fs/btrfs/ctree.c

   1 /*
   2  * Copyright (C) 2007,2008 Oracle.  All rights reserved.
   3  *
   4  * This program is free software; you can redistribute it and/or
   5  * modify it under the terms of the GNU General Public
   6  * License v2 as published by the Free Software Foundation.
   7  *
   8  * This program is distributed in the hope that it will be useful,
   9  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  10  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  11  * General Public License for more details.
  12  *
  13  * You should have received a copy of the GNU General Public
  14  * License along with this program; if not, write to the
  15  * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
  16  * Boston, MA 021110-1307, USA.
  17  */
  18
  19 #include <linux/sched.h>
  20 #include <linux/slab.h>
  21 #include <linux/rbtree.h>
  22 #include <linux/vmalloc.h>
  23 #include "ctree.h"
  24 #include "disk-io.h"
  25 #include "transaction.h"
  26 #include "print-tree.h"
  27 #include "locking.h"
  28
  29 static int split_node(struct btrfs_trans_handle *trans, struct btrfs_root
  30                       *root, struct btrfs_path *path, int level);
  31 static int split_leaf(struct btrfs_trans_handle *trans, struct btrfs_root
  32                       *root, struct btrfs_key *ins_key,
  33                       struct btrfs_path *path, int data_size, int extend);
  34 static int push_node_left(struct btrfs_trans_handle *trans,
  35                           struct btrfs_root *root, struct extent_buffer *dst,
  36                           struct extent_buffer *src, int empty);
  37 static int balance_node_right(struct btrfs_trans_handle *trans,
  38                               struct btrfs_root *root,
  39                               struct extent_buffer *dst_buf,
  40                               struct extent_buffer *src_buf);
  41 static void del_ptr(struct btrfs_root *root, struct btrfs_path *path,
  42                     int level, int slot);
  43 static int tree_mod_log_free_eb(struct btrfs_fs_info *fs_info,
  44                                  struct extent_buffer *eb);
  45
  46 struct btrfs_path *btrfs_alloc_path(void)
  47 {
  48         struct btrfs_path *path;
  49         path = kmem_cache_zalloc(btrfs_path_cachep, GFP_NOFS);
  50         return path;
  51 }
  52
  53 /*
  54  * set all locked nodes in the path to blocking locks.  This should
  55  * be done before scheduling
  56  */
  57 noinline void btrfs_set_path_blocking(struct btrfs_path *p)
  58 {
  59         int i;
  60         for (i = 0; i < BTRFS_MAX_LEVEL; i++) {
  61                 if (!p->nodes[i] || !p->locks[i])
  62                         continue;
  63                 btrfs_set_lock_blocking_rw(p->nodes[i], p->locks[i]);
  64                 if (p->locks[i] == BTRFS_READ_LOCK)
  65                         p->locks[i] = BTRFS_READ_LOCK_BLOCKING;
  66                 else if (p->locks[i] == BTRFS_WRITE_LOCK)
  67                         p->locks[i] = BTRFS_WRITE_LOCK_BLOCKING;
  68         }
  69 }
  70
  71 /*
  72  * reset all the locked nodes in the patch to spinning locks.
  73  *
  74  * held is used to keep lockdep happy, when lockdep is enabled
  75  * we set held to a blocking lock before we go around and
  76  * retake all the spinlocks in the path.  You can safely use NULL
  77  * for held
  78  */
  79 noinline void btrfs_clear_path_blocking(struct btrfs_path *p,
  80                                         struct extent_buffer *held, int held_rw)
  81 {
  82         int i;
  83
  84         if (held) {
  85                 btrfs_set_lock_blocking_rw(held, held_rw);
  86                 if (held_rw == BTRFS_WRITE_LOCK)
  87                         held_rw = BTRFS_WRITE_LOCK_BLOCKING;
  88                 else if (held_rw == BTRFS_READ_LOCK)
  89                         held_rw = BTRFS_READ_LOCK_BLOCKING;
  90         }
  91         btrfs_set_path_blocking(p);
  92
  93         for (i = BTRFS_MAX_LEVEL - 1; i >= 0; i--) {
  94                 if (p->nodes[i] && p->locks[i]) {
  95                         btrfs_clear_lock_blocking_rw(p->nodes[i], p->locks[i]);
  96                         if (p->locks[i] == BTRFS_WRITE_LOCK_BLOCKING)
  97                                 p->locks[i] = BTRFS_WRITE_LOCK;
  98                         else if (p->locks[i] == BTRFS_READ_LOCK_BLOCKING)
  99                                 p->locks[i] = BTRFS_READ_LOCK;
 100                 }
 101         }
 102
 103         if (held)
 104                 btrfs_clear_lock_blocking_rw(held, held_rw);
 105 }
 106
 107 /* this also releases the path */
 108 void btrfs_free_path(struct btrfs_path *p)
 109 {
 110         if (!p)
 111                 return;
 112         btrfs_release_path(p);
 113         kmem_cache_free(btrfs_path_cachep, p);
 114 }
 115
 116 /*
 117  * path release drops references on the extent buffers in the path
 118  * and it drops any locks held by this path
 119  *
 120  * It is safe to call this on paths that no locks or extent buffers held.
 121  */
 122 noinline void btrfs_release_path(struct btrfs_path *p)
 123 {
 124         int i;
 125
 126         for (i = 0; i < BTRFS_MAX_LEVEL; i++) {
 127                 p->slots[i] = 0;
 128                 if (!p->nodes[i])
 129                         continue;
 130                 if (p->locks[i]) {
 131                         btrfs_tree_unlock_rw(p->nodes[i], p->locks[i]);
 132                         p->locks[i] = 0;
 133                 }
 134                 free_extent_buffer(p->nodes[i]);
 135                 p->nodes[i] = NULL;
 136         }
 137 }
 138
 139 /*
 140  * safely gets a reference on the root node of a tree.  A lock
 141  * is not taken, so a concurrent writer may put a different node
 142  * at the root of the tree.  See btrfs_lock_root_node for the
 143  * looping required.
 144  *
 145  * The extent buffer returned by this has a reference taken, so
 146  * it won't disappear.  It may stop being the root of the tree
 147  * at any time because there are no locks held.
 148  */
 149 struct extent_buffer *btrfs_root_node(struct btrfs_root *root)
 150 {
 151         struct extent_buffer *eb;
 152
 153         while (1) {
 154                 rcu_read_lock();
 155                 eb = rcu_dereference(root->node);
 156
 157                 /*
 158                  * RCU really hurts here, we could free up the root node because
 159                  * it was COWed but we may not get the new root node yet so do
 160                  * the inc_not_zero dance and if it doesn't work then
 161                  * synchronize_rcu and try again.
 162                  */
 163                 if (atomic_inc_not_zero(&eb->refs)) {
 164                         rcu_read_unlock();
 165                         break;
 166                 }
 167                 rcu_read_unlock();
 168                 synchronize_rcu();
 169         }
 170         return eb;
 171 }
 172
 173 /* loop around taking references on and locking the root node of the
 174  * tree until you end up with a lock on the root.  A locked buffer
 175  * is returned, with a reference held.
 176  */
 177 struct extent_buffer *btrfs_lock_root_node(struct btrfs_root *root)
 178 {
 179         struct extent_buffer *eb;
 180
 181         while (1) {
 182                 eb = btrfs_root_node(root);
 183                 btrfs_tree_lock(eb);
 184                 if (eb == root->node)
 185                         break;
 186                 btrfs_tree_unlock(eb);
 187                 free_extent_buffer(eb);
 188         }
 189         return eb;
 190 }
 191
 192 /* loop around taking references on and locking the root node of the
 193  * tree until you end up with a lock on the root.  A locked buffer
 194  * is returned, with a reference held.
 195  */
 196 static struct extent_buffer *btrfs_read_lock_root_node(struct btrfs_root *root)
 197 {
 198         struct extent_buffer *eb;
 199
 200         while (1) {
 201                 eb = btrfs_root_node(root);
 202                 btrfs_tree_read_lock(eb);
 203                 if (eb == root->node)
 204                         break;
 205                 btrfs_tree_read_unlock(eb);
 206                 free_extent_buffer(eb);
 207         }
 208         return eb;
 209 }
 210
 211 /* cowonly root (everything not a reference counted cow subvolume), just get
 212  * put onto a simple dirty list.  transaction.c walks this to make sure they
 213  * get properly updated on disk.
 214  */
 215 static void add_root_to_dirty_list(struct btrfs_root *root)
 216 {
 217         if (test_bit(BTRFS_ROOT_DIRTY, &root->state) ||
 218             !test_bit(BTRFS_ROOT_TRACK_DIRTY, &root->state))
 219                 return;
 220
 221         spin_lock(&root->fs_info->trans_lock);
 222         if (!test_and_set_bit(BTRFS_ROOT_DIRTY, &root->state)) {
 223                 /* Want the extent tree to be the last on the list */
 224                 if (root->objectid == BTRFS_EXTENT_TREE_OBJECTID)
 225                         list_move_tail(&root->dirty_list,
 226                                        &root->fs_info->dirty_cowonly_roots);
 227                 else
 228                         list_move(&root->dirty_list,
 229                                   &root->fs_info->dirty_cowonly_roots);
 230         }
 231         spin_unlock(&root->fs_info->trans_lock);
 232 }
 233
 234 /*
 235  * used by snapshot creation to make a copy of a root for a tree with
 236  * a given objectid.  The buffer with the new root node is returned in
 237  * cow_ret, and this func returns zero on success or a negative error code.
 238  */
 239 int btrfs_copy_root(struct btrfs_trans_handle *trans,
 240                       struct btrfs_root *root,
 241                       struct extent_buffer *buf,
 242                       struct extent_buffer **cow_ret, u64 new_root_objectid)
 243 {
 244         struct extent_buffer *cow;
 245         int ret = 0;
 246         int level;
 247         struct btrfs_disk_key disk_key;
 248
 249         WARN_ON(test_bit(BTRFS_ROOT_REF_COWS, &root->state) &&
 250                 trans->transid != root->fs_info->running_transaction->transid);
 251         WARN_ON(test_bit(BTRFS_ROOT_REF_COWS, &root->state) &&
 252                 trans->transid != root->last_trans);
 253
 254         level = btrfs_header_level(buf);
 255         if (level == 0)
 256                 btrfs_item_key(buf, &disk_key, 0);
 257         else
 258                 btrfs_node_key(buf, &disk_key, 0);
 259
 260         cow = btrfs_alloc_tree_block(trans, root, 0, new_root_objectid,
 261                         &disk_key, level, buf->start, 0);
 262         if (IS_ERR(cow))
 263                 return PTR_ERR(cow);
 264
 265         copy_extent_buffer(cow, buf, 0, 0, cow->len);
 266         btrfs_set_header_bytenr(cow, cow->start);
 267         btrfs_set_header_generation(cow, trans->transid);
 268         btrfs_set_header_backref_rev(cow, BTRFS_MIXED_BACKREF_REV);
 269         btrfs_clear_header_flag(cow, BTRFS_HEADER_FLAG_WRITTEN |
 270                                      BTRFS_HEADER_FLAG_RELOC);
 271         if (new_root_objectid == BTRFS_TREE_RELOC_OBJECTID)
 272                 btrfs_set_header_flag(cow, BTRFS_HEADER_FLAG_RELOC);
 273         else
 274                 btrfs_set_header_owner(cow, new_root_objectid);
 275
 276         write_extent_buffer(cow, root->fs_info->fsid, btrfs_header_fsid(),
 277                             BTRFS_FSID_SIZE);
 278
 279         WARN_ON(btrfs_header_generation(buf) > trans->transid);
 280         if (new_root_objectid == BTRFS_TREE_RELOC_OBJECTID)
 281                 ret = btrfs_inc_ref(trans, root, cow, 1);
 282         else
 283                 ret = btrfs_inc_ref(trans, root, cow, 0);
 284
 285         if (ret)
 286                 return ret;
 287
 288         btrfs_mark_buffer_dirty(cow);
 289         *cow_ret = cow;
 290         return 0;
 291 }
 292
 293 enum mod_log_op {
 294         MOD_LOG_KEY_REPLACE,
 295         MOD_LOG_KEY_ADD,
 296         MOD_LOG_KEY_REMOVE,
 297         MOD_LOG_KEY_REMOVE_WHILE_FREEING,
 298         MOD_LOG_KEY_REMOVE_WHILE_MOVING,
 299         MOD_LOG_MOVE_KEYS,
 300         MOD_LOG_ROOT_REPLACE,
 301 };
 302
 303 struct tree_mod_move {
 304         int dst_slot;
 305         int nr_items;
 306 };
 307
 308 struct tree_mod_root {
 309         u64 logical;
 310         u8 level;
 311 };
 312
 313 struct tree_mod_elem {
 314         struct rb_node node;
 315         u64 logical;
 316         u64 seq;
 317         enum mod_log_op op;
 318
 319         /* this is used for MOD_LOG_KEY_* and MOD_LOG_MOVE_KEYS operations */
 320         int slot;
 321
 322         /* this is used for MOD_LOG_KEY* and MOD_LOG_ROOT_REPLACE */
 323         u64 generation;
 324
 325         /* those are used for op == MOD_LOG_KEY_{REPLACE,REMOVE} */
 326         struct btrfs_disk_key key;
 327         u64 blockptr;
 328
 329         /* this is used for op == MOD_LOG_MOVE_KEYS */
 330         struct tree_mod_move move;
 331
 332         /* this is used for op == MOD_LOG_ROOT_REPLACE */
 333         struct tree_mod_root old_root;
 334 };
 335
 336 static inline void tree_mod_log_read_lock(struct btrfs_fs_info *fs_info)
 337 {
 338         read_lock(&fs_info->tree_mod_log_lock);
 339 }
 340
 341 static inline void tree_mod_log_read_unlock(struct btrfs_fs_info *fs_info)
 342 {
 343         read_unlock(&fs_info->tree_mod_log_lock);
 344 }
 345
 346 static inline void tree_mod_log_write_lock(struct btrfs_fs_info *fs_info)
 347 {
 348         write_lock(&fs_info->tree_mod_log_lock);
 349 }
 350
 351 static inline void tree_mod_log_write_unlock(struct btrfs_fs_info *fs_info)
 352 {
 353         write_unlock(&fs_info->tree_mod_log_lock);
 354 }
 355
 356 /*
 357  * Pull a new tree mod seq number for our operation.
 358  */
 359 static inline u64 btrfs_inc_tree_mod_seq(struct btrfs_fs_info *fs_info)
 360 {
 361         return atomic64_inc_return(&fs_info->tree_mod_seq);
 362 }
 363
 364 /*
 365  * This adds a new blocker to the tree mod log's blocker list if the @elem
 366  * passed does not already have a sequence number set. So when a caller expects
 367  * to record tree modifications, it should ensure to set elem->seq to zero
 368  * before calling btrfs_get_tree_mod_seq.
 369  * Returns a fresh, unused tree log modification sequence number, even if no new
 370  * blocker was added.
 371  */
 372 u64 btrfs_get_tree_mod_seq(struct btrfs_fs_info *fs_info,
 373                            struct seq_list *elem)
 374 {
 375         tree_mod_log_write_lock(fs_info);
 376         spin_lock(&fs_info->tree_mod_seq_lock);
 377         if (!elem->seq) {
 378                 elem->seq = btrfs_inc_tree_mod_seq(fs_info);
 379                 list_add_tail(&elem->list, &fs_info->tree_mod_seq_list);
 380         }
 381         spin_unlock(&fs_info->tree_mod_seq_lock);
 382         tree_mod_log_write_unlock(fs_info);
 383
 384         return elem->seq;
 385 }
 386
 387 void btrfs_put_tree_mod_seq(struct btrfs_fs_info *fs_info,
 388                             struct seq_list *elem)
 389 {
 390         struct rb_root *tm_root;
 391         struct rb_node *node;
 392         struct rb_node *next;
 393         struct seq_list *cur_elem;
 394         struct tree_mod_elem *tm;
 395         u64 min_seq = (u64)-1;
 396         u64 seq_putting = elem->seq;
 397
 398         if (!seq_putting)
 399                 return;
 400
 401         spin_lock(&fs_info->tree_mod_seq_lock);
 402         list_del(&elem->list);
 403         elem->seq = 0;
 404
 405         list_for_each_entry(cur_elem, &fs_info->tree_mod_seq_list, list) {
 406                 if (cur_elem->seq < min_seq) {
 407                         if (seq_putting > cur_elem->seq) {
 408                                 /*
 409                                  * blocker with lower sequence number exists, we
 410                                  * cannot remove anything from the log
 411                                  */
 412                                 spin_unlock(&fs_info->tree_mod_seq_lock);
 413                                 return;
 414                         }
 415                         min_seq = cur_elem->seq;
 416                 }
 417         }
 418         spin_unlock(&fs_info->tree_mod_seq_lock);
 419
 420         /*
 421          * anything that's lower than the lowest existing (read: blocked)
 422          * sequence number can be removed from the tree.
 423          */
 424         tree_mod_log_write_lock(fs_info);
 425         tm_root = &fs_info->tree_mod_log;
 426         for (node = rb_first(tm_root); node; node = next) {
 427                 next = rb_next(node);
 428                 tm = container_of(node, struct tree_mod_elem, node);
 429                 if (tm->seq > min_seq)
 430                         continue;
 431                 rb_erase(node, tm_root);
 432                 kfree(tm);
 433         }
 434         tree_mod_log_write_unlock(fs_info);
 435 }
 436
 437 /*
 438  * key order of the log:
 439  *       node/leaf start address -> sequence
 440  *
 441  * The 'start address' is the logical address of the *new* root node
 442  * for root replace operations, or the logical address of the affected
 443  * block for all other operations.
 444  *
 445  * Note: must be called with write lock (tree_mod_log_write_lock).
 446  */
 447 static noinline int
 448 __tree_mod_log_insert(struct btrfs_fs_info *fs_info, struct tree_mod_elem *tm)
 449 {
 450         struct rb_root *tm_root;
 451         struct rb_node **new;
 452         struct rb_node *parent = NULL;
 453         struct tree_mod_elem *cur;
 454
 455         BUG_ON(!tm);
 456
 457         tm->seq = btrfs_inc_tree_mod_seq(fs_info);
 458
 459         tm_root = &fs_info->tree_mod_log;
 460         new = &tm_root->rb_node;
 461         while (*new) {
 462                 cur = container_of(*new, struct tree_mod_elem, node);
 463                 parent = *new;
 464                 if (cur->logical < tm->logical)
 465                         new = &((*new)->rb_left);
 466                 else if (cur->logical > tm->logical)
 467                         new = &((*new)->rb_right);
 468                 else if (cur->seq < tm->seq)
 469                         new = &((*new)->rb_left);
 470                 else if (cur->seq > tm->seq)
 471                         new = &((*new)->rb_right);
 472                 else
 473                         return -EEXIST;
 474         }
 475
 476         rb_link_node(&tm->node, parent, new);
 477         rb_insert_color(&tm->node, tm_root);
 478         return 0;
 479 }
 480
 481 /*
 482  * Determines if logging can be omitted. Returns 1 if it can. Otherwise, it
 483  * returns zero with the tree_mod_log_lock acquired. The caller must hold
 484  * this until all tree mod log insertions are recorded in the rb tree and then
 485  * call tree_mod_log_write_unlock() to release.
 486  */
 487 static inline int tree_mod_dont_log(struct btrfs_fs_info *fs_info,
 488                                     struct extent_buffer *eb) {
 489         smp_mb();
 490         if (list_empty(&(fs_info)->tree_mod_seq_list))
 491                 return 1;
 492         if (eb && btrfs_header_level(eb) == 0)
 493                 return 1;
 494
 495         tree_mod_log_write_lock(fs_info);
 496         if (list_empty(&(fs_info)->tree_mod_seq_list)) {
 497                 tree_mod_log_write_unlock(fs_info);
 498                 return 1;
 499         }
 500
 501         return 0;
 502 }
 503
 504 /* Similar to tree_mod_dont_log, but doesn't acquire any locks. */
 505 static inline int tree_mod_need_log(const struct btrfs_fs_info *fs_info,
 506                                     struct extent_buffer *eb)
 507 {
 508         smp_mb();
 509         if (list_empty(&(fs_info)->tree_mod_seq_list))
 510                 return 0;
 511         if (eb && btrfs_header_level(eb) == 0)
 512                 return 0;
 513
 514         return 1;
 515 }
 516
 517 static struct tree_mod_elem *
 518 alloc_tree_mod_elem(struct extent_buffer *eb, int slot,
 519                     enum mod_log_op op, gfp_t flags)
 520 {
 521         struct tree_mod_elem *tm;
 522
 523         tm = kzalloc(sizeof(*tm), flags);
 524         if (!tm)
 525                 return NULL;
 526
 527         tm->logical = eb->start;
 528         if (op != MOD_LOG_KEY_ADD) {
 529                 btrfs_node_key(eb, &tm->key, slot);
 530                 tm->blockptr = btrfs_node_blockptr(eb, slot);
 531         }
 532         tm->op = op;
 533         tm->slot = slot;
 534         tm->generation = btrfs_node_ptr_generation(eb, slot);
 535         RB_CLEAR_NODE(&tm->node);
 536
 537         return tm;
 538 }
 539
 540 static noinline int
 541 tree_mod_log_insert_key(struct btrfs_fs_info *fs_info,
 542                         struct extent_buffer *eb, int slot,
 543                         enum mod_log_op op, gfp_t flags)
 544 {
 545         struct tree_mod_elem *tm;
 546         int ret;
 547
 548         if (!tree_mod_need_log(fs_info, eb))
 549                 return 0;
 550
 551         tm = alloc_tree_mod_elem(eb, slot, op, flags);
 552         if (!tm)
 553                 return -ENOMEM;
 554
 555         if (tree_mod_dont_log(fs_info, eb)) {
 556                 kfree(tm);
 557                 return 0;
 558         }
 559
 560         ret = __tree_mod_log_insert(fs_info, tm);
 561         tree_mod_log_write_unlock(fs_info);
 562         if (ret)
 563                 kfree(tm);
 564
 565         return ret;
 566 }
 567
 568 static noinline int
 569 tree_mod_log_insert_move(struct btrfs_fs_info *fs_info,
 570                          struct extent_buffer *eb, int dst_slot, int src_slot,
 571                          int nr_items, gfp_t flags)
 572 {
 573         struct tree_mod_elem *tm = NULL;
 574         struct tree_mod_elem **tm_list = NULL;
 575         int ret = 0;
 576         int i;
 577         int locked = 0;
 578
 579         if (!tree_mod_need_log(fs_info, eb))
 580                 return 0;
 581
 582         tm_list = kcalloc(nr_items, sizeof(struct tree_mod_elem *), flags);
 583         if (!tm_list)
 584                 return -ENOMEM;
 585
 586         tm = kzalloc(sizeof(*tm), flags);
 587         if (!tm) {
 588                 ret = -ENOMEM;
 589                 goto free_tms;
 590         }
 591
 592         tm->logical = eb->start;
 593         tm->slot = src_slot;
 594         tm->move.dst_slot = dst_slot;
 595         tm->move.nr_items = nr_items;
 596         tm->op = MOD_LOG_MOVE_KEYS;
 597
 598         for (i = 0; i + dst_slot < src_slot && i < nr_items; i++) {
 599                 tm_list[i] = alloc_tree_mod_elem(eb, i + dst_slot,
 600                     MOD_LOG_KEY_REMOVE_WHILE_MOVING, flags);
 601                 if (!tm_list[i]) {
 602                         ret = -ENOMEM;
 603                         goto free_tms;
 604                 }
 605         }
 606
 607         if (tree_mod_dont_log(fs_info, eb))
 608                 goto free_tms;
 609         locked = 1;
 610
 611         /*
 612          * When we override something during the move, we log these removals.
 613          * This can only happen when we move towards the beginning of the
 614          * buffer, i.e. dst_slot < src_slot.
 615          */
 616         for (i = 0; i + dst_slot < src_slot && i < nr_items; i++) {
 617                 ret = __tree_mod_log_insert(fs_info, tm_list[i]);
 618                 if (ret)
 619                         goto free_tms;
 620         }
 621
 622         ret = __tree_mod_log_insert(fs_info, tm);
 623         if (ret)
 624                 goto free_tms;
 625         tree_mod_log_write_unlock(fs_info);
 626         kfree(tm_list);
 627
 628         return 0;
 629 free_tms:
 630         for (i = 0; i < nr_items; i++) {
 631                 if (tm_list[i] && !RB_EMPTY_NODE(&tm_list[i]->node))
 632                         rb_erase(&tm_list[i]->node, &fs_info->tree_mod_log);
 633                 kfree(tm_list[i]);
 634         }
 635         if (locked)
 636                 tree_mod_log_write_unlock(fs_info);
 637         kfree(tm_list);
 638         kfree(tm);
 639
 640         return ret;
 641 }
 642
 643 static inline int
 644 __tree_mod_log_free_eb(struct btrfs_fs_info *fs_info,
 645                        struct tree_mod_elem **tm_list,
 646                        int nritems)
 647 {
 648         int i, j;
 649         int ret;
 650
 651         for (i = nritems - 1; i >= 0; i--) {
 652                 ret = __tree_mod_log_insert(fs_info, tm_list[i]);
 653                 if (ret) {
 654                         for (j = nritems - 1; j > i; j--)
 655                                 rb_erase(&tm_list[j]->node,
 656                                          &fs_info->tree_mod_log);
 657                         return ret;
 658                 }
 659         }
 660
 661         return 0;
 662 }
 663
 664 static noinline int
 665 tree_mod_log_insert_root(struct btrfs_fs_info *fs_info,
 666                          struct extent_buffer *old_root,
 667                          struct extent_buffer *new_root, gfp_t flags,
 668                          int log_removal)
 669 {
 670         struct tree_mod_elem *tm = NULL;
 671         struct tree_mod_elem **tm_list = NULL;
 672         int nritems = 0;
 673         int ret = 0;
 674         int i;
 675
 676         if (!tree_mod_need_log(fs_info, NULL))
 677                 return 0;
 678
 679         if (log_removal && btrfs_header_level(old_root) > 0) {
 680                 nritems = btrfs_header_nritems(old_root);
 681                 tm_list = kcalloc(nritems, sizeof(struct tree_mod_elem *),
 682                                   flags);
 683                 if (!tm_list) {
 684                         ret = -ENOMEM;
 685                         goto free_tms;
 686                 }
 687                 for (i = 0; i < nritems; i++) {
 688                         tm_list[i] = alloc_tree_mod_elem(old_root, i,
 689                             MOD_LOG_KEY_REMOVE_WHILE_FREEING, flags);
 690                         if (!tm_list[i]) {
 691                                 ret = -ENOMEM;
 692                                 goto free_tms;
 693                         }
 694                 }
 695         }
 696
 697         tm = kzalloc(sizeof(*tm), flags);
 698         if (!tm) {
 699                 ret = -ENOMEM;
 700                 goto free_tms;
 701         }
 702
 703         tm->logical = new_root->start;
 704         tm->old_root.logical = old_root->start;
 705         tm->old_root.level = btrfs_header_level(old_root);
 706         tm->generation = btrfs_header_generation(old_root);
 707         tm->op = MOD_LOG_ROOT_REPLACE;
 708
 709         if (tree_mod_dont_log(fs_info, NULL))
 710                 goto free_tms;
 711
 712         if (tm_list)
 713                 ret = __tree_mod_log_free_eb(fs_info, tm_list, nritems);
 714         if (!ret)
 715                 ret = __tree_mod_log_insert(fs_info, tm);
 716
 717         tree_mod_log_write_unlock(fs_info);
 718         if (ret)
 719                 goto free_tms;
 720         kfree(tm_list);
 721
 722         return ret;
 723
 724 free_tms:
 725         if (tm_list) {
 726                 for (i = 0; i < nritems; i++)
 727                         kfree(tm_list[i]);
 728                 kfree(tm_list);
 729         }
 730         kfree(tm);
 731
 732         return ret;
 733 }
 734
 735 static struct tree_mod_elem *
 736 __tree_mod_log_search(struct btrfs_fs_info *fs_info, u64 start, u64 min_seq,
 737                       int smallest)
 738 {
 739         struct rb_root *tm_root;
 740         struct rb_node *node;
 741         struct tree_mod_elem *cur = NULL;
 742         struct tree_mod_elem *found = NULL;
 743
 744         tree_mod_log_read_lock(fs_info);
 745         tm_root = &fs_info->tree_mod_log;
 746         node = tm_root->rb_node;
 747         while (node) {
 748                 cur = container_of(node, struct tree_mod_elem, node);
 749                 if (cur->logical < start) {
 750                         node = node->rb_left;
 751                 } else if (cur->logical > start) {
 752                         node = node->rb_right;
 753                 } else if (cur->seq < min_seq) {
 754                         node = node->rb_left;
 755                 } else if (!smallest) {
 756                         /* we want the node with the highest seq */
 757                         if (found)
 758                                 BUG_ON(found->seq > cur->seq);
 759                         found = cur;
 760                         node = node->rb_left;
 761                 } else if (cur->seq > min_seq) {
 762                         /* we want the node with the smallest seq */
 763                         if (found)
 764                                 BUG_ON(found->seq < cur->seq);
 765                         found = cur;
 766                         node = node->rb_right;
 767                 } else {
 768                         found = cur;
 769                         break;
 770                 }
 771         }
 772         tree_mod_log_read_unlock(fs_info);
 773
 774         return found;
 775 }
 776
 777 /*
 778  * this returns the element from the log with the smallest time sequence
 779  * value that's in the log (the oldest log item). any element with a time
 780  * sequence lower than min_seq will be ignored.
 781  */
 782 static struct tree_mod_elem *
 783 tree_mod_log_search_oldest(struct btrfs_fs_info *fs_info, u64 start,
 784                            u64 min_seq)
 785 {
 786         return __tree_mod_log_search(fs_info, start, min_seq, 1);
 787 }
 788
 789 /*
 790  * this returns the element from the log with the largest time sequence
 791  * value that's in the log (the most recent log item). any element with
 792  * a time sequence lower than min_seq will be ignored.
 793  */
 794 static struct tree_mod_elem *
 795 tree_mod_log_search(struct btrfs_fs_info *fs_info, u64 start, u64 min_seq)
 796 {
 797         return __tree_mod_log_search(fs_info, start, min_seq, 0);
 798 }
 799
 800 static noinline int
 801 tree_mod_log_eb_copy(struct btrfs_fs_info *fs_info, struct extent_buffer *dst,
 802                      struct extent_buffer *src, unsigned long dst_offset,
 803                      unsigned long src_offset, int nr_items)
 804 {
 805         int ret = 0;
 806         struct tree_mod_elem **tm_list = NULL;
 807         struct tree_mod_elem **tm_list_add, **tm_list_rem;
 808         int i;
 809         int locked = 0;
 810
 811         if (!tree_mod_need_log(fs_info, NULL))
 812                 return 0;
 813
 814         if (btrfs_header_level(dst) == 0 && btrfs_header_level(src) == 0)
 815                 return 0;
 816
 817         tm_list = kcalloc(nr_items * 2, sizeof(struct tree_mod_elem *),
 818                           GFP_NOFS);
 819         if (!tm_list)
 820                 return -ENOMEM;
 821
 822         tm_list_add = tm_list;
 823         tm_list_rem = tm_list + nr_items;
 824         for (i = 0; i < nr_items; i++) {
 825                 tm_list_rem[i] = alloc_tree_mod_elem(src, i + src_offset,
 826                     MOD_LOG_KEY_REMOVE, GFP_NOFS);
 827                 if (!tm_list_rem[i]) {
 828                         ret = -ENOMEM;
 829                         goto free_tms;
 830                 }
 831
 832                 tm_list_add[i] = alloc_tree_mod_elem(dst, i + dst_offset,
 833                     MOD_LOG_KEY_ADD, GFP_NOFS);
 834                 if (!tm_list_add[i]) {
 835                         ret = -ENOMEM;
 836                         goto free_tms;
 837                 }
 838         }
 839
 840         if (tree_mod_dont_log(fs_info, NULL))
 841                 goto free_tms;
 842         locked = 1;
 843
 844         for (i = 0; i < nr_items; i++) {
 845                 ret = __tree_mod_log_insert(fs_info, tm_list_rem[i]);
 846                 if (ret)
 847                         goto free_tms;
 848                 ret = __tree_mod_log_insert(fs_info, tm_list_add[i]);
 849                 if (ret)
 850                         goto free_tms;
 851         }
 852
 853         tree_mod_log_write_unlock(fs_info);
 854         kfree(tm_list);
 855
 856         return 0;
 857
 858 free_tms:
 859         for (i = 0; i < nr_items * 2; i++) {
 860                 if (tm_list[i] && !RB_EMPTY_NODE(&tm_list[i]->node))
 861                         rb_erase(&tm_list[i]->node, &fs_info->tree_mod_log);
 862                 kfree(tm_list[i]);
 863         }
 864         if (locked)
 865                 tree_mod_log_write_unlock(fs_info);
 866         kfree(tm_list);
 867
 868         return ret;
 869 }
 870
 871 static inline void
 872 tree_mod_log_eb_move(struct btrfs_fs_info *fs_info, struct extent_buffer *dst,
 873                      int dst_offset, int src_offset, int nr_items)
 874 {
 875         int ret;
 876         ret = tree_mod_log_insert_move(fs_info, dst, dst_offset, src_offset,
 877                                        nr_items, GFP_NOFS);
 878         BUG_ON(ret < 0);
 879 }
 880
 881 static noinline void
 882 tree_mod_log_set_node_key(struct btrfs_fs_info *fs_info,
 883                           struct extent_buffer *eb, int slot, int atomic)
 884 {
 885         int ret;
 886
 887         ret = tree_mod_log_insert_key(fs_info, eb, slot,
 888                                         MOD_LOG_KEY_REPLACE,
 889                                         atomic ? GFP_ATOMIC : GFP_NOFS);
 890         BUG_ON(ret < 0);
 891 }
 892
 893 static noinline int
 894 tree_mod_log_free_eb(struct btrfs_fs_info *fs_info, struct extent_buffer *eb)
 895 {
 896         struct tree_mod_elem **tm_list = NULL;
 897         int nritems = 0;
 898         int i;
 899         int ret = 0;
 900
 901         if (btrfs_header_level(eb) == 0)
 902                 return 0;
 903
 904         if (!tree_mod_need_log(fs_info, NULL))
 905                 return 0;
 906
 907         nritems = btrfs_header_nritems(eb);
 908         tm_list = kcalloc(nritems, sizeof(struct tree_mod_elem *), GFP_NOFS);
 909         if (!tm_list)
 910                 return -ENOMEM;
 911
 912         for (i = 0; i < nritems; i++) {
 913                 tm_list[i] = alloc_tree_mod_elem(eb, i,
 914                     MOD_LOG_KEY_REMOVE_WHILE_FREEING, GFP_NOFS);
 915                 if (!tm_list[i]) {
 916                         ret = -ENOMEM;
 917                         goto free_tms;
 918                 }
 919         }
 920
 921         if (tree_mod_dont_log(fs_info, eb))
 922                 goto free_tms;
 923
 924         ret = __tree_mod_log_free_eb(fs_info, tm_list, nritems);
 925         tree_mod_log_write_unlock(fs_info);
 926         if (ret)
 927                 goto free_tms;
 928         kfree(tm_list);
 929
 930         return 0;
 931
 932 free_tms:
 933         for (i = 0; i < nritems; i++)
 934                 kfree(tm_list[i]);
 935         kfree(tm_list);
 936
 937         return ret;
 938 }
 939
 940 static noinline void
 941 tree_mod_log_set_root_pointer(struct btrfs_root *root,
 942                               struct extent_buffer *new_root_node,
 943                               int log_removal)
 944 {
 945         int ret;
 946         ret = tree_mod_log_insert_root(root->fs_info, root->node,
 947                                        new_root_node, GFP_NOFS, log_removal);
 948         BUG_ON(ret < 0);
 949 }
 950
 951 /*
 952  * check if the tree block can be shared by multiple trees
 953  */
 954 int btrfs_block_can_be_shared(struct btrfs_root *root,
 955                               struct extent_buffer *buf)
 956 {
 957         /*
 958          * Tree blocks not in reference counted trees and tree roots
 959          * are never shared. If a block was allocated after the last
 960          * snapshot and the block was not allocated by tree relocation,
 961          * we know the block is not shared.
 962          */
 963         if (test_bit(BTRFS_ROOT_REF_COWS, &root->state) &&
 964             buf != root->node && buf != root->commit_root &&
 965             (btrfs_header_generation(buf) <=
 966              btrfs_root_last_snapshot(&root->root_item) ||
 967              btrfs_header_flag(buf, BTRFS_HEADER_FLAG_RELOC)))
 968                 return 1;
 969 #ifdef BTRFS_COMPAT_EXTENT_TREE_V0
 970         if (test_bit(BTRFS_ROOT_REF_COWS, &root->state) &&
 971             btrfs_header_backref_rev(buf) < BTRFS_MIXED_BACKREF_REV)
 972                 return 1;
 973 #endif
 974         return 0;
 975 }
 976
 977 static noinline int update_ref_for_cow(struct btrfs_trans_handle *trans,
 978                                        struct btrfs_root *root,
 979                                        struct extent_buffer *buf,
 980                                        struct extent_buffer *cow,
 981                                        int *last_ref)
 982 {
 983         u64 refs;
 984         u64 owner;
 985         u64 flags;
 986         u64 new_flags = 0;
 987         int ret;
 988
 989         /*
 990          * Backrefs update rules:
 991          *
 992          * Always use full backrefs for extent pointers in tree block
 993          * allocated by tree relocation.
 994          *
 995          * If a shared tree block is no longer referenced by its owner
 996          * tree (btrfs_header_owner(buf) == root->root_key.objectid),
 997          * use full backrefs for extent pointers in tree block.
 998          *
 999          * If a tree block is been relocating
1000          * (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID),
1001          * use full backrefs for extent pointers in tree block.
1002          * The reason for this is some operations (such as drop tree)
1003          * are only allowed for blocks use full backrefs.
1004          */
1005
1006         if (btrfs_block_can_be_shared(root, buf)) {
1007                 ret = btrfs_lookup_extent_info(trans, root, buf->start,
1008                                                btrfs_header_level(buf), 1,
1009                                                &refs, &flags);
1010                 if (ret)
1011                         return ret;
1012                 if (refs == 0) {
1013                         ret = -EROFS;
1014                         btrfs_handle_fs_error(root->fs_info, ret, NULL);
1015                         return ret;
1016                 }
1017         } else {
1018                 refs = 1;
1019                 if (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID ||
1020                     btrfs_header_backref_rev(buf) < BTRFS_MIXED_BACKREF_REV)
1021                         flags = BTRFS_BLOCK_FLAG_FULL_BACKREF;
1022                 else
1023                         flags = 0;
1024         }
1025
1026         owner = btrfs_header_owner(buf);
1027         BUG_ON(owner == BTRFS_TREE_RELOC_OBJECTID &&
1028                !(flags & BTRFS_BLOCK_FLAG_FULL_BACKREF));
1029
1030         if (refs > 1) {
1031                 if ((owner == root->root_key.objectid ||
1032                      root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID) &&
1033                     !(flags & BTRFS_BLOCK_FLAG_FULL_BACKREF)) {
1034                         ret = btrfs_inc_ref(trans, root, buf, 1);
1035                         BUG_ON(ret); /* -ENOMEM */
1036
1037                         if (root->root_key.objectid ==
1038                             BTRFS_TREE_RELOC_OBJECTID) {
1039                                 ret = btrfs_dec_ref(trans, root, buf, 0);
1040                                 BUG_ON(ret); /* -ENOMEM */
1041                                 ret = btrfs_inc_ref(trans, root, cow, 1);
1042                                 BUG_ON(ret); /* -ENOMEM */
1043                         }
1044                         new_flags |= BTRFS_BLOCK_FLAG_FULL_BACKREF;
1045                 } else {
1046
1047                         if (root->root_key.objectid ==
1048                             BTRFS_TREE_RELOC_OBJECTID)
1049                                 ret = btrfs_inc_ref(trans, root, cow, 1);
1050                         else
1051                                 ret = btrfs_inc_ref(trans, root, cow, 0);
1052                         BUG_ON(ret); /* -ENOMEM */
1053                 }
1054                 if (new_flags != 0) {
1055                         int level = btrfs_header_level(buf);
1056
1057                         ret = btrfs_set_disk_extent_flags(trans, root,
1058                                                           buf->start,
1059                                                           buf->len,
1060                                                           new_flags, level, 0);
1061                         if (ret)
1062                                 return ret;
1063                 }
1064         } else {
1065                 if (flags & BTRFS_BLOCK_FLAG_FULL_BACKREF) {
1066                         if (root->root_key.objectid ==
1067                             BTRFS_TREE_RELOC_OBJECTID)
1068                                 ret = btrfs_inc_ref(trans, root, cow, 1);
1069                         else
1070                                 ret = btrfs_inc_ref(trans, root, cow, 0);
1071                         BUG_ON(ret); /* -ENOMEM */
1072                         ret = btrfs_dec_ref(trans, root, buf, 1);
1073                         BUG_ON(ret); /* -ENOMEM */
1074                 }
1075                 clean_tree_block(trans, root->fs_info, buf);
1076                 *last_ref = 1;
1077         }
1078         return 0;
1079 }
1080
1081 /*
1082  * does the dirty work in cow of a single block.  The parent block (if
1083  * supplied) is updated to point to the new cow copy.  The new buffer is marked
1084  * dirty and returned locked.  If you modify the block it needs to be marked
1085  * dirty again.
1086  *
1087  * search_start -- an allocation hint for the new block
1088  *
1089  * empty_size -- a hint that you plan on doing more cow.  This is the size in
1090  * bytes the allocator should try to find free next to the block it returns.
1091  * This is just a hint and may be ignored by the allocator.
1092  */
1093 static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans,
1094                              struct btrfs_root *root,
1095                              struct extent_buffer *buf,
1096                              struct extent_buffer *parent, int parent_slot,
1097                              struct extent_buffer **cow_ret,
1098                              u64 search_start, u64 empty_size)
1099 {
1100         struct btrfs_disk_key disk_key;
1101         struct extent_buffer *cow;
1102         int level, ret;
1103         int last_ref = 0;
1104         int unlock_orig = 0;
1105         u64 parent_start;
1106
1107         if (*cow_ret == buf)
1108                 unlock_orig = 1;
1109
1110         btrfs_assert_tree_locked(buf);
1111
1112         WARN_ON(test_bit(BTRFS_ROOT_REF_COWS, &root->state) &&
1113                 trans->transid != root->fs_info->running_transaction->transid);
1114         WARN_ON(test_bit(BTRFS_ROOT_REF_COWS, &root->state) &&
1115                 trans->transid != root->last_trans);
1116
1117         level = btrfs_header_level(buf);
1118
1119         if (level == 0)
1120                 btrfs_item_key(buf, &disk_key, 0);
1121         else
1122                 btrfs_node_key(buf, &disk_key, 0);
1123
1124         if (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID) {
1125                 if (parent)
1126                         parent_start = parent->start;
1127                 else
1128                         parent_start = 0;
1129         } else
1130                 parent_start = 0;
1131
1132         cow = btrfs_alloc_tree_block(trans, root, parent_start,
1133                         root->root_key.objectid, &disk_key, level,
1134                         search_start, empty_size);
1135         if (IS_ERR(cow))
1136                 return PTR_ERR(cow);
1137
1138         /* cow is set to blocking by btrfs_init_new_buffer */
1139
1140         copy_extent_buffer(cow, buf, 0, 0, cow->len);
1141         btrfs_set_header_bytenr(cow, cow->start);
1142         btrfs_set_header_generation(cow, trans->transid);
1143         btrfs_set_header_backref_rev(cow, BTRFS_MIXED_BACKREF_REV);
1144         btrfs_clear_header_flag(cow, BTRFS_HEADER_FLAG_WRITTEN |
1145                                      BTRFS_HEADER_FLAG_RELOC);
1146         if (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID)
1147                 btrfs_set_header_flag(cow, BTRFS_HEADER_FLAG_RELOC);
1148         else
1149                 btrfs_set_header_owner(cow, root->root_key.objectid);
1150
1151         write_extent_buffer(cow, root->fs_info->fsid, btrfs_header_fsid(),
1152                             BTRFS_FSID_SIZE);
1153
1154         ret = update_ref_for_cow(trans, root, buf, cow, &last_ref);
1155         if (ret) {
1156                 btrfs_abort_transaction(trans, ret);
1157                 return ret;
1158         }
1159
1160         if (test_bit(BTRFS_ROOT_REF_COWS, &root->state)) {
1161                 ret = btrfs_reloc_cow_block(trans, root, buf, cow);
1162                 if (ret) {
1163                         btrfs_abort_transaction(trans, ret);
1164                         return ret;
1165                 }
1166         }
1167
1168         if (buf == root->node) {
1169                 WARN_ON(parent && parent != buf);
1170                 if (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID ||
1171                     btrfs_header_backref_rev(buf) < BTRFS_MIXED_BACKREF_REV)
1172                         parent_start = buf->start;
1173                 else
1174                         parent_start = 0;
1175
1176                 extent_buffer_get(cow);
1177                 tree_mod_log_set_root_pointer(root, cow, 1);
1178                 rcu_assign_pointer(root->node, cow);
1179
1180                 btrfs_free_tree_block(trans, root, buf, parent_start,
1181                                       last_ref);
1182                 free_extent_buffer(buf);
1183                 add_root_to_dirty_list(root);
1184         } else {
1185                 if (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID)
1186                         parent_start = parent->start;
1187                 else
1188                         parent_start = 0;
1189
1190                 WARN_ON(trans->transid != btrfs_header_generation(parent));
1191                 tree_mod_log_insert_key(root->fs_info, parent, parent_slot,
1192                                         MOD_LOG_KEY_REPLACE, GFP_NOFS);
1193                 btrfs_set_node_blockptr(parent, parent_slot,
1194                                         cow->start);
1195                 btrfs_set_node_ptr_generation(parent, parent_slot,
1196                                               trans->transid);
1197                 btrfs_mark_buffer_dirty(parent);
1198                 if (last_ref) {
1199                         ret = tree_mod_log_free_eb(root->fs_info, buf);
1200                         if (ret) {
1201                                 btrfs_abort_transaction(trans, ret);
1202                                 return ret;
1203                         }
1204                 }
1205                 btrfs_free_tree_block(trans, root, buf, parent_start,
1206                                       last_ref);
1207         }
1208         if (unlock_orig)
1209                 btrfs_tree_unlock(buf);
1210         free_extent_buffer_stale(buf);
1211         btrfs_mark_buffer_dirty(cow);
1212         *cow_ret = cow;
1213         return 0;
1214 }
1215
1216 /*
1217  * returns the logical address of the oldest predecessor of the given root.
1218  * entries older than time_seq are ignored.
1219  */
1220 static struct tree_mod_elem *
1221 __tree_mod_log_oldest_root(struct btrfs_fs_info *fs_info,
1222                            struct extent_buffer *eb_root, u64 time_seq)
1223 {
1224         struct tree_mod_elem *tm;
1225         struct tree_mod_elem *found = NULL;
1226         u64 root_logical = eb_root->start;
1227         int looped = 0;
1228
1229         if (!time_seq)
1230                 return NULL;
1231
1232         /*
1233          * the very last operation that's logged for a root is the
1234          * replacement operation (if it is replaced at all). this has
1235          * the logical address of the *new* root, making it the very
1236          * first operation that's logged for this root.
1237          */
1238         while (1) {
1239                 tm = tree_mod_log_search_oldest(fs_info, root_logical,
1240                                                 time_seq);
1241                 if (!looped && !tm)
1242                         return NULL;
1243                 /*
1244                  * if there are no tree operation for the oldest root, we simply
1245                  * return it. this should only happen if that (old) root is at
1246                  * level 0.
1247                  */
1248                 if (!tm)
1249                         break;
1250
1251                 /*
1252                  * if there's an operation that's not a root replacement, we
1253                  * found the oldest version of our root. normally, we'll find a
1254                  * MOD_LOG_KEY_REMOVE_WHILE_FREEING operation here.
1255                  */
1256                 if (tm->op != MOD_LOG_ROOT_REPLACE)
1257                         break;
1258
1259                 found = tm;
1260                 root_logical = tm->old_root.logical;
1261                 looped = 1;
1262         }
1263
1264         /* if there's no old root to return, return what we found instead */
1265         if (!found)
1266                 found = tm;
1267
1268         return found;
1269 }
1270
1271 /*
1272  * tm is a pointer to the first operation to rewind within eb. then, all
1273  * previous operations will be rewound (until we reach something older than
1274  * time_seq).
1275  */
1276 static void
1277 __tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct extent_buffer *eb,
1278                       u64 time_seq, struct tree_mod_elem *first_tm)
1279 {
1280         u32 n;
1281         struct rb_node *next;
1282         struct tree_mod_elem *tm = first_tm;
1283         unsigned long o_dst;
1284         unsigned long o_src;
1285         unsigned long p_size = sizeof(struct btrfs_key_ptr);
1286
1287         n = btrfs_header_nritems(eb);
1288         tree_mod_log_read_lock(fs_info);
1289         while (tm && tm->seq >= time_seq) {
1290                 /*
1291                  * all the operations are recorded with the operator used for
1292                  * the modification. as we're going backwards, we do the
1293                  * opposite of each operation here.
1294                  */
1295                 switch (tm->op) {
1296                 case MOD_LOG_KEY_REMOVE_WHILE_FREEING:
1297                         BUG_ON(tm->slot < n);
1298                         /* Fallthrough */
1299                 case MOD_LOG_KEY_REMOVE_WHILE_MOVING:
1300                 case MOD_LOG_KEY_REMOVE:
1301                         btrfs_set_node_key(eb, &tm->key, tm->slot);
1302                         btrfs_set_node_blockptr(eb, tm->slot, tm->blockptr);
1303                         btrfs_set_node_ptr_generation(eb, tm->slot,
1304                                                       tm->generation);
1305                         n++;
1306                         break;
1307                 case MOD_LOG_KEY_REPLACE:
1308                         BUG_ON(tm->slot >= n);
1309                         btrfs_set_node_key(eb, &tm->key, tm->slot);
1310                         btrfs_set_node_blockptr(eb, tm->slot, tm->blockptr);
1311                         btrfs_set_node_ptr_generation(eb, tm->slot,
1312                                                       tm->generation);
1313                         break;
1314                 case MOD_LOG_KEY_ADD:
1315                         /* if a move operation is needed it's in the log */
1316                         n--;
1317                         break;
1318                 case MOD_LOG_MOVE_KEYS:
1319                         o_dst = btrfs_node_key_ptr_offset(tm->slot);
1320                         o_src = btrfs_node_key_ptr_offset(tm->move.dst_slot);
1321                         memmove_extent_buffer(eb, o_dst, o_src,
1322                                               tm->move.nr_items * p_size);
1323                         break;
1324                 case MOD_LOG_ROOT_REPLACE:
1325                         /*
1326                          * this operation is special. for roots, this must be
1327                          * handled explicitly before rewinding.
1328                          * for non-roots, this operation may exist if the node
1329                          * was a root: root A -> child B; then A gets empty and
1330                          * B is promoted to the new root. in the mod log, we'll
1331                          * have a root-replace operation for B, a tree block
1332                          * that is no root. we simply ignore that operation.
1333                          */
1334                         break;
1335                 }
1336                 next = rb_next(&tm->node);
1337                 if (!next)
1338                         break;
1339                 tm = container_of(next, struct tree_mod_elem, node);
1340                 if (tm->logical != first_tm->logical)
1341                         break;
1342         }
1343         tree_mod_log_read_unlock(fs_info);
1344         btrfs_set_header_nritems(eb, n);
1345 }
1346
1347 /*
1348  * Called with eb read locked. If the buffer cannot be rewound, the same buffer
1349  * is returned. If rewind operations happen, a fresh buffer is returned. The
1350  * returned buffer is always read-locked. If the returned buffer is not the
1351  * input buffer, the lock on the input buffer is released and the input buffer
1352  * is freed (its refcount is decremented).
1353  */
1354 static struct extent_buffer *
1355 tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct btrfs_path *path,
1356                     struct extent_buffer *eb, u64 time_seq)
1357 {
1358         struct extent_buffer *eb_rewin;
1359         struct tree_mod_elem *tm;
1360
1361         if (!time_seq)
1362                 return eb;
1363
1364         if (btrfs_header_level(eb) == 0)
1365                 return eb;
1366
1367         tm = tree_mod_log_search(fs_info, eb->start, time_seq);
1368         if (!tm)
1369                 return eb;
1370
1371         btrfs_set_path_blocking(path);
1372         btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK);
1373
1374         if (tm->op == MOD_LOG_KEY_REMOVE_WHILE_FREEING) {
1375                 BUG_ON(tm->slot != 0);
1376                 eb_rewin = alloc_dummy_extent_buffer(fs_info, eb->start,
1377                                                 eb->len);
1378                 if (!eb_rewin) {
1379                         btrfs_tree_read_unlock_blocking(eb);
1380                         free_extent_buffer(eb);
1381                         return NULL;
1382                 }
1383                 btrfs_set_header_bytenr(eb_rewin, eb->start);
1384                 btrfs_set_header_backref_rev(eb_rewin,
1385                                              btrfs_header_backref_rev(eb));
1386                 btrfs_set_header_owner(eb_rewin, btrfs_header_owner(eb));
1387                 btrfs_set_header_level(eb_rewin, btrfs_header_level(eb));
1388         } else {
1389                 eb_rewin = btrfs_clone_extent_buffer(eb);
1390                 if (!eb_rewin) {
1391                         btrfs_tree_read_unlock_blocking(eb);
1392                         free_extent_buffer(eb);
1393                         return NULL;
1394                 }
1395         }
1396
1397         btrfs_clear_path_blocking(path, NULL, BTRFS_READ_LOCK);
1398         btrfs_tree_read_unlock_blocking(eb);
1399         free_extent_buffer(eb);
1400
1401         extent_buffer_get(eb_rewin);
1402         btrfs_tree_read_lock(eb_rewin);
1403         __tree_mod_log_rewind(fs_info, eb_rewin, time_seq, tm);
1404         WARN_ON(btrfs_header_nritems(eb_rewin) >
1405                 BTRFS_NODEPTRS_PER_BLOCK(fs_info->tree_root));
1406
1407         return eb_rewin;
1408 }
1409
1410 /*
1411  * get_old_root() rewinds the state of @root's root node to the given @time_seq
1412  * value. If there are no changes, the current root->root_node is returned. If
1413  * anything changed in between, there's a fresh buffer allocated on which the
1414  * rewind operations are done. In any case, the returned buffer is read locked.
1415  * Returns NULL on error (with no locks held).
1416  */
1417 static inline struct extent_buffer *
1418 get_old_root(struct btrfs_root *root, u64 time_seq)
1419 {
1420         struct tree_mod_elem *tm;
1421         struct extent_buffer *eb = NULL;
1422         struct extent_buffer *eb_root;
1423         struct extent_buffer *old;
1424         struct tree_mod_root *old_root = NULL;
1425         u64 old_generation = 0;
1426         u64 logical;
1427
1428         eb_root = btrfs_read_lock_root_node(root);
1429         tm = __tree_mod_log_oldest_root(root->fs_info, eb_root, time_seq);
1430         if (!tm)
1431                 return eb_root;
1432
1433         if (tm->op == MOD_LOG_ROOT_REPLACE) {
1434                 old_root = &tm->old_root;
1435                 old_generation = tm->generation;
1436                 logical = old_root->logical;
1437         } else {
1438                 logical = eb_root->start;
1439         }
1440
1441         tm = tree_mod_log_search(root->fs_info, logical, time_seq);
1442         if (old_root && tm && tm->op != MOD_LOG_KEY_REMOVE_WHILE_FREEING) {
1443                 btrfs_tree_read_unlock(eb_root);
1444                 free_extent_buffer(eb_root);
1445                 old = read_tree_block(root, logical, 0);
1446                 if (WARN_ON(IS_ERR(old) || !extent_buffer_uptodate(old))) {
1447                         if (!IS_ERR(old))
1448                                 free_extent_buffer(old);
1449                         btrfs_warn(root->fs_info,
1450                                 "failed to read tree block %llu from get_old_root", logical);
1451                 } else {
1452                         eb = btrfs_clone_extent_buffer(old);
1453                         free_extent_buffer(old);
1454                 }
1455         } else if (old_root) {
1456                 btrfs_tree_read_unlock(eb_root);
1457                 free_extent_buffer(eb_root);
1458                 eb = alloc_dummy_extent_buffer(root->fs_info, logical,
1459                                         root->nodesize);
1460         } else {
1461                 btrfs_set_lock_blocking_rw(eb_root, BTRFS_READ_LOCK);
1462                 eb = btrfs_clone_extent_buffer(eb_root);
1463                 btrfs_tree_read_unlock_blocking(eb_root);
1464                 free_extent_buffer(eb_root);
1465         }
1466
1467         if (!eb)
1468                 return NULL;
1469         extent_buffer_get(eb);
1470         btrfs_tree_read_lock(eb);
1471         if (old_root) {
1472                 btrfs_set_header_bytenr(eb, eb->start);
1473                 btrfs_set_header_backref_rev(eb, BTRFS_MIXED_BACKREF_REV);
1474                 btrfs_set_header_owner(eb, btrfs_header_owner(eb_root));
1475                 btrfs_set_header_level(eb, old_root->level);
1476                 btrfs_set_header_generation(eb, old_generation);
1477         }
1478         if (tm)
1479                 __tree_mod_log_rewind(root->fs_info, eb, time_seq, tm);
1480         else
1481                 WARN_ON(btrfs_header_level(eb) != 0);
1482         WARN_ON(btrfs_header_nritems(eb) > BTRFS_NODEPTRS_PER_BLOCK(root));
1483
1484         return eb;
1485 }
1486
1487 int btrfs_old_root_level(struct btrfs_root *root, u64 time_seq)
1488 {
1489         struct tree_mod_elem *tm;
1490         int level;
1491         struct extent_buffer *eb_root = btrfs_root_node(root);
1492
1493         tm = __tree_mod_log_oldest_root(root->fs_info, eb_root, time_seq);
1494         if (tm && tm->op == MOD_LOG_ROOT_REPLACE) {
1495                 level = tm->old_root.level;
1496         } else {
1497                 level = btrfs_header_level(eb_root);
1498         }
1499         free_extent_buffer(eb_root);
1500
1501         return level;
1502 }
1503
1504 static inline int should_cow_block(struct btrfs_trans_handle *trans,
1505                                    struct btrfs_root *root,
1506                                    struct extent_buffer *buf)
1507 {
1508         if (btrfs_is_testing(root->fs_info))
1509                 return 0;
1510
1511         /* ensure we can see the force_cow */
1512         smp_rmb();
1513
1514         /*
1515          * We do not need to cow a block if
1516          * 1) this block is not created or changed in this transaction;
1517          * 2) this block does not belong to TREE_RELOC tree;
1518          * 3) the root is not forced COW.
1519          *
1520          * What is forced COW:
1521          *    when we create snapshot during committing the transaction,
1522          *    after we've finished coping src root, we must COW the shared
1523          *    block to ensure the metadata consistency.
1524          */
1525         if (btrfs_header_generation(buf) == trans->transid &&
1526             !btrfs_header_flag(buf, BTRFS_HEADER_FLAG_WRITTEN) &&
1527             !(root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID &&
1528               btrfs_header_flag(buf, BTRFS_HEADER_FLAG_RELOC)) &&
1529             !test_bit(BTRFS_ROOT_FORCE_COW, &root->state))
1530                 return 0;
1531         return 1;
1532 }
1533
1534 /*
1535  * cows a single block, see __btrfs_cow_block for the real work.
1536  * This version of it has extra checks so that a block isn't COWed more than
1537  * once per transaction, as long as it hasn't been written yet
1538  */
1539 noinline int btrfs_cow_block(struct btrfs_trans_handle *trans,
1540                     struct btrfs_root *root, struct extent_buffer *buf,
1541                     struct extent_buffer *parent, int parent_slot,
1542                     struct extent_buffer **cow_ret)
1543 {
1544         u64 search_start;
1545         int ret;
1546
1547         if (trans->transaction != root->fs_info->running_transaction)
1548                 WARN(1, KERN_CRIT "trans %llu running %llu\n",
1549                        trans->transid,
1550                        root->fs_info->running_transaction->transid);
1551
1552         if (trans->transid != root->fs_info->generation)
1553                 WARN(1, KERN_CRIT "trans %llu running %llu\n",
1554                        trans->transid, root->fs_info->generation);
1555
1556         if (!should_cow_block(trans, root, buf)) {
1557                 trans->dirty = true;
1558                 *cow_ret = buf;
1559                 return 0;
1560         }
1561
1562         search_start = buf->start & ~((u64)SZ_1G - 1);
1563
1564         if (parent)
1565                 btrfs_set_lock_blocking(parent);
1566         btrfs_set_lock_blocking(buf);
1567
1568         ret = __btrfs_cow_block(trans, root, buf, parent,
1569                                  parent_slot, cow_ret, search_start, 0);
1570
1571         trace_btrfs_cow_block(root, buf, *cow_ret);
1572
1573         return ret;
1574 }
1575
1576 /*
1577  * helper function for defrag to decide if two blocks pointed to by a
1578  * node are actually close by
1579  */
1580 static int close_blocks(u64 blocknr, u64 other, u32 blocksize)
1581 {
1582         if (blocknr < other && other - (blocknr + blocksize) < 32768)
1583                 return 1;
1584         if (blocknr > other && blocknr - (other + blocksize) < 32768)
1585                 return 1;
1586         return 0;
1587 }
1588
1589 /*
1590  * compare two keys in a memcmp fashion
1591  */
1592 static int comp_keys(struct btrfs_disk_key *disk, struct btrfs_key *k2)
1593 {
1594         struct btrfs_key k1;
1595
1596         btrfs_disk_key_to_cpu(&k1, disk);
1597
1598         return btrfs_comp_cpu_keys(&k1, k2);
1599 }
1600
1601 /*
1602  * same as comp_keys only with two btrfs_key's
1603  */
1604 int btrfs_comp_cpu_keys(struct btrfs_key *k1, struct btrfs_key *k2)
1605 {
1606         if (k1->objectid > k2->objectid)
1607                 return 1;
1608         if (k1->objectid < k2->objectid)
1609                 return -1;
1610         if (k1->type > k2->type)
1611                 return 1;
1612         if (k1->type < k2->type)
1613                 return -1;
1614         if (k1->offset > k2->offset)
1615                 return 1;
1616         if (k1->offset < k2->offset)
1617                 return -1;
1618         return 0;
1619 }
1620
1621 /*
1622  * this is used by the defrag code to go through all the
1623  * leaves pointed to by a node and reallocate them so that
1624  * disk order is close to key order
1625  */
1626 int btrfs_realloc_node(struct btrfs_trans_handle *trans,
1627                        struct btrfs_root *root, struct extent_buffer *parent,
1628                        int start_slot, u64 *last_ret,
1629                        struct btrfs_key *progress)
1630 {
1631         struct extent_buffer *cur;
1632         u64 blocknr;
1633         u64 gen;
1634         u64 search_start = *last_ret;
1635         u64 last_block = 0;
1636         u64 other;
1637         u32 parent_nritems;
1638         int end_slot;
1639         int i;
1640         int err = 0;
1641         int parent_level;
1642         int uptodate;
1643         u32 blocksize;
1644         int progress_passed = 0;
1645         struct btrfs_disk_key disk_key;
1646
1647         parent_level = btrfs_header_level(parent);
1648
1649         WARN_ON(trans->transaction != root->fs_info->running_transaction);
1650         WARN_ON(trans->transid != root->fs_info->generation);
1651
1652         parent_nritems = btrfs_header_nritems(parent);
1653         blocksize = root->nodesize;
1654         end_slot = parent_nritems - 1;
1655
1656         if (parent_nritems <= 1)
1657                 return 0;
1658
1659         btrfs_set_lock_blocking(parent);
1660
1661         for (i = start_slot; i <= end_slot; i++) {
1662                 int close = 1;
1663
1664                 btrfs_node_key(parent, &disk_key, i);
1665                 if (!progress_passed && comp_keys(&disk_key, progress) < 0)
1666                         continue;
1667
1668                 progress_passed = 1;
1669                 blocknr = btrfs_node_blockptr(parent, i);
1670                 gen = btrfs_node_ptr_generation(parent, i);
1671                 if (last_block == 0)
1672                         last_block = blocknr;
1673
1674                 if (i > 0) {
1675                         other = btrfs_node_blockptr(parent, i - 1);
1676                         close = close_blocks(blocknr, other, blocksize);
1677                 }
1678                 if (!close && i < end_slot) {
1679                         other = btrfs_node_blockptr(parent, i + 1);
1680                         close = close_blocks(blocknr, other, blocksize);
1681                 }
1682                 if (close) {
1683                         last_block = blocknr;
1684                         continue;
1685                 }
1686
1687                 cur = btrfs_find_tree_block(root->fs_info, blocknr);
1688                 if (cur)
1689                         uptodate = btrfs_buffer_uptodate(cur, gen, 0);
1690                 else
1691                         uptodate = 0;
1692                 if (!cur || !uptodate) {
1693                         if (!cur) {
1694                                 cur = read_tree_block(root, blocknr, gen);
1695                                 if (IS_ERR(cur)) {
1696                                         return PTR_ERR(cur);
1697                                 } else if (!extent_buffer_uptodate(cur)) {
1698                                         free_extent_buffer(cur);
1699                                         return -EIO;
1700                                 }
1701                         } else if (!uptodate) {
1702                                 err = btrfs_read_buffer(cur, gen);
1703                                 if (err) {
1704                                         free_extent_buffer(cur);
1705                                         return err;
1706                                 }
1707                         }
1708                 }
1709                 if (search_start == 0)
1710                         search_start = last_block;
1711
1712                 btrfs_tree_lock(cur);
1713                 btrfs_set_lock_blocking(cur);
1714                 err = __btrfs_cow_block(trans, root, cur, parent, i,
1715                                         &cur, search_start,
1716                                         min(16 * blocksize,
1717                                             (end_slot - i) * blocksize));
1718                 if (err) {
1719                         btrfs_tree_unlock(cur);
1720                         free_extent_buffer(cur);
1721                         break;
1722                 }
1723                 search_start = cur->start;
1724                 last_block = cur->start;
1725                 *last_ret = search_start;
1726                 btrfs_tree_unlock(cur);
1727                 free_extent_buffer(cur);
1728         }
1729         return err;
1730 }
1731
1732 /*
1733  * The leaf data grows from end-to-front in the node.
1734  * this returns the address of the start of the last item,
1735  * which is the stop of the leaf data stack
1736  */
1737 static inline unsigned int leaf_data_end(struct btrfs_root *root,
1738                                          struct extent_buffer *leaf)
1739 {
1740         u32 nr = btrfs_header_nritems(leaf);
1741         if (nr == 0)
1742                 return BTRFS_LEAF_DATA_SIZE(root);
1743         return btrfs_item_offset_nr(leaf, nr - 1);
1744 }
1745
1746
1747 /*
1748  * search for key in the extent_buffer.  The items start at offset p,
1749  * and they are item_size apart.  There are 'max' items in p.
1750  *
1751  * the slot in the array is returned via slot, and it points to
1752  * the place where you would insert key if it is not found in
1753  * the array.
1754  *
1755  * slot may point to max if the key is bigger than all of the keys
1756  */
1757 static noinline int generic_bin_search(struct extent_buffer *eb,
1758                                        unsigned long p,
1759                                        int item_size, struct btrfs_key *key,
1760                                        int max, int *slot)
1761 {
1762         int low = 0;
1763         int high = max;
1764         int mid;
1765         int ret;
1766         struct btrfs_disk_key *tmp = NULL;
1767         struct btrfs_disk_key unaligned;
1768         unsigned long offset;
1769         char *kaddr = NULL;
1770         unsigned long map_start = 0;
1771         unsigned long map_len = 0;
1772         int err;
1773
1774         if (low > high) {
1775                 btrfs_err(eb->fs_info,
1776                  "%s: low (%d) > high (%d) eb %llu owner %llu level %d",
1777                           __func__, low, high, eb->start,
1778                           btrfs_header_owner(eb), btrfs_header_level(eb));
1779                 return -EINVAL;
1780         }
1781
1782         while (low < high) {
1783                 mid = (low + high) / 2;
1784                 offset = p + mid * item_size;
1785
1786                 if (!kaddr || offset < map_start ||
1787                     (offset + sizeof(struct btrfs_disk_key)) >
1788                     map_start + map_len) {
1789
1790                         err = map_private_extent_buffer(eb, offset,
1791                                                 sizeof(struct btrfs_disk_key),
1792                                                 &kaddr, &map_start, &map_len);
1793
1794                         if (!err) {
1795                                 tmp = (struct btrfs_disk_key *)(kaddr + offset -
1796                                                         map_start);
1797                         } else if (err == 1) {
1798                                 read_extent_buffer(eb, &unaligned,
1799                                                    offset, sizeof(unaligned));
1800                                 tmp = &unaligned;
1801                         } else {
1802                                 return err;
1803                         }
1804
1805                 } else {
1806                         tmp = (struct btrfs_disk_key *)(kaddr + offset -
1807                                                         map_start);
1808                 }
1809                 ret = comp_keys(tmp, key);
1810
1811                 if (ret < 0)
1812                         low = mid + 1;
1813                 else if (ret > 0)
1814                         high = mid;
1815                 else {
1816                         *slot = mid;
1817                         return 0;
1818                 }
1819         }
1820         *slot = low;
1821         return 1;
1822 }
1823
1824 /*
1825  * simple bin_search frontend that does the right thing for
1826  * leaves vs nodes
1827  */
1828 static int bin_search(struct extent_buffer *eb, struct btrfs_key *key,
1829                       int level, int *slot)
1830 {
1831         if (level == 0)
1832                 return generic_bin_search(eb,
1833                                           offsetof(struct btrfs_leaf, items),
1834                                           sizeof(struct btrfs_item),
1835                                           key, btrfs_header_nritems(eb),
1836                                           slot);
1837         else
1838                 return generic_bin_search(eb,
1839                                           offsetof(struct btrfs_node, ptrs),
1840                                           sizeof(struct btrfs_key_ptr),
1841                                           key, btrfs_header_nritems(eb),
1842                                           slot);
1843 }
1844
1845 int btrfs_bin_search(struct extent_buffer *eb, struct btrfs_key *key,
1846                      int level, int *slot)
1847 {
1848         return bin_search(eb, key, level, slot);
1849 }
1850
1851 static void root_add_used(struct btrfs_root *root, u32 size)
1852 {
1853         spin_lock(&root->accounting_lock);
1854         btrfs_set_root_used(&root->root_item,
1855                             btrfs_root_used(&root->root_item) + size);
1856         spin_unlock(&root->accounting_lock);
1857 }
1858
1859 static void root_sub_used(struct btrfs_root *root, u32 size)
1860 {
1861         spin_lock(&root->accounting_lock);
1862         btrfs_set_root_used(&root->root_item,
1863                             btrfs_root_used(&root->root_item) - size);
1864         spin_unlock(&root->accounting_lock);
1865 }
1866
1867 /* given a node and slot number, this reads the blocks it points to.  The
1868  * extent buffer is returned with a reference taken (but unlocked).
1869  */
1870 static noinline struct extent_buffer *read_node_slot(struct btrfs_root *root,
1871                                    struct extent_buffer *parent, int slot)
1872 {
1873         int level = btrfs_header_level(parent);
1874         struct extent_buffer *eb;
1875
1876         if (slot < 0 || slot >= btrfs_header_nritems(parent))
1877                 return ERR_PTR(-ENOENT);
1878
1879         BUG_ON(level == 0);
1880
1881         eb = read_tree_block(root, btrfs_node_blockptr(parent, slot),
1882                              btrfs_node_ptr_generation(parent, slot));
1883         if (!IS_ERR(eb) && !extent_buffer_uptodate(eb)) {
1884                 free_extent_buffer(eb);
1885                 eb = ERR_PTR(-EIO);
1886         }
1887
1888         return eb;
1889 }
1890
1891 /*
1892  * node level balancing, used to make sure nodes are in proper order for
1893  * item deletion.  We balance from the top down, so we have to make sure
1894  * that a deletion won't leave an node completely empty later on.
1895  */
1896 static noinline int balance_level(struct btrfs_trans_handle *trans,
1897                          struct btrfs_root *root,
1898                          struct btrfs_path *path, int level)
1899 {
1900         struct extent_buffer *right = NULL;
1901         struct extent_buffer *mid;
1902         struct extent_buffer *left = NULL;
1903         struct extent_buffer *parent = NULL;
1904         int ret = 0;
1905         int wret;
1906         int pslot;
1907         int orig_slot = path->slots[level];
1908         u64 orig_ptr;
1909
1910         if (level == 0)
1911                 return 0;
1912
1913         mid = path->nodes[level];
1914
1915         WARN_ON(path->locks[level] != BTRFS_WRITE_LOCK &&
1916                 path->locks[level] != BTRFS_WRITE_LOCK_BLOCKING);
1917         WARN_ON(btrfs_header_generation(mid) != trans->transid);
1918
1919         orig_ptr = btrfs_node_blockptr(mid, orig_slot);
1920
1921         if (level < BTRFS_MAX_LEVEL - 1) {
1922                 parent = path->nodes[level + 1];
1923                 pslot = path->slots[level + 1];
1924         }
1925
1926         /*
1927          * deal with the case where there is only one pointer in the root
1928          * by promoting the node below to a root
1929          */
1930         if (!parent) {
1931                 struct extent_buffer *child;
1932
1933                 if (btrfs_header_nritems(mid) != 1)
1934                         return 0;
1935
1936                 /* promote the child to a root */
1937                 child = read_node_slot(root, mid, 0);
1938                 if (IS_ERR(child)) {
1939                         ret = PTR_ERR(child);
1940                         btrfs_handle_fs_error(root->fs_info, ret, NULL);
1941                         goto enospc;
1942                 }
1943
1944                 btrfs_tree_lock(child);
1945                 btrfs_set_lock_blocking(child);
1946                 ret = btrfs_cow_block(trans, root, child, mid, 0, &child);
1947                 if (ret) {
1948                         btrfs_tree_unlock(child);
1949                         free_extent_buffer(child);
1950                         goto enospc;
1951                 }
1952
1953                 tree_mod_log_set_root_pointer(root, child, 1);
1954                 rcu_assign_pointer(root->node, child);
1955
1956                 add_root_to_dirty_list(root);
1957                 btrfs_tree_unlock(child);
1958
1959                 path->locks[level] = 0;
1960                 path->nodes[level] = NULL;
1961                 clean_tree_block(trans, root->fs_info, mid);
1962                 btrfs_tree_unlock(mid);
1963                 /* once for the path */
1964                 free_extent_buffer(mid);
1965
1966                 root_sub_used(root, mid->len);
1967                 btrfs_free_tree_block(trans, root, mid, 0, 1);
1968                 /* once for the root ptr */
1969                 free_extent_buffer_stale(mid);
1970                 return 0;
1971         }
1972         if (btrfs_header_nritems(mid) >
1973             BTRFS_NODEPTRS_PER_BLOCK(root) / 4)
1974                 return 0;
1975
1976         left = read_node_slot(root, parent, pslot - 1);
1977         if (IS_ERR(left))
1978                 left = NULL;
1979
1980         if (left) {
1981                 btrfs_tree_lock(left);
1982                 btrfs_set_lock_blocking(left);
1983                 wret = btrfs_cow_block(trans, root, left,
1984                                        parent, pslot - 1, &left);
1985                 if (wret) {
1986                         ret = wret;
1987                         goto enospc;
1988                 }
1989         }
1990
1991         right = read_node_slot(root, parent, pslot + 1);
1992         if (IS_ERR(right))
1993                 right = NULL;
1994
1995         if (right) {
1996                 btrfs_tree_lock(right);
1997                 btrfs_set_lock_blocking(right);
1998                 wret = btrfs_cow_block(trans, root, right,
1999                                        parent, pslot + 1, &right);
2000                 if (wret) {
2001                         ret = wret;
2002                         goto enospc;
2003                 }
2004         }
2005
2006         /* first, try to make some room in the middle buffer */
2007         if (left) {
2008                 orig_slot += btrfs_header_nritems(left);
2009                 wret = push_node_left(trans, root, left, mid, 1);
2010                 if (wret < 0)
2011                         ret = wret;
2012         }
2013
2014         /*
2015          * then try to empty the right most buffer into the middle
2016          */
2017         if (right) {
2018                 wret = push_node_left(trans, root, mid, right, 1);
2019                 if (wret < 0 && wret != -ENOSPC)
2020                         ret = wret;
2021                 if (btrfs_header_nritems(right) == 0) {
2022                         clean_tree_block(trans, root->fs_info, right);
2023                         btrfs_tree_unlock(right);
2024                         del_ptr(root, path, level + 1, pslot + 1);
2025                         root_sub_used(root, right->len);
2026                         btrfs_free_tree_block(trans, root, right, 0, 1);
2027                         free_extent_buffer_stale(right);
2028                         right = NULL;
2029                 } else {
2030                         struct btrfs_disk_key right_key;
2031                         btrfs_node_key(right, &right_key, 0);
2032                         tree_mod_log_set_node_key(root->fs_info, parent,
2033                                                   pslot + 1, 0);
2034                         btrfs_set_node_key(parent, &right_key, pslot + 1);
2035                         btrfs_mark_buffer_dirty(parent);
2036                 }
2037         }
2038         if (btrfs_header_nritems(mid) == 1) {
2039                 /*
2040                  * we're not allowed to leave a node with one item in the
2041                  * tree during a delete.  A deletion from lower in the tree
2042                  * could try to delete the only pointer in this node.
2043                  * So, pull some keys from the left.
2044                  * There has to be a left pointer at this point because
2045                  * otherwise we would have pulled some pointers from the
2046                  * right
2047                  */
2048                 if (!left) {
2049                         ret = -EROFS;
2050                         btrfs_handle_fs_error(root->fs_info, ret, NULL);
2051                         goto enospc;
2052                 }
2053                 wret = balance_node_right(trans, root, mid, left);
2054                 if (wret < 0) {
2055                         ret = wret;
2056                         goto enospc;
2057                 }
2058                 if (wret == 1) {
2059                         wret = push_node_left(trans, root, left, mid, 1);
2060                         if (wret < 0)
2061                                 ret = wret;
2062                 }
2063                 BUG_ON(wret == 1);
2064         }
2065         if (btrfs_header_nritems(mid) == 0) {
2066                 clean_tree_block(trans, root->fs_info, mid);
2067                 btrfs_tree_unlock(mid);
2068                 del_ptr(root, path, level + 1, pslot);
2069                 root_sub_used(root, mid->len);
2070                 btrfs_free_tree_block(trans, root, mid, 0, 1);
2071                 free_extent_buffer_stale(mid);
2072                 mid = NULL;
2073         } else {
2074                 /* update the parent key to reflect our changes */
2075                 struct btrfs_disk_key mid_key;
2076                 btrfs_node_key(mid, &mid_key, 0);
2077                 tree_mod_log_set_node_key(root->fs_info, parent,
2078                                           pslot, 0);
2079                 btrfs_set_node_key(parent, &mid_key, pslot);
2080                 btrfs_mark_buffer_dirty(parent);
2081         }
2082
2083         /* update the path */
2084         if (left) {
2085                 if (btrfs_header_nritems(left) > orig_slot) {
2086                         extent_buffer_get(left);
2087                         /* left was locked after cow */
2088                         path->nodes[level] = left;
2089                         path->slots[level + 1] -= 1;
2090                         path->slots[level] = orig_slot;
2091                         if (mid) {
2092                                 btrfs_tree_unlock(mid);
2093                                 free_extent_buffer(mid);
2094                         }
2095                 } else {
2096                         orig_slot -= btrfs_header_nritems(left);
2097                         path->slots[level] = orig_slot;
2098                 }
2099         }
2100         /* double check we haven't messed things up */
2101         if (orig_ptr !=
2102             btrfs_node_blockptr(path->nodes[level], path->slots[level]))
2103                 BUG();
2104 enospc:
2105         if (right) {
2106                 btrfs_tree_unlock(right);
2107                 free_extent_buffer(right);
2108         }
2109         if (left) {
2110                 if (path->nodes[level] != left)
2111                         btrfs_tree_unlock(left);
2112                 free_extent_buffer(left);
2113         }
2114         return ret;
2115 }
2116
2117 /* Node balancing for insertion.  Here we only split or push nodes around
2118  * when they are completely full.  This is also done top down, so we
2119  * have to be pessimistic.
2120  */
2121 static noinline int push_nodes_for_insert(struct btrfs_trans_handle *trans,
2122                                           struct btrfs_root *root,
2123                                           struct btrfs_path *path, int level)
2124 {
2125         struct extent_buffer *right = NULL;
2126         struct extent_buffer *mid;
2127         struct extent_buffer *left = NULL;
2128         struct extent_buffer *parent = NULL;
2129         int ret = 0;
2130         int wret;
2131         int pslot;
2132         int orig_slot = path->slots[level];
2133
2134         if (level == 0)
2135                 return 1;
2136
2137         mid = path->nodes[level];
2138         WARN_ON(btrfs_header_generation(mid) != trans->transid);
2139
2140         if (level < BTRFS_MAX_LEVEL - 1) {
2141                 parent = path->nodes[level + 1];
2142                 pslot = path->slots[level + 1];
2143         }
2144
2145         if (!parent)
2146                 return 1;
2147
2148         left = read_node_slot(root, parent, pslot - 1);
2149         if (IS_ERR(left))
2150                 left = NULL;
2151
2152         /* first, try to make some room in the middle buffer */
2153         if (left) {
2154                 u32 left_nr;
2155
2156                 btrfs_tree_lock(left);
2157                 btrfs_set_lock_blocking(left);
2158
2159                 left_nr = btrfs_header_nritems(left);
2160                 if (left_nr >= BTRFS_NODEPTRS_PER_BLOCK(root) - 1) {
2161                         wret = 1;
2162                 } else {
2163                         ret = btrfs_cow_block(trans, root, left, parent,
2164                                               pslot - 1, &left);
2165                         if (ret)
2166                                 wret = 1;
2167                         else {
2168                                 wret = push_node_left(trans, root,
2169                                                       left, mid, 0);
2170                         }
2171                 }
2172                 if (wret < 0)
2173                         ret = wret;
2174                 if (wret == 0) {
2175                         struct btrfs_disk_key disk_key;
2176                         orig_slot += left_nr;
2177                         btrfs_node_key(mid, &disk_key, 0);
2178                         tree_mod_log_set_node_key(root->fs_info, parent,
2179                                                   pslot, 0);
2180                         btrfs_set_node_key(parent, &disk_key, pslot);
2181                         btrfs_mark_buffer_dirty(parent);
2182                         if (btrfs_header_nritems(left) > orig_slot) {
2183                                 path->nodes[level] = left;
2184                                 path->slots[level + 1] -= 1;
2185                                 path->slots[level] = orig_slot;
2186                                 btrfs_tree_unlock(mid);
2187                                 free_extent_buffer(mid);
2188                         } else {
2189                                 orig_slot -=
2190                                         btrfs_header_nritems(left);
2191                                 path->slots[level] = orig_slot;
2192                                 btrfs_tree_unlock(left);
2193                                 free_extent_buffer(left);
2194                         }
2195                         return 0;
2196                 }
2197                 btrfs_tree_unlock(left);
2198                 free_extent_buffer(left);
2199         }
2200         right = read_node_slot(root, parent, pslot + 1);
2201         if (IS_ERR(right))
2202                 right = NULL;
2203
2204         /*
2205          * then try to empty the right most buffer into the middle
2206          */
2207         if (right) {
2208                 u32 right_nr;
2209
2210                 btrfs_tree_lock(right);
2211                 btrfs_set_lock_blocking(right);
2212
2213                 right_nr = btrfs_header_nritems(right);
2214                 if (right_nr >= BTRFS_NODEPTRS_PER_BLOCK(root) - 1) {
2215                         wret = 1;
2216                 } else {
2217                         ret = btrfs_cow_block(trans, root, right,
2218                                               parent, pslot + 1,
2219                                               &right);
2220                         if (ret)
2221                                 wret = 1;
2222                         else {
2223                                 wret = balance_node_right(trans, root,
2224                                                           right, mid);
2225                         }
2226                 }
2227                 if (wret < 0)
2228                         ret = wret;
2229                 if (wret == 0) {
2230                         struct btrfs_disk_key disk_key;
2231
2232                         btrfs_node_key(right, &disk_key, 0);
2233                         tree_mod_log_set_node_key(root->fs_info, parent,
2234                                                   pslot + 1, 0);
2235                         btrfs_set_node_key(parent, &disk_key, pslot + 1);
2236                         btrfs_mark_buffer_dirty(parent);
2237
2238                         if (btrfs_header_nritems(mid) <= orig_slot) {
2239                                 path->nodes[level] = right;
2240                                 path->slots[level + 1] += 1;
2241                                 path->slots[level] = orig_slot -
2242                                         btrfs_header_nritems(mid);
2243                                 btrfs_tree_unlock(mid);
2244                                 free_extent_buffer(mid);
2245                         } else {
2246                                 btrfs_tree_unlock(right);
2247                                 free_extent_buffer(right);
2248                         }
2249                         return 0;
2250                 }
2251                 btrfs_tree_unlock(right);
2252                 free_extent_buffer(right);
2253         }
2254         return 1;
2255 }
2256
2257 /*
2258  * readahead one full node of leaves, finding things that are close
2259  * to the block in 'slot', and triggering ra on them.
2260  */
2261 static void reada_for_search(struct btrfs_root *root,
2262                              struct btrfs_path *path,
2263                              int level, int slot, u64 objectid)
2264 {
2265         struct extent_buffer *node;
2266         struct btrfs_disk_key disk_key;
2267         u32 nritems;
2268         u64 search;
2269         u64 target;
2270         u64 nread = 0;
2271         u64 gen;
2272         struct extent_buffer *eb;
2273         u32 nr;
2274         u32 blocksize;
2275         u32 nscan = 0;
2276
2277         if (level != 1)
2278                 return;
2279
2280         if (!path->nodes[level])
2281                 return;
2282
2283         node = path->nodes[level];
2284
2285         search = btrfs_node_blockptr(node, slot);
2286         blocksize = root->nodesize;
2287         eb = btrfs_find_tree_block(root->fs_info, search);
2288         if (eb) {
2289                 free_extent_buffer(eb);
2290                 return;
2291         }
2292
2293         target = search;
2294
2295         nritems = btrfs_header_nritems(node);
2296         nr = slot;
2297
2298         while (1) {
2299                 if (path->reada == READA_BACK) {
2300                         if (nr == 0)
2301                                 break;
2302                         nr--;
2303                 } else if (path->reada == READA_FORWARD) {
2304                         nr++;
2305                         if (nr >= nritems)
2306                                 break;
2307                 }
2308                 if (path->reada == READA_BACK && objectid) {
2309                         btrfs_node_key(node, &disk_key, nr);
2310                         if (btrfs_disk_key_objectid(&disk_key) != objectid)
2311                                 break;
2312                 }
2313                 search = btrfs_node_blockptr(node, nr);
2314                 if ((search <= target && target - search <= 65536) ||
2315                     (search > target && search - target <= 65536)) {
2316                         gen = btrfs_node_ptr_generation(node, nr);
2317                         readahead_tree_block(root, search);
2318                         nread += blocksize;
2319                 }
2320                 nscan++;
2321                 if ((nread > 65536 || nscan > 32))
2322                         break;
2323         }
2324 }
2325
2326 static noinline void reada_for_balance(struct btrfs_root *root,
2327                                        struct btrfs_path *path, int level)
2328 {
2329         int slot;
2330         int nritems;
2331         struct extent_buffer *parent;
2332         struct extent_buffer *eb;
2333         u64 gen;
2334         u64 block1 = 0;
2335         u64 block2 = 0;
2336
2337         parent = path->nodes[level + 1];
2338         if (!parent)
2339                 return;
2340
2341         nritems = btrfs_header_nritems(parent);
2342         slot = path->slots[level + 1];
2343
2344         if (slot > 0) {
2345                 block1 = btrfs_node_blockptr(parent, slot - 1);
2346                 gen = btrfs_node_ptr_generation(parent, slot - 1);
2347                 eb = btrfs_find_tree_block(root->fs_info, block1);
2348                 /*
2349                  * if we get -eagain from btrfs_buffer_uptodate, we
2350                  * don't want to return eagain here.  That will loop
2351                  * forever
2352                  */
2353                 if (eb && btrfs_buffer_uptodate(eb, gen, 1) != 0)
2354                         block1 = 0;
2355                 free_extent_buffer(eb);
2356         }
2357         if (slot + 1 < nritems) {
2358                 block2 = btrfs_node_blockptr(parent, slot + 1);
2359                 gen = btrfs_node_ptr_generation(parent, slot + 1);
2360                 eb = btrfs_find_tree_block(root->fs_info, block2);
2361                 if (eb && btrfs_buffer_uptodate(eb, gen, 1) != 0)
2362                         block2 = 0;
2363                 free_extent_buffer(eb);
2364         }
2365
2366         if (block1)
2367                 readahead_tree_block(root, block1);
2368         if (block2)
2369                 readahead_tree_block(root, block2);
2370 }
2371
2372
2373 /*
2374  * when we walk down the tree, it is usually safe to unlock the higher layers
2375  * in the tree.  The exceptions are when our path goes through slot 0, because
2376  * operations on the tree might require changing key pointers higher up in the
2377  * tree.
2378  *
2379  * callers might also have set path->keep_locks, which tells this code to keep
2380  * the lock if the path points to the last slot in the block.  This is part of
2381  * walking through the tree, and selecting the next slot in the higher block.
2382  *
2383  * lowest_unlock sets the lowest level in the tree we're allowed to unlock.  so
2384  * if lowest_unlock is 1, level 0 won't be unlocked
2385  */
2386 static noinline void unlock_up(struct btrfs_path *path, int level,
2387                                int lowest_unlock, int min_write_lock_level,
2388                                int *write_lock_level)
2389 {
2390         int i;
2391         int skip_level = level;
2392         int no_skips = 0;
2393         struct extent_buffer *t;
2394
2395         for (i = level; i < BTRFS_MAX_LEVEL; i++) {
2396                 if (!path->nodes[i])
2397                         break;
2398                 if (!path->locks[i])
2399                         break;
2400                 if (!no_skips && path->slots[i] == 0) {
2401                         skip_level = i + 1;
2402                         continue;
2403                 }
2404                 if (!no_skips && path->keep_locks) {
2405                         u32 nritems;
2406                         t = path->nodes[i];
2407                         nritems = btrfs_header_nritems(t);
2408                         if (nritems < 1 || path->slots[i] >= nritems - 1) {
2409                                 skip_level = i + 1;
2410                                 continue;
2411                         }
2412                 }
2413                 if (skip_level < i && i >= lowest_unlock)
2414                         no_skips = 1;
2415
2416                 t = path->nodes[i];
2417                 if (i >= lowest_unlock && i > skip_level && path->locks[i]) {
2418                         btrfs_tree_unlock_rw(t, path->locks[i]);
2419                         path->locks[i] = 0;
2420                         if (write_lock_level &&
2421                             i > min_write_lock_level &&
2422                             i <= *write_lock_level) {
2423                                 *write_lock_level = i - 1;
2424                         }
2425                 }
2426         }
2427 }
2428
2429 /*
2430  * This releases any locks held in the path starting at level and
2431  * going all the way up to the root.
2432  *
2433  * btrfs_search_slot will keep the lock held on higher nodes in a few
2434  * corner cases, such as COW of the block at slot zero in the node.  This
2435  * ignores those rules, and it should only be called when there are no
2436  * more updates to be done higher up in the tree.
2437  */
2438 noinline void btrfs_unlock_up_safe(struct btrfs_path *path, int level)
2439 {
2440         int i;
2441
2442         if (path->keep_locks)
2443                 return;
2444
2445         for (i = level; i < BTRFS_MAX_LEVEL; i++) {
2446                 if (!path->nodes[i])
2447                         continue;
2448                 if (!path->locks[i])
2449                         continue;
2450                 btrfs_tree_unlock_rw(path->nodes[i], path->locks[i]);
2451                 path->locks[i] = 0;
2452         }
2453 }
2454
2455 /*
2456  * helper function for btrfs_search_slot.  The goal is to find a block
2457  * in cache without setting the path to blocking.  If we find the block
2458  * we return zero and the path is unchanged.
2459  *
2460  * If we can't find the block, we set the path blocking and do some
2461  * reada.  -EAGAIN is returned and the search must be repeated.
2462  */
2463 static int
2464 read_block_for_search(struct btrfs_trans_handle *trans,
2465                        struct btrfs_root *root, struct btrfs_path *p,
2466                        struct extent_buffer **eb_ret, int level, int slot,
2467                        struct btrfs_key *key, u64 time_seq)
2468 {
2469         u64 blocknr;
2470         u64 gen;
2471         struct extent_buffer *b = *eb_ret;
2472         struct extent_buffer *tmp;
2473         int ret;
2474
2475         blocknr = btrfs_node_blockptr(b, slot);
2476         gen = btrfs_node_ptr_generation(b, slot);
2477
2478         tmp = btrfs_find_tree_block(root->fs_info, blocknr);
2479         if (tmp) {
2480                 /* first we do an atomic uptodate check */
2481                 if (btrfs_buffer_uptodate(tmp, gen, 1) > 0) {
2482                         *eb_ret = tmp;
2483                         return 0;
2484                 }
2485
2486                 /* the pages were up to date, but we failed
2487                  * the generation number check.  Do a full
2488                  * read for the generation number that is correct.
2489                  * We must do this without dropping locks so
2490                  * we can trust our generation number
2491                  */
2492                 btrfs_set_path_blocking(p);
2493
2494                 /* now we're allowed to do a blocking uptodate check */
2495                 ret = btrfs_read_buffer(tmp, gen);
2496                 if (!ret) {
2497                         *eb_ret = tmp;
2498                         return 0;
2499                 }
2500                 free_extent_buffer(tmp);
2501                 btrfs_release_path(p);
2502                 return -EIO;
2503         }
2504
2505         /*
2506          * reduce lock contention at high levels
2507          * of the btree by dropping locks before
2508          * we read.  Don't release the lock on the current
2509          * level because we need to walk this node to figure
2510          * out which blocks to read.
2511          */
2512         btrfs_unlock_up_safe(p, level + 1);
2513         btrfs_set_path_blocking(p);
2514
2515         free_extent_buffer(tmp);
2516         if (p->reada != READA_NONE)
2517                 reada_for_search(root, p, level, slot, key->objectid);
2518
2519         btrfs_release_path(p);
2520
2521         ret = -EAGAIN;
2522         tmp = read_tree_block(root, blocknr, 0);
2523         if (!IS_ERR(tmp)) {
2524                 /*
2525                  * If the read above didn't mark this buffer up to date,
2526                  * it will never end up being up to date.  Set ret to EIO now
2527                  * and give up so that our caller doesn't loop forever
2528                  * on our EAGAINs.
2529                  */
2530                 if (!btrfs_buffer_uptodate(tmp, 0, 0))
2531                         ret = -EIO;
2532                 free_extent_buffer(tmp);
2533         } else {
2534                 ret = PTR_ERR(tmp);
2535         }
2536         return ret;
2537 }
2538
2539 /*
2540  * helper function for btrfs_search_slot.  This does all of the checks
2541  * for node-level blocks and does any balancing required based on
2542  * the ins_len.
2543  *
2544  * If no extra work was required, zero is returned.  If we had to
2545  * drop the path, -EAGAIN is returned and btrfs_search_slot must
2546  * start over
2547  */
2548 static int
2549 setup_nodes_for_search(struct btrfs_trans_handle *trans,
2550                        struct btrfs_root *root, struct btrfs_path *p,
2551                        struct extent_buffer *b, int level, int ins_len,
2552                        int *write_lock_level)
2553 {
2554         int ret;
2555         if ((p->search_for_split || ins_len > 0) && btrfs_header_nritems(b) >=
2556             BTRFS_NODEPTRS_PER_BLOCK(root) - 3) {
2557                 int sret;
2558
2559                 if (*write_lock_level < level + 1) {
2560                         *write_lock_level = level + 1;
2561                         btrfs_release_path(p);
2562                         goto again;
2563                 }
2564
2565                 btrfs_set_path_blocking(p);
2566                 reada_for_balance(root, p, level);
2567                 sret = split_node(trans, root, p, level);
2568                 btrfs_clear_path_blocking(p, NULL, 0);
2569
2570                 BUG_ON(sret > 0);
2571                 if (sret) {
2572                         ret = sret;
2573                         goto done;
2574                 }
2575                 b = p->nodes[level];
2576         } else if (ins_len < 0 && btrfs_header_nritems(b) <
2577                    BTRFS_NODEPTRS_PER_BLOCK(root) / 2) {
2578                 int sret;
2579
2580                 if (*write_lock_level < level + 1) {
2581                         *write_lock_level = level + 1;
2582                         btrfs_release_path(p);
2583                         goto again;
2584                 }
2585
2586                 btrfs_set_path_blocking(p);
2587                 reada_for_balance(root, p, level);
2588                 sret = balance_level(trans, root, p, level);
2589                 btrfs_clear_path_blocking(p, NULL, 0);
2590
2591                 if (sret) {
2592                         ret = sret;
2593                         goto done;
2594                 }
2595                 b = p->nodes[level];
2596                 if (!b) {
2597                         btrfs_release_path(p);
2598                         goto again;
2599                 }
2600                 BUG_ON(btrfs_header_nritems(b) == 1);
2601         }
2602         return 0;
2603
2604 again:
2605         ret = -EAGAIN;
2606 done:
2607         return ret;
2608 }
2609
2610 static void key_search_validate(struct extent_buffer *b,
2611                                 struct btrfs_key *key,
2612                                 int level)
2613 {
2614 #ifdef CONFIG_BTRFS_ASSERT
2615         struct btrfs_disk_key disk_key;
2616
2617         btrfs_cpu_key_to_disk(&disk_key, key);
2618
2619         if (level == 0)
2620                 ASSERT(!memcmp_extent_buffer(b, &disk_key,
2621                     offsetof(struct btrfs_leaf, items[0].key),
2622                     sizeof(disk_key)));
2623         else
2624                 ASSERT(!memcmp_extent_buffer(b, &disk_key,
2625                     offsetof(struct btrfs_node, ptrs[0].key),
2626                     sizeof(disk_key)));
2627 #endif
2628 }
2629
2630 static int key_search(struct extent_buffer *b, struct btrfs_key *key,
2631                       int level, int *prev_cmp, int *slot)
2632 {
2633         if (*prev_cmp != 0) {
2634                 *prev_cmp = bin_search(b, key, level, slot);
2635                 return *prev_cmp;
2636         }
2637
2638         key_search_validate(b, key, level);
2639         *slot = 0;
2640
2641         return 0;
2642 }
2643
2644 int btrfs_find_item(struct btrfs_root *fs_root, struct btrfs_path *path,
2645                 u64 iobjectid, u64 ioff, u8 key_type,
2646                 struct btrfs_key *found_key)
2647 {
2648         int ret;
2649         struct btrfs_key key;
2650         struct extent_buffer *eb;
2651
2652         ASSERT(path);
2653         ASSERT(found_key);
2654
2655         key.type = key_type;
2656         key.objectid = iobjectid;
2657         key.offset = ioff;
2658
2659         ret = btrfs_search_slot(NULL, fs_root, &key, path, 0, 0);
2660         if (ret < 0)
2661                 return ret;
2662
2663         eb = path->nodes[0];
2664         if (ret && path->slots[0] >= btrfs_header_nritems(eb)) {
2665                 ret = btrfs_next_leaf(fs_root, path);
2666                 if (ret)
2667                         return ret;
2668                 eb = path->nodes[0];
2669         }
2670
2671         btrfs_item_key_to_cpu(eb, found_key, path->slots[0]);
2672         if (found_key->type != key.type ||
2673                         found_key->objectid != key.objectid)
2674                 return 1;
2675
2676         return 0;
2677 }
2678
2679 /*
2680  * look for key in the tree.  path is filled in with nodes along the way
2681  * if key is found, we return zero and you can find the item in the leaf
2682  * level of the path (level 0)
2683  *
2684  * If the key isn't found, the path points to the slot where it should
2685  * be inserted, and 1 is returned.  If there are other errors during the
2686  * search a negative error number is returned.
2687  *
2688  * if ins_len > 0, nodes and leaves will be split as we walk down the
2689  * tree.  if ins_len < 0, nodes will be merged as we walk down the tree (if
2690  * possible)
2691  */
2692 int btrfs_search_slot(struct btrfs_trans_handle *trans, struct btrfs_root
2693                       *root, struct btrfs_key *key, struct btrfs_path *p, int
2694                       ins_len, int cow)
2695 {
2696         struct extent_buffer *b;
2697         int slot;
2698         int ret;
2699         int err;
2700         int level;
2701         int lowest_unlock = 1;
2702         int root_lock;
2703         /* everything at write_lock_level or lower must be write locked */
2704         int write_lock_level = 0;
2705         u8 lowest_level = 0;
2706         int min_write_lock_level;
2707         int prev_cmp;
2708
2709         lowest_level = p->lowest_level;
2710         WARN_ON(lowest_level && ins_len > 0);
2711         WARN_ON(p->nodes[0] != NULL);
2712         BUG_ON(!cow && ins_len);
2713
2714         if (ins_len < 0) {
2715                 lowest_unlock = 2;
2716
2717                 /* when we are removing items, we might have to go up to level
2718                  * two as we update tree pointers  Make sure we keep write
2719                  * for those levels as well
2720                  */
2721                 write_lock_level = 2;
2722         } else if (ins_len > 0) {
2723                 /*
2724                  * for inserting items, make sure we have a write lock on
2725                  * level 1 so we can update keys
2726                  */
2727                 write_lock_level = 1;
2728         }
2729
2730         if (!cow)
2731                 write_lock_level = -1;
2732
2733         if (cow && (p->keep_locks || p->lowest_level))
2734                 write_lock_level = BTRFS_MAX_LEVEL;
2735
2736         min_write_lock_level = write_lock_level;
2737
2738 again:
2739         prev_cmp = -1;
2740         /*
2741          * we try very hard to do read locks on the root
2742          */
2743         root_lock = BTRFS_READ_LOCK;
2744         level = 0;
2745         if (p->search_commit_root) {
2746                 /*
2747                  * the commit roots are read only
2748                  * so we always do read locks
2749                  */
2750                 if (p->need_commit_sem)
2751                         down_read(&root->fs_info->commit_root_sem);
2752                 b = root->commit_root;
2753                 extent_buffer_get(b);
2754                 level = btrfs_header_level(b);
2755                 if (p->need_commit_sem)
2756                         up_read(&root->fs_info->commit_root_sem);
2757                 if (!p->skip_locking)
2758                         btrfs_tree_read_lock(b);
2759         } else {
2760                 if (p->skip_locking) {
2761                         b = btrfs_root_node(root);
2762                         level = btrfs_header_level(b);
2763                 } else {
2764                         /* we don't know the level of the root node
2765                          * until we actually have it read locked
2766                          */
2767                         b = btrfs_read_lock_root_node(root);
2768                         level = btrfs_header_level(b);
2769                         if (level <= write_lock_level) {
2770                                 /* whoops, must trade for write lock */
2771                                 btrfs_tree_read_unlock(b);
2772                                 free_extent_buffer(b);
2773                                 b = btrfs_lock_root_node(root);
2774                                 root_lock = BTRFS_WRITE_LOCK;
2775
2776                                 /* the level might have changed, check again */
2777                                 level = btrfs_header_level(b);
2778                         }
2779                 }
2780         }
2781         p->nodes[level] = b;
2782         if (!p->skip_locking)
2783                 p->locks[level] = root_lock;
2784
2785         while (b) {
2786                 level = btrfs_header_level(b);
2787
2788                 /*
2789                  * setup the path here so we can release it under lock
2790                  * contention with the cow code
2791                  */
2792                 if (cow) {
2793                         /*
2794                          * if we don't really need to cow this block
2795                          * then we don't want to set the path blocking,
2796                          * so we test it here
2797                          */
2798                         if (!should_cow_block(trans, root, b)) {
2799                                 trans->dirty = true;
2800                                 goto cow_done;
2801                         }
2802
2803                         /*
2804                          * must have write locks on this node and the
2805                          * parent
2806                          */
2807                         if (level > write_lock_level ||
2808                             (level + 1 > write_lock_level &&
2809                             level + 1 < BTRFS_MAX_LEVEL &&
2810                             p->nodes[level + 1])) {
2811                                 write_lock_level = level + 1;
2812                                 btrfs_release_path(p);
2813                                 goto again;
2814                         }
2815
2816                         btrfs_set_path_blocking(p);
2817                         err = btrfs_cow_block(trans, root, b,
2818                                               p->nodes[level + 1],
2819                                               p->slots[level + 1], &b);
2820                         if (err) {
2821                                 ret = err;
2822                                 goto done;
2823                         }
2824                 }
2825 cow_done:
2826                 p->nodes[level] = b;
2827                 btrfs_clear_path_blocking(p, NULL, 0);
2828
2829                 /*
2830                  * we have a lock on b and as long as we aren't changing
2831                  * the tree, there is no way to for the items in b to change.
2832                  * It is safe to drop the lock on our parent before we
2833                  * go through the expensive btree search on b.
2834                  *
2835                  * If we're inserting or deleting (ins_len != 0), then we might
2836                  * be changing slot zero, which may require changing the parent.
2837                  * So, we can't drop the lock until after we know which slot
2838                  * we're operating on.
2839                  */
2840                 if (!ins_len && !p->keep_locks) {
2841                         int u = level + 1;
2842
2843                         if (u < BTRFS_MAX_LEVEL && p->locks[u]) {
2844                                 btrfs_tree_unlock_rw(p->nodes[u], p->locks[u]);
2845                                 p->locks[u] = 0;
2846                         }
2847                 }
2848
2849                 ret = key_search(b, key, level, &prev_cmp, &slot);
2850                 if (ret < 0)
2851                         goto done;
2852
2853                 if (level != 0) {
2854                         int dec = 0;
2855                         if (ret && slot > 0) {
2856                                 dec = 1;
2857                                 slot -= 1;
2858                         }
2859                         p->slots[level] = slot;
2860                         err = setup_nodes_for_search(trans, root, p, b, level,
2861                                              ins_len, &write_lock_level);
2862                         if (err == -EAGAIN)
2863                                 goto again;
2864                         if (err) {
2865                                 ret = err;
2866                                 goto done;
2867                         }
2868                         b = p->nodes[level];
2869                         slot = p->slots[level];
2870
2871                         /*
2872                          * slot 0 is special, if we change the key
2873                          * we have to update the parent pointer
2874                          * which means we must have a write lock
2875                          * on the parent
2876                          */
2877                         if (slot == 0 && ins_len &&
2878                             write_lock_level < level + 1) {
2879                                 write_lock_level = level + 1;
2880                                 btrfs_release_path(p);
2881                                 goto again;
2882                         }
2883
2884                         unlock_up(p, level, lowest_unlock,
2885                                   min_write_lock_level, &write_lock_level);
2886
2887                         if (level == lowest_level) {
2888                                 if (dec)
2889                                         p->slots[level]++;
2890                                 goto done;
2891                         }
2892
2893                         err = read_block_for_search(trans, root, p,
2894                                                     &b, level, slot, key, 0);
2895                         if (err == -EAGAIN)
2896                                 goto again;
2897                         if (err) {
2898                                 ret = err;
2899                                 goto done;
2900                         }
2901
2902                         if (!p->skip_locking) {
2903                                 level = btrfs_header_level(b);
2904                                 if (level <= write_lock_level) {
2905                                         err = btrfs_try_tree_write_lock(b);
2906                                         if (!err) {
2907                                                 btrfs_set_path_blocking(p);
2908                                                 btrfs_tree_lock(b);
2909                                                 btrfs_clear_path_blocking(p, b,
2910                                                                   BTRFS_WRITE_LOCK);
2911                                         }
2912                                         p->locks[level] = BTRFS_WRITE_LOCK;
2913                                 } else {
2914                                         err = btrfs_tree_read_lock_atomic(b);
2915                                         if (!err) {
2916                                                 btrfs_set_path_blocking(p);
2917                                                 btrfs_tree_read_lock(b);
2918                                                 btrfs_clear_path_blocking(p, b,
2919                                                                   BTRFS_READ_LOCK);
2920                                         }
2921                                         p->locks[level] = BTRFS_READ_LOCK;
2922                                 }
2923                                 p->nodes[level] = b;
2924                         }
2925                 } else {
2926                         p->slots[level] = slot;
2927                         if (ins_len > 0 &&
2928                             btrfs_leaf_free_space(root, b) < ins_len) {
2929                                 if (write_lock_level < 1) {
2930                                         write_lock_level = 1;
2931                                         btrfs_release_path(p);
2932                                         goto again;
2933                                 }
2934
2935                                 btrfs_set_path_blocking(p);
2936                                 err = split_leaf(trans, root, key,
2937                                                  p, ins_len, ret == 0);
2938                                 btrfs_clear_path_blocking(p, NULL, 0);
2939
2940                                 BUG_ON(err > 0);
2941                                 if (err) {
2942                                         ret = err;
2943                                         goto done;
2944                                 }
2945                         }
2946                         if (!p->search_for_split)
2947                                 unlock_up(p, level, lowest_unlock,
2948                                           min_write_lock_level, &write_lock_level);
2949                         goto done;
2950                 }
2951         }
2952         ret = 1;
2953 done:
2954         /*
2955          * we don't really know what they plan on doing with the path
2956          * from here on, so for now just mark it as blocking
2957          */
2958         if (!p->leave_spinning)
2959                 btrfs_set_path_blocking(p);
2960         if (ret < 0 && !p->skip_release_on_error)
2961                 btrfs_release_path(p);
2962         return ret;
2963 }
2964
2965 /*
2966  * Like btrfs_search_slot, this looks for a key in the given tree. It uses the
2967  * current state of the tree together with the operations recorded in the tree
2968  * modification log to search for the key in a previous version of this tree, as
2969  * denoted by the time_seq parameter.
2970  *
2971  * Naturally, there is no support for insert, delete or cow operations.
2972  *
2973  * The resulting path and return value will be set up as if we called
2974  * btrfs_search_slot at that point in time with ins_len and cow both set to 0.
2975  */
2976 int btrfs_search_old_slot(struct btrfs_root *root, struct btrfs_key *key,
2977                           struct btrfs_path *p, u64 time_seq)
2978 {
2979         struct extent_buffer *b;
2980         int slot;
2981         int ret;
2982         int err;
2983         int level;
2984         int lowest_unlock = 1;
2985         u8 lowest_level = 0;
2986         int prev_cmp = -1;
2987
2988         lowest_level = p->lowest_level;
2989         WARN_ON(p->nodes[0] != NULL);
2990
2991         if (p->search_commit_root) {
2992                 BUG_ON(time_seq);
2993                 return btrfs_search_slot(NULL, root, key, p, 0, 0);
2994         }
2995
2996 again:
2997         b = get_old_root(root, time_seq);
2998         level = btrfs_header_level(b);
2999         p->locks[level] = BTRFS_READ_LOCK;
3000
3001         while (b) {
3002                 level = btrfs_header_level(b);
3003                 p->nodes[level] = b;
3004                 btrfs_clear_path_blocking(p, NULL, 0);
3005
3006                 /*
3007                  * we have a lock on b and as long as we aren't changing
3008                  * the tree, there is no way to for the items in b to change.
3009                  * It is safe to drop the lock on our parent before we
3010                  * go through the expensive btree search on b.
3011                  */
3012                 btrfs_unlock_up_safe(p, level + 1);
3013
3014                 /*
3015                  * Since we can unwind ebs we want to do a real search every
3016                  * time.
3017                  */
3018                 prev_cmp = -1;
3019                 ret = key_search(b, key, level, &prev_cmp, &slot);
3020
3021                 if (level != 0) {
3022                         int dec = 0;
3023                         if (ret && slot > 0) {
3024                                 dec = 1;
3025                                 slot -= 1;
3026                         }
3027                         p->slots[level] = slot;
3028                         unlock_up(p, level, lowest_unlock, 0, NULL);
3029
3030                         if (level == lowest_level) {
3031                                 if (dec)
3032                                         p->slots[level]++;
3033                                 goto done;
3034                         }
3035
3036                         err = read_block_for_search(NULL, root, p, &b, level,
3037                                                     slot, key, time_seq);
3038                         if (err == -EAGAIN)
3039                                 goto again;
3040                         if (err) {
3041                                 ret = err;
3042                                 goto done;
3043                         }
3044
3045                         level = btrfs_header_level(b);
3046                         err = btrfs_tree_read_lock_atomic(b);
3047                         if (!err) {
3048                                 btrfs_set_path_blocking(p);
3049                                 btrfs_tree_read_lock(b);
3050                                 btrfs_clear_path_blocking(p, b,
3051                                                           BTRFS_READ_LOCK);
3052                         }
3053                         b = tree_mod_log_rewind(root->fs_info, p, b, time_seq);
3054                         if (!b) {
3055                                 ret = -ENOMEM;
3056                                 goto done;
3057                         }
3058                         p->locks[level] = BTRFS_READ_LOCK;
3059                         p->nodes[level] = b;
3060                 } else {
3061                         p->slots[level] = slot;
3062                         unlock_up(p, level, lowest_unlock, 0, NULL);
3063                         goto done;
3064                 }
3065         }
3066         ret = 1;
3067 done:
3068         if (!p->leave_spinning)
3069                 btrfs_set_path_blocking(p);
3070         if (ret < 0)
3071                 btrfs_release_path(p);
3072
3073         return ret;
3074 }
3075
3076 /*
3077  * helper to use instead of search slot if no exact match is needed but
3078  * instead the next or previous item should be returned.
3079  * When find_higher is true, the next higher item is returned, the next lower
3080  * otherwise.
3081  * When return_any and find_higher are both true, and no higher item is found,
3082  * return the next lower instead.
3083  * When return_any is true and find_higher is false, and no lower item is found,
3084  * return the next higher instead.
3085  * It returns 0 if any item is found, 1 if none is found (tree empty), and
3086  * < 0 on error
3087  */
3088 int btrfs_search_slot_for_read(struct btrfs_root *root,
3089                                struct btrfs_key *key, struct btrfs_path *p,
3090                                int find_higher, int return_any)
3091 {
3092         int ret;
3093         struct extent_buffer *leaf;
3094
3095 again:
3096         ret = btrfs_search_slot(NULL, root, key, p, 0, 0);
3097         if (ret <= 0)
3098                 return ret;
3099         /*
3100          * a return value of 1 means the path is at the position where the
3101          * item should be inserted. Normally this is the next bigger item,
3102          * but in case the previous item is the last in a leaf, path points
3103          * to the first free slot in the previous leaf, i.e. at an invalid
3104          * item.
3105          */
3106         leaf = p->nodes[0];
3107
3108         if (find_higher) {
3109                 if (p->slots[0] >= btrfs_header_nritems(leaf)) {
3110                         ret = btrfs_next_leaf(root, p);
3111                         if (ret <= 0)
3112                                 return ret;
3113                         if (!return_any)
3114                                 return 1;
3115                         /*
3116                          * no higher item found, return the next
3117                          * lower instead
3118                          */
3119                         return_any = 0;
3120                         find_higher = 0;
3121                         btrfs_release_path(p);
3122                         goto again;
3123                 }
3124         } else {
3125                 if (p->slots[0] == 0) {
3126                         ret = btrfs_prev_leaf(root, p);
3127                         if (ret < 0)
3128                                 return ret;
3129                         if (!ret) {
3130                                 leaf = p->nodes[0];
3131                                 if (p->slots[0] == btrfs_header_nritems(leaf))
3132                                         p->slots[0]--;
3133                                 return 0;
3134                         }
3135                         if (!return_any)
3136                                 return 1;
3137                         /*
3138                          * no lower item found, return the next
3139                          * higher instead
3140                          */
3141                         return_any = 0;
3142                         find_higher = 1;
3143                         btrfs_release_path(p);
3144                         goto again;
3145                 } else {
3146                         --p->slots[0];
3147                 }
3148         }
3149         return 0;
3150 }
3151
3152 /*
3153  * adjust the pointers going up the tree, starting at level
3154  * making sure the right key of each node is points to 'key'.
3155  * This is used after shifting pointers to the left, so it stops
3156  * fixing up pointers when a given leaf/node is not in slot 0 of the
3157  * higher levels
3158  *
3159  */
3160 static void fixup_low_keys(struct btrfs_fs_info *fs_info,
3161                            struct btrfs_path *path,
3162                            struct btrfs_disk_key *key, int level)
3163 {
3164         int i;
3165         struct extent_buffer *t;
3166
3167         for (i = level; i < BTRFS_MAX_LEVEL; i++) {
3168                 int tslot = path->slots[i];
3169                 if (!path->nodes[i])
3170                         break;
3171                 t = path->nodes[i];
3172                 tree_mod_log_set_node_key(fs_info, t, tslot, 1);
3173                 btrfs_set_node_key(t, key, tslot);
3174                 btrfs_mark_buffer_dirty(path->nodes[i]);
3175                 if (tslot != 0)
3176                         break;
3177         }
3178 }
3179
3180 /*
3181  * update item key.
3182  *
3183  * This function isn't completely safe. It's the caller's responsibility
3184  * that the new key won't break the order
3185  */
3186 void btrfs_set_item_key_safe(struct btrfs_fs_info *fs_info,
3187                              struct btrfs_path *path,
3188                              struct btrfs_key *new_key)
3189 {
3190         struct btrfs_disk_key disk_key;
3191         struct extent_buffer *eb;
3192         int slot;
3193
3194         eb = path->nodes[0];
3195         slot = path->slots[0];
3196         if (slot > 0) {
3197                 btrfs_item_key(eb, &disk_key, slot - 1);
3198                 BUG_ON(comp_keys(&disk_key, new_key) >= 0);
3199         }
3200         if (slot < btrfs_header_nritems(eb) - 1) {
3201                 btrfs_item_key(eb, &disk_key, slot + 1);
3202                 BUG_ON(comp_keys(&disk_key, new_key) <= 0);
3203         }
3204
3205         btrfs_cpu_key_to_disk(&disk_key, new_key);
3206         btrfs_set_item_key(eb, &disk_key, slot);
3207         btrfs_mark_buffer_dirty(eb);
3208         if (slot == 0)
3209                 fixup_low_keys(fs_info, path, &disk_key, 1);
3210 }
3211
3212 /*
3213  * try to push data from one node into the next node left in the
3214  * tree.
3215  *
3216  * returns 0 if some ptrs were pushed left, < 0 if there was some horrible
3217  * error, and > 0 if there was no room in the left hand block.
3218  */
3219 static int push_node_left(struct btrfs_trans_handle *trans,
3220                           struct btrfs_root *root, struct extent_buffer *dst,
3221                           struct extent_buffer *src, int empty)
3222 {
3223         int push_items = 0;
3224         int src_nritems;
3225         int dst_nritems;
3226         int ret = 0;
3227
3228         src_nritems = btrfs_header_nritems(src);
3229         dst_nritems = btrfs_header_nritems(dst);
3230         push_items = BTRFS_NODEPTRS_PER_BLOCK(root) - dst_nritems;
3231         WARN_ON(btrfs_header_generation(src) != trans->transid);
3232         WARN_ON(btrfs_header_generation(dst) != trans->transid);
3233
3234         if (!empty && src_nritems <= 8)
3235                 return 1;
3236
3237         if (push_items <= 0)
3238                 return 1;
3239
3240         if (empty) {
3241                 push_items = min(src_nritems, push_items);
3242                 if (push_items < src_nritems) {
3243                         /* leave at least 8 pointers in the node if
3244                          * we aren't going to empty it
3245                          */
3246                         if (src_nritems - push_items < 8) {
3247                                 if (push_items <= 8)
3248                                         return 1;
3249                                 push_items -= 8;
3250                         }
3251                 }
3252         } else
3253                 push_items = min(src_nritems - 8, push_items);
3254
3255         ret = tree_mod_log_eb_copy(root->fs_info, dst, src, dst_nritems, 0,
3256                                    push_items);
3257         if (ret) {
3258                 btrfs_abort_transaction(trans, ret);
3259                 return ret;
3260         }
3261         copy_extent_buffer(dst, src,
3262                            btrfs_node_key_ptr_offset(dst_nritems),
3263                            btrfs_node_key_ptr_offset(0),
3264                            push_items * sizeof(struct btrfs_key_ptr));
3265
3266         if (push_items < src_nritems) {
3267                 /*
3268                  * don't call tree_mod_log_eb_move here, key removal was already
3269                  * fully logged by tree_mod_log_eb_copy above.
3270                  */
3271                 memmove_extent_buffer(src, btrfs_node_key_ptr_offset(0),
3272                                       btrfs_node_key_ptr_offset(push_items),
3273                                       (src_nritems - push_items) *
3274                                       sizeof(struct btrfs_key_ptr));
3275         }
3276         btrfs_set_header_nritems(src, src_nritems - push_items);
3277         btrfs_set_header_nritems(dst, dst_nritems + push_items);
3278         btrfs_mark_buffer_dirty(src);
3279         btrfs_mark_buffer_dirty(dst);
3280
3281         return ret;
3282 }
3283
3284 /*
3285  * try to push data from one node into the next node right in the
3286  * tree.
3287  *
3288  * returns 0 if some ptrs were pushed, < 0 if there was some horrible
3289  * error, and > 0 if there was no room in the right hand block.
3290  *
3291  * this will  only push up to 1/2 the contents of the left node over
3292  */
3293 static int balance_node_right(struct btrfs_trans_handle *trans,
3294                               struct btrfs_root *root,
3295                               struct extent_buffer *dst,
3296                               struct extent_buffer *src)
3297 {
3298         int push_items = 0;
3299         int max_push;
3300         int src_nritems;
3301         int dst_nritems;
3302         int ret = 0;
3303
3304         WARN_ON(btrfs_header_generation(src) != trans->transid);
3305         WARN_ON(btrfs_header_generation(dst) != trans->transid);
3306
3307         src_nritems = btrfs_header_nritems(src);
3308         dst_nritems = btrfs_header_nritems(dst);
3309         push_items = BTRFS_NODEPTRS_PER_BLOCK(root) - dst_nritems;
3310         if (push_items <= 0)
3311                 return 1;
3312
3313         if (src_nritems < 4)
3314                 return 1;
3315
3316         max_push = src_nritems / 2 + 1;
3317         /* don't try to empty the node */
3318         if (max_push >= src_nritems)
3319                 return 1;
3320
3321         if (max_push < push_items)
3322                 push_items = max_push;
3323
3324         tree_mod_log_eb_move(root->fs_info, dst, push_items, 0, dst_nritems);
3325         memmove_extent_buffer(dst, btrfs_node_key_ptr_offset(push_items),
3326                                       btrfs_node_key_ptr_offset(0),
3327                                       (dst_nritems) *
3328                                       sizeof(struct btrfs_key_ptr));
3329
3330         ret = tree_mod_log_eb_copy(root->fs_info, dst, src, 0,
3331                                    src_nritems - push_items, push_items);
3332         if (ret) {
3333                 btrfs_abort_transaction(trans, ret);
3334                 return ret;
3335         }
3336         copy_extent_buffer(dst, src,
3337                            btrfs_node_key_ptr_offset(0),
3338                            btrfs_node_key_ptr_offset(src_nritems - push_items),
3339                            push_items * sizeof(struct btrfs_key_ptr));
3340
3341         btrfs_set_header_nritems(src, src_nritems - push_items);
3342         btrfs_set_header_nritems(dst, dst_nritems + push_items);
3343
3344         btrfs_mark_buffer_dirty(src);
3345         btrfs_mark_buffer_dirty(dst);
3346
3347         return ret;
3348 }
3349
3350 /*
3351  * helper function to insert a new root level in the tree.
3352  * A new node is allocated, and a single item is inserted to
3353  * point to the existing root
3354  *
3355  * returns zero on success or < 0 on failure.
3356  */
3357 static noinline int insert_new_root(struct btrfs_trans_handle *trans,
3358                            struct btrfs_root *root,
3359                            struct btrfs_path *path, int level)
3360 {
3361         u64 lower_gen;
3362         struct extent_buffer *lower;
3363         struct extent_buffer *c;
3364         struct extent_buffer *old;
3365         struct btrfs_disk_key lower_key;
3366
3367         BUG_ON(path->nodes[level]);
3368         BUG_ON(path->nodes[level-1] != root->node);
3369
3370         lower = path->nodes[level-1];
3371         if (level == 1)
3372                 btrfs_item_key(lower, &lower_key, 0);
3373         else
3374                 btrfs_node_key(lower, &lower_key, 0);
3375
3376         c = btrfs_alloc_tree_block(trans, root, 0, root->root_key.objectid,
3377                                    &lower_key, level, root->node->start, 0);
3378         if (IS_ERR(c))
3379                 return PTR_ERR(c);
3380
3381         root_add_used(root, root->nodesize);
3382
3383         memset_extent_buffer(c, 0, 0, sizeof(struct btrfs_header));
3384         btrfs_set_header_nritems(c, 1);
3385         btrfs_set_header_level(c, level);
3386         btrfs_set_header_bytenr(c, c->start);
3387         btrfs_set_header_generation(c, trans->transid);
3388         btrfs_set_header_backref_rev(c, BTRFS_MIXED_BACKREF_REV);
3389         btrfs_set_header_owner(c, root->root_key.objectid);
3390
3391         write_extent_buffer(c, root->fs_info->fsid, btrfs_header_fsid(),
3392                             BTRFS_FSID_SIZE);
3393
3394         write_extent_buffer(c, root->fs_info->chunk_tree_uuid,
3395                             btrfs_header_chunk_tree_uuid(c), BTRFS_UUID_SIZE);
3396
3397         btrfs_set_node_key(c, &lower_key, 0);
3398         btrfs_set_node_blockptr(c, 0, lower->start);
3399         lower_gen = btrfs_header_generation(lower);
3400         WARN_ON(lower_gen != trans->transid);
3401
3402         btrfs_set_node_ptr_generation(c, 0, lower_gen);
3403
3404         btrfs_mark_buffer_dirty(c);
3405
3406         old = root->node;
3407         tree_mod_log_set_root_pointer(root, c, 0);
3408         rcu_assign_pointer(root->node, c);
3409
3410         /* the super has an extra ref to root->node */
3411         free_extent_buffer(old);
3412
3413         add_root_to_dirty_list(root);
3414         extent_buffer_get(c);
3415         path->nodes[level] = c;
3416         path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING;
3417         path->slots[level] = 0;
3418         return 0;
3419 }
3420
3421 /*
3422  * worker function to insert a single pointer in a node.
3423  * the node should have enough room for the pointer already
3424  *
3425  * slot and level indicate where you want the key to go, and
3426  * blocknr is the block the key points to.
3427  */
3428 static void insert_ptr(struct btrfs_trans_handle *trans,
3429                        struct btrfs_root *root, struct btrfs_path *path,
3430                        struct btrfs_disk_key *key, u64 bytenr,
3431                        int slot, int level)
3432 {
3433         struct extent_buffer *lower;
3434         int nritems;
3435         int ret;
3436
3437         BUG_ON(!path->nodes[level]);
3438         btrfs_assert_tree_locked(path->nodes[level]);
3439         lower = path->nodes[level];
3440         nritems = btrfs_header_nritems(lower);
3441         BUG_ON(slot > nritems);
3442         BUG_ON(nritems == BTRFS_NODEPTRS_PER_BLOCK(root));
3443         if (slot != nritems) {
3444                 if (level)
3445                         tree_mod_log_eb_move(root->fs_info, lower, slot + 1,
3446                                              slot, nritems - slot);
3447                 memmove_extent_buffer(lower,
3448                               btrfs_node_key_ptr_offset(slot + 1),
3449                               btrfs_node_key_ptr_offset(slot),
3450                               (nritems - slot) * sizeof(struct btrfs_key_ptr));
3451         }
3452         if (level) {
3453                 ret = tree_mod_log_insert_key(root->fs_info, lower, slot,
3454                                               MOD_LOG_KEY_ADD, GFP_NOFS);
3455                 BUG_ON(ret < 0);
3456         }
3457         btrfs_set_node_key(lower, key, slot);
3458         btrfs_set_node_blockptr(lower, slot, bytenr);
3459         WARN_ON(trans->transid == 0);
3460         btrfs_set_node_ptr_generation(lower, slot, trans->transid);
3461         btrfs_set_header_nritems(lower, nritems + 1);
3462         btrfs_mark_buffer_dirty(lower);
3463 }
3464
3465 /*
3466  * split the node at the specified level in path in two.
3467  * The path is corrected to point to the appropriate node after the split
3468  *
3469  * Before splitting this tries to make some room in the node by pushing
3470  * left and right, if either one works, it returns right away.
3471  *
3472  * returns 0 on success and < 0 on failure
3473  */
3474 static noinline int split_node(struct btrfs_trans_handle *trans,
3475                                struct btrfs_root *root,
3476                                struct btrfs_path *path, int level)
3477 {
3478         struct extent_buffer *c;
3479         struct extent_buffer *split;
3480         struct btrfs_disk_key disk_key;
3481         int mid;
3482         int ret;
3483         u32 c_nritems;
3484
3485         c = path->nodes[level];
3486         WARN_ON(btrfs_header_generation(c) != trans->transid);
3487         if (c == root->node) {
3488                 /*
3489                  * trying to split the root, lets make a new one
3490                  *
3491                  * tree mod log: We don't log_removal old root in
3492                  * insert_new_root, because that root buffer will be kept as a
3493                  * normal node. We are going to log removal of half of the
3494                  * elements below with tree_mod_log_eb_copy. We're holding a
3495                  * tree lock on the buffer, which is why we cannot race with
3496                  * other tree_mod_log users.
3497                  */
3498                 ret = insert_new_root(trans, root, path, level + 1);
3499                 if (ret)
3500                         return ret;
3501         } else {
3502                 ret = push_nodes_for_insert(trans, root, path, level);
3503                 c = path->nodes[level];
3504                 if (!ret && btrfs_header_nritems(c) <
3505                     BTRFS_NODEPTRS_PER_BLOCK(root) - 3)
3506                         return 0;
3507                 if (ret < 0)
3508                         return ret;
3509         }
3510
3511         c_nritems = btrfs_header_nritems(c);
3512         mid = (c_nritems + 1) / 2;
3513         btrfs_node_key(c, &disk_key, mid);
3514
3515         split = btrfs_alloc_tree_block(trans, root, 0, root->root_key.objectid,
3516                         &disk_key, level, c->start, 0);
3517         if (IS_ERR(split))
3518                 return PTR_ERR(split);
3519
3520         root_add_used(root, root->nodesize);
3521
3522         memset_extent_buffer(split, 0, 0, sizeof(struct btrfs_header));
3523         btrfs_set_header_level(split, btrfs_header_level(c));
3524         btrfs_set_header_bytenr(split, split->start);
3525         btrfs_set_header_generation(split, trans->transid);
3526         btrfs_set_header_backref_rev(split, BTRFS_MIXED_BACKREF_REV);
3527         btrfs_set_header_owner(split, root->root_key.objectid);
3528         write_extent_buffer(split, root->fs_info->fsid,
3529                             btrfs_header_fsid(), BTRFS_FSID_SIZE);
3530         write_extent_buffer(split, root->fs_info->chunk_tree_uuid,
3531                             btrfs_header_chunk_tree_uuid(split),
3532                             BTRFS_UUID_SIZE);
3533
3534         ret = tree_mod_log_eb_copy(root->fs_info, split, c, 0,
3535                                    mid, c_nritems - mid);
3536         if (ret) {
3537                 btrfs_abort_transaction(trans, ret);
3538                 return ret;
3539         }
3540         copy_extent_buffer(split, c,
3541                            btrfs_node_key_ptr_offset(0),
3542                            btrfs_node_key_ptr_offset(mid),
3543                            (c_nritems - mid) * sizeof(struct btrfs_key_ptr));
3544         btrfs_set_header_nritems(split, c_nritems - mid);
3545         btrfs_set_header_nritems(c, mid);
3546         ret = 0;
3547
3548         btrfs_mark_buffer_dirty(c);
3549         btrfs_mark_buffer_dirty(split);
3550
3551         insert_ptr(trans, root, path, &disk_key, split->start,
3552                    path->slots[level + 1] + 1, level + 1);
3553
3554         if (path->slots[level] >= mid) {
3555                 path->slots[level] -= mid;
3556                 btrfs_tree_unlock(c);
3557                 free_extent_buffer(c);
3558                 path->nodes[level] = split;
3559                 path->slots[level + 1] += 1;
3560         } else {
3561                 btrfs_tree_unlock(split);
3562                 free_extent_buffer(split);
3563         }
3564         return ret;
3565 }
3566
3567 /*
3568  * how many bytes are required to store the items in a leaf.  start
3569  * and nr indicate which items in the leaf to check.  This totals up the
3570  * space used both by the item structs and the item data
3571  */
3572 static int leaf_space_used(struct extent_buffer *l, int start, int nr)
3573 {
3574         struct btrfs_item *start_item;
3575         struct btrfs_item *end_item;
3576         struct btrfs_map_token token;
3577         int data_len;
3578         int nritems = btrfs_header_nritems(l);
3579         int end = min(nritems, start + nr) - 1;
3580
3581         if (!nr)
3582                 return 0;
3583         btrfs_init_map_token(&token);
3584         start_item = btrfs_item_nr(start);
3585         end_item = btrfs_item_nr(end);
3586         data_len = btrfs_token_item_offset(l, start_item, &token) +
3587                 btrfs_token_item_size(l, start_item, &token);
3588         data_len = data_len - btrfs_token_item_offset(l, end_item, &token);
3589         data_len += sizeof(struct btrfs_item) * nr;
3590         WARN_ON(data_len < 0);
3591         return data_len;
3592 }
3593
3594 /*
3595  * The space between the end of the leaf items and
3596  * the start of the leaf data.  IOW, how much room
3597  * the leaf has left for both items and data
3598  */
3599 noinline int btrfs_leaf_free_space(struct btrfs_root *root,
3600                                    struct extent_buffer *leaf)
3601 {
3602         int nritems = btrfs_header_nritems(leaf);
3603         int ret;
3604         ret = BTRFS_LEAF_DATA_SIZE(root) - leaf_space_used(leaf, 0, nritems);
3605         if (ret < 0) {
3606                 btrfs_crit(root->fs_info,
3607                         "leaf free space ret %d, leaf data size %lu, used %d nritems %d",
3608                        ret, (unsigned long) BTRFS_LEAF_DATA_SIZE(root),
3609                        leaf_space_used(leaf, 0, nritems), nritems);
3610         }
3611         return ret;
3612 }
3613
3614 /*
3615  * min slot controls the lowest index we're willing to push to the
3616  * right.  We'll push up to and including min_slot, but no lower
3617  */
3618 static noinline int __push_leaf_right(struct btrfs_trans_handle *trans,
3619                                       struct btrfs_root *root,
3620                                       struct btrfs_path *path,
3621                                       int data_size, int empty,
3622                                       struct extent_buffer *right,
3623                                       int free_space, u32 left_nritems,
3624                                       u32 min_slot)
3625 {
3626         struct extent_buffer *left = path->nodes[0];
3627         struct extent_buffer *upper = path->nodes[1];
3628         struct btrfs_map_token token;
3629         struct btrfs_disk_key disk_key;
3630         int slot;
3631         u32 i;
3632         int push_space = 0;
3633         int push_items = 0;
3634         struct btrfs_item *item;
3635         u32 nr;
3636         u32 right_nritems;
3637         u32 data_end;
3638         u32 this_item_size;
3639
3640         btrfs_init_map_token(&token);
3641
3642         if (empty)
3643                 nr = 0;
3644         else
3645                 nr = max_t(u32, 1, min_slot);
3646
3647         if (path->slots[0] >= left_nritems)
3648                 push_space += data_size;
3649
3650         slot = path->slots[1];
3651         i = left_nritems - 1;
3652         while (i >= nr) {
3653                 item = btrfs_item_nr(i);
3654
3655                 if (!empty && push_items > 0) {
3656                         if (path->slots[0] > i)
3657                                 break;
3658                         if (path->slots[0] == i) {
3659                                 int space = btrfs_leaf_free_space(root, left);
3660                                 if (space + push_space * 2 > free_space)
3661                                         break;
3662                         }
3663                 }
3664
3665                 if (path->slots[0] == i)
3666                         push_space += data_size;
3667
3668                 this_item_size = btrfs_item_size(left, item);
3669                 if (this_item_size + sizeof(*item) + push_space > free_space)
3670                         break;
3671
3672                 push_items++;
3673                 push_space += this_item_size + sizeof(*item);
3674                 if (i == 0)
3675                         break;
3676                 i--;
3677         }
3678
3679         if (push_items == 0)
3680                 goto out_unlock;
3681
3682         WARN_ON(!empty && push_items == left_nritems);
3683
3684         /* push left to right */
3685         right_nritems = btrfs_header_nritems(right);
3686
3687         push_space = btrfs_item_end_nr(left, left_nritems - push_items);
3688         push_space -= leaf_data_end(root, left);
3689
3690         /* make room in the right data area */
3691         data_end = leaf_data_end(root, right);
3692         memmove_extent_buffer(right,
3693                               btrfs_leaf_data(right) + data_end - push_space,
3694                               btrfs_leaf_data(right) + data_end,
3695                               BTRFS_LEAF_DATA_SIZE(root) - data_end);
3696
3697         /* copy from the left data area */
3698         copy_extent_buffer(right, left, btrfs_leaf_data(right) +
3699                      BTRFS_LEAF_DATA_SIZE(root) - push_space,
3700                      btrfs_leaf_data(left) + leaf_data_end(root, left),
3701                      push_space);
3702
3703         memmove_extent_buffer(right, btrfs_item_nr_offset(push_items),
3704                               btrfs_item_nr_offset(0),
3705                               right_nritems * sizeof(struct btrfs_item));
3706
3707         /* copy the items from left to right */
3708         copy_extent_buffer(right, left, btrfs_item_nr_offset(0),
3709                    btrfs_item_nr_offset(left_nritems - push_items),
3710                    push_items * sizeof(struct btrfs_item));
3711
3712         /* update the item pointers */
3713         right_nritems += push_items;
3714         btrfs_set_header_nritems(right, right_nritems);
3715         push_space = BTRFS_LEAF_DATA_SIZE(root);
3716         for (i = 0; i < right_nritems; i++) {
3717                 item = btrfs_item_nr(i);
3718                 push_space -= btrfs_token_item_size(right, item, &token);
3719                 btrfs_set_token_item_offset(right, item, push_space, &token);
3720         }
3721
3722         left_nritems -= push_items;
3723         btrfs_set_header_nritems(left, left_nritems);
3724
3725         if (left_nritems)
3726                 btrfs_mark_buffer_dirty(left);
3727         else
3728                 clean_tree_block(trans, root->fs_info, left);
3729
3730         btrfs_mark_buffer_dirty(right);
3731
3732         btrfs_item_key(right, &disk_key, 0);
3733         btrfs_set_node_key(upper, &disk_key, slot + 1);
3734         btrfs_mark_buffer_dirty(upper);
3735
3736         /* then fixup the leaf pointer in the path */
3737         if (path->slots[0] >= left_nritems) {
3738                 path->slots[0] -= left_nritems;
3739                 if (btrfs_header_nritems(path->nodes[0]) == 0)
3740                         clean_tree_block(trans, root->fs_info, path->nodes[0]);
3741                 btrfs_tree_unlock(path->nodes[0]);
3742                 free_extent_buffer(path->nodes[0]);
3743                 path->nodes[0] = right;
3744                 path->slots[1] += 1;
3745         } else {
3746                 btrfs_tree_unlock(right);
3747                 free_extent_buffer(right);
3748         }
3749         return 0;
3750
3751 out_unlock:
3752         btrfs_tree_unlock(right);
3753         free_extent_buffer(right);
3754         return 1;
3755 }
3756
3757 /*
3758  * push some data in the path leaf to the right, trying to free up at
3759  * least data_size bytes.  returns zero if the push worked, nonzero otherwise
3760  *
3761  * returns 1 if the push failed because the other node didn't have enough
3762  * room, 0 if everything worked out and < 0 if there were major errors.
3763  *
3764  * this will push starting from min_slot to the end of the leaf.  It won't
3765  * push any slot lower than min_slot
3766  */
3767 static int push_leaf_right(struct btrfs_trans_handle *trans, struct btrfs_root
3768                            *root, struct btrfs_path *path,
3769                            int min_data_size, int data_size,
3770                            int empty, u32 min_slot)
3771 {
3772         struct extent_buffer *left = path->nodes[0];
3773         struct extent_buffer *right;
3774         struct extent_buffer *upper;
3775         int slot;
3776         int free_space;
3777         u32 left_nritems;
3778         int ret;
3779
3780         if (!path->nodes[1])
3781                 return 1;
3782
3783         slot = path->slots[1];
3784         upper = path->nodes[1];
3785         if (slot >= btrfs_header_nritems(upper) - 1)
3786                 return 1;
3787
3788         btrfs_assert_tree_locked(path->nodes[1]);
3789
3790         right = read_node_slot(root, upper, slot + 1);
3791         /*
3792          * slot + 1 is not valid or we fail to read the right node,
3793          * no big deal, just return.
3794          */
3795         if (IS_ERR(right))
3796                 return 1;
3797
3798         btrfs_tree_lock(right);
3799         btrfs_set_lock_blocking(right);
3800
3801         free_space = btrfs_leaf_free_space(root, right);
3802         if (free_space < data_size)
3803                 goto out_unlock;
3804
3805         /* cow and double check */
3806         ret = btrfs_cow_block(trans, root, right, upper,
3807                               slot + 1, &right);
3808         if (ret)
3809                 goto out_unlock;
3810
3811         free_space = btrfs_leaf_free_space(root, right);
3812         if (free_space < data_size)
3813                 goto out_unlock;
3814
3815         left_nritems = btrfs_header_nritems(left);
3816         if (left_nritems == 0)
3817                 goto out_unlock;
3818
3819         if (path->slots[0] == left_nritems && !empty) {
3820                 /* Key greater than all keys in the leaf, right neighbor has
3821                  * enough room for it and we're not emptying our leaf to delete
3822                  * it, therefore use right neighbor to insert the new item and
3823                  * no need to touch/dirty our left leaft. */
3824                 btrfs_tree_unlock(left);
3825                 free_extent_buffer(left);
3826                 path->nodes[0] = right;
3827                 path->slots[0] = 0;
3828                 path->slots[1]++;
3829                 return 0;
3830         }
3831
3832         return __push_leaf_right(trans, root, path, min_data_size, empty,
3833                                 right, free_space, left_nritems, min_slot);
3834 out_unlock:
3835         btrfs_tree_unlock(right);
3836         free_extent_buffer(right);
3837         return 1;
3838 }
3839
3840 /*
3841  * push some data in the path leaf to the left, trying to free up at
3842  * least data_size bytes.  returns zero if the push worked, nonzero otherwise
3843  *
3844  * max_slot can put a limit on how far into the leaf we'll push items.  The
3845  * item at 'max_slot' won't be touched.  Use (u32)-1 to make us do all the
3846  * items
3847  */
3848 static noinline int __push_leaf_left(struct btrfs_trans_handle *trans,
3849                                      struct btrfs_root *root,
3850                                      struct btrfs_path *path, int data_size,
3851                                      int empty, struct extent_buffer *left,
3852                                      int free_space, u32 right_nritems,
3853                                      u32 max_slot)
3854 {
3855         struct btrfs_disk_key disk_key;
3856         struct extent_buffer *right = path->nodes[0];
3857         int i;
3858         int push_space = 0;
3859         int push_items = 0;
3860         struct btrfs_item *item;
3861         u32 old_left_nritems;
3862         u32 nr;
3863         int ret = 0;
3864         u32 this_item_size;
3865         u32 old_left_item_size;
3866         struct btrfs_map_token token;
3867
3868         btrfs_init_map_token(&token);
3869
3870         if (empty)
3871                 nr = min(right_nritems, max_slot);
3872         else
3873                 nr = min(right_nritems - 1, max_slot);
3874
3875         for (i = 0; i < nr; i++) {
3876                 item = btrfs_item_nr(i);
3877
3878                 if (!empty && push_items > 0) {
3879                         if (path->slots[0] < i)
3880                                 break;
3881                         if (path->slots[0] == i) {
3882                                 int space = btrfs_leaf_free_space(root, right);
3883                                 if (space + push_space * 2 > free_space)
3884                                         break;
3885                         }
3886                 }
3887
3888                 if (path->slots[0] == i)
3889                         push_space += data_size;
3890
3891                 this_item_size = btrfs_item_size(right, item);
3892                 if (this_item_size + sizeof(*item) + push_space > free_space)
3893                         break;
3894
3895                 push_items++;
3896                 push_space += this_item_size + sizeof(*item);
3897         }
3898
3899         if (push_items == 0) {
3900                 ret = 1;
3901                 goto out;
3902         }
3903         WARN_ON(!empty && push_items == btrfs_header_nritems(right));
3904
3905         /* push data from right to left */
3906         copy_extent_buffer(left, right,
3907                            btrfs_item_nr_offset(btrfs_header_nritems(left)),
3908                            btrfs_item_nr_offset(0),
3909                            push_items * sizeof(struct btrfs_item));
3910
3911         push_space = BTRFS_LEAF_DATA_SIZE(root) -
3912                      btrfs_item_offset_nr(right, push_items - 1);
3913
3914         copy_extent_buffer(left, right, btrfs_leaf_data(left) +
3915                      leaf_data_end(root, left) - push_space,
3916                      btrfs_leaf_data(right) +
3917                      btrfs_item_offset_nr(right, push_items - 1),
3918                      push_space);
3919         old_left_nritems = btrfs_header_nritems(left);
3920         BUG_ON(old_left_nritems <= 0);
3921
3922         old_left_item_size = btrfs_item_offset_nr(left, old_left_nritems - 1);
3923         for (i = old_left_nritems; i < old_left_nritems + push_items; i++) {
3924                 u32 ioff;
3925
3926                 item = btrfs_item_nr(i);
3927
3928                 ioff = btrfs_token_item_offset(left, item, &token);
3929                 btrfs_set_token_item_offset(left, item,
3930                       ioff - (BTRFS_LEAF_DATA_SIZE(root) - old_left_item_size),
3931                       &token);
3932         }
3933         btrfs_set_header_nritems(left, old_left_nritems + push_items);
3934
3935         /* fixup right node */
3936         if (push_items > right_nritems)
3937                 WARN(1, KERN_CRIT "push items %d nr %u\n", push_items,
3938                        right_nritems);
3939
3940         if (push_items < right_nritems) {
3941                 push_space = btrfs_item_offset_nr(right, push_items - 1) -
3942                                                   leaf_data_end(root, right);
3943                 memmove_extent_buffer(right, btrfs_leaf_data(right) +
3944                                       BTRFS_LEAF_DATA_SIZE(root) - push_space,
3945                                       btrfs_leaf_data(right) +
3946                                       leaf_data_end(root, right), push_space);
3947
3948                 memmove_extent_buffer(right, btrfs_item_nr_offset(0),
3949                               btrfs_item_nr_offset(push_items),
3950                              (btrfs_header_nritems(right) - push_items) *
3951                              sizeof(struct btrfs_item));
3952         }
3953         right_nritems -= push_items;
3954         btrfs_set_header_nritems(right, right_nritems);
3955         push_space = BTRFS_LEAF_DATA_SIZE(root);
3956         for (i = 0; i < right_nritems; i++) {
3957                 item = btrfs_item_nr(i);
3958
3959                 push_space = push_space - btrfs_token_item_size(right,
3960                                                                 item, &token);
3961                 btrfs_set_token_item_offset(right, item, push_space, &token);
3962         }
3963
3964         btrfs_mark_buffer_dirty(left);
3965         if (right_nritems)
3966                 btrfs_mark_buffer_dirty(right);
3967         else
3968                 clean_tree_block(trans, root->fs_info, right);
3969
3970         btrfs_item_key(right, &disk_key, 0);
3971         fixup_low_keys(root->fs_info, path, &disk_key, 1);
3972
3973         /* then fixup the leaf pointer in the path */
3974         if (path->slots[0] < push_items) {
3975                 path->slots[0] += old_left_nritems;
3976                 btrfs_tree_unlock(path->nodes[0]);
3977                 free_extent_buffer(path->nodes[0]);
3978                 path->nodes[0] = left;
3979                 path->slots[1] -= 1;
3980         } else {
3981                 btrfs_tree_unlock(left);
3982                 free_extent_buffer(left);
3983                 path->slots[0] -= push_items;
3984         }
3985         BUG_ON(path->slots[0] < 0);
3986         return ret;
3987 out:
3988         btrfs_tree_unlock(left);
3989         free_extent_buffer(left);
3990         return ret;
3991 }
3992
3993 /*
3994  * push some data in the path leaf to the left, trying to free up at
3995  * least data_size bytes.  returns zero if the push worked, nonzero otherwise
3996  *
3997  * max_slot can put a limit on how far into the leaf we'll push items.  The
3998  * item at 'max_slot' won't be touched.  Use (u32)-1 to make us push all the
3999  * items
4000  */
4001 static int push_leaf_left(struct btrfs_trans_handle *trans, struct btrfs_root
4002                           *root, struct btrfs_path *path, int min_data_size,
4003                           int data_size, int empty, u32 max_slot)
4004 {
4005         struct extent_buffer *right = path->nodes[0];
4006         struct extent_buffer *left;
4007         int slot;
4008         int free_space;
4009         u32 right_nritems;
4010         int ret = 0;
4011
4012         slot = path->slots[1];
4013         if (slot == 0)
4014                 return 1;
4015         if (!path->nodes[1])
4016                 return 1;
4017
4018         right_nritems = btrfs_header_nritems(right);
4019         if (right_nritems == 0)
4020                 return 1;
4021
4022         btrfs_assert_tree_locked(path->nodes[1]);
4023
4024         left = read_node_slot(root, path->nodes[1], slot - 1);
4025         /*
4026          * slot - 1 is not valid or we fail to read the left node,
4027          * no big deal, just return.
4028          */
4029         if (IS_ERR(left))
4030                 return 1;
4031
4032         btrfs_tree_lock(left);
4033         btrfs_set_lock_blocking(left);
4034
4035         free_space = btrfs_leaf_free_space(root, left);
4036         if (free_space < data_size) {
4037                 ret = 1;
4038                 goto out;
4039         }
4040
4041         /* cow and double check */
4042         ret = btrfs_cow_block(trans, root, left,
4043                               path->nodes[1], slot - 1, &left);
4044         if (ret) {
4045                 /* we hit -ENOSPC, but it isn't fatal here */
4046                 if (ret == -ENOSPC)
4047                         ret = 1;
4048                 goto out;
4049         }
4050
4051         free_space = btrfs_leaf_free_space(root, left);
4052         if (free_space < data_size) {
4053                 ret = 1;
4054                 goto out;
4055         }
4056
4057         return __push_leaf_left(trans, root, path, min_data_size,
4058                                empty, left, free_space, right_nritems,
4059                                max_slot);
4060 out:
4061         btrfs_tree_unlock(left);
4062         free_extent_buffer(left);
4063         return ret;
4064 }
4065
4066 /*
4067  * split the path's leaf in two, making sure there is at least data_size
4068  * available for the resulting leaf level of the path.
4069  */
4070 static noinline void copy_for_split(struct btrfs_trans_handle *trans,
4071                                     struct btrfs_root *root,
4072                                     struct btrfs_path *path,
4073                                     struct extent_buffer *l,
4074                                     struct extent_buffer *right,
4075                                     int slot, int mid, int nritems)
4076 {
4077         int data_copy_size;
4078         int rt_data_off;
4079         int i;
4080         struct btrfs_disk_key disk_key;
4081         struct btrfs_map_token token;
4082
4083         btrfs_init_map_token(&token);
4084
4085         nritems = nritems - mid;
4086         btrfs_set_header_nritems(right, nritems);
4087         data_copy_size = btrfs_item_end_nr(l, mid) - leaf_data_end(root, l);
4088
4089         copy_extent_buffer(right, l, btrfs_item_nr_offset(0),
4090                            btrfs_item_nr_offset(mid),
4091                            nritems * sizeof(struct btrfs_item));
4092
4093         copy_extent_buffer(right, l,
4094                      btrfs_leaf_data(right) + BTRFS_LEAF_DATA_SIZE(root) -
4095                      data_copy_size, btrfs_leaf_data(l) +
4096                      leaf_data_end(root, l), data_copy_size);
4097
4098         rt_data_off = BTRFS_LEAF_DATA_SIZE(root) -
4099                       btrfs_item_end_nr(l, mid);
4100
4101         for (i = 0; i < nritems; i++) {
4102                 struct btrfs_item *item = btrfs_item_nr(i);
4103                 u32 ioff;
4104
4105                 ioff = btrfs_token_item_offset(right, item, &token);
4106                 btrfs_set_token_item_offset(right, item,
4107                                             ioff + rt_data_off, &token);
4108         }
4109
4110         btrfs_set_header_nritems(l, mid);
4111         btrfs_item_key(right, &disk_key, 0);
4112         insert_ptr(trans, root, path, &disk_key, right->start,
4113                    path->slots[1] + 1, 1);
4114
4115         btrfs_mark_buffer_dirty(right);
4116         btrfs_mark_buffer_dirty(l);
4117         BUG_ON(path->slots[0] != slot);
4118
4119         if (mid <= slot) {
4120                 btrfs_tree_unlock(path->nodes[0]);
4121                 free_extent_buffer(path->nodes[0]);
4122                 path->nodes[0] = right;
4123                 path->slots[0] -= mid;
4124                 path->slots[1] += 1;
4125         } else {
4126                 btrfs_tree_unlock(right);
4127                 free_extent_buffer(right);
4128         }
4129
4130         BUG_ON(path->slots[0] < 0);
4131 }
4132
4133 /*
4134  * double splits happen when we need to insert a big item in the middle
4135  * of a leaf.  A double split can leave us with 3 mostly empty leaves:
4136  * leaf: [ slots 0 - N] [ our target ] [ N + 1 - total in leaf ]
4137  *          A                 B                 C
4138  *
4139  * We avoid this by trying to push the items on either side of our target
4140  * into the adjacent leaves.  If all goes well we can avoid the double split
4141  * completely.
4142  */
4143 static noinline int push_for_double_split(struct btrfs_trans_handle *trans,
4144                                           struct btrfs_root *root,
4145                                           struct btrfs_path *path,
4146                                           int data_size)
4147 {
4148         int ret;
4149         int progress = 0;
4150         int slot;
4151         u32 nritems;
4152         int space_needed = data_size;
4153
4154         slot = path->slots[0];
4155         if (slot < btrfs_header_nritems(path->nodes[0]))
4156                 space_needed -= btrfs_leaf_free_space(root, path->nodes[0]);
4157
4158         /*
4159          * try to push all the items after our slot into the
4160          * right leaf
4161          */
4162         ret = push_leaf_right(trans, root, path, 1, space_needed, 0, slot);
4163         if (ret < 0)
4164                 return ret;
4165
4166         if (ret == 0)
4167                 progress++;
4168
4169         nritems = btrfs_header_nritems(path->nodes[0]);
4170         /*
4171          * our goal is to get our slot at the start or end of a leaf.  If
4172          * we've done so we're done
4173          */
4174         if (path->slots[0] == 0 || path->slots[0] == nritems)
4175                 return 0;
4176
4177         if (btrfs_leaf_free_space(root, path->nodes[0]) >= data_size)
4178                 return 0;
4179
4180         /* try to push all the items before our slot into the next leaf */
4181         slot = path->slots[0];
4182         ret = push_leaf_left(trans, root, path, 1, space_needed, 0, slot);
4183         if (ret < 0)
4184                 return ret;
4185
4186         if (ret == 0)
4187                 progress++;
4188
4189         if (progress)
4190                 return 0;
4191         return 1;
4192 }
4193
4194 /*
4195  * split the path's leaf in two, making sure there is at least data_size
4196  * available for the resulting leaf level of the path.
4197  *
4198  * returns 0 if all went well and < 0 on failure.
4199  */
4200 static noinline int split_leaf(struct btrfs_trans_handle *trans,
4201                                struct btrfs_root *root,
4202                                struct btrfs_key *ins_key,
4203                                struct btrfs_path *path, int data_size,
4204                                int extend)
4205 {
4206         struct btrfs_disk_key disk_key;
4207         struct extent_buffer *l;
4208         u32 nritems;
4209         int mid;
4210         int slot;
4211         struct extent_buffer *right;
4212         struct btrfs_fs_info *fs_info = root->fs_info;
4213         int ret = 0;
4214         int wret;
4215         int split;
4216         int num_doubles = 0;
4217         int tried_avoid_double = 0;
4218
4219         l = path->nodes[0];
4220         slot = path->slots[0];
4221         if (extend && data_size + btrfs_item_size_nr(l, slot) +
4222             sizeof(struct btrfs_item) > BTRFS_LEAF_DATA_SIZE(root))
4223                 return -EOVERFLOW;
4224
4225         /* first try to make some room by pushing left and right */
4226         if (data_size && path->nodes[1]) {
4227                 int space_needed = data_size;
4228
4229                 if (slot < btrfs_header_nritems(l))
4230                         space_needed -= btrfs_leaf_free_space(root, l);
4231
4232                 wret = push_leaf_right(trans, root, path, space_needed,
4233                                        space_needed, 0, 0);
4234                 if (wret < 0)
4235                         return wret;
4236                 if (wret) {
4237                         wret = push_leaf_left(trans, root, path, space_needed,
4238                                               space_needed, 0, (u32)-1);
4239                         if (wret < 0)
4240                                 return wret;
4241                 }
4242                 l = path->nodes[0];
4243
4244                 /* did the pushes work? */
4245                 if (btrfs_leaf_free_space(root, l) >= data_size)
4246                         return 0;
4247         }
4248
4249         if (!path->nodes[1]) {
4250                 ret = insert_new_root(trans, root, path, 1);
4251                 if (ret)
4252                         return ret;
4253         }
4254 again:
4255         split = 1;
4256         l = path->nodes[0];
4257         slot = path->slots[0];
4258         nritems = btrfs_header_nritems(l);
4259         mid = (nritems + 1) / 2;
4260
4261         if (mid <= slot) {
4262                 if (nritems == 1 ||
4263                     leaf_space_used(l, mid, nritems - mid) + data_size >
4264                         BTRFS_LEAF_DATA_SIZE(root)) {
4265                         if (slot >= nritems) {
4266                                 split = 0;
4267                         } else {
4268                                 mid = slot;
4269                                 if (mid != nritems &&
4270                                     leaf_space_used(l, mid, nritems - mid) +
4271                                     data_size > BTRFS_LEAF_DATA_SIZE(root)) {
4272                                         if (data_size && !tried_avoid_double)
4273                                                 goto push_for_double;
4274                                         split = 2;
4275                                 }
4276                         }
4277                 }
4278         } else {
4279                 if (leaf_space_used(l, 0, mid) + data_size >
4280                         BTRFS_LEAF_DATA_SIZE(root)) {
4281                         if (!extend && data_size && slot == 0) {
4282                                 split = 0;
4283                         } else if ((extend || !data_size) && slot == 0) {
4284                                 mid = 1;
4285                         } else {
4286                                 mid = slot;
4287                                 if (mid != nritems &&
4288                                     leaf_space_used(l, mid, nritems - mid) +
4289                                     data_size > BTRFS_LEAF_DATA_SIZE(root)) {
4290                                         if (data_size && !tried_avoid_double)
4291                                                 goto push_for_double;
4292                                         split = 2;
4293                                 }
4294                         }
4295                 }
4296         }
4297
4298         if (split == 0)
4299                 btrfs_cpu_key_to_disk(&disk_key, ins_key);
4300         else
4301                 btrfs_item_key(l, &disk_key, mid);
4302
4303         right = btrfs_alloc_tree_block(trans, root, 0, root->root_key.objectid,
4304                         &disk_key, 0, l->start, 0);
4305         if (IS_ERR(right))
4306                 return PTR_ERR(right);
4307
4308         root_add_used(root, root->nodesize);
4309
4310         memset_extent_buffer(right, 0, 0, sizeof(struct btrfs_header));
4311         btrfs_set_header_bytenr(right, right->start);
4312         btrfs_set_header_generation(right, trans->transid);
4313         btrfs_set_header_backref_rev(right, BTRFS_MIXED_BACKREF_REV);
4314         btrfs_set_header_owner(right, root->root_key.objectid);
4315         btrfs_set_header_level(right, 0);
4316         write_extent_buffer(right, fs_info->fsid,
4317                             btrfs_header_fsid(), BTRFS_FSID_SIZE);
4318
4319         write_extent_buffer(right, fs_info->chunk_tree_uuid,
4320                             btrfs_header_chunk_tree_uuid(right),
4321                             BTRFS_UUID_SIZE);
4322
4323         if (split == 0) {
4324                 if (mid <= slot) {
4325                         btrfs_set_header_nritems(right, 0);
4326                         insert_ptr(trans, root, path, &disk_key, right->start,
4327                                    path->slots[1] + 1, 1);
4328                         btrfs_tree_unlock(path->nodes[0]);
4329                         free_extent_buffer(path->nodes[0]);
4330                         path->nodes[0] = right;
4331                         path->slots[0] = 0;
4332                         path->slots[1] += 1;
4333                 } else {
4334                         btrfs_set_header_nritems(right, 0);
4335                         insert_ptr(trans, root, path, &disk_key, right->start,
4336                                           path->slots[1], 1);
4337                         btrfs_tree_unlock(path->nodes[0]);
4338                         free_extent_buffer(path->nodes[0]);
4339                         path->nodes[0] = right;
4340                         path->slots[0] = 0;
4341                         if (path->slots[1] == 0)
4342                                 fixup_low_keys(fs_info, path, &disk_key, 1);
4343                 }
4344                 btrfs_mark_buffer_dirty(right);
4345                 return ret;
4346         }
4347
4348         copy_for_split(trans, root, path, l, right, slot, mid, nritems);
4349
4350         if (split == 2) {
4351                 BUG_ON(num_doubles != 0);
4352                 num_doubles++;
4353                 goto again;
4354         }
4355
4356         return 0;
4357
4358 push_for_double:
4359         push_for_double_split(trans, root, path, data_size);
4360         tried_avoid_double = 1;
4361         if (btrfs_leaf_free_space(root, path->nodes[0]) >= data_size)
4362                 return 0;
4363         goto again;
4364 }
4365
4366 static noinline int setup_leaf_for_split(struct btrfs_trans_handle *trans,
4367                                          struct btrfs_root *root,
4368                                          struct btrfs_path *path, int ins_len)
4369 {
4370         struct btrfs_key key;
4371         struct extent_buffer *leaf;
4372         struct btrfs_file_extent_item *fi;
4373         u64 extent_len = 0;
4374         u32 item_size;
4375         int ret;
4376
4377         leaf = path->nodes[0];
4378         btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
4379
4380         BUG_ON(key.type != BTRFS_EXTENT_DATA_KEY &&
4381                key.type != BTRFS_EXTENT_CSUM_KEY);
4382
4383         if (btrfs_leaf_free_space(root, leaf) >= ins_len)
4384                 return 0;
4385
4386         item_size = btrfs_item_size_nr(leaf, path->slots[0]);
4387         if (key.type == BTRFS_EXTENT_DATA_KEY) {
4388                 fi = btrfs_item_ptr(leaf, path->slots[0],
4389                                     struct btrfs_file_extent_item);
4390                 extent_len = btrfs_file_extent_num_bytes(leaf, fi);
4391         }
4392         btrfs_release_path(path);
4393
4394         path->keep_locks = 1;
4395         path->search_for_split = 1;
4396         ret = btrfs_search_slot(trans, root, &key, path, 0, 1);
4397         path->search_for_split = 0;
4398         if (ret > 0)
4399                 ret = -EAGAIN;
4400         if (ret < 0)
4401                 goto err;
4402
4403         ret = -EAGAIN;
4404         leaf = path->nodes[0];
4405         /* if our item isn't there, return now */
4406         if (item_size != btrfs_item_size_nr(leaf, path->slots[0]))
4407                 goto err;
4408
4409         /* the leaf has  changed, it now has room.  return now */
4410         if (btrfs_leaf_free_space(root, path->nodes[0]) >= ins_len)
4411                 goto err;
4412
4413         if (key.type == BTRFS_EXTENT_DATA_KEY) {
4414                 fi = btrfs_item_ptr(leaf, path->slots[0],
4415                                     struct btrfs_file_extent_item);
4416                 if (extent_len != btrfs_file_extent_num_bytes(leaf, fi))
4417                         goto err;
4418         }
4419
4420         btrfs_set_path_blocking(path);
4421         ret = split_leaf(trans, root, &key, path, ins_len, 1);
4422         if (ret)
4423                 goto err;
4424
4425         path->keep_locks = 0;
4426         btrfs_unlock_up_safe(path, 1);
4427         return 0;
4428 err:
4429         path->keep_locks = 0;
4430         return ret;
4431 }
4432
4433 static noinline int split_item(struct btrfs_trans_handle *trans,
4434                                struct btrfs_root *root,
4435                                struct btrfs_path *path,
4436                                struct btrfs_key *new_key,
4437                                unsigned long split_offset)
4438 {
4439         struct extent_buffer *leaf;
4440         struct btrfs_item *item;
4441         struct btrfs_item *new_item;
4442         int slot;
4443         char *buf;
4444         u32 nritems;
4445         u32 item_size;
4446         u32 orig_offset;
4447         struct btrfs_disk_key disk_key;
4448
4449         leaf = path->nodes[0];
4450         BUG_ON(btrfs_leaf_free_space(root, leaf) < sizeof(struct btrfs_item));
4451
4452         btrfs_set_path_blocking(path);
4453
4454         item = btrfs_item_nr(path->slots[0]);
4455         orig_offset = btrfs_item_offset(leaf, item);
4456         item_size = btrfs_item_size(leaf, item);
4457
4458         buf = kmalloc(item_size, GFP_NOFS);
4459         if (!buf)
4460                 return -ENOMEM;
4461
4462         read_extent_buffer(leaf, buf, btrfs_item_ptr_offset(leaf,
4463                             path->slots[0]), item_size);
4464
4465         slot = path->slots[0] + 1;
4466         nritems = btrfs_header_nritems(leaf);
4467         if (slot != nritems) {
4468                 /* shift the items */
4469                 memmove_extent_buffer(leaf, btrfs_item_nr_offset(slot + 1),
4470                                 btrfs_item_nr_offset(slot),
4471                                 (nritems - slot) * sizeof(struct btrfs_item));
4472         }
4473
4474         btrfs_cpu_key_to_disk(&disk_key, new_key);
4475         btrfs_set_item_key(leaf, &disk_key, slot);
4476
4477         new_item = btrfs_item_nr(slot);
4478
4479         btrfs_set_item_offset(leaf, new_item, orig_offset);
4480         btrfs_set_item_size(leaf, new_item, item_size - split_offset);
4481
4482         btrfs_set_item_offset(leaf, item,
4483                               orig_offset + item_size - split_offset);
4484         btrfs_set_item_size(leaf, item, split_offset);
4485
4486         btrfs_set_header_nritems(leaf, nritems + 1);
4487
4488         /* write the data for the start of the original item */
4489         write_extent_buffer(leaf, buf,
4490                             btrfs_item_ptr_offset(leaf, path->slots[0]),
4491                             split_offset);
4492
4493         /* write the data for the new item */
4494         write_extent_buffer(leaf, buf + split_offset,
4495                             btrfs_item_ptr_offset(leaf, slot),
4496                             item_size - split_offset);
4497         btrfs_mark_buffer_dirty(leaf);
4498
4499         BUG_ON(btrfs_leaf_free_space(root, leaf) < 0);
4500         kfree(buf);
4501         return 0;
4502 }
4503
4504 /*
4505  * This function splits a single item into two items,
4506  * giving 'new_key' to the new item and splitting the
4507  * old one at split_offset (from the start of the item).
4508  *
4509  * The path may be released by this operation.  After
4510  * the split, the path is pointing to the old item.  The
4511  * new item is going to be in the same node as the old one.
4512  *
4513  * Note, the item being split must be smaller enough to live alone on
4514  * a tree block with room for one extra struct btrfs_item
4515  *
4516  * This allows us to split the item in place, keeping a lock on the
4517  * leaf the entire time.
4518  */
4519 int btrfs_split_item(struct btrfs_trans_handle *trans,
4520                      struct btrfs_root *root,
4521                      struct btrfs_path *path,
4522                      struct btrfs_key *new_key,
4523                      unsigned long split_offset)
4524 {
4525         int ret;
4526         ret = setup_leaf_for_split(trans, root, path,
4527                                    sizeof(struct btrfs_item));
4528         if (ret)
4529                 return ret;
4530
4531         ret = split_item(trans, root, path, new_key, split_offset);
4532         return ret;
4533 }
4534
4535 /*
4536  * This function duplicate a item, giving 'new_key' to the new item.
4537  * It guarantees both items live in the same tree leaf and the new item
4538  * is contiguous with the original item.
4539  *
4540  * This allows us to split file extent in place, keeping a lock on the
4541  * leaf the entire time.
4542  */
4543 int btrfs_duplicate_item(struct btrfs_trans_handle *trans,
4544                          struct btrfs_root *root,
4545                          struct btrfs_path *path,
4546                          struct btrfs_key *new_key)
4547 {
4548         struct extent_buffer *leaf;
4549         int ret;
4550         u32 item_size;
4551
4552         leaf = path->nodes[0];
4553         item_size = btrfs_item_size_nr(leaf, path->slots[0]);
4554         ret = setup_leaf_for_split(trans, root, path,
4555                                    item_size + sizeof(struct btrfs_item));
4556         if (ret)
4557                 return ret;
4558
4559         path->slots[0]++;
4560         setup_items_for_insert(root, path, new_key, &item_size,
4561                                item_size, item_size +
4562                                sizeof(struct btrfs_item), 1);
4563         leaf = path->nodes[0];
4564         memcpy_extent_buffer(leaf,
4565                              btrfs_item_ptr_offset(leaf, path->slots[0]),
4566                              btrfs_item_ptr_offset(leaf, path->slots[0] - 1),
4567                              item_size);
4568         return 0;
4569 }
4570
4571 /*
4572  * make the item pointed to by the path smaller.  new_size indicates
4573  * how small to make it, and from_end tells us if we just chop bytes
4574  * off the end of the item or if we shift the item to chop bytes off
4575  * the front.
4576  */
4577 void btrfs_truncate_item(struct btrfs_root *root, struct btrfs_path *path,
4578                          u32 new_size, int from_end)
4579 {
4580         int slot;
4581         struct extent_buffer *leaf;
4582         struct btrfs_item *item;
4583         u32 nritems;
4584         unsigned int data_end;
4585         unsigned int old_data_start;
4586         unsigned int old_size;
4587         unsigned int size_diff;
4588         int i;
4589         struct btrfs_map_token token;
4590
4591         btrfs_init_map_token(&token);
4592
4593         leaf = path->nodes[0];
4594         slot = path->slots[0];
4595
4596         old_size = btrfs_item_size_nr(leaf, slot);
4597         if (old_size == new_size)
4598                 return;
4599
4600         nritems = btrfs_header_nritems(leaf);
4601         data_end = leaf_data_end(root, leaf);
4602
4603         old_data_start = btrfs_item_offset_nr(leaf, slot);
4604
4605         size_diff = old_size - new_size;
4606
4607         BUG_ON(slot < 0);
4608         BUG_ON(slot >= nritems);
4609
4610         /*
4611          * item0..itemN ... dataN.offset..dataN.size .. data0.size
4612          */
4613         /* first correct the data pointers */
4614         for (i = slot; i < nritems; i++) {
4615                 u32 ioff;
4616                 item = btrfs_item_nr(i);
4617
4618                 ioff = btrfs_token_item_offset(leaf, item, &token);
4619                 btrfs_set_token_item_offset(leaf, item,
4620                                             ioff + size_diff, &token);
4621         }
4622
4623         /* shift the data */
4624         if (from_end) {
4625                 memmove_extent_buffer(leaf, btrfs_leaf_data(leaf) +
4626                               data_end + size_diff, btrfs_leaf_data(leaf) +
4627                               data_end, old_data_start + new_size - data_end);
4628         } else {
4629                 struct btrfs_disk_key disk_key;
4630                 u64 offset;
4631
4632                 btrfs_item_key(leaf, &disk_key, slot);
4633
4634                 if (btrfs_disk_key_type(&disk_key) == BTRFS_EXTENT_DATA_KEY) {
4635                         unsigned long ptr;
4636                         struct btrfs_file_extent_item *fi;
4637
4638                         fi = btrfs_item_ptr(leaf, slot,
4639                                             struct btrfs_file_extent_item);
4640                         fi = (struct btrfs_file_extent_item *)(
4641                              (unsigned long)fi - size_diff);
4642
4643                         if (btrfs_file_extent_type(leaf, fi) ==
4644                             BTRFS_FILE_EXTENT_INLINE) {
4645                                 ptr = btrfs_item_ptr_offset(leaf, slot);
4646                                 memmove_extent_buffer(leaf, ptr,
4647                                       (unsigned long)fi,
4648                                       BTRFS_FILE_EXTENT_INLINE_DATA_START);
4649                         }
4650                 }
4651
4652                 memmove_extent_buffer(leaf, btrfs_leaf_data(leaf) +
4653                               data_end + size_diff, btrfs_leaf_data(leaf) +
4654                               data_end, old_data_start - data_end);
4655
4656                 offset = btrfs_disk_key_offset(&disk_key);
4657                 btrfs_set_disk_key_offset(&disk_key, offset + size_diff);
4658                 btrfs_set_item_key(leaf, &disk_key, slot);
4659                 if (slot == 0)
4660                         fixup_low_keys(root->fs_info, path, &disk_key, 1);
4661         }
4662
4663         item = btrfs_item_nr(slot);
4664         btrfs_set_item_size(leaf, item, new_size);
4665         btrfs_mark_buffer_dirty(leaf);
4666
4667         if (btrfs_leaf_free_space(root, leaf) < 0) {
4668                 btrfs_print_leaf(root, leaf);
4669                 BUG();
4670         }
4671 }
4672
4673 /*
4674  * make the item pointed to by the path bigger, data_size is the added size.
4675  */
4676 void btrfs_extend_item(struct btrfs_root *root, struct btrfs_path *path,
4677                        u32 data_size)
4678 {
4679         int slot;
4680         struct extent_buffer *leaf;
4681         struct btrfs_item *item;
4682         u32 nritems;
4683         unsigned int data_end;
4684         unsigned int old_data;
4685         unsigned int old_size;
4686         int i;
4687         struct btrfs_map_token token;
4688
4689         btrfs_init_map_token(&token);
4690
4691         leaf = path->nodes[0];
4692
4693         nritems = btrfs_header_nritems(leaf);
4694         data_end = leaf_data_end(root, leaf);
4695
4696         if (btrfs_leaf_free_space(root, leaf) < data_size) {
4697                 btrfs_print_leaf(root, leaf);
4698                 BUG();
4699         }
4700         slot = path->slots[0];
4701         old_data = btrfs_item_end_nr(leaf, slot);
4702
4703         BUG_ON(slot < 0);
4704         if (slot >= nritems) {
4705                 btrfs_print_leaf(root, leaf);
4706                 btrfs_crit(root->fs_info, "slot %d too large, nritems %d",
4707                        slot, nritems);
4708                 BUG_ON(1);
4709         }
4710
4711         /*
4712          * item0..itemN ... dataN.offset..dataN.size .. data0.size
4713          */
4714         /* first correct the data pointers */
4715         for (i = slot; i < nritems; i++) {
4716                 u32 ioff;
4717                 item = btrfs_item_nr(i);
4718
4719                 ioff = btrfs_token_item_offset(leaf, item, &token);
4720                 btrfs_set_token_item_offset(leaf, item,
4721                                             ioff - data_size, &token);
4722         }
4723
4724         /* shift the data */
4725         memmove_extent_buffer(leaf, btrfs_leaf_data(leaf) +
4726                       data_end - data_size, btrfs_leaf_data(leaf) +
4727                       data_end, old_data - data_end);
4728
4729         data_end = old_data;
4730         old_size = btrfs_item_size_nr(leaf, slot);
4731         item = btrfs_item_nr(slot);
4732         btrfs_set_item_size(leaf, item, old_size + data_size);
4733         btrfs_mark_buffer_dirty(leaf);
4734
4735         if (btrfs_leaf_free_space(root, leaf) < 0) {
4736                 btrfs_print_leaf(root, leaf);
4737                 BUG();
4738         }
4739 }
4740
4741 /*
4742  * this is a helper for btrfs_insert_empty_items, the main goal here is
4743  * to save stack depth by doing the bulk of the work in a function
4744  * that doesn't call btrfs_search_slot
4745  */
4746 void setup_items_for_insert(struct btrfs_root *root, struct btrfs_path *path,
4747                             struct btrfs_key *cpu_key, u32 *data_size,
4748                             u32 total_data, u32 total_size, int nr)
4749 {
4750         struct btrfs_item *item;
4751         int i;
4752         u32 nritems;
4753         unsigned int data_end;
4754         struct btrfs_disk_key disk_key;
4755         struct extent_buffer *leaf;
4756         int slot;
4757         struct btrfs_map_token token;
4758
4759         if (path->slots[0] == 0) {
4760                 btrfs_cpu_key_to_disk(&disk_key, cpu_key);
4761                 fixup_low_keys(root->fs_info, path, &disk_key, 1);
4762         }
4763         btrfs_unlock_up_safe(path, 1);
4764
4765         btrfs_init_map_token(&token);
4766
4767         leaf = path->nodes[0];
4768         slot = path->slots[0];
4769
4770         nritems = btrfs_header_nritems(leaf);
4771         data_end = leaf_data_end(root, leaf);
4772
4773         if (btrfs_leaf_free_space(root, leaf) < total_size) {
4774                 btrfs_print_leaf(root, leaf);
4775                 btrfs_crit(root->fs_info, "not enough freespace need %u have %d",
4776                        total_size, btrfs_leaf_free_space(root, leaf));
4777                 BUG();
4778         }
4779
4780         if (slot != nritems) {
4781                 unsigned int old_data = btrfs_item_end_nr(leaf, slot);
4782
4783                 if (old_data < data_end) {
4784                         btrfs_print_leaf(root, leaf);
4785                         btrfs_crit(root->fs_info, "slot %d old_data %d data_end %d",
4786                                slot, old_data, data_end);
4787                         BUG_ON(1);
4788                 }
4789                 /*
4790                  * item0..itemN ... dataN.offset..dataN.size .. data0.size
4791                  */
4792                 /* first correct the data pointers */
4793                 for (i = slot; i < nritems; i++) {
4794                         u32 ioff;
4795
4796                         item = btrfs_item_nr( i);
4797                         ioff = btrfs_token_item_offset(leaf, item, &token);
4798                         btrfs_set_token_item_offset(leaf, item,
4799                                                     ioff - total_data, &token);
4800                 }
4801                 /* shift the items */
4802                 memmove_extent_buffer(leaf, btrfs_item_nr_offset(slot + nr),
4803                               btrfs_item_nr_offset(slot),
4804                               (nritems - slot) * sizeof(struct btrfs_item));
4805
4806                 /* shift the data */
4807                 memmove_extent_buffer(leaf, btrfs_leaf_data(leaf) +
4808                               data_end - total_data, btrfs_leaf_data(leaf) +
4809                               data_end, old_data - data_end);
4810                 data_end = old_data;
4811         }
4812
4813         /* setup the item for the new data */
4814         for (i = 0; i < nr; i++) {
4815                 btrfs_cpu_key_to_disk(&disk_key, cpu_key + i);
4816                 btrfs_set_item_key(leaf, &disk_key, slot + i);
4817                 item = btrfs_item_nr(slot + i);
4818                 btrfs_set_token_item_offset(leaf, item,
4819                                             data_end - data_size[i], &token);
4820                 data_end -= data_size[i];
4821                 btrfs_set_token_item_size(leaf, item, data_size[i], &token);
4822         }
4823
4824         btrfs_set_header_nritems(leaf, nritems + nr);
4825         btrfs_mark_buffer_dirty(leaf);
4826
4827         if (btrfs_leaf_free_space(root, leaf) < 0) {
4828                 btrfs_print_leaf(root, leaf);
4829                 BUG();
4830         }
4831 }
4832
4833 /*
4834  * Given a key and some data, insert items into the tree.
4835  * This does all the path init required, making room in the tree if needed.
4836  */
4837 int btrfs_insert_empty_items(struct btrfs_trans_handle *trans,
4838                             struct btrfs_root *root,
4839                             struct btrfs_path *path,
4840                             struct btrfs_key *cpu_key, u32 *data_size,
4841                             int nr)
4842 {
4843         int ret = 0;
4844         int slot;
4845         int i;
4846         u32 total_size = 0;
4847         u32 total_data = 0;
4848
4849         for (i = 0; i < nr; i++)
4850                 total_data += data_size[i];
4851
4852         total_size = total_data + (nr * sizeof(struct btrfs_item));
4853         ret = btrfs_search_slot(trans, root, cpu_key, path, total_size, 1);
4854         if (ret == 0)
4855                 return -EEXIST;
4856         if (ret < 0)
4857                 return ret;
4858
4859         slot = path->slots[0];
4860         BUG_ON(slot < 0);
4861
4862         setup_items_for_insert(root, path, cpu_key, data_size,
4863                                total_data, total_size, nr);
4864         return 0;
4865 }
4866
4867 /*
4868  * Given a key and some data, insert an item into the tree.
4869  * This does all the path init required, making room in the tree if needed.
4870  */
4871 int btrfs_insert_item(struct btrfs_trans_handle *trans, struct btrfs_root
4872                       *root, struct btrfs_key *cpu_key, void *data, u32
4873                       data_size)
4874 {
4875         int ret = 0;
4876         struct btrfs_path *path;
4877         struct extent_buffer *leaf;
4878         unsigned long ptr;
4879
4880         path = btrfs_alloc_path();
4881         if (!path)
4882                 return -ENOMEM;
4883         ret = btrfs_insert_empty_item(trans, root, path, cpu_key, data_size);
4884         if (!ret) {
4885                 leaf = path->nodes[0];
4886                 ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
4887                 write_extent_buffer(leaf, data, ptr, data_size);
4888                 btrfs_mark_buffer_dirty(leaf);
4889         }
4890         btrfs_free_path(path);
4891         return ret;
4892 }
4893
4894 /*
4895  * delete the pointer from a given node.
4896  *
4897  * the tree should have been previously balanced so the deletion does not
4898  * empty a node.
4899  */
4900 static void del_ptr(struct btrfs_root *root, struct btrfs_path *path,
4901                     int level, int slot)
4902 {
4903         struct extent_buffer *parent = path->nodes[level];
4904         u32 nritems;
4905         int ret;
4906
4907         nritems = btrfs_header_nritems(parent);
4908         if (slot != nritems - 1) {
4909                 if (level)
4910                         tree_mod_log_eb_move(root->fs_info, parent, slot,
4911                                              slot + 1, nritems - slot - 1);
4912                 memmove_extent_buffer(parent,
4913                               btrfs_node_key_ptr_offset(slot),
4914                               btrfs_node_key_ptr_offset(slot + 1),
4915                               sizeof(struct btrfs_key_ptr) *
4916                               (nritems - slot - 1));
4917         } else if (level) {
4918                 ret = tree_mod_log_insert_key(root->fs_info, parent, slot,
4919                                               MOD_LOG_KEY_REMOVE, GFP_NOFS);
4920                 BUG_ON(ret < 0);
4921         }
4922
4923         nritems--;
4924         btrfs_set_header_nritems(parent, nritems);
4925         if (nritems == 0 && parent == root->node) {
4926                 BUG_ON(btrfs_header_level(root->node) != 1);
4927                 /* just turn the root into a leaf and break */
4928                 btrfs_set_header_level(root->node, 0);
4929         } else if (slot == 0) {
4930                 struct btrfs_disk_key disk_key;
4931
4932                 btrfs_node_key(parent, &disk_key, 0);
4933                 fixup_low_keys(root->fs_info, path, &disk_key, level + 1);
4934         }
4935         btrfs_mark_buffer_dirty(parent);
4936 }
4937
4938 /*
4939  * a helper function to delete the leaf pointed to by path->slots[1] and
4940  * path->nodes[1].
4941  *
4942  * This deletes the pointer in path->nodes[1] and frees the leaf
4943  * block extent.  zero is returned if it all worked out, < 0 otherwise.
4944  *
4945  * The path must have already been setup for deleting the leaf, including
4946  * all the proper balancing.  path->nodes[1] must be locked.
4947  */
4948 static noinline void btrfs_del_leaf(struct btrfs_trans_handle *trans,
4949                                     struct btrfs_root *root,
4950                                     struct btrfs_path *path,
4951                                     struct extent_buffer *leaf)
4952 {
4953         WARN_ON(btrfs_header_generation(leaf) != trans->transid);
4954         del_ptr(root, path, 1, path->slots[1]);
4955
4956         /*
4957          * btrfs_free_extent is expensive, we want to make sure we
4958          * aren't holding any locks when we call it
4959          */
4960         btrfs_unlock_up_safe(path, 0);
4961
4962         root_sub_used(root, leaf->len);
4963
4964         extent_buffer_get(leaf);
4965         btrfs_free_tree_block(trans, root, leaf, 0, 1);
4966         free_extent_buffer_stale(leaf);
4967 }
4968 /*
4969  * delete the item at the leaf level in path.  If that empties
4970  * the leaf, remove it from the tree
4971  */
4972 int btrfs_del_items(struct btrfs_trans_handle *trans, struct btrfs_root *root,
4973                     struct btrfs_path *path, int slot, int nr)
4974 {
4975         struct extent_buffer *leaf;
4976         struct btrfs_item *item;
4977         u32 last_off;
4978         u32 dsize = 0;
4979         int ret = 0;
4980         int wret;
4981         int i;
4982         u32 nritems;
4983         struct btrfs_map_token token;
4984
4985         btrfs_init_map_token(&token);
4986
4987         leaf = path->nodes[0];
4988         last_off = btrfs_item_offset_nr(leaf, slot + nr - 1);
4989
4990         for (i = 0; i < nr; i++)
4991                 dsize += btrfs_item_size_nr(leaf, slot + i);
4992
4993         nritems = btrfs_header_nritems(leaf);
4994
4995         if (slot + nr != nritems) {
4996                 int data_end = leaf_data_end(root, leaf);
4997
4998                 memmove_extent_buffer(leaf, btrfs_leaf_data(leaf) +
4999                               data_end + dsize,
5000                               btrfs_leaf_data(leaf) + data_end,
5001                               last_off - data_end);
5002
5003                 for (i = slot + nr; i < nritems; i++) {
5004                         u32 ioff;
5005
5006                         item = btrfs_item_nr(i);
5007                         ioff = btrfs_token_item_offset(leaf, item, &token);
5008                         btrfs_set_token_item_offset(leaf, item,
5009                                                     ioff + dsize, &token);
5010                 }
5011
5012                 memmove_extent_buffer(leaf, btrfs_item_nr_offset(slot),
5013                               btrfs_item_nr_offset(slot + nr),
5014                               sizeof(struct btrfs_item) *
5015                               (nritems - slot - nr));
5016         }
5017         btrfs_set_header_nritems(leaf, nritems - nr);
5018         nritems -= nr;
5019
5020         /* delete the leaf if we've emptied it */
5021         if (nritems == 0) {
5022                 if (leaf == root->node) {
5023                         btrfs_set_header_level(leaf, 0);
5024                 } else {
5025                         btrfs_set_path_blocking(path);
5026                         clean_tree_block(trans, root->fs_info, leaf);
5027                         btrfs_del_leaf(trans, root, path, leaf);
5028                 }
5029         } else {
5030                 int used = leaf_space_used(leaf, 0, nritems);
5031                 if (slot == 0) {
5032                         struct btrfs_disk_key disk_key;
5033
5034                         btrfs_item_key(leaf, &disk_key, 0);
5035                         fixup_low_keys(root->fs_info, path, &disk_key, 1);
5036                 }
5037
5038                 /* delete the leaf if it is mostly empty */
5039                 if (used < BTRFS_LEAF_DATA_SIZE(root) / 3) {
5040                         /* push_leaf_left fixes the path.
5041                          * make sure the path still points to our leaf
5042                          * for possible call to del_ptr below
5043                          */
5044                         slot = path->slots[1];
5045                         extent_buffer_get(leaf);
5046
5047                         btrfs_set_path_blocking(path);
5048                         wret = push_leaf_left(trans, root, path, 1, 1,
5049                                               1, (u32)-1);
5050                         if (wret < 0 && wret != -ENOSPC)
5051                                 ret = wret;
5052
5053                         if (path->nodes[0] == leaf &&
5054                             btrfs_header_nritems(leaf)) {
5055                                 wret = push_leaf_right(trans, root, path, 1,
5056                                                        1, 1, 0);
5057                                 if (wret < 0 && wret != -ENOSPC)
5058                                         ret = wret;
5059                         }
5060
5061                         if (btrfs_header_nritems(leaf) == 0) {
5062                                 path->slots[1] = slot;
5063                                 btrfs_del_leaf(trans, root, path, leaf);
5064                                 free_extent_buffer(leaf);
5065                                 ret = 0;
5066                         } else {
5067                                 /* if we're still in the path, make sure
5068                                  * we're dirty.  Otherwise, one of the
5069                                  * push_leaf functions must have already
5070                                  * dirtied this buffer
5071                                  */
5072                                 if (path->nodes[0] == leaf)
5073                                         btrfs_mark_buffer_dirty(leaf);
5074                                 free_extent_buffer(leaf);
5075                         }
5076                 } else {
5077                         btrfs_mark_buffer_dirty(leaf);
5078                 }
5079         }
5080         return ret;
5081 }
5082
5083 /*
5084  * search the tree again to find a leaf with lesser keys
5085  * returns 0 if it found something or 1 if there are no lesser leaves.
5086  * returns < 0 on io errors.
5087  *
5088  * This may release the path, and so you may lose any locks held at the
5089  * time you call it.
5090  */
5091 int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path)
5092 {
5093         struct btrfs_key key;
5094         struct btrfs_disk_key found_key;
5095         int ret;
5096
5097         btrfs_item_key_to_cpu(path->nodes[0], &key, 0);
5098
5099         if (key.offset > 0) {
5100                 key.offset--;
5101         } else if (key.type > 0) {
5102                 key.type--;
5103                 key.offset = (u64)-1;
5104         } else if (key.objectid > 0) {
5105                 key.objectid--;
5106                 key.type = (u8)-1;
5107                 key.offset = (u64)-1;
5108         } else {
5109                 return 1;
5110         }
5111
5112         btrfs_release_path(path);
5113         ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
5114         if (ret < 0)
5115                 return ret;
5116         btrfs_item_key(path->nodes[0], &found_key, 0);
5117         ret = comp_keys(&found_key, &key);
5118         /*
5119          * We might have had an item with the previous key in the tree right
5120          * before we released our path. And after we released our path, that
5121          * item might have been pushed to the first slot (0) of the leaf we
5122          * were holding due to a tree balance. Alternatively, an item with the
5123          * previous key can exist as the only element of a leaf (big fat item).
5124          * Therefore account for these 2 cases, so that our callers (like
5125          * btrfs_previous_item) don't miss an existing item with a key matching
5126          * the previous key we computed above.
5127          */
5128         if (ret <= 0)
5129                 return 0;
5130         return 1;
5131 }
5132
5133 /*
5134  * A helper function to walk down the tree starting at min_key, and looking
5135  * for nodes or leaves that are have a minimum transaction id.
5136  * This is used by the btree defrag code, and tree logging
5137  *
5138  * This does not cow, but it does stuff the starting key it finds back
5139  * into min_key, so you can call btrfs_search_slot with cow=1 on the
5140  * key and get a writable path.
5141  *
5142  * This does lock as it descends, and path->keep_locks should be set
5143  * to 1 by the caller.
5144  *
5145  * This honors path->lowest_level to prevent descent past a given level
5146  * of the tree.
5147  *
5148  * min_trans indicates the oldest transaction that you are interested
5149  * in walking through.  Any nodes or leaves older than min_trans are
5150  * skipped over (without reading them).
5151  *
5152  * returns zero if something useful was found, < 0 on error and 1 if there
5153  * was nothing in the tree that matched the search criteria.
5154  */
5155 int btrfs_search_forward(struct btrfs_root *root, struct btrfs_key *min_key,
5156                          struct btrfs_path *path,
5157                          u64 min_trans)
5158 {
5159         struct extent_buffer *cur;
5160         struct btrfs_key found_key;
5161         int slot;
5162         int sret;
5163         u32 nritems;
5164         int level;
5165         int ret = 1;
5166         int keep_locks = path->keep_locks;
5167
5168         path->keep_locks = 1;
5169 again:
5170         cur = btrfs_read_lock_root_node(root);
5171         level = btrfs_header_level(cur);
5172         WARN_ON(path->nodes[level]);
5173         path->nodes[level] = cur;
5174         path->locks[level] = BTRFS_READ_LOCK;
5175
5176         if (btrfs_header_generation(cur) < min_trans) {
5177                 ret = 1;
5178                 goto out;
5179         }
5180         while (1) {
5181                 nritems = btrfs_header_nritems(cur);
5182                 level = btrfs_header_level(cur);
5183                 sret = bin_search(cur, min_key, level, &slot);
5184
5185                 /* at the lowest level, we're done, setup the path and exit */
5186                 if (level == path->lowest_level) {
5187                         if (slot >= nritems)
5188                                 goto find_next_key;
5189                         ret = 0;
5190                         path->slots[level] = slot;
5191                         btrfs_item_key_to_cpu(cur, &found_key, slot);
5192                         goto out;
5193                 }
5194                 if (sret && slot > 0)
5195                         slot--;
5196                 /*
5197                  * check this node pointer against the min_trans parameters.
5198                  * If it is too old, old, skip to the next one.
5199                  */
5200                 while (slot < nritems) {
5201                         u64 gen;
5202
5203                         gen = btrfs_node_ptr_generation(cur, slot);
5204                         if (gen < min_trans) {
5205                                 slot++;
5206                                 continue;
5207                         }
5208                         break;
5209                 }
5210 find_next_key:
5211                 /*
5212                  * we didn't find a candidate key in this node, walk forward
5213                  * and find another one
5214                  */
5215                 if (slot >= nritems) {
5216                         path->slots[level] = slot;
5217                         btrfs_set_path_blocking(path);
5218                         sret = btrfs_find_next_key(root, path, min_key, level,
5219                                                   min_trans);
5220                         if (sret == 0) {
5221                                 btrfs_release_path(path);
5222                                 goto again;
5223                         } else {
5224                                 goto out;
5225                         }
5226                 }
5227                 /* save our key for returning back */
5228                 btrfs_node_key_to_cpu(cur, &found_key, slot);
5229                 path->slots[level] = slot;
5230                 if (level == path->lowest_level) {
5231                         ret = 0;
5232                         goto out;
5233                 }
5234                 btrfs_set_path_blocking(path);
5235                 cur = read_node_slot(root, cur, slot);
5236                 if (IS_ERR(cur)) {
5237                         ret = PTR_ERR(cur);
5238                         goto out;
5239                 }
5240
5241                 btrfs_tree_read_lock(cur);
5242
5243                 path->locks[level - 1] = BTRFS_READ_LOCK;
5244                 path->nodes[level - 1] = cur;
5245                 unlock_up(path, level, 1, 0, NULL);
5246                 btrfs_clear_path_blocking(path, NULL, 0);
5247         }
5248 out:
5249         path->keep_locks = keep_locks;
5250         if (ret == 0) {
5251                 btrfs_unlock_up_safe(path, path->lowest_level + 1);
5252                 btrfs_set_path_blocking(path);
5253                 memcpy(min_key, &found_key, sizeof(found_key));
5254         }
5255         return ret;
5256 }
5257
5258 static int tree_move_down(struct btrfs_root *root,
5259                            struct btrfs_path *path,
5260                            int *level, int root_level)
5261 {
5262         struct extent_buffer *eb;
5263
5264         BUG_ON(*level == 0);
5265         eb = read_node_slot(root, path->nodes[*level], path->slots[*level]);
5266         if (IS_ERR(eb))
5267                 return PTR_ERR(eb);
5268
5269         path->nodes[*level - 1] = eb;
5270         path->slots[*level - 1] = 0;
5271         (*level)--;
5272         return 0;
5273 }
5274
5275 static int tree_move_next_or_upnext(struct btrfs_root *root,
5276                                     struct btrfs_path *path,
5277                                     int *level, int root_level)
5278 {
5279         int ret = 0;
5280         int nritems;
5281         nritems = btrfs_header_nritems(path->nodes[*level]);
5282
5283         path->slots[*level]++;
5284
5285         while (path->slots[*level] >= nritems) {
5286                 if (*level == root_level)
5287                         return -1;
5288
5289                 /* move upnext */
5290                 path->slots[*level] = 0;
5291                 free_extent_buffer(path->nodes[*level]);
5292                 path->nodes[*level] = NULL;
5293                 (*level)++;
5294                 path->slots[*level]++;
5295
5296                 nritems = btrfs_header_nritems(path->nodes[*level]);
5297                 ret = 1;
5298         }
5299         return ret;
5300 }
5301
5302 /*
5303  * Returns 1 if it had to move up and next. 0 is returned if it moved only next
5304  * or down.
5305  */
5306 static int tree_advance(struct btrfs_root *root,
5307                         struct btrfs_path *path,
5308                         int *level, int root_level,
5309                         int allow_down,
5310                         struct btrfs_key *key)
5311 {
5312         int ret;
5313
5314         if (*level == 0 || !allow_down) {
5315                 ret = tree_move_next_or_upnext(root, path, level, root_level);
5316         } else {
5317                 ret = tree_move_down(root, path, level, root_level);
5318         }
5319         if (ret >= 0) {
5320                 if (*level == 0)
5321                         btrfs_item_key_to_cpu(path->nodes[*level], key,
5322                                         path->slots[*level]);
5323                 else
5324                         btrfs_node_key_to_cpu(path->nodes[*level], key,
5325                                         path->slots[*level]);
5326         }
5327         return ret;
5328 }
5329
5330 static int tree_compare_item(struct btrfs_root *left_root,
5331                              struct btrfs_path *left_path,
5332                              struct btrfs_path *right_path,
5333                              char *tmp_buf)
5334 {
5335         int cmp;
5336         int len1, len2;
5337         unsigned long off1, off2;
5338
5339         len1 = btrfs_item_size_nr(left_path->nodes[0], left_path->slots[0]);
5340         len2 = btrfs_item_size_nr(right_path->nodes[0], right_path->slots[0]);
5341         if (len1 != len2)
5342                 return 1;
5343
5344         off1 = btrfs_item_ptr_offset(left_path->nodes[0], left_path->slots[0]);
5345         off2 = btrfs_item_ptr_offset(right_path->nodes[0],
5346                                 right_path->slots[0]);
5347
5348         read_extent_buffer(left_path->nodes[0], tmp_buf, off1, len1);
5349
5350         cmp = memcmp_extent_buffer(right_path->nodes[0], tmp_buf, off2, len1);
5351         if (cmp)
5352                 return 1;
5353         return 0;
5354 }
5355
5356 #define ADVANCE 1
5357 #define ADVANCE_ONLY_NEXT -1
5358
5359 /*
5360  * This function compares two trees and calls the provided callback for
5361  * every changed/new/deleted item it finds.
5362  * If shared tree blocks are encountered, whole subtrees are skipped, making
5363  * the compare pretty fast on snapshotted subvolumes.
5364  *
5365  * This currently works on commit roots only. As commit roots are read only,
5366  * we don't do any locking. The commit roots are protected with transactions.
5367  * Transactions are ended and rejoined when a commit is tried in between.
5368  *
5369  * This function checks for modifications done to the trees while comparing.
5370  * If it detects a change, it aborts immediately.
5371  */
5372 int btrfs_compare_trees(struct btrfs_root *left_root,
5373                         struct btrfs_root *right_root,
5374                         btrfs_changed_cb_t changed_cb, void *ctx)
5375 {
5376         int ret;
5377         int cmp;
5378         struct btrfs_path *left_path = NULL;
5379         struct btrfs_path *right_path = NULL;
5380         struct btrfs_key left_key;
5381         struct btrfs_key right_key;
5382         char *tmp_buf = NULL;
5383         int left_root_level;
5384         int right_root_level;
5385         int left_level;
5386         int right_level;
5387         int left_end_reached;
5388         int right_end_reached;
5389         int advance_left;
5390         int advance_right;
5391         u64 left_blockptr;
5392         u64 right_blockptr;
5393         u64 left_gen;
5394         u64 right_gen;
5395
5396         left_path = btrfs_alloc_path();
5397         if (!left_path) {
5398                 ret = -ENOMEM;
5399                 goto out;
5400         }
5401         right_path = btrfs_alloc_path();
5402         if (!right_path) {
5403                 ret = -ENOMEM;
5404                 goto out;
5405         }
5406
5407         tmp_buf = kmalloc(left_root->nodesize, GFP_KERNEL | __GFP_NOWARN);
5408         if (!tmp_buf) {
5409                 tmp_buf = vmalloc(left_root->nodesize);
5410                 if (!tmp_buf) {
5411                         ret = -ENOMEM;
5412                         goto out;
5413                 }
5414         }
5415
5416         left_path->search_commit_root = 1;
5417         left_path->skip_locking = 1;
5418         right_path->search_commit_root = 1;
5419         right_path->skip_locking = 1;
5420
5421         /*
5422          * Strategy: Go to the first items of both trees. Then do
5423          *
5424          * If both trees are at level 0
5425          *   Compare keys of current items
5426          *     If left < right treat left item as new, advance left tree
5427          *       and repeat
5428          *     If left > right treat right item as deleted, advance right tree
5429          *       and repeat
5430          *     If left == right do deep compare of items, treat as changed if
5431          *       needed, advance both trees and repeat
5432          * If both trees are at the same level but not at level 0
5433          *   Compare keys of current nodes/leafs
5434          *     If left < right advance left tree and repeat
5435          *     If left > right advance right tree and repeat
5436          *     If left == right compare blockptrs of the next nodes/leafs
5437          *       If they match advance both trees but stay at the same level
5438          *         and repeat
5439          *       If they don't match advance both trees while allowing to go
5440          *         deeper and repeat
5441          * If tree levels are different
5442          *   Advance the tree that needs it and repeat
5443          *
5444          * Advancing a tree means:
5445          *   If we are at level 0, try to go to the next slot. If that's not
5446          *   possible, go one level up and repeat. Stop when we found a level
5447          *   where we could go to the next slot. We may at this point be on a
5448          *   node or a leaf.
5449          *
5450          *   If we are not at level 0 and not on shared tree blocks, go one
5451          *   level deeper.
5452          *
5453          *   If we are not at level 0 and on shared tree blocks, go one slot to
5454          *   the right if possible or go up and right.
5455          */
5456
5457         down_read(&left_root->fs_info->commit_root_sem);
5458         left_level = btrfs_header_level(left_root->commit_root);
5459         left_root_level = left_level;
5460         left_path->nodes[left_level] = left_root->commit_root;
5461         extent_buffer_get(left_path->nodes[left_level]);
5462
5463         right_level = btrfs_header_level(right_root->commit_root);
5464         right_root_level = right_level;
5465         right_path->nodes[right_level] = right_root->commit_root;
5466         extent_buffer_get(right_path->nodes[right_level]);
5467         up_read(&left_root->fs_info->commit_root_sem);
5468
5469         if (left_level == 0)
5470                 btrfs_item_key_to_cpu(left_path->nodes[left_level],
5471                                 &left_key, left_path->slots[left_level]);
5472         else
5473                 btrfs_node_key_to_cpu(left_path->nodes[left_level],
5474                                 &left_key, left_path->slots[left_level]);
5475         if (right_level == 0)
5476                 btrfs_item_key_to_cpu(right_path->nodes[right_level],
5477                                 &right_key, right_path->slots[right_level]);
5478         else
5479                 btrfs_node_key_to_cpu(right_path->nodes[right_level],
5480                                 &right_key, right_path->slots[right_level]);
5481
5482         left_end_reached = right_end_reached = 0;
5483         advance_left = advance_right = 0;
5484
5485         while (1) {
5486                 if (advance_left && !left_end_reached) {
5487                         ret = tree_advance(left_root, left_path, &left_level,
5488                                         left_root_level,
5489                                         advance_left != ADVANCE_ONLY_NEXT,
5490                                         &left_key);
5491                         if (ret == -1)
5492                                 left_end_reached = ADVANCE;
5493                         else if (ret < 0)
5494                                 goto out;
5495                         advance_left = 0;
5496                 }
5497                 if (advance_right && !right_end_reached) {
5498                         ret = tree_advance(right_root, right_path, &right_level,
5499                                         right_root_level,
5500                                         advance_right != ADVANCE_ONLY_NEXT,
5501                                         &right_key);
5502                         if (ret == -1)
5503                                 right_end_reached = ADVANCE;
5504                         else if (ret < 0)
5505                                 goto out;
5506                         advance_right = 0;
5507                 }
5508
5509                 if (left_end_reached && right_end_reached) {
5510                         ret = 0;
5511                         goto out;
5512                 } else if (left_end_reached) {
5513                         if (right_level == 0) {
5514                                 ret = changed_cb(left_root, right_root,
5515                                                 left_path, right_path,
5516                                                 &right_key,
5517                                                 BTRFS_COMPARE_TREE_DELETED,
5518                                                 ctx);
5519                                 if (ret < 0)
5520                                         goto out;
5521                         }
5522                         advance_right = ADVANCE;
5523                         continue;
5524                 } else if (right_end_reached) {
5525                         if (left_level == 0) {
5526                                 ret = changed_cb(left_root, right_root,
5527                                                 left_path, right_path,
5528                                                 &left_key,
5529                                                 BTRFS_COMPARE_TREE_NEW,
5530                                                 ctx);
5531                                 if (ret < 0)
5532                                         goto out;
5533                         }
5534                         advance_left = ADVANCE;
5535                         continue;
5536                 }
5537
5538                 if (left_level == 0 && right_level == 0) {
5539                         cmp = btrfs_comp_cpu_keys(&left_key, &right_key);
5540                         if (cmp < 0) {
5541                                 ret = changed_cb(left_root, right_root,
5542                                                 left_path, right_path,
5543                                                 &left_key,
5544                                                 BTRFS_COMPARE_TREE_NEW,
5545                                                 ctx);
5546                                 if (ret < 0)
5547                                         goto out;
5548                                 advance_left = ADVANCE;
5549                         } else if (cmp > 0) {
5550                                 ret = changed_cb(left_root, right_root,
5551                                                 left_path, right_path,
5552                                                 &right_key,
5553                                                 BTRFS_COMPARE_TREE_DELETED,
5554                                                 ctx);
5555                                 if (ret < 0)
5556                                         goto out;
5557                                 advance_right = ADVANCE;
5558                         } else {
5559                                 enum btrfs_compare_tree_result result;
5560
5561                                 WARN_ON(!extent_buffer_uptodate(left_path->nodes[0]));
5562                                 ret = tree_compare_item(left_root, left_path,
5563                                                 right_path, tmp_buf);
5564                                 if (ret)
5565                                         result = BTRFS_COMPARE_TREE_CHANGED;
5566                                 else
5567                                         result = BTRFS_COMPARE_TREE_SAME;
5568                                 ret = changed_cb(left_root, right_root,
5569                                                  left_path, right_path,
5570                                                  &left_key, result, ctx);
5571                                 if (ret < 0)
5572                                         goto out;
5573                                 advance_left = ADVANCE;
5574                                 advance_right = ADVANCE;
5575                         }
5576                 } else if (left_level == right_level) {
5577                         cmp = btrfs_comp_cpu_keys(&left_key, &right_key);
5578                         if (cmp < 0) {
5579                                 advance_left = ADVANCE;
5580                         } else if (cmp > 0) {
5581                                 advance_right = ADVANCE;
5582                         } else {
5583                                 left_blockptr = btrfs_node_blockptr(
5584                                                 left_path->nodes[left_level],
5585                                                 left_path->slots[left_level]);
5586                                 right_blockptr = btrfs_node_blockptr(
5587                                                 right_path->nodes[right_level],
5588                                                 right_path->slots[right_level]);
5589                                 left_gen = btrfs_node_ptr_generation(
5590                                                 left_path->nodes[left_level],
5591                                                 left_path->slots[left_level]);
5592                                 right_gen = btrfs_node_ptr_generation(
5593                                                 right_path->nodes[right_level],
5594                                                 right_path->slots[right_level]);
5595                                 if (left_blockptr == right_blockptr &&
5596                                     left_gen == right_gen) {
5597                                         /*
5598                                          * As we're on a shared block, don't
5599                                          * allow to go deeper.
5600                                          */
5601                                         advance_left = ADVANCE_ONLY_NEXT;
5602                                         advance_right = ADVANCE_ONLY_NEXT;
5603                                 } else {
5604                                         advance_left = ADVANCE;
5605                                         advance_right = ADVANCE;
5606                                 }
5607                         }
5608                 } else if (left_level < right_level) {
5609                         advance_right = ADVANCE;
5610                 } else {
5611                         advance_left = ADVANCE;
5612                 }
5613         }
5614
5615 out:
5616         btrfs_free_path(left_path);
5617         btrfs_free_path(right_path);
5618         kvfree(tmp_buf);
5619         return ret;
5620 }
5621
5622 /*
5623  * this is similar to btrfs_next_leaf, but does not try to preserve
5624  * and fixup the path.  It looks for and returns the next key in the
5625  * tree based on the current path and the min_trans parameters.
5626  *
5627  * 0 is returned if another key is found, < 0 if there are any errors
5628  * and 1 is returned if there are no higher keys in the tree
5629  *
5630  * path->keep_locks should be set to 1 on the search made before
5631  * calling this function.
5632  */
5633 int btrfs_find_next_key(struct btrfs_root *root, struct btrfs_path *path,
5634                         struct btrfs_key *key, int level, u64 min_trans)
5635 {
5636         int slot;
5637         struct extent_buffer *c;
5638
5639         WARN_ON(!path->keep_locks);
5640         while (level < BTRFS_MAX_LEVEL) {
5641                 if (!path->nodes[level])
5642                         return 1;
5643
5644                 slot = path->slots[level] + 1;
5645                 c = path->nodes[level];
5646 next:
5647                 if (slot >= btrfs_header_nritems(c)) {
5648                         int ret;
5649                         int orig_lowest;
5650                         struct btrfs_key cur_key;
5651                         if (level + 1 >= BTRFS_MAX_LEVEL ||
5652                             !path->nodes[level + 1])
5653                                 return 1;
5654
5655                         if (path->locks[level + 1]) {
5656                                 level++;
5657                                 continue;
5658                         }
5659
5660                         slot = btrfs_header_nritems(c) - 1;
5661                         if (level == 0)
5662                                 btrfs_item_key_to_cpu(c, &cur_key, slot);
5663                         else
5664                                 btrfs_node_key_to_cpu(c, &cur_key, slot);
5665
5666                         orig_lowest = path->lowest_level;
5667                         btrfs_release_path(path);
5668                         path->lowest_level = level;
5669                         ret = btrfs_search_slot(NULL, root, &cur_key, path,
5670                                                 0, 0);
5671                         path->lowest_level = orig_lowest;
5672                         if (ret < 0)
5673                                 return ret;
5674
5675                         c = path->nodes[level];
5676                         slot = path->slots[level];
5677                         if (ret == 0)
5678                                 slot++;
5679                         goto next;
5680                 }
5681
5682                 if (level == 0)
5683                         btrfs_item_key_to_cpu(c, key, slot);
5684                 else {
5685                         u64 gen = btrfs_node_ptr_generation(c, slot);
5686
5687                         if (gen < min_trans) {
5688                                 slot++;
5689                                 goto next;
5690                         }
5691                         btrfs_node_key_to_cpu(c, key, slot);
5692                 }
5693                 return 0;
5694         }
5695         return 1;
5696 }
5697
5698 /*
5699  * search the tree again to find a leaf with greater keys
5700  * returns 0 if it found something or 1 if there are no greater leaves.
5701  * returns < 0 on io errors.
5702  */
5703 int btrfs_next_leaf(struct btrfs_root *root, struct btrfs_path *path)
5704 {
5705         return btrfs_next_old_leaf(root, path, 0);
5706 }
5707
5708 int btrfs_next_old_leaf(struct btrfs_root *root, struct btrfs_path *path,
5709                         u64 time_seq)
5710 {
5711         int slot;
5712         int level;
5713         struct extent_buffer *c;
5714         struct extent_buffer *next;
5715         struct btrfs_key key;
5716         u32 nritems;
5717         int ret;
5718         int old_spinning = path->leave_spinning;
5719         int next_rw_lock = 0;
5720
5721         nritems = btrfs_header_nritems(path->nodes[0]);
5722         if (nritems == 0)
5723                 return 1;
5724
5725         btrfs_item_key_to_cpu(path->nodes[0], &key, nritems - 1);
5726 again:
5727         level = 1;
5728         next = NULL;
5729         next_rw_lock = 0;
5730         btrfs_release_path(path);
5731
5732         path->keep_locks = 1;
5733         path->leave_spinning = 1;
5734
5735         if (time_seq)
5736                 ret = btrfs_search_old_slot(root, &key, path, time_seq);
5737         else
5738                 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
5739         path->keep_locks = 0;
5740
5741         if (ret < 0)
5742                 return ret;
5743
5744         nritems = btrfs_header_nritems(path->nodes[0]);
5745         /*
5746          * by releasing the path above we dropped all our locks.  A balance
5747          * could have added more items next to the key that used to be
5748          * at the very end of the block.  So, check again here and
5749          * advance the path if there are now more items available.
5750          */
5751         if (nritems > 0 && path->slots[0] < nritems - 1) {
5752                 if (ret == 0)
5753                         path->slots[0]++;
5754                 ret = 0;
5755                 goto done;
5756         }
5757         /*
5758          * So the above check misses one case:
5759          * - after releasing the path above, someone has removed the item that
5760          *   used to be at the very end of the block, and balance between leafs
5761          *   gets another one with bigger key.offset to replace it.
5762          *
5763          * This one should be returned as well, or we can get leaf corruption
5764          * later(esp. in __btrfs_drop_extents()).
5765          *
5766          * And a bit more explanation about this check,
5767          * with ret > 0, the key isn't found, the path points to the slot
5768          * where it should be inserted, so the path->slots[0] item must be the
5769          * bigger one.
5770          */
5771         if (nritems > 0 && ret > 0 && path->slots[0] == nritems - 1) {
5772                 ret = 0;
5773                 goto done;
5774         }
5775
5776         while (level < BTRFS_MAX_LEVEL) {
5777                 if (!path->nodes[level]) {
5778                         ret = 1;
5779                         goto done;
5780                 }
5781
5782                 slot = path->slots[level] + 1;
5783                 c = path->nodes[level];
5784                 if (slot >= btrfs_header_nritems(c)) {
5785                         level++;
5786                         if (level == BTRFS_MAX_LEVEL) {
5787                                 ret = 1;
5788                                 goto done;
5789                         }
5790                         continue;
5791                 }
5792
5793                 if (next) {
5794                         btrfs_tree_unlock_rw(next, next_rw_lock);
5795                         free_extent_buffer(next);
5796                 }
5797
5798                 next = c;
5799                 next_rw_lock = path->locks[level];
5800                 ret = read_block_for_search(NULL, root, path, &next, level,
5801                                             slot, &key, 0);
5802                 if (ret == -EAGAIN)
5803                         goto again;
5804
5805                 if (ret < 0) {
5806                         btrfs_release_path(path);
5807                         goto done;
5808                 }
5809
5810                 if (!path->skip_locking) {
5811                         ret = btrfs_try_tree_read_lock(next);
5812                         if (!ret && time_seq) {
5813                                 /*
5814                                  * If we don't get the lock, we may be racing
5815                                  * with push_leaf_left, holding that lock while
5816                                  * itself waiting for the leaf we've currently
5817                                  * locked. To solve this situation, we give up
5818                                  * on our lock and cycle.
5819                                  */
5820                                 free_extent_buffer(next);
5821                                 btrfs_release_path(path);
5822                                 cond_resched();
5823                                 goto again;
5824                         }
5825                         if (!ret) {
5826                                 btrfs_set_path_blocking(path);
5827                                 btrfs_tree_read_lock(next);
5828                                 btrfs_clear_path_blocking(path, next,
5829                                                           BTRFS_READ_LOCK);
5830                         }
5831                         next_rw_lock = BTRFS_READ_LOCK;
5832                 }
5833                 break;
5834         }
5835         path->slots[level] = slot;
5836         while (1) {
5837                 level--;
5838                 c = path->nodes[level];
5839                 if (path->locks[level])
5840                         btrfs_tree_unlock_rw(c, path->locks[level]);
5841
5842                 free_extent_buffer(c);
5843                 path->nodes[level] = next;
5844                 path->slots[level] = 0;
5845                 if (!path->skip_locking)
5846                         path->locks[level] = next_rw_lock;
5847                 if (!level)
5848                         break;
5849
5850                 ret = read_block_for_search(NULL, root, path, &next, level,
5851                                             0, &key, 0);
5852                 if (ret == -EAGAIN)
5853                         goto again;
5854
5855                 if (ret < 0) {
5856                         btrfs_release_path(path);
5857                         goto done;
5858                 }
5859
5860                 if (!path->skip_locking) {
5861                         ret = btrfs_try_tree_read_lock(next);
5862                         if (!ret) {
5863                                 btrfs_set_path_blocking(path);
5864                                 btrfs_tree_read_lock(next);
5865                                 btrfs_clear_path_blocking(path, next,
5866                                                           BTRFS_READ_LOCK);
5867                         }
5868                         next_rw_lock = BTRFS_READ_LOCK;
5869                 }
5870         }
5871         ret = 0;
5872 done:
5873         unlock_up(path, 0, 1, 0, NULL);
5874         path->leave_spinning = old_spinning;
5875         if (!old_spinning)
5876                 btrfs_set_path_blocking(path);
5877
5878         return ret;
5879 }
5880
5881 /*
5882  * this uses btrfs_prev_leaf to walk backwards in the tree, and keeps
5883  * searching until it gets past min_objectid or finds an item of 'type'
5884  *
5885  * returns 0 if something is found, 1 if nothing was found and < 0 on error
5886  */
5887 int btrfs_previous_item(struct btrfs_root *root,
5888                         struct btrfs_path *path, u64 min_objectid,
5889                         int type)
5890 {
5891         struct btrfs_key found_key;
5892         struct extent_buffer *leaf;
5893         u32 nritems;
5894         int ret;
5895
5896         while (1) {
5897                 if (path->slots[0] == 0) {
5898                         btrfs_set_path_blocking(path);
5899                         ret = btrfs_prev_leaf(root, path);
5900                         if (ret != 0)
5901                                 return ret;
5902                 } else {
5903                         path->slots[0]--;
5904                 }
5905                 leaf = path->nodes[0];
5906                 nritems = btrfs_header_nritems(leaf);
5907                 if (nritems == 0)
5908                         return 1;
5909                 if (path->slots[0] == nritems)
5910                         path->slots[0]--;
5911
5912                 btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
5913                 if (found_key.objectid < min_objectid)
5914                         break;
5915                 if (found_key.type == type)
5916                         return 0;
5917                 if (found_key.objectid == min_objectid &&
5918                     found_key.type < type)
5919                         break;
5920         }
5921         return 1;
5922 }
5923
5924 /*
5925  * search in extent tree to find a previous Metadata/Data extent item with
5926  * min objecitd.
5927  *
5928  * returns 0 if something is found, 1 if nothing was found and < 0 on error
5929  */
5930 int btrfs_previous_extent_item(struct btrfs_root *root,
5931                         struct btrfs_path *path, u64 min_objectid)
5932 {
5933         struct btrfs_key found_key;
5934         struct extent_buffer *leaf;
5935         u32 nritems;
5936         int ret;
5937
5938         while (1) {
5939                 if (path->slots[0] == 0) {
5940                         btrfs_set_path_blocking(path);
5941                         ret = btrfs_prev_leaf(root, path);
5942                         if (ret != 0)
5943                                 return ret;
5944                 } else {
5945                         path->slots[0]--;
5946                 }
5947                 leaf = path->nodes[0];
5948                 nritems = btrfs_header_nritems(leaf);
5949                 if (nritems == 0)
5950                         return 1;
5951                 if (path->slots[0] == nritems)
5952                         path->slots[0]--;
5953
5954                 btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
5955                 if (found_key.objectid < min_objectid)
5956                         break;
5957                 if (found_key.type == BTRFS_EXTENT_ITEM_KEY ||
5958                     found_key.type == BTRFS_METADATA_ITEM_KEY)
5959                         return 0;
5960                 if (found_key.objectid == min_objectid &&
5961                     found_key.type < BTRFS_EXTENT_ITEM_KEY)
5962                         break;
5963         }
5964         return 1;
5965 }