1 // SPDX-License-Identifier: GPL-2.0 2 /* 3 * Copyright 2012 Google, Inc. 4 * 5 * Foreground allocator code: allocate buckets from freelist, and allocate in 6 * sector granularity from writepoints. 7 * 8 * bch2_bucket_alloc() allocates a single bucket from a specific device. 9 * 10 * bch2_bucket_alloc_set() allocates one or more buckets from different devices 11 * in a given filesystem. 12 */ 13 14 #include "bcachefs.h" 15 #include "alloc_background.h" 16 #include "alloc_foreground.h" 17 #include "backpointers.h" 18 #include "btree_iter.h" 19 #include "btree_update.h" 20 #include "btree_gc.h" 21 #include "buckets.h" 22 #include "buckets_waiting_for_journal.h" 23 #include "clock.h" 24 #include "debug.h" 25 #include "disk_groups.h" 26 #include "ec.h" 27 #include "error.h" 28 #include "io_write.h" 29 #include "journal.h" 30 #include "movinggc.h" 31 #include "nocow_locking.h" 32 #include "trace.h" 33 34 #include <linux/math64.h> 35 #include <linux/rculist.h> 36 #include <linux/rcupdate.h> 37 38 static void bch2_trans_mutex_lock_norelock(struct btree_trans *trans, 39 struct mutex *lock) 40 { 41 if (!mutex_trylock(lock)) { 42 bch2_trans_unlock(trans); 43 mutex_lock(lock); 44 } 45 } 46 47 const char * const bch2_watermarks[] = { 48 #define x(t) #t, 49 BCH_WATERMARKS() 50 #undef x 51 NULL 52 }; 53 54 /* 55 * Open buckets represent a bucket that's currently being allocated from. They 56 * serve two purposes: 57 * 58 * - They track buckets that have been partially allocated, allowing for 59 * sub-bucket sized allocations - they're used by the sector allocator below 60 * 61 * - They provide a reference to the buckets they own that mark and sweep GC 62 * can find, until the new allocation has a pointer to it inserted into the 63 * btree 64 * 65 * When allocating some space with the sector allocator, the allocation comes 66 * with a reference to an open bucket - the caller is required to put that 67 * reference _after_ doing the index update that makes its allocation reachable. 68 */ 69 70 void bch2_reset_alloc_cursors(struct bch_fs *c) 71 { 72 rcu_read_lock(); 73 for_each_member_device_rcu(c, ca, NULL) 74 ca->alloc_cursor = 0; 75 rcu_read_unlock(); 76 } 77 78 static void bch2_open_bucket_hash_add(struct bch_fs *c, struct open_bucket *ob) 79 { 80 open_bucket_idx_t idx = ob - c->open_buckets; 81 open_bucket_idx_t *slot = open_bucket_hashslot(c, ob->dev, ob->bucket); 82 83 ob->hash = *slot; 84 *slot = idx; 85 } 86 87 static void bch2_open_bucket_hash_remove(struct bch_fs *c, struct open_bucket *ob) 88 { 89 open_bucket_idx_t idx = ob - c->open_buckets; 90 open_bucket_idx_t *slot = open_bucket_hashslot(c, ob->dev, ob->bucket); 91 92 while (*slot != idx) { 93 BUG_ON(!*slot); 94 slot = &c->open_buckets[*slot].hash; 95 } 96 97 *slot = ob->hash; 98 ob->hash = 0; 99 } 100 101 void __bch2_open_bucket_put(struct bch_fs *c, struct open_bucket *ob) 102 { 103 struct bch_dev *ca = bch_dev_bkey_exists(c, ob->dev); 104 105 if (ob->ec) { 106 ec_stripe_new_put(c, ob->ec, STRIPE_REF_io); 107 return; 108 } 109 110 percpu_down_read(&c->mark_lock); 111 spin_lock(&ob->lock); 112 113 ob->valid = false; 114 ob->data_type = 0; 115 116 spin_unlock(&ob->lock); 117 percpu_up_read(&c->mark_lock); 118 119 spin_lock(&c->freelist_lock); 120 bch2_open_bucket_hash_remove(c, ob); 121 122 ob->freelist = c->open_buckets_freelist; 123 c->open_buckets_freelist = ob - c->open_buckets; 124 125 c->open_buckets_nr_free++; 126 ca->nr_open_buckets--; 127 spin_unlock(&c->freelist_lock); 128 129 closure_wake_up(&c->open_buckets_wait); 130 } 131 132 void bch2_open_bucket_write_error(struct bch_fs *c, 133 struct open_buckets *obs, 134 unsigned dev) 135 { 136 struct open_bucket *ob; 137 unsigned i; 138 139 open_bucket_for_each(c, obs, ob, i) 140 if (ob->dev == dev && ob->ec) 141 bch2_ec_bucket_cancel(c, ob); 142 } 143 144 static struct open_bucket *bch2_open_bucket_alloc(struct bch_fs *c) 145 { 146 struct open_bucket *ob; 147 148 BUG_ON(!c->open_buckets_freelist || !c->open_buckets_nr_free); 149 150 ob = c->open_buckets + c->open_buckets_freelist; 151 c->open_buckets_freelist = ob->freelist; 152 atomic_set(&ob->pin, 1); 153 ob->data_type = 0; 154 155 c->open_buckets_nr_free--; 156 return ob; 157 } 158 159 static void open_bucket_free_unused(struct bch_fs *c, struct open_bucket *ob) 160 { 161 BUG_ON(c->open_buckets_partial_nr >= 162 ARRAY_SIZE(c->open_buckets_partial)); 163 164 spin_lock(&c->freelist_lock); 165 ob->on_partial_list = true; 166 c->open_buckets_partial[c->open_buckets_partial_nr++] = 167 ob - c->open_buckets; 168 spin_unlock(&c->freelist_lock); 169 170 closure_wake_up(&c->open_buckets_wait); 171 closure_wake_up(&c->freelist_wait); 172 } 173 174 /* _only_ for allocating the journal on a new device: */ 175 long bch2_bucket_alloc_new_fs(struct bch_dev *ca) 176 { 177 while (ca->new_fs_bucket_idx < ca->mi.nbuckets) { 178 u64 b = ca->new_fs_bucket_idx++; 179 180 if (!is_superblock_bucket(ca, b) && 181 (!ca->buckets_nouse || !test_bit(b, ca->buckets_nouse))) 182 return b; 183 } 184 185 return -1; 186 } 187 188 static inline unsigned open_buckets_reserved(enum bch_watermark watermark) 189 { 190 switch (watermark) { 191 case BCH_WATERMARK_reclaim: 192 return 0; 193 case BCH_WATERMARK_btree: 194 case BCH_WATERMARK_btree_copygc: 195 return OPEN_BUCKETS_COUNT / 4; 196 case BCH_WATERMARK_copygc: 197 return OPEN_BUCKETS_COUNT / 3; 198 default: 199 return OPEN_BUCKETS_COUNT / 2; 200 } 201 } 202 203 static struct open_bucket *__try_alloc_bucket(struct bch_fs *c, struct bch_dev *ca, 204 u64 bucket, 205 enum bch_watermark watermark, 206 const struct bch_alloc_v4 *a, 207 struct bucket_alloc_state *s, 208 struct closure *cl) 209 { 210 struct open_bucket *ob; 211 212 if (unlikely(ca->buckets_nouse && test_bit(bucket, ca->buckets_nouse))) { 213 s->skipped_nouse++; 214 return NULL; 215 } 216 217 if (bch2_bucket_is_open(c, ca->dev_idx, bucket)) { 218 s->skipped_open++; 219 return NULL; 220 } 221 222 if (bch2_bucket_needs_journal_commit(&c->buckets_waiting_for_journal, 223 c->journal.flushed_seq_ondisk, ca->dev_idx, bucket)) { 224 s->skipped_need_journal_commit++; 225 return NULL; 226 } 227 228 if (bch2_bucket_nocow_is_locked(&c->nocow_locks, POS(ca->dev_idx, bucket))) { 229 s->skipped_nocow++; 230 return NULL; 231 } 232 233 spin_lock(&c->freelist_lock); 234 235 if (unlikely(c->open_buckets_nr_free <= open_buckets_reserved(watermark))) { 236 if (cl) 237 closure_wait(&c->open_buckets_wait, cl); 238 239 track_event_change(&c->times[BCH_TIME_blocked_allocate_open_bucket], true); 240 spin_unlock(&c->freelist_lock); 241 return ERR_PTR(-BCH_ERR_open_buckets_empty); 242 } 243 244 /* Recheck under lock: */ 245 if (bch2_bucket_is_open(c, ca->dev_idx, bucket)) { 246 spin_unlock(&c->freelist_lock); 247 s->skipped_open++; 248 return NULL; 249 } 250 251 ob = bch2_open_bucket_alloc(c); 252 253 spin_lock(&ob->lock); 254 255 ob->valid = true; 256 ob->sectors_free = ca->mi.bucket_size; 257 ob->dev = ca->dev_idx; 258 ob->gen = a->gen; 259 ob->bucket = bucket; 260 spin_unlock(&ob->lock); 261 262 ca->nr_open_buckets++; 263 bch2_open_bucket_hash_add(c, ob); 264 265 track_event_change(&c->times[BCH_TIME_blocked_allocate_open_bucket], false); 266 track_event_change(&c->times[BCH_TIME_blocked_allocate], false); 267 268 spin_unlock(&c->freelist_lock); 269 return ob; 270 } 271 272 static struct open_bucket *try_alloc_bucket(struct btree_trans *trans, struct bch_dev *ca, 273 enum bch_watermark watermark, u64 free_entry, 274 struct bucket_alloc_state *s, 275 struct bkey_s_c freespace_k, 276 struct closure *cl) 277 { 278 struct bch_fs *c = trans->c; 279 struct btree_iter iter = { NULL }; 280 struct bkey_s_c k; 281 struct open_bucket *ob; 282 struct bch_alloc_v4 a_convert; 283 const struct bch_alloc_v4 *a; 284 u64 b = free_entry & ~(~0ULL << 56); 285 unsigned genbits = free_entry >> 56; 286 struct printbuf buf = PRINTBUF; 287 int ret; 288 289 if (b < ca->mi.first_bucket || b >= ca->mi.nbuckets) { 290 prt_printf(&buf, "freespace btree has bucket outside allowed range %u-%llu\n" 291 " freespace key ", 292 ca->mi.first_bucket, ca->mi.nbuckets); 293 bch2_bkey_val_to_text(&buf, c, freespace_k); 294 bch2_trans_inconsistent(trans, "%s", buf.buf); 295 ob = ERR_PTR(-EIO); 296 goto err; 297 } 298 299 k = bch2_bkey_get_iter(trans, &iter, 300 BTREE_ID_alloc, POS(ca->dev_idx, b), 301 BTREE_ITER_CACHED); 302 ret = bkey_err(k); 303 if (ret) { 304 ob = ERR_PTR(ret); 305 goto err; 306 } 307 308 a = bch2_alloc_to_v4(k, &a_convert); 309 310 if (a->data_type != BCH_DATA_free) { 311 if (c->curr_recovery_pass <= BCH_RECOVERY_PASS_check_alloc_info) { 312 ob = NULL; 313 goto err; 314 } 315 316 prt_printf(&buf, "non free bucket in freespace btree\n" 317 " freespace key "); 318 bch2_bkey_val_to_text(&buf, c, freespace_k); 319 prt_printf(&buf, "\n "); 320 bch2_bkey_val_to_text(&buf, c, k); 321 bch2_trans_inconsistent(trans, "%s", buf.buf); 322 ob = ERR_PTR(-EIO); 323 goto err; 324 } 325 326 if (genbits != (alloc_freespace_genbits(*a) >> 56) && 327 c->curr_recovery_pass > BCH_RECOVERY_PASS_check_alloc_info) { 328 prt_printf(&buf, "bucket in freespace btree with wrong genbits (got %u should be %llu)\n" 329 " freespace key ", 330 genbits, alloc_freespace_genbits(*a) >> 56); 331 bch2_bkey_val_to_text(&buf, c, freespace_k); 332 prt_printf(&buf, "\n "); 333 bch2_bkey_val_to_text(&buf, c, k); 334 bch2_trans_inconsistent(trans, "%s", buf.buf); 335 ob = ERR_PTR(-EIO); 336 goto err; 337 } 338 339 if (c->curr_recovery_pass <= BCH_RECOVERY_PASS_check_extents_to_backpointers) { 340 struct bch_backpointer bp; 341 struct bpos bp_pos = POS_MIN; 342 343 ret = bch2_get_next_backpointer(trans, POS(ca->dev_idx, b), -1, 344 &bp_pos, &bp, 345 BTREE_ITER_NOPRESERVE); 346 if (ret) { 347 ob = ERR_PTR(ret); 348 goto err; 349 } 350 351 if (!bkey_eq(bp_pos, POS_MAX)) { 352 /* 353 * Bucket may have data in it - we don't call 354 * bc2h_trans_inconnsistent() because fsck hasn't 355 * finished yet 356 */ 357 ob = NULL; 358 goto err; 359 } 360 } 361 362 ob = __try_alloc_bucket(c, ca, b, watermark, a, s, cl); 363 if (!ob) 364 set_btree_iter_dontneed(&iter); 365 err: 366 if (iter.path) 367 set_btree_iter_dontneed(&iter); 368 bch2_trans_iter_exit(trans, &iter); 369 printbuf_exit(&buf); 370 return ob; 371 } 372 373 /* 374 * This path is for before the freespace btree is initialized: 375 * 376 * If ca->new_fs_bucket_idx is nonzero, we haven't yet marked superblock & 377 * journal buckets - journal buckets will be < ca->new_fs_bucket_idx 378 */ 379 static noinline struct open_bucket * 380 bch2_bucket_alloc_early(struct btree_trans *trans, 381 struct bch_dev *ca, 382 enum bch_watermark watermark, 383 struct bucket_alloc_state *s, 384 struct closure *cl) 385 { 386 struct btree_iter iter, citer; 387 struct bkey_s_c k, ck; 388 struct open_bucket *ob = NULL; 389 u64 first_bucket = max_t(u64, ca->mi.first_bucket, ca->new_fs_bucket_idx); 390 u64 alloc_start = max(first_bucket, READ_ONCE(ca->alloc_cursor)); 391 u64 alloc_cursor = alloc_start; 392 int ret; 393 394 /* 395 * Scan with an uncached iterator to avoid polluting the key cache. An 396 * uncached iter will return a cached key if one exists, but if not 397 * there is no other underlying protection for the associated key cache 398 * slot. To avoid racing bucket allocations, look up the cached key slot 399 * of any likely allocation candidate before attempting to proceed with 400 * the allocation. This provides proper exclusion on the associated 401 * bucket. 402 */ 403 again: 404 for_each_btree_key_norestart(trans, iter, BTREE_ID_alloc, POS(ca->dev_idx, alloc_cursor), 405 BTREE_ITER_SLOTS, k, ret) { 406 struct bch_alloc_v4 a_convert; 407 const struct bch_alloc_v4 *a; 408 409 if (bkey_ge(k.k->p, POS(ca->dev_idx, ca->mi.nbuckets))) 410 break; 411 412 if (ca->new_fs_bucket_idx && 413 is_superblock_bucket(ca, k.k->p.offset)) 414 continue; 415 416 a = bch2_alloc_to_v4(k, &a_convert); 417 if (a->data_type != BCH_DATA_free) 418 continue; 419 420 /* now check the cached key to serialize concurrent allocs of the bucket */ 421 ck = bch2_bkey_get_iter(trans, &citer, BTREE_ID_alloc, k.k->p, BTREE_ITER_CACHED); 422 ret = bkey_err(ck); 423 if (ret) 424 break; 425 426 a = bch2_alloc_to_v4(ck, &a_convert); 427 if (a->data_type != BCH_DATA_free) 428 goto next; 429 430 s->buckets_seen++; 431 432 ob = __try_alloc_bucket(trans->c, ca, k.k->p.offset, watermark, a, s, cl); 433 next: 434 set_btree_iter_dontneed(&citer); 435 bch2_trans_iter_exit(trans, &citer); 436 if (ob) 437 break; 438 } 439 bch2_trans_iter_exit(trans, &iter); 440 441 alloc_cursor = iter.pos.offset; 442 ca->alloc_cursor = alloc_cursor; 443 444 if (!ob && ret) 445 ob = ERR_PTR(ret); 446 447 if (!ob && alloc_start > first_bucket) { 448 alloc_cursor = alloc_start = first_bucket; 449 goto again; 450 } 451 452 return ob; 453 } 454 455 static struct open_bucket *bch2_bucket_alloc_freelist(struct btree_trans *trans, 456 struct bch_dev *ca, 457 enum bch_watermark watermark, 458 struct bucket_alloc_state *s, 459 struct closure *cl) 460 { 461 struct btree_iter iter; 462 struct bkey_s_c k; 463 struct open_bucket *ob = NULL; 464 u64 alloc_start = max_t(u64, ca->mi.first_bucket, READ_ONCE(ca->alloc_cursor)); 465 u64 alloc_cursor = alloc_start; 466 int ret; 467 468 BUG_ON(ca->new_fs_bucket_idx); 469 again: 470 for_each_btree_key_norestart(trans, iter, BTREE_ID_freespace, 471 POS(ca->dev_idx, alloc_cursor), 0, k, ret) { 472 if (k.k->p.inode != ca->dev_idx) 473 break; 474 475 for (alloc_cursor = max(alloc_cursor, bkey_start_offset(k.k)); 476 alloc_cursor < k.k->p.offset; 477 alloc_cursor++) { 478 ret = btree_trans_too_many_iters(trans); 479 if (ret) { 480 ob = ERR_PTR(ret); 481 break; 482 } 483 484 s->buckets_seen++; 485 486 ob = try_alloc_bucket(trans, ca, watermark, 487 alloc_cursor, s, k, cl); 488 if (ob) { 489 set_btree_iter_dontneed(&iter); 490 break; 491 } 492 } 493 494 if (ob || ret) 495 break; 496 } 497 bch2_trans_iter_exit(trans, &iter); 498 499 ca->alloc_cursor = alloc_cursor; 500 501 if (!ob && ret) 502 ob = ERR_PTR(ret); 503 504 if (!ob && alloc_start > ca->mi.first_bucket) { 505 alloc_cursor = alloc_start = ca->mi.first_bucket; 506 goto again; 507 } 508 509 return ob; 510 } 511 512 /** 513 * bch2_bucket_alloc_trans - allocate a single bucket from a specific device 514 * @trans: transaction object 515 * @ca: device to allocate from 516 * @watermark: how important is this allocation? 517 * @cl: if not NULL, closure to be used to wait if buckets not available 518 * @usage: for secondarily also returning the current device usage 519 * 520 * Returns: an open_bucket on success, or an ERR_PTR() on failure. 521 */ 522 static struct open_bucket *bch2_bucket_alloc_trans(struct btree_trans *trans, 523 struct bch_dev *ca, 524 enum bch_watermark watermark, 525 struct closure *cl, 526 struct bch_dev_usage *usage) 527 { 528 struct bch_fs *c = trans->c; 529 struct open_bucket *ob = NULL; 530 bool freespace = READ_ONCE(ca->mi.freespace_initialized); 531 u64 avail; 532 struct bucket_alloc_state s = { 0 }; 533 bool waiting = false; 534 again: 535 bch2_dev_usage_read_fast(ca, usage); 536 avail = dev_buckets_free(ca, *usage, watermark); 537 538 if (usage->d[BCH_DATA_need_discard].buckets > avail) 539 bch2_do_discards(c); 540 541 if (usage->d[BCH_DATA_need_gc_gens].buckets > avail) 542 bch2_do_gc_gens(c); 543 544 if (should_invalidate_buckets(ca, *usage)) 545 bch2_do_invalidates(c); 546 547 if (!avail) { 548 if (cl && !waiting) { 549 closure_wait(&c->freelist_wait, cl); 550 waiting = true; 551 goto again; 552 } 553 554 track_event_change(&c->times[BCH_TIME_blocked_allocate], true); 555 556 ob = ERR_PTR(-BCH_ERR_freelist_empty); 557 goto err; 558 } 559 560 if (waiting) 561 closure_wake_up(&c->freelist_wait); 562 alloc: 563 ob = likely(freespace) 564 ? bch2_bucket_alloc_freelist(trans, ca, watermark, &s, cl) 565 : bch2_bucket_alloc_early(trans, ca, watermark, &s, cl); 566 567 if (s.skipped_need_journal_commit * 2 > avail) 568 bch2_journal_flush_async(&c->journal, NULL); 569 570 if (!ob && freespace && c->curr_recovery_pass <= BCH_RECOVERY_PASS_check_alloc_info) { 571 freespace = false; 572 goto alloc; 573 } 574 err: 575 if (!ob) 576 ob = ERR_PTR(-BCH_ERR_no_buckets_found); 577 578 if (!IS_ERR(ob)) 579 trace_and_count(c, bucket_alloc, ca, 580 bch2_watermarks[watermark], 581 ob->bucket, 582 usage->d[BCH_DATA_free].buckets, 583 avail, 584 bch2_copygc_wait_amount(c), 585 c->copygc_wait - atomic64_read(&c->io_clock[WRITE].now), 586 &s, 587 cl == NULL, 588 ""); 589 else if (!bch2_err_matches(PTR_ERR(ob), BCH_ERR_transaction_restart)) 590 trace_and_count(c, bucket_alloc_fail, ca, 591 bch2_watermarks[watermark], 592 0, 593 usage->d[BCH_DATA_free].buckets, 594 avail, 595 bch2_copygc_wait_amount(c), 596 c->copygc_wait - atomic64_read(&c->io_clock[WRITE].now), 597 &s, 598 cl == NULL, 599 bch2_err_str(PTR_ERR(ob))); 600 601 return ob; 602 } 603 604 struct open_bucket *bch2_bucket_alloc(struct bch_fs *c, struct bch_dev *ca, 605 enum bch_watermark watermark, 606 struct closure *cl) 607 { 608 struct bch_dev_usage usage; 609 struct open_bucket *ob; 610 611 bch2_trans_do(c, NULL, NULL, 0, 612 PTR_ERR_OR_ZERO(ob = bch2_bucket_alloc_trans(trans, ca, watermark, 613 cl, &usage))); 614 return ob; 615 } 616 617 static int __dev_stripe_cmp(struct dev_stripe_state *stripe, 618 unsigned l, unsigned r) 619 { 620 return ((stripe->next_alloc[l] > stripe->next_alloc[r]) - 621 (stripe->next_alloc[l] < stripe->next_alloc[r])); 622 } 623 624 #define dev_stripe_cmp(l, r) __dev_stripe_cmp(stripe, l, r) 625 626 struct dev_alloc_list bch2_dev_alloc_list(struct bch_fs *c, 627 struct dev_stripe_state *stripe, 628 struct bch_devs_mask *devs) 629 { 630 struct dev_alloc_list ret = { .nr = 0 }; 631 unsigned i; 632 633 for_each_set_bit(i, devs->d, BCH_SB_MEMBERS_MAX) 634 ret.devs[ret.nr++] = i; 635 636 bubble_sort(ret.devs, ret.nr, dev_stripe_cmp); 637 return ret; 638 } 639 640 static inline void bch2_dev_stripe_increment_inlined(struct bch_dev *ca, 641 struct dev_stripe_state *stripe, 642 struct bch_dev_usage *usage) 643 { 644 u64 *v = stripe->next_alloc + ca->dev_idx; 645 u64 free_space = dev_buckets_available(ca, BCH_WATERMARK_normal); 646 u64 free_space_inv = free_space 647 ? div64_u64(1ULL << 48, free_space) 648 : 1ULL << 48; 649 u64 scale = *v / 4; 650 651 if (*v + free_space_inv >= *v) 652 *v += free_space_inv; 653 else 654 *v = U64_MAX; 655 656 for (v = stripe->next_alloc; 657 v < stripe->next_alloc + ARRAY_SIZE(stripe->next_alloc); v++) 658 *v = *v < scale ? 0 : *v - scale; 659 } 660 661 void bch2_dev_stripe_increment(struct bch_dev *ca, 662 struct dev_stripe_state *stripe) 663 { 664 struct bch_dev_usage usage; 665 666 bch2_dev_usage_read_fast(ca, &usage); 667 bch2_dev_stripe_increment_inlined(ca, stripe, &usage); 668 } 669 670 static int add_new_bucket(struct bch_fs *c, 671 struct open_buckets *ptrs, 672 struct bch_devs_mask *devs_may_alloc, 673 unsigned nr_replicas, 674 unsigned *nr_effective, 675 bool *have_cache, 676 unsigned flags, 677 struct open_bucket *ob) 678 { 679 unsigned durability = 680 bch_dev_bkey_exists(c, ob->dev)->mi.durability; 681 682 BUG_ON(*nr_effective >= nr_replicas); 683 684 __clear_bit(ob->dev, devs_may_alloc->d); 685 *nr_effective += durability; 686 *have_cache |= !durability; 687 688 ob_push(c, ptrs, ob); 689 690 if (*nr_effective >= nr_replicas) 691 return 1; 692 if (ob->ec) 693 return 1; 694 return 0; 695 } 696 697 int bch2_bucket_alloc_set_trans(struct btree_trans *trans, 698 struct open_buckets *ptrs, 699 struct dev_stripe_state *stripe, 700 struct bch_devs_mask *devs_may_alloc, 701 unsigned nr_replicas, 702 unsigned *nr_effective, 703 bool *have_cache, 704 unsigned flags, 705 enum bch_data_type data_type, 706 enum bch_watermark watermark, 707 struct closure *cl) 708 { 709 struct bch_fs *c = trans->c; 710 struct dev_alloc_list devs_sorted = 711 bch2_dev_alloc_list(c, stripe, devs_may_alloc); 712 unsigned dev; 713 struct bch_dev *ca; 714 int ret = -BCH_ERR_insufficient_devices; 715 unsigned i; 716 717 BUG_ON(*nr_effective >= nr_replicas); 718 719 for (i = 0; i < devs_sorted.nr; i++) { 720 struct bch_dev_usage usage; 721 struct open_bucket *ob; 722 723 dev = devs_sorted.devs[i]; 724 725 rcu_read_lock(); 726 ca = rcu_dereference(c->devs[dev]); 727 if (ca) 728 percpu_ref_get(&ca->ref); 729 rcu_read_unlock(); 730 731 if (!ca) 732 continue; 733 734 if (!ca->mi.durability && *have_cache) { 735 percpu_ref_put(&ca->ref); 736 continue; 737 } 738 739 ob = bch2_bucket_alloc_trans(trans, ca, watermark, cl, &usage); 740 if (!IS_ERR(ob)) 741 bch2_dev_stripe_increment_inlined(ca, stripe, &usage); 742 percpu_ref_put(&ca->ref); 743 744 if (IS_ERR(ob)) { 745 ret = PTR_ERR(ob); 746 if (bch2_err_matches(ret, BCH_ERR_transaction_restart) || cl) 747 break; 748 continue; 749 } 750 751 ob->data_type = data_type; 752 753 if (add_new_bucket(c, ptrs, devs_may_alloc, 754 nr_replicas, nr_effective, 755 have_cache, flags, ob)) { 756 ret = 0; 757 break; 758 } 759 } 760 761 return ret; 762 } 763 764 /* Allocate from stripes: */ 765 766 /* 767 * if we can't allocate a new stripe because there are already too many 768 * partially filled stripes, force allocating from an existing stripe even when 769 * it's to a device we don't want: 770 */ 771 772 static int bucket_alloc_from_stripe(struct btree_trans *trans, 773 struct open_buckets *ptrs, 774 struct write_point *wp, 775 struct bch_devs_mask *devs_may_alloc, 776 u16 target, 777 unsigned nr_replicas, 778 unsigned *nr_effective, 779 bool *have_cache, 780 enum bch_watermark watermark, 781 unsigned flags, 782 struct closure *cl) 783 { 784 struct bch_fs *c = trans->c; 785 struct dev_alloc_list devs_sorted; 786 struct ec_stripe_head *h; 787 struct open_bucket *ob; 788 unsigned i, ec_idx; 789 int ret = 0; 790 791 if (nr_replicas < 2) 792 return 0; 793 794 if (ec_open_bucket(c, ptrs)) 795 return 0; 796 797 h = bch2_ec_stripe_head_get(trans, target, 0, nr_replicas - 1, watermark, cl); 798 if (IS_ERR(h)) 799 return PTR_ERR(h); 800 if (!h) 801 return 0; 802 803 devs_sorted = bch2_dev_alloc_list(c, &wp->stripe, devs_may_alloc); 804 805 for (i = 0; i < devs_sorted.nr; i++) 806 for (ec_idx = 0; ec_idx < h->s->nr_data; ec_idx++) { 807 if (!h->s->blocks[ec_idx]) 808 continue; 809 810 ob = c->open_buckets + h->s->blocks[ec_idx]; 811 if (ob->dev == devs_sorted.devs[i] && 812 !test_and_set_bit(ec_idx, h->s->blocks_allocated)) 813 goto got_bucket; 814 } 815 goto out_put_head; 816 got_bucket: 817 ob->ec_idx = ec_idx; 818 ob->ec = h->s; 819 ec_stripe_new_get(h->s, STRIPE_REF_io); 820 821 ret = add_new_bucket(c, ptrs, devs_may_alloc, 822 nr_replicas, nr_effective, 823 have_cache, flags, ob); 824 out_put_head: 825 bch2_ec_stripe_head_put(c, h); 826 return ret; 827 } 828 829 /* Sector allocator */ 830 831 static bool want_bucket(struct bch_fs *c, 832 struct write_point *wp, 833 struct bch_devs_mask *devs_may_alloc, 834 bool *have_cache, bool ec, 835 struct open_bucket *ob) 836 { 837 struct bch_dev *ca = bch_dev_bkey_exists(c, ob->dev); 838 839 if (!test_bit(ob->dev, devs_may_alloc->d)) 840 return false; 841 842 if (ob->data_type != wp->data_type) 843 return false; 844 845 if (!ca->mi.durability && 846 (wp->data_type == BCH_DATA_btree || ec || *have_cache)) 847 return false; 848 849 if (ec != (ob->ec != NULL)) 850 return false; 851 852 return true; 853 } 854 855 static int bucket_alloc_set_writepoint(struct bch_fs *c, 856 struct open_buckets *ptrs, 857 struct write_point *wp, 858 struct bch_devs_mask *devs_may_alloc, 859 unsigned nr_replicas, 860 unsigned *nr_effective, 861 bool *have_cache, 862 bool ec, unsigned flags) 863 { 864 struct open_buckets ptrs_skip = { .nr = 0 }; 865 struct open_bucket *ob; 866 unsigned i; 867 int ret = 0; 868 869 open_bucket_for_each(c, &wp->ptrs, ob, i) { 870 if (!ret && want_bucket(c, wp, devs_may_alloc, 871 have_cache, ec, ob)) 872 ret = add_new_bucket(c, ptrs, devs_may_alloc, 873 nr_replicas, nr_effective, 874 have_cache, flags, ob); 875 else 876 ob_push(c, &ptrs_skip, ob); 877 } 878 wp->ptrs = ptrs_skip; 879 880 return ret; 881 } 882 883 static int bucket_alloc_set_partial(struct bch_fs *c, 884 struct open_buckets *ptrs, 885 struct write_point *wp, 886 struct bch_devs_mask *devs_may_alloc, 887 unsigned nr_replicas, 888 unsigned *nr_effective, 889 bool *have_cache, bool ec, 890 enum bch_watermark watermark, 891 unsigned flags) 892 { 893 int i, ret = 0; 894 895 if (!c->open_buckets_partial_nr) 896 return 0; 897 898 spin_lock(&c->freelist_lock); 899 900 if (!c->open_buckets_partial_nr) 901 goto unlock; 902 903 for (i = c->open_buckets_partial_nr - 1; i >= 0; --i) { 904 struct open_bucket *ob = c->open_buckets + c->open_buckets_partial[i]; 905 906 if (want_bucket(c, wp, devs_may_alloc, have_cache, ec, ob)) { 907 struct bch_dev *ca = bch_dev_bkey_exists(c, ob->dev); 908 struct bch_dev_usage usage; 909 u64 avail; 910 911 bch2_dev_usage_read_fast(ca, &usage); 912 avail = dev_buckets_free(ca, usage, watermark); 913 if (!avail) 914 continue; 915 916 array_remove_item(c->open_buckets_partial, 917 c->open_buckets_partial_nr, 918 i); 919 ob->on_partial_list = false; 920 921 ret = add_new_bucket(c, ptrs, devs_may_alloc, 922 nr_replicas, nr_effective, 923 have_cache, flags, ob); 924 if (ret) 925 break; 926 } 927 } 928 unlock: 929 spin_unlock(&c->freelist_lock); 930 return ret; 931 } 932 933 static int __open_bucket_add_buckets(struct btree_trans *trans, 934 struct open_buckets *ptrs, 935 struct write_point *wp, 936 struct bch_devs_list *devs_have, 937 u16 target, 938 bool erasure_code, 939 unsigned nr_replicas, 940 unsigned *nr_effective, 941 bool *have_cache, 942 enum bch_watermark watermark, 943 unsigned flags, 944 struct closure *_cl) 945 { 946 struct bch_fs *c = trans->c; 947 struct bch_devs_mask devs; 948 struct open_bucket *ob; 949 struct closure *cl = NULL; 950 unsigned i; 951 int ret; 952 953 devs = target_rw_devs(c, wp->data_type, target); 954 955 /* Don't allocate from devices we already have pointers to: */ 956 darray_for_each(*devs_have, i) 957 __clear_bit(*i, devs.d); 958 959 open_bucket_for_each(c, ptrs, ob, i) 960 __clear_bit(ob->dev, devs.d); 961 962 if (erasure_code && ec_open_bucket(c, ptrs)) 963 return 0; 964 965 ret = bucket_alloc_set_writepoint(c, ptrs, wp, &devs, 966 nr_replicas, nr_effective, 967 have_cache, erasure_code, flags); 968 if (ret) 969 return ret; 970 971 ret = bucket_alloc_set_partial(c, ptrs, wp, &devs, 972 nr_replicas, nr_effective, 973 have_cache, erasure_code, watermark, flags); 974 if (ret) 975 return ret; 976 977 if (erasure_code) { 978 ret = bucket_alloc_from_stripe(trans, ptrs, wp, &devs, 979 target, 980 nr_replicas, nr_effective, 981 have_cache, 982 watermark, flags, _cl); 983 } else { 984 retry_blocking: 985 /* 986 * Try nonblocking first, so that if one device is full we'll try from 987 * other devices: 988 */ 989 ret = bch2_bucket_alloc_set_trans(trans, ptrs, &wp->stripe, &devs, 990 nr_replicas, nr_effective, have_cache, 991 flags, wp->data_type, watermark, cl); 992 if (ret && 993 !bch2_err_matches(ret, BCH_ERR_transaction_restart) && 994 !bch2_err_matches(ret, BCH_ERR_insufficient_devices) && 995 !cl && _cl) { 996 cl = _cl; 997 goto retry_blocking; 998 } 999 } 1000 1001 return ret; 1002 } 1003 1004 static int open_bucket_add_buckets(struct btree_trans *trans, 1005 struct open_buckets *ptrs, 1006 struct write_point *wp, 1007 struct bch_devs_list *devs_have, 1008 u16 target, 1009 unsigned erasure_code, 1010 unsigned nr_replicas, 1011 unsigned *nr_effective, 1012 bool *have_cache, 1013 enum bch_watermark watermark, 1014 unsigned flags, 1015 struct closure *cl) 1016 { 1017 int ret; 1018 1019 if (erasure_code) { 1020 ret = __open_bucket_add_buckets(trans, ptrs, wp, 1021 devs_have, target, erasure_code, 1022 nr_replicas, nr_effective, have_cache, 1023 watermark, flags, cl); 1024 if (bch2_err_matches(ret, BCH_ERR_transaction_restart) || 1025 bch2_err_matches(ret, BCH_ERR_operation_blocked) || 1026 bch2_err_matches(ret, BCH_ERR_freelist_empty) || 1027 bch2_err_matches(ret, BCH_ERR_open_buckets_empty)) 1028 return ret; 1029 if (*nr_effective >= nr_replicas) 1030 return 0; 1031 } 1032 1033 ret = __open_bucket_add_buckets(trans, ptrs, wp, 1034 devs_have, target, false, 1035 nr_replicas, nr_effective, have_cache, 1036 watermark, flags, cl); 1037 return ret < 0 ? ret : 0; 1038 } 1039 1040 /** 1041 * should_drop_bucket - check if this is open_bucket should go away 1042 * @ob: open_bucket to predicate on 1043 * @c: filesystem handle 1044 * @ca: if set, we're killing buckets for a particular device 1045 * @ec: if true, we're shutting down erasure coding and killing all ec 1046 * open_buckets 1047 * otherwise, return true 1048 * Returns: true if we should kill this open_bucket 1049 * 1050 * We're killing open_buckets because we're shutting down a device, erasure 1051 * coding, or the entire filesystem - check if this open_bucket matches: 1052 */ 1053 static bool should_drop_bucket(struct open_bucket *ob, struct bch_fs *c, 1054 struct bch_dev *ca, bool ec) 1055 { 1056 if (ec) { 1057 return ob->ec != NULL; 1058 } else if (ca) { 1059 bool drop = ob->dev == ca->dev_idx; 1060 struct open_bucket *ob2; 1061 unsigned i; 1062 1063 if (!drop && ob->ec) { 1064 unsigned nr_blocks; 1065 1066 mutex_lock(&ob->ec->lock); 1067 nr_blocks = bkey_i_to_stripe(&ob->ec->new_stripe.key)->v.nr_blocks; 1068 1069 for (i = 0; i < nr_blocks; i++) { 1070 if (!ob->ec->blocks[i]) 1071 continue; 1072 1073 ob2 = c->open_buckets + ob->ec->blocks[i]; 1074 drop |= ob2->dev == ca->dev_idx; 1075 } 1076 mutex_unlock(&ob->ec->lock); 1077 } 1078 1079 return drop; 1080 } else { 1081 return true; 1082 } 1083 } 1084 1085 static void bch2_writepoint_stop(struct bch_fs *c, struct bch_dev *ca, 1086 bool ec, struct write_point *wp) 1087 { 1088 struct open_buckets ptrs = { .nr = 0 }; 1089 struct open_bucket *ob; 1090 unsigned i; 1091 1092 mutex_lock(&wp->lock); 1093 open_bucket_for_each(c, &wp->ptrs, ob, i) 1094 if (should_drop_bucket(ob, c, ca, ec)) 1095 bch2_open_bucket_put(c, ob); 1096 else 1097 ob_push(c, &ptrs, ob); 1098 wp->ptrs = ptrs; 1099 mutex_unlock(&wp->lock); 1100 } 1101 1102 void bch2_open_buckets_stop(struct bch_fs *c, struct bch_dev *ca, 1103 bool ec) 1104 { 1105 unsigned i; 1106 1107 /* Next, close write points that point to this device... */ 1108 for (i = 0; i < ARRAY_SIZE(c->write_points); i++) 1109 bch2_writepoint_stop(c, ca, ec, &c->write_points[i]); 1110 1111 bch2_writepoint_stop(c, ca, ec, &c->copygc_write_point); 1112 bch2_writepoint_stop(c, ca, ec, &c->rebalance_write_point); 1113 bch2_writepoint_stop(c, ca, ec, &c->btree_write_point); 1114 1115 mutex_lock(&c->btree_reserve_cache_lock); 1116 while (c->btree_reserve_cache_nr) { 1117 struct btree_alloc *a = 1118 &c->btree_reserve_cache[--c->btree_reserve_cache_nr]; 1119 1120 bch2_open_buckets_put(c, &a->ob); 1121 } 1122 mutex_unlock(&c->btree_reserve_cache_lock); 1123 1124 spin_lock(&c->freelist_lock); 1125 i = 0; 1126 while (i < c->open_buckets_partial_nr) { 1127 struct open_bucket *ob = 1128 c->open_buckets + c->open_buckets_partial[i]; 1129 1130 if (should_drop_bucket(ob, c, ca, ec)) { 1131 --c->open_buckets_partial_nr; 1132 swap(c->open_buckets_partial[i], 1133 c->open_buckets_partial[c->open_buckets_partial_nr]); 1134 ob->on_partial_list = false; 1135 spin_unlock(&c->freelist_lock); 1136 bch2_open_bucket_put(c, ob); 1137 spin_lock(&c->freelist_lock); 1138 } else { 1139 i++; 1140 } 1141 } 1142 spin_unlock(&c->freelist_lock); 1143 1144 bch2_ec_stop_dev(c, ca); 1145 } 1146 1147 static inline struct hlist_head *writepoint_hash(struct bch_fs *c, 1148 unsigned long write_point) 1149 { 1150 unsigned hash = 1151 hash_long(write_point, ilog2(ARRAY_SIZE(c->write_points_hash))); 1152 1153 return &c->write_points_hash[hash]; 1154 } 1155 1156 static struct write_point *__writepoint_find(struct hlist_head *head, 1157 unsigned long write_point) 1158 { 1159 struct write_point *wp; 1160 1161 rcu_read_lock(); 1162 hlist_for_each_entry_rcu(wp, head, node) 1163 if (wp->write_point == write_point) 1164 goto out; 1165 wp = NULL; 1166 out: 1167 rcu_read_unlock(); 1168 return wp; 1169 } 1170 1171 static inline bool too_many_writepoints(struct bch_fs *c, unsigned factor) 1172 { 1173 u64 stranded = c->write_points_nr * c->bucket_size_max; 1174 u64 free = bch2_fs_usage_read_short(c).free; 1175 1176 return stranded * factor > free; 1177 } 1178 1179 static bool try_increase_writepoints(struct bch_fs *c) 1180 { 1181 struct write_point *wp; 1182 1183 if (c->write_points_nr == ARRAY_SIZE(c->write_points) || 1184 too_many_writepoints(c, 32)) 1185 return false; 1186 1187 wp = c->write_points + c->write_points_nr++; 1188 hlist_add_head_rcu(&wp->node, writepoint_hash(c, wp->write_point)); 1189 return true; 1190 } 1191 1192 static bool try_decrease_writepoints(struct btree_trans *trans, unsigned old_nr) 1193 { 1194 struct bch_fs *c = trans->c; 1195 struct write_point *wp; 1196 struct open_bucket *ob; 1197 unsigned i; 1198 1199 mutex_lock(&c->write_points_hash_lock); 1200 if (c->write_points_nr < old_nr) { 1201 mutex_unlock(&c->write_points_hash_lock); 1202 return true; 1203 } 1204 1205 if (c->write_points_nr == 1 || 1206 !too_many_writepoints(c, 8)) { 1207 mutex_unlock(&c->write_points_hash_lock); 1208 return false; 1209 } 1210 1211 wp = c->write_points + --c->write_points_nr; 1212 1213 hlist_del_rcu(&wp->node); 1214 mutex_unlock(&c->write_points_hash_lock); 1215 1216 bch2_trans_mutex_lock_norelock(trans, &wp->lock); 1217 open_bucket_for_each(c, &wp->ptrs, ob, i) 1218 open_bucket_free_unused(c, ob); 1219 wp->ptrs.nr = 0; 1220 mutex_unlock(&wp->lock); 1221 return true; 1222 } 1223 1224 static struct write_point *writepoint_find(struct btree_trans *trans, 1225 unsigned long write_point) 1226 { 1227 struct bch_fs *c = trans->c; 1228 struct write_point *wp, *oldest; 1229 struct hlist_head *head; 1230 1231 if (!(write_point & 1UL)) { 1232 wp = (struct write_point *) write_point; 1233 bch2_trans_mutex_lock_norelock(trans, &wp->lock); 1234 return wp; 1235 } 1236 1237 head = writepoint_hash(c, write_point); 1238 restart_find: 1239 wp = __writepoint_find(head, write_point); 1240 if (wp) { 1241 lock_wp: 1242 bch2_trans_mutex_lock_norelock(trans, &wp->lock); 1243 if (wp->write_point == write_point) 1244 goto out; 1245 mutex_unlock(&wp->lock); 1246 goto restart_find; 1247 } 1248 restart_find_oldest: 1249 oldest = NULL; 1250 for (wp = c->write_points; 1251 wp < c->write_points + c->write_points_nr; wp++) 1252 if (!oldest || time_before64(wp->last_used, oldest->last_used)) 1253 oldest = wp; 1254 1255 bch2_trans_mutex_lock_norelock(trans, &oldest->lock); 1256 bch2_trans_mutex_lock_norelock(trans, &c->write_points_hash_lock); 1257 if (oldest >= c->write_points + c->write_points_nr || 1258 try_increase_writepoints(c)) { 1259 mutex_unlock(&c->write_points_hash_lock); 1260 mutex_unlock(&oldest->lock); 1261 goto restart_find_oldest; 1262 } 1263 1264 wp = __writepoint_find(head, write_point); 1265 if (wp && wp != oldest) { 1266 mutex_unlock(&c->write_points_hash_lock); 1267 mutex_unlock(&oldest->lock); 1268 goto lock_wp; 1269 } 1270 1271 wp = oldest; 1272 hlist_del_rcu(&wp->node); 1273 wp->write_point = write_point; 1274 hlist_add_head_rcu(&wp->node, head); 1275 mutex_unlock(&c->write_points_hash_lock); 1276 out: 1277 wp->last_used = local_clock(); 1278 return wp; 1279 } 1280 1281 static noinline void 1282 deallocate_extra_replicas(struct bch_fs *c, 1283 struct open_buckets *ptrs, 1284 struct open_buckets *ptrs_no_use, 1285 unsigned extra_replicas) 1286 { 1287 struct open_buckets ptrs2 = { 0 }; 1288 struct open_bucket *ob; 1289 unsigned i; 1290 1291 open_bucket_for_each(c, ptrs, ob, i) { 1292 unsigned d = bch_dev_bkey_exists(c, ob->dev)->mi.durability; 1293 1294 if (d && d <= extra_replicas) { 1295 extra_replicas -= d; 1296 ob_push(c, ptrs_no_use, ob); 1297 } else { 1298 ob_push(c, &ptrs2, ob); 1299 } 1300 } 1301 1302 *ptrs = ptrs2; 1303 } 1304 1305 /* 1306 * Get us an open_bucket we can allocate from, return with it locked: 1307 */ 1308 int bch2_alloc_sectors_start_trans(struct btree_trans *trans, 1309 unsigned target, 1310 unsigned erasure_code, 1311 struct write_point_specifier write_point, 1312 struct bch_devs_list *devs_have, 1313 unsigned nr_replicas, 1314 unsigned nr_replicas_required, 1315 enum bch_watermark watermark, 1316 unsigned flags, 1317 struct closure *cl, 1318 struct write_point **wp_ret) 1319 { 1320 struct bch_fs *c = trans->c; 1321 struct write_point *wp; 1322 struct open_bucket *ob; 1323 struct open_buckets ptrs; 1324 unsigned nr_effective, write_points_nr; 1325 bool have_cache; 1326 int ret; 1327 int i; 1328 1329 if (!IS_ENABLED(CONFIG_BCACHEFS_ERASURE_CODING)) 1330 erasure_code = false; 1331 1332 BUG_ON(flags & BCH_WRITE_ONLY_SPECIFIED_DEVS); 1333 1334 BUG_ON(!nr_replicas || !nr_replicas_required); 1335 retry: 1336 ptrs.nr = 0; 1337 nr_effective = 0; 1338 write_points_nr = c->write_points_nr; 1339 have_cache = false; 1340 1341 *wp_ret = wp = writepoint_find(trans, write_point.v); 1342 1343 /* metadata may not allocate on cache devices: */ 1344 if (wp->data_type != BCH_DATA_user) 1345 have_cache = true; 1346 1347 if (target && !(flags & BCH_WRITE_ONLY_SPECIFIED_DEVS)) { 1348 ret = open_bucket_add_buckets(trans, &ptrs, wp, devs_have, 1349 target, erasure_code, 1350 nr_replicas, &nr_effective, 1351 &have_cache, watermark, 1352 flags, NULL); 1353 if (!ret || 1354 bch2_err_matches(ret, BCH_ERR_transaction_restart)) 1355 goto alloc_done; 1356 1357 /* Don't retry from all devices if we're out of open buckets: */ 1358 if (bch2_err_matches(ret, BCH_ERR_open_buckets_empty)) { 1359 int ret2 = open_bucket_add_buckets(trans, &ptrs, wp, devs_have, 1360 target, erasure_code, 1361 nr_replicas, &nr_effective, 1362 &have_cache, watermark, 1363 flags, cl); 1364 if (!ret2 || 1365 bch2_err_matches(ret2, BCH_ERR_transaction_restart) || 1366 bch2_err_matches(ret2, BCH_ERR_open_buckets_empty)) { 1367 ret = ret2; 1368 goto alloc_done; 1369 } 1370 } 1371 1372 /* 1373 * Only try to allocate cache (durability = 0 devices) from the 1374 * specified target: 1375 */ 1376 have_cache = true; 1377 1378 ret = open_bucket_add_buckets(trans, &ptrs, wp, devs_have, 1379 0, erasure_code, 1380 nr_replicas, &nr_effective, 1381 &have_cache, watermark, 1382 flags, cl); 1383 } else { 1384 ret = open_bucket_add_buckets(trans, &ptrs, wp, devs_have, 1385 target, erasure_code, 1386 nr_replicas, &nr_effective, 1387 &have_cache, watermark, 1388 flags, cl); 1389 } 1390 alloc_done: 1391 BUG_ON(!ret && nr_effective < nr_replicas); 1392 1393 if (erasure_code && !ec_open_bucket(c, &ptrs)) 1394 pr_debug("failed to get ec bucket: ret %u", ret); 1395 1396 if (ret == -BCH_ERR_insufficient_devices && 1397 nr_effective >= nr_replicas_required) 1398 ret = 0; 1399 1400 if (ret) 1401 goto err; 1402 1403 if (nr_effective > nr_replicas) 1404 deallocate_extra_replicas(c, &ptrs, &wp->ptrs, nr_effective - nr_replicas); 1405 1406 /* Free buckets we didn't use: */ 1407 open_bucket_for_each(c, &wp->ptrs, ob, i) 1408 open_bucket_free_unused(c, ob); 1409 1410 wp->ptrs = ptrs; 1411 1412 wp->sectors_free = UINT_MAX; 1413 1414 open_bucket_for_each(c, &wp->ptrs, ob, i) 1415 wp->sectors_free = min(wp->sectors_free, ob->sectors_free); 1416 1417 BUG_ON(!wp->sectors_free || wp->sectors_free == UINT_MAX); 1418 1419 return 0; 1420 err: 1421 open_bucket_for_each(c, &wp->ptrs, ob, i) 1422 if (ptrs.nr < ARRAY_SIZE(ptrs.v)) 1423 ob_push(c, &ptrs, ob); 1424 else 1425 open_bucket_free_unused(c, ob); 1426 wp->ptrs = ptrs; 1427 1428 mutex_unlock(&wp->lock); 1429 1430 if (bch2_err_matches(ret, BCH_ERR_freelist_empty) && 1431 try_decrease_writepoints(trans, write_points_nr)) 1432 goto retry; 1433 1434 if (bch2_err_matches(ret, BCH_ERR_open_buckets_empty) || 1435 bch2_err_matches(ret, BCH_ERR_freelist_empty)) 1436 return cl 1437 ? -BCH_ERR_bucket_alloc_blocked 1438 : -BCH_ERR_ENOSPC_bucket_alloc; 1439 1440 return ret; 1441 } 1442 1443 struct bch_extent_ptr bch2_ob_ptr(struct bch_fs *c, struct open_bucket *ob) 1444 { 1445 struct bch_dev *ca = bch_dev_bkey_exists(c, ob->dev); 1446 1447 return (struct bch_extent_ptr) { 1448 .type = 1 << BCH_EXTENT_ENTRY_ptr, 1449 .gen = ob->gen, 1450 .dev = ob->dev, 1451 .offset = bucket_to_sector(ca, ob->bucket) + 1452 ca->mi.bucket_size - 1453 ob->sectors_free, 1454 }; 1455 } 1456 1457 void bch2_alloc_sectors_append_ptrs(struct bch_fs *c, struct write_point *wp, 1458 struct bkey_i *k, unsigned sectors, 1459 bool cached) 1460 { 1461 bch2_alloc_sectors_append_ptrs_inlined(c, wp, k, sectors, cached); 1462 } 1463 1464 /* 1465 * Append pointers to the space we just allocated to @k, and mark @sectors space 1466 * as allocated out of @ob 1467 */ 1468 void bch2_alloc_sectors_done(struct bch_fs *c, struct write_point *wp) 1469 { 1470 bch2_alloc_sectors_done_inlined(c, wp); 1471 } 1472 1473 static inline void writepoint_init(struct write_point *wp, 1474 enum bch_data_type type) 1475 { 1476 mutex_init(&wp->lock); 1477 wp->data_type = type; 1478 1479 INIT_WORK(&wp->index_update_work, bch2_write_point_do_index_updates); 1480 INIT_LIST_HEAD(&wp->writes); 1481 spin_lock_init(&wp->writes_lock); 1482 } 1483 1484 void bch2_fs_allocator_foreground_init(struct bch_fs *c) 1485 { 1486 struct open_bucket *ob; 1487 struct write_point *wp; 1488 1489 mutex_init(&c->write_points_hash_lock); 1490 c->write_points_nr = ARRAY_SIZE(c->write_points); 1491 1492 /* open bucket 0 is a sentinal NULL: */ 1493 spin_lock_init(&c->open_buckets[0].lock); 1494 1495 for (ob = c->open_buckets + 1; 1496 ob < c->open_buckets + ARRAY_SIZE(c->open_buckets); ob++) { 1497 spin_lock_init(&ob->lock); 1498 c->open_buckets_nr_free++; 1499 1500 ob->freelist = c->open_buckets_freelist; 1501 c->open_buckets_freelist = ob - c->open_buckets; 1502 } 1503 1504 writepoint_init(&c->btree_write_point, BCH_DATA_btree); 1505 writepoint_init(&c->rebalance_write_point, BCH_DATA_user); 1506 writepoint_init(&c->copygc_write_point, BCH_DATA_user); 1507 1508 for (wp = c->write_points; 1509 wp < c->write_points + c->write_points_nr; wp++) { 1510 writepoint_init(wp, BCH_DATA_user); 1511 1512 wp->last_used = local_clock(); 1513 wp->write_point = (unsigned long) wp; 1514 hlist_add_head_rcu(&wp->node, 1515 writepoint_hash(c, wp->write_point)); 1516 } 1517 } 1518 1519 static void bch2_open_bucket_to_text(struct printbuf *out, struct bch_fs *c, struct open_bucket *ob) 1520 { 1521 struct bch_dev *ca = bch_dev_bkey_exists(c, ob->dev); 1522 unsigned data_type = ob->data_type; 1523 barrier(); /* READ_ONCE() doesn't work on bitfields */ 1524 1525 prt_printf(out, "%zu ref %u ", 1526 ob - c->open_buckets, 1527 atomic_read(&ob->pin)); 1528 bch2_prt_data_type(out, data_type); 1529 prt_printf(out, " %u:%llu gen %u allocated %u/%u", 1530 ob->dev, ob->bucket, ob->gen, 1531 ca->mi.bucket_size - ob->sectors_free, ca->mi.bucket_size); 1532 if (ob->ec) 1533 prt_printf(out, " ec idx %llu", ob->ec->idx); 1534 if (ob->on_partial_list) 1535 prt_str(out, " partial"); 1536 prt_newline(out); 1537 } 1538 1539 void bch2_open_buckets_to_text(struct printbuf *out, struct bch_fs *c) 1540 { 1541 struct open_bucket *ob; 1542 1543 out->atomic++; 1544 1545 for (ob = c->open_buckets; 1546 ob < c->open_buckets + ARRAY_SIZE(c->open_buckets); 1547 ob++) { 1548 spin_lock(&ob->lock); 1549 if (ob->valid && !ob->on_partial_list) 1550 bch2_open_bucket_to_text(out, c, ob); 1551 spin_unlock(&ob->lock); 1552 } 1553 1554 --out->atomic; 1555 } 1556 1557 void bch2_open_buckets_partial_to_text(struct printbuf *out, struct bch_fs *c) 1558 { 1559 unsigned i; 1560 1561 out->atomic++; 1562 spin_lock(&c->freelist_lock); 1563 1564 for (i = 0; i < c->open_buckets_partial_nr; i++) 1565 bch2_open_bucket_to_text(out, c, 1566 c->open_buckets + c->open_buckets_partial[i]); 1567 1568 spin_unlock(&c->freelist_lock); 1569 --out->atomic; 1570 } 1571 1572 static const char * const bch2_write_point_states[] = { 1573 #define x(n) #n, 1574 WRITE_POINT_STATES() 1575 #undef x 1576 NULL 1577 }; 1578 1579 static void bch2_write_point_to_text(struct printbuf *out, struct bch_fs *c, 1580 struct write_point *wp) 1581 { 1582 struct open_bucket *ob; 1583 unsigned i; 1584 1585 prt_printf(out, "%lu: ", wp->write_point); 1586 prt_human_readable_u64(out, wp->sectors_allocated); 1587 1588 prt_printf(out, " last wrote: "); 1589 bch2_pr_time_units(out, sched_clock() - wp->last_used); 1590 1591 for (i = 0; i < WRITE_POINT_STATE_NR; i++) { 1592 prt_printf(out, " %s: ", bch2_write_point_states[i]); 1593 bch2_pr_time_units(out, wp->time[i]); 1594 } 1595 1596 prt_newline(out); 1597 1598 printbuf_indent_add(out, 2); 1599 open_bucket_for_each(c, &wp->ptrs, ob, i) 1600 bch2_open_bucket_to_text(out, c, ob); 1601 printbuf_indent_sub(out, 2); 1602 } 1603 1604 void bch2_write_points_to_text(struct printbuf *out, struct bch_fs *c) 1605 { 1606 struct write_point *wp; 1607 1608 prt_str(out, "Foreground write points\n"); 1609 for (wp = c->write_points; 1610 wp < c->write_points + ARRAY_SIZE(c->write_points); 1611 wp++) 1612 bch2_write_point_to_text(out, c, wp); 1613 1614 prt_str(out, "Copygc write point\n"); 1615 bch2_write_point_to_text(out, c, &c->copygc_write_point); 1616 1617 prt_str(out, "Rebalance write point\n"); 1618 bch2_write_point_to_text(out, c, &c->rebalance_write_point); 1619 1620 prt_str(out, "Btree write point\n"); 1621 bch2_write_point_to_text(out, c, &c->btree_write_point); 1622 } 1623