1 // SPDX-License-Identifier: GPL-2.0 2 /* 3 * Copyright (c) 2017 Christoph Hellwig. 4 */ 5 6 #include "xfs.h" 7 #include "xfs_shared.h" 8 #include "xfs_format.h" 9 #include "xfs_bit.h" 10 #include "xfs_log_format.h" 11 #include "xfs_trans_resv.h" 12 #include "xfs_mount.h" 13 #include "xfs_inode.h" 14 #include "xfs_trace.h" 15 16 /* 17 * In-core extent record layout: 18 * 19 * +-------+----------------------------+ 20 * | 00:53 | all 54 bits of startoff | 21 * | 54:63 | low 10 bits of startblock | 22 * +-------+----------------------------+ 23 * | 00:20 | all 21 bits of length | 24 * | 21 | unwritten extent bit | 25 * | 22:63 | high 42 bits of startblock | 26 * +-------+----------------------------+ 27 */ 28 #define XFS_IEXT_STARTOFF_MASK xfs_mask64lo(BMBT_STARTOFF_BITLEN) 29 #define XFS_IEXT_LENGTH_MASK xfs_mask64lo(BMBT_BLOCKCOUNT_BITLEN) 30 #define XFS_IEXT_STARTBLOCK_MASK xfs_mask64lo(BMBT_STARTBLOCK_BITLEN) 31 32 struct xfs_iext_rec { 33 uint64_t lo; 34 uint64_t hi; 35 }; 36 37 /* 38 * Given that the length can't be a zero, only an empty hi value indicates an 39 * unused record. 40 */ 41 static bool xfs_iext_rec_is_empty(struct xfs_iext_rec *rec) 42 { 43 return rec->hi == 0; 44 } 45 46 static inline void xfs_iext_rec_clear(struct xfs_iext_rec *rec) 47 { 48 rec->lo = 0; 49 rec->hi = 0; 50 } 51 52 static void 53 xfs_iext_set( 54 struct xfs_iext_rec *rec, 55 struct xfs_bmbt_irec *irec) 56 { 57 ASSERT((irec->br_startoff & ~XFS_IEXT_STARTOFF_MASK) == 0); 58 ASSERT((irec->br_blockcount & ~XFS_IEXT_LENGTH_MASK) == 0); 59 ASSERT((irec->br_startblock & ~XFS_IEXT_STARTBLOCK_MASK) == 0); 60 61 rec->lo = irec->br_startoff & XFS_IEXT_STARTOFF_MASK; 62 rec->hi = irec->br_blockcount & XFS_IEXT_LENGTH_MASK; 63 64 rec->lo |= (irec->br_startblock << 54); 65 rec->hi |= ((irec->br_startblock & ~xfs_mask64lo(10)) << (22 - 10)); 66 67 if (irec->br_state == XFS_EXT_UNWRITTEN) 68 rec->hi |= (1 << 21); 69 } 70 71 static void 72 xfs_iext_get( 73 struct xfs_bmbt_irec *irec, 74 struct xfs_iext_rec *rec) 75 { 76 irec->br_startoff = rec->lo & XFS_IEXT_STARTOFF_MASK; 77 irec->br_blockcount = rec->hi & XFS_IEXT_LENGTH_MASK; 78 79 irec->br_startblock = rec->lo >> 54; 80 irec->br_startblock |= (rec->hi & xfs_mask64hi(42)) >> (22 - 10); 81 82 if (rec->hi & (1 << 21)) 83 irec->br_state = XFS_EXT_UNWRITTEN; 84 else 85 irec->br_state = XFS_EXT_NORM; 86 } 87 88 enum { 89 NODE_SIZE = 256, 90 KEYS_PER_NODE = NODE_SIZE / (sizeof(uint64_t) + sizeof(void *)), 91 RECS_PER_LEAF = (NODE_SIZE - (2 * sizeof(struct xfs_iext_leaf *))) / 92 sizeof(struct xfs_iext_rec), 93 }; 94 95 /* 96 * In-core extent btree block layout: 97 * 98 * There are two types of blocks in the btree: leaf and inner (non-leaf) blocks. 99 * 100 * The leaf blocks are made up by %KEYS_PER_NODE extent records, which each 101 * contain the startoffset, blockcount, startblock and unwritten extent flag. 102 * See above for the exact format, followed by pointers to the previous and next 103 * leaf blocks (if there are any). 104 * 105 * The inner (non-leaf) blocks first contain KEYS_PER_NODE lookup keys, followed 106 * by an equal number of pointers to the btree blocks at the next lower level. 107 * 108 * +-------+-------+-------+-------+-------+----------+----------+ 109 * Leaf: | rec 1 | rec 2 | rec 3 | rec 4 | rec N | prev-ptr | next-ptr | 110 * +-------+-------+-------+-------+-------+----------+----------+ 111 * 112 * +-------+-------+-------+-------+-------+-------+------+-------+ 113 * Inner: | key 1 | key 2 | key 3 | key N | ptr 1 | ptr 2 | ptr3 | ptr N | 114 * +-------+-------+-------+-------+-------+-------+------+-------+ 115 */ 116 struct xfs_iext_node { 117 uint64_t keys[KEYS_PER_NODE]; 118 #define XFS_IEXT_KEY_INVALID (1ULL << 63) 119 void *ptrs[KEYS_PER_NODE]; 120 }; 121 122 struct xfs_iext_leaf { 123 struct xfs_iext_rec recs[RECS_PER_LEAF]; 124 struct xfs_iext_leaf *prev; 125 struct xfs_iext_leaf *next; 126 }; 127 128 inline xfs_extnum_t xfs_iext_count(struct xfs_ifork *ifp) 129 { 130 return ifp->if_bytes / sizeof(struct xfs_iext_rec); 131 } 132 133 static inline int xfs_iext_max_recs(struct xfs_ifork *ifp) 134 { 135 if (ifp->if_height == 1) 136 return xfs_iext_count(ifp); 137 return RECS_PER_LEAF; 138 } 139 140 static inline struct xfs_iext_rec *cur_rec(struct xfs_iext_cursor *cur) 141 { 142 return &cur->leaf->recs[cur->pos]; 143 } 144 145 static inline bool xfs_iext_valid(struct xfs_ifork *ifp, 146 struct xfs_iext_cursor *cur) 147 { 148 if (!cur->leaf) 149 return false; 150 if (cur->pos < 0 || cur->pos >= xfs_iext_max_recs(ifp)) 151 return false; 152 if (xfs_iext_rec_is_empty(cur_rec(cur))) 153 return false; 154 return true; 155 } 156 157 static void * 158 xfs_iext_find_first_leaf( 159 struct xfs_ifork *ifp) 160 { 161 struct xfs_iext_node *node = ifp->if_data; 162 int height; 163 164 if (!ifp->if_height) 165 return NULL; 166 167 for (height = ifp->if_height; height > 1; height--) { 168 node = node->ptrs[0]; 169 ASSERT(node); 170 } 171 172 return node; 173 } 174 175 static void * 176 xfs_iext_find_last_leaf( 177 struct xfs_ifork *ifp) 178 { 179 struct xfs_iext_node *node = ifp->if_data; 180 int height, i; 181 182 if (!ifp->if_height) 183 return NULL; 184 185 for (height = ifp->if_height; height > 1; height--) { 186 for (i = 1; i < KEYS_PER_NODE; i++) 187 if (!node->ptrs[i]) 188 break; 189 node = node->ptrs[i - 1]; 190 ASSERT(node); 191 } 192 193 return node; 194 } 195 196 void 197 xfs_iext_first( 198 struct xfs_ifork *ifp, 199 struct xfs_iext_cursor *cur) 200 { 201 cur->pos = 0; 202 cur->leaf = xfs_iext_find_first_leaf(ifp); 203 } 204 205 void 206 xfs_iext_last( 207 struct xfs_ifork *ifp, 208 struct xfs_iext_cursor *cur) 209 { 210 int i; 211 212 cur->leaf = xfs_iext_find_last_leaf(ifp); 213 if (!cur->leaf) { 214 cur->pos = 0; 215 return; 216 } 217 218 for (i = 1; i < xfs_iext_max_recs(ifp); i++) { 219 if (xfs_iext_rec_is_empty(&cur->leaf->recs[i])) 220 break; 221 } 222 cur->pos = i - 1; 223 } 224 225 void 226 xfs_iext_next( 227 struct xfs_ifork *ifp, 228 struct xfs_iext_cursor *cur) 229 { 230 if (!cur->leaf) { 231 ASSERT(cur->pos <= 0 || cur->pos >= RECS_PER_LEAF); 232 xfs_iext_first(ifp, cur); 233 return; 234 } 235 236 ASSERT(cur->pos >= 0); 237 ASSERT(cur->pos < xfs_iext_max_recs(ifp)); 238 239 cur->pos++; 240 if (ifp->if_height > 1 && !xfs_iext_valid(ifp, cur) && 241 cur->leaf->next) { 242 cur->leaf = cur->leaf->next; 243 cur->pos = 0; 244 } 245 } 246 247 void 248 xfs_iext_prev( 249 struct xfs_ifork *ifp, 250 struct xfs_iext_cursor *cur) 251 { 252 if (!cur->leaf) { 253 ASSERT(cur->pos <= 0 || cur->pos >= RECS_PER_LEAF); 254 xfs_iext_last(ifp, cur); 255 return; 256 } 257 258 ASSERT(cur->pos >= 0); 259 ASSERT(cur->pos <= RECS_PER_LEAF); 260 261 recurse: 262 do { 263 cur->pos--; 264 if (xfs_iext_valid(ifp, cur)) 265 return; 266 } while (cur->pos > 0); 267 268 if (ifp->if_height > 1 && cur->leaf->prev) { 269 cur->leaf = cur->leaf->prev; 270 cur->pos = RECS_PER_LEAF; 271 goto recurse; 272 } 273 } 274 275 static inline int 276 xfs_iext_key_cmp( 277 struct xfs_iext_node *node, 278 int n, 279 xfs_fileoff_t offset) 280 { 281 if (node->keys[n] > offset) 282 return 1; 283 if (node->keys[n] < offset) 284 return -1; 285 return 0; 286 } 287 288 static inline int 289 xfs_iext_rec_cmp( 290 struct xfs_iext_rec *rec, 291 xfs_fileoff_t offset) 292 { 293 uint64_t rec_offset = rec->lo & XFS_IEXT_STARTOFF_MASK; 294 uint32_t rec_len = rec->hi & XFS_IEXT_LENGTH_MASK; 295 296 if (rec_offset > offset) 297 return 1; 298 if (rec_offset + rec_len <= offset) 299 return -1; 300 return 0; 301 } 302 303 static void * 304 xfs_iext_find_level( 305 struct xfs_ifork *ifp, 306 xfs_fileoff_t offset, 307 int level) 308 { 309 struct xfs_iext_node *node = ifp->if_data; 310 int height, i; 311 312 if (!ifp->if_height) 313 return NULL; 314 315 for (height = ifp->if_height; height > level; height--) { 316 for (i = 1; i < KEYS_PER_NODE; i++) 317 if (xfs_iext_key_cmp(node, i, offset) > 0) 318 break; 319 320 node = node->ptrs[i - 1]; 321 if (!node) 322 break; 323 } 324 325 return node; 326 } 327 328 static int 329 xfs_iext_node_pos( 330 struct xfs_iext_node *node, 331 xfs_fileoff_t offset) 332 { 333 int i; 334 335 for (i = 1; i < KEYS_PER_NODE; i++) { 336 if (xfs_iext_key_cmp(node, i, offset) > 0) 337 break; 338 } 339 340 return i - 1; 341 } 342 343 static int 344 xfs_iext_node_insert_pos( 345 struct xfs_iext_node *node, 346 xfs_fileoff_t offset) 347 { 348 int i; 349 350 for (i = 0; i < KEYS_PER_NODE; i++) { 351 if (xfs_iext_key_cmp(node, i, offset) > 0) 352 return i; 353 } 354 355 return KEYS_PER_NODE; 356 } 357 358 static int 359 xfs_iext_node_nr_entries( 360 struct xfs_iext_node *node, 361 int start) 362 { 363 int i; 364 365 for (i = start; i < KEYS_PER_NODE; i++) { 366 if (node->keys[i] == XFS_IEXT_KEY_INVALID) 367 break; 368 } 369 370 return i; 371 } 372 373 static int 374 xfs_iext_leaf_nr_entries( 375 struct xfs_ifork *ifp, 376 struct xfs_iext_leaf *leaf, 377 int start) 378 { 379 int i; 380 381 for (i = start; i < xfs_iext_max_recs(ifp); i++) { 382 if (xfs_iext_rec_is_empty(&leaf->recs[i])) 383 break; 384 } 385 386 return i; 387 } 388 389 static inline uint64_t 390 xfs_iext_leaf_key( 391 struct xfs_iext_leaf *leaf, 392 int n) 393 { 394 return leaf->recs[n].lo & XFS_IEXT_STARTOFF_MASK; 395 } 396 397 static inline void * 398 xfs_iext_alloc_node( 399 int size) 400 { 401 return kzalloc(size, GFP_KERNEL | __GFP_NOLOCKDEP | __GFP_NOFAIL); 402 } 403 404 static void 405 xfs_iext_grow( 406 struct xfs_ifork *ifp) 407 { 408 struct xfs_iext_node *node = xfs_iext_alloc_node(NODE_SIZE); 409 int i; 410 411 if (ifp->if_height == 1) { 412 struct xfs_iext_leaf *prev = ifp->if_data; 413 414 node->keys[0] = xfs_iext_leaf_key(prev, 0); 415 node->ptrs[0] = prev; 416 } else { 417 struct xfs_iext_node *prev = ifp->if_data; 418 419 ASSERT(ifp->if_height > 1); 420 421 node->keys[0] = prev->keys[0]; 422 node->ptrs[0] = prev; 423 } 424 425 for (i = 1; i < KEYS_PER_NODE; i++) 426 node->keys[i] = XFS_IEXT_KEY_INVALID; 427 428 ifp->if_data = node; 429 ifp->if_height++; 430 } 431 432 static void 433 xfs_iext_update_node( 434 struct xfs_ifork *ifp, 435 xfs_fileoff_t old_offset, 436 xfs_fileoff_t new_offset, 437 int level, 438 void *ptr) 439 { 440 struct xfs_iext_node *node = ifp->if_data; 441 int height, i; 442 443 for (height = ifp->if_height; height > level; height--) { 444 for (i = 0; i < KEYS_PER_NODE; i++) { 445 if (i > 0 && xfs_iext_key_cmp(node, i, old_offset) > 0) 446 break; 447 if (node->keys[i] == old_offset) 448 node->keys[i] = new_offset; 449 } 450 node = node->ptrs[i - 1]; 451 ASSERT(node); 452 } 453 454 ASSERT(node == ptr); 455 } 456 457 static struct xfs_iext_node * 458 xfs_iext_split_node( 459 struct xfs_iext_node **nodep, 460 int *pos, 461 int *nr_entries) 462 { 463 struct xfs_iext_node *node = *nodep; 464 struct xfs_iext_node *new = xfs_iext_alloc_node(NODE_SIZE); 465 const int nr_move = KEYS_PER_NODE / 2; 466 int nr_keep = nr_move + (KEYS_PER_NODE & 1); 467 int i = 0; 468 469 /* for sequential append operations just spill over into the new node */ 470 if (*pos == KEYS_PER_NODE) { 471 *nodep = new; 472 *pos = 0; 473 *nr_entries = 0; 474 goto done; 475 } 476 477 478 for (i = 0; i < nr_move; i++) { 479 new->keys[i] = node->keys[nr_keep + i]; 480 new->ptrs[i] = node->ptrs[nr_keep + i]; 481 482 node->keys[nr_keep + i] = XFS_IEXT_KEY_INVALID; 483 node->ptrs[nr_keep + i] = NULL; 484 } 485 486 if (*pos >= nr_keep) { 487 *nodep = new; 488 *pos -= nr_keep; 489 *nr_entries = nr_move; 490 } else { 491 *nr_entries = nr_keep; 492 } 493 done: 494 for (; i < KEYS_PER_NODE; i++) 495 new->keys[i] = XFS_IEXT_KEY_INVALID; 496 return new; 497 } 498 499 static void 500 xfs_iext_insert_node( 501 struct xfs_ifork *ifp, 502 uint64_t offset, 503 void *ptr, 504 int level) 505 { 506 struct xfs_iext_node *node, *new; 507 int i, pos, nr_entries; 508 509 again: 510 if (ifp->if_height < level) 511 xfs_iext_grow(ifp); 512 513 new = NULL; 514 node = xfs_iext_find_level(ifp, offset, level); 515 pos = xfs_iext_node_insert_pos(node, offset); 516 nr_entries = xfs_iext_node_nr_entries(node, pos); 517 518 ASSERT(pos >= nr_entries || xfs_iext_key_cmp(node, pos, offset) != 0); 519 ASSERT(nr_entries <= KEYS_PER_NODE); 520 521 if (nr_entries == KEYS_PER_NODE) 522 new = xfs_iext_split_node(&node, &pos, &nr_entries); 523 524 /* 525 * Update the pointers in higher levels if the first entry changes 526 * in an existing node. 527 */ 528 if (node != new && pos == 0 && nr_entries > 0) 529 xfs_iext_update_node(ifp, node->keys[0], offset, level, node); 530 531 for (i = nr_entries; i > pos; i--) { 532 node->keys[i] = node->keys[i - 1]; 533 node->ptrs[i] = node->ptrs[i - 1]; 534 } 535 node->keys[pos] = offset; 536 node->ptrs[pos] = ptr; 537 538 if (new) { 539 offset = new->keys[0]; 540 ptr = new; 541 level++; 542 goto again; 543 } 544 } 545 546 static struct xfs_iext_leaf * 547 xfs_iext_split_leaf( 548 struct xfs_iext_cursor *cur, 549 int *nr_entries) 550 { 551 struct xfs_iext_leaf *leaf = cur->leaf; 552 struct xfs_iext_leaf *new = xfs_iext_alloc_node(NODE_SIZE); 553 const int nr_move = RECS_PER_LEAF / 2; 554 int nr_keep = nr_move + (RECS_PER_LEAF & 1); 555 int i; 556 557 /* for sequential append operations just spill over into the new node */ 558 if (cur->pos == RECS_PER_LEAF) { 559 cur->leaf = new; 560 cur->pos = 0; 561 *nr_entries = 0; 562 goto done; 563 } 564 565 for (i = 0; i < nr_move; i++) { 566 new->recs[i] = leaf->recs[nr_keep + i]; 567 xfs_iext_rec_clear(&leaf->recs[nr_keep + i]); 568 } 569 570 if (cur->pos >= nr_keep) { 571 cur->leaf = new; 572 cur->pos -= nr_keep; 573 *nr_entries = nr_move; 574 } else { 575 *nr_entries = nr_keep; 576 } 577 done: 578 if (leaf->next) 579 leaf->next->prev = new; 580 new->next = leaf->next; 581 new->prev = leaf; 582 leaf->next = new; 583 return new; 584 } 585 586 static void 587 xfs_iext_alloc_root( 588 struct xfs_ifork *ifp, 589 struct xfs_iext_cursor *cur) 590 { 591 ASSERT(ifp->if_bytes == 0); 592 593 ifp->if_data = xfs_iext_alloc_node(sizeof(struct xfs_iext_rec)); 594 ifp->if_height = 1; 595 596 /* now that we have a node step into it */ 597 cur->leaf = ifp->if_data; 598 cur->pos = 0; 599 } 600 601 static void 602 xfs_iext_realloc_root( 603 struct xfs_ifork *ifp, 604 struct xfs_iext_cursor *cur) 605 { 606 int64_t new_size = ifp->if_bytes + sizeof(struct xfs_iext_rec); 607 void *new; 608 609 /* account for the prev/next pointers */ 610 if (new_size / sizeof(struct xfs_iext_rec) == RECS_PER_LEAF) 611 new_size = NODE_SIZE; 612 613 new = krealloc(ifp->if_data, new_size, 614 GFP_KERNEL | __GFP_NOLOCKDEP | __GFP_NOFAIL); 615 memset(new + ifp->if_bytes, 0, new_size - ifp->if_bytes); 616 ifp->if_data = new; 617 cur->leaf = new; 618 } 619 620 /* 621 * Increment the sequence counter on extent tree changes. If we are on a COW 622 * fork, this allows the writeback code to skip looking for a COW extent if the 623 * COW fork hasn't changed. We use WRITE_ONCE here to ensure the update to the 624 * sequence counter is seen before the modifications to the extent tree itself 625 * take effect. 626 */ 627 static inline void xfs_iext_inc_seq(struct xfs_ifork *ifp) 628 { 629 WRITE_ONCE(ifp->if_seq, READ_ONCE(ifp->if_seq) + 1); 630 } 631 632 void 633 xfs_iext_insert_raw( 634 struct xfs_ifork *ifp, 635 struct xfs_iext_cursor *cur, 636 struct xfs_bmbt_irec *irec) 637 { 638 xfs_fileoff_t offset = irec->br_startoff; 639 struct xfs_iext_leaf *new = NULL; 640 int nr_entries, i; 641 642 xfs_iext_inc_seq(ifp); 643 644 if (ifp->if_height == 0) 645 xfs_iext_alloc_root(ifp, cur); 646 else if (ifp->if_height == 1) 647 xfs_iext_realloc_root(ifp, cur); 648 649 nr_entries = xfs_iext_leaf_nr_entries(ifp, cur->leaf, cur->pos); 650 ASSERT(nr_entries <= RECS_PER_LEAF); 651 ASSERT(cur->pos >= nr_entries || 652 xfs_iext_rec_cmp(cur_rec(cur), irec->br_startoff) != 0); 653 654 if (nr_entries == RECS_PER_LEAF) 655 new = xfs_iext_split_leaf(cur, &nr_entries); 656 657 /* 658 * Update the pointers in higher levels if the first entry changes 659 * in an existing node. 660 */ 661 if (cur->leaf != new && cur->pos == 0 && nr_entries > 0) { 662 xfs_iext_update_node(ifp, xfs_iext_leaf_key(cur->leaf, 0), 663 offset, 1, cur->leaf); 664 } 665 666 for (i = nr_entries; i > cur->pos; i--) 667 cur->leaf->recs[i] = cur->leaf->recs[i - 1]; 668 xfs_iext_set(cur_rec(cur), irec); 669 ifp->if_bytes += sizeof(struct xfs_iext_rec); 670 671 if (new) 672 xfs_iext_insert_node(ifp, xfs_iext_leaf_key(new, 0), new, 2); 673 } 674 675 void 676 xfs_iext_insert( 677 struct xfs_inode *ip, 678 struct xfs_iext_cursor *cur, 679 struct xfs_bmbt_irec *irec, 680 int state) 681 { 682 struct xfs_ifork *ifp = xfs_iext_state_to_fork(ip, state); 683 684 xfs_iext_insert_raw(ifp, cur, irec); 685 trace_xfs_iext_insert(ip, cur, state, _RET_IP_); 686 } 687 688 static struct xfs_iext_node * 689 xfs_iext_rebalance_node( 690 struct xfs_iext_node *parent, 691 int *pos, 692 struct xfs_iext_node *node, 693 int nr_entries) 694 { 695 /* 696 * If the neighbouring nodes are completely full, or have different 697 * parents, we might never be able to merge our node, and will only 698 * delete it once the number of entries hits zero. 699 */ 700 if (nr_entries == 0) 701 return node; 702 703 if (*pos > 0) { 704 struct xfs_iext_node *prev = parent->ptrs[*pos - 1]; 705 int nr_prev = xfs_iext_node_nr_entries(prev, 0), i; 706 707 if (nr_prev + nr_entries <= KEYS_PER_NODE) { 708 for (i = 0; i < nr_entries; i++) { 709 prev->keys[nr_prev + i] = node->keys[i]; 710 prev->ptrs[nr_prev + i] = node->ptrs[i]; 711 } 712 return node; 713 } 714 } 715 716 if (*pos + 1 < xfs_iext_node_nr_entries(parent, *pos)) { 717 struct xfs_iext_node *next = parent->ptrs[*pos + 1]; 718 int nr_next = xfs_iext_node_nr_entries(next, 0), i; 719 720 if (nr_entries + nr_next <= KEYS_PER_NODE) { 721 /* 722 * Merge the next node into this node so that we don't 723 * have to do an additional update of the keys in the 724 * higher levels. 725 */ 726 for (i = 0; i < nr_next; i++) { 727 node->keys[nr_entries + i] = next->keys[i]; 728 node->ptrs[nr_entries + i] = next->ptrs[i]; 729 } 730 731 ++*pos; 732 return next; 733 } 734 } 735 736 return NULL; 737 } 738 739 static void 740 xfs_iext_remove_node( 741 struct xfs_ifork *ifp, 742 xfs_fileoff_t offset, 743 void *victim) 744 { 745 struct xfs_iext_node *node, *parent; 746 int level = 2, pos, nr_entries, i; 747 748 ASSERT(level <= ifp->if_height); 749 node = xfs_iext_find_level(ifp, offset, level); 750 pos = xfs_iext_node_pos(node, offset); 751 again: 752 ASSERT(node->ptrs[pos]); 753 ASSERT(node->ptrs[pos] == victim); 754 kfree(victim); 755 756 nr_entries = xfs_iext_node_nr_entries(node, pos) - 1; 757 offset = node->keys[0]; 758 for (i = pos; i < nr_entries; i++) { 759 node->keys[i] = node->keys[i + 1]; 760 node->ptrs[i] = node->ptrs[i + 1]; 761 } 762 node->keys[nr_entries] = XFS_IEXT_KEY_INVALID; 763 node->ptrs[nr_entries] = NULL; 764 765 if (pos == 0 && nr_entries > 0) { 766 xfs_iext_update_node(ifp, offset, node->keys[0], level, node); 767 offset = node->keys[0]; 768 } 769 770 if (nr_entries >= KEYS_PER_NODE / 2) 771 return; 772 773 if (level < ifp->if_height) { 774 /* 775 * If we aren't at the root yet try to find a neighbour node to 776 * merge with (or delete the node if it is empty), and then 777 * recurse up to the next level. 778 */ 779 level++; 780 parent = xfs_iext_find_level(ifp, offset, level); 781 pos = xfs_iext_node_pos(parent, offset); 782 783 ASSERT(pos != KEYS_PER_NODE); 784 ASSERT(parent->ptrs[pos] == node); 785 786 node = xfs_iext_rebalance_node(parent, &pos, node, nr_entries); 787 if (node) { 788 victim = node; 789 node = parent; 790 goto again; 791 } 792 } else if (nr_entries == 1) { 793 /* 794 * If we are at the root and only one entry is left we can just 795 * free this node and update the root pointer. 796 */ 797 ASSERT(node == ifp->if_data); 798 ifp->if_data = node->ptrs[0]; 799 ifp->if_height--; 800 kfree(node); 801 } 802 } 803 804 static void 805 xfs_iext_rebalance_leaf( 806 struct xfs_ifork *ifp, 807 struct xfs_iext_cursor *cur, 808 struct xfs_iext_leaf *leaf, 809 xfs_fileoff_t offset, 810 int nr_entries) 811 { 812 /* 813 * If the neighbouring nodes are completely full we might never be able 814 * to merge our node, and will only delete it once the number of 815 * entries hits zero. 816 */ 817 if (nr_entries == 0) 818 goto remove_node; 819 820 if (leaf->prev) { 821 int nr_prev = xfs_iext_leaf_nr_entries(ifp, leaf->prev, 0), i; 822 823 if (nr_prev + nr_entries <= RECS_PER_LEAF) { 824 for (i = 0; i < nr_entries; i++) 825 leaf->prev->recs[nr_prev + i] = leaf->recs[i]; 826 827 if (cur->leaf == leaf) { 828 cur->leaf = leaf->prev; 829 cur->pos += nr_prev; 830 } 831 goto remove_node; 832 } 833 } 834 835 if (leaf->next) { 836 int nr_next = xfs_iext_leaf_nr_entries(ifp, leaf->next, 0), i; 837 838 if (nr_entries + nr_next <= RECS_PER_LEAF) { 839 /* 840 * Merge the next node into this node so that we don't 841 * have to do an additional update of the keys in the 842 * higher levels. 843 */ 844 for (i = 0; i < nr_next; i++) { 845 leaf->recs[nr_entries + i] = 846 leaf->next->recs[i]; 847 } 848 849 if (cur->leaf == leaf->next) { 850 cur->leaf = leaf; 851 cur->pos += nr_entries; 852 } 853 854 offset = xfs_iext_leaf_key(leaf->next, 0); 855 leaf = leaf->next; 856 goto remove_node; 857 } 858 } 859 860 return; 861 remove_node: 862 if (leaf->prev) 863 leaf->prev->next = leaf->next; 864 if (leaf->next) 865 leaf->next->prev = leaf->prev; 866 xfs_iext_remove_node(ifp, offset, leaf); 867 } 868 869 static void 870 xfs_iext_free_last_leaf( 871 struct xfs_ifork *ifp) 872 { 873 ifp->if_height--; 874 kfree(ifp->if_data); 875 ifp->if_data = NULL; 876 } 877 878 void 879 xfs_iext_remove( 880 struct xfs_inode *ip, 881 struct xfs_iext_cursor *cur, 882 int state) 883 { 884 struct xfs_ifork *ifp = xfs_iext_state_to_fork(ip, state); 885 struct xfs_iext_leaf *leaf = cur->leaf; 886 xfs_fileoff_t offset = xfs_iext_leaf_key(leaf, 0); 887 int i, nr_entries; 888 889 trace_xfs_iext_remove(ip, cur, state, _RET_IP_); 890 891 ASSERT(ifp->if_height > 0); 892 ASSERT(ifp->if_data != NULL); 893 ASSERT(xfs_iext_valid(ifp, cur)); 894 895 xfs_iext_inc_seq(ifp); 896 897 nr_entries = xfs_iext_leaf_nr_entries(ifp, leaf, cur->pos) - 1; 898 for (i = cur->pos; i < nr_entries; i++) 899 leaf->recs[i] = leaf->recs[i + 1]; 900 xfs_iext_rec_clear(&leaf->recs[nr_entries]); 901 ifp->if_bytes -= sizeof(struct xfs_iext_rec); 902 903 if (cur->pos == 0 && nr_entries > 0) { 904 xfs_iext_update_node(ifp, offset, xfs_iext_leaf_key(leaf, 0), 1, 905 leaf); 906 offset = xfs_iext_leaf_key(leaf, 0); 907 } else if (cur->pos == nr_entries) { 908 if (ifp->if_height > 1 && leaf->next) 909 cur->leaf = leaf->next; 910 else 911 cur->leaf = NULL; 912 cur->pos = 0; 913 } 914 915 if (nr_entries >= RECS_PER_LEAF / 2) 916 return; 917 918 if (ifp->if_height > 1) 919 xfs_iext_rebalance_leaf(ifp, cur, leaf, offset, nr_entries); 920 else if (nr_entries == 0) 921 xfs_iext_free_last_leaf(ifp); 922 } 923 924 /* 925 * Lookup the extent covering bno. 926 * 927 * If there is an extent covering bno return the extent index, and store the 928 * expanded extent structure in *gotp, and the extent cursor in *cur. 929 * If there is no extent covering bno, but there is an extent after it (e.g. 930 * it lies in a hole) return that extent in *gotp and its cursor in *cur 931 * instead. 932 * If bno is beyond the last extent return false, and return an invalid 933 * cursor value. 934 */ 935 bool 936 xfs_iext_lookup_extent( 937 struct xfs_inode *ip, 938 struct xfs_ifork *ifp, 939 xfs_fileoff_t offset, 940 struct xfs_iext_cursor *cur, 941 struct xfs_bmbt_irec *gotp) 942 { 943 XFS_STATS_INC(ip->i_mount, xs_look_exlist); 944 945 cur->leaf = xfs_iext_find_level(ifp, offset, 1); 946 if (!cur->leaf) { 947 cur->pos = 0; 948 return false; 949 } 950 951 for (cur->pos = 0; cur->pos < xfs_iext_max_recs(ifp); cur->pos++) { 952 struct xfs_iext_rec *rec = cur_rec(cur); 953 954 if (xfs_iext_rec_is_empty(rec)) 955 break; 956 if (xfs_iext_rec_cmp(rec, offset) >= 0) 957 goto found; 958 } 959 960 /* Try looking in the next node for an entry > offset */ 961 if (ifp->if_height == 1 || !cur->leaf->next) 962 return false; 963 cur->leaf = cur->leaf->next; 964 cur->pos = 0; 965 if (!xfs_iext_valid(ifp, cur)) 966 return false; 967 found: 968 xfs_iext_get(gotp, cur_rec(cur)); 969 return true; 970 } 971 972 /* 973 * Returns the last extent before end, and if this extent doesn't cover 974 * end, update end to the end of the extent. 975 */ 976 bool 977 xfs_iext_lookup_extent_before( 978 struct xfs_inode *ip, 979 struct xfs_ifork *ifp, 980 xfs_fileoff_t *end, 981 struct xfs_iext_cursor *cur, 982 struct xfs_bmbt_irec *gotp) 983 { 984 /* could be optimized to not even look up the next on a match.. */ 985 if (xfs_iext_lookup_extent(ip, ifp, *end - 1, cur, gotp) && 986 gotp->br_startoff <= *end - 1) 987 return true; 988 if (!xfs_iext_prev_extent(ifp, cur, gotp)) 989 return false; 990 *end = gotp->br_startoff + gotp->br_blockcount; 991 return true; 992 } 993 994 void 995 xfs_iext_update_extent( 996 struct xfs_inode *ip, 997 int state, 998 struct xfs_iext_cursor *cur, 999 struct xfs_bmbt_irec *new) 1000 { 1001 struct xfs_ifork *ifp = xfs_iext_state_to_fork(ip, state); 1002 1003 xfs_iext_inc_seq(ifp); 1004 1005 if (cur->pos == 0) { 1006 struct xfs_bmbt_irec old; 1007 1008 xfs_iext_get(&old, cur_rec(cur)); 1009 if (new->br_startoff != old.br_startoff) { 1010 xfs_iext_update_node(ifp, old.br_startoff, 1011 new->br_startoff, 1, cur->leaf); 1012 } 1013 } 1014 1015 trace_xfs_bmap_pre_update(ip, cur, state, _RET_IP_); 1016 xfs_iext_set(cur_rec(cur), new); 1017 trace_xfs_bmap_post_update(ip, cur, state, _RET_IP_); 1018 } 1019 1020 /* 1021 * Return true if the cursor points at an extent and return the extent structure 1022 * in gotp. Else return false. 1023 */ 1024 bool 1025 xfs_iext_get_extent( 1026 struct xfs_ifork *ifp, 1027 struct xfs_iext_cursor *cur, 1028 struct xfs_bmbt_irec *gotp) 1029 { 1030 if (!xfs_iext_valid(ifp, cur)) 1031 return false; 1032 xfs_iext_get(gotp, cur_rec(cur)); 1033 return true; 1034 } 1035 1036 /* 1037 * This is a recursive function, because of that we need to be extremely 1038 * careful with stack usage. 1039 */ 1040 static void 1041 xfs_iext_destroy_node( 1042 struct xfs_iext_node *node, 1043 int level) 1044 { 1045 int i; 1046 1047 if (level > 1) { 1048 for (i = 0; i < KEYS_PER_NODE; i++) { 1049 if (node->keys[i] == XFS_IEXT_KEY_INVALID) 1050 break; 1051 xfs_iext_destroy_node(node->ptrs[i], level - 1); 1052 } 1053 } 1054 1055 kfree(node); 1056 } 1057 1058 void 1059 xfs_iext_destroy( 1060 struct xfs_ifork *ifp) 1061 { 1062 xfs_iext_destroy_node(ifp->if_data, ifp->if_height); 1063 1064 ifp->if_bytes = 0; 1065 ifp->if_height = 0; 1066 ifp->if_data = NULL; 1067 } 1068