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->excluded_extents, start, end,
  68                         EXTENT_UPTODATE);
  69         return 0;
  70 }
  71
  72 void btrfs_free_excluded_extents(struct btrfs_block_group *cache)
  73 {
  74         struct btrfs_fs_info *fs_info = cache->fs_info;
  75         u64 start, end;
  76
  77         start = cache->start;
  78         end = start + cache->length - 1;
  79
  80         clear_extent_bits(&fs_info->excluded_extents, start, end,
  81                           EXTENT_UPTODATE);
  82 }
  83
  84 static u64 generic_ref_to_space_flags(struct btrfs_ref *ref)
  85 {
  86         if (ref->type == BTRFS_REF_METADATA) {
  87                 if (ref->tree_ref.root == BTRFS_CHUNK_TREE_OBJECTID)
  88                         return BTRFS_BLOCK_GROUP_SYSTEM;
  89                 else
  90                         return BTRFS_BLOCK_GROUP_METADATA;
  91         }
  92         return BTRFS_BLOCK_GROUP_DATA;
  93 }
  94
  95 static void add_pinned_bytes(struct btrfs_fs_info *fs_info,
  96                              struct btrfs_ref *ref)
  97 {
  98         struct btrfs_space_info *space_info;
  99         u64 flags = generic_ref_to_space_flags(ref);
 100
 101         space_info = btrfs_find_space_info(fs_info, flags);
 102         ASSERT(space_info);
 103         percpu_counter_add_batch(&space_info->total_bytes_pinned, ref->len,
 104                     BTRFS_TOTAL_BYTES_PINNED_BATCH);
 105 }
 106
 107 static void sub_pinned_bytes(struct btrfs_fs_info *fs_info,
 108                              struct btrfs_ref *ref)
 109 {
 110         struct btrfs_space_info *space_info;
 111         u64 flags = generic_ref_to_space_flags(ref);
 112
 113         space_info = btrfs_find_space_info(fs_info, flags);
 114         ASSERT(space_info);
 115         percpu_counter_add_batch(&space_info->total_bytes_pinned, -ref->len,
 116                     BTRFS_TOTAL_BYTES_PINNED_BATCH);
 117 }
 118
 119 /* simple helper to search for an existing data extent at a given offset */
 120 int btrfs_lookup_data_extent(struct btrfs_fs_info *fs_info, u64 start, u64 len)
 121 {
 122         int ret;
 123         struct btrfs_key key;
 124         struct btrfs_path *path;
 125
 126         path = btrfs_alloc_path();
 127         if (!path)
 128                 return -ENOMEM;
 129
 130         key.objectid = start;
 131         key.offset = len;
 132         key.type = BTRFS_EXTENT_ITEM_KEY;
 133         ret = btrfs_search_slot(NULL, fs_info->extent_root, &key, path, 0, 0);
 134         btrfs_free_path(path);
 135         return ret;
 136 }
 137
 138 /*
 139  * helper function to lookup reference count and flags of a tree block.
 140  *
 141  * the head node for delayed ref is used to store the sum of all the
 142  * reference count modifications queued up in the rbtree. the head
 143  * node may also store the extent flags to set. This way you can check
 144  * to see what the reference count and extent flags would be if all of
 145  * the delayed refs are not processed.
 146  */
 147 int btrfs_lookup_extent_info(struct btrfs_trans_handle *trans,
 148                              struct btrfs_fs_info *fs_info, u64 bytenr,
 149                              u64 offset, int metadata, u64 *refs, u64 *flags)
 150 {
 151         struct btrfs_delayed_ref_head *head;
 152         struct btrfs_delayed_ref_root *delayed_refs;
 153         struct btrfs_path *path;
 154         struct btrfs_extent_item *ei;
 155         struct extent_buffer *leaf;
 156         struct btrfs_key key;
 157         u32 item_size;
 158         u64 num_refs;
 159         u64 extent_flags;
 160         int ret;
 161
 162         /*
 163          * If we don't have skinny metadata, don't bother doing anything
 164          * different
 165          */
 166         if (metadata && !btrfs_fs_incompat(fs_info, SKINNY_METADATA)) {
 167                 offset = fs_info->nodesize;
 168                 metadata = 0;
 169         }
 170
 171         path = btrfs_alloc_path();
 172         if (!path)
 173                 return -ENOMEM;
 174
 175         if (!trans) {
 176                 path->skip_locking = 1;
 177                 path->search_commit_root = 1;
 178         }
 179
 180 search_again:
 181         key.objectid = bytenr;
 182         key.offset = offset;
 183         if (metadata)
 184                 key.type = BTRFS_METADATA_ITEM_KEY;
 185         else
 186                 key.type = BTRFS_EXTENT_ITEM_KEY;
 187
 188         ret = btrfs_search_slot(trans, fs_info->extent_root, &key, path, 0, 0);
 189         if (ret < 0)
 190                 goto out_free;
 191
 192         if (ret > 0 && metadata && key.type == BTRFS_METADATA_ITEM_KEY) {
 193                 if (path->slots[0]) {
 194                         path->slots[0]--;
 195                         btrfs_item_key_to_cpu(path->nodes[0], &key,
 196                                               path->slots[0]);
 197                         if (key.objectid == bytenr &&
 198                             key.type == BTRFS_EXTENT_ITEM_KEY &&
 199                             key.offset == fs_info->nodesize)
 200                                 ret = 0;
 201                 }
 202         }
 203
 204         if (ret == 0) {
 205                 leaf = path->nodes[0];
 206                 item_size = btrfs_item_size_nr(leaf, path->slots[0]);
 207                 if (item_size >= sizeof(*ei)) {
 208                         ei = btrfs_item_ptr(leaf, path->slots[0],
 209                                             struct btrfs_extent_item);
 210                         num_refs = btrfs_extent_refs(leaf, ei);
 211                         extent_flags = btrfs_extent_flags(leaf, ei);
 212                 } else {
 213                         ret = -EINVAL;
 214                         btrfs_print_v0_err(fs_info);
 215                         if (trans)
 216                                 btrfs_abort_transaction(trans, ret);
 217                         else
 218                                 btrfs_handle_fs_error(fs_info, ret, NULL);
 219
 220                         goto out_free;
 221                 }
 222
 223                 BUG_ON(num_refs == 0);
 224         } else {
 225                 num_refs = 0;
 226                 extent_flags = 0;
 227                 ret = 0;
 228         }
 229
 230         if (!trans)
 231                 goto out;
 232
 233         delayed_refs = &trans->transaction->delayed_refs;
 234         spin_lock(&delayed_refs->lock);
 235         head = btrfs_find_delayed_ref_head(delayed_refs, bytenr);
 236         if (head) {
 237                 if (!mutex_trylock(&head->mutex)) {
 238                         refcount_inc(&head->refs);
 239                         spin_unlock(&delayed_refs->lock);
 240
 241                         btrfs_release_path(path);
 242
 243                         /*
 244                          * Mutex was contended, block until it's released and try
 245                          * again
 246                          */
 247                         mutex_lock(&head->mutex);
 248                         mutex_unlock(&head->mutex);
 249                         btrfs_put_delayed_ref_head(head);
 250                         goto search_again;
 251                 }
 252                 spin_lock(&head->lock);
 253                 if (head->extent_op && head->extent_op->update_flags)
 254                         extent_flags |= head->extent_op->flags_to_set;
 255                 else
 256                         BUG_ON(num_refs == 0);
 257
 258                 num_refs += head->ref_mod;
 259                 spin_unlock(&head->lock);
 260                 mutex_unlock(&head->mutex);
 261         }
 262         spin_unlock(&delayed_refs->lock);
 263 out:
 264         WARN_ON(num_refs == 0);
 265         if (refs)
 266                 *refs = num_refs;
 267         if (flags)
 268                 *flags = extent_flags;
 269 out_free:
 270         btrfs_free_path(path);
 271         return ret;
 272 }
 273
 274 /*
 275  * Back reference rules.  Back refs have three main goals:
 276  *
 277  * 1) differentiate between all holders of references to an extent so that
 278  *    when a reference is dropped we can make sure it was a valid reference
 279  *    before freeing the extent.
 280  *
 281  * 2) Provide enough information to quickly find the holders of an extent
 282  *    if we notice a given block is corrupted or bad.
 283  *
 284  * 3) Make it easy to migrate blocks for FS shrinking or storage pool
 285  *    maintenance.  This is actually the same as #2, but with a slightly
 286  *    different use case.
 287  *
 288  * There are two kinds of back refs. The implicit back refs is optimized
 289  * for pointers in non-shared tree blocks. For a given pointer in a block,
 290  * back refs of this kind provide information about the block's owner tree
 291  * and the pointer's key. These information allow us to find the block by
 292  * b-tree searching. The full back refs is for pointers in tree blocks not
 293  * referenced by their owner trees. The location of tree block is recorded
 294  * in the back refs. Actually the full back refs is generic, and can be
 295  * used in all cases the implicit back refs is used. The major shortcoming
 296  * of the full back refs is its overhead. Every time a tree block gets
 297  * COWed, we have to update back refs entry for all pointers in it.
 298  *
 299  * For a newly allocated tree block, we use implicit back refs for
 300  * pointers in it. This means most tree related operations only involve
 301  * implicit back refs. For a tree block created in old transaction, the
 302  * only way to drop a reference to it is COW it. So we can detect the
 303  * event that tree block loses its owner tree's reference and do the
 304  * back refs conversion.
 305  *
 306  * When a tree block is COWed through a tree, there are four cases:
 307  *
 308  * The reference count of the block is one and the tree is the block's
 309  * owner tree. Nothing to do in this case.
 310  *
 311  * The reference count of the block is one and the tree is not the
 312  * block's owner tree. In this case, full back refs is used for pointers
 313  * in the block. Remove these full back refs, add implicit back refs for
 314  * every pointers in the new block.
 315  *
 316  * The reference count of the block is greater than one and the tree is
 317  * the block's owner tree. In this case, implicit back refs is used for
 318  * pointers in the block. Add full back refs for every pointers in the
 319  * block, increase lower level extents' reference counts. The original
 320  * implicit back refs are entailed to the new block.
 321  *
 322  * The reference count of the block is greater than one and the tree is
 323  * not the block's owner tree. Add implicit back refs for every pointer in
 324  * the new block, increase lower level extents' reference count.
 325  *
 326  * Back Reference Key composing:
 327  *
 328  * The key objectid corresponds to the first byte in the extent,
 329  * The key type is used to differentiate between types of back refs.
 330  * There are different meanings of the key offset for different types
 331  * of back refs.
 332  *
 333  * File extents can be referenced by:
 334  *
 335  * - multiple snapshots, subvolumes, or different generations in one subvol
 336  * - different files inside a single subvolume
 337  * - different offsets inside a file (bookend extents in file.c)
 338  *
 339  * The extent ref structure for the implicit back refs has fields for:
 340  *
 341  * - Objectid of the subvolume root
 342  * - objectid of the file holding the reference
 343  * - original offset in the file
 344  * - how many bookend extents
 345  *
 346  * The key offset for the implicit back refs is hash of the first
 347  * three fields.
 348  *
 349  * The extent ref structure for the full back refs has field for:
 350  *
 351  * - number of pointers in the tree leaf
 352  *
 353  * The key offset for the implicit back refs is the first byte of
 354  * the tree leaf
 355  *
 356  * When a file extent is allocated, The implicit back refs is used.
 357  * the fields are filled in:
 358  *
 359  *     (root_key.objectid, inode objectid, offset in file, 1)
 360  *
 361  * When a file extent is removed file truncation, we find the
 362  * corresponding implicit back refs and check the following fields:
 363  *
 364  *     (btrfs_header_owner(leaf), inode objectid, offset in file)
 365  *
 366  * Btree extents can be referenced by:
 367  *
 368  * - Different subvolumes
 369  *
 370  * Both the implicit back refs and the full back refs for tree blocks
 371  * only consist of key. The key offset for the implicit back refs is
 372  * objectid of block's owner tree. The key offset for the full back refs
 373  * is the first byte of parent block.
 374  *
 375  * When implicit back refs is used, information about the lowest key and
 376  * level of the tree block are required. These information are stored in
 377  * tree block info structure.
 378  */
 379
 380 /*
 381  * is_data == BTRFS_REF_TYPE_BLOCK, tree block type is required,
 382  * is_data == BTRFS_REF_TYPE_DATA, data type is requiried,
 383  * is_data == BTRFS_REF_TYPE_ANY, either type is OK.
 384  */
 385 int btrfs_get_extent_inline_ref_type(const struct extent_buffer *eb,
 386                                      struct btrfs_extent_inline_ref *iref,
 387                                      enum btrfs_inline_ref_type is_data)
 388 {
 389         int type = btrfs_extent_inline_ref_type(eb, iref);
 390         u64 offset = btrfs_extent_inline_ref_offset(eb, iref);
 391
 392         if (type == BTRFS_TREE_BLOCK_REF_KEY ||
 393             type == BTRFS_SHARED_BLOCK_REF_KEY ||
 394             type == BTRFS_SHARED_DATA_REF_KEY ||
 395             type == BTRFS_EXTENT_DATA_REF_KEY) {
 396                 if (is_data == BTRFS_REF_TYPE_BLOCK) {
 397                         if (type == BTRFS_TREE_BLOCK_REF_KEY)
 398                                 return type;
 399                         if (type == BTRFS_SHARED_BLOCK_REF_KEY) {
 400                                 ASSERT(eb->fs_info);
 401                                 /*
 402                                  * Every shared one has parent tree
 403                                  * block, which must be aligned to
 404                                  * nodesize.
 405                                  */
 406                                 if (offset &&
 407                                     IS_ALIGNED(offset, eb->fs_info->nodesize))
 408                                         return type;
 409                         }
 410                 } else if (is_data == BTRFS_REF_TYPE_DATA) {
 411                         if (type == BTRFS_EXTENT_DATA_REF_KEY)
 412                                 return type;
 413                         if (type == BTRFS_SHARED_DATA_REF_KEY) {
 414                                 ASSERT(eb->fs_info);
 415                                 /*
 416                                  * Every shared one has parent tree
 417                                  * block, which must be aligned to
 418                                  * nodesize.
 419                                  */
 420                                 if (offset &&
 421                                     IS_ALIGNED(offset, eb->fs_info->nodesize))
 422                                         return type;
 423                         }
 424                 } else {
 425                         ASSERT(is_data == BTRFS_REF_TYPE_ANY);
 426                         return type;
 427                 }
 428         }
 429
 430         btrfs_print_leaf((struct extent_buffer *)eb);
 431         btrfs_err(eb->fs_info, "eb %llu invalid extent inline ref type %d",
 432                   eb->start, type);
 433         WARN_ON(1);
 434
 435         return BTRFS_REF_TYPE_INVALID;
 436 }
 437
 438 u64 hash_extent_data_ref(u64 root_objectid, u64 owner, u64 offset)
 439 {
 440         u32 high_crc = ~(u32)0;
 441         u32 low_crc = ~(u32)0;
 442         __le64 lenum;
 443
 444         lenum = cpu_to_le64(root_objectid);
 445         high_crc = btrfs_crc32c(high_crc, &lenum, sizeof(lenum));
 446         lenum = cpu_to_le64(owner);
 447         low_crc = btrfs_crc32c(low_crc, &lenum, sizeof(lenum));
 448         lenum = cpu_to_le64(offset);
 449         low_crc = btrfs_crc32c(low_crc, &lenum, sizeof(lenum));
 450
 451         return ((u64)high_crc << 31) ^ (u64)low_crc;
 452 }
 453
 454 static u64 hash_extent_data_ref_item(struct extent_buffer *leaf,
 455                                      struct btrfs_extent_data_ref *ref)
 456 {
 457         return hash_extent_data_ref(btrfs_extent_data_ref_root(leaf, ref),
 458                                     btrfs_extent_data_ref_objectid(leaf, ref),
 459                                     btrfs_extent_data_ref_offset(leaf, ref));
 460 }
 461
 462 static int match_extent_data_ref(struct extent_buffer *leaf,
 463                                  struct btrfs_extent_data_ref *ref,
 464                                  u64 root_objectid, u64 owner, u64 offset)
 465 {
 466         if (btrfs_extent_data_ref_root(leaf, ref) != root_objectid ||
 467             btrfs_extent_data_ref_objectid(leaf, ref) != owner ||
 468             btrfs_extent_data_ref_offset(leaf, ref) != offset)
 469                 return 0;
 470         return 1;
 471 }
 472
 473 static noinline int lookup_extent_data_ref(struct btrfs_trans_handle *trans,
 474                                            struct btrfs_path *path,
 475                                            u64 bytenr, u64 parent,
 476                                            u64 root_objectid,
 477                                            u64 owner, u64 offset)
 478 {
 479         struct btrfs_root *root = trans->fs_info->extent_root;
 480         struct btrfs_key key;
 481         struct btrfs_extent_data_ref *ref;
 482         struct extent_buffer *leaf;
 483         u32 nritems;
 484         int ret;
 485         int recow;
 486         int err = -ENOENT;
 487
 488         key.objectid = bytenr;
 489         if (parent) {
 490                 key.type = BTRFS_SHARED_DATA_REF_KEY;
 491                 key.offset = parent;
 492         } else {
 493                 key.type = BTRFS_EXTENT_DATA_REF_KEY;
 494                 key.offset = hash_extent_data_ref(root_objectid,
 495                                                   owner, offset);
 496         }
 497 again:
 498         recow = 0;
 499         ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
 500         if (ret < 0) {
 501                 err = ret;
 502                 goto fail;
 503         }
 504
 505         if (parent) {
 506                 if (!ret)
 507                         return 0;
 508                 goto fail;
 509         }
 510
 511         leaf = path->nodes[0];
 512         nritems = btrfs_header_nritems(leaf);
 513         while (1) {
 514                 if (path->slots[0] >= nritems) {
 515                         ret = btrfs_next_leaf(root, path);
 516                         if (ret < 0)
 517                                 err = ret;
 518                         if (ret)
 519                                 goto fail;
 520
 521                         leaf = path->nodes[0];
 522                         nritems = btrfs_header_nritems(leaf);
 523                         recow = 1;
 524                 }
 525
 526                 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
 527                 if (key.objectid != bytenr ||
 528                     key.type != BTRFS_EXTENT_DATA_REF_KEY)
 529                         goto fail;
 530
 531                 ref = btrfs_item_ptr(leaf, path->slots[0],
 532                                      struct btrfs_extent_data_ref);
 533
 534                 if (match_extent_data_ref(leaf, ref, root_objectid,
 535                                           owner, offset)) {
 536                         if (recow) {
 537                                 btrfs_release_path(path);
 538                                 goto again;
 539                         }
 540                         err = 0;
 541                         break;
 542                 }
 543                 path->slots[0]++;
 544         }
 545 fail:
 546         return err;
 547 }
 548
 549 static noinline int insert_extent_data_ref(struct btrfs_trans_handle *trans,
 550                                            struct btrfs_path *path,
 551                                            u64 bytenr, u64 parent,
 552                                            u64 root_objectid, u64 owner,
 553                                            u64 offset, int refs_to_add)
 554 {
 555         struct btrfs_root *root = trans->fs_info->extent_root;
 556         struct btrfs_key key;
 557         struct extent_buffer *leaf;
 558         u32 size;
 559         u32 num_refs;
 560         int ret;
 561
 562         key.objectid = bytenr;
 563         if (parent) {
 564                 key.type = BTRFS_SHARED_DATA_REF_KEY;
 565                 key.offset = parent;
 566                 size = sizeof(struct btrfs_shared_data_ref);
 567         } else {
 568                 key.type = BTRFS_EXTENT_DATA_REF_KEY;
 569                 key.offset = hash_extent_data_ref(root_objectid,
 570                                                   owner, offset);
 571                 size = sizeof(struct btrfs_extent_data_ref);
 572         }
 573
 574         ret = btrfs_insert_empty_item(trans, root, path, &key, size);
 575         if (ret && ret != -EEXIST)
 576                 goto fail;
 577
 578         leaf = path->nodes[0];
 579         if (parent) {
 580                 struct btrfs_shared_data_ref *ref;
 581                 ref = btrfs_item_ptr(leaf, path->slots[0],
 582                                      struct btrfs_shared_data_ref);
 583                 if (ret == 0) {
 584                         btrfs_set_shared_data_ref_count(leaf, ref, refs_to_add);
 585                 } else {
 586                         num_refs = btrfs_shared_data_ref_count(leaf, ref);
 587                         num_refs += refs_to_add;
 588                         btrfs_set_shared_data_ref_count(leaf, ref, num_refs);
 589                 }
 590         } else {
 591                 struct btrfs_extent_data_ref *ref;
 592                 while (ret == -EEXIST) {
 593                         ref = btrfs_item_ptr(leaf, path->slots[0],
 594                                              struct btrfs_extent_data_ref);
 595                         if (match_extent_data_ref(leaf, ref, root_objectid,
 596                                                   owner, offset))
 597                                 break;
 598                         btrfs_release_path(path);
 599                         key.offset++;
 600                         ret = btrfs_insert_empty_item(trans, root, path, &key,
 601                                                       size);
 602                         if (ret && ret != -EEXIST)
 603                                 goto fail;
 604
 605                         leaf = path->nodes[0];
 606                 }
 607                 ref = btrfs_item_ptr(leaf, path->slots[0],
 608                                      struct btrfs_extent_data_ref);
 609                 if (ret == 0) {
 610                         btrfs_set_extent_data_ref_root(leaf, ref,
 611                                                        root_objectid);
 612                         btrfs_set_extent_data_ref_objectid(leaf, ref, owner);
 613                         btrfs_set_extent_data_ref_offset(leaf, ref, offset);
 614                         btrfs_set_extent_data_ref_count(leaf, ref, refs_to_add);
 615                 } else {
 616                         num_refs = btrfs_extent_data_ref_count(leaf, ref);
 617                         num_refs += refs_to_add;
 618                         btrfs_set_extent_data_ref_count(leaf, ref, num_refs);
 619                 }
 620         }
 621         btrfs_mark_buffer_dirty(leaf);
 622         ret = 0;
 623 fail:
 624         btrfs_release_path(path);
 625         return ret;
 626 }
 627
 628 static noinline int remove_extent_data_ref(struct btrfs_trans_handle *trans,
 629                                            struct btrfs_path *path,
 630                                            int refs_to_drop, int *last_ref)
 631 {
 632         struct btrfs_key key;
 633         struct btrfs_extent_data_ref *ref1 = NULL;
 634         struct btrfs_shared_data_ref *ref2 = NULL;
 635         struct extent_buffer *leaf;
 636         u32 num_refs = 0;
 637         int ret = 0;
 638
 639         leaf = path->nodes[0];
 640         btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
 641
 642         if (key.type == BTRFS_EXTENT_DATA_REF_KEY) {
 643                 ref1 = btrfs_item_ptr(leaf, path->slots[0],
 644                                       struct btrfs_extent_data_ref);
 645                 num_refs = btrfs_extent_data_ref_count(leaf, ref1);
 646         } else if (key.type == BTRFS_SHARED_DATA_REF_KEY) {
 647                 ref2 = btrfs_item_ptr(leaf, path->slots[0],
 648                                       struct btrfs_shared_data_ref);
 649                 num_refs = btrfs_shared_data_ref_count(leaf, ref2);
 650         } else if (unlikely(key.type == BTRFS_EXTENT_REF_V0_KEY)) {
 651                 btrfs_print_v0_err(trans->fs_info);
 652                 btrfs_abort_transaction(trans, -EINVAL);
 653                 return -EINVAL;
 654         } else {
 655                 BUG();
 656         }
 657
 658         BUG_ON(num_refs < refs_to_drop);
 659         num_refs -= refs_to_drop;
 660
 661         if (num_refs == 0) {
 662                 ret = btrfs_del_item(trans, trans->fs_info->extent_root, path);
 663                 *last_ref = 1;
 664         } else {
 665                 if (key.type == BTRFS_EXTENT_DATA_REF_KEY)
 666                         btrfs_set_extent_data_ref_count(leaf, ref1, num_refs);
 667                 else if (key.type == BTRFS_SHARED_DATA_REF_KEY)
 668                         btrfs_set_shared_data_ref_count(leaf, ref2, num_refs);
 669                 btrfs_mark_buffer_dirty(leaf);
 670         }
 671         return ret;
 672 }
 673
 674 static noinline u32 extent_data_ref_count(struct btrfs_path *path,
 675                                           struct btrfs_extent_inline_ref *iref)
 676 {
 677         struct btrfs_key key;
 678         struct extent_buffer *leaf;
 679         struct btrfs_extent_data_ref *ref1;
 680         struct btrfs_shared_data_ref *ref2;
 681         u32 num_refs = 0;
 682         int type;
 683
 684         leaf = path->nodes[0];
 685         btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
 686
 687         BUG_ON(key.type == BTRFS_EXTENT_REF_V0_KEY);
 688         if (iref) {
 689                 /*
 690                  * If type is invalid, we should have bailed out earlier than
 691                  * this call.
 692                  */
 693                 type = btrfs_get_extent_inline_ref_type(leaf, iref, BTRFS_REF_TYPE_DATA);
 694                 ASSERT(type != BTRFS_REF_TYPE_INVALID);
 695                 if (type == BTRFS_EXTENT_DATA_REF_KEY) {
 696                         ref1 = (struct btrfs_extent_data_ref *)(&iref->offset);
 697                         num_refs = btrfs_extent_data_ref_count(leaf, ref1);
 698                 } else {
 699                         ref2 = (struct btrfs_shared_data_ref *)(iref + 1);
 700                         num_refs = btrfs_shared_data_ref_count(leaf, ref2);
 701                 }
 702         } else if (key.type == BTRFS_EXTENT_DATA_REF_KEY) {
 703                 ref1 = btrfs_item_ptr(leaf, path->slots[0],
 704                                       struct btrfs_extent_data_ref);
 705                 num_refs = btrfs_extent_data_ref_count(leaf, ref1);
 706         } else if (key.type == BTRFS_SHARED_DATA_REF_KEY) {
 707                 ref2 = btrfs_item_ptr(leaf, path->slots[0],
 708                                       struct btrfs_shared_data_ref);
 709                 num_refs = btrfs_shared_data_ref_count(leaf, ref2);
 710         } else {
 711                 WARN_ON(1);
 712         }
 713         return num_refs;
 714 }
 715
 716 static noinline int lookup_tree_block_ref(struct btrfs_trans_handle *trans,
 717                                           struct btrfs_path *path,
 718                                           u64 bytenr, u64 parent,
 719                                           u64 root_objectid)
 720 {
 721         struct btrfs_root *root = trans->fs_info->extent_root;
 722         struct btrfs_key key;
 723         int ret;
 724
 725         key.objectid = bytenr;
 726         if (parent) {
 727                 key.type = BTRFS_SHARED_BLOCK_REF_KEY;
 728                 key.offset = parent;
 729         } else {
 730                 key.type = BTRFS_TREE_BLOCK_REF_KEY;
 731                 key.offset = root_objectid;
 732         }
 733
 734         ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
 735         if (ret > 0)
 736                 ret = -ENOENT;
 737         return ret;
 738 }
 739
 740 static noinline int insert_tree_block_ref(struct btrfs_trans_handle *trans,
 741                                           struct btrfs_path *path,
 742                                           u64 bytenr, u64 parent,
 743                                           u64 root_objectid)
 744 {
 745         struct btrfs_key key;
 746         int ret;
 747
 748         key.objectid = bytenr;
 749         if (parent) {
 750                 key.type = BTRFS_SHARED_BLOCK_REF_KEY;
 751                 key.offset = parent;
 752         } else {
 753                 key.type = BTRFS_TREE_BLOCK_REF_KEY;
 754                 key.offset = root_objectid;
 755         }
 756
 757         ret = btrfs_insert_empty_item(trans, trans->fs_info->extent_root,
 758                                       path, &key, 0);
 759         btrfs_release_path(path);
 760         return ret;
 761 }
 762
 763 static inline int extent_ref_type(u64 parent, u64 owner)
 764 {
 765         int type;
 766         if (owner < BTRFS_FIRST_FREE_OBJECTID) {
 767                 if (parent > 0)
 768                         type = BTRFS_SHARED_BLOCK_REF_KEY;
 769                 else
 770                         type = BTRFS_TREE_BLOCK_REF_KEY;
 771         } else {
 772                 if (parent > 0)
 773                         type = BTRFS_SHARED_DATA_REF_KEY;
 774                 else
 775                         type = BTRFS_EXTENT_DATA_REF_KEY;
 776         }
 777         return type;
 778 }
 779
 780 static int find_next_key(struct btrfs_path *path, int level,
 781                          struct btrfs_key *key)
 782
 783 {
 784         for (; level < BTRFS_MAX_LEVEL; level++) {
 785                 if (!path->nodes[level])
 786                         break;
 787                 if (path->slots[level] + 1 >=
 788                     btrfs_header_nritems(path->nodes[level]))
 789                         continue;
 790                 if (level == 0)
 791                         btrfs_item_key_to_cpu(path->nodes[level], key,
 792                                               path->slots[level] + 1);
 793                 else
 794                         btrfs_node_key_to_cpu(path->nodes[level], key,
 795                                               path->slots[level] + 1);
 796                 return 0;
 797         }
 798         return 1;
 799 }
 800
 801 /*
 802  * look for inline back ref. if back ref is found, *ref_ret is set
 803  * to the address of inline back ref, and 0 is returned.
 804  *
 805  * if back ref isn't found, *ref_ret is set to the address where it
 806  * should be inserted, and -ENOENT is returned.
 807  *
 808  * if insert is true and there are too many inline back refs, the path
 809  * points to the extent item, and -EAGAIN is returned.
 810  *
 811  * NOTE: inline back refs are ordered in the same way that back ref
 812  *       items in the tree are ordered.
 813  */
 814 static noinline_for_stack
 815 int lookup_inline_extent_backref(struct btrfs_trans_handle *trans,
 816                                  struct btrfs_path *path,
 817                                  struct btrfs_extent_inline_ref **ref_ret,
 818                                  u64 bytenr, u64 num_bytes,
 819                                  u64 parent, u64 root_objectid,
 820                                  u64 owner, u64 offset, int insert)
 821 {
 822         struct btrfs_fs_info *fs_info = trans->fs_info;
 823         struct btrfs_root *root = fs_info->extent_root;
 824         struct btrfs_key key;
 825         struct extent_buffer *leaf;
 826         struct btrfs_extent_item *ei;
 827         struct btrfs_extent_inline_ref *iref;
 828         u64 flags;
 829         u64 item_size;
 830         unsigned long ptr;
 831         unsigned long end;
 832         int extra_size;
 833         int type;
 834         int want;
 835         int ret;
 836         int err = 0;
 837         bool skinny_metadata = btrfs_fs_incompat(fs_info, SKINNY_METADATA);
 838         int needed;
 839
 840         key.objectid = bytenr;
 841         key.type = BTRFS_EXTENT_ITEM_KEY;
 842         key.offset = num_bytes;
 843
 844         want = extent_ref_type(parent, owner);
 845         if (insert) {
 846                 extra_size = btrfs_extent_inline_ref_size(want);
 847                 path->keep_locks = 1;
 848         } else
 849                 extra_size = -1;
 850
 851         /*
 852          * Owner is our level, so we can just add one to get the level for the
 853          * block we are interested in.
 854          */
 855         if (skinny_metadata && owner < BTRFS_FIRST_FREE_OBJECTID) {
 856                 key.type = BTRFS_METADATA_ITEM_KEY;
 857                 key.offset = owner;
 858         }
 859
 860 again:
 861         ret = btrfs_search_slot(trans, root, &key, path, extra_size, 1);
 862         if (ret < 0) {
 863                 err = ret;
 864                 goto out;
 865         }
 866
 867         /*
 868          * We may be a newly converted file system which still has the old fat
 869          * extent entries for metadata, so try and see if we have one of those.
 870          */
 871         if (ret > 0 && skinny_metadata) {
 872                 skinny_metadata = false;
 873                 if (path->slots[0]) {
 874                         path->slots[0]--;
 875                         btrfs_item_key_to_cpu(path->nodes[0], &key,
 876                                               path->slots[0]);
 877                         if (key.objectid == bytenr &&
 878                             key.type == BTRFS_EXTENT_ITEM_KEY &&
 879                             key.offset == num_bytes)
 880                                 ret = 0;
 881                 }
 882                 if (ret) {
 883                         key.objectid = bytenr;
 884                         key.type = BTRFS_EXTENT_ITEM_KEY;
 885                         key.offset = num_bytes;
 886                         btrfs_release_path(path);
 887                         goto again;
 888                 }
 889         }
 890
 891         if (ret && !insert) {
 892                 err = -ENOENT;
 893                 goto out;
 894         } else if (WARN_ON(ret)) {
 895                 err = -EIO;
 896                 goto out;
 897         }
 898
 899         leaf = path->nodes[0];
 900         item_size = btrfs_item_size_nr(leaf, path->slots[0]);
 901         if (unlikely(item_size < sizeof(*ei))) {
 902                 err = -EINVAL;
 903                 btrfs_print_v0_err(fs_info);
 904                 btrfs_abort_transaction(trans, err);
 905                 goto out;
 906         }
 907
 908         ei = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
 909         flags = btrfs_extent_flags(leaf, ei);
 910
 911         ptr = (unsigned long)(ei + 1);
 912         end = (unsigned long)ei + item_size;
 913
 914         if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK && !skinny_metadata) {
 915                 ptr += sizeof(struct btrfs_tree_block_info);
 916                 BUG_ON(ptr > end);
 917         }
 918
 919         if (owner >= BTRFS_FIRST_FREE_OBJECTID)
 920                 needed = BTRFS_REF_TYPE_DATA;
 921         else
 922                 needed = BTRFS_REF_TYPE_BLOCK;
 923
 924         err = -ENOENT;
 925         while (1) {
 926                 if (ptr >= end) {
 927                         WARN_ON(ptr > end);
 928                         break;
 929                 }
 930                 iref = (struct btrfs_extent_inline_ref *)ptr;
 931                 type = btrfs_get_extent_inline_ref_type(leaf, iref, needed);
 932                 if (type == BTRFS_REF_TYPE_INVALID) {
 933                         err = -EUCLEAN;
 934                         goto out;
 935                 }
 936
 937                 if (want < type)
 938                         break;
 939                 if (want > type) {
 940                         ptr += btrfs_extent_inline_ref_size(type);
 941                         continue;
 942                 }
 943
 944                 if (type == BTRFS_EXTENT_DATA_REF_KEY) {
 945                         struct btrfs_extent_data_ref *dref;
 946                         dref = (struct btrfs_extent_data_ref *)(&iref->offset);
 947                         if (match_extent_data_ref(leaf, dref, root_objectid,
 948                                                   owner, offset)) {
 949                                 err = 0;
 950                                 break;
 951                         }
 952                         if (hash_extent_data_ref_item(leaf, dref) <
 953                             hash_extent_data_ref(root_objectid, owner, offset))
 954                                 break;
 955                 } else {
 956                         u64 ref_offset;
 957                         ref_offset = btrfs_extent_inline_ref_offset(leaf, iref);
 958                         if (parent > 0) {
 959                                 if (parent == ref_offset) {
 960                                         err = 0;
 961                                         break;
 962                                 }
 963                                 if (ref_offset < parent)
 964                                         break;
 965                         } else {
 966                                 if (root_objectid == ref_offset) {
 967                                         err = 0;
 968                                         break;
 969                                 }
 970                                 if (ref_offset < root_objectid)
 971                                         break;
 972                         }
 973                 }
 974                 ptr += btrfs_extent_inline_ref_size(type);
 975         }
 976         if (err == -ENOENT && insert) {
 977                 if (item_size + extra_size >=
 978                     BTRFS_MAX_EXTENT_ITEM_SIZE(root)) {
 979                         err = -EAGAIN;
 980                         goto out;
 981                 }
 982                 /*
 983                  * To add new inline back ref, we have to make sure
 984                  * there is no corresponding back ref item.
 985                  * For simplicity, we just do not add new inline back
 986                  * ref if there is any kind of item for this block
 987                  */
 988                 if (find_next_key(path, 0, &key) == 0 &&
 989                     key.objectid == bytenr &&
 990                     key.type < BTRFS_BLOCK_GROUP_ITEM_KEY) {
 991                         err = -EAGAIN;
 992                         goto out;
 993                 }
 994         }
 995         *ref_ret = (struct btrfs_extent_inline_ref *)ptr;
 996 out:
 997         if (insert) {
 998                 path->keep_locks = 0;
 999                 btrfs_unlock_up_safe(path, 1);
1000         }
1001         return err;
1002 }
1003
1004 /*
1005  * helper to add new inline back ref
1006  */
1007 static noinline_for_stack
1008 void setup_inline_extent_backref(struct btrfs_fs_info *fs_info,
1009                                  struct btrfs_path *path,
1010                                  struct btrfs_extent_inline_ref *iref,
1011                                  u64 parent, u64 root_objectid,
1012                                  u64 owner, u64 offset, int refs_to_add,
1013                                  struct btrfs_delayed_extent_op *extent_op)
1014 {
1015         struct extent_buffer *leaf;
1016         struct btrfs_extent_item *ei;
1017         unsigned long ptr;
1018         unsigned long end;
1019         unsigned long item_offset;
1020         u64 refs;
1021         int size;
1022         int type;
1023
1024         leaf = path->nodes[0];
1025         ei = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
1026         item_offset = (unsigned long)iref - (unsigned long)ei;
1027
1028         type = extent_ref_type(parent, owner);
1029         size = btrfs_extent_inline_ref_size(type);
1030
1031         btrfs_extend_item(path, size);
1032
1033         ei = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
1034         refs = btrfs_extent_refs(leaf, ei);
1035         refs += refs_to_add;
1036         btrfs_set_extent_refs(leaf, ei, refs);
1037         if (extent_op)
1038                 __run_delayed_extent_op(extent_op, leaf, ei);
1039
1040         ptr = (unsigned long)ei + item_offset;
1041         end = (unsigned long)ei + btrfs_item_size_nr(leaf, path->slots[0]);
1042         if (ptr < end - size)
1043                 memmove_extent_buffer(leaf, ptr + size, ptr,
1044                                       end - size - ptr);
1045
1046         iref = (struct btrfs_extent_inline_ref *)ptr;
1047         btrfs_set_extent_inline_ref_type(leaf, iref, type);
1048         if (type == BTRFS_EXTENT_DATA_REF_KEY) {
1049                 struct btrfs_extent_data_ref *dref;
1050                 dref = (struct btrfs_extent_data_ref *)(&iref->offset);
1051                 btrfs_set_extent_data_ref_root(leaf, dref, root_objectid);
1052                 btrfs_set_extent_data_ref_objectid(leaf, dref, owner);
1053                 btrfs_set_extent_data_ref_offset(leaf, dref, offset);
1054                 btrfs_set_extent_data_ref_count(leaf, dref, refs_to_add);
1055         } else if (type == BTRFS_SHARED_DATA_REF_KEY) {
1056                 struct btrfs_shared_data_ref *sref;
1057                 sref = (struct btrfs_shared_data_ref *)(iref + 1);
1058                 btrfs_set_shared_data_ref_count(leaf, sref, refs_to_add);
1059                 btrfs_set_extent_inline_ref_offset(leaf, iref, parent);
1060         } else if (type == BTRFS_SHARED_BLOCK_REF_KEY) {
1061                 btrfs_set_extent_inline_ref_offset(leaf, iref, parent);
1062         } else {
1063                 btrfs_set_extent_inline_ref_offset(leaf, iref, root_objectid);
1064         }
1065         btrfs_mark_buffer_dirty(leaf);
1066 }
1067
1068 static int lookup_extent_backref(struct btrfs_trans_handle *trans,
1069                                  struct btrfs_path *path,
1070                                  struct btrfs_extent_inline_ref **ref_ret,
1071                                  u64 bytenr, u64 num_bytes, u64 parent,
1072                                  u64 root_objectid, u64 owner, u64 offset)
1073 {
1074         int ret;
1075
1076         ret = lookup_inline_extent_backref(trans, path, ref_ret, bytenr,
1077                                            num_bytes, parent, root_objectid,
1078                                            owner, offset, 0);
1079         if (ret != -ENOENT)
1080                 return ret;
1081
1082         btrfs_release_path(path);
1083         *ref_ret = NULL;
1084
1085         if (owner < BTRFS_FIRST_FREE_OBJECTID) {
1086                 ret = lookup_tree_block_ref(trans, path, bytenr, parent,
1087                                             root_objectid);
1088         } else {
1089                 ret = lookup_extent_data_ref(trans, path, bytenr, parent,
1090                                              root_objectid, owner, offset);
1091         }
1092         return ret;
1093 }
1094
1095 /*
1096  * helper to update/remove inline back ref
1097  */
1098 static noinline_for_stack
1099 void update_inline_extent_backref(struct btrfs_path *path,
1100                                   struct btrfs_extent_inline_ref *iref,
1101                                   int refs_to_mod,
1102                                   struct btrfs_delayed_extent_op *extent_op,
1103                                   int *last_ref)
1104 {
1105         struct extent_buffer *leaf = path->nodes[0];
1106         struct btrfs_extent_item *ei;
1107         struct btrfs_extent_data_ref *dref = NULL;
1108         struct btrfs_shared_data_ref *sref = NULL;
1109         unsigned long ptr;
1110         unsigned long end;
1111         u32 item_size;
1112         int size;
1113         int type;
1114         u64 refs;
1115
1116         ei = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
1117         refs = btrfs_extent_refs(leaf, ei);
1118         WARN_ON(refs_to_mod < 0 && refs + refs_to_mod <= 0);
1119         refs += refs_to_mod;
1120         btrfs_set_extent_refs(leaf, ei, refs);
1121         if (extent_op)
1122                 __run_delayed_extent_op(extent_op, leaf, ei);
1123
1124         /*
1125          * If type is invalid, we should have bailed out after
1126          * lookup_inline_extent_backref().
1127          */
1128         type = btrfs_get_extent_inline_ref_type(leaf, iref, BTRFS_REF_TYPE_ANY);
1129         ASSERT(type != BTRFS_REF_TYPE_INVALID);
1130
1131         if (type == BTRFS_EXTENT_DATA_REF_KEY) {
1132                 dref = (struct btrfs_extent_data_ref *)(&iref->offset);
1133                 refs = btrfs_extent_data_ref_count(leaf, dref);
1134         } else if (type == BTRFS_SHARED_DATA_REF_KEY) {
1135                 sref = (struct btrfs_shared_data_ref *)(iref + 1);
1136                 refs = btrfs_shared_data_ref_count(leaf, sref);
1137         } else {
1138                 refs = 1;
1139                 BUG_ON(refs_to_mod != -1);
1140         }
1141
1142         BUG_ON(refs_to_mod < 0 && refs < -refs_to_mod);
1143         refs += refs_to_mod;
1144
1145         if (refs > 0) {
1146                 if (type == BTRFS_EXTENT_DATA_REF_KEY)
1147                         btrfs_set_extent_data_ref_count(leaf, dref, refs);
1148                 else
1149                         btrfs_set_shared_data_ref_count(leaf, sref, refs);
1150         } else {
1151                 *last_ref = 1;
1152                 size =  btrfs_extent_inline_ref_size(type);
1153                 item_size = btrfs_item_size_nr(leaf, path->slots[0]);
1154                 ptr = (unsigned long)iref;
1155                 end = (unsigned long)ei + item_size;
1156                 if (ptr + size < end)
1157                         memmove_extent_buffer(leaf, ptr, ptr + size,
1158                                               end - ptr - size);
1159                 item_size -= size;
1160                 btrfs_truncate_item(path, item_size, 1);
1161         }
1162         btrfs_mark_buffer_dirty(leaf);
1163 }
1164
1165 static noinline_for_stack
1166 int insert_inline_extent_backref(struct btrfs_trans_handle *trans,
1167                                  struct btrfs_path *path,
1168                                  u64 bytenr, u64 num_bytes, u64 parent,
1169                                  u64 root_objectid, u64 owner,
1170                                  u64 offset, int refs_to_add,
1171                                  struct btrfs_delayed_extent_op *extent_op)
1172 {
1173         struct btrfs_extent_inline_ref *iref;
1174         int ret;
1175
1176         ret = lookup_inline_extent_backref(trans, path, &iref, bytenr,
1177                                            num_bytes, parent, root_objectid,
1178                                            owner, offset, 1);
1179         if (ret == 0) {
1180                 BUG_ON(owner < BTRFS_FIRST_FREE_OBJECTID);
1181                 update_inline_extent_backref(path, iref, refs_to_add,
1182                                              extent_op, NULL);
1183         } else if (ret == -ENOENT) {
1184                 setup_inline_extent_backref(trans->fs_info, path, iref, parent,
1185                                             root_objectid, owner, offset,
1186                                             refs_to_add, extent_op);
1187                 ret = 0;
1188         }
1189         return ret;
1190 }
1191
1192 static int remove_extent_backref(struct btrfs_trans_handle *trans,
1193                                  struct btrfs_path *path,
1194                                  struct btrfs_extent_inline_ref *iref,
1195                                  int refs_to_drop, int is_data, int *last_ref)
1196 {
1197         int ret = 0;
1198
1199         BUG_ON(!is_data && refs_to_drop != 1);
1200         if (iref) {
1201                 update_inline_extent_backref(path, iref, -refs_to_drop, NULL,
1202                                              last_ref);
1203         } else if (is_data) {
1204                 ret = remove_extent_data_ref(trans, path, refs_to_drop,
1205                                              last_ref);
1206         } else {
1207                 *last_ref = 1;
1208                 ret = btrfs_del_item(trans, trans->fs_info->extent_root, path);
1209         }
1210         return ret;
1211 }
1212
1213 static int btrfs_issue_discard(struct block_device *bdev, u64 start, u64 len,
1214                                u64 *discarded_bytes)
1215 {
1216         int j, ret = 0;
1217         u64 bytes_left, end;
1218         u64 aligned_start = ALIGN(start, 1 << 9);
1219
1220         if (WARN_ON(start != aligned_start)) {
1221                 len -= aligned_start - start;
1222                 len = round_down(len, 1 << 9);
1223                 start = aligned_start;
1224         }
1225
1226         *discarded_bytes = 0;
1227
1228         if (!len)
1229                 return 0;
1230
1231         end = start + len;
1232         bytes_left = len;
1233
1234         /* Skip any superblocks on this device. */
1235         for (j = 0; j < BTRFS_SUPER_MIRROR_MAX; j++) {
1236                 u64 sb_start = btrfs_sb_offset(j);
1237                 u64 sb_end = sb_start + BTRFS_SUPER_INFO_SIZE;
1238                 u64 size = sb_start - start;
1239
1240                 if (!in_range(sb_start, start, bytes_left) &&
1241                     !in_range(sb_end, start, bytes_left) &&
1242                     !in_range(start, sb_start, BTRFS_SUPER_INFO_SIZE))
1243                         continue;
1244
1245                 /*
1246                  * Superblock spans beginning of range.  Adjust start and
1247                  * try again.
1248                  */
1249                 if (sb_start <= start) {
1250                         start += sb_end - start;
1251                         if (start > end) {
1252                                 bytes_left = 0;
1253                                 break;
1254                         }
1255                         bytes_left = end - start;
1256                         continue;
1257                 }
1258
1259                 if (size) {
1260                         ret = blkdev_issue_discard(bdev, start >> 9, size >> 9,
1261                                                    GFP_NOFS, 0);
1262                         if (!ret)
1263                                 *discarded_bytes += size;
1264                         else if (ret != -EOPNOTSUPP)
1265                                 return ret;
1266                 }
1267
1268                 start = sb_end;
1269                 if (start > end) {
1270                         bytes_left = 0;
1271                         break;
1272                 }
1273                 bytes_left = end - start;
1274         }
1275
1276         if (bytes_left) {
1277                 ret = blkdev_issue_discard(bdev, start >> 9, bytes_left >> 9,
1278                                            GFP_NOFS, 0);
1279                 if (!ret)
1280                         *discarded_bytes += bytes_left;
1281         }
1282         return ret;
1283 }
1284
1285 int btrfs_discard_extent(struct btrfs_fs_info *fs_info, u64 bytenr,
1286                          u64 num_bytes, u64 *actual_bytes)
1287 {
1288         int ret = 0;
1289         u64 discarded_bytes = 0;
1290         u64 end = bytenr + num_bytes;
1291         u64 cur = bytenr;
1292         struct btrfs_bio *bbio = NULL;
1293
1294
1295         /*
1296          * Avoid races with device replace and make sure our bbio has devices
1297          * associated to its stripes that don't go away while we are discarding.
1298          */
1299         btrfs_bio_counter_inc_blocked(fs_info);
1300         while (cur < end) {
1301                 struct btrfs_bio_stripe *stripe;
1302                 int i;
1303
1304                 num_bytes = end - cur;
1305                 /* Tell the block device(s) that the sectors can be discarded */
1306                 ret = btrfs_map_block(fs_info, BTRFS_MAP_DISCARD, cur,
1307                                       &num_bytes, &bbio, 0);
1308                 /*
1309                  * Error can be -ENOMEM, -ENOENT (no such chunk mapping) or
1310                  * -EOPNOTSUPP. For any such error, @num_bytes is not updated,
1311                  * thus we can't continue anyway.
1312                  */
1313                 if (ret < 0)
1314                         goto out;
1315
1316                 stripe = bbio->stripes;
1317                 for (i = 0; i < bbio->num_stripes; i++, stripe++) {
1318                         u64 bytes;
1319                         struct request_queue *req_q;
1320
1321                         if (!stripe->dev->bdev) {
1322                                 ASSERT(btrfs_test_opt(fs_info, DEGRADED));
1323                                 continue;
1324                         }
1325                         req_q = bdev_get_queue(stripe->dev->bdev);
1326                         if (!blk_queue_discard(req_q))
1327                                 continue;
1328
1329                         ret = btrfs_issue_discard(stripe->dev->bdev,
1330                                                   stripe->physical,
1331                                                   stripe->length,
1332                                                   &bytes);
1333                         if (!ret) {
1334                                 discarded_bytes += bytes;
1335                         } else if (ret != -EOPNOTSUPP) {
1336                                 /*
1337                                  * Logic errors or -ENOMEM, or -EIO, but
1338                                  * unlikely to happen.
1339                                  *
1340                                  * And since there are two loops, explicitly
1341                                  * go to out to avoid confusion.
1342                                  */
1343                                 btrfs_put_bbio(bbio);
1344                                 goto out;
1345                         }
1346
1347                         /*
1348                          * Just in case we get back EOPNOTSUPP for some reason,
1349                          * just ignore the return value so we don't screw up
1350                          * people calling discard_extent.
1351                          */
1352                         ret = 0;
1353                 }
1354                 btrfs_put_bbio(bbio);
1355                 cur += num_bytes;
1356         }
1357 out:
1358         btrfs_bio_counter_dec(fs_info);
1359
1360         if (actual_bytes)
1361                 *actual_bytes = discarded_bytes;
1362
1363
1364         if (ret == -EOPNOTSUPP)
1365                 ret = 0;
1366         return ret;
1367 }
1368
1369 /* Can return -ENOMEM */
1370 int btrfs_inc_extent_ref(struct btrfs_trans_handle *trans,
1371                          struct btrfs_ref *generic_ref)
1372 {
1373         struct btrfs_fs_info *fs_info = trans->fs_info;
1374         int old_ref_mod, new_ref_mod;
1375         int ret;
1376
1377         ASSERT(generic_ref->type != BTRFS_REF_NOT_SET &&
1378                generic_ref->action);
1379         BUG_ON(generic_ref->type == BTRFS_REF_METADATA &&
1380                generic_ref->tree_ref.root == BTRFS_TREE_LOG_OBJECTID);
1381
1382         if (generic_ref->type == BTRFS_REF_METADATA)
1383                 ret = btrfs_add_delayed_tree_ref(trans, generic_ref,
1384                                 NULL, &old_ref_mod, &new_ref_mod);
1385         else
1386                 ret = btrfs_add_delayed_data_ref(trans, generic_ref, 0,
1387                                                  &old_ref_mod, &new_ref_mod);
1388
1389         btrfs_ref_tree_mod(fs_info, generic_ref);
1390
1391         if (ret == 0 && old_ref_mod < 0 && new_ref_mod >= 0)
1392                 sub_pinned_bytes(fs_info, generic_ref);
1393
1394         return ret;
1395 }
1396
1397 /*
1398  * __btrfs_inc_extent_ref - insert backreference for a given extent
1399  *
1400  * @trans:          Handle of transaction
1401  *
1402  * @node:           The delayed ref node used to get the bytenr/length for
1403  *                  extent whose references are incremented.
1404  *
1405  * @parent:         If this is a shared extent (BTRFS_SHARED_DATA_REF_KEY/
1406  *                  BTRFS_SHARED_BLOCK_REF_KEY) then it holds the logical
1407  *                  bytenr of the parent block. Since new extents are always
1408  *                  created with indirect references, this will only be the case
1409  *                  when relocating a shared extent. In that case, root_objectid
1410  *                  will be BTRFS_TREE_RELOC_OBJECTID. Otheriwse, parent must
1411  *                  be 0
1412  *
1413  * @root_objectid:  The id of the root where this modification has originated,
1414  *                  this can be either one of the well-known metadata trees or
1415  *                  the subvolume id which references this extent.
1416  *
1417  * @owner:          For data extents it is the inode number of the owning file.
1418  *                  For metadata extents this parameter holds the level in the
1419  *                  tree of the extent.
1420  *
1421  * @offset:         For metadata extents the offset is ignored and is currently
1422  *                  always passed as 0. For data extents it is the fileoffset
1423  *                  this extent belongs to.
1424  *
1425  * @refs_to_add     Number of references to add
1426  *
1427  * @extent_op       Pointer to a structure, holding information necessary when
1428  *                  updating a tree block's flags
1429  *
1430  */
1431 static int __btrfs_inc_extent_ref(struct btrfs_trans_handle *trans,
1432                                   struct btrfs_delayed_ref_node *node,
1433                                   u64 parent, u64 root_objectid,
1434                                   u64 owner, u64 offset, int refs_to_add,
1435                                   struct btrfs_delayed_extent_op *extent_op)
1436 {
1437         struct btrfs_path *path;
1438         struct extent_buffer *leaf;
1439         struct btrfs_extent_item *item;
1440         struct btrfs_key key;
1441         u64 bytenr = node->bytenr;
1442         u64 num_bytes = node->num_bytes;
1443         u64 refs;
1444         int ret;
1445
1446         path = btrfs_alloc_path();
1447         if (!path)
1448                 return -ENOMEM;
1449
1450         path->leave_spinning = 1;
1451         /* this will setup the path even if it fails to insert the back ref */
1452         ret = insert_inline_extent_backref(trans, path, bytenr, num_bytes,
1453                                            parent, root_objectid, owner,
1454                                            offset, refs_to_add, extent_op);
1455         if ((ret < 0 && ret != -EAGAIN) || !ret)
1456                 goto out;
1457
1458         /*
1459          * Ok we had -EAGAIN which means we didn't have space to insert and
1460          * inline extent ref, so just update the reference count and add a
1461          * normal backref.
1462          */
1463         leaf = path->nodes[0];
1464         btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
1465         item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
1466         refs = btrfs_extent_refs(leaf, item);
1467         btrfs_set_extent_refs(leaf, item, refs + refs_to_add);
1468         if (extent_op)
1469                 __run_delayed_extent_op(extent_op, leaf, item);
1470
1471         btrfs_mark_buffer_dirty(leaf);
1472         btrfs_release_path(path);
1473
1474         path->leave_spinning = 1;
1475         /* now insert the actual backref */
1476         if (owner < BTRFS_FIRST_FREE_OBJECTID) {
1477                 BUG_ON(refs_to_add != 1);
1478                 ret = insert_tree_block_ref(trans, path, bytenr, parent,
1479                                             root_objectid);
1480         } else {
1481                 ret = insert_extent_data_ref(trans, path, bytenr, parent,
1482                                              root_objectid, owner, offset,
1483                                              refs_to_add);
1484         }
1485         if (ret)
1486                 btrfs_abort_transaction(trans, ret);
1487 out:
1488         btrfs_free_path(path);
1489         return ret;
1490 }
1491
1492 static int run_delayed_data_ref(struct btrfs_trans_handle *trans,
1493                                 struct btrfs_delayed_ref_node *node,
1494                                 struct btrfs_delayed_extent_op *extent_op,
1495                                 int insert_reserved)
1496 {
1497         int ret = 0;
1498         struct btrfs_delayed_data_ref *ref;
1499         struct btrfs_key ins;
1500         u64 parent = 0;
1501         u64 ref_root = 0;
1502         u64 flags = 0;
1503
1504         ins.objectid = node->bytenr;
1505         ins.offset = node->num_bytes;
1506         ins.type = BTRFS_EXTENT_ITEM_KEY;
1507
1508         ref = btrfs_delayed_node_to_data_ref(node);
1509         trace_run_delayed_data_ref(trans->fs_info, node, ref, node->action);
1510
1511         if (node->type == BTRFS_SHARED_DATA_REF_KEY)
1512                 parent = ref->parent;
1513         ref_root = ref->root;
1514
1515         if (node->action == BTRFS_ADD_DELAYED_REF && insert_reserved) {
1516                 if (extent_op)
1517                         flags |= extent_op->flags_to_set;
1518                 ret = alloc_reserved_file_extent(trans, parent, ref_root,
1519                                                  flags, ref->objectid,
1520                                                  ref->offset, &ins,
1521                                                  node->ref_mod);
1522         } else if (node->action == BTRFS_ADD_DELAYED_REF) {
1523                 ret = __btrfs_inc_extent_ref(trans, node, parent, ref_root,
1524                                              ref->objectid, ref->offset,
1525                                              node->ref_mod, extent_op);
1526         } else if (node->action == BTRFS_DROP_DELAYED_REF) {
1527                 ret = __btrfs_free_extent(trans, node, parent,
1528                                           ref_root, ref->objectid,
1529                                           ref->offset, node->ref_mod,
1530                                           extent_op);
1531         } else {
1532                 BUG();
1533         }
1534         return ret;
1535 }
1536
1537 static void __run_delayed_extent_op(struct btrfs_delayed_extent_op *extent_op,
1538                                     struct extent_buffer *leaf,
1539                                     struct btrfs_extent_item *ei)
1540 {
1541         u64 flags = btrfs_extent_flags(leaf, ei);
1542         if (extent_op->update_flags) {
1543                 flags |= extent_op->flags_to_set;
1544                 btrfs_set_extent_flags(leaf, ei, flags);
1545         }
1546
1547         if (extent_op->update_key) {
1548                 struct btrfs_tree_block_info *bi;
1549                 BUG_ON(!(flags & BTRFS_EXTENT_FLAG_TREE_BLOCK));
1550                 bi = (struct btrfs_tree_block_info *)(ei + 1);
1551                 btrfs_set_tree_block_key(leaf, bi, &extent_op->key);
1552         }
1553 }
1554
1555 static int run_delayed_extent_op(struct btrfs_trans_handle *trans,
1556                                  struct btrfs_delayed_ref_head *head,
1557                                  struct btrfs_delayed_extent_op *extent_op)
1558 {
1559         struct btrfs_fs_info *fs_info = trans->fs_info;
1560         struct btrfs_key key;
1561         struct btrfs_path *path;
1562         struct btrfs_extent_item *ei;
1563         struct extent_buffer *leaf;
1564         u32 item_size;
1565         int ret;
1566         int err = 0;
1567         int metadata = !extent_op->is_data;
1568
1569         if (TRANS_ABORTED(trans))
1570                 return 0;
1571
1572         if (metadata && !btrfs_fs_incompat(fs_info, SKINNY_METADATA))
1573                 metadata = 0;
1574
1575         path = btrfs_alloc_path();
1576         if (!path)
1577                 return -ENOMEM;
1578
1579         key.objectid = head->bytenr;
1580
1581         if (metadata) {
1582                 key.type = BTRFS_METADATA_ITEM_KEY;
1583                 key.offset = extent_op->level;
1584         } else {
1585                 key.type = BTRFS_EXTENT_ITEM_KEY;
1586                 key.offset = head->num_bytes;
1587         }
1588
1589 again:
1590         path->leave_spinning = 1;
1591         ret = btrfs_search_slot(trans, fs_info->extent_root, &key, path, 0, 1);
1592         if (ret < 0) {
1593                 err = ret;
1594                 goto out;
1595         }
1596         if (ret > 0) {
1597                 if (metadata) {
1598                         if (path->slots[0] > 0) {
1599                                 path->slots[0]--;
1600                                 btrfs_item_key_to_cpu(path->nodes[0], &key,
1601                                                       path->slots[0]);
1602                                 if (key.objectid == head->bytenr &&
1603                                     key.type == BTRFS_EXTENT_ITEM_KEY &&
1604                                     key.offset == head->num_bytes)
1605                                         ret = 0;
1606                         }
1607                         if (ret > 0) {
1608                                 btrfs_release_path(path);
1609                                 metadata = 0;
1610
1611                                 key.objectid = head->bytenr;
1612                                 key.offset = head->num_bytes;
1613                                 key.type = BTRFS_EXTENT_ITEM_KEY;
1614                                 goto again;
1615                         }
1616                 } else {
1617                         err = -EIO;
1618                         goto out;
1619                 }
1620         }
1621
1622         leaf = path->nodes[0];
1623         item_size = btrfs_item_size_nr(leaf, path->slots[0]);
1624
1625         if (unlikely(item_size < sizeof(*ei))) {
1626                 err = -EINVAL;
1627                 btrfs_print_v0_err(fs_info);
1628                 btrfs_abort_transaction(trans, err);
1629                 goto out;
1630         }
1631
1632         ei = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
1633         __run_delayed_extent_op(extent_op, leaf, ei);
1634
1635         btrfs_mark_buffer_dirty(leaf);
1636 out:
1637         btrfs_free_path(path);
1638         return err;
1639 }
1640
1641 static int run_delayed_tree_ref(struct btrfs_trans_handle *trans,
1642                                 struct btrfs_delayed_ref_node *node,
1643                                 struct btrfs_delayed_extent_op *extent_op,
1644                                 int insert_reserved)
1645 {
1646         int ret = 0;
1647         struct btrfs_delayed_tree_ref *ref;
1648         u64 parent = 0;
1649         u64 ref_root = 0;
1650
1651         ref = btrfs_delayed_node_to_tree_ref(node);
1652         trace_run_delayed_tree_ref(trans->fs_info, node, ref, node->action);
1653
1654         if (node->type == BTRFS_SHARED_BLOCK_REF_KEY)
1655                 parent = ref->parent;
1656         ref_root = ref->root;
1657
1658         if (node->ref_mod != 1) {
1659                 btrfs_err(trans->fs_info,
1660         "btree block(%llu) has %d references rather than 1: action %d ref_root %llu parent %llu",
1661                           node->bytenr, node->ref_mod, node->action, ref_root,
1662                           parent);
1663                 return -EIO;
1664         }
1665         if (node->action == BTRFS_ADD_DELAYED_REF && insert_reserved) {
1666                 BUG_ON(!extent_op || !extent_op->update_flags);
1667                 ret = alloc_reserved_tree_block(trans, node, extent_op);
1668         } else if (node->action == BTRFS_ADD_DELAYED_REF) {
1669                 ret = __btrfs_inc_extent_ref(trans, node, parent, ref_root,
1670                                              ref->level, 0, 1, extent_op);
1671         } else if (node->action == BTRFS_DROP_DELAYED_REF) {
1672                 ret = __btrfs_free_extent(trans, node, parent, ref_root,
1673                                           ref->level, 0, 1, extent_op);
1674         } else {
1675                 BUG();
1676         }
1677         return ret;
1678 }
1679
1680 /* helper function to actually process a single delayed ref entry */
1681 static int run_one_delayed_ref(struct btrfs_trans_handle *trans,
1682                                struct btrfs_delayed_ref_node *node,
1683                                struct btrfs_delayed_extent_op *extent_op,
1684                                int insert_reserved)
1685 {
1686         int ret = 0;
1687
1688         if (TRANS_ABORTED(trans)) {
1689                 if (insert_reserved)
1690                         btrfs_pin_extent(trans, node->bytenr, node->num_bytes, 1);
1691                 return 0;
1692         }
1693
1694         if (node->type == BTRFS_TREE_BLOCK_REF_KEY ||
1695             node->type == BTRFS_SHARED_BLOCK_REF_KEY)
1696                 ret = run_delayed_tree_ref(trans, node, extent_op,
1697                                            insert_reserved);
1698         else if (node->type == BTRFS_EXTENT_DATA_REF_KEY ||
1699                  node->type == BTRFS_SHARED_DATA_REF_KEY)
1700                 ret = run_delayed_data_ref(trans, node, extent_op,
1701                                            insert_reserved);
1702         else
1703                 BUG();
1704         if (ret && insert_reserved)
1705                 btrfs_pin_extent(trans, node->bytenr, node->num_bytes, 1);
1706         return ret;
1707 }
1708
1709 static inline struct btrfs_delayed_ref_node *
1710 select_delayed_ref(struct btrfs_delayed_ref_head *head)
1711 {
1712         struct btrfs_delayed_ref_node *ref;
1713
1714         if (RB_EMPTY_ROOT(&head->ref_tree.rb_root))
1715                 return NULL;
1716
1717         /*
1718          * Select a delayed ref of type BTRFS_ADD_DELAYED_REF first.
1719          * This is to prevent a ref count from going down to zero, which deletes
1720          * the extent item from the extent tree, when there still are references
1721          * to add, which would fail because they would not find the extent item.
1722          */
1723         if (!list_empty(&head->ref_add_list))
1724                 return list_first_entry(&head->ref_add_list,
1725                                 struct btrfs_delayed_ref_node, add_list);
1726
1727         ref = rb_entry(rb_first_cached(&head->ref_tree),
1728                        struct btrfs_delayed_ref_node, ref_node);
1729         ASSERT(list_empty(&ref->add_list));
1730         return ref;
1731 }
1732
1733 static void unselect_delayed_ref_head(struct btrfs_delayed_ref_root *delayed_refs,
1734                                       struct btrfs_delayed_ref_head *head)
1735 {
1736         spin_lock(&delayed_refs->lock);
1737         head->processing = 0;
1738         delayed_refs->num_heads_ready++;
1739         spin_unlock(&delayed_refs->lock);
1740         btrfs_delayed_ref_unlock(head);
1741 }
1742
1743 static struct btrfs_delayed_extent_op *cleanup_extent_op(
1744                                 struct btrfs_delayed_ref_head *head)
1745 {
1746         struct btrfs_delayed_extent_op *extent_op = head->extent_op;
1747
1748         if (!extent_op)
1749                 return NULL;
1750
1751         if (head->must_insert_reserved) {
1752                 head->extent_op = NULL;
1753                 btrfs_free_delayed_extent_op(extent_op);
1754                 return NULL;
1755         }
1756         return extent_op;
1757 }
1758
1759 static int run_and_cleanup_extent_op(struct btrfs_trans_handle *trans,
1760                                      struct btrfs_delayed_ref_head *head)
1761 {
1762         struct btrfs_delayed_extent_op *extent_op;
1763         int ret;
1764
1765         extent_op = cleanup_extent_op(head);
1766         if (!extent_op)
1767                 return 0;
1768         head->extent_op = NULL;
1769         spin_unlock(&head->lock);
1770         ret = run_delayed_extent_op(trans, head, extent_op);
1771         btrfs_free_delayed_extent_op(extent_op);
1772         return ret ? ret : 1;
1773 }
1774
1775 void btrfs_cleanup_ref_head_accounting(struct btrfs_fs_info *fs_info,
1776                                   struct btrfs_delayed_ref_root *delayed_refs,
1777                                   struct btrfs_delayed_ref_head *head)
1778 {
1779         int nr_items = 1;       /* Dropping this ref head update. */
1780
1781         if (head->total_ref_mod < 0) {
1782                 struct btrfs_space_info *space_info;
1783                 u64 flags;
1784
1785                 if (head->is_data)
1786                         flags = BTRFS_BLOCK_GROUP_DATA;
1787                 else if (head->is_system)
1788                         flags = BTRFS_BLOCK_GROUP_SYSTEM;
1789                 else
1790                         flags = BTRFS_BLOCK_GROUP_METADATA;
1791                 space_info = btrfs_find_space_info(fs_info, flags);
1792                 ASSERT(space_info);
1793                 percpu_counter_add_batch(&space_info->total_bytes_pinned,
1794                                    -head->num_bytes,
1795                                    BTRFS_TOTAL_BYTES_PINNED_BATCH);
1796
1797                 /*
1798                  * We had csum deletions accounted for in our delayed refs rsv,
1799                  * we need to drop the csum leaves for this update from our
1800                  * delayed_refs_rsv.
1801                  */
1802                 if (head->is_data) {
1803                         spin_lock(&delayed_refs->lock);
1804                         delayed_refs->pending_csums -= head->num_bytes;
1805                         spin_unlock(&delayed_refs->lock);
1806                         nr_items += btrfs_csum_bytes_to_leaves(fs_info,
1807                                 head->num_bytes);
1808                 }
1809         }
1810
1811         btrfs_delayed_refs_rsv_release(fs_info, nr_items);
1812 }
1813
1814 static int cleanup_ref_head(struct btrfs_trans_handle *trans,
1815                             struct btrfs_delayed_ref_head *head)
1816 {
1817
1818         struct btrfs_fs_info *fs_info = trans->fs_info;
1819         struct btrfs_delayed_ref_root *delayed_refs;
1820         int ret;
1821
1822         delayed_refs = &trans->transaction->delayed_refs;
1823
1824         ret = run_and_cleanup_extent_op(trans, head);
1825         if (ret < 0) {
1826                 unselect_delayed_ref_head(delayed_refs, head);
1827                 btrfs_debug(fs_info, "run_delayed_extent_op returned %d", ret);
1828                 return ret;
1829         } else if (ret) {
1830                 return ret;
1831         }
1832
1833         /*
1834          * Need to drop our head ref lock and re-acquire the delayed ref lock
1835          * and then re-check to make sure nobody got added.
1836          */
1837         spin_unlock(&head->lock);
1838         spin_lock(&delayed_refs->lock);
1839         spin_lock(&head->lock);
1840         if (!RB_EMPTY_ROOT(&head->ref_tree.rb_root) || head->extent_op) {
1841                 spin_unlock(&head->lock);
1842                 spin_unlock(&delayed_refs->lock);
1843                 return 1;
1844         }
1845         btrfs_delete_ref_head(delayed_refs, head);
1846         spin_unlock(&head->lock);
1847         spin_unlock(&delayed_refs->lock);
1848
1849         if (head->must_insert_reserved) {
1850                 btrfs_pin_extent(trans, head->bytenr, head->num_bytes, 1);
1851                 if (head->is_data) {
1852                         ret = btrfs_del_csums(trans, fs_info->csum_root,
1853                                               head->bytenr, head->num_bytes);
1854                 }
1855         }
1856
1857         btrfs_cleanup_ref_head_accounting(fs_info, delayed_refs, head);
1858
1859         trace_run_delayed_ref_head(fs_info, head, 0);
1860         btrfs_delayed_ref_unlock(head);
1861         btrfs_put_delayed_ref_head(head);
1862         return 0;
1863 }
1864
1865 static struct btrfs_delayed_ref_head *btrfs_obtain_ref_head(
1866                                         struct btrfs_trans_handle *trans)
1867 {
1868         struct btrfs_delayed_ref_root *delayed_refs =
1869                 &trans->transaction->delayed_refs;
1870         struct btrfs_delayed_ref_head *head = NULL;
1871         int ret;
1872
1873         spin_lock(&delayed_refs->lock);
1874         head = btrfs_select_ref_head(delayed_refs);
1875         if (!head) {
1876                 spin_unlock(&delayed_refs->lock);
1877                 return head;
1878         }
1879
1880         /*
1881          * Grab the lock that says we are going to process all the refs for
1882          * this head
1883          */
1884         ret = btrfs_delayed_ref_lock(delayed_refs, head);
1885         spin_unlock(&delayed_refs->lock);
1886
1887         /*
1888          * We may have dropped the spin lock to get the head mutex lock, and
1889          * that might have given someone else time to free the head.  If that's
1890          * true, it has been removed from our list and we can move on.
1891          */
1892         if (ret == -EAGAIN)
1893                 head = ERR_PTR(-EAGAIN);
1894
1895         return head;
1896 }
1897
1898 static int btrfs_run_delayed_refs_for_head(struct btrfs_trans_handle *trans,
1899                                     struct btrfs_delayed_ref_head *locked_ref,
1900                                     unsigned long *run_refs)
1901 {
1902         struct btrfs_fs_info *fs_info = trans->fs_info;
1903         struct btrfs_delayed_ref_root *delayed_refs;
1904         struct btrfs_delayed_extent_op *extent_op;
1905         struct btrfs_delayed_ref_node *ref;
1906         int must_insert_reserved = 0;
1907         int ret;
1908
1909         delayed_refs = &trans->transaction->delayed_refs;
1910
1911         lockdep_assert_held(&locked_ref->mutex);
1912         lockdep_assert_held(&locked_ref->lock);
1913
1914         while ((ref = select_delayed_ref(locked_ref))) {
1915                 if (ref->seq &&
1916                     btrfs_check_delayed_seq(fs_info, ref->seq)) {
1917                         spin_unlock(&locked_ref->lock);
1918                         unselect_delayed_ref_head(delayed_refs, locked_ref);
1919                         return -EAGAIN;
1920                 }
1921
1922                 (*run_refs)++;
1923                 ref->in_tree = 0;
1924                 rb_erase_cached(&ref->ref_node, &locked_ref->ref_tree);
1925                 RB_CLEAR_NODE(&ref->ref_node);
1926                 if (!list_empty(&ref->add_list))
1927                         list_del(&ref->add_list);
1928                 /*
1929                  * When we play the delayed ref, also correct the ref_mod on
1930                  * head
1931                  */
1932                 switch (ref->action) {
1933                 case BTRFS_ADD_DELAYED_REF:
1934                 case BTRFS_ADD_DELAYED_EXTENT:
1935                         locked_ref->ref_mod -= ref->ref_mod;
1936                         break;
1937                 case BTRFS_DROP_DELAYED_REF:
1938                         locked_ref->ref_mod += ref->ref_mod;
1939                         break;
1940                 default:
1941                         WARN_ON(1);
1942                 }
1943                 atomic_dec(&delayed_refs->num_entries);
1944
1945                 /*
1946                  * Record the must_insert_reserved flag before we drop the
1947                  * spin lock.
1948                  */
1949                 must_insert_reserved = locked_ref->must_insert_reserved;
1950                 locked_ref->must_insert_reserved = 0;
1951
1952                 extent_op = locked_ref->extent_op;
1953                 locked_ref->extent_op = NULL;
1954                 spin_unlock(&locked_ref->lock);
1955
1956                 ret = run_one_delayed_ref(trans, ref, extent_op,
1957                                           must_insert_reserved);
1958
1959                 btrfs_free_delayed_extent_op(extent_op);
1960                 if (ret) {
1961                         unselect_delayed_ref_head(delayed_refs, locked_ref);
1962                         btrfs_put_delayed_ref(ref);
1963                         btrfs_debug(fs_info, "run_one_delayed_ref returned %d",
1964                                     ret);
1965                         return ret;
1966                 }
1967
1968                 btrfs_put_delayed_ref(ref);
1969                 cond_resched();
1970
1971                 spin_lock(&locked_ref->lock);
1972                 btrfs_merge_delayed_refs(trans, delayed_refs, locked_ref);
1973         }
1974
1975         return 0;
1976 }
1977
1978 /*
1979  * Returns 0 on success or if called with an already aborted transaction.
1980  * Returns -ENOMEM or -EIO on failure and will abort the transaction.
1981  */
1982 static noinline int __btrfs_run_delayed_refs(struct btrfs_trans_handle *trans,
1983                                              unsigned long nr)
1984 {
1985         struct btrfs_fs_info *fs_info = trans->fs_info;
1986         struct btrfs_delayed_ref_root *delayed_refs;
1987         struct btrfs_delayed_ref_head *locked_ref = NULL;
1988         ktime_t start = ktime_get();
1989         int ret;
1990         unsigned long count = 0;
1991         unsigned long actual_count = 0;
1992
1993         delayed_refs = &trans->transaction->delayed_refs;
1994         do {
1995                 if (!locked_ref) {
1996                         locked_ref = btrfs_obtain_ref_head(trans);
1997                         if (IS_ERR_OR_NULL(locked_ref)) {
1998                                 if (PTR_ERR(locked_ref) == -EAGAIN) {
1999                                         continue;
2000                                 } else {
2001                                         break;
2002                                 }
2003                         }
2004                         count++;
2005                 }
2006                 /*
2007                  * We need to try and merge add/drops of the same ref since we
2008                  * can run into issues with relocate dropping the implicit ref
2009                  * and then it being added back again before the drop can
2010                  * finish.  If we merged anything we need to re-loop so we can
2011                  * get a good ref.
2012                  * Or we can get node references of the same type that weren't
2013                  * merged when created due to bumps in the tree mod seq, and
2014                  * we need to merge them to prevent adding an inline extent
2015                  * backref before dropping it (triggering a BUG_ON at
2016                  * insert_inline_extent_backref()).
2017                  */
2018                 spin_lock(&locked_ref->lock);
2019                 btrfs_merge_delayed_refs(trans, delayed_refs, locked_ref);
2020
2021                 ret = btrfs_run_delayed_refs_for_head(trans, locked_ref,
2022                                                       &actual_count);
2023                 if (ret < 0 && ret != -EAGAIN) {
2024                         /*
2025                          * Error, btrfs_run_delayed_refs_for_head already
2026                          * unlocked everything so just bail out
2027                          */
2028                         return ret;
2029                 } else if (!ret) {
2030                         /*
2031                          * Success, perform the usual cleanup of a processed
2032                          * head
2033                          */
2034                         ret = cleanup_ref_head(trans, locked_ref);
2035                         if (ret > 0 ) {
2036                                 /* We dropped our lock, we need to loop. */
2037                                 ret = 0;
2038                                 continue;
2039                         } else if (ret) {
2040                                 return ret;
2041                         }
2042                 }
2043
2044                 /*
2045                  * Either success case or btrfs_run_delayed_refs_for_head
2046                  * returned -EAGAIN, meaning we need to select another head
2047                  */
2048
2049                 locked_ref = NULL;
2050                 cond_resched();
2051         } while ((nr != -1 && count < nr) || locked_ref);
2052
2053         /*
2054          * We don't want to include ref heads since we can have empty ref heads
2055          * and those will drastically skew our runtime down since we just do
2056          * accounting, no actual extent tree updates.
2057          */
2058         if (actual_count > 0) {
2059                 u64 runtime = ktime_to_ns(ktime_sub(ktime_get(), start));
2060                 u64 avg;
2061
2062                 /*
2063                  * We weigh the current average higher than our current runtime
2064                  * to avoid large swings in the average.
2065                  */
2066                 spin_lock(&delayed_refs->lock);
2067                 avg = fs_info->avg_delayed_ref_runtime * 3 + runtime;
2068                 fs_info->avg_delayed_ref_runtime = avg >> 2;    /* div by 4 */
2069                 spin_unlock(&delayed_refs->lock);
2070         }
2071         return 0;
2072 }
2073
2074 #ifdef SCRAMBLE_DELAYED_REFS
2075 /*
2076  * Normally delayed refs get processed in ascending bytenr order. This
2077  * correlates in most cases to the order added. To expose dependencies on this
2078  * order, we start to process the tree in the middle instead of the beginning
2079  */
2080 static u64 find_middle(struct rb_root *root)
2081 {
2082         struct rb_node *n = root->rb_node;
2083         struct btrfs_delayed_ref_node *entry;
2084         int alt = 1;
2085         u64 middle;
2086         u64 first = 0, last = 0;
2087
2088         n = rb_first(root);
2089         if (n) {
2090                 entry = rb_entry(n, struct btrfs_delayed_ref_node, rb_node);
2091                 first = entry->bytenr;
2092         }
2093         n = rb_last(root);
2094         if (n) {
2095                 entry = rb_entry(n, struct btrfs_delayed_ref_node, rb_node);
2096                 last = entry->bytenr;
2097         }
2098         n = root->rb_node;
2099
2100         while (n) {
2101                 entry = rb_entry(n, struct btrfs_delayed_ref_node, rb_node);
2102                 WARN_ON(!entry->in_tree);
2103
2104                 middle = entry->bytenr;
2105
2106                 if (alt)
2107                         n = n->rb_left;
2108                 else
2109                         n = n->rb_right;
2110
2111                 alt = 1 - alt;
2112         }
2113         return middle;
2114 }
2115 #endif
2116
2117 static inline u64 heads_to_leaves(struct btrfs_fs_info *fs_info, u64 heads)
2118 {
2119         u64 num_bytes;
2120
2121         num_bytes = heads * (sizeof(struct btrfs_extent_item) +
2122                              sizeof(struct btrfs_extent_inline_ref));
2123         if (!btrfs_fs_incompat(fs_info, SKINNY_METADATA))
2124                 num_bytes += heads * sizeof(struct btrfs_tree_block_info);
2125
2126         /*
2127          * We don't ever fill up leaves all the way so multiply by 2 just to be
2128          * closer to what we're really going to want to use.
2129          */
2130         return div_u64(num_bytes, BTRFS_LEAF_DATA_SIZE(fs_info));
2131 }
2132
2133 /*
2134  * Takes the number of bytes to be csumm'ed and figures out how many leaves it
2135  * would require to store the csums for that many bytes.
2136  */
2137 u64 btrfs_csum_bytes_to_leaves(struct btrfs_fs_info *fs_info, u64 csum_bytes)
2138 {
2139         u64 csum_size;
2140         u64 num_csums_per_leaf;
2141         u64 num_csums;
2142
2143         csum_size = BTRFS_MAX_ITEM_SIZE(fs_info);
2144         num_csums_per_leaf = div64_u64(csum_size,
2145                         (u64)btrfs_super_csum_size(fs_info->super_copy));
2146         num_csums = div64_u64(csum_bytes, fs_info->sectorsize);
2147         num_csums += num_csums_per_leaf - 1;
2148         num_csums = div64_u64(num_csums, num_csums_per_leaf);
2149         return num_csums;
2150 }
2151
2152 /*
2153  * this starts processing the delayed reference count updates and
2154  * extent insertions we have queued up so far.  count can be
2155  * 0, which means to process everything in the tree at the start
2156  * of the run (but not newly added entries), or it can be some target
2157  * number you'd like to process.
2158  *
2159  * Returns 0 on success or if called with an aborted transaction
2160  * Returns <0 on error and aborts the transaction
2161  */
2162 int btrfs_run_delayed_refs(struct btrfs_trans_handle *trans,
2163                            unsigned long count)
2164 {
2165         struct btrfs_fs_info *fs_info = trans->fs_info;
2166         struct rb_node *node;
2167         struct btrfs_delayed_ref_root *delayed_refs;
2168         struct btrfs_delayed_ref_head *head;
2169         int ret;
2170         int run_all = count == (unsigned long)-1;
2171
2172         /* We'll clean this up in btrfs_cleanup_transaction */
2173         if (TRANS_ABORTED(trans))
2174                 return 0;
2175
2176         if (test_bit(BTRFS_FS_CREATING_FREE_SPACE_TREE, &fs_info->flags))
2177                 return 0;
2178
2179         delayed_refs = &trans->transaction->delayed_refs;
2180         if (count == 0)
2181                 count = atomic_read(&delayed_refs->num_entries) * 2;
2182
2183 again:
2184 #ifdef SCRAMBLE_DELAYED_REFS
2185         delayed_refs->run_delayed_start = find_middle(&delayed_refs->root);
2186 #endif
2187         ret = __btrfs_run_delayed_refs(trans, count);
2188         if (ret < 0) {
2189                 btrfs_abort_transaction(trans, ret);
2190                 return ret;
2191         }
2192
2193         if (run_all) {
2194                 btrfs_create_pending_block_groups(trans);
2195
2196                 spin_lock(&delayed_refs->lock);
2197                 node = rb_first_cached(&delayed_refs->href_root);
2198                 if (!node) {
2199                         spin_unlock(&delayed_refs->lock);
2200                         goto out;
2201                 }
2202                 head = rb_entry(node, struct btrfs_delayed_ref_head,
2203                                 href_node);
2204                 refcount_inc(&head->refs);
2205                 spin_unlock(&delayed_refs->lock);
2206
2207                 /* Mutex was contended, block until it's released and retry. */
2208                 mutex_lock(&head->mutex);
2209                 mutex_unlock(&head->mutex);
2210
2211                 btrfs_put_delayed_ref_head(head);
2212                 cond_resched();
2213                 goto again;
2214         }
2215 out:
2216         return 0;
2217 }
2218
2219 int btrfs_set_disk_extent_flags(struct btrfs_trans_handle *trans,
2220                                 struct extent_buffer *eb, u64 flags,
2221                                 int level, int is_data)
2222 {
2223         struct btrfs_delayed_extent_op *extent_op;
2224         int ret;
2225
2226         extent_op = btrfs_alloc_delayed_extent_op();
2227         if (!extent_op)
2228                 return -ENOMEM;
2229
2230         extent_op->flags_to_set = flags;
2231         extent_op->update_flags = true;
2232         extent_op->update_key = false;
2233         extent_op->is_data = is_data ? true : false;
2234         extent_op->level = level;
2235
2236         ret = btrfs_add_delayed_extent_op(trans, eb->start, eb->len, extent_op);
2237         if (ret)
2238                 btrfs_free_delayed_extent_op(extent_op);
2239         return ret;
2240 }
2241
2242 static noinline int check_delayed_ref(struct btrfs_root *root,
2243                                       struct btrfs_path *path,
2244                                       u64 objectid, u64 offset, u64 bytenr)
2245 {
2246         struct btrfs_delayed_ref_head *head;
2247         struct btrfs_delayed_ref_node *ref;
2248         struct btrfs_delayed_data_ref *data_ref;
2249         struct btrfs_delayed_ref_root *delayed_refs;
2250         struct btrfs_transaction *cur_trans;
2251         struct rb_node *node;
2252         int ret = 0;
2253
2254         spin_lock(&root->fs_info->trans_lock);
2255         cur_trans = root->fs_info->running_transaction;
2256         if (cur_trans)
2257                 refcount_inc(&cur_trans->use_count);
2258         spin_unlock(&root->fs_info->trans_lock);
2259         if (!cur_trans)
2260                 return 0;
2261
2262         delayed_refs = &cur_trans->delayed_refs;
2263         spin_lock(&delayed_refs->lock);
2264         head = btrfs_find_delayed_ref_head(delayed_refs, bytenr);
2265         if (!head) {
2266                 spin_unlock(&delayed_refs->lock);
2267                 btrfs_put_transaction(cur_trans);
2268                 return 0;
2269         }
2270
2271         if (!mutex_trylock(&head->mutex)) {
2272                 refcount_inc(&head->refs);
2273                 spin_unlock(&delayed_refs->lock);
2274
2275                 btrfs_release_path(path);
2276
2277                 /*
2278                  * Mutex was contended, block until it's released and let
2279                  * caller try again
2280                  */
2281                 mutex_lock(&head->mutex);
2282                 mutex_unlock(&head->mutex);
2283                 btrfs_put_delayed_ref_head(head);
2284                 btrfs_put_transaction(cur_trans);
2285                 return -EAGAIN;
2286         }
2287         spin_unlock(&delayed_refs->lock);
2288
2289         spin_lock(&head->lock);
2290         /*
2291          * XXX: We should replace this with a proper search function in the
2292          * future.
2293          */
2294         for (node = rb_first_cached(&head->ref_tree); node;
2295              node = rb_next(node)) {
2296                 ref = rb_entry(node, struct btrfs_delayed_ref_node, ref_node);
2297                 /* If it's a shared ref we know a cross reference exists */
2298                 if (ref->type != BTRFS_EXTENT_DATA_REF_KEY) {
2299                         ret = 1;
2300                         break;
2301                 }
2302
2303                 data_ref = btrfs_delayed_node_to_data_ref(ref);
2304
2305                 /*
2306                  * If our ref doesn't match the one we're currently looking at
2307                  * then we have a cross reference.
2308                  */
2309                 if (data_ref->root != root->root_key.objectid ||
2310                     data_ref->objectid != objectid ||
2311                     data_ref->offset != offset) {
2312                         ret = 1;
2313                         break;
2314                 }
2315         }
2316         spin_unlock(&head->lock);
2317         mutex_unlock(&head->mutex);
2318         btrfs_put_transaction(cur_trans);
2319         return ret;
2320 }
2321
2322 static noinline int check_committed_ref(struct btrfs_root *root,
2323                                         struct btrfs_path *path,
2324                                         u64 objectid, u64 offset, u64 bytenr)
2325 {
2326         struct btrfs_fs_info *fs_info = root->fs_info;
2327         struct btrfs_root *extent_root = fs_info->extent_root;
2328         struct extent_buffer *leaf;
2329         struct btrfs_extent_data_ref *ref;
2330         struct btrfs_extent_inline_ref *iref;
2331         struct btrfs_extent_item *ei;
2332         struct btrfs_key key;
2333         u32 item_size;
2334         int type;
2335         int ret;
2336
2337         key.objectid = bytenr;
2338         key.offset = (u64)-1;
2339         key.type = BTRFS_EXTENT_ITEM_KEY;
2340
2341         ret = btrfs_search_slot(NULL, extent_root, &key, path, 0, 0);
2342         if (ret < 0)
2343                 goto out;
2344         BUG_ON(ret == 0); /* Corruption */
2345
2346         ret = -ENOENT;
2347         if (path->slots[0] == 0)
2348                 goto out;
2349
2350         path->slots[0]--;
2351         leaf = path->nodes[0];
2352         btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
2353
2354         if (key.objectid != bytenr || key.type != BTRFS_EXTENT_ITEM_KEY)
2355                 goto out;
2356
2357         ret = 1;
2358         item_size = btrfs_item_size_nr(leaf, path->slots[0]);
2359         ei = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
2360
2361         /* If extent item has more than 1 inline ref then it's shared */
2362         if (item_size != sizeof(*ei) +
2363             btrfs_extent_inline_ref_size(BTRFS_EXTENT_DATA_REF_KEY))
2364                 goto out;
2365
2366         /* If extent created before last snapshot => it's definitely shared */
2367         if (btrfs_extent_generation(leaf, ei) <=
2368             btrfs_root_last_snapshot(&root->root_item))
2369                 goto out;
2370
2371         iref = (struct btrfs_extent_inline_ref *)(ei + 1);
2372
2373         /* If this extent has SHARED_DATA_REF then it's shared */
2374         type = btrfs_get_extent_inline_ref_type(leaf, iref, BTRFS_REF_TYPE_DATA);
2375         if (type != BTRFS_EXTENT_DATA_REF_KEY)
2376                 goto out;
2377
2378         ref = (struct btrfs_extent_data_ref *)(&iref->offset);
2379         if (btrfs_extent_refs(leaf, ei) !=
2380             btrfs_extent_data_ref_count(leaf, ref) ||
2381             btrfs_extent_data_ref_root(leaf, ref) !=
2382             root->root_key.objectid ||
2383             btrfs_extent_data_ref_objectid(leaf, ref) != objectid ||
2384             btrfs_extent_data_ref_offset(leaf, ref) != offset)
2385                 goto out;
2386
2387         ret = 0;
2388 out:
2389         return ret;
2390 }
2391
2392 int btrfs_cross_ref_exist(struct btrfs_root *root, u64 objectid, u64 offset,
2393                           u64 bytenr)
2394 {
2395         struct btrfs_path *path;
2396         int ret;
2397
2398         path = btrfs_alloc_path();
2399         if (!path)
2400                 return -ENOMEM;
2401
2402         do {
2403                 ret = check_committed_ref(root, path, objectid,
2404                                           offset, bytenr);
2405                 if (ret && ret != -ENOENT)
2406                         goto out;
2407
2408                 ret = check_delayed_ref(root, path, objectid, offset, bytenr);
2409         } while (ret == -EAGAIN);
2410
2411 out:
2412         btrfs_free_path(path);
2413         if (root->root_key.objectid == BTRFS_DATA_RELOC_TREE_OBJECTID)
2414                 WARN_ON(ret > 0);
2415         return ret;
2416 }
2417
2418 static int __btrfs_mod_ref(struct btrfs_trans_handle *trans,
2419                            struct btrfs_root *root,
2420                            struct extent_buffer *buf,
2421                            int full_backref, int inc)
2422 {
2423         struct btrfs_fs_info *fs_info = root->fs_info;
2424         u64 bytenr;
2425         u64 num_bytes;
2426         u64 parent;
2427         u64 ref_root;
2428         u32 nritems;
2429         struct btrfs_key key;
2430         struct btrfs_file_extent_item *fi;
2431         struct btrfs_ref generic_ref = { 0 };
2432         bool for_reloc = btrfs_header_flag(buf, BTRFS_HEADER_FLAG_RELOC);
2433         int i;
2434         int action;
2435         int level;
2436         int ret = 0;
2437
2438         if (btrfs_is_testing(fs_info))
2439                 return 0;
2440
2441         ref_root = btrfs_header_owner(buf);
2442         nritems = btrfs_header_nritems(buf);
2443         level = btrfs_header_level(buf);
2444
2445         if (!test_bit(BTRFS_ROOT_REF_COWS, &root->state) && level == 0)
2446                 return 0;
2447
2448         if (full_backref)
2449                 parent = buf->start;
2450         else
2451                 parent = 0;
2452         if (inc)
2453                 action = BTRFS_ADD_DELAYED_REF;
2454         else
2455                 action = BTRFS_DROP_DELAYED_REF;
2456
2457         for (i = 0; i < nritems; i++) {
2458                 if (level == 0) {
2459                         btrfs_item_key_to_cpu(buf, &key, i);
2460                         if (key.type != BTRFS_EXTENT_DATA_KEY)
2461                                 continue;
2462                         fi = btrfs_item_ptr(buf, i,
2463                                             struct btrfs_file_extent_item);
2464                         if (btrfs_file_extent_type(buf, fi) ==
2465                             BTRFS_FILE_EXTENT_INLINE)
2466                                 continue;
2467                         bytenr = btrfs_file_extent_disk_bytenr(buf, fi);
2468                         if (bytenr == 0)
2469                                 continue;
2470
2471                         num_bytes = btrfs_file_extent_disk_num_bytes(buf, fi);
2472                         key.offset -= btrfs_file_extent_offset(buf, fi);
2473                         btrfs_init_generic_ref(&generic_ref, action, bytenr,
2474                                                num_bytes, parent);
2475                         generic_ref.real_root = root->root_key.objectid;
2476                         btrfs_init_data_ref(&generic_ref, ref_root, key.objectid,
2477                                             key.offset);
2478                         generic_ref.skip_qgroup = for_reloc;
2479                         if (inc)
2480                                 ret = btrfs_inc_extent_ref(trans, &generic_ref);
2481                         else
2482                                 ret = btrfs_free_extent(trans, &generic_ref);
2483                         if (ret)
2484                                 goto fail;
2485                 } else {
2486                         bytenr = btrfs_node_blockptr(buf, i);
2487                         num_bytes = fs_info->nodesize;
2488                         btrfs_init_generic_ref(&generic_ref, action, bytenr,
2489                                                num_bytes, parent);
2490                         generic_ref.real_root = root->root_key.objectid;
2491                         btrfs_init_tree_ref(&generic_ref, level - 1, ref_root);
2492                         generic_ref.skip_qgroup = for_reloc;
2493                         if (inc)
2494                                 ret = btrfs_inc_extent_ref(trans, &generic_ref);
2495                         else
2496                                 ret = btrfs_free_extent(trans, &generic_ref);
2497                         if (ret)
2498                                 goto fail;
2499                 }
2500         }
2501         return 0;
2502 fail:
2503         return ret;
2504 }
2505
2506 int btrfs_inc_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root,
2507                   struct extent_buffer *buf, int full_backref)
2508 {
2509         return __btrfs_mod_ref(trans, root, buf, full_backref, 1);
2510 }
2511
2512 int btrfs_dec_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root,
2513                   struct extent_buffer *buf, int full_backref)
2514 {
2515         return __btrfs_mod_ref(trans, root, buf, full_backref, 0);
2516 }
2517
2518 int btrfs_extent_readonly(struct btrfs_fs_info *fs_info, u64 bytenr)
2519 {
2520         struct btrfs_block_group *block_group;
2521         int readonly = 0;
2522
2523         block_group = btrfs_lookup_block_group(fs_info, bytenr);
2524         if (!block_group || block_group->ro)
2525                 readonly = 1;
2526         if (block_group)
2527                 btrfs_put_block_group(block_group);
2528         return readonly;
2529 }
2530
2531 static u64 get_alloc_profile_by_root(struct btrfs_root *root, int data)
2532 {
2533         struct btrfs_fs_info *fs_info = root->fs_info;
2534         u64 flags;
2535         u64 ret;
2536
2537         if (data)
2538                 flags = BTRFS_BLOCK_GROUP_DATA;
2539         else if (root == fs_info->chunk_root)
2540                 flags = BTRFS_BLOCK_GROUP_SYSTEM;
2541         else
2542                 flags = BTRFS_BLOCK_GROUP_METADATA;
2543
2544         ret = btrfs_get_alloc_profile(fs_info, flags);
2545         return ret;
2546 }
2547
2548 static u64 first_logical_byte(struct btrfs_fs_info *fs_info, u64 search_start)
2549 {
2550         struct btrfs_block_group *cache;
2551         u64 bytenr;
2552
2553         spin_lock(&fs_info->block_group_cache_lock);
2554         bytenr = fs_info->first_logical_byte;
2555         spin_unlock(&fs_info->block_group_cache_lock);
2556
2557         if (bytenr < (u64)-1)
2558                 return bytenr;
2559
2560         cache = btrfs_lookup_first_block_group(fs_info, search_start);
2561         if (!cache)
2562                 return 0;
2563
2564         bytenr = cache->start;
2565         btrfs_put_block_group(cache);
2566
2567         return bytenr;
2568 }
2569
2570 static int pin_down_extent(struct btrfs_trans_handle *trans,
2571                            struct btrfs_block_group *cache,
2572                            u64 bytenr, u64 num_bytes, int reserved)
2573 {
2574         struct btrfs_fs_info *fs_info = cache->fs_info;
2575
2576         spin_lock(&cache->space_info->lock);
2577         spin_lock(&cache->lock);
2578         cache->pinned += num_bytes;
2579         btrfs_space_info_update_bytes_pinned(fs_info, cache->space_info,
2580                                              num_bytes);
2581         if (reserved) {
2582                 cache->reserved -= num_bytes;
2583                 cache->space_info->bytes_reserved -= num_bytes;
2584         }
2585         spin_unlock(&cache->lock);
2586         spin_unlock(&cache->space_info->lock);
2587
2588         percpu_counter_add_batch(&cache->space_info->total_bytes_pinned,
2589                     num_bytes, BTRFS_TOTAL_BYTES_PINNED_BATCH);
2590         set_extent_dirty(&trans->transaction->pinned_extents, bytenr,
2591                          bytenr + num_bytes - 1, GFP_NOFS | __GFP_NOFAIL);
2592         return 0;
2593 }
2594
2595 int btrfs_pin_extent(struct btrfs_trans_handle *trans,
2596                      u64 bytenr, u64 num_bytes, int reserved)
2597 {
2598         struct btrfs_block_group *cache;
2599
2600         cache = btrfs_lookup_block_group(trans->fs_info, bytenr);
2601         BUG_ON(!cache); /* Logic error */
2602
2603         pin_down_extent(trans, cache, bytenr, num_bytes, reserved);
2604
2605         btrfs_put_block_group(cache);
2606         return 0;
2607 }
2608
2609 /*
2610  * this function must be called within transaction
2611  */
2612 int btrfs_pin_extent_for_log_replay(struct btrfs_trans_handle *trans,
2613                                     u64 bytenr, u64 num_bytes)
2614 {
2615         struct btrfs_block_group *cache;
2616         int ret;
2617
2618         btrfs_add_excluded_extent(trans->fs_info, bytenr, num_bytes);
2619
2620         cache = btrfs_lookup_block_group(trans->fs_info, bytenr);
2621         if (!cache)
2622                 return -EINVAL;
2623
2624         /*
2625          * pull in the free space cache (if any) so that our pin
2626          * removes the free space from the cache.  We have load_only set
2627          * to one because the slow code to read in the free extents does check
2628          * the pinned extents.
2629          */
2630         btrfs_cache_block_group(cache, 1);
2631
2632         pin_down_extent(trans, cache, bytenr, num_bytes, 0);
2633
2634         /* remove us from the free space cache (if we're there at all) */
2635         ret = btrfs_remove_free_space(cache, bytenr, num_bytes);
2636         btrfs_put_block_group(cache);
2637         return ret;
2638 }
2639
2640 static int __exclude_logged_extent(struct btrfs_fs_info *fs_info,
2641                                    u64 start, u64 num_bytes)
2642 {
2643         int ret;
2644         struct btrfs_block_group *block_group;
2645         struct btrfs_caching_control *caching_ctl;
2646
2647         block_group = btrfs_lookup_block_group(fs_info, start);
2648         if (!block_group)
2649                 return -EINVAL;
2650
2651         btrfs_cache_block_group(block_group, 0);
2652         caching_ctl = btrfs_get_caching_control(block_group);
2653
2654         if (!caching_ctl) {
2655                 /* Logic error */
2656                 BUG_ON(!btrfs_block_group_done(block_group));
2657                 ret = btrfs_remove_free_space(block_group, start, num_bytes);
2658         } else {
2659                 mutex_lock(&caching_ctl->mutex);
2660
2661                 if (start >= caching_ctl->progress) {
2662                         ret = btrfs_add_excluded_extent(fs_info, start,
2663                                                         num_bytes);
2664                 } else if (start + num_bytes <= caching_ctl->progress) {
2665                         ret = btrfs_remove_free_space(block_group,
2666                                                       start, num_bytes);
2667                 } else {
2668                         num_bytes = caching_ctl->progress - start;
2669                         ret = btrfs_remove_free_space(block_group,
2670                                                       start, num_bytes);
2671                         if (ret)
2672                                 goto out_lock;
2673
2674                         num_bytes = (start + num_bytes) -
2675                                 caching_ctl->progress;
2676                         start = caching_ctl->progress;
2677                         ret = btrfs_add_excluded_extent(fs_info, start,
2678                                                         num_bytes);
2679                 }
2680 out_lock:
2681                 mutex_unlock(&caching_ctl->mutex);
2682                 btrfs_put_caching_control(caching_ctl);
2683         }
2684         btrfs_put_block_group(block_group);
2685         return ret;
2686 }
2687
2688 int btrfs_exclude_logged_extents(struct extent_buffer *eb)
2689 {
2690         struct btrfs_fs_info *fs_info = eb->fs_info;
2691         struct btrfs_file_extent_item *item;
2692         struct btrfs_key key;
2693         int found_type;
2694         int i;
2695         int ret = 0;
2696
2697         if (!btrfs_fs_incompat(fs_info, MIXED_GROUPS))
2698                 return 0;
2699
2700         for (i = 0; i < btrfs_header_nritems(eb); i++) {
2701                 btrfs_item_key_to_cpu(eb, &key, i);
2702                 if (key.type != BTRFS_EXTENT_DATA_KEY)
2703                         continue;
2704                 item = btrfs_item_ptr(eb, i, struct btrfs_file_extent_item);
2705                 found_type = btrfs_file_extent_type(eb, item);
2706                 if (found_type == BTRFS_FILE_EXTENT_INLINE)
2707                         continue;
2708                 if (btrfs_file_extent_disk_bytenr(eb, item) == 0)
2709                         continue;
2710                 key.objectid = btrfs_file_extent_disk_bytenr(eb, item);
2711                 key.offset = btrfs_file_extent_disk_num_bytes(eb, item);
2712                 ret = __exclude_logged_extent(fs_info, key.objectid, key.offset);
2713                 if (ret)
2714                         break;
2715         }
2716
2717         return ret;
2718 }
2719
2720 static void
2721 btrfs_inc_block_group_reservations(struct btrfs_block_group *bg)
2722 {
2723         atomic_inc(&bg->reservations);
2724 }
2725
2726 void btrfs_prepare_extent_commit(struct btrfs_fs_info *fs_info)
2727 {
2728         struct btrfs_caching_control *next;
2729         struct btrfs_caching_control *caching_ctl;
2730         struct btrfs_block_group *cache;
2731
2732         down_write(&fs_info->commit_root_sem);
2733
2734         list_for_each_entry_safe(caching_ctl, next,
2735                                  &fs_info->caching_block_groups, list) {
2736                 cache = caching_ctl->block_group;
2737                 if (btrfs_block_group_done(cache)) {
2738                         cache->last_byte_to_unpin = (u64)-1;
2739                         list_del_init(&caching_ctl->list);
2740                         btrfs_put_caching_control(caching_ctl);
2741                 } else {
2742                         cache->last_byte_to_unpin = caching_ctl->progress;
2743                 }
2744         }
2745
2746         up_write(&fs_info->commit_root_sem);
2747
2748         btrfs_update_global_block_rsv(fs_info);
2749 }
2750
2751 /*
2752  * Returns the free cluster for the given space info and sets empty_cluster to
2753  * what it should be based on the mount options.
2754  */
2755 static struct btrfs_free_cluster *
2756 fetch_cluster_info(struct btrfs_fs_info *fs_info,
2757                    struct btrfs_space_info *space_info, u64 *empty_cluster)
2758 {
2759         struct btrfs_free_cluster *ret = NULL;
2760
2761         *empty_cluster = 0;
2762         if (btrfs_mixed_space_info(space_info))
2763                 return ret;
2764
2765         if (space_info->flags & BTRFS_BLOCK_GROUP_METADATA) {
2766                 ret = &fs_info->meta_alloc_cluster;
2767                 if (btrfs_test_opt(fs_info, SSD))
2768                         *empty_cluster = SZ_2M;
2769                 else
2770                         *empty_cluster = SZ_64K;
2771         } else if ((space_info->flags & BTRFS_BLOCK_GROUP_DATA) &&
2772                    btrfs_test_opt(fs_info, SSD_SPREAD)) {
2773                 *empty_cluster = SZ_2M;
2774                 ret = &fs_info->data_alloc_cluster;
2775         }
2776
2777         return ret;
2778 }
2779
2780 static int unpin_extent_range(struct btrfs_fs_info *fs_info,
2781                               u64 start, u64 end,
2782                               const bool return_free_space)
2783 {
2784         struct btrfs_block_group *cache = NULL;
2785         struct btrfs_space_info *space_info;
2786         struct btrfs_block_rsv *global_rsv = &fs_info->global_block_rsv;
2787         struct btrfs_free_cluster *cluster = NULL;
2788         u64 len;
2789         u64 total_unpinned = 0;
2790         u64 empty_cluster = 0;
2791         bool readonly;
2792
2793         while (start <= end) {
2794                 readonly = false;
2795                 if (!cache ||
2796                     start >= cache->start + cache->length) {
2797                         if (cache)
2798                                 btrfs_put_block_group(cache);
2799                         total_unpinned = 0;
2800                         cache = btrfs_lookup_block_group(fs_info, start);
2801                         BUG_ON(!cache); /* Logic error */
2802
2803                         cluster = fetch_cluster_info(fs_info,
2804                                                      cache->space_info,
2805                                                      &empty_cluster);
2806                         empty_cluster <<= 1;
2807                 }
2808
2809                 len = cache->start + cache->length - start;
2810                 len = min(len, end + 1 - start);
2811
2812                 if (start < cache->last_byte_to_unpin) {
2813                         len = min(len, cache->last_byte_to_unpin - start);
2814                         if (return_free_space)
2815                                 btrfs_add_free_space(cache, start, len);
2816                 }
2817
2818                 start += len;
2819                 total_unpinned += len;
2820                 space_info = cache->space_info;
2821
2822                 /*
2823                  * If this space cluster has been marked as fragmented and we've
2824                  * unpinned enough in this block group to potentially allow a
2825                  * cluster to be created inside of it go ahead and clear the
2826                  * fragmented check.
2827                  */
2828                 if (cluster && cluster->fragmented &&
2829                     total_unpinned > empty_cluster) {
2830                         spin_lock(&cluster->lock);
2831                         cluster->fragmented = 0;
2832                         spin_unlock(&cluster->lock);
2833                 }
2834
2835                 spin_lock(&space_info->lock);
2836                 spin_lock(&cache->lock);
2837                 cache->pinned -= len;
2838                 btrfs_space_info_update_bytes_pinned(fs_info, space_info, -len);
2839                 space_info->max_extent_size = 0;
2840                 percpu_counter_add_batch(&space_info->total_bytes_pinned,
2841                             -len, BTRFS_TOTAL_BYTES_PINNED_BATCH);
2842                 if (cache->ro) {
2843                         space_info->bytes_readonly += len;
2844                         readonly = true;
2845                 }
2846                 spin_unlock(&cache->lock);
2847                 if (!readonly && return_free_space &&
2848                     global_rsv->space_info == space_info) {
2849                         u64 to_add = len;
2850
2851                         spin_lock(&global_rsv->lock);
2852                         if (!global_rsv->full) {
2853                                 to_add = min(len, global_rsv->size -
2854                                              global_rsv->reserved);
2855                                 global_rsv->reserved += to_add;
2856                                 btrfs_space_info_update_bytes_may_use(fs_info,
2857                                                 space_info, to_add);
2858                                 if (global_rsv->reserved >= global_rsv->size)
2859                                         global_rsv->full = 1;
2860                                 len -= to_add;
2861                         }
2862                         spin_unlock(&global_rsv->lock);
2863                         /* Add to any tickets we may have */
2864                         if (len)
2865                                 btrfs_try_granting_tickets(fs_info,
2866                                                            space_info);
2867                 }
2868                 spin_unlock(&space_info->lock);
2869         }
2870
2871         if (cache)
2872                 btrfs_put_block_group(cache);
2873         return 0;
2874 }
2875
2876 int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans)
2877 {
2878         struct btrfs_fs_info *fs_info = trans->fs_info;
2879         struct btrfs_block_group *block_group, *tmp;
2880         struct list_head *deleted_bgs;
2881         struct extent_io_tree *unpin;
2882         u64 start;
2883         u64 end;
2884         int ret;
2885
2886         unpin = &trans->transaction->pinned_extents;
2887
2888         while (!TRANS_ABORTED(trans)) {
2889                 struct extent_state *cached_state = NULL;
2890
2891                 mutex_lock(&fs_info->unused_bg_unpin_mutex);
2892                 ret = find_first_extent_bit(unpin, 0, &start, &end,
2893                                             EXTENT_DIRTY, &cached_state);
2894                 if (ret) {
2895                         mutex_unlock(&fs_info->unused_bg_unpin_mutex);
2896                         break;
2897                 }
2898                 if (test_bit(BTRFS_FS_LOG_RECOVERING, &fs_info->flags))
2899                         clear_extent_bits(&fs_info->excluded_extents, start,
2900                                           end, EXTENT_UPTODATE);
2901
2902                 if (btrfs_test_opt(fs_info, DISCARD_SYNC))
2903                         ret = btrfs_discard_extent(fs_info, start,
2904                                                    end + 1 - start, NULL);
2905
2906                 clear_extent_dirty(unpin, start, end, &cached_state);
2907                 unpin_extent_range(fs_info, start, end, true);
2908                 mutex_unlock(&fs_info->unused_bg_unpin_mutex);
2909                 free_extent_state(cached_state);
2910                 cond_resched();
2911         }
2912
2913         if (btrfs_test_opt(fs_info, DISCARD_ASYNC)) {
2914                 btrfs_discard_calc_delay(&fs_info->discard_ctl);
2915                 btrfs_discard_schedule_work(&fs_info->discard_ctl, true);
2916         }
2917
2918         /*
2919          * Transaction is finished.  We don't need the lock anymore.  We
2920          * do need to clean up the block groups in case of a transaction
2921          * abort.
2922          */
2923         deleted_bgs = &trans->transaction->deleted_bgs;
2924         list_for_each_entry_safe(block_group, tmp, deleted_bgs, bg_list) {
2925                 u64 trimmed = 0;
2926
2927                 ret = -EROFS;
2928                 if (!TRANS_ABORTED(trans))
2929                         ret = btrfs_discard_extent(fs_info,
2930                                                    block_group->start,
2931                                                    block_group->length,
2932                                                    &trimmed);
2933
2934                 list_del_init(&block_group->bg_list);
2935                 btrfs_put_block_group_trimming(block_group);
2936                 btrfs_put_block_group(block_group);
2937
2938                 if (ret) {
2939                         const char *errstr = btrfs_decode_error(ret);
2940                         btrfs_warn(fs_info,
2941                            "discard failed while removing blockgroup: errno=%d %s",
2942                                    ret, errstr);
2943                 }
2944         }
2945
2946         return 0;
2947 }
2948
2949 static int __btrfs_free_extent(struct btrfs_trans_handle *trans,
2950                                struct btrfs_delayed_ref_node *node, u64 parent,
2951                                u64 root_objectid, u64 owner_objectid,
2952                                u64 owner_offset, int refs_to_drop,
2953                                struct btrfs_delayed_extent_op *extent_op)
2954 {
2955         struct btrfs_fs_info *info = trans->fs_info;
2956         struct btrfs_key key;
2957         struct btrfs_path *path;
2958         struct btrfs_root *extent_root = info->extent_root;
2959         struct extent_buffer *leaf;
2960         struct btrfs_extent_item *ei;
2961         struct btrfs_extent_inline_ref *iref;
2962         int ret;
2963         int is_data;
2964         int extent_slot = 0;
2965         int found_extent = 0;
2966         int num_to_del = 1;
2967         u32 item_size;
2968         u64 refs;
2969         u64 bytenr = node->bytenr;
2970         u64 num_bytes = node->num_bytes;
2971         int last_ref = 0;
2972         bool skinny_metadata = btrfs_fs_incompat(info, SKINNY_METADATA);
2973
2974         path = btrfs_alloc_path();
2975         if (!path)
2976                 return -ENOMEM;
2977
2978         path->leave_spinning = 1;
2979
2980         is_data = owner_objectid >= BTRFS_FIRST_FREE_OBJECTID;
2981         BUG_ON(!is_data && refs_to_drop != 1);
2982
2983         if (is_data)
2984                 skinny_metadata = false;
2985
2986         ret = lookup_extent_backref(trans, path, &iref, bytenr, num_bytes,
2987                                     parent, root_objectid, owner_objectid,
2988                                     owner_offset);
2989         if (ret == 0) {
2990                 extent_slot = path->slots[0];
2991                 while (extent_slot >= 0) {
2992                         btrfs_item_key_to_cpu(path->nodes[0], &key,
2993                                               extent_slot);
2994                         if (key.objectid != bytenr)
2995                                 break;
2996                         if (key.type == BTRFS_EXTENT_ITEM_KEY &&
2997                             key.offset == num_bytes) {
2998                                 found_extent = 1;
2999                                 break;
3000                         }
3001                         if (key.type == BTRFS_METADATA_ITEM_KEY &&
3002                             key.offset == owner_objectid) {
3003                                 found_extent = 1;
3004                                 break;
3005                         }
3006                         if (path->slots[0] - extent_slot > 5)
3007                                 break;
3008                         extent_slot--;
3009                 }
3010
3011                 if (!found_extent) {
3012                         BUG_ON(iref);
3013                         ret = remove_extent_backref(trans, path, NULL,
3014                                                     refs_to_drop,
3015                                                     is_data, &last_ref);
3016                         if (ret) {
3017                                 btrfs_abort_transaction(trans, ret);
3018                                 goto out;
3019                         }
3020                         btrfs_release_path(path);
3021                         path->leave_spinning = 1;
3022
3023                         key.objectid = bytenr;
3024                         key.type = BTRFS_EXTENT_ITEM_KEY;
3025                         key.offset = num_bytes;
3026
3027                         if (!is_data && skinny_metadata) {
3028                                 key.type = BTRFS_METADATA_ITEM_KEY;
3029                                 key.offset = owner_objectid;
3030                         }
3031
3032                         ret = btrfs_search_slot(trans, extent_root,
3033                                                 &key, path, -1, 1);
3034                         if (ret > 0 && skinny_metadata && path->slots[0]) {
3035                                 /*
3036                                  * Couldn't find our skinny metadata item,
3037                                  * see if we have ye olde extent item.
3038                                  */
3039                                 path->slots[0]--;
3040                                 btrfs_item_key_to_cpu(path->nodes[0], &key,
3041                                                       path->slots[0]);
3042                                 if (key.objectid == bytenr &&
3043                                     key.type == BTRFS_EXTENT_ITEM_KEY &&
3044                                     key.offset == num_bytes)
3045                                         ret = 0;
3046                         }
3047
3048                         if (ret > 0 && skinny_metadata) {
3049                                 skinny_metadata = false;
3050                                 key.objectid = bytenr;
3051                                 key.type = BTRFS_EXTENT_ITEM_KEY;
3052                                 key.offset = num_bytes;
3053                                 btrfs_release_path(path);
3054                                 ret = btrfs_search_slot(trans, extent_root,
3055                                                         &key, path, -1, 1);
3056                         }
3057
3058                         if (ret) {
3059                                 btrfs_err(info,
3060                                           "umm, got %d back from search, was looking for %llu",
3061                                           ret, bytenr);
3062                                 if (ret > 0)
3063                                         btrfs_print_leaf(path->nodes[0]);
3064                         }
3065                         if (ret < 0) {
3066                                 btrfs_abort_transaction(trans, ret);
3067                                 goto out;
3068                         }
3069                         extent_slot = path->slots[0];
3070                 }
3071         } else if (WARN_ON(ret == -ENOENT)) {
3072                 btrfs_print_leaf(path->nodes[0]);
3073                 btrfs_err(info,
3074                         "unable to find ref byte nr %llu parent %llu root %llu  owner %llu offset %llu",
3075                         bytenr, parent, root_objectid, owner_objectid,
3076                         owner_offset);
3077                 btrfs_abort_transaction(trans, ret);
3078                 goto out;
3079         } else {
3080                 btrfs_abort_transaction(trans, ret);
3081                 goto out;
3082         }
3083
3084         leaf = path->nodes[0];
3085         item_size = btrfs_item_size_nr(leaf, extent_slot);
3086         if (unlikely(item_size < sizeof(*ei))) {
3087                 ret = -EINVAL;
3088                 btrfs_print_v0_err(info);
3089                 btrfs_abort_transaction(trans, ret);
3090                 goto out;
3091         }
3092         ei = btrfs_item_ptr(leaf, extent_slot,
3093                             struct btrfs_extent_item);
3094         if (owner_objectid < BTRFS_FIRST_FREE_OBJECTID &&
3095             key.type == BTRFS_EXTENT_ITEM_KEY) {
3096                 struct btrfs_tree_block_info *bi;
3097                 BUG_ON(item_size < sizeof(*ei) + sizeof(*bi));
3098                 bi = (struct btrfs_tree_block_info *)(ei + 1);
3099                 WARN_ON(owner_objectid != btrfs_tree_block_level(leaf, bi));
3100         }
3101
3102         refs = btrfs_extent_refs(leaf, ei);
3103         if (refs < refs_to_drop) {
3104                 btrfs_err(info,
3105                           "trying to drop %d refs but we only have %Lu for bytenr %Lu",
3106                           refs_to_drop, refs, bytenr);
3107                 ret = -EINVAL;
3108                 btrfs_abort_transaction(trans, ret);
3109                 goto out;
3110         }
3111         refs -= refs_to_drop;
3112
3113         if (refs > 0) {
3114                 if (extent_op)
3115                         __run_delayed_extent_op(extent_op, leaf, ei);
3116                 /*
3117                  * In the case of inline back ref, reference count will
3118                  * be updated by remove_extent_backref
3119                  */
3120                 if (iref) {
3121                         BUG_ON(!found_extent);
3122                 } else {
3123                         btrfs_set_extent_refs(leaf, ei, refs);
3124                         btrfs_mark_buffer_dirty(leaf);
3125                 }
3126                 if (found_extent) {
3127                         ret = remove_extent_backref(trans, path, iref,
3128                                                     refs_to_drop, is_data,
3129                                                     &last_ref);
3130                         if (ret) {
3131                                 btrfs_abort_transaction(trans, ret);
3132                                 goto out;
3133                         }
3134                 }
3135         } else {
3136                 if (found_extent) {
3137                         BUG_ON(is_data && refs_to_drop !=
3138                                extent_data_ref_count(path, iref));
3139                         if (iref) {
3140                                 BUG_ON(path->slots[0] != extent_slot);
3141                         } else {
3142                                 BUG_ON(path->slots[0] != extent_slot + 1);
3143                                 path->slots[0] = extent_slot;
3144                                 num_to_del = 2;
3145                         }
3146                 }
3147
3148                 last_ref = 1;
3149                 ret = btrfs_del_items(trans, extent_root, path, path->slots[0],
3150                                       num_to_del);
3151                 if (ret) {
3152                         btrfs_abort_transaction(trans, ret);
3153                         goto out;
3154                 }
3155                 btrfs_release_path(path);
3156
3157                 if (is_data) {
3158                         ret = btrfs_del_csums(trans, info->csum_root, bytenr,
3159                                               num_bytes);
3160                         if (ret) {
3161                                 btrfs_abort_transaction(trans, ret);
3162                                 goto out;
3163                         }
3164                 }
3165
3166                 ret = add_to_free_space_tree(trans, bytenr, num_bytes);
3167                 if (ret) {
3168                         btrfs_abort_transaction(trans, ret);
3169                         goto out;
3170                 }
3171
3172                 ret = btrfs_update_block_group(trans, bytenr, num_bytes, 0);
3173                 if (ret) {
3174                         btrfs_abort_transaction(trans, ret);
3175                         goto out;
3176                 }
3177         }
3178         btrfs_release_path(path);
3179
3180 out:
3181         btrfs_free_path(path);
3182         return ret;
3183 }
3184
3185 /*
3186  * when we free an block, it is possible (and likely) that we free the last
3187  * delayed ref for that extent as well.  This searches the delayed ref tree for
3188  * a given extent, and if there are no other delayed refs to be processed, it
3189  * removes it from the tree.
3190  */
3191 static noinline int check_ref_cleanup(struct btrfs_trans_handle *trans,
3192                                       u64 bytenr)
3193 {
3194         struct btrfs_delayed_ref_head *head;
3195         struct btrfs_delayed_ref_root *delayed_refs;
3196         int ret = 0;
3197
3198         delayed_refs = &trans->transaction->delayed_refs;
3199         spin_lock(&delayed_refs->lock);
3200         head = btrfs_find_delayed_ref_head(delayed_refs, bytenr);
3201         if (!head)
3202                 goto out_delayed_unlock;
3203
3204         spin_lock(&head->lock);
3205         if (!RB_EMPTY_ROOT(&head->ref_tree.rb_root))
3206                 goto out;
3207
3208         if (cleanup_extent_op(head) != NULL)
3209                 goto out;
3210
3211         /*
3212          * waiting for the lock here would deadlock.  If someone else has it
3213          * locked they are already in the process of dropping it anyway
3214          */
3215         if (!mutex_trylock(&head->mutex))
3216                 goto out;
3217
3218         btrfs_delete_ref_head(delayed_refs, head);
3219         head->processing = 0;
3220
3221         spin_unlock(&head->lock);
3222         spin_unlock(&delayed_refs->lock);
3223
3224         BUG_ON(head->extent_op);
3225         if (head->must_insert_reserved)
3226                 ret = 1;
3227
3228         btrfs_cleanup_ref_head_accounting(trans->fs_info, delayed_refs, head);
3229         mutex_unlock(&head->mutex);
3230         btrfs_put_delayed_ref_head(head);
3231         return ret;
3232 out:
3233         spin_unlock(&head->lock);
3234
3235 out_delayed_unlock:
3236         spin_unlock(&delayed_refs->lock);
3237         return 0;
3238 }
3239
3240 void btrfs_free_tree_block(struct btrfs_trans_handle *trans,
3241                            struct btrfs_root *root,
3242                            struct extent_buffer *buf,
3243                            u64 parent, int last_ref)
3244 {
3245         struct btrfs_fs_info *fs_info = root->fs_info;
3246         struct btrfs_ref generic_ref = { 0 };
3247         int pin = 1;
3248         int ret;
3249
3250         btrfs_init_generic_ref(&generic_ref, BTRFS_DROP_DELAYED_REF,
3251                                buf->start, buf->len, parent);
3252         btrfs_init_tree_ref(&generic_ref, btrfs_header_level(buf),
3253                             root->root_key.objectid);
3254
3255         if (root->root_key.objectid != BTRFS_TREE_LOG_OBJECTID) {
3256                 int old_ref_mod, new_ref_mod;
3257
3258                 btrfs_ref_tree_mod(fs_info, &generic_ref);
3259                 ret = btrfs_add_delayed_tree_ref(trans, &generic_ref, NULL,
3260                                                  &old_ref_mod, &new_ref_mod);
3261                 BUG_ON(ret); /* -ENOMEM */
3262                 pin = old_ref_mod >= 0 && new_ref_mod < 0;
3263         }
3264
3265         if (last_ref && btrfs_header_generation(buf) == trans->transid) {
3266                 struct btrfs_block_group *cache;
3267
3268                 if (root->root_key.objectid != BTRFS_TREE_LOG_OBJECTID) {
3269                         ret = check_ref_cleanup(trans, buf->start);
3270                         if (!ret)
3271                                 goto out;
3272                 }
3273
3274                 pin = 0;
3275                 cache = btrfs_lookup_block_group(fs_info, buf->start);
3276
3277                 if (btrfs_header_flag(buf, BTRFS_HEADER_FLAG_WRITTEN)) {
3278                         pin_down_extent(trans, cache, buf->start, buf->len, 1);
3279                         btrfs_put_block_group(cache);
3280                         goto out;
3281                 }
3282
3283                 WARN_ON(test_bit(EXTENT_BUFFER_DIRTY, &buf->bflags));
3284
3285                 btrfs_add_free_space(cache, buf->start, buf->len);
3286                 btrfs_free_reserved_bytes(cache, buf->len, 0);
3287                 btrfs_put_block_group(cache);
3288                 trace_btrfs_reserved_extent_free(fs_info, buf->start, buf->len);
3289         }
3290 out:
3291         if (pin)
3292                 add_pinned_bytes(fs_info, &generic_ref);
3293
3294         if (last_ref) {
3295                 /*
3296                  * Deleting the buffer, clear the corrupt flag since it doesn't
3297                  * matter anymore.
3298                  */
3299                 clear_bit(EXTENT_BUFFER_CORRUPT, &buf->bflags);
3300         }
3301 }
3302
3303 /* Can return -ENOMEM */
3304 int btrfs_free_extent(struct btrfs_trans_handle *trans, struct btrfs_ref *ref)
3305 {
3306         struct btrfs_fs_info *fs_info = trans->fs_info;
3307         int old_ref_mod, new_ref_mod;
3308         int ret;
3309
3310         if (btrfs_is_testing(fs_info))
3311                 return 0;
3312
3313         /*
3314          * tree log blocks never actually go into the extent allocation
3315          * tree, just update pinning info and exit early.
3316          */
3317         if ((ref->type == BTRFS_REF_METADATA &&
3318              ref->tree_ref.root == BTRFS_TREE_LOG_OBJECTID) ||
3319             (ref->type == BTRFS_REF_DATA &&
3320              ref->data_ref.ref_root == BTRFS_TREE_LOG_OBJECTID)) {
3321                 /* unlocks the pinned mutex */
3322                 btrfs_pin_extent(trans, ref->bytenr, ref->len, 1);
3323                 old_ref_mod = new_ref_mod = 0;
3324                 ret = 0;
3325         } else if (ref->type == BTRFS_REF_METADATA) {
3326                 ret = btrfs_add_delayed_tree_ref(trans, ref, NULL,
3327                                                  &old_ref_mod, &new_ref_mod);
3328         } else {
3329                 ret = btrfs_add_delayed_data_ref(trans, ref, 0,
3330                                                  &old_ref_mod, &new_ref_mod);
3331         }
3332
3333         if (!((ref->type == BTRFS_REF_METADATA &&
3334                ref->tree_ref.root == BTRFS_TREE_LOG_OBJECTID) ||
3335               (ref->type == BTRFS_REF_DATA &&
3336                ref->data_ref.ref_root == BTRFS_TREE_LOG_OBJECTID)))
3337                 btrfs_ref_tree_mod(fs_info, ref);
3338
3339         if (ret == 0 && old_ref_mod >= 0 && new_ref_mod < 0)
3340                 add_pinned_bytes(fs_info, ref);
3341
3342         return ret;
3343 }
3344
3345 enum btrfs_loop_type {
3346         LOOP_CACHING_NOWAIT,
3347         LOOP_CACHING_WAIT,
3348         LOOP_ALLOC_CHUNK,
3349         LOOP_NO_EMPTY_SIZE,
3350 };
3351
3352 static inline void
3353 btrfs_lock_block_group(struct btrfs_block_group *cache,
3354                        int delalloc)
3355 {
3356         if (delalloc)
3357                 down_read(&cache->data_rwsem);
3358 }
3359
3360 static inline void btrfs_grab_block_group(struct btrfs_block_group *cache,
3361                        int delalloc)
3362 {
3363         btrfs_get_block_group(cache);
3364         if (delalloc)
3365                 down_read(&cache->data_rwsem);
3366 }
3367
3368 static struct btrfs_block_group *btrfs_lock_cluster(
3369                    struct btrfs_block_group *block_group,
3370                    struct btrfs_free_cluster *cluster,
3371                    int delalloc)
3372 {
3373         struct btrfs_block_group *used_bg = NULL;
3374
3375         spin_lock(&cluster->refill_lock);
3376         while (1) {
3377                 used_bg = cluster->block_group;
3378                 if (!used_bg)
3379                         return NULL;
3380
3381                 if (used_bg == block_group)
3382                         return used_bg;
3383
3384                 btrfs_get_block_group(used_bg);
3385
3386                 if (!delalloc)
3387                         return used_bg;
3388
3389                 if (down_read_trylock(&used_bg->data_rwsem))
3390                         return used_bg;
3391
3392                 spin_unlock(&cluster->refill_lock);
3393
3394                 /* We should only have one-level nested. */
3395                 down_read_nested(&used_bg->data_rwsem, SINGLE_DEPTH_NESTING);
3396
3397                 spin_lock(&cluster->refill_lock);
3398                 if (used_bg == cluster->block_group)
3399                         return used_bg;
3400
3401                 up_read(&used_bg->data_rwsem);
3402                 btrfs_put_block_group(used_bg);
3403         }
3404 }
3405
3406 static inline void
3407 btrfs_release_block_group(struct btrfs_block_group *cache,
3408                          int delalloc)
3409 {
3410         if (delalloc)
3411                 up_read(&cache->data_rwsem);
3412         btrfs_put_block_group(cache);
3413 }
3414
3415 enum btrfs_extent_allocation_policy {
3416         BTRFS_EXTENT_ALLOC_CLUSTERED,
3417 };
3418
3419 /*
3420  * Structure used internally for find_free_extent() function.  Wraps needed
3421  * parameters.
3422  */
3423 struct find_free_extent_ctl {
3424         /* Basic allocation info */
3425         u64 num_bytes;
3426         u64 empty_size;
3427         u64 flags;
3428         int delalloc;
3429
3430         /* Where to start the search inside the bg */
3431         u64 search_start;
3432
3433         /* For clustered allocation */
3434         u64 empty_cluster;
3435         struct btrfs_free_cluster *last_ptr;
3436         bool use_cluster;
3437
3438         bool have_caching_bg;
3439         bool orig_have_caching_bg;
3440
3441         /* RAID index, converted from flags */
3442         int index;
3443
3444         /*
3445          * Current loop number, check find_free_extent_update_loop() for details
3446          */
3447         int loop;
3448
3449         /*
3450          * Whether we're refilling a cluster, if true we need to re-search
3451          * current block group but don't try to refill the cluster again.
3452          */
3453         bool retry_clustered;
3454
3455         /*
3456          * Whether we're updating free space cache, if true we need to re-search
3457          * current block group but don't try updating free space cache again.
3458          */
3459         bool retry_unclustered;
3460
3461         /* If current block group is cached */
3462         int cached;
3463
3464         /* Max contiguous hole found */
3465         u64 max_extent_size;
3466
3467         /* Total free space from free space cache, not always contiguous */
3468         u64 total_free_space;
3469
3470         /* Found result */
3471         u64 found_offset;
3472
3473         /* Hint where to start looking for an empty space */
3474         u64 hint_byte;
3475
3476         /* Allocation policy */
3477         enum btrfs_extent_allocation_policy policy;
3478 };
3479
3480
3481 /*
3482  * Helper function for find_free_extent().
3483  *
3484  * Return -ENOENT to inform caller that we need fallback to unclustered mode.
3485  * Return -EAGAIN to inform caller that we need to re-search this block group
3486  * Return >0 to inform caller that we find nothing
3487  * Return 0 means we have found a location and set ffe_ctl->found_offset.
3488  */
3489 static int find_free_extent_clustered(struct btrfs_block_group *bg,
3490                                       struct find_free_extent_ctl *ffe_ctl,
3491                                       struct btrfs_block_group **cluster_bg_ret)
3492 {
3493         struct btrfs_block_group *cluster_bg;
3494         struct btrfs_free_cluster *last_ptr = ffe_ctl->last_ptr;
3495         u64 aligned_cluster;
3496         u64 offset;
3497         int ret;
3498
3499         cluster_bg = btrfs_lock_cluster(bg, last_ptr, ffe_ctl->delalloc);
3500         if (!cluster_bg)
3501                 goto refill_cluster;
3502         if (cluster_bg != bg && (cluster_bg->ro ||
3503             !block_group_bits(cluster_bg, ffe_ctl->flags)))
3504                 goto release_cluster;
3505
3506         offset = btrfs_alloc_from_cluster(cluster_bg, last_ptr,
3507                         ffe_ctl->num_bytes, cluster_bg->start,
3508                         &ffe_ctl->max_extent_size);
3509         if (offset) {
3510                 /* We have a block, we're done */
3511                 spin_unlock(&last_ptr->refill_lock);
3512                 trace_btrfs_reserve_extent_cluster(cluster_bg,
3513                                 ffe_ctl->search_start, ffe_ctl->num_bytes);
3514                 *cluster_bg_ret = cluster_bg;
3515                 ffe_ctl->found_offset = offset;
3516                 return 0;
3517         }
3518         WARN_ON(last_ptr->block_group != cluster_bg);
3519
3520 release_cluster:
3521         /*
3522          * If we are on LOOP_NO_EMPTY_SIZE, we can't set up a new clusters, so
3523          * lets just skip it and let the allocator find whatever block it can
3524          * find. If we reach this point, we will have tried the cluster
3525          * allocator plenty of times and not have found anything, so we are
3526          * likely way too fragmented for the clustering stuff to find anything.
3527          *
3528          * However, if the cluster is taken from the current block group,
3529          * release the cluster first, so that we stand a better chance of
3530          * succeeding in the unclustered allocation.
3531          */
3532         if (ffe_ctl->loop >= LOOP_NO_EMPTY_SIZE && cluster_bg != bg) {
3533                 spin_unlock(&last_ptr->refill_lock);
3534                 btrfs_release_block_group(cluster_bg, ffe_ctl->delalloc);
3535                 return -ENOENT;
3536         }
3537
3538         /* This cluster didn't work out, free it and start over */
3539         btrfs_return_cluster_to_free_space(NULL, last_ptr);
3540
3541         if (cluster_bg != bg)
3542                 btrfs_release_block_group(cluster_bg, ffe_ctl->delalloc);
3543
3544 refill_cluster:
3545         if (ffe_ctl->loop >= LOOP_NO_EMPTY_SIZE) {
3546                 spin_unlock(&last_ptr->refill_lock);
3547                 return -ENOENT;
3548         }
3549
3550         aligned_cluster = max_t(u64,
3551                         ffe_ctl->empty_cluster + ffe_ctl->empty_size,
3552                         bg->full_stripe_len);
3553         ret = btrfs_find_space_cluster(bg, last_ptr, ffe_ctl->search_start,
3554                         ffe_ctl->num_bytes, aligned_cluster);
3555         if (ret == 0) {
3556                 /* Now pull our allocation out of this cluster */
3557                 offset = btrfs_alloc_from_cluster(bg, last_ptr,
3558                                 ffe_ctl->num_bytes, ffe_ctl->search_start,
3559                                 &ffe_ctl->max_extent_size);
3560                 if (offset) {
3561                         /* We found one, proceed */
3562                         spin_unlock(&last_ptr->refill_lock);
3563                         trace_btrfs_reserve_extent_cluster(bg,
3564                                         ffe_ctl->search_start,
3565                                         ffe_ctl->num_bytes);
3566                         ffe_ctl->found_offset = offset;
3567                         return 0;
3568                 }
3569         } else if (!ffe_ctl->cached && ffe_ctl->loop > LOOP_CACHING_NOWAIT &&
3570                    !ffe_ctl->retry_clustered) {
3571                 spin_unlock(&last_ptr->refill_lock);
3572
3573                 ffe_ctl->retry_clustered = true;
3574                 btrfs_wait_block_group_cache_progress(bg, ffe_ctl->num_bytes +
3575                                 ffe_ctl->empty_cluster + ffe_ctl->empty_size);
3576                 return -EAGAIN;
3577         }
3578         /*
3579          * At this point we either didn't find a cluster or we weren't able to
3580          * allocate a block from our cluster.  Free the cluster we've been
3581          * trying to use, and go to the next block group.
3582          */
3583         btrfs_return_cluster_to_free_space(NULL, last_ptr);
3584         spin_unlock(&last_ptr->refill_lock);
3585         return 1;
3586 }
3587
3588 /*
3589  * Return >0 to inform caller that we find nothing
3590  * Return 0 when we found an free extent and set ffe_ctrl->found_offset
3591  * Return -EAGAIN to inform caller that we need to re-search this block group
3592  */
3593 static int find_free_extent_unclustered(struct btrfs_block_group *bg,
3594                                         struct find_free_extent_ctl *ffe_ctl)
3595 {
3596         struct btrfs_free_cluster *last_ptr = ffe_ctl->last_ptr;
3597         u64 offset;
3598
3599         /*
3600          * We are doing an unclustered allocation, set the fragmented flag so
3601          * we don't bother trying to setup a cluster again until we get more
3602          * space.
3603          */
3604         if (unlikely(last_ptr)) {
3605                 spin_lock(&last_ptr->lock);
3606                 last_ptr->fragmented = 1;
3607                 spin_unlock(&last_ptr->lock);
3608         }
3609         if (ffe_ctl->cached) {
3610                 struct btrfs_free_space_ctl *free_space_ctl;
3611
3612                 free_space_ctl = bg->free_space_ctl;
3613                 spin_lock(&free_space_ctl->tree_lock);
3614                 if (free_space_ctl->free_space <
3615                     ffe_ctl->num_bytes + ffe_ctl->empty_cluster +
3616                     ffe_ctl->empty_size) {
3617                         ffe_ctl->total_free_space = max_t(u64,
3618                                         ffe_ctl->total_free_space,
3619                                         free_space_ctl->free_space);
3620                         spin_unlock(&free_space_ctl->tree_lock);
3621                         return 1;
3622                 }
3623                 spin_unlock(&free_space_ctl->tree_lock);
3624         }
3625
3626         offset = btrfs_find_space_for_alloc(bg, ffe_ctl->search_start,
3627                         ffe_ctl->num_bytes, ffe_ctl->empty_size,
3628                         &ffe_ctl->max_extent_size);
3629
3630         /*
3631          * If we didn't find a chunk, and we haven't failed on this block group
3632          * before, and this block group is in the middle of caching and we are
3633          * ok with waiting, then go ahead and wait for progress to be made, and
3634          * set @retry_unclustered to true.
3635          *
3636          * If @retry_unclustered is true then we've already waited on this
3637          * block group once and should move on to the next block group.
3638          */
3639         if (!offset && !ffe_ctl->retry_unclustered && !ffe_ctl->cached &&
3640             ffe_ctl->loop > LOOP_CACHING_NOWAIT) {
3641                 btrfs_wait_block_group_cache_progress(bg, ffe_ctl->num_bytes +
3642                                                       ffe_ctl->empty_size);
3643                 ffe_ctl->retry_unclustered = true;
3644                 return -EAGAIN;
3645         } else if (!offset) {
3646                 return 1;
3647         }
3648         ffe_ctl->found_offset = offset;
3649         return 0;
3650 }
3651
3652 static int do_allocation_clustered(struct btrfs_block_group *block_group,
3653                                    struct find_free_extent_ctl *ffe_ctl,
3654                                    struct btrfs_block_group **bg_ret)
3655 {
3656         int ret;
3657
3658         /* We want to try and use the cluster allocator, so lets look there */
3659         if (ffe_ctl->last_ptr && ffe_ctl->use_cluster) {
3660                 ret = find_free_extent_clustered(block_group, ffe_ctl, bg_ret);
3661                 if (ret >= 0 || ret == -EAGAIN)
3662                         return ret;
3663                 /* ret == -ENOENT case falls through */
3664         }
3665
3666         return find_free_extent_unclustered(block_group, ffe_ctl);
3667 }
3668
3669 static int do_allocation(struct btrfs_block_group *block_group,
3670                          struct find_free_extent_ctl *ffe_ctl,
3671                          struct btrfs_block_group **bg_ret)
3672 {
3673         switch (ffe_ctl->policy) {
3674         case BTRFS_EXTENT_ALLOC_CLUSTERED:
3675                 return do_allocation_clustered(block_group, ffe_ctl, bg_ret);
3676         default:
3677                 BUG();
3678         }
3679 }
3680
3681 static void release_block_group(struct btrfs_block_group *block_group,
3682                                 struct find_free_extent_ctl *ffe_ctl,
3683                                 int delalloc)
3684 {
3685         switch (ffe_ctl->policy) {
3686         case BTRFS_EXTENT_ALLOC_CLUSTERED:
3687                 ffe_ctl->retry_clustered = false;
3688                 ffe_ctl->retry_unclustered = false;
3689                 break;
3690         default:
3691                 BUG();
3692         }
3693
3694         BUG_ON(btrfs_bg_flags_to_raid_index(block_group->flags) !=
3695                ffe_ctl->index);
3696         btrfs_release_block_group(block_group, delalloc);
3697 }
3698
3699 static void found_extent_clustered(struct find_free_extent_ctl *ffe_ctl,
3700                                    struct btrfs_key *ins)
3701 {
3702         struct btrfs_free_cluster *last_ptr = ffe_ctl->last_ptr;
3703
3704         if (!ffe_ctl->use_cluster && last_ptr) {
3705                 spin_lock(&last_ptr->lock);
3706                 last_ptr->window_start = ins->objectid;
3707                 spin_unlock(&last_ptr->lock);
3708         }
3709 }
3710
3711 static void found_extent(struct find_free_extent_ctl *ffe_ctl,
3712                          struct btrfs_key *ins)
3713 {
3714         switch (ffe_ctl->policy) {
3715         case BTRFS_EXTENT_ALLOC_CLUSTERED:
3716                 found_extent_clustered(ffe_ctl, ins);
3717                 break;
3718         default:
3719                 BUG();
3720         }
3721 }
3722
3723 static int chunk_allocation_failed(struct find_free_extent_ctl *ffe_ctl)
3724 {
3725         switch (ffe_ctl->policy) {
3726         case BTRFS_EXTENT_ALLOC_CLUSTERED:
3727                 /*
3728                  * If we can't allocate a new chunk we've already looped through
3729                  * at least once, move on to the NO_EMPTY_SIZE case.
3730                  */
3731                 ffe_ctl->loop = LOOP_NO_EMPTY_SIZE;
3732                 return 0;
3733         default:
3734                 BUG();
3735         }
3736 }
3737
3738 /*
3739  * Return >0 means caller needs to re-search for free extent
3740  * Return 0 means we have the needed free extent.
3741  * Return <0 means we failed to locate any free extent.
3742  */
3743 static int find_free_extent_update_loop(struct btrfs_fs_info *fs_info,
3744                                         struct btrfs_key *ins,
3745                                         struct find_free_extent_ctl *ffe_ctl,
3746                                         bool full_search)
3747 {
3748         struct btrfs_root *root = fs_info->extent_root;
3749         int ret;
3750
3751         if ((ffe_ctl->loop == LOOP_CACHING_NOWAIT) &&
3752             ffe_ctl->have_caching_bg && !ffe_ctl->orig_have_caching_bg)
3753                 ffe_ctl->orig_have_caching_bg = true;
3754
3755         if (!ins->objectid && ffe_ctl->loop >= LOOP_CACHING_WAIT &&
3756             ffe_ctl->have_caching_bg)
3757                 return 1;
3758
3759         if (!ins->objectid && ++(ffe_ctl->index) < BTRFS_NR_RAID_TYPES)
3760                 return 1;
3761
3762         if (ins->objectid) {
3763                 found_extent(ffe_ctl, ins);
3764                 return 0;
3765         }
3766
3767         /*
3768          * LOOP_CACHING_NOWAIT, search partially cached block groups, kicking
3769          *                      caching kthreads as we move along
3770          * LOOP_CACHING_WAIT, search everything, and wait if our bg is caching
3771          * LOOP_ALLOC_CHUNK, force a chunk allocation and try again
3772          * LOOP_NO_EMPTY_SIZE, set empty_size and empty_cluster to 0 and try
3773          *                     again
3774          */
3775         if (ffe_ctl->loop < LOOP_NO_EMPTY_SIZE) {
3776                 ffe_ctl->index = 0;
3777                 if (ffe_ctl->loop == LOOP_CACHING_NOWAIT) {
3778                         /*
3779                          * We want to skip the LOOP_CACHING_WAIT step if we
3780                          * don't have any uncached bgs and we've already done a
3781                          * full search through.
3782                          */
3783                         if (ffe_ctl->orig_have_caching_bg || !full_search)
3784                                 ffe_ctl->loop = LOOP_CACHING_WAIT;
3785                         else
3786                                 ffe_ctl->loop = LOOP_ALLOC_CHUNK;
3787                 } else {
3788                         ffe_ctl->loop++;
3789                 }
3790
3791                 if (ffe_ctl->loop == LOOP_ALLOC_CHUNK) {
3792                         struct btrfs_trans_handle *trans;
3793                         int exist = 0;
3794
3795                         trans = current->journal_info;
3796                         if (trans)
3797                                 exist = 1;
3798                         else
3799                                 trans = btrfs_join_transaction(root);
3800
3801                         if (IS_ERR(trans)) {
3802                                 ret = PTR_ERR(trans);
3803                                 return ret;
3804                         }
3805
3806                         ret = btrfs_chunk_alloc(trans, ffe_ctl->flags,
3807                                                 CHUNK_ALLOC_FORCE);
3808
3809                         /* Do not bail out on ENOSPC since we can do more. */
3810                         if (ret == -ENOSPC)
3811                                 ret = chunk_allocation_failed(ffe_ctl);
3812                         else if (ret < 0)
3813                                 btrfs_abort_transaction(trans, ret);
3814                         else
3815                                 ret = 0;
3816                         if (!exist)
3817                                 btrfs_end_transaction(trans);
3818                         if (ret)
3819                                 return ret;
3820                 }
3821
3822                 if (ffe_ctl->loop == LOOP_NO_EMPTY_SIZE) {
3823                         if (ffe_ctl->policy != BTRFS_EXTENT_ALLOC_CLUSTERED)
3824                                 return -ENOSPC;
3825
3826                         /*
3827                          * Don't loop again if we already have no empty_size and
3828                          * no empty_cluster.
3829                          */
3830                         if (ffe_ctl->empty_size == 0 &&
3831                             ffe_ctl->empty_cluster == 0)
3832                                 return -ENOSPC;
3833                         ffe_ctl->empty_size = 0;
3834                         ffe_ctl->empty_cluster = 0;
3835                 }
3836                 return 1;
3837         }
3838         return -ENOSPC;
3839 }
3840
3841 static int prepare_allocation_clustered(struct btrfs_fs_info *fs_info,
3842                                         struct find_free_extent_ctl *ffe_ctl,
3843                                         struct btrfs_space_info *space_info,
3844                                         struct btrfs_key *ins)
3845 {
3846         /*
3847          * If our free space is heavily fragmented we may not be able to make
3848          * big contiguous allocations, so instead of doing the expensive search
3849          * for free space, simply return ENOSPC with our max_extent_size so we
3850          * can go ahead and search for a more manageable chunk.
3851          *
3852          * If our max_extent_size is large enough for our allocation simply
3853          * disable clustering since we will likely not be able to find enough
3854          * space to create a cluster and induce latency trying.
3855          */
3856         if (space_info->max_extent_size) {
3857                 spin_lock(&space_info->lock);
3858                 if (space_info->max_extent_size &&
3859                     ffe_ctl->num_bytes > space_info->max_extent_size) {
3860                         ins->offset = space_info->max_extent_size;
3861                         spin_unlock(&space_info->lock);
3862                         return -ENOSPC;
3863                 } else if (space_info->max_extent_size) {
3864                         ffe_ctl->use_cluster = false;
3865                 }
3866                 spin_unlock(&space_info->lock);
3867         }
3868
3869         ffe_ctl->last_ptr = fetch_cluster_info(fs_info, space_info,
3870                                                &ffe_ctl->empty_cluster);
3871         if (ffe_ctl->last_ptr) {
3872                 struct btrfs_free_cluster *last_ptr = ffe_ctl->last_ptr;
3873
3874                 spin_lock(&last_ptr->lock);
3875                 if (last_ptr->block_group)
3876                         ffe_ctl->hint_byte = last_ptr->window_start;
3877                 if (last_ptr->fragmented) {
3878                         /*
3879                          * We still set window_start so we can keep track of the
3880                          * last place we found an allocation to try and save
3881                          * some time.
3882                          */
3883                         ffe_ctl->hint_byte = last_ptr->window_start;
3884                         ffe_ctl->use_cluster = false;
3885                 }
3886                 spin_unlock(&last_ptr->lock);
3887         }
3888
3889         return 0;
3890 }
3891
3892 static int prepare_allocation(struct btrfs_fs_info *fs_info,
3893                               struct find_free_extent_ctl *ffe_ctl,
3894                               struct btrfs_space_info *space_info,
3895                               struct btrfs_key *ins)
3896 {
3897         switch (ffe_ctl->policy) {
3898         case BTRFS_EXTENT_ALLOC_CLUSTERED:
3899                 return prepare_allocation_clustered(fs_info, ffe_ctl,
3900                                                     space_info, ins);
3901         default:
3902                 BUG();
3903         }
3904 }
3905
3906 /*
3907  * walks the btree of allocated extents and find a hole of a given size.
3908  * The key ins is changed to record the hole:
3909  * ins->objectid == start position
3910  * ins->flags = BTRFS_EXTENT_ITEM_KEY
3911  * ins->offset == the size of the hole.
3912  * Any available blocks before search_start are skipped.
3913  *
3914  * If there is no suitable free space, we will record the max size of
3915  * the free space extent currently.
3916  *
3917  * The overall logic and call chain:
3918  *
3919  * find_free_extent()
3920  * |- Iterate through all block groups
3921  * |  |- Get a valid block group
3922  * |  |- Try to do clustered allocation in that block group
3923  * |  |- Try to do unclustered allocation in that block group
3924  * |  |- Check if the result is valid
3925  * |  |  |- If valid, then exit
3926  * |  |- Jump to next block group
3927  * |
3928  * |- Push harder to find free extents
3929  *    |- If not found, re-iterate all block groups
3930  */
3931 static noinline int find_free_extent(struct btrfs_fs_info *fs_info,
3932                                 u64 ram_bytes, u64 num_bytes, u64 empty_size,
3933                                 u64 hint_byte_orig, struct btrfs_key *ins,
3934                                 u64 flags, int delalloc)
3935 {
3936         int ret = 0;
3937         int cache_block_group_error = 0;
3938         struct btrfs_block_group *block_group = NULL;
3939         struct find_free_extent_ctl ffe_ctl = {0};
3940         struct btrfs_space_info *space_info;
3941         bool full_search = false;
3942
3943         WARN_ON(num_bytes < fs_info->sectorsize);
3944
3945         ffe_ctl.num_bytes = num_bytes;
3946         ffe_ctl.empty_size = empty_size;
3947         ffe_ctl.flags = flags;
3948         ffe_ctl.search_start = 0;
3949         ffe_ctl.delalloc = delalloc;
3950         ffe_ctl.index = btrfs_bg_flags_to_raid_index(flags);
3951         ffe_ctl.have_caching_bg = false;
3952         ffe_ctl.orig_have_caching_bg = false;
3953         ffe_ctl.found_offset = 0;
3954         ffe_ctl.hint_byte = hint_byte_orig;
3955         ffe_ctl.policy = BTRFS_EXTENT_ALLOC_CLUSTERED;
3956
3957         /* For clustered allocation */
3958         ffe_ctl.retry_clustered = false;
3959         ffe_ctl.retry_unclustered = false;
3960         ffe_ctl.last_ptr = NULL;
3961         ffe_ctl.use_cluster = true;
3962
3963         ins->type = BTRFS_EXTENT_ITEM_KEY;
3964         ins->objectid = 0;
3965         ins->offset = 0;
3966
3967         trace_find_free_extent(fs_info, num_bytes, empty_size, flags);
3968
3969         space_info = btrfs_find_space_info(fs_info, flags);
3970         if (!space_info) {
3971                 btrfs_err(fs_info, "No space info for %llu", flags);
3972                 return -ENOSPC;
3973         }
3974
3975         ret = prepare_allocation(fs_info, &ffe_ctl, space_info, ins);
3976         if (ret < 0)
3977                 return ret;
3978
3979         ffe_ctl.search_start = max(ffe_ctl.search_start,
3980                                    first_logical_byte(fs_info, 0));
3981         ffe_ctl.search_start = max(ffe_ctl.search_start, ffe_ctl.hint_byte);
3982         if (ffe_ctl.search_start == ffe_ctl.hint_byte) {
3983                 block_group = btrfs_lookup_block_group(fs_info,
3984                                                        ffe_ctl.search_start);
3985                 /*
3986                  * we don't want to use the block group if it doesn't match our
3987                  * allocation bits, or if its not cached.
3988                  *
3989                  * However if we are re-searching with an ideal block group
3990                  * picked out then we don't care that the block group is cached.
3991                  */
3992                 if (block_group && block_group_bits(block_group, flags) &&
3993                     block_group->cached != BTRFS_CACHE_NO) {
3994                         down_read(&space_info->groups_sem);
3995                         if (list_empty(&block_group->list) ||
3996                             block_group->ro) {
3997                                 /*
3998                                  * someone is removing this block group,
3999                                  * we can't jump into the have_block_group
4000                                  * target because our list pointers are not
4001                                  * valid
4002                                  */
4003                                 btrfs_put_block_group(block_group);
4004                                 up_read(&space_info->groups_sem);
4005                         } else {
4006                                 ffe_ctl.index = btrfs_bg_flags_to_raid_index(
4007                                                 block_group->flags);
4008                                 btrfs_lock_block_group(block_group, delalloc);
4009                                 goto have_block_group;
4010                         }
4011                 } else if (block_group) {
4012                         btrfs_put_block_group(block_group);
4013                 }
4014         }
4015 search:
4016         ffe_ctl.have_caching_bg = false;
4017         if (ffe_ctl.index == btrfs_bg_flags_to_raid_index(flags) ||
4018             ffe_ctl.index == 0)
4019                 full_search = true;
4020         down_read(&space_info->groups_sem);
4021         list_for_each_entry(block_group,
4022                             &space_info->block_groups[ffe_ctl.index], list) {
4023                 struct btrfs_block_group *bg_ret;
4024
4025                 /* If the block group is read-only, we can skip it entirely. */
4026                 if (unlikely(block_group->ro))
4027                         continue;
4028
4029                 btrfs_grab_block_group(block_group, delalloc);
4030                 ffe_ctl.search_start = block_group->start;
4031
4032                 /*
4033                  * this can happen if we end up cycling through all the
4034                  * raid types, but we want to make sure we only allocate
4035                  * for the proper type.
4036                  */
4037                 if (!block_group_bits(block_group, flags)) {
4038                         u64 extra = BTRFS_BLOCK_GROUP_DUP |
4039                                 BTRFS_BLOCK_GROUP_RAID1_MASK |
4040                                 BTRFS_BLOCK_GROUP_RAID56_MASK |
4041                                 BTRFS_BLOCK_GROUP_RAID10;
4042
4043                         /*
4044                          * if they asked for extra copies and this block group
4045                          * doesn't provide them, bail.  This does allow us to
4046                          * fill raid0 from raid1.
4047                          */
4048                         if ((flags & extra) && !(block_group->flags & extra))
4049                                 goto loop;
4050
4051                         /*
4052                          * This block group has different flags than we want.
4053                          * It's possible that we have MIXED_GROUP flag but no
4054                          * block group is mixed.  Just skip such block group.
4055                          */
4056                         btrfs_release_block_group(block_group, delalloc);
4057                         continue;
4058                 }
4059
4060 have_block_group:
4061                 ffe_ctl.cached = btrfs_block_group_done(block_group);
4062                 if (unlikely(!ffe_ctl.cached)) {
4063                         ffe_ctl.have_caching_bg = true;
4064                         ret = btrfs_cache_block_group(block_group, 0);
4065
4066                         /*
4067                          * If we get ENOMEM here or something else we want to
4068                          * try other block groups, because it may not be fatal.
4069                          * However if we can't find anything else we need to
4070                          * save our return here so that we return the actual
4071                          * error that caused problems, not ENOSPC.
4072                          */
4073                         if (ret < 0) {
4074                                 if (!cache_block_group_error)
4075                                         cache_block_group_error = ret;
4076                                 ret = 0;
4077                                 goto loop;
4078                         }
4079                         ret = 0;
4080                 }
4081
4082                 if (unlikely(block_group->cached == BTRFS_CACHE_ERROR))
4083                         goto loop;
4084
4085                 bg_ret = NULL;
4086                 ret = do_allocation(block_group, &ffe_ctl, &bg_ret);
4087                 if (ret == 0) {
4088                         if (bg_ret && bg_ret != block_group) {
4089                                 btrfs_release_block_group(block_group, delalloc);
4090                                 block_group = bg_ret;
4091                         }
4092                 } else if (ret == -EAGAIN) {
4093                         goto have_block_group;
4094                 } else if (ret > 0) {
4095                         goto loop;
4096                 }
4097
4098                 /* Checks */
4099                 ffe_ctl.search_start = round_up(ffe_ctl.found_offset,
4100                                              fs_info->stripesize);
4101
4102                 /* move on to the next group */
4103                 if (ffe_ctl.search_start + num_bytes >
4104                     block_group->start + block_group->length) {
4105                         btrfs_add_free_space(block_group, ffe_ctl.found_offset,
4106                                              num_bytes);
4107                         goto loop;
4108                 }
4109
4110                 if (ffe_ctl.found_offset < ffe_ctl.search_start)
4111                         btrfs_add_free_space(block_group, ffe_ctl.found_offset,
4112                                 ffe_ctl.search_start - ffe_ctl.found_offset);
4113
4114                 ret = btrfs_add_reserved_bytes(block_group, ram_bytes,
4115                                 num_bytes, delalloc);
4116                 if (ret == -EAGAIN) {
4117                         btrfs_add_free_space(block_group, ffe_ctl.found_offset,
4118                                              num_bytes);
4119                         goto loop;
4120                 }
4121                 btrfs_inc_block_group_reservations(block_group);
4122
4123                 /* we are all good, lets return */
4124                 ins->objectid = ffe_ctl.search_start;
4125                 ins->offset = num_bytes;
4126
4127                 trace_btrfs_reserve_extent(block_group, ffe_ctl.search_start,
4128                                            num_bytes);
4129                 btrfs_release_block_group(block_group, delalloc);
4130                 break;
4131 loop:
4132                 release_block_group(block_group, &ffe_ctl, delalloc);
4133                 cond_resched();
4134         }
4135         up_read(&space_info->groups_sem);
4136
4137         ret = find_free_extent_update_loop(fs_info, ins, &ffe_ctl, full_search);
4138         if (ret > 0)
4139                 goto search;
4140
4141         if (ret == -ENOSPC && !cache_block_group_error) {
4142                 /*
4143                  * Use ffe_ctl->total_free_space as fallback if we can't find
4144                  * any contiguous hole.
4145                  */
4146                 if (!ffe_ctl.max_extent_size)
4147                         ffe_ctl.max_extent_size = ffe_ctl.total_free_space;
4148                 spin_lock(&space_info->lock);
4149                 space_info->max_extent_size = ffe_ctl.max_extent_size;
4150                 spin_unlock(&space_info->lock);
4151                 ins->offset = ffe_ctl.max_extent_size;
4152         } else if (ret == -ENOSPC) {
4153                 ret = cache_block_group_error;
4154         }
4155         return ret;
4156 }
4157
4158 /*
4159  * btrfs_reserve_extent - entry point to the extent allocator. Tries to find a
4160  *                        hole that is at least as big as @num_bytes.
4161  *
4162  * @root           -    The root that will contain this extent
4163  *
4164  * @ram_bytes      -    The amount of space in ram that @num_bytes take. This
4165  *                      is used for accounting purposes. This value differs
4166  *                      from @num_bytes only in the case of compressed extents.
4167  *
4168  * @num_bytes      -    Number of bytes to allocate on-disk.
4169  *
4170  * @min_alloc_size -    Indicates the minimum amount of space that the
4171  *                      allocator should try to satisfy. In some cases
4172  *                      @num_bytes may be larger than what is required and if
4173  *                      the filesystem is fragmented then allocation fails.
4174  *                      However, the presence of @min_alloc_size gives a
4175  *                      chance to try and satisfy the smaller allocation.
4176  *
4177  * @empty_size     -    A hint that you plan on doing more COW. This is the
4178  *                      size in bytes the allocator should try to find free
4179  *                      next to the block it returns.  This is just a hint and
4180  *                      may be ignored by the allocator.
4181  *
4182  * @hint_byte      -    Hint to the allocator to start searching above the byte
4183  *                      address passed. It might be ignored.
4184  *
4185  * @ins            -    This key is modified to record the found hole. It will
4186  *                      have the following values:
4187  *                      ins->objectid == start position
4188  *                      ins->flags = BTRFS_EXTENT_ITEM_KEY
4189  *                      ins->offset == the size of the hole.
4190  *
4191  * @is_data        -    Boolean flag indicating whether an extent is
4192  *                      allocated for data (true) or metadata (false)
4193  *
4194  * @delalloc       -    Boolean flag indicating whether this allocation is for
4195  *                      delalloc or not. If 'true' data_rwsem of block groups
4196  *                      is going to be acquired.
4197  *
4198  *
4199  * Returns 0 when an allocation succeeded or < 0 when an error occurred. In
4200  * case -ENOSPC is returned then @ins->offset will contain the size of the
4201  * largest available hole the allocator managed to find.
4202  */
4203 int btrfs_reserve_extent(struct btrfs_root *root, u64 ram_bytes,
4204                          u64 num_bytes, u64 min_alloc_size,
4205                          u64 empty_size, u64 hint_byte,
4206                          struct btrfs_key *ins, int is_data, int delalloc)
4207 {
4208         struct btrfs_fs_info *fs_info = root->fs_info;
4209         bool final_tried = num_bytes == min_alloc_size;
4210         u64 flags;
4211         int ret;
4212
4213         flags = get_alloc_profile_by_root(root, is_data);
4214 again:
4215         WARN_ON(num_bytes < fs_info->sectorsize);
4216         ret = find_free_extent(fs_info, ram_bytes, num_bytes, empty_size,
4217                                hint_byte, ins, flags, delalloc);
4218         if (!ret && !is_data) {
4219                 btrfs_dec_block_group_reservations(fs_info, ins->objectid);
4220         } else if (ret == -ENOSPC) {
4221                 if (!final_tried && ins->offset) {
4222                         num_bytes = min(num_bytes >> 1, ins->offset);
4223                         num_bytes = round_down(num_bytes,
4224                                                fs_info->sectorsize);
4225                         num_bytes = max(num_bytes, min_alloc_size);
4226                         ram_bytes = num_bytes;
4227                         if (num_bytes == min_alloc_size)
4228                                 final_tried = true;
4229                         goto again;
4230                 } else if (btrfs_test_opt(fs_info, ENOSPC_DEBUG)) {
4231                         struct btrfs_space_info *sinfo;
4232
4233                         sinfo = btrfs_find_space_info(fs_info, flags);
4234                         btrfs_err(fs_info,
4235                                   "allocation failed flags %llu, wanted %llu",
4236                                   flags, num_bytes);
4237                         if (sinfo)
4238                                 btrfs_dump_space_info(fs_info, sinfo,
4239                                                       num_bytes, 1);
4240                 }
4241         }
4242
4243         return ret;
4244 }
4245
4246 int btrfs_free_reserved_extent(struct btrfs_fs_info *fs_info,
4247                                u64 start, u64 len, int delalloc)
4248 {
4249         struct btrfs_block_group *cache;
4250
4251         cache = btrfs_lookup_block_group(fs_info, start);
4252         if (!cache) {
4253                 btrfs_err(fs_info, "Unable to find block group for %llu",
4254                           start);
4255                 return -ENOSPC;
4256         }
4257
4258         btrfs_add_free_space(cache, start, len);
4259         btrfs_free_reserved_bytes(cache, len, delalloc);
4260         trace_btrfs_reserved_extent_free(fs_info, start, len);
4261
4262         btrfs_put_block_group(cache);
4263         return 0;
4264 }
4265
4266 int btrfs_pin_reserved_extent(struct btrfs_trans_handle *trans, u64 start,
4267                               u64 len)
4268 {
4269         struct btrfs_block_group *cache;
4270         int ret = 0;
4271
4272         cache = btrfs_lookup_block_group(trans->fs_info, start);
4273         if (!cache) {
4274                 btrfs_err(trans->fs_info, "unable to find block group for %llu",
4275                           start);
4276                 return -ENOSPC;
4277         }
4278
4279         ret = pin_down_extent(trans, cache, start, len, 1);
4280         btrfs_put_block_group(cache);
4281         return ret;
4282 }
4283
4284 static int alloc_reserved_file_extent(struct btrfs_trans_handle *trans,
4285                                       u64 parent, u64 root_objectid,
4286                                       u64 flags, u64 owner, u64 offset,
4287                                       struct btrfs_key *ins, int ref_mod)
4288 {
4289         struct btrfs_fs_info *fs_info = trans->fs_info;
4290         int ret;
4291         struct btrfs_extent_item *extent_item;
4292         struct btrfs_extent_inline_ref *iref;
4293         struct btrfs_path *path;
4294         struct extent_buffer *leaf;
4295         int type;
4296         u32 size;
4297
4298         if (parent > 0)
4299                 type = BTRFS_SHARED_DATA_REF_KEY;
4300         else
4301                 type = BTRFS_EXTENT_DATA_REF_KEY;
4302
4303         size = sizeof(*extent_item) + btrfs_extent_inline_ref_size(type);
4304
4305         path = btrfs_alloc_path();
4306         if (!path)
4307                 return -ENOMEM;
4308
4309         path->leave_spinning = 1;
4310         ret = btrfs_insert_empty_item(trans, fs_info->extent_root, path,
4311                                       ins, size);
4312         if (ret) {
4313                 btrfs_free_path(path);
4314                 return ret;
4315         }
4316
4317         leaf = path->nodes[0];
4318         extent_item = btrfs_item_ptr(leaf, path->slots[0],
4319                                      struct btrfs_extent_item);
4320         btrfs_set_extent_refs(leaf, extent_item, ref_mod);
4321         btrfs_set_extent_generation(leaf, extent_item, trans->transid);
4322         btrfs_set_extent_flags(leaf, extent_item,
4323                                flags | BTRFS_EXTENT_FLAG_DATA);
4324
4325         iref = (struct btrfs_extent_inline_ref *)(extent_item + 1);
4326         btrfs_set_extent_inline_ref_type(leaf, iref, type);
4327         if (parent > 0) {
4328                 struct btrfs_shared_data_ref *ref;
4329                 ref = (struct btrfs_shared_data_ref *)(iref + 1);
4330                 btrfs_set_extent_inline_ref_offset(leaf, iref, parent);
4331                 btrfs_set_shared_data_ref_count(leaf, ref, ref_mod);
4332         } else {
4333                 struct btrfs_extent_data_ref *ref;
4334                 ref = (struct btrfs_extent_data_ref *)(&iref->offset);
4335                 btrfs_set_extent_data_ref_root(leaf, ref, root_objectid);
4336                 btrfs_set_extent_data_ref_objectid(leaf, ref, owner);
4337                 btrfs_set_extent_data_ref_offset(leaf, ref, offset);
4338                 btrfs_set_extent_data_ref_count(leaf, ref, ref_mod);
4339         }
4340
4341         btrfs_mark_buffer_dirty(path->nodes[0]);
4342         btrfs_free_path(path);
4343
4344         ret = remove_from_free_space_tree(trans, ins->objectid, ins->offset);
4345         if (ret)
4346                 return ret;
4347
4348         ret = btrfs_update_block_group(trans, ins->objectid, ins->offset, 1);
4349         if (ret) { /* -ENOENT, logic error */
4350                 btrfs_err(fs_info, "update block group failed for %llu %llu",
4351                         ins->objectid, ins->offset);
4352                 BUG();
4353         }
4354         trace_btrfs_reserved_extent_alloc(fs_info, ins->objectid, ins->offset);
4355         return ret;
4356 }
4357
4358 static int alloc_reserved_tree_block(struct btrfs_trans_handle *trans,
4359                                      struct btrfs_delayed_ref_node *node,
4360                                      struct btrfs_delayed_extent_op *extent_op)
4361 {
4362         struct btrfs_fs_info *fs_info = trans->fs_info;
4363         int ret;
4364         struct btrfs_extent_item *extent_item;
4365         struct btrfs_key extent_key;
4366         struct btrfs_tree_block_info *block_info;
4367         struct btrfs_extent_inline_ref *iref;
4368         struct btrfs_path *path;
4369         struct extent_buffer *leaf;
4370         struct btrfs_delayed_tree_ref *ref;
4371         u32 size = sizeof(*extent_item) + sizeof(*iref);
4372         u64 num_bytes;
4373         u64 flags = extent_op->flags_to_set;
4374         bool skinny_metadata = btrfs_fs_incompat(fs_info, SKINNY_METADATA);
4375
4376         ref = btrfs_delayed_node_to_tree_ref(node);
4377
4378         extent_key.objectid = node->bytenr;
4379         if (skinny_metadata) {
4380                 extent_key.offset = ref->level;
4381                 extent_key.type = BTRFS_METADATA_ITEM_KEY;
4382                 num_bytes = fs_info->nodesize;
4383         } else {
4384                 extent_key.offset = node->num_bytes;
4385                 extent_key.type = BTRFS_EXTENT_ITEM_KEY;
4386                 size += sizeof(*block_info);
4387                 num_bytes = node->num_bytes;
4388         }
4389
4390         path = btrfs_alloc_path();
4391         if (!path)
4392                 return -ENOMEM;
4393
4394         path->leave_spinning = 1;
4395         ret = btrfs_insert_empty_item(trans, fs_info->extent_root, path,
4396                                       &extent_key, size);
4397         if (ret) {
4398                 btrfs_free_path(path);
4399                 return ret;
4400         }
4401
4402         leaf = path->nodes[0];
4403         extent_item = btrfs_item_ptr(leaf, path->slots[0],
4404                                      struct btrfs_extent_item);
4405         btrfs_set_extent_refs(leaf, extent_item, 1);
4406         btrfs_set_extent_generation(leaf, extent_item, trans->transid);
4407         btrfs_set_extent_flags(leaf, extent_item,
4408                                flags | BTRFS_EXTENT_FLAG_TREE_BLOCK);
4409
4410         if (skinny_metadata) {
4411                 iref = (struct btrfs_extent_inline_ref *)(extent_item + 1);
4412         } else {
4413                 block_info = (struct btrfs_tree_block_info *)(extent_item + 1);
4414                 btrfs_set_tree_block_key(leaf, block_info, &extent_op->key);
4415                 btrfs_set_tree_block_level(leaf, block_info, ref->level);
4416                 iref = (struct btrfs_extent_inline_ref *)(block_info + 1);
4417         }
4418
4419         if (node->type == BTRFS_SHARED_BLOCK_REF_KEY) {
4420                 BUG_ON(!(flags & BTRFS_BLOCK_FLAG_FULL_BACKREF));
4421                 btrfs_set_extent_inline_ref_type(leaf, iref,
4422                                                  BTRFS_SHARED_BLOCK_REF_KEY);
4423                 btrfs_set_extent_inline_ref_offset(leaf, iref, ref->parent);
4424         } else {
4425                 btrfs_set_extent_inline_ref_type(leaf, iref,
4426                                                  BTRFS_TREE_BLOCK_REF_KEY);
4427                 btrfs_set_extent_inline_ref_offset(leaf, iref, ref->root);
4428         }
4429
4430         btrfs_mark_buffer_dirty(leaf);
4431         btrfs_free_path(path);
4432
4433         ret = remove_from_free_space_tree(trans, extent_key.objectid,
4434                                           num_bytes);
4435         if (ret)
4436                 return ret;
4437
4438         ret = btrfs_update_block_group(trans, extent_key.objectid,
4439                                        fs_info->nodesize, 1);
4440         if (ret) { /* -ENOENT, logic error */
4441                 btrfs_err(fs_info, "update block group failed for %llu %llu",
4442                         extent_key.objectid, extent_key.offset);
4443                 BUG();
4444         }
4445
4446         trace_btrfs_reserved_extent_alloc(fs_info, extent_key.objectid,
4447                                           fs_info->nodesize);
4448         return ret;
4449 }
4450
4451 int btrfs_alloc_reserved_file_extent(struct btrfs_trans_handle *trans,
4452                                      struct btrfs_root *root, u64 owner,
4453                                      u64 offset, u64 ram_bytes,
4454                                      struct btrfs_key *ins)
4455 {
4456         struct btrfs_ref generic_ref = { 0 };
4457         int ret;
4458
4459         BUG_ON(root->root_key.objectid == BTRFS_TREE_LOG_OBJECTID);
4460
4461         btrfs_init_generic_ref(&generic_ref, BTRFS_ADD_DELAYED_EXTENT,
4462                                ins->objectid, ins->offset, 0);
4463         btrfs_init_data_ref(&generic_ref, root->root_key.objectid, owner, offset);
4464         btrfs_ref_tree_mod(root->fs_info, &generic_ref);
4465         ret = btrfs_add_delayed_data_ref(trans, &generic_ref,
4466                                          ram_bytes, NULL, NULL);
4467         return ret;
4468 }
4469
4470 /*
4471  * this is used by the tree logging recovery code.  It records that
4472  * an extent has been allocated and makes sure to clear the free
4473  * space cache bits as well
4474  */
4475 int btrfs_alloc_logged_file_extent(struct btrfs_trans_handle *trans,
4476                                    u64 root_objectid, u64 owner, u64 offset,
4477                                    struct btrfs_key *ins)
4478 {
4479         struct btrfs_fs_info *fs_info = trans->fs_info;
4480         int ret;
4481         struct btrfs_block_group *block_group;
4482         struct btrfs_space_info *space_info;
4483
4484         /*
4485          * Mixed block groups will exclude before processing the log so we only
4486          * need to do the exclude dance if this fs isn't mixed.
4487          */
4488         if (!btrfs_fs_incompat(fs_info, MIXED_GROUPS)) {
4489                 ret = __exclude_logged_extent(fs_info, ins->objectid,
4490                                               ins->offset);
4491                 if (ret)
4492                         return ret;
4493         }
4494
4495         block_group = btrfs_lookup_block_group(fs_info, ins->objectid);
4496         if (!block_group)
4497                 return -EINVAL;
4498
4499         space_info = block_group->space_info;
4500         spin_lock(&space_info->lock);
4501         spin_lock(&block_group->lock);
4502         space_info->bytes_reserved += ins->offset;
4503         block_group->reserved += ins->offset;
4504         spin_unlock(&block_group->lock);
4505         spin_unlock(&space_info->lock);
4506
4507         ret = alloc_reserved_file_extent(trans, 0, root_objectid, 0, owner,
4508                                          offset, ins, 1);
4509         if (ret)
4510                 btrfs_pin_extent(trans, ins->objectid, ins->offset, 1);
4511         btrfs_put_block_group(block_group);
4512         return ret;
4513 }
4514
4515 static struct extent_buffer *
4516 btrfs_init_new_buffer(struct btrfs_trans_handle *trans, struct btrfs_root *root,
4517                       u64 bytenr, int level, u64 owner)
4518 {
4519         struct btrfs_fs_info *fs_info = root->fs_info;
4520         struct extent_buffer *buf;
4521
4522         buf = btrfs_find_create_tree_block(fs_info, bytenr);
4523         if (IS_ERR(buf))
4524                 return buf;
4525
4526         /*
4527          * Extra safety check in case the extent tree is corrupted and extent
4528          * allocator chooses to use a tree block which is already used and
4529          * locked.
4530          */
4531         if (buf->lock_owner == current->pid) {
4532                 btrfs_err_rl(fs_info,
4533 "tree block %llu owner %llu already locked by pid=%d, extent tree corruption detected",
4534                         buf->start, btrfs_header_owner(buf), current->pid);
4535                 free_extent_buffer(buf);
4536                 return ERR_PTR(-EUCLEAN);
4537         }
4538
4539         btrfs_set_buffer_lockdep_class(root->root_key.objectid, buf, level);
4540         btrfs_tree_lock(buf);
4541         btrfs_clean_tree_block(buf);
4542         clear_bit(EXTENT_BUFFER_STALE, &buf->bflags);
4543
4544         btrfs_set_lock_blocking_write(buf);
4545         set_extent_buffer_uptodate(buf);
4546
4547         memzero_extent_buffer(buf, 0, sizeof(struct btrfs_header));
4548         btrfs_set_header_level(buf, level);
4549         btrfs_set_header_bytenr(buf, buf->start);
4550         btrfs_set_header_generation(buf, trans->transid);
4551         btrfs_set_header_backref_rev(buf, BTRFS_MIXED_BACKREF_REV);
4552         btrfs_set_header_owner(buf, owner);
4553         write_extent_buffer_fsid(buf, fs_info->fs_devices->metadata_uuid);
4554         write_extent_buffer_chunk_tree_uuid(buf, fs_info->chunk_tree_uuid);
4555         if (root->root_key.objectid == BTRFS_TREE_LOG_OBJECTID) {
4556                 buf->log_index = root->log_transid % 2;
4557                 /*
4558                  * we allow two log transactions at a time, use different
4559                  * EXTENT bit to differentiate dirty pages.
4560                  */
4561                 if (buf->log_index == 0)
4562                         set_extent_dirty(&root->dirty_log_pages, buf->start,
4563                                         buf->start + buf->len - 1, GFP_NOFS);
4564                 else
4565                         set_extent_new(&root->dirty_log_pages, buf->start,
4566                                         buf->start + buf->len - 1);
4567         } else {
4568                 buf->log_index = -1;
4569                 set_extent_dirty(&trans->transaction->dirty_pages, buf->start,
4570                          buf->start + buf->len - 1, GFP_NOFS);
4571         }
4572         trans->dirty = true;
4573         /* this returns a buffer locked for blocking */
4574         return buf;
4575 }
4576
4577 /*
4578  * finds a free extent and does all the dirty work required for allocation
4579  * returns the tree buffer or an ERR_PTR on error.
4580  */
4581 struct extent_buffer *btrfs_alloc_tree_block(struct btrfs_trans_handle *trans,
4582                                              struct btrfs_root *root,
4583                                              u64 parent, u64 root_objectid,
4584                                              const struct btrfs_disk_key *key,
4585                                              int level, u64 hint,
4586                                              u64 empty_size)
4587 {
4588         struct btrfs_fs_info *fs_info = root->fs_info;
4589         struct btrfs_key ins;
4590         struct btrfs_block_rsv *block_rsv;
4591         struct extent_buffer *buf;
4592         struct btrfs_delayed_extent_op *extent_op;
4593         struct btrfs_ref generic_ref = { 0 };
4594         u64 flags = 0;
4595         int ret;
4596         u32 blocksize = fs_info->nodesize;
4597         bool skinny_metadata = btrfs_fs_incompat(fs_info, SKINNY_METADATA);
4598
4599 #ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
4600         if (btrfs_is_testing(fs_info)) {
4601                 buf = btrfs_init_new_buffer(trans, root, root->alloc_bytenr,
4602                                             level, root_objectid);
4603                 if (!IS_ERR(buf))
4604                         root->alloc_bytenr += blocksize;
4605                 return buf;
4606         }
4607 #endif
4608
4609         block_rsv = btrfs_use_block_rsv(trans, root, blocksize);
4610         if (IS_ERR(block_rsv))
4611                 return ERR_CAST(block_rsv);
4612
4613         ret = btrfs_reserve_extent(root, blocksize, blocksize, blocksize,
4614                                    empty_size, hint, &ins, 0, 0);
4615         if (ret)
4616                 goto out_unuse;
4617
4618         buf = btrfs_init_new_buffer(trans, root, ins.objectid, level,
4619                                     root_objectid);
4620         if (IS_ERR(buf)) {
4621                 ret = PTR_ERR(buf);
4622                 goto out_free_reserved;
4623         }
4624
4625         if (root_objectid == BTRFS_TREE_RELOC_OBJECTID) {
4626                 if (parent == 0)
4627                         parent = ins.objectid;
4628                 flags |= BTRFS_BLOCK_FLAG_FULL_BACKREF;
4629         } else
4630                 BUG_ON(parent > 0);
4631
4632         if (root_objectid != BTRFS_TREE_LOG_OBJECTID) {
4633                 extent_op = btrfs_alloc_delayed_extent_op();
4634                 if (!extent_op) {
4635                         ret = -ENOMEM;
4636                         goto out_free_buf;
4637                 }
4638                 if (key)
4639                         memcpy(&extent_op->key, key, sizeof(extent_op->key));
4640                 else
4641                         memset(&extent_op->key, 0, sizeof(extent_op->key));
4642                 extent_op->flags_to_set = flags;
4643                 extent_op->update_key = skinny_metadata ? false : true;
4644                 extent_op->update_flags = true;
4645                 extent_op->is_data = false;
4646                 extent_op->level = level;
4647
4648                 btrfs_init_generic_ref(&generic_ref, BTRFS_ADD_DELAYED_EXTENT,
4649                                        ins.objectid, ins.offset, parent);
4650                 generic_ref.real_root = root->root_key.objectid;
4651                 btrfs_init_tree_ref(&generic_ref, level, root_objectid);
4652                 btrfs_ref_tree_mod(fs_info, &generic_ref);
4653                 ret = btrfs_add_delayed_tree_ref(trans, &generic_ref,
4654                                                  extent_op, NULL, NULL);
4655                 if (ret)
4656                         goto out_free_delayed;
4657         }
4658         return buf;
4659
4660 out_free_delayed:
4661         btrfs_free_delayed_extent_op(extent_op);
4662 out_free_buf:
4663         free_extent_buffer(buf);
4664 out_free_reserved:
4665         btrfs_free_reserved_extent(fs_info, ins.objectid, ins.offset, 0);
4666 out_unuse:
4667         btrfs_unuse_block_rsv(fs_info, block_rsv, blocksize);
4668         return ERR_PTR(ret);
4669 }
4670
4671 struct walk_control {
4672         u64 refs[BTRFS_MAX_LEVEL];
4673         u64 flags[BTRFS_MAX_LEVEL];
4674         struct btrfs_key update_progress;
4675         struct btrfs_key drop_progress;
4676         int drop_level;
4677         int stage;
4678         int level;
4679         int shared_level;
4680         int update_ref;
4681         int keep_locks;
4682         int reada_slot;
4683         int reada_count;
4684         int restarted;
4685 };
4686
4687 #define DROP_REFERENCE  1
4688 #define UPDATE_BACKREF  2
4689
4690 static noinline void reada_walk_down(struct btrfs_trans_handle *trans,
4691                                      struct btrfs_root *root,
4692                                      struct walk_control *wc,
4693                                      struct btrfs_path *path)
4694 {
4695         struct btrfs_fs_info *fs_info = root->fs_info;
4696         u64 bytenr;
4697         u64 generation;
4698         u64 refs;
4699         u64 flags;
4700         u32 nritems;
4701         struct btrfs_key key;
4702         struct extent_buffer *eb;
4703         int ret;
4704         int slot;
4705         int nread = 0;
4706
4707         if (path->slots[wc->level] < wc->reada_slot) {
4708                 wc->reada_count = wc->reada_count * 2 / 3;
4709                 wc->reada_count = max(wc->reada_count, 2);
4710         } else {
4711                 wc->reada_count = wc->reada_count * 3 / 2;
4712                 wc->reada_count = min_t(int, wc->reada_count,
4713                                         BTRFS_NODEPTRS_PER_BLOCK(fs_info));
4714         }
4715
4716         eb = path->nodes[wc->level];
4717         nritems = btrfs_header_nritems(eb);
4718
4719         for (slot = path->slots[wc->level]; slot < nritems; slot++) {
4720                 if (nread >= wc->reada_count)
4721                         break;
4722
4723                 cond_resched();
4724                 bytenr = btrfs_node_blockptr(eb, slot);
4725                 generation = btrfs_node_ptr_generation(eb, slot);
4726
4727                 if (slot == path->slots[wc->level])
4728                         goto reada;
4729
4730                 if (wc->stage == UPDATE_BACKREF &&
4731                     generation <= root->root_key.offset)
4732                         continue;
4733
4734                 /* We don't lock the tree block, it's OK to be racy here */
4735                 ret = btrfs_lookup_extent_info(trans, fs_info, bytenr,
4736                                                wc->level - 1, 1, &refs,
4737                                                &flags);
4738                 /* We don't care about errors in readahead. */
4739                 if (ret < 0)
4740                         continue;
4741                 BUG_ON(refs == 0);
4742
4743                 if (wc->stage == DROP_REFERENCE) {
4744                         if (refs == 1)
4745                                 goto reada;
4746
4747                         if (wc->level == 1 &&
4748                             (flags & BTRFS_BLOCK_FLAG_FULL_BACKREF))
4749                                 continue;
4750                         if (!wc->update_ref ||
4751                             generation <= root->root_key.offset)
4752                                 continue;
4753                         btrfs_node_key_to_cpu(eb, &key, slot);
4754                         ret = btrfs_comp_cpu_keys(&key,
4755                                                   &wc->update_progress);
4756                         if (ret < 0)
4757                                 continue;
4758                 } else {
4759                         if (wc->level == 1 &&
4760                             (flags & BTRFS_BLOCK_FLAG_FULL_BACKREF))
4761                                 continue;
4762                 }
4763 reada:
4764                 readahead_tree_block(fs_info, bytenr);
4765                 nread++;
4766         }
4767         wc->reada_slot = slot;
4768 }
4769
4770 /*
4771  * helper to process tree block while walking down the tree.
4772  *
4773  * when wc->stage == UPDATE_BACKREF, this function updates
4774  * back refs for pointers in the block.
4775  *
4776  * NOTE: return value 1 means we should stop walking down.
4777  */
4778 static noinline int walk_down_proc(struct btrfs_trans_handle *trans,
4779                                    struct btrfs_root *root,
4780                                    struct btrfs_path *path,
4781                                    struct walk_control *wc, int lookup_info)
4782 {
4783         struct btrfs_fs_info *fs_info = root->fs_info;
4784         int level = wc->level;
4785         struct extent_buffer *eb = path->nodes[level];
4786         u64 flag = BTRFS_BLOCK_FLAG_FULL_BACKREF;
4787         int ret;
4788
4789         if (wc->stage == UPDATE_BACKREF &&
4790             btrfs_header_owner(eb) != root->root_key.objectid)
4791                 return 1;
4792
4793         /*
4794          * when reference count of tree block is 1, it won't increase
4795          * again. once full backref flag is set, we never clear it.
4796          */
4797         if (lookup_info &&
4798             ((wc->stage == DROP_REFERENCE && wc->refs[level] != 1) ||
4799              (wc->stage == UPDATE_BACKREF && !(wc->flags[level] & flag)))) {
4800                 BUG_ON(!path->locks[level]);
4801                 ret = btrfs_lookup_extent_info(trans, fs_info,
4802                                                eb->start, level, 1,
4803                                                &wc->refs[level],
4804                                                &wc->flags[level]);
4805                 BUG_ON(ret == -ENOMEM);
4806                 if (ret)
4807                         return ret;
4808                 BUG_ON(wc->refs[level] == 0);
4809         }
4810
4811         if (wc->stage == DROP_REFERENCE) {
4812                 if (wc->refs[level] > 1)
4813                         return 1;
4814
4815                 if (path->locks[level] && !wc->keep_locks) {
4816                         btrfs_tree_unlock_rw(eb, path->locks[level]);
4817                         path->locks[level] = 0;
4818                 }
4819                 return 0;
4820         }
4821
4822         /* wc->stage == UPDATE_BACKREF */
4823         if (!(wc->flags[level] & flag)) {
4824                 BUG_ON(!path->locks[level]);
4825                 ret = btrfs_inc_ref(trans, root, eb, 1);
4826                 BUG_ON(ret); /* -ENOMEM */
4827                 ret = btrfs_dec_ref(trans, root, eb, 0);
4828                 BUG_ON(ret); /* -ENOMEM */
4829                 ret = btrfs_set_disk_extent_flags(trans, eb, flag,
4830                                                   btrfs_header_level(eb), 0);
4831                 BUG_ON(ret); /* -ENOMEM */
4832                 wc->flags[level] |= flag;
4833         }
4834
4835         /*
4836          * the block is shared by multiple trees, so it's not good to
4837          * keep the tree lock
4838          */
4839         if (path->locks[level] && level > 0) {
4840                 btrfs_tree_unlock_rw(eb, path->locks[level]);
4841                 path->locks[level] = 0;
4842         }
4843         return 0;
4844 }
4845
4846 /*
4847  * This is used to verify a ref exists for this root to deal with a bug where we
4848  * would have a drop_progress key that hadn't been updated properly.
4849  */
4850 static int check_ref_exists(struct btrfs_trans_handle *trans,
4851                             struct btrfs_root *root, u64 bytenr, u64 parent,
4852                             int level)
4853 {
4854         struct btrfs_path *path;
4855         struct btrfs_extent_inline_ref *iref;
4856         int ret;
4857
4858         path = btrfs_alloc_path();
4859         if (!path)
4860                 return -ENOMEM;
4861
4862         ret = lookup_extent_backref(trans, path, &iref, bytenr,
4863                                     root->fs_info->nodesize, parent,
4864                                     root->root_key.objectid, level, 0);
4865         btrfs_free_path(path);
4866         if (ret == -ENOENT)
4867                 return 0;
4868         if (ret < 0)
4869                 return ret;
4870         return 1;
4871 }
4872
4873 /*
4874  * helper to process tree block pointer.
4875  *
4876  * when wc->stage == DROP_REFERENCE, this function checks
4877  * reference count of the block pointed to. if the block
4878  * is shared and we need update back refs for the subtree
4879  * rooted at the block, this function changes wc->stage to
4880  * UPDATE_BACKREF. if the block is shared and there is no
4881  * need to update back, this function drops the reference
4882  * to the block.
4883  *
4884  * NOTE: return value 1 means we should stop walking down.
4885  */
4886 static noinline int do_walk_down(struct btrfs_trans_handle *trans,
4887                                  struct btrfs_root *root,
4888                                  struct btrfs_path *path,
4889                                  struct walk_control *wc, int *lookup_info)
4890 {
4891         struct btrfs_fs_info *fs_info = root->fs_info;
4892         u64 bytenr;
4893         u64 generation;
4894         u64 parent;
4895         struct btrfs_key key;
4896         struct btrfs_key first_key;
4897         struct btrfs_ref ref = { 0 };
4898         struct extent_buffer *next;
4899         int level = wc->level;
4900         int reada = 0;
4901         int ret = 0;
4902         bool need_account = false;
4903
4904         generation = btrfs_node_ptr_generation(path->nodes[level],
4905                                                path->slots[level]);
4906         /*
4907          * if the lower level block was created before the snapshot
4908          * was created, we know there is no need to update back refs
4909          * for the subtree
4910          */
4911         if (wc->stage == UPDATE_BACKREF &&
4912             generation <= root->root_key.offset) {
4913                 *lookup_info = 1;
4914                 return 1;
4915         }
4916
4917         bytenr = btrfs_node_blockptr(path->nodes[level], path->slots[level]);
4918         btrfs_node_key_to_cpu(path->nodes[level], &first_key,
4919                               path->slots[level]);
4920
4921         next = find_extent_buffer(fs_info, bytenr);
4922         if (!next) {
4923                 next = btrfs_find_create_tree_block(fs_info, bytenr);
4924                 if (IS_ERR(next))
4925                         return PTR_ERR(next);
4926
4927                 btrfs_set_buffer_lockdep_class(root->root_key.objectid, next,
4928                                                level - 1);
4929                 reada = 1;
4930         }
4931         btrfs_tree_lock(next);
4932         btrfs_set_lock_blocking_write(next);
4933
4934         ret = btrfs_lookup_extent_info(trans, fs_info, bytenr, level - 1, 1,
4935                                        &wc->refs[level - 1],
4936                                        &wc->flags[level - 1]);
4937         if (ret < 0)
4938                 goto out_unlock;
4939
4940         if (unlikely(wc->refs[level - 1] == 0)) {
4941                 btrfs_err(fs_info, "Missing references.");
4942                 ret = -EIO;
4943                 goto out_unlock;
4944         }
4945         *lookup_info = 0;
4946
4947         if (wc->stage == DROP_REFERENCE) {
4948                 if (wc->refs[level - 1] > 1) {
4949                         need_account = true;
4950                         if (level == 1 &&
4951                             (wc->flags[0] & BTRFS_BLOCK_FLAG_FULL_BACKREF))
4952                                 goto skip;
4953
4954                         if (!wc->update_ref ||
4955                             generation <= root->root_key.offset)
4956                                 goto skip;
4957
4958                         btrfs_node_key_to_cpu(path->nodes[level], &key,
4959                                               path->slots[level]);
4960                         ret = btrfs_comp_cpu_keys(&key, &wc->update_progress);
4961                         if (ret < 0)
4962                                 goto skip;
4963
4964                         wc->stage = UPDATE_BACKREF;
4965                         wc->shared_level = level - 1;
4966                 }
4967         } else {
4968                 if (level == 1 &&
4969                     (wc->flags[0] & BTRFS_BLOCK_FLAG_FULL_BACKREF))
4970                         goto skip;
4971         }
4972
4973         if (!btrfs_buffer_uptodate(next, generation, 0)) {
4974                 btrfs_tree_unlock(next);
4975                 free_extent_buffer(next);
4976                 next = NULL;
4977                 *lookup_info = 1;
4978         }
4979
4980         if (!next) {
4981                 if (reada && level == 1)
4982                         reada_walk_down(trans, root, wc, path);
4983                 next = read_tree_block(fs_info, bytenr, generation, level - 1,
4984                                        &first_key);
4985                 if (IS_ERR(next)) {
4986                         return PTR_ERR(next);
4987                 } else if (!extent_buffer_uptodate(next)) {
4988                         free_extent_buffer(next);
4989                         return -EIO;
4990                 }
4991                 btrfs_tree_lock(next);
4992                 btrfs_set_lock_blocking_write(next);
4993         }
4994
4995         level--;
4996         ASSERT(level == btrfs_header_level(next));
4997         if (level != btrfs_header_level(next)) {
4998                 btrfs_err(root->fs_info, "mismatched level");
4999                 ret = -EIO;
5000                 goto out_unlock;
5001         }
5002         path->nodes[level] = next;
5003         path->slots[level] = 0;
5004         path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING;
5005         wc->level = level;
5006         if (wc->level == 1)
5007                 wc->reada_slot = 0;
5008         return 0;
5009 skip:
5010         wc->refs[level - 1] = 0;
5011         wc->flags[level - 1] = 0;
5012         if (wc->stage == DROP_REFERENCE) {
5013                 if (wc->flags[level] & BTRFS_BLOCK_FLAG_FULL_BACKREF) {
5014                         parent = path->nodes[level]->start;
5015                 } else {
5016                         ASSERT(root->root_key.objectid ==
5017                                btrfs_header_owner(path->nodes[level]));
5018                         if (root->root_key.objectid !=
5019                             btrfs_header_owner(path->nodes[level])) {
5020                                 btrfs_err(root->fs_info,
5021                                                 "mismatched block owner");
5022                                 ret = -EIO;
5023                                 goto out_unlock;
5024                         }
5025                         parent = 0;
5026                 }
5027
5028                 /*
5029                  * If we had a drop_progress we need to verify the refs are set
5030                  * as expected.  If we find our ref then we know that from here
5031                  * on out everything should be correct, and we can clear the
5032                  * ->restarted flag.
5033                  */
5034                 if (wc->restarted) {
5035                         ret = check_ref_exists(trans, root, bytenr, parent,
5036                                                level - 1);
5037                         if (ret < 0)
5038                                 goto out_unlock;
5039                         if (ret == 0)
5040                                 goto no_delete;
5041                         ret = 0;
5042                         wc->restarted = 0;
5043                 }
5044
5045                 /*
5046                  * Reloc tree doesn't contribute to qgroup numbers, and we have
5047                  * already accounted them at merge time (replace_path),
5048                  * thus we could skip expensive subtree trace here.
5049                  */
5050                 if (root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID &&
5051                     need_account) {
5052                         ret = btrfs_qgroup_trace_subtree(trans, next,
5053                                                          generation, level - 1);
5054                         if (ret) {
5055                                 btrfs_err_rl(fs_info,
5056                                              "Error %d accounting shared subtree. Quota is out of sync, rescan required.",
5057                                              ret);
5058                         }
5059                 }
5060
5061                 /*
5062                  * We need to update the next key in our walk control so we can
5063                  * update the drop_progress key accordingly.  We don't care if
5064                  * find_next_key doesn't find a key because that means we're at
5065                  * the end and are going to clean up now.
5066                  */
5067                 wc->drop_level = level;
5068                 find_next_key(path, level, &wc->drop_progress);
5069
5070                 btrfs_init_generic_ref(&ref, BTRFS_DROP_DELAYED_REF, bytenr,
5071                                        fs_info->nodesize, parent);
5072                 btrfs_init_tree_ref(&ref, level - 1, root->root_key.objectid);
5073                 ret = btrfs_free_extent(trans, &ref);
5074                 if (ret)
5075                         goto out_unlock;
5076         }
5077 no_delete:
5078         *lookup_info = 1;
5079         ret = 1;
5080
5081 out_unlock:
5082         btrfs_tree_unlock(next);
5083         free_extent_buffer(next);
5084
5085         return ret;
5086 }
5087
5088 /*
5089  * helper to process tree block while walking up the tree.
5090  *
5091  * when wc->stage == DROP_REFERENCE, this function drops
5092  * reference count on the block.
5093  *
5094  * when wc->stage == UPDATE_BACKREF, this function changes
5095  * wc->stage back to DROP_REFERENCE if we changed wc->stage
5096  * to UPDATE_BACKREF previously while processing the block.
5097  *
5098  * NOTE: return value 1 means we should stop walking up.
5099  */
5100 static noinline int walk_up_proc(struct btrfs_trans_handle *trans,
5101                                  struct btrfs_root *root,
5102                                  struct btrfs_path *path,
5103                                  struct walk_control *wc)
5104 {
5105         struct btrfs_fs_info *fs_info = root->fs_info;
5106         int ret;
5107         int level = wc->level;
5108         struct extent_buffer *eb = path->nodes[level];
5109         u64 parent = 0;
5110
5111         if (wc->stage == UPDATE_BACKREF) {
5112                 BUG_ON(wc->shared_level < level);
5113                 if (level < wc->shared_level)
5114                         goto out;
5115
5116                 ret = find_next_key(path, level + 1, &wc->update_progress);
5117                 if (ret > 0)
5118                         wc->update_ref = 0;
5119
5120                 wc->stage = DROP_REFERENCE;
5121                 wc->shared_level = -1;
5122                 path->slots[level] = 0;
5123
5124                 /*
5125                  * check reference count again if the block isn't locked.
5126                  * we should start walking down the tree again if reference
5127                  * count is one.
5128                  */
5129                 if (!path->locks[level]) {
5130                         BUG_ON(level == 0);
5131                         btrfs_tree_lock(eb);
5132                         btrfs_set_lock_blocking_write(eb);
5133                         path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING;
5134
5135                         ret = btrfs_lookup_extent_info(trans, fs_info,
5136                                                        eb->start, level, 1,
5137                                                        &wc->refs[level],
5138                                                        &wc->flags[level]);
5139                         if (ret < 0) {
5140                                 btrfs_tree_unlock_rw(eb, path->locks[level]);
5141                                 path->locks[level] = 0;
5142                                 return ret;
5143                         }
5144                         BUG_ON(wc->refs[level] == 0);
5145                         if (wc->refs[level] == 1) {
5146                                 btrfs_tree_unlock_rw(eb, path->locks[level]);
5147                                 path->locks[level] = 0;
5148                                 return 1;
5149                         }
5150                 }
5151         }
5152
5153         /* wc->stage == DROP_REFERENCE */
5154         BUG_ON(wc->refs[level] > 1 && !path->locks[level]);
5155
5156         if (wc->refs[level] == 1) {
5157                 if (level == 0) {
5158                         if (wc->flags[level] & BTRFS_BLOCK_FLAG_FULL_BACKREF)
5159                                 ret = btrfs_dec_ref(trans, root, eb, 1);
5160                         else
5161                                 ret = btrfs_dec_ref(trans, root, eb, 0);
5162                         BUG_ON(ret); /* -ENOMEM */
5163                         if (is_fstree(root->root_key.objectid)) {
5164                                 ret = btrfs_qgroup_trace_leaf_items(trans, eb);
5165                                 if (ret) {
5166                                         btrfs_err_rl(fs_info,
5167         "error %d accounting leaf items, quota is out of sync, rescan required",
5168                                              ret);
5169                                 }
5170                         }
5171                 }
5172                 /* make block locked assertion in btrfs_clean_tree_block happy */
5173                 if (!path->locks[level] &&
5174                     btrfs_header_generation(eb) == trans->transid) {
5175                         btrfs_tree_lock(eb);
5176                         btrfs_set_lock_blocking_write(eb);
5177                         path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING;
5178                 }
5179                 btrfs_clean_tree_block(eb);
5180         }
5181
5182         if (eb == root->node) {
5183                 if (wc->flags[level] & BTRFS_BLOCK_FLAG_FULL_BACKREF)
5184                         parent = eb->start;
5185                 else if (root->root_key.objectid != btrfs_header_owner(eb))
5186                         goto owner_mismatch;
5187         } else {
5188                 if (wc->flags[level + 1] & BTRFS_BLOCK_FLAG_FULL_BACKREF)
5189                         parent = path->nodes[level + 1]->start;
5190                 else if (root->root_key.objectid !=
5191                          btrfs_header_owner(path->nodes[level + 1]))
5192                         goto owner_mismatch;
5193         }
5194
5195         btrfs_free_tree_block(trans, root, eb, parent, wc->refs[level] == 1);
5196 out:
5197         wc->refs[level] = 0;
5198         wc->flags[level] = 0;
5199         return 0;
5200
5201 owner_mismatch:
5202         btrfs_err_rl(fs_info, "unexpected tree owner, have %llu expect %llu",
5203                      btrfs_header_owner(eb), root->root_key.objectid);
5204         return -EUCLEAN;
5205 }
5206
5207 static noinline int walk_down_tree(struct btrfs_trans_handle *trans,
5208                                    struct btrfs_root *root,
5209                                    struct btrfs_path *path,
5210                                    struct walk_control *wc)
5211 {
5212         int level = wc->level;
5213         int lookup_info = 1;
5214         int ret;
5215
5216         while (level >= 0) {
5217                 ret = walk_down_proc(trans, root, path, wc, lookup_info);
5218                 if (ret > 0)
5219                         break;
5220
5221                 if (level == 0)
5222                         break;
5223
5224                 if (path->slots[level] >=
5225                     btrfs_header_nritems(path->nodes[level]))
5226                         break;
5227
5228                 ret = do_walk_down(trans, root, path, wc, &lookup_info);
5229                 if (ret > 0) {
5230                         path->slots[level]++;
5231                         continue;
5232                 } else if (ret < 0)
5233                         return ret;
5234                 level = wc->level;
5235         }
5236         return 0;
5237 }
5238
5239 static noinline int walk_up_tree(struct btrfs_trans_handle *trans,
5240                                  struct btrfs_root *root,
5241                                  struct btrfs_path *path,
5242                                  struct walk_control *wc, int max_level)
5243 {
5244         int level = wc->level;
5245         int ret;
5246
5247         path->slots[level] = btrfs_header_nritems(path->nodes[level]);
5248         while (level < max_level && path->nodes[level]) {
5249                 wc->level = level;
5250                 if (path->slots[level] + 1 <
5251                     btrfs_header_nritems(path->nodes[level])) {
5252                         path->slots[level]++;
5253                         return 0;
5254                 } else {
5255                         ret = walk_up_proc(trans, root, path, wc);
5256                         if (ret > 0)
5257                                 return 0;
5258                         if (ret < 0)
5259                                 return ret;
5260
5261                         if (path->locks[level]) {
5262                                 btrfs_tree_unlock_rw(path->nodes[level],
5263                                                      path->locks[level]);
5264                                 path->locks[level] = 0;
5265                         }
5266                         free_extent_buffer(path->nodes[level]);
5267                         path->nodes[level] = NULL;
5268                         level++;
5269                 }
5270         }
5271         return 1;
5272 }
5273
5274 /*
5275  * drop a subvolume tree.
5276  *
5277  * this function traverses the tree freeing any blocks that only
5278  * referenced by the tree.
5279  *
5280  * when a shared tree block is found. this function decreases its
5281  * reference count by one. if update_ref is true, this function
5282  * also make sure backrefs for the shared block and all lower level
5283  * blocks are properly updated.
5284  *
5285  * If called with for_reloc == 0, may exit early with -EAGAIN
5286  */
5287 int btrfs_drop_snapshot(struct btrfs_root *root, int update_ref, int for_reloc)
5288 {
5289         struct btrfs_fs_info *fs_info = root->fs_info;
5290         struct btrfs_path *path;
5291         struct btrfs_trans_handle *trans;
5292         struct btrfs_root *tree_root = fs_info->tree_root;
5293         struct btrfs_root_item *root_item = &root->root_item;
5294         struct walk_control *wc;
5295         struct btrfs_key key;
5296         int err = 0;
5297         int ret;
5298         int level;
5299         bool root_dropped = false;
5300
5301         btrfs_debug(fs_info, "Drop subvolume %llu", root->root_key.objectid);
5302
5303         path = btrfs_alloc_path();
5304         if (!path) {
5305                 err = -ENOMEM;
5306                 goto out;
5307         }
5308
5309         wc = kzalloc(sizeof(*wc), GFP_NOFS);
5310         if (!wc) {
5311                 btrfs_free_path(path);
5312                 err = -ENOMEM;
5313                 goto out;
5314         }
5315
5316         trans = btrfs_start_transaction(tree_root, 0);
5317         if (IS_ERR(trans)) {
5318                 err = PTR_ERR(trans);
5319                 goto out_free;
5320         }
5321
5322         err = btrfs_run_delayed_items(trans);
5323         if (err)
5324                 goto out_end_trans;
5325
5326         /*
5327          * This will help us catch people modifying the fs tree while we're
5328          * dropping it.  It is unsafe to mess with the fs tree while it's being
5329          * dropped as we unlock the root node and parent nodes as we walk down
5330          * the tree, assuming nothing will change.  If something does change
5331          * then we'll have stale information and drop references to blocks we've
5332          * already dropped.
5333          */
5334         set_bit(BTRFS_ROOT_DELETING, &root->state);
5335         if (btrfs_disk_key_objectid(&root_item->drop_progress) == 0) {
5336                 level = btrfs_header_level(root->node);
5337                 path->nodes[level] = btrfs_lock_root_node(root);
5338                 btrfs_set_lock_blocking_write(path->nodes[level]);
5339                 path->slots[level] = 0;
5340                 path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING;
5341                 memset(&wc->update_progress, 0,
5342                        sizeof(wc->update_progress));
5343         } else {
5344                 btrfs_disk_key_to_cpu(&key, &root_item->drop_progress);
5345                 memcpy(&wc->update_progress, &key,
5346                        sizeof(wc->update_progress));
5347
5348                 level = root_item->drop_level;
5349                 BUG_ON(level == 0);
5350                 path->lowest_level = level;
5351                 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
5352                 path->lowest_level = 0;
5353                 if (ret < 0) {
5354                         err = ret;
5355                         goto out_end_trans;
5356                 }
5357                 WARN_ON(ret > 0);
5358
5359                 /*
5360                  * unlock our path, this is safe because only this
5361                  * function is allowed to delete this snapshot
5362                  */
5363                 btrfs_unlock_up_safe(path, 0);
5364
5365                 level = btrfs_header_level(root->node);
5366                 while (1) {
5367                         btrfs_tree_lock(path->nodes[level]);
5368                         btrfs_set_lock_blocking_write(path->nodes[level]);
5369                         path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING;
5370
5371                         ret = btrfs_lookup_extent_info(trans, fs_info,
5372                                                 path->nodes[level]->start,
5373                                                 level, 1, &wc->refs[level],
5374                                                 &wc->flags[level]);
5375                         if (ret < 0) {
5376                                 err = ret;
5377                                 goto out_end_trans;
5378                         }
5379                         BUG_ON(wc->refs[level] == 0);
5380
5381                         if (level == root_item->drop_level)
5382                                 break;
5383
5384                         btrfs_tree_unlock(path->nodes[level]);
5385                         path->locks[level] = 0;
5386                         WARN_ON(wc->refs[level] != 1);
5387                         level--;
5388                 }
5389         }
5390
5391         wc->restarted = test_bit(BTRFS_ROOT_DEAD_TREE, &root->state);
5392         wc->level = level;
5393         wc->shared_level = -1;
5394         wc->stage = DROP_REFERENCE;
5395         wc->update_ref = update_ref;
5396         wc->keep_locks = 0;
5397         wc->reada_count = BTRFS_NODEPTRS_PER_BLOCK(fs_info);
5398
5399         while (1) {
5400
5401                 ret = walk_down_tree(trans, root, path, wc);
5402                 if (ret < 0) {
5403                         err = ret;
5404                         break;
5405                 }
5406
5407                 ret = walk_up_tree(trans, root, path, wc, BTRFS_MAX_LEVEL);
5408                 if (ret < 0) {
5409                         err = ret;
5410                         break;
5411                 }
5412
5413                 if (ret > 0) {
5414                         BUG_ON(wc->stage != DROP_REFERENCE);
5415                         break;
5416                 }
5417
5418                 if (wc->stage == DROP_REFERENCE) {
5419                         wc->drop_level = wc->level;
5420                         btrfs_node_key_to_cpu(path->nodes[wc->drop_level],
5421                                               &wc->drop_progress,
5422                                               path->slots[wc->drop_level]);
5423                 }
5424                 btrfs_cpu_key_to_disk(&root_item->drop_progress,
5425                                       &wc->drop_progress);
5426                 root_item->drop_level = wc->drop_level;
5427
5428                 BUG_ON(wc->level == 0);
5429                 if (btrfs_should_end_transaction(trans) ||
5430                     (!for_reloc && btrfs_need_cleaner_sleep(fs_info))) {
5431                         ret = btrfs_update_root(trans, tree_root,
5432                                                 &root->root_key,
5433                                                 root_item);
5434                         if (ret) {
5435                                 btrfs_abort_transaction(trans, ret);
5436                                 err = ret;
5437                                 goto out_end_trans;
5438                         }
5439
5440                         btrfs_end_transaction_throttle(trans);
5441                         if (!for_reloc && btrfs_need_cleaner_sleep(fs_info)) {
5442                                 btrfs_debug(fs_info,
5443                                             "drop snapshot early exit");
5444                                 err = -EAGAIN;
5445                                 goto out_free;
5446                         }
5447
5448                         trans = btrfs_start_transaction(tree_root, 0);
5449                         if (IS_ERR(trans)) {
5450                                 err = PTR_ERR(trans);
5451                                 goto out_free;
5452                         }
5453                 }
5454         }
5455         btrfs_release_path(path);
5456         if (err)
5457                 goto out_end_trans;
5458
5459         ret = btrfs_del_root(trans, &root->root_key);
5460         if (ret) {
5461                 btrfs_abort_transaction(trans, ret);
5462                 err = ret;
5463                 goto out_end_trans;
5464         }
5465
5466         if (root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID) {
5467                 ret = btrfs_find_root(tree_root, &root->root_key, path,
5468                                       NULL, NULL);
5469                 if (ret < 0) {
5470                         btrfs_abort_transaction(trans, ret);
5471                         err = ret;
5472                         goto out_end_trans;
5473                 } else if (ret > 0) {
5474                         /* if we fail to delete the orphan item this time
5475                          * around, it'll get picked up the next time.
5476                          *
5477                          * The most common failure here is just -ENOENT.
5478                          */
5479                         btrfs_del_orphan_item(trans, tree_root,
5480                                               root->root_key.objectid);
5481                 }
5482         }
5483
5484         if (test_bit(BTRFS_ROOT_IN_RADIX, &root->state))
5485                 btrfs_add_dropped_root(trans, root);
5486         else
5487                 btrfs_put_root(root);
5488         root_dropped = true;
5489 out_end_trans:
5490         btrfs_end_transaction_throttle(trans);
5491 out_free:
5492         kfree(wc);
5493         btrfs_free_path(path);
5494 out:
5495         /*
5496          * So if we need to stop dropping the snapshot for whatever reason we
5497          * need to make sure to add it back to the dead root list so that we
5498          * keep trying to do the work later.  This also cleans up roots if we
5499          * don't have it in the radix (like when we recover after a power fail
5500          * or unmount) so we don't leak memory.
5501          */
5502         if (!for_reloc && !root_dropped)
5503                 btrfs_add_dead_root(root);
5504         if (err && err != -EAGAIN)
5505                 btrfs_handle_fs_error(fs_info, err, NULL);
5506         return err;
5507 }
5508
5509 /*
5510  * drop subtree rooted at tree block 'node'.
5511  *
5512  * NOTE: this function will unlock and release tree block 'node'
5513  * only used by relocation code
5514  */
5515 int btrfs_drop_subtree(struct btrfs_trans_handle *trans,
5516                         struct btrfs_root *root,
5517                         struct extent_buffer *node,
5518                         struct extent_buffer *parent)
5519 {
5520         struct btrfs_fs_info *fs_info = root->fs_info;
5521         struct btrfs_path *path;
5522         struct walk_control *wc;
5523         int level;
5524         int parent_level;
5525         int ret = 0;
5526         int wret;
5527
5528         BUG_ON(root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID);
5529
5530         path = btrfs_alloc_path();
5531         if (!path)
5532                 return -ENOMEM;
5533
5534         wc = kzalloc(sizeof(*wc), GFP_NOFS);
5535         if (!wc) {
5536                 btrfs_free_path(path);
5537                 return -ENOMEM;
5538         }
5539
5540         btrfs_assert_tree_locked(parent);
5541         parent_level = btrfs_header_level(parent);
5542         atomic_inc(&parent->refs);
5543         path->nodes[parent_level] = parent;
5544         path->slots[parent_level] = btrfs_header_nritems(parent);
5545
5546         btrfs_assert_tree_locked(node);
5547         level = btrfs_header_level(node);
5548         path->nodes[level] = node;
5549         path->slots[level] = 0;
5550         path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING;
5551
5552         wc->refs[parent_level] = 1;
5553         wc->flags[parent_level] = BTRFS_BLOCK_FLAG_FULL_BACKREF;
5554         wc->level = level;
5555         wc->shared_level = -1;
5556         wc->stage = DROP_REFERENCE;
5557         wc->update_ref = 0;
5558         wc->keep_locks = 1;
5559         wc->reada_count = BTRFS_NODEPTRS_PER_BLOCK(fs_info);
5560
5561         while (1) {
5562                 wret = walk_down_tree(trans, root, path, wc);
5563                 if (wret < 0) {
5564                         ret = wret;
5565                         break;
5566                 }
5567
5568                 wret = walk_up_tree(trans, root, path, wc, parent_level);
5569                 if (wret < 0)
5570                         ret = wret;
5571                 if (wret != 0)
5572                         break;
5573         }
5574
5575         kfree(wc);
5576         btrfs_free_path(path);
5577         return ret;
5578 }
5579
5580 /*
5581  * helper to account the unused space of all the readonly block group in the
5582  * space_info. takes mirrors into account.
5583  */
5584 u64 btrfs_account_ro_block_groups_free_space(struct btrfs_space_info *sinfo)
5585 {
5586         struct btrfs_block_group *block_group;
5587         u64 free_bytes = 0;
5588         int factor;
5589
5590         /* It's df, we don't care if it's racy */
5591         if (list_empty(&sinfo->ro_bgs))
5592                 return 0;
5593
5594         spin_lock(&sinfo->lock);
5595         list_for_each_entry(block_group, &sinfo->ro_bgs, ro_list) {
5596                 spin_lock(&block_group->lock);
5597
5598                 if (!block_group->ro) {
5599                         spin_unlock(&block_group->lock);
5600                         continue;
5601                 }
5602
5603                 factor = btrfs_bg_type_to_factor(block_group->flags);
5604                 free_bytes += (block_group->length -
5605                                block_group->used) * factor;
5606
5607                 spin_unlock(&block_group->lock);
5608         }
5609         spin_unlock(&sinfo->lock);
5610
5611         return free_bytes;
5612 }
5613
5614 int btrfs_error_unpin_extent_range(struct btrfs_fs_info *fs_info,
5615                                    u64 start, u64 end)
5616 {
5617         return unpin_extent_range(fs_info, start, end, false);
5618 }
5619
5620 /*
5621  * It used to be that old block groups would be left around forever.
5622  * Iterating over them would be enough to trim unused space.  Since we
5623  * now automatically remove them, we also need to iterate over unallocated
5624  * space.
5625  *
5626  * We don't want a transaction for this since the discard may take a
5627  * substantial amount of time.  We don't require that a transaction be
5628  * running, but we do need to take a running transaction into account
5629  * to ensure that we're not discarding chunks that were released or
5630  * allocated in the current transaction.
5631  *
5632  * Holding the chunks lock will prevent other threads from allocating
5633  * or releasing chunks, but it won't prevent a running transaction
5634  * from committing and releasing the memory that the pending chunks
5635  * list head uses.  For that, we need to take a reference to the
5636  * transaction and hold the commit root sem.  We only need to hold
5637  * it while performing the free space search since we have already
5638  * held back allocations.
5639  */
5640 static int btrfs_trim_free_extents(struct btrfs_device *device, u64 *trimmed)
5641 {
5642         u64 start = SZ_1M, len = 0, end = 0;
5643         int ret;
5644
5645         *trimmed = 0;
5646
5647         /* Discard not supported = nothing to do. */
5648         if (!blk_queue_discard(bdev_get_queue(device->bdev)))
5649                 return 0;
5650
5651         /* Not writable = nothing to do. */
5652         if (!test_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state))
5653                 return 0;
5654
5655         /* No free space = nothing to do. */
5656         if (device->total_bytes <= device->bytes_used)
5657                 return 0;
5658
5659         ret = 0;
5660
5661         while (1) {
5662                 struct btrfs_fs_info *fs_info = device->fs_info;
5663                 u64 bytes;
5664
5665                 ret = mutex_lock_interruptible(&fs_info->chunk_mutex);
5666                 if (ret)
5667                         break;
5668
5669                 find_first_clear_extent_bit(&device->alloc_state, start,
5670                                             &start, &end,
5671                                             CHUNK_TRIMMED | CHUNK_ALLOCATED);
5672
5673                 /* Ensure we skip the reserved area in the first 1M */
5674                 start = max_t(u64, start, SZ_1M);
5675
5676                 /*
5677                  * If find_first_clear_extent_bit find a range that spans the
5678                  * end of the device it will set end to -1, in this case it's up
5679                  * to the caller to trim the value to the size of the device.
5680                  */
5681                 end = min(end, device->total_bytes - 1);
5682
5683                 len = end - start + 1;
5684
5685                 /* We didn't find any extents */
5686                 if (!len) {
5687                         mutex_unlock(&fs_info->chunk_mutex);
5688                         ret = 0;
5689                         break;
5690                 }
5691
5692                 ret = btrfs_issue_discard(device->bdev, start, len,
5693                                           &bytes);
5694                 if (!ret)
5695                         set_extent_bits(&device->alloc_state, start,
5696                                         start + bytes - 1,
5697                                         CHUNK_TRIMMED);
5698                 mutex_unlock(&fs_info->chunk_mutex);
5699
5700                 if (ret)
5701                         break;
5702
5703                 start += len;
5704                 *trimmed += bytes;
5705
5706                 if (fatal_signal_pending(current)) {
5707                         ret = -ERESTARTSYS;
5708                         break;
5709                 }
5710
5711                 cond_resched();
5712         }
5713
5714         return ret;
5715 }
5716
5717 /*
5718  * Trim the whole filesystem by:
5719  * 1) trimming the free space in each block group
5720  * 2) trimming the unallocated space on each device
5721  *
5722  * This will also continue trimming even if a block group or device encounters
5723  * an error.  The return value will be the last error, or 0 if nothing bad
5724  * happens.
5725  */
5726 int btrfs_trim_fs(struct btrfs_fs_info *fs_info, struct fstrim_range *range)
5727 {
5728         struct btrfs_block_group *cache = NULL;
5729         struct btrfs_device *device;
5730         struct list_head *devices;
5731         u64 group_trimmed;
5732         u64 range_end = U64_MAX;
5733         u64 start;
5734         u64 end;
5735         u64 trimmed = 0;
5736         u64 bg_failed = 0;
5737         u64 dev_failed = 0;
5738         int bg_ret = 0;
5739         int dev_ret = 0;
5740         int ret = 0;
5741
5742         /*
5743          * Check range overflow if range->len is set.
5744          * The default range->len is U64_MAX.
5745          */
5746         if (range->len != U64_MAX &&
5747             check_add_overflow(range->start, range->len, &range_end))
5748                 return -EINVAL;
5749
5750         cache = btrfs_lookup_first_block_group(fs_info, range->start);
5751         for (; cache; cache = btrfs_next_block_group(cache)) {
5752                 if (cache->start >= range_end) {
5753                         btrfs_put_block_group(cache);
5754                         break;
5755                 }
5756
5757                 start = max(range->start, cache->start);
5758                 end = min(range_end, cache->start + cache->length);
5759
5760                 if (end - start >= range->minlen) {
5761                         if (!btrfs_block_group_done(cache)) {
5762                                 ret = btrfs_cache_block_group(cache, 0);
5763                                 if (ret) {
5764                                         bg_failed++;
5765                                         bg_ret = ret;
5766                                         continue;
5767                                 }
5768                                 ret = btrfs_wait_block_group_cache_done(cache);
5769                                 if (ret) {
5770                                         bg_failed++;
5771                                         bg_ret = ret;
5772                                         continue;
5773                                 }
5774                         }
5775                         ret = btrfs_trim_block_group(cache,
5776                                                      &group_trimmed,
5777                                                      start,
5778                                                      end,
5779                                                      range->minlen);
5780
5781                         trimmed += group_trimmed;
5782                         if (ret) {
5783                                 bg_failed++;
5784                                 bg_ret = ret;
5785                                 continue;
5786                         }
5787                 }
5788         }
5789
5790         if (bg_failed)
5791                 btrfs_warn(fs_info,
5792                         "failed to trim %llu block group(s), last error %d",
5793                         bg_failed, bg_ret);
5794         mutex_lock(&fs_info->fs_devices->device_list_mutex);
5795         devices = &fs_info->fs_devices->devices;
5796         list_for_each_entry(device, devices, dev_list) {
5797                 ret = btrfs_trim_free_extents(device, &group_trimmed);
5798                 if (ret) {
5799                         dev_failed++;
5800                         dev_ret = ret;
5801                         break;
5802                 }
5803
5804                 trimmed += group_trimmed;
5805         }
5806         mutex_unlock(&fs_info->fs_devices->device_list_mutex);
5807
5808         if (dev_failed)
5809                 btrfs_warn(fs_info,
5810                         "failed to trim %llu device(s), last error %d",
5811                         dev_failed, dev_ret);
5812         range->len = trimmed;
5813         if (bg_ret)
5814                 return bg_ret;
5815         return dev_ret;
5816 }