fs/btrfs/ctree.c

   1 // SPDX-License-Identifier: GPL-2.0
   2 /*
   3  * Copyright (C) 2007,2008 Oracle.  All rights reserved.
   4  */
   5
   6 #include <linux/sched.h>
   7 #include <linux/slab.h>
   8 #include <linux/rbtree.h>
   9 #include <linux/mm.h>
  10 #include <linux/error-injection.h>
  11 #include "ctree.h"
  12 #include "disk-io.h"
  13 #include "transaction.h"
  14 #include "print-tree.h"
  15 #include "locking.h"
  16 #include "volumes.h"
  17 #include "qgroup.h"
  18 #include "tree-mod-log.h"
  19
  20 static int split_node(struct btrfs_trans_handle *trans, struct btrfs_root
  21                       *root, struct btrfs_path *path, int level);
  22 static int split_leaf(struct btrfs_trans_handle *trans, struct btrfs_root *root,
  23                       const struct btrfs_key *ins_key, struct btrfs_path *path,
  24                       int data_size, int extend);
  25 static int push_node_left(struct btrfs_trans_handle *trans,
  26                           struct extent_buffer *dst,
  27                           struct extent_buffer *src, int empty);
  28 static int balance_node_right(struct btrfs_trans_handle *trans,
  29                               struct extent_buffer *dst_buf,
  30                               struct extent_buffer *src_buf);
  31 static void del_ptr(struct btrfs_root *root, struct btrfs_path *path,
  32                     int level, int slot);
  33
  34 static const struct btrfs_csums {
  35         u16             size;
  36         const char      name[10];
  37         const char      driver[12];
  38 } btrfs_csums[] = {
  39         [BTRFS_CSUM_TYPE_CRC32] = { .size = 4, .name = "crc32c" },
  40         [BTRFS_CSUM_TYPE_XXHASH] = { .size = 8, .name = "xxhash64" },
  41         [BTRFS_CSUM_TYPE_SHA256] = { .size = 32, .name = "sha256" },
  42         [BTRFS_CSUM_TYPE_BLAKE2] = { .size = 32, .name = "blake2b",
  43                                      .driver = "blake2b-256" },
  44 };
  45
  46 int btrfs_super_csum_size(const struct btrfs_super_block *s)
  47 {
  48         u16 t = btrfs_super_csum_type(s);
  49         /*
  50          * csum type is validated at mount time
  51          */
  52         return btrfs_csums[t].size;
  53 }
  54
  55 const char *btrfs_super_csum_name(u16 csum_type)
  56 {
  57         /* csum type is validated at mount time */
  58         return btrfs_csums[csum_type].name;
  59 }
  60
  61 /*
  62  * Return driver name if defined, otherwise the name that's also a valid driver
  63  * name
  64  */
  65 const char *btrfs_super_csum_driver(u16 csum_type)
  66 {
  67         /* csum type is validated at mount time */
  68         return btrfs_csums[csum_type].driver[0] ?
  69                 btrfs_csums[csum_type].driver :
  70                 btrfs_csums[csum_type].name;
  71 }
  72
  73 size_t __attribute_const__ btrfs_get_num_csums(void)
  74 {
  75         return ARRAY_SIZE(btrfs_csums);
  76 }
  77
  78 struct btrfs_path *btrfs_alloc_path(void)
  79 {
  80         return kmem_cache_zalloc(btrfs_path_cachep, GFP_NOFS);
  81 }
  82
  83 /* this also releases the path */
  84 void btrfs_free_path(struct btrfs_path *p)
  85 {
  86         if (!p)
  87                 return;
  88         btrfs_release_path(p);
  89         kmem_cache_free(btrfs_path_cachep, p);
  90 }
  91
  92 /*
  93  * path release drops references on the extent buffers in the path
  94  * and it drops any locks held by this path
  95  *
  96  * It is safe to call this on paths that no locks or extent buffers held.
  97  */
  98 noinline void btrfs_release_path(struct btrfs_path *p)
  99 {
 100         int i;
 101
 102         for (i = 0; i < BTRFS_MAX_LEVEL; i++) {
 103                 p->slots[i] = 0;
 104                 if (!p->nodes[i])
 105                         continue;
 106                 if (p->locks[i]) {
 107                         btrfs_tree_unlock_rw(p->nodes[i], p->locks[i]);
 108                         p->locks[i] = 0;
 109                 }
 110                 free_extent_buffer(p->nodes[i]);
 111                 p->nodes[i] = NULL;
 112         }
 113 }
 114
 115 /*
 116  * safely gets a reference on the root node of a tree.  A lock
 117  * is not taken, so a concurrent writer may put a different node
 118  * at the root of the tree.  See btrfs_lock_root_node for the
 119  * looping required.
 120  *
 121  * The extent buffer returned by this has a reference taken, so
 122  * it won't disappear.  It may stop being the root of the tree
 123  * at any time because there are no locks held.
 124  */
 125 struct extent_buffer *btrfs_root_node(struct btrfs_root *root)
 126 {
 127         struct extent_buffer *eb;
 128
 129         while (1) {
 130                 rcu_read_lock();
 131                 eb = rcu_dereference(root->node);
 132
 133                 /*
 134                  * RCU really hurts here, we could free up the root node because
 135                  * it was COWed but we may not get the new root node yet so do
 136                  * the inc_not_zero dance and if it doesn't work then
 137                  * synchronize_rcu and try again.
 138                  */
 139                 if (atomic_inc_not_zero(&eb->refs)) {
 140                         rcu_read_unlock();
 141                         break;
 142                 }
 143                 rcu_read_unlock();
 144                 synchronize_rcu();
 145         }
 146         return eb;
 147 }
 148
 149 /*
 150  * Cowonly root (not-shareable trees, everything not subvolume or reloc roots),
 151  * just get put onto a simple dirty list.  Transaction walks this list to make
 152  * sure they get properly updated on disk.
 153  */
 154 static void add_root_to_dirty_list(struct btrfs_root *root)
 155 {
 156         struct btrfs_fs_info *fs_info = root->fs_info;
 157
 158         if (test_bit(BTRFS_ROOT_DIRTY, &root->state) ||
 159             !test_bit(BTRFS_ROOT_TRACK_DIRTY, &root->state))
 160                 return;
 161
 162         spin_lock(&fs_info->trans_lock);
 163         if (!test_and_set_bit(BTRFS_ROOT_DIRTY, &root->state)) {
 164                 /* Want the extent tree to be the last on the list */
 165                 if (root->root_key.objectid == BTRFS_EXTENT_TREE_OBJECTID)
 166                         list_move_tail(&root->dirty_list,
 167                                        &fs_info->dirty_cowonly_roots);
 168                 else
 169                         list_move(&root->dirty_list,
 170                                   &fs_info->dirty_cowonly_roots);
 171         }
 172         spin_unlock(&fs_info->trans_lock);
 173 }
 174
 175 /*
 176  * used by snapshot creation to make a copy of a root for a tree with
 177  * a given objectid.  The buffer with the new root node is returned in
 178  * cow_ret, and this func returns zero on success or a negative error code.
 179  */
 180 int btrfs_copy_root(struct btrfs_trans_handle *trans,
 181                       struct btrfs_root *root,
 182                       struct extent_buffer *buf,
 183                       struct extent_buffer **cow_ret, u64 new_root_objectid)
 184 {
 185         struct btrfs_fs_info *fs_info = root->fs_info;
 186         struct extent_buffer *cow;
 187         int ret = 0;
 188         int level;
 189         struct btrfs_disk_key disk_key;
 190
 191         WARN_ON(test_bit(BTRFS_ROOT_SHAREABLE, &root->state) &&
 192                 trans->transid != fs_info->running_transaction->transid);
 193         WARN_ON(test_bit(BTRFS_ROOT_SHAREABLE, &root->state) &&
 194                 trans->transid != root->last_trans);
 195
 196         level = btrfs_header_level(buf);
 197         if (level == 0)
 198                 btrfs_item_key(buf, &disk_key, 0);
 199         else
 200                 btrfs_node_key(buf, &disk_key, 0);
 201
 202         cow = btrfs_alloc_tree_block(trans, root, 0, new_root_objectid,
 203                                      &disk_key, level, buf->start, 0,
 204                                      BTRFS_NESTING_NEW_ROOT);
 205         if (IS_ERR(cow))
 206                 return PTR_ERR(cow);
 207
 208         copy_extent_buffer_full(cow, buf);
 209         btrfs_set_header_bytenr(cow, cow->start);
 210         btrfs_set_header_generation(cow, trans->transid);
 211         btrfs_set_header_backref_rev(cow, BTRFS_MIXED_BACKREF_REV);
 212         btrfs_clear_header_flag(cow, BTRFS_HEADER_FLAG_WRITTEN |
 213                                      BTRFS_HEADER_FLAG_RELOC);
 214         if (new_root_objectid == BTRFS_TREE_RELOC_OBJECTID)
 215                 btrfs_set_header_flag(cow, BTRFS_HEADER_FLAG_RELOC);
 216         else
 217                 btrfs_set_header_owner(cow, new_root_objectid);
 218
 219         write_extent_buffer_fsid(cow, fs_info->fs_devices->metadata_uuid);
 220
 221         WARN_ON(btrfs_header_generation(buf) > trans->transid);
 222         if (new_root_objectid == BTRFS_TREE_RELOC_OBJECTID)
 223                 ret = btrfs_inc_ref(trans, root, cow, 1);
 224         else
 225                 ret = btrfs_inc_ref(trans, root, cow, 0);
 226         if (ret) {
 227                 btrfs_tree_unlock(cow);
 228                 free_extent_buffer(cow);
 229                 btrfs_abort_transaction(trans, ret);
 230                 return ret;
 231         }
 232
 233         btrfs_mark_buffer_dirty(cow);
 234         *cow_ret = cow;
 235         return 0;
 236 }
 237
 238 /*
 239  * check if the tree block can be shared by multiple trees
 240  */
 241 int btrfs_block_can_be_shared(struct btrfs_root *root,
 242                               struct extent_buffer *buf)
 243 {
 244         /*
 245          * Tree blocks not in shareable trees and tree roots are never shared.
 246          * If a block was allocated after the last snapshot and the block was
 247          * not allocated by tree relocation, we know the block is not shared.
 248          */
 249         if (test_bit(BTRFS_ROOT_SHAREABLE, &root->state) &&
 250             buf != root->node && buf != root->commit_root &&
 251             (btrfs_header_generation(buf) <=
 252              btrfs_root_last_snapshot(&root->root_item) ||
 253              btrfs_header_flag(buf, BTRFS_HEADER_FLAG_RELOC)))
 254                 return 1;
 255
 256         return 0;
 257 }
 258
 259 static noinline int update_ref_for_cow(struct btrfs_trans_handle *trans,
 260                                        struct btrfs_root *root,
 261                                        struct extent_buffer *buf,
 262                                        struct extent_buffer *cow,
 263                                        int *last_ref)
 264 {
 265         struct btrfs_fs_info *fs_info = root->fs_info;
 266         u64 refs;
 267         u64 owner;
 268         u64 flags;
 269         u64 new_flags = 0;
 270         int ret;
 271
 272         /*
 273          * Backrefs update rules:
 274          *
 275          * Always use full backrefs for extent pointers in tree block
 276          * allocated by tree relocation.
 277          *
 278          * If a shared tree block is no longer referenced by its owner
 279          * tree (btrfs_header_owner(buf) == root->root_key.objectid),
 280          * use full backrefs for extent pointers in tree block.
 281          *
 282          * If a tree block is been relocating
 283          * (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID),
 284          * use full backrefs for extent pointers in tree block.
 285          * The reason for this is some operations (such as drop tree)
 286          * are only allowed for blocks use full backrefs.
 287          */
 288
 289         if (btrfs_block_can_be_shared(root, buf)) {
 290                 ret = btrfs_lookup_extent_info(trans, fs_info, buf->start,
 291                                                btrfs_header_level(buf), 1,
 292                                                &refs, &flags);
 293                 if (ret)
 294                         return ret;
 295                 if (refs == 0) {
 296                         ret = -EROFS;
 297                         btrfs_handle_fs_error(fs_info, ret, NULL);
 298                         return ret;
 299                 }
 300         } else {
 301                 refs = 1;
 302                 if (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID ||
 303                     btrfs_header_backref_rev(buf) < BTRFS_MIXED_BACKREF_REV)
 304                         flags = BTRFS_BLOCK_FLAG_FULL_BACKREF;
 305                 else
 306                         flags = 0;
 307         }
 308
 309         owner = btrfs_header_owner(buf);
 310         BUG_ON(owner == BTRFS_TREE_RELOC_OBJECTID &&
 311                !(flags & BTRFS_BLOCK_FLAG_FULL_BACKREF));
 312
 313         if (refs > 1) {
 314                 if ((owner == root->root_key.objectid ||
 315                      root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID) &&
 316                     !(flags & BTRFS_BLOCK_FLAG_FULL_BACKREF)) {
 317                         ret = btrfs_inc_ref(trans, root, buf, 1);
 318                         if (ret)
 319                                 return ret;
 320
 321                         if (root->root_key.objectid ==
 322                             BTRFS_TREE_RELOC_OBJECTID) {
 323                                 ret = btrfs_dec_ref(trans, root, buf, 0);
 324                                 if (ret)
 325                                         return ret;
 326                                 ret = btrfs_inc_ref(trans, root, cow, 1);
 327                                 if (ret)
 328                                         return ret;
 329                         }
 330                         new_flags |= BTRFS_BLOCK_FLAG_FULL_BACKREF;
 331                 } else {
 332
 333                         if (root->root_key.objectid ==
 334                             BTRFS_TREE_RELOC_OBJECTID)
 335                                 ret = btrfs_inc_ref(trans, root, cow, 1);
 336                         else
 337                                 ret = btrfs_inc_ref(trans, root, cow, 0);
 338                         if (ret)
 339                                 return ret;
 340                 }
 341                 if (new_flags != 0) {
 342                         int level = btrfs_header_level(buf);
 343
 344                         ret = btrfs_set_disk_extent_flags(trans, buf,
 345                                                           new_flags, level, 0);
 346                         if (ret)
 347                                 return ret;
 348                 }
 349         } else {
 350                 if (flags & BTRFS_BLOCK_FLAG_FULL_BACKREF) {
 351                         if (root->root_key.objectid ==
 352                             BTRFS_TREE_RELOC_OBJECTID)
 353                                 ret = btrfs_inc_ref(trans, root, cow, 1);
 354                         else
 355                                 ret = btrfs_inc_ref(trans, root, cow, 0);
 356                         if (ret)
 357                                 return ret;
 358                         ret = btrfs_dec_ref(trans, root, buf, 1);
 359                         if (ret)
 360                                 return ret;
 361                 }
 362                 btrfs_clean_tree_block(buf);
 363                 *last_ref = 1;
 364         }
 365         return 0;
 366 }
 367
 368 /*
 369  * does the dirty work in cow of a single block.  The parent block (if
 370  * supplied) is updated to point to the new cow copy.  The new buffer is marked
 371  * dirty and returned locked.  If you modify the block it needs to be marked
 372  * dirty again.
 373  *
 374  * search_start -- an allocation hint for the new block
 375  *
 376  * empty_size -- a hint that you plan on doing more cow.  This is the size in
 377  * bytes the allocator should try to find free next to the block it returns.
 378  * This is just a hint and may be ignored by the allocator.
 379  */
 380 static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans,
 381                              struct btrfs_root *root,
 382                              struct extent_buffer *buf,
 383                              struct extent_buffer *parent, int parent_slot,
 384                              struct extent_buffer **cow_ret,
 385                              u64 search_start, u64 empty_size,
 386                              enum btrfs_lock_nesting nest)
 387 {
 388         struct btrfs_fs_info *fs_info = root->fs_info;
 389         struct btrfs_disk_key disk_key;
 390         struct extent_buffer *cow;
 391         int level, ret;
 392         int last_ref = 0;
 393         int unlock_orig = 0;
 394         u64 parent_start = 0;
 395
 396         if (*cow_ret == buf)
 397                 unlock_orig = 1;
 398
 399         btrfs_assert_tree_write_locked(buf);
 400
 401         WARN_ON(test_bit(BTRFS_ROOT_SHAREABLE, &root->state) &&
 402                 trans->transid != fs_info->running_transaction->transid);
 403         WARN_ON(test_bit(BTRFS_ROOT_SHAREABLE, &root->state) &&
 404                 trans->transid != root->last_trans);
 405
 406         level = btrfs_header_level(buf);
 407
 408         if (level == 0)
 409                 btrfs_item_key(buf, &disk_key, 0);
 410         else
 411                 btrfs_node_key(buf, &disk_key, 0);
 412
 413         if ((root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID) && parent)
 414                 parent_start = parent->start;
 415
 416         cow = btrfs_alloc_tree_block(trans, root, parent_start,
 417                                      root->root_key.objectid, &disk_key, level,
 418                                      search_start, empty_size, nest);
 419         if (IS_ERR(cow))
 420                 return PTR_ERR(cow);
 421
 422         /* cow is set to blocking by btrfs_init_new_buffer */
 423
 424         copy_extent_buffer_full(cow, buf);
 425         btrfs_set_header_bytenr(cow, cow->start);
 426         btrfs_set_header_generation(cow, trans->transid);
 427         btrfs_set_header_backref_rev(cow, BTRFS_MIXED_BACKREF_REV);
 428         btrfs_clear_header_flag(cow, BTRFS_HEADER_FLAG_WRITTEN |
 429                                      BTRFS_HEADER_FLAG_RELOC);
 430         if (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID)
 431                 btrfs_set_header_flag(cow, BTRFS_HEADER_FLAG_RELOC);
 432         else
 433                 btrfs_set_header_owner(cow, root->root_key.objectid);
 434
 435         write_extent_buffer_fsid(cow, fs_info->fs_devices->metadata_uuid);
 436
 437         ret = update_ref_for_cow(trans, root, buf, cow, &last_ref);
 438         if (ret) {
 439                 btrfs_tree_unlock(cow);
 440                 free_extent_buffer(cow);
 441                 btrfs_abort_transaction(trans, ret);
 442                 return ret;
 443         }
 444
 445         if (test_bit(BTRFS_ROOT_SHAREABLE, &root->state)) {
 446                 ret = btrfs_reloc_cow_block(trans, root, buf, cow);
 447                 if (ret) {
 448                         btrfs_tree_unlock(cow);
 449                         free_extent_buffer(cow);
 450                         btrfs_abort_transaction(trans, ret);
 451                         return ret;
 452                 }
 453         }
 454
 455         if (buf == root->node) {
 456                 WARN_ON(parent && parent != buf);
 457                 if (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID ||
 458                     btrfs_header_backref_rev(buf) < BTRFS_MIXED_BACKREF_REV)
 459                         parent_start = buf->start;
 460
 461                 atomic_inc(&cow->refs);
 462                 ret = btrfs_tree_mod_log_insert_root(root->node, cow, true);
 463                 BUG_ON(ret < 0);
 464                 rcu_assign_pointer(root->node, cow);
 465
 466                 btrfs_free_tree_block(trans, root, buf, parent_start,
 467                                       last_ref);
 468                 free_extent_buffer(buf);
 469                 add_root_to_dirty_list(root);
 470         } else {
 471                 WARN_ON(trans->transid != btrfs_header_generation(parent));
 472                 btrfs_tree_mod_log_insert_key(parent, parent_slot,
 473                                               BTRFS_MOD_LOG_KEY_REPLACE, GFP_NOFS);
 474                 btrfs_set_node_blockptr(parent, parent_slot,
 475                                         cow->start);
 476                 btrfs_set_node_ptr_generation(parent, parent_slot,
 477                                               trans->transid);
 478                 btrfs_mark_buffer_dirty(parent);
 479                 if (last_ref) {
 480                         ret = btrfs_tree_mod_log_free_eb(buf);
 481                         if (ret) {
 482                                 btrfs_tree_unlock(cow);
 483                                 free_extent_buffer(cow);
 484                                 btrfs_abort_transaction(trans, ret);
 485                                 return ret;
 486                         }
 487                 }
 488                 btrfs_free_tree_block(trans, root, buf, parent_start,
 489                                       last_ref);
 490         }
 491         if (unlock_orig)
 492                 btrfs_tree_unlock(buf);
 493         free_extent_buffer_stale(buf);
 494         btrfs_mark_buffer_dirty(cow);
 495         *cow_ret = cow;
 496         return 0;
 497 }
 498
 499 static inline int should_cow_block(struct btrfs_trans_handle *trans,
 500                                    struct btrfs_root *root,
 501                                    struct extent_buffer *buf)
 502 {
 503         if (btrfs_is_testing(root->fs_info))
 504                 return 0;
 505
 506         /* Ensure we can see the FORCE_COW bit */
 507         smp_mb__before_atomic();
 508
 509         /*
 510          * We do not need to cow a block if
 511          * 1) this block is not created or changed in this transaction;
 512          * 2) this block does not belong to TREE_RELOC tree;
 513          * 3) the root is not forced COW.
 514          *
 515          * What is forced COW:
 516          *    when we create snapshot during committing the transaction,
 517          *    after we've finished copying src root, we must COW the shared
 518          *    block to ensure the metadata consistency.
 519          */
 520         if (btrfs_header_generation(buf) == trans->transid &&
 521             !btrfs_header_flag(buf, BTRFS_HEADER_FLAG_WRITTEN) &&
 522             !(root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID &&
 523               btrfs_header_flag(buf, BTRFS_HEADER_FLAG_RELOC)) &&
 524             !test_bit(BTRFS_ROOT_FORCE_COW, &root->state))
 525                 return 0;
 526         return 1;
 527 }
 528
 529 /*
 530  * cows a single block, see __btrfs_cow_block for the real work.
 531  * This version of it has extra checks so that a block isn't COWed more than
 532  * once per transaction, as long as it hasn't been written yet
 533  */
 534 noinline int btrfs_cow_block(struct btrfs_trans_handle *trans,
 535                     struct btrfs_root *root, struct extent_buffer *buf,
 536                     struct extent_buffer *parent, int parent_slot,
 537                     struct extent_buffer **cow_ret,
 538                     enum btrfs_lock_nesting nest)
 539 {
 540         struct btrfs_fs_info *fs_info = root->fs_info;
 541         u64 search_start;
 542         int ret;
 543
 544         if (test_bit(BTRFS_ROOT_DELETING, &root->state))
 545                 btrfs_err(fs_info,
 546                         "COW'ing blocks on a fs root that's being dropped");
 547
 548         if (trans->transaction != fs_info->running_transaction)
 549                 WARN(1, KERN_CRIT "trans %llu running %llu\n",
 550                        trans->transid,
 551                        fs_info->running_transaction->transid);
 552
 553         if (trans->transid != fs_info->generation)
 554                 WARN(1, KERN_CRIT "trans %llu running %llu\n",
 555                        trans->transid, fs_info->generation);
 556
 557         if (!should_cow_block(trans, root, buf)) {
 558                 *cow_ret = buf;
 559                 return 0;
 560         }
 561
 562         search_start = buf->start & ~((u64)SZ_1G - 1);
 563
 564         /*
 565          * Before CoWing this block for later modification, check if it's
 566          * the subtree root and do the delayed subtree trace if needed.
 567          *
 568          * Also We don't care about the error, as it's handled internally.
 569          */
 570         btrfs_qgroup_trace_subtree_after_cow(trans, root, buf);
 571         ret = __btrfs_cow_block(trans, root, buf, parent,
 572                                  parent_slot, cow_ret, search_start, 0, nest);
 573
 574         trace_btrfs_cow_block(root, buf, *cow_ret);
 575
 576         return ret;
 577 }
 578 ALLOW_ERROR_INJECTION(btrfs_cow_block, ERRNO);
 579
 580 /*
 581  * helper function for defrag to decide if two blocks pointed to by a
 582  * node are actually close by
 583  */
 584 static int close_blocks(u64 blocknr, u64 other, u32 blocksize)
 585 {
 586         if (blocknr < other && other - (blocknr + blocksize) < 32768)
 587                 return 1;
 588         if (blocknr > other && blocknr - (other + blocksize) < 32768)
 589                 return 1;
 590         return 0;
 591 }
 592
 593 #ifdef __LITTLE_ENDIAN
 594
 595 /*
 596  * Compare two keys, on little-endian the disk order is same as CPU order and
 597  * we can avoid the conversion.
 598  */
 599 static int comp_keys(const struct btrfs_disk_key *disk_key,
 600                      const struct btrfs_key *k2)
 601 {
 602         const struct btrfs_key *k1 = (const struct btrfs_key *)disk_key;
 603
 604         return btrfs_comp_cpu_keys(k1, k2);
 605 }
 606
 607 #else
 608
 609 /*
 610  * compare two keys in a memcmp fashion
 611  */
 612 static int comp_keys(const struct btrfs_disk_key *disk,
 613                      const struct btrfs_key *k2)
 614 {
 615         struct btrfs_key k1;
 616
 617         btrfs_disk_key_to_cpu(&k1, disk);
 618
 619         return btrfs_comp_cpu_keys(&k1, k2);
 620 }
 621 #endif
 622
 623 /*
 624  * same as comp_keys only with two btrfs_key's
 625  */
 626 int __pure btrfs_comp_cpu_keys(const struct btrfs_key *k1, const struct btrfs_key *k2)
 627 {
 628         if (k1->objectid > k2->objectid)
 629                 return 1;
 630         if (k1->objectid < k2->objectid)
 631                 return -1;
 632         if (k1->type > k2->type)
 633                 return 1;
 634         if (k1->type < k2->type)
 635                 return -1;
 636         if (k1->offset > k2->offset)
 637                 return 1;
 638         if (k1->offset < k2->offset)
 639                 return -1;
 640         return 0;
 641 }
 642
 643 /*
 644  * this is used by the defrag code to go through all the
 645  * leaves pointed to by a node and reallocate them so that
 646  * disk order is close to key order
 647  */
 648 int btrfs_realloc_node(struct btrfs_trans_handle *trans,
 649                        struct btrfs_root *root, struct extent_buffer *parent,
 650                        int start_slot, u64 *last_ret,
 651                        struct btrfs_key *progress)
 652 {
 653         struct btrfs_fs_info *fs_info = root->fs_info;
 654         struct extent_buffer *cur;
 655         u64 blocknr;
 656         u64 search_start = *last_ret;
 657         u64 last_block = 0;
 658         u64 other;
 659         u32 parent_nritems;
 660         int end_slot;
 661         int i;
 662         int err = 0;
 663         u32 blocksize;
 664         int progress_passed = 0;
 665         struct btrfs_disk_key disk_key;
 666
 667         WARN_ON(trans->transaction != fs_info->running_transaction);
 668         WARN_ON(trans->transid != fs_info->generation);
 669
 670         parent_nritems = btrfs_header_nritems(parent);
 671         blocksize = fs_info->nodesize;
 672         end_slot = parent_nritems - 1;
 673
 674         if (parent_nritems <= 1)
 675                 return 0;
 676
 677         for (i = start_slot; i <= end_slot; i++) {
 678                 int close = 1;
 679
 680                 btrfs_node_key(parent, &disk_key, i);
 681                 if (!progress_passed && comp_keys(&disk_key, progress) < 0)
 682                         continue;
 683
 684                 progress_passed = 1;
 685                 blocknr = btrfs_node_blockptr(parent, i);
 686                 if (last_block == 0)
 687                         last_block = blocknr;
 688
 689                 if (i > 0) {
 690                         other = btrfs_node_blockptr(parent, i - 1);
 691                         close = close_blocks(blocknr, other, blocksize);
 692                 }
 693                 if (!close && i < end_slot) {
 694                         other = btrfs_node_blockptr(parent, i + 1);
 695                         close = close_blocks(blocknr, other, blocksize);
 696                 }
 697                 if (close) {
 698                         last_block = blocknr;
 699                         continue;
 700                 }
 701
 702                 cur = btrfs_read_node_slot(parent, i);
 703                 if (IS_ERR(cur))
 704                         return PTR_ERR(cur);
 705                 if (search_start == 0)
 706                         search_start = last_block;
 707
 708                 btrfs_tree_lock(cur);
 709                 err = __btrfs_cow_block(trans, root, cur, parent, i,
 710                                         &cur, search_start,
 711                                         min(16 * blocksize,
 712                                             (end_slot - i) * blocksize),
 713                                         BTRFS_NESTING_COW);
 714                 if (err) {
 715                         btrfs_tree_unlock(cur);
 716                         free_extent_buffer(cur);
 717                         break;
 718                 }
 719                 search_start = cur->start;
 720                 last_block = cur->start;
 721                 *last_ret = search_start;
 722                 btrfs_tree_unlock(cur);
 723                 free_extent_buffer(cur);
 724         }
 725         return err;
 726 }
 727
 728 /*
 729  * search for key in the extent_buffer.  The items start at offset p,
 730  * and they are item_size apart.
 731  *
 732  * the slot in the array is returned via slot, and it points to
 733  * the place where you would insert key if it is not found in
 734  * the array.
 735  *
 736  * Slot may point to total number of items if the key is bigger than
 737  * all of the keys
 738  */
 739 static noinline int generic_bin_search(struct extent_buffer *eb,
 740                                        unsigned long p, int item_size,
 741                                        const struct btrfs_key *key, int *slot)
 742 {
 743         int low = 0;
 744         int high = btrfs_header_nritems(eb);
 745         int ret;
 746         const int key_size = sizeof(struct btrfs_disk_key);
 747
 748         if (low > high) {
 749                 btrfs_err(eb->fs_info,
 750                  "%s: low (%d) > high (%d) eb %llu owner %llu level %d",
 751                           __func__, low, high, eb->start,
 752                           btrfs_header_owner(eb), btrfs_header_level(eb));
 753                 return -EINVAL;
 754         }
 755
 756         while (low < high) {
 757                 unsigned long oip;
 758                 unsigned long offset;
 759                 struct btrfs_disk_key *tmp;
 760                 struct btrfs_disk_key unaligned;
 761                 int mid;
 762
 763                 mid = (low + high) / 2;
 764                 offset = p + mid * item_size;
 765                 oip = offset_in_page(offset);
 766
 767                 if (oip + key_size <= PAGE_SIZE) {
 768                         const unsigned long idx = get_eb_page_index(offset);
 769                         char *kaddr = page_address(eb->pages[idx]);
 770
 771                         oip = get_eb_offset_in_page(eb, offset);
 772                         tmp = (struct btrfs_disk_key *)(kaddr + oip);
 773                 } else {
 774                         read_extent_buffer(eb, &unaligned, offset, key_size);
 775                         tmp = &unaligned;
 776                 }
 777
 778                 ret = comp_keys(tmp, key);
 779
 780                 if (ret < 0)
 781                         low = mid + 1;
 782                 else if (ret > 0)
 783                         high = mid;
 784                 else {
 785                         *slot = mid;
 786                         return 0;
 787                 }
 788         }
 789         *slot = low;
 790         return 1;
 791 }
 792
 793 /*
 794  * simple bin_search frontend that does the right thing for
 795  * leaves vs nodes
 796  */
 797 int btrfs_bin_search(struct extent_buffer *eb, const struct btrfs_key *key,
 798                      int *slot)
 799 {
 800         if (btrfs_header_level(eb) == 0)
 801                 return generic_bin_search(eb,
 802                                           offsetof(struct btrfs_leaf, items),
 803                                           sizeof(struct btrfs_item), key, slot);
 804         else
 805                 return generic_bin_search(eb,
 806                                           offsetof(struct btrfs_node, ptrs),
 807                                           sizeof(struct btrfs_key_ptr), key, slot);
 808 }
 809
 810 static void root_add_used(struct btrfs_root *root, u32 size)
 811 {
 812         spin_lock(&root->accounting_lock);
 813         btrfs_set_root_used(&root->root_item,
 814                             btrfs_root_used(&root->root_item) + size);
 815         spin_unlock(&root->accounting_lock);
 816 }
 817
 818 static void root_sub_used(struct btrfs_root *root, u32 size)
 819 {
 820         spin_lock(&root->accounting_lock);
 821         btrfs_set_root_used(&root->root_item,
 822                             btrfs_root_used(&root->root_item) - size);
 823         spin_unlock(&root->accounting_lock);
 824 }
 825
 826 /* given a node and slot number, this reads the blocks it points to.  The
 827  * extent buffer is returned with a reference taken (but unlocked).
 828  */
 829 struct extent_buffer *btrfs_read_node_slot(struct extent_buffer *parent,
 830                                            int slot)
 831 {
 832         int level = btrfs_header_level(parent);
 833         struct extent_buffer *eb;
 834         struct btrfs_key first_key;
 835
 836         if (slot < 0 || slot >= btrfs_header_nritems(parent))
 837                 return ERR_PTR(-ENOENT);
 838
 839         BUG_ON(level == 0);
 840
 841         btrfs_node_key_to_cpu(parent, &first_key, slot);
 842         eb = read_tree_block(parent->fs_info, btrfs_node_blockptr(parent, slot),
 843                              btrfs_header_owner(parent),
 844                              btrfs_node_ptr_generation(parent, slot),
 845                              level - 1, &first_key);
 846         if (!IS_ERR(eb) && !extent_buffer_uptodate(eb)) {
 847                 free_extent_buffer(eb);
 848                 eb = ERR_PTR(-EIO);
 849         }
 850
 851         return eb;
 852 }
 853
 854 /*
 855  * node level balancing, used to make sure nodes are in proper order for
 856  * item deletion.  We balance from the top down, so we have to make sure
 857  * that a deletion won't leave an node completely empty later on.
 858  */
 859 static noinline int balance_level(struct btrfs_trans_handle *trans,
 860                          struct btrfs_root *root,
 861                          struct btrfs_path *path, int level)
 862 {
 863         struct btrfs_fs_info *fs_info = root->fs_info;
 864         struct extent_buffer *right = NULL;
 865         struct extent_buffer *mid;
 866         struct extent_buffer *left = NULL;
 867         struct extent_buffer *parent = NULL;
 868         int ret = 0;
 869         int wret;
 870         int pslot;
 871         int orig_slot = path->slots[level];
 872         u64 orig_ptr;
 873
 874         ASSERT(level > 0);
 875
 876         mid = path->nodes[level];
 877
 878         WARN_ON(path->locks[level] != BTRFS_WRITE_LOCK);
 879         WARN_ON(btrfs_header_generation(mid) != trans->transid);
 880
 881         orig_ptr = btrfs_node_blockptr(mid, orig_slot);
 882
 883         if (level < BTRFS_MAX_LEVEL - 1) {
 884                 parent = path->nodes[level + 1];
 885                 pslot = path->slots[level + 1];
 886         }
 887
 888         /*
 889          * deal with the case where there is only one pointer in the root
 890          * by promoting the node below to a root
 891          */
 892         if (!parent) {
 893                 struct extent_buffer *child;
 894
 895                 if (btrfs_header_nritems(mid) != 1)
 896                         return 0;
 897
 898                 /* promote the child to a root */
 899                 child = btrfs_read_node_slot(mid, 0);
 900                 if (IS_ERR(child)) {
 901                         ret = PTR_ERR(child);
 902                         btrfs_handle_fs_error(fs_info, ret, NULL);
 903                         goto enospc;
 904                 }
 905
 906                 btrfs_tree_lock(child);
 907                 ret = btrfs_cow_block(trans, root, child, mid, 0, &child,
 908                                       BTRFS_NESTING_COW);
 909                 if (ret) {
 910                         btrfs_tree_unlock(child);
 911                         free_extent_buffer(child);
 912                         goto enospc;
 913                 }
 914
 915                 ret = btrfs_tree_mod_log_insert_root(root->node, child, true);
 916                 BUG_ON(ret < 0);
 917                 rcu_assign_pointer(root->node, child);
 918
 919                 add_root_to_dirty_list(root);
 920                 btrfs_tree_unlock(child);
 921
 922                 path->locks[level] = 0;
 923                 path->nodes[level] = NULL;
 924                 btrfs_clean_tree_block(mid);
 925                 btrfs_tree_unlock(mid);
 926                 /* once for the path */
 927                 free_extent_buffer(mid);
 928
 929                 root_sub_used(root, mid->len);
 930                 btrfs_free_tree_block(trans, root, mid, 0, 1);
 931                 /* once for the root ptr */
 932                 free_extent_buffer_stale(mid);
 933                 return 0;
 934         }
 935         if (btrfs_header_nritems(mid) >
 936             BTRFS_NODEPTRS_PER_BLOCK(fs_info) / 4)
 937                 return 0;
 938
 939         left = btrfs_read_node_slot(parent, pslot - 1);
 940         if (IS_ERR(left))
 941                 left = NULL;
 942
 943         if (left) {
 944                 __btrfs_tree_lock(left, BTRFS_NESTING_LEFT);
 945                 wret = btrfs_cow_block(trans, root, left,
 946                                        parent, pslot - 1, &left,
 947                                        BTRFS_NESTING_LEFT_COW);
 948                 if (wret) {
 949                         ret = wret;
 950                         goto enospc;
 951                 }
 952         }
 953
 954         right = btrfs_read_node_slot(parent, pslot + 1);
 955         if (IS_ERR(right))
 956                 right = NULL;
 957
 958         if (right) {
 959                 __btrfs_tree_lock(right, BTRFS_NESTING_RIGHT);
 960                 wret = btrfs_cow_block(trans, root, right,
 961                                        parent, pslot + 1, &right,
 962                                        BTRFS_NESTING_RIGHT_COW);
 963                 if (wret) {
 964                         ret = wret;
 965                         goto enospc;
 966                 }
 967         }
 968
 969         /* first, try to make some room in the middle buffer */
 970         if (left) {
 971                 orig_slot += btrfs_header_nritems(left);
 972                 wret = push_node_left(trans, left, mid, 1);
 973                 if (wret < 0)
 974                         ret = wret;
 975         }
 976
 977         /*
 978          * then try to empty the right most buffer into the middle
 979          */
 980         if (right) {
 981                 wret = push_node_left(trans, mid, right, 1);
 982                 if (wret < 0 && wret != -ENOSPC)
 983                         ret = wret;
 984                 if (btrfs_header_nritems(right) == 0) {
 985                         btrfs_clean_tree_block(right);
 986                         btrfs_tree_unlock(right);
 987                         del_ptr(root, path, level + 1, pslot + 1);
 988                         root_sub_used(root, right->len);
 989                         btrfs_free_tree_block(trans, root, right, 0, 1);
 990                         free_extent_buffer_stale(right);
 991                         right = NULL;
 992                 } else {
 993                         struct btrfs_disk_key right_key;
 994                         btrfs_node_key(right, &right_key, 0);
 995                         ret = btrfs_tree_mod_log_insert_key(parent, pslot + 1,
 996                                         BTRFS_MOD_LOG_KEY_REPLACE, GFP_NOFS);
 997                         BUG_ON(ret < 0);
 998                         btrfs_set_node_key(parent, &right_key, pslot + 1);
 999                         btrfs_mark_buffer_dirty(parent);
1000                 }
1001         }
1002         if (btrfs_header_nritems(mid) == 1) {
1003                 /*
1004                  * we're not allowed to leave a node with one item in the
1005                  * tree during a delete.  A deletion from lower in the tree
1006                  * could try to delete the only pointer in this node.
1007                  * So, pull some keys from the left.
1008                  * There has to be a left pointer at this point because
1009                  * otherwise we would have pulled some pointers from the
1010                  * right
1011                  */
1012                 if (!left) {
1013                         ret = -EROFS;
1014                         btrfs_handle_fs_error(fs_info, ret, NULL);
1015                         goto enospc;
1016                 }
1017                 wret = balance_node_right(trans, mid, left);
1018                 if (wret < 0) {
1019                         ret = wret;
1020                         goto enospc;
1021                 }
1022                 if (wret == 1) {
1023                         wret = push_node_left(trans, left, mid, 1);
1024                         if (wret < 0)
1025                                 ret = wret;
1026                 }
1027                 BUG_ON(wret == 1);
1028         }
1029         if (btrfs_header_nritems(mid) == 0) {
1030                 btrfs_clean_tree_block(mid);
1031                 btrfs_tree_unlock(mid);
1032                 del_ptr(root, path, level + 1, pslot);
1033                 root_sub_used(root, mid->len);
1034                 btrfs_free_tree_block(trans, root, mid, 0, 1);
1035                 free_extent_buffer_stale(mid);
1036                 mid = NULL;
1037         } else {
1038                 /* update the parent key to reflect our changes */
1039                 struct btrfs_disk_key mid_key;
1040                 btrfs_node_key(mid, &mid_key, 0);
1041                 ret = btrfs_tree_mod_log_insert_key(parent, pslot,
1042                                 BTRFS_MOD_LOG_KEY_REPLACE, GFP_NOFS);
1043                 BUG_ON(ret < 0);
1044                 btrfs_set_node_key(parent, &mid_key, pslot);
1045                 btrfs_mark_buffer_dirty(parent);
1046         }
1047
1048         /* update the path */
1049         if (left) {
1050                 if (btrfs_header_nritems(left) > orig_slot) {
1051                         atomic_inc(&left->refs);
1052                         /* left was locked after cow */
1053                         path->nodes[level] = left;
1054                         path->slots[level + 1] -= 1;
1055                         path->slots[level] = orig_slot;
1056                         if (mid) {
1057                                 btrfs_tree_unlock(mid);
1058                                 free_extent_buffer(mid);
1059                         }
1060                 } else {
1061                         orig_slot -= btrfs_header_nritems(left);
1062                         path->slots[level] = orig_slot;
1063                 }
1064         }
1065         /* double check we haven't messed things up */
1066         if (orig_ptr !=
1067             btrfs_node_blockptr(path->nodes[level], path->slots[level]))
1068                 BUG();
1069 enospc:
1070         if (right) {
1071                 btrfs_tree_unlock(right);
1072                 free_extent_buffer(right);
1073         }
1074         if (left) {
1075                 if (path->nodes[level] != left)
1076                         btrfs_tree_unlock(left);
1077                 free_extent_buffer(left);
1078         }
1079         return ret;
1080 }
1081
1082 /* Node balancing for insertion.  Here we only split or push nodes around
1083  * when they are completely full.  This is also done top down, so we
1084  * have to be pessimistic.
1085  */
1086 static noinline int push_nodes_for_insert(struct btrfs_trans_handle *trans,
1087                                           struct btrfs_root *root,
1088                                           struct btrfs_path *path, int level)
1089 {
1090         struct btrfs_fs_info *fs_info = root->fs_info;
1091         struct extent_buffer *right = NULL;
1092         struct extent_buffer *mid;
1093         struct extent_buffer *left = NULL;
1094         struct extent_buffer *parent = NULL;
1095         int ret = 0;
1096         int wret;
1097         int pslot;
1098         int orig_slot = path->slots[level];
1099
1100         if (level == 0)
1101                 return 1;
1102
1103         mid = path->nodes[level];
1104         WARN_ON(btrfs_header_generation(mid) != trans->transid);
1105
1106         if (level < BTRFS_MAX_LEVEL - 1) {
1107                 parent = path->nodes[level + 1];
1108                 pslot = path->slots[level + 1];
1109         }
1110
1111         if (!parent)
1112                 return 1;
1113
1114         left = btrfs_read_node_slot(parent, pslot - 1);
1115         if (IS_ERR(left))
1116                 left = NULL;
1117
1118         /* first, try to make some room in the middle buffer */
1119         if (left) {
1120                 u32 left_nr;
1121
1122                 __btrfs_tree_lock(left, BTRFS_NESTING_LEFT);
1123
1124                 left_nr = btrfs_header_nritems(left);
1125                 if (left_nr >= BTRFS_NODEPTRS_PER_BLOCK(fs_info) - 1) {
1126                         wret = 1;
1127                 } else {
1128                         ret = btrfs_cow_block(trans, root, left, parent,
1129                                               pslot - 1, &left,
1130                                               BTRFS_NESTING_LEFT_COW);
1131                         if (ret)
1132                                 wret = 1;
1133                         else {
1134                                 wret = push_node_left(trans, left, mid, 0);
1135                         }
1136                 }
1137                 if (wret < 0)
1138                         ret = wret;
1139                 if (wret == 0) {
1140                         struct btrfs_disk_key disk_key;
1141                         orig_slot += left_nr;
1142                         btrfs_node_key(mid, &disk_key, 0);
1143                         ret = btrfs_tree_mod_log_insert_key(parent, pslot,
1144                                         BTRFS_MOD_LOG_KEY_REPLACE, GFP_NOFS);
1145                         BUG_ON(ret < 0);
1146                         btrfs_set_node_key(parent, &disk_key, pslot);
1147                         btrfs_mark_buffer_dirty(parent);
1148                         if (btrfs_header_nritems(left) > orig_slot) {
1149                                 path->nodes[level] = left;
1150                                 path->slots[level + 1] -= 1;
1151                                 path->slots[level] = orig_slot;
1152                                 btrfs_tree_unlock(mid);
1153                                 free_extent_buffer(mid);
1154                         } else {
1155                                 orig_slot -=
1156                                         btrfs_header_nritems(left);
1157                                 path->slots[level] = orig_slot;
1158                                 btrfs_tree_unlock(left);
1159                                 free_extent_buffer(left);
1160                         }
1161                         return 0;
1162                 }
1163                 btrfs_tree_unlock(left);
1164                 free_extent_buffer(left);
1165         }
1166         right = btrfs_read_node_slot(parent, pslot + 1);
1167         if (IS_ERR(right))
1168                 right = NULL;
1169
1170         /*
1171          * then try to empty the right most buffer into the middle
1172          */
1173         if (right) {
1174                 u32 right_nr;
1175
1176                 __btrfs_tree_lock(right, BTRFS_NESTING_RIGHT);
1177
1178                 right_nr = btrfs_header_nritems(right);
1179                 if (right_nr >= BTRFS_NODEPTRS_PER_BLOCK(fs_info) - 1) {
1180                         wret = 1;
1181                 } else {
1182                         ret = btrfs_cow_block(trans, root, right,
1183                                               parent, pslot + 1,
1184                                               &right, BTRFS_NESTING_RIGHT_COW);
1185                         if (ret)
1186                                 wret = 1;
1187                         else {
1188                                 wret = balance_node_right(trans, right, mid);
1189                         }
1190                 }
1191                 if (wret < 0)
1192                         ret = wret;
1193                 if (wret == 0) {
1194                         struct btrfs_disk_key disk_key;
1195
1196                         btrfs_node_key(right, &disk_key, 0);
1197                         ret = btrfs_tree_mod_log_insert_key(parent, pslot + 1,
1198                                         BTRFS_MOD_LOG_KEY_REPLACE, GFP_NOFS);
1199                         BUG_ON(ret < 0);
1200                         btrfs_set_node_key(parent, &disk_key, pslot + 1);
1201                         btrfs_mark_buffer_dirty(parent);
1202
1203                         if (btrfs_header_nritems(mid) <= orig_slot) {
1204                                 path->nodes[level] = right;
1205                                 path->slots[level + 1] += 1;
1206                                 path->slots[level] = orig_slot -
1207                                         btrfs_header_nritems(mid);
1208                                 btrfs_tree_unlock(mid);
1209                                 free_extent_buffer(mid);
1210                         } else {
1211                                 btrfs_tree_unlock(right);
1212                                 free_extent_buffer(right);
1213                         }
1214                         return 0;
1215                 }
1216                 btrfs_tree_unlock(right);
1217                 free_extent_buffer(right);
1218         }
1219         return 1;
1220 }
1221
1222 /*
1223  * readahead one full node of leaves, finding things that are close
1224  * to the block in 'slot', and triggering ra on them.
1225  */
1226 static void reada_for_search(struct btrfs_fs_info *fs_info,
1227                              struct btrfs_path *path,
1228                              int level, int slot, u64 objectid)
1229 {
1230         struct extent_buffer *node;
1231         struct btrfs_disk_key disk_key;
1232         u32 nritems;
1233         u64 search;
1234         u64 target;
1235         u64 nread = 0;
1236         u64 nread_max;
1237         u32 nr;
1238         u32 blocksize;
1239         u32 nscan = 0;
1240
1241         if (level != 1 && path->reada != READA_FORWARD_ALWAYS)
1242                 return;
1243
1244         if (!path->nodes[level])
1245                 return;
1246
1247         node = path->nodes[level];
1248
1249         /*
1250          * Since the time between visiting leaves is much shorter than the time
1251          * between visiting nodes, limit read ahead of nodes to 1, to avoid too
1252          * much IO at once (possibly random).
1253          */
1254         if (path->reada == READA_FORWARD_ALWAYS) {
1255                 if (level > 1)
1256                         nread_max = node->fs_info->nodesize;
1257                 else
1258                         nread_max = SZ_128K;
1259         } else {
1260                 nread_max = SZ_64K;
1261         }
1262
1263         search = btrfs_node_blockptr(node, slot);
1264         blocksize = fs_info->nodesize;
1265         if (path->reada != READA_FORWARD_ALWAYS) {
1266                 struct extent_buffer *eb;
1267
1268                 eb = find_extent_buffer(fs_info, search);
1269                 if (eb) {
1270                         free_extent_buffer(eb);
1271                         return;
1272                 }
1273         }
1274
1275         target = search;
1276
1277         nritems = btrfs_header_nritems(node);
1278         nr = slot;
1279
1280         while (1) {
1281                 if (path->reada == READA_BACK) {
1282                         if (nr == 0)
1283                                 break;
1284                         nr--;
1285                 } else if (path->reada == READA_FORWARD ||
1286                            path->reada == READA_FORWARD_ALWAYS) {
1287                         nr++;
1288                         if (nr >= nritems)
1289                                 break;
1290                 }
1291                 if (path->reada == READA_BACK && objectid) {
1292                         btrfs_node_key(node, &disk_key, nr);
1293                         if (btrfs_disk_key_objectid(&disk_key) != objectid)
1294                                 break;
1295                 }
1296                 search = btrfs_node_blockptr(node, nr);
1297                 if (path->reada == READA_FORWARD_ALWAYS ||
1298                     (search <= target && target - search <= 65536) ||
1299                     (search > target && search - target <= 65536)) {
1300                         btrfs_readahead_node_child(node, nr);
1301                         nread += blocksize;
1302                 }
1303                 nscan++;
1304                 if (nread > nread_max || nscan > 32)
1305                         break;
1306         }
1307 }
1308
1309 static noinline void reada_for_balance(struct btrfs_path *path, int level)
1310 {
1311         struct extent_buffer *parent;
1312         int slot;
1313         int nritems;
1314
1315         parent = path->nodes[level + 1];
1316         if (!parent)
1317                 return;
1318
1319         nritems = btrfs_header_nritems(parent);
1320         slot = path->slots[level + 1];
1321
1322         if (slot > 0)
1323                 btrfs_readahead_node_child(parent, slot - 1);
1324         if (slot + 1 < nritems)
1325                 btrfs_readahead_node_child(parent, slot + 1);
1326 }
1327
1328
1329 /*
1330  * when we walk down the tree, it is usually safe to unlock the higher layers
1331  * in the tree.  The exceptions are when our path goes through slot 0, because
1332  * operations on the tree might require changing key pointers higher up in the
1333  * tree.
1334  *
1335  * callers might also have set path->keep_locks, which tells this code to keep
1336  * the lock if the path points to the last slot in the block.  This is part of
1337  * walking through the tree, and selecting the next slot in the higher block.
1338  *
1339  * lowest_unlock sets the lowest level in the tree we're allowed to unlock.  so
1340  * if lowest_unlock is 1, level 0 won't be unlocked
1341  */
1342 static noinline void unlock_up(struct btrfs_path *path, int level,
1343                                int lowest_unlock, int min_write_lock_level,
1344                                int *write_lock_level)
1345 {
1346         int i;
1347         int skip_level = level;
1348         int no_skips = 0;
1349         struct extent_buffer *t;
1350
1351         for (i = level; i < BTRFS_MAX_LEVEL; i++) {
1352                 if (!path->nodes[i])
1353                         break;
1354                 if (!path->locks[i])
1355                         break;
1356                 if (!no_skips && path->slots[i] == 0) {
1357                         skip_level = i + 1;
1358                         continue;
1359                 }
1360                 if (!no_skips && path->keep_locks) {
1361                         u32 nritems;
1362                         t = path->nodes[i];
1363                         nritems = btrfs_header_nritems(t);
1364                         if (nritems < 1 || path->slots[i] >= nritems - 1) {
1365                                 skip_level = i + 1;
1366                                 continue;
1367                         }
1368                 }
1369                 if (skip_level < i && i >= lowest_unlock)
1370                         no_skips = 1;
1371
1372                 t = path->nodes[i];
1373                 if (i >= lowest_unlock && i > skip_level) {
1374                         btrfs_tree_unlock_rw(t, path->locks[i]);
1375                         path->locks[i] = 0;
1376                         if (write_lock_level &&
1377                             i > min_write_lock_level &&
1378                             i <= *write_lock_level) {
1379                                 *write_lock_level = i - 1;
1380                         }
1381                 }
1382         }
1383 }
1384
1385 /*
1386  * helper function for btrfs_search_slot.  The goal is to find a block
1387  * in cache without setting the path to blocking.  If we find the block
1388  * we return zero and the path is unchanged.
1389  *
1390  * If we can't find the block, we set the path blocking and do some
1391  * reada.  -EAGAIN is returned and the search must be repeated.
1392  */
1393 static int
1394 read_block_for_search(struct btrfs_root *root, struct btrfs_path *p,
1395                       struct extent_buffer **eb_ret, int level, int slot,
1396                       const struct btrfs_key *key)
1397 {
1398         struct btrfs_fs_info *fs_info = root->fs_info;
1399         u64 blocknr;
1400         u64 gen;
1401         struct extent_buffer *tmp;
1402         struct btrfs_key first_key;
1403         int ret;
1404         int parent_level;
1405
1406         blocknr = btrfs_node_blockptr(*eb_ret, slot);
1407         gen = btrfs_node_ptr_generation(*eb_ret, slot);
1408         parent_level = btrfs_header_level(*eb_ret);
1409         btrfs_node_key_to_cpu(*eb_ret, &first_key, slot);
1410
1411         tmp = find_extent_buffer(fs_info, blocknr);
1412         if (tmp) {
1413                 if (p->reada == READA_FORWARD_ALWAYS)
1414                         reada_for_search(fs_info, p, level, slot, key->objectid);
1415
1416                 /* first we do an atomic uptodate check */
1417                 if (btrfs_buffer_uptodate(tmp, gen, 1) > 0) {
1418                         /*
1419                          * Do extra check for first_key, eb can be stale due to
1420                          * being cached, read from scrub, or have multiple
1421                          * parents (shared tree blocks).
1422                          */
1423                         if (btrfs_verify_level_key(tmp,
1424                                         parent_level - 1, &first_key, gen)) {
1425                                 free_extent_buffer(tmp);
1426                                 return -EUCLEAN;
1427                         }
1428                         *eb_ret = tmp;
1429                         return 0;
1430                 }
1431
1432                 /* now we're allowed to do a blocking uptodate check */
1433                 ret = btrfs_read_buffer(tmp, gen, parent_level - 1, &first_key);
1434                 if (!ret) {
1435                         *eb_ret = tmp;
1436                         return 0;
1437                 }
1438                 free_extent_buffer(tmp);
1439                 btrfs_release_path(p);
1440                 return -EIO;
1441         }
1442
1443         /*
1444          * reduce lock contention at high levels
1445          * of the btree by dropping locks before
1446          * we read.  Don't release the lock on the current
1447          * level because we need to walk this node to figure
1448          * out which blocks to read.
1449          */
1450         btrfs_unlock_up_safe(p, level + 1);
1451
1452         if (p->reada != READA_NONE)
1453                 reada_for_search(fs_info, p, level, slot, key->objectid);
1454
1455         ret = -EAGAIN;
1456         tmp = read_tree_block(fs_info, blocknr, root->root_key.objectid,
1457                               gen, parent_level - 1, &first_key);
1458         if (!IS_ERR(tmp)) {
1459                 /*
1460                  * If the read above didn't mark this buffer up to date,
1461                  * it will never end up being up to date.  Set ret to EIO now
1462                  * and give up so that our caller doesn't loop forever
1463                  * on our EAGAINs.
1464                  */
1465                 if (!extent_buffer_uptodate(tmp))
1466                         ret = -EIO;
1467                 free_extent_buffer(tmp);
1468         } else {
1469                 ret = PTR_ERR(tmp);
1470         }
1471
1472         btrfs_release_path(p);
1473         return ret;
1474 }
1475
1476 /*
1477  * helper function for btrfs_search_slot.  This does all of the checks
1478  * for node-level blocks and does any balancing required based on
1479  * the ins_len.
1480  *
1481  * If no extra work was required, zero is returned.  If we had to
1482  * drop the path, -EAGAIN is returned and btrfs_search_slot must
1483  * start over
1484  */
1485 static int
1486 setup_nodes_for_search(struct btrfs_trans_handle *trans,
1487                        struct btrfs_root *root, struct btrfs_path *p,
1488                        struct extent_buffer *b, int level, int ins_len,
1489                        int *write_lock_level)
1490 {
1491         struct btrfs_fs_info *fs_info = root->fs_info;
1492         int ret = 0;
1493
1494         if ((p->search_for_split || ins_len > 0) && btrfs_header_nritems(b) >=
1495             BTRFS_NODEPTRS_PER_BLOCK(fs_info) - 3) {
1496
1497                 if (*write_lock_level < level + 1) {
1498                         *write_lock_level = level + 1;
1499                         btrfs_release_path(p);
1500                         return -EAGAIN;
1501                 }
1502
1503                 reada_for_balance(p, level);
1504                 ret = split_node(trans, root, p, level);
1505
1506                 b = p->nodes[level];
1507         } else if (ins_len < 0 && btrfs_header_nritems(b) <
1508                    BTRFS_NODEPTRS_PER_BLOCK(fs_info) / 2) {
1509
1510                 if (*write_lock_level < level + 1) {
1511                         *write_lock_level = level + 1;
1512                         btrfs_release_path(p);
1513                         return -EAGAIN;
1514                 }
1515
1516                 reada_for_balance(p, level);
1517                 ret = balance_level(trans, root, p, level);
1518                 if (ret)
1519                         return ret;
1520
1521                 b = p->nodes[level];
1522                 if (!b) {
1523                         btrfs_release_path(p);
1524                         return -EAGAIN;
1525                 }
1526                 BUG_ON(btrfs_header_nritems(b) == 1);
1527         }
1528         return ret;
1529 }
1530
1531 int btrfs_find_item(struct btrfs_root *fs_root, struct btrfs_path *path,
1532                 u64 iobjectid, u64 ioff, u8 key_type,
1533                 struct btrfs_key *found_key)
1534 {
1535         int ret;
1536         struct btrfs_key key;
1537         struct extent_buffer *eb;
1538
1539         ASSERT(path);
1540         ASSERT(found_key);
1541
1542         key.type = key_type;
1543         key.objectid = iobjectid;
1544         key.offset = ioff;
1545
1546         ret = btrfs_search_slot(NULL, fs_root, &key, path, 0, 0);
1547         if (ret < 0)
1548                 return ret;
1549
1550         eb = path->nodes[0];
1551         if (ret && path->slots[0] >= btrfs_header_nritems(eb)) {
1552                 ret = btrfs_next_leaf(fs_root, path);
1553                 if (ret)
1554                         return ret;
1555                 eb = path->nodes[0];
1556         }
1557
1558         btrfs_item_key_to_cpu(eb, found_key, path->slots[0]);
1559         if (found_key->type != key.type ||
1560                         found_key->objectid != key.objectid)
1561                 return 1;
1562
1563         return 0;
1564 }
1565
1566 static struct extent_buffer *btrfs_search_slot_get_root(struct btrfs_root *root,
1567                                                         struct btrfs_path *p,
1568                                                         int write_lock_level)
1569 {
1570         struct btrfs_fs_info *fs_info = root->fs_info;
1571         struct extent_buffer *b;
1572         int root_lock;
1573         int level = 0;
1574
1575         /* We try very hard to do read locks on the root */
1576         root_lock = BTRFS_READ_LOCK;
1577
1578         if (p->search_commit_root) {
1579                 /*
1580                  * The commit roots are read only so we always do read locks,
1581                  * and we always must hold the commit_root_sem when doing
1582                  * searches on them, the only exception is send where we don't
1583                  * want to block transaction commits for a long time, so
1584                  * we need to clone the commit root in order to avoid races
1585                  * with transaction commits that create a snapshot of one of
1586                  * the roots used by a send operation.
1587                  */
1588                 if (p->need_commit_sem) {
1589                         down_read(&fs_info->commit_root_sem);
1590                         b = btrfs_clone_extent_buffer(root->commit_root);
1591                         up_read(&fs_info->commit_root_sem);
1592                         if (!b)
1593                                 return ERR_PTR(-ENOMEM);
1594
1595                 } else {
1596                         b = root->commit_root;
1597                         atomic_inc(&b->refs);
1598                 }
1599                 level = btrfs_header_level(b);
1600                 /*
1601                  * Ensure that all callers have set skip_locking when
1602                  * p->search_commit_root = 1.
1603                  */
1604                 ASSERT(p->skip_locking == 1);
1605
1606                 goto out;
1607         }
1608
1609         if (p->skip_locking) {
1610                 b = btrfs_root_node(root);
1611                 level = btrfs_header_level(b);
1612                 goto out;
1613         }
1614
1615         /*
1616          * If the level is set to maximum, we can skip trying to get the read
1617          * lock.
1618          */
1619         if (write_lock_level < BTRFS_MAX_LEVEL) {
1620                 /*
1621                  * We don't know the level of the root node until we actually
1622                  * have it read locked
1623                  */
1624                 b = btrfs_read_lock_root_node(root);
1625                 level = btrfs_header_level(b);
1626                 if (level > write_lock_level)
1627                         goto out;
1628
1629                 /* Whoops, must trade for write lock */
1630                 btrfs_tree_read_unlock(b);
1631                 free_extent_buffer(b);
1632         }
1633
1634         b = btrfs_lock_root_node(root);
1635         root_lock = BTRFS_WRITE_LOCK;
1636
1637         /* The level might have changed, check again */
1638         level = btrfs_header_level(b);
1639
1640 out:
1641         p->nodes[level] = b;
1642         if (!p->skip_locking)
1643                 p->locks[level] = root_lock;
1644         /*
1645          * Callers are responsible for dropping b's references.
1646          */
1647         return b;
1648 }
1649
1650
1651 /*
1652  * btrfs_search_slot - look for a key in a tree and perform necessary
1653  * modifications to preserve tree invariants.
1654  *
1655  * @trans:      Handle of transaction, used when modifying the tree
1656  * @p:          Holds all btree nodes along the search path
1657  * @root:       The root node of the tree
1658  * @key:        The key we are looking for
1659  * @ins_len:    Indicates purpose of search:
1660  *              >0  for inserts it's size of item inserted (*)
1661  *              <0  for deletions
1662  *               0  for plain searches, not modifying the tree
1663  *
1664  *              (*) If size of item inserted doesn't include
1665  *              sizeof(struct btrfs_item), then p->search_for_extension must
1666  *              be set.
1667  * @cow:        boolean should CoW operations be performed. Must always be 1
1668  *              when modifying the tree.
1669  *
1670  * If @ins_len > 0, nodes and leaves will be split as we walk down the tree.
1671  * If @ins_len < 0, nodes will be merged as we walk down the tree (if possible)
1672  *
1673  * If @key is found, 0 is returned and you can find the item in the leaf level
1674  * of the path (level 0)
1675  *
1676  * If @key isn't found, 1 is returned and the leaf level of the path (level 0)
1677  * points to the slot where it should be inserted
1678  *
1679  * If an error is encountered while searching the tree a negative error number
1680  * is returned
1681  */
1682 int btrfs_search_slot(struct btrfs_trans_handle *trans, struct btrfs_root *root,
1683                       const struct btrfs_key *key, struct btrfs_path *p,
1684                       int ins_len, int cow)
1685 {
1686         struct extent_buffer *b;
1687         int slot;
1688         int ret;
1689         int err;
1690         int level;
1691         int lowest_unlock = 1;
1692         /* everything at write_lock_level or lower must be write locked */
1693         int write_lock_level = 0;
1694         u8 lowest_level = 0;
1695         int min_write_lock_level;
1696         int prev_cmp;
1697
1698         lowest_level = p->lowest_level;
1699         WARN_ON(lowest_level && ins_len > 0);
1700         WARN_ON(p->nodes[0] != NULL);
1701         BUG_ON(!cow && ins_len);
1702
1703         if (ins_len < 0) {
1704                 lowest_unlock = 2;
1705
1706                 /* when we are removing items, we might have to go up to level
1707                  * two as we update tree pointers  Make sure we keep write
1708                  * for those levels as well
1709                  */
1710                 write_lock_level = 2;
1711         } else if (ins_len > 0) {
1712                 /*
1713                  * for inserting items, make sure we have a write lock on
1714                  * level 1 so we can update keys
1715                  */
1716                 write_lock_level = 1;
1717         }
1718
1719         if (!cow)
1720                 write_lock_level = -1;
1721
1722         if (cow && (p->keep_locks || p->lowest_level))
1723                 write_lock_level = BTRFS_MAX_LEVEL;
1724
1725         min_write_lock_level = write_lock_level;
1726
1727 again:
1728         prev_cmp = -1;
1729         b = btrfs_search_slot_get_root(root, p, write_lock_level);
1730         if (IS_ERR(b)) {
1731                 ret = PTR_ERR(b);
1732                 goto done;
1733         }
1734
1735         while (b) {
1736                 int dec = 0;
1737
1738                 level = btrfs_header_level(b);
1739
1740                 if (cow) {
1741                         bool last_level = (level == (BTRFS_MAX_LEVEL - 1));
1742
1743                         /*
1744                          * if we don't really need to cow this block
1745                          * then we don't want to set the path blocking,
1746                          * so we test it here
1747                          */
1748                         if (!should_cow_block(trans, root, b))
1749                                 goto cow_done;
1750
1751                         /*
1752                          * must have write locks on this node and the
1753                          * parent
1754                          */
1755                         if (level > write_lock_level ||
1756                             (level + 1 > write_lock_level &&
1757                             level + 1 < BTRFS_MAX_LEVEL &&
1758                             p->nodes[level + 1])) {
1759                                 write_lock_level = level + 1;
1760                                 btrfs_release_path(p);
1761                                 goto again;
1762                         }
1763
1764                         if (last_level)
1765                                 err = btrfs_cow_block(trans, root, b, NULL, 0,
1766                                                       &b,
1767                                                       BTRFS_NESTING_COW);
1768                         else
1769                                 err = btrfs_cow_block(trans, root, b,
1770                                                       p->nodes[level + 1],
1771                                                       p->slots[level + 1], &b,
1772                                                       BTRFS_NESTING_COW);
1773                         if (err) {
1774                                 ret = err;
1775                                 goto done;
1776                         }
1777                 }
1778 cow_done:
1779                 p->nodes[level] = b;
1780                 /*
1781                  * Leave path with blocking locks to avoid massive
1782                  * lock context switch, this is made on purpose.
1783                  */
1784
1785                 /*
1786                  * we have a lock on b and as long as we aren't changing
1787                  * the tree, there is no way to for the items in b to change.
1788                  * It is safe to drop the lock on our parent before we
1789                  * go through the expensive btree search on b.
1790                  *
1791                  * If we're inserting or deleting (ins_len != 0), then we might
1792                  * be changing slot zero, which may require changing the parent.
1793                  * So, we can't drop the lock until after we know which slot
1794                  * we're operating on.
1795                  */
1796                 if (!ins_len && !p->keep_locks) {
1797                         int u = level + 1;
1798
1799                         if (u < BTRFS_MAX_LEVEL && p->locks[u]) {
1800                                 btrfs_tree_unlock_rw(p->nodes[u], p->locks[u]);
1801                                 p->locks[u] = 0;
1802                         }
1803                 }
1804
1805                 /*
1806                  * If btrfs_bin_search returns an exact match (prev_cmp == 0)
1807                  * we can safely assume the target key will always be in slot 0
1808                  * on lower levels due to the invariants BTRFS' btree provides,
1809                  * namely that a btrfs_key_ptr entry always points to the
1810                  * lowest key in the child node, thus we can skip searching
1811                  * lower levels
1812                  */
1813                 if (prev_cmp == 0) {
1814                         slot = 0;
1815                         ret = 0;
1816                 } else {
1817                         ret = btrfs_bin_search(b, key, &slot);
1818                         prev_cmp = ret;
1819                         if (ret < 0)
1820                                 goto done;
1821                 }
1822
1823                 if (level == 0) {
1824                         p->slots[level] = slot;
1825                         /*
1826                          * Item key already exists. In this case, if we are
1827                          * allowed to insert the item (for example, in dir_item
1828                          * case, item key collision is allowed), it will be
1829                          * merged with the original item. Only the item size
1830                          * grows, no new btrfs item will be added. If
1831                          * search_for_extension is not set, ins_len already
1832                          * accounts the size btrfs_item, deduct it here so leaf
1833                          * space check will be correct.
1834                          */
1835                         if (ret == 0 && ins_len > 0 && !p->search_for_extension) {
1836                                 ASSERT(ins_len >= sizeof(struct btrfs_item));
1837                                 ins_len -= sizeof(struct btrfs_item);
1838                         }
1839                         if (ins_len > 0 &&
1840                             btrfs_leaf_free_space(b) < ins_len) {
1841                                 if (write_lock_level < 1) {
1842                                         write_lock_level = 1;
1843                                         btrfs_release_path(p);
1844                                         goto again;
1845                                 }
1846
1847                                 err = split_leaf(trans, root, key,
1848                                                  p, ins_len, ret == 0);
1849
1850                                 BUG_ON(err > 0);
1851                                 if (err) {
1852                                         ret = err;
1853                                         goto done;
1854                                 }
1855                         }
1856                         if (!p->search_for_split)
1857                                 unlock_up(p, level, lowest_unlock,
1858                                           min_write_lock_level, NULL);
1859                         goto done;
1860                 }
1861                 if (ret && slot > 0) {
1862                         dec = 1;
1863                         slot--;
1864                 }
1865                 p->slots[level] = slot;
1866                 err = setup_nodes_for_search(trans, root, p, b, level, ins_len,
1867                                              &write_lock_level);
1868                 if (err == -EAGAIN)
1869                         goto again;
1870                 if (err) {
1871                         ret = err;
1872                         goto done;
1873                 }
1874                 b = p->nodes[level];
1875                 slot = p->slots[level];
1876
1877                 /*
1878                  * Slot 0 is special, if we change the key we have to update
1879                  * the parent pointer which means we must have a write lock on
1880                  * the parent
1881                  */
1882                 if (slot == 0 && ins_len && write_lock_level < level + 1) {
1883                         write_lock_level = level + 1;
1884                         btrfs_release_path(p);
1885                         goto again;
1886                 }
1887
1888                 unlock_up(p, level, lowest_unlock, min_write_lock_level,
1889                           &write_lock_level);
1890
1891                 if (level == lowest_level) {
1892                         if (dec)
1893                                 p->slots[level]++;
1894                         goto done;
1895                 }
1896
1897                 err = read_block_for_search(root, p, &b, level, slot, key);
1898                 if (err == -EAGAIN)
1899                         goto again;
1900                 if (err) {
1901                         ret = err;
1902                         goto done;
1903                 }
1904
1905                 if (!p->skip_locking) {
1906                         level = btrfs_header_level(b);
1907                         if (level <= write_lock_level) {
1908                                 btrfs_tree_lock(b);
1909                                 p->locks[level] = BTRFS_WRITE_LOCK;
1910                         } else {
1911                                 btrfs_tree_read_lock(b);
1912                                 p->locks[level] = BTRFS_READ_LOCK;
1913                         }
1914                         p->nodes[level] = b;
1915                 }
1916         }
1917         ret = 1;
1918 done:
1919         if (ret < 0 && !p->skip_release_on_error)
1920                 btrfs_release_path(p);
1921         return ret;
1922 }
1923 ALLOW_ERROR_INJECTION(btrfs_search_slot, ERRNO);
1924
1925 /*
1926  * Like btrfs_search_slot, this looks for a key in the given tree. It uses the
1927  * current state of the tree together with the operations recorded in the tree
1928  * modification log to search for the key in a previous version of this tree, as
1929  * denoted by the time_seq parameter.
1930  *
1931  * Naturally, there is no support for insert, delete or cow operations.
1932  *
1933  * The resulting path and return value will be set up as if we called
1934  * btrfs_search_slot at that point in time with ins_len and cow both set to 0.
1935  */
1936 int btrfs_search_old_slot(struct btrfs_root *root, const struct btrfs_key *key,
1937                           struct btrfs_path *p, u64 time_seq)
1938 {
1939         struct btrfs_fs_info *fs_info = root->fs_info;
1940         struct extent_buffer *b;
1941         int slot;
1942         int ret;
1943         int err;
1944         int level;
1945         int lowest_unlock = 1;
1946         u8 lowest_level = 0;
1947
1948         lowest_level = p->lowest_level;
1949         WARN_ON(p->nodes[0] != NULL);
1950
1951         if (p->search_commit_root) {
1952                 BUG_ON(time_seq);
1953                 return btrfs_search_slot(NULL, root, key, p, 0, 0);
1954         }
1955
1956 again:
1957         b = btrfs_get_old_root(root, time_seq);
1958         if (!b) {
1959                 ret = -EIO;
1960                 goto done;
1961         }
1962         level = btrfs_header_level(b);
1963         p->locks[level] = BTRFS_READ_LOCK;
1964
1965         while (b) {
1966                 int dec = 0;
1967
1968                 level = btrfs_header_level(b);
1969                 p->nodes[level] = b;
1970
1971                 /*
1972                  * we have a lock on b and as long as we aren't changing
1973                  * the tree, there is no way to for the items in b to change.
1974                  * It is safe to drop the lock on our parent before we
1975                  * go through the expensive btree search on b.
1976                  */
1977                 btrfs_unlock_up_safe(p, level + 1);
1978
1979                 ret = btrfs_bin_search(b, key, &slot);
1980                 if (ret < 0)
1981                         goto done;
1982
1983                 if (level == 0) {
1984                         p->slots[level] = slot;
1985                         unlock_up(p, level, lowest_unlock, 0, NULL);
1986                         goto done;
1987                 }
1988
1989                 if (ret && slot > 0) {
1990                         dec = 1;
1991                         slot--;
1992                 }
1993                 p->slots[level] = slot;
1994                 unlock_up(p, level, lowest_unlock, 0, NULL);
1995
1996                 if (level == lowest_level) {
1997                         if (dec)
1998                                 p->slots[level]++;
1999                         goto done;
2000                 }
2001
2002                 err = read_block_for_search(root, p, &b, level, slot, key);
2003                 if (err == -EAGAIN)
2004                         goto again;
2005                 if (err) {
2006                         ret = err;
2007                         goto done;
2008                 }
2009
2010                 level = btrfs_header_level(b);
2011                 btrfs_tree_read_lock(b);
2012                 b = btrfs_tree_mod_log_rewind(fs_info, p, b, time_seq);
2013                 if (!b) {
2014                         ret = -ENOMEM;
2015                         goto done;
2016                 }
2017                 p->locks[level] = BTRFS_READ_LOCK;
2018                 p->nodes[level] = b;
2019         }
2020         ret = 1;
2021 done:
2022         if (ret < 0)
2023                 btrfs_release_path(p);
2024
2025         return ret;
2026 }
2027
2028 /*
2029  * helper to use instead of search slot if no exact match is needed but
2030  * instead the next or previous item should be returned.
2031  * When find_higher is true, the next higher item is returned, the next lower
2032  * otherwise.
2033  * When return_any and find_higher are both true, and no higher item is found,
2034  * return the next lower instead.
2035  * When return_any is true and find_higher is false, and no lower item is found,
2036  * return the next higher instead.
2037  * It returns 0 if any item is found, 1 if none is found (tree empty), and
2038  * < 0 on error
2039  */
2040 int btrfs_search_slot_for_read(struct btrfs_root *root,
2041                                const struct btrfs_key *key,
2042                                struct btrfs_path *p, int find_higher,
2043                                int return_any)
2044 {
2045         int ret;
2046         struct extent_buffer *leaf;
2047
2048 again:
2049         ret = btrfs_search_slot(NULL, root, key, p, 0, 0);
2050         if (ret <= 0)
2051                 return ret;
2052         /*
2053          * a return value of 1 means the path is at the position where the
2054          * item should be inserted. Normally this is the next bigger item,
2055          * but in case the previous item is the last in a leaf, path points
2056          * to the first free slot in the previous leaf, i.e. at an invalid
2057          * item.
2058          */
2059         leaf = p->nodes[0];
2060
2061         if (find_higher) {
2062                 if (p->slots[0] >= btrfs_header_nritems(leaf)) {
2063                         ret = btrfs_next_leaf(root, p);
2064                         if (ret <= 0)
2065                                 return ret;
2066                         if (!return_any)
2067                                 return 1;
2068                         /*
2069                          * no higher item found, return the next
2070                          * lower instead
2071                          */
2072                         return_any = 0;
2073                         find_higher = 0;
2074                         btrfs_release_path(p);
2075                         goto again;
2076                 }
2077         } else {
2078                 if (p->slots[0] == 0) {
2079                         ret = btrfs_prev_leaf(root, p);
2080                         if (ret < 0)
2081                                 return ret;
2082                         if (!ret) {
2083                                 leaf = p->nodes[0];
2084                                 if (p->slots[0] == btrfs_header_nritems(leaf))
2085                                         p->slots[0]--;
2086                                 return 0;
2087                         }
2088                         if (!return_any)
2089                                 return 1;
2090                         /*
2091                          * no lower item found, return the next
2092                          * higher instead
2093                          */
2094                         return_any = 0;
2095                         find_higher = 1;
2096                         btrfs_release_path(p);
2097                         goto again;
2098                 } else {
2099                         --p->slots[0];
2100                 }
2101         }
2102         return 0;
2103 }
2104
2105 /*
2106  * Execute search and call btrfs_previous_item to traverse backwards if the item
2107  * was not found.
2108  *
2109  * Return 0 if found, 1 if not found and < 0 if error.
2110  */
2111 int btrfs_search_backwards(struct btrfs_root *root, struct btrfs_key *key,
2112                            struct btrfs_path *path)
2113 {
2114         int ret;
2115
2116         ret = btrfs_search_slot(NULL, root, key, path, 0, 0);
2117         if (ret > 0)
2118                 ret = btrfs_previous_item(root, path, key->objectid, key->type);
2119
2120         if (ret == 0)
2121                 btrfs_item_key_to_cpu(path->nodes[0], key, path->slots[0]);
2122
2123         return ret;
2124 }
2125
2126 /*
2127  * adjust the pointers going up the tree, starting at level
2128  * making sure the right key of each node is points to 'key'.
2129  * This is used after shifting pointers to the left, so it stops
2130  * fixing up pointers when a given leaf/node is not in slot 0 of the
2131  * higher levels
2132  *
2133  */
2134 static void fixup_low_keys(struct btrfs_path *path,
2135                            struct btrfs_disk_key *key, int level)
2136 {
2137         int i;
2138         struct extent_buffer *t;
2139         int ret;
2140
2141         for (i = level; i < BTRFS_MAX_LEVEL; i++) {
2142                 int tslot = path->slots[i];
2143
2144                 if (!path->nodes[i])
2145                         break;
2146                 t = path->nodes[i];
2147                 ret = btrfs_tree_mod_log_insert_key(t, tslot,
2148                                 BTRFS_MOD_LOG_KEY_REPLACE, GFP_ATOMIC);
2149                 BUG_ON(ret < 0);
2150                 btrfs_set_node_key(t, key, tslot);
2151                 btrfs_mark_buffer_dirty(path->nodes[i]);
2152                 if (tslot != 0)
2153                         break;
2154         }
2155 }
2156
2157 /*
2158  * update item key.
2159  *
2160  * This function isn't completely safe. It's the caller's responsibility
2161  * that the new key won't break the order
2162  */
2163 void btrfs_set_item_key_safe(struct btrfs_fs_info *fs_info,
2164                              struct btrfs_path *path,
2165                              const struct btrfs_key *new_key)
2166 {
2167         struct btrfs_disk_key disk_key;
2168         struct extent_buffer *eb;
2169         int slot;
2170
2171         eb = path->nodes[0];
2172         slot = path->slots[0];
2173         if (slot > 0) {
2174                 btrfs_item_key(eb, &disk_key, slot - 1);
2175                 if (unlikely(comp_keys(&disk_key, new_key) >= 0)) {
2176                         btrfs_crit(fs_info,
2177                 "slot %u key (%llu %u %llu) new key (%llu %u %llu)",
2178                                    slot, btrfs_disk_key_objectid(&disk_key),
2179                                    btrfs_disk_key_type(&disk_key),
2180                                    btrfs_disk_key_offset(&disk_key),
2181                                    new_key->objectid, new_key->type,
2182                                    new_key->offset);
2183                         btrfs_print_leaf(eb);
2184                         BUG();
2185                 }
2186         }
2187         if (slot < btrfs_header_nritems(eb) - 1) {
2188                 btrfs_item_key(eb, &disk_key, slot + 1);
2189                 if (unlikely(comp_keys(&disk_key, new_key) <= 0)) {
2190                         btrfs_crit(fs_info,
2191                 "slot %u key (%llu %u %llu) new key (%llu %u %llu)",
2192                                    slot, btrfs_disk_key_objectid(&disk_key),
2193                                    btrfs_disk_key_type(&disk_key),
2194                                    btrfs_disk_key_offset(&disk_key),
2195                                    new_key->objectid, new_key->type,
2196                                    new_key->offset);
2197                         btrfs_print_leaf(eb);
2198                         BUG();
2199                 }
2200         }
2201
2202         btrfs_cpu_key_to_disk(&disk_key, new_key);
2203         btrfs_set_item_key(eb, &disk_key, slot);
2204         btrfs_mark_buffer_dirty(eb);
2205         if (slot == 0)
2206                 fixup_low_keys(path, &disk_key, 1);
2207 }
2208
2209 /*
2210  * Check key order of two sibling extent buffers.
2211  *
2212  * Return true if something is wrong.
2213  * Return false if everything is fine.
2214  *
2215  * Tree-checker only works inside one tree block, thus the following
2216  * corruption can not be detected by tree-checker:
2217  *
2218  * Leaf @left                   | Leaf @right
2219  * --------------------------------------------------------------
2220  * | 1 | 2 | 3 | 4 | 5 | f6 |   | 7 | 8 |
2221  *
2222  * Key f6 in leaf @left itself is valid, but not valid when the next
2223  * key in leaf @right is 7.
2224  * This can only be checked at tree block merge time.
2225  * And since tree checker has ensured all key order in each tree block
2226  * is correct, we only need to bother the last key of @left and the first
2227  * key of @right.
2228  */
2229 static bool check_sibling_keys(struct extent_buffer *left,
2230                                struct extent_buffer *right)
2231 {
2232         struct btrfs_key left_last;
2233         struct btrfs_key right_first;
2234         int level = btrfs_header_level(left);
2235         int nr_left = btrfs_header_nritems(left);
2236         int nr_right = btrfs_header_nritems(right);
2237
2238         /* No key to check in one of the tree blocks */
2239         if (!nr_left || !nr_right)
2240                 return false;
2241
2242         if (level) {
2243                 btrfs_node_key_to_cpu(left, &left_last, nr_left - 1);
2244                 btrfs_node_key_to_cpu(right, &right_first, 0);
2245         } else {
2246                 btrfs_item_key_to_cpu(left, &left_last, nr_left - 1);
2247                 btrfs_item_key_to_cpu(right, &right_first, 0);
2248         }
2249
2250         if (btrfs_comp_cpu_keys(&left_last, &right_first) >= 0) {
2251                 btrfs_crit(left->fs_info,
2252 "bad key order, sibling blocks, left last (%llu %u %llu) right first (%llu %u %llu)",
2253                            left_last.objectid, left_last.type,
2254                            left_last.offset, right_first.objectid,
2255                            right_first.type, right_first.offset);
2256                 return true;
2257         }
2258         return false;
2259 }
2260
2261 /*
2262  * try to push data from one node into the next node left in the
2263  * tree.
2264  *
2265  * returns 0 if some ptrs were pushed left, < 0 if there was some horrible
2266  * error, and > 0 if there was no room in the left hand block.
2267  */
2268 static int push_node_left(struct btrfs_trans_handle *trans,
2269                           struct extent_buffer *dst,
2270                           struct extent_buffer *src, int empty)
2271 {
2272         struct btrfs_fs_info *fs_info = trans->fs_info;
2273         int push_items = 0;
2274         int src_nritems;
2275         int dst_nritems;
2276         int ret = 0;
2277
2278         src_nritems = btrfs_header_nritems(src);
2279         dst_nritems = btrfs_header_nritems(dst);
2280         push_items = BTRFS_NODEPTRS_PER_BLOCK(fs_info) - dst_nritems;
2281         WARN_ON(btrfs_header_generation(src) != trans->transid);
2282         WARN_ON(btrfs_header_generation(dst) != trans->transid);
2283
2284         if (!empty && src_nritems <= 8)
2285                 return 1;
2286
2287         if (push_items <= 0)
2288                 return 1;
2289
2290         if (empty) {
2291                 push_items = min(src_nritems, push_items);
2292                 if (push_items < src_nritems) {
2293                         /* leave at least 8 pointers in the node if
2294                          * we aren't going to empty it
2295                          */
2296                         if (src_nritems - push_items < 8) {
2297                                 if (push_items <= 8)
2298                                         return 1;
2299                                 push_items -= 8;
2300                         }
2301                 }
2302         } else
2303                 push_items = min(src_nritems - 8, push_items);
2304
2305         /* dst is the left eb, src is the middle eb */
2306         if (check_sibling_keys(dst, src)) {
2307                 ret = -EUCLEAN;
2308                 btrfs_abort_transaction(trans, ret);
2309                 return ret;
2310         }
2311         ret = btrfs_tree_mod_log_eb_copy(dst, src, dst_nritems, 0, push_items);
2312         if (ret) {
2313                 btrfs_abort_transaction(trans, ret);
2314                 return ret;
2315         }
2316         copy_extent_buffer(dst, src,
2317                            btrfs_node_key_ptr_offset(dst_nritems),
2318                            btrfs_node_key_ptr_offset(0),
2319                            push_items * sizeof(struct btrfs_key_ptr));
2320
2321         if (push_items < src_nritems) {
2322                 /*
2323                  * Don't call btrfs_tree_mod_log_insert_move() here, key removal
2324                  * was already fully logged by btrfs_tree_mod_log_eb_copy() above.
2325                  */
2326                 memmove_extent_buffer(src, btrfs_node_key_ptr_offset(0),
2327                                       btrfs_node_key_ptr_offset(push_items),
2328                                       (src_nritems - push_items) *
2329                                       sizeof(struct btrfs_key_ptr));
2330         }
2331         btrfs_set_header_nritems(src, src_nritems - push_items);
2332         btrfs_set_header_nritems(dst, dst_nritems + push_items);
2333         btrfs_mark_buffer_dirty(src);
2334         btrfs_mark_buffer_dirty(dst);
2335
2336         return ret;
2337 }
2338
2339 /*
2340  * try to push data from one node into the next node right in the
2341  * tree.
2342  *
2343  * returns 0 if some ptrs were pushed, < 0 if there was some horrible
2344  * error, and > 0 if there was no room in the right hand block.
2345  *
2346  * this will  only push up to 1/2 the contents of the left node over
2347  */
2348 static int balance_node_right(struct btrfs_trans_handle *trans,
2349                               struct extent_buffer *dst,
2350                               struct extent_buffer *src)
2351 {
2352         struct btrfs_fs_info *fs_info = trans->fs_info;
2353         int push_items = 0;
2354         int max_push;
2355         int src_nritems;
2356         int dst_nritems;
2357         int ret = 0;
2358
2359         WARN_ON(btrfs_header_generation(src) != trans->transid);
2360         WARN_ON(btrfs_header_generation(dst) != trans->transid);
2361
2362         src_nritems = btrfs_header_nritems(src);
2363         dst_nritems = btrfs_header_nritems(dst);
2364         push_items = BTRFS_NODEPTRS_PER_BLOCK(fs_info) - dst_nritems;
2365         if (push_items <= 0)
2366                 return 1;
2367
2368         if (src_nritems < 4)
2369                 return 1;
2370
2371         max_push = src_nritems / 2 + 1;
2372         /* don't try to empty the node */
2373         if (max_push >= src_nritems)
2374                 return 1;
2375
2376         if (max_push < push_items)
2377                 push_items = max_push;
2378
2379         /* dst is the right eb, src is the middle eb */
2380         if (check_sibling_keys(src, dst)) {
2381                 ret = -EUCLEAN;
2382                 btrfs_abort_transaction(trans, ret);
2383                 return ret;
2384         }
2385         ret = btrfs_tree_mod_log_insert_move(dst, push_items, 0, dst_nritems);
2386         BUG_ON(ret < 0);
2387         memmove_extent_buffer(dst, btrfs_node_key_ptr_offset(push_items),
2388                                       btrfs_node_key_ptr_offset(0),
2389                                       (dst_nritems) *
2390                                       sizeof(struct btrfs_key_ptr));
2391
2392         ret = btrfs_tree_mod_log_eb_copy(dst, src, 0, src_nritems - push_items,
2393                                          push_items);
2394         if (ret) {
2395                 btrfs_abort_transaction(trans, ret);
2396                 return ret;
2397         }
2398         copy_extent_buffer(dst, src,
2399                            btrfs_node_key_ptr_offset(0),
2400                            btrfs_node_key_ptr_offset(src_nritems - push_items),
2401                            push_items * sizeof(struct btrfs_key_ptr));
2402
2403         btrfs_set_header_nritems(src, src_nritems - push_items);
2404         btrfs_set_header_nritems(dst, dst_nritems + push_items);
2405
2406         btrfs_mark_buffer_dirty(src);
2407         btrfs_mark_buffer_dirty(dst);
2408
2409         return ret;
2410 }
2411
2412 /*
2413  * helper function to insert a new root level in the tree.
2414  * A new node is allocated, and a single item is inserted to
2415  * point to the existing root
2416  *
2417  * returns zero on success or < 0 on failure.
2418  */
2419 static noinline int insert_new_root(struct btrfs_trans_handle *trans,
2420                            struct btrfs_root *root,
2421                            struct btrfs_path *path, int level)
2422 {
2423         struct btrfs_fs_info *fs_info = root->fs_info;
2424         u64 lower_gen;
2425         struct extent_buffer *lower;
2426         struct extent_buffer *c;
2427         struct extent_buffer *old;
2428         struct btrfs_disk_key lower_key;
2429         int ret;
2430
2431         BUG_ON(path->nodes[level]);
2432         BUG_ON(path->nodes[level-1] != root->node);
2433
2434         lower = path->nodes[level-1];
2435         if (level == 1)
2436                 btrfs_item_key(lower, &lower_key, 0);
2437         else
2438                 btrfs_node_key(lower, &lower_key, 0);
2439
2440         c = btrfs_alloc_tree_block(trans, root, 0, root->root_key.objectid,
2441                                    &lower_key, level, root->node->start, 0,
2442                                    BTRFS_NESTING_NEW_ROOT);
2443         if (IS_ERR(c))
2444                 return PTR_ERR(c);
2445
2446         root_add_used(root, fs_info->nodesize);
2447
2448         btrfs_set_header_nritems(c, 1);
2449         btrfs_set_node_key(c, &lower_key, 0);
2450         btrfs_set_node_blockptr(c, 0, lower->start);
2451         lower_gen = btrfs_header_generation(lower);
2452         WARN_ON(lower_gen != trans->transid);
2453
2454         btrfs_set_node_ptr_generation(c, 0, lower_gen);
2455
2456         btrfs_mark_buffer_dirty(c);
2457
2458         old = root->node;
2459         ret = btrfs_tree_mod_log_insert_root(root->node, c, false);
2460         BUG_ON(ret < 0);
2461         rcu_assign_pointer(root->node, c);
2462
2463         /* the super has an extra ref to root->node */
2464         free_extent_buffer(old);
2465
2466         add_root_to_dirty_list(root);
2467         atomic_inc(&c->refs);
2468         path->nodes[level] = c;
2469         path->locks[level] = BTRFS_WRITE_LOCK;
2470         path->slots[level] = 0;
2471         return 0;
2472 }
2473
2474 /*
2475  * worker function to insert a single pointer in a node.
2476  * the node should have enough room for the pointer already
2477  *
2478  * slot and level indicate where you want the key to go, and
2479  * blocknr is the block the key points to.
2480  */
2481 static void insert_ptr(struct btrfs_trans_handle *trans,
2482                        struct btrfs_path *path,
2483                        struct btrfs_disk_key *key, u64 bytenr,
2484                        int slot, int level)
2485 {
2486         struct extent_buffer *lower;
2487         int nritems;
2488         int ret;
2489
2490         BUG_ON(!path->nodes[level]);
2491         btrfs_assert_tree_write_locked(path->nodes[level]);
2492         lower = path->nodes[level];
2493         nritems = btrfs_header_nritems(lower);
2494         BUG_ON(slot > nritems);
2495         BUG_ON(nritems == BTRFS_NODEPTRS_PER_BLOCK(trans->fs_info));
2496         if (slot != nritems) {
2497                 if (level) {
2498                         ret = btrfs_tree_mod_log_insert_move(lower, slot + 1,
2499                                         slot, nritems - slot);
2500                         BUG_ON(ret < 0);
2501                 }
2502                 memmove_extent_buffer(lower,
2503                               btrfs_node_key_ptr_offset(slot + 1),
2504                               btrfs_node_key_ptr_offset(slot),
2505                               (nritems - slot) * sizeof(struct btrfs_key_ptr));
2506         }
2507         if (level) {
2508                 ret = btrfs_tree_mod_log_insert_key(lower, slot,
2509                                             BTRFS_MOD_LOG_KEY_ADD, GFP_NOFS);
2510                 BUG_ON(ret < 0);
2511         }
2512         btrfs_set_node_key(lower, key, slot);
2513         btrfs_set_node_blockptr(lower, slot, bytenr);
2514         WARN_ON(trans->transid == 0);
2515         btrfs_set_node_ptr_generation(lower, slot, trans->transid);
2516         btrfs_set_header_nritems(lower, nritems + 1);
2517         btrfs_mark_buffer_dirty(lower);
2518 }
2519
2520 /*
2521  * split the node at the specified level in path in two.
2522  * The path is corrected to point to the appropriate node after the split
2523  *
2524  * Before splitting this tries to make some room in the node by pushing
2525  * left and right, if either one works, it returns right away.
2526  *
2527  * returns 0 on success and < 0 on failure
2528  */
2529 static noinline int split_node(struct btrfs_trans_handle *trans,
2530                                struct btrfs_root *root,
2531                                struct btrfs_path *path, int level)
2532 {
2533         struct btrfs_fs_info *fs_info = root->fs_info;
2534         struct extent_buffer *c;
2535         struct extent_buffer *split;
2536         struct btrfs_disk_key disk_key;
2537         int mid;
2538         int ret;
2539         u32 c_nritems;
2540
2541         c = path->nodes[level];
2542         WARN_ON(btrfs_header_generation(c) != trans->transid);
2543         if (c == root->node) {
2544                 /*
2545                  * trying to split the root, lets make a new one
2546                  *
2547                  * tree mod log: We don't log_removal old root in
2548                  * insert_new_root, because that root buffer will be kept as a
2549                  * normal node. We are going to log removal of half of the
2550                  * elements below with btrfs_tree_mod_log_eb_copy(). We're
2551                  * holding a tree lock on the buffer, which is why we cannot
2552                  * race with other tree_mod_log users.
2553                  */
2554                 ret = insert_new_root(trans, root, path, level + 1);
2555                 if (ret)
2556                         return ret;
2557         } else {
2558                 ret = push_nodes_for_insert(trans, root, path, level);
2559                 c = path->nodes[level];
2560                 if (!ret && btrfs_header_nritems(c) <
2561                     BTRFS_NODEPTRS_PER_BLOCK(fs_info) - 3)
2562                         return 0;
2563                 if (ret < 0)
2564                         return ret;
2565         }
2566
2567         c_nritems = btrfs_header_nritems(c);
2568         mid = (c_nritems + 1) / 2;
2569         btrfs_node_key(c, &disk_key, mid);
2570
2571         split = btrfs_alloc_tree_block(trans, root, 0, root->root_key.objectid,
2572                                        &disk_key, level, c->start, 0,
2573                                        BTRFS_NESTING_SPLIT);
2574         if (IS_ERR(split))
2575                 return PTR_ERR(split);
2576
2577         root_add_used(root, fs_info->nodesize);
2578         ASSERT(btrfs_header_level(c) == level);
2579
2580         ret = btrfs_tree_mod_log_eb_copy(split, c, 0, mid, c_nritems - mid);
2581         if (ret) {
2582                 btrfs_abort_transaction(trans, ret);
2583                 return ret;
2584         }
2585         copy_extent_buffer(split, c,
2586                            btrfs_node_key_ptr_offset(0),
2587                            btrfs_node_key_ptr_offset(mid),
2588                            (c_nritems - mid) * sizeof(struct btrfs_key_ptr));
2589         btrfs_set_header_nritems(split, c_nritems - mid);
2590         btrfs_set_header_nritems(c, mid);
2591
2592         btrfs_mark_buffer_dirty(c);
2593         btrfs_mark_buffer_dirty(split);
2594
2595         insert_ptr(trans, path, &disk_key, split->start,
2596                    path->slots[level + 1] + 1, level + 1);
2597
2598         if (path->slots[level] >= mid) {
2599                 path->slots[level] -= mid;
2600                 btrfs_tree_unlock(c);
2601                 free_extent_buffer(c);
2602                 path->nodes[level] = split;
2603                 path->slots[level + 1] += 1;
2604         } else {
2605                 btrfs_tree_unlock(split);
2606                 free_extent_buffer(split);
2607         }
2608         return 0;
2609 }
2610
2611 /*
2612  * how many bytes are required to store the items in a leaf.  start
2613  * and nr indicate which items in the leaf to check.  This totals up the
2614  * space used both by the item structs and the item data
2615  */
2616 static int leaf_space_used(struct extent_buffer *l, int start, int nr)
2617 {
2618         struct btrfs_item *start_item;
2619         struct btrfs_item *end_item;
2620         int data_len;
2621         int nritems = btrfs_header_nritems(l);
2622         int end = min(nritems, start + nr) - 1;
2623
2624         if (!nr)
2625                 return 0;
2626         start_item = btrfs_item_nr(start);
2627         end_item = btrfs_item_nr(end);
2628         data_len = btrfs_item_offset(l, start_item) +
2629                    btrfs_item_size(l, start_item);
2630         data_len = data_len - btrfs_item_offset(l, end_item);
2631         data_len += sizeof(struct btrfs_item) * nr;
2632         WARN_ON(data_len < 0);
2633         return data_len;
2634 }
2635
2636 /*
2637  * The space between the end of the leaf items and
2638  * the start of the leaf data.  IOW, how much room
2639  * the leaf has left for both items and data
2640  */
2641 noinline int btrfs_leaf_free_space(struct extent_buffer *leaf)
2642 {
2643         struct btrfs_fs_info *fs_info = leaf->fs_info;
2644         int nritems = btrfs_header_nritems(leaf);
2645         int ret;
2646
2647         ret = BTRFS_LEAF_DATA_SIZE(fs_info) - leaf_space_used(leaf, 0, nritems);
2648         if (ret < 0) {
2649                 btrfs_crit(fs_info,
2650                            "leaf free space ret %d, leaf data size %lu, used %d nritems %d",
2651                            ret,
2652                            (unsigned long) BTRFS_LEAF_DATA_SIZE(fs_info),
2653                            leaf_space_used(leaf, 0, nritems), nritems);
2654         }
2655         return ret;
2656 }
2657
2658 /*
2659  * min slot controls the lowest index we're willing to push to the
2660  * right.  We'll push up to and including min_slot, but no lower
2661  */
2662 static noinline int __push_leaf_right(struct btrfs_path *path,
2663                                       int data_size, int empty,
2664                                       struct extent_buffer *right,
2665                                       int free_space, u32 left_nritems,
2666                                       u32 min_slot)
2667 {
2668         struct btrfs_fs_info *fs_info = right->fs_info;
2669         struct extent_buffer *left = path->nodes[0];
2670         struct extent_buffer *upper = path->nodes[1];
2671         struct btrfs_map_token token;
2672         struct btrfs_disk_key disk_key;
2673         int slot;
2674         u32 i;
2675         int push_space = 0;
2676         int push_items = 0;
2677         struct btrfs_item *item;
2678         u32 nr;
2679         u32 right_nritems;
2680         u32 data_end;
2681         u32 this_item_size;
2682
2683         if (empty)
2684                 nr = 0;
2685         else
2686                 nr = max_t(u32, 1, min_slot);
2687
2688         if (path->slots[0] >= left_nritems)
2689                 push_space += data_size;
2690
2691         slot = path->slots[1];
2692         i = left_nritems - 1;
2693         while (i >= nr) {
2694                 item = btrfs_item_nr(i);
2695
2696                 if (!empty && push_items > 0) {
2697                         if (path->slots[0] > i)
2698                                 break;
2699                         if (path->slots[0] == i) {
2700                                 int space = btrfs_leaf_free_space(left);
2701
2702                                 if (space + push_space * 2 > free_space)
2703                                         break;
2704                         }
2705                 }
2706
2707                 if (path->slots[0] == i)
2708                         push_space += data_size;
2709
2710                 this_item_size = btrfs_item_size(left, item);
2711                 if (this_item_size + sizeof(*item) + push_space > free_space)
2712                         break;
2713
2714                 push_items++;
2715                 push_space += this_item_size + sizeof(*item);
2716                 if (i == 0)
2717                         break;
2718                 i--;
2719         }
2720
2721         if (push_items == 0)
2722                 goto out_unlock;
2723
2724         WARN_ON(!empty && push_items == left_nritems);
2725
2726         /* push left to right */
2727         right_nritems = btrfs_header_nritems(right);
2728
2729         push_space = btrfs_item_end_nr(left, left_nritems - push_items);
2730         push_space -= leaf_data_end(left);
2731
2732         /* make room in the right data area */
2733         data_end = leaf_data_end(right);
2734         memmove_extent_buffer(right,
2735                               BTRFS_LEAF_DATA_OFFSET + data_end - push_space,
2736                               BTRFS_LEAF_DATA_OFFSET + data_end,
2737                               BTRFS_LEAF_DATA_SIZE(fs_info) - data_end);
2738
2739         /* copy from the left data area */
2740         copy_extent_buffer(right, left, BTRFS_LEAF_DATA_OFFSET +
2741                      BTRFS_LEAF_DATA_SIZE(fs_info) - push_space,
2742                      BTRFS_LEAF_DATA_OFFSET + leaf_data_end(left),
2743                      push_space);
2744
2745         memmove_extent_buffer(right, btrfs_item_nr_offset(push_items),
2746                               btrfs_item_nr_offset(0),
2747                               right_nritems * sizeof(struct btrfs_item));
2748
2749         /* copy the items from left to right */
2750         copy_extent_buffer(right, left, btrfs_item_nr_offset(0),
2751                    btrfs_item_nr_offset(left_nritems - push_items),
2752                    push_items * sizeof(struct btrfs_item));
2753
2754         /* update the item pointers */
2755         btrfs_init_map_token(&token, right);
2756         right_nritems += push_items;
2757         btrfs_set_header_nritems(right, right_nritems);
2758         push_space = BTRFS_LEAF_DATA_SIZE(fs_info);
2759         for (i = 0; i < right_nritems; i++) {
2760                 item = btrfs_item_nr(i);
2761                 push_space -= btrfs_token_item_size(&token, item);
2762                 btrfs_set_token_item_offset(&token, item, push_space);
2763         }
2764
2765         left_nritems -= push_items;
2766         btrfs_set_header_nritems(left, left_nritems);
2767
2768         if (left_nritems)
2769                 btrfs_mark_buffer_dirty(left);
2770         else
2771                 btrfs_clean_tree_block(left);
2772
2773         btrfs_mark_buffer_dirty(right);
2774
2775         btrfs_item_key(right, &disk_key, 0);
2776         btrfs_set_node_key(upper, &disk_key, slot + 1);
2777         btrfs_mark_buffer_dirty(upper);
2778
2779         /* then fixup the leaf pointer in the path */
2780         if (path->slots[0] >= left_nritems) {
2781                 path->slots[0] -= left_nritems;
2782                 if (btrfs_header_nritems(path->nodes[0]) == 0)
2783                         btrfs_clean_tree_block(path->nodes[0]);
2784                 btrfs_tree_unlock(path->nodes[0]);
2785                 free_extent_buffer(path->nodes[0]);
2786                 path->nodes[0] = right;
2787                 path->slots[1] += 1;
2788         } else {
2789                 btrfs_tree_unlock(right);
2790                 free_extent_buffer(right);
2791         }
2792         return 0;
2793
2794 out_unlock:
2795         btrfs_tree_unlock(right);
2796         free_extent_buffer(right);
2797         return 1;
2798 }
2799
2800 /*
2801  * push some data in the path leaf to the right, trying to free up at
2802  * least data_size bytes.  returns zero if the push worked, nonzero otherwise
2803  *
2804  * returns 1 if the push failed because the other node didn't have enough
2805  * room, 0 if everything worked out and < 0 if there were major errors.
2806  *
2807  * this will push starting from min_slot to the end of the leaf.  It won't
2808  * push any slot lower than min_slot
2809  */
2810 static int push_leaf_right(struct btrfs_trans_handle *trans, struct btrfs_root
2811                            *root, struct btrfs_path *path,
2812                            int min_data_size, int data_size,
2813                            int empty, u32 min_slot)
2814 {
2815         struct extent_buffer *left = path->nodes[0];
2816         struct extent_buffer *right;
2817         struct extent_buffer *upper;
2818         int slot;
2819         int free_space;
2820         u32 left_nritems;
2821         int ret;
2822
2823         if (!path->nodes[1])
2824                 return 1;
2825
2826         slot = path->slots[1];
2827         upper = path->nodes[1];
2828         if (slot >= btrfs_header_nritems(upper) - 1)
2829                 return 1;
2830
2831         btrfs_assert_tree_write_locked(path->nodes[1]);
2832
2833         right = btrfs_read_node_slot(upper, slot + 1);
2834         /*
2835          * slot + 1 is not valid or we fail to read the right node,
2836          * no big deal, just return.
2837          */
2838         if (IS_ERR(right))
2839                 return 1;
2840
2841         __btrfs_tree_lock(right, BTRFS_NESTING_RIGHT);
2842
2843         free_space = btrfs_leaf_free_space(right);
2844         if (free_space < data_size)
2845                 goto out_unlock;
2846
2847         /* cow and double check */
2848         ret = btrfs_cow_block(trans, root, right, upper,
2849                               slot + 1, &right, BTRFS_NESTING_RIGHT_COW);
2850         if (ret)
2851                 goto out_unlock;
2852
2853         free_space = btrfs_leaf_free_space(right);
2854         if (free_space < data_size)
2855                 goto out_unlock;
2856
2857         left_nritems = btrfs_header_nritems(left);
2858         if (left_nritems == 0)
2859                 goto out_unlock;
2860
2861         if (check_sibling_keys(left, right)) {
2862                 ret = -EUCLEAN;
2863                 btrfs_tree_unlock(right);
2864                 free_extent_buffer(right);
2865                 return ret;
2866         }
2867         if (path->slots[0] == left_nritems && !empty) {
2868                 /* Key greater than all keys in the leaf, right neighbor has
2869                  * enough room for it and we're not emptying our leaf to delete
2870                  * it, therefore use right neighbor to insert the new item and
2871                  * no need to touch/dirty our left leaf. */
2872                 btrfs_tree_unlock(left);
2873                 free_extent_buffer(left);
2874                 path->nodes[0] = right;
2875                 path->slots[0] = 0;
2876                 path->slots[1]++;
2877                 return 0;
2878         }
2879
2880         return __push_leaf_right(path, min_data_size, empty,
2881                                 right, free_space, left_nritems, min_slot);
2882 out_unlock:
2883         btrfs_tree_unlock(right);
2884         free_extent_buffer(right);
2885         return 1;
2886 }
2887
2888 /*
2889  * push some data in the path leaf to the left, trying to free up at
2890  * least data_size bytes.  returns zero if the push worked, nonzero otherwise
2891  *
2892  * max_slot can put a limit on how far into the leaf we'll push items.  The
2893  * item at 'max_slot' won't be touched.  Use (u32)-1 to make us do all the
2894  * items
2895  */
2896 static noinline int __push_leaf_left(struct btrfs_path *path, int data_size,
2897                                      int empty, struct extent_buffer *left,
2898                                      int free_space, u32 right_nritems,
2899                                      u32 max_slot)
2900 {
2901         struct btrfs_fs_info *fs_info = left->fs_info;
2902         struct btrfs_disk_key disk_key;
2903         struct extent_buffer *right = path->nodes[0];
2904         int i;
2905         int push_space = 0;
2906         int push_items = 0;
2907         struct btrfs_item *item;
2908         u32 old_left_nritems;
2909         u32 nr;
2910         int ret = 0;
2911         u32 this_item_size;
2912         u32 old_left_item_size;
2913         struct btrfs_map_token token;
2914
2915         if (empty)
2916                 nr = min(right_nritems, max_slot);
2917         else
2918                 nr = min(right_nritems - 1, max_slot);
2919
2920         for (i = 0; i < nr; i++) {
2921                 item = btrfs_item_nr(i);
2922
2923                 if (!empty && push_items > 0) {
2924                         if (path->slots[0] < i)
2925                                 break;
2926                         if (path->slots[0] == i) {
2927                                 int space = btrfs_leaf_free_space(right);
2928
2929                                 if (space + push_space * 2 > free_space)
2930                                         break;
2931                         }
2932                 }
2933
2934                 if (path->slots[0] == i)
2935                         push_space += data_size;
2936
2937                 this_item_size = btrfs_item_size(right, item);
2938                 if (this_item_size + sizeof(*item) + push_space > free_space)
2939                         break;
2940
2941                 push_items++;
2942                 push_space += this_item_size + sizeof(*item);
2943         }
2944
2945         if (push_items == 0) {
2946                 ret = 1;
2947                 goto out;
2948         }
2949         WARN_ON(!empty && push_items == btrfs_header_nritems(right));
2950
2951         /* push data from right to left */
2952         copy_extent_buffer(left, right,
2953                            btrfs_item_nr_offset(btrfs_header_nritems(left)),
2954                            btrfs_item_nr_offset(0),
2955                            push_items * sizeof(struct btrfs_item));
2956
2957         push_space = BTRFS_LEAF_DATA_SIZE(fs_info) -
2958                      btrfs_item_offset_nr(right, push_items - 1);
2959
2960         copy_extent_buffer(left, right, BTRFS_LEAF_DATA_OFFSET +
2961                      leaf_data_end(left) - push_space,
2962                      BTRFS_LEAF_DATA_OFFSET +
2963                      btrfs_item_offset_nr(right, push_items - 1),
2964                      push_space);
2965         old_left_nritems = btrfs_header_nritems(left);
2966         BUG_ON(old_left_nritems <= 0);
2967
2968         btrfs_init_map_token(&token, left);
2969         old_left_item_size = btrfs_item_offset_nr(left, old_left_nritems - 1);
2970         for (i = old_left_nritems; i < old_left_nritems + push_items; i++) {
2971                 u32 ioff;
2972
2973                 item = btrfs_item_nr(i);
2974
2975                 ioff = btrfs_token_item_offset(&token, item);
2976                 btrfs_set_token_item_offset(&token, item,
2977                       ioff - (BTRFS_LEAF_DATA_SIZE(fs_info) - old_left_item_size));
2978         }
2979         btrfs_set_header_nritems(left, old_left_nritems + push_items);
2980
2981         /* fixup right node */
2982         if (push_items > right_nritems)
2983                 WARN(1, KERN_CRIT "push items %d nr %u\n", push_items,
2984                        right_nritems);
2985
2986         if (push_items < right_nritems) {
2987                 push_space = btrfs_item_offset_nr(right, push_items - 1) -
2988                                                   leaf_data_end(right);
2989                 memmove_extent_buffer(right, BTRFS_LEAF_DATA_OFFSET +
2990                                       BTRFS_LEAF_DATA_SIZE(fs_info) - push_space,
2991                                       BTRFS_LEAF_DATA_OFFSET +
2992                                       leaf_data_end(right), push_space);
2993
2994                 memmove_extent_buffer(right, btrfs_item_nr_offset(0),
2995                               btrfs_item_nr_offset(push_items),
2996                              (btrfs_header_nritems(right) - push_items) *
2997                              sizeof(struct btrfs_item));
2998         }
2999
3000         btrfs_init_map_token(&token, right);
3001         right_nritems -= push_items;
3002         btrfs_set_header_nritems(right, right_nritems);
3003         push_space = BTRFS_LEAF_DATA_SIZE(fs_info);
3004         for (i = 0; i < right_nritems; i++) {
3005                 item = btrfs_item_nr(i);
3006
3007                 push_space = push_space - btrfs_token_item_size(&token, item);
3008                 btrfs_set_token_item_offset(&token, item, push_space);
3009         }
3010
3011         btrfs_mark_buffer_dirty(left);
3012         if (right_nritems)
3013                 btrfs_mark_buffer_dirty(right);
3014         else
3015                 btrfs_clean_tree_block(right);
3016
3017         btrfs_item_key(right, &disk_key, 0);
3018         fixup_low_keys(path, &disk_key, 1);
3019
3020         /* then fixup the leaf pointer in the path */
3021         if (path->slots[0] < push_items) {
3022                 path->slots[0] += old_left_nritems;
3023                 btrfs_tree_unlock(path->nodes[0]);
3024                 free_extent_buffer(path->nodes[0]);
3025                 path->nodes[0] = left;
3026                 path->slots[1] -= 1;
3027         } else {
3028                 btrfs_tree_unlock(left);
3029                 free_extent_buffer(left);
3030                 path->slots[0] -= push_items;
3031         }
3032         BUG_ON(path->slots[0] < 0);
3033         return ret;
3034 out:
3035         btrfs_tree_unlock(left);
3036         free_extent_buffer(left);
3037         return ret;
3038 }
3039
3040 /*
3041  * push some data in the path leaf to the left, trying to free up at
3042  * least data_size bytes.  returns zero if the push worked, nonzero otherwise
3043  *
3044  * max_slot can put a limit on how far into the leaf we'll push items.  The
3045  * item at 'max_slot' won't be touched.  Use (u32)-1 to make us push all the
3046  * items
3047  */
3048 static int push_leaf_left(struct btrfs_trans_handle *trans, struct btrfs_root
3049                           *root, struct btrfs_path *path, int min_data_size,
3050                           int data_size, int empty, u32 max_slot)
3051 {
3052         struct extent_buffer *right = path->nodes[0];
3053         struct extent_buffer *left;
3054         int slot;
3055         int free_space;
3056         u32 right_nritems;
3057         int ret = 0;
3058
3059         slot = path->slots[1];
3060         if (slot == 0)
3061                 return 1;
3062         if (!path->nodes[1])
3063                 return 1;
3064
3065         right_nritems = btrfs_header_nritems(right);
3066         if (right_nritems == 0)
3067                 return 1;
3068
3069         btrfs_assert_tree_write_locked(path->nodes[1]);
3070
3071         left = btrfs_read_node_slot(path->nodes[1], slot - 1);
3072         /*
3073          * slot - 1 is not valid or we fail to read the left node,
3074          * no big deal, just return.
3075          */
3076         if (IS_ERR(left))
3077                 return 1;
3078
3079         __btrfs_tree_lock(left, BTRFS_NESTING_LEFT);
3080
3081         free_space = btrfs_leaf_free_space(left);
3082         if (free_space < data_size) {
3083                 ret = 1;
3084                 goto out;
3085         }
3086
3087         /* cow and double check */
3088         ret = btrfs_cow_block(trans, root, left,
3089                               path->nodes[1], slot - 1, &left,
3090                               BTRFS_NESTING_LEFT_COW);
3091         if (ret) {
3092                 /* we hit -ENOSPC, but it isn't fatal here */
3093                 if (ret == -ENOSPC)
3094                         ret = 1;
3095                 goto out;
3096         }
3097
3098         free_space = btrfs_leaf_free_space(left);
3099         if (free_space < data_size) {
3100                 ret = 1;
3101                 goto out;
3102         }
3103
3104         if (check_sibling_keys(left, right)) {
3105                 ret = -EUCLEAN;
3106                 goto out;
3107         }
3108         return __push_leaf_left(path, min_data_size,
3109                                empty, left, free_space, right_nritems,
3110                                max_slot);
3111 out:
3112         btrfs_tree_unlock(left);
3113         free_extent_buffer(left);
3114         return ret;
3115 }
3116
3117 /*
3118  * split the path's leaf in two, making sure there is at least data_size
3119  * available for the resulting leaf level of the path.
3120  */
3121 static noinline void copy_for_split(struct btrfs_trans_handle *trans,
3122                                     struct btrfs_path *path,
3123                                     struct extent_buffer *l,
3124                                     struct extent_buffer *right,
3125                                     int slot, int mid, int nritems)
3126 {
3127         struct btrfs_fs_info *fs_info = trans->fs_info;
3128         int data_copy_size;
3129         int rt_data_off;
3130         int i;
3131         struct btrfs_disk_key disk_key;
3132         struct btrfs_map_token token;
3133
3134         nritems = nritems - mid;
3135         btrfs_set_header_nritems(right, nritems);
3136         data_copy_size = btrfs_item_end_nr(l, mid) - leaf_data_end(l);
3137
3138         copy_extent_buffer(right, l, btrfs_item_nr_offset(0),
3139                            btrfs_item_nr_offset(mid),
3140                            nritems * sizeof(struct btrfs_item));
3141
3142         copy_extent_buffer(right, l,
3143                      BTRFS_LEAF_DATA_OFFSET + BTRFS_LEAF_DATA_SIZE(fs_info) -
3144                      data_copy_size, BTRFS_LEAF_DATA_OFFSET +
3145                      leaf_data_end(l), data_copy_size);
3146
3147         rt_data_off = BTRFS_LEAF_DATA_SIZE(fs_info) - btrfs_item_end_nr(l, mid);
3148
3149         btrfs_init_map_token(&token, right);
3150         for (i = 0; i < nritems; i++) {
3151                 struct btrfs_item *item = btrfs_item_nr(i);
3152                 u32 ioff;
3153
3154                 ioff = btrfs_token_item_offset(&token, item);
3155                 btrfs_set_token_item_offset(&token, item, ioff + rt_data_off);
3156         }
3157
3158         btrfs_set_header_nritems(l, mid);
3159         btrfs_item_key(right, &disk_key, 0);
3160         insert_ptr(trans, path, &disk_key, right->start, path->slots[1] + 1, 1);
3161
3162         btrfs_mark_buffer_dirty(right);
3163         btrfs_mark_buffer_dirty(l);
3164         BUG_ON(path->slots[0] != slot);
3165
3166         if (mid <= slot) {
3167                 btrfs_tree_unlock(path->nodes[0]);
3168                 free_extent_buffer(path->nodes[0]);
3169                 path->nodes[0] = right;
3170                 path->slots[0] -= mid;
3171                 path->slots[1] += 1;
3172         } else {
3173                 btrfs_tree_unlock(right);
3174                 free_extent_buffer(right);
3175         }
3176
3177         BUG_ON(path->slots[0] < 0);
3178 }
3179
3180 /*
3181  * double splits happen when we need to insert a big item in the middle
3182  * of a leaf.  A double split can leave us with 3 mostly empty leaves:
3183  * leaf: [ slots 0 - N] [ our target ] [ N + 1 - total in leaf ]
3184  *          A                 B                 C
3185  *
3186  * We avoid this by trying to push the items on either side of our target
3187  * into the adjacent leaves.  If all goes well we can avoid the double split
3188  * completely.
3189  */
3190 static noinline int push_for_double_split(struct btrfs_trans_handle *trans,
3191                                           struct btrfs_root *root,
3192                                           struct btrfs_path *path,
3193                                           int data_size)
3194 {
3195         int ret;
3196         int progress = 0;
3197         int slot;
3198         u32 nritems;
3199         int space_needed = data_size;
3200
3201         slot = path->slots[0];
3202         if (slot < btrfs_header_nritems(path->nodes[0]))
3203                 space_needed -= btrfs_leaf_free_space(path->nodes[0]);
3204
3205         /*
3206          * try to push all the items after our slot into the
3207          * right leaf
3208          */
3209         ret = push_leaf_right(trans, root, path, 1, space_needed, 0, slot);
3210         if (ret < 0)
3211                 return ret;
3212
3213         if (ret == 0)
3214                 progress++;
3215
3216         nritems = btrfs_header_nritems(path->nodes[0]);
3217         /*
3218          * our goal is to get our slot at the start or end of a leaf.  If
3219          * we've done so we're done
3220          */
3221         if (path->slots[0] == 0 || path->slots[0] == nritems)
3222                 return 0;
3223
3224         if (btrfs_leaf_free_space(path->nodes[0]) >= data_size)
3225                 return 0;
3226
3227         /* try to push all the items before our slot into the next leaf */
3228         slot = path->slots[0];
3229         space_needed = data_size;
3230         if (slot > 0)
3231                 space_needed -= btrfs_leaf_free_space(path->nodes[0]);
3232         ret = push_leaf_left(trans, root, path, 1, space_needed, 0, slot);
3233         if (ret < 0)
3234                 return ret;
3235
3236         if (ret == 0)
3237                 progress++;
3238
3239         if (progress)
3240                 return 0;
3241         return 1;
3242 }
3243
3244 /*
3245  * split the path's leaf in two, making sure there is at least data_size
3246  * available for the resulting leaf level of the path.
3247  *
3248  * returns 0 if all went well and < 0 on failure.
3249  */
3250 static noinline int split_leaf(struct btrfs_trans_handle *trans,
3251                                struct btrfs_root *root,
3252                                const struct btrfs_key *ins_key,
3253                                struct btrfs_path *path, int data_size,
3254                                int extend)
3255 {
3256         struct btrfs_disk_key disk_key;
3257         struct extent_buffer *l;
3258         u32 nritems;
3259         int mid;
3260         int slot;
3261         struct extent_buffer *right;
3262         struct btrfs_fs_info *fs_info = root->fs_info;
3263         int ret = 0;
3264         int wret;
3265         int split;
3266         int num_doubles = 0;
3267         int tried_avoid_double = 0;
3268
3269         l = path->nodes[0];
3270         slot = path->slots[0];
3271         if (extend && data_size + btrfs_item_size_nr(l, slot) +
3272             sizeof(struct btrfs_item) > BTRFS_LEAF_DATA_SIZE(fs_info))
3273                 return -EOVERFLOW;
3274
3275         /* first try to make some room by pushing left and right */
3276         if (data_size && path->nodes[1]) {
3277                 int space_needed = data_size;
3278
3279                 if (slot < btrfs_header_nritems(l))
3280                         space_needed -= btrfs_leaf_free_space(l);
3281
3282                 wret = push_leaf_right(trans, root, path, space_needed,
3283                                        space_needed, 0, 0);
3284                 if (wret < 0)
3285                         return wret;
3286                 if (wret) {
3287                         space_needed = data_size;
3288                         if (slot > 0)
3289                                 space_needed -= btrfs_leaf_free_space(l);
3290                         wret = push_leaf_left(trans, root, path, space_needed,
3291                                               space_needed, 0, (u32)-1);
3292                         if (wret < 0)
3293                                 return wret;
3294                 }
3295                 l = path->nodes[0];
3296
3297                 /* did the pushes work? */
3298                 if (btrfs_leaf_free_space(l) >= data_size)
3299                         return 0;
3300         }
3301
3302         if (!path->nodes[1]) {
3303                 ret = insert_new_root(trans, root, path, 1);
3304                 if (ret)
3305                         return ret;
3306         }
3307 again:
3308         split = 1;
3309         l = path->nodes[0];
3310         slot = path->slots[0];
3311         nritems = btrfs_header_nritems(l);
3312         mid = (nritems + 1) / 2;
3313
3314         if (mid <= slot) {
3315                 if (nritems == 1 ||
3316                     leaf_space_used(l, mid, nritems - mid) + data_size >
3317                         BTRFS_LEAF_DATA_SIZE(fs_info)) {
3318                         if (slot >= nritems) {
3319                                 split = 0;
3320                         } else {
3321                                 mid = slot;
3322                                 if (mid != nritems &&
3323                                     leaf_space_used(l, mid, nritems - mid) +
3324                                     data_size > BTRFS_LEAF_DATA_SIZE(fs_info)) {
3325                                         if (data_size && !tried_avoid_double)
3326                                                 goto push_for_double;
3327                                         split = 2;
3328                                 }
3329                         }
3330                 }
3331         } else {
3332                 if (leaf_space_used(l, 0, mid) + data_size >
3333                         BTRFS_LEAF_DATA_SIZE(fs_info)) {
3334                         if (!extend && data_size && slot == 0) {
3335                                 split = 0;
3336                         } else if ((extend || !data_size) && slot == 0) {
3337                                 mid = 1;
3338                         } else {
3339                                 mid = slot;
3340                                 if (mid != nritems &&
3341                                     leaf_space_used(l, mid, nritems - mid) +
3342                                     data_size > BTRFS_LEAF_DATA_SIZE(fs_info)) {
3343                                         if (data_size && !tried_avoid_double)
3344                                                 goto push_for_double;
3345                                         split = 2;
3346                                 }
3347                         }
3348                 }
3349         }
3350
3351         if (split == 0)
3352                 btrfs_cpu_key_to_disk(&disk_key, ins_key);
3353         else
3354                 btrfs_item_key(l, &disk_key, mid);
3355
3356         /*
3357          * We have to about BTRFS_NESTING_NEW_ROOT here if we've done a double
3358          * split, because we're only allowed to have MAX_LOCKDEP_SUBCLASSES
3359          * subclasses, which is 8 at the time of this patch, and we've maxed it
3360          * out.  In the future we could add a
3361          * BTRFS_NESTING_SPLIT_THE_SPLITTENING if we need to, but for now just
3362          * use BTRFS_NESTING_NEW_ROOT.
3363          */
3364         right = btrfs_alloc_tree_block(trans, root, 0, root->root_key.objectid,
3365                                        &disk_key, 0, l->start, 0,
3366                                        num_doubles ? BTRFS_NESTING_NEW_ROOT :
3367                                        BTRFS_NESTING_SPLIT);
3368         if (IS_ERR(right))
3369                 return PTR_ERR(right);
3370
3371         root_add_used(root, fs_info->nodesize);
3372
3373         if (split == 0) {
3374                 if (mid <= slot) {
3375                         btrfs_set_header_nritems(right, 0);
3376                         insert_ptr(trans, path, &disk_key,
3377                                    right->start, path->slots[1] + 1, 1);
3378                         btrfs_tree_unlock(path->nodes[0]);
3379                         free_extent_buffer(path->nodes[0]);
3380                         path->nodes[0] = right;
3381                         path->slots[0] = 0;
3382                         path->slots[1] += 1;
3383                 } else {
3384                         btrfs_set_header_nritems(right, 0);
3385                         insert_ptr(trans, path, &disk_key,
3386                                    right->start, path->slots[1], 1);
3387                         btrfs_tree_unlock(path->nodes[0]);
3388                         free_extent_buffer(path->nodes[0]);
3389                         path->nodes[0] = right;
3390                         path->slots[0] = 0;
3391                         if (path->slots[1] == 0)
3392                                 fixup_low_keys(path, &disk_key, 1);
3393                 }
3394                 /*
3395                  * We create a new leaf 'right' for the required ins_len and
3396                  * we'll do btrfs_mark_buffer_dirty() on this leaf after copying
3397                  * the content of ins_len to 'right'.
3398                  */
3399                 return ret;
3400         }
3401
3402         copy_for_split(trans, path, l, right, slot, mid, nritems);
3403
3404         if (split == 2) {
3405                 BUG_ON(num_doubles != 0);
3406                 num_doubles++;
3407                 goto again;
3408         }
3409
3410         return 0;
3411
3412 push_for_double:
3413         push_for_double_split(trans, root, path, data_size);
3414         tried_avoid_double = 1;
3415         if (btrfs_leaf_free_space(path->nodes[0]) >= data_size)
3416                 return 0;
3417         goto again;
3418 }
3419
3420 static noinline int setup_leaf_for_split(struct btrfs_trans_handle *trans,
3421                                          struct btrfs_root *root,
3422                                          struct btrfs_path *path, int ins_len)
3423 {
3424         struct btrfs_key key;
3425         struct extent_buffer *leaf;
3426         struct btrfs_file_extent_item *fi;
3427         u64 extent_len = 0;
3428         u32 item_size;
3429         int ret;
3430
3431         leaf = path->nodes[0];
3432         btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
3433
3434         BUG_ON(key.type != BTRFS_EXTENT_DATA_KEY &&
3435                key.type != BTRFS_EXTENT_CSUM_KEY);
3436
3437         if (btrfs_leaf_free_space(leaf) >= ins_len)
3438                 return 0;
3439
3440         item_size = btrfs_item_size_nr(leaf, path->slots[0]);
3441         if (key.type == BTRFS_EXTENT_DATA_KEY) {
3442                 fi = btrfs_item_ptr(leaf, path->slots[0],
3443                                     struct btrfs_file_extent_item);
3444                 extent_len = btrfs_file_extent_num_bytes(leaf, fi);
3445         }
3446         btrfs_release_path(path);
3447
3448         path->keep_locks = 1;
3449         path->search_for_split = 1;
3450         ret = btrfs_search_slot(trans, root, &key, path, 0, 1);
3451         path->search_for_split = 0;
3452         if (ret > 0)
3453                 ret = -EAGAIN;
3454         if (ret < 0)
3455                 goto err;
3456
3457         ret = -EAGAIN;
3458         leaf = path->nodes[0];
3459         /* if our item isn't there, return now */
3460         if (item_size != btrfs_item_size_nr(leaf, path->slots[0]))
3461                 goto err;
3462
3463         /* the leaf has  changed, it now has room.  return now */
3464         if (btrfs_leaf_free_space(path->nodes[0]) >= ins_len)
3465                 goto err;
3466
3467         if (key.type == BTRFS_EXTENT_DATA_KEY) {
3468                 fi = btrfs_item_ptr(leaf, path->slots[0],
3469                                     struct btrfs_file_extent_item);
3470                 if (extent_len != btrfs_file_extent_num_bytes(leaf, fi))
3471                         goto err;
3472         }
3473
3474         ret = split_leaf(trans, root, &key, path, ins_len, 1);
3475         if (ret)
3476                 goto err;
3477
3478         path->keep_locks = 0;
3479         btrfs_unlock_up_safe(path, 1);
3480         return 0;
3481 err:
3482         path->keep_locks = 0;
3483         return ret;
3484 }
3485
3486 static noinline int split_item(struct btrfs_path *path,
3487                                const struct btrfs_key *new_key,
3488                                unsigned long split_offset)
3489 {
3490         struct extent_buffer *leaf;
3491         struct btrfs_item *item;
3492         struct btrfs_item *new_item;
3493         int slot;
3494         char *buf;
3495         u32 nritems;
3496         u32 item_size;
3497         u32 orig_offset;
3498         struct btrfs_disk_key disk_key;
3499
3500         leaf = path->nodes[0];
3501         BUG_ON(btrfs_leaf_free_space(leaf) < sizeof(struct btrfs_item));
3502
3503         item = btrfs_item_nr(path->slots[0]);
3504         orig_offset = btrfs_item_offset(leaf, item);
3505         item_size = btrfs_item_size(leaf, item);
3506
3507         buf = kmalloc(item_size, GFP_NOFS);
3508         if (!buf)
3509                 return -ENOMEM;
3510
3511         read_extent_buffer(leaf, buf, btrfs_item_ptr_offset(leaf,
3512                             path->slots[0]), item_size);
3513
3514         slot = path->slots[0] + 1;
3515         nritems = btrfs_header_nritems(leaf);
3516         if (slot != nritems) {
3517                 /* shift the items */
3518                 memmove_extent_buffer(leaf, btrfs_item_nr_offset(slot + 1),
3519                                 btrfs_item_nr_offset(slot),
3520                                 (nritems - slot) * sizeof(struct btrfs_item));
3521         }
3522
3523         btrfs_cpu_key_to_disk(&disk_key, new_key);
3524         btrfs_set_item_key(leaf, &disk_key, slot);
3525
3526         new_item = btrfs_item_nr(slot);
3527
3528         btrfs_set_item_offset(leaf, new_item, orig_offset);
3529         btrfs_set_item_size(leaf, new_item, item_size - split_offset);
3530
3531         btrfs_set_item_offset(leaf, item,
3532                               orig_offset + item_size - split_offset);
3533         btrfs_set_item_size(leaf, item, split_offset);
3534
3535         btrfs_set_header_nritems(leaf, nritems + 1);
3536
3537         /* write the data for the start of the original item */
3538         write_extent_buffer(leaf, buf,
3539                             btrfs_item_ptr_offset(leaf, path->slots[0]),
3540                             split_offset);
3541
3542         /* write the data for the new item */
3543         write_extent_buffer(leaf, buf + split_offset,
3544                             btrfs_item_ptr_offset(leaf, slot),
3545                             item_size - split_offset);
3546         btrfs_mark_buffer_dirty(leaf);
3547
3548         BUG_ON(btrfs_leaf_free_space(leaf) < 0);
3549         kfree(buf);
3550         return 0;
3551 }
3552
3553 /*
3554  * This function splits a single item into two items,
3555  * giving 'new_key' to the new item and splitting the
3556  * old one at split_offset (from the start of the item).
3557  *
3558  * The path may be released by this operation.  After
3559  * the split, the path is pointing to the old item.  The
3560  * new item is going to be in the same node as the old one.
3561  *
3562  * Note, the item being split must be smaller enough to live alone on
3563  * a tree block with room for one extra struct btrfs_item
3564  *
3565  * This allows us to split the item in place, keeping a lock on the
3566  * leaf the entire time.
3567  */
3568 int btrfs_split_item(struct btrfs_trans_handle *trans,
3569                      struct btrfs_root *root,
3570                      struct btrfs_path *path,
3571                      const struct btrfs_key *new_key,
3572                      unsigned long split_offset)
3573 {
3574         int ret;
3575         ret = setup_leaf_for_split(trans, root, path,
3576                                    sizeof(struct btrfs_item));
3577         if (ret)
3578                 return ret;
3579
3580         ret = split_item(path, new_key, split_offset);
3581         return ret;
3582 }
3583
3584 /*
3585  * make the item pointed to by the path smaller.  new_size indicates
3586  * how small to make it, and from_end tells us if we just chop bytes
3587  * off the end of the item or if we shift the item to chop bytes off
3588  * the front.
3589  */
3590 void btrfs_truncate_item(struct btrfs_path *path, u32 new_size, int from_end)
3591 {
3592         int slot;
3593         struct extent_buffer *leaf;
3594         struct btrfs_item *item;
3595         u32 nritems;
3596         unsigned int data_end;
3597         unsigned int old_data_start;
3598         unsigned int old_size;
3599         unsigned int size_diff;
3600         int i;
3601         struct btrfs_map_token token;
3602
3603         leaf = path->nodes[0];
3604         slot = path->slots[0];
3605
3606         old_size = btrfs_item_size_nr(leaf, slot);
3607         if (old_size == new_size)
3608                 return;
3609
3610         nritems = btrfs_header_nritems(leaf);
3611         data_end = leaf_data_end(leaf);
3612
3613         old_data_start = btrfs_item_offset_nr(leaf, slot);
3614
3615         size_diff = old_size - new_size;
3616
3617         BUG_ON(slot < 0);
3618         BUG_ON(slot >= nritems);
3619
3620         /*
3621          * item0..itemN ... dataN.offset..dataN.size .. data0.size
3622          */
3623         /* first correct the data pointers */
3624         btrfs_init_map_token(&token, leaf);
3625         for (i = slot; i < nritems; i++) {
3626                 u32 ioff;
3627                 item = btrfs_item_nr(i);
3628
3629                 ioff = btrfs_token_item_offset(&token, item);
3630                 btrfs_set_token_item_offset(&token, item, ioff + size_diff);
3631         }
3632
3633         /* shift the data */
3634         if (from_end) {
3635                 memmove_extent_buffer(leaf, BTRFS_LEAF_DATA_OFFSET +
3636                               data_end + size_diff, BTRFS_LEAF_DATA_OFFSET +
3637                               data_end, old_data_start + new_size - data_end);
3638         } else {
3639                 struct btrfs_disk_key disk_key;
3640                 u64 offset;
3641
3642                 btrfs_item_key(leaf, &disk_key, slot);
3643
3644                 if (btrfs_disk_key_type(&disk_key) == BTRFS_EXTENT_DATA_KEY) {
3645                         unsigned long ptr;
3646                         struct btrfs_file_extent_item *fi;
3647
3648                         fi = btrfs_item_ptr(leaf, slot,
3649                                             struct btrfs_file_extent_item);
3650                         fi = (struct btrfs_file_extent_item *)(
3651                              (unsigned long)fi - size_diff);
3652
3653                         if (btrfs_file_extent_type(leaf, fi) ==
3654                             BTRFS_FILE_EXTENT_INLINE) {
3655                                 ptr = btrfs_item_ptr_offset(leaf, slot);
3656                                 memmove_extent_buffer(leaf, ptr,
3657                                       (unsigned long)fi,
3658                                       BTRFS_FILE_EXTENT_INLINE_DATA_START);
3659                         }
3660                 }
3661
3662                 memmove_extent_buffer(leaf, BTRFS_LEAF_DATA_OFFSET +
3663                               data_end + size_diff, BTRFS_LEAF_DATA_OFFSET +
3664                               data_end, old_data_start - data_end);
3665
3666                 offset = btrfs_disk_key_offset(&disk_key);
3667                 btrfs_set_disk_key_offset(&disk_key, offset + size_diff);
3668                 btrfs_set_item_key(leaf, &disk_key, slot);
3669                 if (slot == 0)
3670                         fixup_low_keys(path, &disk_key, 1);
3671         }
3672
3673         item = btrfs_item_nr(slot);
3674         btrfs_set_item_size(leaf, item, new_size);
3675         btrfs_mark_buffer_dirty(leaf);
3676
3677         if (btrfs_leaf_free_space(leaf) < 0) {
3678                 btrfs_print_leaf(leaf);
3679                 BUG();
3680         }
3681 }
3682
3683 /*
3684  * make the item pointed to by the path bigger, data_size is the added size.
3685  */
3686 void btrfs_extend_item(struct btrfs_path *path, u32 data_size)
3687 {
3688         int slot;
3689         struct extent_buffer *leaf;
3690         struct btrfs_item *item;
3691         u32 nritems;
3692         unsigned int data_end;
3693         unsigned int old_data;
3694         unsigned int old_size;
3695         int i;
3696         struct btrfs_map_token token;
3697
3698         leaf = path->nodes[0];
3699
3700         nritems = btrfs_header_nritems(leaf);
3701         data_end = leaf_data_end(leaf);
3702
3703         if (btrfs_leaf_free_space(leaf) < data_size) {
3704                 btrfs_print_leaf(leaf);
3705                 BUG();
3706         }
3707         slot = path->slots[0];
3708         old_data = btrfs_item_end_nr(leaf, slot);
3709
3710         BUG_ON(slot < 0);
3711         if (slot >= nritems) {
3712                 btrfs_print_leaf(leaf);
3713                 btrfs_crit(leaf->fs_info, "slot %d too large, nritems %d",
3714                            slot, nritems);
3715                 BUG();
3716         }
3717
3718         /*
3719          * item0..itemN ... dataN.offset..dataN.size .. data0.size
3720          */
3721         /* first correct the data pointers */
3722         btrfs_init_map_token(&token, leaf);
3723         for (i = slot; i < nritems; i++) {
3724                 u32 ioff;
3725                 item = btrfs_item_nr(i);
3726
3727                 ioff = btrfs_token_item_offset(&token, item);
3728                 btrfs_set_token_item_offset(&token, item, ioff - data_size);
3729         }
3730
3731         /* shift the data */
3732         memmove_extent_buffer(leaf, BTRFS_LEAF_DATA_OFFSET +
3733                       data_end - data_size, BTRFS_LEAF_DATA_OFFSET +
3734                       data_end, old_data - data_end);
3735
3736         data_end = old_data;
3737         old_size = btrfs_item_size_nr(leaf, slot);
3738         item = btrfs_item_nr(slot);
3739         btrfs_set_item_size(leaf, item, old_size + data_size);
3740         btrfs_mark_buffer_dirty(leaf);
3741
3742         if (btrfs_leaf_free_space(leaf) < 0) {
3743                 btrfs_print_leaf(leaf);
3744                 BUG();
3745         }
3746 }
3747
3748 /**
3749  * setup_items_for_insert - Helper called before inserting one or more items
3750  * to a leaf. Main purpose is to save stack depth by doing the bulk of the work
3751  * in a function that doesn't call btrfs_search_slot
3752  *
3753  * @root:       root we are inserting items to
3754  * @path:       points to the leaf/slot where we are going to insert new items
3755  * @batch:      information about the batch of items to insert
3756  */
3757 static void setup_items_for_insert(struct btrfs_root *root, struct btrfs_path *path,
3758                                    const struct btrfs_item_batch *batch)
3759 {
3760         struct btrfs_fs_info *fs_info = root->fs_info;
3761         struct btrfs_item *item;
3762         int i;
3763         u32 nritems;
3764         unsigned int data_end;
3765         struct btrfs_disk_key disk_key;
3766         struct extent_buffer *leaf;
3767         int slot;
3768         struct btrfs_map_token token;
3769         u32 total_size;
3770
3771         /*
3772          * Before anything else, update keys in the parent and other ancestors
3773          * if needed, then release the write locks on them, so that other tasks
3774          * can use them while we modify the leaf.
3775          */
3776         if (path->slots[0] == 0) {
3777                 btrfs_cpu_key_to_disk(&disk_key, &batch->keys[0]);
3778                 fixup_low_keys(path, &disk_key, 1);
3779         }
3780         btrfs_unlock_up_safe(path, 1);
3781
3782         leaf = path->nodes[0];
3783         slot = path->slots[0];
3784
3785         nritems = btrfs_header_nritems(leaf);
3786         data_end = leaf_data_end(leaf);
3787         total_size = batch->total_data_size + (batch->nr * sizeof(struct btrfs_item));
3788
3789         if (btrfs_leaf_free_space(leaf) < total_size) {
3790                 btrfs_print_leaf(leaf);
3791                 btrfs_crit(fs_info, "not enough freespace need %u have %d",
3792                            total_size, btrfs_leaf_free_space(leaf));
3793                 BUG();
3794         }
3795
3796         btrfs_init_map_token(&token, leaf);
3797         if (slot != nritems) {
3798                 unsigned int old_data = btrfs_item_end_nr(leaf, slot);
3799
3800                 if (old_data < data_end) {
3801                         btrfs_print_leaf(leaf);
3802                         btrfs_crit(fs_info,
3803                 "item at slot %d with data offset %u beyond data end of leaf %u",
3804                                    slot, old_data, data_end);
3805                         BUG();
3806                 }
3807                 /*
3808                  * item0..itemN ... dataN.offset..dataN.size .. data0.size
3809                  */
3810                 /* first correct the data pointers */
3811                 for (i = slot; i < nritems; i++) {
3812                         u32 ioff;
3813
3814                         item = btrfs_item_nr(i);
3815                         ioff = btrfs_token_item_offset(&token, item);
3816                         btrfs_set_token_item_offset(&token, item,
3817                                                     ioff - batch->total_data_size);
3818                 }
3819                 /* shift the items */
3820                 memmove_extent_buffer(leaf, btrfs_item_nr_offset(slot + batch->nr),
3821                               btrfs_item_nr_offset(slot),
3822                               (nritems - slot) * sizeof(struct btrfs_item));
3823
3824                 /* shift the data */
3825                 memmove_extent_buffer(leaf, BTRFS_LEAF_DATA_OFFSET +
3826                                       data_end - batch->total_data_size,
3827                                       BTRFS_LEAF_DATA_OFFSET + data_end,
3828                                       old_data - data_end);
3829                 data_end = old_data;
3830         }
3831
3832         /* setup the item for the new data */
3833         for (i = 0; i < batch->nr; i++) {
3834                 btrfs_cpu_key_to_disk(&disk_key, &batch->keys[i]);
3835                 btrfs_set_item_key(leaf, &disk_key, slot + i);
3836                 item = btrfs_item_nr(slot + i);
3837                 data_end -= batch->data_sizes[i];
3838                 btrfs_set_token_item_offset(&token, item, data_end);
3839                 btrfs_set_token_item_size(&token, item, batch->data_sizes[i]);
3840         }
3841
3842         btrfs_set_header_nritems(leaf, nritems + batch->nr);
3843         btrfs_mark_buffer_dirty(leaf);
3844
3845         if (btrfs_leaf_free_space(leaf) < 0) {
3846                 btrfs_print_leaf(leaf);
3847                 BUG();
3848         }
3849 }
3850
3851 /*
3852  * Insert a new item into a leaf.
3853  *
3854  * @root:      The root of the btree.
3855  * @path:      A path pointing to the target leaf and slot.
3856  * @key:       The key of the new item.
3857  * @data_size: The size of the data associated with the new key.
3858  */
3859 void btrfs_setup_item_for_insert(struct btrfs_root *root,
3860                                  struct btrfs_path *path,
3861                                  const struct btrfs_key *key,
3862                                  u32 data_size)
3863 {
3864         struct btrfs_item_batch batch;
3865
3866         batch.keys = key;
3867         batch.data_sizes = &data_size;
3868         batch.total_data_size = data_size;
3869         batch.nr = 1;
3870
3871         setup_items_for_insert(root, path, &batch);
3872 }
3873
3874 /*
3875  * Given a key and some data, insert items into the tree.
3876  * This does all the path init required, making room in the tree if needed.
3877  */
3878 int btrfs_insert_empty_items(struct btrfs_trans_handle *trans,
3879                             struct btrfs_root *root,
3880                             struct btrfs_path *path,
3881                             const struct btrfs_item_batch *batch)
3882 {
3883         int ret = 0;
3884         int slot;
3885         u32 total_size;
3886
3887         total_size = batch->total_data_size + (batch->nr * sizeof(struct btrfs_item));
3888         ret = btrfs_search_slot(trans, root, &batch->keys[0], path, total_size, 1);
3889         if (ret == 0)
3890                 return -EEXIST;
3891         if (ret < 0)
3892                 return ret;
3893
3894         slot = path->slots[0];
3895         BUG_ON(slot < 0);
3896
3897         setup_items_for_insert(root, path, batch);
3898         return 0;
3899 }
3900
3901 /*
3902  * Given a key and some data, insert an item into the tree.
3903  * This does all the path init required, making room in the tree if needed.
3904  */
3905 int btrfs_insert_item(struct btrfs_trans_handle *trans, struct btrfs_root *root,
3906                       const struct btrfs_key *cpu_key, void *data,
3907                       u32 data_size)
3908 {
3909         int ret = 0;
3910         struct btrfs_path *path;
3911         struct extent_buffer *leaf;
3912         unsigned long ptr;
3913
3914         path = btrfs_alloc_path();
3915         if (!path)
3916                 return -ENOMEM;
3917         ret = btrfs_insert_empty_item(trans, root, path, cpu_key, data_size);
3918         if (!ret) {
3919                 leaf = path->nodes[0];
3920                 ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
3921                 write_extent_buffer(leaf, data, ptr, data_size);
3922                 btrfs_mark_buffer_dirty(leaf);
3923         }
3924         btrfs_free_path(path);
3925         return ret;
3926 }
3927
3928 /*
3929  * This function duplicates an item, giving 'new_key' to the new item.
3930  * It guarantees both items live in the same tree leaf and the new item is
3931  * contiguous with the original item.
3932  *
3933  * This allows us to split a file extent in place, keeping a lock on the leaf
3934  * the entire time.
3935  */
3936 int btrfs_duplicate_item(struct btrfs_trans_handle *trans,
3937                          struct btrfs_root *root,
3938                          struct btrfs_path *path,
3939                          const struct btrfs_key *new_key)
3940 {
3941         struct extent_buffer *leaf;
3942         int ret;
3943         u32 item_size;
3944
3945         leaf = path->nodes[0];
3946         item_size = btrfs_item_size_nr(leaf, path->slots[0]);
3947         ret = setup_leaf_for_split(trans, root, path,
3948                                    item_size + sizeof(struct btrfs_item));
3949         if (ret)
3950                 return ret;
3951
3952         path->slots[0]++;
3953         btrfs_setup_item_for_insert(root, path, new_key, item_size);
3954         leaf = path->nodes[0];
3955         memcpy_extent_buffer(leaf,
3956                              btrfs_item_ptr_offset(leaf, path->slots[0]),
3957                              btrfs_item_ptr_offset(leaf, path->slots[0] - 1),
3958                              item_size);
3959         return 0;
3960 }
3961
3962 /*
3963  * delete the pointer from a given node.
3964  *
3965  * the tree should have been previously balanced so the deletion does not
3966  * empty a node.
3967  */
3968 static void del_ptr(struct btrfs_root *root, struct btrfs_path *path,
3969                     int level, int slot)
3970 {
3971         struct extent_buffer *parent = path->nodes[level];
3972         u32 nritems;
3973         int ret;
3974
3975         nritems = btrfs_header_nritems(parent);
3976         if (slot != nritems - 1) {
3977                 if (level) {
3978                         ret = btrfs_tree_mod_log_insert_move(parent, slot,
3979                                         slot + 1, nritems - slot - 1);
3980                         BUG_ON(ret < 0);
3981                 }
3982                 memmove_extent_buffer(parent,
3983                               btrfs_node_key_ptr_offset(slot),
3984                               btrfs_node_key_ptr_offset(slot + 1),
3985                               sizeof(struct btrfs_key_ptr) *
3986                               (nritems - slot - 1));
3987         } else if (level) {
3988                 ret = btrfs_tree_mod_log_insert_key(parent, slot,
3989                                 BTRFS_MOD_LOG_KEY_REMOVE, GFP_NOFS);
3990                 BUG_ON(ret < 0);
3991         }
3992
3993         nritems--;
3994         btrfs_set_header_nritems(parent, nritems);
3995         if (nritems == 0 && parent == root->node) {
3996                 BUG_ON(btrfs_header_level(root->node) != 1);
3997                 /* just turn the root into a leaf and break */
3998                 btrfs_set_header_level(root->node, 0);
3999         } else if (slot == 0) {
4000                 struct btrfs_disk_key disk_key;
4001
4002                 btrfs_node_key(parent, &disk_key, 0);
4003                 fixup_low_keys(path, &disk_key, level + 1);
4004         }
4005         btrfs_mark_buffer_dirty(parent);
4006 }
4007
4008 /*
4009  * a helper function to delete the leaf pointed to by path->slots[1] and
4010  * path->nodes[1].
4011  *
4012  * This deletes the pointer in path->nodes[1] and frees the leaf
4013  * block extent.  zero is returned if it all worked out, < 0 otherwise.
4014  *
4015  * The path must have already been setup for deleting the leaf, including
4016  * all the proper balancing.  path->nodes[1] must be locked.
4017  */
4018 static noinline void btrfs_del_leaf(struct btrfs_trans_handle *trans,
4019                                     struct btrfs_root *root,
4020                                     struct btrfs_path *path,
4021                                     struct extent_buffer *leaf)
4022 {
4023         WARN_ON(btrfs_header_generation(leaf) != trans->transid);
4024         del_ptr(root, path, 1, path->slots[1]);
4025
4026         /*
4027          * btrfs_free_extent is expensive, we want to make sure we
4028          * aren't holding any locks when we call it
4029          */
4030         btrfs_unlock_up_safe(path, 0);
4031
4032         root_sub_used(root, leaf->len);
4033
4034         atomic_inc(&leaf->refs);
4035         btrfs_free_tree_block(trans, root, leaf, 0, 1);
4036         free_extent_buffer_stale(leaf);
4037 }
4038 /*
4039  * delete the item at the leaf level in path.  If that empties
4040  * the leaf, remove it from the tree
4041  */
4042 int btrfs_del_items(struct btrfs_trans_handle *trans, struct btrfs_root *root,
4043                     struct btrfs_path *path, int slot, int nr)
4044 {
4045         struct btrfs_fs_info *fs_info = root->fs_info;
4046         struct extent_buffer *leaf;
4047         struct btrfs_item *item;
4048         u32 last_off;
4049         u32 dsize = 0;
4050         int ret = 0;
4051         int wret;
4052         int i;
4053         u32 nritems;
4054
4055         leaf = path->nodes[0];
4056         last_off = btrfs_item_offset_nr(leaf, slot + nr - 1);
4057
4058         for (i = 0; i < nr; i++)
4059                 dsize += btrfs_item_size_nr(leaf, slot + i);
4060
4061         nritems = btrfs_header_nritems(leaf);
4062
4063         if (slot + nr != nritems) {
4064                 int data_end = leaf_data_end(leaf);
4065                 struct btrfs_map_token token;
4066
4067                 memmove_extent_buffer(leaf, BTRFS_LEAF_DATA_OFFSET +
4068                               data_end + dsize,
4069                               BTRFS_LEAF_DATA_OFFSET + data_end,
4070                               last_off - data_end);
4071
4072                 btrfs_init_map_token(&token, leaf);
4073                 for (i = slot + nr; i < nritems; i++) {
4074                         u32 ioff;
4075
4076                         item = btrfs_item_nr(i);
4077                         ioff = btrfs_token_item_offset(&token, item);
4078                         btrfs_set_token_item_offset(&token, item, ioff + dsize);
4079                 }
4080
4081                 memmove_extent_buffer(leaf, btrfs_item_nr_offset(slot),
4082                               btrfs_item_nr_offset(slot + nr),
4083                               sizeof(struct btrfs_item) *
4084                               (nritems - slot - nr));
4085         }
4086         btrfs_set_header_nritems(leaf, nritems - nr);
4087         nritems -= nr;
4088
4089         /* delete the leaf if we've emptied it */
4090         if (nritems == 0) {
4091                 if (leaf == root->node) {
4092                         btrfs_set_header_level(leaf, 0);
4093                 } else {
4094                         btrfs_clean_tree_block(leaf);
4095                         btrfs_del_leaf(trans, root, path, leaf);
4096                 }
4097         } else {
4098                 int used = leaf_space_used(leaf, 0, nritems);
4099                 if (slot == 0) {
4100                         struct btrfs_disk_key disk_key;
4101
4102                         btrfs_item_key(leaf, &disk_key, 0);
4103                         fixup_low_keys(path, &disk_key, 1);
4104                 }
4105
4106                 /* delete the leaf if it is mostly empty */
4107                 if (used < BTRFS_LEAF_DATA_SIZE(fs_info) / 3) {
4108                         /* push_leaf_left fixes the path.
4109                          * make sure the path still points to our leaf
4110                          * for possible call to del_ptr below
4111                          */
4112                         slot = path->slots[1];
4113                         atomic_inc(&leaf->refs);
4114
4115                         wret = push_leaf_left(trans, root, path, 1, 1,
4116                                               1, (u32)-1);
4117                         if (wret < 0 && wret != -ENOSPC)
4118                                 ret = wret;
4119
4120                         if (path->nodes[0] == leaf &&
4121                             btrfs_header_nritems(leaf)) {
4122                                 wret = push_leaf_right(trans, root, path, 1,
4123                                                        1, 1, 0);
4124                                 if (wret < 0 && wret != -ENOSPC)
4125                                         ret = wret;
4126                         }
4127
4128                         if (btrfs_header_nritems(leaf) == 0) {
4129                                 path->slots[1] = slot;
4130                                 btrfs_del_leaf(trans, root, path, leaf);
4131                                 free_extent_buffer(leaf);
4132                                 ret = 0;
4133                         } else {
4134                                 /* if we're still in the path, make sure
4135                                  * we're dirty.  Otherwise, one of the
4136                                  * push_leaf functions must have already
4137                                  * dirtied this buffer
4138                                  */
4139                                 if (path->nodes[0] == leaf)
4140                                         btrfs_mark_buffer_dirty(leaf);
4141                                 free_extent_buffer(leaf);
4142                         }
4143                 } else {
4144                         btrfs_mark_buffer_dirty(leaf);
4145                 }
4146         }
4147         return ret;
4148 }
4149
4150 /*
4151  * search the tree again to find a leaf with lesser keys
4152  * returns 0 if it found something or 1 if there are no lesser leaves.
4153  * returns < 0 on io errors.
4154  *
4155  * This may release the path, and so you may lose any locks held at the
4156  * time you call it.
4157  */
4158 int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path)
4159 {
4160         struct btrfs_key key;
4161         struct btrfs_disk_key found_key;
4162         int ret;
4163
4164         btrfs_item_key_to_cpu(path->nodes[0], &key, 0);
4165
4166         if (key.offset > 0) {
4167                 key.offset--;
4168         } else if (key.type > 0) {
4169                 key.type--;
4170                 key.offset = (u64)-1;
4171         } else if (key.objectid > 0) {
4172                 key.objectid--;
4173                 key.type = (u8)-1;
4174                 key.offset = (u64)-1;
4175         } else {
4176                 return 1;
4177         }
4178
4179         btrfs_release_path(path);
4180         ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
4181         if (ret < 0)
4182                 return ret;
4183         btrfs_item_key(path->nodes[0], &found_key, 0);
4184         ret = comp_keys(&found_key, &key);
4185         /*
4186          * We might have had an item with the previous key in the tree right
4187          * before we released our path. And after we released our path, that
4188          * item might have been pushed to the first slot (0) of the leaf we
4189          * were holding due to a tree balance. Alternatively, an item with the
4190          * previous key can exist as the only element of a leaf (big fat item).
4191          * Therefore account for these 2 cases, so that our callers (like
4192          * btrfs_previous_item) don't miss an existing item with a key matching
4193          * the previous key we computed above.
4194          */
4195         if (ret <= 0)
4196                 return 0;
4197         return 1;
4198 }
4199
4200 /*
4201  * A helper function to walk down the tree starting at min_key, and looking
4202  * for nodes or leaves that are have a minimum transaction id.
4203  * This is used by the btree defrag code, and tree logging
4204  *
4205  * This does not cow, but it does stuff the starting key it finds back
4206  * into min_key, so you can call btrfs_search_slot with cow=1 on the
4207  * key and get a writable path.
4208  *
4209  * This honors path->lowest_level to prevent descent past a given level
4210  * of the tree.
4211  *
4212  * min_trans indicates the oldest transaction that you are interested
4213  * in walking through.  Any nodes or leaves older than min_trans are
4214  * skipped over (without reading them).
4215  *
4216  * returns zero if something useful was found, < 0 on error and 1 if there
4217  * was nothing in the tree that matched the search criteria.
4218  */
4219 int btrfs_search_forward(struct btrfs_root *root, struct btrfs_key *min_key,
4220                          struct btrfs_path *path,
4221                          u64 min_trans)
4222 {
4223         struct extent_buffer *cur;
4224         struct btrfs_key found_key;
4225         int slot;
4226         int sret;
4227         u32 nritems;
4228         int level;
4229         int ret = 1;
4230         int keep_locks = path->keep_locks;
4231
4232         path->keep_locks = 1;
4233 again:
4234         cur = btrfs_read_lock_root_node(root);
4235         level = btrfs_header_level(cur);
4236         WARN_ON(path->nodes[level]);
4237         path->nodes[level] = cur;
4238         path->locks[level] = BTRFS_READ_LOCK;
4239
4240         if (btrfs_header_generation(cur) < min_trans) {
4241                 ret = 1;
4242                 goto out;
4243         }
4244         while (1) {
4245                 nritems = btrfs_header_nritems(cur);
4246                 level = btrfs_header_level(cur);
4247                 sret = btrfs_bin_search(cur, min_key, &slot);
4248                 if (sret < 0) {
4249                         ret = sret;
4250                         goto out;
4251                 }
4252
4253                 /* at the lowest level, we're done, setup the path and exit */
4254                 if (level == path->lowest_level) {
4255                         if (slot >= nritems)
4256                                 goto find_next_key;
4257                         ret = 0;
4258                         path->slots[level] = slot;
4259                         btrfs_item_key_to_cpu(cur, &found_key, slot);
4260                         goto out;
4261                 }
4262                 if (sret && slot > 0)
4263                         slot--;
4264                 /*
4265                  * check this node pointer against the min_trans parameters.
4266                  * If it is too old, skip to the next one.
4267                  */
4268                 while (slot < nritems) {
4269                         u64 gen;
4270
4271                         gen = btrfs_node_ptr_generation(cur, slot);
4272                         if (gen < min_trans) {
4273                                 slot++;
4274                                 continue;
4275                         }
4276                         break;
4277                 }
4278 find_next_key:
4279                 /*
4280                  * we didn't find a candidate key in this node, walk forward
4281                  * and find another one
4282                  */
4283                 if (slot >= nritems) {
4284                         path->slots[level] = slot;
4285                         sret = btrfs_find_next_key(root, path, min_key, level,
4286                                                   min_trans);
4287                         if (sret == 0) {
4288                                 btrfs_release_path(path);
4289                                 goto again;
4290                         } else {
4291                                 goto out;
4292                         }
4293                 }
4294                 /* save our key for returning back */
4295                 btrfs_node_key_to_cpu(cur, &found_key, slot);
4296                 path->slots[level] = slot;
4297                 if (level == path->lowest_level) {
4298                         ret = 0;
4299                         goto out;
4300                 }
4301                 cur = btrfs_read_node_slot(cur, slot);
4302                 if (IS_ERR(cur)) {
4303                         ret = PTR_ERR(cur);
4304                         goto out;
4305                 }
4306
4307                 btrfs_tree_read_lock(cur);
4308
4309                 path->locks[level - 1] = BTRFS_READ_LOCK;
4310                 path->nodes[level - 1] = cur;
4311                 unlock_up(path, level, 1, 0, NULL);
4312         }
4313 out:
4314         path->keep_locks = keep_locks;
4315         if (ret == 0) {
4316                 btrfs_unlock_up_safe(path, path->lowest_level + 1);
4317                 memcpy(min_key, &found_key, sizeof(found_key));
4318         }
4319         return ret;
4320 }
4321
4322 /*
4323  * this is similar to btrfs_next_leaf, but does not try to preserve
4324  * and fixup the path.  It looks for and returns the next key in the
4325  * tree based on the current path and the min_trans parameters.
4326  *
4327  * 0 is returned if another key is found, < 0 if there are any errors
4328  * and 1 is returned if there are no higher keys in the tree
4329  *
4330  * path->keep_locks should be set to 1 on the search made before
4331  * calling this function.
4332  */
4333 int btrfs_find_next_key(struct btrfs_root *root, struct btrfs_path *path,
4334                         struct btrfs_key *key, int level, u64 min_trans)
4335 {
4336         int slot;
4337         struct extent_buffer *c;
4338
4339         WARN_ON(!path->keep_locks && !path->skip_locking);
4340         while (level < BTRFS_MAX_LEVEL) {
4341                 if (!path->nodes[level])
4342                         return 1;
4343
4344                 slot = path->slots[level] + 1;
4345                 c = path->nodes[level];
4346 next:
4347                 if (slot >= btrfs_header_nritems(c)) {
4348                         int ret;
4349                         int orig_lowest;
4350                         struct btrfs_key cur_key;
4351                         if (level + 1 >= BTRFS_MAX_LEVEL ||
4352                             !path->nodes[level + 1])
4353                                 return 1;
4354
4355                         if (path->locks[level + 1] || path->skip_locking) {
4356                                 level++;
4357                                 continue;
4358                         }
4359
4360                         slot = btrfs_header_nritems(c) - 1;
4361                         if (level == 0)
4362                                 btrfs_item_key_to_cpu(c, &cur_key, slot);
4363                         else
4364                                 btrfs_node_key_to_cpu(c, &cur_key, slot);
4365
4366                         orig_lowest = path->lowest_level;
4367                         btrfs_release_path(path);
4368                         path->lowest_level = level;
4369                         ret = btrfs_search_slot(NULL, root, &cur_key, path,
4370                                                 0, 0);
4371                         path->lowest_level = orig_lowest;
4372                         if (ret < 0)
4373                                 return ret;
4374
4375                         c = path->nodes[level];
4376                         slot = path->slots[level];
4377                         if (ret == 0)
4378                                 slot++;
4379                         goto next;
4380                 }
4381
4382                 if (level == 0)
4383                         btrfs_item_key_to_cpu(c, key, slot);
4384                 else {
4385                         u64 gen = btrfs_node_ptr_generation(c, slot);
4386
4387                         if (gen < min_trans) {
4388                                 slot++;
4389                                 goto next;
4390                         }
4391                         btrfs_node_key_to_cpu(c, key, slot);
4392                 }
4393                 return 0;
4394         }
4395         return 1;
4396 }
4397
4398 int btrfs_next_old_leaf(struct btrfs_root *root, struct btrfs_path *path,
4399                         u64 time_seq)
4400 {
4401         int slot;
4402         int level;
4403         struct extent_buffer *c;
4404         struct extent_buffer *next;
4405         struct btrfs_key key;
4406         u32 nritems;
4407         int ret;
4408         int i;
4409
4410         nritems = btrfs_header_nritems(path->nodes[0]);
4411         if (nritems == 0)
4412                 return 1;
4413
4414         btrfs_item_key_to_cpu(path->nodes[0], &key, nritems - 1);
4415 again:
4416         level = 1;
4417         next = NULL;
4418         btrfs_release_path(path);
4419
4420         path->keep_locks = 1;
4421
4422         if (time_seq)
4423                 ret = btrfs_search_old_slot(root, &key, path, time_seq);
4424         else
4425                 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
4426         path->keep_locks = 0;
4427
4428         if (ret < 0)
4429                 return ret;
4430
4431         nritems = btrfs_header_nritems(path->nodes[0]);
4432         /*
4433          * by releasing the path above we dropped all our locks.  A balance
4434          * could have added more items next to the key that used to be
4435          * at the very end of the block.  So, check again here and
4436          * advance the path if there are now more items available.
4437          */
4438         if (nritems > 0 && path->slots[0] < nritems - 1) {
4439                 if (ret == 0)
4440                         path->slots[0]++;
4441                 ret = 0;
4442                 goto done;
4443         }
4444         /*
4445          * So the above check misses one case:
4446          * - after releasing the path above, someone has removed the item that
4447          *   used to be at the very end of the block, and balance between leafs
4448          *   gets another one with bigger key.offset to replace it.
4449          *
4450          * This one should be returned as well, or we can get leaf corruption
4451          * later(esp. in __btrfs_drop_extents()).
4452          *
4453          * And a bit more explanation about this check,
4454          * with ret > 0, the key isn't found, the path points to the slot
4455          * where it should be inserted, so the path->slots[0] item must be the
4456          * bigger one.
4457          */
4458         if (nritems > 0 && ret > 0 && path->slots[0] == nritems - 1) {
4459                 ret = 0;
4460                 goto done;
4461         }
4462
4463         while (level < BTRFS_MAX_LEVEL) {
4464                 if (!path->nodes[level]) {
4465                         ret = 1;
4466                         goto done;
4467                 }
4468
4469                 slot = path->slots[level] + 1;
4470                 c = path->nodes[level];
4471                 if (slot >= btrfs_header_nritems(c)) {
4472                         level++;
4473                         if (level == BTRFS_MAX_LEVEL) {
4474                                 ret = 1;
4475                                 goto done;
4476                         }
4477                         continue;
4478                 }
4479
4480
4481                 /*
4482                  * Our current level is where we're going to start from, and to
4483                  * make sure lockdep doesn't complain we need to drop our locks
4484                  * and nodes from 0 to our current level.
4485                  */
4486                 for (i = 0; i < level; i++) {
4487                         if (path->locks[level]) {
4488                                 btrfs_tree_read_unlock(path->nodes[i]);
4489                                 path->locks[i] = 0;
4490                         }
4491                         free_extent_buffer(path->nodes[i]);
4492                         path->nodes[i] = NULL;
4493                 }
4494
4495                 next = c;
4496                 ret = read_block_for_search(root, path, &next, level,
4497                                             slot, &key);
4498                 if (ret == -EAGAIN)
4499                         goto again;
4500
4501                 if (ret < 0) {
4502                         btrfs_release_path(path);
4503                         goto done;
4504                 }
4505
4506                 if (!path->skip_locking) {
4507                         ret = btrfs_try_tree_read_lock(next);
4508                         if (!ret && time_seq) {
4509                                 /*
4510                                  * If we don't get the lock, we may be racing
4511                                  * with push_leaf_left, holding that lock while
4512                                  * itself waiting for the leaf we've currently
4513                                  * locked. To solve this situation, we give up
4514                                  * on our lock and cycle.
4515                                  */
4516                                 free_extent_buffer(next);
4517                                 btrfs_release_path(path);
4518                                 cond_resched();
4519                                 goto again;
4520                         }
4521                         if (!ret)
4522                                 btrfs_tree_read_lock(next);
4523                 }
4524                 break;
4525         }
4526         path->slots[level] = slot;
4527         while (1) {
4528                 level--;
4529                 path->nodes[level] = next;
4530                 path->slots[level] = 0;
4531                 if (!path->skip_locking)
4532                         path->locks[level] = BTRFS_READ_LOCK;
4533                 if (!level)
4534                         break;
4535
4536                 ret = read_block_for_search(root, path, &next, level,
4537                                             0, &key);
4538                 if (ret == -EAGAIN)
4539                         goto again;
4540
4541                 if (ret < 0) {
4542                         btrfs_release_path(path);
4543                         goto done;
4544                 }
4545
4546                 if (!path->skip_locking)
4547                         btrfs_tree_read_lock(next);
4548         }
4549         ret = 0;
4550 done:
4551         unlock_up(path, 0, 1, 0, NULL);
4552
4553         return ret;
4554 }
4555
4556 /*
4557  * this uses btrfs_prev_leaf to walk backwards in the tree, and keeps
4558  * searching until it gets past min_objectid or finds an item of 'type'
4559  *
4560  * returns 0 if something is found, 1 if nothing was found and < 0 on error
4561  */
4562 int btrfs_previous_item(struct btrfs_root *root,
4563                         struct btrfs_path *path, u64 min_objectid,
4564                         int type)
4565 {
4566         struct btrfs_key found_key;
4567         struct extent_buffer *leaf;
4568         u32 nritems;
4569         int ret;
4570
4571         while (1) {
4572                 if (path->slots[0] == 0) {
4573                         ret = btrfs_prev_leaf(root, path);
4574                         if (ret != 0)
4575                                 return ret;
4576                 } else {
4577                         path->slots[0]--;
4578                 }
4579                 leaf = path->nodes[0];
4580                 nritems = btrfs_header_nritems(leaf);
4581                 if (nritems == 0)
4582                         return 1;
4583                 if (path->slots[0] == nritems)
4584                         path->slots[0]--;
4585
4586                 btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
4587                 if (found_key.objectid < min_objectid)
4588                         break;
4589                 if (found_key.type == type)
4590                         return 0;
4591                 if (found_key.objectid == min_objectid &&
4592                     found_key.type < type)
4593                         break;
4594         }
4595         return 1;
4596 }
4597
4598 /*
4599  * search in extent tree to find a previous Metadata/Data extent item with
4600  * min objecitd.
4601  *
4602  * returns 0 if something is found, 1 if nothing was found and < 0 on error
4603  */
4604 int btrfs_previous_extent_item(struct btrfs_root *root,
4605                         struct btrfs_path *path, u64 min_objectid)
4606 {
4607         struct btrfs_key found_key;
4608         struct extent_buffer *leaf;
4609         u32 nritems;
4610         int ret;
4611
4612         while (1) {
4613                 if (path->slots[0] == 0) {
4614                         ret = btrfs_prev_leaf(root, path);
4615                         if (ret != 0)
4616                                 return ret;
4617                 } else {
4618                         path->slots[0]--;
4619                 }
4620                 leaf = path->nodes[0];
4621                 nritems = btrfs_header_nritems(leaf);
4622                 if (nritems == 0)
4623                         return 1;
4624                 if (path->slots[0] == nritems)
4625                         path->slots[0]--;
4626
4627                 btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
4628                 if (found_key.objectid < min_objectid)
4629                         break;
4630                 if (found_key.type == BTRFS_EXTENT_ITEM_KEY ||
4631                     found_key.type == BTRFS_METADATA_ITEM_KEY)
4632                         return 0;
4633                 if (found_key.objectid == min_objectid &&
4634                     found_key.type < BTRFS_EXTENT_ITEM_KEY)
4635                         break;
4636         }
4637         return 1;
4638 }