs_print_leaf(path->nodes[0]);
3120 "unable to find ref byte nr %llu parent %llu root %llu owner %llu offset %llu",
3121 bytenr, parent, root_objectid, owner_objectid,
3123 btrfs_abort_transaction(trans, ret);
3126 btrfs_abort_transaction(trans, ret);
3130 leaf = path->nodes[0];
3131 item_size = btrfs_item_size_nr(leaf, extent_slot);
3132 if (unlikely(item_size < sizeof(*ei))) {
3134 btrfs_print_v0_err(info);
3135 btrfs_abort_transaction(trans, ret);
3138 ei = btrfs_item_ptr(leaf, extent_slot,
3139 struct btrfs_extent_item);
3140 if (owner_objectid < BTRFS_FIRST_FREE_OBJECTID &&
3141 key.type == BTRFS_EXTENT_ITEM_KEY) {
3142 struct btrfs_tree_block_info *bi;
3143 if (item_size < sizeof(*ei) + sizeof(*bi)) {
3145 "invalid extent item size for key (%llu, %u, %llu) owner %llu, has %u expect >= %zu",
3146 key.objectid, key.type, key.offset,
3147 owner_objectid, item_size,
3148 sizeof(*ei) + sizeof(*bi));
3149 btrfs_abort_transaction(trans, -EUCLEAN);
3152 bi = (struct btrfs_tree_block_info *)(ei + 1);
3153 WARN_ON(owner_objectid != btrfs_tree_block_level(leaf, bi));
3156 refs = btrfs_extent_refs(leaf, ei);
3157 if (refs < refs_to_drop) {
3159 "trying to drop %d refs but we only have %llu for bytenr %llu",
3160 refs_to_drop, refs, bytenr);
3161 btrfs_abort_transaction(trans, -EUCLEAN);
3164 refs -= refs_to_drop;
3168 __run_delayed_extent_op(extent_op, leaf, ei);
3170 * In the case of inline back ref, reference count will
3171 * be updated by remove_extent_backref
3174 if (!found_extent) {
3176 "invalid iref, got inlined extent ref but no EXTENT/METADATA_ITEM found");
3177 btrfs_abort_transaction(trans, -EUCLEAN);
3181 btrfs_set_extent_refs(leaf, ei, refs);
3182 btrfs_mark_buffer_dirty(leaf);
3185 ret = remove_extent_backref(trans, path, iref,
3186 refs_to_drop, is_data,
3189 btrfs_abort_transaction(trans, ret);
3194 /* In this branch refs == 1 */
3196 if (is_data && refs_to_drop !=
3197 extent_data_ref_count(path, iref)) {
3199 "invalid refs_to_drop, current refs %u refs_to_drop %u",
3200 extent_data_ref_count(path, iref),
3202 btrfs_abort_transaction(trans, -EUCLEAN);
3206 if (path->slots[0] != extent_slot) {
3208 "invalid iref, extent item key (%llu %u %llu) doesn't have wanted iref",
3209 key.objectid, key.type,
3211 btrfs_abort_transaction(trans, -EUCLEAN);
3216 * No inline ref, we must be at SHARED_* item,
3217 * And it's single ref, it must be:
3218 * | extent_slot ||extent_slot + 1|
3219 * [ EXTENT/METADATA_ITEM ][ SHARED_* ITEM ]
3221 if (path->slots[0] != extent_slot + 1) {
3223 "invalid SHARED_* item, previous item is not EXTENT/METADATA_ITEM");
3224 btrfs_abort_transaction(trans, -EUCLEAN);
3227 path->slots[0] = extent_slot;
3233 ret = btrfs_del_items(trans, extent_root, path, path->slots[0],
3236 btrfs_abort_transaction(trans, ret);
3239 btrfs_release_path(path);
3242 ret = btrfs_del_csums(trans, info->csum_root, bytenr,
3245 btrfs_abort_transaction(trans, ret);
3250 ret = add_to_free_space_tree(trans, bytenr, num_bytes);
3252 btrfs_abort_transaction(trans, ret);
3256 ret = btrfs_update_block_group(trans, bytenr, num_bytes, 0);
3258 btrfs_abort_transaction(trans, ret);
3262 btrfs_release_path(path);
3265 btrfs_free_path(path);
3269 * Leaf dump can take up a lot of log buffer, so we only do full leaf
3270 * dump for debug build.
3272 if (IS_ENABLED(CONFIG_BTRFS_DEBUG)) {
3273 btrfs_crit(info, "path->slots[0]=%d extent_slot=%d",
3274 path->slots[0], extent_slot);
3275 btrfs_print_leaf(path->nodes[0]);
3278 btrfs_free_path(path);
3283 * when we free an block, it is possible (and likely) that we free the last
3284 * delayed ref for that extent as well. This searches the delayed ref tree for
3285 * a given extent, and if there are no other delayed refs to be processed, it
3286 * removes it from the tree.
3288 static noinline int check_ref_cleanup(struct btrfs_trans_handle *trans,
3291 struct btrfs_delayed_ref_head *head;
3292 struct btrfs_delayed_ref_root *delayed_refs;
3295 delayed_refs = &trans->transaction->delayed_refs;
3296 spin_lock(&delayed_refs->lock);
3297 head = btrfs_find_delayed_ref_head(delayed_refs, bytenr);
3299 goto out_delayed_unlock;
3301 spin_lock(&head->lock);
3302 if (!RB_EMPTY_ROOT(&head->ref_tree.rb_root))
3305 if (cleanup_extent_op(head) != NULL)
3309 * waiting for the lock here would deadlock. If someone else has it
3310 * locked they are already in the process of dropping it anyway
3312 if (!mutex_trylock(&head->mutex))
3315 btrfs_delete_ref_head(delayed_refs, head);
3316 head->processing = 0;
3318 spin_unlock(&head->lock);
3319 spin_unlock(&delayed_refs->lock);
3321 BUG_ON(head->extent_op);
3322 if (head->must_insert_reserved)
3325 btrfs_cleanup_ref_head_accounting(trans->fs_info, delayed_refs, head);
3326 mutex_unlock(&head->mutex);
3327 btrfs_put_delayed_ref_head(head);
3330 spin_unlock(&head->lock);
3333 spin_unlock(&delayed_refs->lock);
3337 void btrfs_free_tree_block(struct btrfs_trans_handle *trans,
3338 struct btrfs_root *root,
3339 struct extent_buffer *buf,
3340 u64 parent, int last_ref)
3342 struct btrfs_fs_info *fs_info = root->fs_info;
3343 struct btrfs_ref generic_ref = { 0 };
3347 btrfs_init_generic_ref(&generic_ref, BTRFS_DROP_DELAYED_REF,
3348 buf->start, buf->len, parent);
3349 btrfs_init_tree_ref(&generic_ref, btrfs_header_level(buf),
3350 root->root_key.objectid);
3352 if (root->root_key.objectid != BTRFS_TREE_LOG_OBJECTID) {
3353 int old_ref_mod, new_ref_mod;
3355 btrfs_ref_tree_mod(fs_info, &generic_ref);
3356 ret = btrfs_add_delayed_tree_ref(trans, &generic_ref, NULL,
3357 &old_ref_mod, &new_ref_mod);
3358 BUG_ON(ret); /* -ENOMEM */
3359 pin = old_ref_mod >= 0 && new_ref_mod < 0;
3362 if (last_ref && btrfs_header_generation(buf) == trans->transid) {
3363 struct btrfs_block_group *cache;
3365 if (root->root_key.objectid != BTRFS_TREE_LOG_OBJECTID) {
3366 ret = check_ref_cleanup(trans, buf->start);
3372 cache = btrfs_lookup_block_group(fs_info, buf->start);
3374 if (btrfs_header_flag(buf, BTRFS_HEADER_FLAG_WRITTEN)) {
3375 pin_down_extent(trans, cache, buf->start, buf->len, 1);
3376 btrfs_put_block_group(cache);
3380 WARN_ON(test_bit(EXTENT_BUFFER_DIRTY, &buf->bflags));
3382 btrfs_add_free_space(cache, buf->start, buf->len);
3383 btrfs_free_reserved_bytes(cache, buf->len, 0);
3384 btrfs_put_block_group(cache);
3385 trace_btrfs_reserved_extent_free(fs_info, buf->start, buf->len);
3389 add_pinned_bytes(fs_info, &generic_ref);
3393 * Deleting the buffer, clear the corrupt flag since it doesn't
3396 clear_bit(EXTENT_BUFFER_CORRUPT, &buf->bflags);
3400 /* Can return -ENOMEM */
3401 int btrfs_free_extent(struct btrfs_trans_handle *trans, struct btrfs_ref *ref)
3403 struct btrfs_fs_info *fs_info = trans->fs_info;
3404 int old_ref_mod, new_ref_mod;
3407 if (btrfs_is_testing(fs_info))
3411 * tree log blocks never actually go into the extent allocation
3412 * tree, just update pinning info and exit early.
3414 if ((ref->type == BTRFS_REF_METADATA &&
3415 ref->tree_ref.root == BTRFS_TREE_LOG_OBJECTID) ||
3416 (ref->type == BTRFS_REF_DATA &&
3417 ref->data_ref.ref_root == BTRFS_TREE_LOG_OBJECTID)) {
3418 /* unlocks the pinned mutex */
3419 btrfs_pin_extent(trans, ref->bytenr, ref->len, 1);
3420 old_ref_mod = new_ref_mod = 0;
3422 } else if (ref->type == BTRFS_REF_METADATA) {
3423 ret = btrfs_add_delayed_tree_ref(trans, ref, NULL,
3424 &old_ref_mod, &new_ref_mod);
3426 ret = btrfs_add_delayed_data_ref(trans, ref, 0,
3427 &old_ref_mod, &new_ref_mod);
3430 if (!((ref->type == BTRFS_REF_METADATA &&
3431 ref->tree_ref.root == BTRFS_TREE_LOG_OBJECTID) ||
3432 (ref->type == BTRFS_REF_DATA &&
3433 ref->data_ref.ref_root == BTRFS_TREE_LOG_OBJECTID)))
3434 btrfs_ref_tree_mod(fs_info, ref);
3436 if (ret == 0 && old_ref_mod >= 0 && new_ref_mod < 0)
3437 add_pinned_bytes(fs_info, ref);
3442 enum btrfs_loop_type {
3443 LOOP_CACHING_NOWAIT,
3450 btrfs_lock_block_group(struct btrfs_block_group *cache,
3454 down_read(&cache->data_rwsem);
3457 static inline void btrfs_grab_block_group(struct btrfs_block_group *cache,
3460 btrfs_get_block_group(cache);
3462 down_read(&cache->data_rwsem);
3465 static struct btrfs_block_group *btrfs_lock_cluster(
3466 struct btrfs_block_group *block_group,
3467 struct btrfs_free_cluster *cluster,
3469 __acquires(&cluster->refill_lock)
3471 struct btrfs_block_group *used_bg = NULL;
3473 spin_lock(&cluster->refill_lock);
3475 used_bg = cluster->block_group;
3479 if (used_bg == block_group)
3482 btrfs_get_block_group(used_bg);
3487 if (down_read_trylock(&used_bg->data_rwsem))
3490 spin_unlock(&cluster->refill_lock);
3492 /* We should only have one-level nested. */
3493 down_read_nested(&used_bg->data_rwsem, SINGLE_DEPTH_NESTING);
3495 spin_lock(&cluster->refill_lock);
3496 if (used_bg == cluster->block_group)
3499 up_read(&used_bg->data_rwsem);
3500 btrfs_put_block_group(used_bg);
3505 btrfs_release_block_group(struct btrfs_block_group *cache,
3509 up_read(&cache->data_rwsem);
3510 btrfs_put_block_group(cache);
3513 enum btrfs_extent_allocation_policy {
3514 BTRFS_EXTENT_ALLOC_CLUSTERED,
3518 * Structure used internally for find_free_extent() function. Wraps needed
3521 struct find_free_extent_ctl {
3522 /* Basic allocation info */
3528 /* Where to start the search inside the bg */
3531 /* For clustered allocation */
3533 struct btrfs_free_cluster *last_ptr;
3536 bool have_caching_bg;
3537 bool orig_have_caching_bg;
3539 /* RAID index, converted from flags */
3543 * Current loop number, check find_free_extent_update_loop() for details
3548 * Whether we're refilling a cluster, if true we need to re-search
3549 * current block group but don't try to refill the cluster again.
3551 bool retry_clustered;
3554 * Whether we're updating free space cache, if true we need to re-search
3555 * current block group but don't try updating free space cache again.
3557 bool retry_unclustered;
3559 /* If current block group is cached */
3562 /* Max contiguous hole found */
3563 u64 max_extent_size;
3565 /* Total free space from free space cache, not always contiguous */
3566 u64 total_free_space;
3571 /* Hint where to start looking for an empty space */
3574 /* Allocation policy */
3575 enum btrfs_extent_allocation_policy policy;
3580 * Helper function for find_free_extent().
3582 * Return -ENOENT to inform caller that we need fallback to unclustered mode.
3583 * Return -EAGAIN to inform caller that we need to re-search this block group
3584 * Return >0 to inform caller that we find nothing
3585 * Return 0 means we have found a location and set ffe_ctl->found_offset.
3587 static int find_free_extent_clustered(struct btrfs_block_group *bg,
3588 struct find_free_extent_ctl *ffe_ctl,
3589 struct btrfs_block_group **cluster_bg_ret)
3591 struct btrfs_block_group *cluster_bg;
3592 struct btrfs_free_cluster *last_ptr = ffe_ctl->last_ptr;
3593 u64 aligned_cluster;
3597 cluster_bg = btrfs_lock_cluster(bg, last_ptr, ffe_ctl->delalloc);
3599 goto refill_cluster;
3600 if (cluster_bg != bg && (cluster_bg->ro ||
3601 !block_group_bits(cluster_bg, ffe_ctl->flags)))
3602 goto release_cluster;
3604 offset = btrfs_alloc_from_cluster(cluster_bg, last_ptr,
3605 ffe_ctl->num_bytes, cluster_bg->start,
3606 &ffe_ctl->max_extent_size);
3608 /* We have a block, we're done */
3609 spin_unlock(&last_ptr->refill_lock);
3610 trace_btrfs_reserve_extent_cluster(cluster_bg,
3611 ffe_ctl->search_start, ffe_ctl->num_bytes);
3612 *cluster_bg_ret = cluster_bg;
3613 ffe_ctl->found_offset = offset;
3616 WARN_ON(last_ptr->block_group != cluster_bg);
3620 * If we are on LOOP_NO_EMPTY_SIZE, we can't set up a new clusters, so
3621 * lets just skip it and let the allocator find whatever block it can
3622 * find. If we reach this point, we will have tried the cluster
3623 * allocator plenty of times and not have found anything, so we are
3624 * likely way too fragmented for the clustering stuff to find anything.
3626 * However, if the cluster is taken from the current block group,
3627 * release the cluster first, so that we stand a better chance of
3628 * succeeding in the unclustered allocation.
3630 if (ffe_ctl->loop >= LOOP_NO_EMPTY_SIZE && cluster_bg != bg) {
3631 spin_unlock(&last_ptr->refill_lock);
3632 btrfs_release_block_group(cluster_bg, ffe_ctl->delalloc);
3636 /* This cluster didn't work out, free it and start over */
3637 btrfs_return_cluster_to_free_space(NULL, last_ptr);
3639 if (cluster_bg != bg)
3640 btrfs_release_block_group(cluster_bg, ffe_ctl->delalloc);
3643 if (ffe_ctl->loop >= LOOP_NO_EMPTY_SIZE) {
3644 spin_unlock(&last_ptr->refill_lock);
3648 aligned_cluster = max_t(u64,
3649 ffe_ctl->empty_cluster + ffe_ctl->empty_size,
3650 bg->full_stripe_len);
3651 ret = btrfs_find_space_cluster(bg, last_ptr, ffe_ctl->search_start,
3652 ffe_ctl->num_bytes, aligned_cluster);
3654 /* Now pull our allocation out of this cluster */
3655 offset = btrfs_alloc_from_cluster(bg, last_ptr,
3656 ffe_ctl->num_bytes, ffe_ctl->search_start,
3657 &ffe_ctl->max_extent_size);
3659 /* We found one, proceed */
3660 spin_unlock(&last_ptr->refill_lock);
3661 trace_btrfs_reserve_extent_cluster(bg,
3662 ffe_ctl->search_start,
3663 ffe_ctl->num_bytes);
3664 ffe_ctl->found_offset = offset;
3667 } else if (!ffe_ctl->cached && ffe_ctl->loop > LOOP_CACHING_NOWAIT &&
3668 !ffe_ctl->retry_clustered) {
3669 spin_unlock(&last_ptr->refill_lock);
3671 ffe_ctl->retry_clustered = true;
3672 btrfs_wait_block_group_cache_progress(bg, ffe_ctl->num_bytes +
3673 ffe_ctl->empty_cluster + ffe_ctl->empty_size);
3677 * At this point we either didn't find a cluster or we weren't able to
3678 * allocate a block from our cluster. Free the cluster we've been
3679 * trying to use, and go to the next block group.
3681 btrfs_return_cluster_to_free_space(NULL, last_ptr);
3682 spin_unlock(&last_ptr->refill_lock);
3687 * Return >0 to inform caller that we find nothing
3688 * Return 0 when we found an free extent and set ffe_ctrl->found_offset
3689 * Return -EAGAIN to inform caller that we need to re-search this block group
3691 static int find_free_extent_unclustered(struct btrfs_block_group *bg,
3692 struct find_free_extent_ctl *ffe_ctl)
3694 struct btrfs_free_cluster *last_ptr = ffe_ctl->last_ptr;
3698 * We are doing an unclustered allocation, set the fragmented flag so
3699 * we don't bother trying to setup a cluster again until we get more
3702 if (unlikely(last_ptr)) {
3703 spin_lock(&last_ptr->lock);
3704 last_ptr->fragmented = 1;
3705 spin_unlock(&last_ptr->lock);
3707 if (ffe_ctl->cached) {
3708 struct btrfs_free_space_ctl *free_space_ctl;
3710 free_space_ctl = bg->free_space_ctl;
3711 spin_lock(&free_space_ctl->tree_lock);
3712 if (free_space_ctl->free_space <
3713 ffe_ctl->num_bytes + ffe_ctl->empty_cluster +
3714 ffe_ctl->empty_size) {
3715 ffe_ctl->total_free_space = max_t(u64,
3716 ffe_ctl->total_free_space,
3717 free_space_ctl->free_space);
3718 spin_unlock(&free_space_ctl->tree_lock);
3721 spin_unlock(&free_space_ctl->tree_lock);
3724 offset = btrfs_find_space_for_alloc(bg, ffe_ctl->search_start,
3725 ffe_ctl->num_bytes, ffe_ctl->empty_size,
3726 &ffe_ctl->max_extent_size);
3729 * If we didn't find a chunk, and we haven't failed on this block group
3730 * before, and this block group is in the middle of caching and we are
3731 * ok with waiting, then go ahead and wait for progress to be made, and
3732 * set @retry_unclustered to true.
3734 * If @retry_unclustered is true then we've already waited on this
3735 * block group once and should move on to the next block group.
3737 if (!offset && !ffe_ctl->retry_unclustered && !ffe_ctl->cached &&
3738 ffe_ctl->loop > LOOP_CACHING_NOWAIT) {
3739 btrfs_wait_block_group_cache_progress(bg, ffe_ctl->num_bytes +
3740 ffe_ctl->empty_size);
3741 ffe_ctl->retry_unclustered = true;
3743 } else if (!offset) {
3746 ffe_ctl->found_offset = offset;
3750 static int do_allocation_clustered(struct btrfs_block_group *block_group,
3751 struct find_free_extent_ctl *ffe_ctl,
3752 struct btrfs_block_group **bg_ret)
3756 /* We want to try and use the cluster allocator, so lets look there */
3757 if (ffe_ctl->last_ptr && ffe_ctl->use_cluster) {
3758 ret = find_free_extent_clustered(block_group, ffe_ctl, bg_ret);
3759 if (ret >= 0 || ret == -EAGAIN)
3761 /* ret == -ENOENT case falls through */
3764 return find_free_extent_unclustered(block_group, ffe_ctl);
3767 static int do_allocation(struct btrfs_block_group *block_group,
3768 struct find_free_extent_ctl *ffe_ctl,
3769 struct btrfs_block_group **bg_ret)
3771 switch (ffe_ctl->policy) {
3772 case BTRFS_EXTENT_ALLOC_CLUSTERED:
3773 return do_allocation_clustered(block_group, ffe_ctl, bg_ret);
3779 static void release_block_group(struct btrfs_block_group *block_group,
3780 struct find_free_extent_ctl *ffe_ctl,
3783 switch (ffe_ctl->policy) {
3784 case BTRFS_EXTENT_ALLOC_CLUSTERED:
3785 ffe_ctl->retry_clustered = false;
3786 ffe_ctl->retry_unclustered = false;
3792 BUG_ON(btrfs_bg_flags_to_raid_index(block_group->flags) !=
3794 btrfs_release_block_group(block_group, delalloc);
3797 static void found_extent_clustered(struct find_free_extent_ctl *ffe_ctl,
3798 struct btrfs_key *ins)
3800 struct btrfs_free_cluster *last_ptr = ffe_ctl->last_ptr;
3802 if (!ffe_ctl->use_cluster && last_ptr) {
3803 spin_lock(&last_ptr->lock);
3804 last_ptr->window_start = ins->objectid;
3805 spin_unlock(&last_ptr->lock);
3809 static void found_extent(struct find_free_extent_ctl *ffe_ctl,
3810 struct btrfs_key *ins)
3812 switch (ffe_ctl->policy) {
3813 case BTRFS_EXTENT_ALLOC_CLUSTERED:
3814 found_extent_clustered(ffe_ctl, ins);
3821 static int chunk_allocation_failed(struct find_free_extent_ctl *ffe_ctl)
3823 switch (ffe_ctl->policy) {
3824 case BTRFS_EXTENT_ALLOC_CLUSTERED:
3826 * If we can't allocate a new chunk we've already looped through
3827 * at least once, move on to the NO_EMPTY_SIZE case.
3829 ffe_ctl->loop = LOOP_NO_EMPTY_SIZE;
3837 * Return >0 means caller needs to re-search for free extent
3838 * Return 0 means we have the needed free extent.
3839 * Return <0 means we failed to locate any free extent.
3841 static int find_free_extent_update_loop(struct btrfs_fs_info *fs_info,
3842 struct btrfs_key *ins,
3843 struct find_free_extent_ctl *ffe_ctl,
3846 struct btrfs_root *root = fs_info->extent_root;
3849 if ((ffe_ctl->loop == LOOP_CACHING_NOWAIT) &&
3850 ffe_ctl->have_caching_bg && !ffe_ctl->orig_have_caching_bg)
3851 ffe_ctl->orig_have_caching_bg = true;
3853 if (!ins->objectid && ffe_ctl->loop >= LOOP_CACHING_WAIT &&
3854 ffe_ctl->have_caching_bg)
3857 if (!ins->objectid && ++(ffe_ctl->index) < BTRFS_NR_RAID_TYPES)
3860 if (ins->objectid) {
3861 found_extent(ffe_ctl, ins);
3866 * LOOP_CACHING_NOWAIT, search partially cached block groups, kicking
3867 * caching kthreads as we move along
3868 * LOOP_CACHING_WAIT, search everything, and wait if our bg is caching
3869 * LOOP_ALLOC_CHUNK, force a chunk allocation and try again
3870 * LOOP_NO_EMPTY_SIZE, set empty_size and empty_cluster to 0 and try
3873 if (ffe_ctl->loop < LOOP_NO_EMPTY_SIZE) {
3875 if (ffe_ctl->loop == LOOP_CACHING_NOWAIT) {
3877 * We want to skip the LOOP_CACHING_WAIT step if we
3878 * don't have any uncached bgs and we've already done a
3879 * full search through.
3881 if (ffe_ctl->orig_have_caching_bg || !full_search)
3882 ffe_ctl->loop = LOOP_CACHING_WAIT;
3884 ffe_ctl->loop = LOOP_ALLOC_CHUNK;
3889 if (ffe_ctl->loop == LOOP_ALLOC_CHUNK) {
3890 struct btrfs_trans_handle *trans;
3893 trans = current->journal_info;
3897 trans = btrfs_join_transaction(root);
3899 if (IS_ERR(trans)) {
3900 ret = PTR_ERR(trans);
3904 ret = btrfs_chunk_alloc(trans, ffe_ctl->flags,
3907 /* Do not bail out on ENOSPC since we can do more. */
3909 ret = chunk_allocation_failed(ffe_ctl);
3911 btrfs_abort_transaction(trans, ret);
3915 btrfs_end_transaction(trans);
3920 if (ffe_ctl->loop == LOOP_NO_EMPTY_SIZE) {
3921 if (ffe_ctl->policy != BTRFS_EXTENT_ALLOC_CLUSTERED)
3925 * Don't loop again if we already have no empty_size and
3928 if (ffe_ctl->empty_size == 0 &&
3929 ffe_ctl->empty_cluster == 0)
3931 ffe_ctl->empty_size = 0;
3932 ffe_ctl->empty_cluster = 0;
3939 static int prepare_allocation_clustered(struct btrfs_fs_info *fs_info,
3940 struct find_free_extent_ctl *ffe_ctl,
3941 struct btrfs_space_info *space_info,
3942 struct btrfs_key *ins)
3945 * If our free space is heavily fragmented we may not be able to make
3946 * big contiguous allocations, so instead of doing the expensive search
3947 * for free space, simply return ENOSPC with our max_extent_size so we
3948 * can go ahead and search for a more manageable chunk.
3950 * If our max_extent_size is large enough for our allocation simply
3951 * disable clustering since we will likely not be able to find enough
3952 * space to create a cluster and induce latency trying.
3954 if (space_info->max_extent_size) {
3955 spin_lock(&space_info->lock);
3956 if (space_info->max_extent_size &&
3957 ffe_ctl->num_bytes > space_info->max_extent_size) {
3958 ins->offset = space_info->max_extent_size;
3959 spin_unlock(&space_info->lock);
3961 } else if (space_info->max_extent_size) {
3962 ffe_ctl->use_cluster = false;
3964 spin_unlock(&space_info->lock);
3967 ffe_ctl->last_ptr = fetch_cluster_info(fs_info, space_info,
3968 &ffe_ctl->empty_cluster);
3969 if (ffe_ctl->last_ptr) {
3970 struct btrfs_free_cluster *last_ptr = ffe_ctl->last_ptr;
3972 spin_lock(&last_ptr->lock);
3973 if (last_ptr->block_group)
3974 ffe_ctl->hint_byte = last_ptr->window_start;
3975 if (last_ptr->fragmented) {
3977 * We still set window_start so we can keep track of the
3978 * last place we found an allocation to try and save
3981 ffe_ctl->hint_byte = last_ptr->window_start;
3982 ffe_ctl->use_cluster = false;
3984 spin_unlock(&last_ptr->lock);
3990 static int prepare_allocation(struct btrfs_fs_info *fs_info,
3991 struct find_free_extent_ctl *ffe_ctl,
3992 struct btrfs_space_info *space_info,
3993 struct btrfs_key *ins)
3995 switch (ffe_ctl->policy) {
3996 case BTRFS_EXTENT_ALLOC_CLUSTERED:
3997 return prepare_allocation_clustered(fs_info, ffe_ctl,
4005 * walks the btree of allocated extents and find a hole of a given size.
4006 * The key ins is changed to record the hole:
4007 * ins->objectid == start position
4008 * ins->flags = BTRFS_EXTENT_ITEM_KEY
4009 * ins->offset == the size of the hole.
4010 * Any available blocks before search_start are skipped.
4012 * If there is no suitable free space, we will record the max size of
4013 * the free space extent currently.
4015 * The overall logic and call chain:
4017 * find_free_extent()
4018 * |- Iterate through all block groups
4019 * | |- Get a valid block group
4020 * | |- Try to do clustered allocation in that block group
4021 * | |- Try to do unclustered allocation in that block group
4022 * | |- Check if the result is valid
4023 * | | |- If valid, then exit
4024 * | |- Jump to next block group
4026 * |- Push harder to find free extents
4027 * |- If not found, re-iterate all block groups
4029 static noinline int find_free_extent(struct btrfs_root *root,
4030 u64 ram_bytes, u64 num_bytes, u64 empty_size,
4031 u64 hint_byte_orig, struct btrfs_key *ins,
4032 u64 flags, int delalloc)
4034 struct btrfs_fs_info *fs_info = root->fs_info;
4036 int cache_block_group_error = 0;
4037 struct btrfs_block_group *block_group = NULL;
4038 struct find_free_extent_ctl ffe_ctl = {0};
4039 struct btrfs_space_info *space_info;
4040 bool full_search = false;
4042 WARN_ON(num_bytes < fs_info->sectorsize);
4044 ffe_ctl.num_bytes = num_bytes;
4045 ffe_ctl.empty_size = empty_size;
4046 ffe_ctl.flags = flags;
4047 ffe_ctl.search_start = 0;
4048 ffe_ctl.delalloc = delalloc;
4049 ffe_ctl.index = btrfs_bg_flags_to_raid_index(flags);
4050 ffe_ctl.have_caching_bg = false;
4051 ffe_ctl.orig_have_caching_bg = false;
4052 ffe_ctl.found_offset = 0;
4053 ffe_ctl.hint_byte = hint_byte_orig;
4054 ffe_ctl.policy = BTRFS_EXTENT_ALLOC_CLUSTERED;
4056 /* For clustered allocation */
4057 ffe_ctl.retry_clustered = false;
4058 ffe_ctl.retry_unclustered = false;
4059 ffe_ctl.last_ptr = NULL;
4060 ffe_ctl.use_cluster = true;
4062 ins->type = BTRFS_EXTENT_ITEM_KEY;
4066 trace_find_free_extent(root, num_bytes, empty_size, flags);
4068 space_info = btrfs_find_space_info(fs_info, flags);
4070 btrfs_err(fs_info, "No space info for %llu", flags);
4074 ret = prepare_allocation(fs_info, &ffe_ctl, space_info, ins);
4078 ffe_ctl.search_start = max(ffe_ctl.search_start,
4079 first_logical_byte(fs_info, 0));
4080 ffe_ctl.search_start = max(ffe_ctl.search_start, ffe_ctl.hint_byte);
4081 if (ffe_ctl.search_start == ffe_ctl.hint_byte) {
4082 block_group = btrfs_lookup_block_group(fs_info,
4083 ffe_ctl.search_start);
4085 * we don't want to use the block group if it doesn't match our
4086 * allocation bits, or if its not cached.
4088 * However if we are re-searching with an ideal block group
4089 * picked out then we don't care that the block group is cached.
4091 if (block_group && block_group_bits(block_group, flags) &&
4092 block_group->cached != BTRFS_CACHE_NO) {
4093 down_read(&space_info->groups_sem);
4094 if (list_empty(&block_group->list) ||
4097 * someone is removing this block group,
4098 * we can't jump into the have_block_group
4099 * target because our list pointers are not
4102 btrfs_put_block_group(block_group);
4103 up_read(&space_info->groups_sem);
4105 ffe_ctl.index = btrfs_bg_flags_to_raid_index(
4106 block_group->flags);
4107 btrfs_lock_block_group(block_group, delalloc);
4108 goto have_block_group;
4110 } else if (block_group) {
4111 btrfs_put_block_group(block_group);
4115 ffe_ctl.have_caching_bg = false;
4116 if (ffe_ctl.index == btrfs_bg_flags_to_raid_index(flags) ||
4119 down_read(&space_info->groups_sem);
4120 list_for_each_entry(block_group,
4121 &space_info->block_groups[ffe_ctl.index], list) {
4122 struct btrfs_block_group *bg_ret;
4124 /* If the block group is read-only, we can skip it entirely. */
4125 if (unlikely(block_group->ro))
4128 btrfs_grab_block_group(block_group, delalloc);
4129 ffe_ctl.search_start = block_group->start;
4132 * this can happen if we end up cycling through all the
4133 * raid types, but we want to make sure we only allocate
4134 * for the proper type.
4136 if (!block_group_bits(block_group, flags)) {
4137 u64 extra = BTRFS_BLOCK_GROUP_DUP |
4138 BTRFS_BLOCK_GROUP_RAID1_MASK |
4139 BTRFS_BLOCK_GROUP_RAID56_MASK |
4140 BTRFS_BLOCK_GROUP_RAID10;
4143 * if they asked for extra copies and this block group
4144 * doesn't provide them, bail. This does allow us to
4145 * fill raid0 from raid1.
4147 if ((flags & extra) && !(block_group->flags & extra))
4151 * This block group has different flags than we want.
4152 * It's possible that we have MIXED_GROUP flag but no
4153 * block group is mixed. Just skip such block group.
4155 btrfs_release_block_group(block_group, delalloc);
4160 ffe_ctl.cached = btrfs_block_group_done(block_group);
4161 if (unlikely(!ffe_ctl.cached)) {
4162 ffe_ctl.have_caching_bg = true;
4163 ret = btrfs_cache_block_group(block_group, 0);
4166 * If we get ENOMEM here or something else we want to
4167 * try other block groups, because it may not be fatal.
4168 * However if we can't find anything else we need to
4169 * save our return here so that we return the actual
4170 * error that caused problems, not ENOSPC.
4173 if (!cache_block_group_error)
4174 cache_block_group_error = ret;
4181 if (unlikely(block_group->cached == BTRFS_CACHE_ERROR))
4185 ret = do_allocation(block_group, &ffe_ctl, &bg_ret);
4187 if (bg_ret && bg_ret != block_group) {
4188 btrfs_release_block_group(block_group, delalloc);
4189 block_group = bg_ret;
4191 } else if (ret == -EAGAIN) {
4192 goto have_block_group;
4193 } else if (ret > 0) {
4198 ffe_ctl.search_start = round_up(ffe_ctl.found_offset,
4199 fs_info->stripesize);
4201 /* move on to the next group */
4202 if (ffe_ctl.search_start + num_bytes >
4203 block_group->start + block_group->length) {
4204 btrfs_add_free_space(block_group, ffe_ctl.found_offset,
4209 if (ffe_ctl.found_offset < ffe_ctl.search_start)
4210 btrfs_add_free_space(block_group, ffe_ctl.found_offset,
4211 ffe_ctl.search_start - ffe_ctl.found_offset);
4213 ret = btrfs_add_reserved_bytes(block_group, ram_bytes,
4214 num_bytes, delalloc);
4215 if (ret == -EAGAIN) {
4216 btrfs_add_free_space(block_group, ffe_ctl.found_offset,
4220 btrfs_inc_block_group_reservations(block_group);
4222 /* we are all good, lets return */
4223 ins->objectid = ffe_ctl.search_start;
4224 ins->offset = num_bytes;
4226 trace_btrfs_reserve_extent(block_group, ffe_ctl.search_start,
4228 btrfs_release_block_group(block_group, delalloc);
4231 release_block_group(block_group, &ffe_ctl, delalloc);
4234 up_read(&space_info->groups_sem);
4236 ret = find_free_extent_update_loop(fs_info, ins, &ffe_ctl, full_search);
4240 if (ret == -ENOSPC && !cache_block_group_error) {
4242 * Use ffe_ctl->total_free_space as fallback if we can't find
4243 * any contiguous hole.
4245 if (!ffe_ctl.max_extent_size)
4246 ffe_ctl.max_extent_size = ffe_ctl.total_free_space;
4247 spin_lock(&space_info->lock);
4248 space_info->max_extent_size = ffe_ctl.max_extent_size;
4249 spin_unlock(&space_info->lock);
4250 ins->offset = ffe_ctl.max_extent_size;
4251 } else if (ret == -ENOSPC) {
4252 ret = cache_block_group_error;
4258 * btrfs_reserve_extent - entry point to the extent allocator. Tries to find a
4259 * hole that is at least as big as @num_bytes.
4261 * @root - The root that will contain this extent
4263 * @ram_bytes - The amount of space in ram that @num_bytes take. This
4264 * is used for accounting purposes. This value differs
4265 * from @num_bytes only in the case of compressed extents.
4267 * @num_bytes - Number of bytes to allocate on-disk.
4269 * @min_alloc_size - Indicates the minimum amount of space that the
4270 * allocator should try to satisfy. In some cases
4271 * @num_bytes may be larger than what is required and if
4272 * the filesystem is fragmented then allocation fails.
4273 * However, the presence of @min_alloc_size gives a
4274 * chance to try and satisfy the smaller allocation.
4276 * @empty_size - A hint that you plan on doing more COW. This is the
4277 * size in bytes the allocator should try to find free
4278 * next to the block it returns. This is just a hint and
4279 * may be ignored by the allocator.
4281 * @hint_byte - Hint to the allocator to start searching above the byte
4282 * address passed. It might be ignored.
4284 * @ins - This key is modified to record the found hole. It will
4285 * have the following values:
4286 * ins->objectid == start position
4287 * ins->flags = BTRFS_EXTENT_ITEM_KEY
4288 * ins->offset == the size of the hole.
4290 * @is_data - Boolean flag indicating whether an extent is
4291 * allocated for data (true) or metadata (false)
4293 * @delalloc - Boolean flag indicating whether this allocation is for
4294 * delalloc or not. If 'true' data_rwsem of block groups
4295 * is going to be acquired.
4298 * Returns 0 when an allocation succeeded or < 0 when an error occurred. In
4299 * case -ENOSPC is returned then @ins->offset will contain the size of the
4300 * largest available hole the allocator managed to find.
4302 int btrfs_reserve_extent(struct btrfs_root *root, u64 ram_bytes,
4303 u64 num_bytes, u64 min_alloc_size,
4304 u64 empty_size, u64 hint_byte,
4305 struct btrfs_key *ins, int is_data, int delalloc)
4307 struct btrfs_fs_info *fs_info = root->fs_info;
4308 bool final_tried = num_bytes == min_alloc_size;
4312 flags = get_alloc_profile_by_root(root, is_data);
4314 WARN_ON(num_bytes < fs_info->sectorsize);
4315 ret = find_free_extent(root, ram_bytes, num_bytes, empty_size,
4316 hint_byte, ins, flags, delalloc);
4317 if (!ret && !is_data) {
4318 btrfs_dec_block_group_reservations(fs_info, ins->objectid);
4319 } else if (ret == -ENOSPC) {
4320 if (!final_tried && ins->offset) {
4321 num_bytes = min(num_bytes >> 1, ins->offset);
4322 num_bytes = round_down(num_bytes,
4323 fs_info->sectorsize);
4324 num_bytes = max(num_bytes, min_alloc_size);
4325 ram_bytes = num_bytes;
4326 if (num_bytes == min_alloc_size)
4329 } else if (btrfs_test_opt(fs_info, ENOSPC_DEBUG)) {
4330 struct btrfs_space_info *sinfo;
4332 sinfo = btrfs_find_space_info(fs_info, flags);
4334 "allocation failed flags %llu, wanted %llu",
4337 btrfs_dump_space_info(fs_info, sinfo,
4345 int btrfs_free_reserved_extent(struct btrfs_fs_info *fs_info,
4346 u64 start, u64 len, int delalloc)
4348 struct btrfs_block_group *cache;
4350 cache = btrfs_lookup_block_group(fs_info, start);
4352 btrfs_err(fs_info, "Unable to find block group for %llu",
4357 btrfs_add_free_space(cache, start, len);
4358 btrfs_free_reserved_bytes(cache, len, delalloc);
4359 trace_btrfs_reserved_extent_free(fs_info, start, len);
4361 btrfs_put_block_group(cache);
4365 int btrfs_pin_reserved_extent(struct btrfs_trans_handle *trans, u64 start,
4368 struct btrfs_block_group *cache;
4371 cache = btrfs_lookup_block_group(trans->fs_info, start);
4373 btrfs_err(trans->fs_info, "unable to find block group for %llu",
4378 ret = pin_down_extent(trans, cache, start, len, 1);
4379 btrfs_put_block_group(cache);
4383 static int alloc_reserved_file_extent(struct btrfs_trans_handle *trans,
4384 u64 parent, u64 root_objectid,
4385 u64 flags, u64 owner, u64 offset,
4386 struct btrfs_key *ins, int ref_mod)
4388 struct btrfs_fs_info *fs_info = trans->fs_info;
4390 struct btrfs_extent_item *extent_item;
4391 struct btrfs_extent_inline_ref *iref;
4392 struct btrfs_path *path;
4393 struct extent_buffer *leaf;
4398 type = BTRFS_SHARED_DATA_REF_KEY;
4400 type = BTRFS_EXTENT_DATA_REF_KEY;
4402 size = sizeof(*extent_item) + btrfs_extent_inline_ref_size(type);
4404 path = btrfs_alloc_path();
4408 ret = btrfs_insert_empty_item(trans, fs_info->extent_root, path,
4411 btrfs_free_path(path);
4415 leaf = path->nodes[0];
4416 extent_item = btrfs_item_ptr(leaf, path->slots[0],
4417 struct btrfs_extent_item);
4418 btrfs_set_extent_refs(leaf, extent_item, ref_mod);
4419 btrfs_set_extent_generation(leaf, extent_item, trans->transid);
4420 btrfs_set_extent_flags(leaf, extent_item,
4421 flags | BTRFS_EXTENT_FLAG_DATA);
4423 iref = (struct btrfs_extent_inline_ref *)(extent_item + 1);
4424 btrfs_set_extent_inline_ref_type(leaf, iref, type);
4426 struct btrfs_shared_data_ref *ref;
4427 ref = (struct btrfs_shared_data_ref *)(iref + 1);
4428 btrfs_set_extent_inline_ref_offset(leaf, iref, parent);
4429 btrfs_set_shared_data_ref_count(leaf, ref, ref_mod);
4431 struct btrfs_extent_data_ref *ref;
4432 ref = (struct btrfs_extent_data_ref *)(&iref->offset);
4433 btrfs_set_extent_data_ref_root(leaf, ref, root_objectid);
4434 btrfs_set_extent_data_ref_objectid(leaf, ref, owner);
4435 btrfs_set_extent_data_ref_offset(leaf, ref, offset);
4436 btrfs_set_extent_data_ref_count(leaf, ref, ref_mod);
4439 btrfs_mark_buffer_dirty(path->nodes[0]);
4440 btrfs_free_path(path);
4442 ret = remove_from_free_space_tree(trans, ins->objectid, ins->offset);
4446 ret = btrfs_update_block_group(trans, ins->objectid, ins->offset, 1);
4447 if (ret) { /* -ENOENT, logic error */
4448 btrfs_err(fs_info, "update block group failed for %llu %llu",
4449 ins->objectid, ins->offset);
4452 trace_btrfs_reserved_extent_alloc(fs_info, ins->objectid, ins->offset);
4456 static int alloc_reserved_tree_block(struct btrfs_trans_handle *trans,
4457 struct btrfs_delayed_ref_node *node,
4458 struct btrfs_delayed_extent_op *extent_op)
4460 struct btrfs_fs_info *fs_info = trans->fs_info;
4462 struct btrfs_extent_item *extent_item;
4463 struct btrfs_key extent_key;
4464 struct btrfs_tree_block_info *block_info;
4465 struct btrfs_extent_inline_ref *iref;
4466 struct btrfs_path *path;
4467 struct extent_buffer *leaf;
4468 struct btrfs_delayed_tree_ref *ref;
4469 u32 size = sizeof(*extent_item) + sizeof(*iref);
4471 u64 flags = extent_op->flags_to_set;
4472 bool skinny_metadata = btrfs_fs_incompat(fs_info, SKINNY_METADATA);
4474 ref = btrfs_delayed_node_to_tree_ref(node);
4476 extent_key.objectid = node->bytenr;
4477 if (skinny_metadata) {
4478 extent_key.offset = ref->level;
4479 extent_key.type = BTRFS_METADATA_ITEM_KEY;
4480 num_bytes = fs_info->nodesize;
4482 extent_key.offset = node->num_bytes;
4483 extent_key.type = BTRFS_EXTENT_ITEM_KEY;
4484 size += sizeof(*block_info);
4485 num_bytes = node->num_bytes;
4488 path = btrfs_alloc_path();
4492 ret = btrfs_insert_empty_item(trans, fs_info->extent_root, path,
4495 btrfs_free_path(path);
4499 leaf = path->nodes[0];
4500 extent_item = btrfs_item_ptr(leaf, path->slots[0],
4501 struct btrfs_extent_item);
4502 btrfs_set_extent_refs(leaf, extent_item, 1);
4503 btrfs_set_extent_generation(leaf, extent_item, trans->transid);
4504 btrfs_set_extent_flags(leaf, extent_item,
4505 flags | BTRFS_EXTENT_FLAG_TREE_BLOCK);
4507 if (skinny_metadata) {
4508 iref = (struct btrfs_extent_inline_ref *)(extent_item + 1);
4510 block_info = (struct btrfs_tree_block_info *)(extent_item + 1);
4511 btrfs_set_tree_block_key(leaf, block_info, &extent_op->key);
4512 btrfs_set_tree_block_level(leaf, block_info, ref->level);
4513 iref = (struct btrfs_extent_inline_ref *)(block_info + 1);
4516 if (node->type == BTRFS_SHARED_BLOCK_REF_KEY) {
4517 BUG_ON(!(flags & BTRFS_BLOCK_FLAG_FULL_BACKREF));
4518 btrfs_set_extent_inline_ref_type(leaf, iref,
4519 BTRFS_SHARED_BLOCK_REF_KEY);
4520 btrfs_set_extent_inline_ref_offset(leaf, iref, ref->parent);
4522 btrfs_set_extent_inline_ref_type(leaf, iref,
4523 BTRFS_TREE_BLOCK_REF_KEY);
4524 btrfs_set_extent_inline_ref_offset(leaf, iref, ref->root);
4527 btrfs_mark_buffer_dirty(leaf);
4528 btrfs_free_path(path);
4530 ret = remove_from_free_space_tree(trans, extent_key.objectid,
4535 ret = btrfs_update_block_group(trans, extent_key.objectid,
4536 fs_info->nodesize, 1);
4537 if (ret) { /* -ENOENT, logic error */
4538 btrfs_err(fs_info, "update block group failed for %llu %llu",
4539 extent_key.objectid, extent_key.offset);
4543 trace_btrfs_reserved_extent_alloc(fs_info, extent_key.objectid,
4548 int btrfs_alloc_reserved_file_extent(struct btrfs_trans_handle *trans,
4549 struct btrfs_root *root, u64 owner,
4550 u64 offset, u64 ram_bytes,
4551 struct btrfs_key *ins)
4553 struct btrfs_ref generic_ref = { 0 };
4556 BUG_ON(root->root_key.objectid == BTRFS_TREE_LOG_OBJECTID);
4558 btrfs_init_generic_ref(&generic_ref, BTRFS_ADD_DELAYED_EXTENT,
4559 ins->objectid, ins->offset, 0);
4560 btrfs_init_data_ref(&generic_ref, root->root_key.objectid, owner, offset);
4561 btrfs_ref_tree_mod(root->fs_info, &generic_ref);
4562 ret = btrfs_add_delayed_data_ref(trans, &generic_ref,
4563 ram_bytes, NULL, NULL);
4568 * this is used by the tree logging recovery code. It records that
4569 * an extent has been allocated and makes sure to clear the free
4570 * space cache bits as well
4572 int btrfs_alloc_logged_file_extent(struct btrfs_trans_handle *trans,
4573 u64 root_objectid, u64 owner, u64 offset,
4574 struct btrfs_key *ins)
4576 struct btrfs_fs_info *fs_info = trans->fs_info;
4578 struct btrfs_block_group *block_group;
4579 struct btrfs_space_info *space_info;
4582 * Mixed block groups will exclude before processing the log so we only
4583 * need to do the exclude dance if this fs isn't mixed.
4585 if (!btrfs_fs_incompat(fs_info, MIXED_GROUPS)) {
4586 ret = __exclude_logged_extent(fs_info, ins->objectid,
4592 block_group = btrfs_lookup_block_group(fs_info, ins->objectid);
4596 space_info = block_group->space_info;
4597 spin_lock(&space_info->lock);
4598 spin_lock(&block_group->lock);
4599 space_info->bytes_reserved += ins->offset;
4600 block_group->reserved += ins->offset;
4601 spin_unlock(&block_group->lock);
4602 spin_unlock(&space_info->lock);
4604 ret = alloc_reserved_file_extent(trans, 0, root_objectid, 0, owner,
4607 btrfs_pin_extent(trans, ins->objectid, ins->offset, 1);
4608 btrfs_put_block_group(block_group);
4612 static struct extent_buffer *
4613 btrfs_init_new_buffer(struct btrfs_trans_handle *trans, struct btrfs_root *root,
4614 u64 bytenr, int level, u64 owner,
4615 enum btrfs_lock_nesting nest)
4617 struct btrfs_fs_info *fs_info = root->fs_info;
4618 struct extent_buffer *buf;
4620 buf = btrfs_find_create_tree_block(fs_info, bytenr, owner, level);
4625 * Extra safety check in case the extent tree is corrupted and extent
4626 * allocator chooses to use a tree block which is already used and
4629 if (buf->lock_owner == current->pid) {
4630 btrfs_err_rl(fs_info,
4631 "tree block %llu owner %llu already locked by pid=%d, extent tree corruption detected",
4632 buf->start, btrfs_header_owner(buf), current->pid);
4633 free_extent_buffer(buf);
4634 return ERR_PTR(-EUCLEAN);
4638 * This needs to stay, because we could allocate a freed block from an
4639 * old tree into a new tree, so we need to make sure this new block is
4640 * set to the appropriate level and owner.
4642 btrfs_set_buffer_lockdep_class(owner, buf, level);
4643 __btrfs_tree_lock(buf, nest);
4644 btrfs_clean_tree_block(buf);
4645 clear_bit(EXTENT_BUFFER_STALE, &buf->bflags);
4647 set_extent_buffer_uptodate(buf);
4649 memzero_extent_buffer(buf, 0, sizeof(struct btrfs_header));
4650 btrfs_set_header_level(buf, level);
4651 btrfs_set_header_bytenr(buf, buf->start);
4652 btrfs_set_header_generation(buf, trans->transid);
4653 btrfs_set_header_backref_rev(buf, BTRFS_MIXED_BACKREF_REV);
4654 btrfs_set_header_owner(buf, owner);
4655 write_extent_buffer_fsid(buf, fs_info->fs_devices->metadata_uuid);
4656 write_extent_buffer_chunk_tree_uuid(buf, fs_info->chunk_tree_uuid);
4657 if (root->root_key.objectid == BTRFS_TREE_LOG_OBJECTID) {
4658 buf->log_index = root->log_transid % 2;
4660 * we allow two log transactions at a time, use different
4661 * EXTENT bit to differentiate dirty pages.
4663 if (buf->log_index == 0)
4664 set_extent_dirty(&root->dirty_log_pages, buf->start,
4665 buf->start + buf->len - 1, GFP_NOFS);
4667 set_extent_new(&root->dirty_log_pages, buf->start,
4668 buf->start + buf->len - 1);
4670 buf->log_index = -1;
4671 set_extent_dirty(&trans->transaction->dirty_pages, buf->start,
4672 buf->start + buf->len - 1, GFP_NOFS);
4674 trans->dirty = true;
4675 /* this returns a buffer locked for blocking */
4680 * finds a free extent and does all the dirty work required for allocation
4681 * returns the tree buffer or an ERR_PTR on error.
4683 struct extent_buffer *btrfs_alloc_tree_block(struct btrfs_trans_handle *trans,
4684 struct btrfs_root *root,
4685 u64 parent, u64 root_objectid,
4686 const struct btrfs_disk_key *key,
4687 int level, u64 hint,
4689 enum btrfs_lock_nesting nest)
4691 struct btrfs_fs_info *fs_info = root->fs_info;
4692 struct btrfs_key ins;
4693 struct btrfs_block_rsv *block_rsv;
4694 struct extent_buffer *buf;
4695 struct btrfs_delayed_extent_op *extent_op;
4696 struct btrfs_ref generic_ref = { 0 };
4699 u32 blocksize = fs_info->nodesize;
4700 bool skinny_metadata = btrfs_fs_incompat(fs_info, SKINNY_METADATA);
4702 #ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
4703 if (btrfs_is_testing(fs_info)) {
4704 buf = btrfs_init_new_buffer(trans, root, root->alloc_bytenr,
4705 level, root_objectid, nest);
4707 root->alloc_bytenr += blocksize;
4712 block_rsv = btrfs_use_block_rsv(trans, root, blocksize);
4713 if (IS_ERR(block_rsv))
4714 return ERR_CAST(block_rsv);
4716 ret = btrfs_reserve_extent(root, blocksize, blocksize, blocksize,
4717 empty_size, hint, &ins, 0, 0);
4721 buf = btrfs_init_new_buffer(trans, root, ins.objectid, level,
4722 root_objectid, nest);
4725 goto out_free_reserved;
4728 if (root_objectid == BTRFS_TREE_RELOC_OBJECTID) {
4730 parent = ins.objectid;
4731 flags |= BTRFS_BLOCK_FLAG_FULL_BACKREF;
4735 if (root_objectid != BTRFS_TREE_LOG_OBJECTID) {
4736 extent_op = btrfs_alloc_delayed_extent_op();
4742 memcpy(&extent_op->key, key, sizeof(extent_op->key));
4744 memset(&extent_op->key, 0, sizeof(extent_op->key));
4745 extent_op->flags_to_set = flags;
4746 extent_op->update_key = skinny_metadata ? false : true;
4747 extent_op->update_flags = true;
4748 extent_op->is_data = false;
4749 extent_op->level = level;
4751 btrfs_init_generic_ref(&generic_ref, BTRFS_ADD_DELAYED_EXTENT,
4752 ins.objectid, ins.offset, parent);
4753 generic_ref.real_root = root->root_key.objectid;
4754 btrfs_init_tree_ref(&generic_ref, level, root_objectid);
4755 btrfs_ref_tree_mod(fs_info, &generic_ref);
4756 ret = btrfs_add_delayed_tree_ref(trans, &generic_ref,
4757 extent_op, NULL, NULL);
4759 goto out_free_delayed;
4764 btrfs_free_delayed_extent_op(extent_op);
4766 free_extent_buffer(buf);
4768 btrfs_free_reserved_extent(fs_info, ins.objectid, ins.offset, 0);
4770 btrfs_unuse_block_rsv(fs_info, block_rsv, blocksize);
4771 return ERR_PTR(ret);
4774 struct walk_control {
4775 u64 refs[BTRFS_MAX_LEVEL];
4776 u64 flags[BTRFS_MAX_LEVEL];
4777 struct btrfs_key update_progress;
4778 struct btrfs_key drop_progress;
4790 #define DROP_REFERENCE 1
4791 #define UPDATE_BACKREF 2
4793 static noinline void reada_walk_down(struct btrfs_trans_handle *trans,
4794 struct btrfs_root *root,
4795 struct walk_control *wc,
4796 struct btrfs_path *path)
4798 struct btrfs_fs_info *fs_info = root->fs_info;
4804 struct btrfs_key key;
4805 struct extent_buffer *eb;
4810 if (path->slots[wc->level] < wc->reada_slot) {
4811 wc->reada_count = wc->reada_count * 2 / 3;
4812 wc->reada_count = max(wc->reada_count, 2);
4814 wc->reada_count = wc->reada_count * 3 / 2;
4815 wc->reada_count = min_t(int, wc->reada_count,
4816 BTRFS_NODEPTRS_PER_BLOCK(fs_info));
4819 eb = path->nodes[wc->level];
4820 nritems = btrfs_header_nritems(eb);
4822 for (slot = path->slots[wc->level]; slot < nritems; slot++) {
4823 if (nread >= wc->reada_count)
4827 bytenr = btrfs_node_blockptr(eb, slot);
4828 generation = btrfs_node_ptr_generation(eb, slot);
4830 if (slot == path->slots[wc->level])
4833 if (wc->stage == UPDATE_BACKREF &&
4834 generation <= root->root_key.offset)
4837 /* We don't lock the tree block, it's OK to be racy here */
4838 ret = btrfs_lookup_extent_info(trans, fs_info, bytenr,
4839 wc->level - 1, 1, &refs,
4841 /* We don't care about errors in readahead. */
4846 if (wc->stage == DROP_REFERENCE) {
4850 if (wc->level == 1 &&
4851 (flags & BTRFS_BLOCK_FLAG_FULL_BACKREF))
4853 if (!wc->update_ref ||
4854 generation <= root->root_key.offset)
4856 btrfs_node_key_to_cpu(eb, &key, slot);
4857 ret = btrfs_comp_cpu_keys(&key,
4858 &wc->update_progress);
4862 if (wc->level == 1 &&
4863 (flags & BTRFS_BLOCK_FLAG_FULL_BACKREF))
4867 btrfs_readahead_node_child(eb, slot);
4870 wc->reada_slot = slot;
4874 * helper to process tree block while walking down the tree.
4876 * when wc->stage == UPDATE_BACKREF, this function updates
4877 * back refs for pointers in the block.
4879 * NOTE: return value 1 means we should stop walking down.
4881 static noinline int walk_down_proc(struct btrfs_trans_handle *trans,
4882 struct btrfs_root *root,
4883 struct btrfs_path *path,
4884 struct walk_control *wc, int lookup_info)
4886 struct btrfs_fs_info *fs_info = root->fs_info;
4887 int level = wc->level;
4888 struct extent_buffer *eb = path->nodes[level];
4889 u64 flag = BTRFS_BLOCK_FLAG_FULL_BACKREF;
4892 if (wc->stage == UPDATE_BACKREF &&
4893 btrfs_header_owner(eb) != root->root_key.objectid)
4897 * when reference count of tree block is 1, it won't increase
4898 * again. once full backref flag is set, we never clear it.
4901 ((wc->stage == DROP_REFERENCE && wc->refs[level] != 1) ||
4902 (wc->stage == UPDATE_BACKREF && !(wc->flags[level] & flag)))) {
4903 BUG_ON(!path->locks[level]);
4904 ret = btrfs_lookup_extent_info(trans, fs_info,
4905 eb->start, level, 1,
4908 BUG_ON(ret == -ENOMEM);
4911 BUG_ON(wc->refs[level] == 0);
4914 if (wc->stage == DROP_REFERENCE) {
4915 if (wc->refs[level] > 1)
4918 if (path->locks[level] && !wc->keep_locks) {
4919 btrfs_tree_unlock_rw(eb, path->locks[level]);
4920 path->locks[level] = 0;
4925 /* wc->stage == UPDATE_BACKREF */
4926 if (!(wc->flags[level] & flag)) {
4927 BUG_ON(!path->locks[level]);
4928 ret = btrfs_inc_ref(trans, root, eb, 1);
4929 BUG_ON(ret); /* -ENOMEM */
4930 ret = btrfs_dec_ref(trans, root, eb, 0);
4931 BUG_ON(ret); /* -ENOMEM */
4932 ret = btrfs_set_disk_extent_flags(trans, eb, flag,
4933 btrfs_header_level(eb), 0);
4934 BUG_ON(ret); /* -ENOMEM */
4935 wc->flags[level] |= flag;
4939 * the block is shared by multiple trees, so it's not good to
4940 * keep the tree lock
4942 if (path->locks[level] && level > 0) {
4943 btrfs_tree_unlock_rw(eb, path->locks[level]);
4944 path->locks[level] = 0;
4950 * This is used to verify a ref exists for this root to deal with a bug where we
4951 * would have a drop_progress key that hadn't been updated properly.
4953 static int check_ref_exists(struct btrfs_trans_handle *trans,
4954 struct btrfs_root *root, u64 bytenr, u64 parent,
4957 struct btrfs_path *path;
4958 struct btrfs_extent_inline_ref *iref;
4961 path = btrfs_alloc_path();
4965 ret = lookup_extent_backref(trans, path, &iref, bytenr,
4966 root->fs_info->nodesize, parent,
4967 root->root_key.objectid, level, 0);
4968 btrfs_free_path(path);
4977 * helper to process tree block pointer.
4979 * when wc->stage == DROP_REFERENCE, this function checks
4980 * reference count of the block pointed to. if the block
4981 * is shared and we need update back refs for the subtree
4982 * rooted at the block, this function changes wc->stage to
4983 * UPDATE_BACKREF. if the block is shared and there is no
4984 * need to update back, this function drops the reference
4987 * NOTE: return value 1 means we should stop walking down.
4989 static noinline int do_walk_down(struct btrfs_trans_handle *trans,
4990 struct btrfs_root *root,
4991 struct btrfs_path *path,
4992 struct walk_control *wc, int *lookup_info)
4994 struct btrfs_fs_info *fs_info = root->fs_info;
4998 struct btrfs_key key;
4999 struct btrfs_key first_key;
5000 struct btrfs_ref ref = { 0 };
5001 struct extent_buffer *next;
5002 int level = wc->level;
5005 bool need_account = false;
5007 generation = btrfs_node_ptr_generation(path->nodes[level],
5008 path->slots[level]);
5010 * if the lower level block was created before the snapshot
5011 * was created, we know there is no need to update back refs
5014 if (wc->stage == UPDATE_BACKREF &&
5015 generation <= root->root_key.offset) {
5020 bytenr = btrfs_node_blockptr(path->nodes[level], path->slots[level]);
5021 btrfs_node_key_to_cpu(path->nodes[level], &first_key,
5022 path->slots[level]);
5024 next = find_extent_buffer(fs_info, bytenr);
5026 next = btrfs_find_create_tree_block(fs_info, bytenr,
5027 root->root_key.objectid, level - 1);
5029 return PTR_ERR(next);
5032 btrfs_tree_lock(next);
5034 ret = btrfs_lookup_extent_info(trans, fs_info, bytenr, level - 1, 1,
5035 &wc->refs[level - 1],
5036 &wc->flags[level - 1]);
5040 if (unlikely(wc->refs[level - 1] == 0)) {
5041 btrfs_err(fs_info, "Missing references.");
5047 if (wc->stage == DROP_REFERENCE) {
5048 if (wc->refs[level - 1] > 1) {
5049 need_account = true;
5051 (wc->flags[0] & BTRFS_BLOCK_FLAG_FULL_BACKREF))
5054 if (!wc->update_ref ||
5055 generation <= root->root_key.offset)
5058 btrfs_node_key_to_cpu(path->nodes[level], &key,
5059 path->slots[level]);
5060 ret = btrfs_comp_cpu_keys(&key, &wc->update_progress);
5064 wc->stage = UPDATE_BACKREF;
5065 wc->shared_level = level - 1;
5069 (wc->flags[0] & BTRFS_BLOCK_FLAG_FULL_BACKREF))
5073 if (!btrfs_buffer_uptodate(next, generation, 0)) {
5074 btrfs_tree_unlock(next);
5075 free_extent_buffer(next);
5081 if (reada && level == 1)
5082 reada_walk_down(trans, root, wc, path);
5083 next = read_tree_block(fs_info, bytenr, root->root_key.objectid,
5084 generation, level - 1, &first_key);
5086 return PTR_ERR(next);
5087 } else if (!extent_buffer_uptodate(next)) {
5088 free_extent_buffer(next);
5091 btrfs_tree_lock(next);
5095 ASSERT(level == btrfs_header_level(next));
5096 if (level != btrfs_header_level(next)) {
5097 btrfs_err(root->fs_info, "mismatched level");
5101 path->nodes[level] = next;
5102 path->slots[level] = 0;
5103 path->locks[level] = BTRFS_WRITE_LOCK;
5109 wc->refs[level - 1] = 0;
5110 wc->flags[level - 1] = 0;
5111 if (wc->stage == DROP_REFERENCE) {
5112 if (wc->flags[level] & BTRFS_BLOCK_FLAG_FULL_BACKREF) {
5113 parent = path->nodes[level]->start;
5115 ASSERT(root->root_key.objectid ==
5116 btrfs_header_owner(path->nodes[level]));
5117 if (root->root_key.objectid !=
5118 btrfs_header_owner(path->nodes[level])) {
5119 btrfs_err(root->fs_info,
5120 "mismatched block owner");
5128 * If we had a drop_progress we need to verify the refs are set
5129 * as expected. If we find our ref then we know that from here
5130 * on out everything should be correct, and we can clear the
5133 if (wc->restarted) {
5134 ret = check_ref_exists(trans, root, bytenr, parent,
5145 * Reloc tree doesn't contribute to qgroup numbers, and we have
5146 * already accounted them at merge time (replace_path),
5147 * thus we could skip expensive subtree trace here.
5149 if (root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID &&
5151 ret = btrfs_qgroup_trace_subtree(trans, next,
5152 generation, level - 1);
5154 btrfs_err_rl(fs_info,
5155 "Error %d accounting shared subtree. Quota is out of sync, rescan required.",
5161 * We need to update the next key in our walk control so we can
5162 * update the drop_progress key accordingly. We don't care if
5163 * find_next_key doesn't find a key because that means we're at
5164 * the end and are going to clean up now.
5166 wc->drop_level = level;
5167 find_next_key(path, level, &wc->drop_progress);
5169 btrfs_init_generic_ref(&ref, BTRFS_DROP_DELAYED_REF, bytenr,
5170 fs_info->nodesize, parent);
5171 btrfs_init_tree_ref(&ref, level - 1, root->root_key.objectid);
5172 ret = btrfs_free_extent(trans, &ref);
5181 btrfs_tree_unlock(next);
5182 free_extent_buffer(next);
5188 * helper to process tree block while walking up the tree.
5190 * when wc->stage == DROP_REFERENCE, this function drops
5191 * reference count on the block.
5193 * when wc->stage == UPDATE_BACKREF, this function changes
5194 * wc->stage back to DROP_REFERENCE if we changed wc->stage
5195 * to UPDATE_BACKREF previously while processing the block.
5197 * NOTE: return value 1 means we should stop walking up.
5199 static noinline int walk_up_proc(struct btrfs_trans_handle *trans,
5200 struct btrfs_root *root,
5201 struct btrfs_path *path,
5202 struct walk_control *wc)
5204 struct btrfs_fs_info *fs_info = root->fs_info;
5206 int level = wc->level;
5207 struct extent_buffer *eb = path->nodes[level];
5210 if (wc->stage == UPDATE_BACKREF) {
5211 BUG_ON(wc->shared_level < level);
5212 if (level < wc->shared_level)
5215 ret = find_next_key(path, level + 1, &wc->update_progress);
5219 wc->stage = DROP_REFERENCE;
5220 wc->shared_level = -1;
5221 path->slots[level] = 0;
5224 * check reference count again if the block isn't locked.
5225 * we should start walking down the tree again if reference
5228 if (!path->locks[level]) {
5230 btrfs_tree_lock(eb);
5231 path->locks[level] = BTRFS_WRITE_LOCK;
5233 ret = btrfs_lookup_extent_info(trans, fs_info,
5234 eb->start, level, 1,
5238 btrfs_tree_unlock_rw(eb, path->locks[level]);
5239 path->locks[level] = 0;
5242 BUG_ON(wc->refs[level] == 0);
5243 if (wc->refs[level] == 1) {
5244 btrfs_tree_unlock_rw(eb, path->locks[level]);
5245 path->locks[level] = 0;
5251 /* wc->stage == DROP_REFERENCE */
5252 BUG_ON(wc->refs[level] > 1 && !path->locks[level]);
5254 if (wc->refs[level] == 1) {
5256 if (wc->flags[level] & BTRFS_BLOCK_FLAG_FULL_BACKREF)
5257 ret = btrfs_dec_ref(trans, root, eb, 1);
5259 ret = btrfs_dec_ref(trans, root, eb, 0);
5260 BUG_ON(ret); /* -ENOMEM */
5261 if (is_fstree(root->root_key.objectid)) {
5262 ret = btrfs_qgroup_trace_leaf_items(trans, eb);
5264 btrfs_err_rl(fs_info,
5265 "error %d accounting leaf items, quota is out of sync, rescan required",
5270 /* make block locked assertion in btrfs_clean_tree_block happy */
5271 if (!path->locks[level] &&
5272 btrfs_header_generation(eb) == trans->transid) {
5273 btrfs_tree_lock(eb);
5274 path->locks[level] = BTRFS_WRITE_LOCK;
5276 btrfs_clean_tree_block(eb);
5279 if (eb == root->node) {
5280 if (wc->flags[level] & BTRFS_BLOCK_FLAG_FULL_BACKREF)
5282 else if (root->root_key.objectid != btrfs_header_owner(eb))
5283 goto owner_mismatch;
5285 if (wc->flags[level + 1] & BTRFS_BLOCK_FLAG_FULL_BACKREF)
5286 parent = path->nodes[level + 1]->start;
5287 else if (root->root_key.objectid !=
5288 btrfs_header_owner(path->nodes[level + 1]))
5289 goto owner_mismatch;
5292 btrfs_free_tree_block(trans, root, eb, parent, wc->refs[level] == 1);
5294 wc->refs[level] = 0;
5295 wc->flags[level] = 0;
5299 btrfs_err_rl(fs_info, "unexpected tree owner, have %llu expect %llu",
5300 btrfs_header_owner(eb), root->root_key.objectid);
5304 static noinline int walk_down_tree(struct btrfs_trans_handle *trans,
5305 struct btrfs_root *root,
5306 struct btrfs_path *path,
5307 struct walk_control *wc)
5309 int level = wc->level;
5310 int lookup_info = 1;
5313 while (level >= 0) {
5314 ret = walk_down_proc(trans, root, path, wc, lookup_info);
5321 if (path->slots[level] >=
5322 btrfs_header_nritems(path->nodes[level]))
5325 ret = do_walk_down(trans, root, path, wc, &lookup_info);
5327 path->slots[level]++;
5336 static noinline int walk_up_tree(struct btrfs_trans_handle *trans,
5337 struct btrfs_root *root,
5338 struct btrfs_path *path,
5339 struct walk_control *wc, int max_level)
5341 int level = wc->level;
5344 path->slots[level] = btrfs_header_nritems(path->nodes[level]);
5345 while (level < max_level && path->nodes[level]) {
5347 if (path->slots[level] + 1 <
5348 btrfs_header_nritems(path->nodes[level])) {
5349 path->slots[level]++;
5352 ret = walk_up_proc(trans, root, path, wc);
5358 if (path->locks[level]) {
5359 btrfs_tree_unlock_rw(path->nodes[level],
5360 path->locks[level]);
5361 path->locks[level] = 0;
5363 free_extent_buffer(path->nodes[level]);
5364 path->nodes[level] = NULL;
5372 * drop a subvolume tree.
5374 * this function traverses the tree freeing any blocks that only
5375 * referenced by the tree.
5377 * when a shared tree block is found. this function decreases its
5378 * reference count by one. if update_ref is true, this function
5379 * also make sure backrefs for the shared block and all lower level
5380 * blocks are properly updated.
5382 * If called with for_reloc == 0, may exit early with -EAGAIN
5384 int btrfs_drop_snapshot(struct btrfs_root *root, int update_ref, int for_reloc)
5386 struct btrfs_fs_info *fs_info = root->fs_info;
5387 struct btrfs_path *path;
5388 struct btrfs_trans_handle *trans;
5389 struct btrfs_root *tree_root = fs_info->tree_root;
5390 struct btrfs_root_item *root_item = &root->root_item;
5391 struct walk_control *wc;
5392 struct btrfs_key key;
5396 bool root_dropped = false;
5398 btrfs_debug(fs_info, "Drop subvolume %llu", root->root_key.objectid);
5400 path = btrfs_alloc_path();
5406 wc = kzalloc(sizeof(*wc), GFP_NOFS);
5408 btrfs_free_path(path);
5414 * Use join to avoid potential EINTR from transaction start. See
5415 * wait_reserve_ticket and the whole reservation callchain.
5418 trans = btrfs_join_transaction(tree_root);
5420 trans = btrfs_start_transaction(tree_root, 0);
5421 if (IS_ERR(trans)) {
5422 err = PTR_ERR(trans);
5426 err = btrfs_run_delayed_items(trans);
5431 * This will help us catch people modifying the fs tree while we're
5432 * dropping it. It is unsafe to mess with the fs tree while it's being
5433 * dropped as we unlock the root node and parent nodes as we walk down
5434 * the tree, assuming nothing will change. If something does change
5435 * then we'll have stale information and drop references to blocks we've
5438 set_bit(BTRFS_ROOT_DELETING, &root->state);
5439 if (btrfs_disk_key_objectid(&root_item->drop_progress) == 0) {
5440 level = btrfs_header_level(root->node);
5441 path->nodes[level] = btrfs_lock_root_node(root);
5442 path->slots[level] = 0;
5443 path->locks[level] = BTRFS_WRITE_LOCK;
5444 memset(&wc->update_progress, 0,
5445 sizeof(wc->update_progress));
5447 btrfs_disk_key_to_cpu(&key, &root_item->drop_progress);
5448 memcpy(&wc->update_progress, &key,
5449 sizeof(wc->update_progress));
5451 level = btrfs_root_drop_level(root_item);
5453 path->lowest_level = level;
5454 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
5455 path->lowest_level = 0;
5463 * unlock our path, this is safe because only this
5464 * function is allowed to delete this snapshot
5466 btrfs_unlock_up_safe(path, 0);
5468 level = btrfs_header_level(root->node);
5470 btrfs_tree_lock(path->nodes[level]);
5471 path->locks[level] = BTRFS_WRITE_LOCK;
5473 ret = btrfs_lookup_extent_info(trans, fs_info,
5474 path->nodes[level]->start,
5475 level, 1, &wc->refs[level],
5481 BUG_ON(wc->refs[level] == 0);
5483 if (level == btrfs_root_drop_level(root_item))
5486 btrfs_tree_unlock(path->nodes[level]);
5487 path->locks[level] = 0;
5488 WARN_ON(wc->refs[level] != 1);
5493 wc->restarted = test_bit(BTRFS_ROOT_DEAD_TREE, &root->state);
5495 wc->shared_level = -1;
5496 wc->stage = DROP_REFERENCE;
5497 wc->update_ref = update_ref;
5499 wc->reada_count = BTRFS_NODEPTRS_PER_BLOCK(fs_info);
5503 ret = walk_down_tree(trans, root, path, wc);
5509 ret = walk_up_tree(trans, root, path, wc, BTRFS_MAX_LEVEL);
5516 BUG_ON(wc->stage != DROP_REFERENCE);
5520 if (wc->stage == DROP_REFERENCE) {
5521 wc->drop_level = wc->level;
5522 btrfs_node_key_to_cpu(path->nodes[wc->drop_level],
5524 path->slots[wc->drop_level]);
5526 btrfs_cpu_key_to_disk(&root_item->drop_progress,
5527 &wc->drop_progress);
5528 btrfs_set_root_drop_level(root_item, wc->drop_level);
5530 BUG_ON(wc->level == 0);
5531 if (btrfs_should_end_transaction(trans) ||
5532 (!for_reloc && btrfs_need_cleaner_sleep(fs_info))) {
5533 ret = btrfs_update_root(trans, tree_root,
5537 btrfs_abort_transaction(trans, ret);
5542 btrfs_end_transaction_throttle(trans);
5543 if (!for_reloc && btrfs_need_cleaner_sleep(fs_info)) {
5544 btrfs_debug(fs_info,
5545 "drop snapshot early exit");
5550 trans = btrfs_start_transaction(tree_root, 0);
5551 if (IS_ERR(trans)) {
5552 err = PTR_ERR(trans);
5557 btrfs_release_path(path);
5561 ret = btrfs_del_root(trans, &root->root_key);
5563 btrfs_abort_transaction(trans, ret);
5568 if (root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID) {
5569 ret = btrfs_find_root(tree_root, &root->root_key, path,
5572 btrfs_abort_transaction(trans, ret);
5575 } else if (ret > 0) {
5576 /* if we fail to delete the orphan item this time
5577 * around, it'll get picked up the next time.
5579 * The most common failure here is just -ENOENT.
5581 btrfs_del_orphan_item(trans, tree_root,
5582 root->root_key.objectid);
5587 * This subvolume is going to be completely dropped, and won't be
5588 * recorded as dirty roots, thus pertrans meta rsv will not be freed at
5589 * commit transaction time. So free it here manually.
5591 btrfs_qgroup_convert_reserved_meta(root, INT_MAX);
5592 btrfs_qgroup_free_meta_all_pertrans(root);
5594 if (test_bit(BTRFS_ROOT_IN_RADIX, &root->state))
5595 btrfs_add_dropped_root(trans, root);
5597 btrfs_put_root(root);
5598 root_dropped = true;
5600 btrfs_end_transaction_throttle(trans);
5603 btrfs_free_path(path);
5606 * So if we need to stop dropping the snapshot for whatever reason we
5607 * need to make sure to add it back to the dead root list so that we
5608 * keep trying to do the work later. This also cleans up roots if we
5609 * don't have it in the radix (like when we recover after a power fail
5610 * or unmount) so we don't leak memory.
5612 if (!for_reloc && !root_dropped)
5613 btrfs_add_dead_root(root);
5618 * drop subtree rooted at tree block 'node'.
5620 * NOTE: this function will unlock and release tree block 'node'
5621 * only used by relocation code
5623 int btrfs_drop_subtree(struct btrfs_trans_handle *trans,
5624 struct btrfs_root *root,
5625 struct extent_buffer *node,
5626 struct extent_buffer *parent)
5628 struct btrfs_fs_info *fs_info = root->fs_info;
5629 struct btrfs_path *path;
5630 struct walk_control *wc;
5636 BUG_ON(root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID);
5638 path = btrfs_alloc_path();
5642 wc = kzalloc(sizeof(*wc), GFP_NOFS);
5644 btrfs_free_path(path);
5648 btrfs_assert_tree_locked(parent);
5649 parent_level = btrfs_header_level(parent);
5650 atomic_inc(&parent->refs);
5651 path->nodes[parent_level] = parent;
5652 path->slots[parent_level] = btrfs_header_nritems(parent);
5654 btrfs_assert_tree_locked(node);
5655 level = btrfs_header_level(node);
5656 path->nodes[level] = node;
5657 path->slots[level] = 0;
5658 path->locks[level] = BTRFS_WRITE_LOCK;
5660 wc->refs[parent_level] = 1;
5661 wc->flags[parent_level] = BTRFS_BLOCK_FLAG_FULL_BACKREF;
5663 wc->shared_level = -1;
5664 wc->stage = DROP_REFERENCE;
5667 wc->reada_count = BTRFS_NODEPTRS_PER_BLOCK(fs_info);
5670 wret = walk_down_tree(trans, root, path, wc);
5676 wret = walk_up_tree(trans, root, path, wc, parent_level);
5684 btrfs_free_path(path);
5689 * helper to account the unused space of all the readonly block group in the
5690 * space_info. takes mirrors into account.
5692 u64 btrfs_account_ro_block_groups_free_space(struct btrfs_space_info *sinfo)
5694 struct btrfs_block_group *block_group;
5698 /* It's df, we don't care if it's racy */
5699 if (list_empty(&sinfo->ro_bgs))
5702 spin_lock(&sinfo->lock);
5703 list_for_each_entry(block_group, &sinfo->ro_bgs, ro_list) {
5704 spin_lock(&block_group->lock);
5706 if (!block_group->ro) {
5707 spin_unlock(&block_group->lock);
5711 factor = btrfs_bg_type_to_factor(block_group->flags);
5712 free_bytes += (block_group->length -
5713 block_group->used) * factor;
5715 spin_unlock(&block_group->lock);
5717 spin_unlock(&sinfo->lock);
5722 int btrfs_error_unpin_extent_range(struct btrfs_fs_info *fs_info,
5725 return unpin_extent_range(fs_info, start, end, false);
5729 * It used to be that old block groups would be left around forever.
5730 * Iterating over them would be enough to trim unused space. Since we
5731 * now automatically remove them, we also need to iterate over unallocated
5734 * We don't want a transaction for this since the discard may take a
5735 * substantial amount of time. We don't require that a transaction be
5736 * running, but we do need to take a running transaction into account
5737 * to ensure that we're not discarding chunks that were released or
5738 * allocated in the current transaction.
5740 * Holding the chunks lock will prevent other threads from allocating
5741 * or releasing chunks, but it won't prevent a running transaction
5742 * from committing and releasing the memory that the pending chunks
5743 * list head uses. For that, we need to take a reference to the
5744 * transaction and hold the commit root sem. We only need to hold
5745 * it while performing the free space search since we have already
5746 * held back allocations.
5748 static int btrfs_trim_free_extents(struct btrfs_device *device, u64 *trimmed)
5750 u64 start = SZ_1M, len = 0, end = 0;
5755 /* Discard not supported = nothing to do. */
5756 if (!blk_queue_discard(bdev_get_queue(device->bdev)))
5759 /* Not writable = nothing to do. */
5760 if (!test_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state))
5763 /* No free space = nothing to do. */
5764 if (device->total_bytes <= device->bytes_used)
5770 struct btrfs_fs_info *fs_info = device->fs_info;
5773 ret = mutex_lock_interruptible(&fs_info->chunk_mutex);
5777 find_first_clear_extent_bit(&device->alloc_state, start,
5779 CHUNK_TRIMMED | CHUNK_ALLOCATED);
5781 /* Check if there are any CHUNK_* bits left */
5782 if (start > device->total_bytes) {
5783 WARN_ON(IS_ENABLED(CONFIG_BTRFS_DEBUG));
5784 btrfs_warn_in_rcu(fs_info,
5785 "ignoring attempt to trim beyond device size: offset %llu length %llu device %s device size %llu",
5786 start, end - start + 1,
5787 rcu_str_deref(device->name),
5788 device->total_bytes);
5789 mutex_unlock(&fs_info->chunk_mutex);
5794 /* Ensure we skip the reserved area in the first 1M */
5795 start = max_t(u64, start, SZ_1M);
5798 * If find_first_clear_extent_bit find a range that spans the
5799 * end of the device it will set end to -1, in this case it's up
5800 * to the caller to trim the value to the size of the device.
5802 end = min(end, device->total_bytes - 1);
5804 len = end - start + 1;
5806 /* We didn't find any extents */
5808 mutex_unlock(&fs_info->chunk_mutex);
5813 ret = btrfs_issue_discard(device->bdev, start, len,
5816 set_extent_bits(&device->alloc_state, start,
5819 mutex_unlock(&fs_info->chunk_mutex);
5827 if (fatal_signal_pending(current)) {
5839 * Trim the whole filesystem by:
5840 * 1) trimming the free space in each block group
5841 * 2) trimming the unallocated space on each device
5843 * This will also continue trimming even if a block group or device encounters
5844 * an error. The return value will be the last error, or 0 if nothing bad
5847 int btrfs_trim_fs(struct btrfs_fs_info *fs_info, struct fstrim_range *range)
5849 struct btrfs_block_group *cache = NULL;
5850 struct btrfs_device *device;
5851 struct list_head *devices;
5853 u64 range_end = U64_MAX;
5864 * Check range overflow if range->len is set.
5865 * The default range->len is U64_MAX.
5867 if (range->len != U64_MAX &&
5868 check_add_overflow(range->start, range->len, &range_end))
5871 cache = btrfs_lookup_first_block_group(fs_info, range->start);
5872 for (; cache; cache = btrfs_next_block_group(cache)) {
5873 if (cache->start >= range_end) {
5874 btrfs_put_block_group(cache);
5878 start = max(range->start, cache->start);
5879 end = min(range_end, cache->start + cache->length);
5881 if (end - start >= range->minlen) {
5882 if (!btrfs_block_group_done(cache)) {
5883 ret = btrfs_cache_block_group(cache, 0);
5889 ret = btrfs_wait_block_group_cache_done(cache);
5896 ret = btrfs_trim_block_group(cache,
5902 trimmed += group_trimmed;
5913 "failed to trim %llu block group(s), last error %d",
5915 mutex_lock(&fs_info->fs_devices->device_list_mutex);
5916 devices = &fs_info->fs_devices->devices;
5917 list_for_each_entry(device, devices, dev_list) {
5918 ret = btrfs_trim_free_extents(device, &group_trimmed);
5925 trimmed += group_trimmed;
5927 mutex_unlock(&fs_info->fs_devices->device_list_mutex);
5931 "failed to trim %llu device(s), last error %d",
5932 dev_failed, dev_ret);
5933 range->len = trimmed;