Commit | Line | Data |
---|---|---|
c1d7c514 | 1 | // SPDX-License-Identifier: GPL-2.0 |
e02119d5 CM |
2 | /* |
3 | * Copyright (C) 2008 Oracle. All rights reserved. | |
e02119d5 CM |
4 | */ |
5 | ||
6 | #include <linux/sched.h> | |
5a0e3ad6 | 7 | #include <linux/slab.h> |
c6adc9cc | 8 | #include <linux/blkdev.h> |
5dc562c5 | 9 | #include <linux/list_sort.h> |
c7f88c4e | 10 | #include <linux/iversion.h> |
602cbe91 | 11 | #include "misc.h" |
9678c543 | 12 | #include "ctree.h" |
995946dd | 13 | #include "tree-log.h" |
e02119d5 CM |
14 | #include "disk-io.h" |
15 | #include "locking.h" | |
16 | #include "print-tree.h" | |
f186373f | 17 | #include "backref.h" |
ebb8765b | 18 | #include "compression.h" |
df2c95f3 | 19 | #include "qgroup.h" |
6787bb9f NB |
20 | #include "block-group.h" |
21 | #include "space-info.h" | |
d3575156 | 22 | #include "zoned.h" |
26c2c454 | 23 | #include "inode-item.h" |
e02119d5 CM |
24 | |
25 | /* magic values for the inode_only field in btrfs_log_inode: | |
26 | * | |
27 | * LOG_INODE_ALL means to log everything | |
28 | * LOG_INODE_EXISTS means to log just enough to recreate the inode | |
29 | * during log replay | |
30 | */ | |
e13976cf DS |
31 | enum { |
32 | LOG_INODE_ALL, | |
33 | LOG_INODE_EXISTS, | |
34 | LOG_OTHER_INODE, | |
35 | LOG_OTHER_INODE_ALL, | |
36 | }; | |
e02119d5 | 37 | |
12fcfd22 CM |
38 | /* |
39 | * directory trouble cases | |
40 | * | |
41 | * 1) on rename or unlink, if the inode being unlinked isn't in the fsync | |
42 | * log, we must force a full commit before doing an fsync of the directory | |
43 | * where the unlink was done. | |
44 | * ---> record transid of last unlink/rename per directory | |
45 | * | |
46 | * mkdir foo/some_dir | |
47 | * normal commit | |
48 | * rename foo/some_dir foo2/some_dir | |
49 | * mkdir foo/some_dir | |
50 | * fsync foo/some_dir/some_file | |
51 | * | |
52 | * The fsync above will unlink the original some_dir without recording | |
53 | * it in its new location (foo2). After a crash, some_dir will be gone | |
54 | * unless the fsync of some_file forces a full commit | |
55 | * | |
56 | * 2) we must log any new names for any file or dir that is in the fsync | |
57 | * log. ---> check inode while renaming/linking. | |
58 | * | |
59 | * 2a) we must log any new names for any file or dir during rename | |
60 | * when the directory they are being removed from was logged. | |
61 | * ---> check inode and old parent dir during rename | |
62 | * | |
63 | * 2a is actually the more important variant. With the extra logging | |
64 | * a crash might unlink the old name without recreating the new one | |
65 | * | |
66 | * 3) after a crash, we must go through any directories with a link count | |
67 | * of zero and redo the rm -rf | |
68 | * | |
69 | * mkdir f1/foo | |
70 | * normal commit | |
71 | * rm -rf f1/foo | |
72 | * fsync(f1) | |
73 | * | |
74 | * The directory f1 was fully removed from the FS, but fsync was never | |
75 | * called on f1, only its parent dir. After a crash the rm -rf must | |
76 | * be replayed. This must be able to recurse down the entire | |
77 | * directory tree. The inode link count fixup code takes care of the | |
78 | * ugly details. | |
79 | */ | |
80 | ||
e02119d5 CM |
81 | /* |
82 | * stages for the tree walking. The first | |
83 | * stage (0) is to only pin down the blocks we find | |
84 | * the second stage (1) is to make sure that all the inodes | |
85 | * we find in the log are created in the subvolume. | |
86 | * | |
87 | * The last stage is to deal with directories and links and extents | |
88 | * and all the other fun semantics | |
89 | */ | |
e13976cf DS |
90 | enum { |
91 | LOG_WALK_PIN_ONLY, | |
92 | LOG_WALK_REPLAY_INODES, | |
93 | LOG_WALK_REPLAY_DIR_INDEX, | |
94 | LOG_WALK_REPLAY_ALL, | |
95 | }; | |
e02119d5 | 96 | |
12fcfd22 | 97 | static int btrfs_log_inode(struct btrfs_trans_handle *trans, |
90d04510 | 98 | struct btrfs_inode *inode, |
49dae1bc | 99 | int inode_only, |
8407f553 | 100 | struct btrfs_log_ctx *ctx); |
ec051c0f YZ |
101 | static int link_to_fixup_dir(struct btrfs_trans_handle *trans, |
102 | struct btrfs_root *root, | |
103 | struct btrfs_path *path, u64 objectid); | |
12fcfd22 CM |
104 | static noinline int replay_dir_deletes(struct btrfs_trans_handle *trans, |
105 | struct btrfs_root *root, | |
106 | struct btrfs_root *log, | |
107 | struct btrfs_path *path, | |
108 | u64 dirid, int del_all); | |
fa1a0f42 | 109 | static void wait_log_commit(struct btrfs_root *root, int transid); |
e02119d5 CM |
110 | |
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 | ||
e02119d5 CM |
134 | /* |
135 | * 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 | */ | |
139 | static int start_log_trans(struct btrfs_trans_handle *trans, | |
8b050d35 MX |
140 | struct btrfs_root *root, |
141 | struct btrfs_log_ctx *ctx) | |
e02119d5 | 142 | { |
0b246afa | 143 | struct btrfs_fs_info *fs_info = root->fs_info; |
47876f7c | 144 | struct btrfs_root *tree_root = fs_info->tree_root; |
fa1a0f42 | 145 | const bool zoned = btrfs_is_zoned(fs_info); |
34eb2a52 | 146 | int ret = 0; |
fa1a0f42 | 147 | bool created = false; |
7237f183 | 148 | |
47876f7c FM |
149 | /* |
150 | * First check if the log root tree was already created. If not, create | |
151 | * it before locking the root's log_mutex, just to keep lockdep happy. | |
152 | */ | |
153 | if (!test_bit(BTRFS_ROOT_HAS_LOG_TREE, &tree_root->state)) { | |
154 | mutex_lock(&tree_root->log_mutex); | |
155 | if (!fs_info->log_root_tree) { | |
156 | ret = btrfs_init_log_root_tree(trans, fs_info); | |
fa1a0f42 | 157 | if (!ret) { |
47876f7c | 158 | set_bit(BTRFS_ROOT_HAS_LOG_TREE, &tree_root->state); |
fa1a0f42 NA |
159 | created = true; |
160 | } | |
47876f7c FM |
161 | } |
162 | mutex_unlock(&tree_root->log_mutex); | |
163 | if (ret) | |
164 | return ret; | |
165 | } | |
166 | ||
7237f183 | 167 | mutex_lock(&root->log_mutex); |
34eb2a52 | 168 | |
fa1a0f42 | 169 | again: |
7237f183 | 170 | if (root->log_root) { |
fa1a0f42 NA |
171 | int index = (root->log_transid + 1) % 2; |
172 | ||
4884b8e8 | 173 | if (btrfs_need_log_full_commit(trans)) { |
f31f09f6 | 174 | ret = BTRFS_LOG_FORCE_COMMIT; |
50471a38 MX |
175 | goto out; |
176 | } | |
34eb2a52 | 177 | |
fa1a0f42 NA |
178 | if (zoned && atomic_read(&root->log_commit[index])) { |
179 | wait_log_commit(root, root->log_transid - 1); | |
180 | goto again; | |
181 | } | |
182 | ||
ff782e0a | 183 | if (!root->log_start_pid) { |
27cdeb70 | 184 | clear_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state); |
34eb2a52 | 185 | root->log_start_pid = current->pid; |
ff782e0a | 186 | } else if (root->log_start_pid != current->pid) { |
27cdeb70 | 187 | set_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state); |
ff782e0a | 188 | } |
34eb2a52 | 189 | } else { |
fa1a0f42 NA |
190 | /* |
191 | * This means fs_info->log_root_tree was already created | |
192 | * for some other FS trees. Do the full commit not to mix | |
193 | * nodes from multiple log transactions to do sequential | |
194 | * writing. | |
195 | */ | |
196 | if (zoned && !created) { | |
f31f09f6 | 197 | ret = BTRFS_LOG_FORCE_COMMIT; |
fa1a0f42 NA |
198 | goto out; |
199 | } | |
200 | ||
e02119d5 | 201 | ret = btrfs_add_log_tree(trans, root); |
4a500fd1 | 202 | if (ret) |
e87ac136 | 203 | goto out; |
34eb2a52 | 204 | |
e7a79811 | 205 | set_bit(BTRFS_ROOT_HAS_LOG_TREE, &root->state); |
34eb2a52 Z |
206 | clear_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state); |
207 | root->log_start_pid = current->pid; | |
e02119d5 | 208 | } |
34eb2a52 | 209 | |
7237f183 | 210 | atomic_inc(&root->log_writers); |
289cffcb | 211 | if (!ctx->logging_new_name) { |
34eb2a52 | 212 | int index = root->log_transid % 2; |
8b050d35 | 213 | list_add_tail(&ctx->list, &root->log_ctxs[index]); |
d1433deb | 214 | ctx->log_transid = root->log_transid; |
8b050d35 | 215 | } |
34eb2a52 | 216 | |
e87ac136 | 217 | out: |
7237f183 | 218 | mutex_unlock(&root->log_mutex); |
e87ac136 | 219 | return ret; |
e02119d5 CM |
220 | } |
221 | ||
222 | /* | |
223 | * returns 0 if there was a log transaction running and we were able | |
224 | * to join, or returns -ENOENT if there were not transactions | |
225 | * in progress | |
226 | */ | |
227 | static int join_running_log_trans(struct btrfs_root *root) | |
228 | { | |
fa1a0f42 | 229 | const bool zoned = btrfs_is_zoned(root->fs_info); |
e02119d5 CM |
230 | int ret = -ENOENT; |
231 | ||
e7a79811 FM |
232 | if (!test_bit(BTRFS_ROOT_HAS_LOG_TREE, &root->state)) |
233 | return ret; | |
234 | ||
7237f183 | 235 | mutex_lock(&root->log_mutex); |
fa1a0f42 | 236 | again: |
e02119d5 | 237 | if (root->log_root) { |
fa1a0f42 NA |
238 | int index = (root->log_transid + 1) % 2; |
239 | ||
e02119d5 | 240 | ret = 0; |
fa1a0f42 NA |
241 | if (zoned && atomic_read(&root->log_commit[index])) { |
242 | wait_log_commit(root, root->log_transid - 1); | |
243 | goto again; | |
244 | } | |
7237f183 | 245 | atomic_inc(&root->log_writers); |
e02119d5 | 246 | } |
7237f183 | 247 | mutex_unlock(&root->log_mutex); |
e02119d5 CM |
248 | return ret; |
249 | } | |
250 | ||
12fcfd22 CM |
251 | /* |
252 | * This either makes the current running log transaction wait | |
253 | * until you call btrfs_end_log_trans() or it makes any future | |
254 | * log transactions wait until you call btrfs_end_log_trans() | |
255 | */ | |
45128b08 | 256 | void btrfs_pin_log_trans(struct btrfs_root *root) |
12fcfd22 | 257 | { |
12fcfd22 | 258 | atomic_inc(&root->log_writers); |
12fcfd22 CM |
259 | } |
260 | ||
e02119d5 CM |
261 | /* |
262 | * indicate we're done making changes to the log tree | |
263 | * and wake up anyone waiting to do a sync | |
264 | */ | |
143bede5 | 265 | void btrfs_end_log_trans(struct btrfs_root *root) |
e02119d5 | 266 | { |
7237f183 | 267 | if (atomic_dec_and_test(&root->log_writers)) { |
093258e6 DS |
268 | /* atomic_dec_and_test implies a barrier */ |
269 | cond_wake_up_nomb(&root->log_writer_wait); | |
7237f183 | 270 | } |
e02119d5 CM |
271 | } |
272 | ||
247462a5 DS |
273 | static void btrfs_wait_tree_block_writeback(struct extent_buffer *buf) |
274 | { | |
275 | filemap_fdatawait_range(buf->pages[0]->mapping, | |
276 | buf->start, buf->start + buf->len - 1); | |
277 | } | |
e02119d5 CM |
278 | |
279 | /* | |
280 | * the walk control struct is used to pass state down the chain when | |
281 | * processing the log tree. The stage field tells us which part | |
282 | * of the log tree processing we are currently doing. The others | |
283 | * are state fields used for that specific part | |
284 | */ | |
285 | struct walk_control { | |
286 | /* should we free the extent on disk when done? This is used | |
287 | * at transaction commit time while freeing a log tree | |
288 | */ | |
289 | int free; | |
290 | ||
e02119d5 CM |
291 | /* pin only walk, we record which extents on disk belong to the |
292 | * log trees | |
293 | */ | |
294 | int pin; | |
295 | ||
296 | /* what stage of the replay code we're currently in */ | |
297 | int stage; | |
298 | ||
f2d72f42 FM |
299 | /* |
300 | * Ignore any items from the inode currently being processed. Needs | |
301 | * to be set every time we find a BTRFS_INODE_ITEM_KEY and we are in | |
302 | * the LOG_WALK_REPLAY_INODES stage. | |
303 | */ | |
304 | bool ignore_cur_inode; | |
305 | ||
e02119d5 CM |
306 | /* the root we are currently replaying */ |
307 | struct btrfs_root *replay_dest; | |
308 | ||
309 | /* the trans handle for the current replay */ | |
310 | struct btrfs_trans_handle *trans; | |
311 | ||
312 | /* the function that gets used to process blocks we find in the | |
313 | * tree. Note the extent_buffer might not be up to date when it is | |
314 | * passed in, and it must be checked or read if you need the data | |
315 | * inside it | |
316 | */ | |
317 | int (*process_func)(struct btrfs_root *log, struct extent_buffer *eb, | |
581c1760 | 318 | struct walk_control *wc, u64 gen, int level); |
e02119d5 CM |
319 | }; |
320 | ||
321 | /* | |
322 | * process_func used to pin down extents, write them or wait on them | |
323 | */ | |
324 | static int process_one_buffer(struct btrfs_root *log, | |
325 | struct extent_buffer *eb, | |
581c1760 | 326 | struct walk_control *wc, u64 gen, int level) |
e02119d5 | 327 | { |
0b246afa | 328 | struct btrfs_fs_info *fs_info = log->fs_info; |
b50c6e25 JB |
329 | int ret = 0; |
330 | ||
8c2a1a30 JB |
331 | /* |
332 | * If this fs is mixed then we need to be able to process the leaves to | |
333 | * pin down any logged extents, so we have to read the block. | |
334 | */ | |
0b246afa | 335 | if (btrfs_fs_incompat(fs_info, MIXED_GROUPS)) { |
6a2e9dc4 | 336 | ret = btrfs_read_extent_buffer(eb, gen, level, NULL); |
8c2a1a30 JB |
337 | if (ret) |
338 | return ret; | |
339 | } | |
340 | ||
c816d705 | 341 | if (wc->pin) { |
9fce5704 | 342 | ret = btrfs_pin_extent_for_log_replay(wc->trans, eb->start, |
2ff7e61e | 343 | eb->len); |
c816d705 FM |
344 | if (ret) |
345 | return ret; | |
e02119d5 | 346 | |
c816d705 FM |
347 | if (btrfs_buffer_uptodate(eb, gen, 0) && |
348 | btrfs_header_level(eb) == 0) | |
bcdc428c | 349 | ret = btrfs_exclude_logged_extents(eb); |
e02119d5 | 350 | } |
b50c6e25 | 351 | return ret; |
e02119d5 CM |
352 | } |
353 | ||
086dcbfa FM |
354 | static int do_overwrite_item(struct btrfs_trans_handle *trans, |
355 | struct btrfs_root *root, | |
356 | struct btrfs_path *path, | |
357 | struct extent_buffer *eb, int slot, | |
358 | struct btrfs_key *key) | |
e02119d5 CM |
359 | { |
360 | int ret; | |
361 | u32 item_size; | |
362 | u64 saved_i_size = 0; | |
363 | int save_old_i_size = 0; | |
364 | unsigned long src_ptr; | |
365 | unsigned long dst_ptr; | |
366 | int overwrite_root = 0; | |
4bc4bee4 | 367 | bool inode_item = key->type == BTRFS_INODE_ITEM_KEY; |
e02119d5 CM |
368 | |
369 | if (root->root_key.objectid != BTRFS_TREE_LOG_OBJECTID) | |
370 | overwrite_root = 1; | |
371 | ||
3212fa14 | 372 | item_size = btrfs_item_size(eb, slot); |
e02119d5 CM |
373 | src_ptr = btrfs_item_ptr_offset(eb, slot); |
374 | ||
086dcbfa FM |
375 | /* Our caller must have done a search for the key for us. */ |
376 | ASSERT(path->nodes[0] != NULL); | |
377 | ||
378 | /* | |
379 | * And the slot must point to the exact key or the slot where the key | |
380 | * should be at (the first item with a key greater than 'key') | |
381 | */ | |
382 | if (path->slots[0] < btrfs_header_nritems(path->nodes[0])) { | |
383 | struct btrfs_key found_key; | |
384 | ||
385 | btrfs_item_key_to_cpu(path->nodes[0], &found_key, path->slots[0]); | |
386 | ret = btrfs_comp_cpu_keys(&found_key, key); | |
387 | ASSERT(ret >= 0); | |
388 | } else { | |
389 | ret = 1; | |
390 | } | |
4bc4bee4 | 391 | |
e02119d5 CM |
392 | if (ret == 0) { |
393 | char *src_copy; | |
394 | char *dst_copy; | |
3212fa14 | 395 | u32 dst_size = btrfs_item_size(path->nodes[0], |
e02119d5 CM |
396 | path->slots[0]); |
397 | if (dst_size != item_size) | |
398 | goto insert; | |
399 | ||
400 | if (item_size == 0) { | |
b3b4aa74 | 401 | btrfs_release_path(path); |
e02119d5 CM |
402 | return 0; |
403 | } | |
404 | dst_copy = kmalloc(item_size, GFP_NOFS); | |
405 | src_copy = kmalloc(item_size, GFP_NOFS); | |
2a29edc6 | 406 | if (!dst_copy || !src_copy) { |
b3b4aa74 | 407 | btrfs_release_path(path); |
2a29edc6 | 408 | kfree(dst_copy); |
409 | kfree(src_copy); | |
410 | return -ENOMEM; | |
411 | } | |
e02119d5 CM |
412 | |
413 | read_extent_buffer(eb, src_copy, src_ptr, item_size); | |
414 | ||
415 | dst_ptr = btrfs_item_ptr_offset(path->nodes[0], path->slots[0]); | |
416 | read_extent_buffer(path->nodes[0], dst_copy, dst_ptr, | |
417 | item_size); | |
418 | ret = memcmp(dst_copy, src_copy, item_size); | |
419 | ||
420 | kfree(dst_copy); | |
421 | kfree(src_copy); | |
422 | /* | |
423 | * they have the same contents, just return, this saves | |
424 | * us from cowing blocks in the destination tree and doing | |
425 | * extra writes that may not have been done by a previous | |
426 | * sync | |
427 | */ | |
428 | if (ret == 0) { | |
b3b4aa74 | 429 | btrfs_release_path(path); |
e02119d5 CM |
430 | return 0; |
431 | } | |
432 | ||
4bc4bee4 JB |
433 | /* |
434 | * We need to load the old nbytes into the inode so when we | |
435 | * replay the extents we've logged we get the right nbytes. | |
436 | */ | |
437 | if (inode_item) { | |
438 | struct btrfs_inode_item *item; | |
439 | u64 nbytes; | |
d555438b | 440 | u32 mode; |
4bc4bee4 JB |
441 | |
442 | item = btrfs_item_ptr(path->nodes[0], path->slots[0], | |
443 | struct btrfs_inode_item); | |
444 | nbytes = btrfs_inode_nbytes(path->nodes[0], item); | |
445 | item = btrfs_item_ptr(eb, slot, | |
446 | struct btrfs_inode_item); | |
447 | btrfs_set_inode_nbytes(eb, item, nbytes); | |
d555438b JB |
448 | |
449 | /* | |
450 | * If this is a directory we need to reset the i_size to | |
451 | * 0 so that we can set it up properly when replaying | |
452 | * the rest of the items in this log. | |
453 | */ | |
454 | mode = btrfs_inode_mode(eb, item); | |
455 | if (S_ISDIR(mode)) | |
456 | btrfs_set_inode_size(eb, item, 0); | |
4bc4bee4 JB |
457 | } |
458 | } else if (inode_item) { | |
459 | struct btrfs_inode_item *item; | |
d555438b | 460 | u32 mode; |
4bc4bee4 JB |
461 | |
462 | /* | |
463 | * New inode, set nbytes to 0 so that the nbytes comes out | |
464 | * properly when we replay the extents. | |
465 | */ | |
466 | item = btrfs_item_ptr(eb, slot, struct btrfs_inode_item); | |
467 | btrfs_set_inode_nbytes(eb, item, 0); | |
d555438b JB |
468 | |
469 | /* | |
470 | * If this is a directory we need to reset the i_size to 0 so | |
471 | * that we can set it up properly when replaying the rest of | |
472 | * the items in this log. | |
473 | */ | |
474 | mode = btrfs_inode_mode(eb, item); | |
475 | if (S_ISDIR(mode)) | |
476 | btrfs_set_inode_size(eb, item, 0); | |
e02119d5 CM |
477 | } |
478 | insert: | |
b3b4aa74 | 479 | btrfs_release_path(path); |
e02119d5 | 480 | /* try to insert the key into the destination tree */ |
df8d116f | 481 | path->skip_release_on_error = 1; |
e02119d5 CM |
482 | ret = btrfs_insert_empty_item(trans, root, path, |
483 | key, item_size); | |
df8d116f | 484 | path->skip_release_on_error = 0; |
e02119d5 CM |
485 | |
486 | /* make sure any existing item is the correct size */ | |
df8d116f | 487 | if (ret == -EEXIST || ret == -EOVERFLOW) { |
e02119d5 | 488 | u32 found_size; |
3212fa14 | 489 | found_size = btrfs_item_size(path->nodes[0], |
e02119d5 | 490 | path->slots[0]); |
143bede5 | 491 | if (found_size > item_size) |
78ac4f9e | 492 | btrfs_truncate_item(path, item_size, 1); |
143bede5 | 493 | else if (found_size < item_size) |
c71dd880 | 494 | btrfs_extend_item(path, item_size - found_size); |
e02119d5 | 495 | } else if (ret) { |
4a500fd1 | 496 | return ret; |
e02119d5 CM |
497 | } |
498 | dst_ptr = btrfs_item_ptr_offset(path->nodes[0], | |
499 | path->slots[0]); | |
500 | ||
501 | /* don't overwrite an existing inode if the generation number | |
502 | * was logged as zero. This is done when the tree logging code | |
503 | * is just logging an inode to make sure it exists after recovery. | |
504 | * | |
505 | * Also, don't overwrite i_size on directories during replay. | |
506 | * log replay inserts and removes directory items based on the | |
507 | * state of the tree found in the subvolume, and i_size is modified | |
508 | * as it goes | |
509 | */ | |
510 | if (key->type == BTRFS_INODE_ITEM_KEY && ret == -EEXIST) { | |
511 | struct btrfs_inode_item *src_item; | |
512 | struct btrfs_inode_item *dst_item; | |
513 | ||
514 | src_item = (struct btrfs_inode_item *)src_ptr; | |
515 | dst_item = (struct btrfs_inode_item *)dst_ptr; | |
516 | ||
1a4bcf47 FM |
517 | if (btrfs_inode_generation(eb, src_item) == 0) { |
518 | struct extent_buffer *dst_eb = path->nodes[0]; | |
2f2ff0ee | 519 | const u64 ino_size = btrfs_inode_size(eb, src_item); |
1a4bcf47 | 520 | |
2f2ff0ee FM |
521 | /* |
522 | * For regular files an ino_size == 0 is used only when | |
523 | * logging that an inode exists, as part of a directory | |
524 | * fsync, and the inode wasn't fsynced before. In this | |
525 | * case don't set the size of the inode in the fs/subvol | |
526 | * tree, otherwise we would be throwing valid data away. | |
527 | */ | |
1a4bcf47 | 528 | if (S_ISREG(btrfs_inode_mode(eb, src_item)) && |
2f2ff0ee | 529 | S_ISREG(btrfs_inode_mode(dst_eb, dst_item)) && |
60d48e2e DS |
530 | ino_size != 0) |
531 | btrfs_set_inode_size(dst_eb, dst_item, ino_size); | |
e02119d5 | 532 | goto no_copy; |
1a4bcf47 | 533 | } |
e02119d5 CM |
534 | |
535 | if (overwrite_root && | |
536 | S_ISDIR(btrfs_inode_mode(eb, src_item)) && | |
537 | S_ISDIR(btrfs_inode_mode(path->nodes[0], dst_item))) { | |
538 | save_old_i_size = 1; | |
539 | saved_i_size = btrfs_inode_size(path->nodes[0], | |
540 | dst_item); | |
541 | } | |
542 | } | |
543 | ||
544 | copy_extent_buffer(path->nodes[0], eb, dst_ptr, | |
545 | src_ptr, item_size); | |
546 | ||
547 | if (save_old_i_size) { | |
548 | struct btrfs_inode_item *dst_item; | |
549 | dst_item = (struct btrfs_inode_item *)dst_ptr; | |
550 | btrfs_set_inode_size(path->nodes[0], dst_item, saved_i_size); | |
551 | } | |
552 | ||
553 | /* make sure the generation is filled in */ | |
554 | if (key->type == BTRFS_INODE_ITEM_KEY) { | |
555 | struct btrfs_inode_item *dst_item; | |
556 | dst_item = (struct btrfs_inode_item *)dst_ptr; | |
557 | if (btrfs_inode_generation(path->nodes[0], dst_item) == 0) { | |
558 | btrfs_set_inode_generation(path->nodes[0], dst_item, | |
559 | trans->transid); | |
560 | } | |
561 | } | |
562 | no_copy: | |
563 | btrfs_mark_buffer_dirty(path->nodes[0]); | |
b3b4aa74 | 564 | btrfs_release_path(path); |
e02119d5 CM |
565 | return 0; |
566 | } | |
567 | ||
086dcbfa FM |
568 | /* |
569 | * Item overwrite used by replay and tree logging. eb, slot and key all refer | |
570 | * to the src data we are copying out. | |
571 | * | |
572 | * root is the tree we are copying into, and path is a scratch | |
573 | * path for use in this function (it should be released on entry and | |
574 | * will be released on exit). | |
575 | * | |
576 | * If the key is already in the destination tree the existing item is | |
577 | * overwritten. If the existing item isn't big enough, it is extended. | |
578 | * If it is too large, it is truncated. | |
579 | * | |
580 | * If the key isn't in the destination yet, a new item is inserted. | |
581 | */ | |
582 | static int overwrite_item(struct btrfs_trans_handle *trans, | |
583 | struct btrfs_root *root, | |
584 | struct btrfs_path *path, | |
585 | struct extent_buffer *eb, int slot, | |
586 | struct btrfs_key *key) | |
587 | { | |
588 | int ret; | |
589 | ||
590 | /* Look for the key in the destination tree. */ | |
591 | ret = btrfs_search_slot(NULL, root, key, path, 0, 0); | |
592 | if (ret < 0) | |
593 | return ret; | |
594 | ||
595 | return do_overwrite_item(trans, root, path, eb, slot, key); | |
596 | } | |
597 | ||
e02119d5 CM |
598 | /* |
599 | * simple helper to read an inode off the disk from a given root | |
600 | * This can only be called for subvolume roots and not for the log | |
601 | */ | |
602 | static noinline struct inode *read_one_inode(struct btrfs_root *root, | |
603 | u64 objectid) | |
604 | { | |
605 | struct inode *inode; | |
e02119d5 | 606 | |
0202e83f | 607 | inode = btrfs_iget(root->fs_info->sb, objectid, root); |
2e19f1f9 | 608 | if (IS_ERR(inode)) |
5d4f98a2 | 609 | inode = NULL; |
e02119d5 CM |
610 | return inode; |
611 | } | |
612 | ||
613 | /* replays a single extent in 'eb' at 'slot' with 'key' into the | |
614 | * subvolume 'root'. path is released on entry and should be released | |
615 | * on exit. | |
616 | * | |
617 | * extents in the log tree have not been allocated out of the extent | |
618 | * tree yet. So, this completes the allocation, taking a reference | |
619 | * as required if the extent already exists or creating a new extent | |
620 | * if it isn't in the extent allocation tree yet. | |
621 | * | |
622 | * The extent is inserted into the file, dropping any existing extents | |
623 | * from the file that overlap the new one. | |
624 | */ | |
625 | static noinline int replay_one_extent(struct btrfs_trans_handle *trans, | |
626 | struct btrfs_root *root, | |
627 | struct btrfs_path *path, | |
628 | struct extent_buffer *eb, int slot, | |
629 | struct btrfs_key *key) | |
630 | { | |
5893dfb9 | 631 | struct btrfs_drop_extents_args drop_args = { 0 }; |
0b246afa | 632 | struct btrfs_fs_info *fs_info = root->fs_info; |
e02119d5 | 633 | int found_type; |
e02119d5 | 634 | u64 extent_end; |
e02119d5 | 635 | u64 start = key->offset; |
4bc4bee4 | 636 | u64 nbytes = 0; |
e02119d5 CM |
637 | struct btrfs_file_extent_item *item; |
638 | struct inode *inode = NULL; | |
639 | unsigned long size; | |
640 | int ret = 0; | |
641 | ||
642 | item = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item); | |
643 | found_type = btrfs_file_extent_type(eb, item); | |
644 | ||
d899e052 | 645 | if (found_type == BTRFS_FILE_EXTENT_REG || |
4bc4bee4 JB |
646 | found_type == BTRFS_FILE_EXTENT_PREALLOC) { |
647 | nbytes = btrfs_file_extent_num_bytes(eb, item); | |
648 | extent_end = start + nbytes; | |
649 | ||
650 | /* | |
651 | * We don't add to the inodes nbytes if we are prealloc or a | |
652 | * hole. | |
653 | */ | |
654 | if (btrfs_file_extent_disk_bytenr(eb, item) == 0) | |
655 | nbytes = 0; | |
656 | } else if (found_type == BTRFS_FILE_EXTENT_INLINE) { | |
e41ca589 | 657 | size = btrfs_file_extent_ram_bytes(eb, item); |
4bc4bee4 | 658 | nbytes = btrfs_file_extent_ram_bytes(eb, item); |
da17066c | 659 | extent_end = ALIGN(start + size, |
0b246afa | 660 | fs_info->sectorsize); |
e02119d5 CM |
661 | } else { |
662 | ret = 0; | |
663 | goto out; | |
664 | } | |
665 | ||
666 | inode = read_one_inode(root, key->objectid); | |
667 | if (!inode) { | |
668 | ret = -EIO; | |
669 | goto out; | |
670 | } | |
671 | ||
672 | /* | |
673 | * first check to see if we already have this extent in the | |
674 | * file. This must be done before the btrfs_drop_extents run | |
675 | * so we don't try to drop this extent. | |
676 | */ | |
f85b7379 DS |
677 | ret = btrfs_lookup_file_extent(trans, root, path, |
678 | btrfs_ino(BTRFS_I(inode)), start, 0); | |
e02119d5 | 679 | |
d899e052 YZ |
680 | if (ret == 0 && |
681 | (found_type == BTRFS_FILE_EXTENT_REG || | |
682 | found_type == BTRFS_FILE_EXTENT_PREALLOC)) { | |
e02119d5 CM |
683 | struct btrfs_file_extent_item cmp1; |
684 | struct btrfs_file_extent_item cmp2; | |
685 | struct btrfs_file_extent_item *existing; | |
686 | struct extent_buffer *leaf; | |
687 | ||
688 | leaf = path->nodes[0]; | |
689 | existing = btrfs_item_ptr(leaf, path->slots[0], | |
690 | struct btrfs_file_extent_item); | |
691 | ||
692 | read_extent_buffer(eb, &cmp1, (unsigned long)item, | |
693 | sizeof(cmp1)); | |
694 | read_extent_buffer(leaf, &cmp2, (unsigned long)existing, | |
695 | sizeof(cmp2)); | |
696 | ||
697 | /* | |
698 | * we already have a pointer to this exact extent, | |
699 | * we don't have to do anything | |
700 | */ | |
701 | if (memcmp(&cmp1, &cmp2, sizeof(cmp1)) == 0) { | |
b3b4aa74 | 702 | btrfs_release_path(path); |
e02119d5 CM |
703 | goto out; |
704 | } | |
705 | } | |
b3b4aa74 | 706 | btrfs_release_path(path); |
e02119d5 CM |
707 | |
708 | /* drop any overlapping extents */ | |
5893dfb9 FM |
709 | drop_args.start = start; |
710 | drop_args.end = extent_end; | |
711 | drop_args.drop_cache = true; | |
712 | ret = btrfs_drop_extents(trans, root, BTRFS_I(inode), &drop_args); | |
3650860b JB |
713 | if (ret) |
714 | goto out; | |
e02119d5 | 715 | |
07d400a6 YZ |
716 | if (found_type == BTRFS_FILE_EXTENT_REG || |
717 | found_type == BTRFS_FILE_EXTENT_PREALLOC) { | |
5d4f98a2 | 718 | u64 offset; |
07d400a6 YZ |
719 | unsigned long dest_offset; |
720 | struct btrfs_key ins; | |
721 | ||
3168021c FM |
722 | if (btrfs_file_extent_disk_bytenr(eb, item) == 0 && |
723 | btrfs_fs_incompat(fs_info, NO_HOLES)) | |
724 | goto update_inode; | |
725 | ||
07d400a6 YZ |
726 | ret = btrfs_insert_empty_item(trans, root, path, key, |
727 | sizeof(*item)); | |
3650860b JB |
728 | if (ret) |
729 | goto out; | |
07d400a6 YZ |
730 | dest_offset = btrfs_item_ptr_offset(path->nodes[0], |
731 | path->slots[0]); | |
732 | copy_extent_buffer(path->nodes[0], eb, dest_offset, | |
733 | (unsigned long)item, sizeof(*item)); | |
734 | ||
735 | ins.objectid = btrfs_file_extent_disk_bytenr(eb, item); | |
736 | ins.offset = btrfs_file_extent_disk_num_bytes(eb, item); | |
737 | ins.type = BTRFS_EXTENT_ITEM_KEY; | |
5d4f98a2 | 738 | offset = key->offset - btrfs_file_extent_offset(eb, item); |
07d400a6 | 739 | |
df2c95f3 QW |
740 | /* |
741 | * Manually record dirty extent, as here we did a shallow | |
742 | * file extent item copy and skip normal backref update, | |
743 | * but modifying extent tree all by ourselves. | |
744 | * So need to manually record dirty extent for qgroup, | |
745 | * as the owner of the file extent changed from log tree | |
746 | * (doesn't affect qgroup) to fs/file tree(affects qgroup) | |
747 | */ | |
a95f3aaf | 748 | ret = btrfs_qgroup_trace_extent(trans, |
df2c95f3 QW |
749 | btrfs_file_extent_disk_bytenr(eb, item), |
750 | btrfs_file_extent_disk_num_bytes(eb, item), | |
751 | GFP_NOFS); | |
752 | if (ret < 0) | |
753 | goto out; | |
754 | ||
07d400a6 | 755 | if (ins.objectid > 0) { |
82fa113f | 756 | struct btrfs_ref ref = { 0 }; |
07d400a6 YZ |
757 | u64 csum_start; |
758 | u64 csum_end; | |
759 | LIST_HEAD(ordered_sums); | |
82fa113f | 760 | |
07d400a6 YZ |
761 | /* |
762 | * is this extent already allocated in the extent | |
763 | * allocation tree? If so, just add a reference | |
764 | */ | |
2ff7e61e | 765 | ret = btrfs_lookup_data_extent(fs_info, ins.objectid, |
07d400a6 | 766 | ins.offset); |
3736127a MPS |
767 | if (ret < 0) { |
768 | goto out; | |
769 | } else if (ret == 0) { | |
82fa113f QW |
770 | btrfs_init_generic_ref(&ref, |
771 | BTRFS_ADD_DELAYED_REF, | |
772 | ins.objectid, ins.offset, 0); | |
773 | btrfs_init_data_ref(&ref, | |
774 | root->root_key.objectid, | |
f42c5da6 | 775 | key->objectid, offset, 0, false); |
82fa113f | 776 | ret = btrfs_inc_extent_ref(trans, &ref); |
b50c6e25 JB |
777 | if (ret) |
778 | goto out; | |
07d400a6 YZ |
779 | } else { |
780 | /* | |
781 | * insert the extent pointer in the extent | |
782 | * allocation tree | |
783 | */ | |
5d4f98a2 | 784 | ret = btrfs_alloc_logged_file_extent(trans, |
2ff7e61e | 785 | root->root_key.objectid, |
5d4f98a2 | 786 | key->objectid, offset, &ins); |
b50c6e25 JB |
787 | if (ret) |
788 | goto out; | |
07d400a6 | 789 | } |
b3b4aa74 | 790 | btrfs_release_path(path); |
07d400a6 YZ |
791 | |
792 | if (btrfs_file_extent_compression(eb, item)) { | |
793 | csum_start = ins.objectid; | |
794 | csum_end = csum_start + ins.offset; | |
795 | } else { | |
796 | csum_start = ins.objectid + | |
797 | btrfs_file_extent_offset(eb, item); | |
798 | csum_end = csum_start + | |
799 | btrfs_file_extent_num_bytes(eb, item); | |
800 | } | |
801 | ||
802 | ret = btrfs_lookup_csums_range(root->log_root, | |
803 | csum_start, csum_end - 1, | |
a2de733c | 804 | &ordered_sums, 0); |
3650860b JB |
805 | if (ret) |
806 | goto out; | |
b84b8390 FM |
807 | /* |
808 | * Now delete all existing cums in the csum root that | |
809 | * cover our range. We do this because we can have an | |
810 | * extent that is completely referenced by one file | |
811 | * extent item and partially referenced by another | |
812 | * file extent item (like after using the clone or | |
813 | * extent_same ioctls). In this case if we end up doing | |
814 | * the replay of the one that partially references the | |
815 | * extent first, and we do not do the csum deletion | |
816 | * below, we can get 2 csum items in the csum tree that | |
817 | * overlap each other. For example, imagine our log has | |
818 | * the two following file extent items: | |
819 | * | |
820 | * key (257 EXTENT_DATA 409600) | |
821 | * extent data disk byte 12845056 nr 102400 | |
822 | * extent data offset 20480 nr 20480 ram 102400 | |
823 | * | |
824 | * key (257 EXTENT_DATA 819200) | |
825 | * extent data disk byte 12845056 nr 102400 | |
826 | * extent data offset 0 nr 102400 ram 102400 | |
827 | * | |
828 | * Where the second one fully references the 100K extent | |
829 | * that starts at disk byte 12845056, and the log tree | |
830 | * has a single csum item that covers the entire range | |
831 | * of the extent: | |
832 | * | |
833 | * key (EXTENT_CSUM EXTENT_CSUM 12845056) itemsize 100 | |
834 | * | |
835 | * After the first file extent item is replayed, the | |
836 | * csum tree gets the following csum item: | |
837 | * | |
838 | * key (EXTENT_CSUM EXTENT_CSUM 12865536) itemsize 20 | |
839 | * | |
840 | * Which covers the 20K sub-range starting at offset 20K | |
841 | * of our extent. Now when we replay the second file | |
842 | * extent item, if we do not delete existing csum items | |
843 | * that cover any of its blocks, we end up getting two | |
844 | * csum items in our csum tree that overlap each other: | |
845 | * | |
846 | * key (EXTENT_CSUM EXTENT_CSUM 12845056) itemsize 100 | |
847 | * key (EXTENT_CSUM EXTENT_CSUM 12865536) itemsize 20 | |
848 | * | |
849 | * Which is a problem, because after this anyone trying | |
850 | * to lookup up for the checksum of any block of our | |
851 | * extent starting at an offset of 40K or higher, will | |
852 | * end up looking at the second csum item only, which | |
853 | * does not contain the checksum for any block starting | |
854 | * at offset 40K or higher of our extent. | |
855 | */ | |
07d400a6 YZ |
856 | while (!list_empty(&ordered_sums)) { |
857 | struct btrfs_ordered_sum *sums; | |
fc28b25e JB |
858 | struct btrfs_root *csum_root; |
859 | ||
07d400a6 YZ |
860 | sums = list_entry(ordered_sums.next, |
861 | struct btrfs_ordered_sum, | |
862 | list); | |
fc28b25e JB |
863 | csum_root = btrfs_csum_root(fs_info, |
864 | sums->bytenr); | |
b84b8390 | 865 | if (!ret) |
fc28b25e | 866 | ret = btrfs_del_csums(trans, csum_root, |
5b4aacef JM |
867 | sums->bytenr, |
868 | sums->len); | |
3650860b JB |
869 | if (!ret) |
870 | ret = btrfs_csum_file_blocks(trans, | |
fc28b25e JB |
871 | csum_root, |
872 | sums); | |
07d400a6 YZ |
873 | list_del(&sums->list); |
874 | kfree(sums); | |
875 | } | |
3650860b JB |
876 | if (ret) |
877 | goto out; | |
07d400a6 | 878 | } else { |
b3b4aa74 | 879 | btrfs_release_path(path); |
07d400a6 YZ |
880 | } |
881 | } else if (found_type == BTRFS_FILE_EXTENT_INLINE) { | |
882 | /* inline extents are easy, we just overwrite them */ | |
883 | ret = overwrite_item(trans, root, path, eb, slot, key); | |
3650860b JB |
884 | if (ret) |
885 | goto out; | |
07d400a6 | 886 | } |
e02119d5 | 887 | |
9ddc959e JB |
888 | ret = btrfs_inode_set_file_extent_range(BTRFS_I(inode), start, |
889 | extent_end - start); | |
890 | if (ret) | |
891 | goto out; | |
892 | ||
3168021c | 893 | update_inode: |
2766ff61 | 894 | btrfs_update_inode_bytes(BTRFS_I(inode), nbytes, drop_args.bytes_found); |
9a56fcd1 | 895 | ret = btrfs_update_inode(trans, root, BTRFS_I(inode)); |
e02119d5 | 896 | out: |
8aa1e49e | 897 | iput(inode); |
e02119d5 CM |
898 | return ret; |
899 | } | |
900 | ||
313ab753 FM |
901 | static int unlink_inode_for_log_replay(struct btrfs_trans_handle *trans, |
902 | struct btrfs_inode *dir, | |
903 | struct btrfs_inode *inode, | |
904 | const char *name, | |
905 | int name_len) | |
906 | { | |
907 | int ret; | |
908 | ||
909 | ret = btrfs_unlink_inode(trans, dir, inode, name, name_len); | |
910 | if (ret) | |
911 | return ret; | |
912 | /* | |
913 | * Whenever we need to check if a name exists or not, we check the | |
914 | * fs/subvolume tree. So after an unlink we must run delayed items, so | |
915 | * that future checks for a name during log replay see that the name | |
916 | * does not exists anymore. | |
917 | */ | |
918 | return btrfs_run_delayed_items(trans); | |
919 | } | |
920 | ||
e02119d5 CM |
921 | /* |
922 | * when cleaning up conflicts between the directory names in the | |
923 | * subvolume, directory names in the log and directory names in the | |
924 | * inode back references, we may have to unlink inodes from directories. | |
925 | * | |
926 | * This is a helper function to do the unlink of a specific directory | |
927 | * item | |
928 | */ | |
929 | static noinline int drop_one_dir_item(struct btrfs_trans_handle *trans, | |
e02119d5 | 930 | struct btrfs_path *path, |
207e7d92 | 931 | struct btrfs_inode *dir, |
e02119d5 CM |
932 | struct btrfs_dir_item *di) |
933 | { | |
9798ba24 | 934 | struct btrfs_root *root = dir->root; |
e02119d5 CM |
935 | struct inode *inode; |
936 | char *name; | |
937 | int name_len; | |
938 | struct extent_buffer *leaf; | |
939 | struct btrfs_key location; | |
940 | int ret; | |
941 | ||
942 | leaf = path->nodes[0]; | |
943 | ||
944 | btrfs_dir_item_key_to_cpu(leaf, di, &location); | |
945 | name_len = btrfs_dir_name_len(leaf, di); | |
946 | name = kmalloc(name_len, GFP_NOFS); | |
2a29edc6 | 947 | if (!name) |
948 | return -ENOMEM; | |
949 | ||
e02119d5 | 950 | read_extent_buffer(leaf, name, (unsigned long)(di + 1), name_len); |
b3b4aa74 | 951 | btrfs_release_path(path); |
e02119d5 CM |
952 | |
953 | inode = read_one_inode(root, location.objectid); | |
c00e9493 | 954 | if (!inode) { |
3650860b JB |
955 | ret = -EIO; |
956 | goto out; | |
c00e9493 | 957 | } |
e02119d5 | 958 | |
ec051c0f | 959 | ret = link_to_fixup_dir(trans, root, path, location.objectid); |
3650860b JB |
960 | if (ret) |
961 | goto out; | |
12fcfd22 | 962 | |
313ab753 | 963 | ret = unlink_inode_for_log_replay(trans, dir, BTRFS_I(inode), name, |
207e7d92 | 964 | name_len); |
3650860b | 965 | out: |
e02119d5 | 966 | kfree(name); |
e02119d5 CM |
967 | iput(inode); |
968 | return ret; | |
969 | } | |
970 | ||
971 | /* | |
77a5b9e3 FM |
972 | * See if a given name and sequence number found in an inode back reference are |
973 | * already in a directory and correctly point to this inode. | |
974 | * | |
975 | * Returns: < 0 on error, 0 if the directory entry does not exists and 1 if it | |
976 | * exists. | |
e02119d5 CM |
977 | */ |
978 | static noinline int inode_in_dir(struct btrfs_root *root, | |
979 | struct btrfs_path *path, | |
980 | u64 dirid, u64 objectid, u64 index, | |
981 | const char *name, int name_len) | |
982 | { | |
983 | struct btrfs_dir_item *di; | |
984 | struct btrfs_key location; | |
77a5b9e3 | 985 | int ret = 0; |
e02119d5 CM |
986 | |
987 | di = btrfs_lookup_dir_index_item(NULL, root, path, dirid, | |
988 | index, name, name_len, 0); | |
77a5b9e3 | 989 | if (IS_ERR(di)) { |
8dcbc261 | 990 | ret = PTR_ERR(di); |
77a5b9e3 FM |
991 | goto out; |
992 | } else if (di) { | |
e02119d5 CM |
993 | btrfs_dir_item_key_to_cpu(path->nodes[0], di, &location); |
994 | if (location.objectid != objectid) | |
995 | goto out; | |
77a5b9e3 | 996 | } else { |
e02119d5 | 997 | goto out; |
77a5b9e3 | 998 | } |
e02119d5 | 999 | |
77a5b9e3 | 1000 | btrfs_release_path(path); |
e02119d5 | 1001 | di = btrfs_lookup_dir_item(NULL, root, path, dirid, name, name_len, 0); |
77a5b9e3 FM |
1002 | if (IS_ERR(di)) { |
1003 | ret = PTR_ERR(di); | |
e02119d5 | 1004 | goto out; |
77a5b9e3 FM |
1005 | } else if (di) { |
1006 | btrfs_dir_item_key_to_cpu(path->nodes[0], di, &location); | |
1007 | if (location.objectid == objectid) | |
1008 | ret = 1; | |
1009 | } | |
e02119d5 | 1010 | out: |
b3b4aa74 | 1011 | btrfs_release_path(path); |
77a5b9e3 | 1012 | return ret; |
e02119d5 CM |
1013 | } |
1014 | ||
1015 | /* | |
1016 | * helper function to check a log tree for a named back reference in | |
1017 | * an inode. This is used to decide if a back reference that is | |
1018 | * found in the subvolume conflicts with what we find in the log. | |
1019 | * | |
1020 | * inode backreferences may have multiple refs in a single item, | |
1021 | * during replay we process one reference at a time, and we don't | |
1022 | * want to delete valid links to a file from the subvolume if that | |
1023 | * link is also in the log. | |
1024 | */ | |
1025 | static noinline int backref_in_log(struct btrfs_root *log, | |
1026 | struct btrfs_key *key, | |
f186373f | 1027 | u64 ref_objectid, |
df8d116f | 1028 | const char *name, int namelen) |
e02119d5 CM |
1029 | { |
1030 | struct btrfs_path *path; | |
e02119d5 | 1031 | int ret; |
e02119d5 CM |
1032 | |
1033 | path = btrfs_alloc_path(); | |
2a29edc6 | 1034 | if (!path) |
1035 | return -ENOMEM; | |
1036 | ||
e02119d5 | 1037 | ret = btrfs_search_slot(NULL, log, key, path, 0, 0); |
d3316c82 NB |
1038 | if (ret < 0) { |
1039 | goto out; | |
1040 | } else if (ret == 1) { | |
89cbf5f6 | 1041 | ret = 0; |
f186373f MF |
1042 | goto out; |
1043 | } | |
1044 | ||
89cbf5f6 NB |
1045 | if (key->type == BTRFS_INODE_EXTREF_KEY) |
1046 | ret = !!btrfs_find_name_in_ext_backref(path->nodes[0], | |
1047 | path->slots[0], | |
1048 | ref_objectid, | |
1049 | name, namelen); | |
1050 | else | |
1051 | ret = !!btrfs_find_name_in_backref(path->nodes[0], | |
1052 | path->slots[0], | |
1053 | name, namelen); | |
e02119d5 CM |
1054 | out: |
1055 | btrfs_free_path(path); | |
89cbf5f6 | 1056 | return ret; |
e02119d5 CM |
1057 | } |
1058 | ||
5a1d7843 | 1059 | static inline int __add_inode_ref(struct btrfs_trans_handle *trans, |
e02119d5 | 1060 | struct btrfs_root *root, |
e02119d5 | 1061 | struct btrfs_path *path, |
5a1d7843 | 1062 | struct btrfs_root *log_root, |
94c91a1f NB |
1063 | struct btrfs_inode *dir, |
1064 | struct btrfs_inode *inode, | |
f186373f MF |
1065 | u64 inode_objectid, u64 parent_objectid, |
1066 | u64 ref_index, char *name, int namelen, | |
1067 | int *search_done) | |
e02119d5 | 1068 | { |
34f3e4f2 | 1069 | int ret; |
f186373f MF |
1070 | char *victim_name; |
1071 | int victim_name_len; | |
1072 | struct extent_buffer *leaf; | |
5a1d7843 | 1073 | struct btrfs_dir_item *di; |
f186373f MF |
1074 | struct btrfs_key search_key; |
1075 | struct btrfs_inode_extref *extref; | |
c622ae60 | 1076 | |
f186373f MF |
1077 | again: |
1078 | /* Search old style refs */ | |
1079 | search_key.objectid = inode_objectid; | |
1080 | search_key.type = BTRFS_INODE_REF_KEY; | |
1081 | search_key.offset = parent_objectid; | |
1082 | ret = btrfs_search_slot(NULL, root, &search_key, path, 0, 0); | |
e02119d5 | 1083 | if (ret == 0) { |
e02119d5 CM |
1084 | struct btrfs_inode_ref *victim_ref; |
1085 | unsigned long ptr; | |
1086 | unsigned long ptr_end; | |
f186373f MF |
1087 | |
1088 | leaf = path->nodes[0]; | |
e02119d5 CM |
1089 | |
1090 | /* are we trying to overwrite a back ref for the root directory | |
1091 | * if so, just jump out, we're done | |
1092 | */ | |
f186373f | 1093 | if (search_key.objectid == search_key.offset) |
5a1d7843 | 1094 | return 1; |
e02119d5 CM |
1095 | |
1096 | /* check all the names in this back reference to see | |
1097 | * if they are in the log. if so, we allow them to stay | |
1098 | * otherwise they must be unlinked as a conflict | |
1099 | */ | |
1100 | ptr = btrfs_item_ptr_offset(leaf, path->slots[0]); | |
3212fa14 | 1101 | ptr_end = ptr + btrfs_item_size(leaf, path->slots[0]); |
d397712b | 1102 | while (ptr < ptr_end) { |
e02119d5 CM |
1103 | victim_ref = (struct btrfs_inode_ref *)ptr; |
1104 | victim_name_len = btrfs_inode_ref_name_len(leaf, | |
1105 | victim_ref); | |
1106 | victim_name = kmalloc(victim_name_len, GFP_NOFS); | |
3650860b JB |
1107 | if (!victim_name) |
1108 | return -ENOMEM; | |
e02119d5 CM |
1109 | |
1110 | read_extent_buffer(leaf, victim_name, | |
1111 | (unsigned long)(victim_ref + 1), | |
1112 | victim_name_len); | |
1113 | ||
d3316c82 NB |
1114 | ret = backref_in_log(log_root, &search_key, |
1115 | parent_objectid, victim_name, | |
1116 | victim_name_len); | |
1117 | if (ret < 0) { | |
1118 | kfree(victim_name); | |
1119 | return ret; | |
1120 | } else if (!ret) { | |
94c91a1f | 1121 | inc_nlink(&inode->vfs_inode); |
b3b4aa74 | 1122 | btrfs_release_path(path); |
12fcfd22 | 1123 | |
313ab753 | 1124 | ret = unlink_inode_for_log_replay(trans, dir, inode, |
4ec5934e | 1125 | victim_name, victim_name_len); |
f186373f | 1126 | kfree(victim_name); |
ada9af21 FDBM |
1127 | if (ret) |
1128 | return ret; | |
f186373f MF |
1129 | *search_done = 1; |
1130 | goto again; | |
e02119d5 CM |
1131 | } |
1132 | kfree(victim_name); | |
f186373f | 1133 | |
e02119d5 CM |
1134 | ptr = (unsigned long)(victim_ref + 1) + victim_name_len; |
1135 | } | |
e02119d5 | 1136 | |
c622ae60 | 1137 | /* |
1138 | * NOTE: we have searched root tree and checked the | |
bb7ab3b9 | 1139 | * corresponding ref, it does not need to check again. |
c622ae60 | 1140 | */ |
5a1d7843 | 1141 | *search_done = 1; |
e02119d5 | 1142 | } |
b3b4aa74 | 1143 | btrfs_release_path(path); |
e02119d5 | 1144 | |
f186373f MF |
1145 | /* Same search but for extended refs */ |
1146 | extref = btrfs_lookup_inode_extref(NULL, root, path, name, namelen, | |
1147 | inode_objectid, parent_objectid, 0, | |
1148 | 0); | |
7a6b75b7 FM |
1149 | if (IS_ERR(extref)) { |
1150 | return PTR_ERR(extref); | |
1151 | } else if (extref) { | |
f186373f MF |
1152 | u32 item_size; |
1153 | u32 cur_offset = 0; | |
1154 | unsigned long base; | |
1155 | struct inode *victim_parent; | |
1156 | ||
1157 | leaf = path->nodes[0]; | |
1158 | ||
3212fa14 | 1159 | item_size = btrfs_item_size(leaf, path->slots[0]); |
f186373f MF |
1160 | base = btrfs_item_ptr_offset(leaf, path->slots[0]); |
1161 | ||
1162 | while (cur_offset < item_size) { | |
dd9ef135 | 1163 | extref = (struct btrfs_inode_extref *)(base + cur_offset); |
f186373f MF |
1164 | |
1165 | victim_name_len = btrfs_inode_extref_name_len(leaf, extref); | |
1166 | ||
1167 | if (btrfs_inode_extref_parent(leaf, extref) != parent_objectid) | |
1168 | goto next; | |
1169 | ||
1170 | victim_name = kmalloc(victim_name_len, GFP_NOFS); | |
3650860b JB |
1171 | if (!victim_name) |
1172 | return -ENOMEM; | |
f186373f MF |
1173 | read_extent_buffer(leaf, victim_name, (unsigned long)&extref->name, |
1174 | victim_name_len); | |
1175 | ||
1176 | search_key.objectid = inode_objectid; | |
1177 | search_key.type = BTRFS_INODE_EXTREF_KEY; | |
1178 | search_key.offset = btrfs_extref_hash(parent_objectid, | |
1179 | victim_name, | |
1180 | victim_name_len); | |
d3316c82 NB |
1181 | ret = backref_in_log(log_root, &search_key, |
1182 | parent_objectid, victim_name, | |
1183 | victim_name_len); | |
1184 | if (ret < 0) { | |
f35838a6 | 1185 | kfree(victim_name); |
d3316c82 NB |
1186 | return ret; |
1187 | } else if (!ret) { | |
f186373f MF |
1188 | ret = -ENOENT; |
1189 | victim_parent = read_one_inode(root, | |
94c91a1f | 1190 | parent_objectid); |
f186373f | 1191 | if (victim_parent) { |
94c91a1f | 1192 | inc_nlink(&inode->vfs_inode); |
f186373f MF |
1193 | btrfs_release_path(path); |
1194 | ||
313ab753 | 1195 | ret = unlink_inode_for_log_replay(trans, |
4ec5934e | 1196 | BTRFS_I(victim_parent), |
94c91a1f | 1197 | inode, |
4ec5934e NB |
1198 | victim_name, |
1199 | victim_name_len); | |
f186373f | 1200 | } |
f186373f MF |
1201 | iput(victim_parent); |
1202 | kfree(victim_name); | |
3650860b JB |
1203 | if (ret) |
1204 | return ret; | |
f186373f MF |
1205 | *search_done = 1; |
1206 | goto again; | |
1207 | } | |
1208 | kfree(victim_name); | |
f186373f MF |
1209 | next: |
1210 | cur_offset += victim_name_len + sizeof(*extref); | |
1211 | } | |
1212 | *search_done = 1; | |
1213 | } | |
1214 | btrfs_release_path(path); | |
1215 | ||
34f3e4f2 | 1216 | /* look for a conflicting sequence number */ |
94c91a1f | 1217 | di = btrfs_lookup_dir_index_item(trans, root, path, btrfs_ino(dir), |
f186373f | 1218 | ref_index, name, namelen, 0); |
52db7779 | 1219 | if (IS_ERR(di)) { |
8dcbc261 | 1220 | return PTR_ERR(di); |
52db7779 | 1221 | } else if (di) { |
9798ba24 | 1222 | ret = drop_one_dir_item(trans, path, dir, di); |
3650860b JB |
1223 | if (ret) |
1224 | return ret; | |
34f3e4f2 | 1225 | } |
1226 | btrfs_release_path(path); | |
1227 | ||
52042d8e | 1228 | /* look for a conflicting name */ |
94c91a1f | 1229 | di = btrfs_lookup_dir_item(trans, root, path, btrfs_ino(dir), |
34f3e4f2 | 1230 | name, namelen, 0); |
52db7779 FM |
1231 | if (IS_ERR(di)) { |
1232 | return PTR_ERR(di); | |
1233 | } else if (di) { | |
9798ba24 | 1234 | ret = drop_one_dir_item(trans, path, dir, di); |
3650860b JB |
1235 | if (ret) |
1236 | return ret; | |
34f3e4f2 | 1237 | } |
1238 | btrfs_release_path(path); | |
1239 | ||
5a1d7843 JS |
1240 | return 0; |
1241 | } | |
e02119d5 | 1242 | |
bae15d95 QW |
1243 | static int extref_get_fields(struct extent_buffer *eb, unsigned long ref_ptr, |
1244 | u32 *namelen, char **name, u64 *index, | |
1245 | u64 *parent_objectid) | |
f186373f MF |
1246 | { |
1247 | struct btrfs_inode_extref *extref; | |
1248 | ||
1249 | extref = (struct btrfs_inode_extref *)ref_ptr; | |
1250 | ||
1251 | *namelen = btrfs_inode_extref_name_len(eb, extref); | |
1252 | *name = kmalloc(*namelen, GFP_NOFS); | |
1253 | if (*name == NULL) | |
1254 | return -ENOMEM; | |
1255 | ||
1256 | read_extent_buffer(eb, *name, (unsigned long)&extref->name, | |
1257 | *namelen); | |
1258 | ||
1f250e92 FM |
1259 | if (index) |
1260 | *index = btrfs_inode_extref_index(eb, extref); | |
f186373f MF |
1261 | if (parent_objectid) |
1262 | *parent_objectid = btrfs_inode_extref_parent(eb, extref); | |
1263 | ||
1264 | return 0; | |
1265 | } | |
1266 | ||
bae15d95 QW |
1267 | static int ref_get_fields(struct extent_buffer *eb, unsigned long ref_ptr, |
1268 | u32 *namelen, char **name, u64 *index) | |
f186373f MF |
1269 | { |
1270 | struct btrfs_inode_ref *ref; | |
1271 | ||
1272 | ref = (struct btrfs_inode_ref *)ref_ptr; | |
1273 | ||
1274 | *namelen = btrfs_inode_ref_name_len(eb, ref); | |
1275 | *name = kmalloc(*namelen, GFP_NOFS); | |
1276 | if (*name == NULL) | |
1277 | return -ENOMEM; | |
1278 | ||
1279 | read_extent_buffer(eb, *name, (unsigned long)(ref + 1), *namelen); | |
1280 | ||
1f250e92 FM |
1281 | if (index) |
1282 | *index = btrfs_inode_ref_index(eb, ref); | |
f186373f MF |
1283 | |
1284 | return 0; | |
1285 | } | |
1286 | ||
1f250e92 FM |
1287 | /* |
1288 | * Take an inode reference item from the log tree and iterate all names from the | |
1289 | * inode reference item in the subvolume tree with the same key (if it exists). | |
1290 | * For any name that is not in the inode reference item from the log tree, do a | |
1291 | * proper unlink of that name (that is, remove its entry from the inode | |
1292 | * reference item and both dir index keys). | |
1293 | */ | |
1294 | static int unlink_old_inode_refs(struct btrfs_trans_handle *trans, | |
1295 | struct btrfs_root *root, | |
1296 | struct btrfs_path *path, | |
1297 | struct btrfs_inode *inode, | |
1298 | struct extent_buffer *log_eb, | |
1299 | int log_slot, | |
1300 | struct btrfs_key *key) | |
1301 | { | |
1302 | int ret; | |
1303 | unsigned long ref_ptr; | |
1304 | unsigned long ref_end; | |
1305 | struct extent_buffer *eb; | |
1306 | ||
1307 | again: | |
1308 | btrfs_release_path(path); | |
1309 | ret = btrfs_search_slot(NULL, root, key, path, 0, 0); | |
1310 | if (ret > 0) { | |
1311 | ret = 0; | |
1312 | goto out; | |
1313 | } | |
1314 | if (ret < 0) | |
1315 | goto out; | |
1316 | ||
1317 | eb = path->nodes[0]; | |
1318 | ref_ptr = btrfs_item_ptr_offset(eb, path->slots[0]); | |
3212fa14 | 1319 | ref_end = ref_ptr + btrfs_item_size(eb, path->slots[0]); |
1f250e92 FM |
1320 | while (ref_ptr < ref_end) { |
1321 | char *name = NULL; | |
1322 | int namelen; | |
1323 | u64 parent_id; | |
1324 | ||
1325 | if (key->type == BTRFS_INODE_EXTREF_KEY) { | |
1326 | ret = extref_get_fields(eb, ref_ptr, &namelen, &name, | |
1327 | NULL, &parent_id); | |
1328 | } else { | |
1329 | parent_id = key->offset; | |
1330 | ret = ref_get_fields(eb, ref_ptr, &namelen, &name, | |
1331 | NULL); | |
1332 | } | |
1333 | if (ret) | |
1334 | goto out; | |
1335 | ||
1336 | if (key->type == BTRFS_INODE_EXTREF_KEY) | |
6ff49c6a NB |
1337 | ret = !!btrfs_find_name_in_ext_backref(log_eb, log_slot, |
1338 | parent_id, name, | |
1339 | namelen); | |
1f250e92 | 1340 | else |
9bb8407f NB |
1341 | ret = !!btrfs_find_name_in_backref(log_eb, log_slot, |
1342 | name, namelen); | |
1f250e92 FM |
1343 | |
1344 | if (!ret) { | |
1345 | struct inode *dir; | |
1346 | ||
1347 | btrfs_release_path(path); | |
1348 | dir = read_one_inode(root, parent_id); | |
1349 | if (!dir) { | |
1350 | ret = -ENOENT; | |
1351 | kfree(name); | |
1352 | goto out; | |
1353 | } | |
313ab753 | 1354 | ret = unlink_inode_for_log_replay(trans, BTRFS_I(dir), |
1f250e92 FM |
1355 | inode, name, namelen); |
1356 | kfree(name); | |
1357 | iput(dir); | |
1358 | if (ret) | |
1359 | goto out; | |
1360 | goto again; | |
1361 | } | |
1362 | ||
1363 | kfree(name); | |
1364 | ref_ptr += namelen; | |
1365 | if (key->type == BTRFS_INODE_EXTREF_KEY) | |
1366 | ref_ptr += sizeof(struct btrfs_inode_extref); | |
1367 | else | |
1368 | ref_ptr += sizeof(struct btrfs_inode_ref); | |
1369 | } | |
1370 | ret = 0; | |
1371 | out: | |
1372 | btrfs_release_path(path); | |
1373 | return ret; | |
1374 | } | |
1375 | ||
0d836392 FM |
1376 | static int btrfs_inode_ref_exists(struct inode *inode, struct inode *dir, |
1377 | const u8 ref_type, const char *name, | |
1378 | const int namelen) | |
1379 | { | |
1380 | struct btrfs_key key; | |
1381 | struct btrfs_path *path; | |
1382 | const u64 parent_id = btrfs_ino(BTRFS_I(dir)); | |
1383 | int ret; | |
1384 | ||
1385 | path = btrfs_alloc_path(); | |
1386 | if (!path) | |
1387 | return -ENOMEM; | |
1388 | ||
1389 | key.objectid = btrfs_ino(BTRFS_I(inode)); | |
1390 | key.type = ref_type; | |
1391 | if (key.type == BTRFS_INODE_REF_KEY) | |
1392 | key.offset = parent_id; | |
1393 | else | |
1394 | key.offset = btrfs_extref_hash(parent_id, name, namelen); | |
1395 | ||
1396 | ret = btrfs_search_slot(NULL, BTRFS_I(inode)->root, &key, path, 0, 0); | |
1397 | if (ret < 0) | |
1398 | goto out; | |
1399 | if (ret > 0) { | |
1400 | ret = 0; | |
1401 | goto out; | |
1402 | } | |
1403 | if (key.type == BTRFS_INODE_EXTREF_KEY) | |
6ff49c6a NB |
1404 | ret = !!btrfs_find_name_in_ext_backref(path->nodes[0], |
1405 | path->slots[0], parent_id, name, namelen); | |
0d836392 | 1406 | else |
9bb8407f NB |
1407 | ret = !!btrfs_find_name_in_backref(path->nodes[0], path->slots[0], |
1408 | name, namelen); | |
0d836392 FM |
1409 | |
1410 | out: | |
1411 | btrfs_free_path(path); | |
1412 | return ret; | |
1413 | } | |
1414 | ||
6d9cc072 | 1415 | static int add_link(struct btrfs_trans_handle *trans, |
6b5fc433 FM |
1416 | struct inode *dir, struct inode *inode, const char *name, |
1417 | int namelen, u64 ref_index) | |
1418 | { | |
6d9cc072 | 1419 | struct btrfs_root *root = BTRFS_I(dir)->root; |
6b5fc433 FM |
1420 | struct btrfs_dir_item *dir_item; |
1421 | struct btrfs_key key; | |
1422 | struct btrfs_path *path; | |
1423 | struct inode *other_inode = NULL; | |
1424 | int ret; | |
1425 | ||
1426 | path = btrfs_alloc_path(); | |
1427 | if (!path) | |
1428 | return -ENOMEM; | |
1429 | ||
1430 | dir_item = btrfs_lookup_dir_item(NULL, root, path, | |
1431 | btrfs_ino(BTRFS_I(dir)), | |
1432 | name, namelen, 0); | |
1433 | if (!dir_item) { | |
1434 | btrfs_release_path(path); | |
1435 | goto add_link; | |
1436 | } else if (IS_ERR(dir_item)) { | |
1437 | ret = PTR_ERR(dir_item); | |
1438 | goto out; | |
1439 | } | |
1440 | ||
1441 | /* | |
1442 | * Our inode's dentry collides with the dentry of another inode which is | |
1443 | * in the log but not yet processed since it has a higher inode number. | |
1444 | * So delete that other dentry. | |
1445 | */ | |
1446 | btrfs_dir_item_key_to_cpu(path->nodes[0], dir_item, &key); | |
1447 | btrfs_release_path(path); | |
1448 | other_inode = read_one_inode(root, key.objectid); | |
1449 | if (!other_inode) { | |
1450 | ret = -ENOENT; | |
1451 | goto out; | |
1452 | } | |
313ab753 FM |
1453 | ret = unlink_inode_for_log_replay(trans, BTRFS_I(dir), BTRFS_I(other_inode), |
1454 | name, namelen); | |
6b5fc433 FM |
1455 | if (ret) |
1456 | goto out; | |
1457 | /* | |
1458 | * If we dropped the link count to 0, bump it so that later the iput() | |
1459 | * on the inode will not free it. We will fixup the link count later. | |
1460 | */ | |
1461 | if (other_inode->i_nlink == 0) | |
769030e1 | 1462 | set_nlink(other_inode, 1); |
6b5fc433 FM |
1463 | add_link: |
1464 | ret = btrfs_add_link(trans, BTRFS_I(dir), BTRFS_I(inode), | |
1465 | name, namelen, 0, ref_index); | |
1466 | out: | |
1467 | iput(other_inode); | |
1468 | btrfs_free_path(path); | |
1469 | ||
1470 | return ret; | |
1471 | } | |
1472 | ||
5a1d7843 JS |
1473 | /* |
1474 | * replay one inode back reference item found in the log tree. | |
1475 | * eb, slot and key refer to the buffer and key found in the log tree. | |
1476 | * root is the destination we are replaying into, and path is for temp | |
1477 | * use by this function. (it should be released on return). | |
1478 | */ | |
1479 | static noinline int add_inode_ref(struct btrfs_trans_handle *trans, | |
1480 | struct btrfs_root *root, | |
1481 | struct btrfs_root *log, | |
1482 | struct btrfs_path *path, | |
1483 | struct extent_buffer *eb, int slot, | |
1484 | struct btrfs_key *key) | |
1485 | { | |
03b2f08b GB |
1486 | struct inode *dir = NULL; |
1487 | struct inode *inode = NULL; | |
5a1d7843 JS |
1488 | unsigned long ref_ptr; |
1489 | unsigned long ref_end; | |
03b2f08b | 1490 | char *name = NULL; |
5a1d7843 JS |
1491 | int namelen; |
1492 | int ret; | |
1493 | int search_done = 0; | |
f186373f MF |
1494 | int log_ref_ver = 0; |
1495 | u64 parent_objectid; | |
1496 | u64 inode_objectid; | |
f46dbe3d | 1497 | u64 ref_index = 0; |
f186373f MF |
1498 | int ref_struct_size; |
1499 | ||
1500 | ref_ptr = btrfs_item_ptr_offset(eb, slot); | |
3212fa14 | 1501 | ref_end = ref_ptr + btrfs_item_size(eb, slot); |
f186373f MF |
1502 | |
1503 | if (key->type == BTRFS_INODE_EXTREF_KEY) { | |
1504 | struct btrfs_inode_extref *r; | |
1505 | ||
1506 | ref_struct_size = sizeof(struct btrfs_inode_extref); | |
1507 | log_ref_ver = 1; | |
1508 | r = (struct btrfs_inode_extref *)ref_ptr; | |
1509 | parent_objectid = btrfs_inode_extref_parent(eb, r); | |
1510 | } else { | |
1511 | ref_struct_size = sizeof(struct btrfs_inode_ref); | |
1512 | parent_objectid = key->offset; | |
1513 | } | |
1514 | inode_objectid = key->objectid; | |
e02119d5 | 1515 | |
5a1d7843 JS |
1516 | /* |
1517 | * it is possible that we didn't log all the parent directories | |
1518 | * for a given inode. If we don't find the dir, just don't | |
1519 | * copy the back ref in. The link count fixup code will take | |
1520 | * care of the rest | |
1521 | */ | |
f186373f | 1522 | dir = read_one_inode(root, parent_objectid); |
03b2f08b GB |
1523 | if (!dir) { |
1524 | ret = -ENOENT; | |
1525 | goto out; | |
1526 | } | |
5a1d7843 | 1527 | |
f186373f | 1528 | inode = read_one_inode(root, inode_objectid); |
5a1d7843 | 1529 | if (!inode) { |
03b2f08b GB |
1530 | ret = -EIO; |
1531 | goto out; | |
5a1d7843 JS |
1532 | } |
1533 | ||
5a1d7843 | 1534 | while (ref_ptr < ref_end) { |
f186373f | 1535 | if (log_ref_ver) { |
bae15d95 QW |
1536 | ret = extref_get_fields(eb, ref_ptr, &namelen, &name, |
1537 | &ref_index, &parent_objectid); | |
f186373f MF |
1538 | /* |
1539 | * parent object can change from one array | |
1540 | * item to another. | |
1541 | */ | |
1542 | if (!dir) | |
1543 | dir = read_one_inode(root, parent_objectid); | |
03b2f08b GB |
1544 | if (!dir) { |
1545 | ret = -ENOENT; | |
1546 | goto out; | |
1547 | } | |
f186373f | 1548 | } else { |
bae15d95 QW |
1549 | ret = ref_get_fields(eb, ref_ptr, &namelen, &name, |
1550 | &ref_index); | |
f186373f MF |
1551 | } |
1552 | if (ret) | |
03b2f08b | 1553 | goto out; |
5a1d7843 | 1554 | |
77a5b9e3 FM |
1555 | ret = inode_in_dir(root, path, btrfs_ino(BTRFS_I(dir)), |
1556 | btrfs_ino(BTRFS_I(inode)), ref_index, | |
1557 | name, namelen); | |
1558 | if (ret < 0) { | |
1559 | goto out; | |
1560 | } else if (ret == 0) { | |
5a1d7843 JS |
1561 | /* |
1562 | * look for a conflicting back reference in the | |
1563 | * metadata. if we find one we have to unlink that name | |
1564 | * of the file before we add our new link. Later on, we | |
1565 | * overwrite any existing back reference, and we don't | |
1566 | * want to create dangling pointers in the directory. | |
1567 | */ | |
1568 | ||
1569 | if (!search_done) { | |
1570 | ret = __add_inode_ref(trans, root, path, log, | |
94c91a1f | 1571 | BTRFS_I(dir), |
d75eefdf | 1572 | BTRFS_I(inode), |
f186373f MF |
1573 | inode_objectid, |
1574 | parent_objectid, | |
1575 | ref_index, name, namelen, | |
5a1d7843 | 1576 | &search_done); |
03b2f08b GB |
1577 | if (ret) { |
1578 | if (ret == 1) | |
1579 | ret = 0; | |
3650860b JB |
1580 | goto out; |
1581 | } | |
5a1d7843 JS |
1582 | } |
1583 | ||
0d836392 FM |
1584 | /* |
1585 | * If a reference item already exists for this inode | |
1586 | * with the same parent and name, but different index, | |
1587 | * drop it and the corresponding directory index entries | |
1588 | * from the parent before adding the new reference item | |
1589 | * and dir index entries, otherwise we would fail with | |
1590 | * -EEXIST returned from btrfs_add_link() below. | |
1591 | */ | |
1592 | ret = btrfs_inode_ref_exists(inode, dir, key->type, | |
1593 | name, namelen); | |
1594 | if (ret > 0) { | |
313ab753 | 1595 | ret = unlink_inode_for_log_replay(trans, |
0d836392 FM |
1596 | BTRFS_I(dir), |
1597 | BTRFS_I(inode), | |
1598 | name, namelen); | |
1599 | /* | |
1600 | * If we dropped the link count to 0, bump it so | |
1601 | * that later the iput() on the inode will not | |
1602 | * free it. We will fixup the link count later. | |
1603 | */ | |
1604 | if (!ret && inode->i_nlink == 0) | |
769030e1 | 1605 | set_nlink(inode, 1); |
0d836392 FM |
1606 | } |
1607 | if (ret < 0) | |
1608 | goto out; | |
1609 | ||
5a1d7843 | 1610 | /* insert our name */ |
6d9cc072 | 1611 | ret = add_link(trans, dir, inode, name, namelen, |
6b5fc433 | 1612 | ref_index); |
3650860b JB |
1613 | if (ret) |
1614 | goto out; | |
5a1d7843 | 1615 | |
f96d4474 JB |
1616 | ret = btrfs_update_inode(trans, root, BTRFS_I(inode)); |
1617 | if (ret) | |
1618 | goto out; | |
5a1d7843 | 1619 | } |
77a5b9e3 | 1620 | /* Else, ret == 1, we already have a perfect match, we're done. */ |
5a1d7843 | 1621 | |
f186373f | 1622 | ref_ptr = (unsigned long)(ref_ptr + ref_struct_size) + namelen; |
5a1d7843 | 1623 | kfree(name); |
03b2f08b | 1624 | name = NULL; |
f186373f MF |
1625 | if (log_ref_ver) { |
1626 | iput(dir); | |
1627 | dir = NULL; | |
1628 | } | |
5a1d7843 | 1629 | } |
e02119d5 | 1630 | |
1f250e92 FM |
1631 | /* |
1632 | * Before we overwrite the inode reference item in the subvolume tree | |
1633 | * with the item from the log tree, we must unlink all names from the | |
1634 | * parent directory that are in the subvolume's tree inode reference | |
1635 | * item, otherwise we end up with an inconsistent subvolume tree where | |
1636 | * dir index entries exist for a name but there is no inode reference | |
1637 | * item with the same name. | |
1638 | */ | |
1639 | ret = unlink_old_inode_refs(trans, root, path, BTRFS_I(inode), eb, slot, | |
1640 | key); | |
1641 | if (ret) | |
1642 | goto out; | |
1643 | ||
e02119d5 CM |
1644 | /* finally write the back reference in the inode */ |
1645 | ret = overwrite_item(trans, root, path, eb, slot, key); | |
5a1d7843 | 1646 | out: |
b3b4aa74 | 1647 | btrfs_release_path(path); |
03b2f08b | 1648 | kfree(name); |
e02119d5 CM |
1649 | iput(dir); |
1650 | iput(inode); | |
3650860b | 1651 | return ret; |
e02119d5 CM |
1652 | } |
1653 | ||
f186373f | 1654 | static int count_inode_extrefs(struct btrfs_root *root, |
36283658 | 1655 | struct btrfs_inode *inode, struct btrfs_path *path) |
f186373f MF |
1656 | { |
1657 | int ret = 0; | |
1658 | int name_len; | |
1659 | unsigned int nlink = 0; | |
1660 | u32 item_size; | |
1661 | u32 cur_offset = 0; | |
36283658 | 1662 | u64 inode_objectid = btrfs_ino(inode); |
f186373f MF |
1663 | u64 offset = 0; |
1664 | unsigned long ptr; | |
1665 | struct btrfs_inode_extref *extref; | |
1666 | struct extent_buffer *leaf; | |
1667 | ||
1668 | while (1) { | |
1669 | ret = btrfs_find_one_extref(root, inode_objectid, offset, path, | |
1670 | &extref, &offset); | |
1671 | if (ret) | |
1672 | break; | |
c71bf099 | 1673 | |
f186373f | 1674 | leaf = path->nodes[0]; |
3212fa14 | 1675 | item_size = btrfs_item_size(leaf, path->slots[0]); |
f186373f | 1676 | ptr = btrfs_item_ptr_offset(leaf, path->slots[0]); |
2c2c452b | 1677 | cur_offset = 0; |
f186373f MF |
1678 | |
1679 | while (cur_offset < item_size) { | |
1680 | extref = (struct btrfs_inode_extref *) (ptr + cur_offset); | |
1681 | name_len = btrfs_inode_extref_name_len(leaf, extref); | |
1682 | ||
1683 | nlink++; | |
1684 | ||
1685 | cur_offset += name_len + sizeof(*extref); | |
1686 | } | |
1687 | ||
1688 | offset++; | |
1689 | btrfs_release_path(path); | |
1690 | } | |
1691 | btrfs_release_path(path); | |
1692 | ||
2c2c452b | 1693 | if (ret < 0 && ret != -ENOENT) |
f186373f MF |
1694 | return ret; |
1695 | return nlink; | |
1696 | } | |
1697 | ||
1698 | static int count_inode_refs(struct btrfs_root *root, | |
f329e319 | 1699 | struct btrfs_inode *inode, struct btrfs_path *path) |
e02119d5 | 1700 | { |
e02119d5 CM |
1701 | int ret; |
1702 | struct btrfs_key key; | |
f186373f | 1703 | unsigned int nlink = 0; |
e02119d5 CM |
1704 | unsigned long ptr; |
1705 | unsigned long ptr_end; | |
1706 | int name_len; | |
f329e319 | 1707 | u64 ino = btrfs_ino(inode); |
e02119d5 | 1708 | |
33345d01 | 1709 | key.objectid = ino; |
e02119d5 CM |
1710 | key.type = BTRFS_INODE_REF_KEY; |
1711 | key.offset = (u64)-1; | |
1712 | ||
d397712b | 1713 | while (1) { |
e02119d5 CM |
1714 | ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); |
1715 | if (ret < 0) | |
1716 | break; | |
1717 | if (ret > 0) { | |
1718 | if (path->slots[0] == 0) | |
1719 | break; | |
1720 | path->slots[0]--; | |
1721 | } | |
e93ae26f | 1722 | process_slot: |
e02119d5 CM |
1723 | btrfs_item_key_to_cpu(path->nodes[0], &key, |
1724 | path->slots[0]); | |
33345d01 | 1725 | if (key.objectid != ino || |
e02119d5 CM |
1726 | key.type != BTRFS_INODE_REF_KEY) |
1727 | break; | |
1728 | ptr = btrfs_item_ptr_offset(path->nodes[0], path->slots[0]); | |
3212fa14 | 1729 | ptr_end = ptr + btrfs_item_size(path->nodes[0], |
e02119d5 | 1730 | path->slots[0]); |
d397712b | 1731 | while (ptr < ptr_end) { |
e02119d5 CM |
1732 | struct btrfs_inode_ref *ref; |
1733 | ||
1734 | ref = (struct btrfs_inode_ref *)ptr; | |
1735 | name_len = btrfs_inode_ref_name_len(path->nodes[0], | |
1736 | ref); | |
1737 | ptr = (unsigned long)(ref + 1) + name_len; | |
1738 | nlink++; | |
1739 | } | |
1740 | ||
1741 | if (key.offset == 0) | |
1742 | break; | |
e93ae26f FDBM |
1743 | if (path->slots[0] > 0) { |
1744 | path->slots[0]--; | |
1745 | goto process_slot; | |
1746 | } | |
e02119d5 | 1747 | key.offset--; |
b3b4aa74 | 1748 | btrfs_release_path(path); |
e02119d5 | 1749 | } |
b3b4aa74 | 1750 | btrfs_release_path(path); |
f186373f MF |
1751 | |
1752 | return nlink; | |
1753 | } | |
1754 | ||
1755 | /* | |
1756 | * There are a few corners where the link count of the file can't | |
1757 | * be properly maintained during replay. So, instead of adding | |
1758 | * lots of complexity to the log code, we just scan the backrefs | |
1759 | * for any file that has been through replay. | |
1760 | * | |
1761 | * The scan will update the link count on the inode to reflect the | |
1762 | * number of back refs found. If it goes down to zero, the iput | |
1763 | * will free the inode. | |
1764 | */ | |
1765 | static noinline int fixup_inode_link_count(struct btrfs_trans_handle *trans, | |
1766 | struct btrfs_root *root, | |
1767 | struct inode *inode) | |
1768 | { | |
1769 | struct btrfs_path *path; | |
1770 | int ret; | |
1771 | u64 nlink = 0; | |
4a0cc7ca | 1772 | u64 ino = btrfs_ino(BTRFS_I(inode)); |
f186373f MF |
1773 | |
1774 | path = btrfs_alloc_path(); | |
1775 | if (!path) | |
1776 | return -ENOMEM; | |
1777 | ||
f329e319 | 1778 | ret = count_inode_refs(root, BTRFS_I(inode), path); |
f186373f MF |
1779 | if (ret < 0) |
1780 | goto out; | |
1781 | ||
1782 | nlink = ret; | |
1783 | ||
36283658 | 1784 | ret = count_inode_extrefs(root, BTRFS_I(inode), path); |
f186373f MF |
1785 | if (ret < 0) |
1786 | goto out; | |
1787 | ||
1788 | nlink += ret; | |
1789 | ||
1790 | ret = 0; | |
1791 | ||
e02119d5 | 1792 | if (nlink != inode->i_nlink) { |
bfe86848 | 1793 | set_nlink(inode, nlink); |
f96d4474 JB |
1794 | ret = btrfs_update_inode(trans, root, BTRFS_I(inode)); |
1795 | if (ret) | |
1796 | goto out; | |
e02119d5 | 1797 | } |
8d5bf1cb | 1798 | BTRFS_I(inode)->index_cnt = (u64)-1; |
e02119d5 | 1799 | |
c71bf099 YZ |
1800 | if (inode->i_nlink == 0) { |
1801 | if (S_ISDIR(inode->i_mode)) { | |
1802 | ret = replay_dir_deletes(trans, root, NULL, path, | |
33345d01 | 1803 | ino, 1); |
3650860b JB |
1804 | if (ret) |
1805 | goto out; | |
c71bf099 | 1806 | } |
ecdcf3c2 NB |
1807 | ret = btrfs_insert_orphan_item(trans, root, ino); |
1808 | if (ret == -EEXIST) | |
1809 | ret = 0; | |
12fcfd22 | 1810 | } |
12fcfd22 | 1811 | |
f186373f MF |
1812 | out: |
1813 | btrfs_free_path(path); | |
1814 | return ret; | |
e02119d5 CM |
1815 | } |
1816 | ||
1817 | static noinline int fixup_inode_link_counts(struct btrfs_trans_handle *trans, | |
1818 | struct btrfs_root *root, | |
1819 | struct btrfs_path *path) | |
1820 | { | |
1821 | int ret; | |
1822 | struct btrfs_key key; | |
1823 | struct inode *inode; | |
1824 | ||
1825 | key.objectid = BTRFS_TREE_LOG_FIXUP_OBJECTID; | |
1826 | key.type = BTRFS_ORPHAN_ITEM_KEY; | |
1827 | key.offset = (u64)-1; | |
d397712b | 1828 | while (1) { |
e02119d5 CM |
1829 | ret = btrfs_search_slot(trans, root, &key, path, -1, 1); |
1830 | if (ret < 0) | |
1831 | break; | |
1832 | ||
1833 | if (ret == 1) { | |
011b28ac | 1834 | ret = 0; |
e02119d5 CM |
1835 | if (path->slots[0] == 0) |
1836 | break; | |
1837 | path->slots[0]--; | |
1838 | } | |
1839 | ||
1840 | btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]); | |
1841 | if (key.objectid != BTRFS_TREE_LOG_FIXUP_OBJECTID || | |
1842 | key.type != BTRFS_ORPHAN_ITEM_KEY) | |
1843 | break; | |
1844 | ||
1845 | ret = btrfs_del_item(trans, root, path); | |
65a246c5 | 1846 | if (ret) |
011b28ac | 1847 | break; |
e02119d5 | 1848 | |
b3b4aa74 | 1849 | btrfs_release_path(path); |
e02119d5 | 1850 | inode = read_one_inode(root, key.offset); |
011b28ac JB |
1851 | if (!inode) { |
1852 | ret = -EIO; | |
1853 | break; | |
1854 | } | |
e02119d5 CM |
1855 | |
1856 | ret = fixup_inode_link_count(trans, root, inode); | |
e02119d5 | 1857 | iput(inode); |
3650860b | 1858 | if (ret) |
011b28ac | 1859 | break; |
e02119d5 | 1860 | |
12fcfd22 CM |
1861 | /* |
1862 | * fixup on a directory may create new entries, | |
1863 | * make sure we always look for the highset possible | |
1864 | * offset | |
1865 | */ | |
1866 | key.offset = (u64)-1; | |
e02119d5 | 1867 | } |
b3b4aa74 | 1868 | btrfs_release_path(path); |
65a246c5 | 1869 | return ret; |
e02119d5 CM |
1870 | } |
1871 | ||
1872 | ||
1873 | /* | |
1874 | * record a given inode in the fixup dir so we can check its link | |
1875 | * count when replay is done. The link count is incremented here | |
1876 | * so the inode won't go away until we check it | |
1877 | */ | |
1878 | static noinline int link_to_fixup_dir(struct btrfs_trans_handle *trans, | |
1879 | struct btrfs_root *root, | |
1880 | struct btrfs_path *path, | |
1881 | u64 objectid) | |
1882 | { | |
1883 | struct btrfs_key key; | |
1884 | int ret = 0; | |
1885 | struct inode *inode; | |
1886 | ||
1887 | inode = read_one_inode(root, objectid); | |
c00e9493 TI |
1888 | if (!inode) |
1889 | return -EIO; | |
e02119d5 CM |
1890 | |
1891 | key.objectid = BTRFS_TREE_LOG_FIXUP_OBJECTID; | |
962a298f | 1892 | key.type = BTRFS_ORPHAN_ITEM_KEY; |
e02119d5 CM |
1893 | key.offset = objectid; |
1894 | ||
1895 | ret = btrfs_insert_empty_item(trans, root, path, &key, 0); | |
1896 | ||
b3b4aa74 | 1897 | btrfs_release_path(path); |
e02119d5 | 1898 | if (ret == 0) { |
9bf7a489 JB |
1899 | if (!inode->i_nlink) |
1900 | set_nlink(inode, 1); | |
1901 | else | |
8b558c5f | 1902 | inc_nlink(inode); |
9a56fcd1 | 1903 | ret = btrfs_update_inode(trans, root, BTRFS_I(inode)); |
e02119d5 CM |
1904 | } else if (ret == -EEXIST) { |
1905 | ret = 0; | |
e02119d5 CM |
1906 | } |
1907 | iput(inode); | |
1908 | ||
1909 | return ret; | |
1910 | } | |
1911 | ||
1912 | /* | |
1913 | * when replaying the log for a directory, we only insert names | |
1914 | * for inodes that actually exist. This means an fsync on a directory | |
1915 | * does not implicitly fsync all the new files in it | |
1916 | */ | |
1917 | static noinline int insert_one_name(struct btrfs_trans_handle *trans, | |
1918 | struct btrfs_root *root, | |
e02119d5 | 1919 | u64 dirid, u64 index, |
60d53eb3 | 1920 | char *name, int name_len, |
e02119d5 CM |
1921 | struct btrfs_key *location) |
1922 | { | |
1923 | struct inode *inode; | |
1924 | struct inode *dir; | |
1925 | int ret; | |
1926 | ||
1927 | inode = read_one_inode(root, location->objectid); | |
1928 | if (!inode) | |
1929 | return -ENOENT; | |
1930 | ||
1931 | dir = read_one_inode(root, dirid); | |
1932 | if (!dir) { | |
1933 | iput(inode); | |
1934 | return -EIO; | |
1935 | } | |
d555438b | 1936 | |
db0a669f NB |
1937 | ret = btrfs_add_link(trans, BTRFS_I(dir), BTRFS_I(inode), name, |
1938 | name_len, 1, index); | |
e02119d5 CM |
1939 | |
1940 | /* FIXME, put inode into FIXUP list */ | |
1941 | ||
1942 | iput(inode); | |
1943 | iput(dir); | |
1944 | return ret; | |
1945 | } | |
1946 | ||
339d0354 FM |
1947 | static int delete_conflicting_dir_entry(struct btrfs_trans_handle *trans, |
1948 | struct btrfs_inode *dir, | |
1949 | struct btrfs_path *path, | |
1950 | struct btrfs_dir_item *dst_di, | |
1951 | const struct btrfs_key *log_key, | |
1952 | u8 log_type, | |
1953 | bool exists) | |
1954 | { | |
1955 | struct btrfs_key found_key; | |
1956 | ||
1957 | btrfs_dir_item_key_to_cpu(path->nodes[0], dst_di, &found_key); | |
1958 | /* The existing dentry points to the same inode, don't delete it. */ | |
1959 | if (found_key.objectid == log_key->objectid && | |
1960 | found_key.type == log_key->type && | |
1961 | found_key.offset == log_key->offset && | |
1962 | btrfs_dir_type(path->nodes[0], dst_di) == log_type) | |
1963 | return 1; | |
1964 | ||
1965 | /* | |
1966 | * Don't drop the conflicting directory entry if the inode for the new | |
1967 | * entry doesn't exist. | |
1968 | */ | |
1969 | if (!exists) | |
1970 | return 0; | |
1971 | ||
1972 | return drop_one_dir_item(trans, path, dir, dst_di); | |
1973 | } | |
1974 | ||
e02119d5 CM |
1975 | /* |
1976 | * take a single entry in a log directory item and replay it into | |
1977 | * the subvolume. | |
1978 | * | |
1979 | * if a conflicting item exists in the subdirectory already, | |
1980 | * the inode it points to is unlinked and put into the link count | |
1981 | * fix up tree. | |
1982 | * | |
1983 | * If a name from the log points to a file or directory that does | |
1984 | * not exist in the FS, it is skipped. fsyncs on directories | |
1985 | * do not force down inodes inside that directory, just changes to the | |
1986 | * names or unlinks in a directory. | |
bb53eda9 FM |
1987 | * |
1988 | * Returns < 0 on error, 0 if the name wasn't replayed (dentry points to a | |
1989 | * non-existing inode) and 1 if the name was replayed. | |
e02119d5 CM |
1990 | */ |
1991 | static noinline int replay_one_name(struct btrfs_trans_handle *trans, | |
1992 | struct btrfs_root *root, | |
1993 | struct btrfs_path *path, | |
1994 | struct extent_buffer *eb, | |
1995 | struct btrfs_dir_item *di, | |
1996 | struct btrfs_key *key) | |
1997 | { | |
1998 | char *name; | |
1999 | int name_len; | |
339d0354 FM |
2000 | struct btrfs_dir_item *dir_dst_di; |
2001 | struct btrfs_dir_item *index_dst_di; | |
2002 | bool dir_dst_matches = false; | |
2003 | bool index_dst_matches = false; | |
e02119d5 | 2004 | struct btrfs_key log_key; |
339d0354 | 2005 | struct btrfs_key search_key; |
e02119d5 | 2006 | struct inode *dir; |
e02119d5 | 2007 | u8 log_type; |
cfd31269 FM |
2008 | bool exists; |
2009 | int ret; | |
339d0354 | 2010 | bool update_size = true; |
bb53eda9 | 2011 | bool name_added = false; |
e02119d5 CM |
2012 | |
2013 | dir = read_one_inode(root, key->objectid); | |
c00e9493 TI |
2014 | if (!dir) |
2015 | return -EIO; | |
e02119d5 CM |
2016 | |
2017 | name_len = btrfs_dir_name_len(eb, di); | |
2018 | name = kmalloc(name_len, GFP_NOFS); | |
2bac325e FDBM |
2019 | if (!name) { |
2020 | ret = -ENOMEM; | |
2021 | goto out; | |
2022 | } | |
2a29edc6 | 2023 | |
e02119d5 CM |
2024 | log_type = btrfs_dir_type(eb, di); |
2025 | read_extent_buffer(eb, name, (unsigned long)(di + 1), | |
2026 | name_len); | |
2027 | ||
2028 | btrfs_dir_item_key_to_cpu(eb, di, &log_key); | |
cfd31269 | 2029 | ret = btrfs_lookup_inode(trans, root, path, &log_key, 0); |
b3b4aa74 | 2030 | btrfs_release_path(path); |
cfd31269 FM |
2031 | if (ret < 0) |
2032 | goto out; | |
2033 | exists = (ret == 0); | |
2034 | ret = 0; | |
4bef0848 | 2035 | |
339d0354 FM |
2036 | dir_dst_di = btrfs_lookup_dir_item(trans, root, path, key->objectid, |
2037 | name, name_len, 1); | |
2038 | if (IS_ERR(dir_dst_di)) { | |
2039 | ret = PTR_ERR(dir_dst_di); | |
3650860b | 2040 | goto out; |
339d0354 FM |
2041 | } else if (dir_dst_di) { |
2042 | ret = delete_conflicting_dir_entry(trans, BTRFS_I(dir), path, | |
2043 | dir_dst_di, &log_key, log_type, | |
2044 | exists); | |
2045 | if (ret < 0) | |
2046 | goto out; | |
2047 | dir_dst_matches = (ret == 1); | |
e02119d5 | 2048 | } |
e15ac641 | 2049 | |
339d0354 FM |
2050 | btrfs_release_path(path); |
2051 | ||
2052 | index_dst_di = btrfs_lookup_dir_index_item(trans, root, path, | |
2053 | key->objectid, key->offset, | |
2054 | name, name_len, 1); | |
2055 | if (IS_ERR(index_dst_di)) { | |
2056 | ret = PTR_ERR(index_dst_di); | |
e15ac641 | 2057 | goto out; |
339d0354 FM |
2058 | } else if (index_dst_di) { |
2059 | ret = delete_conflicting_dir_entry(trans, BTRFS_I(dir), path, | |
2060 | index_dst_di, &log_key, | |
2061 | log_type, exists); | |
2062 | if (ret < 0) | |
e02119d5 | 2063 | goto out; |
339d0354 | 2064 | index_dst_matches = (ret == 1); |
e02119d5 CM |
2065 | } |
2066 | ||
339d0354 FM |
2067 | btrfs_release_path(path); |
2068 | ||
2069 | if (dir_dst_matches && index_dst_matches) { | |
2070 | ret = 0; | |
a2cc11db | 2071 | update_size = false; |
e02119d5 CM |
2072 | goto out; |
2073 | } | |
2074 | ||
725af92a NB |
2075 | /* |
2076 | * Check if the inode reference exists in the log for the given name, | |
2077 | * inode and parent inode | |
2078 | */ | |
339d0354 FM |
2079 | search_key.objectid = log_key.objectid; |
2080 | search_key.type = BTRFS_INODE_REF_KEY; | |
2081 | search_key.offset = key->objectid; | |
2082 | ret = backref_in_log(root->log_root, &search_key, 0, name, name_len); | |
725af92a NB |
2083 | if (ret < 0) { |
2084 | goto out; | |
2085 | } else if (ret) { | |
2086 | /* The dentry will be added later. */ | |
2087 | ret = 0; | |
2088 | update_size = false; | |
2089 | goto out; | |
2090 | } | |
2091 | ||
339d0354 FM |
2092 | search_key.objectid = log_key.objectid; |
2093 | search_key.type = BTRFS_INODE_EXTREF_KEY; | |
2094 | search_key.offset = key->objectid; | |
2095 | ret = backref_in_log(root->log_root, &search_key, key->objectid, name, | |
725af92a NB |
2096 | name_len); |
2097 | if (ret < 0) { | |
2098 | goto out; | |
2099 | } else if (ret) { | |
df8d116f FM |
2100 | /* The dentry will be added later. */ |
2101 | ret = 0; | |
2102 | update_size = false; | |
2103 | goto out; | |
2104 | } | |
b3b4aa74 | 2105 | btrfs_release_path(path); |
60d53eb3 Z |
2106 | ret = insert_one_name(trans, root, key->objectid, key->offset, |
2107 | name, name_len, &log_key); | |
df8d116f | 2108 | if (ret && ret != -ENOENT && ret != -EEXIST) |
3650860b | 2109 | goto out; |
bb53eda9 FM |
2110 | if (!ret) |
2111 | name_added = true; | |
d555438b | 2112 | update_size = false; |
3650860b | 2113 | ret = 0; |
339d0354 FM |
2114 | |
2115 | out: | |
2116 | if (!ret && update_size) { | |
2117 | btrfs_i_size_write(BTRFS_I(dir), dir->i_size + name_len * 2); | |
2118 | ret = btrfs_update_inode(trans, root, BTRFS_I(dir)); | |
2119 | } | |
2120 | kfree(name); | |
2121 | iput(dir); | |
2122 | if (!ret && name_added) | |
2123 | ret = 1; | |
2124 | return ret; | |
e02119d5 CM |
2125 | } |
2126 | ||
339d0354 | 2127 | /* Replay one dir item from a BTRFS_DIR_INDEX_KEY key. */ |
e02119d5 CM |
2128 | static noinline int replay_one_dir_item(struct btrfs_trans_handle *trans, |
2129 | struct btrfs_root *root, | |
2130 | struct btrfs_path *path, | |
2131 | struct extent_buffer *eb, int slot, | |
2132 | struct btrfs_key *key) | |
2133 | { | |
339d0354 | 2134 | int ret; |
e02119d5 | 2135 | struct btrfs_dir_item *di; |
e02119d5 | 2136 | |
339d0354 FM |
2137 | /* We only log dir index keys, which only contain a single dir item. */ |
2138 | ASSERT(key->type == BTRFS_DIR_INDEX_KEY); | |
bb53eda9 | 2139 | |
339d0354 FM |
2140 | di = btrfs_item_ptr(eb, slot, struct btrfs_dir_item); |
2141 | ret = replay_one_name(trans, root, path, eb, di, key); | |
2142 | if (ret < 0) | |
2143 | return ret; | |
bb53eda9 | 2144 | |
339d0354 FM |
2145 | /* |
2146 | * If this entry refers to a non-directory (directories can not have a | |
2147 | * link count > 1) and it was added in the transaction that was not | |
2148 | * committed, make sure we fixup the link count of the inode the entry | |
2149 | * points to. Otherwise something like the following would result in a | |
2150 | * directory pointing to an inode with a wrong link that does not account | |
2151 | * for this dir entry: | |
2152 | * | |
2153 | * mkdir testdir | |
2154 | * touch testdir/foo | |
2155 | * touch testdir/bar | |
2156 | * sync | |
2157 | * | |
2158 | * ln testdir/bar testdir/bar_link | |
2159 | * ln testdir/foo testdir/foo_link | |
2160 | * xfs_io -c "fsync" testdir/bar | |
2161 | * | |
2162 | * <power failure> | |
2163 | * | |
2164 | * mount fs, log replay happens | |
2165 | * | |
2166 | * File foo would remain with a link count of 1 when it has two entries | |
2167 | * pointing to it in the directory testdir. This would make it impossible | |
2168 | * to ever delete the parent directory has it would result in stale | |
2169 | * dentries that can never be deleted. | |
2170 | */ | |
2171 | if (ret == 1 && btrfs_dir_type(eb, di) != BTRFS_FT_DIR) { | |
2172 | struct btrfs_path *fixup_path; | |
2173 | struct btrfs_key di_key; | |
bb53eda9 | 2174 | |
339d0354 FM |
2175 | fixup_path = btrfs_alloc_path(); |
2176 | if (!fixup_path) | |
2177 | return -ENOMEM; | |
2178 | ||
2179 | btrfs_dir_item_key_to_cpu(eb, di, &di_key); | |
2180 | ret = link_to_fixup_dir(trans, root, fixup_path, di_key.objectid); | |
2181 | btrfs_free_path(fixup_path); | |
e02119d5 | 2182 | } |
339d0354 | 2183 | |
bb53eda9 | 2184 | return ret; |
e02119d5 CM |
2185 | } |
2186 | ||
2187 | /* | |
2188 | * directory replay has two parts. There are the standard directory | |
2189 | * items in the log copied from the subvolume, and range items | |
2190 | * created in the log while the subvolume was logged. | |
2191 | * | |
2192 | * The range items tell us which parts of the key space the log | |
2193 | * is authoritative for. During replay, if a key in the subvolume | |
2194 | * directory is in a logged range item, but not actually in the log | |
2195 | * that means it was deleted from the directory before the fsync | |
2196 | * and should be removed. | |
2197 | */ | |
2198 | static noinline int find_dir_range(struct btrfs_root *root, | |
2199 | struct btrfs_path *path, | |
ccae4a19 | 2200 | u64 dirid, |
e02119d5 CM |
2201 | u64 *start_ret, u64 *end_ret) |
2202 | { | |
2203 | struct btrfs_key key; | |
2204 | u64 found_end; | |
2205 | struct btrfs_dir_log_item *item; | |
2206 | int ret; | |
2207 | int nritems; | |
2208 | ||
2209 | if (*start_ret == (u64)-1) | |
2210 | return 1; | |
2211 | ||
2212 | key.objectid = dirid; | |
ccae4a19 | 2213 | key.type = BTRFS_DIR_LOG_INDEX_KEY; |
e02119d5 CM |
2214 | key.offset = *start_ret; |
2215 | ||
2216 | ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); | |
2217 | if (ret < 0) | |
2218 | goto out; | |
2219 | if (ret > 0) { | |
2220 | if (path->slots[0] == 0) | |
2221 | goto out; | |
2222 | path->slots[0]--; | |
2223 | } | |
2224 | if (ret != 0) | |
2225 | btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]); | |
2226 | ||
ccae4a19 | 2227 | if (key.type != BTRFS_DIR_LOG_INDEX_KEY || key.objectid != dirid) { |
e02119d5 CM |
2228 | ret = 1; |
2229 | goto next; | |
2230 | } | |
2231 | item = btrfs_item_ptr(path->nodes[0], path->slots[0], | |
2232 | struct btrfs_dir_log_item); | |
2233 | found_end = btrfs_dir_log_end(path->nodes[0], item); | |
2234 | ||
2235 | if (*start_ret >= key.offset && *start_ret <= found_end) { | |
2236 | ret = 0; | |
2237 | *start_ret = key.offset; | |
2238 | *end_ret = found_end; | |
2239 | goto out; | |
2240 | } | |
2241 | ret = 1; | |
2242 | next: | |
2243 | /* check the next slot in the tree to see if it is a valid item */ | |
2244 | nritems = btrfs_header_nritems(path->nodes[0]); | |
2a7bf53f | 2245 | path->slots[0]++; |
e02119d5 CM |
2246 | if (path->slots[0] >= nritems) { |
2247 | ret = btrfs_next_leaf(root, path); | |
2248 | if (ret) | |
2249 | goto out; | |
e02119d5 CM |
2250 | } |
2251 | ||
2252 | btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]); | |
2253 | ||
ccae4a19 | 2254 | if (key.type != BTRFS_DIR_LOG_INDEX_KEY || key.objectid != dirid) { |
e02119d5 CM |
2255 | ret = 1; |
2256 | goto out; | |
2257 | } | |
2258 | item = btrfs_item_ptr(path->nodes[0], path->slots[0], | |
2259 | struct btrfs_dir_log_item); | |
2260 | found_end = btrfs_dir_log_end(path->nodes[0], item); | |
2261 | *start_ret = key.offset; | |
2262 | *end_ret = found_end; | |
2263 | ret = 0; | |
2264 | out: | |
b3b4aa74 | 2265 | btrfs_release_path(path); |
e02119d5 CM |
2266 | return ret; |
2267 | } | |
2268 | ||
2269 | /* | |
2270 | * this looks for a given directory item in the log. If the directory | |
2271 | * item is not in the log, the item is removed and the inode it points | |
2272 | * to is unlinked | |
2273 | */ | |
2274 | static noinline int check_item_in_log(struct btrfs_trans_handle *trans, | |
e02119d5 CM |
2275 | struct btrfs_root *log, |
2276 | struct btrfs_path *path, | |
2277 | struct btrfs_path *log_path, | |
2278 | struct inode *dir, | |
2279 | struct btrfs_key *dir_key) | |
2280 | { | |
d1ed82f3 | 2281 | struct btrfs_root *root = BTRFS_I(dir)->root; |
e02119d5 CM |
2282 | int ret; |
2283 | struct extent_buffer *eb; | |
2284 | int slot; | |
e02119d5 | 2285 | struct btrfs_dir_item *di; |
e02119d5 | 2286 | int name_len; |
e02119d5 | 2287 | char *name; |
ccae4a19 | 2288 | struct inode *inode = NULL; |
e02119d5 CM |
2289 | struct btrfs_key location; |
2290 | ||
ccae4a19 | 2291 | /* |
143823cf | 2292 | * Currently we only log dir index keys. Even if we replay a log created |
ccae4a19 FM |
2293 | * by an older kernel that logged both dir index and dir item keys, all |
2294 | * we need to do is process the dir index keys, we (and our caller) can | |
2295 | * safely ignore dir item keys (key type BTRFS_DIR_ITEM_KEY). | |
2296 | */ | |
2297 | ASSERT(dir_key->type == BTRFS_DIR_INDEX_KEY); | |
2298 | ||
e02119d5 CM |
2299 | eb = path->nodes[0]; |
2300 | slot = path->slots[0]; | |
ccae4a19 FM |
2301 | di = btrfs_item_ptr(eb, slot, struct btrfs_dir_item); |
2302 | name_len = btrfs_dir_name_len(eb, di); | |
2303 | name = kmalloc(name_len, GFP_NOFS); | |
2304 | if (!name) { | |
2305 | ret = -ENOMEM; | |
2306 | goto out; | |
2307 | } | |
e02119d5 | 2308 | |
ccae4a19 | 2309 | read_extent_buffer(eb, name, (unsigned long)(di + 1), name_len); |
3650860b | 2310 | |
ccae4a19 FM |
2311 | if (log) { |
2312 | struct btrfs_dir_item *log_di; | |
e02119d5 | 2313 | |
ccae4a19 FM |
2314 | log_di = btrfs_lookup_dir_index_item(trans, log, log_path, |
2315 | dir_key->objectid, | |
2316 | dir_key->offset, | |
2317 | name, name_len, 0); | |
2318 | if (IS_ERR(log_di)) { | |
2319 | ret = PTR_ERR(log_di); | |
2320 | goto out; | |
2321 | } else if (log_di) { | |
2322 | /* The dentry exists in the log, we have nothing to do. */ | |
e02119d5 CM |
2323 | ret = 0; |
2324 | goto out; | |
2325 | } | |
ccae4a19 | 2326 | } |
e02119d5 | 2327 | |
ccae4a19 FM |
2328 | btrfs_dir_item_key_to_cpu(eb, di, &location); |
2329 | btrfs_release_path(path); | |
2330 | btrfs_release_path(log_path); | |
2331 | inode = read_one_inode(root, location.objectid); | |
2332 | if (!inode) { | |
2333 | ret = -EIO; | |
2334 | goto out; | |
e02119d5 | 2335 | } |
ccae4a19 FM |
2336 | |
2337 | ret = link_to_fixup_dir(trans, root, path, location.objectid); | |
2338 | if (ret) | |
2339 | goto out; | |
2340 | ||
2341 | inc_nlink(inode); | |
313ab753 FM |
2342 | ret = unlink_inode_for_log_replay(trans, BTRFS_I(dir), BTRFS_I(inode), |
2343 | name, name_len); | |
ccae4a19 FM |
2344 | /* |
2345 | * Unlike dir item keys, dir index keys can only have one name (entry) in | |
2346 | * them, as there are no key collisions since each key has a unique offset | |
2347 | * (an index number), so we're done. | |
2348 | */ | |
e02119d5 | 2349 | out: |
b3b4aa74 DS |
2350 | btrfs_release_path(path); |
2351 | btrfs_release_path(log_path); | |
ccae4a19 FM |
2352 | kfree(name); |
2353 | iput(inode); | |
e02119d5 CM |
2354 | return ret; |
2355 | } | |
2356 | ||
4f764e51 FM |
2357 | static int replay_xattr_deletes(struct btrfs_trans_handle *trans, |
2358 | struct btrfs_root *root, | |
2359 | struct btrfs_root *log, | |
2360 | struct btrfs_path *path, | |
2361 | const u64 ino) | |
2362 | { | |
2363 | struct btrfs_key search_key; | |
2364 | struct btrfs_path *log_path; | |
2365 | int i; | |
2366 | int nritems; | |
2367 | int ret; | |
2368 | ||
2369 | log_path = btrfs_alloc_path(); | |
2370 | if (!log_path) | |
2371 | return -ENOMEM; | |
2372 | ||
2373 | search_key.objectid = ino; | |
2374 | search_key.type = BTRFS_XATTR_ITEM_KEY; | |
2375 | search_key.offset = 0; | |
2376 | again: | |
2377 | ret = btrfs_search_slot(NULL, root, &search_key, path, 0, 0); | |
2378 | if (ret < 0) | |
2379 | goto out; | |
2380 | process_leaf: | |
2381 | nritems = btrfs_header_nritems(path->nodes[0]); | |
2382 | for (i = path->slots[0]; i < nritems; i++) { | |
2383 | struct btrfs_key key; | |
2384 | struct btrfs_dir_item *di; | |
2385 | struct btrfs_dir_item *log_di; | |
2386 | u32 total_size; | |
2387 | u32 cur; | |
2388 | ||
2389 | btrfs_item_key_to_cpu(path->nodes[0], &key, i); | |
2390 | if (key.objectid != ino || key.type != BTRFS_XATTR_ITEM_KEY) { | |
2391 | ret = 0; | |
2392 | goto out; | |
2393 | } | |
2394 | ||
2395 | di = btrfs_item_ptr(path->nodes[0], i, struct btrfs_dir_item); | |
3212fa14 | 2396 | total_size = btrfs_item_size(path->nodes[0], i); |
4f764e51 FM |
2397 | cur = 0; |
2398 | while (cur < total_size) { | |
2399 | u16 name_len = btrfs_dir_name_len(path->nodes[0], di); | |
2400 | u16 data_len = btrfs_dir_data_len(path->nodes[0], di); | |
2401 | u32 this_len = sizeof(*di) + name_len + data_len; | |
2402 | char *name; | |
2403 | ||
2404 | name = kmalloc(name_len, GFP_NOFS); | |
2405 | if (!name) { | |
2406 | ret = -ENOMEM; | |
2407 | goto out; | |
2408 | } | |
2409 | read_extent_buffer(path->nodes[0], name, | |
2410 | (unsigned long)(di + 1), name_len); | |
2411 | ||
2412 | log_di = btrfs_lookup_xattr(NULL, log, log_path, ino, | |
2413 | name, name_len, 0); | |
2414 | btrfs_release_path(log_path); | |
2415 | if (!log_di) { | |
2416 | /* Doesn't exist in log tree, so delete it. */ | |
2417 | btrfs_release_path(path); | |
2418 | di = btrfs_lookup_xattr(trans, root, path, ino, | |
2419 | name, name_len, -1); | |
2420 | kfree(name); | |
2421 | if (IS_ERR(di)) { | |
2422 | ret = PTR_ERR(di); | |
2423 | goto out; | |
2424 | } | |
2425 | ASSERT(di); | |
2426 | ret = btrfs_delete_one_dir_name(trans, root, | |
2427 | path, di); | |
2428 | if (ret) | |
2429 | goto out; | |
2430 | btrfs_release_path(path); | |
2431 | search_key = key; | |
2432 | goto again; | |
2433 | } | |
2434 | kfree(name); | |
2435 | if (IS_ERR(log_di)) { | |
2436 | ret = PTR_ERR(log_di); | |
2437 | goto out; | |
2438 | } | |
2439 | cur += this_len; | |
2440 | di = (struct btrfs_dir_item *)((char *)di + this_len); | |
2441 | } | |
2442 | } | |
2443 | ret = btrfs_next_leaf(root, path); | |
2444 | if (ret > 0) | |
2445 | ret = 0; | |
2446 | else if (ret == 0) | |
2447 | goto process_leaf; | |
2448 | out: | |
2449 | btrfs_free_path(log_path); | |
2450 | btrfs_release_path(path); | |
2451 | return ret; | |
2452 | } | |
2453 | ||
2454 | ||
e02119d5 CM |
2455 | /* |
2456 | * deletion replay happens before we copy any new directory items | |
2457 | * out of the log or out of backreferences from inodes. It | |
2458 | * scans the log to find ranges of keys that log is authoritative for, | |
2459 | * and then scans the directory to find items in those ranges that are | |
2460 | * not present in the log. | |
2461 | * | |
2462 | * Anything we don't find in the log is unlinked and removed from the | |
2463 | * directory. | |
2464 | */ | |
2465 | static noinline int replay_dir_deletes(struct btrfs_trans_handle *trans, | |
2466 | struct btrfs_root *root, | |
2467 | struct btrfs_root *log, | |
2468 | struct btrfs_path *path, | |
12fcfd22 | 2469 | u64 dirid, int del_all) |
e02119d5 CM |
2470 | { |
2471 | u64 range_start; | |
2472 | u64 range_end; | |
e02119d5 CM |
2473 | int ret = 0; |
2474 | struct btrfs_key dir_key; | |
2475 | struct btrfs_key found_key; | |
2476 | struct btrfs_path *log_path; | |
2477 | struct inode *dir; | |
2478 | ||
2479 | dir_key.objectid = dirid; | |
ccae4a19 | 2480 | dir_key.type = BTRFS_DIR_INDEX_KEY; |
e02119d5 CM |
2481 | log_path = btrfs_alloc_path(); |
2482 | if (!log_path) | |
2483 | return -ENOMEM; | |
2484 | ||
2485 | dir = read_one_inode(root, dirid); | |
2486 | /* it isn't an error if the inode isn't there, that can happen | |
2487 | * because we replay the deletes before we copy in the inode item | |
2488 | * from the log | |
2489 | */ | |
2490 | if (!dir) { | |
2491 | btrfs_free_path(log_path); | |
2492 | return 0; | |
2493 | } | |
ccae4a19 | 2494 | |
e02119d5 CM |
2495 | range_start = 0; |
2496 | range_end = 0; | |
d397712b | 2497 | while (1) { |
12fcfd22 CM |
2498 | if (del_all) |
2499 | range_end = (u64)-1; | |
2500 | else { | |
ccae4a19 | 2501 | ret = find_dir_range(log, path, dirid, |
12fcfd22 | 2502 | &range_start, &range_end); |
10adb115 FM |
2503 | if (ret < 0) |
2504 | goto out; | |
2505 | else if (ret > 0) | |
12fcfd22 CM |
2506 | break; |
2507 | } | |
e02119d5 CM |
2508 | |
2509 | dir_key.offset = range_start; | |
d397712b | 2510 | while (1) { |
e02119d5 CM |
2511 | int nritems; |
2512 | ret = btrfs_search_slot(NULL, root, &dir_key, path, | |
2513 | 0, 0); | |
2514 | if (ret < 0) | |
2515 | goto out; | |
2516 | ||
2517 | nritems = btrfs_header_nritems(path->nodes[0]); | |
2518 | if (path->slots[0] >= nritems) { | |
2519 | ret = btrfs_next_leaf(root, path); | |
b98def7c | 2520 | if (ret == 1) |
e02119d5 | 2521 | break; |
b98def7c LB |
2522 | else if (ret < 0) |
2523 | goto out; | |
e02119d5 CM |
2524 | } |
2525 | btrfs_item_key_to_cpu(path->nodes[0], &found_key, | |
2526 | path->slots[0]); | |
2527 | if (found_key.objectid != dirid || | |
ccae4a19 FM |
2528 | found_key.type != dir_key.type) { |
2529 | ret = 0; | |
2530 | goto out; | |
2531 | } | |
e02119d5 CM |
2532 | |
2533 | if (found_key.offset > range_end) | |
2534 | break; | |
2535 | ||
d1ed82f3 | 2536 | ret = check_item_in_log(trans, log, path, |
12fcfd22 CM |
2537 | log_path, dir, |
2538 | &found_key); | |
3650860b JB |
2539 | if (ret) |
2540 | goto out; | |
e02119d5 CM |
2541 | if (found_key.offset == (u64)-1) |
2542 | break; | |
2543 | dir_key.offset = found_key.offset + 1; | |
2544 | } | |
b3b4aa74 | 2545 | btrfs_release_path(path); |
e02119d5 CM |
2546 | if (range_end == (u64)-1) |
2547 | break; | |
2548 | range_start = range_end + 1; | |
2549 | } | |
e02119d5 | 2550 | ret = 0; |
e02119d5 | 2551 | out: |
b3b4aa74 | 2552 | btrfs_release_path(path); |
e02119d5 CM |
2553 | btrfs_free_path(log_path); |
2554 | iput(dir); | |
2555 | return ret; | |
2556 | } | |
2557 | ||
2558 | /* | |
2559 | * the process_func used to replay items from the log tree. This | |
2560 | * gets called in two different stages. The first stage just looks | |
2561 | * for inodes and makes sure they are all copied into the subvolume. | |
2562 | * | |
2563 | * The second stage copies all the other item types from the log into | |
2564 | * the subvolume. The two stage approach is slower, but gets rid of | |
2565 | * lots of complexity around inodes referencing other inodes that exist | |
2566 | * only in the log (references come from either directory items or inode | |
2567 | * back refs). | |
2568 | */ | |
2569 | static int replay_one_buffer(struct btrfs_root *log, struct extent_buffer *eb, | |
581c1760 | 2570 | struct walk_control *wc, u64 gen, int level) |
e02119d5 CM |
2571 | { |
2572 | int nritems; | |
2573 | struct btrfs_path *path; | |
2574 | struct btrfs_root *root = wc->replay_dest; | |
2575 | struct btrfs_key key; | |
e02119d5 CM |
2576 | int i; |
2577 | int ret; | |
2578 | ||
6a2e9dc4 | 2579 | ret = btrfs_read_extent_buffer(eb, gen, level, NULL); |
018642a1 TI |
2580 | if (ret) |
2581 | return ret; | |
e02119d5 CM |
2582 | |
2583 | level = btrfs_header_level(eb); | |
2584 | ||
2585 | if (level != 0) | |
2586 | return 0; | |
2587 | ||
2588 | path = btrfs_alloc_path(); | |
1e5063d0 MF |
2589 | if (!path) |
2590 | return -ENOMEM; | |
e02119d5 CM |
2591 | |
2592 | nritems = btrfs_header_nritems(eb); | |
2593 | for (i = 0; i < nritems; i++) { | |
2594 | btrfs_item_key_to_cpu(eb, &key, i); | |
e02119d5 CM |
2595 | |
2596 | /* inode keys are done during the first stage */ | |
2597 | if (key.type == BTRFS_INODE_ITEM_KEY && | |
2598 | wc->stage == LOG_WALK_REPLAY_INODES) { | |
e02119d5 CM |
2599 | struct btrfs_inode_item *inode_item; |
2600 | u32 mode; | |
2601 | ||
2602 | inode_item = btrfs_item_ptr(eb, i, | |
2603 | struct btrfs_inode_item); | |
f2d72f42 FM |
2604 | /* |
2605 | * If we have a tmpfile (O_TMPFILE) that got fsync'ed | |
2606 | * and never got linked before the fsync, skip it, as | |
2607 | * replaying it is pointless since it would be deleted | |
2608 | * later. We skip logging tmpfiles, but it's always | |
2609 | * possible we are replaying a log created with a kernel | |
2610 | * that used to log tmpfiles. | |
2611 | */ | |
2612 | if (btrfs_inode_nlink(eb, inode_item) == 0) { | |
2613 | wc->ignore_cur_inode = true; | |
2614 | continue; | |
2615 | } else { | |
2616 | wc->ignore_cur_inode = false; | |
2617 | } | |
4f764e51 FM |
2618 | ret = replay_xattr_deletes(wc->trans, root, log, |
2619 | path, key.objectid); | |
2620 | if (ret) | |
2621 | break; | |
e02119d5 CM |
2622 | mode = btrfs_inode_mode(eb, inode_item); |
2623 | if (S_ISDIR(mode)) { | |
2624 | ret = replay_dir_deletes(wc->trans, | |
12fcfd22 | 2625 | root, log, path, key.objectid, 0); |
b50c6e25 JB |
2626 | if (ret) |
2627 | break; | |
e02119d5 CM |
2628 | } |
2629 | ret = overwrite_item(wc->trans, root, path, | |
2630 | eb, i, &key); | |
b50c6e25 JB |
2631 | if (ret) |
2632 | break; | |
e02119d5 | 2633 | |
471d557a FM |
2634 | /* |
2635 | * Before replaying extents, truncate the inode to its | |
2636 | * size. We need to do it now and not after log replay | |
2637 | * because before an fsync we can have prealloc extents | |
2638 | * added beyond the inode's i_size. If we did it after, | |
2639 | * through orphan cleanup for example, we would drop | |
2640 | * those prealloc extents just after replaying them. | |
e02119d5 CM |
2641 | */ |
2642 | if (S_ISREG(mode)) { | |
5893dfb9 | 2643 | struct btrfs_drop_extents_args drop_args = { 0 }; |
471d557a FM |
2644 | struct inode *inode; |
2645 | u64 from; | |
2646 | ||
2647 | inode = read_one_inode(root, key.objectid); | |
2648 | if (!inode) { | |
2649 | ret = -EIO; | |
2650 | break; | |
2651 | } | |
2652 | from = ALIGN(i_size_read(inode), | |
2653 | root->fs_info->sectorsize); | |
5893dfb9 FM |
2654 | drop_args.start = from; |
2655 | drop_args.end = (u64)-1; | |
2656 | drop_args.drop_cache = true; | |
2657 | ret = btrfs_drop_extents(wc->trans, root, | |
2658 | BTRFS_I(inode), | |
2659 | &drop_args); | |
471d557a | 2660 | if (!ret) { |
2766ff61 FM |
2661 | inode_sub_bytes(inode, |
2662 | drop_args.bytes_found); | |
f2d72f42 | 2663 | /* Update the inode's nbytes. */ |
471d557a | 2664 | ret = btrfs_update_inode(wc->trans, |
9a56fcd1 | 2665 | root, BTRFS_I(inode)); |
471d557a FM |
2666 | } |
2667 | iput(inode); | |
b50c6e25 JB |
2668 | if (ret) |
2669 | break; | |
e02119d5 | 2670 | } |
c71bf099 | 2671 | |
e02119d5 CM |
2672 | ret = link_to_fixup_dir(wc->trans, root, |
2673 | path, key.objectid); | |
b50c6e25 JB |
2674 | if (ret) |
2675 | break; | |
e02119d5 | 2676 | } |
dd8e7217 | 2677 | |
f2d72f42 FM |
2678 | if (wc->ignore_cur_inode) |
2679 | continue; | |
2680 | ||
dd8e7217 JB |
2681 | if (key.type == BTRFS_DIR_INDEX_KEY && |
2682 | wc->stage == LOG_WALK_REPLAY_DIR_INDEX) { | |
2683 | ret = replay_one_dir_item(wc->trans, root, path, | |
2684 | eb, i, &key); | |
2685 | if (ret) | |
2686 | break; | |
2687 | } | |
2688 | ||
e02119d5 CM |
2689 | if (wc->stage < LOG_WALK_REPLAY_ALL) |
2690 | continue; | |
2691 | ||
2692 | /* these keys are simply copied */ | |
2693 | if (key.type == BTRFS_XATTR_ITEM_KEY) { | |
2694 | ret = overwrite_item(wc->trans, root, path, | |
2695 | eb, i, &key); | |
b50c6e25 JB |
2696 | if (ret) |
2697 | break; | |
2da1c669 LB |
2698 | } else if (key.type == BTRFS_INODE_REF_KEY || |
2699 | key.type == BTRFS_INODE_EXTREF_KEY) { | |
f186373f MF |
2700 | ret = add_inode_ref(wc->trans, root, log, path, |
2701 | eb, i, &key); | |
b50c6e25 JB |
2702 | if (ret && ret != -ENOENT) |
2703 | break; | |
2704 | ret = 0; | |
e02119d5 CM |
2705 | } else if (key.type == BTRFS_EXTENT_DATA_KEY) { |
2706 | ret = replay_one_extent(wc->trans, root, path, | |
2707 | eb, i, &key); | |
b50c6e25 JB |
2708 | if (ret) |
2709 | break; | |
e02119d5 | 2710 | } |
339d0354 FM |
2711 | /* |
2712 | * We don't log BTRFS_DIR_ITEM_KEY keys anymore, only the | |
2713 | * BTRFS_DIR_INDEX_KEY items which we use to derive the | |
2714 | * BTRFS_DIR_ITEM_KEY items. If we are replaying a log from an | |
2715 | * older kernel with such keys, ignore them. | |
2716 | */ | |
e02119d5 CM |
2717 | } |
2718 | btrfs_free_path(path); | |
b50c6e25 | 2719 | return ret; |
e02119d5 CM |
2720 | } |
2721 | ||
6787bb9f NB |
2722 | /* |
2723 | * Correctly adjust the reserved bytes occupied by a log tree extent buffer | |
2724 | */ | |
2725 | static void unaccount_log_buffer(struct btrfs_fs_info *fs_info, u64 start) | |
2726 | { | |
2727 | struct btrfs_block_group *cache; | |
2728 | ||
2729 | cache = btrfs_lookup_block_group(fs_info, start); | |
2730 | if (!cache) { | |
2731 | btrfs_err(fs_info, "unable to find block group for %llu", start); | |
2732 | return; | |
2733 | } | |
2734 | ||
2735 | spin_lock(&cache->space_info->lock); | |
2736 | spin_lock(&cache->lock); | |
2737 | cache->reserved -= fs_info->nodesize; | |
2738 | cache->space_info->bytes_reserved -= fs_info->nodesize; | |
2739 | spin_unlock(&cache->lock); | |
2740 | spin_unlock(&cache->space_info->lock); | |
2741 | ||
2742 | btrfs_put_block_group(cache); | |
2743 | } | |
2744 | ||
d397712b | 2745 | static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans, |
e02119d5 CM |
2746 | struct btrfs_root *root, |
2747 | struct btrfs_path *path, int *level, | |
2748 | struct walk_control *wc) | |
2749 | { | |
0b246afa | 2750 | struct btrfs_fs_info *fs_info = root->fs_info; |
e02119d5 CM |
2751 | u64 bytenr; |
2752 | u64 ptr_gen; | |
2753 | struct extent_buffer *next; | |
2754 | struct extent_buffer *cur; | |
e02119d5 CM |
2755 | u32 blocksize; |
2756 | int ret = 0; | |
2757 | ||
d397712b | 2758 | while (*level > 0) { |
581c1760 QW |
2759 | struct btrfs_key first_key; |
2760 | ||
e02119d5 CM |
2761 | cur = path->nodes[*level]; |
2762 | ||
fae7f21c | 2763 | WARN_ON(btrfs_header_level(cur) != *level); |
e02119d5 CM |
2764 | |
2765 | if (path->slots[*level] >= | |
2766 | btrfs_header_nritems(cur)) | |
2767 | break; | |
2768 | ||
2769 | bytenr = btrfs_node_blockptr(cur, path->slots[*level]); | |
2770 | ptr_gen = btrfs_node_ptr_generation(cur, path->slots[*level]); | |
581c1760 | 2771 | btrfs_node_key_to_cpu(cur, &first_key, path->slots[*level]); |
0b246afa | 2772 | blocksize = fs_info->nodesize; |
e02119d5 | 2773 | |
3fbaf258 JB |
2774 | next = btrfs_find_create_tree_block(fs_info, bytenr, |
2775 | btrfs_header_owner(cur), | |
2776 | *level - 1); | |
c871b0f2 LB |
2777 | if (IS_ERR(next)) |
2778 | return PTR_ERR(next); | |
e02119d5 | 2779 | |
e02119d5 | 2780 | if (*level == 1) { |
581c1760 QW |
2781 | ret = wc->process_func(root, next, wc, ptr_gen, |
2782 | *level - 1); | |
b50c6e25 JB |
2783 | if (ret) { |
2784 | free_extent_buffer(next); | |
1e5063d0 | 2785 | return ret; |
b50c6e25 | 2786 | } |
4a500fd1 | 2787 | |
e02119d5 CM |
2788 | path->slots[*level]++; |
2789 | if (wc->free) { | |
6a2e9dc4 | 2790 | ret = btrfs_read_extent_buffer(next, ptr_gen, |
581c1760 | 2791 | *level - 1, &first_key); |
018642a1 TI |
2792 | if (ret) { |
2793 | free_extent_buffer(next); | |
2794 | return ret; | |
2795 | } | |
e02119d5 | 2796 | |
681ae509 JB |
2797 | if (trans) { |
2798 | btrfs_tree_lock(next); | |
6a884d7d | 2799 | btrfs_clean_tree_block(next); |
681ae509 JB |
2800 | btrfs_wait_tree_block_writeback(next); |
2801 | btrfs_tree_unlock(next); | |
7bfc1007 | 2802 | ret = btrfs_pin_reserved_extent(trans, |
10e958d5 NB |
2803 | bytenr, blocksize); |
2804 | if (ret) { | |
2805 | free_extent_buffer(next); | |
2806 | return ret; | |
2807 | } | |
d3575156 NA |
2808 | btrfs_redirty_list_add( |
2809 | trans->transaction, next); | |
1846430c LB |
2810 | } else { |
2811 | if (test_and_clear_bit(EXTENT_BUFFER_DIRTY, &next->bflags)) | |
2812 | clear_extent_buffer_dirty(next); | |
10e958d5 | 2813 | unaccount_log_buffer(fs_info, bytenr); |
3650860b | 2814 | } |
e02119d5 CM |
2815 | } |
2816 | free_extent_buffer(next); | |
2817 | continue; | |
2818 | } | |
6a2e9dc4 | 2819 | ret = btrfs_read_extent_buffer(next, ptr_gen, *level - 1, &first_key); |
018642a1 TI |
2820 | if (ret) { |
2821 | free_extent_buffer(next); | |
2822 | return ret; | |
2823 | } | |
e02119d5 | 2824 | |
e02119d5 CM |
2825 | if (path->nodes[*level-1]) |
2826 | free_extent_buffer(path->nodes[*level-1]); | |
2827 | path->nodes[*level-1] = next; | |
2828 | *level = btrfs_header_level(next); | |
2829 | path->slots[*level] = 0; | |
2830 | cond_resched(); | |
2831 | } | |
4a500fd1 | 2832 | path->slots[*level] = btrfs_header_nritems(path->nodes[*level]); |
e02119d5 CM |
2833 | |
2834 | cond_resched(); | |
2835 | return 0; | |
2836 | } | |
2837 | ||
d397712b | 2838 | static noinline int walk_up_log_tree(struct btrfs_trans_handle *trans, |
e02119d5 CM |
2839 | struct btrfs_root *root, |
2840 | struct btrfs_path *path, int *level, | |
2841 | struct walk_control *wc) | |
2842 | { | |
0b246afa | 2843 | struct btrfs_fs_info *fs_info = root->fs_info; |
e02119d5 CM |
2844 | int i; |
2845 | int slot; | |
2846 | int ret; | |
2847 | ||
d397712b | 2848 | for (i = *level; i < BTRFS_MAX_LEVEL - 1 && path->nodes[i]; i++) { |
e02119d5 | 2849 | slot = path->slots[i]; |
4a500fd1 | 2850 | if (slot + 1 < btrfs_header_nritems(path->nodes[i])) { |
e02119d5 CM |
2851 | path->slots[i]++; |
2852 | *level = i; | |
2853 | WARN_ON(*level == 0); | |
2854 | return 0; | |
2855 | } else { | |
1e5063d0 | 2856 | ret = wc->process_func(root, path->nodes[*level], wc, |
581c1760 QW |
2857 | btrfs_header_generation(path->nodes[*level]), |
2858 | *level); | |
1e5063d0 MF |
2859 | if (ret) |
2860 | return ret; | |
2861 | ||
e02119d5 CM |
2862 | if (wc->free) { |
2863 | struct extent_buffer *next; | |
2864 | ||
2865 | next = path->nodes[*level]; | |
2866 | ||
681ae509 JB |
2867 | if (trans) { |
2868 | btrfs_tree_lock(next); | |
6a884d7d | 2869 | btrfs_clean_tree_block(next); |
681ae509 JB |
2870 | btrfs_wait_tree_block_writeback(next); |
2871 | btrfs_tree_unlock(next); | |
7bfc1007 | 2872 | ret = btrfs_pin_reserved_extent(trans, |
10e958d5 NB |
2873 | path->nodes[*level]->start, |
2874 | path->nodes[*level]->len); | |
2875 | if (ret) | |
2876 | return ret; | |
84c25448 NA |
2877 | btrfs_redirty_list_add(trans->transaction, |
2878 | next); | |
1846430c LB |
2879 | } else { |
2880 | if (test_and_clear_bit(EXTENT_BUFFER_DIRTY, &next->bflags)) | |
2881 | clear_extent_buffer_dirty(next); | |
e02119d5 | 2882 | |
10e958d5 NB |
2883 | unaccount_log_buffer(fs_info, |
2884 | path->nodes[*level]->start); | |
2885 | } | |
e02119d5 CM |
2886 | } |
2887 | free_extent_buffer(path->nodes[*level]); | |
2888 | path->nodes[*level] = NULL; | |
2889 | *level = i + 1; | |
2890 | } | |
2891 | } | |
2892 | return 1; | |
2893 | } | |
2894 | ||
2895 | /* | |
2896 | * drop the reference count on the tree rooted at 'snap'. This traverses | |
2897 | * the tree freeing any blocks that have a ref count of zero after being | |
2898 | * decremented. | |
2899 | */ | |
2900 | static int walk_log_tree(struct btrfs_trans_handle *trans, | |
2901 | struct btrfs_root *log, struct walk_control *wc) | |
2902 | { | |
2ff7e61e | 2903 | struct btrfs_fs_info *fs_info = log->fs_info; |
e02119d5 CM |
2904 | int ret = 0; |
2905 | int wret; | |
2906 | int level; | |
2907 | struct btrfs_path *path; | |
e02119d5 CM |
2908 | int orig_level; |
2909 | ||
2910 | path = btrfs_alloc_path(); | |
db5b493a TI |
2911 | if (!path) |
2912 | return -ENOMEM; | |
e02119d5 CM |
2913 | |
2914 | level = btrfs_header_level(log->node); | |
2915 | orig_level = level; | |
2916 | path->nodes[level] = log->node; | |
67439dad | 2917 | atomic_inc(&log->node->refs); |
e02119d5 CM |
2918 | path->slots[level] = 0; |
2919 | ||
d397712b | 2920 | while (1) { |
e02119d5 CM |
2921 | wret = walk_down_log_tree(trans, log, path, &level, wc); |
2922 | if (wret > 0) | |
2923 | break; | |
79787eaa | 2924 | if (wret < 0) { |
e02119d5 | 2925 | ret = wret; |
79787eaa JM |
2926 | goto out; |
2927 | } | |
e02119d5 CM |
2928 | |
2929 | wret = walk_up_log_tree(trans, log, path, &level, wc); | |
2930 | if (wret > 0) | |
2931 | break; | |
79787eaa | 2932 | if (wret < 0) { |
e02119d5 | 2933 | ret = wret; |
79787eaa JM |
2934 | goto out; |
2935 | } | |
e02119d5 CM |
2936 | } |
2937 | ||
2938 | /* was the root node processed? if not, catch it here */ | |
2939 | if (path->nodes[orig_level]) { | |
79787eaa | 2940 | ret = wc->process_func(log, path->nodes[orig_level], wc, |
581c1760 QW |
2941 | btrfs_header_generation(path->nodes[orig_level]), |
2942 | orig_level); | |
79787eaa JM |
2943 | if (ret) |
2944 | goto out; | |
e02119d5 CM |
2945 | if (wc->free) { |
2946 | struct extent_buffer *next; | |
2947 | ||
2948 | next = path->nodes[orig_level]; | |
2949 | ||
681ae509 JB |
2950 | if (trans) { |
2951 | btrfs_tree_lock(next); | |
6a884d7d | 2952 | btrfs_clean_tree_block(next); |
681ae509 JB |
2953 | btrfs_wait_tree_block_writeback(next); |
2954 | btrfs_tree_unlock(next); | |
7bfc1007 | 2955 | ret = btrfs_pin_reserved_extent(trans, |
10e958d5 NB |
2956 | next->start, next->len); |
2957 | if (ret) | |
2958 | goto out; | |
84c25448 | 2959 | btrfs_redirty_list_add(trans->transaction, next); |
1846430c LB |
2960 | } else { |
2961 | if (test_and_clear_bit(EXTENT_BUFFER_DIRTY, &next->bflags)) | |
2962 | clear_extent_buffer_dirty(next); | |
10e958d5 | 2963 | unaccount_log_buffer(fs_info, next->start); |
681ae509 | 2964 | } |
e02119d5 CM |
2965 | } |
2966 | } | |
2967 | ||
79787eaa | 2968 | out: |
e02119d5 | 2969 | btrfs_free_path(path); |
e02119d5 CM |
2970 | return ret; |
2971 | } | |
2972 | ||
7237f183 YZ |
2973 | /* |
2974 | * helper function to update the item for a given subvolumes log root | |
2975 | * in the tree of log roots | |
2976 | */ | |
2977 | static int update_log_root(struct btrfs_trans_handle *trans, | |
4203e968 JB |
2978 | struct btrfs_root *log, |
2979 | struct btrfs_root_item *root_item) | |
7237f183 | 2980 | { |
0b246afa | 2981 | struct btrfs_fs_info *fs_info = log->fs_info; |
7237f183 YZ |
2982 | int ret; |
2983 | ||
2984 | if (log->log_transid == 1) { | |
2985 | /* insert root item on the first sync */ | |
0b246afa | 2986 | ret = btrfs_insert_root(trans, fs_info->log_root_tree, |
4203e968 | 2987 | &log->root_key, root_item); |
7237f183 | 2988 | } else { |
0b246afa | 2989 | ret = btrfs_update_root(trans, fs_info->log_root_tree, |
4203e968 | 2990 | &log->root_key, root_item); |
7237f183 YZ |
2991 | } |
2992 | return ret; | |
2993 | } | |
2994 | ||
60d53eb3 | 2995 | static void wait_log_commit(struct btrfs_root *root, int transid) |
e02119d5 CM |
2996 | { |
2997 | DEFINE_WAIT(wait); | |
7237f183 | 2998 | int index = transid % 2; |
e02119d5 | 2999 | |
7237f183 YZ |
3000 | /* |
3001 | * we only allow two pending log transactions at a time, | |
3002 | * so we know that if ours is more than 2 older than the | |
3003 | * current transaction, we're done | |
3004 | */ | |
49e83f57 | 3005 | for (;;) { |
7237f183 YZ |
3006 | prepare_to_wait(&root->log_commit_wait[index], |
3007 | &wait, TASK_UNINTERRUPTIBLE); | |
12fcfd22 | 3008 | |
49e83f57 LB |
3009 | if (!(root->log_transid_committed < transid && |
3010 | atomic_read(&root->log_commit[index]))) | |
3011 | break; | |
12fcfd22 | 3012 | |
49e83f57 LB |
3013 | mutex_unlock(&root->log_mutex); |
3014 | schedule(); | |
7237f183 | 3015 | mutex_lock(&root->log_mutex); |
49e83f57 LB |
3016 | } |
3017 | finish_wait(&root->log_commit_wait[index], &wait); | |
7237f183 YZ |
3018 | } |
3019 | ||
60d53eb3 | 3020 | static void wait_for_writer(struct btrfs_root *root) |
7237f183 YZ |
3021 | { |
3022 | DEFINE_WAIT(wait); | |
8b050d35 | 3023 | |
49e83f57 LB |
3024 | for (;;) { |
3025 | prepare_to_wait(&root->log_writer_wait, &wait, | |
3026 | TASK_UNINTERRUPTIBLE); | |
3027 | if (!atomic_read(&root->log_writers)) | |
3028 | break; | |
3029 | ||
7237f183 | 3030 | mutex_unlock(&root->log_mutex); |
49e83f57 | 3031 | schedule(); |
575849ec | 3032 | mutex_lock(&root->log_mutex); |
7237f183 | 3033 | } |
49e83f57 | 3034 | finish_wait(&root->log_writer_wait, &wait); |
e02119d5 CM |
3035 | } |
3036 | ||
8b050d35 MX |
3037 | static inline void btrfs_remove_log_ctx(struct btrfs_root *root, |
3038 | struct btrfs_log_ctx *ctx) | |
3039 | { | |
8b050d35 MX |
3040 | mutex_lock(&root->log_mutex); |
3041 | list_del_init(&ctx->list); | |
3042 | mutex_unlock(&root->log_mutex); | |
3043 | } | |
3044 | ||
3045 | /* | |
3046 | * Invoked in log mutex context, or be sure there is no other task which | |
3047 | * can access the list. | |
3048 | */ | |
3049 | static inline void btrfs_remove_all_log_ctxs(struct btrfs_root *root, | |
3050 | int index, int error) | |
3051 | { | |
3052 | struct btrfs_log_ctx *ctx; | |
570dd450 | 3053 | struct btrfs_log_ctx *safe; |
8b050d35 | 3054 | |
570dd450 CM |
3055 | list_for_each_entry_safe(ctx, safe, &root->log_ctxs[index], list) { |
3056 | list_del_init(&ctx->list); | |
8b050d35 | 3057 | ctx->log_ret = error; |
570dd450 | 3058 | } |
8b050d35 MX |
3059 | } |
3060 | ||
e02119d5 CM |
3061 | /* |
3062 | * btrfs_sync_log does sends a given tree log down to the disk and | |
3063 | * updates the super blocks to record it. When this call is done, | |
12fcfd22 CM |
3064 | * you know that any inodes previously logged are safely on disk only |
3065 | * if it returns 0. | |
3066 | * | |
3067 | * Any other return value means you need to call btrfs_commit_transaction. | |
3068 | * Some of the edge cases for fsyncing directories that have had unlinks | |
3069 | * or renames done in the past mean that sometimes the only safe | |
3070 | * fsync is to commit the whole FS. When btrfs_sync_log returns -EAGAIN, | |
3071 | * that has happened. | |
e02119d5 CM |
3072 | */ |
3073 | int btrfs_sync_log(struct btrfs_trans_handle *trans, | |
8b050d35 | 3074 | struct btrfs_root *root, struct btrfs_log_ctx *ctx) |
e02119d5 | 3075 | { |
7237f183 YZ |
3076 | int index1; |
3077 | int index2; | |
8cef4e16 | 3078 | int mark; |
e02119d5 | 3079 | int ret; |
0b246afa | 3080 | struct btrfs_fs_info *fs_info = root->fs_info; |
e02119d5 | 3081 | struct btrfs_root *log = root->log_root; |
0b246afa | 3082 | struct btrfs_root *log_root_tree = fs_info->log_root_tree; |
4203e968 | 3083 | struct btrfs_root_item new_root_item; |
bb14a59b | 3084 | int log_transid = 0; |
8b050d35 | 3085 | struct btrfs_log_ctx root_log_ctx; |
c6adc9cc | 3086 | struct blk_plug plug; |
47876f7c FM |
3087 | u64 log_root_start; |
3088 | u64 log_root_level; | |
e02119d5 | 3089 | |
7237f183 | 3090 | mutex_lock(&root->log_mutex); |
d1433deb MX |
3091 | log_transid = ctx->log_transid; |
3092 | if (root->log_transid_committed >= log_transid) { | |
3093 | mutex_unlock(&root->log_mutex); | |
3094 | return ctx->log_ret; | |
3095 | } | |
3096 | ||
3097 | index1 = log_transid % 2; | |
7237f183 | 3098 | if (atomic_read(&root->log_commit[index1])) { |
60d53eb3 | 3099 | wait_log_commit(root, log_transid); |
7237f183 | 3100 | mutex_unlock(&root->log_mutex); |
8b050d35 | 3101 | return ctx->log_ret; |
e02119d5 | 3102 | } |
d1433deb | 3103 | ASSERT(log_transid == root->log_transid); |
7237f183 YZ |
3104 | atomic_set(&root->log_commit[index1], 1); |
3105 | ||
3106 | /* wait for previous tree log sync to complete */ | |
3107 | if (atomic_read(&root->log_commit[(index1 + 1) % 2])) | |
60d53eb3 | 3108 | wait_log_commit(root, log_transid - 1); |
48cab2e0 | 3109 | |
86df7eb9 | 3110 | while (1) { |
2ecb7923 | 3111 | int batch = atomic_read(&root->log_batch); |
cd354ad6 | 3112 | /* when we're on an ssd, just kick the log commit out */ |
0b246afa | 3113 | if (!btrfs_test_opt(fs_info, SSD) && |
27cdeb70 | 3114 | test_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state)) { |
86df7eb9 YZ |
3115 | mutex_unlock(&root->log_mutex); |
3116 | schedule_timeout_uninterruptible(1); | |
3117 | mutex_lock(&root->log_mutex); | |
3118 | } | |
60d53eb3 | 3119 | wait_for_writer(root); |
2ecb7923 | 3120 | if (batch == atomic_read(&root->log_batch)) |
e02119d5 CM |
3121 | break; |
3122 | } | |
e02119d5 | 3123 | |
12fcfd22 | 3124 | /* bail out if we need to do a full commit */ |
4884b8e8 | 3125 | if (btrfs_need_log_full_commit(trans)) { |
f31f09f6 | 3126 | ret = BTRFS_LOG_FORCE_COMMIT; |
12fcfd22 CM |
3127 | mutex_unlock(&root->log_mutex); |
3128 | goto out; | |
3129 | } | |
3130 | ||
8cef4e16 YZ |
3131 | if (log_transid % 2 == 0) |
3132 | mark = EXTENT_DIRTY; | |
3133 | else | |
3134 | mark = EXTENT_NEW; | |
3135 | ||
690587d1 CM |
3136 | /* we start IO on all the marked extents here, but we don't actually |
3137 | * wait for them until later. | |
3138 | */ | |
c6adc9cc | 3139 | blk_start_plug(&plug); |
2ff7e61e | 3140 | ret = btrfs_write_marked_extents(fs_info, &log->dirty_log_pages, mark); |
b528f467 NA |
3141 | /* |
3142 | * -EAGAIN happens when someone, e.g., a concurrent transaction | |
3143 | * commit, writes a dirty extent in this tree-log commit. This | |
3144 | * concurrent write will create a hole writing out the extents, | |
3145 | * and we cannot proceed on a zoned filesystem, requiring | |
3146 | * sequential writing. While we can bail out to a full commit | |
3147 | * here, but we can continue hoping the concurrent writing fills | |
3148 | * the hole. | |
3149 | */ | |
3150 | if (ret == -EAGAIN && btrfs_is_zoned(fs_info)) | |
3151 | ret = 0; | |
79787eaa | 3152 | if (ret) { |
c6adc9cc | 3153 | blk_finish_plug(&plug); |
66642832 | 3154 | btrfs_abort_transaction(trans, ret); |
90787766 | 3155 | btrfs_set_log_full_commit(trans); |
79787eaa JM |
3156 | mutex_unlock(&root->log_mutex); |
3157 | goto out; | |
3158 | } | |
7237f183 | 3159 | |
4203e968 JB |
3160 | /* |
3161 | * We _must_ update under the root->log_mutex in order to make sure we | |
3162 | * have a consistent view of the log root we are trying to commit at | |
3163 | * this moment. | |
3164 | * | |
3165 | * We _must_ copy this into a local copy, because we are not holding the | |
3166 | * log_root_tree->log_mutex yet. This is important because when we | |
3167 | * commit the log_root_tree we must have a consistent view of the | |
3168 | * log_root_tree when we update the super block to point at the | |
3169 | * log_root_tree bytenr. If we update the log_root_tree here we'll race | |
3170 | * with the commit and possibly point at the new block which we may not | |
3171 | * have written out. | |
3172 | */ | |
5d4f98a2 | 3173 | btrfs_set_root_node(&log->root_item, log->node); |
4203e968 | 3174 | memcpy(&new_root_item, &log->root_item, sizeof(new_root_item)); |
7237f183 | 3175 | |
7237f183 YZ |
3176 | root->log_transid++; |
3177 | log->log_transid = root->log_transid; | |
ff782e0a | 3178 | root->log_start_pid = 0; |
7237f183 | 3179 | /* |
8cef4e16 YZ |
3180 | * IO has been started, blocks of the log tree have WRITTEN flag set |
3181 | * in their headers. new modifications of the log will be written to | |
3182 | * new positions. so it's safe to allow log writers to go in. | |
7237f183 YZ |
3183 | */ |
3184 | mutex_unlock(&root->log_mutex); | |
3185 | ||
3ddebf27 | 3186 | if (btrfs_is_zoned(fs_info)) { |
e75f9fd1 | 3187 | mutex_lock(&fs_info->tree_root->log_mutex); |
3ddebf27 NA |
3188 | if (!log_root_tree->node) { |
3189 | ret = btrfs_alloc_log_tree_node(trans, log_root_tree); | |
3190 | if (ret) { | |
ea32af47 | 3191 | mutex_unlock(&fs_info->tree_root->log_mutex); |
50ff5788 | 3192 | blk_finish_plug(&plug); |
3ddebf27 NA |
3193 | goto out; |
3194 | } | |
3195 | } | |
e75f9fd1 | 3196 | mutex_unlock(&fs_info->tree_root->log_mutex); |
3ddebf27 NA |
3197 | } |
3198 | ||
e75f9fd1 NA |
3199 | btrfs_init_log_ctx(&root_log_ctx, NULL); |
3200 | ||
3201 | mutex_lock(&log_root_tree->log_mutex); | |
3202 | ||
e3d3b415 FM |
3203 | index2 = log_root_tree->log_transid % 2; |
3204 | list_add_tail(&root_log_ctx.list, &log_root_tree->log_ctxs[index2]); | |
3205 | root_log_ctx.log_transid = log_root_tree->log_transid; | |
3206 | ||
4203e968 JB |
3207 | /* |
3208 | * Now we are safe to update the log_root_tree because we're under the | |
3209 | * log_mutex, and we're a current writer so we're holding the commit | |
3210 | * open until we drop the log_mutex. | |
3211 | */ | |
3212 | ret = update_log_root(trans, log, &new_root_item); | |
4a500fd1 | 3213 | if (ret) { |
d1433deb MX |
3214 | if (!list_empty(&root_log_ctx.list)) |
3215 | list_del_init(&root_log_ctx.list); | |
3216 | ||
c6adc9cc | 3217 | blk_finish_plug(&plug); |
90787766 | 3218 | btrfs_set_log_full_commit(trans); |
995946dd | 3219 | |
79787eaa | 3220 | if (ret != -ENOSPC) { |
66642832 | 3221 | btrfs_abort_transaction(trans, ret); |
79787eaa JM |
3222 | mutex_unlock(&log_root_tree->log_mutex); |
3223 | goto out; | |
3224 | } | |
bf89d38f | 3225 | btrfs_wait_tree_log_extents(log, mark); |
4a500fd1 | 3226 | mutex_unlock(&log_root_tree->log_mutex); |
f31f09f6 | 3227 | ret = BTRFS_LOG_FORCE_COMMIT; |
4a500fd1 YZ |
3228 | goto out; |
3229 | } | |
3230 | ||
d1433deb | 3231 | if (log_root_tree->log_transid_committed >= root_log_ctx.log_transid) { |
3da5ab56 | 3232 | blk_finish_plug(&plug); |
cbd60aa7 | 3233 | list_del_init(&root_log_ctx.list); |
d1433deb MX |
3234 | mutex_unlock(&log_root_tree->log_mutex); |
3235 | ret = root_log_ctx.log_ret; | |
3236 | goto out; | |
3237 | } | |
8b050d35 | 3238 | |
d1433deb | 3239 | index2 = root_log_ctx.log_transid % 2; |
7237f183 | 3240 | if (atomic_read(&log_root_tree->log_commit[index2])) { |
c6adc9cc | 3241 | blk_finish_plug(&plug); |
bf89d38f | 3242 | ret = btrfs_wait_tree_log_extents(log, mark); |
60d53eb3 | 3243 | wait_log_commit(log_root_tree, |
d1433deb | 3244 | root_log_ctx.log_transid); |
7237f183 | 3245 | mutex_unlock(&log_root_tree->log_mutex); |
5ab5e44a FM |
3246 | if (!ret) |
3247 | ret = root_log_ctx.log_ret; | |
7237f183 YZ |
3248 | goto out; |
3249 | } | |
d1433deb | 3250 | ASSERT(root_log_ctx.log_transid == log_root_tree->log_transid); |
7237f183 YZ |
3251 | atomic_set(&log_root_tree->log_commit[index2], 1); |
3252 | ||
12fcfd22 | 3253 | if (atomic_read(&log_root_tree->log_commit[(index2 + 1) % 2])) { |
60d53eb3 | 3254 | wait_log_commit(log_root_tree, |
d1433deb | 3255 | root_log_ctx.log_transid - 1); |
12fcfd22 CM |
3256 | } |
3257 | ||
12fcfd22 CM |
3258 | /* |
3259 | * now that we've moved on to the tree of log tree roots, | |
3260 | * check the full commit flag again | |
3261 | */ | |
4884b8e8 | 3262 | if (btrfs_need_log_full_commit(trans)) { |
c6adc9cc | 3263 | blk_finish_plug(&plug); |
bf89d38f | 3264 | btrfs_wait_tree_log_extents(log, mark); |
12fcfd22 | 3265 | mutex_unlock(&log_root_tree->log_mutex); |
f31f09f6 | 3266 | ret = BTRFS_LOG_FORCE_COMMIT; |
12fcfd22 CM |
3267 | goto out_wake_log_root; |
3268 | } | |
7237f183 | 3269 | |
2ff7e61e | 3270 | ret = btrfs_write_marked_extents(fs_info, |
c6adc9cc MX |
3271 | &log_root_tree->dirty_log_pages, |
3272 | EXTENT_DIRTY | EXTENT_NEW); | |
3273 | blk_finish_plug(&plug); | |
b528f467 NA |
3274 | /* |
3275 | * As described above, -EAGAIN indicates a hole in the extents. We | |
3276 | * cannot wait for these write outs since the waiting cause a | |
3277 | * deadlock. Bail out to the full commit instead. | |
3278 | */ | |
3279 | if (ret == -EAGAIN && btrfs_is_zoned(fs_info)) { | |
3280 | btrfs_set_log_full_commit(trans); | |
3281 | btrfs_wait_tree_log_extents(log, mark); | |
3282 | mutex_unlock(&log_root_tree->log_mutex); | |
3283 | goto out_wake_log_root; | |
3284 | } else if (ret) { | |
90787766 | 3285 | btrfs_set_log_full_commit(trans); |
66642832 | 3286 | btrfs_abort_transaction(trans, ret); |
79787eaa JM |
3287 | mutex_unlock(&log_root_tree->log_mutex); |
3288 | goto out_wake_log_root; | |
3289 | } | |
bf89d38f | 3290 | ret = btrfs_wait_tree_log_extents(log, mark); |
5ab5e44a | 3291 | if (!ret) |
bf89d38f JM |
3292 | ret = btrfs_wait_tree_log_extents(log_root_tree, |
3293 | EXTENT_NEW | EXTENT_DIRTY); | |
5ab5e44a | 3294 | if (ret) { |
90787766 | 3295 | btrfs_set_log_full_commit(trans); |
5ab5e44a FM |
3296 | mutex_unlock(&log_root_tree->log_mutex); |
3297 | goto out_wake_log_root; | |
3298 | } | |
e02119d5 | 3299 | |
47876f7c FM |
3300 | log_root_start = log_root_tree->node->start; |
3301 | log_root_level = btrfs_header_level(log_root_tree->node); | |
7237f183 | 3302 | log_root_tree->log_transid++; |
7237f183 YZ |
3303 | mutex_unlock(&log_root_tree->log_mutex); |
3304 | ||
3305 | /* | |
47876f7c FM |
3306 | * Here we are guaranteed that nobody is going to write the superblock |
3307 | * for the current transaction before us and that neither we do write | |
3308 | * our superblock before the previous transaction finishes its commit | |
3309 | * and writes its superblock, because: | |
3310 | * | |
3311 | * 1) We are holding a handle on the current transaction, so no body | |
3312 | * can commit it until we release the handle; | |
3313 | * | |
3314 | * 2) Before writing our superblock we acquire the tree_log_mutex, so | |
3315 | * if the previous transaction is still committing, and hasn't yet | |
3316 | * written its superblock, we wait for it to do it, because a | |
3317 | * transaction commit acquires the tree_log_mutex when the commit | |
3318 | * begins and releases it only after writing its superblock. | |
7237f183 | 3319 | */ |
47876f7c | 3320 | mutex_lock(&fs_info->tree_log_mutex); |
165ea85f JB |
3321 | |
3322 | /* | |
3323 | * The previous transaction writeout phase could have failed, and thus | |
3324 | * marked the fs in an error state. We must not commit here, as we | |
3325 | * could have updated our generation in the super_for_commit and | |
3326 | * writing the super here would result in transid mismatches. If there | |
3327 | * is an error here just bail. | |
3328 | */ | |
84961539 | 3329 | if (BTRFS_FS_ERROR(fs_info)) { |
165ea85f JB |
3330 | ret = -EIO; |
3331 | btrfs_set_log_full_commit(trans); | |
3332 | btrfs_abort_transaction(trans, ret); | |
3333 | mutex_unlock(&fs_info->tree_log_mutex); | |
3334 | goto out_wake_log_root; | |
3335 | } | |
3336 | ||
47876f7c FM |
3337 | btrfs_set_super_log_root(fs_info->super_for_commit, log_root_start); |
3338 | btrfs_set_super_log_root_level(fs_info->super_for_commit, log_root_level); | |
eece6a9c | 3339 | ret = write_all_supers(fs_info, 1); |
47876f7c | 3340 | mutex_unlock(&fs_info->tree_log_mutex); |
5af3e8cc | 3341 | if (ret) { |
90787766 | 3342 | btrfs_set_log_full_commit(trans); |
66642832 | 3343 | btrfs_abort_transaction(trans, ret); |
5af3e8cc SB |
3344 | goto out_wake_log_root; |
3345 | } | |
7237f183 | 3346 | |
e1a6d264 FM |
3347 | /* |
3348 | * We know there can only be one task here, since we have not yet set | |
3349 | * root->log_commit[index1] to 0 and any task attempting to sync the | |
3350 | * log must wait for the previous log transaction to commit if it's | |
3351 | * still in progress or wait for the current log transaction commit if | |
3352 | * someone else already started it. We use <= and not < because the | |
3353 | * first log transaction has an ID of 0. | |
3354 | */ | |
3355 | ASSERT(root->last_log_commit <= log_transid); | |
3356 | root->last_log_commit = log_transid; | |
257c62e1 | 3357 | |
12fcfd22 | 3358 | out_wake_log_root: |
570dd450 | 3359 | mutex_lock(&log_root_tree->log_mutex); |
8b050d35 MX |
3360 | btrfs_remove_all_log_ctxs(log_root_tree, index2, ret); |
3361 | ||
d1433deb | 3362 | log_root_tree->log_transid_committed++; |
7237f183 | 3363 | atomic_set(&log_root_tree->log_commit[index2], 0); |
d1433deb MX |
3364 | mutex_unlock(&log_root_tree->log_mutex); |
3365 | ||
33a9eca7 | 3366 | /* |
093258e6 DS |
3367 | * The barrier before waitqueue_active (in cond_wake_up) is needed so |
3368 | * all the updates above are seen by the woken threads. It might not be | |
3369 | * necessary, but proving that seems to be hard. | |
33a9eca7 | 3370 | */ |
093258e6 | 3371 | cond_wake_up(&log_root_tree->log_commit_wait[index2]); |
e02119d5 | 3372 | out: |
d1433deb | 3373 | mutex_lock(&root->log_mutex); |
570dd450 | 3374 | btrfs_remove_all_log_ctxs(root, index1, ret); |
d1433deb | 3375 | root->log_transid_committed++; |
7237f183 | 3376 | atomic_set(&root->log_commit[index1], 0); |
d1433deb | 3377 | mutex_unlock(&root->log_mutex); |
8b050d35 | 3378 | |
33a9eca7 | 3379 | /* |
093258e6 DS |
3380 | * The barrier before waitqueue_active (in cond_wake_up) is needed so |
3381 | * all the updates above are seen by the woken threads. It might not be | |
3382 | * necessary, but proving that seems to be hard. | |
33a9eca7 | 3383 | */ |
093258e6 | 3384 | cond_wake_up(&root->log_commit_wait[index1]); |
b31eabd8 | 3385 | return ret; |
e02119d5 CM |
3386 | } |
3387 | ||
4a500fd1 YZ |
3388 | static void free_log_tree(struct btrfs_trans_handle *trans, |
3389 | struct btrfs_root *log) | |
e02119d5 CM |
3390 | { |
3391 | int ret; | |
e02119d5 CM |
3392 | struct walk_control wc = { |
3393 | .free = 1, | |
3394 | .process_func = process_one_buffer | |
3395 | }; | |
3396 | ||
3ddebf27 NA |
3397 | if (log->node) { |
3398 | ret = walk_log_tree(trans, log, &wc); | |
3399 | if (ret) { | |
40cdc509 FM |
3400 | /* |
3401 | * We weren't able to traverse the entire log tree, the | |
3402 | * typical scenario is getting an -EIO when reading an | |
3403 | * extent buffer of the tree, due to a previous writeback | |
3404 | * failure of it. | |
3405 | */ | |
3406 | set_bit(BTRFS_FS_STATE_LOG_CLEANUP_ERROR, | |
3407 | &log->fs_info->fs_state); | |
3408 | ||
3409 | /* | |
3410 | * Some extent buffers of the log tree may still be dirty | |
3411 | * and not yet written back to storage, because we may | |
3412 | * have updates to a log tree without syncing a log tree, | |
3413 | * such as during rename and link operations. So flush | |
3414 | * them out and wait for their writeback to complete, so | |
3415 | * that we properly cleanup their state and pages. | |
3416 | */ | |
3417 | btrfs_write_marked_extents(log->fs_info, | |
3418 | &log->dirty_log_pages, | |
3419 | EXTENT_DIRTY | EXTENT_NEW); | |
3420 | btrfs_wait_tree_log_extents(log, | |
3421 | EXTENT_DIRTY | EXTENT_NEW); | |
3422 | ||
3ddebf27 NA |
3423 | if (trans) |
3424 | btrfs_abort_transaction(trans, ret); | |
3425 | else | |
3426 | btrfs_handle_fs_error(log->fs_info, ret, NULL); | |
3427 | } | |
374b0e2d | 3428 | } |
e02119d5 | 3429 | |
59b0713a FM |
3430 | clear_extent_bits(&log->dirty_log_pages, 0, (u64)-1, |
3431 | EXTENT_DIRTY | EXTENT_NEW | EXTENT_NEED_WAIT); | |
e289f03e | 3432 | extent_io_tree_release(&log->log_csum_range); |
d3575156 | 3433 | |
00246528 | 3434 | btrfs_put_root(log); |
4a500fd1 YZ |
3435 | } |
3436 | ||
3437 | /* | |
3438 | * free all the extents used by the tree log. This should be called | |
3439 | * at commit time of the full transaction | |
3440 | */ | |
3441 | int btrfs_free_log(struct btrfs_trans_handle *trans, struct btrfs_root *root) | |
3442 | { | |
3443 | if (root->log_root) { | |
3444 | free_log_tree(trans, root->log_root); | |
3445 | root->log_root = NULL; | |
e7a79811 | 3446 | clear_bit(BTRFS_ROOT_HAS_LOG_TREE, &root->state); |
4a500fd1 YZ |
3447 | } |
3448 | return 0; | |
3449 | } | |
3450 | ||
3451 | int btrfs_free_log_root_tree(struct btrfs_trans_handle *trans, | |
3452 | struct btrfs_fs_info *fs_info) | |
3453 | { | |
3454 | if (fs_info->log_root_tree) { | |
3455 | free_log_tree(trans, fs_info->log_root_tree); | |
3456 | fs_info->log_root_tree = NULL; | |
47876f7c | 3457 | clear_bit(BTRFS_ROOT_HAS_LOG_TREE, &fs_info->tree_root->state); |
4a500fd1 | 3458 | } |
e02119d5 CM |
3459 | return 0; |
3460 | } | |
3461 | ||
803f0f64 | 3462 | /* |
0f8ce498 FM |
3463 | * Check if an inode was logged in the current transaction. This correctly deals |
3464 | * with the case where the inode was logged but has a logged_trans of 0, which | |
3465 | * happens if the inode is evicted and loaded again, as logged_trans is an in | |
3466 | * memory only field (not persisted). | |
3467 | * | |
3468 | * Returns 1 if the inode was logged before in the transaction, 0 if it was not, | |
3469 | * and < 0 on error. | |
803f0f64 | 3470 | */ |
0f8ce498 FM |
3471 | static int inode_logged(struct btrfs_trans_handle *trans, |
3472 | struct btrfs_inode *inode, | |
3473 | struct btrfs_path *path_in) | |
803f0f64 | 3474 | { |
0f8ce498 FM |
3475 | struct btrfs_path *path = path_in; |
3476 | struct btrfs_key key; | |
3477 | int ret; | |
3478 | ||
803f0f64 | 3479 | if (inode->logged_trans == trans->transid) |
0f8ce498 | 3480 | return 1; |
803f0f64 | 3481 | |
0f8ce498 FM |
3482 | /* |
3483 | * If logged_trans is not 0, then we know the inode logged was not logged | |
3484 | * in this transaction, so we can return false right away. | |
3485 | */ | |
3486 | if (inode->logged_trans > 0) | |
3487 | return 0; | |
3488 | ||
3489 | /* | |
3490 | * If no log tree was created for this root in this transaction, then | |
3491 | * the inode can not have been logged in this transaction. In that case | |
3492 | * set logged_trans to anything greater than 0 and less than the current | |
3493 | * transaction's ID, to avoid the search below in a future call in case | |
3494 | * a log tree gets created after this. | |
3495 | */ | |
3496 | if (!test_bit(BTRFS_ROOT_HAS_LOG_TREE, &inode->root->state)) { | |
3497 | inode->logged_trans = trans->transid - 1; | |
3498 | return 0; | |
3499 | } | |
3500 | ||
3501 | /* | |
3502 | * We have a log tree and the inode's logged_trans is 0. We can't tell | |
3503 | * for sure if the inode was logged before in this transaction by looking | |
3504 | * only at logged_trans. We could be pessimistic and assume it was, but | |
3505 | * that can lead to unnecessarily logging an inode during rename and link | |
3506 | * operations, and then further updating the log in followup rename and | |
3507 | * link operations, specially if it's a directory, which adds latency | |
3508 | * visible to applications doing a series of rename or link operations. | |
3509 | * | |
3510 | * A logged_trans of 0 here can mean several things: | |
3511 | * | |
3512 | * 1) The inode was never logged since the filesystem was mounted, and may | |
3513 | * or may have not been evicted and loaded again; | |
3514 | * | |
3515 | * 2) The inode was logged in a previous transaction, then evicted and | |
3516 | * then loaded again; | |
3517 | * | |
3518 | * 3) The inode was logged in the current transaction, then evicted and | |
3519 | * then loaded again. | |
3520 | * | |
3521 | * For cases 1) and 2) we don't want to return true, but we need to detect | |
3522 | * case 3) and return true. So we do a search in the log root for the inode | |
3523 | * item. | |
3524 | */ | |
3525 | key.objectid = btrfs_ino(inode); | |
3526 | key.type = BTRFS_INODE_ITEM_KEY; | |
3527 | key.offset = 0; | |
3528 | ||
3529 | if (!path) { | |
3530 | path = btrfs_alloc_path(); | |
3531 | if (!path) | |
3532 | return -ENOMEM; | |
3533 | } | |
3534 | ||
3535 | ret = btrfs_search_slot(NULL, inode->root->log_root, &key, path, 0, 0); | |
3536 | ||
3537 | if (path_in) | |
3538 | btrfs_release_path(path); | |
3539 | else | |
3540 | btrfs_free_path(path); | |
1e0860f3 | 3541 | |
6e8e777d | 3542 | /* |
0f8ce498 FM |
3543 | * Logging an inode always results in logging its inode item. So if we |
3544 | * did not find the item we know the inode was not logged for sure. | |
6e8e777d | 3545 | */ |
0f8ce498 FM |
3546 | if (ret < 0) { |
3547 | return ret; | |
3548 | } else if (ret > 0) { | |
3549 | /* | |
3550 | * Set logged_trans to a value greater than 0 and less then the | |
3551 | * current transaction to avoid doing the search in future calls. | |
3552 | */ | |
3553 | inode->logged_trans = trans->transid - 1; | |
3554 | return 0; | |
3555 | } | |
3556 | ||
3557 | /* | |
3558 | * The inode was previously logged and then evicted, set logged_trans to | |
3559 | * the current transacion's ID, to avoid future tree searches as long as | |
3560 | * the inode is not evicted again. | |
3561 | */ | |
3562 | inode->logged_trans = trans->transid; | |
3563 | ||
3564 | /* | |
3565 | * If it's a directory, then we must set last_dir_index_offset to the | |
3566 | * maximum possible value, so that the next attempt to log the inode does | |
3567 | * not skip checking if dir index keys found in modified subvolume tree | |
3568 | * leaves have been logged before, otherwise it would result in attempts | |
3569 | * to insert duplicate dir index keys in the log tree. This must be done | |
3570 | * because last_dir_index_offset is an in-memory only field, not persisted | |
3571 | * in the inode item or any other on-disk structure, so its value is lost | |
3572 | * once the inode is evicted. | |
3573 | */ | |
3574 | if (S_ISDIR(inode->vfs_inode.i_mode)) | |
3575 | inode->last_dir_index_offset = (u64)-1; | |
803f0f64 | 3576 | |
0f8ce498 | 3577 | return 1; |
803f0f64 FM |
3578 | } |
3579 | ||
839061fe FM |
3580 | /* |
3581 | * Delete a directory entry from the log if it exists. | |
3582 | * | |
3583 | * Returns < 0 on error | |
3584 | * 1 if the entry does not exists | |
3585 | * 0 if the entry existed and was successfully deleted | |
3586 | */ | |
3587 | static int del_logged_dentry(struct btrfs_trans_handle *trans, | |
3588 | struct btrfs_root *log, | |
3589 | struct btrfs_path *path, | |
3590 | u64 dir_ino, | |
3591 | const char *name, int name_len, | |
3592 | u64 index) | |
3593 | { | |
3594 | struct btrfs_dir_item *di; | |
3595 | ||
3596 | /* | |
3597 | * We only log dir index items of a directory, so we don't need to look | |
3598 | * for dir item keys. | |
3599 | */ | |
3600 | di = btrfs_lookup_dir_index_item(trans, log, path, dir_ino, | |
3601 | index, name, name_len, -1); | |
3602 | if (IS_ERR(di)) | |
3603 | return PTR_ERR(di); | |
3604 | else if (!di) | |
3605 | return 1; | |
3606 | ||
3607 | /* | |
3608 | * We do not need to update the size field of the directory's | |
3609 | * inode item because on log replay we update the field to reflect | |
3610 | * all existing entries in the directory (see overwrite_item()). | |
3611 | */ | |
3612 | return btrfs_delete_one_dir_name(trans, log, path, di); | |
3613 | } | |
3614 | ||
e02119d5 CM |
3615 | /* |
3616 | * If both a file and directory are logged, and unlinks or renames are | |
3617 | * mixed in, we have a few interesting corners: | |
3618 | * | |
3619 | * create file X in dir Y | |
3620 | * link file X to X.link in dir Y | |
3621 | * fsync file X | |
3622 | * unlink file X but leave X.link | |
3623 | * fsync dir Y | |
3624 | * | |
3625 | * After a crash we would expect only X.link to exist. But file X | |
3626 | * didn't get fsync'd again so the log has back refs for X and X.link. | |
3627 | * | |
3628 | * We solve this by removing directory entries and inode backrefs from the | |
3629 | * log when a file that was logged in the current transaction is | |
3630 | * unlinked. Any later fsync will include the updated log entries, and | |
3631 | * we'll be able to reconstruct the proper directory items from backrefs. | |
3632 | * | |
3633 | * This optimizations allows us to avoid relogging the entire inode | |
3634 | * or the entire directory. | |
3635 | */ | |
9a35fc95 JB |
3636 | void btrfs_del_dir_entries_in_log(struct btrfs_trans_handle *trans, |
3637 | struct btrfs_root *root, | |
3638 | const char *name, int name_len, | |
3639 | struct btrfs_inode *dir, u64 index) | |
e02119d5 | 3640 | { |
e02119d5 CM |
3641 | struct btrfs_path *path; |
3642 | int ret; | |
e02119d5 | 3643 | |
0f8ce498 FM |
3644 | ret = inode_logged(trans, dir, NULL); |
3645 | if (ret == 0) | |
3646 | return; | |
3647 | else if (ret < 0) { | |
3648 | btrfs_set_log_full_commit(trans); | |
9a35fc95 | 3649 | return; |
0f8ce498 | 3650 | } |
3a5f1d45 | 3651 | |
e02119d5 CM |
3652 | ret = join_running_log_trans(root); |
3653 | if (ret) | |
9a35fc95 | 3654 | return; |
e02119d5 | 3655 | |
49f34d1f | 3656 | mutex_lock(&dir->log_mutex); |
e02119d5 | 3657 | |
e02119d5 | 3658 | path = btrfs_alloc_path(); |
a62f44a5 | 3659 | if (!path) { |
839061fe | 3660 | ret = -ENOMEM; |
a62f44a5 TI |
3661 | goto out_unlock; |
3662 | } | |
2a29edc6 | 3663 | |
839061fe FM |
3664 | ret = del_logged_dentry(trans, root->log_root, path, btrfs_ino(dir), |
3665 | name, name_len, index); | |
e02119d5 | 3666 | btrfs_free_path(path); |
a62f44a5 | 3667 | out_unlock: |
49f34d1f | 3668 | mutex_unlock(&dir->log_mutex); |
839061fe | 3669 | if (ret < 0) |
90787766 | 3670 | btrfs_set_log_full_commit(trans); |
12fcfd22 | 3671 | btrfs_end_log_trans(root); |
e02119d5 CM |
3672 | } |
3673 | ||
3674 | /* see comments for btrfs_del_dir_entries_in_log */ | |
9a35fc95 JB |
3675 | void btrfs_del_inode_ref_in_log(struct btrfs_trans_handle *trans, |
3676 | struct btrfs_root *root, | |
3677 | const char *name, int name_len, | |
3678 | struct btrfs_inode *inode, u64 dirid) | |
e02119d5 CM |
3679 | { |
3680 | struct btrfs_root *log; | |
3681 | u64 index; | |
3682 | int ret; | |
3683 | ||
0f8ce498 FM |
3684 | ret = inode_logged(trans, inode, NULL); |
3685 | if (ret == 0) | |
9a35fc95 | 3686 | return; |
0f8ce498 FM |
3687 | else if (ret < 0) { |
3688 | btrfs_set_log_full_commit(trans); | |
3689 | return; | |
3690 | } | |
3a5f1d45 | 3691 | |
e02119d5 CM |
3692 | ret = join_running_log_trans(root); |
3693 | if (ret) | |
9a35fc95 | 3694 | return; |
e02119d5 | 3695 | log = root->log_root; |
a491abb2 | 3696 | mutex_lock(&inode->log_mutex); |
e02119d5 | 3697 | |
a491abb2 | 3698 | ret = btrfs_del_inode_ref(trans, log, name, name_len, btrfs_ino(inode), |
e02119d5 | 3699 | dirid, &index); |
a491abb2 | 3700 | mutex_unlock(&inode->log_mutex); |
9a35fc95 | 3701 | if (ret < 0 && ret != -ENOENT) |
90787766 | 3702 | btrfs_set_log_full_commit(trans); |
12fcfd22 | 3703 | btrfs_end_log_trans(root); |
e02119d5 CM |
3704 | } |
3705 | ||
3706 | /* | |
3707 | * creates a range item in the log for 'dirid'. first_offset and | |
3708 | * last_offset tell us which parts of the key space the log should | |
3709 | * be considered authoritative for. | |
3710 | */ | |
3711 | static noinline int insert_dir_log_key(struct btrfs_trans_handle *trans, | |
3712 | struct btrfs_root *log, | |
3713 | struct btrfs_path *path, | |
339d0354 | 3714 | u64 dirid, |
e02119d5 CM |
3715 | u64 first_offset, u64 last_offset) |
3716 | { | |
3717 | int ret; | |
3718 | struct btrfs_key key; | |
3719 | struct btrfs_dir_log_item *item; | |
3720 | ||
3721 | key.objectid = dirid; | |
3722 | key.offset = first_offset; | |
339d0354 | 3723 | key.type = BTRFS_DIR_LOG_INDEX_KEY; |
e02119d5 | 3724 | ret = btrfs_insert_empty_item(trans, log, path, &key, sizeof(*item)); |
750ee454 FM |
3725 | /* |
3726 | * -EEXIST is fine and can happen sporadically when we are logging a | |
3727 | * directory and have concurrent insertions in the subvolume's tree for | |
3728 | * items from other inodes and that result in pushing off some dir items | |
3729 | * from one leaf to another in order to accommodate for the new items. | |
3730 | * This results in logging the same dir index range key. | |
3731 | */ | |
3732 | if (ret && ret != -EEXIST) | |
4a500fd1 | 3733 | return ret; |
e02119d5 CM |
3734 | |
3735 | item = btrfs_item_ptr(path->nodes[0], path->slots[0], | |
3736 | struct btrfs_dir_log_item); | |
750ee454 FM |
3737 | if (ret == -EEXIST) { |
3738 | const u64 curr_end = btrfs_dir_log_end(path->nodes[0], item); | |
3739 | ||
3740 | /* | |
3741 | * btrfs_del_dir_entries_in_log() might have been called during | |
3742 | * an unlink between the initial insertion of this key and the | |
3743 | * current update, or we might be logging a single entry deletion | |
3744 | * during a rename, so set the new last_offset to the max value. | |
3745 | */ | |
3746 | last_offset = max(last_offset, curr_end); | |
3747 | } | |
e02119d5 CM |
3748 | btrfs_set_dir_log_end(path->nodes[0], item, last_offset); |
3749 | btrfs_mark_buffer_dirty(path->nodes[0]); | |
b3b4aa74 | 3750 | btrfs_release_path(path); |
e02119d5 CM |
3751 | return 0; |
3752 | } | |
3753 | ||
086dcbfa FM |
3754 | static int flush_dir_items_batch(struct btrfs_trans_handle *trans, |
3755 | struct btrfs_root *log, | |
3756 | struct extent_buffer *src, | |
3757 | struct btrfs_path *dst_path, | |
3758 | int start_slot, | |
3759 | int count) | |
3760 | { | |
3761 | char *ins_data = NULL; | |
b7ef5f3a | 3762 | struct btrfs_item_batch batch; |
086dcbfa | 3763 | struct extent_buffer *dst; |
da1b811f FM |
3764 | unsigned long src_offset; |
3765 | unsigned long dst_offset; | |
086dcbfa FM |
3766 | struct btrfs_key key; |
3767 | u32 item_size; | |
3768 | int ret; | |
3769 | int i; | |
3770 | ||
3771 | ASSERT(count > 0); | |
b7ef5f3a | 3772 | batch.nr = count; |
086dcbfa FM |
3773 | |
3774 | if (count == 1) { | |
3775 | btrfs_item_key_to_cpu(src, &key, start_slot); | |
3212fa14 | 3776 | item_size = btrfs_item_size(src, start_slot); |
b7ef5f3a FM |
3777 | batch.keys = &key; |
3778 | batch.data_sizes = &item_size; | |
3779 | batch.total_data_size = item_size; | |
086dcbfa | 3780 | } else { |
b7ef5f3a FM |
3781 | struct btrfs_key *ins_keys; |
3782 | u32 *ins_sizes; | |
3783 | ||
086dcbfa FM |
3784 | ins_data = kmalloc(count * sizeof(u32) + |
3785 | count * sizeof(struct btrfs_key), GFP_NOFS); | |
3786 | if (!ins_data) | |
3787 | return -ENOMEM; | |
3788 | ||
3789 | ins_sizes = (u32 *)ins_data; | |
3790 | ins_keys = (struct btrfs_key *)(ins_data + count * sizeof(u32)); | |
b7ef5f3a FM |
3791 | batch.keys = ins_keys; |
3792 | batch.data_sizes = ins_sizes; | |
3793 | batch.total_data_size = 0; | |
086dcbfa FM |
3794 | |
3795 | for (i = 0; i < count; i++) { | |
3796 | const int slot = start_slot + i; | |
3797 | ||
3798 | btrfs_item_key_to_cpu(src, &ins_keys[i], slot); | |
3212fa14 | 3799 | ins_sizes[i] = btrfs_item_size(src, slot); |
b7ef5f3a | 3800 | batch.total_data_size += ins_sizes[i]; |
086dcbfa FM |
3801 | } |
3802 | } | |
3803 | ||
b7ef5f3a | 3804 | ret = btrfs_insert_empty_items(trans, log, dst_path, &batch); |
086dcbfa FM |
3805 | if (ret) |
3806 | goto out; | |
3807 | ||
3808 | dst = dst_path->nodes[0]; | |
da1b811f FM |
3809 | /* |
3810 | * Copy all the items in bulk, in a single copy operation. Item data is | |
3811 | * organized such that it's placed at the end of a leaf and from right | |
3812 | * to left. For example, the data for the second item ends at an offset | |
3813 | * that matches the offset where the data for the first item starts, the | |
3814 | * data for the third item ends at an offset that matches the offset | |
3815 | * where the data of the second items starts, and so on. | |
3816 | * Therefore our source and destination start offsets for copy match the | |
3817 | * offsets of the last items (highest slots). | |
3818 | */ | |
3819 | dst_offset = btrfs_item_ptr_offset(dst, dst_path->slots[0] + count - 1); | |
3820 | src_offset = btrfs_item_ptr_offset(src, start_slot + count - 1); | |
3821 | copy_extent_buffer(dst, src, dst_offset, src_offset, batch.total_data_size); | |
086dcbfa FM |
3822 | btrfs_release_path(dst_path); |
3823 | out: | |
3824 | kfree(ins_data); | |
3825 | ||
3826 | return ret; | |
3827 | } | |
3828 | ||
eb10d85e FM |
3829 | static int process_dir_items_leaf(struct btrfs_trans_handle *trans, |
3830 | struct btrfs_inode *inode, | |
3831 | struct btrfs_path *path, | |
3832 | struct btrfs_path *dst_path, | |
732d591a FM |
3833 | struct btrfs_log_ctx *ctx, |
3834 | u64 *last_old_dentry_offset) | |
eb10d85e FM |
3835 | { |
3836 | struct btrfs_root *log = inode->root->log_root; | |
3837 | struct extent_buffer *src = path->nodes[0]; | |
3838 | const int nritems = btrfs_header_nritems(src); | |
3839 | const u64 ino = btrfs_ino(inode); | |
086dcbfa FM |
3840 | bool last_found = false; |
3841 | int batch_start = 0; | |
3842 | int batch_size = 0; | |
eb10d85e FM |
3843 | int i; |
3844 | ||
3845 | for (i = path->slots[0]; i < nritems; i++) { | |
732d591a | 3846 | struct btrfs_dir_item *di; |
eb10d85e | 3847 | struct btrfs_key key; |
eb10d85e FM |
3848 | int ret; |
3849 | ||
3850 | btrfs_item_key_to_cpu(src, &key, i); | |
3851 | ||
339d0354 | 3852 | if (key.objectid != ino || key.type != BTRFS_DIR_INDEX_KEY) { |
086dcbfa FM |
3853 | last_found = true; |
3854 | break; | |
3855 | } | |
eb10d85e | 3856 | |
732d591a | 3857 | di = btrfs_item_ptr(src, i, struct btrfs_dir_item); |
dc287224 | 3858 | ctx->last_dir_item_offset = key.offset; |
732d591a FM |
3859 | |
3860 | /* | |
3861 | * Skip ranges of items that consist only of dir item keys created | |
3862 | * in past transactions. However if we find a gap, we must log a | |
3863 | * dir index range item for that gap, so that index keys in that | |
3864 | * gap are deleted during log replay. | |
3865 | */ | |
3866 | if (btrfs_dir_transid(src, di) < trans->transid) { | |
3867 | if (key.offset > *last_old_dentry_offset + 1) { | |
3868 | ret = insert_dir_log_key(trans, log, dst_path, | |
3869 | ino, *last_old_dentry_offset + 1, | |
3870 | key.offset - 1); | |
732d591a FM |
3871 | if (ret < 0) |
3872 | return ret; | |
3873 | } | |
3874 | ||
3875 | *last_old_dentry_offset = key.offset; | |
3876 | continue; | |
3877 | } | |
eb10d85e FM |
3878 | /* |
3879 | * We must make sure that when we log a directory entry, the | |
3880 | * corresponding inode, after log replay, has a matching link | |
3881 | * count. For example: | |
3882 | * | |
3883 | * touch foo | |
3884 | * mkdir mydir | |
3885 | * sync | |
3886 | * ln foo mydir/bar | |
3887 | * xfs_io -c "fsync" mydir | |
3888 | * <crash> | |
3889 | * <mount fs and log replay> | |
3890 | * | |
3891 | * Would result in a fsync log that when replayed, our file inode | |
3892 | * would have a link count of 1, but we get two directory entries | |
3893 | * pointing to the same inode. After removing one of the names, | |
3894 | * it would not be possible to remove the other name, which | |
3895 | * resulted always in stale file handle errors, and would not be | |
3896 | * possible to rmdir the parent directory, since its i_size could | |
3897 | * never be decremented to the value BTRFS_EMPTY_DIR_SIZE, | |
3898 | * resulting in -ENOTEMPTY errors. | |
3899 | */ | |
086dcbfa | 3900 | if (!ctx->log_new_dentries) { |
086dcbfa FM |
3901 | struct btrfs_key di_key; |
3902 | ||
086dcbfa | 3903 | btrfs_dir_item_key_to_cpu(src, di, &di_key); |
732d591a | 3904 | if (di_key.type != BTRFS_ROOT_ITEM_KEY) |
086dcbfa FM |
3905 | ctx->log_new_dentries = true; |
3906 | } | |
3907 | ||
0f8ce498 | 3908 | if (!ctx->logged_before) |
086dcbfa | 3909 | goto add_to_batch; |
dc287224 FM |
3910 | |
3911 | /* | |
3912 | * If we were logged before and have logged dir items, we can skip | |
3913 | * checking if any item with a key offset larger than the last one | |
3914 | * we logged is in the log tree, saving time and avoiding adding | |
0f8ce498 FM |
3915 | * contention on the log tree. We can only rely on the value of |
3916 | * last_dir_index_offset when we know for sure that the inode was | |
3917 | * previously logged in the current transaction. | |
dc287224 | 3918 | */ |
339d0354 | 3919 | if (key.offset > inode->last_dir_index_offset) |
dc287224 | 3920 | goto add_to_batch; |
086dcbfa FM |
3921 | /* |
3922 | * Check if the key was already logged before. If not we can add | |
3923 | * it to a batch for bulk insertion. | |
3924 | */ | |
3925 | ret = btrfs_search_slot(NULL, log, &key, dst_path, 0, 0); | |
3926 | if (ret < 0) { | |
3927 | return ret; | |
3928 | } else if (ret > 0) { | |
3929 | btrfs_release_path(dst_path); | |
3930 | goto add_to_batch; | |
3931 | } | |
3932 | ||
3933 | /* | |
3934 | * Item exists in the log. Overwrite the item in the log if it | |
3935 | * has different content or do nothing if it has exactly the same | |
3936 | * content. And then flush the current batch if any - do it after | |
3937 | * overwriting the current item, or we would deadlock otherwise, | |
3938 | * since we are holding a path for the existing item. | |
3939 | */ | |
3940 | ret = do_overwrite_item(trans, log, dst_path, src, i, &key); | |
3941 | if (ret < 0) | |
3942 | return ret; | |
3943 | ||
3944 | if (batch_size > 0) { | |
3945 | ret = flush_dir_items_batch(trans, log, src, dst_path, | |
3946 | batch_start, batch_size); | |
3947 | if (ret < 0) | |
3948 | return ret; | |
3949 | batch_size = 0; | |
3950 | } | |
3951 | continue; | |
3952 | add_to_batch: | |
3953 | if (batch_size == 0) | |
3954 | batch_start = i; | |
3955 | batch_size++; | |
eb10d85e FM |
3956 | } |
3957 | ||
086dcbfa FM |
3958 | if (batch_size > 0) { |
3959 | int ret; | |
3960 | ||
3961 | ret = flush_dir_items_batch(trans, log, src, dst_path, | |
3962 | batch_start, batch_size); | |
3963 | if (ret < 0) | |
3964 | return ret; | |
3965 | } | |
3966 | ||
3967 | return last_found ? 1 : 0; | |
eb10d85e FM |
3968 | } |
3969 | ||
e02119d5 CM |
3970 | /* |
3971 | * log all the items included in the current transaction for a given | |
3972 | * directory. This also creates the range items in the log tree required | |
3973 | * to replay anything deleted before the fsync | |
3974 | */ | |
3975 | static noinline int log_dir_items(struct btrfs_trans_handle *trans, | |
90d04510 | 3976 | struct btrfs_inode *inode, |
e02119d5 | 3977 | struct btrfs_path *path, |
339d0354 | 3978 | struct btrfs_path *dst_path, |
2f2ff0ee | 3979 | struct btrfs_log_ctx *ctx, |
e02119d5 CM |
3980 | u64 min_offset, u64 *last_offset_ret) |
3981 | { | |
3982 | struct btrfs_key min_key; | |
90d04510 | 3983 | struct btrfs_root *root = inode->root; |
e02119d5 | 3984 | struct btrfs_root *log = root->log_root; |
4a500fd1 | 3985 | int err = 0; |
e02119d5 | 3986 | int ret; |
732d591a | 3987 | u64 last_old_dentry_offset = min_offset - 1; |
e02119d5 | 3988 | u64 last_offset = (u64)-1; |
684a5773 | 3989 | u64 ino = btrfs_ino(inode); |
e02119d5 | 3990 | |
33345d01 | 3991 | min_key.objectid = ino; |
339d0354 | 3992 | min_key.type = BTRFS_DIR_INDEX_KEY; |
e02119d5 CM |
3993 | min_key.offset = min_offset; |
3994 | ||
6174d3cb | 3995 | ret = btrfs_search_forward(root, &min_key, path, trans->transid); |
e02119d5 CM |
3996 | |
3997 | /* | |
3998 | * we didn't find anything from this transaction, see if there | |
3999 | * is anything at all | |
4000 | */ | |
339d0354 FM |
4001 | if (ret != 0 || min_key.objectid != ino || |
4002 | min_key.type != BTRFS_DIR_INDEX_KEY) { | |
33345d01 | 4003 | min_key.objectid = ino; |
339d0354 | 4004 | min_key.type = BTRFS_DIR_INDEX_KEY; |
e02119d5 | 4005 | min_key.offset = (u64)-1; |
b3b4aa74 | 4006 | btrfs_release_path(path); |
e02119d5 CM |
4007 | ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0); |
4008 | if (ret < 0) { | |
b3b4aa74 | 4009 | btrfs_release_path(path); |
e02119d5 CM |
4010 | return ret; |
4011 | } | |
339d0354 | 4012 | ret = btrfs_previous_item(root, path, ino, BTRFS_DIR_INDEX_KEY); |
e02119d5 CM |
4013 | |
4014 | /* if ret == 0 there are items for this type, | |
4015 | * create a range to tell us the last key of this type. | |
4016 | * otherwise, there are no items in this directory after | |
4017 | * *min_offset, and we create a range to indicate that. | |
4018 | */ | |
4019 | if (ret == 0) { | |
4020 | struct btrfs_key tmp; | |
732d591a | 4021 | |
e02119d5 CM |
4022 | btrfs_item_key_to_cpu(path->nodes[0], &tmp, |
4023 | path->slots[0]); | |
339d0354 | 4024 | if (tmp.type == BTRFS_DIR_INDEX_KEY) |
732d591a | 4025 | last_old_dentry_offset = tmp.offset; |
e02119d5 CM |
4026 | } |
4027 | goto done; | |
4028 | } | |
4029 | ||
4030 | /* go backward to find any previous key */ | |
339d0354 | 4031 | ret = btrfs_previous_item(root, path, ino, BTRFS_DIR_INDEX_KEY); |
e02119d5 CM |
4032 | if (ret == 0) { |
4033 | struct btrfs_key tmp; | |
a450a4af | 4034 | |
e02119d5 | 4035 | btrfs_item_key_to_cpu(path->nodes[0], &tmp, path->slots[0]); |
a450a4af FM |
4036 | /* |
4037 | * The dir index key before the first one we found that needs to | |
4038 | * be logged might be in a previous leaf, and there might be a | |
4039 | * gap between these keys, meaning that we had deletions that | |
4040 | * happened. So the key range item we log (key type | |
4041 | * BTRFS_DIR_LOG_INDEX_KEY) must cover a range that starts at the | |
4042 | * previous key's offset plus 1, so that those deletes are replayed. | |
4043 | */ | |
4044 | if (tmp.type == BTRFS_DIR_INDEX_KEY) | |
732d591a | 4045 | last_old_dentry_offset = tmp.offset; |
e02119d5 | 4046 | } |
b3b4aa74 | 4047 | btrfs_release_path(path); |
e02119d5 | 4048 | |
2cc83342 JB |
4049 | /* |
4050 | * Find the first key from this transaction again. See the note for | |
4051 | * log_new_dir_dentries, if we're logging a directory recursively we | |
4052 | * won't be holding its i_mutex, which means we can modify the directory | |
4053 | * while we're logging it. If we remove an entry between our first | |
4054 | * search and this search we'll not find the key again and can just | |
4055 | * bail. | |
4056 | */ | |
bb56f02f | 4057 | search: |
e02119d5 | 4058 | ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0); |
2cc83342 | 4059 | if (ret != 0) |
e02119d5 | 4060 | goto done; |
e02119d5 CM |
4061 | |
4062 | /* | |
4063 | * we have a block from this transaction, log every item in it | |
4064 | * from our directory | |
4065 | */ | |
d397712b | 4066 | while (1) { |
732d591a FM |
4067 | ret = process_dir_items_leaf(trans, inode, path, dst_path, ctx, |
4068 | &last_old_dentry_offset); | |
eb10d85e FM |
4069 | if (ret != 0) { |
4070 | if (ret < 0) | |
4a500fd1 | 4071 | err = ret; |
eb10d85e | 4072 | goto done; |
e02119d5 | 4073 | } |
eb10d85e | 4074 | path->slots[0] = btrfs_header_nritems(path->nodes[0]); |
e02119d5 CM |
4075 | |
4076 | /* | |
4077 | * look ahead to the next item and see if it is also | |
4078 | * from this directory and from this transaction | |
4079 | */ | |
4080 | ret = btrfs_next_leaf(root, path); | |
80c0b421 LB |
4081 | if (ret) { |
4082 | if (ret == 1) | |
4083 | last_offset = (u64)-1; | |
4084 | else | |
4085 | err = ret; | |
e02119d5 CM |
4086 | goto done; |
4087 | } | |
eb10d85e | 4088 | btrfs_item_key_to_cpu(path->nodes[0], &min_key, path->slots[0]); |
339d0354 | 4089 | if (min_key.objectid != ino || min_key.type != BTRFS_DIR_INDEX_KEY) { |
e02119d5 CM |
4090 | last_offset = (u64)-1; |
4091 | goto done; | |
4092 | } | |
4093 | if (btrfs_header_generation(path->nodes[0]) != trans->transid) { | |
a450a4af FM |
4094 | /* |
4095 | * The next leaf was not changed in the current transaction | |
4096 | * and has at least one dir index key. | |
4097 | * We check for the next key because there might have been | |
4098 | * one or more deletions between the last key we logged and | |
4099 | * that next key. So the key range item we log (key type | |
4100 | * BTRFS_DIR_LOG_INDEX_KEY) must end at the next key's | |
4101 | * offset minus 1, so that those deletes are replayed. | |
4102 | */ | |
4103 | last_offset = min_key.offset - 1; | |
e02119d5 CM |
4104 | goto done; |
4105 | } | |
eb10d85e FM |
4106 | if (need_resched()) { |
4107 | btrfs_release_path(path); | |
4108 | cond_resched(); | |
4109 | goto search; | |
4110 | } | |
e02119d5 CM |
4111 | } |
4112 | done: | |
b3b4aa74 DS |
4113 | btrfs_release_path(path); |
4114 | btrfs_release_path(dst_path); | |
e02119d5 | 4115 | |
4a500fd1 YZ |
4116 | if (err == 0) { |
4117 | *last_offset_ret = last_offset; | |
4118 | /* | |
732d591a FM |
4119 | * In case the leaf was changed in the current transaction but |
4120 | * all its dir items are from a past transaction, the last item | |
4121 | * in the leaf is a dir item and there's no gap between that last | |
4122 | * dir item and the first one on the next leaf (which did not | |
4123 | * change in the current transaction), then we don't need to log | |
4124 | * a range, last_old_dentry_offset is == to last_offset. | |
4a500fd1 | 4125 | */ |
732d591a FM |
4126 | ASSERT(last_old_dentry_offset <= last_offset); |
4127 | if (last_old_dentry_offset < last_offset) { | |
4128 | ret = insert_dir_log_key(trans, log, path, ino, | |
4129 | last_old_dentry_offset + 1, | |
4130 | last_offset); | |
4131 | if (ret) | |
4132 | err = ret; | |
4133 | } | |
4a500fd1 YZ |
4134 | } |
4135 | return err; | |
e02119d5 CM |
4136 | } |
4137 | ||
4138 | /* | |
4139 | * logging directories is very similar to logging inodes, We find all the items | |
4140 | * from the current transaction and write them to the log. | |
4141 | * | |
4142 | * The recovery code scans the directory in the subvolume, and if it finds a | |
4143 | * key in the range logged that is not present in the log tree, then it means | |
4144 | * that dir entry was unlinked during the transaction. | |
4145 | * | |
4146 | * In order for that scan to work, we must include one key smaller than | |
4147 | * the smallest logged by this transaction and one key larger than the largest | |
4148 | * key logged by this transaction. | |
4149 | */ | |
4150 | static noinline int log_directory_changes(struct btrfs_trans_handle *trans, | |
90d04510 | 4151 | struct btrfs_inode *inode, |
e02119d5 | 4152 | struct btrfs_path *path, |
2f2ff0ee FM |
4153 | struct btrfs_path *dst_path, |
4154 | struct btrfs_log_ctx *ctx) | |
e02119d5 CM |
4155 | { |
4156 | u64 min_key; | |
4157 | u64 max_key; | |
4158 | int ret; | |
e02119d5 | 4159 | |
732d591a | 4160 | min_key = BTRFS_DIR_START_INDEX; |
e02119d5 | 4161 | max_key = 0; |
339d0354 | 4162 | ctx->last_dir_item_offset = inode->last_dir_index_offset; |
dc287224 | 4163 | |
d397712b | 4164 | while (1) { |
339d0354 | 4165 | ret = log_dir_items(trans, inode, path, dst_path, |
dbf39ea4 | 4166 | ctx, min_key, &max_key); |
4a500fd1 YZ |
4167 | if (ret) |
4168 | return ret; | |
e02119d5 CM |
4169 | if (max_key == (u64)-1) |
4170 | break; | |
4171 | min_key = max_key + 1; | |
4172 | } | |
4173 | ||
339d0354 FM |
4174 | inode->last_dir_index_offset = ctx->last_dir_item_offset; |
4175 | ||
e02119d5 CM |
4176 | return 0; |
4177 | } | |
4178 | ||
4179 | /* | |
4180 | * a helper function to drop items from the log before we relog an | |
4181 | * inode. max_key_type indicates the highest item type to remove. | |
4182 | * This cannot be run for file data extents because it does not | |
4183 | * free the extents they point to. | |
4184 | */ | |
88e221cd | 4185 | static int drop_inode_items(struct btrfs_trans_handle *trans, |
e02119d5 CM |
4186 | struct btrfs_root *log, |
4187 | struct btrfs_path *path, | |
88e221cd FM |
4188 | struct btrfs_inode *inode, |
4189 | int max_key_type) | |
e02119d5 CM |
4190 | { |
4191 | int ret; | |
4192 | struct btrfs_key key; | |
4193 | struct btrfs_key found_key; | |
18ec90d6 | 4194 | int start_slot; |
e02119d5 | 4195 | |
88e221cd | 4196 | key.objectid = btrfs_ino(inode); |
e02119d5 CM |
4197 | key.type = max_key_type; |
4198 | key.offset = (u64)-1; | |
4199 | ||
d397712b | 4200 | while (1) { |
e02119d5 | 4201 | ret = btrfs_search_slot(trans, log, &key, path, -1, 1); |
3650860b | 4202 | BUG_ON(ret == 0); /* Logic error */ |
4a500fd1 | 4203 | if (ret < 0) |
e02119d5 CM |
4204 | break; |
4205 | ||
4206 | if (path->slots[0] == 0) | |
4207 | break; | |
4208 | ||
4209 | path->slots[0]--; | |
4210 | btrfs_item_key_to_cpu(path->nodes[0], &found_key, | |
4211 | path->slots[0]); | |
4212 | ||
88e221cd | 4213 | if (found_key.objectid != key.objectid) |
e02119d5 CM |
4214 | break; |
4215 | ||
18ec90d6 JB |
4216 | found_key.offset = 0; |
4217 | found_key.type = 0; | |
e3b83361 | 4218 | ret = btrfs_bin_search(path->nodes[0], &found_key, &start_slot); |
cbca7d59 FM |
4219 | if (ret < 0) |
4220 | break; | |
18ec90d6 JB |
4221 | |
4222 | ret = btrfs_del_items(trans, log, path, start_slot, | |
4223 | path->slots[0] - start_slot + 1); | |
4224 | /* | |
4225 | * If start slot isn't 0 then we don't need to re-search, we've | |
4226 | * found the last guy with the objectid in this tree. | |
4227 | */ | |
4228 | if (ret || start_slot != 0) | |
65a246c5 | 4229 | break; |
b3b4aa74 | 4230 | btrfs_release_path(path); |
e02119d5 | 4231 | } |
b3b4aa74 | 4232 | btrfs_release_path(path); |
5bdbeb21 JB |
4233 | if (ret > 0) |
4234 | ret = 0; | |
4a500fd1 | 4235 | return ret; |
e02119d5 CM |
4236 | } |
4237 | ||
8a2b3da1 FM |
4238 | static int truncate_inode_items(struct btrfs_trans_handle *trans, |
4239 | struct btrfs_root *log_root, | |
4240 | struct btrfs_inode *inode, | |
4241 | u64 new_size, u32 min_type) | |
4242 | { | |
d9ac19c3 JB |
4243 | struct btrfs_truncate_control control = { |
4244 | .new_size = new_size, | |
487e81d2 | 4245 | .ino = btrfs_ino(inode), |
d9ac19c3 | 4246 | .min_type = min_type, |
5caa490e | 4247 | .skip_ref_updates = true, |
d9ac19c3 | 4248 | }; |
8a2b3da1 | 4249 | |
8697b8f8 | 4250 | return btrfs_truncate_inode_items(trans, log_root, &control); |
8a2b3da1 FM |
4251 | } |
4252 | ||
94edf4ae JB |
4253 | static void fill_inode_item(struct btrfs_trans_handle *trans, |
4254 | struct extent_buffer *leaf, | |
4255 | struct btrfs_inode_item *item, | |
1a4bcf47 FM |
4256 | struct inode *inode, int log_inode_only, |
4257 | u64 logged_isize) | |
94edf4ae | 4258 | { |
0b1c6cca | 4259 | struct btrfs_map_token token; |
77eea05e | 4260 | u64 flags; |
0b1c6cca | 4261 | |
c82f823c | 4262 | btrfs_init_map_token(&token, leaf); |
94edf4ae JB |
4263 | |
4264 | if (log_inode_only) { | |
4265 | /* set the generation to zero so the recover code | |
4266 | * can tell the difference between an logging | |
4267 | * just to say 'this inode exists' and a logging | |
4268 | * to say 'update this inode with these values' | |
4269 | */ | |
cc4c13d5 DS |
4270 | btrfs_set_token_inode_generation(&token, item, 0); |
4271 | btrfs_set_token_inode_size(&token, item, logged_isize); | |
94edf4ae | 4272 | } else { |
cc4c13d5 DS |
4273 | btrfs_set_token_inode_generation(&token, item, |
4274 | BTRFS_I(inode)->generation); | |
4275 | btrfs_set_token_inode_size(&token, item, inode->i_size); | |
0b1c6cca JB |
4276 | } |
4277 | ||
cc4c13d5 DS |
4278 | btrfs_set_token_inode_uid(&token, item, i_uid_read(inode)); |
4279 | btrfs_set_token_inode_gid(&token, item, i_gid_read(inode)); | |
4280 | btrfs_set_token_inode_mode(&token, item, inode->i_mode); | |
4281 | btrfs_set_token_inode_nlink(&token, item, inode->i_nlink); | |
4282 | ||
4283 | btrfs_set_token_timespec_sec(&token, &item->atime, | |
4284 | inode->i_atime.tv_sec); | |
4285 | btrfs_set_token_timespec_nsec(&token, &item->atime, | |
4286 | inode->i_atime.tv_nsec); | |
4287 | ||
4288 | btrfs_set_token_timespec_sec(&token, &item->mtime, | |
4289 | inode->i_mtime.tv_sec); | |
4290 | btrfs_set_token_timespec_nsec(&token, &item->mtime, | |
4291 | inode->i_mtime.tv_nsec); | |
4292 | ||
4293 | btrfs_set_token_timespec_sec(&token, &item->ctime, | |
4294 | inode->i_ctime.tv_sec); | |
4295 | btrfs_set_token_timespec_nsec(&token, &item->ctime, | |
4296 | inode->i_ctime.tv_nsec); | |
4297 | ||
e593e54e FM |
4298 | /* |
4299 | * We do not need to set the nbytes field, in fact during a fast fsync | |
4300 | * its value may not even be correct, since a fast fsync does not wait | |
4301 | * for ordered extent completion, which is where we update nbytes, it | |
4302 | * only waits for writeback to complete. During log replay as we find | |
4303 | * file extent items and replay them, we adjust the nbytes field of the | |
4304 | * inode item in subvolume tree as needed (see overwrite_item()). | |
4305 | */ | |
cc4c13d5 DS |
4306 | |
4307 | btrfs_set_token_inode_sequence(&token, item, inode_peek_iversion(inode)); | |
4308 | btrfs_set_token_inode_transid(&token, item, trans->transid); | |
4309 | btrfs_set_token_inode_rdev(&token, item, inode->i_rdev); | |
77eea05e BB |
4310 | flags = btrfs_inode_combine_flags(BTRFS_I(inode)->flags, |
4311 | BTRFS_I(inode)->ro_flags); | |
4312 | btrfs_set_token_inode_flags(&token, item, flags); | |
cc4c13d5 | 4313 | btrfs_set_token_inode_block_group(&token, item, 0); |
94edf4ae JB |
4314 | } |
4315 | ||
a95249b3 JB |
4316 | static int log_inode_item(struct btrfs_trans_handle *trans, |
4317 | struct btrfs_root *log, struct btrfs_path *path, | |
2ac691d8 | 4318 | struct btrfs_inode *inode, bool inode_item_dropped) |
a95249b3 JB |
4319 | { |
4320 | struct btrfs_inode_item *inode_item; | |
a95249b3 JB |
4321 | int ret; |
4322 | ||
2ac691d8 FM |
4323 | /* |
4324 | * If we are doing a fast fsync and the inode was logged before in the | |
4325 | * current transaction, then we know the inode was previously logged and | |
4326 | * it exists in the log tree. For performance reasons, in this case use | |
4327 | * btrfs_search_slot() directly with ins_len set to 0 so that we never | |
4328 | * attempt a write lock on the leaf's parent, which adds unnecessary lock | |
4329 | * contention in case there are concurrent fsyncs for other inodes of the | |
4330 | * same subvolume. Using btrfs_insert_empty_item() when the inode item | |
4331 | * already exists can also result in unnecessarily splitting a leaf. | |
4332 | */ | |
4333 | if (!inode_item_dropped && inode->logged_trans == trans->transid) { | |
4334 | ret = btrfs_search_slot(trans, log, &inode->location, path, 0, 1); | |
4335 | ASSERT(ret <= 0); | |
4336 | if (ret > 0) | |
4337 | ret = -ENOENT; | |
4338 | } else { | |
4339 | /* | |
4340 | * This means it is the first fsync in the current transaction, | |
4341 | * so the inode item is not in the log and we need to insert it. | |
4342 | * We can never get -EEXIST because we are only called for a fast | |
4343 | * fsync and in case an inode eviction happens after the inode was | |
4344 | * logged before in the current transaction, when we load again | |
4345 | * the inode, we set BTRFS_INODE_NEEDS_FULL_SYNC on its runtime | |
4346 | * flags and set ->logged_trans to 0. | |
4347 | */ | |
4348 | ret = btrfs_insert_empty_item(trans, log, path, &inode->location, | |
4349 | sizeof(*inode_item)); | |
4350 | ASSERT(ret != -EEXIST); | |
4351 | } | |
4352 | if (ret) | |
a95249b3 JB |
4353 | return ret; |
4354 | inode_item = btrfs_item_ptr(path->nodes[0], path->slots[0], | |
4355 | struct btrfs_inode_item); | |
6d889a3b NB |
4356 | fill_inode_item(trans, path->nodes[0], inode_item, &inode->vfs_inode, |
4357 | 0, 0); | |
a95249b3 JB |
4358 | btrfs_release_path(path); |
4359 | return 0; | |
4360 | } | |
4361 | ||
40e046ac | 4362 | static int log_csums(struct btrfs_trans_handle *trans, |
3ebac17c | 4363 | struct btrfs_inode *inode, |
40e046ac FM |
4364 | struct btrfs_root *log_root, |
4365 | struct btrfs_ordered_sum *sums) | |
4366 | { | |
e289f03e FM |
4367 | const u64 lock_end = sums->bytenr + sums->len - 1; |
4368 | struct extent_state *cached_state = NULL; | |
40e046ac FM |
4369 | int ret; |
4370 | ||
3ebac17c FM |
4371 | /* |
4372 | * If this inode was not used for reflink operations in the current | |
4373 | * transaction with new extents, then do the fast path, no need to | |
4374 | * worry about logging checksum items with overlapping ranges. | |
4375 | */ | |
4376 | if (inode->last_reflink_trans < trans->transid) | |
4377 | return btrfs_csum_file_blocks(trans, log_root, sums); | |
4378 | ||
e289f03e FM |
4379 | /* |
4380 | * Serialize logging for checksums. This is to avoid racing with the | |
4381 | * same checksum being logged by another task that is logging another | |
4382 | * file which happens to refer to the same extent as well. Such races | |
4383 | * can leave checksum items in the log with overlapping ranges. | |
4384 | */ | |
4385 | ret = lock_extent_bits(&log_root->log_csum_range, sums->bytenr, | |
4386 | lock_end, &cached_state); | |
4387 | if (ret) | |
4388 | return ret; | |
40e046ac FM |
4389 | /* |
4390 | * Due to extent cloning, we might have logged a csum item that covers a | |
4391 | * subrange of a cloned extent, and later we can end up logging a csum | |
4392 | * item for a larger subrange of the same extent or the entire range. | |
4393 | * This would leave csum items in the log tree that cover the same range | |
4394 | * and break the searches for checksums in the log tree, resulting in | |
4395 | * some checksums missing in the fs/subvolume tree. So just delete (or | |
4396 | * trim and adjust) any existing csum items in the log for this range. | |
4397 | */ | |
4398 | ret = btrfs_del_csums(trans, log_root, sums->bytenr, sums->len); | |
e289f03e FM |
4399 | if (!ret) |
4400 | ret = btrfs_csum_file_blocks(trans, log_root, sums); | |
40e046ac | 4401 | |
e289f03e FM |
4402 | unlock_extent_cached(&log_root->log_csum_range, sums->bytenr, lock_end, |
4403 | &cached_state); | |
4404 | ||
4405 | return ret; | |
40e046ac FM |
4406 | } |
4407 | ||
31ff1cd2 | 4408 | static noinline int copy_items(struct btrfs_trans_handle *trans, |
44d70e19 | 4409 | struct btrfs_inode *inode, |
31ff1cd2 | 4410 | struct btrfs_path *dst_path, |
0e56315c | 4411 | struct btrfs_path *src_path, |
1a4bcf47 FM |
4412 | int start_slot, int nr, int inode_only, |
4413 | u64 logged_isize) | |
31ff1cd2 | 4414 | { |
44d70e19 | 4415 | struct btrfs_root *log = inode->root->log_root; |
31ff1cd2 | 4416 | struct btrfs_file_extent_item *extent; |
16e7549f | 4417 | struct extent_buffer *src = src_path->nodes[0]; |
7f30c072 | 4418 | int ret = 0; |
31ff1cd2 CM |
4419 | struct btrfs_key *ins_keys; |
4420 | u32 *ins_sizes; | |
b7ef5f3a | 4421 | struct btrfs_item_batch batch; |
31ff1cd2 CM |
4422 | char *ins_data; |
4423 | int i; | |
7f30c072 | 4424 | int dst_index; |
7f30c072 FM |
4425 | const bool skip_csum = (inode->flags & BTRFS_INODE_NODATASUM); |
4426 | const u64 i_size = i_size_read(&inode->vfs_inode); | |
d20f7043 | 4427 | |
31ff1cd2 CM |
4428 | ins_data = kmalloc(nr * sizeof(struct btrfs_key) + |
4429 | nr * sizeof(u32), GFP_NOFS); | |
2a29edc6 | 4430 | if (!ins_data) |
4431 | return -ENOMEM; | |
4432 | ||
31ff1cd2 CM |
4433 | ins_sizes = (u32 *)ins_data; |
4434 | ins_keys = (struct btrfs_key *)(ins_data + nr * sizeof(u32)); | |
b7ef5f3a FM |
4435 | batch.keys = ins_keys; |
4436 | batch.data_sizes = ins_sizes; | |
4437 | batch.total_data_size = 0; | |
7f30c072 | 4438 | batch.nr = 0; |
31ff1cd2 | 4439 | |
7f30c072 | 4440 | dst_index = 0; |
31ff1cd2 | 4441 | for (i = 0; i < nr; i++) { |
7f30c072 FM |
4442 | const int src_slot = start_slot + i; |
4443 | struct btrfs_root *csum_root; | |
5b7ce5e2 FM |
4444 | struct btrfs_ordered_sum *sums; |
4445 | struct btrfs_ordered_sum *sums_next; | |
4446 | LIST_HEAD(ordered_sums); | |
7f30c072 FM |
4447 | u64 disk_bytenr; |
4448 | u64 disk_num_bytes; | |
4449 | u64 extent_offset; | |
4450 | u64 extent_num_bytes; | |
4451 | bool is_old_extent; | |
4452 | ||
4453 | btrfs_item_key_to_cpu(src, &ins_keys[dst_index], src_slot); | |
4454 | ||
4455 | if (ins_keys[dst_index].type != BTRFS_EXTENT_DATA_KEY) | |
4456 | goto add_to_batch; | |
4457 | ||
4458 | extent = btrfs_item_ptr(src, src_slot, | |
4459 | struct btrfs_file_extent_item); | |
4460 | ||
4461 | is_old_extent = (btrfs_file_extent_generation(src, extent) < | |
4462 | trans->transid); | |
4463 | ||
4464 | /* | |
4465 | * Don't copy extents from past generations. That would make us | |
4466 | * log a lot more metadata for common cases like doing only a | |
4467 | * few random writes into a file and then fsync it for the first | |
4468 | * time or after the full sync flag is set on the inode. We can | |
4469 | * get leaves full of extent items, most of which are from past | |
4470 | * generations, so we can skip them - as long as the inode has | |
4471 | * not been the target of a reflink operation in this transaction, | |
4472 | * as in that case it might have had file extent items with old | |
4473 | * generations copied into it. We also must always log prealloc | |
4474 | * extents that start at or beyond eof, otherwise we would lose | |
4475 | * them on log replay. | |
4476 | */ | |
4477 | if (is_old_extent && | |
4478 | ins_keys[dst_index].offset < i_size && | |
4479 | inode->last_reflink_trans < trans->transid) | |
4480 | continue; | |
4481 | ||
4482 | if (skip_csum) | |
4483 | goto add_to_batch; | |
4484 | ||
4485 | /* Only regular extents have checksums. */ | |
4486 | if (btrfs_file_extent_type(src, extent) != BTRFS_FILE_EXTENT_REG) | |
4487 | goto add_to_batch; | |
4488 | ||
4489 | /* | |
4490 | * If it's an extent created in a past transaction, then its | |
4491 | * checksums are already accessible from the committed csum tree, | |
4492 | * no need to log them. | |
4493 | */ | |
4494 | if (is_old_extent) | |
4495 | goto add_to_batch; | |
4496 | ||
4497 | disk_bytenr = btrfs_file_extent_disk_bytenr(src, extent); | |
4498 | /* If it's an explicit hole, there are no checksums. */ | |
4499 | if (disk_bytenr == 0) | |
4500 | goto add_to_batch; | |
4501 | ||
4502 | disk_num_bytes = btrfs_file_extent_disk_num_bytes(src, extent); | |
4503 | ||
4504 | if (btrfs_file_extent_compression(src, extent)) { | |
4505 | extent_offset = 0; | |
4506 | extent_num_bytes = disk_num_bytes; | |
4507 | } else { | |
4508 | extent_offset = btrfs_file_extent_offset(src, extent); | |
4509 | extent_num_bytes = btrfs_file_extent_num_bytes(src, extent); | |
4510 | } | |
4511 | ||
4512 | csum_root = btrfs_csum_root(trans->fs_info, disk_bytenr); | |
4513 | disk_bytenr += extent_offset; | |
4514 | ret = btrfs_lookup_csums_range(csum_root, disk_bytenr, | |
4515 | disk_bytenr + extent_num_bytes - 1, | |
4516 | &ordered_sums, 0); | |
4517 | if (ret) | |
4518 | goto out; | |
4519 | ||
5b7ce5e2 FM |
4520 | list_for_each_entry_safe(sums, sums_next, &ordered_sums, list) { |
4521 | if (!ret) | |
4522 | ret = log_csums(trans, inode, log, sums); | |
4523 | list_del(&sums->list); | |
4524 | kfree(sums); | |
4525 | } | |
4526 | if (ret) | |
4527 | goto out; | |
4528 | ||
7f30c072 FM |
4529 | add_to_batch: |
4530 | ins_sizes[dst_index] = btrfs_item_size(src, src_slot); | |
4531 | batch.total_data_size += ins_sizes[dst_index]; | |
4532 | batch.nr++; | |
4533 | dst_index++; | |
31ff1cd2 | 4534 | } |
7f30c072 FM |
4535 | |
4536 | /* | |
4537 | * We have a leaf full of old extent items that don't need to be logged, | |
4538 | * so we don't need to do anything. | |
4539 | */ | |
4540 | if (batch.nr == 0) | |
4541 | goto out; | |
4542 | ||
b7ef5f3a | 4543 | ret = btrfs_insert_empty_items(trans, log, dst_path, &batch); |
7f30c072 FM |
4544 | if (ret) |
4545 | goto out; | |
4546 | ||
4547 | dst_index = 0; | |
4548 | for (i = 0; i < nr; i++) { | |
4549 | const int src_slot = start_slot + i; | |
4550 | const int dst_slot = dst_path->slots[0] + dst_index; | |
4551 | struct btrfs_key key; | |
4552 | unsigned long src_offset; | |
4553 | unsigned long dst_offset; | |
4554 | ||
4555 | /* | |
4556 | * We're done, all the remaining items in the source leaf | |
4557 | * correspond to old file extent items. | |
4558 | */ | |
4559 | if (dst_index >= batch.nr) | |
4560 | break; | |
4561 | ||
4562 | btrfs_item_key_to_cpu(src, &key, src_slot); | |
4563 | ||
4564 | if (key.type != BTRFS_EXTENT_DATA_KEY) | |
4565 | goto copy_item; | |
31ff1cd2 | 4566 | |
7f30c072 FM |
4567 | extent = btrfs_item_ptr(src, src_slot, |
4568 | struct btrfs_file_extent_item); | |
31ff1cd2 | 4569 | |
7f30c072 FM |
4570 | /* See the comment in the previous loop, same logic. */ |
4571 | if (btrfs_file_extent_generation(src, extent) < trans->transid && | |
4572 | key.offset < i_size && | |
4573 | inode->last_reflink_trans < trans->transid) | |
4574 | continue; | |
4575 | ||
4576 | copy_item: | |
4577 | dst_offset = btrfs_item_ptr_offset(dst_path->nodes[0], dst_slot); | |
4578 | src_offset = btrfs_item_ptr_offset(src, src_slot); | |
31ff1cd2 | 4579 | |
7f30c072 FM |
4580 | if (key.type == BTRFS_INODE_ITEM_KEY) { |
4581 | struct btrfs_inode_item *inode_item; | |
4582 | ||
4583 | inode_item = btrfs_item_ptr(dst_path->nodes[0], dst_slot, | |
31ff1cd2 | 4584 | struct btrfs_inode_item); |
94edf4ae | 4585 | fill_inode_item(trans, dst_path->nodes[0], inode_item, |
f85b7379 DS |
4586 | &inode->vfs_inode, |
4587 | inode_only == LOG_INODE_EXISTS, | |
1a4bcf47 | 4588 | logged_isize); |
94edf4ae JB |
4589 | } else { |
4590 | copy_extent_buffer(dst_path->nodes[0], src, dst_offset, | |
7f30c072 | 4591 | src_offset, ins_sizes[dst_index]); |
31ff1cd2 | 4592 | } |
94edf4ae | 4593 | |
7f30c072 | 4594 | dst_index++; |
31ff1cd2 CM |
4595 | } |
4596 | ||
4597 | btrfs_mark_buffer_dirty(dst_path->nodes[0]); | |
b3b4aa74 | 4598 | btrfs_release_path(dst_path); |
7f30c072 | 4599 | out: |
31ff1cd2 | 4600 | kfree(ins_data); |
d20f7043 | 4601 | |
4a500fd1 | 4602 | return ret; |
31ff1cd2 CM |
4603 | } |
4604 | ||
4f0f586b ST |
4605 | static int extent_cmp(void *priv, const struct list_head *a, |
4606 | const struct list_head *b) | |
5dc562c5 | 4607 | { |
214cc184 | 4608 | const struct extent_map *em1, *em2; |
5dc562c5 JB |
4609 | |
4610 | em1 = list_entry(a, struct extent_map, list); | |
4611 | em2 = list_entry(b, struct extent_map, list); | |
4612 | ||
4613 | if (em1->start < em2->start) | |
4614 | return -1; | |
4615 | else if (em1->start > em2->start) | |
4616 | return 1; | |
4617 | return 0; | |
4618 | } | |
4619 | ||
e7175a69 JB |
4620 | static int log_extent_csums(struct btrfs_trans_handle *trans, |
4621 | struct btrfs_inode *inode, | |
a9ecb653 | 4622 | struct btrfs_root *log_root, |
48778179 FM |
4623 | const struct extent_map *em, |
4624 | struct btrfs_log_ctx *ctx) | |
5dc562c5 | 4625 | { |
48778179 | 4626 | struct btrfs_ordered_extent *ordered; |
fc28b25e | 4627 | struct btrfs_root *csum_root; |
2ab28f32 JB |
4628 | u64 csum_offset; |
4629 | u64 csum_len; | |
48778179 FM |
4630 | u64 mod_start = em->mod_start; |
4631 | u64 mod_len = em->mod_len; | |
8407f553 FM |
4632 | LIST_HEAD(ordered_sums); |
4633 | int ret = 0; | |
0aa4a17d | 4634 | |
e7175a69 JB |
4635 | if (inode->flags & BTRFS_INODE_NODATASUM || |
4636 | test_bit(EXTENT_FLAG_PREALLOC, &em->flags) || | |
8407f553 | 4637 | em->block_start == EXTENT_MAP_HOLE) |
70c8a91c | 4638 | return 0; |
5dc562c5 | 4639 | |
48778179 FM |
4640 | list_for_each_entry(ordered, &ctx->ordered_extents, log_list) { |
4641 | const u64 ordered_end = ordered->file_offset + ordered->num_bytes; | |
4642 | const u64 mod_end = mod_start + mod_len; | |
4643 | struct btrfs_ordered_sum *sums; | |
4644 | ||
4645 | if (mod_len == 0) | |
4646 | break; | |
4647 | ||
4648 | if (ordered_end <= mod_start) | |
4649 | continue; | |
4650 | if (mod_end <= ordered->file_offset) | |
4651 | break; | |
4652 | ||
4653 | /* | |
4654 | * We are going to copy all the csums on this ordered extent, so | |
4655 | * go ahead and adjust mod_start and mod_len in case this ordered | |
4656 | * extent has already been logged. | |
4657 | */ | |
4658 | if (ordered->file_offset > mod_start) { | |
4659 | if (ordered_end >= mod_end) | |
4660 | mod_len = ordered->file_offset - mod_start; | |
4661 | /* | |
4662 | * If we have this case | |
4663 | * | |
4664 | * |--------- logged extent ---------| | |
4665 | * |----- ordered extent ----| | |
4666 | * | |
4667 | * Just don't mess with mod_start and mod_len, we'll | |
4668 | * just end up logging more csums than we need and it | |
4669 | * will be ok. | |
4670 | */ | |
4671 | } else { | |
4672 | if (ordered_end < mod_end) { | |
4673 | mod_len = mod_end - ordered_end; | |
4674 | mod_start = ordered_end; | |
4675 | } else { | |
4676 | mod_len = 0; | |
4677 | } | |
4678 | } | |
4679 | ||
4680 | /* | |
4681 | * To keep us from looping for the above case of an ordered | |
4682 | * extent that falls inside of the logged extent. | |
4683 | */ | |
4684 | if (test_and_set_bit(BTRFS_ORDERED_LOGGED_CSUM, &ordered->flags)) | |
4685 | continue; | |
4686 | ||
4687 | list_for_each_entry(sums, &ordered->list, list) { | |
4688 | ret = log_csums(trans, inode, log_root, sums); | |
4689 | if (ret) | |
4690 | return ret; | |
4691 | } | |
4692 | } | |
4693 | ||
4694 | /* We're done, found all csums in the ordered extents. */ | |
4695 | if (mod_len == 0) | |
4696 | return 0; | |
4697 | ||
e7175a69 | 4698 | /* If we're compressed we have to save the entire range of csums. */ |
488111aa FDBM |
4699 | if (em->compress_type) { |
4700 | csum_offset = 0; | |
8407f553 | 4701 | csum_len = max(em->block_len, em->orig_block_len); |
488111aa | 4702 | } else { |
48778179 FM |
4703 | csum_offset = mod_start - em->start; |
4704 | csum_len = mod_len; | |
488111aa | 4705 | } |
2ab28f32 | 4706 | |
70c8a91c | 4707 | /* block start is already adjusted for the file extent offset. */ |
fc28b25e JB |
4708 | csum_root = btrfs_csum_root(trans->fs_info, em->block_start); |
4709 | ret = btrfs_lookup_csums_range(csum_root, | |
70c8a91c JB |
4710 | em->block_start + csum_offset, |
4711 | em->block_start + csum_offset + | |
4712 | csum_len - 1, &ordered_sums, 0); | |
4713 | if (ret) | |
4714 | return ret; | |
5dc562c5 | 4715 | |
70c8a91c JB |
4716 | while (!list_empty(&ordered_sums)) { |
4717 | struct btrfs_ordered_sum *sums = list_entry(ordered_sums.next, | |
4718 | struct btrfs_ordered_sum, | |
4719 | list); | |
4720 | if (!ret) | |
3ebac17c | 4721 | ret = log_csums(trans, inode, log_root, sums); |
70c8a91c JB |
4722 | list_del(&sums->list); |
4723 | kfree(sums); | |
5dc562c5 JB |
4724 | } |
4725 | ||
70c8a91c | 4726 | return ret; |
5dc562c5 JB |
4727 | } |
4728 | ||
8407f553 | 4729 | static int log_one_extent(struct btrfs_trans_handle *trans, |
90d04510 | 4730 | struct btrfs_inode *inode, |
8407f553 FM |
4731 | const struct extent_map *em, |
4732 | struct btrfs_path *path, | |
8407f553 FM |
4733 | struct btrfs_log_ctx *ctx) |
4734 | { | |
5893dfb9 | 4735 | struct btrfs_drop_extents_args drop_args = { 0 }; |
90d04510 | 4736 | struct btrfs_root *log = inode->root->log_root; |
e1f53ed8 | 4737 | struct btrfs_file_extent_item fi = { 0 }; |
8407f553 | 4738 | struct extent_buffer *leaf; |
8407f553 FM |
4739 | struct btrfs_key key; |
4740 | u64 extent_offset = em->start - em->orig_start; | |
4741 | u64 block_len; | |
4742 | int ret; | |
8407f553 | 4743 | |
e1f53ed8 FM |
4744 | btrfs_set_stack_file_extent_generation(&fi, trans->transid); |
4745 | if (test_bit(EXTENT_FLAG_PREALLOC, &em->flags)) | |
4746 | btrfs_set_stack_file_extent_type(&fi, BTRFS_FILE_EXTENT_PREALLOC); | |
4747 | else | |
4748 | btrfs_set_stack_file_extent_type(&fi, BTRFS_FILE_EXTENT_REG); | |
4749 | ||
4750 | block_len = max(em->block_len, em->orig_block_len); | |
4751 | if (em->compress_type != BTRFS_COMPRESS_NONE) { | |
4752 | btrfs_set_stack_file_extent_disk_bytenr(&fi, em->block_start); | |
4753 | btrfs_set_stack_file_extent_disk_num_bytes(&fi, block_len); | |
4754 | } else if (em->block_start < EXTENT_MAP_LAST_BYTE) { | |
4755 | btrfs_set_stack_file_extent_disk_bytenr(&fi, em->block_start - | |
4756 | extent_offset); | |
4757 | btrfs_set_stack_file_extent_disk_num_bytes(&fi, block_len); | |
4758 | } | |
4759 | ||
4760 | btrfs_set_stack_file_extent_offset(&fi, extent_offset); | |
4761 | btrfs_set_stack_file_extent_num_bytes(&fi, em->len); | |
4762 | btrfs_set_stack_file_extent_ram_bytes(&fi, em->ram_bytes); | |
4763 | btrfs_set_stack_file_extent_compression(&fi, em->compress_type); | |
4764 | ||
48778179 | 4765 | ret = log_extent_csums(trans, inode, log, em, ctx); |
8407f553 FM |
4766 | if (ret) |
4767 | return ret; | |
4768 | ||
5328b2a7 FM |
4769 | /* |
4770 | * If this is the first time we are logging the inode in the current | |
4771 | * transaction, we can avoid btrfs_drop_extents(), which is expensive | |
4772 | * because it does a deletion search, which always acquires write locks | |
4773 | * for extent buffers at levels 2, 1 and 0. This not only wastes time | |
4774 | * but also adds significant contention in a log tree, since log trees | |
4775 | * are small, with a root at level 2 or 3 at most, due to their short | |
4776 | * life span. | |
4777 | */ | |
0f8ce498 | 4778 | if (ctx->logged_before) { |
5328b2a7 FM |
4779 | drop_args.path = path; |
4780 | drop_args.start = em->start; | |
4781 | drop_args.end = em->start + em->len; | |
4782 | drop_args.replace_extent = true; | |
e1f53ed8 | 4783 | drop_args.extent_item_size = sizeof(fi); |
5328b2a7 FM |
4784 | ret = btrfs_drop_extents(trans, log, inode, &drop_args); |
4785 | if (ret) | |
4786 | return ret; | |
4787 | } | |
8407f553 | 4788 | |
5893dfb9 | 4789 | if (!drop_args.extent_inserted) { |
9d122629 | 4790 | key.objectid = btrfs_ino(inode); |
8407f553 FM |
4791 | key.type = BTRFS_EXTENT_DATA_KEY; |
4792 | key.offset = em->start; | |
4793 | ||
4794 | ret = btrfs_insert_empty_item(trans, log, path, &key, | |
e1f53ed8 | 4795 | sizeof(fi)); |
8407f553 FM |
4796 | if (ret) |
4797 | return ret; | |
4798 | } | |
4799 | leaf = path->nodes[0]; | |
e1f53ed8 FM |
4800 | write_extent_buffer(leaf, &fi, |
4801 | btrfs_item_ptr_offset(leaf, path->slots[0]), | |
4802 | sizeof(fi)); | |
8407f553 FM |
4803 | btrfs_mark_buffer_dirty(leaf); |
4804 | ||
4805 | btrfs_release_path(path); | |
4806 | ||
4807 | return ret; | |
4808 | } | |
4809 | ||
31d11b83 FM |
4810 | /* |
4811 | * Log all prealloc extents beyond the inode's i_size to make sure we do not | |
d9947887 | 4812 | * lose them after doing a full/fast fsync and replaying the log. We scan the |
31d11b83 FM |
4813 | * subvolume's root instead of iterating the inode's extent map tree because |
4814 | * otherwise we can log incorrect extent items based on extent map conversion. | |
4815 | * That can happen due to the fact that extent maps are merged when they | |
4816 | * are not in the extent map tree's list of modified extents. | |
4817 | */ | |
4818 | static int btrfs_log_prealloc_extents(struct btrfs_trans_handle *trans, | |
4819 | struct btrfs_inode *inode, | |
4820 | struct btrfs_path *path) | |
4821 | { | |
4822 | struct btrfs_root *root = inode->root; | |
4823 | struct btrfs_key key; | |
4824 | const u64 i_size = i_size_read(&inode->vfs_inode); | |
4825 | const u64 ino = btrfs_ino(inode); | |
4826 | struct btrfs_path *dst_path = NULL; | |
0e56315c | 4827 | bool dropped_extents = false; |
f135cea3 FM |
4828 | u64 truncate_offset = i_size; |
4829 | struct extent_buffer *leaf; | |
4830 | int slot; | |
31d11b83 FM |
4831 | int ins_nr = 0; |
4832 | int start_slot; | |
4833 | int ret; | |
4834 | ||
4835 | if (!(inode->flags & BTRFS_INODE_PREALLOC)) | |
4836 | return 0; | |
4837 | ||
4838 | key.objectid = ino; | |
4839 | key.type = BTRFS_EXTENT_DATA_KEY; | |
4840 | key.offset = i_size; | |
4841 | ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); | |
4842 | if (ret < 0) | |
4843 | goto out; | |
4844 | ||
f135cea3 FM |
4845 | /* |
4846 | * We must check if there is a prealloc extent that starts before the | |
4847 | * i_size and crosses the i_size boundary. This is to ensure later we | |
4848 | * truncate down to the end of that extent and not to the i_size, as | |
4849 | * otherwise we end up losing part of the prealloc extent after a log | |
4850 | * replay and with an implicit hole if there is another prealloc extent | |
4851 | * that starts at an offset beyond i_size. | |
4852 | */ | |
4853 | ret = btrfs_previous_item(root, path, ino, BTRFS_EXTENT_DATA_KEY); | |
4854 | if (ret < 0) | |
4855 | goto out; | |
4856 | ||
4857 | if (ret == 0) { | |
4858 | struct btrfs_file_extent_item *ei; | |
4859 | ||
4860 | leaf = path->nodes[0]; | |
4861 | slot = path->slots[0]; | |
4862 | ei = btrfs_item_ptr(leaf, slot, struct btrfs_file_extent_item); | |
4863 | ||
4864 | if (btrfs_file_extent_type(leaf, ei) == | |
4865 | BTRFS_FILE_EXTENT_PREALLOC) { | |
4866 | u64 extent_end; | |
4867 | ||
4868 | btrfs_item_key_to_cpu(leaf, &key, slot); | |
4869 | extent_end = key.offset + | |
4870 | btrfs_file_extent_num_bytes(leaf, ei); | |
4871 | ||
4872 | if (extent_end > i_size) | |
4873 | truncate_offset = extent_end; | |
4874 | } | |
4875 | } else { | |
4876 | ret = 0; | |
4877 | } | |
4878 | ||
31d11b83 | 4879 | while (true) { |
f135cea3 FM |
4880 | leaf = path->nodes[0]; |
4881 | slot = path->slots[0]; | |
31d11b83 FM |
4882 | |
4883 | if (slot >= btrfs_header_nritems(leaf)) { | |
4884 | if (ins_nr > 0) { | |
4885 | ret = copy_items(trans, inode, dst_path, path, | |
0e56315c | 4886 | start_slot, ins_nr, 1, 0); |
31d11b83 FM |
4887 | if (ret < 0) |
4888 | goto out; | |
4889 | ins_nr = 0; | |
4890 | } | |
4891 | ret = btrfs_next_leaf(root, path); | |
4892 | if (ret < 0) | |
4893 | goto out; | |
4894 | if (ret > 0) { | |
4895 | ret = 0; | |
4896 | break; | |
4897 | } | |
4898 | continue; | |
4899 | } | |
4900 | ||
4901 | btrfs_item_key_to_cpu(leaf, &key, slot); | |
4902 | if (key.objectid > ino) | |
4903 | break; | |
4904 | if (WARN_ON_ONCE(key.objectid < ino) || | |
4905 | key.type < BTRFS_EXTENT_DATA_KEY || | |
4906 | key.offset < i_size) { | |
4907 | path->slots[0]++; | |
4908 | continue; | |
4909 | } | |
0e56315c | 4910 | if (!dropped_extents) { |
31d11b83 FM |
4911 | /* |
4912 | * Avoid logging extent items logged in past fsync calls | |
4913 | * and leading to duplicate keys in the log tree. | |
4914 | */ | |
8a2b3da1 FM |
4915 | ret = truncate_inode_items(trans, root->log_root, inode, |
4916 | truncate_offset, | |
4917 | BTRFS_EXTENT_DATA_KEY); | |
31d11b83 FM |
4918 | if (ret) |
4919 | goto out; | |
0e56315c | 4920 | dropped_extents = true; |
31d11b83 FM |
4921 | } |
4922 | if (ins_nr == 0) | |
4923 | start_slot = slot; | |
4924 | ins_nr++; | |
4925 | path->slots[0]++; | |
4926 | if (!dst_path) { | |
4927 | dst_path = btrfs_alloc_path(); | |
4928 | if (!dst_path) { | |
4929 | ret = -ENOMEM; | |
4930 | goto out; | |
4931 | } | |
4932 | } | |
4933 | } | |
0bc2d3c0 | 4934 | if (ins_nr > 0) |
0e56315c | 4935 | ret = copy_items(trans, inode, dst_path, path, |
31d11b83 | 4936 | start_slot, ins_nr, 1, 0); |
31d11b83 FM |
4937 | out: |
4938 | btrfs_release_path(path); | |
4939 | btrfs_free_path(dst_path); | |
4940 | return ret; | |
4941 | } | |
4942 | ||
5dc562c5 | 4943 | static int btrfs_log_changed_extents(struct btrfs_trans_handle *trans, |
9d122629 | 4944 | struct btrfs_inode *inode, |
827463c4 | 4945 | struct btrfs_path *path, |
48778179 | 4946 | struct btrfs_log_ctx *ctx) |
5dc562c5 | 4947 | { |
48778179 FM |
4948 | struct btrfs_ordered_extent *ordered; |
4949 | struct btrfs_ordered_extent *tmp; | |
5dc562c5 JB |
4950 | struct extent_map *em, *n; |
4951 | struct list_head extents; | |
9d122629 | 4952 | struct extent_map_tree *tree = &inode->extent_tree; |
5dc562c5 | 4953 | int ret = 0; |
2ab28f32 | 4954 | int num = 0; |
5dc562c5 JB |
4955 | |
4956 | INIT_LIST_HEAD(&extents); | |
4957 | ||
5dc562c5 | 4958 | write_lock(&tree->lock); |
5dc562c5 JB |
4959 | |
4960 | list_for_each_entry_safe(em, n, &tree->modified_extents, list) { | |
4961 | list_del_init(&em->list); | |
2ab28f32 JB |
4962 | /* |
4963 | * Just an arbitrary number, this can be really CPU intensive | |
4964 | * once we start getting a lot of extents, and really once we | |
4965 | * have a bunch of extents we just want to commit since it will | |
4966 | * be faster. | |
4967 | */ | |
4968 | if (++num > 32768) { | |
4969 | list_del_init(&tree->modified_extents); | |
4970 | ret = -EFBIG; | |
4971 | goto process; | |
4972 | } | |
4973 | ||
5f96bfb7 | 4974 | if (em->generation < trans->transid) |
5dc562c5 | 4975 | continue; |
8c6c5928 | 4976 | |
31d11b83 FM |
4977 | /* We log prealloc extents beyond eof later. */ |
4978 | if (test_bit(EXTENT_FLAG_PREALLOC, &em->flags) && | |
4979 | em->start >= i_size_read(&inode->vfs_inode)) | |
4980 | continue; | |
4981 | ||
ff44c6e3 | 4982 | /* Need a ref to keep it from getting evicted from cache */ |
490b54d6 | 4983 | refcount_inc(&em->refs); |
ff44c6e3 | 4984 | set_bit(EXTENT_FLAG_LOGGING, &em->flags); |
5dc562c5 | 4985 | list_add_tail(&em->list, &extents); |
2ab28f32 | 4986 | num++; |
5dc562c5 JB |
4987 | } |
4988 | ||
4989 | list_sort(NULL, &extents, extent_cmp); | |
2ab28f32 | 4990 | process: |
5dc562c5 JB |
4991 | while (!list_empty(&extents)) { |
4992 | em = list_entry(extents.next, struct extent_map, list); | |
4993 | ||
4994 | list_del_init(&em->list); | |
4995 | ||
4996 | /* | |
4997 | * If we had an error we just need to delete everybody from our | |
4998 | * private list. | |
4999 | */ | |
ff44c6e3 | 5000 | if (ret) { |
201a9038 | 5001 | clear_em_logging(tree, em); |
ff44c6e3 | 5002 | free_extent_map(em); |
5dc562c5 | 5003 | continue; |
ff44c6e3 JB |
5004 | } |
5005 | ||
5006 | write_unlock(&tree->lock); | |
5dc562c5 | 5007 | |
90d04510 | 5008 | ret = log_one_extent(trans, inode, em, path, ctx); |
ff44c6e3 | 5009 | write_lock(&tree->lock); |
201a9038 JB |
5010 | clear_em_logging(tree, em); |
5011 | free_extent_map(em); | |
5dc562c5 | 5012 | } |
ff44c6e3 JB |
5013 | WARN_ON(!list_empty(&extents)); |
5014 | write_unlock(&tree->lock); | |
5dc562c5 | 5015 | |
31d11b83 FM |
5016 | if (!ret) |
5017 | ret = btrfs_log_prealloc_extents(trans, inode, path); | |
48778179 FM |
5018 | if (ret) |
5019 | return ret; | |
31d11b83 | 5020 | |
48778179 FM |
5021 | /* |
5022 | * We have logged all extents successfully, now make sure the commit of | |
5023 | * the current transaction waits for the ordered extents to complete | |
5024 | * before it commits and wipes out the log trees, otherwise we would | |
5025 | * lose data if an ordered extents completes after the transaction | |
5026 | * commits and a power failure happens after the transaction commit. | |
5027 | */ | |
5028 | list_for_each_entry_safe(ordered, tmp, &ctx->ordered_extents, log_list) { | |
5029 | list_del_init(&ordered->log_list); | |
5030 | set_bit(BTRFS_ORDERED_LOGGED, &ordered->flags); | |
5031 | ||
5032 | if (!test_bit(BTRFS_ORDERED_COMPLETE, &ordered->flags)) { | |
5033 | spin_lock_irq(&inode->ordered_tree.lock); | |
5034 | if (!test_bit(BTRFS_ORDERED_COMPLETE, &ordered->flags)) { | |
5035 | set_bit(BTRFS_ORDERED_PENDING, &ordered->flags); | |
5036 | atomic_inc(&trans->transaction->pending_ordered); | |
5037 | } | |
5038 | spin_unlock_irq(&inode->ordered_tree.lock); | |
5039 | } | |
5040 | btrfs_put_ordered_extent(ordered); | |
5041 | } | |
5042 | ||
5043 | return 0; | |
5dc562c5 JB |
5044 | } |
5045 | ||
481b01c0 | 5046 | static int logged_inode_size(struct btrfs_root *log, struct btrfs_inode *inode, |
1a4bcf47 FM |
5047 | struct btrfs_path *path, u64 *size_ret) |
5048 | { | |
5049 | struct btrfs_key key; | |
5050 | int ret; | |
5051 | ||
481b01c0 | 5052 | key.objectid = btrfs_ino(inode); |
1a4bcf47 FM |
5053 | key.type = BTRFS_INODE_ITEM_KEY; |
5054 | key.offset = 0; | |
5055 | ||
5056 | ret = btrfs_search_slot(NULL, log, &key, path, 0, 0); | |
5057 | if (ret < 0) { | |
5058 | return ret; | |
5059 | } else if (ret > 0) { | |
2f2ff0ee | 5060 | *size_ret = 0; |
1a4bcf47 FM |
5061 | } else { |
5062 | struct btrfs_inode_item *item; | |
5063 | ||
5064 | item = btrfs_item_ptr(path->nodes[0], path->slots[0], | |
5065 | struct btrfs_inode_item); | |
5066 | *size_ret = btrfs_inode_size(path->nodes[0], item); | |
bf504110 FM |
5067 | /* |
5068 | * If the in-memory inode's i_size is smaller then the inode | |
5069 | * size stored in the btree, return the inode's i_size, so | |
5070 | * that we get a correct inode size after replaying the log | |
5071 | * when before a power failure we had a shrinking truncate | |
5072 | * followed by addition of a new name (rename / new hard link). | |
5073 | * Otherwise return the inode size from the btree, to avoid | |
5074 | * data loss when replaying a log due to previously doing a | |
5075 | * write that expands the inode's size and logging a new name | |
5076 | * immediately after. | |
5077 | */ | |
5078 | if (*size_ret > inode->vfs_inode.i_size) | |
5079 | *size_ret = inode->vfs_inode.i_size; | |
1a4bcf47 FM |
5080 | } |
5081 | ||
5082 | btrfs_release_path(path); | |
5083 | return 0; | |
5084 | } | |
5085 | ||
36283bf7 FM |
5086 | /* |
5087 | * At the moment we always log all xattrs. This is to figure out at log replay | |
5088 | * time which xattrs must have their deletion replayed. If a xattr is missing | |
5089 | * in the log tree and exists in the fs/subvol tree, we delete it. This is | |
5090 | * because if a xattr is deleted, the inode is fsynced and a power failure | |
5091 | * happens, causing the log to be replayed the next time the fs is mounted, | |
5092 | * we want the xattr to not exist anymore (same behaviour as other filesystems | |
5093 | * with a journal, ext3/4, xfs, f2fs, etc). | |
5094 | */ | |
5095 | static int btrfs_log_all_xattrs(struct btrfs_trans_handle *trans, | |
1a93c36a | 5096 | struct btrfs_inode *inode, |
36283bf7 FM |
5097 | struct btrfs_path *path, |
5098 | struct btrfs_path *dst_path) | |
5099 | { | |
90d04510 | 5100 | struct btrfs_root *root = inode->root; |
36283bf7 FM |
5101 | int ret; |
5102 | struct btrfs_key key; | |
1a93c36a | 5103 | const u64 ino = btrfs_ino(inode); |
36283bf7 FM |
5104 | int ins_nr = 0; |
5105 | int start_slot = 0; | |
f2f121ab FM |
5106 | bool found_xattrs = false; |
5107 | ||
5108 | if (test_bit(BTRFS_INODE_NO_XATTRS, &inode->runtime_flags)) | |
5109 | return 0; | |
36283bf7 FM |
5110 | |
5111 | key.objectid = ino; | |
5112 | key.type = BTRFS_XATTR_ITEM_KEY; | |
5113 | key.offset = 0; | |
5114 | ||
5115 | ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); | |
5116 | if (ret < 0) | |
5117 | return ret; | |
5118 | ||
5119 | while (true) { | |
5120 | int slot = path->slots[0]; | |
5121 | struct extent_buffer *leaf = path->nodes[0]; | |
5122 | int nritems = btrfs_header_nritems(leaf); | |
5123 | ||
5124 | if (slot >= nritems) { | |
5125 | if (ins_nr > 0) { | |
1a93c36a | 5126 | ret = copy_items(trans, inode, dst_path, path, |
0e56315c | 5127 | start_slot, ins_nr, 1, 0); |
36283bf7 FM |
5128 | if (ret < 0) |
5129 | return ret; | |
5130 | ins_nr = 0; | |
5131 | } | |
5132 | ret = btrfs_next_leaf(root, path); | |
5133 | if (ret < 0) | |
5134 | return ret; | |
5135 | else if (ret > 0) | |
5136 | break; | |
5137 | continue; | |
5138 | } | |
5139 | ||
5140 | btrfs_item_key_to_cpu(leaf, &key, slot); | |
5141 | if (key.objectid != ino || key.type != BTRFS_XATTR_ITEM_KEY) | |
5142 | break; | |
5143 | ||
5144 | if (ins_nr == 0) | |
5145 | start_slot = slot; | |
5146 | ins_nr++; | |
5147 | path->slots[0]++; | |
f2f121ab | 5148 | found_xattrs = true; |
36283bf7 FM |
5149 | cond_resched(); |
5150 | } | |
5151 | if (ins_nr > 0) { | |
1a93c36a | 5152 | ret = copy_items(trans, inode, dst_path, path, |
0e56315c | 5153 | start_slot, ins_nr, 1, 0); |
36283bf7 FM |
5154 | if (ret < 0) |
5155 | return ret; | |
5156 | } | |
5157 | ||
f2f121ab FM |
5158 | if (!found_xattrs) |
5159 | set_bit(BTRFS_INODE_NO_XATTRS, &inode->runtime_flags); | |
5160 | ||
36283bf7 FM |
5161 | return 0; |
5162 | } | |
5163 | ||
a89ca6f2 | 5164 | /* |
0e56315c FM |
5165 | * When using the NO_HOLES feature if we punched a hole that causes the |
5166 | * deletion of entire leafs or all the extent items of the first leaf (the one | |
5167 | * that contains the inode item and references) we may end up not processing | |
5168 | * any extents, because there are no leafs with a generation matching the | |
5169 | * current transaction that have extent items for our inode. So we need to find | |
5170 | * if any holes exist and then log them. We also need to log holes after any | |
5171 | * truncate operation that changes the inode's size. | |
a89ca6f2 | 5172 | */ |
0e56315c | 5173 | static int btrfs_log_holes(struct btrfs_trans_handle *trans, |
0e56315c | 5174 | struct btrfs_inode *inode, |
7af59743 | 5175 | struct btrfs_path *path) |
a89ca6f2 | 5176 | { |
90d04510 | 5177 | struct btrfs_root *root = inode->root; |
0b246afa | 5178 | struct btrfs_fs_info *fs_info = root->fs_info; |
a89ca6f2 | 5179 | struct btrfs_key key; |
a0308dd7 NB |
5180 | const u64 ino = btrfs_ino(inode); |
5181 | const u64 i_size = i_size_read(&inode->vfs_inode); | |
7af59743 | 5182 | u64 prev_extent_end = 0; |
0e56315c | 5183 | int ret; |
a89ca6f2 | 5184 | |
0e56315c | 5185 | if (!btrfs_fs_incompat(fs_info, NO_HOLES) || i_size == 0) |
a89ca6f2 FM |
5186 | return 0; |
5187 | ||
5188 | key.objectid = ino; | |
5189 | key.type = BTRFS_EXTENT_DATA_KEY; | |
7af59743 | 5190 | key.offset = 0; |
a89ca6f2 FM |
5191 | |
5192 | ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); | |
a89ca6f2 FM |
5193 | if (ret < 0) |
5194 | return ret; | |
5195 | ||
0e56315c | 5196 | while (true) { |
0e56315c | 5197 | struct extent_buffer *leaf = path->nodes[0]; |
a89ca6f2 | 5198 | |
0e56315c FM |
5199 | if (path->slots[0] >= btrfs_header_nritems(path->nodes[0])) { |
5200 | ret = btrfs_next_leaf(root, path); | |
5201 | if (ret < 0) | |
5202 | return ret; | |
5203 | if (ret > 0) { | |
5204 | ret = 0; | |
5205 | break; | |
5206 | } | |
5207 | leaf = path->nodes[0]; | |
5208 | } | |
5209 | ||
5210 | btrfs_item_key_to_cpu(leaf, &key, path->slots[0]); | |
5211 | if (key.objectid != ino || key.type != BTRFS_EXTENT_DATA_KEY) | |
5212 | break; | |
5213 | ||
5214 | /* We have a hole, log it. */ | |
5215 | if (prev_extent_end < key.offset) { | |
7af59743 | 5216 | const u64 hole_len = key.offset - prev_extent_end; |
0e56315c FM |
5217 | |
5218 | /* | |
5219 | * Release the path to avoid deadlocks with other code | |
5220 | * paths that search the root while holding locks on | |
5221 | * leafs from the log root. | |
5222 | */ | |
5223 | btrfs_release_path(path); | |
d1f68ba0 OS |
5224 | ret = btrfs_insert_hole_extent(trans, root->log_root, |
5225 | ino, prev_extent_end, | |
5226 | hole_len); | |
0e56315c FM |
5227 | if (ret < 0) |
5228 | return ret; | |
5229 | ||
5230 | /* | |
5231 | * Search for the same key again in the root. Since it's | |
5232 | * an extent item and we are holding the inode lock, the | |
5233 | * key must still exist. If it doesn't just emit warning | |
5234 | * and return an error to fall back to a transaction | |
5235 | * commit. | |
5236 | */ | |
5237 | ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); | |
5238 | if (ret < 0) | |
5239 | return ret; | |
5240 | if (WARN_ON(ret > 0)) | |
5241 | return -ENOENT; | |
5242 | leaf = path->nodes[0]; | |
5243 | } | |
a89ca6f2 | 5244 | |
7af59743 | 5245 | prev_extent_end = btrfs_file_extent_end(path); |
0e56315c FM |
5246 | path->slots[0]++; |
5247 | cond_resched(); | |
a89ca6f2 | 5248 | } |
a89ca6f2 | 5249 | |
7af59743 | 5250 | if (prev_extent_end < i_size) { |
0e56315c | 5251 | u64 hole_len; |
a89ca6f2 | 5252 | |
0e56315c | 5253 | btrfs_release_path(path); |
7af59743 | 5254 | hole_len = ALIGN(i_size - prev_extent_end, fs_info->sectorsize); |
d1f68ba0 OS |
5255 | ret = btrfs_insert_hole_extent(trans, root->log_root, ino, |
5256 | prev_extent_end, hole_len); | |
0e56315c FM |
5257 | if (ret < 0) |
5258 | return ret; | |
5259 | } | |
5260 | ||
5261 | return 0; | |
a89ca6f2 FM |
5262 | } |
5263 | ||
56f23fdb FM |
5264 | /* |
5265 | * When we are logging a new inode X, check if it doesn't have a reference that | |
5266 | * matches the reference from some other inode Y created in a past transaction | |
5267 | * and that was renamed in the current transaction. If we don't do this, then at | |
5268 | * log replay time we can lose inode Y (and all its files if it's a directory): | |
5269 | * | |
5270 | * mkdir /mnt/x | |
5271 | * echo "hello world" > /mnt/x/foobar | |
5272 | * sync | |
5273 | * mv /mnt/x /mnt/y | |
5274 | * mkdir /mnt/x # or touch /mnt/x | |
5275 | * xfs_io -c fsync /mnt/x | |
5276 | * <power fail> | |
5277 | * mount fs, trigger log replay | |
5278 | * | |
5279 | * After the log replay procedure, we would lose the first directory and all its | |
5280 | * files (file foobar). | |
5281 | * For the case where inode Y is not a directory we simply end up losing it: | |
5282 | * | |
5283 | * echo "123" > /mnt/foo | |
5284 | * sync | |
5285 | * mv /mnt/foo /mnt/bar | |
5286 | * echo "abc" > /mnt/foo | |
5287 | * xfs_io -c fsync /mnt/foo | |
5288 | * <power fail> | |
5289 | * | |
5290 | * We also need this for cases where a snapshot entry is replaced by some other | |
5291 | * entry (file or directory) otherwise we end up with an unreplayable log due to | |
5292 | * attempts to delete the snapshot entry (entry of type BTRFS_ROOT_ITEM_KEY) as | |
5293 | * if it were a regular entry: | |
5294 | * | |
5295 | * mkdir /mnt/x | |
5296 | * btrfs subvolume snapshot /mnt /mnt/x/snap | |
5297 | * btrfs subvolume delete /mnt/x/snap | |
5298 | * rmdir /mnt/x | |
5299 | * mkdir /mnt/x | |
5300 | * fsync /mnt/x or fsync some new file inside it | |
5301 | * <power fail> | |
5302 | * | |
5303 | * The snapshot delete, rmdir of x, mkdir of a new x and the fsync all happen in | |
5304 | * the same transaction. | |
5305 | */ | |
5306 | static int btrfs_check_ref_name_override(struct extent_buffer *eb, | |
5307 | const int slot, | |
5308 | const struct btrfs_key *key, | |
4791c8f1 | 5309 | struct btrfs_inode *inode, |
a3baaf0d | 5310 | u64 *other_ino, u64 *other_parent) |
56f23fdb FM |
5311 | { |
5312 | int ret; | |
5313 | struct btrfs_path *search_path; | |
5314 | char *name = NULL; | |
5315 | u32 name_len = 0; | |
3212fa14 | 5316 | u32 item_size = btrfs_item_size(eb, slot); |
56f23fdb FM |
5317 | u32 cur_offset = 0; |
5318 | unsigned long ptr = btrfs_item_ptr_offset(eb, slot); | |
5319 | ||
5320 | search_path = btrfs_alloc_path(); | |
5321 | if (!search_path) | |
5322 | return -ENOMEM; | |
5323 | search_path->search_commit_root = 1; | |
5324 | search_path->skip_locking = 1; | |
5325 | ||
5326 | while (cur_offset < item_size) { | |
5327 | u64 parent; | |
5328 | u32 this_name_len; | |
5329 | u32 this_len; | |
5330 | unsigned long name_ptr; | |
5331 | struct btrfs_dir_item *di; | |
5332 | ||
5333 | if (key->type == BTRFS_INODE_REF_KEY) { | |
5334 | struct btrfs_inode_ref *iref; | |
5335 | ||
5336 | iref = (struct btrfs_inode_ref *)(ptr + cur_offset); | |
5337 | parent = key->offset; | |
5338 | this_name_len = btrfs_inode_ref_name_len(eb, iref); | |
5339 | name_ptr = (unsigned long)(iref + 1); | |
5340 | this_len = sizeof(*iref) + this_name_len; | |
5341 | } else { | |
5342 | struct btrfs_inode_extref *extref; | |
5343 | ||
5344 | extref = (struct btrfs_inode_extref *)(ptr + | |
5345 | cur_offset); | |
5346 | parent = btrfs_inode_extref_parent(eb, extref); | |
5347 | this_name_len = btrfs_inode_extref_name_len(eb, extref); | |
5348 | name_ptr = (unsigned long)&extref->name; | |
5349 | this_len = sizeof(*extref) + this_name_len; | |
5350 | } | |
5351 | ||
5352 | if (this_name_len > name_len) { | |
5353 | char *new_name; | |
5354 | ||
5355 | new_name = krealloc(name, this_name_len, GFP_NOFS); | |
5356 | if (!new_name) { | |
5357 | ret = -ENOMEM; | |
5358 | goto out; | |
5359 | } | |
5360 | name_len = this_name_len; | |
5361 | name = new_name; | |
5362 | } | |
5363 | ||
5364 | read_extent_buffer(eb, name, name_ptr, this_name_len); | |
4791c8f1 NB |
5365 | di = btrfs_lookup_dir_item(NULL, inode->root, search_path, |
5366 | parent, name, this_name_len, 0); | |
56f23fdb | 5367 | if (di && !IS_ERR(di)) { |
44f714da FM |
5368 | struct btrfs_key di_key; |
5369 | ||
5370 | btrfs_dir_item_key_to_cpu(search_path->nodes[0], | |
5371 | di, &di_key); | |
5372 | if (di_key.type == BTRFS_INODE_ITEM_KEY) { | |
6b5fc433 FM |
5373 | if (di_key.objectid != key->objectid) { |
5374 | ret = 1; | |
5375 | *other_ino = di_key.objectid; | |
a3baaf0d | 5376 | *other_parent = parent; |
6b5fc433 FM |
5377 | } else { |
5378 | ret = 0; | |
5379 | } | |
44f714da FM |
5380 | } else { |
5381 | ret = -EAGAIN; | |
5382 | } | |
56f23fdb FM |
5383 | goto out; |
5384 | } else if (IS_ERR(di)) { | |
5385 | ret = PTR_ERR(di); | |
5386 | goto out; | |
5387 | } | |
5388 | btrfs_release_path(search_path); | |
5389 | ||
5390 | cur_offset += this_len; | |
5391 | } | |
5392 | ret = 0; | |
5393 | out: | |
5394 | btrfs_free_path(search_path); | |
5395 | kfree(name); | |
5396 | return ret; | |
5397 | } | |
5398 | ||
6b5fc433 FM |
5399 | struct btrfs_ino_list { |
5400 | u64 ino; | |
a3baaf0d | 5401 | u64 parent; |
6b5fc433 FM |
5402 | struct list_head list; |
5403 | }; | |
5404 | ||
5405 | static int log_conflicting_inodes(struct btrfs_trans_handle *trans, | |
5406 | struct btrfs_root *root, | |
5407 | struct btrfs_path *path, | |
5408 | struct btrfs_log_ctx *ctx, | |
a3baaf0d | 5409 | u64 ino, u64 parent) |
6b5fc433 FM |
5410 | { |
5411 | struct btrfs_ino_list *ino_elem; | |
5412 | LIST_HEAD(inode_list); | |
5413 | int ret = 0; | |
5414 | ||
5415 | ino_elem = kmalloc(sizeof(*ino_elem), GFP_NOFS); | |
5416 | if (!ino_elem) | |
5417 | return -ENOMEM; | |
5418 | ino_elem->ino = ino; | |
a3baaf0d | 5419 | ino_elem->parent = parent; |
6b5fc433 FM |
5420 | list_add_tail(&ino_elem->list, &inode_list); |
5421 | ||
5422 | while (!list_empty(&inode_list)) { | |
5423 | struct btrfs_fs_info *fs_info = root->fs_info; | |
5424 | struct btrfs_key key; | |
5425 | struct inode *inode; | |
5426 | ||
5427 | ino_elem = list_first_entry(&inode_list, struct btrfs_ino_list, | |
5428 | list); | |
5429 | ino = ino_elem->ino; | |
a3baaf0d | 5430 | parent = ino_elem->parent; |
6b5fc433 FM |
5431 | list_del(&ino_elem->list); |
5432 | kfree(ino_elem); | |
5433 | if (ret) | |
5434 | continue; | |
5435 | ||
5436 | btrfs_release_path(path); | |
5437 | ||
0202e83f | 5438 | inode = btrfs_iget(fs_info->sb, ino, root); |
6b5fc433 FM |
5439 | /* |
5440 | * If the other inode that had a conflicting dir entry was | |
a3baaf0d FM |
5441 | * deleted in the current transaction, we need to log its parent |
5442 | * directory. | |
6b5fc433 FM |
5443 | */ |
5444 | if (IS_ERR(inode)) { | |
5445 | ret = PTR_ERR(inode); | |
a3baaf0d | 5446 | if (ret == -ENOENT) { |
0202e83f | 5447 | inode = btrfs_iget(fs_info->sb, parent, root); |
a3baaf0d FM |
5448 | if (IS_ERR(inode)) { |
5449 | ret = PTR_ERR(inode); | |
5450 | } else { | |
90d04510 | 5451 | ret = btrfs_log_inode(trans, |
a3baaf0d FM |
5452 | BTRFS_I(inode), |
5453 | LOG_OTHER_INODE_ALL, | |
48778179 | 5454 | ctx); |
410f954c | 5455 | btrfs_add_delayed_iput(inode); |
a3baaf0d FM |
5456 | } |
5457 | } | |
6b5fc433 FM |
5458 | continue; |
5459 | } | |
b5e4ff9d FM |
5460 | /* |
5461 | * If the inode was already logged skip it - otherwise we can | |
5462 | * hit an infinite loop. Example: | |
5463 | * | |
5464 | * From the commit root (previous transaction) we have the | |
5465 | * following inodes: | |
5466 | * | |
5467 | * inode 257 a directory | |
5468 | * inode 258 with references "zz" and "zz_link" on inode 257 | |
5469 | * inode 259 with reference "a" on inode 257 | |
5470 | * | |
5471 | * And in the current (uncommitted) transaction we have: | |
5472 | * | |
5473 | * inode 257 a directory, unchanged | |
5474 | * inode 258 with references "a" and "a2" on inode 257 | |
5475 | * inode 259 with reference "zz_link" on inode 257 | |
5476 | * inode 261 with reference "zz" on inode 257 | |
5477 | * | |
5478 | * When logging inode 261 the following infinite loop could | |
5479 | * happen if we don't skip already logged inodes: | |
5480 | * | |
5481 | * - we detect inode 258 as a conflicting inode, with inode 261 | |
5482 | * on reference "zz", and log it; | |
5483 | * | |
5484 | * - we detect inode 259 as a conflicting inode, with inode 258 | |
5485 | * on reference "a", and log it; | |
5486 | * | |
5487 | * - we detect inode 258 as a conflicting inode, with inode 259 | |
5488 | * on reference "zz_link", and log it - again! After this we | |
5489 | * repeat the above steps forever. | |
5490 | */ | |
5491 | spin_lock(&BTRFS_I(inode)->lock); | |
5492 | /* | |
5493 | * Check the inode's logged_trans only instead of | |
5494 | * btrfs_inode_in_log(). This is because the last_log_commit of | |
1f295373 FM |
5495 | * the inode is not updated when we only log that it exists (see |
5496 | * btrfs_log_inode()). | |
b5e4ff9d FM |
5497 | */ |
5498 | if (BTRFS_I(inode)->logged_trans == trans->transid) { | |
5499 | spin_unlock(&BTRFS_I(inode)->lock); | |
5500 | btrfs_add_delayed_iput(inode); | |
5501 | continue; | |
5502 | } | |
5503 | spin_unlock(&BTRFS_I(inode)->lock); | |
6b5fc433 FM |
5504 | /* |
5505 | * We are safe logging the other inode without acquiring its | |
5506 | * lock as long as we log with the LOG_INODE_EXISTS mode. We | |
5507 | * are safe against concurrent renames of the other inode as | |
5508 | * well because during a rename we pin the log and update the | |
5509 | * log with the new name before we unpin it. | |
5510 | */ | |
90d04510 | 5511 | ret = btrfs_log_inode(trans, BTRFS_I(inode), LOG_OTHER_INODE, ctx); |
6b5fc433 | 5512 | if (ret) { |
410f954c | 5513 | btrfs_add_delayed_iput(inode); |
6b5fc433 FM |
5514 | continue; |
5515 | } | |
5516 | ||
5517 | key.objectid = ino; | |
5518 | key.type = BTRFS_INODE_REF_KEY; | |
5519 | key.offset = 0; | |
5520 | ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); | |
5521 | if (ret < 0) { | |
410f954c | 5522 | btrfs_add_delayed_iput(inode); |
6b5fc433 FM |
5523 | continue; |
5524 | } | |
5525 | ||
5526 | while (true) { | |
5527 | struct extent_buffer *leaf = path->nodes[0]; | |
5528 | int slot = path->slots[0]; | |
5529 | u64 other_ino = 0; | |
a3baaf0d | 5530 | u64 other_parent = 0; |
6b5fc433 FM |
5531 | |
5532 | if (slot >= btrfs_header_nritems(leaf)) { | |
5533 | ret = btrfs_next_leaf(root, path); | |
5534 | if (ret < 0) { | |
5535 | break; | |
5536 | } else if (ret > 0) { | |
5537 | ret = 0; | |
5538 | break; | |
5539 | } | |
5540 | continue; | |
5541 | } | |
5542 | ||
5543 | btrfs_item_key_to_cpu(leaf, &key, slot); | |
5544 | if (key.objectid != ino || | |
5545 | (key.type != BTRFS_INODE_REF_KEY && | |
5546 | key.type != BTRFS_INODE_EXTREF_KEY)) { | |
5547 | ret = 0; | |
5548 | break; | |
5549 | } | |
5550 | ||
5551 | ret = btrfs_check_ref_name_override(leaf, slot, &key, | |
a3baaf0d FM |
5552 | BTRFS_I(inode), &other_ino, |
5553 | &other_parent); | |
6b5fc433 FM |
5554 | if (ret < 0) |
5555 | break; | |
5556 | if (ret > 0) { | |
5557 | ino_elem = kmalloc(sizeof(*ino_elem), GFP_NOFS); | |
5558 | if (!ino_elem) { | |
5559 | ret = -ENOMEM; | |
5560 | break; | |
5561 | } | |
5562 | ino_elem->ino = other_ino; | |
a3baaf0d | 5563 | ino_elem->parent = other_parent; |
6b5fc433 FM |
5564 | list_add_tail(&ino_elem->list, &inode_list); |
5565 | ret = 0; | |
5566 | } | |
5567 | path->slots[0]++; | |
5568 | } | |
410f954c | 5569 | btrfs_add_delayed_iput(inode); |
6b5fc433 FM |
5570 | } |
5571 | ||
5572 | return ret; | |
5573 | } | |
5574 | ||
da447009 FM |
5575 | static int copy_inode_items_to_log(struct btrfs_trans_handle *trans, |
5576 | struct btrfs_inode *inode, | |
5577 | struct btrfs_key *min_key, | |
5578 | const struct btrfs_key *max_key, | |
5579 | struct btrfs_path *path, | |
5580 | struct btrfs_path *dst_path, | |
5581 | const u64 logged_isize, | |
5582 | const bool recursive_logging, | |
5583 | const int inode_only, | |
5584 | struct btrfs_log_ctx *ctx, | |
5585 | bool *need_log_inode_item) | |
5586 | { | |
d9947887 | 5587 | const u64 i_size = i_size_read(&inode->vfs_inode); |
da447009 FM |
5588 | struct btrfs_root *root = inode->root; |
5589 | int ins_start_slot = 0; | |
5590 | int ins_nr = 0; | |
5591 | int ret; | |
5592 | ||
5593 | while (1) { | |
5594 | ret = btrfs_search_forward(root, min_key, path, trans->transid); | |
5595 | if (ret < 0) | |
5596 | return ret; | |
5597 | if (ret > 0) { | |
5598 | ret = 0; | |
5599 | break; | |
5600 | } | |
5601 | again: | |
5602 | /* Note, ins_nr might be > 0 here, cleanup outside the loop */ | |
5603 | if (min_key->objectid != max_key->objectid) | |
5604 | break; | |
5605 | if (min_key->type > max_key->type) | |
5606 | break; | |
5607 | ||
d9947887 | 5608 | if (min_key->type == BTRFS_INODE_ITEM_KEY) { |
da447009 | 5609 | *need_log_inode_item = false; |
d9947887 FM |
5610 | } else if (min_key->type == BTRFS_EXTENT_DATA_KEY && |
5611 | min_key->offset >= i_size) { | |
5612 | /* | |
5613 | * Extents at and beyond eof are logged with | |
5614 | * btrfs_log_prealloc_extents(). | |
5615 | * Only regular files have BTRFS_EXTENT_DATA_KEY keys, | |
5616 | * and no keys greater than that, so bail out. | |
5617 | */ | |
5618 | break; | |
5619 | } else if ((min_key->type == BTRFS_INODE_REF_KEY || | |
5620 | min_key->type == BTRFS_INODE_EXTREF_KEY) && | |
5621 | inode->generation == trans->transid && | |
5622 | !recursive_logging) { | |
da447009 FM |
5623 | u64 other_ino = 0; |
5624 | u64 other_parent = 0; | |
5625 | ||
5626 | ret = btrfs_check_ref_name_override(path->nodes[0], | |
5627 | path->slots[0], min_key, inode, | |
5628 | &other_ino, &other_parent); | |
5629 | if (ret < 0) { | |
5630 | return ret; | |
289cffcb | 5631 | } else if (ret > 0 && |
da447009 FM |
5632 | other_ino != btrfs_ino(BTRFS_I(ctx->inode))) { |
5633 | if (ins_nr > 0) { | |
5634 | ins_nr++; | |
5635 | } else { | |
5636 | ins_nr = 1; | |
5637 | ins_start_slot = path->slots[0]; | |
5638 | } | |
5639 | ret = copy_items(trans, inode, dst_path, path, | |
5640 | ins_start_slot, ins_nr, | |
5641 | inode_only, logged_isize); | |
5642 | if (ret < 0) | |
5643 | return ret; | |
5644 | ins_nr = 0; | |
5645 | ||
5646 | ret = log_conflicting_inodes(trans, root, path, | |
5647 | ctx, other_ino, other_parent); | |
5648 | if (ret) | |
5649 | return ret; | |
5650 | btrfs_release_path(path); | |
5651 | goto next_key; | |
5652 | } | |
d9947887 FM |
5653 | } else if (min_key->type == BTRFS_XATTR_ITEM_KEY) { |
5654 | /* Skip xattrs, logged later with btrfs_log_all_xattrs() */ | |
da447009 FM |
5655 | if (ins_nr == 0) |
5656 | goto next_slot; | |
5657 | ret = copy_items(trans, inode, dst_path, path, | |
5658 | ins_start_slot, | |
5659 | ins_nr, inode_only, logged_isize); | |
5660 | if (ret < 0) | |
5661 | return ret; | |
5662 | ins_nr = 0; | |
5663 | goto next_slot; | |
5664 | } | |
5665 | ||
5666 | if (ins_nr && ins_start_slot + ins_nr == path->slots[0]) { | |
5667 | ins_nr++; | |
5668 | goto next_slot; | |
5669 | } else if (!ins_nr) { | |
5670 | ins_start_slot = path->slots[0]; | |
5671 | ins_nr = 1; | |
5672 | goto next_slot; | |
5673 | } | |
5674 | ||
5675 | ret = copy_items(trans, inode, dst_path, path, ins_start_slot, | |
5676 | ins_nr, inode_only, logged_isize); | |
5677 | if (ret < 0) | |
5678 | return ret; | |
5679 | ins_nr = 1; | |
5680 | ins_start_slot = path->slots[0]; | |
5681 | next_slot: | |
5682 | path->slots[0]++; | |
5683 | if (path->slots[0] < btrfs_header_nritems(path->nodes[0])) { | |
5684 | btrfs_item_key_to_cpu(path->nodes[0], min_key, | |
5685 | path->slots[0]); | |
5686 | goto again; | |
5687 | } | |
5688 | if (ins_nr) { | |
5689 | ret = copy_items(trans, inode, dst_path, path, | |
5690 | ins_start_slot, ins_nr, inode_only, | |
5691 | logged_isize); | |
5692 | if (ret < 0) | |
5693 | return ret; | |
5694 | ins_nr = 0; | |
5695 | } | |
5696 | btrfs_release_path(path); | |
5697 | next_key: | |
5698 | if (min_key->offset < (u64)-1) { | |
5699 | min_key->offset++; | |
5700 | } else if (min_key->type < max_key->type) { | |
5701 | min_key->type++; | |
5702 | min_key->offset = 0; | |
5703 | } else { | |
5704 | break; | |
5705 | } | |
96acb375 FM |
5706 | |
5707 | /* | |
5708 | * We may process many leaves full of items for our inode, so | |
5709 | * avoid monopolizing a cpu for too long by rescheduling while | |
5710 | * not holding locks on any tree. | |
5711 | */ | |
5712 | cond_resched(); | |
da447009 | 5713 | } |
d9947887 | 5714 | if (ins_nr) { |
da447009 FM |
5715 | ret = copy_items(trans, inode, dst_path, path, ins_start_slot, |
5716 | ins_nr, inode_only, logged_isize); | |
d9947887 FM |
5717 | if (ret) |
5718 | return ret; | |
5719 | } | |
5720 | ||
5721 | if (inode_only == LOG_INODE_ALL && S_ISREG(inode->vfs_inode.i_mode)) { | |
5722 | /* | |
5723 | * Release the path because otherwise we might attempt to double | |
5724 | * lock the same leaf with btrfs_log_prealloc_extents() below. | |
5725 | */ | |
5726 | btrfs_release_path(path); | |
5727 | ret = btrfs_log_prealloc_extents(trans, inode, dst_path); | |
5728 | } | |
da447009 FM |
5729 | |
5730 | return ret; | |
5731 | } | |
5732 | ||
e02119d5 CM |
5733 | /* log a single inode in the tree log. |
5734 | * At least one parent directory for this inode must exist in the tree | |
5735 | * or be logged already. | |
5736 | * | |
5737 | * Any items from this inode changed by the current transaction are copied | |
5738 | * to the log tree. An extra reference is taken on any extents in this | |
5739 | * file, allowing us to avoid a whole pile of corner cases around logging | |
5740 | * blocks that have been removed from the tree. | |
5741 | * | |
5742 | * See LOG_INODE_ALL and related defines for a description of what inode_only | |
5743 | * does. | |
5744 | * | |
5745 | * This handles both files and directories. | |
5746 | */ | |
12fcfd22 | 5747 | static int btrfs_log_inode(struct btrfs_trans_handle *trans, |
90d04510 | 5748 | struct btrfs_inode *inode, |
49dae1bc | 5749 | int inode_only, |
8407f553 | 5750 | struct btrfs_log_ctx *ctx) |
e02119d5 CM |
5751 | { |
5752 | struct btrfs_path *path; | |
5753 | struct btrfs_path *dst_path; | |
5754 | struct btrfs_key min_key; | |
5755 | struct btrfs_key max_key; | |
90d04510 | 5756 | struct btrfs_root *log = inode->root->log_root; |
65faced5 | 5757 | int ret; |
5dc562c5 | 5758 | bool fast_search = false; |
a59108a7 NB |
5759 | u64 ino = btrfs_ino(inode); |
5760 | struct extent_map_tree *em_tree = &inode->extent_tree; | |
1a4bcf47 | 5761 | u64 logged_isize = 0; |
e4545de5 | 5762 | bool need_log_inode_item = true; |
9a8fca62 | 5763 | bool xattrs_logged = false; |
a3baaf0d | 5764 | bool recursive_logging = false; |
2ac691d8 | 5765 | bool inode_item_dropped = true; |
0f8ce498 | 5766 | const bool orig_logged_before = ctx->logged_before; |
e02119d5 | 5767 | |
e02119d5 | 5768 | path = btrfs_alloc_path(); |
5df67083 TI |
5769 | if (!path) |
5770 | return -ENOMEM; | |
e02119d5 | 5771 | dst_path = btrfs_alloc_path(); |
5df67083 TI |
5772 | if (!dst_path) { |
5773 | btrfs_free_path(path); | |
5774 | return -ENOMEM; | |
5775 | } | |
e02119d5 | 5776 | |
33345d01 | 5777 | min_key.objectid = ino; |
e02119d5 CM |
5778 | min_key.type = BTRFS_INODE_ITEM_KEY; |
5779 | min_key.offset = 0; | |
5780 | ||
33345d01 | 5781 | max_key.objectid = ino; |
12fcfd22 | 5782 | |
12fcfd22 | 5783 | |
5dc562c5 | 5784 | /* today the code can only do partial logging of directories */ |
a59108a7 | 5785 | if (S_ISDIR(inode->vfs_inode.i_mode) || |
5269b67e | 5786 | (!test_bit(BTRFS_INODE_NEEDS_FULL_SYNC, |
a59108a7 | 5787 | &inode->runtime_flags) && |
781feef7 | 5788 | inode_only >= LOG_INODE_EXISTS)) |
e02119d5 CM |
5789 | max_key.type = BTRFS_XATTR_ITEM_KEY; |
5790 | else | |
5791 | max_key.type = (u8)-1; | |
5792 | max_key.offset = (u64)-1; | |
5793 | ||
2c2c452b | 5794 | /* |
5aa7d1a7 FM |
5795 | * Only run delayed items if we are a directory. We want to make sure |
5796 | * all directory indexes hit the fs/subvolume tree so we can find them | |
5797 | * and figure out which index ranges have to be logged. | |
2c2c452b | 5798 | */ |
f6df27dd | 5799 | if (S_ISDIR(inode->vfs_inode.i_mode)) { |
65faced5 FM |
5800 | ret = btrfs_commit_inode_delayed_items(trans, inode); |
5801 | if (ret) | |
f6df27dd | 5802 | goto out; |
16cdcec7 MX |
5803 | } |
5804 | ||
a3baaf0d FM |
5805 | if (inode_only == LOG_OTHER_INODE || inode_only == LOG_OTHER_INODE_ALL) { |
5806 | recursive_logging = true; | |
5807 | if (inode_only == LOG_OTHER_INODE) | |
5808 | inode_only = LOG_INODE_EXISTS; | |
5809 | else | |
5810 | inode_only = LOG_INODE_ALL; | |
a59108a7 | 5811 | mutex_lock_nested(&inode->log_mutex, SINGLE_DEPTH_NESTING); |
781feef7 | 5812 | } else { |
a59108a7 | 5813 | mutex_lock(&inode->log_mutex); |
781feef7 | 5814 | } |
e02119d5 | 5815 | |
d0e64a98 FM |
5816 | /* |
5817 | * For symlinks, we must always log their content, which is stored in an | |
5818 | * inline extent, otherwise we could end up with an empty symlink after | |
5819 | * log replay, which is invalid on linux (symlink(2) returns -ENOENT if | |
5820 | * one attempts to create an empty symlink). | |
5821 | * We don't need to worry about flushing delalloc, because when we create | |
5822 | * the inline extent when the symlink is created (we never have delalloc | |
5823 | * for symlinks). | |
5824 | */ | |
5825 | if (S_ISLNK(inode->vfs_inode.i_mode)) | |
5826 | inode_only = LOG_INODE_ALL; | |
5827 | ||
0f8ce498 FM |
5828 | /* |
5829 | * Before logging the inode item, cache the value returned by | |
5830 | * inode_logged(), because after that we have the need to figure out if | |
5831 | * the inode was previously logged in this transaction. | |
5832 | */ | |
5833 | ret = inode_logged(trans, inode, path); | |
65faced5 | 5834 | if (ret < 0) |
0f8ce498 | 5835 | goto out_unlock; |
0f8ce498 | 5836 | ctx->logged_before = (ret == 1); |
65faced5 | 5837 | ret = 0; |
0f8ce498 | 5838 | |
64d6b281 FM |
5839 | /* |
5840 | * This is for cases where logging a directory could result in losing a | |
5841 | * a file after replaying the log. For example, if we move a file from a | |
5842 | * directory A to a directory B, then fsync directory A, we have no way | |
5843 | * to known the file was moved from A to B, so logging just A would | |
5844 | * result in losing the file after a log replay. | |
5845 | */ | |
5846 | if (S_ISDIR(inode->vfs_inode.i_mode) && | |
5847 | inode_only == LOG_INODE_ALL && | |
5848 | inode->last_unlink_trans >= trans->transid) { | |
5849 | btrfs_set_log_full_commit(trans); | |
f31f09f6 | 5850 | ret = BTRFS_LOG_FORCE_COMMIT; |
64d6b281 FM |
5851 | goto out_unlock; |
5852 | } | |
5853 | ||
e02119d5 CM |
5854 | /* |
5855 | * a brute force approach to making sure we get the most uptodate | |
5856 | * copies of everything. | |
5857 | */ | |
a59108a7 | 5858 | if (S_ISDIR(inode->vfs_inode.i_mode)) { |
e02119d5 CM |
5859 | int max_key_type = BTRFS_DIR_LOG_INDEX_KEY; |
5860 | ||
ab12313a | 5861 | clear_bit(BTRFS_INODE_COPY_EVERYTHING, &inode->runtime_flags); |
4f764e51 FM |
5862 | if (inode_only == LOG_INODE_EXISTS) |
5863 | max_key_type = BTRFS_XATTR_ITEM_KEY; | |
0f8ce498 FM |
5864 | if (ctx->logged_before) |
5865 | ret = drop_inode_items(trans, log, path, inode, | |
5866 | max_key_type); | |
e02119d5 | 5867 | } else { |
0f8ce498 | 5868 | if (inode_only == LOG_INODE_EXISTS && ctx->logged_before) { |
1a4bcf47 FM |
5869 | /* |
5870 | * Make sure the new inode item we write to the log has | |
5871 | * the same isize as the current one (if it exists). | |
5872 | * This is necessary to prevent data loss after log | |
5873 | * replay, and also to prevent doing a wrong expanding | |
5874 | * truncate - for e.g. create file, write 4K into offset | |
5875 | * 0, fsync, write 4K into offset 4096, add hard link, | |
5876 | * fsync some other file (to sync log), power fail - if | |
5877 | * we use the inode's current i_size, after log replay | |
5878 | * we get a 8Kb file, with the last 4Kb extent as a hole | |
5879 | * (zeroes), as if an expanding truncate happened, | |
5880 | * instead of getting a file of 4Kb only. | |
5881 | */ | |
65faced5 FM |
5882 | ret = logged_inode_size(log, inode, path, &logged_isize); |
5883 | if (ret) | |
1a4bcf47 FM |
5884 | goto out_unlock; |
5885 | } | |
a742994a | 5886 | if (test_bit(BTRFS_INODE_NEEDS_FULL_SYNC, |
a59108a7 | 5887 | &inode->runtime_flags)) { |
a742994a | 5888 | if (inode_only == LOG_INODE_EXISTS) { |
4f764e51 | 5889 | max_key.type = BTRFS_XATTR_ITEM_KEY; |
0f8ce498 FM |
5890 | if (ctx->logged_before) |
5891 | ret = drop_inode_items(trans, log, path, | |
5892 | inode, max_key.type); | |
a742994a FM |
5893 | } else { |
5894 | clear_bit(BTRFS_INODE_NEEDS_FULL_SYNC, | |
a59108a7 | 5895 | &inode->runtime_flags); |
a742994a | 5896 | clear_bit(BTRFS_INODE_COPY_EVERYTHING, |
a59108a7 | 5897 | &inode->runtime_flags); |
0f8ce498 | 5898 | if (ctx->logged_before) |
4934a815 FM |
5899 | ret = truncate_inode_items(trans, log, |
5900 | inode, 0, 0); | |
a742994a | 5901 | } |
4f764e51 | 5902 | } else if (test_and_clear_bit(BTRFS_INODE_COPY_EVERYTHING, |
a59108a7 | 5903 | &inode->runtime_flags) || |
6cfab851 | 5904 | inode_only == LOG_INODE_EXISTS) { |
4f764e51 | 5905 | if (inode_only == LOG_INODE_ALL) |
183f37fa | 5906 | fast_search = true; |
4f764e51 | 5907 | max_key.type = BTRFS_XATTR_ITEM_KEY; |
0f8ce498 FM |
5908 | if (ctx->logged_before) |
5909 | ret = drop_inode_items(trans, log, path, inode, | |
5910 | max_key.type); | |
a95249b3 JB |
5911 | } else { |
5912 | if (inode_only == LOG_INODE_ALL) | |
5913 | fast_search = true; | |
2ac691d8 | 5914 | inode_item_dropped = false; |
a95249b3 | 5915 | goto log_extents; |
5dc562c5 | 5916 | } |
a95249b3 | 5917 | |
e02119d5 | 5918 | } |
65faced5 | 5919 | if (ret) |
4a500fd1 | 5920 | goto out_unlock; |
e02119d5 | 5921 | |
65faced5 | 5922 | ret = copy_inode_items_to_log(trans, inode, &min_key, &max_key, |
da447009 | 5923 | path, dst_path, logged_isize, |
7af59743 FM |
5924 | recursive_logging, inode_only, ctx, |
5925 | &need_log_inode_item); | |
65faced5 | 5926 | if (ret) |
da447009 | 5927 | goto out_unlock; |
5dc562c5 | 5928 | |
36283bf7 FM |
5929 | btrfs_release_path(path); |
5930 | btrfs_release_path(dst_path); | |
65faced5 FM |
5931 | ret = btrfs_log_all_xattrs(trans, inode, path, dst_path); |
5932 | if (ret) | |
36283bf7 | 5933 | goto out_unlock; |
9a8fca62 | 5934 | xattrs_logged = true; |
a89ca6f2 FM |
5935 | if (max_key.type >= BTRFS_EXTENT_DATA_KEY && !fast_search) { |
5936 | btrfs_release_path(path); | |
5937 | btrfs_release_path(dst_path); | |
65faced5 FM |
5938 | ret = btrfs_log_holes(trans, inode, path); |
5939 | if (ret) | |
a89ca6f2 FM |
5940 | goto out_unlock; |
5941 | } | |
a95249b3 | 5942 | log_extents: |
f3b15ccd JB |
5943 | btrfs_release_path(path); |
5944 | btrfs_release_path(dst_path); | |
e4545de5 | 5945 | if (need_log_inode_item) { |
65faced5 FM |
5946 | ret = log_inode_item(trans, log, dst_path, inode, inode_item_dropped); |
5947 | if (ret) | |
b590b839 FM |
5948 | goto out_unlock; |
5949 | /* | |
5950 | * If we are doing a fast fsync and the inode was logged before | |
5951 | * in this transaction, we don't need to log the xattrs because | |
5952 | * they were logged before. If xattrs were added, changed or | |
5953 | * deleted since the last time we logged the inode, then we have | |
5954 | * already logged them because the inode had the runtime flag | |
5955 | * BTRFS_INODE_COPY_EVERYTHING set. | |
5956 | */ | |
5957 | if (!xattrs_logged && inode->logged_trans < trans->transid) { | |
65faced5 FM |
5958 | ret = btrfs_log_all_xattrs(trans, inode, path, dst_path); |
5959 | if (ret) | |
b590b839 | 5960 | goto out_unlock; |
9a8fca62 FM |
5961 | btrfs_release_path(path); |
5962 | } | |
e4545de5 | 5963 | } |
5dc562c5 | 5964 | if (fast_search) { |
90d04510 | 5965 | ret = btrfs_log_changed_extents(trans, inode, dst_path, ctx); |
65faced5 | 5966 | if (ret) |
5dc562c5 | 5967 | goto out_unlock; |
d006a048 | 5968 | } else if (inode_only == LOG_INODE_ALL) { |
06d3d22b LB |
5969 | struct extent_map *em, *n; |
5970 | ||
49dae1bc | 5971 | write_lock(&em_tree->lock); |
48778179 FM |
5972 | list_for_each_entry_safe(em, n, &em_tree->modified_extents, list) |
5973 | list_del_init(&em->list); | |
49dae1bc | 5974 | write_unlock(&em_tree->lock); |
5dc562c5 JB |
5975 | } |
5976 | ||
a59108a7 | 5977 | if (inode_only == LOG_INODE_ALL && S_ISDIR(inode->vfs_inode.i_mode)) { |
90d04510 | 5978 | ret = log_directory_changes(trans, inode, path, dst_path, ctx); |
65faced5 | 5979 | if (ret) |
4a500fd1 | 5980 | goto out_unlock; |
e02119d5 | 5981 | } |
49dae1bc | 5982 | |
130341be FM |
5983 | spin_lock(&inode->lock); |
5984 | inode->logged_trans = trans->transid; | |
d1d832a0 | 5985 | /* |
130341be FM |
5986 | * Don't update last_log_commit if we logged that an inode exists. |
5987 | * We do this for three reasons: | |
5988 | * | |
5989 | * 1) We might have had buffered writes to this inode that were | |
5990 | * flushed and had their ordered extents completed in this | |
5991 | * transaction, but we did not previously log the inode with | |
5992 | * LOG_INODE_ALL. Later the inode was evicted and after that | |
5993 | * it was loaded again and this LOG_INODE_EXISTS log operation | |
5994 | * happened. We must make sure that if an explicit fsync against | |
5995 | * the inode is performed later, it logs the new extents, an | |
5996 | * updated inode item, etc, and syncs the log. The same logic | |
5997 | * applies to direct IO writes instead of buffered writes. | |
5998 | * | |
5999 | * 2) When we log the inode with LOG_INODE_EXISTS, its inode item | |
6000 | * is logged with an i_size of 0 or whatever value was logged | |
6001 | * before. If later the i_size of the inode is increased by a | |
6002 | * truncate operation, the log is synced through an fsync of | |
6003 | * some other inode and then finally an explicit fsync against | |
6004 | * this inode is made, we must make sure this fsync logs the | |
6005 | * inode with the new i_size, the hole between old i_size and | |
6006 | * the new i_size, and syncs the log. | |
6007 | * | |
6008 | * 3) If we are logging that an ancestor inode exists as part of | |
6009 | * logging a new name from a link or rename operation, don't update | |
6010 | * its last_log_commit - otherwise if an explicit fsync is made | |
6011 | * against an ancestor, the fsync considers the inode in the log | |
6012 | * and doesn't sync the log, resulting in the ancestor missing after | |
6013 | * a power failure unless the log was synced as part of an fsync | |
6014 | * against any other unrelated inode. | |
d1d832a0 | 6015 | */ |
130341be FM |
6016 | if (inode_only != LOG_INODE_EXISTS) |
6017 | inode->last_log_commit = inode->last_sub_trans; | |
6018 | spin_unlock(&inode->lock); | |
23e3337f FM |
6019 | |
6020 | /* | |
6021 | * Reset the last_reflink_trans so that the next fsync does not need to | |
6022 | * go through the slower path when logging extents and their checksums. | |
6023 | */ | |
6024 | if (inode_only == LOG_INODE_ALL) | |
6025 | inode->last_reflink_trans = 0; | |
6026 | ||
4a500fd1 | 6027 | out_unlock: |
a59108a7 | 6028 | mutex_unlock(&inode->log_mutex); |
f6df27dd | 6029 | out: |
e02119d5 CM |
6030 | btrfs_free_path(path); |
6031 | btrfs_free_path(dst_path); | |
0f8ce498 FM |
6032 | |
6033 | if (recursive_logging) | |
6034 | ctx->logged_before = orig_logged_before; | |
6035 | ||
65faced5 | 6036 | return ret; |
e02119d5 CM |
6037 | } |
6038 | ||
ab12313a FM |
6039 | /* |
6040 | * Check if we need to log an inode. This is used in contexts where while | |
6041 | * logging an inode we need to log another inode (either that it exists or in | |
6042 | * full mode). This is used instead of btrfs_inode_in_log() because the later | |
6043 | * requires the inode to be in the log and have the log transaction committed, | |
6044 | * while here we do not care if the log transaction was already committed - our | |
6045 | * caller will commit the log later - and we want to avoid logging an inode | |
6046 | * multiple times when multiple tasks have joined the same log transaction. | |
6047 | */ | |
6048 | static bool need_log_inode(struct btrfs_trans_handle *trans, | |
6049 | struct btrfs_inode *inode) | |
6050 | { | |
8be2ba2e FM |
6051 | /* |
6052 | * If a directory was not modified, no dentries added or removed, we can | |
6053 | * and should avoid logging it. | |
6054 | */ | |
6055 | if (S_ISDIR(inode->vfs_inode.i_mode) && inode->last_trans < trans->transid) | |
6056 | return false; | |
6057 | ||
ab12313a FM |
6058 | /* |
6059 | * If this inode does not have new/updated/deleted xattrs since the last | |
6060 | * time it was logged and is flagged as logged in the current transaction, | |
6061 | * we can skip logging it. As for new/deleted names, those are updated in | |
6062 | * the log by link/unlink/rename operations. | |
6063 | * In case the inode was logged and then evicted and reloaded, its | |
6064 | * logged_trans will be 0, in which case we have to fully log it since | |
6065 | * logged_trans is a transient field, not persisted. | |
6066 | */ | |
6067 | if (inode->logged_trans == trans->transid && | |
6068 | !test_bit(BTRFS_INODE_COPY_EVERYTHING, &inode->runtime_flags)) | |
6069 | return false; | |
6070 | ||
6071 | return true; | |
6072 | } | |
6073 | ||
2f2ff0ee FM |
6074 | struct btrfs_dir_list { |
6075 | u64 ino; | |
6076 | struct list_head list; | |
6077 | }; | |
6078 | ||
6079 | /* | |
6080 | * Log the inodes of the new dentries of a directory. See log_dir_items() for | |
6081 | * details about the why it is needed. | |
6082 | * This is a recursive operation - if an existing dentry corresponds to a | |
6083 | * directory, that directory's new entries are logged too (same behaviour as | |
6084 | * ext3/4, xfs, f2fs, reiserfs, nilfs2). Note that when logging the inodes | |
6085 | * the dentries point to we do not lock their i_mutex, otherwise lockdep | |
6086 | * complains about the following circular lock dependency / possible deadlock: | |
6087 | * | |
6088 | * CPU0 CPU1 | |
6089 | * ---- ---- | |
6090 | * lock(&type->i_mutex_dir_key#3/2); | |
6091 | * lock(sb_internal#2); | |
6092 | * lock(&type->i_mutex_dir_key#3/2); | |
6093 | * lock(&sb->s_type->i_mutex_key#14); | |
6094 | * | |
6095 | * Where sb_internal is the lock (a counter that works as a lock) acquired by | |
6096 | * sb_start_intwrite() in btrfs_start_transaction(). | |
6097 | * Not locking i_mutex of the inodes is still safe because: | |
6098 | * | |
6099 | * 1) For regular files we log with a mode of LOG_INODE_EXISTS. It's possible | |
6100 | * that while logging the inode new references (names) are added or removed | |
6101 | * from the inode, leaving the logged inode item with a link count that does | |
6102 | * not match the number of logged inode reference items. This is fine because | |
6103 | * at log replay time we compute the real number of links and correct the | |
6104 | * link count in the inode item (see replay_one_buffer() and | |
6105 | * link_to_fixup_dir()); | |
6106 | * | |
6107 | * 2) For directories we log with a mode of LOG_INODE_ALL. It's possible that | |
339d0354 FM |
6108 | * while logging the inode's items new index items (key type |
6109 | * BTRFS_DIR_INDEX_KEY) are added to fs/subvol tree and the logged inode item | |
2f2ff0ee | 6110 | * has a size that doesn't match the sum of the lengths of all the logged |
339d0354 FM |
6111 | * names - this is ok, not a problem, because at log replay time we set the |
6112 | * directory's i_size to the correct value (see replay_one_name() and | |
6113 | * do_overwrite_item()). | |
2f2ff0ee FM |
6114 | */ |
6115 | static int log_new_dir_dentries(struct btrfs_trans_handle *trans, | |
6116 | struct btrfs_root *root, | |
51cc0d32 | 6117 | struct btrfs_inode *start_inode, |
2f2ff0ee FM |
6118 | struct btrfs_log_ctx *ctx) |
6119 | { | |
0b246afa | 6120 | struct btrfs_fs_info *fs_info = root->fs_info; |
2f2ff0ee FM |
6121 | struct btrfs_path *path; |
6122 | LIST_HEAD(dir_list); | |
6123 | struct btrfs_dir_list *dir_elem; | |
6124 | int ret = 0; | |
6125 | ||
c48792c6 FM |
6126 | /* |
6127 | * If we are logging a new name, as part of a link or rename operation, | |
6128 | * don't bother logging new dentries, as we just want to log the names | |
6129 | * of an inode and that any new parents exist. | |
6130 | */ | |
6131 | if (ctx->logging_new_name) | |
6132 | return 0; | |
6133 | ||
2f2ff0ee FM |
6134 | path = btrfs_alloc_path(); |
6135 | if (!path) | |
6136 | return -ENOMEM; | |
6137 | ||
6138 | dir_elem = kmalloc(sizeof(*dir_elem), GFP_NOFS); | |
6139 | if (!dir_elem) { | |
6140 | btrfs_free_path(path); | |
6141 | return -ENOMEM; | |
6142 | } | |
51cc0d32 | 6143 | dir_elem->ino = btrfs_ino(start_inode); |
2f2ff0ee FM |
6144 | list_add_tail(&dir_elem->list, &dir_list); |
6145 | ||
6146 | while (!list_empty(&dir_list)) { | |
6147 | struct extent_buffer *leaf; | |
6148 | struct btrfs_key min_key; | |
6149 | int nritems; | |
6150 | int i; | |
6151 | ||
6152 | dir_elem = list_first_entry(&dir_list, struct btrfs_dir_list, | |
6153 | list); | |
6154 | if (ret) | |
6155 | goto next_dir_inode; | |
6156 | ||
6157 | min_key.objectid = dir_elem->ino; | |
339d0354 | 6158 | min_key.type = BTRFS_DIR_INDEX_KEY; |
2f2ff0ee FM |
6159 | min_key.offset = 0; |
6160 | again: | |
6161 | btrfs_release_path(path); | |
732d591a | 6162 | ret = btrfs_search_forward(root, &min_key, path, trans->transid); |
2f2ff0ee FM |
6163 | if (ret < 0) { |
6164 | goto next_dir_inode; | |
6165 | } else if (ret > 0) { | |
6166 | ret = 0; | |
6167 | goto next_dir_inode; | |
6168 | } | |
6169 | ||
2f2ff0ee FM |
6170 | leaf = path->nodes[0]; |
6171 | nritems = btrfs_header_nritems(leaf); | |
6172 | for (i = path->slots[0]; i < nritems; i++) { | |
6173 | struct btrfs_dir_item *di; | |
6174 | struct btrfs_key di_key; | |
6175 | struct inode *di_inode; | |
6176 | struct btrfs_dir_list *new_dir_elem; | |
6177 | int log_mode = LOG_INODE_EXISTS; | |
6178 | int type; | |
6179 | ||
6180 | btrfs_item_key_to_cpu(leaf, &min_key, i); | |
6181 | if (min_key.objectid != dir_elem->ino || | |
339d0354 | 6182 | min_key.type != BTRFS_DIR_INDEX_KEY) |
2f2ff0ee FM |
6183 | goto next_dir_inode; |
6184 | ||
6185 | di = btrfs_item_ptr(leaf, i, struct btrfs_dir_item); | |
6186 | type = btrfs_dir_type(leaf, di); | |
de6bc7f5 | 6187 | if (btrfs_dir_transid(leaf, di) < trans->transid) |
2f2ff0ee FM |
6188 | continue; |
6189 | btrfs_dir_item_key_to_cpu(leaf, di, &di_key); | |
6190 | if (di_key.type == BTRFS_ROOT_ITEM_KEY) | |
6191 | continue; | |
6192 | ||
ec125cfb | 6193 | btrfs_release_path(path); |
0202e83f | 6194 | di_inode = btrfs_iget(fs_info->sb, di_key.objectid, root); |
2f2ff0ee FM |
6195 | if (IS_ERR(di_inode)) { |
6196 | ret = PTR_ERR(di_inode); | |
6197 | goto next_dir_inode; | |
6198 | } | |
6199 | ||
0e44cb3f | 6200 | if (!need_log_inode(trans, BTRFS_I(di_inode))) { |
410f954c | 6201 | btrfs_add_delayed_iput(di_inode); |
ec125cfb | 6202 | break; |
2f2ff0ee FM |
6203 | } |
6204 | ||
6205 | ctx->log_new_dentries = false; | |
d0e64a98 | 6206 | if (type == BTRFS_FT_DIR) |
2f2ff0ee | 6207 | log_mode = LOG_INODE_ALL; |
90d04510 | 6208 | ret = btrfs_log_inode(trans, BTRFS_I(di_inode), |
48778179 | 6209 | log_mode, ctx); |
410f954c | 6210 | btrfs_add_delayed_iput(di_inode); |
2f2ff0ee FM |
6211 | if (ret) |
6212 | goto next_dir_inode; | |
6213 | if (ctx->log_new_dentries) { | |
6214 | new_dir_elem = kmalloc(sizeof(*new_dir_elem), | |
6215 | GFP_NOFS); | |
6216 | if (!new_dir_elem) { | |
6217 | ret = -ENOMEM; | |
6218 | goto next_dir_inode; | |
6219 | } | |
6220 | new_dir_elem->ino = di_key.objectid; | |
6221 | list_add_tail(&new_dir_elem->list, &dir_list); | |
6222 | } | |
6223 | break; | |
6224 | } | |
2f2ff0ee FM |
6225 | if (min_key.offset < (u64)-1) { |
6226 | min_key.offset++; | |
6227 | goto again; | |
6228 | } | |
6229 | next_dir_inode: | |
6230 | list_del(&dir_elem->list); | |
6231 | kfree(dir_elem); | |
6232 | } | |
6233 | ||
6234 | btrfs_free_path(path); | |
6235 | return ret; | |
6236 | } | |
6237 | ||
18aa0922 | 6238 | static int btrfs_log_all_parents(struct btrfs_trans_handle *trans, |
d0a0b78d | 6239 | struct btrfs_inode *inode, |
18aa0922 FM |
6240 | struct btrfs_log_ctx *ctx) |
6241 | { | |
3ffbd68c | 6242 | struct btrfs_fs_info *fs_info = trans->fs_info; |
18aa0922 FM |
6243 | int ret; |
6244 | struct btrfs_path *path; | |
6245 | struct btrfs_key key; | |
d0a0b78d NB |
6246 | struct btrfs_root *root = inode->root; |
6247 | const u64 ino = btrfs_ino(inode); | |
18aa0922 FM |
6248 | |
6249 | path = btrfs_alloc_path(); | |
6250 | if (!path) | |
6251 | return -ENOMEM; | |
6252 | path->skip_locking = 1; | |
6253 | path->search_commit_root = 1; | |
6254 | ||
6255 | key.objectid = ino; | |
6256 | key.type = BTRFS_INODE_REF_KEY; | |
6257 | key.offset = 0; | |
6258 | ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); | |
6259 | if (ret < 0) | |
6260 | goto out; | |
6261 | ||
6262 | while (true) { | |
6263 | struct extent_buffer *leaf = path->nodes[0]; | |
6264 | int slot = path->slots[0]; | |
6265 | u32 cur_offset = 0; | |
6266 | u32 item_size; | |
6267 | unsigned long ptr; | |
6268 | ||
6269 | if (slot >= btrfs_header_nritems(leaf)) { | |
6270 | ret = btrfs_next_leaf(root, path); | |
6271 | if (ret < 0) | |
6272 | goto out; | |
6273 | else if (ret > 0) | |
6274 | break; | |
6275 | continue; | |
6276 | } | |
6277 | ||
6278 | btrfs_item_key_to_cpu(leaf, &key, slot); | |
6279 | /* BTRFS_INODE_EXTREF_KEY is BTRFS_INODE_REF_KEY + 1 */ | |
6280 | if (key.objectid != ino || key.type > BTRFS_INODE_EXTREF_KEY) | |
6281 | break; | |
6282 | ||
3212fa14 | 6283 | item_size = btrfs_item_size(leaf, slot); |
18aa0922 FM |
6284 | ptr = btrfs_item_ptr_offset(leaf, slot); |
6285 | while (cur_offset < item_size) { | |
6286 | struct btrfs_key inode_key; | |
6287 | struct inode *dir_inode; | |
6288 | ||
6289 | inode_key.type = BTRFS_INODE_ITEM_KEY; | |
6290 | inode_key.offset = 0; | |
6291 | ||
6292 | if (key.type == BTRFS_INODE_EXTREF_KEY) { | |
6293 | struct btrfs_inode_extref *extref; | |
6294 | ||
6295 | extref = (struct btrfs_inode_extref *) | |
6296 | (ptr + cur_offset); | |
6297 | inode_key.objectid = btrfs_inode_extref_parent( | |
6298 | leaf, extref); | |
6299 | cur_offset += sizeof(*extref); | |
6300 | cur_offset += btrfs_inode_extref_name_len(leaf, | |
6301 | extref); | |
6302 | } else { | |
6303 | inode_key.objectid = key.offset; | |
6304 | cur_offset = item_size; | |
6305 | } | |
6306 | ||
0202e83f DS |
6307 | dir_inode = btrfs_iget(fs_info->sb, inode_key.objectid, |
6308 | root); | |
0f375eed FM |
6309 | /* |
6310 | * If the parent inode was deleted, return an error to | |
6311 | * fallback to a transaction commit. This is to prevent | |
6312 | * getting an inode that was moved from one parent A to | |
6313 | * a parent B, got its former parent A deleted and then | |
6314 | * it got fsync'ed, from existing at both parents after | |
6315 | * a log replay (and the old parent still existing). | |
6316 | * Example: | |
6317 | * | |
6318 | * mkdir /mnt/A | |
6319 | * mkdir /mnt/B | |
6320 | * touch /mnt/B/bar | |
6321 | * sync | |
6322 | * mv /mnt/B/bar /mnt/A/bar | |
6323 | * mv -T /mnt/A /mnt/B | |
6324 | * fsync /mnt/B/bar | |
6325 | * <power fail> | |
6326 | * | |
6327 | * If we ignore the old parent B which got deleted, | |
6328 | * after a log replay we would have file bar linked | |
6329 | * at both parents and the old parent B would still | |
6330 | * exist. | |
6331 | */ | |
6332 | if (IS_ERR(dir_inode)) { | |
6333 | ret = PTR_ERR(dir_inode); | |
6334 | goto out; | |
6335 | } | |
18aa0922 | 6336 | |
3e6a86a1 FM |
6337 | if (!need_log_inode(trans, BTRFS_I(dir_inode))) { |
6338 | btrfs_add_delayed_iput(dir_inode); | |
6339 | continue; | |
6340 | } | |
6341 | ||
289cffcb | 6342 | ctx->log_new_dentries = false; |
90d04510 | 6343 | ret = btrfs_log_inode(trans, BTRFS_I(dir_inode), |
48778179 | 6344 | LOG_INODE_ALL, ctx); |
289cffcb | 6345 | if (!ret && ctx->log_new_dentries) |
657ed1aa | 6346 | ret = log_new_dir_dentries(trans, root, |
f85b7379 | 6347 | BTRFS_I(dir_inode), ctx); |
410f954c | 6348 | btrfs_add_delayed_iput(dir_inode); |
18aa0922 FM |
6349 | if (ret) |
6350 | goto out; | |
6351 | } | |
6352 | path->slots[0]++; | |
6353 | } | |
6354 | ret = 0; | |
6355 | out: | |
6356 | btrfs_free_path(path); | |
6357 | return ret; | |
6358 | } | |
6359 | ||
b8aa330d FM |
6360 | static int log_new_ancestors(struct btrfs_trans_handle *trans, |
6361 | struct btrfs_root *root, | |
6362 | struct btrfs_path *path, | |
6363 | struct btrfs_log_ctx *ctx) | |
6364 | { | |
6365 | struct btrfs_key found_key; | |
6366 | ||
6367 | btrfs_item_key_to_cpu(path->nodes[0], &found_key, path->slots[0]); | |
6368 | ||
6369 | while (true) { | |
6370 | struct btrfs_fs_info *fs_info = root->fs_info; | |
b8aa330d FM |
6371 | struct extent_buffer *leaf = path->nodes[0]; |
6372 | int slot = path->slots[0]; | |
6373 | struct btrfs_key search_key; | |
6374 | struct inode *inode; | |
0202e83f | 6375 | u64 ino; |
b8aa330d FM |
6376 | int ret = 0; |
6377 | ||
6378 | btrfs_release_path(path); | |
6379 | ||
0202e83f DS |
6380 | ino = found_key.offset; |
6381 | ||
b8aa330d FM |
6382 | search_key.objectid = found_key.offset; |
6383 | search_key.type = BTRFS_INODE_ITEM_KEY; | |
6384 | search_key.offset = 0; | |
0202e83f | 6385 | inode = btrfs_iget(fs_info->sb, ino, root); |
b8aa330d FM |
6386 | if (IS_ERR(inode)) |
6387 | return PTR_ERR(inode); | |
6388 | ||
ab12313a FM |
6389 | if (BTRFS_I(inode)->generation >= trans->transid && |
6390 | need_log_inode(trans, BTRFS_I(inode))) | |
90d04510 | 6391 | ret = btrfs_log_inode(trans, BTRFS_I(inode), |
48778179 | 6392 | LOG_INODE_EXISTS, ctx); |
410f954c | 6393 | btrfs_add_delayed_iput(inode); |
b8aa330d FM |
6394 | if (ret) |
6395 | return ret; | |
6396 | ||
6397 | if (search_key.objectid == BTRFS_FIRST_FREE_OBJECTID) | |
6398 | break; | |
6399 | ||
6400 | search_key.type = BTRFS_INODE_REF_KEY; | |
6401 | ret = btrfs_search_slot(NULL, root, &search_key, path, 0, 0); | |
6402 | if (ret < 0) | |
6403 | return ret; | |
6404 | ||
6405 | leaf = path->nodes[0]; | |
6406 | slot = path->slots[0]; | |
6407 | if (slot >= btrfs_header_nritems(leaf)) { | |
6408 | ret = btrfs_next_leaf(root, path); | |
6409 | if (ret < 0) | |
6410 | return ret; | |
6411 | else if (ret > 0) | |
6412 | return -ENOENT; | |
6413 | leaf = path->nodes[0]; | |
6414 | slot = path->slots[0]; | |
6415 | } | |
6416 | ||
6417 | btrfs_item_key_to_cpu(leaf, &found_key, slot); | |
6418 | if (found_key.objectid != search_key.objectid || | |
6419 | found_key.type != BTRFS_INODE_REF_KEY) | |
6420 | return -ENOENT; | |
6421 | } | |
6422 | return 0; | |
6423 | } | |
6424 | ||
6425 | static int log_new_ancestors_fast(struct btrfs_trans_handle *trans, | |
6426 | struct btrfs_inode *inode, | |
6427 | struct dentry *parent, | |
6428 | struct btrfs_log_ctx *ctx) | |
6429 | { | |
6430 | struct btrfs_root *root = inode->root; | |
b8aa330d FM |
6431 | struct dentry *old_parent = NULL; |
6432 | struct super_block *sb = inode->vfs_inode.i_sb; | |
6433 | int ret = 0; | |
6434 | ||
6435 | while (true) { | |
6436 | if (!parent || d_really_is_negative(parent) || | |
6437 | sb != parent->d_sb) | |
6438 | break; | |
6439 | ||
6440 | inode = BTRFS_I(d_inode(parent)); | |
6441 | if (root != inode->root) | |
6442 | break; | |
6443 | ||
ab12313a FM |
6444 | if (inode->generation >= trans->transid && |
6445 | need_log_inode(trans, inode)) { | |
90d04510 | 6446 | ret = btrfs_log_inode(trans, inode, |
48778179 | 6447 | LOG_INODE_EXISTS, ctx); |
b8aa330d FM |
6448 | if (ret) |
6449 | break; | |
6450 | } | |
6451 | if (IS_ROOT(parent)) | |
6452 | break; | |
6453 | ||
6454 | parent = dget_parent(parent); | |
6455 | dput(old_parent); | |
6456 | old_parent = parent; | |
6457 | } | |
6458 | dput(old_parent); | |
6459 | ||
6460 | return ret; | |
6461 | } | |
6462 | ||
6463 | static int log_all_new_ancestors(struct btrfs_trans_handle *trans, | |
6464 | struct btrfs_inode *inode, | |
6465 | struct dentry *parent, | |
6466 | struct btrfs_log_ctx *ctx) | |
6467 | { | |
6468 | struct btrfs_root *root = inode->root; | |
6469 | const u64 ino = btrfs_ino(inode); | |
6470 | struct btrfs_path *path; | |
6471 | struct btrfs_key search_key; | |
6472 | int ret; | |
6473 | ||
6474 | /* | |
6475 | * For a single hard link case, go through a fast path that does not | |
6476 | * need to iterate the fs/subvolume tree. | |
6477 | */ | |
6478 | if (inode->vfs_inode.i_nlink < 2) | |
6479 | return log_new_ancestors_fast(trans, inode, parent, ctx); | |
6480 | ||
6481 | path = btrfs_alloc_path(); | |
6482 | if (!path) | |
6483 | return -ENOMEM; | |
6484 | ||
6485 | search_key.objectid = ino; | |
6486 | search_key.type = BTRFS_INODE_REF_KEY; | |
6487 | search_key.offset = 0; | |
6488 | again: | |
6489 | ret = btrfs_search_slot(NULL, root, &search_key, path, 0, 0); | |
6490 | if (ret < 0) | |
6491 | goto out; | |
6492 | if (ret == 0) | |
6493 | path->slots[0]++; | |
6494 | ||
6495 | while (true) { | |
6496 | struct extent_buffer *leaf = path->nodes[0]; | |
6497 | int slot = path->slots[0]; | |
6498 | struct btrfs_key found_key; | |
6499 | ||
6500 | if (slot >= btrfs_header_nritems(leaf)) { | |
6501 | ret = btrfs_next_leaf(root, path); | |
6502 | if (ret < 0) | |
6503 | goto out; | |
6504 | else if (ret > 0) | |
6505 | break; | |
6506 | continue; | |
6507 | } | |
6508 | ||
6509 | btrfs_item_key_to_cpu(leaf, &found_key, slot); | |
6510 | if (found_key.objectid != ino || | |
6511 | found_key.type > BTRFS_INODE_EXTREF_KEY) | |
6512 | break; | |
6513 | ||
6514 | /* | |
6515 | * Don't deal with extended references because they are rare | |
6516 | * cases and too complex to deal with (we would need to keep | |
6517 | * track of which subitem we are processing for each item in | |
6518 | * this loop, etc). So just return some error to fallback to | |
6519 | * a transaction commit. | |
6520 | */ | |
6521 | if (found_key.type == BTRFS_INODE_EXTREF_KEY) { | |
6522 | ret = -EMLINK; | |
6523 | goto out; | |
6524 | } | |
6525 | ||
6526 | /* | |
6527 | * Logging ancestors needs to do more searches on the fs/subvol | |
6528 | * tree, so it releases the path as needed to avoid deadlocks. | |
6529 | * Keep track of the last inode ref key and resume from that key | |
6530 | * after logging all new ancestors for the current hard link. | |
6531 | */ | |
6532 | memcpy(&search_key, &found_key, sizeof(search_key)); | |
6533 | ||
6534 | ret = log_new_ancestors(trans, root, path, ctx); | |
6535 | if (ret) | |
6536 | goto out; | |
6537 | btrfs_release_path(path); | |
6538 | goto again; | |
6539 | } | |
6540 | ret = 0; | |
6541 | out: | |
6542 | btrfs_free_path(path); | |
6543 | return ret; | |
6544 | } | |
6545 | ||
e02119d5 CM |
6546 | /* |
6547 | * helper function around btrfs_log_inode to make sure newly created | |
6548 | * parent directories also end up in the log. A minimal inode and backref | |
6549 | * only logging is done of any parent directories that are older than | |
6550 | * the last committed transaction | |
6551 | */ | |
48a3b636 | 6552 | static int btrfs_log_inode_parent(struct btrfs_trans_handle *trans, |
19df27a9 | 6553 | struct btrfs_inode *inode, |
49dae1bc | 6554 | struct dentry *parent, |
41a1eada | 6555 | int inode_only, |
8b050d35 | 6556 | struct btrfs_log_ctx *ctx) |
e02119d5 | 6557 | { |
f882274b | 6558 | struct btrfs_root *root = inode->root; |
0b246afa | 6559 | struct btrfs_fs_info *fs_info = root->fs_info; |
12fcfd22 | 6560 | int ret = 0; |
2f2ff0ee | 6561 | bool log_dentries = false; |
12fcfd22 | 6562 | |
0b246afa | 6563 | if (btrfs_test_opt(fs_info, NOTREELOG)) { |
f31f09f6 | 6564 | ret = BTRFS_LOG_FORCE_COMMIT; |
3a5e1404 SW |
6565 | goto end_no_trans; |
6566 | } | |
6567 | ||
f882274b | 6568 | if (btrfs_root_refs(&root->root_item) == 0) { |
f31f09f6 | 6569 | ret = BTRFS_LOG_FORCE_COMMIT; |
76dda93c YZ |
6570 | goto end_no_trans; |
6571 | } | |
6572 | ||
f2d72f42 FM |
6573 | /* |
6574 | * Skip already logged inodes or inodes corresponding to tmpfiles | |
6575 | * (since logging them is pointless, a link count of 0 means they | |
6576 | * will never be accessible). | |
6577 | */ | |
626e9f41 FM |
6578 | if ((btrfs_inode_in_log(inode, trans->transid) && |
6579 | list_empty(&ctx->ordered_extents)) || | |
f2d72f42 | 6580 | inode->vfs_inode.i_nlink == 0) { |
257c62e1 CM |
6581 | ret = BTRFS_NO_LOG_SYNC; |
6582 | goto end_no_trans; | |
6583 | } | |
6584 | ||
8b050d35 | 6585 | ret = start_log_trans(trans, root, ctx); |
4a500fd1 | 6586 | if (ret) |
e87ac136 | 6587 | goto end_no_trans; |
e02119d5 | 6588 | |
90d04510 | 6589 | ret = btrfs_log_inode(trans, inode, inode_only, ctx); |
4a500fd1 YZ |
6590 | if (ret) |
6591 | goto end_trans; | |
12fcfd22 | 6592 | |
af4176b4 CM |
6593 | /* |
6594 | * for regular files, if its inode is already on disk, we don't | |
6595 | * have to worry about the parents at all. This is because | |
6596 | * we can use the last_unlink_trans field to record renames | |
6597 | * and other fun in this file. | |
6598 | */ | |
19df27a9 | 6599 | if (S_ISREG(inode->vfs_inode.i_mode) && |
47d3db41 FM |
6600 | inode->generation < trans->transid && |
6601 | inode->last_unlink_trans < trans->transid) { | |
4a500fd1 YZ |
6602 | ret = 0; |
6603 | goto end_trans; | |
6604 | } | |
af4176b4 | 6605 | |
289cffcb | 6606 | if (S_ISDIR(inode->vfs_inode.i_mode) && ctx->log_new_dentries) |
2f2ff0ee FM |
6607 | log_dentries = true; |
6608 | ||
18aa0922 | 6609 | /* |
01327610 | 6610 | * On unlink we must make sure all our current and old parent directory |
18aa0922 FM |
6611 | * inodes are fully logged. This is to prevent leaving dangling |
6612 | * directory index entries in directories that were our parents but are | |
6613 | * not anymore. Not doing this results in old parent directory being | |
6614 | * impossible to delete after log replay (rmdir will always fail with | |
6615 | * error -ENOTEMPTY). | |
6616 | * | |
6617 | * Example 1: | |
6618 | * | |
6619 | * mkdir testdir | |
6620 | * touch testdir/foo | |
6621 | * ln testdir/foo testdir/bar | |
6622 | * sync | |
6623 | * unlink testdir/bar | |
6624 | * xfs_io -c fsync testdir/foo | |
6625 | * <power failure> | |
6626 | * mount fs, triggers log replay | |
6627 | * | |
6628 | * If we don't log the parent directory (testdir), after log replay the | |
6629 | * directory still has an entry pointing to the file inode using the bar | |
6630 | * name, but a matching BTRFS_INODE_[REF|EXTREF]_KEY does not exist and | |
6631 | * the file inode has a link count of 1. | |
6632 | * | |
6633 | * Example 2: | |
6634 | * | |
6635 | * mkdir testdir | |
6636 | * touch foo | |
6637 | * ln foo testdir/foo2 | |
6638 | * ln foo testdir/foo3 | |
6639 | * sync | |
6640 | * unlink testdir/foo3 | |
6641 | * xfs_io -c fsync foo | |
6642 | * <power failure> | |
6643 | * mount fs, triggers log replay | |
6644 | * | |
6645 | * Similar as the first example, after log replay the parent directory | |
6646 | * testdir still has an entry pointing to the inode file with name foo3 | |
6647 | * but the file inode does not have a matching BTRFS_INODE_REF_KEY item | |
6648 | * and has a link count of 2. | |
6649 | */ | |
47d3db41 | 6650 | if (inode->last_unlink_trans >= trans->transid) { |
b8aa330d | 6651 | ret = btrfs_log_all_parents(trans, inode, ctx); |
18aa0922 FM |
6652 | if (ret) |
6653 | goto end_trans; | |
6654 | } | |
6655 | ||
b8aa330d FM |
6656 | ret = log_all_new_ancestors(trans, inode, parent, ctx); |
6657 | if (ret) | |
41bd6067 | 6658 | goto end_trans; |
76dda93c | 6659 | |
2f2ff0ee | 6660 | if (log_dentries) |
b8aa330d | 6661 | ret = log_new_dir_dentries(trans, root, inode, ctx); |
2f2ff0ee FM |
6662 | else |
6663 | ret = 0; | |
4a500fd1 YZ |
6664 | end_trans: |
6665 | if (ret < 0) { | |
90787766 | 6666 | btrfs_set_log_full_commit(trans); |
f31f09f6 | 6667 | ret = BTRFS_LOG_FORCE_COMMIT; |
4a500fd1 | 6668 | } |
8b050d35 MX |
6669 | |
6670 | if (ret) | |
6671 | btrfs_remove_log_ctx(root, ctx); | |
12fcfd22 CM |
6672 | btrfs_end_log_trans(root); |
6673 | end_no_trans: | |
6674 | return ret; | |
e02119d5 CM |
6675 | } |
6676 | ||
6677 | /* | |
6678 | * it is not safe to log dentry if the chunk root has added new | |
6679 | * chunks. This returns 0 if the dentry was logged, and 1 otherwise. | |
6680 | * If this returns 1, you must commit the transaction to safely get your | |
6681 | * data on disk. | |
6682 | */ | |
6683 | int btrfs_log_dentry_safe(struct btrfs_trans_handle *trans, | |
e5b84f7a | 6684 | struct dentry *dentry, |
8b050d35 | 6685 | struct btrfs_log_ctx *ctx) |
e02119d5 | 6686 | { |
6a912213 JB |
6687 | struct dentry *parent = dget_parent(dentry); |
6688 | int ret; | |
6689 | ||
f882274b | 6690 | ret = btrfs_log_inode_parent(trans, BTRFS_I(d_inode(dentry)), parent, |
48778179 | 6691 | LOG_INODE_ALL, ctx); |
6a912213 JB |
6692 | dput(parent); |
6693 | ||
6694 | return ret; | |
e02119d5 CM |
6695 | } |
6696 | ||
6697 | /* | |
6698 | * should be called during mount to recover any replay any log trees | |
6699 | * from the FS | |
6700 | */ | |
6701 | int btrfs_recover_log_trees(struct btrfs_root *log_root_tree) | |
6702 | { | |
6703 | int ret; | |
6704 | struct btrfs_path *path; | |
6705 | struct btrfs_trans_handle *trans; | |
6706 | struct btrfs_key key; | |
6707 | struct btrfs_key found_key; | |
e02119d5 CM |
6708 | struct btrfs_root *log; |
6709 | struct btrfs_fs_info *fs_info = log_root_tree->fs_info; | |
6710 | struct walk_control wc = { | |
6711 | .process_func = process_one_buffer, | |
430a6626 | 6712 | .stage = LOG_WALK_PIN_ONLY, |
e02119d5 CM |
6713 | }; |
6714 | ||
e02119d5 | 6715 | path = btrfs_alloc_path(); |
db5b493a TI |
6716 | if (!path) |
6717 | return -ENOMEM; | |
6718 | ||
afcdd129 | 6719 | set_bit(BTRFS_FS_LOG_RECOVERING, &fs_info->flags); |
e02119d5 | 6720 | |
4a500fd1 | 6721 | trans = btrfs_start_transaction(fs_info->tree_root, 0); |
79787eaa JM |
6722 | if (IS_ERR(trans)) { |
6723 | ret = PTR_ERR(trans); | |
6724 | goto error; | |
6725 | } | |
e02119d5 CM |
6726 | |
6727 | wc.trans = trans; | |
6728 | wc.pin = 1; | |
6729 | ||
db5b493a | 6730 | ret = walk_log_tree(trans, log_root_tree, &wc); |
79787eaa | 6731 | if (ret) { |
ba51e2a1 | 6732 | btrfs_abort_transaction(trans, ret); |
79787eaa JM |
6733 | goto error; |
6734 | } | |
e02119d5 CM |
6735 | |
6736 | again: | |
6737 | key.objectid = BTRFS_TREE_LOG_OBJECTID; | |
6738 | key.offset = (u64)-1; | |
962a298f | 6739 | key.type = BTRFS_ROOT_ITEM_KEY; |
e02119d5 | 6740 | |
d397712b | 6741 | while (1) { |
e02119d5 | 6742 | ret = btrfs_search_slot(NULL, log_root_tree, &key, path, 0, 0); |
79787eaa JM |
6743 | |
6744 | if (ret < 0) { | |
ba51e2a1 | 6745 | btrfs_abort_transaction(trans, ret); |
79787eaa JM |
6746 | goto error; |
6747 | } | |
e02119d5 CM |
6748 | if (ret > 0) { |
6749 | if (path->slots[0] == 0) | |
6750 | break; | |
6751 | path->slots[0]--; | |
6752 | } | |
6753 | btrfs_item_key_to_cpu(path->nodes[0], &found_key, | |
6754 | path->slots[0]); | |
b3b4aa74 | 6755 | btrfs_release_path(path); |
e02119d5 CM |
6756 | if (found_key.objectid != BTRFS_TREE_LOG_OBJECTID) |
6757 | break; | |
6758 | ||
62a2c73e | 6759 | log = btrfs_read_tree_root(log_root_tree, &found_key); |
79787eaa JM |
6760 | if (IS_ERR(log)) { |
6761 | ret = PTR_ERR(log); | |
ba51e2a1 | 6762 | btrfs_abort_transaction(trans, ret); |
79787eaa JM |
6763 | goto error; |
6764 | } | |
e02119d5 | 6765 | |
56e9357a DS |
6766 | wc.replay_dest = btrfs_get_fs_root(fs_info, found_key.offset, |
6767 | true); | |
79787eaa JM |
6768 | if (IS_ERR(wc.replay_dest)) { |
6769 | ret = PTR_ERR(wc.replay_dest); | |
9bc574de JB |
6770 | |
6771 | /* | |
6772 | * We didn't find the subvol, likely because it was | |
6773 | * deleted. This is ok, simply skip this log and go to | |
6774 | * the next one. | |
6775 | * | |
6776 | * We need to exclude the root because we can't have | |
6777 | * other log replays overwriting this log as we'll read | |
6778 | * it back in a few more times. This will keep our | |
6779 | * block from being modified, and we'll just bail for | |
6780 | * each subsequent pass. | |
6781 | */ | |
6782 | if (ret == -ENOENT) | |
9fce5704 | 6783 | ret = btrfs_pin_extent_for_log_replay(trans, |
9bc574de JB |
6784 | log->node->start, |
6785 | log->node->len); | |
00246528 | 6786 | btrfs_put_root(log); |
9bc574de JB |
6787 | |
6788 | if (!ret) | |
6789 | goto next; | |
ba51e2a1 | 6790 | btrfs_abort_transaction(trans, ret); |
79787eaa JM |
6791 | goto error; |
6792 | } | |
e02119d5 | 6793 | |
07d400a6 | 6794 | wc.replay_dest->log_root = log; |
2002ae11 JB |
6795 | ret = btrfs_record_root_in_trans(trans, wc.replay_dest); |
6796 | if (ret) | |
6797 | /* The loop needs to continue due to the root refs */ | |
ba51e2a1 | 6798 | btrfs_abort_transaction(trans, ret); |
2002ae11 JB |
6799 | else |
6800 | ret = walk_log_tree(trans, log, &wc); | |
e02119d5 | 6801 | |
b50c6e25 | 6802 | if (!ret && wc.stage == LOG_WALK_REPLAY_ALL) { |
e02119d5 CM |
6803 | ret = fixup_inode_link_counts(trans, wc.replay_dest, |
6804 | path); | |
ba51e2a1 JB |
6805 | if (ret) |
6806 | btrfs_abort_transaction(trans, ret); | |
e02119d5 CM |
6807 | } |
6808 | ||
900c9981 LB |
6809 | if (!ret && wc.stage == LOG_WALK_REPLAY_ALL) { |
6810 | struct btrfs_root *root = wc.replay_dest; | |
6811 | ||
6812 | btrfs_release_path(path); | |
6813 | ||
6814 | /* | |
6815 | * We have just replayed everything, and the highest | |
6816 | * objectid of fs roots probably has changed in case | |
6817 | * some inode_item's got replayed. | |
6818 | * | |
6819 | * root->objectid_mutex is not acquired as log replay | |
6820 | * could only happen during mount. | |
6821 | */ | |
453e4873 | 6822 | ret = btrfs_init_root_free_objectid(root); |
ba51e2a1 JB |
6823 | if (ret) |
6824 | btrfs_abort_transaction(trans, ret); | |
900c9981 LB |
6825 | } |
6826 | ||
07d400a6 | 6827 | wc.replay_dest->log_root = NULL; |
00246528 | 6828 | btrfs_put_root(wc.replay_dest); |
00246528 | 6829 | btrfs_put_root(log); |
e02119d5 | 6830 | |
b50c6e25 JB |
6831 | if (ret) |
6832 | goto error; | |
9bc574de | 6833 | next: |
e02119d5 CM |
6834 | if (found_key.offset == 0) |
6835 | break; | |
9bc574de | 6836 | key.offset = found_key.offset - 1; |
e02119d5 | 6837 | } |
b3b4aa74 | 6838 | btrfs_release_path(path); |
e02119d5 CM |
6839 | |
6840 | /* step one is to pin it all, step two is to replay just inodes */ | |
6841 | if (wc.pin) { | |
6842 | wc.pin = 0; | |
6843 | wc.process_func = replay_one_buffer; | |
6844 | wc.stage = LOG_WALK_REPLAY_INODES; | |
6845 | goto again; | |
6846 | } | |
6847 | /* step three is to replay everything */ | |
6848 | if (wc.stage < LOG_WALK_REPLAY_ALL) { | |
6849 | wc.stage++; | |
6850 | goto again; | |
6851 | } | |
6852 | ||
6853 | btrfs_free_path(path); | |
6854 | ||
abefa55a | 6855 | /* step 4: commit the transaction, which also unpins the blocks */ |
3a45bb20 | 6856 | ret = btrfs_commit_transaction(trans); |
abefa55a JB |
6857 | if (ret) |
6858 | return ret; | |
6859 | ||
e02119d5 | 6860 | log_root_tree->log_root = NULL; |
afcdd129 | 6861 | clear_bit(BTRFS_FS_LOG_RECOVERING, &fs_info->flags); |
00246528 | 6862 | btrfs_put_root(log_root_tree); |
79787eaa | 6863 | |
abefa55a | 6864 | return 0; |
79787eaa | 6865 | error: |
b50c6e25 | 6866 | if (wc.trans) |
3a45bb20 | 6867 | btrfs_end_transaction(wc.trans); |
1aeb6b56 | 6868 | clear_bit(BTRFS_FS_LOG_RECOVERING, &fs_info->flags); |
79787eaa JM |
6869 | btrfs_free_path(path); |
6870 | return ret; | |
e02119d5 | 6871 | } |
12fcfd22 CM |
6872 | |
6873 | /* | |
6874 | * there are some corner cases where we want to force a full | |
6875 | * commit instead of allowing a directory to be logged. | |
6876 | * | |
6877 | * They revolve around files there were unlinked from the directory, and | |
6878 | * this function updates the parent directory so that a full commit is | |
6879 | * properly done if it is fsync'd later after the unlinks are done. | |
2be63d5c FM |
6880 | * |
6881 | * Must be called before the unlink operations (updates to the subvolume tree, | |
6882 | * inodes, etc) are done. | |
12fcfd22 CM |
6883 | */ |
6884 | void btrfs_record_unlink_dir(struct btrfs_trans_handle *trans, | |
4176bdbf | 6885 | struct btrfs_inode *dir, struct btrfs_inode *inode, |
12fcfd22 CM |
6886 | int for_rename) |
6887 | { | |
af4176b4 CM |
6888 | /* |
6889 | * when we're logging a file, if it hasn't been renamed | |
6890 | * or unlinked, and its inode is fully committed on disk, | |
6891 | * we don't have to worry about walking up the directory chain | |
6892 | * to log its parents. | |
6893 | * | |
6894 | * So, we use the last_unlink_trans field to put this transid | |
6895 | * into the file. When the file is logged we check it and | |
6896 | * don't log the parents if the file is fully on disk. | |
6897 | */ | |
4176bdbf NB |
6898 | mutex_lock(&inode->log_mutex); |
6899 | inode->last_unlink_trans = trans->transid; | |
6900 | mutex_unlock(&inode->log_mutex); | |
af4176b4 | 6901 | |
12fcfd22 CM |
6902 | /* |
6903 | * if this directory was already logged any new | |
6904 | * names for this file/dir will get recorded | |
6905 | */ | |
4176bdbf | 6906 | if (dir->logged_trans == trans->transid) |
12fcfd22 CM |
6907 | return; |
6908 | ||
6909 | /* | |
6910 | * if the inode we're about to unlink was logged, | |
6911 | * the log will be properly updated for any new names | |
6912 | */ | |
4176bdbf | 6913 | if (inode->logged_trans == trans->transid) |
12fcfd22 CM |
6914 | return; |
6915 | ||
6916 | /* | |
6917 | * when renaming files across directories, if the directory | |
6918 | * there we're unlinking from gets fsync'd later on, there's | |
6919 | * no way to find the destination directory later and fsync it | |
6920 | * properly. So, we have to be conservative and force commits | |
6921 | * so the new name gets discovered. | |
6922 | */ | |
6923 | if (for_rename) | |
6924 | goto record; | |
6925 | ||
6926 | /* we can safely do the unlink without any special recording */ | |
6927 | return; | |
6928 | ||
6929 | record: | |
4176bdbf NB |
6930 | mutex_lock(&dir->log_mutex); |
6931 | dir->last_unlink_trans = trans->transid; | |
6932 | mutex_unlock(&dir->log_mutex); | |
1ec9a1ae FM |
6933 | } |
6934 | ||
6935 | /* | |
6936 | * Make sure that if someone attempts to fsync the parent directory of a deleted | |
6937 | * snapshot, it ends up triggering a transaction commit. This is to guarantee | |
6938 | * that after replaying the log tree of the parent directory's root we will not | |
6939 | * see the snapshot anymore and at log replay time we will not see any log tree | |
6940 | * corresponding to the deleted snapshot's root, which could lead to replaying | |
6941 | * it after replaying the log tree of the parent directory (which would replay | |
6942 | * the snapshot delete operation). | |
2be63d5c FM |
6943 | * |
6944 | * Must be called before the actual snapshot destroy operation (updates to the | |
6945 | * parent root and tree of tree roots trees, etc) are done. | |
1ec9a1ae FM |
6946 | */ |
6947 | void btrfs_record_snapshot_destroy(struct btrfs_trans_handle *trans, | |
43663557 | 6948 | struct btrfs_inode *dir) |
1ec9a1ae | 6949 | { |
43663557 NB |
6950 | mutex_lock(&dir->log_mutex); |
6951 | dir->last_unlink_trans = trans->transid; | |
6952 | mutex_unlock(&dir->log_mutex); | |
12fcfd22 CM |
6953 | } |
6954 | ||
d5f5bd54 FM |
6955 | /** |
6956 | * Update the log after adding a new name for an inode. | |
6957 | * | |
6958 | * @trans: Transaction handle. | |
6959 | * @old_dentry: The dentry associated with the old name and the old | |
6960 | * parent directory. | |
6961 | * @old_dir: The inode of the previous parent directory for the case | |
6962 | * of a rename. For a link operation, it must be NULL. | |
88d2beec FM |
6963 | * @old_dir_index: The index number associated with the old name, meaningful |
6964 | * only for rename operations (when @old_dir is not NULL). | |
6965 | * Ignored for link operations. | |
d5f5bd54 FM |
6966 | * @parent: The dentry associated with the directory under which the |
6967 | * new name is located. | |
6968 | * | |
6969 | * Call this after adding a new name for an inode, as a result of a link or | |
6970 | * rename operation, and it will properly update the log to reflect the new name. | |
12fcfd22 | 6971 | */ |
75b463d2 | 6972 | void btrfs_log_new_name(struct btrfs_trans_handle *trans, |
d5f5bd54 | 6973 | struct dentry *old_dentry, struct btrfs_inode *old_dir, |
88d2beec | 6974 | u64 old_dir_index, struct dentry *parent) |
12fcfd22 | 6975 | { |
d5f5bd54 | 6976 | struct btrfs_inode *inode = BTRFS_I(d_inode(old_dentry)); |
259c4b96 | 6977 | struct btrfs_root *root = inode->root; |
75b463d2 | 6978 | struct btrfs_log_ctx ctx; |
259c4b96 | 6979 | bool log_pinned = false; |
0f8ce498 | 6980 | int ret; |
12fcfd22 | 6981 | |
af4176b4 CM |
6982 | /* |
6983 | * this will force the logging code to walk the dentry chain | |
6984 | * up for the file | |
6985 | */ | |
9a6509c4 | 6986 | if (!S_ISDIR(inode->vfs_inode.i_mode)) |
9ca5fbfb | 6987 | inode->last_unlink_trans = trans->transid; |
af4176b4 | 6988 | |
12fcfd22 CM |
6989 | /* |
6990 | * if this inode hasn't been logged and directory we're renaming it | |
6991 | * from hasn't been logged, we don't need to log it | |
6992 | */ | |
0f8ce498 FM |
6993 | ret = inode_logged(trans, inode, NULL); |
6994 | if (ret < 0) { | |
6995 | goto out; | |
6996 | } else if (ret == 0) { | |
6997 | if (!old_dir) | |
6998 | return; | |
6999 | /* | |
7000 | * If the inode was not logged and we are doing a rename (old_dir is not | |
7001 | * NULL), check if old_dir was logged - if it was not we can return and | |
7002 | * do nothing. | |
7003 | */ | |
7004 | ret = inode_logged(trans, old_dir, NULL); | |
7005 | if (ret < 0) | |
7006 | goto out; | |
7007 | else if (ret == 0) | |
7008 | return; | |
7009 | } | |
7010 | ret = 0; | |
12fcfd22 | 7011 | |
54a40fc3 FM |
7012 | /* |
7013 | * If we are doing a rename (old_dir is not NULL) from a directory that | |
88d2beec FM |
7014 | * was previously logged, make sure that on log replay we get the old |
7015 | * dir entry deleted. This is needed because we will also log the new | |
7016 | * name of the renamed inode, so we need to make sure that after log | |
7017 | * replay we don't end up with both the new and old dir entries existing. | |
54a40fc3 | 7018 | */ |
88d2beec FM |
7019 | if (old_dir && old_dir->logged_trans == trans->transid) { |
7020 | struct btrfs_root *log = old_dir->root->log_root; | |
7021 | struct btrfs_path *path; | |
88d2beec FM |
7022 | |
7023 | ASSERT(old_dir_index >= BTRFS_DIR_START_INDEX); | |
7024 | ||
259c4b96 FM |
7025 | /* |
7026 | * We have two inodes to update in the log, the old directory and | |
7027 | * the inode that got renamed, so we must pin the log to prevent | |
7028 | * anyone from syncing the log until we have updated both inodes | |
7029 | * in the log. | |
7030 | */ | |
723df2bc FM |
7031 | ret = join_running_log_trans(root); |
7032 | /* | |
7033 | * At least one of the inodes was logged before, so this should | |
7034 | * not fail, but if it does, it's not serious, just bail out and | |
7035 | * mark the log for a full commit. | |
7036 | */ | |
7037 | if (WARN_ON_ONCE(ret < 0)) | |
7038 | goto out; | |
259c4b96 | 7039 | log_pinned = true; |
259c4b96 | 7040 | |
88d2beec FM |
7041 | path = btrfs_alloc_path(); |
7042 | if (!path) { | |
259c4b96 FM |
7043 | ret = -ENOMEM; |
7044 | goto out; | |
88d2beec FM |
7045 | } |
7046 | ||
7047 | /* | |
7048 | * Other concurrent task might be logging the old directory, | |
7049 | * as it can be triggered when logging other inode that had or | |
750ee454 FM |
7050 | * still has a dentry in the old directory. We lock the old |
7051 | * directory's log_mutex to ensure the deletion of the old | |
7052 | * name is persisted, because during directory logging we | |
7053 | * delete all BTRFS_DIR_LOG_INDEX_KEY keys and the deletion of | |
7054 | * the old name's dir index item is in the delayed items, so | |
7055 | * it could be missed by an in progress directory logging. | |
88d2beec FM |
7056 | */ |
7057 | mutex_lock(&old_dir->log_mutex); | |
7058 | ret = del_logged_dentry(trans, log, path, btrfs_ino(old_dir), | |
7059 | old_dentry->d_name.name, | |
7060 | old_dentry->d_name.len, old_dir_index); | |
7061 | if (ret > 0) { | |
7062 | /* | |
7063 | * The dentry does not exist in the log, so record its | |
7064 | * deletion. | |
7065 | */ | |
7066 | btrfs_release_path(path); | |
7067 | ret = insert_dir_log_key(trans, log, path, | |
7068 | btrfs_ino(old_dir), | |
7069 | old_dir_index, old_dir_index); | |
7070 | } | |
7071 | mutex_unlock(&old_dir->log_mutex); | |
7072 | ||
7073 | btrfs_free_path(path); | |
259c4b96 FM |
7074 | if (ret < 0) |
7075 | goto out; | |
88d2beec | 7076 | } |
54a40fc3 | 7077 | |
75b463d2 FM |
7078 | btrfs_init_log_ctx(&ctx, &inode->vfs_inode); |
7079 | ctx.logging_new_name = true; | |
7080 | /* | |
7081 | * We don't care about the return value. If we fail to log the new name | |
7082 | * then we know the next attempt to sync the log will fallback to a full | |
7083 | * transaction commit (due to a call to btrfs_set_log_full_commit()), so | |
7084 | * we don't need to worry about getting a log committed that has an | |
7085 | * inconsistent state after a rename operation. | |
7086 | */ | |
48778179 | 7087 | btrfs_log_inode_parent(trans, inode, parent, LOG_INODE_EXISTS, &ctx); |
259c4b96 | 7088 | out: |
0f8ce498 FM |
7089 | /* |
7090 | * If an error happened mark the log for a full commit because it's not | |
7091 | * consistent and up to date or we couldn't find out if one of the | |
7092 | * inodes was logged before in this transaction. Do it before unpinning | |
7093 | * the log, to avoid any races with someone else trying to commit it. | |
7094 | */ | |
7095 | if (ret < 0) | |
7096 | btrfs_set_log_full_commit(trans); | |
7097 | if (log_pinned) | |
259c4b96 | 7098 | btrfs_end_log_trans(root); |
12fcfd22 CM |
7099 | } |
7100 |