1 // SPDX-License-Identifier: GPL-2.0 2 3 #include "bcachefs.h" 4 #include "btree_cache.h" 5 #include "btree_io.h" 6 #include "btree_journal_iter.h" 7 #include "btree_node_scan.h" 8 #include "btree_update_interior.h" 9 #include "buckets.h" 10 #include "error.h" 11 #include "journal_io.h" 12 #include "recovery_passes.h" 13 14 #include <linux/kthread.h> 15 #include <linux/sort.h> 16 17 struct find_btree_nodes_worker { 18 struct closure *cl; 19 struct find_btree_nodes *f; 20 struct bch_dev *ca; 21 }; 22 23 static void found_btree_node_to_text(struct printbuf *out, struct bch_fs *c, const struct found_btree_node *n) 24 { 25 prt_printf(out, "%s l=%u seq=%u cookie=%llx ", bch2_btree_id_str(n->btree_id), n->level, n->seq, n->cookie); 26 bch2_bpos_to_text(out, n->min_key); 27 prt_str(out, "-"); 28 bch2_bpos_to_text(out, n->max_key); 29 30 if (n->range_updated) 31 prt_str(out, " range updated"); 32 if (n->overwritten) 33 prt_str(out, " overwritten"); 34 35 for (unsigned i = 0; i < n->nr_ptrs; i++) { 36 prt_char(out, ' '); 37 bch2_extent_ptr_to_text(out, c, n->ptrs + i); 38 } 39 } 40 41 static void found_btree_nodes_to_text(struct printbuf *out, struct bch_fs *c, found_btree_nodes nodes) 42 { 43 printbuf_indent_add(out, 2); 44 darray_for_each(nodes, i) { 45 found_btree_node_to_text(out, c, i); 46 prt_newline(out); 47 } 48 printbuf_indent_sub(out, 2); 49 } 50 51 static void found_btree_node_to_key(struct bkey_i *k, const struct found_btree_node *f) 52 { 53 struct bkey_i_btree_ptr_v2 *bp = bkey_btree_ptr_v2_init(k); 54 55 set_bkey_val_u64s(&bp->k, sizeof(struct bch_btree_ptr_v2) / sizeof(u64) + f->nr_ptrs); 56 bp->k.p = f->max_key; 57 bp->v.seq = cpu_to_le64(f->cookie); 58 bp->v.sectors_written = 0; 59 bp->v.flags = 0; 60 bp->v.min_key = f->min_key; 61 SET_BTREE_PTR_RANGE_UPDATED(&bp->v, f->range_updated); 62 memcpy(bp->v.start, f->ptrs, sizeof(struct bch_extent_ptr) * f->nr_ptrs); 63 } 64 65 static bool found_btree_node_is_readable(struct btree_trans *trans, 66 const struct found_btree_node *f) 67 { 68 struct { __BKEY_PADDED(k, BKEY_BTREE_PTR_VAL_U64s_MAX); } k; 69 70 found_btree_node_to_key(&k.k, f); 71 72 struct btree *b = bch2_btree_node_get_noiter(trans, &k.k, f->btree_id, f->level, false); 73 bool ret = !IS_ERR_OR_NULL(b); 74 if (ret) 75 six_unlock_read(&b->c.lock); 76 77 /* 78 * We might update this node's range; if that happens, we need the node 79 * to be re-read so the read path can trim keys that are no longer in 80 * this node 81 */ 82 if (b != btree_node_root(trans->c, b)) 83 bch2_btree_node_evict(trans, &k.k); 84 return ret; 85 } 86 87 static int found_btree_node_cmp_cookie(const void *_l, const void *_r) 88 { 89 const struct found_btree_node *l = _l; 90 const struct found_btree_node *r = _r; 91 92 return cmp_int(l->btree_id, r->btree_id) ?: 93 cmp_int(l->level, r->level) ?: 94 cmp_int(l->cookie, r->cookie); 95 } 96 97 /* 98 * Given two found btree nodes, if their sequence numbers are equal, take the 99 * one that's readable: 100 */ 101 static int found_btree_node_cmp_time(const struct found_btree_node *l, 102 const struct found_btree_node *r) 103 { 104 return cmp_int(l->seq, r->seq); 105 } 106 107 static int found_btree_node_cmp_pos(const void *_l, const void *_r) 108 { 109 const struct found_btree_node *l = _l; 110 const struct found_btree_node *r = _r; 111 112 return cmp_int(l->btree_id, r->btree_id) ?: 113 -cmp_int(l->level, r->level) ?: 114 bpos_cmp(l->min_key, r->min_key) ?: 115 -found_btree_node_cmp_time(l, r); 116 } 117 118 static void try_read_btree_node(struct find_btree_nodes *f, struct bch_dev *ca, 119 struct bio *bio, struct btree_node *bn, u64 offset) 120 { 121 struct bch_fs *c = container_of(f, struct bch_fs, found_btree_nodes); 122 123 bio_reset(bio, ca->disk_sb.bdev, REQ_OP_READ); 124 bio->bi_iter.bi_sector = offset; 125 bch2_bio_map(bio, bn, PAGE_SIZE); 126 127 submit_bio_wait(bio); 128 if (bch2_dev_io_err_on(bio->bi_status, ca, BCH_MEMBER_ERROR_read, 129 "IO error in try_read_btree_node() at %llu: %s", 130 offset, bch2_blk_status_to_str(bio->bi_status))) 131 return; 132 133 if (le64_to_cpu(bn->magic) != bset_magic(c)) 134 return; 135 136 if (bch2_csum_type_is_encryption(BSET_CSUM_TYPE(&bn->keys))) { 137 struct nonce nonce = btree_nonce(&bn->keys, 0); 138 unsigned bytes = (void *) &bn->keys - (void *) &bn->flags; 139 140 bch2_encrypt(c, BSET_CSUM_TYPE(&bn->keys), nonce, &bn->flags, bytes); 141 } 142 143 if (btree_id_is_alloc(BTREE_NODE_ID(bn))) 144 return; 145 146 if (BTREE_NODE_LEVEL(bn) >= BTREE_MAX_DEPTH) 147 return; 148 149 rcu_read_lock(); 150 struct found_btree_node n = { 151 .btree_id = BTREE_NODE_ID(bn), 152 .level = BTREE_NODE_LEVEL(bn), 153 .seq = BTREE_NODE_SEQ(bn), 154 .cookie = le64_to_cpu(bn->keys.seq), 155 .min_key = bn->min_key, 156 .max_key = bn->max_key, 157 .nr_ptrs = 1, 158 .ptrs[0].type = 1 << BCH_EXTENT_ENTRY_ptr, 159 .ptrs[0].offset = offset, 160 .ptrs[0].dev = ca->dev_idx, 161 .ptrs[0].gen = *bucket_gen(ca, sector_to_bucket(ca, offset)), 162 }; 163 rcu_read_unlock(); 164 165 if (bch2_trans_run(c, found_btree_node_is_readable(trans, &n))) { 166 mutex_lock(&f->lock); 167 if (BSET_BIG_ENDIAN(&bn->keys) != CPU_BIG_ENDIAN) { 168 bch_err(c, "try_read_btree_node() can't handle endian conversion"); 169 f->ret = -EINVAL; 170 goto unlock; 171 } 172 173 if (darray_push(&f->nodes, n)) 174 f->ret = -ENOMEM; 175 unlock: 176 mutex_unlock(&f->lock); 177 } 178 } 179 180 static int read_btree_nodes_worker(void *p) 181 { 182 struct find_btree_nodes_worker *w = p; 183 struct bch_fs *c = container_of(w->f, struct bch_fs, found_btree_nodes); 184 struct bch_dev *ca = w->ca; 185 void *buf = (void *) __get_free_page(GFP_KERNEL); 186 struct bio *bio = bio_alloc(NULL, 1, 0, GFP_KERNEL); 187 unsigned long last_print = jiffies; 188 189 if (!buf || !bio) { 190 bch_err(c, "read_btree_nodes_worker: error allocating bio/buf"); 191 w->f->ret = -ENOMEM; 192 goto err; 193 } 194 195 for (u64 bucket = ca->mi.first_bucket; bucket < ca->mi.nbuckets; bucket++) 196 for (unsigned bucket_offset = 0; 197 bucket_offset + btree_sectors(c) <= ca->mi.bucket_size; 198 bucket_offset += btree_sectors(c)) { 199 if (time_after(jiffies, last_print + HZ * 30)) { 200 u64 cur_sector = bucket * ca->mi.bucket_size + bucket_offset; 201 u64 end_sector = ca->mi.nbuckets * ca->mi.bucket_size; 202 203 bch_info(ca, "%s: %2u%% done", __func__, 204 (unsigned) div64_u64(cur_sector * 100, end_sector)); 205 last_print = jiffies; 206 } 207 208 u64 sector = bucket * ca->mi.bucket_size + bucket_offset; 209 210 if (c->sb.version_upgrade_complete >= bcachefs_metadata_version_mi_btree_bitmap && 211 !bch2_dev_btree_bitmap_marked_sectors(ca, sector, btree_sectors(c))) 212 continue; 213 214 try_read_btree_node(w->f, ca, bio, buf, sector); 215 } 216 err: 217 bio_put(bio); 218 free_page((unsigned long) buf); 219 percpu_ref_get(&ca->io_ref); 220 closure_put(w->cl); 221 kfree(w); 222 return 0; 223 } 224 225 static int read_btree_nodes(struct find_btree_nodes *f) 226 { 227 struct bch_fs *c = container_of(f, struct bch_fs, found_btree_nodes); 228 struct closure cl; 229 int ret = 0; 230 231 closure_init_stack(&cl); 232 233 for_each_online_member(c, ca) { 234 if (!(ca->mi.data_allowed & BIT(BCH_DATA_btree))) 235 continue; 236 237 struct find_btree_nodes_worker *w = kmalloc(sizeof(*w), GFP_KERNEL); 238 struct task_struct *t; 239 240 if (!w) { 241 percpu_ref_put(&ca->io_ref); 242 ret = -ENOMEM; 243 goto err; 244 } 245 246 percpu_ref_get(&ca->io_ref); 247 closure_get(&cl); 248 w->cl = &cl; 249 w->f = f; 250 w->ca = ca; 251 252 t = kthread_run(read_btree_nodes_worker, w, "read_btree_nodes/%s", ca->name); 253 ret = IS_ERR_OR_NULL(t); 254 if (ret) { 255 percpu_ref_put(&ca->io_ref); 256 closure_put(&cl); 257 f->ret = ret; 258 bch_err(c, "error starting kthread: %i", ret); 259 break; 260 } 261 } 262 err: 263 closure_sync(&cl); 264 return f->ret ?: ret; 265 } 266 267 static void bubble_up(struct found_btree_node *n, struct found_btree_node *end) 268 { 269 while (n + 1 < end && 270 found_btree_node_cmp_pos(n, n + 1) > 0) { 271 swap(n[0], n[1]); 272 n++; 273 } 274 } 275 276 static int handle_overwrites(struct bch_fs *c, 277 struct found_btree_node *start, 278 struct found_btree_node *end) 279 { 280 struct found_btree_node *n; 281 again: 282 for (n = start + 1; 283 n < end && 284 n->btree_id == start->btree_id && 285 n->level == start->level && 286 bpos_lt(n->min_key, start->max_key); 287 n++) { 288 int cmp = found_btree_node_cmp_time(start, n); 289 290 if (cmp > 0) { 291 if (bpos_cmp(start->max_key, n->max_key) >= 0) 292 n->overwritten = true; 293 else { 294 n->range_updated = true; 295 n->min_key = bpos_successor(start->max_key); 296 n->range_updated = true; 297 bubble_up(n, end); 298 goto again; 299 } 300 } else if (cmp < 0) { 301 BUG_ON(bpos_cmp(n->min_key, start->min_key) <= 0); 302 303 start->max_key = bpos_predecessor(n->min_key); 304 start->range_updated = true; 305 } else { 306 struct printbuf buf = PRINTBUF; 307 308 prt_str(&buf, "overlapping btree nodes with same seq! halting\n "); 309 found_btree_node_to_text(&buf, c, start); 310 prt_str(&buf, "\n "); 311 found_btree_node_to_text(&buf, c, n); 312 bch_err(c, "%s", buf.buf); 313 printbuf_exit(&buf); 314 return -BCH_ERR_fsck_repair_unimplemented; 315 } 316 } 317 318 return 0; 319 } 320 321 int bch2_scan_for_btree_nodes(struct bch_fs *c) 322 { 323 struct find_btree_nodes *f = &c->found_btree_nodes; 324 struct printbuf buf = PRINTBUF; 325 size_t dst; 326 int ret = 0; 327 328 if (f->nodes.nr) 329 return 0; 330 331 mutex_init(&f->lock); 332 333 ret = read_btree_nodes(f); 334 if (ret) 335 return ret; 336 337 if (!f->nodes.nr) { 338 bch_err(c, "%s: no btree nodes found", __func__); 339 ret = -EINVAL; 340 goto err; 341 } 342 343 if (0 && c->opts.verbose) { 344 printbuf_reset(&buf); 345 prt_printf(&buf, "%s: nodes found:\n", __func__); 346 found_btree_nodes_to_text(&buf, c, f->nodes); 347 bch2_print_string_as_lines(KERN_INFO, buf.buf); 348 } 349 350 sort(f->nodes.data, f->nodes.nr, sizeof(f->nodes.data[0]), found_btree_node_cmp_cookie, NULL); 351 352 dst = 0; 353 darray_for_each(f->nodes, i) { 354 struct found_btree_node *prev = dst ? f->nodes.data + dst - 1 : NULL; 355 356 if (prev && 357 prev->cookie == i->cookie) { 358 if (prev->nr_ptrs == ARRAY_SIZE(prev->ptrs)) { 359 bch_err(c, "%s: found too many replicas for btree node", __func__); 360 ret = -EINVAL; 361 goto err; 362 } 363 prev->ptrs[prev->nr_ptrs++] = i->ptrs[0]; 364 } else { 365 f->nodes.data[dst++] = *i; 366 } 367 } 368 f->nodes.nr = dst; 369 370 sort(f->nodes.data, f->nodes.nr, sizeof(f->nodes.data[0]), found_btree_node_cmp_pos, NULL); 371 372 if (0 && c->opts.verbose) { 373 printbuf_reset(&buf); 374 prt_printf(&buf, "%s: nodes after merging replicas:\n", __func__); 375 found_btree_nodes_to_text(&buf, c, f->nodes); 376 bch2_print_string_as_lines(KERN_INFO, buf.buf); 377 } 378 379 dst = 0; 380 darray_for_each(f->nodes, i) { 381 if (i->overwritten) 382 continue; 383 384 ret = handle_overwrites(c, i, &darray_top(f->nodes)); 385 if (ret) 386 goto err; 387 388 BUG_ON(i->overwritten); 389 f->nodes.data[dst++] = *i; 390 } 391 f->nodes.nr = dst; 392 393 if (c->opts.verbose) { 394 printbuf_reset(&buf); 395 prt_printf(&buf, "%s: nodes found after overwrites:\n", __func__); 396 found_btree_nodes_to_text(&buf, c, f->nodes); 397 bch2_print_string_as_lines(KERN_INFO, buf.buf); 398 } 399 400 eytzinger0_sort(f->nodes.data, f->nodes.nr, sizeof(f->nodes.data[0]), found_btree_node_cmp_pos, NULL); 401 err: 402 printbuf_exit(&buf); 403 return ret; 404 } 405 406 static int found_btree_node_range_start_cmp(const void *_l, const void *_r) 407 { 408 const struct found_btree_node *l = _l; 409 const struct found_btree_node *r = _r; 410 411 return cmp_int(l->btree_id, r->btree_id) ?: 412 -cmp_int(l->level, r->level) ?: 413 bpos_cmp(l->max_key, r->min_key); 414 } 415 416 #define for_each_found_btree_node_in_range(_f, _search, _idx) \ 417 for (size_t _idx = eytzinger0_find_gt((_f)->nodes.data, (_f)->nodes.nr, \ 418 sizeof((_f)->nodes.data[0]), \ 419 found_btree_node_range_start_cmp, &search); \ 420 _idx < (_f)->nodes.nr && \ 421 (_f)->nodes.data[_idx].btree_id == _search.btree_id && \ 422 (_f)->nodes.data[_idx].level == _search.level && \ 423 bpos_lt((_f)->nodes.data[_idx].min_key, _search.max_key); \ 424 _idx = eytzinger0_next(_idx, (_f)->nodes.nr)) 425 426 bool bch2_btree_node_is_stale(struct bch_fs *c, struct btree *b) 427 { 428 struct find_btree_nodes *f = &c->found_btree_nodes; 429 430 struct found_btree_node search = { 431 .btree_id = b->c.btree_id, 432 .level = b->c.level, 433 .min_key = b->data->min_key, 434 .max_key = b->key.k.p, 435 }; 436 437 for_each_found_btree_node_in_range(f, search, idx) 438 if (f->nodes.data[idx].seq > BTREE_NODE_SEQ(b->data)) 439 return true; 440 return false; 441 } 442 443 bool bch2_btree_has_scanned_nodes(struct bch_fs *c, enum btree_id btree) 444 { 445 struct found_btree_node search = { 446 .btree_id = btree, 447 .level = 0, 448 .min_key = POS_MIN, 449 .max_key = SPOS_MAX, 450 }; 451 452 for_each_found_btree_node_in_range(&c->found_btree_nodes, search, idx) 453 return true; 454 return false; 455 } 456 457 int bch2_get_scanned_nodes(struct bch_fs *c, enum btree_id btree, 458 unsigned level, struct bpos node_min, struct bpos node_max) 459 { 460 if (btree_id_is_alloc(btree)) 461 return 0; 462 463 struct find_btree_nodes *f = &c->found_btree_nodes; 464 465 int ret = bch2_run_explicit_recovery_pass(c, BCH_RECOVERY_PASS_scan_for_btree_nodes); 466 if (ret) 467 return ret; 468 469 if (c->opts.verbose) { 470 struct printbuf buf = PRINTBUF; 471 472 prt_printf(&buf, "recovering %s l=%u ", bch2_btree_id_str(btree), level); 473 bch2_bpos_to_text(&buf, node_min); 474 prt_str(&buf, " - "); 475 bch2_bpos_to_text(&buf, node_max); 476 477 bch_info(c, "%s(): %s", __func__, buf.buf); 478 printbuf_exit(&buf); 479 } 480 481 struct found_btree_node search = { 482 .btree_id = btree, 483 .level = level, 484 .min_key = node_min, 485 .max_key = node_max, 486 }; 487 488 for_each_found_btree_node_in_range(f, search, idx) { 489 struct found_btree_node n = f->nodes.data[idx]; 490 491 n.range_updated |= bpos_lt(n.min_key, node_min); 492 n.min_key = bpos_max(n.min_key, node_min); 493 494 n.range_updated |= bpos_gt(n.max_key, node_max); 495 n.max_key = bpos_min(n.max_key, node_max); 496 497 struct { __BKEY_PADDED(k, BKEY_BTREE_PTR_VAL_U64s_MAX); } tmp; 498 499 found_btree_node_to_key(&tmp.k, &n); 500 501 struct printbuf buf = PRINTBUF; 502 bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(&tmp.k)); 503 bch_verbose(c, "%s(): recovering %s", __func__, buf.buf); 504 printbuf_exit(&buf); 505 506 BUG_ON(bch2_bkey_invalid(c, bkey_i_to_s_c(&tmp.k), BKEY_TYPE_btree, 0, NULL)); 507 508 ret = bch2_journal_key_insert(c, btree, level + 1, &tmp.k); 509 if (ret) 510 return ret; 511 } 512 513 return 0; 514 } 515 516 void bch2_find_btree_nodes_exit(struct find_btree_nodes *f) 517 { 518 darray_exit(&f->nodes); 519 } 520