fs/btrfs/extent-tree.c

   1 // SPDX-License-Identifier: GPL-2.0
   2 /*
   3  * Copyright (C) 2007 Oracle.  All rights reserved.
   4  */
   5
   6 #include <linux/sched.h>
   7 #include <linux/sched/signal.h>
   8 #include <linux/pagemap.h>
   9 #include <linux/writeback.h>
  10 #include <linux/blkdev.h>
  11 #include <linux/sort.h>
  12 #include <linux/rcupdate.h>
  13 #include <linux/kthread.h>
  14 #include <linux/slab.h>
  15 #include <linux/ratelimit.h>
  16 #include <linux/percpu_counter.h>
  17 #include <linux/lockdep.h>
  18 #include <linux/crc32c.h>
  19 #include "misc.h"
  20 #include "tree-log.h"
  21 #include "disk-io.h"
  22 #include "print-tree.h"
  23 #include "volumes.h"
  24 #include "raid56.h"
  25 #include "locking.h"
  26 #include "free-space-cache.h"
  27 #include "free-space-tree.h"
  28 #include "sysfs.h"
  29 #include "qgroup.h"
  30 #include "ref-verify.h"
  31 #include "space-info.h"
  32 #include "block-rsv.h"
  33 #include "delalloc-space.h"
  34 #include "block-group.h"
  35 #include "discard.h"
  36
  37 #undef SCRAMBLE_DELAYED_REFS
  38
  39
  40 static int __btrfs_free_extent(struct btrfs_trans_handle *trans,
  41                                struct btrfs_delayed_ref_node *node, u64 parent,
  42                                u64 root_objectid, u64 owner_objectid,
  43                                u64 owner_offset, int refs_to_drop,
  44                                struct btrfs_delayed_extent_op *extra_op);
  45 static void __run_delayed_extent_op(struct btrfs_delayed_extent_op *extent_op,
  46                                     struct extent_buffer *leaf,
  47                                     struct btrfs_extent_item *ei);
  48 static int alloc_reserved_file_extent(struct btrfs_trans_handle *trans,
  49                                       u64 parent, u64 root_objectid,
  50                                       u64 flags, u64 owner, u64 offset,
  51                                       struct btrfs_key *ins, int ref_mod);
  52 static int alloc_reserved_tree_block(struct btrfs_trans_handle *trans,
  53                                      struct btrfs_delayed_ref_node *node,
  54                                      struct btrfs_delayed_extent_op *extent_op);
  55 static int find_next_key(struct btrfs_path *path, int level,
  56                          struct btrfs_key *key);
  57
  58 static int block_group_bits(struct btrfs_block_group *cache, u64 bits)
  59 {
  60         return (cache->flags & bits) == bits;
  61 }
  62
  63 int btrfs_add_excluded_extent(struct btrfs_fs_info *fs_info,
  64                               u64 start, u64 num_bytes)
  65 {
  66         u64 end = start + num_bytes - 1;
  67         set_extent_bits(&fs_info->freed_extents[0],
  68                         start, end, EXTENT_UPTODATE);
  69         set_extent_bits(&fs_info->freed_extents[1],
  70                         start, end, EXTENT_UPTODATE);
  71         return 0;
  72 }
  73
  74 void btrfs_free_excluded_extents(struct btrfs_block_group *cache)
  75 {
  76         struct btrfs_fs_info *fs_info = cache->fs_info;
  77         u64 start, end;
  78
  79         start = cache->start;
  80         end = start + cache->length - 1;
  81
  82         clear_extent_bits(&fs_info->freed_extents[0],
  83                           start, end, EXTENT_UPTODATE);
  84         clear_extent_bits(&fs_info->freed_extents[1],
  85                           start, end, EXTENT_UPTODATE);
  86 }
  87
  88 static u64 generic_ref_to_space_flags(struct btrfs_ref *ref)
  89 {
  90         if (ref->type == BTRFS_REF_METADATA) {
  91                 if (ref->tree_ref.root == BTRFS_CHUNK_TREE_OBJECTID)
  92                         return BTRFS_BLOCK_GROUP_SYSTEM;
  93                 else
  94                         return BTRFS_BLOCK_GROUP_METADATA;
  95         }
  96         return BTRFS_BLOCK_GROUP_DATA;
  97 }
  98
  99 static void add_pinned_bytes(struct btrfs_fs_info *fs_info,
 100                              struct btrfs_ref *ref)
 101 {
 102         struct btrfs_space_info *space_info;
 103         u64 flags = generic_ref_to_space_flags(ref);
 104
 105         space_info = btrfs_find_space_info(fs_info, flags);
 106         ASSERT(space_info);
 107         percpu_counter_add_batch(&space_info->total_bytes_pinned, ref->len,
 108                     BTRFS_TOTAL_BYTES_PINNED_BATCH);
 109 }
 110
 111 static void sub_pinned_bytes(struct btrfs_fs_info *fs_info,
 112                              struct btrfs_ref *ref)
 113 {
 114         struct btrfs_space_info *space_info;
 115         u64 flags = generic_ref_to_space_flags(ref);
 116
 117         space_info = btrfs_find_space_info(fs_info, flags);
 118         ASSERT(space_info);
 119         percpu_counter_add_batch(&space_info->total_bytes_pinned, -ref->len,
 120                     BTRFS_TOTAL_BYTES_PINNED_BATCH);
 121 }
 122
 123 /* simple helper to search for an existing data extent at a given offset */
 124 int btrfs_lookup_data_extent(struct btrfs_fs_info *fs_info, u64 start, u64 len)
 125 {
 126         int ret;
 127         struct btrfs_key key;
 128         struct btrfs_path *path;
 129
 130         path = btrfs_alloc_path();
 131         if (!path)
 132                 return -ENOMEM;
 133
 134         key.objectid = start;
 135         key.offset = len;
 136         key.type = BTRFS_EXTENT_ITEM_KEY;
 137         ret = btrfs_search_slot(NULL, fs_info->extent_root, &key, path, 0, 0);
 138         btrfs_free_path(path);
 139         return ret;
 140 }
 141
 142 /*
 143  * helper function to lookup reference count and flags of a tree block.
 144  *
 145  * the head node for delayed ref is used to store the sum of all the
 146  * reference count modifications queued up in the rbtree. the head
 147  * node may also store the extent flags to set. This way you can check
 148  * to see what the reference count and extent flags would be if all of
 149  * the delayed refs are not processed.
 150  */
 151 int btrfs_lookup_extent_info(struct btrfs_trans_handle *trans,
 152                              struct btrfs_fs_info *fs_info, u64 bytenr,
 153                              u64 offset, int metadata, u64 *refs, u64 *flags)
 154 {
 155         struct btrfs_delayed_ref_head *head;
 156         struct btrfs_delayed_ref_root *delayed_refs;
 157         struct btrfs_path *path;
 158         struct btrfs_extent_item *ei;
 159         struct extent_buffer *leaf;
 160         struct btrfs_key key;
 161         u32 item_size;
 162         u64 num_refs;
 163         u64 extent_flags;
 164         int ret;
 165
 166         /*
 167          * If we don't have skinny metadata, don't bother doing anything
 168          * different
 169          */
 170         if (metadata && !btrfs_fs_incompat(fs_info, SKINNY_METADATA)) {
 171                 offset = fs_info->nodesize;
 172                 metadata = 0;
 173         }
 174
 175         path = btrfs_alloc_path();
 176         if (!path)
 177                 return -ENOMEM;
 178
 179         if (!trans) {
 180                 path->skip_locking = 1;
 181                 path->search_commit_root = 1;
 182         }
 183
 184 search_again:
 185         key.objectid = bytenr;
 186         key.offset = offset;
 187         if (metadata)
 188                 key.type = BTRFS_METADATA_ITEM_KEY;
 189         else
 190                 key.type = BTRFS_EXTENT_ITEM_KEY;
 191
 192         ret = btrfs_search_slot(trans, fs_info->extent_root, &key, path, 0, 0);
 193         if (ret < 0)
 194                 goto out_free;
 195
 196         if (ret > 0 && metadata && key.type == BTRFS_METADATA_ITEM_KEY) {
 197                 if (path->slots[0]) {
 198                         path->slots[0]--;
 199                         btrfs_item_key_to_cpu(path->nodes[0], &key,
 200                                               path->slots[0]);
 201                         if (key.objectid == bytenr &&
 202                             key.type == BTRFS_EXTENT_ITEM_KEY &&
 203                             key.offset == fs_info->nodesize)
 204                                 ret = 0;
 205                 }
 206         }
 207
 208         if (ret == 0) {
 209                 leaf = path->nodes[0];
 210                 item_size = btrfs_item_size_nr(leaf, path->slots[0]);
 211                 if (item_size >= sizeof(*ei)) {
 212                         ei = btrfs_item_ptr(leaf, path->slots[0],
 213                                             struct btrfs_extent_item);
 214                         num_refs = btrfs_extent_refs(leaf, ei);
 215                         extent_flags = btrfs_extent_flags(leaf, ei);
 216                 } else {
 217                         ret = -EINVAL;
 218                         btrfs_print_v0_err(fs_info);
 219                         if (trans)
 220                                 btrfs_abort_transaction(trans, ret);
 221                         else
 222                                 btrfs_handle_fs_error(fs_info, ret, NULL);
 223
 224                         goto out_free;
 225                 }
 226
 227                 BUG_ON(num_refs == 0);
 228         } else {
 229                 num_refs = 0;
 230                 extent_flags = 0;
 231                 ret = 0;
 232         }
 233
 234         if (!trans)
 235                 goto out;
 236
 237         delayed_refs = &trans->transaction->delayed_refs;
 238         spin_lock(&delayed_refs->lock);
 239         head = btrfs_find_delayed_ref_head(delayed_refs, bytenr);
 240         if (head) {
 241                 if (!mutex_trylock(&head->mutex)) {
 242                         refcount_inc(&head->refs);
 243                         spin_unlock(&delayed_refs->lock);
 244
 245                         btrfs_release_path(path);
 246
 247                         /*
 248                          * Mutex was contended, block until it's released and try
 249                          * again
 250                          */
 251                         mutex_lock(&head->mutex);
 252                         mutex_unlock(&head->mutex);
 253                         btrfs_put_delayed_ref_head(head);
 254                         goto search_again;
 255                 }
 256                 spin_lock(&head->lock);
 257                 if (head->extent_op && head->extent_op->update_flags)
 258                         extent_flags |= head->extent_op->flags_to_set;
 259                 else
 260                         BUG_ON(num_refs == 0);
 261
 262                 num_refs += head->ref_mod;
 263                 spin_unlock(&head->lock);
 264                 mutex_unlock(&head->mutex);
 265         }
 266         spin_unlock(&delayed_refs->lock);
 267 out:
 268         WARN_ON(num_refs == 0);
 269         if (refs)
 270                 *refs = num_refs;
 271         if (flags)
 272                 *flags = extent_flags;
 273 out_free:
 274         btrfs_free_path(path);
 275         return ret;
 276 }
 277
 278 /*
 279  * Back reference rules.  Back refs have three main goals:
 280  *
 281  * 1) differentiate between all holders of references to an extent so that
 282  *    when a reference is dropped we can make sure it was a valid reference
 283  *    before freeing the extent.
 284  *
 285  * 2) Provide enough information to quickly find the holders of an extent
 286  *    if we notice a given block is corrupted or bad.
 287  *
 288  * 3) Make it easy to migrate blocks for FS shrinking or storage pool
 289  *    maintenance.  This is actually the same as #2, but with a slightly
 290  *    different use case.
 291  *
 292  * There are two kinds of back refs. The implicit back refs is optimized
 293  * for pointers in non-shared tree blocks. For a given pointer in a block,
 294  * back refs of this kind provide information about the block's owner tree
 295  * and the pointer's key. These information allow us to find the block by
 296  * b-tree searching. The full back refs is for pointers in tree blocks not
 297  * referenced by their owner trees. The location of tree block is recorded
 298  * in the back refs. Actually the full back refs is generic, and can be
 299  * used in all cases the implicit back refs is used. The major shortcoming
 300  * of the full back refs is its overhead. Every time a tree block gets
 301  * COWed, we have to update back refs entry for all pointers in it.
 302  *
 303  * For a newly allocated tree block, we use implicit back refs for
 304  * pointers in it. This means most tree related operations only involve
 305  * implicit back refs. For a tree block created in old transaction, the
 306  * only way to drop a reference to it is COW it. So we can detect the
 307  * event that tree block loses its owner tree's reference and do the
 308  * back refs conversion.
 309  *
 310  * When a tree block is COWed through a tree, there are four cases:
 311  *
 312  * The reference count of the block is one and the tree is the block's
 313  * owner tree. Nothing to do in this case.
 314  *
 315  * The reference count of the block is one and the tree is not the
 316  * block's owner tree. In this case, full back refs is used for pointers
 317  * in the block. Remove these full back refs, add implicit back refs for
 318  * every pointers in the new block.
 319  *
 320  * The reference count of the block is greater than one and the tree is
 321  * the block's owner tree. In this case, implicit back refs is used for
 322  * pointers in the block. Add full back refs for every pointers in the
 323  * block, increase lower level extents' reference counts. The original
 324  * implicit back refs are entailed to the new block.
 325  *
 326  * The reference count of the block is greater than one and the tree is
 327  * not the block's owner tree. Add implicit back refs for every pointer in
 328  * the new block, increase lower level extents' reference count.
 329  *
 330  * Back Reference Key composing:
 331  *
 332  * The key objectid corresponds to the first byte in the extent,
 333  * The key type is used to differentiate between types of back refs.
 334  * There are different meanings of the key offset for different types
 335  * of back refs.
 336  *
 337  * File extents can be referenced by:
 338  *
 339  * - multiple snapshots, subvolumes, or different generations in one subvol
 340  * - different files inside a single subvolume
 341  * - different offsets inside a file (bookend extents in file.c)
 342  *
 343  * The extent ref structure for the implicit back refs has fields for:
 344  *
 345  * - Objectid of the subvolume root
 346  * - objectid of the file holding the reference
 347  * - original offset in the file
 348  * - how many bookend extents
 349  *
 350  * The key offset for the implicit back refs is hash of the first
 351  * three fields.
 352  *
 353  * The extent ref structure for the full back refs has field for:
 354  *
 355  * - number of pointers in the tree leaf
 356  *
 357  * The key offset for the implicit back refs is the first byte of
 358  * the tree leaf
 359  *
 360  * When a file extent is allocated, The implicit back refs is used.
 361  * the fields are filled in:
 362  *
 363  *     (root_key.objectid, inode objectid, offset in file, 1)
 364  *
 365  * When a file extent is removed file truncation, we find the
 366  * corresponding implicit back refs and check the following fields:
 367  *
 368  *     (btrfs_header_owner(leaf), inode objectid, offset in file)
 369  *
 370  * Btree extents can be referenced by:
 371  *
 372  * - Different subvolumes
 373  *
 374  * Both the implicit back refs and the full back refs for tree blocks
 375  * only consist of key. The key offset for the implicit back refs is
 376  * objectid of block's owner tree. The key offset for the full back refs
 377  * is the first byte of parent block.
 378  *
 379  * When implicit back refs is used, information about the lowest key and
 380  * level of the tree block are required. These information are stored in
 381  * tree block info structure.
 382  */
 383
 384 /*
 385  * is_data == BTRFS_REF_TYPE_BLOCK, tree block type is required,
 386  * is_data == BTRFS_REF_TYPE_DATA, data type is requiried,
 387  * is_data == BTRFS_REF_TYPE_ANY, either type is OK.
 388  */
 389 int btrfs_get_extent_inline_ref_type(const struct extent_buffer *eb,
 390                                      struct btrfs_extent_inline_ref *iref,
 391                                      enum btrfs_inline_ref_type is_data)
 392 {
 393         int type = btrfs_extent_inline_ref_type(eb, iref);
 394         u64 offset = btrfs_extent_inline_ref_offset(eb, iref);
 395
 396         if (type == BTRFS_TREE_BLOCK_REF_KEY ||
 397             type == BTRFS_SHARED_BLOCK_REF_KEY ||
 398             type == BTRFS_SHARED_DATA_REF_KEY ||
 399             type == BTRFS_EXTENT_DATA_REF_KEY) {
 400                 if (is_data == BTRFS_REF_TYPE_BLOCK) {
 401                         if (type == BTRFS_TREE_BLOCK_REF_KEY)
 402                                 return type;
 403                         if (type == BTRFS_SHARED_BLOCK_REF_KEY) {
 404                                 ASSERT(eb->fs_info);
 405                                 /*
 406                                  * Every shared one has parent tree
 407                                  * block, which must be aligned to
 408                                  * nodesize.
 409                                  */
 410                                 if (offset &&
 411                                     IS_ALIGNED(offset, eb->fs_info->nodesize))
 412                                         return type;
 413                         }
 414                 } else if (is_data == BTRFS_REF_TYPE_DATA) {
 415                         if (type == BTRFS_EXTENT_DATA_REF_KEY)
 416                                 return type;
 417                         if (type == BTRFS_SHARED_DATA_REF_KEY) {
 418                                 ASSERT(eb->fs_info);
 419                                 /*
 420                                  * Every shared one has parent tree
 421                                  * block, which must be aligned to
 422                                  * nodesize.
 423                                  */
 424                                 if (offset &&
 425                                     IS_ALIGNED(offset, eb->fs_info->nodesize))
 426                                         return type;
 427                         }
 428                 } else {
 429                         ASSERT(is_data == BTRFS_REF_TYPE_ANY);
 430                         return type;
 431                 }
 432         }
 433
 434         btrfs_print_leaf((struct extent_buffer *)eb);
 435         btrfs_err(eb->fs_info, "eb %llu invalid extent inline ref type %d",
 436                   eb->start, type);
 437         WARN_ON(1);
 438
 439         return BTRFS_REF_TYPE_INVALID;
 440 }
 441
 442 u64 hash_extent_data_ref(u64 root_objectid, u64 owner, u64 offset)
 443 {
 444         u32 high_crc = ~(u32)0;
 445         u32 low_crc = ~(u32)0;
 446         __le64 lenum;
 447
 448         lenum = cpu_to_le64(root_objectid);
 449         high_crc = btrfs_crc32c(high_crc, &lenum, sizeof(lenum));
 450         lenum = cpu_to_le64(owner);
 451         low_crc = btrfs_crc32c(low_crc, &lenum, sizeof(lenum));
 452         lenum = cpu_to_le64(offset);
 453         low_crc = btrfs_crc32c(low_crc, &lenum, sizeof(lenum));
 454
 455         return ((u64)high_crc << 31) ^ (u64)low_crc;
 456 }
 457
 458 static u64 hash_extent_data_ref_item(struct extent_buffer *leaf,
 459                                      struct btrfs_extent_data_ref *ref)
 460 {
 461         return hash_extent_data_ref(btrfs_extent_data_ref_root(leaf, ref),
 462                                     btrfs_extent_data_ref_objectid(leaf, ref),
 463                                     btrfs_extent_data_ref_offset(leaf, ref));
 464 }
 465
 466 static int match_extent_data_ref(struct extent_buffer *leaf,
 467                                  struct btrfs_extent_data_ref *ref,
 468                                  u64 root_objectid, u64 owner, u64 offset)
 469 {
 470         if (btrfs_extent_data_ref_root(leaf, ref) != root_objectid ||
 471             btrfs_extent_data_ref_objectid(leaf, ref) != owner ||
 472             btrfs_extent_data_ref_offset(leaf, ref) != offset)
 473                 return 0;
 474         return 1;
 475 }
 476
 477 static noinline int lookup_extent_data_ref(struct btrfs_trans_handle *trans,
 478                                            struct btrfs_path *path,
 479                                            u64 bytenr, u64 parent,
 480                                            u64 root_objectid,
 481                                            u64 owner, u64 offset)
 482 {
 483         struct btrfs_root *root = trans->fs_info->extent_root;
 484         struct btrfs_key key;
 485         struct btrfs_extent_data_ref *ref;
 486         struct extent_buffer *leaf;
 487         u32 nritems;
 488         int ret;
 489         int recow;
 490         int err = -ENOENT;
 491
 492         key.objectid = bytenr;
 493         if (parent) {
 494                 key.type = BTRFS_SHARED_DATA_REF_KEY;
 495                 key.offset = parent;
 496         } else {
 497                 key.type = BTRFS_EXTENT_DATA_REF_KEY;
 498                 key.offset = hash_extent_data_ref(root_objectid,
 499                                                   owner, offset);
 500         }
 501 again:
 502         recow = 0;
 503         ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
 504         if (ret < 0) {
 505                 err = ret;
 506                 goto fail;
 507         }
 508
 509         if (parent) {
 510                 if (!ret)
 511                         return 0;
 512                 goto fail;
 513         }
 514
 515         leaf = path->nodes[0];
 516         nritems = btrfs_header_nritems(leaf);
 517         while (1) {
 518                 if (path->slots[0] >= nritems) {
 519                         ret = btrfs_next_leaf(root, path);
 520                         if (ret < 0)
 521                                 err = ret;
 522                         if (ret)
 523                                 goto fail;
 524
 525                         leaf = path->nodes[0];
 526                         nritems = btrfs_header_nritems(leaf);
 527                         recow = 1;
 528                 }
 529
 530                 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
 531                 if (key.objectid != bytenr ||
 532                     key.type != BTRFS_EXTENT_DATA_REF_KEY)
 533                         goto fail;
 534
 535                 ref = btrfs_item_ptr(leaf, path->slots[0],
 536                                      struct btrfs_extent_data_ref);
 537
 538                 if (match_extent_data_ref(leaf, ref, root_objectid,
 539                                           owner, offset)) {
 540                         if (recow) {
 541                                 btrfs_release_path(path);
 542                                 goto again;
 543                         }
 544                         err = 0;
 545                         break;
 546                 }
 547                 path->slots[0]++;
 548         }
 549 fail:
 550         return err;
 551 }
 552
 553 static noinline int insert_extent_data_ref(struct btrfs_trans_handle *trans,
 554                                            struct btrfs_path *path,
 555                                            u64 bytenr, u64 parent,
 556                                            u64 root_objectid, u64 owner,
 557                                            u64 offset, int refs_to_add)
 558 {
 559         struct btrfs_root *root = trans->fs_info->extent_root;
 560         struct btrfs_key key;
 561         struct extent_buffer *leaf;
 562         u32 size;
 563         u32 num_refs;
 564         int ret;
 565
 566         key.objectid = bytenr;
 567         if (parent) {
 568                 key.type = BTRFS_SHARED_DATA_REF_KEY;
 569                 key.offset = parent;
 570                 size = sizeof(struct btrfs_shared_data_ref);
 571         } else {
 572                 key.type = BTRFS_EXTENT_DATA_REF_KEY;
 573                 key.offset = hash_extent_data_ref(root_objectid,
 574                                                   owner, offset);
 575                 size = sizeof(struct btrfs_extent_data_ref);
 576         }
 577
 578         ret = btrfs_insert_empty_item(trans, root, path, &key, size);
 579         if (ret && ret != -EEXIST)
 580                 goto fail;
 581
 582         leaf = path->nodes[0];
 583         if (parent) {
 584                 struct btrfs_shared_data_ref *ref;
 585                 ref = btrfs_item_ptr(leaf, path->slots[0],
 586                                      struct btrfs_shared_data_ref);
 587                 if (ret == 0) {
 588                         btrfs_set_shared_data_ref_count(leaf, ref, refs_to_add);
 589                 } else {
 590                         num_refs = btrfs_shared_data_ref_count(leaf, ref);
 591                         num_refs += refs_to_add;
 592                         btrfs_set_shared_data_ref_count(leaf, ref, num_refs);
 593                 }
 594         } else {
 595                 struct btrfs_extent_data_ref *ref;
 596                 while (ret == -EEXIST) {
 597                         ref = btrfs_item_ptr(leaf, path->slots[0],
 598                                              struct btrfs_extent_data_ref);
 599                         if (match_extent_data_ref(leaf, ref, root_objectid,
 600                                                   owner, offset))
 601                                 break;
 602                         btrfs_release_path(path);
 603                         key.offset++;
 604                         ret = btrfs_insert_empty_item(trans, root, path, &key,
 605                                                       size);
 606                         if (ret && ret != -EEXIST)
 607                                 goto fail;
 608
 609                         leaf = path->nodes[0];
 610                 }
 611                 ref = btrfs_item_ptr(leaf, path->slots[0],
 612                                      struct btrfs_extent_data_ref);
 613                 if (ret == 0) {
 614                         btrfs_set_extent_data_ref_root(leaf, ref,
 615                                                        root_objectid);
 616                         btrfs_set_extent_data_ref_objectid(leaf, ref, owner);
 617                         btrfs_set_extent_data_ref_offset(leaf, ref, offset);
 618                         btrfs_set_extent_data_ref_count(leaf, ref, refs_to_add);
 619                 } else {
 620                         num_refs = btrfs_extent_data_ref_count(leaf, ref);
 621                         num_refs += refs_to_add;
 622                         btrfs_set_extent_data_ref_count(leaf, ref, num_refs);
 623                 }
 624         }
 625         btrfs_mark_buffer_dirty(leaf);
 626         ret = 0;
 627 fail:
 628         btrfs_release_path(path);
 629         return ret;
 630 }
 631
 632 static noinline int remove_extent_data_ref(struct btrfs_trans_handle *trans,
 633                                            struct btrfs_path *path,
 634                                            int refs_to_drop, int *last_ref)
 635 {
 636         struct btrfs_key key;
 637         struct btrfs_extent_data_ref *ref1 = NULL;
 638         struct btrfs_shared_data_ref *ref2 = NULL;
 639         struct extent_buffer *leaf;
 640         u32 num_refs = 0;
 641         int ret = 0;
 642
 643         leaf = path->nodes[0];
 644         btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
 645
 646         if (key.type == BTRFS_EXTENT_DATA_REF_KEY) {
 647                 ref1 = btrfs_item_ptr(leaf, path->slots[0],
 648                                       struct btrfs_extent_data_ref);
 649                 num_refs = btrfs_extent_data_ref_count(leaf, ref1);
 650         } else if (key.type == BTRFS_SHARED_DATA_REF_KEY) {
 651                 ref2 = btrfs_item_ptr(leaf, path->slots[0],
 652                                       struct btrfs_shared_data_ref);
 653                 num_refs = btrfs_shared_data_ref_count(leaf, ref2);
 654         } else if (unlikely(key.type == BTRFS_EXTENT_REF_V0_KEY)) {
 655                 btrfs_print_v0_err(trans->fs_info);
 656                 btrfs_abort_transaction(trans, -EINVAL);
 657                 return -EINVAL;
 658         } else {
 659                 BUG();
 660         }
 661
 662         BUG_ON(num_refs < refs_to_drop);
 663         num_refs -= refs_to_drop;
 664
 665         if (num_refs == 0) {
 666                 ret = btrfs_del_item(trans, trans->fs_info->extent_root, path);
 667                 *last_ref = 1;
 668         } else {
 669                 if (key.type == BTRFS_EXTENT_DATA_REF_KEY)
 670                         btrfs_set_extent_data_ref_count(leaf, ref1, num_refs);
 671                 else if (key.type == BTRFS_SHARED_DATA_REF_KEY)
 672                         btrfs_set_shared_data_ref_count(leaf, ref2, num_refs);
 673                 btrfs_mark_buffer_dirty(leaf);
 674         }
 675         return ret;
 676 }
 677
 678 static noinline u32 extent_data_ref_count(struct btrfs_path *path,
 679                                           struct btrfs_extent_inline_ref *iref)
 680 {
 681         struct btrfs_key key;
 682         struct extent_buffer *leaf;
 683         struct btrfs_extent_data_ref *ref1;
 684         struct btrfs_shared_data_ref *ref2;
 685         u32 num_refs = 0;
 686         int type;
 687
 688         leaf = path->nodes[0];
 689         btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
 690
 691         BUG_ON(key.type == BTRFS_EXTENT_REF_V0_KEY);
 692         if (iref) {
 693                 /*
 694                  * If type is invalid, we should have bailed out earlier than
 695                  * this call.
 696                  */
 697                 type = btrfs_get_extent_inline_ref_type(leaf, iref, BTRFS_REF_TYPE_DATA);
 698                 ASSERT(type != BTRFS_REF_TYPE_INVALID);
 699                 if (type == BTRFS_EXTENT_DATA_REF_KEY) {
 700                         ref1 = (struct btrfs_extent_data_ref *)(&iref->offset);
 701                         num_refs = btrfs_extent_data_ref_count(leaf, ref1);
 702                 } else {
 703                         ref2 = (struct btrfs_shared_data_ref *)(iref + 1);
 704                         num_refs = btrfs_shared_data_ref_count(leaf, ref2);
 705                 }
 706         } else if (key.type == BTRFS_EXTENT_DATA_REF_KEY) {
 707                 ref1 = btrfs_item_ptr(leaf, path->slots[0],
 708                                       struct btrfs_extent_data_ref);
 709                 num_refs = btrfs_extent_data_ref_count(leaf, ref1);
 710         } else if (key.type == BTRFS_SHARED_DATA_REF_KEY) {
 711                 ref2 = btrfs_item_ptr(leaf, path->slots[0],
 712                                       struct btrfs_shared_data_ref);
 713                 num_refs = btrfs_shared_data_ref_count(leaf, ref2);
 714         } else {
 715                 WARN_ON(1);
 716         }
 717         return num_refs;
 718 }
 719
 720 static noinline int lookup_tree_block_ref(struct btrfs_trans_handle *trans,
 721                                           struct btrfs_path *path,
 722                                           u64 bytenr, u64 parent,
 723                                           u64 root_objectid)
 724 {
 725         struct btrfs_root *root = trans->fs_info->extent_root;
 726         struct btrfs_key key;
 727         int ret;
 728
 729         key.objectid = bytenr;
 730         if (parent) {
 731                 key.type = BTRFS_SHARED_BLOCK_REF_KEY;
 732                 key.offset = parent;
 733         } else {
 734                 key.type = BTRFS_TREE_BLOCK_REF_KEY;
 735                 key.offset = root_objectid;
 736         }
 737
 738         ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
 739         if (ret > 0)
 740                 ret = -ENOENT;
 741         return ret;
 742 }
 743
 744 static noinline int insert_tree_block_ref(struct btrfs_trans_handle *trans,
 745                                           struct btrfs_path *path,
 746                                           u64 bytenr, u64 parent,
 747                                           u64 root_objectid)
 748 {
 749         struct btrfs_key key;
 750         int ret;
 751
 752         key.objectid = bytenr;
 753         if (parent) {
 754                 key.type = BTRFS_SHARED_BLOCK_REF_KEY;
 755                 key.offset = parent;
 756         } else {
 757                 key.type = BTRFS_TREE_BLOCK_REF_KEY;
 758                 key.offset = root_objectid;
 759         }
 760
 761         ret = btrfs_insert_empty_item(trans, trans->fs_info->extent_root,
 762                                       path, &key, 0);
 763         btrfs_release_path(path);
 764         return ret;
 765 }
 766
 767 static inline int extent_ref_type(u64 parent, u64 owner)
 768 {
 769         int type;
 770         if (owner < BTRFS_FIRST_FREE_OBJECTID) {
 771                 if (parent > 0)
 772                         type = BTRFS_SHARED_BLOCK_REF_KEY;
 773                 else
 774                         type = BTRFS_TREE_BLOCK_REF_KEY;
 775         } else {
 776                 if (parent > 0)
 777                         type = BTRFS_SHARED_DATA_REF_KEY;
 778                 else
 779                         type = BTRFS_EXTENT_DATA_REF_KEY;
 780         }
 781         return type;
 782 }
 783
 784 static int find_next_key(struct btrfs_path *path, int level,
 785                          struct btrfs_key *key)
 786
 787 {
 788         for (; level < BTRFS_MAX_LEVEL; level++) {
 789                 if (!path->nodes[level])
 790                         break;
 791                 if (path->slots[level] + 1 >=
 792                     btrfs_header_nritems(path->nodes[level]))
 793                         continue;
 794                 if (level == 0)
 795                         btrfs_item_key_to_cpu(path->nodes[level], key,
 796                                               path->slots[level] + 1);
 797                 else
 798                         btrfs_node_key_to_cpu(path->nodes[level], key,
 799                                               path->slots[level] + 1);
 800                 return 0;
 801         }
 802         return 1;
 803 }
 804
 805 /*
 806  * look for inline back ref. if back ref is found, *ref_ret is set
 807  * to the address of inline back ref, and 0 is returned.
 808  *
 809  * if back ref isn't found, *ref_ret is set to the address where it
 810  * should be inserted, and -ENOENT is returned.
 811  *
 812  * if insert is true and there are too many inline back refs, the path
 813  * points to the extent item, and -EAGAIN is returned.
 814  *
 815  * NOTE: inline back refs are ordered in the same way that back ref
 816  *       items in the tree are ordered.
 817  */
 818 static noinline_for_stack
 819 int lookup_inline_extent_backref(struct btrfs_trans_handle *trans,
 820                                  struct btrfs_path *path,
 821                                  struct btrfs_extent_inline_ref **ref_ret,
 822                                  u64 bytenr, u64 num_bytes,
 823                                  u64 parent, u64 root_objectid,
 824                                  u64 owner, u64 offset, int insert)
 825 {
 826         struct btrfs_fs_info *fs_info = trans->fs_info;
 827         struct btrfs_root *root = fs_info->extent_root;
 828         struct btrfs_key key;
 829         struct extent_buffer *leaf;
 830         struct btrfs_extent_item *ei;
 831         struct btrfs_extent_inline_ref *iref;
 832         u64 flags;
 833         u64 item_size;
 834         unsigned long ptr;
 835         unsigned long end;
 836         int extra_size;
 837         int type;
 838         int want;
 839         int ret;
 840         int err = 0;
 841         bool skinny_metadata = btrfs_fs_incompat(fs_info, SKINNY_METADATA);
 842         int needed;
 843
 844         key.objectid = bytenr;
 845         key.type = BTRFS_EXTENT_ITEM_KEY;
 846         key.offset = num_bytes;
 847
 848         want = extent_ref_type(parent, owner);
 849         if (insert) {
 850                 extra_size = btrfs_extent_inline_ref_size(want);
 851                 path->keep_locks = 1;
 852         } else
 853                 extra_size = -1;
 854
 855         /*
 856          * Owner is our level, so we can just add one to get the level for the
 857          * block we are interested in.
 858          */
 859         if (skinny_metadata && owner < BTRFS_FIRST_FREE_OBJECTID) {
 860                 key.type = BTRFS_METADATA_ITEM_KEY;
 861                 key.offset = owner;
 862         }
 863
 864 again:
 865         ret = btrfs_search_slot(trans, root, &key, path, extra_size, 1);
 866         if (ret < 0) {
 867                 err = ret;
 868                 goto out;
 869         }
 870
 871         /*
 872          * We may be a newly converted file system which still has the old fat
 873          * extent entries for metadata, so try and see if we have one of those.
 874          */
 875         if (ret > 0 && skinny_metadata) {
 876                 skinny_metadata = false;
 877                 if (path->slots[0]) {
 878                         path->slots[0]--;
 879                         btrfs_item_key_to_cpu(path->nodes[0], &key,
 880                                               path->slots[0]);
 881                         if (key.objectid == bytenr &&
 882                             key.type == BTRFS_EXTENT_ITEM_KEY &&
 883                             key.offset == num_bytes)
 884                                 ret = 0;
 885                 }
 886                 if (ret) {
 887                         key.objectid = bytenr;
 888                         key.type = BTRFS_EXTENT_ITEM_KEY;
 889                         key.offset = num_bytes;
 890                         btrfs_release_path(path);
 891                         goto again;
 892                 }
 893         }
 894
 895         if (ret && !insert) {
 896                 err = -ENOENT;
 897                 goto out;
 898         } else if (WARN_ON(ret)) {
 899                 err = -EIO;
 900                 goto out;
 901         }
 902
 903         leaf = path->nodes[0];
 904         item_size = btrfs_item_size_nr(leaf, path->slots[0]);
 905         if (unlikely(item_size < sizeof(*ei))) {
 906                 err = -EINVAL;
 907                 btrfs_print_v0_err(fs_info);
 908                 btrfs_abort_transaction(trans, err);
 909                 goto out;
 910         }
 911
 912         ei = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
 913         flags = btrfs_extent_flags(leaf, ei);
 914
 915         ptr = (unsigned long)(ei + 1);
 916         end = (unsigned long)ei + item_size;
 917
 918         if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK && !skinny_metadata) {
 919                 ptr += sizeof(struct btrfs_tree_block_info);
 920                 BUG_ON(ptr > end);
 921         }
 922
 923         if (owner >= BTRFS_FIRST_FREE_OBJECTID)
 924                 needed = BTRFS_REF_TYPE_DATA;
 925         else
 926                 needed = BTRFS_REF_TYPE_BLOCK;
 927
 928         err = -ENOENT;
 929         while (1) {
 930                 if (ptr >= end) {
 931                         WARN_ON(ptr > end);
 932                         break;
 933                 }
 934                 iref = (struct btrfs_extent_inline_ref *)ptr;
 935                 type = btrfs_get_extent_inline_ref_type(leaf, iref, needed);
 936                 if (type == BTRFS_REF_TYPE_INVALID) {
 937                         err = -EUCLEAN;
 938                         goto out;
 939                 }
 940
 941                 if (want < type)
 942                         break;
 943                 if (want > type) {
 944                         ptr += btrfs_extent_inline_ref_size(type);
 945                         continue;
 946                 }
 947
 948                 if (type == BTRFS_EXTENT_DATA_REF_KEY) {
 949                         struct btrfs_extent_data_ref *dref;
 950                         dref = (struct btrfs_extent_data_ref *)(&iref->offset);
 951                         if (match_extent_data_ref(leaf, dref, root_objectid,
 952                                                   owner, offset)) {
 953                                 err = 0;
 954                                 break;
 955                         }
 956                         if (hash_extent_data_ref_item(leaf, dref) <
 957                             hash_extent_data_ref(root_objectid, owner, offset))
 958                                 break;
 959                 } else {
 960                         u64 ref_offset;
 961                         ref_offset = btrfs_extent_inline_ref_offset(leaf, iref);
 962                         if (parent > 0) {
 963                                 if (parent == ref_offset) {
 964                                         err = 0;
 965                                         break;
 966                                 }
 967                                 if (ref_offset < parent)
 968                                         break;
 969                         } else {
 970                                 if (root_objectid == ref_offset) {
 971                                         err = 0;
 972                                         break;
 973                                 }
 974                                 if (ref_offset < root_objectid)
 975                                         break;
 976                         }
 977                 }
 978                 ptr += btrfs_extent_inline_ref_size(type);
 979         }
 980         if (err == -ENOENT && insert) {
 981                 if (item_size + extra_size >=
 982                     BTRFS_MAX_EXTENT_ITEM_SIZE(root)) {
 983                         err = -EAGAIN;
 984                         goto out;
 985                 }
 986                 /*
 987                  * To add new inline back ref, we have to make sure
 988                  * there is no corresponding back ref item.
 989                  * For simplicity, we just do not add new inline back
 990                  * ref if there is any kind of item for this block
 991                  */
 992                 if (find_next_key(path, 0, &key) == 0 &&
 993                     key.objectid == bytenr &&
 994                     key.type < BTRFS_BLOCK_GROUP_ITEM_KEY) {
 995                         err = -EAGAIN;
 996                         goto out;
 997                 }
 998         }
 999         *ref_ret = (struct btrfs_extent_inline_ref *)ptr;
1000 out:
1001         if (insert) {
1002                 path->keep_locks = 0;
1003                 btrfs_unlock_up_safe(path, 1);
1004         }
1005         return err;
1006 }
1007
1008 /*
1009  * helper to add new inline back ref
1010  */
1011 static noinline_for_stack
1012 void setup_inline_extent_backref(struct btrfs_fs_info *fs_info,
1013                                  struct btrfs_path *path,
1014                                  struct btrfs_extent_inline_ref *iref,
1015                                  u64 parent, u64 root_objectid,
1016                                  u64 owner, u64 offset, int refs_to_add,
1017                                  struct btrfs_delayed_extent_op *extent_op)
1018 {
1019         struct extent_buffer *leaf;
1020         struct btrfs_extent_item *ei;
1021         unsigned long ptr;
1022         unsigned long end;
1023         unsigned long item_offset;
1024         u64 refs;
1025         int size;
1026         int type;
1027
1028         leaf = path->nodes[0];
1029         ei = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
1030         item_offset = (unsigned long)iref - (unsigned long)ei;
1031
1032         type = extent_ref_type(parent, owner);
1033         size = btrfs_extent_inline_ref_size(type);
1034
1035         btrfs_extend_item(path, size);
1036
1037         ei = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
1038         refs = btrfs_extent_refs(leaf, ei);
1039         refs += refs_to_add;
1040         btrfs_set_extent_refs(leaf, ei, refs);
1041         if (extent_op)
1042                 __run_delayed_extent_op(extent_op, leaf, ei);
1043
1044         ptr = (unsigned long)ei + item_offset;
1045         end = (unsigned long)ei + btrfs_item_size_nr(leaf, path->slots[0]);
1046         if (ptr < end - size)
1047                 memmove_extent_buffer(leaf, ptr + size, ptr,
1048                                       end - size - ptr);
1049
1050         iref = (struct btrfs_extent_inline_ref *)ptr;
1051         btrfs_set_extent_inline_ref_type(leaf, iref, type);
1052         if (type == BTRFS_EXTENT_DATA_REF_KEY) {
1053                 struct btrfs_extent_data_ref *dref;
1054                 dref = (struct btrfs_extent_data_ref *)(&iref->offset);
1055                 btrfs_set_extent_data_ref_root(leaf, dref, root_objectid);
1056                 btrfs_set_extent_data_ref_objectid(leaf, dref, owner);
1057                 btrfs_set_extent_data_ref_offset(leaf, dref, offset);
1058                 btrfs_set_extent_data_ref_count(leaf, dref, refs_to_add);
1059         } else if (type == BTRFS_SHARED_DATA_REF_KEY) {
1060                 struct btrfs_shared_data_ref *sref;
1061                 sref = (struct btrfs_shared_data_ref *)(iref + 1);
1062                 btrfs_set_shared_data_ref_count(leaf, sref, refs_to_add);
1063                 btrfs_set_extent_inline_ref_offset(leaf, iref, parent);
1064         } else if (type == BTRFS_SHARED_BLOCK_REF_KEY) {
1065                 btrfs_set_extent_inline_ref_offset(leaf, iref, parent);
1066         } else {
1067                 btrfs_set_extent_inline_ref_offset(leaf, iref, root_objectid);
1068         }
1069         btrfs_mark_buffer_dirty(leaf);
1070 }
1071
1072 static int lookup_extent_backref(struct btrfs_trans_handle *trans,
1073                                  struct btrfs_path *path,
1074                                  struct btrfs_extent_inline_ref **ref_ret,
1075                                  u64 bytenr, u64 num_bytes, u64 parent,
1076                                  u64 root_objectid, u64 owner, u64 offset)
1077 {
1078         int ret;
1079
1080         ret = lookup_inline_extent_backref(trans, path, ref_ret, bytenr,
1081                                            num_bytes, parent, root_objectid,
1082                                            owner, offset, 0);
1083         if (ret != -ENOENT)
1084                 return ret;
1085
1086         btrfs_release_path(path);
1087         *ref_ret = NULL;
1088
1089         if (owner < BTRFS_FIRST_FREE_OBJECTID) {
1090                 ret = lookup_tree_block_ref(trans, path, bytenr, parent,
1091                                             root_objectid);
1092         } else {
1093                 ret = lookup_extent_data_ref(trans, path, bytenr, parent,
1094                                              root_objectid, owner, offset);
1095         }
1096         return ret;
1097 }
1098
1099 /*
1100  * helper to update/remove inline back ref
1101  */
1102 static noinline_for_stack
1103 void update_inline_extent_backref(struct btrfs_path *path,
1104                                   struct btrfs_extent_inline_ref *iref,
1105                                   int refs_to_mod,
1106                                   struct btrfs_delayed_extent_op *extent_op,
1107                                   int *last_ref)
1108 {
1109         struct extent_buffer *leaf = path->nodes[0];
1110         struct btrfs_extent_item *ei;
1111         struct btrfs_extent_data_ref *dref = NULL;
1112         struct btrfs_shared_data_ref *sref = NULL;
1113         unsigned long ptr;
1114         unsigned long end;
1115         u32 item_size;
1116         int size;
1117         int type;
1118         u64 refs;
1119
1120         ei = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
1121         refs = btrfs_extent_refs(leaf, ei);
1122         WARN_ON(refs_to_mod < 0 && refs + refs_to_mod <= 0);
1123         refs += refs_to_mod;
1124         btrfs_set_extent_refs(leaf, ei, refs);
1125         if (extent_op)
1126                 __run_delayed_extent_op(extent_op, leaf, ei);
1127
1128         /*
1129          * If type is invalid, we should have bailed out after
1130          * lookup_inline_extent_backref().
1131          */
1132         type = btrfs_get_extent_inline_ref_type(leaf, iref, BTRFS_REF_TYPE_ANY);
1133         ASSERT(type != BTRFS_REF_TYPE_INVALID);
1134
1135         if (type == BTRFS_EXTENT_DATA_REF_KEY) {
1136                 dref = (struct btrfs_extent_data_ref *)(&iref->offset);
1137                 refs = btrfs_extent_data_ref_count(leaf, dref);
1138         } else if (type == BTRFS_SHARED_DATA_REF_KEY) {
1139                 sref = (struct btrfs_shared_data_ref *)(iref + 1);
1140                 refs = btrfs_shared_data_ref_count(leaf, sref);
1141         } else {
1142                 refs = 1;
1143                 BUG_ON(refs_to_mod != -1);
1144         }
1145
1146         BUG_ON(refs_to_mod < 0 && refs < -refs_to_mod);
1147         refs += refs_to_mod;
1148
1149         if (refs > 0) {
1150                 if (type == BTRFS_EXTENT_DATA_REF_KEY)
1151                         btrfs_set_extent_data_ref_count(leaf, dref, refs);
1152                 else
1153                         btrfs_set_shared_data_ref_count(leaf, sref, refs);
1154         } else {
1155                 *last_ref = 1;
1156                 size =  btrfs_extent_inline_ref_size(type);
1157                 item_size = btrfs_item_size_nr(leaf, path->slots[0]);
1158                 ptr = (unsigned long)iref;
1159                 end = (unsigned long)ei + item_size;
1160                 if (ptr + size < end)
1161                         memmove_extent_buffer(leaf, ptr, ptr + size,
1162                                               end - ptr - size);
1163                 item_size -= size;
1164                 btrfs_truncate_item(path, item_size, 1);
1165         }
1166         btrfs_mark_buffer_dirty(leaf);
1167 }
1168
1169 static noinline_for_stack
1170 int insert_inline_extent_backref(struct btrfs_trans_handle *trans,
1171                                  struct btrfs_path *path,
1172                                  u64 bytenr, u64 num_bytes, u64 parent,
1173                                  u64 root_objectid, u64 owner,
1174                                  u64 offset, int refs_to_add,
1175                                  struct btrfs_delayed_extent_op *extent_op)
1176 {
1177         struct btrfs_extent_inline_ref *iref;
1178         int ret;
1179
1180         ret = lookup_inline_extent_backref(trans, path, &iref, bytenr,
1181                                            num_bytes, parent, root_objectid,
1182                                            owner, offset, 1);
1183         if (ret == 0) {
1184                 BUG_ON(owner < BTRFS_FIRST_FREE_OBJECTID);
1185                 update_inline_extent_backref(path, iref, refs_to_add,
1186                                              extent_op, NULL);
1187         } else if (ret == -ENOENT) {
1188                 setup_inline_extent_backref(trans->fs_info, path, iref, parent,
1189                                             root_objectid, owner, offset,
1190                                             refs_to_add, extent_op);
1191                 ret = 0;
1192         }
1193         return ret;
1194 }
1195
1196 static int insert_extent_backref(struct btrfs_trans_handle *trans,
1197                                  struct btrfs_path *path,
1198                                  u64 bytenr, u64 parent, u64 root_objectid,
1199                                  u64 owner, u64 offset, int refs_to_add)
1200 {
1201         int ret;
1202         if (owner < BTRFS_FIRST_FREE_OBJECTID) {
1203                 BUG_ON(refs_to_add != 1);
1204                 ret = insert_tree_block_ref(trans, path, bytenr, parent,
1205                                             root_objectid);
1206         } else {
1207                 ret = insert_extent_data_ref(trans, path, bytenr, parent,
1208                                              root_objectid, owner, offset,
1209                                              refs_to_add);
1210         }
1211         return ret;
1212 }
1213
1214 static int remove_extent_backref(struct btrfs_trans_handle *trans,
1215                                  struct btrfs_path *path,
1216                                  struct btrfs_extent_inline_ref *iref,
1217                                  int refs_to_drop, int is_data, int *last_ref)
1218 {
1219         int ret = 0;
1220
1221         BUG_ON(!is_data && refs_to_drop != 1);
1222         if (iref) {
1223                 update_inline_extent_backref(path, iref, -refs_to_drop, NULL,
1224                                              last_ref);
1225         } else if (is_data) {
1226                 ret = remove_extent_data_ref(trans, path, refs_to_drop,
1227                                              last_ref);
1228         } else {
1229                 *last_ref = 1;
1230                 ret = btrfs_del_item(trans, trans->fs_info->extent_root, path);
1231         }
1232         return ret;
1233 }
1234
1235 static int btrfs_issue_discard(struct block_device *bdev, u64 start, u64 len,
1236                                u64 *discarded_bytes)
1237 {
1238         int j, ret = 0;
1239         u64 bytes_left, end;
1240         u64 aligned_start = ALIGN(start, 1 << 9);
1241
1242         if (WARN_ON(start != aligned_start)) {
1243                 len -= aligned_start - start;
1244                 len = round_down(len, 1 << 9);
1245                 start = aligned_start;
1246         }
1247
1248         *discarded_bytes = 0;
1249
1250         if (!len)
1251                 return 0;
1252
1253         end = start + len;
1254         bytes_left = len;
1255
1256         /* Skip any superblocks on this device. */
1257         for (j = 0; j < BTRFS_SUPER_MIRROR_MAX; j++) {
1258                 u64 sb_start = btrfs_sb_offset(j);
1259                 u64 sb_end = sb_start + BTRFS_SUPER_INFO_SIZE;
1260                 u64 size = sb_start - start;
1261
1262                 if (!in_range(sb_start, start, bytes_left) &&
1263                     !in_range(sb_end, start, bytes_left) &&
1264                     !in_range(start, sb_start, BTRFS_SUPER_INFO_SIZE))
1265                         continue;
1266
1267                 /*
1268                  * Superblock spans beginning of range.  Adjust start and
1269                  * try again.
1270                  */
1271                 if (sb_start <= start) {
1272                         start += sb_end - start;
1273                         if (start > end) {
1274                                 bytes_left = 0;
1275                                 break;
1276                         }
1277                         bytes_left = end - start;
1278                         continue;
1279                 }
1280
1281                 if (size) {
1282                         ret = blkdev_issue_discard(bdev, start >> 9, size >> 9,
1283                                                    GFP_NOFS, 0);
1284                         if (!ret)
1285                                 *discarded_bytes += size;
1286                         else if (ret != -EOPNOTSUPP)
1287                                 return ret;
1288                 }
1289
1290                 start = sb_end;
1291                 if (start > end) {
1292                         bytes_left = 0;
1293                         break;
1294                 }
1295                 bytes_left = end - start;
1296         }
1297
1298         if (bytes_left) {
1299                 ret = blkdev_issue_discard(bdev, start >> 9, bytes_left >> 9,
1300                                            GFP_NOFS, 0);
1301                 if (!ret)
1302                         *discarded_bytes += bytes_left;
1303         }
1304         return ret;
1305 }
1306
1307 int btrfs_discard_extent(struct btrfs_fs_info *fs_info, u64 bytenr,
1308                          u64 num_bytes, u64 *actual_bytes)
1309 {
1310         int ret = 0;
1311         u64 discarded_bytes = 0;
1312         u64 end = bytenr + num_bytes;
1313         u64 cur = bytenr;
1314         struct btrfs_bio *bbio = NULL;
1315
1316
1317         /*
1318          * Avoid races with device replace and make sure our bbio has devices
1319          * associated to its stripes that don't go away while we are discarding.
1320          */
1321         btrfs_bio_counter_inc_blocked(fs_info);
1322         while (cur < end) {
1323                 struct btrfs_bio_stripe *stripe;
1324                 int i;
1325
1326                 num_bytes = end - cur;
1327                 /* Tell the block device(s) that the sectors can be discarded */
1328                 ret = btrfs_map_block(fs_info, BTRFS_MAP_DISCARD, cur,
1329                                       &num_bytes, &bbio, 0);
1330                 /*
1331                  * Error can be -ENOMEM, -ENOENT (no such chunk mapping) or
1332                  * -EOPNOTSUPP. For any such error, @num_bytes is not updated,
1333                  * thus we can't continue anyway.
1334                  */
1335                 if (ret < 0)
1336                         goto out;
1337
1338                 stripe = bbio->stripes;
1339                 for (i = 0; i < bbio->num_stripes; i++, stripe++) {
1340                         u64 bytes;
1341                         struct request_queue *req_q;
1342
1343                         if (!stripe->dev->bdev) {
1344                                 ASSERT(btrfs_test_opt(fs_info, DEGRADED));
1345                                 continue;
1346                         }
1347                         req_q = bdev_get_queue(stripe->dev->bdev);
1348                         if (!blk_queue_discard(req_q))
1349                                 continue;
1350
1351                         ret = btrfs_issue_discard(stripe->dev->bdev,
1352                                                   stripe->physical,
1353                                                   stripe->length,
1354                                                   &bytes);
1355                         if (!ret) {
1356                                 discarded_bytes += bytes;
1357                         } else if (ret != -EOPNOTSUPP) {
1358                                 /*
1359                                  * Logic errors or -ENOMEM, or -EIO, but
1360                                  * unlikely to happen.
1361                                  *
1362                                  * And since there are two loops, explicitly
1363                                  * go to out to avoid confusion.
1364                                  */
1365                                 btrfs_put_bbio(bbio);
1366                                 goto out;
1367                         }
1368
1369                         /*
1370                          * Just in case we get back EOPNOTSUPP for some reason,
1371                          * just ignore the return value so we don't screw up
1372                          * people calling discard_extent.
1373                          */
1374                         ret = 0;
1375                 }
1376                 btrfs_put_bbio(bbio);
1377                 cur += num_bytes;
1378         }
1379 out:
1380         btrfs_bio_counter_dec(fs_info);
1381
1382         if (actual_bytes)
1383                 *actual_bytes = discarded_bytes;
1384
1385
1386         if (ret == -EOPNOTSUPP)
1387                 ret = 0;
1388         return ret;
1389 }
1390
1391 /* Can return -ENOMEM */
1392 int btrfs_inc_extent_ref(struct btrfs_trans_handle *trans,
1393                          struct btrfs_ref *generic_ref)
1394 {
1395         struct btrfs_fs_info *fs_info = trans->fs_info;
1396         int old_ref_mod, new_ref_mod;
1397         int ret;
1398
1399         ASSERT(generic_ref->type != BTRFS_REF_NOT_SET &&
1400                generic_ref->action);
1401         BUG_ON(generic_ref->type == BTRFS_REF_METADATA &&
1402                generic_ref->tree_ref.root == BTRFS_TREE_LOG_OBJECTID);
1403
1404         if (generic_ref->type == BTRFS_REF_METADATA)
1405                 ret = btrfs_add_delayed_tree_ref(trans, generic_ref,
1406                                 NULL, &old_ref_mod, &new_ref_mod);
1407         else
1408                 ret = btrfs_add_delayed_data_ref(trans, generic_ref, 0,
1409                                                  &old_ref_mod, &new_ref_mod);
1410
1411         btrfs_ref_tree_mod(fs_info, generic_ref);
1412
1413         if (ret == 0 && old_ref_mod < 0 && new_ref_mod >= 0)
1414                 sub_pinned_bytes(fs_info, generic_ref);
1415
1416         return ret;
1417 }
1418
1419 /*
1420  * __btrfs_inc_extent_ref - insert backreference for a given extent
1421  *
1422  * @trans:          Handle of transaction
1423  *
1424  * @node:           The delayed ref node used to get the bytenr/length for
1425  *                  extent whose references are incremented.
1426  *
1427  * @parent:         If this is a shared extent (BTRFS_SHARED_DATA_REF_KEY/
1428  *                  BTRFS_SHARED_BLOCK_REF_KEY) then it holds the logical
1429  *                  bytenr of the parent block. Since new extents are always
1430  *                  created with indirect references, this will only be the case
1431  *                  when relocating a shared extent. In that case, root_objectid
1432  *                  will be BTRFS_TREE_RELOC_OBJECTID. Otheriwse, parent must
1433  *                  be 0
1434  *
1435  * @root_objectid:  The id of the root where this modification has originated,
1436  *                  this can be either one of the well-known metadata trees or
1437  *                  the subvolume id which references this extent.
1438  *
1439  * @owner:          For data extents it is the inode number of the owning file.
1440  *                  For metadata extents this parameter holds the level in the
1441  *                  tree of the extent.
1442  *
1443  * @offset:         For metadata extents the offset is ignored and is currently
1444  *                  always passed as 0. For data extents it is the fileoffset
1445  *                  this extent belongs to.
1446  *
1447  * @refs_to_add     Number of references to add
1448  *
1449  * @extent_op       Pointer to a structure, holding information necessary when
1450  *                  updating a tree block's flags
1451  *
1452  */
1453 static int __btrfs_inc_extent_ref(struct btrfs_trans_handle *trans,
1454                                   struct btrfs_delayed_ref_node *node,
1455                                   u64 parent, u64 root_objectid,
1456                                   u64 owner, u64 offset, int refs_to_add,
1457                                   struct btrfs_delayed_extent_op *extent_op)
1458 {
1459         struct btrfs_path *path;
1460         struct extent_buffer *leaf;
1461         struct btrfs_extent_item *item;
1462         struct btrfs_key key;
1463         u64 bytenr = node->bytenr;
1464         u64 num_bytes = node->num_bytes;
1465         u64 refs;
1466         int ret;
1467
1468         path = btrfs_alloc_path();
1469         if (!path)
1470                 return -ENOMEM;
1471
1472         path->reada = READA_FORWARD;
1473         path->leave_spinning = 1;
1474         /* this will setup the path even if it fails to insert the back ref */
1475         ret = insert_inline_extent_backref(trans, path, bytenr, num_bytes,
1476                                            parent, root_objectid, owner,
1477                                            offset, refs_to_add, extent_op);
1478         if ((ret < 0 && ret != -EAGAIN) || !ret)
1479                 goto out;
1480
1481         /*
1482          * Ok we had -EAGAIN which means we didn't have space to insert and
1483          * inline extent ref, so just update the reference count and add a
1484          * normal backref.
1485          */
1486         leaf = path->nodes[0];
1487         btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
1488         item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
1489         refs = btrfs_extent_refs(leaf, item);
1490         btrfs_set_extent_refs(leaf, item, refs + refs_to_add);
1491         if (extent_op)
1492                 __run_delayed_extent_op(extent_op, leaf, item);
1493
1494         btrfs_mark_buffer_dirty(leaf);
1495         btrfs_release_path(path);
1496
1497         path->reada = READA_FORWARD;
1498         path->leave_spinning = 1;
1499         /* now insert the actual backref */
1500         ret = insert_extent_backref(trans, path, bytenr, parent, root_objectid,
1501                                     owner, offset, refs_to_add);
1502         if (ret)
1503                 btrfs_abort_transaction(trans, ret);
1504 out:
1505         btrfs_free_path(path);
1506         return ret;
1507 }
1508
1509 static int run_delayed_data_ref(struct btrfs_trans_handle *trans,
1510                                 struct btrfs_delayed_ref_node *node,
1511                                 struct btrfs_delayed_extent_op *extent_op,
1512                                 int insert_reserved)
1513 {
1514         int ret = 0;
1515         struct btrfs_delayed_data_ref *ref;
1516         struct btrfs_key ins;
1517         u64 parent = 0;
1518         u64 ref_root = 0;
1519         u64 flags = 0;
1520
1521         ins.objectid = node->bytenr;
1522         ins.offset = node->num_bytes;
1523         ins.type = BTRFS_EXTENT_ITEM_KEY;
1524
1525         ref = btrfs_delayed_node_to_data_ref(node);
1526         trace_run_delayed_data_ref(trans->fs_info, node, ref, node->action);
1527
1528         if (node->type == BTRFS_SHARED_DATA_REF_KEY)
1529                 parent = ref->parent;
1530         ref_root = ref->root;
1531
1532         if (node->action == BTRFS_ADD_DELAYED_REF && insert_reserved) {
1533                 if (extent_op)
1534                         flags |= extent_op->flags_to_set;
1535                 ret = alloc_reserved_file_extent(trans, parent, ref_root,
1536                                                  flags, ref->objectid,
1537                                                  ref->offset, &ins,
1538                                                  node->ref_mod);
1539         } else if (node->action == BTRFS_ADD_DELAYED_REF) {
1540                 ret = __btrfs_inc_extent_ref(trans, node, parent, ref_root,
1541                                              ref->objectid, ref->offset,
1542                                              node->ref_mod, extent_op);
1543         } else if (node->action == BTRFS_DROP_DELAYED_REF) {
1544                 ret = __btrfs_free_extent(trans, node, parent,
1545                                           ref_root, ref->objectid,
1546                                           ref->offset, node->ref_mod,
1547                                           extent_op);
1548         } else {
1549                 BUG();
1550         }
1551         return ret;
1552 }
1553
1554 static void __run_delayed_extent_op(struct btrfs_delayed_extent_op *extent_op,
1555                                     struct extent_buffer *leaf,
1556                                     struct btrfs_extent_item *ei)
1557 {
1558         u64 flags = btrfs_extent_flags(leaf, ei);
1559         if (extent_op->update_flags) {
1560                 flags |= extent_op->flags_to_set;
1561                 btrfs_set_extent_flags(leaf, ei, flags);
1562         }
1563
1564         if (extent_op->update_key) {
1565                 struct btrfs_tree_block_info *bi;
1566                 BUG_ON(!(flags & BTRFS_EXTENT_FLAG_TREE_BLOCK));
1567                 bi = (struct btrfs_tree_block_info *)(ei + 1);
1568                 btrfs_set_tree_block_key(leaf, bi, &extent_op->key);
1569         }
1570 }
1571
1572 static int run_delayed_extent_op(struct btrfs_trans_handle *trans,
1573                                  struct btrfs_delayed_ref_head *head,
1574                                  struct btrfs_delayed_extent_op *extent_op)
1575 {
1576         struct btrfs_fs_info *fs_info = trans->fs_info;
1577         struct btrfs_key key;
1578         struct btrfs_path *path;
1579         struct btrfs_extent_item *ei;
1580         struct extent_buffer *leaf;
1581         u32 item_size;
1582         int ret;
1583         int err = 0;
1584         int metadata = !extent_op->is_data;
1585
1586         if (trans->aborted)
1587                 return 0;
1588
1589         if (metadata && !btrfs_fs_incompat(fs_info, SKINNY_METADATA))
1590                 metadata = 0;
1591
1592         path = btrfs_alloc_path();
1593         if (!path)
1594                 return -ENOMEM;
1595
1596         key.objectid = head->bytenr;
1597
1598         if (metadata) {
1599                 key.type = BTRFS_METADATA_ITEM_KEY;
1600                 key.offset = extent_op->level;
1601         } else {
1602                 key.type = BTRFS_EXTENT_ITEM_KEY;
1603                 key.offset = head->num_bytes;
1604         }
1605
1606 again:
1607         path->reada = READA_FORWARD;
1608         path->leave_spinning = 1;
1609         ret = btrfs_search_slot(trans, fs_info->extent_root, &key, path, 0, 1);
1610         if (ret < 0) {
1611                 err = ret;
1612                 goto out;
1613         }
1614         if (ret > 0) {
1615                 if (metadata) {
1616                         if (path->slots[0] > 0) {
1617                                 path->slots[0]--;
1618                                 btrfs_item_key_to_cpu(path->nodes[0], &key,
1619                                                       path->slots[0]);
1620                                 if (key.objectid == head->bytenr &&
1621                                     key.type == BTRFS_EXTENT_ITEM_KEY &&
1622                                     key.offset == head->num_bytes)
1623                                         ret = 0;
1624                         }
1625                         if (ret > 0) {
1626                                 btrfs_release_path(path);
1627                                 metadata = 0;
1628
1629                                 key.objectid = head->bytenr;
1630                                 key.offset = head->num_bytes;
1631                                 key.type = BTRFS_EXTENT_ITEM_KEY;
1632                                 goto again;
1633                         }
1634                 } else {
1635                         err = -EIO;
1636                         goto out;
1637                 }
1638         }
1639
1640         leaf = path->nodes[0];
1641         item_size = btrfs_item_size_nr(leaf, path->slots[0]);
1642
1643         if (unlikely(item_size < sizeof(*ei))) {
1644                 err = -EINVAL;
1645                 btrfs_print_v0_err(fs_info);
1646                 btrfs_abort_transaction(trans, err);
1647                 goto out;
1648         }
1649
1650         ei = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
1651         __run_delayed_extent_op(extent_op, leaf, ei);
1652
1653         btrfs_mark_buffer_dirty(leaf);
1654 out:
1655         btrfs_free_path(path);
1656         return err;
1657 }
1658
1659 static int run_delayed_tree_ref(struct btrfs_trans_handle *trans,
1660                                 struct btrfs_delayed_ref_node *node,
1661                                 struct btrfs_delayed_extent_op *extent_op,
1662                                 int insert_reserved)
1663 {
1664         int ret = 0;
1665         struct btrfs_delayed_tree_ref *ref;
1666         u64 parent = 0;
1667         u64 ref_root = 0;
1668
1669         ref = btrfs_delayed_node_to_tree_ref(node);
1670         trace_run_delayed_tree_ref(trans->fs_info, node, ref, node->action);
1671
1672         if (node->type == BTRFS_SHARED_BLOCK_REF_KEY)
1673                 parent = ref->parent;
1674         ref_root = ref->root;
1675
1676         if (node->ref_mod != 1) {
1677                 btrfs_err(trans->fs_info,
1678         "btree block(%llu) has %d references rather than 1: action %d ref_root %llu parent %llu",
1679                           node->bytenr, node->ref_mod, node->action, ref_root,
1680                           parent);
1681                 return -EIO;
1682         }
1683         if (node->action == BTRFS_ADD_DELAYED_REF && insert_reserved) {
1684                 BUG_ON(!extent_op || !extent_op->update_flags);
1685                 ret = alloc_reserved_tree_block(trans, node, extent_op);
1686         } else if (node->action == BTRFS_ADD_DELAYED_REF) {
1687                 ret = __btrfs_inc_extent_ref(trans, node, parent, ref_root,
1688                                              ref->level, 0, 1, extent_op);
1689         } else if (node->action == BTRFS_DROP_DELAYED_REF) {
1690                 ret = __btrfs_free_extent(trans, node, parent, ref_root,
1691                                           ref->level, 0, 1, extent_op);
1692         } else {
1693                 BUG();
1694         }
1695         return ret;
1696 }
1697
1698 /* helper function to actually process a single delayed ref entry */
1699 static int run_one_delayed_ref(struct btrfs_trans_handle *trans,
1700                                struct btrfs_delayed_ref_node *node,
1701                                struct btrfs_delayed_extent_op *extent_op,
1702                                int insert_reserved)
1703 {
1704         int ret = 0;
1705
1706         if (trans->aborted) {
1707                 if (insert_reserved)
1708                         btrfs_pin_extent(trans->fs_info, node->bytenr,
1709                                          node->num_bytes, 1);
1710                 return 0;
1711         }
1712
1713         if (node->type == BTRFS_TREE_BLOCK_REF_KEY ||
1714             node->type == BTRFS_SHARED_BLOCK_REF_KEY)
1715                 ret = run_delayed_tree_ref(trans, node, extent_op,
1716                                            insert_reserved);
1717         else if (node->type == BTRFS_EXTENT_DATA_REF_KEY ||
1718                  node->type == BTRFS_SHARED_DATA_REF_KEY)
1719                 ret = run_delayed_data_ref(trans, node, extent_op,
1720                                            insert_reserved);
1721         else
1722                 BUG();
1723         if (ret && insert_reserved)
1724                 btrfs_pin_extent(trans->fs_info, node->bytenr,
1725                                  node->num_bytes, 1);
1726         return ret;
1727 }
1728
1729 static inline struct btrfs_delayed_ref_node *
1730 select_delayed_ref(struct btrfs_delayed_ref_head *head)
1731 {
1732         struct btrfs_delayed_ref_node *ref;
1733
1734         if (RB_EMPTY_ROOT(&head->ref_tree.rb_root))
1735                 return NULL;
1736
1737         /*
1738          * Select a delayed ref of type BTRFS_ADD_DELAYED_REF first.
1739          * This is to prevent a ref count from going down to zero, which deletes
1740          * the extent item from the extent tree, when there still are references
1741          * to add, which would fail because they would not find the extent item.
1742          */
1743         if (!list_empty(&head->ref_add_list))
1744                 return list_first_entry(&head->ref_add_list,
1745                                 struct btrfs_delayed_ref_node, add_list);
1746
1747         ref = rb_entry(rb_first_cached(&head->ref_tree),
1748                        struct btrfs_delayed_ref_node, ref_node);
1749         ASSERT(list_empty(&ref->add_list));
1750         return ref;
1751 }
1752
1753 static void unselect_delayed_ref_head(struct btrfs_delayed_ref_root *delayed_refs,
1754                                       struct btrfs_delayed_ref_head *head)
1755 {
1756         spin_lock(&delayed_refs->lock);
1757         head->processing = 0;
1758         delayed_refs->num_heads_ready++;
1759         spin_unlock(&delayed_refs->lock);
1760         btrfs_delayed_ref_unlock(head);
1761 }
1762
1763 static struct btrfs_delayed_extent_op *cleanup_extent_op(
1764                                 struct btrfs_delayed_ref_head *head)
1765 {
1766         struct btrfs_delayed_extent_op *extent_op = head->extent_op;
1767
1768         if (!extent_op)
1769                 return NULL;
1770
1771         if (head->must_insert_reserved) {
1772                 head->extent_op = NULL;
1773                 btrfs_free_delayed_extent_op(extent_op);
1774                 return NULL;
1775         }
1776         return extent_op;
1777 }
1778
1779 static int run_and_cleanup_extent_op(struct btrfs_trans_handle *trans,
1780                                      struct btrfs_delayed_ref_head *head)
1781 {
1782         struct btrfs_delayed_extent_op *extent_op;
1783         int ret;
1784
1785         extent_op = cleanup_extent_op(head);
1786         if (!extent_op)
1787                 return 0;
1788         head->extent_op = NULL;
1789         spin_unlock(&head->lock);
1790         ret = run_delayed_extent_op(trans, head, extent_op);
1791         btrfs_free_delayed_extent_op(extent_op);
1792         return ret ? ret : 1;
1793 }
1794
1795 void btrfs_cleanup_ref_head_accounting(struct btrfs_fs_info *fs_info,
1796                                   struct btrfs_delayed_ref_root *delayed_refs,
1797                                   struct btrfs_delayed_ref_head *head)
1798 {
1799         int nr_items = 1;       /* Dropping this ref head update. */
1800
1801         if (head->total_ref_mod < 0) {
1802                 struct btrfs_space_info *space_info;
1803                 u64 flags;
1804
1805                 if (head->is_data)
1806                         flags = BTRFS_BLOCK_GROUP_DATA;
1807                 else if (head->is_system)
1808                         flags = BTRFS_BLOCK_GROUP_SYSTEM;
1809                 else
1810                         flags = BTRFS_BLOCK_GROUP_METADATA;
1811                 space_info = btrfs_find_space_info(fs_info, flags);
1812                 ASSERT(space_info);
1813                 percpu_counter_add_batch(&space_info->total_bytes_pinned,
1814                                    -head->num_bytes,
1815                                    BTRFS_TOTAL_BYTES_PINNED_BATCH);
1816
1817                 /*
1818                  * We had csum deletions accounted for in our delayed refs rsv,
1819                  * we need to drop the csum leaves for this update from our
1820                  * delayed_refs_rsv.
1821                  */
1822                 if (head->is_data) {
1823                         spin_lock(&delayed_refs->lock);
1824                         delayed_refs->pending_csums -= head->num_bytes;
1825                         spin_unlock(&delayed_refs->lock);
1826                         nr_items += btrfs_csum_bytes_to_leaves(fs_info,
1827                                 head->num_bytes);
1828                 }
1829         }
1830
1831         btrfs_delayed_refs_rsv_release(fs_info, nr_items);
1832 }
1833
1834 static int cleanup_ref_head(struct btrfs_trans_handle *trans,
1835                             struct btrfs_delayed_ref_head *head)
1836 {
1837
1838         struct btrfs_fs_info *fs_info = trans->fs_info;
1839         struct btrfs_delayed_ref_root *delayed_refs;
1840         int ret;
1841
1842         delayed_refs = &trans->transaction->delayed_refs;
1843
1844         ret = run_and_cleanup_extent_op(trans, head);
1845         if (ret < 0) {
1846                 unselect_delayed_ref_head(delayed_refs, head);
1847                 btrfs_debug(fs_info, "run_delayed_extent_op returned %d", ret);
1848                 return ret;
1849         } else if (ret) {
1850                 return ret;
1851         }
1852
1853         /*
1854          * Need to drop our head ref lock and re-acquire the delayed ref lock
1855          * and then re-check to make sure nobody got added.
1856          */
1857         spin_unlock(&head->lock);
1858         spin_lock(&delayed_refs->lock);
1859         spin_lock(&head->lock);
1860         if (!RB_EMPTY_ROOT(&head->ref_tree.rb_root) || head->extent_op) {
1861                 spin_unlock(&head->lock);
1862                 spin_unlock(&delayed_refs->lock);
1863                 return 1;
1864         }
1865         btrfs_delete_ref_head(delayed_refs, head);
1866         spin_unlock(&head->lock);
1867         spin_unlock(&delayed_refs->lock);
1868
1869         if (head->must_insert_reserved) {
1870                 btrfs_pin_extent(fs_info, head->bytenr,
1871                                  head->num_bytes, 1);
1872                 if (head->is_data) {
1873                         ret = btrfs_del_csums(trans, fs_info->csum_root,
1874                                               head->bytenr, head->num_bytes);
1875                 }
1876         }
1877
1878         btrfs_cleanup_ref_head_accounting(fs_info, delayed_refs, head);
1879
1880         trace_run_delayed_ref_head(fs_info, head, 0);
1881         btrfs_delayed_ref_unlock(head);
1882         btrfs_put_delayed_ref_head(head);
1883         return 0;
1884 }
1885
1886 static struct btrfs_delayed_ref_head *btrfs_obtain_ref_head(
1887                                         struct btrfs_trans_handle *trans)
1888 {
1889         struct btrfs_delayed_ref_root *delayed_refs =
1890                 &trans->transaction->delayed_refs;
1891         struct btrfs_delayed_ref_head *head = NULL;
1892         int ret;
1893
1894         spin_lock(&delayed_refs->lock);
1895         head = btrfs_select_ref_head(delayed_refs);
1896         if (!head) {
1897                 spin_unlock(&delayed_refs->lock);
1898                 return head;
1899         }
1900
1901         /*
1902          * Grab the lock that says we are going to process all the refs for
1903          * this head
1904          */
1905         ret = btrfs_delayed_ref_lock(delayed_refs, head);
1906         spin_unlock(&delayed_refs->lock);
1907
1908         /*
1909          * We may have dropped the spin lock to get the head mutex lock, and
1910          * that might have given someone else time to free the head.  If that's
1911          * true, it has been removed from our list and we can move on.
1912          */
1913         if (ret == -EAGAIN)
1914                 head = ERR_PTR(-EAGAIN);
1915
1916         return head;
1917 }
1918
1919 static int btrfs_run_delayed_refs_for_head(struct btrfs_trans_handle *trans,
1920                                     struct btrfs_delayed_ref_head *locked_ref,
1921                                     unsigned long *run_refs)
1922 {
1923         struct btrfs_fs_info *fs_info = trans->fs_info;
1924         struct btrfs_delayed_ref_root *delayed_refs;
1925         struct btrfs_delayed_extent_op *extent_op;
1926         struct btrfs_delayed_ref_node *ref;
1927         int must_insert_reserved = 0;
1928         int ret;
1929
1930         delayed_refs = &trans->transaction->delayed_refs;
1931
1932         lockdep_assert_held(&locked_ref->mutex);
1933         lockdep_assert_held(&locked_ref->lock);
1934
1935         while ((ref = select_delayed_ref(locked_ref))) {
1936                 if (ref->seq &&
1937                     btrfs_check_delayed_seq(fs_info, ref->seq)) {
1938                         spin_unlock(&locked_ref->lock);
1939                         unselect_delayed_ref_head(delayed_refs, locked_ref);
1940                         return -EAGAIN;
1941                 }
1942
1943                 (*run_refs)++;
1944                 ref->in_tree = 0;
1945                 rb_erase_cached(&ref->ref_node, &locked_ref->ref_tree);
1946                 RB_CLEAR_NODE(&ref->ref_node);
1947                 if (!list_empty(&ref->add_list))
1948                         list_del(&ref->add_list);
1949                 /*
1950                  * When we play the delayed ref, also correct the ref_mod on
1951                  * head
1952                  */
1953                 switch (ref->action) {
1954                 case BTRFS_ADD_DELAYED_REF:
1955                 case BTRFS_ADD_DELAYED_EXTENT:
1956                         locked_ref->ref_mod -= ref->ref_mod;
1957                         break;
1958                 case BTRFS_DROP_DELAYED_REF:
1959                         locked_ref->ref_mod += ref->ref_mod;
1960                         break;
1961                 default:
1962                         WARN_ON(1);
1963                 }
1964                 atomic_dec(&delayed_refs->num_entries);
1965
1966                 /*
1967                  * Record the must_insert_reserved flag before we drop the
1968                  * spin lock.
1969                  */
1970                 must_insert_reserved = locked_ref->must_insert_reserved;
1971                 locked_ref->must_insert_reserved = 0;
1972
1973                 extent_op = locked_ref->extent_op;
1974                 locked_ref->extent_op = NULL;
1975                 spin_unlock(&locked_ref->lock);
1976
1977                 ret = run_one_delayed_ref(trans, ref, extent_op,
1978                                           must_insert_reserved);
1979
1980                 btrfs_free_delayed_extent_op(extent_op);
1981                 if (ret) {
1982                         unselect_delayed_ref_head(delayed_refs, locked_ref);
1983                         btrfs_put_delayed_ref(ref);
1984                         btrfs_debug(fs_info, "run_one_delayed_ref returned %d",
1985                                     ret);
1986                         return ret;
1987                 }
1988
1989                 btrfs_put_delayed_ref(ref);
1990                 cond_resched();
1991
1992                 spin_lock(&locked_ref->lock);
1993                 btrfs_merge_delayed_refs(trans, delayed_refs, locked_ref);
1994         }
1995
1996         return 0;
1997 }
1998
1999 /*
2000  * Returns 0 on success or if called with an already aborted transaction.
2001  * Returns -ENOMEM or -EIO on failure and will abort the transaction.
2002  */
2003 static noinline int __btrfs_run_delayed_refs(struct btrfs_trans_handle *trans,
2004                                              unsigned long nr)
2005 {
2006         struct btrfs_fs_info *fs_info = trans->fs_info;
2007         struct btrfs_delayed_ref_root *delayed_refs;
2008         struct btrfs_delayed_ref_head *locked_ref = NULL;
2009         ktime_t start = ktime_get();
2010         int ret;
2011         unsigned long count = 0;
2012         unsigned long actual_count = 0;
2013
2014         delayed_refs = &trans->transaction->delayed_refs;
2015         do {
2016                 if (!locked_ref) {
2017                         locked_ref = btrfs_obtain_ref_head(trans);
2018                         if (IS_ERR_OR_NULL(locked_ref)) {
2019                                 if (PTR_ERR(locked_ref) == -EAGAIN) {
2020                                         continue;
2021                                 } else {
2022                                         break;
2023                                 }
2024                         }
2025                         count++;
2026                 }
2027                 /*
2028                  * We need to try and merge add/drops of the same ref since we
2029                  * can run into issues with relocate dropping the implicit ref
2030                  * and then it being added back again before the drop can
2031                  * finish.  If we merged anything we need to re-loop so we can
2032                  * get a good ref.
2033                  * Or we can get node references of the same type that weren't
2034                  * merged when created due to bumps in the tree mod seq, and
2035                  * we need to merge them to prevent adding an inline extent
2036                  * backref before dropping it (triggering a BUG_ON at
2037                  * insert_inline_extent_backref()).
2038                  */
2039                 spin_lock(&locked_ref->lock);
2040                 btrfs_merge_delayed_refs(trans, delayed_refs, locked_ref);
2041
2042                 ret = btrfs_run_delayed_refs_for_head(trans, locked_ref,
2043                                                       &actual_count);
2044                 if (ret < 0 && ret != -EAGAIN) {
2045                         /*
2046                          * Error, btrfs_run_delayed_refs_for_head already
2047                          * unlocked everything so just bail out
2048                          */
2049                         return ret;
2050                 } else if (!ret) {
2051                         /*
2052                          * Success, perform the usual cleanup of a processed
2053                          * head
2054                          */
2055                         ret = cleanup_ref_head(trans, locked_ref);
2056                         if (ret > 0 ) {
2057                                 /* We dropped our lock, we need to loop. */
2058                                 ret = 0;
2059                                 continue;
2060                         } else if (ret) {
2061                                 return ret;
2062                         }
2063                 }
2064
2065                 /*
2066                  * Either success case or btrfs_run_delayed_refs_for_head
2067                  * returned -EAGAIN, meaning we need to select another head
2068                  */
2069
2070                 locked_ref = NULL;
2071                 cond_resched();
2072         } while ((nr != -1 && count < nr) || locked_ref);
2073
2074         /*
2075          * We don't want to include ref heads since we can have empty ref heads
2076          * and those will drastically skew our runtime down since we just do
2077          * accounting, no actual extent tree updates.
2078          */
2079         if (actual_count > 0) {
2080                 u64 runtime = ktime_to_ns(ktime_sub(ktime_get(), start));
2081                 u64 avg;
2082
2083                 /*
2084                  * We weigh the current average higher than our current runtime
2085                  * to avoid large swings in the average.
2086                  */
2087                 spin_lock(&delayed_refs->lock);
2088                 avg = fs_info->avg_delayed_ref_runtime * 3 + runtime;
2089                 fs_info->avg_delayed_ref_runtime = avg >> 2;    /* div by 4 */
2090                 spin_unlock(&delayed_refs->lock);
2091         }
2092         return 0;
2093 }
2094
2095 #ifdef SCRAMBLE_DELAYED_REFS
2096 /*
2097  * Normally delayed refs get processed in ascending bytenr order. This
2098  * correlates in most cases to the order added. To expose dependencies on this
2099  * order, we start to process the tree in the middle instead of the beginning
2100  */
2101 static u64 find_middle(struct rb_root *root)
2102 {
2103         struct rb_node *n = root->rb_node;
2104         struct btrfs_delayed_ref_node *entry;
2105         int alt = 1;
2106         u64 middle;
2107         u64 first = 0, last = 0;
2108
2109         n = rb_first(root);
2110         if (n) {
2111                 entry = rb_entry(n, struct btrfs_delayed_ref_node, rb_node);
2112                 first = entry->bytenr;
2113         }
2114         n = rb_last(root);
2115         if (n) {
2116                 entry = rb_entry(n, struct btrfs_delayed_ref_node, rb_node);
2117                 last = entry->bytenr;
2118         }
2119         n = root->rb_node;
2120
2121         while (n) {
2122                 entry = rb_entry(n, struct btrfs_delayed_ref_node, rb_node);
2123                 WARN_ON(!entry->in_tree);
2124
2125                 middle = entry->bytenr;
2126
2127                 if (alt)
2128                         n = n->rb_left;
2129                 else
2130                         n = n->rb_right;
2131
2132                 alt = 1 - alt;
2133         }
2134         return middle;
2135 }
2136 #endif
2137
2138 static inline u64 heads_to_leaves(struct btrfs_fs_info *fs_info, u64 heads)
2139 {
2140         u64 num_bytes;
2141
2142         num_bytes = heads * (sizeof(struct btrfs_extent_item) +
2143                              sizeof(struct btrfs_extent_inline_ref));
2144         if (!btrfs_fs_incompat(fs_info, SKINNY_METADATA))
2145                 num_bytes += heads * sizeof(struct btrfs_tree_block_info);
2146
2147         /*
2148          * We don't ever fill up leaves all the way so multiply by 2 just to be
2149          * closer to what we're really going to want to use.
2150          */
2151         return div_u64(num_bytes, BTRFS_LEAF_DATA_SIZE(fs_info));
2152 }
2153
2154 /*
2155  * Takes the number of bytes to be csumm'ed and figures out how many leaves it
2156  * would require to store the csums for that many bytes.
2157  */
2158 u64 btrfs_csum_bytes_to_leaves(struct btrfs_fs_info *fs_info, u64 csum_bytes)
2159 {
2160         u64 csum_size;
2161         u64 num_csums_per_leaf;
2162         u64 num_csums;
2163
2164         csum_size = BTRFS_MAX_ITEM_SIZE(fs_info);
2165         num_csums_per_leaf = div64_u64(csum_size,
2166                         (u64)btrfs_super_csum_size(fs_info->super_copy));
2167         num_csums = div64_u64(csum_bytes, fs_info->sectorsize);
2168         num_csums += num_csums_per_leaf - 1;
2169         num_csums = div64_u64(num_csums, num_csums_per_leaf);
2170         return num_csums;
2171 }
2172
2173 /*
2174  * this starts processing the delayed reference count updates and
2175  * extent insertions we have queued up so far.  count can be
2176  * 0, which means to process everything in the tree at the start
2177  * of the run (but not newly added entries), or it can be some target
2178  * number you'd like to process.
2179  *
2180  * Returns 0 on success or if called with an aborted transaction
2181  * Returns <0 on error and aborts the transaction
2182  */
2183 int btrfs_run_delayed_refs(struct btrfs_trans_handle *trans,
2184                            unsigned long count)
2185 {
2186         struct btrfs_fs_info *fs_info = trans->fs_info;
2187         struct rb_node *node;
2188         struct btrfs_delayed_ref_root *delayed_refs;
2189         struct btrfs_delayed_ref_head *head;
2190         int ret;
2191         int run_all = count == (unsigned long)-1;
2192
2193         /* We'll clean this up in btrfs_cleanup_transaction */
2194         if (trans->aborted)
2195                 return 0;
2196
2197         if (test_bit(BTRFS_FS_CREATING_FREE_SPACE_TREE, &fs_info->flags))
2198                 return 0;
2199
2200         delayed_refs = &trans->transaction->delayed_refs;
2201         if (count == 0)
2202                 count = atomic_read(&delayed_refs->num_entries) * 2;
2203
2204 again:
2205 #ifdef SCRAMBLE_DELAYED_REFS
2206         delayed_refs->run_delayed_start = find_middle(&delayed_refs->root);
2207 #endif
2208         ret = __btrfs_run_delayed_refs(trans, count);
2209         if (ret < 0) {
2210                 btrfs_abort_transaction(trans, ret);
2211                 return ret;
2212         }
2213
2214         if (run_all) {
2215                 btrfs_create_pending_block_groups(trans);
2216
2217                 spin_lock(&delayed_refs->lock);
2218                 node = rb_first_cached(&delayed_refs->href_root);
2219                 if (!node) {
2220                         spin_unlock(&delayed_refs->lock);
2221                         goto out;
2222                 }
2223                 head = rb_entry(node, struct btrfs_delayed_ref_head,
2224                                 href_node);
2225                 refcount_inc(&head->refs);
2226                 spin_unlock(&delayed_refs->lock);
2227
2228                 /* Mutex was contended, block until it's released and retry. */
2229                 mutex_lock(&head->mutex);
2230                 mutex_unlock(&head->mutex);
2231
2232                 btrfs_put_delayed_ref_head(head);
2233                 cond_resched();
2234                 goto again;
2235         }
2236 out:
2237         return 0;
2238 }
2239
2240 int btrfs_set_disk_extent_flags(struct btrfs_trans_handle *trans,
2241                                 u64 bytenr, u64 num_bytes, u64 flags,
2242                                 int level, int is_data)
2243 {
2244         struct btrfs_delayed_extent_op *extent_op;
2245         int ret;
2246
2247         extent_op = btrfs_alloc_delayed_extent_op();
2248         if (!extent_op)
2249                 return -ENOMEM;
2250
2251         extent_op->flags_to_set = flags;
2252         extent_op->update_flags = true;
2253         extent_op->update_key = false;
2254         extent_op->is_data = is_data ? true : false;
2255         extent_op->level = level;
2256
2257         ret = btrfs_add_delayed_extent_op(trans, bytenr, num_bytes, extent_op);
2258         if (ret)
2259                 btrfs_free_delayed_extent_op(extent_op);
2260         return ret;
2261 }
2262
2263 static noinline int check_delayed_ref(struct btrfs_root *root,
2264                                       struct btrfs_path *path,
2265                                       u64 objectid, u64 offset, u64 bytenr)
2266 {
2267         struct btrfs_delayed_ref_head *head;
2268         struct btrfs_delayed_ref_node *ref;
2269         struct btrfs_delayed_data_ref *data_ref;
2270         struct btrfs_delayed_ref_root *delayed_refs;
2271         struct btrfs_transaction *cur_trans;
2272         struct rb_node *node;
2273         int ret = 0;
2274
2275         spin_lock(&root->fs_info->trans_lock);
2276         cur_trans = root->fs_info->running_transaction;
2277         if (cur_trans)
2278                 refcount_inc(&cur_trans->use_count);
2279         spin_unlock(&root->fs_info->trans_lock);
2280         if (!cur_trans)
2281                 return 0;
2282
2283         delayed_refs = &cur_trans->delayed_refs;
2284         spin_lock(&delayed_refs->lock);
2285         head = btrfs_find_delayed_ref_head(delayed_refs, bytenr);
2286         if (!head) {
2287                 spin_unlock(&delayed_refs->lock);
2288                 btrfs_put_transaction(cur_trans);
2289                 return 0;
2290         }
2291
2292         if (!mutex_trylock(&head->mutex)) {
2293                 refcount_inc(&head->refs);
2294                 spin_unlock(&delayed_refs->lock);
2295
2296                 btrfs_release_path(path);
2297
2298                 /*
2299                  * Mutex was contended, block until it's released and let
2300                  * caller try again
2301                  */
2302                 mutex_lock(&head->mutex);
2303                 mutex_unlock(&head->mutex);
2304                 btrfs_put_delayed_ref_head(head);
2305                 btrfs_put_transaction(cur_trans);
2306                 return -EAGAIN;
2307         }
2308         spin_unlock(&delayed_refs->lock);
2309
2310         spin_lock(&head->lock);
2311         /*
2312          * XXX: We should replace this with a proper search function in the
2313          * future.
2314          */
2315         for (node = rb_first_cached(&head->ref_tree); node;
2316              node = rb_next(node)) {
2317                 ref = rb_entry(node, struct btrfs_delayed_ref_node, ref_node);
2318                 /* If it's a shared ref we know a cross reference exists */
2319                 if (ref->type != BTRFS_EXTENT_DATA_REF_KEY) {
2320                         ret = 1;
2321                         break;
2322                 }
2323
2324                 data_ref = btrfs_delayed_node_to_data_ref(ref);
2325
2326                 /*
2327                  * If our ref doesn't match the one we're currently looking at
2328                  * then we have a cross reference.
2329                  */
2330                 if (data_ref->root != root->root_key.objectid ||
2331                     data_ref->objectid != objectid ||
2332                     data_ref->offset != offset) {
2333                         ret = 1;
2334                         break;
2335                 }
2336         }
2337         spin_unlock(&head->lock);
2338         mutex_unlock(&head->mutex);
2339         btrfs_put_transaction(cur_trans);
2340         return ret;
2341 }
2342
2343 static noinline int check_committed_ref(struct btrfs_root *root,
2344                                         struct btrfs_path *path,
2345                                         u64 objectid, u64 offset, u64 bytenr)
2346 {
2347         struct btrfs_fs_info *fs_info = root->fs_info;
2348         struct btrfs_root *extent_root = fs_info->extent_root;
2349         struct extent_buffer *leaf;
2350         struct btrfs_extent_data_ref *ref;
2351         struct btrfs_extent_inline_ref *iref;
2352         struct btrfs_extent_item *ei;
2353         struct btrfs_key key;
2354         u32 item_size;
2355         int type;
2356         int ret;
2357
2358         key.objectid = bytenr;
2359         key.offset = (u64)-1;
2360         key.type = BTRFS_EXTENT_ITEM_KEY;
2361
2362         ret = btrfs_search_slot(NULL, extent_root, &key, path, 0, 0);
2363         if (ret < 0)
2364                 goto out;
2365         BUG_ON(ret == 0); /* Corruption */
2366
2367         ret = -ENOENT;
2368         if (path->slots[0] == 0)
2369                 goto out;
2370
2371         path->slots[0]--;
2372         leaf = path->nodes[0];
2373         btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
2374
2375         if (key.objectid != bytenr || key.type != BTRFS_EXTENT_ITEM_KEY)
2376                 goto out;
2377
2378         ret = 1;
2379         item_size = btrfs_item_size_nr(leaf, path->slots[0]);
2380         ei = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
2381
2382         /* If extent item has more than 1 inline ref then it's shared */
2383         if (item_size != sizeof(*ei) +
2384             btrfs_extent_inline_ref_size(BTRFS_EXTENT_DATA_REF_KEY))
2385                 goto out;
2386
2387         /* If extent created before last snapshot => it's definitely shared */
2388         if (btrfs_extent_generation(leaf, ei) <=
2389             btrfs_root_last_snapshot(&root->root_item))
2390                 goto out;
2391
2392         iref = (struct btrfs_extent_inline_ref *)(ei + 1);
2393
2394         /* If this extent has SHARED_DATA_REF then it's shared */
2395         type = btrfs_get_extent_inline_ref_type(leaf, iref, BTRFS_REF_TYPE_DATA);
2396         if (type != BTRFS_EXTENT_DATA_REF_KEY)
2397                 goto out;
2398
2399         ref = (struct btrfs_extent_data_ref *)(&iref->offset);
2400         if (btrfs_extent_refs(leaf, ei) !=
2401             btrfs_extent_data_ref_count(leaf, ref) ||
2402             btrfs_extent_data_ref_root(leaf, ref) !=
2403             root->root_key.objectid ||
2404             btrfs_extent_data_ref_objectid(leaf, ref) != objectid ||
2405             btrfs_extent_data_ref_offset(leaf, ref) != offset)
2406                 goto out;
2407
2408         ret = 0;
2409 out:
2410         return ret;
2411 }
2412
2413 int btrfs_cross_ref_exist(struct btrfs_root *root, u64 objectid, u64 offset,
2414                           u64 bytenr)
2415 {
2416         struct btrfs_path *path;
2417         int ret;
2418
2419         path = btrfs_alloc_path();
2420         if (!path)
2421                 return -ENOMEM;
2422
2423         do {
2424                 ret = check_committed_ref(root, path, objectid,
2425                                           offset, bytenr);
2426                 if (ret && ret != -ENOENT)
2427                         goto out;
2428
2429                 ret = check_delayed_ref(root, path, objectid, offset, bytenr);
2430         } while (ret == -EAGAIN);
2431
2432 out:
2433         btrfs_free_path(path);
2434         if (root->root_key.objectid == BTRFS_DATA_RELOC_TREE_OBJECTID)
2435                 WARN_ON(ret > 0);
2436         return ret;
2437 }
2438
2439 static int __btrfs_mod_ref(struct btrfs_trans_handle *trans,
2440                            struct btrfs_root *root,
2441                            struct extent_buffer *buf,
2442                            int full_backref, int inc)
2443 {
2444         struct btrfs_fs_info *fs_info = root->fs_info;
2445         u64 bytenr;
2446         u64 num_bytes;
2447         u64 parent;
2448         u64 ref_root;
2449         u32 nritems;
2450         struct btrfs_key key;
2451         struct btrfs_file_extent_item *fi;
2452         struct btrfs_ref generic_ref = { 0 };
2453         bool for_reloc = btrfs_header_flag(buf, BTRFS_HEADER_FLAG_RELOC);
2454         int i;
2455         int action;
2456         int level;
2457         int ret = 0;
2458
2459         if (btrfs_is_testing(fs_info))
2460                 return 0;
2461
2462         ref_root = btrfs_header_owner(buf);
2463         nritems = btrfs_header_nritems(buf);
2464         level = btrfs_header_level(buf);
2465
2466         if (!test_bit(BTRFS_ROOT_REF_COWS, &root->state) && level == 0)
2467                 return 0;
2468
2469         if (full_backref)
2470                 parent = buf->start;
2471         else
2472                 parent = 0;
2473         if (inc)
2474                 action = BTRFS_ADD_DELAYED_REF;
2475         else
2476                 action = BTRFS_DROP_DELAYED_REF;
2477
2478         for (i = 0; i < nritems; i++) {
2479                 if (level == 0) {
2480                         btrfs_item_key_to_cpu(buf, &key, i);
2481                         if (key.type != BTRFS_EXTENT_DATA_KEY)
2482                                 continue;
2483                         fi = btrfs_item_ptr(buf, i,
2484                                             struct btrfs_file_extent_item);
2485                         if (btrfs_file_extent_type(buf, fi) ==
2486                             BTRFS_FILE_EXTENT_INLINE)
2487                                 continue;
2488                         bytenr = btrfs_file_extent_disk_bytenr(buf, fi);
2489                         if (bytenr == 0)
2490                                 continue;
2491
2492                         num_bytes = btrfs_file_extent_disk_num_bytes(buf, fi);
2493                         key.offset -= btrfs_file_extent_offset(buf, fi);
2494                         btrfs_init_generic_ref(&generic_ref, action, bytenr,
2495                                                num_bytes, parent);
2496                         generic_ref.real_root = root->root_key.objectid;
2497                         btrfs_init_data_ref(&generic_ref, ref_root, key.objectid,
2498                                             key.offset);
2499                         generic_ref.skip_qgroup = for_reloc;
2500                         if (inc)
2501                                 ret = btrfs_inc_extent_ref(trans, &generic_ref);
2502                         else
2503                                 ret = btrfs_free_extent(trans, &generic_ref);
2504                         if (ret)
2505                                 goto fail;
2506                 } else {
2507                         bytenr = btrfs_node_blockptr(buf, i);
2508                         num_bytes = fs_info->nodesize;
2509                         btrfs_init_generic_ref(&generic_ref, action, bytenr,
2510                                                num_bytes, parent);
2511                         generic_ref.real_root = root->root_key.objectid;
2512                         btrfs_init_tree_ref(&generic_ref, level - 1, ref_root);
2513                         generic_ref.skip_qgroup = for_reloc;
2514                         if (inc)
2515                                 ret = btrfs_inc_extent_ref(trans, &generic_ref);
2516                         else
2517                                 ret = btrfs_free_extent(trans, &generic_ref);
2518                         if (ret)
2519                                 goto fail;
2520                 }
2521         }
2522         return 0;
2523 fail:
2524         return ret;
2525 }
2526
2527 int btrfs_inc_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root,
2528                   struct extent_buffer *buf, int full_backref)
2529 {
2530         return __btrfs_mod_ref(trans, root, buf, full_backref, 1);
2531 }
2532
2533 int btrfs_dec_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root,
2534                   struct extent_buffer *buf, int full_backref)
2535 {
2536         return __btrfs_mod_ref(trans, root, buf, full_backref, 0);
2537 }
2538
2539 int btrfs_extent_readonly(struct btrfs_fs_info *fs_info, u64 bytenr)
2540 {
2541         struct btrfs_block_group *block_group;
2542         int readonly = 0;
2543
2544         block_group = btrfs_lookup_block_group(fs_info, bytenr);
2545         if (!block_group || block_group->ro)
2546                 readonly = 1;
2547         if (block_group)
2548                 btrfs_put_block_group(block_group);
2549         return readonly;
2550 }
2551
2552 static u64 get_alloc_profile_by_root(struct btrfs_root *root, int data)
2553 {
2554         struct btrfs_fs_info *fs_info = root->fs_info;
2555         u64 flags;
2556         u64 ret;
2557
2558         if (data)
2559                 flags = BTRFS_BLOCK_GROUP_DATA;
2560         else if (root == fs_info->chunk_root)
2561                 flags = BTRFS_BLOCK_GROUP_SYSTEM;
2562         else
2563                 flags = BTRFS_BLOCK_GROUP_METADATA;
2564
2565         ret = btrfs_get_alloc_profile(fs_info, flags);
2566         return ret;
2567 }
2568
2569 static u64 first_logical_byte(struct btrfs_fs_info *fs_info, u64 search_start)
2570 {
2571         struct btrfs_block_group *cache;
2572         u64 bytenr;
2573
2574         spin_lock(&fs_info->block_group_cache_lock);
2575         bytenr = fs_info->first_logical_byte;
2576         spin_unlock(&fs_info->block_group_cache_lock);
2577
2578         if (bytenr < (u64)-1)
2579                 return bytenr;
2580
2581         cache = btrfs_lookup_first_block_group(fs_info, search_start);
2582         if (!cache)
2583                 return 0;
2584
2585         bytenr = cache->start;
2586         btrfs_put_block_group(cache);
2587
2588         return bytenr;
2589 }
2590
2591 static int pin_down_extent(struct btrfs_block_group *cache,
2592                            u64 bytenr, u64 num_bytes, int reserved)
2593 {
2594         struct btrfs_fs_info *fs_info = cache->fs_info;
2595
2596         spin_lock(&cache->space_info->lock);
2597         spin_lock(&cache->lock);
2598         cache->pinned += num_bytes;
2599         btrfs_space_info_update_bytes_pinned(fs_info, cache->space_info,
2600                                              num_bytes);
2601         if (reserved) {
2602                 cache->reserved -= num_bytes;
2603                 cache->space_info->bytes_reserved -= num_bytes;
2604         }
2605         spin_unlock(&cache->lock);
2606         spin_unlock(&cache->space_info->lock);
2607
2608         percpu_counter_add_batch(&cache->space_info->total_bytes_pinned,
2609                     num_bytes, BTRFS_TOTAL_BYTES_PINNED_BATCH);
2610         set_extent_dirty(fs_info->pinned_extents, bytenr,
2611                          bytenr + num_bytes - 1, GFP_NOFS | __GFP_NOFAIL);
2612         return 0;
2613 }
2614
2615 int btrfs_pin_extent(struct btrfs_fs_info *fs_info,
2616                      u64 bytenr, u64 num_bytes, int reserved)
2617 {
2618         struct btrfs_block_group *cache;
2619
2620         ASSERT(fs_info->running_transaction);
2621
2622         cache = btrfs_lookup_block_group(fs_info, bytenr);
2623         BUG_ON(!cache); /* Logic error */
2624
2625         pin_down_extent(cache, bytenr, num_bytes, reserved);
2626
2627         btrfs_put_block_group(cache);
2628         return 0;
2629 }
2630
2631 /*
2632  * this function must be called within transaction
2633  */
2634 int btrfs_pin_extent_for_log_replay(struct btrfs_fs_info *fs_info,
2635                                     u64 bytenr, u64 num_bytes)
2636 {
2637         struct btrfs_block_group *cache;
2638         int ret;
2639
2640         cache = btrfs_lookup_block_group(fs_info, bytenr);
2641         if (!cache)
2642                 return -EINVAL;
2643
2644         /*
2645          * pull in the free space cache (if any) so that our pin
2646          * removes the free space from the cache.  We have load_only set
2647          * to one because the slow code to read in the free extents does check
2648          * the pinned extents.
2649          */
2650         btrfs_cache_block_group(cache, 1);
2651
2652         pin_down_extent(cache, bytenr, num_bytes, 0);
2653
2654         /* remove us from the free space cache (if we're there at all) */
2655         ret = btrfs_remove_free_space(cache, bytenr, num_bytes);
2656         btrfs_put_block_group(cache);
2657         return ret;
2658 }
2659
2660 static int __exclude_logged_extent(struct btrfs_fs_info *fs_info,
2661                                    u64 start, u64 num_bytes)
2662 {
2663         int ret;
2664         struct btrfs_block_group *block_group;
2665         struct btrfs_caching_control *caching_ctl;
2666
2667         block_group = btrfs_lookup_block_group(fs_info, start);
2668         if (!block_group)
2669                 return -EINVAL;
2670
2671         btrfs_cache_block_group(block_group, 0);
2672         caching_ctl = btrfs_get_caching_control(block_group);
2673
2674         if (!caching_ctl) {
2675                 /* Logic error */
2676                 BUG_ON(!btrfs_block_group_done(block_group));
2677                 ret = btrfs_remove_free_space(block_group, start, num_bytes);
2678         } else {
2679                 mutex_lock(&caching_ctl->mutex);
2680
2681                 if (start >= caching_ctl->progress) {
2682                         ret = btrfs_add_excluded_extent(fs_info, start,
2683                                                         num_bytes);
2684                 } else if (start + num_bytes <= caching_ctl->progress) {
2685                         ret = btrfs_remove_free_space(block_group,
2686                                                       start, num_bytes);
2687                 } else {
2688                         num_bytes = caching_ctl->progress - start;
2689                         ret = btrfs_remove_free_space(block_group,
2690                                                       start, num_bytes);
2691                         if (ret)
2692                                 goto out_lock;
2693
2694                         num_bytes = (start + num_bytes) -
2695                                 caching_ctl->progress;
2696                         start = caching_ctl->progress;
2697                         ret = btrfs_add_excluded_extent(fs_info, start,
2698                                                         num_bytes);
2699                 }
2700 out_lock:
2701                 mutex_unlock(&caching_ctl->mutex);
2702                 btrfs_put_caching_control(caching_ctl);
2703         }
2704         btrfs_put_block_group(block_group);
2705         return ret;
2706 }
2707
2708 int btrfs_exclude_logged_extents(struct extent_buffer *eb)
2709 {
2710         struct btrfs_fs_info *fs_info = eb->fs_info;
2711         struct btrfs_file_extent_item *item;
2712         struct btrfs_key key;
2713         int found_type;
2714         int i;
2715         int ret = 0;
2716
2717         if (!btrfs_fs_incompat(fs_info, MIXED_GROUPS))
2718                 return 0;
2719
2720         for (i = 0; i < btrfs_header_nritems(eb); i++) {
2721                 btrfs_item_key_to_cpu(eb, &key, i);
2722                 if (key.type != BTRFS_EXTENT_DATA_KEY)
2723                         continue;
2724                 item = btrfs_item_ptr(eb, i, struct btrfs_file_extent_item);
2725                 found_type = btrfs_file_extent_type(eb, item);
2726                 if (found_type == BTRFS_FILE_EXTENT_INLINE)
2727                         continue;
2728                 if (btrfs_file_extent_disk_bytenr(eb, item) == 0)
2729                         continue;
2730                 key.objectid = btrfs_file_extent_disk_bytenr(eb, item);
2731                 key.offset = btrfs_file_extent_disk_num_bytes(eb, item);
2732                 ret = __exclude_logged_extent(fs_info, key.objectid, key.offset);
2733                 if (ret)
2734                         break;
2735         }
2736
2737         return ret;
2738 }
2739
2740 static void
2741 btrfs_inc_block_group_reservations(struct btrfs_block_group *bg)
2742 {
2743         atomic_inc(&bg->reservations);
2744 }
2745
2746 void btrfs_prepare_extent_commit(struct btrfs_fs_info *fs_info)
2747 {
2748         struct btrfs_caching_control *next;
2749         struct btrfs_caching_control *caching_ctl;
2750         struct btrfs_block_group *cache;
2751
2752         down_write(&fs_info->commit_root_sem);
2753
2754         list_for_each_entry_safe(caching_ctl, next,
2755                                  &fs_info->caching_block_groups, list) {
2756                 cache = caching_ctl->block_group;
2757                 if (btrfs_block_group_done(cache)) {
2758                         cache->last_byte_to_unpin = (u64)-1;
2759                         list_del_init(&caching_ctl->list);
2760                         btrfs_put_caching_control(caching_ctl);
2761                 } else {
2762                         cache->last_byte_to_unpin = caching_ctl->progress;
2763                 }
2764         }
2765
2766         if (fs_info->pinned_extents == &fs_info->freed_extents[0])
2767                 fs_info->pinned_extents = &fs_info->freed_extents[1];
2768         else
2769                 fs_info->pinned_extents = &fs_info->freed_extents[0];
2770
2771         up_write(&fs_info->commit_root_sem);
2772
2773         btrfs_update_global_block_rsv(fs_info);
2774 }
2775
2776 /*
2777  * Returns the free cluster for the given space info and sets empty_cluster to
2778  * what it should be based on the mount options.
2779  */
2780 static struct btrfs_free_cluster *
2781 fetch_cluster_info(struct btrfs_fs_info *fs_info,
2782                    struct btrfs_space_info *space_info, u64 *empty_cluster)
2783 {
2784         struct btrfs_free_cluster *ret = NULL;
2785
2786         *empty_cluster = 0;
2787         if (btrfs_mixed_space_info(space_info))
2788                 return ret;
2789
2790         if (space_info->flags & BTRFS_BLOCK_GROUP_METADATA) {
2791                 ret = &fs_info->meta_alloc_cluster;
2792                 if (btrfs_test_opt(fs_info, SSD))
2793                         *empty_cluster = SZ_2M;
2794                 else
2795                         *empty_cluster = SZ_64K;
2796         } else if ((space_info->flags & BTRFS_BLOCK_GROUP_DATA) &&
2797                    btrfs_test_opt(fs_info, SSD_SPREAD)) {
2798                 *empty_cluster = SZ_2M;
2799                 ret = &fs_info->data_alloc_cluster;
2800         }
2801
2802         return ret;
2803 }
2804
2805 static int unpin_extent_range(struct btrfs_fs_info *fs_info,
2806                               u64 start, u64 end,
2807                               const bool return_free_space)
2808 {
2809         struct btrfs_block_group *cache = NULL;
2810         struct btrfs_space_info *space_info;
2811         struct btrfs_block_rsv *global_rsv = &fs_info->global_block_rsv;
2812         struct btrfs_free_cluster *cluster = NULL;
2813         u64 len;
2814         u64 total_unpinned = 0;
2815         u64 empty_cluster = 0;
2816         bool readonly;
2817
2818         while (start <= end) {
2819                 readonly = false;
2820                 if (!cache ||
2821                     start >= cache->start + cache->length) {
2822                         if (cache)
2823                                 btrfs_put_block_group(cache);
2824                         total_unpinned = 0;
2825                         cache = btrfs_lookup_block_group(fs_info, start);
2826                         BUG_ON(!cache); /* Logic error */
2827
2828                         cluster = fetch_cluster_info(fs_info,
2829                                                      cache->space_info,
2830                                                      &empty_cluster);
2831                         empty_cluster <<= 1;
2832                 }
2833
2834                 len = cache->start + cache->length - start;
2835                 len = min(len, end + 1 - start);
2836
2837                 if (start < cache->last_byte_to_unpin) {
2838                         len = min(len, cache->last_byte_to_unpin - start);
2839                         if (return_free_space)
2840                                 btrfs_add_free_space(cache, start, len);
2841                 }
2842
2843                 start += len;
2844                 total_unpinned += len;
2845                 space_info = cache->space_info;
2846
2847                 /*
2848                  * If this space cluster has been marked as fragmented and we've
2849                  * unpinned enough in this block group to potentially allow a
2850                  * cluster to be created inside of it go ahead and clear the
2851                  * fragmented check.
2852                  */
2853                 if (cluster && cluster->fragmented &&
2854                     total_unpinned > empty_cluster) {
2855                         spin_lock(&cluster->lock);
2856                         cluster->fragmented = 0;
2857                         spin_unlock(&cluster->lock);
2858                 }
2859
2860                 spin_lock(&space_info->lock);
2861                 spin_lock(&cache->lock);
2862                 cache->pinned -= len;
2863                 btrfs_space_info_update_bytes_pinned(fs_info, space_info, -len);
2864                 space_info->max_extent_size = 0;
2865                 percpu_counter_add_batch(&space_info->total_bytes_pinned,
2866                             -len, BTRFS_TOTAL_BYTES_PINNED_BATCH);
2867                 if (cache->ro) {
2868                         space_info->bytes_readonly += len;
2869                         readonly = true;
2870                 }
2871                 spin_unlock(&cache->lock);
2872                 if (!readonly && return_free_space &&
2873                     global_rsv->space_info == space_info) {
2874                         u64 to_add = len;
2875
2876                         spin_lock(&global_rsv->lock);
2877                         if (!global_rsv->full) {
2878                                 to_add = min(len, global_rsv->size -
2879                                              global_rsv->reserved);
2880                                 global_rsv->reserved += to_add;
2881                                 btrfs_space_info_update_bytes_may_use(fs_info,
2882                                                 space_info, to_add);
2883                                 if (global_rsv->reserved >= global_rsv->size)
2884                                         global_rsv->full = 1;
2885                                 len -= to_add;
2886                         }
2887                         spin_unlock(&global_rsv->lock);
2888                         /* Add to any tickets we may have */
2889                         if (len)
2890                                 btrfs_try_granting_tickets(fs_info,
2891                                                            space_info);
2892                 }
2893                 spin_unlock(&space_info->lock);
2894         }
2895
2896         if (cache)
2897                 btrfs_put_block_group(cache);
2898         return 0;
2899 }
2900
2901 int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans)
2902 {
2903         struct btrfs_fs_info *fs_info = trans->fs_info;
2904         struct btrfs_block_group *block_group, *tmp;
2905         struct list_head *deleted_bgs;
2906         struct extent_io_tree *unpin;
2907         u64 start;
2908         u64 end;
2909         int ret;
2910
2911         if (fs_info->pinned_extents == &fs_info->freed_extents[0])
2912                 unpin = &fs_info->freed_extents[1];
2913         else
2914                 unpin = &fs_info->freed_extents[0];
2915
2916         while (!trans->aborted) {
2917                 struct extent_state *cached_state = NULL;
2918
2919                 mutex_lock(&fs_info->unused_bg_unpin_mutex);
2920                 ret = find_first_extent_bit(unpin, 0, &start, &end,
2921                                             EXTENT_DIRTY, &cached_state);
2922                 if (ret) {
2923                         mutex_unlock(&fs_info->unused_bg_unpin_mutex);
2924                         break;
2925                 }
2926
2927                 if (btrfs_test_opt(fs_info, DISCARD_SYNC))
2928                         ret = btrfs_discard_extent(fs_info, start,
2929                                                    end + 1 - start, NULL);
2930
2931                 clear_extent_dirty(unpin, start, end, &cached_state);
2932                 unpin_extent_range(fs_info, start, end, true);
2933                 mutex_unlock(&fs_info->unused_bg_unpin_mutex);
2934                 free_extent_state(cached_state);
2935                 cond_resched();
2936         }
2937
2938         if (btrfs_test_opt(fs_info, DISCARD_ASYNC)) {
2939                 btrfs_discard_calc_delay(&fs_info->discard_ctl);
2940                 btrfs_discard_schedule_work(&fs_info->discard_ctl, true);
2941         }
2942
2943         /*
2944          * Transaction is finished.  We don't need the lock anymore.  We
2945          * do need to clean up the block groups in case of a transaction
2946          * abort.
2947          */
2948         deleted_bgs = &trans->transaction->deleted_bgs;
2949         list_for_each_entry_safe(block_group, tmp, deleted_bgs, bg_list) {
2950                 u64 trimmed = 0;
2951
2952                 ret = -EROFS;
2953                 if (!trans->aborted)
2954                         ret = btrfs_discard_extent(fs_info,
2955                                                    block_group->start,
2956                                                    block_group->length,
2957                                                    &trimmed);
2958
2959                 list_del_init(&block_group->bg_list);
2960                 btrfs_put_block_group_trimming(block_group);
2961                 btrfs_put_block_group(block_group);
2962
2963                 if (ret) {
2964                         const char *errstr = btrfs_decode_error(ret);
2965                         btrfs_warn(fs_info,
2966                            "discard failed while removing blockgroup: errno=%d %s",
2967                                    ret, errstr);
2968                 }
2969         }
2970
2971         return 0;
2972 }
2973
2974 static int __btrfs_free_extent(struct btrfs_trans_handle *trans,
2975                                struct btrfs_delayed_ref_node *node, u64 parent,
2976                                u64 root_objectid, u64 owner_objectid,
2977                                u64 owner_offset, int refs_to_drop,
2978                                struct btrfs_delayed_extent_op *extent_op)
2979 {
2980         struct btrfs_fs_info *info = trans->fs_info;
2981         struct btrfs_key key;
2982         struct btrfs_path *path;
2983         struct btrfs_root *extent_root = info->extent_root;
2984         struct extent_buffer *leaf;
2985         struct btrfs_extent_item *ei;
2986         struct btrfs_extent_inline_ref *iref;
2987         int ret;
2988         int is_data;
2989         int extent_slot = 0;
2990         int found_extent = 0;
2991         int num_to_del = 1;
2992         u32 item_size;
2993         u64 refs;
2994         u64 bytenr = node->bytenr;
2995         u64 num_bytes = node->num_bytes;
2996         int last_ref = 0;
2997         bool skinny_metadata = btrfs_fs_incompat(info, SKINNY_METADATA);
2998
2999         path = btrfs_alloc_path();
3000         if (!path)
3001                 return -ENOMEM;
3002
3003         path->reada = READA_FORWARD;
3004         path->leave_spinning = 1;
3005
3006         is_data = owner_objectid >= BTRFS_FIRST_FREE_OBJECTID;
3007         BUG_ON(!is_data && refs_to_drop != 1);
3008
3009         if (is_data)
3010                 skinny_metadata = false;
3011
3012         ret = lookup_extent_backref(trans, path, &iref, bytenr, num_bytes,
3013                                     parent, root_objectid, owner_objectid,
3014                                     owner_offset);
3015         if (ret == 0) {
3016                 extent_slot = path->slots[0];
3017                 while (extent_slot >= 0) {
3018                         btrfs_item_key_to_cpu(path->nodes[0], &key,
3019                                               extent_slot);
3020                         if (key.objectid != bytenr)
3021                                 break;
3022                         if (key.type == BTRFS_EXTENT_ITEM_KEY &&
3023                             key.offset == num_bytes) {
3024                                 found_extent = 1;
3025                                 break;
3026                         }
3027                         if (key.type == BTRFS_METADATA_ITEM_KEY &&
3028                             key.offset == owner_objectid) {
3029                                 found_extent = 1;
3030                                 break;
3031                         }
3032                         if (path->slots[0] - extent_slot > 5)
3033                                 break;
3034                         extent_slot--;
3035                 }
3036
3037                 if (!found_extent) {
3038                         BUG_ON(iref);
3039                         ret = remove_extent_backref(trans, path, NULL,
3040                                                     refs_to_drop,
3041                                                     is_data, &last_ref);
3042                         if (ret) {
3043                                 btrfs_abort_transaction(trans, ret);
3044                                 goto out;
3045                         }
3046                         btrfs_release_path(path);
3047                         path->leave_spinning = 1;
3048
3049                         key.objectid = bytenr;
3050                         key.type = BTRFS_EXTENT_ITEM_KEY;
3051                         key.offset = num_bytes;
3052
3053                         if (!is_data && skinny_metadata) {
3054                                 key.type = BTRFS_METADATA_ITEM_KEY;
3055                                 key.offset = owner_objectid;
3056                         }
3057
3058                         ret = btrfs_search_slot(trans, extent_root,
3059                                                 &key, path, -1, 1);
3060                         if (ret > 0 && skinny_metadata && path->slots[0]) {
3061                                 /*
3062                                  * Couldn't find our skinny metadata item,
3063                                  * see if we have ye olde extent item.
3064                                  */
3065                                 path->slots[0]--;
3066                                 btrfs_item_key_to_cpu(path->nodes[0], &key,
3067                                                       path->slots[0]);
3068                                 if (key.objectid == bytenr &&
3069                                     key.type == BTRFS_EXTENT_ITEM_KEY &&
3070                                     key.offset == num_bytes)
3071                                         ret = 0;
3072                         }
3073
3074                         if (ret > 0 && skinny_metadata) {
3075                                 skinny_metadata = false;
3076                                 key.objectid = bytenr;
3077                                 key.type = BTRFS_EXTENT_ITEM_KEY;
3078                                 key.offset = num_bytes;
3079                                 btrfs_release_path(path);
3080                                 ret = btrfs_search_slot(trans, extent_root,
3081                                                         &key, path, -1, 1);
3082                         }
3083
3084                         if (ret) {
3085                                 btrfs_err(info,
3086                                           "umm, got %d back from search, was looking for %llu",
3087                                           ret, bytenr);
3088                                 if (ret > 0)
3089                                         btrfs_print_leaf(path->nodes[0]);
3090                         }
3091                         if (ret < 0) {
3092                                 btrfs_abort_transaction(trans, ret);
3093                                 goto out;
3094                         }
3095                         extent_slot = path->slots[0];
3096                 }
3097         } else if (WARN_ON(ret == -ENOENT)) {
3098                 btrfs_print_leaf(path->nodes[0]);
3099                 btrfs_err(info,
3100                         "unable to find ref byte nr %llu parent %llu root %llu  owner %llu offset %llu",
3101                         bytenr, parent, root_objectid, owner_objectid,
3102                         owner_offset);
3103                 btrfs_abort_transaction(trans, ret);
3104                 goto out;
3105         } else {
3106                 btrfs_abort_transaction(trans, ret);
3107                 goto out;
3108         }
3109
3110         leaf = path->nodes[0];
3111         item_size = btrfs_item_size_nr(leaf, extent_slot);
3112         if (unlikely(item_size < sizeof(*ei))) {
3113                 ret = -EINVAL;
3114                 btrfs_print_v0_err(info);
3115                 btrfs_abort_transaction(trans, ret);
3116                 goto out;
3117         }
3118         ei = btrfs_item_ptr(leaf, extent_slot,
3119                             struct btrfs_extent_item);
3120         if (owner_objectid < BTRFS_FIRST_FREE_OBJECTID &&
3121             key.type == BTRFS_EXTENT_ITEM_KEY) {
3122                 struct btrfs_tree_block_info *bi;
3123                 BUG_ON(item_size < sizeof(*ei) + sizeof(*bi));
3124                 bi = (struct btrfs_tree_block_info *)(ei + 1);
3125                 WARN_ON(owner_objectid != btrfs_tree_block_level(leaf, bi));
3126         }
3127
3128         refs = btrfs_extent_refs(leaf, ei);
3129         if (refs < refs_to_drop) {
3130                 btrfs_err(info,
3131                           "trying to drop %d refs but we only have %Lu for bytenr %Lu",
3132                           refs_to_drop, refs, bytenr);
3133                 ret = -EINVAL;
3134                 btrfs_abort_transaction(trans, ret);
3135                 goto out;
3136         }
3137         refs -= refs_to_drop;
3138
3139         if (refs > 0) {
3140                 if (extent_op)
3141                         __run_delayed_extent_op(extent_op, leaf, ei);
3142                 /*
3143                  * In the case of inline back ref, reference count will
3144                  * be updated by remove_extent_backref
3145                  */
3146                 if (iref) {
3147                         BUG_ON(!found_extent);
3148                 } else {
3149                         btrfs_set_extent_refs(leaf, ei, refs);
3150                         btrfs_mark_buffer_dirty(leaf);
3151                 }
3152                 if (found_extent) {
3153                         ret = remove_extent_backref(trans, path, iref,
3154                                                     refs_to_drop, is_data,
3155                                                     &last_ref);
3156                         if (ret) {
3157                                 btrfs_abort_transaction(trans, ret);
3158                                 goto out;
3159                         }
3160                 }
3161         } else {
3162                 if (found_extent) {
3163                         BUG_ON(is_data && refs_to_drop !=
3164                                extent_data_ref_count(path, iref));
3165                         if (iref) {
3166                                 BUG_ON(path->slots[0] != extent_slot);
3167                         } else {
3168                                 BUG_ON(path->slots[0] != extent_slot + 1);
3169                                 path->slots[0] = extent_slot;
3170                                 num_to_del = 2;
3171                         }
3172                 }
3173
3174                 last_ref = 1;
3175                 ret = btrfs_del_items(trans, extent_root, path, path->slots[0],
3176                                       num_to_del);
3177                 if (ret) {
3178                         btrfs_abort_transaction(trans, ret);
3179                         goto out;
3180                 }
3181                 btrfs_release_path(path);
3182
3183                 if (is_data) {
3184                         ret = btrfs_del_csums(trans, info->csum_root, bytenr,
3185                                               num_bytes);
3186                         if (ret) {
3187                                 btrfs_abort_transaction(trans, ret);
3188                                 goto out;
3189                         }
3190                 }
3191
3192                 ret = add_to_free_space_tree(trans, bytenr, num_bytes);
3193                 if (ret) {
3194                         btrfs_abort_transaction(trans, ret);
3195                         goto out;
3196                 }
3197
3198                 ret = btrfs_update_block_group(trans, bytenr, num_bytes, 0);
3199                 if (ret) {
3200                         btrfs_abort_transaction(trans, ret);
3201                         goto out;
3202                 }
3203         }
3204         btrfs_release_path(path);
3205
3206 out:
3207         btrfs_free_path(path);
3208         return ret;
3209 }
3210
3211 /*
3212  * when we free an block, it is possible (and likely) that we free the last
3213  * delayed ref for that extent as well.  This searches the delayed ref tree for
3214  * a given extent, and if there are no other delayed refs to be processed, it
3215  * removes it from the tree.
3216  */
3217 static noinline int check_ref_cleanup(struct btrfs_trans_handle *trans,
3218                                       u64 bytenr)
3219 {
3220         struct btrfs_delayed_ref_head *head;
3221         struct btrfs_delayed_ref_root *delayed_refs;
3222         int ret = 0;
3223
3224         delayed_refs = &trans->transaction->delayed_refs;
3225         spin_lock(&delayed_refs->lock);
3226         head = btrfs_find_delayed_ref_head(delayed_refs, bytenr);
3227         if (!head)
3228                 goto out_delayed_unlock;
3229
3230         spin_lock(&head->lock);
3231         if (!RB_EMPTY_ROOT(&head->ref_tree.rb_root))
3232                 goto out;
3233
3234         if (cleanup_extent_op(head) != NULL)
3235                 goto out;
3236
3237         /*
3238          * waiting for the lock here would deadlock.  If someone else has it
3239          * locked they are already in the process of dropping it anyway
3240          */
3241         if (!mutex_trylock(&head->mutex))
3242                 goto out;
3243
3244         btrfs_delete_ref_head(delayed_refs, head);
3245         head->processing = 0;
3246
3247         spin_unlock(&head->lock);
3248         spin_unlock(&delayed_refs->lock);
3249
3250         BUG_ON(head->extent_op);
3251         if (head->must_insert_reserved)
3252                 ret = 1;
3253
3254         btrfs_cleanup_ref_head_accounting(trans->fs_info, delayed_refs, head);
3255         mutex_unlock(&head->mutex);
3256         btrfs_put_delayed_ref_head(head);
3257         return ret;
3258 out:
3259         spin_unlock(&head->lock);
3260
3261 out_delayed_unlock:
3262         spin_unlock(&delayed_refs->lock);
3263         return 0;
3264 }
3265
3266 void btrfs_free_tree_block(struct btrfs_trans_handle *trans,
3267                            struct btrfs_root *root,
3268                            struct extent_buffer *buf,
3269                            u64 parent, int last_ref)
3270 {
3271         struct btrfs_fs_info *fs_info = root->fs_info;
3272         struct btrfs_ref generic_ref = { 0 };
3273         int pin = 1;
3274         int ret;
3275
3276         btrfs_init_generic_ref(&generic_ref, BTRFS_DROP_DELAYED_REF,
3277                                buf->start, buf->len, parent);
3278         btrfs_init_tree_ref(&generic_ref, btrfs_header_level(buf),
3279                             root->root_key.objectid);
3280
3281         if (root->root_key.objectid != BTRFS_TREE_LOG_OBJECTID) {
3282                 int old_ref_mod, new_ref_mod;
3283
3284                 btrfs_ref_tree_mod(fs_info, &generic_ref);
3285                 ret = btrfs_add_delayed_tree_ref(trans, &generic_ref, NULL,
3286                                                  &old_ref_mod, &new_ref_mod);
3287                 BUG_ON(ret); /* -ENOMEM */
3288                 pin = old_ref_mod >= 0 && new_ref_mod < 0;
3289         }
3290
3291         if (last_ref && btrfs_header_generation(buf) == trans->transid) {
3292                 struct btrfs_block_group *cache;
3293
3294                 if (root->root_key.objectid != BTRFS_TREE_LOG_OBJECTID) {
3295                         ret = check_ref_cleanup(trans, buf->start);
3296                         if (!ret)
3297                                 goto out;
3298                 }
3299
3300                 pin = 0;
3301                 cache = btrfs_lookup_block_group(fs_info, buf->start);
3302
3303                 if (btrfs_header_flag(buf, BTRFS_HEADER_FLAG_WRITTEN)) {
3304                         pin_down_extent(cache, buf->start, buf->len, 1);
3305                         btrfs_put_block_group(cache);
3306                         goto out;
3307                 }
3308
3309                 WARN_ON(test_bit(EXTENT_BUFFER_DIRTY, &buf->bflags));
3310
3311                 btrfs_add_free_space(cache, buf->start, buf->len);
3312                 btrfs_free_reserved_bytes(cache, buf->len, 0);
3313                 btrfs_put_block_group(cache);
3314                 trace_btrfs_reserved_extent_free(fs_info, buf->start, buf->len);
3315         }
3316 out:
3317         if (pin)
3318                 add_pinned_bytes(fs_info, &generic_ref);
3319
3320         if (last_ref) {
3321                 /*
3322                  * Deleting the buffer, clear the corrupt flag since it doesn't
3323                  * matter anymore.
3324                  */
3325                 clear_bit(EXTENT_BUFFER_CORRUPT, &buf->bflags);
3326         }
3327 }
3328
3329 /* Can return -ENOMEM */
3330 int btrfs_free_extent(struct btrfs_trans_handle *trans, struct btrfs_ref *ref)
3331 {
3332         struct btrfs_fs_info *fs_info = trans->fs_info;
3333         int old_ref_mod, new_ref_mod;
3334         int ret;
3335
3336         if (btrfs_is_testing(fs_info))
3337                 return 0;
3338
3339         /*
3340          * tree log blocks never actually go into the extent allocation
3341          * tree, just update pinning info and exit early.
3342          */
3343         if ((ref->type == BTRFS_REF_METADATA &&
3344              ref->tree_ref.root == BTRFS_TREE_LOG_OBJECTID) ||
3345             (ref->type == BTRFS_REF_DATA &&
3346              ref->data_ref.ref_root == BTRFS_TREE_LOG_OBJECTID)) {
3347                 /* unlocks the pinned mutex */
3348                 btrfs_pin_extent(fs_info, ref->bytenr, ref->len, 1);
3349                 old_ref_mod = new_ref_mod = 0;
3350                 ret = 0;
3351         } else if (ref->type == BTRFS_REF_METADATA) {
3352                 ret = btrfs_add_delayed_tree_ref(trans, ref, NULL,
3353                                                  &old_ref_mod, &new_ref_mod);
3354         } else {
3355                 ret = btrfs_add_delayed_data_ref(trans, ref, 0,
3356                                                  &old_ref_mod, &new_ref_mod);
3357         }
3358
3359         if (!((ref->type == BTRFS_REF_METADATA &&
3360                ref->tree_ref.root == BTRFS_TREE_LOG_OBJECTID) ||
3361               (ref->type == BTRFS_REF_DATA &&
3362                ref->data_ref.ref_root == BTRFS_TREE_LOG_OBJECTID)))
3363                 btrfs_ref_tree_mod(fs_info, ref);
3364
3365         if (ret == 0 && old_ref_mod >= 0 && new_ref_mod < 0)
3366                 add_pinned_bytes(fs_info, ref);
3367
3368         return ret;
3369 }
3370
3371 enum btrfs_loop_type {
3372         LOOP_CACHING_NOWAIT,
3373         LOOP_CACHING_WAIT,
3374         LOOP_ALLOC_CHUNK,
3375         LOOP_NO_EMPTY_SIZE,
3376 };
3377
3378 static inline void
3379 btrfs_lock_block_group(struct btrfs_block_group *cache,
3380                        int delalloc)
3381 {
3382         if (delalloc)
3383                 down_read(&cache->data_rwsem);
3384 }
3385
3386 static inline void btrfs_grab_block_group(struct btrfs_block_group *cache,
3387                        int delalloc)
3388 {
3389         btrfs_get_block_group(cache);
3390         if (delalloc)
3391                 down_read(&cache->data_rwsem);
3392 }
3393
3394 static struct btrfs_block_group *btrfs_lock_cluster(
3395                    struct btrfs_block_group *block_group,
3396                    struct btrfs_free_cluster *cluster,
3397                    int delalloc)
3398 {
3399         struct btrfs_block_group *used_bg = NULL;
3400
3401         spin_lock(&cluster->refill_lock);
3402         while (1) {
3403                 used_bg = cluster->block_group;
3404                 if (!used_bg)
3405                         return NULL;
3406
3407                 if (used_bg == block_group)
3408                         return used_bg;
3409
3410                 btrfs_get_block_group(used_bg);
3411
3412                 if (!delalloc)
3413                         return used_bg;
3414
3415                 if (down_read_trylock(&used_bg->data_rwsem))
3416                         return used_bg;
3417
3418                 spin_unlock(&cluster->refill_lock);
3419
3420                 /* We should only have one-level nested. */
3421                 down_read_nested(&used_bg->data_rwsem, SINGLE_DEPTH_NESTING);
3422
3423                 spin_lock(&cluster->refill_lock);
3424                 if (used_bg == cluster->block_group)
3425                         return used_bg;
3426
3427                 up_read(&used_bg->data_rwsem);
3428                 btrfs_put_block_group(used_bg);
3429         }
3430 }
3431
3432 static inline void
3433 btrfs_release_block_group(struct btrfs_block_group *cache,
3434                          int delalloc)
3435 {
3436         if (delalloc)
3437                 up_read(&cache->data_rwsem);
3438         btrfs_put_block_group(cache);
3439 }
3440
3441 /*
3442  * Structure used internally for find_free_extent() function.  Wraps needed
3443  * parameters.
3444  */
3445 struct find_free_extent_ctl {
3446         /* Basic allocation info */
3447         u64 num_bytes;
3448         u64 empty_size;
3449         u64 flags;
3450         int delalloc;
3451
3452         /* Where to start the search inside the bg */
3453         u64 search_start;
3454
3455         /* For clustered allocation */
3456         u64 empty_cluster;
3457
3458         bool have_caching_bg;
3459         bool orig_have_caching_bg;
3460
3461         /* RAID index, converted from flags */
3462         int index;
3463
3464         /*
3465          * Current loop number, check find_free_extent_update_loop() for details
3466          */
3467         int loop;
3468
3469         /*
3470          * Whether we're refilling a cluster, if true we need to re-search
3471          * current block group but don't try to refill the cluster again.
3472          */
3473         bool retry_clustered;
3474
3475         /*
3476          * Whether we're updating free space cache, if true we need to re-search
3477          * current block group but don't try updating free space cache again.
3478          */
3479         bool retry_unclustered;
3480
3481         /* If current block group is cached */
3482         int cached;
3483
3484         /* Max contiguous hole found */
3485         u64 max_extent_size;
3486
3487         /* Total free space from free space cache, not always contiguous */
3488         u64 total_free_space;
3489
3490         /* Found result */
3491         u64 found_offset;
3492 };
3493
3494
3495 /*
3496  * Helper function for find_free_extent().
3497  *
3498  * Return -ENOENT to inform caller that we need fallback to unclustered mode.
3499  * Return -EAGAIN to inform caller that we need to re-search this block group
3500  * Return >0 to inform caller that we find nothing
3501  * Return 0 means we have found a location and set ffe_ctl->found_offset.
3502  */
3503 static int find_free_extent_clustered(struct btrfs_block_group *bg,
3504                 struct btrfs_free_cluster *last_ptr,
3505                 struct find_free_extent_ctl *ffe_ctl,
3506                 struct btrfs_block_group **cluster_bg_ret)
3507 {
3508         struct btrfs_block_group *cluster_bg;
3509         u64 aligned_cluster;
3510         u64 offset;
3511         int ret;
3512
3513         cluster_bg = btrfs_lock_cluster(bg, last_ptr, ffe_ctl->delalloc);
3514         if (!cluster_bg)
3515                 goto refill_cluster;
3516         if (cluster_bg != bg && (cluster_bg->ro ||
3517             !block_group_bits(cluster_bg, ffe_ctl->flags)))
3518                 goto release_cluster;
3519
3520         offset = btrfs_alloc_from_cluster(cluster_bg, last_ptr,
3521                         ffe_ctl->num_bytes, cluster_bg->start,
3522                         &ffe_ctl->max_extent_size);
3523         if (offset) {
3524                 /* We have a block, we're done */
3525                 spin_unlock(&last_ptr->refill_lock);
3526                 trace_btrfs_reserve_extent_cluster(cluster_bg,
3527                                 ffe_ctl->search_start, ffe_ctl->num_bytes);
3528                 *cluster_bg_ret = cluster_bg;
3529                 ffe_ctl->found_offset = offset;
3530                 return 0;
3531         }
3532         WARN_ON(last_ptr->block_group != cluster_bg);
3533
3534 release_cluster:
3535         /*
3536          * If we are on LOOP_NO_EMPTY_SIZE, we can't set up a new clusters, so
3537          * lets just skip it and let the allocator find whatever block it can
3538          * find. If we reach this point, we will have tried the cluster
3539          * allocator plenty of times and not have found anything, so we are
3540          * likely way too fragmented for the clustering stuff to find anything.
3541          *
3542          * However, if the cluster is taken from the current block group,
3543          * release the cluster first, so that we stand a better chance of
3544          * succeeding in the unclustered allocation.
3545          */
3546         if (ffe_ctl->loop >= LOOP_NO_EMPTY_SIZE && cluster_bg != bg) {
3547                 spin_unlock(&last_ptr->refill_lock);
3548                 btrfs_release_block_group(cluster_bg, ffe_ctl->delalloc);
3549                 return -ENOENT;
3550         }
3551
3552         /* This cluster didn't work out, free it and start over */
3553         btrfs_return_cluster_to_free_space(NULL, last_ptr);
3554
3555         if (cluster_bg != bg)
3556                 btrfs_release_block_group(cluster_bg, ffe_ctl->delalloc);
3557
3558 refill_cluster:
3559         if (ffe_ctl->loop >= LOOP_NO_EMPTY_SIZE) {
3560                 spin_unlock(&last_ptr->refill_lock);
3561                 return -ENOENT;
3562         }
3563
3564         aligned_cluster = max_t(u64,
3565                         ffe_ctl->empty_cluster + ffe_ctl->empty_size,
3566                         bg->full_stripe_len);
3567         ret = btrfs_find_space_cluster(bg, last_ptr, ffe_ctl->search_start,
3568                         ffe_ctl->num_bytes, aligned_cluster);
3569         if (ret == 0) {
3570                 /* Now pull our allocation out of this cluster */
3571                 offset = btrfs_alloc_from_cluster(bg, last_ptr,
3572                                 ffe_ctl->num_bytes, ffe_ctl->search_start,
3573                                 &ffe_ctl->max_extent_size);
3574                 if (offset) {
3575                         /* We found one, proceed */
3576                         spin_unlock(&last_ptr->refill_lock);
3577                         trace_btrfs_reserve_extent_cluster(bg,
3578                                         ffe_ctl->search_start,
3579                                         ffe_ctl->num_bytes);
3580                         ffe_ctl->found_offset = offset;
3581                         return 0;
3582                 }
3583         } else if (!ffe_ctl->cached && ffe_ctl->loop > LOOP_CACHING_NOWAIT &&
3584                    !ffe_ctl->retry_clustered) {
3585                 spin_unlock(&last_ptr->refill_lock);
3586
3587                 ffe_ctl->retry_clustered = true;
3588                 btrfs_wait_block_group_cache_progress(bg, ffe_ctl->num_bytes +
3589                                 ffe_ctl->empty_cluster + ffe_ctl->empty_size);
3590                 return -EAGAIN;
3591         }
3592         /*
3593          * At this point we either didn't find a cluster or we weren't able to
3594          * allocate a block from our cluster.  Free the cluster we've been
3595          * trying to use, and go to the next block group.
3596          */
3597         btrfs_return_cluster_to_free_space(NULL, last_ptr);
3598         spin_unlock(&last_ptr->refill_lock);
3599         return 1;
3600 }
3601
3602 /*
3603  * Return >0 to inform caller that we find nothing
3604  * Return 0 when we found an free extent and set ffe_ctrl->found_offset
3605  * Return -EAGAIN to inform caller that we need to re-search this block group
3606  */
3607 static int find_free_extent_unclustered(struct btrfs_block_group *bg,
3608                 struct btrfs_free_cluster *last_ptr,
3609                 struct find_free_extent_ctl *ffe_ctl)
3610 {
3611         u64 offset;
3612
3613         /*
3614          * We are doing an unclustered allocation, set the fragmented flag so
3615          * we don't bother trying to setup a cluster again until we get more
3616          * space.
3617          */
3618         if (unlikely(last_ptr)) {
3619                 spin_lock(&last_ptr->lock);
3620                 last_ptr->fragmented = 1;
3621                 spin_unlock(&last_ptr->lock);
3622         }
3623         if (ffe_ctl->cached) {
3624                 struct btrfs_free_space_ctl *free_space_ctl;
3625
3626                 free_space_ctl = bg->free_space_ctl;
3627                 spin_lock(&free_space_ctl->tree_lock);
3628                 if (free_space_ctl->free_space <
3629                     ffe_ctl->num_bytes + ffe_ctl->empty_cluster +
3630                     ffe_ctl->empty_size) {
3631                         ffe_ctl->total_free_space = max_t(u64,
3632                                         ffe_ctl->total_free_space,
3633                                         free_space_ctl->free_space);
3634                         spin_unlock(&free_space_ctl->tree_lock);
3635                         return 1;
3636                 }
3637                 spin_unlock(&free_space_ctl->tree_lock);
3638         }
3639
3640         offset = btrfs_find_space_for_alloc(bg, ffe_ctl->search_start,
3641                         ffe_ctl->num_bytes, ffe_ctl->empty_size,
3642                         &ffe_ctl->max_extent_size);
3643
3644         /*
3645          * If we didn't find a chunk, and we haven't failed on this block group
3646          * before, and this block group is in the middle of caching and we are
3647          * ok with waiting, then go ahead and wait for progress to be made, and
3648          * set @retry_unclustered to true.
3649          *
3650          * If @retry_unclustered is true then we've already waited on this
3651          * block group once and should move on to the next block group.
3652          */
3653         if (!offset && !ffe_ctl->retry_unclustered && !ffe_ctl->cached &&
3654             ffe_ctl->loop > LOOP_CACHING_NOWAIT) {
3655                 btrfs_wait_block_group_cache_progress(bg, ffe_ctl->num_bytes +
3656                                                       ffe_ctl->empty_size);
3657                 ffe_ctl->retry_unclustered = true;
3658                 return -EAGAIN;
3659         } else if (!offset) {
3660                 return 1;
3661         }
3662         ffe_ctl->found_offset = offset;
3663         return 0;
3664 }
3665
3666 /*
3667  * Return >0 means caller needs to re-search for free extent
3668  * Return 0 means we have the needed free extent.
3669  * Return <0 means we failed to locate any free extent.
3670  */
3671 static int find_free_extent_update_loop(struct btrfs_fs_info *fs_info,
3672                                         struct btrfs_free_cluster *last_ptr,
3673                                         struct btrfs_key *ins,
3674                                         struct find_free_extent_ctl *ffe_ctl,
3675                                         int full_search, bool use_cluster)
3676 {
3677         struct btrfs_root *root = fs_info->extent_root;
3678         int ret;
3679
3680         if ((ffe_ctl->loop == LOOP_CACHING_NOWAIT) &&
3681             ffe_ctl->have_caching_bg && !ffe_ctl->orig_have_caching_bg)
3682                 ffe_ctl->orig_have_caching_bg = true;
3683
3684         if (!ins->objectid && ffe_ctl->loop >= LOOP_CACHING_WAIT &&
3685             ffe_ctl->have_caching_bg)
3686                 return 1;
3687
3688         if (!ins->objectid && ++(ffe_ctl->index) < BTRFS_NR_RAID_TYPES)
3689                 return 1;
3690
3691         if (ins->objectid) {
3692                 if (!use_cluster && last_ptr) {
3693                         spin_lock(&last_ptr->lock);
3694                         last_ptr->window_start = ins->objectid;
3695                         spin_unlock(&last_ptr->lock);
3696                 }
3697                 return 0;
3698         }
3699
3700         /*
3701          * LOOP_CACHING_NOWAIT, search partially cached block groups, kicking
3702          *                      caching kthreads as we move along
3703          * LOOP_CACHING_WAIT, search everything, and wait if our bg is caching
3704          * LOOP_ALLOC_CHUNK, force a chunk allocation and try again
3705          * LOOP_NO_EMPTY_SIZE, set empty_size and empty_cluster to 0 and try
3706          *                     again
3707          */
3708         if (ffe_ctl->loop < LOOP_NO_EMPTY_SIZE) {
3709                 ffe_ctl->index = 0;
3710                 if (ffe_ctl->loop == LOOP_CACHING_NOWAIT) {
3711                         /*
3712                          * We want to skip the LOOP_CACHING_WAIT step if we
3713                          * don't have any uncached bgs and we've already done a
3714                          * full search through.
3715                          */
3716                         if (ffe_ctl->orig_have_caching_bg || !full_search)
3717                                 ffe_ctl->loop = LOOP_CACHING_WAIT;
3718                         else
3719                                 ffe_ctl->loop = LOOP_ALLOC_CHUNK;
3720                 } else {
3721                         ffe_ctl->loop++;
3722                 }
3723
3724                 if (ffe_ctl->loop == LOOP_ALLOC_CHUNK) {
3725                         struct btrfs_trans_handle *trans;
3726                         int exist = 0;
3727
3728                         trans = current->journal_info;
3729                         if (trans)
3730                                 exist = 1;
3731                         else
3732                                 trans = btrfs_join_transaction(root);
3733
3734                         if (IS_ERR(trans)) {
3735                                 ret = PTR_ERR(trans);
3736                                 return ret;
3737                         }
3738
3739                         ret = btrfs_chunk_alloc(trans, ffe_ctl->flags,
3740                                                 CHUNK_ALLOC_FORCE);
3741
3742                         /*
3743                          * If we can't allocate a new chunk we've already looped
3744                          * through at least once, move on to the NO_EMPTY_SIZE
3745                          * case.
3746                          */
3747                         if (ret == -ENOSPC)
3748                                 ffe_ctl->loop = LOOP_NO_EMPTY_SIZE;
3749
3750                         /* Do not bail out on ENOSPC since we can do more. */
3751                         if (ret < 0 && ret != -ENOSPC)
3752                                 btrfs_abort_transaction(trans, ret);
3753                         else
3754                                 ret = 0;
3755                         if (!exist)
3756                                 btrfs_end_transaction(trans);
3757                         if (ret)
3758                                 return ret;
3759                 }
3760
3761                 if (ffe_ctl->loop == LOOP_NO_EMPTY_SIZE) {
3762                         /*
3763                          * Don't loop again if we already have no empty_size and
3764                          * no empty_cluster.
3765                          */
3766                         if (ffe_ctl->empty_size == 0 &&
3767                             ffe_ctl->empty_cluster == 0)
3768                                 return -ENOSPC;
3769                         ffe_ctl->empty_size = 0;
3770                         ffe_ctl->empty_cluster = 0;
3771                 }
3772                 return 1;
3773         }
3774         return -ENOSPC;
3775 }
3776
3777 /*
3778  * walks the btree of allocated extents and find a hole of a given size.
3779  * The key ins is changed to record the hole:
3780  * ins->objectid == start position
3781  * ins->flags = BTRFS_EXTENT_ITEM_KEY
3782  * ins->offset == the size of the hole.
3783  * Any available blocks before search_start are skipped.
3784  *
3785  * If there is no suitable free space, we will record the max size of
3786  * the free space extent currently.
3787  *
3788  * The overall logic and call chain:
3789  *
3790  * find_free_extent()
3791  * |- Iterate through all block groups
3792  * |  |- Get a valid block group
3793  * |  |- Try to do clustered allocation in that block group
3794  * |  |- Try to do unclustered allocation in that block group
3795  * |  |- Check if the result is valid
3796  * |  |  |- If valid, then exit
3797  * |  |- Jump to next block group
3798  * |
3799  * |- Push harder to find free extents
3800  *    |- If not found, re-iterate all block groups
3801  */
3802 static noinline int find_free_extent(struct btrfs_fs_info *fs_info,
3803                                 u64 ram_bytes, u64 num_bytes, u64 empty_size,
3804                                 u64 hint_byte, struct btrfs_key *ins,
3805                                 u64 flags, int delalloc)
3806 {
3807         int ret = 0;
3808         int cache_block_group_error = 0;
3809         struct btrfs_free_cluster *last_ptr = NULL;
3810         struct btrfs_block_group *block_group = NULL;
3811         struct find_free_extent_ctl ffe_ctl = {0};
3812         struct btrfs_space_info *space_info;
3813         bool use_cluster = true;
3814         bool full_search = false;
3815
3816         WARN_ON(num_bytes < fs_info->sectorsize);
3817
3818         ffe_ctl.num_bytes = num_bytes;
3819         ffe_ctl.empty_size = empty_size;
3820         ffe_ctl.flags = flags;
3821         ffe_ctl.search_start = 0;
3822         ffe_ctl.retry_clustered = false;
3823         ffe_ctl.retry_unclustered = false;
3824         ffe_ctl.delalloc = delalloc;
3825         ffe_ctl.index = btrfs_bg_flags_to_raid_index(flags);
3826         ffe_ctl.have_caching_bg = false;
3827         ffe_ctl.orig_have_caching_bg = false;
3828         ffe_ctl.found_offset = 0;
3829
3830         ins->type = BTRFS_EXTENT_ITEM_KEY;
3831         ins->objectid = 0;
3832         ins->offset = 0;
3833
3834         trace_find_free_extent(fs_info, num_bytes, empty_size, flags);
3835
3836         space_info = btrfs_find_space_info(fs_info, flags);
3837         if (!space_info) {
3838                 btrfs_err(fs_info, "No space info for %llu", flags);
3839                 return -ENOSPC;
3840         }
3841
3842         /*
3843          * If our free space is heavily fragmented we may not be able to make
3844          * big contiguous allocations, so instead of doing the expensive search
3845          * for free space, simply return ENOSPC with our max_extent_size so we
3846          * can go ahead and search for a more manageable chunk.
3847          *
3848          * If our max_extent_size is large enough for our allocation simply
3849          * disable clustering since we will likely not be able to find enough
3850          * space to create a cluster and induce latency trying.
3851          */
3852         if (unlikely(space_info->max_extent_size)) {
3853                 spin_lock(&space_info->lock);
3854                 if (space_info->max_extent_size &&
3855                     num_bytes > space_info->max_extent_size) {
3856                         ins->offset = space_info->max_extent_size;
3857                         spin_unlock(&space_info->lock);
3858                         return -ENOSPC;
3859                 } else if (space_info->max_extent_size) {
3860                         use_cluster = false;
3861                 }
3862                 spin_unlock(&space_info->lock);
3863         }
3864
3865         last_ptr = fetch_cluster_info(fs_info, space_info,
3866                                       &ffe_ctl.empty_cluster);
3867         if (last_ptr) {
3868                 spin_lock(&last_ptr->lock);
3869                 if (last_ptr->block_group)
3870                         hint_byte = last_ptr->window_start;
3871                 if (last_ptr->fragmented) {
3872                         /*
3873                          * We still set window_start so we can keep track of the
3874                          * last place we found an allocation to try and save
3875                          * some time.
3876                          */
3877                         hint_byte = last_ptr->window_start;
3878                         use_cluster = false;
3879                 }
3880                 spin_unlock(&last_ptr->lock);
3881         }
3882
3883         ffe_ctl.search_start = max(ffe_ctl.search_start,
3884                                    first_logical_byte(fs_info, 0));
3885         ffe_ctl.search_start = max(ffe_ctl.search_start, hint_byte);
3886         if (ffe_ctl.search_start == hint_byte) {
3887                 block_group = btrfs_lookup_block_group(fs_info,
3888                                                        ffe_ctl.search_start);
3889                 /*
3890                  * we don't want to use the block group if it doesn't match our
3891                  * allocation bits, or if its not cached.
3892                  *
3893                  * However if we are re-searching with an ideal block group
3894                  * picked out then we don't care that the block group is cached.
3895                  */
3896                 if (block_group && block_group_bits(block_group, flags) &&
3897                     block_group->cached != BTRFS_CACHE_NO) {
3898                         down_read(&space_info->groups_sem);
3899                         if (list_empty(&block_group->list) ||
3900                             block_group->ro) {
3901                                 /*
3902                                  * someone is removing this block group,
3903                                  * we can't jump into the have_block_group
3904                                  * target because our list pointers are not
3905                                  * valid
3906                                  */
3907                                 btrfs_put_block_group(block_group);
3908                                 up_read(&space_info->groups_sem);
3909                         } else {
3910                                 ffe_ctl.index = btrfs_bg_flags_to_raid_index(
3911                                                 block_group->flags);
3912                                 btrfs_lock_block_group(block_group, delalloc);
3913                                 goto have_block_group;
3914                         }
3915                 } else if (block_group) {
3916                         btrfs_put_block_group(block_group);
3917                 }
3918         }
3919 search:
3920         ffe_ctl.have_caching_bg = false;
3921         if (ffe_ctl.index == btrfs_bg_flags_to_raid_index(flags) ||
3922             ffe_ctl.index == 0)
3923                 full_search = true;
3924         down_read(&space_info->groups_sem);
3925         list_for_each_entry(block_group,
3926                             &space_info->block_groups[ffe_ctl.index], list) {
3927                 /* If the block group is read-only, we can skip it entirely. */
3928                 if (unlikely(block_group->ro))
3929                         continue;
3930
3931                 btrfs_grab_block_group(block_group, delalloc);
3932                 ffe_ctl.search_start = block_group->start;
3933
3934                 /*
3935                  * this can happen if we end up cycling through all the
3936                  * raid types, but we want to make sure we only allocate
3937                  * for the proper type.
3938                  */
3939                 if (!block_group_bits(block_group, flags)) {
3940                         u64 extra = BTRFS_BLOCK_GROUP_DUP |
3941                                 BTRFS_BLOCK_GROUP_RAID1_MASK |
3942                                 BTRFS_BLOCK_GROUP_RAID56_MASK |
3943                                 BTRFS_BLOCK_GROUP_RAID10;
3944
3945                         /*
3946                          * if they asked for extra copies and this block group
3947                          * doesn't provide them, bail.  This does allow us to
3948                          * fill raid0 from raid1.
3949                          */
3950                         if ((flags & extra) && !(block_group->flags & extra))
3951                                 goto loop;
3952
3953                         /*
3954                          * This block group has different flags than we want.
3955                          * It's possible that we have MIXED_GROUP flag but no
3956                          * block group is mixed.  Just skip such block group.
3957                          */
3958                         btrfs_release_block_group(block_group, delalloc);
3959                         continue;
3960                 }
3961
3962 have_block_group:
3963                 ffe_ctl.cached = btrfs_block_group_done(block_group);
3964                 if (unlikely(!ffe_ctl.cached)) {
3965                         ffe_ctl.have_caching_bg = true;
3966                         ret = btrfs_cache_block_group(block_group, 0);
3967
3968                         /*
3969                          * If we get ENOMEM here or something else we want to
3970                          * try other block groups, because it may not be fatal.
3971                          * However if we can't find anything else we need to
3972                          * save our return here so that we return the actual
3973                          * error that caused problems, not ENOSPC.
3974                          */
3975                         if (ret < 0) {
3976                                 if (!cache_block_group_error)
3977                                         cache_block_group_error = ret;
3978                                 ret = 0;
3979                                 goto loop;
3980                         }
3981                         ret = 0;
3982                 }
3983
3984                 if (unlikely(block_group->cached == BTRFS_CACHE_ERROR))
3985                         goto loop;
3986
3987                 /*
3988                  * Ok we want to try and use the cluster allocator, so
3989                  * lets look there
3990                  */
3991                 if (last_ptr && use_cluster) {
3992                         struct btrfs_block_group *cluster_bg = NULL;
3993
3994                         ret = find_free_extent_clustered(block_group, last_ptr,
3995                                                          &ffe_ctl, &cluster_bg);
3996
3997                         if (ret == 0) {
3998                                 if (cluster_bg && cluster_bg != block_group) {
3999                                         btrfs_release_block_group(block_group,
4000                                                                   delalloc);
4001                                         block_group = cluster_bg;
4002                                 }
4003                                 goto checks;
4004                         } else if (ret == -EAGAIN) {
4005                                 goto have_block_group;
4006                         } else if (ret > 0) {
4007                                 goto loop;
4008                         }
4009                         /* ret == -ENOENT case falls through */
4010                 }
4011
4012                 ret = find_free_extent_unclustered(block_group, last_ptr,
4013                                                    &ffe_ctl);
4014                 if (ret == -EAGAIN)
4015                         goto have_block_group;
4016                 else if (ret > 0)
4017                         goto loop;
4018                 /* ret == 0 case falls through */
4019 checks:
4020                 ffe_ctl.search_start = round_up(ffe_ctl.found_offset,
4021                                              fs_info->stripesize);
4022
4023                 /* move on to the next group */
4024                 if (ffe_ctl.search_start + num_bytes >
4025                     block_group->start + block_group->length) {
4026                         btrfs_add_free_space(block_group, ffe_ctl.found_offset,
4027                                              num_bytes);
4028                         goto loop;
4029                 }
4030
4031                 if (ffe_ctl.found_offset < ffe_ctl.search_start)
4032                         btrfs_add_free_space(block_group, ffe_ctl.found_offset,
4033                                 ffe_ctl.search_start - ffe_ctl.found_offset);
4034
4035                 ret = btrfs_add_reserved_bytes(block_group, ram_bytes,
4036                                 num_bytes, delalloc);
4037                 if (ret == -EAGAIN) {
4038                         btrfs_add_free_space(block_group, ffe_ctl.found_offset,
4039                                              num_bytes);
4040                         goto loop;
4041                 }
4042                 btrfs_inc_block_group_reservations(block_group);
4043
4044                 /* we are all good, lets return */
4045                 ins->objectid = ffe_ctl.search_start;
4046                 ins->offset = num_bytes;
4047
4048                 trace_btrfs_reserve_extent(block_group, ffe_ctl.search_start,
4049                                            num_bytes);
4050                 btrfs_release_block_group(block_group, delalloc);
4051                 break;
4052 loop:
4053                 ffe_ctl.retry_clustered = false;
4054                 ffe_ctl.retry_unclustered = false;
4055                 BUG_ON(btrfs_bg_flags_to_raid_index(block_group->flags) !=
4056                        ffe_ctl.index);
4057                 btrfs_release_block_group(block_group, delalloc);
4058                 cond_resched();
4059         }
4060         up_read(&space_info->groups_sem);
4061
4062         ret = find_free_extent_update_loop(fs_info, last_ptr, ins, &ffe_ctl,
4063                                            full_search, use_cluster);
4064         if (ret > 0)
4065                 goto search;
4066
4067         if (ret == -ENOSPC && !cache_block_group_error) {
4068                 /*
4069                  * Use ffe_ctl->total_free_space as fallback if we can't find
4070                  * any contiguous hole.
4071                  */
4072                 if (!ffe_ctl.max_extent_size)
4073                         ffe_ctl.max_extent_size = ffe_ctl.total_free_space;
4074                 spin_lock(&space_info->lock);
4075                 space_info->max_extent_size = ffe_ctl.max_extent_size;
4076                 spin_unlock(&space_info->lock);
4077                 ins->offset = ffe_ctl.max_extent_size;
4078         } else if (ret == -ENOSPC) {
4079                 ret = cache_block_group_error;
4080         }
4081         return ret;
4082 }
4083
4084 /*
4085  * btrfs_reserve_extent - entry point to the extent allocator. Tries to find a
4086  *                        hole that is at least as big as @num_bytes.
4087  *
4088  * @root           -    The root that will contain this extent
4089  *
4090  * @ram_bytes      -    The amount of space in ram that @num_bytes take. This
4091  *                      is used for accounting purposes. This value differs
4092  *                      from @num_bytes only in the case of compressed extents.
4093  *
4094  * @num_bytes      -    Number of bytes to allocate on-disk.
4095  *
4096  * @min_alloc_size -    Indicates the minimum amount of space that the
4097  *                      allocator should try to satisfy. In some cases
4098  *                      @num_bytes may be larger than what is required and if
4099  *                      the filesystem is fragmented then allocation fails.
4100  *                      However, the presence of @min_alloc_size gives a
4101  *                      chance to try and satisfy the smaller allocation.
4102  *
4103  * @empty_size     -    A hint that you plan on doing more COW. This is the
4104  *                      size in bytes the allocator should try to find free
4105  *                      next to the block it returns.  This is just a hint and
4106  *                      may be ignored by the allocator.
4107  *
4108  * @hint_byte      -    Hint to the allocator to start searching above the byte
4109  *                      address passed. It might be ignored.
4110  *
4111  * @ins            -    This key is modified to record the found hole. It will
4112  *                      have the following values:
4113  *                      ins->objectid == start position
4114  *                      ins->flags = BTRFS_EXTENT_ITEM_KEY
4115  *                      ins->offset == the size of the hole.
4116  *
4117  * @is_data        -    Boolean flag indicating whether an extent is
4118  *                      allocated for data (true) or metadata (false)
4119  *
4120  * @delalloc       -    Boolean flag indicating whether this allocation is for
4121  *                      delalloc or not. If 'true' data_rwsem of block groups
4122  *                      is going to be acquired.
4123  *
4124  *
4125  * Returns 0 when an allocation succeeded or < 0 when an error occurred. In
4126  * case -ENOSPC is returned then @ins->offset will contain the size of the
4127  * largest available hole the allocator managed to find.
4128  */
4129 int btrfs_reserve_extent(struct btrfs_root *root, u64 ram_bytes,
4130                          u64 num_bytes, u64 min_alloc_size,
4131                          u64 empty_size, u64 hint_byte,
4132                          struct btrfs_key *ins, int is_data, int delalloc)
4133 {
4134         struct btrfs_fs_info *fs_info = root->fs_info;
4135         bool final_tried = num_bytes == min_alloc_size;
4136         u64 flags;
4137         int ret;
4138
4139         flags = get_alloc_profile_by_root(root, is_data);
4140 again:
4141         WARN_ON(num_bytes < fs_info->sectorsize);
4142         ret = find_free_extent(fs_info, ram_bytes, num_bytes, empty_size,
4143                                hint_byte, ins, flags, delalloc);
4144         if (!ret && !is_data) {
4145                 btrfs_dec_block_group_reservations(fs_info, ins->objectid);
4146         } else if (ret == -ENOSPC) {
4147                 if (!final_tried && ins->offset) {
4148                         num_bytes = min(num_bytes >> 1, ins->offset);
4149                         num_bytes = round_down(num_bytes,
4150                                                fs_info->sectorsize);
4151                         num_bytes = max(num_bytes, min_alloc_size);
4152                         ram_bytes = num_bytes;
4153                         if (num_bytes == min_alloc_size)
4154                                 final_tried = true;
4155                         goto again;
4156                 } else if (btrfs_test_opt(fs_info, ENOSPC_DEBUG)) {
4157                         struct btrfs_space_info *sinfo;
4158
4159                         sinfo = btrfs_find_space_info(fs_info, flags);
4160                         btrfs_err(fs_info,
4161                                   "allocation failed flags %llu, wanted %llu",
4162                                   flags, num_bytes);
4163                         if (sinfo)
4164                                 btrfs_dump_space_info(fs_info, sinfo,
4165                                                       num_bytes, 1);
4166                 }
4167         }
4168
4169         return ret;
4170 }
4171
4172 int btrfs_free_reserved_extent(struct btrfs_fs_info *fs_info,
4173                                u64 start, u64 len, int delalloc)
4174 {
4175         struct btrfs_block_group *cache;
4176
4177         cache = btrfs_lookup_block_group(fs_info, start);
4178         if (!cache) {
4179                 btrfs_err(fs_info, "Unable to find block group for %llu",
4180                           start);
4181                 return -ENOSPC;
4182         }
4183
4184         btrfs_add_free_space(cache, start, len);
4185         btrfs_free_reserved_bytes(cache, len, delalloc);
4186         trace_btrfs_reserved_extent_free(fs_info, start, len);
4187
4188         btrfs_put_block_group(cache);
4189         return 0;
4190 }
4191
4192 int btrfs_pin_reserved_extent(struct btrfs_fs_info *fs_info, u64 start, u64 len)
4193 {
4194         struct btrfs_block_group *cache;
4195         int ret = 0;
4196
4197         cache = btrfs_lookup_block_group(fs_info, start);
4198         if (!cache) {
4199                 btrfs_err(fs_info, "unable to find block group for %llu", start);
4200                 return -ENOSPC;
4201         }
4202
4203         ret = pin_down_extent(cache, start, len, 1);
4204         btrfs_put_block_group(cache);
4205         return ret;
4206 }
4207
4208 static int alloc_reserved_file_extent(struct btrfs_trans_handle *trans,
4209                                       u64 parent, u64 root_objectid,
4210                                       u64 flags, u64 owner, u64 offset,
4211                                       struct btrfs_key *ins, int ref_mod)
4212 {
4213         struct btrfs_fs_info *fs_info = trans->fs_info;
4214         int ret;
4215         struct btrfs_extent_item *extent_item;
4216         struct btrfs_extent_inline_ref *iref;
4217         struct btrfs_path *path;
4218         struct extent_buffer *leaf;
4219         int type;
4220         u32 size;
4221
4222         if (parent > 0)
4223                 type = BTRFS_SHARED_DATA_REF_KEY;
4224         else
4225                 type = BTRFS_EXTENT_DATA_REF_KEY;
4226
4227         size = sizeof(*extent_item) + btrfs_extent_inline_ref_size(type);
4228
4229         path = btrfs_alloc_path();
4230         if (!path)
4231                 return -ENOMEM;
4232
4233         path->leave_spinning = 1;
4234         ret = btrfs_insert_empty_item(trans, fs_info->extent_root, path,
4235                                       ins, size);
4236         if (ret) {
4237                 btrfs_free_path(path);
4238                 return ret;
4239         }
4240
4241         leaf = path->nodes[0];
4242         extent_item = btrfs_item_ptr(leaf, path->slots[0],
4243                                      struct btrfs_extent_item);
4244         btrfs_set_extent_refs(leaf, extent_item, ref_mod);
4245         btrfs_set_extent_generation(leaf, extent_item, trans->transid);
4246         btrfs_set_extent_flags(leaf, extent_item,
4247                                flags | BTRFS_EXTENT_FLAG_DATA);
4248
4249         iref = (struct btrfs_extent_inline_ref *)(extent_item + 1);
4250         btrfs_set_extent_inline_ref_type(leaf, iref, type);
4251         if (parent > 0) {
4252                 struct btrfs_shared_data_ref *ref;
4253                 ref = (struct btrfs_shared_data_ref *)(iref + 1);
4254                 btrfs_set_extent_inline_ref_offset(leaf, iref, parent);
4255                 btrfs_set_shared_data_ref_count(leaf, ref, ref_mod);
4256         } else {
4257                 struct btrfs_extent_data_ref *ref;
4258                 ref = (struct btrfs_extent_data_ref *)(&iref->offset);
4259                 btrfs_set_extent_data_ref_root(leaf, ref, root_objectid);
4260                 btrfs_set_extent_data_ref_objectid(leaf, ref, owner);
4261                 btrfs_set_extent_data_ref_offset(leaf, ref, offset);
4262                 btrfs_set_extent_data_ref_count(leaf, ref, ref_mod);
4263         }
4264
4265         btrfs_mark_buffer_dirty(path->nodes[0]);
4266         btrfs_free_path(path);
4267
4268         ret = remove_from_free_space_tree(trans, ins->objectid, ins->offset);
4269         if (ret)
4270                 return ret;
4271
4272         ret = btrfs_update_block_group(trans, ins->objectid, ins->offset, 1);
4273         if (ret) { /* -ENOENT, logic error */
4274                 btrfs_err(fs_info, "update block group failed for %llu %llu",
4275                         ins->objectid, ins->offset);
4276                 BUG();
4277         }
4278         trace_btrfs_reserved_extent_alloc(fs_info, ins->objectid, ins->offset);
4279         return ret;
4280 }
4281
4282 static int alloc_reserved_tree_block(struct btrfs_trans_handle *trans,
4283                                      struct btrfs_delayed_ref_node *node,
4284                                      struct btrfs_delayed_extent_op *extent_op)
4285 {
4286         struct btrfs_fs_info *fs_info = trans->fs_info;
4287         int ret;
4288         struct btrfs_extent_item *extent_item;
4289         struct btrfs_key extent_key;
4290         struct btrfs_tree_block_info *block_info;
4291         struct btrfs_extent_inline_ref *iref;
4292         struct btrfs_path *path;
4293         struct extent_buffer *leaf;
4294         struct btrfs_delayed_tree_ref *ref;
4295         u32 size = sizeof(*extent_item) + sizeof(*iref);
4296         u64 num_bytes;
4297         u64 flags = extent_op->flags_to_set;
4298         bool skinny_metadata = btrfs_fs_incompat(fs_info, SKINNY_METADATA);
4299
4300         ref = btrfs_delayed_node_to_tree_ref(node);
4301
4302         extent_key.objectid = node->bytenr;
4303         if (skinny_metadata) {
4304                 extent_key.offset = ref->level;
4305                 extent_key.type = BTRFS_METADATA_ITEM_KEY;
4306                 num_bytes = fs_info->nodesize;
4307         } else {
4308                 extent_key.offset = node->num_bytes;
4309                 extent_key.type = BTRFS_EXTENT_ITEM_KEY;
4310                 size += sizeof(*block_info);
4311                 num_bytes = node->num_bytes;
4312         }
4313
4314         path = btrfs_alloc_path();
4315         if (!path)
4316                 return -ENOMEM;
4317
4318         path->leave_spinning = 1;
4319         ret = btrfs_insert_empty_item(trans, fs_info->extent_root, path,
4320                                       &extent_key, size);
4321         if (ret) {
4322                 btrfs_free_path(path);
4323                 return ret;
4324         }
4325
4326         leaf = path->nodes[0];
4327         extent_item = btrfs_item_ptr(leaf, path->slots[0],
4328                                      struct btrfs_extent_item);
4329         btrfs_set_extent_refs(leaf, extent_item, 1);
4330         btrfs_set_extent_generation(leaf, extent_item, trans->transid);
4331         btrfs_set_extent_flags(leaf, extent_item,
4332                                flags | BTRFS_EXTENT_FLAG_TREE_BLOCK);
4333
4334         if (skinny_metadata) {
4335                 iref = (struct btrfs_extent_inline_ref *)(extent_item + 1);
4336         } else {
4337                 block_info = (struct btrfs_tree_block_info *)(extent_item + 1);
4338                 btrfs_set_tree_block_key(leaf, block_info, &extent_op->key);
4339                 btrfs_set_tree_block_level(leaf, block_info, ref->level);
4340                 iref = (struct btrfs_extent_inline_ref *)(block_info + 1);
4341         }
4342
4343         if (node->type == BTRFS_SHARED_BLOCK_REF_KEY) {
4344                 BUG_ON(!(flags & BTRFS_BLOCK_FLAG_FULL_BACKREF));
4345                 btrfs_set_extent_inline_ref_type(leaf, iref,
4346                                                  BTRFS_SHARED_BLOCK_REF_KEY);
4347                 btrfs_set_extent_inline_ref_offset(leaf, iref, ref->parent);
4348         } else {
4349                 btrfs_set_extent_inline_ref_type(leaf, iref,
4350                                                  BTRFS_TREE_BLOCK_REF_KEY);
4351                 btrfs_set_extent_inline_ref_offset(leaf, iref, ref->root);
4352         }
4353
4354         btrfs_mark_buffer_dirty(leaf);
4355         btrfs_free_path(path);
4356
4357         ret = remove_from_free_space_tree(trans, extent_key.objectid,
4358                                           num_bytes);
4359         if (ret)
4360                 return ret;
4361
4362         ret = btrfs_update_block_group(trans, extent_key.objectid,
4363                                        fs_info->nodesize, 1);
4364         if (ret) { /* -ENOENT, logic error */
4365                 btrfs_err(fs_info, "update block group failed for %llu %llu",
4366                         extent_key.objectid, extent_key.offset);
4367                 BUG();
4368         }
4369
4370         trace_btrfs_reserved_extent_alloc(fs_info, extent_key.objectid,
4371                                           fs_info->nodesize);
4372         return ret;
4373 }
4374
4375 int btrfs_alloc_reserved_file_extent(struct btrfs_trans_handle *trans,
4376                                      struct btrfs_root *root, u64 owner,
4377                                      u64 offset, u64 ram_bytes,
4378                                      struct btrfs_key *ins)
4379 {
4380         struct btrfs_ref generic_ref = { 0 };
4381         int ret;
4382
4383         BUG_ON(root->root_key.objectid == BTRFS_TREE_LOG_OBJECTID);
4384
4385         btrfs_init_generic_ref(&generic_ref, BTRFS_ADD_DELAYED_EXTENT,
4386                                ins->objectid, ins->offset, 0);
4387         btrfs_init_data_ref(&generic_ref, root->root_key.objectid, owner, offset);
4388         btrfs_ref_tree_mod(root->fs_info, &generic_ref);
4389         ret = btrfs_add_delayed_data_ref(trans, &generic_ref,
4390                                          ram_bytes, NULL, NULL);
4391         return ret;
4392 }
4393
4394 /*
4395  * this is used by the tree logging recovery code.  It records that
4396  * an extent has been allocated and makes sure to clear the free
4397  * space cache bits as well
4398  */
4399 int btrfs_alloc_logged_file_extent(struct btrfs_trans_handle *trans,
4400                                    u64 root_objectid, u64 owner, u64 offset,
4401                                    struct btrfs_key *ins)
4402 {
4403         struct btrfs_fs_info *fs_info = trans->fs_info;
4404         int ret;
4405         struct btrfs_block_group *block_group;
4406         struct btrfs_space_info *space_info;
4407
4408         /*
4409          * Mixed block groups will exclude before processing the log so we only
4410          * need to do the exclude dance if this fs isn't mixed.
4411          */
4412         if (!btrfs_fs_incompat(fs_info, MIXED_GROUPS)) {
4413                 ret = __exclude_logged_extent(fs_info, ins->objectid,
4414                                               ins->offset);
4415                 if (ret)
4416                         return ret;
4417         }
4418
4419         block_group = btrfs_lookup_block_group(fs_info, ins->objectid);
4420         if (!block_group)
4421                 return -EINVAL;
4422
4423         space_info = block_group->space_info;
4424         spin_lock(&space_info->lock);
4425         spin_lock(&block_group->lock);
4426         space_info->bytes_reserved += ins->offset;
4427         block_group->reserved += ins->offset;
4428         spin_unlock(&block_group->lock);
4429         spin_unlock(&space_info->lock);
4430
4431         ret = alloc_reserved_file_extent(trans, 0, root_objectid, 0, owner,
4432                                          offset, ins, 1);
4433         btrfs_put_block_group(block_group);
4434         return ret;
4435 }
4436
4437 static struct extent_buffer *
4438 btrfs_init_new_buffer(struct btrfs_trans_handle *trans, struct btrfs_root *root,
4439                       u64 bytenr, int level, u64 owner)
4440 {
4441         struct btrfs_fs_info *fs_info = root->fs_info;
4442         struct extent_buffer *buf;
4443
4444         buf = btrfs_find_create_tree_block(fs_info, bytenr);
4445         if (IS_ERR(buf))
4446                 return buf;
4447
4448         /*
4449          * Extra safety check in case the extent tree is corrupted and extent
4450          * allocator chooses to use a tree block which is already used and
4451          * locked.
4452          */
4453         if (buf->lock_owner == current->pid) {
4454                 btrfs_err_rl(fs_info,
4455 "tree block %llu owner %llu already locked by pid=%d, extent tree corruption detected",
4456                         buf->start, btrfs_header_owner(buf), current->pid);
4457                 free_extent_buffer(buf);
4458                 return ERR_PTR(-EUCLEAN);
4459         }
4460
4461         btrfs_set_buffer_lockdep_class(root->root_key.objectid, buf, level);
4462         btrfs_tree_lock(buf);
4463         btrfs_clean_tree_block(buf);
4464         clear_bit(EXTENT_BUFFER_STALE, &buf->bflags);
4465
4466         btrfs_set_lock_blocking_write(buf);
4467         set_extent_buffer_uptodate(buf);
4468
4469         memzero_extent_buffer(buf, 0, sizeof(struct btrfs_header));
4470         btrfs_set_header_level(buf, level);
4471         btrfs_set_header_bytenr(buf, buf->start);
4472         btrfs_set_header_generation(buf, trans->transid);
4473         btrfs_set_header_backref_rev(buf, BTRFS_MIXED_BACKREF_REV);
4474         btrfs_set_header_owner(buf, owner);
4475         write_extent_buffer_fsid(buf, fs_info->fs_devices->metadata_uuid);
4476         write_extent_buffer_chunk_tree_uuid(buf, fs_info->chunk_tree_uuid);
4477         if (root->root_key.objectid == BTRFS_TREE_LOG_OBJECTID) {
4478                 buf->log_index = root->log_transid % 2;
4479                 /*
4480                  * we allow two log transactions at a time, use different
4481                  * EXTENT bit to differentiate dirty pages.
4482                  */
4483                 if (buf->log_index == 0)
4484                         set_extent_dirty(&root->dirty_log_pages, buf->start,
4485                                         buf->start + buf->len - 1, GFP_NOFS);
4486                 else
4487                         set_extent_new(&root->dirty_log_pages, buf->start,
4488                                         buf->start + buf->len - 1);
4489         } else {
4490                 buf->log_index = -1;
4491                 set_extent_dirty(&trans->transaction->dirty_pages, buf->start,
4492                          buf->start + buf->len - 1, GFP_NOFS);
4493         }
4494         trans->dirty = true;
4495         /* this returns a buffer locked for blocking */
4496         return buf;
4497 }
4498
4499 /*
4500  * finds a free extent and does all the dirty work required for allocation
4501  * returns the tree buffer or an ERR_PTR on error.
4502  */
4503 struct extent_buffer *btrfs_alloc_tree_block(struct btrfs_trans_handle *trans,
4504                                              struct btrfs_root *root,
4505                                              u64 parent, u64 root_objectid,
4506                                              const struct btrfs_disk_key *key,
4507                                              int level, u64 hint,
4508                                              u64 empty_size)
4509 {
4510         struct btrfs_fs_info *fs_info = root->fs_info;
4511         struct btrfs_key ins;
4512         struct btrfs_block_rsv *block_rsv;
4513         struct extent_buffer *buf;
4514         struct btrfs_delayed_extent_op *extent_op;
4515         struct btrfs_ref generic_ref = { 0 };
4516         u64 flags = 0;
4517         int ret;
4518         u32 blocksize = fs_info->nodesize;
4519         bool skinny_metadata = btrfs_fs_incompat(fs_info, SKINNY_METADATA);
4520
4521 #ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
4522         if (btrfs_is_testing(fs_info)) {
4523                 buf = btrfs_init_new_buffer(trans, root, root->alloc_bytenr,
4524                                             level, root_objectid);
4525                 if (!IS_ERR(buf))
4526                         root->alloc_bytenr += blocksize;
4527                 return buf;
4528         }
4529 #endif
4530
4531         block_rsv = btrfs_use_block_rsv(trans, root, blocksize);
4532         if (IS_ERR(block_rsv))
4533                 return ERR_CAST(block_rsv);
4534
4535         ret = btrfs_reserve_extent(root, blocksize, blocksize, blocksize,
4536                                    empty_size, hint, &ins, 0, 0);
4537         if (ret)
4538                 goto out_unuse;
4539
4540         buf = btrfs_init_new_buffer(trans, root, ins.objectid, level,
4541                                     root_objectid);
4542         if (IS_ERR(buf)) {
4543                 ret = PTR_ERR(buf);
4544                 goto out_free_reserved;
4545         }
4546
4547         if (root_objectid == BTRFS_TREE_RELOC_OBJECTID) {
4548                 if (parent == 0)
4549                         parent = ins.objectid;
4550                 flags |= BTRFS_BLOCK_FLAG_FULL_BACKREF;
4551         } else
4552                 BUG_ON(parent > 0);
4553
4554         if (root_objectid != BTRFS_TREE_LOG_OBJECTID) {
4555                 extent_op = btrfs_alloc_delayed_extent_op();
4556                 if (!extent_op) {
4557                         ret = -ENOMEM;
4558                         goto out_free_buf;
4559                 }
4560                 if (key)
4561                         memcpy(&extent_op->key, key, sizeof(extent_op->key));
4562                 else
4563                         memset(&extent_op->key, 0, sizeof(extent_op->key));
4564                 extent_op->flags_to_set = flags;
4565                 extent_op->update_key = skinny_metadata ? false : true;
4566                 extent_op->update_flags = true;
4567                 extent_op->is_data = false;
4568                 extent_op->level = level;
4569
4570                 btrfs_init_generic_ref(&generic_ref, BTRFS_ADD_DELAYED_EXTENT,
4571                                        ins.objectid, ins.offset, parent);
4572                 generic_ref.real_root = root->root_key.objectid;
4573                 btrfs_init_tree_ref(&generic_ref, level, root_objectid);
4574                 btrfs_ref_tree_mod(fs_info, &generic_ref);
4575                 ret = btrfs_add_delayed_tree_ref(trans, &generic_ref,
4576                                                  extent_op, NULL, NULL);
4577                 if (ret)
4578                         goto out_free_delayed;
4579         }
4580         return buf;
4581
4582 out_free_delayed:
4583         btrfs_free_delayed_extent_op(extent_op);
4584 out_free_buf:
4585         free_extent_buffer(buf);
4586 out_free_reserved:
4587         btrfs_free_reserved_extent(fs_info, ins.objectid, ins.offset, 0);
4588 out_unuse:
4589         btrfs_unuse_block_rsv(fs_info, block_rsv, blocksize);
4590         return ERR_PTR(ret);
4591 }
4592
4593 struct walk_control {
4594         u64 refs[BTRFS_MAX_LEVEL];
4595         u64 flags[BTRFS_MAX_LEVEL];
4596         struct btrfs_key update_progress;
4597         struct btrfs_key drop_progress;
4598         int drop_level;
4599         int stage;
4600         int level;
4601         int shared_level;
4602         int update_ref;
4603         int keep_locks;
4604         int reada_slot;
4605         int reada_count;
4606         int restarted;
4607 };
4608
4609 #define DROP_REFERENCE  1
4610 #define UPDATE_BACKREF  2
4611
4612 static noinline void reada_walk_down(struct btrfs_trans_handle *trans,
4613                                      struct btrfs_root *root,
4614                                      struct walk_control *wc,
4615                                      struct btrfs_path *path)
4616 {
4617         struct btrfs_fs_info *fs_info = root->fs_info;
4618         u64 bytenr;
4619         u64 generation;
4620         u64 refs;
4621         u64 flags;
4622         u32 nritems;
4623         struct btrfs_key key;
4624         struct extent_buffer *eb;
4625         int ret;
4626         int slot;
4627         int nread = 0;
4628
4629         if (path->slots[wc->level] < wc->reada_slot) {
4630                 wc->reada_count = wc->reada_count * 2 / 3;
4631                 wc->reada_count = max(wc->reada_count, 2);
4632         } else {
4633                 wc->reada_count = wc->reada_count * 3 / 2;
4634                 wc->reada_count = min_t(int, wc->reada_count,
4635                                         BTRFS_NODEPTRS_PER_BLOCK(fs_info));
4636         }
4637
4638         eb = path->nodes[wc->level];
4639         nritems = btrfs_header_nritems(eb);
4640
4641         for (slot = path->slots[wc->level]; slot < nritems; slot++) {
4642                 if (nread >= wc->reada_count)
4643                         break;
4644
4645                 cond_resched();
4646                 bytenr = btrfs_node_blockptr(eb, slot);
4647                 generation = btrfs_node_ptr_generation(eb, slot);
4648
4649                 if (slot == path->slots[wc->level])
4650                         goto reada;
4651
4652                 if (wc->stage == UPDATE_BACKREF &&
4653                     generation <= root->root_key.offset)
4654                         continue;
4655
4656                 /* We don't lock the tree block, it's OK to be racy here */
4657                 ret = btrfs_lookup_extent_info(trans, fs_info, bytenr,
4658                                                wc->level - 1, 1, &refs,
4659                                                &flags);
4660                 /* We don't care about errors in readahead. */
4661                 if (ret < 0)
4662                         continue;
4663                 BUG_ON(refs == 0);
4664
4665                 if (wc->stage == DROP_REFERENCE) {
4666                         if (refs == 1)
4667                                 goto reada;
4668
4669                         if (wc->level == 1 &&
4670                             (flags & BTRFS_BLOCK_FLAG_FULL_BACKREF))
4671                                 continue;
4672                         if (!wc->update_ref ||
4673                             generation <= root->root_key.offset)
4674                                 continue;
4675                         btrfs_node_key_to_cpu(eb, &key, slot);
4676                         ret = btrfs_comp_cpu_keys(&key,
4677                                                   &wc->update_progress);
4678                         if (ret < 0)
4679                                 continue;
4680                 } else {
4681                         if (wc->level == 1 &&
4682                             (flags & BTRFS_BLOCK_FLAG_FULL_BACKREF))
4683                                 continue;
4684                 }
4685 reada:
4686                 readahead_tree_block(fs_info, bytenr);
4687                 nread++;
4688         }
4689         wc->reada_slot = slot;
4690 }
4691
4692 /*
4693  * helper to process tree block while walking down the tree.
4694  *
4695  * when wc->stage == UPDATE_BACKREF, this function updates
4696  * back refs for pointers in the block.
4697  *
4698  * NOTE: return value 1 means we should stop walking down.
4699  */
4700 static noinline int walk_down_proc(struct btrfs_trans_handle *trans,
4701                                    struct btrfs_root *root,
4702                                    struct btrfs_path *path,
4703                                    struct walk_control *wc, int lookup_info)
4704 {
4705         struct btrfs_fs_info *fs_info = root->fs_info;
4706         int level = wc->level;
4707         struct extent_buffer *eb = path->nodes[level];
4708         u64 flag = BTRFS_BLOCK_FLAG_FULL_BACKREF;
4709         int ret;
4710
4711         if (wc->stage == UPDATE_BACKREF &&
4712             btrfs_header_owner(eb) != root->root_key.objectid)
4713                 return 1;
4714
4715         /*
4716          * when reference count of tree block is 1, it won't increase
4717          * again. once full backref flag is set, we never clear it.
4718          */
4719         if (lookup_info &&
4720             ((wc->stage == DROP_REFERENCE && wc->refs[level] != 1) ||
4721              (wc->stage == UPDATE_BACKREF && !(wc->flags[level] & flag)))) {
4722                 BUG_ON(!path->locks[level]);
4723                 ret = btrfs_lookup_extent_info(trans, fs_info,
4724                                                eb->start, level, 1,
4725                                                &wc->refs[level],
4726                                                &wc->flags[level]);
4727                 BUG_ON(ret == -ENOMEM);
4728                 if (ret)
4729                         return ret;
4730                 BUG_ON(wc->refs[level] == 0);
4731         }
4732
4733         if (wc->stage == DROP_REFERENCE) {
4734                 if (wc->refs[level] > 1)
4735                         return 1;
4736
4737                 if (path->locks[level] && !wc->keep_locks) {
4738                         btrfs_tree_unlock_rw(eb, path->locks[level]);
4739                         path->locks[level] = 0;
4740                 }
4741                 return 0;
4742         }
4743
4744         /* wc->stage == UPDATE_BACKREF */
4745         if (!(wc->flags[level] & flag)) {
4746                 BUG_ON(!path->locks[level]);
4747                 ret = btrfs_inc_ref(trans, root, eb, 1);
4748                 BUG_ON(ret); /* -ENOMEM */
4749                 ret = btrfs_dec_ref(trans, root, eb, 0);
4750                 BUG_ON(ret); /* -ENOMEM */
4751                 ret = btrfs_set_disk_extent_flags(trans, eb->start,
4752                                                   eb->len, flag,
4753                                                   btrfs_header_level(eb), 0);
4754                 BUG_ON(ret); /* -ENOMEM */
4755                 wc->flags[level] |= flag;
4756         }
4757
4758         /*
4759          * the block is shared by multiple trees, so it's not good to
4760          * keep the tree lock
4761          */
4762         if (path->locks[level] && level > 0) {
4763                 btrfs_tree_unlock_rw(eb, path->locks[level]);
4764                 path->locks[level] = 0;
4765         }
4766         return 0;
4767 }
4768
4769 /*
4770  * This is used to verify a ref exists for this root to deal with a bug where we
4771  * would have a drop_progress key that hadn't been updated properly.
4772  */
4773 static int check_ref_exists(struct btrfs_trans_handle *trans,
4774                             struct btrfs_root *root, u64 bytenr, u64 parent,
4775                             int level)
4776 {
4777         struct btrfs_path *path;
4778         struct btrfs_extent_inline_ref *iref;
4779         int ret;
4780
4781         path = btrfs_alloc_path();
4782         if (!path)
4783                 return -ENOMEM;
4784
4785         ret = lookup_extent_backref(trans, path, &iref, bytenr,
4786                                     root->fs_info->nodesize, parent,
4787                                     root->root_key.objectid, level, 0);
4788         btrfs_free_path(path);
4789         if (ret == -ENOENT)
4790                 return 0;
4791         if (ret < 0)
4792                 return ret;
4793         return 1;
4794 }
4795
4796 /*
4797  * helper to process tree block pointer.
4798  *
4799  * when wc->stage == DROP_REFERENCE, this function checks
4800  * reference count of the block pointed to. if the block
4801  * is shared and we need update back refs for the subtree
4802  * rooted at the block, this function changes wc->stage to
4803  * UPDATE_BACKREF. if the block is shared and there is no
4804  * need to update back, this function drops the reference
4805  * to the block.
4806  *
4807  * NOTE: return value 1 means we should stop walking down.
4808  */
4809 static noinline int do_walk_down(struct btrfs_trans_handle *trans,
4810                                  struct btrfs_root *root,
4811                                  struct btrfs_path *path,
4812                                  struct walk_control *wc, int *lookup_info)
4813 {
4814         struct btrfs_fs_info *fs_info = root->fs_info;
4815         u64 bytenr;
4816         u64 generation;
4817         u64 parent;
4818         struct btrfs_key key;
4819         struct btrfs_key first_key;
4820         struct btrfs_ref ref = { 0 };
4821         struct extent_buffer *next;
4822         int level = wc->level;
4823         int reada = 0;
4824         int ret = 0;
4825         bool need_account = false;
4826
4827         generation = btrfs_node_ptr_generation(path->nodes[level],
4828                                                path->slots[level]);
4829         /*
4830          * if the lower level block was created before the snapshot
4831          * was created, we know there is no need to update back refs
4832          * for the subtree
4833          */
4834         if (wc->stage == UPDATE_BACKREF &&
4835             generation <= root->root_key.offset) {
4836                 *lookup_info = 1;
4837                 return 1;
4838         }
4839
4840         bytenr = btrfs_node_blockptr(path->nodes[level], path->slots[level]);
4841         btrfs_node_key_to_cpu(path->nodes[level], &first_key,
4842                               path->slots[level]);
4843
4844         next = find_extent_buffer(fs_info, bytenr);
4845         if (!next) {
4846                 next = btrfs_find_create_tree_block(fs_info, bytenr);
4847                 if (IS_ERR(next))
4848                         return PTR_ERR(next);
4849
4850                 btrfs_set_buffer_lockdep_class(root->root_key.objectid, next,
4851                                                level - 1);
4852                 reada = 1;
4853         }
4854         btrfs_tree_lock(next);
4855         btrfs_set_lock_blocking_write(next);
4856
4857         ret = btrfs_lookup_extent_info(trans, fs_info, bytenr, level - 1, 1,
4858                                        &wc->refs[level - 1],
4859                                        &wc->flags[level - 1]);
4860         if (ret < 0)
4861                 goto out_unlock;
4862
4863         if (unlikely(wc->refs[level - 1] == 0)) {
4864                 btrfs_err(fs_info, "Missing references.");
4865                 ret = -EIO;
4866                 goto out_unlock;
4867         }
4868         *lookup_info = 0;
4869
4870         if (wc->stage == DROP_REFERENCE) {
4871                 if (wc->refs[level - 1] > 1) {
4872                         need_account = true;
4873                         if (level == 1 &&
4874                             (wc->flags[0] & BTRFS_BLOCK_FLAG_FULL_BACKREF))
4875                                 goto skip;
4876
4877                         if (!wc->update_ref ||
4878                             generation <= root->root_key.offset)
4879                                 goto skip;
4880
4881                         btrfs_node_key_to_cpu(path->nodes[level], &key,
4882                                               path->slots[level]);
4883                         ret = btrfs_comp_cpu_keys(&key, &wc->update_progress);
4884                         if (ret < 0)
4885                                 goto skip;
4886
4887                         wc->stage = UPDATE_BACKREF;
4888                         wc->shared_level = level - 1;
4889                 }
4890         } else {
4891                 if (level == 1 &&
4892                     (wc->flags[0] & BTRFS_BLOCK_FLAG_FULL_BACKREF))
4893                         goto skip;
4894         }
4895
4896         if (!btrfs_buffer_uptodate(next, generation, 0)) {
4897                 btrfs_tree_unlock(next);
4898                 free_extent_buffer(next);
4899                 next = NULL;
4900                 *lookup_info = 1;
4901         }
4902
4903         if (!next) {
4904                 if (reada && level == 1)
4905                         reada_walk_down(trans, root, wc, path);
4906                 next = read_tree_block(fs_info, bytenr, generation, level - 1,
4907                                        &first_key);
4908                 if (IS_ERR(next)) {
4909                         return PTR_ERR(next);
4910                 } else if (!extent_buffer_uptodate(next)) {
4911                         free_extent_buffer(next);
4912                         return -EIO;
4913                 }
4914                 btrfs_tree_lock(next);
4915                 btrfs_set_lock_blocking_write(next);
4916         }
4917
4918         level--;
4919         ASSERT(level == btrfs_header_level(next));
4920         if (level != btrfs_header_level(next)) {
4921                 btrfs_err(root->fs_info, "mismatched level");
4922                 ret = -EIO;
4923                 goto out_unlock;
4924         }
4925         path->nodes[level] = next;
4926         path->slots[level] = 0;
4927         path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING;
4928         wc->level = level;
4929         if (wc->level == 1)
4930                 wc->reada_slot = 0;
4931         return 0;
4932 skip:
4933         wc->refs[level - 1] = 0;
4934         wc->flags[level - 1] = 0;
4935         if (wc->stage == DROP_REFERENCE) {
4936                 if (wc->flags[level] & BTRFS_BLOCK_FLAG_FULL_BACKREF) {
4937                         parent = path->nodes[level]->start;
4938                 } else {
4939                         ASSERT(root->root_key.objectid ==
4940                                btrfs_header_owner(path->nodes[level]));
4941                         if (root->root_key.objectid !=
4942                             btrfs_header_owner(path->nodes[level])) {
4943                                 btrfs_err(root->fs_info,
4944                                                 "mismatched block owner");
4945                                 ret = -EIO;
4946                                 goto out_unlock;
4947                         }
4948                         parent = 0;
4949                 }
4950
4951                 /*
4952                  * If we had a drop_progress we need to verify the refs are set
4953                  * as expected.  If we find our ref then we know that from here
4954                  * on out everything should be correct, and we can clear the
4955                  * ->restarted flag.
4956                  */
4957                 if (wc->restarted) {
4958                         ret = check_ref_exists(trans, root, bytenr, parent,
4959                                                level - 1);
4960                         if (ret < 0)
4961                                 goto out_unlock;
4962                         if (ret == 0)
4963                                 goto no_delete;
4964                         ret = 0;
4965                         wc->restarted = 0;
4966                 }
4967
4968                 /*
4969                  * Reloc tree doesn't contribute to qgroup numbers, and we have
4970                  * already accounted them at merge time (replace_path),
4971                  * thus we could skip expensive subtree trace here.
4972                  */
4973                 if (root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID &&
4974                     need_account) {
4975                         ret = btrfs_qgroup_trace_subtree(trans, next,
4976                                                          generation, level - 1);
4977                         if (ret) {
4978                                 btrfs_err_rl(fs_info,
4979                                              "Error %d accounting shared subtree. Quota is out of sync, rescan required.",
4980                                              ret);
4981                         }
4982                 }
4983
4984                 /*
4985                  * We need to update the next key in our walk control so we can
4986                  * update the drop_progress key accordingly.  We don't care if
4987                  * find_next_key doesn't find a key because that means we're at
4988                  * the end and are going to clean up now.
4989                  */
4990                 wc->drop_level = level;
4991                 find_next_key(path, level, &wc->drop_progress);
4992
4993                 btrfs_init_generic_ref(&ref, BTRFS_DROP_DELAYED_REF, bytenr,
4994                                        fs_info->nodesize, parent);
4995                 btrfs_init_tree_ref(&ref, level - 1, root->root_key.objectid);
4996                 ret = btrfs_free_extent(trans, &ref);
4997                 if (ret)
4998                         goto out_unlock;
4999         }
5000 no_delete:
5001         *lookup_info = 1;
5002         ret = 1;
5003
5004 out_unlock:
5005         btrfs_tree_unlock(next);
5006         free_extent_buffer(next);
5007
5008         return ret;
5009 }
5010
5011 /*
5012  * helper to process tree block while walking up the tree.
5013  *
5014  * when wc->stage == DROP_REFERENCE, this function drops
5015  * reference count on the block.
5016  *
5017  * when wc->stage == UPDATE_BACKREF, this function changes
5018  * wc->stage back to DROP_REFERENCE if we changed wc->stage
5019  * to UPDATE_BACKREF previously while processing the block.
5020  *
5021  * NOTE: return value 1 means we should stop walking up.
5022  */
5023 static noinline int walk_up_proc(struct btrfs_trans_handle *trans,
5024                                  struct btrfs_root *root,
5025                                  struct btrfs_path *path,
5026                                  struct walk_control *wc)
5027 {
5028         struct btrfs_fs_info *fs_info = root->fs_info;
5029         int ret;
5030         int level = wc->level;
5031         struct extent_buffer *eb = path->nodes[level];
5032         u64 parent = 0;
5033
5034         if (wc->stage == UPDATE_BACKREF) {
5035                 BUG_ON(wc->shared_level < level);
5036                 if (level < wc->shared_level)
5037                         goto out;
5038
5039                 ret = find_next_key(path, level + 1, &wc->update_progress);
5040                 if (ret > 0)
5041                         wc->update_ref = 0;
5042
5043                 wc->stage = DROP_REFERENCE;
5044                 wc->shared_level = -1;
5045                 path->slots[level] = 0;
5046
5047                 /*
5048                  * check reference count again if the block isn't locked.
5049                  * we should start walking down the tree again if reference
5050                  * count is one.
5051                  */
5052                 if (!path->locks[level]) {
5053                         BUG_ON(level == 0);
5054                         btrfs_tree_lock(eb);
5055                         btrfs_set_lock_blocking_write(eb);
5056                         path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING;
5057
5058                         ret = btrfs_lookup_extent_info(trans, fs_info,
5059                                                        eb->start, level, 1,
5060                                                        &wc->refs[level],
5061                                                        &wc->flags[level]);
5062                         if (ret < 0) {
5063                                 btrfs_tree_unlock_rw(eb, path->locks[level]);
5064                                 path->locks[level] = 0;
5065                                 return ret;
5066                         }
5067                         BUG_ON(wc->refs[level] == 0);
5068                         if (wc->refs[level] == 1) {
5069                                 btrfs_tree_unlock_rw(eb, path->locks[level]);
5070                                 path->locks[level] = 0;
5071                                 return 1;
5072                         }
5073                 }
5074         }
5075
5076         /* wc->stage == DROP_REFERENCE */
5077         BUG_ON(wc->refs[level] > 1 && !path->locks[level]);
5078
5079         if (wc->refs[level] == 1) {
5080                 if (level == 0) {
5081                         if (wc->flags[level] & BTRFS_BLOCK_FLAG_FULL_BACKREF)
5082                                 ret = btrfs_dec_ref(trans, root, eb, 1);
5083                         else
5084                                 ret = btrfs_dec_ref(trans, root, eb, 0);
5085                         BUG_ON(ret); /* -ENOMEM */
5086                         if (is_fstree(root->root_key.objectid)) {
5087                                 ret = btrfs_qgroup_trace_leaf_items(trans, eb);
5088                                 if (ret) {
5089                                         btrfs_err_rl(fs_info,
5090         "error %d accounting leaf items, quota is out of sync, rescan required",
5091                                              ret);
5092                                 }
5093                         }
5094                 }
5095                 /* make block locked assertion in btrfs_clean_tree_block happy */
5096                 if (!path->locks[level] &&
5097                     btrfs_header_generation(eb) == trans->transid) {
5098                         btrfs_tree_lock(eb);
5099                         btrfs_set_lock_blocking_write(eb);
5100                         path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING;
5101                 }
5102                 btrfs_clean_tree_block(eb);
5103         }
5104
5105         if (eb == root->node) {
5106                 if (wc->flags[level] & BTRFS_BLOCK_FLAG_FULL_BACKREF)
5107                         parent = eb->start;
5108                 else if (root->root_key.objectid != btrfs_header_owner(eb))
5109                         goto owner_mismatch;
5110         } else {
5111                 if (wc->flags[level + 1] & BTRFS_BLOCK_FLAG_FULL_BACKREF)
5112                         parent = path->nodes[level + 1]->start;
5113                 else if (root->root_key.objectid !=
5114                          btrfs_header_owner(path->nodes[level + 1]))
5115                         goto owner_mismatch;
5116         }
5117
5118         btrfs_free_tree_block(trans, root, eb, parent, wc->refs[level] == 1);
5119 out:
5120         wc->refs[level] = 0;
5121         wc->flags[level] = 0;
5122         return 0;
5123
5124 owner_mismatch:
5125         btrfs_err_rl(fs_info, "unexpected tree owner, have %llu expect %llu",
5126                      btrfs_header_owner(eb), root->root_key.objectid);
5127         return -EUCLEAN;
5128 }
5129
5130 static noinline int walk_down_tree(struct btrfs_trans_handle *trans,
5131                                    struct btrfs_root *root,
5132                                    struct btrfs_path *path,
5133                                    struct walk_control *wc)
5134 {
5135         int level = wc->level;
5136         int lookup_info = 1;
5137         int ret;
5138
5139         while (level >= 0) {
5140                 ret = walk_down_proc(trans, root, path, wc, lookup_info);
5141                 if (ret > 0)
5142                         break;
5143
5144                 if (level == 0)
5145                         break;
5146
5147                 if (path->slots[level] >=
5148                     btrfs_header_nritems(path->nodes[level]))
5149                         break;
5150
5151                 ret = do_walk_down(trans, root, path, wc, &lookup_info);
5152                 if (ret > 0) {
5153                         path->slots[level]++;
5154                         continue;
5155                 } else if (ret < 0)
5156                         return ret;
5157                 level = wc->level;
5158         }
5159         return 0;
5160 }
5161
5162 static noinline int walk_up_tree(struct btrfs_trans_handle *trans,
5163                                  struct btrfs_root *root,
5164                                  struct btrfs_path *path,
5165                                  struct walk_control *wc, int max_level)
5166 {
5167         int level = wc->level;
5168         int ret;
5169
5170         path->slots[level] = btrfs_header_nritems(path->nodes[level]);
5171         while (level < max_level && path->nodes[level]) {
5172                 wc->level = level;
5173                 if (path->slots[level] + 1 <
5174                     btrfs_header_nritems(path->nodes[level])) {
5175                         path->slots[level]++;
5176                         return 0;
5177                 } else {
5178                         ret = walk_up_proc(trans, root, path, wc);
5179                         if (ret > 0)
5180                                 return 0;
5181                         if (ret < 0)
5182                                 return ret;
5183
5184                         if (path->locks[level]) {
5185                                 btrfs_tree_unlock_rw(path->nodes[level],
5186                                                      path->locks[level]);
5187                                 path->locks[level] = 0;
5188                         }
5189                         free_extent_buffer(path->nodes[level]);
5190                         path->nodes[level] = NULL;
5191                         level++;
5192                 }
5193         }
5194         return 1;
5195 }
5196
5197 /*
5198  * drop a subvolume tree.
5199  *
5200  * this function traverses the tree freeing any blocks that only
5201  * referenced by the tree.
5202  *
5203  * when a shared tree block is found. this function decreases its
5204  * reference count by one. if update_ref is true, this function
5205  * also make sure backrefs for the shared block and all lower level
5206  * blocks are properly updated.
5207  *
5208  * If called with for_reloc == 0, may exit early with -EAGAIN
5209  */
5210 int btrfs_drop_snapshot(struct btrfs_root *root,
5211                          struct btrfs_block_rsv *block_rsv, int update_ref,
5212                          int for_reloc)
5213 {
5214         struct btrfs_fs_info *fs_info = root->fs_info;
5215         struct btrfs_path *path;
5216         struct btrfs_trans_handle *trans;
5217         struct btrfs_root *tree_root = fs_info->tree_root;
5218         struct btrfs_root_item *root_item = &root->root_item;
5219         struct walk_control *wc;
5220         struct btrfs_key key;
5221         int err = 0;
5222         int ret;
5223         int level;
5224         bool root_dropped = false;
5225
5226         btrfs_debug(fs_info, "Drop subvolume %llu", root->root_key.objectid);
5227
5228         path = btrfs_alloc_path();
5229         if (!path) {
5230                 err = -ENOMEM;
5231                 goto out;
5232         }
5233
5234         wc = kzalloc(sizeof(*wc), GFP_NOFS);
5235         if (!wc) {
5236                 btrfs_free_path(path);
5237                 err = -ENOMEM;
5238                 goto out;
5239         }
5240
5241         trans = btrfs_start_transaction(tree_root, 0);
5242         if (IS_ERR(trans)) {
5243                 err = PTR_ERR(trans);
5244                 goto out_free;
5245         }
5246
5247         err = btrfs_run_delayed_items(trans);
5248         if (err)
5249                 goto out_end_trans;
5250
5251         if (block_rsv)
5252                 trans->block_rsv = block_rsv;
5253
5254         /*
5255          * This will help us catch people modifying the fs tree while we're
5256          * dropping it.  It is unsafe to mess with the fs tree while it's being
5257          * dropped as we unlock the root node and parent nodes as we walk down
5258          * the tree, assuming nothing will change.  If something does change
5259          * then we'll have stale information and drop references to blocks we've
5260          * already dropped.
5261          */
5262         set_bit(BTRFS_ROOT_DELETING, &root->state);
5263         if (btrfs_disk_key_objectid(&root_item->drop_progress) == 0) {
5264                 level = btrfs_header_level(root->node);
5265                 path->nodes[level] = btrfs_lock_root_node(root);
5266                 btrfs_set_lock_blocking_write(path->nodes[level]);
5267                 path->slots[level] = 0;
5268                 path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING;
5269                 memset(&wc->update_progress, 0,
5270                        sizeof(wc->update_progress));
5271         } else {
5272                 btrfs_disk_key_to_cpu(&key, &root_item->drop_progress);
5273                 memcpy(&wc->update_progress, &key,
5274                        sizeof(wc->update_progress));
5275
5276                 level = root_item->drop_level;
5277                 BUG_ON(level == 0);
5278                 path->lowest_level = level;
5279                 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
5280                 path->lowest_level = 0;
5281                 if (ret < 0) {
5282                         err = ret;
5283                         goto out_end_trans;
5284                 }
5285                 WARN_ON(ret > 0);
5286
5287                 /*
5288                  * unlock our path, this is safe because only this
5289                  * function is allowed to delete this snapshot
5290                  */
5291                 btrfs_unlock_up_safe(path, 0);
5292
5293                 level = btrfs_header_level(root->node);
5294                 while (1) {
5295                         btrfs_tree_lock(path->nodes[level]);
5296                         btrfs_set_lock_blocking_write(path->nodes[level]);
5297                         path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING;
5298
5299                         ret = btrfs_lookup_extent_info(trans, fs_info,
5300                                                 path->nodes[level]->start,
5301                                                 level, 1, &wc->refs[level],
5302                                                 &wc->flags[level]);
5303                         if (ret < 0) {
5304                                 err = ret;
5305                                 goto out_end_trans;
5306                         }
5307                         BUG_ON(wc->refs[level] == 0);
5308
5309                         if (level == root_item->drop_level)
5310                                 break;
5311
5312                         btrfs_tree_unlock(path->nodes[level]);
5313                         path->locks[level] = 0;
5314                         WARN_ON(wc->refs[level] != 1);
5315                         level--;
5316                 }
5317         }
5318
5319         wc->restarted = test_bit(BTRFS_ROOT_DEAD_TREE, &root->state);
5320         wc->level = level;
5321         wc->shared_level = -1;
5322         wc->stage = DROP_REFERENCE;
5323         wc->update_ref = update_ref;
5324         wc->keep_locks = 0;
5325         wc->reada_count = BTRFS_NODEPTRS_PER_BLOCK(fs_info);
5326
5327         while (1) {
5328
5329                 ret = walk_down_tree(trans, root, path, wc);
5330                 if (ret < 0) {
5331                         err = ret;
5332                         break;
5333                 }
5334
5335                 ret = walk_up_tree(trans, root, path, wc, BTRFS_MAX_LEVEL);
5336                 if (ret < 0) {
5337                         err = ret;
5338                         break;
5339                 }
5340
5341                 if (ret > 0) {
5342                         BUG_ON(wc->stage != DROP_REFERENCE);
5343                         break;
5344                 }
5345
5346                 if (wc->stage == DROP_REFERENCE) {
5347                         wc->drop_level = wc->level;
5348                         btrfs_node_key_to_cpu(path->nodes[wc->drop_level],
5349                                               &wc->drop_progress,
5350                                               path->slots[wc->drop_level]);
5351                 }
5352                 btrfs_cpu_key_to_disk(&root_item->drop_progress,
5353                                       &wc->drop_progress);
5354                 root_item->drop_level = wc->drop_level;
5355
5356                 BUG_ON(wc->level == 0);
5357                 if (btrfs_should_end_transaction(trans) ||
5358                     (!for_reloc && btrfs_need_cleaner_sleep(fs_info))) {
5359                         ret = btrfs_update_root(trans, tree_root,
5360                                                 &root->root_key,
5361                                                 root_item);
5362                         if (ret) {
5363                                 btrfs_abort_transaction(trans, ret);
5364                                 err = ret;
5365                                 goto out_end_trans;
5366                         }
5367
5368                         btrfs_end_transaction_throttle(trans);
5369                         if (!for_reloc && btrfs_need_cleaner_sleep(fs_info)) {
5370                                 btrfs_debug(fs_info,
5371                                             "drop snapshot early exit");
5372                                 err = -EAGAIN;
5373                                 goto out_free;
5374                         }
5375
5376                         trans = btrfs_start_transaction(tree_root, 0);
5377                         if (IS_ERR(trans)) {
5378                                 err = PTR_ERR(trans);
5379                                 goto out_free;
5380                         }
5381                         if (block_rsv)
5382                                 trans->block_rsv = block_rsv;
5383                 }
5384         }
5385         btrfs_release_path(path);
5386         if (err)
5387                 goto out_end_trans;
5388
5389         ret = btrfs_del_root(trans, &root->root_key);
5390         if (ret) {
5391                 btrfs_abort_transaction(trans, ret);
5392                 err = ret;
5393                 goto out_end_trans;
5394         }
5395
5396         if (root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID) {
5397                 ret = btrfs_find_root(tree_root, &root->root_key, path,
5398                                       NULL, NULL);
5399                 if (ret < 0) {
5400                         btrfs_abort_transaction(trans, ret);
5401                         err = ret;
5402                         goto out_end_trans;
5403                 } else if (ret > 0) {
5404                         /* if we fail to delete the orphan item this time
5405                          * around, it'll get picked up the next time.
5406                          *
5407                          * The most common failure here is just -ENOENT.
5408                          */
5409                         btrfs_del_orphan_item(trans, tree_root,
5410                                               root->root_key.objectid);
5411                 }
5412         }
5413
5414         if (test_bit(BTRFS_ROOT_IN_RADIX, &root->state)) {
5415                 btrfs_add_dropped_root(trans, root);
5416         } else {
5417                 free_extent_buffer(root->node);
5418                 free_extent_buffer(root->commit_root);
5419                 btrfs_put_fs_root(root);
5420         }
5421         root_dropped = true;
5422 out_end_trans:
5423         btrfs_end_transaction_throttle(trans);
5424 out_free:
5425         kfree(wc);
5426         btrfs_free_path(path);
5427 out:
5428         /*
5429          * So if we need to stop dropping the snapshot for whatever reason we
5430          * need to make sure to add it back to the dead root list so that we
5431          * keep trying to do the work later.  This also cleans up roots if we
5432          * don't have it in the radix (like when we recover after a power fail
5433          * or unmount) so we don't leak memory.
5434          */
5435         if (!for_reloc && !root_dropped)
5436                 btrfs_add_dead_root(root);
5437         if (err && err != -EAGAIN)
5438                 btrfs_handle_fs_error(fs_info, err, NULL);
5439         return err;
5440 }
5441
5442 /*
5443  * drop subtree rooted at tree block 'node'.
5444  *
5445  * NOTE: this function will unlock and release tree block 'node'
5446  * only used by relocation code
5447  */
5448 int btrfs_drop_subtree(struct btrfs_trans_handle *trans,
5449                         struct btrfs_root *root,
5450                         struct extent_buffer *node,
5451                         struct extent_buffer *parent)
5452 {
5453         struct btrfs_fs_info *fs_info = root->fs_info;
5454         struct btrfs_path *path;
5455         struct walk_control *wc;
5456         int level;
5457         int parent_level;
5458         int ret = 0;
5459         int wret;
5460
5461         BUG_ON(root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID);
5462
5463         path = btrfs_alloc_path();
5464         if (!path)
5465                 return -ENOMEM;
5466
5467         wc = kzalloc(sizeof(*wc), GFP_NOFS);
5468         if (!wc) {
5469                 btrfs_free_path(path);
5470                 return -ENOMEM;
5471         }
5472
5473         btrfs_assert_tree_locked(parent);
5474         parent_level = btrfs_header_level(parent);
5475         atomic_inc(&parent->refs);
5476         path->nodes[parent_level] = parent;
5477         path->slots[parent_level] = btrfs_header_nritems(parent);
5478
5479         btrfs_assert_tree_locked(node);
5480         level = btrfs_header_level(node);
5481         path->nodes[level] = node;
5482         path->slots[level] = 0;
5483         path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING;
5484
5485         wc->refs[parent_level] = 1;
5486         wc->flags[parent_level] = BTRFS_BLOCK_FLAG_FULL_BACKREF;
5487         wc->level = level;
5488         wc->shared_level = -1;
5489         wc->stage = DROP_REFERENCE;
5490         wc->update_ref = 0;
5491         wc->keep_locks = 1;
5492         wc->reada_count = BTRFS_NODEPTRS_PER_BLOCK(fs_info);
5493
5494         while (1) {
5495                 wret = walk_down_tree(trans, root, path, wc);
5496                 if (wret < 0) {
5497                         ret = wret;
5498                         break;
5499                 }
5500
5501                 wret = walk_up_tree(trans, root, path, wc, parent_level);
5502                 if (wret < 0)
5503                         ret = wret;
5504                 if (wret != 0)
5505                         break;
5506         }
5507
5508         kfree(wc);
5509         btrfs_free_path(path);
5510         return ret;
5511 }
5512
5513 /*
5514  * helper to account the unused space of all the readonly block group in the
5515  * space_info. takes mirrors into account.
5516  */
5517 u64 btrfs_account_ro_block_groups_free_space(struct btrfs_space_info *sinfo)
5518 {
5519         struct btrfs_block_group *block_group;
5520         u64 free_bytes = 0;
5521         int factor;
5522
5523         /* It's df, we don't care if it's racy */
5524         if (list_empty(&sinfo->ro_bgs))
5525                 return 0;
5526
5527         spin_lock(&sinfo->lock);
5528         list_for_each_entry(block_group, &sinfo->ro_bgs, ro_list) {
5529                 spin_lock(&block_group->lock);
5530
5531                 if (!block_group->ro) {
5532                         spin_unlock(&block_group->lock);
5533                         continue;
5534                 }
5535
5536                 factor = btrfs_bg_type_to_factor(block_group->flags);
5537                 free_bytes += (block_group->length -
5538                                block_group->used) * factor;
5539
5540                 spin_unlock(&block_group->lock);
5541         }
5542         spin_unlock(&sinfo->lock);
5543
5544         return free_bytes;
5545 }
5546
5547 int btrfs_error_unpin_extent_range(struct btrfs_fs_info *fs_info,
5548                                    u64 start, u64 end)
5549 {
5550         return unpin_extent_range(fs_info, start, end, false);
5551 }
5552
5553 /*
5554  * It used to be that old block groups would be left around forever.
5555  * Iterating over them would be enough to trim unused space.  Since we
5556  * now automatically remove them, we also need to iterate over unallocated
5557  * space.
5558  *
5559  * We don't want a transaction for this since the discard may take a
5560  * substantial amount of time.  We don't require that a transaction be
5561  * running, but we do need to take a running transaction into account
5562  * to ensure that we're not discarding chunks that were released or
5563  * allocated in the current transaction.
5564  *
5565  * Holding the chunks lock will prevent other threads from allocating
5566  * or releasing chunks, but it won't prevent a running transaction
5567  * from committing and releasing the memory that the pending chunks
5568  * list head uses.  For that, we need to take a reference to the
5569  * transaction and hold the commit root sem.  We only need to hold
5570  * it while performing the free space search since we have already
5571  * held back allocations.
5572  */
5573 static int btrfs_trim_free_extents(struct btrfs_device *device, u64 *trimmed)
5574 {
5575         u64 start = SZ_1M, len = 0, end = 0;
5576         int ret;
5577
5578         *trimmed = 0;
5579
5580         /* Discard not supported = nothing to do. */
5581         if (!blk_queue_discard(bdev_get_queue(device->bdev)))
5582                 return 0;
5583
5584         /* Not writable = nothing to do. */
5585         if (!test_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state))
5586                 return 0;
5587
5588         /* No free space = nothing to do. */
5589         if (device->total_bytes <= device->bytes_used)
5590                 return 0;
5591
5592         ret = 0;
5593
5594         while (1) {
5595                 struct btrfs_fs_info *fs_info = device->fs_info;
5596                 u64 bytes;
5597
5598                 ret = mutex_lock_interruptible(&fs_info->chunk_mutex);
5599                 if (ret)
5600                         break;
5601
5602                 find_first_clear_extent_bit(&device->alloc_state, start,
5603                                             &start, &end,
5604                                             CHUNK_TRIMMED | CHUNK_ALLOCATED);
5605
5606                 /* Ensure we skip the reserved area in the first 1M */
5607                 start = max_t(u64, start, SZ_1M);
5608
5609                 /*
5610                  * If find_first_clear_extent_bit find a range that spans the
5611                  * end of the device it will set end to -1, in this case it's up
5612                  * to the caller to trim the value to the size of the device.
5613                  */
5614                 end = min(end, device->total_bytes - 1);
5615
5616                 len = end - start + 1;
5617
5618                 /* We didn't find any extents */
5619                 if (!len) {
5620                         mutex_unlock(&fs_info->chunk_mutex);
5621                         ret = 0;
5622                         break;
5623                 }
5624
5625                 ret = btrfs_issue_discard(device->bdev, start, len,
5626                                           &bytes);
5627                 if (!ret)
5628                         set_extent_bits(&device->alloc_state, start,
5629                                         start + bytes - 1,
5630                                         CHUNK_TRIMMED);
5631                 mutex_unlock(&fs_info->chunk_mutex);
5632
5633                 if (ret)
5634                         break;
5635
5636                 start += len;
5637                 *trimmed += bytes;
5638
5639                 if (fatal_signal_pending(current)) {
5640                         ret = -ERESTARTSYS;
5641                         break;
5642                 }
5643
5644                 cond_resched();
5645         }
5646
5647         return ret;
5648 }
5649
5650 /*
5651  * Trim the whole filesystem by:
5652  * 1) trimming the free space in each block group
5653  * 2) trimming the unallocated space on each device
5654  *
5655  * This will also continue trimming even if a block group or device encounters
5656  * an error.  The return value will be the last error, or 0 if nothing bad
5657  * happens.
5658  */
5659 int btrfs_trim_fs(struct btrfs_fs_info *fs_info, struct fstrim_range *range)
5660 {
5661         struct btrfs_block_group *cache = NULL;
5662         struct btrfs_device *device;
5663         struct list_head *devices;
5664         u64 group_trimmed;
5665         u64 range_end = U64_MAX;
5666         u64 start;
5667         u64 end;
5668         u64 trimmed = 0;
5669         u64 bg_failed = 0;
5670         u64 dev_failed = 0;
5671         int bg_ret = 0;
5672         int dev_ret = 0;
5673         int ret = 0;
5674
5675         /*
5676          * Check range overflow if range->len is set.
5677          * The default range->len is U64_MAX.
5678          */
5679         if (range->len != U64_MAX &&
5680             check_add_overflow(range->start, range->len, &range_end))
5681                 return -EINVAL;
5682
5683         cache = btrfs_lookup_first_block_group(fs_info, range->start);
5684         for (; cache; cache = btrfs_next_block_group(cache)) {
5685                 if (cache->start >= range_end) {
5686                         btrfs_put_block_group(cache);
5687                         break;
5688                 }
5689
5690                 start = max(range->start, cache->start);
5691                 end = min(range_end, cache->start + cache->length);
5692
5693                 if (end - start >= range->minlen) {
5694                         if (!btrfs_block_group_done(cache)) {
5695                                 ret = btrfs_cache_block_group(cache, 0);
5696                                 if (ret) {
5697                                         bg_failed++;
5698                                         bg_ret = ret;
5699                                         continue;
5700                                 }
5701                                 ret = btrfs_wait_block_group_cache_done(cache);
5702                                 if (ret) {
5703                                         bg_failed++;
5704                                         bg_ret = ret;
5705                                         continue;
5706                                 }
5707                         }
5708                         ret = btrfs_trim_block_group(cache,
5709                                                      &group_trimmed,
5710                                                      start,
5711                                                      end,
5712                                                      range->minlen);
5713
5714                         trimmed += group_trimmed;
5715                         if (ret) {
5716                                 bg_failed++;
5717                                 bg_ret = ret;
5718                                 continue;
5719                         }
5720                 }
5721         }
5722
5723         if (bg_failed)
5724                 btrfs_warn(fs_info,
5725                         "failed to trim %llu block group(s), last error %d",
5726                         bg_failed, bg_ret);
5727         mutex_lock(&fs_info->fs_devices->device_list_mutex);
5728         devices = &fs_info->fs_devices->devices;
5729         list_for_each_entry(device, devices, dev_list) {
5730                 ret = btrfs_trim_free_extents(device, &group_trimmed);
5731                 if (ret) {
5732                         dev_failed++;
5733                         dev_ret = ret;
5734                         break;
5735                 }
5736
5737                 trimmed += group_trimmed;
5738         }
5739         mutex_unlock(&fs_info->fs_devices->device_list_mutex);
5740
5741         if (dev_failed)
5742                 btrfs_warn(fs_info,
5743                         "failed to trim %llu device(s), last error %d",
5744                         dev_failed, dev_ret);
5745         range->len = trimmed;
5746         if (bg_ret)
5747                 return bg_ret;
5748         return dev_ret;
5749 }
5750
5751 /*
5752  * btrfs_{start,end}_write_no_snapshotting() are similar to
5753  * mnt_{want,drop}_write(), they are used to prevent some tasks from writing
5754  * data into the page cache through nocow before the subvolume is snapshoted,
5755  * but flush the data into disk after the snapshot creation, or to prevent
5756  * operations while snapshotting is ongoing and that cause the snapshot to be
5757  * inconsistent (writes followed by expanding truncates for example).
5758  */
5759 void btrfs_end_write_no_snapshotting(struct btrfs_root *root)
5760 {
5761         percpu_counter_dec(&root->subv_writers->counter);
5762         cond_wake_up(&root->subv_writers->wait);
5763 }
5764
5765 int btrfs_start_write_no_snapshotting(struct btrfs_root *root)
5766 {
5767         if (atomic_read(&root->will_be_snapshotted))
5768                 return 0;
5769
5770         percpu_counter_inc(&root->subv_writers->counter);
5771         /*
5772          * Make sure counter is updated before we check for snapshot creation.
5773          */
5774         smp_mb();
5775         if (atomic_read(&root->will_be_snapshotted)) {
5776                 btrfs_end_write_no_snapshotting(root);
5777                 return 0;
5778         }
5779         return 1;
5780 }
5781
5782 void btrfs_wait_for_snapshot_creation(struct btrfs_root *root)
5783 {
5784         while (true) {
5785                 int ret;
5786
5787                 ret = btrfs_start_write_no_snapshotting(root);
5788                 if (ret)
5789                         break;
5790                 wait_var_event(&root->will_be_snapshotted,
5791                                !atomic_read(&root->will_be_snapshotted));
5792         }
5793 }