fs/bcachefs/btree_io.c

   1 // SPDX-License-Identifier: GPL-2.0
   2
   3 #include "bcachefs.h"
   4 #include "bkey_methods.h"
   5 #include "bkey_sort.h"
   6 #include "btree_cache.h"
   7 #include "btree_io.h"
   8 #include "btree_iter.h"
   9 #include "btree_locking.h"
  10 #include "btree_update.h"
  11 #include "btree_update_interior.h"
  12 #include "buckets.h"
  13 #include "checksum.h"
  14 #include "debug.h"
  15 #include "error.h"
  16 #include "extents.h"
  17 #include "io_write.h"
  18 #include "journal_reclaim.h"
  19 #include "journal_seq_blacklist.h"
  20 #include "recovery.h"
  21 #include "super-io.h"
  22 #include "trace.h"
  23
  24 #include <linux/sched/mm.h>
  25
  26 void bch2_btree_node_io_unlock(struct btree *b)
  27 {
  28         EBUG_ON(!btree_node_write_in_flight(b));
  29
  30         clear_btree_node_write_in_flight_inner(b);
  31         clear_btree_node_write_in_flight(b);
  32         wake_up_bit(&b->flags, BTREE_NODE_write_in_flight);
  33 }
  34
  35 void bch2_btree_node_io_lock(struct btree *b)
  36 {
  37         bch2_assert_btree_nodes_not_locked();
  38
  39         wait_on_bit_lock_io(&b->flags, BTREE_NODE_write_in_flight,
  40                             TASK_UNINTERRUPTIBLE);
  41 }
  42
  43 void __bch2_btree_node_wait_on_read(struct btree *b)
  44 {
  45         wait_on_bit_io(&b->flags, BTREE_NODE_read_in_flight,
  46                        TASK_UNINTERRUPTIBLE);
  47 }
  48
  49 void __bch2_btree_node_wait_on_write(struct btree *b)
  50 {
  51         wait_on_bit_io(&b->flags, BTREE_NODE_write_in_flight,
  52                        TASK_UNINTERRUPTIBLE);
  53 }
  54
  55 void bch2_btree_node_wait_on_read(struct btree *b)
  56 {
  57         bch2_assert_btree_nodes_not_locked();
  58
  59         wait_on_bit_io(&b->flags, BTREE_NODE_read_in_flight,
  60                        TASK_UNINTERRUPTIBLE);
  61 }
  62
  63 void bch2_btree_node_wait_on_write(struct btree *b)
  64 {
  65         bch2_assert_btree_nodes_not_locked();
  66
  67         wait_on_bit_io(&b->flags, BTREE_NODE_write_in_flight,
  68                        TASK_UNINTERRUPTIBLE);
  69 }
  70
  71 static void verify_no_dups(struct btree *b,
  72                            struct bkey_packed *start,
  73                            struct bkey_packed *end)
  74 {
  75 #ifdef CONFIG_BCACHEFS_DEBUG
  76         struct bkey_packed *k, *p;
  77
  78         if (start == end)
  79                 return;
  80
  81         for (p = start, k = bkey_p_next(start);
  82              k != end;
  83              p = k, k = bkey_p_next(k)) {
  84                 struct bkey l = bkey_unpack_key(b, p);
  85                 struct bkey r = bkey_unpack_key(b, k);
  86
  87                 BUG_ON(bpos_ge(l.p, bkey_start_pos(&r)));
  88         }
  89 #endif
  90 }
  91
  92 static void set_needs_whiteout(struct bset *i, int v)
  93 {
  94         struct bkey_packed *k;
  95
  96         for (k = i->start; k != vstruct_last(i); k = bkey_p_next(k))
  97                 k->needs_whiteout = v;
  98 }
  99
 100 static void btree_bounce_free(struct bch_fs *c, size_t size,
 101                               bool used_mempool, void *p)
 102 {
 103         if (used_mempool)
 104                 mempool_free(p, &c->btree_bounce_pool);
 105         else
 106                 kvfree(p);
 107 }
 108
 109 static void *btree_bounce_alloc(struct bch_fs *c, size_t size,
 110                                 bool *used_mempool)
 111 {
 112         unsigned flags = memalloc_nofs_save();
 113         void *p;
 114
 115         BUG_ON(size > c->opts.btree_node_size);
 116
 117         *used_mempool = false;
 118         p = kvmalloc(size, __GFP_NOWARN|GFP_NOWAIT);
 119         if (!p) {
 120                 *used_mempool = true;
 121                 p = mempool_alloc(&c->btree_bounce_pool, GFP_NOFS);
 122         }
 123         memalloc_nofs_restore(flags);
 124         return p;
 125 }
 126
 127 static void sort_bkey_ptrs(const struct btree *bt,
 128                            struct bkey_packed **ptrs, unsigned nr)
 129 {
 130         unsigned n = nr, a = nr / 2, b, c, d;
 131
 132         if (!a)
 133                 return;
 134
 135         /* Heap sort: see lib/sort.c: */
 136         while (1) {
 137                 if (a)
 138                         a--;
 139                 else if (--n)
 140                         swap(ptrs[0], ptrs[n]);
 141                 else
 142                         break;
 143
 144                 for (b = a; c = 2 * b + 1, (d = c + 1) < n;)
 145                         b = bch2_bkey_cmp_packed(bt,
 146                                             ptrs[c],
 147                                             ptrs[d]) >= 0 ? c : d;
 148                 if (d == n)
 149                         b = c;
 150
 151                 while (b != a &&
 152                        bch2_bkey_cmp_packed(bt,
 153                                        ptrs[a],
 154                                        ptrs[b]) >= 0)
 155                         b = (b - 1) / 2;
 156                 c = b;
 157                 while (b != a) {
 158                         b = (b - 1) / 2;
 159                         swap(ptrs[b], ptrs[c]);
 160                 }
 161         }
 162 }
 163
 164 static void bch2_sort_whiteouts(struct bch_fs *c, struct btree *b)
 165 {
 166         struct bkey_packed *new_whiteouts, **ptrs, **ptrs_end, *k;
 167         bool used_mempool = false;
 168         size_t bytes = b->whiteout_u64s * sizeof(u64);
 169
 170         if (!b->whiteout_u64s)
 171                 return;
 172
 173         new_whiteouts = btree_bounce_alloc(c, bytes, &used_mempool);
 174
 175         ptrs = ptrs_end = ((void *) new_whiteouts + bytes);
 176
 177         for (k = unwritten_whiteouts_start(b);
 178              k != unwritten_whiteouts_end(b);
 179              k = bkey_p_next(k))
 180                 *--ptrs = k;
 181
 182         sort_bkey_ptrs(b, ptrs, ptrs_end - ptrs);
 183
 184         k = new_whiteouts;
 185
 186         while (ptrs != ptrs_end) {
 187                 bkey_p_copy(k, *ptrs);
 188                 k = bkey_p_next(k);
 189                 ptrs++;
 190         }
 191
 192         verify_no_dups(b, new_whiteouts,
 193                        (void *) ((u64 *) new_whiteouts + b->whiteout_u64s));
 194
 195         memcpy_u64s(unwritten_whiteouts_start(b),
 196                     new_whiteouts, b->whiteout_u64s);
 197
 198         btree_bounce_free(c, bytes, used_mempool, new_whiteouts);
 199 }
 200
 201 static bool should_compact_bset(struct btree *b, struct bset_tree *t,
 202                                 bool compacting, enum compact_mode mode)
 203 {
 204         if (!bset_dead_u64s(b, t))
 205                 return false;
 206
 207         switch (mode) {
 208         case COMPACT_LAZY:
 209                 return should_compact_bset_lazy(b, t) ||
 210                         (compacting && !bset_written(b, bset(b, t)));
 211         case COMPACT_ALL:
 212                 return true;
 213         default:
 214                 BUG();
 215         }
 216 }
 217
 218 static bool bch2_drop_whiteouts(struct btree *b, enum compact_mode mode)
 219 {
 220         struct bset_tree *t;
 221         bool ret = false;
 222
 223         for_each_bset(b, t) {
 224                 struct bset *i = bset(b, t);
 225                 struct bkey_packed *k, *n, *out, *start, *end;
 226                 struct btree_node_entry *src = NULL, *dst = NULL;
 227
 228                 if (t != b->set && !bset_written(b, i)) {
 229                         src = container_of(i, struct btree_node_entry, keys);
 230                         dst = max(write_block(b),
 231                                   (void *) btree_bkey_last(b, t - 1));
 232                 }
 233
 234                 if (src != dst)
 235                         ret = true;
 236
 237                 if (!should_compact_bset(b, t, ret, mode)) {
 238                         if (src != dst) {
 239                                 memmove(dst, src, sizeof(*src) +
 240                                         le16_to_cpu(src->keys.u64s) *
 241                                         sizeof(u64));
 242                                 i = &dst->keys;
 243                                 set_btree_bset(b, t, i);
 244                         }
 245                         continue;
 246                 }
 247
 248                 start   = btree_bkey_first(b, t);
 249                 end     = btree_bkey_last(b, t);
 250
 251                 if (src != dst) {
 252                         memmove(dst, src, sizeof(*src));
 253                         i = &dst->keys;
 254                         set_btree_bset(b, t, i);
 255                 }
 256
 257                 out = i->start;
 258
 259                 for (k = start; k != end; k = n) {
 260                         n = bkey_p_next(k);
 261
 262                         if (!bkey_deleted(k)) {
 263                                 bkey_p_copy(out, k);
 264                                 out = bkey_p_next(out);
 265                         } else {
 266                                 BUG_ON(k->needs_whiteout);
 267                         }
 268                 }
 269
 270                 i->u64s = cpu_to_le16((u64 *) out - i->_data);
 271                 set_btree_bset_end(b, t);
 272                 bch2_bset_set_no_aux_tree(b, t);
 273                 ret = true;
 274         }
 275
 276         bch2_verify_btree_nr_keys(b);
 277
 278         bch2_btree_build_aux_trees(b);
 279
 280         return ret;
 281 }
 282
 283 bool bch2_compact_whiteouts(struct bch_fs *c, struct btree *b,
 284                             enum compact_mode mode)
 285 {
 286         return bch2_drop_whiteouts(b, mode);
 287 }
 288
 289 static void btree_node_sort(struct bch_fs *c, struct btree *b,
 290                             unsigned start_idx,
 291                             unsigned end_idx,
 292                             bool filter_whiteouts)
 293 {
 294         struct btree_node *out;
 295         struct sort_iter_stack sort_iter;
 296         struct bset_tree *t;
 297         struct bset *start_bset = bset(b, &b->set[start_idx]);
 298         bool used_mempool = false;
 299         u64 start_time, seq = 0;
 300         unsigned i, u64s = 0, bytes, shift = end_idx - start_idx - 1;
 301         bool sorting_entire_node = start_idx == 0 &&
 302                 end_idx == b->nsets;
 303
 304         sort_iter_stack_init(&sort_iter, b);
 305
 306         for (t = b->set + start_idx;
 307              t < b->set + end_idx;
 308              t++) {
 309                 u64s += le16_to_cpu(bset(b, t)->u64s);
 310                 sort_iter_add(&sort_iter.iter,
 311                               btree_bkey_first(b, t),
 312                               btree_bkey_last(b, t));
 313         }
 314
 315         bytes = sorting_entire_node
 316                 ? btree_buf_bytes(b)
 317                 : __vstruct_bytes(struct btree_node, u64s);
 318
 319         out = btree_bounce_alloc(c, bytes, &used_mempool);
 320
 321         start_time = local_clock();
 322
 323         u64s = bch2_sort_keys(out->keys.start, &sort_iter.iter, filter_whiteouts);
 324
 325         out->keys.u64s = cpu_to_le16(u64s);
 326
 327         BUG_ON(vstruct_end(&out->keys) > (void *) out + bytes);
 328
 329         if (sorting_entire_node)
 330                 bch2_time_stats_update(&c->times[BCH_TIME_btree_node_sort],
 331                                        start_time);
 332
 333         /* Make sure we preserve bset journal_seq: */
 334         for (t = b->set + start_idx; t < b->set + end_idx; t++)
 335                 seq = max(seq, le64_to_cpu(bset(b, t)->journal_seq));
 336         start_bset->journal_seq = cpu_to_le64(seq);
 337
 338         if (sorting_entire_node) {
 339                 u64s = le16_to_cpu(out->keys.u64s);
 340
 341                 BUG_ON(bytes != btree_buf_bytes(b));
 342
 343                 /*
 344                  * Our temporary buffer is the same size as the btree node's
 345                  * buffer, we can just swap buffers instead of doing a big
 346                  * memcpy()
 347                  */
 348                 *out = *b->data;
 349                 out->keys.u64s = cpu_to_le16(u64s);
 350                 swap(out, b->data);
 351                 set_btree_bset(b, b->set, &b->data->keys);
 352         } else {
 353                 start_bset->u64s = out->keys.u64s;
 354                 memcpy_u64s(start_bset->start,
 355                             out->keys.start,
 356                             le16_to_cpu(out->keys.u64s));
 357         }
 358
 359         for (i = start_idx + 1; i < end_idx; i++)
 360                 b->nr.bset_u64s[start_idx] +=
 361                         b->nr.bset_u64s[i];
 362
 363         b->nsets -= shift;
 364
 365         for (i = start_idx + 1; i < b->nsets; i++) {
 366                 b->nr.bset_u64s[i]      = b->nr.bset_u64s[i + shift];
 367                 b->set[i]               = b->set[i + shift];
 368         }
 369
 370         for (i = b->nsets; i < MAX_BSETS; i++)
 371                 b->nr.bset_u64s[i] = 0;
 372
 373         set_btree_bset_end(b, &b->set[start_idx]);
 374         bch2_bset_set_no_aux_tree(b, &b->set[start_idx]);
 375
 376         btree_bounce_free(c, bytes, used_mempool, out);
 377
 378         bch2_verify_btree_nr_keys(b);
 379 }
 380
 381 void bch2_btree_sort_into(struct bch_fs *c,
 382                          struct btree *dst,
 383                          struct btree *src)
 384 {
 385         struct btree_nr_keys nr;
 386         struct btree_node_iter src_iter;
 387         u64 start_time = local_clock();
 388
 389         BUG_ON(dst->nsets != 1);
 390
 391         bch2_bset_set_no_aux_tree(dst, dst->set);
 392
 393         bch2_btree_node_iter_init_from_start(&src_iter, src);
 394
 395         nr = bch2_sort_repack(btree_bset_first(dst),
 396                         src, &src_iter,
 397                         &dst->format,
 398                         true);
 399
 400         bch2_time_stats_update(&c->times[BCH_TIME_btree_node_sort],
 401                                start_time);
 402
 403         set_btree_bset_end(dst, dst->set);
 404
 405         dst->nr.live_u64s       += nr.live_u64s;
 406         dst->nr.bset_u64s[0]    += nr.bset_u64s[0];
 407         dst->nr.packed_keys     += nr.packed_keys;
 408         dst->nr.unpacked_keys   += nr.unpacked_keys;
 409
 410         bch2_verify_btree_nr_keys(dst);
 411 }
 412
 413 /*
 414  * We're about to add another bset to the btree node, so if there's currently
 415  * too many bsets - sort some of them together:
 416  */
 417 static bool btree_node_compact(struct bch_fs *c, struct btree *b)
 418 {
 419         unsigned unwritten_idx;
 420         bool ret = false;
 421
 422         for (unwritten_idx = 0;
 423              unwritten_idx < b->nsets;
 424              unwritten_idx++)
 425                 if (!bset_written(b, bset(b, &b->set[unwritten_idx])))
 426                         break;
 427
 428         if (b->nsets - unwritten_idx > 1) {
 429                 btree_node_sort(c, b, unwritten_idx,
 430                                 b->nsets, false);
 431                 ret = true;
 432         }
 433
 434         if (unwritten_idx > 1) {
 435                 btree_node_sort(c, b, 0, unwritten_idx, false);
 436                 ret = true;
 437         }
 438
 439         return ret;
 440 }
 441
 442 void bch2_btree_build_aux_trees(struct btree *b)
 443 {
 444         struct bset_tree *t;
 445
 446         for_each_bset(b, t)
 447                 bch2_bset_build_aux_tree(b, t,
 448                                 !bset_written(b, bset(b, t)) &&
 449                                 t == bset_tree_last(b));
 450 }
 451
 452 /*
 453  * If we have MAX_BSETS (3) bsets, should we sort them all down to just one?
 454  *
 455  * The first bset is going to be of similar order to the size of the node, the
 456  * last bset is bounded by btree_write_set_buffer(), which is set to keep the
 457  * memmove on insert from being too expensive: the middle bset should, ideally,
 458  * be the geometric mean of the first and the last.
 459  *
 460  * Returns true if the middle bset is greater than that geometric mean:
 461  */
 462 static inline bool should_compact_all(struct bch_fs *c, struct btree *b)
 463 {
 464         unsigned mid_u64s_bits =
 465                 (ilog2(btree_max_u64s(c)) + BTREE_WRITE_SET_U64s_BITS) / 2;
 466
 467         return bset_u64s(&b->set[1]) > 1U << mid_u64s_bits;
 468 }
 469
 470 /*
 471  * @bch_btree_init_next - initialize a new (unwritten) bset that can then be
 472  * inserted into
 473  *
 474  * Safe to call if there already is an unwritten bset - will only add a new bset
 475  * if @b doesn't already have one.
 476  *
 477  * Returns true if we sorted (i.e. invalidated iterators
 478  */
 479 void bch2_btree_init_next(struct btree_trans *trans, struct btree *b)
 480 {
 481         struct bch_fs *c = trans->c;
 482         struct btree_node_entry *bne;
 483         bool reinit_iter = false;
 484
 485         EBUG_ON(!six_lock_counts(&b->c.lock).n[SIX_LOCK_write]);
 486         BUG_ON(bset_written(b, bset(b, &b->set[1])));
 487         BUG_ON(btree_node_just_written(b));
 488
 489         if (b->nsets == MAX_BSETS &&
 490             !btree_node_write_in_flight(b) &&
 491             should_compact_all(c, b)) {
 492                 bch2_btree_node_write(c, b, SIX_LOCK_write,
 493                                       BTREE_WRITE_init_next_bset);
 494                 reinit_iter = true;
 495         }
 496
 497         if (b->nsets == MAX_BSETS &&
 498             btree_node_compact(c, b))
 499                 reinit_iter = true;
 500
 501         BUG_ON(b->nsets >= MAX_BSETS);
 502
 503         bne = want_new_bset(c, b);
 504         if (bne)
 505                 bch2_bset_init_next(b, bne);
 506
 507         bch2_btree_build_aux_trees(b);
 508
 509         if (reinit_iter)
 510                 bch2_trans_node_reinit_iter(trans, b);
 511 }
 512
 513 static void btree_err_msg(struct printbuf *out, struct bch_fs *c,
 514                           struct bch_dev *ca,
 515                           struct btree *b, struct bset *i,
 516                           unsigned offset, int write)
 517 {
 518         prt_printf(out, bch2_log_msg(c, "%s"),
 519                    write == READ
 520                    ? "error validating btree node "
 521                    : "corrupt btree node before write ");
 522         if (ca)
 523                 prt_printf(out, "on %s ", ca->name);
 524         prt_printf(out, "at btree ");
 525         bch2_btree_pos_to_text(out, c, b);
 526
 527         prt_printf(out, "\n  node offset %u/%u",
 528                    b->written, btree_ptr_sectors_written(&b->key));
 529         if (i)
 530                 prt_printf(out, " bset u64s %u", le16_to_cpu(i->u64s));
 531         prt_str(out, ": ");
 532 }
 533
 534 __printf(9, 10)
 535 static int __btree_err(int ret,
 536                        struct bch_fs *c,
 537                        struct bch_dev *ca,
 538                        struct btree *b,
 539                        struct bset *i,
 540                        int write,
 541                        bool have_retry,
 542                        enum bch_sb_error_id err_type,
 543                        const char *fmt, ...)
 544 {
 545         struct printbuf out = PRINTBUF;
 546         va_list args;
 547
 548         btree_err_msg(&out, c, ca, b, i, b->written, write);
 549
 550         va_start(args, fmt);
 551         prt_vprintf(&out, fmt, args);
 552         va_end(args);
 553
 554         if (write == WRITE) {
 555                 bch2_print_string_as_lines(KERN_ERR, out.buf);
 556                 ret = c->opts.errors == BCH_ON_ERROR_continue
 557                         ? 0
 558                         : -BCH_ERR_fsck_errors_not_fixed;
 559                 goto out;
 560         }
 561
 562         if (!have_retry && ret == -BCH_ERR_btree_node_read_err_want_retry)
 563                 ret = -BCH_ERR_btree_node_read_err_fixable;
 564         if (!have_retry && ret == -BCH_ERR_btree_node_read_err_must_retry)
 565                 ret = -BCH_ERR_btree_node_read_err_bad_node;
 566
 567         if (ret != -BCH_ERR_btree_node_read_err_fixable)
 568                 bch2_sb_error_count(c, err_type);
 569
 570         switch (ret) {
 571         case -BCH_ERR_btree_node_read_err_fixable:
 572                 ret = bch2_fsck_err(c, FSCK_CAN_FIX, err_type, "%s", out.buf);
 573                 if (ret != -BCH_ERR_fsck_fix &&
 574                     ret != -BCH_ERR_fsck_ignore)
 575                         goto fsck_err;
 576                 ret = -BCH_ERR_fsck_fix;
 577                 break;
 578         case -BCH_ERR_btree_node_read_err_want_retry:
 579         case -BCH_ERR_btree_node_read_err_must_retry:
 580                 bch2_print_string_as_lines(KERN_ERR, out.buf);
 581                 break;
 582         case -BCH_ERR_btree_node_read_err_bad_node:
 583                 bch2_print_string_as_lines(KERN_ERR, out.buf);
 584                 ret = bch2_topology_error(c);
 585                 break;
 586         case -BCH_ERR_btree_node_read_err_incompatible:
 587                 bch2_print_string_as_lines(KERN_ERR, out.buf);
 588                 ret = -BCH_ERR_fsck_errors_not_fixed;
 589                 break;
 590         default:
 591                 BUG();
 592         }
 593 out:
 594 fsck_err:
 595         printbuf_exit(&out);
 596         return ret;
 597 }
 598
 599 #define btree_err(type, c, ca, b, i, _err_type, msg, ...)               \
 600 ({                                                                      \
 601         int _ret = __btree_err(type, c, ca, b, i, write, have_retry,    \
 602                                BCH_FSCK_ERR_##_err_type,                \
 603                                msg, ##__VA_ARGS__);                     \
 604                                                                         \
 605         if (_ret != -BCH_ERR_fsck_fix) {                                \
 606                 ret = _ret;                                             \
 607                 goto fsck_err;                                          \
 608         }                                                               \
 609                                                                         \
 610         *saw_error = true;                                              \
 611 })
 612
 613 #define btree_err_on(cond, ...) ((cond) ? btree_err(__VA_ARGS__) : false)
 614
 615 /*
 616  * When btree topology repair changes the start or end of a node, that might
 617  * mean we have to drop keys that are no longer inside the node:
 618  */
 619 __cold
 620 void bch2_btree_node_drop_keys_outside_node(struct btree *b)
 621 {
 622         struct bset_tree *t;
 623
 624         for_each_bset(b, t) {
 625                 struct bset *i = bset(b, t);
 626                 struct bkey_packed *k;
 627
 628                 for (k = i->start; k != vstruct_last(i); k = bkey_p_next(k))
 629                         if (bkey_cmp_left_packed(b, k, &b->data->min_key) >= 0)
 630                                 break;
 631
 632                 if (k != i->start) {
 633                         unsigned shift = (u64 *) k - (u64 *) i->start;
 634
 635                         memmove_u64s_down(i->start, k,
 636                                           (u64 *) vstruct_end(i) - (u64 *) k);
 637                         i->u64s = cpu_to_le16(le16_to_cpu(i->u64s) - shift);
 638                         set_btree_bset_end(b, t);
 639                 }
 640
 641                 for (k = i->start; k != vstruct_last(i); k = bkey_p_next(k))
 642                         if (bkey_cmp_left_packed(b, k, &b->data->max_key) > 0)
 643                                 break;
 644
 645                 if (k != vstruct_last(i)) {
 646                         i->u64s = cpu_to_le16((u64 *) k - (u64 *) i->start);
 647                         set_btree_bset_end(b, t);
 648                 }
 649         }
 650
 651         /*
 652          * Always rebuild search trees: eytzinger search tree nodes directly
 653          * depend on the values of min/max key:
 654          */
 655         bch2_bset_set_no_aux_tree(b, b->set);
 656         bch2_btree_build_aux_trees(b);
 657         b->nr = bch2_btree_node_count_keys(b);
 658
 659         struct bkey_s_c k;
 660         struct bkey unpacked;
 661         struct btree_node_iter iter;
 662         for_each_btree_node_key_unpack(b, k, &iter, &unpacked) {
 663                 BUG_ON(bpos_lt(k.k->p, b->data->min_key));
 664                 BUG_ON(bpos_gt(k.k->p, b->data->max_key));
 665         }
 666 }
 667
 668 static int validate_bset(struct bch_fs *c, struct bch_dev *ca,
 669                          struct btree *b, struct bset *i,
 670                          unsigned offset, unsigned sectors,
 671                          int write, bool have_retry, bool *saw_error)
 672 {
 673         unsigned version = le16_to_cpu(i->version);
 674         struct printbuf buf1 = PRINTBUF;
 675         struct printbuf buf2 = PRINTBUF;
 676         int ret = 0;
 677
 678         btree_err_on(!bch2_version_compatible(version),
 679                      -BCH_ERR_btree_node_read_err_incompatible,
 680                      c, ca, b, i,
 681                      btree_node_unsupported_version,
 682                      "unsupported bset version %u.%u",
 683                      BCH_VERSION_MAJOR(version),
 684                      BCH_VERSION_MINOR(version));
 685
 686         if (btree_err_on(version < c->sb.version_min,
 687                          -BCH_ERR_btree_node_read_err_fixable,
 688                          c, NULL, b, i,
 689                          btree_node_bset_older_than_sb_min,
 690                          "bset version %u older than superblock version_min %u",
 691                          version, c->sb.version_min)) {
 692                 mutex_lock(&c->sb_lock);
 693                 c->disk_sb.sb->version_min = cpu_to_le16(version);
 694                 bch2_write_super(c);
 695                 mutex_unlock(&c->sb_lock);
 696         }
 697
 698         if (btree_err_on(BCH_VERSION_MAJOR(version) >
 699                          BCH_VERSION_MAJOR(c->sb.version),
 700                          -BCH_ERR_btree_node_read_err_fixable,
 701                          c, NULL, b, i,
 702                          btree_node_bset_newer_than_sb,
 703                          "bset version %u newer than superblock version %u",
 704                          version, c->sb.version)) {
 705                 mutex_lock(&c->sb_lock);
 706                 c->disk_sb.sb->version = cpu_to_le16(version);
 707                 bch2_write_super(c);
 708                 mutex_unlock(&c->sb_lock);
 709         }
 710
 711         btree_err_on(BSET_SEPARATE_WHITEOUTS(i),
 712                      -BCH_ERR_btree_node_read_err_incompatible,
 713                      c, ca, b, i,
 714                      btree_node_unsupported_version,
 715                      "BSET_SEPARATE_WHITEOUTS no longer supported");
 716
 717         if (btree_err_on(offset + sectors > btree_sectors(c),
 718                          -BCH_ERR_btree_node_read_err_fixable,
 719                          c, ca, b, i,
 720                          bset_past_end_of_btree_node,
 721                          "bset past end of btree node")) {
 722                 i->u64s = 0;
 723                 ret = 0;
 724                 goto out;
 725         }
 726
 727         btree_err_on(offset && !i->u64s,
 728                      -BCH_ERR_btree_node_read_err_fixable,
 729                      c, ca, b, i,
 730                      bset_empty,
 731                      "empty bset");
 732
 733         btree_err_on(BSET_OFFSET(i) && BSET_OFFSET(i) != offset,
 734                      -BCH_ERR_btree_node_read_err_want_retry,
 735                      c, ca, b, i,
 736                      bset_wrong_sector_offset,
 737                      "bset at wrong sector offset");
 738
 739         if (!offset) {
 740                 struct btree_node *bn =
 741                         container_of(i, struct btree_node, keys);
 742                 /* These indicate that we read the wrong btree node: */
 743
 744                 if (b->key.k.type == KEY_TYPE_btree_ptr_v2) {
 745                         struct bch_btree_ptr_v2 *bp =
 746                                 &bkey_i_to_btree_ptr_v2(&b->key)->v;
 747
 748                         /* XXX endianness */
 749                         btree_err_on(bp->seq != bn->keys.seq,
 750                                      -BCH_ERR_btree_node_read_err_must_retry,
 751                                      c, ca, b, NULL,
 752                                      bset_bad_seq,
 753                                      "incorrect sequence number (wrong btree node)");
 754                 }
 755
 756                 btree_err_on(BTREE_NODE_ID(bn) != b->c.btree_id,
 757                              -BCH_ERR_btree_node_read_err_must_retry,
 758                              c, ca, b, i,
 759                              btree_node_bad_btree,
 760                              "incorrect btree id");
 761
 762                 btree_err_on(BTREE_NODE_LEVEL(bn) != b->c.level,
 763                              -BCH_ERR_btree_node_read_err_must_retry,
 764                              c, ca, b, i,
 765                              btree_node_bad_level,
 766                              "incorrect level");
 767
 768                 if (!write)
 769                         compat_btree_node(b->c.level, b->c.btree_id, version,
 770                                           BSET_BIG_ENDIAN(i), write, bn);
 771
 772                 if (b->key.k.type == KEY_TYPE_btree_ptr_v2) {
 773                         struct bch_btree_ptr_v2 *bp =
 774                                 &bkey_i_to_btree_ptr_v2(&b->key)->v;
 775
 776                         if (BTREE_PTR_RANGE_UPDATED(bp)) {
 777                                 b->data->min_key = bp->min_key;
 778                                 b->data->max_key = b->key.k.p;
 779                         }
 780
 781                         btree_err_on(!bpos_eq(b->data->min_key, bp->min_key),
 782                                      -BCH_ERR_btree_node_read_err_must_retry,
 783                                      c, ca, b, NULL,
 784                                      btree_node_bad_min_key,
 785                                      "incorrect min_key: got %s should be %s",
 786                                      (printbuf_reset(&buf1),
 787                                       bch2_bpos_to_text(&buf1, bn->min_key), buf1.buf),
 788                                      (printbuf_reset(&buf2),
 789                                       bch2_bpos_to_text(&buf2, bp->min_key), buf2.buf));
 790                 }
 791
 792                 btree_err_on(!bpos_eq(bn->max_key, b->key.k.p),
 793                              -BCH_ERR_btree_node_read_err_must_retry,
 794                              c, ca, b, i,
 795                              btree_node_bad_max_key,
 796                              "incorrect max key %s",
 797                              (printbuf_reset(&buf1),
 798                               bch2_bpos_to_text(&buf1, bn->max_key), buf1.buf));
 799
 800                 if (write)
 801                         compat_btree_node(b->c.level, b->c.btree_id, version,
 802                                           BSET_BIG_ENDIAN(i), write, bn);
 803
 804                 btree_err_on(bch2_bkey_format_invalid(c, &bn->format, write, &buf1),
 805                              -BCH_ERR_btree_node_read_err_bad_node,
 806                              c, ca, b, i,
 807                              btree_node_bad_format,
 808                              "invalid bkey format: %s\n  %s", buf1.buf,
 809                              (printbuf_reset(&buf2),
 810                               bch2_bkey_format_to_text(&buf2, &bn->format), buf2.buf));
 811                 printbuf_reset(&buf1);
 812
 813                 compat_bformat(b->c.level, b->c.btree_id, version,
 814                                BSET_BIG_ENDIAN(i), write,
 815                                &bn->format);
 816         }
 817 out:
 818 fsck_err:
 819         printbuf_exit(&buf2);
 820         printbuf_exit(&buf1);
 821         return ret;
 822 }
 823
 824 static int bset_key_invalid(struct bch_fs *c, struct btree *b,
 825                             struct bkey_s_c k,
 826                             bool updated_range, int rw,
 827                             struct printbuf *err)
 828 {
 829         return __bch2_bkey_invalid(c, k, btree_node_type(b), READ, err) ?:
 830                 (!updated_range ? bch2_bkey_in_btree_node(c, b, k, err) : 0) ?:
 831                 (rw == WRITE ? bch2_bkey_val_invalid(c, k, READ, err) : 0);
 832 }
 833
 834 static bool bkey_packed_valid(struct bch_fs *c, struct btree *b,
 835                          struct bset *i, struct bkey_packed *k)
 836 {
 837         if (bkey_p_next(k) > vstruct_last(i))
 838                 return false;
 839
 840         if (k->format > KEY_FORMAT_CURRENT)
 841                 return false;
 842
 843         if (!bkeyp_u64s_valid(&b->format, k))
 844                 return false;
 845
 846         struct printbuf buf = PRINTBUF;
 847         struct bkey tmp;
 848         struct bkey_s u = __bkey_disassemble(b, k, &tmp);
 849         bool ret = __bch2_bkey_invalid(c, u.s_c, btree_node_type(b), READ, &buf);
 850         printbuf_exit(&buf);
 851         return ret;
 852 }
 853
 854 static int validate_bset_keys(struct bch_fs *c, struct btree *b,
 855                          struct bset *i, int write,
 856                          bool have_retry, bool *saw_error)
 857 {
 858         unsigned version = le16_to_cpu(i->version);
 859         struct bkey_packed *k, *prev = NULL;
 860         struct printbuf buf = PRINTBUF;
 861         bool updated_range = b->key.k.type == KEY_TYPE_btree_ptr_v2 &&
 862                 BTREE_PTR_RANGE_UPDATED(&bkey_i_to_btree_ptr_v2(&b->key)->v);
 863         int ret = 0;
 864
 865         for (k = i->start;
 866              k != vstruct_last(i);) {
 867                 struct bkey_s u;
 868                 struct bkey tmp;
 869                 unsigned next_good_key;
 870
 871                 if (btree_err_on(bkey_p_next(k) > vstruct_last(i),
 872                                  -BCH_ERR_btree_node_read_err_fixable,
 873                                  c, NULL, b, i,
 874                                  btree_node_bkey_past_bset_end,
 875                                  "key extends past end of bset")) {
 876                         i->u64s = cpu_to_le16((u64 *) k - i->_data);
 877                         break;
 878                 }
 879
 880                 if (btree_err_on(k->format > KEY_FORMAT_CURRENT,
 881                                  -BCH_ERR_btree_node_read_err_fixable,
 882                                  c, NULL, b, i,
 883                                  btree_node_bkey_bad_format,
 884                                  "invalid bkey format %u", k->format))
 885                         goto drop_this_key;
 886
 887                 if (btree_err_on(!bkeyp_u64s_valid(&b->format, k),
 888                                  -BCH_ERR_btree_node_read_err_fixable,
 889                                  c, NULL, b, i,
 890                                  btree_node_bkey_bad_u64s,
 891                                  "bad k->u64s %u (min %u max %zu)", k->u64s,
 892                                  bkeyp_key_u64s(&b->format, k),
 893                                  U8_MAX - BKEY_U64s + bkeyp_key_u64s(&b->format, k)))
 894                         goto drop_this_key;
 895
 896                 if (!write)
 897                         bch2_bkey_compat(b->c.level, b->c.btree_id, version,
 898                                     BSET_BIG_ENDIAN(i), write,
 899                                     &b->format, k);
 900
 901                 u = __bkey_disassemble(b, k, &tmp);
 902
 903                 printbuf_reset(&buf);
 904                 if (bset_key_invalid(c, b, u.s_c, updated_range, write, &buf)) {
 905                         printbuf_reset(&buf);
 906                         bset_key_invalid(c, b, u.s_c, updated_range, write, &buf);
 907                         prt_printf(&buf, "\n  ");
 908                         bch2_bkey_val_to_text(&buf, c, u.s_c);
 909
 910                         btree_err(-BCH_ERR_btree_node_read_err_fixable,
 911                                   c, NULL, b, i,
 912                                   btree_node_bad_bkey,
 913                                   "invalid bkey: %s", buf.buf);
 914                         goto drop_this_key;
 915                 }
 916
 917                 if (write)
 918                         bch2_bkey_compat(b->c.level, b->c.btree_id, version,
 919                                     BSET_BIG_ENDIAN(i), write,
 920                                     &b->format, k);
 921
 922                 if (prev && bkey_iter_cmp(b, prev, k) > 0) {
 923                         struct bkey up = bkey_unpack_key(b, prev);
 924
 925                         printbuf_reset(&buf);
 926                         prt_printf(&buf, "keys out of order: ");
 927                         bch2_bkey_to_text(&buf, &up);
 928                         prt_printf(&buf, " > ");
 929                         bch2_bkey_to_text(&buf, u.k);
 930
 931                         if (btree_err(-BCH_ERR_btree_node_read_err_fixable,
 932                                       c, NULL, b, i,
 933                                       btree_node_bkey_out_of_order,
 934                                       "%s", buf.buf))
 935                                 goto drop_this_key;
 936                 }
 937
 938                 prev = k;
 939                 k = bkey_p_next(k);
 940                 continue;
 941 drop_this_key:
 942                 next_good_key = k->u64s;
 943
 944                 if (!next_good_key ||
 945                     (BSET_BIG_ENDIAN(i) == CPU_BIG_ENDIAN &&
 946                      version >= bcachefs_metadata_version_snapshot)) {
 947                         /*
 948                          * only do scanning if bch2_bkey_compat() has nothing to
 949                          * do
 950                          */
 951
 952                         if (!bkey_packed_valid(c, b, i, (void *) ((u64 *) k + next_good_key))) {
 953                                 for (next_good_key = 1;
 954                                      next_good_key < (u64 *) vstruct_last(i) - (u64 *) k;
 955                                      next_good_key++)
 956                                         if (bkey_packed_valid(c, b, i, (void *) ((u64 *) k + next_good_key)))
 957                                                 goto got_good_key;
 958                         }
 959
 960                         /*
 961                          * didn't find a good key, have to truncate the rest of
 962                          * the bset
 963                          */
 964                         next_good_key = (u64 *) vstruct_last(i) - (u64 *) k;
 965                 }
 966 got_good_key:
 967                 le16_add_cpu(&i->u64s, -next_good_key);
 968                 memmove_u64s_down(k, bkey_p_next(k), (u64 *) vstruct_end(i) - (u64 *) k);
 969         }
 970 fsck_err:
 971         printbuf_exit(&buf);
 972         return ret;
 973 }
 974
 975 int bch2_btree_node_read_done(struct bch_fs *c, struct bch_dev *ca,
 976                               struct btree *b, bool have_retry, bool *saw_error)
 977 {
 978         struct btree_node_entry *bne;
 979         struct sort_iter *iter;
 980         struct btree_node *sorted;
 981         struct bkey_packed *k;
 982         struct bset *i;
 983         bool used_mempool, blacklisted;
 984         bool updated_range = b->key.k.type == KEY_TYPE_btree_ptr_v2 &&
 985                 BTREE_PTR_RANGE_UPDATED(&bkey_i_to_btree_ptr_v2(&b->key)->v);
 986         unsigned u64s;
 987         unsigned ptr_written = btree_ptr_sectors_written(&b->key);
 988         struct printbuf buf = PRINTBUF;
 989         int ret = 0, retry_read = 0, write = READ;
 990         u64 start_time = local_clock();
 991
 992         b->version_ondisk = U16_MAX;
 993         /* We might get called multiple times on read retry: */
 994         b->written = 0;
 995
 996         iter = mempool_alloc(&c->fill_iter, GFP_NOFS);
 997         sort_iter_init(iter, b, (btree_blocks(c) + 1) * 2);
 998
 999         if (bch2_meta_read_fault("btree"))
1000                 btree_err(-BCH_ERR_btree_node_read_err_must_retry,
1001                           c, ca, b, NULL,
1002                           btree_node_fault_injected,
1003                           "dynamic fault");
1004
1005         btree_err_on(le64_to_cpu(b->data->magic) != bset_magic(c),
1006                      -BCH_ERR_btree_node_read_err_must_retry,
1007                      c, ca, b, NULL,
1008                      btree_node_bad_magic,
1009                      "bad magic: want %llx, got %llx",
1010                      bset_magic(c), le64_to_cpu(b->data->magic));
1011
1012         if (b->key.k.type == KEY_TYPE_btree_ptr_v2) {
1013                 struct bch_btree_ptr_v2 *bp =
1014                         &bkey_i_to_btree_ptr_v2(&b->key)->v;
1015
1016                 bch2_bpos_to_text(&buf, b->data->min_key);
1017                 prt_str(&buf, "-");
1018                 bch2_bpos_to_text(&buf, b->data->max_key);
1019
1020                 btree_err_on(b->data->keys.seq != bp->seq,
1021                              -BCH_ERR_btree_node_read_err_must_retry,
1022                              c, ca, b, NULL,
1023                              btree_node_bad_seq,
1024                              "got wrong btree node (want %llx got %llx)\n"
1025                              "got btree %s level %llu pos %s",
1026                              bp->seq, b->data->keys.seq,
1027                              bch2_btree_id_str(BTREE_NODE_ID(b->data)),
1028                              BTREE_NODE_LEVEL(b->data),
1029                              buf.buf);
1030         } else {
1031                 btree_err_on(!b->data->keys.seq,
1032                              -BCH_ERR_btree_node_read_err_must_retry,
1033                              c, ca, b, NULL,
1034                              btree_node_bad_seq,
1035                              "bad btree header: seq 0");
1036         }
1037
1038         while (b->written < (ptr_written ?: btree_sectors(c))) {
1039                 unsigned sectors;
1040                 struct nonce nonce;
1041                 bool first = !b->written;
1042                 bool csum_bad;
1043
1044                 if (!b->written) {
1045                         i = &b->data->keys;
1046
1047                         btree_err_on(!bch2_checksum_type_valid(c, BSET_CSUM_TYPE(i)),
1048                                      -BCH_ERR_btree_node_read_err_want_retry,
1049                                      c, ca, b, i,
1050                                      bset_unknown_csum,
1051                                      "unknown checksum type %llu", BSET_CSUM_TYPE(i));
1052
1053                         nonce = btree_nonce(i, b->written << 9);
1054
1055                         struct bch_csum csum = csum_vstruct(c, BSET_CSUM_TYPE(i), nonce, b->data);
1056                         csum_bad = bch2_crc_cmp(b->data->csum, csum);
1057                         if (csum_bad)
1058                                 bch2_io_error(ca, BCH_MEMBER_ERROR_checksum);
1059
1060                         btree_err_on(csum_bad,
1061                                      -BCH_ERR_btree_node_read_err_want_retry,
1062                                      c, ca, b, i,
1063                                      bset_bad_csum,
1064                                      "%s",
1065                                      (printbuf_reset(&buf),
1066                                       bch2_csum_err_msg(&buf, BSET_CSUM_TYPE(i), b->data->csum, csum),
1067                                       buf.buf));
1068
1069                         ret = bset_encrypt(c, i, b->written << 9);
1070                         if (bch2_fs_fatal_err_on(ret, c,
1071                                         "decrypting btree node: %s", bch2_err_str(ret)))
1072                                 goto fsck_err;
1073
1074                         btree_err_on(btree_node_type_is_extents(btree_node_type(b)) &&
1075                                      !BTREE_NODE_NEW_EXTENT_OVERWRITE(b->data),
1076                                      -BCH_ERR_btree_node_read_err_incompatible,
1077                                      c, NULL, b, NULL,
1078                                      btree_node_unsupported_version,
1079                                      "btree node does not have NEW_EXTENT_OVERWRITE set");
1080
1081                         sectors = vstruct_sectors(b->data, c->block_bits);
1082                 } else {
1083                         bne = write_block(b);
1084                         i = &bne->keys;
1085
1086                         if (i->seq != b->data->keys.seq)
1087                                 break;
1088
1089                         btree_err_on(!bch2_checksum_type_valid(c, BSET_CSUM_TYPE(i)),
1090                                      -BCH_ERR_btree_node_read_err_want_retry,
1091                                      c, ca, b, i,
1092                                      bset_unknown_csum,
1093                                      "unknown checksum type %llu", BSET_CSUM_TYPE(i));
1094
1095                         nonce = btree_nonce(i, b->written << 9);
1096                         struct bch_csum csum = csum_vstruct(c, BSET_CSUM_TYPE(i), nonce, bne);
1097                         csum_bad = bch2_crc_cmp(bne->csum, csum);
1098                         if (csum_bad)
1099                                 bch2_io_error(ca, BCH_MEMBER_ERROR_checksum);
1100
1101                         btree_err_on(csum_bad,
1102                                      -BCH_ERR_btree_node_read_err_want_retry,
1103                                      c, ca, b, i,
1104                                      bset_bad_csum,
1105                                      "%s",
1106                                      (printbuf_reset(&buf),
1107                                       bch2_csum_err_msg(&buf, BSET_CSUM_TYPE(i), bne->csum, csum),
1108                                       buf.buf));
1109
1110                         ret = bset_encrypt(c, i, b->written << 9);
1111                         if (bch2_fs_fatal_err_on(ret, c,
1112                                         "decrypting btree node: %s", bch2_err_str(ret)))
1113                                 goto fsck_err;
1114
1115                         sectors = vstruct_sectors(bne, c->block_bits);
1116                 }
1117
1118                 b->version_ondisk = min(b->version_ondisk,
1119                                         le16_to_cpu(i->version));
1120
1121                 ret = validate_bset(c, ca, b, i, b->written, sectors,
1122                                     READ, have_retry, saw_error);
1123                 if (ret)
1124                         goto fsck_err;
1125
1126                 if (!b->written)
1127                         btree_node_set_format(b, b->data->format);
1128
1129                 ret = validate_bset_keys(c, b, i, READ, have_retry, saw_error);
1130                 if (ret)
1131                         goto fsck_err;
1132
1133                 SET_BSET_BIG_ENDIAN(i, CPU_BIG_ENDIAN);
1134
1135                 blacklisted = bch2_journal_seq_is_blacklisted(c,
1136                                         le64_to_cpu(i->journal_seq),
1137                                         true);
1138
1139                 btree_err_on(blacklisted && first,
1140                              -BCH_ERR_btree_node_read_err_fixable,
1141                              c, ca, b, i,
1142                              bset_blacklisted_journal_seq,
1143                              "first btree node bset has blacklisted journal seq (%llu)",
1144                              le64_to_cpu(i->journal_seq));
1145
1146                 btree_err_on(blacklisted && ptr_written,
1147                              -BCH_ERR_btree_node_read_err_fixable,
1148                              c, ca, b, i,
1149                              first_bset_blacklisted_journal_seq,
1150                              "found blacklisted bset (journal seq %llu) in btree node at offset %u-%u/%u",
1151                              le64_to_cpu(i->journal_seq),
1152                              b->written, b->written + sectors, ptr_written);
1153
1154                 b->written += sectors;
1155
1156                 if (blacklisted && !first)
1157                         continue;
1158
1159                 sort_iter_add(iter,
1160                               vstruct_idx(i, 0),
1161                               vstruct_last(i));
1162         }
1163
1164         if (ptr_written) {
1165                 btree_err_on(b->written < ptr_written,
1166                              -BCH_ERR_btree_node_read_err_want_retry,
1167                              c, ca, b, NULL,
1168                              btree_node_data_missing,
1169                              "btree node data missing: expected %u sectors, found %u",
1170                              ptr_written, b->written);
1171         } else {
1172                 for (bne = write_block(b);
1173                      bset_byte_offset(b, bne) < btree_buf_bytes(b);
1174                      bne = (void *) bne + block_bytes(c))
1175                         btree_err_on(bne->keys.seq == b->data->keys.seq &&
1176                                      !bch2_journal_seq_is_blacklisted(c,
1177                                                                       le64_to_cpu(bne->keys.journal_seq),
1178                                                                       true),
1179                                      -BCH_ERR_btree_node_read_err_want_retry,
1180                                      c, ca, b, NULL,
1181                                      btree_node_bset_after_end,
1182                                      "found bset signature after last bset");
1183         }
1184
1185         sorted = btree_bounce_alloc(c, btree_buf_bytes(b), &used_mempool);
1186         sorted->keys.u64s = 0;
1187
1188         set_btree_bset(b, b->set, &b->data->keys);
1189
1190         b->nr = bch2_key_sort_fix_overlapping(c, &sorted->keys, iter);
1191
1192         u64s = le16_to_cpu(sorted->keys.u64s);
1193         *sorted = *b->data;
1194         sorted->keys.u64s = cpu_to_le16(u64s);
1195         swap(sorted, b->data);
1196         set_btree_bset(b, b->set, &b->data->keys);
1197         b->nsets = 1;
1198
1199         BUG_ON(b->nr.live_u64s != u64s);
1200
1201         btree_bounce_free(c, btree_buf_bytes(b), used_mempool, sorted);
1202
1203         if (updated_range)
1204                 bch2_btree_node_drop_keys_outside_node(b);
1205
1206         i = &b->data->keys;
1207         for (k = i->start; k != vstruct_last(i);) {
1208                 struct bkey tmp;
1209                 struct bkey_s u = __bkey_disassemble(b, k, &tmp);
1210
1211                 printbuf_reset(&buf);
1212
1213                 if (bch2_bkey_val_invalid(c, u.s_c, READ, &buf) ||
1214                     (bch2_inject_invalid_keys &&
1215                      !bversion_cmp(u.k->version, MAX_VERSION))) {
1216                         printbuf_reset(&buf);
1217
1218                         prt_printf(&buf, "invalid bkey: ");
1219                         bch2_bkey_val_invalid(c, u.s_c, READ, &buf);
1220                         prt_printf(&buf, "\n  ");
1221                         bch2_bkey_val_to_text(&buf, c, u.s_c);
1222
1223                         btree_err(-BCH_ERR_btree_node_read_err_fixable,
1224                                   c, NULL, b, i,
1225                                   btree_node_bad_bkey,
1226                                   "%s", buf.buf);
1227
1228                         btree_keys_account_key_drop(&b->nr, 0, k);
1229
1230                         i->u64s = cpu_to_le16(le16_to_cpu(i->u64s) - k->u64s);
1231                         memmove_u64s_down(k, bkey_p_next(k),
1232                                           (u64 *) vstruct_end(i) - (u64 *) k);
1233                         set_btree_bset_end(b, b->set);
1234                         continue;
1235                 }
1236
1237                 if (u.k->type == KEY_TYPE_btree_ptr_v2) {
1238                         struct bkey_s_btree_ptr_v2 bp = bkey_s_to_btree_ptr_v2(u);
1239
1240                         bp.v->mem_ptr = 0;
1241                 }
1242
1243                 k = bkey_p_next(k);
1244         }
1245
1246         bch2_bset_build_aux_tree(b, b->set, false);
1247
1248         set_needs_whiteout(btree_bset_first(b), true);
1249
1250         btree_node_reset_sib_u64s(b);
1251
1252         bkey_for_each_ptr(bch2_bkey_ptrs(bkey_i_to_s(&b->key)), ptr) {
1253                 struct bch_dev *ca2 = bch_dev_bkey_exists(c, ptr->dev);
1254
1255                 if (ca2->mi.state != BCH_MEMBER_STATE_rw)
1256                         set_btree_node_need_rewrite(b);
1257         }
1258
1259         if (!ptr_written)
1260                 set_btree_node_need_rewrite(b);
1261 out:
1262         mempool_free(iter, &c->fill_iter);
1263         printbuf_exit(&buf);
1264         bch2_time_stats_update(&c->times[BCH_TIME_btree_node_read_done], start_time);
1265         return retry_read;
1266 fsck_err:
1267         if (ret == -BCH_ERR_btree_node_read_err_want_retry ||
1268             ret == -BCH_ERR_btree_node_read_err_must_retry) {
1269                 retry_read = 1;
1270         } else {
1271                 set_btree_node_read_error(b);
1272                 bch2_btree_lost_data(c, b->c.btree_id);
1273         }
1274         goto out;
1275 }
1276
1277 static void btree_node_read_work(struct work_struct *work)
1278 {
1279         struct btree_read_bio *rb =
1280                 container_of(work, struct btree_read_bio, work);
1281         struct bch_fs *c        = rb->c;
1282         struct btree *b         = rb->b;
1283         struct bch_dev *ca      = bch_dev_bkey_exists(c, rb->pick.ptr.dev);
1284         struct bio *bio         = &rb->bio;
1285         struct bch_io_failures failed = { .nr = 0 };
1286         struct printbuf buf = PRINTBUF;
1287         bool saw_error = false;
1288         bool retry = false;
1289         bool can_retry;
1290
1291         goto start;
1292         while (1) {
1293                 retry = true;
1294                 bch_info(c, "retrying read");
1295                 ca = bch_dev_bkey_exists(c, rb->pick.ptr.dev);
1296                 rb->have_ioref          = bch2_dev_get_ioref(ca, READ);
1297                 bio_reset(bio, NULL, REQ_OP_READ|REQ_SYNC|REQ_META);
1298                 bio->bi_iter.bi_sector  = rb->pick.ptr.offset;
1299                 bio->bi_iter.bi_size    = btree_buf_bytes(b);
1300
1301                 if (rb->have_ioref) {
1302                         bio_set_dev(bio, ca->disk_sb.bdev);
1303                         submit_bio_wait(bio);
1304                 } else {
1305                         bio->bi_status = BLK_STS_REMOVED;
1306                 }
1307 start:
1308                 printbuf_reset(&buf);
1309                 bch2_btree_pos_to_text(&buf, c, b);
1310                 bch2_dev_io_err_on(bio->bi_status, ca, BCH_MEMBER_ERROR_read,
1311                                    "btree read error %s for %s",
1312                                    bch2_blk_status_to_str(bio->bi_status), buf.buf);
1313                 if (rb->have_ioref)
1314                         percpu_ref_put(&ca->io_ref);
1315                 rb->have_ioref = false;
1316
1317                 bch2_mark_io_failure(&failed, &rb->pick);
1318
1319                 can_retry = bch2_bkey_pick_read_device(c,
1320                                 bkey_i_to_s_c(&b->key),
1321                                 &failed, &rb->pick) > 0;
1322
1323                 if (!bio->bi_status &&
1324                     !bch2_btree_node_read_done(c, ca, b, can_retry, &saw_error)) {
1325                         if (retry)
1326                                 bch_info(c, "retry success");
1327                         break;
1328                 }
1329
1330                 saw_error = true;
1331
1332                 if (!can_retry) {
1333                         set_btree_node_read_error(b);
1334                         bch2_btree_lost_data(c, b->c.btree_id);
1335                         break;
1336                 }
1337         }
1338
1339         bch2_time_stats_update(&c->times[BCH_TIME_btree_node_read],
1340                                rb->start_time);
1341         bio_put(&rb->bio);
1342
1343         if (saw_error &&
1344             !btree_node_read_error(b) &&
1345             c->curr_recovery_pass != BCH_RECOVERY_PASS_scan_for_btree_nodes) {
1346                 printbuf_reset(&buf);
1347                 bch2_bpos_to_text(&buf, b->key.k.p);
1348                 bch_err_ratelimited(c, "%s: rewriting btree node at btree=%s level=%u %s due to error",
1349                          __func__, bch2_btree_id_str(b->c.btree_id), b->c.level, buf.buf);
1350
1351                 bch2_btree_node_rewrite_async(c, b);
1352         }
1353
1354         printbuf_exit(&buf);
1355         clear_btree_node_read_in_flight(b);
1356         wake_up_bit(&b->flags, BTREE_NODE_read_in_flight);
1357 }
1358
1359 static void btree_node_read_endio(struct bio *bio)
1360 {
1361         struct btree_read_bio *rb =
1362                 container_of(bio, struct btree_read_bio, bio);
1363         struct bch_fs *c        = rb->c;
1364
1365         if (rb->have_ioref) {
1366                 struct bch_dev *ca = bch_dev_bkey_exists(c, rb->pick.ptr.dev);
1367
1368                 bch2_latency_acct(ca, rb->start_time, READ);
1369         }
1370
1371         queue_work(c->io_complete_wq, &rb->work);
1372 }
1373
1374 struct btree_node_read_all {
1375         struct closure          cl;
1376         struct bch_fs           *c;
1377         struct btree            *b;
1378         unsigned                nr;
1379         void                    *buf[BCH_REPLICAS_MAX];
1380         struct bio              *bio[BCH_REPLICAS_MAX];
1381         blk_status_t            err[BCH_REPLICAS_MAX];
1382 };
1383
1384 static unsigned btree_node_sectors_written(struct bch_fs *c, void *data)
1385 {
1386         struct btree_node *bn = data;
1387         struct btree_node_entry *bne;
1388         unsigned offset = 0;
1389
1390         if (le64_to_cpu(bn->magic) !=  bset_magic(c))
1391                 return 0;
1392
1393         while (offset < btree_sectors(c)) {
1394                 if (!offset) {
1395                         offset += vstruct_sectors(bn, c->block_bits);
1396                 } else {
1397                         bne = data + (offset << 9);
1398                         if (bne->keys.seq != bn->keys.seq)
1399                                 break;
1400                         offset += vstruct_sectors(bne, c->block_bits);
1401                 }
1402         }
1403
1404         return offset;
1405 }
1406
1407 static bool btree_node_has_extra_bsets(struct bch_fs *c, unsigned offset, void *data)
1408 {
1409         struct btree_node *bn = data;
1410         struct btree_node_entry *bne;
1411
1412         if (!offset)
1413                 return false;
1414
1415         while (offset < btree_sectors(c)) {
1416                 bne = data + (offset << 9);
1417                 if (bne->keys.seq == bn->keys.seq)
1418                         return true;
1419                 offset++;
1420         }
1421
1422         return false;
1423         return offset;
1424 }
1425
1426 static CLOSURE_CALLBACK(btree_node_read_all_replicas_done)
1427 {
1428         closure_type(ra, struct btree_node_read_all, cl);
1429         struct bch_fs *c = ra->c;
1430         struct btree *b = ra->b;
1431         struct printbuf buf = PRINTBUF;
1432         bool dump_bset_maps = false;
1433         bool have_retry = false;
1434         int ret = 0, best = -1, write = READ;
1435         unsigned i, written = 0, written2 = 0;
1436         __le64 seq = b->key.k.type == KEY_TYPE_btree_ptr_v2
1437                 ? bkey_i_to_btree_ptr_v2(&b->key)->v.seq : 0;
1438         bool _saw_error = false, *saw_error = &_saw_error;
1439
1440         for (i = 0; i < ra->nr; i++) {
1441                 struct btree_node *bn = ra->buf[i];
1442
1443                 if (ra->err[i])
1444                         continue;
1445
1446                 if (le64_to_cpu(bn->magic) != bset_magic(c) ||
1447                     (seq && seq != bn->keys.seq))
1448                         continue;
1449
1450                 if (best < 0) {
1451                         best = i;
1452                         written = btree_node_sectors_written(c, bn);
1453                         continue;
1454                 }
1455
1456                 written2 = btree_node_sectors_written(c, ra->buf[i]);
1457                 if (btree_err_on(written2 != written, -BCH_ERR_btree_node_read_err_fixable,
1458                                  c, NULL, b, NULL,
1459                                  btree_node_replicas_sectors_written_mismatch,
1460                                  "btree node sectors written mismatch: %u != %u",
1461                                  written, written2) ||
1462                     btree_err_on(btree_node_has_extra_bsets(c, written2, ra->buf[i]),
1463                                  -BCH_ERR_btree_node_read_err_fixable,
1464                                  c, NULL, b, NULL,
1465                                  btree_node_bset_after_end,
1466                                  "found bset signature after last bset") ||
1467                     btree_err_on(memcmp(ra->buf[best], ra->buf[i], written << 9),
1468                                  -BCH_ERR_btree_node_read_err_fixable,
1469                                  c, NULL, b, NULL,
1470                                  btree_node_replicas_data_mismatch,
1471                                  "btree node replicas content mismatch"))
1472                         dump_bset_maps = true;
1473
1474                 if (written2 > written) {
1475                         written = written2;
1476                         best = i;
1477                 }
1478         }
1479 fsck_err:
1480         if (dump_bset_maps) {
1481                 for (i = 0; i < ra->nr; i++) {
1482                         struct btree_node *bn = ra->buf[i];
1483                         struct btree_node_entry *bne = NULL;
1484                         unsigned offset = 0, sectors;
1485                         bool gap = false;
1486
1487                         if (ra->err[i])
1488                                 continue;
1489
1490                         printbuf_reset(&buf);
1491
1492                         while (offset < btree_sectors(c)) {
1493                                 if (!offset) {
1494                                         sectors = vstruct_sectors(bn, c->block_bits);
1495                                 } else {
1496                                         bne = ra->buf[i] + (offset << 9);
1497                                         if (bne->keys.seq != bn->keys.seq)
1498                                                 break;
1499                                         sectors = vstruct_sectors(bne, c->block_bits);
1500                                 }
1501
1502                                 prt_printf(&buf, " %u-%u", offset, offset + sectors);
1503                                 if (bne && bch2_journal_seq_is_blacklisted(c,
1504                                                         le64_to_cpu(bne->keys.journal_seq), false))
1505                                         prt_printf(&buf, "*");
1506                                 offset += sectors;
1507                         }
1508
1509                         while (offset < btree_sectors(c)) {
1510                                 bne = ra->buf[i] + (offset << 9);
1511                                 if (bne->keys.seq == bn->keys.seq) {
1512                                         if (!gap)
1513                                                 prt_printf(&buf, " GAP");
1514                                         gap = true;
1515
1516                                         sectors = vstruct_sectors(bne, c->block_bits);
1517                                         prt_printf(&buf, " %u-%u", offset, offset + sectors);
1518                                         if (bch2_journal_seq_is_blacklisted(c,
1519                                                         le64_to_cpu(bne->keys.journal_seq), false))
1520                                                 prt_printf(&buf, "*");
1521                                 }
1522                                 offset++;
1523                         }
1524
1525                         bch_err(c, "replica %u:%s", i, buf.buf);
1526                 }
1527         }
1528
1529         if (best >= 0) {
1530                 memcpy(b->data, ra->buf[best], btree_buf_bytes(b));
1531                 ret = bch2_btree_node_read_done(c, NULL, b, false, saw_error);
1532         } else {
1533                 ret = -1;
1534         }
1535
1536         if (ret) {
1537                 set_btree_node_read_error(b);
1538                 bch2_btree_lost_data(c, b->c.btree_id);
1539         } else if (*saw_error)
1540                 bch2_btree_node_rewrite_async(c, b);
1541
1542         for (i = 0; i < ra->nr; i++) {
1543                 mempool_free(ra->buf[i], &c->btree_bounce_pool);
1544                 bio_put(ra->bio[i]);
1545         }
1546
1547         closure_debug_destroy(&ra->cl);
1548         kfree(ra);
1549         printbuf_exit(&buf);
1550
1551         clear_btree_node_read_in_flight(b);
1552         wake_up_bit(&b->flags, BTREE_NODE_read_in_flight);
1553 }
1554
1555 static void btree_node_read_all_replicas_endio(struct bio *bio)
1556 {
1557         struct btree_read_bio *rb =
1558                 container_of(bio, struct btree_read_bio, bio);
1559         struct bch_fs *c        = rb->c;
1560         struct btree_node_read_all *ra = rb->ra;
1561
1562         if (rb->have_ioref) {
1563                 struct bch_dev *ca = bch_dev_bkey_exists(c, rb->pick.ptr.dev);
1564
1565                 bch2_latency_acct(ca, rb->start_time, READ);
1566         }
1567
1568         ra->err[rb->idx] = bio->bi_status;
1569         closure_put(&ra->cl);
1570 }
1571
1572 /*
1573  * XXX This allocates multiple times from the same mempools, and can deadlock
1574  * under sufficient memory pressure (but is only a debug path)
1575  */
1576 static int btree_node_read_all_replicas(struct bch_fs *c, struct btree *b, bool sync)
1577 {
1578         struct bkey_s_c k = bkey_i_to_s_c(&b->key);
1579         struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k);
1580         const union bch_extent_entry *entry;
1581         struct extent_ptr_decoded pick;
1582         struct btree_node_read_all *ra;
1583         unsigned i;
1584
1585         ra = kzalloc(sizeof(*ra), GFP_NOFS);
1586         if (!ra)
1587                 return -BCH_ERR_ENOMEM_btree_node_read_all_replicas;
1588
1589         closure_init(&ra->cl, NULL);
1590         ra->c   = c;
1591         ra->b   = b;
1592         ra->nr  = bch2_bkey_nr_ptrs(k);
1593
1594         for (i = 0; i < ra->nr; i++) {
1595                 ra->buf[i] = mempool_alloc(&c->btree_bounce_pool, GFP_NOFS);
1596                 ra->bio[i] = bio_alloc_bioset(NULL,
1597                                               buf_pages(ra->buf[i], btree_buf_bytes(b)),
1598                                               REQ_OP_READ|REQ_SYNC|REQ_META,
1599                                               GFP_NOFS,
1600                                               &c->btree_bio);
1601         }
1602
1603         i = 0;
1604         bkey_for_each_ptr_decode(k.k, ptrs, pick, entry) {
1605                 struct bch_dev *ca = bch_dev_bkey_exists(c, pick.ptr.dev);
1606                 struct btree_read_bio *rb =
1607                         container_of(ra->bio[i], struct btree_read_bio, bio);
1608                 rb->c                   = c;
1609                 rb->b                   = b;
1610                 rb->ra                  = ra;
1611                 rb->start_time          = local_clock();
1612                 rb->have_ioref          = bch2_dev_get_ioref(ca, READ);
1613                 rb->idx                 = i;
1614                 rb->pick                = pick;
1615                 rb->bio.bi_iter.bi_sector = pick.ptr.offset;
1616                 rb->bio.bi_end_io       = btree_node_read_all_replicas_endio;
1617                 bch2_bio_map(&rb->bio, ra->buf[i], btree_buf_bytes(b));
1618
1619                 if (rb->have_ioref) {
1620                         this_cpu_add(ca->io_done->sectors[READ][BCH_DATA_btree],
1621                                      bio_sectors(&rb->bio));
1622                         bio_set_dev(&rb->bio, ca->disk_sb.bdev);
1623
1624                         closure_get(&ra->cl);
1625                         submit_bio(&rb->bio);
1626                 } else {
1627                         ra->err[i] = BLK_STS_REMOVED;
1628                 }
1629
1630                 i++;
1631         }
1632
1633         if (sync) {
1634                 closure_sync(&ra->cl);
1635                 btree_node_read_all_replicas_done(&ra->cl.work);
1636         } else {
1637                 continue_at(&ra->cl, btree_node_read_all_replicas_done,
1638                             c->io_complete_wq);
1639         }
1640
1641         return 0;
1642 }
1643
1644 void bch2_btree_node_read(struct btree_trans *trans, struct btree *b,
1645                           bool sync)
1646 {
1647         struct bch_fs *c = trans->c;
1648         struct extent_ptr_decoded pick;
1649         struct btree_read_bio *rb;
1650         struct bch_dev *ca;
1651         struct bio *bio;
1652         int ret;
1653
1654         trace_and_count(c, btree_node_read, trans, b);
1655
1656         if (bch2_verify_all_btree_replicas &&
1657             !btree_node_read_all_replicas(c, b, sync))
1658                 return;
1659
1660         ret = bch2_bkey_pick_read_device(c, bkey_i_to_s_c(&b->key),
1661                                          NULL, &pick);
1662
1663         if (ret <= 0) {
1664                 struct printbuf buf = PRINTBUF;
1665
1666                 prt_str(&buf, "btree node read error: no device to read from\n at ");
1667                 bch2_btree_pos_to_text(&buf, c, b);
1668                 bch_err_ratelimited(c, "%s", buf.buf);
1669
1670                 if (c->recovery_passes_explicit & BIT_ULL(BCH_RECOVERY_PASS_check_topology) &&
1671                     c->curr_recovery_pass > BCH_RECOVERY_PASS_check_topology)
1672                         bch2_fatal_error(c);
1673
1674                 set_btree_node_read_error(b);
1675                 bch2_btree_lost_data(c, b->c.btree_id);
1676                 clear_btree_node_read_in_flight(b);
1677                 wake_up_bit(&b->flags, BTREE_NODE_read_in_flight);
1678                 printbuf_exit(&buf);
1679                 return;
1680         }
1681
1682         ca = bch_dev_bkey_exists(c, pick.ptr.dev);
1683
1684         bio = bio_alloc_bioset(NULL,
1685                                buf_pages(b->data, btree_buf_bytes(b)),
1686                                REQ_OP_READ|REQ_SYNC|REQ_META,
1687                                GFP_NOFS,
1688                                &c->btree_bio);
1689         rb = container_of(bio, struct btree_read_bio, bio);
1690         rb->c                   = c;
1691         rb->b                   = b;
1692         rb->ra                  = NULL;
1693         rb->start_time          = local_clock();
1694         rb->have_ioref          = bch2_dev_get_ioref(ca, READ);
1695         rb->pick                = pick;
1696         INIT_WORK(&rb->work, btree_node_read_work);
1697         bio->bi_iter.bi_sector  = pick.ptr.offset;
1698         bio->bi_end_io          = btree_node_read_endio;
1699         bch2_bio_map(bio, b->data, btree_buf_bytes(b));
1700
1701         if (rb->have_ioref) {
1702                 this_cpu_add(ca->io_done->sectors[READ][BCH_DATA_btree],
1703                              bio_sectors(bio));
1704                 bio_set_dev(bio, ca->disk_sb.bdev);
1705
1706                 if (sync) {
1707                         submit_bio_wait(bio);
1708                         bch2_latency_acct(ca, rb->start_time, READ);
1709                         btree_node_read_work(&rb->work);
1710                 } else {
1711                         submit_bio(bio);
1712                 }
1713         } else {
1714                 bio->bi_status = BLK_STS_REMOVED;
1715
1716                 if (sync)
1717                         btree_node_read_work(&rb->work);
1718                 else
1719                         queue_work(c->io_complete_wq, &rb->work);
1720         }
1721 }
1722
1723 static int __bch2_btree_root_read(struct btree_trans *trans, enum btree_id id,
1724                                   const struct bkey_i *k, unsigned level)
1725 {
1726         struct bch_fs *c = trans->c;
1727         struct closure cl;
1728         struct btree *b;
1729         int ret;
1730
1731         closure_init_stack(&cl);
1732
1733         do {
1734                 ret = bch2_btree_cache_cannibalize_lock(trans, &cl);
1735                 closure_sync(&cl);
1736         } while (ret);
1737
1738         b = bch2_btree_node_mem_alloc(trans, level != 0);
1739         bch2_btree_cache_cannibalize_unlock(trans);
1740
1741         BUG_ON(IS_ERR(b));
1742
1743         bkey_copy(&b->key, k);
1744         BUG_ON(bch2_btree_node_hash_insert(&c->btree_cache, b, level, id));
1745
1746         set_btree_node_read_in_flight(b);
1747
1748         bch2_btree_node_read(trans, b, true);
1749
1750         if (btree_node_read_error(b)) {
1751                 bch2_btree_node_hash_remove(&c->btree_cache, b);
1752
1753                 mutex_lock(&c->btree_cache.lock);
1754                 list_move(&b->list, &c->btree_cache.freeable);
1755                 mutex_unlock(&c->btree_cache.lock);
1756
1757                 ret = -BCH_ERR_btree_node_read_error;
1758                 goto err;
1759         }
1760
1761         bch2_btree_set_root_for_read(c, b);
1762 err:
1763         six_unlock_write(&b->c.lock);
1764         six_unlock_intent(&b->c.lock);
1765
1766         return ret;
1767 }
1768
1769 int bch2_btree_root_read(struct bch_fs *c, enum btree_id id,
1770                         const struct bkey_i *k, unsigned level)
1771 {
1772         return bch2_trans_run(c, __bch2_btree_root_read(trans, id, k, level));
1773 }
1774
1775 static void bch2_btree_complete_write(struct bch_fs *c, struct btree *b,
1776                                       struct btree_write *w)
1777 {
1778         unsigned long old, new, v = READ_ONCE(b->will_make_reachable);
1779
1780         do {
1781                 old = new = v;
1782                 if (!(old & 1))
1783                         break;
1784
1785                 new &= ~1UL;
1786         } while ((v = cmpxchg(&b->will_make_reachable, old, new)) != old);
1787
1788         if (old & 1)
1789                 closure_put(&((struct btree_update *) new)->cl);
1790
1791         bch2_journal_pin_drop(&c->journal, &w->journal);
1792 }
1793
1794 static void __btree_node_write_done(struct bch_fs *c, struct btree *b)
1795 {
1796         struct btree_write *w = btree_prev_write(b);
1797         unsigned long old, new, v;
1798         unsigned type = 0;
1799
1800         bch2_btree_complete_write(c, b, w);
1801
1802         v = READ_ONCE(b->flags);
1803         do {
1804                 old = new = v;
1805
1806                 if ((old & (1U << BTREE_NODE_dirty)) &&
1807                     (old & (1U << BTREE_NODE_need_write)) &&
1808                     !(old & (1U << BTREE_NODE_never_write)) &&
1809                     !(old & (1U << BTREE_NODE_write_blocked)) &&
1810                     !(old & (1U << BTREE_NODE_will_make_reachable))) {
1811                         new &= ~(1U << BTREE_NODE_dirty);
1812                         new &= ~(1U << BTREE_NODE_need_write);
1813                         new |=  (1U << BTREE_NODE_write_in_flight);
1814                         new |=  (1U << BTREE_NODE_write_in_flight_inner);
1815                         new |=  (1U << BTREE_NODE_just_written);
1816                         new ^=  (1U << BTREE_NODE_write_idx);
1817
1818                         type = new & BTREE_WRITE_TYPE_MASK;
1819                         new &= ~BTREE_WRITE_TYPE_MASK;
1820                 } else {
1821                         new &= ~(1U << BTREE_NODE_write_in_flight);
1822                         new &= ~(1U << BTREE_NODE_write_in_flight_inner);
1823                 }
1824         } while ((v = cmpxchg(&b->flags, old, new)) != old);
1825
1826         if (new & (1U << BTREE_NODE_write_in_flight))
1827                 __bch2_btree_node_write(c, b, BTREE_WRITE_ALREADY_STARTED|type);
1828         else
1829                 wake_up_bit(&b->flags, BTREE_NODE_write_in_flight);
1830 }
1831
1832 static void btree_node_write_done(struct bch_fs *c, struct btree *b)
1833 {
1834         struct btree_trans *trans = bch2_trans_get(c);
1835
1836         btree_node_lock_nopath_nofail(trans, &b->c, SIX_LOCK_read);
1837         __btree_node_write_done(c, b);
1838         six_unlock_read(&b->c.lock);
1839
1840         bch2_trans_put(trans);
1841 }
1842
1843 static void btree_node_write_work(struct work_struct *work)
1844 {
1845         struct btree_write_bio *wbio =
1846                 container_of(work, struct btree_write_bio, work);
1847         struct bch_fs *c        = wbio->wbio.c;
1848         struct btree *b         = wbio->wbio.bio.bi_private;
1849         struct bch_extent_ptr *ptr;
1850         int ret = 0;
1851
1852         btree_bounce_free(c,
1853                 wbio->data_bytes,
1854                 wbio->wbio.used_mempool,
1855                 wbio->data);
1856
1857         bch2_bkey_drop_ptrs(bkey_i_to_s(&wbio->key), ptr,
1858                 bch2_dev_list_has_dev(wbio->wbio.failed, ptr->dev));
1859
1860         if (!bch2_bkey_nr_ptrs(bkey_i_to_s_c(&wbio->key))) {
1861                 ret = -BCH_ERR_btree_node_write_all_failed;
1862                 goto err;
1863         }
1864
1865         if (wbio->wbio.first_btree_write) {
1866                 if (wbio->wbio.failed.nr) {
1867
1868                 }
1869         } else {
1870                 ret = bch2_trans_do(c, NULL, NULL, 0,
1871                         bch2_btree_node_update_key_get_iter(trans, b, &wbio->key,
1872                                         BCH_WATERMARK_interior_updates|
1873                                         BCH_TRANS_COMMIT_journal_reclaim|
1874                                         BCH_TRANS_COMMIT_no_enospc|
1875                                         BCH_TRANS_COMMIT_no_check_rw,
1876                                         !wbio->wbio.failed.nr));
1877                 if (ret)
1878                         goto err;
1879         }
1880 out:
1881         bio_put(&wbio->wbio.bio);
1882         btree_node_write_done(c, b);
1883         return;
1884 err:
1885         set_btree_node_noevict(b);
1886         bch2_fs_fatal_err_on(!bch2_err_matches(ret, EROFS), c,
1887                              "writing btree node: %s", bch2_err_str(ret));
1888         goto out;
1889 }
1890
1891 static void btree_node_write_endio(struct bio *bio)
1892 {
1893         struct bch_write_bio *wbio      = to_wbio(bio);
1894         struct bch_write_bio *parent    = wbio->split ? wbio->parent : NULL;
1895         struct bch_write_bio *orig      = parent ?: wbio;
1896         struct btree_write_bio *wb      = container_of(orig, struct btree_write_bio, wbio);
1897         struct bch_fs *c                = wbio->c;
1898         struct btree *b                 = wbio->bio.bi_private;
1899         struct bch_dev *ca              = bch_dev_bkey_exists(c, wbio->dev);
1900         unsigned long flags;
1901
1902         if (wbio->have_ioref)
1903                 bch2_latency_acct(ca, wbio->submit_time, WRITE);
1904
1905         if (bch2_dev_io_err_on(bio->bi_status, ca, BCH_MEMBER_ERROR_write,
1906                                "btree write error: %s",
1907                                bch2_blk_status_to_str(bio->bi_status)) ||
1908             bch2_meta_write_fault("btree")) {
1909                 spin_lock_irqsave(&c->btree_write_error_lock, flags);
1910                 bch2_dev_list_add_dev(&orig->failed, wbio->dev);
1911                 spin_unlock_irqrestore(&c->btree_write_error_lock, flags);
1912         }
1913
1914         if (wbio->have_ioref)
1915                 percpu_ref_put(&ca->io_ref);
1916
1917         if (parent) {
1918                 bio_put(bio);
1919                 bio_endio(&parent->bio);
1920                 return;
1921         }
1922
1923         clear_btree_node_write_in_flight_inner(b);
1924         wake_up_bit(&b->flags, BTREE_NODE_write_in_flight_inner);
1925         INIT_WORK(&wb->work, btree_node_write_work);
1926         queue_work(c->btree_io_complete_wq, &wb->work);
1927 }
1928
1929 static int validate_bset_for_write(struct bch_fs *c, struct btree *b,
1930                                    struct bset *i, unsigned sectors)
1931 {
1932         struct printbuf buf = PRINTBUF;
1933         bool saw_error;
1934         int ret;
1935
1936         ret = bch2_bkey_invalid(c, bkey_i_to_s_c(&b->key),
1937                                 BKEY_TYPE_btree, WRITE, &buf);
1938
1939         if (ret)
1940                 bch2_fs_inconsistent(c, "invalid btree node key before write: %s", buf.buf);
1941         printbuf_exit(&buf);
1942         if (ret)
1943                 return ret;
1944
1945         ret = validate_bset_keys(c, b, i, WRITE, false, &saw_error) ?:
1946                 validate_bset(c, NULL, b, i, b->written, sectors, WRITE, false, &saw_error);
1947         if (ret) {
1948                 bch2_inconsistent_error(c);
1949                 dump_stack();
1950         }
1951
1952         return ret;
1953 }
1954
1955 static void btree_write_submit(struct work_struct *work)
1956 {
1957         struct btree_write_bio *wbio = container_of(work, struct btree_write_bio, work);
1958         BKEY_PADDED_ONSTACK(k, BKEY_BTREE_PTR_VAL_U64s_MAX) tmp;
1959
1960         bkey_copy(&tmp.k, &wbio->key);
1961
1962         bkey_for_each_ptr(bch2_bkey_ptrs(bkey_i_to_s(&tmp.k)), ptr)
1963                 ptr->offset += wbio->sector_offset;
1964
1965         bch2_submit_wbio_replicas(&wbio->wbio, wbio->wbio.c, BCH_DATA_btree,
1966                                   &tmp.k, false);
1967 }
1968
1969 void __bch2_btree_node_write(struct bch_fs *c, struct btree *b, unsigned flags)
1970 {
1971         struct btree_write_bio *wbio;
1972         struct bset_tree *t;
1973         struct bset *i;
1974         struct btree_node *bn = NULL;
1975         struct btree_node_entry *bne = NULL;
1976         struct sort_iter_stack sort_iter;
1977         struct nonce nonce;
1978         unsigned bytes_to_write, sectors_to_write, bytes, u64s;
1979         u64 seq = 0;
1980         bool used_mempool;
1981         unsigned long old, new;
1982         bool validate_before_checksum = false;
1983         enum btree_write_type type = flags & BTREE_WRITE_TYPE_MASK;
1984         void *data;
1985         int ret;
1986
1987         if (flags & BTREE_WRITE_ALREADY_STARTED)
1988                 goto do_write;
1989
1990         /*
1991          * We may only have a read lock on the btree node - the dirty bit is our
1992          * "lock" against racing with other threads that may be trying to start
1993          * a write, we do a write iff we clear the dirty bit. Since setting the
1994          * dirty bit requires a write lock, we can't race with other threads
1995          * redirtying it:
1996          */
1997         do {
1998                 old = new = READ_ONCE(b->flags);
1999
2000                 if (!(old & (1 << BTREE_NODE_dirty)))
2001                         return;
2002
2003                 if ((flags & BTREE_WRITE_ONLY_IF_NEED) &&
2004                     !(old & (1 << BTREE_NODE_need_write)))
2005                         return;
2006
2007                 if (old &
2008                     ((1 << BTREE_NODE_never_write)|
2009                      (1 << BTREE_NODE_write_blocked)))
2010                         return;
2011
2012                 if (b->written &&
2013                     (old & (1 << BTREE_NODE_will_make_reachable)))
2014                         return;
2015
2016                 if (old & (1 << BTREE_NODE_write_in_flight))
2017                         return;
2018
2019                 if (flags & BTREE_WRITE_ONLY_IF_NEED)
2020                         type = new & BTREE_WRITE_TYPE_MASK;
2021                 new &= ~BTREE_WRITE_TYPE_MASK;
2022
2023                 new &= ~(1 << BTREE_NODE_dirty);
2024                 new &= ~(1 << BTREE_NODE_need_write);
2025                 new |=  (1 << BTREE_NODE_write_in_flight);
2026                 new |=  (1 << BTREE_NODE_write_in_flight_inner);
2027                 new |=  (1 << BTREE_NODE_just_written);
2028                 new ^=  (1 << BTREE_NODE_write_idx);
2029         } while (cmpxchg_acquire(&b->flags, old, new) != old);
2030
2031         if (new & (1U << BTREE_NODE_need_write))
2032                 return;
2033 do_write:
2034         BUG_ON((type == BTREE_WRITE_initial) != (b->written == 0));
2035
2036         atomic_dec(&c->btree_cache.dirty);
2037
2038         BUG_ON(btree_node_fake(b));
2039         BUG_ON((b->will_make_reachable != 0) != !b->written);
2040
2041         BUG_ON(b->written >= btree_sectors(c));
2042         BUG_ON(b->written & (block_sectors(c) - 1));
2043         BUG_ON(bset_written(b, btree_bset_last(b)));
2044         BUG_ON(le64_to_cpu(b->data->magic) != bset_magic(c));
2045         BUG_ON(memcmp(&b->data->format, &b->format, sizeof(b->format)));
2046
2047         bch2_sort_whiteouts(c, b);
2048
2049         sort_iter_stack_init(&sort_iter, b);
2050
2051         bytes = !b->written
2052                 ? sizeof(struct btree_node)
2053                 : sizeof(struct btree_node_entry);
2054
2055         bytes += b->whiteout_u64s * sizeof(u64);
2056
2057         for_each_bset(b, t) {
2058                 i = bset(b, t);
2059
2060                 if (bset_written(b, i))
2061                         continue;
2062
2063                 bytes += le16_to_cpu(i->u64s) * sizeof(u64);
2064                 sort_iter_add(&sort_iter.iter,
2065                               btree_bkey_first(b, t),
2066                               btree_bkey_last(b, t));
2067                 seq = max(seq, le64_to_cpu(i->journal_seq));
2068         }
2069
2070         BUG_ON(b->written && !seq);
2071
2072         /* bch2_varint_decode may read up to 7 bytes past the end of the buffer: */
2073         bytes += 8;
2074
2075         /* buffer must be a multiple of the block size */
2076         bytes = round_up(bytes, block_bytes(c));
2077
2078         data = btree_bounce_alloc(c, bytes, &used_mempool);
2079
2080         if (!b->written) {
2081                 bn = data;
2082                 *bn = *b->data;
2083                 i = &bn->keys;
2084         } else {
2085                 bne = data;
2086                 bne->keys = b->data->keys;
2087                 i = &bne->keys;
2088         }
2089
2090         i->journal_seq  = cpu_to_le64(seq);
2091         i->u64s         = 0;
2092
2093         sort_iter_add(&sort_iter.iter,
2094                       unwritten_whiteouts_start(b),
2095                       unwritten_whiteouts_end(b));
2096         SET_BSET_SEPARATE_WHITEOUTS(i, false);
2097
2098         b->whiteout_u64s = 0;
2099
2100         u64s = bch2_sort_keys(i->start, &sort_iter.iter, false);
2101         le16_add_cpu(&i->u64s, u64s);
2102
2103         BUG_ON(!b->written && i->u64s != b->data->keys.u64s);
2104
2105         set_needs_whiteout(i, false);
2106
2107         /* do we have data to write? */
2108         if (b->written && !i->u64s)
2109                 goto nowrite;
2110
2111         bytes_to_write = vstruct_end(i) - data;
2112         sectors_to_write = round_up(bytes_to_write, block_bytes(c)) >> 9;
2113
2114         if (!b->written &&
2115             b->key.k.type == KEY_TYPE_btree_ptr_v2)
2116                 BUG_ON(btree_ptr_sectors_written(&b->key) != sectors_to_write);
2117
2118         memset(data + bytes_to_write, 0,
2119                (sectors_to_write << 9) - bytes_to_write);
2120
2121         BUG_ON(b->written + sectors_to_write > btree_sectors(c));
2122         BUG_ON(BSET_BIG_ENDIAN(i) != CPU_BIG_ENDIAN);
2123         BUG_ON(i->seq != b->data->keys.seq);
2124
2125         i->version = cpu_to_le16(c->sb.version);
2126         SET_BSET_OFFSET(i, b->written);
2127         SET_BSET_CSUM_TYPE(i, bch2_meta_checksum_type(c));
2128
2129         if (bch2_csum_type_is_encryption(BSET_CSUM_TYPE(i)))
2130                 validate_before_checksum = true;
2131
2132         /* validate_bset will be modifying: */
2133         if (le16_to_cpu(i->version) < bcachefs_metadata_version_current)
2134                 validate_before_checksum = true;
2135
2136         /* if we're going to be encrypting, check metadata validity first: */
2137         if (validate_before_checksum &&
2138             validate_bset_for_write(c, b, i, sectors_to_write))
2139                 goto err;
2140
2141         ret = bset_encrypt(c, i, b->written << 9);
2142         if (bch2_fs_fatal_err_on(ret, c,
2143                         "encrypting btree node: %s", bch2_err_str(ret)))
2144                 goto err;
2145
2146         nonce = btree_nonce(i, b->written << 9);
2147
2148         if (bn)
2149                 bn->csum = csum_vstruct(c, BSET_CSUM_TYPE(i), nonce, bn);
2150         else
2151                 bne->csum = csum_vstruct(c, BSET_CSUM_TYPE(i), nonce, bne);
2152
2153         /* if we're not encrypting, check metadata after checksumming: */
2154         if (!validate_before_checksum &&
2155             validate_bset_for_write(c, b, i, sectors_to_write))
2156                 goto err;
2157
2158         /*
2159          * We handle btree write errors by immediately halting the journal -
2160          * after we've done that, we can't issue any subsequent btree writes
2161          * because they might have pointers to new nodes that failed to write.
2162          *
2163          * Furthermore, there's no point in doing any more btree writes because
2164          * with the journal stopped, we're never going to update the journal to
2165          * reflect that those writes were done and the data flushed from the
2166          * journal:
2167          *
2168          * Also on journal error, the pending write may have updates that were
2169          * never journalled (interior nodes, see btree_update_nodes_written()) -
2170          * it's critical that we don't do the write in that case otherwise we
2171          * will have updates visible that weren't in the journal:
2172          *
2173          * Make sure to update b->written so bch2_btree_init_next() doesn't
2174          * break:
2175          */
2176         if (bch2_journal_error(&c->journal) ||
2177             c->opts.nochanges)
2178                 goto err;
2179
2180         trace_and_count(c, btree_node_write, b, bytes_to_write, sectors_to_write);
2181
2182         wbio = container_of(bio_alloc_bioset(NULL,
2183                                 buf_pages(data, sectors_to_write << 9),
2184                                 REQ_OP_WRITE|REQ_META,
2185                                 GFP_NOFS,
2186                                 &c->btree_bio),
2187                             struct btree_write_bio, wbio.bio);
2188         wbio_init(&wbio->wbio.bio);
2189         wbio->data                      = data;
2190         wbio->data_bytes                = bytes;
2191         wbio->sector_offset             = b->written;
2192         wbio->wbio.c                    = c;
2193         wbio->wbio.used_mempool         = used_mempool;
2194         wbio->wbio.first_btree_write    = !b->written;
2195         wbio->wbio.bio.bi_end_io        = btree_node_write_endio;
2196         wbio->wbio.bio.bi_private       = b;
2197
2198         bch2_bio_map(&wbio->wbio.bio, data, sectors_to_write << 9);
2199
2200         bkey_copy(&wbio->key, &b->key);
2201
2202         b->written += sectors_to_write;
2203
2204         if (wbio->key.k.type == KEY_TYPE_btree_ptr_v2)
2205                 bkey_i_to_btree_ptr_v2(&wbio->key)->v.sectors_written =
2206                         cpu_to_le16(b->written);
2207
2208         atomic64_inc(&c->btree_write_stats[type].nr);
2209         atomic64_add(bytes_to_write, &c->btree_write_stats[type].bytes);
2210
2211         INIT_WORK(&wbio->work, btree_write_submit);
2212         queue_work(c->io_complete_wq, &wbio->work);
2213         return;
2214 err:
2215         set_btree_node_noevict(b);
2216         b->written += sectors_to_write;
2217 nowrite:
2218         btree_bounce_free(c, bytes, used_mempool, data);
2219         __btree_node_write_done(c, b);
2220 }
2221
2222 /*
2223  * Work that must be done with write lock held:
2224  */
2225 bool bch2_btree_post_write_cleanup(struct bch_fs *c, struct btree *b)
2226 {
2227         bool invalidated_iter = false;
2228         struct btree_node_entry *bne;
2229         struct bset_tree *t;
2230
2231         if (!btree_node_just_written(b))
2232                 return false;
2233
2234         BUG_ON(b->whiteout_u64s);
2235
2236         clear_btree_node_just_written(b);
2237
2238         /*
2239          * Note: immediately after write, bset_written() doesn't work - the
2240          * amount of data we had to write after compaction might have been
2241          * smaller than the offset of the last bset.
2242          *
2243          * However, we know that all bsets have been written here, as long as
2244          * we're still holding the write lock:
2245          */
2246
2247         /*
2248          * XXX: decide if we really want to unconditionally sort down to a
2249          * single bset:
2250          */
2251         if (b->nsets > 1) {
2252                 btree_node_sort(c, b, 0, b->nsets, true);
2253                 invalidated_iter = true;
2254         } else {
2255                 invalidated_iter = bch2_drop_whiteouts(b, COMPACT_ALL);
2256         }
2257
2258         for_each_bset(b, t)
2259                 set_needs_whiteout(bset(b, t), true);
2260
2261         bch2_btree_verify(c, b);
2262
2263         /*
2264          * If later we don't unconditionally sort down to a single bset, we have
2265          * to ensure this is still true:
2266          */
2267         BUG_ON((void *) btree_bkey_last(b, bset_tree_last(b)) > write_block(b));
2268
2269         bne = want_new_bset(c, b);
2270         if (bne)
2271                 bch2_bset_init_next(b, bne);
2272
2273         bch2_btree_build_aux_trees(b);
2274
2275         return invalidated_iter;
2276 }
2277
2278 /*
2279  * Use this one if the node is intent locked:
2280  */
2281 void bch2_btree_node_write(struct bch_fs *c, struct btree *b,
2282                            enum six_lock_type lock_type_held,
2283                            unsigned flags)
2284 {
2285         if (lock_type_held == SIX_LOCK_intent ||
2286             (lock_type_held == SIX_LOCK_read &&
2287              six_lock_tryupgrade(&b->c.lock))) {
2288                 __bch2_btree_node_write(c, b, flags);
2289
2290                 /* don't cycle lock unnecessarily: */
2291                 if (btree_node_just_written(b) &&
2292                     six_trylock_write(&b->c.lock)) {
2293                         bch2_btree_post_write_cleanup(c, b);
2294                         six_unlock_write(&b->c.lock);
2295                 }
2296
2297                 if (lock_type_held == SIX_LOCK_read)
2298                         six_lock_downgrade(&b->c.lock);
2299         } else {
2300                 __bch2_btree_node_write(c, b, flags);
2301                 if (lock_type_held == SIX_LOCK_write &&
2302                     btree_node_just_written(b))
2303                         bch2_btree_post_write_cleanup(c, b);
2304         }
2305 }
2306
2307 static bool __bch2_btree_flush_all(struct bch_fs *c, unsigned flag)
2308 {
2309         struct bucket_table *tbl;
2310         struct rhash_head *pos;
2311         struct btree *b;
2312         unsigned i;
2313         bool ret = false;
2314 restart:
2315         rcu_read_lock();
2316         for_each_cached_btree(b, c, tbl, i, pos)
2317                 if (test_bit(flag, &b->flags)) {
2318                         rcu_read_unlock();
2319                         wait_on_bit_io(&b->flags, flag, TASK_UNINTERRUPTIBLE);
2320                         ret = true;
2321                         goto restart;
2322                 }
2323         rcu_read_unlock();
2324
2325         return ret;
2326 }
2327
2328 bool bch2_btree_flush_all_reads(struct bch_fs *c)
2329 {
2330         return __bch2_btree_flush_all(c, BTREE_NODE_read_in_flight);
2331 }
2332
2333 bool bch2_btree_flush_all_writes(struct bch_fs *c)
2334 {
2335         return __bch2_btree_flush_all(c, BTREE_NODE_write_in_flight);
2336 }
2337
2338 static const char * const bch2_btree_write_types[] = {
2339 #define x(t, n) [n] = #t,
2340         BCH_BTREE_WRITE_TYPES()
2341         NULL
2342 };
2343
2344 void bch2_btree_write_stats_to_text(struct printbuf *out, struct bch_fs *c)
2345 {
2346         printbuf_tabstop_push(out, 20);
2347         printbuf_tabstop_push(out, 10);
2348
2349         prt_tab(out);
2350         prt_str(out, "nr");
2351         prt_tab(out);
2352         prt_str(out, "size");
2353         prt_newline(out);
2354
2355         for (unsigned i = 0; i < BTREE_WRITE_TYPE_NR; i++) {
2356                 u64 nr          = atomic64_read(&c->btree_write_stats[i].nr);
2357                 u64 bytes       = atomic64_read(&c->btree_write_stats[i].bytes);
2358
2359                 prt_printf(out, "%s:", bch2_btree_write_types[i]);
2360                 prt_tab(out);
2361                 prt_u64(out, nr);
2362                 prt_tab(out);
2363                 prt_human_readable_u64(out, nr ? div64_u64(bytes, nr) : 0);
2364                 prt_newline(out);
2365         }
2366 }