2 * Copyright (C) 2008 Oracle. All rights reserved.
4 * This program is free software; you can redistribute it and/or
5 * modify it under the terms of the GNU General Public
6 * License v2 as published by the Free Software Foundation.
8 * This program is distributed in the hope that it will be useful,
9 * but WITHOUT ANY WARRANTY; without even the implied warranty of
10 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
11 * General Public License for more details.
13 * You should have received a copy of the GNU General Public
14 * License along with this program; if not, write to the
15 * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
16 * Boston, MA 021110-1307, USA.
19 #include <linux/sched.h>
20 #include <linux/slab.h>
21 #include <linux/list_sort.h>
23 #include "transaction.h"
26 #include "print-tree.h"
32 /* magic values for the inode_only field in btrfs_log_inode:
34 * LOG_INODE_ALL means to log everything
35 * LOG_INODE_EXISTS means to log just enough to recreate the inode
38 #define LOG_INODE_ALL 0
39 #define LOG_INODE_EXISTS 1
42 * directory trouble cases
44 * 1) on rename or unlink, if the inode being unlinked isn't in the fsync
45 * log, we must force a full commit before doing an fsync of the directory
46 * where the unlink was done.
47 * ---> record transid of last unlink/rename per directory
51 * rename foo/some_dir foo2/some_dir
53 * fsync foo/some_dir/some_file
55 * The fsync above will unlink the original some_dir without recording
56 * it in its new location (foo2). After a crash, some_dir will be gone
57 * unless the fsync of some_file forces a full commit
59 * 2) we must log any new names for any file or dir that is in the fsync
60 * log. ---> check inode while renaming/linking.
62 * 2a) we must log any new names for any file or dir during rename
63 * when the directory they are being removed from was logged.
64 * ---> check inode and old parent dir during rename
66 * 2a is actually the more important variant. With the extra logging
67 * a crash might unlink the old name without recreating the new one
69 * 3) after a crash, we must go through any directories with a link count
70 * of zero and redo the rm -rf
77 * The directory f1 was fully removed from the FS, but fsync was never
78 * called on f1, only its parent dir. After a crash the rm -rf must
79 * be replayed. This must be able to recurse down the entire
80 * directory tree. The inode link count fixup code takes care of the
85 * stages for the tree walking. The first
86 * stage (0) is to only pin down the blocks we find
87 * the second stage (1) is to make sure that all the inodes
88 * we find in the log are created in the subvolume.
90 * The last stage is to deal with directories and links and extents
91 * and all the other fun semantics
93 #define LOG_WALK_PIN_ONLY 0
94 #define LOG_WALK_REPLAY_INODES 1
95 #define LOG_WALK_REPLAY_ALL 2
97 static int btrfs_log_inode(struct btrfs_trans_handle *trans,
98 struct btrfs_root *root, struct inode *inode,
100 static int link_to_fixup_dir(struct btrfs_trans_handle *trans,
101 struct btrfs_root *root,
102 struct btrfs_path *path, u64 objectid);
103 static noinline int replay_dir_deletes(struct btrfs_trans_handle *trans,
104 struct btrfs_root *root,
105 struct btrfs_root *log,
106 struct btrfs_path *path,
107 u64 dirid, int del_all);
110 * tree logging is a special write ahead log used to make sure that
111 * fsyncs and O_SYNCs can happen without doing full tree commits.
113 * Full tree commits are expensive because they require commonly
114 * modified blocks to be recowed, creating many dirty pages in the
115 * extent tree an 4x-6x higher write load than ext3.
117 * Instead of doing a tree commit on every fsync, we use the
118 * key ranges and transaction ids to find items for a given file or directory
119 * that have changed in this transaction. Those items are copied into
120 * a special tree (one per subvolume root), that tree is written to disk
121 * and then the fsync is considered complete.
123 * After a crash, items are copied out of the log-tree back into the
124 * subvolume tree. Any file data extents found are recorded in the extent
125 * allocation tree, and the log-tree freed.
127 * The log tree is read three times, once to pin down all the extents it is
128 * using in ram and once, once to create all the inodes logged in the tree
129 * and once to do all the other items.
133 * start a sub transaction and setup the log tree
134 * this increments the log tree writer count to make the people
135 * syncing the tree wait for us to finish
137 static int start_log_trans(struct btrfs_trans_handle *trans,
138 struct btrfs_root *root)
143 mutex_lock(&root->log_mutex);
144 if (root->log_root) {
145 if (!root->log_start_pid) {
146 root->log_start_pid = current->pid;
147 root->log_multiple_pids = false;
148 } else if (root->log_start_pid != current->pid) {
149 root->log_multiple_pids = true;
152 atomic_inc(&root->log_batch);
153 atomic_inc(&root->log_writers);
154 mutex_unlock(&root->log_mutex);
157 root->log_multiple_pids = false;
158 root->log_start_pid = current->pid;
159 mutex_lock(&root->fs_info->tree_log_mutex);
160 if (!root->fs_info->log_root_tree) {
161 ret = btrfs_init_log_root_tree(trans, root->fs_info);
165 if (err == 0 && !root->log_root) {
166 ret = btrfs_add_log_tree(trans, root);
170 mutex_unlock(&root->fs_info->tree_log_mutex);
171 atomic_inc(&root->log_batch);
172 atomic_inc(&root->log_writers);
173 mutex_unlock(&root->log_mutex);
178 * returns 0 if there was a log transaction running and we were able
179 * to join, or returns -ENOENT if there were not transactions
182 static int join_running_log_trans(struct btrfs_root *root)
190 mutex_lock(&root->log_mutex);
191 if (root->log_root) {
193 atomic_inc(&root->log_writers);
195 mutex_unlock(&root->log_mutex);
200 * This either makes the current running log transaction wait
201 * until you call btrfs_end_log_trans() or it makes any future
202 * log transactions wait until you call btrfs_end_log_trans()
204 int btrfs_pin_log_trans(struct btrfs_root *root)
208 mutex_lock(&root->log_mutex);
209 atomic_inc(&root->log_writers);
210 mutex_unlock(&root->log_mutex);
215 * indicate we're done making changes to the log tree
216 * and wake up anyone waiting to do a sync
218 void btrfs_end_log_trans(struct btrfs_root *root)
220 if (atomic_dec_and_test(&root->log_writers)) {
222 if (waitqueue_active(&root->log_writer_wait))
223 wake_up(&root->log_writer_wait);
229 * the walk control struct is used to pass state down the chain when
230 * processing the log tree. The stage field tells us which part
231 * of the log tree processing we are currently doing. The others
232 * are state fields used for that specific part
234 struct walk_control {
235 /* should we free the extent on disk when done? This is used
236 * at transaction commit time while freeing a log tree
240 /* should we write out the extent buffer? This is used
241 * while flushing the log tree to disk during a sync
245 /* should we wait for the extent buffer io to finish? Also used
246 * while flushing the log tree to disk for a sync
250 /* pin only walk, we record which extents on disk belong to the
255 /* what stage of the replay code we're currently in */
258 /* the root we are currently replaying */
259 struct btrfs_root *replay_dest;
261 /* the trans handle for the current replay */
262 struct btrfs_trans_handle *trans;
264 /* the function that gets used to process blocks we find in the
265 * tree. Note the extent_buffer might not be up to date when it is
266 * passed in, and it must be checked or read if you need the data
269 int (*process_func)(struct btrfs_root *log, struct extent_buffer *eb,
270 struct walk_control *wc, u64 gen);
274 * process_func used to pin down extents, write them or wait on them
276 static int process_one_buffer(struct btrfs_root *log,
277 struct extent_buffer *eb,
278 struct walk_control *wc, u64 gen)
281 btrfs_pin_extent_for_log_replay(log->fs_info->extent_root,
284 if (btrfs_buffer_uptodate(eb, gen, 0)) {
286 btrfs_write_tree_block(eb);
288 btrfs_wait_tree_block_writeback(eb);
294 * Item overwrite used by replay and tree logging. eb, slot and key all refer
295 * to the src data we are copying out.
297 * root is the tree we are copying into, and path is a scratch
298 * path for use in this function (it should be released on entry and
299 * will be released on exit).
301 * If the key is already in the destination tree the existing item is
302 * overwritten. If the existing item isn't big enough, it is extended.
303 * If it is too large, it is truncated.
305 * If the key isn't in the destination yet, a new item is inserted.
307 static noinline int overwrite_item(struct btrfs_trans_handle *trans,
308 struct btrfs_root *root,
309 struct btrfs_path *path,
310 struct extent_buffer *eb, int slot,
311 struct btrfs_key *key)
315 u64 saved_i_size = 0;
316 int save_old_i_size = 0;
317 unsigned long src_ptr;
318 unsigned long dst_ptr;
319 int overwrite_root = 0;
321 if (root->root_key.objectid != BTRFS_TREE_LOG_OBJECTID)
324 item_size = btrfs_item_size_nr(eb, slot);
325 src_ptr = btrfs_item_ptr_offset(eb, slot);
327 /* look for the key in the destination tree */
328 ret = btrfs_search_slot(NULL, root, key, path, 0, 0);
332 u32 dst_size = btrfs_item_size_nr(path->nodes[0],
334 if (dst_size != item_size)
337 if (item_size == 0) {
338 btrfs_release_path(path);
341 dst_copy = kmalloc(item_size, GFP_NOFS);
342 src_copy = kmalloc(item_size, GFP_NOFS);
343 if (!dst_copy || !src_copy) {
344 btrfs_release_path(path);
350 read_extent_buffer(eb, src_copy, src_ptr, item_size);
352 dst_ptr = btrfs_item_ptr_offset(path->nodes[0], path->slots[0]);
353 read_extent_buffer(path->nodes[0], dst_copy, dst_ptr,
355 ret = memcmp(dst_copy, src_copy, item_size);
360 * they have the same contents, just return, this saves
361 * us from cowing blocks in the destination tree and doing
362 * extra writes that may not have been done by a previous
366 btrfs_release_path(path);
372 btrfs_release_path(path);
373 /* try to insert the key into the destination tree */
374 ret = btrfs_insert_empty_item(trans, root, path,
377 /* make sure any existing item is the correct size */
378 if (ret == -EEXIST) {
380 found_size = btrfs_item_size_nr(path->nodes[0],
382 if (found_size > item_size)
383 btrfs_truncate_item(trans, root, path, item_size, 1);
384 else if (found_size < item_size)
385 btrfs_extend_item(trans, root, path,
386 item_size - found_size);
390 dst_ptr = btrfs_item_ptr_offset(path->nodes[0],
393 /* don't overwrite an existing inode if the generation number
394 * was logged as zero. This is done when the tree logging code
395 * is just logging an inode to make sure it exists after recovery.
397 * Also, don't overwrite i_size on directories during replay.
398 * log replay inserts and removes directory items based on the
399 * state of the tree found in the subvolume, and i_size is modified
402 if (key->type == BTRFS_INODE_ITEM_KEY && ret == -EEXIST) {
403 struct btrfs_inode_item *src_item;
404 struct btrfs_inode_item *dst_item;
406 src_item = (struct btrfs_inode_item *)src_ptr;
407 dst_item = (struct btrfs_inode_item *)dst_ptr;
409 if (btrfs_inode_generation(eb, src_item) == 0)
412 if (overwrite_root &&
413 S_ISDIR(btrfs_inode_mode(eb, src_item)) &&
414 S_ISDIR(btrfs_inode_mode(path->nodes[0], dst_item))) {
416 saved_i_size = btrfs_inode_size(path->nodes[0],
421 copy_extent_buffer(path->nodes[0], eb, dst_ptr,
424 if (save_old_i_size) {
425 struct btrfs_inode_item *dst_item;
426 dst_item = (struct btrfs_inode_item *)dst_ptr;
427 btrfs_set_inode_size(path->nodes[0], dst_item, saved_i_size);
430 /* make sure the generation is filled in */
431 if (key->type == BTRFS_INODE_ITEM_KEY) {
432 struct btrfs_inode_item *dst_item;
433 dst_item = (struct btrfs_inode_item *)dst_ptr;
434 if (btrfs_inode_generation(path->nodes[0], dst_item) == 0) {
435 btrfs_set_inode_generation(path->nodes[0], dst_item,
440 btrfs_mark_buffer_dirty(path->nodes[0]);
441 btrfs_release_path(path);
446 * simple helper to read an inode off the disk from a given root
447 * This can only be called for subvolume roots and not for the log
449 static noinline struct inode *read_one_inode(struct btrfs_root *root,
452 struct btrfs_key key;
455 key.objectid = objectid;
456 key.type = BTRFS_INODE_ITEM_KEY;
458 inode = btrfs_iget(root->fs_info->sb, &key, root, NULL);
461 } else if (is_bad_inode(inode)) {
468 /* replays a single extent in 'eb' at 'slot' with 'key' into the
469 * subvolume 'root'. path is released on entry and should be released
472 * extents in the log tree have not been allocated out of the extent
473 * tree yet. So, this completes the allocation, taking a reference
474 * as required if the extent already exists or creating a new extent
475 * if it isn't in the extent allocation tree yet.
477 * The extent is inserted into the file, dropping any existing extents
478 * from the file that overlap the new one.
480 static noinline int replay_one_extent(struct btrfs_trans_handle *trans,
481 struct btrfs_root *root,
482 struct btrfs_path *path,
483 struct extent_buffer *eb, int slot,
484 struct btrfs_key *key)
488 u64 start = key->offset;
490 struct btrfs_file_extent_item *item;
491 struct inode *inode = NULL;
495 item = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item);
496 found_type = btrfs_file_extent_type(eb, item);
498 if (found_type == BTRFS_FILE_EXTENT_REG ||
499 found_type == BTRFS_FILE_EXTENT_PREALLOC)
500 extent_end = start + btrfs_file_extent_num_bytes(eb, item);
501 else if (found_type == BTRFS_FILE_EXTENT_INLINE) {
502 size = btrfs_file_extent_inline_len(eb, item);
503 extent_end = ALIGN(start + size, root->sectorsize);
509 inode = read_one_inode(root, key->objectid);
516 * first check to see if we already have this extent in the
517 * file. This must be done before the btrfs_drop_extents run
518 * so we don't try to drop this extent.
520 ret = btrfs_lookup_file_extent(trans, root, path, btrfs_ino(inode),
524 (found_type == BTRFS_FILE_EXTENT_REG ||
525 found_type == BTRFS_FILE_EXTENT_PREALLOC)) {
526 struct btrfs_file_extent_item cmp1;
527 struct btrfs_file_extent_item cmp2;
528 struct btrfs_file_extent_item *existing;
529 struct extent_buffer *leaf;
531 leaf = path->nodes[0];
532 existing = btrfs_item_ptr(leaf, path->slots[0],
533 struct btrfs_file_extent_item);
535 read_extent_buffer(eb, &cmp1, (unsigned long)item,
537 read_extent_buffer(leaf, &cmp2, (unsigned long)existing,
541 * we already have a pointer to this exact extent,
542 * we don't have to do anything
544 if (memcmp(&cmp1, &cmp2, sizeof(cmp1)) == 0) {
545 btrfs_release_path(path);
549 btrfs_release_path(path);
551 saved_nbytes = inode_get_bytes(inode);
552 /* drop any overlapping extents */
553 ret = btrfs_drop_extents(trans, root, inode, start, extent_end, 1);
556 if (found_type == BTRFS_FILE_EXTENT_REG ||
557 found_type == BTRFS_FILE_EXTENT_PREALLOC) {
559 unsigned long dest_offset;
560 struct btrfs_key ins;
562 ret = btrfs_insert_empty_item(trans, root, path, key,
565 dest_offset = btrfs_item_ptr_offset(path->nodes[0],
567 copy_extent_buffer(path->nodes[0], eb, dest_offset,
568 (unsigned long)item, sizeof(*item));
570 ins.objectid = btrfs_file_extent_disk_bytenr(eb, item);
571 ins.offset = btrfs_file_extent_disk_num_bytes(eb, item);
572 ins.type = BTRFS_EXTENT_ITEM_KEY;
573 offset = key->offset - btrfs_file_extent_offset(eb, item);
575 if (ins.objectid > 0) {
578 LIST_HEAD(ordered_sums);
580 * is this extent already allocated in the extent
581 * allocation tree? If so, just add a reference
583 ret = btrfs_lookup_extent(root, ins.objectid,
586 ret = btrfs_inc_extent_ref(trans, root,
587 ins.objectid, ins.offset,
588 0, root->root_key.objectid,
589 key->objectid, offset, 0);
593 * insert the extent pointer in the extent
596 ret = btrfs_alloc_logged_file_extent(trans,
597 root, root->root_key.objectid,
598 key->objectid, offset, &ins);
601 btrfs_release_path(path);
603 if (btrfs_file_extent_compression(eb, item)) {
604 csum_start = ins.objectid;
605 csum_end = csum_start + ins.offset;
607 csum_start = ins.objectid +
608 btrfs_file_extent_offset(eb, item);
609 csum_end = csum_start +
610 btrfs_file_extent_num_bytes(eb, item);
613 ret = btrfs_lookup_csums_range(root->log_root,
614 csum_start, csum_end - 1,
617 while (!list_empty(&ordered_sums)) {
618 struct btrfs_ordered_sum *sums;
619 sums = list_entry(ordered_sums.next,
620 struct btrfs_ordered_sum,
622 ret = btrfs_csum_file_blocks(trans,
623 root->fs_info->csum_root,
626 list_del(&sums->list);
630 btrfs_release_path(path);
632 } else if (found_type == BTRFS_FILE_EXTENT_INLINE) {
633 /* inline extents are easy, we just overwrite them */
634 ret = overwrite_item(trans, root, path, eb, slot, key);
638 inode_set_bytes(inode, saved_nbytes);
639 ret = btrfs_update_inode(trans, root, inode);
647 * when cleaning up conflicts between the directory names in the
648 * subvolume, directory names in the log and directory names in the
649 * inode back references, we may have to unlink inodes from directories.
651 * This is a helper function to do the unlink of a specific directory
654 static noinline int drop_one_dir_item(struct btrfs_trans_handle *trans,
655 struct btrfs_root *root,
656 struct btrfs_path *path,
658 struct btrfs_dir_item *di)
663 struct extent_buffer *leaf;
664 struct btrfs_key location;
667 leaf = path->nodes[0];
669 btrfs_dir_item_key_to_cpu(leaf, di, &location);
670 name_len = btrfs_dir_name_len(leaf, di);
671 name = kmalloc(name_len, GFP_NOFS);
675 read_extent_buffer(leaf, name, (unsigned long)(di + 1), name_len);
676 btrfs_release_path(path);
678 inode = read_one_inode(root, location.objectid);
684 ret = link_to_fixup_dir(trans, root, path, location.objectid);
687 ret = btrfs_unlink_inode(trans, root, dir, inode, name, name_len);
693 btrfs_run_delayed_items(trans, root);
698 * helper function to see if a given name and sequence number found
699 * in an inode back reference are already in a directory and correctly
700 * point to this inode
702 static noinline int inode_in_dir(struct btrfs_root *root,
703 struct btrfs_path *path,
704 u64 dirid, u64 objectid, u64 index,
705 const char *name, int name_len)
707 struct btrfs_dir_item *di;
708 struct btrfs_key location;
711 di = btrfs_lookup_dir_index_item(NULL, root, path, dirid,
712 index, name, name_len, 0);
713 if (di && !IS_ERR(di)) {
714 btrfs_dir_item_key_to_cpu(path->nodes[0], di, &location);
715 if (location.objectid != objectid)
719 btrfs_release_path(path);
721 di = btrfs_lookup_dir_item(NULL, root, path, dirid, name, name_len, 0);
722 if (di && !IS_ERR(di)) {
723 btrfs_dir_item_key_to_cpu(path->nodes[0], di, &location);
724 if (location.objectid != objectid)
730 btrfs_release_path(path);
735 * helper function to check a log tree for a named back reference in
736 * an inode. This is used to decide if a back reference that is
737 * found in the subvolume conflicts with what we find in the log.
739 * inode backreferences may have multiple refs in a single item,
740 * during replay we process one reference at a time, and we don't
741 * want to delete valid links to a file from the subvolume if that
742 * link is also in the log.
744 static noinline int backref_in_log(struct btrfs_root *log,
745 struct btrfs_key *key,
747 char *name, int namelen)
749 struct btrfs_path *path;
750 struct btrfs_inode_ref *ref;
752 unsigned long ptr_end;
753 unsigned long name_ptr;
759 path = btrfs_alloc_path();
763 ret = btrfs_search_slot(NULL, log, key, path, 0, 0);
767 ptr = btrfs_item_ptr_offset(path->nodes[0], path->slots[0]);
769 if (key->type == BTRFS_INODE_EXTREF_KEY) {
770 if (btrfs_find_name_in_ext_backref(path, ref_objectid,
771 name, namelen, NULL))
777 item_size = btrfs_item_size_nr(path->nodes[0], path->slots[0]);
778 ptr_end = ptr + item_size;
779 while (ptr < ptr_end) {
780 ref = (struct btrfs_inode_ref *)ptr;
781 found_name_len = btrfs_inode_ref_name_len(path->nodes[0], ref);
782 if (found_name_len == namelen) {
783 name_ptr = (unsigned long)(ref + 1);
784 ret = memcmp_extent_buffer(path->nodes[0], name,
791 ptr = (unsigned long)(ref + 1) + found_name_len;
794 btrfs_free_path(path);
798 static inline int __add_inode_ref(struct btrfs_trans_handle *trans,
799 struct btrfs_root *root,
800 struct btrfs_path *path,
801 struct btrfs_root *log_root,
802 struct inode *dir, struct inode *inode,
803 struct extent_buffer *eb,
804 u64 inode_objectid, u64 parent_objectid,
805 u64 ref_index, char *name, int namelen,
811 struct extent_buffer *leaf;
812 struct btrfs_dir_item *di;
813 struct btrfs_key search_key;
814 struct btrfs_inode_extref *extref;
817 /* Search old style refs */
818 search_key.objectid = inode_objectid;
819 search_key.type = BTRFS_INODE_REF_KEY;
820 search_key.offset = parent_objectid;
821 ret = btrfs_search_slot(NULL, root, &search_key, path, 0, 0);
823 struct btrfs_inode_ref *victim_ref;
825 unsigned long ptr_end;
827 leaf = path->nodes[0];
829 /* are we trying to overwrite a back ref for the root directory
830 * if so, just jump out, we're done
832 if (search_key.objectid == search_key.offset)
835 /* check all the names in this back reference to see
836 * if they are in the log. if so, we allow them to stay
837 * otherwise they must be unlinked as a conflict
839 ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
840 ptr_end = ptr + btrfs_item_size_nr(leaf, path->slots[0]);
841 while (ptr < ptr_end) {
842 victim_ref = (struct btrfs_inode_ref *)ptr;
843 victim_name_len = btrfs_inode_ref_name_len(leaf,
845 victim_name = kmalloc(victim_name_len, GFP_NOFS);
846 BUG_ON(!victim_name);
848 read_extent_buffer(leaf, victim_name,
849 (unsigned long)(victim_ref + 1),
852 if (!backref_in_log(log_root, &search_key,
856 btrfs_inc_nlink(inode);
857 btrfs_release_path(path);
859 ret = btrfs_unlink_inode(trans, root, dir,
863 btrfs_run_delayed_items(trans, root);
870 ptr = (unsigned long)(victim_ref + 1) + victim_name_len;
875 * NOTE: we have searched root tree and checked the
876 * coresponding ref, it does not need to check again.
880 btrfs_release_path(path);
882 /* Same search but for extended refs */
883 extref = btrfs_lookup_inode_extref(NULL, root, path, name, namelen,
884 inode_objectid, parent_objectid, 0,
886 if (!IS_ERR_OR_NULL(extref)) {
890 struct inode *victim_parent;
892 leaf = path->nodes[0];
894 item_size = btrfs_item_size_nr(leaf, path->slots[0]);
895 base = btrfs_item_ptr_offset(leaf, path->slots[0]);
897 while (cur_offset < item_size) {
898 extref = (struct btrfs_inode_extref *)base + cur_offset;
900 victim_name_len = btrfs_inode_extref_name_len(leaf, extref);
902 if (btrfs_inode_extref_parent(leaf, extref) != parent_objectid)
905 victim_name = kmalloc(victim_name_len, GFP_NOFS);
906 read_extent_buffer(leaf, victim_name, (unsigned long)&extref->name,
909 search_key.objectid = inode_objectid;
910 search_key.type = BTRFS_INODE_EXTREF_KEY;
911 search_key.offset = btrfs_extref_hash(parent_objectid,
915 if (!backref_in_log(log_root, &search_key,
916 parent_objectid, victim_name,
919 victim_parent = read_one_inode(root,
922 btrfs_inc_nlink(inode);
923 btrfs_release_path(path);
925 ret = btrfs_unlink_inode(trans, root,
930 btrfs_run_delayed_items(trans, root);
941 cur_offset += victim_name_len + sizeof(*extref);
945 btrfs_release_path(path);
947 /* look for a conflicting sequence number */
948 di = btrfs_lookup_dir_index_item(trans, root, path, btrfs_ino(dir),
949 ref_index, name, namelen, 0);
950 if (di && !IS_ERR(di)) {
951 ret = drop_one_dir_item(trans, root, path, dir, di);
954 btrfs_release_path(path);
956 /* look for a conflicing name */
957 di = btrfs_lookup_dir_item(trans, root, path, btrfs_ino(dir),
959 if (di && !IS_ERR(di)) {
960 ret = drop_one_dir_item(trans, root, path, dir, di);
963 btrfs_release_path(path);
968 static int extref_get_fields(struct extent_buffer *eb, unsigned long ref_ptr,
969 u32 *namelen, char **name, u64 *index,
970 u64 *parent_objectid)
972 struct btrfs_inode_extref *extref;
974 extref = (struct btrfs_inode_extref *)ref_ptr;
976 *namelen = btrfs_inode_extref_name_len(eb, extref);
977 *name = kmalloc(*namelen, GFP_NOFS);
981 read_extent_buffer(eb, *name, (unsigned long)&extref->name,
984 *index = btrfs_inode_extref_index(eb, extref);
986 *parent_objectid = btrfs_inode_extref_parent(eb, extref);
991 static int ref_get_fields(struct extent_buffer *eb, unsigned long ref_ptr,
992 u32 *namelen, char **name, u64 *index)
994 struct btrfs_inode_ref *ref;
996 ref = (struct btrfs_inode_ref *)ref_ptr;
998 *namelen = btrfs_inode_ref_name_len(eb, ref);
999 *name = kmalloc(*namelen, GFP_NOFS);
1003 read_extent_buffer(eb, *name, (unsigned long)(ref + 1), *namelen);
1005 *index = btrfs_inode_ref_index(eb, ref);
1011 * replay one inode back reference item found in the log tree.
1012 * eb, slot and key refer to the buffer and key found in the log tree.
1013 * root is the destination we are replaying into, and path is for temp
1014 * use by this function. (it should be released on return).
1016 static noinline int add_inode_ref(struct btrfs_trans_handle *trans,
1017 struct btrfs_root *root,
1018 struct btrfs_root *log,
1019 struct btrfs_path *path,
1020 struct extent_buffer *eb, int slot,
1021 struct btrfs_key *key)
1024 struct inode *inode;
1025 unsigned long ref_ptr;
1026 unsigned long ref_end;
1030 int search_done = 0;
1031 int log_ref_ver = 0;
1032 u64 parent_objectid;
1035 int ref_struct_size;
1037 ref_ptr = btrfs_item_ptr_offset(eb, slot);
1038 ref_end = ref_ptr + btrfs_item_size_nr(eb, slot);
1040 if (key->type == BTRFS_INODE_EXTREF_KEY) {
1041 struct btrfs_inode_extref *r;
1043 ref_struct_size = sizeof(struct btrfs_inode_extref);
1045 r = (struct btrfs_inode_extref *)ref_ptr;
1046 parent_objectid = btrfs_inode_extref_parent(eb, r);
1048 ref_struct_size = sizeof(struct btrfs_inode_ref);
1049 parent_objectid = key->offset;
1051 inode_objectid = key->objectid;
1054 * it is possible that we didn't log all the parent directories
1055 * for a given inode. If we don't find the dir, just don't
1056 * copy the back ref in. The link count fixup code will take
1059 dir = read_one_inode(root, parent_objectid);
1063 inode = read_one_inode(root, inode_objectid);
1069 while (ref_ptr < ref_end) {
1071 ret = extref_get_fields(eb, ref_ptr, &namelen, &name,
1072 &ref_index, &parent_objectid);
1074 * parent object can change from one array
1078 dir = read_one_inode(root, parent_objectid);
1082 ret = ref_get_fields(eb, ref_ptr, &namelen, &name,
1088 /* if we already have a perfect match, we're done */
1089 if (!inode_in_dir(root, path, btrfs_ino(dir), btrfs_ino(inode),
1090 ref_index, name, namelen)) {
1092 * look for a conflicting back reference in the
1093 * metadata. if we find one we have to unlink that name
1094 * of the file before we add our new link. Later on, we
1095 * overwrite any existing back reference, and we don't
1096 * want to create dangling pointers in the directory.
1100 ret = __add_inode_ref(trans, root, path, log,
1104 ref_index, name, namelen,
1111 /* insert our name */
1112 ret = btrfs_add_link(trans, dir, inode, name, namelen,
1116 btrfs_update_inode(trans, root, inode);
1119 ref_ptr = (unsigned long)(ref_ptr + ref_struct_size) + namelen;
1127 /* finally write the back reference in the inode */
1128 ret = overwrite_item(trans, root, path, eb, slot, key);
1132 btrfs_release_path(path);
1138 static int insert_orphan_item(struct btrfs_trans_handle *trans,
1139 struct btrfs_root *root, u64 offset)
1142 ret = btrfs_find_orphan_item(root, offset);
1144 ret = btrfs_insert_orphan_item(trans, root, offset);
1148 static int count_inode_extrefs(struct btrfs_root *root,
1149 struct inode *inode, struct btrfs_path *path)
1153 unsigned int nlink = 0;
1156 u64 inode_objectid = btrfs_ino(inode);
1159 struct btrfs_inode_extref *extref;
1160 struct extent_buffer *leaf;
1163 ret = btrfs_find_one_extref(root, inode_objectid, offset, path,
1168 leaf = path->nodes[0];
1169 item_size = btrfs_item_size_nr(leaf, path->slots[0]);
1170 ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
1172 while (cur_offset < item_size) {
1173 extref = (struct btrfs_inode_extref *) (ptr + cur_offset);
1174 name_len = btrfs_inode_extref_name_len(leaf, extref);
1178 cur_offset += name_len + sizeof(*extref);
1182 btrfs_release_path(path);
1184 btrfs_release_path(path);
1191 static int count_inode_refs(struct btrfs_root *root,
1192 struct inode *inode, struct btrfs_path *path)
1195 struct btrfs_key key;
1196 unsigned int nlink = 0;
1198 unsigned long ptr_end;
1200 u64 ino = btrfs_ino(inode);
1203 key.type = BTRFS_INODE_REF_KEY;
1204 key.offset = (u64)-1;
1207 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
1211 if (path->slots[0] == 0)
1215 btrfs_item_key_to_cpu(path->nodes[0], &key,
1217 if (key.objectid != ino ||
1218 key.type != BTRFS_INODE_REF_KEY)
1220 ptr = btrfs_item_ptr_offset(path->nodes[0], path->slots[0]);
1221 ptr_end = ptr + btrfs_item_size_nr(path->nodes[0],
1223 while (ptr < ptr_end) {
1224 struct btrfs_inode_ref *ref;
1226 ref = (struct btrfs_inode_ref *)ptr;
1227 name_len = btrfs_inode_ref_name_len(path->nodes[0],
1229 ptr = (unsigned long)(ref + 1) + name_len;
1233 if (key.offset == 0)
1236 btrfs_release_path(path);
1238 btrfs_release_path(path);
1244 * There are a few corners where the link count of the file can't
1245 * be properly maintained during replay. So, instead of adding
1246 * lots of complexity to the log code, we just scan the backrefs
1247 * for any file that has been through replay.
1249 * The scan will update the link count on the inode to reflect the
1250 * number of back refs found. If it goes down to zero, the iput
1251 * will free the inode.
1253 static noinline int fixup_inode_link_count(struct btrfs_trans_handle *trans,
1254 struct btrfs_root *root,
1255 struct inode *inode)
1257 struct btrfs_path *path;
1260 u64 ino = btrfs_ino(inode);
1262 path = btrfs_alloc_path();
1266 ret = count_inode_refs(root, inode, path);
1272 ret = count_inode_extrefs(root, inode, path);
1283 if (nlink != inode->i_nlink) {
1284 set_nlink(inode, nlink);
1285 btrfs_update_inode(trans, root, inode);
1287 BTRFS_I(inode)->index_cnt = (u64)-1;
1289 if (inode->i_nlink == 0) {
1290 if (S_ISDIR(inode->i_mode)) {
1291 ret = replay_dir_deletes(trans, root, NULL, path,
1295 ret = insert_orphan_item(trans, root, ino);
1300 btrfs_free_path(path);
1304 static noinline int fixup_inode_link_counts(struct btrfs_trans_handle *trans,
1305 struct btrfs_root *root,
1306 struct btrfs_path *path)
1309 struct btrfs_key key;
1310 struct inode *inode;
1312 key.objectid = BTRFS_TREE_LOG_FIXUP_OBJECTID;
1313 key.type = BTRFS_ORPHAN_ITEM_KEY;
1314 key.offset = (u64)-1;
1316 ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
1321 if (path->slots[0] == 0)
1326 btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
1327 if (key.objectid != BTRFS_TREE_LOG_FIXUP_OBJECTID ||
1328 key.type != BTRFS_ORPHAN_ITEM_KEY)
1331 ret = btrfs_del_item(trans, root, path);
1335 btrfs_release_path(path);
1336 inode = read_one_inode(root, key.offset);
1340 ret = fixup_inode_link_count(trans, root, inode);
1346 * fixup on a directory may create new entries,
1347 * make sure we always look for the highset possible
1350 key.offset = (u64)-1;
1354 btrfs_release_path(path);
1360 * record a given inode in the fixup dir so we can check its link
1361 * count when replay is done. The link count is incremented here
1362 * so the inode won't go away until we check it
1364 static noinline int link_to_fixup_dir(struct btrfs_trans_handle *trans,
1365 struct btrfs_root *root,
1366 struct btrfs_path *path,
1369 struct btrfs_key key;
1371 struct inode *inode;
1373 inode = read_one_inode(root, objectid);
1377 key.objectid = BTRFS_TREE_LOG_FIXUP_OBJECTID;
1378 btrfs_set_key_type(&key, BTRFS_ORPHAN_ITEM_KEY);
1379 key.offset = objectid;
1381 ret = btrfs_insert_empty_item(trans, root, path, &key, 0);
1383 btrfs_release_path(path);
1385 if (!inode->i_nlink)
1386 set_nlink(inode, 1);
1388 btrfs_inc_nlink(inode);
1389 ret = btrfs_update_inode(trans, root, inode);
1390 } else if (ret == -EEXIST) {
1401 * when replaying the log for a directory, we only insert names
1402 * for inodes that actually exist. This means an fsync on a directory
1403 * does not implicitly fsync all the new files in it
1405 static noinline int insert_one_name(struct btrfs_trans_handle *trans,
1406 struct btrfs_root *root,
1407 struct btrfs_path *path,
1408 u64 dirid, u64 index,
1409 char *name, int name_len, u8 type,
1410 struct btrfs_key *location)
1412 struct inode *inode;
1416 inode = read_one_inode(root, location->objectid);
1420 dir = read_one_inode(root, dirid);
1425 ret = btrfs_add_link(trans, dir, inode, name, name_len, 1, index);
1427 /* FIXME, put inode into FIXUP list */
1435 * take a single entry in a log directory item and replay it into
1438 * if a conflicting item exists in the subdirectory already,
1439 * the inode it points to is unlinked and put into the link count
1442 * If a name from the log points to a file or directory that does
1443 * not exist in the FS, it is skipped. fsyncs on directories
1444 * do not force down inodes inside that directory, just changes to the
1445 * names or unlinks in a directory.
1447 static noinline int replay_one_name(struct btrfs_trans_handle *trans,
1448 struct btrfs_root *root,
1449 struct btrfs_path *path,
1450 struct extent_buffer *eb,
1451 struct btrfs_dir_item *di,
1452 struct btrfs_key *key)
1456 struct btrfs_dir_item *dst_di;
1457 struct btrfs_key found_key;
1458 struct btrfs_key log_key;
1464 dir = read_one_inode(root, key->objectid);
1468 name_len = btrfs_dir_name_len(eb, di);
1469 name = kmalloc(name_len, GFP_NOFS);
1473 log_type = btrfs_dir_type(eb, di);
1474 read_extent_buffer(eb, name, (unsigned long)(di + 1),
1477 btrfs_dir_item_key_to_cpu(eb, di, &log_key);
1478 exists = btrfs_lookup_inode(trans, root, path, &log_key, 0);
1483 btrfs_release_path(path);
1485 if (key->type == BTRFS_DIR_ITEM_KEY) {
1486 dst_di = btrfs_lookup_dir_item(trans, root, path, key->objectid,
1488 } else if (key->type == BTRFS_DIR_INDEX_KEY) {
1489 dst_di = btrfs_lookup_dir_index_item(trans, root, path,
1496 if (IS_ERR_OR_NULL(dst_di)) {
1497 /* we need a sequence number to insert, so we only
1498 * do inserts for the BTRFS_DIR_INDEX_KEY types
1500 if (key->type != BTRFS_DIR_INDEX_KEY)
1505 btrfs_dir_item_key_to_cpu(path->nodes[0], dst_di, &found_key);
1506 /* the existing item matches the logged item */
1507 if (found_key.objectid == log_key.objectid &&
1508 found_key.type == log_key.type &&
1509 found_key.offset == log_key.offset &&
1510 btrfs_dir_type(path->nodes[0], dst_di) == log_type) {
1515 * don't drop the conflicting directory entry if the inode
1516 * for the new entry doesn't exist
1521 ret = drop_one_dir_item(trans, root, path, dir, dst_di);
1524 if (key->type == BTRFS_DIR_INDEX_KEY)
1527 btrfs_release_path(path);
1533 btrfs_release_path(path);
1534 ret = insert_one_name(trans, root, path, key->objectid, key->offset,
1535 name, name_len, log_type, &log_key);
1537 BUG_ON(ret && ret != -ENOENT);
1542 * find all the names in a directory item and reconcile them into
1543 * the subvolume. Only BTRFS_DIR_ITEM_KEY types will have more than
1544 * one name in a directory item, but the same code gets used for
1545 * both directory index types
1547 static noinline int replay_one_dir_item(struct btrfs_trans_handle *trans,
1548 struct btrfs_root *root,
1549 struct btrfs_path *path,
1550 struct extent_buffer *eb, int slot,
1551 struct btrfs_key *key)
1554 u32 item_size = btrfs_item_size_nr(eb, slot);
1555 struct btrfs_dir_item *di;
1558 unsigned long ptr_end;
1560 ptr = btrfs_item_ptr_offset(eb, slot);
1561 ptr_end = ptr + item_size;
1562 while (ptr < ptr_end) {
1563 di = (struct btrfs_dir_item *)ptr;
1564 if (verify_dir_item(root, eb, di))
1566 name_len = btrfs_dir_name_len(eb, di);
1567 ret = replay_one_name(trans, root, path, eb, di, key);
1569 ptr = (unsigned long)(di + 1);
1576 * directory replay has two parts. There are the standard directory
1577 * items in the log copied from the subvolume, and range items
1578 * created in the log while the subvolume was logged.
1580 * The range items tell us which parts of the key space the log
1581 * is authoritative for. During replay, if a key in the subvolume
1582 * directory is in a logged range item, but not actually in the log
1583 * that means it was deleted from the directory before the fsync
1584 * and should be removed.
1586 static noinline int find_dir_range(struct btrfs_root *root,
1587 struct btrfs_path *path,
1588 u64 dirid, int key_type,
1589 u64 *start_ret, u64 *end_ret)
1591 struct btrfs_key key;
1593 struct btrfs_dir_log_item *item;
1597 if (*start_ret == (u64)-1)
1600 key.objectid = dirid;
1601 key.type = key_type;
1602 key.offset = *start_ret;
1604 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
1608 if (path->slots[0] == 0)
1613 btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
1615 if (key.type != key_type || key.objectid != dirid) {
1619 item = btrfs_item_ptr(path->nodes[0], path->slots[0],
1620 struct btrfs_dir_log_item);
1621 found_end = btrfs_dir_log_end(path->nodes[0], item);
1623 if (*start_ret >= key.offset && *start_ret <= found_end) {
1625 *start_ret = key.offset;
1626 *end_ret = found_end;
1631 /* check the next slot in the tree to see if it is a valid item */
1632 nritems = btrfs_header_nritems(path->nodes[0]);
1633 if (path->slots[0] >= nritems) {
1634 ret = btrfs_next_leaf(root, path);
1641 btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
1643 if (key.type != key_type || key.objectid != dirid) {
1647 item = btrfs_item_ptr(path->nodes[0], path->slots[0],
1648 struct btrfs_dir_log_item);
1649 found_end = btrfs_dir_log_end(path->nodes[0], item);
1650 *start_ret = key.offset;
1651 *end_ret = found_end;
1654 btrfs_release_path(path);
1659 * this looks for a given directory item in the log. If the directory
1660 * item is not in the log, the item is removed and the inode it points
1663 static noinline int check_item_in_log(struct btrfs_trans_handle *trans,
1664 struct btrfs_root *root,
1665 struct btrfs_root *log,
1666 struct btrfs_path *path,
1667 struct btrfs_path *log_path,
1669 struct btrfs_key *dir_key)
1672 struct extent_buffer *eb;
1675 struct btrfs_dir_item *di;
1676 struct btrfs_dir_item *log_di;
1679 unsigned long ptr_end;
1681 struct inode *inode;
1682 struct btrfs_key location;
1685 eb = path->nodes[0];
1686 slot = path->slots[0];
1687 item_size = btrfs_item_size_nr(eb, slot);
1688 ptr = btrfs_item_ptr_offset(eb, slot);
1689 ptr_end = ptr + item_size;
1690 while (ptr < ptr_end) {
1691 di = (struct btrfs_dir_item *)ptr;
1692 if (verify_dir_item(root, eb, di)) {
1697 name_len = btrfs_dir_name_len(eb, di);
1698 name = kmalloc(name_len, GFP_NOFS);
1703 read_extent_buffer(eb, name, (unsigned long)(di + 1),
1706 if (log && dir_key->type == BTRFS_DIR_ITEM_KEY) {
1707 log_di = btrfs_lookup_dir_item(trans, log, log_path,
1710 } else if (log && dir_key->type == BTRFS_DIR_INDEX_KEY) {
1711 log_di = btrfs_lookup_dir_index_item(trans, log,
1717 if (IS_ERR_OR_NULL(log_di)) {
1718 btrfs_dir_item_key_to_cpu(eb, di, &location);
1719 btrfs_release_path(path);
1720 btrfs_release_path(log_path);
1721 inode = read_one_inode(root, location.objectid);
1727 ret = link_to_fixup_dir(trans, root,
1728 path, location.objectid);
1730 btrfs_inc_nlink(inode);
1731 ret = btrfs_unlink_inode(trans, root, dir, inode,
1735 btrfs_run_delayed_items(trans, root);
1740 /* there might still be more names under this key
1741 * check and repeat if required
1743 ret = btrfs_search_slot(NULL, root, dir_key, path,
1750 btrfs_release_path(log_path);
1753 ptr = (unsigned long)(di + 1);
1758 btrfs_release_path(path);
1759 btrfs_release_path(log_path);
1764 * deletion replay happens before we copy any new directory items
1765 * out of the log or out of backreferences from inodes. It
1766 * scans the log to find ranges of keys that log is authoritative for,
1767 * and then scans the directory to find items in those ranges that are
1768 * not present in the log.
1770 * Anything we don't find in the log is unlinked and removed from the
1773 static noinline int replay_dir_deletes(struct btrfs_trans_handle *trans,
1774 struct btrfs_root *root,
1775 struct btrfs_root *log,
1776 struct btrfs_path *path,
1777 u64 dirid, int del_all)
1781 int key_type = BTRFS_DIR_LOG_ITEM_KEY;
1783 struct btrfs_key dir_key;
1784 struct btrfs_key found_key;
1785 struct btrfs_path *log_path;
1788 dir_key.objectid = dirid;
1789 dir_key.type = BTRFS_DIR_ITEM_KEY;
1790 log_path = btrfs_alloc_path();
1794 dir = read_one_inode(root, dirid);
1795 /* it isn't an error if the inode isn't there, that can happen
1796 * because we replay the deletes before we copy in the inode item
1800 btrfs_free_path(log_path);
1808 range_end = (u64)-1;
1810 ret = find_dir_range(log, path, dirid, key_type,
1811 &range_start, &range_end);
1816 dir_key.offset = range_start;
1819 ret = btrfs_search_slot(NULL, root, &dir_key, path,
1824 nritems = btrfs_header_nritems(path->nodes[0]);
1825 if (path->slots[0] >= nritems) {
1826 ret = btrfs_next_leaf(root, path);
1830 btrfs_item_key_to_cpu(path->nodes[0], &found_key,
1832 if (found_key.objectid != dirid ||
1833 found_key.type != dir_key.type)
1836 if (found_key.offset > range_end)
1839 ret = check_item_in_log(trans, root, log, path,
1843 if (found_key.offset == (u64)-1)
1845 dir_key.offset = found_key.offset + 1;
1847 btrfs_release_path(path);
1848 if (range_end == (u64)-1)
1850 range_start = range_end + 1;
1855 if (key_type == BTRFS_DIR_LOG_ITEM_KEY) {
1856 key_type = BTRFS_DIR_LOG_INDEX_KEY;
1857 dir_key.type = BTRFS_DIR_INDEX_KEY;
1858 btrfs_release_path(path);
1862 btrfs_release_path(path);
1863 btrfs_free_path(log_path);
1869 * the process_func used to replay items from the log tree. This
1870 * gets called in two different stages. The first stage just looks
1871 * for inodes and makes sure they are all copied into the subvolume.
1873 * The second stage copies all the other item types from the log into
1874 * the subvolume. The two stage approach is slower, but gets rid of
1875 * lots of complexity around inodes referencing other inodes that exist
1876 * only in the log (references come from either directory items or inode
1879 static int replay_one_buffer(struct btrfs_root *log, struct extent_buffer *eb,
1880 struct walk_control *wc, u64 gen)
1883 struct btrfs_path *path;
1884 struct btrfs_root *root = wc->replay_dest;
1885 struct btrfs_key key;
1890 ret = btrfs_read_buffer(eb, gen);
1894 level = btrfs_header_level(eb);
1899 path = btrfs_alloc_path();
1903 nritems = btrfs_header_nritems(eb);
1904 for (i = 0; i < nritems; i++) {
1905 btrfs_item_key_to_cpu(eb, &key, i);
1907 /* inode keys are done during the first stage */
1908 if (key.type == BTRFS_INODE_ITEM_KEY &&
1909 wc->stage == LOG_WALK_REPLAY_INODES) {
1910 struct btrfs_inode_item *inode_item;
1913 inode_item = btrfs_item_ptr(eb, i,
1914 struct btrfs_inode_item);
1915 mode = btrfs_inode_mode(eb, inode_item);
1916 if (S_ISDIR(mode)) {
1917 ret = replay_dir_deletes(wc->trans,
1918 root, log, path, key.objectid, 0);
1921 ret = overwrite_item(wc->trans, root, path,
1925 /* for regular files, make sure corresponding
1926 * orhpan item exist. extents past the new EOF
1927 * will be truncated later by orphan cleanup.
1929 if (S_ISREG(mode)) {
1930 ret = insert_orphan_item(wc->trans, root,
1935 ret = link_to_fixup_dir(wc->trans, root,
1936 path, key.objectid);
1939 if (wc->stage < LOG_WALK_REPLAY_ALL)
1942 /* these keys are simply copied */
1943 if (key.type == BTRFS_XATTR_ITEM_KEY) {
1944 ret = overwrite_item(wc->trans, root, path,
1947 } else if (key.type == BTRFS_INODE_REF_KEY) {
1948 ret = add_inode_ref(wc->trans, root, log, path,
1950 BUG_ON(ret && ret != -ENOENT);
1951 } else if (key.type == BTRFS_INODE_EXTREF_KEY) {
1952 ret = add_inode_ref(wc->trans, root, log, path,
1954 BUG_ON(ret && ret != -ENOENT);
1955 } else if (key.type == BTRFS_EXTENT_DATA_KEY) {
1956 ret = replay_one_extent(wc->trans, root, path,
1959 } else if (key.type == BTRFS_DIR_ITEM_KEY ||
1960 key.type == BTRFS_DIR_INDEX_KEY) {
1961 ret = replay_one_dir_item(wc->trans, root, path,
1966 btrfs_free_path(path);
1970 static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans,
1971 struct btrfs_root *root,
1972 struct btrfs_path *path, int *level,
1973 struct walk_control *wc)
1978 struct extent_buffer *next;
1979 struct extent_buffer *cur;
1980 struct extent_buffer *parent;
1984 WARN_ON(*level < 0);
1985 WARN_ON(*level >= BTRFS_MAX_LEVEL);
1987 while (*level > 0) {
1988 WARN_ON(*level < 0);
1989 WARN_ON(*level >= BTRFS_MAX_LEVEL);
1990 cur = path->nodes[*level];
1992 if (btrfs_header_level(cur) != *level)
1995 if (path->slots[*level] >=
1996 btrfs_header_nritems(cur))
1999 bytenr = btrfs_node_blockptr(cur, path->slots[*level]);
2000 ptr_gen = btrfs_node_ptr_generation(cur, path->slots[*level]);
2001 blocksize = btrfs_level_size(root, *level - 1);
2003 parent = path->nodes[*level];
2004 root_owner = btrfs_header_owner(parent);
2006 next = btrfs_find_create_tree_block(root, bytenr, blocksize);
2011 ret = wc->process_func(root, next, wc, ptr_gen);
2015 path->slots[*level]++;
2017 ret = btrfs_read_buffer(next, ptr_gen);
2019 free_extent_buffer(next);
2023 btrfs_tree_lock(next);
2024 btrfs_set_lock_blocking(next);
2025 clean_tree_block(trans, root, next);
2026 btrfs_wait_tree_block_writeback(next);
2027 btrfs_tree_unlock(next);
2029 WARN_ON(root_owner !=
2030 BTRFS_TREE_LOG_OBJECTID);
2031 ret = btrfs_free_and_pin_reserved_extent(root,
2033 BUG_ON(ret); /* -ENOMEM or logic errors */
2035 free_extent_buffer(next);
2038 ret = btrfs_read_buffer(next, ptr_gen);
2040 free_extent_buffer(next);
2044 WARN_ON(*level <= 0);
2045 if (path->nodes[*level-1])
2046 free_extent_buffer(path->nodes[*level-1]);
2047 path->nodes[*level-1] = next;
2048 *level = btrfs_header_level(next);
2049 path->slots[*level] = 0;
2052 WARN_ON(*level < 0);
2053 WARN_ON(*level >= BTRFS_MAX_LEVEL);
2055 path->slots[*level] = btrfs_header_nritems(path->nodes[*level]);
2061 static noinline int walk_up_log_tree(struct btrfs_trans_handle *trans,
2062 struct btrfs_root *root,
2063 struct btrfs_path *path, int *level,
2064 struct walk_control *wc)
2071 for (i = *level; i < BTRFS_MAX_LEVEL - 1 && path->nodes[i]; i++) {
2072 slot = path->slots[i];
2073 if (slot + 1 < btrfs_header_nritems(path->nodes[i])) {
2076 WARN_ON(*level == 0);
2079 struct extent_buffer *parent;
2080 if (path->nodes[*level] == root->node)
2081 parent = path->nodes[*level];
2083 parent = path->nodes[*level + 1];
2085 root_owner = btrfs_header_owner(parent);
2086 ret = wc->process_func(root, path->nodes[*level], wc,
2087 btrfs_header_generation(path->nodes[*level]));
2092 struct extent_buffer *next;
2094 next = path->nodes[*level];
2096 btrfs_tree_lock(next);
2097 btrfs_set_lock_blocking(next);
2098 clean_tree_block(trans, root, next);
2099 btrfs_wait_tree_block_writeback(next);
2100 btrfs_tree_unlock(next);
2102 WARN_ON(root_owner != BTRFS_TREE_LOG_OBJECTID);
2103 ret = btrfs_free_and_pin_reserved_extent(root,
2104 path->nodes[*level]->start,
2105 path->nodes[*level]->len);
2108 free_extent_buffer(path->nodes[*level]);
2109 path->nodes[*level] = NULL;
2117 * drop the reference count on the tree rooted at 'snap'. This traverses
2118 * the tree freeing any blocks that have a ref count of zero after being
2121 static int walk_log_tree(struct btrfs_trans_handle *trans,
2122 struct btrfs_root *log, struct walk_control *wc)
2127 struct btrfs_path *path;
2131 path = btrfs_alloc_path();
2135 level = btrfs_header_level(log->node);
2137 path->nodes[level] = log->node;
2138 extent_buffer_get(log->node);
2139 path->slots[level] = 0;
2142 wret = walk_down_log_tree(trans, log, path, &level, wc);
2150 wret = walk_up_log_tree(trans, log, path, &level, wc);
2159 /* was the root node processed? if not, catch it here */
2160 if (path->nodes[orig_level]) {
2161 ret = wc->process_func(log, path->nodes[orig_level], wc,
2162 btrfs_header_generation(path->nodes[orig_level]));
2166 struct extent_buffer *next;
2168 next = path->nodes[orig_level];
2170 btrfs_tree_lock(next);
2171 btrfs_set_lock_blocking(next);
2172 clean_tree_block(trans, log, next);
2173 btrfs_wait_tree_block_writeback(next);
2174 btrfs_tree_unlock(next);
2176 WARN_ON(log->root_key.objectid !=
2177 BTRFS_TREE_LOG_OBJECTID);
2178 ret = btrfs_free_and_pin_reserved_extent(log, next->start,
2180 BUG_ON(ret); /* -ENOMEM or logic errors */
2185 for (i = 0; i <= orig_level; i++) {
2186 if (path->nodes[i]) {
2187 free_extent_buffer(path->nodes[i]);
2188 path->nodes[i] = NULL;
2191 btrfs_free_path(path);
2196 * helper function to update the item for a given subvolumes log root
2197 * in the tree of log roots
2199 static int update_log_root(struct btrfs_trans_handle *trans,
2200 struct btrfs_root *log)
2204 if (log->log_transid == 1) {
2205 /* insert root item on the first sync */
2206 ret = btrfs_insert_root(trans, log->fs_info->log_root_tree,
2207 &log->root_key, &log->root_item);
2209 ret = btrfs_update_root(trans, log->fs_info->log_root_tree,
2210 &log->root_key, &log->root_item);
2215 static int wait_log_commit(struct btrfs_trans_handle *trans,
2216 struct btrfs_root *root, unsigned long transid)
2219 int index = transid % 2;
2222 * we only allow two pending log transactions at a time,
2223 * so we know that if ours is more than 2 older than the
2224 * current transaction, we're done
2227 prepare_to_wait(&root->log_commit_wait[index],
2228 &wait, TASK_UNINTERRUPTIBLE);
2229 mutex_unlock(&root->log_mutex);
2231 if (root->fs_info->last_trans_log_full_commit !=
2232 trans->transid && root->log_transid < transid + 2 &&
2233 atomic_read(&root->log_commit[index]))
2236 finish_wait(&root->log_commit_wait[index], &wait);
2237 mutex_lock(&root->log_mutex);
2238 } while (root->fs_info->last_trans_log_full_commit !=
2239 trans->transid && root->log_transid < transid + 2 &&
2240 atomic_read(&root->log_commit[index]));
2244 static void wait_for_writer(struct btrfs_trans_handle *trans,
2245 struct btrfs_root *root)
2248 while (root->fs_info->last_trans_log_full_commit !=
2249 trans->transid && atomic_read(&root->log_writers)) {
2250 prepare_to_wait(&root->log_writer_wait,
2251 &wait, TASK_UNINTERRUPTIBLE);
2252 mutex_unlock(&root->log_mutex);
2253 if (root->fs_info->last_trans_log_full_commit !=
2254 trans->transid && atomic_read(&root->log_writers))
2256 mutex_lock(&root->log_mutex);
2257 finish_wait(&root->log_writer_wait, &wait);
2262 * btrfs_sync_log does sends a given tree log down to the disk and
2263 * updates the super blocks to record it. When this call is done,
2264 * you know that any inodes previously logged are safely on disk only
2267 * Any other return value means you need to call btrfs_commit_transaction.
2268 * Some of the edge cases for fsyncing directories that have had unlinks
2269 * or renames done in the past mean that sometimes the only safe
2270 * fsync is to commit the whole FS. When btrfs_sync_log returns -EAGAIN,
2271 * that has happened.
2273 int btrfs_sync_log(struct btrfs_trans_handle *trans,
2274 struct btrfs_root *root)
2280 struct btrfs_root *log = root->log_root;
2281 struct btrfs_root *log_root_tree = root->fs_info->log_root_tree;
2282 unsigned long log_transid = 0;
2284 mutex_lock(&root->log_mutex);
2285 log_transid = root->log_transid;
2286 index1 = root->log_transid % 2;
2287 if (atomic_read(&root->log_commit[index1])) {
2288 wait_log_commit(trans, root, root->log_transid);
2289 mutex_unlock(&root->log_mutex);
2292 atomic_set(&root->log_commit[index1], 1);
2294 /* wait for previous tree log sync to complete */
2295 if (atomic_read(&root->log_commit[(index1 + 1) % 2]))
2296 wait_log_commit(trans, root, root->log_transid - 1);
2298 int batch = atomic_read(&root->log_batch);
2299 /* when we're on an ssd, just kick the log commit out */
2300 if (!btrfs_test_opt(root, SSD) && root->log_multiple_pids) {
2301 mutex_unlock(&root->log_mutex);
2302 schedule_timeout_uninterruptible(1);
2303 mutex_lock(&root->log_mutex);
2305 wait_for_writer(trans, root);
2306 if (batch == atomic_read(&root->log_batch))
2310 /* bail out if we need to do a full commit */
2311 if (root->fs_info->last_trans_log_full_commit == trans->transid) {
2313 btrfs_free_logged_extents(log, log_transid);
2314 mutex_unlock(&root->log_mutex);
2318 if (log_transid % 2 == 0)
2319 mark = EXTENT_DIRTY;
2323 /* we start IO on all the marked extents here, but we don't actually
2324 * wait for them until later.
2326 ret = btrfs_write_marked_extents(log, &log->dirty_log_pages, mark);
2328 btrfs_abort_transaction(trans, root, ret);
2329 btrfs_free_logged_extents(log, log_transid);
2330 mutex_unlock(&root->log_mutex);
2334 btrfs_set_root_node(&log->root_item, log->node);
2336 root->log_transid++;
2337 log->log_transid = root->log_transid;
2338 root->log_start_pid = 0;
2341 * IO has been started, blocks of the log tree have WRITTEN flag set
2342 * in their headers. new modifications of the log will be written to
2343 * new positions. so it's safe to allow log writers to go in.
2345 mutex_unlock(&root->log_mutex);
2347 mutex_lock(&log_root_tree->log_mutex);
2348 atomic_inc(&log_root_tree->log_batch);
2349 atomic_inc(&log_root_tree->log_writers);
2350 mutex_unlock(&log_root_tree->log_mutex);
2352 ret = update_log_root(trans, log);
2354 mutex_lock(&log_root_tree->log_mutex);
2355 if (atomic_dec_and_test(&log_root_tree->log_writers)) {
2357 if (waitqueue_active(&log_root_tree->log_writer_wait))
2358 wake_up(&log_root_tree->log_writer_wait);
2362 if (ret != -ENOSPC) {
2363 btrfs_abort_transaction(trans, root, ret);
2364 mutex_unlock(&log_root_tree->log_mutex);
2367 root->fs_info->last_trans_log_full_commit = trans->transid;
2368 btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark);
2369 btrfs_free_logged_extents(log, log_transid);
2370 mutex_unlock(&log_root_tree->log_mutex);
2375 index2 = log_root_tree->log_transid % 2;
2376 if (atomic_read(&log_root_tree->log_commit[index2])) {
2377 btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark);
2378 wait_log_commit(trans, log_root_tree,
2379 log_root_tree->log_transid);
2380 btrfs_free_logged_extents(log, log_transid);
2381 mutex_unlock(&log_root_tree->log_mutex);
2385 atomic_set(&log_root_tree->log_commit[index2], 1);
2387 if (atomic_read(&log_root_tree->log_commit[(index2 + 1) % 2])) {
2388 wait_log_commit(trans, log_root_tree,
2389 log_root_tree->log_transid - 1);
2392 wait_for_writer(trans, log_root_tree);
2395 * now that we've moved on to the tree of log tree roots,
2396 * check the full commit flag again
2398 if (root->fs_info->last_trans_log_full_commit == trans->transid) {
2399 btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark);
2400 btrfs_free_logged_extents(log, log_transid);
2401 mutex_unlock(&log_root_tree->log_mutex);
2403 goto out_wake_log_root;
2406 ret = btrfs_write_and_wait_marked_extents(log_root_tree,
2407 &log_root_tree->dirty_log_pages,
2408 EXTENT_DIRTY | EXTENT_NEW);
2410 btrfs_abort_transaction(trans, root, ret);
2411 btrfs_free_logged_extents(log, log_transid);
2412 mutex_unlock(&log_root_tree->log_mutex);
2413 goto out_wake_log_root;
2415 btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark);
2416 btrfs_wait_logged_extents(log, log_transid);
2418 btrfs_set_super_log_root(root->fs_info->super_for_commit,
2419 log_root_tree->node->start);
2420 btrfs_set_super_log_root_level(root->fs_info->super_for_commit,
2421 btrfs_header_level(log_root_tree->node));
2423 log_root_tree->log_transid++;
2426 mutex_unlock(&log_root_tree->log_mutex);
2429 * nobody else is going to jump in and write the the ctree
2430 * super here because the log_commit atomic below is protecting
2431 * us. We must be called with a transaction handle pinning
2432 * the running transaction open, so a full commit can't hop
2433 * in and cause problems either.
2435 btrfs_scrub_pause_super(root);
2436 ret = write_ctree_super(trans, root->fs_info->tree_root, 1);
2437 btrfs_scrub_continue_super(root);
2439 btrfs_abort_transaction(trans, root, ret);
2440 goto out_wake_log_root;
2443 mutex_lock(&root->log_mutex);
2444 if (root->last_log_commit < log_transid)
2445 root->last_log_commit = log_transid;
2446 mutex_unlock(&root->log_mutex);
2449 atomic_set(&log_root_tree->log_commit[index2], 0);
2451 if (waitqueue_active(&log_root_tree->log_commit_wait[index2]))
2452 wake_up(&log_root_tree->log_commit_wait[index2]);
2454 atomic_set(&root->log_commit[index1], 0);
2456 if (waitqueue_active(&root->log_commit_wait[index1]))
2457 wake_up(&root->log_commit_wait[index1]);
2461 static void free_log_tree(struct btrfs_trans_handle *trans,
2462 struct btrfs_root *log)
2467 struct walk_control wc = {
2469 .process_func = process_one_buffer
2473 ret = walk_log_tree(trans, log, &wc);
2478 ret = find_first_extent_bit(&log->dirty_log_pages,
2479 0, &start, &end, EXTENT_DIRTY | EXTENT_NEW,
2484 clear_extent_bits(&log->dirty_log_pages, start, end,
2485 EXTENT_DIRTY | EXTENT_NEW, GFP_NOFS);
2489 * We may have short-circuited the log tree with the full commit logic
2490 * and left ordered extents on our list, so clear these out to keep us
2491 * from leaking inodes and memory.
2493 btrfs_free_logged_extents(log, 0);
2494 btrfs_free_logged_extents(log, 1);
2496 free_extent_buffer(log->node);
2501 * free all the extents used by the tree log. This should be called
2502 * at commit time of the full transaction
2504 int btrfs_free_log(struct btrfs_trans_handle *trans, struct btrfs_root *root)
2506 if (root->log_root) {
2507 free_log_tree(trans, root->log_root);
2508 root->log_root = NULL;
2513 int btrfs_free_log_root_tree(struct btrfs_trans_handle *trans,
2514 struct btrfs_fs_info *fs_info)
2516 if (fs_info->log_root_tree) {
2517 free_log_tree(trans, fs_info->log_root_tree);
2518 fs_info->log_root_tree = NULL;
2524 * If both a file and directory are logged, and unlinks or renames are
2525 * mixed in, we have a few interesting corners:
2527 * create file X in dir Y
2528 * link file X to X.link in dir Y
2530 * unlink file X but leave X.link
2533 * After a crash we would expect only X.link to exist. But file X
2534 * didn't get fsync'd again so the log has back refs for X and X.link.
2536 * We solve this by removing directory entries and inode backrefs from the
2537 * log when a file that was logged in the current transaction is
2538 * unlinked. Any later fsync will include the updated log entries, and
2539 * we'll be able to reconstruct the proper directory items from backrefs.
2541 * This optimizations allows us to avoid relogging the entire inode
2542 * or the entire directory.
2544 int btrfs_del_dir_entries_in_log(struct btrfs_trans_handle *trans,
2545 struct btrfs_root *root,
2546 const char *name, int name_len,
2547 struct inode *dir, u64 index)
2549 struct btrfs_root *log;
2550 struct btrfs_dir_item *di;
2551 struct btrfs_path *path;
2555 u64 dir_ino = btrfs_ino(dir);
2557 if (BTRFS_I(dir)->logged_trans < trans->transid)
2560 ret = join_running_log_trans(root);
2564 mutex_lock(&BTRFS_I(dir)->log_mutex);
2566 log = root->log_root;
2567 path = btrfs_alloc_path();
2573 di = btrfs_lookup_dir_item(trans, log, path, dir_ino,
2574 name, name_len, -1);
2580 ret = btrfs_delete_one_dir_name(trans, log, path, di);
2581 bytes_del += name_len;
2584 btrfs_release_path(path);
2585 di = btrfs_lookup_dir_index_item(trans, log, path, dir_ino,
2586 index, name, name_len, -1);
2592 ret = btrfs_delete_one_dir_name(trans, log, path, di);
2593 bytes_del += name_len;
2597 /* update the directory size in the log to reflect the names
2601 struct btrfs_key key;
2603 key.objectid = dir_ino;
2605 key.type = BTRFS_INODE_ITEM_KEY;
2606 btrfs_release_path(path);
2608 ret = btrfs_search_slot(trans, log, &key, path, 0, 1);
2614 struct btrfs_inode_item *item;
2617 item = btrfs_item_ptr(path->nodes[0], path->slots[0],
2618 struct btrfs_inode_item);
2619 i_size = btrfs_inode_size(path->nodes[0], item);
2620 if (i_size > bytes_del)
2621 i_size -= bytes_del;
2624 btrfs_set_inode_size(path->nodes[0], item, i_size);
2625 btrfs_mark_buffer_dirty(path->nodes[0]);
2628 btrfs_release_path(path);
2631 btrfs_free_path(path);
2633 mutex_unlock(&BTRFS_I(dir)->log_mutex);
2634 if (ret == -ENOSPC) {
2635 root->fs_info->last_trans_log_full_commit = trans->transid;
2638 btrfs_abort_transaction(trans, root, ret);
2640 btrfs_end_log_trans(root);
2645 /* see comments for btrfs_del_dir_entries_in_log */
2646 int btrfs_del_inode_ref_in_log(struct btrfs_trans_handle *trans,
2647 struct btrfs_root *root,
2648 const char *name, int name_len,
2649 struct inode *inode, u64 dirid)
2651 struct btrfs_root *log;
2655 if (BTRFS_I(inode)->logged_trans < trans->transid)
2658 ret = join_running_log_trans(root);
2661 log = root->log_root;
2662 mutex_lock(&BTRFS_I(inode)->log_mutex);
2664 ret = btrfs_del_inode_ref(trans, log, name, name_len, btrfs_ino(inode),
2666 mutex_unlock(&BTRFS_I(inode)->log_mutex);
2667 if (ret == -ENOSPC) {
2668 root->fs_info->last_trans_log_full_commit = trans->transid;
2670 } else if (ret < 0 && ret != -ENOENT)
2671 btrfs_abort_transaction(trans, root, ret);
2672 btrfs_end_log_trans(root);
2678 * creates a range item in the log for 'dirid'. first_offset and
2679 * last_offset tell us which parts of the key space the log should
2680 * be considered authoritative for.
2682 static noinline int insert_dir_log_key(struct btrfs_trans_handle *trans,
2683 struct btrfs_root *log,
2684 struct btrfs_path *path,
2685 int key_type, u64 dirid,
2686 u64 first_offset, u64 last_offset)
2689 struct btrfs_key key;
2690 struct btrfs_dir_log_item *item;
2692 key.objectid = dirid;
2693 key.offset = first_offset;
2694 if (key_type == BTRFS_DIR_ITEM_KEY)
2695 key.type = BTRFS_DIR_LOG_ITEM_KEY;
2697 key.type = BTRFS_DIR_LOG_INDEX_KEY;
2698 ret = btrfs_insert_empty_item(trans, log, path, &key, sizeof(*item));
2702 item = btrfs_item_ptr(path->nodes[0], path->slots[0],
2703 struct btrfs_dir_log_item);
2704 btrfs_set_dir_log_end(path->nodes[0], item, last_offset);
2705 btrfs_mark_buffer_dirty(path->nodes[0]);
2706 btrfs_release_path(path);
2711 * log all the items included in the current transaction for a given
2712 * directory. This also creates the range items in the log tree required
2713 * to replay anything deleted before the fsync
2715 static noinline int log_dir_items(struct btrfs_trans_handle *trans,
2716 struct btrfs_root *root, struct inode *inode,
2717 struct btrfs_path *path,
2718 struct btrfs_path *dst_path, int key_type,
2719 u64 min_offset, u64 *last_offset_ret)
2721 struct btrfs_key min_key;
2722 struct btrfs_key max_key;
2723 struct btrfs_root *log = root->log_root;
2724 struct extent_buffer *src;
2729 u64 first_offset = min_offset;
2730 u64 last_offset = (u64)-1;
2731 u64 ino = btrfs_ino(inode);
2733 log = root->log_root;
2734 max_key.objectid = ino;
2735 max_key.offset = (u64)-1;
2736 max_key.type = key_type;
2738 min_key.objectid = ino;
2739 min_key.type = key_type;
2740 min_key.offset = min_offset;
2742 path->keep_locks = 1;
2744 ret = btrfs_search_forward(root, &min_key, &max_key,
2745 path, trans->transid);
2748 * we didn't find anything from this transaction, see if there
2749 * is anything at all
2751 if (ret != 0 || min_key.objectid != ino || min_key.type != key_type) {
2752 min_key.objectid = ino;
2753 min_key.type = key_type;
2754 min_key.offset = (u64)-1;
2755 btrfs_release_path(path);
2756 ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0);
2758 btrfs_release_path(path);
2761 ret = btrfs_previous_item(root, path, ino, key_type);
2763 /* if ret == 0 there are items for this type,
2764 * create a range to tell us the last key of this type.
2765 * otherwise, there are no items in this directory after
2766 * *min_offset, and we create a range to indicate that.
2769 struct btrfs_key tmp;
2770 btrfs_item_key_to_cpu(path->nodes[0], &tmp,
2772 if (key_type == tmp.type)
2773 first_offset = max(min_offset, tmp.offset) + 1;
2778 /* go backward to find any previous key */
2779 ret = btrfs_previous_item(root, path, ino, key_type);
2781 struct btrfs_key tmp;
2782 btrfs_item_key_to_cpu(path->nodes[0], &tmp, path->slots[0]);
2783 if (key_type == tmp.type) {
2784 first_offset = tmp.offset;
2785 ret = overwrite_item(trans, log, dst_path,
2786 path->nodes[0], path->slots[0],
2794 btrfs_release_path(path);
2796 /* find the first key from this transaction again */
2797 ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0);
2804 * we have a block from this transaction, log every item in it
2805 * from our directory
2808 struct btrfs_key tmp;
2809 src = path->nodes[0];
2810 nritems = btrfs_header_nritems(src);
2811 for (i = path->slots[0]; i < nritems; i++) {
2812 btrfs_item_key_to_cpu(src, &min_key, i);
2814 if (min_key.objectid != ino || min_key.type != key_type)
2816 ret = overwrite_item(trans, log, dst_path, src, i,
2823 path->slots[0] = nritems;
2826 * look ahead to the next item and see if it is also
2827 * from this directory and from this transaction
2829 ret = btrfs_next_leaf(root, path);
2831 last_offset = (u64)-1;
2834 btrfs_item_key_to_cpu(path->nodes[0], &tmp, path->slots[0]);
2835 if (tmp.objectid != ino || tmp.type != key_type) {
2836 last_offset = (u64)-1;
2839 if (btrfs_header_generation(path->nodes[0]) != trans->transid) {
2840 ret = overwrite_item(trans, log, dst_path,
2841 path->nodes[0], path->slots[0],
2846 last_offset = tmp.offset;
2851 btrfs_release_path(path);
2852 btrfs_release_path(dst_path);
2855 *last_offset_ret = last_offset;
2857 * insert the log range keys to indicate where the log
2860 ret = insert_dir_log_key(trans, log, path, key_type,
2861 ino, first_offset, last_offset);
2869 * logging directories is very similar to logging inodes, We find all the items
2870 * from the current transaction and write them to the log.
2872 * The recovery code scans the directory in the subvolume, and if it finds a
2873 * key in the range logged that is not present in the log tree, then it means
2874 * that dir entry was unlinked during the transaction.
2876 * In order for that scan to work, we must include one key smaller than
2877 * the smallest logged by this transaction and one key larger than the largest
2878 * key logged by this transaction.
2880 static noinline int log_directory_changes(struct btrfs_trans_handle *trans,
2881 struct btrfs_root *root, struct inode *inode,
2882 struct btrfs_path *path,
2883 struct btrfs_path *dst_path)
2888 int key_type = BTRFS_DIR_ITEM_KEY;
2894 ret = log_dir_items(trans, root, inode, path,
2895 dst_path, key_type, min_key,
2899 if (max_key == (u64)-1)
2901 min_key = max_key + 1;
2904 if (key_type == BTRFS_DIR_ITEM_KEY) {
2905 key_type = BTRFS_DIR_INDEX_KEY;
2912 * a helper function to drop items from the log before we relog an
2913 * inode. max_key_type indicates the highest item type to remove.
2914 * This cannot be run for file data extents because it does not
2915 * free the extents they point to.
2917 static int drop_objectid_items(struct btrfs_trans_handle *trans,
2918 struct btrfs_root *log,
2919 struct btrfs_path *path,
2920 u64 objectid, int max_key_type)
2923 struct btrfs_key key;
2924 struct btrfs_key found_key;
2927 key.objectid = objectid;
2928 key.type = max_key_type;
2929 key.offset = (u64)-1;
2932 ret = btrfs_search_slot(trans, log, &key, path, -1, 1);
2937 if (path->slots[0] == 0)
2941 btrfs_item_key_to_cpu(path->nodes[0], &found_key,
2944 if (found_key.objectid != objectid)
2947 found_key.offset = 0;
2949 ret = btrfs_bin_search(path->nodes[0], &found_key, 0,
2952 ret = btrfs_del_items(trans, log, path, start_slot,
2953 path->slots[0] - start_slot + 1);
2955 * If start slot isn't 0 then we don't need to re-search, we've
2956 * found the last guy with the objectid in this tree.
2958 if (ret || start_slot != 0)
2960 btrfs_release_path(path);
2962 btrfs_release_path(path);
2968 static void fill_inode_item(struct btrfs_trans_handle *trans,
2969 struct extent_buffer *leaf,
2970 struct btrfs_inode_item *item,
2971 struct inode *inode, int log_inode_only)
2973 struct btrfs_map_token token;
2975 btrfs_init_map_token(&token);
2977 if (log_inode_only) {
2978 /* set the generation to zero so the recover code
2979 * can tell the difference between an logging
2980 * just to say 'this inode exists' and a logging
2981 * to say 'update this inode with these values'
2983 btrfs_set_token_inode_generation(leaf, item, 0, &token);
2984 btrfs_set_token_inode_size(leaf, item, 0, &token);
2986 btrfs_set_token_inode_generation(leaf, item,
2987 BTRFS_I(inode)->generation,
2989 btrfs_set_token_inode_size(leaf, item, inode->i_size, &token);
2992 btrfs_set_token_inode_uid(leaf, item, i_uid_read(inode), &token);
2993 btrfs_set_token_inode_gid(leaf, item, i_gid_read(inode), &token);
2994 btrfs_set_token_inode_mode(leaf, item, inode->i_mode, &token);
2995 btrfs_set_token_inode_nlink(leaf, item, inode->i_nlink, &token);
2997 btrfs_set_token_timespec_sec(leaf, btrfs_inode_atime(item),
2998 inode->i_atime.tv_sec, &token);
2999 btrfs_set_token_timespec_nsec(leaf, btrfs_inode_atime(item),
3000 inode->i_atime.tv_nsec, &token);
3002 btrfs_set_token_timespec_sec(leaf, btrfs_inode_mtime(item),
3003 inode->i_mtime.tv_sec, &token);
3004 btrfs_set_token_timespec_nsec(leaf, btrfs_inode_mtime(item),
3005 inode->i_mtime.tv_nsec, &token);
3007 btrfs_set_token_timespec_sec(leaf, btrfs_inode_ctime(item),
3008 inode->i_ctime.tv_sec, &token);
3009 btrfs_set_token_timespec_nsec(leaf, btrfs_inode_ctime(item),
3010 inode->i_ctime.tv_nsec, &token);
3012 btrfs_set_token_inode_nbytes(leaf, item, inode_get_bytes(inode),
3015 btrfs_set_token_inode_sequence(leaf, item, inode->i_version, &token);
3016 btrfs_set_token_inode_transid(leaf, item, trans->transid, &token);
3017 btrfs_set_token_inode_rdev(leaf, item, inode->i_rdev, &token);
3018 btrfs_set_token_inode_flags(leaf, item, BTRFS_I(inode)->flags, &token);
3019 btrfs_set_token_inode_block_group(leaf, item, 0, &token);
3022 static int log_inode_item(struct btrfs_trans_handle *trans,
3023 struct btrfs_root *log, struct btrfs_path *path,
3024 struct inode *inode)
3026 struct btrfs_inode_item *inode_item;
3027 struct btrfs_key key;
3030 memcpy(&key, &BTRFS_I(inode)->location, sizeof(key));
3031 ret = btrfs_insert_empty_item(trans, log, path, &key,
3032 sizeof(*inode_item));
3033 if (ret && ret != -EEXIST)
3035 inode_item = btrfs_item_ptr(path->nodes[0], path->slots[0],
3036 struct btrfs_inode_item);
3037 fill_inode_item(trans, path->nodes[0], inode_item, inode, 0);
3038 btrfs_release_path(path);
3042 static noinline int copy_items(struct btrfs_trans_handle *trans,
3043 struct inode *inode,
3044 struct btrfs_path *dst_path,
3045 struct extent_buffer *src,
3046 int start_slot, int nr, int inode_only)
3048 unsigned long src_offset;
3049 unsigned long dst_offset;
3050 struct btrfs_root *log = BTRFS_I(inode)->root->log_root;
3051 struct btrfs_file_extent_item *extent;
3052 struct btrfs_inode_item *inode_item;
3054 struct btrfs_key *ins_keys;
3058 struct list_head ordered_sums;
3059 int skip_csum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;
3061 INIT_LIST_HEAD(&ordered_sums);
3063 ins_data = kmalloc(nr * sizeof(struct btrfs_key) +
3064 nr * sizeof(u32), GFP_NOFS);
3068 ins_sizes = (u32 *)ins_data;
3069 ins_keys = (struct btrfs_key *)(ins_data + nr * sizeof(u32));
3071 for (i = 0; i < nr; i++) {
3072 ins_sizes[i] = btrfs_item_size_nr(src, i + start_slot);
3073 btrfs_item_key_to_cpu(src, ins_keys + i, i + start_slot);
3075 ret = btrfs_insert_empty_items(trans, log, dst_path,
3076 ins_keys, ins_sizes, nr);
3082 for (i = 0; i < nr; i++, dst_path->slots[0]++) {
3083 dst_offset = btrfs_item_ptr_offset(dst_path->nodes[0],
3084 dst_path->slots[0]);
3086 src_offset = btrfs_item_ptr_offset(src, start_slot + i);
3088 if (ins_keys[i].type == BTRFS_INODE_ITEM_KEY) {
3089 inode_item = btrfs_item_ptr(dst_path->nodes[0],
3091 struct btrfs_inode_item);
3092 fill_inode_item(trans, dst_path->nodes[0], inode_item,
3093 inode, inode_only == LOG_INODE_EXISTS);
3095 copy_extent_buffer(dst_path->nodes[0], src, dst_offset,
3096 src_offset, ins_sizes[i]);
3099 /* take a reference on file data extents so that truncates
3100 * or deletes of this inode don't have to relog the inode
3103 if (btrfs_key_type(ins_keys + i) == BTRFS_EXTENT_DATA_KEY &&
3106 extent = btrfs_item_ptr(src, start_slot + i,
3107 struct btrfs_file_extent_item);
3109 if (btrfs_file_extent_generation(src, extent) < trans->transid)
3112 found_type = btrfs_file_extent_type(src, extent);
3113 if (found_type == BTRFS_FILE_EXTENT_REG) {
3115 ds = btrfs_file_extent_disk_bytenr(src,
3117 /* ds == 0 is a hole */
3121 dl = btrfs_file_extent_disk_num_bytes(src,
3123 cs = btrfs_file_extent_offset(src, extent);
3124 cl = btrfs_file_extent_num_bytes(src,
3126 if (btrfs_file_extent_compression(src,
3132 ret = btrfs_lookup_csums_range(
3133 log->fs_info->csum_root,
3134 ds + cs, ds + cs + cl - 1,
3141 btrfs_mark_buffer_dirty(dst_path->nodes[0]);
3142 btrfs_release_path(dst_path);
3146 * we have to do this after the loop above to avoid changing the
3147 * log tree while trying to change the log tree.
3150 while (!list_empty(&ordered_sums)) {
3151 struct btrfs_ordered_sum *sums = list_entry(ordered_sums.next,
3152 struct btrfs_ordered_sum,
3155 ret = btrfs_csum_file_blocks(trans, log, sums);
3156 list_del(&sums->list);
3162 static int extent_cmp(void *priv, struct list_head *a, struct list_head *b)
3164 struct extent_map *em1, *em2;
3166 em1 = list_entry(a, struct extent_map, list);
3167 em2 = list_entry(b, struct extent_map, list);
3169 if (em1->start < em2->start)
3171 else if (em1->start > em2->start)
3176 static int drop_adjacent_extents(struct btrfs_trans_handle *trans,
3177 struct btrfs_root *root, struct inode *inode,
3178 struct extent_map *em,
3179 struct btrfs_path *path)
3181 struct btrfs_file_extent_item *fi;
3182 struct extent_buffer *leaf;
3183 struct btrfs_key key, new_key;
3184 struct btrfs_map_token token;
3186 u64 extent_offset = 0;
3193 btrfs_init_map_token(&token);
3194 leaf = path->nodes[0];
3196 if (path->slots[0] >= btrfs_header_nritems(leaf)) {
3198 ret = btrfs_del_items(trans, root, path,
3205 ret = btrfs_next_leaf_write(trans, root, path, 1);
3210 leaf = path->nodes[0];
3213 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
3214 if (key.objectid != btrfs_ino(inode) ||
3215 key.type != BTRFS_EXTENT_DATA_KEY ||
3216 key.offset >= em->start + em->len)
3219 fi = btrfs_item_ptr(leaf, path->slots[0],
3220 struct btrfs_file_extent_item);
3221 extent_type = btrfs_token_file_extent_type(leaf, fi, &token);
3222 if (extent_type == BTRFS_FILE_EXTENT_REG ||
3223 extent_type == BTRFS_FILE_EXTENT_PREALLOC) {
3224 extent_offset = btrfs_token_file_extent_offset(leaf,
3226 extent_end = key.offset +
3227 btrfs_token_file_extent_num_bytes(leaf, fi,
3229 } else if (extent_type == BTRFS_FILE_EXTENT_INLINE) {
3230 extent_end = key.offset +
3231 btrfs_file_extent_inline_len(leaf, fi);
3236 if (extent_end <= em->len + em->start) {
3238 del_slot = path->slots[0];
3245 * Ok so we'll ignore previous items if we log a new extent,
3246 * which can lead to overlapping extents, so if we have an
3247 * existing extent we want to adjust we _have_ to check the next
3248 * guy to make sure we even need this extent anymore, this keeps
3249 * us from panicing in set_item_key_safe.
3251 if (path->slots[0] < btrfs_header_nritems(leaf) - 1) {
3252 struct btrfs_key tmp_key;
3254 btrfs_item_key_to_cpu(leaf, &tmp_key,
3255 path->slots[0] + 1);
3256 if (tmp_key.objectid == btrfs_ino(inode) &&
3257 tmp_key.type == BTRFS_EXTENT_DATA_KEY &&
3258 tmp_key.offset <= em->start + em->len) {
3260 del_slot = path->slots[0];
3266 BUG_ON(extent_type == BTRFS_FILE_EXTENT_INLINE);
3267 memcpy(&new_key, &key, sizeof(new_key));
3268 new_key.offset = em->start + em->len;
3269 btrfs_set_item_key_safe(trans, root, path, &new_key);
3270 extent_offset += em->start + em->len - key.offset;
3271 btrfs_set_token_file_extent_offset(leaf, fi, extent_offset,
3273 btrfs_set_token_file_extent_num_bytes(leaf, fi, extent_end -
3274 (em->start + em->len),
3276 btrfs_mark_buffer_dirty(leaf);
3280 ret = btrfs_del_items(trans, root, path, del_slot, del_nr);
3285 static int log_one_extent(struct btrfs_trans_handle *trans,
3286 struct inode *inode, struct btrfs_root *root,
3287 struct extent_map *em, struct btrfs_path *path)
3289 struct btrfs_root *log = root->log_root;
3290 struct btrfs_file_extent_item *fi;
3291 struct extent_buffer *leaf;
3292 struct btrfs_ordered_extent *ordered;
3293 struct list_head ordered_sums;
3294 struct btrfs_map_token token;
3295 struct btrfs_key key;
3296 u64 mod_start = em->mod_start;
3297 u64 mod_len = em->mod_len;
3300 u64 extent_offset = em->start - em->orig_start;
3303 int index = log->log_transid % 2;
3304 bool skip_csum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;
3307 INIT_LIST_HEAD(&ordered_sums);
3308 btrfs_init_map_token(&token);
3309 key.objectid = btrfs_ino(inode);
3310 key.type = BTRFS_EXTENT_DATA_KEY;
3311 key.offset = em->start;
3312 path->really_keep_locks = 1;
3314 ret = btrfs_insert_empty_item(trans, log, path, &key, sizeof(*fi));
3315 if (ret && ret != -EEXIST) {
3316 path->really_keep_locks = 0;
3319 leaf = path->nodes[0];
3320 fi = btrfs_item_ptr(leaf, path->slots[0],
3321 struct btrfs_file_extent_item);
3324 * If we are overwriting an inline extent with a real one then we need
3325 * to just delete the inline extent as it may not be large enough to
3326 * have the entire file_extent_item.
3328 if (ret && btrfs_token_file_extent_type(leaf, fi, &token) ==
3329 BTRFS_FILE_EXTENT_INLINE) {
3330 ret = btrfs_del_item(trans, log, path);
3331 btrfs_release_path(path);
3333 path->really_keep_locks = 0;
3339 btrfs_set_token_file_extent_generation(leaf, fi, em->generation,
3341 if (test_bit(EXTENT_FLAG_PREALLOC, &em->flags)) {
3343 btrfs_set_token_file_extent_type(leaf, fi,
3344 BTRFS_FILE_EXTENT_PREALLOC,
3347 btrfs_set_token_file_extent_type(leaf, fi,
3348 BTRFS_FILE_EXTENT_REG,
3350 if (em->block_start == 0)
3354 block_len = max(em->block_len, em->orig_block_len);
3355 if (em->compress_type != BTRFS_COMPRESS_NONE) {
3356 btrfs_set_token_file_extent_disk_bytenr(leaf, fi,
3359 btrfs_set_token_file_extent_disk_num_bytes(leaf, fi, block_len,
3361 } else if (em->block_start < EXTENT_MAP_LAST_BYTE) {
3362 btrfs_set_token_file_extent_disk_bytenr(leaf, fi,
3364 extent_offset, &token);
3365 btrfs_set_token_file_extent_disk_num_bytes(leaf, fi, block_len,
3368 btrfs_set_token_file_extent_disk_bytenr(leaf, fi, 0, &token);
3369 btrfs_set_token_file_extent_disk_num_bytes(leaf, fi, 0,
3373 btrfs_set_token_file_extent_offset(leaf, fi,
3374 em->start - em->orig_start,
3376 btrfs_set_token_file_extent_num_bytes(leaf, fi, em->len, &token);
3377 btrfs_set_token_file_extent_ram_bytes(leaf, fi, em->len, &token);
3378 btrfs_set_token_file_extent_compression(leaf, fi, em->compress_type,
3380 btrfs_set_token_file_extent_encryption(leaf, fi, 0, &token);
3381 btrfs_set_token_file_extent_other_encoding(leaf, fi, 0, &token);
3382 btrfs_mark_buffer_dirty(leaf);
3385 * Have to check the extent to the right of us to make sure it doesn't
3386 * fall in our current range. We're ok if the previous extent is in our
3387 * range since the recovery stuff will run us in key order and thus just
3388 * drop the part we overwrote.
3390 ret = drop_adjacent_extents(trans, log, inode, em, path);
3391 btrfs_release_path(path);
3392 path->really_keep_locks = 0;
3400 if (em->compress_type) {
3402 csum_len = block_len;
3406 * First check and see if our csums are on our outstanding ordered
3410 spin_lock_irq(&log->log_extents_lock[index]);
3411 list_for_each_entry(ordered, &log->logged_list[index], log_list) {
3412 struct btrfs_ordered_sum *sum;
3417 if (ordered->inode != inode)
3420 if (ordered->file_offset + ordered->len <= mod_start ||
3421 mod_start + mod_len <= ordered->file_offset)
3425 * We are going to copy all the csums on this ordered extent, so
3426 * go ahead and adjust mod_start and mod_len in case this
3427 * ordered extent has already been logged.
3429 if (ordered->file_offset > mod_start) {
3430 if (ordered->file_offset + ordered->len >=
3431 mod_start + mod_len)
3432 mod_len = ordered->file_offset - mod_start;
3434 * If we have this case
3436 * |--------- logged extent ---------|
3437 * |----- ordered extent ----|
3439 * Just don't mess with mod_start and mod_len, we'll
3440 * just end up logging more csums than we need and it
3444 if (ordered->file_offset + ordered->len <
3445 mod_start + mod_len) {
3446 mod_len = (mod_start + mod_len) -
3447 (ordered->file_offset + ordered->len);
3448 mod_start = ordered->file_offset +
3456 * To keep us from looping for the above case of an ordered
3457 * extent that falls inside of the logged extent.
3459 if (test_and_set_bit(BTRFS_ORDERED_LOGGED_CSUM,
3462 atomic_inc(&ordered->refs);
3463 spin_unlock_irq(&log->log_extents_lock[index]);
3465 * we've dropped the lock, we must either break or
3466 * start over after this.
3469 wait_event(ordered->wait, ordered->csum_bytes_left == 0);
3471 list_for_each_entry(sum, &ordered->list, list) {
3472 ret = btrfs_csum_file_blocks(trans, log, sum);
3474 btrfs_put_ordered_extent(ordered);
3478 btrfs_put_ordered_extent(ordered);
3482 spin_unlock_irq(&log->log_extents_lock[index]);
3485 if (!mod_len || ret)
3488 csum_offset = mod_start - em->start;
3491 /* block start is already adjusted for the file extent offset. */
3492 ret = btrfs_lookup_csums_range(log->fs_info->csum_root,
3493 em->block_start + csum_offset,
3494 em->block_start + csum_offset +
3495 csum_len - 1, &ordered_sums, 0);
3499 while (!list_empty(&ordered_sums)) {
3500 struct btrfs_ordered_sum *sums = list_entry(ordered_sums.next,
3501 struct btrfs_ordered_sum,
3504 ret = btrfs_csum_file_blocks(trans, log, sums);
3505 list_del(&sums->list);
3512 static int btrfs_log_changed_extents(struct btrfs_trans_handle *trans,
3513 struct btrfs_root *root,
3514 struct inode *inode,
3515 struct btrfs_path *path)
3517 struct extent_map *em, *n;
3518 struct list_head extents;
3519 struct extent_map_tree *tree = &BTRFS_I(inode)->extent_tree;
3524 INIT_LIST_HEAD(&extents);
3526 write_lock(&tree->lock);
3527 test_gen = root->fs_info->last_trans_committed;
3529 list_for_each_entry_safe(em, n, &tree->modified_extents, list) {
3530 list_del_init(&em->list);
3533 * Just an arbitrary number, this can be really CPU intensive
3534 * once we start getting a lot of extents, and really once we
3535 * have a bunch of extents we just want to commit since it will
3538 if (++num > 32768) {
3539 list_del_init(&tree->modified_extents);
3544 if (em->generation <= test_gen)
3546 /* Need a ref to keep it from getting evicted from cache */
3547 atomic_inc(&em->refs);
3548 set_bit(EXTENT_FLAG_LOGGING, &em->flags);
3549 list_add_tail(&em->list, &extents);
3553 list_sort(NULL, &extents, extent_cmp);
3556 while (!list_empty(&extents)) {
3557 em = list_entry(extents.next, struct extent_map, list);
3559 list_del_init(&em->list);
3562 * If we had an error we just need to delete everybody from our
3566 clear_em_logging(tree, em);
3567 free_extent_map(em);
3571 write_unlock(&tree->lock);
3573 ret = log_one_extent(trans, inode, root, em, path);
3574 write_lock(&tree->lock);
3575 clear_em_logging(tree, em);
3576 free_extent_map(em);
3578 WARN_ON(!list_empty(&extents));
3579 write_unlock(&tree->lock);
3581 btrfs_release_path(path);
3585 /* log a single inode in the tree log.
3586 * At least one parent directory for this inode must exist in the tree
3587 * or be logged already.
3589 * Any items from this inode changed by the current transaction are copied
3590 * to the log tree. An extra reference is taken on any extents in this
3591 * file, allowing us to avoid a whole pile of corner cases around logging
3592 * blocks that have been removed from the tree.
3594 * See LOG_INODE_ALL and related defines for a description of what inode_only
3597 * This handles both files and directories.
3599 static int btrfs_log_inode(struct btrfs_trans_handle *trans,
3600 struct btrfs_root *root, struct inode *inode,
3603 struct btrfs_path *path;
3604 struct btrfs_path *dst_path;
3605 struct btrfs_key min_key;
3606 struct btrfs_key max_key;
3607 struct btrfs_root *log = root->log_root;
3608 struct extent_buffer *src = NULL;
3612 int ins_start_slot = 0;
3614 bool fast_search = false;
3615 u64 ino = btrfs_ino(inode);
3617 log = root->log_root;
3619 path = btrfs_alloc_path();
3622 dst_path = btrfs_alloc_path();
3624 btrfs_free_path(path);
3628 min_key.objectid = ino;
3629 min_key.type = BTRFS_INODE_ITEM_KEY;
3632 max_key.objectid = ino;
3635 /* today the code can only do partial logging of directories */
3636 if (S_ISDIR(inode->i_mode) ||
3637 (!test_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
3638 &BTRFS_I(inode)->runtime_flags) &&
3639 inode_only == LOG_INODE_EXISTS))
3640 max_key.type = BTRFS_XATTR_ITEM_KEY;
3642 max_key.type = (u8)-1;
3643 max_key.offset = (u64)-1;
3645 /* Only run delayed items if we are a dir or a new file */
3646 if (S_ISDIR(inode->i_mode) ||
3647 BTRFS_I(inode)->generation > root->fs_info->last_trans_committed) {
3648 ret = btrfs_commit_inode_delayed_items(trans, inode);
3650 btrfs_free_path(path);
3651 btrfs_free_path(dst_path);
3656 mutex_lock(&BTRFS_I(inode)->log_mutex);
3658 btrfs_get_logged_extents(log, inode);
3661 * a brute force approach to making sure we get the most uptodate
3662 * copies of everything.
3664 if (S_ISDIR(inode->i_mode)) {
3665 int max_key_type = BTRFS_DIR_LOG_INDEX_KEY;
3667 if (inode_only == LOG_INODE_EXISTS)
3668 max_key_type = BTRFS_XATTR_ITEM_KEY;
3669 ret = drop_objectid_items(trans, log, path, ino, max_key_type);
3671 if (test_and_clear_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
3672 &BTRFS_I(inode)->runtime_flags)) {
3673 clear_bit(BTRFS_INODE_COPY_EVERYTHING,
3674 &BTRFS_I(inode)->runtime_flags);
3675 ret = btrfs_truncate_inode_items(trans, log,
3677 } else if (test_and_clear_bit(BTRFS_INODE_COPY_EVERYTHING,
3678 &BTRFS_I(inode)->runtime_flags)) {
3679 if (inode_only == LOG_INODE_ALL)
3681 max_key.type = BTRFS_XATTR_ITEM_KEY;
3682 ret = drop_objectid_items(trans, log, path, ino,
3685 if (inode_only == LOG_INODE_ALL)
3687 ret = log_inode_item(trans, log, dst_path, inode);
3700 path->keep_locks = 1;
3704 ret = btrfs_search_forward(root, &min_key, &max_key,
3705 path, trans->transid);
3709 /* note, ins_nr might be > 0 here, cleanup outside the loop */
3710 if (min_key.objectid != ino)
3712 if (min_key.type > max_key.type)
3715 src = path->nodes[0];
3716 if (ins_nr && ins_start_slot + ins_nr == path->slots[0]) {
3719 } else if (!ins_nr) {
3720 ins_start_slot = path->slots[0];
3725 ret = copy_items(trans, inode, dst_path, src, ins_start_slot,
3726 ins_nr, inode_only);
3732 ins_start_slot = path->slots[0];
3735 nritems = btrfs_header_nritems(path->nodes[0]);
3737 if (path->slots[0] < nritems) {
3738 btrfs_item_key_to_cpu(path->nodes[0], &min_key,
3743 ret = copy_items(trans, inode, dst_path, src,
3745 ins_nr, inode_only);
3752 btrfs_release_path(path);
3754 if (min_key.offset < (u64)-1)
3756 else if (min_key.type < (u8)-1)
3758 else if (min_key.objectid < (u64)-1)
3764 ret = copy_items(trans, inode, dst_path, src, ins_start_slot,
3765 ins_nr, inode_only);
3775 btrfs_release_path(dst_path);
3776 ret = btrfs_log_changed_extents(trans, root, inode, dst_path);
3782 struct extent_map_tree *tree = &BTRFS_I(inode)->extent_tree;
3783 struct extent_map *em, *n;
3785 write_lock(&tree->lock);
3786 list_for_each_entry_safe(em, n, &tree->modified_extents, list)
3787 list_del_init(&em->list);
3788 write_unlock(&tree->lock);
3791 if (inode_only == LOG_INODE_ALL && S_ISDIR(inode->i_mode)) {
3792 btrfs_release_path(path);
3793 btrfs_release_path(dst_path);
3794 ret = log_directory_changes(trans, root, inode, path, dst_path);
3800 BTRFS_I(inode)->logged_trans = trans->transid;
3801 BTRFS_I(inode)->last_log_commit = BTRFS_I(inode)->last_sub_trans;
3804 btrfs_free_logged_extents(log, log->log_transid);
3805 mutex_unlock(&BTRFS_I(inode)->log_mutex);
3807 btrfs_free_path(path);
3808 btrfs_free_path(dst_path);
3813 * follow the dentry parent pointers up the chain and see if any
3814 * of the directories in it require a full commit before they can
3815 * be logged. Returns zero if nothing special needs to be done or 1 if
3816 * a full commit is required.
3818 static noinline int check_parent_dirs_for_sync(struct btrfs_trans_handle *trans,
3819 struct inode *inode,
3820 struct dentry *parent,
3821 struct super_block *sb,
3825 struct btrfs_root *root;
3826 struct dentry *old_parent = NULL;
3829 * for regular files, if its inode is already on disk, we don't
3830 * have to worry about the parents at all. This is because
3831 * we can use the last_unlink_trans field to record renames
3832 * and other fun in this file.
3834 if (S_ISREG(inode->i_mode) &&
3835 BTRFS_I(inode)->generation <= last_committed &&
3836 BTRFS_I(inode)->last_unlink_trans <= last_committed)
3839 if (!S_ISDIR(inode->i_mode)) {
3840 if (!parent || !parent->d_inode || sb != parent->d_inode->i_sb)
3842 inode = parent->d_inode;
3846 BTRFS_I(inode)->logged_trans = trans->transid;
3849 if (BTRFS_I(inode)->last_unlink_trans > last_committed) {
3850 root = BTRFS_I(inode)->root;
3853 * make sure any commits to the log are forced
3854 * to be full commits
3856 root->fs_info->last_trans_log_full_commit =
3862 if (!parent || !parent->d_inode || sb != parent->d_inode->i_sb)
3865 if (IS_ROOT(parent))
3868 parent = dget_parent(parent);
3870 old_parent = parent;
3871 inode = parent->d_inode;
3880 * helper function around btrfs_log_inode to make sure newly created
3881 * parent directories also end up in the log. A minimal inode and backref
3882 * only logging is done of any parent directories that are older than
3883 * the last committed transaction
3885 int btrfs_log_inode_parent(struct btrfs_trans_handle *trans,
3886 struct btrfs_root *root, struct inode *inode,
3887 struct dentry *parent, int exists_only)
3889 int inode_only = exists_only ? LOG_INODE_EXISTS : LOG_INODE_ALL;
3890 struct super_block *sb;
3891 struct dentry *old_parent = NULL;
3893 u64 last_committed = root->fs_info->last_trans_committed;
3897 if (btrfs_test_opt(root, NOTREELOG)) {
3902 if (root->fs_info->last_trans_log_full_commit >
3903 root->fs_info->last_trans_committed) {
3908 if (root != BTRFS_I(inode)->root ||
3909 btrfs_root_refs(&root->root_item) == 0) {
3914 ret = check_parent_dirs_for_sync(trans, inode, parent,
3915 sb, last_committed);
3919 if (btrfs_inode_in_log(inode, trans->transid)) {
3920 ret = BTRFS_NO_LOG_SYNC;
3924 ret = start_log_trans(trans, root);
3928 ret = btrfs_log_inode(trans, root, inode, inode_only);
3933 * for regular files, if its inode is already on disk, we don't
3934 * have to worry about the parents at all. This is because
3935 * we can use the last_unlink_trans field to record renames
3936 * and other fun in this file.
3938 if (S_ISREG(inode->i_mode) &&
3939 BTRFS_I(inode)->generation <= last_committed &&
3940 BTRFS_I(inode)->last_unlink_trans <= last_committed) {
3945 inode_only = LOG_INODE_EXISTS;
3947 if (!parent || !parent->d_inode || sb != parent->d_inode->i_sb)
3950 inode = parent->d_inode;
3951 if (root != BTRFS_I(inode)->root)
3954 if (BTRFS_I(inode)->generation >
3955 root->fs_info->last_trans_committed) {
3956 ret = btrfs_log_inode(trans, root, inode, inode_only);
3960 if (IS_ROOT(parent))
3963 parent = dget_parent(parent);
3965 old_parent = parent;
3971 root->fs_info->last_trans_log_full_commit = trans->transid;
3974 btrfs_end_log_trans(root);
3980 * it is not safe to log dentry if the chunk root has added new
3981 * chunks. This returns 0 if the dentry was logged, and 1 otherwise.
3982 * If this returns 1, you must commit the transaction to safely get your
3985 int btrfs_log_dentry_safe(struct btrfs_trans_handle *trans,
3986 struct btrfs_root *root, struct dentry *dentry)
3988 struct dentry *parent = dget_parent(dentry);
3991 ret = btrfs_log_inode_parent(trans, root, dentry->d_inode, parent, 0);
3998 * should be called during mount to recover any replay any log trees
4001 int btrfs_recover_log_trees(struct btrfs_root *log_root_tree)
4004 struct btrfs_path *path;
4005 struct btrfs_trans_handle *trans;
4006 struct btrfs_key key;
4007 struct btrfs_key found_key;
4008 struct btrfs_key tmp_key;
4009 struct btrfs_root *log;
4010 struct btrfs_fs_info *fs_info = log_root_tree->fs_info;
4011 struct walk_control wc = {
4012 .process_func = process_one_buffer,
4016 path = btrfs_alloc_path();
4020 fs_info->log_root_recovering = 1;
4022 trans = btrfs_start_transaction(fs_info->tree_root, 0);
4023 if (IS_ERR(trans)) {
4024 ret = PTR_ERR(trans);
4031 ret = walk_log_tree(trans, log_root_tree, &wc);
4033 btrfs_error(fs_info, ret, "Failed to pin buffers while "
4034 "recovering log root tree.");
4039 key.objectid = BTRFS_TREE_LOG_OBJECTID;
4040 key.offset = (u64)-1;
4041 btrfs_set_key_type(&key, BTRFS_ROOT_ITEM_KEY);
4044 ret = btrfs_search_slot(NULL, log_root_tree, &key, path, 0, 0);
4047 btrfs_error(fs_info, ret,
4048 "Couldn't find tree log root.");
4052 if (path->slots[0] == 0)
4056 btrfs_item_key_to_cpu(path->nodes[0], &found_key,
4058 btrfs_release_path(path);
4059 if (found_key.objectid != BTRFS_TREE_LOG_OBJECTID)
4062 log = btrfs_read_fs_root_no_radix(log_root_tree,
4066 btrfs_error(fs_info, ret,
4067 "Couldn't read tree log root.");
4071 tmp_key.objectid = found_key.offset;
4072 tmp_key.type = BTRFS_ROOT_ITEM_KEY;
4073 tmp_key.offset = (u64)-1;
4075 wc.replay_dest = btrfs_read_fs_root_no_name(fs_info, &tmp_key);
4076 if (IS_ERR(wc.replay_dest)) {
4077 ret = PTR_ERR(wc.replay_dest);
4078 btrfs_error(fs_info, ret, "Couldn't read target root "
4079 "for tree log recovery.");
4083 wc.replay_dest->log_root = log;
4084 btrfs_record_root_in_trans(trans, wc.replay_dest);
4085 ret = walk_log_tree(trans, log, &wc);
4088 if (wc.stage == LOG_WALK_REPLAY_ALL) {
4089 ret = fixup_inode_link_counts(trans, wc.replay_dest,
4094 key.offset = found_key.offset - 1;
4095 wc.replay_dest->log_root = NULL;
4096 free_extent_buffer(log->node);
4097 free_extent_buffer(log->commit_root);
4100 if (found_key.offset == 0)
4103 btrfs_release_path(path);
4105 /* step one is to pin it all, step two is to replay just inodes */
4108 wc.process_func = replay_one_buffer;
4109 wc.stage = LOG_WALK_REPLAY_INODES;
4112 /* step three is to replay everything */
4113 if (wc.stage < LOG_WALK_REPLAY_ALL) {
4118 btrfs_free_path(path);
4120 free_extent_buffer(log_root_tree->node);
4121 log_root_tree->log_root = NULL;
4122 fs_info->log_root_recovering = 0;
4124 /* step 4: commit the transaction, which also unpins the blocks */
4125 btrfs_commit_transaction(trans, fs_info->tree_root);
4127 kfree(log_root_tree);
4131 btrfs_free_path(path);
4136 * there are some corner cases where we want to force a full
4137 * commit instead of allowing a directory to be logged.
4139 * They revolve around files there were unlinked from the directory, and
4140 * this function updates the parent directory so that a full commit is
4141 * properly done if it is fsync'd later after the unlinks are done.
4143 void btrfs_record_unlink_dir(struct btrfs_trans_handle *trans,
4144 struct inode *dir, struct inode *inode,
4148 * when we're logging a file, if it hasn't been renamed
4149 * or unlinked, and its inode is fully committed on disk,
4150 * we don't have to worry about walking up the directory chain
4151 * to log its parents.
4153 * So, we use the last_unlink_trans field to put this transid
4154 * into the file. When the file is logged we check it and
4155 * don't log the parents if the file is fully on disk.
4157 if (S_ISREG(inode->i_mode))
4158 BTRFS_I(inode)->last_unlink_trans = trans->transid;
4161 * if this directory was already logged any new
4162 * names for this file/dir will get recorded
4165 if (BTRFS_I(dir)->logged_trans == trans->transid)
4169 * if the inode we're about to unlink was logged,
4170 * the log will be properly updated for any new names
4172 if (BTRFS_I(inode)->logged_trans == trans->transid)
4176 * when renaming files across directories, if the directory
4177 * there we're unlinking from gets fsync'd later on, there's
4178 * no way to find the destination directory later and fsync it
4179 * properly. So, we have to be conservative and force commits
4180 * so the new name gets discovered.
4185 /* we can safely do the unlink without any special recording */
4189 BTRFS_I(dir)->last_unlink_trans = trans->transid;
4193 * Call this after adding a new name for a file and it will properly
4194 * update the log to reflect the new name.
4196 * It will return zero if all goes well, and it will return 1 if a
4197 * full transaction commit is required.
4199 int btrfs_log_new_name(struct btrfs_trans_handle *trans,
4200 struct inode *inode, struct inode *old_dir,
4201 struct dentry *parent)
4203 struct btrfs_root * root = BTRFS_I(inode)->root;
4206 * this will force the logging code to walk the dentry chain
4209 if (S_ISREG(inode->i_mode))
4210 BTRFS_I(inode)->last_unlink_trans = trans->transid;
4213 * if this inode hasn't been logged and directory we're renaming it
4214 * from hasn't been logged, we don't need to log it
4216 if (BTRFS_I(inode)->logged_trans <=
4217 root->fs_info->last_trans_committed &&
4218 (!old_dir || BTRFS_I(old_dir)->logged_trans <=
4219 root->fs_info->last_trans_committed))
4222 return btrfs_log_inode_parent(trans, root, inode, parent, 1);