1 /* 2 * Copyright (C) 2011 STRATO. All rights reserved. 3 * 4 * This program is free software; you can redistribute it and/or 5 * modify it under the terms of the GNU General Public 6 * License v2 as published by the Free Software Foundation. 7 * 8 * This program is distributed in the hope that it will be useful, 9 * but WITHOUT ANY WARRANTY; without even the implied warranty of 10 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU 11 * General Public License for more details. 12 * 13 * You should have received a copy of the GNU General Public 14 * License along with this program; if not, write to the 15 * Free Software Foundation, Inc., 59 Temple Place - Suite 330, 16 * Boston, MA 021110-1307, USA. 17 */ 18 19 #include <linux/vmalloc.h> 20 #include "ctree.h" 21 #include "disk-io.h" 22 #include "backref.h" 23 #include "ulist.h" 24 #include "transaction.h" 25 #include "delayed-ref.h" 26 #include "locking.h" 27 28 struct extent_inode_elem { 29 u64 inum; 30 u64 offset; 31 struct extent_inode_elem *next; 32 }; 33 34 static int check_extent_in_eb(struct btrfs_key *key, struct extent_buffer *eb, 35 struct btrfs_file_extent_item *fi, 36 u64 extent_item_pos, 37 struct extent_inode_elem **eie) 38 { 39 u64 data_offset; 40 u64 data_len; 41 struct extent_inode_elem *e; 42 43 data_offset = btrfs_file_extent_offset(eb, fi); 44 data_len = btrfs_file_extent_num_bytes(eb, fi); 45 46 if (extent_item_pos < data_offset || 47 extent_item_pos >= data_offset + data_len) 48 return 1; 49 50 e = kmalloc(sizeof(*e), GFP_NOFS); 51 if (!e) 52 return -ENOMEM; 53 54 e->next = *eie; 55 e->inum = key->objectid; 56 e->offset = key->offset + (extent_item_pos - data_offset); 57 *eie = e; 58 59 return 0; 60 } 61 62 static int find_extent_in_eb(struct extent_buffer *eb, u64 wanted_disk_byte, 63 u64 extent_item_pos, 64 struct extent_inode_elem **eie) 65 { 66 u64 disk_byte; 67 struct btrfs_key key; 68 struct btrfs_file_extent_item *fi; 69 int slot; 70 int nritems; 71 int extent_type; 72 int ret; 73 74 /* 75 * from the shared data ref, we only have the leaf but we need 76 * the key. thus, we must look into all items and see that we 77 * find one (some) with a reference to our extent item. 78 */ 79 nritems = btrfs_header_nritems(eb); 80 for (slot = 0; slot < nritems; ++slot) { 81 btrfs_item_key_to_cpu(eb, &key, slot); 82 if (key.type != BTRFS_EXTENT_DATA_KEY) 83 continue; 84 fi = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item); 85 extent_type = btrfs_file_extent_type(eb, fi); 86 if (extent_type == BTRFS_FILE_EXTENT_INLINE) 87 continue; 88 /* don't skip BTRFS_FILE_EXTENT_PREALLOC, we can handle that */ 89 disk_byte = btrfs_file_extent_disk_bytenr(eb, fi); 90 if (disk_byte != wanted_disk_byte) 91 continue; 92 93 ret = check_extent_in_eb(&key, eb, fi, extent_item_pos, eie); 94 if (ret < 0) 95 return ret; 96 } 97 98 return 0; 99 } 100 101 /* 102 * this structure records all encountered refs on the way up to the root 103 */ 104 struct __prelim_ref { 105 struct list_head list; 106 u64 root_id; 107 struct btrfs_key key_for_search; 108 int level; 109 int count; 110 struct extent_inode_elem *inode_list; 111 u64 parent; 112 u64 wanted_disk_byte; 113 }; 114 115 /* 116 * the rules for all callers of this function are: 117 * - obtaining the parent is the goal 118 * - if you add a key, you must know that it is a correct key 119 * - if you cannot add the parent or a correct key, then we will look into the 120 * block later to set a correct key 121 * 122 * delayed refs 123 * ============ 124 * backref type | shared | indirect | shared | indirect 125 * information | tree | tree | data | data 126 * --------------------+--------+----------+--------+---------- 127 * parent logical | y | - | - | - 128 * key to resolve | - | y | y | y 129 * tree block logical | - | - | - | - 130 * root for resolving | y | y | y | y 131 * 132 * - column 1: we've the parent -> done 133 * - column 2, 3, 4: we use the key to find the parent 134 * 135 * on disk refs (inline or keyed) 136 * ============================== 137 * backref type | shared | indirect | shared | indirect 138 * information | tree | tree | data | data 139 * --------------------+--------+----------+--------+---------- 140 * parent logical | y | - | y | - 141 * key to resolve | - | - | - | y 142 * tree block logical | y | y | y | y 143 * root for resolving | - | y | y | y 144 * 145 * - column 1, 3: we've the parent -> done 146 * - column 2: we take the first key from the block to find the parent 147 * (see __add_missing_keys) 148 * - column 4: we use the key to find the parent 149 * 150 * additional information that's available but not required to find the parent 151 * block might help in merging entries to gain some speed. 152 */ 153 154 static int __add_prelim_ref(struct list_head *head, u64 root_id, 155 struct btrfs_key *key, int level, 156 u64 parent, u64 wanted_disk_byte, int count) 157 { 158 struct __prelim_ref *ref; 159 160 /* in case we're adding delayed refs, we're holding the refs spinlock */ 161 ref = kmalloc(sizeof(*ref), GFP_ATOMIC); 162 if (!ref) 163 return -ENOMEM; 164 165 ref->root_id = root_id; 166 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 ref->level = level; 173 ref->count = count; 174 ref->parent = parent; 175 ref->wanted_disk_byte = wanted_disk_byte; 176 list_add_tail(&ref->list, head); 177 178 return 0; 179 } 180 181 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, 184 u64 wanted_disk_byte, 185 const u64 *extent_item_pos) 186 { 187 int ret = 0; 188 int slot; 189 struct extent_buffer *eb; 190 struct btrfs_key key; 191 struct btrfs_file_extent_item *fi; 192 struct extent_inode_elem *eie = NULL; 193 u64 disk_byte; 194 195 if (level != 0) { 196 eb = path->nodes[level]; 197 ret = ulist_add(parents, eb->start, 0, GFP_NOFS); 198 if (ret < 0) 199 return ret; 200 return 0; 201 } 202 203 /* 204 * We normally enter this function with the path already pointing to 205 * the first item to check. But sometimes, we may enter it with 206 * 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])) 209 ret = btrfs_next_old_leaf(root, path, time_seq); 210 211 while (!ret) { 212 eb = path->nodes[0]; 213 slot = path->slots[0]; 214 215 btrfs_item_key_to_cpu(eb, &key, slot); 216 217 if (key.objectid != key_for_search->objectid || 218 key.type != BTRFS_EXTENT_DATA_KEY) 219 break; 220 221 fi = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item); 222 disk_byte = btrfs_file_extent_disk_bytenr(eb, fi); 223 224 if (disk_byte == wanted_disk_byte) { 225 eie = NULL; 226 if (extent_item_pos) { 227 ret = check_extent_in_eb(&key, eb, fi, 228 *extent_item_pos, 229 &eie); 230 if (ret < 0) 231 break; 232 } 233 if (!ret) { 234 ret = ulist_add(parents, eb->start, 235 (uintptr_t)eie, GFP_NOFS); 236 if (ret < 0) 237 break; 238 if (!extent_item_pos) { 239 ret = btrfs_next_old_leaf(root, path, 240 time_seq); 241 continue; 242 } 243 } 244 } 245 ret = btrfs_next_old_item(root, path, time_seq); 246 } 247 248 if (ret > 0) 249 ret = 0; 250 return ret; 251 } 252 253 /* 254 * resolve an indirect backref in the form (root_id, key, level) 255 * to a logical address 256 */ 257 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 struct ulist *parents, 262 const u64 *extent_item_pos) 263 { 264 struct btrfs_path *path; 265 struct btrfs_root *root; 266 struct btrfs_key root_key; 267 struct extent_buffer *eb; 268 int ret = 0; 269 int root_level; 270 int level = ref->level; 271 272 path = btrfs_alloc_path(); 273 if (!path) 274 return -ENOMEM; 275 path->search_commit_root = !!search_commit_root; 276 277 root_key.objectid = ref->root_id; 278 root_key.type = BTRFS_ROOT_ITEM_KEY; 279 root_key.offset = (u64)-1; 280 root = btrfs_read_fs_root_no_name(fs_info, &root_key); 281 if (IS_ERR(root)) { 282 ret = PTR_ERR(root); 283 goto out; 284 } 285 286 root_level = btrfs_old_root_level(root, time_seq); 287 288 if (root_level + 1 == level) 289 goto out; 290 291 path->lowest_level = level; 292 ret = btrfs_search_old_slot(root, &ref->key_for_search, path, time_seq); 293 pr_debug("search slot in root %llu (level %d, ref count %d) returned " 294 "%d for key (%llu %u %llu)\n", 295 (unsigned long long)ref->root_id, level, ref->count, ret, 296 (unsigned long long)ref->key_for_search.objectid, 297 ref->key_for_search.type, 298 (unsigned long long)ref->key_for_search.offset); 299 if (ret < 0) 300 goto out; 301 302 eb = path->nodes[level]; 303 while (!eb) { 304 if (!level) { 305 WARN_ON(1); 306 ret = 1; 307 goto out; 308 } 309 level--; 310 eb = path->nodes[level]; 311 } 312 313 ret = add_all_parents(root, path, parents, level, &ref->key_for_search, 314 time_seq, ref->wanted_disk_byte, 315 extent_item_pos); 316 out: 317 btrfs_free_path(path); 318 return ret; 319 } 320 321 /* 322 * resolve all indirect backrefs from the list 323 */ 324 static int __resolve_indirect_refs(struct btrfs_fs_info *fs_info, 325 int search_commit_root, u64 time_seq, 326 struct list_head *head, 327 const u64 *extent_item_pos) 328 { 329 int err; 330 int ret = 0; 331 struct __prelim_ref *ref; 332 struct __prelim_ref *ref_safe; 333 struct __prelim_ref *new_ref; 334 struct ulist *parents; 335 struct ulist_node *node; 336 struct ulist_iterator uiter; 337 338 parents = ulist_alloc(GFP_NOFS); 339 if (!parents) 340 return -ENOMEM; 341 342 /* 343 * _safe allows us to insert directly after the current item without 344 * iterating over the newly inserted items. 345 * we're also allowed to re-assign ref during iteration. 346 */ 347 list_for_each_entry_safe(ref, ref_safe, head, list) { 348 if (ref->parent) /* already direct */ 349 continue; 350 if (ref->count == 0) 351 continue; 352 err = __resolve_indirect_ref(fs_info, search_commit_root, 353 time_seq, ref, parents, 354 extent_item_pos); 355 if (err == -ENOMEM) 356 goto out; 357 if (err) 358 continue; 359 360 /* we put the first parent into the ref at hand */ 361 ULIST_ITER_INIT(&uiter); 362 node = ulist_next(parents, &uiter); 363 ref->parent = node ? node->val : 0; 364 ref->inode_list = node ? 365 (struct extent_inode_elem *)(uintptr_t)node->aux : 0; 366 367 /* additional parents require new refs being added here */ 368 while ((node = ulist_next(parents, &uiter))) { 369 new_ref = kmalloc(sizeof(*new_ref), GFP_NOFS); 370 if (!new_ref) { 371 ret = -ENOMEM; 372 goto out; 373 } 374 memcpy(new_ref, ref, sizeof(*ref)); 375 new_ref->parent = node->val; 376 new_ref->inode_list = (struct extent_inode_elem *) 377 (uintptr_t)node->aux; 378 list_add(&new_ref->list, &ref->list); 379 } 380 ulist_reinit(parents); 381 } 382 out: 383 ulist_free(parents); 384 return ret; 385 } 386 387 static inline int ref_for_same_block(struct __prelim_ref *ref1, 388 struct __prelim_ref *ref2) 389 { 390 if (ref1->level != ref2->level) 391 return 0; 392 if (ref1->root_id != ref2->root_id) 393 return 0; 394 if (ref1->key_for_search.type != ref2->key_for_search.type) 395 return 0; 396 if (ref1->key_for_search.objectid != ref2->key_for_search.objectid) 397 return 0; 398 if (ref1->key_for_search.offset != ref2->key_for_search.offset) 399 return 0; 400 if (ref1->parent != ref2->parent) 401 return 0; 402 403 return 1; 404 } 405 406 /* 407 * read tree blocks and add keys where required. 408 */ 409 static int __add_missing_keys(struct btrfs_fs_info *fs_info, 410 struct list_head *head) 411 { 412 struct list_head *pos; 413 struct extent_buffer *eb; 414 415 list_for_each(pos, head) { 416 struct __prelim_ref *ref; 417 ref = list_entry(pos, struct __prelim_ref, list); 418 419 if (ref->parent) 420 continue; 421 if (ref->key_for_search.type) 422 continue; 423 BUG_ON(!ref->wanted_disk_byte); 424 eb = read_tree_block(fs_info->tree_root, ref->wanted_disk_byte, 425 fs_info->tree_root->leafsize, 0); 426 if (!eb || !extent_buffer_uptodate(eb)) { 427 free_extent_buffer(eb); 428 return -EIO; 429 } 430 btrfs_tree_read_lock(eb); 431 if (btrfs_header_level(eb) == 0) 432 btrfs_item_key_to_cpu(eb, &ref->key_for_search, 0); 433 else 434 btrfs_node_key_to_cpu(eb, &ref->key_for_search, 0); 435 btrfs_tree_read_unlock(eb); 436 free_extent_buffer(eb); 437 } 438 return 0; 439 } 440 441 /* 442 * merge two lists of backrefs and adjust counts accordingly 443 * 444 * mode = 1: merge identical keys, if key is set 445 * FIXME: if we add more keys in __add_prelim_ref, we can merge more here. 446 * additionally, we could even add a key range for the blocks we 447 * looked into to merge even more (-> replace unresolved refs by those 448 * having a parent). 449 * mode = 2: merge identical parents 450 */ 451 static void __merge_refs(struct list_head *head, int mode) 452 { 453 struct list_head *pos1; 454 455 list_for_each(pos1, head) { 456 struct list_head *n2; 457 struct list_head *pos2; 458 struct __prelim_ref *ref1; 459 460 ref1 = list_entry(pos1, struct __prelim_ref, list); 461 462 for (pos2 = pos1->next, n2 = pos2->next; pos2 != head; 463 pos2 = n2, n2 = pos2->next) { 464 struct __prelim_ref *ref2; 465 struct __prelim_ref *xchg; 466 struct extent_inode_elem *eie; 467 468 ref2 = list_entry(pos2, struct __prelim_ref, list); 469 470 if (mode == 1) { 471 if (!ref_for_same_block(ref1, ref2)) 472 continue; 473 if (!ref1->parent && ref2->parent) { 474 xchg = ref1; 475 ref1 = ref2; 476 ref2 = xchg; 477 } 478 } else { 479 if (ref1->parent != ref2->parent) 480 continue; 481 } 482 483 eie = ref1->inode_list; 484 while (eie && eie->next) 485 eie = eie->next; 486 if (eie) 487 eie->next = ref2->inode_list; 488 else 489 ref1->inode_list = ref2->inode_list; 490 ref1->count += ref2->count; 491 492 list_del(&ref2->list); 493 kfree(ref2); 494 } 495 496 } 497 } 498 499 /* 500 * add all currently queued delayed refs from this head whose seq nr is 501 * smaller or equal that seq to the list 502 */ 503 static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq, 504 struct list_head *prefs) 505 { 506 struct btrfs_delayed_extent_op *extent_op = head->extent_op; 507 struct rb_node *n = &head->node.rb_node; 508 struct btrfs_key key; 509 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 while ((n = rb_prev(n))) { 517 struct btrfs_delayed_ref_node *node; 518 node = rb_entry(n, struct btrfs_delayed_ref_node, 519 rb_node); 520 if (node->bytenr != head->node.bytenr) 521 break; 522 WARN_ON(node->is_head); 523 524 if (node->seq > seq) 525 continue; 526 527 switch (node->action) { 528 case BTRFS_ADD_DELAYED_EXTENT: 529 case BTRFS_UPDATE_DELAYED_HEAD: 530 WARN_ON(1); 531 continue; 532 case BTRFS_ADD_DELAYED_REF: 533 sgn = 1; 534 break; 535 case BTRFS_DROP_DELAYED_REF: 536 sgn = -1; 537 break; 538 default: 539 BUG_ON(1); 540 } 541 switch (node->type) { 542 case BTRFS_TREE_BLOCK_REF_KEY: { 543 struct btrfs_delayed_tree_ref *ref; 544 545 ref = btrfs_delayed_node_to_tree_ref(node); 546 ret = __add_prelim_ref(prefs, ref->root, &op_key, 547 ref->level + 1, 0, node->bytenr, 548 node->ref_mod * sgn); 549 break; 550 } 551 case BTRFS_SHARED_BLOCK_REF_KEY: { 552 struct btrfs_delayed_tree_ref *ref; 553 554 ref = btrfs_delayed_node_to_tree_ref(node); 555 ret = __add_prelim_ref(prefs, ref->root, NULL, 556 ref->level + 1, ref->parent, 557 node->bytenr, 558 node->ref_mod * sgn); 559 break; 560 } 561 case BTRFS_EXTENT_DATA_REF_KEY: { 562 struct btrfs_delayed_data_ref *ref; 563 ref = btrfs_delayed_node_to_data_ref(node); 564 565 key.objectid = ref->objectid; 566 key.type = BTRFS_EXTENT_DATA_KEY; 567 key.offset = ref->offset; 568 ret = __add_prelim_ref(prefs, ref->root, &key, 0, 0, 569 node->bytenr, 570 node->ref_mod * sgn); 571 break; 572 } 573 case BTRFS_SHARED_DATA_REF_KEY: { 574 struct btrfs_delayed_data_ref *ref; 575 576 ref = btrfs_delayed_node_to_data_ref(node); 577 578 key.objectid = ref->objectid; 579 key.type = BTRFS_EXTENT_DATA_KEY; 580 key.offset = ref->offset; 581 ret = __add_prelim_ref(prefs, ref->root, &key, 0, 582 ref->parent, node->bytenr, 583 node->ref_mod * sgn); 584 break; 585 } 586 default: 587 WARN_ON(1); 588 } 589 if (ret) 590 return ret; 591 } 592 593 return 0; 594 } 595 596 /* 597 * add all inline backrefs for bytenr to the list 598 */ 599 static int __add_inline_refs(struct btrfs_fs_info *fs_info, 600 struct btrfs_path *path, u64 bytenr, 601 int *info_level, struct list_head *prefs) 602 { 603 int ret = 0; 604 int slot; 605 struct extent_buffer *leaf; 606 struct btrfs_key key; 607 unsigned long ptr; 608 unsigned long end; 609 struct btrfs_extent_item *ei; 610 u64 flags; 611 u64 item_size; 612 613 /* 614 * enumerate all inline refs 615 */ 616 leaf = path->nodes[0]; 617 slot = path->slots[0]; 618 619 item_size = btrfs_item_size_nr(leaf, slot); 620 BUG_ON(item_size < sizeof(*ei)); 621 622 ei = btrfs_item_ptr(leaf, slot, struct btrfs_extent_item); 623 flags = btrfs_extent_flags(leaf, ei); 624 625 ptr = (unsigned long)(ei + 1); 626 end = (unsigned long)ei + item_size; 627 628 if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) { 629 struct btrfs_tree_block_info *info; 630 631 info = (struct btrfs_tree_block_info *)ptr; 632 *info_level = btrfs_tree_block_level(leaf, info); 633 ptr += sizeof(struct btrfs_tree_block_info); 634 BUG_ON(ptr > end); 635 } else { 636 BUG_ON(!(flags & BTRFS_EXTENT_FLAG_DATA)); 637 } 638 639 while (ptr < end) { 640 struct btrfs_extent_inline_ref *iref; 641 u64 offset; 642 int type; 643 644 iref = (struct btrfs_extent_inline_ref *)ptr; 645 type = btrfs_extent_inline_ref_type(leaf, iref); 646 offset = btrfs_extent_inline_ref_offset(leaf, iref); 647 648 switch (type) { 649 case BTRFS_SHARED_BLOCK_REF_KEY: 650 ret = __add_prelim_ref(prefs, 0, NULL, 651 *info_level + 1, offset, 652 bytenr, 1); 653 break; 654 case BTRFS_SHARED_DATA_REF_KEY: { 655 struct btrfs_shared_data_ref *sdref; 656 int count; 657 658 sdref = (struct btrfs_shared_data_ref *)(iref + 1); 659 count = btrfs_shared_data_ref_count(leaf, sdref); 660 ret = __add_prelim_ref(prefs, 0, NULL, 0, offset, 661 bytenr, count); 662 break; 663 } 664 case BTRFS_TREE_BLOCK_REF_KEY: 665 ret = __add_prelim_ref(prefs, offset, NULL, 666 *info_level + 1, 0, 667 bytenr, 1); 668 break; 669 case BTRFS_EXTENT_DATA_REF_KEY: { 670 struct btrfs_extent_data_ref *dref; 671 int count; 672 u64 root; 673 674 dref = (struct btrfs_extent_data_ref *)(&iref->offset); 675 count = btrfs_extent_data_ref_count(leaf, dref); 676 key.objectid = btrfs_extent_data_ref_objectid(leaf, 677 dref); 678 key.type = BTRFS_EXTENT_DATA_KEY; 679 key.offset = btrfs_extent_data_ref_offset(leaf, dref); 680 root = btrfs_extent_data_ref_root(leaf, dref); 681 ret = __add_prelim_ref(prefs, root, &key, 0, 0, 682 bytenr, count); 683 break; 684 } 685 default: 686 WARN_ON(1); 687 } 688 if (ret) 689 return ret; 690 ptr += btrfs_extent_inline_ref_size(type); 691 } 692 693 return 0; 694 } 695 696 /* 697 * add all non-inline backrefs for bytenr to the list 698 */ 699 static int __add_keyed_refs(struct btrfs_fs_info *fs_info, 700 struct btrfs_path *path, u64 bytenr, 701 int info_level, struct list_head *prefs) 702 { 703 struct btrfs_root *extent_root = fs_info->extent_root; 704 int ret; 705 int slot; 706 struct extent_buffer *leaf; 707 struct btrfs_key key; 708 709 while (1) { 710 ret = btrfs_next_item(extent_root, path); 711 if (ret < 0) 712 break; 713 if (ret) { 714 ret = 0; 715 break; 716 } 717 718 slot = path->slots[0]; 719 leaf = path->nodes[0]; 720 btrfs_item_key_to_cpu(leaf, &key, slot); 721 722 if (key.objectid != bytenr) 723 break; 724 if (key.type < BTRFS_TREE_BLOCK_REF_KEY) 725 continue; 726 if (key.type > BTRFS_SHARED_DATA_REF_KEY) 727 break; 728 729 switch (key.type) { 730 case BTRFS_SHARED_BLOCK_REF_KEY: 731 ret = __add_prelim_ref(prefs, 0, NULL, 732 info_level + 1, key.offset, 733 bytenr, 1); 734 break; 735 case BTRFS_SHARED_DATA_REF_KEY: { 736 struct btrfs_shared_data_ref *sdref; 737 int count; 738 739 sdref = btrfs_item_ptr(leaf, slot, 740 struct btrfs_shared_data_ref); 741 count = btrfs_shared_data_ref_count(leaf, sdref); 742 ret = __add_prelim_ref(prefs, 0, NULL, 0, key.offset, 743 bytenr, count); 744 break; 745 } 746 case BTRFS_TREE_BLOCK_REF_KEY: 747 ret = __add_prelim_ref(prefs, key.offset, NULL, 748 info_level + 1, 0, 749 bytenr, 1); 750 break; 751 case BTRFS_EXTENT_DATA_REF_KEY: { 752 struct btrfs_extent_data_ref *dref; 753 int count; 754 u64 root; 755 756 dref = btrfs_item_ptr(leaf, slot, 757 struct btrfs_extent_data_ref); 758 count = btrfs_extent_data_ref_count(leaf, dref); 759 key.objectid = btrfs_extent_data_ref_objectid(leaf, 760 dref); 761 key.type = BTRFS_EXTENT_DATA_KEY; 762 key.offset = btrfs_extent_data_ref_offset(leaf, dref); 763 root = btrfs_extent_data_ref_root(leaf, dref); 764 ret = __add_prelim_ref(prefs, root, &key, 0, 0, 765 bytenr, count); 766 break; 767 } 768 default: 769 WARN_ON(1); 770 } 771 if (ret) 772 return ret; 773 774 } 775 776 return ret; 777 } 778 779 /* 780 * this adds all existing backrefs (inline backrefs, backrefs and delayed 781 * refs) for the given bytenr to the refs list, merges duplicates and resolves 782 * indirect refs to their parent bytenr. 783 * When roots are found, they're added to the roots list 784 * 785 * FIXME some caching might speed things up 786 */ 787 static int find_parent_nodes(struct btrfs_trans_handle *trans, 788 struct btrfs_fs_info *fs_info, u64 bytenr, 789 u64 time_seq, struct ulist *refs, 790 struct ulist *roots, const u64 *extent_item_pos) 791 { 792 struct btrfs_key key; 793 struct btrfs_path *path; 794 struct btrfs_delayed_ref_root *delayed_refs = NULL; 795 struct btrfs_delayed_ref_head *head; 796 int info_level = 0; 797 int ret; 798 int search_commit_root = (trans == BTRFS_BACKREF_SEARCH_COMMIT_ROOT); 799 struct list_head prefs_delayed; 800 struct list_head prefs; 801 struct __prelim_ref *ref; 802 803 INIT_LIST_HEAD(&prefs); 804 INIT_LIST_HEAD(&prefs_delayed); 805 806 key.objectid = bytenr; 807 key.type = BTRFS_EXTENT_ITEM_KEY; 808 key.offset = (u64)-1; 809 810 path = btrfs_alloc_path(); 811 if (!path) 812 return -ENOMEM; 813 path->search_commit_root = !!search_commit_root; 814 815 /* 816 * grab both a lock on the path and a lock on the delayed ref head. 817 * We need both to get a consistent picture of how the refs look 818 * at a specified point in time 819 */ 820 again: 821 head = NULL; 822 823 ret = btrfs_search_slot(trans, fs_info->extent_root, &key, path, 0, 0); 824 if (ret < 0) 825 goto out; 826 BUG_ON(ret == 0); 827 828 if (trans != BTRFS_BACKREF_SEARCH_COMMIT_ROOT) { 829 /* 830 * look if there are updates for this ref queued and lock the 831 * head 832 */ 833 delayed_refs = &trans->transaction->delayed_refs; 834 spin_lock(&delayed_refs->lock); 835 head = btrfs_find_delayed_ref_head(trans, bytenr); 836 if (head) { 837 if (!mutex_trylock(&head->mutex)) { 838 atomic_inc(&head->node.refs); 839 spin_unlock(&delayed_refs->lock); 840 841 btrfs_release_path(path); 842 843 /* 844 * Mutex was contended, block until it's 845 * released and try again 846 */ 847 mutex_lock(&head->mutex); 848 mutex_unlock(&head->mutex); 849 btrfs_put_delayed_ref(&head->node); 850 goto again; 851 } 852 ret = __add_delayed_refs(head, time_seq, 853 &prefs_delayed); 854 mutex_unlock(&head->mutex); 855 if (ret) { 856 spin_unlock(&delayed_refs->lock); 857 goto out; 858 } 859 } 860 spin_unlock(&delayed_refs->lock); 861 } 862 863 if (path->slots[0]) { 864 struct extent_buffer *leaf; 865 int slot; 866 867 path->slots[0]--; 868 leaf = path->nodes[0]; 869 slot = path->slots[0]; 870 btrfs_item_key_to_cpu(leaf, &key, slot); 871 if (key.objectid == bytenr && 872 key.type == BTRFS_EXTENT_ITEM_KEY) { 873 ret = __add_inline_refs(fs_info, path, bytenr, 874 &info_level, &prefs); 875 if (ret) 876 goto out; 877 ret = __add_keyed_refs(fs_info, path, bytenr, 878 info_level, &prefs); 879 if (ret) 880 goto out; 881 } 882 } 883 btrfs_release_path(path); 884 885 list_splice_init(&prefs_delayed, &prefs); 886 887 ret = __add_missing_keys(fs_info, &prefs); 888 if (ret) 889 goto out; 890 891 __merge_refs(&prefs, 1); 892 893 ret = __resolve_indirect_refs(fs_info, search_commit_root, time_seq, 894 &prefs, extent_item_pos); 895 if (ret) 896 goto out; 897 898 __merge_refs(&prefs, 2); 899 900 while (!list_empty(&prefs)) { 901 ref = list_first_entry(&prefs, struct __prelim_ref, list); 902 list_del(&ref->list); 903 WARN_ON(ref->count < 0); 904 if (ref->count && ref->root_id && ref->parent == 0) { 905 /* no parent == root of tree */ 906 ret = ulist_add(roots, ref->root_id, 0, GFP_NOFS); 907 if (ret < 0) 908 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 u32 bsz; 914 struct extent_buffer *eb; 915 bsz = btrfs_level_size(fs_info->extent_root, 916 info_level); 917 eb = read_tree_block(fs_info->extent_root, 918 ref->parent, bsz, 0); 919 if (!eb || !extent_buffer_uptodate(eb)) { 920 free_extent_buffer(eb); 921 ret = -EIO; 922 goto out; 923 } 924 ret = find_extent_in_eb(eb, bytenr, 925 *extent_item_pos, &eie); 926 ref->inode_list = eie; 927 free_extent_buffer(eb); 928 } 929 ret = ulist_add_merge(refs, ref->parent, 930 (uintptr_t)ref->inode_list, 931 (u64 *)&eie, GFP_NOFS); 932 if (ret < 0) 933 goto out; 934 if (!ret && extent_item_pos) { 935 /* 936 * we've recorded that parent, so we must extend 937 * its inode list here 938 */ 939 BUG_ON(!eie); 940 while (eie->next) 941 eie = eie->next; 942 eie->next = ref->inode_list; 943 } 944 } 945 kfree(ref); 946 } 947 948 out: 949 btrfs_free_path(path); 950 while (!list_empty(&prefs)) { 951 ref = list_first_entry(&prefs, struct __prelim_ref, list); 952 list_del(&ref->list); 953 kfree(ref); 954 } 955 while (!list_empty(&prefs_delayed)) { 956 ref = list_first_entry(&prefs_delayed, struct __prelim_ref, 957 list); 958 list_del(&ref->list); 959 kfree(ref); 960 } 961 962 return ret; 963 } 964 965 static void free_leaf_list(struct ulist *blocks) 966 { 967 struct ulist_node *node = NULL; 968 struct extent_inode_elem *eie; 969 struct extent_inode_elem *eie_next; 970 struct ulist_iterator uiter; 971 972 ULIST_ITER_INIT(&uiter); 973 while ((node = ulist_next(blocks, &uiter))) { 974 if (!node->aux) 975 continue; 976 eie = (struct extent_inode_elem *)(uintptr_t)node->aux; 977 for (; eie; eie = eie_next) { 978 eie_next = eie->next; 979 kfree(eie); 980 } 981 node->aux = 0; 982 } 983 984 ulist_free(blocks); 985 } 986 987 /* 988 * Finds all leafs with a reference to the specified combination of bytenr and 989 * offset. key_list_head will point to a list of corresponding keys (caller must 990 * free each list element). The leafs will be stored in the leafs ulist, which 991 * must be freed with ulist_free. 992 * 993 * returns 0 on success, <0 on error 994 */ 995 static int btrfs_find_all_leafs(struct btrfs_trans_handle *trans, 996 struct btrfs_fs_info *fs_info, u64 bytenr, 997 u64 time_seq, struct ulist **leafs, 998 const u64 *extent_item_pos) 999 { 1000 struct ulist *tmp; 1001 int ret; 1002 1003 tmp = ulist_alloc(GFP_NOFS); 1004 if (!tmp) 1005 return -ENOMEM; 1006 *leafs = ulist_alloc(GFP_NOFS); 1007 if (!*leafs) { 1008 ulist_free(tmp); 1009 return -ENOMEM; 1010 } 1011 1012 ret = find_parent_nodes(trans, fs_info, bytenr, 1013 time_seq, *leafs, tmp, extent_item_pos); 1014 ulist_free(tmp); 1015 1016 if (ret < 0 && ret != -ENOENT) { 1017 free_leaf_list(*leafs); 1018 return ret; 1019 } 1020 1021 return 0; 1022 } 1023 1024 /* 1025 * walk all backrefs for a given extent to find all roots that reference this 1026 * extent. Walking a backref means finding all extents that reference this 1027 * extent and in turn walk the backrefs of those, too. Naturally this is a 1028 * recursive process, but here it is implemented in an iterative fashion: We 1029 * find all referencing extents for the extent in question and put them on a 1030 * list. In turn, we find all referencing extents for those, further appending 1031 * to the list. The way we iterate the list allows adding more elements after 1032 * the current while iterating. The process stops when we reach the end of the 1033 * list. Found roots are added to the roots list. 1034 * 1035 * returns 0 on success, < 0 on error. 1036 */ 1037 int btrfs_find_all_roots(struct btrfs_trans_handle *trans, 1038 struct btrfs_fs_info *fs_info, u64 bytenr, 1039 u64 time_seq, struct ulist **roots) 1040 { 1041 struct ulist *tmp; 1042 struct ulist_node *node = NULL; 1043 struct ulist_iterator uiter; 1044 int ret; 1045 1046 tmp = ulist_alloc(GFP_NOFS); 1047 if (!tmp) 1048 return -ENOMEM; 1049 *roots = ulist_alloc(GFP_NOFS); 1050 if (!*roots) { 1051 ulist_free(tmp); 1052 return -ENOMEM; 1053 } 1054 1055 ULIST_ITER_INIT(&uiter); 1056 while (1) { 1057 ret = find_parent_nodes(trans, fs_info, bytenr, 1058 time_seq, tmp, *roots, NULL); 1059 if (ret < 0 && ret != -ENOENT) { 1060 ulist_free(tmp); 1061 ulist_free(*roots); 1062 return ret; 1063 } 1064 node = ulist_next(tmp, &uiter); 1065 if (!node) 1066 break; 1067 bytenr = node->val; 1068 } 1069 1070 ulist_free(tmp); 1071 return 0; 1072 } 1073 1074 1075 static int __inode_info(u64 inum, u64 ioff, u8 key_type, 1076 struct btrfs_root *fs_root, struct btrfs_path *path, 1077 struct btrfs_key *found_key) 1078 { 1079 int ret; 1080 struct btrfs_key key; 1081 struct extent_buffer *eb; 1082 1083 key.type = key_type; 1084 key.objectid = inum; 1085 key.offset = ioff; 1086 1087 ret = btrfs_search_slot(NULL, fs_root, &key, path, 0, 0); 1088 if (ret < 0) 1089 return ret; 1090 1091 eb = path->nodes[0]; 1092 if (ret && path->slots[0] >= btrfs_header_nritems(eb)) { 1093 ret = btrfs_next_leaf(fs_root, path); 1094 if (ret) 1095 return ret; 1096 eb = path->nodes[0]; 1097 } 1098 1099 btrfs_item_key_to_cpu(eb, found_key, path->slots[0]); 1100 if (found_key->type != key.type || found_key->objectid != key.objectid) 1101 return 1; 1102 1103 return 0; 1104 } 1105 1106 /* 1107 * this makes the path point to (inum INODE_ITEM ioff) 1108 */ 1109 int inode_item_info(u64 inum, u64 ioff, struct btrfs_root *fs_root, 1110 struct btrfs_path *path) 1111 { 1112 struct btrfs_key key; 1113 return __inode_info(inum, ioff, BTRFS_INODE_ITEM_KEY, fs_root, path, 1114 &key); 1115 } 1116 1117 static int inode_ref_info(u64 inum, u64 ioff, struct btrfs_root *fs_root, 1118 struct btrfs_path *path, 1119 struct btrfs_key *found_key) 1120 { 1121 return __inode_info(inum, ioff, BTRFS_INODE_REF_KEY, fs_root, path, 1122 found_key); 1123 } 1124 1125 int btrfs_find_one_extref(struct btrfs_root *root, u64 inode_objectid, 1126 u64 start_off, struct btrfs_path *path, 1127 struct btrfs_inode_extref **ret_extref, 1128 u64 *found_off) 1129 { 1130 int ret, slot; 1131 struct btrfs_key key; 1132 struct btrfs_key found_key; 1133 struct btrfs_inode_extref *extref; 1134 struct extent_buffer *leaf; 1135 unsigned long ptr; 1136 1137 key.objectid = inode_objectid; 1138 btrfs_set_key_type(&key, BTRFS_INODE_EXTREF_KEY); 1139 key.offset = start_off; 1140 1141 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); 1142 if (ret < 0) 1143 return ret; 1144 1145 while (1) { 1146 leaf = path->nodes[0]; 1147 slot = path->slots[0]; 1148 if (slot >= btrfs_header_nritems(leaf)) { 1149 /* 1150 * If the item at offset is not found, 1151 * btrfs_search_slot will point us to the slot 1152 * where it should be inserted. In our case 1153 * that will be the slot directly before the 1154 * next INODE_REF_KEY_V2 item. In the case 1155 * that we're pointing to the last slot in a 1156 * leaf, we must move one leaf over. 1157 */ 1158 ret = btrfs_next_leaf(root, path); 1159 if (ret) { 1160 if (ret >= 1) 1161 ret = -ENOENT; 1162 break; 1163 } 1164 continue; 1165 } 1166 1167 btrfs_item_key_to_cpu(leaf, &found_key, slot); 1168 1169 /* 1170 * Check that we're still looking at an extended ref key for 1171 * this particular objectid. If we have different 1172 * objectid or type then there are no more to be found 1173 * in the tree and we can exit. 1174 */ 1175 ret = -ENOENT; 1176 if (found_key.objectid != inode_objectid) 1177 break; 1178 if (btrfs_key_type(&found_key) != BTRFS_INODE_EXTREF_KEY) 1179 break; 1180 1181 ret = 0; 1182 ptr = btrfs_item_ptr_offset(leaf, path->slots[0]); 1183 extref = (struct btrfs_inode_extref *)ptr; 1184 *ret_extref = extref; 1185 if (found_off) 1186 *found_off = found_key.offset; 1187 break; 1188 } 1189 1190 return ret; 1191 } 1192 1193 /* 1194 * this iterates to turn a name (from iref/extref) into a full filesystem path. 1195 * Elements of the path are separated by '/' and the path is guaranteed to be 1196 * 0-terminated. the path is only given within the current file system. 1197 * Therefore, it never starts with a '/'. the caller is responsible to provide 1198 * "size" bytes in "dest". the dest buffer will be filled backwards. finally, 1199 * the start point of the resulting string is returned. this pointer is within 1200 * dest, normally. 1201 * in case the path buffer would overflow, the pointer is decremented further 1202 * as if output was written to the buffer, though no more output is actually 1203 * generated. that way, the caller can determine how much space would be 1204 * required for the path to fit into the buffer. in that case, the returned 1205 * value will be smaller than dest. callers must check this! 1206 */ 1207 char *btrfs_ref_to_path(struct btrfs_root *fs_root, struct btrfs_path *path, 1208 u32 name_len, unsigned long name_off, 1209 struct extent_buffer *eb_in, u64 parent, 1210 char *dest, u32 size) 1211 { 1212 int slot; 1213 u64 next_inum; 1214 int ret; 1215 s64 bytes_left = ((s64)size) - 1; 1216 struct extent_buffer *eb = eb_in; 1217 struct btrfs_key found_key; 1218 int leave_spinning = path->leave_spinning; 1219 struct btrfs_inode_ref *iref; 1220 1221 if (bytes_left >= 0) 1222 dest[bytes_left] = '\0'; 1223 1224 path->leave_spinning = 1; 1225 while (1) { 1226 bytes_left -= name_len; 1227 if (bytes_left >= 0) 1228 read_extent_buffer(eb, dest + bytes_left, 1229 name_off, name_len); 1230 if (eb != eb_in) { 1231 btrfs_tree_read_unlock_blocking(eb); 1232 free_extent_buffer(eb); 1233 } 1234 ret = inode_ref_info(parent, 0, fs_root, path, &found_key); 1235 if (ret > 0) 1236 ret = -ENOENT; 1237 if (ret) 1238 break; 1239 1240 next_inum = found_key.offset; 1241 1242 /* regular exit ahead */ 1243 if (parent == next_inum) 1244 break; 1245 1246 slot = path->slots[0]; 1247 eb = path->nodes[0]; 1248 /* make sure we can use eb after releasing the path */ 1249 if (eb != eb_in) { 1250 atomic_inc(&eb->refs); 1251 btrfs_tree_read_lock(eb); 1252 btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK); 1253 } 1254 btrfs_release_path(path); 1255 iref = btrfs_item_ptr(eb, slot, struct btrfs_inode_ref); 1256 1257 name_len = btrfs_inode_ref_name_len(eb, iref); 1258 name_off = (unsigned long)(iref + 1); 1259 1260 parent = next_inum; 1261 --bytes_left; 1262 if (bytes_left >= 0) 1263 dest[bytes_left] = '/'; 1264 } 1265 1266 btrfs_release_path(path); 1267 path->leave_spinning = leave_spinning; 1268 1269 if (ret) 1270 return ERR_PTR(ret); 1271 1272 return dest + bytes_left; 1273 } 1274 1275 /* 1276 * this makes the path point to (logical EXTENT_ITEM *) 1277 * returns BTRFS_EXTENT_FLAG_DATA for data, BTRFS_EXTENT_FLAG_TREE_BLOCK for 1278 * tree blocks and <0 on error. 1279 */ 1280 int extent_from_logical(struct btrfs_fs_info *fs_info, u64 logical, 1281 struct btrfs_path *path, struct btrfs_key *found_key, 1282 u64 *flags_ret) 1283 { 1284 int ret; 1285 u64 flags; 1286 u32 item_size; 1287 struct extent_buffer *eb; 1288 struct btrfs_extent_item *ei; 1289 struct btrfs_key key; 1290 1291 key.type = BTRFS_EXTENT_ITEM_KEY; 1292 key.objectid = logical; 1293 key.offset = (u64)-1; 1294 1295 ret = btrfs_search_slot(NULL, fs_info->extent_root, &key, path, 0, 0); 1296 if (ret < 0) 1297 return ret; 1298 ret = btrfs_previous_item(fs_info->extent_root, path, 1299 0, BTRFS_EXTENT_ITEM_KEY); 1300 if (ret < 0) 1301 return ret; 1302 1303 btrfs_item_key_to_cpu(path->nodes[0], found_key, path->slots[0]); 1304 if (found_key->type != BTRFS_EXTENT_ITEM_KEY || 1305 found_key->objectid > logical || 1306 found_key->objectid + found_key->offset <= logical) { 1307 pr_debug("logical %llu is not within any extent\n", 1308 (unsigned long long)logical); 1309 return -ENOENT; 1310 } 1311 1312 eb = path->nodes[0]; 1313 item_size = btrfs_item_size_nr(eb, path->slots[0]); 1314 BUG_ON(item_size < sizeof(*ei)); 1315 1316 ei = btrfs_item_ptr(eb, path->slots[0], struct btrfs_extent_item); 1317 flags = btrfs_extent_flags(eb, ei); 1318 1319 pr_debug("logical %llu is at position %llu within the extent (%llu " 1320 "EXTENT_ITEM %llu) flags %#llx size %u\n", 1321 (unsigned long long)logical, 1322 (unsigned long long)(logical - found_key->objectid), 1323 (unsigned long long)found_key->objectid, 1324 (unsigned long long)found_key->offset, 1325 (unsigned long long)flags, item_size); 1326 1327 WARN_ON(!flags_ret); 1328 if (flags_ret) { 1329 if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) 1330 *flags_ret = BTRFS_EXTENT_FLAG_TREE_BLOCK; 1331 else if (flags & BTRFS_EXTENT_FLAG_DATA) 1332 *flags_ret = BTRFS_EXTENT_FLAG_DATA; 1333 else 1334 BUG_ON(1); 1335 return 0; 1336 } 1337 1338 return -EIO; 1339 } 1340 1341 /* 1342 * helper function to iterate extent inline refs. ptr must point to a 0 value 1343 * for the first call and may be modified. it is used to track state. 1344 * if more refs exist, 0 is returned and the next call to 1345 * __get_extent_inline_ref must pass the modified ptr parameter to get the 1346 * next ref. after the last ref was processed, 1 is returned. 1347 * returns <0 on error 1348 */ 1349 static int __get_extent_inline_ref(unsigned long *ptr, struct extent_buffer *eb, 1350 struct btrfs_extent_item *ei, u32 item_size, 1351 struct btrfs_extent_inline_ref **out_eiref, 1352 int *out_type) 1353 { 1354 unsigned long end; 1355 u64 flags; 1356 struct btrfs_tree_block_info *info; 1357 1358 if (!*ptr) { 1359 /* first call */ 1360 flags = btrfs_extent_flags(eb, ei); 1361 if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) { 1362 info = (struct btrfs_tree_block_info *)(ei + 1); 1363 *out_eiref = 1364 (struct btrfs_extent_inline_ref *)(info + 1); 1365 } else { 1366 *out_eiref = (struct btrfs_extent_inline_ref *)(ei + 1); 1367 } 1368 *ptr = (unsigned long)*out_eiref; 1369 if ((void *)*ptr >= (void *)ei + item_size) 1370 return -ENOENT; 1371 } 1372 1373 end = (unsigned long)ei + item_size; 1374 *out_eiref = (struct btrfs_extent_inline_ref *)*ptr; 1375 *out_type = btrfs_extent_inline_ref_type(eb, *out_eiref); 1376 1377 *ptr += btrfs_extent_inline_ref_size(*out_type); 1378 WARN_ON(*ptr > end); 1379 if (*ptr == end) 1380 return 1; /* last */ 1381 1382 return 0; 1383 } 1384 1385 /* 1386 * reads the tree block backref for an extent. tree level and root are returned 1387 * through out_level and out_root. ptr must point to a 0 value for the first 1388 * call and may be modified (see __get_extent_inline_ref comment). 1389 * returns 0 if data was provided, 1 if there was no more data to provide or 1390 * <0 on error. 1391 */ 1392 int tree_backref_for_extent(unsigned long *ptr, struct extent_buffer *eb, 1393 struct btrfs_extent_item *ei, u32 item_size, 1394 u64 *out_root, u8 *out_level) 1395 { 1396 int ret; 1397 int type; 1398 struct btrfs_tree_block_info *info; 1399 struct btrfs_extent_inline_ref *eiref; 1400 1401 if (*ptr == (unsigned long)-1) 1402 return 1; 1403 1404 while (1) { 1405 ret = __get_extent_inline_ref(ptr, eb, ei, item_size, 1406 &eiref, &type); 1407 if (ret < 0) 1408 return ret; 1409 1410 if (type == BTRFS_TREE_BLOCK_REF_KEY || 1411 type == BTRFS_SHARED_BLOCK_REF_KEY) 1412 break; 1413 1414 if (ret == 1) 1415 return 1; 1416 } 1417 1418 /* we can treat both ref types equally here */ 1419 info = (struct btrfs_tree_block_info *)(ei + 1); 1420 *out_root = btrfs_extent_inline_ref_offset(eb, eiref); 1421 *out_level = btrfs_tree_block_level(eb, info); 1422 1423 if (ret == 1) 1424 *ptr = (unsigned long)-1; 1425 1426 return 0; 1427 } 1428 1429 static int iterate_leaf_refs(struct extent_inode_elem *inode_list, 1430 u64 root, u64 extent_item_objectid, 1431 iterate_extent_inodes_t *iterate, void *ctx) 1432 { 1433 struct extent_inode_elem *eie; 1434 int ret = 0; 1435 1436 for (eie = inode_list; eie; eie = eie->next) { 1437 pr_debug("ref for %llu resolved, key (%llu EXTEND_DATA %llu), " 1438 "root %llu\n", extent_item_objectid, 1439 eie->inum, eie->offset, root); 1440 ret = iterate(eie->inum, eie->offset, root, ctx); 1441 if (ret) { 1442 pr_debug("stopping iteration for %llu due to ret=%d\n", 1443 extent_item_objectid, ret); 1444 break; 1445 } 1446 } 1447 1448 return ret; 1449 } 1450 1451 /* 1452 * calls iterate() for every inode that references the extent identified by 1453 * the given parameters. 1454 * when the iterator function returns a non-zero value, iteration stops. 1455 */ 1456 int iterate_extent_inodes(struct btrfs_fs_info *fs_info, 1457 u64 extent_item_objectid, u64 extent_item_pos, 1458 int search_commit_root, 1459 iterate_extent_inodes_t *iterate, void *ctx) 1460 { 1461 int ret; 1462 struct btrfs_trans_handle *trans; 1463 struct ulist *refs = NULL; 1464 struct ulist *roots = NULL; 1465 struct ulist_node *ref_node = NULL; 1466 struct ulist_node *root_node = NULL; 1467 struct seq_list tree_mod_seq_elem = {}; 1468 struct ulist_iterator ref_uiter; 1469 struct ulist_iterator root_uiter; 1470 1471 pr_debug("resolving all inodes for extent %llu\n", 1472 extent_item_objectid); 1473 1474 if (search_commit_root) { 1475 trans = BTRFS_BACKREF_SEARCH_COMMIT_ROOT; 1476 } else { 1477 trans = btrfs_join_transaction(fs_info->extent_root); 1478 if (IS_ERR(trans)) 1479 return PTR_ERR(trans); 1480 btrfs_get_tree_mod_seq(fs_info, &tree_mod_seq_elem); 1481 } 1482 1483 ret = btrfs_find_all_leafs(trans, fs_info, extent_item_objectid, 1484 tree_mod_seq_elem.seq, &refs, 1485 &extent_item_pos); 1486 if (ret) 1487 goto out; 1488 1489 ULIST_ITER_INIT(&ref_uiter); 1490 while (!ret && (ref_node = ulist_next(refs, &ref_uiter))) { 1491 ret = btrfs_find_all_roots(trans, fs_info, ref_node->val, 1492 tree_mod_seq_elem.seq, &roots); 1493 if (ret) 1494 break; 1495 ULIST_ITER_INIT(&root_uiter); 1496 while (!ret && (root_node = ulist_next(roots, &root_uiter))) { 1497 pr_debug("root %llu references leaf %llu, data list " 1498 "%#llx\n", root_node->val, ref_node->val, 1499 (long long)ref_node->aux); 1500 ret = iterate_leaf_refs((struct extent_inode_elem *) 1501 (uintptr_t)ref_node->aux, 1502 root_node->val, 1503 extent_item_objectid, 1504 iterate, ctx); 1505 } 1506 ulist_free(roots); 1507 } 1508 1509 free_leaf_list(refs); 1510 out: 1511 if (!search_commit_root) { 1512 btrfs_put_tree_mod_seq(fs_info, &tree_mod_seq_elem); 1513 btrfs_end_transaction(trans, fs_info->extent_root); 1514 } 1515 1516 return ret; 1517 } 1518 1519 int iterate_inodes_from_logical(u64 logical, struct btrfs_fs_info *fs_info, 1520 struct btrfs_path *path, 1521 iterate_extent_inodes_t *iterate, void *ctx) 1522 { 1523 int ret; 1524 u64 extent_item_pos; 1525 u64 flags = 0; 1526 struct btrfs_key found_key; 1527 int search_commit_root = path->search_commit_root; 1528 1529 ret = extent_from_logical(fs_info, logical, path, &found_key, &flags); 1530 btrfs_release_path(path); 1531 if (ret < 0) 1532 return ret; 1533 if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) 1534 return -EINVAL; 1535 1536 extent_item_pos = logical - found_key.objectid; 1537 ret = iterate_extent_inodes(fs_info, found_key.objectid, 1538 extent_item_pos, search_commit_root, 1539 iterate, ctx); 1540 1541 return ret; 1542 } 1543 1544 typedef int (iterate_irefs_t)(u64 parent, u32 name_len, unsigned long name_off, 1545 struct extent_buffer *eb, void *ctx); 1546 1547 static int iterate_inode_refs(u64 inum, struct btrfs_root *fs_root, 1548 struct btrfs_path *path, 1549 iterate_irefs_t *iterate, void *ctx) 1550 { 1551 int ret = 0; 1552 int slot; 1553 u32 cur; 1554 u32 len; 1555 u32 name_len; 1556 u64 parent = 0; 1557 int found = 0; 1558 struct extent_buffer *eb; 1559 struct btrfs_item *item; 1560 struct btrfs_inode_ref *iref; 1561 struct btrfs_key found_key; 1562 1563 while (!ret) { 1564 path->leave_spinning = 1; 1565 ret = inode_ref_info(inum, parent ? parent+1 : 0, fs_root, path, 1566 &found_key); 1567 if (ret < 0) 1568 break; 1569 if (ret) { 1570 ret = found ? 0 : -ENOENT; 1571 break; 1572 } 1573 ++found; 1574 1575 parent = found_key.offset; 1576 slot = path->slots[0]; 1577 eb = path->nodes[0]; 1578 /* make sure we can use eb after releasing the path */ 1579 atomic_inc(&eb->refs); 1580 btrfs_tree_read_lock(eb); 1581 btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK); 1582 btrfs_release_path(path); 1583 1584 item = btrfs_item_nr(eb, slot); 1585 iref = btrfs_item_ptr(eb, slot, struct btrfs_inode_ref); 1586 1587 for (cur = 0; cur < btrfs_item_size(eb, item); cur += len) { 1588 name_len = btrfs_inode_ref_name_len(eb, iref); 1589 /* path must be released before calling iterate()! */ 1590 pr_debug("following ref at offset %u for inode %llu in " 1591 "tree %llu\n", cur, 1592 (unsigned long long)found_key.objectid, 1593 (unsigned long long)fs_root->objectid); 1594 ret = iterate(parent, name_len, 1595 (unsigned long)(iref + 1), eb, ctx); 1596 if (ret) 1597 break; 1598 len = sizeof(*iref) + name_len; 1599 iref = (struct btrfs_inode_ref *)((char *)iref + len); 1600 } 1601 btrfs_tree_read_unlock_blocking(eb); 1602 free_extent_buffer(eb); 1603 } 1604 1605 btrfs_release_path(path); 1606 1607 return ret; 1608 } 1609 1610 static int iterate_inode_extrefs(u64 inum, struct btrfs_root *fs_root, 1611 struct btrfs_path *path, 1612 iterate_irefs_t *iterate, void *ctx) 1613 { 1614 int ret; 1615 int slot; 1616 u64 offset = 0; 1617 u64 parent; 1618 int found = 0; 1619 struct extent_buffer *eb; 1620 struct btrfs_inode_extref *extref; 1621 struct extent_buffer *leaf; 1622 u32 item_size; 1623 u32 cur_offset; 1624 unsigned long ptr; 1625 1626 while (1) { 1627 ret = btrfs_find_one_extref(fs_root, inum, offset, path, &extref, 1628 &offset); 1629 if (ret < 0) 1630 break; 1631 if (ret) { 1632 ret = found ? 0 : -ENOENT; 1633 break; 1634 } 1635 ++found; 1636 1637 slot = path->slots[0]; 1638 eb = path->nodes[0]; 1639 /* make sure we can use eb after releasing the path */ 1640 atomic_inc(&eb->refs); 1641 1642 btrfs_tree_read_lock(eb); 1643 btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK); 1644 btrfs_release_path(path); 1645 1646 leaf = path->nodes[0]; 1647 item_size = btrfs_item_size_nr(leaf, path->slots[0]); 1648 ptr = btrfs_item_ptr_offset(leaf, path->slots[0]); 1649 cur_offset = 0; 1650 1651 while (cur_offset < item_size) { 1652 u32 name_len; 1653 1654 extref = (struct btrfs_inode_extref *)(ptr + cur_offset); 1655 parent = btrfs_inode_extref_parent(eb, extref); 1656 name_len = btrfs_inode_extref_name_len(eb, extref); 1657 ret = iterate(parent, name_len, 1658 (unsigned long)&extref->name, eb, ctx); 1659 if (ret) 1660 break; 1661 1662 cur_offset += btrfs_inode_extref_name_len(leaf, extref); 1663 cur_offset += sizeof(*extref); 1664 } 1665 btrfs_tree_read_unlock_blocking(eb); 1666 free_extent_buffer(eb); 1667 1668 offset++; 1669 } 1670 1671 btrfs_release_path(path); 1672 1673 return ret; 1674 } 1675 1676 static int iterate_irefs(u64 inum, struct btrfs_root *fs_root, 1677 struct btrfs_path *path, iterate_irefs_t *iterate, 1678 void *ctx) 1679 { 1680 int ret; 1681 int found_refs = 0; 1682 1683 ret = iterate_inode_refs(inum, fs_root, path, iterate, ctx); 1684 if (!ret) 1685 ++found_refs; 1686 else if (ret != -ENOENT) 1687 return ret; 1688 1689 ret = iterate_inode_extrefs(inum, fs_root, path, iterate, ctx); 1690 if (ret == -ENOENT && found_refs) 1691 return 0; 1692 1693 return ret; 1694 } 1695 1696 /* 1697 * returns 0 if the path could be dumped (probably truncated) 1698 * returns <0 in case of an error 1699 */ 1700 static int inode_to_path(u64 inum, u32 name_len, unsigned long name_off, 1701 struct extent_buffer *eb, void *ctx) 1702 { 1703 struct inode_fs_paths *ipath = ctx; 1704 char *fspath; 1705 char *fspath_min; 1706 int i = ipath->fspath->elem_cnt; 1707 const int s_ptr = sizeof(char *); 1708 u32 bytes_left; 1709 1710 bytes_left = ipath->fspath->bytes_left > s_ptr ? 1711 ipath->fspath->bytes_left - s_ptr : 0; 1712 1713 fspath_min = (char *)ipath->fspath->val + (i + 1) * s_ptr; 1714 fspath = btrfs_ref_to_path(ipath->fs_root, ipath->btrfs_path, name_len, 1715 name_off, eb, inum, fspath_min, bytes_left); 1716 if (IS_ERR(fspath)) 1717 return PTR_ERR(fspath); 1718 1719 if (fspath > fspath_min) { 1720 ipath->fspath->val[i] = (u64)(unsigned long)fspath; 1721 ++ipath->fspath->elem_cnt; 1722 ipath->fspath->bytes_left = fspath - fspath_min; 1723 } else { 1724 ++ipath->fspath->elem_missed; 1725 ipath->fspath->bytes_missing += fspath_min - fspath; 1726 ipath->fspath->bytes_left = 0; 1727 } 1728 1729 return 0; 1730 } 1731 1732 /* 1733 * this dumps all file system paths to the inode into the ipath struct, provided 1734 * is has been created large enough. each path is zero-terminated and accessed 1735 * from ipath->fspath->val[i]. 1736 * when it returns, there are ipath->fspath->elem_cnt number of paths available 1737 * in ipath->fspath->val[]. when the allocated space wasn't sufficient, the 1738 * number of missed paths in recored in ipath->fspath->elem_missed, otherwise, 1739 * it's zero. ipath->fspath->bytes_missing holds the number of bytes that would 1740 * have been needed to return all paths. 1741 */ 1742 int paths_from_inode(u64 inum, struct inode_fs_paths *ipath) 1743 { 1744 return iterate_irefs(inum, ipath->fs_root, ipath->btrfs_path, 1745 inode_to_path, ipath); 1746 } 1747 1748 struct btrfs_data_container *init_data_container(u32 total_bytes) 1749 { 1750 struct btrfs_data_container *data; 1751 size_t alloc_bytes; 1752 1753 alloc_bytes = max_t(size_t, total_bytes, sizeof(*data)); 1754 data = vmalloc(alloc_bytes); 1755 if (!data) 1756 return ERR_PTR(-ENOMEM); 1757 1758 if (total_bytes >= sizeof(*data)) { 1759 data->bytes_left = total_bytes - sizeof(*data); 1760 data->bytes_missing = 0; 1761 } else { 1762 data->bytes_missing = sizeof(*data) - total_bytes; 1763 data->bytes_left = 0; 1764 } 1765 1766 data->elem_cnt = 0; 1767 data->elem_missed = 0; 1768 1769 return data; 1770 } 1771 1772 /* 1773 * allocates space to return multiple file system paths for an inode. 1774 * total_bytes to allocate are passed, note that space usable for actual path 1775 * information will be total_bytes - sizeof(struct inode_fs_paths). 1776 * the returned pointer must be freed with free_ipath() in the end. 1777 */ 1778 struct inode_fs_paths *init_ipath(s32 total_bytes, struct btrfs_root *fs_root, 1779 struct btrfs_path *path) 1780 { 1781 struct inode_fs_paths *ifp; 1782 struct btrfs_data_container *fspath; 1783 1784 fspath = init_data_container(total_bytes); 1785 if (IS_ERR(fspath)) 1786 return (void *)fspath; 1787 1788 ifp = kmalloc(sizeof(*ifp), GFP_NOFS); 1789 if (!ifp) { 1790 kfree(fspath); 1791 return ERR_PTR(-ENOMEM); 1792 } 1793 1794 ifp->btrfs_path = path; 1795 ifp->fspath = fspath; 1796 ifp->fs_root = fs_root; 1797 1798 return ifp; 1799 } 1800 1801 void free_ipath(struct inode_fs_paths *ipath) 1802 { 1803 if (!ipath) 1804 return; 1805 vfree(ipath->fspath); 1806 kfree(ipath); 1807 } 1808