blob: 1bbaace733838e6fc70dba1babafb6f9265c00f9 [file] [log] [blame]
Chris Masone02119d2008-09-05 16:13:11 -04001/*
2 * Copyright (C) 2008 Oracle. All rights reserved.
3 *
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.
7 *
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.
12 *
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.
17 */
18
19#include <linux/sched.h>
Tejun Heo5a0e3ad2010-03-24 17:04:11 +090020#include <linux/slab.h>
Miao Xiec6adc9c2013-05-28 10:05:39 +000021#include <linux/blkdev.h>
Josef Bacik5dc562c2012-08-17 13:14:17 -040022#include <linux/list_sort.h>
Miao Xie995946d2014-04-02 19:51:06 +080023#include "tree-log.h"
Chris Masone02119d2008-09-05 16:13:11 -040024#include "disk-io.h"
25#include "locking.h"
26#include "print-tree.h"
Mark Fashehf1863732012-08-08 11:32:27 -070027#include "backref.h"
Mark Fashehf1863732012-08-08 11:32:27 -070028#include "hash.h"
Chris Masone02119d2008-09-05 16:13:11 -040029
30/* magic values for the inode_only field in btrfs_log_inode:
31 *
32 * LOG_INODE_ALL means to log everything
33 * LOG_INODE_EXISTS means to log just enough to recreate the inode
34 * during log replay
35 */
36#define LOG_INODE_ALL 0
37#define LOG_INODE_EXISTS 1
38
39/*
Chris Mason12fcfd22009-03-24 10:24:20 -040040 * directory trouble cases
41 *
42 * 1) on rename or unlink, if the inode being unlinked isn't in the fsync
43 * log, we must force a full commit before doing an fsync of the directory
44 * where the unlink was done.
45 * ---> record transid of last unlink/rename per directory
46 *
47 * mkdir foo/some_dir
48 * normal commit
49 * rename foo/some_dir foo2/some_dir
50 * mkdir foo/some_dir
51 * fsync foo/some_dir/some_file
52 *
53 * The fsync above will unlink the original some_dir without recording
54 * it in its new location (foo2). After a crash, some_dir will be gone
55 * unless the fsync of some_file forces a full commit
56 *
57 * 2) we must log any new names for any file or dir that is in the fsync
58 * log. ---> check inode while renaming/linking.
59 *
60 * 2a) we must log any new names for any file or dir during rename
61 * when the directory they are being removed from was logged.
62 * ---> check inode and old parent dir during rename
63 *
64 * 2a is actually the more important variant. With the extra logging
65 * a crash might unlink the old name without recreating the new one
66 *
67 * 3) after a crash, we must go through any directories with a link count
68 * of zero and redo the rm -rf
69 *
70 * mkdir f1/foo
71 * normal commit
72 * rm -rf f1/foo
73 * fsync(f1)
74 *
75 * The directory f1 was fully removed from the FS, but fsync was never
76 * called on f1, only its parent dir. After a crash the rm -rf must
77 * be replayed. This must be able to recurse down the entire
78 * directory tree. The inode link count fixup code takes care of the
79 * ugly details.
80 */
81
82/*
Chris Masone02119d2008-09-05 16:13:11 -040083 * stages for the tree walking. The first
84 * stage (0) is to only pin down the blocks we find
85 * the second stage (1) is to make sure that all the inodes
86 * we find in the log are created in the subvolume.
87 *
88 * The last stage is to deal with directories and links and extents
89 * and all the other fun semantics
90 */
91#define LOG_WALK_PIN_ONLY 0
92#define LOG_WALK_REPLAY_INODES 1
Josef Bacikdd8e7212013-09-11 11:57:23 -040093#define LOG_WALK_REPLAY_DIR_INDEX 2
94#define LOG_WALK_REPLAY_ALL 3
Chris Masone02119d2008-09-05 16:13:11 -040095
Chris Mason12fcfd22009-03-24 10:24:20 -040096static int btrfs_log_inode(struct btrfs_trans_handle *trans,
Filipe Manana49dae1b2014-09-06 22:34:39 +010097 struct btrfs_root *root, struct inode *inode,
98 int inode_only,
99 const loff_t start,
Filipe Manana8407f552014-09-05 15:14:39 +0100100 const loff_t end,
101 struct btrfs_log_ctx *ctx);
Yan Zhengec051c02009-01-05 15:43:42 -0500102static int link_to_fixup_dir(struct btrfs_trans_handle *trans,
103 struct btrfs_root *root,
104 struct btrfs_path *path, u64 objectid);
Chris Mason12fcfd22009-03-24 10:24:20 -0400105static noinline int replay_dir_deletes(struct btrfs_trans_handle *trans,
106 struct btrfs_root *root,
107 struct btrfs_root *log,
108 struct btrfs_path *path,
109 u64 dirid, int del_all);
Chris Masone02119d2008-09-05 16:13:11 -0400110
111/*
112 * tree logging is a special write ahead log used to make sure that
113 * fsyncs and O_SYNCs can happen without doing full tree commits.
114 *
115 * Full tree commits are expensive because they require commonly
116 * modified blocks to be recowed, creating many dirty pages in the
117 * extent tree an 4x-6x higher write load than ext3.
118 *
119 * Instead of doing a tree commit on every fsync, we use the
120 * key ranges and transaction ids to find items for a given file or directory
121 * that have changed in this transaction. Those items are copied into
122 * a special tree (one per subvolume root), that tree is written to disk
123 * and then the fsync is considered complete.
124 *
125 * After a crash, items are copied out of the log-tree back into the
126 * subvolume tree. Any file data extents found are recorded in the extent
127 * allocation tree, and the log-tree freed.
128 *
129 * The log tree is read three times, once to pin down all the extents it is
130 * using in ram and once, once to create all the inodes logged in the tree
131 * and once to do all the other items.
132 */
133
134/*
Chris Masone02119d2008-09-05 16:13:11 -0400135 * start a sub transaction and setup the log tree
136 * this increments the log tree writer count to make the people
137 * syncing the tree wait for us to finish
138 */
139static int start_log_trans(struct btrfs_trans_handle *trans,
Miao Xie8b050d32014-02-20 18:08:58 +0800140 struct btrfs_root *root,
141 struct btrfs_log_ctx *ctx)
Chris Masone02119d2008-09-05 16:13:11 -0400142{
Zhaolei34eb2a52015-08-17 18:44:45 +0800143 int ret = 0;
Yan Zheng7237f182009-01-21 12:54:03 -0500144
145 mutex_lock(&root->log_mutex);
Zhaolei34eb2a52015-08-17 18:44:45 +0800146
Yan Zheng7237f182009-01-21 12:54:03 -0500147 if (root->log_root) {
Miao Xie995946d2014-04-02 19:51:06 +0800148 if (btrfs_need_log_full_commit(root->fs_info, trans)) {
Miao Xie50471a32014-02-20 18:08:57 +0800149 ret = -EAGAIN;
150 goto out;
151 }
Zhaolei34eb2a52015-08-17 18:44:45 +0800152
Josef Bacikff782e02009-10-08 15:30:04 -0400153 if (!root->log_start_pid) {
Miao Xie27cdeb72014-04-02 19:51:05 +0800154 clear_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state);
Zhaolei34eb2a52015-08-17 18:44:45 +0800155 root->log_start_pid = current->pid;
Josef Bacikff782e02009-10-08 15:30:04 -0400156 } else if (root->log_start_pid != current->pid) {
Miao Xie27cdeb72014-04-02 19:51:05 +0800157 set_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state);
Josef Bacikff782e02009-10-08 15:30:04 -0400158 }
Zhaolei34eb2a52015-08-17 18:44:45 +0800159 } else {
160 mutex_lock(&root->fs_info->tree_log_mutex);
161 if (!root->fs_info->log_root_tree)
162 ret = btrfs_init_log_root_tree(trans, root->fs_info);
163 mutex_unlock(&root->fs_info->tree_log_mutex);
164 if (ret)
165 goto out;
Josef Bacikff782e02009-10-08 15:30:04 -0400166
Chris Masone02119d2008-09-05 16:13:11 -0400167 ret = btrfs_add_log_tree(trans, root);
Yan, Zheng4a500fd2010-05-16 10:49:59 -0400168 if (ret)
Miao Xiee87ac132014-02-20 18:08:53 +0800169 goto out;
Zhaolei34eb2a52015-08-17 18:44:45 +0800170
171 clear_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state);
172 root->log_start_pid = current->pid;
Chris Masone02119d2008-09-05 16:13:11 -0400173 }
Zhaolei34eb2a52015-08-17 18:44:45 +0800174
Miao Xie2ecb7922012-09-06 04:04:27 -0600175 atomic_inc(&root->log_batch);
Yan Zheng7237f182009-01-21 12:54:03 -0500176 atomic_inc(&root->log_writers);
Miao Xie8b050d32014-02-20 18:08:58 +0800177 if (ctx) {
Zhaolei34eb2a52015-08-17 18:44:45 +0800178 int index = root->log_transid % 2;
Miao Xie8b050d32014-02-20 18:08:58 +0800179 list_add_tail(&ctx->list, &root->log_ctxs[index]);
Miao Xied1433de2014-02-20 18:08:59 +0800180 ctx->log_transid = root->log_transid;
Miao Xie8b050d32014-02-20 18:08:58 +0800181 }
Zhaolei34eb2a52015-08-17 18:44:45 +0800182
Miao Xiee87ac132014-02-20 18:08:53 +0800183out:
Yan Zheng7237f182009-01-21 12:54:03 -0500184 mutex_unlock(&root->log_mutex);
Miao Xiee87ac132014-02-20 18:08:53 +0800185 return ret;
Chris Masone02119d2008-09-05 16:13:11 -0400186}
187
188/*
189 * returns 0 if there was a log transaction running and we were able
190 * to join, or returns -ENOENT if there were not transactions
191 * in progress
192 */
193static int join_running_log_trans(struct btrfs_root *root)
194{
195 int ret = -ENOENT;
196
197 smp_mb();
198 if (!root->log_root)
199 return -ENOENT;
200
Yan Zheng7237f182009-01-21 12:54:03 -0500201 mutex_lock(&root->log_mutex);
Chris Masone02119d2008-09-05 16:13:11 -0400202 if (root->log_root) {
203 ret = 0;
Yan Zheng7237f182009-01-21 12:54:03 -0500204 atomic_inc(&root->log_writers);
Chris Masone02119d2008-09-05 16:13:11 -0400205 }
Yan Zheng7237f182009-01-21 12:54:03 -0500206 mutex_unlock(&root->log_mutex);
Chris Masone02119d2008-09-05 16:13:11 -0400207 return ret;
208}
209
210/*
Chris Mason12fcfd22009-03-24 10:24:20 -0400211 * This either makes the current running log transaction wait
212 * until you call btrfs_end_log_trans() or it makes any future
213 * log transactions wait until you call btrfs_end_log_trans()
214 */
215int btrfs_pin_log_trans(struct btrfs_root *root)
216{
217 int ret = -ENOENT;
218
219 mutex_lock(&root->log_mutex);
220 atomic_inc(&root->log_writers);
221 mutex_unlock(&root->log_mutex);
222 return ret;
223}
224
225/*
Chris Masone02119d2008-09-05 16:13:11 -0400226 * indicate we're done making changes to the log tree
227 * and wake up anyone waiting to do a sync
228 */
Jeff Mahoney143bede2012-03-01 14:56:26 +0100229void btrfs_end_log_trans(struct btrfs_root *root)
Chris Masone02119d2008-09-05 16:13:11 -0400230{
Yan Zheng7237f182009-01-21 12:54:03 -0500231 if (atomic_dec_and_test(&root->log_writers)) {
232 smp_mb();
233 if (waitqueue_active(&root->log_writer_wait))
234 wake_up(&root->log_writer_wait);
235 }
Chris Masone02119d2008-09-05 16:13:11 -0400236}
237
238
239/*
240 * the walk control struct is used to pass state down the chain when
241 * processing the log tree. The stage field tells us which part
242 * of the log tree processing we are currently doing. The others
243 * are state fields used for that specific part
244 */
245struct walk_control {
246 /* should we free the extent on disk when done? This is used
247 * at transaction commit time while freeing a log tree
248 */
249 int free;
250
251 /* should we write out the extent buffer? This is used
252 * while flushing the log tree to disk during a sync
253 */
254 int write;
255
256 /* should we wait for the extent buffer io to finish? Also used
257 * while flushing the log tree to disk for a sync
258 */
259 int wait;
260
261 /* pin only walk, we record which extents on disk belong to the
262 * log trees
263 */
264 int pin;
265
266 /* what stage of the replay code we're currently in */
267 int stage;
268
269 /* the root we are currently replaying */
270 struct btrfs_root *replay_dest;
271
272 /* the trans handle for the current replay */
273 struct btrfs_trans_handle *trans;
274
275 /* the function that gets used to process blocks we find in the
276 * tree. Note the extent_buffer might not be up to date when it is
277 * passed in, and it must be checked or read if you need the data
278 * inside it
279 */
280 int (*process_func)(struct btrfs_root *log, struct extent_buffer *eb,
281 struct walk_control *wc, u64 gen);
282};
283
284/*
285 * process_func used to pin down extents, write them or wait on them
286 */
287static int process_one_buffer(struct btrfs_root *log,
288 struct extent_buffer *eb,
289 struct walk_control *wc, u64 gen)
290{
Josef Bacikb50c6e22013-04-25 15:55:30 -0400291 int ret = 0;
Chris Masone02119d2008-09-05 16:13:11 -0400292
Josef Bacik8c2a1a32013-06-06 13:19:32 -0400293 /*
294 * If this fs is mixed then we need to be able to process the leaves to
295 * pin down any logged extents, so we have to read the block.
296 */
297 if (btrfs_fs_incompat(log->fs_info, MIXED_GROUPS)) {
298 ret = btrfs_read_buffer(eb, gen);
299 if (ret)
300 return ret;
301 }
302
Josef Bacikb50c6e22013-04-25 15:55:30 -0400303 if (wc->pin)
304 ret = btrfs_pin_extent_for_log_replay(log->fs_info->extent_root,
305 eb->start, eb->len);
306
307 if (!ret && btrfs_buffer_uptodate(eb, gen, 0)) {
Josef Bacik8c2a1a32013-06-06 13:19:32 -0400308 if (wc->pin && btrfs_header_level(eb) == 0)
309 ret = btrfs_exclude_logged_extents(log, eb);
Chris Masone02119d2008-09-05 16:13:11 -0400310 if (wc->write)
311 btrfs_write_tree_block(eb);
312 if (wc->wait)
313 btrfs_wait_tree_block_writeback(eb);
314 }
Josef Bacikb50c6e22013-04-25 15:55:30 -0400315 return ret;
Chris Masone02119d2008-09-05 16:13:11 -0400316}
317
318/*
319 * Item overwrite used by replay and tree logging. eb, slot and key all refer
320 * to the src data we are copying out.
321 *
322 * root is the tree we are copying into, and path is a scratch
323 * path for use in this function (it should be released on entry and
324 * will be released on exit).
325 *
326 * If the key is already in the destination tree the existing item is
327 * overwritten. If the existing item isn't big enough, it is extended.
328 * If it is too large, it is truncated.
329 *
330 * If the key isn't in the destination yet, a new item is inserted.
331 */
332static noinline int overwrite_item(struct btrfs_trans_handle *trans,
333 struct btrfs_root *root,
334 struct btrfs_path *path,
335 struct extent_buffer *eb, int slot,
336 struct btrfs_key *key)
337{
338 int ret;
339 u32 item_size;
340 u64 saved_i_size = 0;
341 int save_old_i_size = 0;
342 unsigned long src_ptr;
343 unsigned long dst_ptr;
344 int overwrite_root = 0;
Josef Bacik4bc4bee2013-04-05 20:50:09 +0000345 bool inode_item = key->type == BTRFS_INODE_ITEM_KEY;
Chris Masone02119d2008-09-05 16:13:11 -0400346
347 if (root->root_key.objectid != BTRFS_TREE_LOG_OBJECTID)
348 overwrite_root = 1;
349
350 item_size = btrfs_item_size_nr(eb, slot);
351 src_ptr = btrfs_item_ptr_offset(eb, slot);
352
353 /* look for the key in the destination tree */
354 ret = btrfs_search_slot(NULL, root, key, path, 0, 0);
Josef Bacik4bc4bee2013-04-05 20:50:09 +0000355 if (ret < 0)
356 return ret;
357
Chris Masone02119d2008-09-05 16:13:11 -0400358 if (ret == 0) {
359 char *src_copy;
360 char *dst_copy;
361 u32 dst_size = btrfs_item_size_nr(path->nodes[0],
362 path->slots[0]);
363 if (dst_size != item_size)
364 goto insert;
365
366 if (item_size == 0) {
David Sterbab3b4aa72011-04-21 01:20:15 +0200367 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -0400368 return 0;
369 }
370 dst_copy = kmalloc(item_size, GFP_NOFS);
371 src_copy = kmalloc(item_size, GFP_NOFS);
liubo2a29edc2011-01-26 06:22:08 +0000372 if (!dst_copy || !src_copy) {
David Sterbab3b4aa72011-04-21 01:20:15 +0200373 btrfs_release_path(path);
liubo2a29edc2011-01-26 06:22:08 +0000374 kfree(dst_copy);
375 kfree(src_copy);
376 return -ENOMEM;
377 }
Chris Masone02119d2008-09-05 16:13:11 -0400378
379 read_extent_buffer(eb, src_copy, src_ptr, item_size);
380
381 dst_ptr = btrfs_item_ptr_offset(path->nodes[0], path->slots[0]);
382 read_extent_buffer(path->nodes[0], dst_copy, dst_ptr,
383 item_size);
384 ret = memcmp(dst_copy, src_copy, item_size);
385
386 kfree(dst_copy);
387 kfree(src_copy);
388 /*
389 * they have the same contents, just return, this saves
390 * us from cowing blocks in the destination tree and doing
391 * extra writes that may not have been done by a previous
392 * sync
393 */
394 if (ret == 0) {
David Sterbab3b4aa72011-04-21 01:20:15 +0200395 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -0400396 return 0;
397 }
398
Josef Bacik4bc4bee2013-04-05 20:50:09 +0000399 /*
400 * We need to load the old nbytes into the inode so when we
401 * replay the extents we've logged we get the right nbytes.
402 */
403 if (inode_item) {
404 struct btrfs_inode_item *item;
405 u64 nbytes;
Josef Bacikd5554382013-09-11 14:17:00 -0400406 u32 mode;
Josef Bacik4bc4bee2013-04-05 20:50:09 +0000407
408 item = btrfs_item_ptr(path->nodes[0], path->slots[0],
409 struct btrfs_inode_item);
410 nbytes = btrfs_inode_nbytes(path->nodes[0], item);
411 item = btrfs_item_ptr(eb, slot,
412 struct btrfs_inode_item);
413 btrfs_set_inode_nbytes(eb, item, nbytes);
Josef Bacikd5554382013-09-11 14:17:00 -0400414
415 /*
416 * If this is a directory we need to reset the i_size to
417 * 0 so that we can set it up properly when replaying
418 * the rest of the items in this log.
419 */
420 mode = btrfs_inode_mode(eb, item);
421 if (S_ISDIR(mode))
422 btrfs_set_inode_size(eb, item, 0);
Josef Bacik4bc4bee2013-04-05 20:50:09 +0000423 }
424 } else if (inode_item) {
425 struct btrfs_inode_item *item;
Josef Bacikd5554382013-09-11 14:17:00 -0400426 u32 mode;
Josef Bacik4bc4bee2013-04-05 20:50:09 +0000427
428 /*
429 * New inode, set nbytes to 0 so that the nbytes comes out
430 * properly when we replay the extents.
431 */
432 item = btrfs_item_ptr(eb, slot, struct btrfs_inode_item);
433 btrfs_set_inode_nbytes(eb, item, 0);
Josef Bacikd5554382013-09-11 14:17:00 -0400434
435 /*
436 * If this is a directory we need to reset the i_size to 0 so
437 * that we can set it up properly when replaying the rest of
438 * the items in this log.
439 */
440 mode = btrfs_inode_mode(eb, item);
441 if (S_ISDIR(mode))
442 btrfs_set_inode_size(eb, item, 0);
Chris Masone02119d2008-09-05 16:13:11 -0400443 }
444insert:
David Sterbab3b4aa72011-04-21 01:20:15 +0200445 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -0400446 /* try to insert the key into the destination tree */
Filipe Mananadf8d1162015-01-14 01:52:25 +0000447 path->skip_release_on_error = 1;
Chris Masone02119d2008-09-05 16:13:11 -0400448 ret = btrfs_insert_empty_item(trans, root, path,
449 key, item_size);
Filipe Mananadf8d1162015-01-14 01:52:25 +0000450 path->skip_release_on_error = 0;
Chris Masone02119d2008-09-05 16:13:11 -0400451
452 /* make sure any existing item is the correct size */
Filipe Mananadf8d1162015-01-14 01:52:25 +0000453 if (ret == -EEXIST || ret == -EOVERFLOW) {
Chris Masone02119d2008-09-05 16:13:11 -0400454 u32 found_size;
455 found_size = btrfs_item_size_nr(path->nodes[0],
456 path->slots[0]);
Jeff Mahoney143bede2012-03-01 14:56:26 +0100457 if (found_size > item_size)
Tsutomu Itohafe5fea2013-04-16 05:18:22 +0000458 btrfs_truncate_item(root, path, item_size, 1);
Jeff Mahoney143bede2012-03-01 14:56:26 +0100459 else if (found_size < item_size)
Tsutomu Itoh4b90c682013-04-16 05:18:49 +0000460 btrfs_extend_item(root, path,
Jeff Mahoney143bede2012-03-01 14:56:26 +0100461 item_size - found_size);
Chris Masone02119d2008-09-05 16:13:11 -0400462 } else if (ret) {
Yan, Zheng4a500fd2010-05-16 10:49:59 -0400463 return ret;
Chris Masone02119d2008-09-05 16:13:11 -0400464 }
465 dst_ptr = btrfs_item_ptr_offset(path->nodes[0],
466 path->slots[0]);
467
468 /* don't overwrite an existing inode if the generation number
469 * was logged as zero. This is done when the tree logging code
470 * is just logging an inode to make sure it exists after recovery.
471 *
472 * Also, don't overwrite i_size on directories during replay.
473 * log replay inserts and removes directory items based on the
474 * state of the tree found in the subvolume, and i_size is modified
475 * as it goes
476 */
477 if (key->type == BTRFS_INODE_ITEM_KEY && ret == -EEXIST) {
478 struct btrfs_inode_item *src_item;
479 struct btrfs_inode_item *dst_item;
480
481 src_item = (struct btrfs_inode_item *)src_ptr;
482 dst_item = (struct btrfs_inode_item *)dst_ptr;
483
Filipe Manana1a4bcf42015-02-13 12:30:56 +0000484 if (btrfs_inode_generation(eb, src_item) == 0) {
485 struct extent_buffer *dst_eb = path->nodes[0];
Filipe Manana2f2ff0e2015-03-20 17:19:46 +0000486 const u64 ino_size = btrfs_inode_size(eb, src_item);
Filipe Manana1a4bcf42015-02-13 12:30:56 +0000487
Filipe Manana2f2ff0e2015-03-20 17:19:46 +0000488 /*
489 * For regular files an ino_size == 0 is used only when
490 * logging that an inode exists, as part of a directory
491 * fsync, and the inode wasn't fsynced before. In this
492 * case don't set the size of the inode in the fs/subvol
493 * tree, otherwise we would be throwing valid data away.
494 */
Filipe Manana1a4bcf42015-02-13 12:30:56 +0000495 if (S_ISREG(btrfs_inode_mode(eb, src_item)) &&
Filipe Manana2f2ff0e2015-03-20 17:19:46 +0000496 S_ISREG(btrfs_inode_mode(dst_eb, dst_item)) &&
497 ino_size != 0) {
Filipe Manana1a4bcf42015-02-13 12:30:56 +0000498 struct btrfs_map_token token;
Filipe Manana1a4bcf42015-02-13 12:30:56 +0000499
500 btrfs_init_map_token(&token);
501 btrfs_set_token_inode_size(dst_eb, dst_item,
502 ino_size, &token);
503 }
Chris Masone02119d2008-09-05 16:13:11 -0400504 goto no_copy;
Filipe Manana1a4bcf42015-02-13 12:30:56 +0000505 }
Chris Masone02119d2008-09-05 16:13:11 -0400506
507 if (overwrite_root &&
508 S_ISDIR(btrfs_inode_mode(eb, src_item)) &&
509 S_ISDIR(btrfs_inode_mode(path->nodes[0], dst_item))) {
510 save_old_i_size = 1;
511 saved_i_size = btrfs_inode_size(path->nodes[0],
512 dst_item);
513 }
514 }
515
516 copy_extent_buffer(path->nodes[0], eb, dst_ptr,
517 src_ptr, item_size);
518
519 if (save_old_i_size) {
520 struct btrfs_inode_item *dst_item;
521 dst_item = (struct btrfs_inode_item *)dst_ptr;
522 btrfs_set_inode_size(path->nodes[0], dst_item, saved_i_size);
523 }
524
525 /* make sure the generation is filled in */
526 if (key->type == BTRFS_INODE_ITEM_KEY) {
527 struct btrfs_inode_item *dst_item;
528 dst_item = (struct btrfs_inode_item *)dst_ptr;
529 if (btrfs_inode_generation(path->nodes[0], dst_item) == 0) {
530 btrfs_set_inode_generation(path->nodes[0], dst_item,
531 trans->transid);
532 }
533 }
534no_copy:
535 btrfs_mark_buffer_dirty(path->nodes[0]);
David Sterbab3b4aa72011-04-21 01:20:15 +0200536 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -0400537 return 0;
538}
539
540/*
541 * simple helper to read an inode off the disk from a given root
542 * This can only be called for subvolume roots and not for the log
543 */
544static noinline struct inode *read_one_inode(struct btrfs_root *root,
545 u64 objectid)
546{
Yan Zheng5d4f98a2009-06-10 10:45:14 -0400547 struct btrfs_key key;
Chris Masone02119d2008-09-05 16:13:11 -0400548 struct inode *inode;
Chris Masone02119d2008-09-05 16:13:11 -0400549
Yan Zheng5d4f98a2009-06-10 10:45:14 -0400550 key.objectid = objectid;
551 key.type = BTRFS_INODE_ITEM_KEY;
552 key.offset = 0;
Josef Bacik73f73412009-12-04 17:38:27 +0000553 inode = btrfs_iget(root->fs_info->sb, &key, root, NULL);
Yan Zheng5d4f98a2009-06-10 10:45:14 -0400554 if (IS_ERR(inode)) {
555 inode = NULL;
556 } else if (is_bad_inode(inode)) {
Chris Masone02119d2008-09-05 16:13:11 -0400557 iput(inode);
558 inode = NULL;
559 }
560 return inode;
561}
562
563/* replays a single extent in 'eb' at 'slot' with 'key' into the
564 * subvolume 'root'. path is released on entry and should be released
565 * on exit.
566 *
567 * extents in the log tree have not been allocated out of the extent
568 * tree yet. So, this completes the allocation, taking a reference
569 * as required if the extent already exists or creating a new extent
570 * if it isn't in the extent allocation tree yet.
571 *
572 * The extent is inserted into the file, dropping any existing extents
573 * from the file that overlap the new one.
574 */
575static noinline int replay_one_extent(struct btrfs_trans_handle *trans,
576 struct btrfs_root *root,
577 struct btrfs_path *path,
578 struct extent_buffer *eb, int slot,
579 struct btrfs_key *key)
580{
581 int found_type;
Chris Masone02119d2008-09-05 16:13:11 -0400582 u64 extent_end;
Chris Masone02119d2008-09-05 16:13:11 -0400583 u64 start = key->offset;
Josef Bacik4bc4bee2013-04-05 20:50:09 +0000584 u64 nbytes = 0;
Chris Masone02119d2008-09-05 16:13:11 -0400585 struct btrfs_file_extent_item *item;
586 struct inode *inode = NULL;
587 unsigned long size;
588 int ret = 0;
589
590 item = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item);
591 found_type = btrfs_file_extent_type(eb, item);
592
Yan Zhengd899e052008-10-30 14:25:28 -0400593 if (found_type == BTRFS_FILE_EXTENT_REG ||
Josef Bacik4bc4bee2013-04-05 20:50:09 +0000594 found_type == BTRFS_FILE_EXTENT_PREALLOC) {
595 nbytes = btrfs_file_extent_num_bytes(eb, item);
596 extent_end = start + nbytes;
597
598 /*
599 * We don't add to the inodes nbytes if we are prealloc or a
600 * hole.
601 */
602 if (btrfs_file_extent_disk_bytenr(eb, item) == 0)
603 nbytes = 0;
604 } else if (found_type == BTRFS_FILE_EXTENT_INLINE) {
Chris Mason514ac8a2014-01-03 21:07:00 -0800605 size = btrfs_file_extent_inline_len(eb, slot, item);
Josef Bacik4bc4bee2013-04-05 20:50:09 +0000606 nbytes = btrfs_file_extent_ram_bytes(eb, item);
Qu Wenruofda28322013-02-26 08:10:22 +0000607 extent_end = ALIGN(start + size, root->sectorsize);
Chris Masone02119d2008-09-05 16:13:11 -0400608 } else {
609 ret = 0;
610 goto out;
611 }
612
613 inode = read_one_inode(root, key->objectid);
614 if (!inode) {
615 ret = -EIO;
616 goto out;
617 }
618
619 /*
620 * first check to see if we already have this extent in the
621 * file. This must be done before the btrfs_drop_extents run
622 * so we don't try to drop this extent.
623 */
Li Zefan33345d012011-04-20 10:31:50 +0800624 ret = btrfs_lookup_file_extent(trans, root, path, btrfs_ino(inode),
Chris Masone02119d2008-09-05 16:13:11 -0400625 start, 0);
626
Yan Zhengd899e052008-10-30 14:25:28 -0400627 if (ret == 0 &&
628 (found_type == BTRFS_FILE_EXTENT_REG ||
629 found_type == BTRFS_FILE_EXTENT_PREALLOC)) {
Chris Masone02119d2008-09-05 16:13:11 -0400630 struct btrfs_file_extent_item cmp1;
631 struct btrfs_file_extent_item cmp2;
632 struct btrfs_file_extent_item *existing;
633 struct extent_buffer *leaf;
634
635 leaf = path->nodes[0];
636 existing = btrfs_item_ptr(leaf, path->slots[0],
637 struct btrfs_file_extent_item);
638
639 read_extent_buffer(eb, &cmp1, (unsigned long)item,
640 sizeof(cmp1));
641 read_extent_buffer(leaf, &cmp2, (unsigned long)existing,
642 sizeof(cmp2));
643
644 /*
645 * we already have a pointer to this exact extent,
646 * we don't have to do anything
647 */
648 if (memcmp(&cmp1, &cmp2, sizeof(cmp1)) == 0) {
David Sterbab3b4aa72011-04-21 01:20:15 +0200649 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -0400650 goto out;
651 }
652 }
David Sterbab3b4aa72011-04-21 01:20:15 +0200653 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -0400654
655 /* drop any overlapping extents */
Josef Bacik26714852012-08-29 12:24:27 -0400656 ret = btrfs_drop_extents(trans, root, inode, start, extent_end, 1);
Josef Bacik36508602013-04-25 16:23:32 -0400657 if (ret)
658 goto out;
Chris Masone02119d2008-09-05 16:13:11 -0400659
Yan Zheng07d400a2009-01-06 11:42:00 -0500660 if (found_type == BTRFS_FILE_EXTENT_REG ||
661 found_type == BTRFS_FILE_EXTENT_PREALLOC) {
Yan Zheng5d4f98a2009-06-10 10:45:14 -0400662 u64 offset;
Yan Zheng07d400a2009-01-06 11:42:00 -0500663 unsigned long dest_offset;
664 struct btrfs_key ins;
Chris Masone02119d2008-09-05 16:13:11 -0400665
Yan Zheng07d400a2009-01-06 11:42:00 -0500666 ret = btrfs_insert_empty_item(trans, root, path, key,
667 sizeof(*item));
Josef Bacik36508602013-04-25 16:23:32 -0400668 if (ret)
669 goto out;
Yan Zheng07d400a2009-01-06 11:42:00 -0500670 dest_offset = btrfs_item_ptr_offset(path->nodes[0],
671 path->slots[0]);
672 copy_extent_buffer(path->nodes[0], eb, dest_offset,
673 (unsigned long)item, sizeof(*item));
674
675 ins.objectid = btrfs_file_extent_disk_bytenr(eb, item);
676 ins.offset = btrfs_file_extent_disk_num_bytes(eb, item);
677 ins.type = BTRFS_EXTENT_ITEM_KEY;
Yan Zheng5d4f98a2009-06-10 10:45:14 -0400678 offset = key->offset - btrfs_file_extent_offset(eb, item);
Yan Zheng07d400a2009-01-06 11:42:00 -0500679
680 if (ins.objectid > 0) {
681 u64 csum_start;
682 u64 csum_end;
683 LIST_HEAD(ordered_sums);
684 /*
685 * is this extent already allocated in the extent
686 * allocation tree? If so, just add a reference
687 */
Filipe Manana1a4ed8f2014-10-27 10:44:24 +0000688 ret = btrfs_lookup_data_extent(root, ins.objectid,
Yan Zheng07d400a2009-01-06 11:42:00 -0500689 ins.offset);
690 if (ret == 0) {
691 ret = btrfs_inc_extent_ref(trans, root,
692 ins.objectid, ins.offset,
Yan Zheng5d4f98a2009-06-10 10:45:14 -0400693 0, root->root_key.objectid,
Arne Jansen66d7e7f2011-09-12 15:26:38 +0200694 key->objectid, offset, 0);
Josef Bacikb50c6e22013-04-25 15:55:30 -0400695 if (ret)
696 goto out;
Yan Zheng07d400a2009-01-06 11:42:00 -0500697 } else {
698 /*
699 * insert the extent pointer in the extent
700 * allocation tree
701 */
Yan Zheng5d4f98a2009-06-10 10:45:14 -0400702 ret = btrfs_alloc_logged_file_extent(trans,
703 root, root->root_key.objectid,
704 key->objectid, offset, &ins);
Josef Bacikb50c6e22013-04-25 15:55:30 -0400705 if (ret)
706 goto out;
Yan Zheng07d400a2009-01-06 11:42:00 -0500707 }
David Sterbab3b4aa72011-04-21 01:20:15 +0200708 btrfs_release_path(path);
Yan Zheng07d400a2009-01-06 11:42:00 -0500709
710 if (btrfs_file_extent_compression(eb, item)) {
711 csum_start = ins.objectid;
712 csum_end = csum_start + ins.offset;
713 } else {
714 csum_start = ins.objectid +
715 btrfs_file_extent_offset(eb, item);
716 csum_end = csum_start +
717 btrfs_file_extent_num_bytes(eb, item);
718 }
719
720 ret = btrfs_lookup_csums_range(root->log_root,
721 csum_start, csum_end - 1,
Arne Jansena2de7332011-03-08 14:14:00 +0100722 &ordered_sums, 0);
Josef Bacik36508602013-04-25 16:23:32 -0400723 if (ret)
724 goto out;
Filipe Mananab84b8392015-08-19 11:09:40 +0100725 /*
726 * Now delete all existing cums in the csum root that
727 * cover our range. We do this because we can have an
728 * extent that is completely referenced by one file
729 * extent item and partially referenced by another
730 * file extent item (like after using the clone or
731 * extent_same ioctls). In this case if we end up doing
732 * the replay of the one that partially references the
733 * extent first, and we do not do the csum deletion
734 * below, we can get 2 csum items in the csum tree that
735 * overlap each other. For example, imagine our log has
736 * the two following file extent items:
737 *
738 * key (257 EXTENT_DATA 409600)
739 * extent data disk byte 12845056 nr 102400
740 * extent data offset 20480 nr 20480 ram 102400
741 *
742 * key (257 EXTENT_DATA 819200)
743 * extent data disk byte 12845056 nr 102400
744 * extent data offset 0 nr 102400 ram 102400
745 *
746 * Where the second one fully references the 100K extent
747 * that starts at disk byte 12845056, and the log tree
748 * has a single csum item that covers the entire range
749 * of the extent:
750 *
751 * key (EXTENT_CSUM EXTENT_CSUM 12845056) itemsize 100
752 *
753 * After the first file extent item is replayed, the
754 * csum tree gets the following csum item:
755 *
756 * key (EXTENT_CSUM EXTENT_CSUM 12865536) itemsize 20
757 *
758 * Which covers the 20K sub-range starting at offset 20K
759 * of our extent. Now when we replay the second file
760 * extent item, if we do not delete existing csum items
761 * that cover any of its blocks, we end up getting two
762 * csum items in our csum tree that overlap each other:
763 *
764 * key (EXTENT_CSUM EXTENT_CSUM 12845056) itemsize 100
765 * key (EXTENT_CSUM EXTENT_CSUM 12865536) itemsize 20
766 *
767 * Which is a problem, because after this anyone trying
768 * to lookup up for the checksum of any block of our
769 * extent starting at an offset of 40K or higher, will
770 * end up looking at the second csum item only, which
771 * does not contain the checksum for any block starting
772 * at offset 40K or higher of our extent.
773 */
Yan Zheng07d400a2009-01-06 11:42:00 -0500774 while (!list_empty(&ordered_sums)) {
775 struct btrfs_ordered_sum *sums;
776 sums = list_entry(ordered_sums.next,
777 struct btrfs_ordered_sum,
778 list);
Josef Bacik36508602013-04-25 16:23:32 -0400779 if (!ret)
Filipe Mananab84b8392015-08-19 11:09:40 +0100780 ret = btrfs_del_csums(trans,
781 root->fs_info->csum_root,
782 sums->bytenr,
783 sums->len);
784 if (!ret)
Josef Bacik36508602013-04-25 16:23:32 -0400785 ret = btrfs_csum_file_blocks(trans,
Yan Zheng07d400a2009-01-06 11:42:00 -0500786 root->fs_info->csum_root,
787 sums);
Yan Zheng07d400a2009-01-06 11:42:00 -0500788 list_del(&sums->list);
789 kfree(sums);
790 }
Josef Bacik36508602013-04-25 16:23:32 -0400791 if (ret)
792 goto out;
Yan Zheng07d400a2009-01-06 11:42:00 -0500793 } else {
David Sterbab3b4aa72011-04-21 01:20:15 +0200794 btrfs_release_path(path);
Yan Zheng07d400a2009-01-06 11:42:00 -0500795 }
796 } else if (found_type == BTRFS_FILE_EXTENT_INLINE) {
797 /* inline extents are easy, we just overwrite them */
798 ret = overwrite_item(trans, root, path, eb, slot, key);
Josef Bacik36508602013-04-25 16:23:32 -0400799 if (ret)
800 goto out;
Yan Zheng07d400a2009-01-06 11:42:00 -0500801 }
802
Josef Bacik4bc4bee2013-04-05 20:50:09 +0000803 inode_add_bytes(inode, nbytes);
Tsutomu Itohb9959292012-06-25 21:25:22 -0600804 ret = btrfs_update_inode(trans, root, inode);
Chris Masone02119d2008-09-05 16:13:11 -0400805out:
806 if (inode)
807 iput(inode);
808 return ret;
809}
810
811/*
812 * when cleaning up conflicts between the directory names in the
813 * subvolume, directory names in the log and directory names in the
814 * inode back references, we may have to unlink inodes from directories.
815 *
816 * This is a helper function to do the unlink of a specific directory
817 * item
818 */
819static noinline int drop_one_dir_item(struct btrfs_trans_handle *trans,
820 struct btrfs_root *root,
821 struct btrfs_path *path,
822 struct inode *dir,
823 struct btrfs_dir_item *di)
824{
825 struct inode *inode;
826 char *name;
827 int name_len;
828 struct extent_buffer *leaf;
829 struct btrfs_key location;
830 int ret;
831
832 leaf = path->nodes[0];
833
834 btrfs_dir_item_key_to_cpu(leaf, di, &location);
835 name_len = btrfs_dir_name_len(leaf, di);
836 name = kmalloc(name_len, GFP_NOFS);
liubo2a29edc2011-01-26 06:22:08 +0000837 if (!name)
838 return -ENOMEM;
839
Chris Masone02119d2008-09-05 16:13:11 -0400840 read_extent_buffer(leaf, name, (unsigned long)(di + 1), name_len);
David Sterbab3b4aa72011-04-21 01:20:15 +0200841 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -0400842
843 inode = read_one_inode(root, location.objectid);
Tsutomu Itohc00e9492011-04-28 09:10:23 +0000844 if (!inode) {
Josef Bacik36508602013-04-25 16:23:32 -0400845 ret = -EIO;
846 goto out;
Tsutomu Itohc00e9492011-04-28 09:10:23 +0000847 }
Chris Masone02119d2008-09-05 16:13:11 -0400848
Yan Zhengec051c02009-01-05 15:43:42 -0500849 ret = link_to_fixup_dir(trans, root, path, location.objectid);
Josef Bacik36508602013-04-25 16:23:32 -0400850 if (ret)
851 goto out;
Chris Mason12fcfd22009-03-24 10:24:20 -0400852
Chris Masone02119d2008-09-05 16:13:11 -0400853 ret = btrfs_unlink_inode(trans, root, dir, inode, name, name_len);
Josef Bacik36508602013-04-25 16:23:32 -0400854 if (ret)
855 goto out;
Filipe David Borba Mananaada9af22013-08-05 09:25:47 +0100856 else
857 ret = btrfs_run_delayed_items(trans, root);
Josef Bacik36508602013-04-25 16:23:32 -0400858out:
859 kfree(name);
860 iput(inode);
Chris Masone02119d2008-09-05 16:13:11 -0400861 return ret;
862}
863
864/*
865 * helper function to see if a given name and sequence number found
866 * in an inode back reference are already in a directory and correctly
867 * point to this inode
868 */
869static noinline int inode_in_dir(struct btrfs_root *root,
870 struct btrfs_path *path,
871 u64 dirid, u64 objectid, u64 index,
872 const char *name, int name_len)
873{
874 struct btrfs_dir_item *di;
875 struct btrfs_key location;
876 int match = 0;
877
878 di = btrfs_lookup_dir_index_item(NULL, root, path, dirid,
879 index, name, name_len, 0);
880 if (di && !IS_ERR(di)) {
881 btrfs_dir_item_key_to_cpu(path->nodes[0], di, &location);
882 if (location.objectid != objectid)
883 goto out;
884 } else
885 goto out;
David Sterbab3b4aa72011-04-21 01:20:15 +0200886 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -0400887
888 di = btrfs_lookup_dir_item(NULL, root, path, dirid, name, name_len, 0);
889 if (di && !IS_ERR(di)) {
890 btrfs_dir_item_key_to_cpu(path->nodes[0], di, &location);
891 if (location.objectid != objectid)
892 goto out;
893 } else
894 goto out;
895 match = 1;
896out:
David Sterbab3b4aa72011-04-21 01:20:15 +0200897 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -0400898 return match;
899}
900
901/*
902 * helper function to check a log tree for a named back reference in
903 * an inode. This is used to decide if a back reference that is
904 * found in the subvolume conflicts with what we find in the log.
905 *
906 * inode backreferences may have multiple refs in a single item,
907 * during replay we process one reference at a time, and we don't
908 * want to delete valid links to a file from the subvolume if that
909 * link is also in the log.
910 */
911static noinline int backref_in_log(struct btrfs_root *log,
912 struct btrfs_key *key,
Mark Fashehf1863732012-08-08 11:32:27 -0700913 u64 ref_objectid,
Filipe Mananadf8d1162015-01-14 01:52:25 +0000914 const char *name, int namelen)
Chris Masone02119d2008-09-05 16:13:11 -0400915{
916 struct btrfs_path *path;
917 struct btrfs_inode_ref *ref;
918 unsigned long ptr;
919 unsigned long ptr_end;
920 unsigned long name_ptr;
921 int found_name_len;
922 int item_size;
923 int ret;
924 int match = 0;
925
926 path = btrfs_alloc_path();
liubo2a29edc2011-01-26 06:22:08 +0000927 if (!path)
928 return -ENOMEM;
929
Chris Masone02119d2008-09-05 16:13:11 -0400930 ret = btrfs_search_slot(NULL, log, key, path, 0, 0);
931 if (ret != 0)
932 goto out;
933
Chris Masone02119d2008-09-05 16:13:11 -0400934 ptr = btrfs_item_ptr_offset(path->nodes[0], path->slots[0]);
Mark Fashehf1863732012-08-08 11:32:27 -0700935
936 if (key->type == BTRFS_INODE_EXTREF_KEY) {
937 if (btrfs_find_name_in_ext_backref(path, ref_objectid,
938 name, namelen, NULL))
939 match = 1;
940
941 goto out;
942 }
943
944 item_size = btrfs_item_size_nr(path->nodes[0], path->slots[0]);
Chris Masone02119d2008-09-05 16:13:11 -0400945 ptr_end = ptr + item_size;
946 while (ptr < ptr_end) {
947 ref = (struct btrfs_inode_ref *)ptr;
948 found_name_len = btrfs_inode_ref_name_len(path->nodes[0], ref);
949 if (found_name_len == namelen) {
950 name_ptr = (unsigned long)(ref + 1);
951 ret = memcmp_extent_buffer(path->nodes[0], name,
952 name_ptr, namelen);
953 if (ret == 0) {
954 match = 1;
955 goto out;
956 }
957 }
958 ptr = (unsigned long)(ref + 1) + found_name_len;
959 }
960out:
961 btrfs_free_path(path);
962 return match;
963}
964
Jan Schmidt5a1d7842012-08-17 14:04:41 -0700965static inline int __add_inode_ref(struct btrfs_trans_handle *trans,
966 struct btrfs_root *root,
967 struct btrfs_path *path,
968 struct btrfs_root *log_root,
969 struct inode *dir, struct inode *inode,
Jan Schmidt5a1d7842012-08-17 14:04:41 -0700970 struct extent_buffer *eb,
Mark Fashehf1863732012-08-08 11:32:27 -0700971 u64 inode_objectid, u64 parent_objectid,
972 u64 ref_index, char *name, int namelen,
973 int *search_done)
Jan Schmidt5a1d7842012-08-17 14:04:41 -0700974{
975 int ret;
Mark Fashehf1863732012-08-08 11:32:27 -0700976 char *victim_name;
977 int victim_name_len;
978 struct extent_buffer *leaf;
Jan Schmidt5a1d7842012-08-17 14:04:41 -0700979 struct btrfs_dir_item *di;
Mark Fashehf1863732012-08-08 11:32:27 -0700980 struct btrfs_key search_key;
981 struct btrfs_inode_extref *extref;
Jan Schmidt5a1d7842012-08-17 14:04:41 -0700982
Mark Fashehf1863732012-08-08 11:32:27 -0700983again:
984 /* Search old style refs */
985 search_key.objectid = inode_objectid;
986 search_key.type = BTRFS_INODE_REF_KEY;
987 search_key.offset = parent_objectid;
988 ret = btrfs_search_slot(NULL, root, &search_key, path, 0, 0);
Jan Schmidt5a1d7842012-08-17 14:04:41 -0700989 if (ret == 0) {
Jan Schmidt5a1d7842012-08-17 14:04:41 -0700990 struct btrfs_inode_ref *victim_ref;
991 unsigned long ptr;
992 unsigned long ptr_end;
Mark Fashehf1863732012-08-08 11:32:27 -0700993
994 leaf = path->nodes[0];
Jan Schmidt5a1d7842012-08-17 14:04:41 -0700995
996 /* are we trying to overwrite a back ref for the root directory
997 * if so, just jump out, we're done
998 */
Mark Fashehf1863732012-08-08 11:32:27 -0700999 if (search_key.objectid == search_key.offset)
Jan Schmidt5a1d7842012-08-17 14:04:41 -07001000 return 1;
1001
1002 /* check all the names in this back reference to see
1003 * if they are in the log. if so, we allow them to stay
1004 * otherwise they must be unlinked as a conflict
1005 */
1006 ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
1007 ptr_end = ptr + btrfs_item_size_nr(leaf, path->slots[0]);
1008 while (ptr < ptr_end) {
1009 victim_ref = (struct btrfs_inode_ref *)ptr;
1010 victim_name_len = btrfs_inode_ref_name_len(leaf,
1011 victim_ref);
1012 victim_name = kmalloc(victim_name_len, GFP_NOFS);
Josef Bacik36508602013-04-25 16:23:32 -04001013 if (!victim_name)
1014 return -ENOMEM;
Jan Schmidt5a1d7842012-08-17 14:04:41 -07001015
1016 read_extent_buffer(leaf, victim_name,
1017 (unsigned long)(victim_ref + 1),
1018 victim_name_len);
1019
Mark Fashehf1863732012-08-08 11:32:27 -07001020 if (!backref_in_log(log_root, &search_key,
1021 parent_objectid,
1022 victim_name,
Jan Schmidt5a1d7842012-08-17 14:04:41 -07001023 victim_name_len)) {
Zach Brown8b558c52013-10-16 12:10:34 -07001024 inc_nlink(inode);
Jan Schmidt5a1d7842012-08-17 14:04:41 -07001025 btrfs_release_path(path);
1026
1027 ret = btrfs_unlink_inode(trans, root, dir,
1028 inode, victim_name,
1029 victim_name_len);
Mark Fashehf1863732012-08-08 11:32:27 -07001030 kfree(victim_name);
Josef Bacik36508602013-04-25 16:23:32 -04001031 if (ret)
1032 return ret;
Filipe David Borba Mananaada9af22013-08-05 09:25:47 +01001033 ret = btrfs_run_delayed_items(trans, root);
1034 if (ret)
1035 return ret;
Mark Fashehf1863732012-08-08 11:32:27 -07001036 *search_done = 1;
1037 goto again;
Jan Schmidt5a1d7842012-08-17 14:04:41 -07001038 }
1039 kfree(victim_name);
Mark Fashehf1863732012-08-08 11:32:27 -07001040
Jan Schmidt5a1d7842012-08-17 14:04:41 -07001041 ptr = (unsigned long)(victim_ref + 1) + victim_name_len;
1042 }
Jan Schmidt5a1d7842012-08-17 14:04:41 -07001043
1044 /*
1045 * NOTE: we have searched root tree and checked the
1046 * coresponding ref, it does not need to check again.
1047 */
1048 *search_done = 1;
1049 }
1050 btrfs_release_path(path);
1051
Mark Fashehf1863732012-08-08 11:32:27 -07001052 /* Same search but for extended refs */
1053 extref = btrfs_lookup_inode_extref(NULL, root, path, name, namelen,
1054 inode_objectid, parent_objectid, 0,
1055 0);
1056 if (!IS_ERR_OR_NULL(extref)) {
1057 u32 item_size;
1058 u32 cur_offset = 0;
1059 unsigned long base;
1060 struct inode *victim_parent;
1061
1062 leaf = path->nodes[0];
1063
1064 item_size = btrfs_item_size_nr(leaf, path->slots[0]);
1065 base = btrfs_item_ptr_offset(leaf, path->slots[0]);
1066
1067 while (cur_offset < item_size) {
Quentin Casasnovasdd9ef132015-03-03 16:31:38 +01001068 extref = (struct btrfs_inode_extref *)(base + cur_offset);
Mark Fashehf1863732012-08-08 11:32:27 -07001069
1070 victim_name_len = btrfs_inode_extref_name_len(leaf, extref);
1071
1072 if (btrfs_inode_extref_parent(leaf, extref) != parent_objectid)
1073 goto next;
1074
1075 victim_name = kmalloc(victim_name_len, GFP_NOFS);
Josef Bacik36508602013-04-25 16:23:32 -04001076 if (!victim_name)
1077 return -ENOMEM;
Mark Fashehf1863732012-08-08 11:32:27 -07001078 read_extent_buffer(leaf, victim_name, (unsigned long)&extref->name,
1079 victim_name_len);
1080
1081 search_key.objectid = inode_objectid;
1082 search_key.type = BTRFS_INODE_EXTREF_KEY;
1083 search_key.offset = btrfs_extref_hash(parent_objectid,
1084 victim_name,
1085 victim_name_len);
1086 ret = 0;
1087 if (!backref_in_log(log_root, &search_key,
1088 parent_objectid, victim_name,
1089 victim_name_len)) {
1090 ret = -ENOENT;
1091 victim_parent = read_one_inode(root,
1092 parent_objectid);
1093 if (victim_parent) {
Zach Brown8b558c52013-10-16 12:10:34 -07001094 inc_nlink(inode);
Mark Fashehf1863732012-08-08 11:32:27 -07001095 btrfs_release_path(path);
1096
1097 ret = btrfs_unlink_inode(trans, root,
1098 victim_parent,
1099 inode,
1100 victim_name,
1101 victim_name_len);
Filipe David Borba Mananaada9af22013-08-05 09:25:47 +01001102 if (!ret)
1103 ret = btrfs_run_delayed_items(
1104 trans, root);
Mark Fashehf1863732012-08-08 11:32:27 -07001105 }
Mark Fashehf1863732012-08-08 11:32:27 -07001106 iput(victim_parent);
1107 kfree(victim_name);
Josef Bacik36508602013-04-25 16:23:32 -04001108 if (ret)
1109 return ret;
Mark Fashehf1863732012-08-08 11:32:27 -07001110 *search_done = 1;
1111 goto again;
1112 }
1113 kfree(victim_name);
Josef Bacik36508602013-04-25 16:23:32 -04001114 if (ret)
1115 return ret;
Mark Fashehf1863732012-08-08 11:32:27 -07001116next:
1117 cur_offset += victim_name_len + sizeof(*extref);
1118 }
1119 *search_done = 1;
1120 }
1121 btrfs_release_path(path);
1122
Jan Schmidt5a1d7842012-08-17 14:04:41 -07001123 /* look for a conflicting sequence number */
1124 di = btrfs_lookup_dir_index_item(trans, root, path, btrfs_ino(dir),
Mark Fashehf1863732012-08-08 11:32:27 -07001125 ref_index, name, namelen, 0);
Jan Schmidt5a1d7842012-08-17 14:04:41 -07001126 if (di && !IS_ERR(di)) {
1127 ret = drop_one_dir_item(trans, root, path, dir, di);
Josef Bacik36508602013-04-25 16:23:32 -04001128 if (ret)
1129 return ret;
Jan Schmidt5a1d7842012-08-17 14:04:41 -07001130 }
1131 btrfs_release_path(path);
1132
1133 /* look for a conflicing name */
1134 di = btrfs_lookup_dir_item(trans, root, path, btrfs_ino(dir),
1135 name, namelen, 0);
1136 if (di && !IS_ERR(di)) {
1137 ret = drop_one_dir_item(trans, root, path, dir, di);
Josef Bacik36508602013-04-25 16:23:32 -04001138 if (ret)
1139 return ret;
Jan Schmidt5a1d7842012-08-17 14:04:41 -07001140 }
1141 btrfs_release_path(path);
1142
1143 return 0;
1144}
Chris Masone02119d2008-09-05 16:13:11 -04001145
Mark Fashehf1863732012-08-08 11:32:27 -07001146static int extref_get_fields(struct extent_buffer *eb, unsigned long ref_ptr,
1147 u32 *namelen, char **name, u64 *index,
1148 u64 *parent_objectid)
1149{
1150 struct btrfs_inode_extref *extref;
1151
1152 extref = (struct btrfs_inode_extref *)ref_ptr;
1153
1154 *namelen = btrfs_inode_extref_name_len(eb, extref);
1155 *name = kmalloc(*namelen, GFP_NOFS);
1156 if (*name == NULL)
1157 return -ENOMEM;
1158
1159 read_extent_buffer(eb, *name, (unsigned long)&extref->name,
1160 *namelen);
1161
1162 *index = btrfs_inode_extref_index(eb, extref);
1163 if (parent_objectid)
1164 *parent_objectid = btrfs_inode_extref_parent(eb, extref);
1165
1166 return 0;
1167}
1168
1169static int ref_get_fields(struct extent_buffer *eb, unsigned long ref_ptr,
1170 u32 *namelen, char **name, u64 *index)
1171{
1172 struct btrfs_inode_ref *ref;
1173
1174 ref = (struct btrfs_inode_ref *)ref_ptr;
1175
1176 *namelen = btrfs_inode_ref_name_len(eb, ref);
1177 *name = kmalloc(*namelen, GFP_NOFS);
1178 if (*name == NULL)
1179 return -ENOMEM;
1180
1181 read_extent_buffer(eb, *name, (unsigned long)(ref + 1), *namelen);
1182
1183 *index = btrfs_inode_ref_index(eb, ref);
1184
1185 return 0;
1186}
1187
Chris Masone02119d2008-09-05 16:13:11 -04001188/*
1189 * replay one inode back reference item found in the log tree.
1190 * eb, slot and key refer to the buffer and key found in the log tree.
1191 * root is the destination we are replaying into, and path is for temp
1192 * use by this function. (it should be released on return).
1193 */
1194static noinline int add_inode_ref(struct btrfs_trans_handle *trans,
1195 struct btrfs_root *root,
1196 struct btrfs_root *log,
1197 struct btrfs_path *path,
1198 struct extent_buffer *eb, int slot,
1199 struct btrfs_key *key)
1200{
Geyslan G. Bem03b2f082013-10-11 15:35:45 -03001201 struct inode *dir = NULL;
1202 struct inode *inode = NULL;
Chris Masone02119d2008-09-05 16:13:11 -04001203 unsigned long ref_ptr;
1204 unsigned long ref_end;
Geyslan G. Bem03b2f082013-10-11 15:35:45 -03001205 char *name = NULL;
liubo34f3e4f2011-08-06 08:35:23 +00001206 int namelen;
1207 int ret;
liuboc622ae62011-03-26 08:01:12 -04001208 int search_done = 0;
Mark Fashehf1863732012-08-08 11:32:27 -07001209 int log_ref_ver = 0;
1210 u64 parent_objectid;
1211 u64 inode_objectid;
Chris Masonf46dbe3de2012-10-09 11:17:20 -04001212 u64 ref_index = 0;
Mark Fashehf1863732012-08-08 11:32:27 -07001213 int ref_struct_size;
1214
1215 ref_ptr = btrfs_item_ptr_offset(eb, slot);
1216 ref_end = ref_ptr + btrfs_item_size_nr(eb, slot);
1217
1218 if (key->type == BTRFS_INODE_EXTREF_KEY) {
1219 struct btrfs_inode_extref *r;
1220
1221 ref_struct_size = sizeof(struct btrfs_inode_extref);
1222 log_ref_ver = 1;
1223 r = (struct btrfs_inode_extref *)ref_ptr;
1224 parent_objectid = btrfs_inode_extref_parent(eb, r);
1225 } else {
1226 ref_struct_size = sizeof(struct btrfs_inode_ref);
1227 parent_objectid = key->offset;
1228 }
1229 inode_objectid = key->objectid;
Chris Masone02119d2008-09-05 16:13:11 -04001230
Chris Masone02119d2008-09-05 16:13:11 -04001231 /*
1232 * it is possible that we didn't log all the parent directories
1233 * for a given inode. If we don't find the dir, just don't
1234 * copy the back ref in. The link count fixup code will take
1235 * care of the rest
1236 */
Mark Fashehf1863732012-08-08 11:32:27 -07001237 dir = read_one_inode(root, parent_objectid);
Geyslan G. Bem03b2f082013-10-11 15:35:45 -03001238 if (!dir) {
1239 ret = -ENOENT;
1240 goto out;
1241 }
Chris Masone02119d2008-09-05 16:13:11 -04001242
Mark Fashehf1863732012-08-08 11:32:27 -07001243 inode = read_one_inode(root, inode_objectid);
Tsutomu Itohc00e9492011-04-28 09:10:23 +00001244 if (!inode) {
Geyslan G. Bem03b2f082013-10-11 15:35:45 -03001245 ret = -EIO;
1246 goto out;
Tsutomu Itohc00e9492011-04-28 09:10:23 +00001247 }
Chris Masone02119d2008-09-05 16:13:11 -04001248
Jan Schmidt5a1d7842012-08-17 14:04:41 -07001249 while (ref_ptr < ref_end) {
Mark Fashehf1863732012-08-08 11:32:27 -07001250 if (log_ref_ver) {
1251 ret = extref_get_fields(eb, ref_ptr, &namelen, &name,
1252 &ref_index, &parent_objectid);
1253 /*
1254 * parent object can change from one array
1255 * item to another.
1256 */
1257 if (!dir)
1258 dir = read_one_inode(root, parent_objectid);
Geyslan G. Bem03b2f082013-10-11 15:35:45 -03001259 if (!dir) {
1260 ret = -ENOENT;
1261 goto out;
1262 }
Mark Fashehf1863732012-08-08 11:32:27 -07001263 } else {
1264 ret = ref_get_fields(eb, ref_ptr, &namelen, &name,
1265 &ref_index);
1266 }
1267 if (ret)
Geyslan G. Bem03b2f082013-10-11 15:35:45 -03001268 goto out;
Chris Masone02119d2008-09-05 16:13:11 -04001269
Jan Schmidt5a1d7842012-08-17 14:04:41 -07001270 /* if we already have a perfect match, we're done */
1271 if (!inode_in_dir(root, path, btrfs_ino(dir), btrfs_ino(inode),
Mark Fashehf1863732012-08-08 11:32:27 -07001272 ref_index, name, namelen)) {
Jan Schmidt5a1d7842012-08-17 14:04:41 -07001273 /*
1274 * look for a conflicting back reference in the
1275 * metadata. if we find one we have to unlink that name
1276 * of the file before we add our new link. Later on, we
1277 * overwrite any existing back reference, and we don't
1278 * want to create dangling pointers in the directory.
1279 */
Chris Masone02119d2008-09-05 16:13:11 -04001280
Jan Schmidt5a1d7842012-08-17 14:04:41 -07001281 if (!search_done) {
1282 ret = __add_inode_ref(trans, root, path, log,
Mark Fashehf1863732012-08-08 11:32:27 -07001283 dir, inode, eb,
1284 inode_objectid,
1285 parent_objectid,
1286 ref_index, name, namelen,
Jan Schmidt5a1d7842012-08-17 14:04:41 -07001287 &search_done);
Geyslan G. Bem03b2f082013-10-11 15:35:45 -03001288 if (ret) {
1289 if (ret == 1)
1290 ret = 0;
Jan Schmidt5a1d7842012-08-17 14:04:41 -07001291 goto out;
Josef Bacik36508602013-04-25 16:23:32 -04001292 }
Chris Masone02119d2008-09-05 16:13:11 -04001293 }
Jan Schmidt5a1d7842012-08-17 14:04:41 -07001294
1295 /* insert our name */
1296 ret = btrfs_add_link(trans, dir, inode, name, namelen,
Mark Fashehf1863732012-08-08 11:32:27 -07001297 0, ref_index);
Josef Bacik36508602013-04-25 16:23:32 -04001298 if (ret)
1299 goto out;
Jan Schmidt5a1d7842012-08-17 14:04:41 -07001300
1301 btrfs_update_inode(trans, root, inode);
Chris Masone02119d2008-09-05 16:13:11 -04001302 }
liuboc622ae62011-03-26 08:01:12 -04001303
Mark Fashehf1863732012-08-08 11:32:27 -07001304 ref_ptr = (unsigned long)(ref_ptr + ref_struct_size) + namelen;
Jan Schmidt5a1d7842012-08-17 14:04:41 -07001305 kfree(name);
Geyslan G. Bem03b2f082013-10-11 15:35:45 -03001306 name = NULL;
Mark Fashehf1863732012-08-08 11:32:27 -07001307 if (log_ref_ver) {
1308 iput(dir);
1309 dir = NULL;
1310 }
Chris Masone02119d2008-09-05 16:13:11 -04001311 }
Chris Masone02119d2008-09-05 16:13:11 -04001312
1313 /* finally write the back reference in the inode */
1314 ret = overwrite_item(trans, root, path, eb, slot, key);
Jan Schmidt5a1d7842012-08-17 14:04:41 -07001315out:
David Sterbab3b4aa72011-04-21 01:20:15 +02001316 btrfs_release_path(path);
Geyslan G. Bem03b2f082013-10-11 15:35:45 -03001317 kfree(name);
Chris Masone02119d2008-09-05 16:13:11 -04001318 iput(dir);
1319 iput(inode);
Josef Bacik36508602013-04-25 16:23:32 -04001320 return ret;
Chris Masone02119d2008-09-05 16:13:11 -04001321}
1322
Yan, Zhengc71bf092009-11-12 09:34:40 +00001323static int insert_orphan_item(struct btrfs_trans_handle *trans,
David Sterba9c4f61f2015-01-02 19:12:57 +01001324 struct btrfs_root *root, u64 ino)
Yan, Zhengc71bf092009-11-12 09:34:40 +00001325{
1326 int ret;
David Sterba381cf652015-01-02 18:45:16 +01001327
David Sterba9c4f61f2015-01-02 19:12:57 +01001328 ret = btrfs_insert_orphan_item(trans, root, ino);
1329 if (ret == -EEXIST)
1330 ret = 0;
David Sterba381cf652015-01-02 18:45:16 +01001331
Yan, Zhengc71bf092009-11-12 09:34:40 +00001332 return ret;
1333}
1334
Mark Fashehf1863732012-08-08 11:32:27 -07001335static int count_inode_extrefs(struct btrfs_root *root,
1336 struct inode *inode, struct btrfs_path *path)
Chris Masone02119d2008-09-05 16:13:11 -04001337{
Mark Fashehf1863732012-08-08 11:32:27 -07001338 int ret = 0;
1339 int name_len;
1340 unsigned int nlink = 0;
1341 u32 item_size;
1342 u32 cur_offset = 0;
1343 u64 inode_objectid = btrfs_ino(inode);
1344 u64 offset = 0;
1345 unsigned long ptr;
1346 struct btrfs_inode_extref *extref;
1347 struct extent_buffer *leaf;
1348
1349 while (1) {
1350 ret = btrfs_find_one_extref(root, inode_objectid, offset, path,
1351 &extref, &offset);
1352 if (ret)
1353 break;
1354
1355 leaf = path->nodes[0];
1356 item_size = btrfs_item_size_nr(leaf, path->slots[0]);
1357 ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
Filipe Manana2c2c4522015-01-13 16:40:04 +00001358 cur_offset = 0;
Mark Fashehf1863732012-08-08 11:32:27 -07001359
1360 while (cur_offset < item_size) {
1361 extref = (struct btrfs_inode_extref *) (ptr + cur_offset);
1362 name_len = btrfs_inode_extref_name_len(leaf, extref);
1363
1364 nlink++;
1365
1366 cur_offset += name_len + sizeof(*extref);
1367 }
1368
1369 offset++;
1370 btrfs_release_path(path);
1371 }
1372 btrfs_release_path(path);
1373
Filipe Manana2c2c4522015-01-13 16:40:04 +00001374 if (ret < 0 && ret != -ENOENT)
Mark Fashehf1863732012-08-08 11:32:27 -07001375 return ret;
1376 return nlink;
1377}
1378
1379static int count_inode_refs(struct btrfs_root *root,
1380 struct inode *inode, struct btrfs_path *path)
1381{
Chris Masone02119d2008-09-05 16:13:11 -04001382 int ret;
1383 struct btrfs_key key;
Mark Fashehf1863732012-08-08 11:32:27 -07001384 unsigned int nlink = 0;
Chris Masone02119d2008-09-05 16:13:11 -04001385 unsigned long ptr;
1386 unsigned long ptr_end;
1387 int name_len;
Li Zefan33345d012011-04-20 10:31:50 +08001388 u64 ino = btrfs_ino(inode);
Chris Masone02119d2008-09-05 16:13:11 -04001389
Li Zefan33345d012011-04-20 10:31:50 +08001390 key.objectid = ino;
Chris Masone02119d2008-09-05 16:13:11 -04001391 key.type = BTRFS_INODE_REF_KEY;
1392 key.offset = (u64)-1;
1393
Chris Masond3977122009-01-05 21:25:51 -05001394 while (1) {
Chris Masone02119d2008-09-05 16:13:11 -04001395 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
1396 if (ret < 0)
1397 break;
1398 if (ret > 0) {
1399 if (path->slots[0] == 0)
1400 break;
1401 path->slots[0]--;
1402 }
Filipe David Borba Mananae93ae262013-10-14 22:49:11 +01001403process_slot:
Chris Masone02119d2008-09-05 16:13:11 -04001404 btrfs_item_key_to_cpu(path->nodes[0], &key,
1405 path->slots[0]);
Li Zefan33345d012011-04-20 10:31:50 +08001406 if (key.objectid != ino ||
Chris Masone02119d2008-09-05 16:13:11 -04001407 key.type != BTRFS_INODE_REF_KEY)
1408 break;
1409 ptr = btrfs_item_ptr_offset(path->nodes[0], path->slots[0]);
1410 ptr_end = ptr + btrfs_item_size_nr(path->nodes[0],
1411 path->slots[0]);
Chris Masond3977122009-01-05 21:25:51 -05001412 while (ptr < ptr_end) {
Chris Masone02119d2008-09-05 16:13:11 -04001413 struct btrfs_inode_ref *ref;
1414
1415 ref = (struct btrfs_inode_ref *)ptr;
1416 name_len = btrfs_inode_ref_name_len(path->nodes[0],
1417 ref);
1418 ptr = (unsigned long)(ref + 1) + name_len;
1419 nlink++;
1420 }
1421
1422 if (key.offset == 0)
1423 break;
Filipe David Borba Mananae93ae262013-10-14 22:49:11 +01001424 if (path->slots[0] > 0) {
1425 path->slots[0]--;
1426 goto process_slot;
1427 }
Chris Masone02119d2008-09-05 16:13:11 -04001428 key.offset--;
David Sterbab3b4aa72011-04-21 01:20:15 +02001429 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -04001430 }
David Sterbab3b4aa72011-04-21 01:20:15 +02001431 btrfs_release_path(path);
Mark Fashehf1863732012-08-08 11:32:27 -07001432
1433 return nlink;
1434}
1435
1436/*
1437 * There are a few corners where the link count of the file can't
1438 * be properly maintained during replay. So, instead of adding
1439 * lots of complexity to the log code, we just scan the backrefs
1440 * for any file that has been through replay.
1441 *
1442 * The scan will update the link count on the inode to reflect the
1443 * number of back refs found. If it goes down to zero, the iput
1444 * will free the inode.
1445 */
1446static noinline int fixup_inode_link_count(struct btrfs_trans_handle *trans,
1447 struct btrfs_root *root,
1448 struct inode *inode)
1449{
1450 struct btrfs_path *path;
1451 int ret;
1452 u64 nlink = 0;
1453 u64 ino = btrfs_ino(inode);
1454
1455 path = btrfs_alloc_path();
1456 if (!path)
1457 return -ENOMEM;
1458
1459 ret = count_inode_refs(root, inode, path);
1460 if (ret < 0)
1461 goto out;
1462
1463 nlink = ret;
1464
1465 ret = count_inode_extrefs(root, inode, path);
Mark Fashehf1863732012-08-08 11:32:27 -07001466 if (ret < 0)
1467 goto out;
1468
1469 nlink += ret;
1470
1471 ret = 0;
1472
Chris Masone02119d2008-09-05 16:13:11 -04001473 if (nlink != inode->i_nlink) {
Miklos Szeredibfe86842011-10-28 14:13:29 +02001474 set_nlink(inode, nlink);
Chris Masone02119d2008-09-05 16:13:11 -04001475 btrfs_update_inode(trans, root, inode);
1476 }
Chris Mason8d5bf1c2008-09-11 15:51:21 -04001477 BTRFS_I(inode)->index_cnt = (u64)-1;
Chris Masone02119d2008-09-05 16:13:11 -04001478
Yan, Zhengc71bf092009-11-12 09:34:40 +00001479 if (inode->i_nlink == 0) {
1480 if (S_ISDIR(inode->i_mode)) {
1481 ret = replay_dir_deletes(trans, root, NULL, path,
Li Zefan33345d012011-04-20 10:31:50 +08001482 ino, 1);
Josef Bacik36508602013-04-25 16:23:32 -04001483 if (ret)
1484 goto out;
Yan, Zhengc71bf092009-11-12 09:34:40 +00001485 }
Li Zefan33345d012011-04-20 10:31:50 +08001486 ret = insert_orphan_item(trans, root, ino);
Chris Mason12fcfd22009-03-24 10:24:20 -04001487 }
Chris Mason12fcfd22009-03-24 10:24:20 -04001488
Mark Fashehf1863732012-08-08 11:32:27 -07001489out:
1490 btrfs_free_path(path);
1491 return ret;
Chris Masone02119d2008-09-05 16:13:11 -04001492}
1493
1494static noinline int fixup_inode_link_counts(struct btrfs_trans_handle *trans,
1495 struct btrfs_root *root,
1496 struct btrfs_path *path)
1497{
1498 int ret;
1499 struct btrfs_key key;
1500 struct inode *inode;
1501
1502 key.objectid = BTRFS_TREE_LOG_FIXUP_OBJECTID;
1503 key.type = BTRFS_ORPHAN_ITEM_KEY;
1504 key.offset = (u64)-1;
Chris Masond3977122009-01-05 21:25:51 -05001505 while (1) {
Chris Masone02119d2008-09-05 16:13:11 -04001506 ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
1507 if (ret < 0)
1508 break;
1509
1510 if (ret == 1) {
1511 if (path->slots[0] == 0)
1512 break;
1513 path->slots[0]--;
1514 }
1515
1516 btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
1517 if (key.objectid != BTRFS_TREE_LOG_FIXUP_OBJECTID ||
1518 key.type != BTRFS_ORPHAN_ITEM_KEY)
1519 break;
1520
1521 ret = btrfs_del_item(trans, root, path);
Tsutomu Itoh65a246c2011-05-19 04:37:44 +00001522 if (ret)
1523 goto out;
Chris Masone02119d2008-09-05 16:13:11 -04001524
David Sterbab3b4aa72011-04-21 01:20:15 +02001525 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -04001526 inode = read_one_inode(root, key.offset);
Tsutomu Itohc00e9492011-04-28 09:10:23 +00001527 if (!inode)
1528 return -EIO;
Chris Masone02119d2008-09-05 16:13:11 -04001529
1530 ret = fixup_inode_link_count(trans, root, inode);
Chris Masone02119d2008-09-05 16:13:11 -04001531 iput(inode);
Josef Bacik36508602013-04-25 16:23:32 -04001532 if (ret)
1533 goto out;
Chris Masone02119d2008-09-05 16:13:11 -04001534
Chris Mason12fcfd22009-03-24 10:24:20 -04001535 /*
1536 * fixup on a directory may create new entries,
1537 * make sure we always look for the highset possible
1538 * offset
1539 */
1540 key.offset = (u64)-1;
Chris Masone02119d2008-09-05 16:13:11 -04001541 }
Tsutomu Itoh65a246c2011-05-19 04:37:44 +00001542 ret = 0;
1543out:
David Sterbab3b4aa72011-04-21 01:20:15 +02001544 btrfs_release_path(path);
Tsutomu Itoh65a246c2011-05-19 04:37:44 +00001545 return ret;
Chris Masone02119d2008-09-05 16:13:11 -04001546}
1547
1548
1549/*
1550 * record a given inode in the fixup dir so we can check its link
1551 * count when replay is done. The link count is incremented here
1552 * so the inode won't go away until we check it
1553 */
1554static noinline int link_to_fixup_dir(struct btrfs_trans_handle *trans,
1555 struct btrfs_root *root,
1556 struct btrfs_path *path,
1557 u64 objectid)
1558{
1559 struct btrfs_key key;
1560 int ret = 0;
1561 struct inode *inode;
1562
1563 inode = read_one_inode(root, objectid);
Tsutomu Itohc00e9492011-04-28 09:10:23 +00001564 if (!inode)
1565 return -EIO;
Chris Masone02119d2008-09-05 16:13:11 -04001566
1567 key.objectid = BTRFS_TREE_LOG_FIXUP_OBJECTID;
David Sterba962a2982014-06-04 18:41:45 +02001568 key.type = BTRFS_ORPHAN_ITEM_KEY;
Chris Masone02119d2008-09-05 16:13:11 -04001569 key.offset = objectid;
1570
1571 ret = btrfs_insert_empty_item(trans, root, path, &key, 0);
1572
David Sterbab3b4aa72011-04-21 01:20:15 +02001573 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -04001574 if (ret == 0) {
Josef Bacik9bf7a482013-03-01 13:35:47 -05001575 if (!inode->i_nlink)
1576 set_nlink(inode, 1);
1577 else
Zach Brown8b558c52013-10-16 12:10:34 -07001578 inc_nlink(inode);
Tsutomu Itohb9959292012-06-25 21:25:22 -06001579 ret = btrfs_update_inode(trans, root, inode);
Chris Masone02119d2008-09-05 16:13:11 -04001580 } else if (ret == -EEXIST) {
1581 ret = 0;
1582 } else {
Josef Bacik36508602013-04-25 16:23:32 -04001583 BUG(); /* Logic Error */
Chris Masone02119d2008-09-05 16:13:11 -04001584 }
1585 iput(inode);
1586
1587 return ret;
1588}
1589
1590/*
1591 * when replaying the log for a directory, we only insert names
1592 * for inodes that actually exist. This means an fsync on a directory
1593 * does not implicitly fsync all the new files in it
1594 */
1595static noinline int insert_one_name(struct btrfs_trans_handle *trans,
1596 struct btrfs_root *root,
Chris Masone02119d2008-09-05 16:13:11 -04001597 u64 dirid, u64 index,
Zhaolei60d53eb2015-08-17 18:44:46 +08001598 char *name, int name_len,
Chris Masone02119d2008-09-05 16:13:11 -04001599 struct btrfs_key *location)
1600{
1601 struct inode *inode;
1602 struct inode *dir;
1603 int ret;
1604
1605 inode = read_one_inode(root, location->objectid);
1606 if (!inode)
1607 return -ENOENT;
1608
1609 dir = read_one_inode(root, dirid);
1610 if (!dir) {
1611 iput(inode);
1612 return -EIO;
1613 }
Josef Bacikd5554382013-09-11 14:17:00 -04001614
Chris Masone02119d2008-09-05 16:13:11 -04001615 ret = btrfs_add_link(trans, dir, inode, name, name_len, 1, index);
1616
1617 /* FIXME, put inode into FIXUP list */
1618
1619 iput(inode);
1620 iput(dir);
1621 return ret;
1622}
1623
1624/*
Filipe Mananadf8d1162015-01-14 01:52:25 +00001625 * Return true if an inode reference exists in the log for the given name,
1626 * inode and parent inode.
1627 */
1628static bool name_in_log_ref(struct btrfs_root *log_root,
1629 const char *name, const int name_len,
1630 const u64 dirid, const u64 ino)
1631{
1632 struct btrfs_key search_key;
1633
1634 search_key.objectid = ino;
1635 search_key.type = BTRFS_INODE_REF_KEY;
1636 search_key.offset = dirid;
1637 if (backref_in_log(log_root, &search_key, dirid, name, name_len))
1638 return true;
1639
1640 search_key.type = BTRFS_INODE_EXTREF_KEY;
1641 search_key.offset = btrfs_extref_hash(dirid, name, name_len);
1642 if (backref_in_log(log_root, &search_key, dirid, name, name_len))
1643 return true;
1644
1645 return false;
1646}
1647
1648/*
Chris Masone02119d2008-09-05 16:13:11 -04001649 * take a single entry in a log directory item and replay it into
1650 * the subvolume.
1651 *
1652 * if a conflicting item exists in the subdirectory already,
1653 * the inode it points to is unlinked and put into the link count
1654 * fix up tree.
1655 *
1656 * If a name from the log points to a file or directory that does
1657 * not exist in the FS, it is skipped. fsyncs on directories
1658 * do not force down inodes inside that directory, just changes to the
1659 * names or unlinks in a directory.
Filipe Mananabb53eda2015-07-15 23:26:43 +01001660 *
1661 * Returns < 0 on error, 0 if the name wasn't replayed (dentry points to a
1662 * non-existing inode) and 1 if the name was replayed.
Chris Masone02119d2008-09-05 16:13:11 -04001663 */
1664static noinline int replay_one_name(struct btrfs_trans_handle *trans,
1665 struct btrfs_root *root,
1666 struct btrfs_path *path,
1667 struct extent_buffer *eb,
1668 struct btrfs_dir_item *di,
1669 struct btrfs_key *key)
1670{
1671 char *name;
1672 int name_len;
1673 struct btrfs_dir_item *dst_di;
1674 struct btrfs_key found_key;
1675 struct btrfs_key log_key;
1676 struct inode *dir;
Chris Masone02119d2008-09-05 16:13:11 -04001677 u8 log_type;
Chris Mason4bef0842008-09-08 11:18:08 -04001678 int exists;
Josef Bacik36508602013-04-25 16:23:32 -04001679 int ret = 0;
Josef Bacikd5554382013-09-11 14:17:00 -04001680 bool update_size = (key->type == BTRFS_DIR_INDEX_KEY);
Filipe Mananabb53eda2015-07-15 23:26:43 +01001681 bool name_added = false;
Chris Masone02119d2008-09-05 16:13:11 -04001682
1683 dir = read_one_inode(root, key->objectid);
Tsutomu Itohc00e9492011-04-28 09:10:23 +00001684 if (!dir)
1685 return -EIO;
Chris Masone02119d2008-09-05 16:13:11 -04001686
1687 name_len = btrfs_dir_name_len(eb, di);
1688 name = kmalloc(name_len, GFP_NOFS);
Filipe David Borba Manana2bac3252013-08-04 19:58:57 +01001689 if (!name) {
1690 ret = -ENOMEM;
1691 goto out;
1692 }
liubo2a29edc2011-01-26 06:22:08 +00001693
Chris Masone02119d2008-09-05 16:13:11 -04001694 log_type = btrfs_dir_type(eb, di);
1695 read_extent_buffer(eb, name, (unsigned long)(di + 1),
1696 name_len);
1697
1698 btrfs_dir_item_key_to_cpu(eb, di, &log_key);
Chris Mason4bef0842008-09-08 11:18:08 -04001699 exists = btrfs_lookup_inode(trans, root, path, &log_key, 0);
1700 if (exists == 0)
1701 exists = 1;
1702 else
1703 exists = 0;
David Sterbab3b4aa72011-04-21 01:20:15 +02001704 btrfs_release_path(path);
Chris Mason4bef0842008-09-08 11:18:08 -04001705
Chris Masone02119d2008-09-05 16:13:11 -04001706 if (key->type == BTRFS_DIR_ITEM_KEY) {
1707 dst_di = btrfs_lookup_dir_item(trans, root, path, key->objectid,
1708 name, name_len, 1);
Chris Masond3977122009-01-05 21:25:51 -05001709 } else if (key->type == BTRFS_DIR_INDEX_KEY) {
Chris Masone02119d2008-09-05 16:13:11 -04001710 dst_di = btrfs_lookup_dir_index_item(trans, root, path,
1711 key->objectid,
1712 key->offset, name,
1713 name_len, 1);
1714 } else {
Josef Bacik36508602013-04-25 16:23:32 -04001715 /* Corruption */
1716 ret = -EINVAL;
1717 goto out;
Chris Masone02119d2008-09-05 16:13:11 -04001718 }
David Sterbac7040052011-04-19 18:00:01 +02001719 if (IS_ERR_OR_NULL(dst_di)) {
Chris Masone02119d2008-09-05 16:13:11 -04001720 /* we need a sequence number to insert, so we only
1721 * do inserts for the BTRFS_DIR_INDEX_KEY types
1722 */
1723 if (key->type != BTRFS_DIR_INDEX_KEY)
1724 goto out;
1725 goto insert;
1726 }
1727
1728 btrfs_dir_item_key_to_cpu(path->nodes[0], dst_di, &found_key);
1729 /* the existing item matches the logged item */
1730 if (found_key.objectid == log_key.objectid &&
1731 found_key.type == log_key.type &&
1732 found_key.offset == log_key.offset &&
1733 btrfs_dir_type(path->nodes[0], dst_di) == log_type) {
Filipe Mananaa2cc11d2014-09-08 22:53:18 +01001734 update_size = false;
Chris Masone02119d2008-09-05 16:13:11 -04001735 goto out;
1736 }
1737
1738 /*
1739 * don't drop the conflicting directory entry if the inode
1740 * for the new entry doesn't exist
1741 */
Chris Mason4bef0842008-09-08 11:18:08 -04001742 if (!exists)
Chris Masone02119d2008-09-05 16:13:11 -04001743 goto out;
1744
Chris Masone02119d2008-09-05 16:13:11 -04001745 ret = drop_one_dir_item(trans, root, path, dir, dst_di);
Josef Bacik36508602013-04-25 16:23:32 -04001746 if (ret)
1747 goto out;
Chris Masone02119d2008-09-05 16:13:11 -04001748
1749 if (key->type == BTRFS_DIR_INDEX_KEY)
1750 goto insert;
1751out:
David Sterbab3b4aa72011-04-21 01:20:15 +02001752 btrfs_release_path(path);
Josef Bacikd5554382013-09-11 14:17:00 -04001753 if (!ret && update_size) {
1754 btrfs_i_size_write(dir, dir->i_size + name_len * 2);
1755 ret = btrfs_update_inode(trans, root, dir);
1756 }
Chris Masone02119d2008-09-05 16:13:11 -04001757 kfree(name);
1758 iput(dir);
Filipe Mananabb53eda2015-07-15 23:26:43 +01001759 if (!ret && name_added)
1760 ret = 1;
Josef Bacik36508602013-04-25 16:23:32 -04001761 return ret;
Chris Masone02119d2008-09-05 16:13:11 -04001762
1763insert:
Filipe Mananadf8d1162015-01-14 01:52:25 +00001764 if (name_in_log_ref(root->log_root, name, name_len,
1765 key->objectid, log_key.objectid)) {
1766 /* The dentry will be added later. */
1767 ret = 0;
1768 update_size = false;
1769 goto out;
1770 }
David Sterbab3b4aa72011-04-21 01:20:15 +02001771 btrfs_release_path(path);
Zhaolei60d53eb2015-08-17 18:44:46 +08001772 ret = insert_one_name(trans, root, key->objectid, key->offset,
1773 name, name_len, &log_key);
Filipe Mananadf8d1162015-01-14 01:52:25 +00001774 if (ret && ret != -ENOENT && ret != -EEXIST)
Josef Bacik36508602013-04-25 16:23:32 -04001775 goto out;
Filipe Mananabb53eda2015-07-15 23:26:43 +01001776 if (!ret)
1777 name_added = true;
Josef Bacikd5554382013-09-11 14:17:00 -04001778 update_size = false;
Josef Bacik36508602013-04-25 16:23:32 -04001779 ret = 0;
Chris Masone02119d2008-09-05 16:13:11 -04001780 goto out;
1781}
1782
1783/*
1784 * find all the names in a directory item and reconcile them into
1785 * the subvolume. Only BTRFS_DIR_ITEM_KEY types will have more than
1786 * one name in a directory item, but the same code gets used for
1787 * both directory index types
1788 */
1789static noinline int replay_one_dir_item(struct btrfs_trans_handle *trans,
1790 struct btrfs_root *root,
1791 struct btrfs_path *path,
1792 struct extent_buffer *eb, int slot,
1793 struct btrfs_key *key)
1794{
Filipe Mananabb53eda2015-07-15 23:26:43 +01001795 int ret = 0;
Chris Masone02119d2008-09-05 16:13:11 -04001796 u32 item_size = btrfs_item_size_nr(eb, slot);
1797 struct btrfs_dir_item *di;
1798 int name_len;
1799 unsigned long ptr;
1800 unsigned long ptr_end;
Filipe Mananabb53eda2015-07-15 23:26:43 +01001801 struct btrfs_path *fixup_path = NULL;
Chris Masone02119d2008-09-05 16:13:11 -04001802
1803 ptr = btrfs_item_ptr_offset(eb, slot);
1804 ptr_end = ptr + item_size;
Chris Masond3977122009-01-05 21:25:51 -05001805 while (ptr < ptr_end) {
Chris Masone02119d2008-09-05 16:13:11 -04001806 di = (struct btrfs_dir_item *)ptr;
Josef Bacik22a94d42011-03-16 16:47:17 -04001807 if (verify_dir_item(root, eb, di))
1808 return -EIO;
Chris Masone02119d2008-09-05 16:13:11 -04001809 name_len = btrfs_dir_name_len(eb, di);
1810 ret = replay_one_name(trans, root, path, eb, di, key);
Filipe Mananabb53eda2015-07-15 23:26:43 +01001811 if (ret < 0)
1812 break;
Chris Masone02119d2008-09-05 16:13:11 -04001813 ptr = (unsigned long)(di + 1);
1814 ptr += name_len;
Filipe Mananabb53eda2015-07-15 23:26:43 +01001815
1816 /*
1817 * If this entry refers to a non-directory (directories can not
1818 * have a link count > 1) and it was added in the transaction
1819 * that was not committed, make sure we fixup the link count of
1820 * the inode it the entry points to. Otherwise something like
1821 * the following would result in a directory pointing to an
1822 * inode with a wrong link that does not account for this dir
1823 * entry:
1824 *
1825 * mkdir testdir
1826 * touch testdir/foo
1827 * touch testdir/bar
1828 * sync
1829 *
1830 * ln testdir/bar testdir/bar_link
1831 * ln testdir/foo testdir/foo_link
1832 * xfs_io -c "fsync" testdir/bar
1833 *
1834 * <power failure>
1835 *
1836 * mount fs, log replay happens
1837 *
1838 * File foo would remain with a link count of 1 when it has two
1839 * entries pointing to it in the directory testdir. This would
1840 * make it impossible to ever delete the parent directory has
1841 * it would result in stale dentries that can never be deleted.
1842 */
1843 if (ret == 1 && btrfs_dir_type(eb, di) != BTRFS_FT_DIR) {
1844 struct btrfs_key di_key;
1845
1846 if (!fixup_path) {
1847 fixup_path = btrfs_alloc_path();
1848 if (!fixup_path) {
1849 ret = -ENOMEM;
1850 break;
1851 }
1852 }
1853
1854 btrfs_dir_item_key_to_cpu(eb, di, &di_key);
1855 ret = link_to_fixup_dir(trans, root, fixup_path,
1856 di_key.objectid);
1857 if (ret)
1858 break;
1859 }
1860 ret = 0;
Chris Masone02119d2008-09-05 16:13:11 -04001861 }
Filipe Mananabb53eda2015-07-15 23:26:43 +01001862 btrfs_free_path(fixup_path);
1863 return ret;
Chris Masone02119d2008-09-05 16:13:11 -04001864}
1865
1866/*
1867 * directory replay has two parts. There are the standard directory
1868 * items in the log copied from the subvolume, and range items
1869 * created in the log while the subvolume was logged.
1870 *
1871 * The range items tell us which parts of the key space the log
1872 * is authoritative for. During replay, if a key in the subvolume
1873 * directory is in a logged range item, but not actually in the log
1874 * that means it was deleted from the directory before the fsync
1875 * and should be removed.
1876 */
1877static noinline int find_dir_range(struct btrfs_root *root,
1878 struct btrfs_path *path,
1879 u64 dirid, int key_type,
1880 u64 *start_ret, u64 *end_ret)
1881{
1882 struct btrfs_key key;
1883 u64 found_end;
1884 struct btrfs_dir_log_item *item;
1885 int ret;
1886 int nritems;
1887
1888 if (*start_ret == (u64)-1)
1889 return 1;
1890
1891 key.objectid = dirid;
1892 key.type = key_type;
1893 key.offset = *start_ret;
1894
1895 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
1896 if (ret < 0)
1897 goto out;
1898 if (ret > 0) {
1899 if (path->slots[0] == 0)
1900 goto out;
1901 path->slots[0]--;
1902 }
1903 if (ret != 0)
1904 btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
1905
1906 if (key.type != key_type || key.objectid != dirid) {
1907 ret = 1;
1908 goto next;
1909 }
1910 item = btrfs_item_ptr(path->nodes[0], path->slots[0],
1911 struct btrfs_dir_log_item);
1912 found_end = btrfs_dir_log_end(path->nodes[0], item);
1913
1914 if (*start_ret >= key.offset && *start_ret <= found_end) {
1915 ret = 0;
1916 *start_ret = key.offset;
1917 *end_ret = found_end;
1918 goto out;
1919 }
1920 ret = 1;
1921next:
1922 /* check the next slot in the tree to see if it is a valid item */
1923 nritems = btrfs_header_nritems(path->nodes[0]);
1924 if (path->slots[0] >= nritems) {
1925 ret = btrfs_next_leaf(root, path);
1926 if (ret)
1927 goto out;
1928 } else {
1929 path->slots[0]++;
1930 }
1931
1932 btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
1933
1934 if (key.type != key_type || key.objectid != dirid) {
1935 ret = 1;
1936 goto out;
1937 }
1938 item = btrfs_item_ptr(path->nodes[0], path->slots[0],
1939 struct btrfs_dir_log_item);
1940 found_end = btrfs_dir_log_end(path->nodes[0], item);
1941 *start_ret = key.offset;
1942 *end_ret = found_end;
1943 ret = 0;
1944out:
David Sterbab3b4aa72011-04-21 01:20:15 +02001945 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -04001946 return ret;
1947}
1948
1949/*
1950 * this looks for a given directory item in the log. If the directory
1951 * item is not in the log, the item is removed and the inode it points
1952 * to is unlinked
1953 */
1954static noinline int check_item_in_log(struct btrfs_trans_handle *trans,
1955 struct btrfs_root *root,
1956 struct btrfs_root *log,
1957 struct btrfs_path *path,
1958 struct btrfs_path *log_path,
1959 struct inode *dir,
1960 struct btrfs_key *dir_key)
1961{
1962 int ret;
1963 struct extent_buffer *eb;
1964 int slot;
1965 u32 item_size;
1966 struct btrfs_dir_item *di;
1967 struct btrfs_dir_item *log_di;
1968 int name_len;
1969 unsigned long ptr;
1970 unsigned long ptr_end;
1971 char *name;
1972 struct inode *inode;
1973 struct btrfs_key location;
1974
1975again:
1976 eb = path->nodes[0];
1977 slot = path->slots[0];
1978 item_size = btrfs_item_size_nr(eb, slot);
1979 ptr = btrfs_item_ptr_offset(eb, slot);
1980 ptr_end = ptr + item_size;
Chris Masond3977122009-01-05 21:25:51 -05001981 while (ptr < ptr_end) {
Chris Masone02119d2008-09-05 16:13:11 -04001982 di = (struct btrfs_dir_item *)ptr;
Josef Bacik22a94d42011-03-16 16:47:17 -04001983 if (verify_dir_item(root, eb, di)) {
1984 ret = -EIO;
1985 goto out;
1986 }
1987
Chris Masone02119d2008-09-05 16:13:11 -04001988 name_len = btrfs_dir_name_len(eb, di);
1989 name = kmalloc(name_len, GFP_NOFS);
1990 if (!name) {
1991 ret = -ENOMEM;
1992 goto out;
1993 }
1994 read_extent_buffer(eb, name, (unsigned long)(di + 1),
1995 name_len);
1996 log_di = NULL;
Chris Mason12fcfd22009-03-24 10:24:20 -04001997 if (log && dir_key->type == BTRFS_DIR_ITEM_KEY) {
Chris Masone02119d2008-09-05 16:13:11 -04001998 log_di = btrfs_lookup_dir_item(trans, log, log_path,
1999 dir_key->objectid,
2000 name, name_len, 0);
Chris Mason12fcfd22009-03-24 10:24:20 -04002001 } else if (log && dir_key->type == BTRFS_DIR_INDEX_KEY) {
Chris Masone02119d2008-09-05 16:13:11 -04002002 log_di = btrfs_lookup_dir_index_item(trans, log,
2003 log_path,
2004 dir_key->objectid,
2005 dir_key->offset,
2006 name, name_len, 0);
2007 }
Filipe David Borba Manana269d0402013-10-28 17:39:21 +00002008 if (!log_di || (IS_ERR(log_di) && PTR_ERR(log_di) == -ENOENT)) {
Chris Masone02119d2008-09-05 16:13:11 -04002009 btrfs_dir_item_key_to_cpu(eb, di, &location);
David Sterbab3b4aa72011-04-21 01:20:15 +02002010 btrfs_release_path(path);
2011 btrfs_release_path(log_path);
Chris Masone02119d2008-09-05 16:13:11 -04002012 inode = read_one_inode(root, location.objectid);
Tsutomu Itohc00e9492011-04-28 09:10:23 +00002013 if (!inode) {
2014 kfree(name);
2015 return -EIO;
2016 }
Chris Masone02119d2008-09-05 16:13:11 -04002017
2018 ret = link_to_fixup_dir(trans, root,
2019 path, location.objectid);
Josef Bacik36508602013-04-25 16:23:32 -04002020 if (ret) {
2021 kfree(name);
2022 iput(inode);
2023 goto out;
2024 }
2025
Zach Brown8b558c52013-10-16 12:10:34 -07002026 inc_nlink(inode);
Chris Masone02119d2008-09-05 16:13:11 -04002027 ret = btrfs_unlink_inode(trans, root, dir, inode,
2028 name, name_len);
Josef Bacik36508602013-04-25 16:23:32 -04002029 if (!ret)
Filipe David Borba Mananaada9af22013-08-05 09:25:47 +01002030 ret = btrfs_run_delayed_items(trans, root);
Chris Masone02119d2008-09-05 16:13:11 -04002031 kfree(name);
2032 iput(inode);
Josef Bacik36508602013-04-25 16:23:32 -04002033 if (ret)
2034 goto out;
Chris Masone02119d2008-09-05 16:13:11 -04002035
2036 /* there might still be more names under this key
2037 * check and repeat if required
2038 */
2039 ret = btrfs_search_slot(NULL, root, dir_key, path,
2040 0, 0);
2041 if (ret == 0)
2042 goto again;
2043 ret = 0;
2044 goto out;
Filipe David Borba Manana269d0402013-10-28 17:39:21 +00002045 } else if (IS_ERR(log_di)) {
2046 kfree(name);
2047 return PTR_ERR(log_di);
Chris Masone02119d2008-09-05 16:13:11 -04002048 }
David Sterbab3b4aa72011-04-21 01:20:15 +02002049 btrfs_release_path(log_path);
Chris Masone02119d2008-09-05 16:13:11 -04002050 kfree(name);
2051
2052 ptr = (unsigned long)(di + 1);
2053 ptr += name_len;
2054 }
2055 ret = 0;
2056out:
David Sterbab3b4aa72011-04-21 01:20:15 +02002057 btrfs_release_path(path);
2058 btrfs_release_path(log_path);
Chris Masone02119d2008-09-05 16:13:11 -04002059 return ret;
2060}
2061
Filipe Manana4f764e52015-02-23 19:53:35 +00002062static int replay_xattr_deletes(struct btrfs_trans_handle *trans,
2063 struct btrfs_root *root,
2064 struct btrfs_root *log,
2065 struct btrfs_path *path,
2066 const u64 ino)
2067{
2068 struct btrfs_key search_key;
2069 struct btrfs_path *log_path;
2070 int i;
2071 int nritems;
2072 int ret;
2073
2074 log_path = btrfs_alloc_path();
2075 if (!log_path)
2076 return -ENOMEM;
2077
2078 search_key.objectid = ino;
2079 search_key.type = BTRFS_XATTR_ITEM_KEY;
2080 search_key.offset = 0;
2081again:
2082 ret = btrfs_search_slot(NULL, root, &search_key, path, 0, 0);
2083 if (ret < 0)
2084 goto out;
2085process_leaf:
2086 nritems = btrfs_header_nritems(path->nodes[0]);
2087 for (i = path->slots[0]; i < nritems; i++) {
2088 struct btrfs_key key;
2089 struct btrfs_dir_item *di;
2090 struct btrfs_dir_item *log_di;
2091 u32 total_size;
2092 u32 cur;
2093
2094 btrfs_item_key_to_cpu(path->nodes[0], &key, i);
2095 if (key.objectid != ino || key.type != BTRFS_XATTR_ITEM_KEY) {
2096 ret = 0;
2097 goto out;
2098 }
2099
2100 di = btrfs_item_ptr(path->nodes[0], i, struct btrfs_dir_item);
2101 total_size = btrfs_item_size_nr(path->nodes[0], i);
2102 cur = 0;
2103 while (cur < total_size) {
2104 u16 name_len = btrfs_dir_name_len(path->nodes[0], di);
2105 u16 data_len = btrfs_dir_data_len(path->nodes[0], di);
2106 u32 this_len = sizeof(*di) + name_len + data_len;
2107 char *name;
2108
2109 name = kmalloc(name_len, GFP_NOFS);
2110 if (!name) {
2111 ret = -ENOMEM;
2112 goto out;
2113 }
2114 read_extent_buffer(path->nodes[0], name,
2115 (unsigned long)(di + 1), name_len);
2116
2117 log_di = btrfs_lookup_xattr(NULL, log, log_path, ino,
2118 name, name_len, 0);
2119 btrfs_release_path(log_path);
2120 if (!log_di) {
2121 /* Doesn't exist in log tree, so delete it. */
2122 btrfs_release_path(path);
2123 di = btrfs_lookup_xattr(trans, root, path, ino,
2124 name, name_len, -1);
2125 kfree(name);
2126 if (IS_ERR(di)) {
2127 ret = PTR_ERR(di);
2128 goto out;
2129 }
2130 ASSERT(di);
2131 ret = btrfs_delete_one_dir_name(trans, root,
2132 path, di);
2133 if (ret)
2134 goto out;
2135 btrfs_release_path(path);
2136 search_key = key;
2137 goto again;
2138 }
2139 kfree(name);
2140 if (IS_ERR(log_di)) {
2141 ret = PTR_ERR(log_di);
2142 goto out;
2143 }
2144 cur += this_len;
2145 di = (struct btrfs_dir_item *)((char *)di + this_len);
2146 }
2147 }
2148 ret = btrfs_next_leaf(root, path);
2149 if (ret > 0)
2150 ret = 0;
2151 else if (ret == 0)
2152 goto process_leaf;
2153out:
2154 btrfs_free_path(log_path);
2155 btrfs_release_path(path);
2156 return ret;
2157}
2158
2159
Chris Masone02119d2008-09-05 16:13:11 -04002160/*
2161 * deletion replay happens before we copy any new directory items
2162 * out of the log or out of backreferences from inodes. It
2163 * scans the log to find ranges of keys that log is authoritative for,
2164 * and then scans the directory to find items in those ranges that are
2165 * not present in the log.
2166 *
2167 * Anything we don't find in the log is unlinked and removed from the
2168 * directory.
2169 */
2170static noinline int replay_dir_deletes(struct btrfs_trans_handle *trans,
2171 struct btrfs_root *root,
2172 struct btrfs_root *log,
2173 struct btrfs_path *path,
Chris Mason12fcfd22009-03-24 10:24:20 -04002174 u64 dirid, int del_all)
Chris Masone02119d2008-09-05 16:13:11 -04002175{
2176 u64 range_start;
2177 u64 range_end;
2178 int key_type = BTRFS_DIR_LOG_ITEM_KEY;
2179 int ret = 0;
2180 struct btrfs_key dir_key;
2181 struct btrfs_key found_key;
2182 struct btrfs_path *log_path;
2183 struct inode *dir;
2184
2185 dir_key.objectid = dirid;
2186 dir_key.type = BTRFS_DIR_ITEM_KEY;
2187 log_path = btrfs_alloc_path();
2188 if (!log_path)
2189 return -ENOMEM;
2190
2191 dir = read_one_inode(root, dirid);
2192 /* it isn't an error if the inode isn't there, that can happen
2193 * because we replay the deletes before we copy in the inode item
2194 * from the log
2195 */
2196 if (!dir) {
2197 btrfs_free_path(log_path);
2198 return 0;
2199 }
2200again:
2201 range_start = 0;
2202 range_end = 0;
Chris Masond3977122009-01-05 21:25:51 -05002203 while (1) {
Chris Mason12fcfd22009-03-24 10:24:20 -04002204 if (del_all)
2205 range_end = (u64)-1;
2206 else {
2207 ret = find_dir_range(log, path, dirid, key_type,
2208 &range_start, &range_end);
2209 if (ret != 0)
2210 break;
2211 }
Chris Masone02119d2008-09-05 16:13:11 -04002212
2213 dir_key.offset = range_start;
Chris Masond3977122009-01-05 21:25:51 -05002214 while (1) {
Chris Masone02119d2008-09-05 16:13:11 -04002215 int nritems;
2216 ret = btrfs_search_slot(NULL, root, &dir_key, path,
2217 0, 0);
2218 if (ret < 0)
2219 goto out;
2220
2221 nritems = btrfs_header_nritems(path->nodes[0]);
2222 if (path->slots[0] >= nritems) {
2223 ret = btrfs_next_leaf(root, path);
2224 if (ret)
2225 break;
2226 }
2227 btrfs_item_key_to_cpu(path->nodes[0], &found_key,
2228 path->slots[0]);
2229 if (found_key.objectid != dirid ||
2230 found_key.type != dir_key.type)
2231 goto next_type;
2232
2233 if (found_key.offset > range_end)
2234 break;
2235
2236 ret = check_item_in_log(trans, root, log, path,
Chris Mason12fcfd22009-03-24 10:24:20 -04002237 log_path, dir,
2238 &found_key);
Josef Bacik36508602013-04-25 16:23:32 -04002239 if (ret)
2240 goto out;
Chris Masone02119d2008-09-05 16:13:11 -04002241 if (found_key.offset == (u64)-1)
2242 break;
2243 dir_key.offset = found_key.offset + 1;
2244 }
David Sterbab3b4aa72011-04-21 01:20:15 +02002245 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -04002246 if (range_end == (u64)-1)
2247 break;
2248 range_start = range_end + 1;
2249 }
2250
2251next_type:
2252 ret = 0;
2253 if (key_type == BTRFS_DIR_LOG_ITEM_KEY) {
2254 key_type = BTRFS_DIR_LOG_INDEX_KEY;
2255 dir_key.type = BTRFS_DIR_INDEX_KEY;
David Sterbab3b4aa72011-04-21 01:20:15 +02002256 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -04002257 goto again;
2258 }
2259out:
David Sterbab3b4aa72011-04-21 01:20:15 +02002260 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -04002261 btrfs_free_path(log_path);
2262 iput(dir);
2263 return ret;
2264}
2265
2266/*
2267 * the process_func used to replay items from the log tree. This
2268 * gets called in two different stages. The first stage just looks
2269 * for inodes and makes sure they are all copied into the subvolume.
2270 *
2271 * The second stage copies all the other item types from the log into
2272 * the subvolume. The two stage approach is slower, but gets rid of
2273 * lots of complexity around inodes referencing other inodes that exist
2274 * only in the log (references come from either directory items or inode
2275 * back refs).
2276 */
2277static int replay_one_buffer(struct btrfs_root *log, struct extent_buffer *eb,
2278 struct walk_control *wc, u64 gen)
2279{
2280 int nritems;
2281 struct btrfs_path *path;
2282 struct btrfs_root *root = wc->replay_dest;
2283 struct btrfs_key key;
Chris Masone02119d2008-09-05 16:13:11 -04002284 int level;
2285 int i;
2286 int ret;
2287
Tsutomu Itoh018642a2012-05-29 18:10:13 +09002288 ret = btrfs_read_buffer(eb, gen);
2289 if (ret)
2290 return ret;
Chris Masone02119d2008-09-05 16:13:11 -04002291
2292 level = btrfs_header_level(eb);
2293
2294 if (level != 0)
2295 return 0;
2296
2297 path = btrfs_alloc_path();
Mark Fasheh1e5063d2011-07-12 10:46:06 -07002298 if (!path)
2299 return -ENOMEM;
Chris Masone02119d2008-09-05 16:13:11 -04002300
2301 nritems = btrfs_header_nritems(eb);
2302 for (i = 0; i < nritems; i++) {
2303 btrfs_item_key_to_cpu(eb, &key, i);
Chris Masone02119d2008-09-05 16:13:11 -04002304
2305 /* inode keys are done during the first stage */
2306 if (key.type == BTRFS_INODE_ITEM_KEY &&
2307 wc->stage == LOG_WALK_REPLAY_INODES) {
Chris Masone02119d2008-09-05 16:13:11 -04002308 struct btrfs_inode_item *inode_item;
2309 u32 mode;
2310
2311 inode_item = btrfs_item_ptr(eb, i,
2312 struct btrfs_inode_item);
Filipe Manana4f764e52015-02-23 19:53:35 +00002313 ret = replay_xattr_deletes(wc->trans, root, log,
2314 path, key.objectid);
2315 if (ret)
2316 break;
Chris Masone02119d2008-09-05 16:13:11 -04002317 mode = btrfs_inode_mode(eb, inode_item);
2318 if (S_ISDIR(mode)) {
2319 ret = replay_dir_deletes(wc->trans,
Chris Mason12fcfd22009-03-24 10:24:20 -04002320 root, log, path, key.objectid, 0);
Josef Bacikb50c6e22013-04-25 15:55:30 -04002321 if (ret)
2322 break;
Chris Masone02119d2008-09-05 16:13:11 -04002323 }
2324 ret = overwrite_item(wc->trans, root, path,
2325 eb, i, &key);
Josef Bacikb50c6e22013-04-25 15:55:30 -04002326 if (ret)
2327 break;
Chris Masone02119d2008-09-05 16:13:11 -04002328
Yan, Zhengc71bf092009-11-12 09:34:40 +00002329 /* for regular files, make sure corresponding
2330 * orhpan item exist. extents past the new EOF
2331 * will be truncated later by orphan cleanup.
Chris Masone02119d2008-09-05 16:13:11 -04002332 */
2333 if (S_ISREG(mode)) {
Yan, Zhengc71bf092009-11-12 09:34:40 +00002334 ret = insert_orphan_item(wc->trans, root,
2335 key.objectid);
Josef Bacikb50c6e22013-04-25 15:55:30 -04002336 if (ret)
2337 break;
Chris Masone02119d2008-09-05 16:13:11 -04002338 }
Yan, Zhengc71bf092009-11-12 09:34:40 +00002339
Chris Masone02119d2008-09-05 16:13:11 -04002340 ret = link_to_fixup_dir(wc->trans, root,
2341 path, key.objectid);
Josef Bacikb50c6e22013-04-25 15:55:30 -04002342 if (ret)
2343 break;
Chris Masone02119d2008-09-05 16:13:11 -04002344 }
Josef Bacikdd8e7212013-09-11 11:57:23 -04002345
2346 if (key.type == BTRFS_DIR_INDEX_KEY &&
2347 wc->stage == LOG_WALK_REPLAY_DIR_INDEX) {
2348 ret = replay_one_dir_item(wc->trans, root, path,
2349 eb, i, &key);
2350 if (ret)
2351 break;
2352 }
2353
Chris Masone02119d2008-09-05 16:13:11 -04002354 if (wc->stage < LOG_WALK_REPLAY_ALL)
2355 continue;
2356
2357 /* these keys are simply copied */
2358 if (key.type == BTRFS_XATTR_ITEM_KEY) {
2359 ret = overwrite_item(wc->trans, root, path,
2360 eb, i, &key);
Josef Bacikb50c6e22013-04-25 15:55:30 -04002361 if (ret)
2362 break;
Liu Bo2da1c662013-05-26 13:50:29 +00002363 } else if (key.type == BTRFS_INODE_REF_KEY ||
2364 key.type == BTRFS_INODE_EXTREF_KEY) {
Mark Fashehf1863732012-08-08 11:32:27 -07002365 ret = add_inode_ref(wc->trans, root, log, path,
2366 eb, i, &key);
Josef Bacikb50c6e22013-04-25 15:55:30 -04002367 if (ret && ret != -ENOENT)
2368 break;
2369 ret = 0;
Chris Masone02119d2008-09-05 16:13:11 -04002370 } else if (key.type == BTRFS_EXTENT_DATA_KEY) {
2371 ret = replay_one_extent(wc->trans, root, path,
2372 eb, i, &key);
Josef Bacikb50c6e22013-04-25 15:55:30 -04002373 if (ret)
2374 break;
Josef Bacikdd8e7212013-09-11 11:57:23 -04002375 } else if (key.type == BTRFS_DIR_ITEM_KEY) {
Chris Masone02119d2008-09-05 16:13:11 -04002376 ret = replay_one_dir_item(wc->trans, root, path,
2377 eb, i, &key);
Josef Bacikb50c6e22013-04-25 15:55:30 -04002378 if (ret)
2379 break;
Chris Masone02119d2008-09-05 16:13:11 -04002380 }
2381 }
2382 btrfs_free_path(path);
Josef Bacikb50c6e22013-04-25 15:55:30 -04002383 return ret;
Chris Masone02119d2008-09-05 16:13:11 -04002384}
2385
Chris Masond3977122009-01-05 21:25:51 -05002386static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans,
Chris Masone02119d2008-09-05 16:13:11 -04002387 struct btrfs_root *root,
2388 struct btrfs_path *path, int *level,
2389 struct walk_control *wc)
2390{
2391 u64 root_owner;
Chris Masone02119d2008-09-05 16:13:11 -04002392 u64 bytenr;
2393 u64 ptr_gen;
2394 struct extent_buffer *next;
2395 struct extent_buffer *cur;
2396 struct extent_buffer *parent;
2397 u32 blocksize;
2398 int ret = 0;
2399
2400 WARN_ON(*level < 0);
2401 WARN_ON(*level >= BTRFS_MAX_LEVEL);
2402
Chris Masond3977122009-01-05 21:25:51 -05002403 while (*level > 0) {
Chris Masone02119d2008-09-05 16:13:11 -04002404 WARN_ON(*level < 0);
2405 WARN_ON(*level >= BTRFS_MAX_LEVEL);
2406 cur = path->nodes[*level];
2407
Dulshani Gunawardhanafae7f212013-10-31 10:30:08 +05302408 WARN_ON(btrfs_header_level(cur) != *level);
Chris Masone02119d2008-09-05 16:13:11 -04002409
2410 if (path->slots[*level] >=
2411 btrfs_header_nritems(cur))
2412 break;
2413
2414 bytenr = btrfs_node_blockptr(cur, path->slots[*level]);
2415 ptr_gen = btrfs_node_ptr_generation(cur, path->slots[*level]);
David Sterba707e8a02014-06-04 19:22:26 +02002416 blocksize = root->nodesize;
Chris Masone02119d2008-09-05 16:13:11 -04002417
2418 parent = path->nodes[*level];
2419 root_owner = btrfs_header_owner(parent);
Chris Masone02119d2008-09-05 16:13:11 -04002420
David Sterbaa83fffb2014-06-15 02:39:54 +02002421 next = btrfs_find_create_tree_block(root, bytenr);
liubo2a29edc2011-01-26 06:22:08 +00002422 if (!next)
2423 return -ENOMEM;
Chris Masone02119d2008-09-05 16:13:11 -04002424
Chris Masone02119d2008-09-05 16:13:11 -04002425 if (*level == 1) {
Mark Fasheh1e5063d2011-07-12 10:46:06 -07002426 ret = wc->process_func(root, next, wc, ptr_gen);
Josef Bacikb50c6e22013-04-25 15:55:30 -04002427 if (ret) {
2428 free_extent_buffer(next);
Mark Fasheh1e5063d2011-07-12 10:46:06 -07002429 return ret;
Josef Bacikb50c6e22013-04-25 15:55:30 -04002430 }
Yan, Zheng4a500fd2010-05-16 10:49:59 -04002431
Chris Masone02119d2008-09-05 16:13:11 -04002432 path->slots[*level]++;
2433 if (wc->free) {
Tsutomu Itoh018642a2012-05-29 18:10:13 +09002434 ret = btrfs_read_buffer(next, ptr_gen);
2435 if (ret) {
2436 free_extent_buffer(next);
2437 return ret;
2438 }
Chris Masone02119d2008-09-05 16:13:11 -04002439
Josef Bacik681ae502013-10-07 15:11:00 -04002440 if (trans) {
2441 btrfs_tree_lock(next);
2442 btrfs_set_lock_blocking(next);
Daniel Dressler01d58472014-11-21 17:15:07 +09002443 clean_tree_block(trans, root->fs_info,
2444 next);
Josef Bacik681ae502013-10-07 15:11:00 -04002445 btrfs_wait_tree_block_writeback(next);
2446 btrfs_tree_unlock(next);
2447 }
Chris Masone02119d2008-09-05 16:13:11 -04002448
Chris Masone02119d2008-09-05 16:13:11 -04002449 WARN_ON(root_owner !=
2450 BTRFS_TREE_LOG_OBJECTID);
Chris Masone688b7252011-10-31 20:52:39 -04002451 ret = btrfs_free_and_pin_reserved_extent(root,
Chris Masond00aff02008-09-11 15:54:42 -04002452 bytenr, blocksize);
Josef Bacik36508602013-04-25 16:23:32 -04002453 if (ret) {
2454 free_extent_buffer(next);
2455 return ret;
2456 }
Chris Masone02119d2008-09-05 16:13:11 -04002457 }
2458 free_extent_buffer(next);
2459 continue;
2460 }
Tsutomu Itoh018642a2012-05-29 18:10:13 +09002461 ret = btrfs_read_buffer(next, ptr_gen);
2462 if (ret) {
2463 free_extent_buffer(next);
2464 return ret;
2465 }
Chris Masone02119d2008-09-05 16:13:11 -04002466
2467 WARN_ON(*level <= 0);
2468 if (path->nodes[*level-1])
2469 free_extent_buffer(path->nodes[*level-1]);
2470 path->nodes[*level-1] = next;
2471 *level = btrfs_header_level(next);
2472 path->slots[*level] = 0;
2473 cond_resched();
2474 }
2475 WARN_ON(*level < 0);
2476 WARN_ON(*level >= BTRFS_MAX_LEVEL);
2477
Yan, Zheng4a500fd2010-05-16 10:49:59 -04002478 path->slots[*level] = btrfs_header_nritems(path->nodes[*level]);
Chris Masone02119d2008-09-05 16:13:11 -04002479
2480 cond_resched();
2481 return 0;
2482}
2483
Chris Masond3977122009-01-05 21:25:51 -05002484static noinline int walk_up_log_tree(struct btrfs_trans_handle *trans,
Chris Masone02119d2008-09-05 16:13:11 -04002485 struct btrfs_root *root,
2486 struct btrfs_path *path, int *level,
2487 struct walk_control *wc)
2488{
2489 u64 root_owner;
Chris Masone02119d2008-09-05 16:13:11 -04002490 int i;
2491 int slot;
2492 int ret;
2493
Chris Masond3977122009-01-05 21:25:51 -05002494 for (i = *level; i < BTRFS_MAX_LEVEL - 1 && path->nodes[i]; i++) {
Chris Masone02119d2008-09-05 16:13:11 -04002495 slot = path->slots[i];
Yan, Zheng4a500fd2010-05-16 10:49:59 -04002496 if (slot + 1 < btrfs_header_nritems(path->nodes[i])) {
Chris Masone02119d2008-09-05 16:13:11 -04002497 path->slots[i]++;
2498 *level = i;
2499 WARN_ON(*level == 0);
2500 return 0;
2501 } else {
Zheng Yan31840ae2008-09-23 13:14:14 -04002502 struct extent_buffer *parent;
2503 if (path->nodes[*level] == root->node)
2504 parent = path->nodes[*level];
2505 else
2506 parent = path->nodes[*level + 1];
2507
2508 root_owner = btrfs_header_owner(parent);
Mark Fasheh1e5063d2011-07-12 10:46:06 -07002509 ret = wc->process_func(root, path->nodes[*level], wc,
Chris Masone02119d2008-09-05 16:13:11 -04002510 btrfs_header_generation(path->nodes[*level]));
Mark Fasheh1e5063d2011-07-12 10:46:06 -07002511 if (ret)
2512 return ret;
2513
Chris Masone02119d2008-09-05 16:13:11 -04002514 if (wc->free) {
2515 struct extent_buffer *next;
2516
2517 next = path->nodes[*level];
2518
Josef Bacik681ae502013-10-07 15:11:00 -04002519 if (trans) {
2520 btrfs_tree_lock(next);
2521 btrfs_set_lock_blocking(next);
Daniel Dressler01d58472014-11-21 17:15:07 +09002522 clean_tree_block(trans, root->fs_info,
2523 next);
Josef Bacik681ae502013-10-07 15:11:00 -04002524 btrfs_wait_tree_block_writeback(next);
2525 btrfs_tree_unlock(next);
2526 }
Chris Masone02119d2008-09-05 16:13:11 -04002527
Chris Masone02119d2008-09-05 16:13:11 -04002528 WARN_ON(root_owner != BTRFS_TREE_LOG_OBJECTID);
Chris Masone688b7252011-10-31 20:52:39 -04002529 ret = btrfs_free_and_pin_reserved_extent(root,
Chris Masone02119d2008-09-05 16:13:11 -04002530 path->nodes[*level]->start,
Chris Masond00aff02008-09-11 15:54:42 -04002531 path->nodes[*level]->len);
Josef Bacik36508602013-04-25 16:23:32 -04002532 if (ret)
2533 return ret;
Chris Masone02119d2008-09-05 16:13:11 -04002534 }
2535 free_extent_buffer(path->nodes[*level]);
2536 path->nodes[*level] = NULL;
2537 *level = i + 1;
2538 }
2539 }
2540 return 1;
2541}
2542
2543/*
2544 * drop the reference count on the tree rooted at 'snap'. This traverses
2545 * the tree freeing any blocks that have a ref count of zero after being
2546 * decremented.
2547 */
2548static int walk_log_tree(struct btrfs_trans_handle *trans,
2549 struct btrfs_root *log, struct walk_control *wc)
2550{
2551 int ret = 0;
2552 int wret;
2553 int level;
2554 struct btrfs_path *path;
Chris Masone02119d2008-09-05 16:13:11 -04002555 int orig_level;
2556
2557 path = btrfs_alloc_path();
Tsutomu Itohdb5b4932011-03-23 08:14:16 +00002558 if (!path)
2559 return -ENOMEM;
Chris Masone02119d2008-09-05 16:13:11 -04002560
2561 level = btrfs_header_level(log->node);
2562 orig_level = level;
2563 path->nodes[level] = log->node;
2564 extent_buffer_get(log->node);
2565 path->slots[level] = 0;
2566
Chris Masond3977122009-01-05 21:25:51 -05002567 while (1) {
Chris Masone02119d2008-09-05 16:13:11 -04002568 wret = walk_down_log_tree(trans, log, path, &level, wc);
2569 if (wret > 0)
2570 break;
Jeff Mahoney79787ea2012-03-12 16:03:00 +01002571 if (wret < 0) {
Chris Masone02119d2008-09-05 16:13:11 -04002572 ret = wret;
Jeff Mahoney79787ea2012-03-12 16:03:00 +01002573 goto out;
2574 }
Chris Masone02119d2008-09-05 16:13:11 -04002575
2576 wret = walk_up_log_tree(trans, log, path, &level, wc);
2577 if (wret > 0)
2578 break;
Jeff Mahoney79787ea2012-03-12 16:03:00 +01002579 if (wret < 0) {
Chris Masone02119d2008-09-05 16:13:11 -04002580 ret = wret;
Jeff Mahoney79787ea2012-03-12 16:03:00 +01002581 goto out;
2582 }
Chris Masone02119d2008-09-05 16:13:11 -04002583 }
2584
2585 /* was the root node processed? if not, catch it here */
2586 if (path->nodes[orig_level]) {
Jeff Mahoney79787ea2012-03-12 16:03:00 +01002587 ret = wc->process_func(log, path->nodes[orig_level], wc,
Chris Masone02119d2008-09-05 16:13:11 -04002588 btrfs_header_generation(path->nodes[orig_level]));
Jeff Mahoney79787ea2012-03-12 16:03:00 +01002589 if (ret)
2590 goto out;
Chris Masone02119d2008-09-05 16:13:11 -04002591 if (wc->free) {
2592 struct extent_buffer *next;
2593
2594 next = path->nodes[orig_level];
2595
Josef Bacik681ae502013-10-07 15:11:00 -04002596 if (trans) {
2597 btrfs_tree_lock(next);
2598 btrfs_set_lock_blocking(next);
Daniel Dressler01d58472014-11-21 17:15:07 +09002599 clean_tree_block(trans, log->fs_info, next);
Josef Bacik681ae502013-10-07 15:11:00 -04002600 btrfs_wait_tree_block_writeback(next);
2601 btrfs_tree_unlock(next);
2602 }
Chris Masone02119d2008-09-05 16:13:11 -04002603
Chris Masone02119d2008-09-05 16:13:11 -04002604 WARN_ON(log->root_key.objectid !=
2605 BTRFS_TREE_LOG_OBJECTID);
Chris Masone688b7252011-10-31 20:52:39 -04002606 ret = btrfs_free_and_pin_reserved_extent(log, next->start,
Chris Masond00aff02008-09-11 15:54:42 -04002607 next->len);
Josef Bacik36508602013-04-25 16:23:32 -04002608 if (ret)
2609 goto out;
Chris Masone02119d2008-09-05 16:13:11 -04002610 }
2611 }
2612
Jeff Mahoney79787ea2012-03-12 16:03:00 +01002613out:
Chris Masone02119d2008-09-05 16:13:11 -04002614 btrfs_free_path(path);
Chris Masone02119d2008-09-05 16:13:11 -04002615 return ret;
2616}
2617
Yan Zheng7237f182009-01-21 12:54:03 -05002618/*
2619 * helper function to update the item for a given subvolumes log root
2620 * in the tree of log roots
2621 */
2622static int update_log_root(struct btrfs_trans_handle *trans,
2623 struct btrfs_root *log)
2624{
2625 int ret;
2626
2627 if (log->log_transid == 1) {
2628 /* insert root item on the first sync */
2629 ret = btrfs_insert_root(trans, log->fs_info->log_root_tree,
2630 &log->root_key, &log->root_item);
2631 } else {
2632 ret = btrfs_update_root(trans, log->fs_info->log_root_tree,
2633 &log->root_key, &log->root_item);
2634 }
2635 return ret;
2636}
2637
Zhaolei60d53eb2015-08-17 18:44:46 +08002638static void wait_log_commit(struct btrfs_root *root, int transid)
Chris Masone02119d2008-09-05 16:13:11 -04002639{
2640 DEFINE_WAIT(wait);
Yan Zheng7237f182009-01-21 12:54:03 -05002641 int index = transid % 2;
Chris Masone02119d2008-09-05 16:13:11 -04002642
Yan Zheng7237f182009-01-21 12:54:03 -05002643 /*
2644 * we only allow two pending log transactions at a time,
2645 * so we know that if ours is more than 2 older than the
2646 * current transaction, we're done
2647 */
Chris Masone02119d2008-09-05 16:13:11 -04002648 do {
Yan Zheng7237f182009-01-21 12:54:03 -05002649 prepare_to_wait(&root->log_commit_wait[index],
2650 &wait, TASK_UNINTERRUPTIBLE);
2651 mutex_unlock(&root->log_mutex);
Chris Mason12fcfd22009-03-24 10:24:20 -04002652
Miao Xied1433de2014-02-20 18:08:59 +08002653 if (root->log_transid_committed < transid &&
Yan Zheng7237f182009-01-21 12:54:03 -05002654 atomic_read(&root->log_commit[index]))
Chris Masone02119d2008-09-05 16:13:11 -04002655 schedule();
Chris Mason12fcfd22009-03-24 10:24:20 -04002656
Yan Zheng7237f182009-01-21 12:54:03 -05002657 finish_wait(&root->log_commit_wait[index], &wait);
2658 mutex_lock(&root->log_mutex);
Miao Xied1433de2014-02-20 18:08:59 +08002659 } while (root->log_transid_committed < transid &&
Yan Zheng7237f182009-01-21 12:54:03 -05002660 atomic_read(&root->log_commit[index]));
Yan Zheng7237f182009-01-21 12:54:03 -05002661}
2662
Zhaolei60d53eb2015-08-17 18:44:46 +08002663static void wait_for_writer(struct btrfs_root *root)
Yan Zheng7237f182009-01-21 12:54:03 -05002664{
2665 DEFINE_WAIT(wait);
Miao Xie8b050d32014-02-20 18:08:58 +08002666
2667 while (atomic_read(&root->log_writers)) {
Yan Zheng7237f182009-01-21 12:54:03 -05002668 prepare_to_wait(&root->log_writer_wait,
2669 &wait, TASK_UNINTERRUPTIBLE);
2670 mutex_unlock(&root->log_mutex);
Miao Xie8b050d32014-02-20 18:08:58 +08002671 if (atomic_read(&root->log_writers))
Yan Zheng7237f182009-01-21 12:54:03 -05002672 schedule();
Yan Zheng7237f182009-01-21 12:54:03 -05002673 finish_wait(&root->log_writer_wait, &wait);
Filipe Manana575849e2015-02-11 11:12:39 +00002674 mutex_lock(&root->log_mutex);
Yan Zheng7237f182009-01-21 12:54:03 -05002675 }
Chris Masone02119d2008-09-05 16:13:11 -04002676}
2677
Miao Xie8b050d32014-02-20 18:08:58 +08002678static inline void btrfs_remove_log_ctx(struct btrfs_root *root,
2679 struct btrfs_log_ctx *ctx)
2680{
2681 if (!ctx)
2682 return;
2683
2684 mutex_lock(&root->log_mutex);
2685 list_del_init(&ctx->list);
2686 mutex_unlock(&root->log_mutex);
2687}
2688
2689/*
2690 * Invoked in log mutex context, or be sure there is no other task which
2691 * can access the list.
2692 */
2693static inline void btrfs_remove_all_log_ctxs(struct btrfs_root *root,
2694 int index, int error)
2695{
2696 struct btrfs_log_ctx *ctx;
2697
2698 if (!error) {
2699 INIT_LIST_HEAD(&root->log_ctxs[index]);
2700 return;
2701 }
2702
2703 list_for_each_entry(ctx, &root->log_ctxs[index], list)
2704 ctx->log_ret = error;
2705
2706 INIT_LIST_HEAD(&root->log_ctxs[index]);
2707}
2708
Chris Masone02119d2008-09-05 16:13:11 -04002709/*
2710 * btrfs_sync_log does sends a given tree log down to the disk and
2711 * updates the super blocks to record it. When this call is done,
Chris Mason12fcfd22009-03-24 10:24:20 -04002712 * you know that any inodes previously logged are safely on disk only
2713 * if it returns 0.
2714 *
2715 * Any other return value means you need to call btrfs_commit_transaction.
2716 * Some of the edge cases for fsyncing directories that have had unlinks
2717 * or renames done in the past mean that sometimes the only safe
2718 * fsync is to commit the whole FS. When btrfs_sync_log returns -EAGAIN,
2719 * that has happened.
Chris Masone02119d2008-09-05 16:13:11 -04002720 */
2721int btrfs_sync_log(struct btrfs_trans_handle *trans,
Miao Xie8b050d32014-02-20 18:08:58 +08002722 struct btrfs_root *root, struct btrfs_log_ctx *ctx)
Chris Masone02119d2008-09-05 16:13:11 -04002723{
Yan Zheng7237f182009-01-21 12:54:03 -05002724 int index1;
2725 int index2;
Yan, Zheng8cef4e12009-11-12 09:33:26 +00002726 int mark;
Chris Masone02119d2008-09-05 16:13:11 -04002727 int ret;
Chris Masone02119d2008-09-05 16:13:11 -04002728 struct btrfs_root *log = root->log_root;
Yan Zheng7237f182009-01-21 12:54:03 -05002729 struct btrfs_root *log_root_tree = root->fs_info->log_root_tree;
Miao Xiebb14a592014-02-20 18:08:56 +08002730 int log_transid = 0;
Miao Xie8b050d32014-02-20 18:08:58 +08002731 struct btrfs_log_ctx root_log_ctx;
Miao Xiec6adc9c2013-05-28 10:05:39 +00002732 struct blk_plug plug;
Chris Masone02119d2008-09-05 16:13:11 -04002733
Yan Zheng7237f182009-01-21 12:54:03 -05002734 mutex_lock(&root->log_mutex);
Miao Xied1433de2014-02-20 18:08:59 +08002735 log_transid = ctx->log_transid;
2736 if (root->log_transid_committed >= log_transid) {
Yan Zheng7237f182009-01-21 12:54:03 -05002737 mutex_unlock(&root->log_mutex);
Miao Xie8b050d32014-02-20 18:08:58 +08002738 return ctx->log_ret;
Chris Masone02119d2008-09-05 16:13:11 -04002739 }
Miao Xied1433de2014-02-20 18:08:59 +08002740
2741 index1 = log_transid % 2;
2742 if (atomic_read(&root->log_commit[index1])) {
Zhaolei60d53eb2015-08-17 18:44:46 +08002743 wait_log_commit(root, log_transid);
Miao Xied1433de2014-02-20 18:08:59 +08002744 mutex_unlock(&root->log_mutex);
2745 return ctx->log_ret;
2746 }
2747 ASSERT(log_transid == root->log_transid);
Yan Zheng7237f182009-01-21 12:54:03 -05002748 atomic_set(&root->log_commit[index1], 1);
2749
2750 /* wait for previous tree log sync to complete */
2751 if (atomic_read(&root->log_commit[(index1 + 1) % 2]))
Zhaolei60d53eb2015-08-17 18:44:46 +08002752 wait_log_commit(root, log_transid - 1);
Miao Xie48cab2e2014-02-20 18:08:52 +08002753
Yan, Zheng86df7eb2009-10-14 09:24:59 -04002754 while (1) {
Miao Xie2ecb7922012-09-06 04:04:27 -06002755 int batch = atomic_read(&root->log_batch);
Chris Masoncd354ad2011-10-20 15:45:37 -04002756 /* when we're on an ssd, just kick the log commit out */
Miao Xie27cdeb72014-04-02 19:51:05 +08002757 if (!btrfs_test_opt(root, SSD) &&
2758 test_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state)) {
Yan, Zheng86df7eb2009-10-14 09:24:59 -04002759 mutex_unlock(&root->log_mutex);
2760 schedule_timeout_uninterruptible(1);
2761 mutex_lock(&root->log_mutex);
2762 }
Zhaolei60d53eb2015-08-17 18:44:46 +08002763 wait_for_writer(root);
Miao Xie2ecb7922012-09-06 04:04:27 -06002764 if (batch == atomic_read(&root->log_batch))
Chris Masone02119d2008-09-05 16:13:11 -04002765 break;
2766 }
Chris Masond0c803c2008-09-11 16:17:57 -04002767
Chris Mason12fcfd22009-03-24 10:24:20 -04002768 /* bail out if we need to do a full commit */
Miao Xie995946d2014-04-02 19:51:06 +08002769 if (btrfs_need_log_full_commit(root->fs_info, trans)) {
Chris Mason12fcfd22009-03-24 10:24:20 -04002770 ret = -EAGAIN;
Josef Bacik2ab28f32012-10-12 15:27:49 -04002771 btrfs_free_logged_extents(log, log_transid);
Chris Mason12fcfd22009-03-24 10:24:20 -04002772 mutex_unlock(&root->log_mutex);
2773 goto out;
2774 }
2775
Yan, Zheng8cef4e12009-11-12 09:33:26 +00002776 if (log_transid % 2 == 0)
2777 mark = EXTENT_DIRTY;
2778 else
2779 mark = EXTENT_NEW;
2780
Chris Mason690587d2009-10-13 13:29:19 -04002781 /* we start IO on all the marked extents here, but we don't actually
2782 * wait for them until later.
2783 */
Miao Xiec6adc9c2013-05-28 10:05:39 +00002784 blk_start_plug(&plug);
Yan, Zheng8cef4e12009-11-12 09:33:26 +00002785 ret = btrfs_write_marked_extents(log, &log->dirty_log_pages, mark);
Jeff Mahoney79787ea2012-03-12 16:03:00 +01002786 if (ret) {
Miao Xiec6adc9c2013-05-28 10:05:39 +00002787 blk_finish_plug(&plug);
Jeff Mahoney79787ea2012-03-12 16:03:00 +01002788 btrfs_abort_transaction(trans, root, ret);
Josef Bacik2ab28f32012-10-12 15:27:49 -04002789 btrfs_free_logged_extents(log, log_transid);
Miao Xie995946d2014-04-02 19:51:06 +08002790 btrfs_set_log_full_commit(root->fs_info, trans);
Jeff Mahoney79787ea2012-03-12 16:03:00 +01002791 mutex_unlock(&root->log_mutex);
2792 goto out;
2793 }
Yan Zheng7237f182009-01-21 12:54:03 -05002794
Yan Zheng5d4f98a2009-06-10 10:45:14 -04002795 btrfs_set_root_node(&log->root_item, log->node);
Yan Zheng7237f182009-01-21 12:54:03 -05002796
Yan Zheng7237f182009-01-21 12:54:03 -05002797 root->log_transid++;
2798 log->log_transid = root->log_transid;
Josef Bacikff782e02009-10-08 15:30:04 -04002799 root->log_start_pid = 0;
Yan Zheng7237f182009-01-21 12:54:03 -05002800 /*
Yan, Zheng8cef4e12009-11-12 09:33:26 +00002801 * IO has been started, blocks of the log tree have WRITTEN flag set
2802 * in their headers. new modifications of the log will be written to
2803 * new positions. so it's safe to allow log writers to go in.
Yan Zheng7237f182009-01-21 12:54:03 -05002804 */
2805 mutex_unlock(&root->log_mutex);
2806
Miao Xied1433de2014-02-20 18:08:59 +08002807 btrfs_init_log_ctx(&root_log_ctx);
2808
Yan Zheng7237f182009-01-21 12:54:03 -05002809 mutex_lock(&log_root_tree->log_mutex);
Miao Xie2ecb7922012-09-06 04:04:27 -06002810 atomic_inc(&log_root_tree->log_batch);
Yan Zheng7237f182009-01-21 12:54:03 -05002811 atomic_inc(&log_root_tree->log_writers);
Miao Xied1433de2014-02-20 18:08:59 +08002812
2813 index2 = log_root_tree->log_transid % 2;
2814 list_add_tail(&root_log_ctx.list, &log_root_tree->log_ctxs[index2]);
2815 root_log_ctx.log_transid = log_root_tree->log_transid;
2816
Yan Zheng7237f182009-01-21 12:54:03 -05002817 mutex_unlock(&log_root_tree->log_mutex);
2818
2819 ret = update_log_root(trans, log);
Yan Zheng7237f182009-01-21 12:54:03 -05002820
2821 mutex_lock(&log_root_tree->log_mutex);
2822 if (atomic_dec_and_test(&log_root_tree->log_writers)) {
2823 smp_mb();
2824 if (waitqueue_active(&log_root_tree->log_writer_wait))
2825 wake_up(&log_root_tree->log_writer_wait);
2826 }
2827
Yan, Zheng4a500fd2010-05-16 10:49:59 -04002828 if (ret) {
Miao Xied1433de2014-02-20 18:08:59 +08002829 if (!list_empty(&root_log_ctx.list))
2830 list_del_init(&root_log_ctx.list);
2831
Miao Xiec6adc9c2013-05-28 10:05:39 +00002832 blk_finish_plug(&plug);
Miao Xie995946d2014-04-02 19:51:06 +08002833 btrfs_set_log_full_commit(root->fs_info, trans);
2834
Jeff Mahoney79787ea2012-03-12 16:03:00 +01002835 if (ret != -ENOSPC) {
2836 btrfs_abort_transaction(trans, root, ret);
2837 mutex_unlock(&log_root_tree->log_mutex);
2838 goto out;
2839 }
Yan, Zheng4a500fd2010-05-16 10:49:59 -04002840 btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark);
Josef Bacik2ab28f32012-10-12 15:27:49 -04002841 btrfs_free_logged_extents(log, log_transid);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04002842 mutex_unlock(&log_root_tree->log_mutex);
2843 ret = -EAGAIN;
2844 goto out;
2845 }
2846
Miao Xied1433de2014-02-20 18:08:59 +08002847 if (log_root_tree->log_transid_committed >= root_log_ctx.log_transid) {
Forrest Liu3da5ab52015-01-30 19:42:12 +08002848 blk_finish_plug(&plug);
Miao Xied1433de2014-02-20 18:08:59 +08002849 mutex_unlock(&log_root_tree->log_mutex);
2850 ret = root_log_ctx.log_ret;
2851 goto out;
2852 }
Miao Xie8b050d32014-02-20 18:08:58 +08002853
Miao Xied1433de2014-02-20 18:08:59 +08002854 index2 = root_log_ctx.log_transid % 2;
Yan Zheng7237f182009-01-21 12:54:03 -05002855 if (atomic_read(&log_root_tree->log_commit[index2])) {
Miao Xiec6adc9c2013-05-28 10:05:39 +00002856 blk_finish_plug(&plug);
Filipe Manana5ab5e442014-11-13 16:59:53 +00002857 ret = btrfs_wait_marked_extents(log, &log->dirty_log_pages,
2858 mark);
Josef Bacik50d9aa92014-11-21 14:52:38 -05002859 btrfs_wait_logged_extents(trans, log, log_transid);
Zhaolei60d53eb2015-08-17 18:44:46 +08002860 wait_log_commit(log_root_tree,
Miao Xied1433de2014-02-20 18:08:59 +08002861 root_log_ctx.log_transid);
Yan Zheng7237f182009-01-21 12:54:03 -05002862 mutex_unlock(&log_root_tree->log_mutex);
Filipe Manana5ab5e442014-11-13 16:59:53 +00002863 if (!ret)
2864 ret = root_log_ctx.log_ret;
Yan Zheng7237f182009-01-21 12:54:03 -05002865 goto out;
2866 }
Miao Xied1433de2014-02-20 18:08:59 +08002867 ASSERT(root_log_ctx.log_transid == log_root_tree->log_transid);
Yan Zheng7237f182009-01-21 12:54:03 -05002868 atomic_set(&log_root_tree->log_commit[index2], 1);
2869
Chris Mason12fcfd22009-03-24 10:24:20 -04002870 if (atomic_read(&log_root_tree->log_commit[(index2 + 1) % 2])) {
Zhaolei60d53eb2015-08-17 18:44:46 +08002871 wait_log_commit(log_root_tree,
Miao Xied1433de2014-02-20 18:08:59 +08002872 root_log_ctx.log_transid - 1);
Chris Mason12fcfd22009-03-24 10:24:20 -04002873 }
Yan Zheng7237f182009-01-21 12:54:03 -05002874
Zhaolei60d53eb2015-08-17 18:44:46 +08002875 wait_for_writer(log_root_tree);
Chris Mason12fcfd22009-03-24 10:24:20 -04002876
2877 /*
2878 * now that we've moved on to the tree of log tree roots,
2879 * check the full commit flag again
2880 */
Miao Xie995946d2014-04-02 19:51:06 +08002881 if (btrfs_need_log_full_commit(root->fs_info, trans)) {
Miao Xiec6adc9c2013-05-28 10:05:39 +00002882 blk_finish_plug(&plug);
Yan, Zheng8cef4e12009-11-12 09:33:26 +00002883 btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark);
Josef Bacik2ab28f32012-10-12 15:27:49 -04002884 btrfs_free_logged_extents(log, log_transid);
Chris Mason12fcfd22009-03-24 10:24:20 -04002885 mutex_unlock(&log_root_tree->log_mutex);
2886 ret = -EAGAIN;
2887 goto out_wake_log_root;
2888 }
Yan Zheng7237f182009-01-21 12:54:03 -05002889
Miao Xiec6adc9c2013-05-28 10:05:39 +00002890 ret = btrfs_write_marked_extents(log_root_tree,
2891 &log_root_tree->dirty_log_pages,
2892 EXTENT_DIRTY | EXTENT_NEW);
2893 blk_finish_plug(&plug);
Jeff Mahoney79787ea2012-03-12 16:03:00 +01002894 if (ret) {
Miao Xie995946d2014-04-02 19:51:06 +08002895 btrfs_set_log_full_commit(root->fs_info, trans);
Jeff Mahoney79787ea2012-03-12 16:03:00 +01002896 btrfs_abort_transaction(trans, root, ret);
Josef Bacik2ab28f32012-10-12 15:27:49 -04002897 btrfs_free_logged_extents(log, log_transid);
Jeff Mahoney79787ea2012-03-12 16:03:00 +01002898 mutex_unlock(&log_root_tree->log_mutex);
2899 goto out_wake_log_root;
2900 }
Filipe Manana5ab5e442014-11-13 16:59:53 +00002901 ret = btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark);
2902 if (!ret)
2903 ret = btrfs_wait_marked_extents(log_root_tree,
2904 &log_root_tree->dirty_log_pages,
2905 EXTENT_NEW | EXTENT_DIRTY);
2906 if (ret) {
2907 btrfs_set_log_full_commit(root->fs_info, trans);
2908 btrfs_free_logged_extents(log, log_transid);
2909 mutex_unlock(&log_root_tree->log_mutex);
2910 goto out_wake_log_root;
2911 }
Josef Bacik50d9aa92014-11-21 14:52:38 -05002912 btrfs_wait_logged_extents(trans, log, log_transid);
Chris Masone02119d2008-09-05 16:13:11 -04002913
David Sterba6c417612011-04-13 15:41:04 +02002914 btrfs_set_super_log_root(root->fs_info->super_for_commit,
Yan Zheng7237f182009-01-21 12:54:03 -05002915 log_root_tree->node->start);
David Sterba6c417612011-04-13 15:41:04 +02002916 btrfs_set_super_log_root_level(root->fs_info->super_for_commit,
Yan Zheng7237f182009-01-21 12:54:03 -05002917 btrfs_header_level(log_root_tree->node));
Chris Masone02119d2008-09-05 16:13:11 -04002918
Yan Zheng7237f182009-01-21 12:54:03 -05002919 log_root_tree->log_transid++;
Yan Zheng7237f182009-01-21 12:54:03 -05002920 mutex_unlock(&log_root_tree->log_mutex);
2921
2922 /*
2923 * nobody else is going to jump in and write the the ctree
2924 * super here because the log_commit atomic below is protecting
2925 * us. We must be called with a transaction handle pinning
2926 * the running transaction open, so a full commit can't hop
2927 * in and cause problems either.
2928 */
Stefan Behrens5af3e8c2012-08-01 18:56:49 +02002929 ret = write_ctree_super(trans, root->fs_info->tree_root, 1);
Stefan Behrens5af3e8c2012-08-01 18:56:49 +02002930 if (ret) {
Miao Xie995946d2014-04-02 19:51:06 +08002931 btrfs_set_log_full_commit(root->fs_info, trans);
Stefan Behrens5af3e8c2012-08-01 18:56:49 +02002932 btrfs_abort_transaction(trans, root, ret);
2933 goto out_wake_log_root;
2934 }
Yan Zheng7237f182009-01-21 12:54:03 -05002935
Chris Mason257c62e2009-10-13 13:21:08 -04002936 mutex_lock(&root->log_mutex);
2937 if (root->last_log_commit < log_transid)
2938 root->last_log_commit = log_transid;
2939 mutex_unlock(&root->log_mutex);
2940
Chris Mason12fcfd22009-03-24 10:24:20 -04002941out_wake_log_root:
Miao Xie8b050d32014-02-20 18:08:58 +08002942 /*
2943 * We needn't get log_mutex here because we are sure all
2944 * the other tasks are blocked.
2945 */
2946 btrfs_remove_all_log_ctxs(log_root_tree, index2, ret);
2947
Miao Xied1433de2014-02-20 18:08:59 +08002948 mutex_lock(&log_root_tree->log_mutex);
2949 log_root_tree->log_transid_committed++;
Yan Zheng7237f182009-01-21 12:54:03 -05002950 atomic_set(&log_root_tree->log_commit[index2], 0);
Miao Xied1433de2014-02-20 18:08:59 +08002951 mutex_unlock(&log_root_tree->log_mutex);
2952
Yan Zheng7237f182009-01-21 12:54:03 -05002953 if (waitqueue_active(&log_root_tree->log_commit_wait[index2]))
2954 wake_up(&log_root_tree->log_commit_wait[index2]);
Chris Masone02119d2008-09-05 16:13:11 -04002955out:
Miao Xie8b050d32014-02-20 18:08:58 +08002956 /* See above. */
2957 btrfs_remove_all_log_ctxs(root, index1, ret);
2958
Miao Xied1433de2014-02-20 18:08:59 +08002959 mutex_lock(&root->log_mutex);
2960 root->log_transid_committed++;
Yan Zheng7237f182009-01-21 12:54:03 -05002961 atomic_set(&root->log_commit[index1], 0);
Miao Xied1433de2014-02-20 18:08:59 +08002962 mutex_unlock(&root->log_mutex);
Miao Xie8b050d32014-02-20 18:08:58 +08002963
Yan Zheng7237f182009-01-21 12:54:03 -05002964 if (waitqueue_active(&root->log_commit_wait[index1]))
2965 wake_up(&root->log_commit_wait[index1]);
Chris Masonb31eabd2011-01-31 16:48:24 -05002966 return ret;
Chris Masone02119d2008-09-05 16:13:11 -04002967}
2968
Yan, Zheng4a500fd2010-05-16 10:49:59 -04002969static void free_log_tree(struct btrfs_trans_handle *trans,
2970 struct btrfs_root *log)
Chris Masone02119d2008-09-05 16:13:11 -04002971{
2972 int ret;
Chris Masond0c803c2008-09-11 16:17:57 -04002973 u64 start;
2974 u64 end;
Chris Masone02119d2008-09-05 16:13:11 -04002975 struct walk_control wc = {
2976 .free = 1,
2977 .process_func = process_one_buffer
2978 };
2979
Josef Bacik681ae502013-10-07 15:11:00 -04002980 ret = walk_log_tree(trans, log, &wc);
2981 /* I don't think this can happen but just in case */
2982 if (ret)
2983 btrfs_abort_transaction(trans, log, ret);
Chris Masone02119d2008-09-05 16:13:11 -04002984
Chris Masond3977122009-01-05 21:25:51 -05002985 while (1) {
Chris Masond0c803c2008-09-11 16:17:57 -04002986 ret = find_first_extent_bit(&log->dirty_log_pages,
Josef Bacike6138872012-09-27 17:07:30 -04002987 0, &start, &end, EXTENT_DIRTY | EXTENT_NEW,
2988 NULL);
Chris Masond0c803c2008-09-11 16:17:57 -04002989 if (ret)
2990 break;
2991
Yan, Zheng8cef4e12009-11-12 09:33:26 +00002992 clear_extent_bits(&log->dirty_log_pages, start, end,
2993 EXTENT_DIRTY | EXTENT_NEW, GFP_NOFS);
Chris Masond0c803c2008-09-11 16:17:57 -04002994 }
2995
Josef Bacik2ab28f32012-10-12 15:27:49 -04002996 /*
2997 * We may have short-circuited the log tree with the full commit logic
2998 * and left ordered extents on our list, so clear these out to keep us
2999 * from leaking inodes and memory.
3000 */
3001 btrfs_free_logged_extents(log, 0);
3002 btrfs_free_logged_extents(log, 1);
3003
Yan Zheng7237f182009-01-21 12:54:03 -05003004 free_extent_buffer(log->node);
3005 kfree(log);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003006}
3007
3008/*
3009 * free all the extents used by the tree log. This should be called
3010 * at commit time of the full transaction
3011 */
3012int btrfs_free_log(struct btrfs_trans_handle *trans, struct btrfs_root *root)
3013{
3014 if (root->log_root) {
3015 free_log_tree(trans, root->log_root);
3016 root->log_root = NULL;
3017 }
3018 return 0;
3019}
3020
3021int btrfs_free_log_root_tree(struct btrfs_trans_handle *trans,
3022 struct btrfs_fs_info *fs_info)
3023{
3024 if (fs_info->log_root_tree) {
3025 free_log_tree(trans, fs_info->log_root_tree);
3026 fs_info->log_root_tree = NULL;
3027 }
Chris Masone02119d2008-09-05 16:13:11 -04003028 return 0;
3029}
3030
3031/*
Chris Masone02119d2008-09-05 16:13:11 -04003032 * If both a file and directory are logged, and unlinks or renames are
3033 * mixed in, we have a few interesting corners:
3034 *
3035 * create file X in dir Y
3036 * link file X to X.link in dir Y
3037 * fsync file X
3038 * unlink file X but leave X.link
3039 * fsync dir Y
3040 *
3041 * After a crash we would expect only X.link to exist. But file X
3042 * didn't get fsync'd again so the log has back refs for X and X.link.
3043 *
3044 * We solve this by removing directory entries and inode backrefs from the
3045 * log when a file that was logged in the current transaction is
3046 * unlinked. Any later fsync will include the updated log entries, and
3047 * we'll be able to reconstruct the proper directory items from backrefs.
3048 *
3049 * This optimizations allows us to avoid relogging the entire inode
3050 * or the entire directory.
3051 */
3052int btrfs_del_dir_entries_in_log(struct btrfs_trans_handle *trans,
3053 struct btrfs_root *root,
3054 const char *name, int name_len,
3055 struct inode *dir, u64 index)
3056{
3057 struct btrfs_root *log;
3058 struct btrfs_dir_item *di;
3059 struct btrfs_path *path;
3060 int ret;
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003061 int err = 0;
Chris Masone02119d2008-09-05 16:13:11 -04003062 int bytes_del = 0;
Li Zefan33345d012011-04-20 10:31:50 +08003063 u64 dir_ino = btrfs_ino(dir);
Chris Masone02119d2008-09-05 16:13:11 -04003064
Chris Mason3a5f1d42008-09-11 15:53:37 -04003065 if (BTRFS_I(dir)->logged_trans < trans->transid)
3066 return 0;
3067
Chris Masone02119d2008-09-05 16:13:11 -04003068 ret = join_running_log_trans(root);
3069 if (ret)
3070 return 0;
3071
3072 mutex_lock(&BTRFS_I(dir)->log_mutex);
3073
3074 log = root->log_root;
3075 path = btrfs_alloc_path();
Tsutomu Itoha62f44a2011-04-25 19:43:51 -04003076 if (!path) {
3077 err = -ENOMEM;
3078 goto out_unlock;
3079 }
liubo2a29edc2011-01-26 06:22:08 +00003080
Li Zefan33345d012011-04-20 10:31:50 +08003081 di = btrfs_lookup_dir_item(trans, log, path, dir_ino,
Chris Masone02119d2008-09-05 16:13:11 -04003082 name, name_len, -1);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003083 if (IS_ERR(di)) {
3084 err = PTR_ERR(di);
3085 goto fail;
3086 }
3087 if (di) {
Chris Masone02119d2008-09-05 16:13:11 -04003088 ret = btrfs_delete_one_dir_name(trans, log, path, di);
3089 bytes_del += name_len;
Josef Bacik36508602013-04-25 16:23:32 -04003090 if (ret) {
3091 err = ret;
3092 goto fail;
3093 }
Chris Masone02119d2008-09-05 16:13:11 -04003094 }
David Sterbab3b4aa72011-04-21 01:20:15 +02003095 btrfs_release_path(path);
Li Zefan33345d012011-04-20 10:31:50 +08003096 di = btrfs_lookup_dir_index_item(trans, log, path, dir_ino,
Chris Masone02119d2008-09-05 16:13:11 -04003097 index, name, name_len, -1);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003098 if (IS_ERR(di)) {
3099 err = PTR_ERR(di);
3100 goto fail;
3101 }
3102 if (di) {
Chris Masone02119d2008-09-05 16:13:11 -04003103 ret = btrfs_delete_one_dir_name(trans, log, path, di);
3104 bytes_del += name_len;
Josef Bacik36508602013-04-25 16:23:32 -04003105 if (ret) {
3106 err = ret;
3107 goto fail;
3108 }
Chris Masone02119d2008-09-05 16:13:11 -04003109 }
3110
3111 /* update the directory size in the log to reflect the names
3112 * we have removed
3113 */
3114 if (bytes_del) {
3115 struct btrfs_key key;
3116
Li Zefan33345d012011-04-20 10:31:50 +08003117 key.objectid = dir_ino;
Chris Masone02119d2008-09-05 16:13:11 -04003118 key.offset = 0;
3119 key.type = BTRFS_INODE_ITEM_KEY;
David Sterbab3b4aa72011-04-21 01:20:15 +02003120 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -04003121
3122 ret = btrfs_search_slot(trans, log, &key, path, 0, 1);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003123 if (ret < 0) {
3124 err = ret;
3125 goto fail;
3126 }
Chris Masone02119d2008-09-05 16:13:11 -04003127 if (ret == 0) {
3128 struct btrfs_inode_item *item;
3129 u64 i_size;
3130
3131 item = btrfs_item_ptr(path->nodes[0], path->slots[0],
3132 struct btrfs_inode_item);
3133 i_size = btrfs_inode_size(path->nodes[0], item);
3134 if (i_size > bytes_del)
3135 i_size -= bytes_del;
3136 else
3137 i_size = 0;
3138 btrfs_set_inode_size(path->nodes[0], item, i_size);
3139 btrfs_mark_buffer_dirty(path->nodes[0]);
3140 } else
3141 ret = 0;
David Sterbab3b4aa72011-04-21 01:20:15 +02003142 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -04003143 }
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003144fail:
Chris Masone02119d2008-09-05 16:13:11 -04003145 btrfs_free_path(path);
Tsutomu Itoha62f44a2011-04-25 19:43:51 -04003146out_unlock:
Chris Masone02119d2008-09-05 16:13:11 -04003147 mutex_unlock(&BTRFS_I(dir)->log_mutex);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003148 if (ret == -ENOSPC) {
Miao Xie995946d2014-04-02 19:51:06 +08003149 btrfs_set_log_full_commit(root->fs_info, trans);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003150 ret = 0;
Jeff Mahoney79787ea2012-03-12 16:03:00 +01003151 } else if (ret < 0)
3152 btrfs_abort_transaction(trans, root, ret);
3153
Chris Mason12fcfd22009-03-24 10:24:20 -04003154 btrfs_end_log_trans(root);
Chris Masone02119d2008-09-05 16:13:11 -04003155
Andi Kleen411fc6b2010-10-29 15:14:31 -04003156 return err;
Chris Masone02119d2008-09-05 16:13:11 -04003157}
3158
3159/* see comments for btrfs_del_dir_entries_in_log */
3160int btrfs_del_inode_ref_in_log(struct btrfs_trans_handle *trans,
3161 struct btrfs_root *root,
3162 const char *name, int name_len,
3163 struct inode *inode, u64 dirid)
3164{
3165 struct btrfs_root *log;
3166 u64 index;
3167 int ret;
3168
Chris Mason3a5f1d42008-09-11 15:53:37 -04003169 if (BTRFS_I(inode)->logged_trans < trans->transid)
3170 return 0;
3171
Chris Masone02119d2008-09-05 16:13:11 -04003172 ret = join_running_log_trans(root);
3173 if (ret)
3174 return 0;
3175 log = root->log_root;
3176 mutex_lock(&BTRFS_I(inode)->log_mutex);
3177
Li Zefan33345d012011-04-20 10:31:50 +08003178 ret = btrfs_del_inode_ref(trans, log, name, name_len, btrfs_ino(inode),
Chris Masone02119d2008-09-05 16:13:11 -04003179 dirid, &index);
3180 mutex_unlock(&BTRFS_I(inode)->log_mutex);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003181 if (ret == -ENOSPC) {
Miao Xie995946d2014-04-02 19:51:06 +08003182 btrfs_set_log_full_commit(root->fs_info, trans);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003183 ret = 0;
Jeff Mahoney79787ea2012-03-12 16:03:00 +01003184 } else if (ret < 0 && ret != -ENOENT)
3185 btrfs_abort_transaction(trans, root, ret);
Chris Mason12fcfd22009-03-24 10:24:20 -04003186 btrfs_end_log_trans(root);
Chris Masone02119d2008-09-05 16:13:11 -04003187
Chris Masone02119d2008-09-05 16:13:11 -04003188 return ret;
3189}
3190
3191/*
3192 * creates a range item in the log for 'dirid'. first_offset and
3193 * last_offset tell us which parts of the key space the log should
3194 * be considered authoritative for.
3195 */
3196static noinline int insert_dir_log_key(struct btrfs_trans_handle *trans,
3197 struct btrfs_root *log,
3198 struct btrfs_path *path,
3199 int key_type, u64 dirid,
3200 u64 first_offset, u64 last_offset)
3201{
3202 int ret;
3203 struct btrfs_key key;
3204 struct btrfs_dir_log_item *item;
3205
3206 key.objectid = dirid;
3207 key.offset = first_offset;
3208 if (key_type == BTRFS_DIR_ITEM_KEY)
3209 key.type = BTRFS_DIR_LOG_ITEM_KEY;
3210 else
3211 key.type = BTRFS_DIR_LOG_INDEX_KEY;
3212 ret = btrfs_insert_empty_item(trans, log, path, &key, sizeof(*item));
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003213 if (ret)
3214 return ret;
Chris Masone02119d2008-09-05 16:13:11 -04003215
3216 item = btrfs_item_ptr(path->nodes[0], path->slots[0],
3217 struct btrfs_dir_log_item);
3218 btrfs_set_dir_log_end(path->nodes[0], item, last_offset);
3219 btrfs_mark_buffer_dirty(path->nodes[0]);
David Sterbab3b4aa72011-04-21 01:20:15 +02003220 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -04003221 return 0;
3222}
3223
3224/*
3225 * log all the items included in the current transaction for a given
3226 * directory. This also creates the range items in the log tree required
3227 * to replay anything deleted before the fsync
3228 */
3229static noinline int log_dir_items(struct btrfs_trans_handle *trans,
3230 struct btrfs_root *root, struct inode *inode,
3231 struct btrfs_path *path,
3232 struct btrfs_path *dst_path, int key_type,
Filipe Manana2f2ff0e2015-03-20 17:19:46 +00003233 struct btrfs_log_ctx *ctx,
Chris Masone02119d2008-09-05 16:13:11 -04003234 u64 min_offset, u64 *last_offset_ret)
3235{
3236 struct btrfs_key min_key;
Chris Masone02119d2008-09-05 16:13:11 -04003237 struct btrfs_root *log = root->log_root;
3238 struct extent_buffer *src;
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003239 int err = 0;
Chris Masone02119d2008-09-05 16:13:11 -04003240 int ret;
3241 int i;
3242 int nritems;
3243 u64 first_offset = min_offset;
3244 u64 last_offset = (u64)-1;
Li Zefan33345d012011-04-20 10:31:50 +08003245 u64 ino = btrfs_ino(inode);
Chris Masone02119d2008-09-05 16:13:11 -04003246
3247 log = root->log_root;
Chris Masone02119d2008-09-05 16:13:11 -04003248
Li Zefan33345d012011-04-20 10:31:50 +08003249 min_key.objectid = ino;
Chris Masone02119d2008-09-05 16:13:11 -04003250 min_key.type = key_type;
3251 min_key.offset = min_offset;
3252
Filipe David Borba Manana6174d3c2013-10-01 16:13:42 +01003253 ret = btrfs_search_forward(root, &min_key, path, trans->transid);
Chris Masone02119d2008-09-05 16:13:11 -04003254
3255 /*
3256 * we didn't find anything from this transaction, see if there
3257 * is anything at all
3258 */
Li Zefan33345d012011-04-20 10:31:50 +08003259 if (ret != 0 || min_key.objectid != ino || min_key.type != key_type) {
3260 min_key.objectid = ino;
Chris Masone02119d2008-09-05 16:13:11 -04003261 min_key.type = key_type;
3262 min_key.offset = (u64)-1;
David Sterbab3b4aa72011-04-21 01:20:15 +02003263 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -04003264 ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0);
3265 if (ret < 0) {
David Sterbab3b4aa72011-04-21 01:20:15 +02003266 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -04003267 return ret;
3268 }
Li Zefan33345d012011-04-20 10:31:50 +08003269 ret = btrfs_previous_item(root, path, ino, key_type);
Chris Masone02119d2008-09-05 16:13:11 -04003270
3271 /* if ret == 0 there are items for this type,
3272 * create a range to tell us the last key of this type.
3273 * otherwise, there are no items in this directory after
3274 * *min_offset, and we create a range to indicate that.
3275 */
3276 if (ret == 0) {
3277 struct btrfs_key tmp;
3278 btrfs_item_key_to_cpu(path->nodes[0], &tmp,
3279 path->slots[0]);
Chris Masond3977122009-01-05 21:25:51 -05003280 if (key_type == tmp.type)
Chris Masone02119d2008-09-05 16:13:11 -04003281 first_offset = max(min_offset, tmp.offset) + 1;
Chris Masone02119d2008-09-05 16:13:11 -04003282 }
3283 goto done;
3284 }
3285
3286 /* go backward to find any previous key */
Li Zefan33345d012011-04-20 10:31:50 +08003287 ret = btrfs_previous_item(root, path, ino, key_type);
Chris Masone02119d2008-09-05 16:13:11 -04003288 if (ret == 0) {
3289 struct btrfs_key tmp;
3290 btrfs_item_key_to_cpu(path->nodes[0], &tmp, path->slots[0]);
3291 if (key_type == tmp.type) {
3292 first_offset = tmp.offset;
3293 ret = overwrite_item(trans, log, dst_path,
3294 path->nodes[0], path->slots[0],
3295 &tmp);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003296 if (ret) {
3297 err = ret;
3298 goto done;
3299 }
Chris Masone02119d2008-09-05 16:13:11 -04003300 }
3301 }
David Sterbab3b4aa72011-04-21 01:20:15 +02003302 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -04003303
3304 /* find the first key from this transaction again */
3305 ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0);
Dulshani Gunawardhanafae7f212013-10-31 10:30:08 +05303306 if (WARN_ON(ret != 0))
Chris Masone02119d2008-09-05 16:13:11 -04003307 goto done;
Chris Masone02119d2008-09-05 16:13:11 -04003308
3309 /*
3310 * we have a block from this transaction, log every item in it
3311 * from our directory
3312 */
Chris Masond3977122009-01-05 21:25:51 -05003313 while (1) {
Chris Masone02119d2008-09-05 16:13:11 -04003314 struct btrfs_key tmp;
3315 src = path->nodes[0];
3316 nritems = btrfs_header_nritems(src);
3317 for (i = path->slots[0]; i < nritems; i++) {
Filipe Manana2f2ff0e2015-03-20 17:19:46 +00003318 struct btrfs_dir_item *di;
3319
Chris Masone02119d2008-09-05 16:13:11 -04003320 btrfs_item_key_to_cpu(src, &min_key, i);
3321
Li Zefan33345d012011-04-20 10:31:50 +08003322 if (min_key.objectid != ino || min_key.type != key_type)
Chris Masone02119d2008-09-05 16:13:11 -04003323 goto done;
3324 ret = overwrite_item(trans, log, dst_path, src, i,
3325 &min_key);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003326 if (ret) {
3327 err = ret;
3328 goto done;
3329 }
Filipe Manana2f2ff0e2015-03-20 17:19:46 +00003330
3331 /*
3332 * We must make sure that when we log a directory entry,
3333 * the corresponding inode, after log replay, has a
3334 * matching link count. For example:
3335 *
3336 * touch foo
3337 * mkdir mydir
3338 * sync
3339 * ln foo mydir/bar
3340 * xfs_io -c "fsync" mydir
3341 * <crash>
3342 * <mount fs and log replay>
3343 *
3344 * Would result in a fsync log that when replayed, our
3345 * file inode would have a link count of 1, but we get
3346 * two directory entries pointing to the same inode.
3347 * After removing one of the names, it would not be
3348 * possible to remove the other name, which resulted
3349 * always in stale file handle errors, and would not
3350 * be possible to rmdir the parent directory, since
3351 * its i_size could never decrement to the value
3352 * BTRFS_EMPTY_DIR_SIZE, resulting in -ENOTEMPTY errors.
3353 */
3354 di = btrfs_item_ptr(src, i, struct btrfs_dir_item);
3355 btrfs_dir_item_key_to_cpu(src, di, &tmp);
3356 if (ctx &&
3357 (btrfs_dir_transid(src, di) == trans->transid ||
3358 btrfs_dir_type(src, di) == BTRFS_FT_DIR) &&
3359 tmp.type != BTRFS_ROOT_ITEM_KEY)
3360 ctx->log_new_dentries = true;
Chris Masone02119d2008-09-05 16:13:11 -04003361 }
3362 path->slots[0] = nritems;
3363
3364 /*
3365 * look ahead to the next item and see if it is also
3366 * from this directory and from this transaction
3367 */
3368 ret = btrfs_next_leaf(root, path);
3369 if (ret == 1) {
3370 last_offset = (u64)-1;
3371 goto done;
3372 }
3373 btrfs_item_key_to_cpu(path->nodes[0], &tmp, path->slots[0]);
Li Zefan33345d012011-04-20 10:31:50 +08003374 if (tmp.objectid != ino || tmp.type != key_type) {
Chris Masone02119d2008-09-05 16:13:11 -04003375 last_offset = (u64)-1;
3376 goto done;
3377 }
3378 if (btrfs_header_generation(path->nodes[0]) != trans->transid) {
3379 ret = overwrite_item(trans, log, dst_path,
3380 path->nodes[0], path->slots[0],
3381 &tmp);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003382 if (ret)
3383 err = ret;
3384 else
3385 last_offset = tmp.offset;
Chris Masone02119d2008-09-05 16:13:11 -04003386 goto done;
3387 }
3388 }
3389done:
David Sterbab3b4aa72011-04-21 01:20:15 +02003390 btrfs_release_path(path);
3391 btrfs_release_path(dst_path);
Chris Masone02119d2008-09-05 16:13:11 -04003392
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003393 if (err == 0) {
3394 *last_offset_ret = last_offset;
3395 /*
3396 * insert the log range keys to indicate where the log
3397 * is valid
3398 */
3399 ret = insert_dir_log_key(trans, log, path, key_type,
Li Zefan33345d012011-04-20 10:31:50 +08003400 ino, first_offset, last_offset);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003401 if (ret)
3402 err = ret;
3403 }
3404 return err;
Chris Masone02119d2008-09-05 16:13:11 -04003405}
3406
3407/*
3408 * logging directories is very similar to logging inodes, We find all the items
3409 * from the current transaction and write them to the log.
3410 *
3411 * The recovery code scans the directory in the subvolume, and if it finds a
3412 * key in the range logged that is not present in the log tree, then it means
3413 * that dir entry was unlinked during the transaction.
3414 *
3415 * In order for that scan to work, we must include one key smaller than
3416 * the smallest logged by this transaction and one key larger than the largest
3417 * key logged by this transaction.
3418 */
3419static noinline int log_directory_changes(struct btrfs_trans_handle *trans,
3420 struct btrfs_root *root, struct inode *inode,
3421 struct btrfs_path *path,
Filipe Manana2f2ff0e2015-03-20 17:19:46 +00003422 struct btrfs_path *dst_path,
3423 struct btrfs_log_ctx *ctx)
Chris Masone02119d2008-09-05 16:13:11 -04003424{
3425 u64 min_key;
3426 u64 max_key;
3427 int ret;
3428 int key_type = BTRFS_DIR_ITEM_KEY;
3429
3430again:
3431 min_key = 0;
3432 max_key = 0;
Chris Masond3977122009-01-05 21:25:51 -05003433 while (1) {
Chris Masone02119d2008-09-05 16:13:11 -04003434 ret = log_dir_items(trans, root, inode, path,
Filipe Manana2f2ff0e2015-03-20 17:19:46 +00003435 dst_path, key_type, ctx, min_key,
Chris Masone02119d2008-09-05 16:13:11 -04003436 &max_key);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003437 if (ret)
3438 return ret;
Chris Masone02119d2008-09-05 16:13:11 -04003439 if (max_key == (u64)-1)
3440 break;
3441 min_key = max_key + 1;
3442 }
3443
3444 if (key_type == BTRFS_DIR_ITEM_KEY) {
3445 key_type = BTRFS_DIR_INDEX_KEY;
3446 goto again;
3447 }
3448 return 0;
3449}
3450
3451/*
3452 * a helper function to drop items from the log before we relog an
3453 * inode. max_key_type indicates the highest item type to remove.
3454 * This cannot be run for file data extents because it does not
3455 * free the extents they point to.
3456 */
3457static int drop_objectid_items(struct btrfs_trans_handle *trans,
3458 struct btrfs_root *log,
3459 struct btrfs_path *path,
3460 u64 objectid, int max_key_type)
3461{
3462 int ret;
3463 struct btrfs_key key;
3464 struct btrfs_key found_key;
Josef Bacik18ec90d2012-09-28 11:56:28 -04003465 int start_slot;
Chris Masone02119d2008-09-05 16:13:11 -04003466
3467 key.objectid = objectid;
3468 key.type = max_key_type;
3469 key.offset = (u64)-1;
3470
Chris Masond3977122009-01-05 21:25:51 -05003471 while (1) {
Chris Masone02119d2008-09-05 16:13:11 -04003472 ret = btrfs_search_slot(trans, log, &key, path, -1, 1);
Josef Bacik36508602013-04-25 16:23:32 -04003473 BUG_ON(ret == 0); /* Logic error */
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003474 if (ret < 0)
Chris Masone02119d2008-09-05 16:13:11 -04003475 break;
3476
3477 if (path->slots[0] == 0)
3478 break;
3479
3480 path->slots[0]--;
3481 btrfs_item_key_to_cpu(path->nodes[0], &found_key,
3482 path->slots[0]);
3483
3484 if (found_key.objectid != objectid)
3485 break;
3486
Josef Bacik18ec90d2012-09-28 11:56:28 -04003487 found_key.offset = 0;
3488 found_key.type = 0;
3489 ret = btrfs_bin_search(path->nodes[0], &found_key, 0,
3490 &start_slot);
3491
3492 ret = btrfs_del_items(trans, log, path, start_slot,
3493 path->slots[0] - start_slot + 1);
3494 /*
3495 * If start slot isn't 0 then we don't need to re-search, we've
3496 * found the last guy with the objectid in this tree.
3497 */
3498 if (ret || start_slot != 0)
Tsutomu Itoh65a246c2011-05-19 04:37:44 +00003499 break;
David Sterbab3b4aa72011-04-21 01:20:15 +02003500 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -04003501 }
David Sterbab3b4aa72011-04-21 01:20:15 +02003502 btrfs_release_path(path);
Josef Bacik5bdbeb22012-05-29 16:59:49 -04003503 if (ret > 0)
3504 ret = 0;
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003505 return ret;
Chris Masone02119d2008-09-05 16:13:11 -04003506}
3507
Josef Bacik94edf4a2012-09-25 14:56:25 -04003508static void fill_inode_item(struct btrfs_trans_handle *trans,
3509 struct extent_buffer *leaf,
3510 struct btrfs_inode_item *item,
Filipe Manana1a4bcf42015-02-13 12:30:56 +00003511 struct inode *inode, int log_inode_only,
3512 u64 logged_isize)
Josef Bacik94edf4a2012-09-25 14:56:25 -04003513{
Josef Bacik0b1c6cc2012-10-23 16:03:44 -04003514 struct btrfs_map_token token;
Josef Bacik94edf4a2012-09-25 14:56:25 -04003515
Josef Bacik0b1c6cc2012-10-23 16:03:44 -04003516 btrfs_init_map_token(&token);
Josef Bacik94edf4a2012-09-25 14:56:25 -04003517
3518 if (log_inode_only) {
3519 /* set the generation to zero so the recover code
3520 * can tell the difference between an logging
3521 * just to say 'this inode exists' and a logging
3522 * to say 'update this inode with these values'
3523 */
Josef Bacik0b1c6cc2012-10-23 16:03:44 -04003524 btrfs_set_token_inode_generation(leaf, item, 0, &token);
Filipe Manana1a4bcf42015-02-13 12:30:56 +00003525 btrfs_set_token_inode_size(leaf, item, logged_isize, &token);
Josef Bacik94edf4a2012-09-25 14:56:25 -04003526 } else {
Josef Bacik0b1c6cc2012-10-23 16:03:44 -04003527 btrfs_set_token_inode_generation(leaf, item,
3528 BTRFS_I(inode)->generation,
3529 &token);
3530 btrfs_set_token_inode_size(leaf, item, inode->i_size, &token);
Josef Bacik94edf4a2012-09-25 14:56:25 -04003531 }
3532
Josef Bacik0b1c6cc2012-10-23 16:03:44 -04003533 btrfs_set_token_inode_uid(leaf, item, i_uid_read(inode), &token);
3534 btrfs_set_token_inode_gid(leaf, item, i_gid_read(inode), &token);
3535 btrfs_set_token_inode_mode(leaf, item, inode->i_mode, &token);
3536 btrfs_set_token_inode_nlink(leaf, item, inode->i_nlink, &token);
3537
David Sterbaa937b972014-12-12 17:39:12 +01003538 btrfs_set_token_timespec_sec(leaf, &item->atime,
Josef Bacik0b1c6cc2012-10-23 16:03:44 -04003539 inode->i_atime.tv_sec, &token);
David Sterbaa937b972014-12-12 17:39:12 +01003540 btrfs_set_token_timespec_nsec(leaf, &item->atime,
Josef Bacik0b1c6cc2012-10-23 16:03:44 -04003541 inode->i_atime.tv_nsec, &token);
3542
David Sterbaa937b972014-12-12 17:39:12 +01003543 btrfs_set_token_timespec_sec(leaf, &item->mtime,
Josef Bacik0b1c6cc2012-10-23 16:03:44 -04003544 inode->i_mtime.tv_sec, &token);
David Sterbaa937b972014-12-12 17:39:12 +01003545 btrfs_set_token_timespec_nsec(leaf, &item->mtime,
Josef Bacik0b1c6cc2012-10-23 16:03:44 -04003546 inode->i_mtime.tv_nsec, &token);
3547
David Sterbaa937b972014-12-12 17:39:12 +01003548 btrfs_set_token_timespec_sec(leaf, &item->ctime,
Josef Bacik0b1c6cc2012-10-23 16:03:44 -04003549 inode->i_ctime.tv_sec, &token);
David Sterbaa937b972014-12-12 17:39:12 +01003550 btrfs_set_token_timespec_nsec(leaf, &item->ctime,
Josef Bacik0b1c6cc2012-10-23 16:03:44 -04003551 inode->i_ctime.tv_nsec, &token);
3552
3553 btrfs_set_token_inode_nbytes(leaf, item, inode_get_bytes(inode),
3554 &token);
3555
3556 btrfs_set_token_inode_sequence(leaf, item, inode->i_version, &token);
3557 btrfs_set_token_inode_transid(leaf, item, trans->transid, &token);
3558 btrfs_set_token_inode_rdev(leaf, item, inode->i_rdev, &token);
3559 btrfs_set_token_inode_flags(leaf, item, BTRFS_I(inode)->flags, &token);
3560 btrfs_set_token_inode_block_group(leaf, item, 0, &token);
Josef Bacik94edf4a2012-09-25 14:56:25 -04003561}
3562
Josef Bacika95249b2012-10-11 16:17:34 -04003563static int log_inode_item(struct btrfs_trans_handle *trans,
3564 struct btrfs_root *log, struct btrfs_path *path,
3565 struct inode *inode)
3566{
3567 struct btrfs_inode_item *inode_item;
Josef Bacika95249b2012-10-11 16:17:34 -04003568 int ret;
3569
Filipe David Borba Mananaefd0c402013-10-07 21:20:44 +01003570 ret = btrfs_insert_empty_item(trans, log, path,
3571 &BTRFS_I(inode)->location,
Josef Bacika95249b2012-10-11 16:17:34 -04003572 sizeof(*inode_item));
3573 if (ret && ret != -EEXIST)
3574 return ret;
3575 inode_item = btrfs_item_ptr(path->nodes[0], path->slots[0],
3576 struct btrfs_inode_item);
Filipe Manana1a4bcf42015-02-13 12:30:56 +00003577 fill_inode_item(trans, path->nodes[0], inode_item, inode, 0, 0);
Josef Bacika95249b2012-10-11 16:17:34 -04003578 btrfs_release_path(path);
3579 return 0;
3580}
3581
Chris Mason31ff1cd2008-09-11 16:17:57 -04003582static noinline int copy_items(struct btrfs_trans_handle *trans,
Liu Bod2794402012-08-29 01:07:56 -06003583 struct inode *inode,
Chris Mason31ff1cd2008-09-11 16:17:57 -04003584 struct btrfs_path *dst_path,
Josef Bacik16e75492013-10-22 12:18:51 -04003585 struct btrfs_path *src_path, u64 *last_extent,
Filipe Manana1a4bcf42015-02-13 12:30:56 +00003586 int start_slot, int nr, int inode_only,
3587 u64 logged_isize)
Chris Mason31ff1cd2008-09-11 16:17:57 -04003588{
3589 unsigned long src_offset;
3590 unsigned long dst_offset;
Liu Bod2794402012-08-29 01:07:56 -06003591 struct btrfs_root *log = BTRFS_I(inode)->root->log_root;
Chris Mason31ff1cd2008-09-11 16:17:57 -04003592 struct btrfs_file_extent_item *extent;
3593 struct btrfs_inode_item *inode_item;
Josef Bacik16e75492013-10-22 12:18:51 -04003594 struct extent_buffer *src = src_path->nodes[0];
3595 struct btrfs_key first_key, last_key, key;
Chris Mason31ff1cd2008-09-11 16:17:57 -04003596 int ret;
3597 struct btrfs_key *ins_keys;
3598 u32 *ins_sizes;
3599 char *ins_data;
3600 int i;
Chris Masond20f7042008-12-08 16:58:54 -05003601 struct list_head ordered_sums;
Liu Bod2794402012-08-29 01:07:56 -06003602 int skip_csum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;
Josef Bacik16e75492013-10-22 12:18:51 -04003603 bool has_extents = false;
Filipe Manana74121f72014-08-07 12:00:44 +01003604 bool need_find_last_extent = true;
Josef Bacik16e75492013-10-22 12:18:51 -04003605 bool done = false;
Chris Masond20f7042008-12-08 16:58:54 -05003606
3607 INIT_LIST_HEAD(&ordered_sums);
Chris Mason31ff1cd2008-09-11 16:17:57 -04003608
3609 ins_data = kmalloc(nr * sizeof(struct btrfs_key) +
3610 nr * sizeof(u32), GFP_NOFS);
liubo2a29edc2011-01-26 06:22:08 +00003611 if (!ins_data)
3612 return -ENOMEM;
3613
Josef Bacik16e75492013-10-22 12:18:51 -04003614 first_key.objectid = (u64)-1;
3615
Chris Mason31ff1cd2008-09-11 16:17:57 -04003616 ins_sizes = (u32 *)ins_data;
3617 ins_keys = (struct btrfs_key *)(ins_data + nr * sizeof(u32));
3618
3619 for (i = 0; i < nr; i++) {
3620 ins_sizes[i] = btrfs_item_size_nr(src, i + start_slot);
3621 btrfs_item_key_to_cpu(src, ins_keys + i, i + start_slot);
3622 }
3623 ret = btrfs_insert_empty_items(trans, log, dst_path,
3624 ins_keys, ins_sizes, nr);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003625 if (ret) {
3626 kfree(ins_data);
3627 return ret;
3628 }
Chris Mason31ff1cd2008-09-11 16:17:57 -04003629
Yan Zheng5d4f98a2009-06-10 10:45:14 -04003630 for (i = 0; i < nr; i++, dst_path->slots[0]++) {
Chris Mason31ff1cd2008-09-11 16:17:57 -04003631 dst_offset = btrfs_item_ptr_offset(dst_path->nodes[0],
3632 dst_path->slots[0]);
3633
3634 src_offset = btrfs_item_ptr_offset(src, start_slot + i);
3635
Josef Bacik16e75492013-10-22 12:18:51 -04003636 if ((i == (nr - 1)))
3637 last_key = ins_keys[i];
3638
Josef Bacik94edf4a2012-09-25 14:56:25 -04003639 if (ins_keys[i].type == BTRFS_INODE_ITEM_KEY) {
Chris Mason31ff1cd2008-09-11 16:17:57 -04003640 inode_item = btrfs_item_ptr(dst_path->nodes[0],
3641 dst_path->slots[0],
3642 struct btrfs_inode_item);
Josef Bacik94edf4a2012-09-25 14:56:25 -04003643 fill_inode_item(trans, dst_path->nodes[0], inode_item,
Filipe Manana1a4bcf42015-02-13 12:30:56 +00003644 inode, inode_only == LOG_INODE_EXISTS,
3645 logged_isize);
Josef Bacik94edf4a2012-09-25 14:56:25 -04003646 } else {
3647 copy_extent_buffer(dst_path->nodes[0], src, dst_offset,
3648 src_offset, ins_sizes[i]);
Chris Mason31ff1cd2008-09-11 16:17:57 -04003649 }
Josef Bacik94edf4a2012-09-25 14:56:25 -04003650
Josef Bacik16e75492013-10-22 12:18:51 -04003651 /*
3652 * We set need_find_last_extent here in case we know we were
3653 * processing other items and then walk into the first extent in
3654 * the inode. If we don't hit an extent then nothing changes,
3655 * we'll do the last search the next time around.
3656 */
3657 if (ins_keys[i].type == BTRFS_EXTENT_DATA_KEY) {
3658 has_extents = true;
Filipe Manana74121f72014-08-07 12:00:44 +01003659 if (first_key.objectid == (u64)-1)
Josef Bacik16e75492013-10-22 12:18:51 -04003660 first_key = ins_keys[i];
3661 } else {
3662 need_find_last_extent = false;
3663 }
3664
Chris Mason31ff1cd2008-09-11 16:17:57 -04003665 /* take a reference on file data extents so that truncates
3666 * or deletes of this inode don't have to relog the inode
3667 * again
3668 */
David Sterba962a2982014-06-04 18:41:45 +02003669 if (ins_keys[i].type == BTRFS_EXTENT_DATA_KEY &&
Liu Bod2794402012-08-29 01:07:56 -06003670 !skip_csum) {
Chris Mason31ff1cd2008-09-11 16:17:57 -04003671 int found_type;
3672 extent = btrfs_item_ptr(src, start_slot + i,
3673 struct btrfs_file_extent_item);
3674
liubo8e531cd2011-05-06 10:36:09 +08003675 if (btrfs_file_extent_generation(src, extent) < trans->transid)
3676 continue;
3677
Chris Mason31ff1cd2008-09-11 16:17:57 -04003678 found_type = btrfs_file_extent_type(src, extent);
Josef Bacik6f1fed72012-09-26 11:07:06 -04003679 if (found_type == BTRFS_FILE_EXTENT_REG) {
Yan Zheng5d4f98a2009-06-10 10:45:14 -04003680 u64 ds, dl, cs, cl;
3681 ds = btrfs_file_extent_disk_bytenr(src,
3682 extent);
3683 /* ds == 0 is a hole */
3684 if (ds == 0)
3685 continue;
3686
3687 dl = btrfs_file_extent_disk_num_bytes(src,
3688 extent);
3689 cs = btrfs_file_extent_offset(src, extent);
3690 cl = btrfs_file_extent_num_bytes(src,
Joe Perchesa419aef2009-08-18 11:18:35 -07003691 extent);
Chris Mason580afd72008-12-08 19:15:39 -05003692 if (btrfs_file_extent_compression(src,
3693 extent)) {
3694 cs = 0;
3695 cl = dl;
3696 }
Yan Zheng5d4f98a2009-06-10 10:45:14 -04003697
3698 ret = btrfs_lookup_csums_range(
3699 log->fs_info->csum_root,
3700 ds + cs, ds + cs + cl - 1,
Arne Jansena2de7332011-03-08 14:14:00 +01003701 &ordered_sums, 0);
Josef Bacik36508602013-04-25 16:23:32 -04003702 if (ret) {
3703 btrfs_release_path(dst_path);
3704 kfree(ins_data);
3705 return ret;
3706 }
Chris Mason31ff1cd2008-09-11 16:17:57 -04003707 }
3708 }
Chris Mason31ff1cd2008-09-11 16:17:57 -04003709 }
3710
3711 btrfs_mark_buffer_dirty(dst_path->nodes[0]);
David Sterbab3b4aa72011-04-21 01:20:15 +02003712 btrfs_release_path(dst_path);
Chris Mason31ff1cd2008-09-11 16:17:57 -04003713 kfree(ins_data);
Chris Masond20f7042008-12-08 16:58:54 -05003714
3715 /*
3716 * we have to do this after the loop above to avoid changing the
3717 * log tree while trying to change the log tree.
3718 */
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003719 ret = 0;
Chris Masond3977122009-01-05 21:25:51 -05003720 while (!list_empty(&ordered_sums)) {
Chris Masond20f7042008-12-08 16:58:54 -05003721 struct btrfs_ordered_sum *sums = list_entry(ordered_sums.next,
3722 struct btrfs_ordered_sum,
3723 list);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003724 if (!ret)
3725 ret = btrfs_csum_file_blocks(trans, log, sums);
Chris Masond20f7042008-12-08 16:58:54 -05003726 list_del(&sums->list);
3727 kfree(sums);
3728 }
Josef Bacik16e75492013-10-22 12:18:51 -04003729
3730 if (!has_extents)
3731 return ret;
3732
Filipe Manana74121f72014-08-07 12:00:44 +01003733 if (need_find_last_extent && *last_extent == first_key.offset) {
3734 /*
3735 * We don't have any leafs between our current one and the one
3736 * we processed before that can have file extent items for our
3737 * inode (and have a generation number smaller than our current
3738 * transaction id).
3739 */
3740 need_find_last_extent = false;
3741 }
3742
Josef Bacik16e75492013-10-22 12:18:51 -04003743 /*
3744 * Because we use btrfs_search_forward we could skip leaves that were
3745 * not modified and then assume *last_extent is valid when it really
3746 * isn't. So back up to the previous leaf and read the end of the last
3747 * extent before we go and fill in holes.
3748 */
3749 if (need_find_last_extent) {
3750 u64 len;
3751
3752 ret = btrfs_prev_leaf(BTRFS_I(inode)->root, src_path);
3753 if (ret < 0)
3754 return ret;
3755 if (ret)
3756 goto fill_holes;
3757 if (src_path->slots[0])
3758 src_path->slots[0]--;
3759 src = src_path->nodes[0];
3760 btrfs_item_key_to_cpu(src, &key, src_path->slots[0]);
3761 if (key.objectid != btrfs_ino(inode) ||
3762 key.type != BTRFS_EXTENT_DATA_KEY)
3763 goto fill_holes;
3764 extent = btrfs_item_ptr(src, src_path->slots[0],
3765 struct btrfs_file_extent_item);
3766 if (btrfs_file_extent_type(src, extent) ==
3767 BTRFS_FILE_EXTENT_INLINE) {
Chris Mason514ac8a2014-01-03 21:07:00 -08003768 len = btrfs_file_extent_inline_len(src,
3769 src_path->slots[0],
3770 extent);
Josef Bacik16e75492013-10-22 12:18:51 -04003771 *last_extent = ALIGN(key.offset + len,
3772 log->sectorsize);
3773 } else {
3774 len = btrfs_file_extent_num_bytes(src, extent);
3775 *last_extent = key.offset + len;
3776 }
3777 }
3778fill_holes:
3779 /* So we did prev_leaf, now we need to move to the next leaf, but a few
3780 * things could have happened
3781 *
3782 * 1) A merge could have happened, so we could currently be on a leaf
3783 * that holds what we were copying in the first place.
3784 * 2) A split could have happened, and now not all of the items we want
3785 * are on the same leaf.
3786 *
3787 * So we need to adjust how we search for holes, we need to drop the
3788 * path and re-search for the first extent key we found, and then walk
3789 * forward until we hit the last one we copied.
3790 */
3791 if (need_find_last_extent) {
3792 /* btrfs_prev_leaf could return 1 without releasing the path */
3793 btrfs_release_path(src_path);
3794 ret = btrfs_search_slot(NULL, BTRFS_I(inode)->root, &first_key,
3795 src_path, 0, 0);
3796 if (ret < 0)
3797 return ret;
3798 ASSERT(ret == 0);
3799 src = src_path->nodes[0];
3800 i = src_path->slots[0];
3801 } else {
3802 i = start_slot;
3803 }
3804
3805 /*
3806 * Ok so here we need to go through and fill in any holes we may have
3807 * to make sure that holes are punched for those areas in case they had
3808 * extents previously.
3809 */
3810 while (!done) {
3811 u64 offset, len;
3812 u64 extent_end;
3813
3814 if (i >= btrfs_header_nritems(src_path->nodes[0])) {
3815 ret = btrfs_next_leaf(BTRFS_I(inode)->root, src_path);
3816 if (ret < 0)
3817 return ret;
3818 ASSERT(ret == 0);
3819 src = src_path->nodes[0];
3820 i = 0;
3821 }
3822
3823 btrfs_item_key_to_cpu(src, &key, i);
3824 if (!btrfs_comp_cpu_keys(&key, &last_key))
3825 done = true;
3826 if (key.objectid != btrfs_ino(inode) ||
3827 key.type != BTRFS_EXTENT_DATA_KEY) {
3828 i++;
3829 continue;
3830 }
3831 extent = btrfs_item_ptr(src, i, struct btrfs_file_extent_item);
3832 if (btrfs_file_extent_type(src, extent) ==
3833 BTRFS_FILE_EXTENT_INLINE) {
Chris Mason514ac8a2014-01-03 21:07:00 -08003834 len = btrfs_file_extent_inline_len(src, i, extent);
Josef Bacik16e75492013-10-22 12:18:51 -04003835 extent_end = ALIGN(key.offset + len, log->sectorsize);
3836 } else {
3837 len = btrfs_file_extent_num_bytes(src, extent);
3838 extent_end = key.offset + len;
3839 }
3840 i++;
3841
3842 if (*last_extent == key.offset) {
3843 *last_extent = extent_end;
3844 continue;
3845 }
3846 offset = *last_extent;
3847 len = key.offset - *last_extent;
3848 ret = btrfs_insert_file_extent(trans, log, btrfs_ino(inode),
3849 offset, 0, 0, len, 0, len, 0,
3850 0, 0);
3851 if (ret)
3852 break;
Filipe Manana74121f72014-08-07 12:00:44 +01003853 *last_extent = extent_end;
Josef Bacik16e75492013-10-22 12:18:51 -04003854 }
3855 /*
3856 * Need to let the callers know we dropped the path so they should
3857 * re-search.
3858 */
3859 if (!ret && need_find_last_extent)
3860 ret = 1;
Yan, Zheng4a500fd2010-05-16 10:49:59 -04003861 return ret;
Chris Mason31ff1cd2008-09-11 16:17:57 -04003862}
3863
Josef Bacik5dc562c2012-08-17 13:14:17 -04003864static int extent_cmp(void *priv, struct list_head *a, struct list_head *b)
3865{
3866 struct extent_map *em1, *em2;
3867
3868 em1 = list_entry(a, struct extent_map, list);
3869 em2 = list_entry(b, struct extent_map, list);
3870
3871 if (em1->start < em2->start)
3872 return -1;
3873 else if (em1->start > em2->start)
3874 return 1;
3875 return 0;
3876}
3877
Filipe Manana8407f552014-09-05 15:14:39 +01003878static int wait_ordered_extents(struct btrfs_trans_handle *trans,
3879 struct inode *inode,
3880 struct btrfs_root *root,
3881 const struct extent_map *em,
3882 const struct list_head *logged_list,
3883 bool *ordered_io_error)
Josef Bacik5dc562c2012-08-17 13:14:17 -04003884{
Josef Bacik2ab28f32012-10-12 15:27:49 -04003885 struct btrfs_ordered_extent *ordered;
Filipe Manana8407f552014-09-05 15:14:39 +01003886 struct btrfs_root *log = root->log_root;
Josef Bacik2ab28f32012-10-12 15:27:49 -04003887 u64 mod_start = em->mod_start;
3888 u64 mod_len = em->mod_len;
Filipe Manana8407f552014-09-05 15:14:39 +01003889 const bool skip_csum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;
Josef Bacik2ab28f32012-10-12 15:27:49 -04003890 u64 csum_offset;
3891 u64 csum_len;
Filipe Manana8407f552014-09-05 15:14:39 +01003892 LIST_HEAD(ordered_sums);
3893 int ret = 0;
Josef Bacik09a2a8f92013-04-05 16:51:15 -04003894
Filipe Manana8407f552014-09-05 15:14:39 +01003895 *ordered_io_error = false;
Josef Bacik70c8a912012-10-11 16:54:30 -04003896
Filipe Manana8407f552014-09-05 15:14:39 +01003897 if (test_bit(EXTENT_FLAG_PREALLOC, &em->flags) ||
3898 em->block_start == EXTENT_MAP_HOLE)
Josef Bacik70c8a912012-10-11 16:54:30 -04003899 return 0;
3900
Josef Bacik2ab28f32012-10-12 15:27:49 -04003901 /*
Filipe Manana8407f552014-09-05 15:14:39 +01003902 * Wait far any ordered extent that covers our extent map. If it
3903 * finishes without an error, first check and see if our csums are on
3904 * our outstanding ordered extents.
Josef Bacik2ab28f32012-10-12 15:27:49 -04003905 */
Miao Xie827463c2014-01-14 20:31:51 +08003906 list_for_each_entry(ordered, logged_list, log_list) {
Josef Bacik2ab28f32012-10-12 15:27:49 -04003907 struct btrfs_ordered_sum *sum;
3908
3909 if (!mod_len)
3910 break;
3911
Josef Bacik2ab28f32012-10-12 15:27:49 -04003912 if (ordered->file_offset + ordered->len <= mod_start ||
3913 mod_start + mod_len <= ordered->file_offset)
3914 continue;
3915
Filipe Manana8407f552014-09-05 15:14:39 +01003916 if (!test_bit(BTRFS_ORDERED_IO_DONE, &ordered->flags) &&
3917 !test_bit(BTRFS_ORDERED_IOERR, &ordered->flags) &&
3918 !test_bit(BTRFS_ORDERED_DIRECT, &ordered->flags)) {
3919 const u64 start = ordered->file_offset;
3920 const u64 end = ordered->file_offset + ordered->len - 1;
3921
3922 WARN_ON(ordered->inode != inode);
3923 filemap_fdatawrite_range(inode->i_mapping, start, end);
3924 }
3925
3926 wait_event(ordered->wait,
3927 (test_bit(BTRFS_ORDERED_IO_DONE, &ordered->flags) ||
3928 test_bit(BTRFS_ORDERED_IOERR, &ordered->flags)));
3929
3930 if (test_bit(BTRFS_ORDERED_IOERR, &ordered->flags)) {
Filipe Mananab38ef712014-11-13 17:01:45 +00003931 /*
3932 * Clear the AS_EIO/AS_ENOSPC flags from the inode's
3933 * i_mapping flags, so that the next fsync won't get
3934 * an outdated io error too.
3935 */
3936 btrfs_inode_check_errors(inode);
Filipe Manana8407f552014-09-05 15:14:39 +01003937 *ordered_io_error = true;
3938 break;
3939 }
Josef Bacik2ab28f32012-10-12 15:27:49 -04003940 /*
3941 * We are going to copy all the csums on this ordered extent, so
3942 * go ahead and adjust mod_start and mod_len in case this
3943 * ordered extent has already been logged.
3944 */
3945 if (ordered->file_offset > mod_start) {
3946 if (ordered->file_offset + ordered->len >=
3947 mod_start + mod_len)
3948 mod_len = ordered->file_offset - mod_start;
3949 /*
3950 * If we have this case
3951 *
3952 * |--------- logged extent ---------|
3953 * |----- ordered extent ----|
3954 *
3955 * Just don't mess with mod_start and mod_len, we'll
3956 * just end up logging more csums than we need and it
3957 * will be ok.
3958 */
3959 } else {
3960 if (ordered->file_offset + ordered->len <
3961 mod_start + mod_len) {
3962 mod_len = (mod_start + mod_len) -
3963 (ordered->file_offset + ordered->len);
3964 mod_start = ordered->file_offset +
3965 ordered->len;
3966 } else {
3967 mod_len = 0;
3968 }
3969 }
3970
Filipe Manana8407f552014-09-05 15:14:39 +01003971 if (skip_csum)
3972 continue;
3973
Josef Bacik2ab28f32012-10-12 15:27:49 -04003974 /*
3975 * To keep us from looping for the above case of an ordered
3976 * extent that falls inside of the logged extent.
3977 */
3978 if (test_and_set_bit(BTRFS_ORDERED_LOGGED_CSUM,
3979 &ordered->flags))
3980 continue;
Josef Bacik2ab28f32012-10-12 15:27:49 -04003981
Josef Bacik2ab28f32012-10-12 15:27:49 -04003982 list_for_each_entry(sum, &ordered->list, list) {
3983 ret = btrfs_csum_file_blocks(trans, log, sum);
Miao Xie827463c2014-01-14 20:31:51 +08003984 if (ret)
Filipe Manana8407f552014-09-05 15:14:39 +01003985 break;
Josef Bacik2ab28f32012-10-12 15:27:49 -04003986 }
Josef Bacik2ab28f32012-10-12 15:27:49 -04003987 }
Josef Bacik2ab28f32012-10-12 15:27:49 -04003988
Filipe Manana8407f552014-09-05 15:14:39 +01003989 if (*ordered_io_error || !mod_len || ret || skip_csum)
Josef Bacik2ab28f32012-10-12 15:27:49 -04003990 return ret;
3991
Filipe David Borba Manana488111a2013-10-28 16:30:29 +00003992 if (em->compress_type) {
3993 csum_offset = 0;
Filipe Manana8407f552014-09-05 15:14:39 +01003994 csum_len = max(em->block_len, em->orig_block_len);
Filipe David Borba Manana488111a2013-10-28 16:30:29 +00003995 } else {
3996 csum_offset = mod_start - em->start;
3997 csum_len = mod_len;
3998 }
Josef Bacik2ab28f32012-10-12 15:27:49 -04003999
Josef Bacik70c8a912012-10-11 16:54:30 -04004000 /* block start is already adjusted for the file extent offset. */
4001 ret = btrfs_lookup_csums_range(log->fs_info->csum_root,
4002 em->block_start + csum_offset,
4003 em->block_start + csum_offset +
4004 csum_len - 1, &ordered_sums, 0);
4005 if (ret)
4006 return ret;
4007
4008 while (!list_empty(&ordered_sums)) {
4009 struct btrfs_ordered_sum *sums = list_entry(ordered_sums.next,
4010 struct btrfs_ordered_sum,
4011 list);
4012 if (!ret)
4013 ret = btrfs_csum_file_blocks(trans, log, sums);
4014 list_del(&sums->list);
4015 kfree(sums);
4016 }
4017
4018 return ret;
Josef Bacik5dc562c2012-08-17 13:14:17 -04004019}
4020
Filipe Manana8407f552014-09-05 15:14:39 +01004021static int log_one_extent(struct btrfs_trans_handle *trans,
4022 struct inode *inode, struct btrfs_root *root,
4023 const struct extent_map *em,
4024 struct btrfs_path *path,
4025 const struct list_head *logged_list,
4026 struct btrfs_log_ctx *ctx)
4027{
4028 struct btrfs_root *log = root->log_root;
4029 struct btrfs_file_extent_item *fi;
4030 struct extent_buffer *leaf;
4031 struct btrfs_map_token token;
4032 struct btrfs_key key;
4033 u64 extent_offset = em->start - em->orig_start;
4034 u64 block_len;
4035 int ret;
4036 int extent_inserted = 0;
4037 bool ordered_io_err = false;
4038
4039 ret = wait_ordered_extents(trans, inode, root, em, logged_list,
4040 &ordered_io_err);
4041 if (ret)
4042 return ret;
4043
4044 if (ordered_io_err) {
4045 ctx->io_err = -EIO;
4046 return 0;
4047 }
4048
4049 btrfs_init_map_token(&token);
4050
4051 ret = __btrfs_drop_extents(trans, log, inode, path, em->start,
4052 em->start + em->len, NULL, 0, 1,
4053 sizeof(*fi), &extent_inserted);
4054 if (ret)
4055 return ret;
4056
4057 if (!extent_inserted) {
4058 key.objectid = btrfs_ino(inode);
4059 key.type = BTRFS_EXTENT_DATA_KEY;
4060 key.offset = em->start;
4061
4062 ret = btrfs_insert_empty_item(trans, log, path, &key,
4063 sizeof(*fi));
4064 if (ret)
4065 return ret;
4066 }
4067 leaf = path->nodes[0];
4068 fi = btrfs_item_ptr(leaf, path->slots[0],
4069 struct btrfs_file_extent_item);
4070
Josef Bacik50d9aa92014-11-21 14:52:38 -05004071 btrfs_set_token_file_extent_generation(leaf, fi, trans->transid,
Filipe Manana8407f552014-09-05 15:14:39 +01004072 &token);
4073 if (test_bit(EXTENT_FLAG_PREALLOC, &em->flags))
4074 btrfs_set_token_file_extent_type(leaf, fi,
4075 BTRFS_FILE_EXTENT_PREALLOC,
4076 &token);
4077 else
4078 btrfs_set_token_file_extent_type(leaf, fi,
4079 BTRFS_FILE_EXTENT_REG,
4080 &token);
4081
4082 block_len = max(em->block_len, em->orig_block_len);
4083 if (em->compress_type != BTRFS_COMPRESS_NONE) {
4084 btrfs_set_token_file_extent_disk_bytenr(leaf, fi,
4085 em->block_start,
4086 &token);
4087 btrfs_set_token_file_extent_disk_num_bytes(leaf, fi, block_len,
4088 &token);
4089 } else if (em->block_start < EXTENT_MAP_LAST_BYTE) {
4090 btrfs_set_token_file_extent_disk_bytenr(leaf, fi,
4091 em->block_start -
4092 extent_offset, &token);
4093 btrfs_set_token_file_extent_disk_num_bytes(leaf, fi, block_len,
4094 &token);
4095 } else {
4096 btrfs_set_token_file_extent_disk_bytenr(leaf, fi, 0, &token);
4097 btrfs_set_token_file_extent_disk_num_bytes(leaf, fi, 0,
4098 &token);
4099 }
4100
4101 btrfs_set_token_file_extent_offset(leaf, fi, extent_offset, &token);
4102 btrfs_set_token_file_extent_num_bytes(leaf, fi, em->len, &token);
4103 btrfs_set_token_file_extent_ram_bytes(leaf, fi, em->ram_bytes, &token);
4104 btrfs_set_token_file_extent_compression(leaf, fi, em->compress_type,
4105 &token);
4106 btrfs_set_token_file_extent_encryption(leaf, fi, 0, &token);
4107 btrfs_set_token_file_extent_other_encoding(leaf, fi, 0, &token);
4108 btrfs_mark_buffer_dirty(leaf);
4109
4110 btrfs_release_path(path);
4111
4112 return ret;
4113}
4114
Josef Bacik5dc562c2012-08-17 13:14:17 -04004115static int btrfs_log_changed_extents(struct btrfs_trans_handle *trans,
4116 struct btrfs_root *root,
4117 struct inode *inode,
Miao Xie827463c2014-01-14 20:31:51 +08004118 struct btrfs_path *path,
Filipe Manana8407f552014-09-05 15:14:39 +01004119 struct list_head *logged_list,
4120 struct btrfs_log_ctx *ctx)
Josef Bacik5dc562c2012-08-17 13:14:17 -04004121{
Josef Bacik5dc562c2012-08-17 13:14:17 -04004122 struct extent_map *em, *n;
4123 struct list_head extents;
4124 struct extent_map_tree *tree = &BTRFS_I(inode)->extent_tree;
4125 u64 test_gen;
4126 int ret = 0;
Josef Bacik2ab28f32012-10-12 15:27:49 -04004127 int num = 0;
Josef Bacik5dc562c2012-08-17 13:14:17 -04004128
4129 INIT_LIST_HEAD(&extents);
4130
Josef Bacik5dc562c2012-08-17 13:14:17 -04004131 write_lock(&tree->lock);
4132 test_gen = root->fs_info->last_trans_committed;
4133
4134 list_for_each_entry_safe(em, n, &tree->modified_extents, list) {
4135 list_del_init(&em->list);
Josef Bacik2ab28f32012-10-12 15:27:49 -04004136
4137 /*
4138 * Just an arbitrary number, this can be really CPU intensive
4139 * once we start getting a lot of extents, and really once we
4140 * have a bunch of extents we just want to commit since it will
4141 * be faster.
4142 */
4143 if (++num > 32768) {
4144 list_del_init(&tree->modified_extents);
4145 ret = -EFBIG;
4146 goto process;
4147 }
4148
Josef Bacik5dc562c2012-08-17 13:14:17 -04004149 if (em->generation <= test_gen)
4150 continue;
Josef Bacikff44c6e2012-09-14 12:59:20 -04004151 /* Need a ref to keep it from getting evicted from cache */
4152 atomic_inc(&em->refs);
4153 set_bit(EXTENT_FLAG_LOGGING, &em->flags);
Josef Bacik5dc562c2012-08-17 13:14:17 -04004154 list_add_tail(&em->list, &extents);
Josef Bacik2ab28f32012-10-12 15:27:49 -04004155 num++;
Josef Bacik5dc562c2012-08-17 13:14:17 -04004156 }
4157
4158 list_sort(NULL, &extents, extent_cmp);
4159
Josef Bacik2ab28f32012-10-12 15:27:49 -04004160process:
Josef Bacik5dc562c2012-08-17 13:14:17 -04004161 while (!list_empty(&extents)) {
4162 em = list_entry(extents.next, struct extent_map, list);
4163
4164 list_del_init(&em->list);
4165
4166 /*
4167 * If we had an error we just need to delete everybody from our
4168 * private list.
4169 */
Josef Bacikff44c6e2012-09-14 12:59:20 -04004170 if (ret) {
Josef Bacik201a9032013-01-24 12:02:07 -05004171 clear_em_logging(tree, em);
Josef Bacikff44c6e2012-09-14 12:59:20 -04004172 free_extent_map(em);
Josef Bacik5dc562c2012-08-17 13:14:17 -04004173 continue;
Josef Bacikff44c6e2012-09-14 12:59:20 -04004174 }
4175
4176 write_unlock(&tree->lock);
Josef Bacik5dc562c2012-08-17 13:14:17 -04004177
Filipe Manana8407f552014-09-05 15:14:39 +01004178 ret = log_one_extent(trans, inode, root, em, path, logged_list,
4179 ctx);
Josef Bacikff44c6e2012-09-14 12:59:20 -04004180 write_lock(&tree->lock);
Josef Bacik201a9032013-01-24 12:02:07 -05004181 clear_em_logging(tree, em);
4182 free_extent_map(em);
Josef Bacik5dc562c2012-08-17 13:14:17 -04004183 }
Josef Bacikff44c6e2012-09-14 12:59:20 -04004184 WARN_ON(!list_empty(&extents));
4185 write_unlock(&tree->lock);
Josef Bacik5dc562c2012-08-17 13:14:17 -04004186
Josef Bacik5dc562c2012-08-17 13:14:17 -04004187 btrfs_release_path(path);
Josef Bacik5dc562c2012-08-17 13:14:17 -04004188 return ret;
4189}
4190
Filipe Manana1a4bcf42015-02-13 12:30:56 +00004191static int logged_inode_size(struct btrfs_root *log, struct inode *inode,
4192 struct btrfs_path *path, u64 *size_ret)
4193{
4194 struct btrfs_key key;
4195 int ret;
4196
4197 key.objectid = btrfs_ino(inode);
4198 key.type = BTRFS_INODE_ITEM_KEY;
4199 key.offset = 0;
4200
4201 ret = btrfs_search_slot(NULL, log, &key, path, 0, 0);
4202 if (ret < 0) {
4203 return ret;
4204 } else if (ret > 0) {
Filipe Manana2f2ff0e2015-03-20 17:19:46 +00004205 *size_ret = 0;
Filipe Manana1a4bcf42015-02-13 12:30:56 +00004206 } else {
4207 struct btrfs_inode_item *item;
4208
4209 item = btrfs_item_ptr(path->nodes[0], path->slots[0],
4210 struct btrfs_inode_item);
4211 *size_ret = btrfs_inode_size(path->nodes[0], item);
4212 }
4213
4214 btrfs_release_path(path);
4215 return 0;
4216}
4217
Filipe Manana36283bf2015-06-20 00:44:51 +01004218/*
4219 * At the moment we always log all xattrs. This is to figure out at log replay
4220 * time which xattrs must have their deletion replayed. If a xattr is missing
4221 * in the log tree and exists in the fs/subvol tree, we delete it. This is
4222 * because if a xattr is deleted, the inode is fsynced and a power failure
4223 * happens, causing the log to be replayed the next time the fs is mounted,
4224 * we want the xattr to not exist anymore (same behaviour as other filesystems
4225 * with a journal, ext3/4, xfs, f2fs, etc).
4226 */
4227static int btrfs_log_all_xattrs(struct btrfs_trans_handle *trans,
4228 struct btrfs_root *root,
4229 struct inode *inode,
4230 struct btrfs_path *path,
4231 struct btrfs_path *dst_path)
4232{
4233 int ret;
4234 struct btrfs_key key;
4235 const u64 ino = btrfs_ino(inode);
4236 int ins_nr = 0;
4237 int start_slot = 0;
4238
4239 key.objectid = ino;
4240 key.type = BTRFS_XATTR_ITEM_KEY;
4241 key.offset = 0;
4242
4243 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
4244 if (ret < 0)
4245 return ret;
4246
4247 while (true) {
4248 int slot = path->slots[0];
4249 struct extent_buffer *leaf = path->nodes[0];
4250 int nritems = btrfs_header_nritems(leaf);
4251
4252 if (slot >= nritems) {
4253 if (ins_nr > 0) {
4254 u64 last_extent = 0;
4255
4256 ret = copy_items(trans, inode, dst_path, path,
4257 &last_extent, start_slot,
4258 ins_nr, 1, 0);
4259 /* can't be 1, extent items aren't processed */
4260 ASSERT(ret <= 0);
4261 if (ret < 0)
4262 return ret;
4263 ins_nr = 0;
4264 }
4265 ret = btrfs_next_leaf(root, path);
4266 if (ret < 0)
4267 return ret;
4268 else if (ret > 0)
4269 break;
4270 continue;
4271 }
4272
4273 btrfs_item_key_to_cpu(leaf, &key, slot);
4274 if (key.objectid != ino || key.type != BTRFS_XATTR_ITEM_KEY)
4275 break;
4276
4277 if (ins_nr == 0)
4278 start_slot = slot;
4279 ins_nr++;
4280 path->slots[0]++;
4281 cond_resched();
4282 }
4283 if (ins_nr > 0) {
4284 u64 last_extent = 0;
4285
4286 ret = copy_items(trans, inode, dst_path, path,
4287 &last_extent, start_slot,
4288 ins_nr, 1, 0);
4289 /* can't be 1, extent items aren't processed */
4290 ASSERT(ret <= 0);
4291 if (ret < 0)
4292 return ret;
4293 }
4294
4295 return 0;
4296}
4297
Filipe Mananaa89ca6f2015-06-25 04:17:46 +01004298/*
4299 * If the no holes feature is enabled we need to make sure any hole between the
4300 * last extent and the i_size of our inode is explicitly marked in the log. This
4301 * is to make sure that doing something like:
4302 *
4303 * 1) create file with 128Kb of data
4304 * 2) truncate file to 64Kb
4305 * 3) truncate file to 256Kb
4306 * 4) fsync file
4307 * 5) <crash/power failure>
4308 * 6) mount fs and trigger log replay
4309 *
4310 * Will give us a file with a size of 256Kb, the first 64Kb of data match what
4311 * the file had in its first 64Kb of data at step 1 and the last 192Kb of the
4312 * file correspond to a hole. The presence of explicit holes in a log tree is
4313 * what guarantees that log replay will remove/adjust file extent items in the
4314 * fs/subvol tree.
4315 *
4316 * Here we do not need to care about holes between extents, that is already done
4317 * by copy_items(). We also only need to do this in the full sync path, where we
4318 * lookup for extents from the fs/subvol tree only. In the fast path case, we
4319 * lookup the list of modified extent maps and if any represents a hole, we
4320 * insert a corresponding extent representing a hole in the log tree.
4321 */
4322static int btrfs_log_trailing_hole(struct btrfs_trans_handle *trans,
4323 struct btrfs_root *root,
4324 struct inode *inode,
4325 struct btrfs_path *path)
4326{
4327 int ret;
4328 struct btrfs_key key;
4329 u64 hole_start;
4330 u64 hole_size;
4331 struct extent_buffer *leaf;
4332 struct btrfs_root *log = root->log_root;
4333 const u64 ino = btrfs_ino(inode);
4334 const u64 i_size = i_size_read(inode);
4335
4336 if (!btrfs_fs_incompat(root->fs_info, NO_HOLES))
4337 return 0;
4338
4339 key.objectid = ino;
4340 key.type = BTRFS_EXTENT_DATA_KEY;
4341 key.offset = (u64)-1;
4342
4343 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
4344 ASSERT(ret != 0);
4345 if (ret < 0)
4346 return ret;
4347
4348 ASSERT(path->slots[0] > 0);
4349 path->slots[0]--;
4350 leaf = path->nodes[0];
4351 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
4352
4353 if (key.objectid != ino || key.type != BTRFS_EXTENT_DATA_KEY) {
4354 /* inode does not have any extents */
4355 hole_start = 0;
4356 hole_size = i_size;
4357 } else {
4358 struct btrfs_file_extent_item *extent;
4359 u64 len;
4360
4361 /*
4362 * If there's an extent beyond i_size, an explicit hole was
4363 * already inserted by copy_items().
4364 */
4365 if (key.offset >= i_size)
4366 return 0;
4367
4368 extent = btrfs_item_ptr(leaf, path->slots[0],
4369 struct btrfs_file_extent_item);
4370
4371 if (btrfs_file_extent_type(leaf, extent) ==
4372 BTRFS_FILE_EXTENT_INLINE) {
4373 len = btrfs_file_extent_inline_len(leaf,
4374 path->slots[0],
4375 extent);
4376 ASSERT(len == i_size);
4377 return 0;
4378 }
4379
4380 len = btrfs_file_extent_num_bytes(leaf, extent);
4381 /* Last extent goes beyond i_size, no need to log a hole. */
4382 if (key.offset + len > i_size)
4383 return 0;
4384 hole_start = key.offset + len;
4385 hole_size = i_size - hole_start;
4386 }
4387 btrfs_release_path(path);
4388
4389 /* Last extent ends at i_size. */
4390 if (hole_size == 0)
4391 return 0;
4392
4393 hole_size = ALIGN(hole_size, root->sectorsize);
4394 ret = btrfs_insert_file_extent(trans, log, ino, hole_start, 0, 0,
4395 hole_size, 0, hole_size, 0, 0, 0);
4396 return ret;
4397}
4398
Chris Masone02119d2008-09-05 16:13:11 -04004399/* log a single inode in the tree log.
4400 * At least one parent directory for this inode must exist in the tree
4401 * or be logged already.
4402 *
4403 * Any items from this inode changed by the current transaction are copied
4404 * to the log tree. An extra reference is taken on any extents in this
4405 * file, allowing us to avoid a whole pile of corner cases around logging
4406 * blocks that have been removed from the tree.
4407 *
4408 * See LOG_INODE_ALL and related defines for a description of what inode_only
4409 * does.
4410 *
4411 * This handles both files and directories.
4412 */
Chris Mason12fcfd22009-03-24 10:24:20 -04004413static int btrfs_log_inode(struct btrfs_trans_handle *trans,
Filipe Manana49dae1b2014-09-06 22:34:39 +01004414 struct btrfs_root *root, struct inode *inode,
4415 int inode_only,
4416 const loff_t start,
Filipe Manana8407f552014-09-05 15:14:39 +01004417 const loff_t end,
4418 struct btrfs_log_ctx *ctx)
Chris Masone02119d2008-09-05 16:13:11 -04004419{
4420 struct btrfs_path *path;
4421 struct btrfs_path *dst_path;
4422 struct btrfs_key min_key;
4423 struct btrfs_key max_key;
4424 struct btrfs_root *log = root->log_root;
Chris Mason31ff1cd2008-09-11 16:17:57 -04004425 struct extent_buffer *src = NULL;
Miao Xie827463c2014-01-14 20:31:51 +08004426 LIST_HEAD(logged_list);
Josef Bacik16e75492013-10-22 12:18:51 -04004427 u64 last_extent = 0;
Yan, Zheng4a500fd2010-05-16 10:49:59 -04004428 int err = 0;
Chris Masone02119d2008-09-05 16:13:11 -04004429 int ret;
Chris Mason3a5f1d42008-09-11 15:53:37 -04004430 int nritems;
Chris Mason31ff1cd2008-09-11 16:17:57 -04004431 int ins_start_slot = 0;
4432 int ins_nr;
Josef Bacik5dc562c2012-08-17 13:14:17 -04004433 bool fast_search = false;
Li Zefan33345d012011-04-20 10:31:50 +08004434 u64 ino = btrfs_ino(inode);
Filipe Manana49dae1b2014-09-06 22:34:39 +01004435 struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
Filipe Manana1a4bcf42015-02-13 12:30:56 +00004436 u64 logged_isize = 0;
Filipe Mananae4545de2015-06-17 12:49:23 +01004437 bool need_log_inode_item = true;
Chris Masone02119d2008-09-05 16:13:11 -04004438
Chris Masone02119d2008-09-05 16:13:11 -04004439 path = btrfs_alloc_path();
Tsutomu Itoh5df67082011-02-01 09:17:35 +00004440 if (!path)
4441 return -ENOMEM;
Chris Masone02119d2008-09-05 16:13:11 -04004442 dst_path = btrfs_alloc_path();
Tsutomu Itoh5df67082011-02-01 09:17:35 +00004443 if (!dst_path) {
4444 btrfs_free_path(path);
4445 return -ENOMEM;
4446 }
Chris Masone02119d2008-09-05 16:13:11 -04004447
Li Zefan33345d012011-04-20 10:31:50 +08004448 min_key.objectid = ino;
Chris Masone02119d2008-09-05 16:13:11 -04004449 min_key.type = BTRFS_INODE_ITEM_KEY;
4450 min_key.offset = 0;
4451
Li Zefan33345d012011-04-20 10:31:50 +08004452 max_key.objectid = ino;
Chris Mason12fcfd22009-03-24 10:24:20 -04004453
Chris Mason12fcfd22009-03-24 10:24:20 -04004454
Josef Bacik5dc562c2012-08-17 13:14:17 -04004455 /* today the code can only do partial logging of directories */
Miao Xie5269b672012-11-01 07:35:23 +00004456 if (S_ISDIR(inode->i_mode) ||
4457 (!test_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
4458 &BTRFS_I(inode)->runtime_flags) &&
4459 inode_only == LOG_INODE_EXISTS))
Chris Masone02119d2008-09-05 16:13:11 -04004460 max_key.type = BTRFS_XATTR_ITEM_KEY;
4461 else
4462 max_key.type = (u8)-1;
4463 max_key.offset = (u64)-1;
4464
Filipe Manana2c2c4522015-01-13 16:40:04 +00004465 /*
4466 * Only run delayed items if we are a dir or a new file.
4467 * Otherwise commit the delayed inode only, which is needed in
4468 * order for the log replay code to mark inodes for link count
4469 * fixup (create temporary BTRFS_TREE_LOG_FIXUP_OBJECTID items).
4470 */
Josef Bacik94edf4a2012-09-25 14:56:25 -04004471 if (S_ISDIR(inode->i_mode) ||
Filipe Manana2c2c4522015-01-13 16:40:04 +00004472 BTRFS_I(inode)->generation > root->fs_info->last_trans_committed)
Josef Bacik94edf4a2012-09-25 14:56:25 -04004473 ret = btrfs_commit_inode_delayed_items(trans, inode);
Filipe Manana2c2c4522015-01-13 16:40:04 +00004474 else
4475 ret = btrfs_commit_inode_delayed_inode(inode);
4476
4477 if (ret) {
4478 btrfs_free_path(path);
4479 btrfs_free_path(dst_path);
4480 return ret;
Miao Xie16cdcec2011-04-22 18:12:22 +08004481 }
4482
Chris Masone02119d2008-09-05 16:13:11 -04004483 mutex_lock(&BTRFS_I(inode)->log_mutex);
4484
Filipe Manana08702952014-11-13 17:00:35 +00004485 btrfs_get_logged_extents(inode, &logged_list, start, end);
Josef Bacik2ab28f32012-10-12 15:27:49 -04004486
Chris Masone02119d2008-09-05 16:13:11 -04004487 /*
4488 * a brute force approach to making sure we get the most uptodate
4489 * copies of everything.
4490 */
4491 if (S_ISDIR(inode->i_mode)) {
4492 int max_key_type = BTRFS_DIR_LOG_INDEX_KEY;
4493
Filipe Manana4f764e52015-02-23 19:53:35 +00004494 if (inode_only == LOG_INODE_EXISTS)
4495 max_key_type = BTRFS_XATTR_ITEM_KEY;
Li Zefan33345d012011-04-20 10:31:50 +08004496 ret = drop_objectid_items(trans, log, path, ino, max_key_type);
Chris Masone02119d2008-09-05 16:13:11 -04004497 } else {
Filipe Manana1a4bcf42015-02-13 12:30:56 +00004498 if (inode_only == LOG_INODE_EXISTS) {
4499 /*
4500 * Make sure the new inode item we write to the log has
4501 * the same isize as the current one (if it exists).
4502 * This is necessary to prevent data loss after log
4503 * replay, and also to prevent doing a wrong expanding
4504 * truncate - for e.g. create file, write 4K into offset
4505 * 0, fsync, write 4K into offset 4096, add hard link,
4506 * fsync some other file (to sync log), power fail - if
4507 * we use the inode's current i_size, after log replay
4508 * we get a 8Kb file, with the last 4Kb extent as a hole
4509 * (zeroes), as if an expanding truncate happened,
4510 * instead of getting a file of 4Kb only.
4511 */
4512 err = logged_inode_size(log, inode, path,
4513 &logged_isize);
4514 if (err)
4515 goto out_unlock;
4516 }
Filipe Mananaa7429942015-02-13 16:56:14 +00004517 if (test_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
4518 &BTRFS_I(inode)->runtime_flags)) {
4519 if (inode_only == LOG_INODE_EXISTS) {
Filipe Manana4f764e52015-02-23 19:53:35 +00004520 max_key.type = BTRFS_XATTR_ITEM_KEY;
Filipe Mananaa7429942015-02-13 16:56:14 +00004521 ret = drop_objectid_items(trans, log, path, ino,
4522 max_key.type);
4523 } else {
4524 clear_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
4525 &BTRFS_I(inode)->runtime_flags);
4526 clear_bit(BTRFS_INODE_COPY_EVERYTHING,
4527 &BTRFS_I(inode)->runtime_flags);
Chris Mason28ed1342014-12-17 09:41:04 -08004528 while(1) {
4529 ret = btrfs_truncate_inode_items(trans,
4530 log, inode, 0, 0);
4531 if (ret != -EAGAIN)
4532 break;
4533 }
Filipe Mananaa7429942015-02-13 16:56:14 +00004534 }
Filipe Manana4f764e52015-02-23 19:53:35 +00004535 } else if (test_and_clear_bit(BTRFS_INODE_COPY_EVERYTHING,
4536 &BTRFS_I(inode)->runtime_flags) ||
Josef Bacik6cfab852013-11-12 16:25:58 -05004537 inode_only == LOG_INODE_EXISTS) {
Filipe Manana4f764e52015-02-23 19:53:35 +00004538 if (inode_only == LOG_INODE_ALL)
Josef Bacika95249b2012-10-11 16:17:34 -04004539 fast_search = true;
Filipe Manana4f764e52015-02-23 19:53:35 +00004540 max_key.type = BTRFS_XATTR_ITEM_KEY;
Josef Bacika95249b2012-10-11 16:17:34 -04004541 ret = drop_objectid_items(trans, log, path, ino,
4542 max_key.type);
Josef Bacik5dc562c2012-08-17 13:14:17 -04004543 } else {
Liu Bo183f37f2012-11-01 06:38:47 +00004544 if (inode_only == LOG_INODE_ALL)
4545 fast_search = true;
Josef Bacika95249b2012-10-11 16:17:34 -04004546 goto log_extents;
Josef Bacik5dc562c2012-08-17 13:14:17 -04004547 }
Josef Bacika95249b2012-10-11 16:17:34 -04004548
Chris Masone02119d2008-09-05 16:13:11 -04004549 }
Yan, Zheng4a500fd2010-05-16 10:49:59 -04004550 if (ret) {
4551 err = ret;
4552 goto out_unlock;
4553 }
Chris Masone02119d2008-09-05 16:13:11 -04004554
Chris Masond3977122009-01-05 21:25:51 -05004555 while (1) {
Chris Mason31ff1cd2008-09-11 16:17:57 -04004556 ins_nr = 0;
Filipe David Borba Manana6174d3c2013-10-01 16:13:42 +01004557 ret = btrfs_search_forward(root, &min_key,
Eric Sandeende78b512013-01-31 18:21:12 +00004558 path, trans->transid);
Chris Masone02119d2008-09-05 16:13:11 -04004559 if (ret != 0)
4560 break;
Chris Mason3a5f1d42008-09-11 15:53:37 -04004561again:
Chris Mason31ff1cd2008-09-11 16:17:57 -04004562 /* note, ins_nr might be > 0 here, cleanup outside the loop */
Li Zefan33345d012011-04-20 10:31:50 +08004563 if (min_key.objectid != ino)
Chris Masone02119d2008-09-05 16:13:11 -04004564 break;
4565 if (min_key.type > max_key.type)
4566 break;
Chris Mason31ff1cd2008-09-11 16:17:57 -04004567
Filipe Mananae4545de2015-06-17 12:49:23 +01004568 if (min_key.type == BTRFS_INODE_ITEM_KEY)
4569 need_log_inode_item = false;
4570
Filipe Manana36283bf2015-06-20 00:44:51 +01004571 /* Skip xattrs, we log them later with btrfs_log_all_xattrs() */
4572 if (min_key.type == BTRFS_XATTR_ITEM_KEY) {
4573 if (ins_nr == 0)
4574 goto next_slot;
4575 ret = copy_items(trans, inode, dst_path, path,
4576 &last_extent, ins_start_slot,
4577 ins_nr, inode_only, logged_isize);
4578 if (ret < 0) {
4579 err = ret;
4580 goto out_unlock;
4581 }
4582 ins_nr = 0;
4583 if (ret) {
4584 btrfs_release_path(path);
4585 continue;
4586 }
4587 goto next_slot;
4588 }
4589
Chris Masone02119d2008-09-05 16:13:11 -04004590 src = path->nodes[0];
Chris Mason31ff1cd2008-09-11 16:17:57 -04004591 if (ins_nr && ins_start_slot + ins_nr == path->slots[0]) {
4592 ins_nr++;
4593 goto next_slot;
4594 } else if (!ins_nr) {
4595 ins_start_slot = path->slots[0];
4596 ins_nr = 1;
4597 goto next_slot;
Chris Masone02119d2008-09-05 16:13:11 -04004598 }
4599
Josef Bacik16e75492013-10-22 12:18:51 -04004600 ret = copy_items(trans, inode, dst_path, path, &last_extent,
Filipe Manana1a4bcf42015-02-13 12:30:56 +00004601 ins_start_slot, ins_nr, inode_only,
4602 logged_isize);
Josef Bacik16e75492013-10-22 12:18:51 -04004603 if (ret < 0) {
Yan, Zheng4a500fd2010-05-16 10:49:59 -04004604 err = ret;
4605 goto out_unlock;
Rasmus Villemoesa71db862014-06-20 21:51:43 +02004606 }
4607 if (ret) {
Josef Bacik16e75492013-10-22 12:18:51 -04004608 ins_nr = 0;
4609 btrfs_release_path(path);
4610 continue;
Yan, Zheng4a500fd2010-05-16 10:49:59 -04004611 }
Chris Mason31ff1cd2008-09-11 16:17:57 -04004612 ins_nr = 1;
4613 ins_start_slot = path->slots[0];
4614next_slot:
Chris Masone02119d2008-09-05 16:13:11 -04004615
Chris Mason3a5f1d42008-09-11 15:53:37 -04004616 nritems = btrfs_header_nritems(path->nodes[0]);
4617 path->slots[0]++;
4618 if (path->slots[0] < nritems) {
4619 btrfs_item_key_to_cpu(path->nodes[0], &min_key,
4620 path->slots[0]);
4621 goto again;
4622 }
Chris Mason31ff1cd2008-09-11 16:17:57 -04004623 if (ins_nr) {
Josef Bacik16e75492013-10-22 12:18:51 -04004624 ret = copy_items(trans, inode, dst_path, path,
4625 &last_extent, ins_start_slot,
Filipe Manana1a4bcf42015-02-13 12:30:56 +00004626 ins_nr, inode_only, logged_isize);
Josef Bacik16e75492013-10-22 12:18:51 -04004627 if (ret < 0) {
Yan, Zheng4a500fd2010-05-16 10:49:59 -04004628 err = ret;
4629 goto out_unlock;
4630 }
Josef Bacik16e75492013-10-22 12:18:51 -04004631 ret = 0;
Chris Mason31ff1cd2008-09-11 16:17:57 -04004632 ins_nr = 0;
4633 }
David Sterbab3b4aa72011-04-21 01:20:15 +02004634 btrfs_release_path(path);
Chris Mason3a5f1d42008-09-11 15:53:37 -04004635
Filipe David Borba Manana3d41d702013-10-01 17:06:53 +01004636 if (min_key.offset < (u64)-1) {
Chris Masone02119d2008-09-05 16:13:11 -04004637 min_key.offset++;
Filipe David Borba Manana3d41d702013-10-01 17:06:53 +01004638 } else if (min_key.type < max_key.type) {
Chris Masone02119d2008-09-05 16:13:11 -04004639 min_key.type++;
Filipe David Borba Manana3d41d702013-10-01 17:06:53 +01004640 min_key.offset = 0;
4641 } else {
Chris Masone02119d2008-09-05 16:13:11 -04004642 break;
Filipe David Borba Manana3d41d702013-10-01 17:06:53 +01004643 }
Chris Masone02119d2008-09-05 16:13:11 -04004644 }
Chris Mason31ff1cd2008-09-11 16:17:57 -04004645 if (ins_nr) {
Josef Bacik16e75492013-10-22 12:18:51 -04004646 ret = copy_items(trans, inode, dst_path, path, &last_extent,
Filipe Manana1a4bcf42015-02-13 12:30:56 +00004647 ins_start_slot, ins_nr, inode_only,
4648 logged_isize);
Josef Bacik16e75492013-10-22 12:18:51 -04004649 if (ret < 0) {
Yan, Zheng4a500fd2010-05-16 10:49:59 -04004650 err = ret;
4651 goto out_unlock;
4652 }
Josef Bacik16e75492013-10-22 12:18:51 -04004653 ret = 0;
Chris Mason31ff1cd2008-09-11 16:17:57 -04004654 ins_nr = 0;
4655 }
Josef Bacik5dc562c2012-08-17 13:14:17 -04004656
Filipe Manana36283bf2015-06-20 00:44:51 +01004657 btrfs_release_path(path);
4658 btrfs_release_path(dst_path);
4659 err = btrfs_log_all_xattrs(trans, root, inode, path, dst_path);
4660 if (err)
4661 goto out_unlock;
Filipe Mananaa89ca6f2015-06-25 04:17:46 +01004662 if (max_key.type >= BTRFS_EXTENT_DATA_KEY && !fast_search) {
4663 btrfs_release_path(path);
4664 btrfs_release_path(dst_path);
4665 err = btrfs_log_trailing_hole(trans, root, inode, path);
4666 if (err)
4667 goto out_unlock;
4668 }
Josef Bacika95249b2012-10-11 16:17:34 -04004669log_extents:
Josef Bacikf3b15cc2013-07-22 12:54:30 -04004670 btrfs_release_path(path);
4671 btrfs_release_path(dst_path);
Filipe Mananae4545de2015-06-17 12:49:23 +01004672 if (need_log_inode_item) {
4673 err = log_inode_item(trans, log, dst_path, inode);
4674 if (err)
4675 goto out_unlock;
4676 }
Josef Bacik5dc562c2012-08-17 13:14:17 -04004677 if (fast_search) {
Filipe Mananab38ef712014-11-13 17:01:45 +00004678 /*
4679 * Some ordered extents started by fsync might have completed
4680 * before we collected the ordered extents in logged_list, which
4681 * means they're gone, not in our logged_list nor in the inode's
4682 * ordered tree. We want the application/user space to know an
4683 * error happened while attempting to persist file data so that
4684 * it can take proper action. If such error happened, we leave
4685 * without writing to the log tree and the fsync must report the
4686 * file data write error and not commit the current transaction.
4687 */
4688 err = btrfs_inode_check_errors(inode);
4689 if (err) {
4690 ctx->io_err = err;
4691 goto out_unlock;
4692 }
Miao Xie827463c2014-01-14 20:31:51 +08004693 ret = btrfs_log_changed_extents(trans, root, inode, dst_path,
Filipe Manana8407f552014-09-05 15:14:39 +01004694 &logged_list, ctx);
Josef Bacik5dc562c2012-08-17 13:14:17 -04004695 if (ret) {
4696 err = ret;
4697 goto out_unlock;
4698 }
Josef Bacikd006a042013-11-12 20:54:09 -05004699 } else if (inode_only == LOG_INODE_ALL) {
Liu Bo06d3d222012-08-27 10:52:19 -06004700 struct extent_map *em, *n;
4701
Filipe Manana49dae1b2014-09-06 22:34:39 +01004702 write_lock(&em_tree->lock);
4703 /*
4704 * We can't just remove every em if we're called for a ranged
4705 * fsync - that is, one that doesn't cover the whole possible
4706 * file range (0 to LLONG_MAX). This is because we can have
4707 * em's that fall outside the range we're logging and therefore
4708 * their ordered operations haven't completed yet
4709 * (btrfs_finish_ordered_io() not invoked yet). This means we
4710 * didn't get their respective file extent item in the fs/subvol
4711 * tree yet, and need to let the next fast fsync (one which
4712 * consults the list of modified extent maps) find the em so
4713 * that it logs a matching file extent item and waits for the
4714 * respective ordered operation to complete (if it's still
4715 * running).
4716 *
4717 * Removing every em outside the range we're logging would make
4718 * the next fast fsync not log their matching file extent items,
4719 * therefore making us lose data after a log replay.
4720 */
4721 list_for_each_entry_safe(em, n, &em_tree->modified_extents,
4722 list) {
4723 const u64 mod_end = em->mod_start + em->mod_len - 1;
4724
4725 if (em->mod_start >= start && mod_end <= end)
4726 list_del_init(&em->list);
4727 }
4728 write_unlock(&em_tree->lock);
Josef Bacik5dc562c2012-08-17 13:14:17 -04004729 }
4730
Chris Mason9623f9a2008-09-11 17:42:42 -04004731 if (inode_only == LOG_INODE_ALL && S_ISDIR(inode->i_mode)) {
Filipe Manana2f2ff0e2015-03-20 17:19:46 +00004732 ret = log_directory_changes(trans, root, inode, path, dst_path,
4733 ctx);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04004734 if (ret) {
4735 err = ret;
4736 goto out_unlock;
4737 }
Chris Masone02119d2008-09-05 16:13:11 -04004738 }
Filipe Manana49dae1b2014-09-06 22:34:39 +01004739
Filipe Manana2f2ff0e2015-03-20 17:19:46 +00004740 spin_lock(&BTRFS_I(inode)->lock);
Filipe Manana125c4cf92014-09-11 21:22:14 +01004741 BTRFS_I(inode)->logged_trans = trans->transid;
4742 BTRFS_I(inode)->last_log_commit = BTRFS_I(inode)->last_sub_trans;
Filipe Manana2f2ff0e2015-03-20 17:19:46 +00004743 spin_unlock(&BTRFS_I(inode)->lock);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04004744out_unlock:
Miao Xie827463c2014-01-14 20:31:51 +08004745 if (unlikely(err))
4746 btrfs_put_logged_extents(&logged_list);
4747 else
4748 btrfs_submit_logged_extents(&logged_list, log);
Chris Masone02119d2008-09-05 16:13:11 -04004749 mutex_unlock(&BTRFS_I(inode)->log_mutex);
4750
4751 btrfs_free_path(path);
4752 btrfs_free_path(dst_path);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04004753 return err;
Chris Masone02119d2008-09-05 16:13:11 -04004754}
4755
Chris Mason12fcfd22009-03-24 10:24:20 -04004756/*
4757 * follow the dentry parent pointers up the chain and see if any
4758 * of the directories in it require a full commit before they can
4759 * be logged. Returns zero if nothing special needs to be done or 1 if
4760 * a full commit is required.
4761 */
4762static noinline int check_parent_dirs_for_sync(struct btrfs_trans_handle *trans,
4763 struct inode *inode,
4764 struct dentry *parent,
4765 struct super_block *sb,
4766 u64 last_committed)
Chris Masone02119d2008-09-05 16:13:11 -04004767{
Chris Mason12fcfd22009-03-24 10:24:20 -04004768 int ret = 0;
4769 struct btrfs_root *root;
Josef Bacik6a912212010-11-20 09:48:00 +00004770 struct dentry *old_parent = NULL;
Josef Bacikde2b5302013-09-11 09:36:30 -04004771 struct inode *orig_inode = inode;
Chris Masone02119d2008-09-05 16:13:11 -04004772
Chris Masonaf4176b2009-03-24 10:24:31 -04004773 /*
4774 * for regular files, if its inode is already on disk, we don't
4775 * have to worry about the parents at all. This is because
4776 * we can use the last_unlink_trans field to record renames
4777 * and other fun in this file.
4778 */
4779 if (S_ISREG(inode->i_mode) &&
4780 BTRFS_I(inode)->generation <= last_committed &&
4781 BTRFS_I(inode)->last_unlink_trans <= last_committed)
4782 goto out;
4783
Chris Mason12fcfd22009-03-24 10:24:20 -04004784 if (!S_ISDIR(inode->i_mode)) {
David Howells2b0143b2015-03-17 22:25:59 +00004785 if (!parent || d_really_is_negative(parent) || sb != d_inode(parent)->i_sb)
Chris Mason12fcfd22009-03-24 10:24:20 -04004786 goto out;
David Howells2b0143b2015-03-17 22:25:59 +00004787 inode = d_inode(parent);
Chris Mason12fcfd22009-03-24 10:24:20 -04004788 }
4789
4790 while (1) {
Josef Bacikde2b5302013-09-11 09:36:30 -04004791 /*
4792 * If we are logging a directory then we start with our inode,
4793 * not our parents inode, so we need to skipp setting the
4794 * logged_trans so that further down in the log code we don't
4795 * think this inode has already been logged.
4796 */
4797 if (inode != orig_inode)
4798 BTRFS_I(inode)->logged_trans = trans->transid;
Chris Mason12fcfd22009-03-24 10:24:20 -04004799 smp_mb();
4800
4801 if (BTRFS_I(inode)->last_unlink_trans > last_committed) {
4802 root = BTRFS_I(inode)->root;
4803
4804 /*
4805 * make sure any commits to the log are forced
4806 * to be full commits
4807 */
Miao Xie995946d2014-04-02 19:51:06 +08004808 btrfs_set_log_full_commit(root->fs_info, trans);
Chris Mason12fcfd22009-03-24 10:24:20 -04004809 ret = 1;
4810 break;
4811 }
4812
David Howells2b0143b2015-03-17 22:25:59 +00004813 if (!parent || d_really_is_negative(parent) || sb != d_inode(parent)->i_sb)
Chris Mason12fcfd22009-03-24 10:24:20 -04004814 break;
4815
Yan, Zheng76dda932009-09-21 16:00:26 -04004816 if (IS_ROOT(parent))
Chris Mason12fcfd22009-03-24 10:24:20 -04004817 break;
4818
Josef Bacik6a912212010-11-20 09:48:00 +00004819 parent = dget_parent(parent);
4820 dput(old_parent);
4821 old_parent = parent;
David Howells2b0143b2015-03-17 22:25:59 +00004822 inode = d_inode(parent);
Chris Mason12fcfd22009-03-24 10:24:20 -04004823
4824 }
Josef Bacik6a912212010-11-20 09:48:00 +00004825 dput(old_parent);
Chris Mason12fcfd22009-03-24 10:24:20 -04004826out:
Chris Masone02119d2008-09-05 16:13:11 -04004827 return ret;
4828}
4829
Filipe Manana2f2ff0e2015-03-20 17:19:46 +00004830struct btrfs_dir_list {
4831 u64 ino;
4832 struct list_head list;
4833};
4834
4835/*
4836 * Log the inodes of the new dentries of a directory. See log_dir_items() for
4837 * details about the why it is needed.
4838 * This is a recursive operation - if an existing dentry corresponds to a
4839 * directory, that directory's new entries are logged too (same behaviour as
4840 * ext3/4, xfs, f2fs, reiserfs, nilfs2). Note that when logging the inodes
4841 * the dentries point to we do not lock their i_mutex, otherwise lockdep
4842 * complains about the following circular lock dependency / possible deadlock:
4843 *
4844 * CPU0 CPU1
4845 * ---- ----
4846 * lock(&type->i_mutex_dir_key#3/2);
4847 * lock(sb_internal#2);
4848 * lock(&type->i_mutex_dir_key#3/2);
4849 * lock(&sb->s_type->i_mutex_key#14);
4850 *
4851 * Where sb_internal is the lock (a counter that works as a lock) acquired by
4852 * sb_start_intwrite() in btrfs_start_transaction().
4853 * Not locking i_mutex of the inodes is still safe because:
4854 *
4855 * 1) For regular files we log with a mode of LOG_INODE_EXISTS. It's possible
4856 * that while logging the inode new references (names) are added or removed
4857 * from the inode, leaving the logged inode item with a link count that does
4858 * not match the number of logged inode reference items. This is fine because
4859 * at log replay time we compute the real number of links and correct the
4860 * link count in the inode item (see replay_one_buffer() and
4861 * link_to_fixup_dir());
4862 *
4863 * 2) For directories we log with a mode of LOG_INODE_ALL. It's possible that
4864 * while logging the inode's items new items with keys BTRFS_DIR_ITEM_KEY and
4865 * BTRFS_DIR_INDEX_KEY are added to fs/subvol tree and the logged inode item
4866 * has a size that doesn't match the sum of the lengths of all the logged
4867 * names. This does not result in a problem because if a dir_item key is
4868 * logged but its matching dir_index key is not logged, at log replay time we
4869 * don't use it to replay the respective name (see replay_one_name()). On the
4870 * other hand if only the dir_index key ends up being logged, the respective
4871 * name is added to the fs/subvol tree with both the dir_item and dir_index
4872 * keys created (see replay_one_name()).
4873 * The directory's inode item with a wrong i_size is not a problem as well,
4874 * since we don't use it at log replay time to set the i_size in the inode
4875 * item of the fs/subvol tree (see overwrite_item()).
4876 */
4877static int log_new_dir_dentries(struct btrfs_trans_handle *trans,
4878 struct btrfs_root *root,
4879 struct inode *start_inode,
4880 struct btrfs_log_ctx *ctx)
4881{
4882 struct btrfs_root *log = root->log_root;
4883 struct btrfs_path *path;
4884 LIST_HEAD(dir_list);
4885 struct btrfs_dir_list *dir_elem;
4886 int ret = 0;
4887
4888 path = btrfs_alloc_path();
4889 if (!path)
4890 return -ENOMEM;
4891
4892 dir_elem = kmalloc(sizeof(*dir_elem), GFP_NOFS);
4893 if (!dir_elem) {
4894 btrfs_free_path(path);
4895 return -ENOMEM;
4896 }
4897 dir_elem->ino = btrfs_ino(start_inode);
4898 list_add_tail(&dir_elem->list, &dir_list);
4899
4900 while (!list_empty(&dir_list)) {
4901 struct extent_buffer *leaf;
4902 struct btrfs_key min_key;
4903 int nritems;
4904 int i;
4905
4906 dir_elem = list_first_entry(&dir_list, struct btrfs_dir_list,
4907 list);
4908 if (ret)
4909 goto next_dir_inode;
4910
4911 min_key.objectid = dir_elem->ino;
4912 min_key.type = BTRFS_DIR_ITEM_KEY;
4913 min_key.offset = 0;
4914again:
4915 btrfs_release_path(path);
4916 ret = btrfs_search_forward(log, &min_key, path, trans->transid);
4917 if (ret < 0) {
4918 goto next_dir_inode;
4919 } else if (ret > 0) {
4920 ret = 0;
4921 goto next_dir_inode;
4922 }
4923
4924process_leaf:
4925 leaf = path->nodes[0];
4926 nritems = btrfs_header_nritems(leaf);
4927 for (i = path->slots[0]; i < nritems; i++) {
4928 struct btrfs_dir_item *di;
4929 struct btrfs_key di_key;
4930 struct inode *di_inode;
4931 struct btrfs_dir_list *new_dir_elem;
4932 int log_mode = LOG_INODE_EXISTS;
4933 int type;
4934
4935 btrfs_item_key_to_cpu(leaf, &min_key, i);
4936 if (min_key.objectid != dir_elem->ino ||
4937 min_key.type != BTRFS_DIR_ITEM_KEY)
4938 goto next_dir_inode;
4939
4940 di = btrfs_item_ptr(leaf, i, struct btrfs_dir_item);
4941 type = btrfs_dir_type(leaf, di);
4942 if (btrfs_dir_transid(leaf, di) < trans->transid &&
4943 type != BTRFS_FT_DIR)
4944 continue;
4945 btrfs_dir_item_key_to_cpu(leaf, di, &di_key);
4946 if (di_key.type == BTRFS_ROOT_ITEM_KEY)
4947 continue;
4948
4949 di_inode = btrfs_iget(root->fs_info->sb, &di_key,
4950 root, NULL);
4951 if (IS_ERR(di_inode)) {
4952 ret = PTR_ERR(di_inode);
4953 goto next_dir_inode;
4954 }
4955
4956 if (btrfs_inode_in_log(di_inode, trans->transid)) {
4957 iput(di_inode);
4958 continue;
4959 }
4960
4961 ctx->log_new_dentries = false;
4962 if (type == BTRFS_FT_DIR)
4963 log_mode = LOG_INODE_ALL;
4964 btrfs_release_path(path);
4965 ret = btrfs_log_inode(trans, root, di_inode,
4966 log_mode, 0, LLONG_MAX, ctx);
4967 iput(di_inode);
4968 if (ret)
4969 goto next_dir_inode;
4970 if (ctx->log_new_dentries) {
4971 new_dir_elem = kmalloc(sizeof(*new_dir_elem),
4972 GFP_NOFS);
4973 if (!new_dir_elem) {
4974 ret = -ENOMEM;
4975 goto next_dir_inode;
4976 }
4977 new_dir_elem->ino = di_key.objectid;
4978 list_add_tail(&new_dir_elem->list, &dir_list);
4979 }
4980 break;
4981 }
4982 if (i == nritems) {
4983 ret = btrfs_next_leaf(log, path);
4984 if (ret < 0) {
4985 goto next_dir_inode;
4986 } else if (ret > 0) {
4987 ret = 0;
4988 goto next_dir_inode;
4989 }
4990 goto process_leaf;
4991 }
4992 if (min_key.offset < (u64)-1) {
4993 min_key.offset++;
4994 goto again;
4995 }
4996next_dir_inode:
4997 list_del(&dir_elem->list);
4998 kfree(dir_elem);
4999 }
5000
5001 btrfs_free_path(path);
5002 return ret;
5003}
5004
Filipe Manana18aa0922015-08-05 16:49:08 +01005005static int btrfs_log_all_parents(struct btrfs_trans_handle *trans,
5006 struct inode *inode,
5007 struct btrfs_log_ctx *ctx)
5008{
5009 int ret;
5010 struct btrfs_path *path;
5011 struct btrfs_key key;
5012 struct btrfs_root *root = BTRFS_I(inode)->root;
5013 const u64 ino = btrfs_ino(inode);
5014
5015 path = btrfs_alloc_path();
5016 if (!path)
5017 return -ENOMEM;
5018 path->skip_locking = 1;
5019 path->search_commit_root = 1;
5020
5021 key.objectid = ino;
5022 key.type = BTRFS_INODE_REF_KEY;
5023 key.offset = 0;
5024 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
5025 if (ret < 0)
5026 goto out;
5027
5028 while (true) {
5029 struct extent_buffer *leaf = path->nodes[0];
5030 int slot = path->slots[0];
5031 u32 cur_offset = 0;
5032 u32 item_size;
5033 unsigned long ptr;
5034
5035 if (slot >= btrfs_header_nritems(leaf)) {
5036 ret = btrfs_next_leaf(root, path);
5037 if (ret < 0)
5038 goto out;
5039 else if (ret > 0)
5040 break;
5041 continue;
5042 }
5043
5044 btrfs_item_key_to_cpu(leaf, &key, slot);
5045 /* BTRFS_INODE_EXTREF_KEY is BTRFS_INODE_REF_KEY + 1 */
5046 if (key.objectid != ino || key.type > BTRFS_INODE_EXTREF_KEY)
5047 break;
5048
5049 item_size = btrfs_item_size_nr(leaf, slot);
5050 ptr = btrfs_item_ptr_offset(leaf, slot);
5051 while (cur_offset < item_size) {
5052 struct btrfs_key inode_key;
5053 struct inode *dir_inode;
5054
5055 inode_key.type = BTRFS_INODE_ITEM_KEY;
5056 inode_key.offset = 0;
5057
5058 if (key.type == BTRFS_INODE_EXTREF_KEY) {
5059 struct btrfs_inode_extref *extref;
5060
5061 extref = (struct btrfs_inode_extref *)
5062 (ptr + cur_offset);
5063 inode_key.objectid = btrfs_inode_extref_parent(
5064 leaf, extref);
5065 cur_offset += sizeof(*extref);
5066 cur_offset += btrfs_inode_extref_name_len(leaf,
5067 extref);
5068 } else {
5069 inode_key.objectid = key.offset;
5070 cur_offset = item_size;
5071 }
5072
5073 dir_inode = btrfs_iget(root->fs_info->sb, &inode_key,
5074 root, NULL);
5075 /* If parent inode was deleted, skip it. */
5076 if (IS_ERR(dir_inode))
5077 continue;
5078
5079 ret = btrfs_log_inode(trans, root, dir_inode,
5080 LOG_INODE_ALL, 0, LLONG_MAX, ctx);
5081 iput(dir_inode);
5082 if (ret)
5083 goto out;
5084 }
5085 path->slots[0]++;
5086 }
5087 ret = 0;
5088out:
5089 btrfs_free_path(path);
5090 return ret;
5091}
5092
Chris Masone02119d2008-09-05 16:13:11 -04005093/*
5094 * helper function around btrfs_log_inode to make sure newly created
5095 * parent directories also end up in the log. A minimal inode and backref
5096 * only logging is done of any parent directories that are older than
5097 * the last committed transaction
5098 */
Eric Sandeen48a3b632013-04-25 20:41:01 +00005099static int btrfs_log_inode_parent(struct btrfs_trans_handle *trans,
5100 struct btrfs_root *root, struct inode *inode,
Filipe Manana49dae1b2014-09-06 22:34:39 +01005101 struct dentry *parent,
5102 const loff_t start,
5103 const loff_t end,
5104 int exists_only,
Miao Xie8b050d32014-02-20 18:08:58 +08005105 struct btrfs_log_ctx *ctx)
Chris Masone02119d2008-09-05 16:13:11 -04005106{
Chris Mason12fcfd22009-03-24 10:24:20 -04005107 int inode_only = exists_only ? LOG_INODE_EXISTS : LOG_INODE_ALL;
Chris Masone02119d2008-09-05 16:13:11 -04005108 struct super_block *sb;
Josef Bacik6a912212010-11-20 09:48:00 +00005109 struct dentry *old_parent = NULL;
Chris Mason12fcfd22009-03-24 10:24:20 -04005110 int ret = 0;
5111 u64 last_committed = root->fs_info->last_trans_committed;
Filipe Manana2f2ff0e2015-03-20 17:19:46 +00005112 bool log_dentries = false;
5113 struct inode *orig_inode = inode;
Chris Mason12fcfd22009-03-24 10:24:20 -04005114
5115 sb = inode->i_sb;
5116
Sage Weil3a5e1402009-04-02 16:49:40 -04005117 if (btrfs_test_opt(root, NOTREELOG)) {
5118 ret = 1;
5119 goto end_no_trans;
5120 }
5121
Miao Xie995946d2014-04-02 19:51:06 +08005122 /*
5123 * The prev transaction commit doesn't complete, we need do
5124 * full commit by ourselves.
5125 */
Chris Mason12fcfd22009-03-24 10:24:20 -04005126 if (root->fs_info->last_trans_log_full_commit >
5127 root->fs_info->last_trans_committed) {
5128 ret = 1;
5129 goto end_no_trans;
5130 }
5131
Yan, Zheng76dda932009-09-21 16:00:26 -04005132 if (root != BTRFS_I(inode)->root ||
5133 btrfs_root_refs(&root->root_item) == 0) {
5134 ret = 1;
5135 goto end_no_trans;
5136 }
5137
Chris Mason12fcfd22009-03-24 10:24:20 -04005138 ret = check_parent_dirs_for_sync(trans, inode, parent,
5139 sb, last_committed);
5140 if (ret)
5141 goto end_no_trans;
Chris Masone02119d2008-09-05 16:13:11 -04005142
Josef Bacik22ee6982012-05-29 16:57:49 -04005143 if (btrfs_inode_in_log(inode, trans->transid)) {
Chris Mason257c62e2009-10-13 13:21:08 -04005144 ret = BTRFS_NO_LOG_SYNC;
5145 goto end_no_trans;
5146 }
5147
Miao Xie8b050d32014-02-20 18:08:58 +08005148 ret = start_log_trans(trans, root, ctx);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04005149 if (ret)
Miao Xiee87ac132014-02-20 18:08:53 +08005150 goto end_no_trans;
Chris Mason12fcfd22009-03-24 10:24:20 -04005151
Filipe Manana8407f552014-09-05 15:14:39 +01005152 ret = btrfs_log_inode(trans, root, inode, inode_only, start, end, ctx);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04005153 if (ret)
5154 goto end_trans;
Chris Mason12fcfd22009-03-24 10:24:20 -04005155
Chris Masonaf4176b2009-03-24 10:24:31 -04005156 /*
5157 * for regular files, if its inode is already on disk, we don't
5158 * have to worry about the parents at all. This is because
5159 * we can use the last_unlink_trans field to record renames
5160 * and other fun in this file.
5161 */
5162 if (S_ISREG(inode->i_mode) &&
5163 BTRFS_I(inode)->generation <= last_committed &&
Yan, Zheng4a500fd2010-05-16 10:49:59 -04005164 BTRFS_I(inode)->last_unlink_trans <= last_committed) {
5165 ret = 0;
5166 goto end_trans;
5167 }
Chris Masonaf4176b2009-03-24 10:24:31 -04005168
Filipe Manana2f2ff0e2015-03-20 17:19:46 +00005169 if (S_ISDIR(inode->i_mode) && ctx && ctx->log_new_dentries)
5170 log_dentries = true;
5171
Filipe Manana18aa0922015-08-05 16:49:08 +01005172 /*
5173 * On unlink we must make sure all our current and old parent directores
5174 * inodes are fully logged. This is to prevent leaving dangling
5175 * directory index entries in directories that were our parents but are
5176 * not anymore. Not doing this results in old parent directory being
5177 * impossible to delete after log replay (rmdir will always fail with
5178 * error -ENOTEMPTY).
5179 *
5180 * Example 1:
5181 *
5182 * mkdir testdir
5183 * touch testdir/foo
5184 * ln testdir/foo testdir/bar
5185 * sync
5186 * unlink testdir/bar
5187 * xfs_io -c fsync testdir/foo
5188 * <power failure>
5189 * mount fs, triggers log replay
5190 *
5191 * If we don't log the parent directory (testdir), after log replay the
5192 * directory still has an entry pointing to the file inode using the bar
5193 * name, but a matching BTRFS_INODE_[REF|EXTREF]_KEY does not exist and
5194 * the file inode has a link count of 1.
5195 *
5196 * Example 2:
5197 *
5198 * mkdir testdir
5199 * touch foo
5200 * ln foo testdir/foo2
5201 * ln foo testdir/foo3
5202 * sync
5203 * unlink testdir/foo3
5204 * xfs_io -c fsync foo
5205 * <power failure>
5206 * mount fs, triggers log replay
5207 *
5208 * Similar as the first example, after log replay the parent directory
5209 * testdir still has an entry pointing to the inode file with name foo3
5210 * but the file inode does not have a matching BTRFS_INODE_REF_KEY item
5211 * and has a link count of 2.
5212 */
5213 if (BTRFS_I(inode)->last_unlink_trans > last_committed) {
5214 ret = btrfs_log_all_parents(trans, orig_inode, ctx);
5215 if (ret)
5216 goto end_trans;
5217 }
5218
Chris Masond3977122009-01-05 21:25:51 -05005219 while (1) {
David Howells2b0143b2015-03-17 22:25:59 +00005220 if (!parent || d_really_is_negative(parent) || sb != d_inode(parent)->i_sb)
Chris Masone02119d2008-09-05 16:13:11 -04005221 break;
5222
David Howells2b0143b2015-03-17 22:25:59 +00005223 inode = d_inode(parent);
Yan, Zheng76dda932009-09-21 16:00:26 -04005224 if (root != BTRFS_I(inode)->root)
5225 break;
5226
Filipe Manana18aa0922015-08-05 16:49:08 +01005227 if (BTRFS_I(inode)->generation > last_committed) {
5228 ret = btrfs_log_inode(trans, root, inode,
5229 LOG_INODE_EXISTS,
Filipe Manana8407f552014-09-05 15:14:39 +01005230 0, LLONG_MAX, ctx);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04005231 if (ret)
5232 goto end_trans;
Chris Mason12fcfd22009-03-24 10:24:20 -04005233 }
Yan, Zheng76dda932009-09-21 16:00:26 -04005234 if (IS_ROOT(parent))
Chris Masone02119d2008-09-05 16:13:11 -04005235 break;
Chris Mason12fcfd22009-03-24 10:24:20 -04005236
Josef Bacik6a912212010-11-20 09:48:00 +00005237 parent = dget_parent(parent);
5238 dput(old_parent);
5239 old_parent = parent;
Chris Masone02119d2008-09-05 16:13:11 -04005240 }
Filipe Manana2f2ff0e2015-03-20 17:19:46 +00005241 if (log_dentries)
5242 ret = log_new_dir_dentries(trans, root, orig_inode, ctx);
5243 else
5244 ret = 0;
Yan, Zheng4a500fd2010-05-16 10:49:59 -04005245end_trans:
Josef Bacik6a912212010-11-20 09:48:00 +00005246 dput(old_parent);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04005247 if (ret < 0) {
Miao Xie995946d2014-04-02 19:51:06 +08005248 btrfs_set_log_full_commit(root->fs_info, trans);
Yan, Zheng4a500fd2010-05-16 10:49:59 -04005249 ret = 1;
5250 }
Miao Xie8b050d32014-02-20 18:08:58 +08005251
5252 if (ret)
5253 btrfs_remove_log_ctx(root, ctx);
Chris Mason12fcfd22009-03-24 10:24:20 -04005254 btrfs_end_log_trans(root);
5255end_no_trans:
5256 return ret;
Chris Masone02119d2008-09-05 16:13:11 -04005257}
5258
5259/*
5260 * it is not safe to log dentry if the chunk root has added new
5261 * chunks. This returns 0 if the dentry was logged, and 1 otherwise.
5262 * If this returns 1, you must commit the transaction to safely get your
5263 * data on disk.
5264 */
5265int btrfs_log_dentry_safe(struct btrfs_trans_handle *trans,
Miao Xie8b050d32014-02-20 18:08:58 +08005266 struct btrfs_root *root, struct dentry *dentry,
Filipe Manana49dae1b2014-09-06 22:34:39 +01005267 const loff_t start,
5268 const loff_t end,
Miao Xie8b050d32014-02-20 18:08:58 +08005269 struct btrfs_log_ctx *ctx)
Chris Masone02119d2008-09-05 16:13:11 -04005270{
Josef Bacik6a912212010-11-20 09:48:00 +00005271 struct dentry *parent = dget_parent(dentry);
5272 int ret;
5273
David Howells2b0143b2015-03-17 22:25:59 +00005274 ret = btrfs_log_inode_parent(trans, root, d_inode(dentry), parent,
Filipe Manana49dae1b2014-09-06 22:34:39 +01005275 start, end, 0, ctx);
Josef Bacik6a912212010-11-20 09:48:00 +00005276 dput(parent);
5277
5278 return ret;
Chris Masone02119d2008-09-05 16:13:11 -04005279}
5280
5281/*
5282 * should be called during mount to recover any replay any log trees
5283 * from the FS
5284 */
5285int btrfs_recover_log_trees(struct btrfs_root *log_root_tree)
5286{
5287 int ret;
5288 struct btrfs_path *path;
5289 struct btrfs_trans_handle *trans;
5290 struct btrfs_key key;
5291 struct btrfs_key found_key;
5292 struct btrfs_key tmp_key;
5293 struct btrfs_root *log;
5294 struct btrfs_fs_info *fs_info = log_root_tree->fs_info;
5295 struct walk_control wc = {
5296 .process_func = process_one_buffer,
5297 .stage = 0,
5298 };
5299
Chris Masone02119d2008-09-05 16:13:11 -04005300 path = btrfs_alloc_path();
Tsutomu Itohdb5b4932011-03-23 08:14:16 +00005301 if (!path)
5302 return -ENOMEM;
5303
5304 fs_info->log_root_recovering = 1;
Chris Masone02119d2008-09-05 16:13:11 -04005305
Yan, Zheng4a500fd2010-05-16 10:49:59 -04005306 trans = btrfs_start_transaction(fs_info->tree_root, 0);
Jeff Mahoney79787ea2012-03-12 16:03:00 +01005307 if (IS_ERR(trans)) {
5308 ret = PTR_ERR(trans);
5309 goto error;
5310 }
Chris Masone02119d2008-09-05 16:13:11 -04005311
5312 wc.trans = trans;
5313 wc.pin = 1;
5314
Tsutomu Itohdb5b4932011-03-23 08:14:16 +00005315 ret = walk_log_tree(trans, log_root_tree, &wc);
Jeff Mahoney79787ea2012-03-12 16:03:00 +01005316 if (ret) {
5317 btrfs_error(fs_info, ret, "Failed to pin buffers while "
5318 "recovering log root tree.");
5319 goto error;
5320 }
Chris Masone02119d2008-09-05 16:13:11 -04005321
5322again:
5323 key.objectid = BTRFS_TREE_LOG_OBJECTID;
5324 key.offset = (u64)-1;
David Sterba962a2982014-06-04 18:41:45 +02005325 key.type = BTRFS_ROOT_ITEM_KEY;
Chris Masone02119d2008-09-05 16:13:11 -04005326
Chris Masond3977122009-01-05 21:25:51 -05005327 while (1) {
Chris Masone02119d2008-09-05 16:13:11 -04005328 ret = btrfs_search_slot(NULL, log_root_tree, &key, path, 0, 0);
Jeff Mahoney79787ea2012-03-12 16:03:00 +01005329
5330 if (ret < 0) {
5331 btrfs_error(fs_info, ret,
5332 "Couldn't find tree log root.");
5333 goto error;
5334 }
Chris Masone02119d2008-09-05 16:13:11 -04005335 if (ret > 0) {
5336 if (path->slots[0] == 0)
5337 break;
5338 path->slots[0]--;
5339 }
5340 btrfs_item_key_to_cpu(path->nodes[0], &found_key,
5341 path->slots[0]);
David Sterbab3b4aa72011-04-21 01:20:15 +02005342 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -04005343 if (found_key.objectid != BTRFS_TREE_LOG_OBJECTID)
5344 break;
5345
Miao Xiecb517ea2013-05-15 07:48:19 +00005346 log = btrfs_read_fs_root(log_root_tree, &found_key);
Jeff Mahoney79787ea2012-03-12 16:03:00 +01005347 if (IS_ERR(log)) {
5348 ret = PTR_ERR(log);
5349 btrfs_error(fs_info, ret,
5350 "Couldn't read tree log root.");
5351 goto error;
5352 }
Chris Masone02119d2008-09-05 16:13:11 -04005353
5354 tmp_key.objectid = found_key.offset;
5355 tmp_key.type = BTRFS_ROOT_ITEM_KEY;
5356 tmp_key.offset = (u64)-1;
5357
5358 wc.replay_dest = btrfs_read_fs_root_no_name(fs_info, &tmp_key);
Jeff Mahoney79787ea2012-03-12 16:03:00 +01005359 if (IS_ERR(wc.replay_dest)) {
5360 ret = PTR_ERR(wc.replay_dest);
Josef Bacikb50c6e22013-04-25 15:55:30 -04005361 free_extent_buffer(log->node);
5362 free_extent_buffer(log->commit_root);
5363 kfree(log);
Jeff Mahoney79787ea2012-03-12 16:03:00 +01005364 btrfs_error(fs_info, ret, "Couldn't read target root "
5365 "for tree log recovery.");
5366 goto error;
5367 }
Chris Masone02119d2008-09-05 16:13:11 -04005368
Yan Zheng07d400a2009-01-06 11:42:00 -05005369 wc.replay_dest->log_root = log;
Yan Zheng5d4f98a2009-06-10 10:45:14 -04005370 btrfs_record_root_in_trans(trans, wc.replay_dest);
Chris Masone02119d2008-09-05 16:13:11 -04005371 ret = walk_log_tree(trans, log, &wc);
Chris Masone02119d2008-09-05 16:13:11 -04005372
Josef Bacikb50c6e22013-04-25 15:55:30 -04005373 if (!ret && wc.stage == LOG_WALK_REPLAY_ALL) {
Chris Masone02119d2008-09-05 16:13:11 -04005374 ret = fixup_inode_link_counts(trans, wc.replay_dest,
5375 path);
Chris Masone02119d2008-09-05 16:13:11 -04005376 }
Chris Masone02119d2008-09-05 16:13:11 -04005377
5378 key.offset = found_key.offset - 1;
Yan Zheng07d400a2009-01-06 11:42:00 -05005379 wc.replay_dest->log_root = NULL;
Chris Masone02119d2008-09-05 16:13:11 -04005380 free_extent_buffer(log->node);
Chris Masonb263c2c2009-06-11 11:24:47 -04005381 free_extent_buffer(log->commit_root);
Chris Masone02119d2008-09-05 16:13:11 -04005382 kfree(log);
5383
Josef Bacikb50c6e22013-04-25 15:55:30 -04005384 if (ret)
5385 goto error;
5386
Chris Masone02119d2008-09-05 16:13:11 -04005387 if (found_key.offset == 0)
5388 break;
5389 }
David Sterbab3b4aa72011-04-21 01:20:15 +02005390 btrfs_release_path(path);
Chris Masone02119d2008-09-05 16:13:11 -04005391
5392 /* step one is to pin it all, step two is to replay just inodes */
5393 if (wc.pin) {
5394 wc.pin = 0;
5395 wc.process_func = replay_one_buffer;
5396 wc.stage = LOG_WALK_REPLAY_INODES;
5397 goto again;
5398 }
5399 /* step three is to replay everything */
5400 if (wc.stage < LOG_WALK_REPLAY_ALL) {
5401 wc.stage++;
5402 goto again;
5403 }
5404
5405 btrfs_free_path(path);
5406
Josef Bacikabefa552013-04-24 16:40:05 -04005407 /* step 4: commit the transaction, which also unpins the blocks */
5408 ret = btrfs_commit_transaction(trans, fs_info->tree_root);
5409 if (ret)
5410 return ret;
5411
Chris Masone02119d2008-09-05 16:13:11 -04005412 free_extent_buffer(log_root_tree->node);
5413 log_root_tree->log_root = NULL;
5414 fs_info->log_root_recovering = 0;
Chris Masone02119d2008-09-05 16:13:11 -04005415 kfree(log_root_tree);
Jeff Mahoney79787ea2012-03-12 16:03:00 +01005416
Josef Bacikabefa552013-04-24 16:40:05 -04005417 return 0;
Jeff Mahoney79787ea2012-03-12 16:03:00 +01005418error:
Josef Bacikb50c6e22013-04-25 15:55:30 -04005419 if (wc.trans)
5420 btrfs_end_transaction(wc.trans, fs_info->tree_root);
Jeff Mahoney79787ea2012-03-12 16:03:00 +01005421 btrfs_free_path(path);
5422 return ret;
Chris Masone02119d2008-09-05 16:13:11 -04005423}
Chris Mason12fcfd22009-03-24 10:24:20 -04005424
5425/*
5426 * there are some corner cases where we want to force a full
5427 * commit instead of allowing a directory to be logged.
5428 *
5429 * They revolve around files there were unlinked from the directory, and
5430 * this function updates the parent directory so that a full commit is
5431 * properly done if it is fsync'd later after the unlinks are done.
5432 */
5433void btrfs_record_unlink_dir(struct btrfs_trans_handle *trans,
5434 struct inode *dir, struct inode *inode,
5435 int for_rename)
5436{
5437 /*
Chris Masonaf4176b2009-03-24 10:24:31 -04005438 * when we're logging a file, if it hasn't been renamed
5439 * or unlinked, and its inode is fully committed on disk,
5440 * we don't have to worry about walking up the directory chain
5441 * to log its parents.
5442 *
5443 * So, we use the last_unlink_trans field to put this transid
5444 * into the file. When the file is logged we check it and
5445 * don't log the parents if the file is fully on disk.
5446 */
5447 if (S_ISREG(inode->i_mode))
5448 BTRFS_I(inode)->last_unlink_trans = trans->transid;
5449
5450 /*
Chris Mason12fcfd22009-03-24 10:24:20 -04005451 * if this directory was already logged any new
5452 * names for this file/dir will get recorded
5453 */
5454 smp_mb();
5455 if (BTRFS_I(dir)->logged_trans == trans->transid)
5456 return;
5457
5458 /*
5459 * if the inode we're about to unlink was logged,
5460 * the log will be properly updated for any new names
5461 */
5462 if (BTRFS_I(inode)->logged_trans == trans->transid)
5463 return;
5464
5465 /*
5466 * when renaming files across directories, if the directory
5467 * there we're unlinking from gets fsync'd later on, there's
5468 * no way to find the destination directory later and fsync it
5469 * properly. So, we have to be conservative and force commits
5470 * so the new name gets discovered.
5471 */
5472 if (for_rename)
5473 goto record;
5474
5475 /* we can safely do the unlink without any special recording */
5476 return;
5477
5478record:
5479 BTRFS_I(dir)->last_unlink_trans = trans->transid;
5480}
5481
5482/*
5483 * Call this after adding a new name for a file and it will properly
5484 * update the log to reflect the new name.
5485 *
5486 * It will return zero if all goes well, and it will return 1 if a
5487 * full transaction commit is required.
5488 */
5489int btrfs_log_new_name(struct btrfs_trans_handle *trans,
5490 struct inode *inode, struct inode *old_dir,
5491 struct dentry *parent)
5492{
5493 struct btrfs_root * root = BTRFS_I(inode)->root;
5494
5495 /*
Chris Masonaf4176b2009-03-24 10:24:31 -04005496 * this will force the logging code to walk the dentry chain
5497 * up for the file
5498 */
5499 if (S_ISREG(inode->i_mode))
5500 BTRFS_I(inode)->last_unlink_trans = trans->transid;
5501
5502 /*
Chris Mason12fcfd22009-03-24 10:24:20 -04005503 * if this inode hasn't been logged and directory we're renaming it
5504 * from hasn't been logged, we don't need to log it
5505 */
5506 if (BTRFS_I(inode)->logged_trans <=
5507 root->fs_info->last_trans_committed &&
5508 (!old_dir || BTRFS_I(old_dir)->logged_trans <=
5509 root->fs_info->last_trans_committed))
5510 return 0;
5511
Filipe Manana49dae1b2014-09-06 22:34:39 +01005512 return btrfs_log_inode_parent(trans, root, inode, parent, 0,
5513 LLONG_MAX, 1, NULL);
Chris Mason12fcfd22009-03-24 10:24:20 -04005514}
5515