1 // SPDX-License-Identifier: GPL-2.0
3 * Copyright (C) 2007 Oracle. All rights reserved.
7 #include <linux/pagemap.h>
8 #include <linux/time.h>
9 #include <linux/init.h>
10 #include <linux/string.h>
11 #include <linux/backing-dev.h>
12 #include <linux/falloc.h>
13 #include <linux/writeback.h>
14 #include <linux/compat.h>
15 #include <linux/slab.h>
16 #include <linux/btrfs.h>
17 #include <linux/uio.h>
18 #include <linux/iversion.h>
21 #include "transaction.h"
22 #include "btrfs_inode.h"
23 #include "print-tree.h"
28 #include "compression.h"
30 static struct kmem_cache *btrfs_inode_defrag_cachep;
32 * when auto defrag is enabled we
33 * queue up these defrag structs to remember which
34 * inodes need defragging passes
37 struct rb_node rb_node;
41 * transid where the defrag was added, we search for
42 * extents newer than this
49 /* last offset we were able to defrag */
52 /* if we've wrapped around back to zero once already */
56 static int __compare_inode_defrag(struct inode_defrag *defrag1,
57 struct inode_defrag *defrag2)
59 if (defrag1->root > defrag2->root)
61 else if (defrag1->root < defrag2->root)
63 else if (defrag1->ino > defrag2->ino)
65 else if (defrag1->ino < defrag2->ino)
71 /* pop a record for an inode into the defrag tree. The lock
72 * must be held already
74 * If you're inserting a record for an older transid than an
75 * existing record, the transid already in the tree is lowered
77 * If an existing record is found the defrag item you
80 static int __btrfs_add_inode_defrag(struct btrfs_inode *inode,
81 struct inode_defrag *defrag)
83 struct btrfs_fs_info *fs_info = inode->root->fs_info;
84 struct inode_defrag *entry;
86 struct rb_node *parent = NULL;
89 p = &fs_info->defrag_inodes.rb_node;
92 entry = rb_entry(parent, struct inode_defrag, rb_node);
94 ret = __compare_inode_defrag(defrag, entry);
98 p = &parent->rb_right;
100 /* if we're reinserting an entry for
101 * an old defrag run, make sure to
102 * lower the transid of our existing record
104 if (defrag->transid < entry->transid)
105 entry->transid = defrag->transid;
106 if (defrag->last_offset > entry->last_offset)
107 entry->last_offset = defrag->last_offset;
111 set_bit(BTRFS_INODE_IN_DEFRAG, &inode->runtime_flags);
112 rb_link_node(&defrag->rb_node, parent, p);
113 rb_insert_color(&defrag->rb_node, &fs_info->defrag_inodes);
117 static inline int __need_auto_defrag(struct btrfs_fs_info *fs_info)
119 if (!btrfs_test_opt(fs_info, AUTO_DEFRAG))
122 if (btrfs_fs_closing(fs_info))
129 * insert a defrag record for this inode if auto defrag is
132 int btrfs_add_inode_defrag(struct btrfs_trans_handle *trans,
133 struct btrfs_inode *inode)
135 struct btrfs_root *root = inode->root;
136 struct btrfs_fs_info *fs_info = root->fs_info;
137 struct inode_defrag *defrag;
141 if (!__need_auto_defrag(fs_info))
144 if (test_bit(BTRFS_INODE_IN_DEFRAG, &inode->runtime_flags))
148 transid = trans->transid;
150 transid = inode->root->last_trans;
152 defrag = kmem_cache_zalloc(btrfs_inode_defrag_cachep, GFP_NOFS);
156 defrag->ino = btrfs_ino(inode);
157 defrag->transid = transid;
158 defrag->root = root->root_key.objectid;
160 spin_lock(&fs_info->defrag_inodes_lock);
161 if (!test_bit(BTRFS_INODE_IN_DEFRAG, &inode->runtime_flags)) {
163 * If we set IN_DEFRAG flag and evict the inode from memory,
164 * and then re-read this inode, this new inode doesn't have
165 * IN_DEFRAG flag. At the case, we may find the existed defrag.
167 ret = __btrfs_add_inode_defrag(inode, defrag);
169 kmem_cache_free(btrfs_inode_defrag_cachep, defrag);
171 kmem_cache_free(btrfs_inode_defrag_cachep, defrag);
173 spin_unlock(&fs_info->defrag_inodes_lock);
178 * Requeue the defrag object. If there is a defrag object that points to
179 * the same inode in the tree, we will merge them together (by
180 * __btrfs_add_inode_defrag()) and free the one that we want to requeue.
182 static void btrfs_requeue_inode_defrag(struct btrfs_inode *inode,
183 struct inode_defrag *defrag)
185 struct btrfs_fs_info *fs_info = inode->root->fs_info;
188 if (!__need_auto_defrag(fs_info))
192 * Here we don't check the IN_DEFRAG flag, because we need merge
195 spin_lock(&fs_info->defrag_inodes_lock);
196 ret = __btrfs_add_inode_defrag(inode, defrag);
197 spin_unlock(&fs_info->defrag_inodes_lock);
202 kmem_cache_free(btrfs_inode_defrag_cachep, defrag);
206 * pick the defragable inode that we want, if it doesn't exist, we will get
209 static struct inode_defrag *
210 btrfs_pick_defrag_inode(struct btrfs_fs_info *fs_info, u64 root, u64 ino)
212 struct inode_defrag *entry = NULL;
213 struct inode_defrag tmp;
215 struct rb_node *parent = NULL;
221 spin_lock(&fs_info->defrag_inodes_lock);
222 p = fs_info->defrag_inodes.rb_node;
225 entry = rb_entry(parent, struct inode_defrag, rb_node);
227 ret = __compare_inode_defrag(&tmp, entry);
231 p = parent->rb_right;
236 if (parent && __compare_inode_defrag(&tmp, entry) > 0) {
237 parent = rb_next(parent);
239 entry = rb_entry(parent, struct inode_defrag, rb_node);
245 rb_erase(parent, &fs_info->defrag_inodes);
246 spin_unlock(&fs_info->defrag_inodes_lock);
250 void btrfs_cleanup_defrag_inodes(struct btrfs_fs_info *fs_info)
252 struct inode_defrag *defrag;
253 struct rb_node *node;
255 spin_lock(&fs_info->defrag_inodes_lock);
256 node = rb_first(&fs_info->defrag_inodes);
258 rb_erase(node, &fs_info->defrag_inodes);
259 defrag = rb_entry(node, struct inode_defrag, rb_node);
260 kmem_cache_free(btrfs_inode_defrag_cachep, defrag);
262 cond_resched_lock(&fs_info->defrag_inodes_lock);
264 node = rb_first(&fs_info->defrag_inodes);
266 spin_unlock(&fs_info->defrag_inodes_lock);
269 #define BTRFS_DEFRAG_BATCH 1024
271 static int __btrfs_run_defrag_inode(struct btrfs_fs_info *fs_info,
272 struct inode_defrag *defrag)
274 struct btrfs_root *inode_root;
276 struct btrfs_key key;
277 struct btrfs_ioctl_defrag_range_args range;
283 key.objectid = defrag->root;
284 key.type = BTRFS_ROOT_ITEM_KEY;
285 key.offset = (u64)-1;
287 index = srcu_read_lock(&fs_info->subvol_srcu);
289 inode_root = btrfs_read_fs_root_no_name(fs_info, &key);
290 if (IS_ERR(inode_root)) {
291 ret = PTR_ERR(inode_root);
295 key.objectid = defrag->ino;
296 key.type = BTRFS_INODE_ITEM_KEY;
298 inode = btrfs_iget(fs_info->sb, &key, inode_root, NULL);
300 ret = PTR_ERR(inode);
303 srcu_read_unlock(&fs_info->subvol_srcu, index);
305 /* do a chunk of defrag */
306 clear_bit(BTRFS_INODE_IN_DEFRAG, &BTRFS_I(inode)->runtime_flags);
307 memset(&range, 0, sizeof(range));
309 range.start = defrag->last_offset;
311 sb_start_write(fs_info->sb);
312 num_defrag = btrfs_defrag_file(inode, NULL, &range, defrag->transid,
314 sb_end_write(fs_info->sb);
316 * if we filled the whole defrag batch, there
317 * must be more work to do. Queue this defrag
320 if (num_defrag == BTRFS_DEFRAG_BATCH) {
321 defrag->last_offset = range.start;
322 btrfs_requeue_inode_defrag(BTRFS_I(inode), defrag);
323 } else if (defrag->last_offset && !defrag->cycled) {
325 * we didn't fill our defrag batch, but
326 * we didn't start at zero. Make sure we loop
327 * around to the start of the file.
329 defrag->last_offset = 0;
331 btrfs_requeue_inode_defrag(BTRFS_I(inode), defrag);
333 kmem_cache_free(btrfs_inode_defrag_cachep, defrag);
339 srcu_read_unlock(&fs_info->subvol_srcu, index);
340 kmem_cache_free(btrfs_inode_defrag_cachep, defrag);
345 * run through the list of inodes in the FS that need
348 int btrfs_run_defrag_inodes(struct btrfs_fs_info *fs_info)
350 struct inode_defrag *defrag;
352 u64 root_objectid = 0;
354 atomic_inc(&fs_info->defrag_running);
356 /* Pause the auto defragger. */
357 if (test_bit(BTRFS_FS_STATE_REMOUNTING,
361 if (!__need_auto_defrag(fs_info))
364 /* find an inode to defrag */
365 defrag = btrfs_pick_defrag_inode(fs_info, root_objectid,
368 if (root_objectid || first_ino) {
377 first_ino = defrag->ino + 1;
378 root_objectid = defrag->root;
380 __btrfs_run_defrag_inode(fs_info, defrag);
382 atomic_dec(&fs_info->defrag_running);
385 * during unmount, we use the transaction_wait queue to
386 * wait for the defragger to stop
388 wake_up(&fs_info->transaction_wait);
392 /* simple helper to fault in pages and copy. This should go away
393 * and be replaced with calls into generic code.
395 static noinline int btrfs_copy_from_user(loff_t pos, size_t write_bytes,
396 struct page **prepared_pages,
400 size_t total_copied = 0;
402 int offset = pos & (PAGE_SIZE - 1);
404 while (write_bytes > 0) {
405 size_t count = min_t(size_t,
406 PAGE_SIZE - offset, write_bytes);
407 struct page *page = prepared_pages[pg];
409 * Copy data from userspace to the current page
411 copied = iov_iter_copy_from_user_atomic(page, i, offset, count);
413 /* Flush processor's dcache for this page */
414 flush_dcache_page(page);
417 * if we get a partial write, we can end up with
418 * partially up to date pages. These add
419 * a lot of complexity, so make sure they don't
420 * happen by forcing this copy to be retried.
422 * The rest of the btrfs_file_write code will fall
423 * back to page at a time copies after we return 0.
425 if (!PageUptodate(page) && copied < count)
428 iov_iter_advance(i, copied);
429 write_bytes -= copied;
430 total_copied += copied;
432 /* Return to btrfs_file_write_iter to fault page */
433 if (unlikely(copied == 0))
436 if (copied < PAGE_SIZE - offset) {
447 * unlocks pages after btrfs_file_write is done with them
449 static void btrfs_drop_pages(struct page **pages, size_t num_pages)
452 for (i = 0; i < num_pages; i++) {
453 /* page checked is some magic around finding pages that
454 * have been modified without going through btrfs_set_page_dirty
455 * clear it here. There should be no need to mark the pages
456 * accessed as prepare_pages should have marked them accessed
457 * in prepare_pages via find_or_create_page()
459 ClearPageChecked(pages[i]);
460 unlock_page(pages[i]);
465 static int btrfs_find_new_delalloc_bytes(struct btrfs_inode *inode,
468 struct extent_state **cached_state)
470 u64 search_start = start;
471 const u64 end = start + len - 1;
473 while (search_start < end) {
474 const u64 search_len = end - search_start + 1;
475 struct extent_map *em;
479 em = btrfs_get_extent(inode, NULL, 0, search_start,
484 if (em->block_start != EXTENT_MAP_HOLE)
488 if (em->start < search_start)
489 em_len -= search_start - em->start;
490 if (em_len > search_len)
493 ret = set_extent_bit(&inode->io_tree, search_start,
494 search_start + em_len - 1,
496 NULL, cached_state, GFP_NOFS);
498 search_start = extent_map_end(em);
507 * after copy_from_user, pages need to be dirtied and we need to make
508 * sure holes are created between the current EOF and the start of
509 * any next extents (if required).
511 * this also makes the decision about creating an inline extent vs
512 * doing real data extents, marking pages dirty and delalloc as required.
514 int btrfs_dirty_pages(struct inode *inode, struct page **pages,
515 size_t num_pages, loff_t pos, size_t write_bytes,
516 struct extent_state **cached)
518 struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
523 u64 end_of_last_block;
524 u64 end_pos = pos + write_bytes;
525 loff_t isize = i_size_read(inode);
526 unsigned int extra_bits = 0;
528 start_pos = pos & ~((u64) fs_info->sectorsize - 1);
529 num_bytes = round_up(write_bytes + pos - start_pos,
530 fs_info->sectorsize);
532 end_of_last_block = start_pos + num_bytes - 1;
534 if (!btrfs_is_free_space_inode(BTRFS_I(inode))) {
535 if (start_pos >= isize &&
536 !(BTRFS_I(inode)->flags & BTRFS_INODE_PREALLOC)) {
538 * There can't be any extents following eof in this case
539 * so just set the delalloc new bit for the range
542 extra_bits |= EXTENT_DELALLOC_NEW;
544 err = btrfs_find_new_delalloc_bytes(BTRFS_I(inode),
552 err = btrfs_set_extent_delalloc(inode, start_pos, end_of_last_block,
553 extra_bits, cached, 0);
557 for (i = 0; i < num_pages; i++) {
558 struct page *p = pages[i];
565 * we've only changed i_size in ram, and we haven't updated
566 * the disk i_size. There is no need to log the inode
570 i_size_write(inode, end_pos);
575 * this drops all the extents in the cache that intersect the range
576 * [start, end]. Existing extents are split as required.
578 void btrfs_drop_extent_cache(struct btrfs_inode *inode, u64 start, u64 end,
581 struct extent_map *em;
582 struct extent_map *split = NULL;
583 struct extent_map *split2 = NULL;
584 struct extent_map_tree *em_tree = &inode->extent_tree;
585 u64 len = end - start + 1;
593 WARN_ON(end < start);
594 if (end == (u64)-1) {
603 split = alloc_extent_map();
605 split2 = alloc_extent_map();
606 if (!split || !split2)
609 write_lock(&em_tree->lock);
610 em = lookup_extent_mapping(em_tree, start, len);
612 write_unlock(&em_tree->lock);
616 gen = em->generation;
617 if (skip_pinned && test_bit(EXTENT_FLAG_PINNED, &em->flags)) {
618 if (testend && em->start + em->len >= start + len) {
620 write_unlock(&em_tree->lock);
623 start = em->start + em->len;
625 len = start + len - (em->start + em->len);
627 write_unlock(&em_tree->lock);
630 compressed = test_bit(EXTENT_FLAG_COMPRESSED, &em->flags);
631 clear_bit(EXTENT_FLAG_PINNED, &em->flags);
632 clear_bit(EXTENT_FLAG_LOGGING, &flags);
633 modified = !list_empty(&em->list);
637 if (em->start < start) {
638 split->start = em->start;
639 split->len = start - em->start;
641 if (em->block_start < EXTENT_MAP_LAST_BYTE) {
642 split->orig_start = em->orig_start;
643 split->block_start = em->block_start;
646 split->block_len = em->block_len;
648 split->block_len = split->len;
649 split->orig_block_len = max(split->block_len,
651 split->ram_bytes = em->ram_bytes;
653 split->orig_start = split->start;
654 split->block_len = 0;
655 split->block_start = em->block_start;
656 split->orig_block_len = 0;
657 split->ram_bytes = split->len;
660 split->generation = gen;
661 split->bdev = em->bdev;
662 split->flags = flags;
663 split->compress_type = em->compress_type;
664 replace_extent_mapping(em_tree, em, split, modified);
665 free_extent_map(split);
669 if (testend && em->start + em->len > start + len) {
670 u64 diff = start + len - em->start;
672 split->start = start + len;
673 split->len = em->start + em->len - (start + len);
674 split->bdev = em->bdev;
675 split->flags = flags;
676 split->compress_type = em->compress_type;
677 split->generation = gen;
679 if (em->block_start < EXTENT_MAP_LAST_BYTE) {
680 split->orig_block_len = max(em->block_len,
683 split->ram_bytes = em->ram_bytes;
685 split->block_len = em->block_len;
686 split->block_start = em->block_start;
687 split->orig_start = em->orig_start;
689 split->block_len = split->len;
690 split->block_start = em->block_start
692 split->orig_start = em->orig_start;
695 split->ram_bytes = split->len;
696 split->orig_start = split->start;
697 split->block_len = 0;
698 split->block_start = em->block_start;
699 split->orig_block_len = 0;
702 if (extent_map_in_tree(em)) {
703 replace_extent_mapping(em_tree, em, split,
706 ret = add_extent_mapping(em_tree, split,
708 ASSERT(ret == 0); /* Logic error */
710 free_extent_map(split);
714 if (extent_map_in_tree(em))
715 remove_extent_mapping(em_tree, em);
716 write_unlock(&em_tree->lock);
720 /* once for the tree*/
724 free_extent_map(split);
726 free_extent_map(split2);
730 * this is very complex, but the basic idea is to drop all extents
731 * in the range start - end. hint_block is filled in with a block number
732 * that would be a good hint to the block allocator for this file.
734 * If an extent intersects the range but is not entirely inside the range
735 * it is either truncated or split. Anything entirely inside the range
736 * is deleted from the tree.
738 int __btrfs_drop_extents(struct btrfs_trans_handle *trans,
739 struct btrfs_root *root, struct inode *inode,
740 struct btrfs_path *path, u64 start, u64 end,
741 u64 *drop_end, int drop_cache,
743 u32 extent_item_size,
746 struct btrfs_fs_info *fs_info = root->fs_info;
747 struct extent_buffer *leaf;
748 struct btrfs_file_extent_item *fi;
749 struct btrfs_key key;
750 struct btrfs_key new_key;
751 u64 ino = btrfs_ino(BTRFS_I(inode));
752 u64 search_start = start;
755 u64 extent_offset = 0;
757 u64 last_end = start;
763 int modify_tree = -1;
766 int leafs_visited = 0;
769 btrfs_drop_extent_cache(BTRFS_I(inode), start, end - 1, 0);
771 if (start >= BTRFS_I(inode)->disk_i_size && !replace_extent)
774 update_refs = (test_bit(BTRFS_ROOT_REF_COWS, &root->state) ||
775 root == fs_info->tree_root);
778 ret = btrfs_lookup_file_extent(trans, root, path, ino,
779 search_start, modify_tree);
782 if (ret > 0 && path->slots[0] > 0 && search_start == start) {
783 leaf = path->nodes[0];
784 btrfs_item_key_to_cpu(leaf, &key, path->slots[0] - 1);
785 if (key.objectid == ino &&
786 key.type == BTRFS_EXTENT_DATA_KEY)
792 leaf = path->nodes[0];
793 if (path->slots[0] >= btrfs_header_nritems(leaf)) {
795 ret = btrfs_next_leaf(root, path);
803 leaf = path->nodes[0];
807 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
809 if (key.objectid > ino)
811 if (WARN_ON_ONCE(key.objectid < ino) ||
812 key.type < BTRFS_EXTENT_DATA_KEY) {
817 if (key.type > BTRFS_EXTENT_DATA_KEY || key.offset >= end)
820 fi = btrfs_item_ptr(leaf, path->slots[0],
821 struct btrfs_file_extent_item);
822 extent_type = btrfs_file_extent_type(leaf, fi);
824 if (extent_type == BTRFS_FILE_EXTENT_REG ||
825 extent_type == BTRFS_FILE_EXTENT_PREALLOC) {
826 disk_bytenr = btrfs_file_extent_disk_bytenr(leaf, fi);
827 num_bytes = btrfs_file_extent_disk_num_bytes(leaf, fi);
828 extent_offset = btrfs_file_extent_offset(leaf, fi);
829 extent_end = key.offset +
830 btrfs_file_extent_num_bytes(leaf, fi);
831 } else if (extent_type == BTRFS_FILE_EXTENT_INLINE) {
832 extent_end = key.offset +
833 btrfs_file_extent_ram_bytes(leaf, fi);
840 * Don't skip extent items representing 0 byte lengths. They
841 * used to be created (bug) if while punching holes we hit
842 * -ENOSPC condition. So if we find one here, just ensure we
843 * delete it, otherwise we would insert a new file extent item
844 * with the same key (offset) as that 0 bytes length file
845 * extent item in the call to setup_items_for_insert() later
848 if (extent_end == key.offset && extent_end >= search_start) {
849 last_end = extent_end;
850 goto delete_extent_item;
853 if (extent_end <= search_start) {
859 search_start = max(key.offset, start);
860 if (recow || !modify_tree) {
862 btrfs_release_path(path);
867 * | - range to drop - |
868 * | -------- extent -------- |
870 if (start > key.offset && end < extent_end) {
872 if (extent_type == BTRFS_FILE_EXTENT_INLINE) {
877 memcpy(&new_key, &key, sizeof(new_key));
878 new_key.offset = start;
879 ret = btrfs_duplicate_item(trans, root, path,
881 if (ret == -EAGAIN) {
882 btrfs_release_path(path);
888 leaf = path->nodes[0];
889 fi = btrfs_item_ptr(leaf, path->slots[0] - 1,
890 struct btrfs_file_extent_item);
891 btrfs_set_file_extent_num_bytes(leaf, fi,
894 fi = btrfs_item_ptr(leaf, path->slots[0],
895 struct btrfs_file_extent_item);
897 extent_offset += start - key.offset;
898 btrfs_set_file_extent_offset(leaf, fi, extent_offset);
899 btrfs_set_file_extent_num_bytes(leaf, fi,
901 btrfs_mark_buffer_dirty(leaf);
903 if (update_refs && disk_bytenr > 0) {
904 ret = btrfs_inc_extent_ref(trans, root,
905 disk_bytenr, num_bytes, 0,
906 root->root_key.objectid,
908 start - extent_offset);
909 BUG_ON(ret); /* -ENOMEM */
914 * From here on out we will have actually dropped something, so
915 * last_end can be updated.
917 last_end = extent_end;
920 * | ---- range to drop ----- |
921 * | -------- extent -------- |
923 if (start <= key.offset && end < extent_end) {
924 if (extent_type == BTRFS_FILE_EXTENT_INLINE) {
929 memcpy(&new_key, &key, sizeof(new_key));
930 new_key.offset = end;
931 btrfs_set_item_key_safe(fs_info, path, &new_key);
933 extent_offset += end - key.offset;
934 btrfs_set_file_extent_offset(leaf, fi, extent_offset);
935 btrfs_set_file_extent_num_bytes(leaf, fi,
937 btrfs_mark_buffer_dirty(leaf);
938 if (update_refs && disk_bytenr > 0)
939 inode_sub_bytes(inode, end - key.offset);
943 search_start = extent_end;
945 * | ---- range to drop ----- |
946 * | -------- extent -------- |
948 if (start > key.offset && end >= extent_end) {
950 if (extent_type == BTRFS_FILE_EXTENT_INLINE) {
955 btrfs_set_file_extent_num_bytes(leaf, fi,
957 btrfs_mark_buffer_dirty(leaf);
958 if (update_refs && disk_bytenr > 0)
959 inode_sub_bytes(inode, extent_end - start);
960 if (end == extent_end)
968 * | ---- range to drop ----- |
969 * | ------ extent ------ |
971 if (start <= key.offset && end >= extent_end) {
974 del_slot = path->slots[0];
977 BUG_ON(del_slot + del_nr != path->slots[0]);
982 extent_type == BTRFS_FILE_EXTENT_INLINE) {
983 inode_sub_bytes(inode,
984 extent_end - key.offset);
985 extent_end = ALIGN(extent_end,
986 fs_info->sectorsize);
987 } else if (update_refs && disk_bytenr > 0) {
988 ret = btrfs_free_extent(trans, root,
989 disk_bytenr, num_bytes, 0,
990 root->root_key.objectid,
991 key.objectid, key.offset -
993 BUG_ON(ret); /* -ENOMEM */
994 inode_sub_bytes(inode,
995 extent_end - key.offset);
998 if (end == extent_end)
1001 if (path->slots[0] + 1 < btrfs_header_nritems(leaf)) {
1006 ret = btrfs_del_items(trans, root, path, del_slot,
1009 btrfs_abort_transaction(trans, ret);
1016 btrfs_release_path(path);
1023 if (!ret && del_nr > 0) {
1025 * Set path->slots[0] to first slot, so that after the delete
1026 * if items are move off from our leaf to its immediate left or
1027 * right neighbor leafs, we end up with a correct and adjusted
1028 * path->slots[0] for our insertion (if replace_extent != 0).
1030 path->slots[0] = del_slot;
1031 ret = btrfs_del_items(trans, root, path, del_slot, del_nr);
1033 btrfs_abort_transaction(trans, ret);
1036 leaf = path->nodes[0];
1038 * If btrfs_del_items() was called, it might have deleted a leaf, in
1039 * which case it unlocked our path, so check path->locks[0] matches a
1042 if (!ret && replace_extent && leafs_visited == 1 &&
1043 (path->locks[0] == BTRFS_WRITE_LOCK_BLOCKING ||
1044 path->locks[0] == BTRFS_WRITE_LOCK) &&
1045 btrfs_leaf_free_space(fs_info, leaf) >=
1046 sizeof(struct btrfs_item) + extent_item_size) {
1049 key.type = BTRFS_EXTENT_DATA_KEY;
1051 if (!del_nr && path->slots[0] < btrfs_header_nritems(leaf)) {
1052 struct btrfs_key slot_key;
1054 btrfs_item_key_to_cpu(leaf, &slot_key, path->slots[0]);
1055 if (btrfs_comp_cpu_keys(&key, &slot_key) > 0)
1058 setup_items_for_insert(root, path, &key,
1061 sizeof(struct btrfs_item) +
1062 extent_item_size, 1);
1066 if (!replace_extent || !(*key_inserted))
1067 btrfs_release_path(path);
1069 *drop_end = found ? min(end, last_end) : end;
1073 int btrfs_drop_extents(struct btrfs_trans_handle *trans,
1074 struct btrfs_root *root, struct inode *inode, u64 start,
1075 u64 end, int drop_cache)
1077 struct btrfs_path *path;
1080 path = btrfs_alloc_path();
1083 ret = __btrfs_drop_extents(trans, root, inode, path, start, end, NULL,
1084 drop_cache, 0, 0, NULL);
1085 btrfs_free_path(path);
1089 static int extent_mergeable(struct extent_buffer *leaf, int slot,
1090 u64 objectid, u64 bytenr, u64 orig_offset,
1091 u64 *start, u64 *end)
1093 struct btrfs_file_extent_item *fi;
1094 struct btrfs_key key;
1097 if (slot < 0 || slot >= btrfs_header_nritems(leaf))
1100 btrfs_item_key_to_cpu(leaf, &key, slot);
1101 if (key.objectid != objectid || key.type != BTRFS_EXTENT_DATA_KEY)
1104 fi = btrfs_item_ptr(leaf, slot, struct btrfs_file_extent_item);
1105 if (btrfs_file_extent_type(leaf, fi) != BTRFS_FILE_EXTENT_REG ||
1106 btrfs_file_extent_disk_bytenr(leaf, fi) != bytenr ||
1107 btrfs_file_extent_offset(leaf, fi) != key.offset - orig_offset ||
1108 btrfs_file_extent_compression(leaf, fi) ||
1109 btrfs_file_extent_encryption(leaf, fi) ||
1110 btrfs_file_extent_other_encoding(leaf, fi))
1113 extent_end = key.offset + btrfs_file_extent_num_bytes(leaf, fi);
1114 if ((*start && *start != key.offset) || (*end && *end != extent_end))
1117 *start = key.offset;
1123 * Mark extent in the range start - end as written.
1125 * This changes extent type from 'pre-allocated' to 'regular'. If only
1126 * part of extent is marked as written, the extent will be split into
1129 int btrfs_mark_extent_written(struct btrfs_trans_handle *trans,
1130 struct btrfs_inode *inode, u64 start, u64 end)
1132 struct btrfs_fs_info *fs_info = trans->fs_info;
1133 struct btrfs_root *root = inode->root;
1134 struct extent_buffer *leaf;
1135 struct btrfs_path *path;
1136 struct btrfs_file_extent_item *fi;
1137 struct btrfs_key key;
1138 struct btrfs_key new_key;
1150 u64 ino = btrfs_ino(inode);
1152 path = btrfs_alloc_path();
1159 key.type = BTRFS_EXTENT_DATA_KEY;
1162 ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
1165 if (ret > 0 && path->slots[0] > 0)
1168 leaf = path->nodes[0];
1169 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
1170 if (key.objectid != ino ||
1171 key.type != BTRFS_EXTENT_DATA_KEY) {
1173 btrfs_abort_transaction(trans, ret);
1176 fi = btrfs_item_ptr(leaf, path->slots[0],
1177 struct btrfs_file_extent_item);
1178 if (btrfs_file_extent_type(leaf, fi) != BTRFS_FILE_EXTENT_PREALLOC) {
1180 btrfs_abort_transaction(trans, ret);
1183 extent_end = key.offset + btrfs_file_extent_num_bytes(leaf, fi);
1184 if (key.offset > start || extent_end < end) {
1186 btrfs_abort_transaction(trans, ret);
1190 bytenr = btrfs_file_extent_disk_bytenr(leaf, fi);
1191 num_bytes = btrfs_file_extent_disk_num_bytes(leaf, fi);
1192 orig_offset = key.offset - btrfs_file_extent_offset(leaf, fi);
1193 memcpy(&new_key, &key, sizeof(new_key));
1195 if (start == key.offset && end < extent_end) {
1198 if (extent_mergeable(leaf, path->slots[0] - 1,
1199 ino, bytenr, orig_offset,
1200 &other_start, &other_end)) {
1201 new_key.offset = end;
1202 btrfs_set_item_key_safe(fs_info, path, &new_key);
1203 fi = btrfs_item_ptr(leaf, path->slots[0],
1204 struct btrfs_file_extent_item);
1205 btrfs_set_file_extent_generation(leaf, fi,
1207 btrfs_set_file_extent_num_bytes(leaf, fi,
1209 btrfs_set_file_extent_offset(leaf, fi,
1211 fi = btrfs_item_ptr(leaf, path->slots[0] - 1,
1212 struct btrfs_file_extent_item);
1213 btrfs_set_file_extent_generation(leaf, fi,
1215 btrfs_set_file_extent_num_bytes(leaf, fi,
1217 btrfs_mark_buffer_dirty(leaf);
1222 if (start > key.offset && end == extent_end) {
1225 if (extent_mergeable(leaf, path->slots[0] + 1,
1226 ino, bytenr, orig_offset,
1227 &other_start, &other_end)) {
1228 fi = btrfs_item_ptr(leaf, path->slots[0],
1229 struct btrfs_file_extent_item);
1230 btrfs_set_file_extent_num_bytes(leaf, fi,
1231 start - key.offset);
1232 btrfs_set_file_extent_generation(leaf, fi,
1235 new_key.offset = start;
1236 btrfs_set_item_key_safe(fs_info, path, &new_key);
1238 fi = btrfs_item_ptr(leaf, path->slots[0],
1239 struct btrfs_file_extent_item);
1240 btrfs_set_file_extent_generation(leaf, fi,
1242 btrfs_set_file_extent_num_bytes(leaf, fi,
1244 btrfs_set_file_extent_offset(leaf, fi,
1245 start - orig_offset);
1246 btrfs_mark_buffer_dirty(leaf);
1251 while (start > key.offset || end < extent_end) {
1252 if (key.offset == start)
1255 new_key.offset = split;
1256 ret = btrfs_duplicate_item(trans, root, path, &new_key);
1257 if (ret == -EAGAIN) {
1258 btrfs_release_path(path);
1262 btrfs_abort_transaction(trans, ret);
1266 leaf = path->nodes[0];
1267 fi = btrfs_item_ptr(leaf, path->slots[0] - 1,
1268 struct btrfs_file_extent_item);
1269 btrfs_set_file_extent_generation(leaf, fi, trans->transid);
1270 btrfs_set_file_extent_num_bytes(leaf, fi,
1271 split - key.offset);
1273 fi = btrfs_item_ptr(leaf, path->slots[0],
1274 struct btrfs_file_extent_item);
1276 btrfs_set_file_extent_generation(leaf, fi, trans->transid);
1277 btrfs_set_file_extent_offset(leaf, fi, split - orig_offset);
1278 btrfs_set_file_extent_num_bytes(leaf, fi,
1279 extent_end - split);
1280 btrfs_mark_buffer_dirty(leaf);
1282 ret = btrfs_inc_extent_ref(trans, root, bytenr, num_bytes,
1283 0, root->root_key.objectid,
1286 btrfs_abort_transaction(trans, ret);
1290 if (split == start) {
1293 if (start != key.offset) {
1295 btrfs_abort_transaction(trans, ret);
1306 if (extent_mergeable(leaf, path->slots[0] + 1,
1307 ino, bytenr, orig_offset,
1308 &other_start, &other_end)) {
1310 btrfs_release_path(path);
1313 extent_end = other_end;
1314 del_slot = path->slots[0] + 1;
1316 ret = btrfs_free_extent(trans, root, bytenr, num_bytes,
1317 0, root->root_key.objectid,
1320 btrfs_abort_transaction(trans, ret);
1326 if (extent_mergeable(leaf, path->slots[0] - 1,
1327 ino, bytenr, orig_offset,
1328 &other_start, &other_end)) {
1330 btrfs_release_path(path);
1333 key.offset = other_start;
1334 del_slot = path->slots[0];
1336 ret = btrfs_free_extent(trans, root, bytenr, num_bytes,
1337 0, root->root_key.objectid,
1340 btrfs_abort_transaction(trans, ret);
1345 fi = btrfs_item_ptr(leaf, path->slots[0],
1346 struct btrfs_file_extent_item);
1347 btrfs_set_file_extent_type(leaf, fi,
1348 BTRFS_FILE_EXTENT_REG);
1349 btrfs_set_file_extent_generation(leaf, fi, trans->transid);
1350 btrfs_mark_buffer_dirty(leaf);
1352 fi = btrfs_item_ptr(leaf, del_slot - 1,
1353 struct btrfs_file_extent_item);
1354 btrfs_set_file_extent_type(leaf, fi,
1355 BTRFS_FILE_EXTENT_REG);
1356 btrfs_set_file_extent_generation(leaf, fi, trans->transid);
1357 btrfs_set_file_extent_num_bytes(leaf, fi,
1358 extent_end - key.offset);
1359 btrfs_mark_buffer_dirty(leaf);
1361 ret = btrfs_del_items(trans, root, path, del_slot, del_nr);
1363 btrfs_abort_transaction(trans, ret);
1368 btrfs_free_path(path);
1373 * on error we return an unlocked page and the error value
1374 * on success we return a locked page and 0
1376 static int prepare_uptodate_page(struct inode *inode,
1377 struct page *page, u64 pos,
1378 bool force_uptodate)
1382 if (((pos & (PAGE_SIZE - 1)) || force_uptodate) &&
1383 !PageUptodate(page)) {
1384 ret = btrfs_readpage(NULL, page);
1388 if (!PageUptodate(page)) {
1392 if (page->mapping != inode->i_mapping) {
1401 * this just gets pages into the page cache and locks them down.
1403 static noinline int prepare_pages(struct inode *inode, struct page **pages,
1404 size_t num_pages, loff_t pos,
1405 size_t write_bytes, bool force_uptodate)
1408 unsigned long index = pos >> PAGE_SHIFT;
1409 gfp_t mask = btrfs_alloc_write_mask(inode->i_mapping);
1413 for (i = 0; i < num_pages; i++) {
1415 pages[i] = find_or_create_page(inode->i_mapping, index + i,
1416 mask | __GFP_WRITE);
1424 err = prepare_uptodate_page(inode, pages[i], pos,
1426 if (!err && i == num_pages - 1)
1427 err = prepare_uptodate_page(inode, pages[i],
1428 pos + write_bytes, false);
1431 if (err == -EAGAIN) {
1438 wait_on_page_writeback(pages[i]);
1443 while (faili >= 0) {
1444 unlock_page(pages[faili]);
1445 put_page(pages[faili]);
1453 * This function locks the extent and properly waits for data=ordered extents
1454 * to finish before allowing the pages to be modified if need.
1457 * 1 - the extent is locked
1458 * 0 - the extent is not locked, and everything is OK
1459 * -EAGAIN - need re-prepare the pages
1460 * the other < 0 number - Something wrong happens
1463 lock_and_cleanup_extent_if_need(struct btrfs_inode *inode, struct page **pages,
1464 size_t num_pages, loff_t pos,
1466 u64 *lockstart, u64 *lockend,
1467 struct extent_state **cached_state)
1469 struct btrfs_fs_info *fs_info = inode->root->fs_info;
1475 start_pos = round_down(pos, fs_info->sectorsize);
1476 last_pos = start_pos
1477 + round_up(pos + write_bytes - start_pos,
1478 fs_info->sectorsize) - 1;
1480 if (start_pos < inode->vfs_inode.i_size) {
1481 struct btrfs_ordered_extent *ordered;
1483 lock_extent_bits(&inode->io_tree, start_pos, last_pos,
1485 ordered = btrfs_lookup_ordered_range(inode, start_pos,
1486 last_pos - start_pos + 1);
1488 ordered->file_offset + ordered->len > start_pos &&
1489 ordered->file_offset <= last_pos) {
1490 unlock_extent_cached(&inode->io_tree, start_pos,
1491 last_pos, cached_state);
1492 for (i = 0; i < num_pages; i++) {
1493 unlock_page(pages[i]);
1496 btrfs_start_ordered_extent(&inode->vfs_inode,
1498 btrfs_put_ordered_extent(ordered);
1502 btrfs_put_ordered_extent(ordered);
1503 clear_extent_bit(&inode->io_tree, start_pos, last_pos,
1504 EXTENT_DIRTY | EXTENT_DELALLOC |
1505 EXTENT_DO_ACCOUNTING | EXTENT_DEFRAG,
1506 0, 0, cached_state);
1507 *lockstart = start_pos;
1508 *lockend = last_pos;
1512 for (i = 0; i < num_pages; i++) {
1513 if (clear_page_dirty_for_io(pages[i]))
1514 account_page_redirty(pages[i]);
1515 set_page_extent_mapped(pages[i]);
1516 WARN_ON(!PageLocked(pages[i]));
1522 static noinline int check_can_nocow(struct btrfs_inode *inode, loff_t pos,
1523 size_t *write_bytes)
1525 struct btrfs_fs_info *fs_info = inode->root->fs_info;
1526 struct btrfs_root *root = inode->root;
1527 struct btrfs_ordered_extent *ordered;
1528 u64 lockstart, lockend;
1532 ret = btrfs_start_write_no_snapshotting(root);
1536 lockstart = round_down(pos, fs_info->sectorsize);
1537 lockend = round_up(pos + *write_bytes,
1538 fs_info->sectorsize) - 1;
1541 lock_extent(&inode->io_tree, lockstart, lockend);
1542 ordered = btrfs_lookup_ordered_range(inode, lockstart,
1543 lockend - lockstart + 1);
1547 unlock_extent(&inode->io_tree, lockstart, lockend);
1548 btrfs_start_ordered_extent(&inode->vfs_inode, ordered, 1);
1549 btrfs_put_ordered_extent(ordered);
1552 num_bytes = lockend - lockstart + 1;
1553 ret = can_nocow_extent(&inode->vfs_inode, lockstart, &num_bytes,
1557 btrfs_end_write_no_snapshotting(root);
1559 *write_bytes = min_t(size_t, *write_bytes ,
1560 num_bytes - pos + lockstart);
1563 unlock_extent(&inode->io_tree, lockstart, lockend);
1568 static noinline ssize_t btrfs_buffered_write(struct kiocb *iocb,
1571 struct file *file = iocb->ki_filp;
1572 loff_t pos = iocb->ki_pos;
1573 struct inode *inode = file_inode(file);
1574 struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
1575 struct btrfs_root *root = BTRFS_I(inode)->root;
1576 struct page **pages = NULL;
1577 struct extent_state *cached_state = NULL;
1578 struct extent_changeset *data_reserved = NULL;
1579 u64 release_bytes = 0;
1582 size_t num_written = 0;
1585 bool only_release_metadata = false;
1586 bool force_page_uptodate = false;
1588 nrptrs = min(DIV_ROUND_UP(iov_iter_count(i), PAGE_SIZE),
1589 PAGE_SIZE / (sizeof(struct page *)));
1590 nrptrs = min(nrptrs, current->nr_dirtied_pause - current->nr_dirtied);
1591 nrptrs = max(nrptrs, 8);
1592 pages = kmalloc_array(nrptrs, sizeof(struct page *), GFP_KERNEL);
1596 while (iov_iter_count(i) > 0) {
1597 size_t offset = pos & (PAGE_SIZE - 1);
1598 size_t sector_offset;
1599 size_t write_bytes = min(iov_iter_count(i),
1600 nrptrs * (size_t)PAGE_SIZE -
1602 size_t num_pages = DIV_ROUND_UP(write_bytes + offset,
1604 size_t reserve_bytes;
1607 size_t dirty_sectors;
1611 WARN_ON(num_pages > nrptrs);
1614 * Fault pages before locking them in prepare_pages
1615 * to avoid recursive lock
1617 if (unlikely(iov_iter_fault_in_readable(i, write_bytes))) {
1622 sector_offset = pos & (fs_info->sectorsize - 1);
1623 reserve_bytes = round_up(write_bytes + sector_offset,
1624 fs_info->sectorsize);
1626 extent_changeset_release(data_reserved);
1627 ret = btrfs_check_data_free_space(inode, &data_reserved, pos,
1630 if ((BTRFS_I(inode)->flags & (BTRFS_INODE_NODATACOW |
1631 BTRFS_INODE_PREALLOC)) &&
1632 check_can_nocow(BTRFS_I(inode), pos,
1633 &write_bytes) > 0) {
1635 * For nodata cow case, no need to reserve
1638 only_release_metadata = true;
1640 * our prealloc extent may be smaller than
1641 * write_bytes, so scale down.
1643 num_pages = DIV_ROUND_UP(write_bytes + offset,
1645 reserve_bytes = round_up(write_bytes +
1647 fs_info->sectorsize);
1653 WARN_ON(reserve_bytes == 0);
1654 ret = btrfs_delalloc_reserve_metadata(BTRFS_I(inode),
1657 if (!only_release_metadata)
1658 btrfs_free_reserved_data_space(inode,
1662 btrfs_end_write_no_snapshotting(root);
1666 release_bytes = reserve_bytes;
1669 * This is going to setup the pages array with the number of
1670 * pages we want, so we don't really need to worry about the
1671 * contents of pages from loop to loop
1673 ret = prepare_pages(inode, pages, num_pages,
1675 force_page_uptodate);
1677 btrfs_delalloc_release_extents(BTRFS_I(inode),
1678 reserve_bytes, true);
1682 extents_locked = lock_and_cleanup_extent_if_need(
1683 BTRFS_I(inode), pages,
1684 num_pages, pos, write_bytes, &lockstart,
1685 &lockend, &cached_state);
1686 if (extents_locked < 0) {
1687 if (extents_locked == -EAGAIN)
1689 btrfs_delalloc_release_extents(BTRFS_I(inode),
1690 reserve_bytes, true);
1691 ret = extents_locked;
1695 copied = btrfs_copy_from_user(pos, write_bytes, pages, i);
1697 num_sectors = BTRFS_BYTES_TO_BLKS(fs_info, reserve_bytes);
1698 dirty_sectors = round_up(copied + sector_offset,
1699 fs_info->sectorsize);
1700 dirty_sectors = BTRFS_BYTES_TO_BLKS(fs_info, dirty_sectors);
1703 * if we have trouble faulting in the pages, fall
1704 * back to one page at a time
1706 if (copied < write_bytes)
1710 force_page_uptodate = true;
1714 force_page_uptodate = false;
1715 dirty_pages = DIV_ROUND_UP(copied + offset,
1719 if (num_sectors > dirty_sectors) {
1720 /* release everything except the sectors we dirtied */
1721 release_bytes -= dirty_sectors <<
1722 fs_info->sb->s_blocksize_bits;
1723 if (only_release_metadata) {
1724 btrfs_delalloc_release_metadata(BTRFS_I(inode),
1725 release_bytes, true);
1729 __pos = round_down(pos,
1730 fs_info->sectorsize) +
1731 (dirty_pages << PAGE_SHIFT);
1732 btrfs_delalloc_release_space(inode,
1733 data_reserved, __pos,
1734 release_bytes, true);
1738 release_bytes = round_up(copied + sector_offset,
1739 fs_info->sectorsize);
1742 ret = btrfs_dirty_pages(inode, pages, dirty_pages,
1743 pos, copied, &cached_state);
1745 unlock_extent_cached(&BTRFS_I(inode)->io_tree,
1746 lockstart, lockend, &cached_state);
1747 btrfs_delalloc_release_extents(BTRFS_I(inode), reserve_bytes,
1750 btrfs_drop_pages(pages, num_pages);
1755 if (only_release_metadata)
1756 btrfs_end_write_no_snapshotting(root);
1758 if (only_release_metadata && copied > 0) {
1759 lockstart = round_down(pos,
1760 fs_info->sectorsize);
1761 lockend = round_up(pos + copied,
1762 fs_info->sectorsize) - 1;
1764 set_extent_bit(&BTRFS_I(inode)->io_tree, lockstart,
1765 lockend, EXTENT_NORESERVE, NULL,
1767 only_release_metadata = false;
1770 btrfs_drop_pages(pages, num_pages);
1774 balance_dirty_pages_ratelimited(inode->i_mapping);
1775 if (dirty_pages < (fs_info->nodesize >> PAGE_SHIFT) + 1)
1776 btrfs_btree_balance_dirty(fs_info);
1779 num_written += copied;
1784 if (release_bytes) {
1785 if (only_release_metadata) {
1786 btrfs_end_write_no_snapshotting(root);
1787 btrfs_delalloc_release_metadata(BTRFS_I(inode),
1788 release_bytes, true);
1790 btrfs_delalloc_release_space(inode, data_reserved,
1791 round_down(pos, fs_info->sectorsize),
1792 release_bytes, true);
1796 extent_changeset_free(data_reserved);
1797 return num_written ? num_written : ret;
1800 static ssize_t __btrfs_direct_write(struct kiocb *iocb, struct iov_iter *from)
1802 struct file *file = iocb->ki_filp;
1803 struct inode *inode = file_inode(file);
1806 ssize_t written_buffered;
1810 written = generic_file_direct_write(iocb, from);
1812 if (written < 0 || !iov_iter_count(from))
1816 written_buffered = btrfs_buffered_write(iocb, from);
1817 if (written_buffered < 0) {
1818 err = written_buffered;
1822 * Ensure all data is persisted. We want the next direct IO read to be
1823 * able to read what was just written.
1825 endbyte = pos + written_buffered - 1;
1826 err = btrfs_fdatawrite_range(inode, pos, endbyte);
1829 err = filemap_fdatawait_range(inode->i_mapping, pos, endbyte);
1832 written += written_buffered;
1833 iocb->ki_pos = pos + written_buffered;
1834 invalidate_mapping_pages(file->f_mapping, pos >> PAGE_SHIFT,
1835 endbyte >> PAGE_SHIFT);
1837 return written ? written : err;
1840 static void update_time_for_write(struct inode *inode)
1842 struct timespec64 now;
1844 if (IS_NOCMTIME(inode))
1847 now = current_time(inode);
1848 if (!timespec64_equal(&inode->i_mtime, &now))
1849 inode->i_mtime = now;
1851 if (!timespec64_equal(&inode->i_ctime, &now))
1852 inode->i_ctime = now;
1854 if (IS_I_VERSION(inode))
1855 inode_inc_iversion(inode);
1858 static ssize_t btrfs_file_write_iter(struct kiocb *iocb,
1859 struct iov_iter *from)
1861 struct file *file = iocb->ki_filp;
1862 struct inode *inode = file_inode(file);
1863 struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
1864 struct btrfs_root *root = BTRFS_I(inode)->root;
1867 ssize_t num_written = 0;
1868 bool sync = (file->f_flags & O_DSYNC) || IS_SYNC(file->f_mapping->host);
1871 size_t count = iov_iter_count(from);
1875 if (!(iocb->ki_flags & IOCB_DIRECT) &&
1876 (iocb->ki_flags & IOCB_NOWAIT))
1879 if (!inode_trylock(inode)) {
1880 if (iocb->ki_flags & IOCB_NOWAIT)
1885 err = generic_write_checks(iocb, from);
1887 inode_unlock(inode);
1892 if (iocb->ki_flags & IOCB_NOWAIT) {
1894 * We will allocate space in case nodatacow is not set,
1897 if (!(BTRFS_I(inode)->flags & (BTRFS_INODE_NODATACOW |
1898 BTRFS_INODE_PREALLOC)) ||
1899 check_can_nocow(BTRFS_I(inode), pos, &count) <= 0) {
1900 inode_unlock(inode);
1905 current->backing_dev_info = inode_to_bdi(inode);
1906 err = file_remove_privs(file);
1908 inode_unlock(inode);
1913 * If BTRFS flips readonly due to some impossible error
1914 * (fs_info->fs_state now has BTRFS_SUPER_FLAG_ERROR),
1915 * although we have opened a file as writable, we have
1916 * to stop this write operation to ensure FS consistency.
1918 if (test_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state)) {
1919 inode_unlock(inode);
1925 * We reserve space for updating the inode when we reserve space for the
1926 * extent we are going to write, so we will enospc out there. We don't
1927 * need to start yet another transaction to update the inode as we will
1928 * update the inode when we finish writing whatever data we write.
1930 update_time_for_write(inode);
1932 start_pos = round_down(pos, fs_info->sectorsize);
1933 oldsize = i_size_read(inode);
1934 if (start_pos > oldsize) {
1935 /* Expand hole size to cover write data, preventing empty gap */
1936 end_pos = round_up(pos + count,
1937 fs_info->sectorsize);
1938 err = btrfs_cont_expand(inode, oldsize, end_pos);
1940 inode_unlock(inode);
1943 if (start_pos > round_up(oldsize, fs_info->sectorsize))
1948 atomic_inc(&BTRFS_I(inode)->sync_writers);
1950 if (iocb->ki_flags & IOCB_DIRECT) {
1951 num_written = __btrfs_direct_write(iocb, from);
1953 num_written = btrfs_buffered_write(iocb, from);
1954 if (num_written > 0)
1955 iocb->ki_pos = pos + num_written;
1957 pagecache_isize_extended(inode, oldsize,
1958 i_size_read(inode));
1961 inode_unlock(inode);
1964 * We also have to set last_sub_trans to the current log transid,
1965 * otherwise subsequent syncs to a file that's been synced in this
1966 * transaction will appear to have already occurred.
1968 spin_lock(&BTRFS_I(inode)->lock);
1969 BTRFS_I(inode)->last_sub_trans = root->log_transid;
1970 spin_unlock(&BTRFS_I(inode)->lock);
1971 if (num_written > 0)
1972 num_written = generic_write_sync(iocb, num_written);
1975 atomic_dec(&BTRFS_I(inode)->sync_writers);
1977 current->backing_dev_info = NULL;
1978 return num_written ? num_written : err;
1981 int btrfs_release_file(struct inode *inode, struct file *filp)
1983 struct btrfs_file_private *private = filp->private_data;
1985 if (private && private->filldir_buf)
1986 kfree(private->filldir_buf);
1988 filp->private_data = NULL;
1991 * ordered_data_close is set by settattr when we are about to truncate
1992 * a file from a non-zero size to a zero size. This tries to
1993 * flush down new bytes that may have been written if the
1994 * application were using truncate to replace a file in place.
1996 if (test_and_clear_bit(BTRFS_INODE_ORDERED_DATA_CLOSE,
1997 &BTRFS_I(inode)->runtime_flags))
1998 filemap_flush(inode->i_mapping);
2002 static int start_ordered_ops(struct inode *inode, loff_t start, loff_t end)
2005 struct blk_plug plug;
2008 * This is only called in fsync, which would do synchronous writes, so
2009 * a plug can merge adjacent IOs as much as possible. Esp. in case of
2010 * multiple disks using raid profile, a large IO can be split to
2011 * several segments of stripe length (currently 64K).
2013 blk_start_plug(&plug);
2014 atomic_inc(&BTRFS_I(inode)->sync_writers);
2015 ret = btrfs_fdatawrite_range(inode, start, end);
2016 atomic_dec(&BTRFS_I(inode)->sync_writers);
2017 blk_finish_plug(&plug);
2023 * fsync call for both files and directories. This logs the inode into
2024 * the tree log instead of forcing full commits whenever possible.
2026 * It needs to call filemap_fdatawait so that all ordered extent updates are
2027 * in the metadata btree are up to date for copying to the log.
2029 * It drops the inode mutex before doing the tree log commit. This is an
2030 * important optimization for directories because holding the mutex prevents
2031 * new operations on the dir while we write to disk.
2033 int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
2035 struct dentry *dentry = file_dentry(file);
2036 struct inode *inode = d_inode(dentry);
2037 struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
2038 struct btrfs_root *root = BTRFS_I(inode)->root;
2039 struct btrfs_trans_handle *trans;
2040 struct btrfs_log_ctx ctx;
2045 * The range length can be represented by u64, we have to do the typecasts
2046 * to avoid signed overflow if it's [0, LLONG_MAX] eg. from fsync()
2048 len = (u64)end - (u64)start + 1;
2049 trace_btrfs_sync_file(file, datasync);
2051 btrfs_init_log_ctx(&ctx, inode);
2054 * We write the dirty pages in the range and wait until they complete
2055 * out of the ->i_mutex. If so, we can flush the dirty pages by
2056 * multi-task, and make the performance up. See
2057 * btrfs_wait_ordered_range for an explanation of the ASYNC check.
2059 ret = start_ordered_ops(inode, start, end);
2064 atomic_inc(&root->log_batch);
2067 * We have to do this here to avoid the priority inversion of waiting on
2068 * IO of a lower priority task while holding a transaciton open.
2070 ret = btrfs_wait_ordered_range(inode, start, len);
2072 inode_unlock(inode);
2075 atomic_inc(&root->log_batch);
2078 if (btrfs_inode_in_log(BTRFS_I(inode), fs_info->generation) ||
2079 BTRFS_I(inode)->last_trans <= fs_info->last_trans_committed) {
2081 * We've had everything committed since the last time we were
2082 * modified so clear this flag in case it was set for whatever
2083 * reason, it's no longer relevant.
2085 clear_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
2086 &BTRFS_I(inode)->runtime_flags);
2088 * An ordered extent might have started before and completed
2089 * already with io errors, in which case the inode was not
2090 * updated and we end up here. So check the inode's mapping
2091 * for any errors that might have happened since we last
2092 * checked called fsync.
2094 ret = filemap_check_wb_err(inode->i_mapping, file->f_wb_err);
2095 inode_unlock(inode);
2100 * We use start here because we will need to wait on the IO to complete
2101 * in btrfs_sync_log, which could require joining a transaction (for
2102 * example checking cross references in the nocow path). If we use join
2103 * here we could get into a situation where we're waiting on IO to
2104 * happen that is blocked on a transaction trying to commit. With start
2105 * we inc the extwriter counter, so we wait for all extwriters to exit
2106 * before we start blocking join'ers. This comment is to keep somebody
2107 * from thinking they are super smart and changing this to
2108 * btrfs_join_transaction *cough*Josef*cough*.
2110 trans = btrfs_start_transaction(root, 0);
2111 if (IS_ERR(trans)) {
2112 ret = PTR_ERR(trans);
2113 inode_unlock(inode);
2118 ret = btrfs_log_dentry_safe(trans, dentry, start, end, &ctx);
2120 /* Fallthrough and commit/free transaction. */
2124 /* we've logged all the items and now have a consistent
2125 * version of the file in the log. It is possible that
2126 * someone will come in and modify the file, but that's
2127 * fine because the log is consistent on disk, and we
2128 * have references to all of the file's extents
2130 * It is possible that someone will come in and log the
2131 * file again, but that will end up using the synchronization
2132 * inside btrfs_sync_log to keep things safe.
2134 inode_unlock(inode);
2137 * If any of the ordered extents had an error, just return it to user
2138 * space, so that the application knows some writes didn't succeed and
2139 * can take proper action (retry for e.g.). Blindly committing the
2140 * transaction in this case, would fool userspace that everything was
2141 * successful. And we also want to make sure our log doesn't contain
2142 * file extent items pointing to extents that weren't fully written to -
2143 * just like in the non fast fsync path, where we check for the ordered
2144 * operation's error flag before writing to the log tree and return -EIO
2145 * if any of them had this flag set (btrfs_wait_ordered_range) -
2146 * therefore we need to check for errors in the ordered operations,
2147 * which are indicated by ctx.io_err.
2150 btrfs_end_transaction(trans);
2155 if (ret != BTRFS_NO_LOG_SYNC) {
2157 ret = btrfs_sync_log(trans, root, &ctx);
2159 ret = btrfs_end_transaction(trans);
2163 ret = btrfs_commit_transaction(trans);
2165 ret = btrfs_end_transaction(trans);
2168 ASSERT(list_empty(&ctx.list));
2169 err = file_check_and_advance_wb_err(file);
2172 return ret > 0 ? -EIO : ret;
2175 static const struct vm_operations_struct btrfs_file_vm_ops = {
2176 .fault = filemap_fault,
2177 .map_pages = filemap_map_pages,
2178 .page_mkwrite = btrfs_page_mkwrite,
2181 static int btrfs_file_mmap(struct file *filp, struct vm_area_struct *vma)
2183 struct address_space *mapping = filp->f_mapping;
2185 if (!mapping->a_ops->readpage)
2188 file_accessed(filp);
2189 vma->vm_ops = &btrfs_file_vm_ops;
2194 static int hole_mergeable(struct btrfs_inode *inode, struct extent_buffer *leaf,
2195 int slot, u64 start, u64 end)
2197 struct btrfs_file_extent_item *fi;
2198 struct btrfs_key key;
2200 if (slot < 0 || slot >= btrfs_header_nritems(leaf))
2203 btrfs_item_key_to_cpu(leaf, &key, slot);
2204 if (key.objectid != btrfs_ino(inode) ||
2205 key.type != BTRFS_EXTENT_DATA_KEY)
2208 fi = btrfs_item_ptr(leaf, slot, struct btrfs_file_extent_item);
2210 if (btrfs_file_extent_type(leaf, fi) != BTRFS_FILE_EXTENT_REG)
2213 if (btrfs_file_extent_disk_bytenr(leaf, fi))
2216 if (key.offset == end)
2218 if (key.offset + btrfs_file_extent_num_bytes(leaf, fi) == start)
2223 static int fill_holes(struct btrfs_trans_handle *trans,
2224 struct btrfs_inode *inode,
2225 struct btrfs_path *path, u64 offset, u64 end)
2227 struct btrfs_fs_info *fs_info = trans->fs_info;
2228 struct btrfs_root *root = inode->root;
2229 struct extent_buffer *leaf;
2230 struct btrfs_file_extent_item *fi;
2231 struct extent_map *hole_em;
2232 struct extent_map_tree *em_tree = &inode->extent_tree;
2233 struct btrfs_key key;
2236 if (btrfs_fs_incompat(fs_info, NO_HOLES))
2239 key.objectid = btrfs_ino(inode);
2240 key.type = BTRFS_EXTENT_DATA_KEY;
2241 key.offset = offset;
2243 ret = btrfs_search_slot(trans, root, &key, path, 0, 1);
2246 * We should have dropped this offset, so if we find it then
2247 * something has gone horribly wrong.
2254 leaf = path->nodes[0];
2255 if (hole_mergeable(inode, leaf, path->slots[0] - 1, offset, end)) {
2259 fi = btrfs_item_ptr(leaf, path->slots[0],
2260 struct btrfs_file_extent_item);
2261 num_bytes = btrfs_file_extent_num_bytes(leaf, fi) +
2263 btrfs_set_file_extent_num_bytes(leaf, fi, num_bytes);
2264 btrfs_set_file_extent_ram_bytes(leaf, fi, num_bytes);
2265 btrfs_set_file_extent_offset(leaf, fi, 0);
2266 btrfs_mark_buffer_dirty(leaf);
2270 if (hole_mergeable(inode, leaf, path->slots[0], offset, end)) {
2273 key.offset = offset;
2274 btrfs_set_item_key_safe(fs_info, path, &key);
2275 fi = btrfs_item_ptr(leaf, path->slots[0],
2276 struct btrfs_file_extent_item);
2277 num_bytes = btrfs_file_extent_num_bytes(leaf, fi) + end -
2279 btrfs_set_file_extent_num_bytes(leaf, fi, num_bytes);
2280 btrfs_set_file_extent_ram_bytes(leaf, fi, num_bytes);
2281 btrfs_set_file_extent_offset(leaf, fi, 0);
2282 btrfs_mark_buffer_dirty(leaf);
2285 btrfs_release_path(path);
2287 ret = btrfs_insert_file_extent(trans, root, btrfs_ino(inode),
2288 offset, 0, 0, end - offset, 0, end - offset, 0, 0, 0);
2293 btrfs_release_path(path);
2295 hole_em = alloc_extent_map();
2297 btrfs_drop_extent_cache(inode, offset, end - 1, 0);
2298 set_bit(BTRFS_INODE_NEEDS_FULL_SYNC, &inode->runtime_flags);
2300 hole_em->start = offset;
2301 hole_em->len = end - offset;
2302 hole_em->ram_bytes = hole_em->len;
2303 hole_em->orig_start = offset;
2305 hole_em->block_start = EXTENT_MAP_HOLE;
2306 hole_em->block_len = 0;
2307 hole_em->orig_block_len = 0;
2308 hole_em->bdev = fs_info->fs_devices->latest_bdev;
2309 hole_em->compress_type = BTRFS_COMPRESS_NONE;
2310 hole_em->generation = trans->transid;
2313 btrfs_drop_extent_cache(inode, offset, end - 1, 0);
2314 write_lock(&em_tree->lock);
2315 ret = add_extent_mapping(em_tree, hole_em, 1);
2316 write_unlock(&em_tree->lock);
2317 } while (ret == -EEXIST);
2318 free_extent_map(hole_em);
2320 set_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
2321 &inode->runtime_flags);
2328 * Find a hole extent on given inode and change start/len to the end of hole
2329 * extent.(hole/vacuum extent whose em->start <= start &&
2330 * em->start + em->len > start)
2331 * When a hole extent is found, return 1 and modify start/len.
2333 static int find_first_non_hole(struct inode *inode, u64 *start, u64 *len)
2335 struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
2336 struct extent_map *em;
2339 em = btrfs_get_extent(BTRFS_I(inode), NULL, 0,
2340 round_down(*start, fs_info->sectorsize),
2341 round_up(*len, fs_info->sectorsize), 0);
2345 /* Hole or vacuum extent(only exists in no-hole mode) */
2346 if (em->block_start == EXTENT_MAP_HOLE) {
2348 *len = em->start + em->len > *start + *len ?
2349 0 : *start + *len - em->start - em->len;
2350 *start = em->start + em->len;
2352 free_extent_map(em);
2356 static int btrfs_punch_hole_lock_range(struct inode *inode,
2357 const u64 lockstart,
2359 struct extent_state **cached_state)
2362 struct btrfs_ordered_extent *ordered;
2365 truncate_pagecache_range(inode, lockstart, lockend);
2367 lock_extent_bits(&BTRFS_I(inode)->io_tree, lockstart, lockend,
2369 ordered = btrfs_lookup_first_ordered_extent(inode, lockend);
2372 * We need to make sure we have no ordered extents in this range
2373 * and nobody raced in and read a page in this range, if we did
2374 * we need to try again.
2377 (ordered->file_offset + ordered->len <= lockstart ||
2378 ordered->file_offset > lockend)) &&
2379 !filemap_range_has_page(inode->i_mapping,
2380 lockstart, lockend)) {
2382 btrfs_put_ordered_extent(ordered);
2386 btrfs_put_ordered_extent(ordered);
2387 unlock_extent_cached(&BTRFS_I(inode)->io_tree, lockstart,
2388 lockend, cached_state);
2389 ret = btrfs_wait_ordered_range(inode, lockstart,
2390 lockend - lockstart + 1);
2397 static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
2399 struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
2400 struct btrfs_root *root = BTRFS_I(inode)->root;
2401 struct extent_state *cached_state = NULL;
2402 struct btrfs_path *path;
2403 struct btrfs_block_rsv *rsv;
2404 struct btrfs_trans_handle *trans;
2409 u64 orig_start = offset;
2411 u64 min_size = btrfs_calc_trans_metadata_size(fs_info, 1);
2415 unsigned int rsv_count;
2417 bool no_holes = btrfs_fs_incompat(fs_info, NO_HOLES);
2419 bool truncated_block = false;
2420 bool updated_inode = false;
2422 ret = btrfs_wait_ordered_range(inode, offset, len);
2427 ino_size = round_up(inode->i_size, fs_info->sectorsize);
2428 ret = find_first_non_hole(inode, &offset, &len);
2430 goto out_only_mutex;
2432 /* Already in a large hole */
2434 goto out_only_mutex;
2437 lockstart = round_up(offset, btrfs_inode_sectorsize(inode));
2438 lockend = round_down(offset + len,
2439 btrfs_inode_sectorsize(inode)) - 1;
2440 same_block = (BTRFS_BYTES_TO_BLKS(fs_info, offset))
2441 == (BTRFS_BYTES_TO_BLKS(fs_info, offset + len - 1));
2443 * We needn't truncate any block which is beyond the end of the file
2444 * because we are sure there is no data there.
2447 * Only do this if we are in the same block and we aren't doing the
2450 if (same_block && len < fs_info->sectorsize) {
2451 if (offset < ino_size) {
2452 truncated_block = true;
2453 ret = btrfs_truncate_block(inode, offset, len, 0);
2457 goto out_only_mutex;
2460 /* zero back part of the first block */
2461 if (offset < ino_size) {
2462 truncated_block = true;
2463 ret = btrfs_truncate_block(inode, offset, 0, 0);
2465 inode_unlock(inode);
2470 /* Check the aligned pages after the first unaligned page,
2471 * if offset != orig_start, which means the first unaligned page
2472 * including several following pages are already in holes,
2473 * the extra check can be skipped */
2474 if (offset == orig_start) {
2475 /* after truncate page, check hole again */
2476 len = offset + len - lockstart;
2478 ret = find_first_non_hole(inode, &offset, &len);
2480 goto out_only_mutex;
2483 goto out_only_mutex;
2488 /* Check the tail unaligned part is in a hole */
2489 tail_start = lockend + 1;
2490 tail_len = offset + len - tail_start;
2492 ret = find_first_non_hole(inode, &tail_start, &tail_len);
2493 if (unlikely(ret < 0))
2494 goto out_only_mutex;
2496 /* zero the front end of the last page */
2497 if (tail_start + tail_len < ino_size) {
2498 truncated_block = true;
2499 ret = btrfs_truncate_block(inode,
2500 tail_start + tail_len,
2503 goto out_only_mutex;
2508 if (lockend < lockstart) {
2510 goto out_only_mutex;
2513 ret = btrfs_punch_hole_lock_range(inode, lockstart, lockend,
2516 inode_unlock(inode);
2517 goto out_only_mutex;
2520 path = btrfs_alloc_path();
2526 rsv = btrfs_alloc_block_rsv(fs_info, BTRFS_BLOCK_RSV_TEMP);
2531 rsv->size = btrfs_calc_trans_metadata_size(fs_info, 1);
2535 * 1 - update the inode
2536 * 1 - removing the extents in the range
2537 * 1 - adding the hole extent if no_holes isn't set
2539 rsv_count = no_holes ? 2 : 3;
2540 trans = btrfs_start_transaction(root, rsv_count);
2541 if (IS_ERR(trans)) {
2542 err = PTR_ERR(trans);
2546 ret = btrfs_block_rsv_migrate(&fs_info->trans_block_rsv, rsv,
2549 trans->block_rsv = rsv;
2551 cur_offset = lockstart;
2552 len = lockend - cur_offset;
2553 while (cur_offset < lockend) {
2554 ret = __btrfs_drop_extents(trans, root, inode, path,
2555 cur_offset, lockend + 1,
2556 &drop_end, 1, 0, 0, NULL);
2560 trans->block_rsv = &fs_info->trans_block_rsv;
2562 if (cur_offset < drop_end && cur_offset < ino_size) {
2563 ret = fill_holes(trans, BTRFS_I(inode), path,
2564 cur_offset, drop_end);
2567 * If we failed then we didn't insert our hole
2568 * entries for the area we dropped, so now the
2569 * fs is corrupted, so we must abort the
2572 btrfs_abort_transaction(trans, ret);
2578 cur_offset = drop_end;
2580 ret = btrfs_update_inode(trans, root, inode);
2586 btrfs_end_transaction(trans);
2587 btrfs_btree_balance_dirty(fs_info);
2589 trans = btrfs_start_transaction(root, rsv_count);
2590 if (IS_ERR(trans)) {
2591 ret = PTR_ERR(trans);
2596 ret = btrfs_block_rsv_migrate(&fs_info->trans_block_rsv,
2598 BUG_ON(ret); /* shouldn't happen */
2599 trans->block_rsv = rsv;
2601 ret = find_first_non_hole(inode, &cur_offset, &len);
2602 if (unlikely(ret < 0))
2615 trans->block_rsv = &fs_info->trans_block_rsv;
2617 * If we are using the NO_HOLES feature we might have had already an
2618 * hole that overlaps a part of the region [lockstart, lockend] and
2619 * ends at (or beyond) lockend. Since we have no file extent items to
2620 * represent holes, drop_end can be less than lockend and so we must
2621 * make sure we have an extent map representing the existing hole (the
2622 * call to __btrfs_drop_extents() might have dropped the existing extent
2623 * map representing the existing hole), otherwise the fast fsync path
2624 * will not record the existence of the hole region
2625 * [existing_hole_start, lockend].
2627 if (drop_end <= lockend)
2628 drop_end = lockend + 1;
2630 * Don't insert file hole extent item if it's for a range beyond eof
2631 * (because it's useless) or if it represents a 0 bytes range (when
2632 * cur_offset == drop_end).
2634 if (cur_offset < ino_size && cur_offset < drop_end) {
2635 ret = fill_holes(trans, BTRFS_I(inode), path,
2636 cur_offset, drop_end);
2638 /* Same comment as above. */
2639 btrfs_abort_transaction(trans, ret);
2649 inode_inc_iversion(inode);
2650 inode->i_mtime = inode->i_ctime = current_time(inode);
2652 trans->block_rsv = &fs_info->trans_block_rsv;
2653 ret = btrfs_update_inode(trans, root, inode);
2654 updated_inode = true;
2655 btrfs_end_transaction(trans);
2656 btrfs_btree_balance_dirty(fs_info);
2658 btrfs_free_path(path);
2659 btrfs_free_block_rsv(fs_info, rsv);
2661 unlock_extent_cached(&BTRFS_I(inode)->io_tree, lockstart, lockend,
2664 if (!updated_inode && truncated_block && !ret && !err) {
2666 * If we only end up zeroing part of a page, we still need to
2667 * update the inode item, so that all the time fields are
2668 * updated as well as the necessary btrfs inode in memory fields
2669 * for detecting, at fsync time, if the inode isn't yet in the
2670 * log tree or it's there but not up to date.
2672 trans = btrfs_start_transaction(root, 1);
2673 if (IS_ERR(trans)) {
2674 err = PTR_ERR(trans);
2676 err = btrfs_update_inode(trans, root, inode);
2677 ret = btrfs_end_transaction(trans);
2680 inode_unlock(inode);
2686 /* Helper structure to record which range is already reserved */
2687 struct falloc_range {
2688 struct list_head list;
2694 * Helper function to add falloc range
2696 * Caller should have locked the larger range of extent containing
2699 static int add_falloc_range(struct list_head *head, u64 start, u64 len)
2701 struct falloc_range *prev = NULL;
2702 struct falloc_range *range = NULL;
2704 if (list_empty(head))
2708 * As fallocate iterate by bytenr order, we only need to check
2711 prev = list_entry(head->prev, struct falloc_range, list);
2712 if (prev->start + prev->len == start) {
2717 range = kmalloc(sizeof(*range), GFP_KERNEL);
2720 range->start = start;
2722 list_add_tail(&range->list, head);
2726 static int btrfs_fallocate_update_isize(struct inode *inode,
2730 struct btrfs_trans_handle *trans;
2731 struct btrfs_root *root = BTRFS_I(inode)->root;
2735 if (mode & FALLOC_FL_KEEP_SIZE || end <= i_size_read(inode))
2738 trans = btrfs_start_transaction(root, 1);
2740 return PTR_ERR(trans);
2742 inode->i_ctime = current_time(inode);
2743 i_size_write(inode, end);
2744 btrfs_ordered_update_i_size(inode, end, NULL);
2745 ret = btrfs_update_inode(trans, root, inode);
2746 ret2 = btrfs_end_transaction(trans);
2748 return ret ? ret : ret2;
2752 RANGE_BOUNDARY_WRITTEN_EXTENT = 0,
2753 RANGE_BOUNDARY_PREALLOC_EXTENT = 1,
2754 RANGE_BOUNDARY_HOLE = 2,
2757 static int btrfs_zero_range_check_range_boundary(struct inode *inode,
2760 const u64 sectorsize = btrfs_inode_sectorsize(inode);
2761 struct extent_map *em;
2764 offset = round_down(offset, sectorsize);
2765 em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize, 0);
2769 if (em->block_start == EXTENT_MAP_HOLE)
2770 ret = RANGE_BOUNDARY_HOLE;
2771 else if (test_bit(EXTENT_FLAG_PREALLOC, &em->flags))
2772 ret = RANGE_BOUNDARY_PREALLOC_EXTENT;
2774 ret = RANGE_BOUNDARY_WRITTEN_EXTENT;
2776 free_extent_map(em);
2780 static int btrfs_zero_range(struct inode *inode,
2785 struct btrfs_fs_info *fs_info = BTRFS_I(inode)->root->fs_info;
2786 struct extent_map *em;
2787 struct extent_changeset *data_reserved = NULL;
2790 const u64 sectorsize = btrfs_inode_sectorsize(inode);
2791 u64 alloc_start = round_down(offset, sectorsize);
2792 u64 alloc_end = round_up(offset + len, sectorsize);
2793 u64 bytes_to_reserve = 0;
2794 bool space_reserved = false;
2796 inode_dio_wait(inode);
2798 em = btrfs_get_extent(BTRFS_I(inode), NULL, 0,
2799 alloc_start, alloc_end - alloc_start, 0);
2806 * Avoid hole punching and extent allocation for some cases. More cases
2807 * could be considered, but these are unlikely common and we keep things
2808 * as simple as possible for now. Also, intentionally, if the target
2809 * range contains one or more prealloc extents together with regular
2810 * extents and holes, we drop all the existing extents and allocate a
2811 * new prealloc extent, so that we get a larger contiguous disk extent.
2813 if (em->start <= alloc_start &&
2814 test_bit(EXTENT_FLAG_PREALLOC, &em->flags)) {
2815 const u64 em_end = em->start + em->len;
2817 if (em_end >= offset + len) {
2819 * The whole range is already a prealloc extent,
2820 * do nothing except updating the inode's i_size if
2823 free_extent_map(em);
2824 ret = btrfs_fallocate_update_isize(inode, offset + len,
2829 * Part of the range is already a prealloc extent, so operate
2830 * only on the remaining part of the range.
2832 alloc_start = em_end;
2833 ASSERT(IS_ALIGNED(alloc_start, sectorsize));
2834 len = offset + len - alloc_start;
2835 offset = alloc_start;
2836 alloc_hint = em->block_start + em->len;
2838 free_extent_map(em);
2840 if (BTRFS_BYTES_TO_BLKS(fs_info, offset) ==
2841 BTRFS_BYTES_TO_BLKS(fs_info, offset + len - 1)) {
2842 em = btrfs_get_extent(BTRFS_I(inode), NULL, 0,
2843 alloc_start, sectorsize, 0);
2849 if (test_bit(EXTENT_FLAG_PREALLOC, &em->flags)) {
2850 free_extent_map(em);
2851 ret = btrfs_fallocate_update_isize(inode, offset + len,
2855 if (len < sectorsize && em->block_start != EXTENT_MAP_HOLE) {
2856 free_extent_map(em);
2857 ret = btrfs_truncate_block(inode, offset, len, 0);
2859 ret = btrfs_fallocate_update_isize(inode,
2864 free_extent_map(em);
2865 alloc_start = round_down(offset, sectorsize);
2866 alloc_end = alloc_start + sectorsize;
2870 alloc_start = round_up(offset, sectorsize);
2871 alloc_end = round_down(offset + len, sectorsize);
2874 * For unaligned ranges, check the pages at the boundaries, they might
2875 * map to an extent, in which case we need to partially zero them, or
2876 * they might map to a hole, in which case we need our allocation range
2879 if (!IS_ALIGNED(offset, sectorsize)) {
2880 ret = btrfs_zero_range_check_range_boundary(inode, offset);
2883 if (ret == RANGE_BOUNDARY_HOLE) {
2884 alloc_start = round_down(offset, sectorsize);
2886 } else if (ret == RANGE_BOUNDARY_WRITTEN_EXTENT) {
2887 ret = btrfs_truncate_block(inode, offset, 0, 0);
2895 if (!IS_ALIGNED(offset + len, sectorsize)) {
2896 ret = btrfs_zero_range_check_range_boundary(inode,
2900 if (ret == RANGE_BOUNDARY_HOLE) {
2901 alloc_end = round_up(offset + len, sectorsize);
2903 } else if (ret == RANGE_BOUNDARY_WRITTEN_EXTENT) {
2904 ret = btrfs_truncate_block(inode, offset + len, 0, 1);
2913 if (alloc_start < alloc_end) {
2914 struct extent_state *cached_state = NULL;
2915 const u64 lockstart = alloc_start;
2916 const u64 lockend = alloc_end - 1;
2918 bytes_to_reserve = alloc_end - alloc_start;
2919 ret = btrfs_alloc_data_chunk_ondemand(BTRFS_I(inode),
2923 space_reserved = true;
2924 ret = btrfs_qgroup_reserve_data(inode, &data_reserved,
2925 alloc_start, bytes_to_reserve);
2928 ret = btrfs_punch_hole_lock_range(inode, lockstart, lockend,
2932 ret = btrfs_prealloc_file_range(inode, mode, alloc_start,
2933 alloc_end - alloc_start,
2935 offset + len, &alloc_hint);
2936 unlock_extent_cached(&BTRFS_I(inode)->io_tree, lockstart,
2937 lockend, &cached_state);
2938 /* btrfs_prealloc_file_range releases reserved space on error */
2940 space_reserved = false;
2944 ret = btrfs_fallocate_update_isize(inode, offset + len, mode);
2946 if (ret && space_reserved)
2947 btrfs_free_reserved_data_space(inode, data_reserved,
2948 alloc_start, bytes_to_reserve);
2949 extent_changeset_free(data_reserved);
2954 static long btrfs_fallocate(struct file *file, int mode,
2955 loff_t offset, loff_t len)
2957 struct inode *inode = file_inode(file);
2958 struct extent_state *cached_state = NULL;
2959 struct extent_changeset *data_reserved = NULL;
2960 struct falloc_range *range;
2961 struct falloc_range *tmp;
2962 struct list_head reserve_list;
2970 struct extent_map *em;
2971 int blocksize = btrfs_inode_sectorsize(inode);
2974 alloc_start = round_down(offset, blocksize);
2975 alloc_end = round_up(offset + len, blocksize);
2976 cur_offset = alloc_start;
2978 /* Make sure we aren't being give some crap mode */
2979 if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE |
2980 FALLOC_FL_ZERO_RANGE))
2983 if (mode & FALLOC_FL_PUNCH_HOLE)
2984 return btrfs_punch_hole(inode, offset, len);
2987 * Only trigger disk allocation, don't trigger qgroup reserve
2989 * For qgroup space, it will be checked later.
2991 if (!(mode & FALLOC_FL_ZERO_RANGE)) {
2992 ret = btrfs_alloc_data_chunk_ondemand(BTRFS_I(inode),
2993 alloc_end - alloc_start);
3000 if (!(mode & FALLOC_FL_KEEP_SIZE) && offset + len > inode->i_size) {
3001 ret = inode_newsize_ok(inode, offset + len);
3007 * TODO: Move these two operations after we have checked
3008 * accurate reserved space, or fallocate can still fail but
3009 * with page truncated or size expanded.
3011 * But that's a minor problem and won't do much harm BTW.
3013 if (alloc_start > inode->i_size) {
3014 ret = btrfs_cont_expand(inode, i_size_read(inode),
3018 } else if (offset + len > inode->i_size) {
3020 * If we are fallocating from the end of the file onward we
3021 * need to zero out the end of the block if i_size lands in the
3022 * middle of a block.
3024 ret = btrfs_truncate_block(inode, inode->i_size, 0, 0);
3030 * wait for ordered IO before we have any locks. We'll loop again
3031 * below with the locks held.
3033 ret = btrfs_wait_ordered_range(inode, alloc_start,
3034 alloc_end - alloc_start);
3038 if (mode & FALLOC_FL_ZERO_RANGE) {
3039 ret = btrfs_zero_range(inode, offset, len, mode);
3040 inode_unlock(inode);
3044 locked_end = alloc_end - 1;
3046 struct btrfs_ordered_extent *ordered;
3048 /* the extent lock is ordered inside the running
3051 lock_extent_bits(&BTRFS_I(inode)->io_tree, alloc_start,
3052 locked_end, &cached_state);
3053 ordered = btrfs_lookup_first_ordered_extent(inode, locked_end);
3056 ordered->file_offset + ordered->len > alloc_start &&
3057 ordered->file_offset < alloc_end) {
3058 btrfs_put_ordered_extent(ordered);
3059 unlock_extent_cached(&BTRFS_I(inode)->io_tree,
3060 alloc_start, locked_end,
3063 * we can't wait on the range with the transaction
3064 * running or with the extent lock held
3066 ret = btrfs_wait_ordered_range(inode, alloc_start,
3067 alloc_end - alloc_start);
3072 btrfs_put_ordered_extent(ordered);
3077 /* First, check if we exceed the qgroup limit */
3078 INIT_LIST_HEAD(&reserve_list);
3079 while (cur_offset < alloc_end) {
3080 em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, cur_offset,
3081 alloc_end - cur_offset, 0);
3086 last_byte = min(extent_map_end(em), alloc_end);
3087 actual_end = min_t(u64, extent_map_end(em), offset + len);
3088 last_byte = ALIGN(last_byte, blocksize);
3089 if (em->block_start == EXTENT_MAP_HOLE ||
3090 (cur_offset >= inode->i_size &&
3091 !test_bit(EXTENT_FLAG_PREALLOC, &em->flags))) {
3092 ret = add_falloc_range(&reserve_list, cur_offset,
3093 last_byte - cur_offset);
3095 free_extent_map(em);
3098 ret = btrfs_qgroup_reserve_data(inode, &data_reserved,
3099 cur_offset, last_byte - cur_offset);
3101 free_extent_map(em);
3106 * Do not need to reserve unwritten extent for this
3107 * range, free reserved data space first, otherwise
3108 * it'll result in false ENOSPC error.
3110 btrfs_free_reserved_data_space(inode, data_reserved,
3111 cur_offset, last_byte - cur_offset);
3113 free_extent_map(em);
3114 cur_offset = last_byte;
3118 * If ret is still 0, means we're OK to fallocate.
3119 * Or just cleanup the list and exit.
3121 list_for_each_entry_safe(range, tmp, &reserve_list, list) {
3123 ret = btrfs_prealloc_file_range(inode, mode,
3125 range->len, i_blocksize(inode),
3126 offset + len, &alloc_hint);
3128 btrfs_free_reserved_data_space(inode,
3129 data_reserved, range->start,
3131 list_del(&range->list);
3138 * We didn't need to allocate any more space, but we still extended the
3139 * size of the file so we need to update i_size and the inode item.
3141 ret = btrfs_fallocate_update_isize(inode, actual_end, mode);
3143 unlock_extent_cached(&BTRFS_I(inode)->io_tree, alloc_start, locked_end,
3146 inode_unlock(inode);
3147 /* Let go of our reservation. */
3148 if (ret != 0 && !(mode & FALLOC_FL_ZERO_RANGE))
3149 btrfs_free_reserved_data_space(inode, data_reserved,
3150 alloc_start, alloc_end - cur_offset);
3151 extent_changeset_free(data_reserved);
3155 static int find_desired_extent(struct inode *inode, loff_t *offset, int whence)
3157 struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
3158 struct extent_map *em = NULL;
3159 struct extent_state *cached_state = NULL;
3166 if (inode->i_size == 0)
3170 * *offset can be negative, in this case we start finding DATA/HOLE from
3171 * the very start of the file.
3173 start = max_t(loff_t, 0, *offset);
3175 lockstart = round_down(start, fs_info->sectorsize);
3176 lockend = round_up(i_size_read(inode),
3177 fs_info->sectorsize);
3178 if (lockend <= lockstart)
3179 lockend = lockstart + fs_info->sectorsize;
3181 len = lockend - lockstart + 1;
3183 lock_extent_bits(&BTRFS_I(inode)->io_tree, lockstart, lockend,
3186 while (start < inode->i_size) {
3187 em = btrfs_get_extent_fiemap(BTRFS_I(inode), NULL, 0,
3195 if (whence == SEEK_HOLE &&
3196 (em->block_start == EXTENT_MAP_HOLE ||
3197 test_bit(EXTENT_FLAG_PREALLOC, &em->flags)))
3199 else if (whence == SEEK_DATA &&
3200 (em->block_start != EXTENT_MAP_HOLE &&
3201 !test_bit(EXTENT_FLAG_PREALLOC, &em->flags)))
3204 start = em->start + em->len;
3205 free_extent_map(em);
3209 free_extent_map(em);
3211 if (whence == SEEK_DATA && start >= inode->i_size)
3214 *offset = min_t(loff_t, start, inode->i_size);
3216 unlock_extent_cached(&BTRFS_I(inode)->io_tree, lockstart, lockend,
3221 static loff_t btrfs_file_llseek(struct file *file, loff_t offset, int whence)
3223 struct inode *inode = file->f_mapping->host;
3230 offset = generic_file_llseek(file, offset, whence);
3234 if (offset >= i_size_read(inode)) {
3235 inode_unlock(inode);
3239 ret = find_desired_extent(inode, &offset, whence);
3241 inode_unlock(inode);
3246 offset = vfs_setpos(file, offset, inode->i_sb->s_maxbytes);
3248 inode_unlock(inode);
3252 static int btrfs_file_open(struct inode *inode, struct file *filp)
3254 filp->f_mode |= FMODE_NOWAIT;
3255 return generic_file_open(inode, filp);
3258 const struct file_operations btrfs_file_operations = {
3259 .llseek = btrfs_file_llseek,
3260 .read_iter = generic_file_read_iter,
3261 .splice_read = generic_file_splice_read,
3262 .write_iter = btrfs_file_write_iter,
3263 .mmap = btrfs_file_mmap,
3264 .open = btrfs_file_open,
3265 .release = btrfs_release_file,
3266 .fsync = btrfs_sync_file,
3267 .fallocate = btrfs_fallocate,
3268 .unlocked_ioctl = btrfs_ioctl,
3269 #ifdef CONFIG_COMPAT
3270 .compat_ioctl = btrfs_compat_ioctl,
3272 .clone_file_range = btrfs_clone_file_range,
3273 .dedupe_file_range = btrfs_dedupe_file_range,
3276 void __cold btrfs_auto_defrag_exit(void)
3278 kmem_cache_destroy(btrfs_inode_defrag_cachep);
3281 int __init btrfs_auto_defrag_init(void)
3283 btrfs_inode_defrag_cachep = kmem_cache_create("btrfs_inode_defrag",
3284 sizeof(struct inode_defrag), 0,
3287 if (!btrfs_inode_defrag_cachep)
3293 int btrfs_fdatawrite_range(struct inode *inode, loff_t start, loff_t end)
3298 * So with compression we will find and lock a dirty page and clear the
3299 * first one as dirty, setup an async extent, and immediately return
3300 * with the entire range locked but with nobody actually marked with
3301 * writeback. So we can't just filemap_write_and_wait_range() and
3302 * expect it to work since it will just kick off a thread to do the
3303 * actual work. So we need to call filemap_fdatawrite_range _again_
3304 * since it will wait on the page lock, which won't be unlocked until
3305 * after the pages have been marked as writeback and so we're good to go
3306 * from there. We have to do this otherwise we'll miss the ordered
3307 * extents and that results in badness. Please Josef, do not think you
3308 * know better and pull this out at some point in the future, it is
3309 * right and you are wrong.
3311 ret = filemap_fdatawrite_range(inode->i_mapping, start, end);
3312 if (!ret && test_bit(BTRFS_INODE_HAS_ASYNC_EXTENT,
3313 &BTRFS_I(inode)->runtime_flags))
3314 ret = filemap_fdatawrite_range(inode->i_mapping, start, end);