1 // SPDX-License-Identifier: GPL-2.0 2 /* 3 * Copyright (c) 2014 Red Hat, Inc. 4 * All Rights Reserved. 5 */ 6 #include "xfs.h" 7 #include "xfs_fs.h" 8 #include "xfs_shared.h" 9 #include "xfs_format.h" 10 #include "xfs_log_format.h" 11 #include "xfs_trans_resv.h" 12 #include "xfs_mount.h" 13 #include "xfs_trans.h" 14 #include "xfs_alloc.h" 15 #include "xfs_btree.h" 16 #include "xfs_btree_staging.h" 17 #include "xfs_rmap.h" 18 #include "xfs_rmap_btree.h" 19 #include "xfs_trace.h" 20 #include "xfs_error.h" 21 #include "xfs_extent_busy.h" 22 #include "xfs_ag.h" 23 #include "xfs_ag_resv.h" 24 25 static struct kmem_cache *xfs_rmapbt_cur_cache; 26 27 /* 28 * Reverse map btree. 29 * 30 * This is a per-ag tree used to track the owner(s) of a given extent. With 31 * reflink it is possible for there to be multiple owners, which is a departure 32 * from classic XFS. Owner records for data extents are inserted when the 33 * extent is mapped and removed when an extent is unmapped. Owner records for 34 * all other block types (i.e. metadata) are inserted when an extent is 35 * allocated and removed when an extent is freed. There can only be one owner 36 * of a metadata extent, usually an inode or some other metadata structure like 37 * an AG btree. 38 * 39 * The rmap btree is part of the free space management, so blocks for the tree 40 * are sourced from the agfl. Hence we need transaction reservation support for 41 * this tree so that the freelist is always large enough. This also impacts on 42 * the minimum space we need to leave free in the AG. 43 * 44 * The tree is ordered by [ag block, owner, offset]. This is a large key size, 45 * but it is the only way to enforce unique keys when a block can be owned by 46 * multiple files at any offset. There's no need to order/search by extent 47 * size for online updating/management of the tree. It is intended that most 48 * reverse lookups will be to find the owner(s) of a particular block, or to 49 * try to recover tree and file data from corrupt primary metadata. 50 */ 51 52 static struct xfs_btree_cur * 53 xfs_rmapbt_dup_cursor( 54 struct xfs_btree_cur *cur) 55 { 56 return xfs_rmapbt_init_cursor(cur->bc_mp, cur->bc_tp, 57 cur->bc_ag.agbp, cur->bc_ag.pag); 58 } 59 60 STATIC void 61 xfs_rmapbt_set_root( 62 struct xfs_btree_cur *cur, 63 const union xfs_btree_ptr *ptr, 64 int inc) 65 { 66 struct xfs_buf *agbp = cur->bc_ag.agbp; 67 struct xfs_agf *agf = agbp->b_addr; 68 int btnum = cur->bc_btnum; 69 70 ASSERT(ptr->s != 0); 71 72 agf->agf_roots[btnum] = ptr->s; 73 be32_add_cpu(&agf->agf_levels[btnum], inc); 74 cur->bc_ag.pag->pagf_levels[btnum] += inc; 75 76 xfs_alloc_log_agf(cur->bc_tp, agbp, XFS_AGF_ROOTS | XFS_AGF_LEVELS); 77 } 78 79 STATIC int 80 xfs_rmapbt_alloc_block( 81 struct xfs_btree_cur *cur, 82 const union xfs_btree_ptr *start, 83 union xfs_btree_ptr *new, 84 int *stat) 85 { 86 struct xfs_buf *agbp = cur->bc_ag.agbp; 87 struct xfs_agf *agf = agbp->b_addr; 88 struct xfs_perag *pag = cur->bc_ag.pag; 89 int error; 90 xfs_agblock_t bno; 91 92 /* Allocate the new block from the freelist. If we can't, give up. */ 93 error = xfs_alloc_get_freelist(cur->bc_tp, cur->bc_ag.agbp, 94 &bno, 1); 95 if (error) 96 return error; 97 98 trace_xfs_rmapbt_alloc_block(cur->bc_mp, pag->pag_agno, bno, 1); 99 if (bno == NULLAGBLOCK) { 100 *stat = 0; 101 return 0; 102 } 103 104 xfs_extent_busy_reuse(cur->bc_mp, pag, bno, 1, false); 105 106 new->s = cpu_to_be32(bno); 107 be32_add_cpu(&agf->agf_rmap_blocks, 1); 108 xfs_alloc_log_agf(cur->bc_tp, agbp, XFS_AGF_RMAP_BLOCKS); 109 110 xfs_ag_resv_rmapbt_alloc(cur->bc_mp, pag->pag_agno); 111 112 *stat = 1; 113 return 0; 114 } 115 116 STATIC int 117 xfs_rmapbt_free_block( 118 struct xfs_btree_cur *cur, 119 struct xfs_buf *bp) 120 { 121 struct xfs_buf *agbp = cur->bc_ag.agbp; 122 struct xfs_agf *agf = agbp->b_addr; 123 struct xfs_perag *pag = cur->bc_ag.pag; 124 xfs_agblock_t bno; 125 int error; 126 127 bno = xfs_daddr_to_agbno(cur->bc_mp, xfs_buf_daddr(bp)); 128 trace_xfs_rmapbt_free_block(cur->bc_mp, pag->pag_agno, 129 bno, 1); 130 be32_add_cpu(&agf->agf_rmap_blocks, -1); 131 xfs_alloc_log_agf(cur->bc_tp, agbp, XFS_AGF_RMAP_BLOCKS); 132 error = xfs_alloc_put_freelist(cur->bc_tp, agbp, NULL, bno, 1); 133 if (error) 134 return error; 135 136 xfs_extent_busy_insert(cur->bc_tp, pag, bno, 1, 137 XFS_EXTENT_BUSY_SKIP_DISCARD); 138 139 xfs_ag_resv_free_extent(pag, XFS_AG_RESV_RMAPBT, NULL, 1); 140 return 0; 141 } 142 143 STATIC int 144 xfs_rmapbt_get_minrecs( 145 struct xfs_btree_cur *cur, 146 int level) 147 { 148 return cur->bc_mp->m_rmap_mnr[level != 0]; 149 } 150 151 STATIC int 152 xfs_rmapbt_get_maxrecs( 153 struct xfs_btree_cur *cur, 154 int level) 155 { 156 return cur->bc_mp->m_rmap_mxr[level != 0]; 157 } 158 159 STATIC void 160 xfs_rmapbt_init_key_from_rec( 161 union xfs_btree_key *key, 162 const union xfs_btree_rec *rec) 163 { 164 key->rmap.rm_startblock = rec->rmap.rm_startblock; 165 key->rmap.rm_owner = rec->rmap.rm_owner; 166 key->rmap.rm_offset = rec->rmap.rm_offset; 167 } 168 169 /* 170 * The high key for a reverse mapping record can be computed by shifting 171 * the startblock and offset to the highest value that would still map 172 * to that record. In practice this means that we add blockcount-1 to 173 * the startblock for all records, and if the record is for a data/attr 174 * fork mapping, we add blockcount-1 to the offset too. 175 */ 176 STATIC void 177 xfs_rmapbt_init_high_key_from_rec( 178 union xfs_btree_key *key, 179 const union xfs_btree_rec *rec) 180 { 181 uint64_t off; 182 int adj; 183 184 adj = be32_to_cpu(rec->rmap.rm_blockcount) - 1; 185 186 key->rmap.rm_startblock = rec->rmap.rm_startblock; 187 be32_add_cpu(&key->rmap.rm_startblock, adj); 188 key->rmap.rm_owner = rec->rmap.rm_owner; 189 key->rmap.rm_offset = rec->rmap.rm_offset; 190 if (XFS_RMAP_NON_INODE_OWNER(be64_to_cpu(rec->rmap.rm_owner)) || 191 XFS_RMAP_IS_BMBT_BLOCK(be64_to_cpu(rec->rmap.rm_offset))) 192 return; 193 off = be64_to_cpu(key->rmap.rm_offset); 194 off = (XFS_RMAP_OFF(off) + adj) | (off & ~XFS_RMAP_OFF_MASK); 195 key->rmap.rm_offset = cpu_to_be64(off); 196 } 197 198 STATIC void 199 xfs_rmapbt_init_rec_from_cur( 200 struct xfs_btree_cur *cur, 201 union xfs_btree_rec *rec) 202 { 203 rec->rmap.rm_startblock = cpu_to_be32(cur->bc_rec.r.rm_startblock); 204 rec->rmap.rm_blockcount = cpu_to_be32(cur->bc_rec.r.rm_blockcount); 205 rec->rmap.rm_owner = cpu_to_be64(cur->bc_rec.r.rm_owner); 206 rec->rmap.rm_offset = cpu_to_be64( 207 xfs_rmap_irec_offset_pack(&cur->bc_rec.r)); 208 } 209 210 STATIC void 211 xfs_rmapbt_init_ptr_from_cur( 212 struct xfs_btree_cur *cur, 213 union xfs_btree_ptr *ptr) 214 { 215 struct xfs_agf *agf = cur->bc_ag.agbp->b_addr; 216 217 ASSERT(cur->bc_ag.pag->pag_agno == be32_to_cpu(agf->agf_seqno)); 218 219 ptr->s = agf->agf_roots[cur->bc_btnum]; 220 } 221 222 STATIC int64_t 223 xfs_rmapbt_key_diff( 224 struct xfs_btree_cur *cur, 225 const union xfs_btree_key *key) 226 { 227 struct xfs_rmap_irec *rec = &cur->bc_rec.r; 228 const struct xfs_rmap_key *kp = &key->rmap; 229 __u64 x, y; 230 int64_t d; 231 232 d = (int64_t)be32_to_cpu(kp->rm_startblock) - rec->rm_startblock; 233 if (d) 234 return d; 235 236 x = be64_to_cpu(kp->rm_owner); 237 y = rec->rm_owner; 238 if (x > y) 239 return 1; 240 else if (y > x) 241 return -1; 242 243 x = XFS_RMAP_OFF(be64_to_cpu(kp->rm_offset)); 244 y = rec->rm_offset; 245 if (x > y) 246 return 1; 247 else if (y > x) 248 return -1; 249 return 0; 250 } 251 252 STATIC int64_t 253 xfs_rmapbt_diff_two_keys( 254 struct xfs_btree_cur *cur, 255 const union xfs_btree_key *k1, 256 const union xfs_btree_key *k2) 257 { 258 const struct xfs_rmap_key *kp1 = &k1->rmap; 259 const struct xfs_rmap_key *kp2 = &k2->rmap; 260 int64_t d; 261 __u64 x, y; 262 263 d = (int64_t)be32_to_cpu(kp1->rm_startblock) - 264 be32_to_cpu(kp2->rm_startblock); 265 if (d) 266 return d; 267 268 x = be64_to_cpu(kp1->rm_owner); 269 y = be64_to_cpu(kp2->rm_owner); 270 if (x > y) 271 return 1; 272 else if (y > x) 273 return -1; 274 275 x = XFS_RMAP_OFF(be64_to_cpu(kp1->rm_offset)); 276 y = XFS_RMAP_OFF(be64_to_cpu(kp2->rm_offset)); 277 if (x > y) 278 return 1; 279 else if (y > x) 280 return -1; 281 return 0; 282 } 283 284 static xfs_failaddr_t 285 xfs_rmapbt_verify( 286 struct xfs_buf *bp) 287 { 288 struct xfs_mount *mp = bp->b_mount; 289 struct xfs_btree_block *block = XFS_BUF_TO_BLOCK(bp); 290 struct xfs_perag *pag = bp->b_pag; 291 xfs_failaddr_t fa; 292 unsigned int level; 293 294 /* 295 * magic number and level verification 296 * 297 * During growfs operations, we can't verify the exact level or owner as 298 * the perag is not fully initialised and hence not attached to the 299 * buffer. In this case, check against the maximum tree depth. 300 * 301 * Similarly, during log recovery we will have a perag structure 302 * attached, but the agf information will not yet have been initialised 303 * from the on disk AGF. Again, we can only check against maximum limits 304 * in this case. 305 */ 306 if (!xfs_verify_magic(bp, block->bb_magic)) 307 return __this_address; 308 309 if (!xfs_has_rmapbt(mp)) 310 return __this_address; 311 fa = xfs_btree_sblock_v5hdr_verify(bp); 312 if (fa) 313 return fa; 314 315 level = be16_to_cpu(block->bb_level); 316 if (pag && pag->pagf_init) { 317 if (level >= pag->pagf_levels[XFS_BTNUM_RMAPi]) 318 return __this_address; 319 } else if (level >= mp->m_rmap_maxlevels) 320 return __this_address; 321 322 return xfs_btree_sblock_verify(bp, mp->m_rmap_mxr[level != 0]); 323 } 324 325 static void 326 xfs_rmapbt_read_verify( 327 struct xfs_buf *bp) 328 { 329 xfs_failaddr_t fa; 330 331 if (!xfs_btree_sblock_verify_crc(bp)) 332 xfs_verifier_error(bp, -EFSBADCRC, __this_address); 333 else { 334 fa = xfs_rmapbt_verify(bp); 335 if (fa) 336 xfs_verifier_error(bp, -EFSCORRUPTED, fa); 337 } 338 339 if (bp->b_error) 340 trace_xfs_btree_corrupt(bp, _RET_IP_); 341 } 342 343 static void 344 xfs_rmapbt_write_verify( 345 struct xfs_buf *bp) 346 { 347 xfs_failaddr_t fa; 348 349 fa = xfs_rmapbt_verify(bp); 350 if (fa) { 351 trace_xfs_btree_corrupt(bp, _RET_IP_); 352 xfs_verifier_error(bp, -EFSCORRUPTED, fa); 353 return; 354 } 355 xfs_btree_sblock_calc_crc(bp); 356 357 } 358 359 const struct xfs_buf_ops xfs_rmapbt_buf_ops = { 360 .name = "xfs_rmapbt", 361 .magic = { 0, cpu_to_be32(XFS_RMAP_CRC_MAGIC) }, 362 .verify_read = xfs_rmapbt_read_verify, 363 .verify_write = xfs_rmapbt_write_verify, 364 .verify_struct = xfs_rmapbt_verify, 365 }; 366 367 STATIC int 368 xfs_rmapbt_keys_inorder( 369 struct xfs_btree_cur *cur, 370 const union xfs_btree_key *k1, 371 const union xfs_btree_key *k2) 372 { 373 uint32_t x; 374 uint32_t y; 375 uint64_t a; 376 uint64_t b; 377 378 x = be32_to_cpu(k1->rmap.rm_startblock); 379 y = be32_to_cpu(k2->rmap.rm_startblock); 380 if (x < y) 381 return 1; 382 else if (x > y) 383 return 0; 384 a = be64_to_cpu(k1->rmap.rm_owner); 385 b = be64_to_cpu(k2->rmap.rm_owner); 386 if (a < b) 387 return 1; 388 else if (a > b) 389 return 0; 390 a = XFS_RMAP_OFF(be64_to_cpu(k1->rmap.rm_offset)); 391 b = XFS_RMAP_OFF(be64_to_cpu(k2->rmap.rm_offset)); 392 if (a <= b) 393 return 1; 394 return 0; 395 } 396 397 STATIC int 398 xfs_rmapbt_recs_inorder( 399 struct xfs_btree_cur *cur, 400 const union xfs_btree_rec *r1, 401 const union xfs_btree_rec *r2) 402 { 403 uint32_t x; 404 uint32_t y; 405 uint64_t a; 406 uint64_t b; 407 408 x = be32_to_cpu(r1->rmap.rm_startblock); 409 y = be32_to_cpu(r2->rmap.rm_startblock); 410 if (x < y) 411 return 1; 412 else if (x > y) 413 return 0; 414 a = be64_to_cpu(r1->rmap.rm_owner); 415 b = be64_to_cpu(r2->rmap.rm_owner); 416 if (a < b) 417 return 1; 418 else if (a > b) 419 return 0; 420 a = XFS_RMAP_OFF(be64_to_cpu(r1->rmap.rm_offset)); 421 b = XFS_RMAP_OFF(be64_to_cpu(r2->rmap.rm_offset)); 422 if (a <= b) 423 return 1; 424 return 0; 425 } 426 427 static const struct xfs_btree_ops xfs_rmapbt_ops = { 428 .rec_len = sizeof(struct xfs_rmap_rec), 429 .key_len = 2 * sizeof(struct xfs_rmap_key), 430 431 .dup_cursor = xfs_rmapbt_dup_cursor, 432 .set_root = xfs_rmapbt_set_root, 433 .alloc_block = xfs_rmapbt_alloc_block, 434 .free_block = xfs_rmapbt_free_block, 435 .get_minrecs = xfs_rmapbt_get_minrecs, 436 .get_maxrecs = xfs_rmapbt_get_maxrecs, 437 .init_key_from_rec = xfs_rmapbt_init_key_from_rec, 438 .init_high_key_from_rec = xfs_rmapbt_init_high_key_from_rec, 439 .init_rec_from_cur = xfs_rmapbt_init_rec_from_cur, 440 .init_ptr_from_cur = xfs_rmapbt_init_ptr_from_cur, 441 .key_diff = xfs_rmapbt_key_diff, 442 .buf_ops = &xfs_rmapbt_buf_ops, 443 .diff_two_keys = xfs_rmapbt_diff_two_keys, 444 .keys_inorder = xfs_rmapbt_keys_inorder, 445 .recs_inorder = xfs_rmapbt_recs_inorder, 446 }; 447 448 static struct xfs_btree_cur * 449 xfs_rmapbt_init_common( 450 struct xfs_mount *mp, 451 struct xfs_trans *tp, 452 struct xfs_perag *pag) 453 { 454 struct xfs_btree_cur *cur; 455 456 /* Overlapping btree; 2 keys per pointer. */ 457 cur = xfs_btree_alloc_cursor(mp, tp, XFS_BTNUM_RMAP, 458 mp->m_rmap_maxlevels, xfs_rmapbt_cur_cache); 459 cur->bc_flags = XFS_BTREE_CRC_BLOCKS | XFS_BTREE_OVERLAPPING; 460 cur->bc_statoff = XFS_STATS_CALC_INDEX(xs_rmap_2); 461 cur->bc_ops = &xfs_rmapbt_ops; 462 463 /* take a reference for the cursor */ 464 atomic_inc(&pag->pag_ref); 465 cur->bc_ag.pag = pag; 466 467 return cur; 468 } 469 470 /* Create a new reverse mapping btree cursor. */ 471 struct xfs_btree_cur * 472 xfs_rmapbt_init_cursor( 473 struct xfs_mount *mp, 474 struct xfs_trans *tp, 475 struct xfs_buf *agbp, 476 struct xfs_perag *pag) 477 { 478 struct xfs_agf *agf = agbp->b_addr; 479 struct xfs_btree_cur *cur; 480 481 cur = xfs_rmapbt_init_common(mp, tp, pag); 482 cur->bc_nlevels = be32_to_cpu(agf->agf_levels[XFS_BTNUM_RMAP]); 483 cur->bc_ag.agbp = agbp; 484 return cur; 485 } 486 487 /* Create a new reverse mapping btree cursor with a fake root for staging. */ 488 struct xfs_btree_cur * 489 xfs_rmapbt_stage_cursor( 490 struct xfs_mount *mp, 491 struct xbtree_afakeroot *afake, 492 struct xfs_perag *pag) 493 { 494 struct xfs_btree_cur *cur; 495 496 cur = xfs_rmapbt_init_common(mp, NULL, pag); 497 xfs_btree_stage_afakeroot(cur, afake); 498 return cur; 499 } 500 501 /* 502 * Install a new reverse mapping btree root. Caller is responsible for 503 * invalidating and freeing the old btree blocks. 504 */ 505 void 506 xfs_rmapbt_commit_staged_btree( 507 struct xfs_btree_cur *cur, 508 struct xfs_trans *tp, 509 struct xfs_buf *agbp) 510 { 511 struct xfs_agf *agf = agbp->b_addr; 512 struct xbtree_afakeroot *afake = cur->bc_ag.afake; 513 514 ASSERT(cur->bc_flags & XFS_BTREE_STAGING); 515 516 agf->agf_roots[cur->bc_btnum] = cpu_to_be32(afake->af_root); 517 agf->agf_levels[cur->bc_btnum] = cpu_to_be32(afake->af_levels); 518 agf->agf_rmap_blocks = cpu_to_be32(afake->af_blocks); 519 xfs_alloc_log_agf(tp, agbp, XFS_AGF_ROOTS | XFS_AGF_LEVELS | 520 XFS_AGF_RMAP_BLOCKS); 521 xfs_btree_commit_afakeroot(cur, tp, agbp, &xfs_rmapbt_ops); 522 } 523 524 /* Calculate number of records in a reverse mapping btree block. */ 525 static inline unsigned int 526 xfs_rmapbt_block_maxrecs( 527 unsigned int blocklen, 528 bool leaf) 529 { 530 if (leaf) 531 return blocklen / sizeof(struct xfs_rmap_rec); 532 return blocklen / 533 (2 * sizeof(struct xfs_rmap_key) + sizeof(xfs_rmap_ptr_t)); 534 } 535 536 /* 537 * Calculate number of records in an rmap btree block. 538 */ 539 int 540 xfs_rmapbt_maxrecs( 541 int blocklen, 542 int leaf) 543 { 544 blocklen -= XFS_RMAP_BLOCK_LEN; 545 return xfs_rmapbt_block_maxrecs(blocklen, leaf); 546 } 547 548 /* Compute the max possible height for reverse mapping btrees. */ 549 unsigned int 550 xfs_rmapbt_maxlevels_ondisk(void) 551 { 552 unsigned int minrecs[2]; 553 unsigned int blocklen; 554 555 blocklen = XFS_MIN_CRC_BLOCKSIZE - XFS_BTREE_SBLOCK_CRC_LEN; 556 557 minrecs[0] = xfs_rmapbt_block_maxrecs(blocklen, true) / 2; 558 minrecs[1] = xfs_rmapbt_block_maxrecs(blocklen, false) / 2; 559 560 /* 561 * Compute the asymptotic maxlevels for an rmapbt on any reflink fs. 562 * 563 * On a reflink filesystem, each AG block can have up to 2^32 (per the 564 * refcount record format) owners, which means that theoretically we 565 * could face up to 2^64 rmap records. However, we're likely to run 566 * out of blocks in the AG long before that happens, which means that 567 * we must compute the max height based on what the btree will look 568 * like if it consumes almost all the blocks in the AG due to maximal 569 * sharing factor. 570 */ 571 return xfs_btree_space_to_height(minrecs, XFS_MAX_CRC_AG_BLOCKS); 572 } 573 574 /* Compute the maximum height of an rmap btree. */ 575 void 576 xfs_rmapbt_compute_maxlevels( 577 struct xfs_mount *mp) 578 { 579 if (!xfs_has_rmapbt(mp)) { 580 mp->m_rmap_maxlevels = 0; 581 return; 582 } 583 584 if (xfs_has_reflink(mp)) { 585 /* 586 * Compute the asymptotic maxlevels for an rmap btree on a 587 * filesystem that supports reflink. 588 * 589 * On a reflink filesystem, each AG block can have up to 2^32 590 * (per the refcount record format) owners, which means that 591 * theoretically we could face up to 2^64 rmap records. 592 * However, we're likely to run out of blocks in the AG long 593 * before that happens, which means that we must compute the 594 * max height based on what the btree will look like if it 595 * consumes almost all the blocks in the AG due to maximal 596 * sharing factor. 597 */ 598 mp->m_rmap_maxlevels = xfs_btree_space_to_height(mp->m_rmap_mnr, 599 mp->m_sb.sb_agblocks); 600 } else { 601 /* 602 * If there's no block sharing, compute the maximum rmapbt 603 * height assuming one rmap record per AG block. 604 */ 605 mp->m_rmap_maxlevels = xfs_btree_compute_maxlevels( 606 mp->m_rmap_mnr, mp->m_sb.sb_agblocks); 607 } 608 ASSERT(mp->m_rmap_maxlevels <= xfs_rmapbt_maxlevels_ondisk()); 609 } 610 611 /* Calculate the refcount btree size for some records. */ 612 xfs_extlen_t 613 xfs_rmapbt_calc_size( 614 struct xfs_mount *mp, 615 unsigned long long len) 616 { 617 return xfs_btree_calc_size(mp->m_rmap_mnr, len); 618 } 619 620 /* 621 * Calculate the maximum refcount btree size. 622 */ 623 xfs_extlen_t 624 xfs_rmapbt_max_size( 625 struct xfs_mount *mp, 626 xfs_agblock_t agblocks) 627 { 628 /* Bail out if we're uninitialized, which can happen in mkfs. */ 629 if (mp->m_rmap_mxr[0] == 0) 630 return 0; 631 632 return xfs_rmapbt_calc_size(mp, agblocks); 633 } 634 635 /* 636 * Figure out how many blocks to reserve and how many are used by this btree. 637 */ 638 int 639 xfs_rmapbt_calc_reserves( 640 struct xfs_mount *mp, 641 struct xfs_trans *tp, 642 struct xfs_perag *pag, 643 xfs_extlen_t *ask, 644 xfs_extlen_t *used) 645 { 646 struct xfs_buf *agbp; 647 struct xfs_agf *agf; 648 xfs_agblock_t agblocks; 649 xfs_extlen_t tree_len; 650 int error; 651 652 if (!xfs_has_rmapbt(mp)) 653 return 0; 654 655 error = xfs_alloc_read_agf(mp, tp, pag->pag_agno, 0, &agbp); 656 if (error) 657 return error; 658 659 agf = agbp->b_addr; 660 agblocks = be32_to_cpu(agf->agf_length); 661 tree_len = be32_to_cpu(agf->agf_rmap_blocks); 662 xfs_trans_brelse(tp, agbp); 663 664 /* 665 * The log is permanently allocated, so the space it occupies will 666 * never be available for the kinds of things that would require btree 667 * expansion. We therefore can pretend the space isn't there. 668 */ 669 if (mp->m_sb.sb_logstart && 670 XFS_FSB_TO_AGNO(mp, mp->m_sb.sb_logstart) == pag->pag_agno) 671 agblocks -= mp->m_sb.sb_logblocks; 672 673 /* Reserve 1% of the AG or enough for 1 block per record. */ 674 *ask += max(agblocks / 100, xfs_rmapbt_max_size(mp, agblocks)); 675 *used += tree_len; 676 677 return error; 678 } 679 680 int __init 681 xfs_rmapbt_init_cur_cache(void) 682 { 683 xfs_rmapbt_cur_cache = kmem_cache_create("xfs_rmapbt_cur", 684 xfs_btree_cur_sizeof(xfs_rmapbt_maxlevels_ondisk()), 685 0, 0, NULL); 686 687 if (!xfs_rmapbt_cur_cache) 688 return -ENOMEM; 689 return 0; 690 } 691 692 void 693 xfs_rmapbt_destroy_cur_cache(void) 694 { 695 kmem_cache_destroy(xfs_rmapbt_cur_cache); 696 xfs_rmapbt_cur_cache = NULL; 697 } 698