1 /* -*- mode: c; c-basic-offset: 8; -*- 2 * vim: noexpandtab sw=8 ts=8 sts=0: 3 * 4 * alloc.c 5 * 6 * Extent allocs and frees 7 * 8 * Copyright (C) 2002, 2004 Oracle. All rights reserved. 9 * 10 * This program is free software; you can redistribute it and/or 11 * modify it under the terms of the GNU General Public 12 * License as published by the Free Software Foundation; either 13 * version 2 of the License, or (at your option) any later version. 14 * 15 * This program is distributed in the hope that it will be useful, 16 * but WITHOUT ANY WARRANTY; without even the implied warranty of 17 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU 18 * General Public License for more details. 19 * 20 * You should have received a copy of the GNU General Public 21 * License along with this program; if not, write to the 22 * Free Software Foundation, Inc., 59 Temple Place - Suite 330, 23 * Boston, MA 021110-1307, USA. 24 */ 25 26 #include <linux/fs.h> 27 #include <linux/types.h> 28 #include <linux/slab.h> 29 #include <linux/highmem.h> 30 #include <linux/swap.h> 31 32 #define MLOG_MASK_PREFIX ML_DISK_ALLOC 33 #include <cluster/masklog.h> 34 35 #include "ocfs2.h" 36 37 #include "alloc.h" 38 #include "aops.h" 39 #include "dlmglue.h" 40 #include "extent_map.h" 41 #include "inode.h" 42 #include "journal.h" 43 #include "localalloc.h" 44 #include "suballoc.h" 45 #include "sysfile.h" 46 #include "file.h" 47 #include "super.h" 48 #include "uptodate.h" 49 50 #include "buffer_head_io.h" 51 52 static void ocfs2_free_truncate_context(struct ocfs2_truncate_context *tc); 53 static int ocfs2_cache_extent_block_free(struct ocfs2_cached_dealloc_ctxt *ctxt, 54 struct ocfs2_extent_block *eb); 55 56 /* 57 * Structures which describe a path through a btree, and functions to 58 * manipulate them. 59 * 60 * The idea here is to be as generic as possible with the tree 61 * manipulation code. 62 */ 63 struct ocfs2_path_item { 64 struct buffer_head *bh; 65 struct ocfs2_extent_list *el; 66 }; 67 68 #define OCFS2_MAX_PATH_DEPTH 5 69 70 struct ocfs2_path { 71 int p_tree_depth; 72 struct ocfs2_path_item p_node[OCFS2_MAX_PATH_DEPTH]; 73 }; 74 75 #define path_root_bh(_path) ((_path)->p_node[0].bh) 76 #define path_root_el(_path) ((_path)->p_node[0].el) 77 #define path_leaf_bh(_path) ((_path)->p_node[(_path)->p_tree_depth].bh) 78 #define path_leaf_el(_path) ((_path)->p_node[(_path)->p_tree_depth].el) 79 #define path_num_items(_path) ((_path)->p_tree_depth + 1) 80 81 /* 82 * Reset the actual path elements so that we can re-use the structure 83 * to build another path. Generally, this involves freeing the buffer 84 * heads. 85 */ 86 static void ocfs2_reinit_path(struct ocfs2_path *path, int keep_root) 87 { 88 int i, start = 0, depth = 0; 89 struct ocfs2_path_item *node; 90 91 if (keep_root) 92 start = 1; 93 94 for(i = start; i < path_num_items(path); i++) { 95 node = &path->p_node[i]; 96 97 brelse(node->bh); 98 node->bh = NULL; 99 node->el = NULL; 100 } 101 102 /* 103 * Tree depth may change during truncate, or insert. If we're 104 * keeping the root extent list, then make sure that our path 105 * structure reflects the proper depth. 106 */ 107 if (keep_root) 108 depth = le16_to_cpu(path_root_el(path)->l_tree_depth); 109 110 path->p_tree_depth = depth; 111 } 112 113 static void ocfs2_free_path(struct ocfs2_path *path) 114 { 115 if (path) { 116 ocfs2_reinit_path(path, 0); 117 kfree(path); 118 } 119 } 120 121 /* 122 * All the elements of src into dest. After this call, src could be freed 123 * without affecting dest. 124 * 125 * Both paths should have the same root. Any non-root elements of dest 126 * will be freed. 127 */ 128 static void ocfs2_cp_path(struct ocfs2_path *dest, struct ocfs2_path *src) 129 { 130 int i; 131 132 BUG_ON(path_root_bh(dest) != path_root_bh(src)); 133 BUG_ON(path_root_el(dest) != path_root_el(src)); 134 135 ocfs2_reinit_path(dest, 1); 136 137 for(i = 1; i < OCFS2_MAX_PATH_DEPTH; i++) { 138 dest->p_node[i].bh = src->p_node[i].bh; 139 dest->p_node[i].el = src->p_node[i].el; 140 141 if (dest->p_node[i].bh) 142 get_bh(dest->p_node[i].bh); 143 } 144 } 145 146 /* 147 * Make the *dest path the same as src and re-initialize src path to 148 * have a root only. 149 */ 150 static void ocfs2_mv_path(struct ocfs2_path *dest, struct ocfs2_path *src) 151 { 152 int i; 153 154 BUG_ON(path_root_bh(dest) != path_root_bh(src)); 155 156 for(i = 1; i < OCFS2_MAX_PATH_DEPTH; i++) { 157 brelse(dest->p_node[i].bh); 158 159 dest->p_node[i].bh = src->p_node[i].bh; 160 dest->p_node[i].el = src->p_node[i].el; 161 162 src->p_node[i].bh = NULL; 163 src->p_node[i].el = NULL; 164 } 165 } 166 167 /* 168 * Insert an extent block at given index. 169 * 170 * This will not take an additional reference on eb_bh. 171 */ 172 static inline void ocfs2_path_insert_eb(struct ocfs2_path *path, int index, 173 struct buffer_head *eb_bh) 174 { 175 struct ocfs2_extent_block *eb = (struct ocfs2_extent_block *)eb_bh->b_data; 176 177 /* 178 * Right now, no root bh is an extent block, so this helps 179 * catch code errors with dinode trees. The assertion can be 180 * safely removed if we ever need to insert extent block 181 * structures at the root. 182 */ 183 BUG_ON(index == 0); 184 185 path->p_node[index].bh = eb_bh; 186 path->p_node[index].el = &eb->h_list; 187 } 188 189 static struct ocfs2_path *ocfs2_new_path(struct buffer_head *root_bh, 190 struct ocfs2_extent_list *root_el) 191 { 192 struct ocfs2_path *path; 193 194 BUG_ON(le16_to_cpu(root_el->l_tree_depth) >= OCFS2_MAX_PATH_DEPTH); 195 196 path = kzalloc(sizeof(*path), GFP_NOFS); 197 if (path) { 198 path->p_tree_depth = le16_to_cpu(root_el->l_tree_depth); 199 get_bh(root_bh); 200 path_root_bh(path) = root_bh; 201 path_root_el(path) = root_el; 202 } 203 204 return path; 205 } 206 207 /* 208 * Allocate and initialize a new path based on a disk inode tree. 209 */ 210 static struct ocfs2_path *ocfs2_new_inode_path(struct buffer_head *di_bh) 211 { 212 struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data; 213 struct ocfs2_extent_list *el = &di->id2.i_list; 214 215 return ocfs2_new_path(di_bh, el); 216 } 217 218 /* 219 * Convenience function to journal all components in a path. 220 */ 221 static int ocfs2_journal_access_path(struct inode *inode, handle_t *handle, 222 struct ocfs2_path *path) 223 { 224 int i, ret = 0; 225 226 if (!path) 227 goto out; 228 229 for(i = 0; i < path_num_items(path); i++) { 230 ret = ocfs2_journal_access(handle, inode, path->p_node[i].bh, 231 OCFS2_JOURNAL_ACCESS_WRITE); 232 if (ret < 0) { 233 mlog_errno(ret); 234 goto out; 235 } 236 } 237 238 out: 239 return ret; 240 } 241 242 /* 243 * Return the index of the extent record which contains cluster #v_cluster. 244 * -1 is returned if it was not found. 245 * 246 * Should work fine on interior and exterior nodes. 247 */ 248 int ocfs2_search_extent_list(struct ocfs2_extent_list *el, u32 v_cluster) 249 { 250 int ret = -1; 251 int i; 252 struct ocfs2_extent_rec *rec; 253 u32 rec_end, rec_start, clusters; 254 255 for(i = 0; i < le16_to_cpu(el->l_next_free_rec); i++) { 256 rec = &el->l_recs[i]; 257 258 rec_start = le32_to_cpu(rec->e_cpos); 259 clusters = ocfs2_rec_clusters(el, rec); 260 261 rec_end = rec_start + clusters; 262 263 if (v_cluster >= rec_start && v_cluster < rec_end) { 264 ret = i; 265 break; 266 } 267 } 268 269 return ret; 270 } 271 272 enum ocfs2_contig_type { 273 CONTIG_NONE = 0, 274 CONTIG_LEFT, 275 CONTIG_RIGHT, 276 CONTIG_LEFTRIGHT, 277 }; 278 279 280 /* 281 * NOTE: ocfs2_block_extent_contig(), ocfs2_extents_adjacent() and 282 * ocfs2_extent_contig only work properly against leaf nodes! 283 */ 284 static int ocfs2_block_extent_contig(struct super_block *sb, 285 struct ocfs2_extent_rec *ext, 286 u64 blkno) 287 { 288 u64 blk_end = le64_to_cpu(ext->e_blkno); 289 290 blk_end += ocfs2_clusters_to_blocks(sb, 291 le16_to_cpu(ext->e_leaf_clusters)); 292 293 return blkno == blk_end; 294 } 295 296 static int ocfs2_extents_adjacent(struct ocfs2_extent_rec *left, 297 struct ocfs2_extent_rec *right) 298 { 299 u32 left_range; 300 301 left_range = le32_to_cpu(left->e_cpos) + 302 le16_to_cpu(left->e_leaf_clusters); 303 304 return (left_range == le32_to_cpu(right->e_cpos)); 305 } 306 307 static enum ocfs2_contig_type 308 ocfs2_extent_contig(struct inode *inode, 309 struct ocfs2_extent_rec *ext, 310 struct ocfs2_extent_rec *insert_rec) 311 { 312 u64 blkno = le64_to_cpu(insert_rec->e_blkno); 313 314 /* 315 * Refuse to coalesce extent records with different flag 316 * fields - we don't want to mix unwritten extents with user 317 * data. 318 */ 319 if (ext->e_flags != insert_rec->e_flags) 320 return CONTIG_NONE; 321 322 if (ocfs2_extents_adjacent(ext, insert_rec) && 323 ocfs2_block_extent_contig(inode->i_sb, ext, blkno)) 324 return CONTIG_RIGHT; 325 326 blkno = le64_to_cpu(ext->e_blkno); 327 if (ocfs2_extents_adjacent(insert_rec, ext) && 328 ocfs2_block_extent_contig(inode->i_sb, insert_rec, blkno)) 329 return CONTIG_LEFT; 330 331 return CONTIG_NONE; 332 } 333 334 /* 335 * NOTE: We can have pretty much any combination of contiguousness and 336 * appending. 337 * 338 * The usefulness of APPEND_TAIL is more in that it lets us know that 339 * we'll have to update the path to that leaf. 340 */ 341 enum ocfs2_append_type { 342 APPEND_NONE = 0, 343 APPEND_TAIL, 344 }; 345 346 enum ocfs2_split_type { 347 SPLIT_NONE = 0, 348 SPLIT_LEFT, 349 SPLIT_RIGHT, 350 }; 351 352 struct ocfs2_insert_type { 353 enum ocfs2_split_type ins_split; 354 enum ocfs2_append_type ins_appending; 355 enum ocfs2_contig_type ins_contig; 356 int ins_contig_index; 357 int ins_tree_depth; 358 }; 359 360 struct ocfs2_merge_ctxt { 361 enum ocfs2_contig_type c_contig_type; 362 int c_has_empty_extent; 363 int c_split_covers_rec; 364 }; 365 366 /* 367 * How many free extents have we got before we need more meta data? 368 */ 369 int ocfs2_num_free_extents(struct ocfs2_super *osb, 370 struct inode *inode, 371 struct buffer_head *bh) 372 { 373 int retval; 374 struct ocfs2_extent_list *el; 375 struct ocfs2_extent_block *eb; 376 struct buffer_head *eb_bh = NULL; 377 struct ocfs2_dinode *fe = (struct ocfs2_dinode *)bh->b_data; 378 379 mlog_entry_void(); 380 381 if (!OCFS2_IS_VALID_DINODE(fe)) { 382 OCFS2_RO_ON_INVALID_DINODE(inode->i_sb, fe); 383 retval = -EIO; 384 goto bail; 385 } 386 387 if (fe->i_last_eb_blk) { 388 retval = ocfs2_read_block(osb, le64_to_cpu(fe->i_last_eb_blk), 389 &eb_bh, OCFS2_BH_CACHED, inode); 390 if (retval < 0) { 391 mlog_errno(retval); 392 goto bail; 393 } 394 eb = (struct ocfs2_extent_block *) eb_bh->b_data; 395 el = &eb->h_list; 396 } else 397 el = &fe->id2.i_list; 398 399 BUG_ON(el->l_tree_depth != 0); 400 401 retval = le16_to_cpu(el->l_count) - le16_to_cpu(el->l_next_free_rec); 402 bail: 403 if (eb_bh) 404 brelse(eb_bh); 405 406 mlog_exit(retval); 407 return retval; 408 } 409 410 /* expects array to already be allocated 411 * 412 * sets h_signature, h_blkno, h_suballoc_bit, h_suballoc_slot, and 413 * l_count for you 414 */ 415 static int ocfs2_create_new_meta_bhs(struct ocfs2_super *osb, 416 handle_t *handle, 417 struct inode *inode, 418 int wanted, 419 struct ocfs2_alloc_context *meta_ac, 420 struct buffer_head *bhs[]) 421 { 422 int count, status, i; 423 u16 suballoc_bit_start; 424 u32 num_got; 425 u64 first_blkno; 426 struct ocfs2_extent_block *eb; 427 428 mlog_entry_void(); 429 430 count = 0; 431 while (count < wanted) { 432 status = ocfs2_claim_metadata(osb, 433 handle, 434 meta_ac, 435 wanted - count, 436 &suballoc_bit_start, 437 &num_got, 438 &first_blkno); 439 if (status < 0) { 440 mlog_errno(status); 441 goto bail; 442 } 443 444 for(i = count; i < (num_got + count); i++) { 445 bhs[i] = sb_getblk(osb->sb, first_blkno); 446 if (bhs[i] == NULL) { 447 status = -EIO; 448 mlog_errno(status); 449 goto bail; 450 } 451 ocfs2_set_new_buffer_uptodate(inode, bhs[i]); 452 453 status = ocfs2_journal_access(handle, inode, bhs[i], 454 OCFS2_JOURNAL_ACCESS_CREATE); 455 if (status < 0) { 456 mlog_errno(status); 457 goto bail; 458 } 459 460 memset(bhs[i]->b_data, 0, osb->sb->s_blocksize); 461 eb = (struct ocfs2_extent_block *) bhs[i]->b_data; 462 /* Ok, setup the minimal stuff here. */ 463 strcpy(eb->h_signature, OCFS2_EXTENT_BLOCK_SIGNATURE); 464 eb->h_blkno = cpu_to_le64(first_blkno); 465 eb->h_fs_generation = cpu_to_le32(osb->fs_generation); 466 eb->h_suballoc_slot = cpu_to_le16(osb->slot_num); 467 eb->h_suballoc_bit = cpu_to_le16(suballoc_bit_start); 468 eb->h_list.l_count = 469 cpu_to_le16(ocfs2_extent_recs_per_eb(osb->sb)); 470 471 suballoc_bit_start++; 472 first_blkno++; 473 474 /* We'll also be dirtied by the caller, so 475 * this isn't absolutely necessary. */ 476 status = ocfs2_journal_dirty(handle, bhs[i]); 477 if (status < 0) { 478 mlog_errno(status); 479 goto bail; 480 } 481 } 482 483 count += num_got; 484 } 485 486 status = 0; 487 bail: 488 if (status < 0) { 489 for(i = 0; i < wanted; i++) { 490 if (bhs[i]) 491 brelse(bhs[i]); 492 bhs[i] = NULL; 493 } 494 } 495 mlog_exit(status); 496 return status; 497 } 498 499 /* 500 * Helper function for ocfs2_add_branch() and ocfs2_shift_tree_depth(). 501 * 502 * Returns the sum of the rightmost extent rec logical offset and 503 * cluster count. 504 * 505 * ocfs2_add_branch() uses this to determine what logical cluster 506 * value should be populated into the leftmost new branch records. 507 * 508 * ocfs2_shift_tree_depth() uses this to determine the # clusters 509 * value for the new topmost tree record. 510 */ 511 static inline u32 ocfs2_sum_rightmost_rec(struct ocfs2_extent_list *el) 512 { 513 int i; 514 515 i = le16_to_cpu(el->l_next_free_rec) - 1; 516 517 return le32_to_cpu(el->l_recs[i].e_cpos) + 518 ocfs2_rec_clusters(el, &el->l_recs[i]); 519 } 520 521 /* 522 * Add an entire tree branch to our inode. eb_bh is the extent block 523 * to start at, if we don't want to start the branch at the dinode 524 * structure. 525 * 526 * last_eb_bh is required as we have to update it's next_leaf pointer 527 * for the new last extent block. 528 * 529 * the new branch will be 'empty' in the sense that every block will 530 * contain a single record with cluster count == 0. 531 */ 532 static int ocfs2_add_branch(struct ocfs2_super *osb, 533 handle_t *handle, 534 struct inode *inode, 535 struct buffer_head *fe_bh, 536 struct buffer_head *eb_bh, 537 struct buffer_head **last_eb_bh, 538 struct ocfs2_alloc_context *meta_ac) 539 { 540 int status, new_blocks, i; 541 u64 next_blkno, new_last_eb_blk; 542 struct buffer_head *bh; 543 struct buffer_head **new_eb_bhs = NULL; 544 struct ocfs2_dinode *fe; 545 struct ocfs2_extent_block *eb; 546 struct ocfs2_extent_list *eb_el; 547 struct ocfs2_extent_list *el; 548 u32 new_cpos; 549 550 mlog_entry_void(); 551 552 BUG_ON(!last_eb_bh || !*last_eb_bh); 553 554 fe = (struct ocfs2_dinode *) fe_bh->b_data; 555 556 if (eb_bh) { 557 eb = (struct ocfs2_extent_block *) eb_bh->b_data; 558 el = &eb->h_list; 559 } else 560 el = &fe->id2.i_list; 561 562 /* we never add a branch to a leaf. */ 563 BUG_ON(!el->l_tree_depth); 564 565 new_blocks = le16_to_cpu(el->l_tree_depth); 566 567 /* allocate the number of new eb blocks we need */ 568 new_eb_bhs = kcalloc(new_blocks, sizeof(struct buffer_head *), 569 GFP_KERNEL); 570 if (!new_eb_bhs) { 571 status = -ENOMEM; 572 mlog_errno(status); 573 goto bail; 574 } 575 576 status = ocfs2_create_new_meta_bhs(osb, handle, inode, new_blocks, 577 meta_ac, new_eb_bhs); 578 if (status < 0) { 579 mlog_errno(status); 580 goto bail; 581 } 582 583 eb = (struct ocfs2_extent_block *)(*last_eb_bh)->b_data; 584 new_cpos = ocfs2_sum_rightmost_rec(&eb->h_list); 585 586 /* Note: new_eb_bhs[new_blocks - 1] is the guy which will be 587 * linked with the rest of the tree. 588 * conversly, new_eb_bhs[0] is the new bottommost leaf. 589 * 590 * when we leave the loop, new_last_eb_blk will point to the 591 * newest leaf, and next_blkno will point to the topmost extent 592 * block. */ 593 next_blkno = new_last_eb_blk = 0; 594 for(i = 0; i < new_blocks; i++) { 595 bh = new_eb_bhs[i]; 596 eb = (struct ocfs2_extent_block *) bh->b_data; 597 if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) { 598 OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb); 599 status = -EIO; 600 goto bail; 601 } 602 eb_el = &eb->h_list; 603 604 status = ocfs2_journal_access(handle, inode, bh, 605 OCFS2_JOURNAL_ACCESS_CREATE); 606 if (status < 0) { 607 mlog_errno(status); 608 goto bail; 609 } 610 611 eb->h_next_leaf_blk = 0; 612 eb_el->l_tree_depth = cpu_to_le16(i); 613 eb_el->l_next_free_rec = cpu_to_le16(1); 614 /* 615 * This actually counts as an empty extent as 616 * c_clusters == 0 617 */ 618 eb_el->l_recs[0].e_cpos = cpu_to_le32(new_cpos); 619 eb_el->l_recs[0].e_blkno = cpu_to_le64(next_blkno); 620 /* 621 * eb_el isn't always an interior node, but even leaf 622 * nodes want a zero'd flags and reserved field so 623 * this gets the whole 32 bits regardless of use. 624 */ 625 eb_el->l_recs[0].e_int_clusters = cpu_to_le32(0); 626 if (!eb_el->l_tree_depth) 627 new_last_eb_blk = le64_to_cpu(eb->h_blkno); 628 629 status = ocfs2_journal_dirty(handle, bh); 630 if (status < 0) { 631 mlog_errno(status); 632 goto bail; 633 } 634 635 next_blkno = le64_to_cpu(eb->h_blkno); 636 } 637 638 /* This is a bit hairy. We want to update up to three blocks 639 * here without leaving any of them in an inconsistent state 640 * in case of error. We don't have to worry about 641 * journal_dirty erroring as it won't unless we've aborted the 642 * handle (in which case we would never be here) so reserving 643 * the write with journal_access is all we need to do. */ 644 status = ocfs2_journal_access(handle, inode, *last_eb_bh, 645 OCFS2_JOURNAL_ACCESS_WRITE); 646 if (status < 0) { 647 mlog_errno(status); 648 goto bail; 649 } 650 status = ocfs2_journal_access(handle, inode, fe_bh, 651 OCFS2_JOURNAL_ACCESS_WRITE); 652 if (status < 0) { 653 mlog_errno(status); 654 goto bail; 655 } 656 if (eb_bh) { 657 status = ocfs2_journal_access(handle, inode, eb_bh, 658 OCFS2_JOURNAL_ACCESS_WRITE); 659 if (status < 0) { 660 mlog_errno(status); 661 goto bail; 662 } 663 } 664 665 /* Link the new branch into the rest of the tree (el will 666 * either be on the fe, or the extent block passed in. */ 667 i = le16_to_cpu(el->l_next_free_rec); 668 el->l_recs[i].e_blkno = cpu_to_le64(next_blkno); 669 el->l_recs[i].e_cpos = cpu_to_le32(new_cpos); 670 el->l_recs[i].e_int_clusters = 0; 671 le16_add_cpu(&el->l_next_free_rec, 1); 672 673 /* fe needs a new last extent block pointer, as does the 674 * next_leaf on the previously last-extent-block. */ 675 fe->i_last_eb_blk = cpu_to_le64(new_last_eb_blk); 676 677 eb = (struct ocfs2_extent_block *) (*last_eb_bh)->b_data; 678 eb->h_next_leaf_blk = cpu_to_le64(new_last_eb_blk); 679 680 status = ocfs2_journal_dirty(handle, *last_eb_bh); 681 if (status < 0) 682 mlog_errno(status); 683 status = ocfs2_journal_dirty(handle, fe_bh); 684 if (status < 0) 685 mlog_errno(status); 686 if (eb_bh) { 687 status = ocfs2_journal_dirty(handle, eb_bh); 688 if (status < 0) 689 mlog_errno(status); 690 } 691 692 /* 693 * Some callers want to track the rightmost leaf so pass it 694 * back here. 695 */ 696 brelse(*last_eb_bh); 697 get_bh(new_eb_bhs[0]); 698 *last_eb_bh = new_eb_bhs[0]; 699 700 status = 0; 701 bail: 702 if (new_eb_bhs) { 703 for (i = 0; i < new_blocks; i++) 704 if (new_eb_bhs[i]) 705 brelse(new_eb_bhs[i]); 706 kfree(new_eb_bhs); 707 } 708 709 mlog_exit(status); 710 return status; 711 } 712 713 /* 714 * adds another level to the allocation tree. 715 * returns back the new extent block so you can add a branch to it 716 * after this call. 717 */ 718 static int ocfs2_shift_tree_depth(struct ocfs2_super *osb, 719 handle_t *handle, 720 struct inode *inode, 721 struct buffer_head *fe_bh, 722 struct ocfs2_alloc_context *meta_ac, 723 struct buffer_head **ret_new_eb_bh) 724 { 725 int status, i; 726 u32 new_clusters; 727 struct buffer_head *new_eb_bh = NULL; 728 struct ocfs2_dinode *fe; 729 struct ocfs2_extent_block *eb; 730 struct ocfs2_extent_list *fe_el; 731 struct ocfs2_extent_list *eb_el; 732 733 mlog_entry_void(); 734 735 status = ocfs2_create_new_meta_bhs(osb, handle, inode, 1, meta_ac, 736 &new_eb_bh); 737 if (status < 0) { 738 mlog_errno(status); 739 goto bail; 740 } 741 742 eb = (struct ocfs2_extent_block *) new_eb_bh->b_data; 743 if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) { 744 OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb); 745 status = -EIO; 746 goto bail; 747 } 748 749 eb_el = &eb->h_list; 750 fe = (struct ocfs2_dinode *) fe_bh->b_data; 751 fe_el = &fe->id2.i_list; 752 753 status = ocfs2_journal_access(handle, inode, new_eb_bh, 754 OCFS2_JOURNAL_ACCESS_CREATE); 755 if (status < 0) { 756 mlog_errno(status); 757 goto bail; 758 } 759 760 /* copy the fe data into the new extent block */ 761 eb_el->l_tree_depth = fe_el->l_tree_depth; 762 eb_el->l_next_free_rec = fe_el->l_next_free_rec; 763 for(i = 0; i < le16_to_cpu(fe_el->l_next_free_rec); i++) 764 eb_el->l_recs[i] = fe_el->l_recs[i]; 765 766 status = ocfs2_journal_dirty(handle, new_eb_bh); 767 if (status < 0) { 768 mlog_errno(status); 769 goto bail; 770 } 771 772 status = ocfs2_journal_access(handle, inode, fe_bh, 773 OCFS2_JOURNAL_ACCESS_WRITE); 774 if (status < 0) { 775 mlog_errno(status); 776 goto bail; 777 } 778 779 new_clusters = ocfs2_sum_rightmost_rec(eb_el); 780 781 /* update fe now */ 782 le16_add_cpu(&fe_el->l_tree_depth, 1); 783 fe_el->l_recs[0].e_cpos = 0; 784 fe_el->l_recs[0].e_blkno = eb->h_blkno; 785 fe_el->l_recs[0].e_int_clusters = cpu_to_le32(new_clusters); 786 for(i = 1; i < le16_to_cpu(fe_el->l_next_free_rec); i++) 787 memset(&fe_el->l_recs[i], 0, sizeof(struct ocfs2_extent_rec)); 788 fe_el->l_next_free_rec = cpu_to_le16(1); 789 790 /* If this is our 1st tree depth shift, then last_eb_blk 791 * becomes the allocated extent block */ 792 if (fe_el->l_tree_depth == cpu_to_le16(1)) 793 fe->i_last_eb_blk = eb->h_blkno; 794 795 status = ocfs2_journal_dirty(handle, fe_bh); 796 if (status < 0) { 797 mlog_errno(status); 798 goto bail; 799 } 800 801 *ret_new_eb_bh = new_eb_bh; 802 new_eb_bh = NULL; 803 status = 0; 804 bail: 805 if (new_eb_bh) 806 brelse(new_eb_bh); 807 808 mlog_exit(status); 809 return status; 810 } 811 812 /* 813 * Should only be called when there is no space left in any of the 814 * leaf nodes. What we want to do is find the lowest tree depth 815 * non-leaf extent block with room for new records. There are three 816 * valid results of this search: 817 * 818 * 1) a lowest extent block is found, then we pass it back in 819 * *lowest_eb_bh and return '0' 820 * 821 * 2) the search fails to find anything, but the dinode has room. We 822 * pass NULL back in *lowest_eb_bh, but still return '0' 823 * 824 * 3) the search fails to find anything AND the dinode is full, in 825 * which case we return > 0 826 * 827 * return status < 0 indicates an error. 828 */ 829 static int ocfs2_find_branch_target(struct ocfs2_super *osb, 830 struct inode *inode, 831 struct buffer_head *fe_bh, 832 struct buffer_head **target_bh) 833 { 834 int status = 0, i; 835 u64 blkno; 836 struct ocfs2_dinode *fe; 837 struct ocfs2_extent_block *eb; 838 struct ocfs2_extent_list *el; 839 struct buffer_head *bh = NULL; 840 struct buffer_head *lowest_bh = NULL; 841 842 mlog_entry_void(); 843 844 *target_bh = NULL; 845 846 fe = (struct ocfs2_dinode *) fe_bh->b_data; 847 el = &fe->id2.i_list; 848 849 while(le16_to_cpu(el->l_tree_depth) > 1) { 850 if (le16_to_cpu(el->l_next_free_rec) == 0) { 851 ocfs2_error(inode->i_sb, "Dinode %llu has empty " 852 "extent list (next_free_rec == 0)", 853 (unsigned long long)OCFS2_I(inode)->ip_blkno); 854 status = -EIO; 855 goto bail; 856 } 857 i = le16_to_cpu(el->l_next_free_rec) - 1; 858 blkno = le64_to_cpu(el->l_recs[i].e_blkno); 859 if (!blkno) { 860 ocfs2_error(inode->i_sb, "Dinode %llu has extent " 861 "list where extent # %d has no physical " 862 "block start", 863 (unsigned long long)OCFS2_I(inode)->ip_blkno, i); 864 status = -EIO; 865 goto bail; 866 } 867 868 if (bh) { 869 brelse(bh); 870 bh = NULL; 871 } 872 873 status = ocfs2_read_block(osb, blkno, &bh, OCFS2_BH_CACHED, 874 inode); 875 if (status < 0) { 876 mlog_errno(status); 877 goto bail; 878 } 879 880 eb = (struct ocfs2_extent_block *) bh->b_data; 881 if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) { 882 OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb); 883 status = -EIO; 884 goto bail; 885 } 886 el = &eb->h_list; 887 888 if (le16_to_cpu(el->l_next_free_rec) < 889 le16_to_cpu(el->l_count)) { 890 if (lowest_bh) 891 brelse(lowest_bh); 892 lowest_bh = bh; 893 get_bh(lowest_bh); 894 } 895 } 896 897 /* If we didn't find one and the fe doesn't have any room, 898 * then return '1' */ 899 if (!lowest_bh 900 && (fe->id2.i_list.l_next_free_rec == fe->id2.i_list.l_count)) 901 status = 1; 902 903 *target_bh = lowest_bh; 904 bail: 905 if (bh) 906 brelse(bh); 907 908 mlog_exit(status); 909 return status; 910 } 911 912 /* 913 * Grow a b-tree so that it has more records. 914 * 915 * We might shift the tree depth in which case existing paths should 916 * be considered invalid. 917 * 918 * Tree depth after the grow is returned via *final_depth. 919 * 920 * *last_eb_bh will be updated by ocfs2_add_branch(). 921 */ 922 static int ocfs2_grow_tree(struct inode *inode, handle_t *handle, 923 struct buffer_head *di_bh, int *final_depth, 924 struct buffer_head **last_eb_bh, 925 struct ocfs2_alloc_context *meta_ac) 926 { 927 int ret, shift; 928 struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data; 929 int depth = le16_to_cpu(di->id2.i_list.l_tree_depth); 930 struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); 931 struct buffer_head *bh = NULL; 932 933 BUG_ON(meta_ac == NULL); 934 935 shift = ocfs2_find_branch_target(osb, inode, di_bh, &bh); 936 if (shift < 0) { 937 ret = shift; 938 mlog_errno(ret); 939 goto out; 940 } 941 942 /* We traveled all the way to the bottom of the allocation tree 943 * and didn't find room for any more extents - we need to add 944 * another tree level */ 945 if (shift) { 946 BUG_ON(bh); 947 mlog(0, "need to shift tree depth (current = %d)\n", depth); 948 949 /* ocfs2_shift_tree_depth will return us a buffer with 950 * the new extent block (so we can pass that to 951 * ocfs2_add_branch). */ 952 ret = ocfs2_shift_tree_depth(osb, handle, inode, di_bh, 953 meta_ac, &bh); 954 if (ret < 0) { 955 mlog_errno(ret); 956 goto out; 957 } 958 depth++; 959 if (depth == 1) { 960 /* 961 * Special case: we have room now if we shifted from 962 * tree_depth 0, so no more work needs to be done. 963 * 964 * We won't be calling add_branch, so pass 965 * back *last_eb_bh as the new leaf. At depth 966 * zero, it should always be null so there's 967 * no reason to brelse. 968 */ 969 BUG_ON(*last_eb_bh); 970 get_bh(bh); 971 *last_eb_bh = bh; 972 goto out; 973 } 974 } 975 976 /* call ocfs2_add_branch to add the final part of the tree with 977 * the new data. */ 978 mlog(0, "add branch. bh = %p\n", bh); 979 ret = ocfs2_add_branch(osb, handle, inode, di_bh, bh, last_eb_bh, 980 meta_ac); 981 if (ret < 0) { 982 mlog_errno(ret); 983 goto out; 984 } 985 986 out: 987 if (final_depth) 988 *final_depth = depth; 989 brelse(bh); 990 return ret; 991 } 992 993 /* 994 * This function will discard the rightmost extent record. 995 */ 996 static void ocfs2_shift_records_right(struct ocfs2_extent_list *el) 997 { 998 int next_free = le16_to_cpu(el->l_next_free_rec); 999 int count = le16_to_cpu(el->l_count); 1000 unsigned int num_bytes; 1001 1002 BUG_ON(!next_free); 1003 /* This will cause us to go off the end of our extent list. */ 1004 BUG_ON(next_free >= count); 1005 1006 num_bytes = sizeof(struct ocfs2_extent_rec) * next_free; 1007 1008 memmove(&el->l_recs[1], &el->l_recs[0], num_bytes); 1009 } 1010 1011 static void ocfs2_rotate_leaf(struct ocfs2_extent_list *el, 1012 struct ocfs2_extent_rec *insert_rec) 1013 { 1014 int i, insert_index, next_free, has_empty, num_bytes; 1015 u32 insert_cpos = le32_to_cpu(insert_rec->e_cpos); 1016 struct ocfs2_extent_rec *rec; 1017 1018 next_free = le16_to_cpu(el->l_next_free_rec); 1019 has_empty = ocfs2_is_empty_extent(&el->l_recs[0]); 1020 1021 BUG_ON(!next_free); 1022 1023 /* The tree code before us didn't allow enough room in the leaf. */ 1024 BUG_ON(el->l_next_free_rec == el->l_count && !has_empty); 1025 1026 /* 1027 * The easiest way to approach this is to just remove the 1028 * empty extent and temporarily decrement next_free. 1029 */ 1030 if (has_empty) { 1031 /* 1032 * If next_free was 1 (only an empty extent), this 1033 * loop won't execute, which is fine. We still want 1034 * the decrement above to happen. 1035 */ 1036 for(i = 0; i < (next_free - 1); i++) 1037 el->l_recs[i] = el->l_recs[i+1]; 1038 1039 next_free--; 1040 } 1041 1042 /* 1043 * Figure out what the new record index should be. 1044 */ 1045 for(i = 0; i < next_free; i++) { 1046 rec = &el->l_recs[i]; 1047 1048 if (insert_cpos < le32_to_cpu(rec->e_cpos)) 1049 break; 1050 } 1051 insert_index = i; 1052 1053 mlog(0, "ins %u: index %d, has_empty %d, next_free %d, count %d\n", 1054 insert_cpos, insert_index, has_empty, next_free, le16_to_cpu(el->l_count)); 1055 1056 BUG_ON(insert_index < 0); 1057 BUG_ON(insert_index >= le16_to_cpu(el->l_count)); 1058 BUG_ON(insert_index > next_free); 1059 1060 /* 1061 * No need to memmove if we're just adding to the tail. 1062 */ 1063 if (insert_index != next_free) { 1064 BUG_ON(next_free >= le16_to_cpu(el->l_count)); 1065 1066 num_bytes = next_free - insert_index; 1067 num_bytes *= sizeof(struct ocfs2_extent_rec); 1068 memmove(&el->l_recs[insert_index + 1], 1069 &el->l_recs[insert_index], 1070 num_bytes); 1071 } 1072 1073 /* 1074 * Either we had an empty extent, and need to re-increment or 1075 * there was no empty extent on a non full rightmost leaf node, 1076 * in which case we still need to increment. 1077 */ 1078 next_free++; 1079 el->l_next_free_rec = cpu_to_le16(next_free); 1080 /* 1081 * Make sure none of the math above just messed up our tree. 1082 */ 1083 BUG_ON(le16_to_cpu(el->l_next_free_rec) > le16_to_cpu(el->l_count)); 1084 1085 el->l_recs[insert_index] = *insert_rec; 1086 1087 } 1088 1089 static void ocfs2_remove_empty_extent(struct ocfs2_extent_list *el) 1090 { 1091 int size, num_recs = le16_to_cpu(el->l_next_free_rec); 1092 1093 BUG_ON(num_recs == 0); 1094 1095 if (ocfs2_is_empty_extent(&el->l_recs[0])) { 1096 num_recs--; 1097 size = num_recs * sizeof(struct ocfs2_extent_rec); 1098 memmove(&el->l_recs[0], &el->l_recs[1], size); 1099 memset(&el->l_recs[num_recs], 0, 1100 sizeof(struct ocfs2_extent_rec)); 1101 el->l_next_free_rec = cpu_to_le16(num_recs); 1102 } 1103 } 1104 1105 /* 1106 * Create an empty extent record . 1107 * 1108 * l_next_free_rec may be updated. 1109 * 1110 * If an empty extent already exists do nothing. 1111 */ 1112 static void ocfs2_create_empty_extent(struct ocfs2_extent_list *el) 1113 { 1114 int next_free = le16_to_cpu(el->l_next_free_rec); 1115 1116 BUG_ON(le16_to_cpu(el->l_tree_depth) != 0); 1117 1118 if (next_free == 0) 1119 goto set_and_inc; 1120 1121 if (ocfs2_is_empty_extent(&el->l_recs[0])) 1122 return; 1123 1124 mlog_bug_on_msg(el->l_count == el->l_next_free_rec, 1125 "Asked to create an empty extent in a full list:\n" 1126 "count = %u, tree depth = %u", 1127 le16_to_cpu(el->l_count), 1128 le16_to_cpu(el->l_tree_depth)); 1129 1130 ocfs2_shift_records_right(el); 1131 1132 set_and_inc: 1133 le16_add_cpu(&el->l_next_free_rec, 1); 1134 memset(&el->l_recs[0], 0, sizeof(struct ocfs2_extent_rec)); 1135 } 1136 1137 /* 1138 * For a rotation which involves two leaf nodes, the "root node" is 1139 * the lowest level tree node which contains a path to both leafs. This 1140 * resulting set of information can be used to form a complete "subtree" 1141 * 1142 * This function is passed two full paths from the dinode down to a 1143 * pair of adjacent leaves. It's task is to figure out which path 1144 * index contains the subtree root - this can be the root index itself 1145 * in a worst-case rotation. 1146 * 1147 * The array index of the subtree root is passed back. 1148 */ 1149 static int ocfs2_find_subtree_root(struct inode *inode, 1150 struct ocfs2_path *left, 1151 struct ocfs2_path *right) 1152 { 1153 int i = 0; 1154 1155 /* 1156 * Check that the caller passed in two paths from the same tree. 1157 */ 1158 BUG_ON(path_root_bh(left) != path_root_bh(right)); 1159 1160 do { 1161 i++; 1162 1163 /* 1164 * The caller didn't pass two adjacent paths. 1165 */ 1166 mlog_bug_on_msg(i > left->p_tree_depth, 1167 "Inode %lu, left depth %u, right depth %u\n" 1168 "left leaf blk %llu, right leaf blk %llu\n", 1169 inode->i_ino, left->p_tree_depth, 1170 right->p_tree_depth, 1171 (unsigned long long)path_leaf_bh(left)->b_blocknr, 1172 (unsigned long long)path_leaf_bh(right)->b_blocknr); 1173 } while (left->p_node[i].bh->b_blocknr == 1174 right->p_node[i].bh->b_blocknr); 1175 1176 return i - 1; 1177 } 1178 1179 typedef void (path_insert_t)(void *, struct buffer_head *); 1180 1181 /* 1182 * Traverse a btree path in search of cpos, starting at root_el. 1183 * 1184 * This code can be called with a cpos larger than the tree, in which 1185 * case it will return the rightmost path. 1186 */ 1187 static int __ocfs2_find_path(struct inode *inode, 1188 struct ocfs2_extent_list *root_el, u32 cpos, 1189 path_insert_t *func, void *data) 1190 { 1191 int i, ret = 0; 1192 u32 range; 1193 u64 blkno; 1194 struct buffer_head *bh = NULL; 1195 struct ocfs2_extent_block *eb; 1196 struct ocfs2_extent_list *el; 1197 struct ocfs2_extent_rec *rec; 1198 struct ocfs2_inode_info *oi = OCFS2_I(inode); 1199 1200 el = root_el; 1201 while (el->l_tree_depth) { 1202 if (le16_to_cpu(el->l_next_free_rec) == 0) { 1203 ocfs2_error(inode->i_sb, 1204 "Inode %llu has empty extent list at " 1205 "depth %u\n", 1206 (unsigned long long)oi->ip_blkno, 1207 le16_to_cpu(el->l_tree_depth)); 1208 ret = -EROFS; 1209 goto out; 1210 1211 } 1212 1213 for(i = 0; i < le16_to_cpu(el->l_next_free_rec) - 1; i++) { 1214 rec = &el->l_recs[i]; 1215 1216 /* 1217 * In the case that cpos is off the allocation 1218 * tree, this should just wind up returning the 1219 * rightmost record. 1220 */ 1221 range = le32_to_cpu(rec->e_cpos) + 1222 ocfs2_rec_clusters(el, rec); 1223 if (cpos >= le32_to_cpu(rec->e_cpos) && cpos < range) 1224 break; 1225 } 1226 1227 blkno = le64_to_cpu(el->l_recs[i].e_blkno); 1228 if (blkno == 0) { 1229 ocfs2_error(inode->i_sb, 1230 "Inode %llu has bad blkno in extent list " 1231 "at depth %u (index %d)\n", 1232 (unsigned long long)oi->ip_blkno, 1233 le16_to_cpu(el->l_tree_depth), i); 1234 ret = -EROFS; 1235 goto out; 1236 } 1237 1238 brelse(bh); 1239 bh = NULL; 1240 ret = ocfs2_read_block(OCFS2_SB(inode->i_sb), blkno, 1241 &bh, OCFS2_BH_CACHED, inode); 1242 if (ret) { 1243 mlog_errno(ret); 1244 goto out; 1245 } 1246 1247 eb = (struct ocfs2_extent_block *) bh->b_data; 1248 el = &eb->h_list; 1249 if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) { 1250 OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb); 1251 ret = -EIO; 1252 goto out; 1253 } 1254 1255 if (le16_to_cpu(el->l_next_free_rec) > 1256 le16_to_cpu(el->l_count)) { 1257 ocfs2_error(inode->i_sb, 1258 "Inode %llu has bad count in extent list " 1259 "at block %llu (next free=%u, count=%u)\n", 1260 (unsigned long long)oi->ip_blkno, 1261 (unsigned long long)bh->b_blocknr, 1262 le16_to_cpu(el->l_next_free_rec), 1263 le16_to_cpu(el->l_count)); 1264 ret = -EROFS; 1265 goto out; 1266 } 1267 1268 if (func) 1269 func(data, bh); 1270 } 1271 1272 out: 1273 /* 1274 * Catch any trailing bh that the loop didn't handle. 1275 */ 1276 brelse(bh); 1277 1278 return ret; 1279 } 1280 1281 /* 1282 * Given an initialized path (that is, it has a valid root extent 1283 * list), this function will traverse the btree in search of the path 1284 * which would contain cpos. 1285 * 1286 * The path traveled is recorded in the path structure. 1287 * 1288 * Note that this will not do any comparisons on leaf node extent 1289 * records, so it will work fine in the case that we just added a tree 1290 * branch. 1291 */ 1292 struct find_path_data { 1293 int index; 1294 struct ocfs2_path *path; 1295 }; 1296 static void find_path_ins(void *data, struct buffer_head *bh) 1297 { 1298 struct find_path_data *fp = data; 1299 1300 get_bh(bh); 1301 ocfs2_path_insert_eb(fp->path, fp->index, bh); 1302 fp->index++; 1303 } 1304 static int ocfs2_find_path(struct inode *inode, struct ocfs2_path *path, 1305 u32 cpos) 1306 { 1307 struct find_path_data data; 1308 1309 data.index = 1; 1310 data.path = path; 1311 return __ocfs2_find_path(inode, path_root_el(path), cpos, 1312 find_path_ins, &data); 1313 } 1314 1315 static void find_leaf_ins(void *data, struct buffer_head *bh) 1316 { 1317 struct ocfs2_extent_block *eb =(struct ocfs2_extent_block *)bh->b_data; 1318 struct ocfs2_extent_list *el = &eb->h_list; 1319 struct buffer_head **ret = data; 1320 1321 /* We want to retain only the leaf block. */ 1322 if (le16_to_cpu(el->l_tree_depth) == 0) { 1323 get_bh(bh); 1324 *ret = bh; 1325 } 1326 } 1327 /* 1328 * Find the leaf block in the tree which would contain cpos. No 1329 * checking of the actual leaf is done. 1330 * 1331 * Some paths want to call this instead of allocating a path structure 1332 * and calling ocfs2_find_path(). 1333 * 1334 * This function doesn't handle non btree extent lists. 1335 */ 1336 int ocfs2_find_leaf(struct inode *inode, struct ocfs2_extent_list *root_el, 1337 u32 cpos, struct buffer_head **leaf_bh) 1338 { 1339 int ret; 1340 struct buffer_head *bh = NULL; 1341 1342 ret = __ocfs2_find_path(inode, root_el, cpos, find_leaf_ins, &bh); 1343 if (ret) { 1344 mlog_errno(ret); 1345 goto out; 1346 } 1347 1348 *leaf_bh = bh; 1349 out: 1350 return ret; 1351 } 1352 1353 /* 1354 * Adjust the adjacent records (left_rec, right_rec) involved in a rotation. 1355 * 1356 * Basically, we've moved stuff around at the bottom of the tree and 1357 * we need to fix up the extent records above the changes to reflect 1358 * the new changes. 1359 * 1360 * left_rec: the record on the left. 1361 * left_child_el: is the child list pointed to by left_rec 1362 * right_rec: the record to the right of left_rec 1363 * right_child_el: is the child list pointed to by right_rec 1364 * 1365 * By definition, this only works on interior nodes. 1366 */ 1367 static void ocfs2_adjust_adjacent_records(struct ocfs2_extent_rec *left_rec, 1368 struct ocfs2_extent_list *left_child_el, 1369 struct ocfs2_extent_rec *right_rec, 1370 struct ocfs2_extent_list *right_child_el) 1371 { 1372 u32 left_clusters, right_end; 1373 1374 /* 1375 * Interior nodes never have holes. Their cpos is the cpos of 1376 * the leftmost record in their child list. Their cluster 1377 * count covers the full theoretical range of their child list 1378 * - the range between their cpos and the cpos of the record 1379 * immediately to their right. 1380 */ 1381 left_clusters = le32_to_cpu(right_child_el->l_recs[0].e_cpos); 1382 if (ocfs2_is_empty_extent(&right_child_el->l_recs[0])) { 1383 BUG_ON(le16_to_cpu(right_child_el->l_next_free_rec) <= 1); 1384 left_clusters = le32_to_cpu(right_child_el->l_recs[1].e_cpos); 1385 } 1386 left_clusters -= le32_to_cpu(left_rec->e_cpos); 1387 left_rec->e_int_clusters = cpu_to_le32(left_clusters); 1388 1389 /* 1390 * Calculate the rightmost cluster count boundary before 1391 * moving cpos - we will need to adjust clusters after 1392 * updating e_cpos to keep the same highest cluster count. 1393 */ 1394 right_end = le32_to_cpu(right_rec->e_cpos); 1395 right_end += le32_to_cpu(right_rec->e_int_clusters); 1396 1397 right_rec->e_cpos = left_rec->e_cpos; 1398 le32_add_cpu(&right_rec->e_cpos, left_clusters); 1399 1400 right_end -= le32_to_cpu(right_rec->e_cpos); 1401 right_rec->e_int_clusters = cpu_to_le32(right_end); 1402 } 1403 1404 /* 1405 * Adjust the adjacent root node records involved in a 1406 * rotation. left_el_blkno is passed in as a key so that we can easily 1407 * find it's index in the root list. 1408 */ 1409 static void ocfs2_adjust_root_records(struct ocfs2_extent_list *root_el, 1410 struct ocfs2_extent_list *left_el, 1411 struct ocfs2_extent_list *right_el, 1412 u64 left_el_blkno) 1413 { 1414 int i; 1415 1416 BUG_ON(le16_to_cpu(root_el->l_tree_depth) <= 1417 le16_to_cpu(left_el->l_tree_depth)); 1418 1419 for(i = 0; i < le16_to_cpu(root_el->l_next_free_rec) - 1; i++) { 1420 if (le64_to_cpu(root_el->l_recs[i].e_blkno) == left_el_blkno) 1421 break; 1422 } 1423 1424 /* 1425 * The path walking code should have never returned a root and 1426 * two paths which are not adjacent. 1427 */ 1428 BUG_ON(i >= (le16_to_cpu(root_el->l_next_free_rec) - 1)); 1429 1430 ocfs2_adjust_adjacent_records(&root_el->l_recs[i], left_el, 1431 &root_el->l_recs[i + 1], right_el); 1432 } 1433 1434 /* 1435 * We've changed a leaf block (in right_path) and need to reflect that 1436 * change back up the subtree. 1437 * 1438 * This happens in multiple places: 1439 * - When we've moved an extent record from the left path leaf to the right 1440 * path leaf to make room for an empty extent in the left path leaf. 1441 * - When our insert into the right path leaf is at the leftmost edge 1442 * and requires an update of the path immediately to it's left. This 1443 * can occur at the end of some types of rotation and appending inserts. 1444 * - When we've adjusted the last extent record in the left path leaf and the 1445 * 1st extent record in the right path leaf during cross extent block merge. 1446 */ 1447 static void ocfs2_complete_edge_insert(struct inode *inode, handle_t *handle, 1448 struct ocfs2_path *left_path, 1449 struct ocfs2_path *right_path, 1450 int subtree_index) 1451 { 1452 int ret, i, idx; 1453 struct ocfs2_extent_list *el, *left_el, *right_el; 1454 struct ocfs2_extent_rec *left_rec, *right_rec; 1455 struct buffer_head *root_bh = left_path->p_node[subtree_index].bh; 1456 1457 /* 1458 * Update the counts and position values within all the 1459 * interior nodes to reflect the leaf rotation we just did. 1460 * 1461 * The root node is handled below the loop. 1462 * 1463 * We begin the loop with right_el and left_el pointing to the 1464 * leaf lists and work our way up. 1465 * 1466 * NOTE: within this loop, left_el and right_el always refer 1467 * to the *child* lists. 1468 */ 1469 left_el = path_leaf_el(left_path); 1470 right_el = path_leaf_el(right_path); 1471 for(i = left_path->p_tree_depth - 1; i > subtree_index; i--) { 1472 mlog(0, "Adjust records at index %u\n", i); 1473 1474 /* 1475 * One nice property of knowing that all of these 1476 * nodes are below the root is that we only deal with 1477 * the leftmost right node record and the rightmost 1478 * left node record. 1479 */ 1480 el = left_path->p_node[i].el; 1481 idx = le16_to_cpu(left_el->l_next_free_rec) - 1; 1482 left_rec = &el->l_recs[idx]; 1483 1484 el = right_path->p_node[i].el; 1485 right_rec = &el->l_recs[0]; 1486 1487 ocfs2_adjust_adjacent_records(left_rec, left_el, right_rec, 1488 right_el); 1489 1490 ret = ocfs2_journal_dirty(handle, left_path->p_node[i].bh); 1491 if (ret) 1492 mlog_errno(ret); 1493 1494 ret = ocfs2_journal_dirty(handle, right_path->p_node[i].bh); 1495 if (ret) 1496 mlog_errno(ret); 1497 1498 /* 1499 * Setup our list pointers now so that the current 1500 * parents become children in the next iteration. 1501 */ 1502 left_el = left_path->p_node[i].el; 1503 right_el = right_path->p_node[i].el; 1504 } 1505 1506 /* 1507 * At the root node, adjust the two adjacent records which 1508 * begin our path to the leaves. 1509 */ 1510 1511 el = left_path->p_node[subtree_index].el; 1512 left_el = left_path->p_node[subtree_index + 1].el; 1513 right_el = right_path->p_node[subtree_index + 1].el; 1514 1515 ocfs2_adjust_root_records(el, left_el, right_el, 1516 left_path->p_node[subtree_index + 1].bh->b_blocknr); 1517 1518 root_bh = left_path->p_node[subtree_index].bh; 1519 1520 ret = ocfs2_journal_dirty(handle, root_bh); 1521 if (ret) 1522 mlog_errno(ret); 1523 } 1524 1525 static int ocfs2_rotate_subtree_right(struct inode *inode, 1526 handle_t *handle, 1527 struct ocfs2_path *left_path, 1528 struct ocfs2_path *right_path, 1529 int subtree_index) 1530 { 1531 int ret, i; 1532 struct buffer_head *right_leaf_bh; 1533 struct buffer_head *left_leaf_bh = NULL; 1534 struct buffer_head *root_bh; 1535 struct ocfs2_extent_list *right_el, *left_el; 1536 struct ocfs2_extent_rec move_rec; 1537 1538 left_leaf_bh = path_leaf_bh(left_path); 1539 left_el = path_leaf_el(left_path); 1540 1541 if (left_el->l_next_free_rec != left_el->l_count) { 1542 ocfs2_error(inode->i_sb, 1543 "Inode %llu has non-full interior leaf node %llu" 1544 "(next free = %u)", 1545 (unsigned long long)OCFS2_I(inode)->ip_blkno, 1546 (unsigned long long)left_leaf_bh->b_blocknr, 1547 le16_to_cpu(left_el->l_next_free_rec)); 1548 return -EROFS; 1549 } 1550 1551 /* 1552 * This extent block may already have an empty record, so we 1553 * return early if so. 1554 */ 1555 if (ocfs2_is_empty_extent(&left_el->l_recs[0])) 1556 return 0; 1557 1558 root_bh = left_path->p_node[subtree_index].bh; 1559 BUG_ON(root_bh != right_path->p_node[subtree_index].bh); 1560 1561 ret = ocfs2_journal_access(handle, inode, root_bh, 1562 OCFS2_JOURNAL_ACCESS_WRITE); 1563 if (ret) { 1564 mlog_errno(ret); 1565 goto out; 1566 } 1567 1568 for(i = subtree_index + 1; i < path_num_items(right_path); i++) { 1569 ret = ocfs2_journal_access(handle, inode, 1570 right_path->p_node[i].bh, 1571 OCFS2_JOURNAL_ACCESS_WRITE); 1572 if (ret) { 1573 mlog_errno(ret); 1574 goto out; 1575 } 1576 1577 ret = ocfs2_journal_access(handle, inode, 1578 left_path->p_node[i].bh, 1579 OCFS2_JOURNAL_ACCESS_WRITE); 1580 if (ret) { 1581 mlog_errno(ret); 1582 goto out; 1583 } 1584 } 1585 1586 right_leaf_bh = path_leaf_bh(right_path); 1587 right_el = path_leaf_el(right_path); 1588 1589 /* This is a code error, not a disk corruption. */ 1590 mlog_bug_on_msg(!right_el->l_next_free_rec, "Inode %llu: Rotate fails " 1591 "because rightmost leaf block %llu is empty\n", 1592 (unsigned long long)OCFS2_I(inode)->ip_blkno, 1593 (unsigned long long)right_leaf_bh->b_blocknr); 1594 1595 ocfs2_create_empty_extent(right_el); 1596 1597 ret = ocfs2_journal_dirty(handle, right_leaf_bh); 1598 if (ret) { 1599 mlog_errno(ret); 1600 goto out; 1601 } 1602 1603 /* Do the copy now. */ 1604 i = le16_to_cpu(left_el->l_next_free_rec) - 1; 1605 move_rec = left_el->l_recs[i]; 1606 right_el->l_recs[0] = move_rec; 1607 1608 /* 1609 * Clear out the record we just copied and shift everything 1610 * over, leaving an empty extent in the left leaf. 1611 * 1612 * We temporarily subtract from next_free_rec so that the 1613 * shift will lose the tail record (which is now defunct). 1614 */ 1615 le16_add_cpu(&left_el->l_next_free_rec, -1); 1616 ocfs2_shift_records_right(left_el); 1617 memset(&left_el->l_recs[0], 0, sizeof(struct ocfs2_extent_rec)); 1618 le16_add_cpu(&left_el->l_next_free_rec, 1); 1619 1620 ret = ocfs2_journal_dirty(handle, left_leaf_bh); 1621 if (ret) { 1622 mlog_errno(ret); 1623 goto out; 1624 } 1625 1626 ocfs2_complete_edge_insert(inode, handle, left_path, right_path, 1627 subtree_index); 1628 1629 out: 1630 return ret; 1631 } 1632 1633 /* 1634 * Given a full path, determine what cpos value would return us a path 1635 * containing the leaf immediately to the left of the current one. 1636 * 1637 * Will return zero if the path passed in is already the leftmost path. 1638 */ 1639 static int ocfs2_find_cpos_for_left_leaf(struct super_block *sb, 1640 struct ocfs2_path *path, u32 *cpos) 1641 { 1642 int i, j, ret = 0; 1643 u64 blkno; 1644 struct ocfs2_extent_list *el; 1645 1646 BUG_ON(path->p_tree_depth == 0); 1647 1648 *cpos = 0; 1649 1650 blkno = path_leaf_bh(path)->b_blocknr; 1651 1652 /* Start at the tree node just above the leaf and work our way up. */ 1653 i = path->p_tree_depth - 1; 1654 while (i >= 0) { 1655 el = path->p_node[i].el; 1656 1657 /* 1658 * Find the extent record just before the one in our 1659 * path. 1660 */ 1661 for(j = 0; j < le16_to_cpu(el->l_next_free_rec); j++) { 1662 if (le64_to_cpu(el->l_recs[j].e_blkno) == blkno) { 1663 if (j == 0) { 1664 if (i == 0) { 1665 /* 1666 * We've determined that the 1667 * path specified is already 1668 * the leftmost one - return a 1669 * cpos of zero. 1670 */ 1671 goto out; 1672 } 1673 /* 1674 * The leftmost record points to our 1675 * leaf - we need to travel up the 1676 * tree one level. 1677 */ 1678 goto next_node; 1679 } 1680 1681 *cpos = le32_to_cpu(el->l_recs[j - 1].e_cpos); 1682 *cpos = *cpos + ocfs2_rec_clusters(el, 1683 &el->l_recs[j - 1]); 1684 *cpos = *cpos - 1; 1685 goto out; 1686 } 1687 } 1688 1689 /* 1690 * If we got here, we never found a valid node where 1691 * the tree indicated one should be. 1692 */ 1693 ocfs2_error(sb, 1694 "Invalid extent tree at extent block %llu\n", 1695 (unsigned long long)blkno); 1696 ret = -EROFS; 1697 goto out; 1698 1699 next_node: 1700 blkno = path->p_node[i].bh->b_blocknr; 1701 i--; 1702 } 1703 1704 out: 1705 return ret; 1706 } 1707 1708 /* 1709 * Extend the transaction by enough credits to complete the rotation, 1710 * and still leave at least the original number of credits allocated 1711 * to this transaction. 1712 */ 1713 static int ocfs2_extend_rotate_transaction(handle_t *handle, int subtree_depth, 1714 int op_credits, 1715 struct ocfs2_path *path) 1716 { 1717 int credits = (path->p_tree_depth - subtree_depth) * 2 + 1 + op_credits; 1718 1719 if (handle->h_buffer_credits < credits) 1720 return ocfs2_extend_trans(handle, credits); 1721 1722 return 0; 1723 } 1724 1725 /* 1726 * Trap the case where we're inserting into the theoretical range past 1727 * the _actual_ left leaf range. Otherwise, we'll rotate a record 1728 * whose cpos is less than ours into the right leaf. 1729 * 1730 * It's only necessary to look at the rightmost record of the left 1731 * leaf because the logic that calls us should ensure that the 1732 * theoretical ranges in the path components above the leaves are 1733 * correct. 1734 */ 1735 static int ocfs2_rotate_requires_path_adjustment(struct ocfs2_path *left_path, 1736 u32 insert_cpos) 1737 { 1738 struct ocfs2_extent_list *left_el; 1739 struct ocfs2_extent_rec *rec; 1740 int next_free; 1741 1742 left_el = path_leaf_el(left_path); 1743 next_free = le16_to_cpu(left_el->l_next_free_rec); 1744 rec = &left_el->l_recs[next_free - 1]; 1745 1746 if (insert_cpos > le32_to_cpu(rec->e_cpos)) 1747 return 1; 1748 return 0; 1749 } 1750 1751 static int ocfs2_leftmost_rec_contains(struct ocfs2_extent_list *el, u32 cpos) 1752 { 1753 int next_free = le16_to_cpu(el->l_next_free_rec); 1754 unsigned int range; 1755 struct ocfs2_extent_rec *rec; 1756 1757 if (next_free == 0) 1758 return 0; 1759 1760 rec = &el->l_recs[0]; 1761 if (ocfs2_is_empty_extent(rec)) { 1762 /* Empty list. */ 1763 if (next_free == 1) 1764 return 0; 1765 rec = &el->l_recs[1]; 1766 } 1767 1768 range = le32_to_cpu(rec->e_cpos) + ocfs2_rec_clusters(el, rec); 1769 if (cpos >= le32_to_cpu(rec->e_cpos) && cpos < range) 1770 return 1; 1771 return 0; 1772 } 1773 1774 /* 1775 * Rotate all the records in a btree right one record, starting at insert_cpos. 1776 * 1777 * The path to the rightmost leaf should be passed in. 1778 * 1779 * The array is assumed to be large enough to hold an entire path (tree depth). 1780 * 1781 * Upon succesful return from this function: 1782 * 1783 * - The 'right_path' array will contain a path to the leaf block 1784 * whose range contains e_cpos. 1785 * - That leaf block will have a single empty extent in list index 0. 1786 * - In the case that the rotation requires a post-insert update, 1787 * *ret_left_path will contain a valid path which can be passed to 1788 * ocfs2_insert_path(). 1789 */ 1790 static int ocfs2_rotate_tree_right(struct inode *inode, 1791 handle_t *handle, 1792 enum ocfs2_split_type split, 1793 u32 insert_cpos, 1794 struct ocfs2_path *right_path, 1795 struct ocfs2_path **ret_left_path) 1796 { 1797 int ret, start, orig_credits = handle->h_buffer_credits; 1798 u32 cpos; 1799 struct ocfs2_path *left_path = NULL; 1800 1801 *ret_left_path = NULL; 1802 1803 left_path = ocfs2_new_path(path_root_bh(right_path), 1804 path_root_el(right_path)); 1805 if (!left_path) { 1806 ret = -ENOMEM; 1807 mlog_errno(ret); 1808 goto out; 1809 } 1810 1811 ret = ocfs2_find_cpos_for_left_leaf(inode->i_sb, right_path, &cpos); 1812 if (ret) { 1813 mlog_errno(ret); 1814 goto out; 1815 } 1816 1817 mlog(0, "Insert: %u, first left path cpos: %u\n", insert_cpos, cpos); 1818 1819 /* 1820 * What we want to do here is: 1821 * 1822 * 1) Start with the rightmost path. 1823 * 1824 * 2) Determine a path to the leaf block directly to the left 1825 * of that leaf. 1826 * 1827 * 3) Determine the 'subtree root' - the lowest level tree node 1828 * which contains a path to both leaves. 1829 * 1830 * 4) Rotate the subtree. 1831 * 1832 * 5) Find the next subtree by considering the left path to be 1833 * the new right path. 1834 * 1835 * The check at the top of this while loop also accepts 1836 * insert_cpos == cpos because cpos is only a _theoretical_ 1837 * value to get us the left path - insert_cpos might very well 1838 * be filling that hole. 1839 * 1840 * Stop at a cpos of '0' because we either started at the 1841 * leftmost branch (i.e., a tree with one branch and a 1842 * rotation inside of it), or we've gone as far as we can in 1843 * rotating subtrees. 1844 */ 1845 while (cpos && insert_cpos <= cpos) { 1846 mlog(0, "Rotating a tree: ins. cpos: %u, left path cpos: %u\n", 1847 insert_cpos, cpos); 1848 1849 ret = ocfs2_find_path(inode, left_path, cpos); 1850 if (ret) { 1851 mlog_errno(ret); 1852 goto out; 1853 } 1854 1855 mlog_bug_on_msg(path_leaf_bh(left_path) == 1856 path_leaf_bh(right_path), 1857 "Inode %lu: error during insert of %u " 1858 "(left path cpos %u) results in two identical " 1859 "paths ending at %llu\n", 1860 inode->i_ino, insert_cpos, cpos, 1861 (unsigned long long) 1862 path_leaf_bh(left_path)->b_blocknr); 1863 1864 if (split == SPLIT_NONE && 1865 ocfs2_rotate_requires_path_adjustment(left_path, 1866 insert_cpos)) { 1867 1868 /* 1869 * We've rotated the tree as much as we 1870 * should. The rest is up to 1871 * ocfs2_insert_path() to complete, after the 1872 * record insertion. We indicate this 1873 * situation by returning the left path. 1874 * 1875 * The reason we don't adjust the records here 1876 * before the record insert is that an error 1877 * later might break the rule where a parent 1878 * record e_cpos will reflect the actual 1879 * e_cpos of the 1st nonempty record of the 1880 * child list. 1881 */ 1882 *ret_left_path = left_path; 1883 goto out_ret_path; 1884 } 1885 1886 start = ocfs2_find_subtree_root(inode, left_path, right_path); 1887 1888 mlog(0, "Subtree root at index %d (blk %llu, depth %d)\n", 1889 start, 1890 (unsigned long long) right_path->p_node[start].bh->b_blocknr, 1891 right_path->p_tree_depth); 1892 1893 ret = ocfs2_extend_rotate_transaction(handle, start, 1894 orig_credits, right_path); 1895 if (ret) { 1896 mlog_errno(ret); 1897 goto out; 1898 } 1899 1900 ret = ocfs2_rotate_subtree_right(inode, handle, left_path, 1901 right_path, start); 1902 if (ret) { 1903 mlog_errno(ret); 1904 goto out; 1905 } 1906 1907 if (split != SPLIT_NONE && 1908 ocfs2_leftmost_rec_contains(path_leaf_el(right_path), 1909 insert_cpos)) { 1910 /* 1911 * A rotate moves the rightmost left leaf 1912 * record over to the leftmost right leaf 1913 * slot. If we're doing an extent split 1914 * instead of a real insert, then we have to 1915 * check that the extent to be split wasn't 1916 * just moved over. If it was, then we can 1917 * exit here, passing left_path back - 1918 * ocfs2_split_extent() is smart enough to 1919 * search both leaves. 1920 */ 1921 *ret_left_path = left_path; 1922 goto out_ret_path; 1923 } 1924 1925 /* 1926 * There is no need to re-read the next right path 1927 * as we know that it'll be our current left 1928 * path. Optimize by copying values instead. 1929 */ 1930 ocfs2_mv_path(right_path, left_path); 1931 1932 ret = ocfs2_find_cpos_for_left_leaf(inode->i_sb, right_path, 1933 &cpos); 1934 if (ret) { 1935 mlog_errno(ret); 1936 goto out; 1937 } 1938 } 1939 1940 out: 1941 ocfs2_free_path(left_path); 1942 1943 out_ret_path: 1944 return ret; 1945 } 1946 1947 static void ocfs2_update_edge_lengths(struct inode *inode, handle_t *handle, 1948 struct ocfs2_path *path) 1949 { 1950 int i, idx; 1951 struct ocfs2_extent_rec *rec; 1952 struct ocfs2_extent_list *el; 1953 struct ocfs2_extent_block *eb; 1954 u32 range; 1955 1956 /* Path should always be rightmost. */ 1957 eb = (struct ocfs2_extent_block *)path_leaf_bh(path)->b_data; 1958 BUG_ON(eb->h_next_leaf_blk != 0ULL); 1959 1960 el = &eb->h_list; 1961 BUG_ON(le16_to_cpu(el->l_next_free_rec) == 0); 1962 idx = le16_to_cpu(el->l_next_free_rec) - 1; 1963 rec = &el->l_recs[idx]; 1964 range = le32_to_cpu(rec->e_cpos) + ocfs2_rec_clusters(el, rec); 1965 1966 for (i = 0; i < path->p_tree_depth; i++) { 1967 el = path->p_node[i].el; 1968 idx = le16_to_cpu(el->l_next_free_rec) - 1; 1969 rec = &el->l_recs[idx]; 1970 1971 rec->e_int_clusters = cpu_to_le32(range); 1972 le32_add_cpu(&rec->e_int_clusters, -le32_to_cpu(rec->e_cpos)); 1973 1974 ocfs2_journal_dirty(handle, path->p_node[i].bh); 1975 } 1976 } 1977 1978 static void ocfs2_unlink_path(struct inode *inode, handle_t *handle, 1979 struct ocfs2_cached_dealloc_ctxt *dealloc, 1980 struct ocfs2_path *path, int unlink_start) 1981 { 1982 int ret, i; 1983 struct ocfs2_extent_block *eb; 1984 struct ocfs2_extent_list *el; 1985 struct buffer_head *bh; 1986 1987 for(i = unlink_start; i < path_num_items(path); i++) { 1988 bh = path->p_node[i].bh; 1989 1990 eb = (struct ocfs2_extent_block *)bh->b_data; 1991 /* 1992 * Not all nodes might have had their final count 1993 * decremented by the caller - handle this here. 1994 */ 1995 el = &eb->h_list; 1996 if (le16_to_cpu(el->l_next_free_rec) > 1) { 1997 mlog(ML_ERROR, 1998 "Inode %llu, attempted to remove extent block " 1999 "%llu with %u records\n", 2000 (unsigned long long)OCFS2_I(inode)->ip_blkno, 2001 (unsigned long long)le64_to_cpu(eb->h_blkno), 2002 le16_to_cpu(el->l_next_free_rec)); 2003 2004 ocfs2_journal_dirty(handle, bh); 2005 ocfs2_remove_from_cache(inode, bh); 2006 continue; 2007 } 2008 2009 el->l_next_free_rec = 0; 2010 memset(&el->l_recs[0], 0, sizeof(struct ocfs2_extent_rec)); 2011 2012 ocfs2_journal_dirty(handle, bh); 2013 2014 ret = ocfs2_cache_extent_block_free(dealloc, eb); 2015 if (ret) 2016 mlog_errno(ret); 2017 2018 ocfs2_remove_from_cache(inode, bh); 2019 } 2020 } 2021 2022 static void ocfs2_unlink_subtree(struct inode *inode, handle_t *handle, 2023 struct ocfs2_path *left_path, 2024 struct ocfs2_path *right_path, 2025 int subtree_index, 2026 struct ocfs2_cached_dealloc_ctxt *dealloc) 2027 { 2028 int i; 2029 struct buffer_head *root_bh = left_path->p_node[subtree_index].bh; 2030 struct ocfs2_extent_list *root_el = left_path->p_node[subtree_index].el; 2031 struct ocfs2_extent_list *el; 2032 struct ocfs2_extent_block *eb; 2033 2034 el = path_leaf_el(left_path); 2035 2036 eb = (struct ocfs2_extent_block *)right_path->p_node[subtree_index + 1].bh->b_data; 2037 2038 for(i = 1; i < le16_to_cpu(root_el->l_next_free_rec); i++) 2039 if (root_el->l_recs[i].e_blkno == eb->h_blkno) 2040 break; 2041 2042 BUG_ON(i >= le16_to_cpu(root_el->l_next_free_rec)); 2043 2044 memset(&root_el->l_recs[i], 0, sizeof(struct ocfs2_extent_rec)); 2045 le16_add_cpu(&root_el->l_next_free_rec, -1); 2046 2047 eb = (struct ocfs2_extent_block *)path_leaf_bh(left_path)->b_data; 2048 eb->h_next_leaf_blk = 0; 2049 2050 ocfs2_journal_dirty(handle, root_bh); 2051 ocfs2_journal_dirty(handle, path_leaf_bh(left_path)); 2052 2053 ocfs2_unlink_path(inode, handle, dealloc, right_path, 2054 subtree_index + 1); 2055 } 2056 2057 static int ocfs2_rotate_subtree_left(struct inode *inode, handle_t *handle, 2058 struct ocfs2_path *left_path, 2059 struct ocfs2_path *right_path, 2060 int subtree_index, 2061 struct ocfs2_cached_dealloc_ctxt *dealloc, 2062 int *deleted) 2063 { 2064 int ret, i, del_right_subtree = 0, right_has_empty = 0; 2065 struct buffer_head *root_bh, *di_bh = path_root_bh(right_path); 2066 struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data; 2067 struct ocfs2_extent_list *right_leaf_el, *left_leaf_el; 2068 struct ocfs2_extent_block *eb; 2069 2070 *deleted = 0; 2071 2072 right_leaf_el = path_leaf_el(right_path); 2073 left_leaf_el = path_leaf_el(left_path); 2074 root_bh = left_path->p_node[subtree_index].bh; 2075 BUG_ON(root_bh != right_path->p_node[subtree_index].bh); 2076 2077 if (!ocfs2_is_empty_extent(&left_leaf_el->l_recs[0])) 2078 return 0; 2079 2080 eb = (struct ocfs2_extent_block *)path_leaf_bh(right_path)->b_data; 2081 if (ocfs2_is_empty_extent(&right_leaf_el->l_recs[0])) { 2082 /* 2083 * It's legal for us to proceed if the right leaf is 2084 * the rightmost one and it has an empty extent. There 2085 * are two cases to handle - whether the leaf will be 2086 * empty after removal or not. If the leaf isn't empty 2087 * then just remove the empty extent up front. The 2088 * next block will handle empty leaves by flagging 2089 * them for unlink. 2090 * 2091 * Non rightmost leaves will throw -EAGAIN and the 2092 * caller can manually move the subtree and retry. 2093 */ 2094 2095 if (eb->h_next_leaf_blk != 0ULL) 2096 return -EAGAIN; 2097 2098 if (le16_to_cpu(right_leaf_el->l_next_free_rec) > 1) { 2099 ret = ocfs2_journal_access(handle, inode, 2100 path_leaf_bh(right_path), 2101 OCFS2_JOURNAL_ACCESS_WRITE); 2102 if (ret) { 2103 mlog_errno(ret); 2104 goto out; 2105 } 2106 2107 ocfs2_remove_empty_extent(right_leaf_el); 2108 } else 2109 right_has_empty = 1; 2110 } 2111 2112 if (eb->h_next_leaf_blk == 0ULL && 2113 le16_to_cpu(right_leaf_el->l_next_free_rec) == 1) { 2114 /* 2115 * We have to update i_last_eb_blk during the meta 2116 * data delete. 2117 */ 2118 ret = ocfs2_journal_access(handle, inode, di_bh, 2119 OCFS2_JOURNAL_ACCESS_WRITE); 2120 if (ret) { 2121 mlog_errno(ret); 2122 goto out; 2123 } 2124 2125 del_right_subtree = 1; 2126 } 2127 2128 /* 2129 * Getting here with an empty extent in the right path implies 2130 * that it's the rightmost path and will be deleted. 2131 */ 2132 BUG_ON(right_has_empty && !del_right_subtree); 2133 2134 ret = ocfs2_journal_access(handle, inode, root_bh, 2135 OCFS2_JOURNAL_ACCESS_WRITE); 2136 if (ret) { 2137 mlog_errno(ret); 2138 goto out; 2139 } 2140 2141 for(i = subtree_index + 1; i < path_num_items(right_path); i++) { 2142 ret = ocfs2_journal_access(handle, inode, 2143 right_path->p_node[i].bh, 2144 OCFS2_JOURNAL_ACCESS_WRITE); 2145 if (ret) { 2146 mlog_errno(ret); 2147 goto out; 2148 } 2149 2150 ret = ocfs2_journal_access(handle, inode, 2151 left_path->p_node[i].bh, 2152 OCFS2_JOURNAL_ACCESS_WRITE); 2153 if (ret) { 2154 mlog_errno(ret); 2155 goto out; 2156 } 2157 } 2158 2159 if (!right_has_empty) { 2160 /* 2161 * Only do this if we're moving a real 2162 * record. Otherwise, the action is delayed until 2163 * after removal of the right path in which case we 2164 * can do a simple shift to remove the empty extent. 2165 */ 2166 ocfs2_rotate_leaf(left_leaf_el, &right_leaf_el->l_recs[0]); 2167 memset(&right_leaf_el->l_recs[0], 0, 2168 sizeof(struct ocfs2_extent_rec)); 2169 } 2170 if (eb->h_next_leaf_blk == 0ULL) { 2171 /* 2172 * Move recs over to get rid of empty extent, decrease 2173 * next_free. This is allowed to remove the last 2174 * extent in our leaf (setting l_next_free_rec to 2175 * zero) - the delete code below won't care. 2176 */ 2177 ocfs2_remove_empty_extent(right_leaf_el); 2178 } 2179 2180 ret = ocfs2_journal_dirty(handle, path_leaf_bh(left_path)); 2181 if (ret) 2182 mlog_errno(ret); 2183 ret = ocfs2_journal_dirty(handle, path_leaf_bh(right_path)); 2184 if (ret) 2185 mlog_errno(ret); 2186 2187 if (del_right_subtree) { 2188 ocfs2_unlink_subtree(inode, handle, left_path, right_path, 2189 subtree_index, dealloc); 2190 ocfs2_update_edge_lengths(inode, handle, left_path); 2191 2192 eb = (struct ocfs2_extent_block *)path_leaf_bh(left_path)->b_data; 2193 di->i_last_eb_blk = eb->h_blkno; 2194 2195 /* 2196 * Removal of the extent in the left leaf was skipped 2197 * above so we could delete the right path 2198 * 1st. 2199 */ 2200 if (right_has_empty) 2201 ocfs2_remove_empty_extent(left_leaf_el); 2202 2203 ret = ocfs2_journal_dirty(handle, di_bh); 2204 if (ret) 2205 mlog_errno(ret); 2206 2207 *deleted = 1; 2208 } else 2209 ocfs2_complete_edge_insert(inode, handle, left_path, right_path, 2210 subtree_index); 2211 2212 out: 2213 return ret; 2214 } 2215 2216 /* 2217 * Given a full path, determine what cpos value would return us a path 2218 * containing the leaf immediately to the right of the current one. 2219 * 2220 * Will return zero if the path passed in is already the rightmost path. 2221 * 2222 * This looks similar, but is subtly different to 2223 * ocfs2_find_cpos_for_left_leaf(). 2224 */ 2225 static int ocfs2_find_cpos_for_right_leaf(struct super_block *sb, 2226 struct ocfs2_path *path, u32 *cpos) 2227 { 2228 int i, j, ret = 0; 2229 u64 blkno; 2230 struct ocfs2_extent_list *el; 2231 2232 *cpos = 0; 2233 2234 if (path->p_tree_depth == 0) 2235 return 0; 2236 2237 blkno = path_leaf_bh(path)->b_blocknr; 2238 2239 /* Start at the tree node just above the leaf and work our way up. */ 2240 i = path->p_tree_depth - 1; 2241 while (i >= 0) { 2242 int next_free; 2243 2244 el = path->p_node[i].el; 2245 2246 /* 2247 * Find the extent record just after the one in our 2248 * path. 2249 */ 2250 next_free = le16_to_cpu(el->l_next_free_rec); 2251 for(j = 0; j < le16_to_cpu(el->l_next_free_rec); j++) { 2252 if (le64_to_cpu(el->l_recs[j].e_blkno) == blkno) { 2253 if (j == (next_free - 1)) { 2254 if (i == 0) { 2255 /* 2256 * We've determined that the 2257 * path specified is already 2258 * the rightmost one - return a 2259 * cpos of zero. 2260 */ 2261 goto out; 2262 } 2263 /* 2264 * The rightmost record points to our 2265 * leaf - we need to travel up the 2266 * tree one level. 2267 */ 2268 goto next_node; 2269 } 2270 2271 *cpos = le32_to_cpu(el->l_recs[j + 1].e_cpos); 2272 goto out; 2273 } 2274 } 2275 2276 /* 2277 * If we got here, we never found a valid node where 2278 * the tree indicated one should be. 2279 */ 2280 ocfs2_error(sb, 2281 "Invalid extent tree at extent block %llu\n", 2282 (unsigned long long)blkno); 2283 ret = -EROFS; 2284 goto out; 2285 2286 next_node: 2287 blkno = path->p_node[i].bh->b_blocknr; 2288 i--; 2289 } 2290 2291 out: 2292 return ret; 2293 } 2294 2295 static int ocfs2_rotate_rightmost_leaf_left(struct inode *inode, 2296 handle_t *handle, 2297 struct buffer_head *bh, 2298 struct ocfs2_extent_list *el) 2299 { 2300 int ret; 2301 2302 if (!ocfs2_is_empty_extent(&el->l_recs[0])) 2303 return 0; 2304 2305 ret = ocfs2_journal_access(handle, inode, bh, 2306 OCFS2_JOURNAL_ACCESS_WRITE); 2307 if (ret) { 2308 mlog_errno(ret); 2309 goto out; 2310 } 2311 2312 ocfs2_remove_empty_extent(el); 2313 2314 ret = ocfs2_journal_dirty(handle, bh); 2315 if (ret) 2316 mlog_errno(ret); 2317 2318 out: 2319 return ret; 2320 } 2321 2322 static int __ocfs2_rotate_tree_left(struct inode *inode, 2323 handle_t *handle, int orig_credits, 2324 struct ocfs2_path *path, 2325 struct ocfs2_cached_dealloc_ctxt *dealloc, 2326 struct ocfs2_path **empty_extent_path) 2327 { 2328 int ret, subtree_root, deleted; 2329 u32 right_cpos; 2330 struct ocfs2_path *left_path = NULL; 2331 struct ocfs2_path *right_path = NULL; 2332 2333 BUG_ON(!ocfs2_is_empty_extent(&(path_leaf_el(path)->l_recs[0]))); 2334 2335 *empty_extent_path = NULL; 2336 2337 ret = ocfs2_find_cpos_for_right_leaf(inode->i_sb, path, 2338 &right_cpos); 2339 if (ret) { 2340 mlog_errno(ret); 2341 goto out; 2342 } 2343 2344 left_path = ocfs2_new_path(path_root_bh(path), 2345 path_root_el(path)); 2346 if (!left_path) { 2347 ret = -ENOMEM; 2348 mlog_errno(ret); 2349 goto out; 2350 } 2351 2352 ocfs2_cp_path(left_path, path); 2353 2354 right_path = ocfs2_new_path(path_root_bh(path), 2355 path_root_el(path)); 2356 if (!right_path) { 2357 ret = -ENOMEM; 2358 mlog_errno(ret); 2359 goto out; 2360 } 2361 2362 while (right_cpos) { 2363 ret = ocfs2_find_path(inode, right_path, right_cpos); 2364 if (ret) { 2365 mlog_errno(ret); 2366 goto out; 2367 } 2368 2369 subtree_root = ocfs2_find_subtree_root(inode, left_path, 2370 right_path); 2371 2372 mlog(0, "Subtree root at index %d (blk %llu, depth %d)\n", 2373 subtree_root, 2374 (unsigned long long) 2375 right_path->p_node[subtree_root].bh->b_blocknr, 2376 right_path->p_tree_depth); 2377 2378 ret = ocfs2_extend_rotate_transaction(handle, subtree_root, 2379 orig_credits, left_path); 2380 if (ret) { 2381 mlog_errno(ret); 2382 goto out; 2383 } 2384 2385 /* 2386 * Caller might still want to make changes to the 2387 * tree root, so re-add it to the journal here. 2388 */ 2389 ret = ocfs2_journal_access(handle, inode, 2390 path_root_bh(left_path), 2391 OCFS2_JOURNAL_ACCESS_WRITE); 2392 if (ret) { 2393 mlog_errno(ret); 2394 goto out; 2395 } 2396 2397 ret = ocfs2_rotate_subtree_left(inode, handle, left_path, 2398 right_path, subtree_root, 2399 dealloc, &deleted); 2400 if (ret == -EAGAIN) { 2401 /* 2402 * The rotation has to temporarily stop due to 2403 * the right subtree having an empty 2404 * extent. Pass it back to the caller for a 2405 * fixup. 2406 */ 2407 *empty_extent_path = right_path; 2408 right_path = NULL; 2409 goto out; 2410 } 2411 if (ret) { 2412 mlog_errno(ret); 2413 goto out; 2414 } 2415 2416 /* 2417 * The subtree rotate might have removed records on 2418 * the rightmost edge. If so, then rotation is 2419 * complete. 2420 */ 2421 if (deleted) 2422 break; 2423 2424 ocfs2_mv_path(left_path, right_path); 2425 2426 ret = ocfs2_find_cpos_for_right_leaf(inode->i_sb, left_path, 2427 &right_cpos); 2428 if (ret) { 2429 mlog_errno(ret); 2430 goto out; 2431 } 2432 } 2433 2434 out: 2435 ocfs2_free_path(right_path); 2436 ocfs2_free_path(left_path); 2437 2438 return ret; 2439 } 2440 2441 static int ocfs2_remove_rightmost_path(struct inode *inode, handle_t *handle, 2442 struct ocfs2_path *path, 2443 struct ocfs2_cached_dealloc_ctxt *dealloc) 2444 { 2445 int ret, subtree_index; 2446 u32 cpos; 2447 struct ocfs2_path *left_path = NULL; 2448 struct ocfs2_dinode *di; 2449 struct ocfs2_extent_block *eb; 2450 struct ocfs2_extent_list *el; 2451 2452 /* 2453 * XXX: This code assumes that the root is an inode, which is 2454 * true for now but may change as tree code gets generic. 2455 */ 2456 di = (struct ocfs2_dinode *)path_root_bh(path)->b_data; 2457 if (!OCFS2_IS_VALID_DINODE(di)) { 2458 ret = -EIO; 2459 ocfs2_error(inode->i_sb, 2460 "Inode %llu has invalid path root", 2461 (unsigned long long)OCFS2_I(inode)->ip_blkno); 2462 goto out; 2463 } 2464 2465 /* 2466 * There's two ways we handle this depending on 2467 * whether path is the only existing one. 2468 */ 2469 ret = ocfs2_extend_rotate_transaction(handle, 0, 2470 handle->h_buffer_credits, 2471 path); 2472 if (ret) { 2473 mlog_errno(ret); 2474 goto out; 2475 } 2476 2477 ret = ocfs2_journal_access_path(inode, handle, path); 2478 if (ret) { 2479 mlog_errno(ret); 2480 goto out; 2481 } 2482 2483 ret = ocfs2_find_cpos_for_left_leaf(inode->i_sb, path, &cpos); 2484 if (ret) { 2485 mlog_errno(ret); 2486 goto out; 2487 } 2488 2489 if (cpos) { 2490 /* 2491 * We have a path to the left of this one - it needs 2492 * an update too. 2493 */ 2494 left_path = ocfs2_new_path(path_root_bh(path), 2495 path_root_el(path)); 2496 if (!left_path) { 2497 ret = -ENOMEM; 2498 mlog_errno(ret); 2499 goto out; 2500 } 2501 2502 ret = ocfs2_find_path(inode, left_path, cpos); 2503 if (ret) { 2504 mlog_errno(ret); 2505 goto out; 2506 } 2507 2508 ret = ocfs2_journal_access_path(inode, handle, left_path); 2509 if (ret) { 2510 mlog_errno(ret); 2511 goto out; 2512 } 2513 2514 subtree_index = ocfs2_find_subtree_root(inode, left_path, path); 2515 2516 ocfs2_unlink_subtree(inode, handle, left_path, path, 2517 subtree_index, dealloc); 2518 ocfs2_update_edge_lengths(inode, handle, left_path); 2519 2520 eb = (struct ocfs2_extent_block *)path_leaf_bh(left_path)->b_data; 2521 di->i_last_eb_blk = eb->h_blkno; 2522 } else { 2523 /* 2524 * 'path' is also the leftmost path which 2525 * means it must be the only one. This gets 2526 * handled differently because we want to 2527 * revert the inode back to having extents 2528 * in-line. 2529 */ 2530 ocfs2_unlink_path(inode, handle, dealloc, path, 1); 2531 2532 el = &di->id2.i_list; 2533 el->l_tree_depth = 0; 2534 el->l_next_free_rec = 0; 2535 memset(&el->l_recs[0], 0, sizeof(struct ocfs2_extent_rec)); 2536 2537 di->i_last_eb_blk = 0; 2538 } 2539 2540 ocfs2_journal_dirty(handle, path_root_bh(path)); 2541 2542 out: 2543 ocfs2_free_path(left_path); 2544 return ret; 2545 } 2546 2547 /* 2548 * Left rotation of btree records. 2549 * 2550 * In many ways, this is (unsurprisingly) the opposite of right 2551 * rotation. We start at some non-rightmost path containing an empty 2552 * extent in the leaf block. The code works its way to the rightmost 2553 * path by rotating records to the left in every subtree. 2554 * 2555 * This is used by any code which reduces the number of extent records 2556 * in a leaf. After removal, an empty record should be placed in the 2557 * leftmost list position. 2558 * 2559 * This won't handle a length update of the rightmost path records if 2560 * the rightmost tree leaf record is removed so the caller is 2561 * responsible for detecting and correcting that. 2562 */ 2563 static int ocfs2_rotate_tree_left(struct inode *inode, handle_t *handle, 2564 struct ocfs2_path *path, 2565 struct ocfs2_cached_dealloc_ctxt *dealloc) 2566 { 2567 int ret, orig_credits = handle->h_buffer_credits; 2568 struct ocfs2_path *tmp_path = NULL, *restart_path = NULL; 2569 struct ocfs2_extent_block *eb; 2570 struct ocfs2_extent_list *el; 2571 2572 el = path_leaf_el(path); 2573 if (!ocfs2_is_empty_extent(&el->l_recs[0])) 2574 return 0; 2575 2576 if (path->p_tree_depth == 0) { 2577 rightmost_no_delete: 2578 /* 2579 * In-inode extents. This is trivially handled, so do 2580 * it up front. 2581 */ 2582 ret = ocfs2_rotate_rightmost_leaf_left(inode, handle, 2583 path_leaf_bh(path), 2584 path_leaf_el(path)); 2585 if (ret) 2586 mlog_errno(ret); 2587 goto out; 2588 } 2589 2590 /* 2591 * Handle rightmost branch now. There's several cases: 2592 * 1) simple rotation leaving records in there. That's trivial. 2593 * 2) rotation requiring a branch delete - there's no more 2594 * records left. Two cases of this: 2595 * a) There are branches to the left. 2596 * b) This is also the leftmost (the only) branch. 2597 * 2598 * 1) is handled via ocfs2_rotate_rightmost_leaf_left() 2599 * 2a) we need the left branch so that we can update it with the unlink 2600 * 2b) we need to bring the inode back to inline extents. 2601 */ 2602 2603 eb = (struct ocfs2_extent_block *)path_leaf_bh(path)->b_data; 2604 el = &eb->h_list; 2605 if (eb->h_next_leaf_blk == 0) { 2606 /* 2607 * This gets a bit tricky if we're going to delete the 2608 * rightmost path. Get the other cases out of the way 2609 * 1st. 2610 */ 2611 if (le16_to_cpu(el->l_next_free_rec) > 1) 2612 goto rightmost_no_delete; 2613 2614 if (le16_to_cpu(el->l_next_free_rec) == 0) { 2615 ret = -EIO; 2616 ocfs2_error(inode->i_sb, 2617 "Inode %llu has empty extent block at %llu", 2618 (unsigned long long)OCFS2_I(inode)->ip_blkno, 2619 (unsigned long long)le64_to_cpu(eb->h_blkno)); 2620 goto out; 2621 } 2622 2623 /* 2624 * XXX: The caller can not trust "path" any more after 2625 * this as it will have been deleted. What do we do? 2626 * 2627 * In theory the rotate-for-merge code will never get 2628 * here because it'll always ask for a rotate in a 2629 * nonempty list. 2630 */ 2631 2632 ret = ocfs2_remove_rightmost_path(inode, handle, path, 2633 dealloc); 2634 if (ret) 2635 mlog_errno(ret); 2636 goto out; 2637 } 2638 2639 /* 2640 * Now we can loop, remembering the path we get from -EAGAIN 2641 * and restarting from there. 2642 */ 2643 try_rotate: 2644 ret = __ocfs2_rotate_tree_left(inode, handle, orig_credits, path, 2645 dealloc, &restart_path); 2646 if (ret && ret != -EAGAIN) { 2647 mlog_errno(ret); 2648 goto out; 2649 } 2650 2651 while (ret == -EAGAIN) { 2652 tmp_path = restart_path; 2653 restart_path = NULL; 2654 2655 ret = __ocfs2_rotate_tree_left(inode, handle, orig_credits, 2656 tmp_path, dealloc, 2657 &restart_path); 2658 if (ret && ret != -EAGAIN) { 2659 mlog_errno(ret); 2660 goto out; 2661 } 2662 2663 ocfs2_free_path(tmp_path); 2664 tmp_path = NULL; 2665 2666 if (ret == 0) 2667 goto try_rotate; 2668 } 2669 2670 out: 2671 ocfs2_free_path(tmp_path); 2672 ocfs2_free_path(restart_path); 2673 return ret; 2674 } 2675 2676 static void ocfs2_cleanup_merge(struct ocfs2_extent_list *el, 2677 int index) 2678 { 2679 struct ocfs2_extent_rec *rec = &el->l_recs[index]; 2680 unsigned int size; 2681 2682 if (rec->e_leaf_clusters == 0) { 2683 /* 2684 * We consumed all of the merged-from record. An empty 2685 * extent cannot exist anywhere but the 1st array 2686 * position, so move things over if the merged-from 2687 * record doesn't occupy that position. 2688 * 2689 * This creates a new empty extent so the caller 2690 * should be smart enough to have removed any existing 2691 * ones. 2692 */ 2693 if (index > 0) { 2694 BUG_ON(ocfs2_is_empty_extent(&el->l_recs[0])); 2695 size = index * sizeof(struct ocfs2_extent_rec); 2696 memmove(&el->l_recs[1], &el->l_recs[0], size); 2697 } 2698 2699 /* 2700 * Always memset - the caller doesn't check whether it 2701 * created an empty extent, so there could be junk in 2702 * the other fields. 2703 */ 2704 memset(&el->l_recs[0], 0, sizeof(struct ocfs2_extent_rec)); 2705 } 2706 } 2707 2708 static int ocfs2_get_right_path(struct inode *inode, 2709 struct ocfs2_path *left_path, 2710 struct ocfs2_path **ret_right_path) 2711 { 2712 int ret; 2713 u32 right_cpos; 2714 struct ocfs2_path *right_path = NULL; 2715 struct ocfs2_extent_list *left_el; 2716 2717 *ret_right_path = NULL; 2718 2719 /* This function shouldn't be called for non-trees. */ 2720 BUG_ON(left_path->p_tree_depth == 0); 2721 2722 left_el = path_leaf_el(left_path); 2723 BUG_ON(left_el->l_next_free_rec != left_el->l_count); 2724 2725 ret = ocfs2_find_cpos_for_right_leaf(inode->i_sb, left_path, 2726 &right_cpos); 2727 if (ret) { 2728 mlog_errno(ret); 2729 goto out; 2730 } 2731 2732 /* This function shouldn't be called for the rightmost leaf. */ 2733 BUG_ON(right_cpos == 0); 2734 2735 right_path = ocfs2_new_path(path_root_bh(left_path), 2736 path_root_el(left_path)); 2737 if (!right_path) { 2738 ret = -ENOMEM; 2739 mlog_errno(ret); 2740 goto out; 2741 } 2742 2743 ret = ocfs2_find_path(inode, right_path, right_cpos); 2744 if (ret) { 2745 mlog_errno(ret); 2746 goto out; 2747 } 2748 2749 *ret_right_path = right_path; 2750 out: 2751 if (ret) 2752 ocfs2_free_path(right_path); 2753 return ret; 2754 } 2755 2756 /* 2757 * Remove split_rec clusters from the record at index and merge them 2758 * onto the beginning of the record "next" to it. 2759 * For index < l_count - 1, the next means the extent rec at index + 1. 2760 * For index == l_count - 1, the "next" means the 1st extent rec of the 2761 * next extent block. 2762 */ 2763 static int ocfs2_merge_rec_right(struct inode *inode, 2764 struct ocfs2_path *left_path, 2765 handle_t *handle, 2766 struct ocfs2_extent_rec *split_rec, 2767 int index) 2768 { 2769 int ret, next_free, i; 2770 unsigned int split_clusters = le16_to_cpu(split_rec->e_leaf_clusters); 2771 struct ocfs2_extent_rec *left_rec; 2772 struct ocfs2_extent_rec *right_rec; 2773 struct ocfs2_extent_list *right_el; 2774 struct ocfs2_path *right_path = NULL; 2775 int subtree_index = 0; 2776 struct ocfs2_extent_list *el = path_leaf_el(left_path); 2777 struct buffer_head *bh = path_leaf_bh(left_path); 2778 struct buffer_head *root_bh = NULL; 2779 2780 BUG_ON(index >= le16_to_cpu(el->l_next_free_rec)); 2781 left_rec = &el->l_recs[index]; 2782 2783 if (index == le16_to_cpu(el->l_next_free_rec) - 1 && 2784 le16_to_cpu(el->l_next_free_rec) == le16_to_cpu(el->l_count)) { 2785 /* we meet with a cross extent block merge. */ 2786 ret = ocfs2_get_right_path(inode, left_path, &right_path); 2787 if (ret) { 2788 mlog_errno(ret); 2789 goto out; 2790 } 2791 2792 right_el = path_leaf_el(right_path); 2793 next_free = le16_to_cpu(right_el->l_next_free_rec); 2794 BUG_ON(next_free <= 0); 2795 right_rec = &right_el->l_recs[0]; 2796 if (ocfs2_is_empty_extent(right_rec)) { 2797 BUG_ON(next_free <= 1); 2798 right_rec = &right_el->l_recs[1]; 2799 } 2800 2801 BUG_ON(le32_to_cpu(left_rec->e_cpos) + 2802 le16_to_cpu(left_rec->e_leaf_clusters) != 2803 le32_to_cpu(right_rec->e_cpos)); 2804 2805 subtree_index = ocfs2_find_subtree_root(inode, 2806 left_path, right_path); 2807 2808 ret = ocfs2_extend_rotate_transaction(handle, subtree_index, 2809 handle->h_buffer_credits, 2810 right_path); 2811 if (ret) { 2812 mlog_errno(ret); 2813 goto out; 2814 } 2815 2816 root_bh = left_path->p_node[subtree_index].bh; 2817 BUG_ON(root_bh != right_path->p_node[subtree_index].bh); 2818 2819 ret = ocfs2_journal_access(handle, inode, root_bh, 2820 OCFS2_JOURNAL_ACCESS_WRITE); 2821 if (ret) { 2822 mlog_errno(ret); 2823 goto out; 2824 } 2825 2826 for (i = subtree_index + 1; 2827 i < path_num_items(right_path); i++) { 2828 ret = ocfs2_journal_access(handle, inode, 2829 right_path->p_node[i].bh, 2830 OCFS2_JOURNAL_ACCESS_WRITE); 2831 if (ret) { 2832 mlog_errno(ret); 2833 goto out; 2834 } 2835 2836 ret = ocfs2_journal_access(handle, inode, 2837 left_path->p_node[i].bh, 2838 OCFS2_JOURNAL_ACCESS_WRITE); 2839 if (ret) { 2840 mlog_errno(ret); 2841 goto out; 2842 } 2843 } 2844 2845 } else { 2846 BUG_ON(index == le16_to_cpu(el->l_next_free_rec) - 1); 2847 right_rec = &el->l_recs[index + 1]; 2848 } 2849 2850 ret = ocfs2_journal_access(handle, inode, bh, 2851 OCFS2_JOURNAL_ACCESS_WRITE); 2852 if (ret) { 2853 mlog_errno(ret); 2854 goto out; 2855 } 2856 2857 le16_add_cpu(&left_rec->e_leaf_clusters, -split_clusters); 2858 2859 le32_add_cpu(&right_rec->e_cpos, -split_clusters); 2860 le64_add_cpu(&right_rec->e_blkno, 2861 -ocfs2_clusters_to_blocks(inode->i_sb, split_clusters)); 2862 le16_add_cpu(&right_rec->e_leaf_clusters, split_clusters); 2863 2864 ocfs2_cleanup_merge(el, index); 2865 2866 ret = ocfs2_journal_dirty(handle, bh); 2867 if (ret) 2868 mlog_errno(ret); 2869 2870 if (right_path) { 2871 ret = ocfs2_journal_dirty(handle, path_leaf_bh(right_path)); 2872 if (ret) 2873 mlog_errno(ret); 2874 2875 ocfs2_complete_edge_insert(inode, handle, left_path, 2876 right_path, subtree_index); 2877 } 2878 out: 2879 if (right_path) 2880 ocfs2_free_path(right_path); 2881 return ret; 2882 } 2883 2884 static int ocfs2_get_left_path(struct inode *inode, 2885 struct ocfs2_path *right_path, 2886 struct ocfs2_path **ret_left_path) 2887 { 2888 int ret; 2889 u32 left_cpos; 2890 struct ocfs2_path *left_path = NULL; 2891 2892 *ret_left_path = NULL; 2893 2894 /* This function shouldn't be called for non-trees. */ 2895 BUG_ON(right_path->p_tree_depth == 0); 2896 2897 ret = ocfs2_find_cpos_for_left_leaf(inode->i_sb, 2898 right_path, &left_cpos); 2899 if (ret) { 2900 mlog_errno(ret); 2901 goto out; 2902 } 2903 2904 /* This function shouldn't be called for the leftmost leaf. */ 2905 BUG_ON(left_cpos == 0); 2906 2907 left_path = ocfs2_new_path(path_root_bh(right_path), 2908 path_root_el(right_path)); 2909 if (!left_path) { 2910 ret = -ENOMEM; 2911 mlog_errno(ret); 2912 goto out; 2913 } 2914 2915 ret = ocfs2_find_path(inode, left_path, left_cpos); 2916 if (ret) { 2917 mlog_errno(ret); 2918 goto out; 2919 } 2920 2921 *ret_left_path = left_path; 2922 out: 2923 if (ret) 2924 ocfs2_free_path(left_path); 2925 return ret; 2926 } 2927 2928 /* 2929 * Remove split_rec clusters from the record at index and merge them 2930 * onto the tail of the record "before" it. 2931 * For index > 0, the "before" means the extent rec at index - 1. 2932 * 2933 * For index == 0, the "before" means the last record of the previous 2934 * extent block. And there is also a situation that we may need to 2935 * remove the rightmost leaf extent block in the right_path and change 2936 * the right path to indicate the new rightmost path. 2937 */ 2938 static int ocfs2_merge_rec_left(struct inode *inode, 2939 struct ocfs2_path *right_path, 2940 handle_t *handle, 2941 struct ocfs2_extent_rec *split_rec, 2942 struct ocfs2_cached_dealloc_ctxt *dealloc, 2943 int index) 2944 { 2945 int ret, i, subtree_index = 0, has_empty_extent = 0; 2946 unsigned int split_clusters = le16_to_cpu(split_rec->e_leaf_clusters); 2947 struct ocfs2_extent_rec *left_rec; 2948 struct ocfs2_extent_rec *right_rec; 2949 struct ocfs2_extent_list *el = path_leaf_el(right_path); 2950 struct buffer_head *bh = path_leaf_bh(right_path); 2951 struct buffer_head *root_bh = NULL; 2952 struct ocfs2_path *left_path = NULL; 2953 struct ocfs2_extent_list *left_el; 2954 2955 BUG_ON(index < 0); 2956 2957 right_rec = &el->l_recs[index]; 2958 if (index == 0) { 2959 /* we meet with a cross extent block merge. */ 2960 ret = ocfs2_get_left_path(inode, right_path, &left_path); 2961 if (ret) { 2962 mlog_errno(ret); 2963 goto out; 2964 } 2965 2966 left_el = path_leaf_el(left_path); 2967 BUG_ON(le16_to_cpu(left_el->l_next_free_rec) != 2968 le16_to_cpu(left_el->l_count)); 2969 2970 left_rec = &left_el->l_recs[ 2971 le16_to_cpu(left_el->l_next_free_rec) - 1]; 2972 BUG_ON(le32_to_cpu(left_rec->e_cpos) + 2973 le16_to_cpu(left_rec->e_leaf_clusters) != 2974 le32_to_cpu(split_rec->e_cpos)); 2975 2976 subtree_index = ocfs2_find_subtree_root(inode, 2977 left_path, right_path); 2978 2979 ret = ocfs2_extend_rotate_transaction(handle, subtree_index, 2980 handle->h_buffer_credits, 2981 left_path); 2982 if (ret) { 2983 mlog_errno(ret); 2984 goto out; 2985 } 2986 2987 root_bh = left_path->p_node[subtree_index].bh; 2988 BUG_ON(root_bh != right_path->p_node[subtree_index].bh); 2989 2990 ret = ocfs2_journal_access(handle, inode, root_bh, 2991 OCFS2_JOURNAL_ACCESS_WRITE); 2992 if (ret) { 2993 mlog_errno(ret); 2994 goto out; 2995 } 2996 2997 for (i = subtree_index + 1; 2998 i < path_num_items(right_path); i++) { 2999 ret = ocfs2_journal_access(handle, inode, 3000 right_path->p_node[i].bh, 3001 OCFS2_JOURNAL_ACCESS_WRITE); 3002 if (ret) { 3003 mlog_errno(ret); 3004 goto out; 3005 } 3006 3007 ret = ocfs2_journal_access(handle, inode, 3008 left_path->p_node[i].bh, 3009 OCFS2_JOURNAL_ACCESS_WRITE); 3010 if (ret) { 3011 mlog_errno(ret); 3012 goto out; 3013 } 3014 } 3015 } else { 3016 left_rec = &el->l_recs[index - 1]; 3017 if (ocfs2_is_empty_extent(&el->l_recs[0])) 3018 has_empty_extent = 1; 3019 } 3020 3021 ret = ocfs2_journal_access(handle, inode, bh, 3022 OCFS2_JOURNAL_ACCESS_WRITE); 3023 if (ret) { 3024 mlog_errno(ret); 3025 goto out; 3026 } 3027 3028 if (has_empty_extent && index == 1) { 3029 /* 3030 * The easy case - we can just plop the record right in. 3031 */ 3032 *left_rec = *split_rec; 3033 3034 has_empty_extent = 0; 3035 } else 3036 le16_add_cpu(&left_rec->e_leaf_clusters, split_clusters); 3037 3038 le32_add_cpu(&right_rec->e_cpos, split_clusters); 3039 le64_add_cpu(&right_rec->e_blkno, 3040 ocfs2_clusters_to_blocks(inode->i_sb, split_clusters)); 3041 le16_add_cpu(&right_rec->e_leaf_clusters, -split_clusters); 3042 3043 ocfs2_cleanup_merge(el, index); 3044 3045 ret = ocfs2_journal_dirty(handle, bh); 3046 if (ret) 3047 mlog_errno(ret); 3048 3049 if (left_path) { 3050 ret = ocfs2_journal_dirty(handle, path_leaf_bh(left_path)); 3051 if (ret) 3052 mlog_errno(ret); 3053 3054 /* 3055 * In the situation that the right_rec is empty and the extent 3056 * block is empty also, ocfs2_complete_edge_insert can't handle 3057 * it and we need to delete the right extent block. 3058 */ 3059 if (le16_to_cpu(right_rec->e_leaf_clusters) == 0 && 3060 le16_to_cpu(el->l_next_free_rec) == 1) { 3061 3062 ret = ocfs2_remove_rightmost_path(inode, handle, 3063 right_path, dealloc); 3064 if (ret) { 3065 mlog_errno(ret); 3066 goto out; 3067 } 3068 3069 /* Now the rightmost extent block has been deleted. 3070 * So we use the new rightmost path. 3071 */ 3072 ocfs2_mv_path(right_path, left_path); 3073 left_path = NULL; 3074 } else 3075 ocfs2_complete_edge_insert(inode, handle, left_path, 3076 right_path, subtree_index); 3077 } 3078 out: 3079 if (left_path) 3080 ocfs2_free_path(left_path); 3081 return ret; 3082 } 3083 3084 static int ocfs2_try_to_merge_extent(struct inode *inode, 3085 handle_t *handle, 3086 struct ocfs2_path *path, 3087 int split_index, 3088 struct ocfs2_extent_rec *split_rec, 3089 struct ocfs2_cached_dealloc_ctxt *dealloc, 3090 struct ocfs2_merge_ctxt *ctxt) 3091 3092 { 3093 int ret = 0; 3094 struct ocfs2_extent_list *el = path_leaf_el(path); 3095 struct ocfs2_extent_rec *rec = &el->l_recs[split_index]; 3096 3097 BUG_ON(ctxt->c_contig_type == CONTIG_NONE); 3098 3099 if (ctxt->c_split_covers_rec && ctxt->c_has_empty_extent) { 3100 /* 3101 * The merge code will need to create an empty 3102 * extent to take the place of the newly 3103 * emptied slot. Remove any pre-existing empty 3104 * extents - having more than one in a leaf is 3105 * illegal. 3106 */ 3107 ret = ocfs2_rotate_tree_left(inode, handle, path, 3108 dealloc); 3109 if (ret) { 3110 mlog_errno(ret); 3111 goto out; 3112 } 3113 split_index--; 3114 rec = &el->l_recs[split_index]; 3115 } 3116 3117 if (ctxt->c_contig_type == CONTIG_LEFTRIGHT) { 3118 /* 3119 * Left-right contig implies this. 3120 */ 3121 BUG_ON(!ctxt->c_split_covers_rec); 3122 3123 /* 3124 * Since the leftright insert always covers the entire 3125 * extent, this call will delete the insert record 3126 * entirely, resulting in an empty extent record added to 3127 * the extent block. 3128 * 3129 * Since the adding of an empty extent shifts 3130 * everything back to the right, there's no need to 3131 * update split_index here. 3132 * 3133 * When the split_index is zero, we need to merge it to the 3134 * prevoius extent block. It is more efficient and easier 3135 * if we do merge_right first and merge_left later. 3136 */ 3137 ret = ocfs2_merge_rec_right(inode, path, 3138 handle, split_rec, 3139 split_index); 3140 if (ret) { 3141 mlog_errno(ret); 3142 goto out; 3143 } 3144 3145 /* 3146 * We can only get this from logic error above. 3147 */ 3148 BUG_ON(!ocfs2_is_empty_extent(&el->l_recs[0])); 3149 3150 /* The merge left us with an empty extent, remove it. */ 3151 ret = ocfs2_rotate_tree_left(inode, handle, path, dealloc); 3152 if (ret) { 3153 mlog_errno(ret); 3154 goto out; 3155 } 3156 3157 rec = &el->l_recs[split_index]; 3158 3159 /* 3160 * Note that we don't pass split_rec here on purpose - 3161 * we've merged it into the rec already. 3162 */ 3163 ret = ocfs2_merge_rec_left(inode, path, 3164 handle, rec, 3165 dealloc, 3166 split_index); 3167 3168 if (ret) { 3169 mlog_errno(ret); 3170 goto out; 3171 } 3172 3173 ret = ocfs2_rotate_tree_left(inode, handle, path, 3174 dealloc); 3175 /* 3176 * Error from this last rotate is not critical, so 3177 * print but don't bubble it up. 3178 */ 3179 if (ret) 3180 mlog_errno(ret); 3181 ret = 0; 3182 } else { 3183 /* 3184 * Merge a record to the left or right. 3185 * 3186 * 'contig_type' is relative to the existing record, 3187 * so for example, if we're "right contig", it's to 3188 * the record on the left (hence the left merge). 3189 */ 3190 if (ctxt->c_contig_type == CONTIG_RIGHT) { 3191 ret = ocfs2_merge_rec_left(inode, 3192 path, 3193 handle, split_rec, 3194 dealloc, 3195 split_index); 3196 if (ret) { 3197 mlog_errno(ret); 3198 goto out; 3199 } 3200 } else { 3201 ret = ocfs2_merge_rec_right(inode, 3202 path, 3203 handle, split_rec, 3204 split_index); 3205 if (ret) { 3206 mlog_errno(ret); 3207 goto out; 3208 } 3209 } 3210 3211 if (ctxt->c_split_covers_rec) { 3212 /* 3213 * The merge may have left an empty extent in 3214 * our leaf. Try to rotate it away. 3215 */ 3216 ret = ocfs2_rotate_tree_left(inode, handle, path, 3217 dealloc); 3218 if (ret) 3219 mlog_errno(ret); 3220 ret = 0; 3221 } 3222 } 3223 3224 out: 3225 return ret; 3226 } 3227 3228 static void ocfs2_subtract_from_rec(struct super_block *sb, 3229 enum ocfs2_split_type split, 3230 struct ocfs2_extent_rec *rec, 3231 struct ocfs2_extent_rec *split_rec) 3232 { 3233 u64 len_blocks; 3234 3235 len_blocks = ocfs2_clusters_to_blocks(sb, 3236 le16_to_cpu(split_rec->e_leaf_clusters)); 3237 3238 if (split == SPLIT_LEFT) { 3239 /* 3240 * Region is on the left edge of the existing 3241 * record. 3242 */ 3243 le32_add_cpu(&rec->e_cpos, 3244 le16_to_cpu(split_rec->e_leaf_clusters)); 3245 le64_add_cpu(&rec->e_blkno, len_blocks); 3246 le16_add_cpu(&rec->e_leaf_clusters, 3247 -le16_to_cpu(split_rec->e_leaf_clusters)); 3248 } else { 3249 /* 3250 * Region is on the right edge of the existing 3251 * record. 3252 */ 3253 le16_add_cpu(&rec->e_leaf_clusters, 3254 -le16_to_cpu(split_rec->e_leaf_clusters)); 3255 } 3256 } 3257 3258 /* 3259 * Do the final bits of extent record insertion at the target leaf 3260 * list. If this leaf is part of an allocation tree, it is assumed 3261 * that the tree above has been prepared. 3262 */ 3263 static void ocfs2_insert_at_leaf(struct ocfs2_extent_rec *insert_rec, 3264 struct ocfs2_extent_list *el, 3265 struct ocfs2_insert_type *insert, 3266 struct inode *inode) 3267 { 3268 int i = insert->ins_contig_index; 3269 unsigned int range; 3270 struct ocfs2_extent_rec *rec; 3271 3272 BUG_ON(le16_to_cpu(el->l_tree_depth) != 0); 3273 3274 if (insert->ins_split != SPLIT_NONE) { 3275 i = ocfs2_search_extent_list(el, le32_to_cpu(insert_rec->e_cpos)); 3276 BUG_ON(i == -1); 3277 rec = &el->l_recs[i]; 3278 ocfs2_subtract_from_rec(inode->i_sb, insert->ins_split, rec, 3279 insert_rec); 3280 goto rotate; 3281 } 3282 3283 /* 3284 * Contiguous insert - either left or right. 3285 */ 3286 if (insert->ins_contig != CONTIG_NONE) { 3287 rec = &el->l_recs[i]; 3288 if (insert->ins_contig == CONTIG_LEFT) { 3289 rec->e_blkno = insert_rec->e_blkno; 3290 rec->e_cpos = insert_rec->e_cpos; 3291 } 3292 le16_add_cpu(&rec->e_leaf_clusters, 3293 le16_to_cpu(insert_rec->e_leaf_clusters)); 3294 return; 3295 } 3296 3297 /* 3298 * Handle insert into an empty leaf. 3299 */ 3300 if (le16_to_cpu(el->l_next_free_rec) == 0 || 3301 ((le16_to_cpu(el->l_next_free_rec) == 1) && 3302 ocfs2_is_empty_extent(&el->l_recs[0]))) { 3303 el->l_recs[0] = *insert_rec; 3304 el->l_next_free_rec = cpu_to_le16(1); 3305 return; 3306 } 3307 3308 /* 3309 * Appending insert. 3310 */ 3311 if (insert->ins_appending == APPEND_TAIL) { 3312 i = le16_to_cpu(el->l_next_free_rec) - 1; 3313 rec = &el->l_recs[i]; 3314 range = le32_to_cpu(rec->e_cpos) 3315 + le16_to_cpu(rec->e_leaf_clusters); 3316 BUG_ON(le32_to_cpu(insert_rec->e_cpos) < range); 3317 3318 mlog_bug_on_msg(le16_to_cpu(el->l_next_free_rec) >= 3319 le16_to_cpu(el->l_count), 3320 "inode %lu, depth %u, count %u, next free %u, " 3321 "rec.cpos %u, rec.clusters %u, " 3322 "insert.cpos %u, insert.clusters %u\n", 3323 inode->i_ino, 3324 le16_to_cpu(el->l_tree_depth), 3325 le16_to_cpu(el->l_count), 3326 le16_to_cpu(el->l_next_free_rec), 3327 le32_to_cpu(el->l_recs[i].e_cpos), 3328 le16_to_cpu(el->l_recs[i].e_leaf_clusters), 3329 le32_to_cpu(insert_rec->e_cpos), 3330 le16_to_cpu(insert_rec->e_leaf_clusters)); 3331 i++; 3332 el->l_recs[i] = *insert_rec; 3333 le16_add_cpu(&el->l_next_free_rec, 1); 3334 return; 3335 } 3336 3337 rotate: 3338 /* 3339 * Ok, we have to rotate. 3340 * 3341 * At this point, it is safe to assume that inserting into an 3342 * empty leaf and appending to a leaf have both been handled 3343 * above. 3344 * 3345 * This leaf needs to have space, either by the empty 1st 3346 * extent record, or by virtue of an l_next_rec < l_count. 3347 */ 3348 ocfs2_rotate_leaf(el, insert_rec); 3349 } 3350 3351 static inline void ocfs2_update_dinode_clusters(struct inode *inode, 3352 struct ocfs2_dinode *di, 3353 u32 clusters) 3354 { 3355 le32_add_cpu(&di->i_clusters, clusters); 3356 spin_lock(&OCFS2_I(inode)->ip_lock); 3357 OCFS2_I(inode)->ip_clusters = le32_to_cpu(di->i_clusters); 3358 spin_unlock(&OCFS2_I(inode)->ip_lock); 3359 } 3360 3361 static void ocfs2_adjust_rightmost_records(struct inode *inode, 3362 handle_t *handle, 3363 struct ocfs2_path *path, 3364 struct ocfs2_extent_rec *insert_rec) 3365 { 3366 int ret, i, next_free; 3367 struct buffer_head *bh; 3368 struct ocfs2_extent_list *el; 3369 struct ocfs2_extent_rec *rec; 3370 3371 /* 3372 * Update everything except the leaf block. 3373 */ 3374 for (i = 0; i < path->p_tree_depth; i++) { 3375 bh = path->p_node[i].bh; 3376 el = path->p_node[i].el; 3377 3378 next_free = le16_to_cpu(el->l_next_free_rec); 3379 if (next_free == 0) { 3380 ocfs2_error(inode->i_sb, 3381 "Dinode %llu has a bad extent list", 3382 (unsigned long long)OCFS2_I(inode)->ip_blkno); 3383 ret = -EIO; 3384 return; 3385 } 3386 3387 rec = &el->l_recs[next_free - 1]; 3388 3389 rec->e_int_clusters = insert_rec->e_cpos; 3390 le32_add_cpu(&rec->e_int_clusters, 3391 le16_to_cpu(insert_rec->e_leaf_clusters)); 3392 le32_add_cpu(&rec->e_int_clusters, 3393 -le32_to_cpu(rec->e_cpos)); 3394 3395 ret = ocfs2_journal_dirty(handle, bh); 3396 if (ret) 3397 mlog_errno(ret); 3398 3399 } 3400 } 3401 3402 static int ocfs2_append_rec_to_path(struct inode *inode, handle_t *handle, 3403 struct ocfs2_extent_rec *insert_rec, 3404 struct ocfs2_path *right_path, 3405 struct ocfs2_path **ret_left_path) 3406 { 3407 int ret, next_free; 3408 struct ocfs2_extent_list *el; 3409 struct ocfs2_path *left_path = NULL; 3410 3411 *ret_left_path = NULL; 3412 3413 /* 3414 * This shouldn't happen for non-trees. The extent rec cluster 3415 * count manipulation below only works for interior nodes. 3416 */ 3417 BUG_ON(right_path->p_tree_depth == 0); 3418 3419 /* 3420 * If our appending insert is at the leftmost edge of a leaf, 3421 * then we might need to update the rightmost records of the 3422 * neighboring path. 3423 */ 3424 el = path_leaf_el(right_path); 3425 next_free = le16_to_cpu(el->l_next_free_rec); 3426 if (next_free == 0 || 3427 (next_free == 1 && ocfs2_is_empty_extent(&el->l_recs[0]))) { 3428 u32 left_cpos; 3429 3430 ret = ocfs2_find_cpos_for_left_leaf(inode->i_sb, right_path, 3431 &left_cpos); 3432 if (ret) { 3433 mlog_errno(ret); 3434 goto out; 3435 } 3436 3437 mlog(0, "Append may need a left path update. cpos: %u, " 3438 "left_cpos: %u\n", le32_to_cpu(insert_rec->e_cpos), 3439 left_cpos); 3440 3441 /* 3442 * No need to worry if the append is already in the 3443 * leftmost leaf. 3444 */ 3445 if (left_cpos) { 3446 left_path = ocfs2_new_path(path_root_bh(right_path), 3447 path_root_el(right_path)); 3448 if (!left_path) { 3449 ret = -ENOMEM; 3450 mlog_errno(ret); 3451 goto out; 3452 } 3453 3454 ret = ocfs2_find_path(inode, left_path, left_cpos); 3455 if (ret) { 3456 mlog_errno(ret); 3457 goto out; 3458 } 3459 3460 /* 3461 * ocfs2_insert_path() will pass the left_path to the 3462 * journal for us. 3463 */ 3464 } 3465 } 3466 3467 ret = ocfs2_journal_access_path(inode, handle, right_path); 3468 if (ret) { 3469 mlog_errno(ret); 3470 goto out; 3471 } 3472 3473 ocfs2_adjust_rightmost_records(inode, handle, right_path, insert_rec); 3474 3475 *ret_left_path = left_path; 3476 ret = 0; 3477 out: 3478 if (ret != 0) 3479 ocfs2_free_path(left_path); 3480 3481 return ret; 3482 } 3483 3484 static void ocfs2_split_record(struct inode *inode, 3485 struct ocfs2_path *left_path, 3486 struct ocfs2_path *right_path, 3487 struct ocfs2_extent_rec *split_rec, 3488 enum ocfs2_split_type split) 3489 { 3490 int index; 3491 u32 cpos = le32_to_cpu(split_rec->e_cpos); 3492 struct ocfs2_extent_list *left_el = NULL, *right_el, *insert_el, *el; 3493 struct ocfs2_extent_rec *rec, *tmprec; 3494 3495 right_el = path_leaf_el(right_path);; 3496 if (left_path) 3497 left_el = path_leaf_el(left_path); 3498 3499 el = right_el; 3500 insert_el = right_el; 3501 index = ocfs2_search_extent_list(el, cpos); 3502 if (index != -1) { 3503 if (index == 0 && left_path) { 3504 BUG_ON(ocfs2_is_empty_extent(&el->l_recs[0])); 3505 3506 /* 3507 * This typically means that the record 3508 * started in the left path but moved to the 3509 * right as a result of rotation. We either 3510 * move the existing record to the left, or we 3511 * do the later insert there. 3512 * 3513 * In this case, the left path should always 3514 * exist as the rotate code will have passed 3515 * it back for a post-insert update. 3516 */ 3517 3518 if (split == SPLIT_LEFT) { 3519 /* 3520 * It's a left split. Since we know 3521 * that the rotate code gave us an 3522 * empty extent in the left path, we 3523 * can just do the insert there. 3524 */ 3525 insert_el = left_el; 3526 } else { 3527 /* 3528 * Right split - we have to move the 3529 * existing record over to the left 3530 * leaf. The insert will be into the 3531 * newly created empty extent in the 3532 * right leaf. 3533 */ 3534 tmprec = &right_el->l_recs[index]; 3535 ocfs2_rotate_leaf(left_el, tmprec); 3536 el = left_el; 3537 3538 memset(tmprec, 0, sizeof(*tmprec)); 3539 index = ocfs2_search_extent_list(left_el, cpos); 3540 BUG_ON(index == -1); 3541 } 3542 } 3543 } else { 3544 BUG_ON(!left_path); 3545 BUG_ON(!ocfs2_is_empty_extent(&left_el->l_recs[0])); 3546 /* 3547 * Left path is easy - we can just allow the insert to 3548 * happen. 3549 */ 3550 el = left_el; 3551 insert_el = left_el; 3552 index = ocfs2_search_extent_list(el, cpos); 3553 BUG_ON(index == -1); 3554 } 3555 3556 rec = &el->l_recs[index]; 3557 ocfs2_subtract_from_rec(inode->i_sb, split, rec, split_rec); 3558 ocfs2_rotate_leaf(insert_el, split_rec); 3559 } 3560 3561 /* 3562 * This function only does inserts on an allocation b-tree. For dinode 3563 * lists, ocfs2_insert_at_leaf() is called directly. 3564 * 3565 * right_path is the path we want to do the actual insert 3566 * in. left_path should only be passed in if we need to update that 3567 * portion of the tree after an edge insert. 3568 */ 3569 static int ocfs2_insert_path(struct inode *inode, 3570 handle_t *handle, 3571 struct ocfs2_path *left_path, 3572 struct ocfs2_path *right_path, 3573 struct ocfs2_extent_rec *insert_rec, 3574 struct ocfs2_insert_type *insert) 3575 { 3576 int ret, subtree_index; 3577 struct buffer_head *leaf_bh = path_leaf_bh(right_path); 3578 3579 if (left_path) { 3580 int credits = handle->h_buffer_credits; 3581 3582 /* 3583 * There's a chance that left_path got passed back to 3584 * us without being accounted for in the 3585 * journal. Extend our transaction here to be sure we 3586 * can change those blocks. 3587 */ 3588 credits += left_path->p_tree_depth; 3589 3590 ret = ocfs2_extend_trans(handle, credits); 3591 if (ret < 0) { 3592 mlog_errno(ret); 3593 goto out; 3594 } 3595 3596 ret = ocfs2_journal_access_path(inode, handle, left_path); 3597 if (ret < 0) { 3598 mlog_errno(ret); 3599 goto out; 3600 } 3601 } 3602 3603 /* 3604 * Pass both paths to the journal. The majority of inserts 3605 * will be touching all components anyway. 3606 */ 3607 ret = ocfs2_journal_access_path(inode, handle, right_path); 3608 if (ret < 0) { 3609 mlog_errno(ret); 3610 goto out; 3611 } 3612 3613 if (insert->ins_split != SPLIT_NONE) { 3614 /* 3615 * We could call ocfs2_insert_at_leaf() for some types 3616 * of splits, but it's easier to just let one separate 3617 * function sort it all out. 3618 */ 3619 ocfs2_split_record(inode, left_path, right_path, 3620 insert_rec, insert->ins_split); 3621 3622 /* 3623 * Split might have modified either leaf and we don't 3624 * have a guarantee that the later edge insert will 3625 * dirty this for us. 3626 */ 3627 if (left_path) 3628 ret = ocfs2_journal_dirty(handle, 3629 path_leaf_bh(left_path)); 3630 if (ret) 3631 mlog_errno(ret); 3632 } else 3633 ocfs2_insert_at_leaf(insert_rec, path_leaf_el(right_path), 3634 insert, inode); 3635 3636 ret = ocfs2_journal_dirty(handle, leaf_bh); 3637 if (ret) 3638 mlog_errno(ret); 3639 3640 if (left_path) { 3641 /* 3642 * The rotate code has indicated that we need to fix 3643 * up portions of the tree after the insert. 3644 * 3645 * XXX: Should we extend the transaction here? 3646 */ 3647 subtree_index = ocfs2_find_subtree_root(inode, left_path, 3648 right_path); 3649 ocfs2_complete_edge_insert(inode, handle, left_path, 3650 right_path, subtree_index); 3651 } 3652 3653 ret = 0; 3654 out: 3655 return ret; 3656 } 3657 3658 static int ocfs2_do_insert_extent(struct inode *inode, 3659 handle_t *handle, 3660 struct buffer_head *di_bh, 3661 struct ocfs2_extent_rec *insert_rec, 3662 struct ocfs2_insert_type *type) 3663 { 3664 int ret, rotate = 0; 3665 u32 cpos; 3666 struct ocfs2_path *right_path = NULL; 3667 struct ocfs2_path *left_path = NULL; 3668 struct ocfs2_dinode *di; 3669 struct ocfs2_extent_list *el; 3670 3671 di = (struct ocfs2_dinode *) di_bh->b_data; 3672 el = &di->id2.i_list; 3673 3674 ret = ocfs2_journal_access(handle, inode, di_bh, 3675 OCFS2_JOURNAL_ACCESS_WRITE); 3676 if (ret) { 3677 mlog_errno(ret); 3678 goto out; 3679 } 3680 3681 if (le16_to_cpu(el->l_tree_depth) == 0) { 3682 ocfs2_insert_at_leaf(insert_rec, el, type, inode); 3683 goto out_update_clusters; 3684 } 3685 3686 right_path = ocfs2_new_inode_path(di_bh); 3687 if (!right_path) { 3688 ret = -ENOMEM; 3689 mlog_errno(ret); 3690 goto out; 3691 } 3692 3693 /* 3694 * Determine the path to start with. Rotations need the 3695 * rightmost path, everything else can go directly to the 3696 * target leaf. 3697 */ 3698 cpos = le32_to_cpu(insert_rec->e_cpos); 3699 if (type->ins_appending == APPEND_NONE && 3700 type->ins_contig == CONTIG_NONE) { 3701 rotate = 1; 3702 cpos = UINT_MAX; 3703 } 3704 3705 ret = ocfs2_find_path(inode, right_path, cpos); 3706 if (ret) { 3707 mlog_errno(ret); 3708 goto out; 3709 } 3710 3711 /* 3712 * Rotations and appends need special treatment - they modify 3713 * parts of the tree's above them. 3714 * 3715 * Both might pass back a path immediate to the left of the 3716 * one being inserted to. This will be cause 3717 * ocfs2_insert_path() to modify the rightmost records of 3718 * left_path to account for an edge insert. 3719 * 3720 * XXX: When modifying this code, keep in mind that an insert 3721 * can wind up skipping both of these two special cases... 3722 */ 3723 if (rotate) { 3724 ret = ocfs2_rotate_tree_right(inode, handle, type->ins_split, 3725 le32_to_cpu(insert_rec->e_cpos), 3726 right_path, &left_path); 3727 if (ret) { 3728 mlog_errno(ret); 3729 goto out; 3730 } 3731 3732 /* 3733 * ocfs2_rotate_tree_right() might have extended the 3734 * transaction without re-journaling our tree root. 3735 */ 3736 ret = ocfs2_journal_access(handle, inode, di_bh, 3737 OCFS2_JOURNAL_ACCESS_WRITE); 3738 if (ret) { 3739 mlog_errno(ret); 3740 goto out; 3741 } 3742 } else if (type->ins_appending == APPEND_TAIL 3743 && type->ins_contig != CONTIG_LEFT) { 3744 ret = ocfs2_append_rec_to_path(inode, handle, insert_rec, 3745 right_path, &left_path); 3746 if (ret) { 3747 mlog_errno(ret); 3748 goto out; 3749 } 3750 } 3751 3752 ret = ocfs2_insert_path(inode, handle, left_path, right_path, 3753 insert_rec, type); 3754 if (ret) { 3755 mlog_errno(ret); 3756 goto out; 3757 } 3758 3759 out_update_clusters: 3760 if (type->ins_split == SPLIT_NONE) 3761 ocfs2_update_dinode_clusters(inode, di, 3762 le16_to_cpu(insert_rec->e_leaf_clusters)); 3763 3764 ret = ocfs2_journal_dirty(handle, di_bh); 3765 if (ret) 3766 mlog_errno(ret); 3767 3768 out: 3769 ocfs2_free_path(left_path); 3770 ocfs2_free_path(right_path); 3771 3772 return ret; 3773 } 3774 3775 static enum ocfs2_contig_type 3776 ocfs2_figure_merge_contig_type(struct inode *inode, struct ocfs2_path *path, 3777 struct ocfs2_extent_list *el, int index, 3778 struct ocfs2_extent_rec *split_rec) 3779 { 3780 int status; 3781 enum ocfs2_contig_type ret = CONTIG_NONE; 3782 u32 left_cpos, right_cpos; 3783 struct ocfs2_extent_rec *rec = NULL; 3784 struct ocfs2_extent_list *new_el; 3785 struct ocfs2_path *left_path = NULL, *right_path = NULL; 3786 struct buffer_head *bh; 3787 struct ocfs2_extent_block *eb; 3788 3789 if (index > 0) { 3790 rec = &el->l_recs[index - 1]; 3791 } else if (path->p_tree_depth > 0) { 3792 status = ocfs2_find_cpos_for_left_leaf(inode->i_sb, 3793 path, &left_cpos); 3794 if (status) 3795 goto out; 3796 3797 if (left_cpos != 0) { 3798 left_path = ocfs2_new_path(path_root_bh(path), 3799 path_root_el(path)); 3800 if (!left_path) 3801 goto out; 3802 3803 status = ocfs2_find_path(inode, left_path, left_cpos); 3804 if (status) 3805 goto out; 3806 3807 new_el = path_leaf_el(left_path); 3808 3809 if (le16_to_cpu(new_el->l_next_free_rec) != 3810 le16_to_cpu(new_el->l_count)) { 3811 bh = path_leaf_bh(left_path); 3812 eb = (struct ocfs2_extent_block *)bh->b_data; 3813 OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, 3814 eb); 3815 goto out; 3816 } 3817 rec = &new_el->l_recs[ 3818 le16_to_cpu(new_el->l_next_free_rec) - 1]; 3819 } 3820 } 3821 3822 /* 3823 * We're careful to check for an empty extent record here - 3824 * the merge code will know what to do if it sees one. 3825 */ 3826 if (rec) { 3827 if (index == 1 && ocfs2_is_empty_extent(rec)) { 3828 if (split_rec->e_cpos == el->l_recs[index].e_cpos) 3829 ret = CONTIG_RIGHT; 3830 } else { 3831 ret = ocfs2_extent_contig(inode, rec, split_rec); 3832 } 3833 } 3834 3835 rec = NULL; 3836 if (index < (le16_to_cpu(el->l_next_free_rec) - 1)) 3837 rec = &el->l_recs[index + 1]; 3838 else if (le16_to_cpu(el->l_next_free_rec) == le16_to_cpu(el->l_count) && 3839 path->p_tree_depth > 0) { 3840 status = ocfs2_find_cpos_for_right_leaf(inode->i_sb, 3841 path, &right_cpos); 3842 if (status) 3843 goto out; 3844 3845 if (right_cpos == 0) 3846 goto out; 3847 3848 right_path = ocfs2_new_path(path_root_bh(path), 3849 path_root_el(path)); 3850 if (!right_path) 3851 goto out; 3852 3853 status = ocfs2_find_path(inode, right_path, right_cpos); 3854 if (status) 3855 goto out; 3856 3857 new_el = path_leaf_el(right_path); 3858 rec = &new_el->l_recs[0]; 3859 if (ocfs2_is_empty_extent(rec)) { 3860 if (le16_to_cpu(new_el->l_next_free_rec) <= 1) { 3861 bh = path_leaf_bh(right_path); 3862 eb = (struct ocfs2_extent_block *)bh->b_data; 3863 OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, 3864 eb); 3865 goto out; 3866 } 3867 rec = &new_el->l_recs[1]; 3868 } 3869 } 3870 3871 if (rec) { 3872 enum ocfs2_contig_type contig_type; 3873 3874 contig_type = ocfs2_extent_contig(inode, rec, split_rec); 3875 3876 if (contig_type == CONTIG_LEFT && ret == CONTIG_RIGHT) 3877 ret = CONTIG_LEFTRIGHT; 3878 else if (ret == CONTIG_NONE) 3879 ret = contig_type; 3880 } 3881 3882 out: 3883 if (left_path) 3884 ocfs2_free_path(left_path); 3885 if (right_path) 3886 ocfs2_free_path(right_path); 3887 3888 return ret; 3889 } 3890 3891 static void ocfs2_figure_contig_type(struct inode *inode, 3892 struct ocfs2_insert_type *insert, 3893 struct ocfs2_extent_list *el, 3894 struct ocfs2_extent_rec *insert_rec) 3895 { 3896 int i; 3897 enum ocfs2_contig_type contig_type = CONTIG_NONE; 3898 3899 BUG_ON(le16_to_cpu(el->l_tree_depth) != 0); 3900 3901 for(i = 0; i < le16_to_cpu(el->l_next_free_rec); i++) { 3902 contig_type = ocfs2_extent_contig(inode, &el->l_recs[i], 3903 insert_rec); 3904 if (contig_type != CONTIG_NONE) { 3905 insert->ins_contig_index = i; 3906 break; 3907 } 3908 } 3909 insert->ins_contig = contig_type; 3910 } 3911 3912 /* 3913 * This should only be called against the righmost leaf extent list. 3914 * 3915 * ocfs2_figure_appending_type() will figure out whether we'll have to 3916 * insert at the tail of the rightmost leaf. 3917 * 3918 * This should also work against the dinode list for tree's with 0 3919 * depth. If we consider the dinode list to be the rightmost leaf node 3920 * then the logic here makes sense. 3921 */ 3922 static void ocfs2_figure_appending_type(struct ocfs2_insert_type *insert, 3923 struct ocfs2_extent_list *el, 3924 struct ocfs2_extent_rec *insert_rec) 3925 { 3926 int i; 3927 u32 cpos = le32_to_cpu(insert_rec->e_cpos); 3928 struct ocfs2_extent_rec *rec; 3929 3930 insert->ins_appending = APPEND_NONE; 3931 3932 BUG_ON(le16_to_cpu(el->l_tree_depth) != 0); 3933 3934 if (!el->l_next_free_rec) 3935 goto set_tail_append; 3936 3937 if (ocfs2_is_empty_extent(&el->l_recs[0])) { 3938 /* Were all records empty? */ 3939 if (le16_to_cpu(el->l_next_free_rec) == 1) 3940 goto set_tail_append; 3941 } 3942 3943 i = le16_to_cpu(el->l_next_free_rec) - 1; 3944 rec = &el->l_recs[i]; 3945 3946 if (cpos >= 3947 (le32_to_cpu(rec->e_cpos) + le16_to_cpu(rec->e_leaf_clusters))) 3948 goto set_tail_append; 3949 3950 return; 3951 3952 set_tail_append: 3953 insert->ins_appending = APPEND_TAIL; 3954 } 3955 3956 /* 3957 * Helper function called at the begining of an insert. 3958 * 3959 * This computes a few things that are commonly used in the process of 3960 * inserting into the btree: 3961 * - Whether the new extent is contiguous with an existing one. 3962 * - The current tree depth. 3963 * - Whether the insert is an appending one. 3964 * - The total # of free records in the tree. 3965 * 3966 * All of the information is stored on the ocfs2_insert_type 3967 * structure. 3968 */ 3969 static int ocfs2_figure_insert_type(struct inode *inode, 3970 struct buffer_head *di_bh, 3971 struct buffer_head **last_eb_bh, 3972 struct ocfs2_extent_rec *insert_rec, 3973 int *free_records, 3974 struct ocfs2_insert_type *insert) 3975 { 3976 int ret; 3977 struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data; 3978 struct ocfs2_extent_block *eb; 3979 struct ocfs2_extent_list *el; 3980 struct ocfs2_path *path = NULL; 3981 struct buffer_head *bh = NULL; 3982 3983 insert->ins_split = SPLIT_NONE; 3984 3985 el = &di->id2.i_list; 3986 insert->ins_tree_depth = le16_to_cpu(el->l_tree_depth); 3987 3988 if (el->l_tree_depth) { 3989 /* 3990 * If we have tree depth, we read in the 3991 * rightmost extent block ahead of time as 3992 * ocfs2_figure_insert_type() and ocfs2_add_branch() 3993 * may want it later. 3994 */ 3995 ret = ocfs2_read_block(OCFS2_SB(inode->i_sb), 3996 le64_to_cpu(di->i_last_eb_blk), &bh, 3997 OCFS2_BH_CACHED, inode); 3998 if (ret) { 3999 mlog_exit(ret); 4000 goto out; 4001 } 4002 eb = (struct ocfs2_extent_block *) bh->b_data; 4003 el = &eb->h_list; 4004 } 4005 4006 /* 4007 * Unless we have a contiguous insert, we'll need to know if 4008 * there is room left in our allocation tree for another 4009 * extent record. 4010 * 4011 * XXX: This test is simplistic, we can search for empty 4012 * extent records too. 4013 */ 4014 *free_records = le16_to_cpu(el->l_count) - 4015 le16_to_cpu(el->l_next_free_rec); 4016 4017 if (!insert->ins_tree_depth) { 4018 ocfs2_figure_contig_type(inode, insert, el, insert_rec); 4019 ocfs2_figure_appending_type(insert, el, insert_rec); 4020 return 0; 4021 } 4022 4023 path = ocfs2_new_inode_path(di_bh); 4024 if (!path) { 4025 ret = -ENOMEM; 4026 mlog_errno(ret); 4027 goto out; 4028 } 4029 4030 /* 4031 * In the case that we're inserting past what the tree 4032 * currently accounts for, ocfs2_find_path() will return for 4033 * us the rightmost tree path. This is accounted for below in 4034 * the appending code. 4035 */ 4036 ret = ocfs2_find_path(inode, path, le32_to_cpu(insert_rec->e_cpos)); 4037 if (ret) { 4038 mlog_errno(ret); 4039 goto out; 4040 } 4041 4042 el = path_leaf_el(path); 4043 4044 /* 4045 * Now that we have the path, there's two things we want to determine: 4046 * 1) Contiguousness (also set contig_index if this is so) 4047 * 4048 * 2) Are we doing an append? We can trivially break this up 4049 * into two types of appends: simple record append, or a 4050 * rotate inside the tail leaf. 4051 */ 4052 ocfs2_figure_contig_type(inode, insert, el, insert_rec); 4053 4054 /* 4055 * The insert code isn't quite ready to deal with all cases of 4056 * left contiguousness. Specifically, if it's an insert into 4057 * the 1st record in a leaf, it will require the adjustment of 4058 * cluster count on the last record of the path directly to it's 4059 * left. For now, just catch that case and fool the layers 4060 * above us. This works just fine for tree_depth == 0, which 4061 * is why we allow that above. 4062 */ 4063 if (insert->ins_contig == CONTIG_LEFT && 4064 insert->ins_contig_index == 0) 4065 insert->ins_contig = CONTIG_NONE; 4066 4067 /* 4068 * Ok, so we can simply compare against last_eb to figure out 4069 * whether the path doesn't exist. This will only happen in 4070 * the case that we're doing a tail append, so maybe we can 4071 * take advantage of that information somehow. 4072 */ 4073 if (le64_to_cpu(di->i_last_eb_blk) == path_leaf_bh(path)->b_blocknr) { 4074 /* 4075 * Ok, ocfs2_find_path() returned us the rightmost 4076 * tree path. This might be an appending insert. There are 4077 * two cases: 4078 * 1) We're doing a true append at the tail: 4079 * -This might even be off the end of the leaf 4080 * 2) We're "appending" by rotating in the tail 4081 */ 4082 ocfs2_figure_appending_type(insert, el, insert_rec); 4083 } 4084 4085 out: 4086 ocfs2_free_path(path); 4087 4088 if (ret == 0) 4089 *last_eb_bh = bh; 4090 else 4091 brelse(bh); 4092 return ret; 4093 } 4094 4095 /* 4096 * Insert an extent into an inode btree. 4097 * 4098 * The caller needs to update fe->i_clusters 4099 */ 4100 int ocfs2_insert_extent(struct ocfs2_super *osb, 4101 handle_t *handle, 4102 struct inode *inode, 4103 struct buffer_head *fe_bh, 4104 u32 cpos, 4105 u64 start_blk, 4106 u32 new_clusters, 4107 u8 flags, 4108 struct ocfs2_alloc_context *meta_ac) 4109 { 4110 int status; 4111 int uninitialized_var(free_records); 4112 struct buffer_head *last_eb_bh = NULL; 4113 struct ocfs2_insert_type insert = {0, }; 4114 struct ocfs2_extent_rec rec; 4115 4116 BUG_ON(OCFS2_I(inode)->ip_dyn_features & OCFS2_INLINE_DATA_FL); 4117 4118 mlog(0, "add %u clusters at position %u to inode %llu\n", 4119 new_clusters, cpos, (unsigned long long)OCFS2_I(inode)->ip_blkno); 4120 4121 mlog_bug_on_msg(!ocfs2_sparse_alloc(osb) && 4122 (OCFS2_I(inode)->ip_clusters != cpos), 4123 "Device %s, asking for sparse allocation: inode %llu, " 4124 "cpos %u, clusters %u\n", 4125 osb->dev_str, 4126 (unsigned long long)OCFS2_I(inode)->ip_blkno, cpos, 4127 OCFS2_I(inode)->ip_clusters); 4128 4129 memset(&rec, 0, sizeof(rec)); 4130 rec.e_cpos = cpu_to_le32(cpos); 4131 rec.e_blkno = cpu_to_le64(start_blk); 4132 rec.e_leaf_clusters = cpu_to_le16(new_clusters); 4133 rec.e_flags = flags; 4134 4135 status = ocfs2_figure_insert_type(inode, fe_bh, &last_eb_bh, &rec, 4136 &free_records, &insert); 4137 if (status < 0) { 4138 mlog_errno(status); 4139 goto bail; 4140 } 4141 4142 mlog(0, "Insert.appending: %u, Insert.Contig: %u, " 4143 "Insert.contig_index: %d, Insert.free_records: %d, " 4144 "Insert.tree_depth: %d\n", 4145 insert.ins_appending, insert.ins_contig, insert.ins_contig_index, 4146 free_records, insert.ins_tree_depth); 4147 4148 if (insert.ins_contig == CONTIG_NONE && free_records == 0) { 4149 status = ocfs2_grow_tree(inode, handle, fe_bh, 4150 &insert.ins_tree_depth, &last_eb_bh, 4151 meta_ac); 4152 if (status) { 4153 mlog_errno(status); 4154 goto bail; 4155 } 4156 } 4157 4158 /* Finally, we can add clusters. This might rotate the tree for us. */ 4159 status = ocfs2_do_insert_extent(inode, handle, fe_bh, &rec, &insert); 4160 if (status < 0) 4161 mlog_errno(status); 4162 else 4163 ocfs2_extent_map_insert_rec(inode, &rec); 4164 4165 bail: 4166 if (last_eb_bh) 4167 brelse(last_eb_bh); 4168 4169 mlog_exit(status); 4170 return status; 4171 } 4172 4173 static void ocfs2_make_right_split_rec(struct super_block *sb, 4174 struct ocfs2_extent_rec *split_rec, 4175 u32 cpos, 4176 struct ocfs2_extent_rec *rec) 4177 { 4178 u32 rec_cpos = le32_to_cpu(rec->e_cpos); 4179 u32 rec_range = rec_cpos + le16_to_cpu(rec->e_leaf_clusters); 4180 4181 memset(split_rec, 0, sizeof(struct ocfs2_extent_rec)); 4182 4183 split_rec->e_cpos = cpu_to_le32(cpos); 4184 split_rec->e_leaf_clusters = cpu_to_le16(rec_range - cpos); 4185 4186 split_rec->e_blkno = rec->e_blkno; 4187 le64_add_cpu(&split_rec->e_blkno, 4188 ocfs2_clusters_to_blocks(sb, cpos - rec_cpos)); 4189 4190 split_rec->e_flags = rec->e_flags; 4191 } 4192 4193 static int ocfs2_split_and_insert(struct inode *inode, 4194 handle_t *handle, 4195 struct ocfs2_path *path, 4196 struct buffer_head *di_bh, 4197 struct buffer_head **last_eb_bh, 4198 int split_index, 4199 struct ocfs2_extent_rec *orig_split_rec, 4200 struct ocfs2_alloc_context *meta_ac) 4201 { 4202 int ret = 0, depth; 4203 unsigned int insert_range, rec_range, do_leftright = 0; 4204 struct ocfs2_extent_rec tmprec; 4205 struct ocfs2_extent_list *rightmost_el; 4206 struct ocfs2_extent_rec rec; 4207 struct ocfs2_extent_rec split_rec = *orig_split_rec; 4208 struct ocfs2_insert_type insert; 4209 struct ocfs2_extent_block *eb; 4210 struct ocfs2_dinode *di; 4211 4212 leftright: 4213 /* 4214 * Store a copy of the record on the stack - it might move 4215 * around as the tree is manipulated below. 4216 */ 4217 rec = path_leaf_el(path)->l_recs[split_index]; 4218 4219 di = (struct ocfs2_dinode *)di_bh->b_data; 4220 rightmost_el = &di->id2.i_list; 4221 4222 depth = le16_to_cpu(rightmost_el->l_tree_depth); 4223 if (depth) { 4224 BUG_ON(!(*last_eb_bh)); 4225 eb = (struct ocfs2_extent_block *) (*last_eb_bh)->b_data; 4226 rightmost_el = &eb->h_list; 4227 } 4228 4229 if (le16_to_cpu(rightmost_el->l_next_free_rec) == 4230 le16_to_cpu(rightmost_el->l_count)) { 4231 ret = ocfs2_grow_tree(inode, handle, di_bh, &depth, last_eb_bh, 4232 meta_ac); 4233 if (ret) { 4234 mlog_errno(ret); 4235 goto out; 4236 } 4237 } 4238 4239 memset(&insert, 0, sizeof(struct ocfs2_insert_type)); 4240 insert.ins_appending = APPEND_NONE; 4241 insert.ins_contig = CONTIG_NONE; 4242 insert.ins_tree_depth = depth; 4243 4244 insert_range = le32_to_cpu(split_rec.e_cpos) + 4245 le16_to_cpu(split_rec.e_leaf_clusters); 4246 rec_range = le32_to_cpu(rec.e_cpos) + 4247 le16_to_cpu(rec.e_leaf_clusters); 4248 4249 if (split_rec.e_cpos == rec.e_cpos) { 4250 insert.ins_split = SPLIT_LEFT; 4251 } else if (insert_range == rec_range) { 4252 insert.ins_split = SPLIT_RIGHT; 4253 } else { 4254 /* 4255 * Left/right split. We fake this as a right split 4256 * first and then make a second pass as a left split. 4257 */ 4258 insert.ins_split = SPLIT_RIGHT; 4259 4260 ocfs2_make_right_split_rec(inode->i_sb, &tmprec, insert_range, 4261 &rec); 4262 4263 split_rec = tmprec; 4264 4265 BUG_ON(do_leftright); 4266 do_leftright = 1; 4267 } 4268 4269 ret = ocfs2_do_insert_extent(inode, handle, di_bh, &split_rec, 4270 &insert); 4271 if (ret) { 4272 mlog_errno(ret); 4273 goto out; 4274 } 4275 4276 if (do_leftright == 1) { 4277 u32 cpos; 4278 struct ocfs2_extent_list *el; 4279 4280 do_leftright++; 4281 split_rec = *orig_split_rec; 4282 4283 ocfs2_reinit_path(path, 1); 4284 4285 cpos = le32_to_cpu(split_rec.e_cpos); 4286 ret = ocfs2_find_path(inode, path, cpos); 4287 if (ret) { 4288 mlog_errno(ret); 4289 goto out; 4290 } 4291 4292 el = path_leaf_el(path); 4293 split_index = ocfs2_search_extent_list(el, cpos); 4294 goto leftright; 4295 } 4296 out: 4297 4298 return ret; 4299 } 4300 4301 /* 4302 * Mark part or all of the extent record at split_index in the leaf 4303 * pointed to by path as written. This removes the unwritten 4304 * extent flag. 4305 * 4306 * Care is taken to handle contiguousness so as to not grow the tree. 4307 * 4308 * meta_ac is not strictly necessary - we only truly need it if growth 4309 * of the tree is required. All other cases will degrade into a less 4310 * optimal tree layout. 4311 * 4312 * last_eb_bh should be the rightmost leaf block for any inode with a 4313 * btree. Since a split may grow the tree or a merge might shrink it, the caller cannot trust the contents of that buffer after this call. 4314 * 4315 * This code is optimized for readability - several passes might be 4316 * made over certain portions of the tree. All of those blocks will 4317 * have been brought into cache (and pinned via the journal), so the 4318 * extra overhead is not expressed in terms of disk reads. 4319 */ 4320 static int __ocfs2_mark_extent_written(struct inode *inode, 4321 struct buffer_head *di_bh, 4322 handle_t *handle, 4323 struct ocfs2_path *path, 4324 int split_index, 4325 struct ocfs2_extent_rec *split_rec, 4326 struct ocfs2_alloc_context *meta_ac, 4327 struct ocfs2_cached_dealloc_ctxt *dealloc) 4328 { 4329 int ret = 0; 4330 struct ocfs2_extent_list *el = path_leaf_el(path); 4331 struct buffer_head *last_eb_bh = NULL; 4332 struct ocfs2_extent_rec *rec = &el->l_recs[split_index]; 4333 struct ocfs2_merge_ctxt ctxt; 4334 struct ocfs2_extent_list *rightmost_el; 4335 4336 if (!(rec->e_flags & OCFS2_EXT_UNWRITTEN)) { 4337 ret = -EIO; 4338 mlog_errno(ret); 4339 goto out; 4340 } 4341 4342 if (le32_to_cpu(rec->e_cpos) > le32_to_cpu(split_rec->e_cpos) || 4343 ((le32_to_cpu(rec->e_cpos) + le16_to_cpu(rec->e_leaf_clusters)) < 4344 (le32_to_cpu(split_rec->e_cpos) + le16_to_cpu(split_rec->e_leaf_clusters)))) { 4345 ret = -EIO; 4346 mlog_errno(ret); 4347 goto out; 4348 } 4349 4350 ctxt.c_contig_type = ocfs2_figure_merge_contig_type(inode, path, el, 4351 split_index, 4352 split_rec); 4353 4354 /* 4355 * The core merge / split code wants to know how much room is 4356 * left in this inodes allocation tree, so we pass the 4357 * rightmost extent list. 4358 */ 4359 if (path->p_tree_depth) { 4360 struct ocfs2_extent_block *eb; 4361 struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data; 4362 4363 ret = ocfs2_read_block(OCFS2_SB(inode->i_sb), 4364 le64_to_cpu(di->i_last_eb_blk), 4365 &last_eb_bh, OCFS2_BH_CACHED, inode); 4366 if (ret) { 4367 mlog_exit(ret); 4368 goto out; 4369 } 4370 4371 eb = (struct ocfs2_extent_block *) last_eb_bh->b_data; 4372 if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) { 4373 OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb); 4374 ret = -EROFS; 4375 goto out; 4376 } 4377 4378 rightmost_el = &eb->h_list; 4379 } else 4380 rightmost_el = path_root_el(path); 4381 4382 if (rec->e_cpos == split_rec->e_cpos && 4383 rec->e_leaf_clusters == split_rec->e_leaf_clusters) 4384 ctxt.c_split_covers_rec = 1; 4385 else 4386 ctxt.c_split_covers_rec = 0; 4387 4388 ctxt.c_has_empty_extent = ocfs2_is_empty_extent(&el->l_recs[0]); 4389 4390 mlog(0, "index: %d, contig: %u, has_empty: %u, split_covers: %u\n", 4391 split_index, ctxt.c_contig_type, ctxt.c_has_empty_extent, 4392 ctxt.c_split_covers_rec); 4393 4394 if (ctxt.c_contig_type == CONTIG_NONE) { 4395 if (ctxt.c_split_covers_rec) 4396 el->l_recs[split_index] = *split_rec; 4397 else 4398 ret = ocfs2_split_and_insert(inode, handle, path, di_bh, 4399 &last_eb_bh, split_index, 4400 split_rec, meta_ac); 4401 if (ret) 4402 mlog_errno(ret); 4403 } else { 4404 ret = ocfs2_try_to_merge_extent(inode, handle, path, 4405 split_index, split_rec, 4406 dealloc, &ctxt); 4407 if (ret) 4408 mlog_errno(ret); 4409 } 4410 4411 out: 4412 brelse(last_eb_bh); 4413 return ret; 4414 } 4415 4416 /* 4417 * Mark the already-existing extent at cpos as written for len clusters. 4418 * 4419 * If the existing extent is larger than the request, initiate a 4420 * split. An attempt will be made at merging with adjacent extents. 4421 * 4422 * The caller is responsible for passing down meta_ac if we'll need it. 4423 */ 4424 int ocfs2_mark_extent_written(struct inode *inode, struct buffer_head *di_bh, 4425 handle_t *handle, u32 cpos, u32 len, u32 phys, 4426 struct ocfs2_alloc_context *meta_ac, 4427 struct ocfs2_cached_dealloc_ctxt *dealloc) 4428 { 4429 int ret, index; 4430 u64 start_blkno = ocfs2_clusters_to_blocks(inode->i_sb, phys); 4431 struct ocfs2_extent_rec split_rec; 4432 struct ocfs2_path *left_path = NULL; 4433 struct ocfs2_extent_list *el; 4434 4435 mlog(0, "Inode %lu cpos %u, len %u, phys %u (%llu)\n", 4436 inode->i_ino, cpos, len, phys, (unsigned long long)start_blkno); 4437 4438 if (!ocfs2_writes_unwritten_extents(OCFS2_SB(inode->i_sb))) { 4439 ocfs2_error(inode->i_sb, "Inode %llu has unwritten extents " 4440 "that are being written to, but the feature bit " 4441 "is not set in the super block.", 4442 (unsigned long long)OCFS2_I(inode)->ip_blkno); 4443 ret = -EROFS; 4444 goto out; 4445 } 4446 4447 /* 4448 * XXX: This should be fixed up so that we just re-insert the 4449 * next extent records. 4450 */ 4451 ocfs2_extent_map_trunc(inode, 0); 4452 4453 left_path = ocfs2_new_inode_path(di_bh); 4454 if (!left_path) { 4455 ret = -ENOMEM; 4456 mlog_errno(ret); 4457 goto out; 4458 } 4459 4460 ret = ocfs2_find_path(inode, left_path, cpos); 4461 if (ret) { 4462 mlog_errno(ret); 4463 goto out; 4464 } 4465 el = path_leaf_el(left_path); 4466 4467 index = ocfs2_search_extent_list(el, cpos); 4468 if (index == -1 || index >= le16_to_cpu(el->l_next_free_rec)) { 4469 ocfs2_error(inode->i_sb, 4470 "Inode %llu has an extent at cpos %u which can no " 4471 "longer be found.\n", 4472 (unsigned long long)OCFS2_I(inode)->ip_blkno, cpos); 4473 ret = -EROFS; 4474 goto out; 4475 } 4476 4477 memset(&split_rec, 0, sizeof(struct ocfs2_extent_rec)); 4478 split_rec.e_cpos = cpu_to_le32(cpos); 4479 split_rec.e_leaf_clusters = cpu_to_le16(len); 4480 split_rec.e_blkno = cpu_to_le64(start_blkno); 4481 split_rec.e_flags = path_leaf_el(left_path)->l_recs[index].e_flags; 4482 split_rec.e_flags &= ~OCFS2_EXT_UNWRITTEN; 4483 4484 ret = __ocfs2_mark_extent_written(inode, di_bh, handle, left_path, 4485 index, &split_rec, meta_ac, dealloc); 4486 if (ret) 4487 mlog_errno(ret); 4488 4489 out: 4490 ocfs2_free_path(left_path); 4491 return ret; 4492 } 4493 4494 static int ocfs2_split_tree(struct inode *inode, struct buffer_head *di_bh, 4495 handle_t *handle, struct ocfs2_path *path, 4496 int index, u32 new_range, 4497 struct ocfs2_alloc_context *meta_ac) 4498 { 4499 int ret, depth, credits = handle->h_buffer_credits; 4500 struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data; 4501 struct buffer_head *last_eb_bh = NULL; 4502 struct ocfs2_extent_block *eb; 4503 struct ocfs2_extent_list *rightmost_el, *el; 4504 struct ocfs2_extent_rec split_rec; 4505 struct ocfs2_extent_rec *rec; 4506 struct ocfs2_insert_type insert; 4507 4508 /* 4509 * Setup the record to split before we grow the tree. 4510 */ 4511 el = path_leaf_el(path); 4512 rec = &el->l_recs[index]; 4513 ocfs2_make_right_split_rec(inode->i_sb, &split_rec, new_range, rec); 4514 4515 depth = path->p_tree_depth; 4516 if (depth > 0) { 4517 ret = ocfs2_read_block(OCFS2_SB(inode->i_sb), 4518 le64_to_cpu(di->i_last_eb_blk), 4519 &last_eb_bh, OCFS2_BH_CACHED, inode); 4520 if (ret < 0) { 4521 mlog_errno(ret); 4522 goto out; 4523 } 4524 4525 eb = (struct ocfs2_extent_block *) last_eb_bh->b_data; 4526 rightmost_el = &eb->h_list; 4527 } else 4528 rightmost_el = path_leaf_el(path); 4529 4530 credits += path->p_tree_depth + 4531 ocfs2_extend_meta_needed(&di->id2.i_list); 4532 ret = ocfs2_extend_trans(handle, credits); 4533 if (ret) { 4534 mlog_errno(ret); 4535 goto out; 4536 } 4537 4538 if (le16_to_cpu(rightmost_el->l_next_free_rec) == 4539 le16_to_cpu(rightmost_el->l_count)) { 4540 ret = ocfs2_grow_tree(inode, handle, di_bh, &depth, &last_eb_bh, 4541 meta_ac); 4542 if (ret) { 4543 mlog_errno(ret); 4544 goto out; 4545 } 4546 } 4547 4548 memset(&insert, 0, sizeof(struct ocfs2_insert_type)); 4549 insert.ins_appending = APPEND_NONE; 4550 insert.ins_contig = CONTIG_NONE; 4551 insert.ins_split = SPLIT_RIGHT; 4552 insert.ins_tree_depth = depth; 4553 4554 ret = ocfs2_do_insert_extent(inode, handle, di_bh, &split_rec, &insert); 4555 if (ret) 4556 mlog_errno(ret); 4557 4558 out: 4559 brelse(last_eb_bh); 4560 return ret; 4561 } 4562 4563 static int ocfs2_truncate_rec(struct inode *inode, handle_t *handle, 4564 struct ocfs2_path *path, int index, 4565 struct ocfs2_cached_dealloc_ctxt *dealloc, 4566 u32 cpos, u32 len) 4567 { 4568 int ret; 4569 u32 left_cpos, rec_range, trunc_range; 4570 int wants_rotate = 0, is_rightmost_tree_rec = 0; 4571 struct super_block *sb = inode->i_sb; 4572 struct ocfs2_path *left_path = NULL; 4573 struct ocfs2_extent_list *el = path_leaf_el(path); 4574 struct ocfs2_extent_rec *rec; 4575 struct ocfs2_extent_block *eb; 4576 4577 if (ocfs2_is_empty_extent(&el->l_recs[0]) && index > 0) { 4578 ret = ocfs2_rotate_tree_left(inode, handle, path, dealloc); 4579 if (ret) { 4580 mlog_errno(ret); 4581 goto out; 4582 } 4583 4584 index--; 4585 } 4586 4587 if (index == (le16_to_cpu(el->l_next_free_rec) - 1) && 4588 path->p_tree_depth) { 4589 /* 4590 * Check whether this is the rightmost tree record. If 4591 * we remove all of this record or part of its right 4592 * edge then an update of the record lengths above it 4593 * will be required. 4594 */ 4595 eb = (struct ocfs2_extent_block *)path_leaf_bh(path)->b_data; 4596 if (eb->h_next_leaf_blk == 0) 4597 is_rightmost_tree_rec = 1; 4598 } 4599 4600 rec = &el->l_recs[index]; 4601 if (index == 0 && path->p_tree_depth && 4602 le32_to_cpu(rec->e_cpos) == cpos) { 4603 /* 4604 * Changing the leftmost offset (via partial or whole 4605 * record truncate) of an interior (or rightmost) path 4606 * means we have to update the subtree that is formed 4607 * by this leaf and the one to it's left. 4608 * 4609 * There are two cases we can skip: 4610 * 1) Path is the leftmost one in our inode tree. 4611 * 2) The leaf is rightmost and will be empty after 4612 * we remove the extent record - the rotate code 4613 * knows how to update the newly formed edge. 4614 */ 4615 4616 ret = ocfs2_find_cpos_for_left_leaf(inode->i_sb, path, 4617 &left_cpos); 4618 if (ret) { 4619 mlog_errno(ret); 4620 goto out; 4621 } 4622 4623 if (left_cpos && le16_to_cpu(el->l_next_free_rec) > 1) { 4624 left_path = ocfs2_new_path(path_root_bh(path), 4625 path_root_el(path)); 4626 if (!left_path) { 4627 ret = -ENOMEM; 4628 mlog_errno(ret); 4629 goto out; 4630 } 4631 4632 ret = ocfs2_find_path(inode, left_path, left_cpos); 4633 if (ret) { 4634 mlog_errno(ret); 4635 goto out; 4636 } 4637 } 4638 } 4639 4640 ret = ocfs2_extend_rotate_transaction(handle, 0, 4641 handle->h_buffer_credits, 4642 path); 4643 if (ret) { 4644 mlog_errno(ret); 4645 goto out; 4646 } 4647 4648 ret = ocfs2_journal_access_path(inode, handle, path); 4649 if (ret) { 4650 mlog_errno(ret); 4651 goto out; 4652 } 4653 4654 ret = ocfs2_journal_access_path(inode, handle, left_path); 4655 if (ret) { 4656 mlog_errno(ret); 4657 goto out; 4658 } 4659 4660 rec_range = le32_to_cpu(rec->e_cpos) + ocfs2_rec_clusters(el, rec); 4661 trunc_range = cpos + len; 4662 4663 if (le32_to_cpu(rec->e_cpos) == cpos && rec_range == trunc_range) { 4664 int next_free; 4665 4666 memset(rec, 0, sizeof(*rec)); 4667 ocfs2_cleanup_merge(el, index); 4668 wants_rotate = 1; 4669 4670 next_free = le16_to_cpu(el->l_next_free_rec); 4671 if (is_rightmost_tree_rec && next_free > 1) { 4672 /* 4673 * We skip the edge update if this path will 4674 * be deleted by the rotate code. 4675 */ 4676 rec = &el->l_recs[next_free - 1]; 4677 ocfs2_adjust_rightmost_records(inode, handle, path, 4678 rec); 4679 } 4680 } else if (le32_to_cpu(rec->e_cpos) == cpos) { 4681 /* Remove leftmost portion of the record. */ 4682 le32_add_cpu(&rec->e_cpos, len); 4683 le64_add_cpu(&rec->e_blkno, ocfs2_clusters_to_blocks(sb, len)); 4684 le16_add_cpu(&rec->e_leaf_clusters, -len); 4685 } else if (rec_range == trunc_range) { 4686 /* Remove rightmost portion of the record */ 4687 le16_add_cpu(&rec->e_leaf_clusters, -len); 4688 if (is_rightmost_tree_rec) 4689 ocfs2_adjust_rightmost_records(inode, handle, path, rec); 4690 } else { 4691 /* Caller should have trapped this. */ 4692 mlog(ML_ERROR, "Inode %llu: Invalid record truncate: (%u, %u) " 4693 "(%u, %u)\n", (unsigned long long)OCFS2_I(inode)->ip_blkno, 4694 le32_to_cpu(rec->e_cpos), 4695 le16_to_cpu(rec->e_leaf_clusters), cpos, len); 4696 BUG(); 4697 } 4698 4699 if (left_path) { 4700 int subtree_index; 4701 4702 subtree_index = ocfs2_find_subtree_root(inode, left_path, path); 4703 ocfs2_complete_edge_insert(inode, handle, left_path, path, 4704 subtree_index); 4705 } 4706 4707 ocfs2_journal_dirty(handle, path_leaf_bh(path)); 4708 4709 ret = ocfs2_rotate_tree_left(inode, handle, path, dealloc); 4710 if (ret) { 4711 mlog_errno(ret); 4712 goto out; 4713 } 4714 4715 out: 4716 ocfs2_free_path(left_path); 4717 return ret; 4718 } 4719 4720 int ocfs2_remove_extent(struct inode *inode, struct buffer_head *di_bh, 4721 u32 cpos, u32 len, handle_t *handle, 4722 struct ocfs2_alloc_context *meta_ac, 4723 struct ocfs2_cached_dealloc_ctxt *dealloc) 4724 { 4725 int ret, index; 4726 u32 rec_range, trunc_range; 4727 struct ocfs2_extent_rec *rec; 4728 struct ocfs2_extent_list *el; 4729 struct ocfs2_path *path; 4730 4731 ocfs2_extent_map_trunc(inode, 0); 4732 4733 path = ocfs2_new_inode_path(di_bh); 4734 if (!path) { 4735 ret = -ENOMEM; 4736 mlog_errno(ret); 4737 goto out; 4738 } 4739 4740 ret = ocfs2_find_path(inode, path, cpos); 4741 if (ret) { 4742 mlog_errno(ret); 4743 goto out; 4744 } 4745 4746 el = path_leaf_el(path); 4747 index = ocfs2_search_extent_list(el, cpos); 4748 if (index == -1 || index >= le16_to_cpu(el->l_next_free_rec)) { 4749 ocfs2_error(inode->i_sb, 4750 "Inode %llu has an extent at cpos %u which can no " 4751 "longer be found.\n", 4752 (unsigned long long)OCFS2_I(inode)->ip_blkno, cpos); 4753 ret = -EROFS; 4754 goto out; 4755 } 4756 4757 /* 4758 * We have 3 cases of extent removal: 4759 * 1) Range covers the entire extent rec 4760 * 2) Range begins or ends on one edge of the extent rec 4761 * 3) Range is in the middle of the extent rec (no shared edges) 4762 * 4763 * For case 1 we remove the extent rec and left rotate to 4764 * fill the hole. 4765 * 4766 * For case 2 we just shrink the existing extent rec, with a 4767 * tree update if the shrinking edge is also the edge of an 4768 * extent block. 4769 * 4770 * For case 3 we do a right split to turn the extent rec into 4771 * something case 2 can handle. 4772 */ 4773 rec = &el->l_recs[index]; 4774 rec_range = le32_to_cpu(rec->e_cpos) + ocfs2_rec_clusters(el, rec); 4775 trunc_range = cpos + len; 4776 4777 BUG_ON(cpos < le32_to_cpu(rec->e_cpos) || trunc_range > rec_range); 4778 4779 mlog(0, "Inode %llu, remove (cpos %u, len %u). Existing index %d " 4780 "(cpos %u, len %u)\n", 4781 (unsigned long long)OCFS2_I(inode)->ip_blkno, cpos, len, index, 4782 le32_to_cpu(rec->e_cpos), ocfs2_rec_clusters(el, rec)); 4783 4784 if (le32_to_cpu(rec->e_cpos) == cpos || rec_range == trunc_range) { 4785 ret = ocfs2_truncate_rec(inode, handle, path, index, dealloc, 4786 cpos, len); 4787 if (ret) { 4788 mlog_errno(ret); 4789 goto out; 4790 } 4791 } else { 4792 ret = ocfs2_split_tree(inode, di_bh, handle, path, index, 4793 trunc_range, meta_ac); 4794 if (ret) { 4795 mlog_errno(ret); 4796 goto out; 4797 } 4798 4799 /* 4800 * The split could have manipulated the tree enough to 4801 * move the record location, so we have to look for it again. 4802 */ 4803 ocfs2_reinit_path(path, 1); 4804 4805 ret = ocfs2_find_path(inode, path, cpos); 4806 if (ret) { 4807 mlog_errno(ret); 4808 goto out; 4809 } 4810 4811 el = path_leaf_el(path); 4812 index = ocfs2_search_extent_list(el, cpos); 4813 if (index == -1 || index >= le16_to_cpu(el->l_next_free_rec)) { 4814 ocfs2_error(inode->i_sb, 4815 "Inode %llu: split at cpos %u lost record.", 4816 (unsigned long long)OCFS2_I(inode)->ip_blkno, 4817 cpos); 4818 ret = -EROFS; 4819 goto out; 4820 } 4821 4822 /* 4823 * Double check our values here. If anything is fishy, 4824 * it's easier to catch it at the top level. 4825 */ 4826 rec = &el->l_recs[index]; 4827 rec_range = le32_to_cpu(rec->e_cpos) + 4828 ocfs2_rec_clusters(el, rec); 4829 if (rec_range != trunc_range) { 4830 ocfs2_error(inode->i_sb, 4831 "Inode %llu: error after split at cpos %u" 4832 "trunc len %u, existing record is (%u,%u)", 4833 (unsigned long long)OCFS2_I(inode)->ip_blkno, 4834 cpos, len, le32_to_cpu(rec->e_cpos), 4835 ocfs2_rec_clusters(el, rec)); 4836 ret = -EROFS; 4837 goto out; 4838 } 4839 4840 ret = ocfs2_truncate_rec(inode, handle, path, index, dealloc, 4841 cpos, len); 4842 if (ret) { 4843 mlog_errno(ret); 4844 goto out; 4845 } 4846 } 4847 4848 out: 4849 ocfs2_free_path(path); 4850 return ret; 4851 } 4852 4853 int ocfs2_truncate_log_needs_flush(struct ocfs2_super *osb) 4854 { 4855 struct buffer_head *tl_bh = osb->osb_tl_bh; 4856 struct ocfs2_dinode *di; 4857 struct ocfs2_truncate_log *tl; 4858 4859 di = (struct ocfs2_dinode *) tl_bh->b_data; 4860 tl = &di->id2.i_dealloc; 4861 4862 mlog_bug_on_msg(le16_to_cpu(tl->tl_used) > le16_to_cpu(tl->tl_count), 4863 "slot %d, invalid truncate log parameters: used = " 4864 "%u, count = %u\n", osb->slot_num, 4865 le16_to_cpu(tl->tl_used), le16_to_cpu(tl->tl_count)); 4866 return le16_to_cpu(tl->tl_used) == le16_to_cpu(tl->tl_count); 4867 } 4868 4869 static int ocfs2_truncate_log_can_coalesce(struct ocfs2_truncate_log *tl, 4870 unsigned int new_start) 4871 { 4872 unsigned int tail_index; 4873 unsigned int current_tail; 4874 4875 /* No records, nothing to coalesce */ 4876 if (!le16_to_cpu(tl->tl_used)) 4877 return 0; 4878 4879 tail_index = le16_to_cpu(tl->tl_used) - 1; 4880 current_tail = le32_to_cpu(tl->tl_recs[tail_index].t_start); 4881 current_tail += le32_to_cpu(tl->tl_recs[tail_index].t_clusters); 4882 4883 return current_tail == new_start; 4884 } 4885 4886 int ocfs2_truncate_log_append(struct ocfs2_super *osb, 4887 handle_t *handle, 4888 u64 start_blk, 4889 unsigned int num_clusters) 4890 { 4891 int status, index; 4892 unsigned int start_cluster, tl_count; 4893 struct inode *tl_inode = osb->osb_tl_inode; 4894 struct buffer_head *tl_bh = osb->osb_tl_bh; 4895 struct ocfs2_dinode *di; 4896 struct ocfs2_truncate_log *tl; 4897 4898 mlog_entry("start_blk = %llu, num_clusters = %u\n", 4899 (unsigned long long)start_blk, num_clusters); 4900 4901 BUG_ON(mutex_trylock(&tl_inode->i_mutex)); 4902 4903 start_cluster = ocfs2_blocks_to_clusters(osb->sb, start_blk); 4904 4905 di = (struct ocfs2_dinode *) tl_bh->b_data; 4906 tl = &di->id2.i_dealloc; 4907 if (!OCFS2_IS_VALID_DINODE(di)) { 4908 OCFS2_RO_ON_INVALID_DINODE(osb->sb, di); 4909 status = -EIO; 4910 goto bail; 4911 } 4912 4913 tl_count = le16_to_cpu(tl->tl_count); 4914 mlog_bug_on_msg(tl_count > ocfs2_truncate_recs_per_inode(osb->sb) || 4915 tl_count == 0, 4916 "Truncate record count on #%llu invalid " 4917 "wanted %u, actual %u\n", 4918 (unsigned long long)OCFS2_I(tl_inode)->ip_blkno, 4919 ocfs2_truncate_recs_per_inode(osb->sb), 4920 le16_to_cpu(tl->tl_count)); 4921 4922 /* Caller should have known to flush before calling us. */ 4923 index = le16_to_cpu(tl->tl_used); 4924 if (index >= tl_count) { 4925 status = -ENOSPC; 4926 mlog_errno(status); 4927 goto bail; 4928 } 4929 4930 status = ocfs2_journal_access(handle, tl_inode, tl_bh, 4931 OCFS2_JOURNAL_ACCESS_WRITE); 4932 if (status < 0) { 4933 mlog_errno(status); 4934 goto bail; 4935 } 4936 4937 mlog(0, "Log truncate of %u clusters starting at cluster %u to " 4938 "%llu (index = %d)\n", num_clusters, start_cluster, 4939 (unsigned long long)OCFS2_I(tl_inode)->ip_blkno, index); 4940 4941 if (ocfs2_truncate_log_can_coalesce(tl, start_cluster)) { 4942 /* 4943 * Move index back to the record we are coalescing with. 4944 * ocfs2_truncate_log_can_coalesce() guarantees nonzero 4945 */ 4946 index--; 4947 4948 num_clusters += le32_to_cpu(tl->tl_recs[index].t_clusters); 4949 mlog(0, "Coalesce with index %u (start = %u, clusters = %u)\n", 4950 index, le32_to_cpu(tl->tl_recs[index].t_start), 4951 num_clusters); 4952 } else { 4953 tl->tl_recs[index].t_start = cpu_to_le32(start_cluster); 4954 tl->tl_used = cpu_to_le16(index + 1); 4955 } 4956 tl->tl_recs[index].t_clusters = cpu_to_le32(num_clusters); 4957 4958 status = ocfs2_journal_dirty(handle, tl_bh); 4959 if (status < 0) { 4960 mlog_errno(status); 4961 goto bail; 4962 } 4963 4964 bail: 4965 mlog_exit(status); 4966 return status; 4967 } 4968 4969 static int ocfs2_replay_truncate_records(struct ocfs2_super *osb, 4970 handle_t *handle, 4971 struct inode *data_alloc_inode, 4972 struct buffer_head *data_alloc_bh) 4973 { 4974 int status = 0; 4975 int i; 4976 unsigned int num_clusters; 4977 u64 start_blk; 4978 struct ocfs2_truncate_rec rec; 4979 struct ocfs2_dinode *di; 4980 struct ocfs2_truncate_log *tl; 4981 struct inode *tl_inode = osb->osb_tl_inode; 4982 struct buffer_head *tl_bh = osb->osb_tl_bh; 4983 4984 mlog_entry_void(); 4985 4986 di = (struct ocfs2_dinode *) tl_bh->b_data; 4987 tl = &di->id2.i_dealloc; 4988 i = le16_to_cpu(tl->tl_used) - 1; 4989 while (i >= 0) { 4990 /* Caller has given us at least enough credits to 4991 * update the truncate log dinode */ 4992 status = ocfs2_journal_access(handle, tl_inode, tl_bh, 4993 OCFS2_JOURNAL_ACCESS_WRITE); 4994 if (status < 0) { 4995 mlog_errno(status); 4996 goto bail; 4997 } 4998 4999 tl->tl_used = cpu_to_le16(i); 5000 5001 status = ocfs2_journal_dirty(handle, tl_bh); 5002 if (status < 0) { 5003 mlog_errno(status); 5004 goto bail; 5005 } 5006 5007 /* TODO: Perhaps we can calculate the bulk of the 5008 * credits up front rather than extending like 5009 * this. */ 5010 status = ocfs2_extend_trans(handle, 5011 OCFS2_TRUNCATE_LOG_FLUSH_ONE_REC); 5012 if (status < 0) { 5013 mlog_errno(status); 5014 goto bail; 5015 } 5016 5017 rec = tl->tl_recs[i]; 5018 start_blk = ocfs2_clusters_to_blocks(data_alloc_inode->i_sb, 5019 le32_to_cpu(rec.t_start)); 5020 num_clusters = le32_to_cpu(rec.t_clusters); 5021 5022 /* if start_blk is not set, we ignore the record as 5023 * invalid. */ 5024 if (start_blk) { 5025 mlog(0, "free record %d, start = %u, clusters = %u\n", 5026 i, le32_to_cpu(rec.t_start), num_clusters); 5027 5028 status = ocfs2_free_clusters(handle, data_alloc_inode, 5029 data_alloc_bh, start_blk, 5030 num_clusters); 5031 if (status < 0) { 5032 mlog_errno(status); 5033 goto bail; 5034 } 5035 } 5036 i--; 5037 } 5038 5039 bail: 5040 mlog_exit(status); 5041 return status; 5042 } 5043 5044 /* Expects you to already be holding tl_inode->i_mutex */ 5045 int __ocfs2_flush_truncate_log(struct ocfs2_super *osb) 5046 { 5047 int status; 5048 unsigned int num_to_flush; 5049 handle_t *handle; 5050 struct inode *tl_inode = osb->osb_tl_inode; 5051 struct inode *data_alloc_inode = NULL; 5052 struct buffer_head *tl_bh = osb->osb_tl_bh; 5053 struct buffer_head *data_alloc_bh = NULL; 5054 struct ocfs2_dinode *di; 5055 struct ocfs2_truncate_log *tl; 5056 5057 mlog_entry_void(); 5058 5059 BUG_ON(mutex_trylock(&tl_inode->i_mutex)); 5060 5061 di = (struct ocfs2_dinode *) tl_bh->b_data; 5062 tl = &di->id2.i_dealloc; 5063 if (!OCFS2_IS_VALID_DINODE(di)) { 5064 OCFS2_RO_ON_INVALID_DINODE(osb->sb, di); 5065 status = -EIO; 5066 goto out; 5067 } 5068 5069 num_to_flush = le16_to_cpu(tl->tl_used); 5070 mlog(0, "Flush %u records from truncate log #%llu\n", 5071 num_to_flush, (unsigned long long)OCFS2_I(tl_inode)->ip_blkno); 5072 if (!num_to_flush) { 5073 status = 0; 5074 goto out; 5075 } 5076 5077 data_alloc_inode = ocfs2_get_system_file_inode(osb, 5078 GLOBAL_BITMAP_SYSTEM_INODE, 5079 OCFS2_INVALID_SLOT); 5080 if (!data_alloc_inode) { 5081 status = -EINVAL; 5082 mlog(ML_ERROR, "Could not get bitmap inode!\n"); 5083 goto out; 5084 } 5085 5086 mutex_lock(&data_alloc_inode->i_mutex); 5087 5088 status = ocfs2_inode_lock(data_alloc_inode, &data_alloc_bh, 1); 5089 if (status < 0) { 5090 mlog_errno(status); 5091 goto out_mutex; 5092 } 5093 5094 handle = ocfs2_start_trans(osb, OCFS2_TRUNCATE_LOG_UPDATE); 5095 if (IS_ERR(handle)) { 5096 status = PTR_ERR(handle); 5097 mlog_errno(status); 5098 goto out_unlock; 5099 } 5100 5101 status = ocfs2_replay_truncate_records(osb, handle, data_alloc_inode, 5102 data_alloc_bh); 5103 if (status < 0) 5104 mlog_errno(status); 5105 5106 ocfs2_commit_trans(osb, handle); 5107 5108 out_unlock: 5109 brelse(data_alloc_bh); 5110 ocfs2_inode_unlock(data_alloc_inode, 1); 5111 5112 out_mutex: 5113 mutex_unlock(&data_alloc_inode->i_mutex); 5114 iput(data_alloc_inode); 5115 5116 out: 5117 mlog_exit(status); 5118 return status; 5119 } 5120 5121 int ocfs2_flush_truncate_log(struct ocfs2_super *osb) 5122 { 5123 int status; 5124 struct inode *tl_inode = osb->osb_tl_inode; 5125 5126 mutex_lock(&tl_inode->i_mutex); 5127 status = __ocfs2_flush_truncate_log(osb); 5128 mutex_unlock(&tl_inode->i_mutex); 5129 5130 return status; 5131 } 5132 5133 static void ocfs2_truncate_log_worker(struct work_struct *work) 5134 { 5135 int status; 5136 struct ocfs2_super *osb = 5137 container_of(work, struct ocfs2_super, 5138 osb_truncate_log_wq.work); 5139 5140 mlog_entry_void(); 5141 5142 status = ocfs2_flush_truncate_log(osb); 5143 if (status < 0) 5144 mlog_errno(status); 5145 else 5146 ocfs2_init_inode_steal_slot(osb); 5147 5148 mlog_exit(status); 5149 } 5150 5151 #define OCFS2_TRUNCATE_LOG_FLUSH_INTERVAL (2 * HZ) 5152 void ocfs2_schedule_truncate_log_flush(struct ocfs2_super *osb, 5153 int cancel) 5154 { 5155 if (osb->osb_tl_inode) { 5156 /* We want to push off log flushes while truncates are 5157 * still running. */ 5158 if (cancel) 5159 cancel_delayed_work(&osb->osb_truncate_log_wq); 5160 5161 queue_delayed_work(ocfs2_wq, &osb->osb_truncate_log_wq, 5162 OCFS2_TRUNCATE_LOG_FLUSH_INTERVAL); 5163 } 5164 } 5165 5166 static int ocfs2_get_truncate_log_info(struct ocfs2_super *osb, 5167 int slot_num, 5168 struct inode **tl_inode, 5169 struct buffer_head **tl_bh) 5170 { 5171 int status; 5172 struct inode *inode = NULL; 5173 struct buffer_head *bh = NULL; 5174 5175 inode = ocfs2_get_system_file_inode(osb, 5176 TRUNCATE_LOG_SYSTEM_INODE, 5177 slot_num); 5178 if (!inode) { 5179 status = -EINVAL; 5180 mlog(ML_ERROR, "Could not get load truncate log inode!\n"); 5181 goto bail; 5182 } 5183 5184 status = ocfs2_read_block(osb, OCFS2_I(inode)->ip_blkno, &bh, 5185 OCFS2_BH_CACHED, inode); 5186 if (status < 0) { 5187 iput(inode); 5188 mlog_errno(status); 5189 goto bail; 5190 } 5191 5192 *tl_inode = inode; 5193 *tl_bh = bh; 5194 bail: 5195 mlog_exit(status); 5196 return status; 5197 } 5198 5199 /* called during the 1st stage of node recovery. we stamp a clean 5200 * truncate log and pass back a copy for processing later. if the 5201 * truncate log does not require processing, a *tl_copy is set to 5202 * NULL. */ 5203 int ocfs2_begin_truncate_log_recovery(struct ocfs2_super *osb, 5204 int slot_num, 5205 struct ocfs2_dinode **tl_copy) 5206 { 5207 int status; 5208 struct inode *tl_inode = NULL; 5209 struct buffer_head *tl_bh = NULL; 5210 struct ocfs2_dinode *di; 5211 struct ocfs2_truncate_log *tl; 5212 5213 *tl_copy = NULL; 5214 5215 mlog(0, "recover truncate log from slot %d\n", slot_num); 5216 5217 status = ocfs2_get_truncate_log_info(osb, slot_num, &tl_inode, &tl_bh); 5218 if (status < 0) { 5219 mlog_errno(status); 5220 goto bail; 5221 } 5222 5223 di = (struct ocfs2_dinode *) tl_bh->b_data; 5224 tl = &di->id2.i_dealloc; 5225 if (!OCFS2_IS_VALID_DINODE(di)) { 5226 OCFS2_RO_ON_INVALID_DINODE(tl_inode->i_sb, di); 5227 status = -EIO; 5228 goto bail; 5229 } 5230 5231 if (le16_to_cpu(tl->tl_used)) { 5232 mlog(0, "We'll have %u logs to recover\n", 5233 le16_to_cpu(tl->tl_used)); 5234 5235 *tl_copy = kmalloc(tl_bh->b_size, GFP_KERNEL); 5236 if (!(*tl_copy)) { 5237 status = -ENOMEM; 5238 mlog_errno(status); 5239 goto bail; 5240 } 5241 5242 /* Assuming the write-out below goes well, this copy 5243 * will be passed back to recovery for processing. */ 5244 memcpy(*tl_copy, tl_bh->b_data, tl_bh->b_size); 5245 5246 /* All we need to do to clear the truncate log is set 5247 * tl_used. */ 5248 tl->tl_used = 0; 5249 5250 status = ocfs2_write_block(osb, tl_bh, tl_inode); 5251 if (status < 0) { 5252 mlog_errno(status); 5253 goto bail; 5254 } 5255 } 5256 5257 bail: 5258 if (tl_inode) 5259 iput(tl_inode); 5260 if (tl_bh) 5261 brelse(tl_bh); 5262 5263 if (status < 0 && (*tl_copy)) { 5264 kfree(*tl_copy); 5265 *tl_copy = NULL; 5266 } 5267 5268 mlog_exit(status); 5269 return status; 5270 } 5271 5272 int ocfs2_complete_truncate_log_recovery(struct ocfs2_super *osb, 5273 struct ocfs2_dinode *tl_copy) 5274 { 5275 int status = 0; 5276 int i; 5277 unsigned int clusters, num_recs, start_cluster; 5278 u64 start_blk; 5279 handle_t *handle; 5280 struct inode *tl_inode = osb->osb_tl_inode; 5281 struct ocfs2_truncate_log *tl; 5282 5283 mlog_entry_void(); 5284 5285 if (OCFS2_I(tl_inode)->ip_blkno == le64_to_cpu(tl_copy->i_blkno)) { 5286 mlog(ML_ERROR, "Asked to recover my own truncate log!\n"); 5287 return -EINVAL; 5288 } 5289 5290 tl = &tl_copy->id2.i_dealloc; 5291 num_recs = le16_to_cpu(tl->tl_used); 5292 mlog(0, "cleanup %u records from %llu\n", num_recs, 5293 (unsigned long long)le64_to_cpu(tl_copy->i_blkno)); 5294 5295 mutex_lock(&tl_inode->i_mutex); 5296 for(i = 0; i < num_recs; i++) { 5297 if (ocfs2_truncate_log_needs_flush(osb)) { 5298 status = __ocfs2_flush_truncate_log(osb); 5299 if (status < 0) { 5300 mlog_errno(status); 5301 goto bail_up; 5302 } 5303 } 5304 5305 handle = ocfs2_start_trans(osb, OCFS2_TRUNCATE_LOG_UPDATE); 5306 if (IS_ERR(handle)) { 5307 status = PTR_ERR(handle); 5308 mlog_errno(status); 5309 goto bail_up; 5310 } 5311 5312 clusters = le32_to_cpu(tl->tl_recs[i].t_clusters); 5313 start_cluster = le32_to_cpu(tl->tl_recs[i].t_start); 5314 start_blk = ocfs2_clusters_to_blocks(osb->sb, start_cluster); 5315 5316 status = ocfs2_truncate_log_append(osb, handle, 5317 start_blk, clusters); 5318 ocfs2_commit_trans(osb, handle); 5319 if (status < 0) { 5320 mlog_errno(status); 5321 goto bail_up; 5322 } 5323 } 5324 5325 bail_up: 5326 mutex_unlock(&tl_inode->i_mutex); 5327 5328 mlog_exit(status); 5329 return status; 5330 } 5331 5332 void ocfs2_truncate_log_shutdown(struct ocfs2_super *osb) 5333 { 5334 int status; 5335 struct inode *tl_inode = osb->osb_tl_inode; 5336 5337 mlog_entry_void(); 5338 5339 if (tl_inode) { 5340 cancel_delayed_work(&osb->osb_truncate_log_wq); 5341 flush_workqueue(ocfs2_wq); 5342 5343 status = ocfs2_flush_truncate_log(osb); 5344 if (status < 0) 5345 mlog_errno(status); 5346 5347 brelse(osb->osb_tl_bh); 5348 iput(osb->osb_tl_inode); 5349 } 5350 5351 mlog_exit_void(); 5352 } 5353 5354 int ocfs2_truncate_log_init(struct ocfs2_super *osb) 5355 { 5356 int status; 5357 struct inode *tl_inode = NULL; 5358 struct buffer_head *tl_bh = NULL; 5359 5360 mlog_entry_void(); 5361 5362 status = ocfs2_get_truncate_log_info(osb, 5363 osb->slot_num, 5364 &tl_inode, 5365 &tl_bh); 5366 if (status < 0) 5367 mlog_errno(status); 5368 5369 /* ocfs2_truncate_log_shutdown keys on the existence of 5370 * osb->osb_tl_inode so we don't set any of the osb variables 5371 * until we're sure all is well. */ 5372 INIT_DELAYED_WORK(&osb->osb_truncate_log_wq, 5373 ocfs2_truncate_log_worker); 5374 osb->osb_tl_bh = tl_bh; 5375 osb->osb_tl_inode = tl_inode; 5376 5377 mlog_exit(status); 5378 return status; 5379 } 5380 5381 /* 5382 * Delayed de-allocation of suballocator blocks. 5383 * 5384 * Some sets of block de-allocations might involve multiple suballocator inodes. 5385 * 5386 * The locking for this can get extremely complicated, especially when 5387 * the suballocator inodes to delete from aren't known until deep 5388 * within an unrelated codepath. 5389 * 5390 * ocfs2_extent_block structures are a good example of this - an inode 5391 * btree could have been grown by any number of nodes each allocating 5392 * out of their own suballoc inode. 5393 * 5394 * These structures allow the delay of block de-allocation until a 5395 * later time, when locking of multiple cluster inodes won't cause 5396 * deadlock. 5397 */ 5398 5399 /* 5400 * Describes a single block free from a suballocator 5401 */ 5402 struct ocfs2_cached_block_free { 5403 struct ocfs2_cached_block_free *free_next; 5404 u64 free_blk; 5405 unsigned int free_bit; 5406 }; 5407 5408 struct ocfs2_per_slot_free_list { 5409 struct ocfs2_per_slot_free_list *f_next_suballocator; 5410 int f_inode_type; 5411 int f_slot; 5412 struct ocfs2_cached_block_free *f_first; 5413 }; 5414 5415 static int ocfs2_free_cached_items(struct ocfs2_super *osb, 5416 int sysfile_type, 5417 int slot, 5418 struct ocfs2_cached_block_free *head) 5419 { 5420 int ret; 5421 u64 bg_blkno; 5422 handle_t *handle; 5423 struct inode *inode; 5424 struct buffer_head *di_bh = NULL; 5425 struct ocfs2_cached_block_free *tmp; 5426 5427 inode = ocfs2_get_system_file_inode(osb, sysfile_type, slot); 5428 if (!inode) { 5429 ret = -EINVAL; 5430 mlog_errno(ret); 5431 goto out; 5432 } 5433 5434 mutex_lock(&inode->i_mutex); 5435 5436 ret = ocfs2_inode_lock(inode, &di_bh, 1); 5437 if (ret) { 5438 mlog_errno(ret); 5439 goto out_mutex; 5440 } 5441 5442 handle = ocfs2_start_trans(osb, OCFS2_SUBALLOC_FREE); 5443 if (IS_ERR(handle)) { 5444 ret = PTR_ERR(handle); 5445 mlog_errno(ret); 5446 goto out_unlock; 5447 } 5448 5449 while (head) { 5450 bg_blkno = ocfs2_which_suballoc_group(head->free_blk, 5451 head->free_bit); 5452 mlog(0, "Free bit: (bit %u, blkno %llu)\n", 5453 head->free_bit, (unsigned long long)head->free_blk); 5454 5455 ret = ocfs2_free_suballoc_bits(handle, inode, di_bh, 5456 head->free_bit, bg_blkno, 1); 5457 if (ret) { 5458 mlog_errno(ret); 5459 goto out_journal; 5460 } 5461 5462 ret = ocfs2_extend_trans(handle, OCFS2_SUBALLOC_FREE); 5463 if (ret) { 5464 mlog_errno(ret); 5465 goto out_journal; 5466 } 5467 5468 tmp = head; 5469 head = head->free_next; 5470 kfree(tmp); 5471 } 5472 5473 out_journal: 5474 ocfs2_commit_trans(osb, handle); 5475 5476 out_unlock: 5477 ocfs2_inode_unlock(inode, 1); 5478 brelse(di_bh); 5479 out_mutex: 5480 mutex_unlock(&inode->i_mutex); 5481 iput(inode); 5482 out: 5483 while(head) { 5484 /* Premature exit may have left some dangling items. */ 5485 tmp = head; 5486 head = head->free_next; 5487 kfree(tmp); 5488 } 5489 5490 return ret; 5491 } 5492 5493 int ocfs2_run_deallocs(struct ocfs2_super *osb, 5494 struct ocfs2_cached_dealloc_ctxt *ctxt) 5495 { 5496 int ret = 0, ret2; 5497 struct ocfs2_per_slot_free_list *fl; 5498 5499 if (!ctxt) 5500 return 0; 5501 5502 while (ctxt->c_first_suballocator) { 5503 fl = ctxt->c_first_suballocator; 5504 5505 if (fl->f_first) { 5506 mlog(0, "Free items: (type %u, slot %d)\n", 5507 fl->f_inode_type, fl->f_slot); 5508 ret2 = ocfs2_free_cached_items(osb, fl->f_inode_type, 5509 fl->f_slot, fl->f_first); 5510 if (ret2) 5511 mlog_errno(ret2); 5512 if (!ret) 5513 ret = ret2; 5514 } 5515 5516 ctxt->c_first_suballocator = fl->f_next_suballocator; 5517 kfree(fl); 5518 } 5519 5520 return ret; 5521 } 5522 5523 static struct ocfs2_per_slot_free_list * 5524 ocfs2_find_per_slot_free_list(int type, 5525 int slot, 5526 struct ocfs2_cached_dealloc_ctxt *ctxt) 5527 { 5528 struct ocfs2_per_slot_free_list *fl = ctxt->c_first_suballocator; 5529 5530 while (fl) { 5531 if (fl->f_inode_type == type && fl->f_slot == slot) 5532 return fl; 5533 5534 fl = fl->f_next_suballocator; 5535 } 5536 5537 fl = kmalloc(sizeof(*fl), GFP_NOFS); 5538 if (fl) { 5539 fl->f_inode_type = type; 5540 fl->f_slot = slot; 5541 fl->f_first = NULL; 5542 fl->f_next_suballocator = ctxt->c_first_suballocator; 5543 5544 ctxt->c_first_suballocator = fl; 5545 } 5546 return fl; 5547 } 5548 5549 static int ocfs2_cache_block_dealloc(struct ocfs2_cached_dealloc_ctxt *ctxt, 5550 int type, int slot, u64 blkno, 5551 unsigned int bit) 5552 { 5553 int ret; 5554 struct ocfs2_per_slot_free_list *fl; 5555 struct ocfs2_cached_block_free *item; 5556 5557 fl = ocfs2_find_per_slot_free_list(type, slot, ctxt); 5558 if (fl == NULL) { 5559 ret = -ENOMEM; 5560 mlog_errno(ret); 5561 goto out; 5562 } 5563 5564 item = kmalloc(sizeof(*item), GFP_NOFS); 5565 if (item == NULL) { 5566 ret = -ENOMEM; 5567 mlog_errno(ret); 5568 goto out; 5569 } 5570 5571 mlog(0, "Insert: (type %d, slot %u, bit %u, blk %llu)\n", 5572 type, slot, bit, (unsigned long long)blkno); 5573 5574 item->free_blk = blkno; 5575 item->free_bit = bit; 5576 item->free_next = fl->f_first; 5577 5578 fl->f_first = item; 5579 5580 ret = 0; 5581 out: 5582 return ret; 5583 } 5584 5585 static int ocfs2_cache_extent_block_free(struct ocfs2_cached_dealloc_ctxt *ctxt, 5586 struct ocfs2_extent_block *eb) 5587 { 5588 return ocfs2_cache_block_dealloc(ctxt, EXTENT_ALLOC_SYSTEM_INODE, 5589 le16_to_cpu(eb->h_suballoc_slot), 5590 le64_to_cpu(eb->h_blkno), 5591 le16_to_cpu(eb->h_suballoc_bit)); 5592 } 5593 5594 /* This function will figure out whether the currently last extent 5595 * block will be deleted, and if it will, what the new last extent 5596 * block will be so we can update his h_next_leaf_blk field, as well 5597 * as the dinodes i_last_eb_blk */ 5598 static int ocfs2_find_new_last_ext_blk(struct inode *inode, 5599 unsigned int clusters_to_del, 5600 struct ocfs2_path *path, 5601 struct buffer_head **new_last_eb) 5602 { 5603 int next_free, ret = 0; 5604 u32 cpos; 5605 struct ocfs2_extent_rec *rec; 5606 struct ocfs2_extent_block *eb; 5607 struct ocfs2_extent_list *el; 5608 struct buffer_head *bh = NULL; 5609 5610 *new_last_eb = NULL; 5611 5612 /* we have no tree, so of course, no last_eb. */ 5613 if (!path->p_tree_depth) 5614 goto out; 5615 5616 /* trunc to zero special case - this makes tree_depth = 0 5617 * regardless of what it is. */ 5618 if (OCFS2_I(inode)->ip_clusters == clusters_to_del) 5619 goto out; 5620 5621 el = path_leaf_el(path); 5622 BUG_ON(!el->l_next_free_rec); 5623 5624 /* 5625 * Make sure that this extent list will actually be empty 5626 * after we clear away the data. We can shortcut out if 5627 * there's more than one non-empty extent in the 5628 * list. Otherwise, a check of the remaining extent is 5629 * necessary. 5630 */ 5631 next_free = le16_to_cpu(el->l_next_free_rec); 5632 rec = NULL; 5633 if (ocfs2_is_empty_extent(&el->l_recs[0])) { 5634 if (next_free > 2) 5635 goto out; 5636 5637 /* We may have a valid extent in index 1, check it. */ 5638 if (next_free == 2) 5639 rec = &el->l_recs[1]; 5640 5641 /* 5642 * Fall through - no more nonempty extents, so we want 5643 * to delete this leaf. 5644 */ 5645 } else { 5646 if (next_free > 1) 5647 goto out; 5648 5649 rec = &el->l_recs[0]; 5650 } 5651 5652 if (rec) { 5653 /* 5654 * Check it we'll only be trimming off the end of this 5655 * cluster. 5656 */ 5657 if (le16_to_cpu(rec->e_leaf_clusters) > clusters_to_del) 5658 goto out; 5659 } 5660 5661 ret = ocfs2_find_cpos_for_left_leaf(inode->i_sb, path, &cpos); 5662 if (ret) { 5663 mlog_errno(ret); 5664 goto out; 5665 } 5666 5667 ret = ocfs2_find_leaf(inode, path_root_el(path), cpos, &bh); 5668 if (ret) { 5669 mlog_errno(ret); 5670 goto out; 5671 } 5672 5673 eb = (struct ocfs2_extent_block *) bh->b_data; 5674 el = &eb->h_list; 5675 if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) { 5676 OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb); 5677 ret = -EROFS; 5678 goto out; 5679 } 5680 5681 *new_last_eb = bh; 5682 get_bh(*new_last_eb); 5683 mlog(0, "returning block %llu, (cpos: %u)\n", 5684 (unsigned long long)le64_to_cpu(eb->h_blkno), cpos); 5685 out: 5686 brelse(bh); 5687 5688 return ret; 5689 } 5690 5691 /* 5692 * Trim some clusters off the rightmost edge of a tree. Only called 5693 * during truncate. 5694 * 5695 * The caller needs to: 5696 * - start journaling of each path component. 5697 * - compute and fully set up any new last ext block 5698 */ 5699 static int ocfs2_trim_tree(struct inode *inode, struct ocfs2_path *path, 5700 handle_t *handle, struct ocfs2_truncate_context *tc, 5701 u32 clusters_to_del, u64 *delete_start) 5702 { 5703 int ret, i, index = path->p_tree_depth; 5704 u32 new_edge = 0; 5705 u64 deleted_eb = 0; 5706 struct buffer_head *bh; 5707 struct ocfs2_extent_list *el; 5708 struct ocfs2_extent_rec *rec; 5709 5710 *delete_start = 0; 5711 5712 while (index >= 0) { 5713 bh = path->p_node[index].bh; 5714 el = path->p_node[index].el; 5715 5716 mlog(0, "traveling tree (index = %d, block = %llu)\n", 5717 index, (unsigned long long)bh->b_blocknr); 5718 5719 BUG_ON(le16_to_cpu(el->l_next_free_rec) == 0); 5720 5721 if (index != 5722 (path->p_tree_depth - le16_to_cpu(el->l_tree_depth))) { 5723 ocfs2_error(inode->i_sb, 5724 "Inode %lu has invalid ext. block %llu", 5725 inode->i_ino, 5726 (unsigned long long)bh->b_blocknr); 5727 ret = -EROFS; 5728 goto out; 5729 } 5730 5731 find_tail_record: 5732 i = le16_to_cpu(el->l_next_free_rec) - 1; 5733 rec = &el->l_recs[i]; 5734 5735 mlog(0, "Extent list before: record %d: (%u, %u, %llu), " 5736 "next = %u\n", i, le32_to_cpu(rec->e_cpos), 5737 ocfs2_rec_clusters(el, rec), 5738 (unsigned long long)le64_to_cpu(rec->e_blkno), 5739 le16_to_cpu(el->l_next_free_rec)); 5740 5741 BUG_ON(ocfs2_rec_clusters(el, rec) < clusters_to_del); 5742 5743 if (le16_to_cpu(el->l_tree_depth) == 0) { 5744 /* 5745 * If the leaf block contains a single empty 5746 * extent and no records, we can just remove 5747 * the block. 5748 */ 5749 if (i == 0 && ocfs2_is_empty_extent(rec)) { 5750 memset(rec, 0, 5751 sizeof(struct ocfs2_extent_rec)); 5752 el->l_next_free_rec = cpu_to_le16(0); 5753 5754 goto delete; 5755 } 5756 5757 /* 5758 * Remove any empty extents by shifting things 5759 * left. That should make life much easier on 5760 * the code below. This condition is rare 5761 * enough that we shouldn't see a performance 5762 * hit. 5763 */ 5764 if (ocfs2_is_empty_extent(&el->l_recs[0])) { 5765 le16_add_cpu(&el->l_next_free_rec, -1); 5766 5767 for(i = 0; 5768 i < le16_to_cpu(el->l_next_free_rec); i++) 5769 el->l_recs[i] = el->l_recs[i + 1]; 5770 5771 memset(&el->l_recs[i], 0, 5772 sizeof(struct ocfs2_extent_rec)); 5773 5774 /* 5775 * We've modified our extent list. The 5776 * simplest way to handle this change 5777 * is to being the search from the 5778 * start again. 5779 */ 5780 goto find_tail_record; 5781 } 5782 5783 le16_add_cpu(&rec->e_leaf_clusters, -clusters_to_del); 5784 5785 /* 5786 * We'll use "new_edge" on our way back up the 5787 * tree to know what our rightmost cpos is. 5788 */ 5789 new_edge = le16_to_cpu(rec->e_leaf_clusters); 5790 new_edge += le32_to_cpu(rec->e_cpos); 5791 5792 /* 5793 * The caller will use this to delete data blocks. 5794 */ 5795 *delete_start = le64_to_cpu(rec->e_blkno) 5796 + ocfs2_clusters_to_blocks(inode->i_sb, 5797 le16_to_cpu(rec->e_leaf_clusters)); 5798 5799 /* 5800 * If it's now empty, remove this record. 5801 */ 5802 if (le16_to_cpu(rec->e_leaf_clusters) == 0) { 5803 memset(rec, 0, 5804 sizeof(struct ocfs2_extent_rec)); 5805 le16_add_cpu(&el->l_next_free_rec, -1); 5806 } 5807 } else { 5808 if (le64_to_cpu(rec->e_blkno) == deleted_eb) { 5809 memset(rec, 0, 5810 sizeof(struct ocfs2_extent_rec)); 5811 le16_add_cpu(&el->l_next_free_rec, -1); 5812 5813 goto delete; 5814 } 5815 5816 /* Can this actually happen? */ 5817 if (le16_to_cpu(el->l_next_free_rec) == 0) 5818 goto delete; 5819 5820 /* 5821 * We never actually deleted any clusters 5822 * because our leaf was empty. There's no 5823 * reason to adjust the rightmost edge then. 5824 */ 5825 if (new_edge == 0) 5826 goto delete; 5827 5828 rec->e_int_clusters = cpu_to_le32(new_edge); 5829 le32_add_cpu(&rec->e_int_clusters, 5830 -le32_to_cpu(rec->e_cpos)); 5831 5832 /* 5833 * A deleted child record should have been 5834 * caught above. 5835 */ 5836 BUG_ON(le32_to_cpu(rec->e_int_clusters) == 0); 5837 } 5838 5839 delete: 5840 ret = ocfs2_journal_dirty(handle, bh); 5841 if (ret) { 5842 mlog_errno(ret); 5843 goto out; 5844 } 5845 5846 mlog(0, "extent list container %llu, after: record %d: " 5847 "(%u, %u, %llu), next = %u.\n", 5848 (unsigned long long)bh->b_blocknr, i, 5849 le32_to_cpu(rec->e_cpos), ocfs2_rec_clusters(el, rec), 5850 (unsigned long long)le64_to_cpu(rec->e_blkno), 5851 le16_to_cpu(el->l_next_free_rec)); 5852 5853 /* 5854 * We must be careful to only attempt delete of an 5855 * extent block (and not the root inode block). 5856 */ 5857 if (index > 0 && le16_to_cpu(el->l_next_free_rec) == 0) { 5858 struct ocfs2_extent_block *eb = 5859 (struct ocfs2_extent_block *)bh->b_data; 5860 5861 /* 5862 * Save this for use when processing the 5863 * parent block. 5864 */ 5865 deleted_eb = le64_to_cpu(eb->h_blkno); 5866 5867 mlog(0, "deleting this extent block.\n"); 5868 5869 ocfs2_remove_from_cache(inode, bh); 5870 5871 BUG_ON(ocfs2_rec_clusters(el, &el->l_recs[0])); 5872 BUG_ON(le32_to_cpu(el->l_recs[0].e_cpos)); 5873 BUG_ON(le64_to_cpu(el->l_recs[0].e_blkno)); 5874 5875 ret = ocfs2_cache_extent_block_free(&tc->tc_dealloc, eb); 5876 /* An error here is not fatal. */ 5877 if (ret < 0) 5878 mlog_errno(ret); 5879 } else { 5880 deleted_eb = 0; 5881 } 5882 5883 index--; 5884 } 5885 5886 ret = 0; 5887 out: 5888 return ret; 5889 } 5890 5891 static int ocfs2_do_truncate(struct ocfs2_super *osb, 5892 unsigned int clusters_to_del, 5893 struct inode *inode, 5894 struct buffer_head *fe_bh, 5895 handle_t *handle, 5896 struct ocfs2_truncate_context *tc, 5897 struct ocfs2_path *path) 5898 { 5899 int status; 5900 struct ocfs2_dinode *fe; 5901 struct ocfs2_extent_block *last_eb = NULL; 5902 struct ocfs2_extent_list *el; 5903 struct buffer_head *last_eb_bh = NULL; 5904 u64 delete_blk = 0; 5905 5906 fe = (struct ocfs2_dinode *) fe_bh->b_data; 5907 5908 status = ocfs2_find_new_last_ext_blk(inode, clusters_to_del, 5909 path, &last_eb_bh); 5910 if (status < 0) { 5911 mlog_errno(status); 5912 goto bail; 5913 } 5914 5915 /* 5916 * Each component will be touched, so we might as well journal 5917 * here to avoid having to handle errors later. 5918 */ 5919 status = ocfs2_journal_access_path(inode, handle, path); 5920 if (status < 0) { 5921 mlog_errno(status); 5922 goto bail; 5923 } 5924 5925 if (last_eb_bh) { 5926 status = ocfs2_journal_access(handle, inode, last_eb_bh, 5927 OCFS2_JOURNAL_ACCESS_WRITE); 5928 if (status < 0) { 5929 mlog_errno(status); 5930 goto bail; 5931 } 5932 5933 last_eb = (struct ocfs2_extent_block *) last_eb_bh->b_data; 5934 } 5935 5936 el = &(fe->id2.i_list); 5937 5938 /* 5939 * Lower levels depend on this never happening, but it's best 5940 * to check it up here before changing the tree. 5941 */ 5942 if (el->l_tree_depth && el->l_recs[0].e_int_clusters == 0) { 5943 ocfs2_error(inode->i_sb, 5944 "Inode %lu has an empty extent record, depth %u\n", 5945 inode->i_ino, le16_to_cpu(el->l_tree_depth)); 5946 status = -EROFS; 5947 goto bail; 5948 } 5949 5950 spin_lock(&OCFS2_I(inode)->ip_lock); 5951 OCFS2_I(inode)->ip_clusters = le32_to_cpu(fe->i_clusters) - 5952 clusters_to_del; 5953 spin_unlock(&OCFS2_I(inode)->ip_lock); 5954 le32_add_cpu(&fe->i_clusters, -clusters_to_del); 5955 inode->i_blocks = ocfs2_inode_sector_count(inode); 5956 5957 status = ocfs2_trim_tree(inode, path, handle, tc, 5958 clusters_to_del, &delete_blk); 5959 if (status) { 5960 mlog_errno(status); 5961 goto bail; 5962 } 5963 5964 if (le32_to_cpu(fe->i_clusters) == 0) { 5965 /* trunc to zero is a special case. */ 5966 el->l_tree_depth = 0; 5967 fe->i_last_eb_blk = 0; 5968 } else if (last_eb) 5969 fe->i_last_eb_blk = last_eb->h_blkno; 5970 5971 status = ocfs2_journal_dirty(handle, fe_bh); 5972 if (status < 0) { 5973 mlog_errno(status); 5974 goto bail; 5975 } 5976 5977 if (last_eb) { 5978 /* If there will be a new last extent block, then by 5979 * definition, there cannot be any leaves to the right of 5980 * him. */ 5981 last_eb->h_next_leaf_blk = 0; 5982 status = ocfs2_journal_dirty(handle, last_eb_bh); 5983 if (status < 0) { 5984 mlog_errno(status); 5985 goto bail; 5986 } 5987 } 5988 5989 if (delete_blk) { 5990 status = ocfs2_truncate_log_append(osb, handle, delete_blk, 5991 clusters_to_del); 5992 if (status < 0) { 5993 mlog_errno(status); 5994 goto bail; 5995 } 5996 } 5997 status = 0; 5998 bail: 5999 6000 mlog_exit(status); 6001 return status; 6002 } 6003 6004 static int ocfs2_writeback_zero_func(handle_t *handle, struct buffer_head *bh) 6005 { 6006 set_buffer_uptodate(bh); 6007 mark_buffer_dirty(bh); 6008 return 0; 6009 } 6010 6011 static int ocfs2_ordered_zero_func(handle_t *handle, struct buffer_head *bh) 6012 { 6013 set_buffer_uptodate(bh); 6014 mark_buffer_dirty(bh); 6015 return ocfs2_journal_dirty_data(handle, bh); 6016 } 6017 6018 static void ocfs2_map_and_dirty_page(struct inode *inode, handle_t *handle, 6019 unsigned int from, unsigned int to, 6020 struct page *page, int zero, u64 *phys) 6021 { 6022 int ret, partial = 0; 6023 6024 ret = ocfs2_map_page_blocks(page, phys, inode, from, to, 0); 6025 if (ret) 6026 mlog_errno(ret); 6027 6028 if (zero) 6029 zero_user_segment(page, from, to); 6030 6031 /* 6032 * Need to set the buffers we zero'd into uptodate 6033 * here if they aren't - ocfs2_map_page_blocks() 6034 * might've skipped some 6035 */ 6036 if (ocfs2_should_order_data(inode)) { 6037 ret = walk_page_buffers(handle, 6038 page_buffers(page), 6039 from, to, &partial, 6040 ocfs2_ordered_zero_func); 6041 if (ret < 0) 6042 mlog_errno(ret); 6043 } else { 6044 ret = walk_page_buffers(handle, page_buffers(page), 6045 from, to, &partial, 6046 ocfs2_writeback_zero_func); 6047 if (ret < 0) 6048 mlog_errno(ret); 6049 } 6050 6051 if (!partial) 6052 SetPageUptodate(page); 6053 6054 flush_dcache_page(page); 6055 } 6056 6057 static void ocfs2_zero_cluster_pages(struct inode *inode, loff_t start, 6058 loff_t end, struct page **pages, 6059 int numpages, u64 phys, handle_t *handle) 6060 { 6061 int i; 6062 struct page *page; 6063 unsigned int from, to = PAGE_CACHE_SIZE; 6064 struct super_block *sb = inode->i_sb; 6065 6066 BUG_ON(!ocfs2_sparse_alloc(OCFS2_SB(sb))); 6067 6068 if (numpages == 0) 6069 goto out; 6070 6071 to = PAGE_CACHE_SIZE; 6072 for(i = 0; i < numpages; i++) { 6073 page = pages[i]; 6074 6075 from = start & (PAGE_CACHE_SIZE - 1); 6076 if ((end >> PAGE_CACHE_SHIFT) == page->index) 6077 to = end & (PAGE_CACHE_SIZE - 1); 6078 6079 BUG_ON(from > PAGE_CACHE_SIZE); 6080 BUG_ON(to > PAGE_CACHE_SIZE); 6081 6082 ocfs2_map_and_dirty_page(inode, handle, from, to, page, 1, 6083 &phys); 6084 6085 start = (page->index + 1) << PAGE_CACHE_SHIFT; 6086 } 6087 out: 6088 if (pages) 6089 ocfs2_unlock_and_free_pages(pages, numpages); 6090 } 6091 6092 static int ocfs2_grab_eof_pages(struct inode *inode, loff_t start, loff_t end, 6093 struct page **pages, int *num) 6094 { 6095 int numpages, ret = 0; 6096 struct super_block *sb = inode->i_sb; 6097 struct address_space *mapping = inode->i_mapping; 6098 unsigned long index; 6099 loff_t last_page_bytes; 6100 6101 BUG_ON(start > end); 6102 6103 BUG_ON(start >> OCFS2_SB(sb)->s_clustersize_bits != 6104 (end - 1) >> OCFS2_SB(sb)->s_clustersize_bits); 6105 6106 numpages = 0; 6107 last_page_bytes = PAGE_ALIGN(end); 6108 index = start >> PAGE_CACHE_SHIFT; 6109 do { 6110 pages[numpages] = grab_cache_page(mapping, index); 6111 if (!pages[numpages]) { 6112 ret = -ENOMEM; 6113 mlog_errno(ret); 6114 goto out; 6115 } 6116 6117 numpages++; 6118 index++; 6119 } while (index < (last_page_bytes >> PAGE_CACHE_SHIFT)); 6120 6121 out: 6122 if (ret != 0) { 6123 if (pages) 6124 ocfs2_unlock_and_free_pages(pages, numpages); 6125 numpages = 0; 6126 } 6127 6128 *num = numpages; 6129 6130 return ret; 6131 } 6132 6133 /* 6134 * Zero the area past i_size but still within an allocated 6135 * cluster. This avoids exposing nonzero data on subsequent file 6136 * extends. 6137 * 6138 * We need to call this before i_size is updated on the inode because 6139 * otherwise block_write_full_page() will skip writeout of pages past 6140 * i_size. The new_i_size parameter is passed for this reason. 6141 */ 6142 int ocfs2_zero_range_for_truncate(struct inode *inode, handle_t *handle, 6143 u64 range_start, u64 range_end) 6144 { 6145 int ret = 0, numpages; 6146 struct page **pages = NULL; 6147 u64 phys; 6148 unsigned int ext_flags; 6149 struct super_block *sb = inode->i_sb; 6150 6151 /* 6152 * File systems which don't support sparse files zero on every 6153 * extend. 6154 */ 6155 if (!ocfs2_sparse_alloc(OCFS2_SB(sb))) 6156 return 0; 6157 6158 pages = kcalloc(ocfs2_pages_per_cluster(sb), 6159 sizeof(struct page *), GFP_NOFS); 6160 if (pages == NULL) { 6161 ret = -ENOMEM; 6162 mlog_errno(ret); 6163 goto out; 6164 } 6165 6166 if (range_start == range_end) 6167 goto out; 6168 6169 ret = ocfs2_extent_map_get_blocks(inode, 6170 range_start >> sb->s_blocksize_bits, 6171 &phys, NULL, &ext_flags); 6172 if (ret) { 6173 mlog_errno(ret); 6174 goto out; 6175 } 6176 6177 /* 6178 * Tail is a hole, or is marked unwritten. In either case, we 6179 * can count on read and write to return/push zero's. 6180 */ 6181 if (phys == 0 || ext_flags & OCFS2_EXT_UNWRITTEN) 6182 goto out; 6183 6184 ret = ocfs2_grab_eof_pages(inode, range_start, range_end, pages, 6185 &numpages); 6186 if (ret) { 6187 mlog_errno(ret); 6188 goto out; 6189 } 6190 6191 ocfs2_zero_cluster_pages(inode, range_start, range_end, pages, 6192 numpages, phys, handle); 6193 6194 /* 6195 * Initiate writeout of the pages we zero'd here. We don't 6196 * wait on them - the truncate_inode_pages() call later will 6197 * do that for us. 6198 */ 6199 ret = do_sync_mapping_range(inode->i_mapping, range_start, 6200 range_end - 1, SYNC_FILE_RANGE_WRITE); 6201 if (ret) 6202 mlog_errno(ret); 6203 6204 out: 6205 if (pages) 6206 kfree(pages); 6207 6208 return ret; 6209 } 6210 6211 static void ocfs2_zero_dinode_id2(struct inode *inode, struct ocfs2_dinode *di) 6212 { 6213 unsigned int blocksize = 1 << inode->i_sb->s_blocksize_bits; 6214 6215 memset(&di->id2, 0, blocksize - offsetof(struct ocfs2_dinode, id2)); 6216 } 6217 6218 void ocfs2_dinode_new_extent_list(struct inode *inode, 6219 struct ocfs2_dinode *di) 6220 { 6221 ocfs2_zero_dinode_id2(inode, di); 6222 di->id2.i_list.l_tree_depth = 0; 6223 di->id2.i_list.l_next_free_rec = 0; 6224 di->id2.i_list.l_count = cpu_to_le16(ocfs2_extent_recs_per_inode(inode->i_sb)); 6225 } 6226 6227 void ocfs2_set_inode_data_inline(struct inode *inode, struct ocfs2_dinode *di) 6228 { 6229 struct ocfs2_inode_info *oi = OCFS2_I(inode); 6230 struct ocfs2_inline_data *idata = &di->id2.i_data; 6231 6232 spin_lock(&oi->ip_lock); 6233 oi->ip_dyn_features |= OCFS2_INLINE_DATA_FL; 6234 di->i_dyn_features = cpu_to_le16(oi->ip_dyn_features); 6235 spin_unlock(&oi->ip_lock); 6236 6237 /* 6238 * We clear the entire i_data structure here so that all 6239 * fields can be properly initialized. 6240 */ 6241 ocfs2_zero_dinode_id2(inode, di); 6242 6243 idata->id_count = cpu_to_le16(ocfs2_max_inline_data(inode->i_sb)); 6244 } 6245 6246 int ocfs2_convert_inline_data_to_extents(struct inode *inode, 6247 struct buffer_head *di_bh) 6248 { 6249 int ret, i, has_data, num_pages = 0; 6250 handle_t *handle; 6251 u64 uninitialized_var(block); 6252 struct ocfs2_inode_info *oi = OCFS2_I(inode); 6253 struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); 6254 struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data; 6255 struct ocfs2_alloc_context *data_ac = NULL; 6256 struct page **pages = NULL; 6257 loff_t end = osb->s_clustersize; 6258 6259 has_data = i_size_read(inode) ? 1 : 0; 6260 6261 if (has_data) { 6262 pages = kcalloc(ocfs2_pages_per_cluster(osb->sb), 6263 sizeof(struct page *), GFP_NOFS); 6264 if (pages == NULL) { 6265 ret = -ENOMEM; 6266 mlog_errno(ret); 6267 goto out; 6268 } 6269 6270 ret = ocfs2_reserve_clusters(osb, 1, &data_ac); 6271 if (ret) { 6272 mlog_errno(ret); 6273 goto out; 6274 } 6275 } 6276 6277 handle = ocfs2_start_trans(osb, OCFS2_INLINE_TO_EXTENTS_CREDITS); 6278 if (IS_ERR(handle)) { 6279 ret = PTR_ERR(handle); 6280 mlog_errno(ret); 6281 goto out_unlock; 6282 } 6283 6284 ret = ocfs2_journal_access(handle, inode, di_bh, 6285 OCFS2_JOURNAL_ACCESS_WRITE); 6286 if (ret) { 6287 mlog_errno(ret); 6288 goto out_commit; 6289 } 6290 6291 if (has_data) { 6292 u32 bit_off, num; 6293 unsigned int page_end; 6294 u64 phys; 6295 6296 ret = ocfs2_claim_clusters(osb, handle, data_ac, 1, &bit_off, 6297 &num); 6298 if (ret) { 6299 mlog_errno(ret); 6300 goto out_commit; 6301 } 6302 6303 /* 6304 * Save two copies, one for insert, and one that can 6305 * be changed by ocfs2_map_and_dirty_page() below. 6306 */ 6307 block = phys = ocfs2_clusters_to_blocks(inode->i_sb, bit_off); 6308 6309 /* 6310 * Non sparse file systems zero on extend, so no need 6311 * to do that now. 6312 */ 6313 if (!ocfs2_sparse_alloc(osb) && 6314 PAGE_CACHE_SIZE < osb->s_clustersize) 6315 end = PAGE_CACHE_SIZE; 6316 6317 ret = ocfs2_grab_eof_pages(inode, 0, end, pages, &num_pages); 6318 if (ret) { 6319 mlog_errno(ret); 6320 goto out_commit; 6321 } 6322 6323 /* 6324 * This should populate the 1st page for us and mark 6325 * it up to date. 6326 */ 6327 ret = ocfs2_read_inline_data(inode, pages[0], di_bh); 6328 if (ret) { 6329 mlog_errno(ret); 6330 goto out_commit; 6331 } 6332 6333 page_end = PAGE_CACHE_SIZE; 6334 if (PAGE_CACHE_SIZE > osb->s_clustersize) 6335 page_end = osb->s_clustersize; 6336 6337 for (i = 0; i < num_pages; i++) 6338 ocfs2_map_and_dirty_page(inode, handle, 0, page_end, 6339 pages[i], i > 0, &phys); 6340 } 6341 6342 spin_lock(&oi->ip_lock); 6343 oi->ip_dyn_features &= ~OCFS2_INLINE_DATA_FL; 6344 di->i_dyn_features = cpu_to_le16(oi->ip_dyn_features); 6345 spin_unlock(&oi->ip_lock); 6346 6347 ocfs2_dinode_new_extent_list(inode, di); 6348 6349 ocfs2_journal_dirty(handle, di_bh); 6350 6351 if (has_data) { 6352 /* 6353 * An error at this point should be extremely rare. If 6354 * this proves to be false, we could always re-build 6355 * the in-inode data from our pages. 6356 */ 6357 ret = ocfs2_insert_extent(osb, handle, inode, di_bh, 6358 0, block, 1, 0, NULL); 6359 if (ret) { 6360 mlog_errno(ret); 6361 goto out_commit; 6362 } 6363 6364 inode->i_blocks = ocfs2_inode_sector_count(inode); 6365 } 6366 6367 out_commit: 6368 ocfs2_commit_trans(osb, handle); 6369 6370 out_unlock: 6371 if (data_ac) 6372 ocfs2_free_alloc_context(data_ac); 6373 6374 out: 6375 if (pages) { 6376 ocfs2_unlock_and_free_pages(pages, num_pages); 6377 kfree(pages); 6378 } 6379 6380 return ret; 6381 } 6382 6383 /* 6384 * It is expected, that by the time you call this function, 6385 * inode->i_size and fe->i_size have been adjusted. 6386 * 6387 * WARNING: This will kfree the truncate context 6388 */ 6389 int ocfs2_commit_truncate(struct ocfs2_super *osb, 6390 struct inode *inode, 6391 struct buffer_head *fe_bh, 6392 struct ocfs2_truncate_context *tc) 6393 { 6394 int status, i, credits, tl_sem = 0; 6395 u32 clusters_to_del, new_highest_cpos, range; 6396 struct ocfs2_extent_list *el; 6397 handle_t *handle = NULL; 6398 struct inode *tl_inode = osb->osb_tl_inode; 6399 struct ocfs2_path *path = NULL; 6400 6401 mlog_entry_void(); 6402 6403 new_highest_cpos = ocfs2_clusters_for_bytes(osb->sb, 6404 i_size_read(inode)); 6405 6406 path = ocfs2_new_inode_path(fe_bh); 6407 if (!path) { 6408 status = -ENOMEM; 6409 mlog_errno(status); 6410 goto bail; 6411 } 6412 6413 ocfs2_extent_map_trunc(inode, new_highest_cpos); 6414 6415 start: 6416 /* 6417 * Check that we still have allocation to delete. 6418 */ 6419 if (OCFS2_I(inode)->ip_clusters == 0) { 6420 status = 0; 6421 goto bail; 6422 } 6423 6424 /* 6425 * Truncate always works against the rightmost tree branch. 6426 */ 6427 status = ocfs2_find_path(inode, path, UINT_MAX); 6428 if (status) { 6429 mlog_errno(status); 6430 goto bail; 6431 } 6432 6433 mlog(0, "inode->ip_clusters = %u, tree_depth = %u\n", 6434 OCFS2_I(inode)->ip_clusters, path->p_tree_depth); 6435 6436 /* 6437 * By now, el will point to the extent list on the bottom most 6438 * portion of this tree. Only the tail record is considered in 6439 * each pass. 6440 * 6441 * We handle the following cases, in order: 6442 * - empty extent: delete the remaining branch 6443 * - remove the entire record 6444 * - remove a partial record 6445 * - no record needs to be removed (truncate has completed) 6446 */ 6447 el = path_leaf_el(path); 6448 if (le16_to_cpu(el->l_next_free_rec) == 0) { 6449 ocfs2_error(inode->i_sb, 6450 "Inode %llu has empty extent block at %llu\n", 6451 (unsigned long long)OCFS2_I(inode)->ip_blkno, 6452 (unsigned long long)path_leaf_bh(path)->b_blocknr); 6453 status = -EROFS; 6454 goto bail; 6455 } 6456 6457 i = le16_to_cpu(el->l_next_free_rec) - 1; 6458 range = le32_to_cpu(el->l_recs[i].e_cpos) + 6459 ocfs2_rec_clusters(el, &el->l_recs[i]); 6460 if (i == 0 && ocfs2_is_empty_extent(&el->l_recs[i])) { 6461 clusters_to_del = 0; 6462 } else if (le32_to_cpu(el->l_recs[i].e_cpos) >= new_highest_cpos) { 6463 clusters_to_del = ocfs2_rec_clusters(el, &el->l_recs[i]); 6464 } else if (range > new_highest_cpos) { 6465 clusters_to_del = (ocfs2_rec_clusters(el, &el->l_recs[i]) + 6466 le32_to_cpu(el->l_recs[i].e_cpos)) - 6467 new_highest_cpos; 6468 } else { 6469 status = 0; 6470 goto bail; 6471 } 6472 6473 mlog(0, "clusters_to_del = %u in this pass, tail blk=%llu\n", 6474 clusters_to_del, (unsigned long long)path_leaf_bh(path)->b_blocknr); 6475 6476 mutex_lock(&tl_inode->i_mutex); 6477 tl_sem = 1; 6478 /* ocfs2_truncate_log_needs_flush guarantees us at least one 6479 * record is free for use. If there isn't any, we flush to get 6480 * an empty truncate log. */ 6481 if (ocfs2_truncate_log_needs_flush(osb)) { 6482 status = __ocfs2_flush_truncate_log(osb); 6483 if (status < 0) { 6484 mlog_errno(status); 6485 goto bail; 6486 } 6487 } 6488 6489 credits = ocfs2_calc_tree_trunc_credits(osb->sb, clusters_to_del, 6490 (struct ocfs2_dinode *)fe_bh->b_data, 6491 el); 6492 handle = ocfs2_start_trans(osb, credits); 6493 if (IS_ERR(handle)) { 6494 status = PTR_ERR(handle); 6495 handle = NULL; 6496 mlog_errno(status); 6497 goto bail; 6498 } 6499 6500 status = ocfs2_do_truncate(osb, clusters_to_del, inode, fe_bh, handle, 6501 tc, path); 6502 if (status < 0) { 6503 mlog_errno(status); 6504 goto bail; 6505 } 6506 6507 mutex_unlock(&tl_inode->i_mutex); 6508 tl_sem = 0; 6509 6510 ocfs2_commit_trans(osb, handle); 6511 handle = NULL; 6512 6513 ocfs2_reinit_path(path, 1); 6514 6515 /* 6516 * The check above will catch the case where we've truncated 6517 * away all allocation. 6518 */ 6519 goto start; 6520 6521 bail: 6522 6523 ocfs2_schedule_truncate_log_flush(osb, 1); 6524 6525 if (tl_sem) 6526 mutex_unlock(&tl_inode->i_mutex); 6527 6528 if (handle) 6529 ocfs2_commit_trans(osb, handle); 6530 6531 ocfs2_run_deallocs(osb, &tc->tc_dealloc); 6532 6533 ocfs2_free_path(path); 6534 6535 /* This will drop the ext_alloc cluster lock for us */ 6536 ocfs2_free_truncate_context(tc); 6537 6538 mlog_exit(status); 6539 return status; 6540 } 6541 6542 /* 6543 * Expects the inode to already be locked. 6544 */ 6545 int ocfs2_prepare_truncate(struct ocfs2_super *osb, 6546 struct inode *inode, 6547 struct buffer_head *fe_bh, 6548 struct ocfs2_truncate_context **tc) 6549 { 6550 int status; 6551 unsigned int new_i_clusters; 6552 struct ocfs2_dinode *fe; 6553 struct ocfs2_extent_block *eb; 6554 struct buffer_head *last_eb_bh = NULL; 6555 6556 mlog_entry_void(); 6557 6558 *tc = NULL; 6559 6560 new_i_clusters = ocfs2_clusters_for_bytes(osb->sb, 6561 i_size_read(inode)); 6562 fe = (struct ocfs2_dinode *) fe_bh->b_data; 6563 6564 mlog(0, "fe->i_clusters = %u, new_i_clusters = %u, fe->i_size =" 6565 "%llu\n", le32_to_cpu(fe->i_clusters), new_i_clusters, 6566 (unsigned long long)le64_to_cpu(fe->i_size)); 6567 6568 *tc = kzalloc(sizeof(struct ocfs2_truncate_context), GFP_KERNEL); 6569 if (!(*tc)) { 6570 status = -ENOMEM; 6571 mlog_errno(status); 6572 goto bail; 6573 } 6574 ocfs2_init_dealloc_ctxt(&(*tc)->tc_dealloc); 6575 6576 if (fe->id2.i_list.l_tree_depth) { 6577 status = ocfs2_read_block(osb, le64_to_cpu(fe->i_last_eb_blk), 6578 &last_eb_bh, OCFS2_BH_CACHED, inode); 6579 if (status < 0) { 6580 mlog_errno(status); 6581 goto bail; 6582 } 6583 eb = (struct ocfs2_extent_block *) last_eb_bh->b_data; 6584 if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) { 6585 OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb); 6586 6587 brelse(last_eb_bh); 6588 status = -EIO; 6589 goto bail; 6590 } 6591 } 6592 6593 (*tc)->tc_last_eb_bh = last_eb_bh; 6594 6595 status = 0; 6596 bail: 6597 if (status < 0) { 6598 if (*tc) 6599 ocfs2_free_truncate_context(*tc); 6600 *tc = NULL; 6601 } 6602 mlog_exit_void(); 6603 return status; 6604 } 6605 6606 /* 6607 * 'start' is inclusive, 'end' is not. 6608 */ 6609 int ocfs2_truncate_inline(struct inode *inode, struct buffer_head *di_bh, 6610 unsigned int start, unsigned int end, int trunc) 6611 { 6612 int ret; 6613 unsigned int numbytes; 6614 handle_t *handle; 6615 struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); 6616 struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data; 6617 struct ocfs2_inline_data *idata = &di->id2.i_data; 6618 6619 if (end > i_size_read(inode)) 6620 end = i_size_read(inode); 6621 6622 BUG_ON(start >= end); 6623 6624 if (!(OCFS2_I(inode)->ip_dyn_features & OCFS2_INLINE_DATA_FL) || 6625 !(le16_to_cpu(di->i_dyn_features) & OCFS2_INLINE_DATA_FL) || 6626 !ocfs2_supports_inline_data(osb)) { 6627 ocfs2_error(inode->i_sb, 6628 "Inline data flags for inode %llu don't agree! " 6629 "Disk: 0x%x, Memory: 0x%x, Superblock: 0x%x\n", 6630 (unsigned long long)OCFS2_I(inode)->ip_blkno, 6631 le16_to_cpu(di->i_dyn_features), 6632 OCFS2_I(inode)->ip_dyn_features, 6633 osb->s_feature_incompat); 6634 ret = -EROFS; 6635 goto out; 6636 } 6637 6638 handle = ocfs2_start_trans(osb, OCFS2_INODE_UPDATE_CREDITS); 6639 if (IS_ERR(handle)) { 6640 ret = PTR_ERR(handle); 6641 mlog_errno(ret); 6642 goto out; 6643 } 6644 6645 ret = ocfs2_journal_access(handle, inode, di_bh, 6646 OCFS2_JOURNAL_ACCESS_WRITE); 6647 if (ret) { 6648 mlog_errno(ret); 6649 goto out_commit; 6650 } 6651 6652 numbytes = end - start; 6653 memset(idata->id_data + start, 0, numbytes); 6654 6655 /* 6656 * No need to worry about the data page here - it's been 6657 * truncated already and inline data doesn't need it for 6658 * pushing zero's to disk, so we'll let readpage pick it up 6659 * later. 6660 */ 6661 if (trunc) { 6662 i_size_write(inode, start); 6663 di->i_size = cpu_to_le64(start); 6664 } 6665 6666 inode->i_blocks = ocfs2_inode_sector_count(inode); 6667 inode->i_ctime = inode->i_mtime = CURRENT_TIME; 6668 6669 di->i_ctime = di->i_mtime = cpu_to_le64(inode->i_ctime.tv_sec); 6670 di->i_ctime_nsec = di->i_mtime_nsec = cpu_to_le32(inode->i_ctime.tv_nsec); 6671 6672 ocfs2_journal_dirty(handle, di_bh); 6673 6674 out_commit: 6675 ocfs2_commit_trans(osb, handle); 6676 6677 out: 6678 return ret; 6679 } 6680 6681 static void ocfs2_free_truncate_context(struct ocfs2_truncate_context *tc) 6682 { 6683 /* 6684 * The caller is responsible for completing deallocation 6685 * before freeing the context. 6686 */ 6687 if (tc->tc_dealloc.c_first_suballocator != NULL) 6688 mlog(ML_NOTICE, 6689 "Truncate completion has non-empty dealloc context\n"); 6690 6691 if (tc->tc_last_eb_bh) 6692 brelse(tc->tc_last_eb_bh); 6693 6694 kfree(tc); 6695 } 6696