git.oblomov.eu Git - linux-2.6/blob - fs/btrfs/extent-tree.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 #include <linux/sched.h>
  20 #include <linux/crc32c.h>
  21 #include <linux/pagemap.h>
  22 #include "hash.h"
  23 #include "ctree.h"
  24 #include "disk-io.h"
  25 #include "print-tree.h"
  26 #include "transaction.h"
  27
  28 #define BLOCK_GROUP_DATA EXTENT_WRITEBACK
  29 #define BLOCK_GROUP_METADATA EXTENT_UPTODATE
  30 #define BLOCK_GROUP_DIRTY EXTENT_DIRTY
  31
  32 static int finish_current_insert(struct btrfs_trans_handle *trans, struct
  33                                  btrfs_root *extent_root);
  34 static int del_pending_extents(struct btrfs_trans_handle *trans, struct
  35                                btrfs_root *extent_root);
  36 static int find_previous_extent(struct btrfs_root *root,
  37                                 struct btrfs_path *path)
  38 {
  39         struct btrfs_key found_key;
  40         struct extent_buffer *leaf;
  41         int ret;
  42
  43         while(1) {
  44                 if (path->slots[0] == 0) {
  45                         ret = btrfs_prev_leaf(root, path);
  46                         if (ret != 0)
  47                                 return ret;
  48                 } else {
  49                         path->slots[0]--;
  50                 }
  51                 leaf = path->nodes[0];
  52                 btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
  53                 if (found_key.type == BTRFS_EXTENT_ITEM_KEY)
  54                         return 0;
  55         }
  56         return 1;
  57 }
  58
  59 static int cache_block_group(struct btrfs_root *root,
  60                              struct btrfs_block_group_cache *block_group)
  61 {
  62         struct btrfs_path *path;
  63         int ret;
  64         struct btrfs_key key;
  65         struct extent_buffer *leaf;
  66         struct extent_io_tree *free_space_cache;
  67         int slot;
  68         u64 last = 0;
  69         u64 hole_size;
  70         u64 first_free;
  71         int found = 0;
  72
  73         if (!block_group)
  74                 return 0;
  75
  76         root = root->fs_info->extent_root;
  77         free_space_cache = &root->fs_info->free_space_cache;
  78
  79         if (block_group->cached)
  80                 return 0;
  81
  82         path = btrfs_alloc_path();
  83         if (!path)
  84                 return -ENOMEM;
  85
  86         path->reada = 2;
  87         first_free = block_group->key.objectid;
  88         key.objectid = block_group->key.objectid;
  89         key.offset = 0;
  90         btrfs_set_key_type(&key, BTRFS_EXTENT_ITEM_KEY);
  91         ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
  92         if (ret < 0)
  93                 return ret;
  94         ret = find_previous_extent(root, path);
  95         if (ret < 0)
  96                 return ret;
  97         if (ret == 0) {
  98                 leaf = path->nodes[0];
  99                 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
 100                 if (key.objectid + key.offset > first_free)
 101                         first_free = key.objectid + key.offset;
 102         }
 103         while(1) {
 104                 leaf = path->nodes[0];
 105                 slot = path->slots[0];
 106                 if (slot >= btrfs_header_nritems(leaf)) {
 107                         ret = btrfs_next_leaf(root, path);
 108                         if (ret < 0)
 109                                 goto err;
 110                         if (ret == 0) {
 111                                 continue;
 112                         } else {
 113                                 break;
 114                         }
 115                 }
 116                 btrfs_item_key_to_cpu(leaf, &key, slot);
 117                 if (key.objectid < block_group->key.objectid) {
 118                         goto next;
 119                 }
 120                 if (key.objectid >= block_group->key.objectid +
 121                     block_group->key.offset) {
 122                         break;
 123                 }
 124
 125                 if (btrfs_key_type(&key) == BTRFS_EXTENT_ITEM_KEY) {
 126                         if (!found) {
 127                                 last = first_free;
 128                                 found = 1;
 129                         }
 130                         if (key.objectid > last) {
 131                                 hole_size = key.objectid - last;
 132                                 set_extent_dirty(free_space_cache, last,
 133                                                  last + hole_size - 1,
 134                                                  GFP_NOFS);
 135                         }
 136                         last = key.objectid + key.offset;
 137                 }
 138 next:
 139                 path->slots[0]++;
 140         }
 141
 142         if (!found)
 143                 last = first_free;
 144         if (block_group->key.objectid +
 145             block_group->key.offset > last) {
 146                 hole_size = block_group->key.objectid +
 147                         block_group->key.offset - last;
 148                 set_extent_dirty(free_space_cache, last,
 149                                  last + hole_size - 1, GFP_NOFS);
 150         }
 151         block_group->cached = 1;
 152 err:
 153         btrfs_free_path(path);
 154         return 0;
 155 }
 156
 157 struct btrfs_block_group_cache *btrfs_lookup_block_group(struct
 158                                                          btrfs_fs_info *info,
 159                                                          u64 bytenr)
 160 {
 161         struct extent_io_tree *block_group_cache;
 162         struct btrfs_block_group_cache *block_group = NULL;
 163         u64 ptr;
 164         u64 start;
 165         u64 end;
 166         int ret;
 167
 168         block_group_cache = &info->block_group_cache;
 169         ret = find_first_extent_bit(block_group_cache,
 170                                     bytenr, &start, &end,
 171                                     BLOCK_GROUP_DATA | BLOCK_GROUP_METADATA);
 172         if (ret) {
 173                 return NULL;
 174         }
 175         ret = get_state_private(block_group_cache, start, &ptr);
 176         if (ret)
 177                 return NULL;
 178
 179         block_group = (struct btrfs_block_group_cache *)(unsigned long)ptr;
 180         if (block_group->key.objectid <= bytenr && bytenr <
 181             block_group->key.objectid + block_group->key.offset)
 182                 return block_group;
 183         return NULL;
 184 }
 185 static u64 noinline find_search_start(struct btrfs_root *root,
 186                               struct btrfs_block_group_cache **cache_ret,
 187                               u64 search_start, int num, int data)
 188 {
 189         int ret;
 190         struct btrfs_block_group_cache *cache = *cache_ret;
 191         u64 last;
 192         u64 start = 0;
 193         u64 end = 0;
 194         u64 cache_miss = 0;
 195         u64 total_fs_bytes;
 196         int wrapped = 0;
 197
 198         if (!cache) {
 199                 goto out;
 200         }
 201         total_fs_bytes = btrfs_super_total_bytes(&root->fs_info->super_copy);
 202 again:
 203         ret = cache_block_group(root, cache);
 204         if (ret)
 205                 goto out;
 206
 207         last = max(search_start, cache->key.objectid);
 208
 209         while(1) {
 210                 ret = find_first_extent_bit(&root->fs_info->free_space_cache,
 211                                             last, &start, &end, EXTENT_DIRTY);
 212                 if (ret) {
 213                         if (!cache_miss)
 214                                 cache_miss = last;
 215                         goto new_group;
 216                 }
 217
 218                 start = max(last, start);
 219                 last = end + 1;
 220                 if (last - start < num) {
 221                         if (last == cache->key.objectid + cache->key.offset)
 222                                 cache_miss = start;
 223                         continue;
 224                 }
 225                 if (data != BTRFS_BLOCK_GROUP_MIXED &&
 226                     start + num > cache->key.objectid + cache->key.offset)
 227                         goto new_group;
 228                 if (start + num  > total_fs_bytes)
 229                         goto new_group;
 230                 return start;
 231         }
 232 out:
 233         cache = btrfs_lookup_block_group(root->fs_info, search_start);
 234         if (!cache) {
 235                 printk("Unable to find block group for %Lu\n",
 236                        search_start);
 237                 WARN_ON(1);
 238                 return search_start;
 239         }
 240         return search_start;
 241
 242 new_group:
 243         last = cache->key.objectid + cache->key.offset;
 244 wrapped:
 245         cache = btrfs_lookup_block_group(root->fs_info, last);
 246         if (!cache || cache->key.objectid >= total_fs_bytes) {
 247 no_cache:
 248                 if (!wrapped) {
 249                         wrapped = 1;
 250                         last = search_start;
 251                         data = BTRFS_BLOCK_GROUP_MIXED;
 252                         goto wrapped;
 253                 }
 254                 goto out;
 255         }
 256         if (cache_miss && !cache->cached) {
 257                 cache_block_group(root, cache);
 258                 last = cache_miss;
 259                 cache = btrfs_lookup_block_group(root->fs_info, last);
 260         }
 261         cache = btrfs_find_block_group(root, cache, last, data, 0);
 262         if (!cache)
 263                 goto no_cache;
 264         *cache_ret = cache;
 265         cache_miss = 0;
 266         goto again;
 267 }
 268
 269 static u64 div_factor(u64 num, int factor)
 270 {
 271         if (factor == 10)
 272                 return num;
 273         num *= factor;
 274         do_div(num, 10);
 275         return num;
 276 }
 277
 278 struct btrfs_block_group_cache *btrfs_find_block_group(struct btrfs_root *root,
 279                                                  struct btrfs_block_group_cache
 280                                                  *hint, u64 search_start,
 281                                                  int data, int owner)
 282 {
 283         struct btrfs_block_group_cache *cache;
 284         struct extent_io_tree *block_group_cache;
 285         struct btrfs_block_group_cache *found_group = NULL;
 286         struct btrfs_fs_info *info = root->fs_info;
 287         u64 used;
 288         u64 last = 0;
 289         u64 hint_last;
 290         u64 start;
 291         u64 end;
 292         u64 free_check;
 293         u64 ptr;
 294         u64 total_fs_bytes;
 295         int bit;
 296         int ret;
 297         int full_search = 0;
 298         int factor = 8;
 299         int data_swap = 0;
 300
 301         block_group_cache = &info->block_group_cache;
 302         total_fs_bytes = btrfs_super_total_bytes(&root->fs_info->super_copy);
 303
 304         if (!owner)
 305                 factor = 8;
 306
 307         if (data == BTRFS_BLOCK_GROUP_MIXED) {
 308                 bit = BLOCK_GROUP_DATA | BLOCK_GROUP_METADATA;
 309                 factor = 10;
 310         } else if (data)
 311                 bit = BLOCK_GROUP_DATA;
 312         else
 313                 bit = BLOCK_GROUP_METADATA;
 314
 315         if (search_start && search_start < total_fs_bytes) {
 316                 struct btrfs_block_group_cache *shint;
 317                 shint = btrfs_lookup_block_group(info, search_start);
 318                 if (shint && (shint->data == data ||
 319                               shint->data == BTRFS_BLOCK_GROUP_MIXED)) {
 320                         used = btrfs_block_group_used(&shint->item);
 321                         if (used + shint->pinned <
 322                             div_factor(shint->key.offset, factor)) {
 323                                 return shint;
 324                         }
 325                 }
 326         }
 327         if (hint && hint->key.objectid < total_fs_bytes &&
 328             (hint->data == data || hint->data == BTRFS_BLOCK_GROUP_MIXED)) {
 329                 used = btrfs_block_group_used(&hint->item);
 330                 if (used + hint->pinned <
 331                     div_factor(hint->key.offset, factor)) {
 332                         return hint;
 333                 }
 334                 last = hint->key.objectid + hint->key.offset;
 335                 hint_last = last;
 336         } else {
 337                 if (hint)
 338                         hint_last = max(hint->key.objectid, search_start);
 339                 else
 340                         hint_last = search_start;
 341
 342                 if (hint_last >= total_fs_bytes)
 343                         hint_last = search_start;
 344                 last = hint_last;
 345         }
 346 again:
 347         while(1) {
 348                 ret = find_first_extent_bit(block_group_cache, last,
 349                                             &start, &end, bit);
 350                 if (ret)
 351                         break;
 352
 353                 ret = get_state_private(block_group_cache, start, &ptr);
 354                 if (ret)
 355                         break;
 356
 357                 cache = (struct btrfs_block_group_cache *)(unsigned long)ptr;
 358                 last = cache->key.objectid + cache->key.offset;
 359                 used = btrfs_block_group_used(&cache->item);
 360
 361                 if (cache->key.objectid > total_fs_bytes)
 362                         break;
 363
 364                 if (full_search)
 365                         free_check = cache->key.offset;
 366                 else
 367                         free_check = div_factor(cache->key.offset, factor);
 368                 if (used + cache->pinned < free_check) {
 369                         found_group = cache;
 370                         goto found;
 371                 }
 372                 cond_resched();
 373         }
 374         if (!full_search) {
 375                 last = search_start;
 376                 full_search = 1;
 377                 goto again;
 378         }
 379         if (!data_swap) {
 380                 data_swap = 1;
 381                 bit = BLOCK_GROUP_DATA | BLOCK_GROUP_METADATA;
 382                 last = search_start;
 383                 goto again;
 384         }
 385 found:
 386         return found_group;
 387 }
 388
 389 static u64 hash_extent_ref(u64 root_objectid, u64 ref_generation,
 390                            u64 owner, u64 owner_offset)
 391 {
 392         u32 high_crc = ~(u32)0;
 393         u32 low_crc = ~(u32)0;
 394         __le64 lenum;
 395
 396         lenum = cpu_to_le64(root_objectid);
 397         high_crc = crc32c(high_crc, &lenum, sizeof(lenum));
 398         lenum = cpu_to_le64(ref_generation);
 399         low_crc = crc32c(low_crc, &lenum, sizeof(lenum));
 400
 401 #if 0
 402         lenum = cpu_to_le64(owner);
 403         low_crc = crc32c(low_crc, &lenum, sizeof(lenum));
 404         lenum = cpu_to_le64(owner_offset);
 405         low_crc = crc32c(low_crc, &lenum, sizeof(lenum));
 406 #endif
 407         return ((u64)high_crc << 32) | (u64)low_crc;
 408 }
 409
 410 static int match_extent_ref(struct extent_buffer *leaf,
 411                             struct btrfs_extent_ref *disk_ref,
 412                             struct btrfs_extent_ref *cpu_ref)
 413 {
 414         int ret;
 415         int len;
 416
 417         if (cpu_ref->objectid)
 418                 len = sizeof(*cpu_ref);
 419         else
 420                 len = 2 * sizeof(u64);
 421         ret = memcmp_extent_buffer(leaf, cpu_ref, (unsigned long)disk_ref,
 422                                    len);
 423         return ret == 0;
 424 }
 425
 426 static int noinline lookup_extent_backref(struct btrfs_trans_handle *trans,
 427                                           struct btrfs_root *root,
 428                                           struct btrfs_path *path, u64 bytenr,
 429                                           u64 root_objectid,
 430                                           u64 ref_generation, u64 owner,
 431                                           u64 owner_offset, int del)
 432 {
 433         u64 hash;
 434         struct btrfs_key key;
 435         struct btrfs_key found_key;
 436         struct btrfs_extent_ref ref;
 437         struct extent_buffer *leaf;
 438         struct btrfs_extent_ref *disk_ref;
 439         int ret;
 440         int ret2;
 441
 442         btrfs_set_stack_ref_root(&ref, root_objectid);
 443         btrfs_set_stack_ref_generation(&ref, ref_generation);
 444         btrfs_set_stack_ref_objectid(&ref, owner);
 445         btrfs_set_stack_ref_offset(&ref, owner_offset);
 446
 447         hash = hash_extent_ref(root_objectid, ref_generation, owner,
 448                                owner_offset);
 449         key.offset = hash;
 450         key.objectid = bytenr;
 451         key.type = BTRFS_EXTENT_REF_KEY;
 452
 453         while (1) {
 454                 ret = btrfs_search_slot(trans, root, &key, path,
 455                                         del ? -1 : 0, del);
 456                 if (ret < 0)
 457                         goto out;
 458                 leaf = path->nodes[0];
 459                 if (ret != 0) {
 460                         u32 nritems = btrfs_header_nritems(leaf);
 461                         if (path->slots[0] >= nritems) {
 462                                 ret2 = btrfs_next_leaf(root, path);
 463                                 if (ret2)
 464                                         goto out;
 465                                 leaf = path->nodes[0];
 466                         }
 467                         btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
 468                         if (found_key.objectid != bytenr ||
 469                             found_key.type != BTRFS_EXTENT_REF_KEY)
 470                                 goto out;
 471                         key.offset = found_key.offset;
 472                         if (del) {
 473                                 btrfs_release_path(root, path);
 474                                 continue;
 475                         }
 476                 }
 477                 disk_ref = btrfs_item_ptr(path->nodes[0],
 478                                           path->slots[0],
 479                                           struct btrfs_extent_ref);
 480                 if (match_extent_ref(path->nodes[0], disk_ref, &ref)) {
 481                         ret = 0;
 482                         goto out;
 483                 }
 484                 btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
 485                 key.offset = found_key.offset + 1;
 486                 btrfs_release_path(root, path);
 487         }
 488 out:
 489         return ret;
 490 }
 491
 492 /*
 493  * Back reference rules.  Back refs have three main goals:
 494  *
 495  * 1) differentiate between all holders of references to an extent so that
 496  *    when a reference is dropped we can make sure it was a valid reference
 497  *    before freeing the extent.
 498  *
 499  * 2) Provide enough information to quickly find the holders of an extent
 500  *    if we notice a given block is corrupted or bad.
 501  *
 502  * 3) Make it easy to migrate blocks for FS shrinking or storage pool
 503  *    maintenance.  This is actually the same as #2, but with a slightly
 504  *    different use case.
 505  *
 506  * File extents can be referenced by:
 507  *
 508  * - multiple snapshots, subvolumes, or different generations in one subvol
 509  * - different files inside a single subvolume (in theory, not implemented yet)
 510  * - different offsets inside a file (bookend extents in file.c)
 511  *
 512  * The extent ref structure has fields for:
 513  *
 514  * - Objectid of the subvolume root
 515  * - Generation number of the tree holding the reference
 516  * - objectid of the file holding the reference
 517  * - offset in the file corresponding to the key holding the reference
 518  *
 519  * When a file extent is allocated the fields are filled in:
 520  *     (root_key.objectid, trans->transid, inode objectid, offset in file)
 521  *
 522  * When a leaf is cow'd new references are added for every file extent found
 523  * in the leaf.  It looks the same as the create case, but trans->transid
 524  * will be different when the block is cow'd.
 525  *
 526  *     (root_key.objectid, trans->transid, inode objectid, offset in file)
 527  *
 528  * When a file extent is removed either during snapshot deletion or file
 529  * truncation, the corresponding back reference is found
 530  * by searching for:
 531  *
 532  *     (btrfs_header_owner(leaf), btrfs_header_generation(leaf),
 533  *      inode objectid, offset in file)
 534  *
 535  * Btree extents can be referenced by:
 536  *
 537  * - Different subvolumes
 538  * - Different generations of the same subvolume
 539  *
 540  * Storing sufficient information for a full reverse mapping of a btree
 541  * block would require storing the lowest key of the block in the backref,
 542  * and it would require updating that lowest key either before write out or
 543  * every time it changed.  Instead, the objectid of the lowest key is stored
 544  * along with the level of the tree block.  This provides a hint
 545  * about where in the btree the block can be found.  Searches through the
 546  * btree only need to look for a pointer to that block, so they stop one
 547  * level higher than the level recorded in the backref.
 548  *
 549  * Some btrees do not do reference counting on their extents.  These
 550  * include the extent tree and the tree of tree roots.  Backrefs for these
 551  * trees always have a generation of zero.
 552  *
 553  * When a tree block is created, back references are inserted:
 554  *
 555  * (root->root_key.objectid, trans->transid or zero, level, lowest_key_objectid)
 556  *
 557  * When a tree block is cow'd in a reference counted root,
 558  * new back references are added for all the blocks it points to.
 559  * These are of the form (trans->transid will have increased since creation):
 560  *
 561  * (root->root_key.objectid, trans->transid, level, lowest_key_objectid)
 562  *
 563  * Because the lowest_key_objectid and the level are just hints
 564  * they are not used when backrefs are deleted.  When a backref is deleted:
 565  *
 566  * if backref was for a tree root:
 567  *     root_objectid = root->root_key.objectid
 568  * else
 569  *     root_objectid = btrfs_header_owner(parent)
 570  *
 571  * (root_objectid, btrfs_header_generation(parent) or zero, 0, 0)
 572  *
 573  * Back Reference Key hashing:
 574  *
 575  * Back references have four fields, each 64 bits long.  Unfortunately,
 576  * This is hashed into a single 64 bit number and placed into the key offset.
 577  * The key objectid corresponds to the first byte in the extent, and the
 578  * key type is set to BTRFS_EXTENT_REF_KEY
 579  */
 580 int btrfs_insert_extent_backref(struct btrfs_trans_handle *trans,
 581                                  struct btrfs_root *root,
 582                                  struct btrfs_path *path, u64 bytenr,
 583                                  u64 root_objectid, u64 ref_generation,
 584                                  u64 owner, u64 owner_offset)
 585 {
 586         u64 hash;
 587         struct btrfs_key key;
 588         struct btrfs_extent_ref ref;
 589         struct btrfs_extent_ref *disk_ref;
 590         int ret;
 591
 592         btrfs_set_stack_ref_root(&ref, root_objectid);
 593         btrfs_set_stack_ref_generation(&ref, ref_generation);
 594         btrfs_set_stack_ref_objectid(&ref, owner);
 595         btrfs_set_stack_ref_offset(&ref, owner_offset);
 596
 597         hash = hash_extent_ref(root_objectid, ref_generation, owner,
 598                                owner_offset);
 599         key.offset = hash;
 600         key.objectid = bytenr;
 601         key.type = BTRFS_EXTENT_REF_KEY;
 602
 603         ret = btrfs_insert_empty_item(trans, root, path, &key, sizeof(ref));
 604         while (ret == -EEXIST) {
 605                 disk_ref = btrfs_item_ptr(path->nodes[0], path->slots[0],
 606                                           struct btrfs_extent_ref);
 607                 if (match_extent_ref(path->nodes[0], disk_ref, &ref))
 608                         goto out;
 609                 key.offset++;
 610                 btrfs_release_path(root, path);
 611                 ret = btrfs_insert_empty_item(trans, root, path, &key,
 612                                               sizeof(ref));
 613         }
 614         if (ret)
 615                 goto out;
 616         disk_ref = btrfs_item_ptr(path->nodes[0], path->slots[0],
 617                                   struct btrfs_extent_ref);
 618         write_extent_buffer(path->nodes[0], &ref, (unsigned long)disk_ref,
 619                             sizeof(ref));
 620         btrfs_mark_buffer_dirty(path->nodes[0]);
 621 out:
 622         btrfs_release_path(root, path);
 623         return ret;
 624 }
 625
 626 int btrfs_inc_extent_ref(struct btrfs_trans_handle *trans,
 627                                 struct btrfs_root *root,
 628                                 u64 bytenr, u64 num_bytes,
 629                                 u64 root_objectid, u64 ref_generation,
 630                                 u64 owner, u64 owner_offset)
 631 {
 632         struct btrfs_path *path;
 633         int ret;
 634         struct btrfs_key key;
 635         struct extent_buffer *l;
 636         struct btrfs_extent_item *item;
 637         u32 refs;
 638
 639         WARN_ON(num_bytes < root->sectorsize);
 640         path = btrfs_alloc_path();
 641         if (!path)
 642                 return -ENOMEM;
 643
 644         path->reada = 0;
 645         key.objectid = bytenr;
 646         btrfs_set_key_type(&key, BTRFS_EXTENT_ITEM_KEY);
 647         key.offset = num_bytes;
 648         ret = btrfs_search_slot(trans, root->fs_info->extent_root, &key, path,
 649                                 0, 1);
 650         if (ret < 0)
 651                 return ret;
 652         if (ret != 0) {
 653                 BUG();
 654         }
 655         BUG_ON(ret != 0);
 656         l = path->nodes[0];
 657         item = btrfs_item_ptr(l, path->slots[0], struct btrfs_extent_item);
 658         refs = btrfs_extent_refs(l, item);
 659         btrfs_set_extent_refs(l, item, refs + 1);
 660         btrfs_mark_buffer_dirty(path->nodes[0]);
 661
 662         btrfs_release_path(root->fs_info->extent_root, path);
 663
 664         path->reada = 0;
 665         ret = btrfs_insert_extent_backref(trans, root->fs_info->extent_root,
 666                                           path, bytenr, root_objectid,
 667                                           ref_generation, owner, owner_offset);
 668         BUG_ON(ret);
 669         finish_current_insert(trans, root->fs_info->extent_root);
 670         del_pending_extents(trans, root->fs_info->extent_root);
 671
 672         btrfs_free_path(path);
 673         return 0;
 674 }
 675
 676 int btrfs_extent_post_op(struct btrfs_trans_handle *trans,
 677                          struct btrfs_root *root)
 678 {
 679         finish_current_insert(trans, root->fs_info->extent_root);
 680         del_pending_extents(trans, root->fs_info->extent_root);
 681         return 0;
 682 }
 683
 684 static int lookup_extent_ref(struct btrfs_trans_handle *trans,
 685                              struct btrfs_root *root, u64 bytenr,
 686                              u64 num_bytes, u32 *refs)
 687 {
 688         struct btrfs_path *path;
 689         int ret;
 690         struct btrfs_key key;
 691         struct extent_buffer *l;
 692         struct btrfs_extent_item *item;
 693
 694         WARN_ON(num_bytes < root->sectorsize);
 695         path = btrfs_alloc_path();
 696         path->reada = 0;
 697         key.objectid = bytenr;
 698         key.offset = num_bytes;
 699         btrfs_set_key_type(&key, BTRFS_EXTENT_ITEM_KEY);
 700         ret = btrfs_search_slot(trans, root->fs_info->extent_root, &key, path,
 701                                 0, 0);
 702         if (ret < 0)
 703                 goto out;
 704         if (ret != 0) {
 705                 btrfs_print_leaf(root, path->nodes[0]);
 706                 printk("failed to find block number %Lu\n", bytenr);
 707                 BUG();
 708         }
 709         l = path->nodes[0];
 710         item = btrfs_item_ptr(l, path->slots[0], struct btrfs_extent_item);
 711         *refs = btrfs_extent_refs(l, item);
 712 out:
 713         btrfs_free_path(path);
 714         return 0;
 715 }
 716
 717 u32 btrfs_count_snapshots_in_path(struct btrfs_root *root,
 718                                   struct btrfs_path *count_path,
 719                                   u64 first_extent)
 720 {
 721         struct btrfs_root *extent_root = root->fs_info->extent_root;
 722         struct btrfs_path *path;
 723         u64 bytenr;
 724         u64 found_objectid;
 725         u64 root_objectid = root->root_key.objectid;
 726         u32 total_count = 0;
 727         u32 cur_count;
 728         u32 nritems;
 729         int ret;
 730         struct btrfs_key key;
 731         struct btrfs_key found_key;
 732         struct extent_buffer *l;
 733         struct btrfs_extent_item *item;
 734         struct btrfs_extent_ref *ref_item;
 735         int level = -1;
 736
 737         path = btrfs_alloc_path();
 738 again:
 739         if (level == -1)
 740                 bytenr = first_extent;
 741         else
 742                 bytenr = count_path->nodes[level]->start;
 743
 744         cur_count = 0;
 745         key.objectid = bytenr;
 746         key.offset = 0;
 747
 748         btrfs_set_key_type(&key, BTRFS_EXTENT_ITEM_KEY);
 749         ret = btrfs_search_slot(NULL, extent_root, &key, path, 0, 0);
 750         if (ret < 0)
 751                 goto out;
 752         BUG_ON(ret == 0);
 753
 754         l = path->nodes[0];
 755         btrfs_item_key_to_cpu(l, &found_key, path->slots[0]);
 756
 757         if (found_key.objectid != bytenr ||
 758             found_key.type != BTRFS_EXTENT_ITEM_KEY) {
 759                 goto out;
 760         }
 761
 762         item = btrfs_item_ptr(l, path->slots[0], struct btrfs_extent_item);
 763         while (1) {
 764                 l = path->nodes[0];
 765                 nritems = btrfs_header_nritems(l);
 766                 if (path->slots[0] >= nritems) {
 767                         ret = btrfs_next_leaf(extent_root, path);
 768                         if (ret == 0)
 769                                 continue;
 770                         break;
 771                 }
 772                 btrfs_item_key_to_cpu(l, &found_key, path->slots[0]);
 773                 if (found_key.objectid != bytenr)
 774                         break;
 775
 776                 if (found_key.type != BTRFS_EXTENT_REF_KEY) {
 777                         path->slots[0]++;
 778                         continue;
 779                 }
 780
 781                 cur_count++;
 782                 ref_item = btrfs_item_ptr(l, path->slots[0],
 783                                           struct btrfs_extent_ref);
 784                 found_objectid = btrfs_ref_root(l, ref_item);
 785
 786                 if (found_objectid != root_objectid) {
 787                         total_count = 2;
 788                         goto out;
 789                 }
 790                 total_count = 1;
 791                 path->slots[0]++;
 792         }
 793         if (cur_count == 0) {
 794                 total_count = 0;
 795                 goto out;
 796         }
 797         if (level >= 0 && root->node == count_path->nodes[level])
 798                 goto out;
 799         level++;
 800         btrfs_release_path(root, path);
 801         goto again;
 802
 803 out:
 804         btrfs_free_path(path);
 805         return total_count;
 806 }
 807 int btrfs_inc_root_ref(struct btrfs_trans_handle *trans,
 808                        struct btrfs_root *root, u64 owner_objectid)
 809 {
 810         u64 generation;
 811         u64 key_objectid;
 812         u64 level;
 813         u32 nritems;
 814         struct btrfs_disk_key disk_key;
 815
 816         level = btrfs_header_level(root->node);
 817         generation = trans->transid;
 818         nritems = btrfs_header_nritems(root->node);
 819         if (nritems > 0) {
 820                 if (level == 0)
 821                         btrfs_item_key(root->node, &disk_key, 0);
 822                 else
 823                         btrfs_node_key(root->node, &disk_key, 0);
 824                 key_objectid = btrfs_disk_key_objectid(&disk_key);
 825         } else {
 826                 key_objectid = 0;
 827         }
 828         return btrfs_inc_extent_ref(trans, root, root->node->start,
 829                                     root->node->len, owner_objectid,
 830                                     generation, level, key_objectid);
 831 }
 832
 833 int btrfs_inc_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 834                   struct extent_buffer *buf)
 835 {
 836         u64 bytenr;
 837         u32 nritems;
 838         struct btrfs_key key;
 839         struct btrfs_file_extent_item *fi;
 840         int i;
 841         int level;
 842         int ret;
 843         int faili;
 844
 845         if (!root->ref_cows)
 846                 return 0;
 847
 848         level = btrfs_header_level(buf);
 849         nritems = btrfs_header_nritems(buf);
 850         for (i = 0; i < nritems; i++) {
 851                 if (level == 0) {
 852                         u64 disk_bytenr;
 853                         btrfs_item_key_to_cpu(buf, &key, i);
 854                         if (btrfs_key_type(&key) != BTRFS_EXTENT_DATA_KEY)
 855                                 continue;
 856                         fi = btrfs_item_ptr(buf, i,
 857                                             struct btrfs_file_extent_item);
 858                         if (btrfs_file_extent_type(buf, fi) ==
 859                             BTRFS_FILE_EXTENT_INLINE)
 860                                 continue;
 861                         disk_bytenr = btrfs_file_extent_disk_bytenr(buf, fi);
 862                         if (disk_bytenr == 0)
 863                                 continue;
 864                         ret = btrfs_inc_extent_ref(trans, root, disk_bytenr,
 865                                     btrfs_file_extent_disk_num_bytes(buf, fi),
 866                                     root->root_key.objectid, trans->transid,
 867                                     key.objectid, key.offset);
 868                         if (ret) {
 869                                 faili = i;
 870                                 goto fail;
 871                         }
 872                 } else {
 873                         bytenr = btrfs_node_blockptr(buf, i);
 874                         btrfs_node_key_to_cpu(buf, &key, i);
 875                         ret = btrfs_inc_extent_ref(trans, root, bytenr,
 876                                            btrfs_level_size(root, level - 1),
 877                                            root->root_key.objectid,
 878                                            trans->transid,
 879                                            level - 1, key.objectid);
 880                         if (ret) {
 881                                 faili = i;
 882                                 goto fail;
 883                         }
 884                 }
 885         }
 886         return 0;
 887 fail:
 888         WARN_ON(1);
 889 #if 0
 890         for (i =0; i < faili; i++) {
 891                 if (level == 0) {
 892                         u64 disk_bytenr;
 893                         btrfs_item_key_to_cpu(buf, &key, i);
 894                         if (btrfs_key_type(&key) != BTRFS_EXTENT_DATA_KEY)
 895                                 continue;
 896                         fi = btrfs_item_ptr(buf, i,
 897                                             struct btrfs_file_extent_item);
 898                         if (btrfs_file_extent_type(buf, fi) ==
 899                             BTRFS_FILE_EXTENT_INLINE)
 900                                 continue;
 901                         disk_bytenr = btrfs_file_extent_disk_bytenr(buf, fi);
 902                         if (disk_bytenr == 0)
 903                                 continue;
 904                         err = btrfs_free_extent(trans, root, disk_bytenr,
 905                                     btrfs_file_extent_disk_num_bytes(buf,
 906                                                                       fi), 0);
 907                         BUG_ON(err);
 908                 } else {
 909                         bytenr = btrfs_node_blockptr(buf, i);
 910                         err = btrfs_free_extent(trans, root, bytenr,
 911                                         btrfs_level_size(root, level - 1), 0);
 912                         BUG_ON(err);
 913                 }
 914         }
 915 #endif
 916         return ret;
 917 }
 918
 919 static int write_one_cache_group(struct btrfs_trans_handle *trans,
 920                                  struct btrfs_root *root,
 921                                  struct btrfs_path *path,
 922                                  struct btrfs_block_group_cache *cache)
 923 {
 924         int ret;
 925         int pending_ret;
 926         struct btrfs_root *extent_root = root->fs_info->extent_root;
 927         unsigned long bi;
 928         struct extent_buffer *leaf;
 929
 930         ret = btrfs_search_slot(trans, extent_root, &cache->key, path, 0, 1);
 931         if (ret < 0)
 932                 goto fail;
 933         BUG_ON(ret);
 934
 935         leaf = path->nodes[0];
 936         bi = btrfs_item_ptr_offset(leaf, path->slots[0]);
 937         write_extent_buffer(leaf, &cache->item, bi, sizeof(cache->item));
 938         btrfs_mark_buffer_dirty(leaf);
 939         btrfs_release_path(extent_root, path);
 940 fail:
 941         finish_current_insert(trans, extent_root);
 942         pending_ret = del_pending_extents(trans, extent_root);
 943         if (ret)
 944                 return ret;
 945         if (pending_ret)
 946                 return pending_ret;
 947         return 0;
 948
 949 }
 950
 951 int btrfs_write_dirty_block_groups(struct btrfs_trans_handle *trans,
 952                                    struct btrfs_root *root)
 953 {
 954         struct extent_io_tree *block_group_cache;
 955         struct btrfs_block_group_cache *cache;
 956         int ret;
 957         int err = 0;
 958         int werr = 0;
 959         struct btrfs_path *path;
 960         u64 last = 0;
 961         u64 start;
 962         u64 end;
 963         u64 ptr;
 964
 965         block_group_cache = &root->fs_info->block_group_cache;
 966         path = btrfs_alloc_path();
 967         if (!path)
 968                 return -ENOMEM;
 969
 970         while(1) {
 971                 ret = find_first_extent_bit(block_group_cache, last,
 972                                             &start, &end, BLOCK_GROUP_DIRTY);
 973                 if (ret)
 974                         break;
 975
 976                 last = end + 1;
 977                 ret = get_state_private(block_group_cache, start, &ptr);
 978                 if (ret)
 979                         break;
 980
 981                 cache = (struct btrfs_block_group_cache *)(unsigned long)ptr;
 982                 err = write_one_cache_group(trans, root,
 983                                             path, cache);
 984                 /*
 985                  * if we fail to write the cache group, we want
 986                  * to keep it marked dirty in hopes that a later
 987                  * write will work
 988                  */
 989                 if (err) {
 990                         werr = err;
 991                         continue;
 992                 }
 993                 clear_extent_bits(block_group_cache, start, end,
 994                                   BLOCK_GROUP_DIRTY, GFP_NOFS);
 995         }
 996         btrfs_free_path(path);
 997         return werr;
 998 }
 999
1000 static int update_block_group(struct btrfs_trans_handle *trans,
1001                               struct btrfs_root *root,
1002                               u64 bytenr, u64 num_bytes, int alloc,
1003                               int mark_free, int data)
1004 {
1005         struct btrfs_block_group_cache *cache;
1006         struct btrfs_fs_info *info = root->fs_info;
1007         u64 total = num_bytes;
1008         u64 old_val;
1009         u64 byte_in_group;
1010         u64 start;
1011         u64 end;
1012
1013         while(total) {
1014                 cache = btrfs_lookup_block_group(info, bytenr);
1015                 if (!cache) {
1016                         return -1;
1017                 }
1018                 byte_in_group = bytenr - cache->key.objectid;
1019                 WARN_ON(byte_in_group > cache->key.offset);
1020                 start = cache->key.objectid;
1021                 end = start + cache->key.offset - 1;
1022                 set_extent_bits(&info->block_group_cache, start, end,
1023                                 BLOCK_GROUP_DIRTY, GFP_NOFS);
1024
1025                 old_val = btrfs_block_group_used(&cache->item);
1026                 num_bytes = min(total, cache->key.offset - byte_in_group);
1027                 if (alloc) {
1028                         if (cache->data != data &&
1029                             old_val < (cache->key.offset >> 1)) {
1030                                 int bit_to_clear;
1031                                 int bit_to_set;
1032                                 cache->data = data;
1033                                 if (data) {
1034                                         bit_to_clear = BLOCK_GROUP_METADATA;
1035                                         bit_to_set = BLOCK_GROUP_DATA;
1036                                         cache->item.flags &=
1037                                                 ~BTRFS_BLOCK_GROUP_MIXED;
1038                                         cache->item.flags |=
1039                                                 BTRFS_BLOCK_GROUP_DATA;
1040                                 } else {
1041                                         bit_to_clear = BLOCK_GROUP_DATA;
1042                                         bit_to_set = BLOCK_GROUP_METADATA;
1043                                         cache->item.flags &=
1044                                                 ~BTRFS_BLOCK_GROUP_MIXED;
1045                                         cache->item.flags &=
1046                                                 ~BTRFS_BLOCK_GROUP_DATA;
1047                                 }
1048                                 clear_extent_bits(&info->block_group_cache,
1049                                                   start, end, bit_to_clear,
1050                                                   GFP_NOFS);
1051                                 set_extent_bits(&info->block_group_cache,
1052                                                 start, end, bit_to_set,
1053                                                 GFP_NOFS);
1054                         } else if (cache->data != data &&
1055                                    cache->data != BTRFS_BLOCK_GROUP_MIXED) {
1056                                 cache->data = BTRFS_BLOCK_GROUP_MIXED;
1057                                 set_extent_bits(&info->block_group_cache,
1058                                                 start, end,
1059                                                 BLOCK_GROUP_DATA |
1060                                                 BLOCK_GROUP_METADATA,
1061                                                 GFP_NOFS);
1062                         }
1063                         old_val += num_bytes;
1064                 } else {
1065                         old_val -= num_bytes;
1066                         if (mark_free) {
1067                                 set_extent_dirty(&info->free_space_cache,
1068                                                  bytenr, bytenr + num_bytes - 1,
1069                                                  GFP_NOFS);
1070                         }
1071                 }
1072                 btrfs_set_block_group_used(&cache->item, old_val);
1073                 total -= num_bytes;
1074                 bytenr += num_bytes;
1075         }
1076         return 0;
1077 }
1078 static int update_pinned_extents(struct btrfs_root *root,
1079                                 u64 bytenr, u64 num, int pin)
1080 {
1081         u64 len;
1082         struct btrfs_block_group_cache *cache;
1083         struct btrfs_fs_info *fs_info = root->fs_info;
1084
1085         if (pin) {
1086                 set_extent_dirty(&fs_info->pinned_extents,
1087                                 bytenr, bytenr + num - 1, GFP_NOFS);
1088         } else {
1089                 clear_extent_dirty(&fs_info->pinned_extents,
1090                                 bytenr, bytenr + num - 1, GFP_NOFS);
1091         }
1092         while (num > 0) {
1093                 cache = btrfs_lookup_block_group(fs_info, bytenr);
1094                 WARN_ON(!cache);
1095                 len = min(num, cache->key.offset -
1096                           (bytenr - cache->key.objectid));
1097                 if (pin) {
1098                         cache->pinned += len;
1099                         fs_info->total_pinned += len;
1100                 } else {
1101                         cache->pinned -= len;
1102                         fs_info->total_pinned -= len;
1103                 }
1104                 bytenr += len;
1105                 num -= len;
1106         }
1107         return 0;
1108 }
1109
1110 int btrfs_copy_pinned(struct btrfs_root *root, struct extent_io_tree *copy)
1111 {
1112         u64 last = 0;
1113         u64 start;
1114         u64 end;
1115         struct extent_io_tree *pinned_extents = &root->fs_info->pinned_extents;
1116         int ret;
1117
1118         while(1) {
1119                 ret = find_first_extent_bit(pinned_extents, last,
1120                                             &start, &end, EXTENT_DIRTY);
1121                 if (ret)
1122                         break;
1123                 set_extent_dirty(copy, start, end, GFP_NOFS);
1124                 last = end + 1;
1125         }
1126         return 0;
1127 }
1128
1129 int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans,
1130                                struct btrfs_root *root,
1131                                struct extent_io_tree *unpin)
1132 {
1133         u64 start;
1134         u64 end;
1135         int ret;
1136         struct extent_io_tree *free_space_cache;
1137         free_space_cache = &root->fs_info->free_space_cache;
1138
1139         while(1) {
1140                 ret = find_first_extent_bit(unpin, 0, &start, &end,
1141                                             EXTENT_DIRTY);
1142                 if (ret)
1143                         break;
1144                 update_pinned_extents(root, start, end + 1 - start, 0);
1145                 clear_extent_dirty(unpin, start, end, GFP_NOFS);
1146                 set_extent_dirty(free_space_cache, start, end, GFP_NOFS);
1147         }
1148         return 0;
1149 }
1150
1151 static int finish_current_insert(struct btrfs_trans_handle *trans,
1152                                  struct btrfs_root *extent_root)
1153 {
1154         u64 start;
1155         u64 end;
1156         struct btrfs_fs_info *info = extent_root->fs_info;
1157         struct extent_buffer *eb;
1158         struct btrfs_path *path;
1159         struct btrfs_key ins;
1160         struct btrfs_disk_key first;
1161         struct btrfs_extent_item extent_item;
1162         int ret;
1163         int level;
1164         int err = 0;
1165
1166         btrfs_set_stack_extent_refs(&extent_item, 1);
1167         btrfs_set_key_type(&ins, BTRFS_EXTENT_ITEM_KEY);
1168         path = btrfs_alloc_path();
1169
1170         while(1) {
1171                 ret = find_first_extent_bit(&info->extent_ins, 0, &start,
1172                                             &end, EXTENT_LOCKED);
1173                 if (ret)
1174                         break;
1175
1176                 ins.objectid = start;
1177                 ins.offset = end + 1 - start;
1178                 err = btrfs_insert_item(trans, extent_root, &ins,
1179                                         &extent_item, sizeof(extent_item));
1180                 clear_extent_bits(&info->extent_ins, start, end, EXTENT_LOCKED,
1181                                   GFP_NOFS);
1182                 eb = read_tree_block(extent_root, ins.objectid, ins.offset);
1183                 level = btrfs_header_level(eb);
1184                 if (level == 0) {
1185                         btrfs_item_key(eb, &first, 0);
1186                 } else {
1187                         btrfs_node_key(eb, &first, 0);
1188                 }
1189                 err = btrfs_insert_extent_backref(trans, extent_root, path,
1190                                           start, extent_root->root_key.objectid,
1191                                           0, level,
1192                                           btrfs_disk_key_objectid(&first));
1193                 BUG_ON(err);
1194                 free_extent_buffer(eb);
1195         }
1196         btrfs_free_path(path);
1197         return 0;
1198 }
1199
1200 static int pin_down_bytes(struct btrfs_root *root, u64 bytenr, u32 num_bytes,
1201                           int pending)
1202 {
1203         int err = 0;
1204         struct extent_buffer *buf;
1205
1206         if (!pending) {
1207                 buf = btrfs_find_tree_block(root, bytenr, num_bytes);
1208                 if (buf) {
1209                         if (btrfs_buffer_uptodate(buf)) {
1210                                 u64 transid =
1211                                     root->fs_info->running_transaction->transid;
1212                                 u64 header_transid =
1213                                         btrfs_header_generation(buf);
1214                                 if (header_transid == transid) {
1215                                         clean_tree_block(NULL, root, buf);
1216                                         free_extent_buffer(buf);
1217                                         return 1;
1218                                 }
1219                         }
1220                         free_extent_buffer(buf);
1221                 }
1222                 update_pinned_extents(root, bytenr, num_bytes, 1);
1223         } else {
1224                 set_extent_bits(&root->fs_info->pending_del,
1225                                 bytenr, bytenr + num_bytes - 1,
1226                                 EXTENT_LOCKED, GFP_NOFS);
1227         }
1228         BUG_ON(err < 0);
1229         return 0;
1230 }
1231
1232 /*
1233  * remove an extent from the root, returns 0 on success
1234  */
1235 static int __free_extent(struct btrfs_trans_handle *trans, struct btrfs_root
1236                          *root, u64 bytenr, u64 num_bytes,
1237                          u64 root_objectid, u64 ref_generation,
1238                          u64 owner_objectid, u64 owner_offset, int pin,
1239                          int mark_free)
1240 {
1241         struct btrfs_path *path;
1242         struct btrfs_key key;
1243         struct btrfs_fs_info *info = root->fs_info;
1244         struct btrfs_root *extent_root = info->extent_root;
1245         struct extent_buffer *leaf;
1246         int ret;
1247         struct btrfs_extent_item *ei;
1248         u32 refs;
1249
1250         key.objectid = bytenr;
1251         btrfs_set_key_type(&key, BTRFS_EXTENT_ITEM_KEY);
1252         key.offset = num_bytes;
1253         path = btrfs_alloc_path();
1254         if (!path)
1255                 return -ENOMEM;
1256
1257         path->reada = 0;
1258         ret = lookup_extent_backref(trans, extent_root, path,
1259                                     bytenr, root_objectid,
1260                                     ref_generation,
1261                                     owner_objectid, owner_offset, 1);
1262         if (ret == 0) {
1263                 ret = btrfs_del_item(trans, extent_root, path);
1264         } else {
1265                 btrfs_print_leaf(extent_root, path->nodes[0]);
1266                 WARN_ON(1);
1267                 printk("Unable to find ref byte nr %Lu root %Lu "
1268                        " gen %Lu owner %Lu offset %Lu\n", bytenr,
1269                        root_objectid, ref_generation, owner_objectid,
1270                        owner_offset);
1271         }
1272         btrfs_release_path(extent_root, path);
1273         ret = btrfs_search_slot(trans, extent_root, &key, path, -1, 1);
1274         if (ret < 0)
1275                 return ret;
1276         BUG_ON(ret);
1277
1278         leaf = path->nodes[0];
1279         ei = btrfs_item_ptr(leaf, path->slots[0],
1280                             struct btrfs_extent_item);
1281         refs = btrfs_extent_refs(leaf, ei);
1282         BUG_ON(refs == 0);
1283         refs -= 1;
1284         btrfs_set_extent_refs(leaf, ei, refs);
1285         btrfs_mark_buffer_dirty(leaf);
1286
1287         if (refs == 0) {
1288                 u64 super_used;
1289                 u64 root_used;
1290
1291                 if (pin) {
1292                         ret = pin_down_bytes(root, bytenr, num_bytes, 0);
1293                         if (ret > 0)
1294                                 mark_free = 1;
1295                         BUG_ON(ret < 0);
1296                 }
1297
1298                 /* block accounting for super block */
1299                 super_used = btrfs_super_bytes_used(&info->super_copy);
1300                 btrfs_set_super_bytes_used(&info->super_copy,
1301                                            super_used - num_bytes);
1302
1303                 /* block accounting for root item */
1304                 root_used = btrfs_root_used(&root->root_item);
1305                 btrfs_set_root_used(&root->root_item,
1306                                            root_used - num_bytes);
1307
1308                 ret = btrfs_del_item(trans, extent_root, path);
1309                 if (ret) {
1310                         return ret;
1311                 }
1312                 ret = update_block_group(trans, root, bytenr, num_bytes, 0,
1313                                          mark_free, 0);
1314                 BUG_ON(ret);
1315         }
1316         btrfs_free_path(path);
1317         finish_current_insert(trans, extent_root);
1318         return ret;
1319 }
1320
1321 /*
1322  * find all the blocks marked as pending in the radix tree and remove
1323  * them from the extent map
1324  */
1325 static int del_pending_extents(struct btrfs_trans_handle *trans, struct
1326                                btrfs_root *extent_root)
1327 {
1328         int ret;
1329         int err = 0;
1330         u64 start;
1331         u64 end;
1332         struct extent_io_tree *pending_del;
1333         struct extent_io_tree *pinned_extents;
1334
1335         pending_del = &extent_root->fs_info->pending_del;
1336         pinned_extents = &extent_root->fs_info->pinned_extents;
1337
1338         while(1) {
1339                 ret = find_first_extent_bit(pending_del, 0, &start, &end,
1340                                             EXTENT_LOCKED);
1341                 if (ret)
1342                         break;
1343                 update_pinned_extents(extent_root, start, end + 1 - start, 1);
1344                 clear_extent_bits(pending_del, start, end, EXTENT_LOCKED,
1345                                   GFP_NOFS);
1346                 ret = __free_extent(trans, extent_root,
1347                                      start, end + 1 - start,
1348                                      extent_root->root_key.objectid,
1349                                      0, 0, 0, 0, 0);
1350                 if (ret)
1351                         err = ret;
1352         }
1353         return err;
1354 }
1355
1356 /*
1357  * remove an extent from the root, returns 0 on success
1358  */
1359 int btrfs_free_extent(struct btrfs_trans_handle *trans, struct btrfs_root
1360                       *root, u64 bytenr, u64 num_bytes,
1361                       u64 root_objectid, u64 ref_generation,
1362                       u64 owner_objectid, u64 owner_offset, int pin)
1363 {
1364         struct btrfs_root *extent_root = root->fs_info->extent_root;
1365         int pending_ret;
1366         int ret;
1367
1368         WARN_ON(num_bytes < root->sectorsize);
1369         if (!root->ref_cows)
1370                 ref_generation = 0;
1371
1372         if (root == extent_root) {
1373                 pin_down_bytes(root, bytenr, num_bytes, 1);
1374                 return 0;
1375         }
1376         ret = __free_extent(trans, root, bytenr, num_bytes, root_objectid,
1377                             ref_generation, owner_objectid, owner_offset,
1378                             pin, pin == 0);
1379         pending_ret = del_pending_extents(trans, root->fs_info->extent_root);
1380         return ret ? ret : pending_ret;
1381 }
1382
1383 static u64 stripe_align(struct btrfs_root *root, u64 val)
1384 {
1385         u64 mask = ((u64)root->stripesize - 1);
1386         u64 ret = (val + mask) & ~mask;
1387         return ret;
1388 }
1389
1390 /*
1391  * walks the btree of allocated extents and find a hole of a given size.
1392  * The key ins is changed to record the hole:
1393  * ins->objectid == block start
1394  * ins->flags = BTRFS_EXTENT_ITEM_KEY
1395  * ins->offset == number of blocks
1396  * Any available blocks before search_start are skipped.
1397  */
1398 static int noinline find_free_extent(struct btrfs_trans_handle *trans,
1399                                      struct btrfs_root *orig_root,
1400                                      u64 num_bytes, u64 empty_size,
1401                                      u64 search_start, u64 search_end,
1402                                      u64 hint_byte, struct btrfs_key *ins,
1403                                      u64 exclude_start, u64 exclude_nr,
1404                                      int data)
1405 {
1406         struct btrfs_path *path;
1407         struct btrfs_key key;
1408         u64 hole_size = 0;
1409         u64 aligned;
1410         int ret;
1411         int slot = 0;
1412         u64 last_byte = 0;
1413         u64 orig_search_start = search_start;
1414         int start_found;
1415         struct extent_buffer *l;
1416         struct btrfs_root * root = orig_root->fs_info->extent_root;
1417         struct btrfs_fs_info *info = root->fs_info;
1418         u64 total_needed = num_bytes;
1419         int level;
1420         struct btrfs_block_group_cache *block_group;
1421         int full_scan = 0;
1422         int wrapped = 0;
1423         u64 cached_start;
1424
1425         WARN_ON(num_bytes < root->sectorsize);
1426         btrfs_set_key_type(ins, BTRFS_EXTENT_ITEM_KEY);
1427
1428         level = btrfs_header_level(root->node);
1429
1430         if (num_bytes >= 32 * 1024 * 1024 && hint_byte) {
1431                 data = BTRFS_BLOCK_GROUP_MIXED;
1432         }
1433
1434         /* for SSD, cluster allocations together as much as possible */
1435         if (btrfs_test_opt(root, SSD)) {
1436                 if (!data) {
1437                         if (root->fs_info->last_alloc)
1438                                 hint_byte = root->fs_info->last_alloc;
1439                         else {
1440                                 hint_byte = hint_byte &
1441                                         ~((u64)BTRFS_BLOCK_GROUP_SIZE - 1);
1442                                 empty_size += 16 * 1024 * 1024;
1443                         }
1444                 }
1445         }
1446
1447         search_end = min(search_end,
1448                          btrfs_super_total_bytes(&info->super_copy));
1449         if (hint_byte) {
1450                 block_group = btrfs_lookup_block_group(info, hint_byte);
1451                 if (!block_group)
1452                         hint_byte = search_start;
1453                 block_group = btrfs_find_block_group(root, block_group,
1454                                                      hint_byte, data, 1);
1455         } else {
1456                 block_group = btrfs_find_block_group(root,
1457                                                      trans->block_group,
1458                                                      search_start, data, 1);
1459         }
1460
1461         total_needed += empty_size;
1462         path = btrfs_alloc_path();
1463 check_failed:
1464         if (!block_group) {
1465                 block_group = btrfs_lookup_block_group(info, search_start);
1466                 if (!block_group)
1467                         block_group = btrfs_lookup_block_group(info,
1468                                                        orig_search_start);
1469         }
1470         search_start = find_search_start(root, &block_group, search_start,
1471                                          total_needed, data);
1472
1473         if (!data && btrfs_test_opt(root, SSD) && info->last_alloc &&
1474             search_start != info->last_alloc) {
1475                 info->last_alloc = 0;
1476                 if (!empty_size) {
1477                         empty_size += 16 * 1024 * 1024;
1478                         total_needed += empty_size;
1479                 }
1480                 search_start = find_search_start(root, &block_group,
1481                                                  search_start, total_needed,
1482                                                  data);
1483         }
1484
1485         search_start = stripe_align(root, search_start);
1486         cached_start = search_start;
1487         btrfs_init_path(path);
1488         ins->objectid = search_start;
1489         ins->offset = 0;
1490         start_found = 0;
1491         path->reada = 2;
1492
1493         ret = btrfs_search_slot(trans, root, ins, path, 0, 0);
1494         if (ret < 0)
1495                 goto error;
1496         ret = find_previous_extent(root, path);
1497         if (ret < 0)
1498                 goto error;
1499         l = path->nodes[0];
1500         btrfs_item_key_to_cpu(l, &key, path->slots[0]);
1501         while (1) {
1502                 l = path->nodes[0];
1503                 slot = path->slots[0];
1504                 if (slot >= btrfs_header_nritems(l)) {
1505                         ret = btrfs_next_leaf(root, path);
1506                         if (ret == 0)
1507                                 continue;
1508                         if (ret < 0)
1509                                 goto error;
1510
1511                         search_start = max(search_start,
1512                                            block_group->key.objectid);
1513                         if (!start_found) {
1514                                 aligned = stripe_align(root, search_start);
1515                                 ins->objectid = aligned;
1516                                 if (aligned >= search_end) {
1517                                         ret = -ENOSPC;
1518                                         goto error;
1519                                 }
1520                                 ins->offset = search_end - aligned;
1521                                 start_found = 1;
1522                                 goto check_pending;
1523                         }
1524                         ins->objectid = stripe_align(root,
1525                                                      last_byte > search_start ?
1526                                                      last_byte : search_start);
1527                         if (search_end <= ins->objectid) {
1528                                 ret = -ENOSPC;
1529                                 goto error;
1530                         }
1531                         ins->offset = search_end - ins->objectid;
1532                         BUG_ON(ins->objectid >= search_end);
1533                         goto check_pending;
1534                 }
1535                 btrfs_item_key_to_cpu(l, &key, slot);
1536
1537                 if (key.objectid >= search_start && key.objectid > last_byte &&
1538                     start_found) {
1539                         if (last_byte < search_start)
1540                                 last_byte = search_start;
1541                         aligned = stripe_align(root, last_byte);
1542                         hole_size = key.objectid - aligned;
1543                         if (key.objectid > aligned && hole_size >= num_bytes) {
1544                                 ins->objectid = aligned;
1545                                 ins->offset = hole_size;
1546                                 goto check_pending;
1547                         }
1548                 }
1549                 if (btrfs_key_type(&key) != BTRFS_EXTENT_ITEM_KEY) {
1550                         if (!start_found && btrfs_key_type(&key) ==
1551                             BTRFS_BLOCK_GROUP_ITEM_KEY) {
1552                                 last_byte = key.objectid;
1553                                 start_found = 1;
1554                         }
1555                         goto next;
1556                 }
1557
1558
1559                 start_found = 1;
1560                 last_byte = key.objectid + key.offset;
1561
1562                 if (!full_scan && data != BTRFS_BLOCK_GROUP_MIXED &&
1563                     last_byte >= block_group->key.objectid +
1564                     block_group->key.offset) {
1565                         btrfs_release_path(root, path);
1566                         search_start = block_group->key.objectid +
1567                                 block_group->key.offset;
1568                         goto new_group;
1569                 }
1570 next:
1571                 path->slots[0]++;
1572                 cond_resched();
1573         }
1574 check_pending:
1575         /* we have to make sure we didn't find an extent that has already
1576          * been allocated by the map tree or the original allocation
1577          */
1578         btrfs_release_path(root, path);
1579         BUG_ON(ins->objectid < search_start);
1580
1581         if (ins->objectid + num_bytes >= search_end)
1582                 goto enospc;
1583         if (!full_scan && data != BTRFS_BLOCK_GROUP_MIXED &&
1584             ins->objectid + num_bytes > block_group->
1585             key.objectid + block_group->key.offset) {
1586                 search_start = block_group->key.objectid +
1587                         block_group->key.offset;
1588                 goto new_group;
1589         }
1590         if (test_range_bit(&info->extent_ins, ins->objectid,
1591                            ins->objectid + num_bytes -1, EXTENT_LOCKED, 0)) {
1592                 search_start = ins->objectid + num_bytes;
1593                 goto new_group;
1594         }
1595         if (test_range_bit(&info->pinned_extents, ins->objectid,
1596                            ins->objectid + num_bytes -1, EXTENT_DIRTY, 0)) {
1597                 search_start = ins->objectid + num_bytes;
1598                 goto new_group;
1599         }
1600         if (exclude_nr > 0 && (ins->objectid + num_bytes > exclude_start &&
1601             ins->objectid < exclude_start + exclude_nr)) {
1602                 search_start = exclude_start + exclude_nr;
1603                 goto new_group;
1604         }
1605         if (!data) {
1606                 block_group = btrfs_lookup_block_group(info, ins->objectid);
1607                 if (block_group)
1608                         trans->block_group = block_group;
1609         }
1610         ins->offset = num_bytes;
1611         btrfs_free_path(path);
1612         return 0;
1613
1614 new_group:
1615         if (search_start + num_bytes >= search_end) {
1616 enospc:
1617                 search_start = orig_search_start;
1618                 if (full_scan) {
1619                         ret = -ENOSPC;
1620                         goto error;
1621                 }
1622                 if (wrapped) {
1623                         if (!full_scan)
1624                                 total_needed -= empty_size;
1625                         full_scan = 1;
1626                         data = BTRFS_BLOCK_GROUP_MIXED;
1627                 } else
1628                         wrapped = 1;
1629         }
1630         block_group = btrfs_lookup_block_group(info, search_start);
1631         cond_resched();
1632         block_group = btrfs_find_block_group(root, block_group,
1633                                              search_start, data, 0);
1634         goto check_failed;
1635
1636 error:
1637         btrfs_release_path(root, path);
1638         btrfs_free_path(path);
1639         if (btrfs_test_opt(root, SSD) && !ret && !data)
1640                 info->last_alloc = ins->objectid + ins->offset;
1641         return ret;
1642 }
1643 /*
1644  * finds a free extent and does all the dirty work required for allocation
1645  * returns the key for the extent through ins, and a tree buffer for
1646  * the first block of the extent through buf.
1647  *
1648  * returns 0 if everything worked, non-zero otherwise.
1649  */
1650 int btrfs_alloc_extent(struct btrfs_trans_handle *trans,
1651                        struct btrfs_root *root,
1652                        u64 num_bytes, u64 root_objectid, u64 ref_generation,
1653                        u64 owner, u64 owner_offset,
1654                        u64 empty_size, u64 hint_byte,
1655                        u64 search_end, struct btrfs_key *ins, int data)
1656 {
1657         int ret;
1658         int pending_ret;
1659         u64 super_used;
1660         u64 root_used;
1661         u64 search_start = 0;
1662         u64 new_hint;
1663         struct btrfs_fs_info *info = root->fs_info;
1664         struct btrfs_root *extent_root = info->extent_root;
1665         struct btrfs_extent_item extent_item;
1666         struct btrfs_path *path;
1667
1668         btrfs_set_stack_extent_refs(&extent_item, 1);
1669
1670         new_hint = max(hint_byte, root->fs_info->alloc_start);
1671         if (new_hint < btrfs_super_total_bytes(&info->super_copy))
1672                 hint_byte = new_hint;
1673
1674         WARN_ON(num_bytes < root->sectorsize);
1675         ret = find_free_extent(trans, root, num_bytes, empty_size,
1676                                search_start, search_end, hint_byte, ins,
1677                                trans->alloc_exclude_start,
1678                                trans->alloc_exclude_nr, data);
1679         BUG_ON(ret);
1680         if (ret)
1681                 return ret;
1682
1683         /* block accounting for super block */
1684         super_used = btrfs_super_bytes_used(&info->super_copy);
1685         btrfs_set_super_bytes_used(&info->super_copy, super_used + num_bytes);
1686
1687         /* block accounting for root item */
1688         root_used = btrfs_root_used(&root->root_item);
1689         btrfs_set_root_used(&root->root_item, root_used + num_bytes);
1690
1691         clear_extent_dirty(&root->fs_info->free_space_cache,
1692                            ins->objectid, ins->objectid + ins->offset - 1,
1693                            GFP_NOFS);
1694
1695         if (root == extent_root) {
1696                 set_extent_bits(&root->fs_info->extent_ins, ins->objectid,
1697                                 ins->objectid + ins->offset - 1,
1698                                 EXTENT_LOCKED, GFP_NOFS);
1699                 WARN_ON(data == 1);
1700                 goto update_block;
1701         }
1702
1703         WARN_ON(trans->alloc_exclude_nr);
1704         trans->alloc_exclude_start = ins->objectid;
1705         trans->alloc_exclude_nr = ins->offset;
1706         ret = btrfs_insert_item(trans, extent_root, ins, &extent_item,
1707                                 sizeof(extent_item));
1708
1709         trans->alloc_exclude_start = 0;
1710         trans->alloc_exclude_nr = 0;
1711         BUG_ON(ret);
1712
1713         path = btrfs_alloc_path();
1714         BUG_ON(!path);
1715         ret = btrfs_insert_extent_backref(trans, extent_root, path,
1716                                           ins->objectid, root_objectid,
1717                                           ref_generation, owner, owner_offset);
1718
1719         BUG_ON(ret);
1720         btrfs_free_path(path);
1721         finish_current_insert(trans, extent_root);
1722         pending_ret = del_pending_extents(trans, extent_root);
1723
1724         if (ret) {
1725                 return ret;
1726         }
1727         if (pending_ret) {
1728                 return pending_ret;
1729         }
1730
1731 update_block:
1732         ret = update_block_group(trans, root, ins->objectid, ins->offset, 1, 0,
1733                                  data);
1734         BUG_ON(ret);
1735         return 0;
1736 }
1737
1738 /*
1739  * helper function to allocate a block for a given tree
1740  * returns the tree buffer or NULL.
1741  */
1742 struct extent_buffer *btrfs_alloc_free_block(struct btrfs_trans_handle *trans,
1743                                              struct btrfs_root *root,
1744                                              u32 blocksize,
1745                                              u64 root_objectid, u64 hint,
1746                                              u64 empty_size)
1747 {
1748         u64 ref_generation;
1749
1750         if (root->ref_cows)
1751                 ref_generation = trans->transid;
1752         else
1753                 ref_generation = 0;
1754
1755
1756         return __btrfs_alloc_free_block(trans, root, blocksize, root_objectid,
1757                                         ref_generation, 0, 0, hint, empty_size);
1758 }
1759
1760 /*
1761  * helper function to allocate a block for a given tree
1762  * returns the tree buffer or NULL.
1763  */
1764 struct extent_buffer *__btrfs_alloc_free_block(struct btrfs_trans_handle *trans,
1765                                              struct btrfs_root *root,
1766                                              u32 blocksize,
1767                                              u64 root_objectid,
1768                                              u64 ref_generation,
1769                                              u64 first_objectid,
1770                                              int level,
1771                                              u64 hint,
1772                                              u64 empty_size)
1773 {
1774         struct btrfs_key ins;
1775         int ret;
1776         struct extent_buffer *buf;
1777
1778         ret = btrfs_alloc_extent(trans, root, blocksize,
1779                                  root_objectid, ref_generation,
1780                                  level, first_objectid, empty_size, hint,
1781                                  (u64)-1, &ins, 0);
1782         if (ret) {
1783                 BUG_ON(ret > 0);
1784                 return ERR_PTR(ret);
1785         }
1786         buf = btrfs_find_create_tree_block(root, ins.objectid, blocksize);
1787         if (!buf) {
1788                 btrfs_free_extent(trans, root, ins.objectid, blocksize,
1789                                   root->root_key.objectid, ref_generation,
1790                                   0, 0, 0);
1791                 return ERR_PTR(-ENOMEM);
1792         }
1793         btrfs_set_header_generation(buf, trans->transid);
1794         clean_tree_block(trans, root, buf);
1795         wait_on_tree_block_writeback(root, buf);
1796         btrfs_set_buffer_uptodate(buf);
1797
1798         if (PageDirty(buf->first_page)) {
1799                 printk("page %lu dirty\n", buf->first_page->index);
1800                 WARN_ON(1);
1801         }
1802
1803         set_extent_dirty(&trans->transaction->dirty_pages, buf->start,
1804                          buf->start + buf->len - 1, GFP_NOFS);
1805         set_extent_bits(&BTRFS_I(root->fs_info->btree_inode)->io_tree,
1806                         buf->start, buf->start + buf->len - 1,
1807                         EXTENT_CSUM, GFP_NOFS);
1808         buf->flags |= EXTENT_CSUM;
1809         if (!btrfs_test_opt(root, SSD))
1810                 btrfs_set_buffer_defrag(buf);
1811         trans->blocks_used++;
1812         return buf;
1813 }
1814
1815 static int noinline drop_leaf_ref(struct btrfs_trans_handle *trans,
1816                                   struct btrfs_root *root,
1817                                   struct extent_buffer *leaf)
1818 {
1819         u64 leaf_owner;
1820         u64 leaf_generation;
1821         struct btrfs_key key;
1822         struct btrfs_file_extent_item *fi;
1823         int i;
1824         int nritems;
1825         int ret;
1826
1827         BUG_ON(!btrfs_is_leaf(leaf));
1828         nritems = btrfs_header_nritems(leaf);
1829         leaf_owner = btrfs_header_owner(leaf);
1830         leaf_generation = btrfs_header_generation(leaf);
1831
1832         for (i = 0; i < nritems; i++) {
1833                 u64 disk_bytenr;
1834
1835                 btrfs_item_key_to_cpu(leaf, &key, i);
1836                 if (btrfs_key_type(&key) != BTRFS_EXTENT_DATA_KEY)
1837                         continue;
1838                 fi = btrfs_item_ptr(leaf, i, struct btrfs_file_extent_item);
1839                 if (btrfs_file_extent_type(leaf, fi) ==
1840                     BTRFS_FILE_EXTENT_INLINE)
1841                         continue;
1842                 /*
1843                  * FIXME make sure to insert a trans record that
1844                  * repeats the snapshot del on crash
1845                  */
1846                 disk_bytenr = btrfs_file_extent_disk_bytenr(leaf, fi);
1847                 if (disk_bytenr == 0)
1848                         continue;
1849                 ret = btrfs_free_extent(trans, root, disk_bytenr,
1850                                 btrfs_file_extent_disk_num_bytes(leaf, fi),
1851                                 leaf_owner, leaf_generation,
1852                                 key.objectid, key.offset, 0);
1853                 BUG_ON(ret);
1854         }
1855         return 0;
1856 }
1857
1858 static void noinline reada_walk_down(struct btrfs_root *root,
1859                                      struct extent_buffer *node,
1860                                      int slot)
1861 {
1862         u64 bytenr;
1863         u64 last = 0;
1864         u32 nritems;
1865         u32 refs;
1866         u32 blocksize;
1867         int ret;
1868         int i;
1869         int level;
1870         int skipped = 0;
1871
1872         nritems = btrfs_header_nritems(node);
1873         level = btrfs_header_level(node);
1874         if (level)
1875                 return;
1876
1877         for (i = slot; i < nritems && skipped < 32; i++) {
1878                 bytenr = btrfs_node_blockptr(node, i);
1879                 if (last && ((bytenr > last && bytenr - last > 32 * 1024) ||
1880                              (last > bytenr && last - bytenr > 32 * 1024))) {
1881                         skipped++;
1882                         continue;
1883                 }
1884                 blocksize = btrfs_level_size(root, level - 1);
1885                 if (i != slot) {
1886                         ret = lookup_extent_ref(NULL, root, bytenr,
1887                                                 blocksize, &refs);
1888                         BUG_ON(ret);
1889                         if (refs != 1) {
1890                                 skipped++;
1891                                 continue;
1892                         }
1893                 }
1894                 mutex_unlock(&root->fs_info->fs_mutex);
1895                 ret = readahead_tree_block(root, bytenr, blocksize);
1896                 last = bytenr + blocksize;
1897                 cond_resched();
1898                 mutex_lock(&root->fs_info->fs_mutex);
1899                 if (ret)
1900                         break;
1901         }
1902 }
1903
1904 /*
1905  * helper function for drop_snapshot, this walks down the tree dropping ref
1906  * counts as it goes.
1907  */
1908 static int noinline walk_down_tree(struct btrfs_trans_handle *trans,
1909                                    struct btrfs_root *root,
1910                                    struct btrfs_path *path, int *level)
1911 {
1912         u64 root_owner;
1913         u64 root_gen;
1914         u64 bytenr;
1915         struct extent_buffer *next;
1916         struct extent_buffer *cur;
1917         struct extent_buffer *parent;
1918         u32 blocksize;
1919         int ret;
1920         u32 refs;
1921
1922         WARN_ON(*level < 0);
1923         WARN_ON(*level >= BTRFS_MAX_LEVEL);
1924         ret = lookup_extent_ref(trans, root,
1925                                 path->nodes[*level]->start,
1926                                 path->nodes[*level]->len, &refs);
1927         BUG_ON(ret);
1928         if (refs > 1)
1929                 goto out;
1930
1931         /*
1932          * walk down to the last node level and free all the leaves
1933          */
1934         while(*level >= 0) {
1935                 WARN_ON(*level < 0);
1936                 WARN_ON(*level >= BTRFS_MAX_LEVEL);
1937                 cur = path->nodes[*level];
1938
1939                 if (btrfs_header_level(cur) != *level)
1940                         WARN_ON(1);
1941
1942                 if (path->slots[*level] >=
1943                     btrfs_header_nritems(cur))
1944                         break;
1945                 if (*level == 0) {
1946                         ret = drop_leaf_ref(trans, root, cur);
1947                         BUG_ON(ret);
1948                         break;
1949                 }
1950                 bytenr = btrfs_node_blockptr(cur, path->slots[*level]);
1951                 blocksize = btrfs_level_size(root, *level - 1);
1952                 ret = lookup_extent_ref(trans, root, bytenr, blocksize, &refs);
1953                 BUG_ON(ret);
1954                 if (refs != 1) {
1955                         parent = path->nodes[*level];
1956                         root_owner = btrfs_header_owner(parent);
1957                         root_gen = btrfs_header_generation(parent);
1958                         path->slots[*level]++;
1959                         ret = btrfs_free_extent(trans, root, bytenr,
1960                                                 blocksize, root_owner,
1961                                                 root_gen, 0, 0, 1);
1962                         BUG_ON(ret);
1963                         continue;
1964                 }
1965                 next = btrfs_find_tree_block(root, bytenr, blocksize);
1966                 if (!next || !btrfs_buffer_uptodate(next)) {
1967                         free_extent_buffer(next);
1968                         reada_walk_down(root, cur, path->slots[*level]);
1969                         mutex_unlock(&root->fs_info->fs_mutex);
1970                         next = read_tree_block(root, bytenr, blocksize);
1971                         mutex_lock(&root->fs_info->fs_mutex);
1972
1973                         /* we dropped the lock, check one more time */
1974                         ret = lookup_extent_ref(trans, root, bytenr,
1975                                                 blocksize, &refs);
1976                         BUG_ON(ret);
1977                         if (refs != 1) {
1978                                 parent = path->nodes[*level];
1979                                 root_owner = btrfs_header_owner(parent);
1980                                 root_gen = btrfs_header_generation(parent);
1981
1982                                 path->slots[*level]++;
1983                                 free_extent_buffer(next);
1984                                 ret = btrfs_free_extent(trans, root, bytenr,
1985                                                         blocksize,
1986                                                         root_owner,
1987                                                         root_gen, 0, 0, 1);
1988                                 BUG_ON(ret);
1989                                 continue;
1990                         }
1991                 }
1992                 WARN_ON(*level <= 0);
1993                 if (path->nodes[*level-1])
1994                         free_extent_buffer(path->nodes[*level-1]);
1995                 path->nodes[*level-1] = next;
1996                 *level = btrfs_header_level(next);
1997                 path->slots[*level] = 0;
1998         }
1999 out:
2000         WARN_ON(*level < 0);
2001         WARN_ON(*level >= BTRFS_MAX_LEVEL);
2002
2003         if (path->nodes[*level] == root->node) {
2004                 root_owner = root->root_key.objectid;
2005                 parent = path->nodes[*level];
2006         } else {
2007                 parent = path->nodes[*level + 1];
2008                 root_owner = btrfs_header_owner(parent);
2009         }
2010
2011         root_gen = btrfs_header_generation(parent);
2012         ret = btrfs_free_extent(trans, root, path->nodes[*level]->start,
2013                                 path->nodes[*level]->len,
2014                                 root_owner, root_gen, 0, 0, 1);
2015         free_extent_buffer(path->nodes[*level]);
2016         path->nodes[*level] = NULL;
2017         *level += 1;
2018         BUG_ON(ret);
2019         return 0;
2020 }
2021
2022 /*
2023  * helper for dropping snapshots.  This walks back up the tree in the path
2024  * to find the first node higher up where we haven't yet gone through
2025  * all the slots
2026  */
2027 static int noinline walk_up_tree(struct btrfs_trans_handle *trans,
2028                                  struct btrfs_root *root,
2029                                  struct btrfs_path *path, int *level)
2030 {
2031         u64 root_owner;
2032         u64 root_gen;
2033         struct btrfs_root_item *root_item = &root->root_item;
2034         int i;
2035         int slot;
2036         int ret;
2037
2038         for(i = *level; i < BTRFS_MAX_LEVEL - 1 && path->nodes[i]; i++) {
2039                 slot = path->slots[i];
2040                 if (slot < btrfs_header_nritems(path->nodes[i]) - 1) {
2041                         struct extent_buffer *node;
2042                         struct btrfs_disk_key disk_key;
2043                         node = path->nodes[i];
2044                         path->slots[i]++;
2045                         *level = i;
2046                         WARN_ON(*level == 0);
2047                         btrfs_node_key(node, &disk_key, path->slots[i]);
2048                         memcpy(&root_item->drop_progress,
2049                                &disk_key, sizeof(disk_key));
2050                         root_item->drop_level = i;
2051                         return 0;
2052                 } else {
2053                         if (path->nodes[*level] == root->node) {
2054                                 root_owner = root->root_key.objectid;
2055                                 root_gen =
2056                                    btrfs_header_generation(path->nodes[*level]);
2057                         } else {
2058                                 struct extent_buffer *node;
2059                                 node = path->nodes[*level + 1];
2060                                 root_owner = btrfs_header_owner(node);
2061                                 root_gen = btrfs_header_generation(node);
2062                         }
2063                         ret = btrfs_free_extent(trans, root,
2064                                                 path->nodes[*level]->start,
2065                                                 path->nodes[*level]->len,
2066                                                 root_owner, root_gen, 0, 0, 1);
2067                         BUG_ON(ret);
2068                         free_extent_buffer(path->nodes[*level]);
2069                         path->nodes[*level] = NULL;
2070                         *level = i + 1;
2071                 }
2072         }
2073         return 1;
2074 }
2075
2076 /*
2077  * drop the reference count on the tree rooted at 'snap'.  This traverses
2078  * the tree freeing any blocks that have a ref count of zero after being
2079  * decremented.
2080  */
2081 int btrfs_drop_snapshot(struct btrfs_trans_handle *trans, struct btrfs_root
2082                         *root)
2083 {
2084         int ret = 0;
2085         int wret;
2086         int level;
2087         struct btrfs_path *path;
2088         int i;
2089         int orig_level;
2090         struct btrfs_root_item *root_item = &root->root_item;
2091
2092         path = btrfs_alloc_path();
2093         BUG_ON(!path);
2094
2095         level = btrfs_header_level(root->node);
2096         orig_level = level;
2097         if (btrfs_disk_key_objectid(&root_item->drop_progress) == 0) {
2098                 path->nodes[level] = root->node;
2099                 extent_buffer_get(root->node);
2100                 path->slots[level] = 0;
2101         } else {
2102                 struct btrfs_key key;
2103                 struct btrfs_disk_key found_key;
2104                 struct extent_buffer *node;
2105
2106                 btrfs_disk_key_to_cpu(&key, &root_item->drop_progress);
2107                 level = root_item->drop_level;
2108                 path->lowest_level = level;
2109                 wret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
2110                 if (wret < 0) {
2111                         ret = wret;
2112                         goto out;
2113                 }
2114                 node = path->nodes[level];
2115                 btrfs_node_key(node, &found_key, path->slots[level]);
2116                 WARN_ON(memcmp(&found_key, &root_item->drop_progress,
2117                                sizeof(found_key)));
2118         }
2119         while(1) {
2120                 wret = walk_down_tree(trans, root, path, &level);
2121                 if (wret > 0)
2122                         break;
2123                 if (wret < 0)
2124                         ret = wret;
2125
2126                 wret = walk_up_tree(trans, root, path, &level);
2127                 if (wret > 0)
2128                         break;
2129                 if (wret < 0)
2130                         ret = wret;
2131                 ret = -EAGAIN;
2132                 break;
2133         }
2134         for (i = 0; i <= orig_level; i++) {
2135                 if (path->nodes[i]) {
2136                         free_extent_buffer(path->nodes[i]);
2137                         path->nodes[i] = NULL;
2138                 }
2139         }
2140 out:
2141         btrfs_free_path(path);
2142         return ret;
2143 }
2144
2145 int btrfs_free_block_groups(struct btrfs_fs_info *info)
2146 {
2147         u64 start;
2148         u64 end;
2149         u64 ptr;
2150         int ret;
2151         while(1) {
2152                 ret = find_first_extent_bit(&info->block_group_cache, 0,
2153                                             &start, &end, (unsigned int)-1);
2154                 if (ret)
2155                         break;
2156                 ret = get_state_private(&info->block_group_cache, start, &ptr);
2157                 if (!ret)
2158                         kfree((void *)(unsigned long)ptr);
2159                 clear_extent_bits(&info->block_group_cache, start,
2160                                   end, (unsigned int)-1, GFP_NOFS);
2161         }
2162         while(1) {
2163                 ret = find_first_extent_bit(&info->free_space_cache, 0,
2164                                             &start, &end, EXTENT_DIRTY);
2165                 if (ret)
2166                         break;
2167                 clear_extent_dirty(&info->free_space_cache, start,
2168                                    end, GFP_NOFS);
2169         }
2170         return 0;
2171 }
2172
2173 static int noinline relocate_inode_pages(struct inode *inode, u64 start,
2174                                          u64 len)
2175 {
2176         u64 page_start;
2177         u64 page_end;
2178         u64 delalloc_start;
2179         u64 existing_delalloc;
2180         unsigned long last_index;
2181         unsigned long i;
2182         struct page *page;
2183         struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
2184         struct file_ra_state *ra;
2185
2186         ra = kzalloc(sizeof(*ra), GFP_NOFS);
2187
2188         mutex_lock(&inode->i_mutex);
2189         i = start >> PAGE_CACHE_SHIFT;
2190         last_index = (start + len - 1) >> PAGE_CACHE_SHIFT;
2191
2192         file_ra_state_init(ra, inode->i_mapping);
2193         btrfs_force_ra(inode->i_mapping, ra, NULL, i, last_index);
2194         kfree(ra);
2195
2196         for (; i <= last_index; i++) {
2197                 page = grab_cache_page(inode->i_mapping, i);
2198                 if (!page)
2199                         goto out_unlock;
2200                 if (!PageUptodate(page)) {
2201                         btrfs_readpage(NULL, page);
2202                         lock_page(page);
2203                         if (!PageUptodate(page)) {
2204                                 unlock_page(page);
2205                                 page_cache_release(page);
2206                                 goto out_unlock;
2207                         }
2208                 }
2209                 page_start = (u64)page->index << PAGE_CACHE_SHIFT;
2210                 page_end = page_start + PAGE_CACHE_SIZE - 1;
2211
2212                 lock_extent(io_tree, page_start, page_end, GFP_NOFS);
2213
2214                 delalloc_start = page_start;
2215                 existing_delalloc = count_range_bits(io_tree,
2216                                              &delalloc_start, page_end,
2217                                              PAGE_CACHE_SIZE, EXTENT_DELALLOC);
2218
2219                 set_extent_delalloc(io_tree, page_start,
2220                                     page_end, GFP_NOFS);
2221
2222                 unlock_extent(io_tree, page_start, page_end, GFP_NOFS);
2223                 set_page_dirty(page);
2224                 unlock_page(page);
2225                 page_cache_release(page);
2226         }
2227
2228 out_unlock:
2229         mutex_unlock(&inode->i_mutex);
2230         return 0;
2231 }
2232
2233 /*
2234  * note, this releases the path
2235  */
2236 static int noinline relocate_one_reference(struct btrfs_root *extent_root,
2237                                   struct btrfs_path *path,
2238                                   struct btrfs_key *extent_key)
2239 {
2240         struct inode *inode;
2241         struct btrfs_root *found_root;
2242         struct btrfs_key *root_location;
2243         struct btrfs_extent_ref *ref;
2244         u64 ref_root;
2245         u64 ref_gen;
2246         u64 ref_objectid;
2247         u64 ref_offset;
2248         int ret;
2249
2250         ref = btrfs_item_ptr(path->nodes[0], path->slots[0],
2251                              struct btrfs_extent_ref);
2252         ref_root = btrfs_ref_root(path->nodes[0], ref);
2253         ref_gen = btrfs_ref_generation(path->nodes[0], ref);
2254         ref_objectid = btrfs_ref_objectid(path->nodes[0], ref);
2255         ref_offset = btrfs_ref_offset(path->nodes[0], ref);
2256         btrfs_release_path(extent_root, path);
2257
2258         root_location = kmalloc(sizeof(*root_location), GFP_NOFS);
2259         root_location->objectid = ref_root;
2260         if (ref_gen == 0)
2261                 root_location->offset = 0;
2262         else
2263                 root_location->offset = (u64)-1;
2264         root_location->type = BTRFS_ROOT_ITEM_KEY;
2265
2266         found_root = btrfs_read_fs_root_no_name(extent_root->fs_info,
2267                                                 root_location);
2268         BUG_ON(!found_root);
2269         kfree(root_location);
2270
2271         if (ref_objectid >= BTRFS_FIRST_FREE_OBJECTID) {
2272                 mutex_unlock(&extent_root->fs_info->fs_mutex);
2273                 inode = btrfs_iget_locked(extent_root->fs_info->sb,
2274                                           ref_objectid, found_root);
2275                 if (inode->i_state & I_NEW) {
2276                         /* the inode and parent dir are two different roots */
2277                         BTRFS_I(inode)->root = found_root;
2278                         BTRFS_I(inode)->location.objectid = ref_objectid;
2279                         BTRFS_I(inode)->location.type = BTRFS_INODE_ITEM_KEY;
2280                         BTRFS_I(inode)->location.offset = 0;
2281                         btrfs_read_locked_inode(inode);
2282                         unlock_new_inode(inode);
2283
2284                 }
2285                 /* this can happen if the reference is not against
2286                  * the latest version of the tree root
2287                  */
2288                 if (is_bad_inode(inode)) {
2289                         mutex_lock(&extent_root->fs_info->fs_mutex);
2290                         goto out;
2291                 }
2292                 relocate_inode_pages(inode, ref_offset, extent_key->offset);
2293                 /* FIXME, data=ordered will help get rid of this */
2294                 filemap_fdatawrite(inode->i_mapping);
2295                 iput(inode);
2296                 mutex_lock(&extent_root->fs_info->fs_mutex);
2297         } else {
2298                 struct btrfs_trans_handle *trans;
2299                 struct btrfs_key found_key;
2300                 struct extent_buffer *eb;
2301                 int level;
2302                 int i;
2303
2304                 trans = btrfs_start_transaction(found_root, 1);
2305                 eb = read_tree_block(found_root, extent_key->objectid,
2306                                      extent_key->offset);
2307                 level = btrfs_header_level(eb);
2308
2309                 if (level == 0)
2310                         btrfs_item_key_to_cpu(eb, &found_key, 0);
2311                 else
2312                         btrfs_node_key_to_cpu(eb, &found_key, 0);
2313
2314                 free_extent_buffer(eb);
2315
2316                 path->lowest_level = level;
2317                 path->reada = 2;
2318                 ret = btrfs_search_slot(trans, found_root, &found_key, path,
2319                                         0, 1);
2320                 path->lowest_level = 0;
2321                 for (i = level; i < BTRFS_MAX_LEVEL; i++) {
2322                         if (!path->nodes[i])
2323                                 break;
2324                         free_extent_buffer(path->nodes[i]);
2325                         path->nodes[i] = NULL;
2326                 }
2327                 btrfs_release_path(found_root, path);
2328                 btrfs_end_transaction(trans, found_root);
2329         }
2330
2331 out:
2332         return 0;
2333 }
2334
2335 static int noinline relocate_one_extent(struct btrfs_root *extent_root,
2336                                         struct btrfs_path *path,
2337                                         struct btrfs_key *extent_key)
2338 {
2339         struct btrfs_key key;
2340         struct btrfs_key found_key;
2341         struct extent_buffer *leaf;
2342         u32 nritems;
2343         u32 item_size;
2344         int ret = 0;
2345
2346         key.objectid = extent_key->objectid;
2347         key.type = BTRFS_EXTENT_REF_KEY;
2348         key.offset = 0;
2349
2350         while(1) {
2351                 ret = btrfs_search_slot(NULL, extent_root, &key, path, 0, 0);
2352
2353                 if (ret < 0)
2354                         goto out;
2355
2356                 ret = 0;
2357                 leaf = path->nodes[0];
2358                 nritems = btrfs_header_nritems(leaf);
2359                 if (path->slots[0] == nritems)
2360                         goto out;
2361
2362                 btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
2363                 if (found_key.objectid != extent_key->objectid)
2364                         break;
2365
2366                 if (found_key.type != BTRFS_EXTENT_REF_KEY)
2367                         break;
2368
2369                 key.offset = found_key.offset + 1;
2370                 item_size = btrfs_item_size_nr(leaf, path->slots[0]);
2371
2372                 ret = relocate_one_reference(extent_root, path, extent_key);
2373                 if (ret)
2374                         goto out;
2375         }
2376         ret = 0;
2377 out:
2378         btrfs_release_path(extent_root, path);
2379         return ret;
2380 }
2381
2382 int btrfs_shrink_extent_tree(struct btrfs_root *root, u64 new_size)
2383 {
2384         struct btrfs_trans_handle *trans;
2385         struct btrfs_root *tree_root = root->fs_info->tree_root;
2386         struct btrfs_path *path;
2387         u64 cur_byte;
2388         u64 total_found;
2389         struct btrfs_fs_info *info = root->fs_info;
2390         struct extent_io_tree *block_group_cache;
2391         struct btrfs_key key;
2392         struct btrfs_key found_key;
2393         struct extent_buffer *leaf;
2394         u32 nritems;
2395         int ret;
2396         int progress = 0;
2397
2398         btrfs_set_super_total_bytes(&info->super_copy, new_size);
2399         clear_extent_dirty(&info->free_space_cache, new_size, (u64)-1,
2400                            GFP_NOFS);
2401         block_group_cache = &info->block_group_cache;
2402         path = btrfs_alloc_path();
2403         root = root->fs_info->extent_root;
2404         path->reada = 2;
2405
2406 again:
2407         total_found = 0;
2408         key.objectid = new_size;
2409         key.offset = 0;
2410         key.type = 0;
2411         cur_byte = key.objectid;
2412
2413         ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
2414         if (ret < 0)
2415                 goto out;
2416
2417         ret = find_previous_extent(root, path);
2418         if (ret < 0)
2419                 goto out;
2420         if (ret == 0) {
2421                 leaf = path->nodes[0];
2422                 btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
2423                 if (found_key.objectid + found_key.offset > new_size) {
2424                         cur_byte = found_key.objectid;
2425                         key.objectid = cur_byte;
2426                 }
2427         }
2428         btrfs_release_path(root, path);
2429
2430         while(1) {
2431                 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
2432                 if (ret < 0)
2433                         goto out;
2434
2435                 leaf = path->nodes[0];
2436                 nritems = btrfs_header_nritems(leaf);
2437 next:
2438                 if (path->slots[0] >= nritems) {
2439                         ret = btrfs_next_leaf(root, path);
2440                         if (ret < 0)
2441                                 goto out;
2442                         if (ret == 1) {
2443                                 ret = 0;
2444                                 break;
2445                         }
2446                         leaf = path->nodes[0];
2447                         nritems = btrfs_header_nritems(leaf);
2448                 }
2449
2450                 btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
2451
2452                 if (progress && need_resched()) {
2453                         memcpy(&key, &found_key, sizeof(key));
2454                         mutex_unlock(&root->fs_info->fs_mutex);
2455                         cond_resched();
2456                         mutex_lock(&root->fs_info->fs_mutex);
2457                         btrfs_release_path(root, path);
2458                         btrfs_search_slot(NULL, root, &key, path, 0, 0);
2459                         progress = 0;
2460                         goto next;
2461                 }
2462                 progress = 1;
2463
2464                 if (btrfs_key_type(&found_key) != BTRFS_EXTENT_ITEM_KEY ||
2465                     found_key.objectid + found_key.offset <= cur_byte) {
2466                         path->slots[0]++;
2467                         goto next;
2468                 }
2469
2470                 total_found++;
2471                 cur_byte = found_key.objectid + found_key.offset;
2472                 key.objectid = cur_byte;
2473                 btrfs_release_path(root, path);
2474                 ret = relocate_one_extent(root, path, &found_key);
2475         }
2476
2477         btrfs_release_path(root, path);
2478
2479         if (total_found > 0) {
2480                 trans = btrfs_start_transaction(tree_root, 1);
2481                 btrfs_commit_transaction(trans, tree_root);
2482
2483                 mutex_unlock(&root->fs_info->fs_mutex);
2484                 btrfs_clean_old_snapshots(tree_root);
2485                 mutex_lock(&root->fs_info->fs_mutex);
2486
2487                 trans = btrfs_start_transaction(tree_root, 1);
2488                 btrfs_commit_transaction(trans, tree_root);
2489                 goto again;
2490         }
2491
2492         trans = btrfs_start_transaction(root, 1);
2493         key.objectid = new_size;
2494         key.offset = 0;
2495         key.type = 0;
2496         while(1) {
2497                 u64 ptr;
2498
2499                 ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
2500                 if (ret < 0)
2501                         goto out;
2502
2503                 leaf = path->nodes[0];
2504                 nritems = btrfs_header_nritems(leaf);
2505 bg_next:
2506                 if (path->slots[0] >= nritems) {
2507                         ret = btrfs_next_leaf(root, path);
2508                         if (ret < 0)
2509                                 break;
2510                         if (ret == 1) {
2511                                 ret = 0;
2512                                 break;
2513                         }
2514                         leaf = path->nodes[0];
2515                         btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
2516
2517                         /*
2518                          * btrfs_next_leaf doesn't cow buffers, we have to
2519                          * do the search again
2520                          */
2521                         memcpy(&key, &found_key, sizeof(key));
2522                         btrfs_release_path(root, path);
2523                         goto resched_check;
2524                 }
2525
2526                 btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
2527                 if (btrfs_key_type(&found_key) != BTRFS_BLOCK_GROUP_ITEM_KEY) {
2528                         printk("shrinker found key %Lu %u %Lu\n",
2529                                 found_key.objectid, found_key.type,
2530                                 found_key.offset);
2531                         path->slots[0]++;
2532                         goto bg_next;
2533                 }
2534                 ret = get_state_private(&info->block_group_cache,
2535                                         found_key.objectid, &ptr);
2536                 if (!ret)
2537                         kfree((void *)(unsigned long)ptr);
2538
2539                 clear_extent_bits(&info->block_group_cache, found_key.objectid,
2540                                   found_key.objectid + found_key.offset - 1,
2541                                   (unsigned int)-1, GFP_NOFS);
2542
2543                 key.objectid = found_key.objectid + 1;
2544                 btrfs_del_item(trans, root, path);
2545                 btrfs_release_path(root, path);
2546 resched_check:
2547                 if (need_resched()) {
2548                         mutex_unlock(&root->fs_info->fs_mutex);
2549                         cond_resched();
2550                         mutex_lock(&root->fs_info->fs_mutex);
2551                 }
2552         }
2553         clear_extent_dirty(&info->free_space_cache, new_size, (u64)-1,
2554                            GFP_NOFS);
2555         btrfs_commit_transaction(trans, root);
2556 out:
2557         btrfs_free_path(path);
2558         return ret;
2559 }
2560
2561 int btrfs_grow_extent_tree(struct btrfs_trans_handle *trans,
2562                            struct btrfs_root *root, u64 new_size)
2563 {
2564         struct btrfs_path *path;
2565         u64 nr = 0;
2566         u64 cur_byte;
2567         u64 old_size;
2568         unsigned long rem;
2569         struct btrfs_block_group_cache *cache;
2570         struct btrfs_block_group_item *item;
2571         struct btrfs_fs_info *info = root->fs_info;
2572         struct extent_io_tree *block_group_cache;
2573         struct btrfs_key key;
2574         struct extent_buffer *leaf;
2575         int ret;
2576         int bit;
2577
2578         old_size = btrfs_super_total_bytes(&info->super_copy);
2579         block_group_cache = &info->block_group_cache;
2580
2581         root = info->extent_root;
2582
2583         cache = btrfs_lookup_block_group(root->fs_info, old_size - 1);
2584
2585         cur_byte = cache->key.objectid + cache->key.offset;
2586         if (cur_byte >= new_size)
2587                 goto set_size;
2588
2589         key.offset = BTRFS_BLOCK_GROUP_SIZE;
2590         btrfs_set_key_type(&key, BTRFS_BLOCK_GROUP_ITEM_KEY);
2591
2592         path = btrfs_alloc_path();
2593         if (!path)
2594                 return -ENOMEM;
2595
2596         while(cur_byte < new_size) {
2597                 key.objectid = cur_byte;
2598                 ret = btrfs_insert_empty_item(trans, root, path, &key,
2599                                         sizeof(struct btrfs_block_group_item));
2600                 BUG_ON(ret);
2601                 leaf = path->nodes[0];
2602                 item = btrfs_item_ptr(leaf, path->slots[0],
2603                                       struct btrfs_block_group_item);
2604
2605                 btrfs_set_disk_block_group_used(leaf, item, 0);
2606                 div_long_long_rem(nr, 3, &rem);
2607                 if (rem) {
2608                         btrfs_set_disk_block_group_flags(leaf, item,
2609                                                  BTRFS_BLOCK_GROUP_DATA);
2610                 } else {
2611                         btrfs_set_disk_block_group_flags(leaf, item, 0);
2612                 }
2613                 nr++;
2614
2615                 cache = kmalloc(sizeof(*cache), GFP_NOFS);
2616                 BUG_ON(!cache);
2617
2618                 read_extent_buffer(leaf, &cache->item, (unsigned long)item,
2619                                    sizeof(cache->item));
2620
2621                 memcpy(&cache->key, &key, sizeof(key));
2622                 cache->cached = 0;
2623                 cache->pinned = 0;
2624                 cur_byte = key.objectid + key.offset;
2625                 btrfs_release_path(root, path);
2626
2627                 if (cache->item.flags & BTRFS_BLOCK_GROUP_DATA) {
2628                         bit = BLOCK_GROUP_DATA;
2629                         cache->data = BTRFS_BLOCK_GROUP_DATA;
2630                 } else {
2631                         bit = BLOCK_GROUP_METADATA;
2632                         cache->data = 0;
2633                 }
2634
2635                 /* use EXTENT_LOCKED to prevent merging */
2636                 set_extent_bits(block_group_cache, key.objectid,
2637                                 key.objectid + key.offset - 1,
2638                                 bit | EXTENT_LOCKED, GFP_NOFS);
2639                 set_state_private(block_group_cache, key.objectid,
2640                                   (unsigned long)cache);
2641         }
2642         btrfs_free_path(path);
2643 set_size:
2644         btrfs_set_super_total_bytes(&info->super_copy, new_size);
2645         return 0;
2646 }
2647
2648 int btrfs_read_block_groups(struct btrfs_root *root)
2649 {
2650         struct btrfs_path *path;
2651         int ret;
2652         int err = 0;
2653         int bit;
2654         struct btrfs_block_group_cache *cache;
2655         struct btrfs_fs_info *info = root->fs_info;
2656         struct extent_io_tree *block_group_cache;
2657         struct btrfs_key key;
2658         struct btrfs_key found_key;
2659         struct extent_buffer *leaf;
2660
2661         block_group_cache = &info->block_group_cache;
2662
2663         root = info->extent_root;
2664         key.objectid = 0;
2665         key.offset = BTRFS_BLOCK_GROUP_SIZE;
2666         btrfs_set_key_type(&key, BTRFS_BLOCK_GROUP_ITEM_KEY);
2667
2668         path = btrfs_alloc_path();
2669         if (!path)
2670                 return -ENOMEM;
2671
2672         while(1) {
2673                 ret = btrfs_search_slot(NULL, info->extent_root,
2674                                         &key, path, 0, 0);
2675                 if (ret != 0) {
2676                         err = ret;
2677                         break;
2678                 }
2679                 leaf = path->nodes[0];
2680                 btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
2681                 cache = kmalloc(sizeof(*cache), GFP_NOFS);
2682                 if (!cache) {
2683                         err = -1;
2684                         break;
2685                 }
2686
2687                 read_extent_buffer(leaf, &cache->item,
2688                                    btrfs_item_ptr_offset(leaf, path->slots[0]),
2689                                    sizeof(cache->item));
2690                 memcpy(&cache->key, &found_key, sizeof(found_key));
2691                 cache->cached = 0;
2692                 cache->pinned = 0;
2693                 key.objectid = found_key.objectid + found_key.offset;
2694                 btrfs_release_path(root, path);
2695
2696                 if (cache->item.flags & BTRFS_BLOCK_GROUP_MIXED) {
2697                         bit = BLOCK_GROUP_DATA | BLOCK_GROUP_METADATA;
2698                         cache->data = BTRFS_BLOCK_GROUP_MIXED;
2699                 } else if (cache->item.flags & BTRFS_BLOCK_GROUP_DATA) {
2700                         bit = BLOCK_GROUP_DATA;
2701                         cache->data = BTRFS_BLOCK_GROUP_DATA;
2702                 } else {
2703                         bit = BLOCK_GROUP_METADATA;
2704                         cache->data = 0;
2705                 }
2706
2707                 /* use EXTENT_LOCKED to prevent merging */
2708                 set_extent_bits(block_group_cache, found_key.objectid,
2709                                 found_key.objectid + found_key.offset - 1,
2710                                 bit | EXTENT_LOCKED, GFP_NOFS);
2711                 set_state_private(block_group_cache, found_key.objectid,
2712                                   (unsigned long)cache);
2713
2714                 if (key.objectid >=
2715                     btrfs_super_total_bytes(&info->super_copy))
2716                         break;
2717         }
2718
2719         btrfs_free_path(path);
2720         return 0;
2721 }