git.sesse.net Git - bcachefs-tools-debian/blob - libbcachefs/fsck.c

   1 // SPDX-License-Identifier: GPL-2.0
   2
   3 #include "bcachefs.h"
   4 #include "bkey_buf.h"
   5 #include "btree_update.h"
   6 #include "buckets.h"
   7 #include "darray.h"
   8 #include "dirent.h"
   9 #include "error.h"
  10 #include "fs-common.h"
  11 #include "fsck.h"
  12 #include "inode.h"
  13 #include "keylist.h"
  14 #include "subvolume.h"
  15 #include "super.h"
  16 #include "xattr.h"
  17
  18 #include <linux/bsearch.h>
  19 #include <linux/dcache.h> /* struct qstr */
  20
  21 #define QSTR(n) { { { .len = strlen(n) } }, .name = n }
  22
  23 /*
  24  * XXX: this is handling transaction restarts without returning
  25  * -BCH_ERR_transaction_restart_nested, this is not how we do things anymore:
  26  */
  27 static s64 bch2_count_inode_sectors(struct btree_trans *trans, u64 inum,
  28                                     u32 snapshot)
  29 {
  30         struct btree_iter iter;
  31         struct bkey_s_c k;
  32         u64 sectors = 0;
  33         int ret;
  34
  35         for_each_btree_key_upto(trans, iter, BTREE_ID_extents,
  36                                 SPOS(inum, 0, snapshot),
  37                                 POS(inum, U64_MAX),
  38                                 0, k, ret)
  39                 if (bkey_extent_is_allocation(k.k))
  40                         sectors += k.k->size;
  41
  42         bch2_trans_iter_exit(trans, &iter);
  43
  44         return ret ?: sectors;
  45 }
  46
  47 static s64 bch2_count_subdirs(struct btree_trans *trans, u64 inum,
  48                                     u32 snapshot)
  49 {
  50         struct btree_iter iter;
  51         struct bkey_s_c k;
  52         struct bkey_s_c_dirent d;
  53         u64 subdirs = 0;
  54         int ret;
  55
  56         for_each_btree_key_upto(trans, iter, BTREE_ID_dirents,
  57                                 SPOS(inum, 0, snapshot),
  58                                 POS(inum, U64_MAX),
  59                                 0, k, ret) {
  60                 if (k.k->type != KEY_TYPE_dirent)
  61                         continue;
  62
  63                 d = bkey_s_c_to_dirent(k);
  64                 if (d.v->d_type == DT_DIR)
  65                         subdirs++;
  66         }
  67         bch2_trans_iter_exit(trans, &iter);
  68
  69         return ret ?: subdirs;
  70 }
  71
  72 static int __snapshot_lookup_subvol(struct btree_trans *trans, u32 snapshot,
  73                                     u32 *subvol)
  74 {
  75         struct btree_iter iter;
  76         struct bkey_s_c k;
  77         int ret;
  78
  79         bch2_trans_iter_init(trans, &iter, BTREE_ID_snapshots,
  80                              POS(0, snapshot), 0);
  81         k = bch2_btree_iter_peek_slot(&iter);
  82         ret = bkey_err(k);
  83         if (ret)
  84                 goto err;
  85
  86         if (k.k->type != KEY_TYPE_snapshot) {
  87                 bch_err(trans->c, "snapshot %u not fonud", snapshot);
  88                 ret = -ENOENT;
  89                 goto err;
  90         }
  91
  92         *subvol = le32_to_cpu(bkey_s_c_to_snapshot(k).v->subvol);
  93 err:
  94         bch2_trans_iter_exit(trans, &iter);
  95         return ret;
  96
  97 }
  98
  99 static int __subvol_lookup(struct btree_trans *trans, u32 subvol,
 100                            u32 *snapshot, u64 *inum)
 101 {
 102         struct bch_subvolume s;
 103         int ret;
 104
 105         ret = bch2_subvolume_get(trans, subvol, false, 0, &s);
 106
 107         *snapshot = le32_to_cpu(s.snapshot);
 108         *inum = le64_to_cpu(s.inode);
 109         return ret;
 110 }
 111
 112 static int subvol_lookup(struct btree_trans *trans, u32 subvol,
 113                          u32 *snapshot, u64 *inum)
 114 {
 115         return lockrestart_do(trans, __subvol_lookup(trans, subvol, snapshot, inum));
 116 }
 117
 118 static int lookup_first_inode(struct btree_trans *trans, u64 inode_nr,
 119                               struct bch_inode_unpacked *inode)
 120 {
 121         struct btree_iter iter;
 122         struct bkey_s_c k;
 123         int ret;
 124
 125         bch2_trans_iter_init(trans, &iter, BTREE_ID_inodes,
 126                              POS(0, inode_nr),
 127                              BTREE_ITER_ALL_SNAPSHOTS);
 128         k = bch2_btree_iter_peek(&iter);
 129         ret = bkey_err(k);
 130         if (ret)
 131                 goto err;
 132
 133         if (!k.k || !bkey_eq(k.k->p, POS(0, inode_nr))) {
 134                 ret = -ENOENT;
 135                 goto err;
 136         }
 137
 138         ret = bch2_inode_unpack(k, inode);
 139 err:
 140         if (ret && !bch2_err_matches(ret, BCH_ERR_transaction_restart))
 141                 bch_err(trans->c, "error fetching inode %llu: %s",
 142                         inode_nr, bch2_err_str(ret));
 143         bch2_trans_iter_exit(trans, &iter);
 144         return ret;
 145 }
 146
 147 static int __lookup_inode(struct btree_trans *trans, u64 inode_nr,
 148                           struct bch_inode_unpacked *inode,
 149                           u32 *snapshot)
 150 {
 151         struct btree_iter iter;
 152         struct bkey_s_c k;
 153         int ret;
 154
 155         bch2_trans_iter_init(trans, &iter, BTREE_ID_inodes,
 156                              SPOS(0, inode_nr, *snapshot), 0);
 157         k = bch2_btree_iter_peek_slot(&iter);
 158         ret = bkey_err(k);
 159         if (ret)
 160                 goto err;
 161
 162         ret = bkey_is_inode(k.k)
 163                 ? bch2_inode_unpack(k, inode)
 164                 : -ENOENT;
 165         if (!ret)
 166                 *snapshot = iter.pos.snapshot;
 167 err:
 168         if (ret && !bch2_err_matches(ret, BCH_ERR_transaction_restart))
 169                 bch_err(trans->c, "error fetching inode %llu:%u: %s",
 170                         inode_nr, *snapshot, bch2_err_str(ret));
 171         bch2_trans_iter_exit(trans, &iter);
 172         return ret;
 173 }
 174
 175 static int lookup_inode(struct btree_trans *trans, u64 inode_nr,
 176                         struct bch_inode_unpacked *inode,
 177                         u32 *snapshot)
 178 {
 179         return lockrestart_do(trans, __lookup_inode(trans, inode_nr, inode, snapshot));
 180 }
 181
 182 static int __lookup_dirent(struct btree_trans *trans,
 183                            struct bch_hash_info hash_info,
 184                            subvol_inum dir, struct qstr *name,
 185                            u64 *target, unsigned *type)
 186 {
 187         struct btree_iter iter;
 188         struct bkey_s_c_dirent d;
 189         int ret;
 190
 191         ret = bch2_hash_lookup(trans, &iter, bch2_dirent_hash_desc,
 192                                &hash_info, dir, name, 0);
 193         if (ret)
 194                 return ret;
 195
 196         d = bkey_s_c_to_dirent(bch2_btree_iter_peek_slot(&iter));
 197         *target = le64_to_cpu(d.v->d_inum);
 198         *type = d.v->d_type;
 199         bch2_trans_iter_exit(trans, &iter);
 200         return 0;
 201 }
 202
 203 static int __write_inode(struct btree_trans *trans,
 204                          struct bch_inode_unpacked *inode,
 205                          u32 snapshot)
 206 {
 207         struct btree_iter iter;
 208         int ret;
 209
 210         bch2_trans_iter_init(trans, &iter, BTREE_ID_inodes,
 211                             SPOS(0, inode->bi_inum, snapshot),
 212                             BTREE_ITER_INTENT);
 213
 214         ret   = bch2_btree_iter_traverse(&iter) ?:
 215                 bch2_inode_write(trans, &iter, inode);
 216         bch2_trans_iter_exit(trans, &iter);
 217         return ret;
 218 }
 219
 220 static int write_inode(struct btree_trans *trans,
 221                        struct bch_inode_unpacked *inode,
 222                        u32 snapshot)
 223 {
 224         int ret = commit_do(trans, NULL, NULL,
 225                                   BTREE_INSERT_NOFAIL|
 226                                   BTREE_INSERT_LAZY_RW,
 227                                   __write_inode(trans, inode, snapshot));
 228         if (ret)
 229                 bch_err(trans->c, "error in fsck: error updating inode: %s",
 230                         bch2_err_str(ret));
 231         return ret;
 232 }
 233
 234 static int fsck_inode_rm(struct btree_trans *trans, u64 inum, u32 snapshot)
 235 {
 236         struct bch_fs *c = trans->c;
 237         struct btree_iter iter = { NULL };
 238         struct bkey_i_inode_generation delete;
 239         struct bch_inode_unpacked inode_u;
 240         struct bkey_s_c k;
 241         int ret;
 242
 243         do {
 244                 ret   = bch2_btree_delete_range_trans(trans, BTREE_ID_extents,
 245                                                       SPOS(inum, 0, snapshot),
 246                                                       SPOS(inum, U64_MAX, snapshot),
 247                                                       0, NULL) ?:
 248                         bch2_btree_delete_range_trans(trans, BTREE_ID_dirents,
 249                                                       SPOS(inum, 0, snapshot),
 250                                                       SPOS(inum, U64_MAX, snapshot),
 251                                                       0, NULL) ?:
 252                         bch2_btree_delete_range_trans(trans, BTREE_ID_xattrs,
 253                                                       SPOS(inum, 0, snapshot),
 254                                                       SPOS(inum, U64_MAX, snapshot),
 255                                                       0, NULL);
 256         } while (ret == -BCH_ERR_transaction_restart_nested);
 257         if (ret)
 258                 goto err;
 259 retry:
 260         bch2_trans_begin(trans);
 261
 262         bch2_trans_iter_init(trans, &iter, BTREE_ID_inodes,
 263                              SPOS(0, inum, snapshot), BTREE_ITER_INTENT);
 264         k = bch2_btree_iter_peek_slot(&iter);
 265
 266         ret = bkey_err(k);
 267         if (ret)
 268                 goto err;
 269
 270         if (!bkey_is_inode(k.k)) {
 271                 bch2_fs_inconsistent(c,
 272                                      "inode %llu:%u not found when deleting",
 273                                      inum, snapshot);
 274                 ret = -EIO;
 275                 goto err;
 276         }
 277
 278         bch2_inode_unpack(k, &inode_u);
 279
 280         /* Subvolume root? */
 281         if (inode_u.bi_subvol)
 282                 bch_warn(c, "deleting inode %llu marked as unlinked, but also a subvolume root!?", inode_u.bi_inum);
 283
 284         bkey_inode_generation_init(&delete.k_i);
 285         delete.k.p = iter.pos;
 286         delete.v.bi_generation = cpu_to_le32(inode_u.bi_generation + 1);
 287
 288         ret   = bch2_trans_update(trans, &iter, &delete.k_i, 0) ?:
 289                 bch2_trans_commit(trans, NULL, NULL,
 290                                 BTREE_INSERT_NOFAIL);
 291 err:
 292         bch2_trans_iter_exit(trans, &iter);
 293         if (bch2_err_matches(ret, BCH_ERR_transaction_restart))
 294                 goto retry;
 295
 296         return ret ?: -BCH_ERR_transaction_restart_nested;
 297 }
 298
 299 static int __remove_dirent(struct btree_trans *trans, struct bpos pos)
 300 {
 301         struct bch_fs *c = trans->c;
 302         struct btree_iter iter;
 303         struct bch_inode_unpacked dir_inode;
 304         struct bch_hash_info dir_hash_info;
 305         int ret;
 306
 307         ret = lookup_first_inode(trans, pos.inode, &dir_inode);
 308         if (ret)
 309                 goto err;
 310
 311         dir_hash_info = bch2_hash_info_init(c, &dir_inode);
 312
 313         bch2_trans_iter_init(trans, &iter, BTREE_ID_dirents, pos, BTREE_ITER_INTENT);
 314
 315         ret = bch2_hash_delete_at(trans, bch2_dirent_hash_desc,
 316                                   &dir_hash_info, &iter,
 317                                   BTREE_UPDATE_INTERNAL_SNAPSHOT_NODE);
 318         bch2_trans_iter_exit(trans, &iter);
 319 err:
 320         if (ret && !bch2_err_matches(ret, BCH_ERR_transaction_restart))
 321                 bch_err(c, "%s(): error %s", __func__, bch2_err_str(ret));
 322         return ret;
 323 }
 324
 325 /* Get lost+found, create if it doesn't exist: */
 326 static int lookup_lostfound(struct btree_trans *trans, u32 subvol,
 327                             struct bch_inode_unpacked *lostfound)
 328 {
 329         struct bch_fs *c = trans->c;
 330         struct bch_inode_unpacked root;
 331         struct bch_hash_info root_hash_info;
 332         struct qstr lostfound_str = QSTR("lost+found");
 333         subvol_inum root_inum = { .subvol = subvol };
 334         u64 inum = 0;
 335         unsigned d_type = 0;
 336         u32 snapshot;
 337         int ret;
 338
 339         ret = __subvol_lookup(trans, subvol, &snapshot, &root_inum.inum);
 340         if (ret)
 341                 return ret;
 342
 343         ret = __lookup_inode(trans, root_inum.inum, &root, &snapshot);
 344         if (ret)
 345                 return ret;
 346
 347         root_hash_info = bch2_hash_info_init(c, &root);
 348
 349         ret = __lookup_dirent(trans, root_hash_info, root_inum,
 350                             &lostfound_str, &inum, &d_type);
 351         if (ret == -ENOENT) {
 352                 bch_notice(c, "creating lost+found");
 353                 goto create_lostfound;
 354         }
 355
 356         if (ret && !bch2_err_matches(ret, BCH_ERR_transaction_restart))
 357                 bch_err(c, "error looking up lost+found: %s", bch2_err_str(ret));
 358         if (ret)
 359                 return ret;
 360
 361         if (d_type != DT_DIR) {
 362                 bch_err(c, "error looking up lost+found: not a directory");
 363                 return ret;
 364         }
 365
 366         /*
 367          * The check_dirents pass has already run, dangling dirents
 368          * shouldn't exist here:
 369          */
 370         return __lookup_inode(trans, inum, lostfound, &snapshot);
 371
 372 create_lostfound:
 373         bch2_inode_init_early(c, lostfound);
 374
 375         ret = bch2_create_trans(trans, root_inum, &root,
 376                                 lostfound, &lostfound_str,
 377                                 0, 0, S_IFDIR|0700, 0, NULL, NULL,
 378                                 (subvol_inum) { }, 0);
 379         if (ret && !bch2_err_matches(ret, BCH_ERR_transaction_restart))
 380                 bch_err(c, "error creating lost+found: %s", bch2_err_str(ret));
 381         return ret;
 382 }
 383
 384 static int __reattach_inode(struct btree_trans *trans,
 385                           struct bch_inode_unpacked *inode,
 386                           u32 inode_snapshot)
 387 {
 388         struct bch_hash_info dir_hash;
 389         struct bch_inode_unpacked lostfound;
 390         char name_buf[20];
 391         struct qstr name;
 392         u64 dir_offset = 0;
 393         u32 subvol;
 394         int ret;
 395
 396         ret = __snapshot_lookup_subvol(trans, inode_snapshot, &subvol);
 397         if (ret)
 398                 return ret;
 399
 400         ret = lookup_lostfound(trans, subvol, &lostfound);
 401         if (ret)
 402                 return ret;
 403
 404         if (S_ISDIR(inode->bi_mode)) {
 405                 lostfound.bi_nlink++;
 406
 407                 ret = __write_inode(trans, &lostfound, U32_MAX);
 408                 if (ret)
 409                         return ret;
 410         }
 411
 412         dir_hash = bch2_hash_info_init(trans->c, &lostfound);
 413
 414         snprintf(name_buf, sizeof(name_buf), "%llu", inode->bi_inum);
 415         name = (struct qstr) QSTR(name_buf);
 416
 417         ret = bch2_dirent_create(trans,
 418                                  (subvol_inum) {
 419                                         .subvol = subvol,
 420                                         .inum = lostfound.bi_inum,
 421                                  },
 422                                  &dir_hash,
 423                                  inode_d_type(inode),
 424                                  &name, inode->bi_inum, &dir_offset,
 425                                  BCH_HASH_SET_MUST_CREATE);
 426         if (ret)
 427                 return ret;
 428
 429         inode->bi_dir           = lostfound.bi_inum;
 430         inode->bi_dir_offset    = dir_offset;
 431
 432         return __write_inode(trans, inode, inode_snapshot);
 433 }
 434
 435 static int reattach_inode(struct btree_trans *trans,
 436                           struct bch_inode_unpacked *inode,
 437                           u32 inode_snapshot)
 438 {
 439         int ret = commit_do(trans, NULL, NULL,
 440                                   BTREE_INSERT_LAZY_RW|
 441                                   BTREE_INSERT_NOFAIL,
 442                         __reattach_inode(trans, inode, inode_snapshot));
 443         if (ret) {
 444                 bch_err(trans->c, "error reattaching inode %llu: %s",
 445                         inode->bi_inum, bch2_err_str(ret));
 446                 return ret;
 447         }
 448
 449         return ret;
 450 }
 451
 452 static int remove_backpointer(struct btree_trans *trans,
 453                               struct bch_inode_unpacked *inode)
 454 {
 455         struct btree_iter iter;
 456         struct bkey_s_c k;
 457         int ret;
 458
 459         bch2_trans_iter_init(trans, &iter, BTREE_ID_dirents,
 460                              POS(inode->bi_dir, inode->bi_dir_offset), 0);
 461         k = bch2_btree_iter_peek_slot(&iter);
 462         ret = bkey_err(k);
 463         if (ret)
 464                 goto out;
 465         if (k.k->type != KEY_TYPE_dirent) {
 466                 ret = -ENOENT;
 467                 goto out;
 468         }
 469
 470         ret = __remove_dirent(trans, k.k->p);
 471 out:
 472         bch2_trans_iter_exit(trans, &iter);
 473         return ret;
 474 }
 475
 476 struct snapshots_seen_entry {
 477         u32                             id;
 478         u32                             equiv;
 479 };
 480
 481 struct snapshots_seen {
 482         struct bpos                     pos;
 483         DARRAY(struct snapshots_seen_entry) ids;
 484 };
 485
 486 static inline void snapshots_seen_exit(struct snapshots_seen *s)
 487 {
 488         darray_exit(&s->ids);
 489 }
 490
 491 static inline void snapshots_seen_init(struct snapshots_seen *s)
 492 {
 493         memset(s, 0, sizeof(*s));
 494 }
 495
 496 static int snapshots_seen_add(struct bch_fs *c, struct snapshots_seen *s, u32 id)
 497 {
 498         struct snapshots_seen_entry *i, n = { id, id };
 499         int ret;
 500
 501         darray_for_each(s->ids, i) {
 502                 if (n.equiv < i->equiv)
 503                         break;
 504
 505                 if (i->equiv == n.equiv) {
 506                         bch_err(c, "%s(): adding duplicate snapshot", __func__);
 507                         return -EINVAL;
 508                 }
 509         }
 510
 511         ret = darray_insert_item(&s->ids, i - s->ids.data, n);
 512         if (ret)
 513                 bch_err(c, "error reallocating snapshots_seen table (size %zu)",
 514                         s->ids.size);
 515         return ret;
 516 }
 517
 518 static int snapshots_seen_update(struct bch_fs *c, struct snapshots_seen *s,
 519                                  enum btree_id btree_id, struct bpos pos)
 520 {
 521         struct snapshots_seen_entry *i, n = {
 522                 .id     = pos.snapshot,
 523                 .equiv  = bch2_snapshot_equiv(c, pos.snapshot),
 524         };
 525         int ret = 0;
 526
 527         if (!bkey_eq(s->pos, pos))
 528                 s->ids.nr = 0;
 529
 530         pos.snapshot = n.equiv;
 531         s->pos = pos;
 532
 533         darray_for_each(s->ids, i)
 534                 if (i->equiv == n.equiv) {
 535                         if (fsck_err_on(i->id != n.id, c,
 536                                         "snapshot deletion did not run correctly:\n"
 537                                         "  duplicate keys in btree %s at %llu:%llu snapshots %u, %u (equiv %u)\n",
 538                                         bch2_btree_ids[btree_id],
 539                                         pos.inode, pos.offset,
 540                                         i->id, n.id, n.equiv))
 541                                 return -BCH_ERR_need_snapshot_cleanup;
 542
 543                         return 0;
 544                 }
 545
 546         ret = darray_push(&s->ids, n);
 547         if (ret)
 548                 bch_err(c, "error reallocating snapshots_seen table (size %zu)",
 549                         s->ids.size);
 550 fsck_err:
 551         return ret;
 552 }
 553
 554 /**
 555  * key_visible_in_snapshot - returns true if @id is a descendent of @ancestor,
 556  * and @ancestor hasn't been overwritten in @seen
 557  *
 558  * That is, returns whether key in @ancestor snapshot is visible in @id snapshot
 559  */
 560 static bool key_visible_in_snapshot(struct bch_fs *c, struct snapshots_seen *seen,
 561                                     u32 id, u32 ancestor)
 562 {
 563         ssize_t i;
 564         u32 top = seen->ids.nr ? seen->ids.data[seen->ids.nr - 1].equiv : 0;
 565
 566         BUG_ON(id > ancestor);
 567         BUG_ON(!bch2_snapshot_is_equiv(c, id));
 568         BUG_ON(!bch2_snapshot_is_equiv(c, ancestor));
 569
 570         /* @ancestor should be the snapshot most recently added to @seen */
 571         BUG_ON(ancestor != seen->pos.snapshot);
 572         BUG_ON(ancestor != top);
 573
 574         if (id == ancestor)
 575                 return true;
 576
 577         if (!bch2_snapshot_is_ancestor(c, id, ancestor))
 578                 return false;
 579
 580         for (i = seen->ids.nr - 2;
 581              i >= 0 && seen->ids.data[i].equiv >= id;
 582              --i)
 583                 if (bch2_snapshot_is_ancestor(c, id, seen->ids.data[i].equiv) &&
 584                     bch2_snapshot_is_ancestor(c, seen->ids.data[i].equiv, ancestor))
 585                         return false;
 586
 587         return true;
 588 }
 589
 590 /**
 591  * ref_visible - given a key with snapshot id @src that points to a key with
 592  * snapshot id @dst, test whether there is some snapshot in which @dst is
 593  * visible.
 594  *
 595  * This assumes we're visiting @src keys in natural key order.
 596  *
 597  * @s   - list of snapshot IDs already seen at @src
 598  * @src - snapshot ID of src key
 599  * @dst - snapshot ID of dst key
 600  */
 601 static int ref_visible(struct bch_fs *c, struct snapshots_seen *s,
 602                        u32 src, u32 dst)
 603 {
 604         return dst <= src
 605                 ? key_visible_in_snapshot(c, s, dst, src)
 606                 : bch2_snapshot_is_ancestor(c, src, dst);
 607 }
 608
 609 static int ref_visible2(struct bch_fs *c,
 610                         u32 src, struct snapshots_seen *src_seen,
 611                         u32 dst, struct snapshots_seen *dst_seen)
 612 {
 613         src = bch2_snapshot_equiv(c, src);
 614         dst = bch2_snapshot_equiv(c, dst);
 615
 616         if (dst > src) {
 617                 swap(dst, src);
 618                 swap(dst_seen, src_seen);
 619         }
 620         return key_visible_in_snapshot(c, src_seen, dst, src);
 621 }
 622
 623 #define for_each_visible_inode(_c, _s, _w, _snapshot, _i)                               \
 624         for (_i = (_w)->inodes.data; _i < (_w)->inodes.data + (_w)->inodes.nr &&        \
 625              (_i)->snapshot <= (_snapshot); _i++)                                       \
 626                 if (key_visible_in_snapshot(_c, _s, _i->snapshot, _snapshot))
 627
 628 struct inode_walker_entry {
 629         struct bch_inode_unpacked inode;
 630         u32                     snapshot;
 631         u64                     count;
 632 };
 633
 634 struct inode_walker {
 635         bool                            first_this_inode;
 636         u64                             cur_inum;
 637
 638         DARRAY(struct inode_walker_entry) inodes;
 639 };
 640
 641 static void inode_walker_exit(struct inode_walker *w)
 642 {
 643         darray_exit(&w->inodes);
 644 }
 645
 646 static struct inode_walker inode_walker_init(void)
 647 {
 648         return (struct inode_walker) { 0, };
 649 }
 650
 651 static int add_inode(struct bch_fs *c, struct inode_walker *w,
 652                      struct bkey_s_c inode)
 653 {
 654         struct bch_inode_unpacked u;
 655
 656         BUG_ON(bch2_inode_unpack(inode, &u));
 657
 658         return darray_push(&w->inodes, ((struct inode_walker_entry) {
 659                 .inode          = u,
 660                 .snapshot       = bch2_snapshot_equiv(c, inode.k->p.snapshot),
 661         }));
 662 }
 663
 664 static int __walk_inode(struct btree_trans *trans,
 665                         struct inode_walker *w, struct bpos pos)
 666 {
 667         struct bch_fs *c = trans->c;
 668         struct btree_iter iter;
 669         struct bkey_s_c k;
 670         u32 restart_count = trans->restart_count;
 671         unsigned i;
 672         int ret;
 673
 674         pos.snapshot = bch2_snapshot_equiv(c, pos.snapshot);
 675
 676         if (pos.inode == w->cur_inum) {
 677                 w->first_this_inode = false;
 678                 goto lookup_snapshot;
 679         }
 680
 681         w->inodes.nr = 0;
 682
 683         for_each_btree_key(trans, iter, BTREE_ID_inodes, POS(0, pos.inode),
 684                            BTREE_ITER_ALL_SNAPSHOTS, k, ret) {
 685                 if (k.k->p.offset != pos.inode)
 686                         break;
 687
 688                 if (bkey_is_inode(k.k))
 689                         add_inode(c, w, k);
 690         }
 691         bch2_trans_iter_exit(trans, &iter);
 692
 693         if (ret)
 694                 return ret;
 695
 696         w->cur_inum             = pos.inode;
 697         w->first_this_inode     = true;
 698
 699         if (trans_was_restarted(trans, restart_count))
 700                 return -BCH_ERR_transaction_restart_nested;
 701
 702 lookup_snapshot:
 703         for (i = 0; i < w->inodes.nr; i++)
 704                 if (bch2_snapshot_is_ancestor(c, pos.snapshot, w->inodes.data[i].snapshot))
 705                         goto found;
 706         return INT_MAX;
 707 found:
 708         BUG_ON(pos.snapshot > w->inodes.data[i].snapshot);
 709
 710         if (pos.snapshot != w->inodes.data[i].snapshot) {
 711                 struct inode_walker_entry e = w->inodes.data[i];
 712
 713                 e.snapshot = pos.snapshot;
 714                 e.count = 0;
 715
 716                 bch_info(c, "have key for inode %llu:%u but have inode in ancestor snapshot %u",
 717                          pos.inode, pos.snapshot, w->inodes.data[i].snapshot);
 718
 719                 while (i && w->inodes.data[i - 1].snapshot > pos.snapshot)
 720                         --i;
 721
 722                 ret = darray_insert_item(&w->inodes, i, e);
 723                 if (ret)
 724                         return ret;
 725         }
 726
 727         return i;
 728 }
 729
 730 static int __get_visible_inodes(struct btree_trans *trans,
 731                                 struct inode_walker *w,
 732                                 struct snapshots_seen *s,
 733                                 u64 inum)
 734 {
 735         struct bch_fs *c = trans->c;
 736         struct btree_iter iter;
 737         struct bkey_s_c k;
 738         int ret;
 739
 740         w->inodes.nr = 0;
 741
 742         for_each_btree_key_norestart(trans, iter, BTREE_ID_inodes, POS(0, inum),
 743                            BTREE_ITER_ALL_SNAPSHOTS, k, ret) {
 744                 u32 equiv = bch2_snapshot_equiv(c, k.k->p.snapshot);
 745
 746                 if (k.k->p.offset != inum)
 747                         break;
 748
 749                 if (!ref_visible(c, s, s->pos.snapshot, equiv))
 750                         continue;
 751
 752                 if (bkey_is_inode(k.k))
 753                         add_inode(c, w, k);
 754
 755                 if (equiv >= s->pos.snapshot)
 756                         break;
 757         }
 758         bch2_trans_iter_exit(trans, &iter);
 759
 760         return ret;
 761 }
 762
 763 static int check_key_has_snapshot(struct btree_trans *trans,
 764                                   struct btree_iter *iter,
 765                                   struct bkey_s_c k)
 766 {
 767         struct bch_fs *c = trans->c;
 768         struct printbuf buf = PRINTBUF;
 769         int ret = 0;
 770
 771         if (mustfix_fsck_err_on(!bch2_snapshot_equiv(c, k.k->p.snapshot), c,
 772                         "key in missing snapshot: %s",
 773                         (bch2_bkey_val_to_text(&buf, c, k), buf.buf)))
 774                 ret = bch2_btree_delete_at(trans, iter,
 775                                             BTREE_UPDATE_INTERNAL_SNAPSHOT_NODE) ?: 1;
 776 fsck_err:
 777         printbuf_exit(&buf);
 778         return ret;
 779 }
 780
 781 static int hash_redo_key(struct btree_trans *trans,
 782                          const struct bch_hash_desc desc,
 783                          struct bch_hash_info *hash_info,
 784                          struct btree_iter *k_iter, struct bkey_s_c k)
 785 {
 786         struct bkey_i *delete;
 787         struct bkey_i *tmp;
 788
 789         delete = bch2_trans_kmalloc(trans, sizeof(*delete));
 790         if (IS_ERR(delete))
 791                 return PTR_ERR(delete);
 792
 793         tmp = bch2_bkey_make_mut(trans, k);
 794         if (IS_ERR(tmp))
 795                 return PTR_ERR(tmp);
 796
 797         bkey_init(&delete->k);
 798         delete->k.p = k_iter->pos;
 799         return  bch2_btree_iter_traverse(k_iter) ?:
 800                 bch2_trans_update(trans, k_iter, delete, 0) ?:
 801                 bch2_hash_set_snapshot(trans, desc, hash_info,
 802                                        (subvol_inum) { 0, k.k->p.inode },
 803                                        k.k->p.snapshot, tmp,
 804                                        BCH_HASH_SET_MUST_CREATE,
 805                                        BTREE_UPDATE_INTERNAL_SNAPSHOT_NODE) ?:
 806                 bch2_trans_commit(trans, NULL, NULL,
 807                                   BTREE_INSERT_NOFAIL|
 808                                   BTREE_INSERT_LAZY_RW);
 809 }
 810
 811 static int hash_check_key(struct btree_trans *trans,
 812                           const struct bch_hash_desc desc,
 813                           struct bch_hash_info *hash_info,
 814                           struct btree_iter *k_iter, struct bkey_s_c hash_k)
 815 {
 816         struct bch_fs *c = trans->c;
 817         struct btree_iter iter = { NULL };
 818         struct printbuf buf = PRINTBUF;
 819         struct bkey_s_c k;
 820         u64 hash;
 821         int ret = 0;
 822
 823         if (hash_k.k->type != desc.key_type)
 824                 return 0;
 825
 826         hash = desc.hash_bkey(hash_info, hash_k);
 827
 828         if (likely(hash == hash_k.k->p.offset))
 829                 return 0;
 830
 831         if (hash_k.k->p.offset < hash)
 832                 goto bad_hash;
 833
 834         for_each_btree_key_norestart(trans, iter, desc.btree_id,
 835                                      SPOS(hash_k.k->p.inode, hash, hash_k.k->p.snapshot),
 836                                      BTREE_ITER_SLOTS, k, ret) {
 837                 if (bkey_eq(k.k->p, hash_k.k->p))
 838                         break;
 839
 840                 if (fsck_err_on(k.k->type == desc.key_type &&
 841                                 !desc.cmp_bkey(k, hash_k), c,
 842                                 "duplicate hash table keys:\n%s",
 843                                 (printbuf_reset(&buf),
 844                                  bch2_bkey_val_to_text(&buf, c, hash_k),
 845                                  buf.buf))) {
 846                         ret = bch2_hash_delete_at(trans, desc, hash_info, k_iter, 0) ?: 1;
 847                         break;
 848                 }
 849
 850                 if (bkey_deleted(k.k)) {
 851                         bch2_trans_iter_exit(trans, &iter);
 852                         goto bad_hash;
 853                 }
 854         }
 855 out:
 856         bch2_trans_iter_exit(trans, &iter);
 857         printbuf_exit(&buf);
 858         return ret;
 859 bad_hash:
 860         if (fsck_err(c, "hash table key at wrong offset: btree %s inode %llu offset %llu, hashed to %llu\n%s",
 861                      bch2_btree_ids[desc.btree_id], hash_k.k->p.inode, hash_k.k->p.offset, hash,
 862                      (printbuf_reset(&buf),
 863                       bch2_bkey_val_to_text(&buf, c, hash_k), buf.buf))) {
 864                 ret = hash_redo_key(trans, desc, hash_info, k_iter, hash_k);
 865                 if (ret) {
 866                         bch_err(c, "hash_redo_key err %s", bch2_err_str(ret));
 867                         return ret;
 868                 }
 869                 ret = -BCH_ERR_transaction_restart_nested;
 870         }
 871 fsck_err:
 872         goto out;
 873 }
 874
 875 static int check_inode(struct btree_trans *trans,
 876                        struct btree_iter *iter,
 877                        struct bkey_s_c k,
 878                        struct bch_inode_unpacked *prev,
 879                        struct snapshots_seen *s,
 880                        bool full)
 881 {
 882         struct bch_fs *c = trans->c;
 883         struct bch_inode_unpacked u;
 884         bool do_update = false;
 885         int ret;
 886
 887         ret = check_key_has_snapshot(trans, iter, k);
 888         if (ret < 0)
 889                 goto err;
 890         if (ret)
 891                 return 0;
 892
 893         ret = snapshots_seen_update(c, s, iter->btree_id, k.k->p);
 894         if (ret)
 895                 goto err;
 896
 897         /*
 898          * if snapshot id isn't a leaf node, skip it - deletion in
 899          * particular is not atomic, so on the internal snapshot nodes
 900          * we can see inodes marked for deletion after a clean shutdown
 901          */
 902         if (bch2_snapshot_internal_node(c, k.k->p.snapshot))
 903                 return 0;
 904
 905         if (!bkey_is_inode(k.k))
 906                 return 0;
 907
 908         BUG_ON(bch2_inode_unpack(k, &u));
 909
 910         if (!full &&
 911             !(u.bi_flags & (BCH_INODE_I_SIZE_DIRTY|
 912                             BCH_INODE_I_SECTORS_DIRTY|
 913                             BCH_INODE_UNLINKED)))
 914                 return 0;
 915
 916         if (prev->bi_inum != u.bi_inum)
 917                 *prev = u;
 918
 919         if (fsck_err_on(prev->bi_hash_seed      != u.bi_hash_seed ||
 920                         inode_d_type(prev)      != inode_d_type(&u), c,
 921                         "inodes in different snapshots don't match")) {
 922                 bch_err(c, "repair not implemented yet");
 923                 return -EINVAL;
 924         }
 925
 926         if (u.bi_flags & BCH_INODE_UNLINKED &&
 927             (!c->sb.clean ||
 928              fsck_err(c, "filesystem marked clean, but inode %llu unlinked",
 929                       u.bi_inum))) {
 930                 bch2_trans_unlock(trans);
 931                 bch2_fs_lazy_rw(c);
 932
 933                 ret = fsck_inode_rm(trans, u.bi_inum, iter->pos.snapshot);
 934                 if (ret && !bch2_err_matches(ret, BCH_ERR_transaction_restart))
 935                         bch_err(c, "error in fsck: error while deleting inode: %s",
 936                                 bch2_err_str(ret));
 937                 return ret;
 938         }
 939
 940         if (u.bi_flags & BCH_INODE_I_SIZE_DIRTY &&
 941             (!c->sb.clean ||
 942              fsck_err(c, "filesystem marked clean, but inode %llu has i_size dirty",
 943                       u.bi_inum))) {
 944                 bch_verbose(c, "truncating inode %llu", u.bi_inum);
 945
 946                 bch2_trans_unlock(trans);
 947                 bch2_fs_lazy_rw(c);
 948
 949                 /*
 950                  * XXX: need to truncate partial blocks too here - or ideally
 951                  * just switch units to bytes and that issue goes away
 952                  */
 953                 ret = bch2_btree_delete_range_trans(trans, BTREE_ID_extents,
 954                                 SPOS(u.bi_inum, round_up(u.bi_size, block_bytes(c)) >> 9,
 955                                      iter->pos.snapshot),
 956                                 POS(u.bi_inum, U64_MAX),
 957                                 0, NULL);
 958                 if (ret && !bch2_err_matches(ret, BCH_ERR_transaction_restart))
 959                         bch_err(c, "error in fsck: error truncating inode: %s",
 960                                 bch2_err_str(ret));
 961                 if (ret)
 962                         return ret;
 963
 964                 /*
 965                  * We truncated without our normal sector accounting hook, just
 966                  * make sure we recalculate it:
 967                  */
 968                 u.bi_flags |= BCH_INODE_I_SECTORS_DIRTY;
 969
 970                 u.bi_flags &= ~BCH_INODE_I_SIZE_DIRTY;
 971                 do_update = true;
 972         }
 973
 974         if (u.bi_flags & BCH_INODE_I_SECTORS_DIRTY &&
 975             (!c->sb.clean ||
 976              fsck_err(c, "filesystem marked clean, but inode %llu has i_sectors dirty",
 977                       u.bi_inum))) {
 978                 s64 sectors;
 979
 980                 bch_verbose(c, "recounting sectors for inode %llu",
 981                             u.bi_inum);
 982
 983                 sectors = bch2_count_inode_sectors(trans, u.bi_inum, iter->pos.snapshot);
 984                 if (sectors < 0) {
 985                         bch_err(c, "error in fsck: error recounting inode sectors: %s",
 986                                 bch2_err_str(sectors));
 987                         return sectors;
 988                 }
 989
 990                 u.bi_sectors = sectors;
 991                 u.bi_flags &= ~BCH_INODE_I_SECTORS_DIRTY;
 992                 do_update = true;
 993         }
 994
 995         if (u.bi_flags & BCH_INODE_BACKPTR_UNTRUSTED) {
 996                 u.bi_dir = 0;
 997                 u.bi_dir_offset = 0;
 998                 u.bi_flags &= ~BCH_INODE_BACKPTR_UNTRUSTED;
 999                 do_update = true;
1000         }
1001
1002         if (do_update) {
1003                 ret = __write_inode(trans, &u, iter->pos.snapshot);
1004                 if (ret)
1005                         bch_err(c, "error in fsck: error updating inode: %s",
1006                                 bch2_err_str(ret));
1007         }
1008 err:
1009 fsck_err:
1010         if (ret)
1011                 bch_err(c, "%s(): error %s", __func__, bch2_err_str(ret));
1012         return ret;
1013 }
1014
1015 noinline_for_stack
1016 static int check_inodes(struct bch_fs *c, bool full)
1017 {
1018         struct btree_trans trans;
1019         struct btree_iter iter;
1020         struct bch_inode_unpacked prev = { 0 };
1021         struct snapshots_seen s;
1022         struct bkey_s_c k;
1023         int ret;
1024
1025         snapshots_seen_init(&s);
1026         bch2_trans_init(&trans, c, BTREE_ITER_MAX, 0);
1027
1028         ret = for_each_btree_key_commit(&trans, iter, BTREE_ID_inodes,
1029                         POS_MIN,
1030                         BTREE_ITER_PREFETCH|BTREE_ITER_ALL_SNAPSHOTS, k,
1031                         NULL, NULL, BTREE_INSERT_LAZY_RW|BTREE_INSERT_NOFAIL,
1032                 check_inode(&trans, &iter, k, &prev, &s, full));
1033
1034         bch2_trans_exit(&trans);
1035         snapshots_seen_exit(&s);
1036         if (ret)
1037                 bch_err(c, "%s(): error %s", __func__, bch2_err_str(ret));
1038         return ret;
1039 }
1040
1041 /*
1042  * Checking for overlapping extents needs to be reimplemented
1043  */
1044 #if 0
1045 static int fix_overlapping_extent(struct btree_trans *trans,
1046                                        struct bkey_s_c k, struct bpos cut_at)
1047 {
1048         struct btree_iter iter;
1049         struct bkey_i *u;
1050         int ret;
1051
1052         u = bch2_trans_kmalloc(trans, bkey_bytes(k.k));
1053         ret = PTR_ERR_OR_ZERO(u);
1054         if (ret)
1055                 return ret;
1056
1057         bkey_reassemble(u, k);
1058         bch2_cut_front(cut_at, u);
1059
1060
1061         /*
1062          * We don't want to go through the extent_handle_overwrites path:
1063          *
1064          * XXX: this is going to screw up disk accounting, extent triggers
1065          * assume things about extent overwrites - we should be running the
1066          * triggers manually here
1067          */
1068         bch2_trans_iter_init(trans, &iter, BTREE_ID_extents, u->k.p,
1069                              BTREE_ITER_INTENT|BTREE_ITER_NOT_EXTENTS);
1070
1071         BUG_ON(iter.flags & BTREE_ITER_IS_EXTENTS);
1072         ret   = bch2_btree_iter_traverse(&iter) ?:
1073                 bch2_trans_update(trans, &iter, u, BTREE_TRIGGER_NORUN) ?:
1074                 bch2_trans_commit(trans, NULL, NULL,
1075                                   BTREE_INSERT_NOFAIL|
1076                                   BTREE_INSERT_LAZY_RW);
1077         bch2_trans_iter_exit(trans, &iter);
1078         return ret;
1079 }
1080 #endif
1081
1082 static struct bkey_s_c_dirent dirent_get_by_pos(struct btree_trans *trans,
1083                                                 struct btree_iter *iter,
1084                                                 struct bpos pos)
1085 {
1086         struct bkey_s_c k;
1087         int ret;
1088
1089         bch2_trans_iter_init(trans, iter, BTREE_ID_dirents, pos, 0);
1090         k = bch2_btree_iter_peek_slot(iter);
1091         ret = bkey_err(k);
1092         if (!ret && k.k->type != KEY_TYPE_dirent)
1093                 ret = -ENOENT;
1094         if (ret) {
1095                 bch2_trans_iter_exit(trans, iter);
1096                 return (struct bkey_s_c_dirent) { .k = ERR_PTR(ret) };
1097         }
1098
1099         return bkey_s_c_to_dirent(k);
1100 }
1101
1102 static bool inode_points_to_dirent(struct bch_inode_unpacked *inode,
1103                                    struct bkey_s_c_dirent d)
1104 {
1105         return  inode->bi_dir           == d.k->p.inode &&
1106                 inode->bi_dir_offset    == d.k->p.offset;
1107 }
1108
1109 static bool dirent_points_to_inode(struct bkey_s_c_dirent d,
1110                                    struct bch_inode_unpacked *inode)
1111 {
1112         return d.v->d_type == DT_SUBVOL
1113                 ? le32_to_cpu(d.v->d_child_subvol)      == inode->bi_subvol
1114                 : le64_to_cpu(d.v->d_inum)              == inode->bi_inum;
1115 }
1116
1117 static int inode_backpointer_exists(struct btree_trans *trans,
1118                                     struct bch_inode_unpacked *inode,
1119                                     u32 snapshot)
1120 {
1121         struct btree_iter iter;
1122         struct bkey_s_c_dirent d;
1123         int ret;
1124
1125         d = dirent_get_by_pos(trans, &iter,
1126                         SPOS(inode->bi_dir, inode->bi_dir_offset, snapshot));
1127         ret = bkey_err(d.s_c);
1128         if (ret)
1129                 return ret == -ENOENT ? 0 : ret;
1130
1131         ret = dirent_points_to_inode(d, inode);
1132         bch2_trans_iter_exit(trans, &iter);
1133         return ret;
1134 }
1135
1136 static int check_i_sectors(struct btree_trans *trans, struct inode_walker *w)
1137 {
1138         struct bch_fs *c = trans->c;
1139         struct inode_walker_entry *i;
1140         u32 restart_count = trans->restart_count;
1141         int ret = 0;
1142         s64 count2;
1143
1144         darray_for_each(w->inodes, i) {
1145                 if (i->inode.bi_sectors == i->count)
1146                         continue;
1147
1148                 count2 = bch2_count_inode_sectors(trans, w->cur_inum, i->snapshot);
1149
1150                 if (i->count != count2) {
1151                         bch_err(c, "fsck counted i_sectors wrong: got %llu should be %llu",
1152                                 i->count, count2);
1153                         i->count = count2;
1154                         if (i->inode.bi_sectors == i->count)
1155                                 continue;
1156                 }
1157
1158                 if (fsck_err_on(!(i->inode.bi_flags & BCH_INODE_I_SECTORS_DIRTY), c,
1159                             "inode %llu:%u has incorrect i_sectors: got %llu, should be %llu",
1160                             w->cur_inum, i->snapshot,
1161                             i->inode.bi_sectors, i->count)) {
1162                         i->inode.bi_sectors = i->count;
1163                         ret = write_inode(trans, &i->inode, i->snapshot);
1164                         if (ret)
1165                                 break;
1166                 }
1167         }
1168 fsck_err:
1169         if (ret)
1170                 bch_err(c, "%s(): error %s", __func__, bch2_err_str(ret));
1171         if (!ret && trans_was_restarted(trans, restart_count))
1172                 ret = -BCH_ERR_transaction_restart_nested;
1173         return ret;
1174 }
1175
1176 struct extent_end {
1177         u32                     snapshot;
1178         u64                     offset;
1179         struct snapshots_seen   seen;
1180 };
1181
1182 typedef DARRAY(struct extent_end) extent_ends;
1183
1184 static int check_overlapping_extents(struct btree_trans *trans,
1185                               struct snapshots_seen *seen,
1186                               extent_ends *extent_ends,
1187                               struct bkey_s_c k,
1188                               struct btree_iter *iter)
1189 {
1190         struct bch_fs *c = trans->c;
1191         struct extent_end *i;
1192         struct printbuf buf = PRINTBUF;
1193         int ret = 0;
1194
1195         darray_for_each(*extent_ends, i) {
1196                 /* duplicate, due to transaction restart: */
1197                 if (i->offset   == k.k->p.offset &&
1198                     i->snapshot == k.k->p.snapshot)
1199                         continue;
1200
1201                 if (!ref_visible2(c,
1202                                   k.k->p.snapshot, seen,
1203                                   i->snapshot, &i->seen))
1204                         continue;
1205
1206                 if (fsck_err_on(i->offset > bkey_start_offset(k.k), c,
1207                                 "overlapping extents: extent in snapshot %u ends at %llu overlaps with\n%s",
1208                                 i->snapshot,
1209                                 i->offset,
1210                                 (printbuf_reset(&buf),
1211                                  bch2_bkey_val_to_text(&buf, c, k), buf.buf))) {
1212                         struct bkey_i *update = bch2_trans_kmalloc(trans, bkey_bytes(k.k));
1213                         if ((ret = PTR_ERR_OR_ZERO(update)))
1214                                 goto err;
1215                         bkey_reassemble(update, k);
1216                         ret = bch2_trans_update_extent(trans, iter, update, 0);
1217                         if (!ret)
1218                                 goto err;
1219                 }
1220         }
1221 err:
1222 fsck_err:
1223         printbuf_exit(&buf);
1224         return ret;
1225 }
1226
1227 static int extent_ends_at(extent_ends *extent_ends,
1228                           struct snapshots_seen *seen,
1229                           struct bkey_s_c k)
1230 {
1231         struct extent_end *i, n = (struct extent_end) {
1232                 .snapshot       = k.k->p.snapshot,
1233                 .offset         = k.k->p.offset,
1234                 .seen           = *seen,
1235         };
1236
1237         n.seen.ids.data = kmemdup(seen->ids.data,
1238                               sizeof(seen->ids.data[0]) * seen->ids.size,
1239                               GFP_KERNEL);
1240         if (!n.seen.ids.data)
1241                 return -BCH_ERR_ENOMEM_fsck_extent_ends_at;
1242
1243         darray_for_each(*extent_ends, i) {
1244                 if (i->snapshot == k.k->p.snapshot) {
1245                         snapshots_seen_exit(&i->seen);
1246                         *i = n;
1247                         return 0;
1248                 }
1249
1250                 if (i->snapshot >= k.k->p.snapshot)
1251                         break;
1252         }
1253
1254         return darray_insert_item(extent_ends, i - extent_ends->data, n);
1255 }
1256
1257 static void extent_ends_reset(extent_ends *extent_ends)
1258 {
1259         struct extent_end *i;
1260
1261         darray_for_each(*extent_ends, i)
1262                 snapshots_seen_exit(&i->seen);
1263
1264         extent_ends->nr = 0;
1265 }
1266
1267 static int check_extent(struct btree_trans *trans, struct btree_iter *iter,
1268                         struct bkey_s_c k,
1269                         struct inode_walker *inode,
1270                         struct snapshots_seen *s,
1271                         extent_ends *extent_ends)
1272 {
1273         struct bch_fs *c = trans->c;
1274         struct inode_walker_entry *i;
1275         struct printbuf buf = PRINTBUF;
1276         struct bpos equiv;
1277         int ret = 0;
1278
1279         ret = check_key_has_snapshot(trans, iter, k);
1280         if (ret) {
1281                 ret = ret < 0 ? ret : 0;
1282                 goto out;
1283         }
1284
1285         equiv = k.k->p;
1286         equiv.snapshot = bch2_snapshot_equiv(c, k.k->p.snapshot);
1287
1288         ret = snapshots_seen_update(c, s, iter->btree_id, k.k->p);
1289         if (ret)
1290                 goto err;
1291
1292         if (k.k->type == KEY_TYPE_whiteout)
1293                 goto out;
1294
1295         if (inode->cur_inum != k.k->p.inode) {
1296                 ret = check_i_sectors(trans, inode);
1297                 if (ret)
1298                         goto err;
1299
1300                 extent_ends_reset(extent_ends);
1301         }
1302
1303         BUG_ON(!iter->path->should_be_locked);
1304
1305         ret = check_overlapping_extents(trans, s, extent_ends, k, iter);
1306         if (ret)
1307                 goto err;
1308
1309         ret = extent_ends_at(extent_ends, s, k);
1310         if (ret)
1311                 goto err;
1312
1313         ret = __walk_inode(trans, inode, equiv);
1314         if (ret < 0)
1315                 goto err;
1316
1317         if (fsck_err_on(ret == INT_MAX, c,
1318                         "extent in missing inode:\n  %s",
1319                         (printbuf_reset(&buf),
1320                          bch2_bkey_val_to_text(&buf, c, k), buf.buf))) {
1321                 ret = bch2_btree_delete_at(trans, iter,
1322                                             BTREE_UPDATE_INTERNAL_SNAPSHOT_NODE);
1323                 goto out;
1324         }
1325
1326         if (ret == INT_MAX) {
1327                 ret = 0;
1328                 goto out;
1329         }
1330
1331         i = inode->inodes.data + ret;
1332         ret = 0;
1333
1334         if (fsck_err_on(!S_ISREG(i->inode.bi_mode) &&
1335                         !S_ISLNK(i->inode.bi_mode), c,
1336                         "extent in non regular inode mode %o:\n  %s",
1337                         i->inode.bi_mode,
1338                         (printbuf_reset(&buf),
1339                          bch2_bkey_val_to_text(&buf, c, k), buf.buf))) {
1340                 ret = bch2_btree_delete_at(trans, iter,
1341                                             BTREE_UPDATE_INTERNAL_SNAPSHOT_NODE);
1342                 goto out;
1343         }
1344
1345         /*
1346          * Check inodes in reverse order, from oldest snapshots to newest, so
1347          * that we emit the fewest number of whiteouts necessary:
1348          */
1349         for (i = inode->inodes.data + inode->inodes.nr - 1;
1350              i >= inode->inodes.data;
1351              --i) {
1352                 if (i->snapshot > equiv.snapshot ||
1353                     !key_visible_in_snapshot(c, s, i->snapshot, equiv.snapshot))
1354                         continue;
1355
1356                 if (fsck_err_on(!(i->inode.bi_flags & BCH_INODE_I_SIZE_DIRTY) &&
1357                                 k.k->p.offset > round_up(i->inode.bi_size, block_bytes(c)) >> 9 &&
1358                                 !bkey_extent_is_reservation(k), c,
1359                                 "extent type past end of inode %llu:%u, i_size %llu\n  %s",
1360                                 i->inode.bi_inum, i->snapshot, i->inode.bi_size,
1361                                 (bch2_bkey_val_to_text(&buf, c, k), buf.buf))) {
1362                         struct btree_iter iter2;
1363
1364                         bch2_trans_copy_iter(&iter2, iter);
1365                         bch2_btree_iter_set_snapshot(&iter2, i->snapshot);
1366                         ret =   bch2_btree_iter_traverse(&iter2) ?:
1367                                 bch2_btree_delete_at(trans, &iter2,
1368                                         BTREE_UPDATE_INTERNAL_SNAPSHOT_NODE);
1369                         bch2_trans_iter_exit(trans, &iter2);
1370                         if (ret)
1371                                 goto err;
1372
1373                         if (i->snapshot != equiv.snapshot) {
1374                                 ret = snapshots_seen_add(c, s, i->snapshot);
1375                                 if (ret)
1376                                         goto err;
1377                         }
1378                 }
1379         }
1380
1381         if (bkey_extent_is_allocation(k.k))
1382                 for_each_visible_inode(c, s, inode, equiv.snapshot, i)
1383                         i->count += k.k->size;
1384 #if 0
1385         bch2_bkey_buf_reassemble(&prev, c, k);
1386 #endif
1387
1388 out:
1389 err:
1390 fsck_err:
1391         printbuf_exit(&buf);
1392
1393         if (ret && !bch2_err_matches(ret, BCH_ERR_transaction_restart))
1394                 bch_err(c, "%s(): error %s", __func__, bch2_err_str(ret));
1395         return ret;
1396 }
1397
1398 /*
1399  * Walk extents: verify that extents have a corresponding S_ISREG inode, and
1400  * that i_size an i_sectors are consistent
1401  */
1402 noinline_for_stack
1403 static int check_extents(struct bch_fs *c)
1404 {
1405         struct inode_walker w = inode_walker_init();
1406         struct snapshots_seen s;
1407         struct btree_trans trans;
1408         struct btree_iter iter;
1409         struct bkey_s_c k;
1410         extent_ends extent_ends = { 0 };
1411         struct disk_reservation res = { 0 };
1412         int ret = 0;
1413
1414         snapshots_seen_init(&s);
1415         bch2_trans_init(&trans, c, BTREE_ITER_MAX, 0);
1416
1417         bch_verbose(c, "checking extents");
1418
1419         ret = for_each_btree_key_commit(&trans, iter, BTREE_ID_extents,
1420                         POS(BCACHEFS_ROOT_INO, 0),
1421                         BTREE_ITER_PREFETCH|BTREE_ITER_ALL_SNAPSHOTS, k,
1422                         &res, NULL,
1423                         BTREE_INSERT_LAZY_RW|BTREE_INSERT_NOFAIL, ({
1424                 bch2_disk_reservation_put(c, &res);
1425                 check_extent(&trans, &iter, k, &w, &s, &extent_ends);
1426         }));
1427
1428         bch2_disk_reservation_put(c, &res);
1429         extent_ends_reset(&extent_ends);
1430         darray_exit(&extent_ends);
1431         inode_walker_exit(&w);
1432         bch2_trans_exit(&trans);
1433         snapshots_seen_exit(&s);
1434
1435         if (ret)
1436                 bch_err(c, "%s(): error %s", __func__, bch2_err_str(ret));
1437         return ret;
1438 }
1439
1440 static int check_subdir_count(struct btree_trans *trans, struct inode_walker *w)
1441 {
1442         struct bch_fs *c = trans->c;
1443         struct inode_walker_entry *i;
1444         u32 restart_count = trans->restart_count;
1445         int ret = 0;
1446         s64 count2;
1447
1448         darray_for_each(w->inodes, i) {
1449                 if (i->inode.bi_nlink == i->count)
1450                         continue;
1451
1452                 count2 = bch2_count_subdirs(trans, w->cur_inum, i->snapshot);
1453                 if (count2 < 0)
1454                         return count2;
1455
1456                 if (i->count != count2) {
1457                         bch_err(c, "fsck counted subdirectories wrong: got %llu should be %llu",
1458                                 i->count, count2);
1459                         i->count = count2;
1460                         if (i->inode.bi_nlink == i->count)
1461                                 continue;
1462                 }
1463
1464                 if (fsck_err_on(i->inode.bi_nlink != i->count, c,
1465                                 "directory %llu:%u with wrong i_nlink: got %u, should be %llu",
1466                                 w->cur_inum, i->snapshot, i->inode.bi_nlink, i->count)) {
1467                         i->inode.bi_nlink = i->count;
1468                         ret = write_inode(trans, &i->inode, i->snapshot);
1469                         if (ret)
1470                                 break;
1471                 }
1472         }
1473 fsck_err:
1474         if (ret)
1475                 bch_err(c, "%s(): error %s", __func__, bch2_err_str(ret));
1476         if (!ret && trans_was_restarted(trans, restart_count))
1477                 ret = -BCH_ERR_transaction_restart_nested;
1478         return ret;
1479 }
1480
1481 static int check_dirent_target(struct btree_trans *trans,
1482                                struct btree_iter *iter,
1483                                struct bkey_s_c_dirent d,
1484                                struct bch_inode_unpacked *target,
1485                                u32 target_snapshot)
1486 {
1487         struct bch_fs *c = trans->c;
1488         struct bkey_i_dirent *n;
1489         bool backpointer_exists = true;
1490         struct printbuf buf = PRINTBUF;
1491         int ret = 0;
1492
1493         if (!target->bi_dir &&
1494             !target->bi_dir_offset) {
1495                 target->bi_dir          = d.k->p.inode;
1496                 target->bi_dir_offset   = d.k->p.offset;
1497
1498                 ret = __write_inode(trans, target, target_snapshot);
1499                 if (ret)
1500                         goto err;
1501         }
1502
1503         if (!inode_points_to_dirent(target, d)) {
1504                 ret = inode_backpointer_exists(trans, target, d.k->p.snapshot);
1505                 if (ret < 0)
1506                         goto err;
1507
1508                 backpointer_exists = ret;
1509                 ret = 0;
1510
1511                 if (fsck_err_on(S_ISDIR(target->bi_mode) &&
1512                                 backpointer_exists, c,
1513                                 "directory %llu with multiple links",
1514                                 target->bi_inum)) {
1515                         ret = __remove_dirent(trans, d.k->p);
1516                         goto out;
1517                 }
1518
1519                 if (fsck_err_on(backpointer_exists &&
1520                                 !target->bi_nlink, c,
1521                                 "inode %llu type %s has multiple links but i_nlink 0",
1522                                 target->bi_inum, bch2_d_types[d.v->d_type])) {
1523                         target->bi_nlink++;
1524                         target->bi_flags &= ~BCH_INODE_UNLINKED;
1525
1526                         ret = __write_inode(trans, target, target_snapshot);
1527                         if (ret)
1528                                 goto err;
1529                 }
1530
1531                 if (fsck_err_on(!backpointer_exists, c,
1532                                 "inode %llu:%u has wrong backpointer:\n"
1533                                 "got       %llu:%llu\n"
1534                                 "should be %llu:%llu",
1535                                 target->bi_inum, target_snapshot,
1536                                 target->bi_dir,
1537                                 target->bi_dir_offset,
1538                                 d.k->p.inode,
1539                                 d.k->p.offset)) {
1540                         target->bi_dir          = d.k->p.inode;
1541                         target->bi_dir_offset   = d.k->p.offset;
1542
1543                         ret = __write_inode(trans, target, target_snapshot);
1544                         if (ret)
1545                                 goto err;
1546                 }
1547         }
1548
1549         if (fsck_err_on(d.v->d_type != inode_d_type(target), c,
1550                         "incorrect d_type: got %s, should be %s:\n%s",
1551                         bch2_d_type_str(d.v->d_type),
1552                         bch2_d_type_str(inode_d_type(target)),
1553                         (printbuf_reset(&buf),
1554                          bch2_bkey_val_to_text(&buf, c, d.s_c), buf.buf))) {
1555                 n = bch2_trans_kmalloc(trans, bkey_bytes(d.k));
1556                 ret = PTR_ERR_OR_ZERO(n);
1557                 if (ret)
1558                         goto err;
1559
1560                 bkey_reassemble(&n->k_i, d.s_c);
1561                 n->v.d_type = inode_d_type(target);
1562
1563                 ret = bch2_trans_update(trans, iter, &n->k_i, 0);
1564                 if (ret)
1565                         goto err;
1566
1567                 d = dirent_i_to_s_c(n);
1568         }
1569
1570         if (d.v->d_type == DT_SUBVOL &&
1571             target->bi_parent_subvol != le32_to_cpu(d.v->d_parent_subvol) &&
1572             (c->sb.version < bcachefs_metadata_version_subvol_dirent ||
1573              fsck_err(c, "dirent has wrong d_parent_subvol field: got %u, should be %u",
1574                       le32_to_cpu(d.v->d_parent_subvol),
1575                       target->bi_parent_subvol))) {
1576                 n = bch2_trans_kmalloc(trans, bkey_bytes(d.k));
1577                 ret = PTR_ERR_OR_ZERO(n);
1578                 if (ret)
1579                         goto err;
1580
1581                 bkey_reassemble(&n->k_i, d.s_c);
1582                 n->v.d_parent_subvol = cpu_to_le32(target->bi_parent_subvol);
1583
1584                 ret = bch2_trans_update(trans, iter, &n->k_i, 0);
1585                 if (ret)
1586                         goto err;
1587
1588                 d = dirent_i_to_s_c(n);
1589         }
1590 out:
1591 err:
1592 fsck_err:
1593         printbuf_exit(&buf);
1594
1595         if (ret && !bch2_err_matches(ret, BCH_ERR_transaction_restart))
1596                 bch_err(c, "%s(): error %s", __func__, bch2_err_str(ret));
1597         return ret;
1598 }
1599
1600 static int check_dirent(struct btree_trans *trans, struct btree_iter *iter,
1601                         struct bkey_s_c k,
1602                         struct bch_hash_info *hash_info,
1603                         struct inode_walker *dir,
1604                         struct inode_walker *target,
1605                         struct snapshots_seen *s)
1606 {
1607         struct bch_fs *c = trans->c;
1608         struct bkey_s_c_dirent d;
1609         struct inode_walker_entry *i;
1610         struct printbuf buf = PRINTBUF;
1611         struct bpos equiv;
1612         int ret = 0;
1613
1614         ret = check_key_has_snapshot(trans, iter, k);
1615         if (ret) {
1616                 ret = ret < 0 ? ret : 0;
1617                 goto out;
1618         }
1619
1620         equiv = k.k->p;
1621         equiv.snapshot = bch2_snapshot_equiv(c, k.k->p.snapshot);
1622
1623         ret = snapshots_seen_update(c, s, iter->btree_id, k.k->p);
1624         if (ret)
1625                 goto err;
1626
1627         if (k.k->type == KEY_TYPE_whiteout)
1628                 goto out;
1629
1630         if (dir->cur_inum != k.k->p.inode) {
1631                 ret = check_subdir_count(trans, dir);
1632                 if (ret)
1633                         goto err;
1634         }
1635
1636         BUG_ON(!iter->path->should_be_locked);
1637
1638         ret = __walk_inode(trans, dir, equiv);
1639         if (ret < 0)
1640                 goto err;
1641
1642         if (fsck_err_on(ret == INT_MAX, c,
1643                         "dirent in nonexisting directory:\n%s",
1644                         (printbuf_reset(&buf),
1645                          bch2_bkey_val_to_text(&buf, c, k), buf.buf))) {
1646                 ret = bch2_btree_delete_at(trans, iter,
1647                                 BTREE_UPDATE_INTERNAL_SNAPSHOT_NODE);
1648                 goto out;
1649         }
1650
1651         if (ret == INT_MAX) {
1652                 ret = 0;
1653                 goto out;
1654         }
1655
1656         i = dir->inodes.data + ret;
1657         ret = 0;
1658
1659         if (fsck_err_on(!S_ISDIR(i->inode.bi_mode), c,
1660                         "dirent in non directory inode type %s:\n%s",
1661                         bch2_d_type_str(inode_d_type(&i->inode)),
1662                         (printbuf_reset(&buf),
1663                          bch2_bkey_val_to_text(&buf, c, k), buf.buf))) {
1664                 ret = bch2_btree_delete_at(trans, iter, 0);
1665                 goto out;
1666         }
1667
1668         if (dir->first_this_inode)
1669                 *hash_info = bch2_hash_info_init(c, &dir->inodes.data[0].inode);
1670
1671         ret = hash_check_key(trans, bch2_dirent_hash_desc,
1672                              hash_info, iter, k);
1673         if (ret < 0)
1674                 goto err;
1675         if (ret) {
1676                 /* dirent has been deleted */
1677                 ret = 0;
1678                 goto out;
1679         }
1680
1681         if (k.k->type != KEY_TYPE_dirent)
1682                 goto out;
1683
1684         d = bkey_s_c_to_dirent(k);
1685
1686         if (d.v->d_type == DT_SUBVOL) {
1687                 struct bch_inode_unpacked subvol_root;
1688                 u32 target_subvol = le32_to_cpu(d.v->d_child_subvol);
1689                 u32 target_snapshot;
1690                 u64 target_inum;
1691
1692                 ret = __subvol_lookup(trans, target_subvol,
1693                                       &target_snapshot, &target_inum);
1694                 if (ret && ret != -ENOENT)
1695                         goto err;
1696
1697                 if (fsck_err_on(ret, c,
1698                                 "dirent points to missing subvolume %llu",
1699                                 le64_to_cpu(d.v->d_child_subvol))) {
1700                         ret = __remove_dirent(trans, d.k->p);
1701                         goto err;
1702                 }
1703
1704                 ret = __lookup_inode(trans, target_inum,
1705                                    &subvol_root, &target_snapshot);
1706                 if (ret && ret != -ENOENT)
1707                         goto err;
1708
1709                 if (fsck_err_on(ret, c,
1710                                 "subvolume %u points to missing subvolume root %llu",
1711                                 target_subvol,
1712                                 target_inum)) {
1713                         bch_err(c, "repair not implemented yet");
1714                         ret = -EINVAL;
1715                         goto err;
1716                 }
1717
1718                 if (fsck_err_on(subvol_root.bi_subvol != target_subvol, c,
1719                                 "subvol root %llu has wrong bi_subvol field: got %u, should be %u",
1720                                 target_inum,
1721                                 subvol_root.bi_subvol, target_subvol)) {
1722                         subvol_root.bi_subvol = target_subvol;
1723                         ret = __write_inode(trans, &subvol_root, target_snapshot);
1724                         if (ret)
1725                                 goto err;
1726                 }
1727
1728                 ret = check_dirent_target(trans, iter, d, &subvol_root,
1729                                           target_snapshot);
1730                 if (ret)
1731                         goto err;
1732         } else {
1733                 ret = __get_visible_inodes(trans, target, s, le64_to_cpu(d.v->d_inum));
1734                 if (ret)
1735                         goto err;
1736
1737                 if (fsck_err_on(!target->inodes.nr, c,
1738                                 "dirent points to missing inode: (equiv %u)\n%s",
1739                                 equiv.snapshot,
1740                                 (printbuf_reset(&buf),
1741                                  bch2_bkey_val_to_text(&buf, c, k),
1742                                  buf.buf))) {
1743                         ret = __remove_dirent(trans, d.k->p);
1744                         if (ret)
1745                                 goto err;
1746                 }
1747
1748                 darray_for_each(target->inodes, i) {
1749                         ret = check_dirent_target(trans, iter, d,
1750                                                   &i->inode, i->snapshot);
1751                         if (ret)
1752                                 goto err;
1753                 }
1754         }
1755
1756         if (d.v->d_type == DT_DIR)
1757                 for_each_visible_inode(c, s, dir, equiv.snapshot, i)
1758                         i->count++;
1759
1760 out:
1761 err:
1762 fsck_err:
1763         printbuf_exit(&buf);
1764
1765         if (ret && !bch2_err_matches(ret, BCH_ERR_transaction_restart))
1766                 bch_err(c, "%s(): error %s", __func__, bch2_err_str(ret));
1767         return ret;
1768 }
1769
1770 /*
1771  * Walk dirents: verify that they all have a corresponding S_ISDIR inode,
1772  * validate d_type
1773  */
1774 noinline_for_stack
1775 static int check_dirents(struct bch_fs *c)
1776 {
1777         struct inode_walker dir = inode_walker_init();
1778         struct inode_walker target = inode_walker_init();
1779         struct snapshots_seen s;
1780         struct bch_hash_info hash_info;
1781         struct btree_trans trans;
1782         struct btree_iter iter;
1783         struct bkey_s_c k;
1784         int ret = 0;
1785
1786         bch_verbose(c, "checking dirents");
1787
1788         snapshots_seen_init(&s);
1789         bch2_trans_init(&trans, c, BTREE_ITER_MAX, 0);
1790
1791         ret = for_each_btree_key_commit(&trans, iter, BTREE_ID_dirents,
1792                         POS(BCACHEFS_ROOT_INO, 0),
1793                         BTREE_ITER_PREFETCH|BTREE_ITER_ALL_SNAPSHOTS,
1794                         k,
1795                         NULL, NULL,
1796                         BTREE_INSERT_LAZY_RW|BTREE_INSERT_NOFAIL,
1797                 check_dirent(&trans, &iter, k, &hash_info, &dir, &target, &s));
1798
1799         bch2_trans_exit(&trans);
1800         snapshots_seen_exit(&s);
1801         inode_walker_exit(&dir);
1802         inode_walker_exit(&target);
1803
1804         if (ret)
1805                 bch_err(c, "%s(): error %s", __func__, bch2_err_str(ret));
1806         return ret;
1807 }
1808
1809 static int check_xattr(struct btree_trans *trans, struct btree_iter *iter,
1810                        struct bkey_s_c k,
1811                        struct bch_hash_info *hash_info,
1812                        struct inode_walker *inode)
1813 {
1814         struct bch_fs *c = trans->c;
1815         int ret;
1816
1817         ret = check_key_has_snapshot(trans, iter, k);
1818         if (ret)
1819                 return ret;
1820
1821         ret = __walk_inode(trans, inode, k.k->p);
1822         if (ret < 0)
1823                 return ret;
1824
1825         if (fsck_err_on(ret == INT_MAX, c,
1826                         "xattr for missing inode %llu",
1827                         k.k->p.inode))
1828                 return bch2_btree_delete_at(trans, iter, 0);
1829
1830         if (ret == INT_MAX)
1831                 return 0;
1832
1833         ret = 0;
1834
1835         if (inode->first_this_inode)
1836                 *hash_info = bch2_hash_info_init(c, &inode->inodes.data[0].inode);
1837
1838         ret = hash_check_key(trans, bch2_xattr_hash_desc, hash_info, iter, k);
1839 fsck_err:
1840         if (ret && !bch2_err_matches(ret, BCH_ERR_transaction_restart))
1841                 bch_err(c, "%s(): error %s", __func__, bch2_err_str(ret));
1842         return ret;
1843 }
1844
1845 /*
1846  * Walk xattrs: verify that they all have a corresponding inode
1847  */
1848 noinline_for_stack
1849 static int check_xattrs(struct bch_fs *c)
1850 {
1851         struct inode_walker inode = inode_walker_init();
1852         struct bch_hash_info hash_info;
1853         struct btree_trans trans;
1854         struct btree_iter iter;
1855         struct bkey_s_c k;
1856         int ret = 0;
1857
1858         bch_verbose(c, "checking xattrs");
1859
1860         bch2_trans_init(&trans, c, BTREE_ITER_MAX, 0);
1861
1862         ret = for_each_btree_key_commit(&trans, iter, BTREE_ID_xattrs,
1863                         POS(BCACHEFS_ROOT_INO, 0),
1864                         BTREE_ITER_PREFETCH|BTREE_ITER_ALL_SNAPSHOTS,
1865                         k,
1866                         NULL, NULL,
1867                         BTREE_INSERT_LAZY_RW|BTREE_INSERT_NOFAIL,
1868                 check_xattr(&trans, &iter, k, &hash_info, &inode));
1869
1870         bch2_trans_exit(&trans);
1871
1872         if (ret)
1873                 bch_err(c, "%s(): error %s", __func__, bch2_err_str(ret));
1874         return ret;
1875 }
1876
1877 static int check_root_trans(struct btree_trans *trans)
1878 {
1879         struct bch_fs *c = trans->c;
1880         struct bch_inode_unpacked root_inode;
1881         u32 snapshot;
1882         u64 inum;
1883         int ret;
1884
1885         ret = __subvol_lookup(trans, BCACHEFS_ROOT_SUBVOL, &snapshot, &inum);
1886         if (ret && ret != -ENOENT)
1887                 return ret;
1888
1889         if (mustfix_fsck_err_on(ret, c, "root subvol missing")) {
1890                 struct bkey_i_subvolume root_subvol;
1891
1892                 snapshot        = U32_MAX;
1893                 inum            = BCACHEFS_ROOT_INO;
1894
1895                 bkey_subvolume_init(&root_subvol.k_i);
1896                 root_subvol.k.p.offset = BCACHEFS_ROOT_SUBVOL;
1897                 root_subvol.v.flags     = 0;
1898                 root_subvol.v.snapshot  = cpu_to_le32(snapshot);
1899                 root_subvol.v.inode     = cpu_to_le64(inum);
1900                 ret = commit_do(trans, NULL, NULL,
1901                                       BTREE_INSERT_NOFAIL|
1902                                       BTREE_INSERT_LAZY_RW,
1903                         __bch2_btree_insert(trans, BTREE_ID_subvolumes,
1904                                             &root_subvol.k_i, 0));
1905                 if (ret) {
1906                         bch_err(c, "error writing root subvol: %s", bch2_err_str(ret));
1907                         goto err;
1908                 }
1909
1910         }
1911
1912         ret = __lookup_inode(trans, BCACHEFS_ROOT_INO, &root_inode, &snapshot);
1913         if (ret && ret != -ENOENT)
1914                 return ret;
1915
1916         if (mustfix_fsck_err_on(ret, c, "root directory missing") ||
1917             mustfix_fsck_err_on(!S_ISDIR(root_inode.bi_mode), c,
1918                                 "root inode not a directory")) {
1919                 bch2_inode_init(c, &root_inode, 0, 0, S_IFDIR|0755,
1920                                 0, NULL);
1921                 root_inode.bi_inum = inum;
1922
1923                 ret = __write_inode(trans, &root_inode, snapshot);
1924                 if (ret)
1925                         bch_err(c, "error writing root inode: %s", bch2_err_str(ret));
1926         }
1927 err:
1928 fsck_err:
1929         return ret;
1930 }
1931
1932 /* Get root directory, create if it doesn't exist: */
1933 noinline_for_stack
1934 static int check_root(struct bch_fs *c)
1935 {
1936         bch_verbose(c, "checking root directory");
1937
1938         return bch2_trans_do(c, NULL, NULL,
1939                              BTREE_INSERT_NOFAIL|
1940                              BTREE_INSERT_LAZY_RW,
1941                 check_root_trans(&trans));
1942 }
1943
1944 struct pathbuf_entry {
1945         u64     inum;
1946         u32     snapshot;
1947 };
1948
1949 typedef DARRAY(struct pathbuf_entry) pathbuf;
1950
1951 static bool path_is_dup(pathbuf *p, u64 inum, u32 snapshot)
1952 {
1953         struct pathbuf_entry *i;
1954
1955         darray_for_each(*p, i)
1956                 if (i->inum     == inum &&
1957                     i->snapshot == snapshot)
1958                         return true;
1959
1960         return false;
1961 }
1962
1963 static int path_down(struct bch_fs *c, pathbuf *p,
1964                      u64 inum, u32 snapshot)
1965 {
1966         int ret = darray_push(p, ((struct pathbuf_entry) {
1967                 .inum           = inum,
1968                 .snapshot       = snapshot,
1969         }));
1970
1971         if (ret)
1972                 bch_err(c, "fsck: error allocating memory for pathbuf, size %zu",
1973                         p->size);
1974         return ret;
1975 }
1976
1977 /*
1978  * Check that a given inode is reachable from the root:
1979  *
1980  * XXX: we should also be verifying that inodes are in the right subvolumes
1981  */
1982 static int check_path(struct btree_trans *trans,
1983                       pathbuf *p,
1984                       struct bch_inode_unpacked *inode,
1985                       u32 snapshot)
1986 {
1987         struct bch_fs *c = trans->c;
1988         int ret = 0;
1989
1990         snapshot = bch2_snapshot_equiv(c, snapshot);
1991         p->nr = 0;
1992
1993         while (!(inode->bi_inum == BCACHEFS_ROOT_INO &&
1994                  inode->bi_subvol == BCACHEFS_ROOT_SUBVOL)) {
1995                 struct btree_iter dirent_iter;
1996                 struct bkey_s_c_dirent d;
1997                 u32 parent_snapshot = snapshot;
1998
1999                 if (inode->bi_subvol) {
2000                         u64 inum;
2001
2002                         ret = subvol_lookup(trans, inode->bi_parent_subvol,
2003                                             &parent_snapshot, &inum);
2004                         if (ret)
2005                                 break;
2006                 }
2007
2008                 ret = lockrestart_do(trans,
2009                         PTR_ERR_OR_ZERO((d = dirent_get_by_pos(trans, &dirent_iter,
2010                                           SPOS(inode->bi_dir, inode->bi_dir_offset,
2011                                                parent_snapshot))).k));
2012                 if (ret && ret != -ENOENT)
2013                         break;
2014
2015                 if (!ret && !dirent_points_to_inode(d, inode)) {
2016                         bch2_trans_iter_exit(trans, &dirent_iter);
2017                         ret = -ENOENT;
2018                 }
2019
2020                 if (ret == -ENOENT) {
2021                         if (fsck_err(c,  "unreachable inode %llu:%u, type %s nlink %u backptr %llu:%llu",
2022                                      inode->bi_inum, snapshot,
2023                                      bch2_d_type_str(inode_d_type(inode)),
2024                                      inode->bi_nlink,
2025                                      inode->bi_dir,
2026                                      inode->bi_dir_offset))
2027                                 ret = reattach_inode(trans, inode, snapshot);
2028                         break;
2029                 }
2030
2031                 bch2_trans_iter_exit(trans, &dirent_iter);
2032
2033                 if (!S_ISDIR(inode->bi_mode))
2034                         break;
2035
2036                 ret = path_down(c, p, inode->bi_inum, snapshot);
2037                 if (ret) {
2038                         bch_err(c, "memory allocation failure");
2039                         return ret;
2040                 }
2041
2042                 snapshot = parent_snapshot;
2043
2044                 ret = lookup_inode(trans, inode->bi_dir, inode, &snapshot);
2045                 if (ret) {
2046                         /* Should have been caught in dirents pass */
2047                         bch_err(c, "error looking up parent directory: %i", ret);
2048                         break;
2049                 }
2050
2051                 if (path_is_dup(p, inode->bi_inum, snapshot)) {
2052                         struct pathbuf_entry *i;
2053
2054                         /* XXX print path */
2055                         bch_err(c, "directory structure loop");
2056
2057                         darray_for_each(*p, i)
2058                                 pr_err("%llu:%u", i->inum, i->snapshot);
2059                         pr_err("%llu:%u", inode->bi_inum, snapshot);
2060
2061                         if (!fsck_err(c, "directory structure loop"))
2062                                 return 0;
2063
2064                         ret = commit_do(trans, NULL, NULL,
2065                                               BTREE_INSERT_NOFAIL|
2066                                               BTREE_INSERT_LAZY_RW,
2067                                         remove_backpointer(trans, inode));
2068                         if (ret) {
2069                                 bch_err(c, "error removing dirent: %i", ret);
2070                                 break;
2071                         }
2072
2073                         ret = reattach_inode(trans, inode, snapshot);
2074                 }
2075         }
2076 fsck_err:
2077         if (ret)
2078                 bch_err(c, "%s: err %s", __func__, bch2_err_str(ret));
2079         return ret;
2080 }
2081
2082 /*
2083  * Check for unreachable inodes, as well as loops in the directory structure:
2084  * After check_dirents(), if an inode backpointer doesn't exist that means it's
2085  * unreachable:
2086  */
2087 noinline_for_stack
2088 static int check_directory_structure(struct bch_fs *c)
2089 {
2090         struct btree_trans trans;
2091         struct btree_iter iter;
2092         struct bkey_s_c k;
2093         struct bch_inode_unpacked u;
2094         pathbuf path = { 0, };
2095         int ret;
2096
2097         bch2_trans_init(&trans, c, BTREE_ITER_MAX, 0);
2098
2099         for_each_btree_key(&trans, iter, BTREE_ID_inodes, POS_MIN,
2100                            BTREE_ITER_INTENT|
2101                            BTREE_ITER_PREFETCH|
2102                            BTREE_ITER_ALL_SNAPSHOTS, k, ret) {
2103                 if (!bkey_is_inode(k.k))
2104                         continue;
2105
2106                 ret = bch2_inode_unpack(k, &u);
2107                 if (ret) {
2108                         /* Should have been caught earlier in fsck: */
2109                         bch_err(c, "error unpacking inode %llu: %i", k.k->p.offset, ret);
2110                         break;
2111                 }
2112
2113                 if (u.bi_flags & BCH_INODE_UNLINKED)
2114                         continue;
2115
2116                 ret = check_path(&trans, &path, &u, iter.pos.snapshot);
2117                 if (ret)
2118                         break;
2119         }
2120         bch2_trans_iter_exit(&trans, &iter);
2121
2122         darray_exit(&path);
2123
2124         bch2_trans_exit(&trans);
2125         return ret;
2126 }
2127
2128 struct nlink_table {
2129         size_t          nr;
2130         size_t          size;
2131
2132         struct nlink {
2133                 u64     inum;
2134                 u32     snapshot;
2135                 u32     count;
2136         }               *d;
2137 };
2138
2139 static int add_nlink(struct bch_fs *c, struct nlink_table *t,
2140                      u64 inum, u32 snapshot)
2141 {
2142         if (t->nr == t->size) {
2143                 size_t new_size = max_t(size_t, 128UL, t->size * 2);
2144                 void *d = kvmalloc_array(new_size, sizeof(t->d[0]), GFP_KERNEL);
2145
2146                 if (!d) {
2147                         bch_err(c, "fsck: error allocating memory for nlink_table, size %zu",
2148                                 new_size);
2149                         return -BCH_ERR_ENOMEM_fsck_add_nlink;
2150                 }
2151
2152                 if (t->d)
2153                         memcpy(d, t->d, t->size * sizeof(t->d[0]));
2154                 kvfree(t->d);
2155
2156                 t->d = d;
2157                 t->size = new_size;
2158         }
2159
2160
2161         t->d[t->nr++] = (struct nlink) {
2162                 .inum           = inum,
2163                 .snapshot       = snapshot,
2164         };
2165
2166         return 0;
2167 }
2168
2169 static int nlink_cmp(const void *_l, const void *_r)
2170 {
2171         const struct nlink *l = _l;
2172         const struct nlink *r = _r;
2173
2174         return cmp_int(l->inum, r->inum) ?: cmp_int(l->snapshot, r->snapshot);
2175 }
2176
2177 static void inc_link(struct bch_fs *c, struct snapshots_seen *s,
2178                      struct nlink_table *links,
2179                      u64 range_start, u64 range_end, u64 inum, u32 snapshot)
2180 {
2181         struct nlink *link, key = {
2182                 .inum = inum, .snapshot = U32_MAX,
2183         };
2184
2185         if (inum < range_start || inum >= range_end)
2186                 return;
2187
2188         link = __inline_bsearch(&key, links->d, links->nr,
2189                                 sizeof(links->d[0]), nlink_cmp);
2190         if (!link)
2191                 return;
2192
2193         while (link > links->d && link[0].inum == link[-1].inum)
2194                 --link;
2195
2196         for (; link < links->d + links->nr && link->inum == inum; link++)
2197                 if (ref_visible(c, s, snapshot, link->snapshot)) {
2198                         link->count++;
2199                         if (link->snapshot >= snapshot)
2200                                 break;
2201                 }
2202 }
2203
2204 noinline_for_stack
2205 static int check_nlinks_find_hardlinks(struct bch_fs *c,
2206                                        struct nlink_table *t,
2207                                        u64 start, u64 *end)
2208 {
2209         struct btree_trans trans;
2210         struct btree_iter iter;
2211         struct bkey_s_c k;
2212         struct bch_inode_unpacked u;
2213         int ret = 0;
2214
2215         bch2_trans_init(&trans, c, BTREE_ITER_MAX, 0);
2216
2217         for_each_btree_key(&trans, iter, BTREE_ID_inodes,
2218                            POS(0, start),
2219                            BTREE_ITER_INTENT|
2220                            BTREE_ITER_PREFETCH|
2221                            BTREE_ITER_ALL_SNAPSHOTS, k, ret) {
2222                 if (!bkey_is_inode(k.k))
2223                         continue;
2224
2225                 /* Should never fail, checked by bch2_inode_invalid: */
2226                 BUG_ON(bch2_inode_unpack(k, &u));
2227
2228                 /*
2229                  * Backpointer and directory structure checks are sufficient for
2230                  * directories, since they can't have hardlinks:
2231                  */
2232                 if (S_ISDIR(le16_to_cpu(u.bi_mode)))
2233                         continue;
2234
2235                 if (!u.bi_nlink)
2236                         continue;
2237
2238                 ret = add_nlink(c, t, k.k->p.offset, k.k->p.snapshot);
2239                 if (ret) {
2240                         *end = k.k->p.offset;
2241                         ret = 0;
2242                         break;
2243                 }
2244
2245         }
2246         bch2_trans_iter_exit(&trans, &iter);
2247         bch2_trans_exit(&trans);
2248
2249         if (ret)
2250                 bch_err(c, "error in fsck: btree error %i while walking inodes", ret);
2251
2252         return ret;
2253 }
2254
2255 noinline_for_stack
2256 static int check_nlinks_walk_dirents(struct bch_fs *c, struct nlink_table *links,
2257                                      u64 range_start, u64 range_end)
2258 {
2259         struct btree_trans trans;
2260         struct snapshots_seen s;
2261         struct btree_iter iter;
2262         struct bkey_s_c k;
2263         struct bkey_s_c_dirent d;
2264         int ret;
2265
2266         snapshots_seen_init(&s);
2267
2268         bch2_trans_init(&trans, c, BTREE_ITER_MAX, 0);
2269
2270         for_each_btree_key(&trans, iter, BTREE_ID_dirents, POS_MIN,
2271                            BTREE_ITER_INTENT|
2272                            BTREE_ITER_PREFETCH|
2273                            BTREE_ITER_ALL_SNAPSHOTS, k, ret) {
2274                 ret = snapshots_seen_update(c, &s, iter.btree_id, k.k->p);
2275                 if (ret)
2276                         break;
2277
2278                 switch (k.k->type) {
2279                 case KEY_TYPE_dirent:
2280                         d = bkey_s_c_to_dirent(k);
2281
2282                         if (d.v->d_type != DT_DIR &&
2283                             d.v->d_type != DT_SUBVOL)
2284                                 inc_link(c, &s, links, range_start, range_end,
2285                                          le64_to_cpu(d.v->d_inum),
2286                                          bch2_snapshot_equiv(c, d.k->p.snapshot));
2287                         break;
2288                 }
2289         }
2290         bch2_trans_iter_exit(&trans, &iter);
2291
2292         if (ret)
2293                 bch_err(c, "error in fsck: btree error %i while walking dirents", ret);
2294
2295         bch2_trans_exit(&trans);
2296         snapshots_seen_exit(&s);
2297         return ret;
2298 }
2299
2300 static int check_nlinks_update_inode(struct btree_trans *trans, struct btree_iter *iter,
2301                                      struct bkey_s_c k,
2302                                      struct nlink_table *links,
2303                                      size_t *idx, u64 range_end)
2304 {
2305         struct bch_fs *c = trans->c;
2306         struct bch_inode_unpacked u;
2307         struct nlink *link = &links->d[*idx];
2308         int ret = 0;
2309
2310         if (k.k->p.offset >= range_end)
2311                 return 1;
2312
2313         if (!bkey_is_inode(k.k))
2314                 return 0;
2315
2316         BUG_ON(bch2_inode_unpack(k, &u));
2317
2318         if (S_ISDIR(le16_to_cpu(u.bi_mode)))
2319                 return 0;
2320
2321         if (!u.bi_nlink)
2322                 return 0;
2323
2324         while ((cmp_int(link->inum, k.k->p.offset) ?:
2325                 cmp_int(link->snapshot, k.k->p.snapshot)) < 0) {
2326                 BUG_ON(*idx == links->nr);
2327                 link = &links->d[++*idx];
2328         }
2329
2330         if (fsck_err_on(bch2_inode_nlink_get(&u) != link->count, c,
2331                         "inode %llu type %s has wrong i_nlink (%u, should be %u)",
2332                         u.bi_inum, bch2_d_types[mode_to_type(u.bi_mode)],
2333                         bch2_inode_nlink_get(&u), link->count)) {
2334                 bch2_inode_nlink_set(&u, link->count);
2335                 ret = __write_inode(trans, &u, k.k->p.snapshot);
2336         }
2337 fsck_err:
2338         return ret;
2339 }
2340
2341 noinline_for_stack
2342 static int check_nlinks_update_hardlinks(struct bch_fs *c,
2343                                struct nlink_table *links,
2344                                u64 range_start, u64 range_end)
2345 {
2346         struct btree_trans trans;
2347         struct btree_iter iter;
2348         struct bkey_s_c k;
2349         size_t idx = 0;
2350         int ret = 0;
2351
2352         bch2_trans_init(&trans, c, BTREE_ITER_MAX, 0);
2353
2354         ret = for_each_btree_key_commit(&trans, iter, BTREE_ID_inodes,
2355                         POS(0, range_start),
2356                         BTREE_ITER_INTENT|BTREE_ITER_PREFETCH|BTREE_ITER_ALL_SNAPSHOTS, k,
2357                         NULL, NULL, BTREE_INSERT_LAZY_RW|BTREE_INSERT_NOFAIL,
2358                 check_nlinks_update_inode(&trans, &iter, k, links, &idx, range_end));
2359
2360         bch2_trans_exit(&trans);
2361
2362         if (ret < 0) {
2363                 bch_err(c, "error in fsck: btree error %i while walking inodes", ret);
2364                 return ret;
2365         }
2366
2367         return 0;
2368 }
2369
2370 noinline_for_stack
2371 static int check_nlinks(struct bch_fs *c)
2372 {
2373         struct nlink_table links = { 0 };
2374         u64 this_iter_range_start, next_iter_range_start = 0;
2375         int ret = 0;
2376
2377         bch_verbose(c, "checking inode nlinks");
2378
2379         do {
2380                 this_iter_range_start = next_iter_range_start;
2381                 next_iter_range_start = U64_MAX;
2382
2383                 ret = check_nlinks_find_hardlinks(c, &links,
2384                                                   this_iter_range_start,
2385                                                   &next_iter_range_start);
2386
2387                 ret = check_nlinks_walk_dirents(c, &links,
2388                                           this_iter_range_start,
2389                                           next_iter_range_start);
2390                 if (ret)
2391                         break;
2392
2393                 ret = check_nlinks_update_hardlinks(c, &links,
2394                                          this_iter_range_start,
2395                                          next_iter_range_start);
2396                 if (ret)
2397                         break;
2398
2399                 links.nr = 0;
2400         } while (next_iter_range_start != U64_MAX);
2401
2402         kvfree(links.d);
2403
2404         return ret;
2405 }
2406
2407 static int fix_reflink_p_key(struct btree_trans *trans, struct btree_iter *iter,
2408                              struct bkey_s_c k)
2409 {
2410         struct bkey_s_c_reflink_p p;
2411         struct bkey_i_reflink_p *u;
2412         int ret;
2413
2414         if (k.k->type != KEY_TYPE_reflink_p)
2415                 return 0;
2416
2417         p = bkey_s_c_to_reflink_p(k);
2418
2419         if (!p.v->front_pad && !p.v->back_pad)
2420                 return 0;
2421
2422         u = bch2_trans_kmalloc(trans, sizeof(*u));
2423         ret = PTR_ERR_OR_ZERO(u);
2424         if (ret)
2425                 return ret;
2426
2427         bkey_reassemble(&u->k_i, k);
2428         u->v.front_pad  = 0;
2429         u->v.back_pad   = 0;
2430
2431         return bch2_trans_update(trans, iter, &u->k_i, BTREE_TRIGGER_NORUN);
2432 }
2433
2434 noinline_for_stack
2435 static int fix_reflink_p(struct bch_fs *c)
2436 {
2437         struct btree_trans trans;
2438         struct btree_iter iter;
2439         struct bkey_s_c k;
2440         int ret;
2441
2442         if (c->sb.version >= bcachefs_metadata_version_reflink_p_fix)
2443                 return 0;
2444
2445         bch_verbose(c, "fixing reflink_p keys");
2446
2447         bch2_trans_init(&trans, c, BTREE_ITER_MAX, 0);
2448
2449         ret = for_each_btree_key_commit(&trans, iter,
2450                         BTREE_ID_extents, POS_MIN,
2451                         BTREE_ITER_INTENT|BTREE_ITER_PREFETCH|BTREE_ITER_ALL_SNAPSHOTS, k,
2452                         NULL, NULL, BTREE_INSERT_NOFAIL|BTREE_INSERT_LAZY_RW,
2453                 fix_reflink_p_key(&trans, &iter, k));
2454
2455         bch2_trans_exit(&trans);
2456         return ret;
2457 }
2458
2459 /*
2460  * Checks for inconsistencies that shouldn't happen, unless we have a bug.
2461  * Doesn't fix them yet, mainly because they haven't yet been observed:
2462  */
2463 int bch2_fsck_full(struct bch_fs *c)
2464 {
2465         int ret;
2466 again:
2467         ret =   bch2_fs_check_snapshots(c) ?:
2468                 bch2_fs_check_subvols(c) ?:
2469                 bch2_delete_dead_snapshots(c) ?:
2470                 check_inodes(c, true) ?:
2471                 check_extents(c) ?:
2472                 check_dirents(c) ?:
2473                 check_xattrs(c) ?:
2474                 check_root(c) ?:
2475                 check_directory_structure(c) ?:
2476                 check_nlinks(c) ?:
2477                 fix_reflink_p(c);
2478
2479         if (bch2_err_matches(ret, BCH_ERR_need_snapshot_cleanup)) {
2480                 set_bit(BCH_FS_HAVE_DELETED_SNAPSHOTS, &c->flags);
2481                 goto again;
2482         }
2483
2484         return ret;
2485 }
2486
2487 int bch2_fsck_walk_inodes_only(struct bch_fs *c)
2488 {
2489         return  bch2_fs_check_snapshots(c) ?:
2490                 bch2_fs_check_subvols(c) ?:
2491                 bch2_delete_dead_snapshots(c) ?:
2492                 check_inodes(c, false);
2493 }