Commit | Line | Data |
---|---|---|
31db9f7c AB |
1 | /* |
2 | * Copyright (C) 2012 Alexander Block. All rights reserved. | |
3 | * | |
4 | * This program is free software; you can redistribute it and/or | |
5 | * modify it under the terms of the GNU General Public | |
6 | * License v2 as published by the Free Software Foundation. | |
7 | * | |
8 | * This program is distributed in the hope that it will be useful, | |
9 | * but WITHOUT ANY WARRANTY; without even the implied warranty of | |
10 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU | |
11 | * General Public License for more details. | |
12 | * | |
13 | * You should have received a copy of the GNU General Public | |
14 | * License along with this program; if not, write to the | |
15 | * Free Software Foundation, Inc., 59 Temple Place - Suite 330, | |
16 | * Boston, MA 021110-1307, USA. | |
17 | */ | |
18 | ||
19 | #include <linux/bsearch.h> | |
20 | #include <linux/fs.h> | |
21 | #include <linux/file.h> | |
22 | #include <linux/sort.h> | |
23 | #include <linux/mount.h> | |
24 | #include <linux/xattr.h> | |
25 | #include <linux/posix_acl_xattr.h> | |
26 | #include <linux/radix-tree.h> | |
a1857ebe | 27 | #include <linux/vmalloc.h> |
ed84885d | 28 | #include <linux/string.h> |
31db9f7c AB |
29 | |
30 | #include "send.h" | |
31 | #include "backref.h" | |
0b947aff | 32 | #include "hash.h" |
31db9f7c AB |
33 | #include "locking.h" |
34 | #include "disk-io.h" | |
35 | #include "btrfs_inode.h" | |
36 | #include "transaction.h" | |
37 | ||
38 | static int g_verbose = 0; | |
39 | ||
40 | #define verbose_printk(...) if (g_verbose) printk(__VA_ARGS__) | |
41 | ||
42 | /* | |
43 | * A fs_path is a helper to dynamically build path names with unknown size. | |
44 | * It reallocates the internal buffer on demand. | |
45 | * It allows fast adding of path elements on the right side (normal path) and | |
46 | * fast adding to the left side (reversed path). A reversed path can also be | |
47 | * unreversed if needed. | |
48 | */ | |
49 | struct fs_path { | |
50 | union { | |
51 | struct { | |
52 | char *start; | |
53 | char *end; | |
31db9f7c AB |
54 | |
55 | char *buf; | |
1f5a7ff9 DS |
56 | unsigned short buf_len:15; |
57 | unsigned short reversed:1; | |
31db9f7c AB |
58 | char inline_buf[]; |
59 | }; | |
ace01050 DS |
60 | /* |
61 | * Average path length does not exceed 200 bytes, we'll have | |
62 | * better packing in the slab and higher chance to satisfy | |
63 | * a allocation later during send. | |
64 | */ | |
65 | char pad[256]; | |
31db9f7c AB |
66 | }; |
67 | }; | |
68 | #define FS_PATH_INLINE_SIZE \ | |
69 | (sizeof(struct fs_path) - offsetof(struct fs_path, inline_buf)) | |
70 | ||
71 | ||
72 | /* reused for each extent */ | |
73 | struct clone_root { | |
74 | struct btrfs_root *root; | |
75 | u64 ino; | |
76 | u64 offset; | |
77 | ||
78 | u64 found_refs; | |
79 | }; | |
80 | ||
81 | #define SEND_CTX_MAX_NAME_CACHE_SIZE 128 | |
82 | #define SEND_CTX_NAME_CACHE_CLEAN_SIZE (SEND_CTX_MAX_NAME_CACHE_SIZE * 2) | |
83 | ||
84 | struct send_ctx { | |
85 | struct file *send_filp; | |
86 | loff_t send_off; | |
87 | char *send_buf; | |
88 | u32 send_size; | |
89 | u32 send_max_size; | |
90 | u64 total_send_size; | |
91 | u64 cmd_send_size[BTRFS_SEND_C_MAX + 1]; | |
cb95e7bf | 92 | u64 flags; /* 'flags' member of btrfs_ioctl_send_args is u64 */ |
31db9f7c | 93 | |
31db9f7c AB |
94 | struct btrfs_root *send_root; |
95 | struct btrfs_root *parent_root; | |
96 | struct clone_root *clone_roots; | |
97 | int clone_roots_cnt; | |
98 | ||
99 | /* current state of the compare_tree call */ | |
100 | struct btrfs_path *left_path; | |
101 | struct btrfs_path *right_path; | |
102 | struct btrfs_key *cmp_key; | |
103 | ||
104 | /* | |
105 | * infos of the currently processed inode. In case of deleted inodes, | |
106 | * these are the values from the deleted inode. | |
107 | */ | |
108 | u64 cur_ino; | |
109 | u64 cur_inode_gen; | |
110 | int cur_inode_new; | |
111 | int cur_inode_new_gen; | |
112 | int cur_inode_deleted; | |
31db9f7c AB |
113 | u64 cur_inode_size; |
114 | u64 cur_inode_mode; | |
644d1940 | 115 | u64 cur_inode_rdev; |
16e7549f | 116 | u64 cur_inode_last_extent; |
31db9f7c AB |
117 | |
118 | u64 send_progress; | |
119 | ||
120 | struct list_head new_refs; | |
121 | struct list_head deleted_refs; | |
122 | ||
123 | struct radix_tree_root name_cache; | |
124 | struct list_head name_cache_list; | |
125 | int name_cache_size; | |
126 | ||
2131bcd3 LB |
127 | struct file_ra_state ra; |
128 | ||
31db9f7c | 129 | char *read_buf; |
9f03740a FDBM |
130 | |
131 | /* | |
132 | * We process inodes by their increasing order, so if before an | |
133 | * incremental send we reverse the parent/child relationship of | |
134 | * directories such that a directory with a lower inode number was | |
135 | * the parent of a directory with a higher inode number, and the one | |
136 | * becoming the new parent got renamed too, we can't rename/move the | |
137 | * directory with lower inode number when we finish processing it - we | |
138 | * must process the directory with higher inode number first, then | |
139 | * rename/move it and then rename/move the directory with lower inode | |
140 | * number. Example follows. | |
141 | * | |
142 | * Tree state when the first send was performed: | |
143 | * | |
144 | * . | |
145 | * |-- a (ino 257) | |
146 | * |-- b (ino 258) | |
147 | * | | |
148 | * | | |
149 | * |-- c (ino 259) | |
150 | * | |-- d (ino 260) | |
151 | * | | |
152 | * |-- c2 (ino 261) | |
153 | * | |
154 | * Tree state when the second (incremental) send is performed: | |
155 | * | |
156 | * . | |
157 | * |-- a (ino 257) | |
158 | * |-- b (ino 258) | |
159 | * |-- c2 (ino 261) | |
160 | * |-- d2 (ino 260) | |
161 | * |-- cc (ino 259) | |
162 | * | |
163 | * The sequence of steps that lead to the second state was: | |
164 | * | |
165 | * mv /a/b/c/d /a/b/c2/d2 | |
166 | * mv /a/b/c /a/b/c2/d2/cc | |
167 | * | |
168 | * "c" has lower inode number, but we can't move it (2nd mv operation) | |
169 | * before we move "d", which has higher inode number. | |
170 | * | |
171 | * So we just memorize which move/rename operations must be performed | |
172 | * later when their respective parent is processed and moved/renamed. | |
173 | */ | |
174 | ||
175 | /* Indexed by parent directory inode number. */ | |
176 | struct rb_root pending_dir_moves; | |
177 | ||
178 | /* | |
179 | * Reverse index, indexed by the inode number of a directory that | |
180 | * is waiting for the move/rename of its immediate parent before its | |
181 | * own move/rename can be performed. | |
182 | */ | |
183 | struct rb_root waiting_dir_moves; | |
9dc44214 FM |
184 | |
185 | /* | |
186 | * A directory that is going to be rm'ed might have a child directory | |
187 | * which is in the pending directory moves index above. In this case, | |
188 | * the directory can only be removed after the move/rename of its child | |
189 | * is performed. Example: | |
190 | * | |
191 | * Parent snapshot: | |
192 | * | |
193 | * . (ino 256) | |
194 | * |-- a/ (ino 257) | |
195 | * |-- b/ (ino 258) | |
196 | * |-- c/ (ino 259) | |
197 | * | |-- x/ (ino 260) | |
198 | * | | |
199 | * |-- y/ (ino 261) | |
200 | * | |
201 | * Send snapshot: | |
202 | * | |
203 | * . (ino 256) | |
204 | * |-- a/ (ino 257) | |
205 | * |-- b/ (ino 258) | |
206 | * |-- YY/ (ino 261) | |
207 | * |-- x/ (ino 260) | |
208 | * | |
209 | * Sequence of steps that lead to the send snapshot: | |
210 | * rm -f /a/b/c/foo.txt | |
211 | * mv /a/b/y /a/b/YY | |
212 | * mv /a/b/c/x /a/b/YY | |
213 | * rmdir /a/b/c | |
214 | * | |
215 | * When the child is processed, its move/rename is delayed until its | |
216 | * parent is processed (as explained above), but all other operations | |
217 | * like update utimes, chown, chgrp, etc, are performed and the paths | |
218 | * that it uses for those operations must use the orphanized name of | |
219 | * its parent (the directory we're going to rm later), so we need to | |
220 | * memorize that name. | |
221 | * | |
222 | * Indexed by the inode number of the directory to be deleted. | |
223 | */ | |
224 | struct rb_root orphan_dirs; | |
9f03740a FDBM |
225 | }; |
226 | ||
227 | struct pending_dir_move { | |
228 | struct rb_node node; | |
229 | struct list_head list; | |
230 | u64 parent_ino; | |
231 | u64 ino; | |
232 | u64 gen; | |
84471e24 | 233 | bool is_orphan; |
9f03740a FDBM |
234 | struct list_head update_refs; |
235 | }; | |
236 | ||
237 | struct waiting_dir_move { | |
238 | struct rb_node node; | |
239 | u64 ino; | |
9dc44214 FM |
240 | /* |
241 | * There might be some directory that could not be removed because it | |
242 | * was waiting for this directory inode to be moved first. Therefore | |
243 | * after this directory is moved, we can try to rmdir the ino rmdir_ino. | |
244 | */ | |
245 | u64 rmdir_ino; | |
8b191a68 | 246 | bool orphanized; |
9dc44214 FM |
247 | }; |
248 | ||
249 | struct orphan_dir_info { | |
250 | struct rb_node node; | |
251 | u64 ino; | |
252 | u64 gen; | |
31db9f7c AB |
253 | }; |
254 | ||
255 | struct name_cache_entry { | |
256 | struct list_head list; | |
7e0926fe AB |
257 | /* |
258 | * radix_tree has only 32bit entries but we need to handle 64bit inums. | |
259 | * We use the lower 32bit of the 64bit inum to store it in the tree. If | |
260 | * more then one inum would fall into the same entry, we use radix_list | |
261 | * to store the additional entries. radix_list is also used to store | |
262 | * entries where two entries have the same inum but different | |
263 | * generations. | |
264 | */ | |
265 | struct list_head radix_list; | |
31db9f7c AB |
266 | u64 ino; |
267 | u64 gen; | |
268 | u64 parent_ino; | |
269 | u64 parent_gen; | |
270 | int ret; | |
271 | int need_later_update; | |
272 | int name_len; | |
273 | char name[]; | |
274 | }; | |
275 | ||
9f03740a FDBM |
276 | static int is_waiting_for_move(struct send_ctx *sctx, u64 ino); |
277 | ||
9dc44214 FM |
278 | static struct waiting_dir_move * |
279 | get_waiting_dir_move(struct send_ctx *sctx, u64 ino); | |
280 | ||
281 | static int is_waiting_for_rm(struct send_ctx *sctx, u64 dir_ino); | |
282 | ||
16e7549f JB |
283 | static int need_send_hole(struct send_ctx *sctx) |
284 | { | |
285 | return (sctx->parent_root && !sctx->cur_inode_new && | |
286 | !sctx->cur_inode_new_gen && !sctx->cur_inode_deleted && | |
287 | S_ISREG(sctx->cur_inode_mode)); | |
288 | } | |
289 | ||
31db9f7c AB |
290 | static void fs_path_reset(struct fs_path *p) |
291 | { | |
292 | if (p->reversed) { | |
293 | p->start = p->buf + p->buf_len - 1; | |
294 | p->end = p->start; | |
295 | *p->start = 0; | |
296 | } else { | |
297 | p->start = p->buf; | |
298 | p->end = p->start; | |
299 | *p->start = 0; | |
300 | } | |
301 | } | |
302 | ||
924794c9 | 303 | static struct fs_path *fs_path_alloc(void) |
31db9f7c AB |
304 | { |
305 | struct fs_path *p; | |
306 | ||
307 | p = kmalloc(sizeof(*p), GFP_NOFS); | |
308 | if (!p) | |
309 | return NULL; | |
310 | p->reversed = 0; | |
31db9f7c AB |
311 | p->buf = p->inline_buf; |
312 | p->buf_len = FS_PATH_INLINE_SIZE; | |
313 | fs_path_reset(p); | |
314 | return p; | |
315 | } | |
316 | ||
924794c9 | 317 | static struct fs_path *fs_path_alloc_reversed(void) |
31db9f7c AB |
318 | { |
319 | struct fs_path *p; | |
320 | ||
924794c9 | 321 | p = fs_path_alloc(); |
31db9f7c AB |
322 | if (!p) |
323 | return NULL; | |
324 | p->reversed = 1; | |
325 | fs_path_reset(p); | |
326 | return p; | |
327 | } | |
328 | ||
924794c9 | 329 | static void fs_path_free(struct fs_path *p) |
31db9f7c AB |
330 | { |
331 | if (!p) | |
332 | return; | |
ace01050 DS |
333 | if (p->buf != p->inline_buf) |
334 | kfree(p->buf); | |
31db9f7c AB |
335 | kfree(p); |
336 | } | |
337 | ||
338 | static int fs_path_len(struct fs_path *p) | |
339 | { | |
340 | return p->end - p->start; | |
341 | } | |
342 | ||
343 | static int fs_path_ensure_buf(struct fs_path *p, int len) | |
344 | { | |
345 | char *tmp_buf; | |
346 | int path_len; | |
347 | int old_buf_len; | |
348 | ||
349 | len++; | |
350 | ||
351 | if (p->buf_len >= len) | |
352 | return 0; | |
353 | ||
cfd4a535 CM |
354 | if (len > PATH_MAX) { |
355 | WARN_ON(1); | |
356 | return -ENOMEM; | |
357 | } | |
358 | ||
1b2782c8 DS |
359 | path_len = p->end - p->start; |
360 | old_buf_len = p->buf_len; | |
361 | ||
ace01050 DS |
362 | /* |
363 | * First time the inline_buf does not suffice | |
364 | */ | |
01a9a8a9 | 365 | if (p->buf == p->inline_buf) { |
9c9ca00b | 366 | tmp_buf = kmalloc(len, GFP_NOFS); |
01a9a8a9 FM |
367 | if (tmp_buf) |
368 | memcpy(tmp_buf, p->buf, old_buf_len); | |
369 | } else { | |
9c9ca00b | 370 | tmp_buf = krealloc(p->buf, len, GFP_NOFS); |
01a9a8a9 | 371 | } |
9c9ca00b DS |
372 | if (!tmp_buf) |
373 | return -ENOMEM; | |
374 | p->buf = tmp_buf; | |
375 | /* | |
376 | * The real size of the buffer is bigger, this will let the fast path | |
377 | * happen most of the time | |
378 | */ | |
379 | p->buf_len = ksize(p->buf); | |
ace01050 | 380 | |
31db9f7c AB |
381 | if (p->reversed) { |
382 | tmp_buf = p->buf + old_buf_len - path_len - 1; | |
383 | p->end = p->buf + p->buf_len - 1; | |
384 | p->start = p->end - path_len; | |
385 | memmove(p->start, tmp_buf, path_len + 1); | |
386 | } else { | |
387 | p->start = p->buf; | |
388 | p->end = p->start + path_len; | |
389 | } | |
390 | return 0; | |
391 | } | |
392 | ||
b23ab57d DS |
393 | static int fs_path_prepare_for_add(struct fs_path *p, int name_len, |
394 | char **prepared) | |
31db9f7c AB |
395 | { |
396 | int ret; | |
397 | int new_len; | |
398 | ||
399 | new_len = p->end - p->start + name_len; | |
400 | if (p->start != p->end) | |
401 | new_len++; | |
402 | ret = fs_path_ensure_buf(p, new_len); | |
403 | if (ret < 0) | |
404 | goto out; | |
405 | ||
406 | if (p->reversed) { | |
407 | if (p->start != p->end) | |
408 | *--p->start = '/'; | |
409 | p->start -= name_len; | |
b23ab57d | 410 | *prepared = p->start; |
31db9f7c AB |
411 | } else { |
412 | if (p->start != p->end) | |
413 | *p->end++ = '/'; | |
b23ab57d | 414 | *prepared = p->end; |
31db9f7c AB |
415 | p->end += name_len; |
416 | *p->end = 0; | |
417 | } | |
418 | ||
419 | out: | |
420 | return ret; | |
421 | } | |
422 | ||
423 | static int fs_path_add(struct fs_path *p, const char *name, int name_len) | |
424 | { | |
425 | int ret; | |
b23ab57d | 426 | char *prepared; |
31db9f7c | 427 | |
b23ab57d | 428 | ret = fs_path_prepare_for_add(p, name_len, &prepared); |
31db9f7c AB |
429 | if (ret < 0) |
430 | goto out; | |
b23ab57d | 431 | memcpy(prepared, name, name_len); |
31db9f7c AB |
432 | |
433 | out: | |
434 | return ret; | |
435 | } | |
436 | ||
437 | static int fs_path_add_path(struct fs_path *p, struct fs_path *p2) | |
438 | { | |
439 | int ret; | |
b23ab57d | 440 | char *prepared; |
31db9f7c | 441 | |
b23ab57d | 442 | ret = fs_path_prepare_for_add(p, p2->end - p2->start, &prepared); |
31db9f7c AB |
443 | if (ret < 0) |
444 | goto out; | |
b23ab57d | 445 | memcpy(prepared, p2->start, p2->end - p2->start); |
31db9f7c AB |
446 | |
447 | out: | |
448 | return ret; | |
449 | } | |
450 | ||
451 | static int fs_path_add_from_extent_buffer(struct fs_path *p, | |
452 | struct extent_buffer *eb, | |
453 | unsigned long off, int len) | |
454 | { | |
455 | int ret; | |
b23ab57d | 456 | char *prepared; |
31db9f7c | 457 | |
b23ab57d | 458 | ret = fs_path_prepare_for_add(p, len, &prepared); |
31db9f7c AB |
459 | if (ret < 0) |
460 | goto out; | |
461 | ||
b23ab57d | 462 | read_extent_buffer(eb, prepared, off, len); |
31db9f7c AB |
463 | |
464 | out: | |
465 | return ret; | |
466 | } | |
467 | ||
31db9f7c AB |
468 | static int fs_path_copy(struct fs_path *p, struct fs_path *from) |
469 | { | |
470 | int ret; | |
471 | ||
472 | p->reversed = from->reversed; | |
473 | fs_path_reset(p); | |
474 | ||
475 | ret = fs_path_add_path(p, from); | |
476 | ||
477 | return ret; | |
478 | } | |
479 | ||
480 | ||
481 | static void fs_path_unreverse(struct fs_path *p) | |
482 | { | |
483 | char *tmp; | |
484 | int len; | |
485 | ||
486 | if (!p->reversed) | |
487 | return; | |
488 | ||
489 | tmp = p->start; | |
490 | len = p->end - p->start; | |
491 | p->start = p->buf; | |
492 | p->end = p->start + len; | |
493 | memmove(p->start, tmp, len + 1); | |
494 | p->reversed = 0; | |
495 | } | |
496 | ||
497 | static struct btrfs_path *alloc_path_for_send(void) | |
498 | { | |
499 | struct btrfs_path *path; | |
500 | ||
501 | path = btrfs_alloc_path(); | |
502 | if (!path) | |
503 | return NULL; | |
504 | path->search_commit_root = 1; | |
505 | path->skip_locking = 1; | |
3f8a18cc | 506 | path->need_commit_sem = 1; |
31db9f7c AB |
507 | return path; |
508 | } | |
509 | ||
48a3b636 | 510 | static int write_buf(struct file *filp, const void *buf, u32 len, loff_t *off) |
31db9f7c AB |
511 | { |
512 | int ret; | |
513 | mm_segment_t old_fs; | |
514 | u32 pos = 0; | |
515 | ||
516 | old_fs = get_fs(); | |
517 | set_fs(KERNEL_DS); | |
518 | ||
519 | while (pos < len) { | |
d447d0da FF |
520 | ret = vfs_write(filp, (__force const char __user *)buf + pos, |
521 | len - pos, off); | |
31db9f7c AB |
522 | /* TODO handle that correctly */ |
523 | /*if (ret == -ERESTARTSYS) { | |
524 | continue; | |
525 | }*/ | |
526 | if (ret < 0) | |
527 | goto out; | |
528 | if (ret == 0) { | |
529 | ret = -EIO; | |
530 | goto out; | |
531 | } | |
532 | pos += ret; | |
533 | } | |
534 | ||
535 | ret = 0; | |
536 | ||
537 | out: | |
538 | set_fs(old_fs); | |
539 | return ret; | |
540 | } | |
541 | ||
542 | static int tlv_put(struct send_ctx *sctx, u16 attr, const void *data, int len) | |
543 | { | |
544 | struct btrfs_tlv_header *hdr; | |
545 | int total_len = sizeof(*hdr) + len; | |
546 | int left = sctx->send_max_size - sctx->send_size; | |
547 | ||
548 | if (unlikely(left < total_len)) | |
549 | return -EOVERFLOW; | |
550 | ||
551 | hdr = (struct btrfs_tlv_header *) (sctx->send_buf + sctx->send_size); | |
552 | hdr->tlv_type = cpu_to_le16(attr); | |
553 | hdr->tlv_len = cpu_to_le16(len); | |
554 | memcpy(hdr + 1, data, len); | |
555 | sctx->send_size += total_len; | |
556 | ||
557 | return 0; | |
558 | } | |
559 | ||
95bc79d5 DS |
560 | #define TLV_PUT_DEFINE_INT(bits) \ |
561 | static int tlv_put_u##bits(struct send_ctx *sctx, \ | |
562 | u##bits attr, u##bits value) \ | |
563 | { \ | |
564 | __le##bits __tmp = cpu_to_le##bits(value); \ | |
565 | return tlv_put(sctx, attr, &__tmp, sizeof(__tmp)); \ | |
566 | } | |
31db9f7c | 567 | |
95bc79d5 | 568 | TLV_PUT_DEFINE_INT(64) |
31db9f7c AB |
569 | |
570 | static int tlv_put_string(struct send_ctx *sctx, u16 attr, | |
571 | const char *str, int len) | |
572 | { | |
573 | if (len == -1) | |
574 | len = strlen(str); | |
575 | return tlv_put(sctx, attr, str, len); | |
576 | } | |
577 | ||
578 | static int tlv_put_uuid(struct send_ctx *sctx, u16 attr, | |
579 | const u8 *uuid) | |
580 | { | |
581 | return tlv_put(sctx, attr, uuid, BTRFS_UUID_SIZE); | |
582 | } | |
583 | ||
31db9f7c AB |
584 | static int tlv_put_btrfs_timespec(struct send_ctx *sctx, u16 attr, |
585 | struct extent_buffer *eb, | |
586 | struct btrfs_timespec *ts) | |
587 | { | |
588 | struct btrfs_timespec bts; | |
589 | read_extent_buffer(eb, &bts, (unsigned long)ts, sizeof(bts)); | |
590 | return tlv_put(sctx, attr, &bts, sizeof(bts)); | |
591 | } | |
592 | ||
593 | ||
594 | #define TLV_PUT(sctx, attrtype, attrlen, data) \ | |
595 | do { \ | |
596 | ret = tlv_put(sctx, attrtype, attrlen, data); \ | |
597 | if (ret < 0) \ | |
598 | goto tlv_put_failure; \ | |
599 | } while (0) | |
600 | ||
601 | #define TLV_PUT_INT(sctx, attrtype, bits, value) \ | |
602 | do { \ | |
603 | ret = tlv_put_u##bits(sctx, attrtype, value); \ | |
604 | if (ret < 0) \ | |
605 | goto tlv_put_failure; \ | |
606 | } while (0) | |
607 | ||
608 | #define TLV_PUT_U8(sctx, attrtype, data) TLV_PUT_INT(sctx, attrtype, 8, data) | |
609 | #define TLV_PUT_U16(sctx, attrtype, data) TLV_PUT_INT(sctx, attrtype, 16, data) | |
610 | #define TLV_PUT_U32(sctx, attrtype, data) TLV_PUT_INT(sctx, attrtype, 32, data) | |
611 | #define TLV_PUT_U64(sctx, attrtype, data) TLV_PUT_INT(sctx, attrtype, 64, data) | |
612 | #define TLV_PUT_STRING(sctx, attrtype, str, len) \ | |
613 | do { \ | |
614 | ret = tlv_put_string(sctx, attrtype, str, len); \ | |
615 | if (ret < 0) \ | |
616 | goto tlv_put_failure; \ | |
617 | } while (0) | |
618 | #define TLV_PUT_PATH(sctx, attrtype, p) \ | |
619 | do { \ | |
620 | ret = tlv_put_string(sctx, attrtype, p->start, \ | |
621 | p->end - p->start); \ | |
622 | if (ret < 0) \ | |
623 | goto tlv_put_failure; \ | |
624 | } while(0) | |
625 | #define TLV_PUT_UUID(sctx, attrtype, uuid) \ | |
626 | do { \ | |
627 | ret = tlv_put_uuid(sctx, attrtype, uuid); \ | |
628 | if (ret < 0) \ | |
629 | goto tlv_put_failure; \ | |
630 | } while (0) | |
31db9f7c AB |
631 | #define TLV_PUT_BTRFS_TIMESPEC(sctx, attrtype, eb, ts) \ |
632 | do { \ | |
633 | ret = tlv_put_btrfs_timespec(sctx, attrtype, eb, ts); \ | |
634 | if (ret < 0) \ | |
635 | goto tlv_put_failure; \ | |
636 | } while (0) | |
637 | ||
638 | static int send_header(struct send_ctx *sctx) | |
639 | { | |
640 | struct btrfs_stream_header hdr; | |
641 | ||
642 | strcpy(hdr.magic, BTRFS_SEND_STREAM_MAGIC); | |
643 | hdr.version = cpu_to_le32(BTRFS_SEND_STREAM_VERSION); | |
644 | ||
1bcea355 AJ |
645 | return write_buf(sctx->send_filp, &hdr, sizeof(hdr), |
646 | &sctx->send_off); | |
31db9f7c AB |
647 | } |
648 | ||
649 | /* | |
650 | * For each command/item we want to send to userspace, we call this function. | |
651 | */ | |
652 | static int begin_cmd(struct send_ctx *sctx, int cmd) | |
653 | { | |
654 | struct btrfs_cmd_header *hdr; | |
655 | ||
fae7f21c | 656 | if (WARN_ON(!sctx->send_buf)) |
31db9f7c | 657 | return -EINVAL; |
31db9f7c AB |
658 | |
659 | BUG_ON(sctx->send_size); | |
660 | ||
661 | sctx->send_size += sizeof(*hdr); | |
662 | hdr = (struct btrfs_cmd_header *)sctx->send_buf; | |
663 | hdr->cmd = cpu_to_le16(cmd); | |
664 | ||
665 | return 0; | |
666 | } | |
667 | ||
668 | static int send_cmd(struct send_ctx *sctx) | |
669 | { | |
670 | int ret; | |
671 | struct btrfs_cmd_header *hdr; | |
672 | u32 crc; | |
673 | ||
674 | hdr = (struct btrfs_cmd_header *)sctx->send_buf; | |
675 | hdr->len = cpu_to_le32(sctx->send_size - sizeof(*hdr)); | |
676 | hdr->crc = 0; | |
677 | ||
0b947aff | 678 | crc = btrfs_crc32c(0, (unsigned char *)sctx->send_buf, sctx->send_size); |
31db9f7c AB |
679 | hdr->crc = cpu_to_le32(crc); |
680 | ||
1bcea355 AJ |
681 | ret = write_buf(sctx->send_filp, sctx->send_buf, sctx->send_size, |
682 | &sctx->send_off); | |
31db9f7c AB |
683 | |
684 | sctx->total_send_size += sctx->send_size; | |
685 | sctx->cmd_send_size[le16_to_cpu(hdr->cmd)] += sctx->send_size; | |
686 | sctx->send_size = 0; | |
687 | ||
688 | return ret; | |
689 | } | |
690 | ||
691 | /* | |
692 | * Sends a move instruction to user space | |
693 | */ | |
694 | static int send_rename(struct send_ctx *sctx, | |
695 | struct fs_path *from, struct fs_path *to) | |
696 | { | |
697 | int ret; | |
698 | ||
699 | verbose_printk("btrfs: send_rename %s -> %s\n", from->start, to->start); | |
700 | ||
701 | ret = begin_cmd(sctx, BTRFS_SEND_C_RENAME); | |
702 | if (ret < 0) | |
703 | goto out; | |
704 | ||
705 | TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, from); | |
706 | TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH_TO, to); | |
707 | ||
708 | ret = send_cmd(sctx); | |
709 | ||
710 | tlv_put_failure: | |
711 | out: | |
712 | return ret; | |
713 | } | |
714 | ||
715 | /* | |
716 | * Sends a link instruction to user space | |
717 | */ | |
718 | static int send_link(struct send_ctx *sctx, | |
719 | struct fs_path *path, struct fs_path *lnk) | |
720 | { | |
721 | int ret; | |
722 | ||
723 | verbose_printk("btrfs: send_link %s -> %s\n", path->start, lnk->start); | |
724 | ||
725 | ret = begin_cmd(sctx, BTRFS_SEND_C_LINK); | |
726 | if (ret < 0) | |
727 | goto out; | |
728 | ||
729 | TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, path); | |
730 | TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH_LINK, lnk); | |
731 | ||
732 | ret = send_cmd(sctx); | |
733 | ||
734 | tlv_put_failure: | |
735 | out: | |
736 | return ret; | |
737 | } | |
738 | ||
739 | /* | |
740 | * Sends an unlink instruction to user space | |
741 | */ | |
742 | static int send_unlink(struct send_ctx *sctx, struct fs_path *path) | |
743 | { | |
744 | int ret; | |
745 | ||
746 | verbose_printk("btrfs: send_unlink %s\n", path->start); | |
747 | ||
748 | ret = begin_cmd(sctx, BTRFS_SEND_C_UNLINK); | |
749 | if (ret < 0) | |
750 | goto out; | |
751 | ||
752 | TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, path); | |
753 | ||
754 | ret = send_cmd(sctx); | |
755 | ||
756 | tlv_put_failure: | |
757 | out: | |
758 | return ret; | |
759 | } | |
760 | ||
761 | /* | |
762 | * Sends a rmdir instruction to user space | |
763 | */ | |
764 | static int send_rmdir(struct send_ctx *sctx, struct fs_path *path) | |
765 | { | |
766 | int ret; | |
767 | ||
768 | verbose_printk("btrfs: send_rmdir %s\n", path->start); | |
769 | ||
770 | ret = begin_cmd(sctx, BTRFS_SEND_C_RMDIR); | |
771 | if (ret < 0) | |
772 | goto out; | |
773 | ||
774 | TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, path); | |
775 | ||
776 | ret = send_cmd(sctx); | |
777 | ||
778 | tlv_put_failure: | |
779 | out: | |
780 | return ret; | |
781 | } | |
782 | ||
783 | /* | |
784 | * Helper function to retrieve some fields from an inode item. | |
785 | */ | |
3f8a18cc JB |
786 | static int __get_inode_info(struct btrfs_root *root, struct btrfs_path *path, |
787 | u64 ino, u64 *size, u64 *gen, u64 *mode, u64 *uid, | |
788 | u64 *gid, u64 *rdev) | |
31db9f7c AB |
789 | { |
790 | int ret; | |
791 | struct btrfs_inode_item *ii; | |
792 | struct btrfs_key key; | |
31db9f7c AB |
793 | |
794 | key.objectid = ino; | |
795 | key.type = BTRFS_INODE_ITEM_KEY; | |
796 | key.offset = 0; | |
797 | ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); | |
31db9f7c | 798 | if (ret) { |
3f8a18cc JB |
799 | if (ret > 0) |
800 | ret = -ENOENT; | |
801 | return ret; | |
31db9f7c AB |
802 | } |
803 | ||
804 | ii = btrfs_item_ptr(path->nodes[0], path->slots[0], | |
805 | struct btrfs_inode_item); | |
806 | if (size) | |
807 | *size = btrfs_inode_size(path->nodes[0], ii); | |
808 | if (gen) | |
809 | *gen = btrfs_inode_generation(path->nodes[0], ii); | |
810 | if (mode) | |
811 | *mode = btrfs_inode_mode(path->nodes[0], ii); | |
812 | if (uid) | |
813 | *uid = btrfs_inode_uid(path->nodes[0], ii); | |
814 | if (gid) | |
815 | *gid = btrfs_inode_gid(path->nodes[0], ii); | |
85a7b33b AB |
816 | if (rdev) |
817 | *rdev = btrfs_inode_rdev(path->nodes[0], ii); | |
31db9f7c | 818 | |
3f8a18cc JB |
819 | return ret; |
820 | } | |
821 | ||
822 | static int get_inode_info(struct btrfs_root *root, | |
823 | u64 ino, u64 *size, u64 *gen, | |
824 | u64 *mode, u64 *uid, u64 *gid, | |
825 | u64 *rdev) | |
826 | { | |
827 | struct btrfs_path *path; | |
828 | int ret; | |
829 | ||
830 | path = alloc_path_for_send(); | |
831 | if (!path) | |
832 | return -ENOMEM; | |
833 | ret = __get_inode_info(root, path, ino, size, gen, mode, uid, gid, | |
834 | rdev); | |
31db9f7c AB |
835 | btrfs_free_path(path); |
836 | return ret; | |
837 | } | |
838 | ||
839 | typedef int (*iterate_inode_ref_t)(int num, u64 dir, int index, | |
840 | struct fs_path *p, | |
841 | void *ctx); | |
842 | ||
843 | /* | |
96b5bd77 JS |
844 | * Helper function to iterate the entries in ONE btrfs_inode_ref or |
845 | * btrfs_inode_extref. | |
31db9f7c AB |
846 | * The iterate callback may return a non zero value to stop iteration. This can |
847 | * be a negative value for error codes or 1 to simply stop it. | |
848 | * | |
96b5bd77 | 849 | * path must point to the INODE_REF or INODE_EXTREF when called. |
31db9f7c | 850 | */ |
924794c9 | 851 | static int iterate_inode_ref(struct btrfs_root *root, struct btrfs_path *path, |
31db9f7c AB |
852 | struct btrfs_key *found_key, int resolve, |
853 | iterate_inode_ref_t iterate, void *ctx) | |
854 | { | |
96b5bd77 | 855 | struct extent_buffer *eb = path->nodes[0]; |
31db9f7c AB |
856 | struct btrfs_item *item; |
857 | struct btrfs_inode_ref *iref; | |
96b5bd77 | 858 | struct btrfs_inode_extref *extref; |
31db9f7c AB |
859 | struct btrfs_path *tmp_path; |
860 | struct fs_path *p; | |
96b5bd77 | 861 | u32 cur = 0; |
31db9f7c | 862 | u32 total; |
96b5bd77 | 863 | int slot = path->slots[0]; |
31db9f7c AB |
864 | u32 name_len; |
865 | char *start; | |
866 | int ret = 0; | |
96b5bd77 | 867 | int num = 0; |
31db9f7c | 868 | int index; |
96b5bd77 JS |
869 | u64 dir; |
870 | unsigned long name_off; | |
871 | unsigned long elem_size; | |
872 | unsigned long ptr; | |
31db9f7c | 873 | |
924794c9 | 874 | p = fs_path_alloc_reversed(); |
31db9f7c AB |
875 | if (!p) |
876 | return -ENOMEM; | |
877 | ||
878 | tmp_path = alloc_path_for_send(); | |
879 | if (!tmp_path) { | |
924794c9 | 880 | fs_path_free(p); |
31db9f7c AB |
881 | return -ENOMEM; |
882 | } | |
883 | ||
31db9f7c | 884 | |
96b5bd77 JS |
885 | if (found_key->type == BTRFS_INODE_REF_KEY) { |
886 | ptr = (unsigned long)btrfs_item_ptr(eb, slot, | |
887 | struct btrfs_inode_ref); | |
dd3cc16b | 888 | item = btrfs_item_nr(slot); |
96b5bd77 JS |
889 | total = btrfs_item_size(eb, item); |
890 | elem_size = sizeof(*iref); | |
891 | } else { | |
892 | ptr = btrfs_item_ptr_offset(eb, slot); | |
893 | total = btrfs_item_size_nr(eb, slot); | |
894 | elem_size = sizeof(*extref); | |
895 | } | |
896 | ||
31db9f7c AB |
897 | while (cur < total) { |
898 | fs_path_reset(p); | |
899 | ||
96b5bd77 JS |
900 | if (found_key->type == BTRFS_INODE_REF_KEY) { |
901 | iref = (struct btrfs_inode_ref *)(ptr + cur); | |
902 | name_len = btrfs_inode_ref_name_len(eb, iref); | |
903 | name_off = (unsigned long)(iref + 1); | |
904 | index = btrfs_inode_ref_index(eb, iref); | |
905 | dir = found_key->offset; | |
906 | } else { | |
907 | extref = (struct btrfs_inode_extref *)(ptr + cur); | |
908 | name_len = btrfs_inode_extref_name_len(eb, extref); | |
909 | name_off = (unsigned long)&extref->name; | |
910 | index = btrfs_inode_extref_index(eb, extref); | |
911 | dir = btrfs_inode_extref_parent(eb, extref); | |
912 | } | |
913 | ||
31db9f7c | 914 | if (resolve) { |
96b5bd77 JS |
915 | start = btrfs_ref_to_path(root, tmp_path, name_len, |
916 | name_off, eb, dir, | |
917 | p->buf, p->buf_len); | |
31db9f7c AB |
918 | if (IS_ERR(start)) { |
919 | ret = PTR_ERR(start); | |
920 | goto out; | |
921 | } | |
922 | if (start < p->buf) { | |
923 | /* overflow , try again with larger buffer */ | |
924 | ret = fs_path_ensure_buf(p, | |
925 | p->buf_len + p->buf - start); | |
926 | if (ret < 0) | |
927 | goto out; | |
96b5bd77 JS |
928 | start = btrfs_ref_to_path(root, tmp_path, |
929 | name_len, name_off, | |
930 | eb, dir, | |
931 | p->buf, p->buf_len); | |
31db9f7c AB |
932 | if (IS_ERR(start)) { |
933 | ret = PTR_ERR(start); | |
934 | goto out; | |
935 | } | |
936 | BUG_ON(start < p->buf); | |
937 | } | |
938 | p->start = start; | |
939 | } else { | |
96b5bd77 JS |
940 | ret = fs_path_add_from_extent_buffer(p, eb, name_off, |
941 | name_len); | |
31db9f7c AB |
942 | if (ret < 0) |
943 | goto out; | |
944 | } | |
945 | ||
96b5bd77 JS |
946 | cur += elem_size + name_len; |
947 | ret = iterate(num, dir, index, p, ctx); | |
31db9f7c AB |
948 | if (ret) |
949 | goto out; | |
31db9f7c AB |
950 | num++; |
951 | } | |
952 | ||
953 | out: | |
954 | btrfs_free_path(tmp_path); | |
924794c9 | 955 | fs_path_free(p); |
31db9f7c AB |
956 | return ret; |
957 | } | |
958 | ||
959 | typedef int (*iterate_dir_item_t)(int num, struct btrfs_key *di_key, | |
960 | const char *name, int name_len, | |
961 | const char *data, int data_len, | |
962 | u8 type, void *ctx); | |
963 | ||
964 | /* | |
965 | * Helper function to iterate the entries in ONE btrfs_dir_item. | |
966 | * The iterate callback may return a non zero value to stop iteration. This can | |
967 | * be a negative value for error codes or 1 to simply stop it. | |
968 | * | |
969 | * path must point to the dir item when called. | |
970 | */ | |
924794c9 | 971 | static int iterate_dir_item(struct btrfs_root *root, struct btrfs_path *path, |
31db9f7c AB |
972 | struct btrfs_key *found_key, |
973 | iterate_dir_item_t iterate, void *ctx) | |
974 | { | |
975 | int ret = 0; | |
976 | struct extent_buffer *eb; | |
977 | struct btrfs_item *item; | |
978 | struct btrfs_dir_item *di; | |
31db9f7c AB |
979 | struct btrfs_key di_key; |
980 | char *buf = NULL; | |
7e3ae33e | 981 | int buf_len; |
31db9f7c AB |
982 | u32 name_len; |
983 | u32 data_len; | |
984 | u32 cur; | |
985 | u32 len; | |
986 | u32 total; | |
987 | int slot; | |
988 | int num; | |
989 | u8 type; | |
990 | ||
4395e0c4 FM |
991 | /* |
992 | * Start with a small buffer (1 page). If later we end up needing more | |
993 | * space, which can happen for xattrs on a fs with a leaf size greater | |
994 | * then the page size, attempt to increase the buffer. Typically xattr | |
995 | * values are small. | |
996 | */ | |
997 | buf_len = PATH_MAX; | |
31db9f7c AB |
998 | buf = kmalloc(buf_len, GFP_NOFS); |
999 | if (!buf) { | |
1000 | ret = -ENOMEM; | |
1001 | goto out; | |
1002 | } | |
1003 | ||
31db9f7c AB |
1004 | eb = path->nodes[0]; |
1005 | slot = path->slots[0]; | |
dd3cc16b | 1006 | item = btrfs_item_nr(slot); |
31db9f7c AB |
1007 | di = btrfs_item_ptr(eb, slot, struct btrfs_dir_item); |
1008 | cur = 0; | |
1009 | len = 0; | |
1010 | total = btrfs_item_size(eb, item); | |
1011 | ||
1012 | num = 0; | |
1013 | while (cur < total) { | |
1014 | name_len = btrfs_dir_name_len(eb, di); | |
1015 | data_len = btrfs_dir_data_len(eb, di); | |
1016 | type = btrfs_dir_type(eb, di); | |
1017 | btrfs_dir_item_key_to_cpu(eb, di, &di_key); | |
1018 | ||
7e3ae33e FM |
1019 | if (type == BTRFS_FT_XATTR) { |
1020 | if (name_len > XATTR_NAME_MAX) { | |
1021 | ret = -ENAMETOOLONG; | |
1022 | goto out; | |
1023 | } | |
4395e0c4 | 1024 | if (name_len + data_len > BTRFS_MAX_XATTR_SIZE(root)) { |
7e3ae33e FM |
1025 | ret = -E2BIG; |
1026 | goto out; | |
1027 | } | |
1028 | } else { | |
1029 | /* | |
1030 | * Path too long | |
1031 | */ | |
4395e0c4 | 1032 | if (name_len + data_len > PATH_MAX) { |
7e3ae33e FM |
1033 | ret = -ENAMETOOLONG; |
1034 | goto out; | |
1035 | } | |
31db9f7c AB |
1036 | } |
1037 | ||
4395e0c4 FM |
1038 | if (name_len + data_len > buf_len) { |
1039 | buf_len = name_len + data_len; | |
1040 | if (is_vmalloc_addr(buf)) { | |
1041 | vfree(buf); | |
1042 | buf = NULL; | |
1043 | } else { | |
1044 | char *tmp = krealloc(buf, buf_len, | |
1045 | GFP_NOFS | __GFP_NOWARN); | |
1046 | ||
1047 | if (!tmp) | |
1048 | kfree(buf); | |
1049 | buf = tmp; | |
1050 | } | |
1051 | if (!buf) { | |
1052 | buf = vmalloc(buf_len); | |
1053 | if (!buf) { | |
1054 | ret = -ENOMEM; | |
1055 | goto out; | |
1056 | } | |
1057 | } | |
1058 | } | |
1059 | ||
31db9f7c AB |
1060 | read_extent_buffer(eb, buf, (unsigned long)(di + 1), |
1061 | name_len + data_len); | |
1062 | ||
1063 | len = sizeof(*di) + name_len + data_len; | |
1064 | di = (struct btrfs_dir_item *)((char *)di + len); | |
1065 | cur += len; | |
1066 | ||
1067 | ret = iterate(num, &di_key, buf, name_len, buf + name_len, | |
1068 | data_len, type, ctx); | |
1069 | if (ret < 0) | |
1070 | goto out; | |
1071 | if (ret) { | |
1072 | ret = 0; | |
1073 | goto out; | |
1074 | } | |
1075 | ||
1076 | num++; | |
1077 | } | |
1078 | ||
1079 | out: | |
4395e0c4 | 1080 | kvfree(buf); |
31db9f7c AB |
1081 | return ret; |
1082 | } | |
1083 | ||
1084 | static int __copy_first_ref(int num, u64 dir, int index, | |
1085 | struct fs_path *p, void *ctx) | |
1086 | { | |
1087 | int ret; | |
1088 | struct fs_path *pt = ctx; | |
1089 | ||
1090 | ret = fs_path_copy(pt, p); | |
1091 | if (ret < 0) | |
1092 | return ret; | |
1093 | ||
1094 | /* we want the first only */ | |
1095 | return 1; | |
1096 | } | |
1097 | ||
1098 | /* | |
1099 | * Retrieve the first path of an inode. If an inode has more then one | |
1100 | * ref/hardlink, this is ignored. | |
1101 | */ | |
924794c9 | 1102 | static int get_inode_path(struct btrfs_root *root, |
31db9f7c AB |
1103 | u64 ino, struct fs_path *path) |
1104 | { | |
1105 | int ret; | |
1106 | struct btrfs_key key, found_key; | |
1107 | struct btrfs_path *p; | |
1108 | ||
1109 | p = alloc_path_for_send(); | |
1110 | if (!p) | |
1111 | return -ENOMEM; | |
1112 | ||
1113 | fs_path_reset(path); | |
1114 | ||
1115 | key.objectid = ino; | |
1116 | key.type = BTRFS_INODE_REF_KEY; | |
1117 | key.offset = 0; | |
1118 | ||
1119 | ret = btrfs_search_slot_for_read(root, &key, p, 1, 0); | |
1120 | if (ret < 0) | |
1121 | goto out; | |
1122 | if (ret) { | |
1123 | ret = 1; | |
1124 | goto out; | |
1125 | } | |
1126 | btrfs_item_key_to_cpu(p->nodes[0], &found_key, p->slots[0]); | |
1127 | if (found_key.objectid != ino || | |
96b5bd77 JS |
1128 | (found_key.type != BTRFS_INODE_REF_KEY && |
1129 | found_key.type != BTRFS_INODE_EXTREF_KEY)) { | |
31db9f7c AB |
1130 | ret = -ENOENT; |
1131 | goto out; | |
1132 | } | |
1133 | ||
924794c9 TI |
1134 | ret = iterate_inode_ref(root, p, &found_key, 1, |
1135 | __copy_first_ref, path); | |
31db9f7c AB |
1136 | if (ret < 0) |
1137 | goto out; | |
1138 | ret = 0; | |
1139 | ||
1140 | out: | |
1141 | btrfs_free_path(p); | |
1142 | return ret; | |
1143 | } | |
1144 | ||
1145 | struct backref_ctx { | |
1146 | struct send_ctx *sctx; | |
1147 | ||
3f8a18cc | 1148 | struct btrfs_path *path; |
31db9f7c AB |
1149 | /* number of total found references */ |
1150 | u64 found; | |
1151 | ||
1152 | /* | |
1153 | * used for clones found in send_root. clones found behind cur_objectid | |
1154 | * and cur_offset are not considered as allowed clones. | |
1155 | */ | |
1156 | u64 cur_objectid; | |
1157 | u64 cur_offset; | |
1158 | ||
1159 | /* may be truncated in case it's the last extent in a file */ | |
1160 | u64 extent_len; | |
1161 | ||
619d8c4e FM |
1162 | /* data offset in the file extent item */ |
1163 | u64 data_offset; | |
1164 | ||
31db9f7c | 1165 | /* Just to check for bugs in backref resolving */ |
ee849c04 | 1166 | int found_itself; |
31db9f7c AB |
1167 | }; |
1168 | ||
1169 | static int __clone_root_cmp_bsearch(const void *key, const void *elt) | |
1170 | { | |
995e01b7 | 1171 | u64 root = (u64)(uintptr_t)key; |
31db9f7c AB |
1172 | struct clone_root *cr = (struct clone_root *)elt; |
1173 | ||
1174 | if (root < cr->root->objectid) | |
1175 | return -1; | |
1176 | if (root > cr->root->objectid) | |
1177 | return 1; | |
1178 | return 0; | |
1179 | } | |
1180 | ||
1181 | static int __clone_root_cmp_sort(const void *e1, const void *e2) | |
1182 | { | |
1183 | struct clone_root *cr1 = (struct clone_root *)e1; | |
1184 | struct clone_root *cr2 = (struct clone_root *)e2; | |
1185 | ||
1186 | if (cr1->root->objectid < cr2->root->objectid) | |
1187 | return -1; | |
1188 | if (cr1->root->objectid > cr2->root->objectid) | |
1189 | return 1; | |
1190 | return 0; | |
1191 | } | |
1192 | ||
1193 | /* | |
1194 | * Called for every backref that is found for the current extent. | |
766702ef | 1195 | * Results are collected in sctx->clone_roots->ino/offset/found_refs |
31db9f7c AB |
1196 | */ |
1197 | static int __iterate_backrefs(u64 ino, u64 offset, u64 root, void *ctx_) | |
1198 | { | |
1199 | struct backref_ctx *bctx = ctx_; | |
1200 | struct clone_root *found; | |
1201 | int ret; | |
1202 | u64 i_size; | |
1203 | ||
1204 | /* First check if the root is in the list of accepted clone sources */ | |
995e01b7 | 1205 | found = bsearch((void *)(uintptr_t)root, bctx->sctx->clone_roots, |
31db9f7c AB |
1206 | bctx->sctx->clone_roots_cnt, |
1207 | sizeof(struct clone_root), | |
1208 | __clone_root_cmp_bsearch); | |
1209 | if (!found) | |
1210 | return 0; | |
1211 | ||
1212 | if (found->root == bctx->sctx->send_root && | |
1213 | ino == bctx->cur_objectid && | |
1214 | offset == bctx->cur_offset) { | |
ee849c04 | 1215 | bctx->found_itself = 1; |
31db9f7c AB |
1216 | } |
1217 | ||
1218 | /* | |
766702ef | 1219 | * There are inodes that have extents that lie behind its i_size. Don't |
31db9f7c AB |
1220 | * accept clones from these extents. |
1221 | */ | |
3f8a18cc JB |
1222 | ret = __get_inode_info(found->root, bctx->path, ino, &i_size, NULL, NULL, |
1223 | NULL, NULL, NULL); | |
1224 | btrfs_release_path(bctx->path); | |
31db9f7c AB |
1225 | if (ret < 0) |
1226 | return ret; | |
1227 | ||
619d8c4e | 1228 | if (offset + bctx->data_offset + bctx->extent_len > i_size) |
31db9f7c AB |
1229 | return 0; |
1230 | ||
1231 | /* | |
1232 | * Make sure we don't consider clones from send_root that are | |
1233 | * behind the current inode/offset. | |
1234 | */ | |
1235 | if (found->root == bctx->sctx->send_root) { | |
1236 | /* | |
1237 | * TODO for the moment we don't accept clones from the inode | |
1238 | * that is currently send. We may change this when | |
1239 | * BTRFS_IOC_CLONE_RANGE supports cloning from and to the same | |
1240 | * file. | |
1241 | */ | |
1242 | if (ino >= bctx->cur_objectid) | |
1243 | return 0; | |
e938c8ad AB |
1244 | #if 0 |
1245 | if (ino > bctx->cur_objectid) | |
1246 | return 0; | |
1247 | if (offset + bctx->extent_len > bctx->cur_offset) | |
31db9f7c | 1248 | return 0; |
e938c8ad | 1249 | #endif |
31db9f7c AB |
1250 | } |
1251 | ||
1252 | bctx->found++; | |
1253 | found->found_refs++; | |
1254 | if (ino < found->ino) { | |
1255 | found->ino = ino; | |
1256 | found->offset = offset; | |
1257 | } else if (found->ino == ino) { | |
1258 | /* | |
1259 | * same extent found more then once in the same file. | |
1260 | */ | |
1261 | if (found->offset > offset + bctx->extent_len) | |
1262 | found->offset = offset; | |
1263 | } | |
1264 | ||
1265 | return 0; | |
1266 | } | |
1267 | ||
1268 | /* | |
766702ef AB |
1269 | * Given an inode, offset and extent item, it finds a good clone for a clone |
1270 | * instruction. Returns -ENOENT when none could be found. The function makes | |
1271 | * sure that the returned clone is usable at the point where sending is at the | |
1272 | * moment. This means, that no clones are accepted which lie behind the current | |
1273 | * inode+offset. | |
1274 | * | |
31db9f7c AB |
1275 | * path must point to the extent item when called. |
1276 | */ | |
1277 | static int find_extent_clone(struct send_ctx *sctx, | |
1278 | struct btrfs_path *path, | |
1279 | u64 ino, u64 data_offset, | |
1280 | u64 ino_size, | |
1281 | struct clone_root **found) | |
1282 | { | |
1283 | int ret; | |
1284 | int extent_type; | |
1285 | u64 logical; | |
74dd17fb | 1286 | u64 disk_byte; |
31db9f7c AB |
1287 | u64 num_bytes; |
1288 | u64 extent_item_pos; | |
69917e43 | 1289 | u64 flags = 0; |
31db9f7c AB |
1290 | struct btrfs_file_extent_item *fi; |
1291 | struct extent_buffer *eb = path->nodes[0]; | |
35075bb0 | 1292 | struct backref_ctx *backref_ctx = NULL; |
31db9f7c AB |
1293 | struct clone_root *cur_clone_root; |
1294 | struct btrfs_key found_key; | |
1295 | struct btrfs_path *tmp_path; | |
74dd17fb | 1296 | int compressed; |
31db9f7c AB |
1297 | u32 i; |
1298 | ||
1299 | tmp_path = alloc_path_for_send(); | |
1300 | if (!tmp_path) | |
1301 | return -ENOMEM; | |
1302 | ||
3f8a18cc JB |
1303 | /* We only use this path under the commit sem */ |
1304 | tmp_path->need_commit_sem = 0; | |
1305 | ||
35075bb0 AB |
1306 | backref_ctx = kmalloc(sizeof(*backref_ctx), GFP_NOFS); |
1307 | if (!backref_ctx) { | |
1308 | ret = -ENOMEM; | |
1309 | goto out; | |
1310 | } | |
1311 | ||
3f8a18cc JB |
1312 | backref_ctx->path = tmp_path; |
1313 | ||
31db9f7c AB |
1314 | if (data_offset >= ino_size) { |
1315 | /* | |
1316 | * There may be extents that lie behind the file's size. | |
1317 | * I at least had this in combination with snapshotting while | |
1318 | * writing large files. | |
1319 | */ | |
1320 | ret = 0; | |
1321 | goto out; | |
1322 | } | |
1323 | ||
1324 | fi = btrfs_item_ptr(eb, path->slots[0], | |
1325 | struct btrfs_file_extent_item); | |
1326 | extent_type = btrfs_file_extent_type(eb, fi); | |
1327 | if (extent_type == BTRFS_FILE_EXTENT_INLINE) { | |
1328 | ret = -ENOENT; | |
1329 | goto out; | |
1330 | } | |
74dd17fb | 1331 | compressed = btrfs_file_extent_compression(eb, fi); |
31db9f7c AB |
1332 | |
1333 | num_bytes = btrfs_file_extent_num_bytes(eb, fi); | |
74dd17fb CM |
1334 | disk_byte = btrfs_file_extent_disk_bytenr(eb, fi); |
1335 | if (disk_byte == 0) { | |
31db9f7c AB |
1336 | ret = -ENOENT; |
1337 | goto out; | |
1338 | } | |
74dd17fb | 1339 | logical = disk_byte + btrfs_file_extent_offset(eb, fi); |
31db9f7c | 1340 | |
9e351cc8 | 1341 | down_read(&sctx->send_root->fs_info->commit_root_sem); |
69917e43 LB |
1342 | ret = extent_from_logical(sctx->send_root->fs_info, disk_byte, tmp_path, |
1343 | &found_key, &flags); | |
9e351cc8 | 1344 | up_read(&sctx->send_root->fs_info->commit_root_sem); |
31db9f7c AB |
1345 | btrfs_release_path(tmp_path); |
1346 | ||
1347 | if (ret < 0) | |
1348 | goto out; | |
69917e43 | 1349 | if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) { |
31db9f7c AB |
1350 | ret = -EIO; |
1351 | goto out; | |
1352 | } | |
1353 | ||
1354 | /* | |
1355 | * Setup the clone roots. | |
1356 | */ | |
1357 | for (i = 0; i < sctx->clone_roots_cnt; i++) { | |
1358 | cur_clone_root = sctx->clone_roots + i; | |
1359 | cur_clone_root->ino = (u64)-1; | |
1360 | cur_clone_root->offset = 0; | |
1361 | cur_clone_root->found_refs = 0; | |
1362 | } | |
1363 | ||
35075bb0 AB |
1364 | backref_ctx->sctx = sctx; |
1365 | backref_ctx->found = 0; | |
1366 | backref_ctx->cur_objectid = ino; | |
1367 | backref_ctx->cur_offset = data_offset; | |
1368 | backref_ctx->found_itself = 0; | |
1369 | backref_ctx->extent_len = num_bytes; | |
619d8c4e FM |
1370 | /* |
1371 | * For non-compressed extents iterate_extent_inodes() gives us extent | |
1372 | * offsets that already take into account the data offset, but not for | |
1373 | * compressed extents, since the offset is logical and not relative to | |
1374 | * the physical extent locations. We must take this into account to | |
1375 | * avoid sending clone offsets that go beyond the source file's size, | |
1376 | * which would result in the clone ioctl failing with -EINVAL on the | |
1377 | * receiving end. | |
1378 | */ | |
1379 | if (compressed == BTRFS_COMPRESS_NONE) | |
1380 | backref_ctx->data_offset = 0; | |
1381 | else | |
1382 | backref_ctx->data_offset = btrfs_file_extent_offset(eb, fi); | |
31db9f7c AB |
1383 | |
1384 | /* | |
1385 | * The last extent of a file may be too large due to page alignment. | |
1386 | * We need to adjust extent_len in this case so that the checks in | |
1387 | * __iterate_backrefs work. | |
1388 | */ | |
1389 | if (data_offset + num_bytes >= ino_size) | |
35075bb0 | 1390 | backref_ctx->extent_len = ino_size - data_offset; |
31db9f7c AB |
1391 | |
1392 | /* | |
1393 | * Now collect all backrefs. | |
1394 | */ | |
74dd17fb CM |
1395 | if (compressed == BTRFS_COMPRESS_NONE) |
1396 | extent_item_pos = logical - found_key.objectid; | |
1397 | else | |
1398 | extent_item_pos = 0; | |
31db9f7c AB |
1399 | ret = iterate_extent_inodes(sctx->send_root->fs_info, |
1400 | found_key.objectid, extent_item_pos, 1, | |
35075bb0 | 1401 | __iterate_backrefs, backref_ctx); |
74dd17fb | 1402 | |
31db9f7c AB |
1403 | if (ret < 0) |
1404 | goto out; | |
1405 | ||
35075bb0 | 1406 | if (!backref_ctx->found_itself) { |
31db9f7c AB |
1407 | /* found a bug in backref code? */ |
1408 | ret = -EIO; | |
efe120a0 | 1409 | btrfs_err(sctx->send_root->fs_info, "did not find backref in " |
31db9f7c | 1410 | "send_root. inode=%llu, offset=%llu, " |
351fd353 | 1411 | "disk_byte=%llu found extent=%llu", |
74dd17fb | 1412 | ino, data_offset, disk_byte, found_key.objectid); |
31db9f7c AB |
1413 | goto out; |
1414 | } | |
1415 | ||
1416 | verbose_printk(KERN_DEBUG "btrfs: find_extent_clone: data_offset=%llu, " | |
1417 | "ino=%llu, " | |
1418 | "num_bytes=%llu, logical=%llu\n", | |
1419 | data_offset, ino, num_bytes, logical); | |
1420 | ||
35075bb0 | 1421 | if (!backref_ctx->found) |
31db9f7c AB |
1422 | verbose_printk("btrfs: no clones found\n"); |
1423 | ||
1424 | cur_clone_root = NULL; | |
1425 | for (i = 0; i < sctx->clone_roots_cnt; i++) { | |
1426 | if (sctx->clone_roots[i].found_refs) { | |
1427 | if (!cur_clone_root) | |
1428 | cur_clone_root = sctx->clone_roots + i; | |
1429 | else if (sctx->clone_roots[i].root == sctx->send_root) | |
1430 | /* prefer clones from send_root over others */ | |
1431 | cur_clone_root = sctx->clone_roots + i; | |
31db9f7c AB |
1432 | } |
1433 | ||
1434 | } | |
1435 | ||
1436 | if (cur_clone_root) { | |
1437 | *found = cur_clone_root; | |
1438 | ret = 0; | |
1439 | } else { | |
1440 | ret = -ENOENT; | |
1441 | } | |
1442 | ||
1443 | out: | |
1444 | btrfs_free_path(tmp_path); | |
35075bb0 | 1445 | kfree(backref_ctx); |
31db9f7c AB |
1446 | return ret; |
1447 | } | |
1448 | ||
924794c9 | 1449 | static int read_symlink(struct btrfs_root *root, |
31db9f7c AB |
1450 | u64 ino, |
1451 | struct fs_path *dest) | |
1452 | { | |
1453 | int ret; | |
1454 | struct btrfs_path *path; | |
1455 | struct btrfs_key key; | |
1456 | struct btrfs_file_extent_item *ei; | |
1457 | u8 type; | |
1458 | u8 compression; | |
1459 | unsigned long off; | |
1460 | int len; | |
1461 | ||
1462 | path = alloc_path_for_send(); | |
1463 | if (!path) | |
1464 | return -ENOMEM; | |
1465 | ||
1466 | key.objectid = ino; | |
1467 | key.type = BTRFS_EXTENT_DATA_KEY; | |
1468 | key.offset = 0; | |
1469 | ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); | |
1470 | if (ret < 0) | |
1471 | goto out; | |
1472 | BUG_ON(ret); | |
1473 | ||
1474 | ei = btrfs_item_ptr(path->nodes[0], path->slots[0], | |
1475 | struct btrfs_file_extent_item); | |
1476 | type = btrfs_file_extent_type(path->nodes[0], ei); | |
1477 | compression = btrfs_file_extent_compression(path->nodes[0], ei); | |
1478 | BUG_ON(type != BTRFS_FILE_EXTENT_INLINE); | |
1479 | BUG_ON(compression); | |
1480 | ||
1481 | off = btrfs_file_extent_inline_start(ei); | |
514ac8ad | 1482 | len = btrfs_file_extent_inline_len(path->nodes[0], path->slots[0], ei); |
31db9f7c AB |
1483 | |
1484 | ret = fs_path_add_from_extent_buffer(dest, path->nodes[0], off, len); | |
31db9f7c AB |
1485 | |
1486 | out: | |
1487 | btrfs_free_path(path); | |
1488 | return ret; | |
1489 | } | |
1490 | ||
1491 | /* | |
1492 | * Helper function to generate a file name that is unique in the root of | |
1493 | * send_root and parent_root. This is used to generate names for orphan inodes. | |
1494 | */ | |
1495 | static int gen_unique_name(struct send_ctx *sctx, | |
1496 | u64 ino, u64 gen, | |
1497 | struct fs_path *dest) | |
1498 | { | |
1499 | int ret = 0; | |
1500 | struct btrfs_path *path; | |
1501 | struct btrfs_dir_item *di; | |
1502 | char tmp[64]; | |
1503 | int len; | |
1504 | u64 idx = 0; | |
1505 | ||
1506 | path = alloc_path_for_send(); | |
1507 | if (!path) | |
1508 | return -ENOMEM; | |
1509 | ||
1510 | while (1) { | |
f74b86d8 | 1511 | len = snprintf(tmp, sizeof(tmp), "o%llu-%llu-%llu", |
31db9f7c | 1512 | ino, gen, idx); |
64792f25 | 1513 | ASSERT(len < sizeof(tmp)); |
31db9f7c AB |
1514 | |
1515 | di = btrfs_lookup_dir_item(NULL, sctx->send_root, | |
1516 | path, BTRFS_FIRST_FREE_OBJECTID, | |
1517 | tmp, strlen(tmp), 0); | |
1518 | btrfs_release_path(path); | |
1519 | if (IS_ERR(di)) { | |
1520 | ret = PTR_ERR(di); | |
1521 | goto out; | |
1522 | } | |
1523 | if (di) { | |
1524 | /* not unique, try again */ | |
1525 | idx++; | |
1526 | continue; | |
1527 | } | |
1528 | ||
1529 | if (!sctx->parent_root) { | |
1530 | /* unique */ | |
1531 | ret = 0; | |
1532 | break; | |
1533 | } | |
1534 | ||
1535 | di = btrfs_lookup_dir_item(NULL, sctx->parent_root, | |
1536 | path, BTRFS_FIRST_FREE_OBJECTID, | |
1537 | tmp, strlen(tmp), 0); | |
1538 | btrfs_release_path(path); | |
1539 | if (IS_ERR(di)) { | |
1540 | ret = PTR_ERR(di); | |
1541 | goto out; | |
1542 | } | |
1543 | if (di) { | |
1544 | /* not unique, try again */ | |
1545 | idx++; | |
1546 | continue; | |
1547 | } | |
1548 | /* unique */ | |
1549 | break; | |
1550 | } | |
1551 | ||
1552 | ret = fs_path_add(dest, tmp, strlen(tmp)); | |
1553 | ||
1554 | out: | |
1555 | btrfs_free_path(path); | |
1556 | return ret; | |
1557 | } | |
1558 | ||
1559 | enum inode_state { | |
1560 | inode_state_no_change, | |
1561 | inode_state_will_create, | |
1562 | inode_state_did_create, | |
1563 | inode_state_will_delete, | |
1564 | inode_state_did_delete, | |
1565 | }; | |
1566 | ||
1567 | static int get_cur_inode_state(struct send_ctx *sctx, u64 ino, u64 gen) | |
1568 | { | |
1569 | int ret; | |
1570 | int left_ret; | |
1571 | int right_ret; | |
1572 | u64 left_gen; | |
1573 | u64 right_gen; | |
1574 | ||
1575 | ret = get_inode_info(sctx->send_root, ino, NULL, &left_gen, NULL, NULL, | |
85a7b33b | 1576 | NULL, NULL); |
31db9f7c AB |
1577 | if (ret < 0 && ret != -ENOENT) |
1578 | goto out; | |
1579 | left_ret = ret; | |
1580 | ||
1581 | if (!sctx->parent_root) { | |
1582 | right_ret = -ENOENT; | |
1583 | } else { | |
1584 | ret = get_inode_info(sctx->parent_root, ino, NULL, &right_gen, | |
85a7b33b | 1585 | NULL, NULL, NULL, NULL); |
31db9f7c AB |
1586 | if (ret < 0 && ret != -ENOENT) |
1587 | goto out; | |
1588 | right_ret = ret; | |
1589 | } | |
1590 | ||
1591 | if (!left_ret && !right_ret) { | |
e938c8ad | 1592 | if (left_gen == gen && right_gen == gen) { |
31db9f7c | 1593 | ret = inode_state_no_change; |
e938c8ad | 1594 | } else if (left_gen == gen) { |
31db9f7c AB |
1595 | if (ino < sctx->send_progress) |
1596 | ret = inode_state_did_create; | |
1597 | else | |
1598 | ret = inode_state_will_create; | |
1599 | } else if (right_gen == gen) { | |
1600 | if (ino < sctx->send_progress) | |
1601 | ret = inode_state_did_delete; | |
1602 | else | |
1603 | ret = inode_state_will_delete; | |
1604 | } else { | |
1605 | ret = -ENOENT; | |
1606 | } | |
1607 | } else if (!left_ret) { | |
1608 | if (left_gen == gen) { | |
1609 | if (ino < sctx->send_progress) | |
1610 | ret = inode_state_did_create; | |
1611 | else | |
1612 | ret = inode_state_will_create; | |
1613 | } else { | |
1614 | ret = -ENOENT; | |
1615 | } | |
1616 | } else if (!right_ret) { | |
1617 | if (right_gen == gen) { | |
1618 | if (ino < sctx->send_progress) | |
1619 | ret = inode_state_did_delete; | |
1620 | else | |
1621 | ret = inode_state_will_delete; | |
1622 | } else { | |
1623 | ret = -ENOENT; | |
1624 | } | |
1625 | } else { | |
1626 | ret = -ENOENT; | |
1627 | } | |
1628 | ||
1629 | out: | |
1630 | return ret; | |
1631 | } | |
1632 | ||
1633 | static int is_inode_existent(struct send_ctx *sctx, u64 ino, u64 gen) | |
1634 | { | |
1635 | int ret; | |
1636 | ||
1637 | ret = get_cur_inode_state(sctx, ino, gen); | |
1638 | if (ret < 0) | |
1639 | goto out; | |
1640 | ||
1641 | if (ret == inode_state_no_change || | |
1642 | ret == inode_state_did_create || | |
1643 | ret == inode_state_will_delete) | |
1644 | ret = 1; | |
1645 | else | |
1646 | ret = 0; | |
1647 | ||
1648 | out: | |
1649 | return ret; | |
1650 | } | |
1651 | ||
1652 | /* | |
1653 | * Helper function to lookup a dir item in a dir. | |
1654 | */ | |
1655 | static int lookup_dir_item_inode(struct btrfs_root *root, | |
1656 | u64 dir, const char *name, int name_len, | |
1657 | u64 *found_inode, | |
1658 | u8 *found_type) | |
1659 | { | |
1660 | int ret = 0; | |
1661 | struct btrfs_dir_item *di; | |
1662 | struct btrfs_key key; | |
1663 | struct btrfs_path *path; | |
1664 | ||
1665 | path = alloc_path_for_send(); | |
1666 | if (!path) | |
1667 | return -ENOMEM; | |
1668 | ||
1669 | di = btrfs_lookup_dir_item(NULL, root, path, | |
1670 | dir, name, name_len, 0); | |
1671 | if (!di) { | |
1672 | ret = -ENOENT; | |
1673 | goto out; | |
1674 | } | |
1675 | if (IS_ERR(di)) { | |
1676 | ret = PTR_ERR(di); | |
1677 | goto out; | |
1678 | } | |
1679 | btrfs_dir_item_key_to_cpu(path->nodes[0], di, &key); | |
1af56070 FM |
1680 | if (key.type == BTRFS_ROOT_ITEM_KEY) { |
1681 | ret = -ENOENT; | |
1682 | goto out; | |
1683 | } | |
31db9f7c AB |
1684 | *found_inode = key.objectid; |
1685 | *found_type = btrfs_dir_type(path->nodes[0], di); | |
1686 | ||
1687 | out: | |
1688 | btrfs_free_path(path); | |
1689 | return ret; | |
1690 | } | |
1691 | ||
766702ef AB |
1692 | /* |
1693 | * Looks up the first btrfs_inode_ref of a given ino. It returns the parent dir, | |
1694 | * generation of the parent dir and the name of the dir entry. | |
1695 | */ | |
924794c9 | 1696 | static int get_first_ref(struct btrfs_root *root, u64 ino, |
31db9f7c AB |
1697 | u64 *dir, u64 *dir_gen, struct fs_path *name) |
1698 | { | |
1699 | int ret; | |
1700 | struct btrfs_key key; | |
1701 | struct btrfs_key found_key; | |
1702 | struct btrfs_path *path; | |
31db9f7c | 1703 | int len; |
96b5bd77 | 1704 | u64 parent_dir; |
31db9f7c AB |
1705 | |
1706 | path = alloc_path_for_send(); | |
1707 | if (!path) | |
1708 | return -ENOMEM; | |
1709 | ||
1710 | key.objectid = ino; | |
1711 | key.type = BTRFS_INODE_REF_KEY; | |
1712 | key.offset = 0; | |
1713 | ||
1714 | ret = btrfs_search_slot_for_read(root, &key, path, 1, 0); | |
1715 | if (ret < 0) | |
1716 | goto out; | |
1717 | if (!ret) | |
1718 | btrfs_item_key_to_cpu(path->nodes[0], &found_key, | |
1719 | path->slots[0]); | |
96b5bd77 JS |
1720 | if (ret || found_key.objectid != ino || |
1721 | (found_key.type != BTRFS_INODE_REF_KEY && | |
1722 | found_key.type != BTRFS_INODE_EXTREF_KEY)) { | |
31db9f7c AB |
1723 | ret = -ENOENT; |
1724 | goto out; | |
1725 | } | |
1726 | ||
51a60253 | 1727 | if (found_key.type == BTRFS_INODE_REF_KEY) { |
96b5bd77 JS |
1728 | struct btrfs_inode_ref *iref; |
1729 | iref = btrfs_item_ptr(path->nodes[0], path->slots[0], | |
1730 | struct btrfs_inode_ref); | |
1731 | len = btrfs_inode_ref_name_len(path->nodes[0], iref); | |
1732 | ret = fs_path_add_from_extent_buffer(name, path->nodes[0], | |
1733 | (unsigned long)(iref + 1), | |
1734 | len); | |
1735 | parent_dir = found_key.offset; | |
1736 | } else { | |
1737 | struct btrfs_inode_extref *extref; | |
1738 | extref = btrfs_item_ptr(path->nodes[0], path->slots[0], | |
1739 | struct btrfs_inode_extref); | |
1740 | len = btrfs_inode_extref_name_len(path->nodes[0], extref); | |
1741 | ret = fs_path_add_from_extent_buffer(name, path->nodes[0], | |
1742 | (unsigned long)&extref->name, len); | |
1743 | parent_dir = btrfs_inode_extref_parent(path->nodes[0], extref); | |
1744 | } | |
31db9f7c AB |
1745 | if (ret < 0) |
1746 | goto out; | |
1747 | btrfs_release_path(path); | |
1748 | ||
b46ab97b FM |
1749 | if (dir_gen) { |
1750 | ret = get_inode_info(root, parent_dir, NULL, dir_gen, NULL, | |
1751 | NULL, NULL, NULL); | |
1752 | if (ret < 0) | |
1753 | goto out; | |
1754 | } | |
31db9f7c | 1755 | |
96b5bd77 | 1756 | *dir = parent_dir; |
31db9f7c AB |
1757 | |
1758 | out: | |
1759 | btrfs_free_path(path); | |
1760 | return ret; | |
1761 | } | |
1762 | ||
924794c9 | 1763 | static int is_first_ref(struct btrfs_root *root, |
31db9f7c AB |
1764 | u64 ino, u64 dir, |
1765 | const char *name, int name_len) | |
1766 | { | |
1767 | int ret; | |
1768 | struct fs_path *tmp_name; | |
1769 | u64 tmp_dir; | |
31db9f7c | 1770 | |
924794c9 | 1771 | tmp_name = fs_path_alloc(); |
31db9f7c AB |
1772 | if (!tmp_name) |
1773 | return -ENOMEM; | |
1774 | ||
b46ab97b | 1775 | ret = get_first_ref(root, ino, &tmp_dir, NULL, tmp_name); |
31db9f7c AB |
1776 | if (ret < 0) |
1777 | goto out; | |
1778 | ||
b9291aff | 1779 | if (dir != tmp_dir || name_len != fs_path_len(tmp_name)) { |
31db9f7c AB |
1780 | ret = 0; |
1781 | goto out; | |
1782 | } | |
1783 | ||
e938c8ad | 1784 | ret = !memcmp(tmp_name->start, name, name_len); |
31db9f7c AB |
1785 | |
1786 | out: | |
924794c9 | 1787 | fs_path_free(tmp_name); |
31db9f7c AB |
1788 | return ret; |
1789 | } | |
1790 | ||
766702ef AB |
1791 | /* |
1792 | * Used by process_recorded_refs to determine if a new ref would overwrite an | |
1793 | * already existing ref. In case it detects an overwrite, it returns the | |
1794 | * inode/gen in who_ino/who_gen. | |
1795 | * When an overwrite is detected, process_recorded_refs does proper orphanizing | |
1796 | * to make sure later references to the overwritten inode are possible. | |
1797 | * Orphanizing is however only required for the first ref of an inode. | |
1798 | * process_recorded_refs does an additional is_first_ref check to see if | |
1799 | * orphanizing is really required. | |
1800 | */ | |
31db9f7c AB |
1801 | static int will_overwrite_ref(struct send_ctx *sctx, u64 dir, u64 dir_gen, |
1802 | const char *name, int name_len, | |
1803 | u64 *who_ino, u64 *who_gen) | |
1804 | { | |
1805 | int ret = 0; | |
ebdad913 | 1806 | u64 gen; |
31db9f7c AB |
1807 | u64 other_inode = 0; |
1808 | u8 other_type = 0; | |
1809 | ||
1810 | if (!sctx->parent_root) | |
1811 | goto out; | |
1812 | ||
1813 | ret = is_inode_existent(sctx, dir, dir_gen); | |
1814 | if (ret <= 0) | |
1815 | goto out; | |
1816 | ||
ebdad913 JB |
1817 | /* |
1818 | * If we have a parent root we need to verify that the parent dir was | |
1819 | * not delted and then re-created, if it was then we have no overwrite | |
1820 | * and we can just unlink this entry. | |
1821 | */ | |
1822 | if (sctx->parent_root) { | |
1823 | ret = get_inode_info(sctx->parent_root, dir, NULL, &gen, NULL, | |
1824 | NULL, NULL, NULL); | |
1825 | if (ret < 0 && ret != -ENOENT) | |
1826 | goto out; | |
1827 | if (ret) { | |
1828 | ret = 0; | |
1829 | goto out; | |
1830 | } | |
1831 | if (gen != dir_gen) | |
1832 | goto out; | |
1833 | } | |
1834 | ||
31db9f7c AB |
1835 | ret = lookup_dir_item_inode(sctx->parent_root, dir, name, name_len, |
1836 | &other_inode, &other_type); | |
1837 | if (ret < 0 && ret != -ENOENT) | |
1838 | goto out; | |
1839 | if (ret) { | |
1840 | ret = 0; | |
1841 | goto out; | |
1842 | } | |
1843 | ||
766702ef AB |
1844 | /* |
1845 | * Check if the overwritten ref was already processed. If yes, the ref | |
1846 | * was already unlinked/moved, so we can safely assume that we will not | |
1847 | * overwrite anything at this point in time. | |
1848 | */ | |
31db9f7c AB |
1849 | if (other_inode > sctx->send_progress) { |
1850 | ret = get_inode_info(sctx->parent_root, other_inode, NULL, | |
85a7b33b | 1851 | who_gen, NULL, NULL, NULL, NULL); |
31db9f7c AB |
1852 | if (ret < 0) |
1853 | goto out; | |
1854 | ||
1855 | ret = 1; | |
1856 | *who_ino = other_inode; | |
1857 | } else { | |
1858 | ret = 0; | |
1859 | } | |
1860 | ||
1861 | out: | |
1862 | return ret; | |
1863 | } | |
1864 | ||
766702ef AB |
1865 | /* |
1866 | * Checks if the ref was overwritten by an already processed inode. This is | |
1867 | * used by __get_cur_name_and_parent to find out if the ref was orphanized and | |
1868 | * thus the orphan name needs be used. | |
1869 | * process_recorded_refs also uses it to avoid unlinking of refs that were | |
1870 | * overwritten. | |
1871 | */ | |
31db9f7c AB |
1872 | static int did_overwrite_ref(struct send_ctx *sctx, |
1873 | u64 dir, u64 dir_gen, | |
1874 | u64 ino, u64 ino_gen, | |
1875 | const char *name, int name_len) | |
1876 | { | |
1877 | int ret = 0; | |
1878 | u64 gen; | |
1879 | u64 ow_inode; | |
1880 | u8 other_type; | |
1881 | ||
1882 | if (!sctx->parent_root) | |
1883 | goto out; | |
1884 | ||
1885 | ret = is_inode_existent(sctx, dir, dir_gen); | |
1886 | if (ret <= 0) | |
1887 | goto out; | |
1888 | ||
1889 | /* check if the ref was overwritten by another ref */ | |
1890 | ret = lookup_dir_item_inode(sctx->send_root, dir, name, name_len, | |
1891 | &ow_inode, &other_type); | |
1892 | if (ret < 0 && ret != -ENOENT) | |
1893 | goto out; | |
1894 | if (ret) { | |
1895 | /* was never and will never be overwritten */ | |
1896 | ret = 0; | |
1897 | goto out; | |
1898 | } | |
1899 | ||
1900 | ret = get_inode_info(sctx->send_root, ow_inode, NULL, &gen, NULL, NULL, | |
85a7b33b | 1901 | NULL, NULL); |
31db9f7c AB |
1902 | if (ret < 0) |
1903 | goto out; | |
1904 | ||
1905 | if (ow_inode == ino && gen == ino_gen) { | |
1906 | ret = 0; | |
1907 | goto out; | |
1908 | } | |
1909 | ||
8b191a68 FM |
1910 | /* |
1911 | * We know that it is or will be overwritten. Check this now. | |
1912 | * The current inode being processed might have been the one that caused | |
b786f16a FM |
1913 | * inode 'ino' to be orphanized, therefore check if ow_inode matches |
1914 | * the current inode being processed. | |
8b191a68 | 1915 | */ |
b786f16a FM |
1916 | if ((ow_inode < sctx->send_progress) || |
1917 | (ino != sctx->cur_ino && ow_inode == sctx->cur_ino && | |
1918 | gen == sctx->cur_inode_gen)) | |
31db9f7c AB |
1919 | ret = 1; |
1920 | else | |
1921 | ret = 0; | |
1922 | ||
1923 | out: | |
1924 | return ret; | |
1925 | } | |
1926 | ||
766702ef AB |
1927 | /* |
1928 | * Same as did_overwrite_ref, but also checks if it is the first ref of an inode | |
1929 | * that got overwritten. This is used by process_recorded_refs to determine | |
1930 | * if it has to use the path as returned by get_cur_path or the orphan name. | |
1931 | */ | |
31db9f7c AB |
1932 | static int did_overwrite_first_ref(struct send_ctx *sctx, u64 ino, u64 gen) |
1933 | { | |
1934 | int ret = 0; | |
1935 | struct fs_path *name = NULL; | |
1936 | u64 dir; | |
1937 | u64 dir_gen; | |
1938 | ||
1939 | if (!sctx->parent_root) | |
1940 | goto out; | |
1941 | ||
924794c9 | 1942 | name = fs_path_alloc(); |
31db9f7c AB |
1943 | if (!name) |
1944 | return -ENOMEM; | |
1945 | ||
924794c9 | 1946 | ret = get_first_ref(sctx->parent_root, ino, &dir, &dir_gen, name); |
31db9f7c AB |
1947 | if (ret < 0) |
1948 | goto out; | |
1949 | ||
1950 | ret = did_overwrite_ref(sctx, dir, dir_gen, ino, gen, | |
1951 | name->start, fs_path_len(name)); | |
31db9f7c AB |
1952 | |
1953 | out: | |
924794c9 | 1954 | fs_path_free(name); |
31db9f7c AB |
1955 | return ret; |
1956 | } | |
1957 | ||
766702ef AB |
1958 | /* |
1959 | * Insert a name cache entry. On 32bit kernels the radix tree index is 32bit, | |
1960 | * so we need to do some special handling in case we have clashes. This function | |
1961 | * takes care of this with the help of name_cache_entry::radix_list. | |
5dc67d0b | 1962 | * In case of error, nce is kfreed. |
766702ef | 1963 | */ |
31db9f7c AB |
1964 | static int name_cache_insert(struct send_ctx *sctx, |
1965 | struct name_cache_entry *nce) | |
1966 | { | |
1967 | int ret = 0; | |
7e0926fe AB |
1968 | struct list_head *nce_head; |
1969 | ||
1970 | nce_head = radix_tree_lookup(&sctx->name_cache, | |
1971 | (unsigned long)nce->ino); | |
1972 | if (!nce_head) { | |
1973 | nce_head = kmalloc(sizeof(*nce_head), GFP_NOFS); | |
cfa7a9cc TI |
1974 | if (!nce_head) { |
1975 | kfree(nce); | |
31db9f7c | 1976 | return -ENOMEM; |
cfa7a9cc | 1977 | } |
7e0926fe | 1978 | INIT_LIST_HEAD(nce_head); |
31db9f7c | 1979 | |
7e0926fe | 1980 | ret = radix_tree_insert(&sctx->name_cache, nce->ino, nce_head); |
5dc67d0b AB |
1981 | if (ret < 0) { |
1982 | kfree(nce_head); | |
1983 | kfree(nce); | |
31db9f7c | 1984 | return ret; |
5dc67d0b | 1985 | } |
31db9f7c | 1986 | } |
7e0926fe | 1987 | list_add_tail(&nce->radix_list, nce_head); |
31db9f7c AB |
1988 | list_add_tail(&nce->list, &sctx->name_cache_list); |
1989 | sctx->name_cache_size++; | |
1990 | ||
1991 | return ret; | |
1992 | } | |
1993 | ||
1994 | static void name_cache_delete(struct send_ctx *sctx, | |
1995 | struct name_cache_entry *nce) | |
1996 | { | |
7e0926fe | 1997 | struct list_head *nce_head; |
31db9f7c | 1998 | |
7e0926fe AB |
1999 | nce_head = radix_tree_lookup(&sctx->name_cache, |
2000 | (unsigned long)nce->ino); | |
57fb8910 DS |
2001 | if (!nce_head) { |
2002 | btrfs_err(sctx->send_root->fs_info, | |
2003 | "name_cache_delete lookup failed ino %llu cache size %d, leaking memory", | |
2004 | nce->ino, sctx->name_cache_size); | |
2005 | } | |
31db9f7c | 2006 | |
7e0926fe | 2007 | list_del(&nce->radix_list); |
31db9f7c | 2008 | list_del(&nce->list); |
31db9f7c | 2009 | sctx->name_cache_size--; |
7e0926fe | 2010 | |
57fb8910 DS |
2011 | /* |
2012 | * We may not get to the final release of nce_head if the lookup fails | |
2013 | */ | |
2014 | if (nce_head && list_empty(nce_head)) { | |
7e0926fe AB |
2015 | radix_tree_delete(&sctx->name_cache, (unsigned long)nce->ino); |
2016 | kfree(nce_head); | |
2017 | } | |
31db9f7c AB |
2018 | } |
2019 | ||
2020 | static struct name_cache_entry *name_cache_search(struct send_ctx *sctx, | |
2021 | u64 ino, u64 gen) | |
2022 | { | |
7e0926fe AB |
2023 | struct list_head *nce_head; |
2024 | struct name_cache_entry *cur; | |
31db9f7c | 2025 | |
7e0926fe AB |
2026 | nce_head = radix_tree_lookup(&sctx->name_cache, (unsigned long)ino); |
2027 | if (!nce_head) | |
31db9f7c AB |
2028 | return NULL; |
2029 | ||
7e0926fe AB |
2030 | list_for_each_entry(cur, nce_head, radix_list) { |
2031 | if (cur->ino == ino && cur->gen == gen) | |
2032 | return cur; | |
2033 | } | |
31db9f7c AB |
2034 | return NULL; |
2035 | } | |
2036 | ||
766702ef AB |
2037 | /* |
2038 | * Removes the entry from the list and adds it back to the end. This marks the | |
2039 | * entry as recently used so that name_cache_clean_unused does not remove it. | |
2040 | */ | |
31db9f7c AB |
2041 | static void name_cache_used(struct send_ctx *sctx, struct name_cache_entry *nce) |
2042 | { | |
2043 | list_del(&nce->list); | |
2044 | list_add_tail(&nce->list, &sctx->name_cache_list); | |
2045 | } | |
2046 | ||
766702ef AB |
2047 | /* |
2048 | * Remove some entries from the beginning of name_cache_list. | |
2049 | */ | |
31db9f7c AB |
2050 | static void name_cache_clean_unused(struct send_ctx *sctx) |
2051 | { | |
2052 | struct name_cache_entry *nce; | |
2053 | ||
2054 | if (sctx->name_cache_size < SEND_CTX_NAME_CACHE_CLEAN_SIZE) | |
2055 | return; | |
2056 | ||
2057 | while (sctx->name_cache_size > SEND_CTX_MAX_NAME_CACHE_SIZE) { | |
2058 | nce = list_entry(sctx->name_cache_list.next, | |
2059 | struct name_cache_entry, list); | |
2060 | name_cache_delete(sctx, nce); | |
2061 | kfree(nce); | |
2062 | } | |
2063 | } | |
2064 | ||
2065 | static void name_cache_free(struct send_ctx *sctx) | |
2066 | { | |
2067 | struct name_cache_entry *nce; | |
31db9f7c | 2068 | |
e938c8ad AB |
2069 | while (!list_empty(&sctx->name_cache_list)) { |
2070 | nce = list_entry(sctx->name_cache_list.next, | |
2071 | struct name_cache_entry, list); | |
31db9f7c | 2072 | name_cache_delete(sctx, nce); |
17589bd9 | 2073 | kfree(nce); |
31db9f7c AB |
2074 | } |
2075 | } | |
2076 | ||
766702ef AB |
2077 | /* |
2078 | * Used by get_cur_path for each ref up to the root. | |
2079 | * Returns 0 if it succeeded. | |
2080 | * Returns 1 if the inode is not existent or got overwritten. In that case, the | |
2081 | * name is an orphan name. This instructs get_cur_path to stop iterating. If 1 | |
2082 | * is returned, parent_ino/parent_gen are not guaranteed to be valid. | |
2083 | * Returns <0 in case of error. | |
2084 | */ | |
31db9f7c AB |
2085 | static int __get_cur_name_and_parent(struct send_ctx *sctx, |
2086 | u64 ino, u64 gen, | |
2087 | u64 *parent_ino, | |
2088 | u64 *parent_gen, | |
2089 | struct fs_path *dest) | |
2090 | { | |
2091 | int ret; | |
2092 | int nce_ret; | |
31db9f7c AB |
2093 | struct name_cache_entry *nce = NULL; |
2094 | ||
766702ef AB |
2095 | /* |
2096 | * First check if we already did a call to this function with the same | |
2097 | * ino/gen. If yes, check if the cache entry is still up-to-date. If yes | |
2098 | * return the cached result. | |
2099 | */ | |
31db9f7c AB |
2100 | nce = name_cache_search(sctx, ino, gen); |
2101 | if (nce) { | |
2102 | if (ino < sctx->send_progress && nce->need_later_update) { | |
2103 | name_cache_delete(sctx, nce); | |
2104 | kfree(nce); | |
2105 | nce = NULL; | |
2106 | } else { | |
2107 | name_cache_used(sctx, nce); | |
2108 | *parent_ino = nce->parent_ino; | |
2109 | *parent_gen = nce->parent_gen; | |
2110 | ret = fs_path_add(dest, nce->name, nce->name_len); | |
2111 | if (ret < 0) | |
2112 | goto out; | |
2113 | ret = nce->ret; | |
2114 | goto out; | |
2115 | } | |
2116 | } | |
2117 | ||
766702ef AB |
2118 | /* |
2119 | * If the inode is not existent yet, add the orphan name and return 1. | |
2120 | * This should only happen for the parent dir that we determine in | |
2121 | * __record_new_ref | |
2122 | */ | |
31db9f7c AB |
2123 | ret = is_inode_existent(sctx, ino, gen); |
2124 | if (ret < 0) | |
2125 | goto out; | |
2126 | ||
2127 | if (!ret) { | |
2128 | ret = gen_unique_name(sctx, ino, gen, dest); | |
2129 | if (ret < 0) | |
2130 | goto out; | |
2131 | ret = 1; | |
2132 | goto out_cache; | |
2133 | } | |
2134 | ||
766702ef AB |
2135 | /* |
2136 | * Depending on whether the inode was already processed or not, use | |
2137 | * send_root or parent_root for ref lookup. | |
2138 | */ | |
bf0d1f44 | 2139 | if (ino < sctx->send_progress) |
924794c9 TI |
2140 | ret = get_first_ref(sctx->send_root, ino, |
2141 | parent_ino, parent_gen, dest); | |
31db9f7c | 2142 | else |
924794c9 TI |
2143 | ret = get_first_ref(sctx->parent_root, ino, |
2144 | parent_ino, parent_gen, dest); | |
31db9f7c AB |
2145 | if (ret < 0) |
2146 | goto out; | |
2147 | ||
766702ef AB |
2148 | /* |
2149 | * Check if the ref was overwritten by an inode's ref that was processed | |
2150 | * earlier. If yes, treat as orphan and return 1. | |
2151 | */ | |
31db9f7c AB |
2152 | ret = did_overwrite_ref(sctx, *parent_ino, *parent_gen, ino, gen, |
2153 | dest->start, dest->end - dest->start); | |
2154 | if (ret < 0) | |
2155 | goto out; | |
2156 | if (ret) { | |
2157 | fs_path_reset(dest); | |
2158 | ret = gen_unique_name(sctx, ino, gen, dest); | |
2159 | if (ret < 0) | |
2160 | goto out; | |
2161 | ret = 1; | |
2162 | } | |
2163 | ||
2164 | out_cache: | |
766702ef AB |
2165 | /* |
2166 | * Store the result of the lookup in the name cache. | |
2167 | */ | |
31db9f7c AB |
2168 | nce = kmalloc(sizeof(*nce) + fs_path_len(dest) + 1, GFP_NOFS); |
2169 | if (!nce) { | |
2170 | ret = -ENOMEM; | |
2171 | goto out; | |
2172 | } | |
2173 | ||
2174 | nce->ino = ino; | |
2175 | nce->gen = gen; | |
2176 | nce->parent_ino = *parent_ino; | |
2177 | nce->parent_gen = *parent_gen; | |
2178 | nce->name_len = fs_path_len(dest); | |
2179 | nce->ret = ret; | |
2180 | strcpy(nce->name, dest->start); | |
31db9f7c AB |
2181 | |
2182 | if (ino < sctx->send_progress) | |
2183 | nce->need_later_update = 0; | |
2184 | else | |
2185 | nce->need_later_update = 1; | |
2186 | ||
2187 | nce_ret = name_cache_insert(sctx, nce); | |
2188 | if (nce_ret < 0) | |
2189 | ret = nce_ret; | |
2190 | name_cache_clean_unused(sctx); | |
2191 | ||
2192 | out: | |
31db9f7c AB |
2193 | return ret; |
2194 | } | |
2195 | ||
2196 | /* | |
2197 | * Magic happens here. This function returns the first ref to an inode as it | |
2198 | * would look like while receiving the stream at this point in time. | |
2199 | * We walk the path up to the root. For every inode in between, we check if it | |
2200 | * was already processed/sent. If yes, we continue with the parent as found | |
2201 | * in send_root. If not, we continue with the parent as found in parent_root. | |
2202 | * If we encounter an inode that was deleted at this point in time, we use the | |
2203 | * inodes "orphan" name instead of the real name and stop. Same with new inodes | |
2204 | * that were not created yet and overwritten inodes/refs. | |
2205 | * | |
2206 | * When do we have have orphan inodes: | |
2207 | * 1. When an inode is freshly created and thus no valid refs are available yet | |
2208 | * 2. When a directory lost all it's refs (deleted) but still has dir items | |
2209 | * inside which were not processed yet (pending for move/delete). If anyone | |
2210 | * tried to get the path to the dir items, it would get a path inside that | |
2211 | * orphan directory. | |
2212 | * 3. When an inode is moved around or gets new links, it may overwrite the ref | |
2213 | * of an unprocessed inode. If in that case the first ref would be | |
2214 | * overwritten, the overwritten inode gets "orphanized". Later when we | |
2215 | * process this overwritten inode, it is restored at a new place by moving | |
2216 | * the orphan inode. | |
2217 | * | |
2218 | * sctx->send_progress tells this function at which point in time receiving | |
2219 | * would be. | |
2220 | */ | |
2221 | static int get_cur_path(struct send_ctx *sctx, u64 ino, u64 gen, | |
2222 | struct fs_path *dest) | |
2223 | { | |
2224 | int ret = 0; | |
2225 | struct fs_path *name = NULL; | |
2226 | u64 parent_inode = 0; | |
2227 | u64 parent_gen = 0; | |
2228 | int stop = 0; | |
2229 | ||
924794c9 | 2230 | name = fs_path_alloc(); |
31db9f7c AB |
2231 | if (!name) { |
2232 | ret = -ENOMEM; | |
2233 | goto out; | |
2234 | } | |
2235 | ||
2236 | dest->reversed = 1; | |
2237 | fs_path_reset(dest); | |
2238 | ||
2239 | while (!stop && ino != BTRFS_FIRST_FREE_OBJECTID) { | |
8b191a68 FM |
2240 | struct waiting_dir_move *wdm; |
2241 | ||
31db9f7c AB |
2242 | fs_path_reset(name); |
2243 | ||
9dc44214 FM |
2244 | if (is_waiting_for_rm(sctx, ino)) { |
2245 | ret = gen_unique_name(sctx, ino, gen, name); | |
2246 | if (ret < 0) | |
2247 | goto out; | |
2248 | ret = fs_path_add_path(dest, name); | |
2249 | break; | |
2250 | } | |
2251 | ||
8b191a68 FM |
2252 | wdm = get_waiting_dir_move(sctx, ino); |
2253 | if (wdm && wdm->orphanized) { | |
2254 | ret = gen_unique_name(sctx, ino, gen, name); | |
2255 | stop = 1; | |
2256 | } else if (wdm) { | |
bf0d1f44 FM |
2257 | ret = get_first_ref(sctx->parent_root, ino, |
2258 | &parent_inode, &parent_gen, name); | |
2259 | } else { | |
2260 | ret = __get_cur_name_and_parent(sctx, ino, gen, | |
2261 | &parent_inode, | |
2262 | &parent_gen, name); | |
2263 | if (ret) | |
2264 | stop = 1; | |
2265 | } | |
2266 | ||
31db9f7c AB |
2267 | if (ret < 0) |
2268 | goto out; | |
9f03740a | 2269 | |
31db9f7c AB |
2270 | ret = fs_path_add_path(dest, name); |
2271 | if (ret < 0) | |
2272 | goto out; | |
2273 | ||
2274 | ino = parent_inode; | |
2275 | gen = parent_gen; | |
2276 | } | |
2277 | ||
2278 | out: | |
924794c9 | 2279 | fs_path_free(name); |
31db9f7c AB |
2280 | if (!ret) |
2281 | fs_path_unreverse(dest); | |
2282 | return ret; | |
2283 | } | |
2284 | ||
31db9f7c AB |
2285 | /* |
2286 | * Sends a BTRFS_SEND_C_SUBVOL command/item to userspace | |
2287 | */ | |
2288 | static int send_subvol_begin(struct send_ctx *sctx) | |
2289 | { | |
2290 | int ret; | |
2291 | struct btrfs_root *send_root = sctx->send_root; | |
2292 | struct btrfs_root *parent_root = sctx->parent_root; | |
2293 | struct btrfs_path *path; | |
2294 | struct btrfs_key key; | |
2295 | struct btrfs_root_ref *ref; | |
2296 | struct extent_buffer *leaf; | |
2297 | char *name = NULL; | |
2298 | int namelen; | |
2299 | ||
ffcfaf81 | 2300 | path = btrfs_alloc_path(); |
31db9f7c AB |
2301 | if (!path) |
2302 | return -ENOMEM; | |
2303 | ||
2304 | name = kmalloc(BTRFS_PATH_NAME_MAX, GFP_NOFS); | |
2305 | if (!name) { | |
2306 | btrfs_free_path(path); | |
2307 | return -ENOMEM; | |
2308 | } | |
2309 | ||
2310 | key.objectid = send_root->objectid; | |
2311 | key.type = BTRFS_ROOT_BACKREF_KEY; | |
2312 | key.offset = 0; | |
2313 | ||
2314 | ret = btrfs_search_slot_for_read(send_root->fs_info->tree_root, | |
2315 | &key, path, 1, 0); | |
2316 | if (ret < 0) | |
2317 | goto out; | |
2318 | if (ret) { | |
2319 | ret = -ENOENT; | |
2320 | goto out; | |
2321 | } | |
2322 | ||
2323 | leaf = path->nodes[0]; | |
2324 | btrfs_item_key_to_cpu(leaf, &key, path->slots[0]); | |
2325 | if (key.type != BTRFS_ROOT_BACKREF_KEY || | |
2326 | key.objectid != send_root->objectid) { | |
2327 | ret = -ENOENT; | |
2328 | goto out; | |
2329 | } | |
2330 | ref = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_root_ref); | |
2331 | namelen = btrfs_root_ref_name_len(leaf, ref); | |
2332 | read_extent_buffer(leaf, name, (unsigned long)(ref + 1), namelen); | |
2333 | btrfs_release_path(path); | |
2334 | ||
31db9f7c AB |
2335 | if (parent_root) { |
2336 | ret = begin_cmd(sctx, BTRFS_SEND_C_SNAPSHOT); | |
2337 | if (ret < 0) | |
2338 | goto out; | |
2339 | } else { | |
2340 | ret = begin_cmd(sctx, BTRFS_SEND_C_SUBVOL); | |
2341 | if (ret < 0) | |
2342 | goto out; | |
2343 | } | |
2344 | ||
2345 | TLV_PUT_STRING(sctx, BTRFS_SEND_A_PATH, name, namelen); | |
b96b1db0 RR |
2346 | |
2347 | if (!btrfs_is_empty_uuid(sctx->send_root->root_item.received_uuid)) | |
2348 | TLV_PUT_UUID(sctx, BTRFS_SEND_A_UUID, | |
2349 | sctx->send_root->root_item.received_uuid); | |
2350 | else | |
2351 | TLV_PUT_UUID(sctx, BTRFS_SEND_A_UUID, | |
2352 | sctx->send_root->root_item.uuid); | |
2353 | ||
31db9f7c | 2354 | TLV_PUT_U64(sctx, BTRFS_SEND_A_CTRANSID, |
5a0f4e2c | 2355 | le64_to_cpu(sctx->send_root->root_item.ctransid)); |
31db9f7c | 2356 | if (parent_root) { |
37b8d27d JB |
2357 | if (!btrfs_is_empty_uuid(parent_root->root_item.received_uuid)) |
2358 | TLV_PUT_UUID(sctx, BTRFS_SEND_A_CLONE_UUID, | |
2359 | parent_root->root_item.received_uuid); | |
2360 | else | |
2361 | TLV_PUT_UUID(sctx, BTRFS_SEND_A_CLONE_UUID, | |
2362 | parent_root->root_item.uuid); | |
31db9f7c | 2363 | TLV_PUT_U64(sctx, BTRFS_SEND_A_CLONE_CTRANSID, |
5a0f4e2c | 2364 | le64_to_cpu(sctx->parent_root->root_item.ctransid)); |
31db9f7c AB |
2365 | } |
2366 | ||
2367 | ret = send_cmd(sctx); | |
2368 | ||
2369 | tlv_put_failure: | |
2370 | out: | |
2371 | btrfs_free_path(path); | |
2372 | kfree(name); | |
2373 | return ret; | |
2374 | } | |
2375 | ||
2376 | static int send_truncate(struct send_ctx *sctx, u64 ino, u64 gen, u64 size) | |
2377 | { | |
2378 | int ret = 0; | |
2379 | struct fs_path *p; | |
2380 | ||
2381 | verbose_printk("btrfs: send_truncate %llu size=%llu\n", ino, size); | |
2382 | ||
924794c9 | 2383 | p = fs_path_alloc(); |
31db9f7c AB |
2384 | if (!p) |
2385 | return -ENOMEM; | |
2386 | ||
2387 | ret = begin_cmd(sctx, BTRFS_SEND_C_TRUNCATE); | |
2388 | if (ret < 0) | |
2389 | goto out; | |
2390 | ||
2391 | ret = get_cur_path(sctx, ino, gen, p); | |
2392 | if (ret < 0) | |
2393 | goto out; | |
2394 | TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p); | |
2395 | TLV_PUT_U64(sctx, BTRFS_SEND_A_SIZE, size); | |
2396 | ||
2397 | ret = send_cmd(sctx); | |
2398 | ||
2399 | tlv_put_failure: | |
2400 | out: | |
924794c9 | 2401 | fs_path_free(p); |
31db9f7c AB |
2402 | return ret; |
2403 | } | |
2404 | ||
2405 | static int send_chmod(struct send_ctx *sctx, u64 ino, u64 gen, u64 mode) | |
2406 | { | |
2407 | int ret = 0; | |
2408 | struct fs_path *p; | |
2409 | ||
2410 | verbose_printk("btrfs: send_chmod %llu mode=%llu\n", ino, mode); | |
2411 | ||
924794c9 | 2412 | p = fs_path_alloc(); |
31db9f7c AB |
2413 | if (!p) |
2414 | return -ENOMEM; | |
2415 | ||
2416 | ret = begin_cmd(sctx, BTRFS_SEND_C_CHMOD); | |
2417 | if (ret < 0) | |
2418 | goto out; | |
2419 | ||
2420 | ret = get_cur_path(sctx, ino, gen, p); | |
2421 | if (ret < 0) | |
2422 | goto out; | |
2423 | TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p); | |
2424 | TLV_PUT_U64(sctx, BTRFS_SEND_A_MODE, mode & 07777); | |
2425 | ||
2426 | ret = send_cmd(sctx); | |
2427 | ||
2428 | tlv_put_failure: | |
2429 | out: | |
924794c9 | 2430 | fs_path_free(p); |
31db9f7c AB |
2431 | return ret; |
2432 | } | |
2433 | ||
2434 | static int send_chown(struct send_ctx *sctx, u64 ino, u64 gen, u64 uid, u64 gid) | |
2435 | { | |
2436 | int ret = 0; | |
2437 | struct fs_path *p; | |
2438 | ||
2439 | verbose_printk("btrfs: send_chown %llu uid=%llu, gid=%llu\n", ino, uid, gid); | |
2440 | ||
924794c9 | 2441 | p = fs_path_alloc(); |
31db9f7c AB |
2442 | if (!p) |
2443 | return -ENOMEM; | |
2444 | ||
2445 | ret = begin_cmd(sctx, BTRFS_SEND_C_CHOWN); | |
2446 | if (ret < 0) | |
2447 | goto out; | |
2448 | ||
2449 | ret = get_cur_path(sctx, ino, gen, p); | |
2450 | if (ret < 0) | |
2451 | goto out; | |
2452 | TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p); | |
2453 | TLV_PUT_U64(sctx, BTRFS_SEND_A_UID, uid); | |
2454 | TLV_PUT_U64(sctx, BTRFS_SEND_A_GID, gid); | |
2455 | ||
2456 | ret = send_cmd(sctx); | |
2457 | ||
2458 | tlv_put_failure: | |
2459 | out: | |
924794c9 | 2460 | fs_path_free(p); |
31db9f7c AB |
2461 | return ret; |
2462 | } | |
2463 | ||
2464 | static int send_utimes(struct send_ctx *sctx, u64 ino, u64 gen) | |
2465 | { | |
2466 | int ret = 0; | |
2467 | struct fs_path *p = NULL; | |
2468 | struct btrfs_inode_item *ii; | |
2469 | struct btrfs_path *path = NULL; | |
2470 | struct extent_buffer *eb; | |
2471 | struct btrfs_key key; | |
2472 | int slot; | |
2473 | ||
2474 | verbose_printk("btrfs: send_utimes %llu\n", ino); | |
2475 | ||
924794c9 | 2476 | p = fs_path_alloc(); |
31db9f7c AB |
2477 | if (!p) |
2478 | return -ENOMEM; | |
2479 | ||
2480 | path = alloc_path_for_send(); | |
2481 | if (!path) { | |
2482 | ret = -ENOMEM; | |
2483 | goto out; | |
2484 | } | |
2485 | ||
2486 | key.objectid = ino; | |
2487 | key.type = BTRFS_INODE_ITEM_KEY; | |
2488 | key.offset = 0; | |
2489 | ret = btrfs_search_slot(NULL, sctx->send_root, &key, path, 0, 0); | |
2490 | if (ret < 0) | |
2491 | goto out; | |
2492 | ||
2493 | eb = path->nodes[0]; | |
2494 | slot = path->slots[0]; | |
2495 | ii = btrfs_item_ptr(eb, slot, struct btrfs_inode_item); | |
2496 | ||
2497 | ret = begin_cmd(sctx, BTRFS_SEND_C_UTIMES); | |
2498 | if (ret < 0) | |
2499 | goto out; | |
2500 | ||
2501 | ret = get_cur_path(sctx, ino, gen, p); | |
2502 | if (ret < 0) | |
2503 | goto out; | |
2504 | TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p); | |
a937b979 DS |
2505 | TLV_PUT_BTRFS_TIMESPEC(sctx, BTRFS_SEND_A_ATIME, eb, &ii->atime); |
2506 | TLV_PUT_BTRFS_TIMESPEC(sctx, BTRFS_SEND_A_MTIME, eb, &ii->mtime); | |
2507 | TLV_PUT_BTRFS_TIMESPEC(sctx, BTRFS_SEND_A_CTIME, eb, &ii->ctime); | |
766702ef | 2508 | /* TODO Add otime support when the otime patches get into upstream */ |
31db9f7c AB |
2509 | |
2510 | ret = send_cmd(sctx); | |
2511 | ||
2512 | tlv_put_failure: | |
2513 | out: | |
924794c9 | 2514 | fs_path_free(p); |
31db9f7c AB |
2515 | btrfs_free_path(path); |
2516 | return ret; | |
2517 | } | |
2518 | ||
2519 | /* | |
2520 | * Sends a BTRFS_SEND_C_MKXXX or SYMLINK command to user space. We don't have | |
2521 | * a valid path yet because we did not process the refs yet. So, the inode | |
2522 | * is created as orphan. | |
2523 | */ | |
1f4692da | 2524 | static int send_create_inode(struct send_ctx *sctx, u64 ino) |
31db9f7c AB |
2525 | { |
2526 | int ret = 0; | |
31db9f7c | 2527 | struct fs_path *p; |
31db9f7c | 2528 | int cmd; |
1f4692da | 2529 | u64 gen; |
31db9f7c | 2530 | u64 mode; |
1f4692da | 2531 | u64 rdev; |
31db9f7c | 2532 | |
1f4692da | 2533 | verbose_printk("btrfs: send_create_inode %llu\n", ino); |
31db9f7c | 2534 | |
924794c9 | 2535 | p = fs_path_alloc(); |
31db9f7c AB |
2536 | if (!p) |
2537 | return -ENOMEM; | |
2538 | ||
644d1940 LB |
2539 | if (ino != sctx->cur_ino) { |
2540 | ret = get_inode_info(sctx->send_root, ino, NULL, &gen, &mode, | |
2541 | NULL, NULL, &rdev); | |
2542 | if (ret < 0) | |
2543 | goto out; | |
2544 | } else { | |
2545 | gen = sctx->cur_inode_gen; | |
2546 | mode = sctx->cur_inode_mode; | |
2547 | rdev = sctx->cur_inode_rdev; | |
2548 | } | |
31db9f7c | 2549 | |
e938c8ad | 2550 | if (S_ISREG(mode)) { |
31db9f7c | 2551 | cmd = BTRFS_SEND_C_MKFILE; |
e938c8ad | 2552 | } else if (S_ISDIR(mode)) { |
31db9f7c | 2553 | cmd = BTRFS_SEND_C_MKDIR; |
e938c8ad | 2554 | } else if (S_ISLNK(mode)) { |
31db9f7c | 2555 | cmd = BTRFS_SEND_C_SYMLINK; |
e938c8ad | 2556 | } else if (S_ISCHR(mode) || S_ISBLK(mode)) { |
31db9f7c | 2557 | cmd = BTRFS_SEND_C_MKNOD; |
e938c8ad | 2558 | } else if (S_ISFIFO(mode)) { |
31db9f7c | 2559 | cmd = BTRFS_SEND_C_MKFIFO; |
e938c8ad | 2560 | } else if (S_ISSOCK(mode)) { |
31db9f7c | 2561 | cmd = BTRFS_SEND_C_MKSOCK; |
e938c8ad | 2562 | } else { |
f14d104d | 2563 | btrfs_warn(sctx->send_root->fs_info, "unexpected inode type %o", |
31db9f7c AB |
2564 | (int)(mode & S_IFMT)); |
2565 | ret = -ENOTSUPP; | |
2566 | goto out; | |
2567 | } | |
2568 | ||
2569 | ret = begin_cmd(sctx, cmd); | |
2570 | if (ret < 0) | |
2571 | goto out; | |
2572 | ||
1f4692da | 2573 | ret = gen_unique_name(sctx, ino, gen, p); |
31db9f7c AB |
2574 | if (ret < 0) |
2575 | goto out; | |
2576 | ||
2577 | TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p); | |
1f4692da | 2578 | TLV_PUT_U64(sctx, BTRFS_SEND_A_INO, ino); |
31db9f7c AB |
2579 | |
2580 | if (S_ISLNK(mode)) { | |
2581 | fs_path_reset(p); | |
924794c9 | 2582 | ret = read_symlink(sctx->send_root, ino, p); |
31db9f7c AB |
2583 | if (ret < 0) |
2584 | goto out; | |
2585 | TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH_LINK, p); | |
2586 | } else if (S_ISCHR(mode) || S_ISBLK(mode) || | |
2587 | S_ISFIFO(mode) || S_ISSOCK(mode)) { | |
d79e5043 AJ |
2588 | TLV_PUT_U64(sctx, BTRFS_SEND_A_RDEV, new_encode_dev(rdev)); |
2589 | TLV_PUT_U64(sctx, BTRFS_SEND_A_MODE, mode); | |
31db9f7c AB |
2590 | } |
2591 | ||
2592 | ret = send_cmd(sctx); | |
2593 | if (ret < 0) | |
2594 | goto out; | |
2595 | ||
2596 | ||
2597 | tlv_put_failure: | |
2598 | out: | |
924794c9 | 2599 | fs_path_free(p); |
31db9f7c AB |
2600 | return ret; |
2601 | } | |
2602 | ||
1f4692da AB |
2603 | /* |
2604 | * We need some special handling for inodes that get processed before the parent | |
2605 | * directory got created. See process_recorded_refs for details. | |
2606 | * This function does the check if we already created the dir out of order. | |
2607 | */ | |
2608 | static int did_create_dir(struct send_ctx *sctx, u64 dir) | |
2609 | { | |
2610 | int ret = 0; | |
2611 | struct btrfs_path *path = NULL; | |
2612 | struct btrfs_key key; | |
2613 | struct btrfs_key found_key; | |
2614 | struct btrfs_key di_key; | |
2615 | struct extent_buffer *eb; | |
2616 | struct btrfs_dir_item *di; | |
2617 | int slot; | |
2618 | ||
2619 | path = alloc_path_for_send(); | |
2620 | if (!path) { | |
2621 | ret = -ENOMEM; | |
2622 | goto out; | |
2623 | } | |
2624 | ||
2625 | key.objectid = dir; | |
2626 | key.type = BTRFS_DIR_INDEX_KEY; | |
2627 | key.offset = 0; | |
dff6d0ad FDBM |
2628 | ret = btrfs_search_slot(NULL, sctx->send_root, &key, path, 0, 0); |
2629 | if (ret < 0) | |
2630 | goto out; | |
2631 | ||
1f4692da | 2632 | while (1) { |
dff6d0ad FDBM |
2633 | eb = path->nodes[0]; |
2634 | slot = path->slots[0]; | |
2635 | if (slot >= btrfs_header_nritems(eb)) { | |
2636 | ret = btrfs_next_leaf(sctx->send_root, path); | |
2637 | if (ret < 0) { | |
2638 | goto out; | |
2639 | } else if (ret > 0) { | |
2640 | ret = 0; | |
2641 | break; | |
2642 | } | |
2643 | continue; | |
1f4692da | 2644 | } |
dff6d0ad FDBM |
2645 | |
2646 | btrfs_item_key_to_cpu(eb, &found_key, slot); | |
2647 | if (found_key.objectid != key.objectid || | |
1f4692da AB |
2648 | found_key.type != key.type) { |
2649 | ret = 0; | |
2650 | goto out; | |
2651 | } | |
2652 | ||
2653 | di = btrfs_item_ptr(eb, slot, struct btrfs_dir_item); | |
2654 | btrfs_dir_item_key_to_cpu(eb, di, &di_key); | |
2655 | ||
a0525414 JB |
2656 | if (di_key.type != BTRFS_ROOT_ITEM_KEY && |
2657 | di_key.objectid < sctx->send_progress) { | |
1f4692da AB |
2658 | ret = 1; |
2659 | goto out; | |
2660 | } | |
2661 | ||
dff6d0ad | 2662 | path->slots[0]++; |
1f4692da AB |
2663 | } |
2664 | ||
2665 | out: | |
2666 | btrfs_free_path(path); | |
2667 | return ret; | |
2668 | } | |
2669 | ||
2670 | /* | |
2671 | * Only creates the inode if it is: | |
2672 | * 1. Not a directory | |
2673 | * 2. Or a directory which was not created already due to out of order | |
2674 | * directories. See did_create_dir and process_recorded_refs for details. | |
2675 | */ | |
2676 | static int send_create_inode_if_needed(struct send_ctx *sctx) | |
2677 | { | |
2678 | int ret; | |
2679 | ||
2680 | if (S_ISDIR(sctx->cur_inode_mode)) { | |
2681 | ret = did_create_dir(sctx, sctx->cur_ino); | |
2682 | if (ret < 0) | |
2683 | goto out; | |
2684 | if (ret) { | |
2685 | ret = 0; | |
2686 | goto out; | |
2687 | } | |
2688 | } | |
2689 | ||
2690 | ret = send_create_inode(sctx, sctx->cur_ino); | |
2691 | if (ret < 0) | |
2692 | goto out; | |
2693 | ||
2694 | out: | |
2695 | return ret; | |
2696 | } | |
2697 | ||
31db9f7c AB |
2698 | struct recorded_ref { |
2699 | struct list_head list; | |
2700 | char *dir_path; | |
2701 | char *name; | |
2702 | struct fs_path *full_path; | |
2703 | u64 dir; | |
2704 | u64 dir_gen; | |
2705 | int dir_path_len; | |
2706 | int name_len; | |
2707 | }; | |
2708 | ||
2709 | /* | |
2710 | * We need to process new refs before deleted refs, but compare_tree gives us | |
2711 | * everything mixed. So we first record all refs and later process them. | |
2712 | * This function is a helper to record one ref. | |
2713 | */ | |
a4d96d62 | 2714 | static int __record_ref(struct list_head *head, u64 dir, |
31db9f7c AB |
2715 | u64 dir_gen, struct fs_path *path) |
2716 | { | |
2717 | struct recorded_ref *ref; | |
31db9f7c AB |
2718 | |
2719 | ref = kmalloc(sizeof(*ref), GFP_NOFS); | |
2720 | if (!ref) | |
2721 | return -ENOMEM; | |
2722 | ||
2723 | ref->dir = dir; | |
2724 | ref->dir_gen = dir_gen; | |
2725 | ref->full_path = path; | |
2726 | ||
ed84885d AS |
2727 | ref->name = (char *)kbasename(ref->full_path->start); |
2728 | ref->name_len = ref->full_path->end - ref->name; | |
2729 | ref->dir_path = ref->full_path->start; | |
2730 | if (ref->name == ref->full_path->start) | |
31db9f7c | 2731 | ref->dir_path_len = 0; |
ed84885d | 2732 | else |
31db9f7c AB |
2733 | ref->dir_path_len = ref->full_path->end - |
2734 | ref->full_path->start - 1 - ref->name_len; | |
31db9f7c AB |
2735 | |
2736 | list_add_tail(&ref->list, head); | |
2737 | return 0; | |
2738 | } | |
2739 | ||
ba5e8f2e JB |
2740 | static int dup_ref(struct recorded_ref *ref, struct list_head *list) |
2741 | { | |
2742 | struct recorded_ref *new; | |
2743 | ||
2744 | new = kmalloc(sizeof(*ref), GFP_NOFS); | |
2745 | if (!new) | |
2746 | return -ENOMEM; | |
2747 | ||
2748 | new->dir = ref->dir; | |
2749 | new->dir_gen = ref->dir_gen; | |
2750 | new->full_path = NULL; | |
2751 | INIT_LIST_HEAD(&new->list); | |
2752 | list_add_tail(&new->list, list); | |
2753 | return 0; | |
2754 | } | |
2755 | ||
924794c9 | 2756 | static void __free_recorded_refs(struct list_head *head) |
31db9f7c AB |
2757 | { |
2758 | struct recorded_ref *cur; | |
31db9f7c | 2759 | |
e938c8ad AB |
2760 | while (!list_empty(head)) { |
2761 | cur = list_entry(head->next, struct recorded_ref, list); | |
924794c9 | 2762 | fs_path_free(cur->full_path); |
e938c8ad | 2763 | list_del(&cur->list); |
31db9f7c AB |
2764 | kfree(cur); |
2765 | } | |
31db9f7c AB |
2766 | } |
2767 | ||
2768 | static void free_recorded_refs(struct send_ctx *sctx) | |
2769 | { | |
924794c9 TI |
2770 | __free_recorded_refs(&sctx->new_refs); |
2771 | __free_recorded_refs(&sctx->deleted_refs); | |
31db9f7c AB |
2772 | } |
2773 | ||
2774 | /* | |
766702ef | 2775 | * Renames/moves a file/dir to its orphan name. Used when the first |
31db9f7c AB |
2776 | * ref of an unprocessed inode gets overwritten and for all non empty |
2777 | * directories. | |
2778 | */ | |
2779 | static int orphanize_inode(struct send_ctx *sctx, u64 ino, u64 gen, | |
2780 | struct fs_path *path) | |
2781 | { | |
2782 | int ret; | |
2783 | struct fs_path *orphan; | |
2784 | ||
924794c9 | 2785 | orphan = fs_path_alloc(); |
31db9f7c AB |
2786 | if (!orphan) |
2787 | return -ENOMEM; | |
2788 | ||
2789 | ret = gen_unique_name(sctx, ino, gen, orphan); | |
2790 | if (ret < 0) | |
2791 | goto out; | |
2792 | ||
2793 | ret = send_rename(sctx, path, orphan); | |
2794 | ||
2795 | out: | |
924794c9 | 2796 | fs_path_free(orphan); |
31db9f7c AB |
2797 | return ret; |
2798 | } | |
2799 | ||
9dc44214 FM |
2800 | static struct orphan_dir_info * |
2801 | add_orphan_dir_info(struct send_ctx *sctx, u64 dir_ino) | |
2802 | { | |
2803 | struct rb_node **p = &sctx->orphan_dirs.rb_node; | |
2804 | struct rb_node *parent = NULL; | |
2805 | struct orphan_dir_info *entry, *odi; | |
2806 | ||
2807 | odi = kmalloc(sizeof(*odi), GFP_NOFS); | |
2808 | if (!odi) | |
2809 | return ERR_PTR(-ENOMEM); | |
2810 | odi->ino = dir_ino; | |
2811 | odi->gen = 0; | |
2812 | ||
2813 | while (*p) { | |
2814 | parent = *p; | |
2815 | entry = rb_entry(parent, struct orphan_dir_info, node); | |
2816 | if (dir_ino < entry->ino) { | |
2817 | p = &(*p)->rb_left; | |
2818 | } else if (dir_ino > entry->ino) { | |
2819 | p = &(*p)->rb_right; | |
2820 | } else { | |
2821 | kfree(odi); | |
2822 | return entry; | |
2823 | } | |
2824 | } | |
2825 | ||
2826 | rb_link_node(&odi->node, parent, p); | |
2827 | rb_insert_color(&odi->node, &sctx->orphan_dirs); | |
2828 | return odi; | |
2829 | } | |
2830 | ||
2831 | static struct orphan_dir_info * | |
2832 | get_orphan_dir_info(struct send_ctx *sctx, u64 dir_ino) | |
2833 | { | |
2834 | struct rb_node *n = sctx->orphan_dirs.rb_node; | |
2835 | struct orphan_dir_info *entry; | |
2836 | ||
2837 | while (n) { | |
2838 | entry = rb_entry(n, struct orphan_dir_info, node); | |
2839 | if (dir_ino < entry->ino) | |
2840 | n = n->rb_left; | |
2841 | else if (dir_ino > entry->ino) | |
2842 | n = n->rb_right; | |
2843 | else | |
2844 | return entry; | |
2845 | } | |
2846 | return NULL; | |
2847 | } | |
2848 | ||
2849 | static int is_waiting_for_rm(struct send_ctx *sctx, u64 dir_ino) | |
2850 | { | |
2851 | struct orphan_dir_info *odi = get_orphan_dir_info(sctx, dir_ino); | |
2852 | ||
2853 | return odi != NULL; | |
2854 | } | |
2855 | ||
2856 | static void free_orphan_dir_info(struct send_ctx *sctx, | |
2857 | struct orphan_dir_info *odi) | |
2858 | { | |
2859 | if (!odi) | |
2860 | return; | |
2861 | rb_erase(&odi->node, &sctx->orphan_dirs); | |
2862 | kfree(odi); | |
2863 | } | |
2864 | ||
31db9f7c AB |
2865 | /* |
2866 | * Returns 1 if a directory can be removed at this point in time. | |
2867 | * We check this by iterating all dir items and checking if the inode behind | |
2868 | * the dir item was already processed. | |
2869 | */ | |
9dc44214 FM |
2870 | static int can_rmdir(struct send_ctx *sctx, u64 dir, u64 dir_gen, |
2871 | u64 send_progress) | |
31db9f7c AB |
2872 | { |
2873 | int ret = 0; | |
2874 | struct btrfs_root *root = sctx->parent_root; | |
2875 | struct btrfs_path *path; | |
2876 | struct btrfs_key key; | |
2877 | struct btrfs_key found_key; | |
2878 | struct btrfs_key loc; | |
2879 | struct btrfs_dir_item *di; | |
2880 | ||
6d85ed05 AB |
2881 | /* |
2882 | * Don't try to rmdir the top/root subvolume dir. | |
2883 | */ | |
2884 | if (dir == BTRFS_FIRST_FREE_OBJECTID) | |
2885 | return 0; | |
2886 | ||
31db9f7c AB |
2887 | path = alloc_path_for_send(); |
2888 | if (!path) | |
2889 | return -ENOMEM; | |
2890 | ||
2891 | key.objectid = dir; | |
2892 | key.type = BTRFS_DIR_INDEX_KEY; | |
2893 | key.offset = 0; | |
dff6d0ad FDBM |
2894 | ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); |
2895 | if (ret < 0) | |
2896 | goto out; | |
31db9f7c AB |
2897 | |
2898 | while (1) { | |
9dc44214 FM |
2899 | struct waiting_dir_move *dm; |
2900 | ||
dff6d0ad FDBM |
2901 | if (path->slots[0] >= btrfs_header_nritems(path->nodes[0])) { |
2902 | ret = btrfs_next_leaf(root, path); | |
2903 | if (ret < 0) | |
2904 | goto out; | |
2905 | else if (ret > 0) | |
2906 | break; | |
2907 | continue; | |
31db9f7c | 2908 | } |
dff6d0ad FDBM |
2909 | btrfs_item_key_to_cpu(path->nodes[0], &found_key, |
2910 | path->slots[0]); | |
2911 | if (found_key.objectid != key.objectid || | |
2912 | found_key.type != key.type) | |
31db9f7c | 2913 | break; |
31db9f7c AB |
2914 | |
2915 | di = btrfs_item_ptr(path->nodes[0], path->slots[0], | |
2916 | struct btrfs_dir_item); | |
2917 | btrfs_dir_item_key_to_cpu(path->nodes[0], di, &loc); | |
2918 | ||
9dc44214 FM |
2919 | dm = get_waiting_dir_move(sctx, loc.objectid); |
2920 | if (dm) { | |
2921 | struct orphan_dir_info *odi; | |
2922 | ||
2923 | odi = add_orphan_dir_info(sctx, dir); | |
2924 | if (IS_ERR(odi)) { | |
2925 | ret = PTR_ERR(odi); | |
2926 | goto out; | |
2927 | } | |
2928 | odi->gen = dir_gen; | |
2929 | dm->rmdir_ino = dir; | |
2930 | ret = 0; | |
2931 | goto out; | |
2932 | } | |
2933 | ||
31db9f7c AB |
2934 | if (loc.objectid > send_progress) { |
2935 | ret = 0; | |
2936 | goto out; | |
2937 | } | |
2938 | ||
dff6d0ad | 2939 | path->slots[0]++; |
31db9f7c AB |
2940 | } |
2941 | ||
2942 | ret = 1; | |
2943 | ||
2944 | out: | |
2945 | btrfs_free_path(path); | |
2946 | return ret; | |
2947 | } | |
2948 | ||
9f03740a FDBM |
2949 | static int is_waiting_for_move(struct send_ctx *sctx, u64 ino) |
2950 | { | |
9dc44214 | 2951 | struct waiting_dir_move *entry = get_waiting_dir_move(sctx, ino); |
9f03740a | 2952 | |
9dc44214 | 2953 | return entry != NULL; |
9f03740a FDBM |
2954 | } |
2955 | ||
8b191a68 | 2956 | static int add_waiting_dir_move(struct send_ctx *sctx, u64 ino, bool orphanized) |
9f03740a FDBM |
2957 | { |
2958 | struct rb_node **p = &sctx->waiting_dir_moves.rb_node; | |
2959 | struct rb_node *parent = NULL; | |
2960 | struct waiting_dir_move *entry, *dm; | |
2961 | ||
2962 | dm = kmalloc(sizeof(*dm), GFP_NOFS); | |
2963 | if (!dm) | |
2964 | return -ENOMEM; | |
2965 | dm->ino = ino; | |
9dc44214 | 2966 | dm->rmdir_ino = 0; |
8b191a68 | 2967 | dm->orphanized = orphanized; |
9f03740a FDBM |
2968 | |
2969 | while (*p) { | |
2970 | parent = *p; | |
2971 | entry = rb_entry(parent, struct waiting_dir_move, node); | |
2972 | if (ino < entry->ino) { | |
2973 | p = &(*p)->rb_left; | |
2974 | } else if (ino > entry->ino) { | |
2975 | p = &(*p)->rb_right; | |
2976 | } else { | |
2977 | kfree(dm); | |
2978 | return -EEXIST; | |
2979 | } | |
2980 | } | |
2981 | ||
2982 | rb_link_node(&dm->node, parent, p); | |
2983 | rb_insert_color(&dm->node, &sctx->waiting_dir_moves); | |
2984 | return 0; | |
2985 | } | |
2986 | ||
9dc44214 FM |
2987 | static struct waiting_dir_move * |
2988 | get_waiting_dir_move(struct send_ctx *sctx, u64 ino) | |
9f03740a FDBM |
2989 | { |
2990 | struct rb_node *n = sctx->waiting_dir_moves.rb_node; | |
2991 | struct waiting_dir_move *entry; | |
2992 | ||
2993 | while (n) { | |
2994 | entry = rb_entry(n, struct waiting_dir_move, node); | |
9dc44214 | 2995 | if (ino < entry->ino) |
9f03740a | 2996 | n = n->rb_left; |
9dc44214 | 2997 | else if (ino > entry->ino) |
9f03740a | 2998 | n = n->rb_right; |
9dc44214 FM |
2999 | else |
3000 | return entry; | |
9f03740a | 3001 | } |
9dc44214 FM |
3002 | return NULL; |
3003 | } | |
3004 | ||
3005 | static void free_waiting_dir_move(struct send_ctx *sctx, | |
3006 | struct waiting_dir_move *dm) | |
3007 | { | |
3008 | if (!dm) | |
3009 | return; | |
3010 | rb_erase(&dm->node, &sctx->waiting_dir_moves); | |
3011 | kfree(dm); | |
9f03740a FDBM |
3012 | } |
3013 | ||
bfa7e1f8 FM |
3014 | static int add_pending_dir_move(struct send_ctx *sctx, |
3015 | u64 ino, | |
3016 | u64 ino_gen, | |
f959492f FM |
3017 | u64 parent_ino, |
3018 | struct list_head *new_refs, | |
84471e24 FM |
3019 | struct list_head *deleted_refs, |
3020 | const bool is_orphan) | |
9f03740a FDBM |
3021 | { |
3022 | struct rb_node **p = &sctx->pending_dir_moves.rb_node; | |
3023 | struct rb_node *parent = NULL; | |
73b802f4 | 3024 | struct pending_dir_move *entry = NULL, *pm; |
9f03740a FDBM |
3025 | struct recorded_ref *cur; |
3026 | int exists = 0; | |
3027 | int ret; | |
3028 | ||
3029 | pm = kmalloc(sizeof(*pm), GFP_NOFS); | |
3030 | if (!pm) | |
3031 | return -ENOMEM; | |
3032 | pm->parent_ino = parent_ino; | |
bfa7e1f8 FM |
3033 | pm->ino = ino; |
3034 | pm->gen = ino_gen; | |
84471e24 | 3035 | pm->is_orphan = is_orphan; |
9f03740a FDBM |
3036 | INIT_LIST_HEAD(&pm->list); |
3037 | INIT_LIST_HEAD(&pm->update_refs); | |
3038 | RB_CLEAR_NODE(&pm->node); | |
3039 | ||
3040 | while (*p) { | |
3041 | parent = *p; | |
3042 | entry = rb_entry(parent, struct pending_dir_move, node); | |
3043 | if (parent_ino < entry->parent_ino) { | |
3044 | p = &(*p)->rb_left; | |
3045 | } else if (parent_ino > entry->parent_ino) { | |
3046 | p = &(*p)->rb_right; | |
3047 | } else { | |
3048 | exists = 1; | |
3049 | break; | |
3050 | } | |
3051 | } | |
3052 | ||
f959492f | 3053 | list_for_each_entry(cur, deleted_refs, list) { |
9f03740a FDBM |
3054 | ret = dup_ref(cur, &pm->update_refs); |
3055 | if (ret < 0) | |
3056 | goto out; | |
3057 | } | |
f959492f | 3058 | list_for_each_entry(cur, new_refs, list) { |
9f03740a FDBM |
3059 | ret = dup_ref(cur, &pm->update_refs); |
3060 | if (ret < 0) | |
3061 | goto out; | |
3062 | } | |
3063 | ||
8b191a68 | 3064 | ret = add_waiting_dir_move(sctx, pm->ino, is_orphan); |
9f03740a FDBM |
3065 | if (ret) |
3066 | goto out; | |
3067 | ||
3068 | if (exists) { | |
3069 | list_add_tail(&pm->list, &entry->list); | |
3070 | } else { | |
3071 | rb_link_node(&pm->node, parent, p); | |
3072 | rb_insert_color(&pm->node, &sctx->pending_dir_moves); | |
3073 | } | |
3074 | ret = 0; | |
3075 | out: | |
3076 | if (ret) { | |
3077 | __free_recorded_refs(&pm->update_refs); | |
3078 | kfree(pm); | |
3079 | } | |
3080 | return ret; | |
3081 | } | |
3082 | ||
3083 | static struct pending_dir_move *get_pending_dir_moves(struct send_ctx *sctx, | |
3084 | u64 parent_ino) | |
3085 | { | |
3086 | struct rb_node *n = sctx->pending_dir_moves.rb_node; | |
3087 | struct pending_dir_move *entry; | |
3088 | ||
3089 | while (n) { | |
3090 | entry = rb_entry(n, struct pending_dir_move, node); | |
3091 | if (parent_ino < entry->parent_ino) | |
3092 | n = n->rb_left; | |
3093 | else if (parent_ino > entry->parent_ino) | |
3094 | n = n->rb_right; | |
3095 | else | |
3096 | return entry; | |
3097 | } | |
3098 | return NULL; | |
3099 | } | |
3100 | ||
3101 | static int apply_dir_move(struct send_ctx *sctx, struct pending_dir_move *pm) | |
3102 | { | |
3103 | struct fs_path *from_path = NULL; | |
3104 | struct fs_path *to_path = NULL; | |
2b863a13 | 3105 | struct fs_path *name = NULL; |
9f03740a FDBM |
3106 | u64 orig_progress = sctx->send_progress; |
3107 | struct recorded_ref *cur; | |
2b863a13 | 3108 | u64 parent_ino, parent_gen; |
9dc44214 FM |
3109 | struct waiting_dir_move *dm = NULL; |
3110 | u64 rmdir_ino = 0; | |
9f03740a FDBM |
3111 | int ret; |
3112 | ||
2b863a13 | 3113 | name = fs_path_alloc(); |
9f03740a | 3114 | from_path = fs_path_alloc(); |
2b863a13 FM |
3115 | if (!name || !from_path) { |
3116 | ret = -ENOMEM; | |
3117 | goto out; | |
3118 | } | |
9f03740a | 3119 | |
9dc44214 FM |
3120 | dm = get_waiting_dir_move(sctx, pm->ino); |
3121 | ASSERT(dm); | |
3122 | rmdir_ino = dm->rmdir_ino; | |
3123 | free_waiting_dir_move(sctx, dm); | |
2b863a13 | 3124 | |
84471e24 FM |
3125 | if (pm->is_orphan) { |
3126 | ret = gen_unique_name(sctx, pm->ino, | |
3127 | pm->gen, from_path); | |
3128 | } else { | |
3129 | ret = get_first_ref(sctx->parent_root, pm->ino, | |
3130 | &parent_ino, &parent_gen, name); | |
3131 | if (ret < 0) | |
3132 | goto out; | |
3133 | ret = get_cur_path(sctx, parent_ino, parent_gen, | |
3134 | from_path); | |
3135 | if (ret < 0) | |
3136 | goto out; | |
3137 | ret = fs_path_add_path(from_path, name); | |
3138 | } | |
c992ec94 FM |
3139 | if (ret < 0) |
3140 | goto out; | |
2b863a13 | 3141 | |
f959492f | 3142 | sctx->send_progress = sctx->cur_ino + 1; |
c992ec94 FM |
3143 | fs_path_reset(name); |
3144 | to_path = name; | |
2b863a13 | 3145 | name = NULL; |
9f03740a FDBM |
3146 | ret = get_cur_path(sctx, pm->ino, pm->gen, to_path); |
3147 | if (ret < 0) | |
3148 | goto out; | |
3149 | ||
3150 | ret = send_rename(sctx, from_path, to_path); | |
3151 | if (ret < 0) | |
3152 | goto out; | |
3153 | ||
9dc44214 FM |
3154 | if (rmdir_ino) { |
3155 | struct orphan_dir_info *odi; | |
3156 | ||
3157 | odi = get_orphan_dir_info(sctx, rmdir_ino); | |
3158 | if (!odi) { | |
3159 | /* already deleted */ | |
3160 | goto finish; | |
3161 | } | |
3162 | ret = can_rmdir(sctx, rmdir_ino, odi->gen, sctx->cur_ino + 1); | |
3163 | if (ret < 0) | |
3164 | goto out; | |
3165 | if (!ret) | |
3166 | goto finish; | |
3167 | ||
3168 | name = fs_path_alloc(); | |
3169 | if (!name) { | |
3170 | ret = -ENOMEM; | |
3171 | goto out; | |
3172 | } | |
3173 | ret = get_cur_path(sctx, rmdir_ino, odi->gen, name); | |
3174 | if (ret < 0) | |
3175 | goto out; | |
3176 | ret = send_rmdir(sctx, name); | |
3177 | if (ret < 0) | |
3178 | goto out; | |
3179 | free_orphan_dir_info(sctx, odi); | |
3180 | } | |
3181 | ||
3182 | finish: | |
9f03740a FDBM |
3183 | ret = send_utimes(sctx, pm->ino, pm->gen); |
3184 | if (ret < 0) | |
3185 | goto out; | |
3186 | ||
3187 | /* | |
3188 | * After rename/move, need to update the utimes of both new parent(s) | |
3189 | * and old parent(s). | |
3190 | */ | |
3191 | list_for_each_entry(cur, &pm->update_refs, list) { | |
9dc44214 FM |
3192 | if (cur->dir == rmdir_ino) |
3193 | continue; | |
9f03740a FDBM |
3194 | ret = send_utimes(sctx, cur->dir, cur->dir_gen); |
3195 | if (ret < 0) | |
3196 | goto out; | |
3197 | } | |
3198 | ||
3199 | out: | |
2b863a13 | 3200 | fs_path_free(name); |
9f03740a FDBM |
3201 | fs_path_free(from_path); |
3202 | fs_path_free(to_path); | |
3203 | sctx->send_progress = orig_progress; | |
3204 | ||
3205 | return ret; | |
3206 | } | |
3207 | ||
3208 | static void free_pending_move(struct send_ctx *sctx, struct pending_dir_move *m) | |
3209 | { | |
3210 | if (!list_empty(&m->list)) | |
3211 | list_del(&m->list); | |
3212 | if (!RB_EMPTY_NODE(&m->node)) | |
3213 | rb_erase(&m->node, &sctx->pending_dir_moves); | |
3214 | __free_recorded_refs(&m->update_refs); | |
3215 | kfree(m); | |
3216 | } | |
3217 | ||
3218 | static void tail_append_pending_moves(struct pending_dir_move *moves, | |
3219 | struct list_head *stack) | |
3220 | { | |
3221 | if (list_empty(&moves->list)) { | |
3222 | list_add_tail(&moves->list, stack); | |
3223 | } else { | |
3224 | LIST_HEAD(list); | |
3225 | list_splice_init(&moves->list, &list); | |
3226 | list_add_tail(&moves->list, stack); | |
3227 | list_splice_tail(&list, stack); | |
3228 | } | |
3229 | } | |
3230 | ||
3231 | static int apply_children_dir_moves(struct send_ctx *sctx) | |
3232 | { | |
3233 | struct pending_dir_move *pm; | |
3234 | struct list_head stack; | |
3235 | u64 parent_ino = sctx->cur_ino; | |
3236 | int ret = 0; | |
3237 | ||
3238 | pm = get_pending_dir_moves(sctx, parent_ino); | |
3239 | if (!pm) | |
3240 | return 0; | |
3241 | ||
3242 | INIT_LIST_HEAD(&stack); | |
3243 | tail_append_pending_moves(pm, &stack); | |
3244 | ||
3245 | while (!list_empty(&stack)) { | |
3246 | pm = list_first_entry(&stack, struct pending_dir_move, list); | |
3247 | parent_ino = pm->ino; | |
3248 | ret = apply_dir_move(sctx, pm); | |
3249 | free_pending_move(sctx, pm); | |
3250 | if (ret) | |
3251 | goto out; | |
3252 | pm = get_pending_dir_moves(sctx, parent_ino); | |
3253 | if (pm) | |
3254 | tail_append_pending_moves(pm, &stack); | |
3255 | } | |
3256 | return 0; | |
3257 | ||
3258 | out: | |
3259 | while (!list_empty(&stack)) { | |
3260 | pm = list_first_entry(&stack, struct pending_dir_move, list); | |
3261 | free_pending_move(sctx, pm); | |
3262 | } | |
3263 | return ret; | |
3264 | } | |
3265 | ||
84471e24 FM |
3266 | /* |
3267 | * We might need to delay a directory rename even when no ancestor directory | |
3268 | * (in the send root) with a higher inode number than ours (sctx->cur_ino) was | |
3269 | * renamed. This happens when we rename a directory to the old name (the name | |
3270 | * in the parent root) of some other unrelated directory that got its rename | |
3271 | * delayed due to some ancestor with higher number that got renamed. | |
3272 | * | |
3273 | * Example: | |
3274 | * | |
3275 | * Parent snapshot: | |
3276 | * . (ino 256) | |
3277 | * |---- a/ (ino 257) | |
3278 | * | |---- file (ino 260) | |
3279 | * | | |
3280 | * |---- b/ (ino 258) | |
3281 | * |---- c/ (ino 259) | |
3282 | * | |
3283 | * Send snapshot: | |
3284 | * . (ino 256) | |
3285 | * |---- a/ (ino 258) | |
3286 | * |---- x/ (ino 259) | |
3287 | * |---- y/ (ino 257) | |
3288 | * |----- file (ino 260) | |
3289 | * | |
3290 | * Here we can not rename 258 from 'b' to 'a' without the rename of inode 257 | |
3291 | * from 'a' to 'x/y' happening first, which in turn depends on the rename of | |
3292 | * inode 259 from 'c' to 'x'. So the order of rename commands the send stream | |
3293 | * must issue is: | |
3294 | * | |
3295 | * 1 - rename 259 from 'c' to 'x' | |
3296 | * 2 - rename 257 from 'a' to 'x/y' | |
3297 | * 3 - rename 258 from 'b' to 'a' | |
3298 | * | |
3299 | * Returns 1 if the rename of sctx->cur_ino needs to be delayed, 0 if it can | |
3300 | * be done right away and < 0 on error. | |
3301 | */ | |
3302 | static int wait_for_dest_dir_move(struct send_ctx *sctx, | |
3303 | struct recorded_ref *parent_ref, | |
3304 | const bool is_orphan) | |
3305 | { | |
3306 | struct btrfs_path *path; | |
3307 | struct btrfs_key key; | |
3308 | struct btrfs_key di_key; | |
3309 | struct btrfs_dir_item *di; | |
3310 | u64 left_gen; | |
3311 | u64 right_gen; | |
3312 | int ret = 0; | |
3313 | ||
3314 | if (RB_EMPTY_ROOT(&sctx->waiting_dir_moves)) | |
3315 | return 0; | |
3316 | ||
3317 | path = alloc_path_for_send(); | |
3318 | if (!path) | |
3319 | return -ENOMEM; | |
3320 | ||
3321 | key.objectid = parent_ref->dir; | |
3322 | key.type = BTRFS_DIR_ITEM_KEY; | |
3323 | key.offset = btrfs_name_hash(parent_ref->name, parent_ref->name_len); | |
3324 | ||
3325 | ret = btrfs_search_slot(NULL, sctx->parent_root, &key, path, 0, 0); | |
3326 | if (ret < 0) { | |
3327 | goto out; | |
3328 | } else if (ret > 0) { | |
3329 | ret = 0; | |
3330 | goto out; | |
3331 | } | |
3332 | ||
3333 | di = btrfs_match_dir_item_name(sctx->parent_root, path, | |
3334 | parent_ref->name, parent_ref->name_len); | |
3335 | if (!di) { | |
3336 | ret = 0; | |
3337 | goto out; | |
3338 | } | |
3339 | /* | |
3340 | * di_key.objectid has the number of the inode that has a dentry in the | |
3341 | * parent directory with the same name that sctx->cur_ino is being | |
3342 | * renamed to. We need to check if that inode is in the send root as | |
3343 | * well and if it is currently marked as an inode with a pending rename, | |
3344 | * if it is, we need to delay the rename of sctx->cur_ino as well, so | |
3345 | * that it happens after that other inode is renamed. | |
3346 | */ | |
3347 | btrfs_dir_item_key_to_cpu(path->nodes[0], di, &di_key); | |
3348 | if (di_key.type != BTRFS_INODE_ITEM_KEY) { | |
3349 | ret = 0; | |
3350 | goto out; | |
3351 | } | |
3352 | ||
3353 | ret = get_inode_info(sctx->parent_root, di_key.objectid, NULL, | |
3354 | &left_gen, NULL, NULL, NULL, NULL); | |
3355 | if (ret < 0) | |
3356 | goto out; | |
3357 | ret = get_inode_info(sctx->send_root, di_key.objectid, NULL, | |
3358 | &right_gen, NULL, NULL, NULL, NULL); | |
3359 | if (ret < 0) { | |
3360 | if (ret == -ENOENT) | |
3361 | ret = 0; | |
3362 | goto out; | |
3363 | } | |
3364 | ||
3365 | /* Different inode, no need to delay the rename of sctx->cur_ino */ | |
3366 | if (right_gen != left_gen) { | |
3367 | ret = 0; | |
3368 | goto out; | |
3369 | } | |
3370 | ||
3371 | if (is_waiting_for_move(sctx, di_key.objectid)) { | |
3372 | ret = add_pending_dir_move(sctx, | |
3373 | sctx->cur_ino, | |
3374 | sctx->cur_inode_gen, | |
3375 | di_key.objectid, | |
3376 | &sctx->new_refs, | |
3377 | &sctx->deleted_refs, | |
3378 | is_orphan); | |
3379 | if (!ret) | |
3380 | ret = 1; | |
3381 | } | |
3382 | out: | |
3383 | btrfs_free_path(path); | |
3384 | return ret; | |
3385 | } | |
3386 | ||
80aa6027 FM |
3387 | /* |
3388 | * Check if ino ino1 is an ancestor of inode ino2 in the given root. | |
3389 | * Return 1 if true, 0 if false and < 0 on error. | |
3390 | */ | |
3391 | static int is_ancestor(struct btrfs_root *root, | |
3392 | const u64 ino1, | |
3393 | const u64 ino1_gen, | |
3394 | const u64 ino2, | |
3395 | struct fs_path *fs_path) | |
3396 | { | |
3397 | u64 ino = ino2; | |
3398 | ||
3399 | while (ino > BTRFS_FIRST_FREE_OBJECTID) { | |
3400 | int ret; | |
3401 | u64 parent; | |
3402 | u64 parent_gen; | |
3403 | ||
3404 | fs_path_reset(fs_path); | |
3405 | ret = get_first_ref(root, ino, &parent, &parent_gen, fs_path); | |
3406 | if (ret < 0) { | |
3407 | if (ret == -ENOENT && ino == ino2) | |
3408 | ret = 0; | |
3409 | return ret; | |
3410 | } | |
3411 | if (parent == ino1) | |
3412 | return parent_gen == ino1_gen ? 1 : 0; | |
3413 | ino = parent; | |
3414 | } | |
3415 | return 0; | |
3416 | } | |
3417 | ||
9f03740a | 3418 | static int wait_for_parent_move(struct send_ctx *sctx, |
8b191a68 FM |
3419 | struct recorded_ref *parent_ref, |
3420 | const bool is_orphan) | |
9f03740a | 3421 | { |
f959492f | 3422 | int ret = 0; |
9f03740a FDBM |
3423 | u64 ino = parent_ref->dir; |
3424 | u64 parent_ino_before, parent_ino_after; | |
9f03740a FDBM |
3425 | struct fs_path *path_before = NULL; |
3426 | struct fs_path *path_after = NULL; | |
3427 | int len1, len2; | |
9f03740a FDBM |
3428 | |
3429 | path_after = fs_path_alloc(); | |
f959492f FM |
3430 | path_before = fs_path_alloc(); |
3431 | if (!path_after || !path_before) { | |
9f03740a FDBM |
3432 | ret = -ENOMEM; |
3433 | goto out; | |
3434 | } | |
3435 | ||
bfa7e1f8 | 3436 | /* |
f959492f FM |
3437 | * Our current directory inode may not yet be renamed/moved because some |
3438 | * ancestor (immediate or not) has to be renamed/moved first. So find if | |
3439 | * such ancestor exists and make sure our own rename/move happens after | |
80aa6027 FM |
3440 | * that ancestor is processed to avoid path build infinite loops (done |
3441 | * at get_cur_path()). | |
bfa7e1f8 | 3442 | */ |
f959492f FM |
3443 | while (ino > BTRFS_FIRST_FREE_OBJECTID) { |
3444 | if (is_waiting_for_move(sctx, ino)) { | |
80aa6027 FM |
3445 | /* |
3446 | * If the current inode is an ancestor of ino in the | |
3447 | * parent root, we need to delay the rename of the | |
3448 | * current inode, otherwise don't delayed the rename | |
3449 | * because we can end up with a circular dependency | |
3450 | * of renames, resulting in some directories never | |
3451 | * getting the respective rename operations issued in | |
3452 | * the send stream or getting into infinite path build | |
3453 | * loops. | |
3454 | */ | |
3455 | ret = is_ancestor(sctx->parent_root, | |
3456 | sctx->cur_ino, sctx->cur_inode_gen, | |
3457 | ino, path_before); | |
f959492f FM |
3458 | break; |
3459 | } | |
bfa7e1f8 FM |
3460 | |
3461 | fs_path_reset(path_before); | |
3462 | fs_path_reset(path_after); | |
3463 | ||
3464 | ret = get_first_ref(sctx->send_root, ino, &parent_ino_after, | |
f959492f | 3465 | NULL, path_after); |
bfa7e1f8 FM |
3466 | if (ret < 0) |
3467 | goto out; | |
3468 | ret = get_first_ref(sctx->parent_root, ino, &parent_ino_before, | |
3469 | NULL, path_before); | |
f959492f | 3470 | if (ret < 0 && ret != -ENOENT) { |
bfa7e1f8 | 3471 | goto out; |
f959492f | 3472 | } else if (ret == -ENOENT) { |
bf8e8ca6 | 3473 | ret = 0; |
f959492f | 3474 | break; |
bfa7e1f8 FM |
3475 | } |
3476 | ||
3477 | len1 = fs_path_len(path_before); | |
3478 | len2 = fs_path_len(path_after); | |
f959492f FM |
3479 | if (ino > sctx->cur_ino && |
3480 | (parent_ino_before != parent_ino_after || len1 != len2 || | |
3481 | memcmp(path_before->start, path_after->start, len1))) { | |
bfa7e1f8 | 3482 | ret = 1; |
f959492f | 3483 | break; |
bfa7e1f8 | 3484 | } |
bfa7e1f8 | 3485 | ino = parent_ino_after; |
bfa7e1f8 FM |
3486 | } |
3487 | ||
9f03740a FDBM |
3488 | out: |
3489 | fs_path_free(path_before); | |
3490 | fs_path_free(path_after); | |
3491 | ||
f959492f FM |
3492 | if (ret == 1) { |
3493 | ret = add_pending_dir_move(sctx, | |
3494 | sctx->cur_ino, | |
3495 | sctx->cur_inode_gen, | |
3496 | ino, | |
3497 | &sctx->new_refs, | |
84471e24 | 3498 | &sctx->deleted_refs, |
8b191a68 | 3499 | is_orphan); |
f959492f FM |
3500 | if (!ret) |
3501 | ret = 1; | |
3502 | } | |
3503 | ||
9f03740a FDBM |
3504 | return ret; |
3505 | } | |
3506 | ||
31db9f7c AB |
3507 | /* |
3508 | * This does all the move/link/unlink/rmdir magic. | |
3509 | */ | |
9f03740a | 3510 | static int process_recorded_refs(struct send_ctx *sctx, int *pending_move) |
31db9f7c AB |
3511 | { |
3512 | int ret = 0; | |
3513 | struct recorded_ref *cur; | |
1f4692da | 3514 | struct recorded_ref *cur2; |
ba5e8f2e | 3515 | struct list_head check_dirs; |
31db9f7c | 3516 | struct fs_path *valid_path = NULL; |
b24baf69 | 3517 | u64 ow_inode = 0; |
31db9f7c AB |
3518 | u64 ow_gen; |
3519 | int did_overwrite = 0; | |
3520 | int is_orphan = 0; | |
29d6d30f | 3521 | u64 last_dir_ino_rm = 0; |
84471e24 | 3522 | bool can_rename = true; |
31db9f7c AB |
3523 | |
3524 | verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino); | |
3525 | ||
6d85ed05 AB |
3526 | /* |
3527 | * This should never happen as the root dir always has the same ref | |
3528 | * which is always '..' | |
3529 | */ | |
3530 | BUG_ON(sctx->cur_ino <= BTRFS_FIRST_FREE_OBJECTID); | |
ba5e8f2e | 3531 | INIT_LIST_HEAD(&check_dirs); |
6d85ed05 | 3532 | |
924794c9 | 3533 | valid_path = fs_path_alloc(); |
31db9f7c AB |
3534 | if (!valid_path) { |
3535 | ret = -ENOMEM; | |
3536 | goto out; | |
3537 | } | |
3538 | ||
31db9f7c AB |
3539 | /* |
3540 | * First, check if the first ref of the current inode was overwritten | |
3541 | * before. If yes, we know that the current inode was already orphanized | |
3542 | * and thus use the orphan name. If not, we can use get_cur_path to | |
3543 | * get the path of the first ref as it would like while receiving at | |
3544 | * this point in time. | |
3545 | * New inodes are always orphan at the beginning, so force to use the | |
3546 | * orphan name in this case. | |
3547 | * The first ref is stored in valid_path and will be updated if it | |
3548 | * gets moved around. | |
3549 | */ | |
3550 | if (!sctx->cur_inode_new) { | |
3551 | ret = did_overwrite_first_ref(sctx, sctx->cur_ino, | |
3552 | sctx->cur_inode_gen); | |
3553 | if (ret < 0) | |
3554 | goto out; | |
3555 | if (ret) | |
3556 | did_overwrite = 1; | |
3557 | } | |
3558 | if (sctx->cur_inode_new || did_overwrite) { | |
3559 | ret = gen_unique_name(sctx, sctx->cur_ino, | |
3560 | sctx->cur_inode_gen, valid_path); | |
3561 | if (ret < 0) | |
3562 | goto out; | |
3563 | is_orphan = 1; | |
3564 | } else { | |
3565 | ret = get_cur_path(sctx, sctx->cur_ino, sctx->cur_inode_gen, | |
3566 | valid_path); | |
3567 | if (ret < 0) | |
3568 | goto out; | |
3569 | } | |
3570 | ||
3571 | list_for_each_entry(cur, &sctx->new_refs, list) { | |
1f4692da AB |
3572 | /* |
3573 | * We may have refs where the parent directory does not exist | |
3574 | * yet. This happens if the parent directories inum is higher | |
3575 | * the the current inum. To handle this case, we create the | |
3576 | * parent directory out of order. But we need to check if this | |
3577 | * did already happen before due to other refs in the same dir. | |
3578 | */ | |
3579 | ret = get_cur_inode_state(sctx, cur->dir, cur->dir_gen); | |
3580 | if (ret < 0) | |
3581 | goto out; | |
3582 | if (ret == inode_state_will_create) { | |
3583 | ret = 0; | |
3584 | /* | |
3585 | * First check if any of the current inodes refs did | |
3586 | * already create the dir. | |
3587 | */ | |
3588 | list_for_each_entry(cur2, &sctx->new_refs, list) { | |
3589 | if (cur == cur2) | |
3590 | break; | |
3591 | if (cur2->dir == cur->dir) { | |
3592 | ret = 1; | |
3593 | break; | |
3594 | } | |
3595 | } | |
3596 | ||
3597 | /* | |
3598 | * If that did not happen, check if a previous inode | |
3599 | * did already create the dir. | |
3600 | */ | |
3601 | if (!ret) | |
3602 | ret = did_create_dir(sctx, cur->dir); | |
3603 | if (ret < 0) | |
3604 | goto out; | |
3605 | if (!ret) { | |
3606 | ret = send_create_inode(sctx, cur->dir); | |
3607 | if (ret < 0) | |
3608 | goto out; | |
3609 | } | |
3610 | } | |
3611 | ||
31db9f7c AB |
3612 | /* |
3613 | * Check if this new ref would overwrite the first ref of | |
3614 | * another unprocessed inode. If yes, orphanize the | |
3615 | * overwritten inode. If we find an overwritten ref that is | |
3616 | * not the first ref, simply unlink it. | |
3617 | */ | |
3618 | ret = will_overwrite_ref(sctx, cur->dir, cur->dir_gen, | |
3619 | cur->name, cur->name_len, | |
3620 | &ow_inode, &ow_gen); | |
3621 | if (ret < 0) | |
3622 | goto out; | |
3623 | if (ret) { | |
924794c9 TI |
3624 | ret = is_first_ref(sctx->parent_root, |
3625 | ow_inode, cur->dir, cur->name, | |
3626 | cur->name_len); | |
31db9f7c AB |
3627 | if (ret < 0) |
3628 | goto out; | |
3629 | if (ret) { | |
8996a48c FM |
3630 | struct name_cache_entry *nce; |
3631 | ||
31db9f7c AB |
3632 | ret = orphanize_inode(sctx, ow_inode, ow_gen, |
3633 | cur->full_path); | |
3634 | if (ret < 0) | |
3635 | goto out; | |
8996a48c FM |
3636 | /* |
3637 | * Make sure we clear our orphanized inode's | |
3638 | * name from the name cache. This is because the | |
3639 | * inode ow_inode might be an ancestor of some | |
3640 | * other inode that will be orphanized as well | |
3641 | * later and has an inode number greater than | |
3642 | * sctx->send_progress. We need to prevent | |
3643 | * future name lookups from using the old name | |
3644 | * and get instead the orphan name. | |
3645 | */ | |
3646 | nce = name_cache_search(sctx, ow_inode, ow_gen); | |
3647 | if (nce) { | |
3648 | name_cache_delete(sctx, nce); | |
3649 | kfree(nce); | |
3650 | } | |
31db9f7c AB |
3651 | } else { |
3652 | ret = send_unlink(sctx, cur->full_path); | |
3653 | if (ret < 0) | |
3654 | goto out; | |
3655 | } | |
3656 | } | |
3657 | ||
84471e24 FM |
3658 | if (S_ISDIR(sctx->cur_inode_mode) && sctx->parent_root) { |
3659 | ret = wait_for_dest_dir_move(sctx, cur, is_orphan); | |
3660 | if (ret < 0) | |
3661 | goto out; | |
3662 | if (ret == 1) { | |
3663 | can_rename = false; | |
3664 | *pending_move = 1; | |
3665 | } | |
3666 | } | |
3667 | ||
8b191a68 FM |
3668 | if (S_ISDIR(sctx->cur_inode_mode) && sctx->parent_root && |
3669 | can_rename) { | |
3670 | ret = wait_for_parent_move(sctx, cur, is_orphan); | |
3671 | if (ret < 0) | |
3672 | goto out; | |
3673 | if (ret == 1) { | |
3674 | can_rename = false; | |
3675 | *pending_move = 1; | |
3676 | } | |
3677 | } | |
3678 | ||
31db9f7c AB |
3679 | /* |
3680 | * link/move the ref to the new place. If we have an orphan | |
3681 | * inode, move it and update valid_path. If not, link or move | |
3682 | * it depending on the inode mode. | |
3683 | */ | |
84471e24 | 3684 | if (is_orphan && can_rename) { |
31db9f7c AB |
3685 | ret = send_rename(sctx, valid_path, cur->full_path); |
3686 | if (ret < 0) | |
3687 | goto out; | |
3688 | is_orphan = 0; | |
3689 | ret = fs_path_copy(valid_path, cur->full_path); | |
3690 | if (ret < 0) | |
3691 | goto out; | |
84471e24 | 3692 | } else if (can_rename) { |
31db9f7c AB |
3693 | if (S_ISDIR(sctx->cur_inode_mode)) { |
3694 | /* | |
3695 | * Dirs can't be linked, so move it. For moved | |
3696 | * dirs, we always have one new and one deleted | |
3697 | * ref. The deleted ref is ignored later. | |
3698 | */ | |
8b191a68 FM |
3699 | ret = send_rename(sctx, valid_path, |
3700 | cur->full_path); | |
3701 | if (!ret) | |
3702 | ret = fs_path_copy(valid_path, | |
3703 | cur->full_path); | |
31db9f7c AB |
3704 | if (ret < 0) |
3705 | goto out; | |
3706 | } else { | |
3707 | ret = send_link(sctx, cur->full_path, | |
3708 | valid_path); | |
3709 | if (ret < 0) | |
3710 | goto out; | |
3711 | } | |
3712 | } | |
ba5e8f2e | 3713 | ret = dup_ref(cur, &check_dirs); |
31db9f7c AB |
3714 | if (ret < 0) |
3715 | goto out; | |
3716 | } | |
3717 | ||
3718 | if (S_ISDIR(sctx->cur_inode_mode) && sctx->cur_inode_deleted) { | |
3719 | /* | |
3720 | * Check if we can already rmdir the directory. If not, | |
3721 | * orphanize it. For every dir item inside that gets deleted | |
3722 | * later, we do this check again and rmdir it then if possible. | |
3723 | * See the use of check_dirs for more details. | |
3724 | */ | |
9dc44214 FM |
3725 | ret = can_rmdir(sctx, sctx->cur_ino, sctx->cur_inode_gen, |
3726 | sctx->cur_ino); | |
31db9f7c AB |
3727 | if (ret < 0) |
3728 | goto out; | |
3729 | if (ret) { | |
3730 | ret = send_rmdir(sctx, valid_path); | |
3731 | if (ret < 0) | |
3732 | goto out; | |
3733 | } else if (!is_orphan) { | |
3734 | ret = orphanize_inode(sctx, sctx->cur_ino, | |
3735 | sctx->cur_inode_gen, valid_path); | |
3736 | if (ret < 0) | |
3737 | goto out; | |
3738 | is_orphan = 1; | |
3739 | } | |
3740 | ||
3741 | list_for_each_entry(cur, &sctx->deleted_refs, list) { | |
ba5e8f2e | 3742 | ret = dup_ref(cur, &check_dirs); |
31db9f7c AB |
3743 | if (ret < 0) |
3744 | goto out; | |
3745 | } | |
ccf1626b AB |
3746 | } else if (S_ISDIR(sctx->cur_inode_mode) && |
3747 | !list_empty(&sctx->deleted_refs)) { | |
3748 | /* | |
3749 | * We have a moved dir. Add the old parent to check_dirs | |
3750 | */ | |
3751 | cur = list_entry(sctx->deleted_refs.next, struct recorded_ref, | |
3752 | list); | |
ba5e8f2e | 3753 | ret = dup_ref(cur, &check_dirs); |
ccf1626b AB |
3754 | if (ret < 0) |
3755 | goto out; | |
31db9f7c AB |
3756 | } else if (!S_ISDIR(sctx->cur_inode_mode)) { |
3757 | /* | |
3758 | * We have a non dir inode. Go through all deleted refs and | |
3759 | * unlink them if they were not already overwritten by other | |
3760 | * inodes. | |
3761 | */ | |
3762 | list_for_each_entry(cur, &sctx->deleted_refs, list) { | |
3763 | ret = did_overwrite_ref(sctx, cur->dir, cur->dir_gen, | |
3764 | sctx->cur_ino, sctx->cur_inode_gen, | |
3765 | cur->name, cur->name_len); | |
3766 | if (ret < 0) | |
3767 | goto out; | |
3768 | if (!ret) { | |
1f4692da AB |
3769 | ret = send_unlink(sctx, cur->full_path); |
3770 | if (ret < 0) | |
3771 | goto out; | |
31db9f7c | 3772 | } |
ba5e8f2e | 3773 | ret = dup_ref(cur, &check_dirs); |
31db9f7c AB |
3774 | if (ret < 0) |
3775 | goto out; | |
3776 | } | |
31db9f7c AB |
3777 | /* |
3778 | * If the inode is still orphan, unlink the orphan. This may | |
3779 | * happen when a previous inode did overwrite the first ref | |
3780 | * of this inode and no new refs were added for the current | |
766702ef AB |
3781 | * inode. Unlinking does not mean that the inode is deleted in |
3782 | * all cases. There may still be links to this inode in other | |
3783 | * places. | |
31db9f7c | 3784 | */ |
1f4692da | 3785 | if (is_orphan) { |
31db9f7c AB |
3786 | ret = send_unlink(sctx, valid_path); |
3787 | if (ret < 0) | |
3788 | goto out; | |
3789 | } | |
3790 | } | |
3791 | ||
3792 | /* | |
3793 | * We did collect all parent dirs where cur_inode was once located. We | |
3794 | * now go through all these dirs and check if they are pending for | |
3795 | * deletion and if it's finally possible to perform the rmdir now. | |
3796 | * We also update the inode stats of the parent dirs here. | |
3797 | */ | |
ba5e8f2e | 3798 | list_for_each_entry(cur, &check_dirs, list) { |
766702ef AB |
3799 | /* |
3800 | * In case we had refs into dirs that were not processed yet, | |
3801 | * we don't need to do the utime and rmdir logic for these dirs. | |
3802 | * The dir will be processed later. | |
3803 | */ | |
ba5e8f2e | 3804 | if (cur->dir > sctx->cur_ino) |
31db9f7c AB |
3805 | continue; |
3806 | ||
ba5e8f2e | 3807 | ret = get_cur_inode_state(sctx, cur->dir, cur->dir_gen); |
31db9f7c AB |
3808 | if (ret < 0) |
3809 | goto out; | |
3810 | ||
3811 | if (ret == inode_state_did_create || | |
3812 | ret == inode_state_no_change) { | |
3813 | /* TODO delayed utimes */ | |
ba5e8f2e | 3814 | ret = send_utimes(sctx, cur->dir, cur->dir_gen); |
31db9f7c AB |
3815 | if (ret < 0) |
3816 | goto out; | |
29d6d30f FM |
3817 | } else if (ret == inode_state_did_delete && |
3818 | cur->dir != last_dir_ino_rm) { | |
9dc44214 FM |
3819 | ret = can_rmdir(sctx, cur->dir, cur->dir_gen, |
3820 | sctx->cur_ino); | |
31db9f7c AB |
3821 | if (ret < 0) |
3822 | goto out; | |
3823 | if (ret) { | |
ba5e8f2e JB |
3824 | ret = get_cur_path(sctx, cur->dir, |
3825 | cur->dir_gen, valid_path); | |
31db9f7c AB |
3826 | if (ret < 0) |
3827 | goto out; | |
3828 | ret = send_rmdir(sctx, valid_path); | |
3829 | if (ret < 0) | |
3830 | goto out; | |
29d6d30f | 3831 | last_dir_ino_rm = cur->dir; |
31db9f7c AB |
3832 | } |
3833 | } | |
3834 | } | |
3835 | ||
31db9f7c AB |
3836 | ret = 0; |
3837 | ||
3838 | out: | |
ba5e8f2e | 3839 | __free_recorded_refs(&check_dirs); |
31db9f7c | 3840 | free_recorded_refs(sctx); |
924794c9 | 3841 | fs_path_free(valid_path); |
31db9f7c AB |
3842 | return ret; |
3843 | } | |
3844 | ||
a4d96d62 LB |
3845 | static int record_ref(struct btrfs_root *root, int num, u64 dir, int index, |
3846 | struct fs_path *name, void *ctx, struct list_head *refs) | |
31db9f7c AB |
3847 | { |
3848 | int ret = 0; | |
3849 | struct send_ctx *sctx = ctx; | |
3850 | struct fs_path *p; | |
3851 | u64 gen; | |
3852 | ||
924794c9 | 3853 | p = fs_path_alloc(); |
31db9f7c AB |
3854 | if (!p) |
3855 | return -ENOMEM; | |
3856 | ||
a4d96d62 | 3857 | ret = get_inode_info(root, dir, NULL, &gen, NULL, NULL, |
85a7b33b | 3858 | NULL, NULL); |
31db9f7c AB |
3859 | if (ret < 0) |
3860 | goto out; | |
3861 | ||
31db9f7c AB |
3862 | ret = get_cur_path(sctx, dir, gen, p); |
3863 | if (ret < 0) | |
3864 | goto out; | |
3865 | ret = fs_path_add_path(p, name); | |
3866 | if (ret < 0) | |
3867 | goto out; | |
3868 | ||
a4d96d62 | 3869 | ret = __record_ref(refs, dir, gen, p); |
31db9f7c AB |
3870 | |
3871 | out: | |
3872 | if (ret) | |
924794c9 | 3873 | fs_path_free(p); |
31db9f7c AB |
3874 | return ret; |
3875 | } | |
3876 | ||
a4d96d62 LB |
3877 | static int __record_new_ref(int num, u64 dir, int index, |
3878 | struct fs_path *name, | |
3879 | void *ctx) | |
3880 | { | |
3881 | struct send_ctx *sctx = ctx; | |
3882 | return record_ref(sctx->send_root, num, dir, index, name, | |
3883 | ctx, &sctx->new_refs); | |
3884 | } | |
3885 | ||
3886 | ||
31db9f7c AB |
3887 | static int __record_deleted_ref(int num, u64 dir, int index, |
3888 | struct fs_path *name, | |
3889 | void *ctx) | |
3890 | { | |
31db9f7c | 3891 | struct send_ctx *sctx = ctx; |
a4d96d62 LB |
3892 | return record_ref(sctx->parent_root, num, dir, index, name, |
3893 | ctx, &sctx->deleted_refs); | |
31db9f7c AB |
3894 | } |
3895 | ||
3896 | static int record_new_ref(struct send_ctx *sctx) | |
3897 | { | |
3898 | int ret; | |
3899 | ||
924794c9 TI |
3900 | ret = iterate_inode_ref(sctx->send_root, sctx->left_path, |
3901 | sctx->cmp_key, 0, __record_new_ref, sctx); | |
31db9f7c AB |
3902 | if (ret < 0) |
3903 | goto out; | |
3904 | ret = 0; | |
3905 | ||
3906 | out: | |
3907 | return ret; | |
3908 | } | |
3909 | ||
3910 | static int record_deleted_ref(struct send_ctx *sctx) | |
3911 | { | |
3912 | int ret; | |
3913 | ||
924794c9 TI |
3914 | ret = iterate_inode_ref(sctx->parent_root, sctx->right_path, |
3915 | sctx->cmp_key, 0, __record_deleted_ref, sctx); | |
31db9f7c AB |
3916 | if (ret < 0) |
3917 | goto out; | |
3918 | ret = 0; | |
3919 | ||
3920 | out: | |
3921 | return ret; | |
3922 | } | |
3923 | ||
3924 | struct find_ref_ctx { | |
3925 | u64 dir; | |
ba5e8f2e JB |
3926 | u64 dir_gen; |
3927 | struct btrfs_root *root; | |
31db9f7c AB |
3928 | struct fs_path *name; |
3929 | int found_idx; | |
3930 | }; | |
3931 | ||
3932 | static int __find_iref(int num, u64 dir, int index, | |
3933 | struct fs_path *name, | |
3934 | void *ctx_) | |
3935 | { | |
3936 | struct find_ref_ctx *ctx = ctx_; | |
ba5e8f2e JB |
3937 | u64 dir_gen; |
3938 | int ret; | |
31db9f7c AB |
3939 | |
3940 | if (dir == ctx->dir && fs_path_len(name) == fs_path_len(ctx->name) && | |
3941 | strncmp(name->start, ctx->name->start, fs_path_len(name)) == 0) { | |
ba5e8f2e JB |
3942 | /* |
3943 | * To avoid doing extra lookups we'll only do this if everything | |
3944 | * else matches. | |
3945 | */ | |
3946 | ret = get_inode_info(ctx->root, dir, NULL, &dir_gen, NULL, | |
3947 | NULL, NULL, NULL); | |
3948 | if (ret) | |
3949 | return ret; | |
3950 | if (dir_gen != ctx->dir_gen) | |
3951 | return 0; | |
31db9f7c AB |
3952 | ctx->found_idx = num; |
3953 | return 1; | |
3954 | } | |
3955 | return 0; | |
3956 | } | |
3957 | ||
924794c9 | 3958 | static int find_iref(struct btrfs_root *root, |
31db9f7c AB |
3959 | struct btrfs_path *path, |
3960 | struct btrfs_key *key, | |
ba5e8f2e | 3961 | u64 dir, u64 dir_gen, struct fs_path *name) |
31db9f7c AB |
3962 | { |
3963 | int ret; | |
3964 | struct find_ref_ctx ctx; | |
3965 | ||
3966 | ctx.dir = dir; | |
3967 | ctx.name = name; | |
ba5e8f2e | 3968 | ctx.dir_gen = dir_gen; |
31db9f7c | 3969 | ctx.found_idx = -1; |
ba5e8f2e | 3970 | ctx.root = root; |
31db9f7c | 3971 | |
924794c9 | 3972 | ret = iterate_inode_ref(root, path, key, 0, __find_iref, &ctx); |
31db9f7c AB |
3973 | if (ret < 0) |
3974 | return ret; | |
3975 | ||
3976 | if (ctx.found_idx == -1) | |
3977 | return -ENOENT; | |
3978 | ||
3979 | return ctx.found_idx; | |
3980 | } | |
3981 | ||
3982 | static int __record_changed_new_ref(int num, u64 dir, int index, | |
3983 | struct fs_path *name, | |
3984 | void *ctx) | |
3985 | { | |
ba5e8f2e | 3986 | u64 dir_gen; |
31db9f7c AB |
3987 | int ret; |
3988 | struct send_ctx *sctx = ctx; | |
3989 | ||
ba5e8f2e JB |
3990 | ret = get_inode_info(sctx->send_root, dir, NULL, &dir_gen, NULL, |
3991 | NULL, NULL, NULL); | |
3992 | if (ret) | |
3993 | return ret; | |
3994 | ||
924794c9 | 3995 | ret = find_iref(sctx->parent_root, sctx->right_path, |
ba5e8f2e | 3996 | sctx->cmp_key, dir, dir_gen, name); |
31db9f7c AB |
3997 | if (ret == -ENOENT) |
3998 | ret = __record_new_ref(num, dir, index, name, sctx); | |
3999 | else if (ret > 0) | |
4000 | ret = 0; | |
4001 | ||
4002 | return ret; | |
4003 | } | |
4004 | ||
4005 | static int __record_changed_deleted_ref(int num, u64 dir, int index, | |
4006 | struct fs_path *name, | |
4007 | void *ctx) | |
4008 | { | |
ba5e8f2e | 4009 | u64 dir_gen; |
31db9f7c AB |
4010 | int ret; |
4011 | struct send_ctx *sctx = ctx; | |
4012 | ||
ba5e8f2e JB |
4013 | ret = get_inode_info(sctx->parent_root, dir, NULL, &dir_gen, NULL, |
4014 | NULL, NULL, NULL); | |
4015 | if (ret) | |
4016 | return ret; | |
4017 | ||
924794c9 | 4018 | ret = find_iref(sctx->send_root, sctx->left_path, sctx->cmp_key, |
ba5e8f2e | 4019 | dir, dir_gen, name); |
31db9f7c AB |
4020 | if (ret == -ENOENT) |
4021 | ret = __record_deleted_ref(num, dir, index, name, sctx); | |
4022 | else if (ret > 0) | |
4023 | ret = 0; | |
4024 | ||
4025 | return ret; | |
4026 | } | |
4027 | ||
4028 | static int record_changed_ref(struct send_ctx *sctx) | |
4029 | { | |
4030 | int ret = 0; | |
4031 | ||
924794c9 | 4032 | ret = iterate_inode_ref(sctx->send_root, sctx->left_path, |
31db9f7c AB |
4033 | sctx->cmp_key, 0, __record_changed_new_ref, sctx); |
4034 | if (ret < 0) | |
4035 | goto out; | |
924794c9 | 4036 | ret = iterate_inode_ref(sctx->parent_root, sctx->right_path, |
31db9f7c AB |
4037 | sctx->cmp_key, 0, __record_changed_deleted_ref, sctx); |
4038 | if (ret < 0) | |
4039 | goto out; | |
4040 | ret = 0; | |
4041 | ||
4042 | out: | |
4043 | return ret; | |
4044 | } | |
4045 | ||
4046 | /* | |
4047 | * Record and process all refs at once. Needed when an inode changes the | |
4048 | * generation number, which means that it was deleted and recreated. | |
4049 | */ | |
4050 | static int process_all_refs(struct send_ctx *sctx, | |
4051 | enum btrfs_compare_tree_result cmd) | |
4052 | { | |
4053 | int ret; | |
4054 | struct btrfs_root *root; | |
4055 | struct btrfs_path *path; | |
4056 | struct btrfs_key key; | |
4057 | struct btrfs_key found_key; | |
4058 | struct extent_buffer *eb; | |
4059 | int slot; | |
4060 | iterate_inode_ref_t cb; | |
9f03740a | 4061 | int pending_move = 0; |
31db9f7c AB |
4062 | |
4063 | path = alloc_path_for_send(); | |
4064 | if (!path) | |
4065 | return -ENOMEM; | |
4066 | ||
4067 | if (cmd == BTRFS_COMPARE_TREE_NEW) { | |
4068 | root = sctx->send_root; | |
4069 | cb = __record_new_ref; | |
4070 | } else if (cmd == BTRFS_COMPARE_TREE_DELETED) { | |
4071 | root = sctx->parent_root; | |
4072 | cb = __record_deleted_ref; | |
4073 | } else { | |
4d1a63b2 DS |
4074 | btrfs_err(sctx->send_root->fs_info, |
4075 | "Wrong command %d in process_all_refs", cmd); | |
4076 | ret = -EINVAL; | |
4077 | goto out; | |
31db9f7c AB |
4078 | } |
4079 | ||
4080 | key.objectid = sctx->cmp_key->objectid; | |
4081 | key.type = BTRFS_INODE_REF_KEY; | |
4082 | key.offset = 0; | |
dff6d0ad FDBM |
4083 | ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); |
4084 | if (ret < 0) | |
4085 | goto out; | |
31db9f7c | 4086 | |
dff6d0ad | 4087 | while (1) { |
31db9f7c AB |
4088 | eb = path->nodes[0]; |
4089 | slot = path->slots[0]; | |
dff6d0ad FDBM |
4090 | if (slot >= btrfs_header_nritems(eb)) { |
4091 | ret = btrfs_next_leaf(root, path); | |
4092 | if (ret < 0) | |
4093 | goto out; | |
4094 | else if (ret > 0) | |
4095 | break; | |
4096 | continue; | |
4097 | } | |
4098 | ||
31db9f7c AB |
4099 | btrfs_item_key_to_cpu(eb, &found_key, slot); |
4100 | ||
4101 | if (found_key.objectid != key.objectid || | |
96b5bd77 JS |
4102 | (found_key.type != BTRFS_INODE_REF_KEY && |
4103 | found_key.type != BTRFS_INODE_EXTREF_KEY)) | |
31db9f7c | 4104 | break; |
31db9f7c | 4105 | |
924794c9 | 4106 | ret = iterate_inode_ref(root, path, &found_key, 0, cb, sctx); |
31db9f7c AB |
4107 | if (ret < 0) |
4108 | goto out; | |
4109 | ||
dff6d0ad | 4110 | path->slots[0]++; |
31db9f7c | 4111 | } |
e938c8ad | 4112 | btrfs_release_path(path); |
31db9f7c | 4113 | |
9f03740a FDBM |
4114 | ret = process_recorded_refs(sctx, &pending_move); |
4115 | /* Only applicable to an incremental send. */ | |
4116 | ASSERT(pending_move == 0); | |
31db9f7c AB |
4117 | |
4118 | out: | |
4119 | btrfs_free_path(path); | |
4120 | return ret; | |
4121 | } | |
4122 | ||
4123 | static int send_set_xattr(struct send_ctx *sctx, | |
4124 | struct fs_path *path, | |
4125 | const char *name, int name_len, | |
4126 | const char *data, int data_len) | |
4127 | { | |
4128 | int ret = 0; | |
4129 | ||
4130 | ret = begin_cmd(sctx, BTRFS_SEND_C_SET_XATTR); | |
4131 | if (ret < 0) | |
4132 | goto out; | |
4133 | ||
4134 | TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, path); | |
4135 | TLV_PUT_STRING(sctx, BTRFS_SEND_A_XATTR_NAME, name, name_len); | |
4136 | TLV_PUT(sctx, BTRFS_SEND_A_XATTR_DATA, data, data_len); | |
4137 | ||
4138 | ret = send_cmd(sctx); | |
4139 | ||
4140 | tlv_put_failure: | |
4141 | out: | |
4142 | return ret; | |
4143 | } | |
4144 | ||
4145 | static int send_remove_xattr(struct send_ctx *sctx, | |
4146 | struct fs_path *path, | |
4147 | const char *name, int name_len) | |
4148 | { | |
4149 | int ret = 0; | |
4150 | ||
4151 | ret = begin_cmd(sctx, BTRFS_SEND_C_REMOVE_XATTR); | |
4152 | if (ret < 0) | |
4153 | goto out; | |
4154 | ||
4155 | TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, path); | |
4156 | TLV_PUT_STRING(sctx, BTRFS_SEND_A_XATTR_NAME, name, name_len); | |
4157 | ||
4158 | ret = send_cmd(sctx); | |
4159 | ||
4160 | tlv_put_failure: | |
4161 | out: | |
4162 | return ret; | |
4163 | } | |
4164 | ||
4165 | static int __process_new_xattr(int num, struct btrfs_key *di_key, | |
4166 | const char *name, int name_len, | |
4167 | const char *data, int data_len, | |
4168 | u8 type, void *ctx) | |
4169 | { | |
4170 | int ret; | |
4171 | struct send_ctx *sctx = ctx; | |
4172 | struct fs_path *p; | |
4173 | posix_acl_xattr_header dummy_acl; | |
4174 | ||
924794c9 | 4175 | p = fs_path_alloc(); |
31db9f7c AB |
4176 | if (!p) |
4177 | return -ENOMEM; | |
4178 | ||
4179 | /* | |
4180 | * This hack is needed because empty acl's are stored as zero byte | |
4181 | * data in xattrs. Problem with that is, that receiving these zero byte | |
4182 | * acl's will fail later. To fix this, we send a dummy acl list that | |
4183 | * only contains the version number and no entries. | |
4184 | */ | |
4185 | if (!strncmp(name, XATTR_NAME_POSIX_ACL_ACCESS, name_len) || | |
4186 | !strncmp(name, XATTR_NAME_POSIX_ACL_DEFAULT, name_len)) { | |
4187 | if (data_len == 0) { | |
4188 | dummy_acl.a_version = | |
4189 | cpu_to_le32(POSIX_ACL_XATTR_VERSION); | |
4190 | data = (char *)&dummy_acl; | |
4191 | data_len = sizeof(dummy_acl); | |
4192 | } | |
4193 | } | |
4194 | ||
4195 | ret = get_cur_path(sctx, sctx->cur_ino, sctx->cur_inode_gen, p); | |
4196 | if (ret < 0) | |
4197 | goto out; | |
4198 | ||
4199 | ret = send_set_xattr(sctx, p, name, name_len, data, data_len); | |
4200 | ||
4201 | out: | |
924794c9 | 4202 | fs_path_free(p); |
31db9f7c AB |
4203 | return ret; |
4204 | } | |
4205 | ||
4206 | static int __process_deleted_xattr(int num, struct btrfs_key *di_key, | |
4207 | const char *name, int name_len, | |
4208 | const char *data, int data_len, | |
4209 | u8 type, void *ctx) | |
4210 | { | |
4211 | int ret; | |
4212 | struct send_ctx *sctx = ctx; | |
4213 | struct fs_path *p; | |
4214 | ||
924794c9 | 4215 | p = fs_path_alloc(); |
31db9f7c AB |
4216 | if (!p) |
4217 | return -ENOMEM; | |
4218 | ||
4219 | ret = get_cur_path(sctx, sctx->cur_ino, sctx->cur_inode_gen, p); | |
4220 | if (ret < 0) | |
4221 | goto out; | |
4222 | ||
4223 | ret = send_remove_xattr(sctx, p, name, name_len); | |
4224 | ||
4225 | out: | |
924794c9 | 4226 | fs_path_free(p); |
31db9f7c AB |
4227 | return ret; |
4228 | } | |
4229 | ||
4230 | static int process_new_xattr(struct send_ctx *sctx) | |
4231 | { | |
4232 | int ret = 0; | |
4233 | ||
924794c9 TI |
4234 | ret = iterate_dir_item(sctx->send_root, sctx->left_path, |
4235 | sctx->cmp_key, __process_new_xattr, sctx); | |
31db9f7c AB |
4236 | |
4237 | return ret; | |
4238 | } | |
4239 | ||
4240 | static int process_deleted_xattr(struct send_ctx *sctx) | |
4241 | { | |
4242 | int ret; | |
4243 | ||
924794c9 TI |
4244 | ret = iterate_dir_item(sctx->parent_root, sctx->right_path, |
4245 | sctx->cmp_key, __process_deleted_xattr, sctx); | |
31db9f7c AB |
4246 | |
4247 | return ret; | |
4248 | } | |
4249 | ||
4250 | struct find_xattr_ctx { | |
4251 | const char *name; | |
4252 | int name_len; | |
4253 | int found_idx; | |
4254 | char *found_data; | |
4255 | int found_data_len; | |
4256 | }; | |
4257 | ||
4258 | static int __find_xattr(int num, struct btrfs_key *di_key, | |
4259 | const char *name, int name_len, | |
4260 | const char *data, int data_len, | |
4261 | u8 type, void *vctx) | |
4262 | { | |
4263 | struct find_xattr_ctx *ctx = vctx; | |
4264 | ||
4265 | if (name_len == ctx->name_len && | |
4266 | strncmp(name, ctx->name, name_len) == 0) { | |
4267 | ctx->found_idx = num; | |
4268 | ctx->found_data_len = data_len; | |
a5959bc0 | 4269 | ctx->found_data = kmemdup(data, data_len, GFP_NOFS); |
31db9f7c AB |
4270 | if (!ctx->found_data) |
4271 | return -ENOMEM; | |
31db9f7c AB |
4272 | return 1; |
4273 | } | |
4274 | return 0; | |
4275 | } | |
4276 | ||
924794c9 | 4277 | static int find_xattr(struct btrfs_root *root, |
31db9f7c AB |
4278 | struct btrfs_path *path, |
4279 | struct btrfs_key *key, | |
4280 | const char *name, int name_len, | |
4281 | char **data, int *data_len) | |
4282 | { | |
4283 | int ret; | |
4284 | struct find_xattr_ctx ctx; | |
4285 | ||
4286 | ctx.name = name; | |
4287 | ctx.name_len = name_len; | |
4288 | ctx.found_idx = -1; | |
4289 | ctx.found_data = NULL; | |
4290 | ctx.found_data_len = 0; | |
4291 | ||
924794c9 | 4292 | ret = iterate_dir_item(root, path, key, __find_xattr, &ctx); |
31db9f7c AB |
4293 | if (ret < 0) |
4294 | return ret; | |
4295 | ||
4296 | if (ctx.found_idx == -1) | |
4297 | return -ENOENT; | |
4298 | if (data) { | |
4299 | *data = ctx.found_data; | |
4300 | *data_len = ctx.found_data_len; | |
4301 | } else { | |
4302 | kfree(ctx.found_data); | |
4303 | } | |
4304 | return ctx.found_idx; | |
4305 | } | |
4306 | ||
4307 | ||
4308 | static int __process_changed_new_xattr(int num, struct btrfs_key *di_key, | |
4309 | const char *name, int name_len, | |
4310 | const char *data, int data_len, | |
4311 | u8 type, void *ctx) | |
4312 | { | |
4313 | int ret; | |
4314 | struct send_ctx *sctx = ctx; | |
4315 | char *found_data = NULL; | |
4316 | int found_data_len = 0; | |
31db9f7c | 4317 | |
924794c9 TI |
4318 | ret = find_xattr(sctx->parent_root, sctx->right_path, |
4319 | sctx->cmp_key, name, name_len, &found_data, | |
4320 | &found_data_len); | |
31db9f7c AB |
4321 | if (ret == -ENOENT) { |
4322 | ret = __process_new_xattr(num, di_key, name, name_len, data, | |
4323 | data_len, type, ctx); | |
4324 | } else if (ret >= 0) { | |
4325 | if (data_len != found_data_len || | |
4326 | memcmp(data, found_data, data_len)) { | |
4327 | ret = __process_new_xattr(num, di_key, name, name_len, | |
4328 | data, data_len, type, ctx); | |
4329 | } else { | |
4330 | ret = 0; | |
4331 | } | |
4332 | } | |
4333 | ||
4334 | kfree(found_data); | |
31db9f7c AB |
4335 | return ret; |
4336 | } | |
4337 | ||
4338 | static int __process_changed_deleted_xattr(int num, struct btrfs_key *di_key, | |
4339 | const char *name, int name_len, | |
4340 | const char *data, int data_len, | |
4341 | u8 type, void *ctx) | |
4342 | { | |
4343 | int ret; | |
4344 | struct send_ctx *sctx = ctx; | |
4345 | ||
924794c9 TI |
4346 | ret = find_xattr(sctx->send_root, sctx->left_path, sctx->cmp_key, |
4347 | name, name_len, NULL, NULL); | |
31db9f7c AB |
4348 | if (ret == -ENOENT) |
4349 | ret = __process_deleted_xattr(num, di_key, name, name_len, data, | |
4350 | data_len, type, ctx); | |
4351 | else if (ret >= 0) | |
4352 | ret = 0; | |
4353 | ||
4354 | return ret; | |
4355 | } | |
4356 | ||
4357 | static int process_changed_xattr(struct send_ctx *sctx) | |
4358 | { | |
4359 | int ret = 0; | |
4360 | ||
924794c9 | 4361 | ret = iterate_dir_item(sctx->send_root, sctx->left_path, |
31db9f7c AB |
4362 | sctx->cmp_key, __process_changed_new_xattr, sctx); |
4363 | if (ret < 0) | |
4364 | goto out; | |
924794c9 | 4365 | ret = iterate_dir_item(sctx->parent_root, sctx->right_path, |
31db9f7c AB |
4366 | sctx->cmp_key, __process_changed_deleted_xattr, sctx); |
4367 | ||
4368 | out: | |
4369 | return ret; | |
4370 | } | |
4371 | ||
4372 | static int process_all_new_xattrs(struct send_ctx *sctx) | |
4373 | { | |
4374 | int ret; | |
4375 | struct btrfs_root *root; | |
4376 | struct btrfs_path *path; | |
4377 | struct btrfs_key key; | |
4378 | struct btrfs_key found_key; | |
4379 | struct extent_buffer *eb; | |
4380 | int slot; | |
4381 | ||
4382 | path = alloc_path_for_send(); | |
4383 | if (!path) | |
4384 | return -ENOMEM; | |
4385 | ||
4386 | root = sctx->send_root; | |
4387 | ||
4388 | key.objectid = sctx->cmp_key->objectid; | |
4389 | key.type = BTRFS_XATTR_ITEM_KEY; | |
4390 | key.offset = 0; | |
dff6d0ad FDBM |
4391 | ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); |
4392 | if (ret < 0) | |
4393 | goto out; | |
31db9f7c | 4394 | |
dff6d0ad | 4395 | while (1) { |
31db9f7c AB |
4396 | eb = path->nodes[0]; |
4397 | slot = path->slots[0]; | |
dff6d0ad FDBM |
4398 | if (slot >= btrfs_header_nritems(eb)) { |
4399 | ret = btrfs_next_leaf(root, path); | |
4400 | if (ret < 0) { | |
4401 | goto out; | |
4402 | } else if (ret > 0) { | |
4403 | ret = 0; | |
4404 | break; | |
4405 | } | |
4406 | continue; | |
4407 | } | |
31db9f7c | 4408 | |
dff6d0ad | 4409 | btrfs_item_key_to_cpu(eb, &found_key, slot); |
31db9f7c AB |
4410 | if (found_key.objectid != key.objectid || |
4411 | found_key.type != key.type) { | |
4412 | ret = 0; | |
4413 | goto out; | |
4414 | } | |
4415 | ||
924794c9 TI |
4416 | ret = iterate_dir_item(root, path, &found_key, |
4417 | __process_new_xattr, sctx); | |
31db9f7c AB |
4418 | if (ret < 0) |
4419 | goto out; | |
4420 | ||
dff6d0ad | 4421 | path->slots[0]++; |
31db9f7c AB |
4422 | } |
4423 | ||
4424 | out: | |
4425 | btrfs_free_path(path); | |
4426 | return ret; | |
4427 | } | |
4428 | ||
ed259095 JB |
4429 | static ssize_t fill_read_buf(struct send_ctx *sctx, u64 offset, u32 len) |
4430 | { | |
4431 | struct btrfs_root *root = sctx->send_root; | |
4432 | struct btrfs_fs_info *fs_info = root->fs_info; | |
4433 | struct inode *inode; | |
4434 | struct page *page; | |
4435 | char *addr; | |
4436 | struct btrfs_key key; | |
4437 | pgoff_t index = offset >> PAGE_CACHE_SHIFT; | |
4438 | pgoff_t last_index; | |
4439 | unsigned pg_offset = offset & ~PAGE_CACHE_MASK; | |
4440 | ssize_t ret = 0; | |
4441 | ||
4442 | key.objectid = sctx->cur_ino; | |
4443 | key.type = BTRFS_INODE_ITEM_KEY; | |
4444 | key.offset = 0; | |
4445 | ||
4446 | inode = btrfs_iget(fs_info->sb, &key, root, NULL); | |
4447 | if (IS_ERR(inode)) | |
4448 | return PTR_ERR(inode); | |
4449 | ||
4450 | if (offset + len > i_size_read(inode)) { | |
4451 | if (offset > i_size_read(inode)) | |
4452 | len = 0; | |
4453 | else | |
4454 | len = offset - i_size_read(inode); | |
4455 | } | |
4456 | if (len == 0) | |
4457 | goto out; | |
4458 | ||
4459 | last_index = (offset + len - 1) >> PAGE_CACHE_SHIFT; | |
2131bcd3 LB |
4460 | |
4461 | /* initial readahead */ | |
4462 | memset(&sctx->ra, 0, sizeof(struct file_ra_state)); | |
4463 | file_ra_state_init(&sctx->ra, inode->i_mapping); | |
4464 | btrfs_force_ra(inode->i_mapping, &sctx->ra, NULL, index, | |
4465 | last_index - index + 1); | |
4466 | ||
ed259095 JB |
4467 | while (index <= last_index) { |
4468 | unsigned cur_len = min_t(unsigned, len, | |
4469 | PAGE_CACHE_SIZE - pg_offset); | |
4470 | page = find_or_create_page(inode->i_mapping, index, GFP_NOFS); | |
4471 | if (!page) { | |
4472 | ret = -ENOMEM; | |
4473 | break; | |
4474 | } | |
4475 | ||
4476 | if (!PageUptodate(page)) { | |
4477 | btrfs_readpage(NULL, page); | |
4478 | lock_page(page); | |
4479 | if (!PageUptodate(page)) { | |
4480 | unlock_page(page); | |
4481 | page_cache_release(page); | |
4482 | ret = -EIO; | |
4483 | break; | |
4484 | } | |
4485 | } | |
4486 | ||
4487 | addr = kmap(page); | |
4488 | memcpy(sctx->read_buf + ret, addr + pg_offset, cur_len); | |
4489 | kunmap(page); | |
4490 | unlock_page(page); | |
4491 | page_cache_release(page); | |
4492 | index++; | |
4493 | pg_offset = 0; | |
4494 | len -= cur_len; | |
4495 | ret += cur_len; | |
4496 | } | |
4497 | out: | |
4498 | iput(inode); | |
4499 | return ret; | |
4500 | } | |
4501 | ||
31db9f7c AB |
4502 | /* |
4503 | * Read some bytes from the current inode/file and send a write command to | |
4504 | * user space. | |
4505 | */ | |
4506 | static int send_write(struct send_ctx *sctx, u64 offset, u32 len) | |
4507 | { | |
4508 | int ret = 0; | |
4509 | struct fs_path *p; | |
ed259095 | 4510 | ssize_t num_read = 0; |
31db9f7c | 4511 | |
924794c9 | 4512 | p = fs_path_alloc(); |
31db9f7c AB |
4513 | if (!p) |
4514 | return -ENOMEM; | |
4515 | ||
31db9f7c AB |
4516 | verbose_printk("btrfs: send_write offset=%llu, len=%d\n", offset, len); |
4517 | ||
ed259095 JB |
4518 | num_read = fill_read_buf(sctx, offset, len); |
4519 | if (num_read <= 0) { | |
4520 | if (num_read < 0) | |
4521 | ret = num_read; | |
31db9f7c | 4522 | goto out; |
ed259095 | 4523 | } |
31db9f7c AB |
4524 | |
4525 | ret = begin_cmd(sctx, BTRFS_SEND_C_WRITE); | |
4526 | if (ret < 0) | |
4527 | goto out; | |
4528 | ||
4529 | ret = get_cur_path(sctx, sctx->cur_ino, sctx->cur_inode_gen, p); | |
4530 | if (ret < 0) | |
4531 | goto out; | |
4532 | ||
4533 | TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p); | |
4534 | TLV_PUT_U64(sctx, BTRFS_SEND_A_FILE_OFFSET, offset); | |
e938c8ad | 4535 | TLV_PUT(sctx, BTRFS_SEND_A_DATA, sctx->read_buf, num_read); |
31db9f7c AB |
4536 | |
4537 | ret = send_cmd(sctx); | |
4538 | ||
4539 | tlv_put_failure: | |
4540 | out: | |
924794c9 | 4541 | fs_path_free(p); |
31db9f7c AB |
4542 | if (ret < 0) |
4543 | return ret; | |
e938c8ad | 4544 | return num_read; |
31db9f7c AB |
4545 | } |
4546 | ||
4547 | /* | |
4548 | * Send a clone command to user space. | |
4549 | */ | |
4550 | static int send_clone(struct send_ctx *sctx, | |
4551 | u64 offset, u32 len, | |
4552 | struct clone_root *clone_root) | |
4553 | { | |
4554 | int ret = 0; | |
31db9f7c AB |
4555 | struct fs_path *p; |
4556 | u64 gen; | |
4557 | ||
4558 | verbose_printk("btrfs: send_clone offset=%llu, len=%d, clone_root=%llu, " | |
4559 | "clone_inode=%llu, clone_offset=%llu\n", offset, len, | |
4560 | clone_root->root->objectid, clone_root->ino, | |
4561 | clone_root->offset); | |
4562 | ||
924794c9 | 4563 | p = fs_path_alloc(); |
31db9f7c AB |
4564 | if (!p) |
4565 | return -ENOMEM; | |
4566 | ||
4567 | ret = begin_cmd(sctx, BTRFS_SEND_C_CLONE); | |
4568 | if (ret < 0) | |
4569 | goto out; | |
4570 | ||
4571 | ret = get_cur_path(sctx, sctx->cur_ino, sctx->cur_inode_gen, p); | |
4572 | if (ret < 0) | |
4573 | goto out; | |
4574 | ||
4575 | TLV_PUT_U64(sctx, BTRFS_SEND_A_FILE_OFFSET, offset); | |
4576 | TLV_PUT_U64(sctx, BTRFS_SEND_A_CLONE_LEN, len); | |
4577 | TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p); | |
4578 | ||
e938c8ad | 4579 | if (clone_root->root == sctx->send_root) { |
31db9f7c | 4580 | ret = get_inode_info(sctx->send_root, clone_root->ino, NULL, |
85a7b33b | 4581 | &gen, NULL, NULL, NULL, NULL); |
31db9f7c AB |
4582 | if (ret < 0) |
4583 | goto out; | |
4584 | ret = get_cur_path(sctx, clone_root->ino, gen, p); | |
4585 | } else { | |
924794c9 | 4586 | ret = get_inode_path(clone_root->root, clone_root->ino, p); |
31db9f7c AB |
4587 | } |
4588 | if (ret < 0) | |
4589 | goto out; | |
4590 | ||
37b8d27d JB |
4591 | /* |
4592 | * If the parent we're using has a received_uuid set then use that as | |
4593 | * our clone source as that is what we will look for when doing a | |
4594 | * receive. | |
4595 | * | |
4596 | * This covers the case that we create a snapshot off of a received | |
4597 | * subvolume and then use that as the parent and try to receive on a | |
4598 | * different host. | |
4599 | */ | |
4600 | if (!btrfs_is_empty_uuid(clone_root->root->root_item.received_uuid)) | |
4601 | TLV_PUT_UUID(sctx, BTRFS_SEND_A_CLONE_UUID, | |
4602 | clone_root->root->root_item.received_uuid); | |
4603 | else | |
4604 | TLV_PUT_UUID(sctx, BTRFS_SEND_A_CLONE_UUID, | |
4605 | clone_root->root->root_item.uuid); | |
31db9f7c | 4606 | TLV_PUT_U64(sctx, BTRFS_SEND_A_CLONE_CTRANSID, |
5a0f4e2c | 4607 | le64_to_cpu(clone_root->root->root_item.ctransid)); |
31db9f7c AB |
4608 | TLV_PUT_PATH(sctx, BTRFS_SEND_A_CLONE_PATH, p); |
4609 | TLV_PUT_U64(sctx, BTRFS_SEND_A_CLONE_OFFSET, | |
4610 | clone_root->offset); | |
4611 | ||
4612 | ret = send_cmd(sctx); | |
4613 | ||
4614 | tlv_put_failure: | |
4615 | out: | |
924794c9 | 4616 | fs_path_free(p); |
31db9f7c AB |
4617 | return ret; |
4618 | } | |
4619 | ||
cb95e7bf MF |
4620 | /* |
4621 | * Send an update extent command to user space. | |
4622 | */ | |
4623 | static int send_update_extent(struct send_ctx *sctx, | |
4624 | u64 offset, u32 len) | |
4625 | { | |
4626 | int ret = 0; | |
4627 | struct fs_path *p; | |
4628 | ||
924794c9 | 4629 | p = fs_path_alloc(); |
cb95e7bf MF |
4630 | if (!p) |
4631 | return -ENOMEM; | |
4632 | ||
4633 | ret = begin_cmd(sctx, BTRFS_SEND_C_UPDATE_EXTENT); | |
4634 | if (ret < 0) | |
4635 | goto out; | |
4636 | ||
4637 | ret = get_cur_path(sctx, sctx->cur_ino, sctx->cur_inode_gen, p); | |
4638 | if (ret < 0) | |
4639 | goto out; | |
4640 | ||
4641 | TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p); | |
4642 | TLV_PUT_U64(sctx, BTRFS_SEND_A_FILE_OFFSET, offset); | |
4643 | TLV_PUT_U64(sctx, BTRFS_SEND_A_SIZE, len); | |
4644 | ||
4645 | ret = send_cmd(sctx); | |
4646 | ||
4647 | tlv_put_failure: | |
4648 | out: | |
924794c9 | 4649 | fs_path_free(p); |
cb95e7bf MF |
4650 | return ret; |
4651 | } | |
4652 | ||
16e7549f JB |
4653 | static int send_hole(struct send_ctx *sctx, u64 end) |
4654 | { | |
4655 | struct fs_path *p = NULL; | |
4656 | u64 offset = sctx->cur_inode_last_extent; | |
4657 | u64 len; | |
4658 | int ret = 0; | |
4659 | ||
4660 | p = fs_path_alloc(); | |
4661 | if (!p) | |
4662 | return -ENOMEM; | |
c715e155 FM |
4663 | ret = get_cur_path(sctx, sctx->cur_ino, sctx->cur_inode_gen, p); |
4664 | if (ret < 0) | |
4665 | goto tlv_put_failure; | |
16e7549f JB |
4666 | memset(sctx->read_buf, 0, BTRFS_SEND_READ_SIZE); |
4667 | while (offset < end) { | |
4668 | len = min_t(u64, end - offset, BTRFS_SEND_READ_SIZE); | |
4669 | ||
4670 | ret = begin_cmd(sctx, BTRFS_SEND_C_WRITE); | |
16e7549f JB |
4671 | if (ret < 0) |
4672 | break; | |
4673 | TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p); | |
4674 | TLV_PUT_U64(sctx, BTRFS_SEND_A_FILE_OFFSET, offset); | |
4675 | TLV_PUT(sctx, BTRFS_SEND_A_DATA, sctx->read_buf, len); | |
4676 | ret = send_cmd(sctx); | |
4677 | if (ret < 0) | |
4678 | break; | |
4679 | offset += len; | |
4680 | } | |
4681 | tlv_put_failure: | |
4682 | fs_path_free(p); | |
4683 | return ret; | |
4684 | } | |
4685 | ||
d906d49f FM |
4686 | static int send_extent_data(struct send_ctx *sctx, |
4687 | const u64 offset, | |
4688 | const u64 len) | |
4689 | { | |
4690 | u64 sent = 0; | |
4691 | ||
4692 | if (sctx->flags & BTRFS_SEND_FLAG_NO_FILE_DATA) | |
4693 | return send_update_extent(sctx, offset, len); | |
4694 | ||
4695 | while (sent < len) { | |
4696 | u64 size = len - sent; | |
4697 | int ret; | |
4698 | ||
4699 | if (size > BTRFS_SEND_READ_SIZE) | |
4700 | size = BTRFS_SEND_READ_SIZE; | |
4701 | ret = send_write(sctx, offset + sent, size); | |
4702 | if (ret < 0) | |
4703 | return ret; | |
4704 | if (!ret) | |
4705 | break; | |
4706 | sent += ret; | |
4707 | } | |
4708 | return 0; | |
4709 | } | |
4710 | ||
4711 | static int clone_range(struct send_ctx *sctx, | |
4712 | struct clone_root *clone_root, | |
4713 | const u64 disk_byte, | |
4714 | u64 data_offset, | |
4715 | u64 offset, | |
4716 | u64 len) | |
4717 | { | |
4718 | struct btrfs_path *path; | |
4719 | struct btrfs_key key; | |
4720 | int ret; | |
4721 | ||
4722 | path = alloc_path_for_send(); | |
4723 | if (!path) | |
4724 | return -ENOMEM; | |
4725 | ||
4726 | /* | |
4727 | * We can't send a clone operation for the entire range if we find | |
4728 | * extent items in the respective range in the source file that | |
4729 | * refer to different extents or if we find holes. | |
4730 | * So check for that and do a mix of clone and regular write/copy | |
4731 | * operations if needed. | |
4732 | * | |
4733 | * Example: | |
4734 | * | |
4735 | * mkfs.btrfs -f /dev/sda | |
4736 | * mount /dev/sda /mnt | |
4737 | * xfs_io -f -c "pwrite -S 0xaa 0K 100K" /mnt/foo | |
4738 | * cp --reflink=always /mnt/foo /mnt/bar | |
4739 | * xfs_io -c "pwrite -S 0xbb 50K 50K" /mnt/foo | |
4740 | * btrfs subvolume snapshot -r /mnt /mnt/snap | |
4741 | * | |
4742 | * If when we send the snapshot and we are processing file bar (which | |
4743 | * has a higher inode number than foo) we blindly send a clone operation | |
4744 | * for the [0, 100K[ range from foo to bar, the receiver ends up getting | |
4745 | * a file bar that matches the content of file foo - iow, doesn't match | |
4746 | * the content from bar in the original filesystem. | |
4747 | */ | |
4748 | key.objectid = clone_root->ino; | |
4749 | key.type = BTRFS_EXTENT_DATA_KEY; | |
4750 | key.offset = clone_root->offset; | |
4751 | ret = btrfs_search_slot(NULL, clone_root->root, &key, path, 0, 0); | |
4752 | if (ret < 0) | |
4753 | goto out; | |
4754 | if (ret > 0 && path->slots[0] > 0) { | |
4755 | btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0] - 1); | |
4756 | if (key.objectid == clone_root->ino && | |
4757 | key.type == BTRFS_EXTENT_DATA_KEY) | |
4758 | path->slots[0]--; | |
4759 | } | |
4760 | ||
4761 | while (true) { | |
4762 | struct extent_buffer *leaf = path->nodes[0]; | |
4763 | int slot = path->slots[0]; | |
4764 | struct btrfs_file_extent_item *ei; | |
4765 | u8 type; | |
4766 | u64 ext_len; | |
4767 | u64 clone_len; | |
4768 | ||
4769 | if (slot >= btrfs_header_nritems(leaf)) { | |
4770 | ret = btrfs_next_leaf(clone_root->root, path); | |
4771 | if (ret < 0) | |
4772 | goto out; | |
4773 | else if (ret > 0) | |
4774 | break; | |
4775 | continue; | |
4776 | } | |
4777 | ||
4778 | btrfs_item_key_to_cpu(leaf, &key, slot); | |
4779 | ||
4780 | /* | |
4781 | * We might have an implicit trailing hole (NO_HOLES feature | |
4782 | * enabled). We deal with it after leaving this loop. | |
4783 | */ | |
4784 | if (key.objectid != clone_root->ino || | |
4785 | key.type != BTRFS_EXTENT_DATA_KEY) | |
4786 | break; | |
4787 | ||
4788 | ei = btrfs_item_ptr(leaf, slot, struct btrfs_file_extent_item); | |
4789 | type = btrfs_file_extent_type(leaf, ei); | |
4790 | if (type == BTRFS_FILE_EXTENT_INLINE) { | |
4791 | ext_len = btrfs_file_extent_inline_len(leaf, slot, ei); | |
4792 | ext_len = PAGE_CACHE_ALIGN(ext_len); | |
4793 | } else { | |
4794 | ext_len = btrfs_file_extent_num_bytes(leaf, ei); | |
4795 | } | |
4796 | ||
4797 | if (key.offset + ext_len <= clone_root->offset) | |
4798 | goto next; | |
4799 | ||
4800 | if (key.offset > clone_root->offset) { | |
4801 | /* Implicit hole, NO_HOLES feature enabled. */ | |
4802 | u64 hole_len = key.offset - clone_root->offset; | |
4803 | ||
4804 | if (hole_len > len) | |
4805 | hole_len = len; | |
4806 | ret = send_extent_data(sctx, offset, hole_len); | |
4807 | if (ret < 0) | |
4808 | goto out; | |
4809 | ||
4810 | len -= hole_len; | |
4811 | if (len == 0) | |
4812 | break; | |
4813 | offset += hole_len; | |
4814 | clone_root->offset += hole_len; | |
4815 | data_offset += hole_len; | |
4816 | } | |
4817 | ||
4818 | if (key.offset >= clone_root->offset + len) | |
4819 | break; | |
4820 | ||
4821 | clone_len = min_t(u64, ext_len, len); | |
4822 | ||
4823 | if (btrfs_file_extent_disk_bytenr(leaf, ei) == disk_byte && | |
4824 | btrfs_file_extent_offset(leaf, ei) == data_offset) | |
4825 | ret = send_clone(sctx, offset, clone_len, clone_root); | |
4826 | else | |
4827 | ret = send_extent_data(sctx, offset, clone_len); | |
4828 | ||
4829 | if (ret < 0) | |
4830 | goto out; | |
4831 | ||
4832 | len -= clone_len; | |
4833 | if (len == 0) | |
4834 | break; | |
4835 | offset += clone_len; | |
4836 | clone_root->offset += clone_len; | |
4837 | data_offset += clone_len; | |
4838 | next: | |
4839 | path->slots[0]++; | |
4840 | } | |
4841 | ||
4842 | if (len > 0) | |
4843 | ret = send_extent_data(sctx, offset, len); | |
4844 | else | |
4845 | ret = 0; | |
4846 | out: | |
4847 | btrfs_free_path(path); | |
4848 | return ret; | |
4849 | } | |
4850 | ||
31db9f7c AB |
4851 | static int send_write_or_clone(struct send_ctx *sctx, |
4852 | struct btrfs_path *path, | |
4853 | struct btrfs_key *key, | |
4854 | struct clone_root *clone_root) | |
4855 | { | |
4856 | int ret = 0; | |
4857 | struct btrfs_file_extent_item *ei; | |
4858 | u64 offset = key->offset; | |
31db9f7c | 4859 | u64 len; |
31db9f7c | 4860 | u8 type; |
28e5dd8f | 4861 | u64 bs = sctx->send_root->fs_info->sb->s_blocksize; |
31db9f7c AB |
4862 | |
4863 | ei = btrfs_item_ptr(path->nodes[0], path->slots[0], | |
4864 | struct btrfs_file_extent_item); | |
4865 | type = btrfs_file_extent_type(path->nodes[0], ei); | |
74dd17fb | 4866 | if (type == BTRFS_FILE_EXTENT_INLINE) { |
514ac8ad CM |
4867 | len = btrfs_file_extent_inline_len(path->nodes[0], |
4868 | path->slots[0], ei); | |
74dd17fb CM |
4869 | /* |
4870 | * it is possible the inline item won't cover the whole page, | |
4871 | * but there may be items after this page. Make | |
4872 | * sure to send the whole thing | |
4873 | */ | |
4874 | len = PAGE_CACHE_ALIGN(len); | |
4875 | } else { | |
31db9f7c | 4876 | len = btrfs_file_extent_num_bytes(path->nodes[0], ei); |
74dd17fb | 4877 | } |
31db9f7c AB |
4878 | |
4879 | if (offset + len > sctx->cur_inode_size) | |
4880 | len = sctx->cur_inode_size - offset; | |
4881 | if (len == 0) { | |
4882 | ret = 0; | |
4883 | goto out; | |
4884 | } | |
4885 | ||
28e5dd8f | 4886 | if (clone_root && IS_ALIGNED(offset + len, bs)) { |
d906d49f FM |
4887 | u64 disk_byte; |
4888 | u64 data_offset; | |
4889 | ||
4890 | disk_byte = btrfs_file_extent_disk_bytenr(path->nodes[0], ei); | |
4891 | data_offset = btrfs_file_extent_offset(path->nodes[0], ei); | |
4892 | ret = clone_range(sctx, clone_root, disk_byte, data_offset, | |
4893 | offset, len); | |
cb95e7bf | 4894 | } else { |
d906d49f | 4895 | ret = send_extent_data(sctx, offset, len); |
31db9f7c | 4896 | } |
31db9f7c AB |
4897 | out: |
4898 | return ret; | |
4899 | } | |
4900 | ||
4901 | static int is_extent_unchanged(struct send_ctx *sctx, | |
4902 | struct btrfs_path *left_path, | |
4903 | struct btrfs_key *ekey) | |
4904 | { | |
4905 | int ret = 0; | |
4906 | struct btrfs_key key; | |
4907 | struct btrfs_path *path = NULL; | |
4908 | struct extent_buffer *eb; | |
4909 | int slot; | |
4910 | struct btrfs_key found_key; | |
4911 | struct btrfs_file_extent_item *ei; | |
4912 | u64 left_disknr; | |
4913 | u64 right_disknr; | |
4914 | u64 left_offset; | |
4915 | u64 right_offset; | |
4916 | u64 left_offset_fixed; | |
4917 | u64 left_len; | |
4918 | u64 right_len; | |
74dd17fb CM |
4919 | u64 left_gen; |
4920 | u64 right_gen; | |
31db9f7c AB |
4921 | u8 left_type; |
4922 | u8 right_type; | |
4923 | ||
4924 | path = alloc_path_for_send(); | |
4925 | if (!path) | |
4926 | return -ENOMEM; | |
4927 | ||
4928 | eb = left_path->nodes[0]; | |
4929 | slot = left_path->slots[0]; | |
31db9f7c AB |
4930 | ei = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item); |
4931 | left_type = btrfs_file_extent_type(eb, ei); | |
31db9f7c AB |
4932 | |
4933 | if (left_type != BTRFS_FILE_EXTENT_REG) { | |
4934 | ret = 0; | |
4935 | goto out; | |
4936 | } | |
74dd17fb CM |
4937 | left_disknr = btrfs_file_extent_disk_bytenr(eb, ei); |
4938 | left_len = btrfs_file_extent_num_bytes(eb, ei); | |
4939 | left_offset = btrfs_file_extent_offset(eb, ei); | |
4940 | left_gen = btrfs_file_extent_generation(eb, ei); | |
31db9f7c AB |
4941 | |
4942 | /* | |
4943 | * Following comments will refer to these graphics. L is the left | |
4944 | * extents which we are checking at the moment. 1-8 are the right | |
4945 | * extents that we iterate. | |
4946 | * | |
4947 | * |-----L-----| | |
4948 | * |-1-|-2a-|-3-|-4-|-5-|-6-| | |
4949 | * | |
4950 | * |-----L-----| | |
4951 | * |--1--|-2b-|...(same as above) | |
4952 | * | |
4953 | * Alternative situation. Happens on files where extents got split. | |
4954 | * |-----L-----| | |
4955 | * |-----------7-----------|-6-| | |
4956 | * | |
4957 | * Alternative situation. Happens on files which got larger. | |
4958 | * |-----L-----| | |
4959 | * |-8-| | |
4960 | * Nothing follows after 8. | |
4961 | */ | |
4962 | ||
4963 | key.objectid = ekey->objectid; | |
4964 | key.type = BTRFS_EXTENT_DATA_KEY; | |
4965 | key.offset = ekey->offset; | |
4966 | ret = btrfs_search_slot_for_read(sctx->parent_root, &key, path, 0, 0); | |
4967 | if (ret < 0) | |
4968 | goto out; | |
4969 | if (ret) { | |
4970 | ret = 0; | |
4971 | goto out; | |
4972 | } | |
4973 | ||
4974 | /* | |
4975 | * Handle special case where the right side has no extents at all. | |
4976 | */ | |
4977 | eb = path->nodes[0]; | |
4978 | slot = path->slots[0]; | |
4979 | btrfs_item_key_to_cpu(eb, &found_key, slot); | |
4980 | if (found_key.objectid != key.objectid || | |
4981 | found_key.type != key.type) { | |
57cfd462 JB |
4982 | /* If we're a hole then just pretend nothing changed */ |
4983 | ret = (left_disknr) ? 0 : 1; | |
31db9f7c AB |
4984 | goto out; |
4985 | } | |
4986 | ||
4987 | /* | |
4988 | * We're now on 2a, 2b or 7. | |
4989 | */ | |
4990 | key = found_key; | |
4991 | while (key.offset < ekey->offset + left_len) { | |
4992 | ei = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item); | |
4993 | right_type = btrfs_file_extent_type(eb, ei); | |
31db9f7c AB |
4994 | if (right_type != BTRFS_FILE_EXTENT_REG) { |
4995 | ret = 0; | |
4996 | goto out; | |
4997 | } | |
4998 | ||
007d31f7 JB |
4999 | right_disknr = btrfs_file_extent_disk_bytenr(eb, ei); |
5000 | right_len = btrfs_file_extent_num_bytes(eb, ei); | |
5001 | right_offset = btrfs_file_extent_offset(eb, ei); | |
5002 | right_gen = btrfs_file_extent_generation(eb, ei); | |
5003 | ||
31db9f7c AB |
5004 | /* |
5005 | * Are we at extent 8? If yes, we know the extent is changed. | |
5006 | * This may only happen on the first iteration. | |
5007 | */ | |
d8347fa4 | 5008 | if (found_key.offset + right_len <= ekey->offset) { |
57cfd462 JB |
5009 | /* If we're a hole just pretend nothing changed */ |
5010 | ret = (left_disknr) ? 0 : 1; | |
31db9f7c AB |
5011 | goto out; |
5012 | } | |
5013 | ||
5014 | left_offset_fixed = left_offset; | |
5015 | if (key.offset < ekey->offset) { | |
5016 | /* Fix the right offset for 2a and 7. */ | |
5017 | right_offset += ekey->offset - key.offset; | |
5018 | } else { | |
5019 | /* Fix the left offset for all behind 2a and 2b */ | |
5020 | left_offset_fixed += key.offset - ekey->offset; | |
5021 | } | |
5022 | ||
5023 | /* | |
5024 | * Check if we have the same extent. | |
5025 | */ | |
3954096d | 5026 | if (left_disknr != right_disknr || |
74dd17fb CM |
5027 | left_offset_fixed != right_offset || |
5028 | left_gen != right_gen) { | |
31db9f7c AB |
5029 | ret = 0; |
5030 | goto out; | |
5031 | } | |
5032 | ||
5033 | /* | |
5034 | * Go to the next extent. | |
5035 | */ | |
5036 | ret = btrfs_next_item(sctx->parent_root, path); | |
5037 | if (ret < 0) | |
5038 | goto out; | |
5039 | if (!ret) { | |
5040 | eb = path->nodes[0]; | |
5041 | slot = path->slots[0]; | |
5042 | btrfs_item_key_to_cpu(eb, &found_key, slot); | |
5043 | } | |
5044 | if (ret || found_key.objectid != key.objectid || | |
5045 | found_key.type != key.type) { | |
5046 | key.offset += right_len; | |
5047 | break; | |
adaa4b8e JS |
5048 | } |
5049 | if (found_key.offset != key.offset + right_len) { | |
5050 | ret = 0; | |
5051 | goto out; | |
31db9f7c AB |
5052 | } |
5053 | key = found_key; | |
5054 | } | |
5055 | ||
5056 | /* | |
5057 | * We're now behind the left extent (treat as unchanged) or at the end | |
5058 | * of the right side (treat as changed). | |
5059 | */ | |
5060 | if (key.offset >= ekey->offset + left_len) | |
5061 | ret = 1; | |
5062 | else | |
5063 | ret = 0; | |
5064 | ||
5065 | ||
5066 | out: | |
5067 | btrfs_free_path(path); | |
5068 | return ret; | |
5069 | } | |
5070 | ||
16e7549f JB |
5071 | static int get_last_extent(struct send_ctx *sctx, u64 offset) |
5072 | { | |
5073 | struct btrfs_path *path; | |
5074 | struct btrfs_root *root = sctx->send_root; | |
5075 | struct btrfs_file_extent_item *fi; | |
5076 | struct btrfs_key key; | |
5077 | u64 extent_end; | |
5078 | u8 type; | |
5079 | int ret; | |
5080 | ||
5081 | path = alloc_path_for_send(); | |
5082 | if (!path) | |
5083 | return -ENOMEM; | |
5084 | ||
5085 | sctx->cur_inode_last_extent = 0; | |
5086 | ||
5087 | key.objectid = sctx->cur_ino; | |
5088 | key.type = BTRFS_EXTENT_DATA_KEY; | |
5089 | key.offset = offset; | |
5090 | ret = btrfs_search_slot_for_read(root, &key, path, 0, 1); | |
5091 | if (ret < 0) | |
5092 | goto out; | |
5093 | ret = 0; | |
5094 | btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]); | |
5095 | if (key.objectid != sctx->cur_ino || key.type != BTRFS_EXTENT_DATA_KEY) | |
5096 | goto out; | |
5097 | ||
5098 | fi = btrfs_item_ptr(path->nodes[0], path->slots[0], | |
5099 | struct btrfs_file_extent_item); | |
5100 | type = btrfs_file_extent_type(path->nodes[0], fi); | |
5101 | if (type == BTRFS_FILE_EXTENT_INLINE) { | |
514ac8ad CM |
5102 | u64 size = btrfs_file_extent_inline_len(path->nodes[0], |
5103 | path->slots[0], fi); | |
16e7549f JB |
5104 | extent_end = ALIGN(key.offset + size, |
5105 | sctx->send_root->sectorsize); | |
5106 | } else { | |
5107 | extent_end = key.offset + | |
5108 | btrfs_file_extent_num_bytes(path->nodes[0], fi); | |
5109 | } | |
5110 | sctx->cur_inode_last_extent = extent_end; | |
5111 | out: | |
5112 | btrfs_free_path(path); | |
5113 | return ret; | |
5114 | } | |
5115 | ||
5116 | static int maybe_send_hole(struct send_ctx *sctx, struct btrfs_path *path, | |
5117 | struct btrfs_key *key) | |
5118 | { | |
5119 | struct btrfs_file_extent_item *fi; | |
5120 | u64 extent_end; | |
5121 | u8 type; | |
5122 | int ret = 0; | |
5123 | ||
5124 | if (sctx->cur_ino != key->objectid || !need_send_hole(sctx)) | |
5125 | return 0; | |
5126 | ||
5127 | if (sctx->cur_inode_last_extent == (u64)-1) { | |
5128 | ret = get_last_extent(sctx, key->offset - 1); | |
5129 | if (ret) | |
5130 | return ret; | |
5131 | } | |
5132 | ||
5133 | fi = btrfs_item_ptr(path->nodes[0], path->slots[0], | |
5134 | struct btrfs_file_extent_item); | |
5135 | type = btrfs_file_extent_type(path->nodes[0], fi); | |
5136 | if (type == BTRFS_FILE_EXTENT_INLINE) { | |
514ac8ad CM |
5137 | u64 size = btrfs_file_extent_inline_len(path->nodes[0], |
5138 | path->slots[0], fi); | |
16e7549f JB |
5139 | extent_end = ALIGN(key->offset + size, |
5140 | sctx->send_root->sectorsize); | |
5141 | } else { | |
5142 | extent_end = key->offset + | |
5143 | btrfs_file_extent_num_bytes(path->nodes[0], fi); | |
5144 | } | |
bf54f412 FDBM |
5145 | |
5146 | if (path->slots[0] == 0 && | |
5147 | sctx->cur_inode_last_extent < key->offset) { | |
5148 | /* | |
5149 | * We might have skipped entire leafs that contained only | |
5150 | * file extent items for our current inode. These leafs have | |
5151 | * a generation number smaller (older) than the one in the | |
5152 | * current leaf and the leaf our last extent came from, and | |
5153 | * are located between these 2 leafs. | |
5154 | */ | |
5155 | ret = get_last_extent(sctx, key->offset - 1); | |
5156 | if (ret) | |
5157 | return ret; | |
5158 | } | |
5159 | ||
16e7549f JB |
5160 | if (sctx->cur_inode_last_extent < key->offset) |
5161 | ret = send_hole(sctx, key->offset); | |
5162 | sctx->cur_inode_last_extent = extent_end; | |
5163 | return ret; | |
5164 | } | |
5165 | ||
31db9f7c AB |
5166 | static int process_extent(struct send_ctx *sctx, |
5167 | struct btrfs_path *path, | |
5168 | struct btrfs_key *key) | |
5169 | { | |
31db9f7c | 5170 | struct clone_root *found_clone = NULL; |
57cfd462 | 5171 | int ret = 0; |
31db9f7c AB |
5172 | |
5173 | if (S_ISLNK(sctx->cur_inode_mode)) | |
5174 | return 0; | |
5175 | ||
5176 | if (sctx->parent_root && !sctx->cur_inode_new) { | |
5177 | ret = is_extent_unchanged(sctx, path, key); | |
5178 | if (ret < 0) | |
5179 | goto out; | |
5180 | if (ret) { | |
5181 | ret = 0; | |
16e7549f | 5182 | goto out_hole; |
31db9f7c | 5183 | } |
57cfd462 JB |
5184 | } else { |
5185 | struct btrfs_file_extent_item *ei; | |
5186 | u8 type; | |
5187 | ||
5188 | ei = btrfs_item_ptr(path->nodes[0], path->slots[0], | |
5189 | struct btrfs_file_extent_item); | |
5190 | type = btrfs_file_extent_type(path->nodes[0], ei); | |
5191 | if (type == BTRFS_FILE_EXTENT_PREALLOC || | |
5192 | type == BTRFS_FILE_EXTENT_REG) { | |
5193 | /* | |
5194 | * The send spec does not have a prealloc command yet, | |
5195 | * so just leave a hole for prealloc'ed extents until | |
5196 | * we have enough commands queued up to justify rev'ing | |
5197 | * the send spec. | |
5198 | */ | |
5199 | if (type == BTRFS_FILE_EXTENT_PREALLOC) { | |
5200 | ret = 0; | |
5201 | goto out; | |
5202 | } | |
5203 | ||
5204 | /* Have a hole, just skip it. */ | |
5205 | if (btrfs_file_extent_disk_bytenr(path->nodes[0], ei) == 0) { | |
5206 | ret = 0; | |
5207 | goto out; | |
5208 | } | |
5209 | } | |
31db9f7c AB |
5210 | } |
5211 | ||
5212 | ret = find_extent_clone(sctx, path, key->objectid, key->offset, | |
5213 | sctx->cur_inode_size, &found_clone); | |
5214 | if (ret != -ENOENT && ret < 0) | |
5215 | goto out; | |
5216 | ||
5217 | ret = send_write_or_clone(sctx, path, key, found_clone); | |
16e7549f JB |
5218 | if (ret) |
5219 | goto out; | |
5220 | out_hole: | |
5221 | ret = maybe_send_hole(sctx, path, key); | |
31db9f7c AB |
5222 | out: |
5223 | return ret; | |
5224 | } | |
5225 | ||
5226 | static int process_all_extents(struct send_ctx *sctx) | |
5227 | { | |
5228 | int ret; | |
5229 | struct btrfs_root *root; | |
5230 | struct btrfs_path *path; | |
5231 | struct btrfs_key key; | |
5232 | struct btrfs_key found_key; | |
5233 | struct extent_buffer *eb; | |
5234 | int slot; | |
5235 | ||
5236 | root = sctx->send_root; | |
5237 | path = alloc_path_for_send(); | |
5238 | if (!path) | |
5239 | return -ENOMEM; | |
5240 | ||
5241 | key.objectid = sctx->cmp_key->objectid; | |
5242 | key.type = BTRFS_EXTENT_DATA_KEY; | |
5243 | key.offset = 0; | |
7fdd29d0 FDBM |
5244 | ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); |
5245 | if (ret < 0) | |
5246 | goto out; | |
31db9f7c | 5247 | |
7fdd29d0 | 5248 | while (1) { |
31db9f7c AB |
5249 | eb = path->nodes[0]; |
5250 | slot = path->slots[0]; | |
7fdd29d0 FDBM |
5251 | |
5252 | if (slot >= btrfs_header_nritems(eb)) { | |
5253 | ret = btrfs_next_leaf(root, path); | |
5254 | if (ret < 0) { | |
5255 | goto out; | |
5256 | } else if (ret > 0) { | |
5257 | ret = 0; | |
5258 | break; | |
5259 | } | |
5260 | continue; | |
5261 | } | |
5262 | ||
31db9f7c AB |
5263 | btrfs_item_key_to_cpu(eb, &found_key, slot); |
5264 | ||
5265 | if (found_key.objectid != key.objectid || | |
5266 | found_key.type != key.type) { | |
5267 | ret = 0; | |
5268 | goto out; | |
5269 | } | |
5270 | ||
5271 | ret = process_extent(sctx, path, &found_key); | |
5272 | if (ret < 0) | |
5273 | goto out; | |
5274 | ||
7fdd29d0 | 5275 | path->slots[0]++; |
31db9f7c AB |
5276 | } |
5277 | ||
5278 | out: | |
5279 | btrfs_free_path(path); | |
5280 | return ret; | |
5281 | } | |
5282 | ||
9f03740a FDBM |
5283 | static int process_recorded_refs_if_needed(struct send_ctx *sctx, int at_end, |
5284 | int *pending_move, | |
5285 | int *refs_processed) | |
31db9f7c AB |
5286 | { |
5287 | int ret = 0; | |
5288 | ||
5289 | if (sctx->cur_ino == 0) | |
5290 | goto out; | |
5291 | if (!at_end && sctx->cur_ino == sctx->cmp_key->objectid && | |
96b5bd77 | 5292 | sctx->cmp_key->type <= BTRFS_INODE_EXTREF_KEY) |
31db9f7c AB |
5293 | goto out; |
5294 | if (list_empty(&sctx->new_refs) && list_empty(&sctx->deleted_refs)) | |
5295 | goto out; | |
5296 | ||
9f03740a | 5297 | ret = process_recorded_refs(sctx, pending_move); |
e479d9bb AB |
5298 | if (ret < 0) |
5299 | goto out; | |
5300 | ||
9f03740a | 5301 | *refs_processed = 1; |
31db9f7c AB |
5302 | out: |
5303 | return ret; | |
5304 | } | |
5305 | ||
5306 | static int finish_inode_if_needed(struct send_ctx *sctx, int at_end) | |
5307 | { | |
5308 | int ret = 0; | |
5309 | u64 left_mode; | |
5310 | u64 left_uid; | |
5311 | u64 left_gid; | |
5312 | u64 right_mode; | |
5313 | u64 right_uid; | |
5314 | u64 right_gid; | |
5315 | int need_chmod = 0; | |
5316 | int need_chown = 0; | |
9f03740a FDBM |
5317 | int pending_move = 0; |
5318 | int refs_processed = 0; | |
31db9f7c | 5319 | |
9f03740a FDBM |
5320 | ret = process_recorded_refs_if_needed(sctx, at_end, &pending_move, |
5321 | &refs_processed); | |
31db9f7c AB |
5322 | if (ret < 0) |
5323 | goto out; | |
5324 | ||
9f03740a FDBM |
5325 | /* |
5326 | * We have processed the refs and thus need to advance send_progress. | |
5327 | * Now, calls to get_cur_xxx will take the updated refs of the current | |
5328 | * inode into account. | |
5329 | * | |
5330 | * On the other hand, if our current inode is a directory and couldn't | |
5331 | * be moved/renamed because its parent was renamed/moved too and it has | |
5332 | * a higher inode number, we can only move/rename our current inode | |
5333 | * after we moved/renamed its parent. Therefore in this case operate on | |
5334 | * the old path (pre move/rename) of our current inode, and the | |
5335 | * move/rename will be performed later. | |
5336 | */ | |
5337 | if (refs_processed && !pending_move) | |
5338 | sctx->send_progress = sctx->cur_ino + 1; | |
5339 | ||
31db9f7c AB |
5340 | if (sctx->cur_ino == 0 || sctx->cur_inode_deleted) |
5341 | goto out; | |
5342 | if (!at_end && sctx->cmp_key->objectid == sctx->cur_ino) | |
5343 | goto out; | |
5344 | ||
5345 | ret = get_inode_info(sctx->send_root, sctx->cur_ino, NULL, NULL, | |
85a7b33b | 5346 | &left_mode, &left_uid, &left_gid, NULL); |
31db9f7c AB |
5347 | if (ret < 0) |
5348 | goto out; | |
5349 | ||
e2d044fe AL |
5350 | if (!sctx->parent_root || sctx->cur_inode_new) { |
5351 | need_chown = 1; | |
5352 | if (!S_ISLNK(sctx->cur_inode_mode)) | |
31db9f7c | 5353 | need_chmod = 1; |
e2d044fe AL |
5354 | } else { |
5355 | ret = get_inode_info(sctx->parent_root, sctx->cur_ino, | |
5356 | NULL, NULL, &right_mode, &right_uid, | |
5357 | &right_gid, NULL); | |
5358 | if (ret < 0) | |
5359 | goto out; | |
31db9f7c | 5360 | |
e2d044fe AL |
5361 | if (left_uid != right_uid || left_gid != right_gid) |
5362 | need_chown = 1; | |
5363 | if (!S_ISLNK(sctx->cur_inode_mode) && left_mode != right_mode) | |
5364 | need_chmod = 1; | |
31db9f7c AB |
5365 | } |
5366 | ||
5367 | if (S_ISREG(sctx->cur_inode_mode)) { | |
16e7549f | 5368 | if (need_send_hole(sctx)) { |
766b5e5a FM |
5369 | if (sctx->cur_inode_last_extent == (u64)-1 || |
5370 | sctx->cur_inode_last_extent < | |
5371 | sctx->cur_inode_size) { | |
16e7549f JB |
5372 | ret = get_last_extent(sctx, (u64)-1); |
5373 | if (ret) | |
5374 | goto out; | |
5375 | } | |
5376 | if (sctx->cur_inode_last_extent < | |
5377 | sctx->cur_inode_size) { | |
5378 | ret = send_hole(sctx, sctx->cur_inode_size); | |
5379 | if (ret) | |
5380 | goto out; | |
5381 | } | |
5382 | } | |
31db9f7c AB |
5383 | ret = send_truncate(sctx, sctx->cur_ino, sctx->cur_inode_gen, |
5384 | sctx->cur_inode_size); | |
5385 | if (ret < 0) | |
5386 | goto out; | |
5387 | } | |
5388 | ||
5389 | if (need_chown) { | |
5390 | ret = send_chown(sctx, sctx->cur_ino, sctx->cur_inode_gen, | |
5391 | left_uid, left_gid); | |
5392 | if (ret < 0) | |
5393 | goto out; | |
5394 | } | |
5395 | if (need_chmod) { | |
5396 | ret = send_chmod(sctx, sctx->cur_ino, sctx->cur_inode_gen, | |
5397 | left_mode); | |
5398 | if (ret < 0) | |
5399 | goto out; | |
5400 | } | |
5401 | ||
5402 | /* | |
9f03740a FDBM |
5403 | * If other directory inodes depended on our current directory |
5404 | * inode's move/rename, now do their move/rename operations. | |
31db9f7c | 5405 | */ |
9f03740a FDBM |
5406 | if (!is_waiting_for_move(sctx, sctx->cur_ino)) { |
5407 | ret = apply_children_dir_moves(sctx); | |
5408 | if (ret) | |
5409 | goto out; | |
fcbd2154 FM |
5410 | /* |
5411 | * Need to send that every time, no matter if it actually | |
5412 | * changed between the two trees as we have done changes to | |
5413 | * the inode before. If our inode is a directory and it's | |
5414 | * waiting to be moved/renamed, we will send its utimes when | |
5415 | * it's moved/renamed, therefore we don't need to do it here. | |
5416 | */ | |
5417 | sctx->send_progress = sctx->cur_ino + 1; | |
5418 | ret = send_utimes(sctx, sctx->cur_ino, sctx->cur_inode_gen); | |
5419 | if (ret < 0) | |
5420 | goto out; | |
9f03740a FDBM |
5421 | } |
5422 | ||
31db9f7c AB |
5423 | out: |
5424 | return ret; | |
5425 | } | |
5426 | ||
5427 | static int changed_inode(struct send_ctx *sctx, | |
5428 | enum btrfs_compare_tree_result result) | |
5429 | { | |
5430 | int ret = 0; | |
5431 | struct btrfs_key *key = sctx->cmp_key; | |
5432 | struct btrfs_inode_item *left_ii = NULL; | |
5433 | struct btrfs_inode_item *right_ii = NULL; | |
5434 | u64 left_gen = 0; | |
5435 | u64 right_gen = 0; | |
5436 | ||
31db9f7c AB |
5437 | sctx->cur_ino = key->objectid; |
5438 | sctx->cur_inode_new_gen = 0; | |
16e7549f | 5439 | sctx->cur_inode_last_extent = (u64)-1; |
e479d9bb AB |
5440 | |
5441 | /* | |
5442 | * Set send_progress to current inode. This will tell all get_cur_xxx | |
5443 | * functions that the current inode's refs are not updated yet. Later, | |
5444 | * when process_recorded_refs is finished, it is set to cur_ino + 1. | |
5445 | */ | |
31db9f7c AB |
5446 | sctx->send_progress = sctx->cur_ino; |
5447 | ||
5448 | if (result == BTRFS_COMPARE_TREE_NEW || | |
5449 | result == BTRFS_COMPARE_TREE_CHANGED) { | |
5450 | left_ii = btrfs_item_ptr(sctx->left_path->nodes[0], | |
5451 | sctx->left_path->slots[0], | |
5452 | struct btrfs_inode_item); | |
5453 | left_gen = btrfs_inode_generation(sctx->left_path->nodes[0], | |
5454 | left_ii); | |
5455 | } else { | |
5456 | right_ii = btrfs_item_ptr(sctx->right_path->nodes[0], | |
5457 | sctx->right_path->slots[0], | |
5458 | struct btrfs_inode_item); | |
5459 | right_gen = btrfs_inode_generation(sctx->right_path->nodes[0], | |
5460 | right_ii); | |
5461 | } | |
5462 | if (result == BTRFS_COMPARE_TREE_CHANGED) { | |
5463 | right_ii = btrfs_item_ptr(sctx->right_path->nodes[0], | |
5464 | sctx->right_path->slots[0], | |
5465 | struct btrfs_inode_item); | |
5466 | ||
5467 | right_gen = btrfs_inode_generation(sctx->right_path->nodes[0], | |
5468 | right_ii); | |
6d85ed05 AB |
5469 | |
5470 | /* | |
5471 | * The cur_ino = root dir case is special here. We can't treat | |
5472 | * the inode as deleted+reused because it would generate a | |
5473 | * stream that tries to delete/mkdir the root dir. | |
5474 | */ | |
5475 | if (left_gen != right_gen && | |
5476 | sctx->cur_ino != BTRFS_FIRST_FREE_OBJECTID) | |
31db9f7c AB |
5477 | sctx->cur_inode_new_gen = 1; |
5478 | } | |
5479 | ||
5480 | if (result == BTRFS_COMPARE_TREE_NEW) { | |
5481 | sctx->cur_inode_gen = left_gen; | |
5482 | sctx->cur_inode_new = 1; | |
5483 | sctx->cur_inode_deleted = 0; | |
5484 | sctx->cur_inode_size = btrfs_inode_size( | |
5485 | sctx->left_path->nodes[0], left_ii); | |
5486 | sctx->cur_inode_mode = btrfs_inode_mode( | |
5487 | sctx->left_path->nodes[0], left_ii); | |
644d1940 LB |
5488 | sctx->cur_inode_rdev = btrfs_inode_rdev( |
5489 | sctx->left_path->nodes[0], left_ii); | |
31db9f7c | 5490 | if (sctx->cur_ino != BTRFS_FIRST_FREE_OBJECTID) |
1f4692da | 5491 | ret = send_create_inode_if_needed(sctx); |
31db9f7c AB |
5492 | } else if (result == BTRFS_COMPARE_TREE_DELETED) { |
5493 | sctx->cur_inode_gen = right_gen; | |
5494 | sctx->cur_inode_new = 0; | |
5495 | sctx->cur_inode_deleted = 1; | |
5496 | sctx->cur_inode_size = btrfs_inode_size( | |
5497 | sctx->right_path->nodes[0], right_ii); | |
5498 | sctx->cur_inode_mode = btrfs_inode_mode( | |
5499 | sctx->right_path->nodes[0], right_ii); | |
5500 | } else if (result == BTRFS_COMPARE_TREE_CHANGED) { | |
766702ef AB |
5501 | /* |
5502 | * We need to do some special handling in case the inode was | |
5503 | * reported as changed with a changed generation number. This | |
5504 | * means that the original inode was deleted and new inode | |
5505 | * reused the same inum. So we have to treat the old inode as | |
5506 | * deleted and the new one as new. | |
5507 | */ | |
31db9f7c | 5508 | if (sctx->cur_inode_new_gen) { |
766702ef AB |
5509 | /* |
5510 | * First, process the inode as if it was deleted. | |
5511 | */ | |
31db9f7c AB |
5512 | sctx->cur_inode_gen = right_gen; |
5513 | sctx->cur_inode_new = 0; | |
5514 | sctx->cur_inode_deleted = 1; | |
5515 | sctx->cur_inode_size = btrfs_inode_size( | |
5516 | sctx->right_path->nodes[0], right_ii); | |
5517 | sctx->cur_inode_mode = btrfs_inode_mode( | |
5518 | sctx->right_path->nodes[0], right_ii); | |
5519 | ret = process_all_refs(sctx, | |
5520 | BTRFS_COMPARE_TREE_DELETED); | |
5521 | if (ret < 0) | |
5522 | goto out; | |
5523 | ||
766702ef AB |
5524 | /* |
5525 | * Now process the inode as if it was new. | |
5526 | */ | |
31db9f7c AB |
5527 | sctx->cur_inode_gen = left_gen; |
5528 | sctx->cur_inode_new = 1; | |
5529 | sctx->cur_inode_deleted = 0; | |
5530 | sctx->cur_inode_size = btrfs_inode_size( | |
5531 | sctx->left_path->nodes[0], left_ii); | |
5532 | sctx->cur_inode_mode = btrfs_inode_mode( | |
5533 | sctx->left_path->nodes[0], left_ii); | |
644d1940 LB |
5534 | sctx->cur_inode_rdev = btrfs_inode_rdev( |
5535 | sctx->left_path->nodes[0], left_ii); | |
1f4692da | 5536 | ret = send_create_inode_if_needed(sctx); |
31db9f7c AB |
5537 | if (ret < 0) |
5538 | goto out; | |
5539 | ||
5540 | ret = process_all_refs(sctx, BTRFS_COMPARE_TREE_NEW); | |
5541 | if (ret < 0) | |
5542 | goto out; | |
e479d9bb AB |
5543 | /* |
5544 | * Advance send_progress now as we did not get into | |
5545 | * process_recorded_refs_if_needed in the new_gen case. | |
5546 | */ | |
5547 | sctx->send_progress = sctx->cur_ino + 1; | |
766702ef AB |
5548 | |
5549 | /* | |
5550 | * Now process all extents and xattrs of the inode as if | |
5551 | * they were all new. | |
5552 | */ | |
31db9f7c AB |
5553 | ret = process_all_extents(sctx); |
5554 | if (ret < 0) | |
5555 | goto out; | |
5556 | ret = process_all_new_xattrs(sctx); | |
5557 | if (ret < 0) | |
5558 | goto out; | |
5559 | } else { | |
5560 | sctx->cur_inode_gen = left_gen; | |
5561 | sctx->cur_inode_new = 0; | |
5562 | sctx->cur_inode_new_gen = 0; | |
5563 | sctx->cur_inode_deleted = 0; | |
5564 | sctx->cur_inode_size = btrfs_inode_size( | |
5565 | sctx->left_path->nodes[0], left_ii); | |
5566 | sctx->cur_inode_mode = btrfs_inode_mode( | |
5567 | sctx->left_path->nodes[0], left_ii); | |
5568 | } | |
5569 | } | |
5570 | ||
5571 | out: | |
5572 | return ret; | |
5573 | } | |
5574 | ||
766702ef AB |
5575 | /* |
5576 | * We have to process new refs before deleted refs, but compare_trees gives us | |
5577 | * the new and deleted refs mixed. To fix this, we record the new/deleted refs | |
5578 | * first and later process them in process_recorded_refs. | |
5579 | * For the cur_inode_new_gen case, we skip recording completely because | |
5580 | * changed_inode did already initiate processing of refs. The reason for this is | |
5581 | * that in this case, compare_tree actually compares the refs of 2 different | |
5582 | * inodes. To fix this, process_all_refs is used in changed_inode to handle all | |
5583 | * refs of the right tree as deleted and all refs of the left tree as new. | |
5584 | */ | |
31db9f7c AB |
5585 | static int changed_ref(struct send_ctx *sctx, |
5586 | enum btrfs_compare_tree_result result) | |
5587 | { | |
5588 | int ret = 0; | |
5589 | ||
5590 | BUG_ON(sctx->cur_ino != sctx->cmp_key->objectid); | |
5591 | ||
5592 | if (!sctx->cur_inode_new_gen && | |
5593 | sctx->cur_ino != BTRFS_FIRST_FREE_OBJECTID) { | |
5594 | if (result == BTRFS_COMPARE_TREE_NEW) | |
5595 | ret = record_new_ref(sctx); | |
5596 | else if (result == BTRFS_COMPARE_TREE_DELETED) | |
5597 | ret = record_deleted_ref(sctx); | |
5598 | else if (result == BTRFS_COMPARE_TREE_CHANGED) | |
5599 | ret = record_changed_ref(sctx); | |
5600 | } | |
5601 | ||
5602 | return ret; | |
5603 | } | |
5604 | ||
766702ef AB |
5605 | /* |
5606 | * Process new/deleted/changed xattrs. We skip processing in the | |
5607 | * cur_inode_new_gen case because changed_inode did already initiate processing | |
5608 | * of xattrs. The reason is the same as in changed_ref | |
5609 | */ | |
31db9f7c AB |
5610 | static int changed_xattr(struct send_ctx *sctx, |
5611 | enum btrfs_compare_tree_result result) | |
5612 | { | |
5613 | int ret = 0; | |
5614 | ||
5615 | BUG_ON(sctx->cur_ino != sctx->cmp_key->objectid); | |
5616 | ||
5617 | if (!sctx->cur_inode_new_gen && !sctx->cur_inode_deleted) { | |
5618 | if (result == BTRFS_COMPARE_TREE_NEW) | |
5619 | ret = process_new_xattr(sctx); | |
5620 | else if (result == BTRFS_COMPARE_TREE_DELETED) | |
5621 | ret = process_deleted_xattr(sctx); | |
5622 | else if (result == BTRFS_COMPARE_TREE_CHANGED) | |
5623 | ret = process_changed_xattr(sctx); | |
5624 | } | |
5625 | ||
5626 | return ret; | |
5627 | } | |
5628 | ||
766702ef AB |
5629 | /* |
5630 | * Process new/deleted/changed extents. We skip processing in the | |
5631 | * cur_inode_new_gen case because changed_inode did already initiate processing | |
5632 | * of extents. The reason is the same as in changed_ref | |
5633 | */ | |
31db9f7c AB |
5634 | static int changed_extent(struct send_ctx *sctx, |
5635 | enum btrfs_compare_tree_result result) | |
5636 | { | |
5637 | int ret = 0; | |
5638 | ||
5639 | BUG_ON(sctx->cur_ino != sctx->cmp_key->objectid); | |
5640 | ||
5641 | if (!sctx->cur_inode_new_gen && !sctx->cur_inode_deleted) { | |
5642 | if (result != BTRFS_COMPARE_TREE_DELETED) | |
5643 | ret = process_extent(sctx, sctx->left_path, | |
5644 | sctx->cmp_key); | |
5645 | } | |
5646 | ||
5647 | return ret; | |
5648 | } | |
5649 | ||
ba5e8f2e JB |
5650 | static int dir_changed(struct send_ctx *sctx, u64 dir) |
5651 | { | |
5652 | u64 orig_gen, new_gen; | |
5653 | int ret; | |
5654 | ||
5655 | ret = get_inode_info(sctx->send_root, dir, NULL, &new_gen, NULL, NULL, | |
5656 | NULL, NULL); | |
5657 | if (ret) | |
5658 | return ret; | |
5659 | ||
5660 | ret = get_inode_info(sctx->parent_root, dir, NULL, &orig_gen, NULL, | |
5661 | NULL, NULL, NULL); | |
5662 | if (ret) | |
5663 | return ret; | |
5664 | ||
5665 | return (orig_gen != new_gen) ? 1 : 0; | |
5666 | } | |
5667 | ||
5668 | static int compare_refs(struct send_ctx *sctx, struct btrfs_path *path, | |
5669 | struct btrfs_key *key) | |
5670 | { | |
5671 | struct btrfs_inode_extref *extref; | |
5672 | struct extent_buffer *leaf; | |
5673 | u64 dirid = 0, last_dirid = 0; | |
5674 | unsigned long ptr; | |
5675 | u32 item_size; | |
5676 | u32 cur_offset = 0; | |
5677 | int ref_name_len; | |
5678 | int ret = 0; | |
5679 | ||
5680 | /* Easy case, just check this one dirid */ | |
5681 | if (key->type == BTRFS_INODE_REF_KEY) { | |
5682 | dirid = key->offset; | |
5683 | ||
5684 | ret = dir_changed(sctx, dirid); | |
5685 | goto out; | |
5686 | } | |
5687 | ||
5688 | leaf = path->nodes[0]; | |
5689 | item_size = btrfs_item_size_nr(leaf, path->slots[0]); | |
5690 | ptr = btrfs_item_ptr_offset(leaf, path->slots[0]); | |
5691 | while (cur_offset < item_size) { | |
5692 | extref = (struct btrfs_inode_extref *)(ptr + | |
5693 | cur_offset); | |
5694 | dirid = btrfs_inode_extref_parent(leaf, extref); | |
5695 | ref_name_len = btrfs_inode_extref_name_len(leaf, extref); | |
5696 | cur_offset += ref_name_len + sizeof(*extref); | |
5697 | if (dirid == last_dirid) | |
5698 | continue; | |
5699 | ret = dir_changed(sctx, dirid); | |
5700 | if (ret) | |
5701 | break; | |
5702 | last_dirid = dirid; | |
5703 | } | |
5704 | out: | |
5705 | return ret; | |
5706 | } | |
5707 | ||
766702ef AB |
5708 | /* |
5709 | * Updates compare related fields in sctx and simply forwards to the actual | |
5710 | * changed_xxx functions. | |
5711 | */ | |
31db9f7c AB |
5712 | static int changed_cb(struct btrfs_root *left_root, |
5713 | struct btrfs_root *right_root, | |
5714 | struct btrfs_path *left_path, | |
5715 | struct btrfs_path *right_path, | |
5716 | struct btrfs_key *key, | |
5717 | enum btrfs_compare_tree_result result, | |
5718 | void *ctx) | |
5719 | { | |
5720 | int ret = 0; | |
5721 | struct send_ctx *sctx = ctx; | |
5722 | ||
ba5e8f2e | 5723 | if (result == BTRFS_COMPARE_TREE_SAME) { |
16e7549f JB |
5724 | if (key->type == BTRFS_INODE_REF_KEY || |
5725 | key->type == BTRFS_INODE_EXTREF_KEY) { | |
5726 | ret = compare_refs(sctx, left_path, key); | |
5727 | if (!ret) | |
5728 | return 0; | |
5729 | if (ret < 0) | |
5730 | return ret; | |
5731 | } else if (key->type == BTRFS_EXTENT_DATA_KEY) { | |
5732 | return maybe_send_hole(sctx, left_path, key); | |
5733 | } else { | |
ba5e8f2e | 5734 | return 0; |
16e7549f | 5735 | } |
ba5e8f2e JB |
5736 | result = BTRFS_COMPARE_TREE_CHANGED; |
5737 | ret = 0; | |
5738 | } | |
5739 | ||
31db9f7c AB |
5740 | sctx->left_path = left_path; |
5741 | sctx->right_path = right_path; | |
5742 | sctx->cmp_key = key; | |
5743 | ||
5744 | ret = finish_inode_if_needed(sctx, 0); | |
5745 | if (ret < 0) | |
5746 | goto out; | |
5747 | ||
2981e225 AB |
5748 | /* Ignore non-FS objects */ |
5749 | if (key->objectid == BTRFS_FREE_INO_OBJECTID || | |
5750 | key->objectid == BTRFS_FREE_SPACE_OBJECTID) | |
5751 | goto out; | |
5752 | ||
31db9f7c AB |
5753 | if (key->type == BTRFS_INODE_ITEM_KEY) |
5754 | ret = changed_inode(sctx, result); | |
96b5bd77 JS |
5755 | else if (key->type == BTRFS_INODE_REF_KEY || |
5756 | key->type == BTRFS_INODE_EXTREF_KEY) | |
31db9f7c AB |
5757 | ret = changed_ref(sctx, result); |
5758 | else if (key->type == BTRFS_XATTR_ITEM_KEY) | |
5759 | ret = changed_xattr(sctx, result); | |
5760 | else if (key->type == BTRFS_EXTENT_DATA_KEY) | |
5761 | ret = changed_extent(sctx, result); | |
5762 | ||
5763 | out: | |
5764 | return ret; | |
5765 | } | |
5766 | ||
5767 | static int full_send_tree(struct send_ctx *sctx) | |
5768 | { | |
5769 | int ret; | |
31db9f7c AB |
5770 | struct btrfs_root *send_root = sctx->send_root; |
5771 | struct btrfs_key key; | |
5772 | struct btrfs_key found_key; | |
5773 | struct btrfs_path *path; | |
5774 | struct extent_buffer *eb; | |
5775 | int slot; | |
31db9f7c AB |
5776 | |
5777 | path = alloc_path_for_send(); | |
5778 | if (!path) | |
5779 | return -ENOMEM; | |
5780 | ||
31db9f7c AB |
5781 | key.objectid = BTRFS_FIRST_FREE_OBJECTID; |
5782 | key.type = BTRFS_INODE_ITEM_KEY; | |
5783 | key.offset = 0; | |
5784 | ||
31db9f7c AB |
5785 | ret = btrfs_search_slot_for_read(send_root, &key, path, 1, 0); |
5786 | if (ret < 0) | |
5787 | goto out; | |
5788 | if (ret) | |
5789 | goto out_finish; | |
5790 | ||
5791 | while (1) { | |
31db9f7c AB |
5792 | eb = path->nodes[0]; |
5793 | slot = path->slots[0]; | |
5794 | btrfs_item_key_to_cpu(eb, &found_key, slot); | |
5795 | ||
5796 | ret = changed_cb(send_root, NULL, path, NULL, | |
5797 | &found_key, BTRFS_COMPARE_TREE_NEW, sctx); | |
5798 | if (ret < 0) | |
5799 | goto out; | |
5800 | ||
5801 | key.objectid = found_key.objectid; | |
5802 | key.type = found_key.type; | |
5803 | key.offset = found_key.offset + 1; | |
5804 | ||
5805 | ret = btrfs_next_item(send_root, path); | |
5806 | if (ret < 0) | |
5807 | goto out; | |
5808 | if (ret) { | |
5809 | ret = 0; | |
5810 | break; | |
5811 | } | |
5812 | } | |
5813 | ||
5814 | out_finish: | |
5815 | ret = finish_inode_if_needed(sctx, 1); | |
5816 | ||
5817 | out: | |
5818 | btrfs_free_path(path); | |
31db9f7c AB |
5819 | return ret; |
5820 | } | |
5821 | ||
5822 | static int send_subvol(struct send_ctx *sctx) | |
5823 | { | |
5824 | int ret; | |
5825 | ||
c2c71324 SB |
5826 | if (!(sctx->flags & BTRFS_SEND_FLAG_OMIT_STREAM_HEADER)) { |
5827 | ret = send_header(sctx); | |
5828 | if (ret < 0) | |
5829 | goto out; | |
5830 | } | |
31db9f7c AB |
5831 | |
5832 | ret = send_subvol_begin(sctx); | |
5833 | if (ret < 0) | |
5834 | goto out; | |
5835 | ||
5836 | if (sctx->parent_root) { | |
5837 | ret = btrfs_compare_trees(sctx->send_root, sctx->parent_root, | |
5838 | changed_cb, sctx); | |
5839 | if (ret < 0) | |
5840 | goto out; | |
5841 | ret = finish_inode_if_needed(sctx, 1); | |
5842 | if (ret < 0) | |
5843 | goto out; | |
5844 | } else { | |
5845 | ret = full_send_tree(sctx); | |
5846 | if (ret < 0) | |
5847 | goto out; | |
5848 | } | |
5849 | ||
5850 | out: | |
31db9f7c AB |
5851 | free_recorded_refs(sctx); |
5852 | return ret; | |
5853 | } | |
5854 | ||
e5fa8f86 FM |
5855 | /* |
5856 | * If orphan cleanup did remove any orphans from a root, it means the tree | |
5857 | * was modified and therefore the commit root is not the same as the current | |
5858 | * root anymore. This is a problem, because send uses the commit root and | |
5859 | * therefore can see inode items that don't exist in the current root anymore, | |
5860 | * and for example make calls to btrfs_iget, which will do tree lookups based | |
5861 | * on the current root and not on the commit root. Those lookups will fail, | |
5862 | * returning a -ESTALE error, and making send fail with that error. So make | |
5863 | * sure a send does not see any orphans we have just removed, and that it will | |
5864 | * see the same inodes regardless of whether a transaction commit happened | |
5865 | * before it started (meaning that the commit root will be the same as the | |
5866 | * current root) or not. | |
5867 | */ | |
5868 | static int ensure_commit_roots_uptodate(struct send_ctx *sctx) | |
5869 | { | |
5870 | int i; | |
5871 | struct btrfs_trans_handle *trans = NULL; | |
5872 | ||
5873 | again: | |
5874 | if (sctx->parent_root && | |
5875 | sctx->parent_root->node != sctx->parent_root->commit_root) | |
5876 | goto commit_trans; | |
5877 | ||
5878 | for (i = 0; i < sctx->clone_roots_cnt; i++) | |
5879 | if (sctx->clone_roots[i].root->node != | |
5880 | sctx->clone_roots[i].root->commit_root) | |
5881 | goto commit_trans; | |
5882 | ||
5883 | if (trans) | |
5884 | return btrfs_end_transaction(trans, sctx->send_root); | |
5885 | ||
5886 | return 0; | |
5887 | ||
5888 | commit_trans: | |
5889 | /* Use any root, all fs roots will get their commit roots updated. */ | |
5890 | if (!trans) { | |
5891 | trans = btrfs_join_transaction(sctx->send_root); | |
5892 | if (IS_ERR(trans)) | |
5893 | return PTR_ERR(trans); | |
5894 | goto again; | |
5895 | } | |
5896 | ||
5897 | return btrfs_commit_transaction(trans, sctx->send_root); | |
5898 | } | |
5899 | ||
66ef7d65 DS |
5900 | static void btrfs_root_dec_send_in_progress(struct btrfs_root* root) |
5901 | { | |
5902 | spin_lock(&root->root_item_lock); | |
5903 | root->send_in_progress--; | |
5904 | /* | |
5905 | * Not much left to do, we don't know why it's unbalanced and | |
5906 | * can't blindly reset it to 0. | |
5907 | */ | |
5908 | if (root->send_in_progress < 0) | |
5909 | btrfs_err(root->fs_info, | |
351fd353 | 5910 | "send_in_progres unbalanced %d root %llu", |
66ef7d65 DS |
5911 | root->send_in_progress, root->root_key.objectid); |
5912 | spin_unlock(&root->root_item_lock); | |
5913 | } | |
5914 | ||
31db9f7c AB |
5915 | long btrfs_ioctl_send(struct file *mnt_file, void __user *arg_) |
5916 | { | |
5917 | int ret = 0; | |
5918 | struct btrfs_root *send_root; | |
5919 | struct btrfs_root *clone_root; | |
5920 | struct btrfs_fs_info *fs_info; | |
5921 | struct btrfs_ioctl_send_args *arg = NULL; | |
5922 | struct btrfs_key key; | |
31db9f7c AB |
5923 | struct send_ctx *sctx = NULL; |
5924 | u32 i; | |
5925 | u64 *clone_sources_tmp = NULL; | |
2c686537 | 5926 | int clone_sources_to_rollback = 0; |
896c14f9 | 5927 | int sort_clone_roots = 0; |
18f687d5 | 5928 | int index; |
31db9f7c AB |
5929 | |
5930 | if (!capable(CAP_SYS_ADMIN)) | |
5931 | return -EPERM; | |
5932 | ||
496ad9aa | 5933 | send_root = BTRFS_I(file_inode(mnt_file))->root; |
31db9f7c AB |
5934 | fs_info = send_root->fs_info; |
5935 | ||
2c686537 DS |
5936 | /* |
5937 | * The subvolume must remain read-only during send, protect against | |
521e0546 | 5938 | * making it RW. This also protects against deletion. |
2c686537 DS |
5939 | */ |
5940 | spin_lock(&send_root->root_item_lock); | |
5941 | send_root->send_in_progress++; | |
5942 | spin_unlock(&send_root->root_item_lock); | |
5943 | ||
139f807a JB |
5944 | /* |
5945 | * This is done when we lookup the root, it should already be complete | |
5946 | * by the time we get here. | |
5947 | */ | |
5948 | WARN_ON(send_root->orphan_cleanup_state != ORPHAN_CLEANUP_DONE); | |
5949 | ||
2c686537 DS |
5950 | /* |
5951 | * Userspace tools do the checks and warn the user if it's | |
5952 | * not RO. | |
5953 | */ | |
5954 | if (!btrfs_root_readonly(send_root)) { | |
5955 | ret = -EPERM; | |
5956 | goto out; | |
5957 | } | |
5958 | ||
31db9f7c AB |
5959 | arg = memdup_user(arg_, sizeof(*arg)); |
5960 | if (IS_ERR(arg)) { | |
5961 | ret = PTR_ERR(arg); | |
5962 | arg = NULL; | |
5963 | goto out; | |
5964 | } | |
5965 | ||
5966 | if (!access_ok(VERIFY_READ, arg->clone_sources, | |
700ff4f0 DC |
5967 | sizeof(*arg->clone_sources) * |
5968 | arg->clone_sources_count)) { | |
31db9f7c AB |
5969 | ret = -EFAULT; |
5970 | goto out; | |
5971 | } | |
5972 | ||
c2c71324 | 5973 | if (arg->flags & ~BTRFS_SEND_FLAG_MASK) { |
cb95e7bf MF |
5974 | ret = -EINVAL; |
5975 | goto out; | |
5976 | } | |
5977 | ||
31db9f7c AB |
5978 | sctx = kzalloc(sizeof(struct send_ctx), GFP_NOFS); |
5979 | if (!sctx) { | |
5980 | ret = -ENOMEM; | |
5981 | goto out; | |
5982 | } | |
5983 | ||
5984 | INIT_LIST_HEAD(&sctx->new_refs); | |
5985 | INIT_LIST_HEAD(&sctx->deleted_refs); | |
5986 | INIT_RADIX_TREE(&sctx->name_cache, GFP_NOFS); | |
5987 | INIT_LIST_HEAD(&sctx->name_cache_list); | |
5988 | ||
cb95e7bf MF |
5989 | sctx->flags = arg->flags; |
5990 | ||
31db9f7c | 5991 | sctx->send_filp = fget(arg->send_fd); |
ecc7ada7 TI |
5992 | if (!sctx->send_filp) { |
5993 | ret = -EBADF; | |
31db9f7c AB |
5994 | goto out; |
5995 | } | |
5996 | ||
31db9f7c | 5997 | sctx->send_root = send_root; |
521e0546 DS |
5998 | /* |
5999 | * Unlikely but possible, if the subvolume is marked for deletion but | |
6000 | * is slow to remove the directory entry, send can still be started | |
6001 | */ | |
6002 | if (btrfs_root_dead(sctx->send_root)) { | |
6003 | ret = -EPERM; | |
6004 | goto out; | |
6005 | } | |
6006 | ||
31db9f7c AB |
6007 | sctx->clone_roots_cnt = arg->clone_sources_count; |
6008 | ||
6009 | sctx->send_max_size = BTRFS_SEND_BUF_SIZE; | |
6010 | sctx->send_buf = vmalloc(sctx->send_max_size); | |
6011 | if (!sctx->send_buf) { | |
6012 | ret = -ENOMEM; | |
6013 | goto out; | |
6014 | } | |
6015 | ||
6016 | sctx->read_buf = vmalloc(BTRFS_SEND_READ_SIZE); | |
6017 | if (!sctx->read_buf) { | |
6018 | ret = -ENOMEM; | |
6019 | goto out; | |
6020 | } | |
6021 | ||
9f03740a FDBM |
6022 | sctx->pending_dir_moves = RB_ROOT; |
6023 | sctx->waiting_dir_moves = RB_ROOT; | |
9dc44214 | 6024 | sctx->orphan_dirs = RB_ROOT; |
9f03740a | 6025 | |
31db9f7c AB |
6026 | sctx->clone_roots = vzalloc(sizeof(struct clone_root) * |
6027 | (arg->clone_sources_count + 1)); | |
6028 | if (!sctx->clone_roots) { | |
6029 | ret = -ENOMEM; | |
6030 | goto out; | |
6031 | } | |
6032 | ||
6033 | if (arg->clone_sources_count) { | |
6034 | clone_sources_tmp = vmalloc(arg->clone_sources_count * | |
6035 | sizeof(*arg->clone_sources)); | |
6036 | if (!clone_sources_tmp) { | |
6037 | ret = -ENOMEM; | |
6038 | goto out; | |
6039 | } | |
6040 | ||
6041 | ret = copy_from_user(clone_sources_tmp, arg->clone_sources, | |
6042 | arg->clone_sources_count * | |
6043 | sizeof(*arg->clone_sources)); | |
6044 | if (ret) { | |
6045 | ret = -EFAULT; | |
6046 | goto out; | |
6047 | } | |
6048 | ||
6049 | for (i = 0; i < arg->clone_sources_count; i++) { | |
6050 | key.objectid = clone_sources_tmp[i]; | |
6051 | key.type = BTRFS_ROOT_ITEM_KEY; | |
6052 | key.offset = (u64)-1; | |
18f687d5 WS |
6053 | |
6054 | index = srcu_read_lock(&fs_info->subvol_srcu); | |
6055 | ||
31db9f7c | 6056 | clone_root = btrfs_read_fs_root_no_name(fs_info, &key); |
31db9f7c | 6057 | if (IS_ERR(clone_root)) { |
18f687d5 | 6058 | srcu_read_unlock(&fs_info->subvol_srcu, index); |
31db9f7c AB |
6059 | ret = PTR_ERR(clone_root); |
6060 | goto out; | |
6061 | } | |
2c686537 | 6062 | spin_lock(&clone_root->root_item_lock); |
5cc2b17e FM |
6063 | if (!btrfs_root_readonly(clone_root) || |
6064 | btrfs_root_dead(clone_root)) { | |
2c686537 | 6065 | spin_unlock(&clone_root->root_item_lock); |
18f687d5 | 6066 | srcu_read_unlock(&fs_info->subvol_srcu, index); |
2c686537 DS |
6067 | ret = -EPERM; |
6068 | goto out; | |
6069 | } | |
2f1f465a | 6070 | clone_root->send_in_progress++; |
2c686537 | 6071 | spin_unlock(&clone_root->root_item_lock); |
18f687d5 WS |
6072 | srcu_read_unlock(&fs_info->subvol_srcu, index); |
6073 | ||
31db9f7c | 6074 | sctx->clone_roots[i].root = clone_root; |
2f1f465a | 6075 | clone_sources_to_rollback = i + 1; |
31db9f7c AB |
6076 | } |
6077 | vfree(clone_sources_tmp); | |
6078 | clone_sources_tmp = NULL; | |
6079 | } | |
6080 | ||
6081 | if (arg->parent_root) { | |
6082 | key.objectid = arg->parent_root; | |
6083 | key.type = BTRFS_ROOT_ITEM_KEY; | |
6084 | key.offset = (u64)-1; | |
18f687d5 WS |
6085 | |
6086 | index = srcu_read_lock(&fs_info->subvol_srcu); | |
6087 | ||
31db9f7c | 6088 | sctx->parent_root = btrfs_read_fs_root_no_name(fs_info, &key); |
b1b19596 | 6089 | if (IS_ERR(sctx->parent_root)) { |
18f687d5 | 6090 | srcu_read_unlock(&fs_info->subvol_srcu, index); |
b1b19596 | 6091 | ret = PTR_ERR(sctx->parent_root); |
31db9f7c AB |
6092 | goto out; |
6093 | } | |
18f687d5 | 6094 | |
2c686537 DS |
6095 | spin_lock(&sctx->parent_root->root_item_lock); |
6096 | sctx->parent_root->send_in_progress++; | |
521e0546 DS |
6097 | if (!btrfs_root_readonly(sctx->parent_root) || |
6098 | btrfs_root_dead(sctx->parent_root)) { | |
2c686537 | 6099 | spin_unlock(&sctx->parent_root->root_item_lock); |
18f687d5 | 6100 | srcu_read_unlock(&fs_info->subvol_srcu, index); |
2c686537 DS |
6101 | ret = -EPERM; |
6102 | goto out; | |
6103 | } | |
6104 | spin_unlock(&sctx->parent_root->root_item_lock); | |
18f687d5 WS |
6105 | |
6106 | srcu_read_unlock(&fs_info->subvol_srcu, index); | |
31db9f7c AB |
6107 | } |
6108 | ||
6109 | /* | |
6110 | * Clones from send_root are allowed, but only if the clone source | |
6111 | * is behind the current send position. This is checked while searching | |
6112 | * for possible clone sources. | |
6113 | */ | |
6114 | sctx->clone_roots[sctx->clone_roots_cnt++].root = sctx->send_root; | |
6115 | ||
6116 | /* We do a bsearch later */ | |
6117 | sort(sctx->clone_roots, sctx->clone_roots_cnt, | |
6118 | sizeof(*sctx->clone_roots), __clone_root_cmp_sort, | |
6119 | NULL); | |
896c14f9 | 6120 | sort_clone_roots = 1; |
31db9f7c | 6121 | |
e5fa8f86 FM |
6122 | ret = ensure_commit_roots_uptodate(sctx); |
6123 | if (ret) | |
6124 | goto out; | |
6125 | ||
2755a0de | 6126 | current->journal_info = BTRFS_SEND_TRANS_STUB; |
31db9f7c | 6127 | ret = send_subvol(sctx); |
a26e8c9f | 6128 | current->journal_info = NULL; |
31db9f7c AB |
6129 | if (ret < 0) |
6130 | goto out; | |
6131 | ||
c2c71324 SB |
6132 | if (!(sctx->flags & BTRFS_SEND_FLAG_OMIT_END_CMD)) { |
6133 | ret = begin_cmd(sctx, BTRFS_SEND_C_END); | |
6134 | if (ret < 0) | |
6135 | goto out; | |
6136 | ret = send_cmd(sctx); | |
6137 | if (ret < 0) | |
6138 | goto out; | |
6139 | } | |
31db9f7c AB |
6140 | |
6141 | out: | |
9f03740a FDBM |
6142 | WARN_ON(sctx && !ret && !RB_EMPTY_ROOT(&sctx->pending_dir_moves)); |
6143 | while (sctx && !RB_EMPTY_ROOT(&sctx->pending_dir_moves)) { | |
6144 | struct rb_node *n; | |
6145 | struct pending_dir_move *pm; | |
6146 | ||
6147 | n = rb_first(&sctx->pending_dir_moves); | |
6148 | pm = rb_entry(n, struct pending_dir_move, node); | |
6149 | while (!list_empty(&pm->list)) { | |
6150 | struct pending_dir_move *pm2; | |
6151 | ||
6152 | pm2 = list_first_entry(&pm->list, | |
6153 | struct pending_dir_move, list); | |
6154 | free_pending_move(sctx, pm2); | |
6155 | } | |
6156 | free_pending_move(sctx, pm); | |
6157 | } | |
6158 | ||
6159 | WARN_ON(sctx && !ret && !RB_EMPTY_ROOT(&sctx->waiting_dir_moves)); | |
6160 | while (sctx && !RB_EMPTY_ROOT(&sctx->waiting_dir_moves)) { | |
6161 | struct rb_node *n; | |
6162 | struct waiting_dir_move *dm; | |
6163 | ||
6164 | n = rb_first(&sctx->waiting_dir_moves); | |
6165 | dm = rb_entry(n, struct waiting_dir_move, node); | |
6166 | rb_erase(&dm->node, &sctx->waiting_dir_moves); | |
6167 | kfree(dm); | |
6168 | } | |
6169 | ||
9dc44214 FM |
6170 | WARN_ON(sctx && !ret && !RB_EMPTY_ROOT(&sctx->orphan_dirs)); |
6171 | while (sctx && !RB_EMPTY_ROOT(&sctx->orphan_dirs)) { | |
6172 | struct rb_node *n; | |
6173 | struct orphan_dir_info *odi; | |
6174 | ||
6175 | n = rb_first(&sctx->orphan_dirs); | |
6176 | odi = rb_entry(n, struct orphan_dir_info, node); | |
6177 | free_orphan_dir_info(sctx, odi); | |
6178 | } | |
6179 | ||
896c14f9 WS |
6180 | if (sort_clone_roots) { |
6181 | for (i = 0; i < sctx->clone_roots_cnt; i++) | |
6182 | btrfs_root_dec_send_in_progress( | |
6183 | sctx->clone_roots[i].root); | |
6184 | } else { | |
6185 | for (i = 0; sctx && i < clone_sources_to_rollback; i++) | |
6186 | btrfs_root_dec_send_in_progress( | |
6187 | sctx->clone_roots[i].root); | |
6188 | ||
6189 | btrfs_root_dec_send_in_progress(send_root); | |
6190 | } | |
66ef7d65 DS |
6191 | if (sctx && !IS_ERR_OR_NULL(sctx->parent_root)) |
6192 | btrfs_root_dec_send_in_progress(sctx->parent_root); | |
2c686537 | 6193 | |
31db9f7c AB |
6194 | kfree(arg); |
6195 | vfree(clone_sources_tmp); | |
6196 | ||
6197 | if (sctx) { | |
6198 | if (sctx->send_filp) | |
6199 | fput(sctx->send_filp); | |
6200 | ||
6201 | vfree(sctx->clone_roots); | |
6202 | vfree(sctx->send_buf); | |
6203 | vfree(sctx->read_buf); | |
6204 | ||
6205 | name_cache_free(sctx); | |
6206 | ||
6207 | kfree(sctx); | |
6208 | } | |
6209 | ||
6210 | return ret; | |
6211 | } |