backref.c 46.4 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
/*
 * Copyright (C) 2011 STRATO.  All rights reserved.
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public
 * License v2 as published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * General Public License for more details.
 *
 * You should have received a copy of the GNU General Public
 * License along with this program; if not, write to the
 * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
 * Boston, MA 021110-1307, USA.
 */

19
#include <linux/vmalloc.h>
20 21 22
#include "ctree.h"
#include "disk-io.h"
#include "backref.h"
23 24 25
#include "ulist.h"
#include "transaction.h"
#include "delayed-ref.h"
26
#include "locking.h"
27

28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100
struct extent_inode_elem {
	u64 inum;
	u64 offset;
	struct extent_inode_elem *next;
};

static int check_extent_in_eb(struct btrfs_key *key, struct extent_buffer *eb,
				struct btrfs_file_extent_item *fi,
				u64 extent_item_pos,
				struct extent_inode_elem **eie)
{
	u64 data_offset;
	u64 data_len;
	struct extent_inode_elem *e;

	data_offset = btrfs_file_extent_offset(eb, fi);
	data_len = btrfs_file_extent_num_bytes(eb, fi);

	if (extent_item_pos < data_offset ||
	    extent_item_pos >= data_offset + data_len)
		return 1;

	e = kmalloc(sizeof(*e), GFP_NOFS);
	if (!e)
		return -ENOMEM;

	e->next = *eie;
	e->inum = key->objectid;
	e->offset = key->offset + (extent_item_pos - data_offset);
	*eie = e;

	return 0;
}

static int find_extent_in_eb(struct extent_buffer *eb, u64 wanted_disk_byte,
				u64 extent_item_pos,
				struct extent_inode_elem **eie)
{
	u64 disk_byte;
	struct btrfs_key key;
	struct btrfs_file_extent_item *fi;
	int slot;
	int nritems;
	int extent_type;
	int ret;

	/*
	 * from the shared data ref, we only have the leaf but we need
	 * the key. thus, we must look into all items and see that we
	 * find one (some) with a reference to our extent item.
	 */
	nritems = btrfs_header_nritems(eb);
	for (slot = 0; slot < nritems; ++slot) {
		btrfs_item_key_to_cpu(eb, &key, slot);
		if (key.type != BTRFS_EXTENT_DATA_KEY)
			continue;
		fi = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item);
		extent_type = btrfs_file_extent_type(eb, fi);
		if (extent_type == BTRFS_FILE_EXTENT_INLINE)
			continue;
		/* don't skip BTRFS_FILE_EXTENT_PREALLOC, we can handle that */
		disk_byte = btrfs_file_extent_disk_bytenr(eb, fi);
		if (disk_byte != wanted_disk_byte)
			continue;

		ret = check_extent_in_eb(&key, eb, fi, extent_item_pos, eie);
		if (ret < 0)
			return ret;
	}

	return 0;
}

101 102 103 104 105 106
/*
 * this structure records all encountered refs on the way up to the root
 */
struct __prelim_ref {
	struct list_head list;
	u64 root_id;
107
	struct btrfs_key key_for_search;
108 109
	int level;
	int count;
110
	struct extent_inode_elem *inode_list;
111 112 113 114
	u64 parent;
	u64 wanted_disk_byte;
};

115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153
/*
 * the rules for all callers of this function are:
 * - obtaining the parent is the goal
 * - if you add a key, you must know that it is a correct key
 * - if you cannot add the parent or a correct key, then we will look into the
 *   block later to set a correct key
 *
 * delayed refs
 * ============
 *        backref type | shared | indirect | shared | indirect
 * information         |   tree |     tree |   data |     data
 * --------------------+--------+----------+--------+----------
 *      parent logical |    y   |     -    |    -   |     -
 *      key to resolve |    -   |     y    |    y   |     y
 *  tree block logical |    -   |     -    |    -   |     -
 *  root for resolving |    y   |     y    |    y   |     y
 *
 * - column 1:       we've the parent -> done
 * - column 2, 3, 4: we use the key to find the parent
 *
 * on disk refs (inline or keyed)
 * ==============================
 *        backref type | shared | indirect | shared | indirect
 * information         |   tree |     tree |   data |     data
 * --------------------+--------+----------+--------+----------
 *      parent logical |    y   |     -    |    y   |     -
 *      key to resolve |    -   |     -    |    -   |     y
 *  tree block logical |    y   |     y    |    y   |     y
 *  root for resolving |    -   |     y    |    y   |     y
 *
 * - column 1, 3: we've the parent -> done
 * - column 2:    we take the first key from the block to find the parent
 *                (see __add_missing_keys)
 * - column 4:    we use the key to find the parent
 *
 * additional information that's available but not required to find the parent
 * block might help in merging entries to gain some speed.
 */

154
static int __add_prelim_ref(struct list_head *head, u64 root_id,
155 156
			    struct btrfs_key *key, int level,
			    u64 parent, u64 wanted_disk_byte, int count)
157 158 159 160 161 162 163 164 165 166
{
	struct __prelim_ref *ref;

	/* in case we're adding delayed refs, we're holding the refs spinlock */
	ref = kmalloc(sizeof(*ref), GFP_ATOMIC);
	if (!ref)
		return -ENOMEM;

	ref->root_id = root_id;
	if (key)
167
		ref->key_for_search = *key;
168
	else
169
		memset(&ref->key_for_search, 0, sizeof(ref->key_for_search));
170

171
	ref->inode_list = NULL;
172 173 174 175 176 177 178 179 180 181
	ref->level = level;
	ref->count = count;
	ref->parent = parent;
	ref->wanted_disk_byte = wanted_disk_byte;
	list_add_tail(&ref->list, head);

	return 0;
}

static int add_all_parents(struct btrfs_root *root, struct btrfs_path *path,
182
				struct ulist *parents, int level,
183
				struct btrfs_key *key_for_search, u64 time_seq,
J
Jan Schmidt 已提交
184
				u64 wanted_disk_byte,
185
				const u64 *extent_item_pos)
186
{
187 188 189 190
	int ret = 0;
	int slot;
	struct extent_buffer *eb;
	struct btrfs_key key;
191
	struct btrfs_file_extent_item *fi;
192
	struct extent_inode_elem *eie = NULL;
193 194
	u64 disk_byte;

195 196 197
	if (level != 0) {
		eb = path->nodes[level];
		ret = ulist_add(parents, eb->start, 0, GFP_NOFS);
198 199
		if (ret < 0)
			return ret;
200
		return 0;
201
	}
202 203

	/*
204 205 206
	 * We normally enter this function with the path already pointing to
	 * the first item to check. But sometimes, we may enter it with
	 * slot==nritems. In that case, go to the next leaf before we continue.
207
	 */
208
	if (path->slots[0] >= btrfs_header_nritems(path->nodes[0]))
J
Jan Schmidt 已提交
209
		ret = btrfs_next_old_leaf(root, path, time_seq);
210

211
	while (!ret) {
212
		eb = path->nodes[0];
213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234
		slot = path->slots[0];

		btrfs_item_key_to_cpu(eb, &key, slot);

		if (key.objectid != key_for_search->objectid ||
		    key.type != BTRFS_EXTENT_DATA_KEY)
			break;

		fi = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item);
		disk_byte = btrfs_file_extent_disk_bytenr(eb, fi);

		if (disk_byte == wanted_disk_byte) {
			eie = NULL;
			if (extent_item_pos) {
				ret = check_extent_in_eb(&key, eb, fi,
						*extent_item_pos,
						&eie);
				if (ret < 0)
					break;
			}
			if (!ret) {
				ret = ulist_add(parents, eb->start,
235
						(uintptr_t)eie, GFP_NOFS);
236 237 238 239 240 241 242 243
				if (ret < 0)
					break;
				if (!extent_item_pos) {
					ret = btrfs_next_old_leaf(root, path,
							time_seq);
					continue;
				}
			}
244
		}
245
		ret = btrfs_next_old_item(root, path, time_seq);
246 247
	}

248 249 250
	if (ret > 0)
		ret = 0;
	return ret;
251 252 253 254 255 256 257
}

/*
 * resolve an indirect backref in the form (root_id, key, level)
 * to a logical address
 */
static int __resolve_indirect_ref(struct btrfs_fs_info *fs_info,
258
					int search_commit_root,
259
					u64 time_seq,
260
					struct __prelim_ref *ref,
261 262
					struct ulist *parents,
					const u64 *extent_item_pos)
263 264 265 266 267 268 269 270 271 272 273 274
{
	struct btrfs_path *path;
	struct btrfs_root *root;
	struct btrfs_key root_key;
	struct extent_buffer *eb;
	int ret = 0;
	int root_level;
	int level = ref->level;

	path = btrfs_alloc_path();
	if (!path)
		return -ENOMEM;
275
	path->search_commit_root = !!search_commit_root;
276 277 278 279 280 281 282 283 284 285

	root_key.objectid = ref->root_id;
	root_key.type = BTRFS_ROOT_ITEM_KEY;
	root_key.offset = (u64)-1;
	root = btrfs_read_fs_root_no_name(fs_info, &root_key);
	if (IS_ERR(root)) {
		ret = PTR_ERR(root);
		goto out;
	}

J
Jan Schmidt 已提交
286
	root_level = btrfs_old_root_level(root, time_seq);
287 288 289 290 291

	if (root_level + 1 == level)
		goto out;

	path->lowest_level = level;
292
	ret = btrfs_search_old_slot(root, &ref->key_for_search, path, time_seq);
293 294 295
	pr_debug("search slot in root %llu (level %d, ref count %d) returned "
		 "%d for key (%llu %u %llu)\n",
		 (unsigned long long)ref->root_id, level, ref->count, ret,
296 297 298
		 (unsigned long long)ref->key_for_search.objectid,
		 ref->key_for_search.type,
		 (unsigned long long)ref->key_for_search.offset);
299 300 301 302
	if (ret < 0)
		goto out;

	eb = path->nodes[level];
303 304 305 306 307 308 309 310
	while (!eb) {
		if (!level) {
			WARN_ON(1);
			ret = 1;
			goto out;
		}
		level--;
		eb = path->nodes[level];
311 312
	}

313 314 315
	ret = add_all_parents(root, path, parents, level, &ref->key_for_search,
				time_seq, ref->wanted_disk_byte,
				extent_item_pos);
316 317 318 319 320 321 322 323 324
out:
	btrfs_free_path(path);
	return ret;
}

/*
 * resolve all indirect backrefs from the list
 */
static int __resolve_indirect_refs(struct btrfs_fs_info *fs_info,
325
				   int search_commit_root, u64 time_seq,
326 327
				   struct list_head *head,
				   const u64 *extent_item_pos)
328 329 330 331 332 333 334 335
{
	int err;
	int ret = 0;
	struct __prelim_ref *ref;
	struct __prelim_ref *ref_safe;
	struct __prelim_ref *new_ref;
	struct ulist *parents;
	struct ulist_node *node;
J
Jan Schmidt 已提交
336
	struct ulist_iterator uiter;
337 338 339 340 341 342 343 344 345 346 347 348 349 350 351

	parents = ulist_alloc(GFP_NOFS);
	if (!parents)
		return -ENOMEM;

	/*
	 * _safe allows us to insert directly after the current item without
	 * iterating over the newly inserted items.
	 * we're also allowed to re-assign ref during iteration.
	 */
	list_for_each_entry_safe(ref, ref_safe, head, list) {
		if (ref->parent)	/* already direct */
			continue;
		if (ref->count == 0)
			continue;
352
		err = __resolve_indirect_ref(fs_info, search_commit_root,
353 354
					     time_seq, ref, parents,
					     extent_item_pos);
355 356
		if (err == -ENOMEM)
			goto out;
357
		if (err)
358 359 360
			continue;

		/* we put the first parent into the ref at hand */
J
Jan Schmidt 已提交
361 362
		ULIST_ITER_INIT(&uiter);
		node = ulist_next(parents, &uiter);
363
		ref->parent = node ? node->val : 0;
364 365
		ref->inode_list = node ?
			(struct extent_inode_elem *)(uintptr_t)node->aux : 0;
366 367

		/* additional parents require new refs being added here */
J
Jan Schmidt 已提交
368
		while ((node = ulist_next(parents, &uiter))) {
369 370 371
			new_ref = kmalloc(sizeof(*new_ref), GFP_NOFS);
			if (!new_ref) {
				ret = -ENOMEM;
372
				goto out;
373 374 375
			}
			memcpy(new_ref, ref, sizeof(*ref));
			new_ref->parent = node->val;
376 377
			new_ref->inode_list = (struct extent_inode_elem *)
							(uintptr_t)node->aux;
378 379 380 381
			list_add(&new_ref->list, &ref->list);
		}
		ulist_reinit(parents);
	}
382
out:
383 384 385 386
	ulist_free(parents);
	return ret;
}

387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425
static inline int ref_for_same_block(struct __prelim_ref *ref1,
				     struct __prelim_ref *ref2)
{
	if (ref1->level != ref2->level)
		return 0;
	if (ref1->root_id != ref2->root_id)
		return 0;
	if (ref1->key_for_search.type != ref2->key_for_search.type)
		return 0;
	if (ref1->key_for_search.objectid != ref2->key_for_search.objectid)
		return 0;
	if (ref1->key_for_search.offset != ref2->key_for_search.offset)
		return 0;
	if (ref1->parent != ref2->parent)
		return 0;

	return 1;
}

/*
 * read tree blocks and add keys where required.
 */
static int __add_missing_keys(struct btrfs_fs_info *fs_info,
			      struct list_head *head)
{
	struct list_head *pos;
	struct extent_buffer *eb;

	list_for_each(pos, head) {
		struct __prelim_ref *ref;
		ref = list_entry(pos, struct __prelim_ref, list);

		if (ref->parent)
			continue;
		if (ref->key_for_search.type)
			continue;
		BUG_ON(!ref->wanted_disk_byte);
		eb = read_tree_block(fs_info->tree_root, ref->wanted_disk_byte,
				     fs_info->tree_root->leafsize, 0);
426 427 428 429
		if (!eb || !extent_buffer_uptodate(eb)) {
			free_extent_buffer(eb);
			return -EIO;
		}
430 431 432 433 434 435 436 437 438 439 440
		btrfs_tree_read_lock(eb);
		if (btrfs_header_level(eb) == 0)
			btrfs_item_key_to_cpu(eb, &ref->key_for_search, 0);
		else
			btrfs_node_key_to_cpu(eb, &ref->key_for_search, 0);
		btrfs_tree_read_unlock(eb);
		free_extent_buffer(eb);
	}
	return 0;
}

441 442 443 444
/*
 * merge two lists of backrefs and adjust counts accordingly
 *
 * mode = 1: merge identical keys, if key is set
445 446 447 448
 *    FIXME: if we add more keys in __add_prelim_ref, we can merge more here.
 *           additionally, we could even add a key range for the blocks we
 *           looked into to merge even more (-> replace unresolved refs by those
 *           having a parent).
449 450
 * mode = 2: merge identical parents
 */
451
static void __merge_refs(struct list_head *head, int mode)
452 453 454 455 456 457 458 459 460 461 462 463 464
{
	struct list_head *pos1;

	list_for_each(pos1, head) {
		struct list_head *n2;
		struct list_head *pos2;
		struct __prelim_ref *ref1;

		ref1 = list_entry(pos1, struct __prelim_ref, list);

		for (pos2 = pos1->next, n2 = pos2->next; pos2 != head;
		     pos2 = n2, n2 = pos2->next) {
			struct __prelim_ref *ref2;
465
			struct __prelim_ref *xchg;
466
			struct extent_inode_elem *eie;
467 468 469 470

			ref2 = list_entry(pos2, struct __prelim_ref, list);

			if (mode == 1) {
471
				if (!ref_for_same_block(ref1, ref2))
472
					continue;
473 474 475 476 477
				if (!ref1->parent && ref2->parent) {
					xchg = ref1;
					ref1 = ref2;
					ref2 = xchg;
				}
478 479 480 481
			} else {
				if (ref1->parent != ref2->parent)
					continue;
			}
482 483 484 485 486 487 488 489 490 491

			eie = ref1->inode_list;
			while (eie && eie->next)
				eie = eie->next;
			if (eie)
				eie->next = ref2->inode_list;
			else
				ref1->inode_list = ref2->inode_list;
			ref1->count += ref2->count;

492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507
			list_del(&ref2->list);
			kfree(ref2);
		}

	}
}

/*
 * add all currently queued delayed refs from this head whose seq nr is
 * smaller or equal that seq to the list
 */
static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq,
			      struct list_head *prefs)
{
	struct btrfs_delayed_extent_op *extent_op = head->extent_op;
	struct rb_node *n = &head->node.rb_node;
508 509
	struct btrfs_key key;
	struct btrfs_key op_key = {0};
510
	int sgn;
511
	int ret = 0;
512 513

	if (extent_op && extent_op->update_key)
514
		btrfs_disk_key_to_cpu(&op_key, &extent_op->key);
515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545

	while ((n = rb_prev(n))) {
		struct btrfs_delayed_ref_node *node;
		node = rb_entry(n, struct btrfs_delayed_ref_node,
				rb_node);
		if (node->bytenr != head->node.bytenr)
			break;
		WARN_ON(node->is_head);

		if (node->seq > seq)
			continue;

		switch (node->action) {
		case BTRFS_ADD_DELAYED_EXTENT:
		case BTRFS_UPDATE_DELAYED_HEAD:
			WARN_ON(1);
			continue;
		case BTRFS_ADD_DELAYED_REF:
			sgn = 1;
			break;
		case BTRFS_DROP_DELAYED_REF:
			sgn = -1;
			break;
		default:
			BUG_ON(1);
		}
		switch (node->type) {
		case BTRFS_TREE_BLOCK_REF_KEY: {
			struct btrfs_delayed_tree_ref *ref;

			ref = btrfs_delayed_node_to_tree_ref(node);
546
			ret = __add_prelim_ref(prefs, ref->root, &op_key,
547 548 549 550 551 552 553 554
					       ref->level + 1, 0, node->bytenr,
					       node->ref_mod * sgn);
			break;
		}
		case BTRFS_SHARED_BLOCK_REF_KEY: {
			struct btrfs_delayed_tree_ref *ref;

			ref = btrfs_delayed_node_to_tree_ref(node);
555
			ret = __add_prelim_ref(prefs, ref->root, NULL,
556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588
					       ref->level + 1, ref->parent,
					       node->bytenr,
					       node->ref_mod * sgn);
			break;
		}
		case BTRFS_EXTENT_DATA_REF_KEY: {
			struct btrfs_delayed_data_ref *ref;
			ref = btrfs_delayed_node_to_data_ref(node);

			key.objectid = ref->objectid;
			key.type = BTRFS_EXTENT_DATA_KEY;
			key.offset = ref->offset;
			ret = __add_prelim_ref(prefs, ref->root, &key, 0, 0,
					       node->bytenr,
					       node->ref_mod * sgn);
			break;
		}
		case BTRFS_SHARED_DATA_REF_KEY: {
			struct btrfs_delayed_data_ref *ref;

			ref = btrfs_delayed_node_to_data_ref(node);

			key.objectid = ref->objectid;
			key.type = BTRFS_EXTENT_DATA_KEY;
			key.offset = ref->offset;
			ret = __add_prelim_ref(prefs, ref->root, &key, 0,
					       ref->parent, node->bytenr,
					       node->ref_mod * sgn);
			break;
		}
		default:
			WARN_ON(1);
		}
589 590
		if (ret)
			return ret;
591 592 593 594 595 596 597 598 599 600
	}

	return 0;
}

/*
 * add all inline backrefs for bytenr to the list
 */
static int __add_inline_refs(struct btrfs_fs_info *fs_info,
			     struct btrfs_path *path, u64 bytenr,
601
			     int *info_level, struct list_head *prefs)
602
{
603
	int ret = 0;
604 605 606 607 608 609 610 611 612 613 614 615 616
	int slot;
	struct extent_buffer *leaf;
	struct btrfs_key key;
	unsigned long ptr;
	unsigned long end;
	struct btrfs_extent_item *ei;
	u64 flags;
	u64 item_size;

	/*
	 * enumerate all inline refs
	 */
	leaf = path->nodes[0];
617
	slot = path->slots[0];
618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649

	item_size = btrfs_item_size_nr(leaf, slot);
	BUG_ON(item_size < sizeof(*ei));

	ei = btrfs_item_ptr(leaf, slot, struct btrfs_extent_item);
	flags = btrfs_extent_flags(leaf, ei);

	ptr = (unsigned long)(ei + 1);
	end = (unsigned long)ei + item_size;

	if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) {
		struct btrfs_tree_block_info *info;

		info = (struct btrfs_tree_block_info *)ptr;
		*info_level = btrfs_tree_block_level(leaf, info);
		ptr += sizeof(struct btrfs_tree_block_info);
		BUG_ON(ptr > end);
	} else {
		BUG_ON(!(flags & BTRFS_EXTENT_FLAG_DATA));
	}

	while (ptr < end) {
		struct btrfs_extent_inline_ref *iref;
		u64 offset;
		int type;

		iref = (struct btrfs_extent_inline_ref *)ptr;
		type = btrfs_extent_inline_ref_type(leaf, iref);
		offset = btrfs_extent_inline_ref_offset(leaf, iref);

		switch (type) {
		case BTRFS_SHARED_BLOCK_REF_KEY:
650
			ret = __add_prelim_ref(prefs, 0, NULL,
651 652 653 654 655 656 657 658 659 660 661 662 663 664
						*info_level + 1, offset,
						bytenr, 1);
			break;
		case BTRFS_SHARED_DATA_REF_KEY: {
			struct btrfs_shared_data_ref *sdref;
			int count;

			sdref = (struct btrfs_shared_data_ref *)(iref + 1);
			count = btrfs_shared_data_ref_count(leaf, sdref);
			ret = __add_prelim_ref(prefs, 0, NULL, 0, offset,
					       bytenr, count);
			break;
		}
		case BTRFS_TREE_BLOCK_REF_KEY:
665 666 667
			ret = __add_prelim_ref(prefs, offset, NULL,
					       *info_level + 1, 0,
					       bytenr, 1);
668 669 670 671 672 673 674 675 676 677 678 679 680
			break;
		case BTRFS_EXTENT_DATA_REF_KEY: {
			struct btrfs_extent_data_ref *dref;
			int count;
			u64 root;

			dref = (struct btrfs_extent_data_ref *)(&iref->offset);
			count = btrfs_extent_data_ref_count(leaf, dref);
			key.objectid = btrfs_extent_data_ref_objectid(leaf,
								      dref);
			key.type = BTRFS_EXTENT_DATA_KEY;
			key.offset = btrfs_extent_data_ref_offset(leaf, dref);
			root = btrfs_extent_data_ref_root(leaf, dref);
681 682
			ret = __add_prelim_ref(prefs, root, &key, 0, 0,
					       bytenr, count);
683 684 685 686 687
			break;
		}
		default:
			WARN_ON(1);
		}
688 689
		if (ret)
			return ret;
690 691 692 693 694 695 696 697 698 699 700
		ptr += btrfs_extent_inline_ref_size(type);
	}

	return 0;
}

/*
 * add all non-inline backrefs for bytenr to the list
 */
static int __add_keyed_refs(struct btrfs_fs_info *fs_info,
			    struct btrfs_path *path, u64 bytenr,
701
			    int info_level, struct list_head *prefs)
702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730
{
	struct btrfs_root *extent_root = fs_info->extent_root;
	int ret;
	int slot;
	struct extent_buffer *leaf;
	struct btrfs_key key;

	while (1) {
		ret = btrfs_next_item(extent_root, path);
		if (ret < 0)
			break;
		if (ret) {
			ret = 0;
			break;
		}

		slot = path->slots[0];
		leaf = path->nodes[0];
		btrfs_item_key_to_cpu(leaf, &key, slot);

		if (key.objectid != bytenr)
			break;
		if (key.type < BTRFS_TREE_BLOCK_REF_KEY)
			continue;
		if (key.type > BTRFS_SHARED_DATA_REF_KEY)
			break;

		switch (key.type) {
		case BTRFS_SHARED_BLOCK_REF_KEY:
731
			ret = __add_prelim_ref(prefs, 0, NULL,
732 733 734 735 736 737 738 739 740 741 742 743 744 745 746
						info_level + 1, key.offset,
						bytenr, 1);
			break;
		case BTRFS_SHARED_DATA_REF_KEY: {
			struct btrfs_shared_data_ref *sdref;
			int count;

			sdref = btrfs_item_ptr(leaf, slot,
					      struct btrfs_shared_data_ref);
			count = btrfs_shared_data_ref_count(leaf, sdref);
			ret = __add_prelim_ref(prefs, 0, NULL, 0, key.offset,
						bytenr, count);
			break;
		}
		case BTRFS_TREE_BLOCK_REF_KEY:
747 748 749
			ret = __add_prelim_ref(prefs, key.offset, NULL,
					       info_level + 1, 0,
					       bytenr, 1);
750 751 752 753 754 755 756 757 758 759 760 761 762 763 764
			break;
		case BTRFS_EXTENT_DATA_REF_KEY: {
			struct btrfs_extent_data_ref *dref;
			int count;
			u64 root;

			dref = btrfs_item_ptr(leaf, slot,
					      struct btrfs_extent_data_ref);
			count = btrfs_extent_data_ref_count(leaf, dref);
			key.objectid = btrfs_extent_data_ref_objectid(leaf,
								      dref);
			key.type = BTRFS_EXTENT_DATA_KEY;
			key.offset = btrfs_extent_data_ref_offset(leaf, dref);
			root = btrfs_extent_data_ref_root(leaf, dref);
			ret = __add_prelim_ref(prefs, root, &key, 0, 0,
765
					       bytenr, count);
766 767 768 769 770
			break;
		}
		default:
			WARN_ON(1);
		}
771 772 773
		if (ret)
			return ret;

774 775 776 777 778 779 780 781 782 783 784 785 786 787 788
	}

	return ret;
}

/*
 * this adds all existing backrefs (inline backrefs, backrefs and delayed
 * refs) for the given bytenr to the refs list, merges duplicates and resolves
 * indirect refs to their parent bytenr.
 * When roots are found, they're added to the roots list
 *
 * FIXME some caching might speed things up
 */
static int find_parent_nodes(struct btrfs_trans_handle *trans,
			     struct btrfs_fs_info *fs_info, u64 bytenr,
789 790
			     u64 time_seq, struct ulist *refs,
			     struct ulist *roots, const u64 *extent_item_pos)
791 792 793 794
{
	struct btrfs_key key;
	struct btrfs_path *path;
	struct btrfs_delayed_ref_root *delayed_refs = NULL;
795
	struct btrfs_delayed_ref_head *head;
796 797
	int info_level = 0;
	int ret;
798
	int search_commit_root = (trans == BTRFS_BACKREF_SEARCH_COMMIT_ROOT);
799 800 801 802 803 804 805 806 807 808 809 810 811 812
	struct list_head prefs_delayed;
	struct list_head prefs;
	struct __prelim_ref *ref;

	INIT_LIST_HEAD(&prefs);
	INIT_LIST_HEAD(&prefs_delayed);

	key.objectid = bytenr;
	key.type = BTRFS_EXTENT_ITEM_KEY;
	key.offset = (u64)-1;

	path = btrfs_alloc_path();
	if (!path)
		return -ENOMEM;
813
	path->search_commit_root = !!search_commit_root;
814 815 816 817 818 819 820

	/*
	 * grab both a lock on the path and a lock on the delayed ref head.
	 * We need both to get a consistent picture of how the refs look
	 * at a specified point in time
	 */
again:
821 822
	head = NULL;

823 824 825 826 827
	ret = btrfs_search_slot(trans, fs_info->extent_root, &key, path, 0, 0);
	if (ret < 0)
		goto out;
	BUG_ON(ret == 0);

828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851
	if (trans != BTRFS_BACKREF_SEARCH_COMMIT_ROOT) {
		/*
		 * look if there are updates for this ref queued and lock the
		 * head
		 */
		delayed_refs = &trans->transaction->delayed_refs;
		spin_lock(&delayed_refs->lock);
		head = btrfs_find_delayed_ref_head(trans, bytenr);
		if (head) {
			if (!mutex_trylock(&head->mutex)) {
				atomic_inc(&head->node.refs);
				spin_unlock(&delayed_refs->lock);

				btrfs_release_path(path);

				/*
				 * Mutex was contended, block until it's
				 * released and try again
				 */
				mutex_lock(&head->mutex);
				mutex_unlock(&head->mutex);
				btrfs_put_delayed_ref(&head->node);
				goto again;
			}
852
			ret = __add_delayed_refs(head, time_seq,
853
						 &prefs_delayed);
854
			mutex_unlock(&head->mutex);
855 856 857 858
			if (ret) {
				spin_unlock(&delayed_refs->lock);
				goto out;
			}
859
		}
860
		spin_unlock(&delayed_refs->lock);
861 862 863 864 865 866
	}

	if (path->slots[0]) {
		struct extent_buffer *leaf;
		int slot;

867
		path->slots[0]--;
868
		leaf = path->nodes[0];
869
		slot = path->slots[0];
870 871 872 873
		btrfs_item_key_to_cpu(leaf, &key, slot);
		if (key.objectid == bytenr &&
		    key.type == BTRFS_EXTENT_ITEM_KEY) {
			ret = __add_inline_refs(fs_info, path, bytenr,
874
						&info_level, &prefs);
875 876
			if (ret)
				goto out;
877
			ret = __add_keyed_refs(fs_info, path, bytenr,
878 879 880 881 882 883 884 885 886
					       info_level, &prefs);
			if (ret)
				goto out;
		}
	}
	btrfs_release_path(path);

	list_splice_init(&prefs_delayed, &prefs);

887 888 889 890
	ret = __add_missing_keys(fs_info, &prefs);
	if (ret)
		goto out;

891
	__merge_refs(&prefs, 1);
892

893 894
	ret = __resolve_indirect_refs(fs_info, search_commit_root, time_seq,
				      &prefs, extent_item_pos);
895 896 897
	if (ret)
		goto out;

898
	__merge_refs(&prefs, 2);
899 900 901 902

	while (!list_empty(&prefs)) {
		ref = list_first_entry(&prefs, struct __prelim_ref, list);
		list_del(&ref->list);
J
Julia Lawall 已提交
903
		WARN_ON(ref->count < 0);
904 905 906
		if (ref->count && ref->root_id && ref->parent == 0) {
			/* no parent == root of tree */
			ret = ulist_add(roots, ref->root_id, 0, GFP_NOFS);
907 908
			if (ret < 0)
				goto out;
909 910
		}
		if (ref->count && ref->parent) {
911
			struct extent_inode_elem *eie = NULL;
912
			if (extent_item_pos && !ref->inode_list) {
913 914 915 916 917 918
				u32 bsz;
				struct extent_buffer *eb;
				bsz = btrfs_level_size(fs_info->extent_root,
							info_level);
				eb = read_tree_block(fs_info->extent_root,
							   ref->parent, bsz, 0);
919 920 921 922
				if (!eb || !extent_buffer_uptodate(eb)) {
					free_extent_buffer(eb);
					return -EIO;
				}
923 924
				ret = find_extent_in_eb(eb, bytenr,
							*extent_item_pos, &eie);
925
				ref->inode_list = eie;
926 927
				free_extent_buffer(eb);
			}
928
			ret = ulist_add_merge(refs, ref->parent,
929
					      (uintptr_t)ref->inode_list,
930
					      (u64 *)&eie, GFP_NOFS);
931 932
			if (ret < 0)
				goto out;
933 934 935 936 937 938 939 940 941 942
			if (!ret && extent_item_pos) {
				/*
				 * we've recorded that parent, so we must extend
				 * its inode list here
				 */
				BUG_ON(!eie);
				while (eie->next)
					eie = eie->next;
				eie->next = ref->inode_list;
			}
943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963
		}
		kfree(ref);
	}

out:
	btrfs_free_path(path);
	while (!list_empty(&prefs)) {
		ref = list_first_entry(&prefs, struct __prelim_ref, list);
		list_del(&ref->list);
		kfree(ref);
	}
	while (!list_empty(&prefs_delayed)) {
		ref = list_first_entry(&prefs_delayed, struct __prelim_ref,
				       list);
		list_del(&ref->list);
		kfree(ref);
	}

	return ret;
}

964 965 966 967 968 969 970 971 972 973 974
static void free_leaf_list(struct ulist *blocks)
{
	struct ulist_node *node = NULL;
	struct extent_inode_elem *eie;
	struct extent_inode_elem *eie_next;
	struct ulist_iterator uiter;

	ULIST_ITER_INIT(&uiter);
	while ((node = ulist_next(blocks, &uiter))) {
		if (!node->aux)
			continue;
975
		eie = (struct extent_inode_elem *)(uintptr_t)node->aux;
976 977 978 979 980 981 982 983 984 985
		for (; eie; eie = eie_next) {
			eie_next = eie->next;
			kfree(eie);
		}
		node->aux = 0;
	}

	ulist_free(blocks);
}

986 987 988 989 990 991 992 993 994 995
/*
 * Finds all leafs with a reference to the specified combination of bytenr and
 * offset. key_list_head will point to a list of corresponding keys (caller must
 * free each list element). The leafs will be stored in the leafs ulist, which
 * must be freed with ulist_free.
 *
 * returns 0 on success, <0 on error
 */
static int btrfs_find_all_leafs(struct btrfs_trans_handle *trans,
				struct btrfs_fs_info *fs_info, u64 bytenr,
996
				u64 time_seq, struct ulist **leafs,
997
				const u64 *extent_item_pos)
998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010
{
	struct ulist *tmp;
	int ret;

	tmp = ulist_alloc(GFP_NOFS);
	if (!tmp)
		return -ENOMEM;
	*leafs = ulist_alloc(GFP_NOFS);
	if (!*leafs) {
		ulist_free(tmp);
		return -ENOMEM;
	}

1011
	ret = find_parent_nodes(trans, fs_info, bytenr,
1012
				time_seq, *leafs, tmp, extent_item_pos);
1013 1014 1015
	ulist_free(tmp);

	if (ret < 0 && ret != -ENOENT) {
1016
		free_leaf_list(*leafs);
1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037
		return ret;
	}

	return 0;
}

/*
 * walk all backrefs for a given extent to find all roots that reference this
 * extent. Walking a backref means finding all extents that reference this
 * extent and in turn walk the backrefs of those, too. Naturally this is a
 * recursive process, but here it is implemented in an iterative fashion: We
 * find all referencing extents for the extent in question and put them on a
 * list. In turn, we find all referencing extents for those, further appending
 * to the list. The way we iterate the list allows adding more elements after
 * the current while iterating. The process stops when we reach the end of the
 * list. Found roots are added to the roots list.
 *
 * returns 0 on success, < 0 on error.
 */
int btrfs_find_all_roots(struct btrfs_trans_handle *trans,
				struct btrfs_fs_info *fs_info, u64 bytenr,
1038
				u64 time_seq, struct ulist **roots)
1039 1040 1041
{
	struct ulist *tmp;
	struct ulist_node *node = NULL;
J
Jan Schmidt 已提交
1042
	struct ulist_iterator uiter;
1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053
	int ret;

	tmp = ulist_alloc(GFP_NOFS);
	if (!tmp)
		return -ENOMEM;
	*roots = ulist_alloc(GFP_NOFS);
	if (!*roots) {
		ulist_free(tmp);
		return -ENOMEM;
	}

J
Jan Schmidt 已提交
1054
	ULIST_ITER_INIT(&uiter);
1055
	while (1) {
1056
		ret = find_parent_nodes(trans, fs_info, bytenr,
1057
					time_seq, tmp, *roots, NULL);
1058 1059 1060 1061 1062
		if (ret < 0 && ret != -ENOENT) {
			ulist_free(tmp);
			ulist_free(*roots);
			return ret;
		}
J
Jan Schmidt 已提交
1063
		node = ulist_next(tmp, &uiter);
1064 1065 1066 1067 1068 1069 1070 1071 1072 1073
		if (!node)
			break;
		bytenr = node->val;
	}

	ulist_free(tmp);
	return 0;
}


1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123
static int __inode_info(u64 inum, u64 ioff, u8 key_type,
			struct btrfs_root *fs_root, struct btrfs_path *path,
			struct btrfs_key *found_key)
{
	int ret;
	struct btrfs_key key;
	struct extent_buffer *eb;

	key.type = key_type;
	key.objectid = inum;
	key.offset = ioff;

	ret = btrfs_search_slot(NULL, fs_root, &key, path, 0, 0);
	if (ret < 0)
		return ret;

	eb = path->nodes[0];
	if (ret && path->slots[0] >= btrfs_header_nritems(eb)) {
		ret = btrfs_next_leaf(fs_root, path);
		if (ret)
			return ret;
		eb = path->nodes[0];
	}

	btrfs_item_key_to_cpu(eb, found_key, path->slots[0]);
	if (found_key->type != key.type || found_key->objectid != key.objectid)
		return 1;

	return 0;
}

/*
 * this makes the path point to (inum INODE_ITEM ioff)
 */
int inode_item_info(u64 inum, u64 ioff, struct btrfs_root *fs_root,
			struct btrfs_path *path)
{
	struct btrfs_key key;
	return __inode_info(inum, ioff, BTRFS_INODE_ITEM_KEY, fs_root, path,
				&key);
}

static int inode_ref_info(u64 inum, u64 ioff, struct btrfs_root *fs_root,
				struct btrfs_path *path,
				struct btrfs_key *found_key)
{
	return __inode_info(inum, ioff, BTRFS_INODE_REF_KEY, fs_root, path,
				found_key);
}

M
Mark Fasheh 已提交
1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191
int btrfs_find_one_extref(struct btrfs_root *root, u64 inode_objectid,
			  u64 start_off, struct btrfs_path *path,
			  struct btrfs_inode_extref **ret_extref,
			  u64 *found_off)
{
	int ret, slot;
	struct btrfs_key key;
	struct btrfs_key found_key;
	struct btrfs_inode_extref *extref;
	struct extent_buffer *leaf;
	unsigned long ptr;

	key.objectid = inode_objectid;
	btrfs_set_key_type(&key, BTRFS_INODE_EXTREF_KEY);
	key.offset = start_off;

	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
	if (ret < 0)
		return ret;

	while (1) {
		leaf = path->nodes[0];
		slot = path->slots[0];
		if (slot >= btrfs_header_nritems(leaf)) {
			/*
			 * If the item at offset is not found,
			 * btrfs_search_slot will point us to the slot
			 * where it should be inserted. In our case
			 * that will be the slot directly before the
			 * next INODE_REF_KEY_V2 item. In the case
			 * that we're pointing to the last slot in a
			 * leaf, we must move one leaf over.
			 */
			ret = btrfs_next_leaf(root, path);
			if (ret) {
				if (ret >= 1)
					ret = -ENOENT;
				break;
			}
			continue;
		}

		btrfs_item_key_to_cpu(leaf, &found_key, slot);

		/*
		 * Check that we're still looking at an extended ref key for
		 * this particular objectid. If we have different
		 * objectid or type then there are no more to be found
		 * in the tree and we can exit.
		 */
		ret = -ENOENT;
		if (found_key.objectid != inode_objectid)
			break;
		if (btrfs_key_type(&found_key) != BTRFS_INODE_EXTREF_KEY)
			break;

		ret = 0;
		ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
		extref = (struct btrfs_inode_extref *)ptr;
		*ret_extref = extref;
		if (found_off)
			*found_off = found_key.offset;
		break;
	}

	return ret;
}

1192 1193 1194 1195
char *btrfs_ref_to_path(struct btrfs_root *fs_root, struct btrfs_path *path,
			u32 name_len, unsigned long name_off,
			struct extent_buffer *eb_in, u64 parent,
			char *dest, u32 size)
1196 1197 1198 1199
{
	int slot;
	u64 next_inum;
	int ret;
1200
	s64 bytes_left = ((s64)size) - 1;
1201 1202
	struct extent_buffer *eb = eb_in;
	struct btrfs_key found_key;
1203
	int leave_spinning = path->leave_spinning;
M
Mark Fasheh 已提交
1204
	struct btrfs_inode_ref *iref;
1205 1206 1207 1208

	if (bytes_left >= 0)
		dest[bytes_left] = '\0';

1209
	path->leave_spinning = 1;
1210
	while (1) {
M
Mark Fasheh 已提交
1211
		bytes_left -= name_len;
1212 1213
		if (bytes_left >= 0)
			read_extent_buffer(eb, dest + bytes_left,
M
Mark Fasheh 已提交
1214
					   name_off, name_len);
1215 1216
		if (eb != eb_in) {
			btrfs_tree_read_unlock_blocking(eb);
1217
			free_extent_buffer(eb);
1218
		}
1219
		ret = inode_ref_info(parent, 0, fs_root, path, &found_key);
1220 1221
		if (ret > 0)
			ret = -ENOENT;
1222 1223
		if (ret)
			break;
M
Mark Fasheh 已提交
1224

1225 1226 1227 1228 1229 1230 1231 1232 1233
		next_inum = found_key.offset;

		/* regular exit ahead */
		if (parent == next_inum)
			break;

		slot = path->slots[0];
		eb = path->nodes[0];
		/* make sure we can use eb after releasing the path */
1234
		if (eb != eb_in) {
1235
			atomic_inc(&eb->refs);
1236 1237 1238
			btrfs_tree_read_lock(eb);
			btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK);
		}
1239 1240
		btrfs_release_path(path);
		iref = btrfs_item_ptr(eb, slot, struct btrfs_inode_ref);
M
Mark Fasheh 已提交
1241 1242 1243 1244

		name_len = btrfs_inode_ref_name_len(eb, iref);
		name_off = (unsigned long)(iref + 1);

1245 1246 1247 1248 1249 1250 1251
		parent = next_inum;
		--bytes_left;
		if (bytes_left >= 0)
			dest[bytes_left] = '/';
	}

	btrfs_release_path(path);
1252
	path->leave_spinning = leave_spinning;
1253 1254 1255 1256 1257 1258 1259

	if (ret)
		return ERR_PTR(ret);

	return dest + bytes_left;
}

M
Mark Fasheh 已提交
1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279
/*
 * this iterates to turn a btrfs_inode_ref into a full filesystem path. elements
 * of the path are separated by '/' and the path is guaranteed to be
 * 0-terminated. the path is only given within the current file system.
 * Therefore, it never starts with a '/'. the caller is responsible to provide
 * "size" bytes in "dest". the dest buffer will be filled backwards. finally,
 * the start point of the resulting string is returned. this pointer is within
 * dest, normally.
 * in case the path buffer would overflow, the pointer is decremented further
 * as if output was written to the buffer, though no more output is actually
 * generated. that way, the caller can determine how much space would be
 * required for the path to fit into the buffer. in that case, the returned
 * value will be smaller than dest. callers must check this!
 */
char *btrfs_iref_to_path(struct btrfs_root *fs_root,
			 struct btrfs_path *path,
			 struct btrfs_inode_ref *iref,
			 struct extent_buffer *eb_in, u64 parent,
			 char *dest, u32 size)
{
1280 1281 1282 1283
	return btrfs_ref_to_path(fs_root, path,
				 btrfs_inode_ref_name_len(eb_in, iref),
				 (unsigned long)(iref + 1),
				 eb_in, parent, dest, size);
M
Mark Fasheh 已提交
1284 1285
}

1286 1287 1288 1289 1290 1291
/*
 * this makes the path point to (logical EXTENT_ITEM *)
 * returns BTRFS_EXTENT_FLAG_DATA for data, BTRFS_EXTENT_FLAG_TREE_BLOCK for
 * tree blocks and <0 on error.
 */
int extent_from_logical(struct btrfs_fs_info *fs_info, u64 logical,
1292 1293
			struct btrfs_path *path, struct btrfs_key *found_key,
			u64 *flags_ret)
1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316
{
	int ret;
	u64 flags;
	u32 item_size;
	struct extent_buffer *eb;
	struct btrfs_extent_item *ei;
	struct btrfs_key key;

	key.type = BTRFS_EXTENT_ITEM_KEY;
	key.objectid = logical;
	key.offset = (u64)-1;

	ret = btrfs_search_slot(NULL, fs_info->extent_root, &key, path, 0, 0);
	if (ret < 0)
		return ret;
	ret = btrfs_previous_item(fs_info->extent_root, path,
					0, BTRFS_EXTENT_ITEM_KEY);
	if (ret < 0)
		return ret;

	btrfs_item_key_to_cpu(path->nodes[0], found_key, path->slots[0]);
	if (found_key->type != BTRFS_EXTENT_ITEM_KEY ||
	    found_key->objectid > logical ||
J
Jan Schmidt 已提交
1317 1318 1319
	    found_key->objectid + found_key->offset <= logical) {
		pr_debug("logical %llu is not within any extent\n",
			 (unsigned long long)logical);
1320
		return -ENOENT;
J
Jan Schmidt 已提交
1321
	}
1322 1323 1324 1325 1326 1327 1328 1329

	eb = path->nodes[0];
	item_size = btrfs_item_size_nr(eb, path->slots[0]);
	BUG_ON(item_size < sizeof(*ei));

	ei = btrfs_item_ptr(eb, path->slots[0], struct btrfs_extent_item);
	flags = btrfs_extent_flags(eb, ei);

J
Jan Schmidt 已提交
1330 1331 1332 1333 1334 1335 1336
	pr_debug("logical %llu is at position %llu within the extent (%llu "
		 "EXTENT_ITEM %llu) flags %#llx size %u\n",
		 (unsigned long long)logical,
		 (unsigned long long)(logical - found_key->objectid),
		 (unsigned long long)found_key->objectid,
		 (unsigned long long)found_key->offset,
		 (unsigned long long)flags, item_size);
1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347

	WARN_ON(!flags_ret);
	if (flags_ret) {
		if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK)
			*flags_ret = BTRFS_EXTENT_FLAG_TREE_BLOCK;
		else if (flags & BTRFS_EXTENT_FLAG_DATA)
			*flags_ret = BTRFS_EXTENT_FLAG_DATA;
		else
			BUG_ON(1);
		return 0;
	}
1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439

	return -EIO;
}

/*
 * helper function to iterate extent inline refs. ptr must point to a 0 value
 * for the first call and may be modified. it is used to track state.
 * if more refs exist, 0 is returned and the next call to
 * __get_extent_inline_ref must pass the modified ptr parameter to get the
 * next ref. after the last ref was processed, 1 is returned.
 * returns <0 on error
 */
static int __get_extent_inline_ref(unsigned long *ptr, struct extent_buffer *eb,
				struct btrfs_extent_item *ei, u32 item_size,
				struct btrfs_extent_inline_ref **out_eiref,
				int *out_type)
{
	unsigned long end;
	u64 flags;
	struct btrfs_tree_block_info *info;

	if (!*ptr) {
		/* first call */
		flags = btrfs_extent_flags(eb, ei);
		if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) {
			info = (struct btrfs_tree_block_info *)(ei + 1);
			*out_eiref =
				(struct btrfs_extent_inline_ref *)(info + 1);
		} else {
			*out_eiref = (struct btrfs_extent_inline_ref *)(ei + 1);
		}
		*ptr = (unsigned long)*out_eiref;
		if ((void *)*ptr >= (void *)ei + item_size)
			return -ENOENT;
	}

	end = (unsigned long)ei + item_size;
	*out_eiref = (struct btrfs_extent_inline_ref *)*ptr;
	*out_type = btrfs_extent_inline_ref_type(eb, *out_eiref);

	*ptr += btrfs_extent_inline_ref_size(*out_type);
	WARN_ON(*ptr > end);
	if (*ptr == end)
		return 1; /* last */

	return 0;
}

/*
 * reads the tree block backref for an extent. tree level and root are returned
 * through out_level and out_root. ptr must point to a 0 value for the first
 * call and may be modified (see __get_extent_inline_ref comment).
 * returns 0 if data was provided, 1 if there was no more data to provide or
 * <0 on error.
 */
int tree_backref_for_extent(unsigned long *ptr, struct extent_buffer *eb,
				struct btrfs_extent_item *ei, u32 item_size,
				u64 *out_root, u8 *out_level)
{
	int ret;
	int type;
	struct btrfs_tree_block_info *info;
	struct btrfs_extent_inline_ref *eiref;

	if (*ptr == (unsigned long)-1)
		return 1;

	while (1) {
		ret = __get_extent_inline_ref(ptr, eb, ei, item_size,
						&eiref, &type);
		if (ret < 0)
			return ret;

		if (type == BTRFS_TREE_BLOCK_REF_KEY ||
		    type == BTRFS_SHARED_BLOCK_REF_KEY)
			break;

		if (ret == 1)
			return 1;
	}

	/* we can treat both ref types equally here */
	info = (struct btrfs_tree_block_info *)(ei + 1);
	*out_root = btrfs_extent_inline_ref_offset(eb, eiref);
	*out_level = btrfs_tree_block_level(eb, info);

	if (ret == 1)
		*ptr = (unsigned long)-1;

	return 0;
}

1440 1441
static int iterate_leaf_refs(struct extent_inode_elem *inode_list,
				u64 root, u64 extent_item_objectid,
J
Jan Schmidt 已提交
1442
				iterate_extent_inodes_t *iterate, void *ctx)
1443
{
1444
	struct extent_inode_elem *eie;
J
Jan Schmidt 已提交
1445 1446
	int ret = 0;

1447
	for (eie = inode_list; eie; eie = eie->next) {
J
Jan Schmidt 已提交
1448
		pr_debug("ref for %llu resolved, key (%llu EXTEND_DATA %llu), "
1449 1450 1451
			 "root %llu\n", extent_item_objectid,
			 eie->inum, eie->offset, root);
		ret = iterate(eie->inum, eie->offset, root, ctx);
J
Jan Schmidt 已提交
1452
		if (ret) {
1453 1454
			pr_debug("stopping iteration for %llu due to ret=%d\n",
				 extent_item_objectid, ret);
J
Jan Schmidt 已提交
1455 1456
			break;
		}
1457 1458 1459 1460 1461 1462 1463
	}

	return ret;
}

/*
 * calls iterate() for every inode that references the extent identified by
J
Jan Schmidt 已提交
1464
 * the given parameters.
1465 1466 1467
 * when the iterator function returns a non-zero value, iteration stops.
 */
int iterate_extent_inodes(struct btrfs_fs_info *fs_info,
J
Jan Schmidt 已提交
1468
				u64 extent_item_objectid, u64 extent_item_pos,
1469
				int search_commit_root,
1470 1471 1472
				iterate_extent_inodes_t *iterate, void *ctx)
{
	int ret;
J
Jan Schmidt 已提交
1473
	struct btrfs_trans_handle *trans;
1474 1475
	struct ulist *refs = NULL;
	struct ulist *roots = NULL;
J
Jan Schmidt 已提交
1476 1477
	struct ulist_node *ref_node = NULL;
	struct ulist_node *root_node = NULL;
1478
	struct seq_list tree_mod_seq_elem = {};
J
Jan Schmidt 已提交
1479 1480
	struct ulist_iterator ref_uiter;
	struct ulist_iterator root_uiter;
1481

J
Jan Schmidt 已提交
1482 1483
	pr_debug("resolving all inodes for extent %llu\n",
			extent_item_objectid);
1484

1485 1486 1487 1488 1489 1490
	if (search_commit_root) {
		trans = BTRFS_BACKREF_SEARCH_COMMIT_ROOT;
	} else {
		trans = btrfs_join_transaction(fs_info->extent_root);
		if (IS_ERR(trans))
			return PTR_ERR(trans);
1491
		btrfs_get_tree_mod_seq(fs_info, &tree_mod_seq_elem);
1492
	}
1493

J
Jan Schmidt 已提交
1494
	ret = btrfs_find_all_leafs(trans, fs_info, extent_item_objectid,
1495
				   tree_mod_seq_elem.seq, &refs,
1496
				   &extent_item_pos);
J
Jan Schmidt 已提交
1497 1498
	if (ret)
		goto out;
1499

J
Jan Schmidt 已提交
1500 1501
	ULIST_ITER_INIT(&ref_uiter);
	while (!ret && (ref_node = ulist_next(refs, &ref_uiter))) {
1502
		ret = btrfs_find_all_roots(trans, fs_info, ref_node->val,
1503
					   tree_mod_seq_elem.seq, &roots);
J
Jan Schmidt 已提交
1504 1505
		if (ret)
			break;
J
Jan Schmidt 已提交
1506 1507
		ULIST_ITER_INIT(&root_uiter);
		while (!ret && (root_node = ulist_next(roots, &root_uiter))) {
1508
			pr_debug("root %llu references leaf %llu, data list "
1509
				 "%#llx\n", root_node->val, ref_node->val,
1510 1511 1512 1513 1514 1515
				 (long long)ref_node->aux);
			ret = iterate_leaf_refs((struct extent_inode_elem *)
						(uintptr_t)ref_node->aux,
						root_node->val,
						extent_item_objectid,
						iterate, ctx);
J
Jan Schmidt 已提交
1516
		}
1517
		ulist_free(roots);
1518 1519
	}

1520
	free_leaf_list(refs);
J
Jan Schmidt 已提交
1521
out:
1522
	if (!search_commit_root) {
1523
		btrfs_put_tree_mod_seq(fs_info, &tree_mod_seq_elem);
1524 1525 1526
		btrfs_end_transaction(trans, fs_info->extent_root);
	}

1527 1528 1529 1530 1531 1532 1533 1534
	return ret;
}

int iterate_inodes_from_logical(u64 logical, struct btrfs_fs_info *fs_info,
				struct btrfs_path *path,
				iterate_extent_inodes_t *iterate, void *ctx)
{
	int ret;
J
Jan Schmidt 已提交
1535
	u64 extent_item_pos;
1536
	u64 flags = 0;
1537
	struct btrfs_key found_key;
1538
	int search_commit_root = path->search_commit_root;
1539

1540
	ret = extent_from_logical(fs_info, logical, path, &found_key, &flags);
J
Jan Schmidt 已提交
1541
	btrfs_release_path(path);
1542 1543
	if (ret < 0)
		return ret;
1544
	if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK)
1545
		return -EINVAL;
1546

J
Jan Schmidt 已提交
1547
	extent_item_pos = logical - found_key.objectid;
1548 1549 1550
	ret = iterate_extent_inodes(fs_info, found_key.objectid,
					extent_item_pos, search_commit_root,
					iterate, ctx);
1551 1552 1553 1554

	return ret;
}

M
Mark Fasheh 已提交
1555 1556 1557 1558 1559 1560
typedef int (iterate_irefs_t)(u64 parent, u32 name_len, unsigned long name_off,
			      struct extent_buffer *eb, void *ctx);

static int iterate_inode_refs(u64 inum, struct btrfs_root *fs_root,
			      struct btrfs_path *path,
			      iterate_irefs_t *iterate, void *ctx)
1561
{
1562
	int ret = 0;
1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573
	int slot;
	u32 cur;
	u32 len;
	u32 name_len;
	u64 parent = 0;
	int found = 0;
	struct extent_buffer *eb;
	struct btrfs_item *item;
	struct btrfs_inode_ref *iref;
	struct btrfs_key found_key;

1574
	while (!ret) {
1575
		path->leave_spinning = 1;
1576
		ret = inode_ref_info(inum, parent ? parent+1 : 0, fs_root, path,
M
Mark Fasheh 已提交
1577
				     &found_key);
1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590
		if (ret < 0)
			break;
		if (ret) {
			ret = found ? 0 : -ENOENT;
			break;
		}
		++found;

		parent = found_key.offset;
		slot = path->slots[0];
		eb = path->nodes[0];
		/* make sure we can use eb after releasing the path */
		atomic_inc(&eb->refs);
1591 1592
		btrfs_tree_read_lock(eb);
		btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK);
1593 1594 1595 1596 1597 1598 1599 1600
		btrfs_release_path(path);

		item = btrfs_item_nr(eb, slot);
		iref = btrfs_item_ptr(eb, slot, struct btrfs_inode_ref);

		for (cur = 0; cur < btrfs_item_size(eb, item); cur += len) {
			name_len = btrfs_inode_ref_name_len(eb, iref);
			/* path must be released before calling iterate()! */
J
Jan Schmidt 已提交
1601 1602 1603 1604
			pr_debug("following ref at offset %u for inode %llu in "
				 "tree %llu\n", cur,
				 (unsigned long long)found_key.objectid,
				 (unsigned long long)fs_root->objectid);
M
Mark Fasheh 已提交
1605 1606
			ret = iterate(parent, name_len,
				      (unsigned long)(iref + 1), eb, ctx);
1607
			if (ret)
1608 1609 1610 1611
				break;
			len = sizeof(*iref) + name_len;
			iref = (struct btrfs_inode_ref *)((char *)iref + len);
		}
1612
		btrfs_tree_read_unlock_blocking(eb);
1613 1614 1615 1616 1617 1618 1619 1620
		free_extent_buffer(eb);
	}

	btrfs_release_path(path);

	return ret;
}

M
Mark Fasheh 已提交
1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706
static int iterate_inode_extrefs(u64 inum, struct btrfs_root *fs_root,
				 struct btrfs_path *path,
				 iterate_irefs_t *iterate, void *ctx)
{
	int ret;
	int slot;
	u64 offset = 0;
	u64 parent;
	int found = 0;
	struct extent_buffer *eb;
	struct btrfs_inode_extref *extref;
	struct extent_buffer *leaf;
	u32 item_size;
	u32 cur_offset;
	unsigned long ptr;

	while (1) {
		ret = btrfs_find_one_extref(fs_root, inum, offset, path, &extref,
					    &offset);
		if (ret < 0)
			break;
		if (ret) {
			ret = found ? 0 : -ENOENT;
			break;
		}
		++found;

		slot = path->slots[0];
		eb = path->nodes[0];
		/* make sure we can use eb after releasing the path */
		atomic_inc(&eb->refs);

		btrfs_tree_read_lock(eb);
		btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK);
		btrfs_release_path(path);

		leaf = path->nodes[0];
		item_size = btrfs_item_size_nr(leaf, path->slots[0]);
		ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
		cur_offset = 0;

		while (cur_offset < item_size) {
			u32 name_len;

			extref = (struct btrfs_inode_extref *)(ptr + cur_offset);
			parent = btrfs_inode_extref_parent(eb, extref);
			name_len = btrfs_inode_extref_name_len(eb, extref);
			ret = iterate(parent, name_len,
				      (unsigned long)&extref->name, eb, ctx);
			if (ret)
				break;

			cur_offset += btrfs_inode_extref_name_len(leaf, extref);
			cur_offset += sizeof(*extref);
		}
		btrfs_tree_read_unlock_blocking(eb);
		free_extent_buffer(eb);

		offset++;
	}

	btrfs_release_path(path);

	return ret;
}

static int iterate_irefs(u64 inum, struct btrfs_root *fs_root,
			 struct btrfs_path *path, iterate_irefs_t *iterate,
			 void *ctx)
{
	int ret;
	int found_refs = 0;

	ret = iterate_inode_refs(inum, fs_root, path, iterate, ctx);
	if (!ret)
		++found_refs;
	else if (ret != -ENOENT)
		return ret;

	ret = iterate_inode_extrefs(inum, fs_root, path, iterate, ctx);
	if (ret == -ENOENT && found_refs)
		return 0;

	return ret;
}

1707 1708 1709 1710
/*
 * returns 0 if the path could be dumped (probably truncated)
 * returns <0 in case of an error
 */
M
Mark Fasheh 已提交
1711 1712
static int inode_to_path(u64 inum, u32 name_len, unsigned long name_off,
			 struct extent_buffer *eb, void *ctx)
1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723
{
	struct inode_fs_paths *ipath = ctx;
	char *fspath;
	char *fspath_min;
	int i = ipath->fspath->elem_cnt;
	const int s_ptr = sizeof(char *);
	u32 bytes_left;

	bytes_left = ipath->fspath->bytes_left > s_ptr ?
					ipath->fspath->bytes_left - s_ptr : 0;

1724
	fspath_min = (char *)ipath->fspath->val + (i + 1) * s_ptr;
1725 1726
	fspath = btrfs_ref_to_path(ipath->fs_root, ipath->btrfs_path, name_len,
				   name_off, eb, inum, fspath_min, bytes_left);
1727 1728 1729 1730
	if (IS_ERR(fspath))
		return PTR_ERR(fspath);

	if (fspath > fspath_min) {
1731
		ipath->fspath->val[i] = (u64)(unsigned long)fspath;
1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745
		++ipath->fspath->elem_cnt;
		ipath->fspath->bytes_left = fspath - fspath_min;
	} else {
		++ipath->fspath->elem_missed;
		ipath->fspath->bytes_missing += fspath_min - fspath;
		ipath->fspath->bytes_left = 0;
	}

	return 0;
}

/*
 * this dumps all file system paths to the inode into the ipath struct, provided
 * is has been created large enough. each path is zero-terminated and accessed
1746
 * from ipath->fspath->val[i].
1747
 * when it returns, there are ipath->fspath->elem_cnt number of paths available
1748
 * in ipath->fspath->val[]. when the allocated space wasn't sufficient, the
1749 1750 1751 1752 1753 1754 1755
 * number of missed paths in recored in ipath->fspath->elem_missed, otherwise,
 * it's zero. ipath->fspath->bytes_missing holds the number of bytes that would
 * have been needed to return all paths.
 */
int paths_from_inode(u64 inum, struct inode_fs_paths *ipath)
{
	return iterate_irefs(inum, ipath->fs_root, ipath->btrfs_path,
M
Mark Fasheh 已提交
1756
			     inode_to_path, ipath);
1757 1758 1759 1760 1761 1762 1763 1764
}

struct btrfs_data_container *init_data_container(u32 total_bytes)
{
	struct btrfs_data_container *data;
	size_t alloc_bytes;

	alloc_bytes = max_t(size_t, total_bytes, sizeof(*data));
1765
	data = vmalloc(alloc_bytes);
1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813
	if (!data)
		return ERR_PTR(-ENOMEM);

	if (total_bytes >= sizeof(*data)) {
		data->bytes_left = total_bytes - sizeof(*data);
		data->bytes_missing = 0;
	} else {
		data->bytes_missing = sizeof(*data) - total_bytes;
		data->bytes_left = 0;
	}

	data->elem_cnt = 0;
	data->elem_missed = 0;

	return data;
}

/*
 * allocates space to return multiple file system paths for an inode.
 * total_bytes to allocate are passed, note that space usable for actual path
 * information will be total_bytes - sizeof(struct inode_fs_paths).
 * the returned pointer must be freed with free_ipath() in the end.
 */
struct inode_fs_paths *init_ipath(s32 total_bytes, struct btrfs_root *fs_root,
					struct btrfs_path *path)
{
	struct inode_fs_paths *ifp;
	struct btrfs_data_container *fspath;

	fspath = init_data_container(total_bytes);
	if (IS_ERR(fspath))
		return (void *)fspath;

	ifp = kmalloc(sizeof(*ifp), GFP_NOFS);
	if (!ifp) {
		kfree(fspath);
		return ERR_PTR(-ENOMEM);
	}

	ifp->btrfs_path = path;
	ifp->fspath = fspath;
	ifp->fs_root = fs_root;

	return ifp;
}

void free_ipath(struct inode_fs_paths *ipath)
{
1814 1815
	if (!ipath)
		return;
1816
	vfree(ipath->fspath);
1817 1818
	kfree(ipath);
}