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

   1 // SPDX-License-Identifier: GPL-2.0
   2
   3 #include "bcachefs.h"
   4 #include "alloc_foreground.h"
   5 #include "bkey_methods.h"
   6 #include "btree_cache.h"
   7 #include "btree_gc.h"
   8 #include "btree_journal_iter.h"
   9 #include "btree_update.h"
  10 #include "btree_update_interior.h"
  11 #include "btree_io.h"
  12 #include "btree_iter.h"
  13 #include "btree_locking.h"
  14 #include "buckets.h"
  15 #include "clock.h"
  16 #include "error.h"
  17 #include "extents.h"
  18 #include "journal.h"
  19 #include "journal_reclaim.h"
  20 #include "keylist.h"
  21 #include "replicas.h"
  22 #include "super-io.h"
  23 #include "trace.h"
  24
  25 #include <linux/random.h>
  26
  27 static int bch2_btree_insert_node(struct btree_update *, struct btree_trans *,
  28                                   struct btree_path *, struct btree *,
  29                                   struct keylist *, unsigned);
  30 static void bch2_btree_update_add_new_node(struct btree_update *, struct btree *);
  31
  32 static struct btree_path *get_unlocked_mut_path(struct btree_trans *trans,
  33                                                 enum btree_id btree_id,
  34                                                 unsigned level,
  35                                                 struct bpos pos)
  36 {
  37         struct btree_path *path;
  38
  39         path = bch2_path_get(trans, btree_id, pos, level + 1, level,
  40                              BTREE_ITER_NOPRESERVE|
  41                              BTREE_ITER_INTENT, _RET_IP_);
  42         path = bch2_btree_path_make_mut(trans, path, true, _RET_IP_);
  43         bch2_btree_path_downgrade(trans, path);
  44         __bch2_btree_path_unlock(trans, path);
  45         return path;
  46 }
  47
  48 /* Debug code: */
  49
  50 /*
  51  * Verify that child nodes correctly span parent node's range:
  52  */
  53 static void btree_node_interior_verify(struct bch_fs *c, struct btree *b)
  54 {
  55 #ifdef CONFIG_BCACHEFS_DEBUG
  56         struct bpos next_node = b->data->min_key;
  57         struct btree_node_iter iter;
  58         struct bkey_s_c k;
  59         struct bkey_s_c_btree_ptr_v2 bp;
  60         struct bkey unpacked;
  61         struct printbuf buf1 = PRINTBUF, buf2 = PRINTBUF;
  62
  63         BUG_ON(!b->c.level);
  64
  65         if (!test_bit(JOURNAL_REPLAY_DONE, &c->journal.flags))
  66                 return;
  67
  68         bch2_btree_node_iter_init_from_start(&iter, b);
  69
  70         while (1) {
  71                 k = bch2_btree_node_iter_peek_unpack(&iter, b, &unpacked);
  72                 if (k.k->type != KEY_TYPE_btree_ptr_v2)
  73                         break;
  74                 bp = bkey_s_c_to_btree_ptr_v2(k);
  75
  76                 if (!bpos_eq(next_node, bp.v->min_key)) {
  77                         bch2_dump_btree_node(c, b);
  78                         bch2_bpos_to_text(&buf1, next_node);
  79                         bch2_bpos_to_text(&buf2, bp.v->min_key);
  80                         panic("expected next min_key %s got %s\n", buf1.buf, buf2.buf);
  81                 }
  82
  83                 bch2_btree_node_iter_advance(&iter, b);
  84
  85                 if (bch2_btree_node_iter_end(&iter)) {
  86                         if (!bpos_eq(k.k->p, b->key.k.p)) {
  87                                 bch2_dump_btree_node(c, b);
  88                                 bch2_bpos_to_text(&buf1, b->key.k.p);
  89                                 bch2_bpos_to_text(&buf2, k.k->p);
  90                                 panic("expected end %s got %s\n", buf1.buf, buf2.buf);
  91                         }
  92                         break;
  93                 }
  94
  95                 next_node = bpos_successor(k.k->p);
  96         }
  97 #endif
  98 }
  99
 100 /* Calculate ideal packed bkey format for new btree nodes: */
 101
 102 void __bch2_btree_calc_format(struct bkey_format_state *s, struct btree *b)
 103 {
 104         struct bkey_packed *k;
 105         struct bset_tree *t;
 106         struct bkey uk;
 107
 108         for_each_bset(b, t)
 109                 bset_tree_for_each_key(b, t, k)
 110                         if (!bkey_deleted(k)) {
 111                                 uk = bkey_unpack_key(b, k);
 112                                 bch2_bkey_format_add_key(s, &uk);
 113                         }
 114 }
 115
 116 static struct bkey_format bch2_btree_calc_format(struct btree *b)
 117 {
 118         struct bkey_format_state s;
 119
 120         bch2_bkey_format_init(&s);
 121         bch2_bkey_format_add_pos(&s, b->data->min_key);
 122         bch2_bkey_format_add_pos(&s, b->data->max_key);
 123         __bch2_btree_calc_format(&s, b);
 124
 125         return bch2_bkey_format_done(&s);
 126 }
 127
 128 static size_t btree_node_u64s_with_format(struct btree *b,
 129                                           struct bkey_format *new_f)
 130 {
 131         struct bkey_format *old_f = &b->format;
 132
 133         /* stupid integer promotion rules */
 134         ssize_t delta =
 135             (((int) new_f->key_u64s - old_f->key_u64s) *
 136              (int) b->nr.packed_keys) +
 137             (((int) new_f->key_u64s - BKEY_U64s) *
 138              (int) b->nr.unpacked_keys);
 139
 140         BUG_ON(delta + b->nr.live_u64s < 0);
 141
 142         return b->nr.live_u64s + delta;
 143 }
 144
 145 /**
 146  * bch2_btree_node_format_fits - check if we could rewrite node with a new format
 147  *
 148  * @c:          filesystem handle
 149  * @b:          btree node to rewrite
 150  * @new_f:      bkey format to translate keys to
 151  *
 152  * Returns: true if all re-packed keys will be able to fit in a new node.
 153  *
 154  * Assumes all keys will successfully pack with the new format.
 155  */
 156 bool bch2_btree_node_format_fits(struct bch_fs *c, struct btree *b,
 157                                  struct bkey_format *new_f)
 158 {
 159         size_t u64s = btree_node_u64s_with_format(b, new_f);
 160
 161         return __vstruct_bytes(struct btree_node, u64s) < btree_bytes(c);
 162 }
 163
 164 /* Btree node freeing/allocation: */
 165
 166 static void __btree_node_free(struct bch_fs *c, struct btree *b)
 167 {
 168         trace_and_count(c, btree_node_free, c, b);
 169
 170         BUG_ON(btree_node_write_blocked(b));
 171         BUG_ON(btree_node_dirty(b));
 172         BUG_ON(btree_node_need_write(b));
 173         BUG_ON(b == btree_node_root(c, b));
 174         BUG_ON(b->ob.nr);
 175         BUG_ON(!list_empty(&b->write_blocked));
 176         BUG_ON(b->will_make_reachable);
 177
 178         clear_btree_node_noevict(b);
 179
 180         mutex_lock(&c->btree_cache.lock);
 181         list_move(&b->list, &c->btree_cache.freeable);
 182         mutex_unlock(&c->btree_cache.lock);
 183 }
 184
 185 static void bch2_btree_node_free_inmem(struct btree_trans *trans,
 186                                        struct btree_path *path,
 187                                        struct btree *b)
 188 {
 189         struct bch_fs *c = trans->c;
 190         unsigned level = b->c.level;
 191
 192         bch2_btree_node_lock_write_nofail(trans, path, &b->c);
 193         bch2_btree_node_hash_remove(&c->btree_cache, b);
 194         __btree_node_free(c, b);
 195         six_unlock_write(&b->c.lock);
 196         mark_btree_node_locked_noreset(path, level, BTREE_NODE_INTENT_LOCKED);
 197
 198         trans_for_each_path(trans, path)
 199                 if (path->l[level].b == b) {
 200                         btree_node_unlock(trans, path, level);
 201                         path->l[level].b = ERR_PTR(-BCH_ERR_no_btree_node_init);
 202                 }
 203 }
 204
 205 static void bch2_btree_node_free_never_used(struct btree_update *as,
 206                                             struct btree_trans *trans,
 207                                             struct btree *b)
 208 {
 209         struct bch_fs *c = as->c;
 210         struct prealloc_nodes *p = &as->prealloc_nodes[b->c.lock.readers != NULL];
 211         struct btree_path *path;
 212         unsigned level = b->c.level;
 213
 214         BUG_ON(!list_empty(&b->write_blocked));
 215         BUG_ON(b->will_make_reachable != (1UL|(unsigned long) as));
 216
 217         b->will_make_reachable = 0;
 218         closure_put(&as->cl);
 219
 220         clear_btree_node_will_make_reachable(b);
 221         clear_btree_node_accessed(b);
 222         clear_btree_node_dirty_acct(c, b);
 223         clear_btree_node_need_write(b);
 224
 225         mutex_lock(&c->btree_cache.lock);
 226         list_del_init(&b->list);
 227         bch2_btree_node_hash_remove(&c->btree_cache, b);
 228         mutex_unlock(&c->btree_cache.lock);
 229
 230         BUG_ON(p->nr >= ARRAY_SIZE(p->b));
 231         p->b[p->nr++] = b;
 232
 233         six_unlock_intent(&b->c.lock);
 234
 235         trans_for_each_path(trans, path)
 236                 if (path->l[level].b == b) {
 237                         btree_node_unlock(trans, path, level);
 238                         path->l[level].b = ERR_PTR(-BCH_ERR_no_btree_node_init);
 239                 }
 240 }
 241
 242 static struct btree *__bch2_btree_node_alloc(struct btree_trans *trans,
 243                                              struct disk_reservation *res,
 244                                              struct closure *cl,
 245                                              bool interior_node,
 246                                              unsigned flags)
 247 {
 248         struct bch_fs *c = trans->c;
 249         struct write_point *wp;
 250         struct btree *b;
 251         BKEY_PADDED_ONSTACK(k, BKEY_BTREE_PTR_VAL_U64s_MAX) tmp;
 252         struct open_buckets obs = { .nr = 0 };
 253         struct bch_devs_list devs_have = (struct bch_devs_list) { 0 };
 254         enum bch_watermark watermark = flags & BCH_WATERMARK_MASK;
 255         unsigned nr_reserve = watermark > BCH_WATERMARK_reclaim
 256                 ? BTREE_NODE_RESERVE
 257                 : 0;
 258         int ret;
 259
 260         mutex_lock(&c->btree_reserve_cache_lock);
 261         if (c->btree_reserve_cache_nr > nr_reserve) {
 262                 struct btree_alloc *a =
 263                         &c->btree_reserve_cache[--c->btree_reserve_cache_nr];
 264
 265                 obs = a->ob;
 266                 bkey_copy(&tmp.k, &a->k);
 267                 mutex_unlock(&c->btree_reserve_cache_lock);
 268                 goto mem_alloc;
 269         }
 270         mutex_unlock(&c->btree_reserve_cache_lock);
 271
 272 retry:
 273         ret = bch2_alloc_sectors_start_trans(trans,
 274                                       c->opts.metadata_target ?:
 275                                       c->opts.foreground_target,
 276                                       0,
 277                                       writepoint_ptr(&c->btree_write_point),
 278                                       &devs_have,
 279                                       res->nr_replicas,
 280                                       c->opts.metadata_replicas_required,
 281                                       watermark, 0, cl, &wp);
 282         if (unlikely(ret))
 283                 return ERR_PTR(ret);
 284
 285         if (wp->sectors_free < btree_sectors(c)) {
 286                 struct open_bucket *ob;
 287                 unsigned i;
 288
 289                 open_bucket_for_each(c, &wp->ptrs, ob, i)
 290                         if (ob->sectors_free < btree_sectors(c))
 291                                 ob->sectors_free = 0;
 292
 293                 bch2_alloc_sectors_done(c, wp);
 294                 goto retry;
 295         }
 296
 297         bkey_btree_ptr_v2_init(&tmp.k);
 298         bch2_alloc_sectors_append_ptrs(c, wp, &tmp.k, btree_sectors(c), false);
 299
 300         bch2_open_bucket_get(c, wp, &obs);
 301         bch2_alloc_sectors_done(c, wp);
 302 mem_alloc:
 303         b = bch2_btree_node_mem_alloc(trans, interior_node);
 304         six_unlock_write(&b->c.lock);
 305         six_unlock_intent(&b->c.lock);
 306
 307         /* we hold cannibalize_lock: */
 308         BUG_ON(IS_ERR(b));
 309         BUG_ON(b->ob.nr);
 310
 311         bkey_copy(&b->key, &tmp.k);
 312         b->ob = obs;
 313
 314         return b;
 315 }
 316
 317 static struct btree *bch2_btree_node_alloc(struct btree_update *as,
 318                                            struct btree_trans *trans,
 319                                            unsigned level)
 320 {
 321         struct bch_fs *c = as->c;
 322         struct btree *b;
 323         struct prealloc_nodes *p = &as->prealloc_nodes[!!level];
 324         int ret;
 325
 326         BUG_ON(level >= BTREE_MAX_DEPTH);
 327         BUG_ON(!p->nr);
 328
 329         b = p->b[--p->nr];
 330
 331         btree_node_lock_nopath_nofail(trans, &b->c, SIX_LOCK_intent);
 332         btree_node_lock_nopath_nofail(trans, &b->c, SIX_LOCK_write);
 333
 334         set_btree_node_accessed(b);
 335         set_btree_node_dirty_acct(c, b);
 336         set_btree_node_need_write(b);
 337
 338         bch2_bset_init_first(b, &b->data->keys);
 339         b->c.level      = level;
 340         b->c.btree_id   = as->btree_id;
 341         b->version_ondisk = c->sb.version;
 342
 343         memset(&b->nr, 0, sizeof(b->nr));
 344         b->data->magic = cpu_to_le64(bset_magic(c));
 345         memset(&b->data->_ptr, 0, sizeof(b->data->_ptr));
 346         b->data->flags = 0;
 347         SET_BTREE_NODE_ID(b->data, as->btree_id);
 348         SET_BTREE_NODE_LEVEL(b->data, level);
 349
 350         if (b->key.k.type == KEY_TYPE_btree_ptr_v2) {
 351                 struct bkey_i_btree_ptr_v2 *bp = bkey_i_to_btree_ptr_v2(&b->key);
 352
 353                 bp->v.mem_ptr           = 0;
 354                 bp->v.seq               = b->data->keys.seq;
 355                 bp->v.sectors_written   = 0;
 356         }
 357
 358         SET_BTREE_NODE_NEW_EXTENT_OVERWRITE(b->data, true);
 359
 360         bch2_btree_build_aux_trees(b);
 361
 362         ret = bch2_btree_node_hash_insert(&c->btree_cache, b, level, as->btree_id);
 363         BUG_ON(ret);
 364
 365         trace_and_count(c, btree_node_alloc, c, b);
 366         bch2_increment_clock(c, btree_sectors(c), WRITE);
 367         return b;
 368 }
 369
 370 static void btree_set_min(struct btree *b, struct bpos pos)
 371 {
 372         if (b->key.k.type == KEY_TYPE_btree_ptr_v2)
 373                 bkey_i_to_btree_ptr_v2(&b->key)->v.min_key = pos;
 374         b->data->min_key = pos;
 375 }
 376
 377 static void btree_set_max(struct btree *b, struct bpos pos)
 378 {
 379         b->key.k.p = pos;
 380         b->data->max_key = pos;
 381 }
 382
 383 static struct btree *bch2_btree_node_alloc_replacement(struct btree_update *as,
 384                                                        struct btree_trans *trans,
 385                                                        struct btree *b)
 386 {
 387         struct btree *n = bch2_btree_node_alloc(as, trans, b->c.level);
 388         struct bkey_format format = bch2_btree_calc_format(b);
 389
 390         /*
 391          * The keys might expand with the new format - if they wouldn't fit in
 392          * the btree node anymore, use the old format for now:
 393          */
 394         if (!bch2_btree_node_format_fits(as->c, b, &format))
 395                 format = b->format;
 396
 397         SET_BTREE_NODE_SEQ(n->data, BTREE_NODE_SEQ(b->data) + 1);
 398
 399         btree_set_min(n, b->data->min_key);
 400         btree_set_max(n, b->data->max_key);
 401
 402         n->data->format         = format;
 403         btree_node_set_format(n, format);
 404
 405         bch2_btree_sort_into(as->c, n, b);
 406
 407         btree_node_reset_sib_u64s(n);
 408         return n;
 409 }
 410
 411 static struct btree *__btree_root_alloc(struct btree_update *as,
 412                                 struct btree_trans *trans, unsigned level)
 413 {
 414         struct btree *b = bch2_btree_node_alloc(as, trans, level);
 415
 416         btree_set_min(b, POS_MIN);
 417         btree_set_max(b, SPOS_MAX);
 418         b->data->format = bch2_btree_calc_format(b);
 419
 420         btree_node_set_format(b, b->data->format);
 421         bch2_btree_build_aux_trees(b);
 422
 423         return b;
 424 }
 425
 426 static void bch2_btree_reserve_put(struct btree_update *as, struct btree_trans *trans)
 427 {
 428         struct bch_fs *c = as->c;
 429         struct prealloc_nodes *p;
 430
 431         for (p = as->prealloc_nodes;
 432              p < as->prealloc_nodes + ARRAY_SIZE(as->prealloc_nodes);
 433              p++) {
 434                 while (p->nr) {
 435                         struct btree *b = p->b[--p->nr];
 436
 437                         mutex_lock(&c->btree_reserve_cache_lock);
 438
 439                         if (c->btree_reserve_cache_nr <
 440                             ARRAY_SIZE(c->btree_reserve_cache)) {
 441                                 struct btree_alloc *a =
 442                                         &c->btree_reserve_cache[c->btree_reserve_cache_nr++];
 443
 444                                 a->ob = b->ob;
 445                                 b->ob.nr = 0;
 446                                 bkey_copy(&a->k, &b->key);
 447                         } else {
 448                                 bch2_open_buckets_put(c, &b->ob);
 449                         }
 450
 451                         mutex_unlock(&c->btree_reserve_cache_lock);
 452
 453                         btree_node_lock_nopath_nofail(trans, &b->c, SIX_LOCK_intent);
 454                         btree_node_lock_nopath_nofail(trans, &b->c, SIX_LOCK_write);
 455                         __btree_node_free(c, b);
 456                         six_unlock_write(&b->c.lock);
 457                         six_unlock_intent(&b->c.lock);
 458                 }
 459         }
 460 }
 461
 462 static int bch2_btree_reserve_get(struct btree_trans *trans,
 463                                   struct btree_update *as,
 464                                   unsigned nr_nodes[2],
 465                                   unsigned flags,
 466                                   struct closure *cl)
 467 {
 468         struct bch_fs *c = as->c;
 469         struct btree *b;
 470         unsigned interior;
 471         int ret = 0;
 472
 473         BUG_ON(nr_nodes[0] + nr_nodes[1] > BTREE_RESERVE_MAX);
 474
 475         /*
 476          * Protects reaping from the btree node cache and using the btree node
 477          * open bucket reserve:
 478          *
 479          * BTREE_INSERT_NOWAIT only applies to btree node allocation, not
 480          * blocking on this lock:
 481          */
 482         ret = bch2_btree_cache_cannibalize_lock(c, cl);
 483         if (ret)
 484                 return ret;
 485
 486         for (interior = 0; interior < 2; interior++) {
 487                 struct prealloc_nodes *p = as->prealloc_nodes + interior;
 488
 489                 while (p->nr < nr_nodes[interior]) {
 490                         b = __bch2_btree_node_alloc(trans, &as->disk_res,
 491                                         flags & BTREE_INSERT_NOWAIT ? NULL : cl,
 492                                         interior, flags);
 493                         if (IS_ERR(b)) {
 494                                 ret = PTR_ERR(b);
 495                                 goto err;
 496                         }
 497
 498                         p->b[p->nr++] = b;
 499                 }
 500         }
 501 err:
 502         bch2_btree_cache_cannibalize_unlock(c);
 503         return ret;
 504 }
 505
 506 /* Asynchronous interior node update machinery */
 507
 508 static void bch2_btree_update_free(struct btree_update *as, struct btree_trans *trans)
 509 {
 510         struct bch_fs *c = as->c;
 511
 512         if (as->took_gc_lock)
 513                 up_read(&c->gc_lock);
 514         as->took_gc_lock = false;
 515
 516         bch2_journal_preres_put(&c->journal, &as->journal_preres);
 517
 518         bch2_journal_pin_drop(&c->journal, &as->journal);
 519         bch2_journal_pin_flush(&c->journal, &as->journal);
 520         bch2_disk_reservation_put(c, &as->disk_res);
 521         bch2_btree_reserve_put(as, trans);
 522
 523         bch2_time_stats_update(&c->times[BCH_TIME_btree_interior_update_total],
 524                                as->start_time);
 525
 526         mutex_lock(&c->btree_interior_update_lock);
 527         list_del(&as->unwritten_list);
 528         list_del(&as->list);
 529
 530         closure_debug_destroy(&as->cl);
 531         mempool_free(as, &c->btree_interior_update_pool);
 532
 533         /*
 534          * Have to do the wakeup with btree_interior_update_lock still held,
 535          * since being on btree_interior_update_list is our ref on @c:
 536          */
 537         closure_wake_up(&c->btree_interior_update_wait);
 538
 539         mutex_unlock(&c->btree_interior_update_lock);
 540 }
 541
 542 static void btree_update_add_key(struct btree_update *as,
 543                                  struct keylist *keys, struct btree *b)
 544 {
 545         struct bkey_i *k = &b->key;
 546
 547         BUG_ON(bch2_keylist_u64s(keys) + k->k.u64s >
 548                ARRAY_SIZE(as->_old_keys));
 549
 550         bkey_copy(keys->top, k);
 551         bkey_i_to_btree_ptr_v2(keys->top)->v.mem_ptr = b->c.level + 1;
 552
 553         bch2_keylist_push(keys);
 554 }
 555
 556 /*
 557  * The transactional part of an interior btree node update, where we journal the
 558  * update we did to the interior node and update alloc info:
 559  */
 560 static int btree_update_nodes_written_trans(struct btree_trans *trans,
 561                                             struct btree_update *as)
 562 {
 563         struct bkey_i *k;
 564         int ret;
 565
 566         ret = darray_make_room(&trans->extra_journal_entries, as->journal_u64s);
 567         if (ret)
 568                 return ret;
 569
 570         memcpy(&darray_top(trans->extra_journal_entries),
 571                as->journal_entries,
 572                as->journal_u64s * sizeof(u64));
 573         trans->extra_journal_entries.nr += as->journal_u64s;
 574
 575         trans->journal_pin = &as->journal;
 576
 577         for_each_keylist_key(&as->old_keys, k) {
 578                 unsigned level = bkey_i_to_btree_ptr_v2(k)->v.mem_ptr;
 579
 580                 ret = bch2_trans_mark_old(trans, as->btree_id, level, bkey_i_to_s_c(k), 0);
 581                 if (ret)
 582                         return ret;
 583         }
 584
 585         for_each_keylist_key(&as->new_keys, k) {
 586                 unsigned level = bkey_i_to_btree_ptr_v2(k)->v.mem_ptr;
 587
 588                 ret = bch2_trans_mark_new(trans, as->btree_id, level, k, 0);
 589                 if (ret)
 590                         return ret;
 591         }
 592
 593         return 0;
 594 }
 595
 596 static void btree_update_nodes_written(struct btree_update *as)
 597 {
 598         struct bch_fs *c = as->c;
 599         struct btree *b;
 600         struct btree_trans *trans = bch2_trans_get(c);
 601         u64 journal_seq = 0;
 602         unsigned i;
 603         int ret;
 604
 605         /*
 606          * If we're already in an error state, it might be because a btree node
 607          * was never written, and we might be trying to free that same btree
 608          * node here, but it won't have been marked as allocated and we'll see
 609          * spurious disk usage inconsistencies in the transactional part below
 610          * if we don't skip it:
 611          */
 612         ret = bch2_journal_error(&c->journal);
 613         if (ret)
 614                 goto err;
 615
 616         /*
 617          * Wait for any in flight writes to finish before we free the old nodes
 618          * on disk:
 619          */
 620         for (i = 0; i < as->nr_old_nodes; i++) {
 621                 __le64 seq;
 622
 623                 b = as->old_nodes[i];
 624
 625                 btree_node_lock_nopath_nofail(trans, &b->c, SIX_LOCK_read);
 626                 seq = b->data ? b->data->keys.seq : 0;
 627                 six_unlock_read(&b->c.lock);
 628
 629                 if (seq == as->old_nodes_seq[i])
 630                         wait_on_bit_io(&b->flags, BTREE_NODE_write_in_flight_inner,
 631                                        TASK_UNINTERRUPTIBLE);
 632         }
 633
 634         /*
 635          * We did an update to a parent node where the pointers we added pointed
 636          * to child nodes that weren't written yet: now, the child nodes have
 637          * been written so we can write out the update to the interior node.
 638          */
 639
 640         /*
 641          * We can't call into journal reclaim here: we'd block on the journal
 642          * reclaim lock, but we may need to release the open buckets we have
 643          * pinned in order for other btree updates to make forward progress, and
 644          * journal reclaim does btree updates when flushing bkey_cached entries,
 645          * which may require allocations as well.
 646          */
 647         ret = commit_do(trans, &as->disk_res, &journal_seq,
 648                         BCH_WATERMARK_reclaim|
 649                         BTREE_INSERT_NOFAIL|
 650                         BTREE_INSERT_NOCHECK_RW|
 651                         BTREE_INSERT_JOURNAL_RECLAIM,
 652                         btree_update_nodes_written_trans(trans, as));
 653         bch2_trans_unlock(trans);
 654
 655         bch2_fs_fatal_err_on(ret && !bch2_journal_error(&c->journal), c,
 656                              "%s(): error %s", __func__, bch2_err_str(ret));
 657 err:
 658         if (as->b) {
 659                 struct btree_path *path;
 660
 661                 b = as->b;
 662                 path = get_unlocked_mut_path(trans, as->btree_id, b->c.level, b->key.k.p);
 663                 /*
 664                  * @b is the node we did the final insert into:
 665                  *
 666                  * On failure to get a journal reservation, we still have to
 667                  * unblock the write and allow most of the write path to happen
 668                  * so that shutdown works, but the i->journal_seq mechanism
 669                  * won't work to prevent the btree write from being visible (we
 670                  * didn't get a journal sequence number) - instead
 671                  * __bch2_btree_node_write() doesn't do the actual write if
 672                  * we're in journal error state:
 673                  */
 674
 675                 /*
 676                  * Ensure transaction is unlocked before using
 677                  * btree_node_lock_nopath() (the use of which is always suspect,
 678                  * we need to work on removing this in the future)
 679                  *
 680                  * It should be, but get_unlocked_mut_path() -> bch2_path_get()
 681                  * calls bch2_path_upgrade(), before we call path_make_mut(), so
 682                  * we may rarely end up with a locked path besides the one we
 683                  * have here:
 684                  */
 685                 bch2_trans_unlock(trans);
 686                 btree_node_lock_nopath_nofail(trans, &b->c, SIX_LOCK_intent);
 687                 mark_btree_node_locked(trans, path, b->c.level, BTREE_NODE_INTENT_LOCKED);
 688                 path->l[b->c.level].lock_seq = six_lock_seq(&b->c.lock);
 689                 path->l[b->c.level].b = b;
 690
 691                 bch2_btree_node_lock_write_nofail(trans, path, &b->c);
 692
 693                 mutex_lock(&c->btree_interior_update_lock);
 694
 695                 list_del(&as->write_blocked_list);
 696                 if (list_empty(&b->write_blocked))
 697                         clear_btree_node_write_blocked(b);
 698
 699                 /*
 700                  * Node might have been freed, recheck under
 701                  * btree_interior_update_lock:
 702                  */
 703                 if (as->b == b) {
 704                         BUG_ON(!b->c.level);
 705                         BUG_ON(!btree_node_dirty(b));
 706
 707                         if (!ret) {
 708                                 struct bset *last = btree_bset_last(b);
 709
 710                                 last->journal_seq = cpu_to_le64(
 711                                                              max(journal_seq,
 712                                                                  le64_to_cpu(last->journal_seq)));
 713
 714                                 bch2_btree_add_journal_pin(c, b, journal_seq);
 715                         } else {
 716                                 /*
 717                                  * If we didn't get a journal sequence number we
 718                                  * can't write this btree node, because recovery
 719                                  * won't know to ignore this write:
 720                                  */
 721                                 set_btree_node_never_write(b);
 722                         }
 723                 }
 724
 725                 mutex_unlock(&c->btree_interior_update_lock);
 726
 727                 mark_btree_node_locked_noreset(path, b->c.level, BTREE_NODE_INTENT_LOCKED);
 728                 six_unlock_write(&b->c.lock);
 729
 730                 btree_node_write_if_need(c, b, SIX_LOCK_intent);
 731                 btree_node_unlock(trans, path, b->c.level);
 732                 bch2_path_put(trans, path, true);
 733         }
 734
 735         bch2_journal_pin_drop(&c->journal, &as->journal);
 736
 737         bch2_journal_preres_put(&c->journal, &as->journal_preres);
 738
 739         mutex_lock(&c->btree_interior_update_lock);
 740         for (i = 0; i < as->nr_new_nodes; i++) {
 741                 b = as->new_nodes[i];
 742
 743                 BUG_ON(b->will_make_reachable != (unsigned long) as);
 744                 b->will_make_reachable = 0;
 745                 clear_btree_node_will_make_reachable(b);
 746         }
 747         mutex_unlock(&c->btree_interior_update_lock);
 748
 749         for (i = 0; i < as->nr_new_nodes; i++) {
 750                 b = as->new_nodes[i];
 751
 752                 btree_node_lock_nopath_nofail(trans, &b->c, SIX_LOCK_read);
 753                 btree_node_write_if_need(c, b, SIX_LOCK_read);
 754                 six_unlock_read(&b->c.lock);
 755         }
 756
 757         for (i = 0; i < as->nr_open_buckets; i++)
 758                 bch2_open_bucket_put(c, c->open_buckets + as->open_buckets[i]);
 759
 760         bch2_btree_update_free(as, trans);
 761         bch2_trans_put(trans);
 762 }
 763
 764 static void btree_interior_update_work(struct work_struct *work)
 765 {
 766         struct bch_fs *c =
 767                 container_of(work, struct bch_fs, btree_interior_update_work);
 768         struct btree_update *as;
 769
 770         while (1) {
 771                 mutex_lock(&c->btree_interior_update_lock);
 772                 as = list_first_entry_or_null(&c->btree_interior_updates_unwritten,
 773                                               struct btree_update, unwritten_list);
 774                 if (as && !as->nodes_written)
 775                         as = NULL;
 776                 mutex_unlock(&c->btree_interior_update_lock);
 777
 778                 if (!as)
 779                         break;
 780
 781                 btree_update_nodes_written(as);
 782         }
 783 }
 784
 785 static void btree_update_set_nodes_written(struct closure *cl)
 786 {
 787         struct btree_update *as = container_of(cl, struct btree_update, cl);
 788         struct bch_fs *c = as->c;
 789
 790         mutex_lock(&c->btree_interior_update_lock);
 791         as->nodes_written = true;
 792         mutex_unlock(&c->btree_interior_update_lock);
 793
 794         queue_work(c->btree_interior_update_worker, &c->btree_interior_update_work);
 795 }
 796
 797 /*
 798  * We're updating @b with pointers to nodes that haven't finished writing yet:
 799  * block @b from being written until @as completes
 800  */
 801 static void btree_update_updated_node(struct btree_update *as, struct btree *b)
 802 {
 803         struct bch_fs *c = as->c;
 804
 805         mutex_lock(&c->btree_interior_update_lock);
 806         list_add_tail(&as->unwritten_list, &c->btree_interior_updates_unwritten);
 807
 808         BUG_ON(as->mode != BTREE_INTERIOR_NO_UPDATE);
 809         BUG_ON(!btree_node_dirty(b));
 810         BUG_ON(!b->c.level);
 811
 812         as->mode        = BTREE_INTERIOR_UPDATING_NODE;
 813         as->b           = b;
 814
 815         set_btree_node_write_blocked(b);
 816         list_add(&as->write_blocked_list, &b->write_blocked);
 817
 818         mutex_unlock(&c->btree_interior_update_lock);
 819 }
 820
 821 static void btree_update_reparent(struct btree_update *as,
 822                                   struct btree_update *child)
 823 {
 824         struct bch_fs *c = as->c;
 825
 826         lockdep_assert_held(&c->btree_interior_update_lock);
 827
 828         child->b = NULL;
 829         child->mode = BTREE_INTERIOR_UPDATING_AS;
 830
 831         bch2_journal_pin_copy(&c->journal, &as->journal, &child->journal, NULL);
 832 }
 833
 834 static void btree_update_updated_root(struct btree_update *as, struct btree *b)
 835 {
 836         struct bkey_i *insert = &b->key;
 837         struct bch_fs *c = as->c;
 838
 839         BUG_ON(as->mode != BTREE_INTERIOR_NO_UPDATE);
 840
 841         BUG_ON(as->journal_u64s + jset_u64s(insert->k.u64s) >
 842                ARRAY_SIZE(as->journal_entries));
 843
 844         as->journal_u64s +=
 845                 journal_entry_set((void *) &as->journal_entries[as->journal_u64s],
 846                                   BCH_JSET_ENTRY_btree_root,
 847                                   b->c.btree_id, b->c.level,
 848                                   insert, insert->k.u64s);
 849
 850         mutex_lock(&c->btree_interior_update_lock);
 851         list_add_tail(&as->unwritten_list, &c->btree_interior_updates_unwritten);
 852
 853         as->mode        = BTREE_INTERIOR_UPDATING_ROOT;
 854         mutex_unlock(&c->btree_interior_update_lock);
 855 }
 856
 857 /*
 858  * bch2_btree_update_add_new_node:
 859  *
 860  * This causes @as to wait on @b to be written, before it gets to
 861  * bch2_btree_update_nodes_written
 862  *
 863  * Additionally, it sets b->will_make_reachable to prevent any additional writes
 864  * to @b from happening besides the first until @b is reachable on disk
 865  *
 866  * And it adds @b to the list of @as's new nodes, so that we can update sector
 867  * counts in bch2_btree_update_nodes_written:
 868  */
 869 static void bch2_btree_update_add_new_node(struct btree_update *as, struct btree *b)
 870 {
 871         struct bch_fs *c = as->c;
 872
 873         closure_get(&as->cl);
 874
 875         mutex_lock(&c->btree_interior_update_lock);
 876         BUG_ON(as->nr_new_nodes >= ARRAY_SIZE(as->new_nodes));
 877         BUG_ON(b->will_make_reachable);
 878
 879         as->new_nodes[as->nr_new_nodes++] = b;
 880         b->will_make_reachable = 1UL|(unsigned long) as;
 881         set_btree_node_will_make_reachable(b);
 882
 883         mutex_unlock(&c->btree_interior_update_lock);
 884
 885         btree_update_add_key(as, &as->new_keys, b);
 886
 887         if (b->key.k.type == KEY_TYPE_btree_ptr_v2) {
 888                 unsigned bytes = vstruct_end(&b->data->keys) - (void *) b->data;
 889                 unsigned sectors = round_up(bytes, block_bytes(c)) >> 9;
 890
 891                 bkey_i_to_btree_ptr_v2(&b->key)->v.sectors_written =
 892                         cpu_to_le16(sectors);
 893         }
 894 }
 895
 896 /*
 897  * returns true if @b was a new node
 898  */
 899 static void btree_update_drop_new_node(struct bch_fs *c, struct btree *b)
 900 {
 901         struct btree_update *as;
 902         unsigned long v;
 903         unsigned i;
 904
 905         mutex_lock(&c->btree_interior_update_lock);
 906         /*
 907          * When b->will_make_reachable != 0, it owns a ref on as->cl that's
 908          * dropped when it gets written by bch2_btree_complete_write - the
 909          * xchg() is for synchronization with bch2_btree_complete_write:
 910          */
 911         v = xchg(&b->will_make_reachable, 0);
 912         clear_btree_node_will_make_reachable(b);
 913         as = (struct btree_update *) (v & ~1UL);
 914
 915         if (!as) {
 916                 mutex_unlock(&c->btree_interior_update_lock);
 917                 return;
 918         }
 919
 920         for (i = 0; i < as->nr_new_nodes; i++)
 921                 if (as->new_nodes[i] == b)
 922                         goto found;
 923
 924         BUG();
 925 found:
 926         array_remove_item(as->new_nodes, as->nr_new_nodes, i);
 927         mutex_unlock(&c->btree_interior_update_lock);
 928
 929         if (v & 1)
 930                 closure_put(&as->cl);
 931 }
 932
 933 static void bch2_btree_update_get_open_buckets(struct btree_update *as, struct btree *b)
 934 {
 935         while (b->ob.nr)
 936                 as->open_buckets[as->nr_open_buckets++] =
 937                         b->ob.v[--b->ob.nr];
 938 }
 939
 940 /*
 941  * @b is being split/rewritten: it may have pointers to not-yet-written btree
 942  * nodes and thus outstanding btree_updates - redirect @b's
 943  * btree_updates to point to this btree_update:
 944  */
 945 static void bch2_btree_interior_update_will_free_node(struct btree_update *as,
 946                                                       struct btree *b)
 947 {
 948         struct bch_fs *c = as->c;
 949         struct btree_update *p, *n;
 950         struct btree_write *w;
 951
 952         set_btree_node_dying(b);
 953
 954         if (btree_node_fake(b))
 955                 return;
 956
 957         mutex_lock(&c->btree_interior_update_lock);
 958
 959         /*
 960          * Does this node have any btree_update operations preventing
 961          * it from being written?
 962          *
 963          * If so, redirect them to point to this btree_update: we can
 964          * write out our new nodes, but we won't make them visible until those
 965          * operations complete
 966          */
 967         list_for_each_entry_safe(p, n, &b->write_blocked, write_blocked_list) {
 968                 list_del_init(&p->write_blocked_list);
 969                 btree_update_reparent(as, p);
 970
 971                 /*
 972                  * for flush_held_btree_writes() waiting on updates to flush or
 973                  * nodes to be writeable:
 974                  */
 975                 closure_wake_up(&c->btree_interior_update_wait);
 976         }
 977
 978         clear_btree_node_dirty_acct(c, b);
 979         clear_btree_node_need_write(b);
 980         clear_btree_node_write_blocked(b);
 981
 982         /*
 983          * Does this node have unwritten data that has a pin on the journal?
 984          *
 985          * If so, transfer that pin to the btree_update operation -
 986          * note that if we're freeing multiple nodes, we only need to keep the
 987          * oldest pin of any of the nodes we're freeing. We'll release the pin
 988          * when the new nodes are persistent and reachable on disk:
 989          */
 990         w = btree_current_write(b);
 991         bch2_journal_pin_copy(&c->journal, &as->journal, &w->journal, NULL);
 992         bch2_journal_pin_drop(&c->journal, &w->journal);
 993
 994         w = btree_prev_write(b);
 995         bch2_journal_pin_copy(&c->journal, &as->journal, &w->journal, NULL);
 996         bch2_journal_pin_drop(&c->journal, &w->journal);
 997
 998         mutex_unlock(&c->btree_interior_update_lock);
 999
1000         /*
1001          * Is this a node that isn't reachable on disk yet?
1002          *
1003          * Nodes that aren't reachable yet have writes blocked until they're
1004          * reachable - now that we've cancelled any pending writes and moved
1005          * things waiting on that write to wait on this update, we can drop this
1006          * node from the list of nodes that the other update is making
1007          * reachable, prior to freeing it:
1008          */
1009         btree_update_drop_new_node(c, b);
1010
1011         btree_update_add_key(as, &as->old_keys, b);
1012
1013         as->old_nodes[as->nr_old_nodes] = b;
1014         as->old_nodes_seq[as->nr_old_nodes] = b->data->keys.seq;
1015         as->nr_old_nodes++;
1016 }
1017
1018 static void bch2_btree_update_done(struct btree_update *as, struct btree_trans *trans)
1019 {
1020         struct bch_fs *c = as->c;
1021         u64 start_time = as->start_time;
1022
1023         BUG_ON(as->mode == BTREE_INTERIOR_NO_UPDATE);
1024
1025         if (as->took_gc_lock)
1026                 up_read(&as->c->gc_lock);
1027         as->took_gc_lock = false;
1028
1029         bch2_btree_reserve_put(as, trans);
1030
1031         continue_at(&as->cl, btree_update_set_nodes_written,
1032                     as->c->btree_interior_update_worker);
1033
1034         bch2_time_stats_update(&c->times[BCH_TIME_btree_interior_update_foreground],
1035                                start_time);
1036 }
1037
1038 static struct btree_update *
1039 bch2_btree_update_start(struct btree_trans *trans, struct btree_path *path,
1040                         unsigned level, bool split, unsigned flags)
1041 {
1042         struct bch_fs *c = trans->c;
1043         struct btree_update *as;
1044         u64 start_time = local_clock();
1045         int disk_res_flags = (flags & BTREE_INSERT_NOFAIL)
1046                 ? BCH_DISK_RESERVATION_NOFAIL : 0;
1047         unsigned nr_nodes[2] = { 0, 0 };
1048         unsigned update_level = level;
1049         enum bch_watermark watermark = flags & BCH_WATERMARK_MASK;
1050         unsigned journal_flags = 0;
1051         int ret = 0;
1052         u32 restart_count = trans->restart_count;
1053
1054         BUG_ON(!path->should_be_locked);
1055
1056         if (watermark == BCH_WATERMARK_copygc)
1057                 watermark = BCH_WATERMARK_btree_copygc;
1058         if (watermark < BCH_WATERMARK_btree)
1059                 watermark = BCH_WATERMARK_btree;
1060
1061         flags &= ~BCH_WATERMARK_MASK;
1062         flags |= watermark;
1063
1064         if (flags & BTREE_INSERT_JOURNAL_RECLAIM)
1065                 journal_flags |= JOURNAL_RES_GET_NONBLOCK;
1066         journal_flags |= watermark;
1067
1068         while (1) {
1069                 nr_nodes[!!update_level] += 1 + split;
1070                 update_level++;
1071
1072                 ret = bch2_btree_path_upgrade(trans, path, update_level + 1);
1073                 if (ret)
1074                         return ERR_PTR(ret);
1075
1076                 if (!btree_path_node(path, update_level)) {
1077                         /* Allocating new root? */
1078                         nr_nodes[1] += split;
1079                         update_level = BTREE_MAX_DEPTH;
1080                         break;
1081                 }
1082
1083                 if (bch2_btree_node_insert_fits(c, path->l[update_level].b,
1084                                         BKEY_BTREE_PTR_U64s_MAX * (1 + split)))
1085                         break;
1086
1087                 split = path->l[update_level].b->nr.live_u64s > BTREE_SPLIT_THRESHOLD(c);
1088         }
1089
1090         if (flags & BTREE_INSERT_GC_LOCK_HELD)
1091                 lockdep_assert_held(&c->gc_lock);
1092         else if (!down_read_trylock(&c->gc_lock)) {
1093                 ret = drop_locks_do(trans, (down_read(&c->gc_lock), 0));
1094                 if (ret) {
1095                         up_read(&c->gc_lock);
1096                         return ERR_PTR(ret);
1097                 }
1098         }
1099
1100         as = mempool_alloc(&c->btree_interior_update_pool, GFP_NOFS);
1101         memset(as, 0, sizeof(*as));
1102         closure_init(&as->cl, NULL);
1103         as->c           = c;
1104         as->start_time  = start_time;
1105         as->mode        = BTREE_INTERIOR_NO_UPDATE;
1106         as->took_gc_lock = !(flags & BTREE_INSERT_GC_LOCK_HELD);
1107         as->btree_id    = path->btree_id;
1108         as->update_level = update_level;
1109         INIT_LIST_HEAD(&as->list);
1110         INIT_LIST_HEAD(&as->unwritten_list);
1111         INIT_LIST_HEAD(&as->write_blocked_list);
1112         bch2_keylist_init(&as->old_keys, as->_old_keys);
1113         bch2_keylist_init(&as->new_keys, as->_new_keys);
1114         bch2_keylist_init(&as->parent_keys, as->inline_keys);
1115
1116         mutex_lock(&c->btree_interior_update_lock);
1117         list_add_tail(&as->list, &c->btree_interior_update_list);
1118         mutex_unlock(&c->btree_interior_update_lock);
1119
1120         /*
1121          * We don't want to allocate if we're in an error state, that can cause
1122          * deadlock on emergency shutdown due to open buckets getting stuck in
1123          * the btree_reserve_cache after allocator shutdown has cleared it out.
1124          * This check needs to come after adding us to the btree_interior_update
1125          * list but before calling bch2_btree_reserve_get, to synchronize with
1126          * __bch2_fs_read_only().
1127          */
1128         ret = bch2_journal_error(&c->journal);
1129         if (ret)
1130                 goto err;
1131
1132         ret = bch2_journal_preres_get(&c->journal, &as->journal_preres,
1133                                       BTREE_UPDATE_JOURNAL_RES,
1134                                       journal_flags|JOURNAL_RES_GET_NONBLOCK);
1135         if (ret) {
1136                 if (flags & BTREE_INSERT_JOURNAL_RECLAIM) {
1137                         ret = -BCH_ERR_journal_reclaim_would_deadlock;
1138                         goto err;
1139                 }
1140
1141                 ret = drop_locks_do(trans,
1142                         bch2_journal_preres_get(&c->journal, &as->journal_preres,
1143                                               BTREE_UPDATE_JOURNAL_RES,
1144                                               journal_flags));
1145                 if (ret == -BCH_ERR_journal_preres_get_blocked) {
1146                         trace_and_count(c, trans_restart_journal_preres_get, trans, _RET_IP_, journal_flags);
1147                         ret = btree_trans_restart(trans, BCH_ERR_transaction_restart_journal_preres_get);
1148                 }
1149                 if (ret)
1150                         goto err;
1151         }
1152
1153         ret = bch2_disk_reservation_get(c, &as->disk_res,
1154                         (nr_nodes[0] + nr_nodes[1]) * btree_sectors(c),
1155                         c->opts.metadata_replicas,
1156                         disk_res_flags);
1157         if (ret)
1158                 goto err;
1159
1160         ret = bch2_btree_reserve_get(trans, as, nr_nodes, flags, NULL);
1161         if (bch2_err_matches(ret, ENOSPC) ||
1162             bch2_err_matches(ret, ENOMEM)) {
1163                 struct closure cl;
1164
1165                 /*
1166                  * XXX: this should probably be a separate BTREE_INSERT_NONBLOCK
1167                  * flag
1168                  */
1169                 if (bch2_err_matches(ret, ENOSPC) &&
1170                     (flags & BTREE_INSERT_JOURNAL_RECLAIM) &&
1171                     watermark != BCH_WATERMARK_reclaim) {
1172                         ret = -BCH_ERR_journal_reclaim_would_deadlock;
1173                         goto err;
1174                 }
1175
1176                 closure_init_stack(&cl);
1177
1178                 do {
1179                         ret = bch2_btree_reserve_get(trans, as, nr_nodes, flags, &cl);
1180
1181                         bch2_trans_unlock(trans);
1182                         closure_sync(&cl);
1183                 } while (bch2_err_matches(ret, BCH_ERR_operation_blocked));
1184         }
1185
1186         if (ret) {
1187                 trace_and_count(c, btree_reserve_get_fail, trans->fn,
1188                                 _RET_IP_, nr_nodes[0] + nr_nodes[1], ret);
1189                 goto err;
1190         }
1191
1192         ret = bch2_trans_relock(trans);
1193         if (ret)
1194                 goto err;
1195
1196         bch2_trans_verify_not_restarted(trans, restart_count);
1197         return as;
1198 err:
1199         bch2_btree_update_free(as, trans);
1200         return ERR_PTR(ret);
1201 }
1202
1203 /* Btree root updates: */
1204
1205 static void bch2_btree_set_root_inmem(struct bch_fs *c, struct btree *b)
1206 {
1207         /* Root nodes cannot be reaped */
1208         mutex_lock(&c->btree_cache.lock);
1209         list_del_init(&b->list);
1210         mutex_unlock(&c->btree_cache.lock);
1211
1212         mutex_lock(&c->btree_root_lock);
1213         BUG_ON(btree_node_root(c, b) &&
1214                (b->c.level < btree_node_root(c, b)->c.level ||
1215                 !btree_node_dying(btree_node_root(c, b))));
1216
1217         bch2_btree_id_root(c, b->c.btree_id)->b = b;
1218         mutex_unlock(&c->btree_root_lock);
1219
1220         bch2_recalc_btree_reserve(c);
1221 }
1222
1223 static void bch2_btree_set_root(struct btree_update *as,
1224                                 struct btree_trans *trans,
1225                                 struct btree_path *path,
1226                                 struct btree *b)
1227 {
1228         struct bch_fs *c = as->c;
1229         struct btree *old;
1230
1231         trace_and_count(c, btree_node_set_root, c, b);
1232
1233         old = btree_node_root(c, b);
1234
1235         /*
1236          * Ensure no one is using the old root while we switch to the
1237          * new root:
1238          */
1239         bch2_btree_node_lock_write_nofail(trans, path, &old->c);
1240
1241         bch2_btree_set_root_inmem(c, b);
1242
1243         btree_update_updated_root(as, b);
1244
1245         /*
1246          * Unlock old root after new root is visible:
1247          *
1248          * The new root isn't persistent, but that's ok: we still have
1249          * an intent lock on the new root, and any updates that would
1250          * depend on the new root would have to update the new root.
1251          */
1252         bch2_btree_node_unlock_write(trans, path, old);
1253 }
1254
1255 /* Interior node updates: */
1256
1257 static void bch2_insert_fixup_btree_ptr(struct btree_update *as,
1258                                         struct btree_trans *trans,
1259                                         struct btree_path *path,
1260                                         struct btree *b,
1261                                         struct btree_node_iter *node_iter,
1262                                         struct bkey_i *insert)
1263 {
1264         struct bch_fs *c = as->c;
1265         struct bkey_packed *k;
1266         struct printbuf buf = PRINTBUF;
1267         unsigned long old, new, v;
1268
1269         BUG_ON(insert->k.type == KEY_TYPE_btree_ptr_v2 &&
1270                !btree_ptr_sectors_written(insert));
1271
1272         if (unlikely(!test_bit(JOURNAL_REPLAY_DONE, &c->journal.flags)))
1273                 bch2_journal_key_overwritten(c, b->c.btree_id, b->c.level, insert->k.p);
1274
1275         if (bch2_bkey_invalid(c, bkey_i_to_s_c(insert),
1276                               btree_node_type(b), WRITE, &buf) ?:
1277             bch2_bkey_in_btree_node(b, bkey_i_to_s_c(insert), &buf)) {
1278                 printbuf_reset(&buf);
1279                 prt_printf(&buf, "inserting invalid bkey\n  ");
1280                 bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(insert));
1281                 prt_printf(&buf, "\n  ");
1282                 bch2_bkey_invalid(c, bkey_i_to_s_c(insert),
1283                                   btree_node_type(b), WRITE, &buf);
1284                 bch2_bkey_in_btree_node(b, bkey_i_to_s_c(insert), &buf);
1285
1286                 bch2_fs_inconsistent(c, "%s", buf.buf);
1287                 dump_stack();
1288         }
1289
1290         BUG_ON(as->journal_u64s + jset_u64s(insert->k.u64s) >
1291                ARRAY_SIZE(as->journal_entries));
1292
1293         as->journal_u64s +=
1294                 journal_entry_set((void *) &as->journal_entries[as->journal_u64s],
1295                                   BCH_JSET_ENTRY_btree_keys,
1296                                   b->c.btree_id, b->c.level,
1297                                   insert, insert->k.u64s);
1298
1299         while ((k = bch2_btree_node_iter_peek_all(node_iter, b)) &&
1300                bkey_iter_pos_cmp(b, k, &insert->k.p) < 0)
1301                 bch2_btree_node_iter_advance(node_iter, b);
1302
1303         bch2_btree_bset_insert_key(trans, path, b, node_iter, insert);
1304         set_btree_node_dirty_acct(c, b);
1305
1306         v = READ_ONCE(b->flags);
1307         do {
1308                 old = new = v;
1309
1310                 new &= ~BTREE_WRITE_TYPE_MASK;
1311                 new |= BTREE_WRITE_interior;
1312                 new |= 1 << BTREE_NODE_need_write;
1313         } while ((v = cmpxchg(&b->flags, old, new)) != old);
1314
1315         printbuf_exit(&buf);
1316 }
1317
1318 static void
1319 __bch2_btree_insert_keys_interior(struct btree_update *as,
1320                                   struct btree_trans *trans,
1321                                   struct btree_path *path,
1322                                   struct btree *b,
1323                                   struct btree_node_iter node_iter,
1324                                   struct keylist *keys)
1325 {
1326         struct bkey_i *insert = bch2_keylist_front(keys);
1327         struct bkey_packed *k;
1328
1329         BUG_ON(btree_node_type(b) != BKEY_TYPE_btree);
1330
1331         while ((k = bch2_btree_node_iter_prev_all(&node_iter, b)) &&
1332                (bkey_cmp_left_packed(b, k, &insert->k.p) >= 0))
1333                 ;
1334
1335         while (!bch2_keylist_empty(keys)) {
1336                 insert = bch2_keylist_front(keys);
1337
1338                 if (bpos_gt(insert->k.p, b->key.k.p))
1339                         break;
1340
1341                 bch2_insert_fixup_btree_ptr(as, trans, path, b, &node_iter, insert);
1342                 bch2_keylist_pop_front(keys);
1343         }
1344 }
1345
1346 /*
1347  * Move keys from n1 (original replacement node, now lower node) to n2 (higher
1348  * node)
1349  */
1350 static void __btree_split_node(struct btree_update *as,
1351                                struct btree_trans *trans,
1352                                struct btree *b,
1353                                struct btree *n[2])
1354 {
1355         struct bkey_packed *k;
1356         struct bpos n1_pos = POS_MIN;
1357         struct btree_node_iter iter;
1358         struct bset *bsets[2];
1359         struct bkey_format_state format[2];
1360         struct bkey_packed *out[2];
1361         struct bkey uk;
1362         unsigned u64s, n1_u64s = (b->nr.live_u64s * 3) / 5;
1363         int i;
1364
1365         for (i = 0; i < 2; i++) {
1366                 BUG_ON(n[i]->nsets != 1);
1367
1368                 bsets[i] = btree_bset_first(n[i]);
1369                 out[i] = bsets[i]->start;
1370
1371                 SET_BTREE_NODE_SEQ(n[i]->data, BTREE_NODE_SEQ(b->data) + 1);
1372                 bch2_bkey_format_init(&format[i]);
1373         }
1374
1375         u64s = 0;
1376         for_each_btree_node_key(b, k, &iter) {
1377                 if (bkey_deleted(k))
1378                         continue;
1379
1380                 i = u64s >= n1_u64s;
1381                 u64s += k->u64s;
1382                 uk = bkey_unpack_key(b, k);
1383                 if (!i)
1384                         n1_pos = uk.p;
1385                 bch2_bkey_format_add_key(&format[i], &uk);
1386         }
1387
1388         btree_set_min(n[0], b->data->min_key);
1389         btree_set_max(n[0], n1_pos);
1390         btree_set_min(n[1], bpos_successor(n1_pos));
1391         btree_set_max(n[1], b->data->max_key);
1392
1393         for (i = 0; i < 2; i++) {
1394                 bch2_bkey_format_add_pos(&format[i], n[i]->data->min_key);
1395                 bch2_bkey_format_add_pos(&format[i], n[i]->data->max_key);
1396
1397                 n[i]->data->format = bch2_bkey_format_done(&format[i]);
1398                 btree_node_set_format(n[i], n[i]->data->format);
1399         }
1400
1401         u64s = 0;
1402         for_each_btree_node_key(b, k, &iter) {
1403                 if (bkey_deleted(k))
1404                         continue;
1405
1406                 i = u64s >= n1_u64s;
1407                 u64s += k->u64s;
1408
1409                 if (bch2_bkey_transform(&n[i]->format, out[i], bkey_packed(k)
1410                                         ? &b->format: &bch2_bkey_format_current, k))
1411                         out[i]->format = KEY_FORMAT_LOCAL_BTREE;
1412                 else
1413                         bch2_bkey_unpack(b, (void *) out[i], k);
1414
1415                 out[i]->needs_whiteout = false;
1416
1417                 btree_keys_account_key_add(&n[i]->nr, 0, out[i]);
1418                 out[i] = bkey_p_next(out[i]);
1419         }
1420
1421         for (i = 0; i < 2; i++) {
1422                 bsets[i]->u64s = cpu_to_le16((u64 *) out[i] - bsets[i]->_data);
1423
1424                 BUG_ON(!bsets[i]->u64s);
1425
1426                 set_btree_bset_end(n[i], n[i]->set);
1427
1428                 btree_node_reset_sib_u64s(n[i]);
1429
1430                 bch2_verify_btree_nr_keys(n[i]);
1431
1432                 if (b->c.level)
1433                         btree_node_interior_verify(as->c, n[i]);
1434         }
1435 }
1436
1437 /*
1438  * For updates to interior nodes, we've got to do the insert before we split
1439  * because the stuff we're inserting has to be inserted atomically. Post split,
1440  * the keys might have to go in different nodes and the split would no longer be
1441  * atomic.
1442  *
1443  * Worse, if the insert is from btree node coalescing, if we do the insert after
1444  * we do the split (and pick the pivot) - the pivot we pick might be between
1445  * nodes that were coalesced, and thus in the middle of a child node post
1446  * coalescing:
1447  */
1448 static void btree_split_insert_keys(struct btree_update *as,
1449                                     struct btree_trans *trans,
1450                                     struct btree_path *path,
1451                                     struct btree *b,
1452                                     struct keylist *keys)
1453 {
1454         if (!bch2_keylist_empty(keys) &&
1455             bpos_le(bch2_keylist_front(keys)->k.p, b->data->max_key)) {
1456                 struct btree_node_iter node_iter;
1457
1458                 bch2_btree_node_iter_init(&node_iter, b, &bch2_keylist_front(keys)->k.p);
1459
1460                 __bch2_btree_insert_keys_interior(as, trans, path, b, node_iter, keys);
1461
1462                 btree_node_interior_verify(as->c, b);
1463         }
1464 }
1465
1466 static int btree_split(struct btree_update *as, struct btree_trans *trans,
1467                        struct btree_path *path, struct btree *b,
1468                        struct keylist *keys, unsigned flags)
1469 {
1470         struct bch_fs *c = as->c;
1471         struct btree *parent = btree_node_parent(path, b);
1472         struct btree *n1, *n2 = NULL, *n3 = NULL;
1473         struct btree_path *path1 = NULL, *path2 = NULL;
1474         u64 start_time = local_clock();
1475         int ret = 0;
1476
1477         BUG_ON(!parent && (b != btree_node_root(c, b)));
1478         BUG_ON(parent && !btree_node_intent_locked(path, b->c.level + 1));
1479
1480         bch2_btree_interior_update_will_free_node(as, b);
1481
1482         if (b->nr.live_u64s > BTREE_SPLIT_THRESHOLD(c)) {
1483                 struct btree *n[2];
1484
1485                 trace_and_count(c, btree_node_split, c, b);
1486
1487                 n[0] = n1 = bch2_btree_node_alloc(as, trans, b->c.level);
1488                 n[1] = n2 = bch2_btree_node_alloc(as, trans, b->c.level);
1489
1490                 __btree_split_node(as, trans, b, n);
1491
1492                 if (keys) {
1493                         btree_split_insert_keys(as, trans, path, n1, keys);
1494                         btree_split_insert_keys(as, trans, path, n2, keys);
1495                         BUG_ON(!bch2_keylist_empty(keys));
1496                 }
1497
1498                 bch2_btree_build_aux_trees(n2);
1499                 bch2_btree_build_aux_trees(n1);
1500
1501                 bch2_btree_update_add_new_node(as, n1);
1502                 bch2_btree_update_add_new_node(as, n2);
1503                 six_unlock_write(&n2->c.lock);
1504                 six_unlock_write(&n1->c.lock);
1505
1506                 path1 = get_unlocked_mut_path(trans, path->btree_id, n1->c.level, n1->key.k.p);
1507                 six_lock_increment(&n1->c.lock, SIX_LOCK_intent);
1508                 mark_btree_node_locked(trans, path1, n1->c.level, BTREE_NODE_INTENT_LOCKED);
1509                 bch2_btree_path_level_init(trans, path1, n1);
1510
1511                 path2 = get_unlocked_mut_path(trans, path->btree_id, n2->c.level, n2->key.k.p);
1512                 six_lock_increment(&n2->c.lock, SIX_LOCK_intent);
1513                 mark_btree_node_locked(trans, path2, n2->c.level, BTREE_NODE_INTENT_LOCKED);
1514                 bch2_btree_path_level_init(trans, path2, n2);
1515
1516                 /*
1517                  * Note that on recursive parent_keys == keys, so we
1518                  * can't start adding new keys to parent_keys before emptying it
1519                  * out (which we did with btree_split_insert_keys() above)
1520                  */
1521                 bch2_keylist_add(&as->parent_keys, &n1->key);
1522                 bch2_keylist_add(&as->parent_keys, &n2->key);
1523
1524                 if (!parent) {
1525                         /* Depth increases, make a new root */
1526                         n3 = __btree_root_alloc(as, trans, b->c.level + 1);
1527
1528                         bch2_btree_update_add_new_node(as, n3);
1529                         six_unlock_write(&n3->c.lock);
1530
1531                         path2->locks_want++;
1532                         BUG_ON(btree_node_locked(path2, n3->c.level));
1533                         six_lock_increment(&n3->c.lock, SIX_LOCK_intent);
1534                         mark_btree_node_locked(trans, path2, n3->c.level, BTREE_NODE_INTENT_LOCKED);
1535                         bch2_btree_path_level_init(trans, path2, n3);
1536
1537                         n3->sib_u64s[0] = U16_MAX;
1538                         n3->sib_u64s[1] = U16_MAX;
1539
1540                         btree_split_insert_keys(as, trans, path, n3, &as->parent_keys);
1541                 }
1542         } else {
1543                 trace_and_count(c, btree_node_compact, c, b);
1544
1545                 n1 = bch2_btree_node_alloc_replacement(as, trans, b);
1546
1547                 if (keys) {
1548                         btree_split_insert_keys(as, trans, path, n1, keys);
1549                         BUG_ON(!bch2_keylist_empty(keys));
1550                 }
1551
1552                 bch2_btree_build_aux_trees(n1);
1553                 bch2_btree_update_add_new_node(as, n1);
1554                 six_unlock_write(&n1->c.lock);
1555
1556                 path1 = get_unlocked_mut_path(trans, path->btree_id, n1->c.level, n1->key.k.p);
1557                 six_lock_increment(&n1->c.lock, SIX_LOCK_intent);
1558                 mark_btree_node_locked(trans, path1, n1->c.level, BTREE_NODE_INTENT_LOCKED);
1559                 bch2_btree_path_level_init(trans, path1, n1);
1560
1561                 if (parent)
1562                         bch2_keylist_add(&as->parent_keys, &n1->key);
1563         }
1564
1565         /* New nodes all written, now make them visible: */
1566
1567         if (parent) {
1568                 /* Split a non root node */
1569                 ret = bch2_btree_insert_node(as, trans, path, parent, &as->parent_keys, flags);
1570                 if (ret)
1571                         goto err;
1572         } else if (n3) {
1573                 bch2_btree_set_root(as, trans, path, n3);
1574         } else {
1575                 /* Root filled up but didn't need to be split */
1576                 bch2_btree_set_root(as, trans, path, n1);
1577         }
1578
1579         if (n3) {
1580                 bch2_btree_update_get_open_buckets(as, n3);
1581                 bch2_btree_node_write(c, n3, SIX_LOCK_intent, 0);
1582         }
1583         if (n2) {
1584                 bch2_btree_update_get_open_buckets(as, n2);
1585                 bch2_btree_node_write(c, n2, SIX_LOCK_intent, 0);
1586         }
1587         bch2_btree_update_get_open_buckets(as, n1);
1588         bch2_btree_node_write(c, n1, SIX_LOCK_intent, 0);
1589
1590         /*
1591          * The old node must be freed (in memory) _before_ unlocking the new
1592          * nodes - else another thread could re-acquire a read lock on the old
1593          * node after another thread has locked and updated the new node, thus
1594          * seeing stale data:
1595          */
1596         bch2_btree_node_free_inmem(trans, path, b);
1597
1598         if (n3)
1599                 bch2_trans_node_add(trans, n3);
1600         if (n2)
1601                 bch2_trans_node_add(trans, n2);
1602         bch2_trans_node_add(trans, n1);
1603
1604         if (n3)
1605                 six_unlock_intent(&n3->c.lock);
1606         if (n2)
1607                 six_unlock_intent(&n2->c.lock);
1608         six_unlock_intent(&n1->c.lock);
1609 out:
1610         if (path2) {
1611                 __bch2_btree_path_unlock(trans, path2);
1612                 bch2_path_put(trans, path2, true);
1613         }
1614         if (path1) {
1615                 __bch2_btree_path_unlock(trans, path1);
1616                 bch2_path_put(trans, path1, true);
1617         }
1618
1619         bch2_trans_verify_locks(trans);
1620
1621         bch2_time_stats_update(&c->times[n2
1622                                ? BCH_TIME_btree_node_split
1623                                : BCH_TIME_btree_node_compact],
1624                                start_time);
1625         return ret;
1626 err:
1627         if (n3)
1628                 bch2_btree_node_free_never_used(as, trans, n3);
1629         if (n2)
1630                 bch2_btree_node_free_never_used(as, trans, n2);
1631         bch2_btree_node_free_never_used(as, trans, n1);
1632         goto out;
1633 }
1634
1635 static void
1636 bch2_btree_insert_keys_interior(struct btree_update *as,
1637                                 struct btree_trans *trans,
1638                                 struct btree_path *path,
1639                                 struct btree *b,
1640                                 struct keylist *keys)
1641 {
1642         struct btree_path *linked;
1643
1644         __bch2_btree_insert_keys_interior(as, trans, path, b,
1645                                           path->l[b->c.level].iter, keys);
1646
1647         btree_update_updated_node(as, b);
1648
1649         trans_for_each_path_with_node(trans, b, linked)
1650                 bch2_btree_node_iter_peek(&linked->l[b->c.level].iter, b);
1651
1652         bch2_trans_verify_paths(trans);
1653 }
1654
1655 /**
1656  * bch2_btree_insert_node - insert bkeys into a given btree node
1657  *
1658  * @as:                 btree_update object
1659  * @trans:              btree_trans object
1660  * @path:               path that points to current node
1661  * @b:                  node to insert keys into
1662  * @keys:               list of keys to insert
1663  * @flags:              transaction commit flags
1664  *
1665  * Returns: 0 on success, typically transaction restart error on failure
1666  *
1667  * Inserts as many keys as it can into a given btree node, splitting it if full.
1668  * If a split occurred, this function will return early. This can only happen
1669  * for leaf nodes -- inserts into interior nodes have to be atomic.
1670  */
1671 static int bch2_btree_insert_node(struct btree_update *as, struct btree_trans *trans,
1672                                   struct btree_path *path, struct btree *b,
1673                                   struct keylist *keys, unsigned flags)
1674 {
1675         struct bch_fs *c = as->c;
1676         int old_u64s = le16_to_cpu(btree_bset_last(b)->u64s);
1677         int old_live_u64s = b->nr.live_u64s;
1678         int live_u64s_added, u64s_added;
1679         int ret;
1680
1681         lockdep_assert_held(&c->gc_lock);
1682         BUG_ON(!btree_node_intent_locked(path, b->c.level));
1683         BUG_ON(!b->c.level);
1684         BUG_ON(!as || as->b);
1685         bch2_verify_keylist_sorted(keys);
1686
1687         ret = bch2_btree_node_lock_write(trans, path, &b->c);
1688         if (ret)
1689                 return ret;
1690
1691         bch2_btree_node_prep_for_write(trans, path, b);
1692
1693         if (!bch2_btree_node_insert_fits(c, b, bch2_keylist_u64s(keys))) {
1694                 bch2_btree_node_unlock_write(trans, path, b);
1695                 goto split;
1696         }
1697
1698         btree_node_interior_verify(c, b);
1699
1700         bch2_btree_insert_keys_interior(as, trans, path, b, keys);
1701
1702         live_u64s_added = (int) b->nr.live_u64s - old_live_u64s;
1703         u64s_added = (int) le16_to_cpu(btree_bset_last(b)->u64s) - old_u64s;
1704
1705         if (b->sib_u64s[0] != U16_MAX && live_u64s_added < 0)
1706                 b->sib_u64s[0] = max(0, (int) b->sib_u64s[0] + live_u64s_added);
1707         if (b->sib_u64s[1] != U16_MAX && live_u64s_added < 0)
1708                 b->sib_u64s[1] = max(0, (int) b->sib_u64s[1] + live_u64s_added);
1709
1710         if (u64s_added > live_u64s_added &&
1711             bch2_maybe_compact_whiteouts(c, b))
1712                 bch2_trans_node_reinit_iter(trans, b);
1713
1714         bch2_btree_node_unlock_write(trans, path, b);
1715
1716         btree_node_interior_verify(c, b);
1717         return 0;
1718 split:
1719         /*
1720          * We could attempt to avoid the transaction restart, by calling
1721          * bch2_btree_path_upgrade() and allocating more nodes:
1722          */
1723         if (b->c.level >= as->update_level) {
1724                 trace_and_count(c, trans_restart_split_race, trans, _THIS_IP_, b);
1725                 return btree_trans_restart(trans, BCH_ERR_transaction_restart_split_race);
1726         }
1727
1728         return btree_split(as, trans, path, b, keys, flags);
1729 }
1730
1731 int bch2_btree_split_leaf(struct btree_trans *trans,
1732                           struct btree_path *path,
1733                           unsigned flags)
1734 {
1735         struct btree *b = path_l(path)->b;
1736         struct btree_update *as;
1737         unsigned l;
1738         int ret = 0;
1739
1740         as = bch2_btree_update_start(trans, path, path->level,
1741                                      true, flags);
1742         if (IS_ERR(as))
1743                 return PTR_ERR(as);
1744
1745         ret = btree_split(as, trans, path, b, NULL, flags);
1746         if (ret) {
1747                 bch2_btree_update_free(as, trans);
1748                 return ret;
1749         }
1750
1751         bch2_btree_update_done(as, trans);
1752
1753         for (l = path->level + 1; btree_node_intent_locked(path, l) && !ret; l++)
1754                 ret = bch2_foreground_maybe_merge(trans, path, l, flags);
1755
1756         return ret;
1757 }
1758
1759 int __bch2_foreground_maybe_merge(struct btree_trans *trans,
1760                                   struct btree_path *path,
1761                                   unsigned level,
1762                                   unsigned flags,
1763                                   enum btree_node_sibling sib)
1764 {
1765         struct bch_fs *c = trans->c;
1766         struct btree_path *sib_path = NULL, *new_path = NULL;
1767         struct btree_update *as;
1768         struct bkey_format_state new_s;
1769         struct bkey_format new_f;
1770         struct bkey_i delete;
1771         struct btree *b, *m, *n, *prev, *next, *parent;
1772         struct bpos sib_pos;
1773         size_t sib_u64s;
1774         u64 start_time = local_clock();
1775         int ret = 0;
1776
1777         BUG_ON(!path->should_be_locked);
1778         BUG_ON(!btree_node_locked(path, level));
1779
1780         b = path->l[level].b;
1781
1782         if ((sib == btree_prev_sib && bpos_eq(b->data->min_key, POS_MIN)) ||
1783             (sib == btree_next_sib && bpos_eq(b->data->max_key, SPOS_MAX))) {
1784                 b->sib_u64s[sib] = U16_MAX;
1785                 return 0;
1786         }
1787
1788         sib_pos = sib == btree_prev_sib
1789                 ? bpos_predecessor(b->data->min_key)
1790                 : bpos_successor(b->data->max_key);
1791
1792         sib_path = bch2_path_get(trans, path->btree_id, sib_pos,
1793                                  U8_MAX, level, BTREE_ITER_INTENT, _THIS_IP_);
1794         ret = bch2_btree_path_traverse(trans, sib_path, false);
1795         if (ret)
1796                 goto err;
1797
1798         btree_path_set_should_be_locked(sib_path);
1799
1800         m = sib_path->l[level].b;
1801
1802         if (btree_node_parent(path, b) !=
1803             btree_node_parent(sib_path, m)) {
1804                 b->sib_u64s[sib] = U16_MAX;
1805                 goto out;
1806         }
1807
1808         if (sib == btree_prev_sib) {
1809                 prev = m;
1810                 next = b;
1811         } else {
1812                 prev = b;
1813                 next = m;
1814         }
1815
1816         if (!bpos_eq(bpos_successor(prev->data->max_key), next->data->min_key)) {
1817                 struct printbuf buf1 = PRINTBUF, buf2 = PRINTBUF;
1818
1819                 bch2_bpos_to_text(&buf1, prev->data->max_key);
1820                 bch2_bpos_to_text(&buf2, next->data->min_key);
1821                 bch_err(c,
1822                         "%s(): btree topology error:\n"
1823                         "  prev ends at   %s\n"
1824                         "  next starts at %s",
1825                         __func__, buf1.buf, buf2.buf);
1826                 printbuf_exit(&buf1);
1827                 printbuf_exit(&buf2);
1828                 bch2_topology_error(c);
1829                 ret = -EIO;
1830                 goto err;
1831         }
1832
1833         bch2_bkey_format_init(&new_s);
1834         bch2_bkey_format_add_pos(&new_s, prev->data->min_key);
1835         __bch2_btree_calc_format(&new_s, prev);
1836         __bch2_btree_calc_format(&new_s, next);
1837         bch2_bkey_format_add_pos(&new_s, next->data->max_key);
1838         new_f = bch2_bkey_format_done(&new_s);
1839
1840         sib_u64s = btree_node_u64s_with_format(b, &new_f) +
1841                 btree_node_u64s_with_format(m, &new_f);
1842
1843         if (sib_u64s > BTREE_FOREGROUND_MERGE_HYSTERESIS(c)) {
1844                 sib_u64s -= BTREE_FOREGROUND_MERGE_HYSTERESIS(c);
1845                 sib_u64s /= 2;
1846                 sib_u64s += BTREE_FOREGROUND_MERGE_HYSTERESIS(c);
1847         }
1848
1849         sib_u64s = min(sib_u64s, btree_max_u64s(c));
1850         sib_u64s = min(sib_u64s, (size_t) U16_MAX - 1);
1851         b->sib_u64s[sib] = sib_u64s;
1852
1853         if (b->sib_u64s[sib] > c->btree_foreground_merge_threshold)
1854                 goto out;
1855
1856         parent = btree_node_parent(path, b);
1857         as = bch2_btree_update_start(trans, path, level, false,
1858                                      BTREE_INSERT_NOFAIL|flags);
1859         ret = PTR_ERR_OR_ZERO(as);
1860         if (ret)
1861                 goto err;
1862
1863         trace_and_count(c, btree_node_merge, c, b);
1864
1865         bch2_btree_interior_update_will_free_node(as, b);
1866         bch2_btree_interior_update_will_free_node(as, m);
1867
1868         n = bch2_btree_node_alloc(as, trans, b->c.level);
1869
1870         SET_BTREE_NODE_SEQ(n->data,
1871                            max(BTREE_NODE_SEQ(b->data),
1872                                BTREE_NODE_SEQ(m->data)) + 1);
1873
1874         btree_set_min(n, prev->data->min_key);
1875         btree_set_max(n, next->data->max_key);
1876
1877         n->data->format  = new_f;
1878         btree_node_set_format(n, new_f);
1879
1880         bch2_btree_sort_into(c, n, prev);
1881         bch2_btree_sort_into(c, n, next);
1882
1883         bch2_btree_build_aux_trees(n);
1884         bch2_btree_update_add_new_node(as, n);
1885         six_unlock_write(&n->c.lock);
1886
1887         new_path = get_unlocked_mut_path(trans, path->btree_id, n->c.level, n->key.k.p);
1888         six_lock_increment(&n->c.lock, SIX_LOCK_intent);
1889         mark_btree_node_locked(trans, new_path, n->c.level, BTREE_NODE_INTENT_LOCKED);
1890         bch2_btree_path_level_init(trans, new_path, n);
1891
1892         bkey_init(&delete.k);
1893         delete.k.p = prev->key.k.p;
1894         bch2_keylist_add(&as->parent_keys, &delete);
1895         bch2_keylist_add(&as->parent_keys, &n->key);
1896
1897         bch2_trans_verify_paths(trans);
1898
1899         ret = bch2_btree_insert_node(as, trans, path, parent, &as->parent_keys, flags);
1900         if (ret)
1901                 goto err_free_update;
1902
1903         bch2_trans_verify_paths(trans);
1904
1905         bch2_btree_update_get_open_buckets(as, n);
1906         bch2_btree_node_write(c, n, SIX_LOCK_intent, 0);
1907
1908         bch2_btree_node_free_inmem(trans, path, b);
1909         bch2_btree_node_free_inmem(trans, sib_path, m);
1910
1911         bch2_trans_node_add(trans, n);
1912
1913         bch2_trans_verify_paths(trans);
1914
1915         six_unlock_intent(&n->c.lock);
1916
1917         bch2_btree_update_done(as, trans);
1918
1919         bch2_time_stats_update(&c->times[BCH_TIME_btree_node_merge], start_time);
1920 out:
1921 err:
1922         if (new_path)
1923                 bch2_path_put(trans, new_path, true);
1924         bch2_path_put(trans, sib_path, true);
1925         bch2_trans_verify_locks(trans);
1926         return ret;
1927 err_free_update:
1928         bch2_btree_node_free_never_used(as, trans, n);
1929         bch2_btree_update_free(as, trans);
1930         goto out;
1931 }
1932
1933 int bch2_btree_node_rewrite(struct btree_trans *trans,
1934                             struct btree_iter *iter,
1935                             struct btree *b,
1936                             unsigned flags)
1937 {
1938         struct bch_fs *c = trans->c;
1939         struct btree_path *new_path = NULL;
1940         struct btree *n, *parent;
1941         struct btree_update *as;
1942         int ret;
1943
1944         flags |= BTREE_INSERT_NOFAIL;
1945
1946         parent = btree_node_parent(iter->path, b);
1947         as = bch2_btree_update_start(trans, iter->path, b->c.level,
1948                                      false, flags);
1949         ret = PTR_ERR_OR_ZERO(as);
1950         if (ret)
1951                 goto out;
1952
1953         bch2_btree_interior_update_will_free_node(as, b);
1954
1955         n = bch2_btree_node_alloc_replacement(as, trans, b);
1956
1957         bch2_btree_build_aux_trees(n);
1958         bch2_btree_update_add_new_node(as, n);
1959         six_unlock_write(&n->c.lock);
1960
1961         new_path = get_unlocked_mut_path(trans, iter->btree_id, n->c.level, n->key.k.p);
1962         six_lock_increment(&n->c.lock, SIX_LOCK_intent);
1963         mark_btree_node_locked(trans, new_path, n->c.level, BTREE_NODE_INTENT_LOCKED);
1964         bch2_btree_path_level_init(trans, new_path, n);
1965
1966         trace_and_count(c, btree_node_rewrite, c, b);
1967
1968         if (parent) {
1969                 bch2_keylist_add(&as->parent_keys, &n->key);
1970                 ret = bch2_btree_insert_node(as, trans, iter->path, parent,
1971                                              &as->parent_keys, flags);
1972                 if (ret)
1973                         goto err;
1974         } else {
1975                 bch2_btree_set_root(as, trans, iter->path, n);
1976         }
1977
1978         bch2_btree_update_get_open_buckets(as, n);
1979         bch2_btree_node_write(c, n, SIX_LOCK_intent, 0);
1980
1981         bch2_btree_node_free_inmem(trans, iter->path, b);
1982
1983         bch2_trans_node_add(trans, n);
1984         six_unlock_intent(&n->c.lock);
1985
1986         bch2_btree_update_done(as, trans);
1987 out:
1988         if (new_path)
1989                 bch2_path_put(trans, new_path, true);
1990         bch2_btree_path_downgrade(trans, iter->path);
1991         return ret;
1992 err:
1993         bch2_btree_node_free_never_used(as, trans, n);
1994         bch2_btree_update_free(as, trans);
1995         goto out;
1996 }
1997
1998 struct async_btree_rewrite {
1999         struct bch_fs           *c;
2000         struct work_struct      work;
2001         struct list_head        list;
2002         enum btree_id           btree_id;
2003         unsigned                level;
2004         struct bpos             pos;
2005         __le64                  seq;
2006 };
2007
2008 static int async_btree_node_rewrite_trans(struct btree_trans *trans,
2009                                           struct async_btree_rewrite *a)
2010 {
2011         struct bch_fs *c = trans->c;
2012         struct btree_iter iter;
2013         struct btree *b;
2014         int ret;
2015
2016         bch2_trans_node_iter_init(trans, &iter, a->btree_id, a->pos,
2017                                   BTREE_MAX_DEPTH, a->level, 0);
2018         b = bch2_btree_iter_peek_node(&iter);
2019         ret = PTR_ERR_OR_ZERO(b);
2020         if (ret)
2021                 goto out;
2022
2023         if (!b || b->data->keys.seq != a->seq) {
2024                 struct printbuf buf = PRINTBUF;
2025
2026                 if (b)
2027                         bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(&b->key));
2028                 else
2029                         prt_str(&buf, "(null");
2030                 bch_info(c, "%s: node to rewrite not found:, searching for seq %llu, got\n%s",
2031                          __func__, a->seq, buf.buf);
2032                 printbuf_exit(&buf);
2033                 goto out;
2034         }
2035
2036         ret = bch2_btree_node_rewrite(trans, &iter, b, 0);
2037 out:
2038         bch2_trans_iter_exit(trans, &iter);
2039
2040         return ret;
2041 }
2042
2043 static void async_btree_node_rewrite_work(struct work_struct *work)
2044 {
2045         struct async_btree_rewrite *a =
2046                 container_of(work, struct async_btree_rewrite, work);
2047         struct bch_fs *c = a->c;
2048         int ret;
2049
2050         ret = bch2_trans_do(c, NULL, NULL, 0,
2051                       async_btree_node_rewrite_trans(trans, a));
2052         if (ret)
2053                 bch_err_fn(c, ret);
2054         bch2_write_ref_put(c, BCH_WRITE_REF_node_rewrite);
2055         kfree(a);
2056 }
2057
2058 void bch2_btree_node_rewrite_async(struct bch_fs *c, struct btree *b)
2059 {
2060         struct async_btree_rewrite *a;
2061         int ret;
2062
2063         a = kmalloc(sizeof(*a), GFP_NOFS);
2064         if (!a) {
2065                 bch_err(c, "%s: error allocating memory", __func__);
2066                 return;
2067         }
2068
2069         a->c            = c;
2070         a->btree_id     = b->c.btree_id;
2071         a->level        = b->c.level;
2072         a->pos          = b->key.k.p;
2073         a->seq          = b->data->keys.seq;
2074         INIT_WORK(&a->work, async_btree_node_rewrite_work);
2075
2076         if (unlikely(!test_bit(BCH_FS_MAY_GO_RW, &c->flags))) {
2077                 mutex_lock(&c->pending_node_rewrites_lock);
2078                 list_add(&a->list, &c->pending_node_rewrites);
2079                 mutex_unlock(&c->pending_node_rewrites_lock);
2080                 return;
2081         }
2082
2083         if (!bch2_write_ref_tryget(c, BCH_WRITE_REF_node_rewrite)) {
2084                 if (test_bit(BCH_FS_STARTED, &c->flags)) {
2085                         bch_err(c, "%s: error getting c->writes ref", __func__);
2086                         kfree(a);
2087                         return;
2088                 }
2089
2090                 ret = bch2_fs_read_write_early(c);
2091                 if (ret) {
2092                         bch_err_msg(c, ret, "going read-write");
2093                         kfree(a);
2094                         return;
2095                 }
2096
2097                 bch2_write_ref_get(c, BCH_WRITE_REF_node_rewrite);
2098         }
2099
2100         queue_work(c->btree_interior_update_worker, &a->work);
2101 }
2102
2103 void bch2_do_pending_node_rewrites(struct bch_fs *c)
2104 {
2105         struct async_btree_rewrite *a, *n;
2106
2107         mutex_lock(&c->pending_node_rewrites_lock);
2108         list_for_each_entry_safe(a, n, &c->pending_node_rewrites, list) {
2109                 list_del(&a->list);
2110
2111                 bch2_write_ref_get(c, BCH_WRITE_REF_node_rewrite);
2112                 queue_work(c->btree_interior_update_worker, &a->work);
2113         }
2114         mutex_unlock(&c->pending_node_rewrites_lock);
2115 }
2116
2117 void bch2_free_pending_node_rewrites(struct bch_fs *c)
2118 {
2119         struct async_btree_rewrite *a, *n;
2120
2121         mutex_lock(&c->pending_node_rewrites_lock);
2122         list_for_each_entry_safe(a, n, &c->pending_node_rewrites, list) {
2123                 list_del(&a->list);
2124
2125                 kfree(a);
2126         }
2127         mutex_unlock(&c->pending_node_rewrites_lock);
2128 }
2129
2130 static int __bch2_btree_node_update_key(struct btree_trans *trans,
2131                                         struct btree_iter *iter,
2132                                         struct btree *b, struct btree *new_hash,
2133                                         struct bkey_i *new_key,
2134                                         unsigned commit_flags,
2135                                         bool skip_triggers)
2136 {
2137         struct bch_fs *c = trans->c;
2138         struct btree_iter iter2 = { NULL };
2139         struct btree *parent;
2140         int ret;
2141
2142         if (!skip_triggers) {
2143                 ret = bch2_trans_mark_old(trans, b->c.btree_id, b->c.level + 1,
2144                                           bkey_i_to_s_c(&b->key), 0);
2145                 if (ret)
2146                         return ret;
2147
2148                 ret = bch2_trans_mark_new(trans, b->c.btree_id, b->c.level + 1,
2149                                           new_key, 0);
2150                 if (ret)
2151                         return ret;
2152         }
2153
2154         if (new_hash) {
2155                 bkey_copy(&new_hash->key, new_key);
2156                 ret = bch2_btree_node_hash_insert(&c->btree_cache,
2157                                 new_hash, b->c.level, b->c.btree_id);
2158                 BUG_ON(ret);
2159         }
2160
2161         parent = btree_node_parent(iter->path, b);
2162         if (parent) {
2163                 bch2_trans_copy_iter(&iter2, iter);
2164
2165                 iter2.path = bch2_btree_path_make_mut(trans, iter2.path,
2166                                 iter2.flags & BTREE_ITER_INTENT,
2167                                 _THIS_IP_);
2168
2169                 BUG_ON(iter2.path->level != b->c.level);
2170                 BUG_ON(!bpos_eq(iter2.path->pos, new_key->k.p));
2171
2172                 btree_path_set_level_up(trans, iter2.path);
2173
2174                 trans->paths_sorted = false;
2175
2176                 ret   = bch2_btree_iter_traverse(&iter2) ?:
2177                         bch2_trans_update(trans, &iter2, new_key, BTREE_TRIGGER_NORUN);
2178                 if (ret)
2179                         goto err;
2180         } else {
2181                 BUG_ON(btree_node_root(c, b) != b);
2182
2183                 ret = darray_make_room(&trans->extra_journal_entries,
2184                                        jset_u64s(new_key->k.u64s));
2185                 if (ret)
2186                         return ret;
2187
2188                 journal_entry_set((void *) &darray_top(trans->extra_journal_entries),
2189                                   BCH_JSET_ENTRY_btree_root,
2190                                   b->c.btree_id, b->c.level,
2191                                   new_key, new_key->k.u64s);
2192                 trans->extra_journal_entries.nr += jset_u64s(new_key->k.u64s);
2193         }
2194
2195         ret = bch2_trans_commit(trans, NULL, NULL, commit_flags);
2196         if (ret)
2197                 goto err;
2198
2199         bch2_btree_node_lock_write_nofail(trans, iter->path, &b->c);
2200
2201         if (new_hash) {
2202                 mutex_lock(&c->btree_cache.lock);
2203                 bch2_btree_node_hash_remove(&c->btree_cache, new_hash);
2204                 bch2_btree_node_hash_remove(&c->btree_cache, b);
2205
2206                 bkey_copy(&b->key, new_key);
2207                 ret = __bch2_btree_node_hash_insert(&c->btree_cache, b);
2208                 BUG_ON(ret);
2209                 mutex_unlock(&c->btree_cache.lock);
2210         } else {
2211                 bkey_copy(&b->key, new_key);
2212         }
2213
2214         bch2_btree_node_unlock_write(trans, iter->path, b);
2215 out:
2216         bch2_trans_iter_exit(trans, &iter2);
2217         return ret;
2218 err:
2219         if (new_hash) {
2220                 mutex_lock(&c->btree_cache.lock);
2221                 bch2_btree_node_hash_remove(&c->btree_cache, b);
2222                 mutex_unlock(&c->btree_cache.lock);
2223         }
2224         goto out;
2225 }
2226
2227 int bch2_btree_node_update_key(struct btree_trans *trans, struct btree_iter *iter,
2228                                struct btree *b, struct bkey_i *new_key,
2229                                unsigned commit_flags, bool skip_triggers)
2230 {
2231         struct bch_fs *c = trans->c;
2232         struct btree *new_hash = NULL;
2233         struct btree_path *path = iter->path;
2234         struct closure cl;
2235         int ret = 0;
2236
2237         ret = bch2_btree_path_upgrade(trans, path, b->c.level + 1);
2238         if (ret)
2239                 return ret;
2240
2241         closure_init_stack(&cl);
2242
2243         /*
2244          * check btree_ptr_hash_val() after @b is locked by
2245          * btree_iter_traverse():
2246          */
2247         if (btree_ptr_hash_val(new_key) != b->hash_val) {
2248                 ret = bch2_btree_cache_cannibalize_lock(c, &cl);
2249                 if (ret) {
2250                         ret = drop_locks_do(trans, (closure_sync(&cl), 0));
2251                         if (ret)
2252                                 return ret;
2253                 }
2254
2255                 new_hash = bch2_btree_node_mem_alloc(trans, false);
2256         }
2257
2258         path->intent_ref++;
2259         ret = __bch2_btree_node_update_key(trans, iter, b, new_hash, new_key,
2260                                            commit_flags, skip_triggers);
2261         --path->intent_ref;
2262
2263         if (new_hash) {
2264                 mutex_lock(&c->btree_cache.lock);
2265                 list_move(&new_hash->list, &c->btree_cache.freeable);
2266                 mutex_unlock(&c->btree_cache.lock);
2267
2268                 six_unlock_write(&new_hash->c.lock);
2269                 six_unlock_intent(&new_hash->c.lock);
2270         }
2271         closure_sync(&cl);
2272         bch2_btree_cache_cannibalize_unlock(c);
2273         return ret;
2274 }
2275
2276 int bch2_btree_node_update_key_get_iter(struct btree_trans *trans,
2277                                         struct btree *b, struct bkey_i *new_key,
2278                                         unsigned commit_flags, bool skip_triggers)
2279 {
2280         struct btree_iter iter;
2281         int ret;
2282
2283         bch2_trans_node_iter_init(trans, &iter, b->c.btree_id, b->key.k.p,
2284                                   BTREE_MAX_DEPTH, b->c.level,
2285                                   BTREE_ITER_INTENT);
2286         ret = bch2_btree_iter_traverse(&iter);
2287         if (ret)
2288                 goto out;
2289
2290         /* has node been freed? */
2291         if (iter.path->l[b->c.level].b != b) {
2292                 /* node has been freed: */
2293                 BUG_ON(!btree_node_dying(b));
2294                 goto out;
2295         }
2296
2297         BUG_ON(!btree_node_hashed(b));
2298
2299         ret = bch2_btree_node_update_key(trans, &iter, b, new_key,
2300                                          commit_flags, skip_triggers);
2301 out:
2302         bch2_trans_iter_exit(trans, &iter);
2303         return ret;
2304 }
2305
2306 /* Init code: */
2307
2308 /*
2309  * Only for filesystem bringup, when first reading the btree roots or allocating
2310  * btree roots when initializing a new filesystem:
2311  */
2312 void bch2_btree_set_root_for_read(struct bch_fs *c, struct btree *b)
2313 {
2314         BUG_ON(btree_node_root(c, b));
2315
2316         bch2_btree_set_root_inmem(c, b);
2317 }
2318
2319 static int __bch2_btree_root_alloc(struct btree_trans *trans, enum btree_id id)
2320 {
2321         struct bch_fs *c = trans->c;
2322         struct closure cl;
2323         struct btree *b;
2324         int ret;
2325
2326         closure_init_stack(&cl);
2327
2328         do {
2329                 ret = bch2_btree_cache_cannibalize_lock(c, &cl);
2330                 closure_sync(&cl);
2331         } while (ret);
2332
2333         b = bch2_btree_node_mem_alloc(trans, false);
2334         bch2_btree_cache_cannibalize_unlock(c);
2335
2336         set_btree_node_fake(b);
2337         set_btree_node_need_rewrite(b);
2338         b->c.level      = 0;
2339         b->c.btree_id   = id;
2340
2341         bkey_btree_ptr_init(&b->key);
2342         b->key.k.p = SPOS_MAX;
2343         *((u64 *) bkey_i_to_btree_ptr(&b->key)->v.start) = U64_MAX - id;
2344
2345         bch2_bset_init_first(b, &b->data->keys);
2346         bch2_btree_build_aux_trees(b);
2347
2348         b->data->flags = 0;
2349         btree_set_min(b, POS_MIN);
2350         btree_set_max(b, SPOS_MAX);
2351         b->data->format = bch2_btree_calc_format(b);
2352         btree_node_set_format(b, b->data->format);
2353
2354         ret = bch2_btree_node_hash_insert(&c->btree_cache, b,
2355                                           b->c.level, b->c.btree_id);
2356         BUG_ON(ret);
2357
2358         bch2_btree_set_root_inmem(c, b);
2359
2360         six_unlock_write(&b->c.lock);
2361         six_unlock_intent(&b->c.lock);
2362         return 0;
2363 }
2364
2365 void bch2_btree_root_alloc(struct bch_fs *c, enum btree_id id)
2366 {
2367         bch2_trans_run(c, __bch2_btree_root_alloc(trans, id));
2368 }
2369
2370 void bch2_btree_updates_to_text(struct printbuf *out, struct bch_fs *c)
2371 {
2372         struct btree_update *as;
2373
2374         mutex_lock(&c->btree_interior_update_lock);
2375         list_for_each_entry(as, &c->btree_interior_update_list, list)
2376                 prt_printf(out, "%p m %u w %u r %u j %llu\n",
2377                        as,
2378                        as->mode,
2379                        as->nodes_written,
2380                        closure_nr_remaining(&as->cl),
2381                        as->journal.seq);
2382         mutex_unlock(&c->btree_interior_update_lock);
2383 }
2384
2385 static bool bch2_btree_interior_updates_pending(struct bch_fs *c)
2386 {
2387         bool ret;
2388
2389         mutex_lock(&c->btree_interior_update_lock);
2390         ret = !list_empty(&c->btree_interior_update_list);
2391         mutex_unlock(&c->btree_interior_update_lock);
2392
2393         return ret;
2394 }
2395
2396 bool bch2_btree_interior_updates_flush(struct bch_fs *c)
2397 {
2398         bool ret = bch2_btree_interior_updates_pending(c);
2399
2400         if (ret)
2401                 closure_wait_event(&c->btree_interior_update_wait,
2402                                    !bch2_btree_interior_updates_pending(c));
2403         return ret;
2404 }
2405
2406 void bch2_journal_entry_to_btree_root(struct bch_fs *c, struct jset_entry *entry)
2407 {
2408         struct btree_root *r = bch2_btree_id_root(c, entry->btree_id);
2409
2410         mutex_lock(&c->btree_root_lock);
2411
2412         r->level = entry->level;
2413         r->alive = true;
2414         bkey_copy(&r->key, &entry->start[0]);
2415
2416         mutex_unlock(&c->btree_root_lock);
2417 }
2418
2419 struct jset_entry *
2420 bch2_btree_roots_to_journal_entries(struct bch_fs *c,
2421                                     struct jset_entry *start,
2422                                     struct jset_entry *end)
2423 {
2424         struct jset_entry *entry;
2425         unsigned long have = 0;
2426         unsigned i;
2427
2428         for (entry = start; entry < end; entry = vstruct_next(entry))
2429                 if (entry->type == BCH_JSET_ENTRY_btree_root)
2430                         __set_bit(entry->btree_id, &have);
2431
2432         mutex_lock(&c->btree_root_lock);
2433
2434         for (i = 0; i < btree_id_nr_alive(c); i++) {
2435                 struct btree_root *r = bch2_btree_id_root(c, i);
2436
2437                 if (r->alive && !test_bit(i, &have)) {
2438                         journal_entry_set(end, BCH_JSET_ENTRY_btree_root,
2439                                           i, r->level, &r->key, r->key.k.u64s);
2440                         end = vstruct_next(end);
2441                 }
2442         }
2443
2444         mutex_unlock(&c->btree_root_lock);
2445
2446         return end;
2447 }
2448
2449 void bch2_fs_btree_interior_update_exit(struct bch_fs *c)
2450 {
2451         if (c->btree_interior_update_worker)
2452                 destroy_workqueue(c->btree_interior_update_worker);
2453         mempool_exit(&c->btree_interior_update_pool);
2454 }
2455
2456 void bch2_fs_btree_interior_update_init_early(struct bch_fs *c)
2457 {
2458         mutex_init(&c->btree_reserve_cache_lock);
2459         INIT_LIST_HEAD(&c->btree_interior_update_list);
2460         INIT_LIST_HEAD(&c->btree_interior_updates_unwritten);
2461         mutex_init(&c->btree_interior_update_lock);
2462         INIT_WORK(&c->btree_interior_update_work, btree_interior_update_work);
2463
2464         INIT_LIST_HEAD(&c->pending_node_rewrites);
2465         mutex_init(&c->pending_node_rewrites_lock);
2466 }
2467
2468 int bch2_fs_btree_interior_update_init(struct bch_fs *c)
2469 {
2470         c->btree_interior_update_worker =
2471                 alloc_workqueue("btree_update", WQ_UNBOUND|WQ_MEM_RECLAIM, 1);
2472         if (!c->btree_interior_update_worker)
2473                 return -BCH_ERR_ENOMEM_btree_interior_update_worker_init;
2474
2475         if (mempool_init_kmalloc_pool(&c->btree_interior_update_pool, 1,
2476                                       sizeof(struct btree_update)))
2477                 return -BCH_ERR_ENOMEM_btree_interior_update_pool_init;
2478
2479         return 0;
2480 }