projects / linux-2.6-block.git / blame
? search:
summary | shortlog | log | commit | commitdiff | tree
blob | blame (incremental) | history | HEAD
Merge tag 'ceph-for-4.9-rc2' of git://github.com/ceph/ceph-client
[linux-2.6-block.git] / fs / btrfs / send.c
CommitLineData
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"
ebb8765b 37#include "compression.h"
31db9f7c 38
31db9f7c
AB		 39/*
40 * A fs_path is a helper to dynamically build path names with unknown size.
41 * It reallocates the internal buffer on demand.
42 * It allows fast adding of path elements on the right side (normal path) and
43 * fast adding to the left side (reversed path). A reversed path can also be
44 * unreversed if needed.
45 */
46struct fs_path {
47 union {
48 struct {
49 char *start;
50 char *end;
31db9f7c
AB		 51
52 char *buf;
1f5a7ff9
DS		 53 unsigned short buf_len:15;
54 unsigned short reversed:1;
31db9f7c
AB		 55 char inline_buf[];
56 };
ace01050
DS		 57 /*
58 * Average path length does not exceed 200 bytes, we'll have
59 * better packing in the slab and higher chance to satisfy
60 * a allocation later during send.
61 */
62 char pad[256];
31db9f7c
AB		 63 };
64};
65#define FS_PATH_INLINE_SIZE \
66 (sizeof(struct fs_path) - offsetof(struct fs_path, inline_buf))
67
68
69/* reused for each extent */
70struct clone_root {
71 struct btrfs_root *root;
72 u64 ino;
73 u64 offset;
74
75 u64 found_refs;
76};
77
78#define SEND_CTX_MAX_NAME_CACHE_SIZE 128
79#define SEND_CTX_NAME_CACHE_CLEAN_SIZE (SEND_CTX_MAX_NAME_CACHE_SIZE * 2)
80
81struct send_ctx {
82 struct file *send_filp;
83 loff_t send_off;
84 char *send_buf;
85 u32 send_size;
86 u32 send_max_size;
87 u64 total_send_size;
88 u64 cmd_send_size[BTRFS_SEND_C_MAX + 1];
cb95e7bf 89 u64 flags; /* 'flags' member of btrfs_ioctl_send_args is u64 */
31db9f7c 90
31db9f7c
AB		 91 struct btrfs_root *send_root;
92 struct btrfs_root *parent_root;
93 struct clone_root *clone_roots;
94 int clone_roots_cnt;
95
96 /* current state of the compare_tree call */
97 struct btrfs_path *left_path;
98 struct btrfs_path *right_path;
99 struct btrfs_key *cmp_key;
100
101 /*
102 * infos of the currently processed inode. In case of deleted inodes,
103 * these are the values from the deleted inode.
104 */
105 u64 cur_ino;
106 u64 cur_inode_gen;
107 int cur_inode_new;
108 int cur_inode_new_gen;
109 int cur_inode_deleted;
31db9f7c
AB		 110 u64 cur_inode_size;
111 u64 cur_inode_mode;
644d1940 112 u64 cur_inode_rdev;
16e7549f 113 u64 cur_inode_last_extent;
31db9f7c
AB		 114
115 u64 send_progress;
116
117 struct list_head new_refs;
118 struct list_head deleted_refs;
119
120 struct radix_tree_root name_cache;
121 struct list_head name_cache_list;
122 int name_cache_size;
123
2131bcd3
LB		 124 struct file_ra_state ra;
125
31db9f7c 126 char *read_buf;
9f03740a
FDBM		 127
128 /*
129 * We process inodes by their increasing order, so if before an
130 * incremental send we reverse the parent/child relationship of
131 * directories such that a directory with a lower inode number was
132 * the parent of a directory with a higher inode number, and the one
133 * becoming the new parent got renamed too, we can't rename/move the
134 * directory with lower inode number when we finish processing it - we
135 * must process the directory with higher inode number first, then
136 * rename/move it and then rename/move the directory with lower inode
137 * number. Example follows.
138 *
139 * Tree state when the first send was performed:
140 *
141 * .
142 * |-- a (ino 257)
143 * |-- b (ino 258)
144 * |
145 * |
146 * |-- c (ino 259)
147 * | |-- d (ino 260)
148 * |
149 * |-- c2 (ino 261)
150 *
151 * Tree state when the second (incremental) send is performed:
152 *
153 * .
154 * |-- a (ino 257)
155 * |-- b (ino 258)
156 * |-- c2 (ino 261)
157 * |-- d2 (ino 260)
158 * |-- cc (ino 259)
159 *
160 * The sequence of steps that lead to the second state was:
161 *
162 * mv /a/b/c/d /a/b/c2/d2
163 * mv /a/b/c /a/b/c2/d2/cc
164 *
165 * "c" has lower inode number, but we can't move it (2nd mv operation)
166 * before we move "d", which has higher inode number.
167 *
168 * So we just memorize which move/rename operations must be performed
169 * later when their respective parent is processed and moved/renamed.
170 */
171
172 /* Indexed by parent directory inode number. */
173 struct rb_root pending_dir_moves;
174
175 /*
176 * Reverse index, indexed by the inode number of a directory that
177 * is waiting for the move/rename of its immediate parent before its
178 * own move/rename can be performed.
179 */
180 struct rb_root waiting_dir_moves;
9dc44214
FM		 181
182 /*
183 * A directory that is going to be rm'ed might have a child directory
184 * which is in the pending directory moves index above. In this case,
185 * the directory can only be removed after the move/rename of its child
186 * is performed. Example:
187 *
188 * Parent snapshot:
189 *
190 * . (ino 256)
191 * |-- a/ (ino 257)
192 * |-- b/ (ino 258)
193 * |-- c/ (ino 259)
194 * | |-- x/ (ino 260)
195 * |
196 * |-- y/ (ino 261)
197 *
198 * Send snapshot:
199 *
200 * . (ino 256)
201 * |-- a/ (ino 257)
202 * |-- b/ (ino 258)
203 * |-- YY/ (ino 261)
204 * |-- x/ (ino 260)
205 *
206 * Sequence of steps that lead to the send snapshot:
207 * rm -f /a/b/c/foo.txt
208 * mv /a/b/y /a/b/YY
209 * mv /a/b/c/x /a/b/YY
210 * rmdir /a/b/c
211 *
212 * When the child is processed, its move/rename is delayed until its
213 * parent is processed (as explained above), but all other operations
214 * like update utimes, chown, chgrp, etc, are performed and the paths
215 * that it uses for those operations must use the orphanized name of
216 * its parent (the directory we're going to rm later), so we need to
217 * memorize that name.
218 *
219 * Indexed by the inode number of the directory to be deleted.
220 */
221 struct rb_root orphan_dirs;
9f03740a
FDBM		 222};
223
224struct pending_dir_move {
225 struct rb_node node;
226 struct list_head list;
227 u64 parent_ino;
228 u64 ino;
229 u64 gen;
230 struct list_head update_refs;
231};
232
233struct waiting_dir_move {
234 struct rb_node node;
235 u64 ino;
9dc44214
FM		 236 /*
237 * There might be some directory that could not be removed because it
238 * was waiting for this directory inode to be moved first. Therefore
239 * after this directory is moved, we can try to rmdir the ino rmdir_ino.
240 */
241 u64 rmdir_ino;
8b191a68 242 bool orphanized;
9dc44214
FM		 243};
244
245struct orphan_dir_info {
246 struct rb_node node;
247 u64 ino;
248 u64 gen;
31db9f7c
AB		 249};
250
251struct name_cache_entry {
252 struct list_head list;
7e0926fe
AB		 253 /*
254 * radix_tree has only 32bit entries but we need to handle 64bit inums.
255 * We use the lower 32bit of the 64bit inum to store it in the tree. If
256 * more then one inum would fall into the same entry, we use radix_list
257 * to store the additional entries. radix_list is also used to store
258 * entries where two entries have the same inum but different
259 * generations.
260 */
261 struct list_head radix_list;
31db9f7c
AB		 262 u64 ino;
263 u64 gen;
264 u64 parent_ino;
265 u64 parent_gen;
266 int ret;
267 int need_later_update;
268 int name_len;
269 char name[];
270};
271
95155585
FM		 272static void inconsistent_snapshot_error(struct send_ctx *sctx,
273 enum btrfs_compare_tree_result result,
274 const char *what)
275{
276 const char *result_string;
277
278 switch (result) {
279 case BTRFS_COMPARE_TREE_NEW:
280 result_string = "new";
281 break;
282 case BTRFS_COMPARE_TREE_DELETED:
283 result_string = "deleted";
284 break;
285 case BTRFS_COMPARE_TREE_CHANGED:
286 result_string = "updated";
287 break;
288 case BTRFS_COMPARE_TREE_SAME:
289 ASSERT(0);
290 result_string = "unchanged";
291 break;
292 default:
293 ASSERT(0);
294 result_string = "unexpected";
295 }
296
297 btrfs_err(sctx->send_root->fs_info,
298 "Send: inconsistent snapshot, found %s %s for inode %llu without updated inode item, send root is %llu, parent root is %llu",
299 result_string, what, sctx->cmp_key->objectid,
300 sctx->send_root->root_key.objectid,
301 (sctx->parent_root ?
302 sctx->parent_root->root_key.objectid : 0));
303}
304
9f03740a
FDBM		 305static int is_waiting_for_move(struct send_ctx *sctx, u64 ino);
306
9dc44214
FM		 307static struct waiting_dir_move *
308get_waiting_dir_move(struct send_ctx *sctx, u64 ino);
309
310static int is_waiting_for_rm(struct send_ctx *sctx, u64 dir_ino);
311
16e7549f
JB		 312static int need_send_hole(struct send_ctx *sctx)
313{
314 return (sctx->parent_root && !sctx->cur_inode_new &&
315 !sctx->cur_inode_new_gen && !sctx->cur_inode_deleted &&
316 S_ISREG(sctx->cur_inode_mode));
317}
318
31db9f7c
AB		 319static void fs_path_reset(struct fs_path *p)
320{
321 if (p->reversed) {
322 p->start = p->buf + p->buf_len - 1;
323 p->end = p->start;
324 *p->start = 0;
325 } else {
326 p->start = p->buf;
327 p->end = p->start;
328 *p->start = 0;
329 }
330}
331
924794c9 332static struct fs_path *fs_path_alloc(void)
31db9f7c
AB		 333{
334 struct fs_path *p;
335
e780b0d1 336 p = kmalloc(sizeof(*p), GFP_KERNEL);
31db9f7c
AB		 337 if (!p)
338 return NULL;
339 p->reversed = 0;
31db9f7c
AB		 340 p->buf = p->inline_buf;
341 p->buf_len = FS_PATH_INLINE_SIZE;
342 fs_path_reset(p);
343 return p;
344}
345
924794c9 346static struct fs_path *fs_path_alloc_reversed(void)
31db9f7c
AB		 347{
348 struct fs_path *p;
349
924794c9 350 p = fs_path_alloc();
31db9f7c
AB		 351 if (!p)
352 return NULL;
353 p->reversed = 1;
354 fs_path_reset(p);
355 return p;
356}
357
924794c9 358static void fs_path_free(struct fs_path *p)
31db9f7c
AB		 359{
360 if (!p)
361 return;
ace01050
DS		 362 if (p->buf != p->inline_buf)
363 kfree(p->buf);
31db9f7c
AB		 364 kfree(p);
365}
366
367static int fs_path_len(struct fs_path *p)
368{
369 return p->end - p->start;
370}
371
372static int fs_path_ensure_buf(struct fs_path *p, int len)
373{
374 char *tmp_buf;
375 int path_len;
376 int old_buf_len;
377
378 len++;
379
380 if (p->buf_len >= len)
381 return 0;
382
cfd4a535
CM		 383 if (len > PATH_MAX) {
384 WARN_ON(1);
385 return -ENOMEM;
386 }
387
1b2782c8
DS		 388 path_len = p->end - p->start;
389 old_buf_len = p->buf_len;
390
ace01050
DS		 391 /*
392 * First time the inline_buf does not suffice
393 */
01a9a8a9 394 if (p->buf == p->inline_buf) {
e780b0d1 395 tmp_buf = kmalloc(len, GFP_KERNEL);
01a9a8a9
FM		 396 if (tmp_buf)
397 memcpy(tmp_buf, p->buf, old_buf_len);
398 } else {
e780b0d1 399 tmp_buf = krealloc(p->buf, len, GFP_KERNEL);
01a9a8a9 400 }
9c9ca00b
DS		 401 if (!tmp_buf)
402 return -ENOMEM;
403 p->buf = tmp_buf;
404 /*
405 * The real size of the buffer is bigger, this will let the fast path
406 * happen most of the time
407 */
408 p->buf_len = ksize(p->buf);
ace01050 409
31db9f7c
AB		 410 if (p->reversed) {
411 tmp_buf = p->buf + old_buf_len - path_len - 1;
412 p->end = p->buf + p->buf_len - 1;
413 p->start = p->end - path_len;
414 memmove(p->start, tmp_buf, path_len + 1);
415 } else {
416 p->start = p->buf;
417 p->end = p->start + path_len;
418 }
419 return 0;
420}
421
b23ab57d
DS		 422static int fs_path_prepare_for_add(struct fs_path *p, int name_len,
423 char **prepared)
31db9f7c
AB		 424{
425 int ret;
426 int new_len;
427
428 new_len = p->end - p->start + name_len;
429 if (p->start != p->end)
430 new_len++;
431 ret = fs_path_ensure_buf(p, new_len);
432 if (ret < 0)
433 goto out;
434
435 if (p->reversed) {
436 if (p->start != p->end)
437 *--p->start = '/';
438 p->start -= name_len;
b23ab57d 439 *prepared = p->start;
31db9f7c
AB		 440 } else {
441 if (p->start != p->end)
442 *p->end++ = '/';
b23ab57d 443 *prepared = p->end;
31db9f7c
AB		 444 p->end += name_len;
445 *p->end = 0;
446 }
447
448out:
449 return ret;
450}
451
452static int fs_path_add(struct fs_path *p, const char *name, int name_len)
453{
454 int ret;
b23ab57d 455 char *prepared;
31db9f7c 456
b23ab57d 457 ret = fs_path_prepare_for_add(p, name_len, &prepared);
31db9f7c
AB		 458 if (ret < 0)
459 goto out;
b23ab57d 460 memcpy(prepared, name, name_len);
31db9f7c
AB		 461
462out:
463 return ret;
464}
465
466static int fs_path_add_path(struct fs_path *p, struct fs_path *p2)
467{
468 int ret;
b23ab57d 469 char *prepared;
31db9f7c 470
b23ab57d 471 ret = fs_path_prepare_for_add(p, p2->end - p2->start, &prepared);
31db9f7c
AB		 472 if (ret < 0)
473 goto out;
b23ab57d 474 memcpy(prepared, p2->start, p2->end - p2->start);
31db9f7c
AB		 475
476out:
477 return ret;
478}
479
480static int fs_path_add_from_extent_buffer(struct fs_path *p,
481 struct extent_buffer *eb,
482 unsigned long off, int len)
483{
484 int ret;
b23ab57d 485 char *prepared;
31db9f7c 486
b23ab57d 487 ret = fs_path_prepare_for_add(p, len, &prepared);
31db9f7c
AB		 488 if (ret < 0)
489 goto out;
490
b23ab57d 491 read_extent_buffer(eb, prepared, off, len);
31db9f7c
AB		 492
493out:
494 return ret;
495}
496
31db9f7c
AB		 497static int fs_path_copy(struct fs_path *p, struct fs_path *from)
498{
499 int ret;
500
501 p->reversed = from->reversed;
502 fs_path_reset(p);
503
504 ret = fs_path_add_path(p, from);
505
506 return ret;
507}
508
509
510static void fs_path_unreverse(struct fs_path *p)
511{
512 char *tmp;
513 int len;
514
515 if (!p->reversed)
516 return;
517
518 tmp = p->start;
519 len = p->end - p->start;
520 p->start = p->buf;
521 p->end = p->start + len;
522 memmove(p->start, tmp, len + 1);
523 p->reversed = 0;
524}
525
526static struct btrfs_path *alloc_path_for_send(void)
527{
528 struct btrfs_path *path;
529
530 path = btrfs_alloc_path();
531 if (!path)
532 return NULL;
533 path->search_commit_root = 1;
534 path->skip_locking = 1;
3f8a18cc 535 path->need_commit_sem = 1;
31db9f7c
AB		 536 return path;
537}
538
48a3b636 539static int write_buf(struct file *filp, const void *buf, u32 len, loff_t *off)
31db9f7c
AB		 540{
541 int ret;
542 mm_segment_t old_fs;
543 u32 pos = 0;
544
545 old_fs = get_fs();
546 set_fs(KERNEL_DS);
547
548 while (pos < len) {
d447d0da
FF		 549 ret = vfs_write(filp, (__force const char __user *)buf + pos,
550 len - pos, off);
31db9f7c
AB		 551 /* TODO handle that correctly */
552 /*if (ret == -ERESTARTSYS) {
553 continue;
554 }*/
555 if (ret < 0)
556 goto out;
557 if (ret == 0) {
558 ret = -EIO;
559 goto out;
560 }
561 pos += ret;
562 }
563
564 ret = 0;
565
566out:
567 set_fs(old_fs);
568 return ret;
569}
570
571static int tlv_put(struct send_ctx *sctx, u16 attr, const void *data, int len)
572{
573 struct btrfs_tlv_header *hdr;
574 int total_len = sizeof(*hdr) + len;
575 int left = sctx->send_max_size - sctx->send_size;
576
577 if (unlikely(left < total_len))
578 return -EOVERFLOW;
579
580 hdr = (struct btrfs_tlv_header *) (sctx->send_buf + sctx->send_size);
581 hdr->tlv_type = cpu_to_le16(attr);
582 hdr->tlv_len = cpu_to_le16(len);
583 memcpy(hdr + 1, data, len);
584 sctx->send_size += total_len;
585
586 return 0;
587}
588
95bc79d5
DS		 589#define TLV_PUT_DEFINE_INT(bits) \
590 static int tlv_put_u##bits(struct send_ctx *sctx, \
591 u##bits attr, u##bits value) \
592 { \
593 __le##bits __tmp = cpu_to_le##bits(value); \
594 return tlv_put(sctx, attr, &__tmp, sizeof(__tmp)); \
595 }
31db9f7c 596
95bc79d5 597TLV_PUT_DEFINE_INT(64)
31db9f7c
AB		 598
599static int tlv_put_string(struct send_ctx *sctx, u16 attr,
600 const char *str, int len)
601{
602 if (len == -1)
603 len = strlen(str);
604 return tlv_put(sctx, attr, str, len);
605}
606
607static int tlv_put_uuid(struct send_ctx *sctx, u16 attr,
608 const u8 *uuid)
609{
610 return tlv_put(sctx, attr, uuid, BTRFS_UUID_SIZE);
611}
612
31db9f7c
AB		 613static int tlv_put_btrfs_timespec(struct send_ctx *sctx, u16 attr,
614 struct extent_buffer *eb,
615 struct btrfs_timespec *ts)
616{
617 struct btrfs_timespec bts;
618 read_extent_buffer(eb, &bts, (unsigned long)ts, sizeof(bts));
619 return tlv_put(sctx, attr, &bts, sizeof(bts));
620}
621
622
623#define TLV_PUT(sctx, attrtype, attrlen, data) \
624 do { \
625 ret = tlv_put(sctx, attrtype, attrlen, data); \
626 if (ret < 0) \
627 goto tlv_put_failure; \
628 } while (0)
629
630#define TLV_PUT_INT(sctx, attrtype, bits, value) \
631 do { \
632 ret = tlv_put_u##bits(sctx, attrtype, value); \
633 if (ret < 0) \
634 goto tlv_put_failure; \
635 } while (0)
636
637#define TLV_PUT_U8(sctx, attrtype, data) TLV_PUT_INT(sctx, attrtype, 8, data)
638#define TLV_PUT_U16(sctx, attrtype, data) TLV_PUT_INT(sctx, attrtype, 16, data)
639#define TLV_PUT_U32(sctx, attrtype, data) TLV_PUT_INT(sctx, attrtype, 32, data)
640#define TLV_PUT_U64(sctx, attrtype, data) TLV_PUT_INT(sctx, attrtype, 64, data)
641#define TLV_PUT_STRING(sctx, attrtype, str, len) \
642 do { \
643 ret = tlv_put_string(sctx, attrtype, str, len); \
644 if (ret < 0) \
645 goto tlv_put_failure; \
646 } while (0)
647#define TLV_PUT_PATH(sctx, attrtype, p) \
648 do { \
649 ret = tlv_put_string(sctx, attrtype, p->start, \
650 p->end - p->start); \
651 if (ret < 0) \
652 goto tlv_put_failure; \
653 } while(0)
654#define TLV_PUT_UUID(sctx, attrtype, uuid) \
655 do { \
656 ret = tlv_put_uuid(sctx, attrtype, uuid); \
657 if (ret < 0) \
658 goto tlv_put_failure; \
659 } while (0)
31db9f7c
AB		 660#define TLV_PUT_BTRFS_TIMESPEC(sctx, attrtype, eb, ts) \
661 do { \
662 ret = tlv_put_btrfs_timespec(sctx, attrtype, eb, ts); \
663 if (ret < 0) \
664 goto tlv_put_failure; \
665 } while (0)
666
667static int send_header(struct send_ctx *sctx)
668{
669 struct btrfs_stream_header hdr;
670
671 strcpy(hdr.magic, BTRFS_SEND_STREAM_MAGIC);
672 hdr.version = cpu_to_le32(BTRFS_SEND_STREAM_VERSION);
673
1bcea355
AJ		 674 return write_buf(sctx->send_filp, &hdr, sizeof(hdr),
675 &sctx->send_off);
31db9f7c
AB		 676}
677
678/*
679 * For each command/item we want to send to userspace, we call this function.
680 */
681static int begin_cmd(struct send_ctx *sctx, int cmd)
682{
683 struct btrfs_cmd_header *hdr;
684
fae7f21c 685 if (WARN_ON(!sctx->send_buf))
31db9f7c 686 return -EINVAL;
31db9f7c
AB		 687
688 BUG_ON(sctx->send_size);
689
690 sctx->send_size += sizeof(*hdr);
691 hdr = (struct btrfs_cmd_header *)sctx->send_buf;
692 hdr->cmd = cpu_to_le16(cmd);
693
694 return 0;
695}
696
697static int send_cmd(struct send_ctx *sctx)
698{
699 int ret;
700 struct btrfs_cmd_header *hdr;
701 u32 crc;
702
703 hdr = (struct btrfs_cmd_header *)sctx->send_buf;
704 hdr->len = cpu_to_le32(sctx->send_size - sizeof(*hdr));
705 hdr->crc = 0;
706
0b947aff 707 crc = btrfs_crc32c(0, (unsigned char *)sctx->send_buf, sctx->send_size);
31db9f7c
AB		 708 hdr->crc = cpu_to_le32(crc);
709
1bcea355
AJ		 710 ret = write_buf(sctx->send_filp, sctx->send_buf, sctx->send_size,
711 &sctx->send_off);
31db9f7c
AB		 712
713 sctx->total_send_size += sctx->send_size;
714 sctx->cmd_send_size[le16_to_cpu(hdr->cmd)] += sctx->send_size;
715 sctx->send_size = 0;
716
717 return ret;
718}
719
720/*
721 * Sends a move instruction to user space
722 */
723static int send_rename(struct send_ctx *sctx,
724 struct fs_path *from, struct fs_path *to)
725{
04ab956e 726 struct btrfs_fs_info *fs_info = sctx->send_root->fs_info;
31db9f7c
AB		 727 int ret;
728
04ab956e 729 btrfs_debug(fs_info, "send_rename %s -> %s", from->start, to->start);
31db9f7c
AB		 730
731 ret = begin_cmd(sctx, BTRFS_SEND_C_RENAME);
732 if (ret < 0)
733 goto out;
734
735 TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, from);
736 TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH_TO, to);
737
738 ret = send_cmd(sctx);
739
740tlv_put_failure:
741out:
742 return ret;
743}
744
745/*
746 * Sends a link instruction to user space
747 */
748static int send_link(struct send_ctx *sctx,
749 struct fs_path *path, struct fs_path *lnk)
750{
04ab956e 751 struct btrfs_fs_info *fs_info = sctx->send_root->fs_info;
31db9f7c
AB		 752 int ret;
753
04ab956e 754 btrfs_debug(fs_info, "send_link %s -> %s", path->start, lnk->start);
31db9f7c
AB		 755
756 ret = begin_cmd(sctx, BTRFS_SEND_C_LINK);
757 if (ret < 0)
758 goto out;
759
760 TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, path);
761 TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH_LINK, lnk);
762
763 ret = send_cmd(sctx);
764
765tlv_put_failure:
766out:
767 return ret;
768}
769
770/*
771 * Sends an unlink instruction to user space
772 */
773static int send_unlink(struct send_ctx *sctx, struct fs_path *path)
774{
04ab956e 775 struct btrfs_fs_info *fs_info = sctx->send_root->fs_info;
31db9f7c
AB		 776 int ret;
777
04ab956e 778 btrfs_debug(fs_info, "send_unlink %s", path->start);
31db9f7c
AB		 779
780 ret = begin_cmd(sctx, BTRFS_SEND_C_UNLINK);
781 if (ret < 0)
782 goto out;
783
784 TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, path);
785
786 ret = send_cmd(sctx);
787
788tlv_put_failure:
789out:
790 return ret;
791}
792
793/*
794 * Sends a rmdir instruction to user space
795 */
796static int send_rmdir(struct send_ctx *sctx, struct fs_path *path)
797{
04ab956e 798 struct btrfs_fs_info *fs_info = sctx->send_root->fs_info;
31db9f7c
AB		 799 int ret;
800
04ab956e 801 btrfs_debug(fs_info, "send_rmdir %s", path->start);
31db9f7c
AB		 802
803 ret = begin_cmd(sctx, BTRFS_SEND_C_RMDIR);
804 if (ret < 0)
805 goto out;
806
807 TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, path);
808
809 ret = send_cmd(sctx);
810
811tlv_put_failure:
812out:
813 return ret;
814}
815
816/*
817 * Helper function to retrieve some fields from an inode item.
818 */
3f8a18cc
JB		 819static int __get_inode_info(struct btrfs_root *root, struct btrfs_path *path,
820 u64 ino, u64 *size, u64 *gen, u64 *mode, u64 *uid,
821 u64 *gid, u64 *rdev)
31db9f7c
AB		 822{
823 int ret;
824 struct btrfs_inode_item *ii;
825 struct btrfs_key key;
31db9f7c
AB		 826
827 key.objectid = ino;
828 key.type = BTRFS_INODE_ITEM_KEY;
829 key.offset = 0;
830 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
31db9f7c 831 if (ret) {
3f8a18cc
JB		 832 if (ret > 0)
833 ret = -ENOENT;
834 return ret;
31db9f7c
AB		 835 }
836
837 ii = btrfs_item_ptr(path->nodes[0], path->slots[0],
838 struct btrfs_inode_item);
839 if (size)
840 *size = btrfs_inode_size(path->nodes[0], ii);
841 if (gen)
842 *gen = btrfs_inode_generation(path->nodes[0], ii);
843 if (mode)
844 *mode = btrfs_inode_mode(path->nodes[0], ii);
845 if (uid)
846 *uid = btrfs_inode_uid(path->nodes[0], ii);
847 if (gid)
848 *gid = btrfs_inode_gid(path->nodes[0], ii);
85a7b33b
AB		 849 if (rdev)
850 *rdev = btrfs_inode_rdev(path->nodes[0], ii);
31db9f7c 851
3f8a18cc
JB		 852 return ret;
853}
854
855static int get_inode_info(struct btrfs_root *root,
856 u64 ino, u64 *size, u64 *gen,
857 u64 *mode, u64 *uid, u64 *gid,
858 u64 *rdev)
859{
860 struct btrfs_path *path;
861 int ret;
862
863 path = alloc_path_for_send();
864 if (!path)
865 return -ENOMEM;
866 ret = __get_inode_info(root, path, ino, size, gen, mode, uid, gid,
867 rdev);
31db9f7c
AB		 868 btrfs_free_path(path);
869 return ret;
870}
871
872typedef int (*iterate_inode_ref_t)(int num, u64 dir, int index,
873 struct fs_path *p,
874 void *ctx);
875
876/*
96b5bd77
JS		 877 * Helper function to iterate the entries in ONE btrfs_inode_ref or
878 * btrfs_inode_extref.
31db9f7c
AB		 879 * The iterate callback may return a non zero value to stop iteration. This can
880 * be a negative value for error codes or 1 to simply stop it.
881 *
96b5bd77 882 * path must point to the INODE_REF or INODE_EXTREF when called.
31db9f7c 883 */
924794c9 884static int iterate_inode_ref(struct btrfs_root *root, struct btrfs_path *path,
31db9f7c
AB		 885 struct btrfs_key *found_key, int resolve,
886 iterate_inode_ref_t iterate, void *ctx)
887{
96b5bd77 888 struct extent_buffer *eb = path->nodes[0];
31db9f7c
AB		 889 struct btrfs_item *item;
890 struct btrfs_inode_ref *iref;
96b5bd77 891 struct btrfs_inode_extref *extref;
31db9f7c
AB		 892 struct btrfs_path *tmp_path;
893 struct fs_path *p;
96b5bd77 894 u32 cur = 0;
31db9f7c 895 u32 total;
96b5bd77 896 int slot = path->slots[0];
31db9f7c
AB		 897 u32 name_len;
898 char *start;
899 int ret = 0;
96b5bd77 900 int num = 0;
31db9f7c 901 int index;
96b5bd77
JS		 902 u64 dir;
903 unsigned long name_off;
904 unsigned long elem_size;
905 unsigned long ptr;
31db9f7c 906
924794c9 907 p = fs_path_alloc_reversed();
31db9f7c
AB		 908 if (!p)
909 return -ENOMEM;
910
911 tmp_path = alloc_path_for_send();
912 if (!tmp_path) {
924794c9 913 fs_path_free(p);
31db9f7c
AB		 914 return -ENOMEM;
915 }
916
31db9f7c 917
96b5bd77
JS		 918 if (found_key->type == BTRFS_INODE_REF_KEY) {
919 ptr = (unsigned long)btrfs_item_ptr(eb, slot,
920 struct btrfs_inode_ref);
dd3cc16b 921 item = btrfs_item_nr(slot);
96b5bd77
JS		 922 total = btrfs_item_size(eb, item);
923 elem_size = sizeof(*iref);
924 } else {
925 ptr = btrfs_item_ptr_offset(eb, slot);
926 total = btrfs_item_size_nr(eb, slot);
927 elem_size = sizeof(*extref);
928 }
929
31db9f7c
AB		 930 while (cur < total) {
931 fs_path_reset(p);
932
96b5bd77
JS		 933 if (found_key->type == BTRFS_INODE_REF_KEY) {
934 iref = (struct btrfs_inode_ref *)(ptr + cur);
935 name_len = btrfs_inode_ref_name_len(eb, iref);
936 name_off = (unsigned long)(iref + 1);
937 index = btrfs_inode_ref_index(eb, iref);
938 dir = found_key->offset;
939 } else {
940 extref = (struct btrfs_inode_extref *)(ptr + cur);
941 name_len = btrfs_inode_extref_name_len(eb, extref);
942 name_off = (unsigned long)&extref->name;
943 index = btrfs_inode_extref_index(eb, extref);
944 dir = btrfs_inode_extref_parent(eb, extref);
945 }
946
31db9f7c 947 if (resolve) {
96b5bd77
JS		 948 start = btrfs_ref_to_path(root, tmp_path, name_len,
949 name_off, eb, dir,
950 p->buf, p->buf_len);
31db9f7c
AB		 951 if (IS_ERR(start)) {
952 ret = PTR_ERR(start);
953 goto out;
954 }
955 if (start < p->buf) {
956 /* overflow , try again with larger buffer */
957 ret = fs_path_ensure_buf(p,
958 p->buf_len + p->buf - start);
959 if (ret < 0)
960 goto out;
96b5bd77
JS		 961 start = btrfs_ref_to_path(root, tmp_path,
962 name_len, name_off,
963 eb, dir,
964 p->buf, p->buf_len);
31db9f7c
AB		 965 if (IS_ERR(start)) {
966 ret = PTR_ERR(start);
967 goto out;
968 }
969 BUG_ON(start < p->buf);
970 }
971 p->start = start;
972 } else {
96b5bd77
JS		 973 ret = fs_path_add_from_extent_buffer(p, eb, name_off,
974 name_len);
31db9f7c
AB		 975 if (ret < 0)
976 goto out;
977 }
978
96b5bd77
JS		 979 cur += elem_size + name_len;
980 ret = iterate(num, dir, index, p, ctx);
31db9f7c
AB		 981 if (ret)
982 goto out;
31db9f7c
AB		 983 num++;
984 }
985
986out:
987 btrfs_free_path(tmp_path);
924794c9 988 fs_path_free(p);
31db9f7c
AB		 989 return ret;
990}
991
992typedef int (*iterate_dir_item_t)(int num, struct btrfs_key *di_key,
993 const char *name, int name_len,
994 const char *data, int data_len,
995 u8 type, void *ctx);
996
997/*
998 * Helper function to iterate the entries in ONE btrfs_dir_item.
999 * The iterate callback may return a non zero value to stop iteration. This can
1000 * be a negative value for error codes or 1 to simply stop it.
1001 *
1002 * path must point to the dir item when called.
1003 */
924794c9 1004static int iterate_dir_item(struct btrfs_root *root, struct btrfs_path *path,
31db9f7c
AB		 1005 struct btrfs_key *found_key,
1006 iterate_dir_item_t iterate, void *ctx)
1007{
1008 int ret = 0;
1009 struct extent_buffer *eb;
1010 struct btrfs_item *item;
1011 struct btrfs_dir_item *di;
31db9f7c
AB		 1012 struct btrfs_key di_key;
1013 char *buf = NULL;
7e3ae33e 1014 int buf_len;
31db9f7c
AB		 1015 u32 name_len;
1016 u32 data_len;
1017 u32 cur;
1018 u32 len;
1019 u32 total;
1020 int slot;
1021 int num;
1022 u8 type;
1023
4395e0c4
FM		 1024 /*
1025 * Start with a small buffer (1 page). If later we end up needing more
1026 * space, which can happen for xattrs on a fs with a leaf size greater
1027 * then the page size, attempt to increase the buffer. Typically xattr
1028 * values are small.
1029 */
1030 buf_len = PATH_MAX;
e780b0d1 1031 buf = kmalloc(buf_len, GFP_KERNEL);
31db9f7c
AB		 1032 if (!buf) {
1033 ret = -ENOMEM;
1034 goto out;
1035 }
1036
31db9f7c
AB		 1037 eb = path->nodes[0];
1038 slot = path->slots[0];
dd3cc16b 1039 item = btrfs_item_nr(slot);
31db9f7c
AB		 1040 di = btrfs_item_ptr(eb, slot, struct btrfs_dir_item);
1041 cur = 0;
1042 len = 0;
1043 total = btrfs_item_size(eb, item);
1044
1045 num = 0;
1046 while (cur < total) {
1047 name_len = btrfs_dir_name_len(eb, di);
1048 data_len = btrfs_dir_data_len(eb, di);
1049 type = btrfs_dir_type(eb, di);
1050 btrfs_dir_item_key_to_cpu(eb, di, &di_key);
1051
7e3ae33e
FM		 1052 if (type == BTRFS_FT_XATTR) {
1053 if (name_len > XATTR_NAME_MAX) {
1054 ret = -ENAMETOOLONG;
1055 goto out;
1056 }
4395e0c4 1057 if (name_len + data_len > BTRFS_MAX_XATTR_SIZE(root)) {
7e3ae33e
FM		 1058 ret = -E2BIG;
1059 goto out;
1060 }
1061 } else {
1062 /*
1063 * Path too long
1064 */
4395e0c4 1065 if (name_len + data_len > PATH_MAX) {
7e3ae33e
FM		 1066 ret = -ENAMETOOLONG;
1067 goto out;
1068 }
31db9f7c
AB		 1069 }
1070
4395e0c4
FM		 1071 if (name_len + data_len > buf_len) {
1072 buf_len = name_len + data_len;
1073 if (is_vmalloc_addr(buf)) {
1074 vfree(buf);
1075 buf = NULL;
1076 } else {
1077 char *tmp = krealloc(buf, buf_len,
e780b0d1 1078 GFP_KERNEL | __GFP_NOWARN);
4395e0c4
FM		 1079
1080 if (!tmp)
1081 kfree(buf);
1082 buf = tmp;
1083 }
1084 if (!buf) {
1085 buf = vmalloc(buf_len);
1086 if (!buf) {
1087 ret = -ENOMEM;
1088 goto out;
1089 }
1090 }
1091 }
1092
31db9f7c
AB		 1093 read_extent_buffer(eb, buf, (unsigned long)(di + 1),
1094 name_len + data_len);
1095
1096 len = sizeof(*di) + name_len + data_len;
1097 di = (struct btrfs_dir_item *)((char *)di + len);
1098 cur += len;
1099
1100 ret = iterate(num, &di_key, buf, name_len, buf + name_len,
1101 data_len, type, ctx);
1102 if (ret < 0)
1103 goto out;
1104 if (ret) {
1105 ret = 0;
1106 goto out;
1107 }
1108
1109 num++;
1110 }
1111
1112out:
4395e0c4 1113 kvfree(buf);
31db9f7c
AB		 1114 return ret;
1115}
1116
1117static int __copy_first_ref(int num, u64 dir, int index,
1118 struct fs_path *p, void *ctx)
1119{
1120 int ret;
1121 struct fs_path *pt = ctx;
1122
1123 ret = fs_path_copy(pt, p);
1124 if (ret < 0)
1125 return ret;
1126
1127 /* we want the first only */
1128 return 1;
1129}
1130
1131/*
1132 * Retrieve the first path of an inode. If an inode has more then one
1133 * ref/hardlink, this is ignored.
1134 */
924794c9 1135static int get_inode_path(struct btrfs_root *root,
31db9f7c
AB		 1136 u64 ino, struct fs_path *path)
1137{
1138 int ret;
1139 struct btrfs_key key, found_key;
1140 struct btrfs_path *p;
1141
1142 p = alloc_path_for_send();
1143 if (!p)
1144 return -ENOMEM;
1145
1146 fs_path_reset(path);
1147
1148 key.objectid = ino;
1149 key.type = BTRFS_INODE_REF_KEY;
1150 key.offset = 0;
1151
1152 ret = btrfs_search_slot_for_read(root, &key, p, 1, 0);
1153 if (ret < 0)
1154 goto out;
1155 if (ret) {
1156 ret = 1;
1157 goto out;
1158 }
1159 btrfs_item_key_to_cpu(p->nodes[0], &found_key, p->slots[0]);
1160 if (found_key.objectid != ino ||
96b5bd77
JS		 1161 (found_key.type != BTRFS_INODE_REF_KEY &&
1162 found_key.type != BTRFS_INODE_EXTREF_KEY)) {
31db9f7c
AB		 1163 ret = -ENOENT;
1164 goto out;
1165 }
1166
924794c9
TI		 1167 ret = iterate_inode_ref(root, p, &found_key, 1,
1168 __copy_first_ref, path);
31db9f7c
AB		 1169 if (ret < 0)
1170 goto out;
1171 ret = 0;
1172
1173out:
1174 btrfs_free_path(p);
1175 return ret;
1176}
1177
1178struct backref_ctx {
1179 struct send_ctx *sctx;
1180
3f8a18cc 1181 struct btrfs_path *path;
31db9f7c
AB		 1182 /* number of total found references */
1183 u64 found;
1184
1185 /*
1186 * used for clones found in send_root. clones found behind cur_objectid
1187 * and cur_offset are not considered as allowed clones.
1188 */
1189 u64 cur_objectid;
1190 u64 cur_offset;
1191
1192 /* may be truncated in case it's the last extent in a file */
1193 u64 extent_len;
1194
619d8c4e
FM		 1195 /* data offset in the file extent item */
1196 u64 data_offset;
1197
31db9f7c 1198 /* Just to check for bugs in backref resolving */
ee849c04 1199 int found_itself;
31db9f7c
AB		 1200};
1201
1202static int __clone_root_cmp_bsearch(const void *key, const void *elt)
1203{
995e01b7 1204 u64 root = (u64)(uintptr_t)key;
31db9f7c
AB		 1205 struct clone_root *cr = (struct clone_root *)elt;
1206
1207 if (root < cr->root->objectid)
1208 return -1;
1209 if (root > cr->root->objectid)
1210 return 1;
1211 return 0;
1212}
1213
1214static int __clone_root_cmp_sort(const void *e1, const void *e2)
1215{
1216 struct clone_root *cr1 = (struct clone_root *)e1;
1217 struct clone_root *cr2 = (struct clone_root *)e2;
1218
1219 if (cr1->root->objectid < cr2->root->objectid)
1220 return -1;
1221 if (cr1->root->objectid > cr2->root->objectid)
1222 return 1;
1223 return 0;
1224}
1225
1226/*
1227 * Called for every backref that is found for the current extent.
766702ef 1228 * Results are collected in sctx->clone_roots->ino/offset/found_refs
31db9f7c
AB		 1229 */
1230static int __iterate_backrefs(u64 ino, u64 offset, u64 root, void *ctx_)
1231{
1232 struct backref_ctx *bctx = ctx_;
1233 struct clone_root *found;
1234 int ret;
1235 u64 i_size;
1236
1237 /* First check if the root is in the list of accepted clone sources */
995e01b7 1238 found = bsearch((void *)(uintptr_t)root, bctx->sctx->clone_roots,
31db9f7c
AB		 1239 bctx->sctx->clone_roots_cnt,
1240 sizeof(struct clone_root),
1241 __clone_root_cmp_bsearch);
1242 if (!found)
1243 return 0;
1244
1245 if (found->root == bctx->sctx->send_root &&
1246 ino == bctx->cur_objectid &&
1247 offset == bctx->cur_offset) {
ee849c04 1248 bctx->found_itself = 1;
31db9f7c
AB		 1249 }
1250
1251 /*
766702ef 1252 * There are inodes that have extents that lie behind its i_size. Don't
31db9f7c
AB		 1253 * accept clones from these extents.
1254 */
3f8a18cc
JB		 1255 ret = __get_inode_info(found->root, bctx->path, ino, &i_size, NULL, NULL,
1256 NULL, NULL, NULL);
1257 btrfs_release_path(bctx->path);
31db9f7c
AB		 1258 if (ret < 0)
1259 return ret;
1260
619d8c4e 1261 if (offset + bctx->data_offset + bctx->extent_len > i_size)
31db9f7c
AB		 1262 return 0;
1263
1264 /*
1265 * Make sure we don't consider clones from send_root that are
1266 * behind the current inode/offset.
1267 */
1268 if (found->root == bctx->sctx->send_root) {
1269 /*
1270 * TODO for the moment we don't accept clones from the inode
1271 * that is currently send. We may change this when
1272 * BTRFS_IOC_CLONE_RANGE supports cloning from and to the same
1273 * file.
1274 */
1275 if (ino >= bctx->cur_objectid)
1276 return 0;
e938c8ad
AB		 1277#if 0
1278 if (ino > bctx->cur_objectid)
1279 return 0;
1280 if (offset + bctx->extent_len > bctx->cur_offset)
31db9f7c 1281 return 0;
e938c8ad 1282#endif
31db9f7c
AB		 1283 }
1284
1285 bctx->found++;
1286 found->found_refs++;
1287 if (ino < found->ino) {
1288 found->ino = ino;
1289 found->offset = offset;
1290 } else if (found->ino == ino) {
1291 /*
1292 * same extent found more then once in the same file.
1293 */
1294 if (found->offset > offset + bctx->extent_len)
1295 found->offset = offset;
1296 }
1297
1298 return 0;
1299}
1300
1301/*
766702ef
AB		 1302 * Given an inode, offset and extent item, it finds a good clone for a clone
1303 * instruction. Returns -ENOENT when none could be found. The function makes
1304 * sure that the returned clone is usable at the point where sending is at the
1305 * moment. This means, that no clones are accepted which lie behind the current
1306 * inode+offset.
1307 *
31db9f7c
AB		 1308 * path must point to the extent item when called.
1309 */
1310static int find_extent_clone(struct send_ctx *sctx,
1311 struct btrfs_path *path,
1312 u64 ino, u64 data_offset,
1313 u64 ino_size,
1314 struct clone_root **found)
1315{
04ab956e 1316 struct btrfs_fs_info *fs_info = sctx->send_root->fs_info;
31db9f7c
AB		 1317 int ret;
1318 int extent_type;
1319 u64 logical;
74dd17fb 1320 u64 disk_byte;
31db9f7c
AB		 1321 u64 num_bytes;
1322 u64 extent_item_pos;
69917e43 1323 u64 flags = 0;
31db9f7c
AB		 1324 struct btrfs_file_extent_item *fi;
1325 struct extent_buffer *eb = path->nodes[0];
35075bb0 1326 struct backref_ctx *backref_ctx = NULL;
31db9f7c
AB		 1327 struct clone_root *cur_clone_root;
1328 struct btrfs_key found_key;
1329 struct btrfs_path *tmp_path;
74dd17fb 1330 int compressed;
31db9f7c
AB		 1331 u32 i;
1332
1333 tmp_path = alloc_path_for_send();
1334 if (!tmp_path)
1335 return -ENOMEM;
1336
3f8a18cc
JB		 1337 /* We only use this path under the commit sem */
1338 tmp_path->need_commit_sem = 0;
1339
e780b0d1 1340 backref_ctx = kmalloc(sizeof(*backref_ctx), GFP_KERNEL);
35075bb0
AB		 1341 if (!backref_ctx) {
1342 ret = -ENOMEM;
1343 goto out;
1344 }
1345
3f8a18cc
JB		 1346 backref_ctx->path = tmp_path;
1347
31db9f7c
AB		 1348 if (data_offset >= ino_size) {
1349 /*
1350 * There may be extents that lie behind the file's size.
1351 * I at least had this in combination with snapshotting while
1352 * writing large files.
1353 */
1354 ret = 0;
1355 goto out;
1356 }
1357
1358 fi = btrfs_item_ptr(eb, path->slots[0],
1359 struct btrfs_file_extent_item);
1360 extent_type = btrfs_file_extent_type(eb, fi);
1361 if (extent_type == BTRFS_FILE_EXTENT_INLINE) {
1362 ret = -ENOENT;
1363 goto out;
1364 }
74dd17fb 1365 compressed = btrfs_file_extent_compression(eb, fi);
31db9f7c
AB		 1366
1367 num_bytes = btrfs_file_extent_num_bytes(eb, fi);
74dd17fb
CM		 1368 disk_byte = btrfs_file_extent_disk_bytenr(eb, fi);
1369 if (disk_byte == 0) {
31db9f7c
AB		 1370 ret = -ENOENT;
1371 goto out;
1372 }
74dd17fb 1373 logical = disk_byte + btrfs_file_extent_offset(eb, fi);
31db9f7c 1374
04ab956e
JM		 1375 down_read(&fs_info->commit_root_sem);
1376 ret = extent_from_logical(fs_info, disk_byte, tmp_path,
69917e43 1377 &found_key, &flags);
04ab956e 1378 up_read(&fs_info->commit_root_sem);
31db9f7c
AB		 1379 btrfs_release_path(tmp_path);
1380
1381 if (ret < 0)
1382 goto out;
69917e43 1383 if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) {
31db9f7c
AB		 1384 ret = -EIO;
1385 goto out;
1386 }
1387
1388 /*
1389 * Setup the clone roots.
1390 */
1391 for (i = 0; i < sctx->clone_roots_cnt; i++) {
1392 cur_clone_root = sctx->clone_roots + i;
1393 cur_clone_root->ino = (u64)-1;
1394 cur_clone_root->offset = 0;
1395 cur_clone_root->found_refs = 0;
1396 }
1397
35075bb0
AB		 1398 backref_ctx->sctx = sctx;
1399 backref_ctx->found = 0;
1400 backref_ctx->cur_objectid = ino;
1401 backref_ctx->cur_offset = data_offset;
1402 backref_ctx->found_itself = 0;
1403 backref_ctx->extent_len = num_bytes;
619d8c4e
FM		 1404 /*
1405 * For non-compressed extents iterate_extent_inodes() gives us extent
1406 * offsets that already take into account the data offset, but not for
1407 * compressed extents, since the offset is logical and not relative to
1408 * the physical extent locations. We must take this into account to
1409 * avoid sending clone offsets that go beyond the source file's size,
1410 * which would result in the clone ioctl failing with -EINVAL on the
1411 * receiving end.
1412 */
1413 if (compressed == BTRFS_COMPRESS_NONE)
1414 backref_ctx->data_offset = 0;
1415 else
1416 backref_ctx->data_offset = btrfs_file_extent_offset(eb, fi);
31db9f7c
AB		 1417
1418 /*
1419 * The last extent of a file may be too large due to page alignment.
1420 * We need to adjust extent_len in this case so that the checks in
1421 * __iterate_backrefs work.
1422 */
1423 if (data_offset + num_bytes >= ino_size)
35075bb0 1424 backref_ctx->extent_len = ino_size - data_offset;
31db9f7c
AB		 1425
1426 /*
1427 * Now collect all backrefs.
1428 */
74dd17fb
CM		 1429 if (compressed == BTRFS_COMPRESS_NONE)
1430 extent_item_pos = logical - found_key.objectid;
1431 else
1432 extent_item_pos = 0;
04ab956e 1433 ret = iterate_extent_inodes(fs_info,
31db9f7c 1434 found_key.objectid, extent_item_pos, 1,
35075bb0 1435 __iterate_backrefs, backref_ctx);
74dd17fb 1436
31db9f7c
AB		 1437 if (ret < 0)
1438 goto out;
1439
35075bb0 1440 if (!backref_ctx->found_itself) {
31db9f7c
AB		 1441 /* found a bug in backref code? */
1442 ret = -EIO;
04ab956e 1443 btrfs_err(fs_info,
5d163e0e 1444 "did not find backref in send_root. inode=%llu, offset=%llu, disk_byte=%llu found extent=%llu",
04ab956e 1445 ino, data_offset, disk_byte, found_key.objectid);
31db9f7c
AB		 1446 goto out;
1447 }
1448
04ab956e
JM		 1449 btrfs_debug(fs_info,
1450 "find_extent_clone: data_offset=%llu, ino=%llu, num_bytes=%llu, logical=%llu",
1451 data_offset, ino, num_bytes, logical);
31db9f7c 1452
35075bb0 1453 if (!backref_ctx->found)
04ab956e 1454 btrfs_debug(fs_info, "no clones found");
31db9f7c
AB		 1455
1456 cur_clone_root = NULL;
1457 for (i = 0; i < sctx->clone_roots_cnt; i++) {
1458 if (sctx->clone_roots[i].found_refs) {
1459 if (!cur_clone_root)
1460 cur_clone_root = sctx->clone_roots + i;
1461 else if (sctx->clone_roots[i].root == sctx->send_root)
1462 /* prefer clones from send_root over others */
1463 cur_clone_root = sctx->clone_roots + i;
31db9f7c
AB		 1464 }
1465
1466 }
1467
1468 if (cur_clone_root) {
1469 *found = cur_clone_root;
1470 ret = 0;
1471 } else {
1472 ret = -ENOENT;
1473 }
1474
1475out:
1476 btrfs_free_path(tmp_path);
35075bb0 1477 kfree(backref_ctx);
31db9f7c
AB		 1478 return ret;
1479}
1480
924794c9 1481static int read_symlink(struct btrfs_root *root,
31db9f7c
AB		 1482 u64 ino,
1483 struct fs_path *dest)
1484{
1485 int ret;
1486 struct btrfs_path *path;
1487 struct btrfs_key key;
1488 struct btrfs_file_extent_item *ei;
1489 u8 type;
1490 u8 compression;
1491 unsigned long off;
1492 int len;
1493
1494 path = alloc_path_for_send();
1495 if (!path)
1496 return -ENOMEM;
1497
1498 key.objectid = ino;
1499 key.type = BTRFS_EXTENT_DATA_KEY;
1500 key.offset = 0;
1501 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
1502 if (ret < 0)
1503 goto out;
a879719b
FM		 1504 if (ret) {
1505 /*
1506 * An empty symlink inode. Can happen in rare error paths when
1507 * creating a symlink (transaction committed before the inode
1508 * eviction handler removed the symlink inode items and a crash
1509 * happened in between or the subvol was snapshoted in between).
1510 * Print an informative message to dmesg/syslog so that the user
1511 * can delete the symlink.
1512 */
1513 btrfs_err(root->fs_info,
1514 "Found empty symlink inode %llu at root %llu",
1515 ino, root->root_key.objectid);
1516 ret = -EIO;
1517 goto out;
1518 }
31db9f7c
AB		 1519
1520 ei = btrfs_item_ptr(path->nodes[0], path->slots[0],
1521 struct btrfs_file_extent_item);
1522 type = btrfs_file_extent_type(path->nodes[0], ei);
1523 compression = btrfs_file_extent_compression(path->nodes[0], ei);
1524 BUG_ON(type != BTRFS_FILE_EXTENT_INLINE);
1525 BUG_ON(compression);
1526
1527 off = btrfs_file_extent_inline_start(ei);
514ac8ad 1528 len = btrfs_file_extent_inline_len(path->nodes[0], path->slots[0], ei);
31db9f7c
AB		 1529
1530 ret = fs_path_add_from_extent_buffer(dest, path->nodes[0], off, len);
31db9f7c
AB		 1531
1532out:
1533 btrfs_free_path(path);
1534 return ret;
1535}
1536
1537/*
1538 * Helper function to generate a file name that is unique in the root of
1539 * send_root and parent_root. This is used to generate names for orphan inodes.
1540 */
1541static int gen_unique_name(struct send_ctx *sctx,
1542 u64 ino, u64 gen,
1543 struct fs_path *dest)
1544{
1545 int ret = 0;
1546 struct btrfs_path *path;
1547 struct btrfs_dir_item *di;
1548 char tmp[64];
1549 int len;
1550 u64 idx = 0;
1551
1552 path = alloc_path_for_send();
1553 if (!path)
1554 return -ENOMEM;
1555
1556 while (1) {
f74b86d8 1557 len = snprintf(tmp, sizeof(tmp), "o%llu-%llu-%llu",
31db9f7c 1558 ino, gen, idx);
64792f25 1559 ASSERT(len < sizeof(tmp));
31db9f7c
AB		 1560
1561 di = btrfs_lookup_dir_item(NULL, sctx->send_root,
1562 path, BTRFS_FIRST_FREE_OBJECTID,
1563 tmp, strlen(tmp), 0);
1564 btrfs_release_path(path);
1565 if (IS_ERR(di)) {
1566 ret = PTR_ERR(di);
1567 goto out;
1568 }
1569 if (di) {
1570 /* not unique, try again */
1571 idx++;
1572 continue;
1573 }
1574
1575 if (!sctx->parent_root) {
1576 /* unique */
1577 ret = 0;
1578 break;
1579 }
1580
1581 di = btrfs_lookup_dir_item(NULL, sctx->parent_root,
1582 path, BTRFS_FIRST_FREE_OBJECTID,
1583 tmp, strlen(tmp), 0);
1584 btrfs_release_path(path);
1585 if (IS_ERR(di)) {
1586 ret = PTR_ERR(di);
1587 goto out;
1588 }
1589 if (di) {
1590 /* not unique, try again */
1591 idx++;
1592 continue;
1593 }
1594 /* unique */
1595 break;
1596 }
1597
1598 ret = fs_path_add(dest, tmp, strlen(tmp));
1599
1600out:
1601 btrfs_free_path(path);
1602 return ret;
1603}
1604
1605enum inode_state {
1606 inode_state_no_change,
1607 inode_state_will_create,
1608 inode_state_did_create,
1609 inode_state_will_delete,
1610 inode_state_did_delete,
1611};
1612
1613static int get_cur_inode_state(struct send_ctx *sctx, u64 ino, u64 gen)
1614{
1615 int ret;
1616 int left_ret;
1617 int right_ret;
1618 u64 left_gen;
1619 u64 right_gen;
1620
1621 ret = get_inode_info(sctx->send_root, ino, NULL, &left_gen, NULL, NULL,
85a7b33b 1622 NULL, NULL);
31db9f7c
AB		 1623 if (ret < 0 && ret != -ENOENT)
1624 goto out;
1625 left_ret = ret;
1626
1627 if (!sctx->parent_root) {
1628 right_ret = -ENOENT;
1629 } else {
1630 ret = get_inode_info(sctx->parent_root, ino, NULL, &right_gen,
85a7b33b 1631 NULL, NULL, NULL, NULL);
31db9f7c
AB		 1632 if (ret < 0 && ret != -ENOENT)
1633 goto out;
1634 right_ret = ret;
1635 }
1636
1637 if (!left_ret && !right_ret) {
e938c8ad 1638 if (left_gen == gen && right_gen == gen) {
31db9f7c 1639 ret = inode_state_no_change;
e938c8ad 1640 } else if (left_gen == gen) {
31db9f7c
AB		 1641 if (ino < sctx->send_progress)
1642 ret = inode_state_did_create;
1643 else
1644 ret = inode_state_will_create;
1645 } else if (right_gen == gen) {
1646 if (ino < sctx->send_progress)
1647 ret = inode_state_did_delete;
1648 else
1649 ret = inode_state_will_delete;
1650 } else {
1651 ret = -ENOENT;
1652 }
1653 } else if (!left_ret) {
1654 if (left_gen == gen) {
1655 if (ino < sctx->send_progress)
1656 ret = inode_state_did_create;
1657 else
1658 ret = inode_state_will_create;
1659 } else {
1660 ret = -ENOENT;
1661 }
1662 } else if (!right_ret) {
1663 if (right_gen == gen) {
1664 if (ino < sctx->send_progress)
1665 ret = inode_state_did_delete;
1666 else
1667 ret = inode_state_will_delete;
1668 } else {
1669 ret = -ENOENT;
1670 }
1671 } else {
1672 ret = -ENOENT;
1673 }
1674
1675out:
1676 return ret;
1677}
1678
1679static int is_inode_existent(struct send_ctx *sctx, u64 ino, u64 gen)
1680{
1681 int ret;
1682
1683 ret = get_cur_inode_state(sctx, ino, gen);
1684 if (ret < 0)
1685 goto out;
1686
1687 if (ret == inode_state_no_change ||
1688 ret == inode_state_did_create ||
1689 ret == inode_state_will_delete)
1690 ret = 1;
1691 else
1692 ret = 0;
1693
1694out:
1695 return ret;
1696}
1697
1698/*
1699 * Helper function to lookup a dir item in a dir.
1700 */
1701static int lookup_dir_item_inode(struct btrfs_root *root,
1702 u64 dir, const char *name, int name_len,
1703 u64 *found_inode,
1704 u8 *found_type)
1705{
1706 int ret = 0;
1707 struct btrfs_dir_item *di;
1708 struct btrfs_key key;
1709 struct btrfs_path *path;
1710
1711 path = alloc_path_for_send();
1712 if (!path)
1713 return -ENOMEM;
1714
1715 di = btrfs_lookup_dir_item(NULL, root, path,
1716 dir, name, name_len, 0);
1717 if (!di) {
1718 ret = -ENOENT;
1719 goto out;
1720 }
1721 if (IS_ERR(di)) {
1722 ret = PTR_ERR(di);
1723 goto out;
1724 }
1725 btrfs_dir_item_key_to_cpu(path->nodes[0], di, &key);
1af56070
FM		 1726 if (key.type == BTRFS_ROOT_ITEM_KEY) {
1727 ret = -ENOENT;
1728 goto out;
1729 }
31db9f7c
AB		 1730 *found_inode = key.objectid;
1731 *found_type = btrfs_dir_type(path->nodes[0], di);
1732
1733out:
1734 btrfs_free_path(path);
1735 return ret;
1736}
1737
766702ef
AB		 1738/*
1739 * Looks up the first btrfs_inode_ref of a given ino. It returns the parent dir,
1740 * generation of the parent dir and the name of the dir entry.
1741 */
924794c9 1742static int get_first_ref(struct btrfs_root *root, u64 ino,
31db9f7c
AB		 1743 u64 *dir, u64 *dir_gen, struct fs_path *name)
1744{
1745 int ret;
1746 struct btrfs_key key;
1747 struct btrfs_key found_key;
1748 struct btrfs_path *path;
31db9f7c 1749 int len;
96b5bd77 1750 u64 parent_dir;
31db9f7c
AB		 1751
1752 path = alloc_path_for_send();
1753 if (!path)
1754 return -ENOMEM;
1755
1756 key.objectid = ino;
1757 key.type = BTRFS_INODE_REF_KEY;
1758 key.offset = 0;
1759
1760 ret = btrfs_search_slot_for_read(root, &key, path, 1, 0);
1761 if (ret < 0)
1762 goto out;
1763 if (!ret)
1764 btrfs_item_key_to_cpu(path->nodes[0], &found_key,
1765 path->slots[0]);
96b5bd77
JS		 1766 if (ret || found_key.objectid != ino ||
1767 (found_key.type != BTRFS_INODE_REF_KEY &&
1768 found_key.type != BTRFS_INODE_EXTREF_KEY)) {
31db9f7c
AB		 1769 ret = -ENOENT;
1770 goto out;
1771 }
1772
51a60253 1773 if (found_key.type == BTRFS_INODE_REF_KEY) {
96b5bd77
JS		 1774 struct btrfs_inode_ref *iref;
1775 iref = btrfs_item_ptr(path->nodes[0], path->slots[0],
1776 struct btrfs_inode_ref);
1777 len = btrfs_inode_ref_name_len(path->nodes[0], iref);
1778 ret = fs_path_add_from_extent_buffer(name, path->nodes[0],
1779 (unsigned long)(iref + 1),
1780 len);
1781 parent_dir = found_key.offset;
1782 } else {
1783 struct btrfs_inode_extref *extref;
1784 extref = btrfs_item_ptr(path->nodes[0], path->slots[0],
1785 struct btrfs_inode_extref);
1786 len = btrfs_inode_extref_name_len(path->nodes[0], extref);
1787 ret = fs_path_add_from_extent_buffer(name, path->nodes[0],
1788 (unsigned long)&extref->name, len);
1789 parent_dir = btrfs_inode_extref_parent(path->nodes[0], extref);
1790 }
31db9f7c
AB		 1791 if (ret < 0)
1792 goto out;
1793 btrfs_release_path(path);
1794
b46ab97b
FM		 1795 if (dir_gen) {
1796 ret = get_inode_info(root, parent_dir, NULL, dir_gen, NULL,
1797 NULL, NULL, NULL);
1798 if (ret < 0)
1799 goto out;
1800 }
31db9f7c 1801
96b5bd77 1802 *dir = parent_dir;
31db9f7c
AB		 1803
1804out:
1805 btrfs_free_path(path);
1806 return ret;
1807}
1808
924794c9 1809static int is_first_ref(struct btrfs_root *root,
31db9f7c
AB		 1810 u64 ino, u64 dir,
1811 const char *name, int name_len)
1812{
1813 int ret;
1814 struct fs_path *tmp_name;
1815 u64 tmp_dir;
31db9f7c 1816
924794c9 1817 tmp_name = fs_path_alloc();
31db9f7c
AB		 1818 if (!tmp_name)
1819 return -ENOMEM;
1820
b46ab97b 1821 ret = get_first_ref(root, ino, &tmp_dir, NULL, tmp_name);
31db9f7c
AB		 1822 if (ret < 0)
1823 goto out;
1824
b9291aff 1825 if (dir != tmp_dir || name_len != fs_path_len(tmp_name)) {
31db9f7c
AB		 1826 ret = 0;
1827 goto out;
1828 }
1829
e938c8ad 1830 ret = !memcmp(tmp_name->start, name, name_len);
31db9f7c
AB		 1831
1832out:
924794c9 1833 fs_path_free(tmp_name);
31db9f7c
AB		 1834 return ret;
1835}
1836
766702ef
AB		 1837/*
1838 * Used by process_recorded_refs to determine if a new ref would overwrite an
1839 * already existing ref. In case it detects an overwrite, it returns the
1840 * inode/gen in who_ino/who_gen.
1841 * When an overwrite is detected, process_recorded_refs does proper orphanizing
1842 * to make sure later references to the overwritten inode are possible.
1843 * Orphanizing is however only required for the first ref of an inode.
1844 * process_recorded_refs does an additional is_first_ref check to see if
1845 * orphanizing is really required.
1846 */
31db9f7c
AB		 1847static int will_overwrite_ref(struct send_ctx *sctx, u64 dir, u64 dir_gen,
1848 const char *name, int name_len,
1849 u64 *who_ino, u64 *who_gen)
1850{
1851 int ret = 0;
ebdad913 1852 u64 gen;
31db9f7c
AB		 1853 u64 other_inode = 0;
1854 u8 other_type = 0;
1855
1856 if (!sctx->parent_root)
1857 goto out;
1858
1859 ret = is_inode_existent(sctx, dir, dir_gen);
1860 if (ret <= 0)
1861 goto out;
1862
ebdad913
JB		 1863 /*
1864 * If we have a parent root we need to verify that the parent dir was
01327610 1865 * not deleted and then re-created, if it was then we have no overwrite
ebdad913
JB		 1866 * and we can just unlink this entry.
1867 */
1868 if (sctx->parent_root) {
1869 ret = get_inode_info(sctx->parent_root, dir, NULL, &gen, NULL,
1870 NULL, NULL, NULL);
1871 if (ret < 0 && ret != -ENOENT)
1872 goto out;
1873 if (ret) {
1874 ret = 0;
1875 goto out;
1876 }
1877 if (gen != dir_gen)
1878 goto out;
1879 }
1880
31db9f7c
AB		 1881 ret = lookup_dir_item_inode(sctx->parent_root, dir, name, name_len,
1882 &other_inode, &other_type);
1883 if (ret < 0 && ret != -ENOENT)
1884 goto out;
1885 if (ret) {
1886 ret = 0;
1887 goto out;
1888 }
1889
766702ef
AB		 1890 /*
1891 * Check if the overwritten ref was already processed. If yes, the ref
1892 * was already unlinked/moved, so we can safely assume that we will not
1893 * overwrite anything at this point in time.
1894 */
801bec36
RK		 1895 if (other_inode > sctx->send_progress ||
1896 is_waiting_for_move(sctx, other_inode)) {
31db9f7c 1897 ret = get_inode_info(sctx->parent_root, other_inode, NULL,
85a7b33b 1898 who_gen, NULL, NULL, NULL, NULL);
31db9f7c
AB		 1899 if (ret < 0)
1900 goto out;
1901
1902 ret = 1;
1903 *who_ino = other_inode;
1904 } else {
1905 ret = 0;
1906 }
1907
1908out:
1909 return ret;
1910}
1911
766702ef
AB		 1912/*
1913 * Checks if the ref was overwritten by an already processed inode. This is
1914 * used by __get_cur_name_and_parent to find out if the ref was orphanized and
1915 * thus the orphan name needs be used.
1916 * process_recorded_refs also uses it to avoid unlinking of refs that were
1917 * overwritten.
1918 */
31db9f7c
AB		 1919static int did_overwrite_ref(struct send_ctx *sctx,
1920 u64 dir, u64 dir_gen,
1921 u64 ino, u64 ino_gen,
1922 const char *name, int name_len)
1923{
1924 int ret = 0;
1925 u64 gen;
1926 u64 ow_inode;
1927 u8 other_type;
1928
1929 if (!sctx->parent_root)
1930 goto out;
1931
1932 ret = is_inode_existent(sctx, dir, dir_gen);
1933 if (ret <= 0)
1934 goto out;
1935
1936 /* check if the ref was overwritten by another ref */
1937 ret = lookup_dir_item_inode(sctx->send_root, dir, name, name_len,
1938 &ow_inode, &other_type);
1939 if (ret < 0 && ret != -ENOENT)
1940 goto out;
1941 if (ret) {
1942 /* was never and will never be overwritten */
1943 ret = 0;
1944 goto out;
1945 }
1946
1947 ret = get_inode_info(sctx->send_root, ow_inode, NULL, &gen, NULL, NULL,
85a7b33b 1948 NULL, NULL);
31db9f7c
AB		 1949 if (ret < 0)
1950 goto out;
1951
1952 if (ow_inode == ino && gen == ino_gen) {
1953 ret = 0;
1954 goto out;
1955 }
1956
8b191a68
FM		 1957 /*
1958 * We know that it is or will be overwritten. Check this now.
1959 * The current inode being processed might have been the one that caused
b786f16a
FM		 1960 * inode 'ino' to be orphanized, therefore check if ow_inode matches
1961 * the current inode being processed.
8b191a68 1962 */
b786f16a
FM		 1963 if ((ow_inode < sctx->send_progress) ||
1964 (ino != sctx->cur_ino && ow_inode == sctx->cur_ino &&
1965 gen == sctx->cur_inode_gen))
31db9f7c
AB		 1966 ret = 1;
1967 else
1968 ret = 0;
1969
1970out:
1971 return ret;
1972}
1973
766702ef
AB		 1974/*
1975 * Same as did_overwrite_ref, but also checks if it is the first ref of an inode
1976 * that got overwritten. This is used by process_recorded_refs to determine
1977 * if it has to use the path as returned by get_cur_path or the orphan name.
1978 */
31db9f7c
AB		 1979static int did_overwrite_first_ref(struct send_ctx *sctx, u64 ino, u64 gen)
1980{
1981 int ret = 0;
1982 struct fs_path *name = NULL;
1983 u64 dir;
1984 u64 dir_gen;
1985
1986 if (!sctx->parent_root)
1987 goto out;
1988
924794c9 1989 name = fs_path_alloc();
31db9f7c
AB		 1990 if (!name)
1991 return -ENOMEM;
1992
924794c9 1993 ret = get_first_ref(sctx->parent_root, ino, &dir, &dir_gen, name);
31db9f7c
AB		 1994 if (ret < 0)
1995 goto out;
1996
1997 ret = did_overwrite_ref(sctx, dir, dir_gen, ino, gen,
1998 name->start, fs_path_len(name));
31db9f7c
AB		 1999
2000out:
924794c9 2001 fs_path_free(name);
31db9f7c
AB		 2002 return ret;
2003}
2004
766702ef
AB		 2005/*
2006 * Insert a name cache entry. On 32bit kernels the radix tree index is 32bit,
2007 * so we need to do some special handling in case we have clashes. This function
2008 * takes care of this with the help of name_cache_entry::radix_list.
5dc67d0b 2009 * In case of error, nce is kfreed.
766702ef 2010 */
31db9f7c
AB		 2011static int name_cache_insert(struct send_ctx *sctx,
2012 struct name_cache_entry *nce)
2013{
2014 int ret = 0;
7e0926fe
AB		 2015 struct list_head *nce_head;
2016
2017 nce_head = radix_tree_lookup(&sctx->name_cache,
2018 (unsigned long)nce->ino);
2019 if (!nce_head) {
e780b0d1 2020 nce_head = kmalloc(sizeof(*nce_head), GFP_KERNEL);
cfa7a9cc
TI		 2021 if (!nce_head) {
2022 kfree(nce);
31db9f7c 2023 return -ENOMEM;
cfa7a9cc 2024 }
7e0926fe 2025 INIT_LIST_HEAD(nce_head);
31db9f7c 2026
7e0926fe 2027 ret = radix_tree_insert(&sctx->name_cache, nce->ino, nce_head);
5dc67d0b
AB		 2028 if (ret < 0) {
2029 kfree(nce_head);
2030 kfree(nce);
31db9f7c 2031 return ret;
5dc67d0b 2032 }
31db9f7c 2033 }
7e0926fe 2034 list_add_tail(&nce->radix_list, nce_head);
31db9f7c
AB		 2035 list_add_tail(&nce->list, &sctx->name_cache_list);
2036 sctx->name_cache_size++;
2037
2038 return ret;
2039}
2040
2041static void name_cache_delete(struct send_ctx *sctx,
2042 struct name_cache_entry *nce)
2043{
7e0926fe 2044 struct list_head *nce_head;
31db9f7c 2045
7e0926fe
AB		 2046 nce_head = radix_tree_lookup(&sctx->name_cache,
2047 (unsigned long)nce->ino);
57fb8910
DS		 2048 if (!nce_head) {
2049 btrfs_err(sctx->send_root->fs_info,
2050 "name_cache_delete lookup failed ino %llu cache size %d, leaking memory",
2051 nce->ino, sctx->name_cache_size);
2052 }
31db9f7c 2053
7e0926fe 2054 list_del(&nce->radix_list);
31db9f7c 2055 list_del(&nce->list);
31db9f7c 2056 sctx->name_cache_size--;
7e0926fe 2057
57fb8910
DS		 2058 /*
2059 * We may not get to the final release of nce_head if the lookup fails
2060 */
2061 if (nce_head && list_empty(nce_head)) {
7e0926fe
AB		 2062 radix_tree_delete(&sctx->name_cache, (unsigned long)nce->ino);
2063 kfree(nce_head);
2064 }
31db9f7c
AB		 2065}
2066
2067static struct name_cache_entry *name_cache_search(struct send_ctx *sctx,
2068 u64 ino, u64 gen)
2069{
7e0926fe
AB		 2070 struct list_head *nce_head;
2071 struct name_cache_entry *cur;
31db9f7c 2072
7e0926fe
AB		 2073 nce_head = radix_tree_lookup(&sctx->name_cache, (unsigned long)ino);
2074 if (!nce_head)
31db9f7c
AB		 2075 return NULL;
2076
7e0926fe
AB		 2077 list_for_each_entry(cur, nce_head, radix_list) {
2078 if (cur->ino == ino && cur->gen == gen)
2079 return cur;
2080 }
31db9f7c
AB		 2081 return NULL;
2082}
2083
766702ef
AB		 2084/*
2085 * Removes the entry from the list and adds it back to the end. This marks the
2086 * entry as recently used so that name_cache_clean_unused does not remove it.
2087 */
31db9f7c
AB		 2088static void name_cache_used(struct send_ctx *sctx, struct name_cache_entry *nce)
2089{
2090 list_del(&nce->list);
2091 list_add_tail(&nce->list, &sctx->name_cache_list);
2092}
2093
766702ef
AB		 2094/*
2095 * Remove some entries from the beginning of name_cache_list.
2096 */
31db9f7c
AB		 2097static void name_cache_clean_unused(struct send_ctx *sctx)
2098{
2099 struct name_cache_entry *nce;
2100
2101 if (sctx->name_cache_size < SEND_CTX_NAME_CACHE_CLEAN_SIZE)
2102 return;
2103
2104 while (sctx->name_cache_size > SEND_CTX_MAX_NAME_CACHE_SIZE) {
2105 nce = list_entry(sctx->name_cache_list.next,
2106 struct name_cache_entry, list);
2107 name_cache_delete(sctx, nce);
2108 kfree(nce);
2109 }
2110}
2111
2112static void name_cache_free(struct send_ctx *sctx)
2113{
2114 struct name_cache_entry *nce;
31db9f7c 2115
e938c8ad
AB		 2116 while (!list_empty(&sctx->name_cache_list)) {
2117 nce = list_entry(sctx->name_cache_list.next,
2118 struct name_cache_entry, list);
31db9f7c 2119 name_cache_delete(sctx, nce);
17589bd9 2120 kfree(nce);
31db9f7c
AB		 2121 }
2122}
2123
766702ef
AB		 2124/*
2125 * Used by get_cur_path for each ref up to the root.
2126 * Returns 0 if it succeeded.
2127 * Returns 1 if the inode is not existent or got overwritten. In that case, the
2128 * name is an orphan name. This instructs get_cur_path to stop iterating. If 1
2129 * is returned, parent_ino/parent_gen are not guaranteed to be valid.
2130 * Returns <0 in case of error.
2131 */
31db9f7c
AB		 2132static int __get_cur_name_and_parent(struct send_ctx *sctx,
2133 u64 ino, u64 gen,
2134 u64 *parent_ino,
2135 u64 *parent_gen,
2136 struct fs_path *dest)
2137{
2138 int ret;
2139 int nce_ret;
31db9f7c
AB		 2140 struct name_cache_entry *nce = NULL;
2141
766702ef
AB		 2142 /*
2143 * First check if we already did a call to this function with the same
2144 * ino/gen. If yes, check if the cache entry is still up-to-date. If yes
2145 * return the cached result.
2146 */
31db9f7c
AB		 2147 nce = name_cache_search(sctx, ino, gen);
2148 if (nce) {
2149 if (ino < sctx->send_progress && nce->need_later_update) {
2150 name_cache_delete(sctx, nce);
2151 kfree(nce);
2152 nce = NULL;
2153 } else {
2154 name_cache_used(sctx, nce);
2155 *parent_ino = nce->parent_ino;
2156 *parent_gen = nce->parent_gen;
2157 ret = fs_path_add(dest, nce->name, nce->name_len);
2158 if (ret < 0)
2159 goto out;
2160 ret = nce->ret;
2161 goto out;
2162 }
2163 }
2164
766702ef
AB		 2165 /*
2166 * If the inode is not existent yet, add the orphan name and return 1.
2167 * This should only happen for the parent dir that we determine in
2168 * __record_new_ref
2169 */
31db9f7c
AB		 2170 ret = is_inode_existent(sctx, ino, gen);
2171 if (ret < 0)
2172 goto out;
2173
2174 if (!ret) {
2175 ret = gen_unique_name(sctx, ino, gen, dest);
2176 if (ret < 0)
2177 goto out;
2178 ret = 1;
2179 goto out_cache;
2180 }
2181
766702ef
AB		 2182 /*
2183 * Depending on whether the inode was already processed or not, use
2184 * send_root or parent_root for ref lookup.
2185 */
bf0d1f44 2186 if (ino < sctx->send_progress)
924794c9
TI		 2187 ret = get_first_ref(sctx->send_root, ino,
2188 parent_ino, parent_gen, dest);
31db9f7c 2189 else
924794c9
TI		 2190 ret = get_first_ref(sctx->parent_root, ino,
2191 parent_ino, parent_gen, dest);
31db9f7c
AB		 2192 if (ret < 0)
2193 goto out;
2194
766702ef
AB		 2195 /*
2196 * Check if the ref was overwritten by an inode's ref that was processed
2197 * earlier. If yes, treat as orphan and return 1.
2198 */
31db9f7c
AB		 2199 ret = did_overwrite_ref(sctx, *parent_ino, *parent_gen, ino, gen,
2200 dest->start, dest->end - dest->start);
2201 if (ret < 0)
2202 goto out;
2203 if (ret) {
2204 fs_path_reset(dest);
2205 ret = gen_unique_name(sctx, ino, gen, dest);
2206 if (ret < 0)
2207 goto out;
2208 ret = 1;
2209 }
2210
2211out_cache:
766702ef
AB		 2212 /*
2213 * Store the result of the lookup in the name cache.
2214 */
e780b0d1 2215 nce = kmalloc(sizeof(*nce) + fs_path_len(dest) + 1, GFP_KERNEL);
31db9f7c
AB		 2216 if (!nce) {
2217 ret = -ENOMEM;
2218 goto out;
2219 }
2220
2221 nce->ino = ino;
2222 nce->gen = gen;
2223 nce->parent_ino = *parent_ino;
2224 nce->parent_gen = *parent_gen;
2225 nce->name_len = fs_path_len(dest);
2226 nce->ret = ret;
2227 strcpy(nce->name, dest->start);
31db9f7c
AB		 2228
2229 if (ino < sctx->send_progress)
2230 nce->need_later_update = 0;
2231 else
2232 nce->need_later_update = 1;
2233
2234 nce_ret = name_cache_insert(sctx, nce);
2235 if (nce_ret < 0)
2236 ret = nce_ret;
2237 name_cache_clean_unused(sctx);
2238
2239out:
31db9f7c
AB		 2240 return ret;
2241}
2242
2243/*
2244 * Magic happens here. This function returns the first ref to an inode as it
2245 * would look like while receiving the stream at this point in time.
2246 * We walk the path up to the root. For every inode in between, we check if it
2247 * was already processed/sent. If yes, we continue with the parent as found
2248 * in send_root. If not, we continue with the parent as found in parent_root.
2249 * If we encounter an inode that was deleted at this point in time, we use the
2250 * inodes "orphan" name instead of the real name and stop. Same with new inodes
2251 * that were not created yet and overwritten inodes/refs.
2252 *
2253 * When do we have have orphan inodes:
2254 * 1. When an inode is freshly created and thus no valid refs are available yet
2255 * 2. When a directory lost all it's refs (deleted) but still has dir items
2256 * inside which were not processed yet (pending for move/delete). If anyone
2257 * tried to get the path to the dir items, it would get a path inside that
2258 * orphan directory.
2259 * 3. When an inode is moved around or gets new links, it may overwrite the ref
2260 * of an unprocessed inode. If in that case the first ref would be
2261 * overwritten, the overwritten inode gets "orphanized". Later when we
2262 * process this overwritten inode, it is restored at a new place by moving
2263 * the orphan inode.
2264 *
2265 * sctx->send_progress tells this function at which point in time receiving
2266 * would be.
2267 */
2268static int get_cur_path(struct send_ctx *sctx, u64 ino, u64 gen,
2269 struct fs_path *dest)
2270{
2271 int ret = 0;
2272 struct fs_path *name = NULL;
2273 u64 parent_inode = 0;
2274 u64 parent_gen = 0;
2275 int stop = 0;
2276
924794c9 2277 name = fs_path_alloc();
31db9f7c
AB		 2278 if (!name) {
2279 ret = -ENOMEM;
2280 goto out;
2281 }
2282
2283 dest->reversed = 1;
2284 fs_path_reset(dest);
2285
2286 while (!stop && ino != BTRFS_FIRST_FREE_OBJECTID) {
8b191a68
FM		 2287 struct waiting_dir_move *wdm;
2288
31db9f7c
AB		 2289 fs_path_reset(name);
2290
9dc44214
FM		 2291 if (is_waiting_for_rm(sctx, ino)) {
2292 ret = gen_unique_name(sctx, ino, gen, name);
2293 if (ret < 0)
2294 goto out;
2295 ret = fs_path_add_path(dest, name);
2296 break;
2297 }
2298
8b191a68
FM		 2299 wdm = get_waiting_dir_move(sctx, ino);
2300 if (wdm && wdm->orphanized) {
2301 ret = gen_unique_name(sctx, ino, gen, name);
2302 stop = 1;
2303 } else if (wdm) {
bf0d1f44
FM		 2304 ret = get_first_ref(sctx->parent_root, ino,
2305 &parent_inode, &parent_gen, name);
2306 } else {
2307 ret = __get_cur_name_and_parent(sctx, ino, gen,
2308 &parent_inode,
2309 &parent_gen, name);
2310 if (ret)
2311 stop = 1;
2312 }
2313
31db9f7c
AB		 2314 if (ret < 0)
2315 goto out;
9f03740a 2316
31db9f7c
AB		 2317 ret = fs_path_add_path(dest, name);
2318 if (ret < 0)
2319 goto out;
2320
2321 ino = parent_inode;
2322 gen = parent_gen;
2323 }
2324
2325out:
924794c9 2326 fs_path_free(name);
31db9f7c
AB		 2327 if (!ret)
2328 fs_path_unreverse(dest);
2329 return ret;
2330}
2331
31db9f7c
AB		 2332/*
2333 * Sends a BTRFS_SEND_C_SUBVOL command/item to userspace
2334 */
2335static int send_subvol_begin(struct send_ctx *sctx)
2336{
2337 int ret;
2338 struct btrfs_root *send_root = sctx->send_root;
2339 struct btrfs_root *parent_root = sctx->parent_root;
2340 struct btrfs_path *path;
2341 struct btrfs_key key;
2342 struct btrfs_root_ref *ref;
2343 struct extent_buffer *leaf;
2344 char *name = NULL;
2345 int namelen;
2346
ffcfaf81 2347 path = btrfs_alloc_path();
31db9f7c
AB		 2348 if (!path)
2349 return -ENOMEM;
2350
e780b0d1 2351 name = kmalloc(BTRFS_PATH_NAME_MAX, GFP_KERNEL);
31db9f7c
AB		 2352 if (!name) {
2353 btrfs_free_path(path);
2354 return -ENOMEM;
2355 }
2356
2357 key.objectid = send_root->objectid;
2358 key.type = BTRFS_ROOT_BACKREF_KEY;
2359 key.offset = 0;
2360
2361 ret = btrfs_search_slot_for_read(send_root->fs_info->tree_root,
2362 &key, path, 1, 0);
2363 if (ret < 0)
2364 goto out;
2365 if (ret) {
2366 ret = -ENOENT;
2367 goto out;
2368 }
2369
2370 leaf = path->nodes[0];
2371 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
2372 if (key.type != BTRFS_ROOT_BACKREF_KEY ||
2373 key.objectid != send_root->objectid) {
2374 ret = -ENOENT;
2375 goto out;
2376 }
2377 ref = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_root_ref);
2378 namelen = btrfs_root_ref_name_len(leaf, ref);
2379 read_extent_buffer(leaf, name, (unsigned long)(ref + 1), namelen);
2380 btrfs_release_path(path);
2381
31db9f7c
AB		 2382 if (parent_root) {
2383 ret = begin_cmd(sctx, BTRFS_SEND_C_SNAPSHOT);
2384 if (ret < 0)
2385 goto out;
2386 } else {
2387 ret = begin_cmd(sctx, BTRFS_SEND_C_SUBVOL);
2388 if (ret < 0)
2389 goto out;
2390 }
2391
2392 TLV_PUT_STRING(sctx, BTRFS_SEND_A_PATH, name, namelen);
b96b1db0
RR		 2393
2394 if (!btrfs_is_empty_uuid(sctx->send_root->root_item.received_uuid))
2395 TLV_PUT_UUID(sctx, BTRFS_SEND_A_UUID,
2396 sctx->send_root->root_item.received_uuid);
2397 else
2398 TLV_PUT_UUID(sctx, BTRFS_SEND_A_UUID,
2399 sctx->send_root->root_item.uuid);
2400
31db9f7c 2401 TLV_PUT_U64(sctx, BTRFS_SEND_A_CTRANSID,
5a0f4e2c 2402 le64_to_cpu(sctx->send_root->root_item.ctransid));
31db9f7c 2403 if (parent_root) {
37b8d27d
JB		 2404 if (!btrfs_is_empty_uuid(parent_root->root_item.received_uuid))
2405 TLV_PUT_UUID(sctx, BTRFS_SEND_A_CLONE_UUID,
2406 parent_root->root_item.received_uuid);
2407 else
2408 TLV_PUT_UUID(sctx, BTRFS_SEND_A_CLONE_UUID,
2409 parent_root->root_item.uuid);
31db9f7c 2410 TLV_PUT_U64(sctx, BTRFS_SEND_A_CLONE_CTRANSID,
5a0f4e2c 2411 le64_to_cpu(sctx->parent_root->root_item.ctransid));
31db9f7c
AB		 2412 }
2413
2414 ret = send_cmd(sctx);
2415
2416tlv_put_failure:
2417out:
2418 btrfs_free_path(path);
2419 kfree(name);
2420 return ret;
2421}
2422
2423static int send_truncate(struct send_ctx *sctx, u64 ino, u64 gen, u64 size)
2424{
04ab956e 2425 struct btrfs_fs_info *fs_info = sctx->send_root->fs_info;
31db9f7c
AB		 2426 int ret = 0;
2427 struct fs_path *p;
2428
04ab956e 2429 btrfs_debug(fs_info, "send_truncate %llu size=%llu", ino, size);
31db9f7c 2430
924794c9 2431 p = fs_path_alloc();
31db9f7c
AB		 2432 if (!p)
2433 return -ENOMEM;
2434
2435 ret = begin_cmd(sctx, BTRFS_SEND_C_TRUNCATE);
2436 if (ret < 0)
2437 goto out;
2438
2439 ret = get_cur_path(sctx, ino, gen, p);
2440 if (ret < 0)
2441 goto out;
2442 TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p);
2443 TLV_PUT_U64(sctx, BTRFS_SEND_A_SIZE, size);
2444
2445 ret = send_cmd(sctx);
2446
2447tlv_put_failure:
2448out:
924794c9 2449 fs_path_free(p);
31db9f7c
AB		 2450 return ret;
2451}
2452
2453static int send_chmod(struct send_ctx *sctx, u64 ino, u64 gen, u64 mode)
2454{
04ab956e 2455 struct btrfs_fs_info *fs_info = sctx->send_root->fs_info;
31db9f7c
AB		 2456 int ret = 0;
2457 struct fs_path *p;
2458
04ab956e 2459 btrfs_debug(fs_info, "send_chmod %llu mode=%llu", ino, mode);
31db9f7c 2460
924794c9 2461 p = fs_path_alloc();
31db9f7c
AB		 2462 if (!p)
2463 return -ENOMEM;
2464
2465 ret = begin_cmd(sctx, BTRFS_SEND_C_CHMOD);
2466 if (ret < 0)
2467 goto out;
2468
2469 ret = get_cur_path(sctx, ino, gen, p);
2470 if (ret < 0)
2471 goto out;
2472 TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p);
2473 TLV_PUT_U64(sctx, BTRFS_SEND_A_MODE, mode & 07777);
2474
2475 ret = send_cmd(sctx);
2476
2477tlv_put_failure:
2478out:
924794c9 2479 fs_path_free(p);
31db9f7c
AB		 2480 return ret;
2481}
2482
2483static int send_chown(struct send_ctx *sctx, u64 ino, u64 gen, u64 uid, u64 gid)
2484{
04ab956e 2485 struct btrfs_fs_info *fs_info = sctx->send_root->fs_info;
31db9f7c
AB		 2486 int ret = 0;
2487 struct fs_path *p;
2488
04ab956e
JM		 2489 btrfs_debug(fs_info, "send_chown %llu uid=%llu, gid=%llu",
2490 ino, uid, gid);
31db9f7c 2491
924794c9 2492 p = fs_path_alloc();
31db9f7c
AB		 2493 if (!p)
2494 return -ENOMEM;
2495
2496 ret = begin_cmd(sctx, BTRFS_SEND_C_CHOWN);
2497 if (ret < 0)
2498 goto out;
2499
2500 ret = get_cur_path(sctx, ino, gen, p);
2501 if (ret < 0)
2502 goto out;
2503 TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p);
2504 TLV_PUT_U64(sctx, BTRFS_SEND_A_UID, uid);
2505 TLV_PUT_U64(sctx, BTRFS_SEND_A_GID, gid);
2506
2507 ret = send_cmd(sctx);
2508
2509tlv_put_failure:
2510out:
924794c9 2511 fs_path_free(p);
31db9f7c
AB		 2512 return ret;
2513}
2514
2515static int send_utimes(struct send_ctx *sctx, u64 ino, u64 gen)
2516{
04ab956e 2517 struct btrfs_fs_info *fs_info = sctx->send_root->fs_info;
31db9f7c
AB		 2518 int ret = 0;
2519 struct fs_path *p = NULL;
2520 struct btrfs_inode_item *ii;
2521 struct btrfs_path *path = NULL;
2522 struct extent_buffer *eb;
2523 struct btrfs_key key;
2524 int slot;
2525
04ab956e 2526 btrfs_debug(fs_info, "send_utimes %llu", ino);
31db9f7c 2527
924794c9 2528 p = fs_path_alloc();
31db9f7c
AB		 2529 if (!p)
2530 return -ENOMEM;
2531
2532 path = alloc_path_for_send();
2533 if (!path) {
2534 ret = -ENOMEM;
2535 goto out;
2536 }
2537
2538 key.objectid = ino;
2539 key.type = BTRFS_INODE_ITEM_KEY;
2540 key.offset = 0;
2541 ret = btrfs_search_slot(NULL, sctx->send_root, &key, path, 0, 0);
15b253ea
FM		 2542 if (ret > 0)
2543 ret = -ENOENT;
31db9f7c
AB		 2544 if (ret < 0)
2545 goto out;
2546
2547 eb = path->nodes[0];
2548 slot = path->slots[0];
2549 ii = btrfs_item_ptr(eb, slot, struct btrfs_inode_item);
2550
2551 ret = begin_cmd(sctx, BTRFS_SEND_C_UTIMES);
2552 if (ret < 0)
2553 goto out;
2554
2555 ret = get_cur_path(sctx, ino, gen, p);
2556 if (ret < 0)
2557 goto out;
2558 TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p);
a937b979
DS		 2559 TLV_PUT_BTRFS_TIMESPEC(sctx, BTRFS_SEND_A_ATIME, eb, &ii->atime);
2560 TLV_PUT_BTRFS_TIMESPEC(sctx, BTRFS_SEND_A_MTIME, eb, &ii->mtime);
2561 TLV_PUT_BTRFS_TIMESPEC(sctx, BTRFS_SEND_A_CTIME, eb, &ii->ctime);
766702ef 2562 /* TODO Add otime support when the otime patches get into upstream */
31db9f7c
AB		 2563
2564 ret = send_cmd(sctx);
2565
2566tlv_put_failure:
2567out:
924794c9 2568 fs_path_free(p);
31db9f7c
AB		 2569 btrfs_free_path(path);
2570 return ret;
2571}
2572
2573/*
2574 * Sends a BTRFS_SEND_C_MKXXX or SYMLINK command to user space. We don't have
2575 * a valid path yet because we did not process the refs yet. So, the inode
2576 * is created as orphan.
2577 */
1f4692da 2578static int send_create_inode(struct send_ctx *sctx, u64 ino)
31db9f7c 2579{
04ab956e 2580 struct btrfs_fs_info *fs_info = sctx->send_root->fs_info;
31db9f7c 2581 int ret = 0;
31db9f7c 2582 struct fs_path *p;
31db9f7c 2583 int cmd;
1f4692da 2584 u64 gen;
31db9f7c 2585 u64 mode;
1f4692da 2586 u64 rdev;
31db9f7c 2587
04ab956e 2588 btrfs_debug(fs_info, "send_create_inode %llu", ino);
31db9f7c 2589
924794c9 2590 p = fs_path_alloc();
31db9f7c
AB		 2591 if (!p)
2592 return -ENOMEM;
2593
644d1940
LB		 2594 if (ino != sctx->cur_ino) {
2595 ret = get_inode_info(sctx->send_root, ino, NULL, &gen, &mode,
2596 NULL, NULL, &rdev);
2597 if (ret < 0)
2598 goto out;
2599 } else {
2600 gen = sctx->cur_inode_gen;
2601 mode = sctx->cur_inode_mode;
2602 rdev = sctx->cur_inode_rdev;
2603 }
31db9f7c 2604
e938c8ad 2605 if (S_ISREG(mode)) {
31db9f7c 2606 cmd = BTRFS_SEND_C_MKFILE;
e938c8ad 2607 } else if (S_ISDIR(mode)) {
31db9f7c 2608 cmd = BTRFS_SEND_C_MKDIR;
e938c8ad 2609 } else if (S_ISLNK(mode)) {
31db9f7c 2610 cmd = BTRFS_SEND_C_SYMLINK;
e938c8ad 2611 } else if (S_ISCHR(mode) || S_ISBLK(mode)) {
31db9f7c 2612 cmd = BTRFS_SEND_C_MKNOD;
e938c8ad 2613 } else if (S_ISFIFO(mode)) {
31db9f7c 2614 cmd = BTRFS_SEND_C_MKFIFO;
e938c8ad 2615 } else if (S_ISSOCK(mode)) {
31db9f7c 2616 cmd = BTRFS_SEND_C_MKSOCK;
e938c8ad 2617 } else {
f14d104d 2618 btrfs_warn(sctx->send_root->fs_info, "unexpected inode type %o",
31db9f7c
AB		 2619 (int)(mode & S_IFMT));
2620 ret = -ENOTSUPP;
2621 goto out;
2622 }
2623
2624 ret = begin_cmd(sctx, cmd);
2625 if (ret < 0)
2626 goto out;
2627
1f4692da 2628 ret = gen_unique_name(sctx, ino, gen, p);
31db9f7c
AB		 2629 if (ret < 0)
2630 goto out;
2631
2632 TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p);
1f4692da 2633 TLV_PUT_U64(sctx, BTRFS_SEND_A_INO, ino);
31db9f7c
AB		 2634
2635 if (S_ISLNK(mode)) {
2636 fs_path_reset(p);
924794c9 2637 ret = read_symlink(sctx->send_root, ino, p);
31db9f7c
AB		 2638 if (ret < 0)
2639 goto out;
2640 TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH_LINK, p);
2641 } else if (S_ISCHR(mode) || S_ISBLK(mode) ||
2642 S_ISFIFO(mode) || S_ISSOCK(mode)) {
d79e5043
AJ		 2643 TLV_PUT_U64(sctx, BTRFS_SEND_A_RDEV, new_encode_dev(rdev));
2644 TLV_PUT_U64(sctx, BTRFS_SEND_A_MODE, mode);
31db9f7c
AB		 2645 }
2646
2647 ret = send_cmd(sctx);
2648 if (ret < 0)
2649 goto out;
2650
2651
2652tlv_put_failure:
2653out:
924794c9 2654 fs_path_free(p);
31db9f7c
AB		 2655 return ret;
2656}
2657
1f4692da
AB		 2658/*
2659 * We need some special handling for inodes that get processed before the parent
2660 * directory got created. See process_recorded_refs for details.
2661 * This function does the check if we already created the dir out of order.
2662 */
2663static int did_create_dir(struct send_ctx *sctx, u64 dir)
2664{
2665 int ret = 0;
2666 struct btrfs_path *path = NULL;
2667 struct btrfs_key key;
2668 struct btrfs_key found_key;
2669 struct btrfs_key di_key;
2670 struct extent_buffer *eb;
2671 struct btrfs_dir_item *di;
2672 int slot;
2673
2674 path = alloc_path_for_send();
2675 if (!path) {
2676 ret = -ENOMEM;
2677 goto out;
2678 }
2679
2680 key.objectid = dir;
2681 key.type = BTRFS_DIR_INDEX_KEY;
2682 key.offset = 0;
dff6d0ad
FDBM		 2683 ret = btrfs_search_slot(NULL, sctx->send_root, &key, path, 0, 0);
2684 if (ret < 0)
2685 goto out;
2686
1f4692da 2687 while (1) {
dff6d0ad
FDBM		 2688 eb = path->nodes[0];
2689 slot = path->slots[0];
2690 if (slot >= btrfs_header_nritems(eb)) {
2691 ret = btrfs_next_leaf(sctx->send_root, path);
2692 if (ret < 0) {
2693 goto out;
2694 } else if (ret > 0) {
2695 ret = 0;
2696 break;
2697 }
2698 continue;
1f4692da 2699 }
dff6d0ad
FDBM		 2700
2701 btrfs_item_key_to_cpu(eb, &found_key, slot);
2702 if (found_key.objectid != key.objectid ||
1f4692da
AB		 2703 found_key.type != key.type) {
2704 ret = 0;
2705 goto out;
2706 }
2707
2708 di = btrfs_item_ptr(eb, slot, struct btrfs_dir_item);
2709 btrfs_dir_item_key_to_cpu(eb, di, &di_key);
2710
a0525414
JB		 2711 if (di_key.type != BTRFS_ROOT_ITEM_KEY &&
2712 di_key.objectid < sctx->send_progress) {
1f4692da
AB		 2713 ret = 1;
2714 goto out;
2715 }
2716
dff6d0ad 2717 path->slots[0]++;
1f4692da
AB		 2718 }
2719
2720out:
2721 btrfs_free_path(path);
2722 return ret;
2723}
2724
2725/*
2726 * Only creates the inode if it is:
2727 * 1. Not a directory
2728 * 2. Or a directory which was not created already due to out of order
2729 * directories. See did_create_dir and process_recorded_refs for details.
2730 */
2731static int send_create_inode_if_needed(struct send_ctx *sctx)
2732{
2733 int ret;
2734
2735 if (S_ISDIR(sctx->cur_inode_mode)) {
2736 ret = did_create_dir(sctx, sctx->cur_ino);
2737 if (ret < 0)
2738 goto out;
2739 if (ret) {
2740 ret = 0;
2741 goto out;
2742 }
2743 }
2744
2745 ret = send_create_inode(sctx, sctx->cur_ino);
2746 if (ret < 0)
2747 goto out;
2748
2749out:
2750 return ret;
2751}
2752
31db9f7c
AB		 2753struct recorded_ref {
2754 struct list_head list;
2755 char *dir_path;
2756 char *name;
2757 struct fs_path *full_path;
2758 u64 dir;
2759 u64 dir_gen;
2760 int dir_path_len;
2761 int name_len;
2762};
2763
2764/*
2765 * We need to process new refs before deleted refs, but compare_tree gives us
2766 * everything mixed. So we first record all refs and later process them.
2767 * This function is a helper to record one ref.
2768 */
a4d96d62 2769static int __record_ref(struct list_head *head, u64 dir,
31db9f7c
AB		 2770 u64 dir_gen, struct fs_path *path)
2771{
2772 struct recorded_ref *ref;
31db9f7c 2773
e780b0d1 2774 ref = kmalloc(sizeof(*ref), GFP_KERNEL);
31db9f7c
AB		 2775 if (!ref)
2776 return -ENOMEM;
2777
2778 ref->dir = dir;
2779 ref->dir_gen = dir_gen;
2780 ref->full_path = path;
2781
ed84885d
AS		 2782 ref->name = (char *)kbasename(ref->full_path->start);
2783 ref->name_len = ref->full_path->end - ref->name;
2784 ref->dir_path = ref->full_path->start;
2785 if (ref->name == ref->full_path->start)
31db9f7c 2786 ref->dir_path_len = 0;
ed84885d 2787 else
31db9f7c
AB		 2788 ref->dir_path_len = ref->full_path->end -
2789 ref->full_path->start - 1 - ref->name_len;
31db9f7c
AB		 2790
2791 list_add_tail(&ref->list, head);
2792 return 0;
2793}
2794
ba5e8f2e
JB		 2795static int dup_ref(struct recorded_ref *ref, struct list_head *list)
2796{
2797 struct recorded_ref *new;
2798
e780b0d1 2799 new = kmalloc(sizeof(*ref), GFP_KERNEL);
ba5e8f2e
JB		 2800 if (!new)
2801 return -ENOMEM;
2802
2803 new->dir = ref->dir;
2804 new->dir_gen = ref->dir_gen;
2805 new->full_path = NULL;
2806 INIT_LIST_HEAD(&new->list);
2807 list_add_tail(&new->list, list);
2808 return 0;
2809}
2810
924794c9 2811static void __free_recorded_refs(struct list_head *head)
31db9f7c
AB		 2812{
2813 struct recorded_ref *cur;
31db9f7c 2814
e938c8ad
AB		 2815 while (!list_empty(head)) {
2816 cur = list_entry(head->next, struct recorded_ref, list);
924794c9 2817 fs_path_free(cur->full_path);
e938c8ad 2818 list_del(&cur->list);
31db9f7c
AB		 2819 kfree(cur);
2820 }
31db9f7c
AB		 2821}
2822
2823static void free_recorded_refs(struct send_ctx *sctx)
2824{
924794c9
TI		 2825 __free_recorded_refs(&sctx->new_refs);
2826 __free_recorded_refs(&sctx->deleted_refs);
31db9f7c
AB		 2827}
2828
2829/*
766702ef 2830 * Renames/moves a file/dir to its orphan name. Used when the first
31db9f7c
AB		 2831 * ref of an unprocessed inode gets overwritten and for all non empty
2832 * directories.
2833 */
2834static int orphanize_inode(struct send_ctx *sctx, u64 ino, u64 gen,
2835 struct fs_path *path)
2836{
2837 int ret;
2838 struct fs_path *orphan;
2839
924794c9 2840 orphan = fs_path_alloc();
31db9f7c
AB		 2841 if (!orphan)
2842 return -ENOMEM;
2843
2844 ret = gen_unique_name(sctx, ino, gen, orphan);
2845 if (ret < 0)
2846 goto out;
2847
2848 ret = send_rename(sctx, path, orphan);
2849
2850out:
924794c9 2851 fs_path_free(orphan);
31db9f7c
AB		 2852 return ret;
2853}
2854
9dc44214
FM		 2855static struct orphan_dir_info *
2856add_orphan_dir_info(struct send_ctx *sctx, u64 dir_ino)
2857{
2858 struct rb_node **p = &sctx->orphan_dirs.rb_node;
2859 struct rb_node *parent = NULL;
2860 struct orphan_dir_info *entry, *odi;
2861
e780b0d1 2862 odi = kmalloc(sizeof(*odi), GFP_KERNEL);
9dc44214
FM		 2863 if (!odi)
2864 return ERR_PTR(-ENOMEM);
2865 odi->ino = dir_ino;
2866 odi->gen = 0;
2867
2868 while (*p) {
2869 parent = *p;
2870 entry = rb_entry(parent, struct orphan_dir_info, node);
2871 if (dir_ino < entry->ino) {
2872 p = &(*p)->rb_left;
2873 } else if (dir_ino > entry->ino) {
2874 p = &(*p)->rb_right;
2875 } else {
2876 kfree(odi);
2877 return entry;
2878 }
2879 }
2880
2881 rb_link_node(&odi->node, parent, p);
2882 rb_insert_color(&odi->node, &sctx->orphan_dirs);
2883 return odi;
2884}
2885
2886static struct orphan_dir_info *
2887get_orphan_dir_info(struct send_ctx *sctx, u64 dir_ino)
2888{
2889 struct rb_node *n = sctx->orphan_dirs.rb_node;
2890 struct orphan_dir_info *entry;
2891
2892 while (n) {
2893 entry = rb_entry(n, struct orphan_dir_info, node);
2894 if (dir_ino < entry->ino)
2895 n = n->rb_left;
2896 else if (dir_ino > entry->ino)
2897 n = n->rb_right;
2898 else
2899 return entry;
2900 }
2901 return NULL;
2902}
2903
2904static int is_waiting_for_rm(struct send_ctx *sctx, u64 dir_ino)
2905{
2906 struct orphan_dir_info *odi = get_orphan_dir_info(sctx, dir_ino);
2907
2908 return odi != NULL;
2909}
2910
2911static void free_orphan_dir_info(struct send_ctx *sctx,
2912 struct orphan_dir_info *odi)
2913{
2914 if (!odi)
2915 return;
2916 rb_erase(&odi->node, &sctx->orphan_dirs);
2917 kfree(odi);
2918}
2919
31db9f7c
AB		 2920/*
2921 * Returns 1 if a directory can be removed at this point in time.
2922 * We check this by iterating all dir items and checking if the inode behind
2923 * the dir item was already processed.
2924 */
9dc44214
FM		 2925static int can_rmdir(struct send_ctx *sctx, u64 dir, u64 dir_gen,
2926 u64 send_progress)
31db9f7c
AB		 2927{
2928 int ret = 0;
2929 struct btrfs_root *root = sctx->parent_root;
2930 struct btrfs_path *path;
2931 struct btrfs_key key;
2932 struct btrfs_key found_key;
2933 struct btrfs_key loc;
2934 struct btrfs_dir_item *di;
2935
6d85ed05
AB		 2936 /*
2937 * Don't try to rmdir the top/root subvolume dir.
2938 */
2939 if (dir == BTRFS_FIRST_FREE_OBJECTID)
2940 return 0;
2941
31db9f7c
AB		 2942 path = alloc_path_for_send();
2943 if (!path)
2944 return -ENOMEM;
2945
2946 key.objectid = dir;
2947 key.type = BTRFS_DIR_INDEX_KEY;
2948 key.offset = 0;
dff6d0ad
FDBM		 2949 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
2950 if (ret < 0)
2951 goto out;
31db9f7c
AB		 2952
2953 while (1) {
9dc44214
FM		 2954 struct waiting_dir_move *dm;
2955
dff6d0ad
FDBM		 2956 if (path->slots[0] >= btrfs_header_nritems(path->nodes[0])) {
2957 ret = btrfs_next_leaf(root, path);
2958 if (ret < 0)
2959 goto out;
2960 else if (ret > 0)
2961 break;
2962 continue;
31db9f7c 2963 }
dff6d0ad
FDBM		 2964 btrfs_item_key_to_cpu(path->nodes[0], &found_key,
2965 path->slots[0]);
2966 if (found_key.objectid != key.objectid ||
2967 found_key.type != key.type)
31db9f7c 2968 break;
31db9f7c
AB		 2969
2970 di = btrfs_item_ptr(path->nodes[0], path->slots[0],
2971 struct btrfs_dir_item);
2972 btrfs_dir_item_key_to_cpu(path->nodes[0], di, &loc);
2973
9dc44214
FM		 2974 dm = get_waiting_dir_move(sctx, loc.objectid);
2975 if (dm) {
2976 struct orphan_dir_info *odi;
2977
2978 odi = add_orphan_dir_info(sctx, dir);
2979 if (IS_ERR(odi)) {
2980 ret = PTR_ERR(odi);
2981 goto out;
2982 }
2983 odi->gen = dir_gen;
2984 dm->rmdir_ino = dir;
2985 ret = 0;
2986 goto out;
2987 }
2988
31db9f7c 2989 if (loc.objectid > send_progress) {
443f9d26
RK		 2990 struct orphan_dir_info *odi;
2991
2992 odi = get_orphan_dir_info(sctx, dir);
2993 free_orphan_dir_info(sctx, odi);
31db9f7c
AB		 2994 ret = 0;
2995 goto out;
2996 }
2997
dff6d0ad 2998 path->slots[0]++;
31db9f7c
AB		 2999 }
3000
3001 ret = 1;
3002
3003out:
3004 btrfs_free_path(path);
3005 return ret;
3006}
3007
9f03740a
FDBM		 3008static int is_waiting_for_move(struct send_ctx *sctx, u64 ino)
3009{
9dc44214 3010 struct waiting_dir_move *entry = get_waiting_dir_move(sctx, ino);
9f03740a 3011
9dc44214 3012 return entry != NULL;
9f03740a
FDBM		 3013}
3014
8b191a68 3015static int add_waiting_dir_move(struct send_ctx *sctx, u64 ino, bool orphanized)
9f03740a
FDBM		 3016{
3017 struct rb_node **p = &sctx->waiting_dir_moves.rb_node;
3018 struct rb_node *parent = NULL;
3019 struct waiting_dir_move *entry, *dm;
3020
e780b0d1 3021 dm = kmalloc(sizeof(*dm), GFP_KERNEL);
9f03740a
FDBM		 3022 if (!dm)
3023 return -ENOMEM;
3024 dm->ino = ino;
9dc44214 3025 dm->rmdir_ino = 0;
8b191a68 3026 dm->orphanized = orphanized;
9f03740a
FDBM		 3027
3028 while (*p) {
3029 parent = *p;
3030 entry = rb_entry(parent, struct waiting_dir_move, node);
3031 if (ino < entry->ino) {
3032 p = &(*p)->rb_left;
3033 } else if (ino > entry->ino) {
3034 p = &(*p)->rb_right;
3035 } else {
3036 kfree(dm);
3037 return -EEXIST;
3038 }
3039 }
3040
3041 rb_link_node(&dm->node, parent, p);
3042 rb_insert_color(&dm->node, &sctx->waiting_dir_moves);
3043 return 0;
3044}
3045
9dc44214
FM		 3046static struct waiting_dir_move *
3047get_waiting_dir_move(struct send_ctx *sctx, u64 ino)
9f03740a
FDBM		 3048{
3049 struct rb_node *n = sctx->waiting_dir_moves.rb_node;
3050 struct waiting_dir_move *entry;
3051
3052 while (n) {
3053 entry = rb_entry(n, struct waiting_dir_move, node);
9dc44214 3054 if (ino < entry->ino)
9f03740a 3055 n = n->rb_left;
9dc44214 3056 else if (ino > entry->ino)
9f03740a 3057 n = n->rb_right;
9dc44214
FM		 3058 else
3059 return entry;
9f03740a 3060 }
9dc44214
FM		 3061 return NULL;
3062}
3063
3064static void free_waiting_dir_move(struct send_ctx *sctx,
3065 struct waiting_dir_move *dm)
3066{
3067 if (!dm)
3068 return;
3069 rb_erase(&dm->node, &sctx->waiting_dir_moves);
3070 kfree(dm);
9f03740a
FDBM		 3071}
3072
bfa7e1f8
FM		 3073static int add_pending_dir_move(struct send_ctx *sctx,
3074 u64 ino,
3075 u64 ino_gen,
f959492f
FM		 3076 u64 parent_ino,
3077 struct list_head *new_refs,
84471e24
FM		 3078 struct list_head *deleted_refs,
3079 const bool is_orphan)
9f03740a
FDBM		 3080{
3081 struct rb_node **p = &sctx->pending_dir_moves.rb_node;
3082 struct rb_node *parent = NULL;
73b802f4 3083 struct pending_dir_move *entry = NULL, *pm;
9f03740a
FDBM		 3084 struct recorded_ref *cur;
3085 int exists = 0;
3086 int ret;
3087
e780b0d1 3088 pm = kmalloc(sizeof(*pm), GFP_KERNEL);
9f03740a
FDBM		 3089 if (!pm)
3090 return -ENOMEM;
3091 pm->parent_ino = parent_ino;
bfa7e1f8
FM		 3092 pm->ino = ino;
3093 pm->gen = ino_gen;
9f03740a
FDBM		 3094 INIT_LIST_HEAD(&pm->list);
3095 INIT_LIST_HEAD(&pm->update_refs);
3096 RB_CLEAR_NODE(&pm->node);
3097
3098 while (*p) {
3099 parent = *p;
3100 entry = rb_entry(parent, struct pending_dir_move, node);
3101 if (parent_ino < entry->parent_ino) {
3102 p = &(*p)->rb_left;
3103 } else if (parent_ino > entry->parent_ino) {
3104 p = &(*p)->rb_right;
3105 } else {
3106 exists = 1;
3107 break;
3108 }
3109 }
3110
f959492f 3111 list_for_each_entry(cur, deleted_refs, list) {
9f03740a
FDBM		 3112 ret = dup_ref(cur, &pm->update_refs);
3113 if (ret < 0)
3114 goto out;
3115 }
f959492f 3116 list_for_each_entry(cur, new_refs, list) {
9f03740a
FDBM		 3117 ret = dup_ref(cur, &pm->update_refs);
3118 if (ret < 0)
3119 goto out;
3120 }
3121
8b191a68 3122 ret = add_waiting_dir_move(sctx, pm->ino, is_orphan);
9f03740a
FDBM		 3123 if (ret)
3124 goto out;
3125
3126 if (exists) {
3127 list_add_tail(&pm->list, &entry->list);
3128 } else {
3129 rb_link_node(&pm->node, parent, p);
3130 rb_insert_color(&pm->node, &sctx->pending_dir_moves);
3131 }
3132 ret = 0;
3133out:
3134 if (ret) {
3135 __free_recorded_refs(&pm->update_refs);
3136 kfree(pm);
3137 }
3138 return ret;
3139}
3140
3141static struct pending_dir_move *get_pending_dir_moves(struct send_ctx *sctx,
3142 u64 parent_ino)
3143{
3144 struct rb_node *n = sctx->pending_dir_moves.rb_node;
3145 struct pending_dir_move *entry;
3146
3147 while (n) {
3148 entry = rb_entry(n, struct pending_dir_move, node);
3149 if (parent_ino < entry->parent_ino)
3150 n = n->rb_left;
3151 else if (parent_ino > entry->parent_ino)
3152 n = n->rb_right;
3153 else
3154 return entry;
3155 }
3156 return NULL;
3157}
3158
801bec36
RK		 3159static int path_loop(struct send_ctx *sctx, struct fs_path *name,
3160 u64 ino, u64 gen, u64 *ancestor_ino)
3161{
3162 int ret = 0;
3163 u64 parent_inode = 0;
3164 u64 parent_gen = 0;
3165 u64 start_ino = ino;
3166
3167 *ancestor_ino = 0;
3168 while (ino != BTRFS_FIRST_FREE_OBJECTID) {
3169 fs_path_reset(name);
3170
3171 if (is_waiting_for_rm(sctx, ino))
3172 break;
3173 if (is_waiting_for_move(sctx, ino)) {
3174 if (*ancestor_ino == 0)
3175 *ancestor_ino = ino;
3176 ret = get_first_ref(sctx->parent_root, ino,
3177 &parent_inode, &parent_gen, name);
3178 } else {
3179 ret = __get_cur_name_and_parent(sctx, ino, gen,
3180 &parent_inode,
3181 &parent_gen, name);
3182 if (ret > 0) {
3183 ret = 0;
3184 break;
3185 }
3186 }
3187 if (ret < 0)
3188 break;
3189 if (parent_inode == start_ino) {
3190 ret = 1;
3191 if (*ancestor_ino == 0)
3192 *ancestor_ino = ino;
3193 break;
3194 }
3195 ino = parent_inode;
3196 gen = parent_gen;
3197 }
3198 return ret;
3199}
3200
9f03740a
FDBM		 3201static int apply_dir_move(struct send_ctx *sctx, struct pending_dir_move *pm)
3202{
3203 struct fs_path *from_path = NULL;
3204 struct fs_path *to_path = NULL;
2b863a13 3205 struct fs_path *name = NULL;
9f03740a
FDBM		 3206 u64 orig_progress = sctx->send_progress;
3207 struct recorded_ref *cur;
2b863a13 3208 u64 parent_ino, parent_gen;
9dc44214
FM		 3209 struct waiting_dir_move *dm = NULL;
3210 u64 rmdir_ino = 0;
801bec36
RK		 3211 u64 ancestor;
3212 bool is_orphan;
9f03740a
FDBM		 3213 int ret;
3214
2b863a13 3215 name = fs_path_alloc();
9f03740a 3216 from_path = fs_path_alloc();
2b863a13
FM		 3217 if (!name || !from_path) {
3218 ret = -ENOMEM;
3219 goto out;
3220 }
9f03740a 3221
9dc44214
FM		 3222 dm = get_waiting_dir_move(sctx, pm->ino);
3223 ASSERT(dm);
3224 rmdir_ino = dm->rmdir_ino;
801bec36 3225 is_orphan = dm->orphanized;
9dc44214 3226 free_waiting_dir_move(sctx, dm);
2b863a13 3227
801bec36 3228 if (is_orphan) {
84471e24
FM		 3229 ret = gen_unique_name(sctx, pm->ino,
3230 pm->gen, from_path);
3231 } else {
3232 ret = get_first_ref(sctx->parent_root, pm->ino,
3233 &parent_ino, &parent_gen, name);
3234 if (ret < 0)
3235 goto out;
3236 ret = get_cur_path(sctx, parent_ino, parent_gen,
3237 from_path);
3238 if (ret < 0)
3239 goto out;
3240 ret = fs_path_add_path(from_path, name);
3241 }
c992ec94
FM		 3242 if (ret < 0)
3243 goto out;
2b863a13 3244
f959492f 3245 sctx->send_progress = sctx->cur_ino + 1;
801bec36 3246 ret = path_loop(sctx, name, pm->ino, pm->gen, &ancestor);
7969e77a
FM		 3247 if (ret < 0)
3248 goto out;
801bec36
RK		 3249 if (ret) {
3250 LIST_HEAD(deleted_refs);
3251 ASSERT(ancestor > BTRFS_FIRST_FREE_OBJECTID);
3252 ret = add_pending_dir_move(sctx, pm->ino, pm->gen, ancestor,
3253 &pm->update_refs, &deleted_refs,
3254 is_orphan);
3255 if (ret < 0)
3256 goto out;
3257 if (rmdir_ino) {
3258 dm = get_waiting_dir_move(sctx, pm->ino);
3259 ASSERT(dm);
3260 dm->rmdir_ino = rmdir_ino;
3261 }
3262 goto out;
3263 }
c992ec94
FM		 3264 fs_path_reset(name);
3265 to_path = name;
2b863a13 3266 name = NULL;
9f03740a
FDBM		 3267 ret = get_cur_path(sctx, pm->ino, pm->gen, to_path);
3268 if (ret < 0)
3269 goto out;
3270
3271 ret = send_rename(sctx, from_path, to_path);
3272 if (ret < 0)
3273 goto out;
3274
9dc44214
FM		 3275 if (rmdir_ino) {
3276 struct orphan_dir_info *odi;
3277
3278 odi = get_orphan_dir_info(sctx, rmdir_ino);
3279 if (!odi) {
3280 /* already deleted */
3281 goto finish;
3282 }
99ea42dd 3283 ret = can_rmdir(sctx, rmdir_ino, odi->gen, sctx->cur_ino);
9dc44214
FM		 3284 if (ret < 0)
3285 goto out;
3286 if (!ret)
3287 goto finish;
3288
3289 name = fs_path_alloc();
3290 if (!name) {
3291 ret = -ENOMEM;
3292 goto out;
3293 }
3294 ret = get_cur_path(sctx, rmdir_ino, odi->gen, name);
3295 if (ret < 0)
3296 goto out;
3297 ret = send_rmdir(sctx, name);
3298 if (ret < 0)
3299 goto out;
3300 free_orphan_dir_info(sctx, odi);
3301 }
3302
3303finish:
9f03740a
FDBM		 3304 ret = send_utimes(sctx, pm->ino, pm->gen);
3305 if (ret < 0)
3306 goto out;
3307
3308 /*
3309 * After rename/move, need to update the utimes of both new parent(s)
3310 * and old parent(s).
3311 */
3312 list_for_each_entry(cur, &pm->update_refs, list) {
764433a1
RK		 3313 /*
3314 * The parent inode might have been deleted in the send snapshot
3315 */
3316 ret = get_inode_info(sctx->send_root, cur->dir, NULL,
3317 NULL, NULL, NULL, NULL, NULL);
3318 if (ret == -ENOENT) {
3319 ret = 0;
9dc44214 3320 continue;
764433a1
RK		 3321 }
3322 if (ret < 0)
3323 goto out;
3324
9f03740a
FDBM		 3325 ret = send_utimes(sctx, cur->dir, cur->dir_gen);
3326 if (ret < 0)
3327 goto out;
3328 }
3329
3330out:
2b863a13 3331 fs_path_free(name);
9f03740a
FDBM		 3332 fs_path_free(from_path);
3333 fs_path_free(to_path);
3334 sctx->send_progress = orig_progress;
3335
3336 return ret;
3337}
3338
3339static void free_pending_move(struct send_ctx *sctx, struct pending_dir_move *m)
3340{
3341 if (!list_empty(&m->list))
3342 list_del(&m->list);
3343 if (!RB_EMPTY_NODE(&m->node))
3344 rb_erase(&m->node, &sctx->pending_dir_moves);
3345 __free_recorded_refs(&m->update_refs);
3346 kfree(m);
3347}
3348
3349static void tail_append_pending_moves(struct pending_dir_move *moves,
3350 struct list_head *stack)
3351{
3352 if (list_empty(&moves->list)) {
3353 list_add_tail(&moves->list, stack);
3354 } else {
3355 LIST_HEAD(list);
3356 list_splice_init(&moves->list, &list);
3357 list_add_tail(&moves->list, stack);
3358 list_splice_tail(&list, stack);
3359 }
3360}
3361
3362static int apply_children_dir_moves(struct send_ctx *sctx)
3363{
3364 struct pending_dir_move *pm;
3365 struct list_head stack;
3366 u64 parent_ino = sctx->cur_ino;
3367 int ret = 0;
3368
3369 pm = get_pending_dir_moves(sctx, parent_ino);
3370 if (!pm)
3371 return 0;
3372
3373 INIT_LIST_HEAD(&stack);
3374 tail_append_pending_moves(pm, &stack);
3375
3376 while (!list_empty(&stack)) {
3377 pm = list_first_entry(&stack, struct pending_dir_move, list);
3378 parent_ino = pm->ino;
3379 ret = apply_dir_move(sctx, pm);
3380 free_pending_move(sctx, pm);
3381 if (ret)
3382 goto out;
3383 pm = get_pending_dir_moves(sctx, parent_ino);
3384 if (pm)
3385 tail_append_pending_moves(pm, &stack);
3386 }
3387 return 0;
3388
3389out:
3390 while (!list_empty(&stack)) {
3391 pm = list_first_entry(&stack, struct pending_dir_move, list);
3392 free_pending_move(sctx, pm);
3393 }
3394 return ret;
3395}
3396
84471e24
FM		 3397/*
3398 * We might need to delay a directory rename even when no ancestor directory
3399 * (in the send root) with a higher inode number than ours (sctx->cur_ino) was
3400 * renamed. This happens when we rename a directory to the old name (the name
3401 * in the parent root) of some other unrelated directory that got its rename
3402 * delayed due to some ancestor with higher number that got renamed.
3403 *
3404 * Example:
3405 *
3406 * Parent snapshot:
3407 * . (ino 256)
3408 * |---- a/ (ino 257)
3409 * | |---- file (ino 260)
3410 * |
3411 * |---- b/ (ino 258)
3412 * |---- c/ (ino 259)
3413 *
3414 * Send snapshot:
3415 * . (ino 256)
3416 * |---- a/ (ino 258)
3417 * |---- x/ (ino 259)
3418 * |---- y/ (ino 257)
3419 * |----- file (ino 260)
3420 *
3421 * Here we can not rename 258 from 'b' to 'a' without the rename of inode 257
3422 * from 'a' to 'x/y' happening first, which in turn depends on the rename of
3423 * inode 259 from 'c' to 'x'. So the order of rename commands the send stream
3424 * must issue is:
3425 *
3426 * 1 - rename 259 from 'c' to 'x'
3427 * 2 - rename 257 from 'a' to 'x/y'
3428 * 3 - rename 258 from 'b' to 'a'
3429 *
3430 * Returns 1 if the rename of sctx->cur_ino needs to be delayed, 0 if it can
3431 * be done right away and < 0 on error.
3432 */
3433static int wait_for_dest_dir_move(struct send_ctx *sctx,
3434 struct recorded_ref *parent_ref,
3435 const bool is_orphan)
3436{
3437 struct btrfs_path *path;
3438 struct btrfs_key key;
3439 struct btrfs_key di_key;
3440 struct btrfs_dir_item *di;
3441 u64 left_gen;
3442 u64 right_gen;
3443 int ret = 0;
801bec36 3444 struct waiting_dir_move *wdm;
84471e24
FM		 3445
3446 if (RB_EMPTY_ROOT(&sctx->waiting_dir_moves))
3447 return 0;
3448
3449 path = alloc_path_for_send();
3450 if (!path)
3451 return -ENOMEM;
3452
3453 key.objectid = parent_ref->dir;
3454 key.type = BTRFS_DIR_ITEM_KEY;
3455 key.offset = btrfs_name_hash(parent_ref->name, parent_ref->name_len);
3456
3457 ret = btrfs_search_slot(NULL, sctx->parent_root, &key, path, 0, 0);
3458 if (ret < 0) {
3459 goto out;
3460 } else if (ret > 0) {
3461 ret = 0;
3462 goto out;
3463 }
3464
3465 di = btrfs_match_dir_item_name(sctx->parent_root, path,
3466 parent_ref->name, parent_ref->name_len);
3467 if (!di) {
3468 ret = 0;
3469 goto out;
3470 }
3471 /*
3472 * di_key.objectid has the number of the inode that has a dentry in the
3473 * parent directory with the same name that sctx->cur_ino is being
3474 * renamed to. We need to check if that inode is in the send root as
3475 * well and if it is currently marked as an inode with a pending rename,
3476 * if it is, we need to delay the rename of sctx->cur_ino as well, so
3477 * that it happens after that other inode is renamed.
3478 */
3479 btrfs_dir_item_key_to_cpu(path->nodes[0], di, &di_key);
3480 if (di_key.type != BTRFS_INODE_ITEM_KEY) {
3481 ret = 0;
3482 goto out;
3483 }
3484
3485 ret = get_inode_info(sctx->parent_root, di_key.objectid, NULL,
3486 &left_gen, NULL, NULL, NULL, NULL);
3487 if (ret < 0)
3488 goto out;
3489 ret = get_inode_info(sctx->send_root, di_key.objectid, NULL,
3490 &right_gen, NULL, NULL, NULL, NULL);
3491 if (ret < 0) {
3492 if (ret == -ENOENT)
3493 ret = 0;
3494 goto out;
3495 }
3496
3497 /* Different inode, no need to delay the rename of sctx->cur_ino */
3498 if (right_gen != left_gen) {
3499 ret = 0;
3500 goto out;
3501 }
3502
801bec36
RK		 3503 wdm = get_waiting_dir_move(sctx, di_key.objectid);
3504 if (wdm && !wdm->orphanized) {
84471e24
FM		 3505 ret = add_pending_dir_move(sctx,
3506 sctx->cur_ino,
3507 sctx->cur_inode_gen,
3508 di_key.objectid,
3509 &sctx->new_refs,
3510 &sctx->deleted_refs,
3511 is_orphan);
3512 if (!ret)
3513 ret = 1;
3514 }
3515out:
3516 btrfs_free_path(path);
3517 return ret;
3518}
3519
80aa6027
FM		 3520/*
3521 * Check if ino ino1 is an ancestor of inode ino2 in the given root.
3522 * Return 1 if true, 0 if false and < 0 on error.
3523 */
3524static int is_ancestor(struct btrfs_root *root,
3525 const u64 ino1,
3526 const u64 ino1_gen,
3527 const u64 ino2,
3528 struct fs_path *fs_path)
3529{
3530 u64 ino = ino2;
3531
3532 while (ino > BTRFS_FIRST_FREE_OBJECTID) {
3533 int ret;
3534 u64 parent;
3535 u64 parent_gen;
3536
3537 fs_path_reset(fs_path);
3538 ret = get_first_ref(root, ino, &parent, &parent_gen, fs_path);
3539 if (ret < 0) {
3540 if (ret == -ENOENT && ino == ino2)
3541 ret = 0;
3542 return ret;
3543 }
3544 if (parent == ino1)
3545 return parent_gen == ino1_gen ? 1 : 0;
3546 ino = parent;
3547 }
3548 return 0;
3549}
3550
9f03740a 3551static int wait_for_parent_move(struct send_ctx *sctx,
8b191a68
FM		 3552 struct recorded_ref *parent_ref,
3553 const bool is_orphan)
9f03740a 3554{
f959492f 3555 int ret = 0;
9f03740a
FDBM		 3556 u64 ino = parent_ref->dir;
3557 u64 parent_ino_before, parent_ino_after;
9f03740a
FDBM		 3558 struct fs_path *path_before = NULL;
3559 struct fs_path *path_after = NULL;
3560 int len1, len2;
9f03740a
FDBM		 3561
3562 path_after = fs_path_alloc();
f959492f
FM		 3563 path_before = fs_path_alloc();
3564 if (!path_after || !path_before) {
9f03740a
FDBM		 3565 ret = -ENOMEM;
3566 goto out;
3567 }
3568
bfa7e1f8 3569 /*
f959492f
FM		 3570 * Our current directory inode may not yet be renamed/moved because some
3571 * ancestor (immediate or not) has to be renamed/moved first. So find if
3572 * such ancestor exists and make sure our own rename/move happens after
80aa6027
FM		 3573 * that ancestor is processed to avoid path build infinite loops (done
3574 * at get_cur_path()).
bfa7e1f8 3575 */
f959492f
FM		 3576 while (ino > BTRFS_FIRST_FREE_OBJECTID) {
3577 if (is_waiting_for_move(sctx, ino)) {
80aa6027
FM		 3578 /*
3579 * If the current inode is an ancestor of ino in the
3580 * parent root, we need to delay the rename of the
3581 * current inode, otherwise don't delayed the rename
3582 * because we can end up with a circular dependency
3583 * of renames, resulting in some directories never
3584 * getting the respective rename operations issued in
3585 * the send stream or getting into infinite path build
3586 * loops.
3587 */
3588 ret = is_ancestor(sctx->parent_root,
3589 sctx->cur_ino, sctx->cur_inode_gen,
3590 ino, path_before);
4122ea64
FM		 3591 if (ret)
3592 break;
f959492f 3593 }
bfa7e1f8
FM		 3594
3595 fs_path_reset(path_before);
3596 fs_path_reset(path_after);
3597
3598 ret = get_first_ref(sctx->send_root, ino, &parent_ino_after,
f959492f 3599 NULL, path_after);
bfa7e1f8
FM		 3600 if (ret < 0)
3601 goto out;
3602 ret = get_first_ref(sctx->parent_root, ino, &parent_ino_before,
3603 NULL, path_before);
f959492f 3604 if (ret < 0 && ret != -ENOENT) {
bfa7e1f8 3605 goto out;
f959492f 3606 } else if (ret == -ENOENT) {
bf8e8ca6 3607 ret = 0;
f959492f 3608 break;
bfa7e1f8
FM		 3609 }
3610
3611 len1 = fs_path_len(path_before);
3612 len2 = fs_path_len(path_after);
f959492f
FM		 3613 if (ino > sctx->cur_ino &&
3614 (parent_ino_before != parent_ino_after || len1 != len2 ||
3615 memcmp(path_before->start, path_after->start, len1))) {
bfa7e1f8 3616 ret = 1;
f959492f 3617 break;
bfa7e1f8 3618 }
bfa7e1f8 3619 ino = parent_ino_after;
bfa7e1f8
FM		 3620 }
3621
9f03740a
FDBM		 3622out:
3623 fs_path_free(path_before);
3624 fs_path_free(path_after);
3625
f959492f
FM		 3626 if (ret == 1) {
3627 ret = add_pending_dir_move(sctx,
3628 sctx->cur_ino,
3629 sctx->cur_inode_gen,
3630 ino,
3631 &sctx->new_refs,
84471e24 3632 &sctx->deleted_refs,
8b191a68 3633 is_orphan);
f959492f
FM		 3634 if (!ret)
3635 ret = 1;
3636 }
3637
9f03740a
FDBM		 3638 return ret;
3639}
3640
31db9f7c
AB		 3641/*
3642 * This does all the move/link/unlink/rmdir magic.
3643 */
9f03740a 3644static int process_recorded_refs(struct send_ctx *sctx, int *pending_move)
31db9f7c 3645{
04ab956e 3646 struct btrfs_fs_info *fs_info = sctx->send_root->fs_info;
31db9f7c
AB		 3647 int ret = 0;
3648 struct recorded_ref *cur;
1f4692da 3649 struct recorded_ref *cur2;
ba5e8f2e 3650 struct list_head check_dirs;
31db9f7c 3651 struct fs_path *valid_path = NULL;
b24baf69 3652 u64 ow_inode = 0;
31db9f7c
AB		 3653 u64 ow_gen;
3654 int did_overwrite = 0;
3655 int is_orphan = 0;
29d6d30f 3656 u64 last_dir_ino_rm = 0;
84471e24 3657 bool can_rename = true;
31db9f7c 3658
04ab956e 3659 btrfs_debug(fs_info, "process_recorded_refs %llu", sctx->cur_ino);
31db9f7c 3660
6d85ed05
AB		 3661 /*
3662 * This should never happen as the root dir always has the same ref
3663 * which is always '..'
3664 */
3665 BUG_ON(sctx->cur_ino <= BTRFS_FIRST_FREE_OBJECTID);
ba5e8f2e 3666 INIT_LIST_HEAD(&check_dirs);
6d85ed05 3667
924794c9 3668 valid_path = fs_path_alloc();
31db9f7c
AB		 3669 if (!valid_path) {
3670 ret = -ENOMEM;
3671 goto out;
3672 }
3673
31db9f7c
AB		 3674 /*
3675 * First, check if the first ref of the current inode was overwritten
3676 * before. If yes, we know that the current inode was already orphanized
3677 * and thus use the orphan name. If not, we can use get_cur_path to
3678 * get the path of the first ref as it would like while receiving at
3679 * this point in time.
3680 * New inodes are always orphan at the beginning, so force to use the
3681 * orphan name in this case.
3682 * The first ref is stored in valid_path and will be updated if it
3683 * gets moved around.
3684 */
3685 if (!sctx->cur_inode_new) {
3686 ret = did_overwrite_first_ref(sctx, sctx->cur_ino,
3687 sctx->cur_inode_gen);
3688 if (ret < 0)
3689 goto out;
3690 if (ret)
3691 did_overwrite = 1;
3692 }
3693 if (sctx->cur_inode_new || did_overwrite) {
3694 ret = gen_unique_name(sctx, sctx->cur_ino,
3695 sctx->cur_inode_gen, valid_path);
3696 if (ret < 0)
3697 goto out;
3698 is_orphan = 1;
3699 } else {
3700 ret = get_cur_path(sctx, sctx->cur_ino, sctx->cur_inode_gen,
3701 valid_path);
3702 if (ret < 0)
3703 goto out;
3704 }
3705
3706 list_for_each_entry(cur, &sctx->new_refs, list) {
1f4692da
AB		 3707 /*
3708 * We may have refs where the parent directory does not exist
3709 * yet. This happens if the parent directories inum is higher
3710 * the the current inum. To handle this case, we create the
3711 * parent directory out of order. But we need to check if this
3712 * did already happen before due to other refs in the same dir.
3713 */
3714 ret = get_cur_inode_state(sctx, cur->dir, cur->dir_gen);
3715 if (ret < 0)
3716 goto out;
3717 if (ret == inode_state_will_create) {
3718 ret = 0;
3719 /*
3720 * First check if any of the current inodes refs did
3721 * already create the dir.
3722 */
3723 list_for_each_entry(cur2, &sctx->new_refs, list) {
3724 if (cur == cur2)
3725 break;
3726 if (cur2->dir == cur->dir) {
3727 ret = 1;
3728 break;
3729 }
3730 }
3731
3732 /*
3733 * If that did not happen, check if a previous inode
3734 * did already create the dir.
3735 */
3736 if (!ret)
3737 ret = did_create_dir(sctx, cur->dir);
3738 if (ret < 0)
3739 goto out;
3740 if (!ret) {
3741 ret = send_create_inode(sctx, cur->dir);
3742 if (ret < 0)
3743 goto out;
3744 }
3745 }
3746
31db9f7c
AB		 3747 /*
3748 * Check if this new ref would overwrite the first ref of
3749 * another unprocessed inode. If yes, orphanize the
3750 * overwritten inode. If we find an overwritten ref that is
3751 * not the first ref, simply unlink it.
3752 */
3753 ret = will_overwrite_ref(sctx, cur->dir, cur->dir_gen,
3754 cur->name, cur->name_len,
3755 &ow_inode, &ow_gen);
3756 if (ret < 0)
3757 goto out;
3758 if (ret) {
924794c9
TI		 3759 ret = is_first_ref(sctx->parent_root,
3760 ow_inode, cur->dir, cur->name,
3761 cur->name_len);
31db9f7c
AB		 3762 if (ret < 0)
3763 goto out;
3764 if (ret) {
8996a48c 3765 struct name_cache_entry *nce;
801bec36 3766 struct waiting_dir_move *wdm;
8996a48c 3767
31db9f7c
AB		 3768 ret = orphanize_inode(sctx, ow_inode, ow_gen,
3769 cur->full_path);
3770 if (ret < 0)
3771 goto out;
801bec36
RK		 3772
3773 /*
3774 * If ow_inode has its rename operation delayed
3775 * make sure that its orphanized name is used in
3776 * the source path when performing its rename
3777 * operation.
3778 */
3779 if (is_waiting_for_move(sctx, ow_inode)) {
3780 wdm = get_waiting_dir_move(sctx,
3781 ow_inode);
3782 ASSERT(wdm);
3783 wdm->orphanized = true;
3784 }
3785
8996a48c
FM		 3786 /*
3787 * Make sure we clear our orphanized inode's
3788 * name from the name cache. This is because the
3789 * inode ow_inode might be an ancestor of some
3790 * other inode that will be orphanized as well
3791 * later and has an inode number greater than
3792 * sctx->send_progress. We need to prevent
3793 * future name lookups from using the old name
3794 * and get instead the orphan name.
3795 */
3796 nce = name_cache_search(sctx, ow_inode, ow_gen);
3797 if (nce) {
3798 name_cache_delete(sctx, nce);
3799 kfree(nce);
3800 }
801bec36
RK		 3801
3802 /*
3803 * ow_inode might currently be an ancestor of
3804 * cur_ino, therefore compute valid_path (the
3805 * current path of cur_ino) again because it
3806 * might contain the pre-orphanization name of
3807 * ow_inode, which is no longer valid.
3808 */
3809 fs_path_reset(valid_path);
3810 ret = get_cur_path(sctx, sctx->cur_ino,
3811 sctx->cur_inode_gen, valid_path);
3812 if (ret < 0)
3813 goto out;
31db9f7c
AB		 3814 } else {
3815 ret = send_unlink(sctx, cur->full_path);
3816 if (ret < 0)
3817 goto out;
3818 }
3819 }
3820
84471e24
FM		 3821 if (S_ISDIR(sctx->cur_inode_mode) && sctx->parent_root) {
3822 ret = wait_for_dest_dir_move(sctx, cur, is_orphan);
3823 if (ret < 0)
3824 goto out;
3825 if (ret == 1) {
3826 can_rename = false;
3827 *pending_move = 1;
3828 }
3829 }
3830
8b191a68
FM		 3831 if (S_ISDIR(sctx->cur_inode_mode) && sctx->parent_root &&
3832 can_rename) {
3833 ret = wait_for_parent_move(sctx, cur, is_orphan);
3834 if (ret < 0)
3835 goto out;
3836 if (ret == 1) {
3837 can_rename = false;
3838 *pending_move = 1;
3839 }
3840 }
3841
31db9f7c
AB		 3842 /*
3843 * link/move the ref to the new place. If we have an orphan
3844 * inode, move it and update valid_path. If not, link or move
3845 * it depending on the inode mode.
3846 */
84471e24 3847 if (is_orphan && can_rename) {
31db9f7c
AB		 3848 ret = send_rename(sctx, valid_path, cur->full_path);
3849 if (ret < 0)
3850 goto out;
3851 is_orphan = 0;
3852 ret = fs_path_copy(valid_path, cur->full_path);
3853 if (ret < 0)
3854 goto out;
84471e24 3855 } else if (can_rename) {
31db9f7c
AB		 3856 if (S_ISDIR(sctx->cur_inode_mode)) {
3857 /*
3858 * Dirs can't be linked, so move it. For moved
3859 * dirs, we always have one new and one deleted
3860 * ref. The deleted ref is ignored later.
3861 */
8b191a68
FM		 3862 ret = send_rename(sctx, valid_path,
3863 cur->full_path);
3864 if (!ret)
3865 ret = fs_path_copy(valid_path,
3866 cur->full_path);
31db9f7c
AB		 3867 if (ret < 0)
3868 goto out;
3869 } else {
3870 ret = send_link(sctx, cur->full_path,
3871 valid_path);
3872 if (ret < 0)
3873 goto out;
3874 }
3875 }
ba5e8f2e 3876 ret = dup_ref(cur, &check_dirs);
31db9f7c
AB		 3877 if (ret < 0)
3878 goto out;
3879 }
3880
3881 if (S_ISDIR(sctx->cur_inode_mode) && sctx->cur_inode_deleted) {
3882 /*
3883 * Check if we can already rmdir the directory. If not,
3884 * orphanize it. For every dir item inside that gets deleted
3885 * later, we do this check again and rmdir it then if possible.
3886 * See the use of check_dirs for more details.
3887 */
9dc44214
FM		 3888 ret = can_rmdir(sctx, sctx->cur_ino, sctx->cur_inode_gen,
3889 sctx->cur_ino);
31db9f7c
AB		 3890 if (ret < 0)
3891 goto out;
3892 if (ret) {
3893 ret = send_rmdir(sctx, valid_path);
3894 if (ret < 0)
3895 goto out;
3896 } else if (!is_orphan) {
3897 ret = orphanize_inode(sctx, sctx->cur_ino,
3898 sctx->cur_inode_gen, valid_path);
3899 if (ret < 0)
3900 goto out;
3901 is_orphan = 1;
3902 }
3903
3904 list_for_each_entry(cur, &sctx->deleted_refs, list) {
ba5e8f2e 3905 ret = dup_ref(cur, &check_dirs);
31db9f7c
AB		 3906 if (ret < 0)
3907 goto out;
3908 }
ccf1626b
AB		 3909 } else if (S_ISDIR(sctx->cur_inode_mode) &&
3910 !list_empty(&sctx->deleted_refs)) {
3911 /*
3912 * We have a moved dir. Add the old parent to check_dirs
3913 */
3914 cur = list_entry(sctx->deleted_refs.next, struct recorded_ref,
3915 list);
ba5e8f2e 3916 ret = dup_ref(cur, &check_dirs);
ccf1626b
AB		 3917 if (ret < 0)
3918 goto out;
31db9f7c
AB		 3919 } else if (!S_ISDIR(sctx->cur_inode_mode)) {
3920 /*
3921 * We have a non dir inode. Go through all deleted refs and
3922 * unlink them if they were not already overwritten by other
3923 * inodes.
3924 */
3925 list_for_each_entry(cur, &sctx->deleted_refs, list) {
3926 ret = did_overwrite_ref(sctx, cur->dir, cur->dir_gen,
3927 sctx->cur_ino, sctx->cur_inode_gen,
3928 cur->name, cur->name_len);
3929 if (ret < 0)
3930 goto out;
3931 if (!ret) {
1f4692da
AB		 3932 ret = send_unlink(sctx, cur->full_path);
3933 if (ret < 0)
3934 goto out;
31db9f7c 3935 }
ba5e8f2e 3936 ret = dup_ref(cur, &check_dirs);
31db9f7c
AB		 3937 if (ret < 0)
3938 goto out;
3939 }
31db9f7c
AB		 3940 /*
3941 * If the inode is still orphan, unlink the orphan. This may
3942 * happen when a previous inode did overwrite the first ref
3943 * of this inode and no new refs were added for the current
766702ef
AB		 3944 * inode. Unlinking does not mean that the inode is deleted in
3945 * all cases. There may still be links to this inode in other
3946 * places.
31db9f7c 3947 */
1f4692da 3948 if (is_orphan) {
31db9f7c
AB		 3949 ret = send_unlink(sctx, valid_path);
3950 if (ret < 0)
3951 goto out;
3952 }
3953 }
3954
3955 /*
3956 * We did collect all parent dirs where cur_inode was once located. We
3957 * now go through all these dirs and check if they are pending for
3958 * deletion and if it's finally possible to perform the rmdir now.
3959 * We also update the inode stats of the parent dirs here.
3960 */
ba5e8f2e 3961 list_for_each_entry(cur, &check_dirs, list) {
766702ef
AB		 3962 /*
3963 * In case we had refs into dirs that were not processed yet,
3964 * we don't need to do the utime and rmdir logic for these dirs.
3965 * The dir will be processed later.
3966 */
ba5e8f2e 3967 if (cur->dir > sctx->cur_ino)
31db9f7c
AB		 3968 continue;
3969
ba5e8f2e 3970 ret = get_cur_inode_state(sctx, cur->dir, cur->dir_gen);
31db9f7c
AB		 3971 if (ret < 0)
3972 goto out;
3973
3974 if (ret == inode_state_did_create ||
3975 ret == inode_state_no_change) {
3976 /* TODO delayed utimes */
ba5e8f2e 3977 ret = send_utimes(sctx, cur->dir, cur->dir_gen);
31db9f7c
AB		 3978 if (ret < 0)
3979 goto out;
29d6d30f
FM		 3980 } else if (ret == inode_state_did_delete &&
3981 cur->dir != last_dir_ino_rm) {
9dc44214
FM		 3982 ret = can_rmdir(sctx, cur->dir, cur->dir_gen,
3983 sctx->cur_ino);
31db9f7c
AB		 3984 if (ret < 0)
3985 goto out;
3986 if (ret) {
ba5e8f2e
JB		 3987 ret = get_cur_path(sctx, cur->dir,
3988 cur->dir_gen, valid_path);
31db9f7c
AB		 3989 if (ret < 0)
3990 goto out;
3991 ret = send_rmdir(sctx, valid_path);
3992 if (ret < 0)
3993 goto out;
29d6d30f 3994 last_dir_ino_rm = cur->dir;
31db9f7c
AB		 3995 }
3996 }
3997 }
3998
31db9f7c
AB		 3999 ret = 0;
4000
4001out:
ba5e8f2e 4002 __free_recorded_refs(&check_dirs);
31db9f7c 4003 free_recorded_refs(sctx);
924794c9 4004 fs_path_free(valid_path);
31db9f7c
AB		 4005 return ret;
4006}
4007
a4d96d62
LB		 4008static int record_ref(struct btrfs_root *root, int num, u64 dir, int index,
4009 struct fs_path *name, void *ctx, struct list_head *refs)
31db9f7c
AB		 4010{
4011 int ret = 0;
4012 struct send_ctx *sctx = ctx;
4013 struct fs_path *p;
4014 u64 gen;
4015
924794c9 4016 p = fs_path_alloc();
31db9f7c
AB		 4017 if (!p)
4018 return -ENOMEM;
4019
a4d96d62 4020 ret = get_inode_info(root, dir, NULL, &gen, NULL, NULL,
85a7b33b 4021 NULL, NULL);
31db9f7c
AB		 4022 if (ret < 0)
4023 goto out;
4024
31db9f7c
AB		 4025 ret = get_cur_path(sctx, dir, gen, p);
4026 if (ret < 0)
4027 goto out;
4028 ret = fs_path_add_path(p, name);
4029 if (ret < 0)
4030 goto out;
4031
a4d96d62 4032 ret = __record_ref(refs, dir, gen, p);
31db9f7c
AB		 4033
4034out:
4035 if (ret)
924794c9 4036 fs_path_free(p);
31db9f7c
AB		 4037 return ret;
4038}
4039
a4d96d62
LB		 4040static int __record_new_ref(int num, u64 dir, int index,
4041 struct fs_path *name,
4042 void *ctx)
4043{
4044 struct send_ctx *sctx = ctx;
4045 return record_ref(sctx->send_root, num, dir, index, name,
4046 ctx, &sctx->new_refs);
4047}
4048
4049
31db9f7c
AB		 4050static int __record_deleted_ref(int num, u64 dir, int index,
4051 struct fs_path *name,
4052 void *ctx)
4053{
31db9f7c 4054 struct send_ctx *sctx = ctx;
a4d96d62
LB		 4055 return record_ref(sctx->parent_root, num, dir, index, name,
4056 ctx, &sctx->deleted_refs);
31db9f7c
AB		 4057}
4058
4059static int record_new_ref(struct send_ctx *sctx)
4060{
4061 int ret;
4062
924794c9
TI		 4063 ret = iterate_inode_ref(sctx->send_root, sctx->left_path,
4064 sctx->cmp_key, 0, __record_new_ref, sctx);
31db9f7c
AB		 4065 if (ret < 0)
4066 goto out;
4067 ret = 0;
4068
4069out:
4070 return ret;
4071}
4072
4073static int record_deleted_ref(struct send_ctx *sctx)
4074{
4075 int ret;
4076
924794c9
TI		 4077 ret = iterate_inode_ref(sctx->parent_root, sctx->right_path,
4078 sctx->cmp_key, 0, __record_deleted_ref, sctx);
31db9f7c
AB		 4079 if (ret < 0)
4080 goto out;
4081 ret = 0;
4082
4083out:
4084 return ret;
4085}
4086
4087struct find_ref_ctx {
4088 u64 dir;
ba5e8f2e
JB		 4089 u64 dir_gen;
4090 struct btrfs_root *root;
31db9f7c
AB		 4091 struct fs_path *name;
4092 int found_idx;
4093};
4094
4095static int __find_iref(int num, u64 dir, int index,
4096 struct fs_path *name,
4097 void *ctx_)
4098{
4099 struct find_ref_ctx *ctx = ctx_;
ba5e8f2e
JB		 4100 u64 dir_gen;
4101 int ret;
31db9f7c
AB		 4102
4103 if (dir == ctx->dir && fs_path_len(name) == fs_path_len(ctx->name) &&
4104 strncmp(name->start, ctx->name->start, fs_path_len(name)) == 0) {
ba5e8f2e
JB		 4105 /*
4106 * To avoid doing extra lookups we'll only do this if everything
4107 * else matches.
4108 */
4109 ret = get_inode_info(ctx->root, dir, NULL, &dir_gen, NULL,
4110 NULL, NULL, NULL);
4111 if (ret)
4112 return ret;
4113 if (dir_gen != ctx->dir_gen)
4114 return 0;
31db9f7c
AB		 4115 ctx->found_idx = num;
4116 return 1;
4117 }
4118 return 0;
4119}
4120
924794c9 4121static int find_iref(struct btrfs_root *root,
31db9f7c
AB		 4122 struct btrfs_path *path,
4123 struct btrfs_key *key,
ba5e8f2e 4124 u64 dir, u64 dir_gen, struct fs_path *name)
31db9f7c
AB		 4125{
4126 int ret;
4127 struct find_ref_ctx ctx;
4128
4129 ctx.dir = dir;
4130 ctx.name = name;
ba5e8f2e 4131 ctx.dir_gen = dir_gen;
31db9f7c 4132 ctx.found_idx = -1;
ba5e8f2e 4133 ctx.root = root;
31db9f7c 4134
924794c9 4135 ret = iterate_inode_ref(root, path, key, 0, __find_iref, &ctx);
31db9f7c
AB		 4136 if (ret < 0)
4137 return ret;
4138
4139 if (ctx.found_idx == -1)
4140 return -ENOENT;
4141
4142 return ctx.found_idx;
4143}
4144
4145static int __record_changed_new_ref(int num, u64 dir, int index,
4146 struct fs_path *name,
4147 void *ctx)
4148{
ba5e8f2e 4149 u64 dir_gen;
31db9f7c
AB		 4150 int ret;
4151 struct send_ctx *sctx = ctx;
4152
ba5e8f2e
JB		 4153 ret = get_inode_info(sctx->send_root, dir, NULL, &dir_gen, NULL,
4154 NULL, NULL, NULL);
4155 if (ret)
4156 return ret;
4157
924794c9 4158 ret = find_iref(sctx->parent_root, sctx->right_path,
ba5e8f2e 4159 sctx->cmp_key, dir, dir_gen, name);
31db9f7c
AB		 4160 if (ret == -ENOENT)
4161 ret = __record_new_ref(num, dir, index, name, sctx);
4162 else if (ret > 0)
4163 ret = 0;
4164
4165 return ret;
4166}
4167
4168static int __record_changed_deleted_ref(int num, u64 dir, int index,
4169 struct fs_path *name,
4170 void *ctx)
4171{
ba5e8f2e 4172 u64 dir_gen;
31db9f7c
AB		 4173 int ret;
4174 struct send_ctx *sctx = ctx;
4175
ba5e8f2e
JB		 4176 ret = get_inode_info(sctx->parent_root, dir, NULL, &dir_gen, NULL,
4177 NULL, NULL, NULL);
4178 if (ret)
4179 return ret;
4180
924794c9 4181 ret = find_iref(sctx->send_root, sctx->left_path, sctx->cmp_key,
ba5e8f2e 4182 dir, dir_gen, name);
31db9f7c
AB		 4183 if (ret == -ENOENT)
4184 ret = __record_deleted_ref(num, dir, index, name, sctx);
4185 else if (ret > 0)
4186 ret = 0;
4187
4188 return ret;
4189}
4190
4191static int record_changed_ref(struct send_ctx *sctx)
4192{
4193 int ret = 0;
4194
924794c9 4195 ret = iterate_inode_ref(sctx->send_root, sctx->left_path,
31db9f7c
AB		 4196 sctx->cmp_key, 0, __record_changed_new_ref, sctx);
4197 if (ret < 0)
4198 goto out;
924794c9 4199 ret = iterate_inode_ref(sctx->parent_root, sctx->right_path,
31db9f7c
AB		 4200 sctx->cmp_key, 0, __record_changed_deleted_ref, sctx);
4201 if (ret < 0)
4202 goto out;
4203 ret = 0;
4204
4205out:
4206 return ret;
4207}
4208
4209/*
4210 * Record and process all refs at once. Needed when an inode changes the
4211 * generation number, which means that it was deleted and recreated.
4212 */
4213static int process_all_refs(struct send_ctx *sctx,
4214 enum btrfs_compare_tree_result cmd)
4215{
4216 int ret;
4217 struct btrfs_root *root;
4218 struct btrfs_path *path;
4219 struct btrfs_key key;
4220 struct btrfs_key found_key;
4221 struct extent_buffer *eb;
4222 int slot;
4223 iterate_inode_ref_t cb;
9f03740a 4224 int pending_move = 0;
31db9f7c
AB		 4225
4226 path = alloc_path_for_send();
4227 if (!path)
4228 return -ENOMEM;
4229
4230 if (cmd == BTRFS_COMPARE_TREE_NEW) {
4231 root = sctx->send_root;
4232 cb = __record_new_ref;
4233 } else if (cmd == BTRFS_COMPARE_TREE_DELETED) {
4234 root = sctx->parent_root;
4235 cb = __record_deleted_ref;
4236 } else {
4d1a63b2
DS		 4237 btrfs_err(sctx->send_root->fs_info,
4238 "Wrong command %d in process_all_refs", cmd);
4239 ret = -EINVAL;
4240 goto out;
31db9f7c
AB		 4241 }
4242
4243 key.objectid = sctx->cmp_key->objectid;
4244 key.type = BTRFS_INODE_REF_KEY;
4245 key.offset = 0;
dff6d0ad
FDBM		 4246 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
4247 if (ret < 0)
4248 goto out;
31db9f7c 4249
dff6d0ad 4250 while (1) {
31db9f7c
AB		 4251 eb = path->nodes[0];
4252 slot = path->slots[0];
dff6d0ad
FDBM		 4253 if (slot >= btrfs_header_nritems(eb)) {
4254 ret = btrfs_next_leaf(root, path);
4255 if (ret < 0)
4256 goto out;
4257 else if (ret > 0)
4258 break;
4259 continue;
4260 }
4261
31db9f7c
AB		 4262 btrfs_item_key_to_cpu(eb, &found_key, slot);
4263
4264 if (found_key.objectid != key.objectid ||
96b5bd77
JS		 4265 (found_key.type != BTRFS_INODE_REF_KEY &&
4266 found_key.type != BTRFS_INODE_EXTREF_KEY))
31db9f7c 4267 break;
31db9f7c 4268
924794c9 4269 ret = iterate_inode_ref(root, path, &found_key, 0, cb, sctx);
31db9f7c
AB		 4270 if (ret < 0)
4271 goto out;
4272
dff6d0ad 4273 path->slots[0]++;
31db9f7c 4274 }
e938c8ad 4275 btrfs_release_path(path);
31db9f7c 4276
3dc09ec8
JB		 4277 /*
4278 * We don't actually care about pending_move as we are simply
4279 * re-creating this inode and will be rename'ing it into place once we
4280 * rename the parent directory.
4281 */
9f03740a 4282 ret = process_recorded_refs(sctx, &pending_move);
31db9f7c
AB		 4283out:
4284 btrfs_free_path(path);
4285 return ret;
4286}
4287
4288static int send_set_xattr(struct send_ctx *sctx,
4289 struct fs_path *path,
4290 const char *name, int name_len,
4291 const char *data, int data_len)
4292{
4293 int ret = 0;
4294
4295 ret = begin_cmd(sctx, BTRFS_SEND_C_SET_XATTR);
4296 if (ret < 0)
4297 goto out;
4298
4299 TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, path);
4300 TLV_PUT_STRING(sctx, BTRFS_SEND_A_XATTR_NAME, name, name_len);
4301 TLV_PUT(sctx, BTRFS_SEND_A_XATTR_DATA, data, data_len);
4302
4303 ret = send_cmd(sctx);
4304
4305tlv_put_failure:
4306out:
4307 return ret;
4308}
4309
4310static int send_remove_xattr(struct send_ctx *sctx,
4311 struct fs_path *path,
4312 const char *name, int name_len)
4313{
4314 int ret = 0;
4315
4316 ret = begin_cmd(sctx, BTRFS_SEND_C_REMOVE_XATTR);
4317 if (ret < 0)
4318 goto out;
4319
4320 TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, path);
4321 TLV_PUT_STRING(sctx, BTRFS_SEND_A_XATTR_NAME, name, name_len);
4322
4323 ret = send_cmd(sctx);
4324
4325tlv_put_failure:
4326out:
4327 return ret;
4328}
4329
4330static int __process_new_xattr(int num, struct btrfs_key *di_key,
4331 const char *name, int name_len,
4332 const char *data, int data_len,
4333 u8 type, void *ctx)
4334{
4335 int ret;
4336 struct send_ctx *sctx = ctx;
4337 struct fs_path *p;
2211d5ba 4338 struct posix_acl_xattr_header dummy_acl;
31db9f7c 4339
924794c9 4340 p = fs_path_alloc();
31db9f7c
AB		 4341 if (!p)
4342 return -ENOMEM;
4343
4344 /*
01327610 4345 * This hack is needed because empty acls are stored as zero byte
31db9f7c 4346 * data in xattrs. Problem with that is, that receiving these zero byte
01327610 4347 * acls will fail later. To fix this, we send a dummy acl list that
31db9f7c
AB		 4348 * only contains the version number and no entries.
4349 */
4350 if (!strncmp(name, XATTR_NAME_POSIX_ACL_ACCESS, name_len) ||
4351 !strncmp(name, XATTR_NAME_POSIX_ACL_DEFAULT, name_len)) {
4352 if (data_len == 0) {
4353 dummy_acl.a_version =
4354 cpu_to_le32(POSIX_ACL_XATTR_VERSION);
4355 data = (char *)&dummy_acl;
4356 data_len = sizeof(dummy_acl);
4357 }
4358 }
4359
4360 ret = get_cur_path(sctx, sctx->cur_ino, sctx->cur_inode_gen, p);
4361 if (ret < 0)
4362 goto out;
4363
4364 ret = send_set_xattr(sctx, p, name, name_len, data, data_len);
4365
4366out:
924794c9 4367 fs_path_free(p);
31db9f7c
AB		 4368 return ret;
4369}
4370
4371static int __process_deleted_xattr(int num, struct btrfs_key *di_key,
4372 const char *name, int name_len,
4373 const char *data, int data_len,
4374 u8 type, void *ctx)
4375{
4376 int ret;
4377 struct send_ctx *sctx = ctx;
4378 struct fs_path *p;
4379
924794c9 4380 p = fs_path_alloc();
31db9f7c
AB		 4381 if (!p)
4382 return -ENOMEM;
4383
4384 ret = get_cur_path(sctx, sctx->cur_ino, sctx->cur_inode_gen, p);
4385 if (ret < 0)
4386 goto out;
4387
4388 ret = send_remove_xattr(sctx, p, name, name_len);
4389
4390out:
924794c9 4391 fs_path_free(p);
31db9f7c
AB		 4392 return ret;
4393}
4394
4395static int process_new_xattr(struct send_ctx *sctx)
4396{
4397 int ret = 0;
4398
924794c9
TI		 4399 ret = iterate_dir_item(sctx->send_root, sctx->left_path,
4400 sctx->cmp_key, __process_new_xattr, sctx);
31db9f7c
AB		 4401
4402 return ret;
4403}
4404
4405static int process_deleted_xattr(struct send_ctx *sctx)
4406{
e2c89907
MY		 4407 return iterate_dir_item(sctx->parent_root, sctx->right_path,
4408 sctx->cmp_key, __process_deleted_xattr, sctx);
31db9f7c
AB		 4409}
4410
4411struct find_xattr_ctx {
4412 const char *name;
4413 int name_len;
4414 int found_idx;
4415 char *found_data;
4416 int found_data_len;
4417};
4418
4419static int __find_xattr(int num, struct btrfs_key *di_key,
4420 const char *name, int name_len,
4421 const char *data, int data_len,
4422 u8 type, void *vctx)
4423{
4424 struct find_xattr_ctx *ctx = vctx;
4425
4426 if (name_len == ctx->name_len &&
4427 strncmp(name, ctx->name, name_len) == 0) {
4428 ctx->found_idx = num;
4429 ctx->found_data_len = data_len;
e780b0d1 4430 ctx->found_data = kmemdup(data, data_len, GFP_KERNEL);
31db9f7c
AB		 4431 if (!ctx->found_data)
4432 return -ENOMEM;
31db9f7c
AB		 4433 return 1;
4434 }
4435 return 0;
4436}
4437
924794c9 4438static int find_xattr(struct btrfs_root *root,
31db9f7c
AB		 4439 struct btrfs_path *path,
4440 struct btrfs_key *key,
4441 const char *name, int name_len,
4442 char **data, int *data_len)
4443{
4444 int ret;
4445 struct find_xattr_ctx ctx;
4446
4447 ctx.name = name;
4448 ctx.name_len = name_len;
4449 ctx.found_idx = -1;
4450 ctx.found_data = NULL;
4451 ctx.found_data_len = 0;
4452
924794c9 4453 ret = iterate_dir_item(root, path, key, __find_xattr, &ctx);
31db9f7c
AB		 4454 if (ret < 0)
4455 return ret;
4456
4457 if (ctx.found_idx == -1)
4458 return -ENOENT;
4459 if (data) {
4460 *data = ctx.found_data;
4461 *data_len = ctx.found_data_len;
4462 } else {
4463 kfree(ctx.found_data);
4464 }
4465 return ctx.found_idx;
4466}
4467
4468
4469static int __process_changed_new_xattr(int num, struct btrfs_key *di_key,
4470 const char *name, int name_len,
4471 const char *data, int data_len,
4472 u8 type, void *ctx)
4473{
4474 int ret;
4475 struct send_ctx *sctx = ctx;
4476 char *found_data = NULL;
4477 int found_data_len = 0;
31db9f7c 4478
924794c9
TI		 4479 ret = find_xattr(sctx->parent_root, sctx->right_path,
4480 sctx->cmp_key, name, name_len, &found_data,
4481 &found_data_len);
31db9f7c
AB		 4482 if (ret == -ENOENT) {
4483 ret = __process_new_xattr(num, di_key, name, name_len, data,
4484 data_len, type, ctx);
4485 } else if (ret >= 0) {
4486 if (data_len != found_data_len ||
4487 memcmp(data, found_data, data_len)) {
4488 ret = __process_new_xattr(num, di_key, name, name_len,
4489 data, data_len, type, ctx);
4490 } else {
4491 ret = 0;
4492 }
4493 }
4494
4495 kfree(found_data);
31db9f7c
AB		 4496 return ret;
4497}
4498
4499static int __process_changed_deleted_xattr(int num, struct btrfs_key *di_key,
4500 const char *name, int name_len,
4501 const char *data, int data_len,
4502 u8 type, void *ctx)
4503{
4504 int ret;
4505 struct send_ctx *sctx = ctx;
4506
924794c9
TI		 4507 ret = find_xattr(sctx->send_root, sctx->left_path, sctx->cmp_key,
4508 name, name_len, NULL, NULL);
31db9f7c
AB		 4509 if (ret == -ENOENT)
4510 ret = __process_deleted_xattr(num, di_key, name, name_len, data,
4511 data_len, type, ctx);
4512 else if (ret >= 0)
4513 ret = 0;
4514
4515 return ret;
4516}
4517
4518static int process_changed_xattr(struct send_ctx *sctx)
4519{
4520 int ret = 0;
4521
924794c9 4522 ret = iterate_dir_item(sctx->send_root, sctx->left_path,
31db9f7c
AB		 4523 sctx->cmp_key, __process_changed_new_xattr, sctx);
4524 if (ret < 0)
4525 goto out;
924794c9 4526 ret = iterate_dir_item(sctx->parent_root, sctx->right_path,
31db9f7c
AB		 4527 sctx->cmp_key, __process_changed_deleted_xattr, sctx);
4528
4529out:
4530 return ret;
4531}
4532
4533static int process_all_new_xattrs(struct send_ctx *sctx)
4534{
4535 int ret;
4536 struct btrfs_root *root;
4537 struct btrfs_path *path;
4538 struct btrfs_key key;
4539 struct btrfs_key found_key;
4540 struct extent_buffer *eb;
4541 int slot;
4542
4543 path = alloc_path_for_send();
4544 if (!path)
4545 return -ENOMEM;
4546
4547 root = sctx->send_root;
4548
4549 key.objectid = sctx->cmp_key->objectid;
4550 key.type = BTRFS_XATTR_ITEM_KEY;
4551 key.offset = 0;
dff6d0ad
FDBM		 4552 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
4553 if (ret < 0)
4554 goto out;
31db9f7c 4555
dff6d0ad 4556 while (1) {
31db9f7c
AB		 4557 eb = path->nodes[0];
4558 slot = path->slots[0];
dff6d0ad
FDBM		 4559 if (slot >= btrfs_header_nritems(eb)) {
4560 ret = btrfs_next_leaf(root, path);
4561 if (ret < 0) {
4562 goto out;
4563 } else if (ret > 0) {
4564 ret = 0;
4565 break;
4566 }
4567 continue;
4568 }
31db9f7c 4569
dff6d0ad 4570 btrfs_item_key_to_cpu(eb, &found_key, slot);
31db9f7c
AB		 4571 if (found_key.objectid != key.objectid ||
4572 found_key.type != key.type) {
4573 ret = 0;
4574 goto out;
4575 }
4576
924794c9
TI		 4577 ret = iterate_dir_item(root, path, &found_key,
4578 __process_new_xattr, sctx);
31db9f7c
AB		 4579 if (ret < 0)
4580 goto out;
4581
dff6d0ad 4582 path->slots[0]++;
31db9f7c
AB		 4583 }
4584
4585out:
4586 btrfs_free_path(path);
4587 return ret;
4588}
4589
ed259095
JB		 4590static ssize_t fill_read_buf(struct send_ctx *sctx, u64 offset, u32 len)
4591{
4592 struct btrfs_root *root = sctx->send_root;
4593 struct btrfs_fs_info *fs_info = root->fs_info;
4594 struct inode *inode;
4595 struct page *page;
4596 char *addr;
4597 struct btrfs_key key;
09cbfeaf 4598 pgoff_t index = offset >> PAGE_SHIFT;
ed259095 4599 pgoff_t last_index;
09cbfeaf 4600 unsigned pg_offset = offset & ~PAGE_MASK;
ed259095
JB		 4601 ssize_t ret = 0;
4602
4603 key.objectid = sctx->cur_ino;
4604 key.type = BTRFS_INODE_ITEM_KEY;
4605 key.offset = 0;
4606
4607 inode = btrfs_iget(fs_info->sb, &key, root, NULL);
4608 if (IS_ERR(inode))
4609 return PTR_ERR(inode);
4610
4611 if (offset + len > i_size_read(inode)) {
4612 if (offset > i_size_read(inode))
4613 len = 0;
4614 else
4615 len = offset - i_size_read(inode);
4616 }
4617 if (len == 0)
4618 goto out;
4619
09cbfeaf 4620 last_index = (offset + len - 1) >> PAGE_SHIFT;
2131bcd3
LB		 4621
4622 /* initial readahead */
4623 memset(&sctx->ra, 0, sizeof(struct file_ra_state));
4624 file_ra_state_init(&sctx->ra, inode->i_mapping);
4625 btrfs_force_ra(inode->i_mapping, &sctx->ra, NULL, index,
4626 last_index - index + 1);
4627
ed259095
JB		 4628 while (index <= last_index) {
4629 unsigned cur_len = min_t(unsigned, len,
09cbfeaf 4630 PAGE_SIZE - pg_offset);
e780b0d1 4631 page = find_or_create_page(inode->i_mapping, index, GFP_KERNEL);
ed259095
JB		 4632 if (!page) {
4633 ret = -ENOMEM;
4634 break;
4635 }
4636
4637 if (!PageUptodate(page)) {
4638 btrfs_readpage(NULL, page);
4639 lock_page(page);
4640 if (!PageUptodate(page)) {
4641 unlock_page(page);
09cbfeaf 4642 put_page(page);
ed259095
JB		 4643 ret = -EIO;
4644 break;
4645 }
4646 }
4647
4648 addr = kmap(page);
4649 memcpy(sctx->read_buf + ret, addr + pg_offset, cur_len);
4650 kunmap(page);
4651 unlock_page(page);
09cbfeaf 4652 put_page(page);
ed259095
JB		 4653 index++;
4654 pg_offset = 0;
4655 len -= cur_len;
4656 ret += cur_len;
4657 }
4658out:
4659 iput(inode);
4660 return ret;
4661}
4662
31db9f7c
AB		 4663/*
4664 * Read some bytes from the current inode/file and send a write command to
4665 * user space.
4666 */
4667static int send_write(struct send_ctx *sctx, u64 offset, u32 len)
4668{
04ab956e 4669 struct btrfs_fs_info *fs_info = sctx->send_root->fs_info;
31db9f7c
AB		 4670 int ret = 0;
4671 struct fs_path *p;
ed259095 4672 ssize_t num_read = 0;
31db9f7c 4673
924794c9 4674 p = fs_path_alloc();
31db9f7c
AB		 4675 if (!p)
4676 return -ENOMEM;
4677
04ab956e 4678 btrfs_debug(fs_info, "send_write offset=%llu, len=%d", offset, len);
31db9f7c 4679
ed259095
JB		 4680 num_read = fill_read_buf(sctx, offset, len);
4681 if (num_read <= 0) {
4682 if (num_read < 0)
4683 ret = num_read;
31db9f7c 4684 goto out;
ed259095 4685 }
31db9f7c
AB		 4686
4687 ret = begin_cmd(sctx, BTRFS_SEND_C_WRITE);
4688 if (ret < 0)
4689 goto out;
4690
4691 ret = get_cur_path(sctx, sctx->cur_ino, sctx->cur_inode_gen, p);
4692 if (ret < 0)
4693 goto out;
4694
4695 TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p);
4696 TLV_PUT_U64(sctx, BTRFS_SEND_A_FILE_OFFSET, offset);
e938c8ad 4697 TLV_PUT(sctx, BTRFS_SEND_A_DATA, sctx->read_buf, num_read);
31db9f7c
AB		 4698
4699 ret = send_cmd(sctx);
4700
4701tlv_put_failure:
4702out:
924794c9 4703 fs_path_free(p);
31db9f7c
AB		 4704 if (ret < 0)
4705 return ret;
e938c8ad 4706 return num_read;
31db9f7c
AB		 4707}
4708
4709/*
4710 * Send a clone command to user space.
4711 */
4712static int send_clone(struct send_ctx *sctx,
4713 u64 offset, u32 len,
4714 struct clone_root *clone_root)
4715{
4716 int ret = 0;
31db9f7c
AB		 4717 struct fs_path *p;
4718 u64 gen;
4719
04ab956e
JM		 4720 btrfs_debug(sctx->send_root->fs_info,
4721 "send_clone offset=%llu, len=%d, clone_root=%llu, clone_inode=%llu, clone_offset=%llu",
4722 offset, len, clone_root->root->objectid, clone_root->ino,
4723 clone_root->offset);
31db9f7c 4724
924794c9 4725 p = fs_path_alloc();
31db9f7c
AB		 4726 if (!p)
4727 return -ENOMEM;
4728
4729 ret = begin_cmd(sctx, BTRFS_SEND_C_CLONE);
4730 if (ret < 0)
4731 goto out;
4732
4733 ret = get_cur_path(sctx, sctx->cur_ino, sctx->cur_inode_gen, p);
4734 if (ret < 0)
4735 goto out;
4736
4737 TLV_PUT_U64(sctx, BTRFS_SEND_A_FILE_OFFSET, offset);
4738 TLV_PUT_U64(sctx, BTRFS_SEND_A_CLONE_LEN, len);
4739 TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p);
4740
e938c8ad 4741 if (clone_root->root == sctx->send_root) {
31db9f7c 4742 ret = get_inode_info(sctx->send_root, clone_root->ino, NULL,
85a7b33b 4743 &gen, NULL, NULL, NULL, NULL);
31db9f7c
AB		 4744 if (ret < 0)
4745 goto out;
4746 ret = get_cur_path(sctx, clone_root->ino, gen, p);
4747 } else {
924794c9 4748 ret = get_inode_path(clone_root->root, clone_root->ino, p);
31db9f7c
AB		 4749 }
4750 if (ret < 0)
4751 goto out;
4752
37b8d27d
JB		 4753 /*
4754 * If the parent we're using has a received_uuid set then use that as
4755 * our clone source as that is what we will look for when doing a
4756 * receive.
4757 *
4758 * This covers the case that we create a snapshot off of a received
4759 * subvolume and then use that as the parent and try to receive on a
4760 * different host.
4761 */
4762 if (!btrfs_is_empty_uuid(clone_root->root->root_item.received_uuid))
4763 TLV_PUT_UUID(sctx, BTRFS_SEND_A_CLONE_UUID,
4764 clone_root->root->root_item.received_uuid);
4765 else
4766 TLV_PUT_UUID(sctx, BTRFS_SEND_A_CLONE_UUID,
4767 clone_root->root->root_item.uuid);
31db9f7c 4768 TLV_PUT_U64(sctx, BTRFS_SEND_A_CLONE_CTRANSID,
5a0f4e2c 4769 le64_to_cpu(clone_root->root->root_item.ctransid));
31db9f7c
AB		 4770 TLV_PUT_PATH(sctx, BTRFS_SEND_A_CLONE_PATH, p);
4771 TLV_PUT_U64(sctx, BTRFS_SEND_A_CLONE_OFFSET,
4772 clone_root->offset);
4773
4774 ret = send_cmd(sctx);
4775
4776tlv_put_failure:
4777out:
924794c9 4778 fs_path_free(p);
31db9f7c
AB		 4779 return ret;
4780}
4781
cb95e7bf
MF		 4782/*
4783 * Send an update extent command to user space.
4784 */
4785static int send_update_extent(struct send_ctx *sctx,
4786 u64 offset, u32 len)
4787{
4788 int ret = 0;
4789 struct fs_path *p;
4790
924794c9 4791 p = fs_path_alloc();
cb95e7bf
MF		 4792 if (!p)
4793 return -ENOMEM;
4794
4795 ret = begin_cmd(sctx, BTRFS_SEND_C_UPDATE_EXTENT);
4796 if (ret < 0)
4797 goto out;
4798
4799 ret = get_cur_path(sctx, sctx->cur_ino, sctx->cur_inode_gen, p);
4800 if (ret < 0)
4801 goto out;
4802
4803 TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p);
4804 TLV_PUT_U64(sctx, BTRFS_SEND_A_FILE_OFFSET, offset);
4805 TLV_PUT_U64(sctx, BTRFS_SEND_A_SIZE, len);
4806
4807 ret = send_cmd(sctx);
4808
4809tlv_put_failure:
4810out:
924794c9 4811 fs_path_free(p);
cb95e7bf
MF		 4812 return ret;
4813}
4814
16e7549f
JB		 4815static int send_hole(struct send_ctx *sctx, u64 end)
4816{
4817 struct fs_path *p = NULL;
4818 u64 offset = sctx->cur_inode_last_extent;
4819 u64 len;
4820 int ret = 0;
4821
4822 p = fs_path_alloc();
4823 if (!p)
4824 return -ENOMEM;
c715e155
FM		 4825 ret = get_cur_path(sctx, sctx->cur_ino, sctx->cur_inode_gen, p);
4826 if (ret < 0)
4827 goto tlv_put_failure;
16e7549f
JB		 4828 memset(sctx->read_buf, 0, BTRFS_SEND_READ_SIZE);
4829 while (offset < end) {
4830 len = min_t(u64, end - offset, BTRFS_SEND_READ_SIZE);
4831
4832 ret = begin_cmd(sctx, BTRFS_SEND_C_WRITE);
16e7549f
JB		 4833 if (ret < 0)
4834 break;
4835 TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p);
4836 TLV_PUT_U64(sctx, BTRFS_SEND_A_FILE_OFFSET, offset);
4837 TLV_PUT(sctx, BTRFS_SEND_A_DATA, sctx->read_buf, len);
4838 ret = send_cmd(sctx);
4839 if (ret < 0)
4840 break;
4841 offset += len;
4842 }
4843tlv_put_failure:
4844 fs_path_free(p);
4845 return ret;
4846}
4847
d906d49f
FM		 4848static int send_extent_data(struct send_ctx *sctx,
4849 const u64 offset,
4850 const u64 len)
4851{
4852 u64 sent = 0;
4853
4854 if (sctx->flags & BTRFS_SEND_FLAG_NO_FILE_DATA)
4855 return send_update_extent(sctx, offset, len);
4856
4857 while (sent < len) {
4858 u64 size = len - sent;
4859 int ret;
4860
4861 if (size > BTRFS_SEND_READ_SIZE)
4862 size = BTRFS_SEND_READ_SIZE;
4863 ret = send_write(sctx, offset + sent, size);
4864 if (ret < 0)
4865 return ret;
4866 if (!ret)
4867 break;
4868 sent += ret;
4869 }
4870 return 0;
4871}
4872
4873static int clone_range(struct send_ctx *sctx,
4874 struct clone_root *clone_root,
4875 const u64 disk_byte,
4876 u64 data_offset,
4877 u64 offset,
4878 u64 len)
4879{
4880 struct btrfs_path *path;
4881 struct btrfs_key key;
4882 int ret;
4883
4884 path = alloc_path_for_send();
4885 if (!path)
4886 return -ENOMEM;
4887
4888 /*
4889 * We can't send a clone operation for the entire range if we find
4890 * extent items in the respective range in the source file that
4891 * refer to different extents or if we find holes.
4892 * So check for that and do a mix of clone and regular write/copy
4893 * operations if needed.
4894 *
4895 * Example:
4896 *
4897 * mkfs.btrfs -f /dev/sda
4898 * mount /dev/sda /mnt
4899 * xfs_io -f -c "pwrite -S 0xaa 0K 100K" /mnt/foo
4900 * cp --reflink=always /mnt/foo /mnt/bar
4901 * xfs_io -c "pwrite -S 0xbb 50K 50K" /mnt/foo
4902 * btrfs subvolume snapshot -r /mnt /mnt/snap
4903 *
4904 * If when we send the snapshot and we are processing file bar (which
4905 * has a higher inode number than foo) we blindly send a clone operation
4906 * for the [0, 100K[ range from foo to bar, the receiver ends up getting
4907 * a file bar that matches the content of file foo - iow, doesn't match
4908 * the content from bar in the original filesystem.
4909 */
4910 key.objectid = clone_root->ino;
4911 key.type = BTRFS_EXTENT_DATA_KEY;
4912 key.offset = clone_root->offset;
4913 ret = btrfs_search_slot(NULL, clone_root->root, &key, path, 0, 0);
4914 if (ret < 0)
4915 goto out;
4916 if (ret > 0 && path->slots[0] > 0) {
4917 btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0] - 1);
4918 if (key.objectid == clone_root->ino &&
4919 key.type == BTRFS_EXTENT_DATA_KEY)
4920 path->slots[0]--;
4921 }
4922
4923 while (true) {
4924 struct extent_buffer *leaf = path->nodes[0];
4925 int slot = path->slots[0];
4926 struct btrfs_file_extent_item *ei;
4927 u8 type;
4928 u64 ext_len;
4929 u64 clone_len;
4930
4931 if (slot >= btrfs_header_nritems(leaf)) {
4932 ret = btrfs_next_leaf(clone_root->root, path);
4933 if (ret < 0)
4934 goto out;
4935 else if (ret > 0)
4936 break;
4937 continue;
4938 }
4939
4940 btrfs_item_key_to_cpu(leaf, &key, slot);
4941
4942 /*
4943 * We might have an implicit trailing hole (NO_HOLES feature
4944 * enabled). We deal with it after leaving this loop.
4945 */
4946 if (key.objectid != clone_root->ino ||
4947 key.type != BTRFS_EXTENT_DATA_KEY)
4948 break;
4949
4950 ei = btrfs_item_ptr(leaf, slot, struct btrfs_file_extent_item);
4951 type = btrfs_file_extent_type(leaf, ei);
4952 if (type == BTRFS_FILE_EXTENT_INLINE) {
4953 ext_len = btrfs_file_extent_inline_len(leaf, slot, ei);
09cbfeaf 4954 ext_len = PAGE_ALIGN(ext_len);
d906d49f
FM		 4955 } else {
4956 ext_len = btrfs_file_extent_num_bytes(leaf, ei);
4957 }
4958
4959 if (key.offset + ext_len <= clone_root->offset)
4960 goto next;
4961
4962 if (key.offset > clone_root->offset) {
4963 /* Implicit hole, NO_HOLES feature enabled. */
4964 u64 hole_len = key.offset - clone_root->offset;
4965
4966 if (hole_len > len)
4967 hole_len = len;
4968 ret = send_extent_data(sctx, offset, hole_len);
4969 if (ret < 0)
4970 goto out;
4971
4972 len -= hole_len;
4973 if (len == 0)
4974 break;
4975 offset += hole_len;
4976 clone_root->offset += hole_len;
4977 data_offset += hole_len;
4978 }
4979
4980 if (key.offset >= clone_root->offset + len)
4981 break;
4982
4983 clone_len = min_t(u64, ext_len, len);
4984
4985 if (btrfs_file_extent_disk_bytenr(leaf, ei) == disk_byte &&
4986 btrfs_file_extent_offset(leaf, ei) == data_offset)
4987 ret = send_clone(sctx, offset, clone_len, clone_root);
4988 else
4989 ret = send_extent_data(sctx, offset, clone_len);
4990
4991 if (ret < 0)
4992 goto out;
4993
4994 len -= clone_len;
4995 if (len == 0)
4996 break;
4997 offset += clone_len;
4998 clone_root->offset += clone_len;
4999 data_offset += clone_len;
5000next:
5001 path->slots[0]++;
5002 }
5003
5004 if (len > 0)
5005 ret = send_extent_data(sctx, offset, len);
5006 else
5007 ret = 0;
5008out:
5009 btrfs_free_path(path);
5010 return ret;
5011}
5012
31db9f7c
AB		 5013static int send_write_or_clone(struct send_ctx *sctx,
5014 struct btrfs_path *path,
5015 struct btrfs_key *key,
5016 struct clone_root *clone_root)
5017{
5018 int ret = 0;
5019 struct btrfs_file_extent_item *ei;
5020 u64 offset = key->offset;
31db9f7c 5021 u64 len;
31db9f7c 5022 u8 type;
28e5dd8f 5023 u64 bs = sctx->send_root->fs_info->sb->s_blocksize;
31db9f7c
AB		 5024
5025 ei = btrfs_item_ptr(path->nodes[0], path->slots[0],
5026 struct btrfs_file_extent_item);
5027 type = btrfs_file_extent_type(path->nodes[0], ei);
74dd17fb 5028 if (type == BTRFS_FILE_EXTENT_INLINE) {
514ac8ad
CM		 5029 len = btrfs_file_extent_inline_len(path->nodes[0],
5030 path->slots[0], ei);
74dd17fb
CM		 5031 /*
5032 * it is possible the inline item won't cover the whole page,
5033 * but there may be items after this page. Make
5034 * sure to send the whole thing
5035 */
09cbfeaf 5036 len = PAGE_ALIGN(len);
74dd17fb 5037 } else {
31db9f7c 5038 len = btrfs_file_extent_num_bytes(path->nodes[0], ei);
74dd17fb 5039 }
31db9f7c
AB		 5040
5041 if (offset + len > sctx->cur_inode_size)
5042 len = sctx->cur_inode_size - offset;
5043 if (len == 0) {
5044 ret = 0;
5045 goto out;
5046 }
5047
28e5dd8f 5048 if (clone_root && IS_ALIGNED(offset + len, bs)) {
d906d49f
FM		 5049 u64 disk_byte;
5050 u64 data_offset;
5051
5052 disk_byte = btrfs_file_extent_disk_bytenr(path->nodes[0], ei);
5053 data_offset = btrfs_file_extent_offset(path->nodes[0], ei);
5054 ret = clone_range(sctx, clone_root, disk_byte, data_offset,
5055 offset, len);
cb95e7bf 5056 } else {
d906d49f 5057 ret = send_extent_data(sctx, offset, len);
31db9f7c 5058 }
31db9f7c
AB		 5059out:
5060 return ret;
5061}
5062
5063static int is_extent_unchanged(struct send_ctx *sctx,
5064 struct btrfs_path *left_path,
5065 struct btrfs_key *ekey)
5066{
5067 int ret = 0;
5068 struct btrfs_key key;
5069 struct btrfs_path *path = NULL;
5070 struct extent_buffer *eb;
5071 int slot;
5072 struct btrfs_key found_key;
5073 struct btrfs_file_extent_item *ei;
5074 u64 left_disknr;
5075 u64 right_disknr;
5076 u64 left_offset;
5077 u64 right_offset;
5078 u64 left_offset_fixed;
5079 u64 left_len;
5080 u64 right_len;
74dd17fb
CM		 5081 u64 left_gen;
5082 u64 right_gen;
31db9f7c
AB		 5083 u8 left_type;
5084 u8 right_type;
5085
5086 path = alloc_path_for_send();
5087 if (!path)
5088 return -ENOMEM;
5089
5090 eb = left_path->nodes[0];
5091 slot = left_path->slots[0];
31db9f7c
AB		 5092 ei = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item);
5093 left_type = btrfs_file_extent_type(eb, ei);
31db9f7c
AB		 5094
5095 if (left_type != BTRFS_FILE_EXTENT_REG) {
5096 ret = 0;
5097 goto out;
5098 }
74dd17fb
CM		 5099 left_disknr = btrfs_file_extent_disk_bytenr(eb, ei);
5100 left_len = btrfs_file_extent_num_bytes(eb, ei);
5101 left_offset = btrfs_file_extent_offset(eb, ei);
5102 left_gen = btrfs_file_extent_generation(eb, ei);
31db9f7c
AB		 5103
5104 /*
5105 * Following comments will refer to these graphics. L is the left
5106 * extents which we are checking at the moment. 1-8 are the right
5107 * extents that we iterate.
5108 *
5109 * |-----L-----|
5110 * |-1-|-2a-|-3-|-4-|-5-|-6-|
5111 *
5112 * |-----L-----|
5113 * |--1--|-2b-|...(same as above)
5114 *
5115 * Alternative situation. Happens on files where extents got split.
5116 * |-----L-----|
5117 * |-----------7-----------|-6-|
5118 *
5119 * Alternative situation. Happens on files which got larger.
5120 * |-----L-----|
5121 * |-8-|
5122 * Nothing follows after 8.
5123 */
5124
5125 key.objectid = ekey->objectid;
5126 key.type = BTRFS_EXTENT_DATA_KEY;
5127 key.offset = ekey->offset;
5128 ret = btrfs_search_slot_for_read(sctx->parent_root, &key, path, 0, 0);
5129 if (ret < 0)
5130 goto out;
5131 if (ret) {
5132 ret = 0;
5133 goto out;
5134 }
5135
5136 /*
5137 * Handle special case where the right side has no extents at all.
5138 */
5139 eb = path->nodes[0];
5140 slot = path->slots[0];
5141 btrfs_item_key_to_cpu(eb, &found_key, slot);
5142 if (found_key.objectid != key.objectid ||
5143 found_key.type != key.type) {
57cfd462
JB		 5144 /* If we're a hole then just pretend nothing changed */
5145 ret = (left_disknr) ? 0 : 1;
31db9f7c
AB		 5146 goto out;
5147 }
5148
5149 /*
5150 * We're now on 2a, 2b or 7.
5151 */
5152 key = found_key;
5153 while (key.offset < ekey->offset + left_len) {
5154 ei = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item);
5155 right_type = btrfs_file_extent_type(eb, ei);
31db9f7c
AB		 5156 if (right_type != BTRFS_FILE_EXTENT_REG) {
5157 ret = 0;
5158 goto out;
5159 }
5160
007d31f7
JB		 5161 right_disknr = btrfs_file_extent_disk_bytenr(eb, ei);
5162 right_len = btrfs_file_extent_num_bytes(eb, ei);
5163 right_offset = btrfs_file_extent_offset(eb, ei);
5164 right_gen = btrfs_file_extent_generation(eb, ei);
5165
31db9f7c
AB		 5166 /*
5167 * Are we at extent 8? If yes, we know the extent is changed.
5168 * This may only happen on the first iteration.
5169 */
d8347fa4 5170 if (found_key.offset + right_len <= ekey->offset) {
57cfd462
JB		 5171 /* If we're a hole just pretend nothing changed */
5172 ret = (left_disknr) ? 0 : 1;
31db9f7c
AB		 5173 goto out;
5174 }
5175
5176 left_offset_fixed = left_offset;
5177 if (key.offset < ekey->offset) {
5178 /* Fix the right offset for 2a and 7. */
5179 right_offset += ekey->offset - key.offset;
5180 } else {
5181 /* Fix the left offset for all behind 2a and 2b */
5182 left_offset_fixed += key.offset - ekey->offset;
5183 }
5184
5185 /*
5186 * Check if we have the same extent.
5187 */
3954096d 5188 if (left_disknr != right_disknr ||
74dd17fb
CM		 5189 left_offset_fixed != right_offset ||
5190 left_gen != right_gen) {
31db9f7c
AB		 5191 ret = 0;
5192 goto out;
5193 }
5194
5195 /*
5196 * Go to the next extent.
5197 */
5198 ret = btrfs_next_item(sctx->parent_root, path);
5199 if (ret < 0)
5200 goto out;
5201 if (!ret) {
5202 eb = path->nodes[0];
5203 slot = path->slots[0];
5204 btrfs_item_key_to_cpu(eb, &found_key, slot);
5205 }
5206 if (ret || found_key.objectid != key.objectid ||
5207 found_key.type != key.type) {
5208 key.offset += right_len;
5209 break;
adaa4b8e
JS		 5210 }
5211 if (found_key.offset != key.offset + right_len) {
5212 ret = 0;
5213 goto out;
31db9f7c
AB		 5214 }
5215 key = found_key;
5216 }
5217
5218 /*
5219 * We're now behind the left extent (treat as unchanged) or at the end
5220 * of the right side (treat as changed).
5221 */
5222 if (key.offset >= ekey->offset + left_len)
5223 ret = 1;
5224 else
5225 ret = 0;
5226
5227
5228out:
5229 btrfs_free_path(path);
5230 return ret;
5231}
5232
16e7549f
JB		 5233static int get_last_extent(struct send_ctx *sctx, u64 offset)
5234{
5235 struct btrfs_path *path;
5236 struct btrfs_root *root = sctx->send_root;
5237 struct btrfs_file_extent_item *fi;
5238 struct btrfs_key key;
5239 u64 extent_end;
5240 u8 type;
5241 int ret;
5242
5243 path = alloc_path_for_send();
5244 if (!path)
5245 return -ENOMEM;
5246
5247 sctx->cur_inode_last_extent = 0;
5248
5249 key.objectid = sctx->cur_ino;
5250 key.type = BTRFS_EXTENT_DATA_KEY;
5251 key.offset = offset;
5252 ret = btrfs_search_slot_for_read(root, &key, path, 0, 1);
5253 if (ret < 0)
5254 goto out;
5255 ret = 0;
5256 btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
5257 if (key.objectid != sctx->cur_ino || key.type != BTRFS_EXTENT_DATA_KEY)
5258 goto out;
5259
5260 fi = btrfs_item_ptr(path->nodes[0], path->slots[0],
5261 struct btrfs_file_extent_item);
5262 type = btrfs_file_extent_type(path->nodes[0], fi);
5263 if (type == BTRFS_FILE_EXTENT_INLINE) {
514ac8ad
CM		 5264 u64 size = btrfs_file_extent_inline_len(path->nodes[0],
5265 path->slots[0], fi);
16e7549f
JB		 5266 extent_end = ALIGN(key.offset + size,
5267 sctx->send_root->sectorsize);
5268 } else {
5269 extent_end = key.offset +
5270 btrfs_file_extent_num_bytes(path->nodes[0], fi);
5271 }
5272 sctx->cur_inode_last_extent = extent_end;
5273out:
5274 btrfs_free_path(path);
5275 return ret;
5276}
5277
5278static int maybe_send_hole(struct send_ctx *sctx, struct btrfs_path *path,
5279 struct btrfs_key *key)
5280{
5281 struct btrfs_file_extent_item *fi;
5282 u64 extent_end;
5283 u8 type;
5284 int ret = 0;
5285
5286 if (sctx->cur_ino != key->objectid || !need_send_hole(sctx))
5287 return 0;
5288
5289 if (sctx->cur_inode_last_extent == (u64)-1) {
5290 ret = get_last_extent(sctx, key->offset - 1);
5291 if (ret)
5292 return ret;
5293 }
5294
5295 fi = btrfs_item_ptr(path->nodes[0], path->slots[0],
5296 struct btrfs_file_extent_item);
5297 type = btrfs_file_extent_type(path->nodes[0], fi);
5298 if (type == BTRFS_FILE_EXTENT_INLINE) {
514ac8ad
CM		 5299 u64 size = btrfs_file_extent_inline_len(path->nodes[0],
5300 path->slots[0], fi);
16e7549f
JB		 5301 extent_end = ALIGN(key->offset + size,
5302 sctx->send_root->sectorsize);
5303 } else {
5304 extent_end = key->offset +
5305 btrfs_file_extent_num_bytes(path->nodes[0], fi);
5306 }
bf54f412
FDBM		 5307
5308 if (path->slots[0] == 0 &&
5309 sctx->cur_inode_last_extent < key->offset) {
5310 /*
5311 * We might have skipped entire leafs that contained only
5312 * file extent items for our current inode. These leafs have
5313 * a generation number smaller (older) than the one in the
5314 * current leaf and the leaf our last extent came from, and
5315 * are located between these 2 leafs.
5316 */
5317 ret = get_last_extent(sctx, key->offset - 1);
5318 if (ret)
5319 return ret;
5320 }
5321
16e7549f
JB		 5322 if (sctx->cur_inode_last_extent < key->offset)
5323 ret = send_hole(sctx, key->offset);
5324 sctx->cur_inode_last_extent = extent_end;
5325 return ret;
5326}
5327
31db9f7c
AB		 5328static int process_extent(struct send_ctx *sctx,
5329 struct btrfs_path *path,
5330 struct btrfs_key *key)
5331{
31db9f7c 5332 struct clone_root *found_clone = NULL;
57cfd462 5333 int ret = 0;
31db9f7c
AB		 5334
5335 if (S_ISLNK(sctx->cur_inode_mode))
5336 return 0;
5337
5338 if (sctx->parent_root && !sctx->cur_inode_new) {
5339 ret = is_extent_unchanged(sctx, path, key);
5340 if (ret < 0)
5341 goto out;
5342 if (ret) {
5343 ret = 0;
16e7549f 5344 goto out_hole;
31db9f7c 5345 }
57cfd462
JB		 5346 } else {
5347 struct btrfs_file_extent_item *ei;
5348 u8 type;
5349
5350 ei = btrfs_item_ptr(path->nodes[0], path->slots[0],
5351 struct btrfs_file_extent_item);
5352 type = btrfs_file_extent_type(path->nodes[0], ei);
5353 if (type == BTRFS_FILE_EXTENT_PREALLOC ||
5354 type == BTRFS_FILE_EXTENT_REG) {
5355 /*
5356 * The send spec does not have a prealloc command yet,
5357 * so just leave a hole for prealloc'ed extents until
5358 * we have enough commands queued up to justify rev'ing
5359 * the send spec.
5360 */
5361 if (type == BTRFS_FILE_EXTENT_PREALLOC) {
5362 ret = 0;
5363 goto out;
5364 }
5365
5366 /* Have a hole, just skip it. */
5367 if (btrfs_file_extent_disk_bytenr(path->nodes[0], ei) == 0) {
5368 ret = 0;
5369 goto out;
5370 }
5371 }
31db9f7c
AB		 5372 }
5373
5374 ret = find_extent_clone(sctx, path, key->objectid, key->offset,
5375 sctx->cur_inode_size, &found_clone);
5376 if (ret != -ENOENT && ret < 0)
5377 goto out;
5378
5379 ret = send_write_or_clone(sctx, path, key, found_clone);
16e7549f
JB		 5380 if (ret)
5381 goto out;
5382out_hole:
5383 ret = maybe_send_hole(sctx, path, key);
31db9f7c
AB		 5384out:
5385 return ret;
5386}
5387
5388static int process_all_extents(struct send_ctx *sctx)
5389{
5390 int ret;
5391 struct btrfs_root *root;
5392 struct btrfs_path *path;
5393 struct btrfs_key key;
5394 struct btrfs_key found_key;
5395 struct extent_buffer *eb;
5396 int slot;
5397
5398 root = sctx->send_root;
5399 path = alloc_path_for_send();
5400 if (!path)
5401 return -ENOMEM;
5402
5403 key.objectid = sctx->cmp_key->objectid;
5404 key.type = BTRFS_EXTENT_DATA_KEY;
5405 key.offset = 0;
7fdd29d0
FDBM		 5406 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
5407 if (ret < 0)
5408 goto out;
31db9f7c 5409
7fdd29d0 5410 while (1) {
31db9f7c
AB		 5411 eb = path->nodes[0];
5412 slot = path->slots[0];
7fdd29d0
FDBM		 5413
5414 if (slot >= btrfs_header_nritems(eb)) {
5415 ret = btrfs_next_leaf(root, path);
5416 if (ret < 0) {
5417 goto out;
5418 } else if (ret > 0) {
5419 ret = 0;
5420 break;
5421 }
5422 continue;
5423 }
5424
31db9f7c
AB		 5425 btrfs_item_key_to_cpu(eb, &found_key, slot);
5426
5427 if (found_key.objectid != key.objectid ||
5428 found_key.type != key.type) {
5429 ret = 0;
5430 goto out;
5431 }
5432
5433 ret = process_extent(sctx, path, &found_key);
5434 if (ret < 0)
5435 goto out;
5436
7fdd29d0 5437 path->slots[0]++;
31db9f7c
AB		 5438 }
5439
5440out:
5441 btrfs_free_path(path);
5442 return ret;
5443}
5444
9f03740a
FDBM		 5445static int process_recorded_refs_if_needed(struct send_ctx *sctx, int at_end,
5446 int *pending_move,
5447 int *refs_processed)
31db9f7c
AB		 5448{
5449 int ret = 0;
5450
5451 if (sctx->cur_ino == 0)
5452 goto out;
5453 if (!at_end && sctx->cur_ino == sctx->cmp_key->objectid &&
96b5bd77 5454 sctx->cmp_key->type <= BTRFS_INODE_EXTREF_KEY)
31db9f7c
AB		 5455 goto out;
5456 if (list_empty(&sctx->new_refs) && list_empty(&sctx->deleted_refs))
5457 goto out;
5458
9f03740a 5459 ret = process_recorded_refs(sctx, pending_move);
e479d9bb
AB		 5460 if (ret < 0)
5461 goto out;
5462
9f03740a 5463 *refs_processed = 1;
31db9f7c
AB		 5464out:
5465 return ret;
5466}
5467
5468static int finish_inode_if_needed(struct send_ctx *sctx, int at_end)
5469{
5470 int ret = 0;
5471 u64 left_mode;
5472 u64 left_uid;
5473 u64 left_gid;
5474 u64 right_mode;
5475 u64 right_uid;
5476 u64 right_gid;
5477 int need_chmod = 0;
5478 int need_chown = 0;
9f03740a
FDBM		 5479 int pending_move = 0;
5480 int refs_processed = 0;
31db9f7c 5481
9f03740a
FDBM		 5482 ret = process_recorded_refs_if_needed(sctx, at_end, &pending_move,
5483 &refs_processed);
31db9f7c
AB		 5484 if (ret < 0)
5485 goto out;
5486
9f03740a
FDBM		 5487 /*
5488 * We have processed the refs and thus need to advance send_progress.
5489 * Now, calls to get_cur_xxx will take the updated refs of the current
5490 * inode into account.
5491 *
5492 * On the other hand, if our current inode is a directory and couldn't
5493 * be moved/renamed because its parent was renamed/moved too and it has
5494 * a higher inode number, we can only move/rename our current inode
5495 * after we moved/renamed its parent. Therefore in this case operate on
5496 * the old path (pre move/rename) of our current inode, and the
5497 * move/rename will be performed later.
5498 */
5499 if (refs_processed && !pending_move)
5500 sctx->send_progress = sctx->cur_ino + 1;
5501
31db9f7c
AB		 5502 if (sctx->cur_ino == 0 || sctx->cur_inode_deleted)
5503 goto out;
5504 if (!at_end && sctx->cmp_key->objectid == sctx->cur_ino)
5505 goto out;
5506
5507 ret = get_inode_info(sctx->send_root, sctx->cur_ino, NULL, NULL,
85a7b33b 5508 &left_mode, &left_uid, &left_gid, NULL);
31db9f7c
AB		 5509 if (ret < 0)
5510 goto out;
5511
e2d044fe
AL		 5512 if (!sctx->parent_root || sctx->cur_inode_new) {
5513 need_chown = 1;
5514 if (!S_ISLNK(sctx->cur_inode_mode))
31db9f7c 5515 need_chmod = 1;
e2d044fe
AL		 5516 } else {
5517 ret = get_inode_info(sctx->parent_root, sctx->cur_ino,
5518 NULL, NULL, &right_mode, &right_uid,
5519 &right_gid, NULL);
5520 if (ret < 0)
5521 goto out;
31db9f7c 5522
e2d044fe
AL		 5523 if (left_uid != right_uid || left_gid != right_gid)
5524 need_chown = 1;
5525 if (!S_ISLNK(sctx->cur_inode_mode) && left_mode != right_mode)
5526 need_chmod = 1;
31db9f7c
AB		 5527 }
5528
5529 if (S_ISREG(sctx->cur_inode_mode)) {
16e7549f 5530 if (need_send_hole(sctx)) {
766b5e5a
FM		 5531 if (sctx->cur_inode_last_extent == (u64)-1 ||
5532 sctx->cur_inode_last_extent <
5533 sctx->cur_inode_size) {
16e7549f
JB		 5534 ret = get_last_extent(sctx, (u64)-1);
5535 if (ret)
5536 goto out;
5537 }
5538 if (sctx->cur_inode_last_extent <
5539 sctx->cur_inode_size) {
5540 ret = send_hole(sctx, sctx->cur_inode_size);
5541 if (ret)
5542 goto out;
5543 }
5544 }
31db9f7c
AB		 5545 ret = send_truncate(sctx, sctx->cur_ino, sctx->cur_inode_gen,
5546 sctx->cur_inode_size);
5547 if (ret < 0)
5548 goto out;
5549 }
5550
5551 if (need_chown) {
5552 ret = send_chown(sctx, sctx->cur_ino, sctx->cur_inode_gen,
5553 left_uid, left_gid);
5554 if (ret < 0)
5555 goto out;
5556 }
5557 if (need_chmod) {
5558 ret = send_chmod(sctx, sctx->cur_ino, sctx->cur_inode_gen,
5559 left_mode);
5560 if (ret < 0)
5561 goto out;
5562 }
5563
5564 /*
9f03740a
FDBM		 5565 * If other directory inodes depended on our current directory
5566 * inode's move/rename, now do their move/rename operations.
31db9f7c 5567 */
9f03740a
FDBM		 5568 if (!is_waiting_for_move(sctx, sctx->cur_ino)) {
5569 ret = apply_children_dir_moves(sctx);
5570 if (ret)
5571 goto out;
fcbd2154
FM		 5572 /*
5573 * Need to send that every time, no matter if it actually
5574 * changed between the two trees as we have done changes to
5575 * the inode before. If our inode is a directory and it's
5576 * waiting to be moved/renamed, we will send its utimes when
5577 * it's moved/renamed, therefore we don't need to do it here.
5578 */
5579 sctx->send_progress = sctx->cur_ino + 1;
5580 ret = send_utimes(sctx, sctx->cur_ino, sctx->cur_inode_gen);
5581 if (ret < 0)
5582 goto out;
9f03740a
FDBM		 5583 }
5584
31db9f7c
AB		 5585out:
5586 return ret;
5587}
5588
5589static int changed_inode(struct send_ctx *sctx,
5590 enum btrfs_compare_tree_result result)
5591{
5592 int ret = 0;
5593 struct btrfs_key *key = sctx->cmp_key;
5594 struct btrfs_inode_item *left_ii = NULL;
5595 struct btrfs_inode_item *right_ii = NULL;
5596 u64 left_gen = 0;
5597 u64 right_gen = 0;
5598
31db9f7c
AB		 5599 sctx->cur_ino = key->objectid;
5600 sctx->cur_inode_new_gen = 0;
16e7549f 5601 sctx->cur_inode_last_extent = (u64)-1;
e479d9bb
AB		 5602
5603 /*
5604 * Set send_progress to current inode. This will tell all get_cur_xxx
5605 * functions that the current inode's refs are not updated yet. Later,
5606 * when process_recorded_refs is finished, it is set to cur_ino + 1.
5607 */
31db9f7c
AB		 5608 sctx->send_progress = sctx->cur_ino;
5609
5610 if (result == BTRFS_COMPARE_TREE_NEW ||
5611 result == BTRFS_COMPARE_TREE_CHANGED) {
5612 left_ii = btrfs_item_ptr(sctx->left_path->nodes[0],
5613 sctx->left_path->slots[0],
5614 struct btrfs_inode_item);
5615 left_gen = btrfs_inode_generation(sctx->left_path->nodes[0],
5616 left_ii);
5617 } else {
5618 right_ii = btrfs_item_ptr(sctx->right_path->nodes[0],
5619 sctx->right_path->slots[0],
5620 struct btrfs_inode_item);
5621 right_gen = btrfs_inode_generation(sctx->right_path->nodes[0],
5622 right_ii);
5623 }
5624 if (result == BTRFS_COMPARE_TREE_CHANGED) {
5625 right_ii = btrfs_item_ptr(sctx->right_path->nodes[0],
5626 sctx->right_path->slots[0],
5627 struct btrfs_inode_item);
5628
5629 right_gen = btrfs_inode_generation(sctx->right_path->nodes[0],
5630 right_ii);
6d85ed05
AB		 5631
5632 /*
5633 * The cur_ino = root dir case is special here. We can't treat
5634 * the inode as deleted+reused because it would generate a
5635 * stream that tries to delete/mkdir the root dir.
5636 */
5637 if (left_gen != right_gen &&
5638 sctx->cur_ino != BTRFS_FIRST_FREE_OBJECTID)
31db9f7c
AB		 5639 sctx->cur_inode_new_gen = 1;
5640 }
5641
5642 if (result == BTRFS_COMPARE_TREE_NEW) {
5643 sctx->cur_inode_gen = left_gen;
5644 sctx->cur_inode_new = 1;
5645 sctx->cur_inode_deleted = 0;
5646 sctx->cur_inode_size = btrfs_inode_size(
5647 sctx->left_path->nodes[0], left_ii);
5648 sctx->cur_inode_mode = btrfs_inode_mode(
5649 sctx->left_path->nodes[0], left_ii);
644d1940
LB		 5650 sctx->cur_inode_rdev = btrfs_inode_rdev(
5651 sctx->left_path->nodes[0], left_ii);
31db9f7c 5652 if (sctx->cur_ino != BTRFS_FIRST_FREE_OBJECTID)
1f4692da 5653 ret = send_create_inode_if_needed(sctx);
31db9f7c
AB		 5654 } else if (result == BTRFS_COMPARE_TREE_DELETED) {
5655 sctx->cur_inode_gen = right_gen;
5656 sctx->cur_inode_new = 0;
5657 sctx->cur_inode_deleted = 1;
5658 sctx->cur_inode_size = btrfs_inode_size(
5659 sctx->right_path->nodes[0], right_ii);
5660 sctx->cur_inode_mode = btrfs_inode_mode(
5661 sctx->right_path->nodes[0], right_ii);
5662 } else if (result == BTRFS_COMPARE_TREE_CHANGED) {
766702ef
AB		 5663 /*
5664 * We need to do some special handling in case the inode was
5665 * reported as changed with a changed generation number. This
5666 * means that the original inode was deleted and new inode
5667 * reused the same inum. So we have to treat the old inode as
5668 * deleted and the new one as new.
5669 */
31db9f7c 5670 if (sctx->cur_inode_new_gen) {
766702ef
AB		 5671 /*
5672 * First, process the inode as if it was deleted.
5673 */
31db9f7c
AB		 5674 sctx->cur_inode_gen = right_gen;
5675 sctx->cur_inode_new = 0;
5676 sctx->cur_inode_deleted = 1;
5677 sctx->cur_inode_size = btrfs_inode_size(
5678 sctx->right_path->nodes[0], right_ii);
5679 sctx->cur_inode_mode = btrfs_inode_mode(
5680 sctx->right_path->nodes[0], right_ii);
5681 ret = process_all_refs(sctx,
5682 BTRFS_COMPARE_TREE_DELETED);
5683 if (ret < 0)
5684 goto out;
5685
766702ef
AB		 5686 /*
5687 * Now process the inode as if it was new.
5688 */
31db9f7c
AB		 5689 sctx->cur_inode_gen = left_gen;
5690 sctx->cur_inode_new = 1;
5691 sctx->cur_inode_deleted = 0;
5692 sctx->cur_inode_size = btrfs_inode_size(
5693 sctx->left_path->nodes[0], left_ii);
5694 sctx->cur_inode_mode = btrfs_inode_mode(
5695 sctx->left_path->nodes[0], left_ii);
644d1940
LB		 5696 sctx->cur_inode_rdev = btrfs_inode_rdev(
5697 sctx->left_path->nodes[0], left_ii);
1f4692da 5698 ret = send_create_inode_if_needed(sctx);
31db9f7c
AB		 5699 if (ret < 0)
5700 goto out;
5701
5702 ret = process_all_refs(sctx, BTRFS_COMPARE_TREE_NEW);
5703 if (ret < 0)
5704 goto out;
e479d9bb
AB		 5705 /*
5706 * Advance send_progress now as we did not get into
5707 * process_recorded_refs_if_needed in the new_gen case.
5708 */
5709 sctx->send_progress = sctx->cur_ino + 1;
766702ef
AB		 5710
5711 /*
5712 * Now process all extents and xattrs of the inode as if
5713 * they were all new.
5714 */
31db9f7c
AB		 5715 ret = process_all_extents(sctx);
5716 if (ret < 0)
5717 goto out;
5718 ret = process_all_new_xattrs(sctx);
5719 if (ret < 0)
5720 goto out;
5721 } else {
5722 sctx->cur_inode_gen = left_gen;
5723 sctx->cur_inode_new = 0;
5724 sctx->cur_inode_new_gen = 0;
5725 sctx->cur_inode_deleted = 0;
5726 sctx->cur_inode_size = btrfs_inode_size(
5727 sctx->left_path->nodes[0], left_ii);
5728 sctx->cur_inode_mode = btrfs_inode_mode(
5729 sctx->left_path->nodes[0], left_ii);
5730 }
5731 }
5732
5733out:
5734 return ret;
5735}
5736
766702ef
AB		 5737/*
5738 * We have to process new refs before deleted refs, but compare_trees gives us
5739 * the new and deleted refs mixed. To fix this, we record the new/deleted refs
5740 * first and later process them in process_recorded_refs.
5741 * For the cur_inode_new_gen case, we skip recording completely because
5742 * changed_inode did already initiate processing of refs. The reason for this is
5743 * that in this case, compare_tree actually compares the refs of 2 different
5744 * inodes. To fix this, process_all_refs is used in changed_inode to handle all
5745 * refs of the right tree as deleted and all refs of the left tree as new.
5746 */
31db9f7c
AB		 5747static int changed_ref(struct send_ctx *sctx,
5748 enum btrfs_compare_tree_result result)
5749{
5750 int ret = 0;
5751
95155585
FM		 5752 if (sctx->cur_ino != sctx->cmp_key->objectid) {
5753 inconsistent_snapshot_error(sctx, result, "reference");
5754 return -EIO;
5755 }
31db9f7c
AB		 5756
5757 if (!sctx->cur_inode_new_gen &&
5758 sctx->cur_ino != BTRFS_FIRST_FREE_OBJECTID) {
5759 if (result == BTRFS_COMPARE_TREE_NEW)
5760 ret = record_new_ref(sctx);
5761 else if (result == BTRFS_COMPARE_TREE_DELETED)
5762 ret = record_deleted_ref(sctx);
5763 else if (result == BTRFS_COMPARE_TREE_CHANGED)
5764 ret = record_changed_ref(sctx);
5765 }
5766
5767 return ret;
5768}
5769
766702ef
AB		 5770/*
5771 * Process new/deleted/changed xattrs. We skip processing in the
5772 * cur_inode_new_gen case because changed_inode did already initiate processing
5773 * of xattrs. The reason is the same as in changed_ref
5774 */
31db9f7c
AB		 5775static int changed_xattr(struct send_ctx *sctx,
5776 enum btrfs_compare_tree_result result)
5777{
5778 int ret = 0;
5779
95155585
FM		 5780 if (sctx->cur_ino != sctx->cmp_key->objectid) {
5781 inconsistent_snapshot_error(sctx, result, "xattr");
5782 return -EIO;
5783 }
31db9f7c
AB		 5784
5785 if (!sctx->cur_inode_new_gen && !sctx->cur_inode_deleted) {
5786 if (result == BTRFS_COMPARE_TREE_NEW)
5787 ret = process_new_xattr(sctx);
5788 else if (result == BTRFS_COMPARE_TREE_DELETED)
5789 ret = process_deleted_xattr(sctx);
5790 else if (result == BTRFS_COMPARE_TREE_CHANGED)
5791 ret = process_changed_xattr(sctx);
5792 }
5793
5794 return ret;
5795}
5796
766702ef
AB		 5797/*
5798 * Process new/deleted/changed extents. We skip processing in the
5799 * cur_inode_new_gen case because changed_inode did already initiate processing
5800 * of extents. The reason is the same as in changed_ref
5801 */
31db9f7c
AB		 5802static int changed_extent(struct send_ctx *sctx,
5803 enum btrfs_compare_tree_result result)
5804{
5805 int ret = 0;
5806
95155585
FM		 5807 if (sctx->cur_ino != sctx->cmp_key->objectid) {
5808 inconsistent_snapshot_error(sctx, result, "extent");
5809 return -EIO;
5810 }
31db9f7c
AB		 5811
5812 if (!sctx->cur_inode_new_gen && !sctx->cur_inode_deleted) {
5813 if (result != BTRFS_COMPARE_TREE_DELETED)
5814 ret = process_extent(sctx, sctx->left_path,
5815 sctx->cmp_key);
5816 }
5817
5818 return ret;
5819}
5820
ba5e8f2e
JB		 5821static int dir_changed(struct send_ctx *sctx, u64 dir)
5822{
5823 u64 orig_gen, new_gen;
5824 int ret;
5825
5826 ret = get_inode_info(sctx->send_root, dir, NULL, &new_gen, NULL, NULL,
5827 NULL, NULL);
5828 if (ret)
5829 return ret;
5830
5831 ret = get_inode_info(sctx->parent_root, dir, NULL, &orig_gen, NULL,
5832 NULL, NULL, NULL);
5833 if (ret)
5834 return ret;
5835
5836 return (orig_gen != new_gen) ? 1 : 0;
5837}
5838
5839static int compare_refs(struct send_ctx *sctx, struct btrfs_path *path,
5840 struct btrfs_key *key)
5841{
5842 struct btrfs_inode_extref *extref;
5843 struct extent_buffer *leaf;
5844 u64 dirid = 0, last_dirid = 0;
5845 unsigned long ptr;
5846 u32 item_size;
5847 u32 cur_offset = 0;
5848 int ref_name_len;
5849 int ret = 0;
5850
5851 /* Easy case, just check this one dirid */
5852 if (key->type == BTRFS_INODE_REF_KEY) {
5853 dirid = key->offset;
5854
5855 ret = dir_changed(sctx, dirid);
5856 goto out;
5857 }
5858
5859 leaf = path->nodes[0];
5860 item_size = btrfs_item_size_nr(leaf, path->slots[0]);
5861 ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
5862 while (cur_offset < item_size) {
5863 extref = (struct btrfs_inode_extref *)(ptr +
5864 cur_offset);
5865 dirid = btrfs_inode_extref_parent(leaf, extref);
5866 ref_name_len = btrfs_inode_extref_name_len(leaf, extref);
5867 cur_offset += ref_name_len + sizeof(*extref);
5868 if (dirid == last_dirid)
5869 continue;
5870 ret = dir_changed(sctx, dirid);
5871 if (ret)
5872 break;
5873 last_dirid = dirid;
5874 }
5875out:
5876 return ret;
5877}
5878
766702ef
AB		 5879/*
5880 * Updates compare related fields in sctx and simply forwards to the actual
5881 * changed_xxx functions.
5882 */
31db9f7c
AB		 5883static int changed_cb(struct btrfs_root *left_root,
5884 struct btrfs_root *right_root,
5885 struct btrfs_path *left_path,
5886 struct btrfs_path *right_path,
5887 struct btrfs_key *key,
5888 enum btrfs_compare_tree_result result,
5889 void *ctx)
5890{
5891 int ret = 0;
5892 struct send_ctx *sctx = ctx;
5893
ba5e8f2e 5894 if (result == BTRFS_COMPARE_TREE_SAME) {
16e7549f
JB		 5895 if (key->type == BTRFS_INODE_REF_KEY ||
5896 key->type == BTRFS_INODE_EXTREF_KEY) {
5897 ret = compare_refs(sctx, left_path, key);
5898 if (!ret)
5899 return 0;
5900 if (ret < 0)
5901 return ret;
5902 } else if (key->type == BTRFS_EXTENT_DATA_KEY) {
5903 return maybe_send_hole(sctx, left_path, key);
5904 } else {
ba5e8f2e 5905 return 0;
16e7549f 5906 }
ba5e8f2e
JB		 5907 result = BTRFS_COMPARE_TREE_CHANGED;
5908 ret = 0;
5909 }
5910
31db9f7c
AB		 5911 sctx->left_path = left_path;
5912 sctx->right_path = right_path;
5913 sctx->cmp_key = key;
5914
5915 ret = finish_inode_if_needed(sctx, 0);
5916 if (ret < 0)
5917 goto out;
5918
2981e225
AB		 5919 /* Ignore non-FS objects */
5920 if (key->objectid == BTRFS_FREE_INO_OBJECTID ||
5921 key->objectid == BTRFS_FREE_SPACE_OBJECTID)
5922 goto out;
5923
31db9f7c
AB		 5924 if (key->type == BTRFS_INODE_ITEM_KEY)
5925 ret = changed_inode(sctx, result);
96b5bd77
JS		 5926 else if (key->type == BTRFS_INODE_REF_KEY ||
5927 key->type == BTRFS_INODE_EXTREF_KEY)
31db9f7c
AB		 5928 ret = changed_ref(sctx, result);
5929 else if (key->type == BTRFS_XATTR_ITEM_KEY)
5930 ret = changed_xattr(sctx, result);
5931 else if (key->type == BTRFS_EXTENT_DATA_KEY)
5932 ret = changed_extent(sctx, result);
5933
5934out:
5935 return ret;
5936}
5937
5938static int full_send_tree(struct send_ctx *sctx)
5939{
5940 int ret;
31db9f7c
AB		 5941 struct btrfs_root *send_root = sctx->send_root;
5942 struct btrfs_key key;
5943 struct btrfs_key found_key;
5944 struct btrfs_path *path;
5945 struct extent_buffer *eb;
5946 int slot;
31db9f7c
AB		 5947
5948 path = alloc_path_for_send();
5949 if (!path)
5950 return -ENOMEM;
5951
31db9f7c
AB		 5952 key.objectid = BTRFS_FIRST_FREE_OBJECTID;
5953 key.type = BTRFS_INODE_ITEM_KEY;
5954 key.offset = 0;
5955
31db9f7c
AB		 5956 ret = btrfs_search_slot_for_read(send_root, &key, path, 1, 0);
5957 if (ret < 0)
5958 goto out;
5959 if (ret)
5960 goto out_finish;
5961
5962 while (1) {
31db9f7c
AB		 5963 eb = path->nodes[0];
5964 slot = path->slots[0];
5965 btrfs_item_key_to_cpu(eb, &found_key, slot);
5966
5967 ret = changed_cb(send_root, NULL, path, NULL,
5968 &found_key, BTRFS_COMPARE_TREE_NEW, sctx);
5969 if (ret < 0)
5970 goto out;
5971
5972 key.objectid = found_key.objectid;
5973 key.type = found_key.type;
5974 key.offset = found_key.offset + 1;
5975
5976 ret = btrfs_next_item(send_root, path);
5977 if (ret < 0)
5978 goto out;
5979 if (ret) {
5980 ret = 0;
5981 break;
5982 }
5983 }
5984
5985out_finish:
5986 ret = finish_inode_if_needed(sctx, 1);
5987
5988out:
5989 btrfs_free_path(path);
31db9f7c
AB		 5990 return ret;
5991}
5992
5993static int send_subvol(struct send_ctx *sctx)
5994{
5995 int ret;
5996
c2c71324
SB		 5997 if (!(sctx->flags & BTRFS_SEND_FLAG_OMIT_STREAM_HEADER)) {
5998 ret = send_header(sctx);
5999 if (ret < 0)
6000 goto out;
6001 }
31db9f7c
AB		 6002
6003 ret = send_subvol_begin(sctx);
6004 if (ret < 0)
6005 goto out;
6006
6007 if (sctx->parent_root) {
6008 ret = btrfs_compare_trees(sctx->send_root, sctx->parent_root,
6009 changed_cb, sctx);
6010 if (ret < 0)
6011 goto out;
6012 ret = finish_inode_if_needed(sctx, 1);
6013 if (ret < 0)
6014 goto out;
6015 } else {
6016 ret = full_send_tree(sctx);
6017 if (ret < 0)
6018 goto out;
6019 }
6020
6021out:
31db9f7c
AB		 6022 free_recorded_refs(sctx);
6023 return ret;
6024}
6025
e5fa8f86
FM		 6026/*
6027 * If orphan cleanup did remove any orphans from a root, it means the tree
6028 * was modified and therefore the commit root is not the same as the current
6029 * root anymore. This is a problem, because send uses the commit root and
6030 * therefore can see inode items that don't exist in the current root anymore,
6031 * and for example make calls to btrfs_iget, which will do tree lookups based
6032 * on the current root and not on the commit root. Those lookups will fail,
6033 * returning a -ESTALE error, and making send fail with that error. So make
6034 * sure a send does not see any orphans we have just removed, and that it will
6035 * see the same inodes regardless of whether a transaction commit happened
6036 * before it started (meaning that the commit root will be the same as the
6037 * current root) or not.
6038 */
6039static int ensure_commit_roots_uptodate(struct send_ctx *sctx)
6040{
6041 int i;
6042 struct btrfs_trans_handle *trans = NULL;
6043
6044again:
6045 if (sctx->parent_root &&
6046 sctx->parent_root->node != sctx->parent_root->commit_root)
6047 goto commit_trans;
6048
6049 for (i = 0; i < sctx->clone_roots_cnt; i++)
6050 if (sctx->clone_roots[i].root->node !=
6051 sctx->clone_roots[i].root->commit_root)
6052 goto commit_trans;
6053
6054 if (trans)
6055 return btrfs_end_transaction(trans, sctx->send_root);
6056
6057 return 0;
6058
6059commit_trans:
6060 /* Use any root, all fs roots will get their commit roots updated. */
6061 if (!trans) {
6062 trans = btrfs_join_transaction(sctx->send_root);
6063 if (IS_ERR(trans))
6064 return PTR_ERR(trans);
6065 goto again;
6066 }
6067
6068 return btrfs_commit_transaction(trans, sctx->send_root);
6069}
6070
66ef7d65
DS		 6071static void btrfs_root_dec_send_in_progress(struct btrfs_root* root)
6072{
6073 spin_lock(&root->root_item_lock);
6074 root->send_in_progress--;
6075 /*
6076 * Not much left to do, we don't know why it's unbalanced and
6077 * can't blindly reset it to 0.
6078 */
6079 if (root->send_in_progress < 0)
6080 btrfs_err(root->fs_info,
351fd353 6081 "send_in_progres unbalanced %d root %llu",
66ef7d65
DS		 6082 root->send_in_progress, root->root_key.objectid);
6083 spin_unlock(&root->root_item_lock);
6084}
6085
31db9f7c
AB		 6086long btrfs_ioctl_send(struct file *mnt_file, void __user *arg_)
6087{
6088 int ret = 0;
6089 struct btrfs_root *send_root;
6090 struct btrfs_root *clone_root;
6091 struct btrfs_fs_info *fs_info;
6092 struct btrfs_ioctl_send_args *arg = NULL;
6093 struct btrfs_key key;
31db9f7c
AB		 6094 struct send_ctx *sctx = NULL;
6095 u32 i;
6096 u64 *clone_sources_tmp = NULL;
2c686537 6097 int clone_sources_to_rollback = 0;
e55d1153 6098 unsigned alloc_size;
896c14f9 6099 int sort_clone_roots = 0;
18f687d5 6100 int index;
31db9f7c
AB		 6101
6102 if (!capable(CAP_SYS_ADMIN))
6103 return -EPERM;
6104
496ad9aa 6105 send_root = BTRFS_I(file_inode(mnt_file))->root;
31db9f7c
AB		 6106 fs_info = send_root->fs_info;
6107
2c686537
DS		 6108 /*
6109 * The subvolume must remain read-only during send, protect against
521e0546 6110 * making it RW. This also protects against deletion.
2c686537
DS		 6111 */
6112 spin_lock(&send_root->root_item_lock);
6113 send_root->send_in_progress++;
6114 spin_unlock(&send_root->root_item_lock);
6115
139f807a
JB		 6116 /*
6117 * This is done when we lookup the root, it should already be complete
6118 * by the time we get here.
6119 */
6120 WARN_ON(send_root->orphan_cleanup_state != ORPHAN_CLEANUP_DONE);
6121
2c686537
DS		 6122 /*
6123 * Userspace tools do the checks and warn the user if it's
6124 * not RO.
6125 */
6126 if (!btrfs_root_readonly(send_root)) {
6127 ret = -EPERM;
6128 goto out;
6129 }
6130
31db9f7c
AB		 6131 arg = memdup_user(arg_, sizeof(*arg));
6132 if (IS_ERR(arg)) {
6133 ret = PTR_ERR(arg);
6134 arg = NULL;
6135 goto out;
6136 }
6137
f5ecec3c
DC		 6138 if (arg->clone_sources_count >
6139 ULLONG_MAX / sizeof(*arg->clone_sources)) {
6140 ret = -EINVAL;
6141 goto out;
6142 }
6143
31db9f7c 6144 if (!access_ok(VERIFY_READ, arg->clone_sources,
700ff4f0
DC		 6145 sizeof(*arg->clone_sources) *
6146 arg->clone_sources_count)) {
31db9f7c
AB		 6147 ret = -EFAULT;
6148 goto out;
6149 }
6150
c2c71324 6151 if (arg->flags & ~BTRFS_SEND_FLAG_MASK) {
cb95e7bf
MF		 6152 ret = -EINVAL;
6153 goto out;
6154 }
6155
e780b0d1 6156 sctx = kzalloc(sizeof(struct send_ctx), GFP_KERNEL);
31db9f7c
AB		 6157 if (!sctx) {
6158 ret = -ENOMEM;
6159 goto out;
6160 }
6161
6162 INIT_LIST_HEAD(&sctx->new_refs);
6163 INIT_LIST_HEAD(&sctx->deleted_refs);
e780b0d1 6164 INIT_RADIX_TREE(&sctx->name_cache, GFP_KERNEL);
31db9f7c
AB		 6165 INIT_LIST_HEAD(&sctx->name_cache_list);
6166
cb95e7bf
MF		 6167 sctx->flags = arg->flags;
6168
31db9f7c 6169 sctx->send_filp = fget(arg->send_fd);
ecc7ada7
TI		 6170 if (!sctx->send_filp) {
6171 ret = -EBADF;
31db9f7c
AB		 6172 goto out;
6173 }
6174
31db9f7c 6175 sctx->send_root = send_root;
521e0546
DS		 6176 /*
6177 * Unlikely but possible, if the subvolume is marked for deletion but
6178 * is slow to remove the directory entry, send can still be started
6179 */
6180 if (btrfs_root_dead(sctx->send_root)) {
6181 ret = -EPERM;
6182 goto out;
6183 }
6184
31db9f7c
AB		 6185 sctx->clone_roots_cnt = arg->clone_sources_count;
6186
6187 sctx->send_max_size = BTRFS_SEND_BUF_SIZE;
6ff48ce0 6188 sctx->send_buf = kmalloc(sctx->send_max_size, GFP_KERNEL | __GFP_NOWARN);
31db9f7c 6189 if (!sctx->send_buf) {
6ff48ce0
DS		 6190 sctx->send_buf = vmalloc(sctx->send_max_size);
6191 if (!sctx->send_buf) {
6192 ret = -ENOMEM;
6193 goto out;
6194 }
31db9f7c
AB		 6195 }
6196
eb5b75fe 6197 sctx->read_buf = kmalloc(BTRFS_SEND_READ_SIZE, GFP_KERNEL | __GFP_NOWARN);
31db9f7c 6198 if (!sctx->read_buf) {
eb5b75fe
DS		 6199 sctx->read_buf = vmalloc(BTRFS_SEND_READ_SIZE);
6200 if (!sctx->read_buf) {
6201 ret = -ENOMEM;
6202 goto out;
6203 }
31db9f7c
AB		 6204 }
6205
9f03740a
FDBM		 6206 sctx->pending_dir_moves = RB_ROOT;
6207 sctx->waiting_dir_moves = RB_ROOT;
9dc44214 6208 sctx->orphan_dirs = RB_ROOT;
9f03740a 6209
e55d1153
DS		 6210 alloc_size = sizeof(struct clone_root) * (arg->clone_sources_count + 1);
6211
c03d01f3 6212 sctx->clone_roots = kzalloc(alloc_size, GFP_KERNEL | __GFP_NOWARN);
31db9f7c 6213 if (!sctx->clone_roots) {
c03d01f3
DS		 6214 sctx->clone_roots = vzalloc(alloc_size);
6215 if (!sctx->clone_roots) {
6216 ret = -ENOMEM;
6217 goto out;
6218 }
31db9f7c
AB		 6219 }
6220
e55d1153
DS		 6221 alloc_size = arg->clone_sources_count * sizeof(*arg->clone_sources);
6222
31db9f7c 6223 if (arg->clone_sources_count) {
2f91306a 6224 clone_sources_tmp = kmalloc(alloc_size, GFP_KERNEL | __GFP_NOWARN);
31db9f7c 6225 if (!clone_sources_tmp) {
2f91306a
DS		 6226 clone_sources_tmp = vmalloc(alloc_size);
6227 if (!clone_sources_tmp) {
6228 ret = -ENOMEM;
6229 goto out;
6230 }
31db9f7c
AB		 6231 }
6232
6233 ret = copy_from_user(clone_sources_tmp, arg->clone_sources,
e55d1153 6234 alloc_size);
31db9f7c
AB		 6235 if (ret) {
6236 ret = -EFAULT;
6237 goto out;
6238 }
6239
6240 for (i = 0; i < arg->clone_sources_count; i++) {
6241 key.objectid = clone_sources_tmp[i];
6242 key.type = BTRFS_ROOT_ITEM_KEY;
6243 key.offset = (u64)-1;
18f687d5
WS		 6244
6245 index = srcu_read_lock(&fs_info->subvol_srcu);
6246
31db9f7c 6247 clone_root = btrfs_read_fs_root_no_name(fs_info, &key);
31db9f7c 6248 if (IS_ERR(clone_root)) {
18f687d5 6249 srcu_read_unlock(&fs_info->subvol_srcu, index);
31db9f7c
AB		 6250 ret = PTR_ERR(clone_root);
6251 goto out;
6252 }
2c686537 6253 spin_lock(&clone_root->root_item_lock);
5cc2b17e
FM		 6254 if (!btrfs_root_readonly(clone_root) ||
6255 btrfs_root_dead(clone_root)) {
2c686537 6256 spin_unlock(&clone_root->root_item_lock);
18f687d5 6257 srcu_read_unlock(&fs_info->subvol_srcu, index);
2c686537
DS		 6258 ret = -EPERM;
6259 goto out;
6260 }
2f1f465a 6261 clone_root->send_in_progress++;
2c686537 6262 spin_unlock(&clone_root->root_item_lock);
18f687d5
WS		 6263 srcu_read_unlock(&fs_info->subvol_srcu, index);
6264
31db9f7c 6265 sctx->clone_roots[i].root = clone_root;
2f1f465a 6266 clone_sources_to_rollback = i + 1;
31db9f7c 6267 }
2f91306a 6268 kvfree(clone_sources_tmp);
31db9f7c
AB		 6269 clone_sources_tmp = NULL;
6270 }
6271
6272 if (arg->parent_root) {
6273 key.objectid = arg->parent_root;
6274 key.type = BTRFS_ROOT_ITEM_KEY;
6275 key.offset = (u64)-1;
18f687d5
WS		 6276
6277 index = srcu_read_lock(&fs_info->subvol_srcu);
6278
31db9f7c 6279 sctx->parent_root = btrfs_read_fs_root_no_name(fs_info, &key);
b1b19596 6280 if (IS_ERR(sctx->parent_root)) {
18f687d5 6281 srcu_read_unlock(&fs_info->subvol_srcu, index);
b1b19596 6282 ret = PTR_ERR(sctx->parent_root);
31db9f7c
AB		 6283 goto out;
6284 }
18f687d5 6285
2c686537
DS		 6286 spin_lock(&sctx->parent_root->root_item_lock);
6287 sctx->parent_root->send_in_progress++;
521e0546
DS		 6288 if (!btrfs_root_readonly(sctx->parent_root) ||
6289 btrfs_root_dead(sctx->parent_root)) {
2c686537 6290 spin_unlock(&sctx->parent_root->root_item_lock);
18f687d5 6291 srcu_read_unlock(&fs_info->subvol_srcu, index);
2c686537
DS		 6292 ret = -EPERM;
6293 goto out;
6294 }
6295 spin_unlock(&sctx->parent_root->root_item_lock);
18f687d5
WS		 6296
6297 srcu_read_unlock(&fs_info->subvol_srcu, index);
31db9f7c
AB		 6298 }
6299
6300 /*
6301 * Clones from send_root are allowed, but only if the clone source
6302 * is behind the current send position. This is checked while searching
6303 * for possible clone sources.
6304 */
6305 sctx->clone_roots[sctx->clone_roots_cnt++].root = sctx->send_root;
6306
6307 /* We do a bsearch later */
6308 sort(sctx->clone_roots, sctx->clone_roots_cnt,
6309 sizeof(*sctx->clone_roots), __clone_root_cmp_sort,
6310 NULL);
896c14f9 6311 sort_clone_roots = 1;
31db9f7c 6312
e5fa8f86
FM		 6313 ret = ensure_commit_roots_uptodate(sctx);
6314 if (ret)
6315 goto out;
6316
2755a0de 6317 current->journal_info = BTRFS_SEND_TRANS_STUB;
31db9f7c 6318 ret = send_subvol(sctx);
a26e8c9f 6319 current->journal_info = NULL;
31db9f7c
AB		 6320 if (ret < 0)
6321 goto out;
6322
c2c71324
SB		 6323 if (!(sctx->flags & BTRFS_SEND_FLAG_OMIT_END_CMD)) {
6324 ret = begin_cmd(sctx, BTRFS_SEND_C_END);
6325 if (ret < 0)
6326 goto out;
6327 ret = send_cmd(sctx);
6328 if (ret < 0)
6329 goto out;
6330 }
31db9f7c
AB		 6331
6332out:
9f03740a
FDBM		 6333 WARN_ON(sctx && !ret && !RB_EMPTY_ROOT(&sctx->pending_dir_moves));
6334 while (sctx && !RB_EMPTY_ROOT(&sctx->pending_dir_moves)) {
6335 struct rb_node *n;
6336 struct pending_dir_move *pm;
6337
6338 n = rb_first(&sctx->pending_dir_moves);
6339 pm = rb_entry(n, struct pending_dir_move, node);
6340 while (!list_empty(&pm->list)) {
6341 struct pending_dir_move *pm2;
6342
6343 pm2 = list_first_entry(&pm->list,
6344 struct pending_dir_move, list);
6345 free_pending_move(sctx, pm2);
6346 }
6347 free_pending_move(sctx, pm);
6348 }
6349
6350 WARN_ON(sctx && !ret && !RB_EMPTY_ROOT(&sctx->waiting_dir_moves));
6351 while (sctx && !RB_EMPTY_ROOT(&sctx->waiting_dir_moves)) {
6352 struct rb_node *n;
6353 struct waiting_dir_move *dm;
6354
6355 n = rb_first(&sctx->waiting_dir_moves);
6356 dm = rb_entry(n, struct waiting_dir_move, node);
6357 rb_erase(&dm->node, &sctx->waiting_dir_moves);
6358 kfree(dm);
6359 }
6360
9dc44214
FM		 6361 WARN_ON(sctx && !ret && !RB_EMPTY_ROOT(&sctx->orphan_dirs));
6362 while (sctx && !RB_EMPTY_ROOT(&sctx->orphan_dirs)) {
6363 struct rb_node *n;
6364 struct orphan_dir_info *odi;
6365
6366 n = rb_first(&sctx->orphan_dirs);
6367 odi = rb_entry(n, struct orphan_dir_info, node);
6368 free_orphan_dir_info(sctx, odi);
6369 }
6370
896c14f9
WS		 6371 if (sort_clone_roots) {
6372 for (i = 0; i < sctx->clone_roots_cnt; i++)
6373 btrfs_root_dec_send_in_progress(
6374 sctx->clone_roots[i].root);
6375 } else {
6376 for (i = 0; sctx && i < clone_sources_to_rollback; i++)
6377 btrfs_root_dec_send_in_progress(
6378 sctx->clone_roots[i].root);
6379
6380 btrfs_root_dec_send_in_progress(send_root);
6381 }
66ef7d65
DS		 6382 if (sctx && !IS_ERR_OR_NULL(sctx->parent_root))
6383 btrfs_root_dec_send_in_progress(sctx->parent_root);
2c686537 6384
31db9f7c 6385 kfree(arg);
2f91306a 6386 kvfree(clone_sources_tmp);
31db9f7c
AB		 6387
6388 if (sctx) {
6389 if (sctx->send_filp)
6390 fput(sctx->send_filp);
6391
c03d01f3 6392 kvfree(sctx->clone_roots);
6ff48ce0 6393 kvfree(sctx->send_buf);
eb5b75fe 6394 kvfree(sctx->read_buf);
31db9f7c
AB		 6395
6396 name_cache_free(sctx);
6397
6398 kfree(sctx);
6399 }
6400
6401 return ret;
6402}
Linux 6.x block layer and io_uring tree(s)