xref: /openbmc/linux/fs/btrfs/backref.c (revision 5b448065)
1 // SPDX-License-Identifier: GPL-2.0
2 /*
3  * Copyright (C) 2011 STRATO.  All rights reserved.
4  */
5 
6 #include <linux/mm.h>
7 #include <linux/rbtree.h>
8 #include <trace/events/btrfs.h>
9 #include "ctree.h"
10 #include "disk-io.h"
11 #include "backref.h"
12 #include "ulist.h"
13 #include "transaction.h"
14 #include "delayed-ref.h"
15 #include "locking.h"
16 #include "misc.h"
17 #include "tree-mod-log.h"
18 
19 /* Just an arbitrary number so we can be sure this happened */
20 #define BACKREF_FOUND_SHARED 6
21 
22 struct extent_inode_elem {
23 	u64 inum;
24 	u64 offset;
25 	struct extent_inode_elem *next;
26 };
27 
28 static int check_extent_in_eb(const struct btrfs_key *key,
29 			      const struct extent_buffer *eb,
30 			      const struct btrfs_file_extent_item *fi,
31 			      u64 extent_item_pos,
32 			      struct extent_inode_elem **eie,
33 			      bool ignore_offset)
34 {
35 	u64 offset = 0;
36 	struct extent_inode_elem *e;
37 
38 	if (!ignore_offset &&
39 	    !btrfs_file_extent_compression(eb, fi) &&
40 	    !btrfs_file_extent_encryption(eb, fi) &&
41 	    !btrfs_file_extent_other_encoding(eb, fi)) {
42 		u64 data_offset;
43 		u64 data_len;
44 
45 		data_offset = btrfs_file_extent_offset(eb, fi);
46 		data_len = btrfs_file_extent_num_bytes(eb, fi);
47 
48 		if (extent_item_pos < data_offset ||
49 		    extent_item_pos >= data_offset + data_len)
50 			return 1;
51 		offset = extent_item_pos - data_offset;
52 	}
53 
54 	e = kmalloc(sizeof(*e), GFP_NOFS);
55 	if (!e)
56 		return -ENOMEM;
57 
58 	e->next = *eie;
59 	e->inum = key->objectid;
60 	e->offset = key->offset + offset;
61 	*eie = e;
62 
63 	return 0;
64 }
65 
66 static void free_inode_elem_list(struct extent_inode_elem *eie)
67 {
68 	struct extent_inode_elem *eie_next;
69 
70 	for (; eie; eie = eie_next) {
71 		eie_next = eie->next;
72 		kfree(eie);
73 	}
74 }
75 
76 static int find_extent_in_eb(const struct extent_buffer *eb,
77 			     u64 wanted_disk_byte, u64 extent_item_pos,
78 			     struct extent_inode_elem **eie,
79 			     bool ignore_offset)
80 {
81 	u64 disk_byte;
82 	struct btrfs_key key;
83 	struct btrfs_file_extent_item *fi;
84 	int slot;
85 	int nritems;
86 	int extent_type;
87 	int ret;
88 
89 	/*
90 	 * from the shared data ref, we only have the leaf but we need
91 	 * the key. thus, we must look into all items and see that we
92 	 * find one (some) with a reference to our extent item.
93 	 */
94 	nritems = btrfs_header_nritems(eb);
95 	for (slot = 0; slot < nritems; ++slot) {
96 		btrfs_item_key_to_cpu(eb, &key, slot);
97 		if (key.type != BTRFS_EXTENT_DATA_KEY)
98 			continue;
99 		fi = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item);
100 		extent_type = btrfs_file_extent_type(eb, fi);
101 		if (extent_type == BTRFS_FILE_EXTENT_INLINE)
102 			continue;
103 		/* don't skip BTRFS_FILE_EXTENT_PREALLOC, we can handle that */
104 		disk_byte = btrfs_file_extent_disk_bytenr(eb, fi);
105 		if (disk_byte != wanted_disk_byte)
106 			continue;
107 
108 		ret = check_extent_in_eb(&key, eb, fi, extent_item_pos, eie, ignore_offset);
109 		if (ret < 0)
110 			return ret;
111 	}
112 
113 	return 0;
114 }
115 
116 struct preftree {
117 	struct rb_root_cached root;
118 	unsigned int count;
119 };
120 
121 #define PREFTREE_INIT	{ .root = RB_ROOT_CACHED, .count = 0 }
122 
123 struct preftrees {
124 	struct preftree direct;    /* BTRFS_SHARED_[DATA|BLOCK]_REF_KEY */
125 	struct preftree indirect;  /* BTRFS_[TREE_BLOCK|EXTENT_DATA]_REF_KEY */
126 	struct preftree indirect_missing_keys;
127 };
128 
129 /*
130  * Checks for a shared extent during backref search.
131  *
132  * The share_count tracks prelim_refs (direct and indirect) having a
133  * ref->count >0:
134  *  - incremented when a ref->count transitions to >0
135  *  - decremented when a ref->count transitions to <1
136  */
137 struct share_check {
138 	u64 root_objectid;
139 	u64 inum;
140 	int share_count;
141 };
142 
143 static inline int extent_is_shared(struct share_check *sc)
144 {
145 	return (sc && sc->share_count > 1) ? BACKREF_FOUND_SHARED : 0;
146 }
147 
148 static struct kmem_cache *btrfs_prelim_ref_cache;
149 
150 int __init btrfs_prelim_ref_init(void)
151 {
152 	btrfs_prelim_ref_cache = kmem_cache_create("btrfs_prelim_ref",
153 					sizeof(struct prelim_ref),
154 					0,
155 					SLAB_MEM_SPREAD,
156 					NULL);
157 	if (!btrfs_prelim_ref_cache)
158 		return -ENOMEM;
159 	return 0;
160 }
161 
162 void __cold btrfs_prelim_ref_exit(void)
163 {
164 	kmem_cache_destroy(btrfs_prelim_ref_cache);
165 }
166 
167 static void free_pref(struct prelim_ref *ref)
168 {
169 	kmem_cache_free(btrfs_prelim_ref_cache, ref);
170 }
171 
172 /*
173  * Return 0 when both refs are for the same block (and can be merged).
174  * A -1 return indicates ref1 is a 'lower' block than ref2, while 1
175  * indicates a 'higher' block.
176  */
177 static int prelim_ref_compare(struct prelim_ref *ref1,
178 			      struct prelim_ref *ref2)
179 {
180 	if (ref1->level < ref2->level)
181 		return -1;
182 	if (ref1->level > ref2->level)
183 		return 1;
184 	if (ref1->root_id < ref2->root_id)
185 		return -1;
186 	if (ref1->root_id > ref2->root_id)
187 		return 1;
188 	if (ref1->key_for_search.type < ref2->key_for_search.type)
189 		return -1;
190 	if (ref1->key_for_search.type > ref2->key_for_search.type)
191 		return 1;
192 	if (ref1->key_for_search.objectid < ref2->key_for_search.objectid)
193 		return -1;
194 	if (ref1->key_for_search.objectid > ref2->key_for_search.objectid)
195 		return 1;
196 	if (ref1->key_for_search.offset < ref2->key_for_search.offset)
197 		return -1;
198 	if (ref1->key_for_search.offset > ref2->key_for_search.offset)
199 		return 1;
200 	if (ref1->parent < ref2->parent)
201 		return -1;
202 	if (ref1->parent > ref2->parent)
203 		return 1;
204 
205 	return 0;
206 }
207 
208 static void update_share_count(struct share_check *sc, int oldcount,
209 			       int newcount)
210 {
211 	if ((!sc) || (oldcount == 0 && newcount < 1))
212 		return;
213 
214 	if (oldcount > 0 && newcount < 1)
215 		sc->share_count--;
216 	else if (oldcount < 1 && newcount > 0)
217 		sc->share_count++;
218 }
219 
220 /*
221  * Add @newref to the @root rbtree, merging identical refs.
222  *
223  * Callers should assume that newref has been freed after calling.
224  */
225 static void prelim_ref_insert(const struct btrfs_fs_info *fs_info,
226 			      struct preftree *preftree,
227 			      struct prelim_ref *newref,
228 			      struct share_check *sc)
229 {
230 	struct rb_root_cached *root;
231 	struct rb_node **p;
232 	struct rb_node *parent = NULL;
233 	struct prelim_ref *ref;
234 	int result;
235 	bool leftmost = true;
236 
237 	root = &preftree->root;
238 	p = &root->rb_root.rb_node;
239 
240 	while (*p) {
241 		parent = *p;
242 		ref = rb_entry(parent, struct prelim_ref, rbnode);
243 		result = prelim_ref_compare(ref, newref);
244 		if (result < 0) {
245 			p = &(*p)->rb_left;
246 		} else if (result > 0) {
247 			p = &(*p)->rb_right;
248 			leftmost = false;
249 		} else {
250 			/* Identical refs, merge them and free @newref */
251 			struct extent_inode_elem *eie = ref->inode_list;
252 
253 			while (eie && eie->next)
254 				eie = eie->next;
255 
256 			if (!eie)
257 				ref->inode_list = newref->inode_list;
258 			else
259 				eie->next = newref->inode_list;
260 			trace_btrfs_prelim_ref_merge(fs_info, ref, newref,
261 						     preftree->count);
262 			/*
263 			 * A delayed ref can have newref->count < 0.
264 			 * The ref->count is updated to follow any
265 			 * BTRFS_[ADD|DROP]_DELAYED_REF actions.
266 			 */
267 			update_share_count(sc, ref->count,
268 					   ref->count + newref->count);
269 			ref->count += newref->count;
270 			free_pref(newref);
271 			return;
272 		}
273 	}
274 
275 	update_share_count(sc, 0, newref->count);
276 	preftree->count++;
277 	trace_btrfs_prelim_ref_insert(fs_info, newref, NULL, preftree->count);
278 	rb_link_node(&newref->rbnode, parent, p);
279 	rb_insert_color_cached(&newref->rbnode, root, leftmost);
280 }
281 
282 /*
283  * Release the entire tree.  We don't care about internal consistency so
284  * just free everything and then reset the tree root.
285  */
286 static void prelim_release(struct preftree *preftree)
287 {
288 	struct prelim_ref *ref, *next_ref;
289 
290 	rbtree_postorder_for_each_entry_safe(ref, next_ref,
291 					     &preftree->root.rb_root, rbnode)
292 		free_pref(ref);
293 
294 	preftree->root = RB_ROOT_CACHED;
295 	preftree->count = 0;
296 }
297 
298 /*
299  * the rules for all callers of this function are:
300  * - obtaining the parent is the goal
301  * - if you add a key, you must know that it is a correct key
302  * - if you cannot add the parent or a correct key, then we will look into the
303  *   block later to set a correct key
304  *
305  * delayed refs
306  * ============
307  *        backref type | shared | indirect | shared | indirect
308  * information         |   tree |     tree |   data |     data
309  * --------------------+--------+----------+--------+----------
310  *      parent logical |    y   |     -    |    -   |     -
311  *      key to resolve |    -   |     y    |    y   |     y
312  *  tree block logical |    -   |     -    |    -   |     -
313  *  root for resolving |    y   |     y    |    y   |     y
314  *
315  * - column 1:       we've the parent -> done
316  * - column 2, 3, 4: we use the key to find the parent
317  *
318  * on disk refs (inline or keyed)
319  * ==============================
320  *        backref type | shared | indirect | shared | indirect
321  * information         |   tree |     tree |   data |     data
322  * --------------------+--------+----------+--------+----------
323  *      parent logical |    y   |     -    |    y   |     -
324  *      key to resolve |    -   |     -    |    -   |     y
325  *  tree block logical |    y   |     y    |    y   |     y
326  *  root for resolving |    -   |     y    |    y   |     y
327  *
328  * - column 1, 3: we've the parent -> done
329  * - column 2:    we take the first key from the block to find the parent
330  *                (see add_missing_keys)
331  * - column 4:    we use the key to find the parent
332  *
333  * additional information that's available but not required to find the parent
334  * block might help in merging entries to gain some speed.
335  */
336 static int add_prelim_ref(const struct btrfs_fs_info *fs_info,
337 			  struct preftree *preftree, u64 root_id,
338 			  const struct btrfs_key *key, int level, u64 parent,
339 			  u64 wanted_disk_byte, int count,
340 			  struct share_check *sc, gfp_t gfp_mask)
341 {
342 	struct prelim_ref *ref;
343 
344 	if (root_id == BTRFS_DATA_RELOC_TREE_OBJECTID)
345 		return 0;
346 
347 	ref = kmem_cache_alloc(btrfs_prelim_ref_cache, gfp_mask);
348 	if (!ref)
349 		return -ENOMEM;
350 
351 	ref->root_id = root_id;
352 	if (key)
353 		ref->key_for_search = *key;
354 	else
355 		memset(&ref->key_for_search, 0, sizeof(ref->key_for_search));
356 
357 	ref->inode_list = NULL;
358 	ref->level = level;
359 	ref->count = count;
360 	ref->parent = parent;
361 	ref->wanted_disk_byte = wanted_disk_byte;
362 	prelim_ref_insert(fs_info, preftree, ref, sc);
363 	return extent_is_shared(sc);
364 }
365 
366 /* direct refs use root == 0, key == NULL */
367 static int add_direct_ref(const struct btrfs_fs_info *fs_info,
368 			  struct preftrees *preftrees, int level, u64 parent,
369 			  u64 wanted_disk_byte, int count,
370 			  struct share_check *sc, gfp_t gfp_mask)
371 {
372 	return add_prelim_ref(fs_info, &preftrees->direct, 0, NULL, level,
373 			      parent, wanted_disk_byte, count, sc, gfp_mask);
374 }
375 
376 /* indirect refs use parent == 0 */
377 static int add_indirect_ref(const struct btrfs_fs_info *fs_info,
378 			    struct preftrees *preftrees, u64 root_id,
379 			    const struct btrfs_key *key, int level,
380 			    u64 wanted_disk_byte, int count,
381 			    struct share_check *sc, gfp_t gfp_mask)
382 {
383 	struct preftree *tree = &preftrees->indirect;
384 
385 	if (!key)
386 		tree = &preftrees->indirect_missing_keys;
387 	return add_prelim_ref(fs_info, tree, root_id, key, level, 0,
388 			      wanted_disk_byte, count, sc, gfp_mask);
389 }
390 
391 static int is_shared_data_backref(struct preftrees *preftrees, u64 bytenr)
392 {
393 	struct rb_node **p = &preftrees->direct.root.rb_root.rb_node;
394 	struct rb_node *parent = NULL;
395 	struct prelim_ref *ref = NULL;
396 	struct prelim_ref target = {};
397 	int result;
398 
399 	target.parent = bytenr;
400 
401 	while (*p) {
402 		parent = *p;
403 		ref = rb_entry(parent, struct prelim_ref, rbnode);
404 		result = prelim_ref_compare(ref, &target);
405 
406 		if (result < 0)
407 			p = &(*p)->rb_left;
408 		else if (result > 0)
409 			p = &(*p)->rb_right;
410 		else
411 			return 1;
412 	}
413 	return 0;
414 }
415 
416 static int add_all_parents(struct btrfs_root *root, struct btrfs_path *path,
417 			   struct ulist *parents,
418 			   struct preftrees *preftrees, struct prelim_ref *ref,
419 			   int level, u64 time_seq, const u64 *extent_item_pos,
420 			   bool ignore_offset)
421 {
422 	int ret = 0;
423 	int slot;
424 	struct extent_buffer *eb;
425 	struct btrfs_key key;
426 	struct btrfs_key *key_for_search = &ref->key_for_search;
427 	struct btrfs_file_extent_item *fi;
428 	struct extent_inode_elem *eie = NULL, *old = NULL;
429 	u64 disk_byte;
430 	u64 wanted_disk_byte = ref->wanted_disk_byte;
431 	u64 count = 0;
432 	u64 data_offset;
433 
434 	if (level != 0) {
435 		eb = path->nodes[level];
436 		ret = ulist_add(parents, eb->start, 0, GFP_NOFS);
437 		if (ret < 0)
438 			return ret;
439 		return 0;
440 	}
441 
442 	/*
443 	 * 1. We normally enter this function with the path already pointing to
444 	 *    the first item to check. But sometimes, we may enter it with
445 	 *    slot == nritems.
446 	 * 2. We are searching for normal backref but bytenr of this leaf
447 	 *    matches shared data backref
448 	 * 3. The leaf owner is not equal to the root we are searching
449 	 *
450 	 * For these cases, go to the next leaf before we continue.
451 	 */
452 	eb = path->nodes[0];
453 	if (path->slots[0] >= btrfs_header_nritems(eb) ||
454 	    is_shared_data_backref(preftrees, eb->start) ||
455 	    ref->root_id != btrfs_header_owner(eb)) {
456 		if (time_seq == BTRFS_SEQ_LAST)
457 			ret = btrfs_next_leaf(root, path);
458 		else
459 			ret = btrfs_next_old_leaf(root, path, time_seq);
460 	}
461 
462 	while (!ret && count < ref->count) {
463 		eb = path->nodes[0];
464 		slot = path->slots[0];
465 
466 		btrfs_item_key_to_cpu(eb, &key, slot);
467 
468 		if (key.objectid != key_for_search->objectid ||
469 		    key.type != BTRFS_EXTENT_DATA_KEY)
470 			break;
471 
472 		/*
473 		 * We are searching for normal backref but bytenr of this leaf
474 		 * matches shared data backref, OR
475 		 * the leaf owner is not equal to the root we are searching for
476 		 */
477 		if (slot == 0 &&
478 		    (is_shared_data_backref(preftrees, eb->start) ||
479 		     ref->root_id != btrfs_header_owner(eb))) {
480 			if (time_seq == BTRFS_SEQ_LAST)
481 				ret = btrfs_next_leaf(root, path);
482 			else
483 				ret = btrfs_next_old_leaf(root, path, time_seq);
484 			continue;
485 		}
486 		fi = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item);
487 		disk_byte = btrfs_file_extent_disk_bytenr(eb, fi);
488 		data_offset = btrfs_file_extent_offset(eb, fi);
489 
490 		if (disk_byte == wanted_disk_byte) {
491 			eie = NULL;
492 			old = NULL;
493 			if (ref->key_for_search.offset == key.offset - data_offset)
494 				count++;
495 			else
496 				goto next;
497 			if (extent_item_pos) {
498 				ret = check_extent_in_eb(&key, eb, fi,
499 						*extent_item_pos,
500 						&eie, ignore_offset);
501 				if (ret < 0)
502 					break;
503 			}
504 			if (ret > 0)
505 				goto next;
506 			ret = ulist_add_merge_ptr(parents, eb->start,
507 						  eie, (void **)&old, GFP_NOFS);
508 			if (ret < 0)
509 				break;
510 			if (!ret && extent_item_pos) {
511 				while (old->next)
512 					old = old->next;
513 				old->next = eie;
514 			}
515 			eie = NULL;
516 		}
517 next:
518 		if (time_seq == BTRFS_SEQ_LAST)
519 			ret = btrfs_next_item(root, path);
520 		else
521 			ret = btrfs_next_old_item(root, path, time_seq);
522 	}
523 
524 	if (ret > 0)
525 		ret = 0;
526 	else if (ret < 0)
527 		free_inode_elem_list(eie);
528 	return ret;
529 }
530 
531 /*
532  * resolve an indirect backref in the form (root_id, key, level)
533  * to a logical address
534  */
535 static int resolve_indirect_ref(struct btrfs_fs_info *fs_info,
536 				struct btrfs_path *path, u64 time_seq,
537 				struct preftrees *preftrees,
538 				struct prelim_ref *ref, struct ulist *parents,
539 				const u64 *extent_item_pos, bool ignore_offset)
540 {
541 	struct btrfs_root *root;
542 	struct extent_buffer *eb;
543 	int ret = 0;
544 	int root_level;
545 	int level = ref->level;
546 	struct btrfs_key search_key = ref->key_for_search;
547 
548 	/*
549 	 * If we're search_commit_root we could possibly be holding locks on
550 	 * other tree nodes.  This happens when qgroups does backref walks when
551 	 * adding new delayed refs.  To deal with this we need to look in cache
552 	 * for the root, and if we don't find it then we need to search the
553 	 * tree_root's commit root, thus the btrfs_get_fs_root_commit_root usage
554 	 * here.
555 	 */
556 	if (path->search_commit_root)
557 		root = btrfs_get_fs_root_commit_root(fs_info, path, ref->root_id);
558 	else
559 		root = btrfs_get_fs_root(fs_info, ref->root_id, false);
560 	if (IS_ERR(root)) {
561 		ret = PTR_ERR(root);
562 		goto out_free;
563 	}
564 
565 	if (!path->search_commit_root &&
566 	    test_bit(BTRFS_ROOT_DELETING, &root->state)) {
567 		ret = -ENOENT;
568 		goto out;
569 	}
570 
571 	if (btrfs_is_testing(fs_info)) {
572 		ret = -ENOENT;
573 		goto out;
574 	}
575 
576 	if (path->search_commit_root)
577 		root_level = btrfs_header_level(root->commit_root);
578 	else if (time_seq == BTRFS_SEQ_LAST)
579 		root_level = btrfs_header_level(root->node);
580 	else
581 		root_level = btrfs_old_root_level(root, time_seq);
582 
583 	if (root_level + 1 == level)
584 		goto out;
585 
586 	/*
587 	 * We can often find data backrefs with an offset that is too large
588 	 * (>= LLONG_MAX, maximum allowed file offset) due to underflows when
589 	 * subtracting a file's offset with the data offset of its
590 	 * corresponding extent data item. This can happen for example in the
591 	 * clone ioctl.
592 	 *
593 	 * So if we detect such case we set the search key's offset to zero to
594 	 * make sure we will find the matching file extent item at
595 	 * add_all_parents(), otherwise we will miss it because the offset
596 	 * taken form the backref is much larger then the offset of the file
597 	 * extent item. This can make us scan a very large number of file
598 	 * extent items, but at least it will not make us miss any.
599 	 *
600 	 * This is an ugly workaround for a behaviour that should have never
601 	 * existed, but it does and a fix for the clone ioctl would touch a lot
602 	 * of places, cause backwards incompatibility and would not fix the
603 	 * problem for extents cloned with older kernels.
604 	 */
605 	if (search_key.type == BTRFS_EXTENT_DATA_KEY &&
606 	    search_key.offset >= LLONG_MAX)
607 		search_key.offset = 0;
608 	path->lowest_level = level;
609 	if (time_seq == BTRFS_SEQ_LAST)
610 		ret = btrfs_search_slot(NULL, root, &search_key, path, 0, 0);
611 	else
612 		ret = btrfs_search_old_slot(root, &search_key, path, time_seq);
613 
614 	btrfs_debug(fs_info,
615 		"search slot in root %llu (level %d, ref count %d) returned %d for key (%llu %u %llu)",
616 		 ref->root_id, level, ref->count, ret,
617 		 ref->key_for_search.objectid, ref->key_for_search.type,
618 		 ref->key_for_search.offset);
619 	if (ret < 0)
620 		goto out;
621 
622 	eb = path->nodes[level];
623 	while (!eb) {
624 		if (WARN_ON(!level)) {
625 			ret = 1;
626 			goto out;
627 		}
628 		level--;
629 		eb = path->nodes[level];
630 	}
631 
632 	ret = add_all_parents(root, path, parents, preftrees, ref, level,
633 			      time_seq, extent_item_pos, ignore_offset);
634 out:
635 	btrfs_put_root(root);
636 out_free:
637 	path->lowest_level = 0;
638 	btrfs_release_path(path);
639 	return ret;
640 }
641 
642 static struct extent_inode_elem *
643 unode_aux_to_inode_list(struct ulist_node *node)
644 {
645 	if (!node)
646 		return NULL;
647 	return (struct extent_inode_elem *)(uintptr_t)node->aux;
648 }
649 
650 /*
651  * We maintain three separate rbtrees: one for direct refs, one for
652  * indirect refs which have a key, and one for indirect refs which do not
653  * have a key. Each tree does merge on insertion.
654  *
655  * Once all of the references are located, we iterate over the tree of
656  * indirect refs with missing keys. An appropriate key is located and
657  * the ref is moved onto the tree for indirect refs. After all missing
658  * keys are thus located, we iterate over the indirect ref tree, resolve
659  * each reference, and then insert the resolved reference onto the
660  * direct tree (merging there too).
661  *
662  * New backrefs (i.e., for parent nodes) are added to the appropriate
663  * rbtree as they are encountered. The new backrefs are subsequently
664  * resolved as above.
665  */
666 static int resolve_indirect_refs(struct btrfs_fs_info *fs_info,
667 				 struct btrfs_path *path, u64 time_seq,
668 				 struct preftrees *preftrees,
669 				 const u64 *extent_item_pos,
670 				 struct share_check *sc, bool ignore_offset)
671 {
672 	int err;
673 	int ret = 0;
674 	struct ulist *parents;
675 	struct ulist_node *node;
676 	struct ulist_iterator uiter;
677 	struct rb_node *rnode;
678 
679 	parents = ulist_alloc(GFP_NOFS);
680 	if (!parents)
681 		return -ENOMEM;
682 
683 	/*
684 	 * We could trade memory usage for performance here by iterating
685 	 * the tree, allocating new refs for each insertion, and then
686 	 * freeing the entire indirect tree when we're done.  In some test
687 	 * cases, the tree can grow quite large (~200k objects).
688 	 */
689 	while ((rnode = rb_first_cached(&preftrees->indirect.root))) {
690 		struct prelim_ref *ref;
691 
692 		ref = rb_entry(rnode, struct prelim_ref, rbnode);
693 		if (WARN(ref->parent,
694 			 "BUG: direct ref found in indirect tree")) {
695 			ret = -EINVAL;
696 			goto out;
697 		}
698 
699 		rb_erase_cached(&ref->rbnode, &preftrees->indirect.root);
700 		preftrees->indirect.count--;
701 
702 		if (ref->count == 0) {
703 			free_pref(ref);
704 			continue;
705 		}
706 
707 		if (sc && sc->root_objectid &&
708 		    ref->root_id != sc->root_objectid) {
709 			free_pref(ref);
710 			ret = BACKREF_FOUND_SHARED;
711 			goto out;
712 		}
713 		err = resolve_indirect_ref(fs_info, path, time_seq, preftrees,
714 					   ref, parents, extent_item_pos,
715 					   ignore_offset);
716 		/*
717 		 * we can only tolerate ENOENT,otherwise,we should catch error
718 		 * and return directly.
719 		 */
720 		if (err == -ENOENT) {
721 			prelim_ref_insert(fs_info, &preftrees->direct, ref,
722 					  NULL);
723 			continue;
724 		} else if (err) {
725 			free_pref(ref);
726 			ret = err;
727 			goto out;
728 		}
729 
730 		/* we put the first parent into the ref at hand */
731 		ULIST_ITER_INIT(&uiter);
732 		node = ulist_next(parents, &uiter);
733 		ref->parent = node ? node->val : 0;
734 		ref->inode_list = unode_aux_to_inode_list(node);
735 
736 		/* Add a prelim_ref(s) for any other parent(s). */
737 		while ((node = ulist_next(parents, &uiter))) {
738 			struct prelim_ref *new_ref;
739 
740 			new_ref = kmem_cache_alloc(btrfs_prelim_ref_cache,
741 						   GFP_NOFS);
742 			if (!new_ref) {
743 				free_pref(ref);
744 				ret = -ENOMEM;
745 				goto out;
746 			}
747 			memcpy(new_ref, ref, sizeof(*ref));
748 			new_ref->parent = node->val;
749 			new_ref->inode_list = unode_aux_to_inode_list(node);
750 			prelim_ref_insert(fs_info, &preftrees->direct,
751 					  new_ref, NULL);
752 		}
753 
754 		/*
755 		 * Now it's a direct ref, put it in the direct tree. We must
756 		 * do this last because the ref could be merged/freed here.
757 		 */
758 		prelim_ref_insert(fs_info, &preftrees->direct, ref, NULL);
759 
760 		ulist_reinit(parents);
761 		cond_resched();
762 	}
763 out:
764 	ulist_free(parents);
765 	return ret;
766 }
767 
768 /*
769  * read tree blocks and add keys where required.
770  */
771 static int add_missing_keys(struct btrfs_fs_info *fs_info,
772 			    struct preftrees *preftrees, bool lock)
773 {
774 	struct prelim_ref *ref;
775 	struct extent_buffer *eb;
776 	struct preftree *tree = &preftrees->indirect_missing_keys;
777 	struct rb_node *node;
778 
779 	while ((node = rb_first_cached(&tree->root))) {
780 		ref = rb_entry(node, struct prelim_ref, rbnode);
781 		rb_erase_cached(node, &tree->root);
782 
783 		BUG_ON(ref->parent);	/* should not be a direct ref */
784 		BUG_ON(ref->key_for_search.type);
785 		BUG_ON(!ref->wanted_disk_byte);
786 
787 		eb = read_tree_block(fs_info, ref->wanted_disk_byte,
788 				     ref->root_id, 0, ref->level - 1, NULL);
789 		if (IS_ERR(eb)) {
790 			free_pref(ref);
791 			return PTR_ERR(eb);
792 		} else if (!extent_buffer_uptodate(eb)) {
793 			free_pref(ref);
794 			free_extent_buffer(eb);
795 			return -EIO;
796 		}
797 		if (lock)
798 			btrfs_tree_read_lock(eb);
799 		if (btrfs_header_level(eb) == 0)
800 			btrfs_item_key_to_cpu(eb, &ref->key_for_search, 0);
801 		else
802 			btrfs_node_key_to_cpu(eb, &ref->key_for_search, 0);
803 		if (lock)
804 			btrfs_tree_read_unlock(eb);
805 		free_extent_buffer(eb);
806 		prelim_ref_insert(fs_info, &preftrees->indirect, ref, NULL);
807 		cond_resched();
808 	}
809 	return 0;
810 }
811 
812 /*
813  * add all currently queued delayed refs from this head whose seq nr is
814  * smaller or equal that seq to the list
815  */
816 static int add_delayed_refs(const struct btrfs_fs_info *fs_info,
817 			    struct btrfs_delayed_ref_head *head, u64 seq,
818 			    struct preftrees *preftrees, struct share_check *sc)
819 {
820 	struct btrfs_delayed_ref_node *node;
821 	struct btrfs_delayed_extent_op *extent_op = head->extent_op;
822 	struct btrfs_key key;
823 	struct btrfs_key tmp_op_key;
824 	struct rb_node *n;
825 	int count;
826 	int ret = 0;
827 
828 	if (extent_op && extent_op->update_key)
829 		btrfs_disk_key_to_cpu(&tmp_op_key, &extent_op->key);
830 
831 	spin_lock(&head->lock);
832 	for (n = rb_first_cached(&head->ref_tree); n; n = rb_next(n)) {
833 		node = rb_entry(n, struct btrfs_delayed_ref_node,
834 				ref_node);
835 		if (node->seq > seq)
836 			continue;
837 
838 		switch (node->action) {
839 		case BTRFS_ADD_DELAYED_EXTENT:
840 		case BTRFS_UPDATE_DELAYED_HEAD:
841 			WARN_ON(1);
842 			continue;
843 		case BTRFS_ADD_DELAYED_REF:
844 			count = node->ref_mod;
845 			break;
846 		case BTRFS_DROP_DELAYED_REF:
847 			count = node->ref_mod * -1;
848 			break;
849 		default:
850 			BUG();
851 		}
852 		switch (node->type) {
853 		case BTRFS_TREE_BLOCK_REF_KEY: {
854 			/* NORMAL INDIRECT METADATA backref */
855 			struct btrfs_delayed_tree_ref *ref;
856 
857 			ref = btrfs_delayed_node_to_tree_ref(node);
858 			ret = add_indirect_ref(fs_info, preftrees, ref->root,
859 					       &tmp_op_key, ref->level + 1,
860 					       node->bytenr, count, sc,
861 					       GFP_ATOMIC);
862 			break;
863 		}
864 		case BTRFS_SHARED_BLOCK_REF_KEY: {
865 			/* SHARED DIRECT METADATA backref */
866 			struct btrfs_delayed_tree_ref *ref;
867 
868 			ref = btrfs_delayed_node_to_tree_ref(node);
869 
870 			ret = add_direct_ref(fs_info, preftrees, ref->level + 1,
871 					     ref->parent, node->bytenr, count,
872 					     sc, GFP_ATOMIC);
873 			break;
874 		}
875 		case BTRFS_EXTENT_DATA_REF_KEY: {
876 			/* NORMAL INDIRECT DATA backref */
877 			struct btrfs_delayed_data_ref *ref;
878 			ref = btrfs_delayed_node_to_data_ref(node);
879 
880 			key.objectid = ref->objectid;
881 			key.type = BTRFS_EXTENT_DATA_KEY;
882 			key.offset = ref->offset;
883 
884 			/*
885 			 * Found a inum that doesn't match our known inum, we
886 			 * know it's shared.
887 			 */
888 			if (sc && sc->inum && ref->objectid != sc->inum) {
889 				ret = BACKREF_FOUND_SHARED;
890 				goto out;
891 			}
892 
893 			ret = add_indirect_ref(fs_info, preftrees, ref->root,
894 					       &key, 0, node->bytenr, count, sc,
895 					       GFP_ATOMIC);
896 			break;
897 		}
898 		case BTRFS_SHARED_DATA_REF_KEY: {
899 			/* SHARED DIRECT FULL backref */
900 			struct btrfs_delayed_data_ref *ref;
901 
902 			ref = btrfs_delayed_node_to_data_ref(node);
903 
904 			ret = add_direct_ref(fs_info, preftrees, 0, ref->parent,
905 					     node->bytenr, count, sc,
906 					     GFP_ATOMIC);
907 			break;
908 		}
909 		default:
910 			WARN_ON(1);
911 		}
912 		/*
913 		 * We must ignore BACKREF_FOUND_SHARED until all delayed
914 		 * refs have been checked.
915 		 */
916 		if (ret && (ret != BACKREF_FOUND_SHARED))
917 			break;
918 	}
919 	if (!ret)
920 		ret = extent_is_shared(sc);
921 out:
922 	spin_unlock(&head->lock);
923 	return ret;
924 }
925 
926 /*
927  * add all inline backrefs for bytenr to the list
928  *
929  * Returns 0 on success, <0 on error, or BACKREF_FOUND_SHARED.
930  */
931 static int add_inline_refs(const struct btrfs_fs_info *fs_info,
932 			   struct btrfs_path *path, u64 bytenr,
933 			   int *info_level, struct preftrees *preftrees,
934 			   struct share_check *sc)
935 {
936 	int ret = 0;
937 	int slot;
938 	struct extent_buffer *leaf;
939 	struct btrfs_key key;
940 	struct btrfs_key found_key;
941 	unsigned long ptr;
942 	unsigned long end;
943 	struct btrfs_extent_item *ei;
944 	u64 flags;
945 	u64 item_size;
946 
947 	/*
948 	 * enumerate all inline refs
949 	 */
950 	leaf = path->nodes[0];
951 	slot = path->slots[0];
952 
953 	item_size = btrfs_item_size_nr(leaf, slot);
954 	BUG_ON(item_size < sizeof(*ei));
955 
956 	ei = btrfs_item_ptr(leaf, slot, struct btrfs_extent_item);
957 	flags = btrfs_extent_flags(leaf, ei);
958 	btrfs_item_key_to_cpu(leaf, &found_key, slot);
959 
960 	ptr = (unsigned long)(ei + 1);
961 	end = (unsigned long)ei + item_size;
962 
963 	if (found_key.type == BTRFS_EXTENT_ITEM_KEY &&
964 	    flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) {
965 		struct btrfs_tree_block_info *info;
966 
967 		info = (struct btrfs_tree_block_info *)ptr;
968 		*info_level = btrfs_tree_block_level(leaf, info);
969 		ptr += sizeof(struct btrfs_tree_block_info);
970 		BUG_ON(ptr > end);
971 	} else if (found_key.type == BTRFS_METADATA_ITEM_KEY) {
972 		*info_level = found_key.offset;
973 	} else {
974 		BUG_ON(!(flags & BTRFS_EXTENT_FLAG_DATA));
975 	}
976 
977 	while (ptr < end) {
978 		struct btrfs_extent_inline_ref *iref;
979 		u64 offset;
980 		int type;
981 
982 		iref = (struct btrfs_extent_inline_ref *)ptr;
983 		type = btrfs_get_extent_inline_ref_type(leaf, iref,
984 							BTRFS_REF_TYPE_ANY);
985 		if (type == BTRFS_REF_TYPE_INVALID)
986 			return -EUCLEAN;
987 
988 		offset = btrfs_extent_inline_ref_offset(leaf, iref);
989 
990 		switch (type) {
991 		case BTRFS_SHARED_BLOCK_REF_KEY:
992 			ret = add_direct_ref(fs_info, preftrees,
993 					     *info_level + 1, offset,
994 					     bytenr, 1, NULL, GFP_NOFS);
995 			break;
996 		case BTRFS_SHARED_DATA_REF_KEY: {
997 			struct btrfs_shared_data_ref *sdref;
998 			int count;
999 
1000 			sdref = (struct btrfs_shared_data_ref *)(iref + 1);
1001 			count = btrfs_shared_data_ref_count(leaf, sdref);
1002 
1003 			ret = add_direct_ref(fs_info, preftrees, 0, offset,
1004 					     bytenr, count, sc, GFP_NOFS);
1005 			break;
1006 		}
1007 		case BTRFS_TREE_BLOCK_REF_KEY:
1008 			ret = add_indirect_ref(fs_info, preftrees, offset,
1009 					       NULL, *info_level + 1,
1010 					       bytenr, 1, NULL, GFP_NOFS);
1011 			break;
1012 		case BTRFS_EXTENT_DATA_REF_KEY: {
1013 			struct btrfs_extent_data_ref *dref;
1014 			int count;
1015 			u64 root;
1016 
1017 			dref = (struct btrfs_extent_data_ref *)(&iref->offset);
1018 			count = btrfs_extent_data_ref_count(leaf, dref);
1019 			key.objectid = btrfs_extent_data_ref_objectid(leaf,
1020 								      dref);
1021 			key.type = BTRFS_EXTENT_DATA_KEY;
1022 			key.offset = btrfs_extent_data_ref_offset(leaf, dref);
1023 
1024 			if (sc && sc->inum && key.objectid != sc->inum) {
1025 				ret = BACKREF_FOUND_SHARED;
1026 				break;
1027 			}
1028 
1029 			root = btrfs_extent_data_ref_root(leaf, dref);
1030 
1031 			ret = add_indirect_ref(fs_info, preftrees, root,
1032 					       &key, 0, bytenr, count,
1033 					       sc, GFP_NOFS);
1034 			break;
1035 		}
1036 		default:
1037 			WARN_ON(1);
1038 		}
1039 		if (ret)
1040 			return ret;
1041 		ptr += btrfs_extent_inline_ref_size(type);
1042 	}
1043 
1044 	return 0;
1045 }
1046 
1047 /*
1048  * add all non-inline backrefs for bytenr to the list
1049  *
1050  * Returns 0 on success, <0 on error, or BACKREF_FOUND_SHARED.
1051  */
1052 static int add_keyed_refs(struct btrfs_fs_info *fs_info,
1053 			  struct btrfs_path *path, u64 bytenr,
1054 			  int info_level, struct preftrees *preftrees,
1055 			  struct share_check *sc)
1056 {
1057 	struct btrfs_root *extent_root = fs_info->extent_root;
1058 	int ret;
1059 	int slot;
1060 	struct extent_buffer *leaf;
1061 	struct btrfs_key key;
1062 
1063 	while (1) {
1064 		ret = btrfs_next_item(extent_root, path);
1065 		if (ret < 0)
1066 			break;
1067 		if (ret) {
1068 			ret = 0;
1069 			break;
1070 		}
1071 
1072 		slot = path->slots[0];
1073 		leaf = path->nodes[0];
1074 		btrfs_item_key_to_cpu(leaf, &key, slot);
1075 
1076 		if (key.objectid != bytenr)
1077 			break;
1078 		if (key.type < BTRFS_TREE_BLOCK_REF_KEY)
1079 			continue;
1080 		if (key.type > BTRFS_SHARED_DATA_REF_KEY)
1081 			break;
1082 
1083 		switch (key.type) {
1084 		case BTRFS_SHARED_BLOCK_REF_KEY:
1085 			/* SHARED DIRECT METADATA backref */
1086 			ret = add_direct_ref(fs_info, preftrees,
1087 					     info_level + 1, key.offset,
1088 					     bytenr, 1, NULL, GFP_NOFS);
1089 			break;
1090 		case BTRFS_SHARED_DATA_REF_KEY: {
1091 			/* SHARED DIRECT FULL backref */
1092 			struct btrfs_shared_data_ref *sdref;
1093 			int count;
1094 
1095 			sdref = btrfs_item_ptr(leaf, slot,
1096 					      struct btrfs_shared_data_ref);
1097 			count = btrfs_shared_data_ref_count(leaf, sdref);
1098 			ret = add_direct_ref(fs_info, preftrees, 0,
1099 					     key.offset, bytenr, count,
1100 					     sc, GFP_NOFS);
1101 			break;
1102 		}
1103 		case BTRFS_TREE_BLOCK_REF_KEY:
1104 			/* NORMAL INDIRECT METADATA backref */
1105 			ret = add_indirect_ref(fs_info, preftrees, key.offset,
1106 					       NULL, info_level + 1, bytenr,
1107 					       1, NULL, GFP_NOFS);
1108 			break;
1109 		case BTRFS_EXTENT_DATA_REF_KEY: {
1110 			/* NORMAL INDIRECT DATA backref */
1111 			struct btrfs_extent_data_ref *dref;
1112 			int count;
1113 			u64 root;
1114 
1115 			dref = btrfs_item_ptr(leaf, slot,
1116 					      struct btrfs_extent_data_ref);
1117 			count = btrfs_extent_data_ref_count(leaf, dref);
1118 			key.objectid = btrfs_extent_data_ref_objectid(leaf,
1119 								      dref);
1120 			key.type = BTRFS_EXTENT_DATA_KEY;
1121 			key.offset = btrfs_extent_data_ref_offset(leaf, dref);
1122 
1123 			if (sc && sc->inum && key.objectid != sc->inum) {
1124 				ret = BACKREF_FOUND_SHARED;
1125 				break;
1126 			}
1127 
1128 			root = btrfs_extent_data_ref_root(leaf, dref);
1129 			ret = add_indirect_ref(fs_info, preftrees, root,
1130 					       &key, 0, bytenr, count,
1131 					       sc, GFP_NOFS);
1132 			break;
1133 		}
1134 		default:
1135 			WARN_ON(1);
1136 		}
1137 		if (ret)
1138 			return ret;
1139 
1140 	}
1141 
1142 	return ret;
1143 }
1144 
1145 /*
1146  * this adds all existing backrefs (inline backrefs, backrefs and delayed
1147  * refs) for the given bytenr to the refs list, merges duplicates and resolves
1148  * indirect refs to their parent bytenr.
1149  * When roots are found, they're added to the roots list
1150  *
1151  * If time_seq is set to BTRFS_SEQ_LAST, it will not search delayed_refs, and
1152  * behave much like trans == NULL case, the difference only lies in it will not
1153  * commit root.
1154  * The special case is for qgroup to search roots in commit_transaction().
1155  *
1156  * @sc - if !NULL, then immediately return BACKREF_FOUND_SHARED when a
1157  * shared extent is detected.
1158  *
1159  * Otherwise this returns 0 for success and <0 for an error.
1160  *
1161  * If ignore_offset is set to false, only extent refs whose offsets match
1162  * extent_item_pos are returned.  If true, every extent ref is returned
1163  * and extent_item_pos is ignored.
1164  *
1165  * FIXME some caching might speed things up
1166  */
1167 static int find_parent_nodes(struct btrfs_trans_handle *trans,
1168 			     struct btrfs_fs_info *fs_info, u64 bytenr,
1169 			     u64 time_seq, struct ulist *refs,
1170 			     struct ulist *roots, const u64 *extent_item_pos,
1171 			     struct share_check *sc, bool ignore_offset)
1172 {
1173 	struct btrfs_key key;
1174 	struct btrfs_path *path;
1175 	struct btrfs_delayed_ref_root *delayed_refs = NULL;
1176 	struct btrfs_delayed_ref_head *head;
1177 	int info_level = 0;
1178 	int ret;
1179 	struct prelim_ref *ref;
1180 	struct rb_node *node;
1181 	struct extent_inode_elem *eie = NULL;
1182 	struct preftrees preftrees = {
1183 		.direct = PREFTREE_INIT,
1184 		.indirect = PREFTREE_INIT,
1185 		.indirect_missing_keys = PREFTREE_INIT
1186 	};
1187 
1188 	key.objectid = bytenr;
1189 	key.offset = (u64)-1;
1190 	if (btrfs_fs_incompat(fs_info, SKINNY_METADATA))
1191 		key.type = BTRFS_METADATA_ITEM_KEY;
1192 	else
1193 		key.type = BTRFS_EXTENT_ITEM_KEY;
1194 
1195 	path = btrfs_alloc_path();
1196 	if (!path)
1197 		return -ENOMEM;
1198 	if (!trans) {
1199 		path->search_commit_root = 1;
1200 		path->skip_locking = 1;
1201 	}
1202 
1203 	if (time_seq == BTRFS_SEQ_LAST)
1204 		path->skip_locking = 1;
1205 
1206 	/*
1207 	 * grab both a lock on the path and a lock on the delayed ref head.
1208 	 * We need both to get a consistent picture of how the refs look
1209 	 * at a specified point in time
1210 	 */
1211 again:
1212 	head = NULL;
1213 
1214 	ret = btrfs_search_slot(trans, fs_info->extent_root, &key, path, 0, 0);
1215 	if (ret < 0)
1216 		goto out;
1217 	BUG_ON(ret == 0);
1218 
1219 #ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
1220 	if (trans && likely(trans->type != __TRANS_DUMMY) &&
1221 	    time_seq != BTRFS_SEQ_LAST) {
1222 #else
1223 	if (trans && time_seq != BTRFS_SEQ_LAST) {
1224 #endif
1225 		/*
1226 		 * look if there are updates for this ref queued and lock the
1227 		 * head
1228 		 */
1229 		delayed_refs = &trans->transaction->delayed_refs;
1230 		spin_lock(&delayed_refs->lock);
1231 		head = btrfs_find_delayed_ref_head(delayed_refs, bytenr);
1232 		if (head) {
1233 			if (!mutex_trylock(&head->mutex)) {
1234 				refcount_inc(&head->refs);
1235 				spin_unlock(&delayed_refs->lock);
1236 
1237 				btrfs_release_path(path);
1238 
1239 				/*
1240 				 * Mutex was contended, block until it's
1241 				 * released and try again
1242 				 */
1243 				mutex_lock(&head->mutex);
1244 				mutex_unlock(&head->mutex);
1245 				btrfs_put_delayed_ref_head(head);
1246 				goto again;
1247 			}
1248 			spin_unlock(&delayed_refs->lock);
1249 			ret = add_delayed_refs(fs_info, head, time_seq,
1250 					       &preftrees, sc);
1251 			mutex_unlock(&head->mutex);
1252 			if (ret)
1253 				goto out;
1254 		} else {
1255 			spin_unlock(&delayed_refs->lock);
1256 		}
1257 	}
1258 
1259 	if (path->slots[0]) {
1260 		struct extent_buffer *leaf;
1261 		int slot;
1262 
1263 		path->slots[0]--;
1264 		leaf = path->nodes[0];
1265 		slot = path->slots[0];
1266 		btrfs_item_key_to_cpu(leaf, &key, slot);
1267 		if (key.objectid == bytenr &&
1268 		    (key.type == BTRFS_EXTENT_ITEM_KEY ||
1269 		     key.type == BTRFS_METADATA_ITEM_KEY)) {
1270 			ret = add_inline_refs(fs_info, path, bytenr,
1271 					      &info_level, &preftrees, sc);
1272 			if (ret)
1273 				goto out;
1274 			ret = add_keyed_refs(fs_info, path, bytenr, info_level,
1275 					     &preftrees, sc);
1276 			if (ret)
1277 				goto out;
1278 		}
1279 	}
1280 
1281 	btrfs_release_path(path);
1282 
1283 	ret = add_missing_keys(fs_info, &preftrees, path->skip_locking == 0);
1284 	if (ret)
1285 		goto out;
1286 
1287 	WARN_ON(!RB_EMPTY_ROOT(&preftrees.indirect_missing_keys.root.rb_root));
1288 
1289 	ret = resolve_indirect_refs(fs_info, path, time_seq, &preftrees,
1290 				    extent_item_pos, sc, ignore_offset);
1291 	if (ret)
1292 		goto out;
1293 
1294 	WARN_ON(!RB_EMPTY_ROOT(&preftrees.indirect.root.rb_root));
1295 
1296 	/*
1297 	 * This walks the tree of merged and resolved refs. Tree blocks are
1298 	 * read in as needed. Unique entries are added to the ulist, and
1299 	 * the list of found roots is updated.
1300 	 *
1301 	 * We release the entire tree in one go before returning.
1302 	 */
1303 	node = rb_first_cached(&preftrees.direct.root);
1304 	while (node) {
1305 		ref = rb_entry(node, struct prelim_ref, rbnode);
1306 		node = rb_next(&ref->rbnode);
1307 		/*
1308 		 * ref->count < 0 can happen here if there are delayed
1309 		 * refs with a node->action of BTRFS_DROP_DELAYED_REF.
1310 		 * prelim_ref_insert() relies on this when merging
1311 		 * identical refs to keep the overall count correct.
1312 		 * prelim_ref_insert() will merge only those refs
1313 		 * which compare identically.  Any refs having
1314 		 * e.g. different offsets would not be merged,
1315 		 * and would retain their original ref->count < 0.
1316 		 */
1317 		if (roots && ref->count && ref->root_id && ref->parent == 0) {
1318 			if (sc && sc->root_objectid &&
1319 			    ref->root_id != sc->root_objectid) {
1320 				ret = BACKREF_FOUND_SHARED;
1321 				goto out;
1322 			}
1323 
1324 			/* no parent == root of tree */
1325 			ret = ulist_add(roots, ref->root_id, 0, GFP_NOFS);
1326 			if (ret < 0)
1327 				goto out;
1328 		}
1329 		if (ref->count && ref->parent) {
1330 			if (extent_item_pos && !ref->inode_list &&
1331 			    ref->level == 0) {
1332 				struct extent_buffer *eb;
1333 
1334 				eb = read_tree_block(fs_info, ref->parent, 0,
1335 						     0, ref->level, NULL);
1336 				if (IS_ERR(eb)) {
1337 					ret = PTR_ERR(eb);
1338 					goto out;
1339 				} else if (!extent_buffer_uptodate(eb)) {
1340 					free_extent_buffer(eb);
1341 					ret = -EIO;
1342 					goto out;
1343 				}
1344 
1345 				if (!path->skip_locking)
1346 					btrfs_tree_read_lock(eb);
1347 				ret = find_extent_in_eb(eb, bytenr,
1348 							*extent_item_pos, &eie, ignore_offset);
1349 				if (!path->skip_locking)
1350 					btrfs_tree_read_unlock(eb);
1351 				free_extent_buffer(eb);
1352 				if (ret < 0)
1353 					goto out;
1354 				ref->inode_list = eie;
1355 			}
1356 			ret = ulist_add_merge_ptr(refs, ref->parent,
1357 						  ref->inode_list,
1358 						  (void **)&eie, GFP_NOFS);
1359 			if (ret < 0)
1360 				goto out;
1361 			if (!ret && extent_item_pos) {
1362 				/*
1363 				 * we've recorded that parent, so we must extend
1364 				 * its inode list here
1365 				 */
1366 				BUG_ON(!eie);
1367 				while (eie->next)
1368 					eie = eie->next;
1369 				eie->next = ref->inode_list;
1370 			}
1371 			eie = NULL;
1372 		}
1373 		cond_resched();
1374 	}
1375 
1376 out:
1377 	btrfs_free_path(path);
1378 
1379 	prelim_release(&preftrees.direct);
1380 	prelim_release(&preftrees.indirect);
1381 	prelim_release(&preftrees.indirect_missing_keys);
1382 
1383 	if (ret < 0)
1384 		free_inode_elem_list(eie);
1385 	return ret;
1386 }
1387 
1388 static void free_leaf_list(struct ulist *blocks)
1389 {
1390 	struct ulist_node *node = NULL;
1391 	struct extent_inode_elem *eie;
1392 	struct ulist_iterator uiter;
1393 
1394 	ULIST_ITER_INIT(&uiter);
1395 	while ((node = ulist_next(blocks, &uiter))) {
1396 		if (!node->aux)
1397 			continue;
1398 		eie = unode_aux_to_inode_list(node);
1399 		free_inode_elem_list(eie);
1400 		node->aux = 0;
1401 	}
1402 
1403 	ulist_free(blocks);
1404 }
1405 
1406 /*
1407  * Finds all leafs with a reference to the specified combination of bytenr and
1408  * offset. key_list_head will point to a list of corresponding keys (caller must
1409  * free each list element). The leafs will be stored in the leafs ulist, which
1410  * must be freed with ulist_free.
1411  *
1412  * returns 0 on success, <0 on error
1413  */
1414 int btrfs_find_all_leafs(struct btrfs_trans_handle *trans,
1415 			 struct btrfs_fs_info *fs_info, u64 bytenr,
1416 			 u64 time_seq, struct ulist **leafs,
1417 			 const u64 *extent_item_pos, bool ignore_offset)
1418 {
1419 	int ret;
1420 
1421 	*leafs = ulist_alloc(GFP_NOFS);
1422 	if (!*leafs)
1423 		return -ENOMEM;
1424 
1425 	ret = find_parent_nodes(trans, fs_info, bytenr, time_seq,
1426 				*leafs, NULL, extent_item_pos, NULL, ignore_offset);
1427 	if (ret < 0 && ret != -ENOENT) {
1428 		free_leaf_list(*leafs);
1429 		return ret;
1430 	}
1431 
1432 	return 0;
1433 }
1434 
1435 /*
1436  * walk all backrefs for a given extent to find all roots that reference this
1437  * extent. Walking a backref means finding all extents that reference this
1438  * extent and in turn walk the backrefs of those, too. Naturally this is a
1439  * recursive process, but here it is implemented in an iterative fashion: We
1440  * find all referencing extents for the extent in question and put them on a
1441  * list. In turn, we find all referencing extents for those, further appending
1442  * to the list. The way we iterate the list allows adding more elements after
1443  * the current while iterating. The process stops when we reach the end of the
1444  * list. Found roots are added to the roots list.
1445  *
1446  * returns 0 on success, < 0 on error.
1447  */
1448 static int btrfs_find_all_roots_safe(struct btrfs_trans_handle *trans,
1449 				     struct btrfs_fs_info *fs_info, u64 bytenr,
1450 				     u64 time_seq, struct ulist **roots,
1451 				     bool ignore_offset)
1452 {
1453 	struct ulist *tmp;
1454 	struct ulist_node *node = NULL;
1455 	struct ulist_iterator uiter;
1456 	int ret;
1457 
1458 	tmp = ulist_alloc(GFP_NOFS);
1459 	if (!tmp)
1460 		return -ENOMEM;
1461 	*roots = ulist_alloc(GFP_NOFS);
1462 	if (!*roots) {
1463 		ulist_free(tmp);
1464 		return -ENOMEM;
1465 	}
1466 
1467 	ULIST_ITER_INIT(&uiter);
1468 	while (1) {
1469 		ret = find_parent_nodes(trans, fs_info, bytenr, time_seq,
1470 					tmp, *roots, NULL, NULL, ignore_offset);
1471 		if (ret < 0 && ret != -ENOENT) {
1472 			ulist_free(tmp);
1473 			ulist_free(*roots);
1474 			*roots = NULL;
1475 			return ret;
1476 		}
1477 		node = ulist_next(tmp, &uiter);
1478 		if (!node)
1479 			break;
1480 		bytenr = node->val;
1481 		cond_resched();
1482 	}
1483 
1484 	ulist_free(tmp);
1485 	return 0;
1486 }
1487 
1488 int btrfs_find_all_roots(struct btrfs_trans_handle *trans,
1489 			 struct btrfs_fs_info *fs_info, u64 bytenr,
1490 			 u64 time_seq, struct ulist **roots,
1491 			 bool ignore_offset)
1492 {
1493 	int ret;
1494 
1495 	if (!trans)
1496 		down_read(&fs_info->commit_root_sem);
1497 	ret = btrfs_find_all_roots_safe(trans, fs_info, bytenr,
1498 					time_seq, roots, ignore_offset);
1499 	if (!trans)
1500 		up_read(&fs_info->commit_root_sem);
1501 	return ret;
1502 }
1503 
1504 /**
1505  * Check if an extent is shared or not
1506  *
1507  * @root:   root inode belongs to
1508  * @inum:   inode number of the inode whose extent we are checking
1509  * @bytenr: logical bytenr of the extent we are checking
1510  * @roots:  list of roots this extent is shared among
1511  * @tmp:    temporary list used for iteration
1512  *
1513  * btrfs_check_shared uses the backref walking code but will short
1514  * circuit as soon as it finds a root or inode that doesn't match the
1515  * one passed in. This provides a significant performance benefit for
1516  * callers (such as fiemap) which want to know whether the extent is
1517  * shared but do not need a ref count.
1518  *
1519  * This attempts to attach to the running transaction in order to account for
1520  * delayed refs, but continues on even when no running transaction exists.
1521  *
1522  * Return: 0 if extent is not shared, 1 if it is shared, < 0 on error.
1523  */
1524 int btrfs_check_shared(struct btrfs_root *root, u64 inum, u64 bytenr,
1525 		struct ulist *roots, struct ulist *tmp)
1526 {
1527 	struct btrfs_fs_info *fs_info = root->fs_info;
1528 	struct btrfs_trans_handle *trans;
1529 	struct ulist_iterator uiter;
1530 	struct ulist_node *node;
1531 	struct btrfs_seq_list elem = BTRFS_SEQ_LIST_INIT(elem);
1532 	int ret = 0;
1533 	struct share_check shared = {
1534 		.root_objectid = root->root_key.objectid,
1535 		.inum = inum,
1536 		.share_count = 0,
1537 	};
1538 
1539 	ulist_init(roots);
1540 	ulist_init(tmp);
1541 
1542 	trans = btrfs_join_transaction_nostart(root);
1543 	if (IS_ERR(trans)) {
1544 		if (PTR_ERR(trans) != -ENOENT && PTR_ERR(trans) != -EROFS) {
1545 			ret = PTR_ERR(trans);
1546 			goto out;
1547 		}
1548 		trans = NULL;
1549 		down_read(&fs_info->commit_root_sem);
1550 	} else {
1551 		btrfs_get_tree_mod_seq(fs_info, &elem);
1552 	}
1553 
1554 	ULIST_ITER_INIT(&uiter);
1555 	while (1) {
1556 		ret = find_parent_nodes(trans, fs_info, bytenr, elem.seq, tmp,
1557 					roots, NULL, &shared, false);
1558 		if (ret == BACKREF_FOUND_SHARED) {
1559 			/* this is the only condition under which we return 1 */
1560 			ret = 1;
1561 			break;
1562 		}
1563 		if (ret < 0 && ret != -ENOENT)
1564 			break;
1565 		ret = 0;
1566 		node = ulist_next(tmp, &uiter);
1567 		if (!node)
1568 			break;
1569 		bytenr = node->val;
1570 		shared.share_count = 0;
1571 		cond_resched();
1572 	}
1573 
1574 	if (trans) {
1575 		btrfs_put_tree_mod_seq(fs_info, &elem);
1576 		btrfs_end_transaction(trans);
1577 	} else {
1578 		up_read(&fs_info->commit_root_sem);
1579 	}
1580 out:
1581 	ulist_release(roots);
1582 	ulist_release(tmp);
1583 	return ret;
1584 }
1585 
1586 int btrfs_find_one_extref(struct btrfs_root *root, u64 inode_objectid,
1587 			  u64 start_off, struct btrfs_path *path,
1588 			  struct btrfs_inode_extref **ret_extref,
1589 			  u64 *found_off)
1590 {
1591 	int ret, slot;
1592 	struct btrfs_key key;
1593 	struct btrfs_key found_key;
1594 	struct btrfs_inode_extref *extref;
1595 	const struct extent_buffer *leaf;
1596 	unsigned long ptr;
1597 
1598 	key.objectid = inode_objectid;
1599 	key.type = BTRFS_INODE_EXTREF_KEY;
1600 	key.offset = start_off;
1601 
1602 	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
1603 	if (ret < 0)
1604 		return ret;
1605 
1606 	while (1) {
1607 		leaf = path->nodes[0];
1608 		slot = path->slots[0];
1609 		if (slot >= btrfs_header_nritems(leaf)) {
1610 			/*
1611 			 * If the item at offset is not found,
1612 			 * btrfs_search_slot will point us to the slot
1613 			 * where it should be inserted. In our case
1614 			 * that will be the slot directly before the
1615 			 * next INODE_REF_KEY_V2 item. In the case
1616 			 * that we're pointing to the last slot in a
1617 			 * leaf, we must move one leaf over.
1618 			 */
1619 			ret = btrfs_next_leaf(root, path);
1620 			if (ret) {
1621 				if (ret >= 1)
1622 					ret = -ENOENT;
1623 				break;
1624 			}
1625 			continue;
1626 		}
1627 
1628 		btrfs_item_key_to_cpu(leaf, &found_key, slot);
1629 
1630 		/*
1631 		 * Check that we're still looking at an extended ref key for
1632 		 * this particular objectid. If we have different
1633 		 * objectid or type then there are no more to be found
1634 		 * in the tree and we can exit.
1635 		 */
1636 		ret = -ENOENT;
1637 		if (found_key.objectid != inode_objectid)
1638 			break;
1639 		if (found_key.type != BTRFS_INODE_EXTREF_KEY)
1640 			break;
1641 
1642 		ret = 0;
1643 		ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
1644 		extref = (struct btrfs_inode_extref *)ptr;
1645 		*ret_extref = extref;
1646 		if (found_off)
1647 			*found_off = found_key.offset;
1648 		break;
1649 	}
1650 
1651 	return ret;
1652 }
1653 
1654 /*
1655  * this iterates to turn a name (from iref/extref) into a full filesystem path.
1656  * Elements of the path are separated by '/' and the path is guaranteed to be
1657  * 0-terminated. the path is only given within the current file system.
1658  * Therefore, it never starts with a '/'. the caller is responsible to provide
1659  * "size" bytes in "dest". the dest buffer will be filled backwards. finally,
1660  * the start point of the resulting string is returned. this pointer is within
1661  * dest, normally.
1662  * in case the path buffer would overflow, the pointer is decremented further
1663  * as if output was written to the buffer, though no more output is actually
1664  * generated. that way, the caller can determine how much space would be
1665  * required for the path to fit into the buffer. in that case, the returned
1666  * value will be smaller than dest. callers must check this!
1667  */
1668 char *btrfs_ref_to_path(struct btrfs_root *fs_root, struct btrfs_path *path,
1669 			u32 name_len, unsigned long name_off,
1670 			struct extent_buffer *eb_in, u64 parent,
1671 			char *dest, u32 size)
1672 {
1673 	int slot;
1674 	u64 next_inum;
1675 	int ret;
1676 	s64 bytes_left = ((s64)size) - 1;
1677 	struct extent_buffer *eb = eb_in;
1678 	struct btrfs_key found_key;
1679 	struct btrfs_inode_ref *iref;
1680 
1681 	if (bytes_left >= 0)
1682 		dest[bytes_left] = '\0';
1683 
1684 	while (1) {
1685 		bytes_left -= name_len;
1686 		if (bytes_left >= 0)
1687 			read_extent_buffer(eb, dest + bytes_left,
1688 					   name_off, name_len);
1689 		if (eb != eb_in) {
1690 			if (!path->skip_locking)
1691 				btrfs_tree_read_unlock(eb);
1692 			free_extent_buffer(eb);
1693 		}
1694 		ret = btrfs_find_item(fs_root, path, parent, 0,
1695 				BTRFS_INODE_REF_KEY, &found_key);
1696 		if (ret > 0)
1697 			ret = -ENOENT;
1698 		if (ret)
1699 			break;
1700 
1701 		next_inum = found_key.offset;
1702 
1703 		/* regular exit ahead */
1704 		if (parent == next_inum)
1705 			break;
1706 
1707 		slot = path->slots[0];
1708 		eb = path->nodes[0];
1709 		/* make sure we can use eb after releasing the path */
1710 		if (eb != eb_in) {
1711 			path->nodes[0] = NULL;
1712 			path->locks[0] = 0;
1713 		}
1714 		btrfs_release_path(path);
1715 		iref = btrfs_item_ptr(eb, slot, struct btrfs_inode_ref);
1716 
1717 		name_len = btrfs_inode_ref_name_len(eb, iref);
1718 		name_off = (unsigned long)(iref + 1);
1719 
1720 		parent = next_inum;
1721 		--bytes_left;
1722 		if (bytes_left >= 0)
1723 			dest[bytes_left] = '/';
1724 	}
1725 
1726 	btrfs_release_path(path);
1727 
1728 	if (ret)
1729 		return ERR_PTR(ret);
1730 
1731 	return dest + bytes_left;
1732 }
1733 
1734 /*
1735  * this makes the path point to (logical EXTENT_ITEM *)
1736  * returns BTRFS_EXTENT_FLAG_DATA for data, BTRFS_EXTENT_FLAG_TREE_BLOCK for
1737  * tree blocks and <0 on error.
1738  */
1739 int extent_from_logical(struct btrfs_fs_info *fs_info, u64 logical,
1740 			struct btrfs_path *path, struct btrfs_key *found_key,
1741 			u64 *flags_ret)
1742 {
1743 	int ret;
1744 	u64 flags;
1745 	u64 size = 0;
1746 	u32 item_size;
1747 	const struct extent_buffer *eb;
1748 	struct btrfs_extent_item *ei;
1749 	struct btrfs_key key;
1750 
1751 	if (btrfs_fs_incompat(fs_info, SKINNY_METADATA))
1752 		key.type = BTRFS_METADATA_ITEM_KEY;
1753 	else
1754 		key.type = BTRFS_EXTENT_ITEM_KEY;
1755 	key.objectid = logical;
1756 	key.offset = (u64)-1;
1757 
1758 	ret = btrfs_search_slot(NULL, fs_info->extent_root, &key, path, 0, 0);
1759 	if (ret < 0)
1760 		return ret;
1761 
1762 	ret = btrfs_previous_extent_item(fs_info->extent_root, path, 0);
1763 	if (ret) {
1764 		if (ret > 0)
1765 			ret = -ENOENT;
1766 		return ret;
1767 	}
1768 	btrfs_item_key_to_cpu(path->nodes[0], found_key, path->slots[0]);
1769 	if (found_key->type == BTRFS_METADATA_ITEM_KEY)
1770 		size = fs_info->nodesize;
1771 	else if (found_key->type == BTRFS_EXTENT_ITEM_KEY)
1772 		size = found_key->offset;
1773 
1774 	if (found_key->objectid > logical ||
1775 	    found_key->objectid + size <= logical) {
1776 		btrfs_debug(fs_info,
1777 			"logical %llu is not within any extent", logical);
1778 		return -ENOENT;
1779 	}
1780 
1781 	eb = path->nodes[0];
1782 	item_size = btrfs_item_size_nr(eb, path->slots[0]);
1783 	BUG_ON(item_size < sizeof(*ei));
1784 
1785 	ei = btrfs_item_ptr(eb, path->slots[0], struct btrfs_extent_item);
1786 	flags = btrfs_extent_flags(eb, ei);
1787 
1788 	btrfs_debug(fs_info,
1789 		"logical %llu is at position %llu within the extent (%llu EXTENT_ITEM %llu) flags %#llx size %u",
1790 		 logical, logical - found_key->objectid, found_key->objectid,
1791 		 found_key->offset, flags, item_size);
1792 
1793 	WARN_ON(!flags_ret);
1794 	if (flags_ret) {
1795 		if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK)
1796 			*flags_ret = BTRFS_EXTENT_FLAG_TREE_BLOCK;
1797 		else if (flags & BTRFS_EXTENT_FLAG_DATA)
1798 			*flags_ret = BTRFS_EXTENT_FLAG_DATA;
1799 		else
1800 			BUG();
1801 		return 0;
1802 	}
1803 
1804 	return -EIO;
1805 }
1806 
1807 /*
1808  * helper function to iterate extent inline refs. ptr must point to a 0 value
1809  * for the first call and may be modified. it is used to track state.
1810  * if more refs exist, 0 is returned and the next call to
1811  * get_extent_inline_ref must pass the modified ptr parameter to get the
1812  * next ref. after the last ref was processed, 1 is returned.
1813  * returns <0 on error
1814  */
1815 static int get_extent_inline_ref(unsigned long *ptr,
1816 				 const struct extent_buffer *eb,
1817 				 const struct btrfs_key *key,
1818 				 const struct btrfs_extent_item *ei,
1819 				 u32 item_size,
1820 				 struct btrfs_extent_inline_ref **out_eiref,
1821 				 int *out_type)
1822 {
1823 	unsigned long end;
1824 	u64 flags;
1825 	struct btrfs_tree_block_info *info;
1826 
1827 	if (!*ptr) {
1828 		/* first call */
1829 		flags = btrfs_extent_flags(eb, ei);
1830 		if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) {
1831 			if (key->type == BTRFS_METADATA_ITEM_KEY) {
1832 				/* a skinny metadata extent */
1833 				*out_eiref =
1834 				     (struct btrfs_extent_inline_ref *)(ei + 1);
1835 			} else {
1836 				WARN_ON(key->type != BTRFS_EXTENT_ITEM_KEY);
1837 				info = (struct btrfs_tree_block_info *)(ei + 1);
1838 				*out_eiref =
1839 				   (struct btrfs_extent_inline_ref *)(info + 1);
1840 			}
1841 		} else {
1842 			*out_eiref = (struct btrfs_extent_inline_ref *)(ei + 1);
1843 		}
1844 		*ptr = (unsigned long)*out_eiref;
1845 		if ((unsigned long)(*ptr) >= (unsigned long)ei + item_size)
1846 			return -ENOENT;
1847 	}
1848 
1849 	end = (unsigned long)ei + item_size;
1850 	*out_eiref = (struct btrfs_extent_inline_ref *)(*ptr);
1851 	*out_type = btrfs_get_extent_inline_ref_type(eb, *out_eiref,
1852 						     BTRFS_REF_TYPE_ANY);
1853 	if (*out_type == BTRFS_REF_TYPE_INVALID)
1854 		return -EUCLEAN;
1855 
1856 	*ptr += btrfs_extent_inline_ref_size(*out_type);
1857 	WARN_ON(*ptr > end);
1858 	if (*ptr == end)
1859 		return 1; /* last */
1860 
1861 	return 0;
1862 }
1863 
1864 /*
1865  * reads the tree block backref for an extent. tree level and root are returned
1866  * through out_level and out_root. ptr must point to a 0 value for the first
1867  * call and may be modified (see get_extent_inline_ref comment).
1868  * returns 0 if data was provided, 1 if there was no more data to provide or
1869  * <0 on error.
1870  */
1871 int tree_backref_for_extent(unsigned long *ptr, struct extent_buffer *eb,
1872 			    struct btrfs_key *key, struct btrfs_extent_item *ei,
1873 			    u32 item_size, u64 *out_root, u8 *out_level)
1874 {
1875 	int ret;
1876 	int type;
1877 	struct btrfs_extent_inline_ref *eiref;
1878 
1879 	if (*ptr == (unsigned long)-1)
1880 		return 1;
1881 
1882 	while (1) {
1883 		ret = get_extent_inline_ref(ptr, eb, key, ei, item_size,
1884 					      &eiref, &type);
1885 		if (ret < 0)
1886 			return ret;
1887 
1888 		if (type == BTRFS_TREE_BLOCK_REF_KEY ||
1889 		    type == BTRFS_SHARED_BLOCK_REF_KEY)
1890 			break;
1891 
1892 		if (ret == 1)
1893 			return 1;
1894 	}
1895 
1896 	/* we can treat both ref types equally here */
1897 	*out_root = btrfs_extent_inline_ref_offset(eb, eiref);
1898 
1899 	if (key->type == BTRFS_EXTENT_ITEM_KEY) {
1900 		struct btrfs_tree_block_info *info;
1901 
1902 		info = (struct btrfs_tree_block_info *)(ei + 1);
1903 		*out_level = btrfs_tree_block_level(eb, info);
1904 	} else {
1905 		ASSERT(key->type == BTRFS_METADATA_ITEM_KEY);
1906 		*out_level = (u8)key->offset;
1907 	}
1908 
1909 	if (ret == 1)
1910 		*ptr = (unsigned long)-1;
1911 
1912 	return 0;
1913 }
1914 
1915 static int iterate_leaf_refs(struct btrfs_fs_info *fs_info,
1916 			     struct extent_inode_elem *inode_list,
1917 			     u64 root, u64 extent_item_objectid,
1918 			     iterate_extent_inodes_t *iterate, void *ctx)
1919 {
1920 	struct extent_inode_elem *eie;
1921 	int ret = 0;
1922 
1923 	for (eie = inode_list; eie; eie = eie->next) {
1924 		btrfs_debug(fs_info,
1925 			    "ref for %llu resolved, key (%llu EXTEND_DATA %llu), root %llu",
1926 			    extent_item_objectid, eie->inum,
1927 			    eie->offset, root);
1928 		ret = iterate(eie->inum, eie->offset, root, ctx);
1929 		if (ret) {
1930 			btrfs_debug(fs_info,
1931 				    "stopping iteration for %llu due to ret=%d",
1932 				    extent_item_objectid, ret);
1933 			break;
1934 		}
1935 	}
1936 
1937 	return ret;
1938 }
1939 
1940 /*
1941  * calls iterate() for every inode that references the extent identified by
1942  * the given parameters.
1943  * when the iterator function returns a non-zero value, iteration stops.
1944  */
1945 int iterate_extent_inodes(struct btrfs_fs_info *fs_info,
1946 				u64 extent_item_objectid, u64 extent_item_pos,
1947 				int search_commit_root,
1948 				iterate_extent_inodes_t *iterate, void *ctx,
1949 				bool ignore_offset)
1950 {
1951 	int ret;
1952 	struct btrfs_trans_handle *trans = NULL;
1953 	struct ulist *refs = NULL;
1954 	struct ulist *roots = NULL;
1955 	struct ulist_node *ref_node = NULL;
1956 	struct ulist_node *root_node = NULL;
1957 	struct btrfs_seq_list seq_elem = BTRFS_SEQ_LIST_INIT(seq_elem);
1958 	struct ulist_iterator ref_uiter;
1959 	struct ulist_iterator root_uiter;
1960 
1961 	btrfs_debug(fs_info, "resolving all inodes for extent %llu",
1962 			extent_item_objectid);
1963 
1964 	if (!search_commit_root) {
1965 		trans = btrfs_attach_transaction(fs_info->extent_root);
1966 		if (IS_ERR(trans)) {
1967 			if (PTR_ERR(trans) != -ENOENT &&
1968 			    PTR_ERR(trans) != -EROFS)
1969 				return PTR_ERR(trans);
1970 			trans = NULL;
1971 		}
1972 	}
1973 
1974 	if (trans)
1975 		btrfs_get_tree_mod_seq(fs_info, &seq_elem);
1976 	else
1977 		down_read(&fs_info->commit_root_sem);
1978 
1979 	ret = btrfs_find_all_leafs(trans, fs_info, extent_item_objectid,
1980 				   seq_elem.seq, &refs,
1981 				   &extent_item_pos, ignore_offset);
1982 	if (ret)
1983 		goto out;
1984 
1985 	ULIST_ITER_INIT(&ref_uiter);
1986 	while (!ret && (ref_node = ulist_next(refs, &ref_uiter))) {
1987 		ret = btrfs_find_all_roots_safe(trans, fs_info, ref_node->val,
1988 						seq_elem.seq, &roots,
1989 						ignore_offset);
1990 		if (ret)
1991 			break;
1992 		ULIST_ITER_INIT(&root_uiter);
1993 		while (!ret && (root_node = ulist_next(roots, &root_uiter))) {
1994 			btrfs_debug(fs_info,
1995 				    "root %llu references leaf %llu, data list %#llx",
1996 				    root_node->val, ref_node->val,
1997 				    ref_node->aux);
1998 			ret = iterate_leaf_refs(fs_info,
1999 						(struct extent_inode_elem *)
2000 						(uintptr_t)ref_node->aux,
2001 						root_node->val,
2002 						extent_item_objectid,
2003 						iterate, ctx);
2004 		}
2005 		ulist_free(roots);
2006 	}
2007 
2008 	free_leaf_list(refs);
2009 out:
2010 	if (trans) {
2011 		btrfs_put_tree_mod_seq(fs_info, &seq_elem);
2012 		btrfs_end_transaction(trans);
2013 	} else {
2014 		up_read(&fs_info->commit_root_sem);
2015 	}
2016 
2017 	return ret;
2018 }
2019 
2020 int iterate_inodes_from_logical(u64 logical, struct btrfs_fs_info *fs_info,
2021 				struct btrfs_path *path,
2022 				iterate_extent_inodes_t *iterate, void *ctx,
2023 				bool ignore_offset)
2024 {
2025 	int ret;
2026 	u64 extent_item_pos;
2027 	u64 flags = 0;
2028 	struct btrfs_key found_key;
2029 	int search_commit_root = path->search_commit_root;
2030 
2031 	ret = extent_from_logical(fs_info, logical, path, &found_key, &flags);
2032 	btrfs_release_path(path);
2033 	if (ret < 0)
2034 		return ret;
2035 	if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK)
2036 		return -EINVAL;
2037 
2038 	extent_item_pos = logical - found_key.objectid;
2039 	ret = iterate_extent_inodes(fs_info, found_key.objectid,
2040 					extent_item_pos, search_commit_root,
2041 					iterate, ctx, ignore_offset);
2042 
2043 	return ret;
2044 }
2045 
2046 typedef int (iterate_irefs_t)(u64 parent, u32 name_len, unsigned long name_off,
2047 			      struct extent_buffer *eb, void *ctx);
2048 
2049 static int iterate_inode_refs(u64 inum, struct btrfs_root *fs_root,
2050 			      struct btrfs_path *path,
2051 			      iterate_irefs_t *iterate, void *ctx)
2052 {
2053 	int ret = 0;
2054 	int slot;
2055 	u32 cur;
2056 	u32 len;
2057 	u32 name_len;
2058 	u64 parent = 0;
2059 	int found = 0;
2060 	struct extent_buffer *eb;
2061 	struct btrfs_item *item;
2062 	struct btrfs_inode_ref *iref;
2063 	struct btrfs_key found_key;
2064 
2065 	while (!ret) {
2066 		ret = btrfs_find_item(fs_root, path, inum,
2067 				parent ? parent + 1 : 0, BTRFS_INODE_REF_KEY,
2068 				&found_key);
2069 
2070 		if (ret < 0)
2071 			break;
2072 		if (ret) {
2073 			ret = found ? 0 : -ENOENT;
2074 			break;
2075 		}
2076 		++found;
2077 
2078 		parent = found_key.offset;
2079 		slot = path->slots[0];
2080 		eb = btrfs_clone_extent_buffer(path->nodes[0]);
2081 		if (!eb) {
2082 			ret = -ENOMEM;
2083 			break;
2084 		}
2085 		btrfs_release_path(path);
2086 
2087 		item = btrfs_item_nr(slot);
2088 		iref = btrfs_item_ptr(eb, slot, struct btrfs_inode_ref);
2089 
2090 		for (cur = 0; cur < btrfs_item_size(eb, item); cur += len) {
2091 			name_len = btrfs_inode_ref_name_len(eb, iref);
2092 			/* path must be released before calling iterate()! */
2093 			btrfs_debug(fs_root->fs_info,
2094 				"following ref at offset %u for inode %llu in tree %llu",
2095 				cur, found_key.objectid,
2096 				fs_root->root_key.objectid);
2097 			ret = iterate(parent, name_len,
2098 				      (unsigned long)(iref + 1), eb, ctx);
2099 			if (ret)
2100 				break;
2101 			len = sizeof(*iref) + name_len;
2102 			iref = (struct btrfs_inode_ref *)((char *)iref + len);
2103 		}
2104 		free_extent_buffer(eb);
2105 	}
2106 
2107 	btrfs_release_path(path);
2108 
2109 	return ret;
2110 }
2111 
2112 static int iterate_inode_extrefs(u64 inum, struct btrfs_root *fs_root,
2113 				 struct btrfs_path *path,
2114 				 iterate_irefs_t *iterate, void *ctx)
2115 {
2116 	int ret;
2117 	int slot;
2118 	u64 offset = 0;
2119 	u64 parent;
2120 	int found = 0;
2121 	struct extent_buffer *eb;
2122 	struct btrfs_inode_extref *extref;
2123 	u32 item_size;
2124 	u32 cur_offset;
2125 	unsigned long ptr;
2126 
2127 	while (1) {
2128 		ret = btrfs_find_one_extref(fs_root, inum, offset, path, &extref,
2129 					    &offset);
2130 		if (ret < 0)
2131 			break;
2132 		if (ret) {
2133 			ret = found ? 0 : -ENOENT;
2134 			break;
2135 		}
2136 		++found;
2137 
2138 		slot = path->slots[0];
2139 		eb = btrfs_clone_extent_buffer(path->nodes[0]);
2140 		if (!eb) {
2141 			ret = -ENOMEM;
2142 			break;
2143 		}
2144 		btrfs_release_path(path);
2145 
2146 		item_size = btrfs_item_size_nr(eb, slot);
2147 		ptr = btrfs_item_ptr_offset(eb, slot);
2148 		cur_offset = 0;
2149 
2150 		while (cur_offset < item_size) {
2151 			u32 name_len;
2152 
2153 			extref = (struct btrfs_inode_extref *)(ptr + cur_offset);
2154 			parent = btrfs_inode_extref_parent(eb, extref);
2155 			name_len = btrfs_inode_extref_name_len(eb, extref);
2156 			ret = iterate(parent, name_len,
2157 				      (unsigned long)&extref->name, eb, ctx);
2158 			if (ret)
2159 				break;
2160 
2161 			cur_offset += btrfs_inode_extref_name_len(eb, extref);
2162 			cur_offset += sizeof(*extref);
2163 		}
2164 		free_extent_buffer(eb);
2165 
2166 		offset++;
2167 	}
2168 
2169 	btrfs_release_path(path);
2170 
2171 	return ret;
2172 }
2173 
2174 static int iterate_irefs(u64 inum, struct btrfs_root *fs_root,
2175 			 struct btrfs_path *path, iterate_irefs_t *iterate,
2176 			 void *ctx)
2177 {
2178 	int ret;
2179 	int found_refs = 0;
2180 
2181 	ret = iterate_inode_refs(inum, fs_root, path, iterate, ctx);
2182 	if (!ret)
2183 		++found_refs;
2184 	else if (ret != -ENOENT)
2185 		return ret;
2186 
2187 	ret = iterate_inode_extrefs(inum, fs_root, path, iterate, ctx);
2188 	if (ret == -ENOENT && found_refs)
2189 		return 0;
2190 
2191 	return ret;
2192 }
2193 
2194 /*
2195  * returns 0 if the path could be dumped (probably truncated)
2196  * returns <0 in case of an error
2197  */
2198 static int inode_to_path(u64 inum, u32 name_len, unsigned long name_off,
2199 			 struct extent_buffer *eb, void *ctx)
2200 {
2201 	struct inode_fs_paths *ipath = ctx;
2202 	char *fspath;
2203 	char *fspath_min;
2204 	int i = ipath->fspath->elem_cnt;
2205 	const int s_ptr = sizeof(char *);
2206 	u32 bytes_left;
2207 
2208 	bytes_left = ipath->fspath->bytes_left > s_ptr ?
2209 					ipath->fspath->bytes_left - s_ptr : 0;
2210 
2211 	fspath_min = (char *)ipath->fspath->val + (i + 1) * s_ptr;
2212 	fspath = btrfs_ref_to_path(ipath->fs_root, ipath->btrfs_path, name_len,
2213 				   name_off, eb, inum, fspath_min, bytes_left);
2214 	if (IS_ERR(fspath))
2215 		return PTR_ERR(fspath);
2216 
2217 	if (fspath > fspath_min) {
2218 		ipath->fspath->val[i] = (u64)(unsigned long)fspath;
2219 		++ipath->fspath->elem_cnt;
2220 		ipath->fspath->bytes_left = fspath - fspath_min;
2221 	} else {
2222 		++ipath->fspath->elem_missed;
2223 		ipath->fspath->bytes_missing += fspath_min - fspath;
2224 		ipath->fspath->bytes_left = 0;
2225 	}
2226 
2227 	return 0;
2228 }
2229 
2230 /*
2231  * this dumps all file system paths to the inode into the ipath struct, provided
2232  * is has been created large enough. each path is zero-terminated and accessed
2233  * from ipath->fspath->val[i].
2234  * when it returns, there are ipath->fspath->elem_cnt number of paths available
2235  * in ipath->fspath->val[]. when the allocated space wasn't sufficient, the
2236  * number of missed paths is recorded in ipath->fspath->elem_missed, otherwise,
2237  * it's zero. ipath->fspath->bytes_missing holds the number of bytes that would
2238  * have been needed to return all paths.
2239  */
2240 int paths_from_inode(u64 inum, struct inode_fs_paths *ipath)
2241 {
2242 	return iterate_irefs(inum, ipath->fs_root, ipath->btrfs_path,
2243 			     inode_to_path, ipath);
2244 }
2245 
2246 struct btrfs_data_container *init_data_container(u32 total_bytes)
2247 {
2248 	struct btrfs_data_container *data;
2249 	size_t alloc_bytes;
2250 
2251 	alloc_bytes = max_t(size_t, total_bytes, sizeof(*data));
2252 	data = kvmalloc(alloc_bytes, GFP_KERNEL);
2253 	if (!data)
2254 		return ERR_PTR(-ENOMEM);
2255 
2256 	if (total_bytes >= sizeof(*data)) {
2257 		data->bytes_left = total_bytes - sizeof(*data);
2258 		data->bytes_missing = 0;
2259 	} else {
2260 		data->bytes_missing = sizeof(*data) - total_bytes;
2261 		data->bytes_left = 0;
2262 	}
2263 
2264 	data->elem_cnt = 0;
2265 	data->elem_missed = 0;
2266 
2267 	return data;
2268 }
2269 
2270 /*
2271  * allocates space to return multiple file system paths for an inode.
2272  * total_bytes to allocate are passed, note that space usable for actual path
2273  * information will be total_bytes - sizeof(struct inode_fs_paths).
2274  * the returned pointer must be freed with free_ipath() in the end.
2275  */
2276 struct inode_fs_paths *init_ipath(s32 total_bytes, struct btrfs_root *fs_root,
2277 					struct btrfs_path *path)
2278 {
2279 	struct inode_fs_paths *ifp;
2280 	struct btrfs_data_container *fspath;
2281 
2282 	fspath = init_data_container(total_bytes);
2283 	if (IS_ERR(fspath))
2284 		return ERR_CAST(fspath);
2285 
2286 	ifp = kmalloc(sizeof(*ifp), GFP_KERNEL);
2287 	if (!ifp) {
2288 		kvfree(fspath);
2289 		return ERR_PTR(-ENOMEM);
2290 	}
2291 
2292 	ifp->btrfs_path = path;
2293 	ifp->fspath = fspath;
2294 	ifp->fs_root = fs_root;
2295 
2296 	return ifp;
2297 }
2298 
2299 void free_ipath(struct inode_fs_paths *ipath)
2300 {
2301 	if (!ipath)
2302 		return;
2303 	kvfree(ipath->fspath);
2304 	kfree(ipath);
2305 }
2306 
2307 struct btrfs_backref_iter *btrfs_backref_iter_alloc(
2308 		struct btrfs_fs_info *fs_info, gfp_t gfp_flag)
2309 {
2310 	struct btrfs_backref_iter *ret;
2311 
2312 	ret = kzalloc(sizeof(*ret), gfp_flag);
2313 	if (!ret)
2314 		return NULL;
2315 
2316 	ret->path = btrfs_alloc_path();
2317 	if (!ret->path) {
2318 		kfree(ret);
2319 		return NULL;
2320 	}
2321 
2322 	/* Current backref iterator only supports iteration in commit root */
2323 	ret->path->search_commit_root = 1;
2324 	ret->path->skip_locking = 1;
2325 	ret->fs_info = fs_info;
2326 
2327 	return ret;
2328 }
2329 
2330 int btrfs_backref_iter_start(struct btrfs_backref_iter *iter, u64 bytenr)
2331 {
2332 	struct btrfs_fs_info *fs_info = iter->fs_info;
2333 	struct btrfs_path *path = iter->path;
2334 	struct btrfs_extent_item *ei;
2335 	struct btrfs_key key;
2336 	int ret;
2337 
2338 	key.objectid = bytenr;
2339 	key.type = BTRFS_METADATA_ITEM_KEY;
2340 	key.offset = (u64)-1;
2341 	iter->bytenr = bytenr;
2342 
2343 	ret = btrfs_search_slot(NULL, fs_info->extent_root, &key, path, 0, 0);
2344 	if (ret < 0)
2345 		return ret;
2346 	if (ret == 0) {
2347 		ret = -EUCLEAN;
2348 		goto release;
2349 	}
2350 	if (path->slots[0] == 0) {
2351 		WARN_ON(IS_ENABLED(CONFIG_BTRFS_DEBUG));
2352 		ret = -EUCLEAN;
2353 		goto release;
2354 	}
2355 	path->slots[0]--;
2356 
2357 	btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
2358 	if ((key.type != BTRFS_EXTENT_ITEM_KEY &&
2359 	     key.type != BTRFS_METADATA_ITEM_KEY) || key.objectid != bytenr) {
2360 		ret = -ENOENT;
2361 		goto release;
2362 	}
2363 	memcpy(&iter->cur_key, &key, sizeof(key));
2364 	iter->item_ptr = (u32)btrfs_item_ptr_offset(path->nodes[0],
2365 						    path->slots[0]);
2366 	iter->end_ptr = (u32)(iter->item_ptr +
2367 			btrfs_item_size_nr(path->nodes[0], path->slots[0]));
2368 	ei = btrfs_item_ptr(path->nodes[0], path->slots[0],
2369 			    struct btrfs_extent_item);
2370 
2371 	/*
2372 	 * Only support iteration on tree backref yet.
2373 	 *
2374 	 * This is an extra precaution for non skinny-metadata, where
2375 	 * EXTENT_ITEM is also used for tree blocks, that we can only use
2376 	 * extent flags to determine if it's a tree block.
2377 	 */
2378 	if (btrfs_extent_flags(path->nodes[0], ei) & BTRFS_EXTENT_FLAG_DATA) {
2379 		ret = -ENOTSUPP;
2380 		goto release;
2381 	}
2382 	iter->cur_ptr = (u32)(iter->item_ptr + sizeof(*ei));
2383 
2384 	/* If there is no inline backref, go search for keyed backref */
2385 	if (iter->cur_ptr >= iter->end_ptr) {
2386 		ret = btrfs_next_item(fs_info->extent_root, path);
2387 
2388 		/* No inline nor keyed ref */
2389 		if (ret > 0) {
2390 			ret = -ENOENT;
2391 			goto release;
2392 		}
2393 		if (ret < 0)
2394 			goto release;
2395 
2396 		btrfs_item_key_to_cpu(path->nodes[0], &iter->cur_key,
2397 				path->slots[0]);
2398 		if (iter->cur_key.objectid != bytenr ||
2399 		    (iter->cur_key.type != BTRFS_SHARED_BLOCK_REF_KEY &&
2400 		     iter->cur_key.type != BTRFS_TREE_BLOCK_REF_KEY)) {
2401 			ret = -ENOENT;
2402 			goto release;
2403 		}
2404 		iter->cur_ptr = (u32)btrfs_item_ptr_offset(path->nodes[0],
2405 							   path->slots[0]);
2406 		iter->item_ptr = iter->cur_ptr;
2407 		iter->end_ptr = (u32)(iter->item_ptr + btrfs_item_size_nr(
2408 				      path->nodes[0], path->slots[0]));
2409 	}
2410 
2411 	return 0;
2412 release:
2413 	btrfs_backref_iter_release(iter);
2414 	return ret;
2415 }
2416 
2417 /*
2418  * Go to the next backref item of current bytenr, can be either inlined or
2419  * keyed.
2420  *
2421  * Caller needs to check whether it's inline ref or not by iter->cur_key.
2422  *
2423  * Return 0 if we get next backref without problem.
2424  * Return >0 if there is no extra backref for this bytenr.
2425  * Return <0 if there is something wrong happened.
2426  */
2427 int btrfs_backref_iter_next(struct btrfs_backref_iter *iter)
2428 {
2429 	struct extent_buffer *eb = btrfs_backref_get_eb(iter);
2430 	struct btrfs_path *path = iter->path;
2431 	struct btrfs_extent_inline_ref *iref;
2432 	int ret;
2433 	u32 size;
2434 
2435 	if (btrfs_backref_iter_is_inline_ref(iter)) {
2436 		/* We're still inside the inline refs */
2437 		ASSERT(iter->cur_ptr < iter->end_ptr);
2438 
2439 		if (btrfs_backref_has_tree_block_info(iter)) {
2440 			/* First tree block info */
2441 			size = sizeof(struct btrfs_tree_block_info);
2442 		} else {
2443 			/* Use inline ref type to determine the size */
2444 			int type;
2445 
2446 			iref = (struct btrfs_extent_inline_ref *)
2447 				((unsigned long)iter->cur_ptr);
2448 			type = btrfs_extent_inline_ref_type(eb, iref);
2449 
2450 			size = btrfs_extent_inline_ref_size(type);
2451 		}
2452 		iter->cur_ptr += size;
2453 		if (iter->cur_ptr < iter->end_ptr)
2454 			return 0;
2455 
2456 		/* All inline items iterated, fall through */
2457 	}
2458 
2459 	/* We're at keyed items, there is no inline item, go to the next one */
2460 	ret = btrfs_next_item(iter->fs_info->extent_root, iter->path);
2461 	if (ret)
2462 		return ret;
2463 
2464 	btrfs_item_key_to_cpu(path->nodes[0], &iter->cur_key, path->slots[0]);
2465 	if (iter->cur_key.objectid != iter->bytenr ||
2466 	    (iter->cur_key.type != BTRFS_TREE_BLOCK_REF_KEY &&
2467 	     iter->cur_key.type != BTRFS_SHARED_BLOCK_REF_KEY))
2468 		return 1;
2469 	iter->item_ptr = (u32)btrfs_item_ptr_offset(path->nodes[0],
2470 					path->slots[0]);
2471 	iter->cur_ptr = iter->item_ptr;
2472 	iter->end_ptr = iter->item_ptr + (u32)btrfs_item_size_nr(path->nodes[0],
2473 						path->slots[0]);
2474 	return 0;
2475 }
2476 
2477 void btrfs_backref_init_cache(struct btrfs_fs_info *fs_info,
2478 			      struct btrfs_backref_cache *cache, int is_reloc)
2479 {
2480 	int i;
2481 
2482 	cache->rb_root = RB_ROOT;
2483 	for (i = 0; i < BTRFS_MAX_LEVEL; i++)
2484 		INIT_LIST_HEAD(&cache->pending[i]);
2485 	INIT_LIST_HEAD(&cache->changed);
2486 	INIT_LIST_HEAD(&cache->detached);
2487 	INIT_LIST_HEAD(&cache->leaves);
2488 	INIT_LIST_HEAD(&cache->pending_edge);
2489 	INIT_LIST_HEAD(&cache->useless_node);
2490 	cache->fs_info = fs_info;
2491 	cache->is_reloc = is_reloc;
2492 }
2493 
2494 struct btrfs_backref_node *btrfs_backref_alloc_node(
2495 		struct btrfs_backref_cache *cache, u64 bytenr, int level)
2496 {
2497 	struct btrfs_backref_node *node;
2498 
2499 	ASSERT(level >= 0 && level < BTRFS_MAX_LEVEL);
2500 	node = kzalloc(sizeof(*node), GFP_NOFS);
2501 	if (!node)
2502 		return node;
2503 
2504 	INIT_LIST_HEAD(&node->list);
2505 	INIT_LIST_HEAD(&node->upper);
2506 	INIT_LIST_HEAD(&node->lower);
2507 	RB_CLEAR_NODE(&node->rb_node);
2508 	cache->nr_nodes++;
2509 	node->level = level;
2510 	node->bytenr = bytenr;
2511 
2512 	return node;
2513 }
2514 
2515 struct btrfs_backref_edge *btrfs_backref_alloc_edge(
2516 		struct btrfs_backref_cache *cache)
2517 {
2518 	struct btrfs_backref_edge *edge;
2519 
2520 	edge = kzalloc(sizeof(*edge), GFP_NOFS);
2521 	if (edge)
2522 		cache->nr_edges++;
2523 	return edge;
2524 }
2525 
2526 /*
2527  * Drop the backref node from cache, also cleaning up all its
2528  * upper edges and any uncached nodes in the path.
2529  *
2530  * This cleanup happens bottom up, thus the node should either
2531  * be the lowest node in the cache or a detached node.
2532  */
2533 void btrfs_backref_cleanup_node(struct btrfs_backref_cache *cache,
2534 				struct btrfs_backref_node *node)
2535 {
2536 	struct btrfs_backref_node *upper;
2537 	struct btrfs_backref_edge *edge;
2538 
2539 	if (!node)
2540 		return;
2541 
2542 	BUG_ON(!node->lowest && !node->detached);
2543 	while (!list_empty(&node->upper)) {
2544 		edge = list_entry(node->upper.next, struct btrfs_backref_edge,
2545 				  list[LOWER]);
2546 		upper = edge->node[UPPER];
2547 		list_del(&edge->list[LOWER]);
2548 		list_del(&edge->list[UPPER]);
2549 		btrfs_backref_free_edge(cache, edge);
2550 
2551 		/*
2552 		 * Add the node to leaf node list if no other child block
2553 		 * cached.
2554 		 */
2555 		if (list_empty(&upper->lower)) {
2556 			list_add_tail(&upper->lower, &cache->leaves);
2557 			upper->lowest = 1;
2558 		}
2559 	}
2560 
2561 	btrfs_backref_drop_node(cache, node);
2562 }
2563 
2564 /*
2565  * Release all nodes/edges from current cache
2566  */
2567 void btrfs_backref_release_cache(struct btrfs_backref_cache *cache)
2568 {
2569 	struct btrfs_backref_node *node;
2570 	int i;
2571 
2572 	while (!list_empty(&cache->detached)) {
2573 		node = list_entry(cache->detached.next,
2574 				  struct btrfs_backref_node, list);
2575 		btrfs_backref_cleanup_node(cache, node);
2576 	}
2577 
2578 	while (!list_empty(&cache->leaves)) {
2579 		node = list_entry(cache->leaves.next,
2580 				  struct btrfs_backref_node, lower);
2581 		btrfs_backref_cleanup_node(cache, node);
2582 	}
2583 
2584 	cache->last_trans = 0;
2585 
2586 	for (i = 0; i < BTRFS_MAX_LEVEL; i++)
2587 		ASSERT(list_empty(&cache->pending[i]));
2588 	ASSERT(list_empty(&cache->pending_edge));
2589 	ASSERT(list_empty(&cache->useless_node));
2590 	ASSERT(list_empty(&cache->changed));
2591 	ASSERT(list_empty(&cache->detached));
2592 	ASSERT(RB_EMPTY_ROOT(&cache->rb_root));
2593 	ASSERT(!cache->nr_nodes);
2594 	ASSERT(!cache->nr_edges);
2595 }
2596 
2597 /*
2598  * Handle direct tree backref
2599  *
2600  * Direct tree backref means, the backref item shows its parent bytenr
2601  * directly. This is for SHARED_BLOCK_REF backref (keyed or inlined).
2602  *
2603  * @ref_key:	The converted backref key.
2604  *		For keyed backref, it's the item key.
2605  *		For inlined backref, objectid is the bytenr,
2606  *		type is btrfs_inline_ref_type, offset is
2607  *		btrfs_inline_ref_offset.
2608  */
2609 static int handle_direct_tree_backref(struct btrfs_backref_cache *cache,
2610 				      struct btrfs_key *ref_key,
2611 				      struct btrfs_backref_node *cur)
2612 {
2613 	struct btrfs_backref_edge *edge;
2614 	struct btrfs_backref_node *upper;
2615 	struct rb_node *rb_node;
2616 
2617 	ASSERT(ref_key->type == BTRFS_SHARED_BLOCK_REF_KEY);
2618 
2619 	/* Only reloc root uses backref pointing to itself */
2620 	if (ref_key->objectid == ref_key->offset) {
2621 		struct btrfs_root *root;
2622 
2623 		cur->is_reloc_root = 1;
2624 		/* Only reloc backref cache cares about a specific root */
2625 		if (cache->is_reloc) {
2626 			root = find_reloc_root(cache->fs_info, cur->bytenr);
2627 			if (!root)
2628 				return -ENOENT;
2629 			cur->root = root;
2630 		} else {
2631 			/*
2632 			 * For generic purpose backref cache, reloc root node
2633 			 * is useless.
2634 			 */
2635 			list_add(&cur->list, &cache->useless_node);
2636 		}
2637 		return 0;
2638 	}
2639 
2640 	edge = btrfs_backref_alloc_edge(cache);
2641 	if (!edge)
2642 		return -ENOMEM;
2643 
2644 	rb_node = rb_simple_search(&cache->rb_root, ref_key->offset);
2645 	if (!rb_node) {
2646 		/* Parent node not yet cached */
2647 		upper = btrfs_backref_alloc_node(cache, ref_key->offset,
2648 					   cur->level + 1);
2649 		if (!upper) {
2650 			btrfs_backref_free_edge(cache, edge);
2651 			return -ENOMEM;
2652 		}
2653 
2654 		/*
2655 		 *  Backrefs for the upper level block isn't cached, add the
2656 		 *  block to pending list
2657 		 */
2658 		list_add_tail(&edge->list[UPPER], &cache->pending_edge);
2659 	} else {
2660 		/* Parent node already cached */
2661 		upper = rb_entry(rb_node, struct btrfs_backref_node, rb_node);
2662 		ASSERT(upper->checked);
2663 		INIT_LIST_HEAD(&edge->list[UPPER]);
2664 	}
2665 	btrfs_backref_link_edge(edge, cur, upper, LINK_LOWER);
2666 	return 0;
2667 }
2668 
2669 /*
2670  * Handle indirect tree backref
2671  *
2672  * Indirect tree backref means, we only know which tree the node belongs to.
2673  * We still need to do a tree search to find out the parents. This is for
2674  * TREE_BLOCK_REF backref (keyed or inlined).
2675  *
2676  * @ref_key:	The same as @ref_key in  handle_direct_tree_backref()
2677  * @tree_key:	The first key of this tree block.
2678  * @path:	A clean (released) path, to avoid allocating path everytime
2679  *		the function get called.
2680  */
2681 static int handle_indirect_tree_backref(struct btrfs_backref_cache *cache,
2682 					struct btrfs_path *path,
2683 					struct btrfs_key *ref_key,
2684 					struct btrfs_key *tree_key,
2685 					struct btrfs_backref_node *cur)
2686 {
2687 	struct btrfs_fs_info *fs_info = cache->fs_info;
2688 	struct btrfs_backref_node *upper;
2689 	struct btrfs_backref_node *lower;
2690 	struct btrfs_backref_edge *edge;
2691 	struct extent_buffer *eb;
2692 	struct btrfs_root *root;
2693 	struct rb_node *rb_node;
2694 	int level;
2695 	bool need_check = true;
2696 	int ret;
2697 
2698 	root = btrfs_get_fs_root(fs_info, ref_key->offset, false);
2699 	if (IS_ERR(root))
2700 		return PTR_ERR(root);
2701 	if (!test_bit(BTRFS_ROOT_SHAREABLE, &root->state))
2702 		cur->cowonly = 1;
2703 
2704 	if (btrfs_root_level(&root->root_item) == cur->level) {
2705 		/* Tree root */
2706 		ASSERT(btrfs_root_bytenr(&root->root_item) == cur->bytenr);
2707 		/*
2708 		 * For reloc backref cache, we may ignore reloc root.  But for
2709 		 * general purpose backref cache, we can't rely on
2710 		 * btrfs_should_ignore_reloc_root() as it may conflict with
2711 		 * current running relocation and lead to missing root.
2712 		 *
2713 		 * For general purpose backref cache, reloc root detection is
2714 		 * completely relying on direct backref (key->offset is parent
2715 		 * bytenr), thus only do such check for reloc cache.
2716 		 */
2717 		if (btrfs_should_ignore_reloc_root(root) && cache->is_reloc) {
2718 			btrfs_put_root(root);
2719 			list_add(&cur->list, &cache->useless_node);
2720 		} else {
2721 			cur->root = root;
2722 		}
2723 		return 0;
2724 	}
2725 
2726 	level = cur->level + 1;
2727 
2728 	/* Search the tree to find parent blocks referring to the block */
2729 	path->search_commit_root = 1;
2730 	path->skip_locking = 1;
2731 	path->lowest_level = level;
2732 	ret = btrfs_search_slot(NULL, root, tree_key, path, 0, 0);
2733 	path->lowest_level = 0;
2734 	if (ret < 0) {
2735 		btrfs_put_root(root);
2736 		return ret;
2737 	}
2738 	if (ret > 0 && path->slots[level] > 0)
2739 		path->slots[level]--;
2740 
2741 	eb = path->nodes[level];
2742 	if (btrfs_node_blockptr(eb, path->slots[level]) != cur->bytenr) {
2743 		btrfs_err(fs_info,
2744 "couldn't find block (%llu) (level %d) in tree (%llu) with key (%llu %u %llu)",
2745 			  cur->bytenr, level - 1, root->root_key.objectid,
2746 			  tree_key->objectid, tree_key->type, tree_key->offset);
2747 		btrfs_put_root(root);
2748 		ret = -ENOENT;
2749 		goto out;
2750 	}
2751 	lower = cur;
2752 
2753 	/* Add all nodes and edges in the path */
2754 	for (; level < BTRFS_MAX_LEVEL; level++) {
2755 		if (!path->nodes[level]) {
2756 			ASSERT(btrfs_root_bytenr(&root->root_item) ==
2757 			       lower->bytenr);
2758 			/* Same as previous should_ignore_reloc_root() call */
2759 			if (btrfs_should_ignore_reloc_root(root) &&
2760 			    cache->is_reloc) {
2761 				btrfs_put_root(root);
2762 				list_add(&lower->list, &cache->useless_node);
2763 			} else {
2764 				lower->root = root;
2765 			}
2766 			break;
2767 		}
2768 
2769 		edge = btrfs_backref_alloc_edge(cache);
2770 		if (!edge) {
2771 			btrfs_put_root(root);
2772 			ret = -ENOMEM;
2773 			goto out;
2774 		}
2775 
2776 		eb = path->nodes[level];
2777 		rb_node = rb_simple_search(&cache->rb_root, eb->start);
2778 		if (!rb_node) {
2779 			upper = btrfs_backref_alloc_node(cache, eb->start,
2780 							 lower->level + 1);
2781 			if (!upper) {
2782 				btrfs_put_root(root);
2783 				btrfs_backref_free_edge(cache, edge);
2784 				ret = -ENOMEM;
2785 				goto out;
2786 			}
2787 			upper->owner = btrfs_header_owner(eb);
2788 			if (!test_bit(BTRFS_ROOT_SHAREABLE, &root->state))
2789 				upper->cowonly = 1;
2790 
2791 			/*
2792 			 * If we know the block isn't shared we can avoid
2793 			 * checking its backrefs.
2794 			 */
2795 			if (btrfs_block_can_be_shared(root, eb))
2796 				upper->checked = 0;
2797 			else
2798 				upper->checked = 1;
2799 
2800 			/*
2801 			 * Add the block to pending list if we need to check its
2802 			 * backrefs, we only do this once while walking up a
2803 			 * tree as we will catch anything else later on.
2804 			 */
2805 			if (!upper->checked && need_check) {
2806 				need_check = false;
2807 				list_add_tail(&edge->list[UPPER],
2808 					      &cache->pending_edge);
2809 			} else {
2810 				if (upper->checked)
2811 					need_check = true;
2812 				INIT_LIST_HEAD(&edge->list[UPPER]);
2813 			}
2814 		} else {
2815 			upper = rb_entry(rb_node, struct btrfs_backref_node,
2816 					 rb_node);
2817 			ASSERT(upper->checked);
2818 			INIT_LIST_HEAD(&edge->list[UPPER]);
2819 			if (!upper->owner)
2820 				upper->owner = btrfs_header_owner(eb);
2821 		}
2822 		btrfs_backref_link_edge(edge, lower, upper, LINK_LOWER);
2823 
2824 		if (rb_node) {
2825 			btrfs_put_root(root);
2826 			break;
2827 		}
2828 		lower = upper;
2829 		upper = NULL;
2830 	}
2831 out:
2832 	btrfs_release_path(path);
2833 	return ret;
2834 }
2835 
2836 /*
2837  * Add backref node @cur into @cache.
2838  *
2839  * NOTE: Even if the function returned 0, @cur is not yet cached as its upper
2840  *	 links aren't yet bi-directional. Needs to finish such links.
2841  *	 Use btrfs_backref_finish_upper_links() to finish such linkage.
2842  *
2843  * @path:	Released path for indirect tree backref lookup
2844  * @iter:	Released backref iter for extent tree search
2845  * @node_key:	The first key of the tree block
2846  */
2847 int btrfs_backref_add_tree_node(struct btrfs_backref_cache *cache,
2848 				struct btrfs_path *path,
2849 				struct btrfs_backref_iter *iter,
2850 				struct btrfs_key *node_key,
2851 				struct btrfs_backref_node *cur)
2852 {
2853 	struct btrfs_fs_info *fs_info = cache->fs_info;
2854 	struct btrfs_backref_edge *edge;
2855 	struct btrfs_backref_node *exist;
2856 	int ret;
2857 
2858 	ret = btrfs_backref_iter_start(iter, cur->bytenr);
2859 	if (ret < 0)
2860 		return ret;
2861 	/*
2862 	 * We skip the first btrfs_tree_block_info, as we don't use the key
2863 	 * stored in it, but fetch it from the tree block
2864 	 */
2865 	if (btrfs_backref_has_tree_block_info(iter)) {
2866 		ret = btrfs_backref_iter_next(iter);
2867 		if (ret < 0)
2868 			goto out;
2869 		/* No extra backref? This means the tree block is corrupted */
2870 		if (ret > 0) {
2871 			ret = -EUCLEAN;
2872 			goto out;
2873 		}
2874 	}
2875 	WARN_ON(cur->checked);
2876 	if (!list_empty(&cur->upper)) {
2877 		/*
2878 		 * The backref was added previously when processing backref of
2879 		 * type BTRFS_TREE_BLOCK_REF_KEY
2880 		 */
2881 		ASSERT(list_is_singular(&cur->upper));
2882 		edge = list_entry(cur->upper.next, struct btrfs_backref_edge,
2883 				  list[LOWER]);
2884 		ASSERT(list_empty(&edge->list[UPPER]));
2885 		exist = edge->node[UPPER];
2886 		/*
2887 		 * Add the upper level block to pending list if we need check
2888 		 * its backrefs
2889 		 */
2890 		if (!exist->checked)
2891 			list_add_tail(&edge->list[UPPER], &cache->pending_edge);
2892 	} else {
2893 		exist = NULL;
2894 	}
2895 
2896 	for (; ret == 0; ret = btrfs_backref_iter_next(iter)) {
2897 		struct extent_buffer *eb;
2898 		struct btrfs_key key;
2899 		int type;
2900 
2901 		cond_resched();
2902 		eb = btrfs_backref_get_eb(iter);
2903 
2904 		key.objectid = iter->bytenr;
2905 		if (btrfs_backref_iter_is_inline_ref(iter)) {
2906 			struct btrfs_extent_inline_ref *iref;
2907 
2908 			/* Update key for inline backref */
2909 			iref = (struct btrfs_extent_inline_ref *)
2910 				((unsigned long)iter->cur_ptr);
2911 			type = btrfs_get_extent_inline_ref_type(eb, iref,
2912 							BTRFS_REF_TYPE_BLOCK);
2913 			if (type == BTRFS_REF_TYPE_INVALID) {
2914 				ret = -EUCLEAN;
2915 				goto out;
2916 			}
2917 			key.type = type;
2918 			key.offset = btrfs_extent_inline_ref_offset(eb, iref);
2919 		} else {
2920 			key.type = iter->cur_key.type;
2921 			key.offset = iter->cur_key.offset;
2922 		}
2923 
2924 		/*
2925 		 * Parent node found and matches current inline ref, no need to
2926 		 * rebuild this node for this inline ref
2927 		 */
2928 		if (exist &&
2929 		    ((key.type == BTRFS_TREE_BLOCK_REF_KEY &&
2930 		      exist->owner == key.offset) ||
2931 		     (key.type == BTRFS_SHARED_BLOCK_REF_KEY &&
2932 		      exist->bytenr == key.offset))) {
2933 			exist = NULL;
2934 			continue;
2935 		}
2936 
2937 		/* SHARED_BLOCK_REF means key.offset is the parent bytenr */
2938 		if (key.type == BTRFS_SHARED_BLOCK_REF_KEY) {
2939 			ret = handle_direct_tree_backref(cache, &key, cur);
2940 			if (ret < 0)
2941 				goto out;
2942 			continue;
2943 		} else if (unlikely(key.type == BTRFS_EXTENT_REF_V0_KEY)) {
2944 			ret = -EINVAL;
2945 			btrfs_print_v0_err(fs_info);
2946 			btrfs_handle_fs_error(fs_info, ret, NULL);
2947 			goto out;
2948 		} else if (key.type != BTRFS_TREE_BLOCK_REF_KEY) {
2949 			continue;
2950 		}
2951 
2952 		/*
2953 		 * key.type == BTRFS_TREE_BLOCK_REF_KEY, inline ref offset
2954 		 * means the root objectid. We need to search the tree to get
2955 		 * its parent bytenr.
2956 		 */
2957 		ret = handle_indirect_tree_backref(cache, path, &key, node_key,
2958 						   cur);
2959 		if (ret < 0)
2960 			goto out;
2961 	}
2962 	ret = 0;
2963 	cur->checked = 1;
2964 	WARN_ON(exist);
2965 out:
2966 	btrfs_backref_iter_release(iter);
2967 	return ret;
2968 }
2969 
2970 /*
2971  * Finish the upwards linkage created by btrfs_backref_add_tree_node()
2972  */
2973 int btrfs_backref_finish_upper_links(struct btrfs_backref_cache *cache,
2974 				     struct btrfs_backref_node *start)
2975 {
2976 	struct list_head *useless_node = &cache->useless_node;
2977 	struct btrfs_backref_edge *edge;
2978 	struct rb_node *rb_node;
2979 	LIST_HEAD(pending_edge);
2980 
2981 	ASSERT(start->checked);
2982 
2983 	/* Insert this node to cache if it's not COW-only */
2984 	if (!start->cowonly) {
2985 		rb_node = rb_simple_insert(&cache->rb_root, start->bytenr,
2986 					   &start->rb_node);
2987 		if (rb_node)
2988 			btrfs_backref_panic(cache->fs_info, start->bytenr,
2989 					    -EEXIST);
2990 		list_add_tail(&start->lower, &cache->leaves);
2991 	}
2992 
2993 	/*
2994 	 * Use breadth first search to iterate all related edges.
2995 	 *
2996 	 * The starting points are all the edges of this node
2997 	 */
2998 	list_for_each_entry(edge, &start->upper, list[LOWER])
2999 		list_add_tail(&edge->list[UPPER], &pending_edge);
3000 
3001 	while (!list_empty(&pending_edge)) {
3002 		struct btrfs_backref_node *upper;
3003 		struct btrfs_backref_node *lower;
3004 
3005 		edge = list_first_entry(&pending_edge,
3006 				struct btrfs_backref_edge, list[UPPER]);
3007 		list_del_init(&edge->list[UPPER]);
3008 		upper = edge->node[UPPER];
3009 		lower = edge->node[LOWER];
3010 
3011 		/* Parent is detached, no need to keep any edges */
3012 		if (upper->detached) {
3013 			list_del(&edge->list[LOWER]);
3014 			btrfs_backref_free_edge(cache, edge);
3015 
3016 			/* Lower node is orphan, queue for cleanup */
3017 			if (list_empty(&lower->upper))
3018 				list_add(&lower->list, useless_node);
3019 			continue;
3020 		}
3021 
3022 		/*
3023 		 * All new nodes added in current build_backref_tree() haven't
3024 		 * been linked to the cache rb tree.
3025 		 * So if we have upper->rb_node populated, this means a cache
3026 		 * hit. We only need to link the edge, as @upper and all its
3027 		 * parents have already been linked.
3028 		 */
3029 		if (!RB_EMPTY_NODE(&upper->rb_node)) {
3030 			if (upper->lowest) {
3031 				list_del_init(&upper->lower);
3032 				upper->lowest = 0;
3033 			}
3034 
3035 			list_add_tail(&edge->list[UPPER], &upper->lower);
3036 			continue;
3037 		}
3038 
3039 		/* Sanity check, we shouldn't have any unchecked nodes */
3040 		if (!upper->checked) {
3041 			ASSERT(0);
3042 			return -EUCLEAN;
3043 		}
3044 
3045 		/* Sanity check, COW-only node has non-COW-only parent */
3046 		if (start->cowonly != upper->cowonly) {
3047 			ASSERT(0);
3048 			return -EUCLEAN;
3049 		}
3050 
3051 		/* Only cache non-COW-only (subvolume trees) tree blocks */
3052 		if (!upper->cowonly) {
3053 			rb_node = rb_simple_insert(&cache->rb_root, upper->bytenr,
3054 						   &upper->rb_node);
3055 			if (rb_node) {
3056 				btrfs_backref_panic(cache->fs_info,
3057 						upper->bytenr, -EEXIST);
3058 				return -EUCLEAN;
3059 			}
3060 		}
3061 
3062 		list_add_tail(&edge->list[UPPER], &upper->lower);
3063 
3064 		/*
3065 		 * Also queue all the parent edges of this uncached node
3066 		 * to finish the upper linkage
3067 		 */
3068 		list_for_each_entry(edge, &upper->upper, list[LOWER])
3069 			list_add_tail(&edge->list[UPPER], &pending_edge);
3070 	}
3071 	return 0;
3072 }
3073 
3074 void btrfs_backref_error_cleanup(struct btrfs_backref_cache *cache,
3075 				 struct btrfs_backref_node *node)
3076 {
3077 	struct btrfs_backref_node *lower;
3078 	struct btrfs_backref_node *upper;
3079 	struct btrfs_backref_edge *edge;
3080 
3081 	while (!list_empty(&cache->useless_node)) {
3082 		lower = list_first_entry(&cache->useless_node,
3083 				   struct btrfs_backref_node, list);
3084 		list_del_init(&lower->list);
3085 	}
3086 	while (!list_empty(&cache->pending_edge)) {
3087 		edge = list_first_entry(&cache->pending_edge,
3088 				struct btrfs_backref_edge, list[UPPER]);
3089 		list_del(&edge->list[UPPER]);
3090 		list_del(&edge->list[LOWER]);
3091 		lower = edge->node[LOWER];
3092 		upper = edge->node[UPPER];
3093 		btrfs_backref_free_edge(cache, edge);
3094 
3095 		/*
3096 		 * Lower is no longer linked to any upper backref nodes and
3097 		 * isn't in the cache, we can free it ourselves.
3098 		 */
3099 		if (list_empty(&lower->upper) &&
3100 		    RB_EMPTY_NODE(&lower->rb_node))
3101 			list_add(&lower->list, &cache->useless_node);
3102 
3103 		if (!RB_EMPTY_NODE(&upper->rb_node))
3104 			continue;
3105 
3106 		/* Add this guy's upper edges to the list to process */
3107 		list_for_each_entry(edge, &upper->upper, list[LOWER])
3108 			list_add_tail(&edge->list[UPPER],
3109 				      &cache->pending_edge);
3110 		if (list_empty(&upper->upper))
3111 			list_add(&upper->list, &cache->useless_node);
3112 	}
3113 
3114 	while (!list_empty(&cache->useless_node)) {
3115 		lower = list_first_entry(&cache->useless_node,
3116 				   struct btrfs_backref_node, list);
3117 		list_del_init(&lower->list);
3118 		if (lower == node)
3119 			node = NULL;
3120 		btrfs_backref_drop_node(cache, lower);
3121 	}
3122 
3123 	btrfs_backref_cleanup_node(cache, node);
3124 	ASSERT(list_empty(&cache->useless_node) &&
3125 	       list_empty(&cache->pending_edge));
3126 }
3127