cmds-check.c

   1 /*
   2  * Copyright (C) 2007 Oracle.  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 #define _XOPEN_SOURCE 500
  20 #define _GNU_SOURCE 1
  21 #include <stdio.h>
  22 #include <stdlib.h>
  23 #include <unistd.h>
  24 #include <fcntl.h>
  25 #include <sys/types.h>
  26 #include <sys/stat.h>
  27 #include <unistd.h>
  28 #include <getopt.h>
  29 #include <uuid/uuid.h>
  30 #include "ctree.h"
  31 #include "volumes.h"
  32 #include "repair.h"
  33 #include "disk-io.h"
  34 #include "print-tree.h"
  35 #include "transaction.h"
  36 #include "version.h"
  37 #include "utils.h"
  38 #include "commands.h"
  39 #include "free-space-cache.h"
  40 #include "btrfsck.h"
  41 #include "qgroup-verify.h"
  42
  43 static u64 bytes_used = 0;
  44 static u64 total_csum_bytes = 0;
  45 static u64 total_btree_bytes = 0;
  46 static u64 total_fs_tree_bytes = 0;
  47 static u64 total_extent_tree_bytes = 0;
  48 static u64 btree_space_waste = 0;
  49 static u64 data_bytes_allocated = 0;
  50 static u64 data_bytes_referenced = 0;
  51 static int found_old_backref = 0;
  52 static LIST_HEAD(duplicate_extents);
  53 static LIST_HEAD(delete_items);
  54 static int repair = 0;
  55 static int no_holes = 0;
  56 static int init_extent_tree = 0;
  57 static int check_data_csum = 0;
  58
  59 struct extent_backref {
  60         struct list_head list;
  61         unsigned int is_data:1;
  62         unsigned int found_extent_tree:1;
  63         unsigned int full_backref:1;
  64         unsigned int found_ref:1;
  65         unsigned int broken:1;
  66 };
  67
  68 struct data_backref {
  69         struct extent_backref node;
  70         union {
  71                 u64 parent;
  72                 u64 root;
  73         };
  74         u64 owner;
  75         u64 offset;
  76         u64 disk_bytenr;
  77         u64 bytes;
  78         u64 ram_bytes;
  79         u32 num_refs;
  80         u32 found_ref;
  81 };
  82
  83 struct tree_backref {
  84         struct extent_backref node;
  85         union {
  86                 u64 parent;
  87                 u64 root;
  88         };
  89 };
  90
  91 struct extent_record {
  92         struct list_head backrefs;
  93         struct list_head dups;
  94         struct list_head list;
  95         struct cache_extent cache;
  96         struct btrfs_disk_key parent_key;
  97         u64 start;
  98         u64 max_size;
  99         u64 nr;
 100         u64 refs;
 101         u64 extent_item_refs;
 102         u64 generation;
 103         u64 parent_generation;
 104         u64 info_objectid;
 105         u32 num_duplicates;
 106         u8 info_level;
 107         unsigned int found_rec:1;
 108         unsigned int content_checked:1;
 109         unsigned int owner_ref_checked:1;
 110         unsigned int is_root:1;
 111         unsigned int metadata:1;
 112 };
 113
 114 struct inode_backref {
 115         struct list_head list;
 116         unsigned int found_dir_item:1;
 117         unsigned int found_dir_index:1;
 118         unsigned int found_inode_ref:1;
 119         unsigned int filetype:8;
 120         int errors;
 121         unsigned int ref_type;
 122         u64 dir;
 123         u64 index;
 124         u16 namelen;
 125         char name[0];
 126 };
 127
 128 struct dropping_root_item_record {
 129         struct list_head list;
 130         struct btrfs_root_item ri;
 131         struct btrfs_key found_key;
 132 };
 133
 134 #define REF_ERR_NO_DIR_ITEM             (1 << 0)
 135 #define REF_ERR_NO_DIR_INDEX            (1 << 1)
 136 #define REF_ERR_NO_INODE_REF            (1 << 2)
 137 #define REF_ERR_DUP_DIR_ITEM            (1 << 3)
 138 #define REF_ERR_DUP_DIR_INDEX           (1 << 4)
 139 #define REF_ERR_DUP_INODE_REF           (1 << 5)
 140 #define REF_ERR_INDEX_UNMATCH           (1 << 6)
 141 #define REF_ERR_FILETYPE_UNMATCH        (1 << 7)
 142 #define REF_ERR_NAME_TOO_LONG           (1 << 8) // 100
 143 #define REF_ERR_NO_ROOT_REF             (1 << 9)
 144 #define REF_ERR_NO_ROOT_BACKREF         (1 << 10)
 145 #define REF_ERR_DUP_ROOT_REF            (1 << 11)
 146 #define REF_ERR_DUP_ROOT_BACKREF        (1 << 12)
 147
 148 struct inode_record {
 149         struct list_head backrefs;
 150         unsigned int checked:1;
 151         unsigned int merging:1;
 152         unsigned int found_inode_item:1;
 153         unsigned int found_dir_item:1;
 154         unsigned int found_file_extent:1;
 155         unsigned int found_csum_item:1;
 156         unsigned int some_csum_missing:1;
 157         unsigned int nodatasum:1;
 158         int errors;
 159
 160         u64 ino;
 161         u32 nlink;
 162         u32 imode;
 163         u64 isize;
 164         u64 nbytes;
 165
 166         u32 found_link;
 167         u64 found_size;
 168         u64 extent_start;
 169         u64 extent_end;
 170         u64 first_extent_gap;
 171
 172         u32 refs;
 173 };
 174
 175 #define I_ERR_NO_INODE_ITEM             (1 << 0)
 176 #define I_ERR_NO_ORPHAN_ITEM            (1 << 1)
 177 #define I_ERR_DUP_INODE_ITEM            (1 << 2)
 178 #define I_ERR_DUP_DIR_INDEX             (1 << 3)
 179 #define I_ERR_ODD_DIR_ITEM              (1 << 4)
 180 #define I_ERR_ODD_FILE_EXTENT           (1 << 5)
 181 #define I_ERR_BAD_FILE_EXTENT           (1 << 6)
 182 #define I_ERR_FILE_EXTENT_OVERLAP       (1 << 7)
 183 #define I_ERR_FILE_EXTENT_DISCOUNT      (1 << 8) // 100
 184 #define I_ERR_DIR_ISIZE_WRONG           (1 << 9)
 185 #define I_ERR_FILE_NBYTES_WRONG         (1 << 10) // 400
 186 #define I_ERR_ODD_CSUM_ITEM             (1 << 11)
 187 #define I_ERR_SOME_CSUM_MISSING         (1 << 12)
 188 #define I_ERR_LINK_COUNT_WRONG          (1 << 13)
 189
 190 struct root_backref {
 191         struct list_head list;
 192         unsigned int found_dir_item:1;
 193         unsigned int found_dir_index:1;
 194         unsigned int found_back_ref:1;
 195         unsigned int found_forward_ref:1;
 196         unsigned int reachable:1;
 197         int errors;
 198         u64 ref_root;
 199         u64 dir;
 200         u64 index;
 201         u16 namelen;
 202         char name[0];
 203 };
 204
 205 struct root_record {
 206         struct list_head backrefs;
 207         struct cache_extent cache;
 208         unsigned int found_root_item:1;
 209         u64 objectid;
 210         u32 found_ref;
 211 };
 212
 213 struct ptr_node {
 214         struct cache_extent cache;
 215         void *data;
 216 };
 217
 218 struct shared_node {
 219         struct cache_extent cache;
 220         struct cache_tree root_cache;
 221         struct cache_tree inode_cache;
 222         struct inode_record *current;
 223         u32 refs;
 224 };
 225
 226 struct block_info {
 227         u64 start;
 228         u32 size;
 229 };
 230
 231 struct walk_control {
 232         struct cache_tree shared;
 233         struct shared_node *nodes[BTRFS_MAX_LEVEL];
 234         int active_node;
 235         int root_level;
 236 };
 237
 238 struct bad_item {
 239         struct btrfs_key key;
 240         u64 root_id;
 241         struct list_head list;
 242 };
 243
 244 static void reset_cached_block_groups(struct btrfs_fs_info *fs_info);
 245
 246 static u8 imode_to_type(u32 imode)
 247 {
 248 #define S_SHIFT 12
 249         static unsigned char btrfs_type_by_mode[S_IFMT >> S_SHIFT] = {
 250                 [S_IFREG >> S_SHIFT]    = BTRFS_FT_REG_FILE,
 251                 [S_IFDIR >> S_SHIFT]    = BTRFS_FT_DIR,
 252                 [S_IFCHR >> S_SHIFT]    = BTRFS_FT_CHRDEV,
 253                 [S_IFBLK >> S_SHIFT]    = BTRFS_FT_BLKDEV,
 254                 [S_IFIFO >> S_SHIFT]    = BTRFS_FT_FIFO,
 255                 [S_IFSOCK >> S_SHIFT]   = BTRFS_FT_SOCK,
 256                 [S_IFLNK >> S_SHIFT]    = BTRFS_FT_SYMLINK,
 257         };
 258
 259         return btrfs_type_by_mode[(imode & S_IFMT) >> S_SHIFT];
 260 #undef S_SHIFT
 261 }
 262
 263 static int device_record_compare(struct rb_node *node1, struct rb_node *node2)
 264 {
 265         struct device_record *rec1;
 266         struct device_record *rec2;
 267
 268         rec1 = rb_entry(node1, struct device_record, node);
 269         rec2 = rb_entry(node2, struct device_record, node);
 270         if (rec1->devid > rec2->devid)
 271                 return -1;
 272         else if (rec1->devid < rec2->devid)
 273                 return 1;
 274         else
 275                 return 0;
 276 }
 277
 278 static struct inode_record *clone_inode_rec(struct inode_record *orig_rec)
 279 {
 280         struct inode_record *rec;
 281         struct inode_backref *backref;
 282         struct inode_backref *orig;
 283         size_t size;
 284
 285         rec = malloc(sizeof(*rec));
 286         memcpy(rec, orig_rec, sizeof(*rec));
 287         rec->refs = 1;
 288         INIT_LIST_HEAD(&rec->backrefs);
 289
 290         list_for_each_entry(orig, &orig_rec->backrefs, list) {
 291                 size = sizeof(*orig) + orig->namelen + 1;
 292                 backref = malloc(size);
 293                 memcpy(backref, orig, size);
 294                 list_add_tail(&backref->list, &rec->backrefs);
 295         }
 296         return rec;
 297 }
 298
 299 static void print_inode_error(int errors)
 300 {
 301         if (errors & I_ERR_NO_INODE_ITEM)
 302                 fprintf(stderr, ", no inode item");
 303         if (errors & I_ERR_NO_ORPHAN_ITEM)
 304                 fprintf(stderr, ", no orphan item");
 305         if (errors & I_ERR_DUP_INODE_ITEM)
 306                 fprintf(stderr, ", dup inode item");
 307         if (errors & I_ERR_DUP_DIR_INDEX)
 308                 fprintf(stderr, ", dup dir index");
 309         if (errors & I_ERR_ODD_DIR_ITEM)
 310                 fprintf(stderr, ", odd dir item");
 311         if (errors & I_ERR_ODD_FILE_EXTENT)
 312                 fprintf(stderr, ", odd file extent");
 313         if (errors & I_ERR_BAD_FILE_EXTENT)
 314                 fprintf(stderr, ", bad file extent");
 315         if (errors & I_ERR_FILE_EXTENT_OVERLAP)
 316                 fprintf(stderr, ", file extent overlap");
 317         if (errors & I_ERR_FILE_EXTENT_DISCOUNT)
 318                 fprintf(stderr, ", file extent discount");
 319         if (errors & I_ERR_DIR_ISIZE_WRONG)
 320                 fprintf(stderr, ", dir isize wrong");
 321         if (errors & I_ERR_FILE_NBYTES_WRONG)
 322                 fprintf(stderr, ", nbytes wrong");
 323         if (errors & I_ERR_ODD_CSUM_ITEM)
 324                 fprintf(stderr, ", odd csum item");
 325         if (errors & I_ERR_SOME_CSUM_MISSING)
 326                 fprintf(stderr, ", some csum missing");
 327         if (errors & I_ERR_LINK_COUNT_WRONG)
 328                 fprintf(stderr, ", link count wrong");
 329         fprintf(stderr, "\n");
 330 }
 331
 332 static void print_ref_error(int errors)
 333 {
 334         if (errors & REF_ERR_NO_DIR_ITEM)
 335                 fprintf(stderr, ", no dir item");
 336         if (errors & REF_ERR_NO_DIR_INDEX)
 337                 fprintf(stderr, ", no dir index");
 338         if (errors & REF_ERR_NO_INODE_REF)
 339                 fprintf(stderr, ", no inode ref");
 340         if (errors & REF_ERR_DUP_DIR_ITEM)
 341                 fprintf(stderr, ", dup dir item");
 342         if (errors & REF_ERR_DUP_DIR_INDEX)
 343                 fprintf(stderr, ", dup dir index");
 344         if (errors & REF_ERR_DUP_INODE_REF)
 345                 fprintf(stderr, ", dup inode ref");
 346         if (errors & REF_ERR_INDEX_UNMATCH)
 347                 fprintf(stderr, ", index unmatch");
 348         if (errors & REF_ERR_FILETYPE_UNMATCH)
 349                 fprintf(stderr, ", filetype unmatch");
 350         if (errors & REF_ERR_NAME_TOO_LONG)
 351                 fprintf(stderr, ", name too long");
 352         if (errors & REF_ERR_NO_ROOT_REF)
 353                 fprintf(stderr, ", no root ref");
 354         if (errors & REF_ERR_NO_ROOT_BACKREF)
 355                 fprintf(stderr, ", no root backref");
 356         if (errors & REF_ERR_DUP_ROOT_REF)
 357                 fprintf(stderr, ", dup root ref");
 358         if (errors & REF_ERR_DUP_ROOT_BACKREF)
 359                 fprintf(stderr, ", dup root backref");
 360         fprintf(stderr, "\n");
 361 }
 362
 363 static struct inode_record *get_inode_rec(struct cache_tree *inode_cache,
 364                                           u64 ino, int mod)
 365 {
 366         struct ptr_node *node;
 367         struct cache_extent *cache;
 368         struct inode_record *rec = NULL;
 369         int ret;
 370
 371         cache = lookup_cache_extent(inode_cache, ino, 1);
 372         if (cache) {
 373                 node = container_of(cache, struct ptr_node, cache);
 374                 rec = node->data;
 375                 if (mod && rec->refs > 1) {
 376                         node->data = clone_inode_rec(rec);
 377                         rec->refs--;
 378                         rec = node->data;
 379                 }
 380         } else if (mod) {
 381                 rec = calloc(1, sizeof(*rec));
 382                 rec->ino = ino;
 383                 rec->extent_start = (u64)-1;
 384                 rec->first_extent_gap = (u64)-1;
 385                 rec->refs = 1;
 386                 INIT_LIST_HEAD(&rec->backrefs);
 387
 388                 node = malloc(sizeof(*node));
 389                 node->cache.start = ino;
 390                 node->cache.size = 1;
 391                 node->data = rec;
 392
 393                 if (ino == BTRFS_FREE_INO_OBJECTID)
 394                         rec->found_link = 1;
 395
 396                 ret = insert_cache_extent(inode_cache, &node->cache);
 397                 BUG_ON(ret);
 398         }
 399         return rec;
 400 }
 401
 402 static void free_inode_rec(struct inode_record *rec)
 403 {
 404         struct inode_backref *backref;
 405
 406         if (--rec->refs > 0)
 407                 return;
 408
 409         while (!list_empty(&rec->backrefs)) {
 410                 backref = list_entry(rec->backrefs.next,
 411                                      struct inode_backref, list);
 412                 list_del(&backref->list);
 413                 free(backref);
 414         }
 415         free(rec);
 416 }
 417
 418 static int can_free_inode_rec(struct inode_record *rec)
 419 {
 420         if (!rec->errors && rec->checked && rec->found_inode_item &&
 421             rec->nlink == rec->found_link && list_empty(&rec->backrefs))
 422                 return 1;
 423         return 0;
 424 }
 425
 426 static void maybe_free_inode_rec(struct cache_tree *inode_cache,
 427                                  struct inode_record *rec)
 428 {
 429         struct cache_extent *cache;
 430         struct inode_backref *tmp, *backref;
 431         struct ptr_node *node;
 432         unsigned char filetype;
 433
 434         if (!rec->found_inode_item)
 435                 return;
 436
 437         filetype = imode_to_type(rec->imode);
 438         list_for_each_entry_safe(backref, tmp, &rec->backrefs, list) {
 439                 if (backref->found_dir_item && backref->found_dir_index) {
 440                         if (backref->filetype != filetype)
 441                                 backref->errors |= REF_ERR_FILETYPE_UNMATCH;
 442                         if (!backref->errors && backref->found_inode_ref) {
 443                                 list_del(&backref->list);
 444                                 free(backref);
 445                         }
 446                 }
 447         }
 448
 449         if (!rec->checked || rec->merging)
 450                 return;
 451
 452         if (S_ISDIR(rec->imode)) {
 453                 if (rec->found_size != rec->isize)
 454                         rec->errors |= I_ERR_DIR_ISIZE_WRONG;
 455                 if (rec->found_file_extent)
 456                         rec->errors |= I_ERR_ODD_FILE_EXTENT;
 457         } else if (S_ISREG(rec->imode) || S_ISLNK(rec->imode)) {
 458                 if (rec->found_dir_item)
 459                         rec->errors |= I_ERR_ODD_DIR_ITEM;
 460                 if (rec->found_size != rec->nbytes)
 461                         rec->errors |= I_ERR_FILE_NBYTES_WRONG;
 462                 if (rec->extent_start == (u64)-1 || rec->extent_start > 0)
 463                         rec->first_extent_gap = 0;
 464                 if (rec->nlink > 0 && !no_holes &&
 465                     (rec->extent_end < rec->isize ||
 466                      rec->first_extent_gap < rec->isize))
 467                         rec->errors |= I_ERR_FILE_EXTENT_DISCOUNT;
 468         }
 469
 470         if (S_ISREG(rec->imode) || S_ISLNK(rec->imode)) {
 471                 if (rec->found_csum_item && rec->nodatasum)
 472                         rec->errors |= I_ERR_ODD_CSUM_ITEM;
 473                 if (rec->some_csum_missing && !rec->nodatasum)
 474                         rec->errors |= I_ERR_SOME_CSUM_MISSING;
 475         }
 476
 477         BUG_ON(rec->refs != 1);
 478         if (can_free_inode_rec(rec)) {
 479                 cache = lookup_cache_extent(inode_cache, rec->ino, 1);
 480                 node = container_of(cache, struct ptr_node, cache);
 481                 BUG_ON(node->data != rec);
 482                 remove_cache_extent(inode_cache, &node->cache);
 483                 free(node);
 484                 free_inode_rec(rec);
 485         }
 486 }
 487
 488 static int check_orphan_item(struct btrfs_root *root, u64 ino)
 489 {
 490         struct btrfs_path path;
 491         struct btrfs_key key;
 492         int ret;
 493
 494         key.objectid = BTRFS_ORPHAN_OBJECTID;
 495         key.type = BTRFS_ORPHAN_ITEM_KEY;
 496         key.offset = ino;
 497
 498         btrfs_init_path(&path);
 499         ret = btrfs_search_slot(NULL, root, &key, &path, 0, 0);
 500         btrfs_release_path(&path);
 501         if (ret > 0)
 502                 ret = -ENOENT;
 503         return ret;
 504 }
 505
 506 static int process_inode_item(struct extent_buffer *eb,
 507                               int slot, struct btrfs_key *key,
 508                               struct shared_node *active_node)
 509 {
 510         struct inode_record *rec;
 511         struct btrfs_inode_item *item;
 512
 513         rec = active_node->current;
 514         BUG_ON(rec->ino != key->objectid || rec->refs > 1);
 515         if (rec->found_inode_item) {
 516                 rec->errors |= I_ERR_DUP_INODE_ITEM;
 517                 return 1;
 518         }
 519         item = btrfs_item_ptr(eb, slot, struct btrfs_inode_item);
 520         rec->nlink = btrfs_inode_nlink(eb, item);
 521         rec->isize = btrfs_inode_size(eb, item);
 522         rec->nbytes = btrfs_inode_nbytes(eb, item);
 523         rec->imode = btrfs_inode_mode(eb, item);
 524         if (btrfs_inode_flags(eb, item) & BTRFS_INODE_NODATASUM)
 525                 rec->nodatasum = 1;
 526         rec->found_inode_item = 1;
 527         if (rec->nlink == 0)
 528                 rec->errors |= I_ERR_NO_ORPHAN_ITEM;
 529         maybe_free_inode_rec(&active_node->inode_cache, rec);
 530         return 0;
 531 }
 532
 533 static struct inode_backref *get_inode_backref(struct inode_record *rec,
 534                                                 const char *name,
 535                                                 int namelen, u64 dir)
 536 {
 537         struct inode_backref *backref;
 538
 539         list_for_each_entry(backref, &rec->backrefs, list) {
 540                 if (backref->dir != dir || backref->namelen != namelen)
 541                         continue;
 542                 if (memcmp(name, backref->name, namelen))
 543                         continue;
 544                 return backref;
 545         }
 546
 547         backref = malloc(sizeof(*backref) + namelen + 1);
 548         memset(backref, 0, sizeof(*backref));
 549         backref->dir = dir;
 550         backref->namelen = namelen;
 551         memcpy(backref->name, name, namelen);
 552         backref->name[namelen] = '\0';
 553         list_add_tail(&backref->list, &rec->backrefs);
 554         return backref;
 555 }
 556
 557 static int add_inode_backref(struct cache_tree *inode_cache,
 558                              u64 ino, u64 dir, u64 index,
 559                              const char *name, int namelen,
 560                              int filetype, int itemtype, int errors)
 561 {
 562         struct inode_record *rec;
 563         struct inode_backref *backref;
 564
 565         rec = get_inode_rec(inode_cache, ino, 1);
 566         backref = get_inode_backref(rec, name, namelen, dir);
 567         if (errors)
 568                 backref->errors |= errors;
 569         if (itemtype == BTRFS_DIR_INDEX_KEY) {
 570                 if (backref->found_dir_index)
 571                         backref->errors |= REF_ERR_DUP_DIR_INDEX;
 572                 if (backref->found_inode_ref && backref->index != index)
 573                         backref->errors |= REF_ERR_INDEX_UNMATCH;
 574                 if (backref->found_dir_item && backref->filetype != filetype)
 575                         backref->errors |= REF_ERR_FILETYPE_UNMATCH;
 576
 577                 backref->index = index;
 578                 backref->filetype = filetype;
 579                 backref->found_dir_index = 1;
 580         } else if (itemtype == BTRFS_DIR_ITEM_KEY) {
 581                 rec->found_link++;
 582                 if (backref->found_dir_item)
 583                         backref->errors |= REF_ERR_DUP_DIR_ITEM;
 584                 if (backref->found_dir_index && backref->filetype != filetype)
 585                         backref->errors |= REF_ERR_FILETYPE_UNMATCH;
 586
 587                 backref->filetype = filetype;
 588                 backref->found_dir_item = 1;
 589         } else if ((itemtype == BTRFS_INODE_REF_KEY) ||
 590                    (itemtype == BTRFS_INODE_EXTREF_KEY)) {
 591                 if (backref->found_inode_ref)
 592                         backref->errors |= REF_ERR_DUP_INODE_REF;
 593                 if (backref->found_dir_index && backref->index != index)
 594                         backref->errors |= REF_ERR_INDEX_UNMATCH;
 595
 596                 backref->ref_type = itemtype;
 597                 backref->index = index;
 598                 backref->found_inode_ref = 1;
 599         } else {
 600                 BUG_ON(1);
 601         }
 602
 603         maybe_free_inode_rec(inode_cache, rec);
 604         return 0;
 605 }
 606
 607 static int merge_inode_recs(struct inode_record *src, struct inode_record *dst,
 608                             struct cache_tree *dst_cache)
 609 {
 610         struct inode_backref *backref;
 611         u32 dir_count = 0;
 612
 613         dst->merging = 1;
 614         list_for_each_entry(backref, &src->backrefs, list) {
 615                 if (backref->found_dir_index) {
 616                         add_inode_backref(dst_cache, dst->ino, backref->dir,
 617                                         backref->index, backref->name,
 618                                         backref->namelen, backref->filetype,
 619                                         BTRFS_DIR_INDEX_KEY, backref->errors);
 620                 }
 621                 if (backref->found_dir_item) {
 622                         dir_count++;
 623                         add_inode_backref(dst_cache, dst->ino,
 624                                         backref->dir, 0, backref->name,
 625                                         backref->namelen, backref->filetype,
 626                                         BTRFS_DIR_ITEM_KEY, backref->errors);
 627                 }
 628                 if (backref->found_inode_ref) {
 629                         add_inode_backref(dst_cache, dst->ino,
 630                                         backref->dir, backref->index,
 631                                         backref->name, backref->namelen, 0,
 632                                         backref->ref_type, backref->errors);
 633                 }
 634         }
 635
 636         if (src->found_dir_item)
 637                 dst->found_dir_item = 1;
 638         if (src->found_file_extent)
 639                 dst->found_file_extent = 1;
 640         if (src->found_csum_item)
 641                 dst->found_csum_item = 1;
 642         if (src->some_csum_missing)
 643                 dst->some_csum_missing = 1;
 644         if (dst->first_extent_gap > src->first_extent_gap)
 645                 dst->first_extent_gap = src->first_extent_gap;
 646
 647         BUG_ON(src->found_link < dir_count);
 648         dst->found_link += src->found_link - dir_count;
 649         dst->found_size += src->found_size;
 650         if (src->extent_start != (u64)-1) {
 651                 if (dst->extent_start == (u64)-1) {
 652                         dst->extent_start = src->extent_start;
 653                         dst->extent_end = src->extent_end;
 654                 } else {
 655                         if (dst->extent_end > src->extent_start)
 656                                 dst->errors |= I_ERR_FILE_EXTENT_OVERLAP;
 657                         else if (dst->extent_end < src->extent_start &&
 658                                  dst->extent_end < dst->first_extent_gap)
 659                                 dst->first_extent_gap = dst->extent_end;
 660                         if (dst->extent_end < src->extent_end)
 661                                 dst->extent_end = src->extent_end;
 662                 }
 663         }
 664
 665         dst->errors |= src->errors;
 666         if (src->found_inode_item) {
 667                 if (!dst->found_inode_item) {
 668                         dst->nlink = src->nlink;
 669                         dst->isize = src->isize;
 670                         dst->nbytes = src->nbytes;
 671                         dst->imode = src->imode;
 672                         dst->nodatasum = src->nodatasum;
 673                         dst->found_inode_item = 1;
 674                 } else {
 675                         dst->errors |= I_ERR_DUP_INODE_ITEM;
 676                 }
 677         }
 678         dst->merging = 0;
 679
 680         return 0;
 681 }
 682
 683 static int splice_shared_node(struct shared_node *src_node,
 684                               struct shared_node *dst_node)
 685 {
 686         struct cache_extent *cache;
 687         struct ptr_node *node, *ins;
 688         struct cache_tree *src, *dst;
 689         struct inode_record *rec, *conflict;
 690         u64 current_ino = 0;
 691         int splice = 0;
 692         int ret;
 693
 694         if (--src_node->refs == 0)
 695                 splice = 1;
 696         if (src_node->current)
 697                 current_ino = src_node->current->ino;
 698
 699         src = &src_node->root_cache;
 700         dst = &dst_node->root_cache;
 701 again:
 702         cache = search_cache_extent(src, 0);
 703         while (cache) {
 704                 node = container_of(cache, struct ptr_node, cache);
 705                 rec = node->data;
 706                 cache = next_cache_extent(cache);
 707
 708                 if (splice) {
 709                         remove_cache_extent(src, &node->cache);
 710                         ins = node;
 711                 } else {
 712                         ins = malloc(sizeof(*ins));
 713                         ins->cache.start = node->cache.start;
 714                         ins->cache.size = node->cache.size;
 715                         ins->data = rec;
 716                         rec->refs++;
 717                 }
 718                 ret = insert_cache_extent(dst, &ins->cache);
 719                 if (ret == -EEXIST) {
 720                         conflict = get_inode_rec(dst, rec->ino, 1);
 721                         merge_inode_recs(rec, conflict, dst);
 722                         if (rec->checked) {
 723                                 conflict->checked = 1;
 724                                 if (dst_node->current == conflict)
 725                                         dst_node->current = NULL;
 726                         }
 727                         maybe_free_inode_rec(dst, conflict);
 728                         free_inode_rec(rec);
 729                         free(ins);
 730                 } else {
 731                         BUG_ON(ret);
 732                 }
 733         }
 734
 735         if (src == &src_node->root_cache) {
 736                 src = &src_node->inode_cache;
 737                 dst = &dst_node->inode_cache;
 738                 goto again;
 739         }
 740
 741         if (current_ino > 0 && (!dst_node->current ||
 742             current_ino > dst_node->current->ino)) {
 743                 if (dst_node->current) {
 744                         dst_node->current->checked = 1;
 745                         maybe_free_inode_rec(dst, dst_node->current);
 746                 }
 747                 dst_node->current = get_inode_rec(dst, current_ino, 1);
 748         }
 749         return 0;
 750 }
 751
 752 static void free_inode_ptr(struct cache_extent *cache)
 753 {
 754         struct ptr_node *node;
 755         struct inode_record *rec;
 756
 757         node = container_of(cache, struct ptr_node, cache);
 758         rec = node->data;
 759         free_inode_rec(rec);
 760         free(node);
 761 }
 762
 763 FREE_EXTENT_CACHE_BASED_TREE(inode_recs, free_inode_ptr);
 764
 765 static struct shared_node *find_shared_node(struct cache_tree *shared,
 766                                             u64 bytenr)
 767 {
 768         struct cache_extent *cache;
 769         struct shared_node *node;
 770
 771         cache = lookup_cache_extent(shared, bytenr, 1);
 772         if (cache) {
 773                 node = container_of(cache, struct shared_node, cache);
 774                 return node;
 775         }
 776         return NULL;
 777 }
 778
 779 static int add_shared_node(struct cache_tree *shared, u64 bytenr, u32 refs)
 780 {
 781         int ret;
 782         struct shared_node *node;
 783
 784         node = calloc(1, sizeof(*node));
 785         node->cache.start = bytenr;
 786         node->cache.size = 1;
 787         cache_tree_init(&node->root_cache);
 788         cache_tree_init(&node->inode_cache);
 789         node->refs = refs;
 790
 791         ret = insert_cache_extent(shared, &node->cache);
 792         BUG_ON(ret);
 793         return 0;
 794 }
 795
 796 static int enter_shared_node(struct btrfs_root *root, u64 bytenr, u32 refs,
 797                              struct walk_control *wc, int level)
 798 {
 799         struct shared_node *node;
 800         struct shared_node *dest;
 801
 802         if (level == wc->active_node)
 803                 return 0;
 804
 805         BUG_ON(wc->active_node <= level);
 806         node = find_shared_node(&wc->shared, bytenr);
 807         if (!node) {
 808                 add_shared_node(&wc->shared, bytenr, refs);
 809                 node = find_shared_node(&wc->shared, bytenr);
 810                 wc->nodes[level] = node;
 811                 wc->active_node = level;
 812                 return 0;
 813         }
 814
 815         if (wc->root_level == wc->active_node &&
 816             btrfs_root_refs(&root->root_item) == 0) {
 817                 if (--node->refs == 0) {
 818                         free_inode_recs_tree(&node->root_cache);
 819                         free_inode_recs_tree(&node->inode_cache);
 820                         remove_cache_extent(&wc->shared, &node->cache);
 821                         free(node);
 822                 }
 823                 return 1;
 824         }
 825
 826         dest = wc->nodes[wc->active_node];
 827         splice_shared_node(node, dest);
 828         if (node->refs == 0) {
 829                 remove_cache_extent(&wc->shared, &node->cache);
 830                 free(node);
 831         }
 832         return 1;
 833 }
 834
 835 static int leave_shared_node(struct btrfs_root *root,
 836                              struct walk_control *wc, int level)
 837 {
 838         struct shared_node *node;
 839         struct shared_node *dest;
 840         int i;
 841
 842         if (level == wc->root_level)
 843                 return 0;
 844
 845         for (i = level + 1; i < BTRFS_MAX_LEVEL; i++) {
 846                 if (wc->nodes[i])
 847                         break;
 848         }
 849         BUG_ON(i >= BTRFS_MAX_LEVEL);
 850
 851         node = wc->nodes[wc->active_node];
 852         wc->nodes[wc->active_node] = NULL;
 853         wc->active_node = i;
 854
 855         dest = wc->nodes[wc->active_node];
 856         if (wc->active_node < wc->root_level ||
 857             btrfs_root_refs(&root->root_item) > 0) {
 858                 BUG_ON(node->refs <= 1);
 859                 splice_shared_node(node, dest);
 860         } else {
 861                 BUG_ON(node->refs < 2);
 862                 node->refs--;
 863         }
 864         return 0;
 865 }
 866
 867 static int is_child_root(struct btrfs_root *root, u64 parent_root_id,
 868                          u64 child_root_id)
 869 {
 870         struct btrfs_path path;
 871         struct btrfs_key key;
 872         struct extent_buffer *leaf;
 873         int has_parent = 0;
 874         int ret;
 875
 876         btrfs_init_path(&path);
 877
 878         key.objectid = parent_root_id;
 879         key.type = BTRFS_ROOT_REF_KEY;
 880         key.offset = child_root_id;
 881         ret = btrfs_search_slot(NULL, root->fs_info->tree_root, &key, &path,
 882                                 0, 0);
 883         BUG_ON(ret < 0);
 884         btrfs_release_path(&path);
 885         if (!ret)
 886                 return 1;
 887
 888         key.objectid = child_root_id;
 889         key.type = BTRFS_ROOT_BACKREF_KEY;
 890         key.offset = 0;
 891         ret = btrfs_search_slot(NULL, root->fs_info->tree_root, &key, &path,
 892                                 0, 0);
 893         BUG_ON(ret <= 0);
 894
 895         while (1) {
 896                 leaf = path.nodes[0];
 897                 if (path.slots[0] >= btrfs_header_nritems(leaf)) {
 898                         ret = btrfs_next_leaf(root->fs_info->tree_root, &path);
 899                         BUG_ON(ret < 0);
 900
 901                         if (ret > 0)
 902                                 break;
 903                         leaf = path.nodes[0];
 904                 }
 905
 906                 btrfs_item_key_to_cpu(leaf, &key, path.slots[0]);
 907                 if (key.objectid != child_root_id ||
 908                     key.type != BTRFS_ROOT_BACKREF_KEY)
 909                         break;
 910
 911                 has_parent = 1;
 912
 913                 if (key.offset == parent_root_id) {
 914                         btrfs_release_path(&path);
 915                         return 1;
 916                 }
 917
 918                 path.slots[0]++;
 919         }
 920
 921         btrfs_release_path(&path);
 922         return has_parent? 0 : -1;
 923 }
 924
 925 static int process_dir_item(struct btrfs_root *root,
 926                             struct extent_buffer *eb,
 927                             int slot, struct btrfs_key *key,
 928                             struct shared_node *active_node)
 929 {
 930         u32 total;
 931         u32 cur = 0;
 932         u32 len;
 933         u32 name_len;
 934         u32 data_len;
 935         int error;
 936         int nritems = 0;
 937         int filetype;
 938         struct btrfs_dir_item *di;
 939         struct inode_record *rec;
 940         struct cache_tree *root_cache;
 941         struct cache_tree *inode_cache;
 942         struct btrfs_key location;
 943         char namebuf[BTRFS_NAME_LEN];
 944
 945         root_cache = &active_node->root_cache;
 946         inode_cache = &active_node->inode_cache;
 947         rec = active_node->current;
 948         rec->found_dir_item = 1;
 949
 950         di = btrfs_item_ptr(eb, slot, struct btrfs_dir_item);
 951         total = btrfs_item_size_nr(eb, slot);
 952         while (cur < total) {
 953                 nritems++;
 954                 btrfs_dir_item_key_to_cpu(eb, di, &location);
 955                 name_len = btrfs_dir_name_len(eb, di);
 956                 data_len = btrfs_dir_data_len(eb, di);
 957                 filetype = btrfs_dir_type(eb, di);
 958
 959                 rec->found_size += name_len;
 960                 if (name_len <= BTRFS_NAME_LEN) {
 961                         len = name_len;
 962                         error = 0;
 963                 } else {
 964                         len = BTRFS_NAME_LEN;
 965                         error = REF_ERR_NAME_TOO_LONG;
 966                 }
 967                 read_extent_buffer(eb, namebuf, (unsigned long)(di + 1), len);
 968
 969                 if (location.type == BTRFS_INODE_ITEM_KEY) {
 970                         add_inode_backref(inode_cache, location.objectid,
 971                                           key->objectid, key->offset, namebuf,
 972                                           len, filetype, key->type, error);
 973                 } else if (location.type == BTRFS_ROOT_ITEM_KEY) {
 974                         add_inode_backref(root_cache, location.objectid,
 975                                           key->objectid, key->offset,
 976                                           namebuf, len, filetype,
 977                                           key->type, error);
 978                 } else {
 979                         fprintf(stderr, "warning line %d\n", __LINE__);
 980                 }
 981
 982                 len = sizeof(*di) + name_len + data_len;
 983                 di = (struct btrfs_dir_item *)((char *)di + len);
 984                 cur += len;
 985         }
 986         if (key->type == BTRFS_DIR_INDEX_KEY && nritems > 1)
 987                 rec->errors |= I_ERR_DUP_DIR_INDEX;
 988
 989         return 0;
 990 }
 991
 992 static int process_inode_ref(struct extent_buffer *eb,
 993                              int slot, struct btrfs_key *key,
 994                              struct shared_node *active_node)
 995 {
 996         u32 total;
 997         u32 cur = 0;
 998         u32 len;
 999         u32 name_len;
1000         u64 index;
1001         int error;
1002         struct cache_tree *inode_cache;
1003         struct btrfs_inode_ref *ref;
1004         char namebuf[BTRFS_NAME_LEN];
1005
1006         inode_cache = &active_node->inode_cache;
1007
1008         ref = btrfs_item_ptr(eb, slot, struct btrfs_inode_ref);
1009         total = btrfs_item_size_nr(eb, slot);
1010         while (cur < total) {
1011                 name_len = btrfs_inode_ref_name_len(eb, ref);
1012                 index = btrfs_inode_ref_index(eb, ref);
1013                 if (name_len <= BTRFS_NAME_LEN) {
1014                         len = name_len;
1015                         error = 0;
1016                 } else {
1017                         len = BTRFS_NAME_LEN;
1018                         error = REF_ERR_NAME_TOO_LONG;
1019                 }
1020                 read_extent_buffer(eb, namebuf, (unsigned long)(ref + 1), len);
1021                 add_inode_backref(inode_cache, key->objectid, key->offset,
1022                                   index, namebuf, len, 0, key->type, error);
1023
1024                 len = sizeof(*ref) + name_len;
1025                 ref = (struct btrfs_inode_ref *)((char *)ref + len);
1026                 cur += len;
1027         }
1028         return 0;
1029 }
1030
1031 static int process_inode_extref(struct extent_buffer *eb,
1032                                 int slot, struct btrfs_key *key,
1033                                 struct shared_node *active_node)
1034 {
1035         u32 total;
1036         u32 cur = 0;
1037         u32 len;
1038         u32 name_len;
1039         u64 index;
1040         u64 parent;
1041         int error;
1042         struct cache_tree *inode_cache;
1043         struct btrfs_inode_extref *extref;
1044         char namebuf[BTRFS_NAME_LEN];
1045
1046         inode_cache = &active_node->inode_cache;
1047
1048         extref = btrfs_item_ptr(eb, slot, struct btrfs_inode_extref);
1049         total = btrfs_item_size_nr(eb, slot);
1050         while (cur < total) {
1051                 name_len = btrfs_inode_extref_name_len(eb, extref);
1052                 index = btrfs_inode_extref_index(eb, extref);
1053                 parent = btrfs_inode_extref_parent(eb, extref);
1054                 if (name_len <= BTRFS_NAME_LEN) {
1055                         len = name_len;
1056                         error = 0;
1057                 } else {
1058                         len = BTRFS_NAME_LEN;
1059                         error = REF_ERR_NAME_TOO_LONG;
1060                 }
1061                 read_extent_buffer(eb, namebuf,
1062                                    (unsigned long)(extref + 1), len);
1063                 add_inode_backref(inode_cache, key->objectid, parent,
1064                                   index, namebuf, len, 0, key->type, error);
1065
1066                 len = sizeof(*extref) + name_len;
1067                 extref = (struct btrfs_inode_extref *)((char *)extref + len);
1068                 cur += len;
1069         }
1070         return 0;
1071
1072 }
1073
1074 static u64 count_csum_range(struct btrfs_root *root, u64 start, u64 len)
1075 {
1076         struct btrfs_key key;
1077         struct btrfs_path path;
1078         struct extent_buffer *leaf;
1079         int ret ;
1080         size_t size;
1081         u64 found = 0;
1082         u64 csum_end;
1083         u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy);
1084
1085         btrfs_init_path(&path);
1086
1087         key.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
1088         key.offset = start;
1089         key.type = BTRFS_EXTENT_CSUM_KEY;
1090
1091         ret = btrfs_search_slot(NULL, root->fs_info->csum_root,
1092                                 &key, &path, 0, 0);
1093         BUG_ON(ret < 0);
1094         if (ret > 0 && path.slots[0] > 0) {
1095                 leaf = path.nodes[0];
1096                 btrfs_item_key_to_cpu(leaf, &key, path.slots[0] - 1);
1097                 if (key.objectid == BTRFS_EXTENT_CSUM_OBJECTID &&
1098                     key.type == BTRFS_EXTENT_CSUM_KEY)
1099                         path.slots[0]--;
1100         }
1101
1102         while (len > 0) {
1103                 leaf = path.nodes[0];
1104                 if (path.slots[0] >= btrfs_header_nritems(leaf)) {
1105                         ret = btrfs_next_leaf(root->fs_info->csum_root, &path);
1106                         BUG_ON(ret < 0);
1107                         if (ret > 0)
1108                                 break;
1109                         leaf = path.nodes[0];
1110                 }
1111
1112                 btrfs_item_key_to_cpu(leaf, &key, path.slots[0]);
1113                 if (key.objectid != BTRFS_EXTENT_CSUM_OBJECTID ||
1114                     key.type != BTRFS_EXTENT_CSUM_KEY)
1115                         break;
1116
1117                 btrfs_item_key_to_cpu(leaf, &key, path.slots[0]);
1118                 if (key.offset >= start + len)
1119                         break;
1120
1121                 if (key.offset > start)
1122                         start = key.offset;
1123
1124                 size = btrfs_item_size_nr(leaf, path.slots[0]);
1125                 csum_end = key.offset + (size / csum_size) * root->sectorsize;
1126                 if (csum_end > start) {
1127                         size = min(csum_end - start, len);
1128                         len -= size;
1129                         start += size;
1130                         found += size;
1131                 }
1132
1133                 path.slots[0]++;
1134         }
1135         btrfs_release_path(&path);
1136         return found;
1137 }
1138
1139 static int process_file_extent(struct btrfs_root *root,
1140                                 struct extent_buffer *eb,
1141                                 int slot, struct btrfs_key *key,
1142                                 struct shared_node *active_node)
1143 {
1144         struct inode_record *rec;
1145         struct btrfs_file_extent_item *fi;
1146         u64 num_bytes = 0;
1147         u64 disk_bytenr = 0;
1148         u64 extent_offset = 0;
1149         u64 mask = root->sectorsize - 1;
1150         int extent_type;
1151
1152         rec = active_node->current;
1153         BUG_ON(rec->ino != key->objectid || rec->refs > 1);
1154         rec->found_file_extent = 1;
1155
1156         if (rec->extent_start == (u64)-1) {
1157                 rec->extent_start = key->offset;
1158                 rec->extent_end = key->offset;
1159         }
1160
1161         if (rec->extent_end > key->offset)
1162                 rec->errors |= I_ERR_FILE_EXTENT_OVERLAP;
1163         else if (rec->extent_end < key->offset &&
1164                  rec->extent_end < rec->first_extent_gap)
1165                 rec->first_extent_gap = rec->extent_end;
1166
1167         fi = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item);
1168         extent_type = btrfs_file_extent_type(eb, fi);
1169
1170         if (extent_type == BTRFS_FILE_EXTENT_INLINE) {
1171                 num_bytes = btrfs_file_extent_inline_len(eb, slot, fi);
1172                 if (num_bytes == 0)
1173                         rec->errors |= I_ERR_BAD_FILE_EXTENT;
1174                 rec->found_size += num_bytes;
1175                 num_bytes = (num_bytes + mask) & ~mask;
1176         } else if (extent_type == BTRFS_FILE_EXTENT_REG ||
1177                    extent_type == BTRFS_FILE_EXTENT_PREALLOC) {
1178                 num_bytes = btrfs_file_extent_num_bytes(eb, fi);
1179                 disk_bytenr = btrfs_file_extent_disk_bytenr(eb, fi);
1180                 extent_offset = btrfs_file_extent_offset(eb, fi);
1181                 if (num_bytes == 0 || (num_bytes & mask))
1182                         rec->errors |= I_ERR_BAD_FILE_EXTENT;
1183                 if (num_bytes + extent_offset >
1184                     btrfs_file_extent_ram_bytes(eb, fi))
1185                         rec->errors |= I_ERR_BAD_FILE_EXTENT;
1186                 if (extent_type == BTRFS_FILE_EXTENT_PREALLOC &&
1187                     (btrfs_file_extent_compression(eb, fi) ||
1188                      btrfs_file_extent_encryption(eb, fi) ||
1189                      btrfs_file_extent_other_encoding(eb, fi)))
1190                         rec->errors |= I_ERR_BAD_FILE_EXTENT;
1191                 if (disk_bytenr > 0)
1192                         rec->found_size += num_bytes;
1193         } else {
1194                 rec->errors |= I_ERR_BAD_FILE_EXTENT;
1195         }
1196         rec->extent_end = key->offset + num_bytes;
1197
1198         if (disk_bytenr > 0) {
1199                 u64 found;
1200                 if (btrfs_file_extent_compression(eb, fi))
1201                         num_bytes = btrfs_file_extent_disk_num_bytes(eb, fi);
1202                 else
1203                         disk_bytenr += extent_offset;
1204
1205                 found = count_csum_range(root, disk_bytenr, num_bytes);
1206                 if (extent_type == BTRFS_FILE_EXTENT_REG) {
1207                         if (found > 0)
1208                                 rec->found_csum_item = 1;
1209                         if (found < num_bytes)
1210                                 rec->some_csum_missing = 1;
1211                 } else if (extent_type == BTRFS_FILE_EXTENT_PREALLOC) {
1212                         if (found > 0)
1213                                 rec->errors |= I_ERR_ODD_CSUM_ITEM;
1214                 }
1215         }
1216         return 0;
1217 }
1218
1219 static int process_one_leaf(struct btrfs_root *root, struct extent_buffer *eb,
1220                             struct walk_control *wc)
1221 {
1222         struct btrfs_key key;
1223         u32 nritems;
1224         int i;
1225         int ret = 0;
1226         int error = 0;
1227         struct cache_tree *inode_cache;
1228         struct shared_node *active_node;
1229
1230         if (wc->root_level == wc->active_node &&
1231             btrfs_root_refs(&root->root_item) == 0)
1232                 return 0;
1233
1234         active_node = wc->nodes[wc->active_node];
1235         inode_cache = &active_node->inode_cache;
1236         nritems = btrfs_header_nritems(eb);
1237         for (i = 0; i < nritems; i++) {
1238                 btrfs_item_key_to_cpu(eb, &key, i);
1239
1240                 if (key.objectid == BTRFS_FREE_SPACE_OBJECTID)
1241                         continue;
1242                 if (key.type == BTRFS_ORPHAN_ITEM_KEY)
1243                         continue;
1244
1245                 if (active_node->current == NULL ||
1246                     active_node->current->ino < key.objectid) {
1247                         if (active_node->current) {
1248                                 active_node->current->checked = 1;
1249                                 maybe_free_inode_rec(inode_cache,
1250                                                      active_node->current);
1251                         }
1252                         active_node->current = get_inode_rec(inode_cache,
1253                                                              key.objectid, 1);
1254                 }
1255                 switch (key.type) {
1256                 case BTRFS_DIR_ITEM_KEY:
1257                 case BTRFS_DIR_INDEX_KEY:
1258                         ret = process_dir_item(root, eb, i, &key, active_node);
1259                         break;
1260                 case BTRFS_INODE_REF_KEY:
1261                         ret = process_inode_ref(eb, i, &key, active_node);
1262                         break;
1263                 case BTRFS_INODE_EXTREF_KEY:
1264                         ret = process_inode_extref(eb, i, &key, active_node);
1265                         break;
1266                 case BTRFS_INODE_ITEM_KEY:
1267                         ret = process_inode_item(eb, i, &key, active_node);
1268                         break;
1269                 case BTRFS_EXTENT_DATA_KEY:
1270                         ret = process_file_extent(root, eb, i, &key,
1271                                                   active_node);
1272                         break;
1273                 default:
1274                         break;
1275                 };
1276                 if (ret != 0)
1277                         error = 1;
1278         }
1279         return error;
1280 }
1281
1282 static void reada_walk_down(struct btrfs_root *root,
1283                             struct extent_buffer *node, int slot)
1284 {
1285         u64 bytenr;
1286         u64 ptr_gen;
1287         u32 nritems;
1288         u32 blocksize;
1289         int i;
1290         int ret;
1291         int level;
1292
1293         level = btrfs_header_level(node);
1294         if (level != 1)
1295                 return;
1296
1297         nritems = btrfs_header_nritems(node);
1298         blocksize = btrfs_level_size(root, level - 1);
1299         for (i = slot; i < nritems; i++) {
1300                 bytenr = btrfs_node_blockptr(node, i);
1301                 ptr_gen = btrfs_node_ptr_generation(node, i);
1302                 ret = readahead_tree_block(root, bytenr, blocksize, ptr_gen);
1303                 if (ret)
1304                         break;
1305         }
1306 }
1307
1308 static int walk_down_tree(struct btrfs_root *root, struct btrfs_path *path,
1309                           struct walk_control *wc, int *level)
1310 {
1311         u64 bytenr;
1312         u64 ptr_gen;
1313         struct extent_buffer *next;
1314         struct extent_buffer *cur;
1315         u32 blocksize;
1316         int ret, err = 0;
1317         u64 refs;
1318
1319         WARN_ON(*level < 0);
1320         WARN_ON(*level >= BTRFS_MAX_LEVEL);
1321         ret = btrfs_lookup_extent_info(NULL, root,
1322                                        path->nodes[*level]->start,
1323                                        *level, 1, &refs, NULL);
1324         if (ret < 0) {
1325                 err = ret;
1326                 goto out;
1327         }
1328
1329         if (refs > 1) {
1330                 ret = enter_shared_node(root, path->nodes[*level]->start,
1331                                         refs, wc, *level);
1332                 if (ret > 0) {
1333                         err = ret;
1334                         goto out;
1335                 }
1336         }
1337
1338         while (*level >= 0) {
1339                 WARN_ON(*level < 0);
1340                 WARN_ON(*level >= BTRFS_MAX_LEVEL);
1341                 cur = path->nodes[*level];
1342
1343                 if (btrfs_header_level(cur) != *level)
1344                         WARN_ON(1);
1345
1346                 if (path->slots[*level] >= btrfs_header_nritems(cur))
1347                         break;
1348                 if (*level == 0) {
1349                         ret = process_one_leaf(root, cur, wc);
1350                         break;
1351                 }
1352                 bytenr = btrfs_node_blockptr(cur, path->slots[*level]);
1353                 ptr_gen = btrfs_node_ptr_generation(cur, path->slots[*level]);
1354                 blocksize = btrfs_level_size(root, *level - 1);
1355                 ret = btrfs_lookup_extent_info(NULL, root, bytenr, *level - 1,
1356                                                1, &refs, NULL);
1357                 if (ret < 0)
1358                         refs = 0;
1359
1360                 if (refs > 1) {
1361                         ret = enter_shared_node(root, bytenr, refs,
1362                                                 wc, *level - 1);
1363                         if (ret > 0) {
1364                                 path->slots[*level]++;
1365                                 continue;
1366                         }
1367                 }
1368
1369                 next = btrfs_find_tree_block(root, bytenr, blocksize);
1370                 if (!next || !btrfs_buffer_uptodate(next, ptr_gen)) {
1371                         free_extent_buffer(next);
1372                         reada_walk_down(root, cur, path->slots[*level]);
1373                         next = read_tree_block(root, bytenr, blocksize,
1374                                                ptr_gen);
1375                         if (!next) {
1376                                 err = -EIO;
1377                                 goto out;
1378                         }
1379                 }
1380
1381                 *level = *level - 1;
1382                 free_extent_buffer(path->nodes[*level]);
1383                 path->nodes[*level] = next;
1384                 path->slots[*level] = 0;
1385         }
1386 out:
1387         path->slots[*level] = btrfs_header_nritems(path->nodes[*level]);
1388         return err;
1389 }
1390
1391 static int walk_up_tree(struct btrfs_root *root, struct btrfs_path *path,
1392                         struct walk_control *wc, int *level)
1393 {
1394         int i;
1395         struct extent_buffer *leaf;
1396
1397         for (i = *level; i < BTRFS_MAX_LEVEL - 1 && path->nodes[i]; i++) {
1398                 leaf = path->nodes[i];
1399                 if (path->slots[i] + 1 < btrfs_header_nritems(leaf)) {
1400                         path->slots[i]++;
1401                         *level = i;
1402                         return 0;
1403                 } else {
1404                         free_extent_buffer(path->nodes[*level]);
1405                         path->nodes[*level] = NULL;
1406                         BUG_ON(*level > wc->active_node);
1407                         if (*level == wc->active_node)
1408                                 leave_shared_node(root, wc, *level);
1409                         *level = i + 1;
1410                 }
1411         }
1412         return 1;
1413 }
1414
1415 static int check_root_dir(struct inode_record *rec)
1416 {
1417         struct inode_backref *backref;
1418         int ret = -1;
1419
1420         if (!rec->found_inode_item || rec->errors)
1421                 goto out;
1422         if (rec->nlink != 1 || rec->found_link != 0)
1423                 goto out;
1424         if (list_empty(&rec->backrefs))
1425                 goto out;
1426         backref = list_entry(rec->backrefs.next, struct inode_backref, list);
1427         if (!backref->found_inode_ref)
1428                 goto out;
1429         if (backref->index != 0 || backref->namelen != 2 ||
1430             memcmp(backref->name, "..", 2))
1431                 goto out;
1432         if (backref->found_dir_index || backref->found_dir_item)
1433                 goto out;
1434         ret = 0;
1435 out:
1436         return ret;
1437 }
1438
1439 static int repair_inode_isize(struct btrfs_trans_handle *trans,
1440                               struct btrfs_root *root, struct btrfs_path *path,
1441                               struct inode_record *rec)
1442 {
1443         struct btrfs_inode_item *ei;
1444         struct btrfs_key key;
1445         int ret;
1446
1447         key.objectid = rec->ino;
1448         key.type = BTRFS_INODE_ITEM_KEY;
1449         key.offset = (u64)-1;
1450
1451         ret = btrfs_search_slot(trans, root, &key, path, 0, 1);
1452         if (ret < 0)
1453                 goto out;
1454         if (ret) {
1455                 if (!path->slots[0]) {
1456                         ret = -ENOENT;
1457                         goto out;
1458                 }
1459                 path->slots[0]--;
1460                 ret = 0;
1461         }
1462         btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
1463         if (key.objectid != rec->ino) {
1464                 ret = -ENOENT;
1465                 goto out;
1466         }
1467
1468         ei = btrfs_item_ptr(path->nodes[0], path->slots[0],
1469                             struct btrfs_inode_item);
1470         btrfs_set_inode_size(path->nodes[0], ei, rec->found_size);
1471         btrfs_mark_buffer_dirty(path->nodes[0]);
1472         rec->errors &= ~I_ERR_DIR_ISIZE_WRONG;
1473         printf("reset isize for dir %Lu root %Lu\n", rec->ino,
1474                root->root_key.objectid);
1475 out:
1476         btrfs_release_path(path);
1477         return ret;
1478 }
1479
1480 static int repair_inode_orphan_item(struct btrfs_trans_handle *trans,
1481                                     struct btrfs_root *root,
1482                                     struct btrfs_path *path,
1483                                     struct inode_record *rec)
1484 {
1485         struct btrfs_key key;
1486         int ret;
1487
1488         key.objectid = BTRFS_ORPHAN_OBJECTID;
1489         key.type = BTRFS_ORPHAN_ITEM_KEY;
1490         key.offset = rec->ino;
1491
1492         ret = btrfs_insert_empty_item(trans, root, path, &key, 0);
1493         btrfs_release_path(path);
1494         if (!ret)
1495                 rec->errors &= ~I_ERR_NO_ORPHAN_ITEM;
1496         return ret;
1497 }
1498
1499 static int try_repair_inode(struct btrfs_root *root, struct inode_record *rec)
1500 {
1501         struct btrfs_trans_handle *trans;
1502         struct btrfs_path *path;
1503         int ret = 0;
1504
1505         /* So far we just fix dir isize wrong */
1506         if (!(rec->errors & (I_ERR_DIR_ISIZE_WRONG | I_ERR_NO_ORPHAN_ITEM)))
1507                 return 1;
1508
1509         path = btrfs_alloc_path();
1510         if (!path)
1511                 return -ENOMEM;
1512
1513         trans = btrfs_start_transaction(root, 1);
1514         if (IS_ERR(trans)) {
1515                 btrfs_free_path(path);
1516                 return PTR_ERR(trans);
1517         }
1518
1519         if (rec->errors & I_ERR_DIR_ISIZE_WRONG)
1520                 ret = repair_inode_isize(trans, root, path, rec);
1521         if (!ret && rec->errors & I_ERR_NO_ORPHAN_ITEM)
1522                 ret = repair_inode_orphan_item(trans, root, path, rec);
1523         btrfs_commit_transaction(trans, root);
1524         btrfs_free_path(path);
1525         return ret;
1526 }
1527
1528 static int check_inode_recs(struct btrfs_root *root,
1529                             struct cache_tree *inode_cache)
1530 {
1531         struct cache_extent *cache;
1532         struct ptr_node *node;
1533         struct inode_record *rec;
1534         struct inode_backref *backref;
1535         int ret;
1536         u64 error = 0;
1537         u64 root_dirid = btrfs_root_dirid(&root->root_item);
1538
1539         if (btrfs_root_refs(&root->root_item) == 0) {
1540                 if (!cache_tree_empty(inode_cache))
1541                         fprintf(stderr, "warning line %d\n", __LINE__);
1542                 return 0;
1543         }
1544
1545         rec = get_inode_rec(inode_cache, root_dirid, 0);
1546         if (rec) {
1547                 ret = check_root_dir(rec);
1548                 if (ret) {
1549                         fprintf(stderr, "root %llu root dir %llu error\n",
1550                                 (unsigned long long)root->root_key.objectid,
1551                                 (unsigned long long)root_dirid);
1552                         error++;
1553                 }
1554         } else {
1555                 fprintf(stderr, "root %llu root dir %llu not found\n",
1556                         (unsigned long long)root->root_key.objectid,
1557                         (unsigned long long)root_dirid);
1558         }
1559
1560         while (1) {
1561                 cache = search_cache_extent(inode_cache, 0);
1562                 if (!cache)
1563                         break;
1564                 node = container_of(cache, struct ptr_node, cache);
1565                 rec = node->data;
1566                 remove_cache_extent(inode_cache, &node->cache);
1567                 free(node);
1568                 if (rec->ino == root_dirid ||
1569                     rec->ino == BTRFS_ORPHAN_OBJECTID) {
1570                         free_inode_rec(rec);
1571                         continue;
1572                 }
1573
1574                 if (rec->errors & I_ERR_NO_ORPHAN_ITEM) {
1575                         ret = check_orphan_item(root, rec->ino);
1576                         if (ret == 0)
1577                                 rec->errors &= ~I_ERR_NO_ORPHAN_ITEM;
1578                         if (can_free_inode_rec(rec)) {
1579                                 free_inode_rec(rec);
1580                                 continue;
1581                         }
1582                 }
1583
1584                 if (repair) {
1585                         ret = try_repair_inode(root, rec);
1586                         if (ret == 0 && can_free_inode_rec(rec)) {
1587                                 free_inode_rec(rec);
1588                                 continue;
1589                         }
1590                         ret = 0;
1591                 }
1592
1593                 error++;
1594                 if (!rec->found_inode_item)
1595                         rec->errors |= I_ERR_NO_INODE_ITEM;
1596                 if (rec->found_link != rec->nlink)
1597                         rec->errors |= I_ERR_LINK_COUNT_WRONG;
1598                 fprintf(stderr, "root %llu inode %llu errors %x",
1599                         (unsigned long long) root->root_key.objectid,
1600                         (unsigned long long) rec->ino, rec->errors);
1601                 print_inode_error(rec->errors);
1602                 list_for_each_entry(backref, &rec->backrefs, list) {
1603                         if (!backref->found_dir_item)
1604                                 backref->errors |= REF_ERR_NO_DIR_ITEM;
1605                         if (!backref->found_dir_index)
1606                                 backref->errors |= REF_ERR_NO_DIR_INDEX;
1607                         if (!backref->found_inode_ref)
1608                                 backref->errors |= REF_ERR_NO_INODE_REF;
1609                         fprintf(stderr, "\tunresolved ref dir %llu index %llu"
1610                                 " namelen %u name %s filetype %d errors %x",
1611                                 (unsigned long long)backref->dir,
1612                                 (unsigned long long)backref->index,
1613                                 backref->namelen, backref->name,
1614                                 backref->filetype, backref->errors);
1615                         print_ref_error(backref->errors);
1616                 }
1617                 free_inode_rec(rec);
1618         }
1619         return (error > 0) ? -1 : 0;
1620 }
1621
1622 static struct root_record *get_root_rec(struct cache_tree *root_cache,
1623                                         u64 objectid)
1624 {
1625         struct cache_extent *cache;
1626         struct root_record *rec = NULL;
1627         int ret;
1628
1629         cache = lookup_cache_extent(root_cache, objectid, 1);
1630         if (cache) {
1631                 rec = container_of(cache, struct root_record, cache);
1632         } else {
1633                 rec = calloc(1, sizeof(*rec));
1634                 rec->objectid = objectid;
1635                 INIT_LIST_HEAD(&rec->backrefs);
1636                 rec->cache.start = objectid;
1637                 rec->cache.size = 1;
1638
1639                 ret = insert_cache_extent(root_cache, &rec->cache);
1640                 BUG_ON(ret);
1641         }
1642         return rec;
1643 }
1644
1645 static struct root_backref *get_root_backref(struct root_record *rec,
1646                                              u64 ref_root, u64 dir, u64 index,
1647                                              const char *name, int namelen)
1648 {
1649         struct root_backref *backref;
1650
1651         list_for_each_entry(backref, &rec->backrefs, list) {
1652                 if (backref->ref_root != ref_root || backref->dir != dir ||
1653                     backref->namelen != namelen)
1654                         continue;
1655                 if (memcmp(name, backref->name, namelen))
1656                         continue;
1657                 return backref;
1658         }
1659
1660         backref = malloc(sizeof(*backref) + namelen + 1);
1661         memset(backref, 0, sizeof(*backref));
1662         backref->ref_root = ref_root;
1663         backref->dir = dir;
1664         backref->index = index;
1665         backref->namelen = namelen;
1666         memcpy(backref->name, name, namelen);
1667         backref->name[namelen] = '\0';
1668         list_add_tail(&backref->list, &rec->backrefs);
1669         return backref;
1670 }
1671
1672 static void free_root_record(struct cache_extent *cache)
1673 {
1674         struct root_record *rec;
1675         struct root_backref *backref;
1676
1677         rec = container_of(cache, struct root_record, cache);
1678         while (!list_empty(&rec->backrefs)) {
1679                 backref = list_entry(rec->backrefs.next,
1680                                      struct root_backref, list);
1681                 list_del(&backref->list);
1682                 free(backref);
1683         }
1684
1685         kfree(rec);
1686 }
1687
1688 FREE_EXTENT_CACHE_BASED_TREE(root_recs, free_root_record);
1689
1690 static int add_root_backref(struct cache_tree *root_cache,
1691                             u64 root_id, u64 ref_root, u64 dir, u64 index,
1692                             const char *name, int namelen,
1693                             int item_type, int errors)
1694 {
1695         struct root_record *rec;
1696         struct root_backref *backref;
1697
1698         rec = get_root_rec(root_cache, root_id);
1699         backref = get_root_backref(rec, ref_root, dir, index, name, namelen);
1700
1701         backref->errors |= errors;
1702
1703         if (item_type != BTRFS_DIR_ITEM_KEY) {
1704                 if (backref->found_dir_index || backref->found_back_ref ||
1705                     backref->found_forward_ref) {
1706                         if (backref->index != index)
1707                                 backref->errors |= REF_ERR_INDEX_UNMATCH;
1708                 } else {
1709                         backref->index = index;
1710                 }
1711         }
1712
1713         if (item_type == BTRFS_DIR_ITEM_KEY) {
1714                 if (backref->found_forward_ref)
1715                         rec->found_ref++;
1716                 backref->found_dir_item = 1;
1717         } else if (item_type == BTRFS_DIR_INDEX_KEY) {
1718                 backref->found_dir_index = 1;
1719         } else if (item_type == BTRFS_ROOT_REF_KEY) {
1720                 if (backref->found_forward_ref)
1721                         backref->errors |= REF_ERR_DUP_ROOT_REF;
1722                 else if (backref->found_dir_item)
1723                         rec->found_ref++;
1724                 backref->found_forward_ref = 1;
1725         } else if (item_type == BTRFS_ROOT_BACKREF_KEY) {
1726                 if (backref->found_back_ref)
1727                         backref->errors |= REF_ERR_DUP_ROOT_BACKREF;
1728                 backref->found_back_ref = 1;
1729         } else {
1730                 BUG_ON(1);
1731         }
1732
1733         if (backref->found_forward_ref && backref->found_dir_item)
1734                 backref->reachable = 1;
1735         return 0;
1736 }
1737
1738 static int merge_root_recs(struct btrfs_root *root,
1739                            struct cache_tree *src_cache,
1740                            struct cache_tree *dst_cache)
1741 {
1742         struct cache_extent *cache;
1743         struct ptr_node *node;
1744         struct inode_record *rec;
1745         struct inode_backref *backref;
1746
1747         if (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID) {
1748                 free_inode_recs_tree(src_cache);
1749                 return 0;
1750         }
1751
1752         while (1) {
1753                 cache = search_cache_extent(src_cache, 0);
1754                 if (!cache)
1755                         break;
1756                 node = container_of(cache, struct ptr_node, cache);
1757                 rec = node->data;
1758                 remove_cache_extent(src_cache, &node->cache);
1759                 free(node);
1760
1761                 if (!is_child_root(root, root->objectid, rec->ino))
1762                         goto skip;
1763
1764                 list_for_each_entry(backref, &rec->backrefs, list) {
1765                         BUG_ON(backref->found_inode_ref);
1766                         if (backref->found_dir_item)
1767                                 add_root_backref(dst_cache, rec->ino,
1768                                         root->root_key.objectid, backref->dir,
1769                                         backref->index, backref->name,
1770                                         backref->namelen, BTRFS_DIR_ITEM_KEY,
1771                                         backref->errors);
1772                         if (backref->found_dir_index)
1773                                 add_root_backref(dst_cache, rec->ino,
1774                                         root->root_key.objectid, backref->dir,
1775                                         backref->index, backref->name,
1776                                         backref->namelen, BTRFS_DIR_INDEX_KEY,
1777                                         backref->errors);
1778                 }
1779 skip:
1780                 free_inode_rec(rec);
1781         }
1782         return 0;
1783 }
1784
1785 static int check_root_refs(struct btrfs_root *root,
1786                            struct cache_tree *root_cache)
1787 {
1788         struct root_record *rec;
1789         struct root_record *ref_root;
1790         struct root_backref *backref;
1791         struct cache_extent *cache;
1792         int loop = 1;
1793         int ret;
1794         int error;
1795         int errors = 0;
1796
1797         rec = get_root_rec(root_cache, BTRFS_FS_TREE_OBJECTID);
1798         rec->found_ref = 1;
1799
1800         /* fixme: this can not detect circular references */
1801         while (loop) {
1802                 loop = 0;
1803                 cache = search_cache_extent(root_cache, 0);
1804                 while (1) {
1805                         if (!cache)
1806                                 break;
1807                         rec = container_of(cache, struct root_record, cache);
1808                         cache = next_cache_extent(cache);
1809
1810                         if (rec->found_ref == 0)
1811                                 continue;
1812
1813                         list_for_each_entry(backref, &rec->backrefs, list) {
1814                                 if (!backref->reachable)
1815                                         continue;
1816
1817                                 ref_root = get_root_rec(root_cache,
1818                                                         backref->ref_root);
1819                                 if (ref_root->found_ref > 0)
1820                                         continue;
1821
1822                                 backref->reachable = 0;
1823                                 rec->found_ref--;
1824                                 if (rec->found_ref == 0)
1825                                         loop = 1;
1826                         }
1827                 }
1828         }
1829
1830         cache = search_cache_extent(root_cache, 0);
1831         while (1) {
1832                 if (!cache)
1833                         break;
1834                 rec = container_of(cache, struct root_record, cache);
1835                 cache = next_cache_extent(cache);
1836
1837                 if (rec->found_ref == 0 &&
1838                     rec->objectid >= BTRFS_FIRST_FREE_OBJECTID &&
1839                     rec->objectid <= BTRFS_LAST_FREE_OBJECTID) {
1840                         ret = check_orphan_item(root->fs_info->tree_root,
1841                                                 rec->objectid);
1842                         if (ret == 0)
1843                                 continue;
1844
1845                         /*
1846                          * If we don't have a root item then we likely just have
1847                          * a dir item in a snapshot for this root but no actual
1848                          * ref key or anything so it's meaningless.
1849                          */
1850                         if (!rec->found_root_item)
1851                                 continue;
1852                         errors++;
1853                         fprintf(stderr, "fs tree %llu not referenced\n",
1854                                 (unsigned long long)rec->objectid);
1855                 }
1856
1857                 error = 0;
1858                 if (rec->found_ref > 0 && !rec->found_root_item)
1859                         error = 1;
1860                 list_for_each_entry(backref, &rec->backrefs, list) {
1861                         if (!backref->found_dir_item)
1862                                 backref->errors |= REF_ERR_NO_DIR_ITEM;
1863                         if (!backref->found_dir_index)
1864                                 backref->errors |= REF_ERR_NO_DIR_INDEX;
1865                         if (!backref->found_back_ref)
1866                                 backref->errors |= REF_ERR_NO_ROOT_BACKREF;
1867                         if (!backref->found_forward_ref)
1868                                 backref->errors |= REF_ERR_NO_ROOT_REF;
1869                         if (backref->reachable && backref->errors)
1870                                 error = 1;
1871                 }
1872                 if (!error)
1873                         continue;
1874
1875                 errors++;
1876                 fprintf(stderr, "fs tree %llu refs %u %s\n",
1877                         (unsigned long long)rec->objectid, rec->found_ref,
1878                          rec->found_root_item ? "" : "not found");
1879
1880                 list_for_each_entry(backref, &rec->backrefs, list) {
1881                         if (!backref->reachable)
1882                                 continue;
1883                         if (!backref->errors && rec->found_root_item)
1884                                 continue;
1885                         fprintf(stderr, "\tunresolved ref root %llu dir %llu"
1886                                 " index %llu namelen %u name %s errors %x\n",
1887                                 (unsigned long long)backref->ref_root,
1888                                 (unsigned long long)backref->dir,
1889                                 (unsigned long long)backref->index,
1890                                 backref->namelen, backref->name,
1891                                 backref->errors);
1892                         print_ref_error(backref->errors);
1893                 }
1894         }
1895         return errors > 0 ? 1 : 0;
1896 }
1897
1898 static int process_root_ref(struct extent_buffer *eb, int slot,
1899                             struct btrfs_key *key,
1900                             struct cache_tree *root_cache)
1901 {
1902         u64 dirid;
1903         u64 index;
1904         u32 len;
1905         u32 name_len;
1906         struct btrfs_root_ref *ref;
1907         char namebuf[BTRFS_NAME_LEN];
1908         int error;
1909
1910         ref = btrfs_item_ptr(eb, slot, struct btrfs_root_ref);
1911
1912         dirid = btrfs_root_ref_dirid(eb, ref);
1913         index = btrfs_root_ref_sequence(eb, ref);
1914         name_len = btrfs_root_ref_name_len(eb, ref);
1915
1916         if (name_len <= BTRFS_NAME_LEN) {
1917                 len = name_len;
1918                 error = 0;
1919         } else {
1920                 len = BTRFS_NAME_LEN;
1921                 error = REF_ERR_NAME_TOO_LONG;
1922         }
1923         read_extent_buffer(eb, namebuf, (unsigned long)(ref + 1), len);
1924
1925         if (key->type == BTRFS_ROOT_REF_KEY) {
1926                 add_root_backref(root_cache, key->offset, key->objectid, dirid,
1927                                  index, namebuf, len, key->type, error);
1928         } else {
1929                 add_root_backref(root_cache, key->objectid, key->offset, dirid,
1930                                  index, namebuf, len, key->type, error);
1931         }
1932         return 0;
1933 }
1934
1935 static int check_fs_root(struct btrfs_root *root,
1936                          struct cache_tree *root_cache,
1937                          struct walk_control *wc)
1938 {
1939         int ret = 0;
1940         int wret;
1941         int level;
1942         struct btrfs_path path;
1943         struct shared_node root_node;
1944         struct root_record *rec;
1945         struct btrfs_root_item *root_item = &root->root_item;
1946
1947         if (root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID) {
1948                 rec = get_root_rec(root_cache, root->root_key.objectid);
1949                 if (btrfs_root_refs(root_item) > 0)
1950                         rec->found_root_item = 1;
1951         }
1952
1953         btrfs_init_path(&path);
1954         memset(&root_node, 0, sizeof(root_node));
1955         cache_tree_init(&root_node.root_cache);
1956         cache_tree_init(&root_node.inode_cache);
1957
1958         level = btrfs_header_level(root->node);
1959         memset(wc->nodes, 0, sizeof(wc->nodes));
1960         wc->nodes[level] = &root_node;
1961         wc->active_node = level;
1962         wc->root_level = level;
1963
1964         if (btrfs_root_refs(root_item) > 0 ||
1965             btrfs_disk_key_objectid(&root_item->drop_progress) == 0) {
1966                 path.nodes[level] = root->node;
1967                 extent_buffer_get(root->node);
1968                 path.slots[level] = 0;
1969         } else {
1970                 struct btrfs_key key;
1971                 struct btrfs_disk_key found_key;
1972
1973                 btrfs_disk_key_to_cpu(&key, &root_item->drop_progress);
1974                 level = root_item->drop_level;
1975                 path.lowest_level = level;
1976                 wret = btrfs_search_slot(NULL, root, &key, &path, 0, 0);
1977                 BUG_ON(wret < 0);
1978                 btrfs_node_key(path.nodes[level], &found_key,
1979                                 path.slots[level]);
1980                 WARN_ON(memcmp(&found_key, &root_item->drop_progress,
1981                                         sizeof(found_key)));
1982         }
1983
1984         while (1) {
1985                 wret = walk_down_tree(root, &path, wc, &level);
1986                 if (wret < 0)
1987                         ret = wret;
1988                 if (wret != 0)
1989                         break;
1990
1991                 wret = walk_up_tree(root, &path, wc, &level);
1992                 if (wret < 0)
1993                         ret = wret;
1994                 if (wret != 0)
1995                         break;
1996         }
1997         btrfs_release_path(&path);
1998
1999         merge_root_recs(root, &root_node.root_cache, root_cache);
2000
2001         if (root_node.current) {
2002                 root_node.current->checked = 1;
2003                 maybe_free_inode_rec(&root_node.inode_cache,
2004                                 root_node.current);
2005         }
2006
2007         ret = check_inode_recs(root, &root_node.inode_cache);
2008         return ret;
2009 }
2010
2011 static int fs_root_objectid(u64 objectid)
2012 {
2013         if (objectid == BTRFS_FS_TREE_OBJECTID ||
2014             objectid == BTRFS_TREE_RELOC_OBJECTID ||
2015             objectid == BTRFS_DATA_RELOC_TREE_OBJECTID ||
2016             (objectid >= BTRFS_FIRST_FREE_OBJECTID &&
2017              objectid <= BTRFS_LAST_FREE_OBJECTID))
2018                 return 1;
2019         return 0;
2020 }
2021
2022 static int check_fs_roots(struct btrfs_root *root,
2023                           struct cache_tree *root_cache)
2024 {
2025         struct btrfs_path path;
2026         struct btrfs_key key;
2027         struct walk_control wc;
2028         struct extent_buffer *leaf;
2029         struct btrfs_root *tmp_root;
2030         struct btrfs_root *tree_root = root->fs_info->tree_root;
2031         int ret;
2032         int err = 0;
2033
2034         /*
2035          * Just in case we made any changes to the extent tree that weren't
2036          * reflected into the free space cache yet.
2037          */
2038         if (repair)
2039                 reset_cached_block_groups(root->fs_info);
2040         memset(&wc, 0, sizeof(wc));
2041         cache_tree_init(&wc.shared);
2042         btrfs_init_path(&path);
2043
2044         key.offset = 0;
2045         key.objectid = 0;
2046         key.type = BTRFS_ROOT_ITEM_KEY;
2047         ret = btrfs_search_slot(NULL, tree_root, &key, &path, 0, 0);
2048         BUG_ON(ret < 0);
2049         while (1) {
2050                 leaf = path.nodes[0];
2051                 if (path.slots[0] >= btrfs_header_nritems(leaf)) {
2052                         ret = btrfs_next_leaf(tree_root, &path);
2053                         if (ret != 0)
2054                                 break;
2055                         leaf = path.nodes[0];
2056                 }
2057                 btrfs_item_key_to_cpu(leaf, &key, path.slots[0]);
2058                 if (key.type == BTRFS_ROOT_ITEM_KEY &&
2059                     fs_root_objectid(key.objectid)) {
2060                         key.offset = (u64)-1;
2061                         tmp_root = btrfs_read_fs_root(root->fs_info, &key);
2062                         if (IS_ERR(tmp_root)) {
2063                                 err = 1;
2064                                 goto next;
2065                         }
2066                         ret = check_fs_root(tmp_root, root_cache, &wc);
2067                         if (ret)
2068                                 err = 1;
2069                 } else if (key.type == BTRFS_ROOT_REF_KEY ||
2070                            key.type == BTRFS_ROOT_BACKREF_KEY) {
2071                         process_root_ref(leaf, path.slots[0], &key,
2072                                          root_cache);
2073                 }
2074 next:
2075                 path.slots[0]++;
2076         }
2077         btrfs_release_path(&path);
2078
2079         if (!cache_tree_empty(&wc.shared))
2080                 fprintf(stderr, "warning line %d\n", __LINE__);
2081
2082         return err;
2083 }
2084
2085 static int all_backpointers_checked(struct extent_record *rec, int print_errs)
2086 {
2087         struct list_head *cur = rec->backrefs.next;
2088         struct extent_backref *back;
2089         struct tree_backref *tback;
2090         struct data_backref *dback;
2091         u64 found = 0;
2092         int err = 0;
2093
2094         while(cur != &rec->backrefs) {
2095                 back = list_entry(cur, struct extent_backref, list);
2096                 cur = cur->next;
2097                 if (!back->found_extent_tree) {
2098                         err = 1;
2099                         if (!print_errs)
2100                                 goto out;
2101                         if (back->is_data) {
2102                                 dback = (struct data_backref *)back;
2103                                 fprintf(stderr, "Backref %llu %s %llu"
2104                                         " owner %llu offset %llu num_refs %lu"
2105                                         " not found in extent tree\n",
2106                                         (unsigned long long)rec->start,
2107                                         back->full_backref ?
2108                                         "parent" : "root",
2109                                         back->full_backref ?
2110                                         (unsigned long long)dback->parent:
2111                                         (unsigned long long)dback->root,
2112                                         (unsigned long long)dback->owner,
2113                                         (unsigned long long)dback->offset,
2114                                         (unsigned long)dback->num_refs);
2115                         } else {
2116                                 tback = (struct tree_backref *)back;
2117                                 fprintf(stderr, "Backref %llu parent %llu"
2118                                         " root %llu not found in extent tree\n",
2119                                         (unsigned long long)rec->start,
2120                                         (unsigned long long)tback->parent,
2121                                         (unsigned long long)tback->root);
2122                         }
2123                 }
2124                 if (!back->is_data && !back->found_ref) {
2125                         err = 1;
2126                         if (!print_errs)
2127                                 goto out;
2128                         tback = (struct tree_backref *)back;
2129                         fprintf(stderr, "Backref %llu %s %llu not referenced back %p\n",
2130                                 (unsigned long long)rec->start,
2131                                 back->full_backref ? "parent" : "root",
2132                                 back->full_backref ?
2133                                 (unsigned long long)tback->parent :
2134                                 (unsigned long long)tback->root, back);
2135                 }
2136                 if (back->is_data) {
2137                         dback = (struct data_backref *)back;
2138                         if (dback->found_ref != dback->num_refs) {
2139                                 err = 1;
2140                                 if (!print_errs)
2141                                         goto out;
2142                                 fprintf(stderr, "Incorrect local backref count"
2143                                         " on %llu %s %llu owner %llu"
2144                                         " offset %llu found %u wanted %u back %p\n",
2145                                         (unsigned long long)rec->start,
2146                                         back->full_backref ?
2147                                         "parent" : "root",
2148                                         back->full_backref ?
2149                                         (unsigned long long)dback->parent:
2150                                         (unsigned long long)dback->root,
2151                                         (unsigned long long)dback->owner,
2152                                         (unsigned long long)dback->offset,
2153                                         dback->found_ref, dback->num_refs, back);
2154                         }
2155                         if (dback->disk_bytenr != rec->start) {
2156                                 err = 1;
2157                                 if (!print_errs)
2158                                         goto out;
2159                                 fprintf(stderr, "Backref disk bytenr does not"
2160                                         " match extent record, bytenr=%llu, "
2161                                         "ref bytenr=%llu\n",
2162                                         (unsigned long long)rec->start,
2163                                         (unsigned long long)dback->disk_bytenr);
2164                         }
2165
2166                         if (dback->bytes != rec->nr) {
2167                                 err = 1;
2168                                 if (!print_errs)
2169                                         goto out;
2170                                 fprintf(stderr, "Backref bytes do not match "
2171                                         "extent backref, bytenr=%llu, ref "
2172                                         "bytes=%llu, backref bytes=%llu\n",
2173                                         (unsigned long long)rec->start,
2174                                         (unsigned long long)rec->nr,
2175                                         (unsigned long long)dback->bytes);
2176                         }
2177                 }
2178                 if (!back->is_data) {
2179                         found += 1;
2180                 } else {
2181                         dback = (struct data_backref *)back;
2182                         found += dback->found_ref;
2183                 }
2184         }
2185         if (found != rec->refs) {
2186                 err = 1;
2187                 if (!print_errs)
2188                         goto out;
2189                 fprintf(stderr, "Incorrect global backref count "
2190                         "on %llu found %llu wanted %llu\n",
2191                         (unsigned long long)rec->start,
2192                         (unsigned long long)found,
2193                         (unsigned long long)rec->refs);
2194         }
2195 out:
2196         return err;
2197 }
2198
2199 static int free_all_extent_backrefs(struct extent_record *rec)
2200 {
2201         struct extent_backref *back;
2202         struct list_head *cur;
2203         while (!list_empty(&rec->backrefs)) {
2204                 cur = rec->backrefs.next;
2205                 back = list_entry(cur, struct extent_backref, list);
2206                 list_del(cur);
2207                 free(back);
2208         }
2209         return 0;
2210 }
2211
2212 static void free_extent_record_cache(struct btrfs_fs_info *fs_info,
2213                                      struct cache_tree *extent_cache)
2214 {
2215         struct cache_extent *cache;
2216         struct extent_record *rec;
2217
2218         while (1) {
2219                 cache = first_cache_extent(extent_cache);
2220                 if (!cache)
2221                         break;
2222                 rec = container_of(cache, struct extent_record, cache);
2223                 btrfs_unpin_extent(fs_info, rec->start, rec->max_size);
2224                 remove_cache_extent(extent_cache, cache);
2225                 free_all_extent_backrefs(rec);
2226                 free(rec);
2227         }
2228 }
2229
2230 static int maybe_free_extent_rec(struct cache_tree *extent_cache,
2231                                  struct extent_record *rec)
2232 {
2233         if (rec->content_checked && rec->owner_ref_checked &&
2234             rec->extent_item_refs == rec->refs && rec->refs > 0 &&
2235             rec->num_duplicates == 0 && !all_backpointers_checked(rec, 0)) {
2236                 remove_cache_extent(extent_cache, &rec->cache);
2237                 free_all_extent_backrefs(rec);
2238                 list_del_init(&rec->list);
2239                 free(rec);
2240         }
2241         return 0;
2242 }
2243
2244 static int check_owner_ref(struct btrfs_root *root,
2245                             struct extent_record *rec,
2246                             struct extent_buffer *buf)
2247 {
2248         struct extent_backref *node;
2249         struct tree_backref *back;
2250         struct btrfs_root *ref_root;
2251         struct btrfs_key key;
2252         struct btrfs_path path;
2253         struct extent_buffer *parent;
2254         int level;
2255         int found = 0;
2256         int ret;
2257
2258         list_for_each_entry(node, &rec->backrefs, list) {
2259                 if (node->is_data)
2260                         continue;
2261                 if (!node->found_ref)
2262                         continue;
2263                 if (node->full_backref)
2264                         continue;
2265                 back = (struct tree_backref *)node;
2266                 if (btrfs_header_owner(buf) == back->root)
2267                         return 0;
2268         }
2269         BUG_ON(rec->is_root);
2270
2271         /* try to find the block by search corresponding fs tree */
2272         key.objectid = btrfs_header_owner(buf);
2273         key.type = BTRFS_ROOT_ITEM_KEY;
2274         key.offset = (u64)-1;
2275
2276         ref_root = btrfs_read_fs_root(root->fs_info, &key);
2277         if (IS_ERR(ref_root))
2278                 return 1;
2279
2280         level = btrfs_header_level(buf);
2281         if (level == 0)
2282                 btrfs_item_key_to_cpu(buf, &key, 0);
2283         else
2284                 btrfs_node_key_to_cpu(buf, &key, 0);
2285
2286         btrfs_init_path(&path);
2287         path.lowest_level = level + 1;
2288         ret = btrfs_search_slot(NULL, ref_root, &key, &path, 0, 0);
2289         if (ret < 0)
2290                 return 0;
2291
2292         parent = path.nodes[level + 1];
2293         if (parent && buf->start == btrfs_node_blockptr(parent,
2294                                                         path.slots[level + 1]))
2295                 found = 1;
2296
2297         btrfs_release_path(&path);
2298         return found ? 0 : 1;
2299 }
2300
2301 static int is_extent_tree_record(struct extent_record *rec)
2302 {
2303         struct list_head *cur = rec->backrefs.next;
2304         struct extent_backref *node;
2305         struct tree_backref *back;
2306         int is_extent = 0;
2307
2308         while(cur != &rec->backrefs) {
2309                 node = list_entry(cur, struct extent_backref, list);
2310                 cur = cur->next;
2311                 if (node->is_data)
2312                         return 0;
2313                 back = (struct tree_backref *)node;
2314                 if (node->full_backref)
2315                         return 0;
2316                 if (back->root == BTRFS_EXTENT_TREE_OBJECTID)
2317                         is_extent = 1;
2318         }
2319         return is_extent;
2320 }
2321
2322
2323 static int record_bad_block_io(struct btrfs_fs_info *info,
2324                                struct cache_tree *extent_cache,
2325                                u64 start, u64 len)
2326 {
2327         struct extent_record *rec;
2328         struct cache_extent *cache;
2329         struct btrfs_key key;
2330
2331         cache = lookup_cache_extent(extent_cache, start, len);
2332         if (!cache)
2333                 return 0;
2334
2335         rec = container_of(cache, struct extent_record, cache);
2336         if (!is_extent_tree_record(rec))
2337                 return 0;
2338
2339         btrfs_disk_key_to_cpu(&key, &rec->parent_key);
2340         return btrfs_add_corrupt_extent_record(info, &key, start, len, 0);
2341 }
2342
2343 static int swap_values(struct btrfs_root *root, struct btrfs_path *path,
2344                        struct extent_buffer *buf, int slot)
2345 {
2346         if (btrfs_header_level(buf)) {
2347                 struct btrfs_key_ptr ptr1, ptr2;
2348
2349                 read_extent_buffer(buf, &ptr1, btrfs_node_key_ptr_offset(slot),
2350                                    sizeof(struct btrfs_key_ptr));
2351                 read_extent_buffer(buf, &ptr2,
2352                                    btrfs_node_key_ptr_offset(slot + 1),
2353                                    sizeof(struct btrfs_key_ptr));
2354                 write_extent_buffer(buf, &ptr1,
2355                                     btrfs_node_key_ptr_offset(slot + 1),
2356                                     sizeof(struct btrfs_key_ptr));
2357                 write_extent_buffer(buf, &ptr2,
2358                                     btrfs_node_key_ptr_offset(slot),
2359                                     sizeof(struct btrfs_key_ptr));
2360                 if (slot == 0) {
2361                         struct btrfs_disk_key key;
2362                         btrfs_node_key(buf, &key, 0);
2363                         btrfs_fixup_low_keys(root, path, &key,
2364                                              btrfs_header_level(buf) + 1);
2365                 }
2366         } else {
2367                 struct btrfs_item *item1, *item2;
2368                 struct btrfs_key k1, k2;
2369                 char *item1_data, *item2_data;
2370                 u32 item1_offset, item2_offset, item1_size, item2_size;
2371
2372                 item1 = btrfs_item_nr(slot);
2373                 item2 = btrfs_item_nr(slot + 1);
2374                 btrfs_item_key_to_cpu(buf, &k1, slot);
2375                 btrfs_item_key_to_cpu(buf, &k2, slot + 1);
2376                 item1_offset = btrfs_item_offset(buf, item1);
2377                 item2_offset = btrfs_item_offset(buf, item2);
2378                 item1_size = btrfs_item_size(buf, item1);
2379                 item2_size = btrfs_item_size(buf, item2);
2380
2381                 item1_data = malloc(item1_size);
2382                 if (!item1_data)
2383                         return -ENOMEM;
2384                 item2_data = malloc(item2_size);
2385                 if (!item2_data) {
2386                         free(item1_data);
2387                         return -ENOMEM;
2388                 }
2389
2390                 read_extent_buffer(buf, item1_data, item1_offset, item1_size);
2391                 read_extent_buffer(buf, item2_data, item2_offset, item2_size);
2392
2393                 write_extent_buffer(buf, item1_data, item2_offset, item2_size);
2394                 write_extent_buffer(buf, item2_data, item1_offset, item1_size);
2395                 free(item1_data);
2396                 free(item2_data);
2397
2398                 btrfs_set_item_offset(buf, item1, item2_offset);
2399                 btrfs_set_item_offset(buf, item2, item1_offset);
2400                 btrfs_set_item_size(buf, item1, item2_size);
2401                 btrfs_set_item_size(buf, item2, item1_size);
2402
2403                 path->slots[0] = slot;
2404                 btrfs_set_item_key_unsafe(root, path, &k2);
2405                 path->slots[0] = slot + 1;
2406                 btrfs_set_item_key_unsafe(root, path, &k1);
2407         }
2408         return 0;
2409 }
2410
2411 /*
2412  * Attempt to fix basic block failures.  Currently we only handle bad key
2413  * orders, we will cycle through the keys and swap them if necessary.
2414  */
2415 static int try_to_fix_bad_block(struct btrfs_trans_handle *trans,
2416                                 struct btrfs_root *root,
2417                                 struct extent_buffer *buf,
2418                                 struct btrfs_disk_key *parent_key,
2419                                 enum btrfs_tree_block_status status)
2420 {
2421         struct btrfs_path *path;
2422         struct btrfs_key k1, k2;
2423         int i;
2424         int ret;
2425
2426         if (status != BTRFS_TREE_BLOCK_BAD_KEY_ORDER)
2427                 return -EIO;
2428
2429         k1.objectid = btrfs_header_owner(buf);
2430         k1.type = BTRFS_ROOT_ITEM_KEY;
2431         k1.offset = (u64)-1;
2432
2433         root = btrfs_read_fs_root(root->fs_info, &k1);
2434         if (IS_ERR(root))
2435                 return -EIO;
2436
2437         path = btrfs_alloc_path();
2438         if (!path)
2439                 return -EIO;
2440
2441         path->lowest_level = btrfs_header_level(buf);
2442         path->skip_check_block = 1;
2443         if (btrfs_header_level(buf))
2444                 btrfs_node_key_to_cpu(buf, &k1, 0);
2445         else
2446                 btrfs_item_key_to_cpu(buf, &k1, 0);
2447
2448         ret = btrfs_search_slot(trans, root, &k1, path, 0, 1);
2449         if (ret) {
2450                 btrfs_free_path(path);
2451                 return -EIO;
2452         }
2453
2454         buf = path->nodes[0];
2455         for (i = 0; i < btrfs_header_nritems(buf) - 1; i++) {
2456                 if (btrfs_header_level(buf)) {
2457                         btrfs_node_key_to_cpu(buf, &k1, i);
2458                         btrfs_node_key_to_cpu(buf, &k2, i + 1);
2459                 } else {
2460                         btrfs_item_key_to_cpu(buf, &k1, i);
2461                         btrfs_item_key_to_cpu(buf, &k2, i + 1);
2462                 }
2463                 if (btrfs_comp_cpu_keys(&k1, &k2) < 0)
2464                         continue;
2465                 ret = swap_values(root, path, buf, i);
2466                 if (ret)
2467                         break;
2468                 btrfs_mark_buffer_dirty(buf);
2469                 i = 0;
2470         }
2471
2472         btrfs_free_path(path);
2473         return ret;
2474 }
2475
2476 static int check_block(struct btrfs_trans_handle *trans,
2477                        struct btrfs_root *root,
2478                        struct cache_tree *extent_cache,
2479                        struct extent_buffer *buf, u64 flags)
2480 {
2481         struct extent_record *rec;
2482         struct cache_extent *cache;
2483         struct btrfs_key key;
2484         enum btrfs_tree_block_status status;
2485         int ret = 0;
2486         int level;
2487
2488         cache = lookup_cache_extent(extent_cache, buf->start, buf->len);
2489         if (!cache)
2490                 return 1;
2491         rec = container_of(cache, struct extent_record, cache);
2492         rec->generation = btrfs_header_generation(buf);
2493
2494         level = btrfs_header_level(buf);
2495         if (btrfs_header_nritems(buf) > 0) {
2496
2497                 if (level == 0)
2498                         btrfs_item_key_to_cpu(buf, &key, 0);
2499                 else
2500                         btrfs_node_key_to_cpu(buf, &key, 0);
2501
2502                 rec->info_objectid = key.objectid;
2503         }
2504         rec->info_level = level;
2505
2506         if (btrfs_is_leaf(buf))
2507                 status = btrfs_check_leaf(root, &rec->parent_key, buf);
2508         else
2509                 status = btrfs_check_node(root, &rec->parent_key, buf);
2510
2511         if (status != BTRFS_TREE_BLOCK_CLEAN) {
2512                 if (repair)
2513                         status = try_to_fix_bad_block(trans, root, buf,
2514                                                       &rec->parent_key,
2515                                                       status);
2516                 if (status != BTRFS_TREE_BLOCK_CLEAN) {
2517                         ret = -EIO;
2518                         fprintf(stderr, "bad block %llu\n",
2519                                 (unsigned long long)buf->start);
2520                 } else {
2521                         /*
2522                          * Signal to callers we need to start the scan over
2523                          * again since we'll have cow'ed blocks.
2524                          */
2525                         ret = -EAGAIN;
2526                 }
2527         } else {
2528                 rec->content_checked = 1;
2529                 if (flags & BTRFS_BLOCK_FLAG_FULL_BACKREF)
2530                         rec->owner_ref_checked = 1;
2531                 else {
2532                         ret = check_owner_ref(root, rec, buf);
2533                         if (!ret)
2534                                 rec->owner_ref_checked = 1;
2535                 }
2536         }
2537         if (!ret)
2538                 maybe_free_extent_rec(extent_cache, rec);
2539         return ret;
2540 }
2541
2542 static struct tree_backref *find_tree_backref(struct extent_record *rec,
2543                                                 u64 parent, u64 root)
2544 {
2545         struct list_head *cur = rec->backrefs.next;
2546         struct extent_backref *node;
2547         struct tree_backref *back;
2548
2549         while(cur != &rec->backrefs) {
2550                 node = list_entry(cur, struct extent_backref, list);
2551                 cur = cur->next;
2552                 if (node->is_data)
2553                         continue;
2554                 back = (struct tree_backref *)node;
2555                 if (parent > 0) {
2556                         if (!node->full_backref)
2557                                 continue;
2558                         if (parent == back->parent)
2559                                 return back;
2560                 } else {
2561                         if (node->full_backref)
2562                                 continue;
2563                         if (back->root == root)
2564                                 return back;
2565                 }
2566         }
2567         return NULL;
2568 }
2569
2570 static struct tree_backref *alloc_tree_backref(struct extent_record *rec,
2571                                                 u64 parent, u64 root)
2572 {
2573         struct tree_backref *ref = malloc(sizeof(*ref));
2574         memset(&ref->node, 0, sizeof(ref->node));
2575         if (parent > 0) {
2576                 ref->parent = parent;
2577                 ref->node.full_backref = 1;
2578         } else {
2579                 ref->root = root;
2580                 ref->node.full_backref = 0;
2581         }
2582         list_add_tail(&ref->node.list, &rec->backrefs);
2583
2584         return ref;
2585 }
2586
2587 static struct data_backref *find_data_backref(struct extent_record *rec,
2588                                                 u64 parent, u64 root,
2589                                                 u64 owner, u64 offset,
2590                                                 int found_ref,
2591                                                 u64 disk_bytenr, u64 bytes)
2592 {
2593         struct list_head *cur = rec->backrefs.next;
2594         struct extent_backref *node;
2595         struct data_backref *back;
2596
2597         while(cur != &rec->backrefs) {
2598                 node = list_entry(cur, struct extent_backref, list);
2599                 cur = cur->next;
2600                 if (!node->is_data)
2601                         continue;
2602                 back = (struct data_backref *)node;
2603                 if (parent > 0) {
2604                         if (!node->full_backref)
2605                                 continue;
2606                         if (parent == back->parent)
2607                                 return back;
2608                 } else {
2609                         if (node->full_backref)
2610                                 continue;
2611                         if (back->root == root && back->owner == owner &&
2612                             back->offset == offset) {
2613                                 if (found_ref && node->found_ref &&
2614                                     (back->bytes != bytes ||
2615                                     back->disk_bytenr != disk_bytenr))
2616                                         continue;
2617                                 return back;
2618                         }
2619                 }
2620         }
2621         return NULL;
2622 }
2623
2624 static struct data_backref *alloc_data_backref(struct extent_record *rec,
2625                                                 u64 parent, u64 root,
2626                                                 u64 owner, u64 offset,
2627                                                 u64 max_size)
2628 {
2629         struct data_backref *ref = malloc(sizeof(*ref));
2630         memset(&ref->node, 0, sizeof(ref->node));
2631         ref->node.is_data = 1;
2632
2633         if (parent > 0) {
2634                 ref->parent = parent;
2635                 ref->owner = 0;
2636                 ref->offset = 0;
2637                 ref->node.full_backref = 1;
2638         } else {
2639                 ref->root = root;
2640                 ref->owner = owner;
2641                 ref->offset = offset;
2642                 ref->node.full_backref = 0;
2643         }
2644         ref->bytes = max_size;
2645         ref->found_ref = 0;
2646         ref->num_refs = 0;
2647         list_add_tail(&ref->node.list, &rec->backrefs);
2648         if (max_size > rec->max_size)
2649                 rec->max_size = max_size;
2650         return ref;
2651 }
2652
2653 static int add_extent_rec(struct cache_tree *extent_cache,
2654                           struct btrfs_key *parent_key, u64 parent_gen,
2655                           u64 start, u64 nr, u64 extent_item_refs,
2656                           int is_root, int inc_ref, int set_checked,
2657                           int metadata, int extent_rec, u64 max_size)
2658 {
2659         struct extent_record *rec;
2660         struct cache_extent *cache;
2661         int ret = 0;
2662         int dup = 0;
2663
2664         cache = lookup_cache_extent(extent_cache, start, nr);
2665         if (cache) {
2666                 rec = container_of(cache, struct extent_record, cache);
2667                 if (inc_ref)
2668                         rec->refs++;
2669                 if (rec->nr == 1)
2670                         rec->nr = max(nr, max_size);
2671
2672                 /*
2673                  * We need to make sure to reset nr to whatever the extent
2674                  * record says was the real size, this way we can compare it to
2675                  * the backrefs.
2676                  */
2677                 if (extent_rec) {
2678                         if (start != rec->start || rec->found_rec) {
2679                                 struct extent_record *tmp;
2680
2681                                 dup = 1;
2682                                 if (list_empty(&rec->list))
2683                                         list_add_tail(&rec->list,
2684                                                       &duplicate_extents);
2685
2686                                 /*
2687                                  * We have to do this song and dance in case we
2688                                  * find an extent record that falls inside of
2689                                  * our current extent record but does not have
2690                                  * the same objectid.
2691                                  */
2692                                 tmp = malloc(sizeof(*tmp));
2693                                 if (!tmp)
2694                                         return -ENOMEM;
2695                                 tmp->start = start;
2696                                 tmp->max_size = max_size;
2697                                 tmp->nr = nr;
2698                                 tmp->found_rec = 1;
2699                                 tmp->metadata = metadata;
2700                                 tmp->extent_item_refs = extent_item_refs;
2701                                 INIT_LIST_HEAD(&tmp->list);
2702                                 list_add_tail(&tmp->list, &rec->dups);
2703                                 rec->num_duplicates++;
2704                         } else {
2705                                 rec->nr = nr;
2706                                 rec->found_rec = 1;
2707                         }
2708                 }
2709
2710                 if (extent_item_refs && !dup) {
2711                         if (rec->extent_item_refs) {
2712                                 fprintf(stderr, "block %llu rec "
2713                                         "extent_item_refs %llu, passed %llu\n",
2714                                         (unsigned long long)start,
2715                                         (unsigned long long)
2716                                                         rec->extent_item_refs,
2717                                         (unsigned long long)extent_item_refs);
2718                         }
2719                         rec->extent_item_refs = extent_item_refs;
2720                 }
2721                 if (is_root)
2722                         rec->is_root = 1;
2723                 if (set_checked) {
2724                         rec->content_checked = 1;
2725                         rec->owner_ref_checked = 1;
2726                 }
2727
2728                 if (parent_key)
2729                         btrfs_cpu_key_to_disk(&rec->parent_key, parent_key);
2730                 if (parent_gen)
2731                         rec->parent_generation = parent_gen;
2732
2733                 if (rec->max_size < max_size)
2734                         rec->max_size = max_size;
2735
2736                 maybe_free_extent_rec(extent_cache, rec);
2737                 return ret;
2738         }
2739         rec = malloc(sizeof(*rec));
2740         rec->start = start;
2741         rec->max_size = max_size;
2742         rec->nr = max(nr, max_size);
2743         rec->found_rec = !!extent_rec;
2744         rec->content_checked = 0;
2745         rec->owner_ref_checked = 0;
2746         rec->num_duplicates = 0;
2747         rec->metadata = metadata;
2748         INIT_LIST_HEAD(&rec->backrefs);
2749         INIT_LIST_HEAD(&rec->dups);
2750         INIT_LIST_HEAD(&rec->list);
2751
2752         if (is_root)
2753                 rec->is_root = 1;
2754         else
2755                 rec->is_root = 0;
2756
2757         if (inc_ref)
2758                 rec->refs = 1;
2759         else
2760                 rec->refs = 0;
2761
2762         if (extent_item_refs)
2763                 rec->extent_item_refs = extent_item_refs;
2764         else
2765                 rec->extent_item_refs = 0;
2766
2767         if (parent_key)
2768                 btrfs_cpu_key_to_disk(&rec->parent_key, parent_key);
2769         else
2770                 memset(&rec->parent_key, 0, sizeof(*parent_key));
2771
2772         if (parent_gen)
2773                 rec->parent_generation = parent_gen;
2774         else
2775                 rec->parent_generation = 0;
2776
2777         rec->cache.start = start;
2778         rec->cache.size = nr;
2779         ret = insert_cache_extent(extent_cache, &rec->cache);
2780         BUG_ON(ret);
2781         bytes_used += nr;
2782         if (set_checked) {
2783                 rec->content_checked = 1;
2784                 rec->owner_ref_checked = 1;
2785         }
2786         return ret;
2787 }
2788
2789 static int add_tree_backref(struct cache_tree *extent_cache, u64 bytenr,
2790                             u64 parent, u64 root, int found_ref)
2791 {
2792         struct extent_record *rec;
2793         struct tree_backref *back;
2794         struct cache_extent *cache;
2795
2796         cache = lookup_cache_extent(extent_cache, bytenr, 1);
2797         if (!cache) {
2798                 add_extent_rec(extent_cache, NULL, 0, bytenr,
2799                                1, 0, 0, 0, 0, 1, 0, 0);
2800                 cache = lookup_cache_extent(extent_cache, bytenr, 1);
2801                 if (!cache)
2802                         abort();
2803         }
2804
2805         rec = container_of(cache, struct extent_record, cache);
2806         if (rec->start != bytenr) {
2807                 abort();
2808         }
2809
2810         back = find_tree_backref(rec, parent, root);
2811         if (!back)
2812                 back = alloc_tree_backref(rec, parent, root);
2813
2814         if (found_ref) {
2815                 if (back->node.found_ref) {
2816                         fprintf(stderr, "Extent back ref already exists "
2817                                 "for %llu parent %llu root %llu \n",
2818                                 (unsigned long long)bytenr,
2819                                 (unsigned long long)parent,
2820                                 (unsigned long long)root);
2821                 }
2822                 back->node.found_ref = 1;
2823         } else {
2824                 if (back->node.found_extent_tree) {
2825                         fprintf(stderr, "Extent back ref already exists "
2826                                 "for %llu parent %llu root %llu \n",
2827                                 (unsigned long long)bytenr,
2828                                 (unsigned long long)parent,
2829                                 (unsigned long long)root);
2830                 }
2831                 back->node.found_extent_tree = 1;
2832         }
2833         return 0;
2834 }
2835
2836 static int add_data_backref(struct cache_tree *extent_cache, u64 bytenr,
2837                             u64 parent, u64 root, u64 owner, u64 offset,
2838                             u32 num_refs, int found_ref, u64 max_size)
2839 {
2840         struct extent_record *rec;
2841         struct data_backref *back;
2842         struct cache_extent *cache;
2843
2844         cache = lookup_cache_extent(extent_cache, bytenr, 1);
2845         if (!cache) {
2846                 add_extent_rec(extent_cache, NULL, 0, bytenr, 1, 0, 0, 0, 0,
2847                                0, 0, max_size);
2848                 cache = lookup_cache_extent(extent_cache, bytenr, 1);
2849                 if (!cache)
2850                         abort();
2851         }
2852
2853         rec = container_of(cache, struct extent_record, cache);
2854         if (rec->max_size < max_size)
2855                 rec->max_size = max_size;
2856
2857         /*
2858          * If found_ref is set then max_size is the real size and must match the
2859          * existing refs.  So if we have already found a ref then we need to
2860          * make sure that this ref matches the existing one, otherwise we need
2861          * to add a new backref so we can notice that the backrefs don't match
2862          * and we need to figure out who is telling the truth.  This is to
2863          * account for that awful fsync bug I introduced where we'd end up with
2864          * a btrfs_file_extent_item that would have its length include multiple
2865          * prealloc extents or point inside of a prealloc extent.
2866          */
2867         back = find_data_backref(rec, parent, root, owner, offset, found_ref,
2868                                  bytenr, max_size);
2869         if (!back)
2870                 back = alloc_data_backref(rec, parent, root, owner, offset,
2871                                           max_size);
2872
2873         if (found_ref) {
2874                 BUG_ON(num_refs != 1);
2875                 if (back->node.found_ref)
2876                         BUG_ON(back->bytes != max_size);
2877                 back->node.found_ref = 1;
2878                 back->found_ref += 1;
2879                 back->bytes = max_size;
2880                 back->disk_bytenr = bytenr;
2881                 rec->refs += 1;
2882                 rec->content_checked = 1;
2883                 rec->owner_ref_checked = 1;
2884         } else {
2885                 if (back->node.found_extent_tree) {
2886                         fprintf(stderr, "Extent back ref already exists "
2887                                 "for %llu parent %llu root %llu "
2888                                 "owner %llu offset %llu num_refs %lu\n",
2889                                 (unsigned long long)bytenr,
2890                                 (unsigned long long)parent,
2891                                 (unsigned long long)root,
2892                                 (unsigned long long)owner,
2893                                 (unsigned long long)offset,
2894                                 (unsigned long)num_refs);
2895                 }
2896                 back->num_refs = num_refs;
2897                 back->node.found_extent_tree = 1;
2898         }
2899         return 0;
2900 }
2901
2902 static int add_pending(struct cache_tree *pending,
2903                        struct cache_tree *seen, u64 bytenr, u32 size)
2904 {
2905         int ret;
2906         ret = add_cache_extent(seen, bytenr, size);
2907         if (ret)
2908                 return ret;
2909         add_cache_extent(pending, bytenr, size);
2910         return 0;
2911 }
2912
2913 static int pick_next_pending(struct cache_tree *pending,
2914                         struct cache_tree *reada,
2915                         struct cache_tree *nodes,
2916                         u64 last, struct block_info *bits, int bits_nr,
2917                         int *reada_bits)
2918 {
2919         unsigned long node_start = last;
2920         struct cache_extent *cache;
2921         int ret;
2922
2923         cache = search_cache_extent(reada, 0);
2924         if (cache) {
2925                 bits[0].start = cache->start;
2926                 bits[0].size = cache->size;
2927                 *reada_bits = 1;
2928                 return 1;
2929         }
2930         *reada_bits = 0;
2931         if (node_start > 32768)
2932                 node_start -= 32768;
2933
2934         cache = search_cache_extent(nodes, node_start);
2935         if (!cache)
2936                 cache = search_cache_extent(nodes, 0);
2937
2938         if (!cache) {
2939                  cache = search_cache_extent(pending, 0);
2940                  if (!cache)
2941                          return 0;
2942                  ret = 0;
2943                  do {
2944                          bits[ret].start = cache->start;
2945                          bits[ret].size = cache->size;
2946                          cache = next_cache_extent(cache);
2947                          ret++;
2948                  } while (cache && ret < bits_nr);
2949                  return ret;
2950         }
2951
2952         ret = 0;
2953         do {
2954                 bits[ret].start = cache->start;
2955                 bits[ret].size = cache->size;
2956                 cache = next_cache_extent(cache);
2957                 ret++;
2958         } while (cache && ret < bits_nr);
2959
2960         if (bits_nr - ret > 8) {
2961                 u64 lookup = bits[0].start + bits[0].size;
2962                 struct cache_extent *next;
2963                 next = search_cache_extent(pending, lookup);
2964                 while(next) {
2965                         if (next->start - lookup > 32768)
2966                                 break;
2967                         bits[ret].start = next->start;
2968                         bits[ret].size = next->size;
2969                         lookup = next->start + next->size;
2970                         ret++;
2971                         if (ret == bits_nr)
2972                                 break;
2973                         next = next_cache_extent(next);
2974                         if (!next)
2975                                 break;
2976                 }
2977         }
2978         return ret;
2979 }
2980
2981 static void free_chunk_record(struct cache_extent *cache)
2982 {
2983         struct chunk_record *rec;
2984
2985         rec = container_of(cache, struct chunk_record, cache);
2986         free(rec);
2987 }
2988
2989 void free_chunk_cache_tree(struct cache_tree *chunk_cache)
2990 {
2991         cache_tree_free_extents(chunk_cache, free_chunk_record);
2992 }
2993
2994 static void free_device_record(struct rb_node *node)
2995 {
2996         struct device_record *rec;
2997
2998         rec = container_of(node, struct device_record, node);
2999         free(rec);
3000 }
3001
3002 FREE_RB_BASED_TREE(device_cache, free_device_record);
3003
3004 int insert_block_group_record(struct block_group_tree *tree,
3005                               struct block_group_record *bg_rec)
3006 {
3007         int ret;
3008
3009         ret = insert_cache_extent(&tree->tree, &bg_rec->cache);
3010         if (ret)
3011                 return ret;
3012
3013         list_add_tail(&bg_rec->list, &tree->block_groups);
3014         return 0;
3015 }
3016
3017 static void free_block_group_record(struct cache_extent *cache)
3018 {
3019         struct block_group_record *rec;
3020
3021         rec = container_of(cache, struct block_group_record, cache);
3022         free(rec);
3023 }
3024
3025 void free_block_group_tree(struct block_group_tree *tree)
3026 {
3027         cache_tree_free_extents(&tree->tree, free_block_group_record);
3028 }
3029
3030 int insert_device_extent_record(struct device_extent_tree *tree,
3031                                 struct device_extent_record *de_rec)
3032 {
3033         int ret;
3034
3035         /*
3036          * Device extent is a bit different from the other extents, because
3037          * the extents which belong to the different devices may have the
3038          * same start and size, so we need use the special extent cache
3039          * search/insert functions.
3040          */
3041         ret = insert_cache_extent2(&tree->tree, &de_rec->cache);
3042         if (ret)
3043                 return ret;
3044
3045         list_add_tail(&de_rec->chunk_list, &tree->no_chunk_orphans);
3046         list_add_tail(&de_rec->device_list, &tree->no_device_orphans);
3047         return 0;
3048 }
3049
3050 static void free_device_extent_record(struct cache_extent *cache)
3051 {
3052         struct device_extent_record *rec;
3053
3054         rec = container_of(cache, struct device_extent_record, cache);
3055         free(rec);
3056 }
3057
3058 void free_device_extent_tree(struct device_extent_tree *tree)
3059 {
3060         cache_tree_free_extents(&tree->tree, free_device_extent_record);
3061 }
3062
3063 #ifdef BTRFS_COMPAT_EXTENT_TREE_V0
3064 static int process_extent_ref_v0(struct cache_tree *extent_cache,
3065                                  struct extent_buffer *leaf, int slot)
3066 {
3067         struct btrfs_extent_ref_v0 *ref0;
3068         struct btrfs_key key;
3069
3070         btrfs_item_key_to_cpu(leaf, &key, slot);
3071         ref0 = btrfs_item_ptr(leaf, slot, struct btrfs_extent_ref_v0);
3072         if (btrfs_ref_objectid_v0(leaf, ref0) < BTRFS_FIRST_FREE_OBJECTID) {
3073                 add_tree_backref(extent_cache, key.objectid, key.offset, 0, 0);
3074         } else {
3075                 add_data_backref(extent_cache, key.objectid, key.offset, 0,
3076                                  0, 0, btrfs_ref_count_v0(leaf, ref0), 0, 0);
3077         }
3078         return 0;
3079 }
3080 #endif
3081
3082 struct chunk_record *btrfs_new_chunk_record(struct extent_buffer *leaf,
3083                                             struct btrfs_key *key,
3084                                             int slot)
3085 {
3086         struct btrfs_chunk *ptr;
3087         struct chunk_record *rec;
3088         int num_stripes, i;
3089
3090         ptr = btrfs_item_ptr(leaf, slot, struct btrfs_chunk);
3091         num_stripes = btrfs_chunk_num_stripes(leaf, ptr);
3092
3093         rec = malloc(btrfs_chunk_record_size(num_stripes));
3094         if (!rec) {
3095                 fprintf(stderr, "memory allocation failed\n");
3096                 exit(-1);
3097         }
3098
3099         memset(rec, 0, btrfs_chunk_record_size(num_stripes));
3100
3101         INIT_LIST_HEAD(&rec->list);
3102         INIT_LIST_HEAD(&rec->dextents);
3103         rec->bg_rec = NULL;
3104
3105         rec->cache.start = key->offset;
3106         rec->cache.size = btrfs_chunk_length(leaf, ptr);
3107
3108         rec->generation = btrfs_header_generation(leaf);
3109
3110         rec->objectid = key->objectid;
3111         rec->type = key->type;
3112         rec->offset = key->offset;
3113
3114         rec->length = rec->cache.size;
3115         rec->owner = btrfs_chunk_owner(leaf, ptr);
3116         rec->stripe_len = btrfs_chunk_stripe_len(leaf, ptr);
3117         rec->type_flags = btrfs_chunk_type(leaf, ptr);
3118         rec->io_width = btrfs_chunk_io_width(leaf, ptr);
3119         rec->io_align = btrfs_chunk_io_align(leaf, ptr);
3120         rec->sector_size = btrfs_chunk_sector_size(leaf, ptr);
3121         rec->num_stripes = num_stripes;
3122         rec->sub_stripes = btrfs_chunk_sub_stripes(leaf, ptr);
3123
3124         for (i = 0; i < rec->num_stripes; ++i) {
3125                 rec->stripes[i].devid =
3126                         btrfs_stripe_devid_nr(leaf, ptr, i);
3127                 rec->stripes[i].offset =
3128                         btrfs_stripe_offset_nr(leaf, ptr, i);
3129                 read_extent_buffer(leaf, rec->stripes[i].dev_uuid,
3130                                 (unsigned long)btrfs_stripe_dev_uuid_nr(ptr, i),
3131                                 BTRFS_UUID_SIZE);
3132         }
3133
3134         return rec;
3135 }
3136
3137 static int process_chunk_item(struct cache_tree *chunk_cache,
3138                               struct btrfs_key *key, struct extent_buffer *eb,
3139                               int slot)
3140 {
3141         struct chunk_record *rec;
3142         int ret = 0;
3143
3144         rec = btrfs_new_chunk_record(eb, key, slot);
3145         ret = insert_cache_extent(chunk_cache, &rec->cache);
3146         if (ret) {
3147                 fprintf(stderr, "Chunk[%llu, %llu] existed.\n",
3148                         rec->offset, rec->length);
3149                 free(rec);
3150         }
3151
3152         return ret;
3153 }
3154
3155 static int process_device_item(struct rb_root *dev_cache,
3156                 struct btrfs_key *key, struct extent_buffer *eb, int slot)
3157 {
3158         struct btrfs_dev_item *ptr;
3159         struct device_record *rec;
3160         int ret = 0;
3161
3162         ptr = btrfs_item_ptr(eb,
3163                 slot, struct btrfs_dev_item);
3164
3165         rec = malloc(sizeof(*rec));
3166         if (!rec) {
3167                 fprintf(stderr, "memory allocation failed\n");
3168                 return -ENOMEM;
3169         }
3170
3171         rec->devid = key->offset;
3172         rec->generation = btrfs_header_generation(eb);
3173
3174         rec->objectid = key->objectid;
3175         rec->type = key->type;
3176         rec->offset = key->offset;
3177
3178         rec->devid = btrfs_device_id(eb, ptr);
3179         rec->total_byte = btrfs_device_total_bytes(eb, ptr);
3180         rec->byte_used = btrfs_device_bytes_used(eb, ptr);
3181
3182         ret = rb_insert(dev_cache, &rec->node, device_record_compare);
3183         if (ret) {
3184                 fprintf(stderr, "Device[%llu] existed.\n", rec->devid);
3185                 free(rec);
3186         }
3187
3188         return ret;
3189 }
3190
3191 struct block_group_record *
3192 btrfs_new_block_group_record(struct extent_buffer *leaf, struct btrfs_key *key,
3193                              int slot)
3194 {
3195         struct btrfs_block_group_item *ptr;
3196         struct block_group_record *rec;
3197
3198         rec = malloc(sizeof(*rec));
3199         if (!rec) {
3200                 fprintf(stderr, "memory allocation failed\n");
3201                 exit(-1);
3202         }
3203         memset(rec, 0, sizeof(*rec));
3204
3205         rec->cache.start = key->objectid;
3206         rec->cache.size = key->offset;
3207
3208         rec->generation = btrfs_header_generation(leaf);
3209
3210         rec->objectid = key->objectid;
3211         rec->type = key->type;
3212         rec->offset = key->offset;
3213
3214         ptr = btrfs_item_ptr(leaf, slot, struct btrfs_block_group_item);
3215         rec->flags = btrfs_disk_block_group_flags(leaf, ptr);
3216
3217         INIT_LIST_HEAD(&rec->list);
3218
3219         return rec;
3220 }
3221
3222 static int process_block_group_item(struct block_group_tree *block_group_cache,
3223                                     struct btrfs_key *key,
3224                                     struct extent_buffer *eb, int slot)
3225 {
3226         struct block_group_record *rec;
3227         int ret = 0;
3228
3229         rec = btrfs_new_block_group_record(eb, key, slot);
3230         ret = insert_block_group_record(block_group_cache, rec);
3231         if (ret) {
3232                 fprintf(stderr, "Block Group[%llu, %llu] existed.\n",
3233                         rec->objectid, rec->offset);
3234                 free(rec);
3235         }
3236
3237         return ret;
3238 }
3239
3240 struct device_extent_record *
3241 btrfs_new_device_extent_record(struct extent_buffer *leaf,
3242                                struct btrfs_key *key, int slot)
3243 {
3244         struct device_extent_record *rec;
3245         struct btrfs_dev_extent *ptr;
3246
3247         rec = malloc(sizeof(*rec));
3248         if (!rec) {
3249                 fprintf(stderr, "memory allocation failed\n");
3250                 exit(-1);
3251         }
3252         memset(rec, 0, sizeof(*rec));
3253
3254         rec->cache.objectid = key->objectid;
3255         rec->cache.start = key->offset;
3256
3257         rec->generation = btrfs_header_generation(leaf);
3258
3259         rec->objectid = key->objectid;
3260         rec->type = key->type;
3261         rec->offset = key->offset;
3262
3263         ptr = btrfs_item_ptr(leaf, slot, struct btrfs_dev_extent);
3264         rec->chunk_objecteid =
3265                 btrfs_dev_extent_chunk_objectid(leaf, ptr);
3266         rec->chunk_offset =
3267                 btrfs_dev_extent_chunk_offset(leaf, ptr);
3268         rec->length = btrfs_dev_extent_length(leaf, ptr);
3269         rec->cache.size = rec->length;
3270
3271         INIT_LIST_HEAD(&rec->chunk_list);
3272         INIT_LIST_HEAD(&rec->device_list);
3273
3274         return rec;
3275 }
3276
3277 static int
3278 process_device_extent_item(struct device_extent_tree *dev_extent_cache,
3279                            struct btrfs_key *key, struct extent_buffer *eb,
3280                            int slot)
3281 {
3282         struct device_extent_record *rec;
3283         int ret;
3284
3285         rec = btrfs_new_device_extent_record(eb, key, slot);
3286         ret = insert_device_extent_record(dev_extent_cache, rec);
3287         if (ret) {
3288                 fprintf(stderr,
3289                         "Device extent[%llu, %llu, %llu] existed.\n",
3290                         rec->objectid, rec->offset, rec->length);
3291                 free(rec);
3292         }
3293
3294         return ret;
3295 }
3296
3297 static int process_extent_item(struct btrfs_root *root,
3298                                struct cache_tree *extent_cache,
3299                                struct extent_buffer *eb, int slot)
3300 {
3301         struct btrfs_extent_item *ei;
3302         struct btrfs_extent_inline_ref *iref;
3303         struct btrfs_extent_data_ref *dref;
3304         struct btrfs_shared_data_ref *sref;
3305         struct btrfs_key key;
3306         unsigned long end;
3307         unsigned long ptr;
3308         int type;
3309         u32 item_size = btrfs_item_size_nr(eb, slot);
3310         u64 refs = 0;
3311         u64 offset;
3312         u64 num_bytes;
3313         int metadata = 0;
3314
3315         btrfs_item_key_to_cpu(eb, &key, slot);
3316
3317         if (key.type == BTRFS_METADATA_ITEM_KEY) {
3318                 metadata = 1;
3319                 num_bytes = root->leafsize;
3320         } else {
3321                 num_bytes = key.offset;
3322         }
3323
3324         if (item_size < sizeof(*ei)) {
3325 #ifdef BTRFS_COMPAT_EXTENT_TREE_V0
3326                 struct btrfs_extent_item_v0 *ei0;
3327                 BUG_ON(item_size != sizeof(*ei0));
3328                 ei0 = btrfs_item_ptr(eb, slot, struct btrfs_extent_item_v0);
3329                 refs = btrfs_extent_refs_v0(eb, ei0);
3330 #else
3331                 BUG();
3332 #endif
3333                 return add_extent_rec(extent_cache, NULL, 0, key.objectid,
3334                                       num_bytes, refs, 0, 0, 0, metadata, 1,
3335                                       num_bytes);
3336         }
3337
3338         ei = btrfs_item_ptr(eb, slot, struct btrfs_extent_item);
3339         refs = btrfs_extent_refs(eb, ei);
3340
3341         add_extent_rec(extent_cache, NULL, 0, key.objectid, num_bytes,
3342                        refs, 0, 0, 0, metadata, 1, num_bytes);
3343
3344         ptr = (unsigned long)(ei + 1);
3345         if (btrfs_extent_flags(eb, ei) & BTRFS_EXTENT_FLAG_TREE_BLOCK &&
3346             key.type == BTRFS_EXTENT_ITEM_KEY)
3347                 ptr += sizeof(struct btrfs_tree_block_info);
3348
3349         end = (unsigned long)ei + item_size;
3350         while (ptr < end) {
3351                 iref = (struct btrfs_extent_inline_ref *)ptr;
3352                 type = btrfs_extent_inline_ref_type(eb, iref);
3353                 offset = btrfs_extent_inline_ref_offset(eb, iref);
3354                 switch (type) {
3355                 case BTRFS_TREE_BLOCK_REF_KEY:
3356                         add_tree_backref(extent_cache, key.objectid,
3357                                          0, offset, 0);
3358                         break;
3359                 case BTRFS_SHARED_BLOCK_REF_KEY:
3360                         add_tree_backref(extent_cache, key.objectid,
3361                                          offset, 0, 0);
3362                         break;
3363                 case BTRFS_EXTENT_DATA_REF_KEY:
3364                         dref = (struct btrfs_extent_data_ref *)(&iref->offset);
3365                         add_data_backref(extent_cache, key.objectid, 0,
3366                                         btrfs_extent_data_ref_root(eb, dref),
3367                                         btrfs_extent_data_ref_objectid(eb,
3368                                                                        dref),
3369                                         btrfs_extent_data_ref_offset(eb, dref),
3370                                         btrfs_extent_data_ref_count(eb, dref),
3371                                         0, num_bytes);
3372                         break;
3373                 case BTRFS_SHARED_DATA_REF_KEY:
3374                         sref = (struct btrfs_shared_data_ref *)(iref + 1);
3375                         add_data_backref(extent_cache, key.objectid, offset,
3376                                         0, 0, 0,
3377                                         btrfs_shared_data_ref_count(eb, sref),
3378                                         0, num_bytes);
3379                         break;
3380                 default:
3381                         fprintf(stderr, "corrupt extent record: key %Lu %u %Lu\n",
3382                                 key.objectid, key.type, num_bytes);
3383                         goto out;
3384                 }
3385                 ptr += btrfs_extent_inline_ref_size(type);
3386         }
3387         WARN_ON(ptr > end);
3388 out:
3389         return 0;
3390 }
3391
3392 static int check_cache_range(struct btrfs_root *root,
3393                              struct btrfs_block_group_cache *cache,
3394                              u64 offset, u64 bytes)
3395 {
3396         struct btrfs_free_space *entry;
3397         u64 *logical;
3398         u64 bytenr;
3399         int stripe_len;
3400         int i, nr, ret;
3401
3402         for (i = 0; i < BTRFS_SUPER_MIRROR_MAX; i++) {
3403                 bytenr = btrfs_sb_offset(i);
3404                 ret = btrfs_rmap_block(&root->fs_info->mapping_tree,
3405                                        cache->key.objectid, bytenr, 0,
3406                                        &logical, &nr, &stripe_len);
3407                 if (ret)
3408                         return ret;
3409
3410                 while (nr--) {
3411                         if (logical[nr] + stripe_len <= offset)
3412                                 continue;
3413                         if (offset + bytes <= logical[nr])
3414                                 continue;
3415                         if (logical[nr] == offset) {
3416                                 if (stripe_len >= bytes) {
3417                                         kfree(logical);
3418                                         return 0;
3419                                 }
3420                                 bytes -= stripe_len;
3421                                 offset += stripe_len;
3422                         } else if (logical[nr] < offset) {
3423                                 if (logical[nr] + stripe_len >=
3424                                     offset + bytes) {
3425                                         kfree(logical);
3426                                         return 0;
3427                                 }
3428                                 bytes = (offset + bytes) -
3429                                         (logical[nr] + stripe_len);
3430                                 offset = logical[nr] + stripe_len;
3431                         } else {
3432                                 /*
3433                                  * Could be tricky, the super may land in the
3434                                  * middle of the area we're checking.  First
3435                                  * check the easiest case, it's at the end.
3436                                  */
3437                                 if (logical[nr] + stripe_len >=
3438                                     bytes + offset) {
3439                                         bytes = logical[nr] - offset;
3440                                         continue;
3441                                 }
3442
3443                                 /* Check the left side */
3444                                 ret = check_cache_range(root, cache,
3445                                                         offset,
3446                                                         logical[nr] - offset);
3447                                 if (ret) {
3448                                         kfree(logical);
3449                                         return ret;
3450                                 }
3451
3452                                 /* Now we continue with the right side */
3453                                 bytes = (offset + bytes) -
3454                                         (logical[nr] + stripe_len);
3455                                 offset = logical[nr] + stripe_len;
3456                         }
3457                 }
3458
3459                 kfree(logical);
3460         }
3461
3462         entry = btrfs_find_free_space(cache->free_space_ctl, offset, bytes);
3463         if (!entry) {
3464                 fprintf(stderr, "There is no free space entry for %Lu-%Lu\n",
3465                         offset, offset+bytes);
3466                 return -EINVAL;
3467         }
3468
3469         if (entry->offset != offset) {
3470                 fprintf(stderr, "Wanted offset %Lu, found %Lu\n", offset,
3471                         entry->offset);
3472                 return -EINVAL;
3473         }
3474
3475         if (entry->bytes != bytes) {
3476                 fprintf(stderr, "Wanted bytes %Lu, found %Lu for off %Lu\n",
3477                         bytes, entry->bytes, offset);
3478                 return -EINVAL;
3479         }
3480
3481         unlink_free_space(cache->free_space_ctl, entry);
3482         free(entry);
3483         return 0;
3484 }
3485
3486 static int verify_space_cache(struct btrfs_root *root,
3487                               struct btrfs_block_group_cache *cache)
3488 {
3489         struct btrfs_path *path;
3490         struct extent_buffer *leaf;
3491         struct btrfs_key key;
3492         u64 last;
3493         int ret = 0;
3494
3495         path = btrfs_alloc_path();
3496         if (!path)
3497                 return -ENOMEM;
3498
3499         root = root->fs_info->extent_root;
3500
3501         last = max_t(u64, cache->key.objectid, BTRFS_SUPER_INFO_OFFSET);
3502
3503         key.objectid = last;
3504         key.offset = 0;
3505         key.type = BTRFS_EXTENT_ITEM_KEY;
3506
3507         ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
3508         if (ret < 0)
3509                 goto out;
3510         ret = 0;
3511         while (1) {
3512                 if (path->slots[0] >= btrfs_header_nritems(path->nodes[0])) {
3513                         ret = btrfs_next_leaf(root, path);
3514                         if (ret < 0)
3515                                 goto out;
3516                         if (ret > 0) {
3517                                 ret = 0;
3518                                 break;
3519                         }
3520                 }
3521                 leaf = path->nodes[0];
3522                 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
3523                 if (key.objectid >= cache->key.offset + cache->key.objectid)
3524                         break;
3525                 if (key.type != BTRFS_EXTENT_ITEM_KEY &&
3526                     key.type != BTRFS_METADATA_ITEM_KEY) {
3527                         path->slots[0]++;
3528                         continue;
3529                 }
3530
3531                 if (last == key.objectid) {
3532                         if (key.type == BTRFS_EXTENT_ITEM_KEY)
3533                                 last = key.objectid + key.offset;
3534                         else
3535                                 last = key.objectid + root->leafsize;
3536                         path->slots[0]++;
3537                         continue;
3538                 }
3539
3540                 ret = check_cache_range(root, cache, last,
3541                                         key.objectid - last);
3542                 if (ret)
3543                         break;
3544                 if (key.type == BTRFS_EXTENT_ITEM_KEY)
3545                         last = key.objectid + key.offset;
3546                 else
3547                         last = key.objectid + root->leafsize;
3548                 path->slots[0]++;
3549         }
3550
3551         if (last < cache->key.objectid + cache->key.offset)
3552                 ret = check_cache_range(root, cache, last,
3553                                         cache->key.objectid +
3554                                         cache->key.offset - last);
3555
3556 out:
3557         btrfs_free_path(path);
3558
3559         if (!ret &&
3560             !RB_EMPTY_ROOT(&cache->free_space_ctl->free_space_offset)) {
3561                 fprintf(stderr, "There are still entries left in the space "
3562                         "cache\n");
3563                 ret = -EINVAL;
3564         }
3565
3566         return ret;
3567 }
3568
3569 static int check_space_cache(struct btrfs_root *root)
3570 {
3571         struct btrfs_block_group_cache *cache;
3572         u64 start = BTRFS_SUPER_INFO_OFFSET + BTRFS_SUPER_INFO_SIZE;
3573         int ret;
3574         int error = 0;
3575
3576         if (btrfs_super_cache_generation(root->fs_info->super_copy) != -1ULL &&
3577             btrfs_super_generation(root->fs_info->super_copy) !=
3578             btrfs_super_cache_generation(root->fs_info->super_copy)) {
3579                 printf("cache and super generation don't match, space cache "
3580                        "will be invalidated\n");
3581                 return 0;
3582         }
3583
3584         while (1) {
3585                 cache = btrfs_lookup_first_block_group(root->fs_info, start);
3586                 if (!cache)
3587                         break;
3588
3589                 start = cache->key.objectid + cache->key.offset;
3590                 if (!cache->free_space_ctl) {
3591                         if (btrfs_init_free_space_ctl(cache,
3592                                                       root->sectorsize)) {
3593                                 ret = -ENOMEM;
3594                                 break;
3595                         }
3596                 } else {
3597                         btrfs_remove_free_space_cache(cache);
3598                 }
3599
3600                 ret = load_free_space_cache(root->fs_info, cache);
3601                 if (!ret)
3602                         continue;
3603
3604                 ret = verify_space_cache(root, cache);
3605                 if (ret) {
3606                         fprintf(stderr, "cache appears valid but isnt %Lu\n",
3607                                 cache->key.objectid);
3608                         error++;
3609                 }
3610         }
3611
3612         return error ? -EINVAL : 0;
3613 }
3614
3615 static int read_extent_data(struct btrfs_root *root, char *data,
3616                         u64 logical, u64 *len, int mirror)
3617 {
3618         u64 offset = 0;
3619         struct btrfs_multi_bio *multi = NULL;
3620         struct btrfs_fs_info *info = root->fs_info;
3621         struct btrfs_device *device;
3622         int ret = 0;
3623         u64 max_len = *len;
3624
3625         ret = btrfs_map_block(&info->mapping_tree, READ, logical, len,
3626                               &multi, mirror, NULL);
3627         if (ret) {
3628                 fprintf(stderr, "Couldn't map the block %llu\n",
3629                                 logical + offset);
3630                 goto err;
3631         }
3632         device = multi->stripes[0].dev;
3633
3634         if (device->fd == 0)
3635                 goto err;
3636         if (*len > max_len)
3637                 *len = max_len;
3638
3639         ret = pread64(device->fd, data, *len, multi->stripes[0].physical);
3640         if (ret != *len)
3641                 ret = -EIO;
3642         else
3643                 ret = 0;
3644 err:
3645         kfree(multi);
3646         return ret;
3647 }
3648
3649 static int check_extent_csums(struct btrfs_root *root, u64 bytenr,
3650                         u64 num_bytes, unsigned long leaf_offset,
3651                         struct extent_buffer *eb) {
3652
3653         u64 offset = 0;
3654         u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy);
3655         char *data;
3656         unsigned long csum_offset;
3657         u32 csum;
3658         u32 csum_expected;
3659         u64 read_len;
3660         u64 data_checked = 0;
3661         u64 tmp;
3662         int ret = 0;
3663         int mirror;
3664         int num_copies;
3665
3666         if (num_bytes % root->sectorsize)
3667                 return -EINVAL;
3668
3669         data = malloc(num_bytes);
3670         if (!data)
3671                 return -ENOMEM;
3672
3673         while (offset < num_bytes) {
3674                 mirror = 0;
3675 again:
3676                 read_len = num_bytes - offset;
3677                 /* read as much space once a time */
3678                 ret = read_extent_data(root, data + offset,
3679                                 bytenr + offset, &read_len, mirror);
3680                 if (ret)
3681                         goto out;
3682                 data_checked = 0;
3683                 /* verify every 4k data's checksum */
3684                 while (data_checked < read_len) {
3685                         csum = ~(u32)0;
3686                         tmp = offset + data_checked;
3687
3688                         csum = btrfs_csum_data(NULL, (char *)data + tmp,
3689                                                csum, root->sectorsize);
3690                         btrfs_csum_final(csum, (char *)&csum);
3691
3692                         csum_offset = leaf_offset +
3693                                  tmp / root->sectorsize * csum_size;
3694                         read_extent_buffer(eb, (char *)&csum_expected,
3695                                            csum_offset, csum_size);
3696                         /* try another mirror */
3697                         if (csum != csum_expected) {
3698                                 fprintf(stderr, "mirror %d bytenr %llu csum %u expected csum %u\n",
3699                                                 mirror, bytenr + tmp,
3700                                                 csum, csum_expected);
3701                                 num_copies = btrfs_num_copies(
3702                                                 &root->fs_info->mapping_tree,
3703                                                 bytenr, num_bytes);
3704                                 if (mirror < num_copies - 1) {
3705                                         mirror += 1;
3706                                         goto again;
3707                                 }
3708                         }
3709                         data_checked += root->sectorsize;
3710                 }
3711                 offset += read_len;
3712         }
3713 out:
3714         free(data);
3715         return ret;
3716 }
3717
3718 static int check_extent_exists(struct btrfs_root *root, u64 bytenr,
3719                                u64 num_bytes)
3720 {
3721         struct btrfs_path *path;
3722         struct extent_buffer *leaf;
3723         struct btrfs_key key;
3724         int ret;
3725
3726         path = btrfs_alloc_path();
3727         if (!path) {
3728                 fprintf(stderr, "Error allocing path\n");
3729                 return -ENOMEM;
3730         }
3731
3732         key.objectid = bytenr;
3733         key.type = BTRFS_EXTENT_ITEM_KEY;
3734         key.offset = 0;
3735
3736
3737 again:
3738         ret = btrfs_search_slot(NULL, root->fs_info->extent_root, &key, path,
3739                                 0, 0);
3740         if (ret < 0) {
3741                 fprintf(stderr, "Error looking up extent record %d\n", ret);
3742                 btrfs_free_path(path);
3743                 return ret;
3744         } else if (ret) {
3745                 if (path->slots[0])
3746                         path->slots[0]--;
3747                 else
3748                         btrfs_prev_leaf(root, path);
3749         }
3750
3751         btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
3752
3753         /*
3754          * Block group items come before extent items if they have the same
3755          * bytenr, so walk back one more just in case.  Dear future traveler,
3756          * first congrats on mastering time travel.  Now if it's not too much
3757          * trouble could you go back to 2006 and tell Chris to make the
3758          * BLOCK_GROUP_ITEM_KEY lower than the EXTENT_ITEM_KEY please?
3759          */
3760         if (key.type == BTRFS_BLOCK_GROUP_ITEM_KEY) {
3761                 if (path->slots[0])
3762                         path->slots[0]--;
3763                 else
3764                         btrfs_prev_leaf(root, path);
3765         }
3766
3767         while (num_bytes) {
3768                 if (path->slots[0] >= btrfs_header_nritems(path->nodes[0])) {
3769                         ret = btrfs_next_leaf(root, path);
3770                         if (ret < 0) {
3771                                 fprintf(stderr, "Error going to next leaf "
3772                                         "%d\n", ret);
3773                                 btrfs_free_path(path);
3774                                 return ret;
3775                         } else if (ret) {
3776                                 break;
3777                         }
3778                 }
3779                 leaf = path->nodes[0];
3780                 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
3781                 if (key.type != BTRFS_EXTENT_ITEM_KEY) {
3782                         path->slots[0]++;
3783                         continue;
3784                 }
3785                 if (key.objectid + key.offset < bytenr) {
3786                         path->slots[0]++;
3787                         continue;
3788                 }
3789                 if (key.objectid > bytenr + num_bytes)
3790                         break;
3791
3792                 if (key.objectid == bytenr) {
3793                         if (key.offset >= num_bytes) {
3794                                 num_bytes = 0;
3795                                 break;
3796                         }
3797                         num_bytes -= key.offset;
3798                         bytenr += key.offset;
3799                 } else if (key.objectid < bytenr) {
3800                         if (key.objectid + key.offset >= bytenr + num_bytes) {
3801                                 num_bytes = 0;
3802                                 break;
3803                         }
3804                         num_bytes = (bytenr + num_bytes) -
3805                                 (key.objectid + key.offset);
3806                         bytenr = key.objectid + key.offset;
3807                 } else {
3808                         if (key.objectid + key.offset < bytenr + num_bytes) {
3809                                 u64 new_start = key.objectid + key.offset;
3810                                 u64 new_bytes = bytenr + num_bytes - new_start;
3811
3812                                 /*
3813                                  * Weird case, the extent is in the middle of
3814                                  * our range, we'll have to search one side
3815                                  * and then the other.  Not sure if this happens
3816                                  * in real life, but no harm in coding it up
3817                                  * anyway just in case.
3818                                  */
3819                                 btrfs_release_path(path);
3820                                 ret = check_extent_exists(root, new_start,
3821                                                           new_bytes);
3822                                 if (ret) {
3823                                         fprintf(stderr, "Right section didn't "
3824                                                 "have a record\n");
3825                                         break;
3826                                 }
3827                                 num_bytes = key.objectid - bytenr;
3828                                 goto again;
3829                         }
3830                         num_bytes = key.objectid - bytenr;
3831                 }
3832                 path->slots[0]++;
3833         }
3834         ret = 0;
3835
3836         if (num_bytes) {
3837                 fprintf(stderr, "There are no extents for csum range "
3838                         "%Lu-%Lu\n", bytenr, bytenr+num_bytes);
3839                 ret = 1;
3840         }
3841
3842         btrfs_free_path(path);
3843         return ret;
3844 }
3845
3846 static int check_csums(struct btrfs_root *root)
3847 {
3848         struct btrfs_path *path;
3849         struct extent_buffer *leaf;
3850         struct btrfs_key key;
3851         u64 offset = 0, num_bytes = 0;
3852         u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy);
3853         int errors = 0;
3854         int ret;
3855         u64 data_len;
3856         unsigned long leaf_offset;
3857
3858         root = root->fs_info->csum_root;
3859
3860         key.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
3861         key.type = BTRFS_EXTENT_CSUM_KEY;
3862         key.offset = 0;
3863
3864         path = btrfs_alloc_path();
3865         if (!path)
3866                 return -ENOMEM;
3867
3868         ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
3869         if (ret < 0) {
3870                 fprintf(stderr, "Error searching csum tree %d\n", ret);
3871                 btrfs_free_path(path);
3872                 return ret;
3873         }
3874
3875         if (ret > 0 && path->slots[0])
3876                 path->slots[0]--;
3877         ret = 0;
3878
3879         while (1) {
3880                 if (path->slots[0] >= btrfs_header_nritems(path->nodes[0])) {
3881                         ret = btrfs_next_leaf(root, path);
3882                         if (ret < 0) {
3883                                 fprintf(stderr, "Error going to next leaf "
3884                                         "%d\n", ret);
3885                                 break;
3886                         }
3887                         if (ret)
3888                                 break;
3889                 }
3890                 leaf = path->nodes[0];
3891
3892                 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
3893                 if (key.type != BTRFS_EXTENT_CSUM_KEY) {
3894                         path->slots[0]++;
3895                         continue;
3896                 }
3897
3898                 data_len = (btrfs_item_size_nr(leaf, path->slots[0]) /
3899                               csum_size) * root->sectorsize;
3900                 if (!check_data_csum)
3901                         goto skip_csum_check;
3902                 leaf_offset = btrfs_item_ptr_offset(leaf, path->slots[0]);
3903                 ret = check_extent_csums(root, key.offset, data_len,
3904                                          leaf_offset, leaf);
3905                 if (ret)
3906                         break;
3907 skip_csum_check:
3908                 if (!num_bytes) {
3909                         offset = key.offset;
3910                 } else if (key.offset != offset + num_bytes) {
3911                         ret = check_extent_exists(root, offset, num_bytes);
3912                         if (ret) {
3913                                 fprintf(stderr, "Csum exists for %Lu-%Lu but "
3914                                         "there is no extent record\n",
3915                                         offset, offset+num_bytes);
3916                                 errors++;
3917                         }
3918                         offset = key.offset;
3919                         num_bytes = 0;
3920                 }
3921                 num_bytes += data_len;
3922                 path->slots[0]++;
3923         }
3924
3925         btrfs_free_path(path);
3926         return errors;
3927 }
3928
3929 static int is_dropped_key(struct btrfs_key *key,
3930                           struct btrfs_key *drop_key) {
3931         if (key->objectid < drop_key->objectid)
3932                 return 1;
3933         else if (key->objectid == drop_key->objectid) {
3934                 if (key->type < drop_key->type)
3935                         return 1;
3936                 else if (key->type == drop_key->type) {
3937                         if (key->offset < drop_key->offset)
3938                                 return 1;
3939                 }
3940         }
3941         return 0;
3942 }
3943
3944 static int run_next_block(struct btrfs_trans_handle *trans,
3945                           struct btrfs_root *root,
3946                           struct block_info *bits,
3947                           int bits_nr,
3948                           u64 *last,
3949                           struct cache_tree *pending,
3950                           struct cache_tree *seen,
3951                           struct cache_tree *reada,
3952                           struct cache_tree *nodes,
3953                           struct cache_tree *extent_cache,
3954                           struct cache_tree *chunk_cache,
3955                           struct rb_root *dev_cache,
3956                           struct block_group_tree *block_group_cache,
3957                           struct device_extent_tree *dev_extent_cache,
3958                           struct btrfs_root_item *ri)
3959 {
3960         struct extent_buffer *buf;
3961         u64 bytenr;
3962         u32 size;
3963         u64 parent;
3964         u64 owner;
3965         u64 flags;
3966         u64 ptr;
3967         u64 gen = 0;
3968         int ret = 0;
3969         int i;
3970         int nritems;
3971         struct btrfs_key key;
3972         struct cache_extent *cache;
3973         int reada_bits;
3974
3975         nritems = pick_next_pending(pending, reada, nodes, *last, bits,
3976                                     bits_nr, &reada_bits);
3977         if (nritems == 0)
3978                 return 1;
3979
3980         if (!reada_bits) {
3981                 for(i = 0; i < nritems; i++) {
3982                         ret = add_cache_extent(reada, bits[i].start,
3983                                                bits[i].size);
3984                         if (ret == -EEXIST)
3985                                 continue;
3986
3987                         /* fixme, get the parent transid */
3988                         readahead_tree_block(root, bits[i].start,
3989                                              bits[i].size, 0);
3990                 }
3991         }
3992         *last = bits[0].start;
3993         bytenr = bits[0].start;
3994         size = bits[0].size;
3995
3996         cache = lookup_cache_extent(pending, bytenr, size);
3997         if (cache) {
3998                 remove_cache_extent(pending, cache);
3999                 free(cache);
4000         }
4001         cache = lookup_cache_extent(reada, bytenr, size);
4002         if (cache) {
4003                 remove_cache_extent(reada, cache);
4004                 free(cache);
4005         }
4006         cache = lookup_cache_extent(nodes, bytenr, size);
4007         if (cache) {
4008                 remove_cache_extent(nodes, cache);
4009                 free(cache);
4010         }
4011         cache = lookup_cache_extent(extent_cache, bytenr, size);
4012         if (cache) {
4013                 struct extent_record *rec;
4014
4015                 rec = container_of(cache, struct extent_record, cache);
4016                 gen = rec->parent_generation;
4017         }
4018
4019         /* fixme, get the real parent transid */
4020         buf = read_tree_block(root, bytenr, size, gen);
4021         if (!extent_buffer_uptodate(buf)) {
4022                 record_bad_block_io(root->fs_info,
4023                                     extent_cache, bytenr, size);
4024                 goto out;
4025         }
4026
4027         nritems = btrfs_header_nritems(buf);
4028
4029         /*
4030          * FIXME, this only works only if we don't have any full
4031          * backref mode.
4032          */
4033         if (!init_extent_tree) {
4034                 ret = btrfs_lookup_extent_info(NULL, root, bytenr,
4035                                        btrfs_header_level(buf), 1, NULL,
4036                                        &flags);
4037                 if (ret < 0)
4038                         flags = 0;
4039         } else {
4040                 flags = 0;
4041         }
4042
4043         if (flags & BTRFS_BLOCK_FLAG_FULL_BACKREF) {
4044                 parent = bytenr;
4045                 owner = 0;
4046         } else {
4047                 parent = 0;
4048                 owner = btrfs_header_owner(buf);
4049         }
4050
4051         ret = check_block(trans, root, extent_cache, buf, flags);
4052         if (ret)
4053                 goto out;
4054
4055         if (btrfs_is_leaf(buf)) {
4056                 btree_space_waste += btrfs_leaf_free_space(root, buf);
4057                 for (i = 0; i < nritems; i++) {
4058                         struct btrfs_file_extent_item *fi;
4059                         btrfs_item_key_to_cpu(buf, &key, i);
4060                         if (key.type == BTRFS_EXTENT_ITEM_KEY) {
4061                                 process_extent_item(root, extent_cache, buf,
4062                                                     i);
4063                                 continue;
4064                         }
4065                         if (key.type == BTRFS_METADATA_ITEM_KEY) {
4066                                 process_extent_item(root, extent_cache, buf,
4067                                                     i);
4068                                 continue;
4069                         }
4070                         if (key.type == BTRFS_EXTENT_CSUM_KEY) {
4071                                 total_csum_bytes +=
4072                                         btrfs_item_size_nr(buf, i);
4073                                 continue;
4074                         }
4075                         if (key.type == BTRFS_CHUNK_ITEM_KEY) {
4076                                 process_chunk_item(chunk_cache, &key, buf, i);
4077                                 continue;
4078                         }
4079                         if (key.type == BTRFS_DEV_ITEM_KEY) {
4080                                 process_device_item(dev_cache, &key, buf, i);
4081                                 continue;
4082                         }
4083                         if (key.type == BTRFS_BLOCK_GROUP_ITEM_KEY) {
4084                                 process_block_group_item(block_group_cache,
4085                                         &key, buf, i);
4086                                 continue;
4087                         }
4088                         if (key.type == BTRFS_DEV_EXTENT_KEY) {
4089                                 process_device_extent_item(dev_extent_cache,
4090                                         &key, buf, i);
4091                                 continue;
4092
4093                         }
4094                         if (key.type == BTRFS_EXTENT_REF_V0_KEY) {
4095 #ifdef BTRFS_COMPAT_EXTENT_TREE_V0
4096                                 process_extent_ref_v0(extent_cache, buf, i);
4097 #else
4098                                 BUG();
4099 #endif
4100                                 continue;
4101                         }
4102
4103                         if (key.type == BTRFS_TREE_BLOCK_REF_KEY) {
4104                                 add_tree_backref(extent_cache, key.objectid, 0,
4105                                                  key.offset, 0);
4106                                 continue;
4107                         }
4108                         if (key.type == BTRFS_SHARED_BLOCK_REF_KEY) {
4109                                 add_tree_backref(extent_cache, key.objectid,
4110                                                  key.offset, 0, 0);
4111                                 continue;
4112                         }
4113                         if (key.type == BTRFS_EXTENT_DATA_REF_KEY) {
4114                                 struct btrfs_extent_data_ref *ref;
4115                                 ref = btrfs_item_ptr(buf, i,
4116                                                 struct btrfs_extent_data_ref);
4117                                 add_data_backref(extent_cache,
4118                                         key.objectid, 0,
4119                                         btrfs_extent_data_ref_root(buf, ref),
4120                                         btrfs_extent_data_ref_objectid(buf,
4121                                                                        ref),
4122                                         btrfs_extent_data_ref_offset(buf, ref),
4123                                         btrfs_extent_data_ref_count(buf, ref),
4124                                         0, root->sectorsize);
4125                                 continue;
4126                         }
4127                         if (key.type == BTRFS_SHARED_DATA_REF_KEY) {
4128                                 struct btrfs_shared_data_ref *ref;
4129                                 ref = btrfs_item_ptr(buf, i,
4130                                                 struct btrfs_shared_data_ref);
4131                                 add_data_backref(extent_cache,
4132                                         key.objectid, key.offset, 0, 0, 0,
4133                                         btrfs_shared_data_ref_count(buf, ref),
4134                                         0, root->sectorsize);
4135                                 continue;
4136                         }
4137                         if (key.type == BTRFS_ORPHAN_ITEM_KEY) {
4138                                 struct bad_item *bad;
4139
4140                                 if (key.objectid == BTRFS_ORPHAN_OBJECTID)
4141                                         continue;
4142                                 if (!owner)
4143                                         continue;
4144                                 bad = malloc(sizeof(struct bad_item));
4145                                 if (!bad)
4146                                         continue;
4147                                 INIT_LIST_HEAD(&bad->list);
4148                                 memcpy(&bad->key, &key,
4149                                        sizeof(struct btrfs_key));
4150                                 bad->root_id = owner;
4151                                 list_add_tail(&bad->list, &delete_items);
4152                                 continue;
4153                         }
4154                         if (key.type != BTRFS_EXTENT_DATA_KEY)
4155                                 continue;
4156                         fi = btrfs_item_ptr(buf, i,
4157                                             struct btrfs_file_extent_item);
4158                         if (btrfs_file_extent_type(buf, fi) ==
4159                             BTRFS_FILE_EXTENT_INLINE)
4160                                 continue;
4161                         if (btrfs_file_extent_disk_bytenr(buf, fi) == 0)
4162                                 continue;
4163
4164                         data_bytes_allocated +=
4165                                 btrfs_file_extent_disk_num_bytes(buf, fi);
4166                         if (data_bytes_allocated < root->sectorsize) {
4167                                 abort();
4168                         }
4169                         data_bytes_referenced +=
4170                                 btrfs_file_extent_num_bytes(buf, fi);
4171                         add_data_backref(extent_cache,
4172                                 btrfs_file_extent_disk_bytenr(buf, fi),
4173                                 parent, owner, key.objectid, key.offset -
4174                                 btrfs_file_extent_offset(buf, fi), 1, 1,
4175                                 btrfs_file_extent_disk_num_bytes(buf, fi));
4176                 }
4177         } else {
4178                 int level;
4179                 struct btrfs_key first_key;
4180
4181                 first_key.objectid = 0;
4182
4183                 if (nritems > 0)
4184                         btrfs_item_key_to_cpu(buf, &first_key, 0);
4185                 level = btrfs_header_level(buf);
4186                 for (i = 0; i < nritems; i++) {
4187                         ptr = btrfs_node_blockptr(buf, i);
4188                         size = btrfs_level_size(root, level - 1);
4189                         btrfs_node_key_to_cpu(buf, &key, i);
4190                         if (ri != NULL) {
4191                                 struct btrfs_key drop_key;
4192                                 btrfs_disk_key_to_cpu(&drop_key,
4193                                                       &ri->drop_progress);
4194                                 if ((level == ri->drop_level)
4195                                     && is_dropped_key(&key, &drop_key)) {
4196                                         continue;
4197                                 }
4198                         }
4199                         ret = add_extent_rec(extent_cache, &key,
4200                                              btrfs_node_ptr_generation(buf, i),
4201                                              ptr, size, 0, 0, 1, 0, 1, 0,
4202                                              size);
4203                         BUG_ON(ret);
4204
4205                         add_tree_backref(extent_cache, ptr, parent, owner, 1);
4206
4207                         if (level > 1) {
4208                                 add_pending(nodes, seen, ptr, size);
4209                         } else {
4210                                 add_pending(pending, seen, ptr, size);
4211                         }
4212                 }
4213                 btree_space_waste += (BTRFS_NODEPTRS_PER_BLOCK(root) -
4214                                       nritems) * sizeof(struct btrfs_key_ptr);
4215         }
4216         total_btree_bytes += buf->len;
4217         if (fs_root_objectid(btrfs_header_owner(buf)))
4218                 total_fs_tree_bytes += buf->len;
4219         if (btrfs_header_owner(buf) == BTRFS_EXTENT_TREE_OBJECTID)
4220                 total_extent_tree_bytes += buf->len;
4221         if (!found_old_backref &&
4222             btrfs_header_owner(buf) == BTRFS_TREE_RELOC_OBJECTID &&
4223             btrfs_header_backref_rev(buf) == BTRFS_MIXED_BACKREF_REV &&
4224             !btrfs_header_flag(buf, BTRFS_HEADER_FLAG_RELOC))
4225                 found_old_backref = 1;
4226 out:
4227         free_extent_buffer(buf);
4228         return ret;
4229 }
4230
4231 static int add_root_to_pending(struct extent_buffer *buf,
4232                                struct cache_tree *extent_cache,
4233                                struct cache_tree *pending,
4234                                struct cache_tree *seen,
4235                                struct cache_tree *nodes,
4236                                struct btrfs_key *root_key)
4237 {
4238         if (btrfs_header_level(buf) > 0)
4239                 add_pending(nodes, seen, buf->start, buf->len);
4240         else
4241                 add_pending(pending, seen, buf->start, buf->len);
4242         add_extent_rec(extent_cache, NULL, 0, buf->start, buf->len,
4243                        0, 1, 1, 0, 1, 0, buf->len);
4244
4245         if (root_key->objectid == BTRFS_TREE_RELOC_OBJECTID ||
4246             btrfs_header_backref_rev(buf) < BTRFS_MIXED_BACKREF_REV)
4247                 add_tree_backref(extent_cache, buf->start, buf->start,
4248                                  0, 1);
4249         else
4250                 add_tree_backref(extent_cache, buf->start, 0,
4251                                  root_key->objectid, 1);
4252         return 0;
4253 }
4254
4255 /* as we fix the tree, we might be deleting blocks that
4256  * we're tracking for repair.  This hook makes sure we
4257  * remove any backrefs for blocks as we are fixing them.
4258  */
4259 static int free_extent_hook(struct btrfs_trans_handle *trans,
4260                             struct btrfs_root *root,
4261                             u64 bytenr, u64 num_bytes, u64 parent,
4262                             u64 root_objectid, u64 owner, u64 offset,
4263                             int refs_to_drop)
4264 {
4265         struct extent_record *rec;
4266         struct cache_extent *cache;
4267         int is_data;
4268         struct cache_tree *extent_cache = root->fs_info->fsck_extent_cache;
4269
4270         is_data = owner >= BTRFS_FIRST_FREE_OBJECTID;
4271         cache = lookup_cache_extent(extent_cache, bytenr, num_bytes);
4272         if (!cache)
4273                 return 0;
4274
4275         rec = container_of(cache, struct extent_record, cache);
4276         if (is_data) {
4277                 struct data_backref *back;
4278                 back = find_data_backref(rec, parent, root_objectid, owner,
4279                                          offset, 1, bytenr, num_bytes);
4280                 if (!back)
4281                         goto out;
4282                 if (back->node.found_ref) {
4283                         back->found_ref -= refs_to_drop;
4284                         if (rec->refs)
4285                                 rec->refs -= refs_to_drop;
4286                 }
4287                 if (back->node.found_extent_tree) {
4288                         back->num_refs -= refs_to_drop;
4289                         if (rec->extent_item_refs)
4290                                 rec->extent_item_refs -= refs_to_drop;
4291                 }
4292                 if (back->found_ref == 0)
4293                         back->node.found_ref = 0;
4294                 if (back->num_refs == 0)
4295                         back->node.found_extent_tree = 0;
4296
4297                 if (!back->node.found_extent_tree && back->node.found_ref) {
4298                         list_del(&back->node.list);
4299                         free(back);
4300                 }
4301         } else {
4302                 struct tree_backref *back;
4303                 back = find_tree_backref(rec, parent, root_objectid);
4304                 if (!back)
4305                         goto out;
4306                 if (back->node.found_ref) {
4307                         if (rec->refs)
4308                                 rec->refs--;
4309                         back->node.found_ref = 0;
4310                 }
4311                 if (back->node.found_extent_tree) {
4312                         if (rec->extent_item_refs)
4313                                 rec->extent_item_refs--;
4314                         back->node.found_extent_tree = 0;
4315                 }
4316                 if (!back->node.found_extent_tree && back->node.found_ref) {
4317                         list_del(&back->node.list);
4318                         free(back);
4319                 }
4320         }
4321         maybe_free_extent_rec(extent_cache, rec);
4322 out:
4323         return 0;
4324 }
4325
4326 static int delete_extent_records(struct btrfs_trans_handle *trans,
4327                                  struct btrfs_root *root,
4328                                  struct btrfs_path *path,
4329                                  u64 bytenr, u64 new_len)
4330 {
4331         struct btrfs_key key;
4332         struct btrfs_key found_key;
4333         struct extent_buffer *leaf;
4334         int ret;
4335         int slot;
4336
4337
4338         key.objectid = bytenr;
4339         key.type = (u8)-1;
4340         key.offset = (u64)-1;
4341
4342         while(1) {
4343                 ret = btrfs_search_slot(trans, root->fs_info->extent_root,
4344                                         &key, path, 0, 1);
4345                 if (ret < 0)
4346                         break;
4347
4348                 if (ret > 0) {
4349                         ret = 0;
4350                         if (path->slots[0] == 0)
4351                                 break;
4352                         path->slots[0]--;
4353                 }
4354                 ret = 0;
4355
4356                 leaf = path->nodes[0];
4357                 slot = path->slots[0];
4358
4359                 btrfs_item_key_to_cpu(leaf, &found_key, slot);
4360                 if (found_key.objectid != bytenr)
4361                         break;
4362
4363                 if (found_key.type != BTRFS_EXTENT_ITEM_KEY &&
4364                     found_key.type != BTRFS_METADATA_ITEM_KEY &&
4365                     found_key.type != BTRFS_TREE_BLOCK_REF_KEY &&
4366                     found_key.type != BTRFS_EXTENT_DATA_REF_KEY &&
4367                     found_key.type != BTRFS_EXTENT_REF_V0_KEY &&
4368                     found_key.type != BTRFS_SHARED_BLOCK_REF_KEY &&
4369                     found_key.type != BTRFS_SHARED_DATA_REF_KEY) {
4370                         btrfs_release_path(path);
4371                         if (found_key.type == 0) {
4372                                 if (found_key.offset == 0)
4373                                         break;
4374                                 key.offset = found_key.offset - 1;
4375                                 key.type = found_key.type;
4376                         }
4377                         key.type = found_key.type - 1;
4378                         key.offset = (u64)-1;
4379                         continue;
4380                 }
4381
4382                 fprintf(stderr, "repair deleting extent record: key %Lu %u %Lu\n",
4383                         found_key.objectid, found_key.type, found_key.offset);
4384
4385                 ret = btrfs_del_item(trans, root->fs_info->extent_root, path);
4386                 if (ret)
4387                         break;
4388                 btrfs_release_path(path);
4389
4390                 if (found_key.type == BTRFS_EXTENT_ITEM_KEY ||
4391                     found_key.type == BTRFS_METADATA_ITEM_KEY) {
4392                         u64 bytes = (found_key.type == BTRFS_EXTENT_ITEM_KEY) ?
4393                                 found_key.offset : root->leafsize;
4394
4395                         ret = btrfs_update_block_group(trans, root, bytenr,
4396                                                        bytes, 0, 0);
4397                         if (ret)
4398                                 break;
4399                 }
4400         }
4401
4402         btrfs_release_path(path);
4403         return ret;
4404 }
4405
4406 /*
4407  * for a single backref, this will allocate a new extent
4408  * and add the backref to it.
4409  */
4410 static int record_extent(struct btrfs_trans_handle *trans,
4411                          struct btrfs_fs_info *info,
4412                          struct btrfs_path *path,
4413                          struct extent_record *rec,
4414                          struct extent_backref *back,
4415                          int allocated, u64 flags)
4416 {
4417         int ret;
4418         struct btrfs_root *extent_root = info->extent_root;
4419         struct extent_buffer *leaf;
4420         struct btrfs_key ins_key;
4421         struct btrfs_extent_item *ei;
4422         struct tree_backref *tback;
4423         struct data_backref *dback;
4424         struct btrfs_tree_block_info *bi;
4425
4426         if (!back->is_data)
4427                 rec->max_size = max_t(u64, rec->max_size,
4428                                     info->extent_root->leafsize);
4429
4430         if (!allocated) {
4431                 u32 item_size = sizeof(*ei);
4432
4433                 if (!back->is_data)
4434                         item_size += sizeof(*bi);
4435
4436                 ins_key.objectid = rec->start;
4437                 ins_key.offset = rec->max_size;
4438                 ins_key.type = BTRFS_EXTENT_ITEM_KEY;
4439
4440                 ret = btrfs_insert_empty_item(trans, extent_root, path,
4441                                         &ins_key, item_size);
4442                 if (ret)
4443                         goto fail;
4444
4445                 leaf = path->nodes[0];
4446                 ei = btrfs_item_ptr(leaf, path->slots[0],
4447                                     struct btrfs_extent_item);
4448
4449                 btrfs_set_extent_refs(leaf, ei, 0);
4450                 btrfs_set_extent_generation(leaf, ei, rec->generation);
4451
4452                 if (back->is_data) {
4453                         btrfs_set_extent_flags(leaf, ei,
4454                                                BTRFS_EXTENT_FLAG_DATA);
4455                 } else {
4456                         struct btrfs_disk_key copy_key;;
4457
4458                         tback = (struct tree_backref *)back;
4459                         bi = (struct btrfs_tree_block_info *)(ei + 1);
4460                         memset_extent_buffer(leaf, 0, (unsigned long)bi,
4461                                              sizeof(*bi));
4462
4463                         btrfs_set_disk_key_objectid(&copy_key,
4464                                                     rec->info_objectid);
4465                         btrfs_set_disk_key_type(&copy_key, 0);
4466                         btrfs_set_disk_key_offset(&copy_key, 0);
4467
4468                         btrfs_set_tree_block_level(leaf, bi, rec->info_level);
4469                         btrfs_set_tree_block_key(leaf, bi, &copy_key);
4470
4471                         btrfs_set_extent_flags(leaf, ei,
4472                                                BTRFS_EXTENT_FLAG_TREE_BLOCK | flags);
4473                 }
4474
4475                 btrfs_mark_buffer_dirty(leaf);
4476                 ret = btrfs_update_block_group(trans, extent_root, rec->start,
4477                                                rec->max_size, 1, 0);
4478                 if (ret)
4479                         goto fail;
4480                 btrfs_release_path(path);
4481         }
4482
4483         if (back->is_data) {
4484                 u64 parent;
4485                 int i;
4486
4487                 dback = (struct data_backref *)back;
4488                 if (back->full_backref)
4489                         parent = dback->parent;
4490                 else
4491                         parent = 0;
4492
4493                 for (i = 0; i < dback->found_ref; i++) {
4494                         /* if parent != 0, we're doing a full backref
4495                          * passing BTRFS_FIRST_FREE_OBJECTID as the owner
4496                          * just makes the backref allocator create a data
4497                          * backref
4498                          */
4499                         ret = btrfs_inc_extent_ref(trans, info->extent_root,
4500                                                    rec->start, rec->max_size,
4501                                                    parent,
4502                                                    dback->root,
4503                                                    parent ?
4504                                                    BTRFS_FIRST_FREE_OBJECTID :
4505                                                    dback->owner,
4506                                                    dback->offset);
4507                         if (ret)
4508                                 break;
4509                 }
4510                 fprintf(stderr, "adding new data backref"
4511                                 " on %llu %s %llu owner %llu"
4512                                 " offset %llu found %d\n",
4513                                 (unsigned long long)rec->start,
4514                                 back->full_backref ?
4515                                 "parent" : "root",
4516                                 back->full_backref ?
4517                                 (unsigned long long)parent :
4518                                 (unsigned long long)dback->root,
4519                                 (unsigned long long)dback->owner,
4520                                 (unsigned long long)dback->offset,
4521                                 dback->found_ref);
4522         } else {
4523                 u64 parent;
4524
4525                 tback = (struct tree_backref *)back;
4526                 if (back->full_backref)
4527                         parent = tback->parent;
4528                 else
4529                         parent = 0;
4530
4531                 ret = btrfs_inc_extent_ref(trans, info->extent_root,
4532                                            rec->start, rec->max_size,
4533                                            parent, tback->root, 0, 0);
4534                 fprintf(stderr, "adding new tree backref on "
4535                         "start %llu len %llu parent %llu root %llu\n",
4536                         rec->start, rec->max_size, tback->parent, tback->root);
4537         }
4538         if (ret)
4539                 goto fail;
4540 fail:
4541         btrfs_release_path(path);
4542         return ret;
4543 }
4544
4545 struct extent_entry {
4546         u64 bytenr;
4547         u64 bytes;
4548         int count;
4549         int broken;
4550         struct list_head list;
4551 };
4552
4553 static struct extent_entry *find_entry(struct list_head *entries,
4554                                        u64 bytenr, u64 bytes)
4555 {
4556         struct extent_entry *entry = NULL;
4557
4558         list_for_each_entry(entry, entries, list) {
4559                 if (entry->bytenr == bytenr && entry->bytes == bytes)
4560                         return entry;
4561         }
4562
4563         return NULL;
4564 }
4565
4566 static struct extent_entry *find_most_right_entry(struct list_head *entries)
4567 {
4568         struct extent_entry *entry, *best = NULL, *prev = NULL;
4569
4570         list_for_each_entry(entry, entries, list) {
4571                 if (!prev) {
4572                         prev = entry;
4573                         continue;
4574                 }
4575
4576                 /*
4577                  * If there are as many broken entries as entries then we know
4578                  * not to trust this particular entry.
4579                  */
4580                 if (entry->broken == entry->count)
4581                         continue;
4582
4583                 /*
4584                  * If our current entry == best then we can't be sure our best
4585                  * is really the best, so we need to keep searching.
4586                  */
4587                 if (best && best->count == entry->count) {
4588                         prev = entry;
4589                         best = NULL;
4590                         continue;
4591                 }
4592
4593                 /* Prev == entry, not good enough, have to keep searching */
4594                 if (!prev->broken && prev->count == entry->count)
4595                         continue;
4596
4597                 if (!best)
4598                         best = (prev->count > entry->count) ? prev : entry;
4599                 else if (best->count < entry->count)
4600                         best = entry;
4601                 prev = entry;
4602         }
4603
4604         return best;
4605 }
4606
4607 static int repair_ref(struct btrfs_trans_handle *trans,
4608                       struct btrfs_fs_info *info, struct btrfs_path *path,
4609                       struct data_backref *dback, struct extent_entry *entry)
4610 {
4611         struct btrfs_root *root;
4612         struct btrfs_file_extent_item *fi;
4613         struct extent_buffer *leaf;
4614         struct btrfs_key key;
4615         u64 bytenr, bytes;
4616         int ret;
4617
4618         key.objectid = dback->root;
4619         key.type = BTRFS_ROOT_ITEM_KEY;
4620         key.offset = (u64)-1;
4621         root = btrfs_read_fs_root(info, &key);
4622         if (IS_ERR(root)) {
4623                 fprintf(stderr, "Couldn't find root for our ref\n");
4624                 return -EINVAL;
4625         }
4626
4627         /*
4628          * The backref points to the original offset of the extent if it was
4629          * split, so we need to search down to the offset we have and then walk
4630          * forward until we find the backref we're looking for.
4631          */
4632         key.objectid = dback->owner;
4633         key.type = BTRFS_EXTENT_DATA_KEY;
4634         key.offset = dback->offset;
4635         ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
4636         if (ret < 0) {
4637                 fprintf(stderr, "Error looking up ref %d\n", ret);
4638                 return ret;
4639         }
4640
4641         while (1) {
4642                 if (path->slots[0] >= btrfs_header_nritems(path->nodes[0])) {
4643                         ret = btrfs_next_leaf(root, path);
4644                         if (ret) {
4645                                 fprintf(stderr, "Couldn't find our ref, next\n");
4646                                 return -EINVAL;
4647                         }
4648                 }
4649                 leaf = path->nodes[0];
4650                 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
4651                 if (key.objectid != dback->owner ||
4652                     key.type != BTRFS_EXTENT_DATA_KEY) {
4653                         fprintf(stderr, "Couldn't find our ref, search\n");
4654                         return -EINVAL;
4655                 }
4656                 fi = btrfs_item_ptr(leaf, path->slots[0],
4657                                     struct btrfs_file_extent_item);
4658                 bytenr = btrfs_file_extent_disk_bytenr(leaf, fi);
4659                 bytes = btrfs_file_extent_disk_num_bytes(leaf, fi);
4660
4661                 if (bytenr == dback->disk_bytenr && bytes == dback->bytes)
4662                         break;
4663                 path->slots[0]++;
4664         }
4665
4666         btrfs_release_path(path);
4667
4668         /*
4669          * Have to make sure that this root gets updated when we commit the
4670          * transaction
4671          */
4672         root->track_dirty = 1;
4673         if (root->last_trans != trans->transid) {
4674                 root->last_trans = trans->transid;
4675                 root->commit_root = root->node;
4676                 extent_buffer_get(root->node);
4677         }
4678
4679         /*
4680          * Ok we have the key of the file extent we want to fix, now we can cow
4681          * down to the thing and fix it.
4682          */
4683         ret = btrfs_search_slot(trans, root, &key, path, 0, 1);
4684         if (ret < 0) {
4685                 fprintf(stderr, "Error cowing down to ref [%Lu, %u, %Lu]: %d\n",
4686                         key.objectid, key.type, key.offset, ret);
4687                 return ret;
4688         }
4689         if (ret > 0) {
4690                 fprintf(stderr, "Well that's odd, we just found this key "
4691                         "[%Lu, %u, %Lu]\n", key.objectid, key.type,
4692                         key.offset);
4693                 return -EINVAL;
4694         }
4695         leaf = path->nodes[0];
4696         fi = btrfs_item_ptr(leaf, path->slots[0],
4697                             struct btrfs_file_extent_item);
4698
4699         if (btrfs_file_extent_compression(leaf, fi) &&
4700             dback->disk_bytenr != entry->bytenr) {
4701                 fprintf(stderr, "Ref doesn't match the record start and is "
4702                         "compressed, please take a btrfs-image of this file "
4703                         "system and send it to a btrfs developer so they can "
4704                         "complete this functionality for bytenr %Lu\n",
4705                         dback->disk_bytenr);
4706                 return -EINVAL;
4707         }
4708
4709         if (dback->node.broken && dback->disk_bytenr != entry->bytenr) {
4710                 btrfs_set_file_extent_disk_bytenr(leaf, fi, entry->bytenr);
4711         } else if (dback->disk_bytenr > entry->bytenr) {
4712                 u64 off_diff, offset;
4713
4714                 off_diff = dback->disk_bytenr - entry->bytenr;
4715                 offset = btrfs_file_extent_offset(leaf, fi);
4716                 if (dback->disk_bytenr + offset +
4717                     btrfs_file_extent_num_bytes(leaf, fi) >
4718                     entry->bytenr + entry->bytes) {
4719                         fprintf(stderr, "Ref is past the entry end, please "
4720                                 "take a btrfs-image of this file system and "
4721                                 "send it to a btrfs developer, ref %Lu\n",
4722                                 dback->disk_bytenr);
4723                         return -EINVAL;
4724                 }
4725                 offset += off_diff;
4726                 btrfs_set_file_extent_disk_bytenr(leaf, fi, entry->bytenr);
4727                 btrfs_set_file_extent_offset(leaf, fi, offset);
4728         } else if (dback->disk_bytenr < entry->bytenr) {
4729                 u64 offset;
4730
4731                 offset = btrfs_file_extent_offset(leaf, fi);
4732                 if (dback->disk_bytenr + offset < entry->bytenr) {
4733                         fprintf(stderr, "Ref is before the entry start, please"
4734                                 " take a btrfs-image of this file system and "
4735                                 "send it to a btrfs developer, ref %Lu\n",
4736                                 dback->disk_bytenr);
4737                         return -EINVAL;
4738                 }
4739
4740                 offset += dback->disk_bytenr;
4741                 offset -= entry->bytenr;
4742                 btrfs_set_file_extent_disk_bytenr(leaf, fi, entry->bytenr);
4743                 btrfs_set_file_extent_offset(leaf, fi, offset);
4744         }
4745
4746         btrfs_set_file_extent_disk_num_bytes(leaf, fi, entry->bytes);
4747
4748         /*
4749          * Chances are if disk_num_bytes were wrong then so is ram_bytes, but
4750          * only do this if we aren't using compression, otherwise it's a
4751          * trickier case.
4752          */
4753         if (!btrfs_file_extent_compression(leaf, fi))
4754                 btrfs_set_file_extent_ram_bytes(leaf, fi, entry->bytes);
4755         else
4756                 printf("ram bytes may be wrong?\n");
4757         btrfs_mark_buffer_dirty(leaf);
4758         btrfs_release_path(path);
4759         return 0;
4760 }
4761
4762 static int verify_backrefs(struct btrfs_trans_handle *trans,
4763                            struct btrfs_fs_info *info, struct btrfs_path *path,
4764                            struct extent_record *rec)
4765 {
4766         struct extent_backref *back;
4767         struct data_backref *dback;
4768         struct extent_entry *entry, *best = NULL;
4769         LIST_HEAD(entries);
4770         int nr_entries = 0;
4771         int broken_entries = 0;
4772         int ret = 0;
4773         short mismatch = 0;
4774
4775         /*
4776          * Metadata is easy and the backrefs should always agree on bytenr and
4777          * size, if not we've got bigger issues.
4778          */
4779         if (rec->metadata)
4780                 return 0;
4781
4782         list_for_each_entry(back, &rec->backrefs, list) {
4783                 dback = (struct data_backref *)back;
4784                 /*
4785                  * We only pay attention to backrefs that we found a real
4786                  * backref for.
4787                  */
4788                 if (dback->found_ref == 0)
4789                         continue;
4790                 if (back->full_backref)
4791                         continue;
4792
4793                 /*
4794                  * For now we only catch when the bytes don't match, not the
4795                  * bytenr.  We can easily do this at the same time, but I want
4796                  * to have a fs image to test on before we just add repair
4797                  * functionality willy-nilly so we know we won't screw up the
4798                  * repair.
4799                  */
4800
4801                 entry = find_entry(&entries, dback->disk_bytenr,
4802                                    dback->bytes);
4803                 if (!entry) {
4804                         entry = malloc(sizeof(struct extent_entry));
4805                         if (!entry) {
4806                                 ret = -ENOMEM;
4807                                 goto out;
4808                         }
4809                         memset(entry, 0, sizeof(*entry));
4810                         entry->bytenr = dback->disk_bytenr;
4811                         entry->bytes = dback->bytes;
4812                         list_add_tail(&entry->list, &entries);
4813                         nr_entries++;
4814                 }
4815
4816                 /*
4817                  * If we only have on entry we may think the entries agree when
4818                  * in reality they don't so we have to do some extra checking.
4819                  */
4820                 if (dback->disk_bytenr != rec->start ||
4821                     dback->bytes != rec->nr || back->broken)
4822                         mismatch = 1;
4823
4824                 if (back->broken) {
4825                         entry->broken++;
4826                         broken_entries++;
4827                 }
4828
4829                 entry->count++;
4830         }
4831
4832         /* Yay all the backrefs agree, carry on good sir */
4833         if (nr_entries <= 1 && !mismatch)
4834                 goto out;
4835
4836         fprintf(stderr, "attempting to repair backref discrepency for bytenr "
4837                 "%Lu\n", rec->start);
4838
4839         /*
4840          * First we want to see if the backrefs can agree amongst themselves who
4841          * is right, so figure out which one of the entries has the highest
4842          * count.
4843          */
4844         best = find_most_right_entry(&entries);
4845
4846         /*
4847          * Ok so we may have an even split between what the backrefs think, so
4848          * this is where we use the extent ref to see what it thinks.
4849          */
4850         if (!best) {
4851                 entry = find_entry(&entries, rec->start, rec->nr);
4852                 if (!entry && (!broken_entries || !rec->found_rec)) {
4853                         fprintf(stderr, "Backrefs don't agree with each other "
4854                                 "and extent record doesn't agree with anybody,"
4855                                 " so we can't fix bytenr %Lu bytes %Lu\n",
4856                                 rec->start, rec->nr);
4857                         ret = -EINVAL;
4858                         goto out;
4859                 } else if (!entry) {
4860                         /*
4861                          * Ok our backrefs were broken, we'll assume this is the
4862                          * correct value and add an entry for this range.
4863                          */
4864                         entry = malloc(sizeof(struct extent_entry));
4865                         if (!entry) {
4866                                 ret = -ENOMEM;
4867                                 goto out;
4868                         }
4869                         memset(entry, 0, sizeof(*entry));
4870                         entry->bytenr = rec->start;
4871                         entry->bytes = rec->nr;
4872                         list_add_tail(&entry->list, &entries);
4873                         nr_entries++;
4874                 }
4875                 entry->count++;
4876                 best = find_most_right_entry(&entries);
4877                 if (!best) {
4878                         fprintf(stderr, "Backrefs and extent record evenly "
4879                                 "split on who is right, this is going to "
4880                                 "require user input to fix bytenr %Lu bytes "
4881                                 "%Lu\n", rec->start, rec->nr);
4882                         ret = -EINVAL;
4883                         goto out;
4884                 }
4885         }
4886
4887         /*
4888          * I don't think this can happen currently as we'll abort() if we catch
4889          * this case higher up, but in case somebody removes that we still can't
4890          * deal with it properly here yet, so just bail out of that's the case.
4891          */
4892         if (best->bytenr != rec->start) {
4893                 fprintf(stderr, "Extent start and backref starts don't match, "
4894                         "please use btrfs-image on this file system and send "
4895                         "it to a btrfs developer so they can make fsck fix "
4896                         "this particular case.  bytenr is %Lu, bytes is %Lu\n",
4897                         rec->start, rec->nr);
4898                 ret = -EINVAL;
4899                 goto out;
4900         }
4901
4902         /*
4903          * Ok great we all agreed on an extent record, let's go find the real
4904          * references and fix up the ones that don't match.
4905          */
4906         list_for_each_entry(back, &rec->backrefs, list) {
4907                 dback = (struct data_backref *)back;
4908
4909                 /*
4910                  * Still ignoring backrefs that don't have a real ref attached
4911                  * to them.
4912                  */
4913                 if (dback->found_ref == 0)
4914                         continue;
4915                 if (back->full_backref)
4916                         continue;
4917
4918                 if (dback->bytes == best->bytes &&
4919                     dback->disk_bytenr == best->bytenr)
4920                         continue;
4921
4922                 ret = repair_ref(trans, info, path, dback, best);
4923                 if (ret)
4924                         goto out;
4925         }
4926
4927         /*
4928          * Ok we messed with the actual refs, which means we need to drop our
4929          * entire cache and go back and rescan.  I know this is a huge pain and
4930          * adds a lot of extra work, but it's the only way to be safe.  Once all
4931          * the backrefs agree we may not need to do anything to the extent
4932          * record itself.
4933          */
4934         ret = -EAGAIN;
4935 out:
4936         while (!list_empty(&entries)) {
4937                 entry = list_entry(entries.next, struct extent_entry, list);
4938                 list_del_init(&entry->list);
4939                 free(entry);
4940         }
4941         return ret;
4942 }
4943
4944 static int process_duplicates(struct btrfs_root *root,
4945                               struct cache_tree *extent_cache,
4946                               struct extent_record *rec)
4947 {
4948         struct extent_record *good, *tmp;
4949         struct cache_extent *cache;
4950         int ret;
4951
4952         /*
4953          * If we found a extent record for this extent then return, or if we
4954          * have more than one duplicate we are likely going to need to delete
4955          * something.
4956          */
4957         if (rec->found_rec || rec->num_duplicates > 1)
4958                 return 0;
4959
4960         /* Shouldn't happen but just in case */
4961         BUG_ON(!rec->num_duplicates);
4962
4963         /*
4964          * So this happens if we end up with a backref that doesn't match the
4965          * actual extent entry.  So either the backref is bad or the extent
4966          * entry is bad.  Either way we want to have the extent_record actually
4967          * reflect what we found in the extent_tree, so we need to take the
4968          * duplicate out and use that as the extent_record since the only way we
4969          * get a duplicate is if we find a real life BTRFS_EXTENT_ITEM_KEY.
4970          */
4971         remove_cache_extent(extent_cache, &rec->cache);
4972
4973         good = list_entry(rec->dups.next, struct extent_record, list);
4974         list_del_init(&good->list);
4975         INIT_LIST_HEAD(&good->backrefs);
4976         INIT_LIST_HEAD(&good->dups);
4977         good->cache.start = good->start;
4978         good->cache.size = good->nr;
4979         good->content_checked = 0;
4980         good->owner_ref_checked = 0;
4981         good->num_duplicates = 0;
4982         good->refs = rec->refs;
4983         list_splice_init(&rec->backrefs, &good->backrefs);
4984         while (1) {
4985                 cache = lookup_cache_extent(extent_cache, good->start,
4986                                             good->nr);
4987                 if (!cache)
4988                         break;
4989                 tmp = container_of(cache, struct extent_record, cache);
4990
4991                 /*
4992                  * If we find another overlapping extent and it's found_rec is
4993                  * set then it's a duplicate and we need to try and delete
4994                  * something.
4995                  */
4996                 if (tmp->found_rec || tmp->num_duplicates > 0) {
4997                         if (list_empty(&good->list))
4998                                 list_add_tail(&good->list,
4999                                               &duplicate_extents);
5000                         good->num_duplicates += tmp->num_duplicates + 1;
5001                         list_splice_init(&tmp->dups, &good->dups);
5002                         list_del_init(&tmp->list);
5003                         list_add_tail(&tmp->list, &good->dups);
5004                         remove_cache_extent(extent_cache, &tmp->cache);
5005                         continue;
5006                 }
5007
5008                 /*
5009                  * Ok we have another non extent item backed extent rec, so lets
5010                  * just add it to this extent and carry on like we did above.
5011                  */
5012                 good->refs += tmp->refs;
5013                 list_splice_init(&tmp->backrefs, &good->backrefs);
5014                 remove_cache_extent(extent_cache, &tmp->cache);
5015                 free(tmp);
5016         }
5017         ret = insert_cache_extent(extent_cache, &good->cache);
5018         BUG_ON(ret);
5019         free(rec);
5020         return good->num_duplicates ? 0 : 1;
5021 }
5022
5023 static int delete_duplicate_records(struct btrfs_trans_handle *trans,
5024                                     struct btrfs_root *root,
5025                                     struct extent_record *rec)
5026 {
5027         LIST_HEAD(delete_list);
5028         struct btrfs_path *path;
5029         struct extent_record *tmp, *good, *n;
5030         int nr_del = 0;
5031         int ret = 0;
5032         struct btrfs_key key;
5033
5034         path = btrfs_alloc_path();
5035         if (!path) {
5036                 ret = -ENOMEM;
5037                 goto out;
5038         }
5039
5040         good = rec;
5041         /* Find the record that covers all of the duplicates. */
5042         list_for_each_entry(tmp, &rec->dups, list) {
5043                 if (good->start < tmp->start)
5044                         continue;
5045                 if (good->nr > tmp->nr)
5046                         continue;
5047
5048                 if (tmp->start + tmp->nr < good->start + good->nr) {
5049                         fprintf(stderr, "Ok we have overlapping extents that "
5050                                 "aren't completely covered by eachother, this "
5051                                 "is going to require more careful thought.  "
5052                                 "The extents are [%Lu-%Lu] and [%Lu-%Lu]\n",
5053                                 tmp->start, tmp->nr, good->start, good->nr);
5054                         abort();
5055                 }
5056                 good = tmp;
5057         }
5058
5059         if (good != rec)
5060                 list_add_tail(&rec->list, &delete_list);
5061
5062         list_for_each_entry_safe(tmp, n, &rec->dups, list) {
5063                 if (tmp == good)
5064                         continue;
5065                 list_move_tail(&tmp->list, &delete_list);
5066         }
5067
5068         root = root->fs_info->extent_root;
5069         list_for_each_entry(tmp, &delete_list, list) {
5070                 if (tmp->found_rec == 0)
5071                         continue;
5072                 key.objectid = tmp->start;
5073                 key.type = BTRFS_EXTENT_ITEM_KEY;
5074                 key.offset = tmp->nr;
5075
5076                 /* Shouldn't happen but just in case */
5077                 if (tmp->metadata) {
5078                         fprintf(stderr, "Well this shouldn't happen, extent "
5079                                 "record overlaps but is metadata? "
5080                                 "[%Lu, %Lu]\n", tmp->start, tmp->nr);
5081                         abort();
5082                 }
5083
5084                 ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
5085                 if (ret) {
5086                         if (ret > 0)
5087                                 ret = -EINVAL;
5088                         goto out;
5089                 }
5090                 ret = btrfs_del_item(trans, root, path);
5091                 if (ret)
5092                         goto out;
5093                 btrfs_release_path(path);
5094                 nr_del++;
5095         }
5096
5097 out:
5098         while (!list_empty(&delete_list)) {
5099                 tmp = list_entry(delete_list.next, struct extent_record, list);
5100                 list_del_init(&tmp->list);
5101                 if (tmp == rec)
5102                         continue;
5103                 free(tmp);
5104         }
5105
5106         while (!list_empty(&rec->dups)) {
5107                 tmp = list_entry(rec->dups.next, struct extent_record, list);
5108                 list_del_init(&tmp->list);
5109                 free(tmp);
5110         }
5111
5112         btrfs_free_path(path);
5113
5114         if (!ret && !nr_del)
5115                 rec->num_duplicates = 0;
5116
5117         return ret ? ret : nr_del;
5118 }
5119
5120 static int find_possible_backrefs(struct btrfs_trans_handle *trans,
5121                                   struct btrfs_fs_info *info,
5122                                   struct btrfs_path *path,
5123                                   struct cache_tree *extent_cache,
5124                                   struct extent_record *rec)
5125 {
5126         struct btrfs_root *root;
5127         struct extent_backref *back;
5128         struct data_backref *dback;
5129         struct cache_extent *cache;
5130         struct btrfs_file_extent_item *fi;
5131         struct btrfs_key key;
5132         u64 bytenr, bytes;
5133         int ret;
5134
5135         list_for_each_entry(back, &rec->backrefs, list) {
5136                 dback = (struct data_backref *)back;
5137
5138                 /* We found this one, we don't need to do a lookup */
5139                 if (dback->found_ref)
5140                         continue;
5141                 /* Don't care about full backrefs (poor unloved backrefs) */
5142                 if (back->full_backref)
5143                         continue;
5144                 key.objectid = dback->root;
5145                 key.type = BTRFS_ROOT_ITEM_KEY;
5146                 key.offset = (u64)-1;
5147
5148                 root = btrfs_read_fs_root(info, &key);
5149
5150                 /* No root, definitely a bad ref, skip */
5151                 if (IS_ERR(root) && PTR_ERR(root) == -ENOENT)
5152                         continue;
5153                 /* Other err, exit */
5154                 if (IS_ERR(root))
5155                         return PTR_ERR(root);
5156
5157                 key.objectid = dback->owner;
5158                 key.type = BTRFS_EXTENT_DATA_KEY;
5159                 key.offset = dback->offset;
5160                 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
5161                 if (ret) {
5162                         btrfs_release_path(path);
5163                         if (ret < 0)
5164                                 return ret;
5165                         /* Didn't find it, we can carry on */
5166                         ret = 0;
5167                         continue;
5168                 }
5169
5170                 fi = btrfs_item_ptr(path->nodes[0], path->slots[0],
5171                                     struct btrfs_file_extent_item);
5172                 bytenr = btrfs_file_extent_disk_bytenr(path->nodes[0], fi);
5173                 bytes = btrfs_file_extent_disk_num_bytes(path->nodes[0], fi);
5174                 btrfs_release_path(path);
5175                 cache = lookup_cache_extent(extent_cache, bytenr, 1);
5176                 if (cache) {
5177                         struct extent_record *tmp;
5178                         tmp = container_of(cache, struct extent_record, cache);
5179
5180                         /*
5181                          * If we found an extent record for the bytenr for this
5182                          * particular backref then we can't add it to our
5183                          * current extent record.  We only want to add backrefs
5184                          * that don't have a corresponding extent item in the
5185                          * extent tree since they likely belong to this record
5186                          * and we need to fix it if it doesn't match bytenrs.
5187                          */
5188                         if  (tmp->found_rec)
5189                                 continue;
5190                 }
5191
5192                 dback->found_ref += 1;
5193                 dback->disk_bytenr = bytenr;
5194                 dback->bytes = bytes;
5195
5196                 /*
5197                  * Set this so the verify backref code knows not to trust the
5198                  * values in this backref.
5199                  */
5200                 back->broken = 1;
5201         }
5202
5203         return 0;
5204 }
5205
5206 /*
5207  * when an incorrect extent item is found, this will delete
5208  * all of the existing entries for it and recreate them
5209  * based on what the tree scan found.
5210  */
5211 static int fixup_extent_refs(struct btrfs_trans_handle *trans,
5212                              struct btrfs_fs_info *info,
5213                              struct cache_tree *extent_cache,
5214                              struct extent_record *rec)
5215 {
5216         int ret;
5217         struct btrfs_path *path;
5218         struct list_head *cur = rec->backrefs.next;
5219         struct cache_extent *cache;
5220         struct extent_backref *back;
5221         int allocated = 0;
5222         u64 flags = 0;
5223
5224         /*
5225          * remember our flags for recreating the extent.
5226          * FIXME, if we have cleared extent tree, we can not
5227          * lookup extent info in extent tree.
5228          */
5229         if (!init_extent_tree) {
5230                 ret = btrfs_lookup_extent_info(NULL, info->extent_root,
5231                                         rec->start, rec->max_size,
5232                                         rec->metadata, NULL, &flags);
5233                 if (ret < 0)
5234                         flags = 0;
5235         } else {
5236                 flags = 0;
5237         }
5238
5239         path = btrfs_alloc_path();
5240         if (!path)
5241                 return -ENOMEM;
5242
5243         if (rec->refs != rec->extent_item_refs && !rec->metadata) {
5244                 /*
5245                  * Sometimes the backrefs themselves are so broken they don't
5246                  * get attached to any meaningful rec, so first go back and
5247                  * check any of our backrefs that we couldn't find and throw
5248                  * them into the list if we find the backref so that
5249                  * verify_backrefs can figure out what to do.
5250                  */
5251                 ret = find_possible_backrefs(trans, info, path, extent_cache,
5252                                              rec);
5253                 if (ret < 0)
5254                         goto out;
5255         }
5256
5257         /* step one, make sure all of the backrefs agree */
5258         ret = verify_backrefs(trans, info, path, rec);
5259         if (ret < 0)
5260                 goto out;
5261
5262         /* step two, delete all the existing records */
5263         ret = delete_extent_records(trans, info->extent_root, path,
5264                                     rec->start, rec->max_size);
5265
5266         if (ret < 0)
5267                 goto out;
5268
5269         /* was this block corrupt?  If so, don't add references to it */
5270         cache = lookup_cache_extent(info->corrupt_blocks,
5271                                     rec->start, rec->max_size);
5272         if (cache) {
5273                 ret = 0;
5274                 goto out;
5275         }
5276
5277         /* step three, recreate all the refs we did find */
5278         while(cur != &rec->backrefs) {
5279                 back = list_entry(cur, struct extent_backref, list);
5280                 cur = cur->next;
5281
5282                 /*
5283                  * if we didn't find any references, don't create a
5284                  * new extent record
5285                  */
5286                 if (!back->found_ref)
5287                         continue;
5288
5289                 ret = record_extent(trans, info, path, rec, back, allocated, flags);
5290                 allocated = 1;
5291
5292                 if (ret)
5293                         goto out;
5294         }
5295 out:
5296         btrfs_free_path(path);
5297         return ret;
5298 }
5299
5300 /* right now we only prune from the extent allocation tree */
5301 static int prune_one_block(struct btrfs_trans_handle *trans,
5302                            struct btrfs_fs_info *info,
5303                            struct btrfs_corrupt_block *corrupt)
5304 {
5305         int ret;
5306         struct btrfs_path path;
5307         struct extent_buffer *eb;
5308         u64 found;
5309         int slot;
5310         int nritems;
5311         int level = corrupt->level + 1;
5312
5313         btrfs_init_path(&path);
5314 again:
5315         /* we want to stop at the parent to our busted block */
5316         path.lowest_level = level;
5317
5318         ret = btrfs_search_slot(trans, info->extent_root,
5319                                 &corrupt->key, &path, -1, 1);
5320
5321         if (ret < 0)
5322                 goto out;
5323
5324         eb = path.nodes[level];
5325         if (!eb) {
5326                 ret = -ENOENT;
5327                 goto out;
5328         }
5329
5330         /*
5331          * hopefully the search gave us the block we want to prune,
5332          * lets try that first
5333          */
5334         slot = path.slots[level];
5335         found =  btrfs_node_blockptr(eb, slot);
5336         if (found == corrupt->cache.start)
5337                 goto del_ptr;
5338
5339         nritems = btrfs_header_nritems(eb);
5340
5341         /* the search failed, lets scan this node and hope we find it */
5342         for (slot = 0; slot < nritems; slot++) {
5343                 found =  btrfs_node_blockptr(eb, slot);
5344                 if (found == corrupt->cache.start)
5345                         goto del_ptr;
5346         }
5347         /*
5348          * we couldn't find the bad block.  TODO, search all the nodes for pointers
5349          * to this block
5350          */
5351         if (eb == info->extent_root->node) {
5352                 ret = -ENOENT;
5353                 goto out;
5354         } else {
5355                 level++;
5356                 btrfs_release_path(&path);
5357                 goto again;
5358         }
5359
5360 del_ptr:
5361         printk("deleting pointer to block %Lu\n", corrupt->cache.start);
5362         ret = btrfs_del_ptr(trans, info->extent_root, &path, level, slot);
5363
5364 out:
5365         btrfs_release_path(&path);
5366         return ret;
5367 }
5368
5369 static int prune_corrupt_blocks(struct btrfs_trans_handle *trans,
5370                                 struct btrfs_fs_info *info)
5371 {
5372         struct cache_extent *cache;
5373         struct btrfs_corrupt_block *corrupt;
5374
5375         cache = search_cache_extent(info->corrupt_blocks, 0);
5376         while (1) {
5377                 if (!cache)
5378                         break;
5379                 corrupt = container_of(cache, struct btrfs_corrupt_block, cache);
5380                 prune_one_block(trans, info, corrupt);
5381                 cache = next_cache_extent(cache);
5382         }
5383         return 0;
5384 }
5385
5386 static void free_corrupt_block(struct cache_extent *cache)
5387 {
5388         struct btrfs_corrupt_block *corrupt;
5389
5390         corrupt = container_of(cache, struct btrfs_corrupt_block, cache);
5391         free(corrupt);
5392 }
5393
5394 FREE_EXTENT_CACHE_BASED_TREE(corrupt_blocks, free_corrupt_block);
5395
5396 static void reset_cached_block_groups(struct btrfs_fs_info *fs_info)
5397 {
5398         struct btrfs_block_group_cache *cache;
5399         u64 start, end;
5400         int ret;
5401
5402         while (1) {
5403                 ret = find_first_extent_bit(&fs_info->free_space_cache, 0,
5404                                             &start, &end, EXTENT_DIRTY);
5405                 if (ret)
5406                         break;
5407                 clear_extent_dirty(&fs_info->free_space_cache, start, end,
5408                                    GFP_NOFS);
5409         }
5410
5411         start = 0;
5412         while (1) {
5413                 cache = btrfs_lookup_first_block_group(fs_info, start);
5414                 if (!cache)
5415                         break;
5416                 if (cache->cached)
5417                         cache->cached = 0;
5418                 start = cache->key.objectid + cache->key.offset;
5419         }
5420 }
5421
5422 static int check_extent_refs(struct btrfs_trans_handle *trans,
5423                              struct btrfs_root *root,
5424                              struct cache_tree *extent_cache)
5425 {
5426         struct extent_record *rec;
5427         struct cache_extent *cache;
5428         int err = 0;
5429         int ret = 0;
5430         int fixed = 0;
5431         int had_dups = 0;
5432
5433         if (repair) {
5434                 /*
5435                  * if we're doing a repair, we have to make sure
5436                  * we don't allocate from the problem extents.
5437                  * In the worst case, this will be all the
5438                  * extents in the FS
5439                  */
5440                 cache = search_cache_extent(extent_cache, 0);
5441                 while(cache) {
5442                         rec = container_of(cache, struct extent_record, cache);
5443                         btrfs_pin_extent(root->fs_info,
5444                                          rec->start, rec->max_size);
5445                         cache = next_cache_extent(cache);
5446                 }
5447
5448                 /* pin down all the corrupted blocks too */
5449                 cache = search_cache_extent(root->fs_info->corrupt_blocks, 0);
5450                 while(cache) {
5451                         btrfs_pin_extent(root->fs_info,
5452                                          cache->start, cache->size);
5453                         cache = next_cache_extent(cache);
5454                 }
5455                 prune_corrupt_blocks(trans, root->fs_info);
5456                 reset_cached_block_groups(root->fs_info);
5457         }
5458
5459         /*
5460          * We need to delete any duplicate entries we find first otherwise we
5461          * could mess up the extent tree when we have backrefs that actually
5462          * belong to a different extent item and not the weird duplicate one.
5463          */
5464         while (repair && !list_empty(&duplicate_extents)) {
5465                 rec = list_entry(duplicate_extents.next, struct extent_record,
5466                                  list);
5467                 list_del_init(&rec->list);
5468
5469                 /* Sometimes we can find a backref before we find an actual
5470                  * extent, so we need to process it a little bit to see if there
5471                  * truly are multiple EXTENT_ITEM_KEY's for the same range, or
5472                  * if this is a backref screwup.  If we need to delete stuff
5473                  * process_duplicates() will return 0, otherwise it will return
5474                  * 1 and we
5475                  */
5476                 if (process_duplicates(root, extent_cache, rec))
5477                         continue;
5478                 ret = delete_duplicate_records(trans, root, rec);
5479                 if (ret < 0)
5480                         return ret;
5481                 /*
5482                  * delete_duplicate_records will return the number of entries
5483                  * deleted, so if it's greater than 0 then we know we actually
5484                  * did something and we need to remove.
5485                  */
5486                 if (ret)
5487                         had_dups = 1;
5488         }
5489
5490         if (had_dups)
5491                 return -EAGAIN;
5492
5493         while(1) {
5494                 fixed = 0;
5495                 cache = search_cache_extent(extent_cache, 0);
5496                 if (!cache)
5497                         break;
5498                 rec = container_of(cache, struct extent_record, cache);
5499                 if (rec->num_duplicates) {
5500                         fprintf(stderr, "extent item %llu has multiple extent "
5501                                 "items\n", (unsigned long long)rec->start);
5502                         err = 1;
5503                 }
5504
5505                 if (rec->refs != rec->extent_item_refs) {
5506                         fprintf(stderr, "ref mismatch on [%llu %llu] ",
5507                                 (unsigned long long)rec->start,
5508                                 (unsigned long long)rec->nr);
5509                         fprintf(stderr, "extent item %llu, found %llu\n",
5510                                 (unsigned long long)rec->extent_item_refs,
5511                                 (unsigned long long)rec->refs);
5512                         if (!fixed && repair) {
5513                                 ret = fixup_extent_refs(trans, root->fs_info,
5514                                                         extent_cache, rec);
5515                                 if (ret)
5516                                         goto repair_abort;
5517                                 fixed = 1;
5518                         }
5519                         err = 1;
5520
5521                 }
5522                 if (all_backpointers_checked(rec, 1)) {
5523                         fprintf(stderr, "backpointer mismatch on [%llu %llu]\n",
5524                                 (unsigned long long)rec->start,
5525                                 (unsigned long long)rec->nr);
5526
5527                         if (!fixed && repair) {
5528                                 ret = fixup_extent_refs(trans, root->fs_info,
5529                                                         extent_cache, rec);
5530                                 if (ret)
5531                                         goto repair_abort;
5532                                 fixed = 1;
5533                         }
5534
5535                         err = 1;
5536                 }
5537                 if (!rec->owner_ref_checked) {
5538                         fprintf(stderr, "owner ref check failed [%llu %llu]\n",
5539                                 (unsigned long long)rec->start,
5540                                 (unsigned long long)rec->nr);
5541                         if (!fixed && repair) {
5542                                 ret = fixup_extent_refs(trans, root->fs_info,
5543                                                         extent_cache, rec);
5544                                 if (ret)
5545                                         goto repair_abort;
5546                                 fixed = 1;
5547                         }
5548                         err = 1;
5549                 }
5550
5551                 remove_cache_extent(extent_cache, cache);
5552                 free_all_extent_backrefs(rec);
5553                 free(rec);
5554         }
5555 repair_abort:
5556         if (repair) {
5557                 if (ret && ret != -EAGAIN) {
5558                         fprintf(stderr, "failed to repair damaged filesystem, aborting\n");
5559                         exit(1);
5560                 } else if (!ret) {
5561                         btrfs_fix_block_accounting(trans, root);
5562                 }
5563                 if (err)
5564                         fprintf(stderr, "repaired damaged extent references\n");
5565                 return ret;
5566         }
5567         return err;
5568 }
5569
5570 u64 calc_stripe_length(u64 type, u64 length, int num_stripes)
5571 {
5572         u64 stripe_size;
5573
5574         if (type & BTRFS_BLOCK_GROUP_RAID0) {
5575                 stripe_size = length;
5576                 stripe_size /= num_stripes;
5577         } else if (type & BTRFS_BLOCK_GROUP_RAID10) {
5578                 stripe_size = length * 2;
5579                 stripe_size /= num_stripes;
5580         } else if (type & BTRFS_BLOCK_GROUP_RAID5) {
5581                 stripe_size = length;
5582                 stripe_size /= (num_stripes - 1);
5583         } else if (type & BTRFS_BLOCK_GROUP_RAID6) {
5584                 stripe_size = length;
5585                 stripe_size /= (num_stripes - 2);
5586         } else {
5587                 stripe_size = length;
5588         }
5589         return stripe_size;
5590 }
5591
5592 static int check_chunk_refs(struct chunk_record *chunk_rec,
5593                             struct block_group_tree *block_group_cache,
5594                             struct device_extent_tree *dev_extent_cache,
5595                             int silent)
5596 {
5597         struct cache_extent *block_group_item;
5598         struct block_group_record *block_group_rec;
5599         struct cache_extent *dev_extent_item;
5600         struct device_extent_record *dev_extent_rec;
5601         u64 devid;
5602         u64 offset;
5603         u64 length;
5604         int i;
5605         int ret = 0;
5606
5607         block_group_item = lookup_cache_extent(&block_group_cache->tree,
5608                                                chunk_rec->offset,
5609                                                chunk_rec->length);
5610         if (block_group_item) {
5611                 block_group_rec = container_of(block_group_item,
5612                                                struct block_group_record,
5613                                                cache);
5614                 if (chunk_rec->length != block_group_rec->offset ||
5615                     chunk_rec->offset != block_group_rec->objectid ||
5616                     chunk_rec->type_flags != block_group_rec->flags) {
5617                         if (!silent)
5618                                 fprintf(stderr,
5619                                         "Chunk[%llu, %u, %llu]: length(%llu), offset(%llu), type(%llu) mismatch with block group[%llu, %u, %llu]: offset(%llu), objectid(%llu), flags(%llu)\n",
5620                                         chunk_rec->objectid,
5621                                         chunk_rec->type,
5622                                         chunk_rec->offset,
5623                                         chunk_rec->length,
5624                                         chunk_rec->offset,
5625                                         chunk_rec->type_flags,
5626                                         block_group_rec->objectid,
5627                                         block_group_rec->type,
5628                                         block_group_rec->offset,
5629                                         block_group_rec->offset,
5630                                         block_group_rec->objectid,
5631                                         block_group_rec->flags);
5632                         ret = -1;
5633                 } else {
5634                         list_del_init(&block_group_rec->list);
5635                         chunk_rec->bg_rec = block_group_rec;
5636                 }
5637         } else {
5638                 if (!silent)
5639                         fprintf(stderr,
5640                                 "Chunk[%llu, %u, %llu]: length(%llu), offset(%llu), type(%llu) is not found in block group\n",
5641                                 chunk_rec->objectid,
5642                                 chunk_rec->type,
5643                                 chunk_rec->offset,
5644                                 chunk_rec->length,
5645                                 chunk_rec->offset,
5646                                 chunk_rec->type_flags);
5647                 ret = -1;
5648         }
5649
5650         length = calc_stripe_length(chunk_rec->type_flags, chunk_rec->length,
5651                                     chunk_rec->num_stripes);
5652         for (i = 0; i < chunk_rec->num_stripes; ++i) {
5653                 devid = chunk_rec->stripes[i].devid;
5654                 offset = chunk_rec->stripes[i].offset;
5655                 dev_extent_item = lookup_cache_extent2(&dev_extent_cache->tree,
5656                                                        devid, offset, length);
5657                 if (dev_extent_item) {
5658                         dev_extent_rec = container_of(dev_extent_item,
5659                                                 struct device_extent_record,
5660                                                 cache);
5661                         if (dev_extent_rec->objectid != devid ||
5662                             dev_extent_rec->offset != offset ||
5663                             dev_extent_rec->chunk_offset != chunk_rec->offset ||
5664                             dev_extent_rec->length != length) {
5665                                 if (!silent)
5666                                         fprintf(stderr,
5667                                                 "Chunk[%llu, %u, %llu] stripe[%llu, %llu] dismatch dev extent[%llu, %llu, %llu]\n",
5668                                                 chunk_rec->objectid,
5669                                                 chunk_rec->type,
5670                                                 chunk_rec->offset,
5671                                                 chunk_rec->stripes[i].devid,
5672                                                 chunk_rec->stripes[i].offset,
5673                                                 dev_extent_rec->objectid,
5674                                                 dev_extent_rec->offset,
5675                                                 dev_extent_rec->length);
5676                                 ret = -1;
5677                         } else {
5678                                 list_move(&dev_extent_rec->chunk_list,
5679                                           &chunk_rec->dextents);
5680                         }
5681                 } else {
5682                         if (!silent)
5683                                 fprintf(stderr,
5684                                         "Chunk[%llu, %u, %llu] stripe[%llu, %llu] is not found in dev extent\n",
5685                                         chunk_rec->objectid,
5686                                         chunk_rec->type,
5687                                         chunk_rec->offset,
5688                                         chunk_rec->stripes[i].devid,
5689                                         chunk_rec->stripes[i].offset);
5690                         ret = -1;
5691                 }
5692         }
5693         return ret;
5694 }
5695
5696 /* check btrfs_chunk -> btrfs_dev_extent / btrfs_block_group_item */
5697 int check_chunks(struct cache_tree *chunk_cache,
5698                  struct block_group_tree *block_group_cache,
5699                  struct device_extent_tree *dev_extent_cache,
5700                  struct list_head *good, struct list_head *bad, int silent)
5701 {
5702         struct cache_extent *chunk_item;
5703         struct chunk_record *chunk_rec;
5704         struct block_group_record *bg_rec;
5705         struct device_extent_record *dext_rec;
5706         int err;
5707         int ret = 0;
5708
5709         chunk_item = first_cache_extent(chunk_cache);
5710         while (chunk_item) {
5711                 chunk_rec = container_of(chunk_item, struct chunk_record,
5712                                          cache);
5713                 err = check_chunk_refs(chunk_rec, block_group_cache,
5714                                        dev_extent_cache, silent);
5715                 if (err) {
5716                         ret = err;
5717                         if (bad)
5718                                 list_add_tail(&chunk_rec->list, bad);
5719                 } else {
5720                         if (good)
5721                                 list_add_tail(&chunk_rec->list, good);
5722                 }
5723
5724                 chunk_item = next_cache_extent(chunk_item);
5725         }
5726
5727         list_for_each_entry(bg_rec, &block_group_cache->block_groups, list) {
5728                 if (!silent)
5729                         fprintf(stderr,
5730                                 "Block group[%llu, %llu] (flags = %llu) didn't find the relative chunk.\n",
5731                                 bg_rec->objectid,
5732                                 bg_rec->offset,
5733                                 bg_rec->flags);
5734                 if (!ret)
5735                         ret = 1;
5736         }
5737
5738         list_for_each_entry(dext_rec, &dev_extent_cache->no_chunk_orphans,
5739                             chunk_list) {
5740                 if (!silent)
5741                         fprintf(stderr,
5742                                 "Device extent[%llu, %llu, %llu] didn't find the relative chunk.\n",
5743                                 dext_rec->objectid,
5744                                 dext_rec->offset,
5745                                 dext_rec->length);
5746                 if (!ret)
5747                         ret = 1;
5748         }
5749         return ret;
5750 }
5751
5752
5753 static int check_device_used(struct device_record *dev_rec,
5754                              struct device_extent_tree *dext_cache)
5755 {
5756         struct cache_extent *cache;
5757         struct device_extent_record *dev_extent_rec;
5758         u64 total_byte = 0;
5759
5760         cache = search_cache_extent2(&dext_cache->tree, dev_rec->devid, 0);
5761         while (cache) {
5762                 dev_extent_rec = container_of(cache,
5763                                               struct device_extent_record,
5764                                               cache);
5765                 if (dev_extent_rec->objectid != dev_rec->devid)
5766                         break;
5767
5768                 list_del(&dev_extent_rec->device_list);
5769                 total_byte += dev_extent_rec->length;
5770                 cache = next_cache_extent(cache);
5771         }
5772
5773         if (total_byte != dev_rec->byte_used) {
5774                 fprintf(stderr,
5775                         "Dev extent's total-byte(%llu) is not equal to byte-used(%llu) in dev[%llu, %u, %llu]\n",
5776                         total_byte, dev_rec->byte_used, dev_rec->objectid,
5777                         dev_rec->type, dev_rec->offset);
5778                 return -1;
5779         } else {
5780                 return 0;
5781         }
5782 }
5783
5784 /* check btrfs_dev_item -> btrfs_dev_extent */
5785 static int check_devices(struct rb_root *dev_cache,
5786                          struct device_extent_tree *dev_extent_cache)
5787 {
5788         struct rb_node *dev_node;
5789         struct device_record *dev_rec;
5790         struct device_extent_record *dext_rec;
5791         int err;
5792         int ret = 0;
5793
5794         dev_node = rb_first(dev_cache);
5795         while (dev_node) {
5796                 dev_rec = container_of(dev_node, struct device_record, node);
5797                 err = check_device_used(dev_rec, dev_extent_cache);
5798                 if (err)
5799                         ret = err;
5800
5801                 dev_node = rb_next(dev_node);
5802         }
5803         list_for_each_entry(dext_rec, &dev_extent_cache->no_device_orphans,
5804                             device_list) {
5805                 fprintf(stderr,
5806                         "Device extent[%llu, %llu, %llu] didn't find its device.\n",
5807                         dext_rec->objectid, dext_rec->offset, dext_rec->length);
5808                 if (!ret)
5809                         ret = 1;
5810         }
5811         return ret;
5812 }
5813
5814 static int check_chunks_and_extents(struct btrfs_root *root)
5815 {
5816         struct rb_root dev_cache;
5817         struct cache_tree chunk_cache;
5818         struct block_group_tree block_group_cache;
5819         struct device_extent_tree dev_extent_cache;
5820         struct cache_tree extent_cache;
5821         struct cache_tree seen;
5822         struct cache_tree pending;
5823         struct cache_tree reada;
5824         struct cache_tree nodes;
5825         struct cache_tree corrupt_blocks;
5826         struct btrfs_path path;
5827         struct btrfs_key key;
5828         struct btrfs_key found_key;
5829         int ret, err = 0;
5830         u64 last = 0;
5831         struct block_info *bits;
5832         int bits_nr;
5833         struct extent_buffer *leaf;
5834         struct btrfs_trans_handle *trans = NULL;
5835         int slot;
5836         struct btrfs_root_item ri;
5837         struct list_head dropping_trees;
5838
5839         dev_cache = RB_ROOT;
5840         cache_tree_init(&chunk_cache);
5841         block_group_tree_init(&block_group_cache);
5842         device_extent_tree_init(&dev_extent_cache);
5843
5844         cache_tree_init(&extent_cache);
5845         cache_tree_init(&seen);
5846         cache_tree_init(&pending);
5847         cache_tree_init(&nodes);
5848         cache_tree_init(&reada);
5849         cache_tree_init(&corrupt_blocks);
5850         INIT_LIST_HEAD(&dropping_trees);
5851
5852         if (repair) {
5853                 trans = btrfs_start_transaction(root, 1);
5854                 if (IS_ERR(trans)) {
5855                         fprintf(stderr, "Error starting transaction\n");
5856                         return PTR_ERR(trans);
5857                 }
5858                 root->fs_info->fsck_extent_cache = &extent_cache;
5859                 root->fs_info->free_extent_hook = free_extent_hook;
5860                 root->fs_info->corrupt_blocks = &corrupt_blocks;
5861         }
5862
5863         bits_nr = 1024;
5864         bits = malloc(bits_nr * sizeof(struct block_info));
5865         if (!bits) {
5866                 perror("malloc");
5867                 exit(1);
5868         }
5869
5870 again:
5871         add_root_to_pending(root->fs_info->tree_root->node,
5872                             &extent_cache, &pending, &seen, &nodes,
5873                             &root->fs_info->tree_root->root_key);
5874
5875         add_root_to_pending(root->fs_info->chunk_root->node,
5876                             &extent_cache, &pending, &seen, &nodes,
5877                             &root->fs_info->chunk_root->root_key);
5878
5879         btrfs_init_path(&path);
5880         key.offset = 0;
5881         key.objectid = 0;
5882         btrfs_set_key_type(&key, BTRFS_ROOT_ITEM_KEY);
5883         ret = btrfs_search_slot(NULL, root->fs_info->tree_root,
5884                                         &key, &path, 0, 0);
5885         BUG_ON(ret < 0);
5886         while(1) {
5887                 leaf = path.nodes[0];
5888                 slot = path.slots[0];
5889                 if (slot >= btrfs_header_nritems(path.nodes[0])) {
5890                         ret = btrfs_next_leaf(root, &path);
5891                         if (ret != 0)
5892                                 break;
5893                         leaf = path.nodes[0];
5894                         slot = path.slots[0];
5895                 }
5896                 btrfs_item_key_to_cpu(leaf, &found_key, path.slots[0]);
5897                 if (btrfs_key_type(&found_key) == BTRFS_ROOT_ITEM_KEY) {
5898                         unsigned long offset;
5899                         struct extent_buffer *buf;
5900
5901                         offset = btrfs_item_ptr_offset(leaf, path.slots[0]);
5902                         read_extent_buffer(leaf, &ri, offset, sizeof(ri));
5903                         if (btrfs_disk_key_objectid(&ri.drop_progress) == 0) {
5904                                 buf = read_tree_block(root->fs_info->tree_root,
5905                                                       btrfs_root_bytenr(&ri),
5906                                                       btrfs_level_size(root,
5907                                                       btrfs_root_level(&ri)),
5908                                                       0);
5909                                 if (!buf) {
5910                                         ret = -EIO;
5911                                         goto out;
5912                                 }
5913                                 add_root_to_pending(buf, &extent_cache,
5914                                                     &pending, &seen, &nodes,
5915                                                     &found_key);
5916                                 free_extent_buffer(buf);
5917                         } else {
5918                                 struct dropping_root_item_record *dri_rec;
5919                                 dri_rec = malloc(sizeof(*dri_rec));
5920                                 if (!dri_rec) {
5921                                         perror("malloc");
5922                                         exit(1);
5923                                 }
5924                                 memcpy(&dri_rec->ri, &ri, sizeof(ri));
5925                                 memcpy(&dri_rec->found_key, &found_key,
5926                                        sizeof(found_key));
5927                                 list_add_tail(&dri_rec->list, &dropping_trees);
5928                         }
5929                 }
5930                 path.slots[0]++;
5931         }
5932         btrfs_release_path(&path);
5933         while (1) {
5934                 ret = run_next_block(trans, root, bits, bits_nr, &last,
5935                                      &pending, &seen, &reada, &nodes,
5936                                      &extent_cache, &chunk_cache, &dev_cache,
5937                                      &block_group_cache, &dev_extent_cache,
5938                                      NULL);
5939                 if (ret != 0)
5940                         break;
5941         }
5942
5943         while (!list_empty(&dropping_trees)) {
5944                 struct dropping_root_item_record *rec;
5945                 struct extent_buffer *buf;
5946                 rec = list_entry(dropping_trees.next,
5947                                  struct dropping_root_item_record, list);
5948                 last = 0;
5949                 if (!bits) {
5950                         perror("realloc");
5951                         exit(1);
5952                 }
5953                 buf = read_tree_block(root->fs_info->tree_root,
5954                                       btrfs_root_bytenr(&rec->ri),
5955                                       btrfs_level_size(root,
5956                                       btrfs_root_level(&rec->ri)), 0);
5957                 if (!buf) {
5958                         ret = -EIO;
5959                         goto out;
5960                 }
5961                 add_root_to_pending(buf, &extent_cache, &pending,
5962                                     &seen, &nodes, &rec->found_key);
5963                 while (1) {
5964                         ret = run_next_block(trans, root, bits, bits_nr, &last,
5965                                              &pending, &seen, &reada,
5966                                              &nodes, &extent_cache,
5967                                              &chunk_cache, &dev_cache,
5968                                              &block_group_cache,
5969                                              &dev_extent_cache,
5970                                              &rec->ri);
5971                         if (ret != 0)
5972                                 break;
5973                 }
5974                 free_extent_buffer(buf);
5975                 list_del(&rec->list);
5976                 free(rec);
5977         }
5978
5979         if (ret >= 0)
5980                 ret = check_extent_refs(trans, root, &extent_cache);
5981         if (ret == -EAGAIN) {
5982                 ret = btrfs_commit_transaction(trans, root);
5983                 if (ret)
5984                         goto out;
5985
5986                 trans = btrfs_start_transaction(root, 1);
5987                 if (IS_ERR(trans)) {
5988                         ret = PTR_ERR(trans);
5989                         goto out;
5990                 }
5991
5992                 free_corrupt_blocks_tree(root->fs_info->corrupt_blocks);
5993                 free_extent_cache_tree(&seen);
5994                 free_extent_cache_tree(&pending);
5995                 free_extent_cache_tree(&reada);
5996                 free_extent_cache_tree(&nodes);
5997                 free_extent_record_cache(root->fs_info, &extent_cache);
5998                 goto again;
5999         }
6000
6001         err = check_chunks(&chunk_cache, &block_group_cache,
6002                            &dev_extent_cache, NULL, NULL, 0);
6003         if (err && !ret)
6004                 ret = err;
6005
6006         err = check_devices(&dev_cache, &dev_extent_cache);
6007         if (err && !ret)
6008                 ret = err;
6009
6010         if (trans) {
6011                 err = btrfs_commit_transaction(trans, root);
6012                 if (!ret)
6013                         ret = err;
6014         }
6015 out:
6016         if (repair) {
6017                 free_corrupt_blocks_tree(root->fs_info->corrupt_blocks);
6018                 root->fs_info->fsck_extent_cache = NULL;
6019                 root->fs_info->free_extent_hook = NULL;
6020                 root->fs_info->corrupt_blocks = NULL;
6021         }
6022         free(bits);
6023         free_chunk_cache_tree(&chunk_cache);
6024         free_device_cache_tree(&dev_cache);
6025         free_block_group_tree(&block_group_cache);
6026         free_device_extent_tree(&dev_extent_cache);
6027         free_extent_cache_tree(&seen);
6028         free_extent_cache_tree(&pending);
6029         free_extent_cache_tree(&reada);
6030         free_extent_cache_tree(&nodes);
6031         return ret;
6032 }
6033
6034 static int btrfs_fsck_reinit_root(struct btrfs_trans_handle *trans,
6035                            struct btrfs_root *root, int overwrite)
6036 {
6037         struct extent_buffer *c;
6038         struct extent_buffer *old = root->node;
6039         int level;
6040         int ret;
6041         struct btrfs_disk_key disk_key = {0,0,0};
6042
6043         level = 0;
6044
6045         if (overwrite) {
6046                 c = old;
6047                 extent_buffer_get(c);
6048                 goto init;
6049         }
6050         c = btrfs_alloc_free_block(trans, root,
6051                                    btrfs_level_size(root, 0),
6052                                    root->root_key.objectid,
6053                                    &disk_key, level, 0, 0);
6054         if (IS_ERR(c)) {
6055                 c = old;
6056                 extent_buffer_get(c);
6057                 overwrite = 1;
6058         }
6059 init:
6060         memset_extent_buffer(c, 0, 0, sizeof(struct btrfs_header));
6061         btrfs_set_header_level(c, level);
6062         btrfs_set_header_bytenr(c, c->start);
6063         btrfs_set_header_generation(c, trans->transid);
6064         btrfs_set_header_backref_rev(c, BTRFS_MIXED_BACKREF_REV);
6065         btrfs_set_header_owner(c, root->root_key.objectid);
6066
6067         write_extent_buffer(c, root->fs_info->fsid,
6068                             btrfs_header_fsid(), BTRFS_FSID_SIZE);
6069
6070         write_extent_buffer(c, root->fs_info->chunk_tree_uuid,
6071                             btrfs_header_chunk_tree_uuid(c),
6072                             BTRFS_UUID_SIZE);
6073
6074         btrfs_mark_buffer_dirty(c);
6075         /*
6076          * this case can happen in the following case:
6077          *
6078          * 1.overwrite previous root.
6079          *
6080          * 2.reinit reloc data root, this is because we skip pin
6081          * down reloc data tree before which means we can allocate
6082          * same block bytenr here.
6083          */
6084         if (old->start == c->start) {
6085                 btrfs_set_root_generation(&root->root_item,
6086                                           trans->transid);
6087                 root->root_item.level = btrfs_header_level(root->node);
6088                 ret = btrfs_update_root(trans, root->fs_info->tree_root,
6089                                         &root->root_key, &root->root_item);
6090                 if (ret) {
6091                         free_extent_buffer(c);
6092                         return ret;
6093                 }
6094         }
6095         free_extent_buffer(old);
6096         root->node = c;
6097         add_root_to_dirty_list(root);
6098         return 0;
6099 }
6100
6101 static int pin_down_tree_blocks(struct btrfs_fs_info *fs_info,
6102                                 struct extent_buffer *eb, int tree_root)
6103 {
6104         struct extent_buffer *tmp;
6105         struct btrfs_root_item *ri;
6106         struct btrfs_key key;
6107         u64 bytenr;
6108         u32 leafsize;
6109         int level = btrfs_header_level(eb);
6110         int nritems;
6111         int ret;
6112         int i;
6113
6114         btrfs_pin_extent(fs_info, eb->start, eb->len);
6115
6116         leafsize = btrfs_super_leafsize(fs_info->super_copy);
6117         nritems = btrfs_header_nritems(eb);
6118         for (i = 0; i < nritems; i++) {
6119                 if (level == 0) {
6120                         btrfs_item_key_to_cpu(eb, &key, i);
6121                         if (key.type != BTRFS_ROOT_ITEM_KEY)
6122                                 continue;
6123                         /* Skip the extent root and reloc roots */
6124                         if (key.objectid == BTRFS_EXTENT_TREE_OBJECTID ||
6125                             key.objectid == BTRFS_TREE_RELOC_OBJECTID ||
6126                             key.objectid == BTRFS_DATA_RELOC_TREE_OBJECTID)
6127                                 continue;
6128                         ri = btrfs_item_ptr(eb, i, struct btrfs_root_item);
6129                         bytenr = btrfs_disk_root_bytenr(eb, ri);
6130
6131                         /*
6132                          * If at any point we start needing the real root we
6133                          * will have to build a stump root for the root we are
6134                          * in, but for now this doesn't actually use the root so
6135                          * just pass in extent_root.
6136                          */
6137                         tmp = read_tree_block(fs_info->extent_root, bytenr,
6138                                               leafsize, 0);
6139                         if (!tmp) {
6140                                 fprintf(stderr, "Error reading root block\n");
6141                                 return -EIO;
6142                         }
6143                         ret = pin_down_tree_blocks(fs_info, tmp, 0);
6144                         free_extent_buffer(tmp);
6145                         if (ret)
6146                                 return ret;
6147                 } else {
6148                         bytenr = btrfs_node_blockptr(eb, i);
6149
6150                         /* If we aren't the tree root don't read the block */
6151                         if (level == 1 && !tree_root) {
6152                                 btrfs_pin_extent(fs_info, bytenr, leafsize);
6153                                 continue;
6154                         }
6155
6156                         tmp = read_tree_block(fs_info->extent_root, bytenr,
6157                                               leafsize, 0);
6158                         if (!tmp) {
6159                                 fprintf(stderr, "Error reading tree block\n");
6160                                 return -EIO;
6161                         }
6162                         ret = pin_down_tree_blocks(fs_info, tmp, tree_root);
6163                         free_extent_buffer(tmp);
6164                         if (ret)
6165                                 return ret;
6166                 }
6167         }
6168
6169         return 0;
6170 }
6171
6172 static int pin_metadata_blocks(struct btrfs_fs_info *fs_info)
6173 {
6174         int ret;
6175
6176         ret = pin_down_tree_blocks(fs_info, fs_info->chunk_root->node, 0);
6177         if (ret)
6178                 return ret;
6179
6180         return pin_down_tree_blocks(fs_info, fs_info->tree_root->node, 1);
6181 }
6182
6183 static int reset_block_groups(struct btrfs_fs_info *fs_info)
6184 {
6185         struct btrfs_block_group_cache *cache;
6186         struct btrfs_path *path;
6187         struct extent_buffer *leaf;
6188         struct btrfs_chunk *chunk;
6189         struct btrfs_key key;
6190         int ret;
6191         u64 start;
6192
6193         path = btrfs_alloc_path();
6194         if (!path)
6195                 return -ENOMEM;
6196
6197         key.objectid = 0;
6198         key.type = BTRFS_CHUNK_ITEM_KEY;
6199         key.offset = 0;
6200
6201         ret = btrfs_search_slot(NULL, fs_info->chunk_root, &key, path, 0, 0);
6202         if (ret < 0) {
6203                 btrfs_free_path(path);
6204                 return ret;
6205         }
6206
6207         /*
6208          * We do this in case the block groups were screwed up and had alloc
6209          * bits that aren't actually set on the chunks.  This happens with
6210          * restored images every time and could happen in real life I guess.
6211          */
6212         fs_info->avail_data_alloc_bits = 0;
6213         fs_info->avail_metadata_alloc_bits = 0;
6214         fs_info->avail_system_alloc_bits = 0;
6215
6216         /* First we need to create the in-memory block groups */
6217         while (1) {
6218                 if (path->slots[0] >= btrfs_header_nritems(path->nodes[0])) {
6219                         ret = btrfs_next_leaf(fs_info->chunk_root, path);
6220                         if (ret < 0) {
6221                                 btrfs_free_path(path);
6222                                 return ret;
6223                         }
6224                         if (ret) {
6225                                 ret = 0;
6226                                 break;
6227                         }
6228                 }
6229                 leaf = path->nodes[0];
6230                 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
6231                 if (key.type != BTRFS_CHUNK_ITEM_KEY) {
6232                         path->slots[0]++;
6233                         continue;
6234                 }
6235
6236                 chunk = btrfs_item_ptr(leaf, path->slots[0],
6237                                        struct btrfs_chunk);
6238                 btrfs_add_block_group(fs_info, 0,
6239                                       btrfs_chunk_type(leaf, chunk),
6240                                       key.objectid, key.offset,
6241                                       btrfs_chunk_length(leaf, chunk));
6242                 set_extent_dirty(&fs_info->free_space_cache, key.offset,
6243                                  key.offset + btrfs_chunk_length(leaf, chunk),
6244                                  GFP_NOFS);
6245                 path->slots[0]++;
6246         }
6247         start = 0;
6248         while (1) {
6249                 cache = btrfs_lookup_first_block_group(fs_info, start);
6250                 if (!cache)
6251                         break;
6252                 cache->cached = 1;
6253                 start = cache->key.objectid + cache->key.offset;
6254         }
6255
6256         btrfs_free_path(path);
6257         return 0;
6258 }
6259
6260 static int reset_balance(struct btrfs_trans_handle *trans,
6261                          struct btrfs_fs_info *fs_info)
6262 {
6263         struct btrfs_root *root = fs_info->tree_root;
6264         struct btrfs_path *path;
6265         struct extent_buffer *leaf;
6266         struct btrfs_key key;
6267         int del_slot, del_nr = 0;
6268         int ret;
6269         int found = 0;
6270
6271         path = btrfs_alloc_path();
6272         if (!path)
6273                 return -ENOMEM;
6274
6275         key.objectid = BTRFS_BALANCE_OBJECTID;
6276         key.type = BTRFS_BALANCE_ITEM_KEY;
6277         key.offset = 0;
6278
6279         ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
6280         if (ret) {
6281                 if (ret > 0)
6282                         ret = 0;
6283                 if (!ret)
6284                         goto reinit_data_reloc;
6285                 else
6286                         goto out;
6287         }
6288
6289         ret = btrfs_del_item(trans, root, path);
6290         if (ret)
6291                 goto out;
6292         btrfs_release_path(path);
6293
6294         key.objectid = BTRFS_TREE_RELOC_OBJECTID;
6295         key.type = BTRFS_ROOT_ITEM_KEY;
6296         key.offset = 0;
6297
6298         ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
6299         if (ret < 0)
6300                 goto out;
6301         while (1) {
6302                 if (path->slots[0] >= btrfs_header_nritems(path->nodes[0])) {
6303                         if (!found)
6304                                 break;
6305
6306                         if (del_nr) {
6307                                 ret = btrfs_del_items(trans, root, path,
6308                                                       del_slot, del_nr);
6309                                 del_nr = 0;
6310                                 if (ret)
6311                                         goto out;
6312                         }
6313                         key.offset++;
6314                         btrfs_release_path(path);
6315
6316                         found = 0;
6317                         ret = btrfs_search_slot(trans, root, &key, path,
6318                                                 -1, 1);
6319                         if (ret < 0)
6320                                 goto out;
6321                         continue;
6322                 }
6323                 found = 1;
6324                 leaf = path->nodes[0];
6325                 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
6326                 if (key.objectid > BTRFS_TREE_RELOC_OBJECTID)
6327                         break;
6328                 if (key.objectid != BTRFS_TREE_RELOC_OBJECTID) {
6329                         path->slots[0]++;
6330                         continue;
6331                 }
6332                 if (!del_nr) {
6333                         del_slot = path->slots[0];
6334                         del_nr = 1;
6335                 } else {
6336                         del_nr++;
6337                 }
6338                 path->slots[0]++;
6339         }
6340
6341         if (del_nr) {
6342                 ret = btrfs_del_items(trans, root, path, del_slot, del_nr);
6343                 if (ret)
6344                         goto out;
6345         }
6346         btrfs_release_path(path);
6347
6348 reinit_data_reloc:
6349         key.objectid = BTRFS_DATA_RELOC_TREE_OBJECTID;
6350         key.type = BTRFS_ROOT_ITEM_KEY;
6351         key.offset = (u64)-1;
6352         root = btrfs_read_fs_root(fs_info, &key);
6353         if (IS_ERR(root)) {
6354                 fprintf(stderr, "Error reading data reloc tree\n");
6355                 return PTR_ERR(root);
6356         }
6357         root->track_dirty = 1;
6358         if (root->last_trans != trans->transid) {
6359                 root->last_trans = trans->transid;
6360                 root->commit_root = root->node;
6361                 extent_buffer_get(root->node);
6362         }
6363         ret = btrfs_fsck_reinit_root(trans, root, 0);
6364         if (ret)
6365                 goto out;
6366         ret = btrfs_make_root_dir(trans, root, BTRFS_FIRST_FREE_OBJECTID);
6367 out:
6368         btrfs_free_path(path);
6369         return ret;
6370 }
6371
6372 static int reinit_extent_tree(struct btrfs_trans_handle *trans,
6373                               struct btrfs_fs_info *fs_info)
6374 {
6375         u64 start = 0;
6376         int ret;
6377
6378         /*
6379          * The only reason we don't do this is because right now we're just
6380          * walking the trees we find and pinning down their bytes, we don't look
6381          * at any of the leaves.  In order to do mixed groups we'd have to check
6382          * the leaves of any fs roots and pin down the bytes for any file
6383          * extents we find.  Not hard but why do it if we don't have to?
6384          */
6385         if (btrfs_fs_incompat(fs_info, BTRFS_FEATURE_INCOMPAT_MIXED_GROUPS)) {
6386                 fprintf(stderr, "We don't support re-initing the extent tree "
6387                         "for mixed block groups yet, please notify a btrfs "
6388                         "developer you want to do this so they can add this "
6389                         "functionality.\n");
6390                 return -EINVAL;
6391         }
6392
6393         /*
6394          * first we need to walk all of the trees except the extent tree and pin
6395          * down the bytes that are in use so we don't overwrite any existing
6396          * metadata.
6397          */
6398         ret = pin_metadata_blocks(fs_info);
6399         if (ret) {
6400                 fprintf(stderr, "error pinning down used bytes\n");
6401                 return ret;
6402         }
6403
6404         /*
6405          * Need to drop all the block groups since we're going to recreate all
6406          * of them again.
6407          */
6408         btrfs_free_block_groups(fs_info);
6409         ret = reset_block_groups(fs_info);
6410         if (ret) {
6411                 fprintf(stderr, "error resetting the block groups\n");
6412                 return ret;
6413         }
6414
6415         /* Ok we can allocate now, reinit the extent root */
6416         ret = btrfs_fsck_reinit_root(trans, fs_info->extent_root, 0);
6417         if (ret) {
6418                 fprintf(stderr, "extent root initialization failed\n");
6419                 /*
6420                  * When the transaction code is updated we should end the
6421                  * transaction, but for now progs only knows about commit so
6422                  * just return an error.
6423                  */
6424                 return ret;
6425         }
6426
6427         /*
6428          * Now we have all the in-memory block groups setup so we can make
6429          * allocations properly, and the metadata we care about is safe since we
6430          * pinned all of it above.
6431          */
6432         while (1) {
6433                 struct btrfs_block_group_cache *cache;
6434
6435                 cache = btrfs_lookup_first_block_group(fs_info, start);
6436                 if (!cache)
6437                         break;
6438                 start = cache->key.objectid + cache->key.offset;
6439                 ret = btrfs_insert_item(trans, fs_info->extent_root,
6440                                         &cache->key, &cache->item,
6441                                         sizeof(cache->item));
6442                 if (ret) {
6443                         fprintf(stderr, "Error adding block group\n");
6444                         return ret;
6445                 }
6446                 btrfs_extent_post_op(trans, fs_info->extent_root);
6447         }
6448
6449         ret = reset_balance(trans, fs_info);
6450         if (ret)
6451                 fprintf(stderr, "error reseting the pending balance\n");
6452
6453         return ret;
6454 }
6455
6456 static int recow_extent_buffer(struct btrfs_root *root, struct extent_buffer *eb)
6457 {
6458         struct btrfs_path *path;
6459         struct btrfs_trans_handle *trans;
6460         struct btrfs_key key;
6461         int ret;
6462
6463         printf("Recowing metadata block %llu\n", eb->start);
6464         key.objectid = btrfs_header_owner(eb);
6465         key.type = BTRFS_ROOT_ITEM_KEY;
6466         key.offset = (u64)-1;
6467
6468         root = btrfs_read_fs_root(root->fs_info, &key);
6469         if (IS_ERR(root)) {
6470                 fprintf(stderr, "Couldn't find owner root %llu\n",
6471                         key.objectid);
6472                 return PTR_ERR(root);
6473         }
6474
6475         path = btrfs_alloc_path();
6476         if (!path)
6477                 return -ENOMEM;
6478
6479         trans = btrfs_start_transaction(root, 1);
6480         if (IS_ERR(trans)) {
6481                 btrfs_free_path(path);
6482                 return PTR_ERR(trans);
6483         }
6484
6485         path->lowest_level = btrfs_header_level(eb);
6486         if (path->lowest_level)
6487                 btrfs_node_key_to_cpu(eb, &key, 0);
6488         else
6489                 btrfs_item_key_to_cpu(eb, &key, 0);
6490
6491         ret = btrfs_search_slot(trans, root, &key, path, 0, 1);
6492         btrfs_commit_transaction(trans, root);
6493         btrfs_free_path(path);
6494         return ret;
6495 }
6496
6497 static int delete_bad_item(struct btrfs_root *root, struct bad_item *bad)
6498 {
6499         struct btrfs_path *path;
6500         struct btrfs_trans_handle *trans;
6501         struct btrfs_key key;
6502         int ret;
6503
6504         printf("Deleting bad item [%llu,%u,%llu]\n", bad->key.objectid,
6505                bad->key.type, bad->key.offset);
6506         key.objectid = bad->root_id;
6507         key.type = BTRFS_ROOT_ITEM_KEY;
6508         key.offset = (u64)-1;
6509
6510         root = btrfs_read_fs_root(root->fs_info, &key);
6511         if (IS_ERR(root)) {
6512                 fprintf(stderr, "Couldn't find owner root %llu\n",
6513                         key.objectid);
6514                 return PTR_ERR(root);
6515         }
6516
6517         path = btrfs_alloc_path();
6518         if (!path)
6519                 return -ENOMEM;
6520
6521         trans = btrfs_start_transaction(root, 1);
6522         if (IS_ERR(trans)) {
6523                 btrfs_free_path(path);
6524                 return PTR_ERR(trans);
6525         }
6526
6527         ret = btrfs_search_slot(trans, root, &bad->key, path, -1, 1);
6528         if (ret) {
6529                 if (ret > 0)
6530                         ret = 0;
6531                 goto out;
6532         }
6533         ret = btrfs_del_item(trans, root, path);
6534 out:
6535         btrfs_commit_transaction(trans, root);
6536         btrfs_free_path(path);
6537         return ret;
6538 }
6539
6540 static struct option long_options[] = {
6541         { "super", 1, NULL, 's' },
6542         { "repair", 0, NULL, 0 },
6543         { "init-csum-tree", 0, NULL, 0 },
6544         { "init-extent-tree", 0, NULL, 0 },
6545         { "check-data-csum", 0, NULL, 0 },
6546         { "backup", 0, NULL, 0 },
6547         { "qgroup-report", 0, NULL, 'Q' },
6548         { NULL, 0, NULL, 0}
6549 };
6550
6551 const char * const cmd_check_usage[] = {
6552         "btrfs check [options] <device>",
6553         "Check an unmounted btrfs filesystem.",
6554         "",
6555         "-s|--super <superblock>     use this superblock copy",
6556         "-b|--backup                 use the backup root copy",
6557         "--repair                    try to repair the filesystem",
6558         "--init-csum-tree            create a new CRC tree",
6559         "--init-extent-tree          create a new extent tree",
6560         "--check-data-csum           verify checkums of data blocks",
6561         "--qgroup-report             print a report on qgroup consistency",
6562         NULL
6563 };
6564
6565 int cmd_check(int argc, char **argv)
6566 {
6567         struct cache_tree root_cache;
6568         struct btrfs_root *root;
6569         struct btrfs_fs_info *info;
6570         u64 bytenr = 0;
6571         char uuidbuf[BTRFS_UUID_UNPARSED_SIZE];
6572         int ret;
6573         u64 num;
6574         int option_index = 0;
6575         int init_csum_tree = 0;
6576         int qgroup_report = 0;
6577         enum btrfs_open_ctree_flags ctree_flags =
6578                 OPEN_CTREE_PARTIAL | OPEN_CTREE_EXCLUSIVE;
6579
6580         while(1) {
6581                 int c;
6582                 c = getopt_long(argc, argv, "as:b", long_options,
6583                                 &option_index);
6584                 if (c < 0)
6585                         break;
6586                 switch(c) {
6587                         case 'a': /* ignored */ break;
6588                         case 'b':
6589                                 ctree_flags |= OPEN_CTREE_BACKUP_ROOT;
6590                                 break;
6591                         case 's':
6592                                 num = arg_strtou64(optarg);
6593                                 if (num >= BTRFS_SUPER_MIRROR_MAX) {
6594                                         fprintf(stderr,
6595                                                 "ERROR: super mirror should be less than: %d\n",
6596                                                 BTRFS_SUPER_MIRROR_MAX);
6597                                         exit(1);
6598                                 }
6599                                 bytenr = btrfs_sb_offset(((int)num));
6600                                 printf("using SB copy %llu, bytenr %llu\n", num,
6601                                        (unsigned long long)bytenr);
6602                                 break;
6603                         case 'Q':
6604                                 qgroup_report = 1;
6605                                 break;
6606                         case '?':
6607                         case 'h':
6608                                 usage(cmd_check_usage);
6609                 }
6610                 if (option_index == 1) {
6611                         printf("enabling repair mode\n");
6612                         repair = 1;
6613                         ctree_flags |= OPEN_CTREE_WRITES;
6614                 } else if (option_index == 2) {
6615                         printf("Creating a new CRC tree\n");
6616                         init_csum_tree = 1;
6617                         repair = 1;
6618                         ctree_flags |= OPEN_CTREE_WRITES;
6619                 } else if (option_index == 3) {
6620                         init_extent_tree = 1;
6621                         ctree_flags |= (OPEN_CTREE_WRITES |
6622                                         OPEN_CTREE_NO_BLOCK_GROUPS);
6623                         repair = 1;
6624                 } else if (option_index == 4) {
6625                         check_data_csum = 1;
6626                 }
6627         }
6628         argc = argc - optind;
6629
6630         if (argc != 1)
6631                 usage(cmd_check_usage);
6632
6633         radix_tree_init();
6634         cache_tree_init(&root_cache);
6635
6636         if((ret = check_mounted(argv[optind])) < 0) {
6637                 fprintf(stderr, "Could not check mount status: %s\n", strerror(-ret));
6638                 goto err_out;
6639         } else if(ret) {
6640                 fprintf(stderr, "%s is currently mounted. Aborting.\n", argv[optind]);
6641                 ret = -EBUSY;
6642                 goto err_out;
6643         }
6644
6645         info = open_ctree_fs_info(argv[optind], bytenr, 0, ctree_flags);
6646         if (!info) {
6647                 fprintf(stderr, "Couldn't open file system\n");
6648                 ret = -EIO;
6649                 goto err_out;
6650         }
6651
6652         root = info->fs_root;
6653         uuid_unparse(info->super_copy->fsid, uuidbuf);
6654         if (qgroup_report) {
6655                 printf("Print quota groups for %s\nUUID: %s\n", argv[optind],
6656                        uuidbuf);
6657                 ret = qgroup_verify_all(info);
6658                 if (ret == 0)
6659                         print_qgroup_report(1);
6660                 goto close_out;
6661         }
6662         printf("Checking filesystem on %s\nUUID: %s\n", argv[optind], uuidbuf);
6663
6664         if (!extent_buffer_uptodate(info->tree_root->node) ||
6665             !extent_buffer_uptodate(info->dev_root->node) ||
6666             !extent_buffer_uptodate(info->chunk_root->node)) {
6667                 fprintf(stderr, "Critical roots corrupted, unable to fsck the FS\n");
6668                 ret = -EIO;
6669                 goto close_out;
6670         }
6671
6672         if (init_extent_tree || init_csum_tree) {
6673                 struct btrfs_trans_handle *trans;
6674
6675                 trans = btrfs_start_transaction(info->extent_root, 0);
6676                 if (IS_ERR(trans)) {
6677                         fprintf(stderr, "Error starting transaction\n");
6678                         ret = PTR_ERR(trans);
6679                         goto close_out;
6680                 }
6681
6682                 if (init_extent_tree) {
6683                         printf("Creating a new extent tree\n");
6684                         ret = reinit_extent_tree(trans, info);
6685                         if (ret)
6686                                 goto close_out;
6687                 }
6688
6689                 if (init_csum_tree) {
6690                         fprintf(stderr, "Reinit crc root\n");
6691                         ret = btrfs_fsck_reinit_root(trans, info->csum_root, 0);
6692                         if (ret) {
6693                                 fprintf(stderr, "crc root initialization failed\n");
6694                                 ret = -EIO;
6695                                 goto close_out;
6696                         }
6697                 }
6698                 /*
6699                  * Ok now we commit and run the normal fsck, which will add
6700                  * extent entries for all of the items it finds.
6701                  */
6702                 ret = btrfs_commit_transaction(trans, info->extent_root);
6703                 if (ret)
6704                         goto close_out;
6705         }
6706         if (!extent_buffer_uptodate(info->extent_root->node)) {
6707                 fprintf(stderr, "Critical roots corrupted, unable to fsck the FS\n");
6708                 ret = -EIO;
6709                 goto close_out;
6710         }
6711
6712         fprintf(stderr, "checking extents\n");
6713         ret = check_chunks_and_extents(root);
6714         if (ret)
6715                 fprintf(stderr, "Errors found in extent allocation tree or chunk allocation\n");
6716
6717         fprintf(stderr, "checking free space cache\n");
6718         ret = check_space_cache(root);
6719         if (ret)
6720                 goto out;
6721
6722         /*
6723          * We used to have to have these hole extents in between our real
6724          * extents so if we don't have this flag set we need to make sure there
6725          * are no gaps in the file extents for inodes, otherwise we can just
6726          * ignore it when this happens.
6727          */
6728         no_holes = btrfs_fs_incompat(root->fs_info,
6729                                      BTRFS_FEATURE_INCOMPAT_NO_HOLES);
6730         fprintf(stderr, "checking fs roots\n");
6731         ret = check_fs_roots(root, &root_cache);
6732         if (ret)
6733                 goto out;
6734
6735         fprintf(stderr, "checking csums\n");
6736         ret = check_csums(root);
6737         if (ret)
6738                 goto out;
6739
6740         fprintf(stderr, "checking root refs\n");
6741         ret = check_root_refs(root, &root_cache);
6742         if (ret)
6743                 goto out;
6744
6745         while (repair && !list_empty(&root->fs_info->recow_ebs)) {
6746                 struct extent_buffer *eb;
6747
6748                 eb = list_first_entry(&root->fs_info->recow_ebs,
6749                                       struct extent_buffer, recow);
6750                 ret = recow_extent_buffer(root, eb);
6751                 if (ret)
6752                         break;
6753         }
6754
6755         while (!list_empty(&delete_items)) {
6756                 struct bad_item *bad;
6757
6758                 bad = list_first_entry(&delete_items, struct bad_item, list);
6759                 list_del_init(&bad->list);
6760                 if (repair)
6761                         ret = delete_bad_item(root, bad);
6762                 free(bad);
6763         }
6764
6765         if (info->quota_enabled) {
6766                 int err;
6767                 fprintf(stderr, "checking quota groups\n");
6768                 err = qgroup_verify_all(info);
6769                 if (err)
6770                         goto out;
6771         }
6772
6773         if (!list_empty(&root->fs_info->recow_ebs)) {
6774                 fprintf(stderr, "Transid errors in file system\n");
6775                 ret = 1;
6776         }
6777 out:
6778         print_qgroup_report(0);
6779         if (found_old_backref) { /*
6780                  * there was a disk format change when mixed
6781                  * backref was in testing tree. The old format
6782                  * existed about one week.
6783                  */
6784                 printf("\n * Found old mixed backref format. "
6785                        "The old format is not supported! *"
6786                        "\n * Please mount the FS in readonly mode, "
6787                        "backup data and re-format the FS. *\n\n");
6788                 ret = 1;
6789         }
6790         printf("found %llu bytes used err is %d\n",
6791                (unsigned long long)bytes_used, ret);
6792         printf("total csum bytes: %llu\n",(unsigned long long)total_csum_bytes);
6793         printf("total tree bytes: %llu\n",
6794                (unsigned long long)total_btree_bytes);
6795         printf("total fs tree bytes: %llu\n",
6796                (unsigned long long)total_fs_tree_bytes);
6797         printf("total extent tree bytes: %llu\n",
6798                (unsigned long long)total_extent_tree_bytes);
6799         printf("btree space waste bytes: %llu\n",
6800                (unsigned long long)btree_space_waste);
6801         printf("file data blocks allocated: %llu\n referenced %llu\n",
6802                 (unsigned long long)data_bytes_allocated,
6803                 (unsigned long long)data_bytes_referenced);
6804         printf("%s\n", BTRFS_BUILD_VERSION);
6805
6806         free_root_recs_tree(&root_cache);
6807 close_out:
6808         close_ctree(root);
6809 err_out:
6810         return ret;
6811 }